diff --git "a/qwen2/trainer_state.json" "b/qwen2/trainer_state.json" new file mode 100644--- /dev/null +++ "b/qwen2/trainer_state.json" @@ -0,0 +1,174641 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.9997795193521877, + "eval_steps": 500, + "global_step": 24944, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 8.017478102262934e-05, + "grad_norm": 27.020566940307617, + "learning_rate": 2.67022696929239e-08, + "loss": 6.76, + "step": 1 + }, + { + "epoch": 0.00016034956204525867, + "grad_norm": 25.159984588623047, + "learning_rate": 5.34045393858478e-08, + "loss": 6.6992, + "step": 2 + }, + { + "epoch": 0.000240524343067888, + "grad_norm": 25.81383514404297, + "learning_rate": 8.010680907877171e-08, + "loss": 6.6775, + "step": 3 + }, + { + "epoch": 0.00032069912409051735, + "grad_norm": 26.60355567932129, + "learning_rate": 1.068090787716956e-07, + "loss": 6.6881, + "step": 4 + }, + { + "epoch": 0.00040087390511314665, + "grad_norm": 23.614158630371094, + "learning_rate": 1.3351134846461952e-07, + "loss": 6.6681, + "step": 5 + }, + { + "epoch": 0.000481048686135776, + "grad_norm": 25.451841354370117, + "learning_rate": 1.6021361815754342e-07, + "loss": 6.6313, + "step": 6 + }, + { + "epoch": 0.0005612234671584053, + "grad_norm": 23.309423446655273, + "learning_rate": 1.8691588785046729e-07, + "loss": 6.6921, + "step": 7 + }, + { + "epoch": 0.0006413982481810347, + "grad_norm": 24.529417037963867, + "learning_rate": 2.136181575433912e-07, + "loss": 6.7642, + "step": 8 + }, + { + "epoch": 0.000721573029203664, + "grad_norm": 25.31582260131836, + "learning_rate": 2.403204272363151e-07, + "loss": 6.5289, + "step": 9 + }, + { + "epoch": 0.0008017478102262933, + "grad_norm": 25.33662223815918, + "learning_rate": 2.6702269692923903e-07, + "loss": 6.6685, + "step": 10 + }, + { + "epoch": 0.0008819225912489227, + "grad_norm": 25.33871078491211, + "learning_rate": 2.937249666221629e-07, + "loss": 6.6631, + "step": 11 + }, + { + "epoch": 0.000962097372271552, + "grad_norm": 26.346580505371094, + "learning_rate": 3.2042723631508683e-07, + "loss": 6.5928, + "step": 12 + }, + { + "epoch": 0.0010422721532941813, + "grad_norm": 24.419307708740234, + "learning_rate": 3.4712950600801073e-07, + "loss": 6.5892, + "step": 13 + }, + { + "epoch": 0.0011224469343168106, + "grad_norm": 23.368968963623047, + "learning_rate": 3.7383177570093457e-07, + "loss": 6.6144, + "step": 14 + }, + { + "epoch": 0.0012026217153394399, + "grad_norm": 23.5020694732666, + "learning_rate": 4.005340453938585e-07, + "loss": 6.5679, + "step": 15 + }, + { + "epoch": 0.0012827964963620694, + "grad_norm": 24.370548248291016, + "learning_rate": 4.272363150867824e-07, + "loss": 6.6155, + "step": 16 + }, + { + "epoch": 0.0013629712773846987, + "grad_norm": 25.244121551513672, + "learning_rate": 4.539385847797063e-07, + "loss": 6.5599, + "step": 17 + }, + { + "epoch": 0.001443146058407328, + "grad_norm": 23.976192474365234, + "learning_rate": 4.806408544726302e-07, + "loss": 6.6165, + "step": 18 + }, + { + "epoch": 0.0015233208394299573, + "grad_norm": 26.96253776550293, + "learning_rate": 5.073431241655541e-07, + "loss": 6.6401, + "step": 19 + }, + { + "epoch": 0.0016034956204525866, + "grad_norm": 23.295860290527344, + "learning_rate": 5.340453938584781e-07, + "loss": 6.5972, + "step": 20 + }, + { + "epoch": 0.0016836704014752159, + "grad_norm": 22.348068237304688, + "learning_rate": 5.607476635514019e-07, + "loss": 6.5652, + "step": 21 + }, + { + "epoch": 0.0017638451824978454, + "grad_norm": 21.81495475769043, + "learning_rate": 5.874499332443258e-07, + "loss": 6.5393, + "step": 22 + }, + { + "epoch": 0.0018440199635204747, + "grad_norm": 21.15462303161621, + "learning_rate": 6.141522029372497e-07, + "loss": 6.5342, + "step": 23 + }, + { + "epoch": 0.001924194744543104, + "grad_norm": 33.77826690673828, + "learning_rate": 6.408544726301737e-07, + "loss": 6.4646, + "step": 24 + }, + { + "epoch": 0.0020043695255657335, + "grad_norm": 22.331083297729492, + "learning_rate": 6.675567423230975e-07, + "loss": 6.6572, + "step": 25 + }, + { + "epoch": 0.0020845443065883626, + "grad_norm": 19.642393112182617, + "learning_rate": 6.942590120160215e-07, + "loss": 6.5459, + "step": 26 + }, + { + "epoch": 0.002164719087610992, + "grad_norm": 17.878122329711914, + "learning_rate": 7.209612817089454e-07, + "loss": 6.5112, + "step": 27 + }, + { + "epoch": 0.002244893868633621, + "grad_norm": 16.757993698120117, + "learning_rate": 7.476635514018691e-07, + "loss": 6.4963, + "step": 28 + }, + { + "epoch": 0.0023250686496562507, + "grad_norm": 16.240556716918945, + "learning_rate": 7.743658210947931e-07, + "loss": 6.4954, + "step": 29 + }, + { + "epoch": 0.0024052434306788798, + "grad_norm": 16.22499656677246, + "learning_rate": 8.01068090787717e-07, + "loss": 6.5059, + "step": 30 + }, + { + "epoch": 0.0024854182117015093, + "grad_norm": 26.01296615600586, + "learning_rate": 8.277703604806409e-07, + "loss": 6.3762, + "step": 31 + }, + { + "epoch": 0.002565592992724139, + "grad_norm": 15.304647445678711, + "learning_rate": 8.544726301735648e-07, + "loss": 6.4163, + "step": 32 + }, + { + "epoch": 0.002645767773746768, + "grad_norm": 15.413506507873535, + "learning_rate": 8.811748998664887e-07, + "loss": 6.4504, + "step": 33 + }, + { + "epoch": 0.0027259425547693974, + "grad_norm": 15.258155822753906, + "learning_rate": 9.078771695594126e-07, + "loss": 6.4875, + "step": 34 + }, + { + "epoch": 0.0028061173357920265, + "grad_norm": 14.354201316833496, + "learning_rate": 9.345794392523365e-07, + "loss": 6.4524, + "step": 35 + }, + { + "epoch": 0.002886292116814656, + "grad_norm": 13.485629081726074, + "learning_rate": 9.612817089452603e-07, + "loss": 6.4729, + "step": 36 + }, + { + "epoch": 0.0029664668978372855, + "grad_norm": 18.795286178588867, + "learning_rate": 9.879839786381843e-07, + "loss": 6.2607, + "step": 37 + }, + { + "epoch": 0.0030466416788599146, + "grad_norm": 9.043806076049805, + "learning_rate": 1.0146862483311082e-06, + "loss": 6.3213, + "step": 38 + }, + { + "epoch": 0.003126816459882544, + "grad_norm": 8.909296989440918, + "learning_rate": 1.0413885180240322e-06, + "loss": 6.3542, + "step": 39 + }, + { + "epoch": 0.003206991240905173, + "grad_norm": 7.784614086151123, + "learning_rate": 1.0680907877169561e-06, + "loss": 6.29, + "step": 40 + }, + { + "epoch": 0.0032871660219278027, + "grad_norm": 7.8301520347595215, + "learning_rate": 1.0947930574098799e-06, + "loss": 6.3165, + "step": 41 + }, + { + "epoch": 0.0033673408029504318, + "grad_norm": 7.5845465660095215, + "learning_rate": 1.1214953271028038e-06, + "loss": 6.2848, + "step": 42 + }, + { + "epoch": 0.0034475155839730613, + "grad_norm": 8.239869117736816, + "learning_rate": 1.1481975967957278e-06, + "loss": 6.244, + "step": 43 + }, + { + "epoch": 0.0035276903649956908, + "grad_norm": 7.545753002166748, + "learning_rate": 1.1748998664886515e-06, + "loss": 6.2267, + "step": 44 + }, + { + "epoch": 0.00360786514601832, + "grad_norm": 7.4281158447265625, + "learning_rate": 1.2016021361815755e-06, + "loss": 6.3545, + "step": 45 + }, + { + "epoch": 0.0036880399270409494, + "grad_norm": 7.301125526428223, + "learning_rate": 1.2283044058744994e-06, + "loss": 6.3071, + "step": 46 + }, + { + "epoch": 0.0037682147080635785, + "grad_norm": 7.347575664520264, + "learning_rate": 1.2550066755674234e-06, + "loss": 6.2682, + "step": 47 + }, + { + "epoch": 0.003848389489086208, + "grad_norm": 7.792304515838623, + "learning_rate": 1.2817089452603473e-06, + "loss": 6.2292, + "step": 48 + }, + { + "epoch": 0.003928564270108837, + "grad_norm": 6.83495569229126, + "learning_rate": 1.308411214953271e-06, + "loss": 6.3245, + "step": 49 + }, + { + "epoch": 0.004008739051131467, + "grad_norm": 6.5145955085754395, + "learning_rate": 1.335113484646195e-06, + "loss": 6.3488, + "step": 50 + }, + { + "epoch": 0.004088913832154096, + "grad_norm": 6.870636463165283, + "learning_rate": 1.361815754339119e-06, + "loss": 6.283, + "step": 51 + }, + { + "epoch": 0.004169088613176725, + "grad_norm": 6.131483554840088, + "learning_rate": 1.388518024032043e-06, + "loss": 6.2449, + "step": 52 + }, + { + "epoch": 0.004249263394199354, + "grad_norm": 5.957351207733154, + "learning_rate": 1.4152202937249669e-06, + "loss": 6.197, + "step": 53 + }, + { + "epoch": 0.004329438175221984, + "grad_norm": 5.79727029800415, + "learning_rate": 1.4419225634178908e-06, + "loss": 6.2479, + "step": 54 + }, + { + "epoch": 0.004409612956244613, + "grad_norm": 6.7776031494140625, + "learning_rate": 1.4686248331108148e-06, + "loss": 6.2363, + "step": 55 + }, + { + "epoch": 0.004489787737267242, + "grad_norm": 5.542695045471191, + "learning_rate": 1.4953271028037383e-06, + "loss": 6.1953, + "step": 56 + }, + { + "epoch": 0.004569962518289872, + "grad_norm": 5.969710350036621, + "learning_rate": 1.5220293724966622e-06, + "loss": 6.1204, + "step": 57 + }, + { + "epoch": 0.004650137299312501, + "grad_norm": 5.7074875831604, + "learning_rate": 1.5487316421895862e-06, + "loss": 6.229, + "step": 58 + }, + { + "epoch": 0.0047303120803351304, + "grad_norm": 5.470493793487549, + "learning_rate": 1.5754339118825101e-06, + "loss": 6.2158, + "step": 59 + }, + { + "epoch": 0.0048104868613577595, + "grad_norm": 5.586243152618408, + "learning_rate": 1.602136181575434e-06, + "loss": 6.2046, + "step": 60 + }, + { + "epoch": 0.0048906616423803895, + "grad_norm": 4.666398048400879, + "learning_rate": 1.628838451268358e-06, + "loss": 6.2188, + "step": 61 + }, + { + "epoch": 0.0049708364234030186, + "grad_norm": 4.898543834686279, + "learning_rate": 1.6555407209612818e-06, + "loss": 6.1272, + "step": 62 + }, + { + "epoch": 0.005051011204425648, + "grad_norm": 4.737247943878174, + "learning_rate": 1.6822429906542057e-06, + "loss": 6.2393, + "step": 63 + }, + { + "epoch": 0.005131185985448278, + "grad_norm": 6.187193393707275, + "learning_rate": 1.7089452603471297e-06, + "loss": 6.1753, + "step": 64 + }, + { + "epoch": 0.005211360766470907, + "grad_norm": 4.390422344207764, + "learning_rate": 1.7356475300400536e-06, + "loss": 6.0559, + "step": 65 + }, + { + "epoch": 0.005291535547493536, + "grad_norm": 4.731480121612549, + "learning_rate": 1.7623497997329774e-06, + "loss": 6.1438, + "step": 66 + }, + { + "epoch": 0.005371710328516165, + "grad_norm": 4.61002779006958, + "learning_rate": 1.7890520694259013e-06, + "loss": 6.156, + "step": 67 + }, + { + "epoch": 0.005451885109538795, + "grad_norm": 4.50532865524292, + "learning_rate": 1.8157543391188253e-06, + "loss": 6.1792, + "step": 68 + }, + { + "epoch": 0.005532059890561424, + "grad_norm": 4.2495574951171875, + "learning_rate": 1.842456608811749e-06, + "loss": 6.1805, + "step": 69 + }, + { + "epoch": 0.005612234671584053, + "grad_norm": 4.196827411651611, + "learning_rate": 1.869158878504673e-06, + "loss": 6.2344, + "step": 70 + }, + { + "epoch": 0.005692409452606683, + "grad_norm": 5.999969959259033, + "learning_rate": 1.895861148197597e-06, + "loss": 6.2014, + "step": 71 + }, + { + "epoch": 0.005772584233629312, + "grad_norm": 3.969874620437622, + "learning_rate": 1.9225634178905207e-06, + "loss": 6.1057, + "step": 72 + }, + { + "epoch": 0.005852759014651941, + "grad_norm": 3.919011354446411, + "learning_rate": 1.9492656875834446e-06, + "loss": 6.1555, + "step": 73 + }, + { + "epoch": 0.005932933795674571, + "grad_norm": 4.136414527893066, + "learning_rate": 1.9759679572763686e-06, + "loss": 6.0845, + "step": 74 + }, + { + "epoch": 0.0060131085766972, + "grad_norm": 4.306534290313721, + "learning_rate": 2.0026702269692925e-06, + "loss": 6.2013, + "step": 75 + }, + { + "epoch": 0.006093283357719829, + "grad_norm": 3.7880969047546387, + "learning_rate": 2.0293724966622165e-06, + "loss": 6.1704, + "step": 76 + }, + { + "epoch": 0.006173458138742458, + "grad_norm": 6.807768821716309, + "learning_rate": 2.0560747663551404e-06, + "loss": 6.0063, + "step": 77 + }, + { + "epoch": 0.006253632919765088, + "grad_norm": 4.161741733551025, + "learning_rate": 2.0827770360480644e-06, + "loss": 6.0569, + "step": 78 + }, + { + "epoch": 0.006333807700787717, + "grad_norm": 3.921083450317383, + "learning_rate": 2.1094793057409883e-06, + "loss": 6.2026, + "step": 79 + }, + { + "epoch": 0.006413982481810346, + "grad_norm": 3.725938320159912, + "learning_rate": 2.1361815754339123e-06, + "loss": 6.1499, + "step": 80 + }, + { + "epoch": 0.006494157262832976, + "grad_norm": 3.85404634475708, + "learning_rate": 2.1628838451268362e-06, + "loss": 6.1265, + "step": 81 + }, + { + "epoch": 0.006574332043855605, + "grad_norm": 3.668159008026123, + "learning_rate": 2.1895861148197598e-06, + "loss": 6.1714, + "step": 82 + }, + { + "epoch": 0.006654506824878234, + "grad_norm": 5.945050239562988, + "learning_rate": 2.2162883845126837e-06, + "loss": 5.9871, + "step": 83 + }, + { + "epoch": 0.0067346816059008635, + "grad_norm": 3.724618434906006, + "learning_rate": 2.2429906542056077e-06, + "loss": 6.1617, + "step": 84 + }, + { + "epoch": 0.0068148563869234935, + "grad_norm": 3.676133155822754, + "learning_rate": 2.2696929238985316e-06, + "loss": 6.1614, + "step": 85 + }, + { + "epoch": 0.0068950311679461225, + "grad_norm": 3.6179938316345215, + "learning_rate": 2.2963951935914556e-06, + "loss": 6.1606, + "step": 86 + }, + { + "epoch": 0.006975205948968752, + "grad_norm": 3.6683545112609863, + "learning_rate": 2.3230974632843795e-06, + "loss": 6.1414, + "step": 87 + }, + { + "epoch": 0.0070553807299913816, + "grad_norm": 3.7329607009887695, + "learning_rate": 2.349799732977303e-06, + "loss": 6.1445, + "step": 88 + }, + { + "epoch": 0.007135555511014011, + "grad_norm": 5.664222717285156, + "learning_rate": 2.376502002670227e-06, + "loss": 5.9995, + "step": 89 + }, + { + "epoch": 0.00721573029203664, + "grad_norm": 3.5671427249908447, + "learning_rate": 2.403204272363151e-06, + "loss": 6.0763, + "step": 90 + }, + { + "epoch": 0.007295905073059269, + "grad_norm": 3.6859803199768066, + "learning_rate": 2.429906542056075e-06, + "loss": 6.1296, + "step": 91 + }, + { + "epoch": 0.007376079854081899, + "grad_norm": 3.6766316890716553, + "learning_rate": 2.456608811748999e-06, + "loss": 6.1362, + "step": 92 + }, + { + "epoch": 0.007456254635104528, + "grad_norm": 3.86043381690979, + "learning_rate": 2.4833110814419228e-06, + "loss": 6.2087, + "step": 93 + }, + { + "epoch": 0.007536429416127157, + "grad_norm": 3.5154123306274414, + "learning_rate": 2.5100133511348467e-06, + "loss": 6.095, + "step": 94 + }, + { + "epoch": 0.007616604197149787, + "grad_norm": 3.817383289337158, + "learning_rate": 2.5367156208277703e-06, + "loss": 6.1632, + "step": 95 + }, + { + "epoch": 0.007696778978172416, + "grad_norm": 3.4238457679748535, + "learning_rate": 2.5634178905206946e-06, + "loss": 6.0483, + "step": 96 + }, + { + "epoch": 0.007776953759195045, + "grad_norm": 3.875856399536133, + "learning_rate": 2.590120160213618e-06, + "loss": 6.0945, + "step": 97 + }, + { + "epoch": 0.007857128540217674, + "grad_norm": 3.5402724742889404, + "learning_rate": 2.616822429906542e-06, + "loss": 6.0867, + "step": 98 + }, + { + "epoch": 0.007937303321240303, + "grad_norm": 3.6940817832946777, + "learning_rate": 2.643524699599466e-06, + "loss": 6.0927, + "step": 99 + }, + { + "epoch": 0.008017478102262934, + "grad_norm": 3.3428852558135986, + "learning_rate": 2.67022696929239e-06, + "loss": 6.1257, + "step": 100 + }, + { + "epoch": 0.008097652883285563, + "grad_norm": 3.2528741359710693, + "learning_rate": 2.696929238985314e-06, + "loss": 6.1461, + "step": 101 + }, + { + "epoch": 0.008177827664308192, + "grad_norm": 3.1691038608551025, + "learning_rate": 2.723631508678238e-06, + "loss": 6.1077, + "step": 102 + }, + { + "epoch": 0.008258002445330821, + "grad_norm": 3.5406744480133057, + "learning_rate": 2.750333778371162e-06, + "loss": 6.167, + "step": 103 + }, + { + "epoch": 0.00833817722635345, + "grad_norm": 3.3365511894226074, + "learning_rate": 2.777036048064086e-06, + "loss": 6.1791, + "step": 104 + }, + { + "epoch": 0.00841835200737608, + "grad_norm": 3.6910934448242188, + "learning_rate": 2.8037383177570094e-06, + "loss": 6.047, + "step": 105 + }, + { + "epoch": 0.008498526788398708, + "grad_norm": 3.6962785720825195, + "learning_rate": 2.8304405874499337e-06, + "loss": 6.0857, + "step": 106 + }, + { + "epoch": 0.00857870156942134, + "grad_norm": 5.186034202575684, + "learning_rate": 2.8571428571428573e-06, + "loss": 5.9519, + "step": 107 + }, + { + "epoch": 0.008658876350443968, + "grad_norm": 3.5460336208343506, + "learning_rate": 2.8838451268357816e-06, + "loss": 6.2102, + "step": 108 + }, + { + "epoch": 0.008739051131466597, + "grad_norm": 3.830237627029419, + "learning_rate": 2.910547396528705e-06, + "loss": 6.2102, + "step": 109 + }, + { + "epoch": 0.008819225912489227, + "grad_norm": 3.5025956630706787, + "learning_rate": 2.9372496662216295e-06, + "loss": 6.1163, + "step": 110 + }, + { + "epoch": 0.008899400693511856, + "grad_norm": 3.107740879058838, + "learning_rate": 2.963951935914553e-06, + "loss": 6.0925, + "step": 111 + }, + { + "epoch": 0.008979575474534485, + "grad_norm": 3.040513753890991, + "learning_rate": 2.9906542056074766e-06, + "loss": 6.1516, + "step": 112 + }, + { + "epoch": 0.009059750255557114, + "grad_norm": 3.245520830154419, + "learning_rate": 3.017356475300401e-06, + "loss": 6.0083, + "step": 113 + }, + { + "epoch": 0.009139925036579745, + "grad_norm": 3.1702659130096436, + "learning_rate": 3.0440587449933245e-06, + "loss": 6.0596, + "step": 114 + }, + { + "epoch": 0.009220099817602374, + "grad_norm": 3.476616621017456, + "learning_rate": 3.070761014686249e-06, + "loss": 6.2571, + "step": 115 + }, + { + "epoch": 0.009300274598625003, + "grad_norm": 4.4942169189453125, + "learning_rate": 3.0974632843791724e-06, + "loss": 5.9487, + "step": 116 + }, + { + "epoch": 0.009380449379647632, + "grad_norm": 3.3257246017456055, + "learning_rate": 3.1241655540720968e-06, + "loss": 6.1038, + "step": 117 + }, + { + "epoch": 0.009460624160670261, + "grad_norm": 3.020477294921875, + "learning_rate": 3.1508678237650203e-06, + "loss": 6.1494, + "step": 118 + }, + { + "epoch": 0.00954079894169289, + "grad_norm": 3.195556640625, + "learning_rate": 3.177570093457944e-06, + "loss": 6.1268, + "step": 119 + }, + { + "epoch": 0.009620973722715519, + "grad_norm": 3.216067314147949, + "learning_rate": 3.204272363150868e-06, + "loss": 6.0798, + "step": 120 + }, + { + "epoch": 0.00970114850373815, + "grad_norm": 3.137930393218994, + "learning_rate": 3.2309746328437917e-06, + "loss": 6.1311, + "step": 121 + }, + { + "epoch": 0.009781323284760779, + "grad_norm": 3.1655943393707275, + "learning_rate": 3.257676902536716e-06, + "loss": 6.124, + "step": 122 + }, + { + "epoch": 0.009861498065783408, + "grad_norm": 3.0116324424743652, + "learning_rate": 3.2843791722296396e-06, + "loss": 6.0153, + "step": 123 + }, + { + "epoch": 0.009941672846806037, + "grad_norm": 3.4047839641571045, + "learning_rate": 3.3110814419225636e-06, + "loss": 6.1985, + "step": 124 + }, + { + "epoch": 0.010021847627828666, + "grad_norm": 4.143270969390869, + "learning_rate": 3.3377837116154875e-06, + "loss": 5.9226, + "step": 125 + }, + { + "epoch": 0.010102022408851295, + "grad_norm": 3.3798935413360596, + "learning_rate": 3.3644859813084115e-06, + "loss": 6.0543, + "step": 126 + }, + { + "epoch": 0.010182197189873924, + "grad_norm": 3.395967483520508, + "learning_rate": 3.3911882510013354e-06, + "loss": 6.1383, + "step": 127 + }, + { + "epoch": 0.010262371970896555, + "grad_norm": 2.83474063873291, + "learning_rate": 3.4178905206942594e-06, + "loss": 6.05, + "step": 128 + }, + { + "epoch": 0.010342546751919184, + "grad_norm": 3.3784477710723877, + "learning_rate": 3.4445927903871833e-06, + "loss": 6.1228, + "step": 129 + }, + { + "epoch": 0.010422721532941813, + "grad_norm": 4.242443561553955, + "learning_rate": 3.4712950600801073e-06, + "loss": 6.1528, + "step": 130 + }, + { + "epoch": 0.010502896313964442, + "grad_norm": 4.0656561851501465, + "learning_rate": 3.497997329773031e-06, + "loss": 6.042, + "step": 131 + }, + { + "epoch": 0.010583071094987071, + "grad_norm": 2.8319947719573975, + "learning_rate": 3.5246995994659548e-06, + "loss": 6.1633, + "step": 132 + }, + { + "epoch": 0.0106632458760097, + "grad_norm": 3.240278482437134, + "learning_rate": 3.5514018691588787e-06, + "loss": 6.0614, + "step": 133 + }, + { + "epoch": 0.01074342065703233, + "grad_norm": 3.4302539825439453, + "learning_rate": 3.5781041388518027e-06, + "loss": 6.0405, + "step": 134 + }, + { + "epoch": 0.01082359543805496, + "grad_norm": 3.8803529739379883, + "learning_rate": 3.6048064085447266e-06, + "loss": 5.8923, + "step": 135 + }, + { + "epoch": 0.01090377021907759, + "grad_norm": 3.091491460800171, + "learning_rate": 3.6315086782376506e-06, + "loss": 6.0547, + "step": 136 + }, + { + "epoch": 0.010983945000100219, + "grad_norm": 3.025940179824829, + "learning_rate": 3.6582109479305745e-06, + "loss": 6.0776, + "step": 137 + }, + { + "epoch": 0.011064119781122848, + "grad_norm": 3.2489027976989746, + "learning_rate": 3.684913217623498e-06, + "loss": 6.0486, + "step": 138 + }, + { + "epoch": 0.011144294562145477, + "grad_norm": 3.0891342163085938, + "learning_rate": 3.7116154873164224e-06, + "loss": 6.1169, + "step": 139 + }, + { + "epoch": 0.011224469343168106, + "grad_norm": 3.1752326488494873, + "learning_rate": 3.738317757009346e-06, + "loss": 6.022, + "step": 140 + }, + { + "epoch": 0.011304644124190737, + "grad_norm": 3.433783531188965, + "learning_rate": 3.7650200267022703e-06, + "loss": 6.1328, + "step": 141 + }, + { + "epoch": 0.011384818905213366, + "grad_norm": 3.5388479232788086, + "learning_rate": 3.791722296395194e-06, + "loss": 6.1719, + "step": 142 + }, + { + "epoch": 0.011464993686235995, + "grad_norm": 3.3045847415924072, + "learning_rate": 3.818424566088118e-06, + "loss": 6.0645, + "step": 143 + }, + { + "epoch": 0.011545168467258624, + "grad_norm": 2.962615728378296, + "learning_rate": 3.845126835781041e-06, + "loss": 6.0625, + "step": 144 + }, + { + "epoch": 0.011625343248281253, + "grad_norm": 3.5859076976776123, + "learning_rate": 3.871829105473966e-06, + "loss": 6.0049, + "step": 145 + }, + { + "epoch": 0.011705518029303882, + "grad_norm": 2.9961724281311035, + "learning_rate": 3.898531375166889e-06, + "loss": 6.104, + "step": 146 + }, + { + "epoch": 0.011785692810326511, + "grad_norm": 3.5061142444610596, + "learning_rate": 3.925233644859814e-06, + "loss": 6.0669, + "step": 147 + }, + { + "epoch": 0.011865867591349142, + "grad_norm": 3.187852621078491, + "learning_rate": 3.951935914552737e-06, + "loss": 6.106, + "step": 148 + }, + { + "epoch": 0.011946042372371771, + "grad_norm": 2.947300434112549, + "learning_rate": 3.9786381842456615e-06, + "loss": 6.1318, + "step": 149 + }, + { + "epoch": 0.0120262171533944, + "grad_norm": 3.7340900897979736, + "learning_rate": 4.005340453938585e-06, + "loss": 6.0068, + "step": 150 + }, + { + "epoch": 0.01210639193441703, + "grad_norm": 3.482743740081787, + "learning_rate": 4.032042723631509e-06, + "loss": 6.1985, + "step": 151 + }, + { + "epoch": 0.012186566715439658, + "grad_norm": 3.405768394470215, + "learning_rate": 4.058744993324433e-06, + "loss": 6.021, + "step": 152 + }, + { + "epoch": 0.012266741496462287, + "grad_norm": 3.3816328048706055, + "learning_rate": 4.085447263017357e-06, + "loss": 6.095, + "step": 153 + }, + { + "epoch": 0.012346916277484916, + "grad_norm": 3.51257061958313, + "learning_rate": 4.112149532710281e-06, + "loss": 6.1111, + "step": 154 + }, + { + "epoch": 0.012427091058507547, + "grad_norm": 3.154148817062378, + "learning_rate": 4.138851802403205e-06, + "loss": 6.0039, + "step": 155 + }, + { + "epoch": 0.012507265839530176, + "grad_norm": 3.5267789363861084, + "learning_rate": 4.165554072096129e-06, + "loss": 6.1777, + "step": 156 + }, + { + "epoch": 0.012587440620552805, + "grad_norm": 4.708312034606934, + "learning_rate": 4.192256341789052e-06, + "loss": 6.0571, + "step": 157 + }, + { + "epoch": 0.012667615401575434, + "grad_norm": 2.981755018234253, + "learning_rate": 4.218958611481977e-06, + "loss": 6.125, + "step": 158 + }, + { + "epoch": 0.012747790182598064, + "grad_norm": 3.1070406436920166, + "learning_rate": 4.2456608811749e-06, + "loss": 6.0342, + "step": 159 + }, + { + "epoch": 0.012827964963620693, + "grad_norm": 3.6669411659240723, + "learning_rate": 4.2723631508678245e-06, + "loss": 6.0466, + "step": 160 + }, + { + "epoch": 0.012908139744643322, + "grad_norm": 3.0786447525024414, + "learning_rate": 4.299065420560748e-06, + "loss": 6.0424, + "step": 161 + }, + { + "epoch": 0.012988314525665953, + "grad_norm": 2.858818769454956, + "learning_rate": 4.3257676902536724e-06, + "loss": 6.0072, + "step": 162 + }, + { + "epoch": 0.013068489306688582, + "grad_norm": 3.2203733921051025, + "learning_rate": 4.352469959946596e-06, + "loss": 6.1304, + "step": 163 + }, + { + "epoch": 0.01314866408771121, + "grad_norm": 3.2318503856658936, + "learning_rate": 4.3791722296395195e-06, + "loss": 5.981, + "step": 164 + }, + { + "epoch": 0.01322883886873384, + "grad_norm": 3.169086456298828, + "learning_rate": 4.405874499332444e-06, + "loss": 6.1033, + "step": 165 + }, + { + "epoch": 0.013309013649756469, + "grad_norm": 3.2154383659362793, + "learning_rate": 4.432576769025367e-06, + "loss": 6.0525, + "step": 166 + }, + { + "epoch": 0.013389188430779098, + "grad_norm": 3.109374761581421, + "learning_rate": 4.459279038718292e-06, + "loss": 5.9924, + "step": 167 + }, + { + "epoch": 0.013469363211801727, + "grad_norm": 2.824176549911499, + "learning_rate": 4.485981308411215e-06, + "loss": 6.0513, + "step": 168 + }, + { + "epoch": 0.013549537992824358, + "grad_norm": 3.2132716178894043, + "learning_rate": 4.51268357810414e-06, + "loss": 6.0593, + "step": 169 + }, + { + "epoch": 0.013629712773846987, + "grad_norm": 3.0032589435577393, + "learning_rate": 4.539385847797063e-06, + "loss": 5.9905, + "step": 170 + }, + { + "epoch": 0.013709887554869616, + "grad_norm": 3.6031315326690674, + "learning_rate": 4.566088117489987e-06, + "loss": 6.0459, + "step": 171 + }, + { + "epoch": 0.013790062335892245, + "grad_norm": 3.2631516456604004, + "learning_rate": 4.592790387182911e-06, + "loss": 6.0264, + "step": 172 + }, + { + "epoch": 0.013870237116914874, + "grad_norm": 3.025954008102417, + "learning_rate": 4.619492656875835e-06, + "loss": 6.0911, + "step": 173 + }, + { + "epoch": 0.013950411897937503, + "grad_norm": 3.3582992553710938, + "learning_rate": 4.646194926568759e-06, + "loss": 6.0356, + "step": 174 + }, + { + "epoch": 0.014030586678960132, + "grad_norm": 3.1111085414886475, + "learning_rate": 4.6728971962616825e-06, + "loss": 6.0647, + "step": 175 + }, + { + "epoch": 0.014110761459982763, + "grad_norm": 3.1111905574798584, + "learning_rate": 4.699599465954606e-06, + "loss": 6.056, + "step": 176 + }, + { + "epoch": 0.014190936241005392, + "grad_norm": 3.0012784004211426, + "learning_rate": 4.7263017356475304e-06, + "loss": 6.0081, + "step": 177 + }, + { + "epoch": 0.014271111022028021, + "grad_norm": 3.5163426399230957, + "learning_rate": 4.753004005340454e-06, + "loss": 5.9742, + "step": 178 + }, + { + "epoch": 0.01435128580305065, + "grad_norm": 3.5614511966705322, + "learning_rate": 4.779706275033378e-06, + "loss": 5.9844, + "step": 179 + }, + { + "epoch": 0.01443146058407328, + "grad_norm": 3.287126302719116, + "learning_rate": 4.806408544726302e-06, + "loss": 6.072, + "step": 180 + }, + { + "epoch": 0.014511635365095909, + "grad_norm": 2.595430374145508, + "learning_rate": 4.833110814419226e-06, + "loss": 5.979, + "step": 181 + }, + { + "epoch": 0.014591810146118538, + "grad_norm": 3.512235403060913, + "learning_rate": 4.85981308411215e-06, + "loss": 6.0322, + "step": 182 + }, + { + "epoch": 0.014671984927141168, + "grad_norm": 3.392423152923584, + "learning_rate": 4.886515353805073e-06, + "loss": 6.0334, + "step": 183 + }, + { + "epoch": 0.014752159708163797, + "grad_norm": 4.059785842895508, + "learning_rate": 4.913217623497998e-06, + "loss": 6.1361, + "step": 184 + }, + { + "epoch": 0.014832334489186427, + "grad_norm": 2.7503762245178223, + "learning_rate": 4.939919893190921e-06, + "loss": 6.0994, + "step": 185 + }, + { + "epoch": 0.014912509270209056, + "grad_norm": 4.435471057891846, + "learning_rate": 4.9666221628838456e-06, + "loss": 5.9637, + "step": 186 + }, + { + "epoch": 0.014992684051231685, + "grad_norm": 4.091432571411133, + "learning_rate": 4.993324432576769e-06, + "loss": 6.022, + "step": 187 + }, + { + "epoch": 0.015072858832254314, + "grad_norm": 3.254970073699951, + "learning_rate": 5.0200267022696935e-06, + "loss": 6.0332, + "step": 188 + }, + { + "epoch": 0.015153033613276943, + "grad_norm": 3.826946496963501, + "learning_rate": 5.046728971962617e-06, + "loss": 6.0308, + "step": 189 + }, + { + "epoch": 0.015233208394299574, + "grad_norm": 3.3818719387054443, + "learning_rate": 5.0734312416555405e-06, + "loss": 5.8511, + "step": 190 + }, + { + "epoch": 0.015313383175322203, + "grad_norm": 4.139673233032227, + "learning_rate": 5.100133511348465e-06, + "loss": 5.9436, + "step": 191 + }, + { + "epoch": 0.015393557956344832, + "grad_norm": 3.093797445297241, + "learning_rate": 5.126835781041389e-06, + "loss": 6.0295, + "step": 192 + }, + { + "epoch": 0.015473732737367461, + "grad_norm": 3.0224900245666504, + "learning_rate": 5.153538050734313e-06, + "loss": 6.0261, + "step": 193 + }, + { + "epoch": 0.01555390751839009, + "grad_norm": 3.3340024948120117, + "learning_rate": 5.180240320427236e-06, + "loss": 6.1519, + "step": 194 + }, + { + "epoch": 0.01563408229941272, + "grad_norm": 3.5352375507354736, + "learning_rate": 5.206942590120161e-06, + "loss": 6.0277, + "step": 195 + }, + { + "epoch": 0.015714257080435348, + "grad_norm": 2.964306592941284, + "learning_rate": 5.233644859813084e-06, + "loss": 5.9905, + "step": 196 + }, + { + "epoch": 0.01579443186145798, + "grad_norm": 2.7662911415100098, + "learning_rate": 5.260347129506009e-06, + "loss": 6.0022, + "step": 197 + }, + { + "epoch": 0.015874606642480606, + "grad_norm": 4.254642486572266, + "learning_rate": 5.287049399198932e-06, + "loss": 6.0006, + "step": 198 + }, + { + "epoch": 0.015954781423503237, + "grad_norm": 3.3264899253845215, + "learning_rate": 5.3137516688918565e-06, + "loss": 5.9354, + "step": 199 + }, + { + "epoch": 0.016034956204525868, + "grad_norm": 3.080687999725342, + "learning_rate": 5.34045393858478e-06, + "loss": 6.0056, + "step": 200 + }, + { + "epoch": 0.016115130985548495, + "grad_norm": 3.630162000656128, + "learning_rate": 5.367156208277704e-06, + "loss": 5.937, + "step": 201 + }, + { + "epoch": 0.016195305766571126, + "grad_norm": 2.8518264293670654, + "learning_rate": 5.393858477970628e-06, + "loss": 5.8618, + "step": 202 + }, + { + "epoch": 0.016275480547593753, + "grad_norm": 4.150312423706055, + "learning_rate": 5.4205607476635515e-06, + "loss": 6.1011, + "step": 203 + }, + { + "epoch": 0.016355655328616384, + "grad_norm": 3.1485273838043213, + "learning_rate": 5.447263017356476e-06, + "loss": 6.0383, + "step": 204 + }, + { + "epoch": 0.01643583010963901, + "grad_norm": 3.413485288619995, + "learning_rate": 5.473965287049399e-06, + "loss": 6.0024, + "step": 205 + }, + { + "epoch": 0.016516004890661642, + "grad_norm": 3.266434669494629, + "learning_rate": 5.500667556742324e-06, + "loss": 5.9679, + "step": 206 + }, + { + "epoch": 0.016596179671684273, + "grad_norm": 3.4536993503570557, + "learning_rate": 5.527369826435247e-06, + "loss": 6.0288, + "step": 207 + }, + { + "epoch": 0.0166763544527069, + "grad_norm": 3.0397145748138428, + "learning_rate": 5.554072096128172e-06, + "loss": 5.9945, + "step": 208 + }, + { + "epoch": 0.01675652923372953, + "grad_norm": 3.3408124446868896, + "learning_rate": 5.580774365821095e-06, + "loss": 6.0261, + "step": 209 + }, + { + "epoch": 0.01683670401475216, + "grad_norm": 3.554244041442871, + "learning_rate": 5.607476635514019e-06, + "loss": 6.0284, + "step": 210 + }, + { + "epoch": 0.01691687879577479, + "grad_norm": 2.6615023612976074, + "learning_rate": 5.634178905206943e-06, + "loss": 5.8354, + "step": 211 + }, + { + "epoch": 0.016997053576797417, + "grad_norm": 3.4736216068267822, + "learning_rate": 5.6608811748998675e-06, + "loss": 6.0413, + "step": 212 + }, + { + "epoch": 0.017077228357820048, + "grad_norm": 2.8463640213012695, + "learning_rate": 5.687583444592791e-06, + "loss": 6.0239, + "step": 213 + }, + { + "epoch": 0.01715740313884268, + "grad_norm": 3.506805658340454, + "learning_rate": 5.7142857142857145e-06, + "loss": 6.0391, + "step": 214 + }, + { + "epoch": 0.017237577919865306, + "grad_norm": 2.5320992469787598, + "learning_rate": 5.740987983978639e-06, + "loss": 5.8218, + "step": 215 + }, + { + "epoch": 0.017317752700887937, + "grad_norm": 3.1186769008636475, + "learning_rate": 5.767690253671563e-06, + "loss": 5.9712, + "step": 216 + }, + { + "epoch": 0.017397927481910564, + "grad_norm": 3.564192056655884, + "learning_rate": 5.794392523364486e-06, + "loss": 5.9692, + "step": 217 + }, + { + "epoch": 0.017478102262933195, + "grad_norm": 3.294774293899536, + "learning_rate": 5.82109479305741e-06, + "loss": 6.0161, + "step": 218 + }, + { + "epoch": 0.017558277043955822, + "grad_norm": 3.3062896728515625, + "learning_rate": 5.847797062750335e-06, + "loss": 5.9558, + "step": 219 + }, + { + "epoch": 0.017638451824978453, + "grad_norm": 3.612069606781006, + "learning_rate": 5.874499332443259e-06, + "loss": 6.0112, + "step": 220 + }, + { + "epoch": 0.017718626606001084, + "grad_norm": 2.833406686782837, + "learning_rate": 5.901201602136182e-06, + "loss": 6.064, + "step": 221 + }, + { + "epoch": 0.01779880138702371, + "grad_norm": 4.009668827056885, + "learning_rate": 5.927903871829106e-06, + "loss": 5.9995, + "step": 222 + }, + { + "epoch": 0.017878976168046342, + "grad_norm": 3.409745931625366, + "learning_rate": 5.9546061415220305e-06, + "loss": 5.9816, + "step": 223 + }, + { + "epoch": 0.01795915094906897, + "grad_norm": 3.511864185333252, + "learning_rate": 5.981308411214953e-06, + "loss": 6.0234, + "step": 224 + }, + { + "epoch": 0.0180393257300916, + "grad_norm": 3.0301144123077393, + "learning_rate": 6.0080106809078775e-06, + "loss": 6.0752, + "step": 225 + }, + { + "epoch": 0.018119500511114228, + "grad_norm": 3.5683367252349854, + "learning_rate": 6.034712950600802e-06, + "loss": 5.9509, + "step": 226 + }, + { + "epoch": 0.01819967529213686, + "grad_norm": 3.170880079269409, + "learning_rate": 6.061415220293726e-06, + "loss": 5.9591, + "step": 227 + }, + { + "epoch": 0.01827985007315949, + "grad_norm": 2.677696466445923, + "learning_rate": 6.088117489986649e-06, + "loss": 5.9866, + "step": 228 + }, + { + "epoch": 0.018360024854182116, + "grad_norm": 2.9106850624084473, + "learning_rate": 6.114819759679573e-06, + "loss": 5.9263, + "step": 229 + }, + { + "epoch": 0.018440199635204747, + "grad_norm": 3.869394302368164, + "learning_rate": 6.141522029372498e-06, + "loss": 6.0288, + "step": 230 + }, + { + "epoch": 0.018520374416227375, + "grad_norm": 4.135583400726318, + "learning_rate": 6.16822429906542e-06, + "loss": 5.9596, + "step": 231 + }, + { + "epoch": 0.018600549197250005, + "grad_norm": 2.8029985427856445, + "learning_rate": 6.194926568758345e-06, + "loss": 6.0073, + "step": 232 + }, + { + "epoch": 0.018680723978272633, + "grad_norm": 3.0072453022003174, + "learning_rate": 6.221628838451269e-06, + "loss": 5.8774, + "step": 233 + }, + { + "epoch": 0.018760898759295264, + "grad_norm": 2.924696207046509, + "learning_rate": 6.2483311081441935e-06, + "loss": 5.9709, + "step": 234 + }, + { + "epoch": 0.018841073540317894, + "grad_norm": 2.8260371685028076, + "learning_rate": 6.275033377837116e-06, + "loss": 5.8042, + "step": 235 + }, + { + "epoch": 0.018921248321340522, + "grad_norm": 3.648798942565918, + "learning_rate": 6.301735647530041e-06, + "loss": 5.8909, + "step": 236 + }, + { + "epoch": 0.019001423102363153, + "grad_norm": 3.2967917919158936, + "learning_rate": 6.328437917222965e-06, + "loss": 6.0259, + "step": 237 + }, + { + "epoch": 0.01908159788338578, + "grad_norm": 3.8249049186706543, + "learning_rate": 6.355140186915888e-06, + "loss": 5.998, + "step": 238 + }, + { + "epoch": 0.01916177266440841, + "grad_norm": 3.970259189605713, + "learning_rate": 6.381842456608812e-06, + "loss": 6.0732, + "step": 239 + }, + { + "epoch": 0.019241947445431038, + "grad_norm": 3.506058692932129, + "learning_rate": 6.408544726301736e-06, + "loss": 5.8301, + "step": 240 + }, + { + "epoch": 0.01932212222645367, + "grad_norm": 4.412944793701172, + "learning_rate": 6.43524699599466e-06, + "loss": 6.0637, + "step": 241 + }, + { + "epoch": 0.0194022970074763, + "grad_norm": 3.995609760284424, + "learning_rate": 6.4619492656875834e-06, + "loss": 5.9175, + "step": 242 + }, + { + "epoch": 0.019482471788498927, + "grad_norm": 2.9621944427490234, + "learning_rate": 6.488651535380508e-06, + "loss": 5.8523, + "step": 243 + }, + { + "epoch": 0.019562646569521558, + "grad_norm": 4.315832614898682, + "learning_rate": 6.515353805073432e-06, + "loss": 6.0303, + "step": 244 + }, + { + "epoch": 0.019642821350544185, + "grad_norm": 3.1982245445251465, + "learning_rate": 6.542056074766355e-06, + "loss": 5.918, + "step": 245 + }, + { + "epoch": 0.019722996131566816, + "grad_norm": 3.031961441040039, + "learning_rate": 6.568758344459279e-06, + "loss": 5.8967, + "step": 246 + }, + { + "epoch": 0.019803170912589443, + "grad_norm": 3.9194893836975098, + "learning_rate": 6.595460614152204e-06, + "loss": 6.021, + "step": 247 + }, + { + "epoch": 0.019883345693612074, + "grad_norm": 3.086456060409546, + "learning_rate": 6.622162883845127e-06, + "loss": 5.9891, + "step": 248 + }, + { + "epoch": 0.019963520474634705, + "grad_norm": 5.396167278289795, + "learning_rate": 6.648865153538051e-06, + "loss": 5.9596, + "step": 249 + }, + { + "epoch": 0.020043695255657332, + "grad_norm": 3.500825881958008, + "learning_rate": 6.675567423230975e-06, + "loss": 5.9827, + "step": 250 + }, + { + "epoch": 0.020123870036679963, + "grad_norm": 3.495565176010132, + "learning_rate": 6.7022696929238994e-06, + "loss": 5.9856, + "step": 251 + }, + { + "epoch": 0.02020404481770259, + "grad_norm": 4.299311637878418, + "learning_rate": 6.728971962616823e-06, + "loss": 5.9336, + "step": 252 + }, + { + "epoch": 0.02028421959872522, + "grad_norm": 2.545771360397339, + "learning_rate": 6.7556742323097465e-06, + "loss": 6.1379, + "step": 253 + }, + { + "epoch": 0.02036439437974785, + "grad_norm": 3.0301387310028076, + "learning_rate": 6.782376502002671e-06, + "loss": 5.9316, + "step": 254 + }, + { + "epoch": 0.02044456916077048, + "grad_norm": 3.0030181407928467, + "learning_rate": 6.809078771695594e-06, + "loss": 5.8123, + "step": 255 + }, + { + "epoch": 0.02052474394179311, + "grad_norm": 3.044619083404541, + "learning_rate": 6.835781041388519e-06, + "loss": 5.9857, + "step": 256 + }, + { + "epoch": 0.020604918722815738, + "grad_norm": 3.0375585556030273, + "learning_rate": 6.862483311081442e-06, + "loss": 5.9001, + "step": 257 + }, + { + "epoch": 0.02068509350383837, + "grad_norm": 3.748898983001709, + "learning_rate": 6.889185580774367e-06, + "loss": 5.9524, + "step": 258 + }, + { + "epoch": 0.020765268284860996, + "grad_norm": 3.786522388458252, + "learning_rate": 6.91588785046729e-06, + "loss": 6.0149, + "step": 259 + }, + { + "epoch": 0.020845443065883627, + "grad_norm": 2.939491033554077, + "learning_rate": 6.9425901201602146e-06, + "loss": 5.9749, + "step": 260 + }, + { + "epoch": 0.020925617846906254, + "grad_norm": 3.5087053775787354, + "learning_rate": 6.969292389853138e-06, + "loss": 5.9238, + "step": 261 + }, + { + "epoch": 0.021005792627928885, + "grad_norm": 4.024533748626709, + "learning_rate": 6.995994659546062e-06, + "loss": 5.8486, + "step": 262 + }, + { + "epoch": 0.021085967408951516, + "grad_norm": 3.3081352710723877, + "learning_rate": 7.022696929238986e-06, + "loss": 5.9058, + "step": 263 + }, + { + "epoch": 0.021166142189974143, + "grad_norm": 3.133509874343872, + "learning_rate": 7.0493991989319095e-06, + "loss": 6.0046, + "step": 264 + }, + { + "epoch": 0.021246316970996774, + "grad_norm": 2.961247682571411, + "learning_rate": 7.076101468624834e-06, + "loss": 5.9517, + "step": 265 + }, + { + "epoch": 0.0213264917520194, + "grad_norm": 3.5335538387298584, + "learning_rate": 7.1028037383177574e-06, + "loss": 6.01, + "step": 266 + }, + { + "epoch": 0.021406666533042032, + "grad_norm": 3.5015270709991455, + "learning_rate": 7.129506008010682e-06, + "loss": 5.9231, + "step": 267 + }, + { + "epoch": 0.02148684131406466, + "grad_norm": 3.1063270568847656, + "learning_rate": 7.156208277703605e-06, + "loss": 5.9476, + "step": 268 + }, + { + "epoch": 0.02156701609508729, + "grad_norm": 2.224583864212036, + "learning_rate": 7.182910547396529e-06, + "loss": 5.7954, + "step": 269 + }, + { + "epoch": 0.02164719087610992, + "grad_norm": 3.989407539367676, + "learning_rate": 7.209612817089453e-06, + "loss": 5.9867, + "step": 270 + }, + { + "epoch": 0.021727365657132548, + "grad_norm": 2.9691379070281982, + "learning_rate": 7.236315086782378e-06, + "loss": 5.9825, + "step": 271 + }, + { + "epoch": 0.02180754043815518, + "grad_norm": 4.410327911376953, + "learning_rate": 7.263017356475301e-06, + "loss": 5.8618, + "step": 272 + }, + { + "epoch": 0.021887715219177806, + "grad_norm": 2.8465473651885986, + "learning_rate": 7.289719626168225e-06, + "loss": 5.9121, + "step": 273 + }, + { + "epoch": 0.021967890000200437, + "grad_norm": 3.088122606277466, + "learning_rate": 7.316421895861149e-06, + "loss": 5.9934, + "step": 274 + }, + { + "epoch": 0.022048064781223065, + "grad_norm": 2.9295144081115723, + "learning_rate": 7.343124165554073e-06, + "loss": 5.9357, + "step": 275 + }, + { + "epoch": 0.022128239562245695, + "grad_norm": 2.9746243953704834, + "learning_rate": 7.369826435246996e-06, + "loss": 5.9584, + "step": 276 + }, + { + "epoch": 0.022208414343268326, + "grad_norm": 4.812570571899414, + "learning_rate": 7.3965287049399205e-06, + "loss": 5.8877, + "step": 277 + }, + { + "epoch": 0.022288589124290954, + "grad_norm": 3.2310681343078613, + "learning_rate": 7.423230974632845e-06, + "loss": 5.9873, + "step": 278 + }, + { + "epoch": 0.022368763905313584, + "grad_norm": 3.0134434700012207, + "learning_rate": 7.449933244325769e-06, + "loss": 5.9313, + "step": 279 + }, + { + "epoch": 0.02244893868633621, + "grad_norm": 3.4617674350738525, + "learning_rate": 7.476635514018692e-06, + "loss": 5.9612, + "step": 280 + }, + { + "epoch": 0.022529113467358843, + "grad_norm": 3.7327933311462402, + "learning_rate": 7.503337783711616e-06, + "loss": 5.9009, + "step": 281 + }, + { + "epoch": 0.022609288248381473, + "grad_norm": 5.069031715393066, + "learning_rate": 7.530040053404541e-06, + "loss": 5.967, + "step": 282 + }, + { + "epoch": 0.0226894630294041, + "grad_norm": 2.840934991836548, + "learning_rate": 7.556742323097463e-06, + "loss": 5.9878, + "step": 283 + }, + { + "epoch": 0.02276963781042673, + "grad_norm": 3.9743142127990723, + "learning_rate": 7.583444592790388e-06, + "loss": 6.0393, + "step": 284 + }, + { + "epoch": 0.02284981259144936, + "grad_norm": 3.124641180038452, + "learning_rate": 7.610146862483312e-06, + "loss": 5.8733, + "step": 285 + }, + { + "epoch": 0.02292998737247199, + "grad_norm": 2.92891788482666, + "learning_rate": 7.636849132176236e-06, + "loss": 5.9604, + "step": 286 + }, + { + "epoch": 0.023010162153494617, + "grad_norm": 3.011392831802368, + "learning_rate": 7.663551401869159e-06, + "loss": 5.9341, + "step": 287 + }, + { + "epoch": 0.023090336934517248, + "grad_norm": 3.8497419357299805, + "learning_rate": 7.690253671562083e-06, + "loss": 5.9987, + "step": 288 + }, + { + "epoch": 0.02317051171553988, + "grad_norm": 3.195448160171509, + "learning_rate": 7.716955941255008e-06, + "loss": 5.9371, + "step": 289 + }, + { + "epoch": 0.023250686496562506, + "grad_norm": 3.4858617782592773, + "learning_rate": 7.743658210947931e-06, + "loss": 5.9299, + "step": 290 + }, + { + "epoch": 0.023330861277585137, + "grad_norm": 3.4011356830596924, + "learning_rate": 7.770360480640855e-06, + "loss": 5.8769, + "step": 291 + }, + { + "epoch": 0.023411036058607764, + "grad_norm": 3.0267844200134277, + "learning_rate": 7.797062750333778e-06, + "loss": 5.8655, + "step": 292 + }, + { + "epoch": 0.023491210839630395, + "grad_norm": 4.010112285614014, + "learning_rate": 7.823765020026704e-06, + "loss": 5.9409, + "step": 293 + }, + { + "epoch": 0.023571385620653022, + "grad_norm": 3.274632215499878, + "learning_rate": 7.850467289719627e-06, + "loss": 5.9333, + "step": 294 + }, + { + "epoch": 0.023651560401675653, + "grad_norm": 3.00844407081604, + "learning_rate": 7.87716955941255e-06, + "loss": 5.9319, + "step": 295 + }, + { + "epoch": 0.023731735182698284, + "grad_norm": 3.6330296993255615, + "learning_rate": 7.903871829105474e-06, + "loss": 5.8765, + "step": 296 + }, + { + "epoch": 0.02381190996372091, + "grad_norm": 3.030797243118286, + "learning_rate": 7.930574098798398e-06, + "loss": 5.9042, + "step": 297 + }, + { + "epoch": 0.023892084744743542, + "grad_norm": 3.3744027614593506, + "learning_rate": 7.957276368491323e-06, + "loss": 5.9618, + "step": 298 + }, + { + "epoch": 0.02397225952576617, + "grad_norm": 3.2510626316070557, + "learning_rate": 7.983978638184247e-06, + "loss": 5.9741, + "step": 299 + }, + { + "epoch": 0.0240524343067888, + "grad_norm": 3.2552661895751953, + "learning_rate": 8.01068090787717e-06, + "loss": 5.8994, + "step": 300 + }, + { + "epoch": 0.024132609087811428, + "grad_norm": 3.164968967437744, + "learning_rate": 8.037383177570094e-06, + "loss": 5.8691, + "step": 301 + }, + { + "epoch": 0.02421278386883406, + "grad_norm": 2.7807435989379883, + "learning_rate": 8.064085447263019e-06, + "loss": 5.9983, + "step": 302 + }, + { + "epoch": 0.02429295864985669, + "grad_norm": 5.789491653442383, + "learning_rate": 8.090787716955942e-06, + "loss": 6.0044, + "step": 303 + }, + { + "epoch": 0.024373133430879317, + "grad_norm": 4.088723182678223, + "learning_rate": 8.117489986648866e-06, + "loss": 5.8722, + "step": 304 + }, + { + "epoch": 0.024453308211901947, + "grad_norm": 4.874099254608154, + "learning_rate": 8.14419225634179e-06, + "loss": 6.0114, + "step": 305 + }, + { + "epoch": 0.024533482992924575, + "grad_norm": 3.980386972427368, + "learning_rate": 8.170894526034715e-06, + "loss": 5.8403, + "step": 306 + }, + { + "epoch": 0.024613657773947206, + "grad_norm": 3.265903949737549, + "learning_rate": 8.197596795727636e-06, + "loss": 5.9144, + "step": 307 + }, + { + "epoch": 0.024693832554969833, + "grad_norm": 4.518984794616699, + "learning_rate": 8.224299065420562e-06, + "loss": 5.8799, + "step": 308 + }, + { + "epoch": 0.024774007335992464, + "grad_norm": 3.1703402996063232, + "learning_rate": 8.251001335113485e-06, + "loss": 5.9606, + "step": 309 + }, + { + "epoch": 0.024854182117015094, + "grad_norm": 5.1741862297058105, + "learning_rate": 8.27770360480641e-06, + "loss": 5.9331, + "step": 310 + }, + { + "epoch": 0.024934356898037722, + "grad_norm": 4.099031925201416, + "learning_rate": 8.304405874499332e-06, + "loss": 5.8888, + "step": 311 + }, + { + "epoch": 0.025014531679060353, + "grad_norm": 2.8996481895446777, + "learning_rate": 8.331108144192257e-06, + "loss": 5.9803, + "step": 312 + }, + { + "epoch": 0.02509470646008298, + "grad_norm": 3.3852181434631348, + "learning_rate": 8.357810413885181e-06, + "loss": 5.9194, + "step": 313 + }, + { + "epoch": 0.02517488124110561, + "grad_norm": 4.47113561630249, + "learning_rate": 8.384512683578105e-06, + "loss": 5.9519, + "step": 314 + }, + { + "epoch": 0.025255056022128238, + "grad_norm": 3.2112221717834473, + "learning_rate": 8.411214953271028e-06, + "loss": 5.9128, + "step": 315 + }, + { + "epoch": 0.02533523080315087, + "grad_norm": 2.519045114517212, + "learning_rate": 8.437917222963953e-06, + "loss": 5.9584, + "step": 316 + }, + { + "epoch": 0.0254154055841735, + "grad_norm": 2.5920066833496094, + "learning_rate": 8.464619492656877e-06, + "loss": 5.7656, + "step": 317 + }, + { + "epoch": 0.025495580365196127, + "grad_norm": 4.73217248916626, + "learning_rate": 8.4913217623498e-06, + "loss": 5.832, + "step": 318 + }, + { + "epoch": 0.025575755146218758, + "grad_norm": 3.7621548175811768, + "learning_rate": 8.518024032042724e-06, + "loss": 5.894, + "step": 319 + }, + { + "epoch": 0.025655929927241385, + "grad_norm": 3.5031847953796387, + "learning_rate": 8.544726301735649e-06, + "loss": 5.8987, + "step": 320 + }, + { + "epoch": 0.025736104708264016, + "grad_norm": 4.1335978507995605, + "learning_rate": 8.571428571428571e-06, + "loss": 5.9121, + "step": 321 + }, + { + "epoch": 0.025816279489286643, + "grad_norm": 3.5417020320892334, + "learning_rate": 8.598130841121496e-06, + "loss": 5.9136, + "step": 322 + }, + { + "epoch": 0.025896454270309274, + "grad_norm": 2.7796106338500977, + "learning_rate": 8.62483311081442e-06, + "loss": 5.8706, + "step": 323 + }, + { + "epoch": 0.025976629051331905, + "grad_norm": 2.9711556434631348, + "learning_rate": 8.651535380507345e-06, + "loss": 5.9401, + "step": 324 + }, + { + "epoch": 0.026056803832354532, + "grad_norm": 3.632439613342285, + "learning_rate": 8.678237650200267e-06, + "loss": 5.8008, + "step": 325 + }, + { + "epoch": 0.026136978613377163, + "grad_norm": 4.166197299957275, + "learning_rate": 8.704939919893192e-06, + "loss": 5.8297, + "step": 326 + }, + { + "epoch": 0.02621715339439979, + "grad_norm": 3.3072633743286133, + "learning_rate": 8.731642189586115e-06, + "loss": 5.9941, + "step": 327 + }, + { + "epoch": 0.02629732817542242, + "grad_norm": 4.302679061889648, + "learning_rate": 8.758344459279039e-06, + "loss": 5.8904, + "step": 328 + }, + { + "epoch": 0.02637750295644505, + "grad_norm": 3.6804211139678955, + "learning_rate": 8.785046728971963e-06, + "loss": 6.0, + "step": 329 + }, + { + "epoch": 0.02645767773746768, + "grad_norm": 3.431161642074585, + "learning_rate": 8.811748998664888e-06, + "loss": 5.9258, + "step": 330 + }, + { + "epoch": 0.02653785251849031, + "grad_norm": 1.9185757637023926, + "learning_rate": 8.838451268357811e-06, + "loss": 5.7432, + "step": 331 + }, + { + "epoch": 0.026618027299512938, + "grad_norm": 3.788281202316284, + "learning_rate": 8.865153538050735e-06, + "loss": 5.964, + "step": 332 + }, + { + "epoch": 0.02669820208053557, + "grad_norm": 3.4814445972442627, + "learning_rate": 8.891855807743658e-06, + "loss": 5.8641, + "step": 333 + }, + { + "epoch": 0.026778376861558196, + "grad_norm": 3.3292438983917236, + "learning_rate": 8.918558077436584e-06, + "loss": 5.9293, + "step": 334 + }, + { + "epoch": 0.026858551642580827, + "grad_norm": 3.2235262393951416, + "learning_rate": 8.945260347129507e-06, + "loss": 5.7731, + "step": 335 + }, + { + "epoch": 0.026938726423603454, + "grad_norm": 3.107299327850342, + "learning_rate": 8.97196261682243e-06, + "loss": 5.9431, + "step": 336 + }, + { + "epoch": 0.027018901204626085, + "grad_norm": 3.1460602283477783, + "learning_rate": 8.998664886515354e-06, + "loss": 5.8346, + "step": 337 + }, + { + "epoch": 0.027099075985648716, + "grad_norm": 3.2336511611938477, + "learning_rate": 9.02536715620828e-06, + "loss": 5.9188, + "step": 338 + }, + { + "epoch": 0.027179250766671343, + "grad_norm": 3.237821340560913, + "learning_rate": 9.052069425901203e-06, + "loss": 5.8444, + "step": 339 + }, + { + "epoch": 0.027259425547693974, + "grad_norm": 3.0538432598114014, + "learning_rate": 9.078771695594126e-06, + "loss": 5.85, + "step": 340 + }, + { + "epoch": 0.0273396003287166, + "grad_norm": 3.1638779640197754, + "learning_rate": 9.10547396528705e-06, + "loss": 5.8689, + "step": 341 + }, + { + "epoch": 0.027419775109739232, + "grad_norm": 2.9599223136901855, + "learning_rate": 9.132176234979973e-06, + "loss": 5.8799, + "step": 342 + }, + { + "epoch": 0.02749994989076186, + "grad_norm": 2.553194761276245, + "learning_rate": 9.158878504672899e-06, + "loss": 5.9253, + "step": 343 + }, + { + "epoch": 0.02758012467178449, + "grad_norm": 3.100206136703491, + "learning_rate": 9.185580774365822e-06, + "loss": 5.8259, + "step": 344 + }, + { + "epoch": 0.02766029945280712, + "grad_norm": 3.091893196105957, + "learning_rate": 9.212283044058746e-06, + "loss": 5.7659, + "step": 345 + }, + { + "epoch": 0.02774047423382975, + "grad_norm": 2.835014820098877, + "learning_rate": 9.23898531375167e-06, + "loss": 5.8224, + "step": 346 + }, + { + "epoch": 0.02782064901485238, + "grad_norm": 2.857104539871216, + "learning_rate": 9.265687583444593e-06, + "loss": 5.8082, + "step": 347 + }, + { + "epoch": 0.027900823795875006, + "grad_norm": 2.968951940536499, + "learning_rate": 9.292389853137518e-06, + "loss": 5.877, + "step": 348 + }, + { + "epoch": 0.027980998576897637, + "grad_norm": 3.438591957092285, + "learning_rate": 9.319092122830442e-06, + "loss": 5.9355, + "step": 349 + }, + { + "epoch": 0.028061173357920265, + "grad_norm": 2.5796146392822266, + "learning_rate": 9.345794392523365e-06, + "loss": 5.9634, + "step": 350 + }, + { + "epoch": 0.028141348138942895, + "grad_norm": 2.911288261413574, + "learning_rate": 9.372496662216289e-06, + "loss": 5.8599, + "step": 351 + }, + { + "epoch": 0.028221522919965526, + "grad_norm": 2.6005191802978516, + "learning_rate": 9.399198931909212e-06, + "loss": 5.8506, + "step": 352 + }, + { + "epoch": 0.028301697700988154, + "grad_norm": 3.57360577583313, + "learning_rate": 9.425901201602137e-06, + "loss": 5.9231, + "step": 353 + }, + { + "epoch": 0.028381872482010784, + "grad_norm": 2.73592472076416, + "learning_rate": 9.452603471295061e-06, + "loss": 5.8929, + "step": 354 + }, + { + "epoch": 0.028462047263033412, + "grad_norm": 3.018139123916626, + "learning_rate": 9.479305740987984e-06, + "loss": 6.0208, + "step": 355 + }, + { + "epoch": 0.028542222044056043, + "grad_norm": 3.265169382095337, + "learning_rate": 9.506008010680908e-06, + "loss": 5.8508, + "step": 356 + }, + { + "epoch": 0.02862239682507867, + "grad_norm": 2.748033285140991, + "learning_rate": 9.532710280373833e-06, + "loss": 5.8301, + "step": 357 + }, + { + "epoch": 0.0287025716061013, + "grad_norm": 2.736762762069702, + "learning_rate": 9.559412550066757e-06, + "loss": 6.0155, + "step": 358 + }, + { + "epoch": 0.02878274638712393, + "grad_norm": 3.415748119354248, + "learning_rate": 9.58611481975968e-06, + "loss": 5.8691, + "step": 359 + }, + { + "epoch": 0.02886292116814656, + "grad_norm": 3.1443898677825928, + "learning_rate": 9.612817089452604e-06, + "loss": 5.8541, + "step": 360 + }, + { + "epoch": 0.02894309594916919, + "grad_norm": 3.234187126159668, + "learning_rate": 9.639519359145529e-06, + "loss": 5.8533, + "step": 361 + }, + { + "epoch": 0.029023270730191817, + "grad_norm": 3.1596081256866455, + "learning_rate": 9.666221628838452e-06, + "loss": 5.8533, + "step": 362 + }, + { + "epoch": 0.029103445511214448, + "grad_norm": 3.4751267433166504, + "learning_rate": 9.692923898531376e-06, + "loss": 5.9148, + "step": 363 + }, + { + "epoch": 0.029183620292237075, + "grad_norm": 2.2010979652404785, + "learning_rate": 9.7196261682243e-06, + "loss": 5.7078, + "step": 364 + }, + { + "epoch": 0.029263795073259706, + "grad_norm": 3.0624687671661377, + "learning_rate": 9.746328437917225e-06, + "loss": 5.8875, + "step": 365 + }, + { + "epoch": 0.029343969854282337, + "grad_norm": 3.092122793197632, + "learning_rate": 9.773030707610147e-06, + "loss": 5.8699, + "step": 366 + }, + { + "epoch": 0.029424144635304964, + "grad_norm": 2.0796587467193604, + "learning_rate": 9.799732977303072e-06, + "loss": 5.7107, + "step": 367 + }, + { + "epoch": 0.029504319416327595, + "grad_norm": 2.0502495765686035, + "learning_rate": 9.826435246995995e-06, + "loss": 5.7319, + "step": 368 + }, + { + "epoch": 0.029584494197350222, + "grad_norm": 3.0154435634613037, + "learning_rate": 9.85313751668892e-06, + "loss": 5.8577, + "step": 369 + }, + { + "epoch": 0.029664668978372853, + "grad_norm": 3.3869025707244873, + "learning_rate": 9.879839786381842e-06, + "loss": 5.8351, + "step": 370 + }, + { + "epoch": 0.02974484375939548, + "grad_norm": 2.8004636764526367, + "learning_rate": 9.906542056074768e-06, + "loss": 5.8723, + "step": 371 + }, + { + "epoch": 0.02982501854041811, + "grad_norm": 1.890354037284851, + "learning_rate": 9.933244325767691e-06, + "loss": 5.7129, + "step": 372 + }, + { + "epoch": 0.029905193321440742, + "grad_norm": 2.927586317062378, + "learning_rate": 9.959946595460615e-06, + "loss": 5.8311, + "step": 373 + }, + { + "epoch": 0.02998536810246337, + "grad_norm": 3.1084296703338623, + "learning_rate": 9.986648865153538e-06, + "loss": 5.8729, + "step": 374 + }, + { + "epoch": 0.030065542883486, + "grad_norm": 2.8084921836853027, + "learning_rate": 1.0013351134846463e-05, + "loss": 5.84, + "step": 375 + }, + { + "epoch": 0.030145717664508628, + "grad_norm": 2.6880064010620117, + "learning_rate": 1.0040053404539387e-05, + "loss": 5.8606, + "step": 376 + }, + { + "epoch": 0.03022589244553126, + "grad_norm": 3.119074821472168, + "learning_rate": 1.0066755674232312e-05, + "loss": 5.8872, + "step": 377 + }, + { + "epoch": 0.030306067226553886, + "grad_norm": 3.487644672393799, + "learning_rate": 1.0093457943925234e-05, + "loss": 5.8717, + "step": 378 + }, + { + "epoch": 0.030386242007576517, + "grad_norm": 2.8567705154418945, + "learning_rate": 1.0120160213618158e-05, + "loss": 5.9771, + "step": 379 + }, + { + "epoch": 0.030466416788599147, + "grad_norm": 3.644785165786743, + "learning_rate": 1.0146862483311081e-05, + "loss": 5.938, + "step": 380 + }, + { + "epoch": 0.030546591569621775, + "grad_norm": 3.0018508434295654, + "learning_rate": 1.0173564753004006e-05, + "loss": 5.8047, + "step": 381 + }, + { + "epoch": 0.030626766350644406, + "grad_norm": 3.6069843769073486, + "learning_rate": 1.020026702269693e-05, + "loss": 5.715, + "step": 382 + }, + { + "epoch": 0.030706941131667033, + "grad_norm": 3.8385562896728516, + "learning_rate": 1.0226969292389855e-05, + "loss": 5.9647, + "step": 383 + }, + { + "epoch": 0.030787115912689664, + "grad_norm": 2.984220266342163, + "learning_rate": 1.0253671562082779e-05, + "loss": 5.8123, + "step": 384 + }, + { + "epoch": 0.03086729069371229, + "grad_norm": 3.142799139022827, + "learning_rate": 1.02803738317757e-05, + "loss": 5.8242, + "step": 385 + }, + { + "epoch": 0.030947465474734922, + "grad_norm": 3.4613544940948486, + "learning_rate": 1.0307076101468626e-05, + "loss": 5.7958, + "step": 386 + }, + { + "epoch": 0.031027640255757553, + "grad_norm": 2.875364303588867, + "learning_rate": 1.0333778371161549e-05, + "loss": 5.8516, + "step": 387 + }, + { + "epoch": 0.03110781503678018, + "grad_norm": 3.1782984733581543, + "learning_rate": 1.0360480640854473e-05, + "loss": 5.8766, + "step": 388 + }, + { + "epoch": 0.03118798981780281, + "grad_norm": 3.2158188819885254, + "learning_rate": 1.0387182910547398e-05, + "loss": 5.8273, + "step": 389 + }, + { + "epoch": 0.03126816459882544, + "grad_norm": 2.8391919136047363, + "learning_rate": 1.0413885180240321e-05, + "loss": 5.9104, + "step": 390 + }, + { + "epoch": 0.031348339379848066, + "grad_norm": 2.9867215156555176, + "learning_rate": 1.0440587449933247e-05, + "loss": 5.8098, + "step": 391 + }, + { + "epoch": 0.031428514160870696, + "grad_norm": 3.4908270835876465, + "learning_rate": 1.0467289719626168e-05, + "loss": 5.7443, + "step": 392 + }, + { + "epoch": 0.03150868894189333, + "grad_norm": 1.6793478727340698, + "learning_rate": 1.0493991989319092e-05, + "loss": 5.6897, + "step": 393 + }, + { + "epoch": 0.03158886372291596, + "grad_norm": 2.951779365539551, + "learning_rate": 1.0520694259012017e-05, + "loss": 5.8229, + "step": 394 + }, + { + "epoch": 0.03166903850393859, + "grad_norm": 3.1288816928863525, + "learning_rate": 1.054739652870494e-05, + "loss": 5.8022, + "step": 395 + }, + { + "epoch": 0.03174921328496121, + "grad_norm": 1.642454743385315, + "learning_rate": 1.0574098798397864e-05, + "loss": 5.6821, + "step": 396 + }, + { + "epoch": 0.031829388065983844, + "grad_norm": 4.305239200592041, + "learning_rate": 1.060080106809079e-05, + "loss": 5.7812, + "step": 397 + }, + { + "epoch": 0.031909562847006474, + "grad_norm": 3.7220706939697266, + "learning_rate": 1.0627503337783713e-05, + "loss": 5.8713, + "step": 398 + }, + { + "epoch": 0.031989737628029105, + "grad_norm": 3.4491770267486572, + "learning_rate": 1.0654205607476635e-05, + "loss": 5.8972, + "step": 399 + }, + { + "epoch": 0.032069912409051736, + "grad_norm": 3.2662644386291504, + "learning_rate": 1.068090787716956e-05, + "loss": 5.8953, + "step": 400 + }, + { + "epoch": 0.03215008719007436, + "grad_norm": 2.685154676437378, + "learning_rate": 1.0707610146862484e-05, + "loss": 5.8868, + "step": 401 + }, + { + "epoch": 0.03223026197109699, + "grad_norm": 2.9210684299468994, + "learning_rate": 1.0734312416555409e-05, + "loss": 5.8328, + "step": 402 + }, + { + "epoch": 0.03231043675211962, + "grad_norm": 3.1853251457214355, + "learning_rate": 1.0761014686248332e-05, + "loss": 5.9017, + "step": 403 + }, + { + "epoch": 0.03239061153314225, + "grad_norm": 2.7322890758514404, + "learning_rate": 1.0787716955941256e-05, + "loss": 5.8812, + "step": 404 + }, + { + "epoch": 0.032470786314164876, + "grad_norm": 3.5045242309570312, + "learning_rate": 1.0814419225634181e-05, + "loss": 5.8234, + "step": 405 + }, + { + "epoch": 0.03255096109518751, + "grad_norm": 3.1871867179870605, + "learning_rate": 1.0841121495327103e-05, + "loss": 5.7986, + "step": 406 + }, + { + "epoch": 0.03263113587621014, + "grad_norm": 3.3734397888183594, + "learning_rate": 1.0867823765020026e-05, + "loss": 5.8262, + "step": 407 + }, + { + "epoch": 0.03271131065723277, + "grad_norm": 2.593291759490967, + "learning_rate": 1.0894526034712952e-05, + "loss": 5.8367, + "step": 408 + }, + { + "epoch": 0.0327914854382554, + "grad_norm": 3.121540069580078, + "learning_rate": 1.0921228304405875e-05, + "loss": 5.8997, + "step": 409 + }, + { + "epoch": 0.03287166021927802, + "grad_norm": 3.189849376678467, + "learning_rate": 1.0947930574098799e-05, + "loss": 5.8794, + "step": 410 + }, + { + "epoch": 0.032951835000300654, + "grad_norm": 2.917727470397949, + "learning_rate": 1.0974632843791724e-05, + "loss": 5.837, + "step": 411 + }, + { + "epoch": 0.033032009781323285, + "grad_norm": 3.113450765609741, + "learning_rate": 1.1001335113484647e-05, + "loss": 5.8296, + "step": 412 + }, + { + "epoch": 0.033112184562345916, + "grad_norm": 3.231476306915283, + "learning_rate": 1.102803738317757e-05, + "loss": 5.6744, + "step": 413 + }, + { + "epoch": 0.03319235934336855, + "grad_norm": 3.252725601196289, + "learning_rate": 1.1054739652870495e-05, + "loss": 5.788, + "step": 414 + }, + { + "epoch": 0.03327253412439117, + "grad_norm": 2.677823543548584, + "learning_rate": 1.1081441922563418e-05, + "loss": 5.8589, + "step": 415 + }, + { + "epoch": 0.0333527089054138, + "grad_norm": 2.853236436843872, + "learning_rate": 1.1108144192256343e-05, + "loss": 5.8132, + "step": 416 + }, + { + "epoch": 0.03343288368643643, + "grad_norm": 3.2792272567749023, + "learning_rate": 1.1134846461949267e-05, + "loss": 5.771, + "step": 417 + }, + { + "epoch": 0.03351305846745906, + "grad_norm": 2.9432010650634766, + "learning_rate": 1.116154873164219e-05, + "loss": 5.8062, + "step": 418 + }, + { + "epoch": 0.03359323324848169, + "grad_norm": 2.993941307067871, + "learning_rate": 1.1188251001335116e-05, + "loss": 5.829, + "step": 419 + }, + { + "epoch": 0.03367340802950432, + "grad_norm": 2.753300905227661, + "learning_rate": 1.1214953271028037e-05, + "loss": 5.8444, + "step": 420 + }, + { + "epoch": 0.03375358281052695, + "grad_norm": 3.742130756378174, + "learning_rate": 1.1241655540720961e-05, + "loss": 5.7479, + "step": 421 + }, + { + "epoch": 0.03383375759154958, + "grad_norm": 3.2151618003845215, + "learning_rate": 1.1268357810413886e-05, + "loss": 5.9194, + "step": 422 + }, + { + "epoch": 0.03391393237257221, + "grad_norm": 3.8069663047790527, + "learning_rate": 1.129506008010681e-05, + "loss": 5.775, + "step": 423 + }, + { + "epoch": 0.033994107153594834, + "grad_norm": 3.1353917121887207, + "learning_rate": 1.1321762349799735e-05, + "loss": 5.8093, + "step": 424 + }, + { + "epoch": 0.034074281934617465, + "grad_norm": 1.742499828338623, + "learning_rate": 1.1348464619492658e-05, + "loss": 5.6831, + "step": 425 + }, + { + "epoch": 0.034154456715640095, + "grad_norm": 3.7589921951293945, + "learning_rate": 1.1375166889185582e-05, + "loss": 5.8623, + "step": 426 + }, + { + "epoch": 0.034234631496662726, + "grad_norm": 2.978842258453369, + "learning_rate": 1.1401869158878505e-05, + "loss": 5.871, + "step": 427 + }, + { + "epoch": 0.03431480627768536, + "grad_norm": 3.402665376663208, + "learning_rate": 1.1428571428571429e-05, + "loss": 5.8511, + "step": 428 + }, + { + "epoch": 0.03439498105870798, + "grad_norm": 3.912107467651367, + "learning_rate": 1.1455273698264353e-05, + "loss": 5.8745, + "step": 429 + }, + { + "epoch": 0.03447515583973061, + "grad_norm": 2.616011381149292, + "learning_rate": 1.1481975967957278e-05, + "loss": 5.7941, + "step": 430 + }, + { + "epoch": 0.03455533062075324, + "grad_norm": 2.978306770324707, + "learning_rate": 1.1508678237650201e-05, + "loss": 5.7585, + "step": 431 + }, + { + "epoch": 0.03463550540177587, + "grad_norm": 4.036025047302246, + "learning_rate": 1.1535380507343127e-05, + "loss": 5.8268, + "step": 432 + }, + { + "epoch": 0.0347156801827985, + "grad_norm": 3.3191730976104736, + "learning_rate": 1.156208277703605e-05, + "loss": 5.8163, + "step": 433 + }, + { + "epoch": 0.03479585496382113, + "grad_norm": 3.0583691596984863, + "learning_rate": 1.1588785046728972e-05, + "loss": 5.8048, + "step": 434 + }, + { + "epoch": 0.03487602974484376, + "grad_norm": 3.5711374282836914, + "learning_rate": 1.1615487316421895e-05, + "loss": 5.7694, + "step": 435 + }, + { + "epoch": 0.03495620452586639, + "grad_norm": 3.1512529850006104, + "learning_rate": 1.164218958611482e-05, + "loss": 5.7855, + "step": 436 + }, + { + "epoch": 0.03503637930688902, + "grad_norm": 2.7647790908813477, + "learning_rate": 1.1668891855807744e-05, + "loss": 5.775, + "step": 437 + }, + { + "epoch": 0.035116554087911644, + "grad_norm": 3.3361620903015137, + "learning_rate": 1.169559412550067e-05, + "loss": 5.7871, + "step": 438 + }, + { + "epoch": 0.035196728868934275, + "grad_norm": 3.8849785327911377, + "learning_rate": 1.1722296395193593e-05, + "loss": 5.8777, + "step": 439 + }, + { + "epoch": 0.035276903649956906, + "grad_norm": 2.932619094848633, + "learning_rate": 1.1748998664886518e-05, + "loss": 5.788, + "step": 440 + }, + { + "epoch": 0.03535707843097954, + "grad_norm": 3.3674240112304688, + "learning_rate": 1.177570093457944e-05, + "loss": 5.8298, + "step": 441 + }, + { + "epoch": 0.03543725321200217, + "grad_norm": 2.7395730018615723, + "learning_rate": 1.1802403204272363e-05, + "loss": 5.7606, + "step": 442 + }, + { + "epoch": 0.03551742799302479, + "grad_norm": 3.1544432640075684, + "learning_rate": 1.1829105473965287e-05, + "loss": 5.7745, + "step": 443 + }, + { + "epoch": 0.03559760277404742, + "grad_norm": 3.0222811698913574, + "learning_rate": 1.1855807743658212e-05, + "loss": 5.7861, + "step": 444 + }, + { + "epoch": 0.03567777755507005, + "grad_norm": 2.7550528049468994, + "learning_rate": 1.1882510013351136e-05, + "loss": 5.8774, + "step": 445 + }, + { + "epoch": 0.035757952336092684, + "grad_norm": 1.8239285945892334, + "learning_rate": 1.1909212283044061e-05, + "loss": 5.6633, + "step": 446 + }, + { + "epoch": 0.035838127117115315, + "grad_norm": 2.9397199153900146, + "learning_rate": 1.1935914552736985e-05, + "loss": 5.7924, + "step": 447 + }, + { + "epoch": 0.03591830189813794, + "grad_norm": 3.338512897491455, + "learning_rate": 1.1962616822429906e-05, + "loss": 5.7672, + "step": 448 + }, + { + "epoch": 0.03599847667916057, + "grad_norm": 3.386232376098633, + "learning_rate": 1.1989319092122832e-05, + "loss": 5.8398, + "step": 449 + }, + { + "epoch": 0.0360786514601832, + "grad_norm": 2.602753162384033, + "learning_rate": 1.2016021361815755e-05, + "loss": 5.7942, + "step": 450 + }, + { + "epoch": 0.03615882624120583, + "grad_norm": 2.480952501296997, + "learning_rate": 1.2042723631508679e-05, + "loss": 5.8083, + "step": 451 + }, + { + "epoch": 0.036239001022228455, + "grad_norm": 3.320004940032959, + "learning_rate": 1.2069425901201604e-05, + "loss": 5.7515, + "step": 452 + }, + { + "epoch": 0.036319175803251086, + "grad_norm": 3.420079231262207, + "learning_rate": 1.2096128170894527e-05, + "loss": 5.7893, + "step": 453 + }, + { + "epoch": 0.03639935058427372, + "grad_norm": 3.4026596546173096, + "learning_rate": 1.2122830440587453e-05, + "loss": 5.738, + "step": 454 + }, + { + "epoch": 0.03647952536529635, + "grad_norm": 2.8644778728485107, + "learning_rate": 1.2149532710280374e-05, + "loss": 5.8125, + "step": 455 + }, + { + "epoch": 0.03655970014631898, + "grad_norm": 2.7557766437530518, + "learning_rate": 1.2176234979973298e-05, + "loss": 5.845, + "step": 456 + }, + { + "epoch": 0.0366398749273416, + "grad_norm": 3.074115037918091, + "learning_rate": 1.2202937249666223e-05, + "loss": 5.772, + "step": 457 + }, + { + "epoch": 0.03672004970836423, + "grad_norm": 2.6061112880706787, + "learning_rate": 1.2229639519359147e-05, + "loss": 5.8229, + "step": 458 + }, + { + "epoch": 0.036800224489386864, + "grad_norm": 2.9774856567382812, + "learning_rate": 1.225634178905207e-05, + "loss": 5.7611, + "step": 459 + }, + { + "epoch": 0.036880399270409495, + "grad_norm": 2.614894390106201, + "learning_rate": 1.2283044058744995e-05, + "loss": 5.8041, + "step": 460 + }, + { + "epoch": 0.036960574051432125, + "grad_norm": 3.1521143913269043, + "learning_rate": 1.2309746328437919e-05, + "loss": 5.7771, + "step": 461 + }, + { + "epoch": 0.03704074883245475, + "grad_norm": 3.2416470050811768, + "learning_rate": 1.233644859813084e-05, + "loss": 5.7329, + "step": 462 + }, + { + "epoch": 0.03712092361347738, + "grad_norm": 3.0785107612609863, + "learning_rate": 1.2363150867823766e-05, + "loss": 5.7903, + "step": 463 + }, + { + "epoch": 0.03720109839450001, + "grad_norm": 3.0555460453033447, + "learning_rate": 1.238985313751669e-05, + "loss": 5.8188, + "step": 464 + }, + { + "epoch": 0.03728127317552264, + "grad_norm": 3.6419575214385986, + "learning_rate": 1.2416555407209615e-05, + "loss": 5.7523, + "step": 465 + }, + { + "epoch": 0.037361447956545266, + "grad_norm": 3.1641368865966797, + "learning_rate": 1.2443257676902538e-05, + "loss": 5.7955, + "step": 466 + }, + { + "epoch": 0.037441622737567896, + "grad_norm": 1.962507724761963, + "learning_rate": 1.2469959946595462e-05, + "loss": 5.6509, + "step": 467 + }, + { + "epoch": 0.03752179751859053, + "grad_norm": 4.6689839363098145, + "learning_rate": 1.2496662216288387e-05, + "loss": 5.8453, + "step": 468 + }, + { + "epoch": 0.03760197229961316, + "grad_norm": 3.00679874420166, + "learning_rate": 1.2523364485981309e-05, + "loss": 5.6974, + "step": 469 + }, + { + "epoch": 0.03768214708063579, + "grad_norm": 3.4395318031311035, + "learning_rate": 1.2550066755674232e-05, + "loss": 5.7556, + "step": 470 + }, + { + "epoch": 0.03776232186165841, + "grad_norm": 3.075054883956909, + "learning_rate": 1.2576769025367158e-05, + "loss": 5.7698, + "step": 471 + }, + { + "epoch": 0.037842496642681044, + "grad_norm": 4.284217357635498, + "learning_rate": 1.2603471295060081e-05, + "loss": 5.7427, + "step": 472 + }, + { + "epoch": 0.037922671423703674, + "grad_norm": 3.045259475708008, + "learning_rate": 1.2630173564753005e-05, + "loss": 5.6313, + "step": 473 + }, + { + "epoch": 0.038002846204726305, + "grad_norm": 2.7683544158935547, + "learning_rate": 1.265687583444593e-05, + "loss": 5.7294, + "step": 474 + }, + { + "epoch": 0.038083020985748936, + "grad_norm": 2.697852849960327, + "learning_rate": 1.2683578104138853e-05, + "loss": 5.7313, + "step": 475 + }, + { + "epoch": 0.03816319576677156, + "grad_norm": 2.918475389480591, + "learning_rate": 1.2710280373831775e-05, + "loss": 5.778, + "step": 476 + }, + { + "epoch": 0.03824337054779419, + "grad_norm": 3.687934637069702, + "learning_rate": 1.27369826435247e-05, + "loss": 5.6931, + "step": 477 + }, + { + "epoch": 0.03832354532881682, + "grad_norm": 3.0111849308013916, + "learning_rate": 1.2763684913217624e-05, + "loss": 5.7678, + "step": 478 + }, + { + "epoch": 0.03840372010983945, + "grad_norm": 2.9193224906921387, + "learning_rate": 1.279038718291055e-05, + "loss": 5.6724, + "step": 479 + }, + { + "epoch": 0.038483894890862076, + "grad_norm": 2.777465581893921, + "learning_rate": 1.2817089452603473e-05, + "loss": 5.755, + "step": 480 + }, + { + "epoch": 0.03856406967188471, + "grad_norm": 2.499058723449707, + "learning_rate": 1.2843791722296396e-05, + "loss": 5.7126, + "step": 481 + }, + { + "epoch": 0.03864424445290734, + "grad_norm": 3.237797975540161, + "learning_rate": 1.287049399198932e-05, + "loss": 5.7903, + "step": 482 + }, + { + "epoch": 0.03872441923392997, + "grad_norm": 2.8278467655181885, + "learning_rate": 1.2897196261682243e-05, + "loss": 5.7728, + "step": 483 + }, + { + "epoch": 0.0388045940149526, + "grad_norm": 3.0044004917144775, + "learning_rate": 1.2923898531375167e-05, + "loss": 5.8539, + "step": 484 + }, + { + "epoch": 0.03888476879597522, + "grad_norm": 3.3587164878845215, + "learning_rate": 1.2950600801068092e-05, + "loss": 5.7544, + "step": 485 + }, + { + "epoch": 0.038964943576997854, + "grad_norm": 2.8966126441955566, + "learning_rate": 1.2977303070761016e-05, + "loss": 5.8099, + "step": 486 + }, + { + "epoch": 0.039045118358020485, + "grad_norm": 3.549325466156006, + "learning_rate": 1.3004005340453941e-05, + "loss": 5.8795, + "step": 487 + }, + { + "epoch": 0.039125293139043116, + "grad_norm": 2.8218977451324463, + "learning_rate": 1.3030707610146864e-05, + "loss": 5.9106, + "step": 488 + }, + { + "epoch": 0.03920546792006575, + "grad_norm": 2.841693639755249, + "learning_rate": 1.3057409879839786e-05, + "loss": 5.7531, + "step": 489 + }, + { + "epoch": 0.03928564270108837, + "grad_norm": 2.7372121810913086, + "learning_rate": 1.308411214953271e-05, + "loss": 5.7405, + "step": 490 + }, + { + "epoch": 0.039365817482111, + "grad_norm": 3.6255762577056885, + "learning_rate": 1.3110814419225635e-05, + "loss": 5.7554, + "step": 491 + }, + { + "epoch": 0.03944599226313363, + "grad_norm": 3.3681294918060303, + "learning_rate": 1.3137516688918559e-05, + "loss": 5.7832, + "step": 492 + }, + { + "epoch": 0.03952616704415626, + "grad_norm": 3.041626214981079, + "learning_rate": 1.3164218958611484e-05, + "loss": 5.7531, + "step": 493 + }, + { + "epoch": 0.03960634182517889, + "grad_norm": 1.9795944690704346, + "learning_rate": 1.3190921228304407e-05, + "loss": 5.6213, + "step": 494 + }, + { + "epoch": 0.03968651660620152, + "grad_norm": 3.443002939224243, + "learning_rate": 1.3217623497997332e-05, + "loss": 5.7274, + "step": 495 + }, + { + "epoch": 0.03976669138722415, + "grad_norm": 2.669070243835449, + "learning_rate": 1.3244325767690254e-05, + "loss": 5.803, + "step": 496 + }, + { + "epoch": 0.03984686616824678, + "grad_norm": 3.5843746662139893, + "learning_rate": 1.3271028037383178e-05, + "loss": 5.6417, + "step": 497 + }, + { + "epoch": 0.03992704094926941, + "grad_norm": 3.2044341564178467, + "learning_rate": 1.3297730307076101e-05, + "loss": 5.774, + "step": 498 + }, + { + "epoch": 0.040007215730292034, + "grad_norm": 3.3464596271514893, + "learning_rate": 1.3324432576769027e-05, + "loss": 5.7065, + "step": 499 + }, + { + "epoch": 0.040087390511314665, + "grad_norm": 2.9001669883728027, + "learning_rate": 1.335113484646195e-05, + "loss": 5.9259, + "step": 500 + }, + { + "epoch": 0.040167565292337296, + "grad_norm": 3.134402275085449, + "learning_rate": 1.3377837116154875e-05, + "loss": 5.7413, + "step": 501 + }, + { + "epoch": 0.040247740073359926, + "grad_norm": 3.2592763900756836, + "learning_rate": 1.3404539385847799e-05, + "loss": 5.7871, + "step": 502 + }, + { + "epoch": 0.04032791485438256, + "grad_norm": 3.3365025520324707, + "learning_rate": 1.343124165554072e-05, + "loss": 5.8015, + "step": 503 + }, + { + "epoch": 0.04040808963540518, + "grad_norm": 2.5674452781677246, + "learning_rate": 1.3457943925233646e-05, + "loss": 5.7072, + "step": 504 + }, + { + "epoch": 0.04048826441642781, + "grad_norm": 3.186352014541626, + "learning_rate": 1.348464619492657e-05, + "loss": 5.6763, + "step": 505 + }, + { + "epoch": 0.04056843919745044, + "grad_norm": 3.2650845050811768, + "learning_rate": 1.3511348464619493e-05, + "loss": 5.7151, + "step": 506 + }, + { + "epoch": 0.040648613978473074, + "grad_norm": 1.9040826559066772, + "learning_rate": 1.3538050734312418e-05, + "loss": 5.624, + "step": 507 + }, + { + "epoch": 0.0407287887594957, + "grad_norm": 2.431427001953125, + "learning_rate": 1.3564753004005342e-05, + "loss": 5.8214, + "step": 508 + }, + { + "epoch": 0.04080896354051833, + "grad_norm": 2.9667649269104004, + "learning_rate": 1.3591455273698267e-05, + "loss": 5.8171, + "step": 509 + }, + { + "epoch": 0.04088913832154096, + "grad_norm": 2.8074851036071777, + "learning_rate": 1.3618157543391189e-05, + "loss": 5.7421, + "step": 510 + }, + { + "epoch": 0.04096931310256359, + "grad_norm": 1.6398851871490479, + "learning_rate": 1.3644859813084112e-05, + "loss": 5.6104, + "step": 511 + }, + { + "epoch": 0.04104948788358622, + "grad_norm": 3.299060344696045, + "learning_rate": 1.3671562082777038e-05, + "loss": 5.6616, + "step": 512 + }, + { + "epoch": 0.041129662664608845, + "grad_norm": 1.4230211973190308, + "learning_rate": 1.3698264352469961e-05, + "loss": 5.623, + "step": 513 + }, + { + "epoch": 0.041209837445631475, + "grad_norm": 3.2304651737213135, + "learning_rate": 1.3724966622162885e-05, + "loss": 5.6586, + "step": 514 + }, + { + "epoch": 0.041290012226654106, + "grad_norm": 3.2831902503967285, + "learning_rate": 1.375166889185581e-05, + "loss": 5.749, + "step": 515 + }, + { + "epoch": 0.04137018700767674, + "grad_norm": 2.517174482345581, + "learning_rate": 1.3778371161548733e-05, + "loss": 5.7467, + "step": 516 + }, + { + "epoch": 0.04145036178869937, + "grad_norm": 1.3982349634170532, + "learning_rate": 1.3805073431241655e-05, + "loss": 5.6104, + "step": 517 + }, + { + "epoch": 0.04153053656972199, + "grad_norm": 4.0974249839782715, + "learning_rate": 1.383177570093458e-05, + "loss": 5.7611, + "step": 518 + }, + { + "epoch": 0.04161071135074462, + "grad_norm": 2.9841859340667725, + "learning_rate": 1.3858477970627504e-05, + "loss": 5.7687, + "step": 519 + }, + { + "epoch": 0.04169088613176725, + "grad_norm": 3.30611515045166, + "learning_rate": 1.3885180240320429e-05, + "loss": 5.6864, + "step": 520 + }, + { + "epoch": 0.041771060912789884, + "grad_norm": 3.004779100418091, + "learning_rate": 1.3911882510013353e-05, + "loss": 5.7551, + "step": 521 + }, + { + "epoch": 0.04185123569381251, + "grad_norm": 3.2205677032470703, + "learning_rate": 1.3938584779706276e-05, + "loss": 5.7102, + "step": 522 + }, + { + "epoch": 0.04193141047483514, + "grad_norm": 2.9509010314941406, + "learning_rate": 1.3965287049399201e-05, + "loss": 5.8434, + "step": 523 + }, + { + "epoch": 0.04201158525585777, + "grad_norm": 3.1770517826080322, + "learning_rate": 1.3991989319092123e-05, + "loss": 5.7429, + "step": 524 + }, + { + "epoch": 0.0420917600368804, + "grad_norm": 3.583570718765259, + "learning_rate": 1.4018691588785047e-05, + "loss": 5.7012, + "step": 525 + }, + { + "epoch": 0.04217193481790303, + "grad_norm": 2.6798760890960693, + "learning_rate": 1.4045393858477972e-05, + "loss": 5.7736, + "step": 526 + }, + { + "epoch": 0.042252109598925655, + "grad_norm": 3.168966770172119, + "learning_rate": 1.4072096128170896e-05, + "loss": 5.8883, + "step": 527 + }, + { + "epoch": 0.042332284379948286, + "grad_norm": 3.566660165786743, + "learning_rate": 1.4098798397863819e-05, + "loss": 5.7666, + "step": 528 + }, + { + "epoch": 0.04241245916097092, + "grad_norm": 2.50288462638855, + "learning_rate": 1.4125500667556744e-05, + "loss": 5.6674, + "step": 529 + }, + { + "epoch": 0.04249263394199355, + "grad_norm": 2.677396774291992, + "learning_rate": 1.4152202937249668e-05, + "loss": 5.7971, + "step": 530 + }, + { + "epoch": 0.04257280872301618, + "grad_norm": 2.6807825565338135, + "learning_rate": 1.417890520694259e-05, + "loss": 5.9165, + "step": 531 + }, + { + "epoch": 0.0426529835040388, + "grad_norm": 3.1418497562408447, + "learning_rate": 1.4205607476635515e-05, + "loss": 5.8104, + "step": 532 + }, + { + "epoch": 0.04273315828506143, + "grad_norm": 2.5701024532318115, + "learning_rate": 1.4232309746328438e-05, + "loss": 5.761, + "step": 533 + }, + { + "epoch": 0.042813333066084064, + "grad_norm": 2.8103854656219482, + "learning_rate": 1.4259012016021364e-05, + "loss": 5.7191, + "step": 534 + }, + { + "epoch": 0.042893507847106695, + "grad_norm": 3.075731039047241, + "learning_rate": 1.4285714285714287e-05, + "loss": 5.7206, + "step": 535 + }, + { + "epoch": 0.04297368262812932, + "grad_norm": 3.088728666305542, + "learning_rate": 1.431241655540721e-05, + "loss": 5.6754, + "step": 536 + }, + { + "epoch": 0.04305385740915195, + "grad_norm": 2.8524487018585205, + "learning_rate": 1.4339118825100136e-05, + "loss": 5.6731, + "step": 537 + }, + { + "epoch": 0.04313403219017458, + "grad_norm": 3.838242769241333, + "learning_rate": 1.4365821094793058e-05, + "loss": 5.6796, + "step": 538 + }, + { + "epoch": 0.04321420697119721, + "grad_norm": 2.852478504180908, + "learning_rate": 1.4392523364485981e-05, + "loss": 5.8364, + "step": 539 + }, + { + "epoch": 0.04329438175221984, + "grad_norm": 2.698676109313965, + "learning_rate": 1.4419225634178906e-05, + "loss": 5.8484, + "step": 540 + }, + { + "epoch": 0.043374556533242466, + "grad_norm": 3.683225631713867, + "learning_rate": 1.444592790387183e-05, + "loss": 5.7229, + "step": 541 + }, + { + "epoch": 0.043454731314265096, + "grad_norm": 3.5502076148986816, + "learning_rate": 1.4472630173564755e-05, + "loss": 5.7216, + "step": 542 + }, + { + "epoch": 0.04353490609528773, + "grad_norm": 3.5335068702697754, + "learning_rate": 1.4499332443257679e-05, + "loss": 5.7952, + "step": 543 + }, + { + "epoch": 0.04361508087631036, + "grad_norm": 3.8212661743164062, + "learning_rate": 1.4526034712950602e-05, + "loss": 5.7891, + "step": 544 + }, + { + "epoch": 0.04369525565733299, + "grad_norm": 3.096299886703491, + "learning_rate": 1.4552736982643526e-05, + "loss": 5.7333, + "step": 545 + }, + { + "epoch": 0.04377543043835561, + "grad_norm": 3.0276641845703125, + "learning_rate": 1.457943925233645e-05, + "loss": 5.7434, + "step": 546 + }, + { + "epoch": 0.043855605219378244, + "grad_norm": 1.5645911693572998, + "learning_rate": 1.4606141522029373e-05, + "loss": 5.5662, + "step": 547 + }, + { + "epoch": 0.043935780000400874, + "grad_norm": 3.1979928016662598, + "learning_rate": 1.4632843791722298e-05, + "loss": 5.8456, + "step": 548 + }, + { + "epoch": 0.044015954781423505, + "grad_norm": 2.8296797275543213, + "learning_rate": 1.4659546061415222e-05, + "loss": 5.7776, + "step": 549 + }, + { + "epoch": 0.04409612956244613, + "grad_norm": 1.4147030115127563, + "learning_rate": 1.4686248331108147e-05, + "loss": 5.5837, + "step": 550 + }, + { + "epoch": 0.04417630434346876, + "grad_norm": 3.264031171798706, + "learning_rate": 1.471295060080107e-05, + "loss": 5.6664, + "step": 551 + }, + { + "epoch": 0.04425647912449139, + "grad_norm": 2.677772283554077, + "learning_rate": 1.4739652870493992e-05, + "loss": 5.7369, + "step": 552 + }, + { + "epoch": 0.04433665390551402, + "grad_norm": 3.0715131759643555, + "learning_rate": 1.4766355140186916e-05, + "loss": 5.6743, + "step": 553 + }, + { + "epoch": 0.04441682868653665, + "grad_norm": 2.502671003341675, + "learning_rate": 1.4793057409879841e-05, + "loss": 5.8362, + "step": 554 + }, + { + "epoch": 0.044497003467559276, + "grad_norm": 3.1661720275878906, + "learning_rate": 1.4819759679572764e-05, + "loss": 5.834, + "step": 555 + }, + { + "epoch": 0.04457717824858191, + "grad_norm": 3.3503644466400146, + "learning_rate": 1.484646194926569e-05, + "loss": 5.6837, + "step": 556 + }, + { + "epoch": 0.04465735302960454, + "grad_norm": 2.6028685569763184, + "learning_rate": 1.4873164218958613e-05, + "loss": 5.7771, + "step": 557 + }, + { + "epoch": 0.04473752781062717, + "grad_norm": 2.9700989723205566, + "learning_rate": 1.4899866488651538e-05, + "loss": 5.6896, + "step": 558 + }, + { + "epoch": 0.0448177025916498, + "grad_norm": 2.532801866531372, + "learning_rate": 1.492656875834446e-05, + "loss": 5.6583, + "step": 559 + }, + { + "epoch": 0.04489787737267242, + "grad_norm": 1.7509477138519287, + "learning_rate": 1.4953271028037384e-05, + "loss": 5.5813, + "step": 560 + }, + { + "epoch": 0.044978052153695054, + "grad_norm": 3.442607879638672, + "learning_rate": 1.4979973297730307e-05, + "loss": 5.7908, + "step": 561 + }, + { + "epoch": 0.045058226934717685, + "grad_norm": 3.2586305141448975, + "learning_rate": 1.5006675567423233e-05, + "loss": 5.7633, + "step": 562 + }, + { + "epoch": 0.045138401715740316, + "grad_norm": 2.8028454780578613, + "learning_rate": 1.5033377837116156e-05, + "loss": 5.7549, + "step": 563 + }, + { + "epoch": 0.04521857649676295, + "grad_norm": 3.252156972885132, + "learning_rate": 1.5060080106809081e-05, + "loss": 5.6794, + "step": 564 + }, + { + "epoch": 0.04529875127778557, + "grad_norm": 2.98047137260437, + "learning_rate": 1.5086782376502005e-05, + "loss": 5.6025, + "step": 565 + }, + { + "epoch": 0.0453789260588082, + "grad_norm": 2.563927412033081, + "learning_rate": 1.5113484646194927e-05, + "loss": 5.7935, + "step": 566 + }, + { + "epoch": 0.04545910083983083, + "grad_norm": 2.993072032928467, + "learning_rate": 1.5140186915887852e-05, + "loss": 5.8042, + "step": 567 + }, + { + "epoch": 0.04553927562085346, + "grad_norm": 3.166377305984497, + "learning_rate": 1.5166889185580775e-05, + "loss": 5.5797, + "step": 568 + }, + { + "epoch": 0.04561945040187609, + "grad_norm": 2.621757745742798, + "learning_rate": 1.5193591455273699e-05, + "loss": 5.6403, + "step": 569 + }, + { + "epoch": 0.04569962518289872, + "grad_norm": 3.165621519088745, + "learning_rate": 1.5220293724966624e-05, + "loss": 5.675, + "step": 570 + }, + { + "epoch": 0.04577979996392135, + "grad_norm": 2.857788562774658, + "learning_rate": 1.5246995994659548e-05, + "loss": 5.6987, + "step": 571 + }, + { + "epoch": 0.04585997474494398, + "grad_norm": 3.2950475215911865, + "learning_rate": 1.527369826435247e-05, + "loss": 5.6632, + "step": 572 + }, + { + "epoch": 0.04594014952596661, + "grad_norm": 2.436385154724121, + "learning_rate": 1.5300400534045393e-05, + "loss": 5.7157, + "step": 573 + }, + { + "epoch": 0.046020324306989234, + "grad_norm": 3.5178961753845215, + "learning_rate": 1.5327102803738318e-05, + "loss": 5.6866, + "step": 574 + }, + { + "epoch": 0.046100499088011865, + "grad_norm": 2.791740655899048, + "learning_rate": 1.5353805073431243e-05, + "loss": 5.7679, + "step": 575 + }, + { + "epoch": 0.046180673869034496, + "grad_norm": 3.624936580657959, + "learning_rate": 1.5380507343124165e-05, + "loss": 5.7389, + "step": 576 + }, + { + "epoch": 0.046260848650057126, + "grad_norm": 2.793405294418335, + "learning_rate": 1.540720961281709e-05, + "loss": 5.7979, + "step": 577 + }, + { + "epoch": 0.04634102343107976, + "grad_norm": 2.672008752822876, + "learning_rate": 1.5433911882510016e-05, + "loss": 5.7943, + "step": 578 + }, + { + "epoch": 0.04642119821210238, + "grad_norm": 3.1682028770446777, + "learning_rate": 1.546061415220294e-05, + "loss": 5.6614, + "step": 579 + }, + { + "epoch": 0.04650137299312501, + "grad_norm": 3.0137650966644287, + "learning_rate": 1.5487316421895863e-05, + "loss": 5.6464, + "step": 580 + }, + { + "epoch": 0.04658154777414764, + "grad_norm": 3.9231832027435303, + "learning_rate": 1.5514018691588785e-05, + "loss": 5.6338, + "step": 581 + }, + { + "epoch": 0.046661722555170274, + "grad_norm": 3.0449907779693604, + "learning_rate": 1.554072096128171e-05, + "loss": 5.6365, + "step": 582 + }, + { + "epoch": 0.0467418973361929, + "grad_norm": 2.8056867122650146, + "learning_rate": 1.5567423230974635e-05, + "loss": 5.713, + "step": 583 + }, + { + "epoch": 0.04682207211721553, + "grad_norm": 3.084036111831665, + "learning_rate": 1.5594125500667557e-05, + "loss": 5.7185, + "step": 584 + }, + { + "epoch": 0.04690224689823816, + "grad_norm": 2.721025228500366, + "learning_rate": 1.5620827770360482e-05, + "loss": 5.7417, + "step": 585 + }, + { + "epoch": 0.04698242167926079, + "grad_norm": 1.5305131673812866, + "learning_rate": 1.5647530040053407e-05, + "loss": 5.5557, + "step": 586 + }, + { + "epoch": 0.04706259646028342, + "grad_norm": 3.3642804622650146, + "learning_rate": 1.567423230974633e-05, + "loss": 5.6589, + "step": 587 + }, + { + "epoch": 0.047142771241306045, + "grad_norm": 2.9822998046875, + "learning_rate": 1.5700934579439254e-05, + "loss": 5.7728, + "step": 588 + }, + { + "epoch": 0.047222946022328675, + "grad_norm": 3.126072406768799, + "learning_rate": 1.5727636849132176e-05, + "loss": 5.7007, + "step": 589 + }, + { + "epoch": 0.047303120803351306, + "grad_norm": 3.2527811527252197, + "learning_rate": 1.57543391188251e-05, + "loss": 5.6494, + "step": 590 + }, + { + "epoch": 0.04738329558437394, + "grad_norm": 2.9025869369506836, + "learning_rate": 1.5781041388518027e-05, + "loss": 5.6661, + "step": 591 + }, + { + "epoch": 0.04746347036539657, + "grad_norm": 3.0921216011047363, + "learning_rate": 1.580774365821095e-05, + "loss": 5.6837, + "step": 592 + }, + { + "epoch": 0.04754364514641919, + "grad_norm": 3.6948015689849854, + "learning_rate": 1.583444592790387e-05, + "loss": 5.5767, + "step": 593 + }, + { + "epoch": 0.04762381992744182, + "grad_norm": 2.737215042114258, + "learning_rate": 1.5861148197596796e-05, + "loss": 5.7362, + "step": 594 + }, + { + "epoch": 0.04770399470846445, + "grad_norm": 3.631638526916504, + "learning_rate": 1.588785046728972e-05, + "loss": 5.73, + "step": 595 + }, + { + "epoch": 0.047784169489487084, + "grad_norm": 3.1406736373901367, + "learning_rate": 1.5914552736982646e-05, + "loss": 5.7194, + "step": 596 + }, + { + "epoch": 0.04786434427050971, + "grad_norm": 3.0118014812469482, + "learning_rate": 1.5941255006675568e-05, + "loss": 5.6229, + "step": 597 + }, + { + "epoch": 0.04794451905153234, + "grad_norm": 2.8122332096099854, + "learning_rate": 1.5967957276368493e-05, + "loss": 5.7261, + "step": 598 + }, + { + "epoch": 0.04802469383255497, + "grad_norm": 2.6583218574523926, + "learning_rate": 1.5994659546061418e-05, + "loss": 5.8064, + "step": 599 + }, + { + "epoch": 0.0481048686135776, + "grad_norm": 2.80397629737854, + "learning_rate": 1.602136181575434e-05, + "loss": 5.8204, + "step": 600 + }, + { + "epoch": 0.04818504339460023, + "grad_norm": 2.94516658782959, + "learning_rate": 1.6048064085447262e-05, + "loss": 5.6216, + "step": 601 + }, + { + "epoch": 0.048265218175622855, + "grad_norm": 1.4511168003082275, + "learning_rate": 1.6074766355140187e-05, + "loss": 5.584, + "step": 602 + }, + { + "epoch": 0.048345392956645486, + "grad_norm": 2.796532392501831, + "learning_rate": 1.6101468624833112e-05, + "loss": 5.7455, + "step": 603 + }, + { + "epoch": 0.04842556773766812, + "grad_norm": 2.8301656246185303, + "learning_rate": 1.6128170894526038e-05, + "loss": 5.623, + "step": 604 + }, + { + "epoch": 0.04850574251869075, + "grad_norm": 2.646819591522217, + "learning_rate": 1.615487316421896e-05, + "loss": 5.6578, + "step": 605 + }, + { + "epoch": 0.04858591729971338, + "grad_norm": 2.624462842941284, + "learning_rate": 1.6181575433911885e-05, + "loss": 5.6963, + "step": 606 + }, + { + "epoch": 0.048666092080736, + "grad_norm": 3.043550968170166, + "learning_rate": 1.6208277703604807e-05, + "loss": 5.6732, + "step": 607 + }, + { + "epoch": 0.04874626686175863, + "grad_norm": 2.6910269260406494, + "learning_rate": 1.6234979973297732e-05, + "loss": 5.7618, + "step": 608 + }, + { + "epoch": 0.048826441642781264, + "grad_norm": 3.631330966949463, + "learning_rate": 1.6261682242990654e-05, + "loss": 5.6622, + "step": 609 + }, + { + "epoch": 0.048906616423803895, + "grad_norm": 2.650761365890503, + "learning_rate": 1.628838451268358e-05, + "loss": 5.6912, + "step": 610 + }, + { + "epoch": 0.04898679120482652, + "grad_norm": 2.744473695755005, + "learning_rate": 1.6315086782376504e-05, + "loss": 5.6816, + "step": 611 + }, + { + "epoch": 0.04906696598584915, + "grad_norm": 3.281635284423828, + "learning_rate": 1.634178905206943e-05, + "loss": 5.6591, + "step": 612 + }, + { + "epoch": 0.04914714076687178, + "grad_norm": 2.604524850845337, + "learning_rate": 1.636849132176235e-05, + "loss": 5.6614, + "step": 613 + }, + { + "epoch": 0.04922731554789441, + "grad_norm": 2.8976547718048096, + "learning_rate": 1.6395193591455273e-05, + "loss": 5.6808, + "step": 614 + }, + { + "epoch": 0.04930749032891704, + "grad_norm": 3.599109411239624, + "learning_rate": 1.6421895861148198e-05, + "loss": 5.7207, + "step": 615 + }, + { + "epoch": 0.049387665109939666, + "grad_norm": 3.6181578636169434, + "learning_rate": 1.6448598130841123e-05, + "loss": 5.6385, + "step": 616 + }, + { + "epoch": 0.0494678398909623, + "grad_norm": 3.349648952484131, + "learning_rate": 1.6475300400534045e-05, + "loss": 5.5836, + "step": 617 + }, + { + "epoch": 0.04954801467198493, + "grad_norm": 3.618941068649292, + "learning_rate": 1.650200267022697e-05, + "loss": 5.7096, + "step": 618 + }, + { + "epoch": 0.04962818945300756, + "grad_norm": 3.0044476985931396, + "learning_rate": 1.6528704939919896e-05, + "loss": 5.6582, + "step": 619 + }, + { + "epoch": 0.04970836423403019, + "grad_norm": 3.0283126831054688, + "learning_rate": 1.655540720961282e-05, + "loss": 5.688, + "step": 620 + }, + { + "epoch": 0.04978853901505281, + "grad_norm": 2.6721031665802, + "learning_rate": 1.6582109479305743e-05, + "loss": 5.7151, + "step": 621 + }, + { + "epoch": 0.049868713796075444, + "grad_norm": 3.2523550987243652, + "learning_rate": 1.6608811748998665e-05, + "loss": 5.5447, + "step": 622 + }, + { + "epoch": 0.049948888577098075, + "grad_norm": 3.1374096870422363, + "learning_rate": 1.663551401869159e-05, + "loss": 5.5812, + "step": 623 + }, + { + "epoch": 0.050029063358120705, + "grad_norm": 1.5406527519226074, + "learning_rate": 1.6662216288384515e-05, + "loss": 5.562, + "step": 624 + }, + { + "epoch": 0.05010923813914333, + "grad_norm": 4.722993850708008, + "learning_rate": 1.6688918558077437e-05, + "loss": 5.7384, + "step": 625 + }, + { + "epoch": 0.05018941292016596, + "grad_norm": 2.700014352798462, + "learning_rate": 1.6715620827770362e-05, + "loss": 5.7323, + "step": 626 + }, + { + "epoch": 0.05026958770118859, + "grad_norm": 4.025119304656982, + "learning_rate": 1.6742323097463287e-05, + "loss": 5.7916, + "step": 627 + }, + { + "epoch": 0.05034976248221122, + "grad_norm": 1.9101437330245972, + "learning_rate": 1.676902536715621e-05, + "loss": 5.5564, + "step": 628 + }, + { + "epoch": 0.05042993726323385, + "grad_norm": 2.9118125438690186, + "learning_rate": 1.6795727636849134e-05, + "loss": 5.7229, + "step": 629 + }, + { + "epoch": 0.050510112044256476, + "grad_norm": 3.8520166873931885, + "learning_rate": 1.6822429906542056e-05, + "loss": 5.6951, + "step": 630 + }, + { + "epoch": 0.05059028682527911, + "grad_norm": 3.017000913619995, + "learning_rate": 1.684913217623498e-05, + "loss": 5.7401, + "step": 631 + }, + { + "epoch": 0.05067046160630174, + "grad_norm": 2.7051327228546143, + "learning_rate": 1.6875834445927907e-05, + "loss": 5.616, + "step": 632 + }, + { + "epoch": 0.05075063638732437, + "grad_norm": 3.2354350090026855, + "learning_rate": 1.690253671562083e-05, + "loss": 5.6835, + "step": 633 + }, + { + "epoch": 0.050830811168347, + "grad_norm": 3.069514751434326, + "learning_rate": 1.6929238985313754e-05, + "loss": 5.6433, + "step": 634 + }, + { + "epoch": 0.05091098594936962, + "grad_norm": 3.2677724361419678, + "learning_rate": 1.6955941255006675e-05, + "loss": 5.7504, + "step": 635 + }, + { + "epoch": 0.050991160730392254, + "grad_norm": 2.343015432357788, + "learning_rate": 1.69826435246996e-05, + "loss": 5.6259, + "step": 636 + }, + { + "epoch": 0.051071335511414885, + "grad_norm": 3.1430344581604004, + "learning_rate": 1.7009345794392526e-05, + "loss": 5.5713, + "step": 637 + }, + { + "epoch": 0.051151510292437516, + "grad_norm": 3.1716904640197754, + "learning_rate": 1.7036048064085448e-05, + "loss": 5.7402, + "step": 638 + }, + { + "epoch": 0.05123168507346014, + "grad_norm": 3.031870126724243, + "learning_rate": 1.7062750333778373e-05, + "loss": 5.6171, + "step": 639 + }, + { + "epoch": 0.05131185985448277, + "grad_norm": 2.9386379718780518, + "learning_rate": 1.7089452603471298e-05, + "loss": 5.6558, + "step": 640 + }, + { + "epoch": 0.0513920346355054, + "grad_norm": 1.665239691734314, + "learning_rate": 1.711615487316422e-05, + "loss": 5.5237, + "step": 641 + }, + { + "epoch": 0.05147220941652803, + "grad_norm": 3.5329062938690186, + "learning_rate": 1.7142857142857142e-05, + "loss": 5.7373, + "step": 642 + }, + { + "epoch": 0.05155238419755066, + "grad_norm": 3.0455572605133057, + "learning_rate": 1.7169559412550067e-05, + "loss": 5.7229, + "step": 643 + }, + { + "epoch": 0.05163255897857329, + "grad_norm": 2.7170004844665527, + "learning_rate": 1.7196261682242992e-05, + "loss": 5.542, + "step": 644 + }, + { + "epoch": 0.05171273375959592, + "grad_norm": 1.5600533485412598, + "learning_rate": 1.7222963951935918e-05, + "loss": 5.5354, + "step": 645 + }, + { + "epoch": 0.05179290854061855, + "grad_norm": 3.5648581981658936, + "learning_rate": 1.724966622162884e-05, + "loss": 5.6437, + "step": 646 + }, + { + "epoch": 0.05187308332164118, + "grad_norm": 1.4193692207336426, + "learning_rate": 1.7276368491321765e-05, + "loss": 5.5308, + "step": 647 + }, + { + "epoch": 0.05195325810266381, + "grad_norm": 1.4635350704193115, + "learning_rate": 1.730307076101469e-05, + "loss": 5.5054, + "step": 648 + }, + { + "epoch": 0.052033432883686434, + "grad_norm": 3.2703733444213867, + "learning_rate": 1.732977303070761e-05, + "loss": 5.5812, + "step": 649 + }, + { + "epoch": 0.052113607664709065, + "grad_norm": 3.024962902069092, + "learning_rate": 1.7356475300400533e-05, + "loss": 5.6454, + "step": 650 + }, + { + "epoch": 0.052193782445731696, + "grad_norm": 2.8068461418151855, + "learning_rate": 1.738317757009346e-05, + "loss": 5.8087, + "step": 651 + }, + { + "epoch": 0.052273957226754326, + "grad_norm": 3.6074740886688232, + "learning_rate": 1.7409879839786384e-05, + "loss": 5.6599, + "step": 652 + }, + { + "epoch": 0.05235413200777695, + "grad_norm": 2.9819726943969727, + "learning_rate": 1.743658210947931e-05, + "loss": 5.6338, + "step": 653 + }, + { + "epoch": 0.05243430678879958, + "grad_norm": 3.1805548667907715, + "learning_rate": 1.746328437917223e-05, + "loss": 5.6404, + "step": 654 + }, + { + "epoch": 0.05251448156982221, + "grad_norm": 2.8106613159179688, + "learning_rate": 1.7489986648865156e-05, + "loss": 5.7412, + "step": 655 + }, + { + "epoch": 0.05259465635084484, + "grad_norm": 3.4122767448425293, + "learning_rate": 1.7516688918558078e-05, + "loss": 5.636, + "step": 656 + }, + { + "epoch": 0.052674831131867474, + "grad_norm": 2.8597397804260254, + "learning_rate": 1.7543391188251003e-05, + "loss": 5.6055, + "step": 657 + }, + { + "epoch": 0.0527550059128901, + "grad_norm": 2.7356834411621094, + "learning_rate": 1.7570093457943925e-05, + "loss": 5.6273, + "step": 658 + }, + { + "epoch": 0.05283518069391273, + "grad_norm": 3.558483839035034, + "learning_rate": 1.759679572763685e-05, + "loss": 5.6368, + "step": 659 + }, + { + "epoch": 0.05291535547493536, + "grad_norm": 2.760746479034424, + "learning_rate": 1.7623497997329775e-05, + "loss": 5.5546, + "step": 660 + }, + { + "epoch": 0.05299553025595799, + "grad_norm": 2.7118146419525146, + "learning_rate": 1.76502002670227e-05, + "loss": 5.7057, + "step": 661 + }, + { + "epoch": 0.05307570503698062, + "grad_norm": 1.3325458765029907, + "learning_rate": 1.7676902536715623e-05, + "loss": 5.5364, + "step": 662 + }, + { + "epoch": 0.053155879818003245, + "grad_norm": 3.2511508464813232, + "learning_rate": 1.7703604806408544e-05, + "loss": 5.6724, + "step": 663 + }, + { + "epoch": 0.053236054599025875, + "grad_norm": 2.795154094696045, + "learning_rate": 1.773030707610147e-05, + "loss": 5.6508, + "step": 664 + }, + { + "epoch": 0.053316229380048506, + "grad_norm": 2.8278191089630127, + "learning_rate": 1.7757009345794395e-05, + "loss": 5.6172, + "step": 665 + }, + { + "epoch": 0.05339640416107114, + "grad_norm": 2.816784381866455, + "learning_rate": 1.7783711615487317e-05, + "loss": 5.7201, + "step": 666 + }, + { + "epoch": 0.05347657894209376, + "grad_norm": 2.9500198364257812, + "learning_rate": 1.7810413885180242e-05, + "loss": 5.7573, + "step": 667 + }, + { + "epoch": 0.05355675372311639, + "grad_norm": 3.6587958335876465, + "learning_rate": 1.7837116154873167e-05, + "loss": 5.5933, + "step": 668 + }, + { + "epoch": 0.05363692850413902, + "grad_norm": 2.7161242961883545, + "learning_rate": 1.786381842456609e-05, + "loss": 5.5811, + "step": 669 + }, + { + "epoch": 0.05371710328516165, + "grad_norm": 2.938166618347168, + "learning_rate": 1.7890520694259014e-05, + "loss": 5.6317, + "step": 670 + }, + { + "epoch": 0.053797278066184284, + "grad_norm": 2.7409870624542236, + "learning_rate": 1.7917222963951936e-05, + "loss": 5.5939, + "step": 671 + }, + { + "epoch": 0.05387745284720691, + "grad_norm": 2.843219518661499, + "learning_rate": 1.794392523364486e-05, + "loss": 5.6088, + "step": 672 + }, + { + "epoch": 0.05395762762822954, + "grad_norm": 2.6621828079223633, + "learning_rate": 1.7970627503337786e-05, + "loss": 5.675, + "step": 673 + }, + { + "epoch": 0.05403780240925217, + "grad_norm": 2.5496578216552734, + "learning_rate": 1.7997329773030708e-05, + "loss": 5.6685, + "step": 674 + }, + { + "epoch": 0.0541179771902748, + "grad_norm": 3.0093417167663574, + "learning_rate": 1.8024032042723633e-05, + "loss": 5.6041, + "step": 675 + }, + { + "epoch": 0.05419815197129743, + "grad_norm": 2.8968422412872314, + "learning_rate": 1.805073431241656e-05, + "loss": 5.7122, + "step": 676 + }, + { + "epoch": 0.054278326752320055, + "grad_norm": 2.834254026412964, + "learning_rate": 1.807743658210948e-05, + "loss": 5.5732, + "step": 677 + }, + { + "epoch": 0.054358501533342686, + "grad_norm": 2.6735472679138184, + "learning_rate": 1.8104138851802406e-05, + "loss": 5.6622, + "step": 678 + }, + { + "epoch": 0.05443867631436532, + "grad_norm": 2.7397592067718506, + "learning_rate": 1.8130841121495328e-05, + "loss": 5.6692, + "step": 679 + }, + { + "epoch": 0.05451885109538795, + "grad_norm": 3.093621253967285, + "learning_rate": 1.8157543391188253e-05, + "loss": 5.538, + "step": 680 + }, + { + "epoch": 0.05459902587641058, + "grad_norm": 2.8383216857910156, + "learning_rate": 1.8184245660881178e-05, + "loss": 5.6367, + "step": 681 + }, + { + "epoch": 0.0546792006574332, + "grad_norm": 2.9389822483062744, + "learning_rate": 1.82109479305741e-05, + "loss": 5.5675, + "step": 682 + }, + { + "epoch": 0.05475937543845583, + "grad_norm": 2.9491333961486816, + "learning_rate": 1.8237650200267025e-05, + "loss": 5.6373, + "step": 683 + }, + { + "epoch": 0.054839550219478464, + "grad_norm": 3.1163148880004883, + "learning_rate": 1.8264352469959947e-05, + "loss": 5.5565, + "step": 684 + }, + { + "epoch": 0.054919725000501095, + "grad_norm": 3.2793915271759033, + "learning_rate": 1.8291054739652872e-05, + "loss": 5.5293, + "step": 685 + }, + { + "epoch": 0.05499989978152372, + "grad_norm": 3.1565775871276855, + "learning_rate": 1.8317757009345797e-05, + "loss": 5.6121, + "step": 686 + }, + { + "epoch": 0.05508007456254635, + "grad_norm": 3.2414512634277344, + "learning_rate": 1.834445927903872e-05, + "loss": 5.5354, + "step": 687 + }, + { + "epoch": 0.05516024934356898, + "grad_norm": 3.2542595863342285, + "learning_rate": 1.8371161548731644e-05, + "loss": 5.6089, + "step": 688 + }, + { + "epoch": 0.05524042412459161, + "grad_norm": 1.409609079360962, + "learning_rate": 1.839786381842457e-05, + "loss": 5.519, + "step": 689 + }, + { + "epoch": 0.05532059890561424, + "grad_norm": 3.08117938041687, + "learning_rate": 1.842456608811749e-05, + "loss": 5.5331, + "step": 690 + }, + { + "epoch": 0.055400773686636866, + "grad_norm": 1.3400872945785522, + "learning_rate": 1.8451268357810413e-05, + "loss": 5.5122, + "step": 691 + }, + { + "epoch": 0.0554809484676595, + "grad_norm": 3.151120662689209, + "learning_rate": 1.847797062750334e-05, + "loss": 5.6528, + "step": 692 + }, + { + "epoch": 0.05556112324868213, + "grad_norm": 2.9463346004486084, + "learning_rate": 1.8504672897196264e-05, + "loss": 5.636, + "step": 693 + }, + { + "epoch": 0.05564129802970476, + "grad_norm": 3.140421152114868, + "learning_rate": 1.8531375166889186e-05, + "loss": 5.5509, + "step": 694 + }, + { + "epoch": 0.05572147281072739, + "grad_norm": 1.255446195602417, + "learning_rate": 1.855807743658211e-05, + "loss": 5.5186, + "step": 695 + }, + { + "epoch": 0.05580164759175001, + "grad_norm": 2.8152389526367188, + "learning_rate": 1.8584779706275036e-05, + "loss": 5.6891, + "step": 696 + }, + { + "epoch": 0.055881822372772644, + "grad_norm": 3.1993637084960938, + "learning_rate": 1.8611481975967958e-05, + "loss": 5.6079, + "step": 697 + }, + { + "epoch": 0.055961997153795275, + "grad_norm": 3.279301404953003, + "learning_rate": 1.8638184245660883e-05, + "loss": 5.6825, + "step": 698 + }, + { + "epoch": 0.056042171934817905, + "grad_norm": 3.013810396194458, + "learning_rate": 1.8664886515353805e-05, + "loss": 5.51, + "step": 699 + }, + { + "epoch": 0.05612234671584053, + "grad_norm": 3.221156120300293, + "learning_rate": 1.869158878504673e-05, + "loss": 5.541, + "step": 700 + }, + { + "epoch": 0.05620252149686316, + "grad_norm": 3.1381072998046875, + "learning_rate": 1.8718291054739655e-05, + "loss": 5.4451, + "step": 701 + }, + { + "epoch": 0.05628269627788579, + "grad_norm": 3.2891857624053955, + "learning_rate": 1.8744993324432577e-05, + "loss": 5.5814, + "step": 702 + }, + { + "epoch": 0.05636287105890842, + "grad_norm": 3.1093356609344482, + "learning_rate": 1.8771695594125502e-05, + "loss": 5.6549, + "step": 703 + }, + { + "epoch": 0.05644304583993105, + "grad_norm": 3.392735481262207, + "learning_rate": 1.8798397863818424e-05, + "loss": 5.6106, + "step": 704 + }, + { + "epoch": 0.056523220620953676, + "grad_norm": 3.1392292976379395, + "learning_rate": 1.882510013351135e-05, + "loss": 5.5833, + "step": 705 + }, + { + "epoch": 0.05660339540197631, + "grad_norm": 3.2494406700134277, + "learning_rate": 1.8851802403204275e-05, + "loss": 5.5663, + "step": 706 + }, + { + "epoch": 0.05668357018299894, + "grad_norm": 3.067695140838623, + "learning_rate": 1.8878504672897197e-05, + "loss": 5.5244, + "step": 707 + }, + { + "epoch": 0.05676374496402157, + "grad_norm": 2.974229335784912, + "learning_rate": 1.8905206942590122e-05, + "loss": 5.5486, + "step": 708 + }, + { + "epoch": 0.0568439197450442, + "grad_norm": 3.6237294673919678, + "learning_rate": 1.8931909212283047e-05, + "loss": 5.5161, + "step": 709 + }, + { + "epoch": 0.056924094526066824, + "grad_norm": 3.854116201400757, + "learning_rate": 1.895861148197597e-05, + "loss": 5.4833, + "step": 710 + }, + { + "epoch": 0.057004269307089454, + "grad_norm": 1.2823879718780518, + "learning_rate": 1.8985313751668894e-05, + "loss": 5.49, + "step": 711 + }, + { + "epoch": 0.057084444088112085, + "grad_norm": 3.7311959266662598, + "learning_rate": 1.9012016021361816e-05, + "loss": 5.6587, + "step": 712 + }, + { + "epoch": 0.057164618869134716, + "grad_norm": 3.54887318611145, + "learning_rate": 1.903871829105474e-05, + "loss": 5.5315, + "step": 713 + }, + { + "epoch": 0.05724479365015734, + "grad_norm": 3.2061192989349365, + "learning_rate": 1.9065420560747666e-05, + "loss": 5.509, + "step": 714 + }, + { + "epoch": 0.05732496843117997, + "grad_norm": 3.432830572128296, + "learning_rate": 1.9092122830440588e-05, + "loss": 5.6086, + "step": 715 + }, + { + "epoch": 0.0574051432122026, + "grad_norm": 3.2045154571533203, + "learning_rate": 1.9118825100133513e-05, + "loss": 5.5571, + "step": 716 + }, + { + "epoch": 0.05748531799322523, + "grad_norm": 3.125880479812622, + "learning_rate": 1.914552736982644e-05, + "loss": 5.5618, + "step": 717 + }, + { + "epoch": 0.05756549277424786, + "grad_norm": 3.509491205215454, + "learning_rate": 1.917222963951936e-05, + "loss": 5.5565, + "step": 718 + }, + { + "epoch": 0.05764566755527049, + "grad_norm": 3.6842870712280273, + "learning_rate": 1.9198931909212282e-05, + "loss": 5.6095, + "step": 719 + }, + { + "epoch": 0.05772584233629312, + "grad_norm": 4.6519083976745605, + "learning_rate": 1.9225634178905207e-05, + "loss": 5.4205, + "step": 720 + }, + { + "epoch": 0.05780601711731575, + "grad_norm": 3.868870258331299, + "learning_rate": 1.9252336448598133e-05, + "loss": 5.5352, + "step": 721 + }, + { + "epoch": 0.05788619189833838, + "grad_norm": 3.4870991706848145, + "learning_rate": 1.9279038718291058e-05, + "loss": 5.5354, + "step": 722 + }, + { + "epoch": 0.05796636667936101, + "grad_norm": 4.445798397064209, + "learning_rate": 1.930574098798398e-05, + "loss": 5.52, + "step": 723 + }, + { + "epoch": 0.058046541460383634, + "grad_norm": 3.4152092933654785, + "learning_rate": 1.9332443257676905e-05, + "loss": 5.5522, + "step": 724 + }, + { + "epoch": 0.058126716241406265, + "grad_norm": 4.600973606109619, + "learning_rate": 1.9359145527369827e-05, + "loss": 5.5393, + "step": 725 + }, + { + "epoch": 0.058206891022428896, + "grad_norm": 4.753367900848389, + "learning_rate": 1.9385847797062752e-05, + "loss": 5.5757, + "step": 726 + }, + { + "epoch": 0.05828706580345153, + "grad_norm": 3.0337533950805664, + "learning_rate": 1.9412550066755674e-05, + "loss": 5.4596, + "step": 727 + }, + { + "epoch": 0.05836724058447415, + "grad_norm": 4.212721824645996, + "learning_rate": 1.94392523364486e-05, + "loss": 5.5037, + "step": 728 + }, + { + "epoch": 0.05844741536549678, + "grad_norm": 3.526479482650757, + "learning_rate": 1.9465954606141524e-05, + "loss": 5.5, + "step": 729 + }, + { + "epoch": 0.05852759014651941, + "grad_norm": 3.9972445964813232, + "learning_rate": 1.949265687583445e-05, + "loss": 5.625, + "step": 730 + }, + { + "epoch": 0.05860776492754204, + "grad_norm": 3.8696465492248535, + "learning_rate": 1.951935914552737e-05, + "loss": 5.4739, + "step": 731 + }, + { + "epoch": 0.058687939708564674, + "grad_norm": 4.505939483642578, + "learning_rate": 1.9546061415220293e-05, + "loss": 5.4355, + "step": 732 + }, + { + "epoch": 0.0587681144895873, + "grad_norm": 4.244717121124268, + "learning_rate": 1.957276368491322e-05, + "loss": 5.407, + "step": 733 + }, + { + "epoch": 0.05884828927060993, + "grad_norm": 3.954223155975342, + "learning_rate": 1.9599465954606144e-05, + "loss": 5.4727, + "step": 734 + }, + { + "epoch": 0.05892846405163256, + "grad_norm": 5.416421890258789, + "learning_rate": 1.9626168224299065e-05, + "loss": 5.469, + "step": 735 + }, + { + "epoch": 0.05900863883265519, + "grad_norm": 4.427737236022949, + "learning_rate": 1.965287049399199e-05, + "loss": 5.3743, + "step": 736 + }, + { + "epoch": 0.05908881361367782, + "grad_norm": 4.022106647491455, + "learning_rate": 1.9679572763684916e-05, + "loss": 5.4724, + "step": 737 + }, + { + "epoch": 0.059168988394700445, + "grad_norm": 6.069715976715088, + "learning_rate": 1.970627503337784e-05, + "loss": 5.4343, + "step": 738 + }, + { + "epoch": 0.059249163175723076, + "grad_norm": 3.8946640491485596, + "learning_rate": 1.9732977303070763e-05, + "loss": 5.5112, + "step": 739 + }, + { + "epoch": 0.059329337956745706, + "grad_norm": 5.465697765350342, + "learning_rate": 1.9759679572763685e-05, + "loss": 5.4093, + "step": 740 + }, + { + "epoch": 0.05940951273776834, + "grad_norm": 5.635557174682617, + "learning_rate": 1.978638184245661e-05, + "loss": 5.4513, + "step": 741 + }, + { + "epoch": 0.05948968751879096, + "grad_norm": 6.082085609436035, + "learning_rate": 1.9813084112149535e-05, + "loss": 5.3987, + "step": 742 + }, + { + "epoch": 0.05956986229981359, + "grad_norm": 5.32405424118042, + "learning_rate": 1.9839786381842457e-05, + "loss": 5.4471, + "step": 743 + }, + { + "epoch": 0.05965003708083622, + "grad_norm": 8.123515129089355, + "learning_rate": 1.9866488651535382e-05, + "loss": 5.3632, + "step": 744 + }, + { + "epoch": 0.05973021186185885, + "grad_norm": 12.88020133972168, + "learning_rate": 1.9893190921228308e-05, + "loss": 5.4045, + "step": 745 + }, + { + "epoch": 0.059810386642881484, + "grad_norm": 8.404499053955078, + "learning_rate": 1.991989319092123e-05, + "loss": 5.2935, + "step": 746 + }, + { + "epoch": 0.05989056142390411, + "grad_norm": 4.698278903961182, + "learning_rate": 1.9946595460614155e-05, + "loss": 5.5188, + "step": 747 + }, + { + "epoch": 0.05997073620492674, + "grad_norm": 6.8173112869262695, + "learning_rate": 1.9973297730307076e-05, + "loss": 5.2656, + "step": 748 + }, + { + "epoch": 0.06005091098594937, + "grad_norm": 5.736358165740967, + "learning_rate": 2e-05, + "loss": 5.411, + "step": 749 + }, + { + "epoch": 0.060131085766972, + "grad_norm": 6.891531944274902, + "learning_rate": 1.9999999915701765e-05, + "loss": 5.203, + "step": 750 + }, + { + "epoch": 0.06021126054799463, + "grad_norm": 7.204330921173096, + "learning_rate": 1.999999966280705e-05, + "loss": 5.3196, + "step": 751 + }, + { + "epoch": 0.060291435329017255, + "grad_norm": 6.025188446044922, + "learning_rate": 1.9999999241315867e-05, + "loss": 5.3657, + "step": 752 + }, + { + "epoch": 0.060371610110039886, + "grad_norm": 6.939337730407715, + "learning_rate": 1.9999998651228226e-05, + "loss": 5.1663, + "step": 753 + }, + { + "epoch": 0.06045178489106252, + "grad_norm": 5.026117324829102, + "learning_rate": 1.9999997892544126e-05, + "loss": 5.3188, + "step": 754 + }, + { + "epoch": 0.06053195967208515, + "grad_norm": 6.505256652832031, + "learning_rate": 1.999999696526359e-05, + "loss": 5.1702, + "step": 755 + }, + { + "epoch": 0.06061213445310777, + "grad_norm": 6.212100982666016, + "learning_rate": 1.9999995869386626e-05, + "loss": 5.2842, + "step": 756 + }, + { + "epoch": 0.0606923092341304, + "grad_norm": 7.508810043334961, + "learning_rate": 1.9999994604913257e-05, + "loss": 5.1548, + "step": 757 + }, + { + "epoch": 0.06077248401515303, + "grad_norm": 7.921309471130371, + "learning_rate": 1.9999993171843502e-05, + "loss": 5.0969, + "step": 758 + }, + { + "epoch": 0.060852658796175664, + "grad_norm": 20.673992156982422, + "learning_rate": 1.9999991570177393e-05, + "loss": 5.1829, + "step": 759 + }, + { + "epoch": 0.060932833577198295, + "grad_norm": 10.641246795654297, + "learning_rate": 1.9999989799914945e-05, + "loss": 5.172, + "step": 760 + }, + { + "epoch": 0.06101300835822092, + "grad_norm": 8.045785903930664, + "learning_rate": 1.999998786105619e-05, + "loss": 5.1876, + "step": 761 + }, + { + "epoch": 0.06109318313924355, + "grad_norm": 7.901401996612549, + "learning_rate": 1.9999985753601172e-05, + "loss": 5.1194, + "step": 762 + }, + { + "epoch": 0.06117335792026618, + "grad_norm": 6.831056594848633, + "learning_rate": 1.9999983477549913e-05, + "loss": 5.1931, + "step": 763 + }, + { + "epoch": 0.06125353270128881, + "grad_norm": 6.556690216064453, + "learning_rate": 1.999998103290246e-05, + "loss": 5.0659, + "step": 764 + }, + { + "epoch": 0.06133370748231144, + "grad_norm": 6.755990982055664, + "learning_rate": 1.9999978419658846e-05, + "loss": 4.9895, + "step": 765 + }, + { + "epoch": 0.061413882263334066, + "grad_norm": 6.814157485961914, + "learning_rate": 1.9999975637819126e-05, + "loss": 4.9564, + "step": 766 + }, + { + "epoch": 0.0614940570443567, + "grad_norm": 5.70683479309082, + "learning_rate": 1.999997268738334e-05, + "loss": 4.9606, + "step": 767 + }, + { + "epoch": 0.06157423182537933, + "grad_norm": 6.331331253051758, + "learning_rate": 1.999996956835154e-05, + "loss": 4.9182, + "step": 768 + }, + { + "epoch": 0.06165440660640196, + "grad_norm": 6.776618957519531, + "learning_rate": 1.9999966280723773e-05, + "loss": 4.9543, + "step": 769 + }, + { + "epoch": 0.06173458138742458, + "grad_norm": 6.254655361175537, + "learning_rate": 1.99999628245001e-05, + "loss": 4.7861, + "step": 770 + }, + { + "epoch": 0.06181475616844721, + "grad_norm": 6.0269999504089355, + "learning_rate": 1.999995919968058e-05, + "loss": 4.9156, + "step": 771 + }, + { + "epoch": 0.061894930949469844, + "grad_norm": 5.454859733581543, + "learning_rate": 1.999995540626527e-05, + "loss": 4.7606, + "step": 772 + }, + { + "epoch": 0.061975105730492475, + "grad_norm": 5.605257511138916, + "learning_rate": 1.999995144425424e-05, + "loss": 4.6582, + "step": 773 + }, + { + "epoch": 0.062055280511515105, + "grad_norm": 5.896077632904053, + "learning_rate": 1.999994731364755e-05, + "loss": 4.6582, + "step": 774 + }, + { + "epoch": 0.06213545529253773, + "grad_norm": 11.941583633422852, + "learning_rate": 1.9999943014445272e-05, + "loss": 4.4639, + "step": 775 + }, + { + "epoch": 0.06221563007356036, + "grad_norm": 10.171135902404785, + "learning_rate": 1.999993854664748e-05, + "loss": 4.7159, + "step": 776 + }, + { + "epoch": 0.06229580485458299, + "grad_norm": 6.884143829345703, + "learning_rate": 1.999993391025425e-05, + "loss": 4.7386, + "step": 777 + }, + { + "epoch": 0.06237597963560562, + "grad_norm": 7.063316345214844, + "learning_rate": 1.9999929105265657e-05, + "loss": 4.6045, + "step": 778 + }, + { + "epoch": 0.06245615441662825, + "grad_norm": 6.897941589355469, + "learning_rate": 1.9999924131681782e-05, + "loss": 4.6461, + "step": 779 + }, + { + "epoch": 0.06253632919765088, + "grad_norm": 6.218994617462158, + "learning_rate": 1.9999918989502714e-05, + "loss": 4.63, + "step": 780 + }, + { + "epoch": 0.06261650397867351, + "grad_norm": 6.069316864013672, + "learning_rate": 1.9999913678728534e-05, + "loss": 4.5283, + "step": 781 + }, + { + "epoch": 0.06269667875969613, + "grad_norm": 5.288468837738037, + "learning_rate": 1.9999908199359335e-05, + "loss": 4.3992, + "step": 782 + }, + { + "epoch": 0.06277685354071877, + "grad_norm": 5.8183913230896, + "learning_rate": 1.9999902551395206e-05, + "loss": 4.4675, + "step": 783 + }, + { + "epoch": 0.06285702832174139, + "grad_norm": 5.018249988555908, + "learning_rate": 1.9999896734836244e-05, + "loss": 4.3861, + "step": 784 + }, + { + "epoch": 0.06293720310276403, + "grad_norm": 5.851385116577148, + "learning_rate": 1.999989074968255e-05, + "loss": 4.499, + "step": 785 + }, + { + "epoch": 0.06301737788378665, + "grad_norm": 5.501115798950195, + "learning_rate": 1.999988459593422e-05, + "loss": 4.433, + "step": 786 + }, + { + "epoch": 0.06309755266480928, + "grad_norm": 5.200289249420166, + "learning_rate": 1.999987827359136e-05, + "loss": 4.3176, + "step": 787 + }, + { + "epoch": 0.06317772744583192, + "grad_norm": 5.101953029632568, + "learning_rate": 1.9999871782654078e-05, + "loss": 4.274, + "step": 788 + }, + { + "epoch": 0.06325790222685454, + "grad_norm": 4.636188507080078, + "learning_rate": 1.9999865123122482e-05, + "loss": 4.1945, + "step": 789 + }, + { + "epoch": 0.06333807700787718, + "grad_norm": 4.8744378089904785, + "learning_rate": 1.9999858294996683e-05, + "loss": 4.3066, + "step": 790 + }, + { + "epoch": 0.0634182517888998, + "grad_norm": 10.143219947814941, + "learning_rate": 1.99998512982768e-05, + "loss": 3.8665, + "step": 791 + }, + { + "epoch": 0.06349842656992243, + "grad_norm": 6.044327259063721, + "learning_rate": 1.9999844132962948e-05, + "loss": 4.312, + "step": 792 + }, + { + "epoch": 0.06357860135094506, + "grad_norm": 6.414331912994385, + "learning_rate": 1.999983679905525e-05, + "loss": 4.3757, + "step": 793 + }, + { + "epoch": 0.06365877613196769, + "grad_norm": 5.5477142333984375, + "learning_rate": 1.999982929655382e-05, + "loss": 4.3398, + "step": 794 + }, + { + "epoch": 0.06373895091299032, + "grad_norm": 4.401949405670166, + "learning_rate": 1.9999821625458797e-05, + "loss": 4.2305, + "step": 795 + }, + { + "epoch": 0.06381912569401295, + "grad_norm": 5.3943023681640625, + "learning_rate": 1.9999813785770307e-05, + "loss": 4.1962, + "step": 796 + }, + { + "epoch": 0.06389930047503557, + "grad_norm": 5.194131374359131, + "learning_rate": 1.9999805777488478e-05, + "loss": 4.1407, + "step": 797 + }, + { + "epoch": 0.06397947525605821, + "grad_norm": 4.772655487060547, + "learning_rate": 1.9999797600613454e-05, + "loss": 4.2385, + "step": 798 + }, + { + "epoch": 0.06405965003708083, + "grad_norm": 5.536653518676758, + "learning_rate": 1.999978925514536e-05, + "loss": 4.1591, + "step": 799 + }, + { + "epoch": 0.06413982481810347, + "grad_norm": 5.270934581756592, + "learning_rate": 1.9999780741084348e-05, + "loss": 4.2601, + "step": 800 + }, + { + "epoch": 0.0642199995991261, + "grad_norm": 4.482230186462402, + "learning_rate": 1.9999772058430553e-05, + "loss": 4.2654, + "step": 801 + }, + { + "epoch": 0.06430017438014872, + "grad_norm": 4.52065896987915, + "learning_rate": 1.9999763207184127e-05, + "loss": 4.0619, + "step": 802 + }, + { + "epoch": 0.06438034916117136, + "grad_norm": 4.862673282623291, + "learning_rate": 1.9999754187345218e-05, + "loss": 4.04, + "step": 803 + }, + { + "epoch": 0.06446052394219398, + "grad_norm": 5.118383884429932, + "learning_rate": 1.9999744998913976e-05, + "loss": 4.0206, + "step": 804 + }, + { + "epoch": 0.06454069872321662, + "grad_norm": 4.234626770019531, + "learning_rate": 1.999973564189056e-05, + "loss": 4.1033, + "step": 805 + }, + { + "epoch": 0.06462087350423924, + "grad_norm": 4.7218523025512695, + "learning_rate": 1.9999726116275125e-05, + "loss": 4.1086, + "step": 806 + }, + { + "epoch": 0.06470104828526187, + "grad_norm": 4.869848251342773, + "learning_rate": 1.999971642206783e-05, + "loss": 4.0472, + "step": 807 + }, + { + "epoch": 0.0647812230662845, + "grad_norm": 4.525954246520996, + "learning_rate": 1.9999706559268842e-05, + "loss": 3.9938, + "step": 808 + }, + { + "epoch": 0.06486139784730713, + "grad_norm": 4.849183082580566, + "learning_rate": 1.9999696527878326e-05, + "loss": 4.063, + "step": 809 + }, + { + "epoch": 0.06494157262832975, + "grad_norm": 4.557774543762207, + "learning_rate": 1.9999686327896448e-05, + "loss": 3.9545, + "step": 810 + }, + { + "epoch": 0.06502174740935239, + "grad_norm": 4.29681396484375, + "learning_rate": 1.9999675959323386e-05, + "loss": 3.9725, + "step": 811 + }, + { + "epoch": 0.06510192219037501, + "grad_norm": 5.011354446411133, + "learning_rate": 1.9999665422159312e-05, + "loss": 3.9872, + "step": 812 + }, + { + "epoch": 0.06518209697139765, + "grad_norm": 4.5723676681518555, + "learning_rate": 1.99996547164044e-05, + "loss": 3.9974, + "step": 813 + }, + { + "epoch": 0.06526227175242028, + "grad_norm": 4.363810062408447, + "learning_rate": 1.9999643842058836e-05, + "loss": 3.9861, + "step": 814 + }, + { + "epoch": 0.0653424465334429, + "grad_norm": 4.388919830322266, + "learning_rate": 1.99996327991228e-05, + "loss": 3.8942, + "step": 815 + }, + { + "epoch": 0.06542262131446554, + "grad_norm": 4.021622180938721, + "learning_rate": 1.9999621587596478e-05, + "loss": 4.1438, + "step": 816 + }, + { + "epoch": 0.06550279609548816, + "grad_norm": 9.669590950012207, + "learning_rate": 1.999961020748006e-05, + "loss": 3.3906, + "step": 817 + }, + { + "epoch": 0.0655829708765108, + "grad_norm": 5.015942096710205, + "learning_rate": 1.999959865877374e-05, + "loss": 3.302, + "step": 818 + }, + { + "epoch": 0.06566314565753342, + "grad_norm": 7.398265361785889, + "learning_rate": 1.9999586941477713e-05, + "loss": 3.8933, + "step": 819 + }, + { + "epoch": 0.06574332043855605, + "grad_norm": 5.77410364151001, + "learning_rate": 1.999957505559217e-05, + "loss": 4.0111, + "step": 820 + }, + { + "epoch": 0.06582349521957868, + "grad_norm": 4.8476996421813965, + "learning_rate": 1.999956300111732e-05, + "loss": 3.9471, + "step": 821 + }, + { + "epoch": 0.06590367000060131, + "grad_norm": 5.597156047821045, + "learning_rate": 1.999955077805336e-05, + "loss": 3.9043, + "step": 822 + }, + { + "epoch": 0.06598384478162395, + "grad_norm": 5.034807205200195, + "learning_rate": 1.9999538386400497e-05, + "loss": 3.9528, + "step": 823 + }, + { + "epoch": 0.06606401956264657, + "grad_norm": 5.190936088562012, + "learning_rate": 1.9999525826158944e-05, + "loss": 3.9081, + "step": 824 + }, + { + "epoch": 0.0661441943436692, + "grad_norm": 4.991243362426758, + "learning_rate": 1.9999513097328903e-05, + "loss": 3.8994, + "step": 825 + }, + { + "epoch": 0.06622436912469183, + "grad_norm": 4.245006084442139, + "learning_rate": 1.99995001999106e-05, + "loss": 3.8765, + "step": 826 + }, + { + "epoch": 0.06630454390571446, + "grad_norm": 4.242435455322266, + "learning_rate": 1.999948713390425e-05, + "loss": 3.9388, + "step": 827 + }, + { + "epoch": 0.0663847186867371, + "grad_norm": 4.640059947967529, + "learning_rate": 1.999947389931007e-05, + "loss": 3.8767, + "step": 828 + }, + { + "epoch": 0.06646489346775972, + "grad_norm": 4.40859842300415, + "learning_rate": 1.9999460496128283e-05, + "loss": 3.8619, + "step": 829 + }, + { + "epoch": 0.06654506824878234, + "grad_norm": 4.076308727264404, + "learning_rate": 1.9999446924359113e-05, + "loss": 3.7885, + "step": 830 + }, + { + "epoch": 0.06662524302980498, + "grad_norm": 4.16943883895874, + "learning_rate": 1.9999433184002797e-05, + "loss": 3.9246, + "step": 831 + }, + { + "epoch": 0.0667054178108276, + "grad_norm": 3.935196876525879, + "learning_rate": 1.999941927505956e-05, + "loss": 3.9535, + "step": 832 + }, + { + "epoch": 0.06678559259185024, + "grad_norm": 4.117587089538574, + "learning_rate": 1.999940519752964e-05, + "loss": 3.8635, + "step": 833 + }, + { + "epoch": 0.06686576737287286, + "grad_norm": 3.9258346557617188, + "learning_rate": 1.9999390951413268e-05, + "loss": 3.9613, + "step": 834 + }, + { + "epoch": 0.06694594215389549, + "grad_norm": 3.6288158893585205, + "learning_rate": 1.9999376536710692e-05, + "loss": 3.8485, + "step": 835 + }, + { + "epoch": 0.06702611693491813, + "grad_norm": 4.185365676879883, + "learning_rate": 1.999936195342215e-05, + "loss": 3.8353, + "step": 836 + }, + { + "epoch": 0.06710629171594075, + "grad_norm": 4.066679000854492, + "learning_rate": 1.999934720154789e-05, + "loss": 3.7648, + "step": 837 + }, + { + "epoch": 0.06718646649696337, + "grad_norm": 4.057693958282471, + "learning_rate": 1.999933228108816e-05, + "loss": 3.9264, + "step": 838 + }, + { + "epoch": 0.06726664127798601, + "grad_norm": 3.928955554962158, + "learning_rate": 1.9999317192043213e-05, + "loss": 3.8458, + "step": 839 + }, + { + "epoch": 0.06734681605900864, + "grad_norm": 3.730722427368164, + "learning_rate": 1.9999301934413303e-05, + "loss": 3.9241, + "step": 840 + }, + { + "epoch": 0.06742699084003127, + "grad_norm": 3.792874813079834, + "learning_rate": 1.9999286508198686e-05, + "loss": 3.8076, + "step": 841 + }, + { + "epoch": 0.0675071656210539, + "grad_norm": 3.8320648670196533, + "learning_rate": 1.999927091339962e-05, + "loss": 3.7653, + "step": 842 + }, + { + "epoch": 0.06758734040207652, + "grad_norm": 3.6672396659851074, + "learning_rate": 1.9999255150016373e-05, + "loss": 3.7262, + "step": 843 + }, + { + "epoch": 0.06766751518309916, + "grad_norm": 3.911635160446167, + "learning_rate": 1.9999239218049208e-05, + "loss": 3.7716, + "step": 844 + }, + { + "epoch": 0.06774768996412178, + "grad_norm": 3.5148911476135254, + "learning_rate": 1.999922311749839e-05, + "loss": 3.7797, + "step": 845 + }, + { + "epoch": 0.06782786474514442, + "grad_norm": 3.829529047012329, + "learning_rate": 1.99992068483642e-05, + "loss": 3.7512, + "step": 846 + }, + { + "epoch": 0.06790803952616704, + "grad_norm": 3.7639009952545166, + "learning_rate": 1.9999190410646905e-05, + "loss": 3.72, + "step": 847 + }, + { + "epoch": 0.06798821430718967, + "grad_norm": 4.020934581756592, + "learning_rate": 1.9999173804346782e-05, + "loss": 3.8435, + "step": 848 + }, + { + "epoch": 0.0680683890882123, + "grad_norm": 4.339954853057861, + "learning_rate": 1.9999157029464115e-05, + "loss": 3.0858, + "step": 849 + }, + { + "epoch": 0.06814856386923493, + "grad_norm": 4.017866134643555, + "learning_rate": 1.9999140085999183e-05, + "loss": 3.7728, + "step": 850 + }, + { + "epoch": 0.06822873865025757, + "grad_norm": 3.8777029514312744, + "learning_rate": 1.9999122973952272e-05, + "loss": 3.7949, + "step": 851 + }, + { + "epoch": 0.06830891343128019, + "grad_norm": 3.598007917404175, + "learning_rate": 1.9999105693323674e-05, + "loss": 3.6146, + "step": 852 + }, + { + "epoch": 0.06838908821230281, + "grad_norm": 3.8576879501342773, + "learning_rate": 1.999908824411368e-05, + "loss": 3.6453, + "step": 853 + }, + { + "epoch": 0.06846926299332545, + "grad_norm": 4.4005022048950195, + "learning_rate": 1.9999070626322576e-05, + "loss": 3.7644, + "step": 854 + }, + { + "epoch": 0.06854943777434808, + "grad_norm": 4.182209014892578, + "learning_rate": 1.999905283995067e-05, + "loss": 3.025, + "step": 855 + }, + { + "epoch": 0.06862961255537071, + "grad_norm": 3.373344659805298, + "learning_rate": 1.9999034884998257e-05, + "loss": 3.6385, + "step": 856 + }, + { + "epoch": 0.06870978733639334, + "grad_norm": 4.610391139984131, + "learning_rate": 1.9999016761465638e-05, + "loss": 3.7198, + "step": 857 + }, + { + "epoch": 0.06878996211741596, + "grad_norm": 3.9405598640441895, + "learning_rate": 1.999899846935312e-05, + "loss": 3.7432, + "step": 858 + }, + { + "epoch": 0.0688701368984386, + "grad_norm": 3.703159809112549, + "learning_rate": 1.9998980008661015e-05, + "loss": 3.756, + "step": 859 + }, + { + "epoch": 0.06895031167946122, + "grad_norm": 3.6823079586029053, + "learning_rate": 1.9998961379389628e-05, + "loss": 3.7606, + "step": 860 + }, + { + "epoch": 0.06903048646048386, + "grad_norm": 3.54736328125, + "learning_rate": 1.9998942581539276e-05, + "loss": 3.642, + "step": 861 + }, + { + "epoch": 0.06911066124150649, + "grad_norm": 3.558504343032837, + "learning_rate": 1.999892361511028e-05, + "loss": 3.6051, + "step": 862 + }, + { + "epoch": 0.06919083602252911, + "grad_norm": 3.662184953689575, + "learning_rate": 1.999890448010295e-05, + "loss": 3.7483, + "step": 863 + }, + { + "epoch": 0.06927101080355175, + "grad_norm": 3.8055264949798584, + "learning_rate": 1.999888517651762e-05, + "loss": 3.7789, + "step": 864 + }, + { + "epoch": 0.06935118558457437, + "grad_norm": 3.431898593902588, + "learning_rate": 1.9998865704354605e-05, + "loss": 3.6964, + "step": 865 + }, + { + "epoch": 0.069431360365597, + "grad_norm": 3.4340925216674805, + "learning_rate": 1.9998846063614243e-05, + "loss": 3.7162, + "step": 866 + }, + { + "epoch": 0.06951153514661963, + "grad_norm": 3.4975173473358154, + "learning_rate": 1.9998826254296855e-05, + "loss": 3.6456, + "step": 867 + }, + { + "epoch": 0.06959170992764226, + "grad_norm": 3.653015613555908, + "learning_rate": 1.9998806276402784e-05, + "loss": 3.6356, + "step": 868 + }, + { + "epoch": 0.0696718847086649, + "grad_norm": 3.4071438312530518, + "learning_rate": 1.999878612993236e-05, + "loss": 3.5714, + "step": 869 + }, + { + "epoch": 0.06975205948968752, + "grad_norm": 3.6551873683929443, + "learning_rate": 1.999876581488593e-05, + "loss": 3.6714, + "step": 870 + }, + { + "epoch": 0.06983223427071014, + "grad_norm": 3.4401211738586426, + "learning_rate": 1.999874533126383e-05, + "loss": 3.6515, + "step": 871 + }, + { + "epoch": 0.06991240905173278, + "grad_norm": 3.6195175647735596, + "learning_rate": 1.9998724679066406e-05, + "loss": 3.754, + "step": 872 + }, + { + "epoch": 0.0699925838327554, + "grad_norm": 3.8479857444763184, + "learning_rate": 1.999870385829401e-05, + "loss": 3.7281, + "step": 873 + }, + { + "epoch": 0.07007275861377804, + "grad_norm": 3.4466075897216797, + "learning_rate": 1.999868286894699e-05, + "loss": 3.6569, + "step": 874 + }, + { + "epoch": 0.07015293339480067, + "grad_norm": 3.8395745754241943, + "learning_rate": 1.9998661711025703e-05, + "loss": 3.7096, + "step": 875 + }, + { + "epoch": 0.07023310817582329, + "grad_norm": 3.932075023651123, + "learning_rate": 1.99986403845305e-05, + "loss": 3.7471, + "step": 876 + }, + { + "epoch": 0.07031328295684593, + "grad_norm": 3.591506004333496, + "learning_rate": 1.9998618889461747e-05, + "loss": 3.6201, + "step": 877 + }, + { + "epoch": 0.07039345773786855, + "grad_norm": 3.9209604263305664, + "learning_rate": 1.99985972258198e-05, + "loss": 3.6545, + "step": 878 + }, + { + "epoch": 0.07047363251889119, + "grad_norm": 4.0281877517700195, + "learning_rate": 1.999857539360503e-05, + "loss": 3.6947, + "step": 879 + }, + { + "epoch": 0.07055380729991381, + "grad_norm": 3.166327476501465, + "learning_rate": 1.9998553392817804e-05, + "loss": 3.5326, + "step": 880 + }, + { + "epoch": 0.07063398208093644, + "grad_norm": 3.7718544006347656, + "learning_rate": 1.999853122345849e-05, + "loss": 3.5223, + "step": 881 + }, + { + "epoch": 0.07071415686195907, + "grad_norm": 3.7201082706451416, + "learning_rate": 1.999850888552746e-05, + "loss": 3.686, + "step": 882 + }, + { + "epoch": 0.0707943316429817, + "grad_norm": 3.562041997909546, + "learning_rate": 1.9998486379025096e-05, + "loss": 3.5854, + "step": 883 + }, + { + "epoch": 0.07087450642400434, + "grad_norm": 3.766857147216797, + "learning_rate": 1.999846370395178e-05, + "loss": 3.5754, + "step": 884 + }, + { + "epoch": 0.07095468120502696, + "grad_norm": 3.6891307830810547, + "learning_rate": 1.9998440860307886e-05, + "loss": 3.6851, + "step": 885 + }, + { + "epoch": 0.07103485598604958, + "grad_norm": 3.409726858139038, + "learning_rate": 1.9998417848093803e-05, + "loss": 3.6899, + "step": 886 + }, + { + "epoch": 0.07111503076707222, + "grad_norm": 3.2977826595306396, + "learning_rate": 1.999839466730992e-05, + "loss": 3.6285, + "step": 887 + }, + { + "epoch": 0.07119520554809484, + "grad_norm": 3.467258930206299, + "learning_rate": 1.9998371317956624e-05, + "loss": 3.6989, + "step": 888 + }, + { + "epoch": 0.07127538032911748, + "grad_norm": 3.571730613708496, + "learning_rate": 1.9998347800034314e-05, + "loss": 3.6255, + "step": 889 + }, + { + "epoch": 0.0713555551101401, + "grad_norm": 3.6391701698303223, + "learning_rate": 1.9998324113543384e-05, + "loss": 3.5245, + "step": 890 + }, + { + "epoch": 0.07143572989116273, + "grad_norm": 3.1792144775390625, + "learning_rate": 1.999830025848423e-05, + "loss": 3.6005, + "step": 891 + }, + { + "epoch": 0.07151590467218537, + "grad_norm": 3.255666732788086, + "learning_rate": 1.999827623485726e-05, + "loss": 3.5609, + "step": 892 + }, + { + "epoch": 0.07159607945320799, + "grad_norm": 3.3565661907196045, + "learning_rate": 1.9998252042662874e-05, + "loss": 3.6389, + "step": 893 + }, + { + "epoch": 0.07167625423423063, + "grad_norm": 3.2623322010040283, + "learning_rate": 1.9998227681901487e-05, + "loss": 3.5691, + "step": 894 + }, + { + "epoch": 0.07175642901525325, + "grad_norm": 3.565937042236328, + "learning_rate": 1.9998203152573503e-05, + "loss": 3.6161, + "step": 895 + }, + { + "epoch": 0.07183660379627588, + "grad_norm": 3.2191855907440186, + "learning_rate": 1.999817845467934e-05, + "loss": 3.6274, + "step": 896 + }, + { + "epoch": 0.07191677857729852, + "grad_norm": 3.4061663150787354, + "learning_rate": 1.9998153588219407e-05, + "loss": 3.5818, + "step": 897 + }, + { + "epoch": 0.07199695335832114, + "grad_norm": 3.6970231533050537, + "learning_rate": 1.999812855319413e-05, + "loss": 3.6597, + "step": 898 + }, + { + "epoch": 0.07207712813934376, + "grad_norm": 3.3757662773132324, + "learning_rate": 1.9998103349603927e-05, + "loss": 3.5704, + "step": 899 + }, + { + "epoch": 0.0721573029203664, + "grad_norm": 3.491680860519409, + "learning_rate": 1.999807797744923e-05, + "loss": 3.6274, + "step": 900 + }, + { + "epoch": 0.07223747770138902, + "grad_norm": 3.227346658706665, + "learning_rate": 1.9998052436730458e-05, + "loss": 3.583, + "step": 901 + }, + { + "epoch": 0.07231765248241166, + "grad_norm": 3.4145619869232178, + "learning_rate": 1.9998026727448048e-05, + "loss": 3.606, + "step": 902 + }, + { + "epoch": 0.07239782726343429, + "grad_norm": 3.2290141582489014, + "learning_rate": 1.999800084960243e-05, + "loss": 3.7383, + "step": 903 + }, + { + "epoch": 0.07247800204445691, + "grad_norm": 3.5676074028015137, + "learning_rate": 1.999797480319404e-05, + "loss": 3.5052, + "step": 904 + }, + { + "epoch": 0.07255817682547955, + "grad_norm": 3.4791758060455322, + "learning_rate": 1.999794858822332e-05, + "loss": 3.6471, + "step": 905 + }, + { + "epoch": 0.07263835160650217, + "grad_norm": 3.39568829536438, + "learning_rate": 1.999792220469071e-05, + "loss": 3.6388, + "step": 906 + }, + { + "epoch": 0.07271852638752481, + "grad_norm": 3.391282320022583, + "learning_rate": 1.9997895652596656e-05, + "loss": 3.5653, + "step": 907 + }, + { + "epoch": 0.07279870116854743, + "grad_norm": 3.274445056915283, + "learning_rate": 1.9997868931941603e-05, + "loss": 3.6753, + "step": 908 + }, + { + "epoch": 0.07287887594957006, + "grad_norm": 3.3663668632507324, + "learning_rate": 1.9997842042726007e-05, + "loss": 3.6354, + "step": 909 + }, + { + "epoch": 0.0729590507305927, + "grad_norm": 3.311185121536255, + "learning_rate": 1.9997814984950313e-05, + "loss": 3.6166, + "step": 910 + }, + { + "epoch": 0.07303922551161532, + "grad_norm": 3.6846513748168945, + "learning_rate": 1.9997787758614984e-05, + "loss": 3.576, + "step": 911 + }, + { + "epoch": 0.07311940029263796, + "grad_norm": 3.3655691146850586, + "learning_rate": 1.9997760363720478e-05, + "loss": 3.6304, + "step": 912 + }, + { + "epoch": 0.07319957507366058, + "grad_norm": 3.255951166152954, + "learning_rate": 1.9997732800267256e-05, + "loss": 3.5646, + "step": 913 + }, + { + "epoch": 0.0732797498546832, + "grad_norm": 3.4095821380615234, + "learning_rate": 1.9997705068255782e-05, + "loss": 3.603, + "step": 914 + }, + { + "epoch": 0.07335992463570584, + "grad_norm": 3.035325765609741, + "learning_rate": 1.9997677167686523e-05, + "loss": 2.8574, + "step": 915 + }, + { + "epoch": 0.07344009941672847, + "grad_norm": 3.231660842895508, + "learning_rate": 1.999764909855995e-05, + "loss": 3.5287, + "step": 916 + }, + { + "epoch": 0.0735202741977511, + "grad_norm": 3.2255494594573975, + "learning_rate": 1.999762086087654e-05, + "loss": 3.6689, + "step": 917 + }, + { + "epoch": 0.07360044897877373, + "grad_norm": 3.103349447250366, + "learning_rate": 1.9997592454636763e-05, + "loss": 3.4885, + "step": 918 + }, + { + "epoch": 0.07368062375979635, + "grad_norm": 3.522520065307617, + "learning_rate": 1.9997563879841103e-05, + "loss": 3.6152, + "step": 919 + }, + { + "epoch": 0.07376079854081899, + "grad_norm": 3.143623113632202, + "learning_rate": 1.9997535136490037e-05, + "loss": 3.545, + "step": 920 + }, + { + "epoch": 0.07384097332184161, + "grad_norm": 3.258782386779785, + "learning_rate": 1.9997506224584053e-05, + "loss": 3.6106, + "step": 921 + }, + { + "epoch": 0.07392114810286425, + "grad_norm": 3.4889090061187744, + "learning_rate": 1.999747714412364e-05, + "loss": 3.5615, + "step": 922 + }, + { + "epoch": 0.07400132288388687, + "grad_norm": 3.4011759757995605, + "learning_rate": 1.999744789510928e-05, + "loss": 3.5693, + "step": 923 + }, + { + "epoch": 0.0740814976649095, + "grad_norm": 3.6882898807525635, + "learning_rate": 1.9997418477541477e-05, + "loss": 3.6022, + "step": 924 + }, + { + "epoch": 0.07416167244593214, + "grad_norm": 3.50477933883667, + "learning_rate": 1.999738889142072e-05, + "loss": 3.4524, + "step": 925 + }, + { + "epoch": 0.07424184722695476, + "grad_norm": 3.8227062225341797, + "learning_rate": 1.999735913674751e-05, + "loss": 3.4948, + "step": 926 + }, + { + "epoch": 0.07432202200797738, + "grad_norm": 3.2662062644958496, + "learning_rate": 1.9997329213522347e-05, + "loss": 3.7848, + "step": 927 + }, + { + "epoch": 0.07440219678900002, + "grad_norm": 3.685037612915039, + "learning_rate": 1.999729912174574e-05, + "loss": 3.521, + "step": 928 + }, + { + "epoch": 0.07448237157002265, + "grad_norm": 3.278874397277832, + "learning_rate": 1.999726886141819e-05, + "loss": 3.568, + "step": 929 + }, + { + "epoch": 0.07456254635104528, + "grad_norm": 3.4870665073394775, + "learning_rate": 1.999723843254021e-05, + "loss": 3.606, + "step": 930 + }, + { + "epoch": 0.07464272113206791, + "grad_norm": 3.2301645278930664, + "learning_rate": 1.9997207835112315e-05, + "loss": 3.5557, + "step": 931 + }, + { + "epoch": 0.07472289591309053, + "grad_norm": 3.1822967529296875, + "learning_rate": 1.999717706913502e-05, + "loss": 3.5996, + "step": 932 + }, + { + "epoch": 0.07480307069411317, + "grad_norm": 3.3657948970794678, + "learning_rate": 1.9997146134608844e-05, + "loss": 3.5491, + "step": 933 + }, + { + "epoch": 0.07488324547513579, + "grad_norm": 3.339327812194824, + "learning_rate": 1.9997115031534308e-05, + "loss": 3.5862, + "step": 934 + }, + { + "epoch": 0.07496342025615843, + "grad_norm": 3.1597721576690674, + "learning_rate": 1.999708375991193e-05, + "loss": 3.6782, + "step": 935 + }, + { + "epoch": 0.07504359503718105, + "grad_norm": 3.602196455001831, + "learning_rate": 1.999705231974225e-05, + "loss": 3.5183, + "step": 936 + }, + { + "epoch": 0.07512376981820368, + "grad_norm": 3.1557846069335938, + "learning_rate": 1.999702071102579e-05, + "loss": 3.6274, + "step": 937 + }, + { + "epoch": 0.07520394459922632, + "grad_norm": 3.2448678016662598, + "learning_rate": 1.999698893376308e-05, + "loss": 3.5754, + "step": 938 + }, + { + "epoch": 0.07528411938024894, + "grad_norm": 3.3709611892700195, + "learning_rate": 1.999695698795466e-05, + "loss": 3.6127, + "step": 939 + }, + { + "epoch": 0.07536429416127158, + "grad_norm": 3.3437490463256836, + "learning_rate": 1.999692487360107e-05, + "loss": 3.5459, + "step": 940 + }, + { + "epoch": 0.0754444689422942, + "grad_norm": 3.1302154064178467, + "learning_rate": 1.999689259070285e-05, + "loss": 3.621, + "step": 941 + }, + { + "epoch": 0.07552464372331683, + "grad_norm": 3.4473876953125, + "learning_rate": 1.9996860139260547e-05, + "loss": 3.662, + "step": 942 + }, + { + "epoch": 0.07560481850433946, + "grad_norm": 3.018497943878174, + "learning_rate": 1.9996827519274702e-05, + "loss": 3.5916, + "step": 943 + }, + { + "epoch": 0.07568499328536209, + "grad_norm": 3.145587205886841, + "learning_rate": 1.999679473074587e-05, + "loss": 3.5636, + "step": 944 + }, + { + "epoch": 0.07576516806638472, + "grad_norm": 3.4863686561584473, + "learning_rate": 1.99967617736746e-05, + "loss": 3.5016, + "step": 945 + }, + { + "epoch": 0.07584534284740735, + "grad_norm": 3.6812143325805664, + "learning_rate": 1.999672864806145e-05, + "loss": 3.5173, + "step": 946 + }, + { + "epoch": 0.07592551762842997, + "grad_norm": 3.4606573581695557, + "learning_rate": 1.9996695353906978e-05, + "loss": 3.463, + "step": 947 + }, + { + "epoch": 0.07600569240945261, + "grad_norm": 3.721071481704712, + "learning_rate": 1.9996661891211747e-05, + "loss": 3.6428, + "step": 948 + }, + { + "epoch": 0.07608586719047523, + "grad_norm": 3.52695894241333, + "learning_rate": 1.999662825997632e-05, + "loss": 3.6023, + "step": 949 + }, + { + "epoch": 0.07616604197149787, + "grad_norm": 3.0612258911132812, + "learning_rate": 1.9996594460201262e-05, + "loss": 3.582, + "step": 950 + }, + { + "epoch": 0.0762462167525205, + "grad_norm": 3.1965274810791016, + "learning_rate": 1.9996560491887145e-05, + "loss": 3.6345, + "step": 951 + }, + { + "epoch": 0.07632639153354312, + "grad_norm": 3.0380666255950928, + "learning_rate": 1.9996526355034543e-05, + "loss": 3.5708, + "step": 952 + }, + { + "epoch": 0.07640656631456576, + "grad_norm": 3.607879400253296, + "learning_rate": 1.9996492049644027e-05, + "loss": 3.5344, + "step": 953 + }, + { + "epoch": 0.07648674109558838, + "grad_norm": 3.4285500049591064, + "learning_rate": 1.999645757571618e-05, + "loss": 3.6403, + "step": 954 + }, + { + "epoch": 0.076566915876611, + "grad_norm": 3.700411319732666, + "learning_rate": 1.999642293325158e-05, + "loss": 3.546, + "step": 955 + }, + { + "epoch": 0.07664709065763364, + "grad_norm": 2.941878080368042, + "learning_rate": 1.9996388122250816e-05, + "loss": 3.5254, + "step": 956 + }, + { + "epoch": 0.07672726543865627, + "grad_norm": 3.297187328338623, + "learning_rate": 1.999635314271447e-05, + "loss": 3.479, + "step": 957 + }, + { + "epoch": 0.0768074402196789, + "grad_norm": 3.654954433441162, + "learning_rate": 1.999631799464313e-05, + "loss": 3.5594, + "step": 958 + }, + { + "epoch": 0.07688761500070153, + "grad_norm": 3.186241626739502, + "learning_rate": 1.9996282678037393e-05, + "loss": 3.535, + "step": 959 + }, + { + "epoch": 0.07696778978172415, + "grad_norm": 3.052255153656006, + "learning_rate": 1.9996247192897854e-05, + "loss": 3.5175, + "step": 960 + }, + { + "epoch": 0.07704796456274679, + "grad_norm": 3.533108711242676, + "learning_rate": 1.9996211539225108e-05, + "loss": 3.5764, + "step": 961 + }, + { + "epoch": 0.07712813934376941, + "grad_norm": 3.2217044830322266, + "learning_rate": 1.999617571701976e-05, + "loss": 3.6199, + "step": 962 + }, + { + "epoch": 0.07720831412479205, + "grad_norm": 3.413069725036621, + "learning_rate": 1.9996139726282414e-05, + "loss": 3.5254, + "step": 963 + }, + { + "epoch": 0.07728848890581468, + "grad_norm": 3.347822904586792, + "learning_rate": 1.9996103567013675e-05, + "loss": 3.5336, + "step": 964 + }, + { + "epoch": 0.0773686636868373, + "grad_norm": 3.1913890838623047, + "learning_rate": 1.999606723921415e-05, + "loss": 3.6471, + "step": 965 + }, + { + "epoch": 0.07744883846785994, + "grad_norm": 3.1622889041900635, + "learning_rate": 1.9996030742884458e-05, + "loss": 3.4202, + "step": 966 + }, + { + "epoch": 0.07752901324888256, + "grad_norm": 3.2556209564208984, + "learning_rate": 1.9995994078025206e-05, + "loss": 2.8064, + "step": 967 + }, + { + "epoch": 0.0776091880299052, + "grad_norm": 2.9313602447509766, + "learning_rate": 1.9995957244637022e-05, + "loss": 3.4513, + "step": 968 + }, + { + "epoch": 0.07768936281092782, + "grad_norm": 3.2965564727783203, + "learning_rate": 1.9995920242720517e-05, + "loss": 3.4996, + "step": 969 + }, + { + "epoch": 0.07776953759195045, + "grad_norm": 3.5333592891693115, + "learning_rate": 1.999588307227632e-05, + "loss": 3.5967, + "step": 970 + }, + { + "epoch": 0.07784971237297308, + "grad_norm": 3.0491037368774414, + "learning_rate": 1.999584573330506e-05, + "loss": 3.606, + "step": 971 + }, + { + "epoch": 0.07792988715399571, + "grad_norm": 3.0874195098876953, + "learning_rate": 1.9995808225807364e-05, + "loss": 3.5179, + "step": 972 + }, + { + "epoch": 0.07801006193501835, + "grad_norm": 3.4330170154571533, + "learning_rate": 1.9995770549783865e-05, + "loss": 3.6255, + "step": 973 + }, + { + "epoch": 0.07809023671604097, + "grad_norm": 3.2541003227233887, + "learning_rate": 1.9995732705235193e-05, + "loss": 3.4873, + "step": 974 + }, + { + "epoch": 0.0781704114970636, + "grad_norm": 3.0658514499664307, + "learning_rate": 1.9995694692161995e-05, + "loss": 3.5522, + "step": 975 + }, + { + "epoch": 0.07825058627808623, + "grad_norm": 3.0803277492523193, + "learning_rate": 1.9995656510564903e-05, + "loss": 3.5542, + "step": 976 + }, + { + "epoch": 0.07833076105910886, + "grad_norm": 3.1886305809020996, + "learning_rate": 1.999561816044457e-05, + "loss": 3.4757, + "step": 977 + }, + { + "epoch": 0.0784109358401315, + "grad_norm": 3.154271364212036, + "learning_rate": 1.9995579641801633e-05, + "loss": 3.5425, + "step": 978 + }, + { + "epoch": 0.07849111062115412, + "grad_norm": 3.100525379180908, + "learning_rate": 1.9995540954636748e-05, + "loss": 3.5427, + "step": 979 + }, + { + "epoch": 0.07857128540217674, + "grad_norm": 3.0537590980529785, + "learning_rate": 1.9995502098950566e-05, + "loss": 3.5576, + "step": 980 + }, + { + "epoch": 0.07865146018319938, + "grad_norm": 2.848907232284546, + "learning_rate": 1.999546307474374e-05, + "loss": 3.5616, + "step": 981 + }, + { + "epoch": 0.078731634964222, + "grad_norm": 2.6782970428466797, + "learning_rate": 1.999542388201693e-05, + "loss": 3.4479, + "step": 982 + }, + { + "epoch": 0.07881180974524463, + "grad_norm": 3.039630174636841, + "learning_rate": 1.9995384520770796e-05, + "loss": 3.5139, + "step": 983 + }, + { + "epoch": 0.07889198452626726, + "grad_norm": 3.1018764972686768, + "learning_rate": 1.9995344991006002e-05, + "loss": 3.5796, + "step": 984 + }, + { + "epoch": 0.07897215930728989, + "grad_norm": 2.8964767456054688, + "learning_rate": 1.999530529272321e-05, + "loss": 3.4988, + "step": 985 + }, + { + "epoch": 0.07905233408831253, + "grad_norm": 2.9227821826934814, + "learning_rate": 1.99952654259231e-05, + "loss": 3.4469, + "step": 986 + }, + { + "epoch": 0.07913250886933515, + "grad_norm": 2.850417375564575, + "learning_rate": 1.9995225390606333e-05, + "loss": 3.4857, + "step": 987 + }, + { + "epoch": 0.07921268365035777, + "grad_norm": 2.472820997238159, + "learning_rate": 1.999518518677359e-05, + "loss": 2.7714, + "step": 988 + }, + { + "epoch": 0.07929285843138041, + "grad_norm": 3.547386884689331, + "learning_rate": 1.9995144814425548e-05, + "loss": 3.4542, + "step": 989 + }, + { + "epoch": 0.07937303321240304, + "grad_norm": 2.131681203842163, + "learning_rate": 1.999510427356289e-05, + "loss": 2.7639, + "step": 990 + }, + { + "epoch": 0.07945320799342567, + "grad_norm": 3.2805490493774414, + "learning_rate": 1.999506356418629e-05, + "loss": 3.5472, + "step": 991 + }, + { + "epoch": 0.0795333827744483, + "grad_norm": 3.1687772274017334, + "learning_rate": 1.9995022686296446e-05, + "loss": 3.3832, + "step": 992 + }, + { + "epoch": 0.07961355755547092, + "grad_norm": 2.772780656814575, + "learning_rate": 1.999498163989404e-05, + "loss": 3.4425, + "step": 993 + }, + { + "epoch": 0.07969373233649356, + "grad_norm": 3.4435534477233887, + "learning_rate": 1.9994940424979767e-05, + "loss": 3.5615, + "step": 994 + }, + { + "epoch": 0.07977390711751618, + "grad_norm": 3.197728157043457, + "learning_rate": 1.999489904155432e-05, + "loss": 3.5018, + "step": 995 + }, + { + "epoch": 0.07985408189853882, + "grad_norm": 2.8434808254241943, + "learning_rate": 1.99948574896184e-05, + "loss": 3.5328, + "step": 996 + }, + { + "epoch": 0.07993425667956144, + "grad_norm": 2.8212735652923584, + "learning_rate": 1.99948157691727e-05, + "loss": 3.3975, + "step": 997 + }, + { + "epoch": 0.08001443146058407, + "grad_norm": 3.3418493270874023, + "learning_rate": 1.9994773880217936e-05, + "loss": 3.4076, + "step": 998 + }, + { + "epoch": 0.0800946062416067, + "grad_norm": 3.333120346069336, + "learning_rate": 1.9994731822754803e-05, + "loss": 3.5875, + "step": 999 + }, + { + "epoch": 0.08017478102262933, + "grad_norm": 3.294674873352051, + "learning_rate": 1.9994689596784012e-05, + "loss": 3.3627, + "step": 1000 + }, + { + "epoch": 0.08025495580365197, + "grad_norm": 2.8912737369537354, + "learning_rate": 1.999464720230628e-05, + "loss": 3.5031, + "step": 1001 + }, + { + "epoch": 0.08033513058467459, + "grad_norm": 2.1793015003204346, + "learning_rate": 1.9994604639322316e-05, + "loss": 2.7551, + "step": 1002 + }, + { + "epoch": 0.08041530536569721, + "grad_norm": 2.8094377517700195, + "learning_rate": 1.9994561907832843e-05, + "loss": 3.4365, + "step": 1003 + }, + { + "epoch": 0.08049548014671985, + "grad_norm": 3.0215861797332764, + "learning_rate": 1.9994519007838577e-05, + "loss": 3.4813, + "step": 1004 + }, + { + "epoch": 0.08057565492774248, + "grad_norm": 2.9410006999969482, + "learning_rate": 1.9994475939340244e-05, + "loss": 3.5518, + "step": 1005 + }, + { + "epoch": 0.08065582970876511, + "grad_norm": 2.877200126647949, + "learning_rate": 1.9994432702338568e-05, + "loss": 3.5447, + "step": 1006 + }, + { + "epoch": 0.08073600448978774, + "grad_norm": 3.0519394874572754, + "learning_rate": 1.9994389296834276e-05, + "loss": 3.3915, + "step": 1007 + }, + { + "epoch": 0.08081617927081036, + "grad_norm": 3.2028729915618896, + "learning_rate": 1.999434572282811e-05, + "loss": 3.4962, + "step": 1008 + }, + { + "epoch": 0.080896354051833, + "grad_norm": 3.0041558742523193, + "learning_rate": 1.999430198032079e-05, + "loss": 3.5662, + "step": 1009 + }, + { + "epoch": 0.08097652883285562, + "grad_norm": 3.404007911682129, + "learning_rate": 1.999425806931307e-05, + "loss": 3.5, + "step": 1010 + }, + { + "epoch": 0.08105670361387826, + "grad_norm": 2.648655891418457, + "learning_rate": 1.9994213989805675e-05, + "loss": 3.3844, + "step": 1011 + }, + { + "epoch": 0.08113687839490089, + "grad_norm": 2.7420387268066406, + "learning_rate": 1.999416974179935e-05, + "loss": 3.5482, + "step": 1012 + }, + { + "epoch": 0.08121705317592351, + "grad_norm": 3.1986794471740723, + "learning_rate": 1.9994125325294856e-05, + "loss": 3.4346, + "step": 1013 + }, + { + "epoch": 0.08129722795694615, + "grad_norm": 3.080667495727539, + "learning_rate": 1.9994080740292926e-05, + "loss": 3.4863, + "step": 1014 + }, + { + "epoch": 0.08137740273796877, + "grad_norm": 3.129408359527588, + "learning_rate": 1.9994035986794316e-05, + "loss": 3.5347, + "step": 1015 + }, + { + "epoch": 0.0814575775189914, + "grad_norm": 2.962331771850586, + "learning_rate": 1.9993991064799783e-05, + "loss": 3.6, + "step": 1016 + }, + { + "epoch": 0.08153775230001403, + "grad_norm": 2.798208236694336, + "learning_rate": 1.999394597431008e-05, + "loss": 3.4324, + "step": 1017 + }, + { + "epoch": 0.08161792708103666, + "grad_norm": 3.1316921710968018, + "learning_rate": 1.9993900715325973e-05, + "loss": 3.4149, + "step": 1018 + }, + { + "epoch": 0.0816981018620593, + "grad_norm": 3.1187808513641357, + "learning_rate": 1.999385528784822e-05, + "loss": 3.3979, + "step": 1019 + }, + { + "epoch": 0.08177827664308192, + "grad_norm": 2.9431686401367188, + "learning_rate": 1.9993809691877592e-05, + "loss": 3.5269, + "step": 1020 + }, + { + "epoch": 0.08185845142410454, + "grad_norm": 2.740689277648926, + "learning_rate": 1.9993763927414853e-05, + "loss": 3.5732, + "step": 1021 + }, + { + "epoch": 0.08193862620512718, + "grad_norm": 3.03684139251709, + "learning_rate": 1.9993717994460778e-05, + "loss": 3.4219, + "step": 1022 + }, + { + "epoch": 0.0820188009861498, + "grad_norm": 3.1059415340423584, + "learning_rate": 1.999367189301614e-05, + "loss": 3.4615, + "step": 1023 + }, + { + "epoch": 0.08209897576717244, + "grad_norm": 3.007650852203369, + "learning_rate": 1.999362562308171e-05, + "loss": 3.4423, + "step": 1024 + }, + { + "epoch": 0.08217915054819507, + "grad_norm": 2.8579084873199463, + "learning_rate": 1.9993579184658282e-05, + "loss": 3.4812, + "step": 1025 + }, + { + "epoch": 0.08225932532921769, + "grad_norm": 3.033511161804199, + "learning_rate": 1.9993532577746627e-05, + "loss": 3.4053, + "step": 1026 + }, + { + "epoch": 0.08233950011024033, + "grad_norm": 3.1892552375793457, + "learning_rate": 1.9993485802347534e-05, + "loss": 3.5048, + "step": 1027 + }, + { + "epoch": 0.08241967489126295, + "grad_norm": 3.26967716217041, + "learning_rate": 1.9993438858461795e-05, + "loss": 3.3962, + "step": 1028 + }, + { + "epoch": 0.08249984967228559, + "grad_norm": 2.7852063179016113, + "learning_rate": 1.9993391746090197e-05, + "loss": 3.5234, + "step": 1029 + }, + { + "epoch": 0.08258002445330821, + "grad_norm": 3.046196460723877, + "learning_rate": 1.999334446523354e-05, + "loss": 3.6355, + "step": 1030 + }, + { + "epoch": 0.08266019923433084, + "grad_norm": 2.3845131397247314, + "learning_rate": 1.999329701589261e-05, + "loss": 2.7428, + "step": 1031 + }, + { + "epoch": 0.08274037401535347, + "grad_norm": 2.953970432281494, + "learning_rate": 1.999324939806822e-05, + "loss": 3.4137, + "step": 1032 + }, + { + "epoch": 0.0828205487963761, + "grad_norm": 2.7767481803894043, + "learning_rate": 1.9993201611761164e-05, + "loss": 3.4845, + "step": 1033 + }, + { + "epoch": 0.08290072357739874, + "grad_norm": 3.0759217739105225, + "learning_rate": 1.9993153656972254e-05, + "loss": 3.4413, + "step": 1034 + }, + { + "epoch": 0.08298089835842136, + "grad_norm": 3.26977276802063, + "learning_rate": 1.9993105533702294e-05, + "loss": 3.4397, + "step": 1035 + }, + { + "epoch": 0.08306107313944398, + "grad_norm": 3.1410093307495117, + "learning_rate": 1.9993057241952094e-05, + "loss": 3.4867, + "step": 1036 + }, + { + "epoch": 0.08314124792046662, + "grad_norm": 2.9807026386260986, + "learning_rate": 1.9993008781722474e-05, + "loss": 3.5226, + "step": 1037 + }, + { + "epoch": 0.08322142270148924, + "grad_norm": 2.8735594749450684, + "learning_rate": 1.9992960153014246e-05, + "loss": 3.4359, + "step": 1038 + }, + { + "epoch": 0.08330159748251188, + "grad_norm": 2.2148826122283936, + "learning_rate": 1.9992911355828234e-05, + "loss": 2.7449, + "step": 1039 + }, + { + "epoch": 0.0833817722635345, + "grad_norm": 3.1327192783355713, + "learning_rate": 1.9992862390165255e-05, + "loss": 3.4474, + "step": 1040 + }, + { + "epoch": 0.08346194704455713, + "grad_norm": 1.8916212320327759, + "learning_rate": 1.999281325602614e-05, + "loss": 2.7329, + "step": 1041 + }, + { + "epoch": 0.08354212182557977, + "grad_norm": 3.9809679985046387, + "learning_rate": 1.9992763953411715e-05, + "loss": 3.5127, + "step": 1042 + }, + { + "epoch": 0.08362229660660239, + "grad_norm": 2.902052879333496, + "learning_rate": 1.999271448232281e-05, + "loss": 3.499, + "step": 1043 + }, + { + "epoch": 0.08370247138762502, + "grad_norm": 3.0791444778442383, + "learning_rate": 1.9992664842760266e-05, + "loss": 3.4357, + "step": 1044 + }, + { + "epoch": 0.08378264616864765, + "grad_norm": 3.04748272895813, + "learning_rate": 1.999261503472491e-05, + "loss": 3.5276, + "step": 1045 + }, + { + "epoch": 0.08386282094967028, + "grad_norm": 2.788569927215576, + "learning_rate": 1.9992565058217584e-05, + "loss": 3.4266, + "step": 1046 + }, + { + "epoch": 0.08394299573069292, + "grad_norm": 3.0969550609588623, + "learning_rate": 1.9992514913239137e-05, + "loss": 3.4497, + "step": 1047 + }, + { + "epoch": 0.08402317051171554, + "grad_norm": 3.121492385864258, + "learning_rate": 1.999246459979041e-05, + "loss": 3.5206, + "step": 1048 + }, + { + "epoch": 0.08410334529273816, + "grad_norm": 2.9673492908477783, + "learning_rate": 1.999241411787225e-05, + "loss": 3.4276, + "step": 1049 + }, + { + "epoch": 0.0841835200737608, + "grad_norm": 2.9488697052001953, + "learning_rate": 1.999236346748551e-05, + "loss": 3.4137, + "step": 1050 + }, + { + "epoch": 0.08426369485478342, + "grad_norm": 2.900768518447876, + "learning_rate": 1.9992312648631046e-05, + "loss": 3.4523, + "step": 1051 + }, + { + "epoch": 0.08434386963580606, + "grad_norm": 2.872584342956543, + "learning_rate": 1.999226166130971e-05, + "loss": 3.4506, + "step": 1052 + }, + { + "epoch": 0.08442404441682869, + "grad_norm": 2.858409881591797, + "learning_rate": 1.9992210505522364e-05, + "loss": 3.5669, + "step": 1053 + }, + { + "epoch": 0.08450421919785131, + "grad_norm": 2.8291101455688477, + "learning_rate": 1.999215918126987e-05, + "loss": 3.5255, + "step": 1054 + }, + { + "epoch": 0.08458439397887395, + "grad_norm": 3.2767751216888428, + "learning_rate": 1.9992107688553097e-05, + "loss": 3.5122, + "step": 1055 + }, + { + "epoch": 0.08466456875989657, + "grad_norm": 3.079619884490967, + "learning_rate": 1.9992056027372904e-05, + "loss": 3.4573, + "step": 1056 + }, + { + "epoch": 0.08474474354091921, + "grad_norm": 3.15585994720459, + "learning_rate": 1.999200419773017e-05, + "loss": 3.4895, + "step": 1057 + }, + { + "epoch": 0.08482491832194183, + "grad_norm": 2.6735756397247314, + "learning_rate": 1.999195219962577e-05, + "loss": 3.5409, + "step": 1058 + }, + { + "epoch": 0.08490509310296446, + "grad_norm": 2.73738431930542, + "learning_rate": 1.9991900033060574e-05, + "loss": 3.3625, + "step": 1059 + }, + { + "epoch": 0.0849852678839871, + "grad_norm": 2.1475346088409424, + "learning_rate": 1.999184769803547e-05, + "loss": 2.7228, + "step": 1060 + }, + { + "epoch": 0.08506544266500972, + "grad_norm": 3.3699421882629395, + "learning_rate": 1.999179519455133e-05, + "loss": 3.4731, + "step": 1061 + }, + { + "epoch": 0.08514561744603236, + "grad_norm": 3.120737075805664, + "learning_rate": 1.9991742522609044e-05, + "loss": 3.476, + "step": 1062 + }, + { + "epoch": 0.08522579222705498, + "grad_norm": 2.8785486221313477, + "learning_rate": 1.9991689682209504e-05, + "loss": 3.4016, + "step": 1063 + }, + { + "epoch": 0.0853059670080776, + "grad_norm": 3.165530204772949, + "learning_rate": 1.9991636673353597e-05, + "loss": 3.4178, + "step": 1064 + }, + { + "epoch": 0.08538614178910024, + "grad_norm": 2.7294702529907227, + "learning_rate": 1.9991583496042217e-05, + "loss": 3.3623, + "step": 1065 + }, + { + "epoch": 0.08546631657012287, + "grad_norm": 2.895212411880493, + "learning_rate": 1.9991530150276258e-05, + "loss": 3.489, + "step": 1066 + }, + { + "epoch": 0.0855464913511455, + "grad_norm": 2.991039514541626, + "learning_rate": 1.9991476636056624e-05, + "loss": 3.4369, + "step": 1067 + }, + { + "epoch": 0.08562666613216813, + "grad_norm": 2.7246639728546143, + "learning_rate": 1.9991422953384215e-05, + "loss": 3.3574, + "step": 1068 + }, + { + "epoch": 0.08570684091319075, + "grad_norm": 2.9088134765625, + "learning_rate": 1.9991369102259938e-05, + "loss": 3.5555, + "step": 1069 + }, + { + "epoch": 0.08578701569421339, + "grad_norm": 3.0024759769439697, + "learning_rate": 1.9991315082684694e-05, + "loss": 3.5172, + "step": 1070 + }, + { + "epoch": 0.08586719047523601, + "grad_norm": 2.6072285175323486, + "learning_rate": 1.9991260894659404e-05, + "loss": 3.3992, + "step": 1071 + }, + { + "epoch": 0.08594736525625864, + "grad_norm": 2.973581075668335, + "learning_rate": 1.9991206538184978e-05, + "loss": 3.4774, + "step": 1072 + }, + { + "epoch": 0.08602754003728127, + "grad_norm": 3.119964599609375, + "learning_rate": 1.9991152013262325e-05, + "loss": 3.5796, + "step": 1073 + }, + { + "epoch": 0.0861077148183039, + "grad_norm": 2.7816596031188965, + "learning_rate": 1.9991097319892375e-05, + "loss": 3.4288, + "step": 1074 + }, + { + "epoch": 0.08618788959932654, + "grad_norm": 2.955848217010498, + "learning_rate": 1.9991042458076042e-05, + "loss": 3.5272, + "step": 1075 + }, + { + "epoch": 0.08626806438034916, + "grad_norm": 3.0845136642456055, + "learning_rate": 1.9990987427814258e-05, + "loss": 3.4614, + "step": 1076 + }, + { + "epoch": 0.08634823916137178, + "grad_norm": 2.7191150188446045, + "learning_rate": 1.9990932229107944e-05, + "loss": 3.4891, + "step": 1077 + }, + { + "epoch": 0.08642841394239442, + "grad_norm": 2.1010730266571045, + "learning_rate": 1.9990876861958037e-05, + "loss": 2.6952, + "step": 1078 + }, + { + "epoch": 0.08650858872341705, + "grad_norm": 3.0517544746398926, + "learning_rate": 1.9990821326365463e-05, + "loss": 3.3801, + "step": 1079 + }, + { + "epoch": 0.08658876350443968, + "grad_norm": 2.7581610679626465, + "learning_rate": 1.999076562233117e-05, + "loss": 3.451, + "step": 1080 + }, + { + "epoch": 0.08666893828546231, + "grad_norm": 2.6990597248077393, + "learning_rate": 1.9990709749856082e-05, + "loss": 3.5023, + "step": 1081 + }, + { + "epoch": 0.08674911306648493, + "grad_norm": 2.6615753173828125, + "learning_rate": 1.999065370894115e-05, + "loss": 3.535, + "step": 1082 + }, + { + "epoch": 0.08682928784750757, + "grad_norm": 2.7099411487579346, + "learning_rate": 1.999059749958732e-05, + "loss": 3.4446, + "step": 1083 + }, + { + "epoch": 0.08690946262853019, + "grad_norm": 2.786587715148926, + "learning_rate": 1.9990541121795535e-05, + "loss": 3.4325, + "step": 1084 + }, + { + "epoch": 0.08698963740955283, + "grad_norm": 2.817314386367798, + "learning_rate": 1.999048457556675e-05, + "loss": 3.5464, + "step": 1085 + }, + { + "epoch": 0.08706981219057545, + "grad_norm": 2.6488068103790283, + "learning_rate": 1.9990427860901914e-05, + "loss": 3.5345, + "step": 1086 + }, + { + "epoch": 0.08714998697159808, + "grad_norm": 2.8325307369232178, + "learning_rate": 1.9990370977801985e-05, + "loss": 3.387, + "step": 1087 + }, + { + "epoch": 0.08723016175262072, + "grad_norm": 2.78718900680542, + "learning_rate": 1.9990313926267922e-05, + "loss": 3.2715, + "step": 1088 + }, + { + "epoch": 0.08731033653364334, + "grad_norm": 2.659316301345825, + "learning_rate": 1.999025670630069e-05, + "loss": 3.4407, + "step": 1089 + }, + { + "epoch": 0.08739051131466598, + "grad_norm": 2.6414761543273926, + "learning_rate": 1.9990199317901248e-05, + "loss": 3.4133, + "step": 1090 + }, + { + "epoch": 0.0874706860956886, + "grad_norm": 2.9420320987701416, + "learning_rate": 1.999014176107057e-05, + "loss": 3.4288, + "step": 1091 + }, + { + "epoch": 0.08755086087671123, + "grad_norm": 2.7176671028137207, + "learning_rate": 1.999008403580962e-05, + "loss": 3.4655, + "step": 1092 + }, + { + "epoch": 0.08763103565773386, + "grad_norm": 2.640207290649414, + "learning_rate": 1.9990026142119373e-05, + "loss": 3.408, + "step": 1093 + }, + { + "epoch": 0.08771121043875649, + "grad_norm": 2.7807633876800537, + "learning_rate": 1.9989968080000808e-05, + "loss": 3.4719, + "step": 1094 + }, + { + "epoch": 0.08779138521977913, + "grad_norm": 2.812588691711426, + "learning_rate": 1.99899098494549e-05, + "loss": 3.4362, + "step": 1095 + }, + { + "epoch": 0.08787156000080175, + "grad_norm": 2.628417730331421, + "learning_rate": 1.9989851450482637e-05, + "loss": 3.4501, + "step": 1096 + }, + { + "epoch": 0.08795173478182437, + "grad_norm": 2.8441214561462402, + "learning_rate": 1.998979288308499e-05, + "loss": 3.4844, + "step": 1097 + }, + { + "epoch": 0.08803190956284701, + "grad_norm": 2.627387285232544, + "learning_rate": 1.9989734147262965e-05, + "loss": 3.4512, + "step": 1098 + }, + { + "epoch": 0.08811208434386963, + "grad_norm": 2.8929386138916016, + "learning_rate": 1.998967524301754e-05, + "loss": 3.3217, + "step": 1099 + }, + { + "epoch": 0.08819225912489226, + "grad_norm": 2.741805076599121, + "learning_rate": 1.9989616170349713e-05, + "loss": 3.5137, + "step": 1100 + }, + { + "epoch": 0.0882724339059149, + "grad_norm": 2.8227455615997314, + "learning_rate": 1.9989556929260475e-05, + "loss": 3.4175, + "step": 1101 + }, + { + "epoch": 0.08835260868693752, + "grad_norm": 2.769559621810913, + "learning_rate": 1.998949751975083e-05, + "loss": 3.5461, + "step": 1102 + }, + { + "epoch": 0.08843278346796016, + "grad_norm": 2.6657018661499023, + "learning_rate": 1.9989437941821775e-05, + "loss": 3.4802, + "step": 1103 + }, + { + "epoch": 0.08851295824898278, + "grad_norm": 2.565366744995117, + "learning_rate": 1.9989378195474322e-05, + "loss": 3.3733, + "step": 1104 + }, + { + "epoch": 0.0885931330300054, + "grad_norm": 2.5993964672088623, + "learning_rate": 1.9989318280709468e-05, + "loss": 3.3879, + "step": 1105 + }, + { + "epoch": 0.08867330781102804, + "grad_norm": 2.6624739170074463, + "learning_rate": 1.998925819752823e-05, + "loss": 3.457, + "step": 1106 + }, + { + "epoch": 0.08875348259205067, + "grad_norm": 1.7813247442245483, + "learning_rate": 1.998919794593162e-05, + "loss": 2.6862, + "step": 1107 + }, + { + "epoch": 0.0888336573730733, + "grad_norm": 2.8757896423339844, + "learning_rate": 1.9989137525920652e-05, + "loss": 3.3462, + "step": 1108 + }, + { + "epoch": 0.08891383215409593, + "grad_norm": 2.8118491172790527, + "learning_rate": 1.9989076937496346e-05, + "loss": 3.2435, + "step": 1109 + }, + { + "epoch": 0.08899400693511855, + "grad_norm": 2.574720621109009, + "learning_rate": 1.9989016180659725e-05, + "loss": 3.4532, + "step": 1110 + }, + { + "epoch": 0.08907418171614119, + "grad_norm": 2.933812141418457, + "learning_rate": 1.998895525541181e-05, + "loss": 3.4741, + "step": 1111 + }, + { + "epoch": 0.08915435649716381, + "grad_norm": 2.662581205368042, + "learning_rate": 1.9988894161753628e-05, + "loss": 3.5776, + "step": 1112 + }, + { + "epoch": 0.08923453127818645, + "grad_norm": 2.72416353225708, + "learning_rate": 1.9988832899686214e-05, + "loss": 3.4379, + "step": 1113 + }, + { + "epoch": 0.08931470605920908, + "grad_norm": 2.6875181198120117, + "learning_rate": 1.9988771469210596e-05, + "loss": 3.4833, + "step": 1114 + }, + { + "epoch": 0.0893948808402317, + "grad_norm": 2.8989124298095703, + "learning_rate": 1.9988709870327812e-05, + "loss": 3.4822, + "step": 1115 + }, + { + "epoch": 0.08947505562125434, + "grad_norm": 2.5532772541046143, + "learning_rate": 1.99886481030389e-05, + "loss": 3.4097, + "step": 1116 + }, + { + "epoch": 0.08955523040227696, + "grad_norm": 2.8583784103393555, + "learning_rate": 1.9988586167344903e-05, + "loss": 3.2847, + "step": 1117 + }, + { + "epoch": 0.0896354051832996, + "grad_norm": 3.0705671310424805, + "learning_rate": 1.9988524063246865e-05, + "loss": 3.4189, + "step": 1118 + }, + { + "epoch": 0.08971557996432222, + "grad_norm": 2.631422281265259, + "learning_rate": 1.998846179074583e-05, + "loss": 3.3516, + "step": 1119 + }, + { + "epoch": 0.08979575474534485, + "grad_norm": 2.7575159072875977, + "learning_rate": 1.9988399349842843e-05, + "loss": 3.4487, + "step": 1120 + }, + { + "epoch": 0.08987592952636748, + "grad_norm": 2.6781046390533447, + "learning_rate": 1.998833674053897e-05, + "loss": 3.4187, + "step": 1121 + }, + { + "epoch": 0.08995610430739011, + "grad_norm": 3.020677089691162, + "learning_rate": 1.998827396283526e-05, + "loss": 3.3929, + "step": 1122 + }, + { + "epoch": 0.09003627908841275, + "grad_norm": 2.927504301071167, + "learning_rate": 1.9988211016732766e-05, + "loss": 3.5339, + "step": 1123 + }, + { + "epoch": 0.09011645386943537, + "grad_norm": 2.7965824604034424, + "learning_rate": 1.998814790223256e-05, + "loss": 3.4814, + "step": 1124 + }, + { + "epoch": 0.090196628650458, + "grad_norm": 2.8516793251037598, + "learning_rate": 1.9988084619335697e-05, + "loss": 3.3981, + "step": 1125 + }, + { + "epoch": 0.09027680343148063, + "grad_norm": 2.649077892303467, + "learning_rate": 1.998802116804325e-05, + "loss": 3.4573, + "step": 1126 + }, + { + "epoch": 0.09035697821250326, + "grad_norm": 2.8234009742736816, + "learning_rate": 1.9987957548356284e-05, + "loss": 3.4159, + "step": 1127 + }, + { + "epoch": 0.0904371529935259, + "grad_norm": 2.9950320720672607, + "learning_rate": 1.9987893760275875e-05, + "loss": 3.5602, + "step": 1128 + }, + { + "epoch": 0.09051732777454852, + "grad_norm": 2.861542224884033, + "learning_rate": 1.9987829803803097e-05, + "loss": 3.4899, + "step": 1129 + }, + { + "epoch": 0.09059750255557114, + "grad_norm": 2.7871179580688477, + "learning_rate": 1.9987765678939028e-05, + "loss": 3.4623, + "step": 1130 + }, + { + "epoch": 0.09067767733659378, + "grad_norm": 3.1959004402160645, + "learning_rate": 1.998770138568475e-05, + "loss": 3.4156, + "step": 1131 + }, + { + "epoch": 0.0907578521176164, + "grad_norm": 3.255955696105957, + "learning_rate": 1.998763692404135e-05, + "loss": 3.3566, + "step": 1132 + }, + { + "epoch": 0.09083802689863903, + "grad_norm": 2.5187320709228516, + "learning_rate": 1.9987572294009908e-05, + "loss": 3.4023, + "step": 1133 + }, + { + "epoch": 0.09091820167966166, + "grad_norm": 2.716796398162842, + "learning_rate": 1.9987507495591517e-05, + "loss": 3.3214, + "step": 1134 + }, + { + "epoch": 0.09099837646068429, + "grad_norm": 2.942685604095459, + "learning_rate": 1.998744252878727e-05, + "loss": 3.3536, + "step": 1135 + }, + { + "epoch": 0.09107855124170693, + "grad_norm": 2.7160892486572266, + "learning_rate": 1.9987377393598262e-05, + "loss": 3.3057, + "step": 1136 + }, + { + "epoch": 0.09115872602272955, + "grad_norm": 2.7863523960113525, + "learning_rate": 1.998731209002559e-05, + "loss": 3.4669, + "step": 1137 + }, + { + "epoch": 0.09123890080375217, + "grad_norm": 2.9600768089294434, + "learning_rate": 1.9987246618070356e-05, + "loss": 3.4099, + "step": 1138 + }, + { + "epoch": 0.09131907558477481, + "grad_norm": 2.936235189437866, + "learning_rate": 1.9987180977733668e-05, + "loss": 3.2732, + "step": 1139 + }, + { + "epoch": 0.09139925036579744, + "grad_norm": 2.7747836112976074, + "learning_rate": 1.9987115169016624e-05, + "loss": 3.443, + "step": 1140 + }, + { + "epoch": 0.09147942514682007, + "grad_norm": 2.4684898853302, + "learning_rate": 1.9987049191920337e-05, + "loss": 3.4708, + "step": 1141 + }, + { + "epoch": 0.0915595999278427, + "grad_norm": 2.7313966751098633, + "learning_rate": 1.9986983046445924e-05, + "loss": 3.4482, + "step": 1142 + }, + { + "epoch": 0.09163977470886532, + "grad_norm": 1.9970083236694336, + "learning_rate": 1.9986916732594497e-05, + "loss": 2.6862, + "step": 1143 + }, + { + "epoch": 0.09171994948988796, + "grad_norm": 1.906376838684082, + "learning_rate": 1.9986850250367174e-05, + "loss": 2.6608, + "step": 1144 + }, + { + "epoch": 0.09180012427091058, + "grad_norm": 2.696077346801758, + "learning_rate": 1.9986783599765073e-05, + "loss": 3.446, + "step": 1145 + }, + { + "epoch": 0.09188029905193322, + "grad_norm": 3.0208094120025635, + "learning_rate": 1.998671678078932e-05, + "loss": 3.4526, + "step": 1146 + }, + { + "epoch": 0.09196047383295584, + "grad_norm": 2.735882043838501, + "learning_rate": 1.998664979344104e-05, + "loss": 3.4935, + "step": 1147 + }, + { + "epoch": 0.09204064861397847, + "grad_norm": 2.506574869155884, + "learning_rate": 1.998658263772137e-05, + "loss": 3.3729, + "step": 1148 + }, + { + "epoch": 0.0921208233950011, + "grad_norm": 2.6323013305664062, + "learning_rate": 1.998651531363143e-05, + "loss": 3.4019, + "step": 1149 + }, + { + "epoch": 0.09220099817602373, + "grad_norm": 2.987034320831299, + "learning_rate": 1.9986447821172365e-05, + "loss": 3.5051, + "step": 1150 + }, + { + "epoch": 0.09228117295704637, + "grad_norm": 2.675057888031006, + "learning_rate": 1.998638016034531e-05, + "loss": 3.3776, + "step": 1151 + }, + { + "epoch": 0.09236134773806899, + "grad_norm": 2.5484695434570312, + "learning_rate": 1.99863123311514e-05, + "loss": 3.45, + "step": 1152 + }, + { + "epoch": 0.09244152251909162, + "grad_norm": 2.4683988094329834, + "learning_rate": 1.9986244333591792e-05, + "loss": 3.4893, + "step": 1153 + }, + { + "epoch": 0.09252169730011425, + "grad_norm": 2.5164108276367188, + "learning_rate": 1.9986176167667617e-05, + "loss": 3.5037, + "step": 1154 + }, + { + "epoch": 0.09260187208113688, + "grad_norm": 2.747535228729248, + "learning_rate": 1.9986107833380034e-05, + "loss": 3.5025, + "step": 1155 + }, + { + "epoch": 0.09268204686215951, + "grad_norm": 2.7311151027679443, + "learning_rate": 1.9986039330730192e-05, + "loss": 3.4617, + "step": 1156 + }, + { + "epoch": 0.09276222164318214, + "grad_norm": 2.6680173873901367, + "learning_rate": 1.9985970659719247e-05, + "loss": 3.3208, + "step": 1157 + }, + { + "epoch": 0.09284239642420476, + "grad_norm": 2.8419880867004395, + "learning_rate": 1.9985901820348354e-05, + "loss": 3.4414, + "step": 1158 + }, + { + "epoch": 0.0929225712052274, + "grad_norm": 2.835416316986084, + "learning_rate": 1.9985832812618678e-05, + "loss": 3.444, + "step": 1159 + }, + { + "epoch": 0.09300274598625002, + "grad_norm": 2.584167242050171, + "learning_rate": 1.998576363653138e-05, + "loss": 3.3762, + "step": 1160 + }, + { + "epoch": 0.09308292076727265, + "grad_norm": 2.519923448562622, + "learning_rate": 1.9985694292087624e-05, + "loss": 3.3379, + "step": 1161 + }, + { + "epoch": 0.09316309554829529, + "grad_norm": 3.334561824798584, + "learning_rate": 1.9985624779288587e-05, + "loss": 3.5417, + "step": 1162 + }, + { + "epoch": 0.09324327032931791, + "grad_norm": 3.0051374435424805, + "learning_rate": 1.998555509813543e-05, + "loss": 3.3593, + "step": 1163 + }, + { + "epoch": 0.09332344511034055, + "grad_norm": 2.432004451751709, + "learning_rate": 1.9985485248629333e-05, + "loss": 3.3904, + "step": 1164 + }, + { + "epoch": 0.09340361989136317, + "grad_norm": 2.543151617050171, + "learning_rate": 1.9985415230771478e-05, + "loss": 3.5094, + "step": 1165 + }, + { + "epoch": 0.0934837946723858, + "grad_norm": 2.9887919425964355, + "learning_rate": 1.9985345044563037e-05, + "loss": 3.3882, + "step": 1166 + }, + { + "epoch": 0.09356396945340843, + "grad_norm": 2.359957456588745, + "learning_rate": 1.99852746900052e-05, + "loss": 3.4397, + "step": 1167 + }, + { + "epoch": 0.09364414423443106, + "grad_norm": 2.7307305335998535, + "learning_rate": 1.9985204167099154e-05, + "loss": 3.4573, + "step": 1168 + }, + { + "epoch": 0.0937243190154537, + "grad_norm": 2.851628541946411, + "learning_rate": 1.998513347584608e-05, + "loss": 3.4073, + "step": 1169 + }, + { + "epoch": 0.09380449379647632, + "grad_norm": 2.677166700363159, + "learning_rate": 1.9985062616247178e-05, + "loss": 3.3922, + "step": 1170 + }, + { + "epoch": 0.09388466857749894, + "grad_norm": 2.6323773860931396, + "learning_rate": 1.9984991588303635e-05, + "loss": 3.3693, + "step": 1171 + }, + { + "epoch": 0.09396484335852158, + "grad_norm": 2.661818265914917, + "learning_rate": 1.9984920392016657e-05, + "loss": 3.4663, + "step": 1172 + }, + { + "epoch": 0.0940450181395442, + "grad_norm": 2.5405514240264893, + "learning_rate": 1.9984849027387438e-05, + "loss": 3.4507, + "step": 1173 + }, + { + "epoch": 0.09412519292056684, + "grad_norm": 2.606072187423706, + "learning_rate": 1.9984777494417186e-05, + "loss": 3.4199, + "step": 1174 + }, + { + "epoch": 0.09420536770158947, + "grad_norm": 2.9075191020965576, + "learning_rate": 1.9984705793107102e-05, + "loss": 3.4695, + "step": 1175 + }, + { + "epoch": 0.09428554248261209, + "grad_norm": 2.5503158569335938, + "learning_rate": 1.9984633923458398e-05, + "loss": 3.3833, + "step": 1176 + }, + { + "epoch": 0.09436571726363473, + "grad_norm": 3.3100812435150146, + "learning_rate": 1.9984561885472287e-05, + "loss": 3.4172, + "step": 1177 + }, + { + "epoch": 0.09444589204465735, + "grad_norm": 2.640113353729248, + "learning_rate": 1.998448967914998e-05, + "loss": 3.3381, + "step": 1178 + }, + { + "epoch": 0.09452606682567999, + "grad_norm": 2.7262957096099854, + "learning_rate": 1.9984417304492693e-05, + "loss": 3.4205, + "step": 1179 + }, + { + "epoch": 0.09460624160670261, + "grad_norm": 2.763866424560547, + "learning_rate": 1.9984344761501652e-05, + "loss": 3.4844, + "step": 1180 + }, + { + "epoch": 0.09468641638772524, + "grad_norm": 2.8268187046051025, + "learning_rate": 1.9984272050178077e-05, + "loss": 3.4286, + "step": 1181 + }, + { + "epoch": 0.09476659116874787, + "grad_norm": 2.62931489944458, + "learning_rate": 1.9984199170523194e-05, + "loss": 3.3656, + "step": 1182 + }, + { + "epoch": 0.0948467659497705, + "grad_norm": 2.5164737701416016, + "learning_rate": 1.998412612253823e-05, + "loss": 3.3596, + "step": 1183 + }, + { + "epoch": 0.09492694073079314, + "grad_norm": 2.454423189163208, + "learning_rate": 1.9984052906224422e-05, + "loss": 3.3754, + "step": 1184 + }, + { + "epoch": 0.09500711551181576, + "grad_norm": 2.7251901626586914, + "learning_rate": 1.9983979521582997e-05, + "loss": 3.3696, + "step": 1185 + }, + { + "epoch": 0.09508729029283838, + "grad_norm": 2.5617566108703613, + "learning_rate": 1.9983905968615196e-05, + "loss": 3.4036, + "step": 1186 + }, + { + "epoch": 0.09516746507386102, + "grad_norm": 2.557006597518921, + "learning_rate": 1.9983832247322264e-05, + "loss": 3.3478, + "step": 1187 + }, + { + "epoch": 0.09524763985488365, + "grad_norm": 2.448751926422119, + "learning_rate": 1.9983758357705434e-05, + "loss": 3.3051, + "step": 1188 + }, + { + "epoch": 0.09532781463590627, + "grad_norm": 2.608074426651001, + "learning_rate": 1.998368429976596e-05, + "loss": 3.4709, + "step": 1189 + }, + { + "epoch": 0.0954079894169289, + "grad_norm": 2.4829421043395996, + "learning_rate": 1.9983610073505086e-05, + "loss": 3.4105, + "step": 1190 + }, + { + "epoch": 0.09548816419795153, + "grad_norm": 2.741018772125244, + "learning_rate": 1.9983535678924058e-05, + "loss": 3.3168, + "step": 1191 + }, + { + "epoch": 0.09556833897897417, + "grad_norm": 2.870720624923706, + "learning_rate": 1.9983461116024147e-05, + "loss": 3.3372, + "step": 1192 + }, + { + "epoch": 0.09564851375999679, + "grad_norm": 2.699462890625, + "learning_rate": 1.9983386384806594e-05, + "loss": 3.4193, + "step": 1193 + }, + { + "epoch": 0.09572868854101942, + "grad_norm": 2.7559823989868164, + "learning_rate": 1.998331148527267e-05, + "loss": 3.4771, + "step": 1194 + }, + { + "epoch": 0.09580886332204205, + "grad_norm": 2.654541254043579, + "learning_rate": 1.998323641742363e-05, + "loss": 3.4464, + "step": 1195 + }, + { + "epoch": 0.09588903810306468, + "grad_norm": 2.5075387954711914, + "learning_rate": 1.9983161181260743e-05, + "loss": 3.3666, + "step": 1196 + }, + { + "epoch": 0.09596921288408732, + "grad_norm": 2.7008464336395264, + "learning_rate": 1.9983085776785277e-05, + "loss": 3.3284, + "step": 1197 + }, + { + "epoch": 0.09604938766510994, + "grad_norm": 3.0328269004821777, + "learning_rate": 1.99830102039985e-05, + "loss": 3.4113, + "step": 1198 + }, + { + "epoch": 0.09612956244613256, + "grad_norm": 2.6366615295410156, + "learning_rate": 1.9982934462901694e-05, + "loss": 3.4862, + "step": 1199 + }, + { + "epoch": 0.0962097372271552, + "grad_norm": 2.6945483684539795, + "learning_rate": 1.998285855349613e-05, + "loss": 3.3932, + "step": 1200 + }, + { + "epoch": 0.09628991200817782, + "grad_norm": 2.058791160583496, + "learning_rate": 1.9982782475783087e-05, + "loss": 2.6606, + "step": 1201 + }, + { + "epoch": 0.09637008678920046, + "grad_norm": 2.7556936740875244, + "learning_rate": 1.998270622976385e-05, + "loss": 3.3889, + "step": 1202 + }, + { + "epoch": 0.09645026157022309, + "grad_norm": 2.647449493408203, + "learning_rate": 1.9982629815439706e-05, + "loss": 3.3004, + "step": 1203 + }, + { + "epoch": 0.09653043635124571, + "grad_norm": 2.5971615314483643, + "learning_rate": 1.998255323281194e-05, + "loss": 3.3982, + "step": 1204 + }, + { + "epoch": 0.09661061113226835, + "grad_norm": 2.9947211742401123, + "learning_rate": 1.9982476481881846e-05, + "loss": 3.4534, + "step": 1205 + }, + { + "epoch": 0.09669078591329097, + "grad_norm": 2.6747658252716064, + "learning_rate": 1.9982399562650716e-05, + "loss": 3.5692, + "step": 1206 + }, + { + "epoch": 0.09677096069431361, + "grad_norm": 2.546473979949951, + "learning_rate": 1.9982322475119845e-05, + "loss": 3.3821, + "step": 1207 + }, + { + "epoch": 0.09685113547533623, + "grad_norm": 2.637160062789917, + "learning_rate": 1.9982245219290538e-05, + "loss": 3.4346, + "step": 1208 + }, + { + "epoch": 0.09693131025635886, + "grad_norm": 2.7329368591308594, + "learning_rate": 1.9982167795164095e-05, + "loss": 3.2915, + "step": 1209 + }, + { + "epoch": 0.0970114850373815, + "grad_norm": 1.6576489210128784, + "learning_rate": 1.998209020274182e-05, + "loss": 2.6638, + "step": 1210 + }, + { + "epoch": 0.09709165981840412, + "grad_norm": 2.741424322128296, + "learning_rate": 1.9982012442025027e-05, + "loss": 3.3229, + "step": 1211 + }, + { + "epoch": 0.09717183459942676, + "grad_norm": 2.524683713912964, + "learning_rate": 1.9981934513015017e-05, + "loss": 3.4661, + "step": 1212 + }, + { + "epoch": 0.09725200938044938, + "grad_norm": 2.7510087490081787, + "learning_rate": 1.998185641571311e-05, + "loss": 3.3136, + "step": 1213 + }, + { + "epoch": 0.097332184161472, + "grad_norm": 2.7322490215301514, + "learning_rate": 1.998177815012062e-05, + "loss": 3.5624, + "step": 1214 + }, + { + "epoch": 0.09741235894249464, + "grad_norm": 2.9717769622802734, + "learning_rate": 1.9981699716238872e-05, + "loss": 3.3392, + "step": 1215 + }, + { + "epoch": 0.09749253372351727, + "grad_norm": 2.5197699069976807, + "learning_rate": 1.9981621114069184e-05, + "loss": 3.5853, + "step": 1216 + }, + { + "epoch": 0.09757270850453989, + "grad_norm": 1.6339656114578247, + "learning_rate": 1.9981542343612877e-05, + "loss": 2.6556, + "step": 1217 + }, + { + "epoch": 0.09765288328556253, + "grad_norm": 2.600677728652954, + "learning_rate": 1.998146340487129e-05, + "loss": 3.4047, + "step": 1218 + }, + { + "epoch": 0.09773305806658515, + "grad_norm": 2.623776435852051, + "learning_rate": 1.9981384297845747e-05, + "loss": 3.463, + "step": 1219 + }, + { + "epoch": 0.09781323284760779, + "grad_norm": 2.6529541015625, + "learning_rate": 1.998130502253758e-05, + "loss": 3.3784, + "step": 1220 + }, + { + "epoch": 0.09789340762863041, + "grad_norm": 2.4190664291381836, + "learning_rate": 1.998122557894813e-05, + "loss": 3.3607, + "step": 1221 + }, + { + "epoch": 0.09797358240965304, + "grad_norm": 2.7285614013671875, + "learning_rate": 1.9981145967078733e-05, + "loss": 3.454, + "step": 1222 + }, + { + "epoch": 0.09805375719067567, + "grad_norm": 2.998338460922241, + "learning_rate": 1.9981066186930733e-05, + "loss": 3.39, + "step": 1223 + }, + { + "epoch": 0.0981339319716983, + "grad_norm": 2.6707348823547363, + "learning_rate": 1.9980986238505476e-05, + "loss": 3.3998, + "step": 1224 + }, + { + "epoch": 0.09821410675272094, + "grad_norm": 2.4453365802764893, + "learning_rate": 1.9980906121804307e-05, + "loss": 3.4863, + "step": 1225 + }, + { + "epoch": 0.09829428153374356, + "grad_norm": 2.532430648803711, + "learning_rate": 1.998082583682858e-05, + "loss": 3.2819, + "step": 1226 + }, + { + "epoch": 0.09837445631476618, + "grad_norm": 2.6605496406555176, + "learning_rate": 1.9980745383579646e-05, + "loss": 3.302, + "step": 1227 + }, + { + "epoch": 0.09845463109578882, + "grad_norm": 2.5314526557922363, + "learning_rate": 1.998066476205886e-05, + "loss": 3.3617, + "step": 1228 + }, + { + "epoch": 0.09853480587681145, + "grad_norm": 2.6699249744415283, + "learning_rate": 1.9980583972267585e-05, + "loss": 3.5505, + "step": 1229 + }, + { + "epoch": 0.09861498065783408, + "grad_norm": 2.756866455078125, + "learning_rate": 1.9980503014207183e-05, + "loss": 3.3389, + "step": 1230 + }, + { + "epoch": 0.09869515543885671, + "grad_norm": 2.5802390575408936, + "learning_rate": 1.9980421887879017e-05, + "loss": 3.4526, + "step": 1231 + }, + { + "epoch": 0.09877533021987933, + "grad_norm": 2.658818244934082, + "learning_rate": 1.9980340593284457e-05, + "loss": 3.3669, + "step": 1232 + }, + { + "epoch": 0.09885550500090197, + "grad_norm": 2.681835651397705, + "learning_rate": 1.9980259130424873e-05, + "loss": 3.418, + "step": 1233 + }, + { + "epoch": 0.0989356797819246, + "grad_norm": 2.4428718090057373, + "learning_rate": 1.9980177499301636e-05, + "loss": 3.4121, + "step": 1234 + }, + { + "epoch": 0.09901585456294723, + "grad_norm": 2.5596609115600586, + "learning_rate": 1.998009569991612e-05, + "loss": 3.3751, + "step": 1235 + }, + { + "epoch": 0.09909602934396985, + "grad_norm": 2.8749701976776123, + "learning_rate": 1.998001373226971e-05, + "loss": 3.3691, + "step": 1236 + }, + { + "epoch": 0.09917620412499248, + "grad_norm": 2.661112070083618, + "learning_rate": 1.997993159636379e-05, + "loss": 3.4053, + "step": 1237 + }, + { + "epoch": 0.09925637890601512, + "grad_norm": 2.342190742492676, + "learning_rate": 1.9979849292199735e-05, + "loss": 3.439, + "step": 1238 + }, + { + "epoch": 0.09933655368703774, + "grad_norm": 2.4941747188568115, + "learning_rate": 1.9979766819778942e-05, + "loss": 3.4916, + "step": 1239 + }, + { + "epoch": 0.09941672846806038, + "grad_norm": 2.5288236141204834, + "learning_rate": 1.9979684179102797e-05, + "loss": 3.3821, + "step": 1240 + }, + { + "epoch": 0.099496903249083, + "grad_norm": 2.60280179977417, + "learning_rate": 1.9979601370172694e-05, + "loss": 3.4771, + "step": 1241 + }, + { + "epoch": 0.09957707803010563, + "grad_norm": 3.232548475265503, + "learning_rate": 1.9979518392990026e-05, + "loss": 3.2744, + "step": 1242 + }, + { + "epoch": 0.09965725281112826, + "grad_norm": 2.4601449966430664, + "learning_rate": 1.9979435247556197e-05, + "loss": 3.381, + "step": 1243 + }, + { + "epoch": 0.09973742759215089, + "grad_norm": 1.5825018882751465, + "learning_rate": 1.9979351933872605e-05, + "loss": 2.6616, + "step": 1244 + }, + { + "epoch": 0.09981760237317353, + "grad_norm": 3.099560022354126, + "learning_rate": 1.997926845194066e-05, + "loss": 3.3434, + "step": 1245 + }, + { + "epoch": 0.09989777715419615, + "grad_norm": 2.7764177322387695, + "learning_rate": 1.9979184801761763e-05, + "loss": 3.449, + "step": 1246 + }, + { + "epoch": 0.09997795193521877, + "grad_norm": 2.3630783557891846, + "learning_rate": 1.997910098333733e-05, + "loss": 3.3597, + "step": 1247 + }, + { + "epoch": 0.10005812671624141, + "grad_norm": 2.4909887313842773, + "learning_rate": 1.9979016996668768e-05, + "loss": 3.4081, + "step": 1248 + }, + { + "epoch": 0.10013830149726403, + "grad_norm": 2.6519551277160645, + "learning_rate": 1.9978932841757498e-05, + "loss": 3.3558, + "step": 1249 + }, + { + "epoch": 0.10021847627828666, + "grad_norm": 2.840891122817993, + "learning_rate": 1.9978848518604938e-05, + "loss": 3.4926, + "step": 1250 + }, + { + "epoch": 0.1002986510593093, + "grad_norm": 2.8665146827697754, + "learning_rate": 1.997876402721251e-05, + "loss": 3.2889, + "step": 1251 + }, + { + "epoch": 0.10037882584033192, + "grad_norm": 2.437431812286377, + "learning_rate": 1.9978679367581635e-05, + "loss": 3.3756, + "step": 1252 + }, + { + "epoch": 0.10045900062135456, + "grad_norm": 2.4598426818847656, + "learning_rate": 1.997859453971374e-05, + "loss": 3.413, + "step": 1253 + }, + { + "epoch": 0.10053917540237718, + "grad_norm": 2.557248830795288, + "learning_rate": 1.997850954361026e-05, + "loss": 3.3713, + "step": 1254 + }, + { + "epoch": 0.1006193501833998, + "grad_norm": 2.527427911758423, + "learning_rate": 1.997842437927263e-05, + "loss": 3.4509, + "step": 1255 + }, + { + "epoch": 0.10069952496442244, + "grad_norm": 2.959649085998535, + "learning_rate": 1.9978339046702278e-05, + "loss": 3.3812, + "step": 1256 + }, + { + "epoch": 0.10077969974544507, + "grad_norm": 2.4224560260772705, + "learning_rate": 1.997825354590065e-05, + "loss": 3.3324, + "step": 1257 + }, + { + "epoch": 0.1008598745264677, + "grad_norm": 2.678090810775757, + "learning_rate": 1.997816787686918e-05, + "loss": 3.3813, + "step": 1258 + }, + { + "epoch": 0.10094004930749033, + "grad_norm": 1.952713966369629, + "learning_rate": 1.9978082039609316e-05, + "loss": 2.6487, + "step": 1259 + }, + { + "epoch": 0.10102022408851295, + "grad_norm": 3.0816352367401123, + "learning_rate": 1.9977996034122505e-05, + "loss": 3.337, + "step": 1260 + }, + { + "epoch": 0.10110039886953559, + "grad_norm": 3.006070137023926, + "learning_rate": 1.99779098604102e-05, + "loss": 3.3917, + "step": 1261 + }, + { + "epoch": 0.10118057365055821, + "grad_norm": 2.6150872707366943, + "learning_rate": 1.997782351847385e-05, + "loss": 3.422, + "step": 1262 + }, + { + "epoch": 0.10126074843158085, + "grad_norm": 2.471501111984253, + "learning_rate": 1.9977737008314914e-05, + "loss": 3.3433, + "step": 1263 + }, + { + "epoch": 0.10134092321260348, + "grad_norm": 2.8284499645233154, + "learning_rate": 1.9977650329934847e-05, + "loss": 3.3997, + "step": 1264 + }, + { + "epoch": 0.1014210979936261, + "grad_norm": 2.78549861907959, + "learning_rate": 1.997756348333511e-05, + "loss": 3.4615, + "step": 1265 + }, + { + "epoch": 0.10150127277464874, + "grad_norm": 2.664252281188965, + "learning_rate": 1.9977476468517173e-05, + "loss": 3.4376, + "step": 1266 + }, + { + "epoch": 0.10158144755567136, + "grad_norm": 2.5589962005615234, + "learning_rate": 1.9977389285482497e-05, + "loss": 3.3723, + "step": 1267 + }, + { + "epoch": 0.101661622336694, + "grad_norm": 2.561814308166504, + "learning_rate": 1.9977301934232555e-05, + "loss": 3.4531, + "step": 1268 + }, + { + "epoch": 0.10174179711771662, + "grad_norm": 2.5105552673339844, + "learning_rate": 1.9977214414768816e-05, + "loss": 3.3986, + "step": 1269 + }, + { + "epoch": 0.10182197189873925, + "grad_norm": 2.897493600845337, + "learning_rate": 1.997712672709276e-05, + "loss": 3.3436, + "step": 1270 + }, + { + "epoch": 0.10190214667976188, + "grad_norm": 2.2245118618011475, + "learning_rate": 1.9977038871205864e-05, + "loss": 3.3434, + "step": 1271 + }, + { + "epoch": 0.10198232146078451, + "grad_norm": 2.499974250793457, + "learning_rate": 1.997695084710961e-05, + "loss": 3.3859, + "step": 1272 + }, + { + "epoch": 0.10206249624180715, + "grad_norm": 2.0399973392486572, + "learning_rate": 1.997686265480548e-05, + "loss": 2.6656, + "step": 1273 + }, + { + "epoch": 0.10214267102282977, + "grad_norm": 2.4646685123443604, + "learning_rate": 1.997677429429496e-05, + "loss": 3.3289, + "step": 1274 + }, + { + "epoch": 0.1022228458038524, + "grad_norm": 2.530395269393921, + "learning_rate": 1.9976685765579545e-05, + "loss": 3.3507, + "step": 1275 + }, + { + "epoch": 0.10230302058487503, + "grad_norm": 2.210503578186035, + "learning_rate": 1.997659706866072e-05, + "loss": 3.3092, + "step": 1276 + }, + { + "epoch": 0.10238319536589766, + "grad_norm": 2.4622578620910645, + "learning_rate": 1.9976508203539987e-05, + "loss": 3.3309, + "step": 1277 + }, + { + "epoch": 0.10246337014692028, + "grad_norm": 2.5060410499572754, + "learning_rate": 1.9976419170218843e-05, + "loss": 3.5193, + "step": 1278 + }, + { + "epoch": 0.10254354492794292, + "grad_norm": 2.6796672344207764, + "learning_rate": 1.9976329968698788e-05, + "loss": 3.4048, + "step": 1279 + }, + { + "epoch": 0.10262371970896554, + "grad_norm": 2.6093997955322266, + "learning_rate": 1.997624059898132e-05, + "loss": 3.3969, + "step": 1280 + }, + { + "epoch": 0.10270389448998818, + "grad_norm": 2.3400065898895264, + "learning_rate": 1.9976151061067958e-05, + "loss": 3.3788, + "step": 1281 + }, + { + "epoch": 0.1027840692710108, + "grad_norm": 2.5757131576538086, + "learning_rate": 1.9976061354960202e-05, + "loss": 3.3971, + "step": 1282 + }, + { + "epoch": 0.10286424405203343, + "grad_norm": 2.8175745010375977, + "learning_rate": 1.997597148065957e-05, + "loss": 3.3597, + "step": 1283 + }, + { + "epoch": 0.10294441883305606, + "grad_norm": 2.4353439807891846, + "learning_rate": 1.997588143816757e-05, + "loss": 3.2705, + "step": 1284 + }, + { + "epoch": 0.10302459361407869, + "grad_norm": 2.618496894836426, + "learning_rate": 1.997579122748573e-05, + "loss": 3.4679, + "step": 1285 + }, + { + "epoch": 0.10310476839510133, + "grad_norm": 2.5630645751953125, + "learning_rate": 1.9975700848615563e-05, + "loss": 3.4781, + "step": 1286 + }, + { + "epoch": 0.10318494317612395, + "grad_norm": 2.5198845863342285, + "learning_rate": 1.9975610301558596e-05, + "loss": 3.3861, + "step": 1287 + }, + { + "epoch": 0.10326511795714657, + "grad_norm": 2.607025384902954, + "learning_rate": 1.9975519586316353e-05, + "loss": 3.4639, + "step": 1288 + }, + { + "epoch": 0.10334529273816921, + "grad_norm": 2.4550621509552, + "learning_rate": 1.9975428702890368e-05, + "loss": 3.384, + "step": 1289 + }, + { + "epoch": 0.10342546751919184, + "grad_norm": 2.4641714096069336, + "learning_rate": 1.9975337651282166e-05, + "loss": 3.4482, + "step": 1290 + }, + { + "epoch": 0.10350564230021447, + "grad_norm": 2.5312259197235107, + "learning_rate": 1.9975246431493295e-05, + "loss": 3.4519, + "step": 1291 + }, + { + "epoch": 0.1035858170812371, + "grad_norm": 2.5283548831939697, + "learning_rate": 1.9975155043525278e-05, + "loss": 3.4436, + "step": 1292 + }, + { + "epoch": 0.10366599186225972, + "grad_norm": 2.4686453342437744, + "learning_rate": 1.9975063487379668e-05, + "loss": 3.358, + "step": 1293 + }, + { + "epoch": 0.10374616664328236, + "grad_norm": 2.390125036239624, + "learning_rate": 1.9974971763058e-05, + "loss": 3.3724, + "step": 1294 + }, + { + "epoch": 0.10382634142430498, + "grad_norm": 2.600217342376709, + "learning_rate": 1.9974879870561822e-05, + "loss": 3.3898, + "step": 1295 + }, + { + "epoch": 0.10390651620532762, + "grad_norm": 2.7028918266296387, + "learning_rate": 1.9974787809892685e-05, + "loss": 3.4314, + "step": 1296 + }, + { + "epoch": 0.10398669098635024, + "grad_norm": 2.7929201126098633, + "learning_rate": 1.9974695581052144e-05, + "loss": 3.4866, + "step": 1297 + }, + { + "epoch": 0.10406686576737287, + "grad_norm": 2.9284214973449707, + "learning_rate": 1.997460318404175e-05, + "loss": 3.4196, + "step": 1298 + }, + { + "epoch": 0.1041470405483955, + "grad_norm": 1.531266450881958, + "learning_rate": 1.997451061886306e-05, + "loss": 2.6519, + "step": 1299 + }, + { + "epoch": 0.10422721532941813, + "grad_norm": 2.9318528175354004, + "learning_rate": 1.9974417885517638e-05, + "loss": 3.4021, + "step": 1300 + }, + { + "epoch": 0.10430739011044077, + "grad_norm": 2.7277567386627197, + "learning_rate": 1.9974324984007047e-05, + "loss": 3.3774, + "step": 1301 + }, + { + "epoch": 0.10438756489146339, + "grad_norm": 2.500680446624756, + "learning_rate": 1.997423191433285e-05, + "loss": 3.3766, + "step": 1302 + }, + { + "epoch": 0.10446773967248602, + "grad_norm": 2.5674848556518555, + "learning_rate": 1.997413867649662e-05, + "loss": 3.4338, + "step": 1303 + }, + { + "epoch": 0.10454791445350865, + "grad_norm": 2.9069225788116455, + "learning_rate": 1.9974045270499923e-05, + "loss": 3.4471, + "step": 1304 + }, + { + "epoch": 0.10462808923453128, + "grad_norm": 2.572330951690674, + "learning_rate": 1.9973951696344344e-05, + "loss": 3.3734, + "step": 1305 + }, + { + "epoch": 0.1047082640155539, + "grad_norm": 2.5105929374694824, + "learning_rate": 1.997385795403145e-05, + "loss": 3.3627, + "step": 1306 + }, + { + "epoch": 0.10478843879657654, + "grad_norm": 2.278763771057129, + "learning_rate": 1.997376404356283e-05, + "loss": 3.387, + "step": 1307 + }, + { + "epoch": 0.10486861357759916, + "grad_norm": 2.5784225463867188, + "learning_rate": 1.9973669964940056e-05, + "loss": 3.2981, + "step": 1308 + }, + { + "epoch": 0.1049487883586218, + "grad_norm": 2.57832670211792, + "learning_rate": 1.9973575718164727e-05, + "loss": 3.4779, + "step": 1309 + }, + { + "epoch": 0.10502896313964442, + "grad_norm": 2.4642739295959473, + "learning_rate": 1.9973481303238428e-05, + "loss": 3.3279, + "step": 1310 + }, + { + "epoch": 0.10510913792066705, + "grad_norm": 2.523259401321411, + "learning_rate": 1.9973386720162746e-05, + "loss": 3.5193, + "step": 1311 + }, + { + "epoch": 0.10518931270168969, + "grad_norm": 2.7018113136291504, + "learning_rate": 1.997329196893928e-05, + "loss": 3.4629, + "step": 1312 + }, + { + "epoch": 0.10526948748271231, + "grad_norm": 2.6093311309814453, + "learning_rate": 1.9973197049569624e-05, + "loss": 3.4337, + "step": 1313 + }, + { + "epoch": 0.10534966226373495, + "grad_norm": 2.5116496086120605, + "learning_rate": 1.9973101962055383e-05, + "loss": 3.4623, + "step": 1314 + }, + { + "epoch": 0.10542983704475757, + "grad_norm": 2.321803569793701, + "learning_rate": 1.997300670639816e-05, + "loss": 3.337, + "step": 1315 + }, + { + "epoch": 0.1055100118257802, + "grad_norm": 1.532580018043518, + "learning_rate": 1.9972911282599553e-05, + "loss": 2.6277, + "step": 1316 + }, + { + "epoch": 0.10559018660680283, + "grad_norm": 2.4580867290496826, + "learning_rate": 1.997281569066118e-05, + "loss": 3.4755, + "step": 1317 + }, + { + "epoch": 0.10567036138782546, + "grad_norm": 2.543029546737671, + "learning_rate": 1.997271993058465e-05, + "loss": 3.5215, + "step": 1318 + }, + { + "epoch": 0.1057505361688481, + "grad_norm": 2.394742250442505, + "learning_rate": 1.9972624002371575e-05, + "loss": 3.3391, + "step": 1319 + }, + { + "epoch": 0.10583071094987072, + "grad_norm": 2.90061092376709, + "learning_rate": 1.9972527906023577e-05, + "loss": 3.431, + "step": 1320 + }, + { + "epoch": 0.10591088573089334, + "grad_norm": 2.681318759918213, + "learning_rate": 1.997243164154227e-05, + "loss": 3.4042, + "step": 1321 + }, + { + "epoch": 0.10599106051191598, + "grad_norm": 2.4904983043670654, + "learning_rate": 1.997233520892928e-05, + "loss": 3.3658, + "step": 1322 + }, + { + "epoch": 0.1060712352929386, + "grad_norm": 2.411968946456909, + "learning_rate": 1.9972238608186233e-05, + "loss": 3.3945, + "step": 1323 + }, + { + "epoch": 0.10615141007396124, + "grad_norm": 2.6627252101898193, + "learning_rate": 1.997214183931476e-05, + "loss": 3.3837, + "step": 1324 + }, + { + "epoch": 0.10623158485498387, + "grad_norm": 2.267850637435913, + "learning_rate": 1.997204490231649e-05, + "loss": 3.3916, + "step": 1325 + }, + { + "epoch": 0.10631175963600649, + "grad_norm": 2.5439138412475586, + "learning_rate": 1.9971947797193054e-05, + "loss": 3.3649, + "step": 1326 + }, + { + "epoch": 0.10639193441702913, + "grad_norm": 1.3696142435073853, + "learning_rate": 1.9971850523946095e-05, + "loss": 2.649, + "step": 1327 + }, + { + "epoch": 0.10647210919805175, + "grad_norm": 2.3548388481140137, + "learning_rate": 1.997175308257725e-05, + "loss": 3.3662, + "step": 1328 + }, + { + "epoch": 0.10655228397907439, + "grad_norm": 2.36635422706604, + "learning_rate": 1.9971655473088163e-05, + "loss": 3.3699, + "step": 1329 + }, + { + "epoch": 0.10663245876009701, + "grad_norm": 3.19661808013916, + "learning_rate": 1.997155769548048e-05, + "loss": 3.3256, + "step": 1330 + }, + { + "epoch": 0.10671263354111964, + "grad_norm": 2.6831254959106445, + "learning_rate": 1.9971459749755846e-05, + "loss": 3.366, + "step": 1331 + }, + { + "epoch": 0.10679280832214227, + "grad_norm": 2.515305280685425, + "learning_rate": 1.9971361635915918e-05, + "loss": 3.42, + "step": 1332 + }, + { + "epoch": 0.1068729831031649, + "grad_norm": 2.543701410293579, + "learning_rate": 1.997126335396234e-05, + "loss": 3.4098, + "step": 1333 + }, + { + "epoch": 0.10695315788418752, + "grad_norm": 2.397066354751587, + "learning_rate": 1.9971164903896784e-05, + "loss": 3.2731, + "step": 1334 + }, + { + "epoch": 0.10703333266521016, + "grad_norm": 1.8562690019607544, + "learning_rate": 1.9971066285720903e-05, + "loss": 2.6311, + "step": 1335 + }, + { + "epoch": 0.10711350744623278, + "grad_norm": 2.7264771461486816, + "learning_rate": 1.9970967499436354e-05, + "loss": 3.4377, + "step": 1336 + }, + { + "epoch": 0.10719368222725542, + "grad_norm": 2.552633047103882, + "learning_rate": 1.9970868545044805e-05, + "loss": 3.4326, + "step": 1337 + }, + { + "epoch": 0.10727385700827805, + "grad_norm": 2.880661964416504, + "learning_rate": 1.9970769422547933e-05, + "loss": 3.2391, + "step": 1338 + }, + { + "epoch": 0.10735403178930067, + "grad_norm": 2.4864120483398438, + "learning_rate": 1.9970670131947397e-05, + "loss": 3.4584, + "step": 1339 + }, + { + "epoch": 0.1074342065703233, + "grad_norm": 2.4912374019622803, + "learning_rate": 1.997057067324488e-05, + "loss": 3.391, + "step": 1340 + }, + { + "epoch": 0.10751438135134593, + "grad_norm": 2.5115509033203125, + "learning_rate": 1.9970471046442053e-05, + "loss": 3.3982, + "step": 1341 + }, + { + "epoch": 0.10759455613236857, + "grad_norm": 2.7571310997009277, + "learning_rate": 1.9970371251540598e-05, + "loss": 3.4191, + "step": 1342 + }, + { + "epoch": 0.10767473091339119, + "grad_norm": 2.6171152591705322, + "learning_rate": 1.99702712885422e-05, + "loss": 3.5442, + "step": 1343 + }, + { + "epoch": 0.10775490569441382, + "grad_norm": 2.692840337753296, + "learning_rate": 1.997017115744854e-05, + "loss": 3.454, + "step": 1344 + }, + { + "epoch": 0.10783508047543645, + "grad_norm": 2.6696324348449707, + "learning_rate": 1.997007085826131e-05, + "loss": 3.4053, + "step": 1345 + }, + { + "epoch": 0.10791525525645908, + "grad_norm": 2.5086753368377686, + "learning_rate": 1.9969970390982197e-05, + "loss": 3.4109, + "step": 1346 + }, + { + "epoch": 0.10799543003748172, + "grad_norm": 2.5398173332214355, + "learning_rate": 1.9969869755612897e-05, + "loss": 3.4241, + "step": 1347 + }, + { + "epoch": 0.10807560481850434, + "grad_norm": 2.7130308151245117, + "learning_rate": 1.996976895215511e-05, + "loss": 3.2461, + "step": 1348 + }, + { + "epoch": 0.10815577959952696, + "grad_norm": 2.4182755947113037, + "learning_rate": 1.9969667980610527e-05, + "loss": 3.3525, + "step": 1349 + }, + { + "epoch": 0.1082359543805496, + "grad_norm": 2.5555524826049805, + "learning_rate": 1.996956684098086e-05, + "loss": 3.4386, + "step": 1350 + }, + { + "epoch": 0.10831612916157222, + "grad_norm": 2.315793991088867, + "learning_rate": 1.9969465533267803e-05, + "loss": 3.3822, + "step": 1351 + }, + { + "epoch": 0.10839630394259486, + "grad_norm": 2.200042963027954, + "learning_rate": 1.9969364057473077e-05, + "loss": 3.2727, + "step": 1352 + }, + { + "epoch": 0.10847647872361749, + "grad_norm": 1.8516185283660889, + "learning_rate": 1.9969262413598387e-05, + "loss": 2.6417, + "step": 1353 + }, + { + "epoch": 0.10855665350464011, + "grad_norm": 2.5219016075134277, + "learning_rate": 1.996916060164544e-05, + "loss": 3.2877, + "step": 1354 + }, + { + "epoch": 0.10863682828566275, + "grad_norm": 2.391143321990967, + "learning_rate": 1.9969058621615965e-05, + "loss": 3.3815, + "step": 1355 + }, + { + "epoch": 0.10871700306668537, + "grad_norm": 2.340240478515625, + "learning_rate": 1.9968956473511673e-05, + "loss": 3.4192, + "step": 1356 + }, + { + "epoch": 0.10879717784770801, + "grad_norm": 2.3807363510131836, + "learning_rate": 1.9968854157334288e-05, + "loss": 3.4248, + "step": 1357 + }, + { + "epoch": 0.10887735262873063, + "grad_norm": 1.3394508361816406, + "learning_rate": 1.9968751673085536e-05, + "loss": 2.6255, + "step": 1358 + }, + { + "epoch": 0.10895752740975326, + "grad_norm": 2.383622169494629, + "learning_rate": 1.9968649020767145e-05, + "loss": 3.1595, + "step": 1359 + }, + { + "epoch": 0.1090377021907759, + "grad_norm": 2.440612554550171, + "learning_rate": 1.9968546200380842e-05, + "loss": 3.4183, + "step": 1360 + }, + { + "epoch": 0.10911787697179852, + "grad_norm": 2.433789014816284, + "learning_rate": 1.9968443211928366e-05, + "loss": 3.4541, + "step": 1361 + }, + { + "epoch": 0.10919805175282116, + "grad_norm": 2.465822219848633, + "learning_rate": 1.9968340055411448e-05, + "loss": 3.3381, + "step": 1362 + }, + { + "epoch": 0.10927822653384378, + "grad_norm": 1.3836828470230103, + "learning_rate": 1.9968236730831833e-05, + "loss": 2.6173, + "step": 1363 + }, + { + "epoch": 0.1093584013148664, + "grad_norm": 2.449627637863159, + "learning_rate": 1.9968133238191258e-05, + "loss": 3.3219, + "step": 1364 + }, + { + "epoch": 0.10943857609588904, + "grad_norm": 2.6059556007385254, + "learning_rate": 1.996802957749147e-05, + "loss": 3.3699, + "step": 1365 + }, + { + "epoch": 0.10951875087691167, + "grad_norm": 3.1022603511810303, + "learning_rate": 1.9967925748734215e-05, + "loss": 3.408, + "step": 1366 + }, + { + "epoch": 0.10959892565793429, + "grad_norm": 2.649897336959839, + "learning_rate": 1.9967821751921247e-05, + "loss": 3.3616, + "step": 1367 + }, + { + "epoch": 0.10967910043895693, + "grad_norm": 2.5115790367126465, + "learning_rate": 1.9967717587054316e-05, + "loss": 3.3862, + "step": 1368 + }, + { + "epoch": 0.10975927521997955, + "grad_norm": 2.2289795875549316, + "learning_rate": 1.9967613254135177e-05, + "loss": 3.4608, + "step": 1369 + }, + { + "epoch": 0.10983945000100219, + "grad_norm": 2.293149948120117, + "learning_rate": 1.9967508753165596e-05, + "loss": 3.2631, + "step": 1370 + }, + { + "epoch": 0.10991962478202481, + "grad_norm": 2.356344223022461, + "learning_rate": 1.9967404084147328e-05, + "loss": 3.34, + "step": 1371 + }, + { + "epoch": 0.10999979956304744, + "grad_norm": 2.3485445976257324, + "learning_rate": 1.9967299247082145e-05, + "loss": 3.2335, + "step": 1372 + }, + { + "epoch": 0.11007997434407008, + "grad_norm": 2.6557953357696533, + "learning_rate": 1.9967194241971805e-05, + "loss": 3.5023, + "step": 1373 + }, + { + "epoch": 0.1101601491250927, + "grad_norm": 2.4312496185302734, + "learning_rate": 1.996708906881808e-05, + "loss": 3.4198, + "step": 1374 + }, + { + "epoch": 0.11024032390611534, + "grad_norm": 2.1239020824432373, + "learning_rate": 1.996698372762275e-05, + "loss": 3.3564, + "step": 1375 + }, + { + "epoch": 0.11032049868713796, + "grad_norm": 1.3166372776031494, + "learning_rate": 1.9966878218387588e-05, + "loss": 2.6333, + "step": 1376 + }, + { + "epoch": 0.11040067346816058, + "grad_norm": 2.604729413986206, + "learning_rate": 1.9966772541114367e-05, + "loss": 3.3461, + "step": 1377 + }, + { + "epoch": 0.11048084824918322, + "grad_norm": 2.9982404708862305, + "learning_rate": 1.9966666695804878e-05, + "loss": 3.4166, + "step": 1378 + }, + { + "epoch": 0.11056102303020585, + "grad_norm": 2.4423375129699707, + "learning_rate": 1.99665606824609e-05, + "loss": 3.2896, + "step": 1379 + }, + { + "epoch": 0.11064119781122848, + "grad_norm": 2.503251075744629, + "learning_rate": 1.996645450108422e-05, + "loss": 3.4048, + "step": 1380 + }, + { + "epoch": 0.11072137259225111, + "grad_norm": 2.3042099475860596, + "learning_rate": 1.996634815167663e-05, + "loss": 3.3552, + "step": 1381 + }, + { + "epoch": 0.11080154737327373, + "grad_norm": 2.7453556060791016, + "learning_rate": 1.9966241634239925e-05, + "loss": 3.4054, + "step": 1382 + }, + { + "epoch": 0.11088172215429637, + "grad_norm": 2.3432271480560303, + "learning_rate": 1.9966134948775897e-05, + "loss": 3.3923, + "step": 1383 + }, + { + "epoch": 0.110961896935319, + "grad_norm": 2.9421753883361816, + "learning_rate": 1.9966028095286346e-05, + "loss": 3.2562, + "step": 1384 + }, + { + "epoch": 0.11104207171634163, + "grad_norm": 2.3648738861083984, + "learning_rate": 1.9965921073773072e-05, + "loss": 3.4131, + "step": 1385 + }, + { + "epoch": 0.11112224649736425, + "grad_norm": 2.5557684898376465, + "learning_rate": 1.9965813884237884e-05, + "loss": 3.3649, + "step": 1386 + }, + { + "epoch": 0.11120242127838688, + "grad_norm": 2.9133336544036865, + "learning_rate": 1.9965706526682583e-05, + "loss": 3.4274, + "step": 1387 + }, + { + "epoch": 0.11128259605940952, + "grad_norm": 2.7711498737335205, + "learning_rate": 1.9965599001108988e-05, + "loss": 3.4072, + "step": 1388 + }, + { + "epoch": 0.11136277084043214, + "grad_norm": 2.809382200241089, + "learning_rate": 1.9965491307518902e-05, + "loss": 3.2969, + "step": 1389 + }, + { + "epoch": 0.11144294562145478, + "grad_norm": 2.77720308303833, + "learning_rate": 1.9965383445914143e-05, + "loss": 3.3679, + "step": 1390 + }, + { + "epoch": 0.1115231204024774, + "grad_norm": 2.5588645935058594, + "learning_rate": 1.9965275416296533e-05, + "loss": 3.3229, + "step": 1391 + }, + { + "epoch": 0.11160329518350003, + "grad_norm": 2.464431047439575, + "learning_rate": 1.9965167218667894e-05, + "loss": 3.5211, + "step": 1392 + }, + { + "epoch": 0.11168346996452266, + "grad_norm": 2.5434377193450928, + "learning_rate": 1.9965058853030044e-05, + "loss": 3.4224, + "step": 1393 + }, + { + "epoch": 0.11176364474554529, + "grad_norm": 2.6768412590026855, + "learning_rate": 1.9964950319384814e-05, + "loss": 3.3298, + "step": 1394 + }, + { + "epoch": 0.11184381952656791, + "grad_norm": 2.3887693881988525, + "learning_rate": 1.9964841617734035e-05, + "loss": 3.4292, + "step": 1395 + }, + { + "epoch": 0.11192399430759055, + "grad_norm": 2.481621503829956, + "learning_rate": 1.996473274807954e-05, + "loss": 3.3276, + "step": 1396 + }, + { + "epoch": 0.11200416908861317, + "grad_norm": 2.869351625442505, + "learning_rate": 1.996462371042316e-05, + "loss": 3.4542, + "step": 1397 + }, + { + "epoch": 0.11208434386963581, + "grad_norm": 2.671532154083252, + "learning_rate": 1.996451450476674e-05, + "loss": 3.3904, + "step": 1398 + }, + { + "epoch": 0.11216451865065843, + "grad_norm": 2.3076071739196777, + "learning_rate": 1.9964405131112113e-05, + "loss": 3.4379, + "step": 1399 + }, + { + "epoch": 0.11224469343168106, + "grad_norm": 2.753865957260132, + "learning_rate": 1.9964295589461125e-05, + "loss": 3.3695, + "step": 1400 + }, + { + "epoch": 0.1123248682127037, + "grad_norm": 1.68938148021698, + "learning_rate": 1.9964185879815633e-05, + "loss": 2.601, + "step": 1401 + }, + { + "epoch": 0.11240504299372632, + "grad_norm": 2.513777256011963, + "learning_rate": 1.9964076002177473e-05, + "loss": 3.3744, + "step": 1402 + }, + { + "epoch": 0.11248521777474896, + "grad_norm": 2.603947639465332, + "learning_rate": 1.9963965956548505e-05, + "loss": 3.4982, + "step": 1403 + }, + { + "epoch": 0.11256539255577158, + "grad_norm": 2.5545413494110107, + "learning_rate": 1.9963855742930583e-05, + "loss": 3.4561, + "step": 1404 + }, + { + "epoch": 0.1126455673367942, + "grad_norm": 1.6174110174179077, + "learning_rate": 1.9963745361325563e-05, + "loss": 2.6205, + "step": 1405 + }, + { + "epoch": 0.11272574211781684, + "grad_norm": 2.5304980278015137, + "learning_rate": 1.9963634811735312e-05, + "loss": 3.2065, + "step": 1406 + }, + { + "epoch": 0.11280591689883947, + "grad_norm": 1.3257725238800049, + "learning_rate": 1.9963524094161687e-05, + "loss": 2.6038, + "step": 1407 + }, + { + "epoch": 0.1128860916798621, + "grad_norm": 2.311168909072876, + "learning_rate": 1.9963413208606554e-05, + "loss": 3.3049, + "step": 1408 + }, + { + "epoch": 0.11296626646088473, + "grad_norm": 2.617140531539917, + "learning_rate": 1.9963302155071788e-05, + "loss": 3.3706, + "step": 1409 + }, + { + "epoch": 0.11304644124190735, + "grad_norm": 2.3457748889923096, + "learning_rate": 1.996319093355926e-05, + "loss": 3.328, + "step": 1410 + }, + { + "epoch": 0.11312661602292999, + "grad_norm": 2.6009521484375, + "learning_rate": 1.9963079544070846e-05, + "loss": 3.3661, + "step": 1411 + }, + { + "epoch": 0.11320679080395261, + "grad_norm": 2.4031753540039062, + "learning_rate": 1.996296798660842e-05, + "loss": 3.4691, + "step": 1412 + }, + { + "epoch": 0.11328696558497525, + "grad_norm": 2.3737447261810303, + "learning_rate": 1.9962856261173866e-05, + "loss": 3.4393, + "step": 1413 + }, + { + "epoch": 0.11336714036599788, + "grad_norm": 1.304734706878662, + "learning_rate": 1.9962744367769066e-05, + "loss": 2.6122, + "step": 1414 + }, + { + "epoch": 0.1134473151470205, + "grad_norm": 2.5836005210876465, + "learning_rate": 1.9962632306395905e-05, + "loss": 3.4148, + "step": 1415 + }, + { + "epoch": 0.11352748992804314, + "grad_norm": 2.3444292545318604, + "learning_rate": 1.9962520077056277e-05, + "loss": 3.4084, + "step": 1416 + }, + { + "epoch": 0.11360766470906576, + "grad_norm": 2.307640314102173, + "learning_rate": 1.996240767975207e-05, + "loss": 3.272, + "step": 1417 + }, + { + "epoch": 0.1136878394900884, + "grad_norm": 2.3736917972564697, + "learning_rate": 1.9962295114485183e-05, + "loss": 3.3616, + "step": 1418 + }, + { + "epoch": 0.11376801427111102, + "grad_norm": 2.413748264312744, + "learning_rate": 1.996218238125751e-05, + "loss": 3.3859, + "step": 1419 + }, + { + "epoch": 0.11384818905213365, + "grad_norm": 2.5151097774505615, + "learning_rate": 1.9962069480070954e-05, + "loss": 3.3002, + "step": 1420 + }, + { + "epoch": 0.11392836383315628, + "grad_norm": 2.2940120697021484, + "learning_rate": 1.9961956410927415e-05, + "loss": 3.3932, + "step": 1421 + }, + { + "epoch": 0.11400853861417891, + "grad_norm": 1.5957744121551514, + "learning_rate": 1.9961843173828806e-05, + "loss": 2.6085, + "step": 1422 + }, + { + "epoch": 0.11408871339520153, + "grad_norm": 2.7000553607940674, + "learning_rate": 1.996172976877703e-05, + "loss": 3.3613, + "step": 1423 + }, + { + "epoch": 0.11416888817622417, + "grad_norm": 2.5040647983551025, + "learning_rate": 1.9961616195774e-05, + "loss": 3.3864, + "step": 1424 + }, + { + "epoch": 0.1142490629572468, + "grad_norm": 2.697265625, + "learning_rate": 1.9961502454821633e-05, + "loss": 3.3306, + "step": 1425 + }, + { + "epoch": 0.11432923773826943, + "grad_norm": 2.448762893676758, + "learning_rate": 1.9961388545921844e-05, + "loss": 3.4698, + "step": 1426 + }, + { + "epoch": 0.11440941251929206, + "grad_norm": 2.3564162254333496, + "learning_rate": 1.9961274469076553e-05, + "loss": 3.4611, + "step": 1427 + }, + { + "epoch": 0.11448958730031468, + "grad_norm": 2.3902556896209717, + "learning_rate": 1.996116022428769e-05, + "loss": 3.3685, + "step": 1428 + }, + { + "epoch": 0.11456976208133732, + "grad_norm": 2.2521958351135254, + "learning_rate": 1.9961045811557175e-05, + "loss": 3.3438, + "step": 1429 + }, + { + "epoch": 0.11464993686235994, + "grad_norm": 1.2504287958145142, + "learning_rate": 1.9960931230886936e-05, + "loss": 2.6156, + "step": 1430 + }, + { + "epoch": 0.11473011164338258, + "grad_norm": 2.576855421066284, + "learning_rate": 1.9960816482278907e-05, + "loss": 3.3333, + "step": 1431 + }, + { + "epoch": 0.1148102864244052, + "grad_norm": 2.306137800216675, + "learning_rate": 1.9960701565735026e-05, + "loss": 3.3973, + "step": 1432 + }, + { + "epoch": 0.11489046120542783, + "grad_norm": 1.190265417098999, + "learning_rate": 1.9960586481257224e-05, + "loss": 2.6129, + "step": 1433 + }, + { + "epoch": 0.11497063598645046, + "grad_norm": 2.4402849674224854, + "learning_rate": 1.9960471228847445e-05, + "loss": 3.4332, + "step": 1434 + }, + { + "epoch": 0.11505081076747309, + "grad_norm": 1.2153384685516357, + "learning_rate": 1.9960355808507632e-05, + "loss": 2.5874, + "step": 1435 + }, + { + "epoch": 0.11513098554849573, + "grad_norm": 2.2871861457824707, + "learning_rate": 1.9960240220239732e-05, + "loss": 3.395, + "step": 1436 + }, + { + "epoch": 0.11521116032951835, + "grad_norm": 1.191616177558899, + "learning_rate": 1.996012446404569e-05, + "loss": 2.5911, + "step": 1437 + }, + { + "epoch": 0.11529133511054097, + "grad_norm": 2.471557855606079, + "learning_rate": 1.996000853992746e-05, + "loss": 3.3735, + "step": 1438 + }, + { + "epoch": 0.11537150989156361, + "grad_norm": 2.436279773712158, + "learning_rate": 1.9959892447886997e-05, + "loss": 3.1748, + "step": 1439 + }, + { + "epoch": 0.11545168467258624, + "grad_norm": 2.4538142681121826, + "learning_rate": 1.9959776187926256e-05, + "loss": 3.4141, + "step": 1440 + }, + { + "epoch": 0.11553185945360887, + "grad_norm": 2.393044948577881, + "learning_rate": 1.99596597600472e-05, + "loss": 3.4633, + "step": 1441 + }, + { + "epoch": 0.1156120342346315, + "grad_norm": 2.507683038711548, + "learning_rate": 1.9959543164251793e-05, + "loss": 3.4044, + "step": 1442 + }, + { + "epoch": 0.11569220901565412, + "grad_norm": 2.3354713916778564, + "learning_rate": 1.9959426400541994e-05, + "loss": 3.5493, + "step": 1443 + }, + { + "epoch": 0.11577238379667676, + "grad_norm": 2.4392833709716797, + "learning_rate": 1.995930946891978e-05, + "loss": 3.326, + "step": 1444 + }, + { + "epoch": 0.11585255857769938, + "grad_norm": 2.3564343452453613, + "learning_rate": 1.9959192369387113e-05, + "loss": 3.4249, + "step": 1445 + }, + { + "epoch": 0.11593273335872202, + "grad_norm": 2.439284563064575, + "learning_rate": 1.9959075101945977e-05, + "loss": 3.384, + "step": 1446 + }, + { + "epoch": 0.11601290813974464, + "grad_norm": 2.4163646697998047, + "learning_rate": 1.9958957666598343e-05, + "loss": 3.3715, + "step": 1447 + }, + { + "epoch": 0.11609308292076727, + "grad_norm": 2.3056020736694336, + "learning_rate": 1.9958840063346195e-05, + "loss": 3.4849, + "step": 1448 + }, + { + "epoch": 0.1161732577017899, + "grad_norm": 2.575815439224243, + "learning_rate": 1.995872229219151e-05, + "loss": 3.2683, + "step": 1449 + }, + { + "epoch": 0.11625343248281253, + "grad_norm": 2.256056785583496, + "learning_rate": 1.9958604353136276e-05, + "loss": 3.3661, + "step": 1450 + }, + { + "epoch": 0.11633360726383517, + "grad_norm": 2.4150781631469727, + "learning_rate": 1.9958486246182483e-05, + "loss": 3.4709, + "step": 1451 + }, + { + "epoch": 0.11641378204485779, + "grad_norm": 2.713209629058838, + "learning_rate": 1.9958367971332127e-05, + "loss": 3.374, + "step": 1452 + }, + { + "epoch": 0.11649395682588042, + "grad_norm": 2.3648910522460938, + "learning_rate": 1.9958249528587193e-05, + "loss": 3.4012, + "step": 1453 + }, + { + "epoch": 0.11657413160690305, + "grad_norm": 2.286106824874878, + "learning_rate": 1.995813091794968e-05, + "loss": 3.2878, + "step": 1454 + }, + { + "epoch": 0.11665430638792568, + "grad_norm": 2.772803783416748, + "learning_rate": 1.995801213942159e-05, + "loss": 3.4313, + "step": 1455 + }, + { + "epoch": 0.1167344811689483, + "grad_norm": 2.3082714080810547, + "learning_rate": 1.9957893193004926e-05, + "loss": 3.3141, + "step": 1456 + }, + { + "epoch": 0.11681465594997094, + "grad_norm": 2.5757033824920654, + "learning_rate": 1.9957774078701693e-05, + "loss": 3.4092, + "step": 1457 + }, + { + "epoch": 0.11689483073099356, + "grad_norm": 2.310429096221924, + "learning_rate": 1.9957654796513898e-05, + "loss": 3.4379, + "step": 1458 + }, + { + "epoch": 0.1169750055120162, + "grad_norm": 2.3961076736450195, + "learning_rate": 1.9957535346443552e-05, + "loss": 3.4072, + "step": 1459 + }, + { + "epoch": 0.11705518029303882, + "grad_norm": 2.5865063667297363, + "learning_rate": 1.995741572849267e-05, + "loss": 3.4036, + "step": 1460 + }, + { + "epoch": 0.11713535507406145, + "grad_norm": 2.3182947635650635, + "learning_rate": 1.9957295942663268e-05, + "loss": 3.3793, + "step": 1461 + }, + { + "epoch": 0.11721552985508409, + "grad_norm": 2.4245924949645996, + "learning_rate": 1.9957175988957364e-05, + "loss": 3.3386, + "step": 1462 + }, + { + "epoch": 0.11729570463610671, + "grad_norm": 2.417896032333374, + "learning_rate": 1.995705586737698e-05, + "loss": 3.4236, + "step": 1463 + }, + { + "epoch": 0.11737587941712935, + "grad_norm": 2.332829236984253, + "learning_rate": 1.9956935577924147e-05, + "loss": 3.3749, + "step": 1464 + }, + { + "epoch": 0.11745605419815197, + "grad_norm": 2.5655930042266846, + "learning_rate": 1.995681512060089e-05, + "loss": 3.3018, + "step": 1465 + }, + { + "epoch": 0.1175362289791746, + "grad_norm": 1.644828200340271, + "learning_rate": 1.995669449540924e-05, + "loss": 2.6245, + "step": 1466 + }, + { + "epoch": 0.11761640376019723, + "grad_norm": 2.5023975372314453, + "learning_rate": 1.9956573702351227e-05, + "loss": 3.4038, + "step": 1467 + }, + { + "epoch": 0.11769657854121986, + "grad_norm": 2.3149991035461426, + "learning_rate": 1.995645274142889e-05, + "loss": 3.3379, + "step": 1468 + }, + { + "epoch": 0.1177767533222425, + "grad_norm": 2.605933904647827, + "learning_rate": 1.995633161264427e-05, + "loss": 3.3142, + "step": 1469 + }, + { + "epoch": 0.11785692810326512, + "grad_norm": 2.383366346359253, + "learning_rate": 1.9956210315999408e-05, + "loss": 3.3407, + "step": 1470 + }, + { + "epoch": 0.11793710288428774, + "grad_norm": 2.4652347564697266, + "learning_rate": 1.995608885149635e-05, + "loss": 3.2931, + "step": 1471 + }, + { + "epoch": 0.11801727766531038, + "grad_norm": 2.4868106842041016, + "learning_rate": 1.9955967219137144e-05, + "loss": 3.4863, + "step": 1472 + }, + { + "epoch": 0.118097452446333, + "grad_norm": 2.467053174972534, + "learning_rate": 1.9955845418923837e-05, + "loss": 3.394, + "step": 1473 + }, + { + "epoch": 0.11817762722735564, + "grad_norm": 2.6038031578063965, + "learning_rate": 1.9955723450858486e-05, + "loss": 3.2303, + "step": 1474 + }, + { + "epoch": 0.11825780200837827, + "grad_norm": 2.7078118324279785, + "learning_rate": 1.9955601314943145e-05, + "loss": 3.5552, + "step": 1475 + }, + { + "epoch": 0.11833797678940089, + "grad_norm": 2.3479700088500977, + "learning_rate": 1.9955479011179878e-05, + "loss": 3.3274, + "step": 1476 + }, + { + "epoch": 0.11841815157042353, + "grad_norm": 2.64113187789917, + "learning_rate": 1.995535653957074e-05, + "loss": 3.3573, + "step": 1477 + }, + { + "epoch": 0.11849832635144615, + "grad_norm": 2.845587730407715, + "learning_rate": 1.9955233900117798e-05, + "loss": 3.2391, + "step": 1478 + }, + { + "epoch": 0.11857850113246879, + "grad_norm": 2.6502885818481445, + "learning_rate": 1.9955111092823123e-05, + "loss": 3.5227, + "step": 1479 + }, + { + "epoch": 0.11865867591349141, + "grad_norm": 2.464036703109741, + "learning_rate": 1.9954988117688786e-05, + "loss": 3.4476, + "step": 1480 + }, + { + "epoch": 0.11873885069451404, + "grad_norm": 2.4524855613708496, + "learning_rate": 1.9954864974716855e-05, + "loss": 3.3525, + "step": 1481 + }, + { + "epoch": 0.11881902547553667, + "grad_norm": 2.4806392192840576, + "learning_rate": 1.995474166390941e-05, + "loss": 3.4783, + "step": 1482 + }, + { + "epoch": 0.1188992002565593, + "grad_norm": 2.469050407409668, + "learning_rate": 1.9954618185268528e-05, + "loss": 3.3314, + "step": 1483 + }, + { + "epoch": 0.11897937503758192, + "grad_norm": 2.6576032638549805, + "learning_rate": 1.995449453879629e-05, + "loss": 3.265, + "step": 1484 + }, + { + "epoch": 0.11905954981860456, + "grad_norm": 2.2263023853302, + "learning_rate": 1.9954370724494786e-05, + "loss": 3.3011, + "step": 1485 + }, + { + "epoch": 0.11913972459962718, + "grad_norm": 2.3258743286132812, + "learning_rate": 1.9954246742366098e-05, + "loss": 3.4055, + "step": 1486 + }, + { + "epoch": 0.11921989938064982, + "grad_norm": 2.3953866958618164, + "learning_rate": 1.995412259241232e-05, + "loss": 3.2997, + "step": 1487 + }, + { + "epoch": 0.11930007416167245, + "grad_norm": 2.419456958770752, + "learning_rate": 1.9953998274635543e-05, + "loss": 3.3835, + "step": 1488 + }, + { + "epoch": 0.11938024894269507, + "grad_norm": 2.414754867553711, + "learning_rate": 1.995387378903786e-05, + "loss": 3.3225, + "step": 1489 + }, + { + "epoch": 0.1194604237237177, + "grad_norm": 2.3457694053649902, + "learning_rate": 1.9953749135621375e-05, + "loss": 3.2345, + "step": 1490 + }, + { + "epoch": 0.11954059850474033, + "grad_norm": 2.603195905685425, + "learning_rate": 1.9953624314388188e-05, + "loss": 3.349, + "step": 1491 + }, + { + "epoch": 0.11962077328576297, + "grad_norm": 2.487447738647461, + "learning_rate": 1.9953499325340404e-05, + "loss": 3.4249, + "step": 1492 + }, + { + "epoch": 0.11970094806678559, + "grad_norm": 2.4204392433166504, + "learning_rate": 1.9953374168480125e-05, + "loss": 3.2992, + "step": 1493 + }, + { + "epoch": 0.11978112284780822, + "grad_norm": 2.5331151485443115, + "learning_rate": 1.995324884380947e-05, + "loss": 3.3304, + "step": 1494 + }, + { + "epoch": 0.11986129762883085, + "grad_norm": 2.33488130569458, + "learning_rate": 1.9953123351330547e-05, + "loss": 3.4714, + "step": 1495 + }, + { + "epoch": 0.11994147240985348, + "grad_norm": 1.5478119850158691, + "learning_rate": 1.995299769104547e-05, + "loss": 2.6024, + "step": 1496 + }, + { + "epoch": 0.12002164719087612, + "grad_norm": 1.4209887981414795, + "learning_rate": 1.995287186295636e-05, + "loss": 2.6034, + "step": 1497 + }, + { + "epoch": 0.12010182197189874, + "grad_norm": 2.372980833053589, + "learning_rate": 1.995274586706534e-05, + "loss": 3.3226, + "step": 1498 + }, + { + "epoch": 0.12018199675292136, + "grad_norm": 2.659360408782959, + "learning_rate": 1.995261970337453e-05, + "loss": 3.333, + "step": 1499 + }, + { + "epoch": 0.120262171533944, + "grad_norm": 2.307072877883911, + "learning_rate": 1.9952493371886058e-05, + "loss": 3.3687, + "step": 1500 + }, + { + "epoch": 0.12034234631496663, + "grad_norm": 1.472861886024475, + "learning_rate": 1.995236687260206e-05, + "loss": 2.6307, + "step": 1501 + }, + { + "epoch": 0.12042252109598926, + "grad_norm": 2.3585188388824463, + "learning_rate": 1.9952240205524662e-05, + "loss": 3.4536, + "step": 1502 + }, + { + "epoch": 0.12050269587701189, + "grad_norm": 2.567256212234497, + "learning_rate": 1.9952113370656004e-05, + "loss": 3.4652, + "step": 1503 + }, + { + "epoch": 0.12058287065803451, + "grad_norm": 2.1602983474731445, + "learning_rate": 1.9951986367998217e-05, + "loss": 3.3252, + "step": 1504 + }, + { + "epoch": 0.12066304543905715, + "grad_norm": 2.492642879486084, + "learning_rate": 1.9951859197553453e-05, + "loss": 3.4413, + "step": 1505 + }, + { + "epoch": 0.12074322022007977, + "grad_norm": 2.3143184185028076, + "learning_rate": 1.9951731859323845e-05, + "loss": 3.3533, + "step": 1506 + }, + { + "epoch": 0.12082339500110241, + "grad_norm": 2.1324844360351562, + "learning_rate": 1.9951604353311545e-05, + "loss": 3.3274, + "step": 1507 + }, + { + "epoch": 0.12090356978212503, + "grad_norm": 2.2979938983917236, + "learning_rate": 1.9951476679518705e-05, + "loss": 3.3022, + "step": 1508 + }, + { + "epoch": 0.12098374456314766, + "grad_norm": 2.370359420776367, + "learning_rate": 1.995134883794748e-05, + "loss": 3.2617, + "step": 1509 + }, + { + "epoch": 0.1210639193441703, + "grad_norm": 2.2410943508148193, + "learning_rate": 1.9951220828600014e-05, + "loss": 3.3685, + "step": 1510 + }, + { + "epoch": 0.12114409412519292, + "grad_norm": 2.792457103729248, + "learning_rate": 1.9951092651478475e-05, + "loss": 3.4579, + "step": 1511 + }, + { + "epoch": 0.12122426890621554, + "grad_norm": 2.197460651397705, + "learning_rate": 1.9950964306585023e-05, + "loss": 3.2441, + "step": 1512 + }, + { + "epoch": 0.12130444368723818, + "grad_norm": 2.619368314743042, + "learning_rate": 1.9950835793921818e-05, + "loss": 3.36, + "step": 1513 + }, + { + "epoch": 0.1213846184682608, + "grad_norm": 2.4784393310546875, + "learning_rate": 1.9950707113491025e-05, + "loss": 3.2743, + "step": 1514 + }, + { + "epoch": 0.12146479324928344, + "grad_norm": 1.577335000038147, + "learning_rate": 1.995057826529482e-05, + "loss": 2.6273, + "step": 1515 + }, + { + "epoch": 0.12154496803030607, + "grad_norm": 2.357057809829712, + "learning_rate": 1.995044924933537e-05, + "loss": 3.2402, + "step": 1516 + }, + { + "epoch": 0.12162514281132869, + "grad_norm": 2.3449795246124268, + "learning_rate": 1.9950320065614857e-05, + "loss": 3.3374, + "step": 1517 + }, + { + "epoch": 0.12170531759235133, + "grad_norm": 2.5192902088165283, + "learning_rate": 1.9950190714135453e-05, + "loss": 3.4188, + "step": 1518 + }, + { + "epoch": 0.12178549237337395, + "grad_norm": 2.665738582611084, + "learning_rate": 1.995006119489934e-05, + "loss": 3.366, + "step": 1519 + }, + { + "epoch": 0.12186566715439659, + "grad_norm": 2.2893733978271484, + "learning_rate": 1.9949931507908702e-05, + "loss": 3.252, + "step": 1520 + }, + { + "epoch": 0.12194584193541921, + "grad_norm": 2.423682928085327, + "learning_rate": 1.9949801653165726e-05, + "loss": 3.376, + "step": 1521 + }, + { + "epoch": 0.12202601671644184, + "grad_norm": 2.5157830715179443, + "learning_rate": 1.9949671630672598e-05, + "loss": 3.3856, + "step": 1522 + }, + { + "epoch": 0.12210619149746448, + "grad_norm": 2.5351624488830566, + "learning_rate": 1.9949541440431516e-05, + "loss": 3.3355, + "step": 1523 + }, + { + "epoch": 0.1221863662784871, + "grad_norm": 2.471149444580078, + "learning_rate": 1.994941108244467e-05, + "loss": 3.3583, + "step": 1524 + }, + { + "epoch": 0.12226654105950974, + "grad_norm": 2.4410572052001953, + "learning_rate": 1.9949280556714262e-05, + "loss": 3.309, + "step": 1525 + }, + { + "epoch": 0.12234671584053236, + "grad_norm": 2.4992518424987793, + "learning_rate": 1.9949149863242487e-05, + "loss": 3.4147, + "step": 1526 + }, + { + "epoch": 0.12242689062155498, + "grad_norm": 2.3512754440307617, + "learning_rate": 1.9949019002031555e-05, + "loss": 3.3889, + "step": 1527 + }, + { + "epoch": 0.12250706540257762, + "grad_norm": 2.161555051803589, + "learning_rate": 1.9948887973083667e-05, + "loss": 3.3352, + "step": 1528 + }, + { + "epoch": 0.12258724018360025, + "grad_norm": 2.4207117557525635, + "learning_rate": 1.9948756776401032e-05, + "loss": 3.3108, + "step": 1529 + }, + { + "epoch": 0.12266741496462288, + "grad_norm": 2.3559460639953613, + "learning_rate": 1.9948625411985867e-05, + "loss": 3.3748, + "step": 1530 + }, + { + "epoch": 0.12274758974564551, + "grad_norm": 2.412036418914795, + "learning_rate": 1.9948493879840385e-05, + "loss": 3.4257, + "step": 1531 + }, + { + "epoch": 0.12282776452666813, + "grad_norm": 2.2908694744110107, + "learning_rate": 1.99483621799668e-05, + "loss": 3.4159, + "step": 1532 + }, + { + "epoch": 0.12290793930769077, + "grad_norm": 2.198009490966797, + "learning_rate": 1.9948230312367335e-05, + "loss": 3.3416, + "step": 1533 + }, + { + "epoch": 0.1229881140887134, + "grad_norm": 2.4922361373901367, + "learning_rate": 1.9948098277044213e-05, + "loss": 3.4169, + "step": 1534 + }, + { + "epoch": 0.12306828886973603, + "grad_norm": 2.582519292831421, + "learning_rate": 1.994796607399966e-05, + "loss": 3.3379, + "step": 1535 + }, + { + "epoch": 0.12314846365075865, + "grad_norm": 2.3407375812530518, + "learning_rate": 1.9947833703235907e-05, + "loss": 3.4186, + "step": 1536 + }, + { + "epoch": 0.12322863843178128, + "grad_norm": 2.412700891494751, + "learning_rate": 1.994770116475518e-05, + "loss": 3.3232, + "step": 1537 + }, + { + "epoch": 0.12330881321280392, + "grad_norm": 2.3591983318328857, + "learning_rate": 1.9947568458559722e-05, + "loss": 3.3784, + "step": 1538 + }, + { + "epoch": 0.12338898799382654, + "grad_norm": 2.338805913925171, + "learning_rate": 1.9947435584651763e-05, + "loss": 3.4554, + "step": 1539 + }, + { + "epoch": 0.12346916277484916, + "grad_norm": 2.577728748321533, + "learning_rate": 1.9947302543033548e-05, + "loss": 3.39, + "step": 1540 + }, + { + "epoch": 0.1235493375558718, + "grad_norm": 2.327728033065796, + "learning_rate": 1.9947169333707315e-05, + "loss": 3.3525, + "step": 1541 + }, + { + "epoch": 0.12362951233689443, + "grad_norm": 2.239586591720581, + "learning_rate": 1.9947035956675314e-05, + "loss": 3.3667, + "step": 1542 + }, + { + "epoch": 0.12370968711791706, + "grad_norm": 2.61857008934021, + "learning_rate": 1.9946902411939792e-05, + "loss": 3.3699, + "step": 1543 + }, + { + "epoch": 0.12378986189893969, + "grad_norm": 2.3534903526306152, + "learning_rate": 1.9946768699503002e-05, + "loss": 3.3156, + "step": 1544 + }, + { + "epoch": 0.12387003667996231, + "grad_norm": 2.2205824851989746, + "learning_rate": 1.9946634819367197e-05, + "loss": 3.3456, + "step": 1545 + }, + { + "epoch": 0.12395021146098495, + "grad_norm": 2.502009630203247, + "learning_rate": 1.9946500771534634e-05, + "loss": 3.403, + "step": 1546 + }, + { + "epoch": 0.12403038624200757, + "grad_norm": 2.5394301414489746, + "learning_rate": 1.994636655600757e-05, + "loss": 3.3085, + "step": 1547 + }, + { + "epoch": 0.12411056102303021, + "grad_norm": 2.33440899848938, + "learning_rate": 1.9946232172788277e-05, + "loss": 3.3025, + "step": 1548 + }, + { + "epoch": 0.12419073580405283, + "grad_norm": 2.3171379566192627, + "learning_rate": 1.9946097621879014e-05, + "loss": 3.2827, + "step": 1549 + }, + { + "epoch": 0.12427091058507546, + "grad_norm": 2.6436655521392822, + "learning_rate": 1.994596290328205e-05, + "loss": 3.3896, + "step": 1550 + }, + { + "epoch": 0.1243510853660981, + "grad_norm": 2.467033863067627, + "learning_rate": 1.9945828016999654e-05, + "loss": 3.4703, + "step": 1551 + }, + { + "epoch": 0.12443126014712072, + "grad_norm": 2.261598587036133, + "learning_rate": 1.9945692963034105e-05, + "loss": 3.2631, + "step": 1552 + }, + { + "epoch": 0.12451143492814336, + "grad_norm": 2.4340784549713135, + "learning_rate": 1.9945557741387678e-05, + "loss": 3.5354, + "step": 1553 + }, + { + "epoch": 0.12459160970916598, + "grad_norm": 2.332502841949463, + "learning_rate": 1.9945422352062652e-05, + "loss": 3.3881, + "step": 1554 + }, + { + "epoch": 0.1246717844901886, + "grad_norm": 2.499232053756714, + "learning_rate": 1.9945286795061308e-05, + "loss": 3.3762, + "step": 1555 + }, + { + "epoch": 0.12475195927121124, + "grad_norm": 2.5436654090881348, + "learning_rate": 1.994515107038594e-05, + "loss": 3.3738, + "step": 1556 + }, + { + "epoch": 0.12483213405223387, + "grad_norm": 2.251526117324829, + "learning_rate": 1.9945015178038825e-05, + "loss": 3.3545, + "step": 1557 + }, + { + "epoch": 0.1249123088332565, + "grad_norm": 2.5009636878967285, + "learning_rate": 1.994487911802226e-05, + "loss": 3.3855, + "step": 1558 + }, + { + "epoch": 0.12499248361427913, + "grad_norm": 2.2395153045654297, + "learning_rate": 1.9944742890338537e-05, + "loss": 3.4434, + "step": 1559 + }, + { + "epoch": 0.12507265839530177, + "grad_norm": 2.332310914993286, + "learning_rate": 1.9944606494989956e-05, + "loss": 3.3055, + "step": 1560 + }, + { + "epoch": 0.1251528331763244, + "grad_norm": 2.7892544269561768, + "learning_rate": 1.9944469931978812e-05, + "loss": 3.3398, + "step": 1561 + }, + { + "epoch": 0.12523300795734701, + "grad_norm": 2.4104926586151123, + "learning_rate": 1.994433320130741e-05, + "loss": 3.4331, + "step": 1562 + }, + { + "epoch": 0.12531318273836964, + "grad_norm": 2.5447254180908203, + "learning_rate": 1.9944196302978058e-05, + "loss": 3.3129, + "step": 1563 + }, + { + "epoch": 0.12539335751939226, + "grad_norm": 1.493804931640625, + "learning_rate": 1.994405923699306e-05, + "loss": 2.606, + "step": 1564 + }, + { + "epoch": 0.12547353230041491, + "grad_norm": 2.3590784072875977, + "learning_rate": 1.994392200335473e-05, + "loss": 3.4249, + "step": 1565 + }, + { + "epoch": 0.12555370708143754, + "grad_norm": 2.752704381942749, + "learning_rate": 1.994378460206537e-05, + "loss": 3.2587, + "step": 1566 + }, + { + "epoch": 0.12563388186246016, + "grad_norm": 1.1951255798339844, + "learning_rate": 1.9943647033127315e-05, + "loss": 2.5953, + "step": 1567 + }, + { + "epoch": 0.12571405664348279, + "grad_norm": 2.4162988662719727, + "learning_rate": 1.9943509296542875e-05, + "loss": 3.3693, + "step": 1568 + }, + { + "epoch": 0.1257942314245054, + "grad_norm": 2.333048105239868, + "learning_rate": 1.994337139231437e-05, + "loss": 3.4215, + "step": 1569 + }, + { + "epoch": 0.12587440620552806, + "grad_norm": 2.342242479324341, + "learning_rate": 1.9943233320444127e-05, + "loss": 3.406, + "step": 1570 + }, + { + "epoch": 0.12595458098655068, + "grad_norm": 2.2906148433685303, + "learning_rate": 1.9943095080934478e-05, + "loss": 3.3655, + "step": 1571 + }, + { + "epoch": 0.1260347557675733, + "grad_norm": 3.030108690261841, + "learning_rate": 1.994295667378775e-05, + "loss": 3.3148, + "step": 1572 + }, + { + "epoch": 0.12611493054859593, + "grad_norm": 2.350456476211548, + "learning_rate": 1.994281809900627e-05, + "loss": 3.3738, + "step": 1573 + }, + { + "epoch": 0.12619510532961856, + "grad_norm": 2.3733201026916504, + "learning_rate": 1.994267935659239e-05, + "loss": 3.2842, + "step": 1574 + }, + { + "epoch": 0.1262752801106412, + "grad_norm": 2.316633462905884, + "learning_rate": 1.994254044654843e-05, + "loss": 3.3716, + "step": 1575 + }, + { + "epoch": 0.12635545489166383, + "grad_norm": 2.241450786590576, + "learning_rate": 1.994240136887675e-05, + "loss": 3.3073, + "step": 1576 + }, + { + "epoch": 0.12643562967268646, + "grad_norm": 2.4020497798919678, + "learning_rate": 1.9942262123579688e-05, + "loss": 3.4376, + "step": 1577 + }, + { + "epoch": 0.12651580445370908, + "grad_norm": 2.6542115211486816, + "learning_rate": 1.994212271065959e-05, + "loss": 3.301, + "step": 1578 + }, + { + "epoch": 0.1265959792347317, + "grad_norm": 1.4188299179077148, + "learning_rate": 1.9941983130118803e-05, + "loss": 2.5875, + "step": 1579 + }, + { + "epoch": 0.12667615401575436, + "grad_norm": 2.5918049812316895, + "learning_rate": 1.9941843381959686e-05, + "loss": 3.2521, + "step": 1580 + }, + { + "epoch": 0.12675632879677698, + "grad_norm": 2.6035730838775635, + "learning_rate": 1.9941703466184594e-05, + "loss": 3.3619, + "step": 1581 + }, + { + "epoch": 0.1268365035777996, + "grad_norm": 2.3485822677612305, + "learning_rate": 1.9941563382795884e-05, + "loss": 3.3218, + "step": 1582 + }, + { + "epoch": 0.12691667835882223, + "grad_norm": 2.3335330486297607, + "learning_rate": 1.994142313179592e-05, + "loss": 3.4248, + "step": 1583 + }, + { + "epoch": 0.12699685313984485, + "grad_norm": 2.30454158782959, + "learning_rate": 1.9941282713187067e-05, + "loss": 3.3484, + "step": 1584 + }, + { + "epoch": 0.1270770279208675, + "grad_norm": 1.483797311782837, + "learning_rate": 1.9941142126971692e-05, + "loss": 2.5901, + "step": 1585 + }, + { + "epoch": 0.12715720270189013, + "grad_norm": 2.3260042667388916, + "learning_rate": 1.994100137315216e-05, + "loss": 3.3925, + "step": 1586 + }, + { + "epoch": 0.12723737748291275, + "grad_norm": 2.134154796600342, + "learning_rate": 1.994086045173085e-05, + "loss": 3.2418, + "step": 1587 + }, + { + "epoch": 0.12731755226393537, + "grad_norm": 2.4545624256134033, + "learning_rate": 1.9940719362710136e-05, + "loss": 3.3447, + "step": 1588 + }, + { + "epoch": 0.127397727044958, + "grad_norm": 2.2991998195648193, + "learning_rate": 1.99405781060924e-05, + "loss": 3.3822, + "step": 1589 + }, + { + "epoch": 0.12747790182598065, + "grad_norm": 2.2080280780792236, + "learning_rate": 1.994043668188002e-05, + "loss": 3.2981, + "step": 1590 + }, + { + "epoch": 0.12755807660700327, + "grad_norm": 2.4005000591278076, + "learning_rate": 1.994029509007538e-05, + "loss": 3.3635, + "step": 1591 + }, + { + "epoch": 0.1276382513880259, + "grad_norm": 2.5113418102264404, + "learning_rate": 1.9940153330680867e-05, + "loss": 3.4379, + "step": 1592 + }, + { + "epoch": 0.12771842616904852, + "grad_norm": 1.4569095373153687, + "learning_rate": 1.994001140369887e-05, + "loss": 2.5867, + "step": 1593 + }, + { + "epoch": 0.12779860095007115, + "grad_norm": 2.5470221042633057, + "learning_rate": 1.9939869309131786e-05, + "loss": 3.3608, + "step": 1594 + }, + { + "epoch": 0.1278787757310938, + "grad_norm": 2.625392198562622, + "learning_rate": 1.9939727046982013e-05, + "loss": 3.3588, + "step": 1595 + }, + { + "epoch": 0.12795895051211642, + "grad_norm": 2.1629843711853027, + "learning_rate": 1.993958461725194e-05, + "loss": 3.3401, + "step": 1596 + }, + { + "epoch": 0.12803912529313904, + "grad_norm": 2.4367525577545166, + "learning_rate": 1.9939442019943973e-05, + "loss": 3.3645, + "step": 1597 + }, + { + "epoch": 0.12811930007416167, + "grad_norm": 2.245635986328125, + "learning_rate": 1.993929925506052e-05, + "loss": 3.3047, + "step": 1598 + }, + { + "epoch": 0.1281994748551843, + "grad_norm": 2.2934436798095703, + "learning_rate": 1.993915632260398e-05, + "loss": 3.2721, + "step": 1599 + }, + { + "epoch": 0.12827964963620694, + "grad_norm": 2.4464962482452393, + "learning_rate": 1.993901322257677e-05, + "loss": 3.444, + "step": 1600 + }, + { + "epoch": 0.12835982441722957, + "grad_norm": 2.2942299842834473, + "learning_rate": 1.99388699549813e-05, + "loss": 3.3523, + "step": 1601 + }, + { + "epoch": 0.1284399991982522, + "grad_norm": 2.38696026802063, + "learning_rate": 1.9938726519819987e-05, + "loss": 3.3102, + "step": 1602 + }, + { + "epoch": 0.12852017397927482, + "grad_norm": 2.2965314388275146, + "learning_rate": 1.9938582917095247e-05, + "loss": 3.3375, + "step": 1603 + }, + { + "epoch": 0.12860034876029744, + "grad_norm": 2.464589834213257, + "learning_rate": 1.9938439146809497e-05, + "loss": 3.3351, + "step": 1604 + }, + { + "epoch": 0.1286805235413201, + "grad_norm": 2.4616730213165283, + "learning_rate": 1.993829520896517e-05, + "loss": 3.3923, + "step": 1605 + }, + { + "epoch": 0.12876069832234271, + "grad_norm": 2.4840760231018066, + "learning_rate": 1.993815110356469e-05, + "loss": 3.3392, + "step": 1606 + }, + { + "epoch": 0.12884087310336534, + "grad_norm": 2.426443099975586, + "learning_rate": 1.993800683061048e-05, + "loss": 3.3423, + "step": 1607 + }, + { + "epoch": 0.12892104788438796, + "grad_norm": 2.382187843322754, + "learning_rate": 1.993786239010498e-05, + "loss": 3.453, + "step": 1608 + }, + { + "epoch": 0.1290012226654106, + "grad_norm": 2.2293498516082764, + "learning_rate": 1.9937717782050623e-05, + "loss": 3.2633, + "step": 1609 + }, + { + "epoch": 0.12908139744643324, + "grad_norm": 2.2938578128814697, + "learning_rate": 1.9937573006449847e-05, + "loss": 3.3583, + "step": 1610 + }, + { + "epoch": 0.12916157222745586, + "grad_norm": 2.52374267578125, + "learning_rate": 1.993742806330509e-05, + "loss": 3.324, + "step": 1611 + }, + { + "epoch": 0.12924174700847849, + "grad_norm": 2.4084458351135254, + "learning_rate": 1.99372829526188e-05, + "loss": 3.3539, + "step": 1612 + }, + { + "epoch": 0.1293219217895011, + "grad_norm": 2.3681628704071045, + "learning_rate": 1.993713767439342e-05, + "loss": 3.2723, + "step": 1613 + }, + { + "epoch": 0.12940209657052373, + "grad_norm": 1.1948190927505493, + "learning_rate": 1.9936992228631405e-05, + "loss": 2.5905, + "step": 1614 + }, + { + "epoch": 0.12948227135154636, + "grad_norm": 2.1834006309509277, + "learning_rate": 1.99368466153352e-05, + "loss": 3.3529, + "step": 1615 + }, + { + "epoch": 0.129562446132569, + "grad_norm": 2.4522335529327393, + "learning_rate": 1.9936700834507263e-05, + "loss": 3.25, + "step": 1616 + }, + { + "epoch": 0.12964262091359163, + "grad_norm": 2.411284923553467, + "learning_rate": 1.9936554886150053e-05, + "loss": 3.4064, + "step": 1617 + }, + { + "epoch": 0.12972279569461426, + "grad_norm": 2.4513702392578125, + "learning_rate": 1.9936408770266026e-05, + "loss": 3.369, + "step": 1618 + }, + { + "epoch": 0.12980297047563688, + "grad_norm": 2.287932872772217, + "learning_rate": 1.9936262486857655e-05, + "loss": 3.371, + "step": 1619 + }, + { + "epoch": 0.1298831452566595, + "grad_norm": 2.370670795440674, + "learning_rate": 1.9936116035927398e-05, + "loss": 3.2164, + "step": 1620 + }, + { + "epoch": 0.12996332003768216, + "grad_norm": 2.315488815307617, + "learning_rate": 1.9935969417477724e-05, + "loss": 3.3058, + "step": 1621 + }, + { + "epoch": 0.13004349481870478, + "grad_norm": 2.3929355144500732, + "learning_rate": 1.993582263151111e-05, + "loss": 3.3419, + "step": 1622 + }, + { + "epoch": 0.1301236695997274, + "grad_norm": 2.429497003555298, + "learning_rate": 1.9935675678030028e-05, + "loss": 3.283, + "step": 1623 + }, + { + "epoch": 0.13020384438075003, + "grad_norm": 2.423508644104004, + "learning_rate": 1.9935528557036953e-05, + "loss": 3.2506, + "step": 1624 + }, + { + "epoch": 0.13028401916177265, + "grad_norm": 2.320374011993408, + "learning_rate": 1.9935381268534374e-05, + "loss": 3.2772, + "step": 1625 + }, + { + "epoch": 0.1303641939427953, + "grad_norm": 2.5845797061920166, + "learning_rate": 1.9935233812524763e-05, + "loss": 3.3347, + "step": 1626 + }, + { + "epoch": 0.13044436872381793, + "grad_norm": 2.2326180934906006, + "learning_rate": 1.9935086189010615e-05, + "loss": 3.3173, + "step": 1627 + }, + { + "epoch": 0.13052454350484055, + "grad_norm": 2.420473337173462, + "learning_rate": 1.9934938397994414e-05, + "loss": 3.379, + "step": 1628 + }, + { + "epoch": 0.13060471828586317, + "grad_norm": 2.3737924098968506, + "learning_rate": 1.993479043947865e-05, + "loss": 3.3254, + "step": 1629 + }, + { + "epoch": 0.1306848930668858, + "grad_norm": 2.392913818359375, + "learning_rate": 1.9934642313465825e-05, + "loss": 3.3683, + "step": 1630 + }, + { + "epoch": 0.13076506784790845, + "grad_norm": 2.3190841674804688, + "learning_rate": 1.993449401995843e-05, + "loss": 3.334, + "step": 1631 + }, + { + "epoch": 0.13084524262893107, + "grad_norm": 2.5383074283599854, + "learning_rate": 1.9934345558958966e-05, + "loss": 3.3167, + "step": 1632 + }, + { + "epoch": 0.1309254174099537, + "grad_norm": 2.519054651260376, + "learning_rate": 1.9934196930469937e-05, + "loss": 3.4381, + "step": 1633 + }, + { + "epoch": 0.13100559219097632, + "grad_norm": 2.147096633911133, + "learning_rate": 1.9934048134493854e-05, + "loss": 3.3685, + "step": 1634 + }, + { + "epoch": 0.13108576697199895, + "grad_norm": 2.464359760284424, + "learning_rate": 1.9933899171033214e-05, + "loss": 3.3932, + "step": 1635 + }, + { + "epoch": 0.1311659417530216, + "grad_norm": 2.226935386657715, + "learning_rate": 1.993375004009054e-05, + "loss": 3.2943, + "step": 1636 + }, + { + "epoch": 0.13124611653404422, + "grad_norm": 2.3091824054718018, + "learning_rate": 1.993360074166834e-05, + "loss": 3.2795, + "step": 1637 + }, + { + "epoch": 0.13132629131506685, + "grad_norm": 2.405766725540161, + "learning_rate": 1.9933451275769127e-05, + "loss": 3.2083, + "step": 1638 + }, + { + "epoch": 0.13140646609608947, + "grad_norm": 2.3994593620300293, + "learning_rate": 1.9933301642395432e-05, + "loss": 3.3724, + "step": 1639 + }, + { + "epoch": 0.1314866408771121, + "grad_norm": 2.4579858779907227, + "learning_rate": 1.993315184154977e-05, + "loss": 3.3778, + "step": 1640 + }, + { + "epoch": 0.13156681565813474, + "grad_norm": 2.3424980640411377, + "learning_rate": 1.993300187323467e-05, + "loss": 3.3325, + "step": 1641 + }, + { + "epoch": 0.13164699043915737, + "grad_norm": 2.505296230316162, + "learning_rate": 1.993285173745266e-05, + "loss": 3.5017, + "step": 1642 + }, + { + "epoch": 0.13172716522018, + "grad_norm": 2.4058637619018555, + "learning_rate": 1.993270143420627e-05, + "loss": 3.4329, + "step": 1643 + }, + { + "epoch": 0.13180734000120262, + "grad_norm": 2.3318285942077637, + "learning_rate": 1.9932550963498032e-05, + "loss": 3.278, + "step": 1644 + }, + { + "epoch": 0.13188751478222524, + "grad_norm": 2.2431747913360596, + "learning_rate": 1.9932400325330487e-05, + "loss": 3.4626, + "step": 1645 + }, + { + "epoch": 0.1319676895632479, + "grad_norm": 2.3628222942352295, + "learning_rate": 1.993224951970617e-05, + "loss": 3.359, + "step": 1646 + }, + { + "epoch": 0.13204786434427052, + "grad_norm": 1.392834186553955, + "learning_rate": 1.9932098546627633e-05, + "loss": 2.6191, + "step": 1647 + }, + { + "epoch": 0.13212803912529314, + "grad_norm": 2.4933252334594727, + "learning_rate": 1.993194740609741e-05, + "loss": 3.4147, + "step": 1648 + }, + { + "epoch": 0.13220821390631576, + "grad_norm": 2.501572370529175, + "learning_rate": 1.9931796098118057e-05, + "loss": 3.2225, + "step": 1649 + }, + { + "epoch": 0.1322883886873384, + "grad_norm": 2.302245616912842, + "learning_rate": 1.9931644622692122e-05, + "loss": 3.4117, + "step": 1650 + }, + { + "epoch": 0.13236856346836104, + "grad_norm": 2.558363199234009, + "learning_rate": 1.9931492979822155e-05, + "loss": 3.428, + "step": 1651 + }, + { + "epoch": 0.13244873824938366, + "grad_norm": 2.235873222351074, + "learning_rate": 1.993134116951072e-05, + "loss": 3.2545, + "step": 1652 + }, + { + "epoch": 0.1325289130304063, + "grad_norm": 2.246251344680786, + "learning_rate": 1.993118919176037e-05, + "loss": 3.2905, + "step": 1653 + }, + { + "epoch": 0.1326090878114289, + "grad_norm": 2.600160837173462, + "learning_rate": 1.9931037046573674e-05, + "loss": 3.3116, + "step": 1654 + }, + { + "epoch": 0.13268926259245153, + "grad_norm": 2.916903495788574, + "learning_rate": 1.9930884733953194e-05, + "loss": 3.3506, + "step": 1655 + }, + { + "epoch": 0.1327694373734742, + "grad_norm": 2.395906925201416, + "learning_rate": 1.9930732253901493e-05, + "loss": 3.3234, + "step": 1656 + }, + { + "epoch": 0.1328496121544968, + "grad_norm": 1.1912308931350708, + "learning_rate": 1.993057960642115e-05, + "loss": 2.5762, + "step": 1657 + }, + { + "epoch": 0.13292978693551943, + "grad_norm": 2.2837493419647217, + "learning_rate": 1.9930426791514734e-05, + "loss": 3.3195, + "step": 1658 + }, + { + "epoch": 0.13300996171654206, + "grad_norm": 2.254809617996216, + "learning_rate": 1.993027380918482e-05, + "loss": 3.33, + "step": 1659 + }, + { + "epoch": 0.13309013649756468, + "grad_norm": 2.4115617275238037, + "learning_rate": 1.993012065943399e-05, + "loss": 3.2952, + "step": 1660 + }, + { + "epoch": 0.13317031127858733, + "grad_norm": 2.2441861629486084, + "learning_rate": 1.9929967342264825e-05, + "loss": 3.3064, + "step": 1661 + }, + { + "epoch": 0.13325048605960996, + "grad_norm": 2.2488210201263428, + "learning_rate": 1.9929813857679913e-05, + "loss": 3.3726, + "step": 1662 + }, + { + "epoch": 0.13333066084063258, + "grad_norm": 2.3906874656677246, + "learning_rate": 1.9929660205681836e-05, + "loss": 3.2637, + "step": 1663 + }, + { + "epoch": 0.1334108356216552, + "grad_norm": 2.2746551036834717, + "learning_rate": 1.992950638627319e-05, + "loss": 3.2902, + "step": 1664 + }, + { + "epoch": 0.13349101040267783, + "grad_norm": 2.4925458431243896, + "learning_rate": 1.9929352399456565e-05, + "loss": 3.4329, + "step": 1665 + }, + { + "epoch": 0.13357118518370048, + "grad_norm": 2.3143773078918457, + "learning_rate": 1.9929198245234554e-05, + "loss": 3.3749, + "step": 1666 + }, + { + "epoch": 0.1336513599647231, + "grad_norm": 2.319312572479248, + "learning_rate": 1.9929043923609767e-05, + "loss": 3.3525, + "step": 1667 + }, + { + "epoch": 0.13373153474574573, + "grad_norm": 2.3038926124572754, + "learning_rate": 1.9928889434584794e-05, + "loss": 3.3571, + "step": 1668 + }, + { + "epoch": 0.13381170952676835, + "grad_norm": 2.27983021736145, + "learning_rate": 1.9928734778162248e-05, + "loss": 3.389, + "step": 1669 + }, + { + "epoch": 0.13389188430779098, + "grad_norm": 1.295944094657898, + "learning_rate": 1.992857995434473e-05, + "loss": 2.5829, + "step": 1670 + }, + { + "epoch": 0.1339720590888136, + "grad_norm": 2.608297348022461, + "learning_rate": 1.992842496313485e-05, + "loss": 3.2434, + "step": 1671 + }, + { + "epoch": 0.13405223386983625, + "grad_norm": 2.3766138553619385, + "learning_rate": 1.992826980453523e-05, + "loss": 3.4086, + "step": 1672 + }, + { + "epoch": 0.13413240865085888, + "grad_norm": 2.4136252403259277, + "learning_rate": 1.9928114478548477e-05, + "loss": 3.2992, + "step": 1673 + }, + { + "epoch": 0.1342125834318815, + "grad_norm": 2.283705711364746, + "learning_rate": 1.9927958985177212e-05, + "loss": 3.3959, + "step": 1674 + }, + { + "epoch": 0.13429275821290412, + "grad_norm": 2.6653902530670166, + "learning_rate": 1.9927803324424065e-05, + "loss": 3.3691, + "step": 1675 + }, + { + "epoch": 0.13437293299392675, + "grad_norm": 2.767446756362915, + "learning_rate": 1.9927647496291644e-05, + "loss": 3.4034, + "step": 1676 + }, + { + "epoch": 0.1344531077749494, + "grad_norm": 2.439509868621826, + "learning_rate": 1.992749150078259e-05, + "loss": 3.4274, + "step": 1677 + }, + { + "epoch": 0.13453328255597202, + "grad_norm": 2.2752585411071777, + "learning_rate": 1.9927335337899526e-05, + "loss": 3.4323, + "step": 1678 + }, + { + "epoch": 0.13461345733699465, + "grad_norm": 2.047318458557129, + "learning_rate": 1.9927179007645086e-05, + "loss": 3.2772, + "step": 1679 + }, + { + "epoch": 0.13469363211801727, + "grad_norm": 2.484761953353882, + "learning_rate": 1.992702251002191e-05, + "loss": 3.3605, + "step": 1680 + }, + { + "epoch": 0.1347738068990399, + "grad_norm": 2.463597059249878, + "learning_rate": 1.992686584503263e-05, + "loss": 3.3856, + "step": 1681 + }, + { + "epoch": 0.13485398168006255, + "grad_norm": 1.275891900062561, + "learning_rate": 1.992670901267989e-05, + "loss": 2.5894, + "step": 1682 + }, + { + "epoch": 0.13493415646108517, + "grad_norm": 2.410836696624756, + "learning_rate": 1.992655201296634e-05, + "loss": 3.2758, + "step": 1683 + }, + { + "epoch": 0.1350143312421078, + "grad_norm": 2.584458112716675, + "learning_rate": 1.9926394845894616e-05, + "loss": 3.3864, + "step": 1684 + }, + { + "epoch": 0.13509450602313042, + "grad_norm": 2.24745512008667, + "learning_rate": 1.9926237511467374e-05, + "loss": 3.3575, + "step": 1685 + }, + { + "epoch": 0.13517468080415304, + "grad_norm": 2.31032133102417, + "learning_rate": 1.9926080009687266e-05, + "loss": 3.3221, + "step": 1686 + }, + { + "epoch": 0.1352548555851757, + "grad_norm": 2.3972315788269043, + "learning_rate": 1.992592234055695e-05, + "loss": 3.2889, + "step": 1687 + }, + { + "epoch": 0.13533503036619832, + "grad_norm": 2.35715389251709, + "learning_rate": 1.9925764504079078e-05, + "loss": 3.29, + "step": 1688 + }, + { + "epoch": 0.13541520514722094, + "grad_norm": 2.3095948696136475, + "learning_rate": 1.9925606500256318e-05, + "loss": 3.2522, + "step": 1689 + }, + { + "epoch": 0.13549537992824356, + "grad_norm": 2.4146728515625, + "learning_rate": 1.992544832909133e-05, + "loss": 3.3387, + "step": 1690 + }, + { + "epoch": 0.1355755547092662, + "grad_norm": 2.3118746280670166, + "learning_rate": 1.9925289990586777e-05, + "loss": 3.3105, + "step": 1691 + }, + { + "epoch": 0.13565572949028884, + "grad_norm": 2.4502170085906982, + "learning_rate": 1.9925131484745336e-05, + "loss": 3.3859, + "step": 1692 + }, + { + "epoch": 0.13573590427131146, + "grad_norm": 2.250684976577759, + "learning_rate": 1.9924972811569677e-05, + "loss": 3.4319, + "step": 1693 + }, + { + "epoch": 0.1358160790523341, + "grad_norm": 2.304602861404419, + "learning_rate": 1.9924813971062472e-05, + "loss": 3.403, + "step": 1694 + }, + { + "epoch": 0.1358962538333567, + "grad_norm": 2.2379894256591797, + "learning_rate": 1.9924654963226403e-05, + "loss": 3.3826, + "step": 1695 + }, + { + "epoch": 0.13597642861437934, + "grad_norm": 1.113734483718872, + "learning_rate": 1.9924495788064148e-05, + "loss": 2.5863, + "step": 1696 + }, + { + "epoch": 0.136056603395402, + "grad_norm": 2.510230541229248, + "learning_rate": 1.9924336445578396e-05, + "loss": 3.2745, + "step": 1697 + }, + { + "epoch": 0.1361367781764246, + "grad_norm": 2.4382808208465576, + "learning_rate": 1.992417693577182e-05, + "loss": 3.4032, + "step": 1698 + }, + { + "epoch": 0.13621695295744723, + "grad_norm": 2.1612565517425537, + "learning_rate": 1.9924017258647125e-05, + "loss": 3.4287, + "step": 1699 + }, + { + "epoch": 0.13629712773846986, + "grad_norm": 2.284189224243164, + "learning_rate": 1.9923857414207e-05, + "loss": 3.4562, + "step": 1700 + }, + { + "epoch": 0.13637730251949248, + "grad_norm": 2.4355509281158447, + "learning_rate": 1.992369740245413e-05, + "loss": 3.4363, + "step": 1701 + }, + { + "epoch": 0.13645747730051513, + "grad_norm": 2.448188304901123, + "learning_rate": 1.992353722339122e-05, + "loss": 3.3676, + "step": 1702 + }, + { + "epoch": 0.13653765208153776, + "grad_norm": 2.2230842113494873, + "learning_rate": 1.9923376877020974e-05, + "loss": 3.2971, + "step": 1703 + }, + { + "epoch": 0.13661782686256038, + "grad_norm": 2.461671829223633, + "learning_rate": 1.992321636334609e-05, + "loss": 3.3647, + "step": 1704 + }, + { + "epoch": 0.136698001643583, + "grad_norm": 2.2328429222106934, + "learning_rate": 1.992305568236927e-05, + "loss": 3.3483, + "step": 1705 + }, + { + "epoch": 0.13677817642460563, + "grad_norm": 2.242177724838257, + "learning_rate": 1.9922894834093234e-05, + "loss": 3.3994, + "step": 1706 + }, + { + "epoch": 0.13685835120562828, + "grad_norm": 2.372765302658081, + "learning_rate": 1.992273381852069e-05, + "loss": 3.4163, + "step": 1707 + }, + { + "epoch": 0.1369385259866509, + "grad_norm": 2.4438250064849854, + "learning_rate": 1.9922572635654345e-05, + "loss": 3.3109, + "step": 1708 + }, + { + "epoch": 0.13701870076767353, + "grad_norm": 2.276784896850586, + "learning_rate": 1.9922411285496925e-05, + "loss": 3.4117, + "step": 1709 + }, + { + "epoch": 0.13709887554869615, + "grad_norm": 2.352701187133789, + "learning_rate": 1.9922249768051144e-05, + "loss": 3.3171, + "step": 1710 + }, + { + "epoch": 0.13717905032971878, + "grad_norm": 2.4100778102874756, + "learning_rate": 1.992208808331973e-05, + "loss": 3.332, + "step": 1711 + }, + { + "epoch": 0.13725922511074143, + "grad_norm": 2.51613187789917, + "learning_rate": 1.9921926231305414e-05, + "loss": 3.2845, + "step": 1712 + }, + { + "epoch": 0.13733939989176405, + "grad_norm": 2.2827203273773193, + "learning_rate": 1.9921764212010918e-05, + "loss": 3.2723, + "step": 1713 + }, + { + "epoch": 0.13741957467278668, + "grad_norm": 2.3355214595794678, + "learning_rate": 1.992160202543897e-05, + "loss": 3.3188, + "step": 1714 + }, + { + "epoch": 0.1374997494538093, + "grad_norm": 2.191023349761963, + "learning_rate": 1.9921439671592307e-05, + "loss": 3.265, + "step": 1715 + }, + { + "epoch": 0.13757992423483192, + "grad_norm": 2.423666000366211, + "learning_rate": 1.9921277150473674e-05, + "loss": 3.2386, + "step": 1716 + }, + { + "epoch": 0.13766009901585458, + "grad_norm": 2.317856788635254, + "learning_rate": 1.9921114462085797e-05, + "loss": 3.4142, + "step": 1717 + }, + { + "epoch": 0.1377402737968772, + "grad_norm": 2.1977314949035645, + "learning_rate": 1.9920951606431433e-05, + "loss": 3.4143, + "step": 1718 + }, + { + "epoch": 0.13782044857789982, + "grad_norm": 2.2580618858337402, + "learning_rate": 1.9920788583513318e-05, + "loss": 3.4225, + "step": 1719 + }, + { + "epoch": 0.13790062335892245, + "grad_norm": 2.299143075942993, + "learning_rate": 1.9920625393334205e-05, + "loss": 3.4095, + "step": 1720 + }, + { + "epoch": 0.13798079813994507, + "grad_norm": 2.2678892612457275, + "learning_rate": 1.9920462035896846e-05, + "loss": 3.3048, + "step": 1721 + }, + { + "epoch": 0.13806097292096772, + "grad_norm": 2.6186370849609375, + "learning_rate": 1.992029851120399e-05, + "loss": 3.4569, + "step": 1722 + }, + { + "epoch": 0.13814114770199035, + "grad_norm": 2.2323951721191406, + "learning_rate": 1.99201348192584e-05, + "loss": 3.3538, + "step": 1723 + }, + { + "epoch": 0.13822132248301297, + "grad_norm": 2.2834715843200684, + "learning_rate": 1.991997096006283e-05, + "loss": 3.3268, + "step": 1724 + }, + { + "epoch": 0.1383014972640356, + "grad_norm": 2.2766709327697754, + "learning_rate": 1.991980693362005e-05, + "loss": 3.3181, + "step": 1725 + }, + { + "epoch": 0.13838167204505822, + "grad_norm": 2.2594263553619385, + "learning_rate": 1.9919642739932817e-05, + "loss": 3.311, + "step": 1726 + }, + { + "epoch": 0.13846184682608087, + "grad_norm": 2.297567844390869, + "learning_rate": 1.9919478379003907e-05, + "loss": 3.416, + "step": 1727 + }, + { + "epoch": 0.1385420216071035, + "grad_norm": 2.161713123321533, + "learning_rate": 1.9919313850836088e-05, + "loss": 3.4312, + "step": 1728 + }, + { + "epoch": 0.13862219638812612, + "grad_norm": 2.3506362438201904, + "learning_rate": 1.991914915543213e-05, + "loss": 3.347, + "step": 1729 + }, + { + "epoch": 0.13870237116914874, + "grad_norm": 2.3944127559661865, + "learning_rate": 1.9918984292794814e-05, + "loss": 3.4337, + "step": 1730 + }, + { + "epoch": 0.13878254595017137, + "grad_norm": 2.3973217010498047, + "learning_rate": 1.991881926292692e-05, + "loss": 3.392, + "step": 1731 + }, + { + "epoch": 0.138862720731194, + "grad_norm": 2.3993489742279053, + "learning_rate": 1.9918654065831228e-05, + "loss": 3.3655, + "step": 1732 + }, + { + "epoch": 0.13894289551221664, + "grad_norm": 2.174729824066162, + "learning_rate": 1.9918488701510525e-05, + "loss": 3.2915, + "step": 1733 + }, + { + "epoch": 0.13902307029323926, + "grad_norm": 2.2808873653411865, + "learning_rate": 1.9918323169967598e-05, + "loss": 3.297, + "step": 1734 + }, + { + "epoch": 0.1391032450742619, + "grad_norm": 2.2358362674713135, + "learning_rate": 1.9918157471205236e-05, + "loss": 3.298, + "step": 1735 + }, + { + "epoch": 0.1391834198552845, + "grad_norm": 2.448129892349243, + "learning_rate": 1.9917991605226237e-05, + "loss": 3.4016, + "step": 1736 + }, + { + "epoch": 0.13926359463630714, + "grad_norm": 2.29948091506958, + "learning_rate": 1.9917825572033394e-05, + "loss": 3.3353, + "step": 1737 + }, + { + "epoch": 0.1393437694173298, + "grad_norm": 2.938283681869507, + "learning_rate": 1.9917659371629504e-05, + "loss": 3.3451, + "step": 1738 + }, + { + "epoch": 0.1394239441983524, + "grad_norm": 2.5508086681365967, + "learning_rate": 1.991749300401738e-05, + "loss": 3.2849, + "step": 1739 + }, + { + "epoch": 0.13950411897937504, + "grad_norm": 2.57798433303833, + "learning_rate": 1.991732646919981e-05, + "loss": 3.4436, + "step": 1740 + }, + { + "epoch": 0.13958429376039766, + "grad_norm": 2.529953956604004, + "learning_rate": 1.991715976717962e-05, + "loss": 3.3906, + "step": 1741 + }, + { + "epoch": 0.13966446854142028, + "grad_norm": 2.2499284744262695, + "learning_rate": 1.9916992897959604e-05, + "loss": 3.2998, + "step": 1742 + }, + { + "epoch": 0.13974464332244294, + "grad_norm": 2.4092366695404053, + "learning_rate": 1.9916825861542585e-05, + "loss": 3.3555, + "step": 1743 + }, + { + "epoch": 0.13982481810346556, + "grad_norm": 2.3269553184509277, + "learning_rate": 1.9916658657931382e-05, + "loss": 3.2834, + "step": 1744 + }, + { + "epoch": 0.13990499288448818, + "grad_norm": 2.38739013671875, + "learning_rate": 1.9916491287128804e-05, + "loss": 3.3287, + "step": 1745 + }, + { + "epoch": 0.1399851676655108, + "grad_norm": 2.263183355331421, + "learning_rate": 1.9916323749137677e-05, + "loss": 3.3705, + "step": 1746 + }, + { + "epoch": 0.14006534244653343, + "grad_norm": 2.431453227996826, + "learning_rate": 1.9916156043960827e-05, + "loss": 3.293, + "step": 1747 + }, + { + "epoch": 0.14014551722755608, + "grad_norm": 2.715021848678589, + "learning_rate": 1.9915988171601083e-05, + "loss": 3.3068, + "step": 1748 + }, + { + "epoch": 0.1402256920085787, + "grad_norm": 1.1112844944000244, + "learning_rate": 1.991582013206127e-05, + "loss": 2.5909, + "step": 1749 + }, + { + "epoch": 0.14030586678960133, + "grad_norm": 1.1214396953582764, + "learning_rate": 1.991565192534423e-05, + "loss": 2.6002, + "step": 1750 + }, + { + "epoch": 0.14038604157062395, + "grad_norm": 2.2592666149139404, + "learning_rate": 1.9915483551452787e-05, + "loss": 3.3884, + "step": 1751 + }, + { + "epoch": 0.14046621635164658, + "grad_norm": 2.470088243484497, + "learning_rate": 1.9915315010389787e-05, + "loss": 3.392, + "step": 1752 + }, + { + "epoch": 0.14054639113266923, + "grad_norm": 2.3518691062927246, + "learning_rate": 1.9915146302158073e-05, + "loss": 3.27, + "step": 1753 + }, + { + "epoch": 0.14062656591369185, + "grad_norm": 3.364732265472412, + "learning_rate": 1.9914977426760486e-05, + "loss": 3.3547, + "step": 1754 + }, + { + "epoch": 0.14070674069471448, + "grad_norm": 2.370198965072632, + "learning_rate": 1.9914808384199874e-05, + "loss": 3.3423, + "step": 1755 + }, + { + "epoch": 0.1407869154757371, + "grad_norm": 2.409479856491089, + "learning_rate": 1.9914639174479084e-05, + "loss": 3.3077, + "step": 1756 + }, + { + "epoch": 0.14086709025675972, + "grad_norm": 2.3391494750976562, + "learning_rate": 1.9914469797600974e-05, + "loss": 3.3229, + "step": 1757 + }, + { + "epoch": 0.14094726503778238, + "grad_norm": 2.9073832035064697, + "learning_rate": 1.9914300253568403e-05, + "loss": 3.3702, + "step": 1758 + }, + { + "epoch": 0.141027439818805, + "grad_norm": 2.2789556980133057, + "learning_rate": 1.9914130542384217e-05, + "loss": 3.2119, + "step": 1759 + }, + { + "epoch": 0.14110761459982762, + "grad_norm": 2.466935634613037, + "learning_rate": 1.9913960664051283e-05, + "loss": 3.454, + "step": 1760 + }, + { + "epoch": 0.14118778938085025, + "grad_norm": 2.6597490310668945, + "learning_rate": 1.991379061857247e-05, + "loss": 3.3185, + "step": 1761 + }, + { + "epoch": 0.14126796416187287, + "grad_norm": 2.2989401817321777, + "learning_rate": 1.991362040595064e-05, + "loss": 3.4144, + "step": 1762 + }, + { + "epoch": 0.14134813894289552, + "grad_norm": 2.272409200668335, + "learning_rate": 1.9913450026188665e-05, + "loss": 3.4028, + "step": 1763 + }, + { + "epoch": 0.14142831372391815, + "grad_norm": 2.2841031551361084, + "learning_rate": 1.9913279479289413e-05, + "loss": 3.3018, + "step": 1764 + }, + { + "epoch": 0.14150848850494077, + "grad_norm": 2.0721194744110107, + "learning_rate": 1.9913108765255768e-05, + "loss": 3.3226, + "step": 1765 + }, + { + "epoch": 0.1415886632859634, + "grad_norm": 2.244127035140991, + "learning_rate": 1.9912937884090596e-05, + "loss": 3.2391, + "step": 1766 + }, + { + "epoch": 0.14166883806698602, + "grad_norm": 2.5054001808166504, + "learning_rate": 1.991276683579679e-05, + "loss": 3.4636, + "step": 1767 + }, + { + "epoch": 0.14174901284800867, + "grad_norm": 2.233550548553467, + "learning_rate": 1.991259562037723e-05, + "loss": 3.3154, + "step": 1768 + }, + { + "epoch": 0.1418291876290313, + "grad_norm": 2.254190444946289, + "learning_rate": 1.9912424237834798e-05, + "loss": 3.2728, + "step": 1769 + }, + { + "epoch": 0.14190936241005392, + "grad_norm": 2.1095964908599854, + "learning_rate": 1.991225268817239e-05, + "loss": 3.3948, + "step": 1770 + }, + { + "epoch": 0.14198953719107654, + "grad_norm": 2.1670045852661133, + "learning_rate": 1.9912080971392892e-05, + "loss": 3.3724, + "step": 1771 + }, + { + "epoch": 0.14206971197209917, + "grad_norm": 2.3975741863250732, + "learning_rate": 1.9911909087499204e-05, + "loss": 3.3004, + "step": 1772 + }, + { + "epoch": 0.14214988675312182, + "grad_norm": 2.2282259464263916, + "learning_rate": 1.9911737036494217e-05, + "loss": 3.2689, + "step": 1773 + }, + { + "epoch": 0.14223006153414444, + "grad_norm": 2.1414148807525635, + "learning_rate": 1.9911564818380844e-05, + "loss": 3.3121, + "step": 1774 + }, + { + "epoch": 0.14231023631516707, + "grad_norm": 2.1781363487243652, + "learning_rate": 1.991139243316198e-05, + "loss": 3.4059, + "step": 1775 + }, + { + "epoch": 0.1423904110961897, + "grad_norm": 2.200150489807129, + "learning_rate": 1.9911219880840528e-05, + "loss": 3.4197, + "step": 1776 + }, + { + "epoch": 0.1424705858772123, + "grad_norm": 2.3278613090515137, + "learning_rate": 1.9911047161419406e-05, + "loss": 3.3314, + "step": 1777 + }, + { + "epoch": 0.14255076065823497, + "grad_norm": 1.1608320474624634, + "learning_rate": 1.9910874274901522e-05, + "loss": 2.587, + "step": 1778 + }, + { + "epoch": 0.1426309354392576, + "grad_norm": 2.1420154571533203, + "learning_rate": 1.991070122128979e-05, + "loss": 3.2731, + "step": 1779 + }, + { + "epoch": 0.1427111102202802, + "grad_norm": 2.0371272563934326, + "learning_rate": 1.9910528000587124e-05, + "loss": 3.2418, + "step": 1780 + }, + { + "epoch": 0.14279128500130284, + "grad_norm": 2.1255383491516113, + "learning_rate": 1.9910354612796454e-05, + "loss": 3.1964, + "step": 1781 + }, + { + "epoch": 0.14287145978232546, + "grad_norm": 2.608011484146118, + "learning_rate": 1.99101810579207e-05, + "loss": 3.3922, + "step": 1782 + }, + { + "epoch": 0.1429516345633481, + "grad_norm": 2.1549007892608643, + "learning_rate": 1.9910007335962777e-05, + "loss": 3.282, + "step": 1783 + }, + { + "epoch": 0.14303180934437074, + "grad_norm": 2.26344895362854, + "learning_rate": 1.990983344692563e-05, + "loss": 3.2816, + "step": 1784 + }, + { + "epoch": 0.14311198412539336, + "grad_norm": 2.2591793537139893, + "learning_rate": 1.9909659390812184e-05, + "loss": 3.2847, + "step": 1785 + }, + { + "epoch": 0.14319215890641598, + "grad_norm": 2.255586624145508, + "learning_rate": 1.990948516762537e-05, + "loss": 3.458, + "step": 1786 + }, + { + "epoch": 0.1432723336874386, + "grad_norm": 3.051359176635742, + "learning_rate": 1.990931077736813e-05, + "loss": 3.3213, + "step": 1787 + }, + { + "epoch": 0.14335250846846126, + "grad_norm": 2.1385722160339355, + "learning_rate": 1.9909136220043402e-05, + "loss": 3.345, + "step": 1788 + }, + { + "epoch": 0.14343268324948388, + "grad_norm": 2.3507566452026367, + "learning_rate": 1.9908961495654128e-05, + "loss": 3.2988, + "step": 1789 + }, + { + "epoch": 0.1435128580305065, + "grad_norm": 2.075727701187134, + "learning_rate": 1.990878660420326e-05, + "loss": 3.3025, + "step": 1790 + }, + { + "epoch": 0.14359303281152913, + "grad_norm": 2.379040241241455, + "learning_rate": 1.990861154569374e-05, + "loss": 3.3636, + "step": 1791 + }, + { + "epoch": 0.14367320759255175, + "grad_norm": 2.3608481884002686, + "learning_rate": 1.990843632012852e-05, + "loss": 3.3248, + "step": 1792 + }, + { + "epoch": 0.14375338237357438, + "grad_norm": 2.0256338119506836, + "learning_rate": 1.9908260927510555e-05, + "loss": 3.2939, + "step": 1793 + }, + { + "epoch": 0.14383355715459703, + "grad_norm": 2.0842196941375732, + "learning_rate": 1.9908085367842806e-05, + "loss": 3.4094, + "step": 1794 + }, + { + "epoch": 0.14391373193561965, + "grad_norm": 2.279244899749756, + "learning_rate": 1.9907909641128227e-05, + "loss": 3.4147, + "step": 1795 + }, + { + "epoch": 0.14399390671664228, + "grad_norm": 2.090961456298828, + "learning_rate": 1.9907733747369786e-05, + "loss": 3.3634, + "step": 1796 + }, + { + "epoch": 0.1440740814976649, + "grad_norm": 2.456637144088745, + "learning_rate": 1.9907557686570444e-05, + "loss": 3.2965, + "step": 1797 + }, + { + "epoch": 0.14415425627868753, + "grad_norm": 2.404590129852295, + "learning_rate": 1.990738145873317e-05, + "loss": 3.3038, + "step": 1798 + }, + { + "epoch": 0.14423443105971018, + "grad_norm": 2.2045981884002686, + "learning_rate": 1.990720506386094e-05, + "loss": 3.3777, + "step": 1799 + }, + { + "epoch": 0.1443146058407328, + "grad_norm": 2.4620840549468994, + "learning_rate": 1.990702850195672e-05, + "loss": 3.3784, + "step": 1800 + }, + { + "epoch": 0.14439478062175543, + "grad_norm": 2.4131972789764404, + "learning_rate": 1.9906851773023494e-05, + "loss": 3.4113, + "step": 1801 + }, + { + "epoch": 0.14447495540277805, + "grad_norm": 2.1836938858032227, + "learning_rate": 1.990667487706424e-05, + "loss": 3.3755, + "step": 1802 + }, + { + "epoch": 0.14455513018380067, + "grad_norm": 2.2151541709899902, + "learning_rate": 1.9906497814081934e-05, + "loss": 3.3961, + "step": 1803 + }, + { + "epoch": 0.14463530496482332, + "grad_norm": 2.287973165512085, + "learning_rate": 1.990632058407957e-05, + "loss": 3.3463, + "step": 1804 + }, + { + "epoch": 0.14471547974584595, + "grad_norm": 2.314615249633789, + "learning_rate": 1.990614318706013e-05, + "loss": 3.3325, + "step": 1805 + }, + { + "epoch": 0.14479565452686857, + "grad_norm": 1.246783971786499, + "learning_rate": 1.990596562302661e-05, + "loss": 2.5688, + "step": 1806 + }, + { + "epoch": 0.1448758293078912, + "grad_norm": 2.070171594619751, + "learning_rate": 1.9905787891982e-05, + "loss": 3.3058, + "step": 1807 + }, + { + "epoch": 0.14495600408891382, + "grad_norm": 2.2133216857910156, + "learning_rate": 1.9905609993929297e-05, + "loss": 3.2904, + "step": 1808 + }, + { + "epoch": 0.14503617886993647, + "grad_norm": 2.0085511207580566, + "learning_rate": 1.9905431928871498e-05, + "loss": 3.3174, + "step": 1809 + }, + { + "epoch": 0.1451163536509591, + "grad_norm": 2.20220685005188, + "learning_rate": 1.990525369681161e-05, + "loss": 3.2466, + "step": 1810 + }, + { + "epoch": 0.14519652843198172, + "grad_norm": 2.1308419704437256, + "learning_rate": 1.9905075297752635e-05, + "loss": 3.2698, + "step": 1811 + }, + { + "epoch": 0.14527670321300434, + "grad_norm": 2.1188526153564453, + "learning_rate": 1.990489673169758e-05, + "loss": 3.3459, + "step": 1812 + }, + { + "epoch": 0.14535687799402697, + "grad_norm": 2.324122905731201, + "learning_rate": 1.9904717998649462e-05, + "loss": 3.4594, + "step": 1813 + }, + { + "epoch": 0.14543705277504962, + "grad_norm": 2.2799854278564453, + "learning_rate": 1.9904539098611283e-05, + "loss": 3.2155, + "step": 1814 + }, + { + "epoch": 0.14551722755607224, + "grad_norm": 2.2961649894714355, + "learning_rate": 1.990436003158607e-05, + "loss": 3.3033, + "step": 1815 + }, + { + "epoch": 0.14559740233709487, + "grad_norm": 2.2109014987945557, + "learning_rate": 1.990418079757684e-05, + "loss": 3.2492, + "step": 1816 + }, + { + "epoch": 0.1456775771181175, + "grad_norm": 2.2658886909484863, + "learning_rate": 1.9904001396586607e-05, + "loss": 3.3402, + "step": 1817 + }, + { + "epoch": 0.14575775189914011, + "grad_norm": 2.327906608581543, + "learning_rate": 1.99038218286184e-05, + "loss": 3.3368, + "step": 1818 + }, + { + "epoch": 0.14583792668016277, + "grad_norm": 2.25846266746521, + "learning_rate": 1.990364209367525e-05, + "loss": 3.36, + "step": 1819 + }, + { + "epoch": 0.1459181014611854, + "grad_norm": 2.447394609451294, + "learning_rate": 1.9903462191760185e-05, + "loss": 3.3187, + "step": 1820 + }, + { + "epoch": 0.145998276242208, + "grad_norm": 2.329237461090088, + "learning_rate": 1.9903282122876234e-05, + "loss": 3.3702, + "step": 1821 + }, + { + "epoch": 0.14607845102323064, + "grad_norm": 2.498325824737549, + "learning_rate": 1.990310188702644e-05, + "loss": 3.4268, + "step": 1822 + }, + { + "epoch": 0.14615862580425326, + "grad_norm": 2.8000080585479736, + "learning_rate": 1.990292148421384e-05, + "loss": 3.4438, + "step": 1823 + }, + { + "epoch": 0.1462388005852759, + "grad_norm": 2.311633825302124, + "learning_rate": 1.990274091444147e-05, + "loss": 3.3108, + "step": 1824 + }, + { + "epoch": 0.14631897536629854, + "grad_norm": 2.0674023628234863, + "learning_rate": 1.9902560177712383e-05, + "loss": 3.234, + "step": 1825 + }, + { + "epoch": 0.14639915014732116, + "grad_norm": 2.4241130352020264, + "learning_rate": 1.9902379274029616e-05, + "loss": 3.2284, + "step": 1826 + }, + { + "epoch": 0.14647932492834378, + "grad_norm": 2.367967128753662, + "learning_rate": 1.9902198203396228e-05, + "loss": 3.3597, + "step": 1827 + }, + { + "epoch": 0.1465594997093664, + "grad_norm": 2.5503923892974854, + "learning_rate": 1.9902016965815268e-05, + "loss": 3.2638, + "step": 1828 + }, + { + "epoch": 0.14663967449038906, + "grad_norm": 2.220764398574829, + "learning_rate": 1.990183556128979e-05, + "loss": 3.2902, + "step": 1829 + }, + { + "epoch": 0.14671984927141168, + "grad_norm": 2.3185062408447266, + "learning_rate": 1.9901653989822855e-05, + "loss": 3.3159, + "step": 1830 + }, + { + "epoch": 0.1468000240524343, + "grad_norm": 2.25873064994812, + "learning_rate": 1.9901472251417525e-05, + "loss": 3.3453, + "step": 1831 + }, + { + "epoch": 0.14688019883345693, + "grad_norm": 2.3766491413116455, + "learning_rate": 1.990129034607686e-05, + "loss": 3.2548, + "step": 1832 + }, + { + "epoch": 0.14696037361447956, + "grad_norm": 2.41988205909729, + "learning_rate": 1.9901108273803932e-05, + "loss": 3.3091, + "step": 1833 + }, + { + "epoch": 0.1470405483955022, + "grad_norm": 2.2186930179595947, + "learning_rate": 1.9900926034601803e-05, + "loss": 3.3079, + "step": 1834 + }, + { + "epoch": 0.14712072317652483, + "grad_norm": 2.264514446258545, + "learning_rate": 1.990074362847356e-05, + "loss": 3.4154, + "step": 1835 + }, + { + "epoch": 0.14720089795754746, + "grad_norm": 2.301100015640259, + "learning_rate": 1.9900561055422257e-05, + "loss": 3.4161, + "step": 1836 + }, + { + "epoch": 0.14728107273857008, + "grad_norm": 2.620426893234253, + "learning_rate": 1.9900378315450994e-05, + "loss": 3.3768, + "step": 1837 + }, + { + "epoch": 0.1473612475195927, + "grad_norm": 2.404336929321289, + "learning_rate": 1.9900195408562835e-05, + "loss": 3.4075, + "step": 1838 + }, + { + "epoch": 0.14744142230061535, + "grad_norm": 2.0791378021240234, + "learning_rate": 1.9900012334760874e-05, + "loss": 3.2382, + "step": 1839 + }, + { + "epoch": 0.14752159708163798, + "grad_norm": 2.257167339324951, + "learning_rate": 1.989982909404819e-05, + "loss": 3.319, + "step": 1840 + }, + { + "epoch": 0.1476017718626606, + "grad_norm": 2.5617921352386475, + "learning_rate": 1.9899645686427882e-05, + "loss": 3.4696, + "step": 1841 + }, + { + "epoch": 0.14768194664368323, + "grad_norm": 2.306251049041748, + "learning_rate": 1.9899462111903035e-05, + "loss": 3.3046, + "step": 1842 + }, + { + "epoch": 0.14776212142470585, + "grad_norm": 2.37430739402771, + "learning_rate": 1.9899278370476742e-05, + "loss": 3.339, + "step": 1843 + }, + { + "epoch": 0.1478422962057285, + "grad_norm": 2.6636605262756348, + "learning_rate": 1.989909446215211e-05, + "loss": 3.344, + "step": 1844 + }, + { + "epoch": 0.14792247098675113, + "grad_norm": 2.0909252166748047, + "learning_rate": 1.989891038693223e-05, + "loss": 3.2419, + "step": 1845 + }, + { + "epoch": 0.14800264576777375, + "grad_norm": 2.338792562484741, + "learning_rate": 1.9898726144820212e-05, + "loss": 3.3376, + "step": 1846 + }, + { + "epoch": 0.14808282054879637, + "grad_norm": 2.2952821254730225, + "learning_rate": 1.989854173581916e-05, + "loss": 3.3151, + "step": 1847 + }, + { + "epoch": 0.148162995329819, + "grad_norm": 2.115614652633667, + "learning_rate": 1.989835715993218e-05, + "loss": 3.2592, + "step": 1848 + }, + { + "epoch": 0.14824317011084162, + "grad_norm": 2.292872667312622, + "learning_rate": 1.989817241716239e-05, + "loss": 3.424, + "step": 1849 + }, + { + "epoch": 0.14832334489186427, + "grad_norm": 1.144291639328003, + "learning_rate": 1.9897987507512902e-05, + "loss": 2.5898, + "step": 1850 + }, + { + "epoch": 0.1484035196728869, + "grad_norm": 2.2746002674102783, + "learning_rate": 1.989780243098683e-05, + "loss": 3.3092, + "step": 1851 + }, + { + "epoch": 0.14848369445390952, + "grad_norm": 2.4605021476745605, + "learning_rate": 1.98976171875873e-05, + "loss": 3.3916, + "step": 1852 + }, + { + "epoch": 0.14856386923493214, + "grad_norm": 1.1905325651168823, + "learning_rate": 1.989743177731743e-05, + "loss": 2.5997, + "step": 1853 + }, + { + "epoch": 0.14864404401595477, + "grad_norm": 2.2048327922821045, + "learning_rate": 1.989724620018035e-05, + "loss": 3.274, + "step": 1854 + }, + { + "epoch": 0.14872421879697742, + "grad_norm": 2.2737650871276855, + "learning_rate": 1.9897060456179186e-05, + "loss": 3.2178, + "step": 1855 + }, + { + "epoch": 0.14880439357800004, + "grad_norm": 2.224547863006592, + "learning_rate": 1.9896874545317072e-05, + "loss": 3.2844, + "step": 1856 + }, + { + "epoch": 0.14888456835902267, + "grad_norm": 2.3130970001220703, + "learning_rate": 1.989668846759714e-05, + "loss": 3.4082, + "step": 1857 + }, + { + "epoch": 0.1489647431400453, + "grad_norm": 2.191310167312622, + "learning_rate": 1.989650222302253e-05, + "loss": 3.3218, + "step": 1858 + }, + { + "epoch": 0.14904491792106792, + "grad_norm": 2.04880690574646, + "learning_rate": 1.9896315811596378e-05, + "loss": 3.1951, + "step": 1859 + }, + { + "epoch": 0.14912509270209057, + "grad_norm": 2.207897663116455, + "learning_rate": 1.989612923332183e-05, + "loss": 3.3512, + "step": 1860 + }, + { + "epoch": 0.1492052674831132, + "grad_norm": 2.281933307647705, + "learning_rate": 1.989594248820203e-05, + "loss": 3.3356, + "step": 1861 + }, + { + "epoch": 0.14928544226413581, + "grad_norm": 2.2580156326293945, + "learning_rate": 1.9895755576240133e-05, + "loss": 3.3009, + "step": 1862 + }, + { + "epoch": 0.14936561704515844, + "grad_norm": 2.1454219818115234, + "learning_rate": 1.989556849743928e-05, + "loss": 3.3492, + "step": 1863 + }, + { + "epoch": 0.14944579182618106, + "grad_norm": 2.031308174133301, + "learning_rate": 1.9895381251802628e-05, + "loss": 3.3636, + "step": 1864 + }, + { + "epoch": 0.14952596660720371, + "grad_norm": 2.3000576496124268, + "learning_rate": 1.9895193839333336e-05, + "loss": 3.3224, + "step": 1865 + }, + { + "epoch": 0.14960614138822634, + "grad_norm": 2.310314655303955, + "learning_rate": 1.989500626003457e-05, + "loss": 3.3453, + "step": 1866 + }, + { + "epoch": 0.14968631616924896, + "grad_norm": 2.1267218589782715, + "learning_rate": 1.989481851390948e-05, + "loss": 3.2758, + "step": 1867 + }, + { + "epoch": 0.14976649095027159, + "grad_norm": 2.2483103275299072, + "learning_rate": 1.9894630600961236e-05, + "loss": 3.3772, + "step": 1868 + }, + { + "epoch": 0.1498466657312942, + "grad_norm": 2.140075206756592, + "learning_rate": 1.989444252119301e-05, + "loss": 3.2827, + "step": 1869 + }, + { + "epoch": 0.14992684051231686, + "grad_norm": 2.4762191772460938, + "learning_rate": 1.989425427460797e-05, + "loss": 3.5087, + "step": 1870 + }, + { + "epoch": 0.15000701529333949, + "grad_norm": 2.1972577571868896, + "learning_rate": 1.9894065861209288e-05, + "loss": 3.2621, + "step": 1871 + }, + { + "epoch": 0.1500871900743621, + "grad_norm": 2.182378053665161, + "learning_rate": 1.9893877281000144e-05, + "loss": 3.3533, + "step": 1872 + }, + { + "epoch": 0.15016736485538473, + "grad_norm": 1.1103260517120361, + "learning_rate": 1.9893688533983715e-05, + "loss": 2.5909, + "step": 1873 + }, + { + "epoch": 0.15024753963640736, + "grad_norm": 2.3276009559631348, + "learning_rate": 1.9893499620163184e-05, + "loss": 3.3696, + "step": 1874 + }, + { + "epoch": 0.15032771441743, + "grad_norm": 1.1112494468688965, + "learning_rate": 1.989331053954174e-05, + "loss": 2.5945, + "step": 1875 + }, + { + "epoch": 0.15040788919845263, + "grad_norm": 1.0362688302993774, + "learning_rate": 1.9893121292122566e-05, + "loss": 2.5583, + "step": 1876 + }, + { + "epoch": 0.15048806397947526, + "grad_norm": 2.536043405532837, + "learning_rate": 1.9892931877908852e-05, + "loss": 3.3165, + "step": 1877 + }, + { + "epoch": 0.15056823876049788, + "grad_norm": 2.1679913997650146, + "learning_rate": 1.9892742296903793e-05, + "loss": 3.2594, + "step": 1878 + }, + { + "epoch": 0.1506484135415205, + "grad_norm": 1.1143604516983032, + "learning_rate": 1.9892552549110584e-05, + "loss": 2.5901, + "step": 1879 + }, + { + "epoch": 0.15072858832254316, + "grad_norm": 2.3217058181762695, + "learning_rate": 1.9892362634532427e-05, + "loss": 3.3061, + "step": 1880 + }, + { + "epoch": 0.15080876310356578, + "grad_norm": 2.064689874649048, + "learning_rate": 1.9892172553172524e-05, + "loss": 3.25, + "step": 1881 + }, + { + "epoch": 0.1508889378845884, + "grad_norm": 2.293738603591919, + "learning_rate": 1.9891982305034077e-05, + "loss": 3.3203, + "step": 1882 + }, + { + "epoch": 0.15096911266561103, + "grad_norm": 2.2416775226593018, + "learning_rate": 1.9891791890120294e-05, + "loss": 3.3412, + "step": 1883 + }, + { + "epoch": 0.15104928744663365, + "grad_norm": 2.271165132522583, + "learning_rate": 1.9891601308434383e-05, + "loss": 3.3286, + "step": 1884 + }, + { + "epoch": 0.1511294622276563, + "grad_norm": 2.1550819873809814, + "learning_rate": 1.9891410559979565e-05, + "loss": 3.4154, + "step": 1885 + }, + { + "epoch": 0.15120963700867893, + "grad_norm": 1.0611159801483154, + "learning_rate": 1.989121964475905e-05, + "loss": 2.564, + "step": 1886 + }, + { + "epoch": 0.15128981178970155, + "grad_norm": 2.4183385372161865, + "learning_rate": 1.989102856277606e-05, + "loss": 3.391, + "step": 1887 + }, + { + "epoch": 0.15136998657072417, + "grad_norm": 2.270463228225708, + "learning_rate": 1.989083731403381e-05, + "loss": 3.3414, + "step": 1888 + }, + { + "epoch": 0.1514501613517468, + "grad_norm": 2.228119134902954, + "learning_rate": 1.9890645898535527e-05, + "loss": 3.3186, + "step": 1889 + }, + { + "epoch": 0.15153033613276945, + "grad_norm": 2.363853693008423, + "learning_rate": 1.9890454316284443e-05, + "loss": 3.2915, + "step": 1890 + }, + { + "epoch": 0.15161051091379207, + "grad_norm": 2.341951847076416, + "learning_rate": 1.9890262567283787e-05, + "loss": 3.4464, + "step": 1891 + }, + { + "epoch": 0.1516906856948147, + "grad_norm": 1.146844744682312, + "learning_rate": 1.9890070651536787e-05, + "loss": 2.577, + "step": 1892 + }, + { + "epoch": 0.15177086047583732, + "grad_norm": 1.1215319633483887, + "learning_rate": 1.9889878569046682e-05, + "loss": 2.566, + "step": 1893 + }, + { + "epoch": 0.15185103525685995, + "grad_norm": 2.0947561264038086, + "learning_rate": 1.988968631981671e-05, + "loss": 3.3875, + "step": 1894 + }, + { + "epoch": 0.1519312100378826, + "grad_norm": 2.3140602111816406, + "learning_rate": 1.9889493903850116e-05, + "loss": 3.4216, + "step": 1895 + }, + { + "epoch": 0.15201138481890522, + "grad_norm": 2.259992837905884, + "learning_rate": 1.9889301321150135e-05, + "loss": 3.2839, + "step": 1896 + }, + { + "epoch": 0.15209155959992784, + "grad_norm": 2.4727284908294678, + "learning_rate": 1.988910857172002e-05, + "loss": 3.207, + "step": 1897 + }, + { + "epoch": 0.15217173438095047, + "grad_norm": 2.2338788509368896, + "learning_rate": 1.988891565556302e-05, + "loss": 3.3209, + "step": 1898 + }, + { + "epoch": 0.1522519091619731, + "grad_norm": 2.502636194229126, + "learning_rate": 1.9888722572682386e-05, + "loss": 3.2841, + "step": 1899 + }, + { + "epoch": 0.15233208394299574, + "grad_norm": 2.2307655811309814, + "learning_rate": 1.9888529323081375e-05, + "loss": 3.3409, + "step": 1900 + }, + { + "epoch": 0.15241225872401837, + "grad_norm": 2.12278151512146, + "learning_rate": 1.9888335906763247e-05, + "loss": 3.3422, + "step": 1901 + }, + { + "epoch": 0.152492433505041, + "grad_norm": 2.271554946899414, + "learning_rate": 1.9888142323731258e-05, + "loss": 3.3442, + "step": 1902 + }, + { + "epoch": 0.15257260828606362, + "grad_norm": 2.4834561347961426, + "learning_rate": 1.988794857398868e-05, + "loss": 3.3298, + "step": 1903 + }, + { + "epoch": 0.15265278306708624, + "grad_norm": 2.036832809448242, + "learning_rate": 1.988775465753877e-05, + "loss": 3.2211, + "step": 1904 + }, + { + "epoch": 0.1527329578481089, + "grad_norm": 2.1008219718933105, + "learning_rate": 1.9887560574384798e-05, + "loss": 3.3593, + "step": 1905 + }, + { + "epoch": 0.15281313262913152, + "grad_norm": 2.112394094467163, + "learning_rate": 1.988736632453004e-05, + "loss": 3.2726, + "step": 1906 + }, + { + "epoch": 0.15289330741015414, + "grad_norm": 2.4101402759552, + "learning_rate": 1.9887171907977776e-05, + "loss": 3.3171, + "step": 1907 + }, + { + "epoch": 0.15297348219117676, + "grad_norm": 2.1193697452545166, + "learning_rate": 1.9886977324731275e-05, + "loss": 3.2538, + "step": 1908 + }, + { + "epoch": 0.1530536569721994, + "grad_norm": 2.1130974292755127, + "learning_rate": 1.988678257479382e-05, + "loss": 3.2717, + "step": 1909 + }, + { + "epoch": 0.153133831753222, + "grad_norm": 2.371448516845703, + "learning_rate": 1.9886587658168698e-05, + "loss": 3.3114, + "step": 1910 + }, + { + "epoch": 0.15321400653424466, + "grad_norm": 2.239240884780884, + "learning_rate": 1.988639257485919e-05, + "loss": 3.3365, + "step": 1911 + }, + { + "epoch": 0.15329418131526729, + "grad_norm": 2.0339581966400146, + "learning_rate": 1.9886197324868585e-05, + "loss": 3.2518, + "step": 1912 + }, + { + "epoch": 0.1533743560962899, + "grad_norm": 2.2798500061035156, + "learning_rate": 1.988600190820018e-05, + "loss": 3.285, + "step": 1913 + }, + { + "epoch": 0.15345453087731253, + "grad_norm": 2.090857744216919, + "learning_rate": 1.988580632485727e-05, + "loss": 3.3542, + "step": 1914 + }, + { + "epoch": 0.15353470565833516, + "grad_norm": 2.275472640991211, + "learning_rate": 1.9885610574843144e-05, + "loss": 3.3528, + "step": 1915 + }, + { + "epoch": 0.1536148804393578, + "grad_norm": 2.387042760848999, + "learning_rate": 1.988541465816111e-05, + "loss": 3.4398, + "step": 1916 + }, + { + "epoch": 0.15369505522038043, + "grad_norm": 2.4270613193511963, + "learning_rate": 1.9885218574814468e-05, + "loss": 3.2893, + "step": 1917 + }, + { + "epoch": 0.15377523000140306, + "grad_norm": 2.180915594100952, + "learning_rate": 1.9885022324806527e-05, + "loss": 3.3713, + "step": 1918 + }, + { + "epoch": 0.15385540478242568, + "grad_norm": 2.178072929382324, + "learning_rate": 1.9884825908140596e-05, + "loss": 3.3574, + "step": 1919 + }, + { + "epoch": 0.1539355795634483, + "grad_norm": 2.5830881595611572, + "learning_rate": 1.9884629324819976e-05, + "loss": 3.4056, + "step": 1920 + }, + { + "epoch": 0.15401575434447096, + "grad_norm": 2.083772897720337, + "learning_rate": 1.9884432574847996e-05, + "loss": 3.2544, + "step": 1921 + }, + { + "epoch": 0.15409592912549358, + "grad_norm": 2.4149250984191895, + "learning_rate": 1.9884235658227963e-05, + "loss": 3.2816, + "step": 1922 + }, + { + "epoch": 0.1541761039065162, + "grad_norm": 2.2062301635742188, + "learning_rate": 1.9884038574963202e-05, + "loss": 3.4387, + "step": 1923 + }, + { + "epoch": 0.15425627868753883, + "grad_norm": 2.1696386337280273, + "learning_rate": 1.9883841325057035e-05, + "loss": 3.1684, + "step": 1924 + }, + { + "epoch": 0.15433645346856145, + "grad_norm": 2.3587472438812256, + "learning_rate": 1.9883643908512785e-05, + "loss": 3.4557, + "step": 1925 + }, + { + "epoch": 0.1544166282495841, + "grad_norm": 2.0614070892333984, + "learning_rate": 1.988344632533378e-05, + "loss": 3.2329, + "step": 1926 + }, + { + "epoch": 0.15449680303060673, + "grad_norm": 1.1535799503326416, + "learning_rate": 1.9883248575523354e-05, + "loss": 2.572, + "step": 1927 + }, + { + "epoch": 0.15457697781162935, + "grad_norm": 2.3671274185180664, + "learning_rate": 1.9883050659084843e-05, + "loss": 3.3442, + "step": 1928 + }, + { + "epoch": 0.15465715259265198, + "grad_norm": 2.3223979473114014, + "learning_rate": 1.988285257602158e-05, + "loss": 3.342, + "step": 1929 + }, + { + "epoch": 0.1547373273736746, + "grad_norm": 2.4325671195983887, + "learning_rate": 1.9882654326336905e-05, + "loss": 3.4349, + "step": 1930 + }, + { + "epoch": 0.15481750215469725, + "grad_norm": 1.1848920583724976, + "learning_rate": 1.9882455910034162e-05, + "loss": 2.5739, + "step": 1931 + }, + { + "epoch": 0.15489767693571987, + "grad_norm": 2.26961612701416, + "learning_rate": 1.988225732711669e-05, + "loss": 3.3417, + "step": 1932 + }, + { + "epoch": 0.1549778517167425, + "grad_norm": 2.119812488555908, + "learning_rate": 1.988205857758785e-05, + "loss": 3.2904, + "step": 1933 + }, + { + "epoch": 0.15505802649776512, + "grad_norm": 2.1822309494018555, + "learning_rate": 1.988185966145098e-05, + "loss": 3.4038, + "step": 1934 + }, + { + "epoch": 0.15513820127878775, + "grad_norm": 2.088184118270874, + "learning_rate": 1.9881660578709444e-05, + "loss": 3.2762, + "step": 1935 + }, + { + "epoch": 0.1552183760598104, + "grad_norm": 2.2895774841308594, + "learning_rate": 1.9881461329366586e-05, + "loss": 3.2393, + "step": 1936 + }, + { + "epoch": 0.15529855084083302, + "grad_norm": 2.138101816177368, + "learning_rate": 1.9881261913425777e-05, + "loss": 3.2255, + "step": 1937 + }, + { + "epoch": 0.15537872562185565, + "grad_norm": 2.3741936683654785, + "learning_rate": 1.9881062330890376e-05, + "loss": 3.4791, + "step": 1938 + }, + { + "epoch": 0.15545890040287827, + "grad_norm": 2.007265567779541, + "learning_rate": 1.9880862581763744e-05, + "loss": 3.2682, + "step": 1939 + }, + { + "epoch": 0.1555390751839009, + "grad_norm": 2.1488943099975586, + "learning_rate": 1.9880662666049254e-05, + "loss": 3.257, + "step": 1940 + }, + { + "epoch": 0.15561924996492355, + "grad_norm": 2.3393683433532715, + "learning_rate": 1.9880462583750272e-05, + "loss": 3.3663, + "step": 1941 + }, + { + "epoch": 0.15569942474594617, + "grad_norm": 2.1485841274261475, + "learning_rate": 1.9880262334870172e-05, + "loss": 3.3307, + "step": 1942 + }, + { + "epoch": 0.1557795995269688, + "grad_norm": 2.127168893814087, + "learning_rate": 1.9880061919412332e-05, + "loss": 3.4489, + "step": 1943 + }, + { + "epoch": 0.15585977430799142, + "grad_norm": 2.1947720050811768, + "learning_rate": 1.9879861337380127e-05, + "loss": 3.2874, + "step": 1944 + }, + { + "epoch": 0.15593994908901404, + "grad_norm": 2.158921957015991, + "learning_rate": 1.9879660588776945e-05, + "loss": 3.2053, + "step": 1945 + }, + { + "epoch": 0.1560201238700367, + "grad_norm": 2.3567709922790527, + "learning_rate": 1.987945967360617e-05, + "loss": 3.2426, + "step": 1946 + }, + { + "epoch": 0.15610029865105932, + "grad_norm": 2.1029419898986816, + "learning_rate": 1.9879258591871183e-05, + "loss": 3.4194, + "step": 1947 + }, + { + "epoch": 0.15618047343208194, + "grad_norm": 2.236489772796631, + "learning_rate": 1.9879057343575378e-05, + "loss": 3.3093, + "step": 1948 + }, + { + "epoch": 0.15626064821310456, + "grad_norm": 2.193817615509033, + "learning_rate": 1.987885592872215e-05, + "loss": 3.2614, + "step": 1949 + }, + { + "epoch": 0.1563408229941272, + "grad_norm": 2.258244752883911, + "learning_rate": 1.9878654347314893e-05, + "loss": 3.317, + "step": 1950 + }, + { + "epoch": 0.15642099777514984, + "grad_norm": 2.6576216220855713, + "learning_rate": 1.9878452599357e-05, + "loss": 3.377, + "step": 1951 + }, + { + "epoch": 0.15650117255617246, + "grad_norm": 2.387352705001831, + "learning_rate": 1.9878250684851883e-05, + "loss": 3.3485, + "step": 1952 + }, + { + "epoch": 0.1565813473371951, + "grad_norm": 2.2729077339172363, + "learning_rate": 1.987804860380294e-05, + "loss": 3.3623, + "step": 1953 + }, + { + "epoch": 0.1566615221182177, + "grad_norm": 2.4699060916900635, + "learning_rate": 1.987784635621358e-05, + "loss": 3.311, + "step": 1954 + }, + { + "epoch": 0.15674169689924033, + "grad_norm": 2.376830816268921, + "learning_rate": 1.9877643942087207e-05, + "loss": 3.3809, + "step": 1955 + }, + { + "epoch": 0.156821871680263, + "grad_norm": 2.3643343448638916, + "learning_rate": 1.9877441361427243e-05, + "loss": 3.2986, + "step": 1956 + }, + { + "epoch": 0.1569020464612856, + "grad_norm": 1.388809084892273, + "learning_rate": 1.9877238614237096e-05, + "loss": 2.5795, + "step": 1957 + }, + { + "epoch": 0.15698222124230823, + "grad_norm": 2.220301866531372, + "learning_rate": 1.987703570052019e-05, + "loss": 3.3064, + "step": 1958 + }, + { + "epoch": 0.15706239602333086, + "grad_norm": 2.413954973220825, + "learning_rate": 1.9876832620279944e-05, + "loss": 3.2955, + "step": 1959 + }, + { + "epoch": 0.15714257080435348, + "grad_norm": 2.084969997406006, + "learning_rate": 1.9876629373519773e-05, + "loss": 3.2787, + "step": 1960 + }, + { + "epoch": 0.15722274558537613, + "grad_norm": 2.0168068408966064, + "learning_rate": 1.987642596024312e-05, + "loss": 3.295, + "step": 1961 + }, + { + "epoch": 0.15730292036639876, + "grad_norm": 2.2773163318634033, + "learning_rate": 1.98762223804534e-05, + "loss": 3.3556, + "step": 1962 + }, + { + "epoch": 0.15738309514742138, + "grad_norm": 2.408128499984741, + "learning_rate": 1.987601863415406e-05, + "loss": 3.2795, + "step": 1963 + }, + { + "epoch": 0.157463269928444, + "grad_norm": 2.091891050338745, + "learning_rate": 1.987581472134852e-05, + "loss": 3.4038, + "step": 1964 + }, + { + "epoch": 0.15754344470946663, + "grad_norm": 2.41982364654541, + "learning_rate": 1.987561064204022e-05, + "loss": 3.3298, + "step": 1965 + }, + { + "epoch": 0.15762361949048925, + "grad_norm": 2.195590019226074, + "learning_rate": 1.9875406396232613e-05, + "loss": 3.3672, + "step": 1966 + }, + { + "epoch": 0.1577037942715119, + "grad_norm": 2.036198854446411, + "learning_rate": 1.987520198392913e-05, + "loss": 3.2836, + "step": 1967 + }, + { + "epoch": 0.15778396905253453, + "grad_norm": 2.204434871673584, + "learning_rate": 1.9874997405133227e-05, + "loss": 3.403, + "step": 1968 + }, + { + "epoch": 0.15786414383355715, + "grad_norm": 2.0949745178222656, + "learning_rate": 1.9874792659848342e-05, + "loss": 3.4199, + "step": 1969 + }, + { + "epoch": 0.15794431861457978, + "grad_norm": 1.1438573598861694, + "learning_rate": 1.987458774807794e-05, + "loss": 2.5558, + "step": 1970 + }, + { + "epoch": 0.1580244933956024, + "grad_norm": 2.0612356662750244, + "learning_rate": 1.987438266982546e-05, + "loss": 3.2917, + "step": 1971 + }, + { + "epoch": 0.15810466817662505, + "grad_norm": 2.2056283950805664, + "learning_rate": 1.9874177425094375e-05, + "loss": 3.3624, + "step": 1972 + }, + { + "epoch": 0.15818484295764768, + "grad_norm": 2.289311408996582, + "learning_rate": 1.9873972013888135e-05, + "loss": 3.3674, + "step": 1973 + }, + { + "epoch": 0.1582650177386703, + "grad_norm": 2.354127883911133, + "learning_rate": 1.9873766436210207e-05, + "loss": 3.3171, + "step": 1974 + }, + { + "epoch": 0.15834519251969292, + "grad_norm": 2.364713430404663, + "learning_rate": 1.987356069206406e-05, + "loss": 3.3785, + "step": 1975 + }, + { + "epoch": 0.15842536730071555, + "grad_norm": 2.096499443054199, + "learning_rate": 1.987335478145316e-05, + "loss": 3.3387, + "step": 1976 + }, + { + "epoch": 0.1585055420817382, + "grad_norm": 2.096958875656128, + "learning_rate": 1.9873148704380973e-05, + "loss": 3.3235, + "step": 1977 + }, + { + "epoch": 0.15858571686276082, + "grad_norm": 2.213000774383545, + "learning_rate": 1.987294246085098e-05, + "loss": 3.3488, + "step": 1978 + }, + { + "epoch": 0.15866589164378345, + "grad_norm": 2.2483460903167725, + "learning_rate": 1.9872736050866658e-05, + "loss": 3.3899, + "step": 1979 + }, + { + "epoch": 0.15874606642480607, + "grad_norm": 2.217906951904297, + "learning_rate": 1.9872529474431484e-05, + "loss": 3.3693, + "step": 1980 + }, + { + "epoch": 0.1588262412058287, + "grad_norm": 2.1277811527252197, + "learning_rate": 1.9872322731548946e-05, + "loss": 3.261, + "step": 1981 + }, + { + "epoch": 0.15890641598685135, + "grad_norm": 2.70749568939209, + "learning_rate": 1.9872115822222522e-05, + "loss": 3.4003, + "step": 1982 + }, + { + "epoch": 0.15898659076787397, + "grad_norm": 2.2755682468414307, + "learning_rate": 1.9871908746455707e-05, + "loss": 3.3788, + "step": 1983 + }, + { + "epoch": 0.1590667655488966, + "grad_norm": 2.1479673385620117, + "learning_rate": 1.987170150425199e-05, + "loss": 3.2308, + "step": 1984 + }, + { + "epoch": 0.15914694032991922, + "grad_norm": 2.107785224914551, + "learning_rate": 1.9871494095614863e-05, + "loss": 3.3665, + "step": 1985 + }, + { + "epoch": 0.15922711511094184, + "grad_norm": 2.297151803970337, + "learning_rate": 1.9871286520547827e-05, + "loss": 3.3699, + "step": 1986 + }, + { + "epoch": 0.1593072898919645, + "grad_norm": 2.6228604316711426, + "learning_rate": 1.987107877905438e-05, + "loss": 3.1561, + "step": 1987 + }, + { + "epoch": 0.15938746467298712, + "grad_norm": 2.4018824100494385, + "learning_rate": 1.987087087113802e-05, + "loss": 3.26, + "step": 1988 + }, + { + "epoch": 0.15946763945400974, + "grad_norm": 2.043574810028076, + "learning_rate": 1.9870662796802256e-05, + "loss": 3.3019, + "step": 1989 + }, + { + "epoch": 0.15954781423503236, + "grad_norm": 2.178148031234741, + "learning_rate": 1.9870454556050597e-05, + "loss": 3.3735, + "step": 1990 + }, + { + "epoch": 0.159627989016055, + "grad_norm": 2.3124167919158936, + "learning_rate": 1.9870246148886553e-05, + "loss": 3.2468, + "step": 1991 + }, + { + "epoch": 0.15970816379707764, + "grad_norm": 2.3249435424804688, + "learning_rate": 1.987003757531364e-05, + "loss": 3.3099, + "step": 1992 + }, + { + "epoch": 0.15978833857810026, + "grad_norm": 2.1672139167785645, + "learning_rate": 1.986982883533537e-05, + "loss": 3.2704, + "step": 1993 + }, + { + "epoch": 0.1598685133591229, + "grad_norm": 2.1517319679260254, + "learning_rate": 1.9869619928955262e-05, + "loss": 3.3075, + "step": 1994 + }, + { + "epoch": 0.1599486881401455, + "grad_norm": 2.5968453884124756, + "learning_rate": 1.986941085617684e-05, + "loss": 3.4227, + "step": 1995 + }, + { + "epoch": 0.16002886292116814, + "grad_norm": 2.134185791015625, + "learning_rate": 1.9869201617003636e-05, + "loss": 3.3561, + "step": 1996 + }, + { + "epoch": 0.1601090377021908, + "grad_norm": 2.2576050758361816, + "learning_rate": 1.9868992211439163e-05, + "loss": 3.3223, + "step": 1997 + }, + { + "epoch": 0.1601892124832134, + "grad_norm": 2.1410717964172363, + "learning_rate": 1.9868782639486964e-05, + "loss": 3.119, + "step": 1998 + }, + { + "epoch": 0.16026938726423604, + "grad_norm": 2.0553505420684814, + "learning_rate": 1.9868572901150568e-05, + "loss": 3.2063, + "step": 1999 + }, + { + "epoch": 0.16034956204525866, + "grad_norm": 2.2687201499938965, + "learning_rate": 1.986836299643351e-05, + "loss": 3.349, + "step": 2000 + }, + { + "epoch": 0.16042973682628128, + "grad_norm": 2.112774610519409, + "learning_rate": 1.986815292533933e-05, + "loss": 3.2693, + "step": 2001 + }, + { + "epoch": 0.16050991160730393, + "grad_norm": 2.4449703693389893, + "learning_rate": 1.986794268787157e-05, + "loss": 3.3708, + "step": 2002 + }, + { + "epoch": 0.16059008638832656, + "grad_norm": 1.2707598209381104, + "learning_rate": 1.9867732284033773e-05, + "loss": 2.5734, + "step": 2003 + }, + { + "epoch": 0.16067026116934918, + "grad_norm": 2.4103097915649414, + "learning_rate": 1.986752171382949e-05, + "loss": 3.2511, + "step": 2004 + }, + { + "epoch": 0.1607504359503718, + "grad_norm": 2.2268474102020264, + "learning_rate": 1.9867310977262263e-05, + "loss": 3.233, + "step": 2005 + }, + { + "epoch": 0.16083061073139443, + "grad_norm": 2.1901161670684814, + "learning_rate": 1.9867100074335654e-05, + "loss": 3.3574, + "step": 2006 + }, + { + "epoch": 0.16091078551241708, + "grad_norm": 2.2308497428894043, + "learning_rate": 1.9866889005053217e-05, + "loss": 3.4388, + "step": 2007 + }, + { + "epoch": 0.1609909602934397, + "grad_norm": 2.076124906539917, + "learning_rate": 1.9866677769418506e-05, + "loss": 3.3083, + "step": 2008 + }, + { + "epoch": 0.16107113507446233, + "grad_norm": 2.160792350769043, + "learning_rate": 1.9866466367435087e-05, + "loss": 3.3776, + "step": 2009 + }, + { + "epoch": 0.16115130985548495, + "grad_norm": 2.1140785217285156, + "learning_rate": 1.9866254799106523e-05, + "loss": 3.3281, + "step": 2010 + }, + { + "epoch": 0.16123148463650758, + "grad_norm": 2.386082172393799, + "learning_rate": 1.9866043064436375e-05, + "loss": 3.3972, + "step": 2011 + }, + { + "epoch": 0.16131165941753023, + "grad_norm": 2.282850980758667, + "learning_rate": 1.986583116342822e-05, + "loss": 3.4454, + "step": 2012 + }, + { + "epoch": 0.16139183419855285, + "grad_norm": 2.10508131980896, + "learning_rate": 1.9865619096085632e-05, + "loss": 3.1629, + "step": 2013 + }, + { + "epoch": 0.16147200897957548, + "grad_norm": 2.3074169158935547, + "learning_rate": 1.9865406862412183e-05, + "loss": 3.4196, + "step": 2014 + }, + { + "epoch": 0.1615521837605981, + "grad_norm": 2.217639684677124, + "learning_rate": 1.9865194462411452e-05, + "loss": 3.3384, + "step": 2015 + }, + { + "epoch": 0.16163235854162072, + "grad_norm": 2.4991538524627686, + "learning_rate": 1.9864981896087016e-05, + "loss": 3.2461, + "step": 2016 + }, + { + "epoch": 0.16171253332264338, + "grad_norm": 2.34298038482666, + "learning_rate": 1.986476916344246e-05, + "loss": 3.2736, + "step": 2017 + }, + { + "epoch": 0.161792708103666, + "grad_norm": 2.425368309020996, + "learning_rate": 1.9864556264481378e-05, + "loss": 3.3787, + "step": 2018 + }, + { + "epoch": 0.16187288288468862, + "grad_norm": 2.1697919368743896, + "learning_rate": 1.986434319920735e-05, + "loss": 3.3187, + "step": 2019 + }, + { + "epoch": 0.16195305766571125, + "grad_norm": 2.1045358180999756, + "learning_rate": 1.9864129967623972e-05, + "loss": 3.2, + "step": 2020 + }, + { + "epoch": 0.16203323244673387, + "grad_norm": 2.1341326236724854, + "learning_rate": 1.9863916569734843e-05, + "loss": 3.3447, + "step": 2021 + }, + { + "epoch": 0.16211340722775652, + "grad_norm": 2.421567678451538, + "learning_rate": 1.9863703005543555e-05, + "loss": 3.3405, + "step": 2022 + }, + { + "epoch": 0.16219358200877915, + "grad_norm": 2.0289740562438965, + "learning_rate": 1.986348927505371e-05, + "loss": 3.2267, + "step": 2023 + }, + { + "epoch": 0.16227375678980177, + "grad_norm": 2.1955008506774902, + "learning_rate": 1.9863275378268914e-05, + "loss": 3.3733, + "step": 2024 + }, + { + "epoch": 0.1623539315708244, + "grad_norm": 2.372415542602539, + "learning_rate": 1.986306131519277e-05, + "loss": 3.381, + "step": 2025 + }, + { + "epoch": 0.16243410635184702, + "grad_norm": 2.1045467853546143, + "learning_rate": 1.9862847085828887e-05, + "loss": 3.2915, + "step": 2026 + }, + { + "epoch": 0.16251428113286964, + "grad_norm": 2.1424002647399902, + "learning_rate": 1.986263269018088e-05, + "loss": 3.2236, + "step": 2027 + }, + { + "epoch": 0.1625944559138923, + "grad_norm": 2.4880874156951904, + "learning_rate": 1.986241812825236e-05, + "loss": 3.3815, + "step": 2028 + }, + { + "epoch": 0.16267463069491492, + "grad_norm": 2.1008591651916504, + "learning_rate": 1.9862203400046945e-05, + "loss": 3.2795, + "step": 2029 + }, + { + "epoch": 0.16275480547593754, + "grad_norm": 2.1767754554748535, + "learning_rate": 1.986198850556826e-05, + "loss": 3.2418, + "step": 2030 + }, + { + "epoch": 0.16283498025696017, + "grad_norm": 2.358548879623413, + "learning_rate": 1.9861773444819924e-05, + "loss": 3.2124, + "step": 2031 + }, + { + "epoch": 0.1629151550379828, + "grad_norm": 2.448943614959717, + "learning_rate": 1.9861558217805562e-05, + "loss": 3.2312, + "step": 2032 + }, + { + "epoch": 0.16299532981900544, + "grad_norm": 1.157619833946228, + "learning_rate": 1.9861342824528807e-05, + "loss": 2.5676, + "step": 2033 + }, + { + "epoch": 0.16307550460002807, + "grad_norm": 2.3122055530548096, + "learning_rate": 1.9861127264993284e-05, + "loss": 3.3547, + "step": 2034 + }, + { + "epoch": 0.1631556793810507, + "grad_norm": 2.331251859664917, + "learning_rate": 1.986091153920263e-05, + "loss": 3.283, + "step": 2035 + }, + { + "epoch": 0.1632358541620733, + "grad_norm": 1.227545142173767, + "learning_rate": 1.9860695647160484e-05, + "loss": 2.5504, + "step": 2036 + }, + { + "epoch": 0.16331602894309594, + "grad_norm": 2.2519912719726562, + "learning_rate": 1.9860479588870488e-05, + "loss": 3.3051, + "step": 2037 + }, + { + "epoch": 0.1633962037241186, + "grad_norm": 2.3007304668426514, + "learning_rate": 1.9860263364336276e-05, + "loss": 3.4089, + "step": 2038 + }, + { + "epoch": 0.1634763785051412, + "grad_norm": 2.1929378509521484, + "learning_rate": 1.98600469735615e-05, + "loss": 3.3947, + "step": 2039 + }, + { + "epoch": 0.16355655328616384, + "grad_norm": 2.4946563243865967, + "learning_rate": 1.985983041654981e-05, + "loss": 3.4395, + "step": 2040 + }, + { + "epoch": 0.16363672806718646, + "grad_norm": 2.227773904800415, + "learning_rate": 1.985961369330485e-05, + "loss": 3.4286, + "step": 2041 + }, + { + "epoch": 0.16371690284820908, + "grad_norm": 2.0558924674987793, + "learning_rate": 1.9859396803830283e-05, + "loss": 3.2627, + "step": 2042 + }, + { + "epoch": 0.16379707762923174, + "grad_norm": 2.053602933883667, + "learning_rate": 1.985917974812976e-05, + "loss": 3.3217, + "step": 2043 + }, + { + "epoch": 0.16387725241025436, + "grad_norm": 2.2575020790100098, + "learning_rate": 1.9858962526206937e-05, + "loss": 3.2746, + "step": 2044 + }, + { + "epoch": 0.16395742719127698, + "grad_norm": 2.247929096221924, + "learning_rate": 1.9858745138065483e-05, + "loss": 3.3724, + "step": 2045 + }, + { + "epoch": 0.1640376019722996, + "grad_norm": 2.4402573108673096, + "learning_rate": 1.985852758370906e-05, + "loss": 3.2971, + "step": 2046 + }, + { + "epoch": 0.16411777675332223, + "grad_norm": 2.1058523654937744, + "learning_rate": 1.9858309863141336e-05, + "loss": 3.2975, + "step": 2047 + }, + { + "epoch": 0.16419795153434488, + "grad_norm": 2.193793296813965, + "learning_rate": 1.9858091976365982e-05, + "loss": 3.246, + "step": 2048 + }, + { + "epoch": 0.1642781263153675, + "grad_norm": 2.261012077331543, + "learning_rate": 1.9857873923386674e-05, + "loss": 3.3861, + "step": 2049 + }, + { + "epoch": 0.16435830109639013, + "grad_norm": 2.3174428939819336, + "learning_rate": 1.9857655704207083e-05, + "loss": 3.3748, + "step": 2050 + }, + { + "epoch": 0.16443847587741275, + "grad_norm": 2.2094533443450928, + "learning_rate": 1.985743731883089e-05, + "loss": 3.2692, + "step": 2051 + }, + { + "epoch": 0.16451865065843538, + "grad_norm": 2.3363120555877686, + "learning_rate": 1.9857218767261783e-05, + "loss": 3.3925, + "step": 2052 + }, + { + "epoch": 0.16459882543945803, + "grad_norm": 2.218790054321289, + "learning_rate": 1.9857000049503438e-05, + "loss": 3.261, + "step": 2053 + }, + { + "epoch": 0.16467900022048065, + "grad_norm": 2.1807937622070312, + "learning_rate": 1.985678116555955e-05, + "loss": 3.3173, + "step": 2054 + }, + { + "epoch": 0.16475917500150328, + "grad_norm": 2.124375104904175, + "learning_rate": 1.98565621154338e-05, + "loss": 3.2888, + "step": 2055 + }, + { + "epoch": 0.1648393497825259, + "grad_norm": 2.1718170642852783, + "learning_rate": 1.9856342899129888e-05, + "loss": 3.3777, + "step": 2056 + }, + { + "epoch": 0.16491952456354853, + "grad_norm": 2.290034294128418, + "learning_rate": 1.9856123516651508e-05, + "loss": 3.3337, + "step": 2057 + }, + { + "epoch": 0.16499969934457118, + "grad_norm": 2.108865737915039, + "learning_rate": 1.985590396800236e-05, + "loss": 3.3677, + "step": 2058 + }, + { + "epoch": 0.1650798741255938, + "grad_norm": 1.2583551406860352, + "learning_rate": 1.9855684253186144e-05, + "loss": 2.5961, + "step": 2059 + }, + { + "epoch": 0.16516004890661642, + "grad_norm": 2.1330573558807373, + "learning_rate": 1.985546437220657e-05, + "loss": 3.3137, + "step": 2060 + }, + { + "epoch": 0.16524022368763905, + "grad_norm": 1.044334888458252, + "learning_rate": 1.9855244325067335e-05, + "loss": 2.5748, + "step": 2061 + }, + { + "epoch": 0.16532039846866167, + "grad_norm": 2.310731887817383, + "learning_rate": 1.9855024111772155e-05, + "loss": 3.3483, + "step": 2062 + }, + { + "epoch": 0.16540057324968432, + "grad_norm": 0.9656704068183899, + "learning_rate": 1.985480373232474e-05, + "loss": 2.5532, + "step": 2063 + }, + { + "epoch": 0.16548074803070695, + "grad_norm": 2.1405954360961914, + "learning_rate": 1.985458318672881e-05, + "loss": 3.2794, + "step": 2064 + }, + { + "epoch": 0.16556092281172957, + "grad_norm": 2.156795024871826, + "learning_rate": 1.985436247498808e-05, + "loss": 3.3539, + "step": 2065 + }, + { + "epoch": 0.1656410975927522, + "grad_norm": 2.175935745239258, + "learning_rate": 1.985414159710627e-05, + "loss": 3.3238, + "step": 2066 + }, + { + "epoch": 0.16572127237377482, + "grad_norm": 1.953257441520691, + "learning_rate": 1.9853920553087108e-05, + "loss": 3.2673, + "step": 2067 + }, + { + "epoch": 0.16580144715479747, + "grad_norm": 2.19228196144104, + "learning_rate": 1.9853699342934316e-05, + "loss": 3.3024, + "step": 2068 + }, + { + "epoch": 0.1658816219358201, + "grad_norm": 2.3753414154052734, + "learning_rate": 1.9853477966651627e-05, + "loss": 3.2694, + "step": 2069 + }, + { + "epoch": 0.16596179671684272, + "grad_norm": 2.0246846675872803, + "learning_rate": 1.9853256424242775e-05, + "loss": 3.2856, + "step": 2070 + }, + { + "epoch": 0.16604197149786534, + "grad_norm": 2.1127939224243164, + "learning_rate": 1.9853034715711488e-05, + "loss": 3.348, + "step": 2071 + }, + { + "epoch": 0.16612214627888797, + "grad_norm": 2.0873055458068848, + "learning_rate": 1.9852812841061508e-05, + "loss": 3.335, + "step": 2072 + }, + { + "epoch": 0.16620232105991062, + "grad_norm": 1.9766894578933716, + "learning_rate": 1.985259080029658e-05, + "loss": 3.2485, + "step": 2073 + }, + { + "epoch": 0.16628249584093324, + "grad_norm": 2.403446674346924, + "learning_rate": 1.985236859342044e-05, + "loss": 3.2958, + "step": 2074 + }, + { + "epoch": 0.16636267062195587, + "grad_norm": 2.1874539852142334, + "learning_rate": 1.985214622043684e-05, + "loss": 3.2559, + "step": 2075 + }, + { + "epoch": 0.1664428454029785, + "grad_norm": 0.9778618812561035, + "learning_rate": 1.9851923681349527e-05, + "loss": 2.5515, + "step": 2076 + }, + { + "epoch": 0.1665230201840011, + "grad_norm": 1.0069622993469238, + "learning_rate": 1.985170097616225e-05, + "loss": 2.5417, + "step": 2077 + }, + { + "epoch": 0.16660319496502377, + "grad_norm": 2.3618102073669434, + "learning_rate": 1.985147810487877e-05, + "loss": 3.2417, + "step": 2078 + }, + { + "epoch": 0.1666833697460464, + "grad_norm": 2.126056671142578, + "learning_rate": 1.9851255067502838e-05, + "loss": 3.2349, + "step": 2079 + }, + { + "epoch": 0.166763544527069, + "grad_norm": 2.2200188636779785, + "learning_rate": 1.985103186403822e-05, + "loss": 3.2083, + "step": 2080 + }, + { + "epoch": 0.16684371930809164, + "grad_norm": 2.1996610164642334, + "learning_rate": 1.9850808494488675e-05, + "loss": 3.3423, + "step": 2081 + }, + { + "epoch": 0.16692389408911426, + "grad_norm": 2.1714484691619873, + "learning_rate": 1.985058495885797e-05, + "loss": 3.3453, + "step": 2082 + }, + { + "epoch": 0.16700406887013688, + "grad_norm": 1.082627534866333, + "learning_rate": 1.985036125714988e-05, + "loss": 2.5709, + "step": 2083 + }, + { + "epoch": 0.16708424365115954, + "grad_norm": 1.923313021659851, + "learning_rate": 1.9850137389368164e-05, + "loss": 3.2085, + "step": 2084 + }, + { + "epoch": 0.16716441843218216, + "grad_norm": 2.1219217777252197, + "learning_rate": 1.9849913355516604e-05, + "loss": 3.2494, + "step": 2085 + }, + { + "epoch": 0.16724459321320478, + "grad_norm": 2.083989381790161, + "learning_rate": 1.984968915559898e-05, + "loss": 3.2351, + "step": 2086 + }, + { + "epoch": 0.1673247679942274, + "grad_norm": 2.1317648887634277, + "learning_rate": 1.9849464789619067e-05, + "loss": 3.316, + "step": 2087 + }, + { + "epoch": 0.16740494277525003, + "grad_norm": 2.315096616744995, + "learning_rate": 1.9849240257580645e-05, + "loss": 3.184, + "step": 2088 + }, + { + "epoch": 0.16748511755627268, + "grad_norm": 2.0960352420806885, + "learning_rate": 1.984901555948751e-05, + "loss": 3.2556, + "step": 2089 + }, + { + "epoch": 0.1675652923372953, + "grad_norm": 2.0564985275268555, + "learning_rate": 1.984879069534344e-05, + "loss": 3.3859, + "step": 2090 + }, + { + "epoch": 0.16764546711831793, + "grad_norm": 2.215658664703369, + "learning_rate": 1.9848565665152228e-05, + "loss": 3.4039, + "step": 2091 + }, + { + "epoch": 0.16772564189934056, + "grad_norm": 2.2368218898773193, + "learning_rate": 1.9848340468917674e-05, + "loss": 3.3254, + "step": 2092 + }, + { + "epoch": 0.16780581668036318, + "grad_norm": 2.234654664993286, + "learning_rate": 1.9848115106643567e-05, + "loss": 3.2108, + "step": 2093 + }, + { + "epoch": 0.16788599146138583, + "grad_norm": 2.2954201698303223, + "learning_rate": 1.9847889578333713e-05, + "loss": 3.4087, + "step": 2094 + }, + { + "epoch": 0.16796616624240845, + "grad_norm": 2.206550121307373, + "learning_rate": 1.9847663883991907e-05, + "loss": 3.2737, + "step": 2095 + }, + { + "epoch": 0.16804634102343108, + "grad_norm": 2.0935580730438232, + "learning_rate": 1.9847438023621963e-05, + "loss": 3.3779, + "step": 2096 + }, + { + "epoch": 0.1681265158044537, + "grad_norm": 2.4165406227111816, + "learning_rate": 1.9847211997227683e-05, + "loss": 3.3089, + "step": 2097 + }, + { + "epoch": 0.16820669058547633, + "grad_norm": 2.1449391841888428, + "learning_rate": 1.984698580481288e-05, + "loss": 3.323, + "step": 2098 + }, + { + "epoch": 0.16828686536649898, + "grad_norm": 2.384033441543579, + "learning_rate": 1.9846759446381365e-05, + "loss": 3.2838, + "step": 2099 + }, + { + "epoch": 0.1683670401475216, + "grad_norm": 2.25594425201416, + "learning_rate": 1.984653292193696e-05, + "loss": 3.3776, + "step": 2100 + }, + { + "epoch": 0.16844721492854423, + "grad_norm": 2.361358642578125, + "learning_rate": 1.9846306231483476e-05, + "loss": 3.2949, + "step": 2101 + }, + { + "epoch": 0.16852738970956685, + "grad_norm": 2.0786619186401367, + "learning_rate": 1.984607937502474e-05, + "loss": 3.268, + "step": 2102 + }, + { + "epoch": 0.16860756449058947, + "grad_norm": 2.2302067279815674, + "learning_rate": 1.9845852352564577e-05, + "loss": 3.3639, + "step": 2103 + }, + { + "epoch": 0.16868773927161212, + "grad_norm": 2.208285093307495, + "learning_rate": 1.9845625164106813e-05, + "loss": 3.3293, + "step": 2104 + }, + { + "epoch": 0.16876791405263475, + "grad_norm": 2.250109910964966, + "learning_rate": 1.9845397809655277e-05, + "loss": 3.5032, + "step": 2105 + }, + { + "epoch": 0.16884808883365737, + "grad_norm": 2.3437323570251465, + "learning_rate": 1.9845170289213805e-05, + "loss": 3.348, + "step": 2106 + }, + { + "epoch": 0.16892826361468, + "grad_norm": 2.4466171264648438, + "learning_rate": 1.9844942602786235e-05, + "loss": 3.3577, + "step": 2107 + }, + { + "epoch": 0.16900843839570262, + "grad_norm": 2.544734477996826, + "learning_rate": 1.9844714750376396e-05, + "loss": 3.3267, + "step": 2108 + }, + { + "epoch": 0.16908861317672527, + "grad_norm": 2.177631378173828, + "learning_rate": 1.984448673198814e-05, + "loss": 3.3781, + "step": 2109 + }, + { + "epoch": 0.1691687879577479, + "grad_norm": 2.216435194015503, + "learning_rate": 1.9844258547625307e-05, + "loss": 3.2646, + "step": 2110 + }, + { + "epoch": 0.16924896273877052, + "grad_norm": 2.215095281600952, + "learning_rate": 1.9844030197291745e-05, + "loss": 3.3646, + "step": 2111 + }, + { + "epoch": 0.16932913751979314, + "grad_norm": 1.0283527374267578, + "learning_rate": 1.9843801680991302e-05, + "loss": 2.5664, + "step": 2112 + }, + { + "epoch": 0.16940931230081577, + "grad_norm": 2.1667866706848145, + "learning_rate": 1.984357299872783e-05, + "loss": 3.465, + "step": 2113 + }, + { + "epoch": 0.16948948708183842, + "grad_norm": 2.8201513290405273, + "learning_rate": 1.9843344150505192e-05, + "loss": 3.2565, + "step": 2114 + }, + { + "epoch": 0.16956966186286104, + "grad_norm": 2.2546567916870117, + "learning_rate": 1.9843115136327236e-05, + "loss": 3.3302, + "step": 2115 + }, + { + "epoch": 0.16964983664388367, + "grad_norm": 2.318136215209961, + "learning_rate": 1.9842885956197827e-05, + "loss": 3.335, + "step": 2116 + }, + { + "epoch": 0.1697300114249063, + "grad_norm": 2.387705087661743, + "learning_rate": 1.9842656610120832e-05, + "loss": 3.3943, + "step": 2117 + }, + { + "epoch": 0.16981018620592891, + "grad_norm": 2.0805916786193848, + "learning_rate": 1.9842427098100116e-05, + "loss": 3.3312, + "step": 2118 + }, + { + "epoch": 0.16989036098695157, + "grad_norm": 2.272442579269409, + "learning_rate": 1.984219742013955e-05, + "loss": 3.3054, + "step": 2119 + }, + { + "epoch": 0.1699705357679742, + "grad_norm": 2.3621742725372314, + "learning_rate": 1.9841967576242996e-05, + "loss": 3.3346, + "step": 2120 + }, + { + "epoch": 0.17005071054899681, + "grad_norm": 2.278641939163208, + "learning_rate": 1.9841737566414345e-05, + "loss": 3.264, + "step": 2121 + }, + { + "epoch": 0.17013088533001944, + "grad_norm": 2.263240337371826, + "learning_rate": 1.9841507390657466e-05, + "loss": 3.4131, + "step": 2122 + }, + { + "epoch": 0.17021106011104206, + "grad_norm": 2.1673710346221924, + "learning_rate": 1.984127704897624e-05, + "loss": 3.2527, + "step": 2123 + }, + { + "epoch": 0.1702912348920647, + "grad_norm": 2.239391565322876, + "learning_rate": 1.984104654137455e-05, + "loss": 3.3406, + "step": 2124 + }, + { + "epoch": 0.17037140967308734, + "grad_norm": 2.273508310317993, + "learning_rate": 1.9840815867856286e-05, + "loss": 3.2957, + "step": 2125 + }, + { + "epoch": 0.17045158445410996, + "grad_norm": 2.223045825958252, + "learning_rate": 1.984058502842533e-05, + "loss": 3.3749, + "step": 2126 + }, + { + "epoch": 0.17053175923513259, + "grad_norm": 2.1824095249176025, + "learning_rate": 1.9840354023085583e-05, + "loss": 3.261, + "step": 2127 + }, + { + "epoch": 0.1706119340161552, + "grad_norm": 2.223565101623535, + "learning_rate": 1.9840122851840933e-05, + "loss": 3.3298, + "step": 2128 + }, + { + "epoch": 0.17069210879717786, + "grad_norm": 2.2460639476776123, + "learning_rate": 1.9839891514695283e-05, + "loss": 3.3976, + "step": 2129 + }, + { + "epoch": 0.17077228357820048, + "grad_norm": 2.484243869781494, + "learning_rate": 1.9839660011652527e-05, + "loss": 3.2626, + "step": 2130 + }, + { + "epoch": 0.1708524583592231, + "grad_norm": 2.3119730949401855, + "learning_rate": 1.9839428342716573e-05, + "loss": 3.2146, + "step": 2131 + }, + { + "epoch": 0.17093263314024573, + "grad_norm": 2.2888729572296143, + "learning_rate": 1.9839196507891323e-05, + "loss": 3.3041, + "step": 2132 + }, + { + "epoch": 0.17101280792126836, + "grad_norm": 2.3239595890045166, + "learning_rate": 1.983896450718069e-05, + "loss": 3.3978, + "step": 2133 + }, + { + "epoch": 0.171092982702291, + "grad_norm": 2.4277701377868652, + "learning_rate": 1.983873234058858e-05, + "loss": 3.4093, + "step": 2134 + }, + { + "epoch": 0.17117315748331363, + "grad_norm": 2.8210043907165527, + "learning_rate": 1.9838500008118912e-05, + "loss": 3.3472, + "step": 2135 + }, + { + "epoch": 0.17125333226433626, + "grad_norm": 2.2373273372650146, + "learning_rate": 1.98382675097756e-05, + "loss": 3.3063, + "step": 2136 + }, + { + "epoch": 0.17133350704535888, + "grad_norm": 2.3943910598754883, + "learning_rate": 1.9838034845562568e-05, + "loss": 3.2527, + "step": 2137 + }, + { + "epoch": 0.1714136818263815, + "grad_norm": 2.138254165649414, + "learning_rate": 1.9837802015483732e-05, + "loss": 3.3616, + "step": 2138 + }, + { + "epoch": 0.17149385660740415, + "grad_norm": 2.1793806552886963, + "learning_rate": 1.9837569019543025e-05, + "loss": 3.4116, + "step": 2139 + }, + { + "epoch": 0.17157403138842678, + "grad_norm": 2.179152011871338, + "learning_rate": 1.983733585774437e-05, + "loss": 3.2189, + "step": 2140 + }, + { + "epoch": 0.1716542061694494, + "grad_norm": 2.0316152572631836, + "learning_rate": 1.98371025300917e-05, + "loss": 3.2943, + "step": 2141 + }, + { + "epoch": 0.17173438095047203, + "grad_norm": 2.2746593952178955, + "learning_rate": 1.9836869036588945e-05, + "loss": 3.3744, + "step": 2142 + }, + { + "epoch": 0.17181455573149465, + "grad_norm": 2.326960325241089, + "learning_rate": 1.9836635377240046e-05, + "loss": 3.3075, + "step": 2143 + }, + { + "epoch": 0.17189473051251727, + "grad_norm": 2.095669984817505, + "learning_rate": 1.9836401552048944e-05, + "loss": 3.401, + "step": 2144 + }, + { + "epoch": 0.17197490529353993, + "grad_norm": 2.28186297416687, + "learning_rate": 1.983616756101958e-05, + "loss": 3.3148, + "step": 2145 + }, + { + "epoch": 0.17205508007456255, + "grad_norm": 2.338709592819214, + "learning_rate": 1.983593340415589e-05, + "loss": 3.2963, + "step": 2146 + }, + { + "epoch": 0.17213525485558517, + "grad_norm": 2.1625616550445557, + "learning_rate": 1.9835699081461833e-05, + "loss": 3.3016, + "step": 2147 + }, + { + "epoch": 0.1722154296366078, + "grad_norm": 2.1669716835021973, + "learning_rate": 1.9835464592941357e-05, + "loss": 3.306, + "step": 2148 + }, + { + "epoch": 0.17229560441763042, + "grad_norm": 2.1677019596099854, + "learning_rate": 1.9835229938598414e-05, + "loss": 3.3555, + "step": 2149 + }, + { + "epoch": 0.17237577919865307, + "grad_norm": 1.2142479419708252, + "learning_rate": 1.9834995118436963e-05, + "loss": 2.5624, + "step": 2150 + }, + { + "epoch": 0.1724559539796757, + "grad_norm": 2.0663883686065674, + "learning_rate": 1.983476013246096e-05, + "loss": 3.275, + "step": 2151 + }, + { + "epoch": 0.17253612876069832, + "grad_norm": 2.1005148887634277, + "learning_rate": 1.9834524980674364e-05, + "loss": 3.3484, + "step": 2152 + }, + { + "epoch": 0.17261630354172094, + "grad_norm": 2.270052433013916, + "learning_rate": 1.9834289663081143e-05, + "loss": 3.2936, + "step": 2153 + }, + { + "epoch": 0.17269647832274357, + "grad_norm": 2.083578109741211, + "learning_rate": 1.9834054179685264e-05, + "loss": 3.2761, + "step": 2154 + }, + { + "epoch": 0.17277665310376622, + "grad_norm": 2.238123893737793, + "learning_rate": 1.98338185304907e-05, + "loss": 3.2371, + "step": 2155 + }, + { + "epoch": 0.17285682788478884, + "grad_norm": 1.2235314846038818, + "learning_rate": 1.983358271550142e-05, + "loss": 2.5786, + "step": 2156 + }, + { + "epoch": 0.17293700266581147, + "grad_norm": 2.221524953842163, + "learning_rate": 1.98333467347214e-05, + "loss": 3.381, + "step": 2157 + }, + { + "epoch": 0.1730171774468341, + "grad_norm": 2.0269522666931152, + "learning_rate": 1.9833110588154622e-05, + "loss": 3.3088, + "step": 2158 + }, + { + "epoch": 0.17309735222785672, + "grad_norm": 2.314709186553955, + "learning_rate": 1.9832874275805064e-05, + "loss": 3.3398, + "step": 2159 + }, + { + "epoch": 0.17317752700887937, + "grad_norm": 2.1633646488189697, + "learning_rate": 1.9832637797676712e-05, + "loss": 3.2496, + "step": 2160 + }, + { + "epoch": 0.173257701789902, + "grad_norm": 2.2899601459503174, + "learning_rate": 1.983240115377355e-05, + "loss": 3.3374, + "step": 2161 + }, + { + "epoch": 0.17333787657092461, + "grad_norm": 2.0982954502105713, + "learning_rate": 1.983216434409957e-05, + "loss": 3.2947, + "step": 2162 + }, + { + "epoch": 0.17341805135194724, + "grad_norm": 2.1574344635009766, + "learning_rate": 1.9831927368658767e-05, + "loss": 3.2717, + "step": 2163 + }, + { + "epoch": 0.17349822613296986, + "grad_norm": 2.324868679046631, + "learning_rate": 1.9831690227455134e-05, + "loss": 3.4075, + "step": 2164 + }, + { + "epoch": 0.17357840091399251, + "grad_norm": 1.972855806350708, + "learning_rate": 1.9831452920492666e-05, + "loss": 3.3084, + "step": 2165 + }, + { + "epoch": 0.17365857569501514, + "grad_norm": 2.118952512741089, + "learning_rate": 1.9831215447775367e-05, + "loss": 3.3688, + "step": 2166 + }, + { + "epoch": 0.17373875047603776, + "grad_norm": 2.2138471603393555, + "learning_rate": 1.9830977809307242e-05, + "loss": 3.3391, + "step": 2167 + }, + { + "epoch": 0.17381892525706039, + "grad_norm": 2.15348219871521, + "learning_rate": 1.9830740005092293e-05, + "loss": 3.1699, + "step": 2168 + }, + { + "epoch": 0.173899100038083, + "grad_norm": 1.1536009311676025, + "learning_rate": 1.9830502035134535e-05, + "loss": 2.5938, + "step": 2169 + }, + { + "epoch": 0.17397927481910566, + "grad_norm": 2.493626594543457, + "learning_rate": 1.9830263899437975e-05, + "loss": 3.3444, + "step": 2170 + }, + { + "epoch": 0.17405944960012829, + "grad_norm": 2.2754178047180176, + "learning_rate": 1.9830025598006633e-05, + "loss": 3.3566, + "step": 2171 + }, + { + "epoch": 0.1741396243811509, + "grad_norm": 2.121807098388672, + "learning_rate": 1.9829787130844522e-05, + "loss": 3.2568, + "step": 2172 + }, + { + "epoch": 0.17421979916217353, + "grad_norm": 2.0406978130340576, + "learning_rate": 1.9829548497955665e-05, + "loss": 3.3342, + "step": 2173 + }, + { + "epoch": 0.17429997394319616, + "grad_norm": 2.177631139755249, + "learning_rate": 1.982930969934408e-05, + "loss": 3.3481, + "step": 2174 + }, + { + "epoch": 0.1743801487242188, + "grad_norm": 2.2692313194274902, + "learning_rate": 1.9829070735013803e-05, + "loss": 3.4066, + "step": 2175 + }, + { + "epoch": 0.17446032350524143, + "grad_norm": 2.2897095680236816, + "learning_rate": 1.9828831604968858e-05, + "loss": 3.3867, + "step": 2176 + }, + { + "epoch": 0.17454049828626406, + "grad_norm": 2.202753782272339, + "learning_rate": 1.9828592309213274e-05, + "loss": 3.302, + "step": 2177 + }, + { + "epoch": 0.17462067306728668, + "grad_norm": 2.2775652408599854, + "learning_rate": 1.982835284775109e-05, + "loss": 3.1896, + "step": 2178 + }, + { + "epoch": 0.1747008478483093, + "grad_norm": 2.4741263389587402, + "learning_rate": 1.982811322058634e-05, + "loss": 3.3378, + "step": 2179 + }, + { + "epoch": 0.17478102262933196, + "grad_norm": 2.154062509536743, + "learning_rate": 1.9827873427723063e-05, + "loss": 3.3129, + "step": 2180 + }, + { + "epoch": 0.17486119741035458, + "grad_norm": 2.186631917953491, + "learning_rate": 1.9827633469165302e-05, + "loss": 3.3297, + "step": 2181 + }, + { + "epoch": 0.1749413721913772, + "grad_norm": 2.239222288131714, + "learning_rate": 1.982739334491711e-05, + "loss": 3.3221, + "step": 2182 + }, + { + "epoch": 0.17502154697239983, + "grad_norm": 2.2402164936065674, + "learning_rate": 1.9827153054982525e-05, + "loss": 3.3817, + "step": 2183 + }, + { + "epoch": 0.17510172175342245, + "grad_norm": 2.5237009525299072, + "learning_rate": 1.9826912599365603e-05, + "loss": 3.302, + "step": 2184 + }, + { + "epoch": 0.1751818965344451, + "grad_norm": 2.10577130317688, + "learning_rate": 1.98266719780704e-05, + "loss": 3.4366, + "step": 2185 + }, + { + "epoch": 0.17526207131546773, + "grad_norm": 2.3105103969573975, + "learning_rate": 1.982643119110097e-05, + "loss": 3.312, + "step": 2186 + }, + { + "epoch": 0.17534224609649035, + "grad_norm": 1.1125974655151367, + "learning_rate": 1.9826190238461373e-05, + "loss": 2.5649, + "step": 2187 + }, + { + "epoch": 0.17542242087751297, + "grad_norm": 2.3424408435821533, + "learning_rate": 1.9825949120155676e-05, + "loss": 3.2427, + "step": 2188 + }, + { + "epoch": 0.1755025956585356, + "grad_norm": 2.1868178844451904, + "learning_rate": 1.9825707836187932e-05, + "loss": 3.3341, + "step": 2189 + }, + { + "epoch": 0.17558277043955825, + "grad_norm": 2.0829780101776123, + "learning_rate": 1.982546638656222e-05, + "loss": 3.3779, + "step": 2190 + }, + { + "epoch": 0.17566294522058087, + "grad_norm": 2.3115029335021973, + "learning_rate": 1.9825224771282608e-05, + "loss": 3.2891, + "step": 2191 + }, + { + "epoch": 0.1757431200016035, + "grad_norm": 2.2162296772003174, + "learning_rate": 1.9824982990353173e-05, + "loss": 3.3152, + "step": 2192 + }, + { + "epoch": 0.17582329478262612, + "grad_norm": 2.286350965499878, + "learning_rate": 1.9824741043777982e-05, + "loss": 3.1659, + "step": 2193 + }, + { + "epoch": 0.17590346956364875, + "grad_norm": 2.1412618160247803, + "learning_rate": 1.9824498931561124e-05, + "loss": 3.2982, + "step": 2194 + }, + { + "epoch": 0.1759836443446714, + "grad_norm": 2.025317430496216, + "learning_rate": 1.982425665370667e-05, + "loss": 3.2947, + "step": 2195 + }, + { + "epoch": 0.17606381912569402, + "grad_norm": 2.2632832527160645, + "learning_rate": 1.9824014210218717e-05, + "loss": 3.3163, + "step": 2196 + }, + { + "epoch": 0.17614399390671664, + "grad_norm": 2.276411533355713, + "learning_rate": 1.982377160110134e-05, + "loss": 3.3264, + "step": 2197 + }, + { + "epoch": 0.17622416868773927, + "grad_norm": 2.3689684867858887, + "learning_rate": 1.9823528826358646e-05, + "loss": 3.3473, + "step": 2198 + }, + { + "epoch": 0.1763043434687619, + "grad_norm": 2.0191967487335205, + "learning_rate": 1.982328588599471e-05, + "loss": 3.2338, + "step": 2199 + }, + { + "epoch": 0.17638451824978452, + "grad_norm": 2.189232587814331, + "learning_rate": 1.982304278001364e-05, + "loss": 3.3151, + "step": 2200 + }, + { + "epoch": 0.17646469303080717, + "grad_norm": 2.062009811401367, + "learning_rate": 1.982279950841953e-05, + "loss": 3.3107, + "step": 2201 + }, + { + "epoch": 0.1765448678118298, + "grad_norm": 2.153679847717285, + "learning_rate": 1.9822556071216482e-05, + "loss": 3.4469, + "step": 2202 + }, + { + "epoch": 0.17662504259285242, + "grad_norm": 2.1213436126708984, + "learning_rate": 1.98223124684086e-05, + "loss": 3.396, + "step": 2203 + }, + { + "epoch": 0.17670521737387504, + "grad_norm": 2.1669862270355225, + "learning_rate": 1.982206869999999e-05, + "loss": 3.3735, + "step": 2204 + }, + { + "epoch": 0.17678539215489766, + "grad_norm": 2.0193734169006348, + "learning_rate": 1.9821824765994767e-05, + "loss": 3.2054, + "step": 2205 + }, + { + "epoch": 0.17686556693592032, + "grad_norm": 2.315067768096924, + "learning_rate": 1.9821580666397037e-05, + "loss": 3.2297, + "step": 2206 + }, + { + "epoch": 0.17694574171694294, + "grad_norm": 2.2271363735198975, + "learning_rate": 1.982133640121092e-05, + "loss": 3.3956, + "step": 2207 + }, + { + "epoch": 0.17702591649796556, + "grad_norm": 2.105565071105957, + "learning_rate": 1.9821091970440533e-05, + "loss": 3.3818, + "step": 2208 + }, + { + "epoch": 0.1771060912789882, + "grad_norm": 2.1116905212402344, + "learning_rate": 1.9820847374089997e-05, + "loss": 3.3163, + "step": 2209 + }, + { + "epoch": 0.1771862660600108, + "grad_norm": 2.6919236183166504, + "learning_rate": 1.9820602612163437e-05, + "loss": 3.3341, + "step": 2210 + }, + { + "epoch": 0.17726644084103346, + "grad_norm": 2.180004119873047, + "learning_rate": 1.982035768466497e-05, + "loss": 3.2897, + "step": 2211 + }, + { + "epoch": 0.1773466156220561, + "grad_norm": 2.1113157272338867, + "learning_rate": 1.982011259159874e-05, + "loss": 3.4137, + "step": 2212 + }, + { + "epoch": 0.1774267904030787, + "grad_norm": 2.3363819122314453, + "learning_rate": 1.9819867332968875e-05, + "loss": 3.3988, + "step": 2213 + }, + { + "epoch": 0.17750696518410133, + "grad_norm": 2.1100406646728516, + "learning_rate": 1.9819621908779504e-05, + "loss": 3.223, + "step": 2214 + }, + { + "epoch": 0.17758713996512396, + "grad_norm": 2.2048587799072266, + "learning_rate": 1.9819376319034766e-05, + "loss": 3.2767, + "step": 2215 + }, + { + "epoch": 0.1776673147461466, + "grad_norm": 2.1221280097961426, + "learning_rate": 1.9819130563738807e-05, + "loss": 3.282, + "step": 2216 + }, + { + "epoch": 0.17774748952716923, + "grad_norm": 2.0548899173736572, + "learning_rate": 1.9818884642895768e-05, + "loss": 3.3141, + "step": 2217 + }, + { + "epoch": 0.17782766430819186, + "grad_norm": 2.0845727920532227, + "learning_rate": 1.9818638556509792e-05, + "loss": 3.3342, + "step": 2218 + }, + { + "epoch": 0.17790783908921448, + "grad_norm": 2.205632209777832, + "learning_rate": 1.981839230458503e-05, + "loss": 3.4156, + "step": 2219 + }, + { + "epoch": 0.1779880138702371, + "grad_norm": 2.128239154815674, + "learning_rate": 1.9818145887125638e-05, + "loss": 3.2513, + "step": 2220 + }, + { + "epoch": 0.17806818865125976, + "grad_norm": 2.293337345123291, + "learning_rate": 1.9817899304135765e-05, + "loss": 3.4149, + "step": 2221 + }, + { + "epoch": 0.17814836343228238, + "grad_norm": 2.16721510887146, + "learning_rate": 1.981765255561957e-05, + "loss": 3.2979, + "step": 2222 + }, + { + "epoch": 0.178228538213305, + "grad_norm": 2.634507417678833, + "learning_rate": 1.981740564158121e-05, + "loss": 3.2981, + "step": 2223 + }, + { + "epoch": 0.17830871299432763, + "grad_norm": 2.1502816677093506, + "learning_rate": 1.9817158562024855e-05, + "loss": 3.387, + "step": 2224 + }, + { + "epoch": 0.17838888777535025, + "grad_norm": 2.1729671955108643, + "learning_rate": 1.9816911316954664e-05, + "loss": 3.3323, + "step": 2225 + }, + { + "epoch": 0.1784690625563729, + "grad_norm": 2.274304151535034, + "learning_rate": 1.981666390637481e-05, + "loss": 3.3425, + "step": 2226 + }, + { + "epoch": 0.17854923733739553, + "grad_norm": 2.513932228088379, + "learning_rate": 1.981641633028946e-05, + "loss": 3.2831, + "step": 2227 + }, + { + "epoch": 0.17862941211841815, + "grad_norm": 2.092432737350464, + "learning_rate": 1.981616858870279e-05, + "loss": 3.2263, + "step": 2228 + }, + { + "epoch": 0.17870958689944078, + "grad_norm": 2.091284990310669, + "learning_rate": 1.9815920681618976e-05, + "loss": 3.3629, + "step": 2229 + }, + { + "epoch": 0.1787897616804634, + "grad_norm": 2.1105616092681885, + "learning_rate": 1.9815672609042202e-05, + "loss": 3.2677, + "step": 2230 + }, + { + "epoch": 0.17886993646148605, + "grad_norm": 2.0605227947235107, + "learning_rate": 1.9815424370976645e-05, + "loss": 3.2227, + "step": 2231 + }, + { + "epoch": 0.17895011124250867, + "grad_norm": 2.0790345668792725, + "learning_rate": 1.9815175967426495e-05, + "loss": 3.3146, + "step": 2232 + }, + { + "epoch": 0.1790302860235313, + "grad_norm": 2.1478517055511475, + "learning_rate": 1.9814927398395936e-05, + "loss": 3.2517, + "step": 2233 + }, + { + "epoch": 0.17911046080455392, + "grad_norm": 2.173302173614502, + "learning_rate": 1.9814678663889162e-05, + "loss": 3.3832, + "step": 2234 + }, + { + "epoch": 0.17919063558557655, + "grad_norm": 2.240675687789917, + "learning_rate": 1.9814429763910362e-05, + "loss": 3.337, + "step": 2235 + }, + { + "epoch": 0.1792708103665992, + "grad_norm": 2.1732213497161865, + "learning_rate": 1.9814180698463737e-05, + "loss": 3.3568, + "step": 2236 + }, + { + "epoch": 0.17935098514762182, + "grad_norm": 2.013343572616577, + "learning_rate": 1.981393146755348e-05, + "loss": 3.3943, + "step": 2237 + }, + { + "epoch": 0.17943115992864445, + "grad_norm": 2.119138240814209, + "learning_rate": 1.9813682071183802e-05, + "loss": 3.3112, + "step": 2238 + }, + { + "epoch": 0.17951133470966707, + "grad_norm": 2.117361068725586, + "learning_rate": 1.9813432509358902e-05, + "loss": 3.2536, + "step": 2239 + }, + { + "epoch": 0.1795915094906897, + "grad_norm": 2.029848575592041, + "learning_rate": 1.981318278208299e-05, + "loss": 3.2479, + "step": 2240 + }, + { + "epoch": 0.17967168427171235, + "grad_norm": 2.164531707763672, + "learning_rate": 1.9812932889360273e-05, + "loss": 3.2328, + "step": 2241 + }, + { + "epoch": 0.17975185905273497, + "grad_norm": 2.2784316539764404, + "learning_rate": 1.9812682831194964e-05, + "loss": 3.3737, + "step": 2242 + }, + { + "epoch": 0.1798320338337576, + "grad_norm": 1.883164405822754, + "learning_rate": 1.9812432607591283e-05, + "loss": 3.2412, + "step": 2243 + }, + { + "epoch": 0.17991220861478022, + "grad_norm": 2.3196609020233154, + "learning_rate": 1.9812182218553447e-05, + "loss": 3.2462, + "step": 2244 + }, + { + "epoch": 0.17999238339580284, + "grad_norm": 1.0781139135360718, + "learning_rate": 1.9811931664085676e-05, + "loss": 2.5607, + "step": 2245 + }, + { + "epoch": 0.1800725581768255, + "grad_norm": 2.1548304557800293, + "learning_rate": 1.9811680944192194e-05, + "loss": 3.2924, + "step": 2246 + }, + { + "epoch": 0.18015273295784812, + "grad_norm": 2.1910674571990967, + "learning_rate": 1.9811430058877233e-05, + "loss": 3.3649, + "step": 2247 + }, + { + "epoch": 0.18023290773887074, + "grad_norm": 2.1037375926971436, + "learning_rate": 1.9811179008145015e-05, + "loss": 3.3081, + "step": 2248 + }, + { + "epoch": 0.18031308251989336, + "grad_norm": 2.102463960647583, + "learning_rate": 1.9810927791999778e-05, + "loss": 3.255, + "step": 2249 + }, + { + "epoch": 0.180393257300916, + "grad_norm": 2.2706828117370605, + "learning_rate": 1.9810676410445757e-05, + "loss": 3.4375, + "step": 2250 + }, + { + "epoch": 0.18047343208193864, + "grad_norm": 2.1482977867126465, + "learning_rate": 1.9810424863487187e-05, + "loss": 3.2585, + "step": 2251 + }, + { + "epoch": 0.18055360686296126, + "grad_norm": 2.2218618392944336, + "learning_rate": 1.9810173151128315e-05, + "loss": 3.2898, + "step": 2252 + }, + { + "epoch": 0.1806337816439839, + "grad_norm": 2.022714376449585, + "learning_rate": 1.9809921273373377e-05, + "loss": 3.3116, + "step": 2253 + }, + { + "epoch": 0.1807139564250065, + "grad_norm": 2.2777321338653564, + "learning_rate": 1.9809669230226628e-05, + "loss": 3.2738, + "step": 2254 + }, + { + "epoch": 0.18079413120602914, + "grad_norm": 2.1909279823303223, + "learning_rate": 1.980941702169231e-05, + "loss": 3.2739, + "step": 2255 + }, + { + "epoch": 0.1808743059870518, + "grad_norm": 2.1429951190948486, + "learning_rate": 1.980916464777468e-05, + "loss": 3.285, + "step": 2256 + }, + { + "epoch": 0.1809544807680744, + "grad_norm": 2.354931354522705, + "learning_rate": 1.980891210847799e-05, + "loss": 3.321, + "step": 2257 + }, + { + "epoch": 0.18103465554909703, + "grad_norm": 2.17895770072937, + "learning_rate": 1.9808659403806498e-05, + "loss": 3.269, + "step": 2258 + }, + { + "epoch": 0.18111483033011966, + "grad_norm": 2.0254647731781006, + "learning_rate": 1.9808406533764467e-05, + "loss": 3.3374, + "step": 2259 + }, + { + "epoch": 0.18119500511114228, + "grad_norm": 2.2211973667144775, + "learning_rate": 1.9808153498356157e-05, + "loss": 3.3495, + "step": 2260 + }, + { + "epoch": 0.1812751798921649, + "grad_norm": 2.0644493103027344, + "learning_rate": 1.9807900297585836e-05, + "loss": 3.3668, + "step": 2261 + }, + { + "epoch": 0.18135535467318756, + "grad_norm": 2.066356658935547, + "learning_rate": 1.9807646931457777e-05, + "loss": 3.2518, + "step": 2262 + }, + { + "epoch": 0.18143552945421018, + "grad_norm": 2.011474609375, + "learning_rate": 1.9807393399976242e-05, + "loss": 3.243, + "step": 2263 + }, + { + "epoch": 0.1815157042352328, + "grad_norm": 2.1263468265533447, + "learning_rate": 1.980713970314551e-05, + "loss": 3.2592, + "step": 2264 + }, + { + "epoch": 0.18159587901625543, + "grad_norm": 2.2767140865325928, + "learning_rate": 1.9806885840969862e-05, + "loss": 3.323, + "step": 2265 + }, + { + "epoch": 0.18167605379727805, + "grad_norm": 1.0752187967300415, + "learning_rate": 1.9806631813453575e-05, + "loss": 2.5635, + "step": 2266 + }, + { + "epoch": 0.1817562285783007, + "grad_norm": 1.9324227571487427, + "learning_rate": 1.9806377620600933e-05, + "loss": 3.1234, + "step": 2267 + }, + { + "epoch": 0.18183640335932333, + "grad_norm": 2.2059292793273926, + "learning_rate": 1.9806123262416217e-05, + "loss": 3.2767, + "step": 2268 + }, + { + "epoch": 0.18191657814034595, + "grad_norm": 2.469486951828003, + "learning_rate": 1.980586873890372e-05, + "loss": 3.2598, + "step": 2269 + }, + { + "epoch": 0.18199675292136858, + "grad_norm": 2.163763999938965, + "learning_rate": 1.9805614050067735e-05, + "loss": 3.392, + "step": 2270 + }, + { + "epoch": 0.1820769277023912, + "grad_norm": 2.1846723556518555, + "learning_rate": 1.980535919591255e-05, + "loss": 3.3907, + "step": 2271 + }, + { + "epoch": 0.18215710248341385, + "grad_norm": 2.084188461303711, + "learning_rate": 1.9805104176442464e-05, + "loss": 3.3708, + "step": 2272 + }, + { + "epoch": 0.18223727726443648, + "grad_norm": 2.0985255241394043, + "learning_rate": 1.980484899166178e-05, + "loss": 3.2628, + "step": 2273 + }, + { + "epoch": 0.1823174520454591, + "grad_norm": 2.201442003250122, + "learning_rate": 1.9804593641574793e-05, + "loss": 3.3007, + "step": 2274 + }, + { + "epoch": 0.18239762682648172, + "grad_norm": 2.0725197792053223, + "learning_rate": 1.9804338126185815e-05, + "loss": 3.2572, + "step": 2275 + }, + { + "epoch": 0.18247780160750435, + "grad_norm": 2.1264846324920654, + "learning_rate": 1.9804082445499154e-05, + "loss": 3.4349, + "step": 2276 + }, + { + "epoch": 0.182557976388527, + "grad_norm": 2.1773860454559326, + "learning_rate": 1.9803826599519115e-05, + "loss": 3.2826, + "step": 2277 + }, + { + "epoch": 0.18263815116954962, + "grad_norm": 2.2766196727752686, + "learning_rate": 1.9803570588250015e-05, + "loss": 3.3409, + "step": 2278 + }, + { + "epoch": 0.18271832595057225, + "grad_norm": 2.109577178955078, + "learning_rate": 1.9803314411696175e-05, + "loss": 3.2511, + "step": 2279 + }, + { + "epoch": 0.18279850073159487, + "grad_norm": 2.304473876953125, + "learning_rate": 1.9803058069861904e-05, + "loss": 3.3359, + "step": 2280 + }, + { + "epoch": 0.1828786755126175, + "grad_norm": 2.289285182952881, + "learning_rate": 1.980280156275153e-05, + "loss": 3.2819, + "step": 2281 + }, + { + "epoch": 0.18295885029364015, + "grad_norm": 1.2111084461212158, + "learning_rate": 1.9802544890369378e-05, + "loss": 2.5657, + "step": 2282 + }, + { + "epoch": 0.18303902507466277, + "grad_norm": 2.1157374382019043, + "learning_rate": 1.980228805271977e-05, + "loss": 3.3992, + "step": 2283 + }, + { + "epoch": 0.1831191998556854, + "grad_norm": 2.1747891902923584, + "learning_rate": 1.980203104980704e-05, + "loss": 3.3482, + "step": 2284 + }, + { + "epoch": 0.18319937463670802, + "grad_norm": 2.229341506958008, + "learning_rate": 1.9801773881635527e-05, + "loss": 3.2803, + "step": 2285 + }, + { + "epoch": 0.18327954941773064, + "grad_norm": 2.2051751613616943, + "learning_rate": 1.980151654820956e-05, + "loss": 3.2668, + "step": 2286 + }, + { + "epoch": 0.1833597241987533, + "grad_norm": 2.1356120109558105, + "learning_rate": 1.980125904953347e-05, + "loss": 3.2939, + "step": 2287 + }, + { + "epoch": 0.18343989897977592, + "grad_norm": 2.2116470336914062, + "learning_rate": 1.9801001385611615e-05, + "loss": 3.2266, + "step": 2288 + }, + { + "epoch": 0.18352007376079854, + "grad_norm": 2.130767822265625, + "learning_rate": 1.980074355644833e-05, + "loss": 3.332, + "step": 2289 + }, + { + "epoch": 0.18360024854182116, + "grad_norm": 2.205773115158081, + "learning_rate": 1.9800485562047964e-05, + "loss": 3.4026, + "step": 2290 + }, + { + "epoch": 0.1836804233228438, + "grad_norm": 2.0002248287200928, + "learning_rate": 1.980022740241486e-05, + "loss": 3.2871, + "step": 2291 + }, + { + "epoch": 0.18376059810386644, + "grad_norm": 2.078160524368286, + "learning_rate": 1.9799969077553376e-05, + "loss": 3.3979, + "step": 2292 + }, + { + "epoch": 0.18384077288488906, + "grad_norm": 2.169721841812134, + "learning_rate": 1.979971058746787e-05, + "loss": 3.2371, + "step": 2293 + }, + { + "epoch": 0.1839209476659117, + "grad_norm": 2.189279794692993, + "learning_rate": 1.9799451932162698e-05, + "loss": 3.2832, + "step": 2294 + }, + { + "epoch": 0.1840011224469343, + "grad_norm": 2.2113754749298096, + "learning_rate": 1.9799193111642218e-05, + "loss": 3.3604, + "step": 2295 + }, + { + "epoch": 0.18408129722795694, + "grad_norm": 2.152120590209961, + "learning_rate": 1.9798934125910798e-05, + "loss": 3.3159, + "step": 2296 + }, + { + "epoch": 0.1841614720089796, + "grad_norm": 2.0025157928466797, + "learning_rate": 1.97986749749728e-05, + "loss": 3.3253, + "step": 2297 + }, + { + "epoch": 0.1842416467900022, + "grad_norm": 2.252350091934204, + "learning_rate": 1.97984156588326e-05, + "loss": 3.3029, + "step": 2298 + }, + { + "epoch": 0.18432182157102484, + "grad_norm": 2.0877225399017334, + "learning_rate": 1.979815617749456e-05, + "loss": 3.2787, + "step": 2299 + }, + { + "epoch": 0.18440199635204746, + "grad_norm": 2.06306529045105, + "learning_rate": 1.9797896530963062e-05, + "loss": 3.4041, + "step": 2300 + }, + { + "epoch": 0.18448217113307008, + "grad_norm": 2.1800172328948975, + "learning_rate": 1.9797636719242483e-05, + "loss": 3.3363, + "step": 2301 + }, + { + "epoch": 0.18456234591409273, + "grad_norm": 2.342158555984497, + "learning_rate": 1.9797376742337197e-05, + "loss": 3.3894, + "step": 2302 + }, + { + "epoch": 0.18464252069511536, + "grad_norm": 2.3280560970306396, + "learning_rate": 1.9797116600251597e-05, + "loss": 3.3391, + "step": 2303 + }, + { + "epoch": 0.18472269547613798, + "grad_norm": 2.09956955909729, + "learning_rate": 1.979685629299006e-05, + "loss": 3.2662, + "step": 2304 + }, + { + "epoch": 0.1848028702571606, + "grad_norm": 2.2837367057800293, + "learning_rate": 1.9796595820556983e-05, + "loss": 3.2958, + "step": 2305 + }, + { + "epoch": 0.18488304503818323, + "grad_norm": 2.3097784519195557, + "learning_rate": 1.979633518295675e-05, + "loss": 3.4606, + "step": 2306 + }, + { + "epoch": 0.18496321981920588, + "grad_norm": 2.1373558044433594, + "learning_rate": 1.9796074380193758e-05, + "loss": 3.3745, + "step": 2307 + }, + { + "epoch": 0.1850433946002285, + "grad_norm": 2.0359785556793213, + "learning_rate": 1.9795813412272406e-05, + "loss": 3.2587, + "step": 2308 + }, + { + "epoch": 0.18512356938125113, + "grad_norm": 2.1642818450927734, + "learning_rate": 1.9795552279197092e-05, + "loss": 3.3443, + "step": 2309 + }, + { + "epoch": 0.18520374416227375, + "grad_norm": 2.603459119796753, + "learning_rate": 1.979529098097222e-05, + "loss": 3.2538, + "step": 2310 + }, + { + "epoch": 0.18528391894329638, + "grad_norm": 1.9758354425430298, + "learning_rate": 1.9795029517602192e-05, + "loss": 3.1746, + "step": 2311 + }, + { + "epoch": 0.18536409372431903, + "grad_norm": 2.1100077629089355, + "learning_rate": 1.9794767889091418e-05, + "loss": 3.2346, + "step": 2312 + }, + { + "epoch": 0.18544426850534165, + "grad_norm": 2.247379779815674, + "learning_rate": 1.979450609544431e-05, + "loss": 3.2327, + "step": 2313 + }, + { + "epoch": 0.18552444328636428, + "grad_norm": 2.019476890563965, + "learning_rate": 1.9794244136665284e-05, + "loss": 3.2697, + "step": 2314 + }, + { + "epoch": 0.1856046180673869, + "grad_norm": 2.256378173828125, + "learning_rate": 1.9793982012758754e-05, + "loss": 3.2919, + "step": 2315 + }, + { + "epoch": 0.18568479284840952, + "grad_norm": 2.2409777641296387, + "learning_rate": 1.9793719723729133e-05, + "loss": 3.2598, + "step": 2316 + }, + { + "epoch": 0.18576496762943215, + "grad_norm": 2.0727996826171875, + "learning_rate": 1.9793457269580855e-05, + "loss": 3.2435, + "step": 2317 + }, + { + "epoch": 0.1858451424104548, + "grad_norm": 2.1669158935546875, + "learning_rate": 1.9793194650318337e-05, + "loss": 3.4215, + "step": 2318 + }, + { + "epoch": 0.18592531719147742, + "grad_norm": 2.2803239822387695, + "learning_rate": 1.9792931865946007e-05, + "loss": 3.2147, + "step": 2319 + }, + { + "epoch": 0.18600549197250005, + "grad_norm": 2.0487375259399414, + "learning_rate": 1.9792668916468297e-05, + "loss": 3.2286, + "step": 2320 + }, + { + "epoch": 0.18608566675352267, + "grad_norm": 2.1251466274261475, + "learning_rate": 1.9792405801889646e-05, + "loss": 3.2617, + "step": 2321 + }, + { + "epoch": 0.1861658415345453, + "grad_norm": 2.060173273086548, + "learning_rate": 1.979214252221448e-05, + "loss": 3.3534, + "step": 2322 + }, + { + "epoch": 0.18624601631556795, + "grad_norm": 2.006978750228882, + "learning_rate": 1.979187907744724e-05, + "loss": 3.2356, + "step": 2323 + }, + { + "epoch": 0.18632619109659057, + "grad_norm": 2.187652826309204, + "learning_rate": 1.9791615467592375e-05, + "loss": 3.2471, + "step": 2324 + }, + { + "epoch": 0.1864063658776132, + "grad_norm": 1.1786742210388184, + "learning_rate": 1.979135169265432e-05, + "loss": 2.5459, + "step": 2325 + }, + { + "epoch": 0.18648654065863582, + "grad_norm": 2.519792079925537, + "learning_rate": 1.9791087752637528e-05, + "loss": 3.3053, + "step": 2326 + }, + { + "epoch": 0.18656671543965844, + "grad_norm": 2.1136765480041504, + "learning_rate": 1.9790823647546446e-05, + "loss": 3.3163, + "step": 2327 + }, + { + "epoch": 0.1866468902206811, + "grad_norm": 2.254910945892334, + "learning_rate": 1.979055937738553e-05, + "loss": 3.3223, + "step": 2328 + }, + { + "epoch": 0.18672706500170372, + "grad_norm": 2.154163360595703, + "learning_rate": 1.9790294942159233e-05, + "loss": 3.3258, + "step": 2329 + }, + { + "epoch": 0.18680723978272634, + "grad_norm": 2.1931910514831543, + "learning_rate": 1.9790030341872012e-05, + "loss": 3.4059, + "step": 2330 + }, + { + "epoch": 0.18688741456374897, + "grad_norm": 2.148287773132324, + "learning_rate": 1.978976557652833e-05, + "loss": 3.1869, + "step": 2331 + }, + { + "epoch": 0.1869675893447716, + "grad_norm": 1.1808720827102661, + "learning_rate": 1.978950064613265e-05, + "loss": 2.5586, + "step": 2332 + }, + { + "epoch": 0.18704776412579424, + "grad_norm": 2.350959062576294, + "learning_rate": 1.978923555068944e-05, + "loss": 3.3572, + "step": 2333 + }, + { + "epoch": 0.18712793890681687, + "grad_norm": 2.2941904067993164, + "learning_rate": 1.978897029020317e-05, + "loss": 3.3684, + "step": 2334 + }, + { + "epoch": 0.1872081136878395, + "grad_norm": 2.338872194290161, + "learning_rate": 1.978870486467831e-05, + "loss": 3.4573, + "step": 2335 + }, + { + "epoch": 0.1872882884688621, + "grad_norm": 2.0982003211975098, + "learning_rate": 1.9788439274119335e-05, + "loss": 3.2909, + "step": 2336 + }, + { + "epoch": 0.18736846324988474, + "grad_norm": 2.196716785430908, + "learning_rate": 1.9788173518530725e-05, + "loss": 3.417, + "step": 2337 + }, + { + "epoch": 0.1874486380309074, + "grad_norm": 2.1577954292297363, + "learning_rate": 1.978790759791696e-05, + "loss": 3.2865, + "step": 2338 + }, + { + "epoch": 0.18752881281193, + "grad_norm": 2.074876546859741, + "learning_rate": 1.978764151228252e-05, + "loss": 3.3049, + "step": 2339 + }, + { + "epoch": 0.18760898759295264, + "grad_norm": 2.1687216758728027, + "learning_rate": 1.9787375261631894e-05, + "loss": 3.3004, + "step": 2340 + }, + { + "epoch": 0.18768916237397526, + "grad_norm": 2.1731722354888916, + "learning_rate": 1.9787108845969573e-05, + "loss": 3.2999, + "step": 2341 + }, + { + "epoch": 0.18776933715499788, + "grad_norm": 2.101536750793457, + "learning_rate": 1.9786842265300043e-05, + "loss": 3.3205, + "step": 2342 + }, + { + "epoch": 0.18784951193602054, + "grad_norm": 2.3469491004943848, + "learning_rate": 1.9786575519627804e-05, + "loss": 3.3434, + "step": 2343 + }, + { + "epoch": 0.18792968671704316, + "grad_norm": 2.201876163482666, + "learning_rate": 1.978630860895735e-05, + "loss": 3.2817, + "step": 2344 + }, + { + "epoch": 0.18800986149806578, + "grad_norm": 2.2012808322906494, + "learning_rate": 1.9786041533293182e-05, + "loss": 3.3702, + "step": 2345 + }, + { + "epoch": 0.1880900362790884, + "grad_norm": 2.0670740604400635, + "learning_rate": 1.9785774292639802e-05, + "loss": 3.3391, + "step": 2346 + }, + { + "epoch": 0.18817021106011103, + "grad_norm": 2.1602468490600586, + "learning_rate": 1.9785506887001717e-05, + "loss": 3.2415, + "step": 2347 + }, + { + "epoch": 0.18825038584113368, + "grad_norm": 2.1734583377838135, + "learning_rate": 1.9785239316383436e-05, + "loss": 3.2643, + "step": 2348 + }, + { + "epoch": 0.1883305606221563, + "grad_norm": 2.290832042694092, + "learning_rate": 1.9784971580789464e-05, + "loss": 3.3054, + "step": 2349 + }, + { + "epoch": 0.18841073540317893, + "grad_norm": 2.1303393840789795, + "learning_rate": 1.9784703680224325e-05, + "loss": 3.4381, + "step": 2350 + }, + { + "epoch": 0.18849091018420155, + "grad_norm": 2.3698182106018066, + "learning_rate": 1.9784435614692527e-05, + "loss": 3.1929, + "step": 2351 + }, + { + "epoch": 0.18857108496522418, + "grad_norm": 2.066033363342285, + "learning_rate": 1.9784167384198592e-05, + "loss": 3.3755, + "step": 2352 + }, + { + "epoch": 0.18865125974624683, + "grad_norm": 2.207127571105957, + "learning_rate": 1.9783898988747048e-05, + "loss": 3.2589, + "step": 2353 + }, + { + "epoch": 0.18873143452726945, + "grad_norm": 2.112621545791626, + "learning_rate": 1.9783630428342414e-05, + "loss": 3.2881, + "step": 2354 + }, + { + "epoch": 0.18881160930829208, + "grad_norm": 2.191512107849121, + "learning_rate": 1.9783361702989215e-05, + "loss": 3.3691, + "step": 2355 + }, + { + "epoch": 0.1888917840893147, + "grad_norm": 2.110112190246582, + "learning_rate": 1.978309281269199e-05, + "loss": 3.2838, + "step": 2356 + }, + { + "epoch": 0.18897195887033733, + "grad_norm": 1.033992886543274, + "learning_rate": 1.9782823757455265e-05, + "loss": 2.5533, + "step": 2357 + }, + { + "epoch": 0.18905213365135998, + "grad_norm": 1.028676152229309, + "learning_rate": 1.978255453728358e-05, + "loss": 2.5526, + "step": 2358 + }, + { + "epoch": 0.1891323084323826, + "grad_norm": 2.099520444869995, + "learning_rate": 1.9782285152181474e-05, + "loss": 3.3326, + "step": 2359 + }, + { + "epoch": 0.18921248321340522, + "grad_norm": 2.2180051803588867, + "learning_rate": 1.978201560215349e-05, + "loss": 3.3608, + "step": 2360 + }, + { + "epoch": 0.18929265799442785, + "grad_norm": 2.0741641521453857, + "learning_rate": 1.9781745887204166e-05, + "loss": 3.3981, + "step": 2361 + }, + { + "epoch": 0.18937283277545047, + "grad_norm": 2.344248056411743, + "learning_rate": 1.9781476007338058e-05, + "loss": 3.2739, + "step": 2362 + }, + { + "epoch": 0.18945300755647312, + "grad_norm": 1.138851284980774, + "learning_rate": 1.978120596255971e-05, + "loss": 2.559, + "step": 2363 + }, + { + "epoch": 0.18953318233749575, + "grad_norm": 2.111957311630249, + "learning_rate": 1.9780935752873676e-05, + "loss": 3.2596, + "step": 2364 + }, + { + "epoch": 0.18961335711851837, + "grad_norm": 2.085519790649414, + "learning_rate": 1.9780665378284514e-05, + "loss": 3.3044, + "step": 2365 + }, + { + "epoch": 0.189693531899541, + "grad_norm": 2.0622401237487793, + "learning_rate": 1.978039483879678e-05, + "loss": 3.1864, + "step": 2366 + }, + { + "epoch": 0.18977370668056362, + "grad_norm": 2.2626571655273438, + "learning_rate": 1.9780124134415035e-05, + "loss": 3.2963, + "step": 2367 + }, + { + "epoch": 0.18985388146158627, + "grad_norm": 2.3058645725250244, + "learning_rate": 1.9779853265143845e-05, + "loss": 3.323, + "step": 2368 + }, + { + "epoch": 0.1899340562426089, + "grad_norm": 2.160595178604126, + "learning_rate": 1.9779582230987776e-05, + "loss": 3.261, + "step": 2369 + }, + { + "epoch": 0.19001423102363152, + "grad_norm": 2.0894927978515625, + "learning_rate": 1.9779311031951396e-05, + "loss": 3.1936, + "step": 2370 + }, + { + "epoch": 0.19009440580465414, + "grad_norm": 2.3322126865386963, + "learning_rate": 1.9779039668039278e-05, + "loss": 3.3521, + "step": 2371 + }, + { + "epoch": 0.19017458058567677, + "grad_norm": 2.2800581455230713, + "learning_rate": 1.9778768139255998e-05, + "loss": 3.3113, + "step": 2372 + }, + { + "epoch": 0.19025475536669942, + "grad_norm": 2.2540066242218018, + "learning_rate": 1.9778496445606134e-05, + "loss": 3.254, + "step": 2373 + }, + { + "epoch": 0.19033493014772204, + "grad_norm": 2.113675355911255, + "learning_rate": 1.977822458709427e-05, + "loss": 3.1965, + "step": 2374 + }, + { + "epoch": 0.19041510492874467, + "grad_norm": 1.9738715887069702, + "learning_rate": 1.977795256372498e-05, + "loss": 3.2942, + "step": 2375 + }, + { + "epoch": 0.1904952797097673, + "grad_norm": 2.0294923782348633, + "learning_rate": 1.977768037550286e-05, + "loss": 3.2886, + "step": 2376 + }, + { + "epoch": 0.19057545449078991, + "grad_norm": 2.0673255920410156, + "learning_rate": 1.9777408022432492e-05, + "loss": 3.257, + "step": 2377 + }, + { + "epoch": 0.19065562927181254, + "grad_norm": 2.22782826423645, + "learning_rate": 1.977713550451847e-05, + "loss": 3.2794, + "step": 2378 + }, + { + "epoch": 0.1907358040528352, + "grad_norm": 2.2623071670532227, + "learning_rate": 1.9776862821765392e-05, + "loss": 3.2714, + "step": 2379 + }, + { + "epoch": 0.1908159788338578, + "grad_norm": 2.0365400314331055, + "learning_rate": 1.977658997417785e-05, + "loss": 3.4015, + "step": 2380 + }, + { + "epoch": 0.19089615361488044, + "grad_norm": 2.1056361198425293, + "learning_rate": 1.977631696176045e-05, + "loss": 3.2883, + "step": 2381 + }, + { + "epoch": 0.19097632839590306, + "grad_norm": 2.2555439472198486, + "learning_rate": 1.977604378451779e-05, + "loss": 3.313, + "step": 2382 + }, + { + "epoch": 0.19105650317692568, + "grad_norm": 2.0494799613952637, + "learning_rate": 1.9775770442454476e-05, + "loss": 3.3108, + "step": 2383 + }, + { + "epoch": 0.19113667795794834, + "grad_norm": 2.1030423641204834, + "learning_rate": 1.9775496935575117e-05, + "loss": 3.4261, + "step": 2384 + }, + { + "epoch": 0.19121685273897096, + "grad_norm": 2.2291839122772217, + "learning_rate": 1.9775223263884325e-05, + "loss": 3.2418, + "step": 2385 + }, + { + "epoch": 0.19129702751999358, + "grad_norm": 2.15889835357666, + "learning_rate": 1.9774949427386716e-05, + "loss": 3.3466, + "step": 2386 + }, + { + "epoch": 0.1913772023010162, + "grad_norm": 2.239522695541382, + "learning_rate": 1.9774675426086904e-05, + "loss": 3.3263, + "step": 2387 + }, + { + "epoch": 0.19145737708203883, + "grad_norm": 2.09946608543396, + "learning_rate": 1.977440125998951e-05, + "loss": 3.3713, + "step": 2388 + }, + { + "epoch": 0.19153755186306148, + "grad_norm": 2.0813286304473877, + "learning_rate": 1.9774126929099153e-05, + "loss": 3.3915, + "step": 2389 + }, + { + "epoch": 0.1916177266440841, + "grad_norm": 2.201869010925293, + "learning_rate": 1.9773852433420463e-05, + "loss": 3.3768, + "step": 2390 + }, + { + "epoch": 0.19169790142510673, + "grad_norm": 2.1243691444396973, + "learning_rate": 1.9773577772958066e-05, + "loss": 3.3186, + "step": 2391 + }, + { + "epoch": 0.19177807620612936, + "grad_norm": 2.1926937103271484, + "learning_rate": 1.977330294771659e-05, + "loss": 3.2487, + "step": 2392 + }, + { + "epoch": 0.19185825098715198, + "grad_norm": 1.9485934972763062, + "learning_rate": 1.977302795770067e-05, + "loss": 3.2515, + "step": 2393 + }, + { + "epoch": 0.19193842576817463, + "grad_norm": 2.3582863807678223, + "learning_rate": 1.977275280291495e-05, + "loss": 3.3199, + "step": 2394 + }, + { + "epoch": 0.19201860054919725, + "grad_norm": 2.0358307361602783, + "learning_rate": 1.9772477483364052e-05, + "loss": 3.3728, + "step": 2395 + }, + { + "epoch": 0.19209877533021988, + "grad_norm": 2.111107587814331, + "learning_rate": 1.9772201999052637e-05, + "loss": 3.2495, + "step": 2396 + }, + { + "epoch": 0.1921789501112425, + "grad_norm": 1.995641827583313, + "learning_rate": 1.9771926349985334e-05, + "loss": 3.2556, + "step": 2397 + }, + { + "epoch": 0.19225912489226513, + "grad_norm": 0.9694114327430725, + "learning_rate": 1.97716505361668e-05, + "loss": 2.5409, + "step": 2398 + }, + { + "epoch": 0.19233929967328778, + "grad_norm": 2.1919851303100586, + "learning_rate": 1.9771374557601678e-05, + "loss": 3.2351, + "step": 2399 + }, + { + "epoch": 0.1924194744543104, + "grad_norm": 2.136730194091797, + "learning_rate": 1.977109841429463e-05, + "loss": 3.3212, + "step": 2400 + }, + { + "epoch": 0.19249964923533303, + "grad_norm": 2.2385568618774414, + "learning_rate": 1.9770822106250302e-05, + "loss": 3.2415, + "step": 2401 + }, + { + "epoch": 0.19257982401635565, + "grad_norm": 3.1672165393829346, + "learning_rate": 1.977054563347336e-05, + "loss": 3.3105, + "step": 2402 + }, + { + "epoch": 0.19265999879737827, + "grad_norm": 2.0314102172851562, + "learning_rate": 1.977026899596846e-05, + "loss": 3.2445, + "step": 2403 + }, + { + "epoch": 0.19274017357840093, + "grad_norm": 2.053314447402954, + "learning_rate": 1.9769992193740267e-05, + "loss": 3.2188, + "step": 2404 + }, + { + "epoch": 0.19282034835942355, + "grad_norm": 2.208404541015625, + "learning_rate": 1.976971522679345e-05, + "loss": 3.3547, + "step": 2405 + }, + { + "epoch": 0.19290052314044617, + "grad_norm": 2.256166934967041, + "learning_rate": 1.976943809513268e-05, + "loss": 3.2776, + "step": 2406 + }, + { + "epoch": 0.1929806979214688, + "grad_norm": 2.2864785194396973, + "learning_rate": 1.9769160798762624e-05, + "loss": 3.3108, + "step": 2407 + }, + { + "epoch": 0.19306087270249142, + "grad_norm": 2.262070655822754, + "learning_rate": 1.976888333768796e-05, + "loss": 3.2908, + "step": 2408 + }, + { + "epoch": 0.19314104748351407, + "grad_norm": 1.1729744672775269, + "learning_rate": 1.976860571191337e-05, + "loss": 2.5737, + "step": 2409 + }, + { + "epoch": 0.1932212222645367, + "grad_norm": 2.301281213760376, + "learning_rate": 1.9768327921443523e-05, + "loss": 3.2563, + "step": 2410 + }, + { + "epoch": 0.19330139704555932, + "grad_norm": 2.350431442260742, + "learning_rate": 1.9768049966283116e-05, + "loss": 3.3044, + "step": 2411 + }, + { + "epoch": 0.19338157182658194, + "grad_norm": 2.263662815093994, + "learning_rate": 1.9767771846436827e-05, + "loss": 3.2798, + "step": 2412 + }, + { + "epoch": 0.19346174660760457, + "grad_norm": 2.1144912242889404, + "learning_rate": 1.976749356190935e-05, + "loss": 3.2957, + "step": 2413 + }, + { + "epoch": 0.19354192138862722, + "grad_norm": 2.0395736694335938, + "learning_rate": 1.9767215112705368e-05, + "loss": 3.2375, + "step": 2414 + }, + { + "epoch": 0.19362209616964984, + "grad_norm": 2.269385814666748, + "learning_rate": 1.976693649882959e-05, + "loss": 3.2931, + "step": 2415 + }, + { + "epoch": 0.19370227095067247, + "grad_norm": 2.2585504055023193, + "learning_rate": 1.97666577202867e-05, + "loss": 3.3335, + "step": 2416 + }, + { + "epoch": 0.1937824457316951, + "grad_norm": 2.105492115020752, + "learning_rate": 1.97663787770814e-05, + "loss": 3.2527, + "step": 2417 + }, + { + "epoch": 0.19386262051271771, + "grad_norm": 2.218930721282959, + "learning_rate": 1.9766099669218398e-05, + "loss": 3.4054, + "step": 2418 + }, + { + "epoch": 0.19394279529374037, + "grad_norm": 1.954771637916565, + "learning_rate": 1.9765820396702397e-05, + "loss": 3.245, + "step": 2419 + }, + { + "epoch": 0.194022970074763, + "grad_norm": 2.399782180786133, + "learning_rate": 1.9765540959538105e-05, + "loss": 3.4374, + "step": 2420 + }, + { + "epoch": 0.19410314485578561, + "grad_norm": 2.4057514667510986, + "learning_rate": 1.976526135773024e-05, + "loss": 3.3936, + "step": 2421 + }, + { + "epoch": 0.19418331963680824, + "grad_norm": 2.2035226821899414, + "learning_rate": 1.9764981591283505e-05, + "loss": 3.3654, + "step": 2422 + }, + { + "epoch": 0.19426349441783086, + "grad_norm": 2.0323662757873535, + "learning_rate": 1.9764701660202622e-05, + "loss": 3.4043, + "step": 2423 + }, + { + "epoch": 0.1943436691988535, + "grad_norm": 2.2176146507263184, + "learning_rate": 1.976442156449231e-05, + "loss": 3.3029, + "step": 2424 + }, + { + "epoch": 0.19442384397987614, + "grad_norm": 2.287065267562866, + "learning_rate": 1.9764141304157295e-05, + "loss": 3.3473, + "step": 2425 + }, + { + "epoch": 0.19450401876089876, + "grad_norm": 2.208211660385132, + "learning_rate": 1.9763860879202296e-05, + "loss": 3.2827, + "step": 2426 + }, + { + "epoch": 0.19458419354192139, + "grad_norm": 2.1996259689331055, + "learning_rate": 1.9763580289632044e-05, + "loss": 3.3171, + "step": 2427 + }, + { + "epoch": 0.194664368322944, + "grad_norm": 2.1775994300842285, + "learning_rate": 1.9763299535451273e-05, + "loss": 3.3287, + "step": 2428 + }, + { + "epoch": 0.19474454310396666, + "grad_norm": 2.2453813552856445, + "learning_rate": 1.9763018616664705e-05, + "loss": 3.2936, + "step": 2429 + }, + { + "epoch": 0.19482471788498928, + "grad_norm": 2.2953736782073975, + "learning_rate": 1.976273753327709e-05, + "loss": 3.2325, + "step": 2430 + }, + { + "epoch": 0.1949048926660119, + "grad_norm": 2.3274569511413574, + "learning_rate": 1.976245628529316e-05, + "loss": 3.2491, + "step": 2431 + }, + { + "epoch": 0.19498506744703453, + "grad_norm": 2.2854866981506348, + "learning_rate": 1.976217487271766e-05, + "loss": 3.2994, + "step": 2432 + }, + { + "epoch": 0.19506524222805716, + "grad_norm": 2.1286463737487793, + "learning_rate": 1.9761893295555327e-05, + "loss": 3.1872, + "step": 2433 + }, + { + "epoch": 0.19514541700907978, + "grad_norm": 2.064481735229492, + "learning_rate": 1.9761611553810914e-05, + "loss": 3.4303, + "step": 2434 + }, + { + "epoch": 0.19522559179010243, + "grad_norm": 2.148008108139038, + "learning_rate": 1.9761329647489176e-05, + "loss": 3.2382, + "step": 2435 + }, + { + "epoch": 0.19530576657112506, + "grad_norm": 2.0719282627105713, + "learning_rate": 1.9761047576594858e-05, + "loss": 3.3881, + "step": 2436 + }, + { + "epoch": 0.19538594135214768, + "grad_norm": 2.2188198566436768, + "learning_rate": 1.9760765341132714e-05, + "loss": 3.3033, + "step": 2437 + }, + { + "epoch": 0.1954661161331703, + "grad_norm": 2.0825886726379395, + "learning_rate": 1.9760482941107512e-05, + "loss": 3.2657, + "step": 2438 + }, + { + "epoch": 0.19554629091419293, + "grad_norm": 2.267831563949585, + "learning_rate": 1.9760200376524e-05, + "loss": 3.2749, + "step": 2439 + }, + { + "epoch": 0.19562646569521558, + "grad_norm": 2.139766216278076, + "learning_rate": 1.9759917647386956e-05, + "loss": 3.3118, + "step": 2440 + }, + { + "epoch": 0.1957066404762382, + "grad_norm": 2.0885164737701416, + "learning_rate": 1.9759634753701137e-05, + "loss": 3.3925, + "step": 2441 + }, + { + "epoch": 0.19578681525726083, + "grad_norm": 1.9819320440292358, + "learning_rate": 1.9759351695471318e-05, + "loss": 3.2882, + "step": 2442 + }, + { + "epoch": 0.19586699003828345, + "grad_norm": 2.074800729751587, + "learning_rate": 1.975906847270227e-05, + "loss": 3.2009, + "step": 2443 + }, + { + "epoch": 0.19594716481930607, + "grad_norm": 2.04768443107605, + "learning_rate": 1.9758785085398763e-05, + "loss": 3.1533, + "step": 2444 + }, + { + "epoch": 0.19602733960032873, + "grad_norm": 2.1837527751922607, + "learning_rate": 1.9758501533565583e-05, + "loss": 3.253, + "step": 2445 + }, + { + "epoch": 0.19610751438135135, + "grad_norm": 2.1604151725769043, + "learning_rate": 1.9758217817207504e-05, + "loss": 3.3302, + "step": 2446 + }, + { + "epoch": 0.19618768916237397, + "grad_norm": 2.1781678199768066, + "learning_rate": 1.975793393632931e-05, + "loss": 3.3622, + "step": 2447 + }, + { + "epoch": 0.1962678639433966, + "grad_norm": 2.1198770999908447, + "learning_rate": 1.975764989093579e-05, + "loss": 3.2844, + "step": 2448 + }, + { + "epoch": 0.19634803872441922, + "grad_norm": 2.0578572750091553, + "learning_rate": 1.9757365681031734e-05, + "loss": 3.3189, + "step": 2449 + }, + { + "epoch": 0.19642821350544187, + "grad_norm": 2.0881145000457764, + "learning_rate": 1.975708130662193e-05, + "loss": 3.2042, + "step": 2450 + }, + { + "epoch": 0.1965083882864645, + "grad_norm": 2.1608352661132812, + "learning_rate": 1.9756796767711174e-05, + "loss": 3.2878, + "step": 2451 + }, + { + "epoch": 0.19658856306748712, + "grad_norm": 2.18009877204895, + "learning_rate": 1.9756512064304264e-05, + "loss": 3.2295, + "step": 2452 + }, + { + "epoch": 0.19666873784850974, + "grad_norm": 2.0347297191619873, + "learning_rate": 1.9756227196405998e-05, + "loss": 3.2719, + "step": 2453 + }, + { + "epoch": 0.19674891262953237, + "grad_norm": 2.1383602619171143, + "learning_rate": 1.9755942164021178e-05, + "loss": 3.296, + "step": 2454 + }, + { + "epoch": 0.19682908741055502, + "grad_norm": 2.1948964595794678, + "learning_rate": 1.9755656967154614e-05, + "loss": 3.2031, + "step": 2455 + }, + { + "epoch": 0.19690926219157764, + "grad_norm": 2.0594606399536133, + "learning_rate": 1.9755371605811115e-05, + "loss": 3.3864, + "step": 2456 + }, + { + "epoch": 0.19698943697260027, + "grad_norm": 2.154031276702881, + "learning_rate": 1.9755086079995484e-05, + "loss": 3.3571, + "step": 2457 + }, + { + "epoch": 0.1970696117536229, + "grad_norm": 2.682865858078003, + "learning_rate": 1.975480038971254e-05, + "loss": 3.3448, + "step": 2458 + }, + { + "epoch": 0.19714978653464552, + "grad_norm": 2.186300277709961, + "learning_rate": 1.9754514534967097e-05, + "loss": 3.2171, + "step": 2459 + }, + { + "epoch": 0.19722996131566817, + "grad_norm": 2.0058562755584717, + "learning_rate": 1.9754228515763984e-05, + "loss": 3.3113, + "step": 2460 + }, + { + "epoch": 0.1973101360966908, + "grad_norm": 2.230923652648926, + "learning_rate": 1.975394233210801e-05, + "loss": 3.3589, + "step": 2461 + }, + { + "epoch": 0.19739031087771342, + "grad_norm": 2.3656208515167236, + "learning_rate": 1.9753655984004006e-05, + "loss": 3.1522, + "step": 2462 + }, + { + "epoch": 0.19747048565873604, + "grad_norm": 2.045874834060669, + "learning_rate": 1.9753369471456802e-05, + "loss": 3.2294, + "step": 2463 + }, + { + "epoch": 0.19755066043975866, + "grad_norm": 2.134373664855957, + "learning_rate": 1.9753082794471225e-05, + "loss": 3.2429, + "step": 2464 + }, + { + "epoch": 0.19763083522078131, + "grad_norm": 2.4980125427246094, + "learning_rate": 1.9752795953052112e-05, + "loss": 3.4943, + "step": 2465 + }, + { + "epoch": 0.19771101000180394, + "grad_norm": 2.291318893432617, + "learning_rate": 1.9752508947204293e-05, + "loss": 3.3757, + "step": 2466 + }, + { + "epoch": 0.19779118478282656, + "grad_norm": 2.143497943878174, + "learning_rate": 1.9752221776932614e-05, + "loss": 3.3368, + "step": 2467 + }, + { + "epoch": 0.1978713595638492, + "grad_norm": 2.1891262531280518, + "learning_rate": 1.975193444224191e-05, + "loss": 3.1997, + "step": 2468 + }, + { + "epoch": 0.1979515343448718, + "grad_norm": 2.2555320262908936, + "learning_rate": 1.9751646943137026e-05, + "loss": 3.285, + "step": 2469 + }, + { + "epoch": 0.19803170912589446, + "grad_norm": 2.1074960231781006, + "learning_rate": 1.9751359279622818e-05, + "loss": 3.3633, + "step": 2470 + }, + { + "epoch": 0.19811188390691709, + "grad_norm": 2.0947470664978027, + "learning_rate": 1.9751071451704123e-05, + "loss": 3.3579, + "step": 2471 + }, + { + "epoch": 0.1981920586879397, + "grad_norm": 2.088322401046753, + "learning_rate": 1.9750783459385802e-05, + "loss": 3.3782, + "step": 2472 + }, + { + "epoch": 0.19827223346896233, + "grad_norm": 1.018718957901001, + "learning_rate": 1.975049530267271e-05, + "loss": 2.549, + "step": 2473 + }, + { + "epoch": 0.19835240824998496, + "grad_norm": 2.1588118076324463, + "learning_rate": 1.97502069815697e-05, + "loss": 3.3472, + "step": 2474 + }, + { + "epoch": 0.1984325830310076, + "grad_norm": 2.1948909759521484, + "learning_rate": 1.9749918496081642e-05, + "loss": 3.2208, + "step": 2475 + }, + { + "epoch": 0.19851275781203023, + "grad_norm": 0.988606870174408, + "learning_rate": 1.974962984621339e-05, + "loss": 2.5559, + "step": 2476 + }, + { + "epoch": 0.19859293259305286, + "grad_norm": 2.2820017337799072, + "learning_rate": 1.9749341031969815e-05, + "loss": 3.34, + "step": 2477 + }, + { + "epoch": 0.19867310737407548, + "grad_norm": 2.285501480102539, + "learning_rate": 1.9749052053355787e-05, + "loss": 3.2323, + "step": 2478 + }, + { + "epoch": 0.1987532821550981, + "grad_norm": 1.9778395891189575, + "learning_rate": 1.9748762910376178e-05, + "loss": 3.1882, + "step": 2479 + }, + { + "epoch": 0.19883345693612076, + "grad_norm": 2.3913042545318604, + "learning_rate": 1.9748473603035862e-05, + "loss": 3.3479, + "step": 2480 + }, + { + "epoch": 0.19891363171714338, + "grad_norm": 2.5990443229675293, + "learning_rate": 1.9748184131339717e-05, + "loss": 3.3424, + "step": 2481 + }, + { + "epoch": 0.198993806498166, + "grad_norm": 2.138176202774048, + "learning_rate": 1.9747894495292624e-05, + "loss": 3.2709, + "step": 2482 + }, + { + "epoch": 0.19907398127918863, + "grad_norm": 1.9952454566955566, + "learning_rate": 1.974760469489946e-05, + "loss": 3.2729, + "step": 2483 + }, + { + "epoch": 0.19915415606021125, + "grad_norm": 2.125485897064209, + "learning_rate": 1.9747314730165123e-05, + "loss": 3.313, + "step": 2484 + }, + { + "epoch": 0.1992343308412339, + "grad_norm": 2.019573926925659, + "learning_rate": 1.974702460109449e-05, + "loss": 3.3307, + "step": 2485 + }, + { + "epoch": 0.19931450562225653, + "grad_norm": 2.1667823791503906, + "learning_rate": 1.9746734307692458e-05, + "loss": 3.2781, + "step": 2486 + }, + { + "epoch": 0.19939468040327915, + "grad_norm": 2.116018533706665, + "learning_rate": 1.9746443849963924e-05, + "loss": 3.2922, + "step": 2487 + }, + { + "epoch": 0.19947485518430177, + "grad_norm": 2.242607831954956, + "learning_rate": 1.974615322791378e-05, + "loss": 3.3324, + "step": 2488 + }, + { + "epoch": 0.1995550299653244, + "grad_norm": 2.4382708072662354, + "learning_rate": 1.9745862441546928e-05, + "loss": 3.2931, + "step": 2489 + }, + { + "epoch": 0.19963520474634705, + "grad_norm": 2.103973627090454, + "learning_rate": 1.9745571490868268e-05, + "loss": 3.2471, + "step": 2490 + }, + { + "epoch": 0.19971537952736967, + "grad_norm": 2.2904856204986572, + "learning_rate": 1.974528037588271e-05, + "loss": 3.3148, + "step": 2491 + }, + { + "epoch": 0.1997955543083923, + "grad_norm": 2.087928295135498, + "learning_rate": 1.9744989096595158e-05, + "loss": 3.2618, + "step": 2492 + }, + { + "epoch": 0.19987572908941492, + "grad_norm": 2.0829193592071533, + "learning_rate": 1.9744697653010526e-05, + "loss": 3.3147, + "step": 2493 + }, + { + "epoch": 0.19995590387043755, + "grad_norm": 2.1945457458496094, + "learning_rate": 1.9744406045133724e-05, + "loss": 3.2423, + "step": 2494 + }, + { + "epoch": 0.20003607865146017, + "grad_norm": 2.068563938140869, + "learning_rate": 1.9744114272969667e-05, + "loss": 3.3291, + "step": 2495 + }, + { + "epoch": 0.20011625343248282, + "grad_norm": 2.140835762023926, + "learning_rate": 1.9743822336523282e-05, + "loss": 3.2605, + "step": 2496 + }, + { + "epoch": 0.20019642821350545, + "grad_norm": 2.1885986328125, + "learning_rate": 1.9743530235799485e-05, + "loss": 3.2842, + "step": 2497 + }, + { + "epoch": 0.20027660299452807, + "grad_norm": 1.9866641759872437, + "learning_rate": 1.9743237970803203e-05, + "loss": 3.3025, + "step": 2498 + }, + { + "epoch": 0.2003567777755507, + "grad_norm": 2.1637380123138428, + "learning_rate": 1.974294554153936e-05, + "loss": 3.2444, + "step": 2499 + }, + { + "epoch": 0.20043695255657332, + "grad_norm": 2.0564210414886475, + "learning_rate": 1.9742652948012893e-05, + "loss": 3.1801, + "step": 2500 + }, + { + "epoch": 0.20051712733759597, + "grad_norm": 2.171795606613159, + "learning_rate": 1.9742360190228733e-05, + "loss": 3.2931, + "step": 2501 + }, + { + "epoch": 0.2005973021186186, + "grad_norm": 2.258796453475952, + "learning_rate": 1.9742067268191805e-05, + "loss": 3.4192, + "step": 2502 + }, + { + "epoch": 0.20067747689964122, + "grad_norm": 2.2292332649230957, + "learning_rate": 1.9741774181907065e-05, + "loss": 3.3, + "step": 2503 + }, + { + "epoch": 0.20075765168066384, + "grad_norm": 2.302875280380249, + "learning_rate": 1.974148093137944e-05, + "loss": 3.4223, + "step": 2504 + }, + { + "epoch": 0.20083782646168646, + "grad_norm": 1.9818254709243774, + "learning_rate": 1.9741187516613883e-05, + "loss": 3.2634, + "step": 2505 + }, + { + "epoch": 0.20091800124270912, + "grad_norm": 1.9136534929275513, + "learning_rate": 1.9740893937615335e-05, + "loss": 3.1658, + "step": 2506 + }, + { + "epoch": 0.20099817602373174, + "grad_norm": 2.0675175189971924, + "learning_rate": 1.9740600194388754e-05, + "loss": 3.2882, + "step": 2507 + }, + { + "epoch": 0.20107835080475436, + "grad_norm": 2.0240671634674072, + "learning_rate": 1.974030628693908e-05, + "loss": 3.2926, + "step": 2508 + }, + { + "epoch": 0.201158525585777, + "grad_norm": 2.081272602081299, + "learning_rate": 1.9740012215271276e-05, + "loss": 3.3424, + "step": 2509 + }, + { + "epoch": 0.2012387003667996, + "grad_norm": 2.2321650981903076, + "learning_rate": 1.9739717979390302e-05, + "loss": 3.399, + "step": 2510 + }, + { + "epoch": 0.20131887514782226, + "grad_norm": 2.015320301055908, + "learning_rate": 1.9739423579301112e-05, + "loss": 3.2875, + "step": 2511 + }, + { + "epoch": 0.2013990499288449, + "grad_norm": 2.123265266418457, + "learning_rate": 1.9739129015008673e-05, + "loss": 3.3564, + "step": 2512 + }, + { + "epoch": 0.2014792247098675, + "grad_norm": 2.0343592166900635, + "learning_rate": 1.9738834286517955e-05, + "loss": 3.3596, + "step": 2513 + }, + { + "epoch": 0.20155939949089013, + "grad_norm": 2.3654775619506836, + "learning_rate": 1.973853939383392e-05, + "loss": 3.2698, + "step": 2514 + }, + { + "epoch": 0.20163957427191276, + "grad_norm": 1.9797167778015137, + "learning_rate": 1.9738244336961542e-05, + "loss": 3.2495, + "step": 2515 + }, + { + "epoch": 0.2017197490529354, + "grad_norm": 2.134491205215454, + "learning_rate": 1.97379491159058e-05, + "loss": 3.2705, + "step": 2516 + }, + { + "epoch": 0.20179992383395803, + "grad_norm": 2.007840156555176, + "learning_rate": 1.9737653730671665e-05, + "loss": 3.3022, + "step": 2517 + }, + { + "epoch": 0.20188009861498066, + "grad_norm": 2.2820136547088623, + "learning_rate": 1.9737358181264123e-05, + "loss": 3.5083, + "step": 2518 + }, + { + "epoch": 0.20196027339600328, + "grad_norm": 2.460667610168457, + "learning_rate": 1.9737062467688148e-05, + "loss": 3.1985, + "step": 2519 + }, + { + "epoch": 0.2020404481770259, + "grad_norm": 2.162230968475342, + "learning_rate": 1.9736766589948738e-05, + "loss": 3.3284, + "step": 2520 + }, + { + "epoch": 0.20212062295804856, + "grad_norm": 2.1546454429626465, + "learning_rate": 1.973647054805087e-05, + "loss": 3.2742, + "step": 2521 + }, + { + "epoch": 0.20220079773907118, + "grad_norm": 1.9966987371444702, + "learning_rate": 1.9736174341999542e-05, + "loss": 3.2776, + "step": 2522 + }, + { + "epoch": 0.2022809725200938, + "grad_norm": 2.097165107727051, + "learning_rate": 1.9735877971799747e-05, + "loss": 3.3209, + "step": 2523 + }, + { + "epoch": 0.20236114730111643, + "grad_norm": 2.3593108654022217, + "learning_rate": 1.973558143745648e-05, + "loss": 3.318, + "step": 2524 + }, + { + "epoch": 0.20244132208213905, + "grad_norm": 2.182675838470459, + "learning_rate": 1.9735284738974742e-05, + "loss": 3.3462, + "step": 2525 + }, + { + "epoch": 0.2025214968631617, + "grad_norm": 2.293081521987915, + "learning_rate": 1.9734987876359533e-05, + "loss": 3.3866, + "step": 2526 + }, + { + "epoch": 0.20260167164418433, + "grad_norm": 2.2966976165771484, + "learning_rate": 1.973469084961586e-05, + "loss": 3.3212, + "step": 2527 + }, + { + "epoch": 0.20268184642520695, + "grad_norm": 2.112844705581665, + "learning_rate": 1.973439365874873e-05, + "loss": 3.2533, + "step": 2528 + }, + { + "epoch": 0.20276202120622958, + "grad_norm": 2.2087178230285645, + "learning_rate": 1.9734096303763152e-05, + "loss": 3.3339, + "step": 2529 + }, + { + "epoch": 0.2028421959872522, + "grad_norm": 2.2967031002044678, + "learning_rate": 1.9733798784664145e-05, + "loss": 3.3296, + "step": 2530 + }, + { + "epoch": 0.20292237076827485, + "grad_norm": 1.9921377897262573, + "learning_rate": 1.9733501101456715e-05, + "loss": 3.2728, + "step": 2531 + }, + { + "epoch": 0.20300254554929748, + "grad_norm": 2.0655970573425293, + "learning_rate": 1.973320325414589e-05, + "loss": 3.3049, + "step": 2532 + }, + { + "epoch": 0.2030827203303201, + "grad_norm": 2.1048362255096436, + "learning_rate": 1.973290524273669e-05, + "loss": 3.3165, + "step": 2533 + }, + { + "epoch": 0.20316289511134272, + "grad_norm": 2.237518787384033, + "learning_rate": 1.9732607067234135e-05, + "loss": 3.2097, + "step": 2534 + }, + { + "epoch": 0.20324306989236535, + "grad_norm": 1.141886830329895, + "learning_rate": 1.973230872764326e-05, + "loss": 2.5706, + "step": 2535 + }, + { + "epoch": 0.203323244673388, + "grad_norm": 1.9733738899230957, + "learning_rate": 1.9732010223969085e-05, + "loss": 3.2842, + "step": 2536 + }, + { + "epoch": 0.20340341945441062, + "grad_norm": 2.5164074897766113, + "learning_rate": 1.973171155621665e-05, + "loss": 3.3555, + "step": 2537 + }, + { + "epoch": 0.20348359423543325, + "grad_norm": 2.225975275039673, + "learning_rate": 1.9731412724390988e-05, + "loss": 3.242, + "step": 2538 + }, + { + "epoch": 0.20356376901645587, + "grad_norm": 2.0053791999816895, + "learning_rate": 1.9731113728497134e-05, + "loss": 3.3037, + "step": 2539 + }, + { + "epoch": 0.2036439437974785, + "grad_norm": 2.2097065448760986, + "learning_rate": 1.9730814568540135e-05, + "loss": 3.3196, + "step": 2540 + }, + { + "epoch": 0.20372411857850115, + "grad_norm": 2.438404083251953, + "learning_rate": 1.9730515244525027e-05, + "loss": 3.2784, + "step": 2541 + }, + { + "epoch": 0.20380429335952377, + "grad_norm": 2.1112232208251953, + "learning_rate": 1.973021575645687e-05, + "loss": 3.2935, + "step": 2542 + }, + { + "epoch": 0.2038844681405464, + "grad_norm": 1.931225061416626, + "learning_rate": 1.9729916104340696e-05, + "loss": 3.2603, + "step": 2543 + }, + { + "epoch": 0.20396464292156902, + "grad_norm": 2.0777759552001953, + "learning_rate": 1.972961628818157e-05, + "loss": 3.214, + "step": 2544 + }, + { + "epoch": 0.20404481770259164, + "grad_norm": 1.0611138343811035, + "learning_rate": 1.9729316307984542e-05, + "loss": 2.5468, + "step": 2545 + }, + { + "epoch": 0.2041249924836143, + "grad_norm": 2.558842658996582, + "learning_rate": 1.972901616375467e-05, + "loss": 3.2611, + "step": 2546 + }, + { + "epoch": 0.20420516726463692, + "grad_norm": 2.034534454345703, + "learning_rate": 1.9728715855497013e-05, + "loss": 3.3754, + "step": 2547 + }, + { + "epoch": 0.20428534204565954, + "grad_norm": 1.9832406044006348, + "learning_rate": 1.9728415383216638e-05, + "loss": 3.1482, + "step": 2548 + }, + { + "epoch": 0.20436551682668216, + "grad_norm": 2.3042867183685303, + "learning_rate": 1.9728114746918606e-05, + "loss": 3.31, + "step": 2549 + }, + { + "epoch": 0.2044456916077048, + "grad_norm": 2.2519195079803467, + "learning_rate": 1.972781394660799e-05, + "loss": 3.2855, + "step": 2550 + }, + { + "epoch": 0.2045258663887274, + "grad_norm": 1.971629023551941, + "learning_rate": 1.972751298228986e-05, + "loss": 3.1416, + "step": 2551 + }, + { + "epoch": 0.20460604116975006, + "grad_norm": 2.095221996307373, + "learning_rate": 1.9727211853969286e-05, + "loss": 3.3306, + "step": 2552 + }, + { + "epoch": 0.2046862159507727, + "grad_norm": 1.1579982042312622, + "learning_rate": 1.972691056165135e-05, + "loss": 2.5192, + "step": 2553 + }, + { + "epoch": 0.2047663907317953, + "grad_norm": 2.001316785812378, + "learning_rate": 1.972660910534113e-05, + "loss": 3.3115, + "step": 2554 + }, + { + "epoch": 0.20484656551281794, + "grad_norm": 2.2214512825012207, + "learning_rate": 1.9726307485043707e-05, + "loss": 3.2402, + "step": 2555 + }, + { + "epoch": 0.20492674029384056, + "grad_norm": 2.120171546936035, + "learning_rate": 1.9726005700764168e-05, + "loss": 3.3292, + "step": 2556 + }, + { + "epoch": 0.2050069150748632, + "grad_norm": 2.136404275894165, + "learning_rate": 1.9725703752507602e-05, + "loss": 3.2391, + "step": 2557 + }, + { + "epoch": 0.20508708985588583, + "grad_norm": 2.024425506591797, + "learning_rate": 1.97254016402791e-05, + "loss": 3.4052, + "step": 2558 + }, + { + "epoch": 0.20516726463690846, + "grad_norm": 2.1096203327178955, + "learning_rate": 1.9725099364083752e-05, + "loss": 3.3262, + "step": 2559 + }, + { + "epoch": 0.20524743941793108, + "grad_norm": 2.085055112838745, + "learning_rate": 1.9724796923926656e-05, + "loss": 3.2549, + "step": 2560 + }, + { + "epoch": 0.2053276141989537, + "grad_norm": 1.2128077745437622, + "learning_rate": 1.972449431981291e-05, + "loss": 2.5646, + "step": 2561 + }, + { + "epoch": 0.20540778897997636, + "grad_norm": 2.0312442779541016, + "learning_rate": 1.972419155174762e-05, + "loss": 3.2749, + "step": 2562 + }, + { + "epoch": 0.20548796376099898, + "grad_norm": 2.1263883113861084, + "learning_rate": 1.9723888619735884e-05, + "loss": 3.2408, + "step": 2563 + }, + { + "epoch": 0.2055681385420216, + "grad_norm": 0.991508424282074, + "learning_rate": 1.9723585523782814e-05, + "loss": 2.5251, + "step": 2564 + }, + { + "epoch": 0.20564831332304423, + "grad_norm": 2.267756462097168, + "learning_rate": 1.972328226389352e-05, + "loss": 3.3439, + "step": 2565 + }, + { + "epoch": 0.20572848810406685, + "grad_norm": 2.227605104446411, + "learning_rate": 1.972297884007312e-05, + "loss": 3.1707, + "step": 2566 + }, + { + "epoch": 0.2058086628850895, + "grad_norm": 2.182208299636841, + "learning_rate": 1.9722675252326712e-05, + "loss": 3.2347, + "step": 2567 + }, + { + "epoch": 0.20588883766611213, + "grad_norm": 2.2782957553863525, + "learning_rate": 1.972237150065943e-05, + "loss": 3.3055, + "step": 2568 + }, + { + "epoch": 0.20596901244713475, + "grad_norm": 2.3213155269622803, + "learning_rate": 1.9722067585076396e-05, + "loss": 3.2563, + "step": 2569 + }, + { + "epoch": 0.20604918722815738, + "grad_norm": 2.0648255348205566, + "learning_rate": 1.972176350558272e-05, + "loss": 3.4089, + "step": 2570 + }, + { + "epoch": 0.20612936200918, + "grad_norm": 2.14058256149292, + "learning_rate": 1.9721459262183543e-05, + "loss": 3.2676, + "step": 2571 + }, + { + "epoch": 0.20620953679020265, + "grad_norm": 2.190009832382202, + "learning_rate": 1.972115485488399e-05, + "loss": 3.2764, + "step": 2572 + }, + { + "epoch": 0.20628971157122528, + "grad_norm": 2.0250840187072754, + "learning_rate": 1.9720850283689192e-05, + "loss": 3.2057, + "step": 2573 + }, + { + "epoch": 0.2063698863522479, + "grad_norm": 1.1330113410949707, + "learning_rate": 1.9720545548604282e-05, + "loss": 2.5708, + "step": 2574 + }, + { + "epoch": 0.20645006113327052, + "grad_norm": 2.0048303604125977, + "learning_rate": 1.97202406496344e-05, + "loss": 3.2188, + "step": 2575 + }, + { + "epoch": 0.20653023591429315, + "grad_norm": 2.1817448139190674, + "learning_rate": 1.971993558678469e-05, + "loss": 3.3278, + "step": 2576 + }, + { + "epoch": 0.2066104106953158, + "grad_norm": 0.9996766448020935, + "learning_rate": 1.9719630360060288e-05, + "loss": 2.55, + "step": 2577 + }, + { + "epoch": 0.20669058547633842, + "grad_norm": 2.2083239555358887, + "learning_rate": 1.9719324969466344e-05, + "loss": 3.3665, + "step": 2578 + }, + { + "epoch": 0.20677076025736105, + "grad_norm": 2.0714123249053955, + "learning_rate": 1.9719019415008005e-05, + "loss": 3.3182, + "step": 2579 + }, + { + "epoch": 0.20685093503838367, + "grad_norm": 2.080324411392212, + "learning_rate": 1.9718713696690427e-05, + "loss": 3.2261, + "step": 2580 + }, + { + "epoch": 0.2069311098194063, + "grad_norm": 2.261531352996826, + "learning_rate": 1.971840781451876e-05, + "loss": 3.2986, + "step": 2581 + }, + { + "epoch": 0.20701128460042895, + "grad_norm": 2.081172227859497, + "learning_rate": 1.9718101768498163e-05, + "loss": 3.2899, + "step": 2582 + }, + { + "epoch": 0.20709145938145157, + "grad_norm": 2.1162009239196777, + "learning_rate": 1.9717795558633793e-05, + "loss": 3.3557, + "step": 2583 + }, + { + "epoch": 0.2071716341624742, + "grad_norm": 1.9570914506912231, + "learning_rate": 1.971748918493082e-05, + "loss": 3.2543, + "step": 2584 + }, + { + "epoch": 0.20725180894349682, + "grad_norm": 2.087233781814575, + "learning_rate": 1.97171826473944e-05, + "loss": 3.2899, + "step": 2585 + }, + { + "epoch": 0.20733198372451944, + "grad_norm": 2.135805606842041, + "learning_rate": 1.9716875946029702e-05, + "loss": 3.4044, + "step": 2586 + }, + { + "epoch": 0.2074121585055421, + "grad_norm": 2.09649658203125, + "learning_rate": 1.9716569080841904e-05, + "loss": 3.2515, + "step": 2587 + }, + { + "epoch": 0.20749233328656472, + "grad_norm": 2.5140702724456787, + "learning_rate": 1.971626205183618e-05, + "loss": 3.2115, + "step": 2588 + }, + { + "epoch": 0.20757250806758734, + "grad_norm": 2.137667655944824, + "learning_rate": 1.9715954859017698e-05, + "loss": 3.281, + "step": 2589 + }, + { + "epoch": 0.20765268284860997, + "grad_norm": 1.986762523651123, + "learning_rate": 1.971564750239164e-05, + "loss": 3.3287, + "step": 2590 + }, + { + "epoch": 0.2077328576296326, + "grad_norm": 2.2992124557495117, + "learning_rate": 1.971533998196319e-05, + "loss": 3.1921, + "step": 2591 + }, + { + "epoch": 0.20781303241065524, + "grad_norm": 2.175947904586792, + "learning_rate": 1.9715032297737533e-05, + "loss": 3.2953, + "step": 2592 + }, + { + "epoch": 0.20789320719167786, + "grad_norm": 2.240165948867798, + "learning_rate": 1.9714724449719856e-05, + "loss": 3.4259, + "step": 2593 + }, + { + "epoch": 0.2079733819727005, + "grad_norm": 2.2440578937530518, + "learning_rate": 1.9714416437915345e-05, + "loss": 3.1711, + "step": 2594 + }, + { + "epoch": 0.2080535567537231, + "grad_norm": 2.1044929027557373, + "learning_rate": 1.97141082623292e-05, + "loss": 3.2968, + "step": 2595 + }, + { + "epoch": 0.20813373153474574, + "grad_norm": 2.1674728393554688, + "learning_rate": 1.9713799922966613e-05, + "loss": 3.2594, + "step": 2596 + }, + { + "epoch": 0.2082139063157684, + "grad_norm": 2.14184308052063, + "learning_rate": 1.971349141983278e-05, + "loss": 3.3168, + "step": 2597 + }, + { + "epoch": 0.208294081096791, + "grad_norm": 2.162769079208374, + "learning_rate": 1.9713182752932907e-05, + "loss": 3.3613, + "step": 2598 + }, + { + "epoch": 0.20837425587781364, + "grad_norm": 2.2401371002197266, + "learning_rate": 1.9712873922272196e-05, + "loss": 3.2815, + "step": 2599 + }, + { + "epoch": 0.20845443065883626, + "grad_norm": 2.1278438568115234, + "learning_rate": 1.9712564927855852e-05, + "loss": 3.3044, + "step": 2600 + }, + { + "epoch": 0.20853460543985888, + "grad_norm": 2.107357978820801, + "learning_rate": 1.971225576968909e-05, + "loss": 3.3879, + "step": 2601 + }, + { + "epoch": 0.20861478022088153, + "grad_norm": 2.155402898788452, + "learning_rate": 1.9711946447777116e-05, + "loss": 3.2974, + "step": 2602 + }, + { + "epoch": 0.20869495500190416, + "grad_norm": 2.259550094604492, + "learning_rate": 1.971163696212515e-05, + "loss": 3.3073, + "step": 2603 + }, + { + "epoch": 0.20877512978292678, + "grad_norm": 2.023064613342285, + "learning_rate": 1.9711327312738404e-05, + "loss": 3.2603, + "step": 2604 + }, + { + "epoch": 0.2088553045639494, + "grad_norm": 2.015432119369507, + "learning_rate": 1.9711017499622104e-05, + "loss": 3.3783, + "step": 2605 + }, + { + "epoch": 0.20893547934497203, + "grad_norm": 2.1372828483581543, + "learning_rate": 1.971070752278147e-05, + "loss": 3.2874, + "step": 2606 + }, + { + "epoch": 0.20901565412599468, + "grad_norm": 2.1991846561431885, + "learning_rate": 1.9710397382221732e-05, + "loss": 3.2921, + "step": 2607 + }, + { + "epoch": 0.2090958289070173, + "grad_norm": 1.999212622642517, + "learning_rate": 1.9710087077948116e-05, + "loss": 3.4656, + "step": 2608 + }, + { + "epoch": 0.20917600368803993, + "grad_norm": 2.004061698913574, + "learning_rate": 1.9709776609965857e-05, + "loss": 3.3107, + "step": 2609 + }, + { + "epoch": 0.20925617846906255, + "grad_norm": 2.12135910987854, + "learning_rate": 1.970946597828018e-05, + "loss": 3.3475, + "step": 2610 + }, + { + "epoch": 0.20933635325008518, + "grad_norm": 1.99836266040802, + "learning_rate": 1.9709155182896332e-05, + "loss": 3.3134, + "step": 2611 + }, + { + "epoch": 0.2094165280311078, + "grad_norm": 2.056757688522339, + "learning_rate": 1.9708844223819546e-05, + "loss": 3.3275, + "step": 2612 + }, + { + "epoch": 0.20949670281213045, + "grad_norm": 2.107579469680786, + "learning_rate": 1.9708533101055072e-05, + "loss": 3.2086, + "step": 2613 + }, + { + "epoch": 0.20957687759315308, + "grad_norm": 1.9897958040237427, + "learning_rate": 1.9708221814608153e-05, + "loss": 3.3212, + "step": 2614 + }, + { + "epoch": 0.2096570523741757, + "grad_norm": 1.1335046291351318, + "learning_rate": 1.9707910364484032e-05, + "loss": 2.5581, + "step": 2615 + }, + { + "epoch": 0.20973722715519832, + "grad_norm": 2.2264606952667236, + "learning_rate": 1.9707598750687966e-05, + "loss": 3.2629, + "step": 2616 + }, + { + "epoch": 0.20981740193622095, + "grad_norm": 1.9538463354110718, + "learning_rate": 1.9707286973225205e-05, + "loss": 3.3021, + "step": 2617 + }, + { + "epoch": 0.2098975767172436, + "grad_norm": 0.9583985805511475, + "learning_rate": 1.9706975032101008e-05, + "loss": 2.5736, + "step": 2618 + }, + { + "epoch": 0.20997775149826622, + "grad_norm": 2.0799336433410645, + "learning_rate": 1.9706662927320634e-05, + "loss": 3.3116, + "step": 2619 + }, + { + "epoch": 0.21005792627928885, + "grad_norm": 2.09220027923584, + "learning_rate": 1.970635065888934e-05, + "loss": 3.2748, + "step": 2620 + }, + { + "epoch": 0.21013810106031147, + "grad_norm": 1.9904522895812988, + "learning_rate": 1.97060382268124e-05, + "loss": 3.3936, + "step": 2621 + }, + { + "epoch": 0.2102182758413341, + "grad_norm": 2.1310317516326904, + "learning_rate": 1.970572563109507e-05, + "loss": 3.2971, + "step": 2622 + }, + { + "epoch": 0.21029845062235675, + "grad_norm": 1.0656049251556396, + "learning_rate": 1.9705412871742632e-05, + "loss": 2.5397, + "step": 2623 + }, + { + "epoch": 0.21037862540337937, + "grad_norm": 2.0154359340667725, + "learning_rate": 1.970509994876035e-05, + "loss": 3.2533, + "step": 2624 + }, + { + "epoch": 0.210458800184402, + "grad_norm": 0.9729559421539307, + "learning_rate": 1.9704786862153505e-05, + "loss": 2.5355, + "step": 2625 + }, + { + "epoch": 0.21053897496542462, + "grad_norm": 2.1058433055877686, + "learning_rate": 1.9704473611927372e-05, + "loss": 3.2635, + "step": 2626 + }, + { + "epoch": 0.21061914974644724, + "grad_norm": 2.0428974628448486, + "learning_rate": 1.9704160198087234e-05, + "loss": 3.4769, + "step": 2627 + }, + { + "epoch": 0.2106993245274699, + "grad_norm": 2.1210267543792725, + "learning_rate": 1.970384662063838e-05, + "loss": 3.2235, + "step": 2628 + }, + { + "epoch": 0.21077949930849252, + "grad_norm": 2.224655866622925, + "learning_rate": 1.9703532879586083e-05, + "loss": 3.2908, + "step": 2629 + }, + { + "epoch": 0.21085967408951514, + "grad_norm": 2.0438194274902344, + "learning_rate": 1.9703218974935648e-05, + "loss": 3.2468, + "step": 2630 + }, + { + "epoch": 0.21093984887053777, + "grad_norm": 1.1558232307434082, + "learning_rate": 1.9702904906692363e-05, + "loss": 2.5643, + "step": 2631 + }, + { + "epoch": 0.2110200236515604, + "grad_norm": 2.2528319358825684, + "learning_rate": 1.9702590674861516e-05, + "loss": 3.3171, + "step": 2632 + }, + { + "epoch": 0.21110019843258304, + "grad_norm": 2.1074059009552, + "learning_rate": 1.970227627944841e-05, + "loss": 3.3345, + "step": 2633 + }, + { + "epoch": 0.21118037321360567, + "grad_norm": 2.0773725509643555, + "learning_rate": 1.9701961720458346e-05, + "loss": 3.337, + "step": 2634 + }, + { + "epoch": 0.2112605479946283, + "grad_norm": 2.1457881927490234, + "learning_rate": 1.9701646997896627e-05, + "loss": 3.3668, + "step": 2635 + }, + { + "epoch": 0.2113407227756509, + "grad_norm": 2.0614795684814453, + "learning_rate": 1.9701332111768564e-05, + "loss": 3.3585, + "step": 2636 + }, + { + "epoch": 0.21142089755667354, + "grad_norm": 0.9278883337974548, + "learning_rate": 1.9701017062079454e-05, + "loss": 2.5681, + "step": 2637 + }, + { + "epoch": 0.2115010723376962, + "grad_norm": 2.1507177352905273, + "learning_rate": 1.970070184883462e-05, + "loss": 3.3896, + "step": 2638 + }, + { + "epoch": 0.2115812471187188, + "grad_norm": 2.2337913513183594, + "learning_rate": 1.9700386472039366e-05, + "loss": 3.3643, + "step": 2639 + }, + { + "epoch": 0.21166142189974144, + "grad_norm": 2.07600998878479, + "learning_rate": 1.970007093169902e-05, + "loss": 3.3093, + "step": 2640 + }, + { + "epoch": 0.21174159668076406, + "grad_norm": 1.9798109531402588, + "learning_rate": 1.9699755227818894e-05, + "loss": 3.2264, + "step": 2641 + }, + { + "epoch": 0.21182177146178668, + "grad_norm": 2.3098855018615723, + "learning_rate": 1.9699439360404313e-05, + "loss": 3.3856, + "step": 2642 + }, + { + "epoch": 0.21190194624280934, + "grad_norm": 2.1368396282196045, + "learning_rate": 1.9699123329460602e-05, + "loss": 3.4039, + "step": 2643 + }, + { + "epoch": 0.21198212102383196, + "grad_norm": 2.07951283454895, + "learning_rate": 1.9698807134993095e-05, + "loss": 3.3954, + "step": 2644 + }, + { + "epoch": 0.21206229580485458, + "grad_norm": 1.982140064239502, + "learning_rate": 1.9698490777007114e-05, + "loss": 3.262, + "step": 2645 + }, + { + "epoch": 0.2121424705858772, + "grad_norm": 1.0331603288650513, + "learning_rate": 1.9698174255507996e-05, + "loss": 2.5304, + "step": 2646 + }, + { + "epoch": 0.21222264536689983, + "grad_norm": 1.0560097694396973, + "learning_rate": 1.969785757050108e-05, + "loss": 2.5511, + "step": 2647 + }, + { + "epoch": 0.21230282014792248, + "grad_norm": 2.165438175201416, + "learning_rate": 1.96975407219917e-05, + "loss": 3.3408, + "step": 2648 + }, + { + "epoch": 0.2123829949289451, + "grad_norm": 1.8918850421905518, + "learning_rate": 1.9697223709985206e-05, + "loss": 3.2782, + "step": 2649 + }, + { + "epoch": 0.21246316970996773, + "grad_norm": 2.1222147941589355, + "learning_rate": 1.9696906534486935e-05, + "loss": 3.3701, + "step": 2650 + }, + { + "epoch": 0.21254334449099035, + "grad_norm": 2.101278781890869, + "learning_rate": 1.9696589195502235e-05, + "loss": 3.2759, + "step": 2651 + }, + { + "epoch": 0.21262351927201298, + "grad_norm": 2.1651828289031982, + "learning_rate": 1.9696271693036463e-05, + "loss": 3.2795, + "step": 2652 + }, + { + "epoch": 0.21270369405303563, + "grad_norm": 2.1141445636749268, + "learning_rate": 1.9695954027094964e-05, + "loss": 3.2346, + "step": 2653 + }, + { + "epoch": 0.21278386883405825, + "grad_norm": 2.017155170440674, + "learning_rate": 1.96956361976831e-05, + "loss": 3.3299, + "step": 2654 + }, + { + "epoch": 0.21286404361508088, + "grad_norm": 2.1709654331207275, + "learning_rate": 1.9695318204806228e-05, + "loss": 3.2772, + "step": 2655 + }, + { + "epoch": 0.2129442183961035, + "grad_norm": 2.2067089080810547, + "learning_rate": 1.9695000048469706e-05, + "loss": 3.2493, + "step": 2656 + }, + { + "epoch": 0.21302439317712613, + "grad_norm": 0.9471723437309265, + "learning_rate": 1.9694681728678898e-05, + "loss": 2.5593, + "step": 2657 + }, + { + "epoch": 0.21310456795814878, + "grad_norm": 2.0700902938842773, + "learning_rate": 1.9694363245439175e-05, + "loss": 3.3263, + "step": 2658 + }, + { + "epoch": 0.2131847427391714, + "grad_norm": 2.0607237815856934, + "learning_rate": 1.9694044598755907e-05, + "loss": 3.2366, + "step": 2659 + }, + { + "epoch": 0.21326491752019403, + "grad_norm": 2.0170857906341553, + "learning_rate": 1.9693725788634456e-05, + "loss": 3.2227, + "step": 2660 + }, + { + "epoch": 0.21334509230121665, + "grad_norm": 2.377138376235962, + "learning_rate": 1.969340681508021e-05, + "loss": 3.355, + "step": 2661 + }, + { + "epoch": 0.21342526708223927, + "grad_norm": 1.9547232389450073, + "learning_rate": 1.9693087678098542e-05, + "loss": 3.3058, + "step": 2662 + }, + { + "epoch": 0.21350544186326192, + "grad_norm": 2.122328758239746, + "learning_rate": 1.9692768377694832e-05, + "loss": 3.286, + "step": 2663 + }, + { + "epoch": 0.21358561664428455, + "grad_norm": 1.9395252466201782, + "learning_rate": 1.969244891387446e-05, + "loss": 3.2841, + "step": 2664 + }, + { + "epoch": 0.21366579142530717, + "grad_norm": 2.148092031478882, + "learning_rate": 1.9692129286642817e-05, + "loss": 3.2623, + "step": 2665 + }, + { + "epoch": 0.2137459662063298, + "grad_norm": 2.013430118560791, + "learning_rate": 1.969180949600529e-05, + "loss": 3.2653, + "step": 2666 + }, + { + "epoch": 0.21382614098735242, + "grad_norm": 2.129920721054077, + "learning_rate": 1.969148954196727e-05, + "loss": 3.2623, + "step": 2667 + }, + { + "epoch": 0.21390631576837504, + "grad_norm": 1.9626632928848267, + "learning_rate": 1.9691169424534157e-05, + "loss": 3.2089, + "step": 2668 + }, + { + "epoch": 0.2139864905493977, + "grad_norm": 2.054158926010132, + "learning_rate": 1.9690849143711335e-05, + "loss": 3.2094, + "step": 2669 + }, + { + "epoch": 0.21406666533042032, + "grad_norm": 1.987445592880249, + "learning_rate": 1.9690528699504216e-05, + "loss": 3.2962, + "step": 2670 + }, + { + "epoch": 0.21414684011144294, + "grad_norm": 2.275787830352783, + "learning_rate": 1.9690208091918198e-05, + "loss": 3.2278, + "step": 2671 + }, + { + "epoch": 0.21422701489246557, + "grad_norm": 2.1742987632751465, + "learning_rate": 1.9689887320958687e-05, + "loss": 3.4205, + "step": 2672 + }, + { + "epoch": 0.2143071896734882, + "grad_norm": 2.0049729347229004, + "learning_rate": 1.968956638663109e-05, + "loss": 3.3523, + "step": 2673 + }, + { + "epoch": 0.21438736445451084, + "grad_norm": 2.0623602867126465, + "learning_rate": 1.968924528894082e-05, + "loss": 3.2781, + "step": 2674 + }, + { + "epoch": 0.21446753923553347, + "grad_norm": 2.271434783935547, + "learning_rate": 1.968892402789329e-05, + "loss": 3.3729, + "step": 2675 + }, + { + "epoch": 0.2145477140165561, + "grad_norm": 2.0219175815582275, + "learning_rate": 1.968860260349391e-05, + "loss": 3.3121, + "step": 2676 + }, + { + "epoch": 0.21462788879757871, + "grad_norm": 2.181121587753296, + "learning_rate": 1.968828101574811e-05, + "loss": 3.2282, + "step": 2677 + }, + { + "epoch": 0.21470806357860134, + "grad_norm": 2.1109423637390137, + "learning_rate": 1.9687959264661306e-05, + "loss": 3.4258, + "step": 2678 + }, + { + "epoch": 0.214788238359624, + "grad_norm": 2.473466396331787, + "learning_rate": 1.968763735023892e-05, + "loss": 3.2167, + "step": 2679 + }, + { + "epoch": 0.2148684131406466, + "grad_norm": 2.128265380859375, + "learning_rate": 1.9687315272486387e-05, + "loss": 3.4385, + "step": 2680 + }, + { + "epoch": 0.21494858792166924, + "grad_norm": 2.100177526473999, + "learning_rate": 1.968699303140913e-05, + "loss": 3.2496, + "step": 2681 + }, + { + "epoch": 0.21502876270269186, + "grad_norm": 1.083207130432129, + "learning_rate": 1.9686670627012585e-05, + "loss": 2.5427, + "step": 2682 + }, + { + "epoch": 0.21510893748371449, + "grad_norm": 2.2308189868927, + "learning_rate": 1.968634805930219e-05, + "loss": 3.3101, + "step": 2683 + }, + { + "epoch": 0.21518911226473714, + "grad_norm": 2.739793300628662, + "learning_rate": 1.9686025328283376e-05, + "loss": 3.1848, + "step": 2684 + }, + { + "epoch": 0.21526928704575976, + "grad_norm": 2.169254779815674, + "learning_rate": 1.9685702433961592e-05, + "loss": 3.3059, + "step": 2685 + }, + { + "epoch": 0.21534946182678238, + "grad_norm": 2.4406816959381104, + "learning_rate": 1.9685379376342273e-05, + "loss": 3.3383, + "step": 2686 + }, + { + "epoch": 0.215429636607805, + "grad_norm": 2.1990466117858887, + "learning_rate": 1.9685056155430876e-05, + "loss": 3.1665, + "step": 2687 + }, + { + "epoch": 0.21550981138882763, + "grad_norm": 2.235483169555664, + "learning_rate": 1.9684732771232845e-05, + "loss": 3.3116, + "step": 2688 + }, + { + "epoch": 0.21558998616985028, + "grad_norm": 2.343665361404419, + "learning_rate": 1.9684409223753635e-05, + "loss": 3.2443, + "step": 2689 + }, + { + "epoch": 0.2156701609508729, + "grad_norm": 2.303452968597412, + "learning_rate": 1.9684085512998693e-05, + "loss": 3.2886, + "step": 2690 + }, + { + "epoch": 0.21575033573189553, + "grad_norm": 2.0548367500305176, + "learning_rate": 1.9683761638973486e-05, + "loss": 3.3759, + "step": 2691 + }, + { + "epoch": 0.21583051051291816, + "grad_norm": 2.0278964042663574, + "learning_rate": 1.968343760168347e-05, + "loss": 3.2955, + "step": 2692 + }, + { + "epoch": 0.21591068529394078, + "grad_norm": 2.156745195388794, + "learning_rate": 1.968311340113411e-05, + "loss": 3.327, + "step": 2693 + }, + { + "epoch": 0.21599086007496343, + "grad_norm": 2.264991044998169, + "learning_rate": 1.968278903733087e-05, + "loss": 3.2487, + "step": 2694 + }, + { + "epoch": 0.21607103485598605, + "grad_norm": 2.524709463119507, + "learning_rate": 1.968246451027922e-05, + "loss": 3.2961, + "step": 2695 + }, + { + "epoch": 0.21615120963700868, + "grad_norm": 2.255248546600342, + "learning_rate": 1.9682139819984626e-05, + "loss": 3.3164, + "step": 2696 + }, + { + "epoch": 0.2162313844180313, + "grad_norm": 2.1502416133880615, + "learning_rate": 1.9681814966452572e-05, + "loss": 3.3037, + "step": 2697 + }, + { + "epoch": 0.21631155919905393, + "grad_norm": 2.5039219856262207, + "learning_rate": 1.9681489949688527e-05, + "loss": 3.2314, + "step": 2698 + }, + { + "epoch": 0.21639173398007658, + "grad_norm": 2.131526470184326, + "learning_rate": 1.9681164769697977e-05, + "loss": 3.2381, + "step": 2699 + }, + { + "epoch": 0.2164719087610992, + "grad_norm": 2.0803334712982178, + "learning_rate": 1.9680839426486398e-05, + "loss": 3.3031, + "step": 2700 + }, + { + "epoch": 0.21655208354212183, + "grad_norm": 2.2448298931121826, + "learning_rate": 1.9680513920059277e-05, + "loss": 3.3529, + "step": 2701 + }, + { + "epoch": 0.21663225832314445, + "grad_norm": 2.1045610904693604, + "learning_rate": 1.9680188250422105e-05, + "loss": 3.2791, + "step": 2702 + }, + { + "epoch": 0.21671243310416707, + "grad_norm": 2.1287083625793457, + "learning_rate": 1.9679862417580372e-05, + "loss": 3.2495, + "step": 2703 + }, + { + "epoch": 0.21679260788518973, + "grad_norm": 1.974147915840149, + "learning_rate": 1.967953642153957e-05, + "loss": 3.3141, + "step": 2704 + }, + { + "epoch": 0.21687278266621235, + "grad_norm": 2.1198177337646484, + "learning_rate": 1.9679210262305192e-05, + "loss": 3.3785, + "step": 2705 + }, + { + "epoch": 0.21695295744723497, + "grad_norm": 2.3529787063598633, + "learning_rate": 1.9678883939882743e-05, + "loss": 3.2544, + "step": 2706 + }, + { + "epoch": 0.2170331322282576, + "grad_norm": 2.48124361038208, + "learning_rate": 1.9678557454277723e-05, + "loss": 3.2554, + "step": 2707 + }, + { + "epoch": 0.21711330700928022, + "grad_norm": 2.2475600242614746, + "learning_rate": 1.9678230805495633e-05, + "loss": 3.4398, + "step": 2708 + }, + { + "epoch": 0.21719348179030287, + "grad_norm": 2.146423578262329, + "learning_rate": 1.9677903993541985e-05, + "loss": 3.2802, + "step": 2709 + }, + { + "epoch": 0.2172736565713255, + "grad_norm": 2.0918970108032227, + "learning_rate": 1.9677577018422282e-05, + "loss": 3.2871, + "step": 2710 + }, + { + "epoch": 0.21735383135234812, + "grad_norm": 2.030280590057373, + "learning_rate": 1.9677249880142045e-05, + "loss": 3.4113, + "step": 2711 + }, + { + "epoch": 0.21743400613337074, + "grad_norm": 2.2805535793304443, + "learning_rate": 1.967692257870679e-05, + "loss": 3.3918, + "step": 2712 + }, + { + "epoch": 0.21751418091439337, + "grad_norm": 2.313739061355591, + "learning_rate": 1.967659511412202e-05, + "loss": 3.264, + "step": 2713 + }, + { + "epoch": 0.21759435569541602, + "grad_norm": 2.2950148582458496, + "learning_rate": 1.9676267486393274e-05, + "loss": 3.2554, + "step": 2714 + }, + { + "epoch": 0.21767453047643864, + "grad_norm": 2.106208562850952, + "learning_rate": 1.9675939695526067e-05, + "loss": 3.2333, + "step": 2715 + }, + { + "epoch": 0.21775470525746127, + "grad_norm": 2.1766748428344727, + "learning_rate": 1.9675611741525926e-05, + "loss": 3.2548, + "step": 2716 + }, + { + "epoch": 0.2178348800384839, + "grad_norm": 2.107177734375, + "learning_rate": 1.9675283624398382e-05, + "loss": 3.3076, + "step": 2717 + }, + { + "epoch": 0.21791505481950652, + "grad_norm": 2.031538248062134, + "learning_rate": 1.9674955344148965e-05, + "loss": 3.2288, + "step": 2718 + }, + { + "epoch": 0.21799522960052917, + "grad_norm": 2.356053590774536, + "learning_rate": 1.9674626900783213e-05, + "loss": 3.2538, + "step": 2719 + }, + { + "epoch": 0.2180754043815518, + "grad_norm": 2.243401050567627, + "learning_rate": 1.9674298294306656e-05, + "loss": 3.2501, + "step": 2720 + }, + { + "epoch": 0.21815557916257441, + "grad_norm": 2.127511978149414, + "learning_rate": 1.9673969524724846e-05, + "loss": 3.4111, + "step": 2721 + }, + { + "epoch": 0.21823575394359704, + "grad_norm": 2.088106632232666, + "learning_rate": 1.9673640592043315e-05, + "loss": 3.2789, + "step": 2722 + }, + { + "epoch": 0.21831592872461966, + "grad_norm": 2.1051278114318848, + "learning_rate": 1.9673311496267612e-05, + "loss": 3.3102, + "step": 2723 + }, + { + "epoch": 0.21839610350564231, + "grad_norm": 2.076958179473877, + "learning_rate": 1.967298223740329e-05, + "loss": 3.1901, + "step": 2724 + }, + { + "epoch": 0.21847627828666494, + "grad_norm": 2.0255184173583984, + "learning_rate": 1.9672652815455895e-05, + "loss": 3.1649, + "step": 2725 + }, + { + "epoch": 0.21855645306768756, + "grad_norm": 1.9588623046875, + "learning_rate": 1.9672323230430984e-05, + "loss": 3.3088, + "step": 2726 + }, + { + "epoch": 0.21863662784871019, + "grad_norm": 2.1841654777526855, + "learning_rate": 1.9671993482334108e-05, + "loss": 3.3607, + "step": 2727 + }, + { + "epoch": 0.2187168026297328, + "grad_norm": 2.1193044185638428, + "learning_rate": 1.9671663571170832e-05, + "loss": 3.3296, + "step": 2728 + }, + { + "epoch": 0.21879697741075543, + "grad_norm": 2.206425666809082, + "learning_rate": 1.967133349694672e-05, + "loss": 3.3002, + "step": 2729 + }, + { + "epoch": 0.21887715219177808, + "grad_norm": 1.9684253931045532, + "learning_rate": 1.9671003259667332e-05, + "loss": 3.2815, + "step": 2730 + }, + { + "epoch": 0.2189573269728007, + "grad_norm": 2.562380313873291, + "learning_rate": 1.9670672859338237e-05, + "loss": 3.3283, + "step": 2731 + }, + { + "epoch": 0.21903750175382333, + "grad_norm": 1.9649481773376465, + "learning_rate": 1.9670342295965006e-05, + "loss": 3.2871, + "step": 2732 + }, + { + "epoch": 0.21911767653484596, + "grad_norm": 2.1684889793395996, + "learning_rate": 1.967001156955321e-05, + "loss": 3.2361, + "step": 2733 + }, + { + "epoch": 0.21919785131586858, + "grad_norm": 2.3150691986083984, + "learning_rate": 1.966968068010843e-05, + "loss": 3.3804, + "step": 2734 + }, + { + "epoch": 0.21927802609689123, + "grad_norm": 1.990041732788086, + "learning_rate": 1.9669349627636244e-05, + "loss": 3.2665, + "step": 2735 + }, + { + "epoch": 0.21935820087791386, + "grad_norm": 2.0187416076660156, + "learning_rate": 1.9669018412142224e-05, + "loss": 3.3517, + "step": 2736 + }, + { + "epoch": 0.21943837565893648, + "grad_norm": 1.1233552694320679, + "learning_rate": 1.9668687033631966e-05, + "loss": 2.5737, + "step": 2737 + }, + { + "epoch": 0.2195185504399591, + "grad_norm": 2.2058417797088623, + "learning_rate": 1.966835549211105e-05, + "loss": 3.2306, + "step": 2738 + }, + { + "epoch": 0.21959872522098173, + "grad_norm": 2.3618180751800537, + "learning_rate": 1.966802378758507e-05, + "loss": 3.288, + "step": 2739 + }, + { + "epoch": 0.21967890000200438, + "grad_norm": 2.2378814220428467, + "learning_rate": 1.966769192005962e-05, + "loss": 3.3949, + "step": 2740 + }, + { + "epoch": 0.219759074783027, + "grad_norm": 2.125837802886963, + "learning_rate": 1.966735988954028e-05, + "loss": 3.2983, + "step": 2741 + }, + { + "epoch": 0.21983924956404963, + "grad_norm": 2.083285331726074, + "learning_rate": 1.966702769603267e-05, + "loss": 3.3059, + "step": 2742 + }, + { + "epoch": 0.21991942434507225, + "grad_norm": 2.117121696472168, + "learning_rate": 1.9666695339542375e-05, + "loss": 3.345, + "step": 2743 + }, + { + "epoch": 0.21999959912609487, + "grad_norm": 2.5020172595977783, + "learning_rate": 1.9666362820075004e-05, + "loss": 3.2546, + "step": 2744 + }, + { + "epoch": 0.22007977390711753, + "grad_norm": 2.092958450317383, + "learning_rate": 1.9666030137636166e-05, + "loss": 3.2728, + "step": 2745 + }, + { + "epoch": 0.22015994868814015, + "grad_norm": 2.3924362659454346, + "learning_rate": 1.966569729223146e-05, + "loss": 3.3418, + "step": 2746 + }, + { + "epoch": 0.22024012346916277, + "grad_norm": 1.8578060865402222, + "learning_rate": 1.9665364283866512e-05, + "loss": 3.2904, + "step": 2747 + }, + { + "epoch": 0.2203202982501854, + "grad_norm": 2.123660087585449, + "learning_rate": 1.966503111254692e-05, + "loss": 3.4088, + "step": 2748 + }, + { + "epoch": 0.22040047303120802, + "grad_norm": 2.1381609439849854, + "learning_rate": 1.9664697778278317e-05, + "loss": 3.1948, + "step": 2749 + }, + { + "epoch": 0.22048064781223067, + "grad_norm": 2.0306954383850098, + "learning_rate": 1.9664364281066314e-05, + "loss": 3.1913, + "step": 2750 + }, + { + "epoch": 0.2205608225932533, + "grad_norm": 2.0507450103759766, + "learning_rate": 1.9664030620916535e-05, + "loss": 3.1935, + "step": 2751 + }, + { + "epoch": 0.22064099737427592, + "grad_norm": 2.209730625152588, + "learning_rate": 1.9663696797834602e-05, + "loss": 3.3026, + "step": 2752 + }, + { + "epoch": 0.22072117215529855, + "grad_norm": 2.1892151832580566, + "learning_rate": 1.9663362811826153e-05, + "loss": 3.3572, + "step": 2753 + }, + { + "epoch": 0.22080134693632117, + "grad_norm": 2.191781997680664, + "learning_rate": 1.966302866289681e-05, + "loss": 3.3464, + "step": 2754 + }, + { + "epoch": 0.22088152171734382, + "grad_norm": 2.138375759124756, + "learning_rate": 1.9662694351052206e-05, + "loss": 3.3746, + "step": 2755 + }, + { + "epoch": 0.22096169649836644, + "grad_norm": 2.252967357635498, + "learning_rate": 1.9662359876297984e-05, + "loss": 3.264, + "step": 2756 + }, + { + "epoch": 0.22104187127938907, + "grad_norm": 1.8419150114059448, + "learning_rate": 1.966202523863978e-05, + "loss": 3.2432, + "step": 2757 + }, + { + "epoch": 0.2211220460604117, + "grad_norm": 1.02731454372406, + "learning_rate": 1.9661690438083232e-05, + "loss": 2.5582, + "step": 2758 + }, + { + "epoch": 0.22120222084143432, + "grad_norm": 2.2978057861328125, + "learning_rate": 1.9661355474633992e-05, + "loss": 3.3403, + "step": 2759 + }, + { + "epoch": 0.22128239562245697, + "grad_norm": 2.1595680713653564, + "learning_rate": 1.9661020348297707e-05, + "loss": 3.2844, + "step": 2760 + }, + { + "epoch": 0.2213625704034796, + "grad_norm": 2.180727481842041, + "learning_rate": 1.9660685059080018e-05, + "loss": 3.3257, + "step": 2761 + }, + { + "epoch": 0.22144274518450222, + "grad_norm": 2.2859737873077393, + "learning_rate": 1.9660349606986588e-05, + "loss": 3.25, + "step": 2762 + }, + { + "epoch": 0.22152291996552484, + "grad_norm": 2.1427793502807617, + "learning_rate": 1.9660013992023064e-05, + "loss": 3.3024, + "step": 2763 + }, + { + "epoch": 0.22160309474654746, + "grad_norm": 1.9702917337417603, + "learning_rate": 1.9659678214195107e-05, + "loss": 3.2551, + "step": 2764 + }, + { + "epoch": 0.22168326952757011, + "grad_norm": 2.1473588943481445, + "learning_rate": 1.9659342273508387e-05, + "loss": 3.2303, + "step": 2765 + }, + { + "epoch": 0.22176344430859274, + "grad_norm": 2.0785722732543945, + "learning_rate": 1.9659006169968554e-05, + "loss": 3.3307, + "step": 2766 + }, + { + "epoch": 0.22184361908961536, + "grad_norm": 2.1176624298095703, + "learning_rate": 1.9658669903581284e-05, + "loss": 3.2819, + "step": 2767 + }, + { + "epoch": 0.221923793870638, + "grad_norm": 2.256169557571411, + "learning_rate": 1.965833347435224e-05, + "loss": 3.431, + "step": 2768 + }, + { + "epoch": 0.2220039686516606, + "grad_norm": 2.103431463241577, + "learning_rate": 1.9657996882287102e-05, + "loss": 3.2737, + "step": 2769 + }, + { + "epoch": 0.22208414343268326, + "grad_norm": 2.2219138145446777, + "learning_rate": 1.9657660127391543e-05, + "loss": 3.2305, + "step": 2770 + }, + { + "epoch": 0.22216431821370589, + "grad_norm": 2.0975611209869385, + "learning_rate": 1.9657323209671234e-05, + "loss": 3.3965, + "step": 2771 + }, + { + "epoch": 0.2222444929947285, + "grad_norm": 2.0235800743103027, + "learning_rate": 1.965698612913186e-05, + "loss": 3.2283, + "step": 2772 + }, + { + "epoch": 0.22232466777575113, + "grad_norm": 2.0301430225372314, + "learning_rate": 1.96566488857791e-05, + "loss": 3.2062, + "step": 2773 + }, + { + "epoch": 0.22240484255677376, + "grad_norm": 2.138112783432007, + "learning_rate": 1.9656311479618647e-05, + "loss": 3.3198, + "step": 2774 + }, + { + "epoch": 0.2224850173377964, + "grad_norm": 2.1128172874450684, + "learning_rate": 1.9655973910656183e-05, + "loss": 3.315, + "step": 2775 + }, + { + "epoch": 0.22256519211881903, + "grad_norm": 2.301407814025879, + "learning_rate": 1.9655636178897404e-05, + "loss": 3.3455, + "step": 2776 + }, + { + "epoch": 0.22264536689984166, + "grad_norm": 1.900573968887329, + "learning_rate": 1.9655298284348e-05, + "loss": 3.1985, + "step": 2777 + }, + { + "epoch": 0.22272554168086428, + "grad_norm": 2.10357403755188, + "learning_rate": 1.965496022701367e-05, + "loss": 3.3429, + "step": 2778 + }, + { + "epoch": 0.2228057164618869, + "grad_norm": 2.1885740756988525, + "learning_rate": 1.9654622006900117e-05, + "loss": 3.4822, + "step": 2779 + }, + { + "epoch": 0.22288589124290956, + "grad_norm": 2.158694267272949, + "learning_rate": 1.9654283624013034e-05, + "loss": 3.3763, + "step": 2780 + }, + { + "epoch": 0.22296606602393218, + "grad_norm": 2.1090786457061768, + "learning_rate": 1.965394507835814e-05, + "loss": 3.2994, + "step": 2781 + }, + { + "epoch": 0.2230462408049548, + "grad_norm": 2.066925525665283, + "learning_rate": 1.965360636994113e-05, + "loss": 3.3324, + "step": 2782 + }, + { + "epoch": 0.22312641558597743, + "grad_norm": 1.0002052783966064, + "learning_rate": 1.9653267498767714e-05, + "loss": 2.5618, + "step": 2783 + }, + { + "epoch": 0.22320659036700005, + "grad_norm": 1.9759465456008911, + "learning_rate": 1.9652928464843614e-05, + "loss": 3.3405, + "step": 2784 + }, + { + "epoch": 0.22328676514802268, + "grad_norm": 2.5823428630828857, + "learning_rate": 1.9652589268174543e-05, + "loss": 3.4158, + "step": 2785 + }, + { + "epoch": 0.22336693992904533, + "grad_norm": 1.9021153450012207, + "learning_rate": 1.9652249908766218e-05, + "loss": 3.3273, + "step": 2786 + }, + { + "epoch": 0.22344711471006795, + "grad_norm": 2.097907543182373, + "learning_rate": 1.965191038662436e-05, + "loss": 3.3278, + "step": 2787 + }, + { + "epoch": 0.22352728949109057, + "grad_norm": 2.4557971954345703, + "learning_rate": 1.9651570701754696e-05, + "loss": 3.3478, + "step": 2788 + }, + { + "epoch": 0.2236074642721132, + "grad_norm": 2.6010870933532715, + "learning_rate": 1.9651230854162953e-05, + "loss": 3.1848, + "step": 2789 + }, + { + "epoch": 0.22368763905313582, + "grad_norm": 1.918853521347046, + "learning_rate": 1.9650890843854857e-05, + "loss": 3.2161, + "step": 2790 + }, + { + "epoch": 0.22376781383415847, + "grad_norm": 2.0317471027374268, + "learning_rate": 1.9650550670836143e-05, + "loss": 3.174, + "step": 2791 + }, + { + "epoch": 0.2238479886151811, + "grad_norm": 2.2403366565704346, + "learning_rate": 1.9650210335112544e-05, + "loss": 3.2787, + "step": 2792 + }, + { + "epoch": 0.22392816339620372, + "grad_norm": 1.9977734088897705, + "learning_rate": 1.9649869836689802e-05, + "loss": 3.2184, + "step": 2793 + }, + { + "epoch": 0.22400833817722635, + "grad_norm": 2.161252498626709, + "learning_rate": 1.9649529175573655e-05, + "loss": 3.1805, + "step": 2794 + }, + { + "epoch": 0.22408851295824897, + "grad_norm": 2.0037059783935547, + "learning_rate": 1.964918835176985e-05, + "loss": 3.1567, + "step": 2795 + }, + { + "epoch": 0.22416868773927162, + "grad_norm": 1.9924983978271484, + "learning_rate": 1.9648847365284127e-05, + "loss": 3.349, + "step": 2796 + }, + { + "epoch": 0.22424886252029425, + "grad_norm": 2.0291435718536377, + "learning_rate": 1.964850621612224e-05, + "loss": 3.2424, + "step": 2797 + }, + { + "epoch": 0.22432903730131687, + "grad_norm": 1.9728279113769531, + "learning_rate": 1.9648164904289936e-05, + "loss": 3.2689, + "step": 2798 + }, + { + "epoch": 0.2244092120823395, + "grad_norm": 2.1027612686157227, + "learning_rate": 1.9647823429792972e-05, + "loss": 3.3547, + "step": 2799 + }, + { + "epoch": 0.22448938686336212, + "grad_norm": 1.9541969299316406, + "learning_rate": 1.9647481792637107e-05, + "loss": 3.2488, + "step": 2800 + }, + { + "epoch": 0.22456956164438477, + "grad_norm": 2.1657137870788574, + "learning_rate": 1.9647139992828097e-05, + "loss": 3.3224, + "step": 2801 + }, + { + "epoch": 0.2246497364254074, + "grad_norm": 0.9116299748420715, + "learning_rate": 1.964679803037171e-05, + "loss": 2.5388, + "step": 2802 + }, + { + "epoch": 0.22472991120643002, + "grad_norm": 2.0984556674957275, + "learning_rate": 1.9646455905273705e-05, + "loss": 3.2001, + "step": 2803 + }, + { + "epoch": 0.22481008598745264, + "grad_norm": 1.9823728799819946, + "learning_rate": 1.9646113617539853e-05, + "loss": 3.2388, + "step": 2804 + }, + { + "epoch": 0.22489026076847526, + "grad_norm": 0.9798767566680908, + "learning_rate": 1.964577116717593e-05, + "loss": 2.5613, + "step": 2805 + }, + { + "epoch": 0.22497043554949792, + "grad_norm": 2.1181013584136963, + "learning_rate": 1.96454285541877e-05, + "loss": 3.2827, + "step": 2806 + }, + { + "epoch": 0.22505061033052054, + "grad_norm": 2.1655960083007812, + "learning_rate": 1.9645085778580944e-05, + "loss": 3.2959, + "step": 2807 + }, + { + "epoch": 0.22513078511154316, + "grad_norm": 2.0142030715942383, + "learning_rate": 1.964474284036144e-05, + "loss": 3.2208, + "step": 2808 + }, + { + "epoch": 0.2252109598925658, + "grad_norm": 2.1043102741241455, + "learning_rate": 1.9644399739534974e-05, + "loss": 3.3151, + "step": 2809 + }, + { + "epoch": 0.2252911346735884, + "grad_norm": 2.0525431632995605, + "learning_rate": 1.9644056476107326e-05, + "loss": 3.3503, + "step": 2810 + }, + { + "epoch": 0.22537130945461106, + "grad_norm": 2.2014856338500977, + "learning_rate": 1.9643713050084284e-05, + "loss": 3.2828, + "step": 2811 + }, + { + "epoch": 0.2254514842356337, + "grad_norm": 2.474151611328125, + "learning_rate": 1.964336946147164e-05, + "loss": 3.248, + "step": 2812 + }, + { + "epoch": 0.2255316590166563, + "grad_norm": 1.9975308179855347, + "learning_rate": 1.964302571027519e-05, + "loss": 3.2534, + "step": 2813 + }, + { + "epoch": 0.22561183379767893, + "grad_norm": 2.050823211669922, + "learning_rate": 1.9642681796500716e-05, + "loss": 3.2594, + "step": 2814 + }, + { + "epoch": 0.22569200857870156, + "grad_norm": 2.3727614879608154, + "learning_rate": 1.964233772015403e-05, + "loss": 3.2915, + "step": 2815 + }, + { + "epoch": 0.2257721833597242, + "grad_norm": 2.0937435626983643, + "learning_rate": 1.964199348124093e-05, + "loss": 3.2832, + "step": 2816 + }, + { + "epoch": 0.22585235814074683, + "grad_norm": 2.1313393115997314, + "learning_rate": 1.9641649079767213e-05, + "loss": 3.1921, + "step": 2817 + }, + { + "epoch": 0.22593253292176946, + "grad_norm": 2.210308074951172, + "learning_rate": 1.9641304515738696e-05, + "loss": 3.295, + "step": 2818 + }, + { + "epoch": 0.22601270770279208, + "grad_norm": 2.1030328273773193, + "learning_rate": 1.964095978916118e-05, + "loss": 3.2961, + "step": 2819 + }, + { + "epoch": 0.2260928824838147, + "grad_norm": 2.158252716064453, + "learning_rate": 1.964061490004048e-05, + "loss": 3.2645, + "step": 2820 + }, + { + "epoch": 0.22617305726483736, + "grad_norm": 2.2052738666534424, + "learning_rate": 1.964026984838241e-05, + "loss": 3.3145, + "step": 2821 + }, + { + "epoch": 0.22625323204585998, + "grad_norm": 2.1880111694335938, + "learning_rate": 1.963992463419279e-05, + "loss": 3.2588, + "step": 2822 + }, + { + "epoch": 0.2263334068268826, + "grad_norm": 2.285766839981079, + "learning_rate": 1.9639579257477436e-05, + "loss": 3.3557, + "step": 2823 + }, + { + "epoch": 0.22641358160790523, + "grad_norm": 2.0834288597106934, + "learning_rate": 1.963923371824217e-05, + "loss": 3.3882, + "step": 2824 + }, + { + "epoch": 0.22649375638892785, + "grad_norm": 2.1418426036834717, + "learning_rate": 1.963888801649283e-05, + "loss": 3.3349, + "step": 2825 + }, + { + "epoch": 0.2265739311699505, + "grad_norm": 2.258115530014038, + "learning_rate": 1.9638542152235226e-05, + "loss": 3.267, + "step": 2826 + }, + { + "epoch": 0.22665410595097313, + "grad_norm": 2.0824742317199707, + "learning_rate": 1.9638196125475205e-05, + "loss": 3.3444, + "step": 2827 + }, + { + "epoch": 0.22673428073199575, + "grad_norm": 2.2438182830810547, + "learning_rate": 1.9637849936218585e-05, + "loss": 3.2777, + "step": 2828 + }, + { + "epoch": 0.22681445551301838, + "grad_norm": 2.2010738849639893, + "learning_rate": 1.963750358447122e-05, + "loss": 3.235, + "step": 2829 + }, + { + "epoch": 0.226894630294041, + "grad_norm": 2.1595306396484375, + "learning_rate": 1.9637157070238938e-05, + "loss": 3.3204, + "step": 2830 + }, + { + "epoch": 0.22697480507506365, + "grad_norm": 2.149864912033081, + "learning_rate": 1.963681039352759e-05, + "loss": 3.269, + "step": 2831 + }, + { + "epoch": 0.22705497985608628, + "grad_norm": 2.077648639678955, + "learning_rate": 1.9636463554343006e-05, + "loss": 3.3669, + "step": 2832 + }, + { + "epoch": 0.2271351546371089, + "grad_norm": 2.0011086463928223, + "learning_rate": 1.9636116552691048e-05, + "loss": 3.2584, + "step": 2833 + }, + { + "epoch": 0.22721532941813152, + "grad_norm": 2.004044532775879, + "learning_rate": 1.963576938857756e-05, + "loss": 3.2455, + "step": 2834 + }, + { + "epoch": 0.22729550419915415, + "grad_norm": 2.57122540473938, + "learning_rate": 1.9635422062008397e-05, + "loss": 3.3637, + "step": 2835 + }, + { + "epoch": 0.2273756789801768, + "grad_norm": 2.041649341583252, + "learning_rate": 1.9635074572989413e-05, + "loss": 3.3358, + "step": 2836 + }, + { + "epoch": 0.22745585376119942, + "grad_norm": 2.0546913146972656, + "learning_rate": 1.9634726921526467e-05, + "loss": 3.2836, + "step": 2837 + }, + { + "epoch": 0.22753602854222205, + "grad_norm": 2.1295278072357178, + "learning_rate": 1.963437910762542e-05, + "loss": 3.3016, + "step": 2838 + }, + { + "epoch": 0.22761620332324467, + "grad_norm": 2.068791389465332, + "learning_rate": 1.9634031131292137e-05, + "loss": 3.2424, + "step": 2839 + }, + { + "epoch": 0.2276963781042673, + "grad_norm": 2.374767541885376, + "learning_rate": 1.9633682992532485e-05, + "loss": 3.343, + "step": 2840 + }, + { + "epoch": 0.22777655288528995, + "grad_norm": 2.0141515731811523, + "learning_rate": 1.9633334691352334e-05, + "loss": 3.2739, + "step": 2841 + }, + { + "epoch": 0.22785672766631257, + "grad_norm": 1.9788978099822998, + "learning_rate": 1.9632986227757556e-05, + "loss": 3.3135, + "step": 2842 + }, + { + "epoch": 0.2279369024473352, + "grad_norm": 0.9829619526863098, + "learning_rate": 1.9632637601754022e-05, + "loss": 2.5587, + "step": 2843 + }, + { + "epoch": 0.22801707722835782, + "grad_norm": 0.9169785380363464, + "learning_rate": 1.9632288813347615e-05, + "loss": 2.535, + "step": 2844 + }, + { + "epoch": 0.22809725200938044, + "grad_norm": 2.2033488750457764, + "learning_rate": 1.9631939862544215e-05, + "loss": 3.2968, + "step": 2845 + }, + { + "epoch": 0.22817742679040307, + "grad_norm": 2.23008394241333, + "learning_rate": 1.96315907493497e-05, + "loss": 3.3479, + "step": 2846 + }, + { + "epoch": 0.22825760157142572, + "grad_norm": 2.0656774044036865, + "learning_rate": 1.963124147376996e-05, + "loss": 3.4794, + "step": 2847 + }, + { + "epoch": 0.22833777635244834, + "grad_norm": 1.0041916370391846, + "learning_rate": 1.9630892035810886e-05, + "loss": 2.5253, + "step": 2848 + }, + { + "epoch": 0.22841795113347096, + "grad_norm": 2.3135368824005127, + "learning_rate": 1.9630542435478364e-05, + "loss": 3.2385, + "step": 2849 + }, + { + "epoch": 0.2284981259144936, + "grad_norm": 2.2243380546569824, + "learning_rate": 1.963019267277829e-05, + "loss": 3.3192, + "step": 2850 + }, + { + "epoch": 0.2285783006955162, + "grad_norm": 1.9119293689727783, + "learning_rate": 1.9629842747716563e-05, + "loss": 3.2366, + "step": 2851 + }, + { + "epoch": 0.22865847547653886, + "grad_norm": 2.1235854625701904, + "learning_rate": 1.962949266029908e-05, + "loss": 3.3173, + "step": 2852 + }, + { + "epoch": 0.2287386502575615, + "grad_norm": 2.194995641708374, + "learning_rate": 1.9629142410531746e-05, + "loss": 3.3729, + "step": 2853 + }, + { + "epoch": 0.2288188250385841, + "grad_norm": 2.07537579536438, + "learning_rate": 1.9628791998420465e-05, + "loss": 3.2625, + "step": 2854 + }, + { + "epoch": 0.22889899981960674, + "grad_norm": 1.9502729177474976, + "learning_rate": 1.9628441423971143e-05, + "loss": 3.1903, + "step": 2855 + }, + { + "epoch": 0.22897917460062936, + "grad_norm": 2.044666290283203, + "learning_rate": 1.9628090687189694e-05, + "loss": 3.3258, + "step": 2856 + }, + { + "epoch": 0.229059349381652, + "grad_norm": 2.250727415084839, + "learning_rate": 1.9627739788082024e-05, + "loss": 3.2404, + "step": 2857 + }, + { + "epoch": 0.22913952416267463, + "grad_norm": 2.1176750659942627, + "learning_rate": 1.9627388726654063e-05, + "loss": 3.2231, + "step": 2858 + }, + { + "epoch": 0.22921969894369726, + "grad_norm": 2.1866843700408936, + "learning_rate": 1.9627037502911716e-05, + "loss": 3.1735, + "step": 2859 + }, + { + "epoch": 0.22929987372471988, + "grad_norm": 1.973892092704773, + "learning_rate": 1.9626686116860905e-05, + "loss": 3.2498, + "step": 2860 + }, + { + "epoch": 0.2293800485057425, + "grad_norm": 2.208529472351074, + "learning_rate": 1.9626334568507567e-05, + "loss": 3.1963, + "step": 2861 + }, + { + "epoch": 0.22946022328676516, + "grad_norm": 2.0686566829681396, + "learning_rate": 1.9625982857857617e-05, + "loss": 3.2222, + "step": 2862 + }, + { + "epoch": 0.22954039806778778, + "grad_norm": 2.1326940059661865, + "learning_rate": 1.9625630984916988e-05, + "loss": 3.3357, + "step": 2863 + }, + { + "epoch": 0.2296205728488104, + "grad_norm": 1.981576919555664, + "learning_rate": 1.9625278949691612e-05, + "loss": 3.24, + "step": 2864 + }, + { + "epoch": 0.22970074762983303, + "grad_norm": 2.0694899559020996, + "learning_rate": 1.962492675218743e-05, + "loss": 3.3557, + "step": 2865 + }, + { + "epoch": 0.22978092241085565, + "grad_norm": 1.9889123439788818, + "learning_rate": 1.9624574392410372e-05, + "loss": 3.3164, + "step": 2866 + }, + { + "epoch": 0.2298610971918783, + "grad_norm": 2.18742036819458, + "learning_rate": 1.962422187036638e-05, + "loss": 3.3475, + "step": 2867 + }, + { + "epoch": 0.22994127197290093, + "grad_norm": 2.1363110542297363, + "learning_rate": 1.96238691860614e-05, + "loss": 3.2782, + "step": 2868 + }, + { + "epoch": 0.23002144675392355, + "grad_norm": 2.228419303894043, + "learning_rate": 1.9623516339501383e-05, + "loss": 3.3258, + "step": 2869 + }, + { + "epoch": 0.23010162153494618, + "grad_norm": 2.1472368240356445, + "learning_rate": 1.9623163330692265e-05, + "loss": 3.3625, + "step": 2870 + }, + { + "epoch": 0.2301817963159688, + "grad_norm": 1.9697352647781372, + "learning_rate": 1.962281015964001e-05, + "loss": 3.355, + "step": 2871 + }, + { + "epoch": 0.23026197109699145, + "grad_norm": 1.9844835996627808, + "learning_rate": 1.9622456826350565e-05, + "loss": 3.2249, + "step": 2872 + }, + { + "epoch": 0.23034214587801408, + "grad_norm": 2.1353352069854736, + "learning_rate": 1.9622103330829888e-05, + "loss": 3.2532, + "step": 2873 + }, + { + "epoch": 0.2304223206590367, + "grad_norm": 2.086789846420288, + "learning_rate": 1.962174967308394e-05, + "loss": 3.3146, + "step": 2874 + }, + { + "epoch": 0.23050249544005932, + "grad_norm": 2.098440647125244, + "learning_rate": 1.9621395853118684e-05, + "loss": 3.3706, + "step": 2875 + }, + { + "epoch": 0.23058267022108195, + "grad_norm": 1.9490317106246948, + "learning_rate": 1.9621041870940086e-05, + "loss": 3.302, + "step": 2876 + }, + { + "epoch": 0.2306628450021046, + "grad_norm": 2.088524580001831, + "learning_rate": 1.962068772655411e-05, + "loss": 3.2794, + "step": 2877 + }, + { + "epoch": 0.23074301978312722, + "grad_norm": 2.0708374977111816, + "learning_rate": 1.9620333419966734e-05, + "loss": 3.2976, + "step": 2878 + }, + { + "epoch": 0.23082319456414985, + "grad_norm": 2.177375078201294, + "learning_rate": 1.9619978951183924e-05, + "loss": 3.1619, + "step": 2879 + }, + { + "epoch": 0.23090336934517247, + "grad_norm": 0.9997729063034058, + "learning_rate": 1.9619624320211658e-05, + "loss": 2.5605, + "step": 2880 + }, + { + "epoch": 0.2309835441261951, + "grad_norm": 2.099133014678955, + "learning_rate": 1.961926952705592e-05, + "loss": 3.2589, + "step": 2881 + }, + { + "epoch": 0.23106371890721775, + "grad_norm": 2.1152329444885254, + "learning_rate": 1.9618914571722683e-05, + "loss": 3.3187, + "step": 2882 + }, + { + "epoch": 0.23114389368824037, + "grad_norm": 2.010549306869507, + "learning_rate": 1.9618559454217942e-05, + "loss": 3.3832, + "step": 2883 + }, + { + "epoch": 0.231224068469263, + "grad_norm": 2.285966157913208, + "learning_rate": 1.961820417454768e-05, + "loss": 3.3262, + "step": 2884 + }, + { + "epoch": 0.23130424325028562, + "grad_norm": 2.1411850452423096, + "learning_rate": 1.9617848732717877e-05, + "loss": 3.2727, + "step": 2885 + }, + { + "epoch": 0.23138441803130824, + "grad_norm": 2.155729055404663, + "learning_rate": 1.9617493128734538e-05, + "loss": 3.3176, + "step": 2886 + }, + { + "epoch": 0.2314645928123309, + "grad_norm": 2.05302095413208, + "learning_rate": 1.9617137362603655e-05, + "loss": 3.2064, + "step": 2887 + }, + { + "epoch": 0.23154476759335352, + "grad_norm": 2.003640651702881, + "learning_rate": 1.9616781434331224e-05, + "loss": 3.1761, + "step": 2888 + }, + { + "epoch": 0.23162494237437614, + "grad_norm": 2.070336103439331, + "learning_rate": 1.961642534392325e-05, + "loss": 3.2145, + "step": 2889 + }, + { + "epoch": 0.23170511715539877, + "grad_norm": 1.9320026636123657, + "learning_rate": 1.9616069091385733e-05, + "loss": 3.339, + "step": 2890 + }, + { + "epoch": 0.2317852919364214, + "grad_norm": 2.109764337539673, + "learning_rate": 1.961571267672468e-05, + "loss": 3.2938, + "step": 2891 + }, + { + "epoch": 0.23186546671744404, + "grad_norm": 2.0604817867279053, + "learning_rate": 1.9615356099946098e-05, + "loss": 3.3131, + "step": 2892 + }, + { + "epoch": 0.23194564149846666, + "grad_norm": 1.0592784881591797, + "learning_rate": 1.9614999361056e-05, + "loss": 2.5356, + "step": 2893 + }, + { + "epoch": 0.2320258162794893, + "grad_norm": 2.1197686195373535, + "learning_rate": 1.9614642460060407e-05, + "loss": 3.3195, + "step": 2894 + }, + { + "epoch": 0.2321059910605119, + "grad_norm": 2.0376200675964355, + "learning_rate": 1.9614285396965327e-05, + "loss": 3.257, + "step": 2895 + }, + { + "epoch": 0.23218616584153454, + "grad_norm": 0.8571614027023315, + "learning_rate": 1.9613928171776786e-05, + "loss": 2.5404, + "step": 2896 + }, + { + "epoch": 0.2322663406225572, + "grad_norm": 1.9854453802108765, + "learning_rate": 1.96135707845008e-05, + "loss": 3.4534, + "step": 2897 + }, + { + "epoch": 0.2323465154035798, + "grad_norm": 1.037591814994812, + "learning_rate": 1.9613213235143403e-05, + "loss": 2.5631, + "step": 2898 + }, + { + "epoch": 0.23242669018460244, + "grad_norm": 2.1063003540039062, + "learning_rate": 1.9612855523710615e-05, + "loss": 3.2061, + "step": 2899 + }, + { + "epoch": 0.23250686496562506, + "grad_norm": 2.00832462310791, + "learning_rate": 1.9612497650208478e-05, + "loss": 3.2194, + "step": 2900 + }, + { + "epoch": 0.23258703974664768, + "grad_norm": 2.338132381439209, + "learning_rate": 1.9612139614643012e-05, + "loss": 3.2484, + "step": 2901 + }, + { + "epoch": 0.23266721452767034, + "grad_norm": 0.9096808433532715, + "learning_rate": 1.9611781417020262e-05, + "loss": 2.5537, + "step": 2902 + }, + { + "epoch": 0.23274738930869296, + "grad_norm": 2.1220574378967285, + "learning_rate": 1.9611423057346265e-05, + "loss": 3.2632, + "step": 2903 + }, + { + "epoch": 0.23282756408971558, + "grad_norm": 1.9563207626342773, + "learning_rate": 1.961106453562706e-05, + "loss": 3.2961, + "step": 2904 + }, + { + "epoch": 0.2329077388707382, + "grad_norm": 2.205922842025757, + "learning_rate": 1.9610705851868694e-05, + "loss": 3.321, + "step": 2905 + }, + { + "epoch": 0.23298791365176083, + "grad_norm": 2.047856330871582, + "learning_rate": 1.9610347006077217e-05, + "loss": 3.2295, + "step": 2906 + }, + { + "epoch": 0.23306808843278345, + "grad_norm": 2.245032787322998, + "learning_rate": 1.9609987998258674e-05, + "loss": 3.3816, + "step": 2907 + }, + { + "epoch": 0.2331482632138061, + "grad_norm": 2.081084728240967, + "learning_rate": 1.9609628828419127e-05, + "loss": 3.3525, + "step": 2908 + }, + { + "epoch": 0.23322843799482873, + "grad_norm": 1.9966280460357666, + "learning_rate": 1.960926949656462e-05, + "loss": 3.3014, + "step": 2909 + }, + { + "epoch": 0.23330861277585135, + "grad_norm": 1.0147496461868286, + "learning_rate": 1.960891000270121e-05, + "loss": 2.5538, + "step": 2910 + }, + { + "epoch": 0.23338878755687398, + "grad_norm": 2.032623529434204, + "learning_rate": 1.960855034683497e-05, + "loss": 3.3235, + "step": 2911 + }, + { + "epoch": 0.2334689623378966, + "grad_norm": 1.931504249572754, + "learning_rate": 1.960819052897196e-05, + "loss": 3.2671, + "step": 2912 + }, + { + "epoch": 0.23354913711891925, + "grad_norm": 1.9290568828582764, + "learning_rate": 1.960783054911824e-05, + "loss": 3.2163, + "step": 2913 + }, + { + "epoch": 0.23362931189994188, + "grad_norm": 1.8725523948669434, + "learning_rate": 1.960747040727988e-05, + "loss": 3.2738, + "step": 2914 + }, + { + "epoch": 0.2337094866809645, + "grad_norm": 1.971794843673706, + "learning_rate": 1.9607110103462962e-05, + "loss": 3.2495, + "step": 2915 + }, + { + "epoch": 0.23378966146198712, + "grad_norm": 2.287850856781006, + "learning_rate": 1.9606749637673552e-05, + "loss": 3.2249, + "step": 2916 + }, + { + "epoch": 0.23386983624300975, + "grad_norm": 2.416687488555908, + "learning_rate": 1.9606389009917728e-05, + "loss": 3.2255, + "step": 2917 + }, + { + "epoch": 0.2339500110240324, + "grad_norm": 1.961500883102417, + "learning_rate": 1.9606028220201568e-05, + "loss": 3.3362, + "step": 2918 + }, + { + "epoch": 0.23403018580505502, + "grad_norm": 2.0115015506744385, + "learning_rate": 1.960566726853116e-05, + "loss": 3.323, + "step": 2919 + }, + { + "epoch": 0.23411036058607765, + "grad_norm": 1.957504391670227, + "learning_rate": 1.960530615491259e-05, + "loss": 3.2573, + "step": 2920 + }, + { + "epoch": 0.23419053536710027, + "grad_norm": 2.052816152572632, + "learning_rate": 1.960494487935194e-05, + "loss": 3.2924, + "step": 2921 + }, + { + "epoch": 0.2342707101481229, + "grad_norm": 2.0293867588043213, + "learning_rate": 1.9604583441855305e-05, + "loss": 3.4846, + "step": 2922 + }, + { + "epoch": 0.23435088492914555, + "grad_norm": 2.1877281665802, + "learning_rate": 1.960422184242878e-05, + "loss": 3.2727, + "step": 2923 + }, + { + "epoch": 0.23443105971016817, + "grad_norm": 2.005934238433838, + "learning_rate": 1.9603860081078456e-05, + "loss": 3.335, + "step": 2924 + }, + { + "epoch": 0.2345112344911908, + "grad_norm": 2.216430425643921, + "learning_rate": 1.9603498157810438e-05, + "loss": 3.2297, + "step": 2925 + }, + { + "epoch": 0.23459140927221342, + "grad_norm": 2.1623623371124268, + "learning_rate": 1.9603136072630823e-05, + "loss": 3.2435, + "step": 2926 + }, + { + "epoch": 0.23467158405323604, + "grad_norm": 1.9455180168151855, + "learning_rate": 1.960277382554572e-05, + "loss": 3.3507, + "step": 2927 + }, + { + "epoch": 0.2347517588342587, + "grad_norm": 2.2209227085113525, + "learning_rate": 1.960241141656124e-05, + "loss": 3.3541, + "step": 2928 + }, + { + "epoch": 0.23483193361528132, + "grad_norm": 1.9940037727355957, + "learning_rate": 1.9602048845683482e-05, + "loss": 3.291, + "step": 2929 + }, + { + "epoch": 0.23491210839630394, + "grad_norm": 2.0431904792785645, + "learning_rate": 1.9601686112918563e-05, + "loss": 3.2131, + "step": 2930 + }, + { + "epoch": 0.23499228317732657, + "grad_norm": 0.9450089931488037, + "learning_rate": 1.9601323218272602e-05, + "loss": 2.538, + "step": 2931 + }, + { + "epoch": 0.2350724579583492, + "grad_norm": 2.0922176837921143, + "learning_rate": 1.9600960161751717e-05, + "loss": 3.3411, + "step": 2932 + }, + { + "epoch": 0.23515263273937184, + "grad_norm": 0.8669295907020569, + "learning_rate": 1.9600596943362026e-05, + "loss": 2.5699, + "step": 2933 + }, + { + "epoch": 0.23523280752039447, + "grad_norm": 1.8688716888427734, + "learning_rate": 1.9600233563109654e-05, + "loss": 3.2319, + "step": 2934 + }, + { + "epoch": 0.2353129823014171, + "grad_norm": 1.9522439241409302, + "learning_rate": 1.959987002100073e-05, + "loss": 3.207, + "step": 2935 + }, + { + "epoch": 0.2353931570824397, + "grad_norm": 1.9609640836715698, + "learning_rate": 1.959950631704138e-05, + "loss": 3.238, + "step": 2936 + }, + { + "epoch": 0.23547333186346234, + "grad_norm": 2.0512166023254395, + "learning_rate": 1.9599142451237733e-05, + "loss": 3.198, + "step": 2937 + }, + { + "epoch": 0.235553506644485, + "grad_norm": 2.1876652240753174, + "learning_rate": 1.959877842359593e-05, + "loss": 3.2421, + "step": 2938 + }, + { + "epoch": 0.2356336814255076, + "grad_norm": 1.9179491996765137, + "learning_rate": 1.9598414234122106e-05, + "loss": 3.4292, + "step": 2939 + }, + { + "epoch": 0.23571385620653024, + "grad_norm": 2.1613221168518066, + "learning_rate": 1.9598049882822402e-05, + "loss": 3.3105, + "step": 2940 + }, + { + "epoch": 0.23579403098755286, + "grad_norm": 2.0390450954437256, + "learning_rate": 1.9597685369702955e-05, + "loss": 3.2347, + "step": 2941 + }, + { + "epoch": 0.23587420576857548, + "grad_norm": 2.0767784118652344, + "learning_rate": 1.9597320694769923e-05, + "loss": 3.2739, + "step": 2942 + }, + { + "epoch": 0.23595438054959814, + "grad_norm": 2.1228528022766113, + "learning_rate": 1.959695585802944e-05, + "loss": 3.272, + "step": 2943 + }, + { + "epoch": 0.23603455533062076, + "grad_norm": 2.157351016998291, + "learning_rate": 1.9596590859487663e-05, + "loss": 3.3248, + "step": 2944 + }, + { + "epoch": 0.23611473011164338, + "grad_norm": 1.9902335405349731, + "learning_rate": 1.959622569915075e-05, + "loss": 3.2675, + "step": 2945 + }, + { + "epoch": 0.236194904892666, + "grad_norm": 1.970341444015503, + "learning_rate": 1.959586037702485e-05, + "loss": 3.2416, + "step": 2946 + }, + { + "epoch": 0.23627507967368863, + "grad_norm": 2.029099941253662, + "learning_rate": 1.9595494893116133e-05, + "loss": 3.2551, + "step": 2947 + }, + { + "epoch": 0.23635525445471128, + "grad_norm": 2.287975311279297, + "learning_rate": 1.959512924743075e-05, + "loss": 3.3394, + "step": 2948 + }, + { + "epoch": 0.2364354292357339, + "grad_norm": 2.124652624130249, + "learning_rate": 1.9594763439974866e-05, + "loss": 3.2375, + "step": 2949 + }, + { + "epoch": 0.23651560401675653, + "grad_norm": 1.984645128250122, + "learning_rate": 1.9594397470754652e-05, + "loss": 3.3918, + "step": 2950 + }, + { + "epoch": 0.23659577879777915, + "grad_norm": 2.0193865299224854, + "learning_rate": 1.9594031339776283e-05, + "loss": 3.3054, + "step": 2951 + }, + { + "epoch": 0.23667595357880178, + "grad_norm": 1.9615354537963867, + "learning_rate": 1.9593665047045924e-05, + "loss": 3.2911, + "step": 2952 + }, + { + "epoch": 0.23675612835982443, + "grad_norm": 1.9674359560012817, + "learning_rate": 1.9593298592569754e-05, + "loss": 3.2659, + "step": 2953 + }, + { + "epoch": 0.23683630314084705, + "grad_norm": 2.0418124198913574, + "learning_rate": 1.959293197635395e-05, + "loss": 3.2318, + "step": 2954 + }, + { + "epoch": 0.23691647792186968, + "grad_norm": 2.1029465198516846, + "learning_rate": 1.9592565198404695e-05, + "loss": 3.2538, + "step": 2955 + }, + { + "epoch": 0.2369966527028923, + "grad_norm": 1.0263588428497314, + "learning_rate": 1.9592198258728172e-05, + "loss": 2.5273, + "step": 2956 + }, + { + "epoch": 0.23707682748391493, + "grad_norm": 2.6133458614349365, + "learning_rate": 1.9591831157330565e-05, + "loss": 3.3588, + "step": 2957 + }, + { + "epoch": 0.23715700226493758, + "grad_norm": 2.036123037338257, + "learning_rate": 1.9591463894218065e-05, + "loss": 3.2421, + "step": 2958 + }, + { + "epoch": 0.2372371770459602, + "grad_norm": 0.9828826785087585, + "learning_rate": 1.9591096469396866e-05, + "loss": 2.5359, + "step": 2959 + }, + { + "epoch": 0.23731735182698283, + "grad_norm": 2.0093767642974854, + "learning_rate": 1.9590728882873163e-05, + "loss": 3.2938, + "step": 2960 + }, + { + "epoch": 0.23739752660800545, + "grad_norm": 2.3046274185180664, + "learning_rate": 1.9590361134653147e-05, + "loss": 3.3076, + "step": 2961 + }, + { + "epoch": 0.23747770138902807, + "grad_norm": 2.0813207626342773, + "learning_rate": 1.9589993224743023e-05, + "loss": 3.2433, + "step": 2962 + }, + { + "epoch": 0.2375578761700507, + "grad_norm": 2.2896857261657715, + "learning_rate": 1.9589625153148996e-05, + "loss": 3.3375, + "step": 2963 + }, + { + "epoch": 0.23763805095107335, + "grad_norm": 2.068748712539673, + "learning_rate": 1.958925691987726e-05, + "loss": 3.2593, + "step": 2964 + }, + { + "epoch": 0.23771822573209597, + "grad_norm": 1.9687508344650269, + "learning_rate": 1.958888852493404e-05, + "loss": 3.2845, + "step": 2965 + }, + { + "epoch": 0.2377984005131186, + "grad_norm": 2.1918551921844482, + "learning_rate": 1.958851996832554e-05, + "loss": 3.3074, + "step": 2966 + }, + { + "epoch": 0.23787857529414122, + "grad_norm": 1.9239635467529297, + "learning_rate": 1.958815125005797e-05, + "loss": 3.33, + "step": 2967 + }, + { + "epoch": 0.23795875007516384, + "grad_norm": 2.0435986518859863, + "learning_rate": 1.9587782370137548e-05, + "loss": 3.2595, + "step": 2968 + }, + { + "epoch": 0.2380389248561865, + "grad_norm": 1.9860968589782715, + "learning_rate": 1.9587413328570494e-05, + "loss": 3.3296, + "step": 2969 + }, + { + "epoch": 0.23811909963720912, + "grad_norm": 2.069615602493286, + "learning_rate": 1.9587044125363032e-05, + "loss": 3.2635, + "step": 2970 + }, + { + "epoch": 0.23819927441823174, + "grad_norm": 2.1281650066375732, + "learning_rate": 1.9586674760521382e-05, + "loss": 3.4199, + "step": 2971 + }, + { + "epoch": 0.23827944919925437, + "grad_norm": 2.0541439056396484, + "learning_rate": 1.9586305234051778e-05, + "loss": 3.3116, + "step": 2972 + }, + { + "epoch": 0.238359623980277, + "grad_norm": 2.1159892082214355, + "learning_rate": 1.9585935545960444e-05, + "loss": 3.2618, + "step": 2973 + }, + { + "epoch": 0.23843979876129964, + "grad_norm": 2.0798568725585938, + "learning_rate": 1.9585565696253614e-05, + "loss": 3.3177, + "step": 2974 + }, + { + "epoch": 0.23851997354232227, + "grad_norm": 1.9853514432907104, + "learning_rate": 1.958519568493753e-05, + "loss": 3.2999, + "step": 2975 + }, + { + "epoch": 0.2386001483233449, + "grad_norm": 2.1489598751068115, + "learning_rate": 1.958482551201842e-05, + "loss": 3.179, + "step": 2976 + }, + { + "epoch": 0.23868032310436751, + "grad_norm": 2.225635290145874, + "learning_rate": 1.958445517750253e-05, + "loss": 3.2629, + "step": 2977 + }, + { + "epoch": 0.23876049788539014, + "grad_norm": 2.011554718017578, + "learning_rate": 1.9584084681396106e-05, + "loss": 3.3003, + "step": 2978 + }, + { + "epoch": 0.2388406726664128, + "grad_norm": 2.1025753021240234, + "learning_rate": 1.9583714023705388e-05, + "loss": 3.3412, + "step": 2979 + }, + { + "epoch": 0.2389208474474354, + "grad_norm": 2.0363895893096924, + "learning_rate": 1.9583343204436633e-05, + "loss": 3.283, + "step": 2980 + }, + { + "epoch": 0.23900102222845804, + "grad_norm": 2.1719510555267334, + "learning_rate": 1.9582972223596087e-05, + "loss": 3.292, + "step": 2981 + }, + { + "epoch": 0.23908119700948066, + "grad_norm": 1.9245455265045166, + "learning_rate": 1.958260108119001e-05, + "loss": 3.2178, + "step": 2982 + }, + { + "epoch": 0.23916137179050329, + "grad_norm": 2.311633586883545, + "learning_rate": 1.9582229777224652e-05, + "loss": 3.3019, + "step": 2983 + }, + { + "epoch": 0.23924154657152594, + "grad_norm": 2.127453565597534, + "learning_rate": 1.958185831170628e-05, + "loss": 3.2, + "step": 2984 + }, + { + "epoch": 0.23932172135254856, + "grad_norm": 2.032238245010376, + "learning_rate": 1.9581486684641153e-05, + "loss": 3.3439, + "step": 2985 + }, + { + "epoch": 0.23940189613357118, + "grad_norm": 1.8835654258728027, + "learning_rate": 1.9581114896035534e-05, + "loss": 3.3325, + "step": 2986 + }, + { + "epoch": 0.2394820709145938, + "grad_norm": 1.9999099969863892, + "learning_rate": 1.95807429458957e-05, + "loss": 3.1896, + "step": 2987 + }, + { + "epoch": 0.23956224569561643, + "grad_norm": 2.050292730331421, + "learning_rate": 1.9580370834227914e-05, + "loss": 3.2985, + "step": 2988 + }, + { + "epoch": 0.23964242047663908, + "grad_norm": 2.1132874488830566, + "learning_rate": 1.9579998561038453e-05, + "loss": 3.3191, + "step": 2989 + }, + { + "epoch": 0.2397225952576617, + "grad_norm": 2.1914381980895996, + "learning_rate": 1.957962612633359e-05, + "loss": 3.4141, + "step": 2990 + }, + { + "epoch": 0.23980277003868433, + "grad_norm": 2.1006686687469482, + "learning_rate": 1.957925353011961e-05, + "loss": 3.3508, + "step": 2991 + }, + { + "epoch": 0.23988294481970696, + "grad_norm": 2.278263807296753, + "learning_rate": 1.9578880772402787e-05, + "loss": 3.2603, + "step": 2992 + }, + { + "epoch": 0.23996311960072958, + "grad_norm": 1.8570445775985718, + "learning_rate": 1.9578507853189418e-05, + "loss": 3.1241, + "step": 2993 + }, + { + "epoch": 0.24004329438175223, + "grad_norm": 1.0075626373291016, + "learning_rate": 1.9578134772485775e-05, + "loss": 2.5331, + "step": 2994 + }, + { + "epoch": 0.24012346916277486, + "grad_norm": 2.0795822143554688, + "learning_rate": 1.9577761530298157e-05, + "loss": 3.4764, + "step": 2995 + }, + { + "epoch": 0.24020364394379748, + "grad_norm": 1.9812713861465454, + "learning_rate": 1.9577388126632856e-05, + "loss": 3.2845, + "step": 2996 + }, + { + "epoch": 0.2402838187248201, + "grad_norm": 2.0223326683044434, + "learning_rate": 1.9577014561496166e-05, + "loss": 3.3342, + "step": 2997 + }, + { + "epoch": 0.24036399350584273, + "grad_norm": 1.9956740140914917, + "learning_rate": 1.957664083489439e-05, + "loss": 3.3017, + "step": 2998 + }, + { + "epoch": 0.24044416828686538, + "grad_norm": 2.109856128692627, + "learning_rate": 1.957626694683382e-05, + "loss": 3.2919, + "step": 2999 + }, + { + "epoch": 0.240524343067888, + "grad_norm": 2.0082437992095947, + "learning_rate": 1.9575892897320763e-05, + "loss": 3.1731, + "step": 3000 + }, + { + "epoch": 0.24060451784891063, + "grad_norm": 2.484584093093872, + "learning_rate": 1.957551868636153e-05, + "loss": 3.3143, + "step": 3001 + }, + { + "epoch": 0.24068469262993325, + "grad_norm": 2.0501654148101807, + "learning_rate": 1.9575144313962427e-05, + "loss": 3.3378, + "step": 3002 + }, + { + "epoch": 0.24076486741095587, + "grad_norm": 1.9639469385147095, + "learning_rate": 1.957476978012976e-05, + "loss": 3.3877, + "step": 3003 + }, + { + "epoch": 0.24084504219197853, + "grad_norm": 2.1242551803588867, + "learning_rate": 1.9574395084869855e-05, + "loss": 3.2819, + "step": 3004 + }, + { + "epoch": 0.24092521697300115, + "grad_norm": 2.154097318649292, + "learning_rate": 1.957402022818902e-05, + "loss": 3.2615, + "step": 3005 + }, + { + "epoch": 0.24100539175402377, + "grad_norm": 1.9974793195724487, + "learning_rate": 1.957364521009358e-05, + "loss": 3.2197, + "step": 3006 + }, + { + "epoch": 0.2410855665350464, + "grad_norm": 1.9912935495376587, + "learning_rate": 1.957327003058986e-05, + "loss": 3.3336, + "step": 3007 + }, + { + "epoch": 0.24116574131606902, + "grad_norm": 2.01576828956604, + "learning_rate": 1.9572894689684174e-05, + "loss": 3.269, + "step": 3008 + }, + { + "epoch": 0.24124591609709167, + "grad_norm": 2.026561737060547, + "learning_rate": 1.957251918738286e-05, + "loss": 3.3269, + "step": 3009 + }, + { + "epoch": 0.2413260908781143, + "grad_norm": 2.1193506717681885, + "learning_rate": 1.9572143523692248e-05, + "loss": 3.2717, + "step": 3010 + }, + { + "epoch": 0.24140626565913692, + "grad_norm": 1.9615144729614258, + "learning_rate": 1.9571767698618664e-05, + "loss": 3.2329, + "step": 3011 + }, + { + "epoch": 0.24148644044015954, + "grad_norm": 1.8499579429626465, + "learning_rate": 1.9571391712168455e-05, + "loss": 3.2469, + "step": 3012 + }, + { + "epoch": 0.24156661522118217, + "grad_norm": 2.0811469554901123, + "learning_rate": 1.9571015564347957e-05, + "loss": 3.2405, + "step": 3013 + }, + { + "epoch": 0.24164679000220482, + "grad_norm": 2.1006247997283936, + "learning_rate": 1.9570639255163504e-05, + "loss": 3.3014, + "step": 3014 + }, + { + "epoch": 0.24172696478322744, + "grad_norm": 2.1690895557403564, + "learning_rate": 1.9570262784621446e-05, + "loss": 3.3075, + "step": 3015 + }, + { + "epoch": 0.24180713956425007, + "grad_norm": 2.1960699558258057, + "learning_rate": 1.9569886152728133e-05, + "loss": 3.177, + "step": 3016 + }, + { + "epoch": 0.2418873143452727, + "grad_norm": 2.043360471725464, + "learning_rate": 1.9569509359489915e-05, + "loss": 3.3202, + "step": 3017 + }, + { + "epoch": 0.24196748912629532, + "grad_norm": 2.228943109512329, + "learning_rate": 1.956913240491314e-05, + "loss": 3.3262, + "step": 3018 + }, + { + "epoch": 0.24204766390731797, + "grad_norm": 2.1224095821380615, + "learning_rate": 1.9568755289004164e-05, + "loss": 3.2974, + "step": 3019 + }, + { + "epoch": 0.2421278386883406, + "grad_norm": 2.1471941471099854, + "learning_rate": 1.9568378011769344e-05, + "loss": 3.2705, + "step": 3020 + }, + { + "epoch": 0.24220801346936321, + "grad_norm": 2.027681827545166, + "learning_rate": 1.9568000573215046e-05, + "loss": 3.2638, + "step": 3021 + }, + { + "epoch": 0.24228818825038584, + "grad_norm": 2.2475035190582275, + "learning_rate": 1.9567622973347632e-05, + "loss": 3.2287, + "step": 3022 + }, + { + "epoch": 0.24236836303140846, + "grad_norm": 2.1126224994659424, + "learning_rate": 1.9567245212173464e-05, + "loss": 3.1345, + "step": 3023 + }, + { + "epoch": 0.2424485378124311, + "grad_norm": 1.9886542558670044, + "learning_rate": 1.9566867289698912e-05, + "loss": 3.3345, + "step": 3024 + }, + { + "epoch": 0.24252871259345374, + "grad_norm": 2.0556821823120117, + "learning_rate": 1.956648920593035e-05, + "loss": 3.2872, + "step": 3025 + }, + { + "epoch": 0.24260888737447636, + "grad_norm": 2.161790609359741, + "learning_rate": 1.9566110960874155e-05, + "loss": 3.2339, + "step": 3026 + }, + { + "epoch": 0.24268906215549899, + "grad_norm": 1.941391110420227, + "learning_rate": 1.9565732554536698e-05, + "loss": 3.2902, + "step": 3027 + }, + { + "epoch": 0.2427692369365216, + "grad_norm": 2.0798959732055664, + "learning_rate": 1.9565353986924364e-05, + "loss": 3.2679, + "step": 3028 + }, + { + "epoch": 0.24284941171754423, + "grad_norm": 1.9967162609100342, + "learning_rate": 1.9564975258043527e-05, + "loss": 3.2227, + "step": 3029 + }, + { + "epoch": 0.24292958649856689, + "grad_norm": 1.952695608139038, + "learning_rate": 1.9564596367900583e-05, + "loss": 3.2599, + "step": 3030 + }, + { + "epoch": 0.2430097612795895, + "grad_norm": 2.036403179168701, + "learning_rate": 1.9564217316501916e-05, + "loss": 3.3054, + "step": 3031 + }, + { + "epoch": 0.24308993606061213, + "grad_norm": 2.1079251766204834, + "learning_rate": 1.956383810385391e-05, + "loss": 3.2725, + "step": 3032 + }, + { + "epoch": 0.24317011084163476, + "grad_norm": 2.0236828327178955, + "learning_rate": 1.956345872996297e-05, + "loss": 3.3068, + "step": 3033 + }, + { + "epoch": 0.24325028562265738, + "grad_norm": 2.138733386993408, + "learning_rate": 1.9563079194835482e-05, + "loss": 3.4758, + "step": 3034 + }, + { + "epoch": 0.24333046040368003, + "grad_norm": 1.8691213130950928, + "learning_rate": 1.956269949847785e-05, + "loss": 3.2558, + "step": 3035 + }, + { + "epoch": 0.24341063518470266, + "grad_norm": 2.065993309020996, + "learning_rate": 1.956231964089648e-05, + "loss": 3.3205, + "step": 3036 + }, + { + "epoch": 0.24349080996572528, + "grad_norm": 2.0376884937286377, + "learning_rate": 1.9561939622097768e-05, + "loss": 3.2423, + "step": 3037 + }, + { + "epoch": 0.2435709847467479, + "grad_norm": 1.928606390953064, + "learning_rate": 1.9561559442088122e-05, + "loss": 3.1555, + "step": 3038 + }, + { + "epoch": 0.24365115952777053, + "grad_norm": 2.1982247829437256, + "learning_rate": 1.9561179100873954e-05, + "loss": 3.3574, + "step": 3039 + }, + { + "epoch": 0.24373133430879318, + "grad_norm": 2.3331618309020996, + "learning_rate": 1.9560798598461678e-05, + "loss": 3.3657, + "step": 3040 + }, + { + "epoch": 0.2438115090898158, + "grad_norm": 1.9933456182479858, + "learning_rate": 1.9560417934857704e-05, + "loss": 3.3423, + "step": 3041 + }, + { + "epoch": 0.24389168387083843, + "grad_norm": 2.2694203853607178, + "learning_rate": 1.9560037110068456e-05, + "loss": 3.3102, + "step": 3042 + }, + { + "epoch": 0.24397185865186105, + "grad_norm": 1.948577642440796, + "learning_rate": 1.9559656124100354e-05, + "loss": 3.2797, + "step": 3043 + }, + { + "epoch": 0.24405203343288367, + "grad_norm": 1.9591971635818481, + "learning_rate": 1.9559274976959812e-05, + "loss": 3.299, + "step": 3044 + }, + { + "epoch": 0.24413220821390633, + "grad_norm": 2.1127822399139404, + "learning_rate": 1.9558893668653268e-05, + "loss": 3.3235, + "step": 3045 + }, + { + "epoch": 0.24421238299492895, + "grad_norm": 1.963666319847107, + "learning_rate": 1.955851219918714e-05, + "loss": 3.2289, + "step": 3046 + }, + { + "epoch": 0.24429255777595157, + "grad_norm": 2.056633949279785, + "learning_rate": 1.955813056856787e-05, + "loss": 3.3356, + "step": 3047 + }, + { + "epoch": 0.2443727325569742, + "grad_norm": 2.011052370071411, + "learning_rate": 1.9557748776801887e-05, + "loss": 3.2791, + "step": 3048 + }, + { + "epoch": 0.24445290733799682, + "grad_norm": 0.9495418667793274, + "learning_rate": 1.9557366823895626e-05, + "loss": 2.5474, + "step": 3049 + }, + { + "epoch": 0.24453308211901947, + "grad_norm": 2.1321282386779785, + "learning_rate": 1.9556984709855533e-05, + "loss": 3.3145, + "step": 3050 + }, + { + "epoch": 0.2446132569000421, + "grad_norm": 2.354271411895752, + "learning_rate": 1.9556602434688043e-05, + "loss": 3.2454, + "step": 3051 + }, + { + "epoch": 0.24469343168106472, + "grad_norm": 1.9401400089263916, + "learning_rate": 1.95562199983996e-05, + "loss": 3.2112, + "step": 3052 + }, + { + "epoch": 0.24477360646208735, + "grad_norm": 2.0370032787323, + "learning_rate": 1.955583740099666e-05, + "loss": 3.2762, + "step": 3053 + }, + { + "epoch": 0.24485378124310997, + "grad_norm": 2.007441282272339, + "learning_rate": 1.955545464248567e-05, + "loss": 3.25, + "step": 3054 + }, + { + "epoch": 0.24493395602413262, + "grad_norm": 2.0910966396331787, + "learning_rate": 1.9555071722873085e-05, + "loss": 3.2151, + "step": 3055 + }, + { + "epoch": 0.24501413080515524, + "grad_norm": 2.0467641353607178, + "learning_rate": 1.9554688642165352e-05, + "loss": 3.3391, + "step": 3056 + }, + { + "epoch": 0.24509430558617787, + "grad_norm": 2.085620880126953, + "learning_rate": 1.9554305400368937e-05, + "loss": 3.401, + "step": 3057 + }, + { + "epoch": 0.2451744803672005, + "grad_norm": 1.8480110168457031, + "learning_rate": 1.9553921997490305e-05, + "loss": 3.2059, + "step": 3058 + }, + { + "epoch": 0.24525465514822312, + "grad_norm": 2.1256606578826904, + "learning_rate": 1.9553538433535914e-05, + "loss": 3.2747, + "step": 3059 + }, + { + "epoch": 0.24533482992924577, + "grad_norm": 2.021613597869873, + "learning_rate": 1.9553154708512227e-05, + "loss": 3.183, + "step": 3060 + }, + { + "epoch": 0.2454150047102684, + "grad_norm": 1.93357253074646, + "learning_rate": 1.9552770822425723e-05, + "loss": 3.3175, + "step": 3061 + }, + { + "epoch": 0.24549517949129102, + "grad_norm": 2.329155921936035, + "learning_rate": 1.9552386775282868e-05, + "loss": 3.3249, + "step": 3062 + }, + { + "epoch": 0.24557535427231364, + "grad_norm": 2.0978734493255615, + "learning_rate": 1.9552002567090138e-05, + "loss": 3.2852, + "step": 3063 + }, + { + "epoch": 0.24565552905333626, + "grad_norm": 2.036546468734741, + "learning_rate": 1.955161819785401e-05, + "loss": 3.3351, + "step": 3064 + }, + { + "epoch": 0.24573570383435892, + "grad_norm": 2.0251543521881104, + "learning_rate": 1.9551233667580968e-05, + "loss": 3.213, + "step": 3065 + }, + { + "epoch": 0.24581587861538154, + "grad_norm": 2.238334894180298, + "learning_rate": 1.9550848976277494e-05, + "loss": 3.3751, + "step": 3066 + }, + { + "epoch": 0.24589605339640416, + "grad_norm": 2.446735382080078, + "learning_rate": 1.9550464123950072e-05, + "loss": 3.199, + "step": 3067 + }, + { + "epoch": 0.2459762281774268, + "grad_norm": 2.3758208751678467, + "learning_rate": 1.9550079110605186e-05, + "loss": 3.2424, + "step": 3068 + }, + { + "epoch": 0.2460564029584494, + "grad_norm": 2.1315319538116455, + "learning_rate": 1.9549693936249333e-05, + "loss": 3.2506, + "step": 3069 + }, + { + "epoch": 0.24613657773947206, + "grad_norm": 2.335379123687744, + "learning_rate": 1.9549308600889013e-05, + "loss": 3.2927, + "step": 3070 + }, + { + "epoch": 0.24621675252049469, + "grad_norm": 2.3903393745422363, + "learning_rate": 1.9548923104530708e-05, + "loss": 3.2792, + "step": 3071 + }, + { + "epoch": 0.2462969273015173, + "grad_norm": 2.2387735843658447, + "learning_rate": 1.9548537447180926e-05, + "loss": 3.2684, + "step": 3072 + }, + { + "epoch": 0.24637710208253993, + "grad_norm": 2.0516648292541504, + "learning_rate": 1.954815162884617e-05, + "loss": 3.3082, + "step": 3073 + }, + { + "epoch": 0.24645727686356256, + "grad_norm": 2.1323390007019043, + "learning_rate": 1.9547765649532945e-05, + "loss": 3.3324, + "step": 3074 + }, + { + "epoch": 0.2465374516445852, + "grad_norm": 1.0070278644561768, + "learning_rate": 1.954737950924775e-05, + "loss": 2.548, + "step": 3075 + }, + { + "epoch": 0.24661762642560783, + "grad_norm": 2.1540679931640625, + "learning_rate": 1.9546993207997104e-05, + "loss": 3.2212, + "step": 3076 + }, + { + "epoch": 0.24669780120663046, + "grad_norm": 2.087541341781616, + "learning_rate": 1.954660674578752e-05, + "loss": 3.2094, + "step": 3077 + }, + { + "epoch": 0.24677797598765308, + "grad_norm": 0.9380767941474915, + "learning_rate": 1.9546220122625506e-05, + "loss": 2.5433, + "step": 3078 + }, + { + "epoch": 0.2468581507686757, + "grad_norm": 2.214038610458374, + "learning_rate": 1.9545833338517587e-05, + "loss": 3.3538, + "step": 3079 + }, + { + "epoch": 0.24693832554969833, + "grad_norm": 1.9491291046142578, + "learning_rate": 1.954544639347028e-05, + "loss": 3.308, + "step": 3080 + }, + { + "epoch": 0.24701850033072098, + "grad_norm": 2.1372554302215576, + "learning_rate": 1.9545059287490116e-05, + "loss": 3.2977, + "step": 3081 + }, + { + "epoch": 0.2470986751117436, + "grad_norm": 2.2479679584503174, + "learning_rate": 1.9544672020583615e-05, + "loss": 3.4227, + "step": 3082 + }, + { + "epoch": 0.24717884989276623, + "grad_norm": 2.0155856609344482, + "learning_rate": 1.9544284592757307e-05, + "loss": 3.2882, + "step": 3083 + }, + { + "epoch": 0.24725902467378885, + "grad_norm": 2.031949758529663, + "learning_rate": 1.954389700401772e-05, + "loss": 3.2871, + "step": 3084 + }, + { + "epoch": 0.24733919945481148, + "grad_norm": 2.1044094562530518, + "learning_rate": 1.95435092543714e-05, + "loss": 3.2727, + "step": 3085 + }, + { + "epoch": 0.24741937423583413, + "grad_norm": 2.140598773956299, + "learning_rate": 1.9543121343824873e-05, + "loss": 3.1619, + "step": 3086 + }, + { + "epoch": 0.24749954901685675, + "grad_norm": 2.0558969974517822, + "learning_rate": 1.9542733272384686e-05, + "loss": 3.3463, + "step": 3087 + }, + { + "epoch": 0.24757972379787938, + "grad_norm": 0.9646708369255066, + "learning_rate": 1.9542345040057378e-05, + "loss": 2.5428, + "step": 3088 + }, + { + "epoch": 0.247659898578902, + "grad_norm": 2.096423864364624, + "learning_rate": 1.9541956646849497e-05, + "loss": 3.2366, + "step": 3089 + }, + { + "epoch": 0.24774007335992462, + "grad_norm": 2.214062213897705, + "learning_rate": 1.9541568092767586e-05, + "loss": 3.2337, + "step": 3090 + }, + { + "epoch": 0.24782024814094727, + "grad_norm": 2.2097692489624023, + "learning_rate": 1.9541179377818204e-05, + "loss": 3.2646, + "step": 3091 + }, + { + "epoch": 0.2479004229219699, + "grad_norm": 1.973303198814392, + "learning_rate": 1.9540790502007897e-05, + "loss": 3.2941, + "step": 3092 + }, + { + "epoch": 0.24798059770299252, + "grad_norm": 1.0571990013122559, + "learning_rate": 1.9540401465343228e-05, + "loss": 2.53, + "step": 3093 + }, + { + "epoch": 0.24806077248401515, + "grad_norm": 1.9982311725616455, + "learning_rate": 1.954001226783075e-05, + "loss": 3.2397, + "step": 3094 + }, + { + "epoch": 0.24814094726503777, + "grad_norm": 2.1261181831359863, + "learning_rate": 1.953962290947703e-05, + "loss": 3.2939, + "step": 3095 + }, + { + "epoch": 0.24822112204606042, + "grad_norm": 2.047048568725586, + "learning_rate": 1.953923339028863e-05, + "loss": 3.2775, + "step": 3096 + }, + { + "epoch": 0.24830129682708305, + "grad_norm": 2.1184892654418945, + "learning_rate": 1.9538843710272113e-05, + "loss": 3.2487, + "step": 3097 + }, + { + "epoch": 0.24838147160810567, + "grad_norm": 2.0155372619628906, + "learning_rate": 1.9538453869434052e-05, + "loss": 3.3873, + "step": 3098 + }, + { + "epoch": 0.2484616463891283, + "grad_norm": 2.0987470149993896, + "learning_rate": 1.9538063867781027e-05, + "loss": 3.3713, + "step": 3099 + }, + { + "epoch": 0.24854182117015092, + "grad_norm": 1.9519648551940918, + "learning_rate": 1.9537673705319602e-05, + "loss": 3.3628, + "step": 3100 + }, + { + "epoch": 0.24862199595117357, + "grad_norm": 2.0628597736358643, + "learning_rate": 1.953728338205636e-05, + "loss": 3.2704, + "step": 3101 + }, + { + "epoch": 0.2487021707321962, + "grad_norm": 2.114900827407837, + "learning_rate": 1.9536892897997887e-05, + "loss": 3.2672, + "step": 3102 + }, + { + "epoch": 0.24878234551321882, + "grad_norm": 2.2339022159576416, + "learning_rate": 1.9536502253150752e-05, + "loss": 3.3481, + "step": 3103 + }, + { + "epoch": 0.24886252029424144, + "grad_norm": 2.0470283031463623, + "learning_rate": 1.9536111447521554e-05, + "loss": 3.3168, + "step": 3104 + }, + { + "epoch": 0.24894269507526406, + "grad_norm": 1.9729664325714111, + "learning_rate": 1.953572048111688e-05, + "loss": 3.275, + "step": 3105 + }, + { + "epoch": 0.24902286985628672, + "grad_norm": 2.3678643703460693, + "learning_rate": 1.9535329353943317e-05, + "loss": 3.2729, + "step": 3106 + }, + { + "epoch": 0.24910304463730934, + "grad_norm": 2.0027873516082764, + "learning_rate": 1.9534938066007464e-05, + "loss": 3.2554, + "step": 3107 + }, + { + "epoch": 0.24918321941833196, + "grad_norm": 2.005751371383667, + "learning_rate": 1.953454661731591e-05, + "loss": 3.3062, + "step": 3108 + }, + { + "epoch": 0.2492633941993546, + "grad_norm": 1.9841952323913574, + "learning_rate": 1.953415500787527e-05, + "loss": 3.3202, + "step": 3109 + }, + { + "epoch": 0.2493435689803772, + "grad_norm": 1.8426015377044678, + "learning_rate": 1.9533763237692127e-05, + "loss": 3.182, + "step": 3110 + }, + { + "epoch": 0.24942374376139986, + "grad_norm": 2.1710972785949707, + "learning_rate": 1.95333713067731e-05, + "loss": 3.261, + "step": 3111 + }, + { + "epoch": 0.2495039185424225, + "grad_norm": 1.9988573789596558, + "learning_rate": 1.9532979215124793e-05, + "loss": 3.2311, + "step": 3112 + }, + { + "epoch": 0.2495840933234451, + "grad_norm": 2.2168166637420654, + "learning_rate": 1.9532586962753814e-05, + "loss": 3.246, + "step": 3113 + }, + { + "epoch": 0.24966426810446773, + "grad_norm": 2.131819248199463, + "learning_rate": 1.953219454966678e-05, + "loss": 3.1845, + "step": 3114 + }, + { + "epoch": 0.24974444288549036, + "grad_norm": 2.1353392601013184, + "learning_rate": 1.9531801975870305e-05, + "loss": 3.3151, + "step": 3115 + }, + { + "epoch": 0.249824617666513, + "grad_norm": 2.2288386821746826, + "learning_rate": 1.9531409241371012e-05, + "loss": 3.3168, + "step": 3116 + }, + { + "epoch": 0.24990479244753563, + "grad_norm": 2.0762953758239746, + "learning_rate": 1.953101634617551e-05, + "loss": 3.266, + "step": 3117 + }, + { + "epoch": 0.24998496722855826, + "grad_norm": 2.150319814682007, + "learning_rate": 1.9530623290290438e-05, + "loss": 3.3809, + "step": 3118 + }, + { + "epoch": 0.2500651420095809, + "grad_norm": 2.151093006134033, + "learning_rate": 1.9530230073722417e-05, + "loss": 3.2996, + "step": 3119 + }, + { + "epoch": 0.25014531679060353, + "grad_norm": 2.177433490753174, + "learning_rate": 1.952983669647807e-05, + "loss": 3.3427, + "step": 3120 + }, + { + "epoch": 0.25022549157162616, + "grad_norm": 2.1037745475769043, + "learning_rate": 1.952944315856404e-05, + "loss": 3.2719, + "step": 3121 + }, + { + "epoch": 0.2503056663526488, + "grad_norm": 1.8543568849563599, + "learning_rate": 1.9529049459986954e-05, + "loss": 3.2836, + "step": 3122 + }, + { + "epoch": 0.2503858411336714, + "grad_norm": 2.103740692138672, + "learning_rate": 1.9528655600753454e-05, + "loss": 3.1967, + "step": 3123 + }, + { + "epoch": 0.25046601591469403, + "grad_norm": 1.9304780960083008, + "learning_rate": 1.9528261580870178e-05, + "loss": 3.1712, + "step": 3124 + }, + { + "epoch": 0.25054619069571665, + "grad_norm": 2.218086004257202, + "learning_rate": 1.9527867400343773e-05, + "loss": 3.3408, + "step": 3125 + }, + { + "epoch": 0.2506263654767393, + "grad_norm": 2.155829429626465, + "learning_rate": 1.9527473059180876e-05, + "loss": 3.2398, + "step": 3126 + }, + { + "epoch": 0.2507065402577619, + "grad_norm": 1.8929585218429565, + "learning_rate": 1.9527078557388145e-05, + "loss": 3.3213, + "step": 3127 + }, + { + "epoch": 0.2507867150387845, + "grad_norm": 2.0600247383117676, + "learning_rate": 1.9526683894972228e-05, + "loss": 3.3943, + "step": 3128 + }, + { + "epoch": 0.2508668898198072, + "grad_norm": 2.2157223224639893, + "learning_rate": 1.952628907193977e-05, + "loss": 3.3292, + "step": 3129 + }, + { + "epoch": 0.25094706460082983, + "grad_norm": 1.8688750267028809, + "learning_rate": 1.9525894088297444e-05, + "loss": 3.2036, + "step": 3130 + }, + { + "epoch": 0.25102723938185245, + "grad_norm": 1.9213111400604248, + "learning_rate": 1.95254989440519e-05, + "loss": 3.3617, + "step": 3131 + }, + { + "epoch": 0.2511074141628751, + "grad_norm": 2.480309009552002, + "learning_rate": 1.95251036392098e-05, + "loss": 3.4231, + "step": 3132 + }, + { + "epoch": 0.2511875889438977, + "grad_norm": 2.015784502029419, + "learning_rate": 1.952470817377781e-05, + "loss": 3.2096, + "step": 3133 + }, + { + "epoch": 0.2512677637249203, + "grad_norm": 2.121203660964966, + "learning_rate": 1.9524312547762595e-05, + "loss": 3.2229, + "step": 3134 + }, + { + "epoch": 0.25134793850594295, + "grad_norm": 2.0187718868255615, + "learning_rate": 1.952391676117083e-05, + "loss": 3.1921, + "step": 3135 + }, + { + "epoch": 0.25142811328696557, + "grad_norm": 2.092148780822754, + "learning_rate": 1.952352081400918e-05, + "loss": 3.3556, + "step": 3136 + }, + { + "epoch": 0.2515082880679882, + "grad_norm": 1.0264406204223633, + "learning_rate": 1.952312470628433e-05, + "loss": 2.5624, + "step": 3137 + }, + { + "epoch": 0.2515884628490108, + "grad_norm": 2.0171360969543457, + "learning_rate": 1.952272843800295e-05, + "loss": 3.2583, + "step": 3138 + }, + { + "epoch": 0.2516686376300335, + "grad_norm": 2.2313344478607178, + "learning_rate": 1.952233200917173e-05, + "loss": 3.3774, + "step": 3139 + }, + { + "epoch": 0.2517488124110561, + "grad_norm": 2.254749298095703, + "learning_rate": 1.9521935419797344e-05, + "loss": 3.255, + "step": 3140 + }, + { + "epoch": 0.25182898719207875, + "grad_norm": 2.064992904663086, + "learning_rate": 1.9521538669886482e-05, + "loss": 3.2833, + "step": 3141 + }, + { + "epoch": 0.25190916197310137, + "grad_norm": 2.0861124992370605, + "learning_rate": 1.9521141759445838e-05, + "loss": 3.3848, + "step": 3142 + }, + { + "epoch": 0.251989336754124, + "grad_norm": 2.034432888031006, + "learning_rate": 1.95207446884821e-05, + "loss": 3.1605, + "step": 3143 + }, + { + "epoch": 0.2520695115351466, + "grad_norm": 1.9785919189453125, + "learning_rate": 1.9520347457001958e-05, + "loss": 3.3723, + "step": 3144 + }, + { + "epoch": 0.25214968631616924, + "grad_norm": 2.028207540512085, + "learning_rate": 1.9519950065012116e-05, + "loss": 3.3466, + "step": 3145 + }, + { + "epoch": 0.25222986109719187, + "grad_norm": 2.0854547023773193, + "learning_rate": 1.9519552512519268e-05, + "loss": 3.2615, + "step": 3146 + }, + { + "epoch": 0.2523100358782145, + "grad_norm": 2.031107187271118, + "learning_rate": 1.951915479953012e-05, + "loss": 3.304, + "step": 3147 + }, + { + "epoch": 0.2523902106592371, + "grad_norm": 0.9022152423858643, + "learning_rate": 1.951875692605138e-05, + "loss": 2.5371, + "step": 3148 + }, + { + "epoch": 0.2524703854402598, + "grad_norm": 1.9034452438354492, + "learning_rate": 1.951835889208975e-05, + "loss": 3.3492, + "step": 3149 + }, + { + "epoch": 0.2525505602212824, + "grad_norm": 2.1461970806121826, + "learning_rate": 1.9517960697651944e-05, + "loss": 3.2523, + "step": 3150 + }, + { + "epoch": 0.25263073500230504, + "grad_norm": 1.857839584350586, + "learning_rate": 1.9517562342744676e-05, + "loss": 3.3837, + "step": 3151 + }, + { + "epoch": 0.25271090978332766, + "grad_norm": 1.946492314338684, + "learning_rate": 1.951716382737466e-05, + "loss": 3.3384, + "step": 3152 + }, + { + "epoch": 0.2527910845643503, + "grad_norm": 0.9239338636398315, + "learning_rate": 1.9516765151548618e-05, + "loss": 2.5466, + "step": 3153 + }, + { + "epoch": 0.2528712593453729, + "grad_norm": 1.9523371458053589, + "learning_rate": 1.9516366315273268e-05, + "loss": 3.1659, + "step": 3154 + }, + { + "epoch": 0.25295143412639554, + "grad_norm": 1.921218991279602, + "learning_rate": 1.9515967318555336e-05, + "loss": 3.2329, + "step": 3155 + }, + { + "epoch": 0.25303160890741816, + "grad_norm": 2.1587536334991455, + "learning_rate": 1.951556816140155e-05, + "loss": 3.2749, + "step": 3156 + }, + { + "epoch": 0.2531117836884408, + "grad_norm": 0.8770867586135864, + "learning_rate": 1.951516884381864e-05, + "loss": 2.5162, + "step": 3157 + }, + { + "epoch": 0.2531919584694634, + "grad_norm": 2.0440421104431152, + "learning_rate": 1.9514769365813333e-05, + "loss": 3.4049, + "step": 3158 + }, + { + "epoch": 0.2532721332504861, + "grad_norm": 0.8906217813491821, + "learning_rate": 1.9514369727392372e-05, + "loss": 2.5311, + "step": 3159 + }, + { + "epoch": 0.2533523080315087, + "grad_norm": 2.073481559753418, + "learning_rate": 1.9513969928562482e-05, + "loss": 3.2318, + "step": 3160 + }, + { + "epoch": 0.25343248281253133, + "grad_norm": 2.085277557373047, + "learning_rate": 1.951356996933042e-05, + "loss": 3.3196, + "step": 3161 + }, + { + "epoch": 0.25351265759355396, + "grad_norm": 2.018026351928711, + "learning_rate": 1.951316984970292e-05, + "loss": 3.1499, + "step": 3162 + }, + { + "epoch": 0.2535928323745766, + "grad_norm": 2.185421943664551, + "learning_rate": 1.951276956968673e-05, + "loss": 3.3965, + "step": 3163 + }, + { + "epoch": 0.2536730071555992, + "grad_norm": 2.2354063987731934, + "learning_rate": 1.9512369129288597e-05, + "loss": 3.3851, + "step": 3164 + }, + { + "epoch": 0.25375318193662183, + "grad_norm": 2.082232713699341, + "learning_rate": 1.951196852851527e-05, + "loss": 3.2491, + "step": 3165 + }, + { + "epoch": 0.25383335671764445, + "grad_norm": 2.106057643890381, + "learning_rate": 1.9511567767373508e-05, + "loss": 3.3126, + "step": 3166 + }, + { + "epoch": 0.2539135314986671, + "grad_norm": 1.9727253913879395, + "learning_rate": 1.9511166845870063e-05, + "loss": 3.2273, + "step": 3167 + }, + { + "epoch": 0.2539937062796897, + "grad_norm": 1.884385108947754, + "learning_rate": 1.95107657640117e-05, + "loss": 3.175, + "step": 3168 + }, + { + "epoch": 0.2540738810607123, + "grad_norm": 2.1801328659057617, + "learning_rate": 1.9510364521805178e-05, + "loss": 3.4133, + "step": 3169 + }, + { + "epoch": 0.254154055841735, + "grad_norm": 2.124138832092285, + "learning_rate": 1.9509963119257262e-05, + "loss": 3.3494, + "step": 3170 + }, + { + "epoch": 0.25423423062275763, + "grad_norm": 1.8894789218902588, + "learning_rate": 1.950956155637472e-05, + "loss": 3.2537, + "step": 3171 + }, + { + "epoch": 0.25431440540378025, + "grad_norm": 1.0753247737884521, + "learning_rate": 1.950915983316432e-05, + "loss": 2.5521, + "step": 3172 + }, + { + "epoch": 0.2543945801848029, + "grad_norm": 1.9766608476638794, + "learning_rate": 1.9508757949632836e-05, + "loss": 3.2733, + "step": 3173 + }, + { + "epoch": 0.2544747549658255, + "grad_norm": 2.1683759689331055, + "learning_rate": 1.950835590578705e-05, + "loss": 3.2601, + "step": 3174 + }, + { + "epoch": 0.2545549297468481, + "grad_norm": 2.1976845264434814, + "learning_rate": 1.950795370163373e-05, + "loss": 3.4058, + "step": 3175 + }, + { + "epoch": 0.25463510452787075, + "grad_norm": 2.0311853885650635, + "learning_rate": 1.9507551337179658e-05, + "loss": 3.281, + "step": 3176 + }, + { + "epoch": 0.25471527930889337, + "grad_norm": 2.126213550567627, + "learning_rate": 1.9507148812431627e-05, + "loss": 3.2917, + "step": 3177 + }, + { + "epoch": 0.254795454089916, + "grad_norm": 2.0295932292938232, + "learning_rate": 1.9506746127396415e-05, + "loss": 3.301, + "step": 3178 + }, + { + "epoch": 0.2548756288709386, + "grad_norm": 1.9968270063400269, + "learning_rate": 1.9506343282080814e-05, + "loss": 3.2916, + "step": 3179 + }, + { + "epoch": 0.2549558036519613, + "grad_norm": 2.026062250137329, + "learning_rate": 1.9505940276491614e-05, + "loss": 3.2522, + "step": 3180 + }, + { + "epoch": 0.2550359784329839, + "grad_norm": 1.9882296323776245, + "learning_rate": 1.9505537110635614e-05, + "loss": 3.2422, + "step": 3181 + }, + { + "epoch": 0.25511615321400655, + "grad_norm": 2.056135892868042, + "learning_rate": 1.9505133784519606e-05, + "loss": 3.2592, + "step": 3182 + }, + { + "epoch": 0.25519632799502917, + "grad_norm": 2.1579837799072266, + "learning_rate": 1.9504730298150395e-05, + "loss": 3.2682, + "step": 3183 + }, + { + "epoch": 0.2552765027760518, + "grad_norm": 2.3648548126220703, + "learning_rate": 1.950432665153478e-05, + "loss": 3.167, + "step": 3184 + }, + { + "epoch": 0.2553566775570744, + "grad_norm": 2.1526551246643066, + "learning_rate": 1.9503922844679566e-05, + "loss": 3.4061, + "step": 3185 + }, + { + "epoch": 0.25543685233809704, + "grad_norm": 1.9567867517471313, + "learning_rate": 1.9503518877591565e-05, + "loss": 3.3634, + "step": 3186 + }, + { + "epoch": 0.25551702711911967, + "grad_norm": 2.05570650100708, + "learning_rate": 1.9503114750277585e-05, + "loss": 3.4172, + "step": 3187 + }, + { + "epoch": 0.2555972019001423, + "grad_norm": 2.278972625732422, + "learning_rate": 1.9502710462744436e-05, + "loss": 3.3807, + "step": 3188 + }, + { + "epoch": 0.2556773766811649, + "grad_norm": 0.9257228374481201, + "learning_rate": 1.950230601499894e-05, + "loss": 2.5271, + "step": 3189 + }, + { + "epoch": 0.2557575514621876, + "grad_norm": 2.0273067951202393, + "learning_rate": 1.9501901407047913e-05, + "loss": 3.314, + "step": 3190 + }, + { + "epoch": 0.2558377262432102, + "grad_norm": 2.0895509719848633, + "learning_rate": 1.9501496638898177e-05, + "loss": 3.2109, + "step": 3191 + }, + { + "epoch": 0.25591790102423284, + "grad_norm": 2.032381772994995, + "learning_rate": 1.9501091710556553e-05, + "loss": 3.304, + "step": 3192 + }, + { + "epoch": 0.25599807580525547, + "grad_norm": 1.9635266065597534, + "learning_rate": 1.9500686622029876e-05, + "loss": 3.1676, + "step": 3193 + }, + { + "epoch": 0.2560782505862781, + "grad_norm": 2.065472364425659, + "learning_rate": 1.9500281373324972e-05, + "loss": 3.2513, + "step": 3194 + }, + { + "epoch": 0.2561584253673007, + "grad_norm": 2.1489098072052, + "learning_rate": 1.949987596444867e-05, + "loss": 3.1966, + "step": 3195 + }, + { + "epoch": 0.25623860014832334, + "grad_norm": 2.1003170013427734, + "learning_rate": 1.9499470395407804e-05, + "loss": 3.3522, + "step": 3196 + }, + { + "epoch": 0.25631877492934596, + "grad_norm": 2.046820878982544, + "learning_rate": 1.949906466620922e-05, + "loss": 3.2307, + "step": 3197 + }, + { + "epoch": 0.2563989497103686, + "grad_norm": 2.0183839797973633, + "learning_rate": 1.9498658776859753e-05, + "loss": 3.2757, + "step": 3198 + }, + { + "epoch": 0.2564791244913912, + "grad_norm": 1.9205706119537354, + "learning_rate": 1.9498252727366245e-05, + "loss": 3.2657, + "step": 3199 + }, + { + "epoch": 0.2565592992724139, + "grad_norm": 0.9572913646697998, + "learning_rate": 1.9497846517735544e-05, + "loss": 2.5406, + "step": 3200 + }, + { + "epoch": 0.2566394740534365, + "grad_norm": 1.958573579788208, + "learning_rate": 1.9497440147974498e-05, + "loss": 3.2936, + "step": 3201 + }, + { + "epoch": 0.25671964883445914, + "grad_norm": 2.2103848457336426, + "learning_rate": 1.9497033618089957e-05, + "loss": 3.308, + "step": 3202 + }, + { + "epoch": 0.25679982361548176, + "grad_norm": 2.0293354988098145, + "learning_rate": 1.949662692808878e-05, + "loss": 3.1669, + "step": 3203 + }, + { + "epoch": 0.2568799983965044, + "grad_norm": 1.9944292306900024, + "learning_rate": 1.9496220077977815e-05, + "loss": 3.2891, + "step": 3204 + }, + { + "epoch": 0.256960173177527, + "grad_norm": 2.098160743713379, + "learning_rate": 1.9495813067763933e-05, + "loss": 3.349, + "step": 3205 + }, + { + "epoch": 0.25704034795854963, + "grad_norm": 1.9572306871414185, + "learning_rate": 1.9495405897453983e-05, + "loss": 3.2519, + "step": 3206 + }, + { + "epoch": 0.25712052273957225, + "grad_norm": 1.957535743713379, + "learning_rate": 1.949499856705484e-05, + "loss": 3.2125, + "step": 3207 + }, + { + "epoch": 0.2572006975205949, + "grad_norm": 2.027087688446045, + "learning_rate": 1.9494591076573368e-05, + "loss": 3.2509, + "step": 3208 + }, + { + "epoch": 0.2572808723016175, + "grad_norm": 2.0553855895996094, + "learning_rate": 1.949418342601643e-05, + "loss": 3.2948, + "step": 3209 + }, + { + "epoch": 0.2573610470826402, + "grad_norm": 2.0535507202148438, + "learning_rate": 1.949377561539091e-05, + "loss": 3.1917, + "step": 3210 + }, + { + "epoch": 0.2574412218636628, + "grad_norm": 1.9368243217468262, + "learning_rate": 1.949336764470368e-05, + "loss": 3.313, + "step": 3211 + }, + { + "epoch": 0.25752139664468543, + "grad_norm": 2.033886194229126, + "learning_rate": 1.9492959513961617e-05, + "loss": 3.2643, + "step": 3212 + }, + { + "epoch": 0.25760157142570805, + "grad_norm": 2.0434935092926025, + "learning_rate": 1.9492551223171602e-05, + "loss": 3.275, + "step": 3213 + }, + { + "epoch": 0.2576817462067307, + "grad_norm": 1.9565790891647339, + "learning_rate": 1.949214277234052e-05, + "loss": 3.3416, + "step": 3214 + }, + { + "epoch": 0.2577619209877533, + "grad_norm": 2.154085636138916, + "learning_rate": 1.9491734161475252e-05, + "loss": 3.3612, + "step": 3215 + }, + { + "epoch": 0.2578420957687759, + "grad_norm": 2.3204760551452637, + "learning_rate": 1.9491325390582693e-05, + "loss": 3.2623, + "step": 3216 + }, + { + "epoch": 0.25792227054979855, + "grad_norm": 2.1209771633148193, + "learning_rate": 1.9490916459669734e-05, + "loss": 3.3158, + "step": 3217 + }, + { + "epoch": 0.2580024453308212, + "grad_norm": 1.9715100526809692, + "learning_rate": 1.949050736874327e-05, + "loss": 3.2542, + "step": 3218 + }, + { + "epoch": 0.2580826201118438, + "grad_norm": 2.001354455947876, + "learning_rate": 1.9490098117810197e-05, + "loss": 3.2799, + "step": 3219 + }, + { + "epoch": 0.2581627948928665, + "grad_norm": 1.932990312576294, + "learning_rate": 1.948968870687741e-05, + "loss": 3.2494, + "step": 3220 + }, + { + "epoch": 0.2582429696738891, + "grad_norm": 1.9936939477920532, + "learning_rate": 1.9489279135951816e-05, + "loss": 3.2596, + "step": 3221 + }, + { + "epoch": 0.2583231444549117, + "grad_norm": 0.9547784328460693, + "learning_rate": 1.9488869405040324e-05, + "loss": 2.5265, + "step": 3222 + }, + { + "epoch": 0.25840331923593435, + "grad_norm": 0.9915238618850708, + "learning_rate": 1.948845951414984e-05, + "loss": 2.5476, + "step": 3223 + }, + { + "epoch": 0.25848349401695697, + "grad_norm": 1.966694951057434, + "learning_rate": 1.9488049463287267e-05, + "loss": 3.3293, + "step": 3224 + }, + { + "epoch": 0.2585636687979796, + "grad_norm": 1.89110267162323, + "learning_rate": 1.9487639252459525e-05, + "loss": 3.2582, + "step": 3225 + }, + { + "epoch": 0.2586438435790022, + "grad_norm": 2.043586254119873, + "learning_rate": 1.948722888167353e-05, + "loss": 3.2981, + "step": 3226 + }, + { + "epoch": 0.25872401836002484, + "grad_norm": 2.0053441524505615, + "learning_rate": 1.94868183509362e-05, + "loss": 3.2943, + "step": 3227 + }, + { + "epoch": 0.25880419314104747, + "grad_norm": 1.967488408088684, + "learning_rate": 1.9486407660254457e-05, + "loss": 3.2224, + "step": 3228 + }, + { + "epoch": 0.2588843679220701, + "grad_norm": 2.016645908355713, + "learning_rate": 1.948599680963522e-05, + "loss": 3.3032, + "step": 3229 + }, + { + "epoch": 0.2589645427030927, + "grad_norm": 1.8769257068634033, + "learning_rate": 1.9485585799085425e-05, + "loss": 3.2339, + "step": 3230 + }, + { + "epoch": 0.2590447174841154, + "grad_norm": 2.240509510040283, + "learning_rate": 1.9485174628611993e-05, + "loss": 3.2884, + "step": 3231 + }, + { + "epoch": 0.259124892265138, + "grad_norm": 2.0177090167999268, + "learning_rate": 1.948476329822186e-05, + "loss": 3.1778, + "step": 3232 + }, + { + "epoch": 0.25920506704616064, + "grad_norm": 2.01212739944458, + "learning_rate": 1.9484351807921965e-05, + "loss": 3.2753, + "step": 3233 + }, + { + "epoch": 0.25928524182718327, + "grad_norm": 1.9687937498092651, + "learning_rate": 1.9483940157719233e-05, + "loss": 3.1899, + "step": 3234 + }, + { + "epoch": 0.2593654166082059, + "grad_norm": 2.0847604274749756, + "learning_rate": 1.948352834762062e-05, + "loss": 3.3707, + "step": 3235 + }, + { + "epoch": 0.2594455913892285, + "grad_norm": 2.032503366470337, + "learning_rate": 1.948311637763306e-05, + "loss": 3.2686, + "step": 3236 + }, + { + "epoch": 0.25952576617025114, + "grad_norm": 1.8376723527908325, + "learning_rate": 1.9482704247763497e-05, + "loss": 3.2535, + "step": 3237 + }, + { + "epoch": 0.25960594095127376, + "grad_norm": 2.175421953201294, + "learning_rate": 1.9482291958018885e-05, + "loss": 3.4122, + "step": 3238 + }, + { + "epoch": 0.2596861157322964, + "grad_norm": 2.0743331909179688, + "learning_rate": 1.948187950840617e-05, + "loss": 3.1743, + "step": 3239 + }, + { + "epoch": 0.259766290513319, + "grad_norm": 1.8938519954681396, + "learning_rate": 1.948146689893231e-05, + "loss": 3.2592, + "step": 3240 + }, + { + "epoch": 0.2598464652943417, + "grad_norm": 2.052457094192505, + "learning_rate": 1.948105412960426e-05, + "loss": 3.2357, + "step": 3241 + }, + { + "epoch": 0.2599266400753643, + "grad_norm": 2.000063896179199, + "learning_rate": 1.9480641200428975e-05, + "loss": 3.287, + "step": 3242 + }, + { + "epoch": 0.26000681485638694, + "grad_norm": 1.9330657720565796, + "learning_rate": 1.9480228111413425e-05, + "loss": 3.3492, + "step": 3243 + }, + { + "epoch": 0.26008698963740956, + "grad_norm": 0.9836860299110413, + "learning_rate": 1.947981486256457e-05, + "loss": 2.5306, + "step": 3244 + }, + { + "epoch": 0.2601671644184322, + "grad_norm": 2.134988784790039, + "learning_rate": 1.9479401453889376e-05, + "loss": 3.4268, + "step": 3245 + }, + { + "epoch": 0.2602473391994548, + "grad_norm": 1.9559845924377441, + "learning_rate": 1.9478987885394814e-05, + "loss": 3.316, + "step": 3246 + }, + { + "epoch": 0.26032751398047743, + "grad_norm": 2.0783851146698, + "learning_rate": 1.947857415708786e-05, + "loss": 3.2119, + "step": 3247 + }, + { + "epoch": 0.26040768876150006, + "grad_norm": 1.9763163328170776, + "learning_rate": 1.9478160268975485e-05, + "loss": 3.2471, + "step": 3248 + }, + { + "epoch": 0.2604878635425227, + "grad_norm": 2.069300889968872, + "learning_rate": 1.9477746221064663e-05, + "loss": 3.3284, + "step": 3249 + }, + { + "epoch": 0.2605680383235453, + "grad_norm": 2.0687143802642822, + "learning_rate": 1.9477332013362387e-05, + "loss": 3.2288, + "step": 3250 + }, + { + "epoch": 0.260648213104568, + "grad_norm": 2.0618951320648193, + "learning_rate": 1.947691764587563e-05, + "loss": 3.2395, + "step": 3251 + }, + { + "epoch": 0.2607283878855906, + "grad_norm": 2.1091432571411133, + "learning_rate": 1.9476503118611383e-05, + "loss": 3.3361, + "step": 3252 + }, + { + "epoch": 0.26080856266661323, + "grad_norm": 1.95653235912323, + "learning_rate": 1.947608843157663e-05, + "loss": 3.1962, + "step": 3253 + }, + { + "epoch": 0.26088873744763585, + "grad_norm": 2.225553512573242, + "learning_rate": 1.947567358477837e-05, + "loss": 3.1937, + "step": 3254 + }, + { + "epoch": 0.2609689122286585, + "grad_norm": 1.9231805801391602, + "learning_rate": 1.9475258578223594e-05, + "loss": 3.2482, + "step": 3255 + }, + { + "epoch": 0.2610490870096811, + "grad_norm": 1.9757078886032104, + "learning_rate": 1.9474843411919295e-05, + "loss": 3.2726, + "step": 3256 + }, + { + "epoch": 0.2611292617907037, + "grad_norm": 2.011930465698242, + "learning_rate": 1.9474428085872472e-05, + "loss": 3.2228, + "step": 3257 + }, + { + "epoch": 0.26120943657172635, + "grad_norm": 1.9675387144088745, + "learning_rate": 1.9474012600090134e-05, + "loss": 3.311, + "step": 3258 + }, + { + "epoch": 0.261289611352749, + "grad_norm": 2.0969676971435547, + "learning_rate": 1.9473596954579282e-05, + "loss": 3.3555, + "step": 3259 + }, + { + "epoch": 0.2613697861337716, + "grad_norm": 2.1254019737243652, + "learning_rate": 1.9473181149346925e-05, + "loss": 3.2196, + "step": 3260 + }, + { + "epoch": 0.2614499609147943, + "grad_norm": 2.143796682357788, + "learning_rate": 1.947276518440007e-05, + "loss": 3.3043, + "step": 3261 + }, + { + "epoch": 0.2615301356958169, + "grad_norm": 2.3426172733306885, + "learning_rate": 1.9472349059745737e-05, + "loss": 3.2468, + "step": 3262 + }, + { + "epoch": 0.2616103104768395, + "grad_norm": 2.120041847229004, + "learning_rate": 1.9471932775390934e-05, + "loss": 3.3125, + "step": 3263 + }, + { + "epoch": 0.26169048525786215, + "grad_norm": 2.0162439346313477, + "learning_rate": 1.9471516331342682e-05, + "loss": 3.2272, + "step": 3264 + }, + { + "epoch": 0.2617706600388848, + "grad_norm": 0.9274709820747375, + "learning_rate": 1.9471099727608008e-05, + "loss": 2.5272, + "step": 3265 + }, + { + "epoch": 0.2618508348199074, + "grad_norm": 2.0680959224700928, + "learning_rate": 1.9470682964193925e-05, + "loss": 3.2808, + "step": 3266 + }, + { + "epoch": 0.26193100960093, + "grad_norm": 1.8746289014816284, + "learning_rate": 1.9470266041107467e-05, + "loss": 3.2878, + "step": 3267 + }, + { + "epoch": 0.26201118438195264, + "grad_norm": 2.2737643718719482, + "learning_rate": 1.946984895835566e-05, + "loss": 3.239, + "step": 3268 + }, + { + "epoch": 0.26209135916297527, + "grad_norm": 2.041003704071045, + "learning_rate": 1.9469431715945537e-05, + "loss": 3.2954, + "step": 3269 + }, + { + "epoch": 0.2621715339439979, + "grad_norm": 2.0235538482666016, + "learning_rate": 1.9469014313884133e-05, + "loss": 3.3922, + "step": 3270 + }, + { + "epoch": 0.26225170872502057, + "grad_norm": 1.9470138549804688, + "learning_rate": 1.946859675217849e-05, + "loss": 3.2493, + "step": 3271 + }, + { + "epoch": 0.2623318835060432, + "grad_norm": 1.925431489944458, + "learning_rate": 1.9468179030835635e-05, + "loss": 3.325, + "step": 3272 + }, + { + "epoch": 0.2624120582870658, + "grad_norm": 1.8381977081298828, + "learning_rate": 1.9467761149862622e-05, + "loss": 3.2906, + "step": 3273 + }, + { + "epoch": 0.26249223306808844, + "grad_norm": 2.1000895500183105, + "learning_rate": 1.946734310926649e-05, + "loss": 3.371, + "step": 3274 + }, + { + "epoch": 0.26257240784911107, + "grad_norm": 1.9309800863265991, + "learning_rate": 1.9466924909054296e-05, + "loss": 3.2207, + "step": 3275 + }, + { + "epoch": 0.2626525826301337, + "grad_norm": 2.1417248249053955, + "learning_rate": 1.9466506549233077e-05, + "loss": 3.32, + "step": 3276 + }, + { + "epoch": 0.2627327574111563, + "grad_norm": 1.9590039253234863, + "learning_rate": 1.94660880298099e-05, + "loss": 3.2395, + "step": 3277 + }, + { + "epoch": 0.26281293219217894, + "grad_norm": 2.051755428314209, + "learning_rate": 1.946566935079181e-05, + "loss": 3.2596, + "step": 3278 + }, + { + "epoch": 0.26289310697320156, + "grad_norm": 2.081812620162964, + "learning_rate": 1.9465250512185876e-05, + "loss": 3.3849, + "step": 3279 + }, + { + "epoch": 0.2629732817542242, + "grad_norm": 1.9612772464752197, + "learning_rate": 1.946483151399915e-05, + "loss": 3.2321, + "step": 3280 + }, + { + "epoch": 0.26305345653524687, + "grad_norm": 1.9884487390518188, + "learning_rate": 1.9464412356238706e-05, + "loss": 3.3134, + "step": 3281 + }, + { + "epoch": 0.2631336313162695, + "grad_norm": 1.8726541996002197, + "learning_rate": 1.94639930389116e-05, + "loss": 3.151, + "step": 3282 + }, + { + "epoch": 0.2632138060972921, + "grad_norm": 2.198272228240967, + "learning_rate": 1.9463573562024908e-05, + "loss": 3.2933, + "step": 3283 + }, + { + "epoch": 0.26329398087831474, + "grad_norm": 1.9254552125930786, + "learning_rate": 1.94631539255857e-05, + "loss": 3.2729, + "step": 3284 + }, + { + "epoch": 0.26337415565933736, + "grad_norm": 2.0393664836883545, + "learning_rate": 1.9462734129601054e-05, + "loss": 3.1627, + "step": 3285 + }, + { + "epoch": 0.26345433044036, + "grad_norm": 1.991904616355896, + "learning_rate": 1.9462314174078044e-05, + "loss": 3.2826, + "step": 3286 + }, + { + "epoch": 0.2635345052213826, + "grad_norm": 1.9198484420776367, + "learning_rate": 1.946189405902375e-05, + "loss": 3.202, + "step": 3287 + }, + { + "epoch": 0.26361468000240523, + "grad_norm": 2.2636373043060303, + "learning_rate": 1.946147378444526e-05, + "loss": 3.2561, + "step": 3288 + }, + { + "epoch": 0.26369485478342786, + "grad_norm": 1.845713496208191, + "learning_rate": 1.9461053350349656e-05, + "loss": 3.2781, + "step": 3289 + }, + { + "epoch": 0.2637750295644505, + "grad_norm": 2.0206425189971924, + "learning_rate": 1.9460632756744028e-05, + "loss": 3.1821, + "step": 3290 + }, + { + "epoch": 0.2638552043454731, + "grad_norm": 2.3294246196746826, + "learning_rate": 1.9460212003635462e-05, + "loss": 3.2882, + "step": 3291 + }, + { + "epoch": 0.2639353791264958, + "grad_norm": 2.152379035949707, + "learning_rate": 1.9459791091031063e-05, + "loss": 3.2566, + "step": 3292 + }, + { + "epoch": 0.2640155539075184, + "grad_norm": 2.301854133605957, + "learning_rate": 1.9459370018937912e-05, + "loss": 3.1826, + "step": 3293 + }, + { + "epoch": 0.26409572868854103, + "grad_norm": 1.0172009468078613, + "learning_rate": 1.9458948787363124e-05, + "loss": 2.5304, + "step": 3294 + }, + { + "epoch": 0.26417590346956366, + "grad_norm": 1.0006318092346191, + "learning_rate": 1.9458527396313788e-05, + "loss": 2.5344, + "step": 3295 + }, + { + "epoch": 0.2642560782505863, + "grad_norm": 2.058570623397827, + "learning_rate": 1.9458105845797014e-05, + "loss": 3.3002, + "step": 3296 + }, + { + "epoch": 0.2643362530316089, + "grad_norm": 2.1408138275146484, + "learning_rate": 1.9457684135819908e-05, + "loss": 3.3176, + "step": 3297 + }, + { + "epoch": 0.2644164278126315, + "grad_norm": 2.071549892425537, + "learning_rate": 1.9457262266389582e-05, + "loss": 3.2455, + "step": 3298 + }, + { + "epoch": 0.26449660259365415, + "grad_norm": 1.9879429340362549, + "learning_rate": 1.9456840237513147e-05, + "loss": 3.2874, + "step": 3299 + }, + { + "epoch": 0.2645767773746768, + "grad_norm": 2.073573350906372, + "learning_rate": 1.945641804919772e-05, + "loss": 3.1597, + "step": 3300 + }, + { + "epoch": 0.2646569521556994, + "grad_norm": 1.2196568250656128, + "learning_rate": 1.9455995701450418e-05, + "loss": 2.5492, + "step": 3301 + }, + { + "epoch": 0.2647371269367221, + "grad_norm": 1.9811162948608398, + "learning_rate": 1.945557319427836e-05, + "loss": 3.2504, + "step": 3302 + }, + { + "epoch": 0.2648173017177447, + "grad_norm": 1.0528534650802612, + "learning_rate": 1.9455150527688668e-05, + "loss": 2.5437, + "step": 3303 + }, + { + "epoch": 0.2648974764987673, + "grad_norm": 2.0797884464263916, + "learning_rate": 1.9454727701688472e-05, + "loss": 3.2635, + "step": 3304 + }, + { + "epoch": 0.26497765127978995, + "grad_norm": 1.9812990427017212, + "learning_rate": 1.94543047162849e-05, + "loss": 3.2428, + "step": 3305 + }, + { + "epoch": 0.2650578260608126, + "grad_norm": 1.9734281301498413, + "learning_rate": 1.945388157148508e-05, + "loss": 3.3191, + "step": 3306 + }, + { + "epoch": 0.2651380008418352, + "grad_norm": 2.120250940322876, + "learning_rate": 1.9453458267296152e-05, + "loss": 3.1556, + "step": 3307 + }, + { + "epoch": 0.2652181756228578, + "grad_norm": 2.079688310623169, + "learning_rate": 1.945303480372525e-05, + "loss": 3.2682, + "step": 3308 + }, + { + "epoch": 0.26529835040388045, + "grad_norm": 2.0890917778015137, + "learning_rate": 1.9452611180779512e-05, + "loss": 3.254, + "step": 3309 + }, + { + "epoch": 0.26537852518490307, + "grad_norm": 1.9334325790405273, + "learning_rate": 1.9452187398466075e-05, + "loss": 3.3052, + "step": 3310 + }, + { + "epoch": 0.2654586999659257, + "grad_norm": 0.9965280294418335, + "learning_rate": 1.9451763456792093e-05, + "loss": 2.5599, + "step": 3311 + }, + { + "epoch": 0.2655388747469484, + "grad_norm": 2.0220086574554443, + "learning_rate": 1.9451339355764715e-05, + "loss": 3.2655, + "step": 3312 + }, + { + "epoch": 0.265619049527971, + "grad_norm": 2.1034998893737793, + "learning_rate": 1.945091509539108e-05, + "loss": 3.3337, + "step": 3313 + }, + { + "epoch": 0.2656992243089936, + "grad_norm": 2.0642261505126953, + "learning_rate": 1.945049067567835e-05, + "loss": 3.2596, + "step": 3314 + }, + { + "epoch": 0.26577939909001624, + "grad_norm": 1.973639965057373, + "learning_rate": 1.945006609663368e-05, + "loss": 3.289, + "step": 3315 + }, + { + "epoch": 0.26585957387103887, + "grad_norm": 2.0268967151641846, + "learning_rate": 1.9449641358264223e-05, + "loss": 3.1863, + "step": 3316 + }, + { + "epoch": 0.2659397486520615, + "grad_norm": 2.1861205101013184, + "learning_rate": 1.9449216460577144e-05, + "loss": 3.2891, + "step": 3317 + }, + { + "epoch": 0.2660199234330841, + "grad_norm": 2.080219268798828, + "learning_rate": 1.9448791403579606e-05, + "loss": 3.28, + "step": 3318 + }, + { + "epoch": 0.26610009821410674, + "grad_norm": 2.3109002113342285, + "learning_rate": 1.9448366187278777e-05, + "loss": 3.2964, + "step": 3319 + }, + { + "epoch": 0.26618027299512936, + "grad_norm": 2.1247501373291016, + "learning_rate": 1.9447940811681824e-05, + "loss": 3.3426, + "step": 3320 + }, + { + "epoch": 0.266260447776152, + "grad_norm": 2.124779224395752, + "learning_rate": 1.9447515276795914e-05, + "loss": 3.3273, + "step": 3321 + }, + { + "epoch": 0.26634062255717467, + "grad_norm": 2.142869234085083, + "learning_rate": 1.944708958262823e-05, + "loss": 3.2706, + "step": 3322 + }, + { + "epoch": 0.2664207973381973, + "grad_norm": 1.9680836200714111, + "learning_rate": 1.9446663729185945e-05, + "loss": 3.318, + "step": 3323 + }, + { + "epoch": 0.2665009721192199, + "grad_norm": 1.9263081550598145, + "learning_rate": 1.9446237716476237e-05, + "loss": 3.2554, + "step": 3324 + }, + { + "epoch": 0.26658114690024254, + "grad_norm": 2.063762664794922, + "learning_rate": 1.944581154450629e-05, + "loss": 3.2724, + "step": 3325 + }, + { + "epoch": 0.26666132168126516, + "grad_norm": 1.855160117149353, + "learning_rate": 1.9445385213283293e-05, + "loss": 3.2747, + "step": 3326 + }, + { + "epoch": 0.2667414964622878, + "grad_norm": 1.9199203252792358, + "learning_rate": 1.9444958722814427e-05, + "loss": 3.2642, + "step": 3327 + }, + { + "epoch": 0.2668216712433104, + "grad_norm": 2.0885839462280273, + "learning_rate": 1.9444532073106886e-05, + "loss": 3.3412, + "step": 3328 + }, + { + "epoch": 0.26690184602433303, + "grad_norm": 1.8966991901397705, + "learning_rate": 1.9444105264167864e-05, + "loss": 3.2057, + "step": 3329 + }, + { + "epoch": 0.26698202080535566, + "grad_norm": 2.0469701290130615, + "learning_rate": 1.9443678296004552e-05, + "loss": 3.1348, + "step": 3330 + }, + { + "epoch": 0.2670621955863783, + "grad_norm": 1.9137179851531982, + "learning_rate": 1.9443251168624155e-05, + "loss": 3.2115, + "step": 3331 + }, + { + "epoch": 0.26714237036740096, + "grad_norm": 1.9244226217269897, + "learning_rate": 1.9442823882033872e-05, + "loss": 3.2254, + "step": 3332 + }, + { + "epoch": 0.2672225451484236, + "grad_norm": 2.106968402862549, + "learning_rate": 1.9442396436240905e-05, + "loss": 3.2719, + "step": 3333 + }, + { + "epoch": 0.2673027199294462, + "grad_norm": 1.9336016178131104, + "learning_rate": 1.9441968831252463e-05, + "loss": 3.1442, + "step": 3334 + }, + { + "epoch": 0.26738289471046883, + "grad_norm": 2.120225667953491, + "learning_rate": 1.9441541067075755e-05, + "loss": 3.2042, + "step": 3335 + }, + { + "epoch": 0.26746306949149146, + "grad_norm": 1.944338321685791, + "learning_rate": 1.944111314371799e-05, + "loss": 3.2848, + "step": 3336 + }, + { + "epoch": 0.2675432442725141, + "grad_norm": 1.8522201776504517, + "learning_rate": 1.9440685061186388e-05, + "loss": 3.324, + "step": 3337 + }, + { + "epoch": 0.2676234190535367, + "grad_norm": 1.9649407863616943, + "learning_rate": 1.9440256819488158e-05, + "loss": 3.3408, + "step": 3338 + }, + { + "epoch": 0.26770359383455933, + "grad_norm": 2.0518925189971924, + "learning_rate": 1.943982841863053e-05, + "loss": 3.2968, + "step": 3339 + }, + { + "epoch": 0.26778376861558195, + "grad_norm": 0.8983094692230225, + "learning_rate": 1.943939985862072e-05, + "loss": 2.5216, + "step": 3340 + }, + { + "epoch": 0.2678639433966046, + "grad_norm": 2.0637078285217285, + "learning_rate": 1.9438971139465952e-05, + "loss": 3.3171, + "step": 3341 + }, + { + "epoch": 0.2679441181776272, + "grad_norm": 2.0000598430633545, + "learning_rate": 1.9438542261173463e-05, + "loss": 3.2837, + "step": 3342 + }, + { + "epoch": 0.2680242929586499, + "grad_norm": 1.8247706890106201, + "learning_rate": 1.9438113223750473e-05, + "loss": 3.2404, + "step": 3343 + }, + { + "epoch": 0.2681044677396725, + "grad_norm": 2.093230724334717, + "learning_rate": 1.9437684027204223e-05, + "loss": 3.2114, + "step": 3344 + }, + { + "epoch": 0.2681846425206951, + "grad_norm": 2.167128801345825, + "learning_rate": 1.9437254671541946e-05, + "loss": 3.1797, + "step": 3345 + }, + { + "epoch": 0.26826481730171775, + "grad_norm": 2.0799195766448975, + "learning_rate": 1.9436825156770882e-05, + "loss": 3.2847, + "step": 3346 + }, + { + "epoch": 0.2683449920827404, + "grad_norm": 1.9165552854537964, + "learning_rate": 1.943639548289827e-05, + "loss": 3.2101, + "step": 3347 + }, + { + "epoch": 0.268425166863763, + "grad_norm": 2.0025646686553955, + "learning_rate": 1.9435965649931358e-05, + "loss": 3.3033, + "step": 3348 + }, + { + "epoch": 0.2685053416447856, + "grad_norm": 1.9173903465270996, + "learning_rate": 1.9435535657877388e-05, + "loss": 3.1672, + "step": 3349 + }, + { + "epoch": 0.26858551642580825, + "grad_norm": 2.0954065322875977, + "learning_rate": 1.9435105506743614e-05, + "loss": 3.2568, + "step": 3350 + }, + { + "epoch": 0.26866569120683087, + "grad_norm": 2.1524600982666016, + "learning_rate": 1.9434675196537286e-05, + "loss": 3.3301, + "step": 3351 + }, + { + "epoch": 0.2687458659878535, + "grad_norm": 1.9364001750946045, + "learning_rate": 1.9434244727265658e-05, + "loss": 3.2819, + "step": 3352 + }, + { + "epoch": 0.2688260407688762, + "grad_norm": 1.9916373491287231, + "learning_rate": 1.943381409893599e-05, + "loss": 3.306, + "step": 3353 + }, + { + "epoch": 0.2689062155498988, + "grad_norm": 1.9957356452941895, + "learning_rate": 1.943338331155554e-05, + "loss": 3.2926, + "step": 3354 + }, + { + "epoch": 0.2689863903309214, + "grad_norm": 2.337214708328247, + "learning_rate": 1.9432952365131574e-05, + "loss": 3.2573, + "step": 3355 + }, + { + "epoch": 0.26906656511194404, + "grad_norm": 2.143495798110962, + "learning_rate": 1.9432521259671355e-05, + "loss": 3.2269, + "step": 3356 + }, + { + "epoch": 0.26914673989296667, + "grad_norm": 2.318575859069824, + "learning_rate": 1.9432089995182154e-05, + "loss": 3.3931, + "step": 3357 + }, + { + "epoch": 0.2692269146739893, + "grad_norm": 2.015371561050415, + "learning_rate": 1.9431658571671238e-05, + "loss": 3.2966, + "step": 3358 + }, + { + "epoch": 0.2693070894550119, + "grad_norm": 1.985512614250183, + "learning_rate": 1.9431226989145878e-05, + "loss": 3.2444, + "step": 3359 + }, + { + "epoch": 0.26938726423603454, + "grad_norm": 2.0800671577453613, + "learning_rate": 1.9430795247613362e-05, + "loss": 3.3425, + "step": 3360 + }, + { + "epoch": 0.26946743901705716, + "grad_norm": 2.0659773349761963, + "learning_rate": 1.9430363347080957e-05, + "loss": 3.3333, + "step": 3361 + }, + { + "epoch": 0.2695476137980798, + "grad_norm": 2.080321788787842, + "learning_rate": 1.9429931287555952e-05, + "loss": 3.3175, + "step": 3362 + }, + { + "epoch": 0.26962778857910247, + "grad_norm": 2.0431699752807617, + "learning_rate": 1.9429499069045628e-05, + "loss": 3.2339, + "step": 3363 + }, + { + "epoch": 0.2697079633601251, + "grad_norm": 1.991603970527649, + "learning_rate": 1.9429066691557272e-05, + "loss": 3.2429, + "step": 3364 + }, + { + "epoch": 0.2697881381411477, + "grad_norm": 2.0283210277557373, + "learning_rate": 1.942863415509818e-05, + "loss": 3.2275, + "step": 3365 + }, + { + "epoch": 0.26986831292217034, + "grad_norm": 2.141178607940674, + "learning_rate": 1.9428201459675633e-05, + "loss": 3.2461, + "step": 3366 + }, + { + "epoch": 0.26994848770319296, + "grad_norm": 1.8761247396469116, + "learning_rate": 1.9427768605296933e-05, + "loss": 3.2302, + "step": 3367 + }, + { + "epoch": 0.2700286624842156, + "grad_norm": 1.9518758058547974, + "learning_rate": 1.9427335591969377e-05, + "loss": 3.3347, + "step": 3368 + }, + { + "epoch": 0.2701088372652382, + "grad_norm": 2.062553644180298, + "learning_rate": 1.9426902419700272e-05, + "loss": 3.2739, + "step": 3369 + }, + { + "epoch": 0.27018901204626083, + "grad_norm": 1.0158344507217407, + "learning_rate": 1.9426469088496908e-05, + "loss": 2.5173, + "step": 3370 + }, + { + "epoch": 0.27026918682728346, + "grad_norm": 2.0513603687286377, + "learning_rate": 1.9426035598366597e-05, + "loss": 3.304, + "step": 3371 + }, + { + "epoch": 0.2703493616083061, + "grad_norm": 2.0913586616516113, + "learning_rate": 1.942560194931665e-05, + "loss": 3.2501, + "step": 3372 + }, + { + "epoch": 0.27042953638932876, + "grad_norm": 0.9028023481369019, + "learning_rate": 1.9425168141354377e-05, + "loss": 2.5034, + "step": 3373 + }, + { + "epoch": 0.2705097111703514, + "grad_norm": 0.8841831684112549, + "learning_rate": 1.9424734174487087e-05, + "loss": 2.52, + "step": 3374 + }, + { + "epoch": 0.270589885951374, + "grad_norm": 2.1381967067718506, + "learning_rate": 1.9424300048722104e-05, + "loss": 3.4413, + "step": 3375 + }, + { + "epoch": 0.27067006073239663, + "grad_norm": 2.202509641647339, + "learning_rate": 1.9423865764066744e-05, + "loss": 3.2971, + "step": 3376 + }, + { + "epoch": 0.27075023551341926, + "grad_norm": 2.042388916015625, + "learning_rate": 1.9423431320528325e-05, + "loss": 3.2773, + "step": 3377 + }, + { + "epoch": 0.2708304102944419, + "grad_norm": 1.9632712602615356, + "learning_rate": 1.9422996718114176e-05, + "loss": 3.2433, + "step": 3378 + }, + { + "epoch": 0.2709105850754645, + "grad_norm": 2.008457899093628, + "learning_rate": 1.942256195683163e-05, + "loss": 3.2235, + "step": 3379 + }, + { + "epoch": 0.27099075985648713, + "grad_norm": 1.8693705797195435, + "learning_rate": 1.9422127036688e-05, + "loss": 3.2864, + "step": 3380 + }, + { + "epoch": 0.27107093463750975, + "grad_norm": 1.0257713794708252, + "learning_rate": 1.9421691957690635e-05, + "loss": 2.5468, + "step": 3381 + }, + { + "epoch": 0.2711511094185324, + "grad_norm": 2.0585906505584717, + "learning_rate": 1.942125671984686e-05, + "loss": 3.1909, + "step": 3382 + }, + { + "epoch": 0.27123128419955506, + "grad_norm": 1.9464991092681885, + "learning_rate": 1.942082132316402e-05, + "loss": 3.3309, + "step": 3383 + }, + { + "epoch": 0.2713114589805777, + "grad_norm": 1.9715230464935303, + "learning_rate": 1.9420385767649453e-05, + "loss": 3.2141, + "step": 3384 + }, + { + "epoch": 0.2713916337616003, + "grad_norm": 2.0234127044677734, + "learning_rate": 1.9419950053310502e-05, + "loss": 3.3046, + "step": 3385 + }, + { + "epoch": 0.2714718085426229, + "grad_norm": 2.05839467048645, + "learning_rate": 1.9419514180154514e-05, + "loss": 3.3837, + "step": 3386 + }, + { + "epoch": 0.27155198332364555, + "grad_norm": 2.1191418170928955, + "learning_rate": 1.9419078148188833e-05, + "loss": 3.3169, + "step": 3387 + }, + { + "epoch": 0.2716321581046682, + "grad_norm": 2.134401321411133, + "learning_rate": 1.9418641957420815e-05, + "loss": 3.2233, + "step": 3388 + }, + { + "epoch": 0.2717123328856908, + "grad_norm": 2.020519733428955, + "learning_rate": 1.9418205607857816e-05, + "loss": 3.2844, + "step": 3389 + }, + { + "epoch": 0.2717925076667134, + "grad_norm": 2.157325029373169, + "learning_rate": 1.941776909950719e-05, + "loss": 3.278, + "step": 3390 + }, + { + "epoch": 0.27187268244773605, + "grad_norm": 1.9125216007232666, + "learning_rate": 1.941733243237629e-05, + "loss": 3.3258, + "step": 3391 + }, + { + "epoch": 0.27195285722875867, + "grad_norm": 2.0966506004333496, + "learning_rate": 1.941689560647249e-05, + "loss": 3.3125, + "step": 3392 + }, + { + "epoch": 0.27203303200978135, + "grad_norm": 1.9798874855041504, + "learning_rate": 1.9416458621803147e-05, + "loss": 3.3247, + "step": 3393 + }, + { + "epoch": 0.272113206790804, + "grad_norm": 1.9993674755096436, + "learning_rate": 1.9416021478375632e-05, + "loss": 3.1968, + "step": 3394 + }, + { + "epoch": 0.2721933815718266, + "grad_norm": 2.032743215560913, + "learning_rate": 1.941558417619731e-05, + "loss": 3.3262, + "step": 3395 + }, + { + "epoch": 0.2722735563528492, + "grad_norm": 0.9239661693572998, + "learning_rate": 1.9415146715275562e-05, + "loss": 2.5215, + "step": 3396 + }, + { + "epoch": 0.27235373113387185, + "grad_norm": 1.8490897417068481, + "learning_rate": 1.9414709095617755e-05, + "loss": 3.2413, + "step": 3397 + }, + { + "epoch": 0.27243390591489447, + "grad_norm": 1.9129081964492798, + "learning_rate": 1.941427131723127e-05, + "loss": 3.3092, + "step": 3398 + }, + { + "epoch": 0.2725140806959171, + "grad_norm": 2.0564968585968018, + "learning_rate": 1.941383338012349e-05, + "loss": 3.2612, + "step": 3399 + }, + { + "epoch": 0.2725942554769397, + "grad_norm": 1.9331543445587158, + "learning_rate": 1.9413395284301797e-05, + "loss": 3.2444, + "step": 3400 + }, + { + "epoch": 0.27267443025796234, + "grad_norm": 1.886130452156067, + "learning_rate": 1.9412957029773575e-05, + "loss": 3.3147, + "step": 3401 + }, + { + "epoch": 0.27275460503898497, + "grad_norm": 1.9280744791030884, + "learning_rate": 1.941251861654622e-05, + "loss": 3.3535, + "step": 3402 + }, + { + "epoch": 0.2728347798200076, + "grad_norm": 2.249251127243042, + "learning_rate": 1.9412080044627112e-05, + "loss": 3.269, + "step": 3403 + }, + { + "epoch": 0.27291495460103027, + "grad_norm": 1.9739965200424194, + "learning_rate": 1.9411641314023656e-05, + "loss": 3.3348, + "step": 3404 + }, + { + "epoch": 0.2729951293820529, + "grad_norm": 1.877474308013916, + "learning_rate": 1.9411202424743245e-05, + "loss": 3.2704, + "step": 3405 + }, + { + "epoch": 0.2730753041630755, + "grad_norm": 2.019200086593628, + "learning_rate": 1.9410763376793275e-05, + "loss": 3.3754, + "step": 3406 + }, + { + "epoch": 0.27315547894409814, + "grad_norm": 0.9876356720924377, + "learning_rate": 1.9410324170181152e-05, + "loss": 2.5181, + "step": 3407 + }, + { + "epoch": 0.27323565372512076, + "grad_norm": 0.9290504455566406, + "learning_rate": 1.940988480491428e-05, + "loss": 2.5332, + "step": 3408 + }, + { + "epoch": 0.2733158285061434, + "grad_norm": 2.0900559425354004, + "learning_rate": 1.9409445281000067e-05, + "loss": 3.2983, + "step": 3409 + }, + { + "epoch": 0.273396003287166, + "grad_norm": 1.9661548137664795, + "learning_rate": 1.940900559844592e-05, + "loss": 3.3088, + "step": 3410 + }, + { + "epoch": 0.27347617806818864, + "grad_norm": 2.161686658859253, + "learning_rate": 1.9408565757259263e-05, + "loss": 3.1708, + "step": 3411 + }, + { + "epoch": 0.27355635284921126, + "grad_norm": 0.911653459072113, + "learning_rate": 1.9408125757447497e-05, + "loss": 2.5228, + "step": 3412 + }, + { + "epoch": 0.2736365276302339, + "grad_norm": 2.2401723861694336, + "learning_rate": 1.9407685599018044e-05, + "loss": 3.2579, + "step": 3413 + }, + { + "epoch": 0.27371670241125656, + "grad_norm": 1.9045729637145996, + "learning_rate": 1.940724528197833e-05, + "loss": 3.2233, + "step": 3414 + }, + { + "epoch": 0.2737968771922792, + "grad_norm": 2.028632402420044, + "learning_rate": 1.940680480633578e-05, + "loss": 3.1927, + "step": 3415 + }, + { + "epoch": 0.2738770519733018, + "grad_norm": 2.0744292736053467, + "learning_rate": 1.9406364172097812e-05, + "loss": 3.3192, + "step": 3416 + }, + { + "epoch": 0.27395722675432443, + "grad_norm": 1.954686164855957, + "learning_rate": 1.940592337927186e-05, + "loss": 3.2526, + "step": 3417 + }, + { + "epoch": 0.27403740153534706, + "grad_norm": 2.020751714706421, + "learning_rate": 1.9405482427865354e-05, + "loss": 3.2262, + "step": 3418 + }, + { + "epoch": 0.2741175763163697, + "grad_norm": 2.132394790649414, + "learning_rate": 1.940504131788573e-05, + "loss": 3.3738, + "step": 3419 + }, + { + "epoch": 0.2741977510973923, + "grad_norm": 2.0656557083129883, + "learning_rate": 1.9404600049340425e-05, + "loss": 3.3198, + "step": 3420 + }, + { + "epoch": 0.27427792587841493, + "grad_norm": 2.128838539123535, + "learning_rate": 1.9404158622236876e-05, + "loss": 3.2086, + "step": 3421 + }, + { + "epoch": 0.27435810065943755, + "grad_norm": 2.180300712585449, + "learning_rate": 1.9403717036582525e-05, + "loss": 3.2904, + "step": 3422 + }, + { + "epoch": 0.2744382754404602, + "grad_norm": 0.944881796836853, + "learning_rate": 1.9403275292384826e-05, + "loss": 2.5524, + "step": 3423 + }, + { + "epoch": 0.27451845022148286, + "grad_norm": 1.917997121810913, + "learning_rate": 1.9402833389651212e-05, + "loss": 3.2587, + "step": 3424 + }, + { + "epoch": 0.2745986250025055, + "grad_norm": 2.019705057144165, + "learning_rate": 1.9402391328389142e-05, + "loss": 3.2946, + "step": 3425 + }, + { + "epoch": 0.2746787997835281, + "grad_norm": 1.9274041652679443, + "learning_rate": 1.9401949108606072e-05, + "loss": 3.2454, + "step": 3426 + }, + { + "epoch": 0.27475897456455073, + "grad_norm": 1.9746479988098145, + "learning_rate": 1.9401506730309453e-05, + "loss": 3.1371, + "step": 3427 + }, + { + "epoch": 0.27483914934557335, + "grad_norm": 1.9500023126602173, + "learning_rate": 1.940106419350674e-05, + "loss": 3.2846, + "step": 3428 + }, + { + "epoch": 0.274919324126596, + "grad_norm": 2.0223848819732666, + "learning_rate": 1.9400621498205402e-05, + "loss": 3.433, + "step": 3429 + }, + { + "epoch": 0.2749994989076186, + "grad_norm": 1.9214712381362915, + "learning_rate": 1.9400178644412896e-05, + "loss": 3.2815, + "step": 3430 + }, + { + "epoch": 0.2750796736886412, + "grad_norm": 1.9174988269805908, + "learning_rate": 1.9399735632136693e-05, + "loss": 3.2479, + "step": 3431 + }, + { + "epoch": 0.27515984846966385, + "grad_norm": 2.047649621963501, + "learning_rate": 1.9399292461384258e-05, + "loss": 3.281, + "step": 3432 + }, + { + "epoch": 0.27524002325068647, + "grad_norm": 1.9054391384124756, + "learning_rate": 1.9398849132163068e-05, + "loss": 3.2959, + "step": 3433 + }, + { + "epoch": 0.27532019803170915, + "grad_norm": 2.1878747940063477, + "learning_rate": 1.939840564448059e-05, + "loss": 3.2504, + "step": 3434 + }, + { + "epoch": 0.2754003728127318, + "grad_norm": 2.049407482147217, + "learning_rate": 1.9397961998344307e-05, + "loss": 3.3306, + "step": 3435 + }, + { + "epoch": 0.2754805475937544, + "grad_norm": 1.999101996421814, + "learning_rate": 1.9397518193761698e-05, + "loss": 3.21, + "step": 3436 + }, + { + "epoch": 0.275560722374777, + "grad_norm": 2.066831588745117, + "learning_rate": 1.939707423074024e-05, + "loss": 3.2952, + "step": 3437 + }, + { + "epoch": 0.27564089715579965, + "grad_norm": 2.1052134037017822, + "learning_rate": 1.939663010928743e-05, + "loss": 3.2366, + "step": 3438 + }, + { + "epoch": 0.27572107193682227, + "grad_norm": 2.132476806640625, + "learning_rate": 1.939618582941074e-05, + "loss": 3.4175, + "step": 3439 + }, + { + "epoch": 0.2758012467178449, + "grad_norm": 0.8748933672904968, + "learning_rate": 1.9395741391117674e-05, + "loss": 2.5293, + "step": 3440 + }, + { + "epoch": 0.2758814214988675, + "grad_norm": 2.0225136280059814, + "learning_rate": 1.9395296794415717e-05, + "loss": 3.251, + "step": 3441 + }, + { + "epoch": 0.27596159627989014, + "grad_norm": 2.0446934700012207, + "learning_rate": 1.939485203931237e-05, + "loss": 3.2272, + "step": 3442 + }, + { + "epoch": 0.27604177106091277, + "grad_norm": 2.3137097358703613, + "learning_rate": 1.9394407125815124e-05, + "loss": 3.2816, + "step": 3443 + }, + { + "epoch": 0.27612194584193545, + "grad_norm": 1.9751384258270264, + "learning_rate": 1.9393962053931486e-05, + "loss": 3.2682, + "step": 3444 + }, + { + "epoch": 0.27620212062295807, + "grad_norm": 1.929879069328308, + "learning_rate": 1.939351682366896e-05, + "loss": 3.2924, + "step": 3445 + }, + { + "epoch": 0.2762822954039807, + "grad_norm": 2.131178379058838, + "learning_rate": 1.939307143503505e-05, + "loss": 3.2488, + "step": 3446 + }, + { + "epoch": 0.2763624701850033, + "grad_norm": 2.086270809173584, + "learning_rate": 1.939262588803727e-05, + "loss": 3.2815, + "step": 3447 + }, + { + "epoch": 0.27644264496602594, + "grad_norm": 2.0188546180725098, + "learning_rate": 1.9392180182683122e-05, + "loss": 3.2412, + "step": 3448 + }, + { + "epoch": 0.27652281974704856, + "grad_norm": 2.1176042556762695, + "learning_rate": 1.939173431898013e-05, + "loss": 3.2173, + "step": 3449 + }, + { + "epoch": 0.2766029945280712, + "grad_norm": 2.3412580490112305, + "learning_rate": 1.9391288296935805e-05, + "loss": 3.3857, + "step": 3450 + }, + { + "epoch": 0.2766831693090938, + "grad_norm": 2.0571680068969727, + "learning_rate": 1.9390842116557672e-05, + "loss": 3.2889, + "step": 3451 + }, + { + "epoch": 0.27676334409011644, + "grad_norm": 1.9196234941482544, + "learning_rate": 1.939039577785325e-05, + "loss": 3.2708, + "step": 3452 + }, + { + "epoch": 0.27684351887113906, + "grad_norm": 2.0640060901641846, + "learning_rate": 1.9389949280830063e-05, + "loss": 3.3363, + "step": 3453 + }, + { + "epoch": 0.27692369365216174, + "grad_norm": 1.949411392211914, + "learning_rate": 1.938950262549564e-05, + "loss": 3.4266, + "step": 3454 + }, + { + "epoch": 0.27700386843318436, + "grad_norm": 0.9633325338363647, + "learning_rate": 1.9389055811857513e-05, + "loss": 2.531, + "step": 3455 + }, + { + "epoch": 0.277084043214207, + "grad_norm": 1.892655611038208, + "learning_rate": 1.938860883992321e-05, + "loss": 3.3004, + "step": 3456 + }, + { + "epoch": 0.2771642179952296, + "grad_norm": 2.108046054840088, + "learning_rate": 1.938816170970028e-05, + "loss": 3.3195, + "step": 3457 + }, + { + "epoch": 0.27724439277625224, + "grad_norm": 1.8957128524780273, + "learning_rate": 1.9387714421196246e-05, + "loss": 3.2361, + "step": 3458 + }, + { + "epoch": 0.27732456755727486, + "grad_norm": 2.0528573989868164, + "learning_rate": 1.9387266974418656e-05, + "loss": 3.2222, + "step": 3459 + }, + { + "epoch": 0.2774047423382975, + "grad_norm": 1.8132096529006958, + "learning_rate": 1.9386819369375053e-05, + "loss": 3.2383, + "step": 3460 + }, + { + "epoch": 0.2774849171193201, + "grad_norm": 2.0574357509613037, + "learning_rate": 1.9386371606072984e-05, + "loss": 3.2836, + "step": 3461 + }, + { + "epoch": 0.27756509190034273, + "grad_norm": 1.961341381072998, + "learning_rate": 1.938592368452e-05, + "loss": 3.2697, + "step": 3462 + }, + { + "epoch": 0.27764526668136535, + "grad_norm": 1.8997162580490112, + "learning_rate": 1.938547560472365e-05, + "loss": 3.3053, + "step": 3463 + }, + { + "epoch": 0.277725441462388, + "grad_norm": 0.9392799735069275, + "learning_rate": 1.938502736669149e-05, + "loss": 2.5367, + "step": 3464 + }, + { + "epoch": 0.27780561624341066, + "grad_norm": 2.2947561740875244, + "learning_rate": 1.938457897043107e-05, + "loss": 3.3442, + "step": 3465 + }, + { + "epoch": 0.2778857910244333, + "grad_norm": 1.9361788034439087, + "learning_rate": 1.938413041594996e-05, + "loss": 3.2384, + "step": 3466 + }, + { + "epoch": 0.2779659658054559, + "grad_norm": 2.001110553741455, + "learning_rate": 1.9383681703255718e-05, + "loss": 3.2404, + "step": 3467 + }, + { + "epoch": 0.27804614058647853, + "grad_norm": 1.9323710203170776, + "learning_rate": 1.938323283235591e-05, + "loss": 3.2322, + "step": 3468 + }, + { + "epoch": 0.27812631536750115, + "grad_norm": 1.9828230142593384, + "learning_rate": 1.9382783803258105e-05, + "loss": 3.3098, + "step": 3469 + }, + { + "epoch": 0.2782064901485238, + "grad_norm": 2.0986292362213135, + "learning_rate": 1.938233461596987e-05, + "loss": 3.2599, + "step": 3470 + }, + { + "epoch": 0.2782866649295464, + "grad_norm": 1.887195110321045, + "learning_rate": 1.938188527049878e-05, + "loss": 3.2467, + "step": 3471 + }, + { + "epoch": 0.278366839710569, + "grad_norm": 1.9873982667922974, + "learning_rate": 1.938143576685241e-05, + "loss": 3.3187, + "step": 3472 + }, + { + "epoch": 0.27844701449159165, + "grad_norm": 2.2462704181671143, + "learning_rate": 1.938098610503834e-05, + "loss": 3.1796, + "step": 3473 + }, + { + "epoch": 0.2785271892726143, + "grad_norm": 1.9534735679626465, + "learning_rate": 1.9380536285064155e-05, + "loss": 3.2206, + "step": 3474 + }, + { + "epoch": 0.27860736405363695, + "grad_norm": 1.8023879528045654, + "learning_rate": 1.9380086306937428e-05, + "loss": 3.2389, + "step": 3475 + }, + { + "epoch": 0.2786875388346596, + "grad_norm": 2.0546376705169678, + "learning_rate": 1.9379636170665753e-05, + "loss": 3.239, + "step": 3476 + }, + { + "epoch": 0.2787677136156822, + "grad_norm": 1.985374927520752, + "learning_rate": 1.937918587625672e-05, + "loss": 3.2618, + "step": 3477 + }, + { + "epoch": 0.2788478883967048, + "grad_norm": 1.9810439348220825, + "learning_rate": 1.937873542371792e-05, + "loss": 3.3329, + "step": 3478 + }, + { + "epoch": 0.27892806317772745, + "grad_norm": 2.056457996368408, + "learning_rate": 1.937828481305694e-05, + "loss": 3.2723, + "step": 3479 + }, + { + "epoch": 0.27900823795875007, + "grad_norm": 2.0632638931274414, + "learning_rate": 1.9377834044281387e-05, + "loss": 3.1855, + "step": 3480 + }, + { + "epoch": 0.2790884127397727, + "grad_norm": 2.0296449661254883, + "learning_rate": 1.9377383117398857e-05, + "loss": 3.2839, + "step": 3481 + }, + { + "epoch": 0.2791685875207953, + "grad_norm": 1.9442201852798462, + "learning_rate": 1.9376932032416955e-05, + "loss": 3.1733, + "step": 3482 + }, + { + "epoch": 0.27924876230181794, + "grad_norm": 1.8312766551971436, + "learning_rate": 1.937648078934328e-05, + "loss": 3.17, + "step": 3483 + }, + { + "epoch": 0.27932893708284057, + "grad_norm": 1.9907219409942627, + "learning_rate": 1.9376029388185443e-05, + "loss": 3.3, + "step": 3484 + }, + { + "epoch": 0.27940911186386325, + "grad_norm": 2.1724941730499268, + "learning_rate": 1.937557782895106e-05, + "loss": 3.3837, + "step": 3485 + }, + { + "epoch": 0.27948928664488587, + "grad_norm": 2.06475830078125, + "learning_rate": 1.9375126111647736e-05, + "loss": 3.3259, + "step": 3486 + }, + { + "epoch": 0.2795694614259085, + "grad_norm": 2.1504814624786377, + "learning_rate": 1.9374674236283088e-05, + "loss": 3.2911, + "step": 3487 + }, + { + "epoch": 0.2796496362069311, + "grad_norm": 1.9961501359939575, + "learning_rate": 1.9374222202864738e-05, + "loss": 3.1385, + "step": 3488 + }, + { + "epoch": 0.27972981098795374, + "grad_norm": 1.979219675064087, + "learning_rate": 1.9373770011400307e-05, + "loss": 3.1922, + "step": 3489 + }, + { + "epoch": 0.27980998576897637, + "grad_norm": 1.9254144430160522, + "learning_rate": 1.9373317661897416e-05, + "loss": 3.2681, + "step": 3490 + }, + { + "epoch": 0.279890160549999, + "grad_norm": 2.0856096744537354, + "learning_rate": 1.937286515436369e-05, + "loss": 3.221, + "step": 3491 + }, + { + "epoch": 0.2799703353310216, + "grad_norm": 2.1021807193756104, + "learning_rate": 1.9372412488806763e-05, + "loss": 3.2424, + "step": 3492 + }, + { + "epoch": 0.28005051011204424, + "grad_norm": 1.9561679363250732, + "learning_rate": 1.9371959665234267e-05, + "loss": 3.2852, + "step": 3493 + }, + { + "epoch": 0.28013068489306686, + "grad_norm": 1.977648138999939, + "learning_rate": 1.9371506683653833e-05, + "loss": 3.2123, + "step": 3494 + }, + { + "epoch": 0.28021085967408954, + "grad_norm": 2.041412115097046, + "learning_rate": 1.9371053544073097e-05, + "loss": 3.1809, + "step": 3495 + }, + { + "epoch": 0.28029103445511216, + "grad_norm": 1.9933162927627563, + "learning_rate": 1.93706002464997e-05, + "loss": 3.2462, + "step": 3496 + }, + { + "epoch": 0.2803712092361348, + "grad_norm": 1.8601696491241455, + "learning_rate": 1.937014679094129e-05, + "loss": 3.2032, + "step": 3497 + }, + { + "epoch": 0.2804513840171574, + "grad_norm": 2.0731780529022217, + "learning_rate": 1.9369693177405504e-05, + "loss": 3.3309, + "step": 3498 + }, + { + "epoch": 0.28053155879818004, + "grad_norm": 1.965147852897644, + "learning_rate": 1.936923940589999e-05, + "loss": 3.2456, + "step": 3499 + }, + { + "epoch": 0.28061173357920266, + "grad_norm": 2.031080722808838, + "learning_rate": 1.9368785476432406e-05, + "loss": 3.223, + "step": 3500 + }, + { + "epoch": 0.2806919083602253, + "grad_norm": 1.9784445762634277, + "learning_rate": 1.93683313890104e-05, + "loss": 3.3264, + "step": 3501 + }, + { + "epoch": 0.2807720831412479, + "grad_norm": 2.112637996673584, + "learning_rate": 1.9367877143641635e-05, + "loss": 3.3127, + "step": 3502 + }, + { + "epoch": 0.28085225792227053, + "grad_norm": 2.1509785652160645, + "learning_rate": 1.9367422740333756e-05, + "loss": 3.3674, + "step": 3503 + }, + { + "epoch": 0.28093243270329316, + "grad_norm": 2.0741491317749023, + "learning_rate": 1.936696817909443e-05, + "loss": 3.3051, + "step": 3504 + }, + { + "epoch": 0.28101260748431584, + "grad_norm": 2.1055374145507812, + "learning_rate": 1.9366513459931325e-05, + "loss": 3.312, + "step": 3505 + }, + { + "epoch": 0.28109278226533846, + "grad_norm": 2.042468309402466, + "learning_rate": 1.93660585828521e-05, + "loss": 3.2782, + "step": 3506 + }, + { + "epoch": 0.2811729570463611, + "grad_norm": 2.060643196105957, + "learning_rate": 1.9365603547864432e-05, + "loss": 3.3275, + "step": 3507 + }, + { + "epoch": 0.2812531318273837, + "grad_norm": 1.9538969993591309, + "learning_rate": 1.9365148354975987e-05, + "loss": 3.2123, + "step": 3508 + }, + { + "epoch": 0.28133330660840633, + "grad_norm": 2.0263893604278564, + "learning_rate": 1.9364693004194445e-05, + "loss": 3.2823, + "step": 3509 + }, + { + "epoch": 0.28141348138942895, + "grad_norm": 2.0932862758636475, + "learning_rate": 1.9364237495527476e-05, + "loss": 3.2099, + "step": 3510 + }, + { + "epoch": 0.2814936561704516, + "grad_norm": 2.0069580078125, + "learning_rate": 1.936378182898276e-05, + "loss": 3.3082, + "step": 3511 + }, + { + "epoch": 0.2815738309514742, + "grad_norm": 2.0554051399230957, + "learning_rate": 1.9363326004567987e-05, + "loss": 3.2363, + "step": 3512 + }, + { + "epoch": 0.2816540057324968, + "grad_norm": 2.1138477325439453, + "learning_rate": 1.936287002229084e-05, + "loss": 3.1683, + "step": 3513 + }, + { + "epoch": 0.28173418051351945, + "grad_norm": 2.0535378456115723, + "learning_rate": 1.9362413882159e-05, + "loss": 3.304, + "step": 3514 + }, + { + "epoch": 0.28181435529454213, + "grad_norm": 2.0686144828796387, + "learning_rate": 1.9361957584180163e-05, + "loss": 3.2427, + "step": 3515 + }, + { + "epoch": 0.28189453007556475, + "grad_norm": 1.9435558319091797, + "learning_rate": 1.936150112836202e-05, + "loss": 3.2739, + "step": 3516 + }, + { + "epoch": 0.2819747048565874, + "grad_norm": 2.0996320247650146, + "learning_rate": 1.9361044514712266e-05, + "loss": 3.3032, + "step": 3517 + }, + { + "epoch": 0.28205487963761, + "grad_norm": 1.9719724655151367, + "learning_rate": 1.9360587743238604e-05, + "loss": 3.2605, + "step": 3518 + }, + { + "epoch": 0.2821350544186326, + "grad_norm": 2.2054243087768555, + "learning_rate": 1.9360130813948728e-05, + "loss": 3.4717, + "step": 3519 + }, + { + "epoch": 0.28221522919965525, + "grad_norm": 2.1023917198181152, + "learning_rate": 1.935967372685035e-05, + "loss": 3.3541, + "step": 3520 + }, + { + "epoch": 0.2822954039806779, + "grad_norm": 1.9695813655853271, + "learning_rate": 1.9359216481951166e-05, + "loss": 3.3251, + "step": 3521 + }, + { + "epoch": 0.2823755787617005, + "grad_norm": 1.915657639503479, + "learning_rate": 1.9358759079258892e-05, + "loss": 3.2755, + "step": 3522 + }, + { + "epoch": 0.2824557535427231, + "grad_norm": 2.016951560974121, + "learning_rate": 1.9358301518781243e-05, + "loss": 3.1892, + "step": 3523 + }, + { + "epoch": 0.28253592832374574, + "grad_norm": 2.120049476623535, + "learning_rate": 1.9357843800525923e-05, + "loss": 3.3062, + "step": 3524 + }, + { + "epoch": 0.28261610310476837, + "grad_norm": 2.092970609664917, + "learning_rate": 1.9357385924500657e-05, + "loss": 3.4205, + "step": 3525 + }, + { + "epoch": 0.28269627788579105, + "grad_norm": 1.9640976190567017, + "learning_rate": 1.935692789071316e-05, + "loss": 3.225, + "step": 3526 + }, + { + "epoch": 0.28277645266681367, + "grad_norm": 1.9089560508728027, + "learning_rate": 1.9356469699171158e-05, + "loss": 3.1897, + "step": 3527 + }, + { + "epoch": 0.2828566274478363, + "grad_norm": 1.969322919845581, + "learning_rate": 1.9356011349882374e-05, + "loss": 3.2909, + "step": 3528 + }, + { + "epoch": 0.2829368022288589, + "grad_norm": 1.940090537071228, + "learning_rate": 1.935555284285454e-05, + "loss": 3.2224, + "step": 3529 + }, + { + "epoch": 0.28301697700988154, + "grad_norm": 1.948202133178711, + "learning_rate": 1.9355094178095376e-05, + "loss": 3.3199, + "step": 3530 + }, + { + "epoch": 0.28309715179090417, + "grad_norm": 1.9604023694992065, + "learning_rate": 1.9354635355612626e-05, + "loss": 3.3313, + "step": 3531 + }, + { + "epoch": 0.2831773265719268, + "grad_norm": 1.8579953908920288, + "learning_rate": 1.9354176375414018e-05, + "loss": 3.2412, + "step": 3532 + }, + { + "epoch": 0.2832575013529494, + "grad_norm": 2.0280885696411133, + "learning_rate": 1.9353717237507294e-05, + "loss": 3.2629, + "step": 3533 + }, + { + "epoch": 0.28333767613397204, + "grad_norm": 2.0091733932495117, + "learning_rate": 1.9353257941900196e-05, + "loss": 3.3014, + "step": 3534 + }, + { + "epoch": 0.28341785091499466, + "grad_norm": 0.9610753059387207, + "learning_rate": 1.9352798488600463e-05, + "loss": 2.5192, + "step": 3535 + }, + { + "epoch": 0.28349802569601734, + "grad_norm": 2.012394905090332, + "learning_rate": 1.9352338877615843e-05, + "loss": 3.1566, + "step": 3536 + }, + { + "epoch": 0.28357820047703997, + "grad_norm": 2.0628674030303955, + "learning_rate": 1.9351879108954085e-05, + "loss": 3.2384, + "step": 3537 + }, + { + "epoch": 0.2836583752580626, + "grad_norm": 0.9182224869728088, + "learning_rate": 1.9351419182622947e-05, + "loss": 2.5468, + "step": 3538 + }, + { + "epoch": 0.2837385500390852, + "grad_norm": 1.8328505754470825, + "learning_rate": 1.9350959098630173e-05, + "loss": 3.2277, + "step": 3539 + }, + { + "epoch": 0.28381872482010784, + "grad_norm": 2.09594464302063, + "learning_rate": 1.9350498856983523e-05, + "loss": 3.3025, + "step": 3540 + }, + { + "epoch": 0.28389889960113046, + "grad_norm": 1.9707244634628296, + "learning_rate": 1.935003845769076e-05, + "loss": 3.2351, + "step": 3541 + }, + { + "epoch": 0.2839790743821531, + "grad_norm": 2.159846782684326, + "learning_rate": 1.9349577900759643e-05, + "loss": 3.2665, + "step": 3542 + }, + { + "epoch": 0.2840592491631757, + "grad_norm": 1.94426429271698, + "learning_rate": 1.9349117186197937e-05, + "loss": 3.3601, + "step": 3543 + }, + { + "epoch": 0.28413942394419833, + "grad_norm": 1.9415934085845947, + "learning_rate": 1.934865631401341e-05, + "loss": 3.2332, + "step": 3544 + }, + { + "epoch": 0.28421959872522096, + "grad_norm": 1.9154818058013916, + "learning_rate": 1.9348195284213833e-05, + "loss": 3.2068, + "step": 3545 + }, + { + "epoch": 0.28429977350624364, + "grad_norm": 1.9328539371490479, + "learning_rate": 1.934773409680698e-05, + "loss": 3.2195, + "step": 3546 + }, + { + "epoch": 0.28437994828726626, + "grad_norm": 1.8765854835510254, + "learning_rate": 1.9347272751800623e-05, + "loss": 3.2979, + "step": 3547 + }, + { + "epoch": 0.2844601230682889, + "grad_norm": 1.9696037769317627, + "learning_rate": 1.934681124920254e-05, + "loss": 3.1851, + "step": 3548 + }, + { + "epoch": 0.2845402978493115, + "grad_norm": 2.0391921997070312, + "learning_rate": 1.9346349589020517e-05, + "loss": 3.2811, + "step": 3549 + }, + { + "epoch": 0.28462047263033413, + "grad_norm": 1.9947116374969482, + "learning_rate": 1.934588777126233e-05, + "loss": 3.2457, + "step": 3550 + }, + { + "epoch": 0.28470064741135676, + "grad_norm": 2.042224645614624, + "learning_rate": 1.9345425795935774e-05, + "loss": 3.2495, + "step": 3551 + }, + { + "epoch": 0.2847808221923794, + "grad_norm": 1.9568336009979248, + "learning_rate": 1.9344963663048627e-05, + "loss": 3.2175, + "step": 3552 + }, + { + "epoch": 0.284860996973402, + "grad_norm": 1.9237115383148193, + "learning_rate": 1.934450137260869e-05, + "loss": 3.2976, + "step": 3553 + }, + { + "epoch": 0.2849411717544246, + "grad_norm": 1.0000301599502563, + "learning_rate": 1.9344038924623754e-05, + "loss": 2.525, + "step": 3554 + }, + { + "epoch": 0.28502134653544725, + "grad_norm": 2.0121500492095947, + "learning_rate": 1.934357631910161e-05, + "loss": 3.2523, + "step": 3555 + }, + { + "epoch": 0.28510152131646993, + "grad_norm": 2.190812349319458, + "learning_rate": 1.9343113556050066e-05, + "loss": 3.3567, + "step": 3556 + }, + { + "epoch": 0.28518169609749255, + "grad_norm": 2.0215957164764404, + "learning_rate": 1.934265063547692e-05, + "loss": 3.2551, + "step": 3557 + }, + { + "epoch": 0.2852618708785152, + "grad_norm": 0.9293874502182007, + "learning_rate": 1.9342187557389977e-05, + "loss": 2.54, + "step": 3558 + }, + { + "epoch": 0.2853420456595378, + "grad_norm": 2.1330440044403076, + "learning_rate": 1.9341724321797043e-05, + "loss": 3.0743, + "step": 3559 + }, + { + "epoch": 0.2854222204405604, + "grad_norm": 2.1552631855010986, + "learning_rate": 1.9341260928705932e-05, + "loss": 3.255, + "step": 3560 + }, + { + "epoch": 0.28550239522158305, + "grad_norm": 2.171773672103882, + "learning_rate": 1.934079737812445e-05, + "loss": 3.2826, + "step": 3561 + }, + { + "epoch": 0.2855825700026057, + "grad_norm": 2.0225207805633545, + "learning_rate": 1.934033367006042e-05, + "loss": 3.217, + "step": 3562 + }, + { + "epoch": 0.2856627447836283, + "grad_norm": 1.964112401008606, + "learning_rate": 1.9339869804521654e-05, + "loss": 3.2646, + "step": 3563 + }, + { + "epoch": 0.2857429195646509, + "grad_norm": 1.974989891052246, + "learning_rate": 1.9339405781515976e-05, + "loss": 3.2677, + "step": 3564 + }, + { + "epoch": 0.28582309434567355, + "grad_norm": 1.9512628316879272, + "learning_rate": 1.9338941601051206e-05, + "loss": 3.1888, + "step": 3565 + }, + { + "epoch": 0.2859032691266962, + "grad_norm": 2.0797953605651855, + "learning_rate": 1.9338477263135174e-05, + "loss": 3.3729, + "step": 3566 + }, + { + "epoch": 0.28598344390771885, + "grad_norm": 0.9088650345802307, + "learning_rate": 1.9338012767775708e-05, + "loss": 2.5155, + "step": 3567 + }, + { + "epoch": 0.2860636186887415, + "grad_norm": 1.9704787731170654, + "learning_rate": 1.9337548114980634e-05, + "loss": 3.2558, + "step": 3568 + }, + { + "epoch": 0.2861437934697641, + "grad_norm": 2.034095525741577, + "learning_rate": 1.933708330475779e-05, + "loss": 3.3, + "step": 3569 + }, + { + "epoch": 0.2862239682507867, + "grad_norm": 1.9755115509033203, + "learning_rate": 1.9336618337115013e-05, + "loss": 3.2067, + "step": 3570 + }, + { + "epoch": 0.28630414303180934, + "grad_norm": 1.8405925035476685, + "learning_rate": 1.9336153212060144e-05, + "loss": 3.3352, + "step": 3571 + }, + { + "epoch": 0.28638431781283197, + "grad_norm": 1.9247221946716309, + "learning_rate": 1.933568792960102e-05, + "loss": 3.1705, + "step": 3572 + }, + { + "epoch": 0.2864644925938546, + "grad_norm": 2.038454532623291, + "learning_rate": 1.9335222489745486e-05, + "loss": 3.1528, + "step": 3573 + }, + { + "epoch": 0.2865446673748772, + "grad_norm": 1.952858328819275, + "learning_rate": 1.9334756892501394e-05, + "loss": 3.2859, + "step": 3574 + }, + { + "epoch": 0.28662484215589984, + "grad_norm": 2.0862011909484863, + "learning_rate": 1.933429113787659e-05, + "loss": 3.2735, + "step": 3575 + }, + { + "epoch": 0.2867050169369225, + "grad_norm": 2.0534934997558594, + "learning_rate": 1.9333825225878923e-05, + "loss": 3.4288, + "step": 3576 + }, + { + "epoch": 0.28678519171794514, + "grad_norm": 2.0169005393981934, + "learning_rate": 1.9333359156516257e-05, + "loss": 3.2905, + "step": 3577 + }, + { + "epoch": 0.28686536649896777, + "grad_norm": 1.9928100109100342, + "learning_rate": 1.9332892929796442e-05, + "loss": 3.1432, + "step": 3578 + }, + { + "epoch": 0.2869455412799904, + "grad_norm": 1.858768105506897, + "learning_rate": 1.9332426545727342e-05, + "loss": 3.2175, + "step": 3579 + }, + { + "epoch": 0.287025716061013, + "grad_norm": 2.201171875, + "learning_rate": 1.9331960004316823e-05, + "loss": 3.328, + "step": 3580 + }, + { + "epoch": 0.28710589084203564, + "grad_norm": 2.0260181427001953, + "learning_rate": 1.9331493305572743e-05, + "loss": 3.2463, + "step": 3581 + }, + { + "epoch": 0.28718606562305826, + "grad_norm": 2.147719383239746, + "learning_rate": 1.933102644950298e-05, + "loss": 3.1575, + "step": 3582 + }, + { + "epoch": 0.2872662404040809, + "grad_norm": 1.9845640659332275, + "learning_rate": 1.9330559436115396e-05, + "loss": 3.2256, + "step": 3583 + }, + { + "epoch": 0.2873464151851035, + "grad_norm": 0.9359309077262878, + "learning_rate": 1.9330092265417867e-05, + "loss": 2.5354, + "step": 3584 + }, + { + "epoch": 0.28742658996612613, + "grad_norm": 1.869568943977356, + "learning_rate": 1.9329624937418272e-05, + "loss": 3.2751, + "step": 3585 + }, + { + "epoch": 0.28750676474714876, + "grad_norm": 2.0414233207702637, + "learning_rate": 1.9329157452124486e-05, + "loss": 3.1533, + "step": 3586 + }, + { + "epoch": 0.28758693952817144, + "grad_norm": 2.0980136394500732, + "learning_rate": 1.9328689809544397e-05, + "loss": 3.3904, + "step": 3587 + }, + { + "epoch": 0.28766711430919406, + "grad_norm": 2.046205759048462, + "learning_rate": 1.9328222009685885e-05, + "loss": 3.2391, + "step": 3588 + }, + { + "epoch": 0.2877472890902167, + "grad_norm": 2.1321768760681152, + "learning_rate": 1.9327754052556838e-05, + "loss": 3.3547, + "step": 3589 + }, + { + "epoch": 0.2878274638712393, + "grad_norm": 1.8094245195388794, + "learning_rate": 1.9327285938165144e-05, + "loss": 3.2385, + "step": 3590 + }, + { + "epoch": 0.28790763865226193, + "grad_norm": 2.1044108867645264, + "learning_rate": 1.93268176665187e-05, + "loss": 3.3328, + "step": 3591 + }, + { + "epoch": 0.28798781343328456, + "grad_norm": 1.8891093730926514, + "learning_rate": 1.932634923762539e-05, + "loss": 3.2642, + "step": 3592 + }, + { + "epoch": 0.2880679882143072, + "grad_norm": 0.9050361514091492, + "learning_rate": 1.9325880651493124e-05, + "loss": 2.5358, + "step": 3593 + }, + { + "epoch": 0.2881481629953298, + "grad_norm": 2.1071951389312744, + "learning_rate": 1.9325411908129793e-05, + "loss": 3.2576, + "step": 3594 + }, + { + "epoch": 0.28822833777635243, + "grad_norm": 1.9670759439468384, + "learning_rate": 1.9324943007543306e-05, + "loss": 3.3311, + "step": 3595 + }, + { + "epoch": 0.28830851255737505, + "grad_norm": 1.9070982933044434, + "learning_rate": 1.9324473949741568e-05, + "loss": 3.2034, + "step": 3596 + }, + { + "epoch": 0.28838868733839773, + "grad_norm": 1.964064359664917, + "learning_rate": 1.9324004734732483e-05, + "loss": 3.1985, + "step": 3597 + }, + { + "epoch": 0.28846886211942036, + "grad_norm": 1.9322209358215332, + "learning_rate": 1.9323535362523967e-05, + "loss": 3.248, + "step": 3598 + }, + { + "epoch": 0.288549036900443, + "grad_norm": 1.994705319404602, + "learning_rate": 1.932306583312393e-05, + "loss": 3.3159, + "step": 3599 + }, + { + "epoch": 0.2886292116814656, + "grad_norm": 0.9182929992675781, + "learning_rate": 1.9322596146540284e-05, + "loss": 2.5477, + "step": 3600 + }, + { + "epoch": 0.2887093864624882, + "grad_norm": 2.426112651824951, + "learning_rate": 1.932212630278096e-05, + "loss": 3.3202, + "step": 3601 + }, + { + "epoch": 0.28878956124351085, + "grad_norm": 1.8940216302871704, + "learning_rate": 1.9321656301853867e-05, + "loss": 3.2942, + "step": 3602 + }, + { + "epoch": 0.2888697360245335, + "grad_norm": 2.568276882171631, + "learning_rate": 1.9321186143766935e-05, + "loss": 3.2994, + "step": 3603 + }, + { + "epoch": 0.2889499108055561, + "grad_norm": 0.9049045443534851, + "learning_rate": 1.932071582852809e-05, + "loss": 2.5277, + "step": 3604 + }, + { + "epoch": 0.2890300855865787, + "grad_norm": 1.9477659463882446, + "learning_rate": 1.932024535614526e-05, + "loss": 3.3229, + "step": 3605 + }, + { + "epoch": 0.28911026036760135, + "grad_norm": 1.9730830192565918, + "learning_rate": 1.931977472662638e-05, + "loss": 3.1639, + "step": 3606 + }, + { + "epoch": 0.289190435148624, + "grad_norm": 2.021280288696289, + "learning_rate": 1.9319303939979385e-05, + "loss": 3.2324, + "step": 3607 + }, + { + "epoch": 0.28927060992964665, + "grad_norm": 2.017016649246216, + "learning_rate": 1.9318832996212203e-05, + "loss": 3.3337, + "step": 3608 + }, + { + "epoch": 0.2893507847106693, + "grad_norm": 2.0299220085144043, + "learning_rate": 1.9318361895332786e-05, + "loss": 3.3285, + "step": 3609 + }, + { + "epoch": 0.2894309594916919, + "grad_norm": 1.8686965703964233, + "learning_rate": 1.931789063734907e-05, + "loss": 3.2819, + "step": 3610 + }, + { + "epoch": 0.2895111342727145, + "grad_norm": 1.972066879272461, + "learning_rate": 1.9317419222269002e-05, + "loss": 3.246, + "step": 3611 + }, + { + "epoch": 0.28959130905373714, + "grad_norm": 2.125575065612793, + "learning_rate": 1.931694765010053e-05, + "loss": 3.358, + "step": 3612 + }, + { + "epoch": 0.28967148383475977, + "grad_norm": 2.0578701496124268, + "learning_rate": 1.9316475920851606e-05, + "loss": 3.2923, + "step": 3613 + }, + { + "epoch": 0.2897516586157824, + "grad_norm": 1.8076528310775757, + "learning_rate": 1.931600403453018e-05, + "loss": 3.3096, + "step": 3614 + }, + { + "epoch": 0.289831833396805, + "grad_norm": 1.9081865549087524, + "learning_rate": 1.931553199114421e-05, + "loss": 3.3214, + "step": 3615 + }, + { + "epoch": 0.28991200817782764, + "grad_norm": 1.9693505764007568, + "learning_rate": 1.931505979070165e-05, + "loss": 3.2047, + "step": 3616 + }, + { + "epoch": 0.2899921829588503, + "grad_norm": 0.92295241355896, + "learning_rate": 1.931458743321047e-05, + "loss": 2.5275, + "step": 3617 + }, + { + "epoch": 0.29007235773987294, + "grad_norm": 1.9967759847640991, + "learning_rate": 1.9314114918678623e-05, + "loss": 3.3419, + "step": 3618 + }, + { + "epoch": 0.29015253252089557, + "grad_norm": 0.8938349485397339, + "learning_rate": 1.9313642247114086e-05, + "loss": 2.5308, + "step": 3619 + }, + { + "epoch": 0.2902327073019182, + "grad_norm": 1.927621603012085, + "learning_rate": 1.9313169418524822e-05, + "loss": 3.2009, + "step": 3620 + }, + { + "epoch": 0.2903128820829408, + "grad_norm": 1.9768717288970947, + "learning_rate": 1.93126964329188e-05, + "loss": 3.2264, + "step": 3621 + }, + { + "epoch": 0.29039305686396344, + "grad_norm": 1.8321272134780884, + "learning_rate": 1.9312223290304003e-05, + "loss": 3.1384, + "step": 3622 + }, + { + "epoch": 0.29047323164498606, + "grad_norm": 2.0489864349365234, + "learning_rate": 1.93117499906884e-05, + "loss": 3.2681, + "step": 3623 + }, + { + "epoch": 0.2905534064260087, + "grad_norm": 0.948834240436554, + "learning_rate": 1.9311276534079976e-05, + "loss": 2.522, + "step": 3624 + }, + { + "epoch": 0.2906335812070313, + "grad_norm": 2.0381219387054443, + "learning_rate": 1.9310802920486713e-05, + "loss": 3.3324, + "step": 3625 + }, + { + "epoch": 0.29071375598805393, + "grad_norm": 2.0688118934631348, + "learning_rate": 1.931032914991659e-05, + "loss": 3.3508, + "step": 3626 + }, + { + "epoch": 0.2907939307690766, + "grad_norm": 2.0992205142974854, + "learning_rate": 1.9309855222377604e-05, + "loss": 3.2997, + "step": 3627 + }, + { + "epoch": 0.29087410555009924, + "grad_norm": 2.079784870147705, + "learning_rate": 1.9309381137877736e-05, + "loss": 3.271, + "step": 3628 + }, + { + "epoch": 0.29095428033112186, + "grad_norm": 2.170200824737549, + "learning_rate": 1.9308906896424987e-05, + "loss": 3.3748, + "step": 3629 + }, + { + "epoch": 0.2910344551121445, + "grad_norm": 2.2653732299804688, + "learning_rate": 1.9308432498027344e-05, + "loss": 3.3621, + "step": 3630 + }, + { + "epoch": 0.2911146298931671, + "grad_norm": 1.9253779649734497, + "learning_rate": 1.930795794269281e-05, + "loss": 3.2496, + "step": 3631 + }, + { + "epoch": 0.29119480467418973, + "grad_norm": 1.9877132177352905, + "learning_rate": 1.9307483230429385e-05, + "loss": 3.228, + "step": 3632 + }, + { + "epoch": 0.29127497945521236, + "grad_norm": 1.9797080755233765, + "learning_rate": 1.9307008361245076e-05, + "loss": 3.2506, + "step": 3633 + }, + { + "epoch": 0.291355154236235, + "grad_norm": 0.833375096321106, + "learning_rate": 1.9306533335147887e-05, + "loss": 2.5232, + "step": 3634 + }, + { + "epoch": 0.2914353290172576, + "grad_norm": 2.04636812210083, + "learning_rate": 1.930605815214582e-05, + "loss": 3.3379, + "step": 3635 + }, + { + "epoch": 0.29151550379828023, + "grad_norm": 1.8103270530700684, + "learning_rate": 1.9305582812246898e-05, + "loss": 3.1688, + "step": 3636 + }, + { + "epoch": 0.29159567857930285, + "grad_norm": 1.90645432472229, + "learning_rate": 1.9305107315459127e-05, + "loss": 3.162, + "step": 3637 + }, + { + "epoch": 0.29167585336032553, + "grad_norm": 2.167712688446045, + "learning_rate": 1.9304631661790528e-05, + "loss": 3.3246, + "step": 3638 + }, + { + "epoch": 0.29175602814134816, + "grad_norm": 1.9494202136993408, + "learning_rate": 1.930415585124912e-05, + "loss": 3.3973, + "step": 3639 + }, + { + "epoch": 0.2918362029223708, + "grad_norm": 2.0296826362609863, + "learning_rate": 1.9303679883842918e-05, + "loss": 3.2305, + "step": 3640 + }, + { + "epoch": 0.2919163777033934, + "grad_norm": 2.062734603881836, + "learning_rate": 1.9303203759579953e-05, + "loss": 3.2465, + "step": 3641 + }, + { + "epoch": 0.291996552484416, + "grad_norm": 0.9204874634742737, + "learning_rate": 1.9302727478468256e-05, + "loss": 2.5488, + "step": 3642 + }, + { + "epoch": 0.29207672726543865, + "grad_norm": 1.8071264028549194, + "learning_rate": 1.9302251040515848e-05, + "loss": 3.2126, + "step": 3643 + }, + { + "epoch": 0.2921569020464613, + "grad_norm": 1.9878833293914795, + "learning_rate": 1.9301774445730767e-05, + "loss": 3.2407, + "step": 3644 + }, + { + "epoch": 0.2922370768274839, + "grad_norm": 1.9419364929199219, + "learning_rate": 1.930129769412105e-05, + "loss": 3.2334, + "step": 3645 + }, + { + "epoch": 0.2923172516085065, + "grad_norm": 1.897097110748291, + "learning_rate": 1.9300820785694726e-05, + "loss": 3.2937, + "step": 3646 + }, + { + "epoch": 0.29239742638952915, + "grad_norm": 2.089411973953247, + "learning_rate": 1.9300343720459845e-05, + "loss": 3.1937, + "step": 3647 + }, + { + "epoch": 0.2924776011705518, + "grad_norm": 2.0062272548675537, + "learning_rate": 1.9299866498424447e-05, + "loss": 3.1511, + "step": 3648 + }, + { + "epoch": 0.29255777595157445, + "grad_norm": 2.009218454360962, + "learning_rate": 1.9299389119596577e-05, + "loss": 3.2969, + "step": 3649 + }, + { + "epoch": 0.2926379507325971, + "grad_norm": 1.9194180965423584, + "learning_rate": 1.9298911583984283e-05, + "loss": 3.2776, + "step": 3650 + }, + { + "epoch": 0.2927181255136197, + "grad_norm": 1.9183790683746338, + "learning_rate": 1.9298433891595617e-05, + "loss": 3.1807, + "step": 3651 + }, + { + "epoch": 0.2927983002946423, + "grad_norm": 2.129232883453369, + "learning_rate": 1.9297956042438634e-05, + "loss": 3.2379, + "step": 3652 + }, + { + "epoch": 0.29287847507566495, + "grad_norm": 2.071230888366699, + "learning_rate": 1.9297478036521387e-05, + "loss": 3.2519, + "step": 3653 + }, + { + "epoch": 0.29295864985668757, + "grad_norm": 1.8735127449035645, + "learning_rate": 1.929699987385194e-05, + "loss": 3.2917, + "step": 3654 + }, + { + "epoch": 0.2930388246377102, + "grad_norm": 1.9461742639541626, + "learning_rate": 1.929652155443835e-05, + "loss": 3.2439, + "step": 3655 + }, + { + "epoch": 0.2931189994187328, + "grad_norm": 1.9136289358139038, + "learning_rate": 1.929604307828868e-05, + "loss": 3.1418, + "step": 3656 + }, + { + "epoch": 0.29319917419975544, + "grad_norm": 1.922761082649231, + "learning_rate": 1.9295564445411005e-05, + "loss": 3.1959, + "step": 3657 + }, + { + "epoch": 0.2932793489807781, + "grad_norm": 1.915016770362854, + "learning_rate": 1.929508565581339e-05, + "loss": 3.2446, + "step": 3658 + }, + { + "epoch": 0.29335952376180074, + "grad_norm": 2.483765125274658, + "learning_rate": 1.9294606709503904e-05, + "loss": 3.2506, + "step": 3659 + }, + { + "epoch": 0.29343969854282337, + "grad_norm": 1.8837412595748901, + "learning_rate": 1.9294127606490623e-05, + "loss": 3.2708, + "step": 3660 + }, + { + "epoch": 0.293519873323846, + "grad_norm": 1.9573516845703125, + "learning_rate": 1.9293648346781627e-05, + "loss": 3.1687, + "step": 3661 + }, + { + "epoch": 0.2936000481048686, + "grad_norm": 1.9930808544158936, + "learning_rate": 1.9293168930384997e-05, + "loss": 3.2142, + "step": 3662 + }, + { + "epoch": 0.29368022288589124, + "grad_norm": 1.8500112295150757, + "learning_rate": 1.9292689357308813e-05, + "loss": 3.1985, + "step": 3663 + }, + { + "epoch": 0.29376039766691386, + "grad_norm": 2.01302433013916, + "learning_rate": 1.9292209627561164e-05, + "loss": 3.2216, + "step": 3664 + }, + { + "epoch": 0.2938405724479365, + "grad_norm": 2.0062475204467773, + "learning_rate": 1.9291729741150132e-05, + "loss": 3.2849, + "step": 3665 + }, + { + "epoch": 0.2939207472289591, + "grad_norm": 1.9486925601959229, + "learning_rate": 1.9291249698083813e-05, + "loss": 3.2209, + "step": 3666 + }, + { + "epoch": 0.29400092200998174, + "grad_norm": 2.71112060546875, + "learning_rate": 1.92907694983703e-05, + "loss": 3.2524, + "step": 3667 + }, + { + "epoch": 0.2940810967910044, + "grad_norm": 2.027045488357544, + "learning_rate": 1.9290289142017685e-05, + "loss": 3.2789, + "step": 3668 + }, + { + "epoch": 0.29416127157202704, + "grad_norm": 2.1906323432922363, + "learning_rate": 1.928980862903407e-05, + "loss": 3.3271, + "step": 3669 + }, + { + "epoch": 0.29424144635304966, + "grad_norm": 1.9252856969833374, + "learning_rate": 1.9289327959427554e-05, + "loss": 3.3016, + "step": 3670 + }, + { + "epoch": 0.2943216211340723, + "grad_norm": 1.9942941665649414, + "learning_rate": 1.9288847133206247e-05, + "loss": 3.3365, + "step": 3671 + }, + { + "epoch": 0.2944017959150949, + "grad_norm": 2.184937000274658, + "learning_rate": 1.928836615037825e-05, + "loss": 3.2738, + "step": 3672 + }, + { + "epoch": 0.29448197069611753, + "grad_norm": 1.8692357540130615, + "learning_rate": 1.928788501095167e-05, + "loss": 3.2455, + "step": 3673 + }, + { + "epoch": 0.29456214547714016, + "grad_norm": 1.991847276687622, + "learning_rate": 1.928740371493462e-05, + "loss": 3.243, + "step": 3674 + }, + { + "epoch": 0.2946423202581628, + "grad_norm": 1.927824854850769, + "learning_rate": 1.9286922262335225e-05, + "loss": 3.2845, + "step": 3675 + }, + { + "epoch": 0.2947224950391854, + "grad_norm": 0.8725367784500122, + "learning_rate": 1.9286440653161588e-05, + "loss": 2.5409, + "step": 3676 + }, + { + "epoch": 0.29480266982020803, + "grad_norm": 2.070176601409912, + "learning_rate": 1.9285958887421836e-05, + "loss": 3.2646, + "step": 3677 + }, + { + "epoch": 0.2948828446012307, + "grad_norm": 1.899441123008728, + "learning_rate": 1.9285476965124087e-05, + "loss": 3.3152, + "step": 3678 + }, + { + "epoch": 0.29496301938225333, + "grad_norm": 1.9185020923614502, + "learning_rate": 1.9284994886276474e-05, + "loss": 3.2261, + "step": 3679 + }, + { + "epoch": 0.29504319416327596, + "grad_norm": 2.121978521347046, + "learning_rate": 1.9284512650887114e-05, + "loss": 3.2479, + "step": 3680 + }, + { + "epoch": 0.2951233689442986, + "grad_norm": 1.8716765642166138, + "learning_rate": 1.9284030258964148e-05, + "loss": 3.3026, + "step": 3681 + }, + { + "epoch": 0.2952035437253212, + "grad_norm": 1.9361289739608765, + "learning_rate": 1.92835477105157e-05, + "loss": 3.2418, + "step": 3682 + }, + { + "epoch": 0.29528371850634383, + "grad_norm": 1.8243577480316162, + "learning_rate": 1.928306500554991e-05, + "loss": 3.2318, + "step": 3683 + }, + { + "epoch": 0.29536389328736645, + "grad_norm": 2.058800458908081, + "learning_rate": 1.928258214407492e-05, + "loss": 3.17, + "step": 3684 + }, + { + "epoch": 0.2954440680683891, + "grad_norm": 1.9866634607315063, + "learning_rate": 1.9282099126098862e-05, + "loss": 3.3218, + "step": 3685 + }, + { + "epoch": 0.2955242428494117, + "grad_norm": 1.9039757251739502, + "learning_rate": 1.9281615951629888e-05, + "loss": 3.3163, + "step": 3686 + }, + { + "epoch": 0.2956044176304343, + "grad_norm": 2.202580690383911, + "learning_rate": 1.9281132620676133e-05, + "loss": 3.2631, + "step": 3687 + }, + { + "epoch": 0.295684592411457, + "grad_norm": 2.110893726348877, + "learning_rate": 1.9280649133245757e-05, + "loss": 3.2568, + "step": 3688 + }, + { + "epoch": 0.2957647671924796, + "grad_norm": 1.8936867713928223, + "learning_rate": 1.928016548934691e-05, + "loss": 3.2281, + "step": 3689 + }, + { + "epoch": 0.29584494197350225, + "grad_norm": 2.173821210861206, + "learning_rate": 1.9279681688987742e-05, + "loss": 3.1407, + "step": 3690 + }, + { + "epoch": 0.2959251167545249, + "grad_norm": 2.2932190895080566, + "learning_rate": 1.927919773217641e-05, + "loss": 3.3101, + "step": 3691 + }, + { + "epoch": 0.2960052915355475, + "grad_norm": 2.04888653755188, + "learning_rate": 1.9278713618921074e-05, + "loss": 3.215, + "step": 3692 + }, + { + "epoch": 0.2960854663165701, + "grad_norm": 1.9490412473678589, + "learning_rate": 1.92782293492299e-05, + "loss": 3.1764, + "step": 3693 + }, + { + "epoch": 0.29616564109759275, + "grad_norm": 1.9604358673095703, + "learning_rate": 1.9277744923111045e-05, + "loss": 3.257, + "step": 3694 + }, + { + "epoch": 0.29624581587861537, + "grad_norm": 2.2096354961395264, + "learning_rate": 1.927726034057268e-05, + "loss": 3.2322, + "step": 3695 + }, + { + "epoch": 0.296325990659638, + "grad_norm": 1.9477686882019043, + "learning_rate": 1.927677560162298e-05, + "loss": 3.2402, + "step": 3696 + }, + { + "epoch": 0.2964061654406606, + "grad_norm": 1.8624459505081177, + "learning_rate": 1.927629070627011e-05, + "loss": 3.212, + "step": 3697 + }, + { + "epoch": 0.29648634022168324, + "grad_norm": 2.064135789871216, + "learning_rate": 1.9275805654522246e-05, + "loss": 3.2459, + "step": 3698 + }, + { + "epoch": 0.2965665150027059, + "grad_norm": 0.9775267243385315, + "learning_rate": 1.9275320446387572e-05, + "loss": 2.5179, + "step": 3699 + }, + { + "epoch": 0.29664668978372855, + "grad_norm": 2.100924253463745, + "learning_rate": 1.9274835081874262e-05, + "loss": 3.2416, + "step": 3700 + }, + { + "epoch": 0.29672686456475117, + "grad_norm": 1.9764448404312134, + "learning_rate": 1.92743495609905e-05, + "loss": 3.267, + "step": 3701 + }, + { + "epoch": 0.2968070393457738, + "grad_norm": 1.9340301752090454, + "learning_rate": 1.9273863883744477e-05, + "loss": 3.2388, + "step": 3702 + }, + { + "epoch": 0.2968872141267964, + "grad_norm": 2.069082021713257, + "learning_rate": 1.9273378050144375e-05, + "loss": 3.2126, + "step": 3703 + }, + { + "epoch": 0.29696738890781904, + "grad_norm": 2.0234293937683105, + "learning_rate": 1.9272892060198385e-05, + "loss": 3.252, + "step": 3704 + }, + { + "epoch": 0.29704756368884166, + "grad_norm": 1.9339956045150757, + "learning_rate": 1.9272405913914705e-05, + "loss": 3.2737, + "step": 3705 + }, + { + "epoch": 0.2971277384698643, + "grad_norm": 1.9511550664901733, + "learning_rate": 1.927191961130153e-05, + "loss": 3.2633, + "step": 3706 + }, + { + "epoch": 0.2972079132508869, + "grad_norm": 1.8874980211257935, + "learning_rate": 1.9271433152367058e-05, + "loss": 3.2296, + "step": 3707 + }, + { + "epoch": 0.29728808803190954, + "grad_norm": 2.0199694633483887, + "learning_rate": 1.927094653711949e-05, + "loss": 3.1995, + "step": 3708 + }, + { + "epoch": 0.2973682628129322, + "grad_norm": 2.035698652267456, + "learning_rate": 1.927045976556703e-05, + "loss": 3.2216, + "step": 3709 + }, + { + "epoch": 0.29744843759395484, + "grad_norm": 2.0604665279388428, + "learning_rate": 1.926997283771789e-05, + "loss": 3.1918, + "step": 3710 + }, + { + "epoch": 0.29752861237497746, + "grad_norm": 1.9120900630950928, + "learning_rate": 1.9269485753580268e-05, + "loss": 3.2604, + "step": 3711 + }, + { + "epoch": 0.2976087871560001, + "grad_norm": 2.106476306915283, + "learning_rate": 1.9268998513162388e-05, + "loss": 3.2896, + "step": 3712 + }, + { + "epoch": 0.2976889619370227, + "grad_norm": 1.81393563747406, + "learning_rate": 1.926851111647246e-05, + "loss": 3.256, + "step": 3713 + }, + { + "epoch": 0.29776913671804534, + "grad_norm": 1.9094704389572144, + "learning_rate": 1.9268023563518696e-05, + "loss": 3.3743, + "step": 3714 + }, + { + "epoch": 0.29784931149906796, + "grad_norm": 1.895796537399292, + "learning_rate": 1.9267535854309323e-05, + "loss": 3.1948, + "step": 3715 + }, + { + "epoch": 0.2979294862800906, + "grad_norm": 1.9900290966033936, + "learning_rate": 1.9267047988852564e-05, + "loss": 3.1487, + "step": 3716 + }, + { + "epoch": 0.2980096610611132, + "grad_norm": 1.8849958181381226, + "learning_rate": 1.9266559967156643e-05, + "loss": 3.2845, + "step": 3717 + }, + { + "epoch": 0.29808983584213583, + "grad_norm": 2.0028483867645264, + "learning_rate": 1.9266071789229785e-05, + "loss": 3.2657, + "step": 3718 + }, + { + "epoch": 0.2981700106231585, + "grad_norm": 2.4480276107788086, + "learning_rate": 1.9265583455080222e-05, + "loss": 3.3215, + "step": 3719 + }, + { + "epoch": 0.29825018540418113, + "grad_norm": 2.0615768432617188, + "learning_rate": 1.9265094964716188e-05, + "loss": 3.301, + "step": 3720 + }, + { + "epoch": 0.29833036018520376, + "grad_norm": 2.170224666595459, + "learning_rate": 1.9264606318145915e-05, + "loss": 3.2634, + "step": 3721 + }, + { + "epoch": 0.2984105349662264, + "grad_norm": 2.0411622524261475, + "learning_rate": 1.9264117515377648e-05, + "loss": 3.2847, + "step": 3722 + }, + { + "epoch": 0.298490709747249, + "grad_norm": 1.939189076423645, + "learning_rate": 1.9263628556419626e-05, + "loss": 3.3813, + "step": 3723 + }, + { + "epoch": 0.29857088452827163, + "grad_norm": 2.0771329402923584, + "learning_rate": 1.9263139441280088e-05, + "loss": 3.2198, + "step": 3724 + }, + { + "epoch": 0.29865105930929425, + "grad_norm": 0.9300599694252014, + "learning_rate": 1.9262650169967286e-05, + "loss": 2.5234, + "step": 3725 + }, + { + "epoch": 0.2987312340903169, + "grad_norm": 0.877933919429779, + "learning_rate": 1.9262160742489465e-05, + "loss": 2.5369, + "step": 3726 + }, + { + "epoch": 0.2988114088713395, + "grad_norm": 2.150012493133545, + "learning_rate": 1.926167115885488e-05, + "loss": 3.1454, + "step": 3727 + }, + { + "epoch": 0.2988915836523621, + "grad_norm": 2.1458120346069336, + "learning_rate": 1.9261181419071783e-05, + "loss": 3.2546, + "step": 3728 + }, + { + "epoch": 0.2989717584333848, + "grad_norm": 0.8638997077941895, + "learning_rate": 1.926069152314843e-05, + "loss": 2.5435, + "step": 3729 + }, + { + "epoch": 0.29905193321440743, + "grad_norm": 1.9602584838867188, + "learning_rate": 1.9260201471093083e-05, + "loss": 3.3589, + "step": 3730 + }, + { + "epoch": 0.29913210799543005, + "grad_norm": 2.0209357738494873, + "learning_rate": 1.9259711262914004e-05, + "loss": 3.3488, + "step": 3731 + }, + { + "epoch": 0.2992122827764527, + "grad_norm": 2.0619025230407715, + "learning_rate": 1.9259220898619454e-05, + "loss": 3.3336, + "step": 3732 + }, + { + "epoch": 0.2992924575574753, + "grad_norm": 1.9879363775253296, + "learning_rate": 1.92587303782177e-05, + "loss": 3.238, + "step": 3733 + }, + { + "epoch": 0.2993726323384979, + "grad_norm": 2.035703182220459, + "learning_rate": 1.925823970171702e-05, + "loss": 3.3751, + "step": 3734 + }, + { + "epoch": 0.29945280711952055, + "grad_norm": 1.8556523323059082, + "learning_rate": 1.9257748869125682e-05, + "loss": 3.2078, + "step": 3735 + }, + { + "epoch": 0.29953298190054317, + "grad_norm": 2.1878881454467773, + "learning_rate": 1.925725788045196e-05, + "loss": 3.3318, + "step": 3736 + }, + { + "epoch": 0.2996131566815658, + "grad_norm": 1.915993332862854, + "learning_rate": 1.9256766735704133e-05, + "loss": 3.2188, + "step": 3737 + }, + { + "epoch": 0.2996933314625884, + "grad_norm": 2.109584093093872, + "learning_rate": 1.925627543489048e-05, + "loss": 3.2109, + "step": 3738 + }, + { + "epoch": 0.2997735062436111, + "grad_norm": 1.9951503276824951, + "learning_rate": 1.9255783978019284e-05, + "loss": 3.1691, + "step": 3739 + }, + { + "epoch": 0.2998536810246337, + "grad_norm": 1.9791450500488281, + "learning_rate": 1.925529236509883e-05, + "loss": 3.2009, + "step": 3740 + }, + { + "epoch": 0.29993385580565635, + "grad_norm": 2.0470240116119385, + "learning_rate": 1.925480059613741e-05, + "loss": 3.2815, + "step": 3741 + }, + { + "epoch": 0.30001403058667897, + "grad_norm": 2.129389524459839, + "learning_rate": 1.925430867114332e-05, + "loss": 3.2039, + "step": 3742 + }, + { + "epoch": 0.3000942053677016, + "grad_norm": 1.9857450723648071, + "learning_rate": 1.925381659012484e-05, + "loss": 3.3212, + "step": 3743 + }, + { + "epoch": 0.3001743801487242, + "grad_norm": 1.986749529838562, + "learning_rate": 1.9253324353090278e-05, + "loss": 3.2007, + "step": 3744 + }, + { + "epoch": 0.30025455492974684, + "grad_norm": 1.9852075576782227, + "learning_rate": 1.925283196004793e-05, + "loss": 3.1707, + "step": 3745 + }, + { + "epoch": 0.30033472971076947, + "grad_norm": 1.8732913732528687, + "learning_rate": 1.925233941100609e-05, + "loss": 3.2914, + "step": 3746 + }, + { + "epoch": 0.3004149044917921, + "grad_norm": 1.8991219997406006, + "learning_rate": 1.9251846705973074e-05, + "loss": 3.3848, + "step": 3747 + }, + { + "epoch": 0.3004950792728147, + "grad_norm": 1.908831000328064, + "learning_rate": 1.9251353844957183e-05, + "loss": 3.2611, + "step": 3748 + }, + { + "epoch": 0.3005752540538374, + "grad_norm": 1.969334602355957, + "learning_rate": 1.9250860827966723e-05, + "loss": 3.255, + "step": 3749 + }, + { + "epoch": 0.30065542883486, + "grad_norm": 1.9065648317337036, + "learning_rate": 1.925036765501001e-05, + "loss": 3.2704, + "step": 3750 + }, + { + "epoch": 0.30073560361588264, + "grad_norm": 0.8992530703544617, + "learning_rate": 1.924987432609536e-05, + "loss": 2.5374, + "step": 3751 + }, + { + "epoch": 0.30081577839690526, + "grad_norm": 1.9602587223052979, + "learning_rate": 1.9249380841231088e-05, + "loss": 3.2437, + "step": 3752 + }, + { + "epoch": 0.3008959531779279, + "grad_norm": 1.9731394052505493, + "learning_rate": 1.9248887200425518e-05, + "loss": 3.207, + "step": 3753 + }, + { + "epoch": 0.3009761279589505, + "grad_norm": 1.9137383699417114, + "learning_rate": 1.9248393403686966e-05, + "loss": 3.2375, + "step": 3754 + }, + { + "epoch": 0.30105630273997314, + "grad_norm": 1.8625576496124268, + "learning_rate": 1.924789945102376e-05, + "loss": 3.1562, + "step": 3755 + }, + { + "epoch": 0.30113647752099576, + "grad_norm": 2.1955296993255615, + "learning_rate": 1.924740534244423e-05, + "loss": 3.2874, + "step": 3756 + }, + { + "epoch": 0.3012166523020184, + "grad_norm": 2.1025257110595703, + "learning_rate": 1.9246911077956708e-05, + "loss": 3.2346, + "step": 3757 + }, + { + "epoch": 0.301296827083041, + "grad_norm": 1.987229347229004, + "learning_rate": 1.924641665756952e-05, + "loss": 3.2876, + "step": 3758 + }, + { + "epoch": 0.30137700186406363, + "grad_norm": 1.93221914768219, + "learning_rate": 1.9245922081291007e-05, + "loss": 3.1143, + "step": 3759 + }, + { + "epoch": 0.3014571766450863, + "grad_norm": 1.8843156099319458, + "learning_rate": 1.924542734912951e-05, + "loss": 3.3322, + "step": 3760 + }, + { + "epoch": 0.30153735142610893, + "grad_norm": 1.8312867879867554, + "learning_rate": 1.924493246109336e-05, + "loss": 3.1825, + "step": 3761 + }, + { + "epoch": 0.30161752620713156, + "grad_norm": 2.0228147506713867, + "learning_rate": 1.924443741719091e-05, + "loss": 3.3209, + "step": 3762 + }, + { + "epoch": 0.3016977009881542, + "grad_norm": 2.1703972816467285, + "learning_rate": 1.9243942217430503e-05, + "loss": 3.2918, + "step": 3763 + }, + { + "epoch": 0.3017778757691768, + "grad_norm": 2.4334609508514404, + "learning_rate": 1.9243446861820488e-05, + "loss": 3.3292, + "step": 3764 + }, + { + "epoch": 0.30185805055019943, + "grad_norm": 2.155601739883423, + "learning_rate": 1.9242951350369217e-05, + "loss": 3.277, + "step": 3765 + }, + { + "epoch": 0.30193822533122205, + "grad_norm": 2.0345895290374756, + "learning_rate": 1.9242455683085042e-05, + "loss": 3.3042, + "step": 3766 + }, + { + "epoch": 0.3020184001122447, + "grad_norm": 2.079098701477051, + "learning_rate": 1.9241959859976323e-05, + "loss": 3.3951, + "step": 3767 + }, + { + "epoch": 0.3020985748932673, + "grad_norm": 1.150458812713623, + "learning_rate": 1.9241463881051418e-05, + "loss": 2.5408, + "step": 3768 + }, + { + "epoch": 0.3021787496742899, + "grad_norm": 1.9327616691589355, + "learning_rate": 1.924096774631869e-05, + "loss": 3.2146, + "step": 3769 + }, + { + "epoch": 0.3022589244553126, + "grad_norm": 2.1728155612945557, + "learning_rate": 1.92404714557865e-05, + "loss": 3.2563, + "step": 3770 + }, + { + "epoch": 0.30233909923633523, + "grad_norm": 2.3745951652526855, + "learning_rate": 1.9239975009463222e-05, + "loss": 3.3229, + "step": 3771 + }, + { + "epoch": 0.30241927401735785, + "grad_norm": 1.906211018562317, + "learning_rate": 1.923947840735722e-05, + "loss": 3.1592, + "step": 3772 + }, + { + "epoch": 0.3024994487983805, + "grad_norm": 1.8710403442382812, + "learning_rate": 1.9238981649476866e-05, + "loss": 3.1737, + "step": 3773 + }, + { + "epoch": 0.3025796235794031, + "grad_norm": 1.91925847530365, + "learning_rate": 1.9238484735830537e-05, + "loss": 3.2179, + "step": 3774 + }, + { + "epoch": 0.3026597983604257, + "grad_norm": 1.8824365139007568, + "learning_rate": 1.9237987666426616e-05, + "loss": 3.1467, + "step": 3775 + }, + { + "epoch": 0.30273997314144835, + "grad_norm": 1.9651037454605103, + "learning_rate": 1.9237490441273475e-05, + "loss": 3.2354, + "step": 3776 + }, + { + "epoch": 0.302820147922471, + "grad_norm": 2.0002474784851074, + "learning_rate": 1.9236993060379503e-05, + "loss": 3.2386, + "step": 3777 + }, + { + "epoch": 0.3029003227034936, + "grad_norm": 1.022275686264038, + "learning_rate": 1.9236495523753083e-05, + "loss": 2.5308, + "step": 3778 + }, + { + "epoch": 0.3029804974845162, + "grad_norm": 2.1197965145111084, + "learning_rate": 1.9235997831402603e-05, + "loss": 3.2565, + "step": 3779 + }, + { + "epoch": 0.3030606722655389, + "grad_norm": 2.2449610233306885, + "learning_rate": 1.9235499983336455e-05, + "loss": 3.2803, + "step": 3780 + }, + { + "epoch": 0.3031408470465615, + "grad_norm": 2.067779064178467, + "learning_rate": 1.9235001979563033e-05, + "loss": 3.2911, + "step": 3781 + }, + { + "epoch": 0.30322102182758415, + "grad_norm": 2.1374599933624268, + "learning_rate": 1.9234503820090732e-05, + "loss": 3.3019, + "step": 3782 + }, + { + "epoch": 0.30330119660860677, + "grad_norm": 1.9228827953338623, + "learning_rate": 1.9234005504927947e-05, + "loss": 3.2325, + "step": 3783 + }, + { + "epoch": 0.3033813713896294, + "grad_norm": 0.9707468152046204, + "learning_rate": 1.923350703408309e-05, + "loss": 2.5604, + "step": 3784 + }, + { + "epoch": 0.303461546170652, + "grad_norm": 0.9262839555740356, + "learning_rate": 1.9233008407564555e-05, + "loss": 2.5381, + "step": 3785 + }, + { + "epoch": 0.30354172095167464, + "grad_norm": 1.9282865524291992, + "learning_rate": 1.9232509625380753e-05, + "loss": 3.3257, + "step": 3786 + }, + { + "epoch": 0.30362189573269727, + "grad_norm": 1.9043142795562744, + "learning_rate": 1.923201068754009e-05, + "loss": 3.2671, + "step": 3787 + }, + { + "epoch": 0.3037020705137199, + "grad_norm": 2.01546311378479, + "learning_rate": 1.9231511594050983e-05, + "loss": 3.2854, + "step": 3788 + }, + { + "epoch": 0.3037822452947425, + "grad_norm": 2.0313963890075684, + "learning_rate": 1.9231012344921847e-05, + "loss": 3.2352, + "step": 3789 + }, + { + "epoch": 0.3038624200757652, + "grad_norm": 1.973333716392517, + "learning_rate": 1.9230512940161094e-05, + "loss": 3.197, + "step": 3790 + }, + { + "epoch": 0.3039425948567878, + "grad_norm": 2.0873711109161377, + "learning_rate": 1.9230013379777145e-05, + "loss": 3.3583, + "step": 3791 + }, + { + "epoch": 0.30402276963781044, + "grad_norm": 2.160496473312378, + "learning_rate": 1.9229513663778425e-05, + "loss": 3.2649, + "step": 3792 + }, + { + "epoch": 0.30410294441883307, + "grad_norm": 1.884652018547058, + "learning_rate": 1.9229013792173355e-05, + "loss": 3.1592, + "step": 3793 + }, + { + "epoch": 0.3041831191998557, + "grad_norm": 2.116847276687622, + "learning_rate": 1.9228513764970365e-05, + "loss": 3.2526, + "step": 3794 + }, + { + "epoch": 0.3042632939808783, + "grad_norm": 2.08489990234375, + "learning_rate": 1.9228013582177886e-05, + "loss": 3.2489, + "step": 3795 + }, + { + "epoch": 0.30434346876190094, + "grad_norm": 0.9838817715644836, + "learning_rate": 1.922751324380435e-05, + "loss": 2.5358, + "step": 3796 + }, + { + "epoch": 0.30442364354292356, + "grad_norm": 2.038745403289795, + "learning_rate": 1.9227012749858194e-05, + "loss": 3.3199, + "step": 3797 + }, + { + "epoch": 0.3045038183239462, + "grad_norm": 1.9892791509628296, + "learning_rate": 1.9226512100347853e-05, + "loss": 3.2069, + "step": 3798 + }, + { + "epoch": 0.3045839931049688, + "grad_norm": 1.9058891534805298, + "learning_rate": 1.9226011295281776e-05, + "loss": 3.2421, + "step": 3799 + }, + { + "epoch": 0.3046641678859915, + "grad_norm": 1.8449913263320923, + "learning_rate": 1.922551033466839e-05, + "loss": 3.2522, + "step": 3800 + }, + { + "epoch": 0.3047443426670141, + "grad_norm": 2.0036323070526123, + "learning_rate": 1.9225009218516158e-05, + "loss": 3.2351, + "step": 3801 + }, + { + "epoch": 0.30482451744803674, + "grad_norm": 2.074855327606201, + "learning_rate": 1.922450794683352e-05, + "loss": 3.2604, + "step": 3802 + }, + { + "epoch": 0.30490469222905936, + "grad_norm": 1.8811694383621216, + "learning_rate": 1.922400651962893e-05, + "loss": 3.2452, + "step": 3803 + }, + { + "epoch": 0.304984867010082, + "grad_norm": 1.992087721824646, + "learning_rate": 1.9223504936910838e-05, + "loss": 3.3115, + "step": 3804 + }, + { + "epoch": 0.3050650417911046, + "grad_norm": 2.526201009750366, + "learning_rate": 1.922300319868771e-05, + "loss": 3.2219, + "step": 3805 + }, + { + "epoch": 0.30514521657212723, + "grad_norm": 1.944654941558838, + "learning_rate": 1.9222501304967992e-05, + "loss": 3.343, + "step": 3806 + }, + { + "epoch": 0.30522539135314986, + "grad_norm": 1.9038630723953247, + "learning_rate": 1.9221999255760156e-05, + "loss": 3.2258, + "step": 3807 + }, + { + "epoch": 0.3053055661341725, + "grad_norm": 1.8908532857894897, + "learning_rate": 1.9221497051072662e-05, + "loss": 3.1877, + "step": 3808 + }, + { + "epoch": 0.3053857409151951, + "grad_norm": 2.033170700073242, + "learning_rate": 1.9220994690913975e-05, + "loss": 3.1971, + "step": 3809 + }, + { + "epoch": 0.3054659156962178, + "grad_norm": 1.9051430225372314, + "learning_rate": 1.922049217529257e-05, + "loss": 3.1816, + "step": 3810 + }, + { + "epoch": 0.3055460904772404, + "grad_norm": 2.148223638534546, + "learning_rate": 1.9219989504216916e-05, + "loss": 3.3137, + "step": 3811 + }, + { + "epoch": 0.30562626525826303, + "grad_norm": 2.053778886795044, + "learning_rate": 1.9219486677695485e-05, + "loss": 3.2097, + "step": 3812 + }, + { + "epoch": 0.30570644003928565, + "grad_norm": 2.176570415496826, + "learning_rate": 1.9218983695736763e-05, + "loss": 3.2461, + "step": 3813 + }, + { + "epoch": 0.3057866148203083, + "grad_norm": 2.394219398498535, + "learning_rate": 1.9218480558349226e-05, + "loss": 3.2544, + "step": 3814 + }, + { + "epoch": 0.3058667896013309, + "grad_norm": 2.3810923099517822, + "learning_rate": 1.921797726554135e-05, + "loss": 3.2688, + "step": 3815 + }, + { + "epoch": 0.3059469643823535, + "grad_norm": 1.8874461650848389, + "learning_rate": 1.9217473817321626e-05, + "loss": 3.1818, + "step": 3816 + }, + { + "epoch": 0.30602713916337615, + "grad_norm": 2.213804006576538, + "learning_rate": 1.921697021369855e-05, + "loss": 3.3254, + "step": 3817 + }, + { + "epoch": 0.3061073139443988, + "grad_norm": 1.9854816198349, + "learning_rate": 1.9216466454680596e-05, + "loss": 3.2428, + "step": 3818 + }, + { + "epoch": 0.3061874887254214, + "grad_norm": 2.1366159915924072, + "learning_rate": 1.9215962540276272e-05, + "loss": 3.3051, + "step": 3819 + }, + { + "epoch": 0.306267663506444, + "grad_norm": 1.9413584470748901, + "learning_rate": 1.9215458470494064e-05, + "loss": 3.2808, + "step": 3820 + }, + { + "epoch": 0.3063478382874667, + "grad_norm": 0.9589145183563232, + "learning_rate": 1.9214954245342476e-05, + "loss": 2.5366, + "step": 3821 + }, + { + "epoch": 0.3064280130684893, + "grad_norm": 1.8540257215499878, + "learning_rate": 1.9214449864830005e-05, + "loss": 3.3309, + "step": 3822 + }, + { + "epoch": 0.30650818784951195, + "grad_norm": 1.8717684745788574, + "learning_rate": 1.921394532896516e-05, + "loss": 3.1575, + "step": 3823 + }, + { + "epoch": 0.30658836263053457, + "grad_norm": 2.2201576232910156, + "learning_rate": 1.921344063775644e-05, + "loss": 3.2795, + "step": 3824 + }, + { + "epoch": 0.3066685374115572, + "grad_norm": 1.8784700632095337, + "learning_rate": 1.921293579121236e-05, + "loss": 3.2674, + "step": 3825 + }, + { + "epoch": 0.3067487121925798, + "grad_norm": 2.063958168029785, + "learning_rate": 1.9212430789341426e-05, + "loss": 3.363, + "step": 3826 + }, + { + "epoch": 0.30682888697360244, + "grad_norm": 2.109651803970337, + "learning_rate": 1.921192563215216e-05, + "loss": 3.3242, + "step": 3827 + }, + { + "epoch": 0.30690906175462507, + "grad_norm": 1.873648762702942, + "learning_rate": 1.9211420319653075e-05, + "loss": 3.282, + "step": 3828 + }, + { + "epoch": 0.3069892365356477, + "grad_norm": 1.943225383758545, + "learning_rate": 1.9210914851852688e-05, + "loss": 3.2591, + "step": 3829 + }, + { + "epoch": 0.3070694113166703, + "grad_norm": 2.0823171138763428, + "learning_rate": 1.9210409228759523e-05, + "loss": 3.2656, + "step": 3830 + }, + { + "epoch": 0.307149586097693, + "grad_norm": 1.9994784593582153, + "learning_rate": 1.9209903450382104e-05, + "loss": 3.2666, + "step": 3831 + }, + { + "epoch": 0.3072297608787156, + "grad_norm": 2.04658579826355, + "learning_rate": 1.9209397516728958e-05, + "loss": 3.1804, + "step": 3832 + }, + { + "epoch": 0.30730993565973824, + "grad_norm": 2.0327463150024414, + "learning_rate": 1.9208891427808617e-05, + "loss": 3.3044, + "step": 3833 + }, + { + "epoch": 0.30739011044076087, + "grad_norm": 2.1134467124938965, + "learning_rate": 1.920838518362961e-05, + "loss": 3.2759, + "step": 3834 + }, + { + "epoch": 0.3074702852217835, + "grad_norm": 2.1870687007904053, + "learning_rate": 1.9207878784200476e-05, + "loss": 3.1696, + "step": 3835 + }, + { + "epoch": 0.3075504600028061, + "grad_norm": 1.939650058746338, + "learning_rate": 1.9207372229529752e-05, + "loss": 3.2085, + "step": 3836 + }, + { + "epoch": 0.30763063478382874, + "grad_norm": 2.005215883255005, + "learning_rate": 1.9206865519625977e-05, + "loss": 3.2481, + "step": 3837 + }, + { + "epoch": 0.30771080956485136, + "grad_norm": 2.166874647140503, + "learning_rate": 1.920635865449769e-05, + "loss": 3.3055, + "step": 3838 + }, + { + "epoch": 0.307790984345874, + "grad_norm": 2.0247128009796143, + "learning_rate": 1.9205851634153447e-05, + "loss": 3.2484, + "step": 3839 + }, + { + "epoch": 0.3078711591268966, + "grad_norm": 1.9226045608520508, + "learning_rate": 1.9205344458601785e-05, + "loss": 3.2319, + "step": 3840 + }, + { + "epoch": 0.3079513339079193, + "grad_norm": 1.9849300384521484, + "learning_rate": 1.9204837127851262e-05, + "loss": 3.3464, + "step": 3841 + }, + { + "epoch": 0.3080315086889419, + "grad_norm": 2.215553045272827, + "learning_rate": 1.920432964191043e-05, + "loss": 3.2756, + "step": 3842 + }, + { + "epoch": 0.30811168346996454, + "grad_norm": 1.8921680450439453, + "learning_rate": 1.9203822000787843e-05, + "loss": 3.2052, + "step": 3843 + }, + { + "epoch": 0.30819185825098716, + "grad_norm": 1.9773515462875366, + "learning_rate": 1.9203314204492057e-05, + "loss": 3.1434, + "step": 3844 + }, + { + "epoch": 0.3082720330320098, + "grad_norm": 2.1733853816986084, + "learning_rate": 1.9202806253031643e-05, + "loss": 3.2031, + "step": 3845 + }, + { + "epoch": 0.3083522078130324, + "grad_norm": 1.927605152130127, + "learning_rate": 1.9202298146415156e-05, + "loss": 3.1842, + "step": 3846 + }, + { + "epoch": 0.30843238259405503, + "grad_norm": 2.1499173641204834, + "learning_rate": 1.9201789884651164e-05, + "loss": 3.3853, + "step": 3847 + }, + { + "epoch": 0.30851255737507766, + "grad_norm": 1.942478060722351, + "learning_rate": 1.920128146774824e-05, + "loss": 3.2332, + "step": 3848 + }, + { + "epoch": 0.3085927321561003, + "grad_norm": 2.066302537918091, + "learning_rate": 1.9200772895714954e-05, + "loss": 3.26, + "step": 3849 + }, + { + "epoch": 0.3086729069371229, + "grad_norm": 2.117607831954956, + "learning_rate": 1.920026416855988e-05, + "loss": 3.1945, + "step": 3850 + }, + { + "epoch": 0.3087530817181456, + "grad_norm": 1.9794089794158936, + "learning_rate": 1.9199755286291592e-05, + "loss": 3.1987, + "step": 3851 + }, + { + "epoch": 0.3088332564991682, + "grad_norm": 1.84760582447052, + "learning_rate": 1.9199246248918673e-05, + "loss": 3.3557, + "step": 3852 + }, + { + "epoch": 0.30891343128019083, + "grad_norm": 2.0266196727752686, + "learning_rate": 1.9198737056449705e-05, + "loss": 3.2456, + "step": 3853 + }, + { + "epoch": 0.30899360606121345, + "grad_norm": 1.9325679540634155, + "learning_rate": 1.9198227708893274e-05, + "loss": 3.3278, + "step": 3854 + }, + { + "epoch": 0.3090737808422361, + "grad_norm": 2.051701068878174, + "learning_rate": 1.9197718206257964e-05, + "loss": 3.2581, + "step": 3855 + }, + { + "epoch": 0.3091539556232587, + "grad_norm": 1.8727130889892578, + "learning_rate": 1.9197208548552365e-05, + "loss": 3.4434, + "step": 3856 + }, + { + "epoch": 0.3092341304042813, + "grad_norm": 1.977274775505066, + "learning_rate": 1.9196698735785073e-05, + "loss": 3.1821, + "step": 3857 + }, + { + "epoch": 0.30931430518530395, + "grad_norm": 1.9609431028366089, + "learning_rate": 1.919618876796468e-05, + "loss": 3.2112, + "step": 3858 + }, + { + "epoch": 0.3093944799663266, + "grad_norm": 2.0003840923309326, + "learning_rate": 1.9195678645099787e-05, + "loss": 3.2212, + "step": 3859 + }, + { + "epoch": 0.3094746547473492, + "grad_norm": 1.8660815954208374, + "learning_rate": 1.9195168367198992e-05, + "loss": 3.1709, + "step": 3860 + }, + { + "epoch": 0.3095548295283719, + "grad_norm": 2.020216226577759, + "learning_rate": 1.9194657934270897e-05, + "loss": 3.3435, + "step": 3861 + }, + { + "epoch": 0.3096350043093945, + "grad_norm": 1.8789215087890625, + "learning_rate": 1.9194147346324113e-05, + "loss": 3.303, + "step": 3862 + }, + { + "epoch": 0.3097151790904171, + "grad_norm": 0.9365851879119873, + "learning_rate": 1.9193636603367244e-05, + "loss": 2.5344, + "step": 3863 + }, + { + "epoch": 0.30979535387143975, + "grad_norm": 1.9579684734344482, + "learning_rate": 1.9193125705408902e-05, + "loss": 3.1912, + "step": 3864 + }, + { + "epoch": 0.3098755286524624, + "grad_norm": 2.029078960418701, + "learning_rate": 1.91926146524577e-05, + "loss": 3.2347, + "step": 3865 + }, + { + "epoch": 0.309955703433485, + "grad_norm": 1.9757417440414429, + "learning_rate": 1.9192103444522257e-05, + "loss": 3.3116, + "step": 3866 + }, + { + "epoch": 0.3100358782145076, + "grad_norm": 1.9080798625946045, + "learning_rate": 1.9191592081611187e-05, + "loss": 3.2039, + "step": 3867 + }, + { + "epoch": 0.31011605299553024, + "grad_norm": 1.9913190603256226, + "learning_rate": 1.9191080563733117e-05, + "loss": 3.245, + "step": 3868 + }, + { + "epoch": 0.31019622777655287, + "grad_norm": 1.8633925914764404, + "learning_rate": 1.9190568890896665e-05, + "loss": 3.3505, + "step": 3869 + }, + { + "epoch": 0.3102764025575755, + "grad_norm": 1.8378344774246216, + "learning_rate": 1.919005706311046e-05, + "loss": 3.2467, + "step": 3870 + }, + { + "epoch": 0.3103565773385981, + "grad_norm": 1.794274926185608, + "learning_rate": 1.9189545080383135e-05, + "loss": 3.2128, + "step": 3871 + }, + { + "epoch": 0.3104367521196208, + "grad_norm": 1.8712042570114136, + "learning_rate": 1.918903294272332e-05, + "loss": 3.3032, + "step": 3872 + }, + { + "epoch": 0.3105169269006434, + "grad_norm": 1.9908820390701294, + "learning_rate": 1.9188520650139643e-05, + "loss": 3.2353, + "step": 3873 + }, + { + "epoch": 0.31059710168166604, + "grad_norm": 2.0576012134552, + "learning_rate": 1.9188008202640746e-05, + "loss": 3.314, + "step": 3874 + }, + { + "epoch": 0.31067727646268867, + "grad_norm": 1.9920880794525146, + "learning_rate": 1.9187495600235273e-05, + "loss": 3.2434, + "step": 3875 + }, + { + "epoch": 0.3107574512437113, + "grad_norm": 2.223588705062866, + "learning_rate": 1.918698284293186e-05, + "loss": 3.3923, + "step": 3876 + }, + { + "epoch": 0.3108376260247339, + "grad_norm": 1.9425047636032104, + "learning_rate": 1.9186469930739154e-05, + "loss": 3.2709, + "step": 3877 + }, + { + "epoch": 0.31091780080575654, + "grad_norm": 1.9279680252075195, + "learning_rate": 1.9185956863665803e-05, + "loss": 3.1881, + "step": 3878 + }, + { + "epoch": 0.31099797558677916, + "grad_norm": 1.9871859550476074, + "learning_rate": 1.9185443641720457e-05, + "loss": 3.311, + "step": 3879 + }, + { + "epoch": 0.3110781503678018, + "grad_norm": 2.021566867828369, + "learning_rate": 1.918493026491177e-05, + "loss": 3.2719, + "step": 3880 + }, + { + "epoch": 0.3111583251488244, + "grad_norm": 2.0892598628997803, + "learning_rate": 1.918441673324839e-05, + "loss": 3.22, + "step": 3881 + }, + { + "epoch": 0.3112384999298471, + "grad_norm": 0.8864895105361938, + "learning_rate": 1.9183903046738986e-05, + "loss": 2.5338, + "step": 3882 + }, + { + "epoch": 0.3113186747108697, + "grad_norm": 0.8850840926170349, + "learning_rate": 1.9183389205392215e-05, + "loss": 2.5226, + "step": 3883 + }, + { + "epoch": 0.31139884949189234, + "grad_norm": 1.9804539680480957, + "learning_rate": 1.9182875209216732e-05, + "loss": 3.2655, + "step": 3884 + }, + { + "epoch": 0.31147902427291496, + "grad_norm": 2.1519877910614014, + "learning_rate": 1.9182361058221214e-05, + "loss": 3.2592, + "step": 3885 + }, + { + "epoch": 0.3115591990539376, + "grad_norm": 2.0336694717407227, + "learning_rate": 1.9181846752414325e-05, + "loss": 3.1748, + "step": 3886 + }, + { + "epoch": 0.3116393738349602, + "grad_norm": 1.9146112203598022, + "learning_rate": 1.9181332291804734e-05, + "loss": 3.182, + "step": 3887 + }, + { + "epoch": 0.31171954861598283, + "grad_norm": 2.2367472648620605, + "learning_rate": 1.9180817676401115e-05, + "loss": 3.3417, + "step": 3888 + }, + { + "epoch": 0.31179972339700546, + "grad_norm": 2.014596939086914, + "learning_rate": 1.9180302906212146e-05, + "loss": 3.296, + "step": 3889 + }, + { + "epoch": 0.3118798981780281, + "grad_norm": 2.051079034805298, + "learning_rate": 1.917978798124651e-05, + "loss": 3.2856, + "step": 3890 + }, + { + "epoch": 0.3119600729590507, + "grad_norm": 1.9833160638809204, + "learning_rate": 1.917927290151288e-05, + "loss": 3.2171, + "step": 3891 + }, + { + "epoch": 0.3120402477400734, + "grad_norm": 1.900899052619934, + "learning_rate": 1.9178757667019946e-05, + "loss": 3.2386, + "step": 3892 + }, + { + "epoch": 0.312120422521096, + "grad_norm": 1.973222017288208, + "learning_rate": 1.917824227777639e-05, + "loss": 3.2778, + "step": 3893 + }, + { + "epoch": 0.31220059730211863, + "grad_norm": 1.870059847831726, + "learning_rate": 1.9177726733790906e-05, + "loss": 3.1351, + "step": 3894 + }, + { + "epoch": 0.31228077208314126, + "grad_norm": 0.9719212651252747, + "learning_rate": 1.9177211035072182e-05, + "loss": 2.541, + "step": 3895 + }, + { + "epoch": 0.3123609468641639, + "grad_norm": 2.042266845703125, + "learning_rate": 1.9176695181628914e-05, + "loss": 3.3142, + "step": 3896 + }, + { + "epoch": 0.3124411216451865, + "grad_norm": 1.9259954690933228, + "learning_rate": 1.9176179173469803e-05, + "loss": 3.3433, + "step": 3897 + }, + { + "epoch": 0.3125212964262091, + "grad_norm": 2.015756130218506, + "learning_rate": 1.9175663010603544e-05, + "loss": 3.1661, + "step": 3898 + }, + { + "epoch": 0.31260147120723175, + "grad_norm": 0.8334498405456543, + "learning_rate": 1.917514669303884e-05, + "loss": 2.5266, + "step": 3899 + }, + { + "epoch": 0.3126816459882544, + "grad_norm": 1.9748164415359497, + "learning_rate": 1.9174630220784392e-05, + "loss": 3.194, + "step": 3900 + }, + { + "epoch": 0.312761820769277, + "grad_norm": 1.9607796669006348, + "learning_rate": 1.9174113593848916e-05, + "loss": 3.2542, + "step": 3901 + }, + { + "epoch": 0.3128419955502997, + "grad_norm": 2.2143795490264893, + "learning_rate": 1.917359681224112e-05, + "loss": 3.3397, + "step": 3902 + }, + { + "epoch": 0.3129221703313223, + "grad_norm": 1.948582410812378, + "learning_rate": 1.917307987596971e-05, + "loss": 3.2826, + "step": 3903 + }, + { + "epoch": 0.3130023451123449, + "grad_norm": 2.0027737617492676, + "learning_rate": 1.9172562785043413e-05, + "loss": 3.3154, + "step": 3904 + }, + { + "epoch": 0.31308251989336755, + "grad_norm": 1.9781750440597534, + "learning_rate": 1.9172045539470932e-05, + "loss": 3.2406, + "step": 3905 + }, + { + "epoch": 0.3131626946743902, + "grad_norm": 2.0357913970947266, + "learning_rate": 1.9171528139261003e-05, + "loss": 3.2979, + "step": 3906 + }, + { + "epoch": 0.3132428694554128, + "grad_norm": 2.4269745349884033, + "learning_rate": 1.917101058442234e-05, + "loss": 3.3575, + "step": 3907 + }, + { + "epoch": 0.3133230442364354, + "grad_norm": 1.8764901161193848, + "learning_rate": 1.917049287496367e-05, + "loss": 3.3204, + "step": 3908 + }, + { + "epoch": 0.31340321901745805, + "grad_norm": 2.1206576824188232, + "learning_rate": 1.9169975010893717e-05, + "loss": 3.3639, + "step": 3909 + }, + { + "epoch": 0.31348339379848067, + "grad_norm": 1.9852527379989624, + "learning_rate": 1.916945699222122e-05, + "loss": 3.2764, + "step": 3910 + }, + { + "epoch": 0.3135635685795033, + "grad_norm": 2.05851149559021, + "learning_rate": 1.9168938818954916e-05, + "loss": 3.2039, + "step": 3911 + }, + { + "epoch": 0.313643743360526, + "grad_norm": 2.0867631435394287, + "learning_rate": 1.916842049110353e-05, + "loss": 3.3096, + "step": 3912 + }, + { + "epoch": 0.3137239181415486, + "grad_norm": 1.9979544878005981, + "learning_rate": 1.9167902008675808e-05, + "loss": 3.255, + "step": 3913 + }, + { + "epoch": 0.3138040929225712, + "grad_norm": 2.0606725215911865, + "learning_rate": 1.9167383371680482e-05, + "loss": 3.3104, + "step": 3914 + }, + { + "epoch": 0.31388426770359384, + "grad_norm": 2.105316400527954, + "learning_rate": 1.9166864580126312e-05, + "loss": 3.2108, + "step": 3915 + }, + { + "epoch": 0.31396444248461647, + "grad_norm": 1.9990416765213013, + "learning_rate": 1.9166345634022033e-05, + "loss": 3.272, + "step": 3916 + }, + { + "epoch": 0.3140446172656391, + "grad_norm": 1.9335676431655884, + "learning_rate": 1.9165826533376396e-05, + "loss": 3.1943, + "step": 3917 + }, + { + "epoch": 0.3141247920466617, + "grad_norm": 2.035630464553833, + "learning_rate": 1.9165307278198156e-05, + "loss": 3.2256, + "step": 3918 + }, + { + "epoch": 0.31420496682768434, + "grad_norm": 2.2253475189208984, + "learning_rate": 1.9164787868496062e-05, + "loss": 3.3417, + "step": 3919 + }, + { + "epoch": 0.31428514160870696, + "grad_norm": 2.0126395225524902, + "learning_rate": 1.9164268304278878e-05, + "loss": 3.2817, + "step": 3920 + }, + { + "epoch": 0.3143653163897296, + "grad_norm": 1.8815838098526, + "learning_rate": 1.916374858555536e-05, + "loss": 3.2356, + "step": 3921 + }, + { + "epoch": 0.31444549117075227, + "grad_norm": 1.9759441614151, + "learning_rate": 1.9163228712334268e-05, + "loss": 3.3193, + "step": 3922 + }, + { + "epoch": 0.3145256659517749, + "grad_norm": 2.083965301513672, + "learning_rate": 1.9162708684624372e-05, + "loss": 3.2545, + "step": 3923 + }, + { + "epoch": 0.3146058407327975, + "grad_norm": 1.9027440547943115, + "learning_rate": 1.9162188502434433e-05, + "loss": 3.3558, + "step": 3924 + }, + { + "epoch": 0.31468601551382014, + "grad_norm": 2.0622317790985107, + "learning_rate": 1.9161668165773227e-05, + "loss": 3.1947, + "step": 3925 + }, + { + "epoch": 0.31476619029484276, + "grad_norm": 2.036334753036499, + "learning_rate": 1.9161147674649525e-05, + "loss": 3.2085, + "step": 3926 + }, + { + "epoch": 0.3148463650758654, + "grad_norm": 1.9801055192947388, + "learning_rate": 1.9160627029072103e-05, + "loss": 3.3082, + "step": 3927 + }, + { + "epoch": 0.314926539856888, + "grad_norm": 1.8816975355148315, + "learning_rate": 1.9160106229049736e-05, + "loss": 3.3589, + "step": 3928 + }, + { + "epoch": 0.31500671463791063, + "grad_norm": 1.8751599788665771, + "learning_rate": 1.915958527459121e-05, + "loss": 3.3337, + "step": 3929 + }, + { + "epoch": 0.31508688941893326, + "grad_norm": 2.0623116493225098, + "learning_rate": 1.91590641657053e-05, + "loss": 3.2527, + "step": 3930 + }, + { + "epoch": 0.3151670641999559, + "grad_norm": 1.901236891746521, + "learning_rate": 1.9158542902400795e-05, + "loss": 3.2119, + "step": 3931 + }, + { + "epoch": 0.3152472389809785, + "grad_norm": 1.8505749702453613, + "learning_rate": 1.915802148468649e-05, + "loss": 3.2814, + "step": 3932 + }, + { + "epoch": 0.3153274137620012, + "grad_norm": 1.9843615293502808, + "learning_rate": 1.9157499912571168e-05, + "loss": 3.2358, + "step": 3933 + }, + { + "epoch": 0.3154075885430238, + "grad_norm": 1.9261963367462158, + "learning_rate": 1.915697818606362e-05, + "loss": 3.2618, + "step": 3934 + }, + { + "epoch": 0.31548776332404643, + "grad_norm": 2.154853582382202, + "learning_rate": 1.9156456305172654e-05, + "loss": 3.1917, + "step": 3935 + }, + { + "epoch": 0.31556793810506906, + "grad_norm": 0.949589729309082, + "learning_rate": 1.915593426990706e-05, + "loss": 2.5383, + "step": 3936 + }, + { + "epoch": 0.3156481128860917, + "grad_norm": 1.9402382373809814, + "learning_rate": 1.915541208027564e-05, + "loss": 3.1761, + "step": 3937 + }, + { + "epoch": 0.3157282876671143, + "grad_norm": 1.987862467765808, + "learning_rate": 1.91548897362872e-05, + "loss": 3.299, + "step": 3938 + }, + { + "epoch": 0.31580846244813693, + "grad_norm": 0.830882728099823, + "learning_rate": 1.9154367237950542e-05, + "loss": 2.5276, + "step": 3939 + }, + { + "epoch": 0.31588863722915955, + "grad_norm": 1.7939114570617676, + "learning_rate": 1.915384458527448e-05, + "loss": 3.1611, + "step": 3940 + }, + { + "epoch": 0.3159688120101822, + "grad_norm": 1.9778516292572021, + "learning_rate": 1.9153321778267827e-05, + "loss": 3.3754, + "step": 3941 + }, + { + "epoch": 0.3160489867912048, + "grad_norm": 2.046485185623169, + "learning_rate": 1.9152798816939397e-05, + "loss": 3.2544, + "step": 3942 + }, + { + "epoch": 0.3161291615722275, + "grad_norm": 1.8696320056915283, + "learning_rate": 1.9152275701298e-05, + "loss": 3.2323, + "step": 3943 + }, + { + "epoch": 0.3162093363532501, + "grad_norm": 1.9924615621566772, + "learning_rate": 1.915175243135246e-05, + "loss": 3.2746, + "step": 3944 + }, + { + "epoch": 0.3162895111342727, + "grad_norm": 1.9107283353805542, + "learning_rate": 1.9151229007111604e-05, + "loss": 3.1809, + "step": 3945 + }, + { + "epoch": 0.31636968591529535, + "grad_norm": 1.9195525646209717, + "learning_rate": 1.9150705428584248e-05, + "loss": 3.2227, + "step": 3946 + }, + { + "epoch": 0.316449860696318, + "grad_norm": 0.909594714641571, + "learning_rate": 1.9150181695779226e-05, + "loss": 2.5183, + "step": 3947 + }, + { + "epoch": 0.3165300354773406, + "grad_norm": 1.858176589012146, + "learning_rate": 1.9149657808705363e-05, + "loss": 3.2515, + "step": 3948 + }, + { + "epoch": 0.3166102102583632, + "grad_norm": 1.8029624223709106, + "learning_rate": 1.9149133767371494e-05, + "loss": 3.2312, + "step": 3949 + }, + { + "epoch": 0.31669038503938585, + "grad_norm": 1.9581269025802612, + "learning_rate": 1.9148609571786453e-05, + "loss": 3.2848, + "step": 3950 + }, + { + "epoch": 0.31677055982040847, + "grad_norm": 1.9368178844451904, + "learning_rate": 1.9148085221959085e-05, + "loss": 3.2693, + "step": 3951 + }, + { + "epoch": 0.3168507346014311, + "grad_norm": 1.939805269241333, + "learning_rate": 1.914756071789822e-05, + "loss": 3.3534, + "step": 3952 + }, + { + "epoch": 0.3169309093824538, + "grad_norm": 0.8284518718719482, + "learning_rate": 1.9147036059612705e-05, + "loss": 2.5223, + "step": 3953 + }, + { + "epoch": 0.3170110841634764, + "grad_norm": 2.0234599113464355, + "learning_rate": 1.9146511247111387e-05, + "loss": 3.2184, + "step": 3954 + }, + { + "epoch": 0.317091258944499, + "grad_norm": 1.9298349618911743, + "learning_rate": 1.914598628040311e-05, + "loss": 3.1757, + "step": 3955 + }, + { + "epoch": 0.31717143372552165, + "grad_norm": 1.889039158821106, + "learning_rate": 1.914546115949673e-05, + "loss": 3.2665, + "step": 3956 + }, + { + "epoch": 0.31725160850654427, + "grad_norm": 2.00331974029541, + "learning_rate": 1.91449358844011e-05, + "loss": 3.2018, + "step": 3957 + }, + { + "epoch": 0.3173317832875669, + "grad_norm": 1.929195523262024, + "learning_rate": 1.914441045512507e-05, + "loss": 3.2371, + "step": 3958 + }, + { + "epoch": 0.3174119580685895, + "grad_norm": 1.8104244470596313, + "learning_rate": 1.9143884871677504e-05, + "loss": 3.1995, + "step": 3959 + }, + { + "epoch": 0.31749213284961214, + "grad_norm": 2.062896966934204, + "learning_rate": 1.9143359134067262e-05, + "loss": 3.2162, + "step": 3960 + }, + { + "epoch": 0.31757230763063476, + "grad_norm": 1.9454967975616455, + "learning_rate": 1.9142833242303206e-05, + "loss": 3.2182, + "step": 3961 + }, + { + "epoch": 0.3176524824116574, + "grad_norm": 1.9001903533935547, + "learning_rate": 1.9142307196394208e-05, + "loss": 3.3441, + "step": 3962 + }, + { + "epoch": 0.31773265719268007, + "grad_norm": 1.8822143077850342, + "learning_rate": 1.914178099634913e-05, + "loss": 3.2496, + "step": 3963 + }, + { + "epoch": 0.3178128319737027, + "grad_norm": 2.000977039337158, + "learning_rate": 1.9141254642176847e-05, + "loss": 3.3862, + "step": 3964 + }, + { + "epoch": 0.3178930067547253, + "grad_norm": 1.9047678709030151, + "learning_rate": 1.914072813388623e-05, + "loss": 3.2512, + "step": 3965 + }, + { + "epoch": 0.31797318153574794, + "grad_norm": 0.8591116666793823, + "learning_rate": 1.9140201471486158e-05, + "loss": 2.5135, + "step": 3966 + }, + { + "epoch": 0.31805335631677056, + "grad_norm": 2.107240676879883, + "learning_rate": 1.9139674654985514e-05, + "loss": 3.3126, + "step": 3967 + }, + { + "epoch": 0.3181335310977932, + "grad_norm": 1.8293145895004272, + "learning_rate": 1.9139147684393175e-05, + "loss": 3.1267, + "step": 3968 + }, + { + "epoch": 0.3182137058788158, + "grad_norm": 2.0682175159454346, + "learning_rate": 1.9138620559718028e-05, + "loss": 3.2555, + "step": 3969 + }, + { + "epoch": 0.31829388065983844, + "grad_norm": 1.9182010889053345, + "learning_rate": 1.9138093280968957e-05, + "loss": 3.3333, + "step": 3970 + }, + { + "epoch": 0.31837405544086106, + "grad_norm": 1.9235725402832031, + "learning_rate": 1.9137565848154855e-05, + "loss": 3.1836, + "step": 3971 + }, + { + "epoch": 0.3184542302218837, + "grad_norm": 1.8694939613342285, + "learning_rate": 1.9137038261284613e-05, + "loss": 3.2446, + "step": 3972 + }, + { + "epoch": 0.31853440500290636, + "grad_norm": 2.2103264331817627, + "learning_rate": 1.9136510520367125e-05, + "loss": 3.2748, + "step": 3973 + }, + { + "epoch": 0.318614579783929, + "grad_norm": 1.9701811075210571, + "learning_rate": 1.9135982625411295e-05, + "loss": 3.2368, + "step": 3974 + }, + { + "epoch": 0.3186947545649516, + "grad_norm": 1.937252402305603, + "learning_rate": 1.913545457642601e-05, + "loss": 3.3333, + "step": 3975 + }, + { + "epoch": 0.31877492934597423, + "grad_norm": 1.9662916660308838, + "learning_rate": 1.9134926373420187e-05, + "loss": 3.4359, + "step": 3976 + }, + { + "epoch": 0.31885510412699686, + "grad_norm": 1.972182273864746, + "learning_rate": 1.913439801640272e-05, + "loss": 3.2764, + "step": 3977 + }, + { + "epoch": 0.3189352789080195, + "grad_norm": 2.2060422897338867, + "learning_rate": 1.913386950538252e-05, + "loss": 3.2069, + "step": 3978 + }, + { + "epoch": 0.3190154536890421, + "grad_norm": 1.97021484375, + "learning_rate": 1.9133340840368507e-05, + "loss": 3.2489, + "step": 3979 + }, + { + "epoch": 0.31909562847006473, + "grad_norm": 2.000838279724121, + "learning_rate": 1.913281202136958e-05, + "loss": 3.2037, + "step": 3980 + }, + { + "epoch": 0.31917580325108735, + "grad_norm": 1.87213933467865, + "learning_rate": 1.9132283048394665e-05, + "loss": 3.281, + "step": 3981 + }, + { + "epoch": 0.31925597803211, + "grad_norm": 2.2094368934631348, + "learning_rate": 1.9131753921452673e-05, + "loss": 3.3376, + "step": 3982 + }, + { + "epoch": 0.31933615281313266, + "grad_norm": 2.0300419330596924, + "learning_rate": 1.9131224640552532e-05, + "loss": 3.2617, + "step": 3983 + }, + { + "epoch": 0.3194163275941553, + "grad_norm": 1.8817319869995117, + "learning_rate": 1.9130695205703157e-05, + "loss": 3.1694, + "step": 3984 + }, + { + "epoch": 0.3194965023751779, + "grad_norm": 1.9831349849700928, + "learning_rate": 1.9130165616913482e-05, + "loss": 3.1956, + "step": 3985 + }, + { + "epoch": 0.31957667715620053, + "grad_norm": 1.921326994895935, + "learning_rate": 1.912963587419243e-05, + "loss": 3.2526, + "step": 3986 + }, + { + "epoch": 0.31965685193722315, + "grad_norm": 1.9240220785140991, + "learning_rate": 1.9129105977548937e-05, + "loss": 3.3159, + "step": 3987 + }, + { + "epoch": 0.3197370267182458, + "grad_norm": 2.371868133544922, + "learning_rate": 1.9128575926991936e-05, + "loss": 3.3692, + "step": 3988 + }, + { + "epoch": 0.3198172014992684, + "grad_norm": 0.8776980638504028, + "learning_rate": 1.912804572253036e-05, + "loss": 2.5171, + "step": 3989 + }, + { + "epoch": 0.319897376280291, + "grad_norm": 2.064377546310425, + "learning_rate": 1.912751536417315e-05, + "loss": 3.1849, + "step": 3990 + }, + { + "epoch": 0.31997755106131365, + "grad_norm": 2.1421046257019043, + "learning_rate": 1.912698485192925e-05, + "loss": 3.2737, + "step": 3991 + }, + { + "epoch": 0.32005772584233627, + "grad_norm": 2.1782991886138916, + "learning_rate": 1.91264541858076e-05, + "loss": 3.3574, + "step": 3992 + }, + { + "epoch": 0.3201379006233589, + "grad_norm": 1.9192150831222534, + "learning_rate": 1.9125923365817147e-05, + "loss": 3.2526, + "step": 3993 + }, + { + "epoch": 0.3202180754043816, + "grad_norm": 2.1904208660125732, + "learning_rate": 1.9125392391966847e-05, + "loss": 3.2299, + "step": 3994 + }, + { + "epoch": 0.3202982501854042, + "grad_norm": 2.1338369846343994, + "learning_rate": 1.9124861264265644e-05, + "loss": 3.1864, + "step": 3995 + }, + { + "epoch": 0.3203784249664268, + "grad_norm": 1.8919702768325806, + "learning_rate": 1.9124329982722498e-05, + "loss": 3.3434, + "step": 3996 + }, + { + "epoch": 0.32045859974744945, + "grad_norm": 1.984487533569336, + "learning_rate": 1.912379854734636e-05, + "loss": 3.269, + "step": 3997 + }, + { + "epoch": 0.32053877452847207, + "grad_norm": 1.9277886152267456, + "learning_rate": 1.9123266958146195e-05, + "loss": 3.2887, + "step": 3998 + }, + { + "epoch": 0.3206189493094947, + "grad_norm": 1.958608627319336, + "learning_rate": 1.912273521513097e-05, + "loss": 3.2288, + "step": 3999 + }, + { + "epoch": 0.3206991240905173, + "grad_norm": 1.9119442701339722, + "learning_rate": 1.912220331830964e-05, + "loss": 3.1477, + "step": 4000 + }, + { + "epoch": 0.32077929887153994, + "grad_norm": 1.984132170677185, + "learning_rate": 1.9121671267691177e-05, + "loss": 3.3746, + "step": 4001 + }, + { + "epoch": 0.32085947365256257, + "grad_norm": 2.098263740539551, + "learning_rate": 1.9121139063284554e-05, + "loss": 3.2623, + "step": 4002 + }, + { + "epoch": 0.3209396484335852, + "grad_norm": 1.9762177467346191, + "learning_rate": 1.9120606705098737e-05, + "loss": 3.2046, + "step": 4003 + }, + { + "epoch": 0.32101982321460787, + "grad_norm": 1.8267244100570679, + "learning_rate": 1.912007419314271e-05, + "loss": 3.2216, + "step": 4004 + }, + { + "epoch": 0.3210999979956305, + "grad_norm": 1.9710570573806763, + "learning_rate": 1.9119541527425442e-05, + "loss": 3.2607, + "step": 4005 + }, + { + "epoch": 0.3211801727766531, + "grad_norm": 2.1302378177642822, + "learning_rate": 1.911900870795592e-05, + "loss": 3.3921, + "step": 4006 + }, + { + "epoch": 0.32126034755767574, + "grad_norm": 2.0549821853637695, + "learning_rate": 1.9118475734743126e-05, + "loss": 3.2341, + "step": 4007 + }, + { + "epoch": 0.32134052233869836, + "grad_norm": 2.1437718868255615, + "learning_rate": 1.9117942607796044e-05, + "loss": 3.2491, + "step": 4008 + }, + { + "epoch": 0.321420697119721, + "grad_norm": 1.8747217655181885, + "learning_rate": 1.9117409327123662e-05, + "loss": 3.231, + "step": 4009 + }, + { + "epoch": 0.3215008719007436, + "grad_norm": 1.8455387353897095, + "learning_rate": 1.9116875892734974e-05, + "loss": 3.2183, + "step": 4010 + }, + { + "epoch": 0.32158104668176624, + "grad_norm": 1.9988974332809448, + "learning_rate": 1.9116342304638967e-05, + "loss": 3.2302, + "step": 4011 + }, + { + "epoch": 0.32166122146278886, + "grad_norm": 1.8778868913650513, + "learning_rate": 1.9115808562844648e-05, + "loss": 3.1979, + "step": 4012 + }, + { + "epoch": 0.3217413962438115, + "grad_norm": 1.953726887702942, + "learning_rate": 1.9115274667361005e-05, + "loss": 3.2849, + "step": 4013 + }, + { + "epoch": 0.32182157102483416, + "grad_norm": 2.03391432762146, + "learning_rate": 1.9114740618197047e-05, + "loss": 3.309, + "step": 4014 + }, + { + "epoch": 0.3219017458058568, + "grad_norm": 1.8237638473510742, + "learning_rate": 1.911420641536177e-05, + "loss": 3.2234, + "step": 4015 + }, + { + "epoch": 0.3219819205868794, + "grad_norm": 1.876570463180542, + "learning_rate": 1.911367205886419e-05, + "loss": 3.3391, + "step": 4016 + }, + { + "epoch": 0.32206209536790203, + "grad_norm": 1.9357993602752686, + "learning_rate": 1.9113137548713306e-05, + "loss": 3.1062, + "step": 4017 + }, + { + "epoch": 0.32214227014892466, + "grad_norm": 2.0273916721343994, + "learning_rate": 1.9112602884918135e-05, + "loss": 3.3062, + "step": 4018 + }, + { + "epoch": 0.3222224449299473, + "grad_norm": 1.8574843406677246, + "learning_rate": 1.9112068067487698e-05, + "loss": 3.2231, + "step": 4019 + }, + { + "epoch": 0.3223026197109699, + "grad_norm": 2.061768054962158, + "learning_rate": 1.9111533096431e-05, + "loss": 3.2715, + "step": 4020 + }, + { + "epoch": 0.32238279449199253, + "grad_norm": 1.810805320739746, + "learning_rate": 1.911099797175706e-05, + "loss": 3.2653, + "step": 4021 + }, + { + "epoch": 0.32246296927301515, + "grad_norm": 1.9458271265029907, + "learning_rate": 1.9110462693474912e-05, + "loss": 3.2671, + "step": 4022 + }, + { + "epoch": 0.3225431440540378, + "grad_norm": 2.026522397994995, + "learning_rate": 1.9109927261593575e-05, + "loss": 3.1908, + "step": 4023 + }, + { + "epoch": 0.32262331883506046, + "grad_norm": 1.84330153465271, + "learning_rate": 1.910939167612207e-05, + "loss": 3.2266, + "step": 4024 + }, + { + "epoch": 0.3227034936160831, + "grad_norm": 1.8834437131881714, + "learning_rate": 1.9108855937069435e-05, + "loss": 3.2327, + "step": 4025 + }, + { + "epoch": 0.3227836683971057, + "grad_norm": 2.145433187484741, + "learning_rate": 1.91083200444447e-05, + "loss": 3.3215, + "step": 4026 + }, + { + "epoch": 0.32286384317812833, + "grad_norm": 1.9248862266540527, + "learning_rate": 1.9107783998256897e-05, + "loss": 3.2017, + "step": 4027 + }, + { + "epoch": 0.32294401795915095, + "grad_norm": 0.8949766755104065, + "learning_rate": 1.9107247798515064e-05, + "loss": 2.5208, + "step": 4028 + }, + { + "epoch": 0.3230241927401736, + "grad_norm": 2.058993339538574, + "learning_rate": 1.9106711445228247e-05, + "loss": 3.2549, + "step": 4029 + }, + { + "epoch": 0.3231043675211962, + "grad_norm": 2.015247344970703, + "learning_rate": 1.910617493840548e-05, + "loss": 3.3208, + "step": 4030 + }, + { + "epoch": 0.3231845423022188, + "grad_norm": 1.9229426383972168, + "learning_rate": 1.9105638278055817e-05, + "loss": 3.2756, + "step": 4031 + }, + { + "epoch": 0.32326471708324145, + "grad_norm": 2.197577714920044, + "learning_rate": 1.91051014641883e-05, + "loss": 3.2151, + "step": 4032 + }, + { + "epoch": 0.3233448918642641, + "grad_norm": 1.9496123790740967, + "learning_rate": 1.910456449681198e-05, + "loss": 3.1787, + "step": 4033 + }, + { + "epoch": 0.32342506664528675, + "grad_norm": 1.8682341575622559, + "learning_rate": 1.910402737593591e-05, + "loss": 3.2805, + "step": 4034 + }, + { + "epoch": 0.3235052414263094, + "grad_norm": 2.0536022186279297, + "learning_rate": 1.910349010156915e-05, + "loss": 3.3202, + "step": 4035 + }, + { + "epoch": 0.323585416207332, + "grad_norm": 1.9728161096572876, + "learning_rate": 1.910295267372076e-05, + "loss": 3.3142, + "step": 4036 + }, + { + "epoch": 0.3236655909883546, + "grad_norm": 0.9275935888290405, + "learning_rate": 1.9102415092399788e-05, + "loss": 2.5107, + "step": 4037 + }, + { + "epoch": 0.32374576576937725, + "grad_norm": 2.0297093391418457, + "learning_rate": 1.910187735761531e-05, + "loss": 3.1979, + "step": 4038 + }, + { + "epoch": 0.32382594055039987, + "grad_norm": 2.12320876121521, + "learning_rate": 1.9101339469376385e-05, + "loss": 3.3833, + "step": 4039 + }, + { + "epoch": 0.3239061153314225, + "grad_norm": 0.9170796871185303, + "learning_rate": 1.9100801427692083e-05, + "loss": 2.5328, + "step": 4040 + }, + { + "epoch": 0.3239862901124451, + "grad_norm": 2.0118815898895264, + "learning_rate": 1.9100263232571482e-05, + "loss": 3.2682, + "step": 4041 + }, + { + "epoch": 0.32406646489346774, + "grad_norm": 1.8806120157241821, + "learning_rate": 1.9099724884023647e-05, + "loss": 3.3289, + "step": 4042 + }, + { + "epoch": 0.32414663967449037, + "grad_norm": 1.926888346672058, + "learning_rate": 1.9099186382057657e-05, + "loss": 3.2689, + "step": 4043 + }, + { + "epoch": 0.32422681445551305, + "grad_norm": 2.004946231842041, + "learning_rate": 1.9098647726682592e-05, + "loss": 3.2963, + "step": 4044 + }, + { + "epoch": 0.32430698923653567, + "grad_norm": 1.8734447956085205, + "learning_rate": 1.9098108917907532e-05, + "loss": 3.2222, + "step": 4045 + }, + { + "epoch": 0.3243871640175583, + "grad_norm": 1.8987663984298706, + "learning_rate": 1.9097569955741562e-05, + "loss": 3.2448, + "step": 4046 + }, + { + "epoch": 0.3244673387985809, + "grad_norm": 2.0499210357666016, + "learning_rate": 1.909703084019377e-05, + "loss": 3.2094, + "step": 4047 + }, + { + "epoch": 0.32454751357960354, + "grad_norm": 1.8275845050811768, + "learning_rate": 1.9096491571273244e-05, + "loss": 3.2368, + "step": 4048 + }, + { + "epoch": 0.32462768836062617, + "grad_norm": 2.039437770843506, + "learning_rate": 1.9095952148989077e-05, + "loss": 3.3634, + "step": 4049 + }, + { + "epoch": 0.3247078631416488, + "grad_norm": 1.9354832172393799, + "learning_rate": 1.909541257335036e-05, + "loss": 3.2469, + "step": 4050 + }, + { + "epoch": 0.3247880379226714, + "grad_norm": 1.8279320001602173, + "learning_rate": 1.9094872844366195e-05, + "loss": 3.2427, + "step": 4051 + }, + { + "epoch": 0.32486821270369404, + "grad_norm": 1.9523504972457886, + "learning_rate": 1.9094332962045675e-05, + "loss": 3.2655, + "step": 4052 + }, + { + "epoch": 0.32494838748471666, + "grad_norm": 1.8731943368911743, + "learning_rate": 1.9093792926397913e-05, + "loss": 3.3517, + "step": 4053 + }, + { + "epoch": 0.3250285622657393, + "grad_norm": 0.8607675433158875, + "learning_rate": 1.9093252737432e-05, + "loss": 2.515, + "step": 4054 + }, + { + "epoch": 0.32510873704676196, + "grad_norm": 2.0615861415863037, + "learning_rate": 1.9092712395157056e-05, + "loss": 3.2238, + "step": 4055 + }, + { + "epoch": 0.3251889118277846, + "grad_norm": 1.9249851703643799, + "learning_rate": 1.9092171899582188e-05, + "loss": 3.2504, + "step": 4056 + }, + { + "epoch": 0.3252690866088072, + "grad_norm": 1.9680140018463135, + "learning_rate": 1.90916312507165e-05, + "loss": 3.2898, + "step": 4057 + }, + { + "epoch": 0.32534926138982984, + "grad_norm": 1.964523196220398, + "learning_rate": 1.9091090448569116e-05, + "loss": 3.27, + "step": 4058 + }, + { + "epoch": 0.32542943617085246, + "grad_norm": 2.0316946506500244, + "learning_rate": 1.9090549493149152e-05, + "loss": 3.3311, + "step": 4059 + }, + { + "epoch": 0.3255096109518751, + "grad_norm": 1.8160362243652344, + "learning_rate": 1.9090008384465724e-05, + "loss": 3.1642, + "step": 4060 + }, + { + "epoch": 0.3255897857328977, + "grad_norm": 1.9452582597732544, + "learning_rate": 1.9089467122527958e-05, + "loss": 3.2908, + "step": 4061 + }, + { + "epoch": 0.32566996051392033, + "grad_norm": 1.9312852621078491, + "learning_rate": 1.9088925707344984e-05, + "loss": 3.2908, + "step": 4062 + }, + { + "epoch": 0.32575013529494296, + "grad_norm": 1.9588499069213867, + "learning_rate": 1.908838413892592e-05, + "loss": 3.2162, + "step": 4063 + }, + { + "epoch": 0.3258303100759656, + "grad_norm": 2.0773990154266357, + "learning_rate": 1.9087842417279908e-05, + "loss": 3.291, + "step": 4064 + }, + { + "epoch": 0.32591048485698826, + "grad_norm": 1.9528840780258179, + "learning_rate": 1.9087300542416077e-05, + "loss": 3.2599, + "step": 4065 + }, + { + "epoch": 0.3259906596380109, + "grad_norm": 0.9823713302612305, + "learning_rate": 1.9086758514343556e-05, + "loss": 2.5061, + "step": 4066 + }, + { + "epoch": 0.3260708344190335, + "grad_norm": 1.895416498184204, + "learning_rate": 1.9086216333071492e-05, + "loss": 3.2344, + "step": 4067 + }, + { + "epoch": 0.32615100920005613, + "grad_norm": 0.8461055755615234, + "learning_rate": 1.908567399860902e-05, + "loss": 2.5255, + "step": 4068 + }, + { + "epoch": 0.32623118398107875, + "grad_norm": 1.9717719554901123, + "learning_rate": 1.9085131510965293e-05, + "loss": 3.3366, + "step": 4069 + }, + { + "epoch": 0.3263113587621014, + "grad_norm": 1.8824479579925537, + "learning_rate": 1.9084588870149443e-05, + "loss": 3.2744, + "step": 4070 + }, + { + "epoch": 0.326391533543124, + "grad_norm": 2.0051615238189697, + "learning_rate": 1.9084046076170632e-05, + "loss": 3.3188, + "step": 4071 + }, + { + "epoch": 0.3264717083241466, + "grad_norm": 1.9621719121932983, + "learning_rate": 1.9083503129038005e-05, + "loss": 3.249, + "step": 4072 + }, + { + "epoch": 0.32655188310516925, + "grad_norm": 1.940922498703003, + "learning_rate": 1.9082960028760713e-05, + "loss": 3.3077, + "step": 4073 + }, + { + "epoch": 0.3266320578861919, + "grad_norm": 2.0373470783233643, + "learning_rate": 1.9082416775347918e-05, + "loss": 3.2412, + "step": 4074 + }, + { + "epoch": 0.32671223266721455, + "grad_norm": 2.095144748687744, + "learning_rate": 1.9081873368808778e-05, + "loss": 3.3212, + "step": 4075 + }, + { + "epoch": 0.3267924074482372, + "grad_norm": 0.9657623171806335, + "learning_rate": 1.9081329809152454e-05, + "loss": 2.5441, + "step": 4076 + }, + { + "epoch": 0.3268725822292598, + "grad_norm": 1.8987069129943848, + "learning_rate": 1.9080786096388108e-05, + "loss": 3.2292, + "step": 4077 + }, + { + "epoch": 0.3269527570102824, + "grad_norm": 2.0627684593200684, + "learning_rate": 1.908024223052491e-05, + "loss": 3.2582, + "step": 4078 + }, + { + "epoch": 0.32703293179130505, + "grad_norm": 2.081825017929077, + "learning_rate": 1.907969821157203e-05, + "loss": 3.3085, + "step": 4079 + }, + { + "epoch": 0.32711310657232767, + "grad_norm": 1.8775413036346436, + "learning_rate": 1.9079154039538633e-05, + "loss": 3.2556, + "step": 4080 + }, + { + "epoch": 0.3271932813533503, + "grad_norm": 1.8829246759414673, + "learning_rate": 1.9078609714433902e-05, + "loss": 3.162, + "step": 4081 + }, + { + "epoch": 0.3272734561343729, + "grad_norm": 1.9406505823135376, + "learning_rate": 1.907806523626701e-05, + "loss": 3.3405, + "step": 4082 + }, + { + "epoch": 0.32735363091539554, + "grad_norm": 2.0508792400360107, + "learning_rate": 1.9077520605047137e-05, + "loss": 3.2907, + "step": 4083 + }, + { + "epoch": 0.32743380569641817, + "grad_norm": 1.884426474571228, + "learning_rate": 1.9076975820783466e-05, + "loss": 3.2628, + "step": 4084 + }, + { + "epoch": 0.32751398047744085, + "grad_norm": 2.104001998901367, + "learning_rate": 1.9076430883485188e-05, + "loss": 3.243, + "step": 4085 + }, + { + "epoch": 0.32759415525846347, + "grad_norm": 1.8733597993850708, + "learning_rate": 1.9075885793161476e-05, + "loss": 3.3286, + "step": 4086 + }, + { + "epoch": 0.3276743300394861, + "grad_norm": 1.7998629808425903, + "learning_rate": 1.9075340549821534e-05, + "loss": 3.3367, + "step": 4087 + }, + { + "epoch": 0.3277545048205087, + "grad_norm": 1.9740078449249268, + "learning_rate": 1.9074795153474542e-05, + "loss": 3.2141, + "step": 4088 + }, + { + "epoch": 0.32783467960153134, + "grad_norm": 1.9555858373641968, + "learning_rate": 1.9074249604129707e-05, + "loss": 3.2112, + "step": 4089 + }, + { + "epoch": 0.32791485438255397, + "grad_norm": 1.9146617650985718, + "learning_rate": 1.907370390179622e-05, + "loss": 3.2496, + "step": 4090 + }, + { + "epoch": 0.3279950291635766, + "grad_norm": 1.8501415252685547, + "learning_rate": 1.9073158046483285e-05, + "loss": 3.3376, + "step": 4091 + }, + { + "epoch": 0.3280752039445992, + "grad_norm": 1.9170705080032349, + "learning_rate": 1.90726120382001e-05, + "loss": 3.2514, + "step": 4092 + }, + { + "epoch": 0.32815537872562184, + "grad_norm": 1.9424045085906982, + "learning_rate": 1.9072065876955878e-05, + "loss": 3.2795, + "step": 4093 + }, + { + "epoch": 0.32823555350664446, + "grad_norm": 1.962353229522705, + "learning_rate": 1.907151956275982e-05, + "loss": 3.2751, + "step": 4094 + }, + { + "epoch": 0.32831572828766714, + "grad_norm": 2.1350741386413574, + "learning_rate": 1.9070973095621136e-05, + "loss": 3.3359, + "step": 4095 + }, + { + "epoch": 0.32839590306868977, + "grad_norm": 1.9610991477966309, + "learning_rate": 1.9070426475549048e-05, + "loss": 3.2782, + "step": 4096 + }, + { + "epoch": 0.3284760778497124, + "grad_norm": 1.9526633024215698, + "learning_rate": 1.9069879702552763e-05, + "loss": 3.1691, + "step": 4097 + }, + { + "epoch": 0.328556252630735, + "grad_norm": 0.8975244164466858, + "learning_rate": 1.9069332776641505e-05, + "loss": 2.5219, + "step": 4098 + }, + { + "epoch": 0.32863642741175764, + "grad_norm": 1.9292477369308472, + "learning_rate": 1.906878569782449e-05, + "loss": 3.2515, + "step": 4099 + }, + { + "epoch": 0.32871660219278026, + "grad_norm": 1.9866831302642822, + "learning_rate": 1.9068238466110944e-05, + "loss": 3.2145, + "step": 4100 + }, + { + "epoch": 0.3287967769738029, + "grad_norm": 2.066164493560791, + "learning_rate": 1.9067691081510098e-05, + "loss": 3.3071, + "step": 4101 + }, + { + "epoch": 0.3288769517548255, + "grad_norm": 1.9767537117004395, + "learning_rate": 1.9067143544031173e-05, + "loss": 3.2347, + "step": 4102 + }, + { + "epoch": 0.32895712653584813, + "grad_norm": 1.925125002861023, + "learning_rate": 1.9066595853683404e-05, + "loss": 3.1786, + "step": 4103 + }, + { + "epoch": 0.32903730131687076, + "grad_norm": 1.984283208847046, + "learning_rate": 1.9066048010476024e-05, + "loss": 3.2764, + "step": 4104 + }, + { + "epoch": 0.3291174760978934, + "grad_norm": 1.9480317831039429, + "learning_rate": 1.906550001441827e-05, + "loss": 3.3267, + "step": 4105 + }, + { + "epoch": 0.32919765087891606, + "grad_norm": 1.9142823219299316, + "learning_rate": 1.906495186551938e-05, + "loss": 3.1573, + "step": 4106 + }, + { + "epoch": 0.3292778256599387, + "grad_norm": 1.9006037712097168, + "learning_rate": 1.9064403563788597e-05, + "loss": 3.1857, + "step": 4107 + }, + { + "epoch": 0.3293580004409613, + "grad_norm": 1.8022528886795044, + "learning_rate": 1.9063855109235164e-05, + "loss": 3.2132, + "step": 4108 + }, + { + "epoch": 0.32943817522198393, + "grad_norm": 1.8572440147399902, + "learning_rate": 1.906330650186833e-05, + "loss": 3.2013, + "step": 4109 + }, + { + "epoch": 0.32951835000300655, + "grad_norm": 1.895451307296753, + "learning_rate": 1.9062757741697338e-05, + "loss": 3.2817, + "step": 4110 + }, + { + "epoch": 0.3295985247840292, + "grad_norm": 1.8866065740585327, + "learning_rate": 1.906220882873145e-05, + "loss": 3.1135, + "step": 4111 + }, + { + "epoch": 0.3296786995650518, + "grad_norm": 1.843314528465271, + "learning_rate": 1.906165976297991e-05, + "loss": 3.2708, + "step": 4112 + }, + { + "epoch": 0.3297588743460744, + "grad_norm": 1.889666199684143, + "learning_rate": 1.9061110544451982e-05, + "loss": 3.2649, + "step": 4113 + }, + { + "epoch": 0.32983904912709705, + "grad_norm": 2.00162935256958, + "learning_rate": 1.9060561173156928e-05, + "loss": 3.2811, + "step": 4114 + }, + { + "epoch": 0.3299192239081197, + "grad_norm": 2.0013725757598877, + "learning_rate": 1.9060011649104e-05, + "loss": 3.2198, + "step": 4115 + }, + { + "epoch": 0.32999939868914235, + "grad_norm": 1.9419302940368652, + "learning_rate": 1.9059461972302472e-05, + "loss": 3.2755, + "step": 4116 + }, + { + "epoch": 0.330079573470165, + "grad_norm": 1.8570917844772339, + "learning_rate": 1.9058912142761607e-05, + "loss": 3.2076, + "step": 4117 + }, + { + "epoch": 0.3301597482511876, + "grad_norm": 2.0382654666900635, + "learning_rate": 1.9058362160490676e-05, + "loss": 3.3452, + "step": 4118 + }, + { + "epoch": 0.3302399230322102, + "grad_norm": 2.0147931575775146, + "learning_rate": 1.9057812025498953e-05, + "loss": 3.2976, + "step": 4119 + }, + { + "epoch": 0.33032009781323285, + "grad_norm": 1.8991518020629883, + "learning_rate": 1.905726173779571e-05, + "loss": 3.2518, + "step": 4120 + }, + { + "epoch": 0.3304002725942555, + "grad_norm": 2.183889150619507, + "learning_rate": 1.9056711297390225e-05, + "loss": 3.3248, + "step": 4121 + }, + { + "epoch": 0.3304804473752781, + "grad_norm": 1.8575677871704102, + "learning_rate": 1.905616070429178e-05, + "loss": 3.2638, + "step": 4122 + }, + { + "epoch": 0.3305606221563007, + "grad_norm": 2.3117806911468506, + "learning_rate": 1.9055609958509658e-05, + "loss": 3.2565, + "step": 4123 + }, + { + "epoch": 0.33064079693732334, + "grad_norm": 0.8967247009277344, + "learning_rate": 1.9055059060053143e-05, + "loss": 2.5199, + "step": 4124 + }, + { + "epoch": 0.33072097171834597, + "grad_norm": 1.7987337112426758, + "learning_rate": 1.9054508008931526e-05, + "loss": 3.1768, + "step": 4125 + }, + { + "epoch": 0.33080114649936865, + "grad_norm": 1.9798110723495483, + "learning_rate": 1.905395680515409e-05, + "loss": 3.1725, + "step": 4126 + }, + { + "epoch": 0.33088132128039127, + "grad_norm": 1.9465582370758057, + "learning_rate": 1.9053405448730137e-05, + "loss": 3.2351, + "step": 4127 + }, + { + "epoch": 0.3309614960614139, + "grad_norm": 0.935319185256958, + "learning_rate": 1.905285393966896e-05, + "loss": 2.5214, + "step": 4128 + }, + { + "epoch": 0.3310416708424365, + "grad_norm": 2.0126302242279053, + "learning_rate": 1.905230227797985e-05, + "loss": 3.41, + "step": 4129 + }, + { + "epoch": 0.33112184562345914, + "grad_norm": 2.0413341522216797, + "learning_rate": 1.905175046367212e-05, + "loss": 3.1594, + "step": 4130 + }, + { + "epoch": 0.33120202040448177, + "grad_norm": 2.2132177352905273, + "learning_rate": 1.9051198496755065e-05, + "loss": 3.339, + "step": 4131 + }, + { + "epoch": 0.3312821951855044, + "grad_norm": 2.0477967262268066, + "learning_rate": 1.9050646377237994e-05, + "loss": 3.1819, + "step": 4132 + }, + { + "epoch": 0.331362369966527, + "grad_norm": 1.853579044342041, + "learning_rate": 1.9050094105130215e-05, + "loss": 3.3171, + "step": 4133 + }, + { + "epoch": 0.33144254474754964, + "grad_norm": 1.9351285696029663, + "learning_rate": 1.9049541680441038e-05, + "loss": 3.252, + "step": 4134 + }, + { + "epoch": 0.33152271952857226, + "grad_norm": 2.1759114265441895, + "learning_rate": 1.9048989103179777e-05, + "loss": 3.2395, + "step": 4135 + }, + { + "epoch": 0.33160289430959494, + "grad_norm": 1.9434101581573486, + "learning_rate": 1.9048436373355752e-05, + "loss": 3.2006, + "step": 4136 + }, + { + "epoch": 0.33168306909061757, + "grad_norm": 1.9240435361862183, + "learning_rate": 1.9047883490978275e-05, + "loss": 3.3505, + "step": 4137 + }, + { + "epoch": 0.3317632438716402, + "grad_norm": 2.0838816165924072, + "learning_rate": 1.9047330456056673e-05, + "loss": 3.3181, + "step": 4138 + }, + { + "epoch": 0.3318434186526628, + "grad_norm": 1.9144959449768066, + "learning_rate": 1.904677726860027e-05, + "loss": 3.2677, + "step": 4139 + }, + { + "epoch": 0.33192359343368544, + "grad_norm": 1.9394205808639526, + "learning_rate": 1.904622392861839e-05, + "loss": 3.251, + "step": 4140 + }, + { + "epoch": 0.33200376821470806, + "grad_norm": 1.9349241256713867, + "learning_rate": 1.9045670436120362e-05, + "loss": 3.2281, + "step": 4141 + }, + { + "epoch": 0.3320839429957307, + "grad_norm": 2.0493414402008057, + "learning_rate": 1.9045116791115514e-05, + "loss": 3.235, + "step": 4142 + }, + { + "epoch": 0.3321641177767533, + "grad_norm": 1.9569944143295288, + "learning_rate": 1.9044562993613188e-05, + "loss": 3.3092, + "step": 4143 + }, + { + "epoch": 0.33224429255777593, + "grad_norm": 2.1727335453033447, + "learning_rate": 1.904400904362272e-05, + "loss": 3.2607, + "step": 4144 + }, + { + "epoch": 0.33232446733879856, + "grad_norm": 2.0722603797912598, + "learning_rate": 1.9043454941153445e-05, + "loss": 3.3063, + "step": 4145 + }, + { + "epoch": 0.33240464211982124, + "grad_norm": 2.0578572750091553, + "learning_rate": 1.9042900686214706e-05, + "loss": 3.3296, + "step": 4146 + }, + { + "epoch": 0.33248481690084386, + "grad_norm": 2.636983871459961, + "learning_rate": 1.9042346278815848e-05, + "loss": 3.3898, + "step": 4147 + }, + { + "epoch": 0.3325649916818665, + "grad_norm": 2.0479896068573, + "learning_rate": 1.904179171896622e-05, + "loss": 3.3025, + "step": 4148 + }, + { + "epoch": 0.3326451664628891, + "grad_norm": 1.9649914503097534, + "learning_rate": 1.9041237006675173e-05, + "loss": 3.1459, + "step": 4149 + }, + { + "epoch": 0.33272534124391173, + "grad_norm": 1.9061081409454346, + "learning_rate": 1.9040682141952052e-05, + "loss": 3.351, + "step": 4150 + }, + { + "epoch": 0.33280551602493436, + "grad_norm": 1.8829230070114136, + "learning_rate": 1.904012712480622e-05, + "loss": 3.2448, + "step": 4151 + }, + { + "epoch": 0.332885690805957, + "grad_norm": 2.0045177936553955, + "learning_rate": 1.9039571955247028e-05, + "loss": 3.3751, + "step": 4152 + }, + { + "epoch": 0.3329658655869796, + "grad_norm": 1.9554473161697388, + "learning_rate": 1.9039016633283836e-05, + "loss": 3.172, + "step": 4153 + }, + { + "epoch": 0.3330460403680022, + "grad_norm": 1.996447205543518, + "learning_rate": 1.9038461158926014e-05, + "loss": 3.2599, + "step": 4154 + }, + { + "epoch": 0.33312621514902485, + "grad_norm": 2.030792236328125, + "learning_rate": 1.9037905532182923e-05, + "loss": 3.3425, + "step": 4155 + }, + { + "epoch": 0.33320638993004753, + "grad_norm": 1.9718327522277832, + "learning_rate": 1.9037349753063927e-05, + "loss": 3.2573, + "step": 4156 + }, + { + "epoch": 0.33328656471107015, + "grad_norm": 2.0071120262145996, + "learning_rate": 1.9036793821578402e-05, + "loss": 3.2628, + "step": 4157 + }, + { + "epoch": 0.3333667394920928, + "grad_norm": 1.8595263957977295, + "learning_rate": 1.9036237737735716e-05, + "loss": 3.2416, + "step": 4158 + }, + { + "epoch": 0.3334469142731154, + "grad_norm": 1.9504207372665405, + "learning_rate": 1.9035681501545243e-05, + "loss": 3.1429, + "step": 4159 + }, + { + "epoch": 0.333527089054138, + "grad_norm": 1.821738839149475, + "learning_rate": 1.903512511301637e-05, + "loss": 3.1624, + "step": 4160 + }, + { + "epoch": 0.33360726383516065, + "grad_norm": 1.979789137840271, + "learning_rate": 1.9034568572158472e-05, + "loss": 3.3171, + "step": 4161 + }, + { + "epoch": 0.3336874386161833, + "grad_norm": 1.960848093032837, + "learning_rate": 1.903401187898093e-05, + "loss": 3.3357, + "step": 4162 + }, + { + "epoch": 0.3337676133972059, + "grad_norm": 0.8718219995498657, + "learning_rate": 1.903345503349313e-05, + "loss": 2.5197, + "step": 4163 + }, + { + "epoch": 0.3338477881782285, + "grad_norm": 2.2782628536224365, + "learning_rate": 1.9032898035704465e-05, + "loss": 3.3163, + "step": 4164 + }, + { + "epoch": 0.33392796295925115, + "grad_norm": 2.039733409881592, + "learning_rate": 1.903234088562432e-05, + "loss": 3.2074, + "step": 4165 + }, + { + "epoch": 0.33400813774027377, + "grad_norm": 1.9300884008407593, + "learning_rate": 1.903178358326209e-05, + "loss": 3.1272, + "step": 4166 + }, + { + "epoch": 0.33408831252129645, + "grad_norm": 1.978061318397522, + "learning_rate": 1.9031226128627174e-05, + "loss": 3.2667, + "step": 4167 + }, + { + "epoch": 0.3341684873023191, + "grad_norm": 1.887509822845459, + "learning_rate": 1.903066852172897e-05, + "loss": 3.2129, + "step": 4168 + }, + { + "epoch": 0.3342486620833417, + "grad_norm": 2.114067792892456, + "learning_rate": 1.9030110762576878e-05, + "loss": 3.2512, + "step": 4169 + }, + { + "epoch": 0.3343288368643643, + "grad_norm": 1.941856861114502, + "learning_rate": 1.90295528511803e-05, + "loss": 3.2125, + "step": 4170 + }, + { + "epoch": 0.33440901164538694, + "grad_norm": 2.1383843421936035, + "learning_rate": 1.902899478754864e-05, + "loss": 3.2546, + "step": 4171 + }, + { + "epoch": 0.33448918642640957, + "grad_norm": 1.9523745775222778, + "learning_rate": 1.9028436571691312e-05, + "loss": 3.1984, + "step": 4172 + }, + { + "epoch": 0.3345693612074322, + "grad_norm": 1.980697512626648, + "learning_rate": 1.9027878203617726e-05, + "loss": 3.4023, + "step": 4173 + }, + { + "epoch": 0.3346495359884548, + "grad_norm": 2.03855299949646, + "learning_rate": 1.902731968333729e-05, + "loss": 3.3743, + "step": 4174 + }, + { + "epoch": 0.33472971076947744, + "grad_norm": 2.1094706058502197, + "learning_rate": 1.9026761010859432e-05, + "loss": 3.3416, + "step": 4175 + }, + { + "epoch": 0.33480988555050006, + "grad_norm": 1.8578190803527832, + "learning_rate": 1.9026202186193562e-05, + "loss": 3.3524, + "step": 4176 + }, + { + "epoch": 0.33489006033152274, + "grad_norm": 1.8930827379226685, + "learning_rate": 1.9025643209349106e-05, + "loss": 3.2549, + "step": 4177 + }, + { + "epoch": 0.33497023511254537, + "grad_norm": 1.9801477193832397, + "learning_rate": 1.9025084080335486e-05, + "loss": 3.4057, + "step": 4178 + }, + { + "epoch": 0.335050409893568, + "grad_norm": 1.8594160079956055, + "learning_rate": 1.902452479916213e-05, + "loss": 3.2871, + "step": 4179 + }, + { + "epoch": 0.3351305846745906, + "grad_norm": 1.956687331199646, + "learning_rate": 1.9023965365838463e-05, + "loss": 3.2031, + "step": 4180 + }, + { + "epoch": 0.33521075945561324, + "grad_norm": 2.1666266918182373, + "learning_rate": 1.9023405780373926e-05, + "loss": 3.3522, + "step": 4181 + }, + { + "epoch": 0.33529093423663586, + "grad_norm": 1.9629385471343994, + "learning_rate": 1.902284604277794e-05, + "loss": 3.25, + "step": 4182 + }, + { + "epoch": 0.3353711090176585, + "grad_norm": 2.7201569080352783, + "learning_rate": 1.9022286153059956e-05, + "loss": 3.2676, + "step": 4183 + }, + { + "epoch": 0.3354512837986811, + "grad_norm": 2.0022342205047607, + "learning_rate": 1.9021726111229404e-05, + "loss": 3.2457, + "step": 4184 + }, + { + "epoch": 0.33553145857970373, + "grad_norm": 0.9247967600822449, + "learning_rate": 1.902116591729573e-05, + "loss": 2.5338, + "step": 4185 + }, + { + "epoch": 0.33561163336072636, + "grad_norm": 2.044076442718506, + "learning_rate": 1.902060557126838e-05, + "loss": 3.2123, + "step": 4186 + }, + { + "epoch": 0.33569180814174904, + "grad_norm": 1.7737160921096802, + "learning_rate": 1.9020045073156796e-05, + "loss": 3.2462, + "step": 4187 + }, + { + "epoch": 0.33577198292277166, + "grad_norm": 2.0880215167999268, + "learning_rate": 1.9019484422970435e-05, + "loss": 3.2787, + "step": 4188 + }, + { + "epoch": 0.3358521577037943, + "grad_norm": 2.072136878967285, + "learning_rate": 1.9018923620718742e-05, + "loss": 3.1863, + "step": 4189 + }, + { + "epoch": 0.3359323324848169, + "grad_norm": 2.0342354774475098, + "learning_rate": 1.9018362666411178e-05, + "loss": 3.4438, + "step": 4190 + }, + { + "epoch": 0.33601250726583953, + "grad_norm": 1.9743096828460693, + "learning_rate": 1.9017801560057198e-05, + "loss": 3.2715, + "step": 4191 + }, + { + "epoch": 0.33609268204686216, + "grad_norm": 1.8430280685424805, + "learning_rate": 1.9017240301666255e-05, + "loss": 3.2295, + "step": 4192 + }, + { + "epoch": 0.3361728568278848, + "grad_norm": 1.8562759160995483, + "learning_rate": 1.9016678891247827e-05, + "loss": 3.3674, + "step": 4193 + }, + { + "epoch": 0.3362530316089074, + "grad_norm": 1.8026390075683594, + "learning_rate": 1.901611732881137e-05, + "loss": 3.2284, + "step": 4194 + }, + { + "epoch": 0.33633320638993003, + "grad_norm": 1.8229455947875977, + "learning_rate": 1.9015555614366344e-05, + "loss": 3.1696, + "step": 4195 + }, + { + "epoch": 0.33641338117095265, + "grad_norm": 1.7516486644744873, + "learning_rate": 1.9014993747922236e-05, + "loss": 3.2797, + "step": 4196 + }, + { + "epoch": 0.33649355595197533, + "grad_norm": 2.1087491512298584, + "learning_rate": 1.9014431729488508e-05, + "loss": 3.3341, + "step": 4197 + }, + { + "epoch": 0.33657373073299796, + "grad_norm": 0.8637174963951111, + "learning_rate": 1.9013869559074637e-05, + "loss": 2.5203, + "step": 4198 + }, + { + "epoch": 0.3366539055140206, + "grad_norm": 1.955409049987793, + "learning_rate": 1.9013307236690105e-05, + "loss": 3.1774, + "step": 4199 + }, + { + "epoch": 0.3367340802950432, + "grad_norm": 0.8470594882965088, + "learning_rate": 1.9012744762344386e-05, + "loss": 2.5148, + "step": 4200 + }, + { + "epoch": 0.3368142550760658, + "grad_norm": 1.884417176246643, + "learning_rate": 1.9012182136046967e-05, + "loss": 3.2909, + "step": 4201 + }, + { + "epoch": 0.33689442985708845, + "grad_norm": 1.9228426218032837, + "learning_rate": 1.9011619357807335e-05, + "loss": 3.2435, + "step": 4202 + }, + { + "epoch": 0.3369746046381111, + "grad_norm": 1.9708958864212036, + "learning_rate": 1.901105642763498e-05, + "loss": 3.208, + "step": 4203 + }, + { + "epoch": 0.3370547794191337, + "grad_norm": 1.8915438652038574, + "learning_rate": 1.9010493345539384e-05, + "loss": 3.3612, + "step": 4204 + }, + { + "epoch": 0.3371349542001563, + "grad_norm": 1.927245855331421, + "learning_rate": 1.900993011153005e-05, + "loss": 3.1525, + "step": 4205 + }, + { + "epoch": 0.33721512898117895, + "grad_norm": 2.120292901992798, + "learning_rate": 1.9009366725616465e-05, + "loss": 3.0692, + "step": 4206 + }, + { + "epoch": 0.3372953037622016, + "grad_norm": 1.926589012145996, + "learning_rate": 1.9008803187808134e-05, + "loss": 3.327, + "step": 4207 + }, + { + "epoch": 0.33737547854322425, + "grad_norm": 1.937375783920288, + "learning_rate": 1.900823949811456e-05, + "loss": 3.291, + "step": 4208 + }, + { + "epoch": 0.3374556533242469, + "grad_norm": 1.808634877204895, + "learning_rate": 1.900767565654524e-05, + "loss": 3.1835, + "step": 4209 + }, + { + "epoch": 0.3375358281052695, + "grad_norm": 1.8934895992279053, + "learning_rate": 1.900711166310968e-05, + "loss": 3.231, + "step": 4210 + }, + { + "epoch": 0.3376160028862921, + "grad_norm": 2.0277836322784424, + "learning_rate": 1.9006547517817394e-05, + "loss": 3.192, + "step": 4211 + }, + { + "epoch": 0.33769617766731475, + "grad_norm": 1.918156623840332, + "learning_rate": 1.9005983220677894e-05, + "loss": 3.1511, + "step": 4212 + }, + { + "epoch": 0.33777635244833737, + "grad_norm": 2.0442562103271484, + "learning_rate": 1.9005418771700688e-05, + "loss": 3.3007, + "step": 4213 + }, + { + "epoch": 0.33785652722936, + "grad_norm": 1.9392849206924438, + "learning_rate": 1.9004854170895296e-05, + "loss": 3.3411, + "step": 4214 + }, + { + "epoch": 0.3379367020103826, + "grad_norm": 1.9863499402999878, + "learning_rate": 1.9004289418271236e-05, + "loss": 3.1803, + "step": 4215 + }, + { + "epoch": 0.33801687679140524, + "grad_norm": 1.8461947441101074, + "learning_rate": 1.9003724513838027e-05, + "loss": 3.1435, + "step": 4216 + }, + { + "epoch": 0.3380970515724279, + "grad_norm": 1.8856730461120605, + "learning_rate": 1.9003159457605202e-05, + "loss": 3.2642, + "step": 4217 + }, + { + "epoch": 0.33817722635345054, + "grad_norm": 1.7784897089004517, + "learning_rate": 1.9002594249582276e-05, + "loss": 3.3135, + "step": 4218 + }, + { + "epoch": 0.33825740113447317, + "grad_norm": 1.8047125339508057, + "learning_rate": 1.900202888977879e-05, + "loss": 3.2936, + "step": 4219 + }, + { + "epoch": 0.3383375759154958, + "grad_norm": 1.962091088294983, + "learning_rate": 1.9001463378204263e-05, + "loss": 3.3555, + "step": 4220 + }, + { + "epoch": 0.3384177506965184, + "grad_norm": 1.8916867971420288, + "learning_rate": 1.900089771486824e-05, + "loss": 3.3186, + "step": 4221 + }, + { + "epoch": 0.33849792547754104, + "grad_norm": 1.904486060142517, + "learning_rate": 1.900033189978025e-05, + "loss": 3.2787, + "step": 4222 + }, + { + "epoch": 0.33857810025856366, + "grad_norm": 2.0093934535980225, + "learning_rate": 1.8999765932949842e-05, + "loss": 3.1442, + "step": 4223 + }, + { + "epoch": 0.3386582750395863, + "grad_norm": 1.8013838529586792, + "learning_rate": 1.8999199814386546e-05, + "loss": 3.1442, + "step": 4224 + }, + { + "epoch": 0.3387384498206089, + "grad_norm": 1.926184058189392, + "learning_rate": 1.8998633544099916e-05, + "loss": 3.3112, + "step": 4225 + }, + { + "epoch": 0.33881862460163154, + "grad_norm": 1.9293932914733887, + "learning_rate": 1.8998067122099498e-05, + "loss": 3.1427, + "step": 4226 + }, + { + "epoch": 0.33889879938265416, + "grad_norm": 1.8953192234039307, + "learning_rate": 1.8997500548394835e-05, + "loss": 3.2905, + "step": 4227 + }, + { + "epoch": 0.33897897416367684, + "grad_norm": 0.8983559608459473, + "learning_rate": 1.899693382299549e-05, + "loss": 2.531, + "step": 4228 + }, + { + "epoch": 0.33905914894469946, + "grad_norm": 1.9389266967773438, + "learning_rate": 1.8996366945911003e-05, + "loss": 3.2528, + "step": 4229 + }, + { + "epoch": 0.3391393237257221, + "grad_norm": 1.899242639541626, + "learning_rate": 1.8995799917150947e-05, + "loss": 3.1106, + "step": 4230 + }, + { + "epoch": 0.3392194985067447, + "grad_norm": 2.0231540203094482, + "learning_rate": 1.8995232736724868e-05, + "loss": 3.2722, + "step": 4231 + }, + { + "epoch": 0.33929967328776733, + "grad_norm": 1.9239879846572876, + "learning_rate": 1.8994665404642344e-05, + "loss": 3.308, + "step": 4232 + }, + { + "epoch": 0.33937984806878996, + "grad_norm": 0.8795631527900696, + "learning_rate": 1.8994097920912926e-05, + "loss": 2.5197, + "step": 4233 + }, + { + "epoch": 0.3394600228498126, + "grad_norm": 1.8089241981506348, + "learning_rate": 1.8993530285546183e-05, + "loss": 3.2589, + "step": 4234 + }, + { + "epoch": 0.3395401976308352, + "grad_norm": 2.129528760910034, + "learning_rate": 1.8992962498551696e-05, + "loss": 3.2329, + "step": 4235 + }, + { + "epoch": 0.33962037241185783, + "grad_norm": 1.9618561267852783, + "learning_rate": 1.899239455993903e-05, + "loss": 3.3986, + "step": 4236 + }, + { + "epoch": 0.33970054719288045, + "grad_norm": 1.7989000082015991, + "learning_rate": 1.8991826469717758e-05, + "loss": 3.2101, + "step": 4237 + }, + { + "epoch": 0.33978072197390313, + "grad_norm": 1.995578646659851, + "learning_rate": 1.899125822789746e-05, + "loss": 3.2444, + "step": 4238 + }, + { + "epoch": 0.33986089675492576, + "grad_norm": 1.84404718875885, + "learning_rate": 1.8990689834487723e-05, + "loss": 3.1606, + "step": 4239 + }, + { + "epoch": 0.3399410715359484, + "grad_norm": 1.9692928791046143, + "learning_rate": 1.8990121289498118e-05, + "loss": 3.3298, + "step": 4240 + }, + { + "epoch": 0.340021246316971, + "grad_norm": 1.9288480281829834, + "learning_rate": 1.898955259293824e-05, + "loss": 3.2149, + "step": 4241 + }, + { + "epoch": 0.34010142109799363, + "grad_norm": 1.9650392532348633, + "learning_rate": 1.898898374481767e-05, + "loss": 3.1843, + "step": 4242 + }, + { + "epoch": 0.34018159587901625, + "grad_norm": 0.8235697150230408, + "learning_rate": 1.8988414745146008e-05, + "loss": 2.5133, + "step": 4243 + }, + { + "epoch": 0.3402617706600389, + "grad_norm": 1.9526174068450928, + "learning_rate": 1.8987845593932836e-05, + "loss": 3.2966, + "step": 4244 + }, + { + "epoch": 0.3403419454410615, + "grad_norm": 1.9185587167739868, + "learning_rate": 1.8987276291187757e-05, + "loss": 3.2701, + "step": 4245 + }, + { + "epoch": 0.3404221202220841, + "grad_norm": 0.8121548295021057, + "learning_rate": 1.8986706836920365e-05, + "loss": 2.507, + "step": 4246 + }, + { + "epoch": 0.34050229500310675, + "grad_norm": 2.0554442405700684, + "learning_rate": 1.8986137231140264e-05, + "loss": 3.2167, + "step": 4247 + }, + { + "epoch": 0.3405824697841294, + "grad_norm": 1.990902304649353, + "learning_rate": 1.8985567473857058e-05, + "loss": 3.3082, + "step": 4248 + }, + { + "epoch": 0.34066264456515205, + "grad_norm": 1.8102474212646484, + "learning_rate": 1.898499756508035e-05, + "loss": 3.1964, + "step": 4249 + }, + { + "epoch": 0.3407428193461747, + "grad_norm": 1.9281057119369507, + "learning_rate": 1.8984427504819745e-05, + "loss": 3.1908, + "step": 4250 + }, + { + "epoch": 0.3408229941271973, + "grad_norm": 1.84388267993927, + "learning_rate": 1.8983857293084867e-05, + "loss": 3.368, + "step": 4251 + }, + { + "epoch": 0.3409031689082199, + "grad_norm": 1.9172298908233643, + "learning_rate": 1.898328692988532e-05, + "loss": 3.3254, + "step": 4252 + }, + { + "epoch": 0.34098334368924255, + "grad_norm": 2.0152437686920166, + "learning_rate": 1.8982716415230717e-05, + "loss": 3.3368, + "step": 4253 + }, + { + "epoch": 0.34106351847026517, + "grad_norm": 1.9106706380844116, + "learning_rate": 1.8982145749130683e-05, + "loss": 3.1729, + "step": 4254 + }, + { + "epoch": 0.3411436932512878, + "grad_norm": 1.888407588005066, + "learning_rate": 1.8981574931594835e-05, + "loss": 3.334, + "step": 4255 + }, + { + "epoch": 0.3412238680323104, + "grad_norm": 1.9067463874816895, + "learning_rate": 1.8981003962632804e-05, + "loss": 3.293, + "step": 4256 + }, + { + "epoch": 0.34130404281333304, + "grad_norm": 2.0281620025634766, + "learning_rate": 1.8980432842254207e-05, + "loss": 3.3243, + "step": 4257 + }, + { + "epoch": 0.3413842175943557, + "grad_norm": 1.8968764543533325, + "learning_rate": 1.897986157046868e-05, + "loss": 3.2769, + "step": 4258 + }, + { + "epoch": 0.34146439237537834, + "grad_norm": 2.065948486328125, + "learning_rate": 1.8979290147285852e-05, + "loss": 3.1793, + "step": 4259 + }, + { + "epoch": 0.34154456715640097, + "grad_norm": 0.8822389245033264, + "learning_rate": 1.8978718572715355e-05, + "loss": 2.526, + "step": 4260 + }, + { + "epoch": 0.3416247419374236, + "grad_norm": 0.8806750774383545, + "learning_rate": 1.8978146846766826e-05, + "loss": 2.5333, + "step": 4261 + }, + { + "epoch": 0.3417049167184462, + "grad_norm": 2.065319776535034, + "learning_rate": 1.8977574969449908e-05, + "loss": 3.2699, + "step": 4262 + }, + { + "epoch": 0.34178509149946884, + "grad_norm": 0.8204147219657898, + "learning_rate": 1.897700294077424e-05, + "loss": 2.5326, + "step": 4263 + }, + { + "epoch": 0.34186526628049146, + "grad_norm": 2.166536569595337, + "learning_rate": 1.8976430760749464e-05, + "loss": 3.2626, + "step": 4264 + }, + { + "epoch": 0.3419454410615141, + "grad_norm": 2.026710271835327, + "learning_rate": 1.897585842938523e-05, + "loss": 3.2656, + "step": 4265 + }, + { + "epoch": 0.3420256158425367, + "grad_norm": 1.9344555139541626, + "learning_rate": 1.8975285946691188e-05, + "loss": 3.3304, + "step": 4266 + }, + { + "epoch": 0.34210579062355934, + "grad_norm": 2.002203941345215, + "learning_rate": 1.8974713312676987e-05, + "loss": 3.2575, + "step": 4267 + }, + { + "epoch": 0.342185965404582, + "grad_norm": 1.9005253314971924, + "learning_rate": 1.8974140527352282e-05, + "loss": 3.2502, + "step": 4268 + }, + { + "epoch": 0.34226614018560464, + "grad_norm": 2.0997931957244873, + "learning_rate": 1.897356759072673e-05, + "loss": 3.3417, + "step": 4269 + }, + { + "epoch": 0.34234631496662726, + "grad_norm": 1.9164271354675293, + "learning_rate": 1.8972994502809992e-05, + "loss": 3.2795, + "step": 4270 + }, + { + "epoch": 0.3424264897476499, + "grad_norm": 1.892552375793457, + "learning_rate": 1.897242126361173e-05, + "loss": 3.281, + "step": 4271 + }, + { + "epoch": 0.3425066645286725, + "grad_norm": 1.832094669342041, + "learning_rate": 1.8971847873141605e-05, + "loss": 3.2404, + "step": 4272 + }, + { + "epoch": 0.34258683930969513, + "grad_norm": 1.8231582641601562, + "learning_rate": 1.897127433140929e-05, + "loss": 3.2004, + "step": 4273 + }, + { + "epoch": 0.34266701409071776, + "grad_norm": 1.903483271598816, + "learning_rate": 1.897070063842445e-05, + "loss": 3.2554, + "step": 4274 + }, + { + "epoch": 0.3427471888717404, + "grad_norm": 1.9402614831924438, + "learning_rate": 1.897012679419676e-05, + "loss": 3.2182, + "step": 4275 + }, + { + "epoch": 0.342827363652763, + "grad_norm": 1.9947987794876099, + "learning_rate": 1.8969552798735892e-05, + "loss": 3.363, + "step": 4276 + }, + { + "epoch": 0.34290753843378563, + "grad_norm": 1.9219658374786377, + "learning_rate": 1.8968978652051523e-05, + "loss": 3.329, + "step": 4277 + }, + { + "epoch": 0.3429877132148083, + "grad_norm": 1.9934762716293335, + "learning_rate": 1.896840435415334e-05, + "loss": 3.3138, + "step": 4278 + }, + { + "epoch": 0.34306788799583093, + "grad_norm": 1.905991792678833, + "learning_rate": 1.8967829905051015e-05, + "loss": 3.2391, + "step": 4279 + }, + { + "epoch": 0.34314806277685356, + "grad_norm": 2.0004653930664062, + "learning_rate": 1.8967255304754242e-05, + "loss": 3.3362, + "step": 4280 + }, + { + "epoch": 0.3432282375578762, + "grad_norm": 1.984078288078308, + "learning_rate": 1.8966680553272707e-05, + "loss": 3.2502, + "step": 4281 + }, + { + "epoch": 0.3433084123388988, + "grad_norm": 1.9376482963562012, + "learning_rate": 1.896610565061609e-05, + "loss": 3.293, + "step": 4282 + }, + { + "epoch": 0.34338858711992143, + "grad_norm": 1.83151376247406, + "learning_rate": 1.89655305967941e-05, + "loss": 3.1813, + "step": 4283 + }, + { + "epoch": 0.34346876190094405, + "grad_norm": 1.968063235282898, + "learning_rate": 1.896495539181642e-05, + "loss": 3.2327, + "step": 4284 + }, + { + "epoch": 0.3435489366819667, + "grad_norm": 1.979332447052002, + "learning_rate": 1.896438003569275e-05, + "loss": 3.2691, + "step": 4285 + }, + { + "epoch": 0.3436291114629893, + "grad_norm": 1.9555323123931885, + "learning_rate": 1.8963804528432797e-05, + "loss": 3.2709, + "step": 4286 + }, + { + "epoch": 0.3437092862440119, + "grad_norm": 2.0402724742889404, + "learning_rate": 1.8963228870046255e-05, + "loss": 3.3119, + "step": 4287 + }, + { + "epoch": 0.34378946102503455, + "grad_norm": 1.9047225713729858, + "learning_rate": 1.8962653060542834e-05, + "loss": 3.3378, + "step": 4288 + }, + { + "epoch": 0.34386963580605723, + "grad_norm": 0.8797641396522522, + "learning_rate": 1.8962077099932242e-05, + "loss": 2.5131, + "step": 4289 + }, + { + "epoch": 0.34394981058707985, + "grad_norm": 0.8403259515762329, + "learning_rate": 1.896150098822419e-05, + "loss": 2.5112, + "step": 4290 + }, + { + "epoch": 0.3440299853681025, + "grad_norm": 2.143780469894409, + "learning_rate": 1.8960924725428387e-05, + "loss": 3.3113, + "step": 4291 + }, + { + "epoch": 0.3441101601491251, + "grad_norm": 1.9282593727111816, + "learning_rate": 1.896034831155455e-05, + "loss": 3.2344, + "step": 4292 + }, + { + "epoch": 0.3441903349301477, + "grad_norm": 1.9433367252349854, + "learning_rate": 1.8959771746612402e-05, + "loss": 3.1781, + "step": 4293 + }, + { + "epoch": 0.34427050971117035, + "grad_norm": 1.9270352125167847, + "learning_rate": 1.8959195030611658e-05, + "loss": 3.2899, + "step": 4294 + }, + { + "epoch": 0.34435068449219297, + "grad_norm": 1.9063423871994019, + "learning_rate": 1.8958618163562043e-05, + "loss": 3.1694, + "step": 4295 + }, + { + "epoch": 0.3444308592732156, + "grad_norm": 1.8956776857376099, + "learning_rate": 1.8958041145473283e-05, + "loss": 3.1929, + "step": 4296 + }, + { + "epoch": 0.3445110340542382, + "grad_norm": 1.7914183139801025, + "learning_rate": 1.8957463976355107e-05, + "loss": 3.2031, + "step": 4297 + }, + { + "epoch": 0.34459120883526084, + "grad_norm": 1.9271382093429565, + "learning_rate": 1.8956886656217247e-05, + "loss": 3.2783, + "step": 4298 + }, + { + "epoch": 0.3446713836162835, + "grad_norm": 1.8815562725067139, + "learning_rate": 1.8956309185069433e-05, + "loss": 3.2411, + "step": 4299 + }, + { + "epoch": 0.34475155839730615, + "grad_norm": 2.0445802211761475, + "learning_rate": 1.8955731562921404e-05, + "loss": 3.2435, + "step": 4300 + }, + { + "epoch": 0.34483173317832877, + "grad_norm": 1.8933343887329102, + "learning_rate": 1.8955153789782894e-05, + "loss": 3.2573, + "step": 4301 + }, + { + "epoch": 0.3449119079593514, + "grad_norm": 1.963031530380249, + "learning_rate": 1.895457586566365e-05, + "loss": 3.2635, + "step": 4302 + }, + { + "epoch": 0.344992082740374, + "grad_norm": 1.8408050537109375, + "learning_rate": 1.8953997790573414e-05, + "loss": 3.1617, + "step": 4303 + }, + { + "epoch": 0.34507225752139664, + "grad_norm": 1.8682496547698975, + "learning_rate": 1.8953419564521926e-05, + "loss": 3.2451, + "step": 4304 + }, + { + "epoch": 0.34515243230241927, + "grad_norm": 2.0750842094421387, + "learning_rate": 1.8952841187518944e-05, + "loss": 3.2462, + "step": 4305 + }, + { + "epoch": 0.3452326070834419, + "grad_norm": 2.0472052097320557, + "learning_rate": 1.8952262659574213e-05, + "loss": 3.2841, + "step": 4306 + }, + { + "epoch": 0.3453127818644645, + "grad_norm": 1.8834266662597656, + "learning_rate": 1.8951683980697488e-05, + "loss": 3.1768, + "step": 4307 + }, + { + "epoch": 0.34539295664548714, + "grad_norm": 2.0470998287200928, + "learning_rate": 1.895110515089853e-05, + "loss": 3.2415, + "step": 4308 + }, + { + "epoch": 0.3454731314265098, + "grad_norm": 1.991530179977417, + "learning_rate": 1.895052617018709e-05, + "loss": 3.2655, + "step": 4309 + }, + { + "epoch": 0.34555330620753244, + "grad_norm": 1.8810392618179321, + "learning_rate": 1.894994703857293e-05, + "loss": 3.2405, + "step": 4310 + }, + { + "epoch": 0.34563348098855506, + "grad_norm": 1.8330262899398804, + "learning_rate": 1.8949367756065825e-05, + "loss": 3.1671, + "step": 4311 + }, + { + "epoch": 0.3457136557695777, + "grad_norm": 1.9815809726715088, + "learning_rate": 1.8948788322675532e-05, + "loss": 3.2527, + "step": 4312 + }, + { + "epoch": 0.3457938305506003, + "grad_norm": 2.099463701248169, + "learning_rate": 1.894820873841182e-05, + "loss": 3.1869, + "step": 4313 + }, + { + "epoch": 0.34587400533162294, + "grad_norm": 1.9247206449508667, + "learning_rate": 1.894762900328446e-05, + "loss": 3.1932, + "step": 4314 + }, + { + "epoch": 0.34595418011264556, + "grad_norm": 1.8880367279052734, + "learning_rate": 1.894704911730323e-05, + "loss": 3.2513, + "step": 4315 + }, + { + "epoch": 0.3460343548936682, + "grad_norm": 2.1419670581817627, + "learning_rate": 1.8946469080477906e-05, + "loss": 3.2445, + "step": 4316 + }, + { + "epoch": 0.3461145296746908, + "grad_norm": 1.8477355241775513, + "learning_rate": 1.8945888892818268e-05, + "loss": 3.2856, + "step": 4317 + }, + { + "epoch": 0.34619470445571343, + "grad_norm": 1.7486398220062256, + "learning_rate": 1.8945308554334094e-05, + "loss": 3.188, + "step": 4318 + }, + { + "epoch": 0.3462748792367361, + "grad_norm": 1.9195737838745117, + "learning_rate": 1.894472806503517e-05, + "loss": 3.3113, + "step": 4319 + }, + { + "epoch": 0.34635505401775873, + "grad_norm": 1.9323934316635132, + "learning_rate": 1.8944147424931282e-05, + "loss": 3.2048, + "step": 4320 + }, + { + "epoch": 0.34643522879878136, + "grad_norm": 1.9637969732284546, + "learning_rate": 1.894356663403222e-05, + "loss": 3.2783, + "step": 4321 + }, + { + "epoch": 0.346515403579804, + "grad_norm": 1.9788999557495117, + "learning_rate": 1.894298569234778e-05, + "loss": 3.1924, + "step": 4322 + }, + { + "epoch": 0.3465955783608266, + "grad_norm": 1.898196816444397, + "learning_rate": 1.8942404599887755e-05, + "loss": 3.2552, + "step": 4323 + }, + { + "epoch": 0.34667575314184923, + "grad_norm": 1.923568606376648, + "learning_rate": 1.8941823356661933e-05, + "loss": 3.3163, + "step": 4324 + }, + { + "epoch": 0.34675592792287185, + "grad_norm": 2.107234477996826, + "learning_rate": 1.8941241962680127e-05, + "loss": 3.3525, + "step": 4325 + }, + { + "epoch": 0.3468361027038945, + "grad_norm": 1.9679394960403442, + "learning_rate": 1.894066041795213e-05, + "loss": 3.2264, + "step": 4326 + }, + { + "epoch": 0.3469162774849171, + "grad_norm": 2.0963099002838135, + "learning_rate": 1.894007872248775e-05, + "loss": 3.2388, + "step": 4327 + }, + { + "epoch": 0.3469964522659397, + "grad_norm": 2.127225399017334, + "learning_rate": 1.8939496876296793e-05, + "loss": 3.1785, + "step": 4328 + }, + { + "epoch": 0.3470766270469624, + "grad_norm": 1.9865678548812866, + "learning_rate": 1.893891487938907e-05, + "loss": 3.1709, + "step": 4329 + }, + { + "epoch": 0.34715680182798503, + "grad_norm": 1.942649006843567, + "learning_rate": 1.8938332731774395e-05, + "loss": 3.0929, + "step": 4330 + }, + { + "epoch": 0.34723697660900765, + "grad_norm": 1.9736446142196655, + "learning_rate": 1.8937750433462575e-05, + "loss": 3.294, + "step": 4331 + }, + { + "epoch": 0.3473171513900303, + "grad_norm": 1.88287353515625, + "learning_rate": 1.8937167984463435e-05, + "loss": 3.3047, + "step": 4332 + }, + { + "epoch": 0.3473973261710529, + "grad_norm": 2.0510284900665283, + "learning_rate": 1.8936585384786793e-05, + "loss": 3.3435, + "step": 4333 + }, + { + "epoch": 0.3474775009520755, + "grad_norm": 2.0161526203155518, + "learning_rate": 1.893600263444247e-05, + "loss": 3.1637, + "step": 4334 + }, + { + "epoch": 0.34755767573309815, + "grad_norm": 1.971595287322998, + "learning_rate": 1.8935419733440293e-05, + "loss": 3.2589, + "step": 4335 + }, + { + "epoch": 0.34763785051412077, + "grad_norm": 1.8145651817321777, + "learning_rate": 1.8934836681790086e-05, + "loss": 3.205, + "step": 4336 + }, + { + "epoch": 0.3477180252951434, + "grad_norm": 1.9425443410873413, + "learning_rate": 1.8934253479501686e-05, + "loss": 3.134, + "step": 4337 + }, + { + "epoch": 0.347798200076166, + "grad_norm": 1.8760933876037598, + "learning_rate": 1.8933670126584916e-05, + "loss": 3.1959, + "step": 4338 + }, + { + "epoch": 0.34787837485718864, + "grad_norm": 1.907256007194519, + "learning_rate": 1.8933086623049618e-05, + "loss": 3.2581, + "step": 4339 + }, + { + "epoch": 0.3479585496382113, + "grad_norm": 1.9014538526535034, + "learning_rate": 1.893250296890563e-05, + "loss": 3.1383, + "step": 4340 + }, + { + "epoch": 0.34803872441923395, + "grad_norm": 0.9260676503181458, + "learning_rate": 1.8931919164162787e-05, + "loss": 2.5297, + "step": 4341 + }, + { + "epoch": 0.34811889920025657, + "grad_norm": 1.9960654973983765, + "learning_rate": 1.893133520883094e-05, + "loss": 3.365, + "step": 4342 + }, + { + "epoch": 0.3481990739812792, + "grad_norm": 2.0395472049713135, + "learning_rate": 1.8930751102919925e-05, + "loss": 3.2584, + "step": 4343 + }, + { + "epoch": 0.3482792487623018, + "grad_norm": 2.0468814373016357, + "learning_rate": 1.893016684643959e-05, + "loss": 3.3116, + "step": 4344 + }, + { + "epoch": 0.34835942354332444, + "grad_norm": 1.9561187028884888, + "learning_rate": 1.8929582439399795e-05, + "loss": 3.2866, + "step": 4345 + }, + { + "epoch": 0.34843959832434707, + "grad_norm": 1.84893000125885, + "learning_rate": 1.892899788181039e-05, + "loss": 3.3461, + "step": 4346 + }, + { + "epoch": 0.3485197731053697, + "grad_norm": 1.8306394815444946, + "learning_rate": 1.892841317368122e-05, + "loss": 3.2307, + "step": 4347 + }, + { + "epoch": 0.3485999478863923, + "grad_norm": 2.1486685276031494, + "learning_rate": 1.8927828315022156e-05, + "loss": 3.2322, + "step": 4348 + }, + { + "epoch": 0.34868012266741494, + "grad_norm": 1.848071813583374, + "learning_rate": 1.892724330584305e-05, + "loss": 3.1958, + "step": 4349 + }, + { + "epoch": 0.3487602974484376, + "grad_norm": 2.0348315238952637, + "learning_rate": 1.8926658146153768e-05, + "loss": 3.228, + "step": 4350 + }, + { + "epoch": 0.34884047222946024, + "grad_norm": 1.9879790544509888, + "learning_rate": 1.892607283596418e-05, + "loss": 3.3062, + "step": 4351 + }, + { + "epoch": 0.34892064701048287, + "grad_norm": 1.844519853591919, + "learning_rate": 1.8925487375284146e-05, + "loss": 3.2039, + "step": 4352 + }, + { + "epoch": 0.3490008217915055, + "grad_norm": 1.8460427522659302, + "learning_rate": 1.892490176412354e-05, + "loss": 3.3231, + "step": 4353 + }, + { + "epoch": 0.3490809965725281, + "grad_norm": 1.8080483675003052, + "learning_rate": 1.8924316002492237e-05, + "loss": 3.3312, + "step": 4354 + }, + { + "epoch": 0.34916117135355074, + "grad_norm": 2.0224106311798096, + "learning_rate": 1.892373009040011e-05, + "loss": 3.2363, + "step": 4355 + }, + { + "epoch": 0.34924134613457336, + "grad_norm": 1.8234974145889282, + "learning_rate": 1.8923144027857042e-05, + "loss": 3.2766, + "step": 4356 + }, + { + "epoch": 0.349321520915596, + "grad_norm": 2.2675838470458984, + "learning_rate": 1.892255781487291e-05, + "loss": 3.2628, + "step": 4357 + }, + { + "epoch": 0.3494016956966186, + "grad_norm": 1.9589381217956543, + "learning_rate": 1.8921971451457594e-05, + "loss": 3.2208, + "step": 4358 + }, + { + "epoch": 0.34948187047764123, + "grad_norm": 2.1214921474456787, + "learning_rate": 1.8921384937620984e-05, + "loss": 3.2095, + "step": 4359 + }, + { + "epoch": 0.3495620452586639, + "grad_norm": 1.8824008703231812, + "learning_rate": 1.8920798273372973e-05, + "loss": 3.2499, + "step": 4360 + }, + { + "epoch": 0.34964222003968654, + "grad_norm": 2.155668020248413, + "learning_rate": 1.8920211458723444e-05, + "loss": 3.2592, + "step": 4361 + }, + { + "epoch": 0.34972239482070916, + "grad_norm": 2.0992255210876465, + "learning_rate": 1.8919624493682294e-05, + "loss": 3.2491, + "step": 4362 + }, + { + "epoch": 0.3498025696017318, + "grad_norm": 2.093177556991577, + "learning_rate": 1.891903737825942e-05, + "loss": 3.3706, + "step": 4363 + }, + { + "epoch": 0.3498827443827544, + "grad_norm": 2.036919355392456, + "learning_rate": 1.891845011246472e-05, + "loss": 3.2543, + "step": 4364 + }, + { + "epoch": 0.34996291916377703, + "grad_norm": 2.109431266784668, + "learning_rate": 1.891786269630809e-05, + "loss": 3.2598, + "step": 4365 + }, + { + "epoch": 0.35004309394479965, + "grad_norm": 2.0915660858154297, + "learning_rate": 1.8917275129799438e-05, + "loss": 3.236, + "step": 4366 + }, + { + "epoch": 0.3501232687258223, + "grad_norm": 1.8963706493377686, + "learning_rate": 1.8916687412948673e-05, + "loss": 3.3027, + "step": 4367 + }, + { + "epoch": 0.3502034435068449, + "grad_norm": 1.96126127243042, + "learning_rate": 1.89160995457657e-05, + "loss": 3.3109, + "step": 4368 + }, + { + "epoch": 0.3502836182878675, + "grad_norm": 1.8757075071334839, + "learning_rate": 1.8915511528260435e-05, + "loss": 3.3288, + "step": 4369 + }, + { + "epoch": 0.3503637930688902, + "grad_norm": 1.8929671049118042, + "learning_rate": 1.891492336044278e-05, + "loss": 3.2441, + "step": 4370 + }, + { + "epoch": 0.35044396784991283, + "grad_norm": 1.9564893245697021, + "learning_rate": 1.8914335042322664e-05, + "loss": 3.149, + "step": 4371 + }, + { + "epoch": 0.35052414263093545, + "grad_norm": 1.9257200956344604, + "learning_rate": 1.891374657391e-05, + "loss": 3.1951, + "step": 4372 + }, + { + "epoch": 0.3506043174119581, + "grad_norm": 2.5176286697387695, + "learning_rate": 1.8913157955214707e-05, + "loss": 3.1998, + "step": 4373 + }, + { + "epoch": 0.3506844921929807, + "grad_norm": 1.8871287107467651, + "learning_rate": 1.8912569186246717e-05, + "loss": 3.3025, + "step": 4374 + }, + { + "epoch": 0.3507646669740033, + "grad_norm": 1.8940234184265137, + "learning_rate": 1.891198026701595e-05, + "loss": 3.3027, + "step": 4375 + }, + { + "epoch": 0.35084484175502595, + "grad_norm": 2.0284781455993652, + "learning_rate": 1.8911391197532337e-05, + "loss": 3.1811, + "step": 4376 + }, + { + "epoch": 0.3509250165360486, + "grad_norm": 1.9837427139282227, + "learning_rate": 1.8910801977805806e-05, + "loss": 3.3077, + "step": 4377 + }, + { + "epoch": 0.3510051913170712, + "grad_norm": 1.8949164152145386, + "learning_rate": 1.8910212607846296e-05, + "loss": 3.1043, + "step": 4378 + }, + { + "epoch": 0.3510853660980938, + "grad_norm": 1.9635217189788818, + "learning_rate": 1.8909623087663745e-05, + "loss": 3.2596, + "step": 4379 + }, + { + "epoch": 0.3511655408791165, + "grad_norm": 1.9534724950790405, + "learning_rate": 1.8909033417268083e-05, + "loss": 3.2474, + "step": 4380 + }, + { + "epoch": 0.3512457156601391, + "grad_norm": 2.1701579093933105, + "learning_rate": 1.890844359666926e-05, + "loss": 3.2458, + "step": 4381 + }, + { + "epoch": 0.35132589044116175, + "grad_norm": 2.0336389541625977, + "learning_rate": 1.890785362587722e-05, + "loss": 3.2378, + "step": 4382 + }, + { + "epoch": 0.35140606522218437, + "grad_norm": 2.0396437644958496, + "learning_rate": 1.8907263504901903e-05, + "loss": 3.3452, + "step": 4383 + }, + { + "epoch": 0.351486240003207, + "grad_norm": 1.8932456970214844, + "learning_rate": 1.8906673233753266e-05, + "loss": 3.2725, + "step": 4384 + }, + { + "epoch": 0.3515664147842296, + "grad_norm": 1.932277798652649, + "learning_rate": 1.890608281244126e-05, + "loss": 3.2322, + "step": 4385 + }, + { + "epoch": 0.35164658956525224, + "grad_norm": 1.8352569341659546, + "learning_rate": 1.890549224097583e-05, + "loss": 3.2808, + "step": 4386 + }, + { + "epoch": 0.35172676434627487, + "grad_norm": 1.964613437652588, + "learning_rate": 1.890490151936694e-05, + "loss": 3.4813, + "step": 4387 + }, + { + "epoch": 0.3518069391272975, + "grad_norm": 1.9369502067565918, + "learning_rate": 1.890431064762455e-05, + "loss": 3.2999, + "step": 4388 + }, + { + "epoch": 0.3518871139083201, + "grad_norm": 2.113607883453369, + "learning_rate": 1.8903719625758622e-05, + "loss": 3.249, + "step": 4389 + }, + { + "epoch": 0.3519672886893428, + "grad_norm": 1.7805556058883667, + "learning_rate": 1.890312845377912e-05, + "loss": 3.1556, + "step": 4390 + }, + { + "epoch": 0.3520474634703654, + "grad_norm": 0.8468862771987915, + "learning_rate": 1.8902537131696007e-05, + "loss": 2.5128, + "step": 4391 + }, + { + "epoch": 0.35212763825138804, + "grad_norm": 1.9114363193511963, + "learning_rate": 1.890194565951926e-05, + "loss": 3.2316, + "step": 4392 + }, + { + "epoch": 0.35220781303241067, + "grad_norm": 1.864394187927246, + "learning_rate": 1.8901354037258844e-05, + "loss": 3.2748, + "step": 4393 + }, + { + "epoch": 0.3522879878134333, + "grad_norm": 1.9843083620071411, + "learning_rate": 1.8900762264924736e-05, + "loss": 3.2738, + "step": 4394 + }, + { + "epoch": 0.3523681625944559, + "grad_norm": 1.8663018941879272, + "learning_rate": 1.890017034252691e-05, + "loss": 3.1058, + "step": 4395 + }, + { + "epoch": 0.35244833737547854, + "grad_norm": 1.9558223485946655, + "learning_rate": 1.8899578270075355e-05, + "loss": 3.1505, + "step": 4396 + }, + { + "epoch": 0.35252851215650116, + "grad_norm": 1.7717362642288208, + "learning_rate": 1.889898604758004e-05, + "loss": 3.1934, + "step": 4397 + }, + { + "epoch": 0.3526086869375238, + "grad_norm": 0.8789693713188171, + "learning_rate": 1.889839367505096e-05, + "loss": 2.5397, + "step": 4398 + }, + { + "epoch": 0.3526888617185464, + "grad_norm": 1.9980815649032593, + "learning_rate": 1.8897801152498097e-05, + "loss": 3.2451, + "step": 4399 + }, + { + "epoch": 0.35276903649956903, + "grad_norm": 1.9105437994003296, + "learning_rate": 1.8897208479931447e-05, + "loss": 3.3112, + "step": 4400 + }, + { + "epoch": 0.3528492112805917, + "grad_norm": 1.8671656847000122, + "learning_rate": 1.8896615657361e-05, + "loss": 3.2515, + "step": 4401 + }, + { + "epoch": 0.35292938606161434, + "grad_norm": 1.9155999422073364, + "learning_rate": 1.889602268479674e-05, + "loss": 3.3121, + "step": 4402 + }, + { + "epoch": 0.35300956084263696, + "grad_norm": 1.8654762506484985, + "learning_rate": 1.8895429562248674e-05, + "loss": 3.2198, + "step": 4403 + }, + { + "epoch": 0.3530897356236596, + "grad_norm": 0.8533163666725159, + "learning_rate": 1.8894836289726806e-05, + "loss": 2.5189, + "step": 4404 + }, + { + "epoch": 0.3531699104046822, + "grad_norm": 2.010502576828003, + "learning_rate": 1.889424286724113e-05, + "loss": 3.1615, + "step": 4405 + }, + { + "epoch": 0.35325008518570483, + "grad_norm": 1.9467787742614746, + "learning_rate": 1.8893649294801654e-05, + "loss": 3.3851, + "step": 4406 + }, + { + "epoch": 0.35333025996672746, + "grad_norm": 0.8304028511047363, + "learning_rate": 1.8893055572418385e-05, + "loss": 2.5195, + "step": 4407 + }, + { + "epoch": 0.3534104347477501, + "grad_norm": 1.9511168003082275, + "learning_rate": 1.8892461700101333e-05, + "loss": 3.2911, + "step": 4408 + }, + { + "epoch": 0.3534906095287727, + "grad_norm": 1.9332983493804932, + "learning_rate": 1.889186767786051e-05, + "loss": 3.2496, + "step": 4409 + }, + { + "epoch": 0.3535707843097953, + "grad_norm": 1.9370172023773193, + "learning_rate": 1.8891273505705933e-05, + "loss": 3.287, + "step": 4410 + }, + { + "epoch": 0.353650959090818, + "grad_norm": 1.8981720209121704, + "learning_rate": 1.8890679183647617e-05, + "loss": 3.1727, + "step": 4411 + }, + { + "epoch": 0.35373113387184063, + "grad_norm": 2.090338706970215, + "learning_rate": 1.8890084711695585e-05, + "loss": 3.3344, + "step": 4412 + }, + { + "epoch": 0.35381130865286325, + "grad_norm": 1.8351390361785889, + "learning_rate": 1.8889490089859857e-05, + "loss": 3.2948, + "step": 4413 + }, + { + "epoch": 0.3538914834338859, + "grad_norm": 1.905441403388977, + "learning_rate": 1.888889531815046e-05, + "loss": 3.2327, + "step": 4414 + }, + { + "epoch": 0.3539716582149085, + "grad_norm": 1.9283897876739502, + "learning_rate": 1.888830039657742e-05, + "loss": 3.245, + "step": 4415 + }, + { + "epoch": 0.3540518329959311, + "grad_norm": 1.911538004875183, + "learning_rate": 1.8887705325150767e-05, + "loss": 3.249, + "step": 4416 + }, + { + "epoch": 0.35413200777695375, + "grad_norm": 1.83982515335083, + "learning_rate": 1.8887110103880536e-05, + "loss": 3.2313, + "step": 4417 + }, + { + "epoch": 0.3542121825579764, + "grad_norm": 1.9856336116790771, + "learning_rate": 1.8886514732776762e-05, + "loss": 3.2952, + "step": 4418 + }, + { + "epoch": 0.354292357338999, + "grad_norm": 1.8518283367156982, + "learning_rate": 1.888591921184948e-05, + "loss": 3.1746, + "step": 4419 + }, + { + "epoch": 0.3543725321200216, + "grad_norm": 0.9751653671264648, + "learning_rate": 1.888532354110873e-05, + "loss": 2.5212, + "step": 4420 + }, + { + "epoch": 0.3544527069010443, + "grad_norm": 2.061859607696533, + "learning_rate": 1.8884727720564558e-05, + "loss": 3.296, + "step": 4421 + }, + { + "epoch": 0.3545328816820669, + "grad_norm": 1.9182848930358887, + "learning_rate": 1.8884131750227008e-05, + "loss": 3.2222, + "step": 4422 + }, + { + "epoch": 0.35461305646308955, + "grad_norm": 1.9388083219528198, + "learning_rate": 1.8883535630106128e-05, + "loss": 3.269, + "step": 4423 + }, + { + "epoch": 0.3546932312441122, + "grad_norm": 2.0681700706481934, + "learning_rate": 1.888293936021197e-05, + "loss": 3.3673, + "step": 4424 + }, + { + "epoch": 0.3547734060251348, + "grad_norm": 1.8449684381484985, + "learning_rate": 1.8882342940554582e-05, + "loss": 3.1155, + "step": 4425 + }, + { + "epoch": 0.3548535808061574, + "grad_norm": 2.0087087154388428, + "learning_rate": 1.8881746371144025e-05, + "loss": 3.2421, + "step": 4426 + }, + { + "epoch": 0.35493375558718004, + "grad_norm": 2.0452568531036377, + "learning_rate": 1.8881149651990352e-05, + "loss": 3.2216, + "step": 4427 + }, + { + "epoch": 0.35501393036820267, + "grad_norm": 1.9597810506820679, + "learning_rate": 1.8880552783103626e-05, + "loss": 3.2213, + "step": 4428 + }, + { + "epoch": 0.3550941051492253, + "grad_norm": 1.8759243488311768, + "learning_rate": 1.8879955764493912e-05, + "loss": 3.1545, + "step": 4429 + }, + { + "epoch": 0.3551742799302479, + "grad_norm": 2.132718801498413, + "learning_rate": 1.887935859617127e-05, + "loss": 3.1893, + "step": 4430 + }, + { + "epoch": 0.3552544547112706, + "grad_norm": 1.8979688882827759, + "learning_rate": 1.8878761278145775e-05, + "loss": 3.2191, + "step": 4431 + }, + { + "epoch": 0.3553346294922932, + "grad_norm": 2.0894558429718018, + "learning_rate": 1.8878163810427494e-05, + "loss": 3.3077, + "step": 4432 + }, + { + "epoch": 0.35541480427331584, + "grad_norm": 1.9846090078353882, + "learning_rate": 1.8877566193026497e-05, + "loss": 3.278, + "step": 4433 + }, + { + "epoch": 0.35549497905433847, + "grad_norm": 2.0027472972869873, + "learning_rate": 1.8876968425952867e-05, + "loss": 3.2467, + "step": 4434 + }, + { + "epoch": 0.3555751538353611, + "grad_norm": 1.8660392761230469, + "learning_rate": 1.8876370509216674e-05, + "loss": 3.219, + "step": 4435 + }, + { + "epoch": 0.3556553286163837, + "grad_norm": 1.9145197868347168, + "learning_rate": 1.8875772442828e-05, + "loss": 3.3206, + "step": 4436 + }, + { + "epoch": 0.35573550339740634, + "grad_norm": 2.006042718887329, + "learning_rate": 1.8875174226796938e-05, + "loss": 3.3015, + "step": 4437 + }, + { + "epoch": 0.35581567817842896, + "grad_norm": 1.872869610786438, + "learning_rate": 1.8874575861133564e-05, + "loss": 3.2037, + "step": 4438 + }, + { + "epoch": 0.3558958529594516, + "grad_norm": 1.8494853973388672, + "learning_rate": 1.8873977345847967e-05, + "loss": 3.2984, + "step": 4439 + }, + { + "epoch": 0.3559760277404742, + "grad_norm": 2.005617141723633, + "learning_rate": 1.887337868095024e-05, + "loss": 3.2391, + "step": 4440 + }, + { + "epoch": 0.3560562025214969, + "grad_norm": 2.0333571434020996, + "learning_rate": 1.8872779866450475e-05, + "loss": 3.2979, + "step": 4441 + }, + { + "epoch": 0.3561363773025195, + "grad_norm": 1.947048306465149, + "learning_rate": 1.887218090235877e-05, + "loss": 3.3688, + "step": 4442 + }, + { + "epoch": 0.35621655208354214, + "grad_norm": 1.9437339305877686, + "learning_rate": 1.8871581788685224e-05, + "loss": 3.2253, + "step": 4443 + }, + { + "epoch": 0.35629672686456476, + "grad_norm": 2.0400452613830566, + "learning_rate": 1.887098252543993e-05, + "loss": 3.2339, + "step": 4444 + }, + { + "epoch": 0.3563769016455874, + "grad_norm": 1.9693090915679932, + "learning_rate": 1.8870383112633e-05, + "loss": 3.3895, + "step": 4445 + }, + { + "epoch": 0.35645707642661, + "grad_norm": 1.8470686674118042, + "learning_rate": 1.8869783550274538e-05, + "loss": 3.2513, + "step": 4446 + }, + { + "epoch": 0.35653725120763263, + "grad_norm": 0.9066380262374878, + "learning_rate": 1.8869183838374653e-05, + "loss": 2.5278, + "step": 4447 + }, + { + "epoch": 0.35661742598865526, + "grad_norm": 1.7940289974212646, + "learning_rate": 1.886858397694345e-05, + "loss": 3.2174, + "step": 4448 + }, + { + "epoch": 0.3566976007696779, + "grad_norm": 1.9422909021377563, + "learning_rate": 1.8867983965991047e-05, + "loss": 3.2976, + "step": 4449 + }, + { + "epoch": 0.3567777755507005, + "grad_norm": 1.8817845582962036, + "learning_rate": 1.8867383805527562e-05, + "loss": 3.1956, + "step": 4450 + }, + { + "epoch": 0.3568579503317232, + "grad_norm": 1.9931834936141968, + "learning_rate": 1.886678349556311e-05, + "loss": 3.2883, + "step": 4451 + }, + { + "epoch": 0.3569381251127458, + "grad_norm": 1.9556522369384766, + "learning_rate": 1.8866183036107814e-05, + "loss": 3.2799, + "step": 4452 + }, + { + "epoch": 0.35701829989376843, + "grad_norm": 1.9157310724258423, + "learning_rate": 1.88655824271718e-05, + "loss": 3.391, + "step": 4453 + }, + { + "epoch": 0.35709847467479106, + "grad_norm": 2.0071966648101807, + "learning_rate": 1.8864981668765185e-05, + "loss": 3.2987, + "step": 4454 + }, + { + "epoch": 0.3571786494558137, + "grad_norm": 1.9012686014175415, + "learning_rate": 1.8864380760898103e-05, + "loss": 3.2239, + "step": 4455 + }, + { + "epoch": 0.3572588242368363, + "grad_norm": 1.9088786840438843, + "learning_rate": 1.886377970358069e-05, + "loss": 3.1395, + "step": 4456 + }, + { + "epoch": 0.3573389990178589, + "grad_norm": 1.904780626296997, + "learning_rate": 1.8863178496823075e-05, + "loss": 3.2527, + "step": 4457 + }, + { + "epoch": 0.35741917379888155, + "grad_norm": 1.7530317306518555, + "learning_rate": 1.886257714063539e-05, + "loss": 3.1409, + "step": 4458 + }, + { + "epoch": 0.3574993485799042, + "grad_norm": 2.0022621154785156, + "learning_rate": 1.886197563502778e-05, + "loss": 3.239, + "step": 4459 + }, + { + "epoch": 0.3575795233609268, + "grad_norm": 1.946103572845459, + "learning_rate": 1.8861373980010386e-05, + "loss": 3.312, + "step": 4460 + }, + { + "epoch": 0.3576596981419494, + "grad_norm": 1.9781253337860107, + "learning_rate": 1.8860772175593346e-05, + "loss": 3.1557, + "step": 4461 + }, + { + "epoch": 0.3577398729229721, + "grad_norm": 1.9896271228790283, + "learning_rate": 1.886017022178681e-05, + "loss": 3.3005, + "step": 4462 + }, + { + "epoch": 0.3578200477039947, + "grad_norm": 1.925900936126709, + "learning_rate": 1.8859568118600934e-05, + "loss": 3.2242, + "step": 4463 + }, + { + "epoch": 0.35790022248501735, + "grad_norm": 2.0114905834198, + "learning_rate": 1.8858965866045853e-05, + "loss": 3.068, + "step": 4464 + }, + { + "epoch": 0.35798039726604, + "grad_norm": 1.8125743865966797, + "learning_rate": 1.885836346413173e-05, + "loss": 3.1632, + "step": 4465 + }, + { + "epoch": 0.3580605720470626, + "grad_norm": 1.7865054607391357, + "learning_rate": 1.8857760912868727e-05, + "loss": 3.2454, + "step": 4466 + }, + { + "epoch": 0.3581407468280852, + "grad_norm": 1.860408067703247, + "learning_rate": 1.8857158212266993e-05, + "loss": 3.2697, + "step": 4467 + }, + { + "epoch": 0.35822092160910785, + "grad_norm": 1.8555052280426025, + "learning_rate": 1.88565553623367e-05, + "loss": 3.1971, + "step": 4468 + }, + { + "epoch": 0.35830109639013047, + "grad_norm": 1.8837809562683105, + "learning_rate": 1.8855952363087997e-05, + "loss": 3.1661, + "step": 4469 + }, + { + "epoch": 0.3583812711711531, + "grad_norm": 1.9417997598648071, + "learning_rate": 1.8855349214531056e-05, + "loss": 3.1863, + "step": 4470 + }, + { + "epoch": 0.3584614459521757, + "grad_norm": 1.8410334587097168, + "learning_rate": 1.8854745916676054e-05, + "loss": 3.1866, + "step": 4471 + }, + { + "epoch": 0.3585416207331984, + "grad_norm": 0.9086436629295349, + "learning_rate": 1.8854142469533156e-05, + "loss": 2.5245, + "step": 4472 + }, + { + "epoch": 0.358621795514221, + "grad_norm": 1.778642177581787, + "learning_rate": 1.885353887311254e-05, + "loss": 3.2523, + "step": 4473 + }, + { + "epoch": 0.35870197029524364, + "grad_norm": 0.8629840016365051, + "learning_rate": 1.885293512742437e-05, + "loss": 2.5331, + "step": 4474 + }, + { + "epoch": 0.35878214507626627, + "grad_norm": 1.9319806098937988, + "learning_rate": 1.885233123247884e-05, + "loss": 3.1903, + "step": 4475 + }, + { + "epoch": 0.3588623198572889, + "grad_norm": 1.7944132089614868, + "learning_rate": 1.8851727188286122e-05, + "loss": 3.2568, + "step": 4476 + }, + { + "epoch": 0.3589424946383115, + "grad_norm": 1.8945879936218262, + "learning_rate": 1.885112299485641e-05, + "loss": 3.3326, + "step": 4477 + }, + { + "epoch": 0.35902266941933414, + "grad_norm": 1.9172887802124023, + "learning_rate": 1.8850518652199877e-05, + "loss": 3.2699, + "step": 4478 + }, + { + "epoch": 0.35910284420035676, + "grad_norm": 1.9558783769607544, + "learning_rate": 1.884991416032672e-05, + "loss": 3.2986, + "step": 4479 + }, + { + "epoch": 0.3591830189813794, + "grad_norm": 1.953784465789795, + "learning_rate": 1.8849309519247128e-05, + "loss": 3.3029, + "step": 4480 + }, + { + "epoch": 0.359263193762402, + "grad_norm": 1.9450770616531372, + "learning_rate": 1.88487047289713e-05, + "loss": 3.2582, + "step": 4481 + }, + { + "epoch": 0.3593433685434247, + "grad_norm": 1.9166477918624878, + "learning_rate": 1.8848099789509423e-05, + "loss": 3.274, + "step": 4482 + }, + { + "epoch": 0.3594235433244473, + "grad_norm": 1.9042506217956543, + "learning_rate": 1.8847494700871706e-05, + "loss": 3.2231, + "step": 4483 + }, + { + "epoch": 0.35950371810546994, + "grad_norm": 2.0987472534179688, + "learning_rate": 1.884688946306835e-05, + "loss": 3.3496, + "step": 4484 + }, + { + "epoch": 0.35958389288649256, + "grad_norm": 1.9069545269012451, + "learning_rate": 1.884628407610955e-05, + "loss": 3.3024, + "step": 4485 + }, + { + "epoch": 0.3596640676675152, + "grad_norm": 1.9809916019439697, + "learning_rate": 1.8845678540005518e-05, + "loss": 3.2645, + "step": 4486 + }, + { + "epoch": 0.3597442424485378, + "grad_norm": 1.9480303525924683, + "learning_rate": 1.8845072854766467e-05, + "loss": 3.3167, + "step": 4487 + }, + { + "epoch": 0.35982441722956043, + "grad_norm": 1.8892587423324585, + "learning_rate": 1.8844467020402603e-05, + "loss": 3.1636, + "step": 4488 + }, + { + "epoch": 0.35990459201058306, + "grad_norm": 1.750543475151062, + "learning_rate": 1.8843861036924142e-05, + "loss": 3.2466, + "step": 4489 + }, + { + "epoch": 0.3599847667916057, + "grad_norm": 1.9657001495361328, + "learning_rate": 1.8843254904341303e-05, + "loss": 3.1719, + "step": 4490 + }, + { + "epoch": 0.3600649415726283, + "grad_norm": 2.025745153427124, + "learning_rate": 1.88426486226643e-05, + "loss": 3.2489, + "step": 4491 + }, + { + "epoch": 0.360145116353651, + "grad_norm": 0.8581278920173645, + "learning_rate": 1.884204219190336e-05, + "loss": 2.5355, + "step": 4492 + }, + { + "epoch": 0.3602252911346736, + "grad_norm": 1.9374067783355713, + "learning_rate": 1.8841435612068703e-05, + "loss": 3.2766, + "step": 4493 + }, + { + "epoch": 0.36030546591569623, + "grad_norm": 1.8645446300506592, + "learning_rate": 1.884082888317056e-05, + "loss": 3.1887, + "step": 4494 + }, + { + "epoch": 0.36038564069671886, + "grad_norm": 1.952170968055725, + "learning_rate": 1.8840222005219156e-05, + "loss": 3.2201, + "step": 4495 + }, + { + "epoch": 0.3604658154777415, + "grad_norm": 1.931240439414978, + "learning_rate": 1.8839614978224724e-05, + "loss": 3.2728, + "step": 4496 + }, + { + "epoch": 0.3605459902587641, + "grad_norm": 1.9001437425613403, + "learning_rate": 1.88390078021975e-05, + "loss": 3.2892, + "step": 4497 + }, + { + "epoch": 0.36062616503978673, + "grad_norm": 1.840685248374939, + "learning_rate": 1.883840047714772e-05, + "loss": 3.2797, + "step": 4498 + }, + { + "epoch": 0.36070633982080935, + "grad_norm": 1.7354681491851807, + "learning_rate": 1.883779300308562e-05, + "loss": 3.1847, + "step": 4499 + }, + { + "epoch": 0.360786514601832, + "grad_norm": 1.9775348901748657, + "learning_rate": 1.883718538002145e-05, + "loss": 3.2406, + "step": 4500 + }, + { + "epoch": 0.3608666893828546, + "grad_norm": 1.8991014957427979, + "learning_rate": 1.8836577607965446e-05, + "loss": 3.1965, + "step": 4501 + }, + { + "epoch": 0.3609468641638773, + "grad_norm": 1.978350281715393, + "learning_rate": 1.8835969686927856e-05, + "loss": 3.4241, + "step": 4502 + }, + { + "epoch": 0.3610270389448999, + "grad_norm": 1.9288362264633179, + "learning_rate": 1.8835361616918933e-05, + "loss": 3.1036, + "step": 4503 + }, + { + "epoch": 0.3611072137259225, + "grad_norm": 1.9852700233459473, + "learning_rate": 1.8834753397948926e-05, + "loss": 3.2607, + "step": 4504 + }, + { + "epoch": 0.36118738850694515, + "grad_norm": 1.8613216876983643, + "learning_rate": 1.8834145030028093e-05, + "loss": 3.2976, + "step": 4505 + }, + { + "epoch": 0.3612675632879678, + "grad_norm": 2.0241031646728516, + "learning_rate": 1.8833536513166685e-05, + "loss": 3.3418, + "step": 4506 + }, + { + "epoch": 0.3613477380689904, + "grad_norm": 2.0473203659057617, + "learning_rate": 1.883292784737497e-05, + "loss": 3.2823, + "step": 4507 + }, + { + "epoch": 0.361427912850013, + "grad_norm": 1.770240068435669, + "learning_rate": 1.8832319032663198e-05, + "loss": 3.3292, + "step": 4508 + }, + { + "epoch": 0.36150808763103565, + "grad_norm": 1.9564950466156006, + "learning_rate": 1.8831710069041644e-05, + "loss": 3.1552, + "step": 4509 + }, + { + "epoch": 0.36158826241205827, + "grad_norm": 1.8219215869903564, + "learning_rate": 1.883110095652057e-05, + "loss": 3.1838, + "step": 4510 + }, + { + "epoch": 0.3616684371930809, + "grad_norm": 0.8824937343597412, + "learning_rate": 1.8830491695110243e-05, + "loss": 2.5112, + "step": 4511 + }, + { + "epoch": 0.3617486119741036, + "grad_norm": 1.802751898765564, + "learning_rate": 1.8829882284820942e-05, + "loss": 3.1976, + "step": 4512 + }, + { + "epoch": 0.3618287867551262, + "grad_norm": 1.8011109828948975, + "learning_rate": 1.8829272725662937e-05, + "loss": 3.2064, + "step": 4513 + }, + { + "epoch": 0.3619089615361488, + "grad_norm": 1.9699091911315918, + "learning_rate": 1.8828663017646505e-05, + "loss": 3.2001, + "step": 4514 + }, + { + "epoch": 0.36198913631717144, + "grad_norm": 1.8936394453048706, + "learning_rate": 1.8828053160781925e-05, + "loss": 3.1598, + "step": 4515 + }, + { + "epoch": 0.36206931109819407, + "grad_norm": 1.8725218772888184, + "learning_rate": 1.882744315507948e-05, + "loss": 3.3524, + "step": 4516 + }, + { + "epoch": 0.3621494858792167, + "grad_norm": 0.8518058061599731, + "learning_rate": 1.8826833000549458e-05, + "loss": 2.5206, + "step": 4517 + }, + { + "epoch": 0.3622296606602393, + "grad_norm": 1.9874541759490967, + "learning_rate": 1.882622269720214e-05, + "loss": 3.2772, + "step": 4518 + }, + { + "epoch": 0.36230983544126194, + "grad_norm": 1.8972949981689453, + "learning_rate": 1.882561224504782e-05, + "loss": 3.2645, + "step": 4519 + }, + { + "epoch": 0.36239001022228456, + "grad_norm": 1.8982256650924683, + "learning_rate": 1.8825001644096785e-05, + "loss": 3.1838, + "step": 4520 + }, + { + "epoch": 0.3624701850033072, + "grad_norm": 1.8648825883865356, + "learning_rate": 1.8824390894359332e-05, + "loss": 3.195, + "step": 4521 + }, + { + "epoch": 0.3625503597843298, + "grad_norm": 1.951767086982727, + "learning_rate": 1.882377999584576e-05, + "loss": 3.2042, + "step": 4522 + }, + { + "epoch": 0.3626305345653525, + "grad_norm": 1.896767497062683, + "learning_rate": 1.882316894856637e-05, + "loss": 3.2501, + "step": 4523 + }, + { + "epoch": 0.3627107093463751, + "grad_norm": 1.8246163129806519, + "learning_rate": 1.8822557752531457e-05, + "loss": 3.2227, + "step": 4524 + }, + { + "epoch": 0.36279088412739774, + "grad_norm": 0.767707109451294, + "learning_rate": 1.8821946407751333e-05, + "loss": 2.4937, + "step": 4525 + }, + { + "epoch": 0.36287105890842036, + "grad_norm": 1.8326988220214844, + "learning_rate": 1.88213349142363e-05, + "loss": 3.2464, + "step": 4526 + }, + { + "epoch": 0.362951233689443, + "grad_norm": 1.8578734397888184, + "learning_rate": 1.882072327199667e-05, + "loss": 3.1655, + "step": 4527 + }, + { + "epoch": 0.3630314084704656, + "grad_norm": 1.901087999343872, + "learning_rate": 1.882011148104275e-05, + "loss": 3.2562, + "step": 4528 + }, + { + "epoch": 0.36311158325148823, + "grad_norm": 1.9805262088775635, + "learning_rate": 1.881949954138486e-05, + "loss": 3.2906, + "step": 4529 + }, + { + "epoch": 0.36319175803251086, + "grad_norm": 1.9615426063537598, + "learning_rate": 1.881888745303332e-05, + "loss": 3.3036, + "step": 4530 + }, + { + "epoch": 0.3632719328135335, + "grad_norm": 1.8912369012832642, + "learning_rate": 1.8818275215998447e-05, + "loss": 3.2148, + "step": 4531 + }, + { + "epoch": 0.3633521075945561, + "grad_norm": 2.0128514766693115, + "learning_rate": 1.881766283029056e-05, + "loss": 3.2946, + "step": 4532 + }, + { + "epoch": 0.3634322823755788, + "grad_norm": 1.8939379453659058, + "learning_rate": 1.8817050295919982e-05, + "loss": 3.2812, + "step": 4533 + }, + { + "epoch": 0.3635124571566014, + "grad_norm": 1.898579478263855, + "learning_rate": 1.8816437612897048e-05, + "loss": 3.3109, + "step": 4534 + }, + { + "epoch": 0.36359263193762403, + "grad_norm": 0.8494036197662354, + "learning_rate": 1.8815824781232076e-05, + "loss": 2.5051, + "step": 4535 + }, + { + "epoch": 0.36367280671864666, + "grad_norm": 2.045412063598633, + "learning_rate": 1.8815211800935414e-05, + "loss": 3.2369, + "step": 4536 + }, + { + "epoch": 0.3637529814996693, + "grad_norm": 1.9550939798355103, + "learning_rate": 1.8814598672017382e-05, + "loss": 3.2692, + "step": 4537 + }, + { + "epoch": 0.3638331562806919, + "grad_norm": 1.8679102659225464, + "learning_rate": 1.8813985394488325e-05, + "loss": 3.3602, + "step": 4538 + }, + { + "epoch": 0.36391333106171453, + "grad_norm": 2.0007479190826416, + "learning_rate": 1.8813371968358578e-05, + "loss": 3.3195, + "step": 4539 + }, + { + "epoch": 0.36399350584273715, + "grad_norm": 1.9545258283615112, + "learning_rate": 1.881275839363849e-05, + "loss": 3.2272, + "step": 4540 + }, + { + "epoch": 0.3640736806237598, + "grad_norm": 2.227914810180664, + "learning_rate": 1.8812144670338398e-05, + "loss": 3.2367, + "step": 4541 + }, + { + "epoch": 0.3641538554047824, + "grad_norm": 2.013424873352051, + "learning_rate": 1.8811530798468655e-05, + "loss": 3.2416, + "step": 4542 + }, + { + "epoch": 0.3642340301858051, + "grad_norm": 1.9015127420425415, + "learning_rate": 1.8810916778039607e-05, + "loss": 3.1901, + "step": 4543 + }, + { + "epoch": 0.3643142049668277, + "grad_norm": 0.8406859040260315, + "learning_rate": 1.881030260906161e-05, + "loss": 2.5081, + "step": 4544 + }, + { + "epoch": 0.3643943797478503, + "grad_norm": 1.9952211380004883, + "learning_rate": 1.880968829154501e-05, + "loss": 3.2598, + "step": 4545 + }, + { + "epoch": 0.36447455452887295, + "grad_norm": 2.030416250228882, + "learning_rate": 1.8809073825500176e-05, + "loss": 3.2659, + "step": 4546 + }, + { + "epoch": 0.3645547293098956, + "grad_norm": 1.917624592781067, + "learning_rate": 1.8808459210937458e-05, + "loss": 3.3962, + "step": 4547 + }, + { + "epoch": 0.3646349040909182, + "grad_norm": 1.946649193763733, + "learning_rate": 1.8807844447867224e-05, + "loss": 3.3092, + "step": 4548 + }, + { + "epoch": 0.3647150788719408, + "grad_norm": 2.0303707122802734, + "learning_rate": 1.8807229536299837e-05, + "loss": 3.25, + "step": 4549 + }, + { + "epoch": 0.36479525365296345, + "grad_norm": 2.099742889404297, + "learning_rate": 1.8806614476245662e-05, + "loss": 3.2534, + "step": 4550 + }, + { + "epoch": 0.36487542843398607, + "grad_norm": 1.9451801776885986, + "learning_rate": 1.8805999267715073e-05, + "loss": 3.2308, + "step": 4551 + }, + { + "epoch": 0.3649556032150087, + "grad_norm": 2.0360004901885986, + "learning_rate": 1.880538391071844e-05, + "loss": 3.2212, + "step": 4552 + }, + { + "epoch": 0.3650357779960314, + "grad_norm": 0.8012794852256775, + "learning_rate": 1.8804768405266135e-05, + "loss": 2.5264, + "step": 4553 + }, + { + "epoch": 0.365115952777054, + "grad_norm": 1.8848601579666138, + "learning_rate": 1.8804152751368542e-05, + "loss": 3.2366, + "step": 4554 + }, + { + "epoch": 0.3651961275580766, + "grad_norm": 1.984095573425293, + "learning_rate": 1.880353694903603e-05, + "loss": 3.2254, + "step": 4555 + }, + { + "epoch": 0.36527630233909925, + "grad_norm": 1.960176706314087, + "learning_rate": 1.8802920998278995e-05, + "loss": 3.2238, + "step": 4556 + }, + { + "epoch": 0.36535647712012187, + "grad_norm": 1.8475584983825684, + "learning_rate": 1.880230489910781e-05, + "loss": 3.2996, + "step": 4557 + }, + { + "epoch": 0.3654366519011445, + "grad_norm": 1.8663572072982788, + "learning_rate": 1.8801688651532868e-05, + "loss": 3.2776, + "step": 4558 + }, + { + "epoch": 0.3655168266821671, + "grad_norm": 1.9910551309585571, + "learning_rate": 1.880107225556456e-05, + "loss": 3.2385, + "step": 4559 + }, + { + "epoch": 0.36559700146318974, + "grad_norm": 2.1491518020629883, + "learning_rate": 1.880045571121327e-05, + "loss": 3.2894, + "step": 4560 + }, + { + "epoch": 0.36567717624421237, + "grad_norm": 2.1208810806274414, + "learning_rate": 1.8799839018489404e-05, + "loss": 3.2025, + "step": 4561 + }, + { + "epoch": 0.365757351025235, + "grad_norm": 0.9151220321655273, + "learning_rate": 1.8799222177403348e-05, + "loss": 2.5356, + "step": 4562 + }, + { + "epoch": 0.36583752580625767, + "grad_norm": 1.9609273672103882, + "learning_rate": 1.8798605187965512e-05, + "loss": 3.3446, + "step": 4563 + }, + { + "epoch": 0.3659177005872803, + "grad_norm": 2.003688097000122, + "learning_rate": 1.8797988050186288e-05, + "loss": 3.2974, + "step": 4564 + }, + { + "epoch": 0.3659978753683029, + "grad_norm": 1.844328761100769, + "learning_rate": 1.879737076407609e-05, + "loss": 3.2513, + "step": 4565 + }, + { + "epoch": 0.36607805014932554, + "grad_norm": 0.8904287815093994, + "learning_rate": 1.8796753329645322e-05, + "loss": 2.5182, + "step": 4566 + }, + { + "epoch": 0.36615822493034816, + "grad_norm": 1.9087327718734741, + "learning_rate": 1.879613574690439e-05, + "loss": 3.1833, + "step": 4567 + }, + { + "epoch": 0.3662383997113708, + "grad_norm": 2.1210460662841797, + "learning_rate": 1.8795518015863713e-05, + "loss": 3.2703, + "step": 4568 + }, + { + "epoch": 0.3663185744923934, + "grad_norm": 2.044801712036133, + "learning_rate": 1.87949001365337e-05, + "loss": 3.2296, + "step": 4569 + }, + { + "epoch": 0.36639874927341604, + "grad_norm": 0.9155616760253906, + "learning_rate": 1.8794282108924772e-05, + "loss": 2.5189, + "step": 4570 + }, + { + "epoch": 0.36647892405443866, + "grad_norm": 2.072857141494751, + "learning_rate": 1.8793663933047342e-05, + "loss": 3.2106, + "step": 4571 + }, + { + "epoch": 0.3665590988354613, + "grad_norm": 1.8201792240142822, + "learning_rate": 1.8793045608911844e-05, + "loss": 3.2108, + "step": 4572 + }, + { + "epoch": 0.3666392736164839, + "grad_norm": 1.941070318222046, + "learning_rate": 1.8792427136528694e-05, + "loss": 3.3386, + "step": 4573 + }, + { + "epoch": 0.3667194483975066, + "grad_norm": 2.037782669067383, + "learning_rate": 1.879180851590832e-05, + "loss": 3.2164, + "step": 4574 + }, + { + "epoch": 0.3667996231785292, + "grad_norm": 0.8911155462265015, + "learning_rate": 1.8791189747061152e-05, + "loss": 2.5381, + "step": 4575 + }, + { + "epoch": 0.36687979795955183, + "grad_norm": 1.8166437149047852, + "learning_rate": 1.8790570829997624e-05, + "loss": 3.2301, + "step": 4576 + }, + { + "epoch": 0.36695997274057446, + "grad_norm": 1.8208822011947632, + "learning_rate": 1.8789951764728173e-05, + "loss": 3.2864, + "step": 4577 + }, + { + "epoch": 0.3670401475215971, + "grad_norm": 1.9924194812774658, + "learning_rate": 1.878933255126323e-05, + "loss": 3.3275, + "step": 4578 + }, + { + "epoch": 0.3671203223026197, + "grad_norm": 1.9115023612976074, + "learning_rate": 1.878871318961324e-05, + "loss": 3.3582, + "step": 4579 + }, + { + "epoch": 0.36720049708364233, + "grad_norm": 1.8078248500823975, + "learning_rate": 1.8788093679788638e-05, + "loss": 3.2607, + "step": 4580 + }, + { + "epoch": 0.36728067186466495, + "grad_norm": 1.9433948993682861, + "learning_rate": 1.878747402179988e-05, + "loss": 3.3069, + "step": 4581 + }, + { + "epoch": 0.3673608466456876, + "grad_norm": 1.9923577308654785, + "learning_rate": 1.8786854215657403e-05, + "loss": 3.2964, + "step": 4582 + }, + { + "epoch": 0.3674410214267102, + "grad_norm": 1.8737437725067139, + "learning_rate": 1.878623426137166e-05, + "loss": 3.1843, + "step": 4583 + }, + { + "epoch": 0.3675211962077329, + "grad_norm": 1.9131028652191162, + "learning_rate": 1.878561415895311e-05, + "loss": 3.4002, + "step": 4584 + }, + { + "epoch": 0.3676013709887555, + "grad_norm": 2.0016469955444336, + "learning_rate": 1.8784993908412195e-05, + "loss": 3.2772, + "step": 4585 + }, + { + "epoch": 0.36768154576977813, + "grad_norm": 2.6055104732513428, + "learning_rate": 1.878437350975938e-05, + "loss": 3.3062, + "step": 4586 + }, + { + "epoch": 0.36776172055080075, + "grad_norm": 2.006032943725586, + "learning_rate": 1.8783752963005124e-05, + "loss": 3.3599, + "step": 4587 + }, + { + "epoch": 0.3678418953318234, + "grad_norm": 1.8247907161712646, + "learning_rate": 1.878313226815989e-05, + "loss": 3.2936, + "step": 4588 + }, + { + "epoch": 0.367922070112846, + "grad_norm": 1.9156153202056885, + "learning_rate": 1.8782511425234143e-05, + "loss": 3.2206, + "step": 4589 + }, + { + "epoch": 0.3680022448938686, + "grad_norm": 1.8830639123916626, + "learning_rate": 1.8781890434238345e-05, + "loss": 3.2738, + "step": 4590 + }, + { + "epoch": 0.36808241967489125, + "grad_norm": 1.9101437330245972, + "learning_rate": 1.878126929518297e-05, + "loss": 3.1874, + "step": 4591 + }, + { + "epoch": 0.36816259445591387, + "grad_norm": 1.8105751276016235, + "learning_rate": 1.878064800807849e-05, + "loss": 3.2809, + "step": 4592 + }, + { + "epoch": 0.3682427692369365, + "grad_norm": 0.8333619832992554, + "learning_rate": 1.878002657293538e-05, + "loss": 2.5153, + "step": 4593 + }, + { + "epoch": 0.3683229440179592, + "grad_norm": 2.075005054473877, + "learning_rate": 1.8779404989764117e-05, + "loss": 3.3668, + "step": 4594 + }, + { + "epoch": 0.3684031187989818, + "grad_norm": 2.0878095626831055, + "learning_rate": 1.8778783258575178e-05, + "loss": 3.2504, + "step": 4595 + }, + { + "epoch": 0.3684832935800044, + "grad_norm": 1.8086611032485962, + "learning_rate": 1.877816137937905e-05, + "loss": 3.2765, + "step": 4596 + }, + { + "epoch": 0.36856346836102705, + "grad_norm": 1.966792345046997, + "learning_rate": 1.877753935218621e-05, + "loss": 3.3711, + "step": 4597 + }, + { + "epoch": 0.36864364314204967, + "grad_norm": 1.8450535535812378, + "learning_rate": 1.8776917177007155e-05, + "loss": 3.1353, + "step": 4598 + }, + { + "epoch": 0.3687238179230723, + "grad_norm": 1.9261813163757324, + "learning_rate": 1.8776294853852368e-05, + "loss": 3.2166, + "step": 4599 + }, + { + "epoch": 0.3688039927040949, + "grad_norm": 0.8597177267074585, + "learning_rate": 1.8775672382732343e-05, + "loss": 2.5195, + "step": 4600 + }, + { + "epoch": 0.36888416748511754, + "grad_norm": 1.8698787689208984, + "learning_rate": 1.8775049763657573e-05, + "loss": 3.2568, + "step": 4601 + }, + { + "epoch": 0.36896434226614017, + "grad_norm": 2.2550418376922607, + "learning_rate": 1.8774426996638558e-05, + "loss": 3.201, + "step": 4602 + }, + { + "epoch": 0.3690445170471628, + "grad_norm": 0.9110979437828064, + "learning_rate": 1.8773804081685795e-05, + "loss": 2.5272, + "step": 4603 + }, + { + "epoch": 0.36912469182818547, + "grad_norm": 1.9356518983840942, + "learning_rate": 1.8773181018809793e-05, + "loss": 3.2709, + "step": 4604 + }, + { + "epoch": 0.3692048666092081, + "grad_norm": 1.8971829414367676, + "learning_rate": 1.8772557808021042e-05, + "loss": 3.1689, + "step": 4605 + }, + { + "epoch": 0.3692850413902307, + "grad_norm": 1.9407752752304077, + "learning_rate": 1.8771934449330062e-05, + "loss": 3.2542, + "step": 4606 + }, + { + "epoch": 0.36936521617125334, + "grad_norm": 0.8002964854240417, + "learning_rate": 1.877131094274736e-05, + "loss": 2.5006, + "step": 4607 + }, + { + "epoch": 0.36944539095227596, + "grad_norm": 1.9505810737609863, + "learning_rate": 1.8770687288283446e-05, + "loss": 3.2086, + "step": 4608 + }, + { + "epoch": 0.3695255657332986, + "grad_norm": 2.097822427749634, + "learning_rate": 1.8770063485948837e-05, + "loss": 3.3269, + "step": 4609 + }, + { + "epoch": 0.3696057405143212, + "grad_norm": 1.8311541080474854, + "learning_rate": 1.8769439535754045e-05, + "loss": 3.1679, + "step": 4610 + }, + { + "epoch": 0.36968591529534384, + "grad_norm": 2.06825852394104, + "learning_rate": 1.8768815437709595e-05, + "loss": 3.2568, + "step": 4611 + }, + { + "epoch": 0.36976609007636646, + "grad_norm": 1.8726776838302612, + "learning_rate": 1.876819119182601e-05, + "loss": 3.3361, + "step": 4612 + }, + { + "epoch": 0.3698462648573891, + "grad_norm": 1.8994171619415283, + "learning_rate": 1.8767566798113806e-05, + "loss": 3.1649, + "step": 4613 + }, + { + "epoch": 0.36992643963841176, + "grad_norm": 1.8548119068145752, + "learning_rate": 1.8766942256583518e-05, + "loss": 3.2854, + "step": 4614 + }, + { + "epoch": 0.3700066144194344, + "grad_norm": 1.8762329816818237, + "learning_rate": 1.8766317567245676e-05, + "loss": 3.2329, + "step": 4615 + }, + { + "epoch": 0.370086789200457, + "grad_norm": 1.8421988487243652, + "learning_rate": 1.8765692730110807e-05, + "loss": 3.187, + "step": 4616 + }, + { + "epoch": 0.37016696398147964, + "grad_norm": 1.9002615213394165, + "learning_rate": 1.8765067745189447e-05, + "loss": 3.2646, + "step": 4617 + }, + { + "epoch": 0.37024713876250226, + "grad_norm": 2.006781578063965, + "learning_rate": 1.8764442612492137e-05, + "loss": 3.1364, + "step": 4618 + }, + { + "epoch": 0.3703273135435249, + "grad_norm": 1.8322311639785767, + "learning_rate": 1.8763817332029412e-05, + "loss": 3.2921, + "step": 4619 + }, + { + "epoch": 0.3704074883245475, + "grad_norm": 0.8296230435371399, + "learning_rate": 1.8763191903811814e-05, + "loss": 2.506, + "step": 4620 + }, + { + "epoch": 0.37048766310557013, + "grad_norm": 1.9064009189605713, + "learning_rate": 1.876256632784989e-05, + "loss": 3.2102, + "step": 4621 + }, + { + "epoch": 0.37056783788659275, + "grad_norm": 1.9610190391540527, + "learning_rate": 1.8761940604154185e-05, + "loss": 3.1565, + "step": 4622 + }, + { + "epoch": 0.3706480126676154, + "grad_norm": 2.00272798538208, + "learning_rate": 1.876131473273525e-05, + "loss": 3.1849, + "step": 4623 + }, + { + "epoch": 0.37072818744863806, + "grad_norm": 1.914533257484436, + "learning_rate": 1.8760688713603637e-05, + "loss": 3.2257, + "step": 4624 + }, + { + "epoch": 0.3708083622296607, + "grad_norm": 2.1001369953155518, + "learning_rate": 1.87600625467699e-05, + "loss": 3.3589, + "step": 4625 + }, + { + "epoch": 0.3708885370106833, + "grad_norm": 2.127488136291504, + "learning_rate": 1.8759436232244596e-05, + "loss": 3.3228, + "step": 4626 + }, + { + "epoch": 0.37096871179170593, + "grad_norm": 0.891227662563324, + "learning_rate": 1.875880977003828e-05, + "loss": 2.5165, + "step": 4627 + }, + { + "epoch": 0.37104888657272855, + "grad_norm": 1.9078054428100586, + "learning_rate": 1.8758183160161526e-05, + "loss": 3.3137, + "step": 4628 + }, + { + "epoch": 0.3711290613537512, + "grad_norm": 1.8633302450180054, + "learning_rate": 1.8757556402624885e-05, + "loss": 3.2277, + "step": 4629 + }, + { + "epoch": 0.3712092361347738, + "grad_norm": 1.9433517456054688, + "learning_rate": 1.8756929497438932e-05, + "loss": 3.2455, + "step": 4630 + }, + { + "epoch": 0.3712894109157964, + "grad_norm": 1.8199448585510254, + "learning_rate": 1.8756302444614232e-05, + "loss": 3.2091, + "step": 4631 + }, + { + "epoch": 0.37136958569681905, + "grad_norm": 1.8025505542755127, + "learning_rate": 1.875567524416136e-05, + "loss": 3.298, + "step": 4632 + }, + { + "epoch": 0.3714497604778417, + "grad_norm": 1.8025574684143066, + "learning_rate": 1.8755047896090888e-05, + "loss": 3.1642, + "step": 4633 + }, + { + "epoch": 0.3715299352588643, + "grad_norm": 1.8665813207626343, + "learning_rate": 1.8754420400413392e-05, + "loss": 3.24, + "step": 4634 + }, + { + "epoch": 0.371610110039887, + "grad_norm": 1.8973525762557983, + "learning_rate": 1.875379275713946e-05, + "loss": 3.1594, + "step": 4635 + }, + { + "epoch": 0.3716902848209096, + "grad_norm": 2.1672236919403076, + "learning_rate": 1.8753164966279662e-05, + "loss": 3.2285, + "step": 4636 + }, + { + "epoch": 0.3717704596019322, + "grad_norm": 1.9709011316299438, + "learning_rate": 1.875253702784459e-05, + "loss": 3.3359, + "step": 4637 + }, + { + "epoch": 0.37185063438295485, + "grad_norm": 1.9350595474243164, + "learning_rate": 1.875190894184483e-05, + "loss": 3.3045, + "step": 4638 + }, + { + "epoch": 0.37193080916397747, + "grad_norm": 0.8481950759887695, + "learning_rate": 1.8751280708290963e-05, + "loss": 2.5118, + "step": 4639 + }, + { + "epoch": 0.3720109839450001, + "grad_norm": 1.9318805932998657, + "learning_rate": 1.8750652327193593e-05, + "loss": 3.3113, + "step": 4640 + }, + { + "epoch": 0.3720911587260227, + "grad_norm": 1.9270602464675903, + "learning_rate": 1.8750023798563308e-05, + "loss": 3.4023, + "step": 4641 + }, + { + "epoch": 0.37217133350704534, + "grad_norm": 2.0618739128112793, + "learning_rate": 1.8749395122410705e-05, + "loss": 3.2689, + "step": 4642 + }, + { + "epoch": 0.37225150828806797, + "grad_norm": 2.092862367630005, + "learning_rate": 1.8748766298746382e-05, + "loss": 3.2113, + "step": 4643 + }, + { + "epoch": 0.3723316830690906, + "grad_norm": 1.9653335809707642, + "learning_rate": 1.8748137327580945e-05, + "loss": 3.2684, + "step": 4644 + }, + { + "epoch": 0.37241185785011327, + "grad_norm": 1.9795969724655151, + "learning_rate": 1.8747508208924994e-05, + "loss": 3.2332, + "step": 4645 + }, + { + "epoch": 0.3724920326311359, + "grad_norm": 1.845009684562683, + "learning_rate": 1.8746878942789143e-05, + "loss": 3.1736, + "step": 4646 + }, + { + "epoch": 0.3725722074121585, + "grad_norm": 1.9733963012695312, + "learning_rate": 1.8746249529183987e-05, + "loss": 3.1704, + "step": 4647 + }, + { + "epoch": 0.37265238219318114, + "grad_norm": 1.7648435831069946, + "learning_rate": 1.8745619968120153e-05, + "loss": 3.2081, + "step": 4648 + }, + { + "epoch": 0.37273255697420377, + "grad_norm": 1.9583704471588135, + "learning_rate": 1.874499025960825e-05, + "loss": 3.1464, + "step": 4649 + }, + { + "epoch": 0.3728127317552264, + "grad_norm": 1.948986530303955, + "learning_rate": 1.8744360403658888e-05, + "loss": 3.1953, + "step": 4650 + }, + { + "epoch": 0.372892906536249, + "grad_norm": 0.8541949391365051, + "learning_rate": 1.8743730400282692e-05, + "loss": 2.5229, + "step": 4651 + }, + { + "epoch": 0.37297308131727164, + "grad_norm": 1.7848334312438965, + "learning_rate": 1.8743100249490285e-05, + "loss": 3.2388, + "step": 4652 + }, + { + "epoch": 0.37305325609829426, + "grad_norm": 1.8904905319213867, + "learning_rate": 1.874246995129229e-05, + "loss": 3.1619, + "step": 4653 + }, + { + "epoch": 0.3731334308793169, + "grad_norm": 2.0260581970214844, + "learning_rate": 1.8741839505699328e-05, + "loss": 3.3226, + "step": 4654 + }, + { + "epoch": 0.37321360566033956, + "grad_norm": 1.933266520500183, + "learning_rate": 1.8741208912722038e-05, + "loss": 3.2446, + "step": 4655 + }, + { + "epoch": 0.3732937804413622, + "grad_norm": 1.8996682167053223, + "learning_rate": 1.8740578172371043e-05, + "loss": 3.2273, + "step": 4656 + }, + { + "epoch": 0.3733739552223848, + "grad_norm": 1.8482310771942139, + "learning_rate": 1.8739947284656984e-05, + "loss": 3.1974, + "step": 4657 + }, + { + "epoch": 0.37345413000340744, + "grad_norm": 2.6606054306030273, + "learning_rate": 1.8739316249590488e-05, + "loss": 3.2455, + "step": 4658 + }, + { + "epoch": 0.37353430478443006, + "grad_norm": 1.9000176191329956, + "learning_rate": 1.8738685067182208e-05, + "loss": 3.3026, + "step": 4659 + }, + { + "epoch": 0.3736144795654527, + "grad_norm": 1.9496656656265259, + "learning_rate": 1.873805373744277e-05, + "loss": 3.2498, + "step": 4660 + }, + { + "epoch": 0.3736946543464753, + "grad_norm": 1.9714584350585938, + "learning_rate": 1.8737422260382832e-05, + "loss": 3.2142, + "step": 4661 + }, + { + "epoch": 0.37377482912749793, + "grad_norm": 1.8613991737365723, + "learning_rate": 1.873679063601303e-05, + "loss": 3.2516, + "step": 4662 + }, + { + "epoch": 0.37385500390852056, + "grad_norm": 1.921667218208313, + "learning_rate": 1.8736158864344017e-05, + "loss": 3.2319, + "step": 4663 + }, + { + "epoch": 0.3739351786895432, + "grad_norm": 1.9464619159698486, + "learning_rate": 1.8735526945386447e-05, + "loss": 3.3085, + "step": 4664 + }, + { + "epoch": 0.37401535347056586, + "grad_norm": 1.8915153741836548, + "learning_rate": 1.8734894879150972e-05, + "loss": 3.3275, + "step": 4665 + }, + { + "epoch": 0.3740955282515885, + "grad_norm": 0.8363573551177979, + "learning_rate": 1.8734262665648245e-05, + "loss": 2.5231, + "step": 4666 + }, + { + "epoch": 0.3741757030326111, + "grad_norm": 1.9160640239715576, + "learning_rate": 1.873363030488893e-05, + "loss": 3.2015, + "step": 4667 + }, + { + "epoch": 0.37425587781363373, + "grad_norm": 2.169728994369507, + "learning_rate": 1.8732997796883684e-05, + "loss": 3.2379, + "step": 4668 + }, + { + "epoch": 0.37433605259465635, + "grad_norm": 2.040412664413452, + "learning_rate": 1.8732365141643175e-05, + "loss": 3.3734, + "step": 4669 + }, + { + "epoch": 0.374416227375679, + "grad_norm": 1.880067229270935, + "learning_rate": 1.8731732339178064e-05, + "loss": 3.2445, + "step": 4670 + }, + { + "epoch": 0.3744964021567016, + "grad_norm": 1.8566386699676514, + "learning_rate": 1.8731099389499022e-05, + "loss": 3.2949, + "step": 4671 + }, + { + "epoch": 0.3745765769377242, + "grad_norm": 1.881557822227478, + "learning_rate": 1.8730466292616727e-05, + "loss": 3.2111, + "step": 4672 + }, + { + "epoch": 0.37465675171874685, + "grad_norm": 2.0405044555664062, + "learning_rate": 1.8729833048541844e-05, + "loss": 3.2798, + "step": 4673 + }, + { + "epoch": 0.3747369264997695, + "grad_norm": 1.8856240510940552, + "learning_rate": 1.8729199657285052e-05, + "loss": 3.2698, + "step": 4674 + }, + { + "epoch": 0.37481710128079215, + "grad_norm": 1.9375600814819336, + "learning_rate": 1.872856611885703e-05, + "loss": 3.2227, + "step": 4675 + }, + { + "epoch": 0.3748972760618148, + "grad_norm": 2.0503551959991455, + "learning_rate": 1.8727932433268464e-05, + "loss": 3.2913, + "step": 4676 + }, + { + "epoch": 0.3749774508428374, + "grad_norm": 2.0559709072113037, + "learning_rate": 1.872729860053003e-05, + "loss": 3.2338, + "step": 4677 + }, + { + "epoch": 0.37505762562386, + "grad_norm": 1.8338474035263062, + "learning_rate": 1.872666462065242e-05, + "loss": 3.1445, + "step": 4678 + }, + { + "epoch": 0.37513780040488265, + "grad_norm": 1.8718096017837524, + "learning_rate": 1.8726030493646314e-05, + "loss": 3.2748, + "step": 4679 + }, + { + "epoch": 0.3752179751859053, + "grad_norm": 2.1073265075683594, + "learning_rate": 1.8725396219522416e-05, + "loss": 3.3158, + "step": 4680 + }, + { + "epoch": 0.3752981499669279, + "grad_norm": 2.191246747970581, + "learning_rate": 1.872476179829141e-05, + "loss": 3.3224, + "step": 4681 + }, + { + "epoch": 0.3753783247479505, + "grad_norm": 2.0371580123901367, + "learning_rate": 1.8724127229963993e-05, + "loss": 3.144, + "step": 4682 + }, + { + "epoch": 0.37545849952897314, + "grad_norm": 1.8963851928710938, + "learning_rate": 1.8723492514550868e-05, + "loss": 3.2412, + "step": 4683 + }, + { + "epoch": 0.37553867430999577, + "grad_norm": 1.9929038286209106, + "learning_rate": 1.8722857652062733e-05, + "loss": 3.2744, + "step": 4684 + }, + { + "epoch": 0.37561884909101845, + "grad_norm": 1.8409450054168701, + "learning_rate": 1.872222264251029e-05, + "loss": 3.2815, + "step": 4685 + }, + { + "epoch": 0.37569902387204107, + "grad_norm": 1.9473669528961182, + "learning_rate": 1.872158748590425e-05, + "loss": 3.1985, + "step": 4686 + }, + { + "epoch": 0.3757791986530637, + "grad_norm": 2.0326364040374756, + "learning_rate": 1.8720952182255318e-05, + "loss": 3.2526, + "step": 4687 + }, + { + "epoch": 0.3758593734340863, + "grad_norm": 1.9251680374145508, + "learning_rate": 1.872031673157421e-05, + "loss": 3.2307, + "step": 4688 + }, + { + "epoch": 0.37593954821510894, + "grad_norm": 2.0248682498931885, + "learning_rate": 1.8719681133871627e-05, + "loss": 3.209, + "step": 4689 + }, + { + "epoch": 0.37601972299613157, + "grad_norm": 1.889003038406372, + "learning_rate": 1.8719045389158298e-05, + "loss": 3.2465, + "step": 4690 + }, + { + "epoch": 0.3760998977771542, + "grad_norm": 1.8555598258972168, + "learning_rate": 1.8718409497444935e-05, + "loss": 3.1661, + "step": 4691 + }, + { + "epoch": 0.3761800725581768, + "grad_norm": 1.8639880418777466, + "learning_rate": 1.8717773458742265e-05, + "loss": 3.3804, + "step": 4692 + }, + { + "epoch": 0.37626024733919944, + "grad_norm": 1.911234974861145, + "learning_rate": 1.8717137273061e-05, + "loss": 3.2887, + "step": 4693 + }, + { + "epoch": 0.37634042212022206, + "grad_norm": 2.017909526824951, + "learning_rate": 1.8716500940411878e-05, + "loss": 3.243, + "step": 4694 + }, + { + "epoch": 0.3764205969012447, + "grad_norm": 1.876994013786316, + "learning_rate": 1.871586446080562e-05, + "loss": 3.2183, + "step": 4695 + }, + { + "epoch": 0.37650077168226737, + "grad_norm": 2.0578272342681885, + "learning_rate": 1.871522783425296e-05, + "loss": 3.311, + "step": 4696 + }, + { + "epoch": 0.37658094646329, + "grad_norm": 1.9445046186447144, + "learning_rate": 1.871459106076463e-05, + "loss": 3.1995, + "step": 4697 + }, + { + "epoch": 0.3766611212443126, + "grad_norm": 2.0486652851104736, + "learning_rate": 1.8713954140351365e-05, + "loss": 3.2025, + "step": 4698 + }, + { + "epoch": 0.37674129602533524, + "grad_norm": 0.8802306652069092, + "learning_rate": 1.8713317073023905e-05, + "loss": 2.5201, + "step": 4699 + }, + { + "epoch": 0.37682147080635786, + "grad_norm": 2.0133697986602783, + "learning_rate": 1.8712679858792994e-05, + "loss": 3.299, + "step": 4700 + }, + { + "epoch": 0.3769016455873805, + "grad_norm": 2.074139356613159, + "learning_rate": 1.871204249766936e-05, + "loss": 3.2567, + "step": 4701 + }, + { + "epoch": 0.3769818203684031, + "grad_norm": 1.9181174039840698, + "learning_rate": 1.871140498966377e-05, + "loss": 3.1604, + "step": 4702 + }, + { + "epoch": 0.37706199514942573, + "grad_norm": 1.9144517183303833, + "learning_rate": 1.871076733478696e-05, + "loss": 3.1882, + "step": 4703 + }, + { + "epoch": 0.37714216993044836, + "grad_norm": 1.9698066711425781, + "learning_rate": 1.8710129533049684e-05, + "loss": 3.2819, + "step": 4704 + }, + { + "epoch": 0.377222344711471, + "grad_norm": 1.9767789840698242, + "learning_rate": 1.870949158446269e-05, + "loss": 3.2595, + "step": 4705 + }, + { + "epoch": 0.37730251949249366, + "grad_norm": 0.9177917838096619, + "learning_rate": 1.870885348903674e-05, + "loss": 2.5211, + "step": 4706 + }, + { + "epoch": 0.3773826942735163, + "grad_norm": 1.8790239095687866, + "learning_rate": 1.8708215246782588e-05, + "loss": 3.1225, + "step": 4707 + }, + { + "epoch": 0.3774628690545389, + "grad_norm": 2.080371856689453, + "learning_rate": 1.8707576857710995e-05, + "loss": 3.3241, + "step": 4708 + }, + { + "epoch": 0.37754304383556153, + "grad_norm": 1.8757331371307373, + "learning_rate": 1.870693832183273e-05, + "loss": 3.27, + "step": 4709 + }, + { + "epoch": 0.37762321861658416, + "grad_norm": 0.7958487272262573, + "learning_rate": 1.870629963915855e-05, + "loss": 2.5149, + "step": 4710 + }, + { + "epoch": 0.3777033933976068, + "grad_norm": 0.8059672117233276, + "learning_rate": 1.8705660809699228e-05, + "loss": 2.5397, + "step": 4711 + }, + { + "epoch": 0.3777835681786294, + "grad_norm": 1.9318807125091553, + "learning_rate": 1.8705021833465528e-05, + "loss": 3.2403, + "step": 4712 + }, + { + "epoch": 0.377863742959652, + "grad_norm": 0.8028481602668762, + "learning_rate": 1.870438271046823e-05, + "loss": 2.5337, + "step": 4713 + }, + { + "epoch": 0.37794391774067465, + "grad_norm": 1.893377423286438, + "learning_rate": 1.8703743440718113e-05, + "loss": 3.1904, + "step": 4714 + }, + { + "epoch": 0.3780240925216973, + "grad_norm": 1.9776052236557007, + "learning_rate": 1.8703104024225943e-05, + "loss": 3.2345, + "step": 4715 + }, + { + "epoch": 0.37810426730271995, + "grad_norm": 1.9080531597137451, + "learning_rate": 1.8702464461002507e-05, + "loss": 3.2332, + "step": 4716 + }, + { + "epoch": 0.3781844420837426, + "grad_norm": 2.1444742679595947, + "learning_rate": 1.8701824751058592e-05, + "loss": 3.3298, + "step": 4717 + }, + { + "epoch": 0.3782646168647652, + "grad_norm": 1.8915637731552124, + "learning_rate": 1.8701184894404972e-05, + "loss": 3.3119, + "step": 4718 + }, + { + "epoch": 0.3783447916457878, + "grad_norm": 1.9525878429412842, + "learning_rate": 1.8700544891052445e-05, + "loss": 3.2623, + "step": 4719 + }, + { + "epoch": 0.37842496642681045, + "grad_norm": 1.914596676826477, + "learning_rate": 1.8699904741011796e-05, + "loss": 3.2211, + "step": 4720 + }, + { + "epoch": 0.3785051412078331, + "grad_norm": 1.9905388355255127, + "learning_rate": 1.869926444429382e-05, + "loss": 3.3666, + "step": 4721 + }, + { + "epoch": 0.3785853159888557, + "grad_norm": 1.8703277111053467, + "learning_rate": 1.869862400090931e-05, + "loss": 3.2004, + "step": 4722 + }, + { + "epoch": 0.3786654907698783, + "grad_norm": 1.7734389305114746, + "learning_rate": 1.869798341086907e-05, + "loss": 3.219, + "step": 4723 + }, + { + "epoch": 0.37874566555090095, + "grad_norm": 1.941781997680664, + "learning_rate": 1.869734267418389e-05, + "loss": 3.2891, + "step": 4724 + }, + { + "epoch": 0.37882584033192357, + "grad_norm": 1.9682639837265015, + "learning_rate": 1.869670179086458e-05, + "loss": 3.3405, + "step": 4725 + }, + { + "epoch": 0.37890601511294625, + "grad_norm": 1.886771559715271, + "learning_rate": 1.869606076092194e-05, + "loss": 3.2497, + "step": 4726 + }, + { + "epoch": 0.37898618989396887, + "grad_norm": 2.0138649940490723, + "learning_rate": 1.869541958436678e-05, + "loss": 3.2394, + "step": 4727 + }, + { + "epoch": 0.3790663646749915, + "grad_norm": 1.946999430656433, + "learning_rate": 1.8694778261209917e-05, + "loss": 3.2098, + "step": 4728 + }, + { + "epoch": 0.3791465394560141, + "grad_norm": 2.0613391399383545, + "learning_rate": 1.869413679146215e-05, + "loss": 3.3191, + "step": 4729 + }, + { + "epoch": 0.37922671423703674, + "grad_norm": 1.9246293306350708, + "learning_rate": 1.8693495175134304e-05, + "loss": 3.2169, + "step": 4730 + }, + { + "epoch": 0.37930688901805937, + "grad_norm": 1.9708172082901, + "learning_rate": 1.8692853412237193e-05, + "loss": 3.3784, + "step": 4731 + }, + { + "epoch": 0.379387063799082, + "grad_norm": 1.8312020301818848, + "learning_rate": 1.8692211502781638e-05, + "loss": 3.2635, + "step": 4732 + }, + { + "epoch": 0.3794672385801046, + "grad_norm": 0.8815755844116211, + "learning_rate": 1.8691569446778456e-05, + "loss": 2.5139, + "step": 4733 + }, + { + "epoch": 0.37954741336112724, + "grad_norm": 2.059762954711914, + "learning_rate": 1.8690927244238482e-05, + "loss": 3.2231, + "step": 4734 + }, + { + "epoch": 0.37962758814214986, + "grad_norm": 1.939064860343933, + "learning_rate": 1.8690284895172536e-05, + "loss": 3.3068, + "step": 4735 + }, + { + "epoch": 0.37970776292317254, + "grad_norm": 1.8276172876358032, + "learning_rate": 1.8689642399591448e-05, + "loss": 3.1783, + "step": 4736 + }, + { + "epoch": 0.37978793770419517, + "grad_norm": 1.909903883934021, + "learning_rate": 1.868899975750605e-05, + "loss": 3.2247, + "step": 4737 + }, + { + "epoch": 0.3798681124852178, + "grad_norm": 1.9300698041915894, + "learning_rate": 1.868835696892718e-05, + "loss": 3.2831, + "step": 4738 + }, + { + "epoch": 0.3799482872662404, + "grad_norm": 1.9378260374069214, + "learning_rate": 1.8687714033865677e-05, + "loss": 3.3099, + "step": 4739 + }, + { + "epoch": 0.38002846204726304, + "grad_norm": 1.8779187202453613, + "learning_rate": 1.8687070952332377e-05, + "loss": 3.2108, + "step": 4740 + }, + { + "epoch": 0.38010863682828566, + "grad_norm": 1.9437865018844604, + "learning_rate": 1.8686427724338116e-05, + "loss": 3.1874, + "step": 4741 + }, + { + "epoch": 0.3801888116093083, + "grad_norm": 1.9513473510742188, + "learning_rate": 1.868578434989375e-05, + "loss": 3.3588, + "step": 4742 + }, + { + "epoch": 0.3802689863903309, + "grad_norm": 2.0016937255859375, + "learning_rate": 1.8685140829010124e-05, + "loss": 3.2762, + "step": 4743 + }, + { + "epoch": 0.38034916117135353, + "grad_norm": 1.7744927406311035, + "learning_rate": 1.8684497161698077e-05, + "loss": 3.1543, + "step": 4744 + }, + { + "epoch": 0.38042933595237616, + "grad_norm": 2.059927225112915, + "learning_rate": 1.8683853347968474e-05, + "loss": 3.3121, + "step": 4745 + }, + { + "epoch": 0.38050951073339884, + "grad_norm": 2.0027518272399902, + "learning_rate": 1.868320938783216e-05, + "loss": 3.2137, + "step": 4746 + }, + { + "epoch": 0.38058968551442146, + "grad_norm": 0.8064899444580078, + "learning_rate": 1.8682565281299998e-05, + "loss": 2.5016, + "step": 4747 + }, + { + "epoch": 0.3806698602954441, + "grad_norm": 1.8598119020462036, + "learning_rate": 1.8681921028382846e-05, + "loss": 3.2611, + "step": 4748 + }, + { + "epoch": 0.3807500350764667, + "grad_norm": 2.039538621902466, + "learning_rate": 1.8681276629091564e-05, + "loss": 3.1498, + "step": 4749 + }, + { + "epoch": 0.38083020985748933, + "grad_norm": 1.9967560768127441, + "learning_rate": 1.8680632083437017e-05, + "loss": 3.3912, + "step": 4750 + }, + { + "epoch": 0.38091038463851196, + "grad_norm": 1.9237140417099, + "learning_rate": 1.8679987391430075e-05, + "loss": 3.3519, + "step": 4751 + }, + { + "epoch": 0.3809905594195346, + "grad_norm": 0.927592933177948, + "learning_rate": 1.8679342553081605e-05, + "loss": 2.5427, + "step": 4752 + }, + { + "epoch": 0.3810707342005572, + "grad_norm": 1.8972018957138062, + "learning_rate": 1.867869756840248e-05, + "loss": 3.251, + "step": 4753 + }, + { + "epoch": 0.38115090898157983, + "grad_norm": 1.920628309249878, + "learning_rate": 1.8678052437403567e-05, + "loss": 3.2665, + "step": 4754 + }, + { + "epoch": 0.38123108376260245, + "grad_norm": 1.7856796979904175, + "learning_rate": 1.867740716009575e-05, + "loss": 3.2387, + "step": 4755 + }, + { + "epoch": 0.3813112585436251, + "grad_norm": 1.9178845882415771, + "learning_rate": 1.8676761736489907e-05, + "loss": 3.271, + "step": 4756 + }, + { + "epoch": 0.38139143332464776, + "grad_norm": 1.9105486869812012, + "learning_rate": 1.867611616659692e-05, + "loss": 3.2285, + "step": 4757 + }, + { + "epoch": 0.3814716081056704, + "grad_norm": 1.914016842842102, + "learning_rate": 1.867547045042767e-05, + "loss": 3.281, + "step": 4758 + }, + { + "epoch": 0.381551782886693, + "grad_norm": 1.832573652267456, + "learning_rate": 1.867482458799305e-05, + "loss": 3.2716, + "step": 4759 + }, + { + "epoch": 0.3816319576677156, + "grad_norm": 0.8499783873558044, + "learning_rate": 1.867417857930394e-05, + "loss": 2.5134, + "step": 4760 + }, + { + "epoch": 0.38171213244873825, + "grad_norm": 1.8447428941726685, + "learning_rate": 1.867353242437124e-05, + "loss": 3.145, + "step": 4761 + }, + { + "epoch": 0.3817923072297609, + "grad_norm": 2.0509519577026367, + "learning_rate": 1.867288612320584e-05, + "loss": 3.327, + "step": 4762 + }, + { + "epoch": 0.3818724820107835, + "grad_norm": 2.005187749862671, + "learning_rate": 1.8672239675818634e-05, + "loss": 3.2478, + "step": 4763 + }, + { + "epoch": 0.3819526567918061, + "grad_norm": 0.8204964399337769, + "learning_rate": 1.8671593082220526e-05, + "loss": 2.5198, + "step": 4764 + }, + { + "epoch": 0.38203283157282875, + "grad_norm": 2.043433427810669, + "learning_rate": 1.867094634242241e-05, + "loss": 3.2289, + "step": 4765 + }, + { + "epoch": 0.38211300635385137, + "grad_norm": 1.8134907484054565, + "learning_rate": 1.86702994564352e-05, + "loss": 3.3442, + "step": 4766 + }, + { + "epoch": 0.38219318113487405, + "grad_norm": 1.8908523321151733, + "learning_rate": 1.8669652424269795e-05, + "loss": 3.2363, + "step": 4767 + }, + { + "epoch": 0.3822733559158967, + "grad_norm": 1.8870826959609985, + "learning_rate": 1.8669005245937108e-05, + "loss": 3.2771, + "step": 4768 + }, + { + "epoch": 0.3823535306969193, + "grad_norm": 1.8519285917282104, + "learning_rate": 1.8668357921448042e-05, + "loss": 3.3143, + "step": 4769 + }, + { + "epoch": 0.3824337054779419, + "grad_norm": 1.9999799728393555, + "learning_rate": 1.866771045081352e-05, + "loss": 3.3754, + "step": 4770 + }, + { + "epoch": 0.38251388025896454, + "grad_norm": 2.0278372764587402, + "learning_rate": 1.8667062834044455e-05, + "loss": 3.2537, + "step": 4771 + }, + { + "epoch": 0.38259405503998717, + "grad_norm": 1.8524404764175415, + "learning_rate": 1.8666415071151766e-05, + "loss": 3.3132, + "step": 4772 + }, + { + "epoch": 0.3826742298210098, + "grad_norm": 1.9316328763961792, + "learning_rate": 1.866576716214637e-05, + "loss": 3.2415, + "step": 4773 + }, + { + "epoch": 0.3827544046020324, + "grad_norm": 1.944123387336731, + "learning_rate": 1.8665119107039195e-05, + "loss": 3.2494, + "step": 4774 + }, + { + "epoch": 0.38283457938305504, + "grad_norm": 1.9020634889602661, + "learning_rate": 1.866447090584117e-05, + "loss": 3.2957, + "step": 4775 + }, + { + "epoch": 0.38291475416407766, + "grad_norm": 1.8461612462997437, + "learning_rate": 1.8663822558563215e-05, + "loss": 3.2095, + "step": 4776 + }, + { + "epoch": 0.38299492894510034, + "grad_norm": 1.990213394165039, + "learning_rate": 1.866317406521626e-05, + "loss": 3.1989, + "step": 4777 + }, + { + "epoch": 0.38307510372612297, + "grad_norm": 1.8432822227478027, + "learning_rate": 1.8662525425811253e-05, + "loss": 3.2787, + "step": 4778 + }, + { + "epoch": 0.3831552785071456, + "grad_norm": 2.002772092819214, + "learning_rate": 1.8661876640359117e-05, + "loss": 3.312, + "step": 4779 + }, + { + "epoch": 0.3832354532881682, + "grad_norm": 2.016413927078247, + "learning_rate": 1.866122770887079e-05, + "loss": 3.3049, + "step": 4780 + }, + { + "epoch": 0.38331562806919084, + "grad_norm": 2.0192606449127197, + "learning_rate": 1.8660578631357222e-05, + "loss": 3.262, + "step": 4781 + }, + { + "epoch": 0.38339580285021346, + "grad_norm": 1.8769158124923706, + "learning_rate": 1.8659929407829346e-05, + "loss": 3.1626, + "step": 4782 + }, + { + "epoch": 0.3834759776312361, + "grad_norm": 1.7758147716522217, + "learning_rate": 1.8659280038298115e-05, + "loss": 3.2246, + "step": 4783 + }, + { + "epoch": 0.3835561524122587, + "grad_norm": 1.84335458278656, + "learning_rate": 1.8658630522774474e-05, + "loss": 3.2195, + "step": 4784 + }, + { + "epoch": 0.38363632719328133, + "grad_norm": 1.9102845191955566, + "learning_rate": 1.8657980861269374e-05, + "loss": 3.2906, + "step": 4785 + }, + { + "epoch": 0.38371650197430396, + "grad_norm": 1.8180668354034424, + "learning_rate": 1.8657331053793765e-05, + "loss": 3.0793, + "step": 4786 + }, + { + "epoch": 0.38379667675532664, + "grad_norm": 2.0344622135162354, + "learning_rate": 1.865668110035861e-05, + "loss": 3.3043, + "step": 4787 + }, + { + "epoch": 0.38387685153634926, + "grad_norm": 1.8776072263717651, + "learning_rate": 1.865603100097486e-05, + "loss": 3.3417, + "step": 4788 + }, + { + "epoch": 0.3839570263173719, + "grad_norm": 1.9690228700637817, + "learning_rate": 1.8655380755653477e-05, + "loss": 3.2721, + "step": 4789 + }, + { + "epoch": 0.3840372010983945, + "grad_norm": 2.1011717319488525, + "learning_rate": 1.865473036440543e-05, + "loss": 3.2001, + "step": 4790 + }, + { + "epoch": 0.38411737587941713, + "grad_norm": 1.9697359800338745, + "learning_rate": 1.8654079827241674e-05, + "loss": 3.3678, + "step": 4791 + }, + { + "epoch": 0.38419755066043976, + "grad_norm": 1.9306493997573853, + "learning_rate": 1.8653429144173186e-05, + "loss": 3.2175, + "step": 4792 + }, + { + "epoch": 0.3842777254414624, + "grad_norm": 2.0877881050109863, + "learning_rate": 1.865277831521093e-05, + "loss": 3.2335, + "step": 4793 + }, + { + "epoch": 0.384357900222485, + "grad_norm": 2.1957547664642334, + "learning_rate": 1.8652127340365883e-05, + "loss": 3.2214, + "step": 4794 + }, + { + "epoch": 0.38443807500350763, + "grad_norm": 2.1530349254608154, + "learning_rate": 1.8651476219649013e-05, + "loss": 3.3198, + "step": 4795 + }, + { + "epoch": 0.38451824978453025, + "grad_norm": 2.0258100032806396, + "learning_rate": 1.8650824953071308e-05, + "loss": 3.3588, + "step": 4796 + }, + { + "epoch": 0.38459842456555293, + "grad_norm": 1.8928371667861938, + "learning_rate": 1.8650173540643745e-05, + "loss": 3.2781, + "step": 4797 + }, + { + "epoch": 0.38467859934657556, + "grad_norm": 1.8498615026474, + "learning_rate": 1.86495219823773e-05, + "loss": 3.1755, + "step": 4798 + }, + { + "epoch": 0.3847587741275982, + "grad_norm": 1.8644450902938843, + "learning_rate": 1.8648870278282963e-05, + "loss": 3.2714, + "step": 4799 + }, + { + "epoch": 0.3848389489086208, + "grad_norm": 2.1359119415283203, + "learning_rate": 1.8648218428371722e-05, + "loss": 3.2152, + "step": 4800 + }, + { + "epoch": 0.3849191236896434, + "grad_norm": 1.8349733352661133, + "learning_rate": 1.8647566432654568e-05, + "loss": 3.2294, + "step": 4801 + }, + { + "epoch": 0.38499929847066605, + "grad_norm": 1.8445771932601929, + "learning_rate": 1.864691429114249e-05, + "loss": 3.116, + "step": 4802 + }, + { + "epoch": 0.3850794732516887, + "grad_norm": 1.9151921272277832, + "learning_rate": 1.8646262003846486e-05, + "loss": 3.176, + "step": 4803 + }, + { + "epoch": 0.3851596480327113, + "grad_norm": 1.9289659261703491, + "learning_rate": 1.864560957077755e-05, + "loss": 3.1885, + "step": 4804 + }, + { + "epoch": 0.3852398228137339, + "grad_norm": 2.0240468978881836, + "learning_rate": 1.8644956991946686e-05, + "loss": 3.2806, + "step": 4805 + }, + { + "epoch": 0.38531999759475655, + "grad_norm": 1.9600903987884521, + "learning_rate": 1.864430426736489e-05, + "loss": 3.2122, + "step": 4806 + }, + { + "epoch": 0.3854001723757792, + "grad_norm": 1.9320369958877563, + "learning_rate": 1.8643651397043175e-05, + "loss": 3.2421, + "step": 4807 + }, + { + "epoch": 0.38548034715680185, + "grad_norm": 2.042457342147827, + "learning_rate": 1.8642998380992542e-05, + "loss": 3.2638, + "step": 4808 + }, + { + "epoch": 0.3855605219378245, + "grad_norm": 1.919456124305725, + "learning_rate": 1.8642345219224e-05, + "loss": 3.2018, + "step": 4809 + }, + { + "epoch": 0.3856406967188471, + "grad_norm": 2.0946733951568604, + "learning_rate": 1.864169191174857e-05, + "loss": 3.2604, + "step": 4810 + }, + { + "epoch": 0.3857208714998697, + "grad_norm": 2.1388120651245117, + "learning_rate": 1.8641038458577255e-05, + "loss": 3.3154, + "step": 4811 + }, + { + "epoch": 0.38580104628089235, + "grad_norm": 1.996064305305481, + "learning_rate": 1.8640384859721075e-05, + "loss": 3.273, + "step": 4812 + }, + { + "epoch": 0.38588122106191497, + "grad_norm": 1.9561363458633423, + "learning_rate": 1.8639731115191055e-05, + "loss": 3.214, + "step": 4813 + }, + { + "epoch": 0.3859613958429376, + "grad_norm": 1.9327912330627441, + "learning_rate": 1.8639077224998212e-05, + "loss": 3.2782, + "step": 4814 + }, + { + "epoch": 0.3860415706239602, + "grad_norm": 1.9377431869506836, + "learning_rate": 1.8638423189153574e-05, + "loss": 3.2195, + "step": 4815 + }, + { + "epoch": 0.38612174540498284, + "grad_norm": 1.8375537395477295, + "learning_rate": 1.8637769007668164e-05, + "loss": 3.2504, + "step": 4816 + }, + { + "epoch": 0.38620192018600547, + "grad_norm": 1.7322063446044922, + "learning_rate": 1.8637114680553012e-05, + "loss": 3.0626, + "step": 4817 + }, + { + "epoch": 0.38628209496702814, + "grad_norm": 1.8963444232940674, + "learning_rate": 1.863646020781915e-05, + "loss": 3.2379, + "step": 4818 + }, + { + "epoch": 0.38636226974805077, + "grad_norm": 1.9701311588287354, + "learning_rate": 1.8635805589477613e-05, + "loss": 3.128, + "step": 4819 + }, + { + "epoch": 0.3864424445290734, + "grad_norm": 1.8227274417877197, + "learning_rate": 1.863515082553944e-05, + "loss": 3.2128, + "step": 4820 + }, + { + "epoch": 0.386522619310096, + "grad_norm": 1.8830486536026, + "learning_rate": 1.8634495916015665e-05, + "loss": 3.1146, + "step": 4821 + }, + { + "epoch": 0.38660279409111864, + "grad_norm": 2.0744266510009766, + "learning_rate": 1.8633840860917332e-05, + "loss": 3.3104, + "step": 4822 + }, + { + "epoch": 0.38668296887214126, + "grad_norm": 1.998174786567688, + "learning_rate": 1.8633185660255483e-05, + "loss": 3.3218, + "step": 4823 + }, + { + "epoch": 0.3867631436531639, + "grad_norm": 1.8820037841796875, + "learning_rate": 1.8632530314041172e-05, + "loss": 3.2919, + "step": 4824 + }, + { + "epoch": 0.3868433184341865, + "grad_norm": 1.9559087753295898, + "learning_rate": 1.863187482228544e-05, + "loss": 3.335, + "step": 4825 + }, + { + "epoch": 0.38692349321520914, + "grad_norm": 2.0497660636901855, + "learning_rate": 1.8631219184999335e-05, + "loss": 3.2115, + "step": 4826 + }, + { + "epoch": 0.38700366799623176, + "grad_norm": 2.1543686389923096, + "learning_rate": 1.8630563402193923e-05, + "loss": 3.3077, + "step": 4827 + }, + { + "epoch": 0.38708384277725444, + "grad_norm": 2.0858399868011475, + "learning_rate": 1.8629907473880255e-05, + "loss": 3.2777, + "step": 4828 + }, + { + "epoch": 0.38716401755827706, + "grad_norm": 2.0435028076171875, + "learning_rate": 1.8629251400069383e-05, + "loss": 3.2665, + "step": 4829 + }, + { + "epoch": 0.3872441923392997, + "grad_norm": 1.9040069580078125, + "learning_rate": 1.8628595180772375e-05, + "loss": 3.2819, + "step": 4830 + }, + { + "epoch": 0.3873243671203223, + "grad_norm": 1.9522682428359985, + "learning_rate": 1.8627938816000295e-05, + "loss": 3.3544, + "step": 4831 + }, + { + "epoch": 0.38740454190134493, + "grad_norm": 2.305818796157837, + "learning_rate": 1.8627282305764204e-05, + "loss": 3.3086, + "step": 4832 + }, + { + "epoch": 0.38748471668236756, + "grad_norm": 2.00286602973938, + "learning_rate": 1.862662565007518e-05, + "loss": 3.2748, + "step": 4833 + }, + { + "epoch": 0.3875648914633902, + "grad_norm": 1.8896442651748657, + "learning_rate": 1.8625968848944284e-05, + "loss": 3.2383, + "step": 4834 + }, + { + "epoch": 0.3876450662444128, + "grad_norm": 1.963668704032898, + "learning_rate": 1.8625311902382592e-05, + "loss": 3.3693, + "step": 4835 + }, + { + "epoch": 0.38772524102543543, + "grad_norm": 2.310842752456665, + "learning_rate": 1.8624654810401184e-05, + "loss": 3.3306, + "step": 4836 + }, + { + "epoch": 0.38780541580645805, + "grad_norm": 1.9340578317642212, + "learning_rate": 1.8623997573011133e-05, + "loss": 3.245, + "step": 4837 + }, + { + "epoch": 0.38788559058748073, + "grad_norm": 2.355339765548706, + "learning_rate": 1.8623340190223523e-05, + "loss": 3.2438, + "step": 4838 + }, + { + "epoch": 0.38796576536850336, + "grad_norm": 1.9141252040863037, + "learning_rate": 1.8622682662049437e-05, + "loss": 3.1623, + "step": 4839 + }, + { + "epoch": 0.388045940149526, + "grad_norm": 2.0000388622283936, + "learning_rate": 1.862202498849996e-05, + "loss": 3.2232, + "step": 4840 + }, + { + "epoch": 0.3881261149305486, + "grad_norm": 2.0033633708953857, + "learning_rate": 1.862136716958618e-05, + "loss": 3.3022, + "step": 4841 + }, + { + "epoch": 0.38820628971157123, + "grad_norm": 1.9986017942428589, + "learning_rate": 1.8620709205319188e-05, + "loss": 3.3242, + "step": 4842 + }, + { + "epoch": 0.38828646449259385, + "grad_norm": 2.042996883392334, + "learning_rate": 1.8620051095710082e-05, + "loss": 3.292, + "step": 4843 + }, + { + "epoch": 0.3883666392736165, + "grad_norm": 1.9832892417907715, + "learning_rate": 1.8619392840769946e-05, + "loss": 3.2182, + "step": 4844 + }, + { + "epoch": 0.3884468140546391, + "grad_norm": 1.9349541664123535, + "learning_rate": 1.8618734440509887e-05, + "loss": 3.2629, + "step": 4845 + }, + { + "epoch": 0.3885269888356617, + "grad_norm": 2.1270930767059326, + "learning_rate": 1.8618075894941004e-05, + "loss": 3.3181, + "step": 4846 + }, + { + "epoch": 0.38860716361668435, + "grad_norm": 1.7912122011184692, + "learning_rate": 1.8617417204074397e-05, + "loss": 3.1509, + "step": 4847 + }, + { + "epoch": 0.388687338397707, + "grad_norm": 1.8711289167404175, + "learning_rate": 1.8616758367921174e-05, + "loss": 3.2087, + "step": 4848 + }, + { + "epoch": 0.38876751317872965, + "grad_norm": 1.8916372060775757, + "learning_rate": 1.861609938649244e-05, + "loss": 3.2429, + "step": 4849 + }, + { + "epoch": 0.3888476879597523, + "grad_norm": 0.9217645525932312, + "learning_rate": 1.861544025979931e-05, + "loss": 2.5111, + "step": 4850 + }, + { + "epoch": 0.3889278627407749, + "grad_norm": 2.027437925338745, + "learning_rate": 1.8614780987852888e-05, + "loss": 3.2361, + "step": 4851 + }, + { + "epoch": 0.3890080375217975, + "grad_norm": 1.8519108295440674, + "learning_rate": 1.86141215706643e-05, + "loss": 3.2156, + "step": 4852 + }, + { + "epoch": 0.38908821230282015, + "grad_norm": 1.9153647422790527, + "learning_rate": 1.861346200824466e-05, + "loss": 3.2425, + "step": 4853 + }, + { + "epoch": 0.38916838708384277, + "grad_norm": 1.8723647594451904, + "learning_rate": 1.861280230060508e-05, + "loss": 3.2231, + "step": 4854 + }, + { + "epoch": 0.3892485618648654, + "grad_norm": 0.8171625733375549, + "learning_rate": 1.8612142447756692e-05, + "loss": 2.5094, + "step": 4855 + }, + { + "epoch": 0.389328736645888, + "grad_norm": 1.8117281198501587, + "learning_rate": 1.8611482449710616e-05, + "loss": 3.098, + "step": 4856 + }, + { + "epoch": 0.38940891142691064, + "grad_norm": 1.9022315740585327, + "learning_rate": 1.8610822306477982e-05, + "loss": 3.3177, + "step": 4857 + }, + { + "epoch": 0.3894890862079333, + "grad_norm": 1.8164869546890259, + "learning_rate": 1.861016201806992e-05, + "loss": 3.2925, + "step": 4858 + }, + { + "epoch": 0.38956926098895595, + "grad_norm": 1.89582359790802, + "learning_rate": 1.860950158449756e-05, + "loss": 3.345, + "step": 4859 + }, + { + "epoch": 0.38964943576997857, + "grad_norm": 1.9953771829605103, + "learning_rate": 1.8608841005772038e-05, + "loss": 3.2133, + "step": 4860 + }, + { + "epoch": 0.3897296105510012, + "grad_norm": 1.7791990041732788, + "learning_rate": 1.860818028190449e-05, + "loss": 3.213, + "step": 4861 + }, + { + "epoch": 0.3898097853320238, + "grad_norm": 1.983325719833374, + "learning_rate": 1.860751941290605e-05, + "loss": 3.2134, + "step": 4862 + }, + { + "epoch": 0.38988996011304644, + "grad_norm": 1.8915168046951294, + "learning_rate": 1.8606858398787877e-05, + "loss": 3.2046, + "step": 4863 + }, + { + "epoch": 0.38997013489406906, + "grad_norm": 1.9913724660873413, + "learning_rate": 1.86061972395611e-05, + "loss": 3.2977, + "step": 4864 + }, + { + "epoch": 0.3900503096750917, + "grad_norm": 1.8320496082305908, + "learning_rate": 1.860553593523687e-05, + "loss": 3.1393, + "step": 4865 + }, + { + "epoch": 0.3901304844561143, + "grad_norm": 2.0641956329345703, + "learning_rate": 1.860487448582634e-05, + "loss": 3.39, + "step": 4866 + }, + { + "epoch": 0.39021065923713694, + "grad_norm": 1.882887840270996, + "learning_rate": 1.8604212891340655e-05, + "loss": 3.2964, + "step": 4867 + }, + { + "epoch": 0.39029083401815956, + "grad_norm": 1.893850564956665, + "learning_rate": 1.8603551151790972e-05, + "loss": 3.3904, + "step": 4868 + }, + { + "epoch": 0.39037100879918224, + "grad_norm": 1.8984800577163696, + "learning_rate": 1.8602889267188453e-05, + "loss": 3.1459, + "step": 4869 + }, + { + "epoch": 0.39045118358020486, + "grad_norm": 2.0680670738220215, + "learning_rate": 1.8602227237544254e-05, + "loss": 3.3381, + "step": 4870 + }, + { + "epoch": 0.3905313583612275, + "grad_norm": 1.985320806503296, + "learning_rate": 1.860156506286953e-05, + "loss": 3.3068, + "step": 4871 + }, + { + "epoch": 0.3906115331422501, + "grad_norm": 1.9042991399765015, + "learning_rate": 1.860090274317545e-05, + "loss": 3.2513, + "step": 4872 + }, + { + "epoch": 0.39069170792327274, + "grad_norm": 2.01936936378479, + "learning_rate": 1.8600240278473186e-05, + "loss": 3.3507, + "step": 4873 + }, + { + "epoch": 0.39077188270429536, + "grad_norm": 0.8308403491973877, + "learning_rate": 1.8599577668773902e-05, + "loss": 2.4896, + "step": 4874 + }, + { + "epoch": 0.390852057485318, + "grad_norm": 1.907974123954773, + "learning_rate": 1.8598914914088765e-05, + "loss": 3.2907, + "step": 4875 + }, + { + "epoch": 0.3909322322663406, + "grad_norm": 1.9675241708755493, + "learning_rate": 1.8598252014428955e-05, + "loss": 3.1732, + "step": 4876 + }, + { + "epoch": 0.39101240704736323, + "grad_norm": 1.9537913799285889, + "learning_rate": 1.859758896980565e-05, + "loss": 3.2659, + "step": 4877 + }, + { + "epoch": 0.39109258182838585, + "grad_norm": 1.8530292510986328, + "learning_rate": 1.8596925780230023e-05, + "loss": 3.2905, + "step": 4878 + }, + { + "epoch": 0.39117275660940853, + "grad_norm": 1.9498929977416992, + "learning_rate": 1.8596262445713256e-05, + "loss": 3.1891, + "step": 4879 + }, + { + "epoch": 0.39125293139043116, + "grad_norm": 1.7778469324111938, + "learning_rate": 1.8595598966266533e-05, + "loss": 3.1704, + "step": 4880 + }, + { + "epoch": 0.3913331061714538, + "grad_norm": 1.8299330472946167, + "learning_rate": 1.859493534190104e-05, + "loss": 3.2662, + "step": 4881 + }, + { + "epoch": 0.3914132809524764, + "grad_norm": 1.9368003606796265, + "learning_rate": 1.8594271572627967e-05, + "loss": 3.2654, + "step": 4882 + }, + { + "epoch": 0.39149345573349903, + "grad_norm": 0.7903642058372498, + "learning_rate": 1.8593607658458507e-05, + "loss": 2.4951, + "step": 4883 + }, + { + "epoch": 0.39157363051452165, + "grad_norm": 1.768283486366272, + "learning_rate": 1.859294359940385e-05, + "loss": 3.2606, + "step": 4884 + }, + { + "epoch": 0.3916538052955443, + "grad_norm": 1.8669155836105347, + "learning_rate": 1.859227939547519e-05, + "loss": 3.2424, + "step": 4885 + }, + { + "epoch": 0.3917339800765669, + "grad_norm": 1.8946278095245361, + "learning_rate": 1.8591615046683727e-05, + "loss": 3.2494, + "step": 4886 + }, + { + "epoch": 0.3918141548575895, + "grad_norm": 2.022855281829834, + "learning_rate": 1.8590950553040668e-05, + "loss": 3.1592, + "step": 4887 + }, + { + "epoch": 0.39189432963861215, + "grad_norm": 1.9650520086288452, + "learning_rate": 1.8590285914557204e-05, + "loss": 3.3713, + "step": 4888 + }, + { + "epoch": 0.39197450441963483, + "grad_norm": 2.138174057006836, + "learning_rate": 1.858962113124455e-05, + "loss": 3.1724, + "step": 4889 + }, + { + "epoch": 0.39205467920065745, + "grad_norm": 1.8759820461273193, + "learning_rate": 1.8588956203113914e-05, + "loss": 3.2476, + "step": 4890 + }, + { + "epoch": 0.3921348539816801, + "grad_norm": 1.994803547859192, + "learning_rate": 1.8588291130176497e-05, + "loss": 3.3304, + "step": 4891 + }, + { + "epoch": 0.3922150287627027, + "grad_norm": 1.9165557622909546, + "learning_rate": 1.858762591244352e-05, + "loss": 3.2115, + "step": 4892 + }, + { + "epoch": 0.3922952035437253, + "grad_norm": 1.8340719938278198, + "learning_rate": 1.85869605499262e-05, + "loss": 3.2529, + "step": 4893 + }, + { + "epoch": 0.39237537832474795, + "grad_norm": 2.1570515632629395, + "learning_rate": 1.858629504263575e-05, + "loss": 3.2833, + "step": 4894 + }, + { + "epoch": 0.39245555310577057, + "grad_norm": 2.128566026687622, + "learning_rate": 1.858562939058339e-05, + "loss": 3.1643, + "step": 4895 + }, + { + "epoch": 0.3925357278867932, + "grad_norm": 1.9098381996154785, + "learning_rate": 1.8584963593780346e-05, + "loss": 3.3561, + "step": 4896 + }, + { + "epoch": 0.3926159026678158, + "grad_norm": 1.991115689277649, + "learning_rate": 1.8584297652237842e-05, + "loss": 3.3007, + "step": 4897 + }, + { + "epoch": 0.39269607744883844, + "grad_norm": 1.8580272197723389, + "learning_rate": 1.85836315659671e-05, + "loss": 3.1722, + "step": 4898 + }, + { + "epoch": 0.3927762522298611, + "grad_norm": 1.9684653282165527, + "learning_rate": 1.858296533497936e-05, + "loss": 3.2722, + "step": 4899 + }, + { + "epoch": 0.39285642701088375, + "grad_norm": 2.019402265548706, + "learning_rate": 1.858229895928585e-05, + "loss": 3.2688, + "step": 4900 + }, + { + "epoch": 0.39293660179190637, + "grad_norm": 1.9360204935073853, + "learning_rate": 1.8581632438897802e-05, + "loss": 3.1859, + "step": 4901 + }, + { + "epoch": 0.393016776572929, + "grad_norm": 1.9481818675994873, + "learning_rate": 1.8580965773826456e-05, + "loss": 3.2776, + "step": 4902 + }, + { + "epoch": 0.3930969513539516, + "grad_norm": 2.028139352798462, + "learning_rate": 1.858029896408305e-05, + "loss": 3.2106, + "step": 4903 + }, + { + "epoch": 0.39317712613497424, + "grad_norm": 1.9939963817596436, + "learning_rate": 1.857963200967883e-05, + "loss": 3.272, + "step": 4904 + }, + { + "epoch": 0.39325730091599687, + "grad_norm": 1.9338184595108032, + "learning_rate": 1.8578964910625038e-05, + "loss": 3.2582, + "step": 4905 + }, + { + "epoch": 0.3933374756970195, + "grad_norm": 0.8841027617454529, + "learning_rate": 1.857829766693292e-05, + "loss": 2.531, + "step": 4906 + }, + { + "epoch": 0.3934176504780421, + "grad_norm": 1.8649380207061768, + "learning_rate": 1.8577630278613726e-05, + "loss": 3.4274, + "step": 4907 + }, + { + "epoch": 0.39349782525906474, + "grad_norm": 1.9175139665603638, + "learning_rate": 1.857696274567871e-05, + "loss": 3.2679, + "step": 4908 + }, + { + "epoch": 0.3935780000400874, + "grad_norm": 1.8452844619750977, + "learning_rate": 1.857629506813912e-05, + "loss": 3.2454, + "step": 4909 + }, + { + "epoch": 0.39365817482111004, + "grad_norm": 1.8839675188064575, + "learning_rate": 1.8575627246006224e-05, + "loss": 3.2506, + "step": 4910 + }, + { + "epoch": 0.39373834960213266, + "grad_norm": 2.0281424522399902, + "learning_rate": 1.8574959279291275e-05, + "loss": 3.345, + "step": 4911 + }, + { + "epoch": 0.3938185243831553, + "grad_norm": 1.9288526773452759, + "learning_rate": 1.857429116800553e-05, + "loss": 3.2235, + "step": 4912 + }, + { + "epoch": 0.3938986991641779, + "grad_norm": 1.975879430770874, + "learning_rate": 1.857362291216026e-05, + "loss": 3.2681, + "step": 4913 + }, + { + "epoch": 0.39397887394520054, + "grad_norm": 1.8267561197280884, + "learning_rate": 1.8572954511766733e-05, + "loss": 3.1591, + "step": 4914 + }, + { + "epoch": 0.39405904872622316, + "grad_norm": 1.7195887565612793, + "learning_rate": 1.857228596683621e-05, + "loss": 3.1052, + "step": 4915 + }, + { + "epoch": 0.3941392235072458, + "grad_norm": 1.9958266019821167, + "learning_rate": 1.8571617277379966e-05, + "loss": 3.3928, + "step": 4916 + }, + { + "epoch": 0.3942193982882684, + "grad_norm": 1.8574450016021729, + "learning_rate": 1.857094844340928e-05, + "loss": 3.3041, + "step": 4917 + }, + { + "epoch": 0.39429957306929103, + "grad_norm": 1.9429258108139038, + "learning_rate": 1.857027946493542e-05, + "loss": 3.3119, + "step": 4918 + }, + { + "epoch": 0.3943797478503137, + "grad_norm": 1.9636784791946411, + "learning_rate": 1.8569610341969667e-05, + "loss": 3.2843, + "step": 4919 + }, + { + "epoch": 0.39445992263133633, + "grad_norm": 1.9474328756332397, + "learning_rate": 1.8568941074523307e-05, + "loss": 3.1476, + "step": 4920 + }, + { + "epoch": 0.39454009741235896, + "grad_norm": 0.8641883134841919, + "learning_rate": 1.856827166260762e-05, + "loss": 2.4988, + "step": 4921 + }, + { + "epoch": 0.3946202721933816, + "grad_norm": 1.8545223474502563, + "learning_rate": 1.8567602106233893e-05, + "loss": 3.1753, + "step": 4922 + }, + { + "epoch": 0.3947004469744042, + "grad_norm": 2.0272858142852783, + "learning_rate": 1.8566932405413415e-05, + "loss": 3.2213, + "step": 4923 + }, + { + "epoch": 0.39478062175542683, + "grad_norm": 1.826952338218689, + "learning_rate": 1.8566262560157475e-05, + "loss": 3.2488, + "step": 4924 + }, + { + "epoch": 0.39486079653644945, + "grad_norm": 1.9159266948699951, + "learning_rate": 1.856559257047737e-05, + "loss": 3.2493, + "step": 4925 + }, + { + "epoch": 0.3949409713174721, + "grad_norm": 1.9870924949645996, + "learning_rate": 1.8564922436384387e-05, + "loss": 3.4226, + "step": 4926 + }, + { + "epoch": 0.3950211460984947, + "grad_norm": 1.8584991693496704, + "learning_rate": 1.8564252157889835e-05, + "loss": 3.3011, + "step": 4927 + }, + { + "epoch": 0.3951013208795173, + "grad_norm": 2.0399961471557617, + "learning_rate": 1.8563581735005008e-05, + "loss": 3.3455, + "step": 4928 + }, + { + "epoch": 0.39518149566053995, + "grad_norm": 1.7720595598220825, + "learning_rate": 1.8562911167741212e-05, + "loss": 3.2228, + "step": 4929 + }, + { + "epoch": 0.39526167044156263, + "grad_norm": 1.9027422666549683, + "learning_rate": 1.856224045610975e-05, + "loss": 3.2055, + "step": 4930 + }, + { + "epoch": 0.39534184522258525, + "grad_norm": 2.2113258838653564, + "learning_rate": 1.8561569600121932e-05, + "loss": 3.2453, + "step": 4931 + }, + { + "epoch": 0.3954220200036079, + "grad_norm": 2.0678584575653076, + "learning_rate": 1.8560898599789066e-05, + "loss": 3.3305, + "step": 4932 + }, + { + "epoch": 0.3955021947846305, + "grad_norm": 2.012629508972168, + "learning_rate": 1.856022745512247e-05, + "loss": 3.3418, + "step": 4933 + }, + { + "epoch": 0.3955823695656531, + "grad_norm": 1.873185634613037, + "learning_rate": 1.8559556166133458e-05, + "loss": 3.1873, + "step": 4934 + }, + { + "epoch": 0.39566254434667575, + "grad_norm": 1.9401960372924805, + "learning_rate": 1.8558884732833342e-05, + "loss": 3.3295, + "step": 4935 + }, + { + "epoch": 0.3957427191276984, + "grad_norm": 2.111529588699341, + "learning_rate": 1.8558213155233448e-05, + "loss": 3.2825, + "step": 4936 + }, + { + "epoch": 0.395822893908721, + "grad_norm": 1.8300468921661377, + "learning_rate": 1.8557541433345094e-05, + "loss": 3.2542, + "step": 4937 + }, + { + "epoch": 0.3959030686897436, + "grad_norm": 1.9324584007263184, + "learning_rate": 1.855686956717961e-05, + "loss": 3.2687, + "step": 4938 + }, + { + "epoch": 0.39598324347076624, + "grad_norm": 1.9051806926727295, + "learning_rate": 1.8556197556748318e-05, + "loss": 3.3315, + "step": 4939 + }, + { + "epoch": 0.3960634182517889, + "grad_norm": 0.8706149458885193, + "learning_rate": 1.855552540206255e-05, + "loss": 2.5231, + "step": 4940 + }, + { + "epoch": 0.39614359303281155, + "grad_norm": 2.1671392917633057, + "learning_rate": 1.8554853103133643e-05, + "loss": 3.2079, + "step": 4941 + }, + { + "epoch": 0.39622376781383417, + "grad_norm": 2.0215141773223877, + "learning_rate": 1.8554180659972925e-05, + "loss": 3.3092, + "step": 4942 + }, + { + "epoch": 0.3963039425948568, + "grad_norm": 1.8730905055999756, + "learning_rate": 1.8553508072591736e-05, + "loss": 3.1036, + "step": 4943 + }, + { + "epoch": 0.3963841173758794, + "grad_norm": 1.848848581314087, + "learning_rate": 1.8552835341001416e-05, + "loss": 3.2437, + "step": 4944 + }, + { + "epoch": 0.39646429215690204, + "grad_norm": 2.012235403060913, + "learning_rate": 1.8552162465213305e-05, + "loss": 3.1902, + "step": 4945 + }, + { + "epoch": 0.39654446693792467, + "grad_norm": 2.1450750827789307, + "learning_rate": 1.8551489445238752e-05, + "loss": 3.2947, + "step": 4946 + }, + { + "epoch": 0.3966246417189473, + "grad_norm": 0.8993139266967773, + "learning_rate": 1.8550816281089102e-05, + "loss": 2.5226, + "step": 4947 + }, + { + "epoch": 0.3967048164999699, + "grad_norm": 1.8916687965393066, + "learning_rate": 1.85501429727757e-05, + "loss": 3.2267, + "step": 4948 + }, + { + "epoch": 0.39678499128099254, + "grad_norm": 1.9378933906555176, + "learning_rate": 1.85494695203099e-05, + "loss": 3.1605, + "step": 4949 + }, + { + "epoch": 0.3968651660620152, + "grad_norm": 0.8579550981521606, + "learning_rate": 1.8548795923703057e-05, + "loss": 2.5142, + "step": 4950 + }, + { + "epoch": 0.39694534084303784, + "grad_norm": 2.0111849308013916, + "learning_rate": 1.854812218296653e-05, + "loss": 3.3015, + "step": 4951 + }, + { + "epoch": 0.39702551562406047, + "grad_norm": 1.927487850189209, + "learning_rate": 1.8547448298111673e-05, + "loss": 3.293, + "step": 4952 + }, + { + "epoch": 0.3971056904050831, + "grad_norm": 2.0810985565185547, + "learning_rate": 1.8546774269149854e-05, + "loss": 3.191, + "step": 4953 + }, + { + "epoch": 0.3971858651861057, + "grad_norm": 2.0626943111419678, + "learning_rate": 1.854610009609243e-05, + "loss": 3.224, + "step": 4954 + }, + { + "epoch": 0.39726603996712834, + "grad_norm": 1.918650507926941, + "learning_rate": 1.854542577895077e-05, + "loss": 3.2792, + "step": 4955 + }, + { + "epoch": 0.39734621474815096, + "grad_norm": 1.81293785572052, + "learning_rate": 1.8544751317736246e-05, + "loss": 3.2113, + "step": 4956 + }, + { + "epoch": 0.3974263895291736, + "grad_norm": 1.860312819480896, + "learning_rate": 1.854407671246022e-05, + "loss": 3.1995, + "step": 4957 + }, + { + "epoch": 0.3975065643101962, + "grad_norm": 1.7803088426589966, + "learning_rate": 1.854340196313408e-05, + "loss": 3.299, + "step": 4958 + }, + { + "epoch": 0.39758673909121883, + "grad_norm": 1.89915132522583, + "learning_rate": 1.8542727069769187e-05, + "loss": 3.2676, + "step": 4959 + }, + { + "epoch": 0.3976669138722415, + "grad_norm": 1.8804012537002563, + "learning_rate": 1.854205203237693e-05, + "loss": 3.2358, + "step": 4960 + }, + { + "epoch": 0.39774708865326414, + "grad_norm": 1.9578505754470825, + "learning_rate": 1.8541376850968685e-05, + "loss": 3.2551, + "step": 4961 + }, + { + "epoch": 0.39782726343428676, + "grad_norm": 2.205256700515747, + "learning_rate": 1.854070152555584e-05, + "loss": 3.2606, + "step": 4962 + }, + { + "epoch": 0.3979074382153094, + "grad_norm": 1.9987461566925049, + "learning_rate": 1.8540026056149773e-05, + "loss": 3.2571, + "step": 4963 + }, + { + "epoch": 0.397987612996332, + "grad_norm": 1.9380533695220947, + "learning_rate": 1.8539350442761876e-05, + "loss": 3.256, + "step": 4964 + }, + { + "epoch": 0.39806778777735463, + "grad_norm": 1.8945567607879639, + "learning_rate": 1.8538674685403545e-05, + "loss": 3.3259, + "step": 4965 + }, + { + "epoch": 0.39814796255837726, + "grad_norm": 1.8989691734313965, + "learning_rate": 1.8537998784086164e-05, + "loss": 3.2369, + "step": 4966 + }, + { + "epoch": 0.3982281373393999, + "grad_norm": 1.8672218322753906, + "learning_rate": 1.8537322738821134e-05, + "loss": 3.2451, + "step": 4967 + }, + { + "epoch": 0.3983083121204225, + "grad_norm": 1.8167524337768555, + "learning_rate": 1.853664654961985e-05, + "loss": 3.2975, + "step": 4968 + }, + { + "epoch": 0.3983884869014451, + "grad_norm": 1.864274024963379, + "learning_rate": 1.8535970216493715e-05, + "loss": 3.2418, + "step": 4969 + }, + { + "epoch": 0.3984686616824678, + "grad_norm": 1.981017827987671, + "learning_rate": 1.853529373945413e-05, + "loss": 3.3083, + "step": 4970 + }, + { + "epoch": 0.39854883646349043, + "grad_norm": 2.0033702850341797, + "learning_rate": 1.8534617118512503e-05, + "loss": 3.2557, + "step": 4971 + }, + { + "epoch": 0.39862901124451305, + "grad_norm": 1.844351053237915, + "learning_rate": 1.8533940353680235e-05, + "loss": 3.2682, + "step": 4972 + }, + { + "epoch": 0.3987091860255357, + "grad_norm": 1.962200403213501, + "learning_rate": 1.8533263444968745e-05, + "loss": 3.2104, + "step": 4973 + }, + { + "epoch": 0.3987893608065583, + "grad_norm": 1.8381510972976685, + "learning_rate": 1.8532586392389437e-05, + "loss": 3.2607, + "step": 4974 + }, + { + "epoch": 0.3988695355875809, + "grad_norm": 1.9685626029968262, + "learning_rate": 1.853190919595373e-05, + "loss": 3.2266, + "step": 4975 + }, + { + "epoch": 0.39894971036860355, + "grad_norm": 1.8490054607391357, + "learning_rate": 1.8531231855673044e-05, + "loss": 3.1797, + "step": 4976 + }, + { + "epoch": 0.3990298851496262, + "grad_norm": 1.831096887588501, + "learning_rate": 1.8530554371558794e-05, + "loss": 3.2577, + "step": 4977 + }, + { + "epoch": 0.3991100599306488, + "grad_norm": 1.7932379245758057, + "learning_rate": 1.8529876743622403e-05, + "loss": 3.2357, + "step": 4978 + }, + { + "epoch": 0.3991902347116714, + "grad_norm": 1.998847484588623, + "learning_rate": 1.8529198971875297e-05, + "loss": 3.2961, + "step": 4979 + }, + { + "epoch": 0.3992704094926941, + "grad_norm": 2.353818416595459, + "learning_rate": 1.8528521056328902e-05, + "loss": 3.2273, + "step": 4980 + }, + { + "epoch": 0.3993505842737167, + "grad_norm": 1.8339937925338745, + "learning_rate": 1.8527842996994646e-05, + "loss": 3.2539, + "step": 4981 + }, + { + "epoch": 0.39943075905473935, + "grad_norm": 1.9911363124847412, + "learning_rate": 1.8527164793883963e-05, + "loss": 3.2773, + "step": 4982 + }, + { + "epoch": 0.39951093383576197, + "grad_norm": 1.8089065551757812, + "learning_rate": 1.8526486447008293e-05, + "loss": 3.1959, + "step": 4983 + }, + { + "epoch": 0.3995911086167846, + "grad_norm": 1.9336798191070557, + "learning_rate": 1.8525807956379057e-05, + "loss": 3.2924, + "step": 4984 + }, + { + "epoch": 0.3996712833978072, + "grad_norm": 1.9475103616714478, + "learning_rate": 1.852512932200771e-05, + "loss": 3.1689, + "step": 4985 + }, + { + "epoch": 0.39975145817882984, + "grad_norm": 1.9499201774597168, + "learning_rate": 1.8524450543905685e-05, + "loss": 3.3557, + "step": 4986 + }, + { + "epoch": 0.39983163295985247, + "grad_norm": 1.8808482885360718, + "learning_rate": 1.852377162208443e-05, + "loss": 3.3229, + "step": 4987 + }, + { + "epoch": 0.3999118077408751, + "grad_norm": 1.930956482887268, + "learning_rate": 1.8523092556555385e-05, + "loss": 3.2738, + "step": 4988 + }, + { + "epoch": 0.3999919825218977, + "grad_norm": 1.9065581560134888, + "learning_rate": 1.8522413347330007e-05, + "loss": 3.2701, + "step": 4989 + }, + { + "epoch": 0.40007215730292034, + "grad_norm": 2.0563111305236816, + "learning_rate": 1.8521733994419742e-05, + "loss": 3.2039, + "step": 4990 + }, + { + "epoch": 0.400152332083943, + "grad_norm": 1.8995200395584106, + "learning_rate": 1.8521054497836048e-05, + "loss": 3.1472, + "step": 4991 + }, + { + "epoch": 0.40023250686496564, + "grad_norm": 2.014606475830078, + "learning_rate": 1.8520374857590378e-05, + "loss": 3.3329, + "step": 4992 + }, + { + "epoch": 0.40031268164598827, + "grad_norm": 2.153189182281494, + "learning_rate": 1.851969507369419e-05, + "loss": 3.3794, + "step": 4993 + }, + { + "epoch": 0.4003928564270109, + "grad_norm": 1.8922863006591797, + "learning_rate": 1.851901514615894e-05, + "loss": 3.3027, + "step": 4994 + }, + { + "epoch": 0.4004730312080335, + "grad_norm": 1.9610660076141357, + "learning_rate": 1.8518335074996105e-05, + "loss": 3.2863, + "step": 4995 + }, + { + "epoch": 0.40055320598905614, + "grad_norm": 2.0271799564361572, + "learning_rate": 1.8517654860217142e-05, + "loss": 3.2545, + "step": 4996 + }, + { + "epoch": 0.40063338077007876, + "grad_norm": 1.8713679313659668, + "learning_rate": 1.8516974501833517e-05, + "loss": 3.2893, + "step": 4997 + }, + { + "epoch": 0.4007135555511014, + "grad_norm": 1.9367130994796753, + "learning_rate": 1.8516293999856704e-05, + "loss": 3.3008, + "step": 4998 + }, + { + "epoch": 0.400793730332124, + "grad_norm": 1.9079943895339966, + "learning_rate": 1.8515613354298173e-05, + "loss": 3.2178, + "step": 4999 + }, + { + "epoch": 0.40087390511314663, + "grad_norm": 1.9199473857879639, + "learning_rate": 1.8514932565169407e-05, + "loss": 3.1843, + "step": 5000 + }, + { + "epoch": 0.4009540798941693, + "grad_norm": 2.0114803314208984, + "learning_rate": 1.8514251632481876e-05, + "loss": 3.231, + "step": 5001 + }, + { + "epoch": 0.40103425467519194, + "grad_norm": 1.9948253631591797, + "learning_rate": 1.8513570556247065e-05, + "loss": 3.2591, + "step": 5002 + }, + { + "epoch": 0.40111442945621456, + "grad_norm": 1.9246660470962524, + "learning_rate": 1.8512889336476452e-05, + "loss": 3.2449, + "step": 5003 + }, + { + "epoch": 0.4011946042372372, + "grad_norm": 1.9220058917999268, + "learning_rate": 1.8512207973181525e-05, + "loss": 3.235, + "step": 5004 + }, + { + "epoch": 0.4012747790182598, + "grad_norm": 1.8854644298553467, + "learning_rate": 1.8511526466373774e-05, + "loss": 3.2634, + "step": 5005 + }, + { + "epoch": 0.40135495379928243, + "grad_norm": 1.8058418035507202, + "learning_rate": 1.8510844816064687e-05, + "loss": 3.2631, + "step": 5006 + }, + { + "epoch": 0.40143512858030506, + "grad_norm": 1.7869597673416138, + "learning_rate": 1.851016302226576e-05, + "loss": 3.3132, + "step": 5007 + }, + { + "epoch": 0.4015153033613277, + "grad_norm": 1.91667902469635, + "learning_rate": 1.8509481084988478e-05, + "loss": 3.2166, + "step": 5008 + }, + { + "epoch": 0.4015954781423503, + "grad_norm": 1.8796039819717407, + "learning_rate": 1.8508799004244342e-05, + "loss": 3.2461, + "step": 5009 + }, + { + "epoch": 0.40167565292337293, + "grad_norm": 1.8529490232467651, + "learning_rate": 1.8508116780044856e-05, + "loss": 3.1545, + "step": 5010 + }, + { + "epoch": 0.4017558277043956, + "grad_norm": 1.8435407876968384, + "learning_rate": 1.850743441240152e-05, + "loss": 3.2557, + "step": 5011 + }, + { + "epoch": 0.40183600248541823, + "grad_norm": 1.9758002758026123, + "learning_rate": 1.8506751901325838e-05, + "loss": 3.1406, + "step": 5012 + }, + { + "epoch": 0.40191617726644085, + "grad_norm": 2.0325684547424316, + "learning_rate": 1.8506069246829318e-05, + "loss": 3.2368, + "step": 5013 + }, + { + "epoch": 0.4019963520474635, + "grad_norm": 2.29884672164917, + "learning_rate": 1.8505386448923467e-05, + "loss": 3.2648, + "step": 5014 + }, + { + "epoch": 0.4020765268284861, + "grad_norm": 1.9420766830444336, + "learning_rate": 1.8504703507619795e-05, + "loss": 3.2772, + "step": 5015 + }, + { + "epoch": 0.4021567016095087, + "grad_norm": 2.234372615814209, + "learning_rate": 1.8504020422929824e-05, + "loss": 3.2617, + "step": 5016 + }, + { + "epoch": 0.40223687639053135, + "grad_norm": 1.950652837753296, + "learning_rate": 1.8503337194865067e-05, + "loss": 3.2107, + "step": 5017 + }, + { + "epoch": 0.402317051171554, + "grad_norm": 1.8948252201080322, + "learning_rate": 1.8502653823437038e-05, + "loss": 3.1939, + "step": 5018 + }, + { + "epoch": 0.4023972259525766, + "grad_norm": 0.8816836476325989, + "learning_rate": 1.8501970308657262e-05, + "loss": 2.5234, + "step": 5019 + }, + { + "epoch": 0.4024774007335992, + "grad_norm": 1.955133080482483, + "learning_rate": 1.8501286650537265e-05, + "loss": 3.3937, + "step": 5020 + }, + { + "epoch": 0.4025575755146219, + "grad_norm": 0.8536704778671265, + "learning_rate": 1.850060284908857e-05, + "loss": 2.5162, + "step": 5021 + }, + { + "epoch": 0.4026377502956445, + "grad_norm": 2.058500051498413, + "learning_rate": 1.8499918904322706e-05, + "loss": 3.3198, + "step": 5022 + }, + { + "epoch": 0.40271792507666715, + "grad_norm": 1.9287675619125366, + "learning_rate": 1.8499234816251208e-05, + "loss": 3.184, + "step": 5023 + }, + { + "epoch": 0.4027980998576898, + "grad_norm": 1.8527783155441284, + "learning_rate": 1.8498550584885604e-05, + "loss": 3.2522, + "step": 5024 + }, + { + "epoch": 0.4028782746387124, + "grad_norm": 1.8757412433624268, + "learning_rate": 1.849786621023743e-05, + "loss": 3.2186, + "step": 5025 + }, + { + "epoch": 0.402958449419735, + "grad_norm": 2.0670294761657715, + "learning_rate": 1.849718169231823e-05, + "loss": 3.2876, + "step": 5026 + }, + { + "epoch": 0.40303862420075764, + "grad_norm": 1.9389206171035767, + "learning_rate": 1.8496497031139543e-05, + "loss": 3.3489, + "step": 5027 + }, + { + "epoch": 0.40311879898178027, + "grad_norm": 1.9155367612838745, + "learning_rate": 1.849581222671291e-05, + "loss": 3.1924, + "step": 5028 + }, + { + "epoch": 0.4031989737628029, + "grad_norm": 2.0851335525512695, + "learning_rate": 1.8495127279049874e-05, + "loss": 3.2123, + "step": 5029 + }, + { + "epoch": 0.4032791485438255, + "grad_norm": 1.8910242319107056, + "learning_rate": 1.8494442188161988e-05, + "loss": 3.208, + "step": 5030 + }, + { + "epoch": 0.4033593233248482, + "grad_norm": 2.100604295730591, + "learning_rate": 1.84937569540608e-05, + "loss": 3.2614, + "step": 5031 + }, + { + "epoch": 0.4034394981058708, + "grad_norm": 1.8144022226333618, + "learning_rate": 1.8493071576757864e-05, + "loss": 3.2432, + "step": 5032 + }, + { + "epoch": 0.40351967288689344, + "grad_norm": 2.0997097492218018, + "learning_rate": 1.8492386056264734e-05, + "loss": 3.3583, + "step": 5033 + }, + { + "epoch": 0.40359984766791607, + "grad_norm": 2.000293254852295, + "learning_rate": 1.8491700392592965e-05, + "loss": 3.3007, + "step": 5034 + }, + { + "epoch": 0.4036800224489387, + "grad_norm": 1.9405362606048584, + "learning_rate": 1.8491014585754126e-05, + "loss": 3.1842, + "step": 5035 + }, + { + "epoch": 0.4037601972299613, + "grad_norm": 1.8389054536819458, + "learning_rate": 1.849032863575977e-05, + "loss": 3.2811, + "step": 5036 + }, + { + "epoch": 0.40384037201098394, + "grad_norm": 2.009887218475342, + "learning_rate": 1.8489642542621468e-05, + "loss": 3.3662, + "step": 5037 + }, + { + "epoch": 0.40392054679200656, + "grad_norm": 1.8980783224105835, + "learning_rate": 1.8488956306350782e-05, + "loss": 3.2203, + "step": 5038 + }, + { + "epoch": 0.4040007215730292, + "grad_norm": 2.002704620361328, + "learning_rate": 1.8488269926959283e-05, + "loss": 3.255, + "step": 5039 + }, + { + "epoch": 0.4040808963540518, + "grad_norm": 2.1087806224823, + "learning_rate": 1.848758340445855e-05, + "loss": 3.3186, + "step": 5040 + }, + { + "epoch": 0.4041610711350745, + "grad_norm": 1.8273206949234009, + "learning_rate": 1.8486896738860145e-05, + "loss": 3.3658, + "step": 5041 + }, + { + "epoch": 0.4042412459160971, + "grad_norm": 1.9512652158737183, + "learning_rate": 1.848620993017566e-05, + "loss": 3.1708, + "step": 5042 + }, + { + "epoch": 0.40432142069711974, + "grad_norm": 2.0485291481018066, + "learning_rate": 1.848552297841666e-05, + "loss": 3.3356, + "step": 5043 + }, + { + "epoch": 0.40440159547814236, + "grad_norm": 1.9977073669433594, + "learning_rate": 1.8484835883594737e-05, + "loss": 3.2689, + "step": 5044 + }, + { + "epoch": 0.404481770259165, + "grad_norm": 2.014890432357788, + "learning_rate": 1.848414864572147e-05, + "loss": 3.1979, + "step": 5045 + }, + { + "epoch": 0.4045619450401876, + "grad_norm": 1.8973881006240845, + "learning_rate": 1.8483461264808446e-05, + "loss": 3.3345, + "step": 5046 + }, + { + "epoch": 0.40464211982121023, + "grad_norm": 2.276654005050659, + "learning_rate": 1.8482773740867255e-05, + "loss": 3.2625, + "step": 5047 + }, + { + "epoch": 0.40472229460223286, + "grad_norm": 1.8185304403305054, + "learning_rate": 1.848208607390949e-05, + "loss": 3.1682, + "step": 5048 + }, + { + "epoch": 0.4048024693832555, + "grad_norm": 1.8711042404174805, + "learning_rate": 1.8481398263946742e-05, + "loss": 3.2255, + "step": 5049 + }, + { + "epoch": 0.4048826441642781, + "grad_norm": 1.9824628829956055, + "learning_rate": 1.8480710310990608e-05, + "loss": 3.2238, + "step": 5050 + }, + { + "epoch": 0.40496281894530073, + "grad_norm": 1.9780806303024292, + "learning_rate": 1.848002221505269e-05, + "loss": 3.1475, + "step": 5051 + }, + { + "epoch": 0.4050429937263234, + "grad_norm": 1.9550259113311768, + "learning_rate": 1.847933397614458e-05, + "loss": 3.2587, + "step": 5052 + }, + { + "epoch": 0.40512316850734603, + "grad_norm": 1.9552356004714966, + "learning_rate": 1.8478645594277895e-05, + "loss": 3.2188, + "step": 5053 + }, + { + "epoch": 0.40520334328836866, + "grad_norm": 0.8787267208099365, + "learning_rate": 1.847795706946423e-05, + "loss": 2.5271, + "step": 5054 + }, + { + "epoch": 0.4052835180693913, + "grad_norm": 2.1018102169036865, + "learning_rate": 1.8477268401715197e-05, + "loss": 3.4036, + "step": 5055 + }, + { + "epoch": 0.4053636928504139, + "grad_norm": 1.8481680154800415, + "learning_rate": 1.8476579591042404e-05, + "loss": 3.2826, + "step": 5056 + }, + { + "epoch": 0.4054438676314365, + "grad_norm": 1.8698005676269531, + "learning_rate": 1.847589063745747e-05, + "loss": 3.2303, + "step": 5057 + }, + { + "epoch": 0.40552404241245915, + "grad_norm": 1.8367769718170166, + "learning_rate": 1.8475201540972007e-05, + "loss": 3.2214, + "step": 5058 + }, + { + "epoch": 0.4056042171934818, + "grad_norm": 2.137726068496704, + "learning_rate": 1.8474512301597628e-05, + "loss": 3.2202, + "step": 5059 + }, + { + "epoch": 0.4056843919745044, + "grad_norm": 1.8006271123886108, + "learning_rate": 1.8473822919345962e-05, + "loss": 3.1992, + "step": 5060 + }, + { + "epoch": 0.405764566755527, + "grad_norm": 1.7695406675338745, + "learning_rate": 1.847313339422863e-05, + "loss": 3.2046, + "step": 5061 + }, + { + "epoch": 0.4058447415365497, + "grad_norm": 2.1193642616271973, + "learning_rate": 1.8472443726257254e-05, + "loss": 3.2632, + "step": 5062 + }, + { + "epoch": 0.4059249163175723, + "grad_norm": 2.0505735874176025, + "learning_rate": 1.847175391544346e-05, + "loss": 3.3392, + "step": 5063 + }, + { + "epoch": 0.40600509109859495, + "grad_norm": 1.9115116596221924, + "learning_rate": 1.8471063961798883e-05, + "loss": 3.183, + "step": 5064 + }, + { + "epoch": 0.4060852658796176, + "grad_norm": 1.8409583568572998, + "learning_rate": 1.8470373865335155e-05, + "loss": 3.193, + "step": 5065 + }, + { + "epoch": 0.4061654406606402, + "grad_norm": 1.8248074054718018, + "learning_rate": 1.8469683626063905e-05, + "loss": 3.2491, + "step": 5066 + }, + { + "epoch": 0.4062456154416628, + "grad_norm": 1.9196770191192627, + "learning_rate": 1.8468993243996775e-05, + "loss": 3.3026, + "step": 5067 + }, + { + "epoch": 0.40632579022268545, + "grad_norm": 2.010618209838867, + "learning_rate": 1.8468302719145405e-05, + "loss": 3.2542, + "step": 5068 + }, + { + "epoch": 0.40640596500370807, + "grad_norm": 1.89511239528656, + "learning_rate": 1.846761205152144e-05, + "loss": 3.2267, + "step": 5069 + }, + { + "epoch": 0.4064861397847307, + "grad_norm": 1.9787671566009521, + "learning_rate": 1.846692124113651e-05, + "loss": 3.3103, + "step": 5070 + }, + { + "epoch": 0.4065663145657533, + "grad_norm": 1.8627393245697021, + "learning_rate": 1.846623028800228e-05, + "loss": 3.2133, + "step": 5071 + }, + { + "epoch": 0.406646489346776, + "grad_norm": 1.8366461992263794, + "learning_rate": 1.8465539192130388e-05, + "loss": 3.1869, + "step": 5072 + }, + { + "epoch": 0.4067266641277986, + "grad_norm": 0.8225895166397095, + "learning_rate": 1.846484795353249e-05, + "loss": 2.5267, + "step": 5073 + }, + { + "epoch": 0.40680683890882124, + "grad_norm": 1.9060359001159668, + "learning_rate": 1.8464156572220234e-05, + "loss": 3.1961, + "step": 5074 + }, + { + "epoch": 0.40688701368984387, + "grad_norm": 1.8562297821044922, + "learning_rate": 1.8463465048205287e-05, + "loss": 3.2524, + "step": 5075 + }, + { + "epoch": 0.4069671884708665, + "grad_norm": 1.9554986953735352, + "learning_rate": 1.84627733814993e-05, + "loss": 3.4194, + "step": 5076 + }, + { + "epoch": 0.4070473632518891, + "grad_norm": 1.8684828281402588, + "learning_rate": 1.8462081572113936e-05, + "loss": 3.2479, + "step": 5077 + }, + { + "epoch": 0.40712753803291174, + "grad_norm": 2.0850656032562256, + "learning_rate": 1.846138962006086e-05, + "loss": 3.2566, + "step": 5078 + }, + { + "epoch": 0.40720771281393436, + "grad_norm": 1.893687129020691, + "learning_rate": 1.8460697525351738e-05, + "loss": 3.1625, + "step": 5079 + }, + { + "epoch": 0.407287887594957, + "grad_norm": 1.9412599802017212, + "learning_rate": 1.8460005287998232e-05, + "loss": 3.2194, + "step": 5080 + }, + { + "epoch": 0.4073680623759796, + "grad_norm": 2.9820940494537354, + "learning_rate": 1.8459312908012026e-05, + "loss": 3.2029, + "step": 5081 + }, + { + "epoch": 0.4074482371570023, + "grad_norm": 1.8738751411437988, + "learning_rate": 1.8458620385404783e-05, + "loss": 3.2041, + "step": 5082 + }, + { + "epoch": 0.4075284119380249, + "grad_norm": 2.061969518661499, + "learning_rate": 1.8457927720188177e-05, + "loss": 3.3469, + "step": 5083 + }, + { + "epoch": 0.40760858671904754, + "grad_norm": 0.9306269288063049, + "learning_rate": 1.8457234912373895e-05, + "loss": 2.5096, + "step": 5084 + }, + { + "epoch": 0.40768876150007016, + "grad_norm": 1.9799225330352783, + "learning_rate": 1.8456541961973615e-05, + "loss": 3.3047, + "step": 5085 + }, + { + "epoch": 0.4077689362810928, + "grad_norm": 2.217698812484741, + "learning_rate": 1.8455848868999014e-05, + "loss": 3.2258, + "step": 5086 + }, + { + "epoch": 0.4078491110621154, + "grad_norm": 1.8278111219406128, + "learning_rate": 1.845515563346178e-05, + "loss": 3.2311, + "step": 5087 + }, + { + "epoch": 0.40792928584313803, + "grad_norm": 1.9147721529006958, + "learning_rate": 1.8454462255373607e-05, + "loss": 3.3345, + "step": 5088 + }, + { + "epoch": 0.40800946062416066, + "grad_norm": 2.4056808948516846, + "learning_rate": 1.8453768734746176e-05, + "loss": 3.193, + "step": 5089 + }, + { + "epoch": 0.4080896354051833, + "grad_norm": 1.9065300226211548, + "learning_rate": 1.8453075071591184e-05, + "loss": 3.2467, + "step": 5090 + }, + { + "epoch": 0.4081698101862059, + "grad_norm": 0.8998948335647583, + "learning_rate": 1.8452381265920327e-05, + "loss": 2.5094, + "step": 5091 + }, + { + "epoch": 0.4082499849672286, + "grad_norm": 1.9769200086593628, + "learning_rate": 1.84516873177453e-05, + "loss": 3.3043, + "step": 5092 + }, + { + "epoch": 0.4083301597482512, + "grad_norm": 1.8483531475067139, + "learning_rate": 1.8450993227077804e-05, + "loss": 3.312, + "step": 5093 + }, + { + "epoch": 0.40841033452927383, + "grad_norm": 2.0042521953582764, + "learning_rate": 1.8450298993929538e-05, + "loss": 3.3654, + "step": 5094 + }, + { + "epoch": 0.40849050931029646, + "grad_norm": 1.7857332229614258, + "learning_rate": 1.8449604618312213e-05, + "loss": 3.2725, + "step": 5095 + }, + { + "epoch": 0.4085706840913191, + "grad_norm": 2.104099988937378, + "learning_rate": 1.8448910100237533e-05, + "loss": 3.235, + "step": 5096 + }, + { + "epoch": 0.4086508588723417, + "grad_norm": 1.9724769592285156, + "learning_rate": 1.8448215439717204e-05, + "loss": 3.2552, + "step": 5097 + }, + { + "epoch": 0.40873103365336433, + "grad_norm": 1.9093074798583984, + "learning_rate": 1.8447520636762938e-05, + "loss": 3.1948, + "step": 5098 + }, + { + "epoch": 0.40881120843438695, + "grad_norm": 1.8792978525161743, + "learning_rate": 1.844682569138645e-05, + "loss": 3.1646, + "step": 5099 + }, + { + "epoch": 0.4088913832154096, + "grad_norm": 1.9187347888946533, + "learning_rate": 1.8446130603599464e-05, + "loss": 3.1949, + "step": 5100 + }, + { + "epoch": 0.4089715579964322, + "grad_norm": 1.923895001411438, + "learning_rate": 1.8445435373413686e-05, + "loss": 3.3252, + "step": 5101 + }, + { + "epoch": 0.4090517327774548, + "grad_norm": 2.016376495361328, + "learning_rate": 1.8444740000840852e-05, + "loss": 3.2739, + "step": 5102 + }, + { + "epoch": 0.4091319075584775, + "grad_norm": 1.75747811794281, + "learning_rate": 1.8444044485892673e-05, + "loss": 3.2419, + "step": 5103 + }, + { + "epoch": 0.4092120823395001, + "grad_norm": 2.045462131500244, + "learning_rate": 1.844334882858088e-05, + "loss": 3.2578, + "step": 5104 + }, + { + "epoch": 0.40929225712052275, + "grad_norm": 1.8926270008087158, + "learning_rate": 1.84426530289172e-05, + "loss": 3.233, + "step": 5105 + }, + { + "epoch": 0.4093724319015454, + "grad_norm": 1.916878581047058, + "learning_rate": 1.844195708691337e-05, + "loss": 3.2321, + "step": 5106 + }, + { + "epoch": 0.409452606682568, + "grad_norm": 1.8378511667251587, + "learning_rate": 1.8441261002581115e-05, + "loss": 3.1747, + "step": 5107 + }, + { + "epoch": 0.4095327814635906, + "grad_norm": 0.8029693961143494, + "learning_rate": 1.8440564775932174e-05, + "loss": 2.5164, + "step": 5108 + }, + { + "epoch": 0.40961295624461325, + "grad_norm": 0.9228432774543762, + "learning_rate": 1.8439868406978286e-05, + "loss": 2.5131, + "step": 5109 + }, + { + "epoch": 0.40969313102563587, + "grad_norm": 1.9553686380386353, + "learning_rate": 1.8439171895731194e-05, + "loss": 3.1339, + "step": 5110 + }, + { + "epoch": 0.4097733058066585, + "grad_norm": 1.934869408607483, + "learning_rate": 1.8438475242202635e-05, + "loss": 3.264, + "step": 5111 + }, + { + "epoch": 0.4098534805876811, + "grad_norm": 1.8600444793701172, + "learning_rate": 1.8437778446404362e-05, + "loss": 3.2377, + "step": 5112 + }, + { + "epoch": 0.4099336553687038, + "grad_norm": 1.8848603963851929, + "learning_rate": 1.8437081508348112e-05, + "loss": 3.2291, + "step": 5113 + }, + { + "epoch": 0.4100138301497264, + "grad_norm": 1.9349730014801025, + "learning_rate": 1.8436384428045645e-05, + "loss": 3.3393, + "step": 5114 + }, + { + "epoch": 0.41009400493074905, + "grad_norm": 2.0190529823303223, + "learning_rate": 1.843568720550871e-05, + "loss": 3.2465, + "step": 5115 + }, + { + "epoch": 0.41017417971177167, + "grad_norm": 2.0419464111328125, + "learning_rate": 1.843498984074906e-05, + "loss": 3.2451, + "step": 5116 + }, + { + "epoch": 0.4102543544927943, + "grad_norm": 1.929263949394226, + "learning_rate": 1.8434292333778458e-05, + "loss": 3.2045, + "step": 5117 + }, + { + "epoch": 0.4103345292738169, + "grad_norm": 1.889817714691162, + "learning_rate": 1.8433594684608653e-05, + "loss": 3.3306, + "step": 5118 + }, + { + "epoch": 0.41041470405483954, + "grad_norm": 1.8992871046066284, + "learning_rate": 1.843289689325142e-05, + "loss": 3.2791, + "step": 5119 + }, + { + "epoch": 0.41049487883586216, + "grad_norm": 1.9507625102996826, + "learning_rate": 1.8432198959718515e-05, + "loss": 3.2736, + "step": 5120 + }, + { + "epoch": 0.4105750536168848, + "grad_norm": 1.9258406162261963, + "learning_rate": 1.843150088402171e-05, + "loss": 3.1572, + "step": 5121 + }, + { + "epoch": 0.4106552283979074, + "grad_norm": 2.1295740604400635, + "learning_rate": 1.8430802666172763e-05, + "loss": 3.2523, + "step": 5122 + }, + { + "epoch": 0.4107354031789301, + "grad_norm": 1.929334282875061, + "learning_rate": 1.843010430618346e-05, + "loss": 3.326, + "step": 5123 + }, + { + "epoch": 0.4108155779599527, + "grad_norm": 1.8323837518692017, + "learning_rate": 1.8429405804065572e-05, + "loss": 3.2863, + "step": 5124 + }, + { + "epoch": 0.41089575274097534, + "grad_norm": 1.9435619115829468, + "learning_rate": 1.8428707159830873e-05, + "loss": 3.2299, + "step": 5125 + }, + { + "epoch": 0.41097592752199796, + "grad_norm": 1.8320289850234985, + "learning_rate": 1.8428008373491137e-05, + "loss": 3.1014, + "step": 5126 + }, + { + "epoch": 0.4110561023030206, + "grad_norm": 1.996969223022461, + "learning_rate": 1.8427309445058155e-05, + "loss": 3.256, + "step": 5127 + }, + { + "epoch": 0.4111362770840432, + "grad_norm": 1.964091181755066, + "learning_rate": 1.84266103745437e-05, + "loss": 3.2838, + "step": 5128 + }, + { + "epoch": 0.41121645186506584, + "grad_norm": 1.8462998867034912, + "learning_rate": 1.842591116195957e-05, + "loss": 3.1805, + "step": 5129 + }, + { + "epoch": 0.41129662664608846, + "grad_norm": 1.8675367832183838, + "learning_rate": 1.8425211807317542e-05, + "loss": 3.303, + "step": 5130 + }, + { + "epoch": 0.4113768014271111, + "grad_norm": 1.966556191444397, + "learning_rate": 1.8424512310629417e-05, + "loss": 3.2987, + "step": 5131 + }, + { + "epoch": 0.4114569762081337, + "grad_norm": 1.9459102153778076, + "learning_rate": 1.8423812671906982e-05, + "loss": 3.1816, + "step": 5132 + }, + { + "epoch": 0.4115371509891564, + "grad_norm": 2.0585219860076904, + "learning_rate": 1.842311289116203e-05, + "loss": 3.317, + "step": 5133 + }, + { + "epoch": 0.411617325770179, + "grad_norm": 1.8631409406661987, + "learning_rate": 1.842241296840637e-05, + "loss": 3.2026, + "step": 5134 + }, + { + "epoch": 0.41169750055120163, + "grad_norm": 1.872414231300354, + "learning_rate": 1.842171290365179e-05, + "loss": 3.2159, + "step": 5135 + }, + { + "epoch": 0.41177767533222426, + "grad_norm": 1.910534143447876, + "learning_rate": 1.8421012696910105e-05, + "loss": 3.3076, + "step": 5136 + }, + { + "epoch": 0.4118578501132469, + "grad_norm": 1.847222089767456, + "learning_rate": 1.842031234819311e-05, + "loss": 3.2775, + "step": 5137 + }, + { + "epoch": 0.4119380248942695, + "grad_norm": 1.8195956945419312, + "learning_rate": 1.8419611857512617e-05, + "loss": 3.2943, + "step": 5138 + }, + { + "epoch": 0.41201819967529213, + "grad_norm": 1.9118677377700806, + "learning_rate": 1.8418911224880435e-05, + "loss": 3.2731, + "step": 5139 + }, + { + "epoch": 0.41209837445631475, + "grad_norm": 1.8542367219924927, + "learning_rate": 1.8418210450308376e-05, + "loss": 3.2832, + "step": 5140 + }, + { + "epoch": 0.4121785492373374, + "grad_norm": 1.8996827602386475, + "learning_rate": 1.8417509533808257e-05, + "loss": 3.2513, + "step": 5141 + }, + { + "epoch": 0.41225872401836, + "grad_norm": 1.9301120042800903, + "learning_rate": 1.8416808475391893e-05, + "loss": 3.2457, + "step": 5142 + }, + { + "epoch": 0.4123388987993827, + "grad_norm": 1.9093670845031738, + "learning_rate": 1.8416107275071104e-05, + "loss": 3.2961, + "step": 5143 + }, + { + "epoch": 0.4124190735804053, + "grad_norm": 2.0929667949676514, + "learning_rate": 1.8415405932857714e-05, + "loss": 3.2667, + "step": 5144 + }, + { + "epoch": 0.41249924836142793, + "grad_norm": 2.1198062896728516, + "learning_rate": 1.8414704448763544e-05, + "loss": 3.3374, + "step": 5145 + }, + { + "epoch": 0.41257942314245055, + "grad_norm": 1.9438467025756836, + "learning_rate": 1.8414002822800425e-05, + "loss": 3.235, + "step": 5146 + }, + { + "epoch": 0.4126595979234732, + "grad_norm": 1.7909159660339355, + "learning_rate": 1.841330105498018e-05, + "loss": 3.1702, + "step": 5147 + }, + { + "epoch": 0.4127397727044958, + "grad_norm": 2.0328316688537598, + "learning_rate": 1.841259914531465e-05, + "loss": 3.3878, + "step": 5148 + }, + { + "epoch": 0.4128199474855184, + "grad_norm": 1.967362642288208, + "learning_rate": 1.841189709381566e-05, + "loss": 3.1318, + "step": 5149 + }, + { + "epoch": 0.41290012226654105, + "grad_norm": 2.053654193878174, + "learning_rate": 1.8411194900495052e-05, + "loss": 3.2374, + "step": 5150 + }, + { + "epoch": 0.41298029704756367, + "grad_norm": 0.8781044483184814, + "learning_rate": 1.8410492565364657e-05, + "loss": 2.5145, + "step": 5151 + }, + { + "epoch": 0.4130604718285863, + "grad_norm": 2.012944221496582, + "learning_rate": 1.8409790088436325e-05, + "loss": 3.2834, + "step": 5152 + }, + { + "epoch": 0.413140646609609, + "grad_norm": 1.9076313972473145, + "learning_rate": 1.8409087469721895e-05, + "loss": 3.2638, + "step": 5153 + }, + { + "epoch": 0.4132208213906316, + "grad_norm": 1.8827487230300903, + "learning_rate": 1.8408384709233217e-05, + "loss": 3.1648, + "step": 5154 + }, + { + "epoch": 0.4133009961716542, + "grad_norm": 2.1093733310699463, + "learning_rate": 1.840768180698213e-05, + "loss": 3.2436, + "step": 5155 + }, + { + "epoch": 0.41338117095267685, + "grad_norm": 1.883270263671875, + "learning_rate": 1.8406978762980497e-05, + "loss": 3.2531, + "step": 5156 + }, + { + "epoch": 0.41346134573369947, + "grad_norm": 1.8796600103378296, + "learning_rate": 1.8406275577240165e-05, + "loss": 3.2637, + "step": 5157 + }, + { + "epoch": 0.4135415205147221, + "grad_norm": 1.9023033380508423, + "learning_rate": 1.8405572249772986e-05, + "loss": 3.2383, + "step": 5158 + }, + { + "epoch": 0.4136216952957447, + "grad_norm": 1.810422420501709, + "learning_rate": 1.8404868780590822e-05, + "loss": 3.2598, + "step": 5159 + }, + { + "epoch": 0.41370187007676734, + "grad_norm": 1.9477864503860474, + "learning_rate": 1.840416516970553e-05, + "loss": 3.281, + "step": 5160 + }, + { + "epoch": 0.41378204485778997, + "grad_norm": 1.9244041442871094, + "learning_rate": 1.8403461417128982e-05, + "loss": 3.2689, + "step": 5161 + }, + { + "epoch": 0.4138622196388126, + "grad_norm": 2.0250067710876465, + "learning_rate": 1.840275752287303e-05, + "loss": 3.3387, + "step": 5162 + }, + { + "epoch": 0.4139423944198352, + "grad_norm": 1.8527157306671143, + "learning_rate": 1.840205348694955e-05, + "loss": 3.1294, + "step": 5163 + }, + { + "epoch": 0.4140225692008579, + "grad_norm": 0.8286119699478149, + "learning_rate": 1.840134930937041e-05, + "loss": 2.5117, + "step": 5164 + }, + { + "epoch": 0.4141027439818805, + "grad_norm": 1.8202402591705322, + "learning_rate": 1.840064499014748e-05, + "loss": 3.2883, + "step": 5165 + }, + { + "epoch": 0.41418291876290314, + "grad_norm": 2.079498291015625, + "learning_rate": 1.8399940529292636e-05, + "loss": 3.2792, + "step": 5166 + }, + { + "epoch": 0.41426309354392576, + "grad_norm": 1.9608527421951294, + "learning_rate": 1.8399235926817756e-05, + "loss": 3.3609, + "step": 5167 + }, + { + "epoch": 0.4143432683249484, + "grad_norm": 1.9221917390823364, + "learning_rate": 1.839853118273472e-05, + "loss": 3.2261, + "step": 5168 + }, + { + "epoch": 0.414423443105971, + "grad_norm": 1.8512831926345825, + "learning_rate": 1.8397826297055406e-05, + "loss": 3.2528, + "step": 5169 + }, + { + "epoch": 0.41450361788699364, + "grad_norm": 1.8065991401672363, + "learning_rate": 1.83971212697917e-05, + "loss": 3.1792, + "step": 5170 + }, + { + "epoch": 0.41458379266801626, + "grad_norm": 1.8500858545303345, + "learning_rate": 1.839641610095549e-05, + "loss": 3.1545, + "step": 5171 + }, + { + "epoch": 0.4146639674490389, + "grad_norm": 1.8876029253005981, + "learning_rate": 1.8395710790558667e-05, + "loss": 3.1737, + "step": 5172 + }, + { + "epoch": 0.4147441422300615, + "grad_norm": 1.8665447235107422, + "learning_rate": 1.8395005338613116e-05, + "loss": 3.285, + "step": 5173 + }, + { + "epoch": 0.4148243170110842, + "grad_norm": 1.8916813135147095, + "learning_rate": 1.839429974513073e-05, + "loss": 3.3069, + "step": 5174 + }, + { + "epoch": 0.4149044917921068, + "grad_norm": 1.9178016185760498, + "learning_rate": 1.8393594010123412e-05, + "loss": 3.2236, + "step": 5175 + }, + { + "epoch": 0.41498466657312943, + "grad_norm": 1.855600118637085, + "learning_rate": 1.8392888133603058e-05, + "loss": 3.1932, + "step": 5176 + }, + { + "epoch": 0.41506484135415206, + "grad_norm": 1.8690580129623413, + "learning_rate": 1.8392182115581566e-05, + "loss": 3.1745, + "step": 5177 + }, + { + "epoch": 0.4151450161351747, + "grad_norm": 1.8767280578613281, + "learning_rate": 1.8391475956070845e-05, + "loss": 3.3328, + "step": 5178 + }, + { + "epoch": 0.4152251909161973, + "grad_norm": 1.975469946861267, + "learning_rate": 1.839076965508279e-05, + "loss": 3.1686, + "step": 5179 + }, + { + "epoch": 0.41530536569721993, + "grad_norm": 1.9548721313476562, + "learning_rate": 1.8390063212629323e-05, + "loss": 3.2462, + "step": 5180 + }, + { + "epoch": 0.41538554047824255, + "grad_norm": 1.7719451189041138, + "learning_rate": 1.8389356628722345e-05, + "loss": 3.2994, + "step": 5181 + }, + { + "epoch": 0.4154657152592652, + "grad_norm": 1.8604025840759277, + "learning_rate": 1.8388649903373767e-05, + "loss": 3.2362, + "step": 5182 + }, + { + "epoch": 0.4155458900402878, + "grad_norm": 1.9519809484481812, + "learning_rate": 1.838794303659551e-05, + "loss": 3.1996, + "step": 5183 + }, + { + "epoch": 0.4156260648213105, + "grad_norm": 2.165491819381714, + "learning_rate": 1.8387236028399494e-05, + "loss": 3.2629, + "step": 5184 + }, + { + "epoch": 0.4157062396023331, + "grad_norm": 1.9447261095046997, + "learning_rate": 1.838652887879763e-05, + "loss": 3.2392, + "step": 5185 + }, + { + "epoch": 0.41578641438335573, + "grad_norm": 2.0759530067443848, + "learning_rate": 1.8385821587801848e-05, + "loss": 3.2054, + "step": 5186 + }, + { + "epoch": 0.41586658916437835, + "grad_norm": 1.7983180284500122, + "learning_rate": 1.8385114155424067e-05, + "loss": 3.1559, + "step": 5187 + }, + { + "epoch": 0.415946763945401, + "grad_norm": 1.8284608125686646, + "learning_rate": 1.8384406581676217e-05, + "loss": 3.245, + "step": 5188 + }, + { + "epoch": 0.4160269387264236, + "grad_norm": 1.8813905715942383, + "learning_rate": 1.838369886657023e-05, + "loss": 3.2676, + "step": 5189 + }, + { + "epoch": 0.4161071135074462, + "grad_norm": 2.11040997505188, + "learning_rate": 1.8382991010118027e-05, + "loss": 3.1989, + "step": 5190 + }, + { + "epoch": 0.41618728828846885, + "grad_norm": 0.789310097694397, + "learning_rate": 1.8382283012331555e-05, + "loss": 2.5171, + "step": 5191 + }, + { + "epoch": 0.4162674630694915, + "grad_norm": 1.7714736461639404, + "learning_rate": 1.8381574873222744e-05, + "loss": 3.2609, + "step": 5192 + }, + { + "epoch": 0.4163476378505141, + "grad_norm": 2.1803267002105713, + "learning_rate": 1.838086659280354e-05, + "loss": 3.1575, + "step": 5193 + }, + { + "epoch": 0.4164278126315368, + "grad_norm": 1.9050383567810059, + "learning_rate": 1.8380158171085872e-05, + "loss": 3.2379, + "step": 5194 + }, + { + "epoch": 0.4165079874125594, + "grad_norm": 1.9080930948257446, + "learning_rate": 1.8379449608081693e-05, + "loss": 3.1719, + "step": 5195 + }, + { + "epoch": 0.416588162193582, + "grad_norm": 0.8150086402893066, + "learning_rate": 1.8378740903802952e-05, + "loss": 2.5161, + "step": 5196 + }, + { + "epoch": 0.41666833697460465, + "grad_norm": 1.917351245880127, + "learning_rate": 1.8378032058261586e-05, + "loss": 3.2368, + "step": 5197 + }, + { + "epoch": 0.41674851175562727, + "grad_norm": 0.7937426567077637, + "learning_rate": 1.8377323071469554e-05, + "loss": 2.5208, + "step": 5198 + }, + { + "epoch": 0.4168286865366499, + "grad_norm": 1.8694226741790771, + "learning_rate": 1.8376613943438807e-05, + "loss": 3.2103, + "step": 5199 + }, + { + "epoch": 0.4169088613176725, + "grad_norm": 1.8568072319030762, + "learning_rate": 1.83759046741813e-05, + "loss": 3.3461, + "step": 5200 + }, + { + "epoch": 0.41698903609869514, + "grad_norm": 1.8629486560821533, + "learning_rate": 1.8375195263709e-05, + "loss": 3.2777, + "step": 5201 + }, + { + "epoch": 0.41706921087971777, + "grad_norm": 1.9337276220321655, + "learning_rate": 1.837448571203385e-05, + "loss": 3.2722, + "step": 5202 + }, + { + "epoch": 0.4171493856607404, + "grad_norm": 1.866683840751648, + "learning_rate": 1.8373776019167825e-05, + "loss": 3.1765, + "step": 5203 + }, + { + "epoch": 0.41722956044176307, + "grad_norm": 1.8901655673980713, + "learning_rate": 1.837306618512289e-05, + "loss": 3.2469, + "step": 5204 + }, + { + "epoch": 0.4173097352227857, + "grad_norm": 1.9002408981323242, + "learning_rate": 1.8372356209911007e-05, + "loss": 3.3132, + "step": 5205 + }, + { + "epoch": 0.4173899100038083, + "grad_norm": 1.8966634273529053, + "learning_rate": 1.837164609354415e-05, + "loss": 3.2781, + "step": 5206 + }, + { + "epoch": 0.41747008478483094, + "grad_norm": 1.8637988567352295, + "learning_rate": 1.8370935836034292e-05, + "loss": 3.2166, + "step": 5207 + }, + { + "epoch": 0.41755025956585357, + "grad_norm": 1.8702932596206665, + "learning_rate": 1.8370225437393405e-05, + "loss": 3.3074, + "step": 5208 + }, + { + "epoch": 0.4176304343468762, + "grad_norm": 1.8845173120498657, + "learning_rate": 1.8369514897633465e-05, + "loss": 3.2167, + "step": 5209 + }, + { + "epoch": 0.4177106091278988, + "grad_norm": 1.9050781726837158, + "learning_rate": 1.8368804216766455e-05, + "loss": 3.1364, + "step": 5210 + }, + { + "epoch": 0.41779078390892144, + "grad_norm": 1.8514766693115234, + "learning_rate": 1.8368093394804356e-05, + "loss": 3.3241, + "step": 5211 + }, + { + "epoch": 0.41787095868994406, + "grad_norm": 1.868762493133545, + "learning_rate": 1.836738243175915e-05, + "loss": 3.2385, + "step": 5212 + }, + { + "epoch": 0.4179511334709667, + "grad_norm": 1.8596783876419067, + "learning_rate": 1.8366671327642826e-05, + "loss": 3.2404, + "step": 5213 + }, + { + "epoch": 0.41803130825198936, + "grad_norm": 1.8317162990570068, + "learning_rate": 1.8365960082467372e-05, + "loss": 3.2094, + "step": 5214 + }, + { + "epoch": 0.418111483033012, + "grad_norm": 1.9825457334518433, + "learning_rate": 1.8365248696244782e-05, + "loss": 3.2905, + "step": 5215 + }, + { + "epoch": 0.4181916578140346, + "grad_norm": 1.8558688163757324, + "learning_rate": 1.8364537168987045e-05, + "loss": 3.2905, + "step": 5216 + }, + { + "epoch": 0.41827183259505724, + "grad_norm": 1.9144270420074463, + "learning_rate": 1.836382550070616e-05, + "loss": 3.2667, + "step": 5217 + }, + { + "epoch": 0.41835200737607986, + "grad_norm": 1.8388786315917969, + "learning_rate": 1.8363113691414126e-05, + "loss": 3.274, + "step": 5218 + }, + { + "epoch": 0.4184321821571025, + "grad_norm": 1.9920109510421753, + "learning_rate": 1.836240174112294e-05, + "loss": 3.2348, + "step": 5219 + }, + { + "epoch": 0.4185123569381251, + "grad_norm": 1.8875536918640137, + "learning_rate": 1.8361689649844606e-05, + "loss": 3.2463, + "step": 5220 + }, + { + "epoch": 0.41859253171914773, + "grad_norm": 1.8233051300048828, + "learning_rate": 1.8360977417591136e-05, + "loss": 3.1568, + "step": 5221 + }, + { + "epoch": 0.41867270650017036, + "grad_norm": 2.0683491230010986, + "learning_rate": 1.8360265044374533e-05, + "loss": 3.2366, + "step": 5222 + }, + { + "epoch": 0.418752881281193, + "grad_norm": 1.9524331092834473, + "learning_rate": 1.8359552530206804e-05, + "loss": 3.2413, + "step": 5223 + }, + { + "epoch": 0.4188330560622156, + "grad_norm": 1.8024606704711914, + "learning_rate": 1.835883987509997e-05, + "loss": 3.2421, + "step": 5224 + }, + { + "epoch": 0.4189132308432383, + "grad_norm": 1.8181113004684448, + "learning_rate": 1.835812707906604e-05, + "loss": 3.27, + "step": 5225 + }, + { + "epoch": 0.4189934056242609, + "grad_norm": 1.86308753490448, + "learning_rate": 1.835741414211703e-05, + "loss": 3.2432, + "step": 5226 + }, + { + "epoch": 0.41907358040528353, + "grad_norm": 1.8881263732910156, + "learning_rate": 1.8356701064264965e-05, + "loss": 3.2324, + "step": 5227 + }, + { + "epoch": 0.41915375518630615, + "grad_norm": 1.786920428276062, + "learning_rate": 1.8355987845521864e-05, + "loss": 3.2195, + "step": 5228 + }, + { + "epoch": 0.4192339299673288, + "grad_norm": 1.8331888914108276, + "learning_rate": 1.8355274485899758e-05, + "loss": 3.1924, + "step": 5229 + }, + { + "epoch": 0.4193141047483514, + "grad_norm": 1.9449892044067383, + "learning_rate": 1.8354560985410666e-05, + "loss": 3.2922, + "step": 5230 + }, + { + "epoch": 0.419394279529374, + "grad_norm": 2.1201059818267822, + "learning_rate": 1.835384734406662e-05, + "loss": 3.2227, + "step": 5231 + }, + { + "epoch": 0.41947445431039665, + "grad_norm": 1.950596809387207, + "learning_rate": 1.835313356187965e-05, + "loss": 3.2922, + "step": 5232 + }, + { + "epoch": 0.4195546290914193, + "grad_norm": 1.9367845058441162, + "learning_rate": 1.8352419638861793e-05, + "loss": 3.2286, + "step": 5233 + }, + { + "epoch": 0.4196348038724419, + "grad_norm": 2.0979013442993164, + "learning_rate": 1.8351705575025084e-05, + "loss": 3.2556, + "step": 5234 + }, + { + "epoch": 0.4197149786534646, + "grad_norm": 2.23527193069458, + "learning_rate": 1.8350991370381565e-05, + "loss": 3.255, + "step": 5235 + }, + { + "epoch": 0.4197951534344872, + "grad_norm": 1.8447107076644897, + "learning_rate": 1.8350277024943274e-05, + "loss": 3.2828, + "step": 5236 + }, + { + "epoch": 0.4198753282155098, + "grad_norm": 1.9923877716064453, + "learning_rate": 1.8349562538722256e-05, + "loss": 3.1766, + "step": 5237 + }, + { + "epoch": 0.41995550299653245, + "grad_norm": 2.0829896926879883, + "learning_rate": 1.8348847911730555e-05, + "loss": 3.2912, + "step": 5238 + }, + { + "epoch": 0.42003567777755507, + "grad_norm": 2.1403310298919678, + "learning_rate": 1.834813314398022e-05, + "loss": 3.1142, + "step": 5239 + }, + { + "epoch": 0.4201158525585777, + "grad_norm": 1.9764177799224854, + "learning_rate": 1.8347418235483304e-05, + "loss": 3.2562, + "step": 5240 + }, + { + "epoch": 0.4201960273396003, + "grad_norm": 1.8905502557754517, + "learning_rate": 1.834670318625186e-05, + "loss": 3.2382, + "step": 5241 + }, + { + "epoch": 0.42027620212062294, + "grad_norm": 1.920587182044983, + "learning_rate": 1.8345987996297935e-05, + "loss": 3.2007, + "step": 5242 + }, + { + "epoch": 0.42035637690164557, + "grad_norm": 1.9042823314666748, + "learning_rate": 1.83452726656336e-05, + "loss": 3.2129, + "step": 5243 + }, + { + "epoch": 0.4204365516826682, + "grad_norm": 2.087758779525757, + "learning_rate": 1.8344557194270906e-05, + "loss": 3.2951, + "step": 5244 + }, + { + "epoch": 0.42051672646369087, + "grad_norm": 2.0047736167907715, + "learning_rate": 1.834384158222192e-05, + "loss": 3.3407, + "step": 5245 + }, + { + "epoch": 0.4205969012447135, + "grad_norm": 2.014590263366699, + "learning_rate": 1.8343125829498704e-05, + "loss": 3.3174, + "step": 5246 + }, + { + "epoch": 0.4206770760257361, + "grad_norm": 1.8999816179275513, + "learning_rate": 1.834240993611333e-05, + "loss": 3.2122, + "step": 5247 + }, + { + "epoch": 0.42075725080675874, + "grad_norm": 2.3664324283599854, + "learning_rate": 1.8341693902077863e-05, + "loss": 3.1665, + "step": 5248 + }, + { + "epoch": 0.42083742558778137, + "grad_norm": 2.026663064956665, + "learning_rate": 1.8340977727404377e-05, + "loss": 3.339, + "step": 5249 + }, + { + "epoch": 0.420917600368804, + "grad_norm": 1.8322863578796387, + "learning_rate": 1.834026141210495e-05, + "loss": 3.327, + "step": 5250 + }, + { + "epoch": 0.4209977751498266, + "grad_norm": 1.9438042640686035, + "learning_rate": 1.833954495619165e-05, + "loss": 3.1793, + "step": 5251 + }, + { + "epoch": 0.42107794993084924, + "grad_norm": 1.8931941986083984, + "learning_rate": 1.8338828359676557e-05, + "loss": 3.3446, + "step": 5252 + }, + { + "epoch": 0.42115812471187186, + "grad_norm": 1.9756920337677002, + "learning_rate": 1.8338111622571764e-05, + "loss": 3.1985, + "step": 5253 + }, + { + "epoch": 0.4212382994928945, + "grad_norm": 2.016552686691284, + "learning_rate": 1.8337394744889344e-05, + "loss": 3.2316, + "step": 5254 + }, + { + "epoch": 0.42131847427391717, + "grad_norm": 1.768164873123169, + "learning_rate": 1.833667772664139e-05, + "loss": 3.2682, + "step": 5255 + }, + { + "epoch": 0.4213986490549398, + "grad_norm": 1.8665845394134521, + "learning_rate": 1.8335960567839984e-05, + "loss": 3.3228, + "step": 5256 + }, + { + "epoch": 0.4214788238359624, + "grad_norm": 1.9512513875961304, + "learning_rate": 1.833524326849722e-05, + "loss": 3.2946, + "step": 5257 + }, + { + "epoch": 0.42155899861698504, + "grad_norm": 1.9776833057403564, + "learning_rate": 1.83345258286252e-05, + "loss": 3.3707, + "step": 5258 + }, + { + "epoch": 0.42163917339800766, + "grad_norm": 1.789343237876892, + "learning_rate": 1.8333808248236008e-05, + "loss": 3.349, + "step": 5259 + }, + { + "epoch": 0.4217193481790303, + "grad_norm": 1.9172592163085938, + "learning_rate": 1.8333090527341742e-05, + "loss": 3.2972, + "step": 5260 + }, + { + "epoch": 0.4217995229600529, + "grad_norm": 1.8951290845870972, + "learning_rate": 1.8332372665954512e-05, + "loss": 3.2129, + "step": 5261 + }, + { + "epoch": 0.42187969774107553, + "grad_norm": 1.8910993337631226, + "learning_rate": 1.833165466408641e-05, + "loss": 3.2135, + "step": 5262 + }, + { + "epoch": 0.42195987252209816, + "grad_norm": 1.8231213092803955, + "learning_rate": 1.833093652174955e-05, + "loss": 3.3139, + "step": 5263 + }, + { + "epoch": 0.4220400473031208, + "grad_norm": 1.9537895917892456, + "learning_rate": 1.8330218238956037e-05, + "loss": 3.2552, + "step": 5264 + }, + { + "epoch": 0.42212022208414346, + "grad_norm": 1.9025044441223145, + "learning_rate": 1.832949981571798e-05, + "loss": 3.3167, + "step": 5265 + }, + { + "epoch": 0.4222003968651661, + "grad_norm": 1.772322177886963, + "learning_rate": 1.832878125204749e-05, + "loss": 3.1936, + "step": 5266 + }, + { + "epoch": 0.4222805716461887, + "grad_norm": 2.0202372074127197, + "learning_rate": 1.8328062547956684e-05, + "loss": 3.3334, + "step": 5267 + }, + { + "epoch": 0.42236074642721133, + "grad_norm": 1.9181188344955444, + "learning_rate": 1.832734370345768e-05, + "loss": 3.194, + "step": 5268 + }, + { + "epoch": 0.42244092120823395, + "grad_norm": 1.893818736076355, + "learning_rate": 1.8326624718562597e-05, + "loss": 3.2837, + "step": 5269 + }, + { + "epoch": 0.4225210959892566, + "grad_norm": 1.9198509454727173, + "learning_rate": 1.832590559328355e-05, + "loss": 3.2477, + "step": 5270 + }, + { + "epoch": 0.4226012707702792, + "grad_norm": 1.9132657051086426, + "learning_rate": 1.8325186327632676e-05, + "loss": 3.2904, + "step": 5271 + }, + { + "epoch": 0.4226814455513018, + "grad_norm": 2.054457902908325, + "learning_rate": 1.832446692162209e-05, + "loss": 3.2756, + "step": 5272 + }, + { + "epoch": 0.42276162033232445, + "grad_norm": 2.0003719329833984, + "learning_rate": 1.8323747375263928e-05, + "loss": 3.3916, + "step": 5273 + }, + { + "epoch": 0.4228417951133471, + "grad_norm": 1.870774269104004, + "learning_rate": 1.8323027688570316e-05, + "loss": 3.2554, + "step": 5274 + }, + { + "epoch": 0.42292196989436975, + "grad_norm": 0.9421573877334595, + "learning_rate": 1.832230786155339e-05, + "loss": 2.5, + "step": 5275 + }, + { + "epoch": 0.4230021446753924, + "grad_norm": 1.969717025756836, + "learning_rate": 1.832158789422529e-05, + "loss": 3.3064, + "step": 5276 + }, + { + "epoch": 0.423082319456415, + "grad_norm": 1.9271284341812134, + "learning_rate": 1.8320867786598152e-05, + "loss": 3.2202, + "step": 5277 + }, + { + "epoch": 0.4231624942374376, + "grad_norm": 1.9085512161254883, + "learning_rate": 1.832014753868411e-05, + "loss": 3.1891, + "step": 5278 + }, + { + "epoch": 0.42324266901846025, + "grad_norm": 1.8107869625091553, + "learning_rate": 1.831942715049532e-05, + "loss": 3.1559, + "step": 5279 + }, + { + "epoch": 0.4233228437994829, + "grad_norm": 0.817382276058197, + "learning_rate": 1.8318706622043915e-05, + "loss": 2.5089, + "step": 5280 + }, + { + "epoch": 0.4234030185805055, + "grad_norm": 1.9336642026901245, + "learning_rate": 1.831798595334205e-05, + "loss": 3.2244, + "step": 5281 + }, + { + "epoch": 0.4234831933615281, + "grad_norm": 2.0477921962738037, + "learning_rate": 1.831726514440187e-05, + "loss": 3.2386, + "step": 5282 + }, + { + "epoch": 0.42356336814255074, + "grad_norm": 1.8491089344024658, + "learning_rate": 1.8316544195235538e-05, + "loss": 3.178, + "step": 5283 + }, + { + "epoch": 0.42364354292357337, + "grad_norm": 1.9381483793258667, + "learning_rate": 1.8315823105855196e-05, + "loss": 3.3308, + "step": 5284 + }, + { + "epoch": 0.423723717704596, + "grad_norm": 1.8267301321029663, + "learning_rate": 1.8315101876273012e-05, + "loss": 3.2125, + "step": 5285 + }, + { + "epoch": 0.42380389248561867, + "grad_norm": 1.853724718093872, + "learning_rate": 1.8314380506501136e-05, + "loss": 3.2285, + "step": 5286 + }, + { + "epoch": 0.4238840672666413, + "grad_norm": 1.8082554340362549, + "learning_rate": 1.8313658996551734e-05, + "loss": 3.2277, + "step": 5287 + }, + { + "epoch": 0.4239642420476639, + "grad_norm": 1.821722388267517, + "learning_rate": 1.8312937346436977e-05, + "loss": 3.2477, + "step": 5288 + }, + { + "epoch": 0.42404441682868654, + "grad_norm": 1.8740289211273193, + "learning_rate": 1.831221555616902e-05, + "loss": 3.1362, + "step": 5289 + }, + { + "epoch": 0.42412459160970917, + "grad_norm": 1.9157500267028809, + "learning_rate": 1.8311493625760044e-05, + "loss": 3.2371, + "step": 5290 + }, + { + "epoch": 0.4242047663907318, + "grad_norm": 1.9069604873657227, + "learning_rate": 1.8310771555222214e-05, + "loss": 3.2079, + "step": 5291 + }, + { + "epoch": 0.4242849411717544, + "grad_norm": 1.878183364868164, + "learning_rate": 1.83100493445677e-05, + "loss": 3.2535, + "step": 5292 + }, + { + "epoch": 0.42436511595277704, + "grad_norm": 1.705826759338379, + "learning_rate": 1.8309326993808685e-05, + "loss": 3.1508, + "step": 5293 + }, + { + "epoch": 0.42444529073379966, + "grad_norm": 1.823608636856079, + "learning_rate": 1.8308604502957346e-05, + "loss": 3.2659, + "step": 5294 + }, + { + "epoch": 0.4245254655148223, + "grad_norm": 1.8738819360733032, + "learning_rate": 1.8307881872025867e-05, + "loss": 3.2535, + "step": 5295 + }, + { + "epoch": 0.42460564029584497, + "grad_norm": 1.8529386520385742, + "learning_rate": 1.8307159101026424e-05, + "loss": 3.198, + "step": 5296 + }, + { + "epoch": 0.4246858150768676, + "grad_norm": 1.8280409574508667, + "learning_rate": 1.8306436189971208e-05, + "loss": 3.1856, + "step": 5297 + }, + { + "epoch": 0.4247659898578902, + "grad_norm": 1.7886123657226562, + "learning_rate": 1.8305713138872402e-05, + "loss": 3.2182, + "step": 5298 + }, + { + "epoch": 0.42484616463891284, + "grad_norm": 1.7124954462051392, + "learning_rate": 1.83049899477422e-05, + "loss": 3.2221, + "step": 5299 + }, + { + "epoch": 0.42492633941993546, + "grad_norm": 1.809082269668579, + "learning_rate": 1.8304266616592796e-05, + "loss": 3.1002, + "step": 5300 + }, + { + "epoch": 0.4250065142009581, + "grad_norm": 1.74720299243927, + "learning_rate": 1.8303543145436383e-05, + "loss": 3.2977, + "step": 5301 + }, + { + "epoch": 0.4250866889819807, + "grad_norm": 1.8191426992416382, + "learning_rate": 1.830281953428516e-05, + "loss": 3.432, + "step": 5302 + }, + { + "epoch": 0.42516686376300333, + "grad_norm": 1.89429771900177, + "learning_rate": 1.8302095783151323e-05, + "loss": 3.3025, + "step": 5303 + }, + { + "epoch": 0.42524703854402596, + "grad_norm": 1.87172532081604, + "learning_rate": 1.830137189204708e-05, + "loss": 3.2014, + "step": 5304 + }, + { + "epoch": 0.4253272133250486, + "grad_norm": 1.9829574823379517, + "learning_rate": 1.8300647860984634e-05, + "loss": 3.1924, + "step": 5305 + }, + { + "epoch": 0.42540738810607126, + "grad_norm": 0.8777483105659485, + "learning_rate": 1.8299923689976187e-05, + "loss": 2.5095, + "step": 5306 + }, + { + "epoch": 0.4254875628870939, + "grad_norm": 2.0024304389953613, + "learning_rate": 1.8299199379033955e-05, + "loss": 3.1866, + "step": 5307 + }, + { + "epoch": 0.4255677376681165, + "grad_norm": 1.8955851793289185, + "learning_rate": 1.8298474928170142e-05, + "loss": 3.1439, + "step": 5308 + }, + { + "epoch": 0.42564791244913913, + "grad_norm": 1.8544456958770752, + "learning_rate": 1.829775033739697e-05, + "loss": 3.246, + "step": 5309 + }, + { + "epoch": 0.42572808723016176, + "grad_norm": 2.051123857498169, + "learning_rate": 1.829702560672665e-05, + "loss": 3.275, + "step": 5310 + }, + { + "epoch": 0.4258082620111844, + "grad_norm": 1.8171823024749756, + "learning_rate": 1.8296300736171405e-05, + "loss": 3.1496, + "step": 5311 + }, + { + "epoch": 0.425888436792207, + "grad_norm": 1.9588192701339722, + "learning_rate": 1.829557572574345e-05, + "loss": 3.3234, + "step": 5312 + }, + { + "epoch": 0.4259686115732296, + "grad_norm": 1.918891429901123, + "learning_rate": 1.829485057545501e-05, + "loss": 3.2123, + "step": 5313 + }, + { + "epoch": 0.42604878635425225, + "grad_norm": 1.9918252229690552, + "learning_rate": 1.829412528531832e-05, + "loss": 3.2981, + "step": 5314 + }, + { + "epoch": 0.4261289611352749, + "grad_norm": 1.9585925340652466, + "learning_rate": 1.8293399855345594e-05, + "loss": 3.3351, + "step": 5315 + }, + { + "epoch": 0.42620913591629755, + "grad_norm": 0.8252387046813965, + "learning_rate": 1.8292674285549075e-05, + "loss": 2.516, + "step": 5316 + }, + { + "epoch": 0.4262893106973202, + "grad_norm": 1.7896661758422852, + "learning_rate": 1.8291948575940986e-05, + "loss": 3.1965, + "step": 5317 + }, + { + "epoch": 0.4263694854783428, + "grad_norm": 1.93975830078125, + "learning_rate": 1.8291222726533568e-05, + "loss": 3.1933, + "step": 5318 + }, + { + "epoch": 0.4264496602593654, + "grad_norm": 1.9062386751174927, + "learning_rate": 1.8290496737339056e-05, + "loss": 3.2283, + "step": 5319 + }, + { + "epoch": 0.42652983504038805, + "grad_norm": 1.7944731712341309, + "learning_rate": 1.828977060836969e-05, + "loss": 3.3114, + "step": 5320 + }, + { + "epoch": 0.4266100098214107, + "grad_norm": 1.7612096071243286, + "learning_rate": 1.8289044339637716e-05, + "loss": 3.1555, + "step": 5321 + }, + { + "epoch": 0.4266901846024333, + "grad_norm": 1.927628755569458, + "learning_rate": 1.8288317931155376e-05, + "loss": 3.187, + "step": 5322 + }, + { + "epoch": 0.4267703593834559, + "grad_norm": 2.0217418670654297, + "learning_rate": 1.828759138293491e-05, + "loss": 3.2682, + "step": 5323 + }, + { + "epoch": 0.42685053416447855, + "grad_norm": 1.955268144607544, + "learning_rate": 1.8286864694988583e-05, + "loss": 3.1931, + "step": 5324 + }, + { + "epoch": 0.42693070894550117, + "grad_norm": 1.961210012435913, + "learning_rate": 1.828613786732863e-05, + "loss": 3.2597, + "step": 5325 + }, + { + "epoch": 0.42701088372652385, + "grad_norm": 1.8335357904434204, + "learning_rate": 1.8285410899967323e-05, + "loss": 3.2621, + "step": 5326 + }, + { + "epoch": 0.4270910585075465, + "grad_norm": 2.08918833732605, + "learning_rate": 1.82846837929169e-05, + "loss": 3.3378, + "step": 5327 + }, + { + "epoch": 0.4271712332885691, + "grad_norm": 1.7779215574264526, + "learning_rate": 1.8283956546189632e-05, + "loss": 3.1046, + "step": 5328 + }, + { + "epoch": 0.4272514080695917, + "grad_norm": 1.7263600826263428, + "learning_rate": 1.8283229159797774e-05, + "loss": 3.2263, + "step": 5329 + }, + { + "epoch": 0.42733158285061434, + "grad_norm": 1.8938992023468018, + "learning_rate": 1.8282501633753594e-05, + "loss": 3.238, + "step": 5330 + }, + { + "epoch": 0.42741175763163697, + "grad_norm": 1.799637794494629, + "learning_rate": 1.8281773968069352e-05, + "loss": 3.261, + "step": 5331 + }, + { + "epoch": 0.4274919324126596, + "grad_norm": 1.90720796585083, + "learning_rate": 1.8281046162757322e-05, + "loss": 3.2238, + "step": 5332 + }, + { + "epoch": 0.4275721071936822, + "grad_norm": 1.8228970766067505, + "learning_rate": 1.828031821782977e-05, + "loss": 3.2378, + "step": 5333 + }, + { + "epoch": 0.42765228197470484, + "grad_norm": 1.9967209100723267, + "learning_rate": 1.8279590133298974e-05, + "loss": 3.2272, + "step": 5334 + }, + { + "epoch": 0.42773245675572746, + "grad_norm": 1.9975961446762085, + "learning_rate": 1.8278861909177202e-05, + "loss": 3.4062, + "step": 5335 + }, + { + "epoch": 0.4278126315367501, + "grad_norm": 1.8490842580795288, + "learning_rate": 1.827813354547674e-05, + "loss": 3.2349, + "step": 5336 + }, + { + "epoch": 0.42789280631777277, + "grad_norm": 1.851488471031189, + "learning_rate": 1.8277405042209863e-05, + "loss": 3.3003, + "step": 5337 + }, + { + "epoch": 0.4279729810987954, + "grad_norm": 1.8358516693115234, + "learning_rate": 1.8276676399388854e-05, + "loss": 3.2612, + "step": 5338 + }, + { + "epoch": 0.428053155879818, + "grad_norm": 1.9696496725082397, + "learning_rate": 1.8275947617026e-05, + "loss": 3.2964, + "step": 5339 + }, + { + "epoch": 0.42813333066084064, + "grad_norm": 2.044461250305176, + "learning_rate": 1.8275218695133584e-05, + "loss": 3.3743, + "step": 5340 + }, + { + "epoch": 0.42821350544186326, + "grad_norm": 0.8153588175773621, + "learning_rate": 1.8274489633723894e-05, + "loss": 2.5109, + "step": 5341 + }, + { + "epoch": 0.4282936802228859, + "grad_norm": 1.9987796545028687, + "learning_rate": 1.827376043280923e-05, + "loss": 3.2897, + "step": 5342 + }, + { + "epoch": 0.4283738550039085, + "grad_norm": 1.9259992837905884, + "learning_rate": 1.827303109240188e-05, + "loss": 3.2321, + "step": 5343 + }, + { + "epoch": 0.42845402978493113, + "grad_norm": 0.8124970197677612, + "learning_rate": 1.827230161251414e-05, + "loss": 2.5183, + "step": 5344 + }, + { + "epoch": 0.42853420456595376, + "grad_norm": 2.0997204780578613, + "learning_rate": 1.827157199315831e-05, + "loss": 3.2517, + "step": 5345 + }, + { + "epoch": 0.4286143793469764, + "grad_norm": 0.7888730764389038, + "learning_rate": 1.8270842234346694e-05, + "loss": 2.5101, + "step": 5346 + }, + { + "epoch": 0.42869455412799906, + "grad_norm": 1.829630970954895, + "learning_rate": 1.827011233609159e-05, + "loss": 3.3287, + "step": 5347 + }, + { + "epoch": 0.4287747289090217, + "grad_norm": 1.8006094694137573, + "learning_rate": 1.826938229840531e-05, + "loss": 3.2186, + "step": 5348 + }, + { + "epoch": 0.4288549036900443, + "grad_norm": 1.920696496963501, + "learning_rate": 1.8268652121300155e-05, + "loss": 3.1976, + "step": 5349 + }, + { + "epoch": 0.42893507847106693, + "grad_norm": 1.7828518152236938, + "learning_rate": 1.8267921804788442e-05, + "loss": 3.2542, + "step": 5350 + }, + { + "epoch": 0.42901525325208956, + "grad_norm": 1.731533169746399, + "learning_rate": 1.8267191348882482e-05, + "loss": 3.2034, + "step": 5351 + }, + { + "epoch": 0.4290954280331122, + "grad_norm": 0.8124703168869019, + "learning_rate": 1.8266460753594587e-05, + "loss": 2.5084, + "step": 5352 + }, + { + "epoch": 0.4291756028141348, + "grad_norm": 1.8752483129501343, + "learning_rate": 1.8265730018937077e-05, + "loss": 3.2899, + "step": 5353 + }, + { + "epoch": 0.42925577759515743, + "grad_norm": 0.7854740619659424, + "learning_rate": 1.8264999144922275e-05, + "loss": 2.514, + "step": 5354 + }, + { + "epoch": 0.42933595237618005, + "grad_norm": 1.9176963567733765, + "learning_rate": 1.8264268131562495e-05, + "loss": 3.3006, + "step": 5355 + }, + { + "epoch": 0.4294161271572027, + "grad_norm": 1.9122002124786377, + "learning_rate": 1.8263536978870072e-05, + "loss": 3.1685, + "step": 5356 + }, + { + "epoch": 0.42949630193822536, + "grad_norm": 0.8213628530502319, + "learning_rate": 1.8262805686857324e-05, + "loss": 2.5095, + "step": 5357 + }, + { + "epoch": 0.429576476719248, + "grad_norm": 1.8398743867874146, + "learning_rate": 1.8262074255536587e-05, + "loss": 3.2711, + "step": 5358 + }, + { + "epoch": 0.4296566515002706, + "grad_norm": 1.8826038837432861, + "learning_rate": 1.826134268492019e-05, + "loss": 3.3496, + "step": 5359 + }, + { + "epoch": 0.4297368262812932, + "grad_norm": 1.9511172771453857, + "learning_rate": 1.8260610975020466e-05, + "loss": 3.1481, + "step": 5360 + }, + { + "epoch": 0.42981700106231585, + "grad_norm": 1.8056483268737793, + "learning_rate": 1.8259879125849753e-05, + "loss": 3.3097, + "step": 5361 + }, + { + "epoch": 0.4298971758433385, + "grad_norm": 1.8507217168807983, + "learning_rate": 1.825914713742039e-05, + "loss": 3.3009, + "step": 5362 + }, + { + "epoch": 0.4299773506243611, + "grad_norm": 1.9863691329956055, + "learning_rate": 1.8258415009744713e-05, + "loss": 3.2423, + "step": 5363 + }, + { + "epoch": 0.4300575254053837, + "grad_norm": 1.8529950380325317, + "learning_rate": 1.8257682742835075e-05, + "loss": 3.2395, + "step": 5364 + }, + { + "epoch": 0.43013770018640635, + "grad_norm": 1.9482485055923462, + "learning_rate": 1.8256950336703814e-05, + "loss": 3.1993, + "step": 5365 + }, + { + "epoch": 0.43021787496742897, + "grad_norm": 1.9573606252670288, + "learning_rate": 1.825621779136328e-05, + "loss": 3.2593, + "step": 5366 + }, + { + "epoch": 0.43029804974845165, + "grad_norm": 1.7887136936187744, + "learning_rate": 1.825548510682582e-05, + "loss": 3.1998, + "step": 5367 + }, + { + "epoch": 0.4303782245294743, + "grad_norm": 1.9080986976623535, + "learning_rate": 1.8254752283103797e-05, + "loss": 3.3319, + "step": 5368 + }, + { + "epoch": 0.4304583993104969, + "grad_norm": 1.9683834314346313, + "learning_rate": 1.8254019320209554e-05, + "loss": 3.2535, + "step": 5369 + }, + { + "epoch": 0.4305385740915195, + "grad_norm": 1.9324703216552734, + "learning_rate": 1.825328621815546e-05, + "loss": 3.238, + "step": 5370 + }, + { + "epoch": 0.43061874887254215, + "grad_norm": 1.7807215452194214, + "learning_rate": 1.8252552976953867e-05, + "loss": 3.1859, + "step": 5371 + }, + { + "epoch": 0.43069892365356477, + "grad_norm": 1.9166194200515747, + "learning_rate": 1.8251819596617135e-05, + "loss": 3.2392, + "step": 5372 + }, + { + "epoch": 0.4307790984345874, + "grad_norm": 1.8270134925842285, + "learning_rate": 1.8251086077157638e-05, + "loss": 3.2516, + "step": 5373 + }, + { + "epoch": 0.43085927321561, + "grad_norm": 1.829762578010559, + "learning_rate": 1.8250352418587734e-05, + "loss": 3.29, + "step": 5374 + }, + { + "epoch": 0.43093944799663264, + "grad_norm": 1.9341176748275757, + "learning_rate": 1.8249618620919797e-05, + "loss": 3.1693, + "step": 5375 + }, + { + "epoch": 0.43101962277765526, + "grad_norm": 1.826054334640503, + "learning_rate": 1.8248884684166198e-05, + "loss": 3.2672, + "step": 5376 + }, + { + "epoch": 0.43109979755867794, + "grad_norm": 1.8207452297210693, + "learning_rate": 1.8248150608339313e-05, + "loss": 3.2348, + "step": 5377 + }, + { + "epoch": 0.43117997233970057, + "grad_norm": 1.7564383745193481, + "learning_rate": 1.8247416393451515e-05, + "loss": 3.1836, + "step": 5378 + }, + { + "epoch": 0.4312601471207232, + "grad_norm": 2.0054092407226562, + "learning_rate": 1.824668203951518e-05, + "loss": 3.1529, + "step": 5379 + }, + { + "epoch": 0.4313403219017458, + "grad_norm": 1.8737218379974365, + "learning_rate": 1.8245947546542693e-05, + "loss": 3.2744, + "step": 5380 + }, + { + "epoch": 0.43142049668276844, + "grad_norm": 1.9221858978271484, + "learning_rate": 1.824521291454644e-05, + "loss": 3.3029, + "step": 5381 + }, + { + "epoch": 0.43150067146379106, + "grad_norm": 1.8708090782165527, + "learning_rate": 1.82444781435388e-05, + "loss": 3.1844, + "step": 5382 + }, + { + "epoch": 0.4315808462448137, + "grad_norm": 2.1467251777648926, + "learning_rate": 1.8243743233532164e-05, + "loss": 3.3406, + "step": 5383 + }, + { + "epoch": 0.4316610210258363, + "grad_norm": 1.9447999000549316, + "learning_rate": 1.8243008184538924e-05, + "loss": 3.2422, + "step": 5384 + }, + { + "epoch": 0.43174119580685894, + "grad_norm": 1.7969722747802734, + "learning_rate": 1.8242272996571467e-05, + "loss": 3.2417, + "step": 5385 + }, + { + "epoch": 0.43182137058788156, + "grad_norm": 1.7528271675109863, + "learning_rate": 1.8241537669642197e-05, + "loss": 3.156, + "step": 5386 + }, + { + "epoch": 0.43190154536890424, + "grad_norm": 1.8370378017425537, + "learning_rate": 1.8240802203763505e-05, + "loss": 3.2429, + "step": 5387 + }, + { + "epoch": 0.43198172014992686, + "grad_norm": 1.8568650484085083, + "learning_rate": 1.8240066598947794e-05, + "loss": 3.3475, + "step": 5388 + }, + { + "epoch": 0.4320618949309495, + "grad_norm": 2.0923850536346436, + "learning_rate": 1.823933085520746e-05, + "loss": 3.2609, + "step": 5389 + }, + { + "epoch": 0.4321420697119721, + "grad_norm": 2.005809783935547, + "learning_rate": 1.8238594972554918e-05, + "loss": 3.2704, + "step": 5390 + }, + { + "epoch": 0.43222224449299473, + "grad_norm": 1.8827193975448608, + "learning_rate": 1.8237858951002564e-05, + "loss": 3.3765, + "step": 5391 + }, + { + "epoch": 0.43230241927401736, + "grad_norm": 1.9355063438415527, + "learning_rate": 1.8237122790562808e-05, + "loss": 3.1259, + "step": 5392 + }, + { + "epoch": 0.43238259405504, + "grad_norm": 1.806472659111023, + "learning_rate": 1.823638649124807e-05, + "loss": 3.1497, + "step": 5393 + }, + { + "epoch": 0.4324627688360626, + "grad_norm": 1.8935047388076782, + "learning_rate": 1.8235650053070757e-05, + "loss": 3.2654, + "step": 5394 + }, + { + "epoch": 0.43254294361708523, + "grad_norm": 1.897139310836792, + "learning_rate": 1.8234913476043287e-05, + "loss": 3.2667, + "step": 5395 + }, + { + "epoch": 0.43262311839810785, + "grad_norm": 1.9331250190734863, + "learning_rate": 1.8234176760178077e-05, + "loss": 3.2456, + "step": 5396 + }, + { + "epoch": 0.4327032931791305, + "grad_norm": 1.9414430856704712, + "learning_rate": 1.8233439905487552e-05, + "loss": 3.2401, + "step": 5397 + }, + { + "epoch": 0.43278346796015316, + "grad_norm": 1.8903214931488037, + "learning_rate": 1.823270291198413e-05, + "loss": 3.3833, + "step": 5398 + }, + { + "epoch": 0.4328636427411758, + "grad_norm": 1.8315471410751343, + "learning_rate": 1.8231965779680237e-05, + "loss": 3.2424, + "step": 5399 + }, + { + "epoch": 0.4329438175221984, + "grad_norm": 1.9246793985366821, + "learning_rate": 1.8231228508588305e-05, + "loss": 3.3076, + "step": 5400 + }, + { + "epoch": 0.43302399230322103, + "grad_norm": 1.9194144010543823, + "learning_rate": 1.8230491098720758e-05, + "loss": 3.1776, + "step": 5401 + }, + { + "epoch": 0.43310416708424365, + "grad_norm": 1.9053339958190918, + "learning_rate": 1.822975355009003e-05, + "loss": 3.2155, + "step": 5402 + }, + { + "epoch": 0.4331843418652663, + "grad_norm": 1.8320519924163818, + "learning_rate": 1.8229015862708565e-05, + "loss": 3.4113, + "step": 5403 + }, + { + "epoch": 0.4332645166462889, + "grad_norm": 1.8461240530014038, + "learning_rate": 1.8228278036588784e-05, + "loss": 3.2131, + "step": 5404 + }, + { + "epoch": 0.4333446914273115, + "grad_norm": 2.053156852722168, + "learning_rate": 1.8227540071743142e-05, + "loss": 3.153, + "step": 5405 + }, + { + "epoch": 0.43342486620833415, + "grad_norm": 1.8236725330352783, + "learning_rate": 1.8226801968184072e-05, + "loss": 3.2926, + "step": 5406 + }, + { + "epoch": 0.43350504098935677, + "grad_norm": 1.966921329498291, + "learning_rate": 1.8226063725924018e-05, + "loss": 3.223, + "step": 5407 + }, + { + "epoch": 0.43358521577037945, + "grad_norm": 1.9541940689086914, + "learning_rate": 1.822532534497543e-05, + "loss": 3.3499, + "step": 5408 + }, + { + "epoch": 0.4336653905514021, + "grad_norm": 1.8760372400283813, + "learning_rate": 1.8224586825350757e-05, + "loss": 3.2296, + "step": 5409 + }, + { + "epoch": 0.4337455653324247, + "grad_norm": 2.097543716430664, + "learning_rate": 1.822384816706245e-05, + "loss": 3.3571, + "step": 5410 + }, + { + "epoch": 0.4338257401134473, + "grad_norm": 1.9990965127944946, + "learning_rate": 1.822310937012296e-05, + "loss": 3.141, + "step": 5411 + }, + { + "epoch": 0.43390591489446995, + "grad_norm": 2.0643062591552734, + "learning_rate": 1.8222370434544742e-05, + "loss": 3.2554, + "step": 5412 + }, + { + "epoch": 0.43398608967549257, + "grad_norm": 0.8732900619506836, + "learning_rate": 1.8221631360340257e-05, + "loss": 2.5388, + "step": 5413 + }, + { + "epoch": 0.4340662644565152, + "grad_norm": 1.9298720359802246, + "learning_rate": 1.8220892147521967e-05, + "loss": 3.3616, + "step": 5414 + }, + { + "epoch": 0.4341464392375378, + "grad_norm": 1.8638337850570679, + "learning_rate": 1.8220152796102334e-05, + "loss": 3.2629, + "step": 5415 + }, + { + "epoch": 0.43422661401856044, + "grad_norm": 1.8851754665374756, + "learning_rate": 1.8219413306093817e-05, + "loss": 3.1624, + "step": 5416 + }, + { + "epoch": 0.43430678879958307, + "grad_norm": 0.8454503417015076, + "learning_rate": 1.821867367750889e-05, + "loss": 2.5098, + "step": 5417 + }, + { + "epoch": 0.43438696358060574, + "grad_norm": 1.8737508058547974, + "learning_rate": 1.8217933910360025e-05, + "loss": 3.2015, + "step": 5418 + }, + { + "epoch": 0.43446713836162837, + "grad_norm": 1.8788999319076538, + "learning_rate": 1.8217194004659683e-05, + "loss": 3.2502, + "step": 5419 + }, + { + "epoch": 0.434547313142651, + "grad_norm": 1.983913540840149, + "learning_rate": 1.8216453960420354e-05, + "loss": 3.2499, + "step": 5420 + }, + { + "epoch": 0.4346274879236736, + "grad_norm": 1.866562843322754, + "learning_rate": 1.8215713777654503e-05, + "loss": 3.1536, + "step": 5421 + }, + { + "epoch": 0.43470766270469624, + "grad_norm": 1.7858978509902954, + "learning_rate": 1.8214973456374613e-05, + "loss": 3.3042, + "step": 5422 + }, + { + "epoch": 0.43478783748571886, + "grad_norm": 1.9615485668182373, + "learning_rate": 1.821423299659317e-05, + "loss": 3.1827, + "step": 5423 + }, + { + "epoch": 0.4348680122667415, + "grad_norm": 1.8880449533462524, + "learning_rate": 1.8213492398322647e-05, + "loss": 3.3389, + "step": 5424 + }, + { + "epoch": 0.4349481870477641, + "grad_norm": 1.807623028755188, + "learning_rate": 1.821275166157554e-05, + "loss": 3.213, + "step": 5425 + }, + { + "epoch": 0.43502836182878674, + "grad_norm": 1.9572664499282837, + "learning_rate": 1.8212010786364333e-05, + "loss": 3.2994, + "step": 5426 + }, + { + "epoch": 0.43510853660980936, + "grad_norm": 1.8345366716384888, + "learning_rate": 1.821126977270152e-05, + "loss": 3.2183, + "step": 5427 + }, + { + "epoch": 0.43518871139083204, + "grad_norm": 1.9280768632888794, + "learning_rate": 1.8210528620599595e-05, + "loss": 3.2417, + "step": 5428 + }, + { + "epoch": 0.43526888617185466, + "grad_norm": 1.925991177558899, + "learning_rate": 1.820978733007105e-05, + "loss": 3.4181, + "step": 5429 + }, + { + "epoch": 0.4353490609528773, + "grad_norm": 1.6686562299728394, + "learning_rate": 1.8209045901128382e-05, + "loss": 3.1897, + "step": 5430 + }, + { + "epoch": 0.4354292357338999, + "grad_norm": 1.9064043760299683, + "learning_rate": 1.820830433378409e-05, + "loss": 3.2932, + "step": 5431 + }, + { + "epoch": 0.43550941051492253, + "grad_norm": 2.001760721206665, + "learning_rate": 1.820756262805069e-05, + "loss": 3.311, + "step": 5432 + }, + { + "epoch": 0.43558958529594516, + "grad_norm": 1.878960132598877, + "learning_rate": 1.8206820783940666e-05, + "loss": 3.2758, + "step": 5433 + }, + { + "epoch": 0.4356697600769678, + "grad_norm": 1.8592727184295654, + "learning_rate": 1.8206078801466544e-05, + "loss": 3.3002, + "step": 5434 + }, + { + "epoch": 0.4357499348579904, + "grad_norm": 1.9741356372833252, + "learning_rate": 1.820533668064082e-05, + "loss": 3.257, + "step": 5435 + }, + { + "epoch": 0.43583010963901303, + "grad_norm": 1.8780325651168823, + "learning_rate": 1.8204594421476014e-05, + "loss": 3.1918, + "step": 5436 + }, + { + "epoch": 0.43591028442003565, + "grad_norm": 1.8405870199203491, + "learning_rate": 1.8203852023984635e-05, + "loss": 3.3535, + "step": 5437 + }, + { + "epoch": 0.43599045920105833, + "grad_norm": 1.9159482717514038, + "learning_rate": 1.8203109488179206e-05, + "loss": 3.2081, + "step": 5438 + }, + { + "epoch": 0.43607063398208096, + "grad_norm": 0.8343738913536072, + "learning_rate": 1.820236681407224e-05, + "loss": 2.4983, + "step": 5439 + }, + { + "epoch": 0.4361508087631036, + "grad_norm": 1.9589110612869263, + "learning_rate": 1.8201624001676262e-05, + "loss": 3.2573, + "step": 5440 + }, + { + "epoch": 0.4362309835441262, + "grad_norm": 1.89934504032135, + "learning_rate": 1.820088105100379e-05, + "loss": 3.239, + "step": 5441 + }, + { + "epoch": 0.43631115832514883, + "grad_norm": 1.99652898311615, + "learning_rate": 1.8200137962067357e-05, + "loss": 3.2001, + "step": 5442 + }, + { + "epoch": 0.43639133310617145, + "grad_norm": 1.8055778741836548, + "learning_rate": 1.8199394734879487e-05, + "loss": 3.349, + "step": 5443 + }, + { + "epoch": 0.4364715078871941, + "grad_norm": 2.014122724533081, + "learning_rate": 1.8198651369452714e-05, + "loss": 3.3809, + "step": 5444 + }, + { + "epoch": 0.4365516826682167, + "grad_norm": 0.7774775624275208, + "learning_rate": 1.8197907865799565e-05, + "loss": 2.5212, + "step": 5445 + }, + { + "epoch": 0.4366318574492393, + "grad_norm": 1.9695292711257935, + "learning_rate": 1.819716422393258e-05, + "loss": 3.1878, + "step": 5446 + }, + { + "epoch": 0.43671203223026195, + "grad_norm": 1.8567306995391846, + "learning_rate": 1.8196420443864294e-05, + "loss": 3.1934, + "step": 5447 + }, + { + "epoch": 0.43679220701128463, + "grad_norm": 1.820715069770813, + "learning_rate": 1.8195676525607248e-05, + "loss": 3.1443, + "step": 5448 + }, + { + "epoch": 0.43687238179230725, + "grad_norm": 1.9028055667877197, + "learning_rate": 1.8194932469173987e-05, + "loss": 3.2804, + "step": 5449 + }, + { + "epoch": 0.4369525565733299, + "grad_norm": 1.8385789394378662, + "learning_rate": 1.819418827457705e-05, + "loss": 3.2605, + "step": 5450 + }, + { + "epoch": 0.4370327313543525, + "grad_norm": 1.938271164894104, + "learning_rate": 1.8193443941828987e-05, + "loss": 3.2828, + "step": 5451 + }, + { + "epoch": 0.4371129061353751, + "grad_norm": 1.959112286567688, + "learning_rate": 1.8192699470942346e-05, + "loss": 3.3251, + "step": 5452 + }, + { + "epoch": 0.43719308091639775, + "grad_norm": 1.83502197265625, + "learning_rate": 1.819195486192968e-05, + "loss": 3.1938, + "step": 5453 + }, + { + "epoch": 0.43727325569742037, + "grad_norm": 1.852940320968628, + "learning_rate": 1.8191210114803546e-05, + "loss": 3.2776, + "step": 5454 + }, + { + "epoch": 0.437353430478443, + "grad_norm": 1.9680075645446777, + "learning_rate": 1.8190465229576492e-05, + "loss": 3.1849, + "step": 5455 + }, + { + "epoch": 0.4374336052594656, + "grad_norm": 1.8230371475219727, + "learning_rate": 1.8189720206261082e-05, + "loss": 3.3457, + "step": 5456 + }, + { + "epoch": 0.43751378004048824, + "grad_norm": 1.7410399913787842, + "learning_rate": 1.8188975044869873e-05, + "loss": 3.2314, + "step": 5457 + }, + { + "epoch": 0.43759395482151087, + "grad_norm": 0.8114197254180908, + "learning_rate": 1.8188229745415437e-05, + "loss": 2.4962, + "step": 5458 + }, + { + "epoch": 0.43767412960253355, + "grad_norm": 1.9151922464370728, + "learning_rate": 1.8187484307910325e-05, + "loss": 3.116, + "step": 5459 + }, + { + "epoch": 0.43775430438355617, + "grad_norm": 1.7716584205627441, + "learning_rate": 1.818673873236712e-05, + "loss": 3.3814, + "step": 5460 + }, + { + "epoch": 0.4378344791645788, + "grad_norm": 1.8028603792190552, + "learning_rate": 1.818599301879838e-05, + "loss": 3.296, + "step": 5461 + }, + { + "epoch": 0.4379146539456014, + "grad_norm": 1.8538362979888916, + "learning_rate": 1.818524716721669e-05, + "loss": 3.2118, + "step": 5462 + }, + { + "epoch": 0.43799482872662404, + "grad_norm": 1.8214696645736694, + "learning_rate": 1.818450117763461e-05, + "loss": 3.3028, + "step": 5463 + }, + { + "epoch": 0.43807500350764667, + "grad_norm": 0.793192446231842, + "learning_rate": 1.818375505006473e-05, + "loss": 2.5052, + "step": 5464 + }, + { + "epoch": 0.4381551782886693, + "grad_norm": 1.8854001760482788, + "learning_rate": 1.8183008784519622e-05, + "loss": 3.2283, + "step": 5465 + }, + { + "epoch": 0.4382353530696919, + "grad_norm": 1.8997423648834229, + "learning_rate": 1.818226238101187e-05, + "loss": 3.3088, + "step": 5466 + }, + { + "epoch": 0.43831552785071454, + "grad_norm": 1.9618622064590454, + "learning_rate": 1.818151583955406e-05, + "loss": 3.1925, + "step": 5467 + }, + { + "epoch": 0.43839570263173716, + "grad_norm": 1.829598069190979, + "learning_rate": 1.818076916015877e-05, + "loss": 3.2494, + "step": 5468 + }, + { + "epoch": 0.43847587741275984, + "grad_norm": 1.7785918712615967, + "learning_rate": 1.81800223428386e-05, + "loss": 3.2894, + "step": 5469 + }, + { + "epoch": 0.43855605219378246, + "grad_norm": 1.9132080078125, + "learning_rate": 1.8179275387606138e-05, + "loss": 3.1898, + "step": 5470 + }, + { + "epoch": 0.4386362269748051, + "grad_norm": 1.9247549772262573, + "learning_rate": 1.817852829447397e-05, + "loss": 3.2454, + "step": 5471 + }, + { + "epoch": 0.4387164017558277, + "grad_norm": 1.8489116430282593, + "learning_rate": 1.8177781063454703e-05, + "loss": 3.2634, + "step": 5472 + }, + { + "epoch": 0.43879657653685034, + "grad_norm": 1.901944875717163, + "learning_rate": 1.8177033694560927e-05, + "loss": 3.1812, + "step": 5473 + }, + { + "epoch": 0.43887675131787296, + "grad_norm": 1.9272984266281128, + "learning_rate": 1.8176286187805242e-05, + "loss": 3.1691, + "step": 5474 + }, + { + "epoch": 0.4389569260988956, + "grad_norm": 1.9183725118637085, + "learning_rate": 1.8175538543200257e-05, + "loss": 3.1146, + "step": 5475 + }, + { + "epoch": 0.4390371008799182, + "grad_norm": 1.8704506158828735, + "learning_rate": 1.8174790760758573e-05, + "loss": 3.1569, + "step": 5476 + }, + { + "epoch": 0.43911727566094083, + "grad_norm": 1.8411444425582886, + "learning_rate": 1.81740428404928e-05, + "loss": 3.3021, + "step": 5477 + }, + { + "epoch": 0.43919745044196346, + "grad_norm": 1.9424549341201782, + "learning_rate": 1.817329478241554e-05, + "loss": 3.3243, + "step": 5478 + }, + { + "epoch": 0.43927762522298613, + "grad_norm": 1.8037317991256714, + "learning_rate": 1.8172546586539415e-05, + "loss": 3.2356, + "step": 5479 + }, + { + "epoch": 0.43935780000400876, + "grad_norm": 1.9499531984329224, + "learning_rate": 1.8171798252877033e-05, + "loss": 3.2313, + "step": 5480 + }, + { + "epoch": 0.4394379747850314, + "grad_norm": 1.9565916061401367, + "learning_rate": 1.8171049781441013e-05, + "loss": 3.186, + "step": 5481 + }, + { + "epoch": 0.439518149566054, + "grad_norm": 1.8838820457458496, + "learning_rate": 1.8170301172243974e-05, + "loss": 3.2433, + "step": 5482 + }, + { + "epoch": 0.43959832434707663, + "grad_norm": 1.8785943984985352, + "learning_rate": 1.8169552425298534e-05, + "loss": 3.3075, + "step": 5483 + }, + { + "epoch": 0.43967849912809925, + "grad_norm": 1.799898624420166, + "learning_rate": 1.8168803540617325e-05, + "loss": 3.1809, + "step": 5484 + }, + { + "epoch": 0.4397586739091219, + "grad_norm": 1.822784662246704, + "learning_rate": 1.8168054518212962e-05, + "loss": 3.2496, + "step": 5485 + }, + { + "epoch": 0.4398388486901445, + "grad_norm": 0.8382135033607483, + "learning_rate": 1.816730535809808e-05, + "loss": 2.5028, + "step": 5486 + }, + { + "epoch": 0.4399190234711671, + "grad_norm": 1.8796132802963257, + "learning_rate": 1.8166556060285306e-05, + "loss": 3.2153, + "step": 5487 + }, + { + "epoch": 0.43999919825218975, + "grad_norm": 1.8540804386138916, + "learning_rate": 1.8165806624787278e-05, + "loss": 3.2922, + "step": 5488 + }, + { + "epoch": 0.44007937303321243, + "grad_norm": 2.1919569969177246, + "learning_rate": 1.8165057051616628e-05, + "loss": 3.2703, + "step": 5489 + }, + { + "epoch": 0.44015954781423505, + "grad_norm": 1.81381356716156, + "learning_rate": 1.8164307340785994e-05, + "loss": 3.3031, + "step": 5490 + }, + { + "epoch": 0.4402397225952577, + "grad_norm": 1.918116569519043, + "learning_rate": 1.8163557492308016e-05, + "loss": 3.2122, + "step": 5491 + }, + { + "epoch": 0.4403198973762803, + "grad_norm": 1.765909194946289, + "learning_rate": 1.8162807506195333e-05, + "loss": 3.1147, + "step": 5492 + }, + { + "epoch": 0.4404000721573029, + "grad_norm": 1.7774628400802612, + "learning_rate": 1.8162057382460593e-05, + "loss": 3.1863, + "step": 5493 + }, + { + "epoch": 0.44048024693832555, + "grad_norm": 2.0012378692626953, + "learning_rate": 1.8161307121116446e-05, + "loss": 3.2328, + "step": 5494 + }, + { + "epoch": 0.44056042171934817, + "grad_norm": 2.247084140777588, + "learning_rate": 1.8160556722175534e-05, + "loss": 3.2881, + "step": 5495 + }, + { + "epoch": 0.4406405965003708, + "grad_norm": 1.8708741664886475, + "learning_rate": 1.815980618565051e-05, + "loss": 3.2688, + "step": 5496 + }, + { + "epoch": 0.4407207712813934, + "grad_norm": 1.9973598718643188, + "learning_rate": 1.8159055511554033e-05, + "loss": 3.158, + "step": 5497 + }, + { + "epoch": 0.44080094606241604, + "grad_norm": 0.9129308462142944, + "learning_rate": 1.8158304699898753e-05, + "loss": 2.5009, + "step": 5498 + }, + { + "epoch": 0.4408811208434387, + "grad_norm": 1.7892521619796753, + "learning_rate": 1.8157553750697334e-05, + "loss": 3.2617, + "step": 5499 + }, + { + "epoch": 0.44096129562446135, + "grad_norm": 1.9244171380996704, + "learning_rate": 1.815680266396243e-05, + "loss": 3.2368, + "step": 5500 + }, + { + "epoch": 0.44104147040548397, + "grad_norm": 2.1055948734283447, + "learning_rate": 1.815605143970671e-05, + "loss": 3.2638, + "step": 5501 + }, + { + "epoch": 0.4411216451865066, + "grad_norm": 1.95768404006958, + "learning_rate": 1.8155300077942836e-05, + "loss": 3.2172, + "step": 5502 + }, + { + "epoch": 0.4412018199675292, + "grad_norm": 2.013479709625244, + "learning_rate": 1.815454857868348e-05, + "loss": 3.3732, + "step": 5503 + }, + { + "epoch": 0.44128199474855184, + "grad_norm": 0.8334648609161377, + "learning_rate": 1.8153796941941306e-05, + "loss": 2.5203, + "step": 5504 + }, + { + "epoch": 0.44136216952957447, + "grad_norm": 1.7955316305160522, + "learning_rate": 1.815304516772899e-05, + "loss": 3.1818, + "step": 5505 + }, + { + "epoch": 0.4414423443105971, + "grad_norm": 1.8581308126449585, + "learning_rate": 1.8152293256059205e-05, + "loss": 3.246, + "step": 5506 + }, + { + "epoch": 0.4415225190916197, + "grad_norm": 1.841229796409607, + "learning_rate": 1.8151541206944633e-05, + "loss": 3.1985, + "step": 5507 + }, + { + "epoch": 0.44160269387264234, + "grad_norm": 0.8555517792701721, + "learning_rate": 1.815078902039794e-05, + "loss": 2.5212, + "step": 5508 + }, + { + "epoch": 0.441682868653665, + "grad_norm": 1.9047297239303589, + "learning_rate": 1.8150036696431827e-05, + "loss": 3.1797, + "step": 5509 + }, + { + "epoch": 0.44176304343468764, + "grad_norm": 1.8396905660629272, + "learning_rate": 1.814928423505896e-05, + "loss": 3.1473, + "step": 5510 + }, + { + "epoch": 0.44184321821571027, + "grad_norm": 2.025937557220459, + "learning_rate": 1.814853163629204e-05, + "loss": 3.3167, + "step": 5511 + }, + { + "epoch": 0.4419233929967329, + "grad_norm": 1.7965980768203735, + "learning_rate": 1.8147778900143742e-05, + "loss": 3.2285, + "step": 5512 + }, + { + "epoch": 0.4420035677777555, + "grad_norm": 1.7939972877502441, + "learning_rate": 1.8147026026626767e-05, + "loss": 3.2084, + "step": 5513 + }, + { + "epoch": 0.44208374255877814, + "grad_norm": 0.7879263162612915, + "learning_rate": 1.8146273015753802e-05, + "loss": 2.5082, + "step": 5514 + }, + { + "epoch": 0.44216391733980076, + "grad_norm": 1.9913653135299683, + "learning_rate": 1.8145519867537548e-05, + "loss": 3.2326, + "step": 5515 + }, + { + "epoch": 0.4422440921208234, + "grad_norm": 1.8726856708526611, + "learning_rate": 1.8144766581990697e-05, + "loss": 3.327, + "step": 5516 + }, + { + "epoch": 0.442324266901846, + "grad_norm": 1.763090968132019, + "learning_rate": 1.8144013159125954e-05, + "loss": 3.2304, + "step": 5517 + }, + { + "epoch": 0.44240444168286863, + "grad_norm": 1.87433660030365, + "learning_rate": 1.814325959895602e-05, + "loss": 3.1956, + "step": 5518 + }, + { + "epoch": 0.44248461646389126, + "grad_norm": 1.825669765472412, + "learning_rate": 1.8142505901493596e-05, + "loss": 3.1418, + "step": 5519 + }, + { + "epoch": 0.44256479124491394, + "grad_norm": 1.879016637802124, + "learning_rate": 1.8141752066751395e-05, + "loss": 3.2126, + "step": 5520 + }, + { + "epoch": 0.44264496602593656, + "grad_norm": 1.8870248794555664, + "learning_rate": 1.8140998094742118e-05, + "loss": 3.1716, + "step": 5521 + }, + { + "epoch": 0.4427251408069592, + "grad_norm": 1.7763888835906982, + "learning_rate": 1.814024398547849e-05, + "loss": 3.1782, + "step": 5522 + }, + { + "epoch": 0.4428053155879818, + "grad_norm": 1.8921078443527222, + "learning_rate": 1.8139489738973215e-05, + "loss": 3.2087, + "step": 5523 + }, + { + "epoch": 0.44288549036900443, + "grad_norm": 1.9486950635910034, + "learning_rate": 1.813873535523901e-05, + "loss": 3.3177, + "step": 5524 + }, + { + "epoch": 0.44296566515002705, + "grad_norm": 1.8190041780471802, + "learning_rate": 1.8137980834288595e-05, + "loss": 3.264, + "step": 5525 + }, + { + "epoch": 0.4430458399310497, + "grad_norm": 1.9443974494934082, + "learning_rate": 1.813722617613469e-05, + "loss": 3.3439, + "step": 5526 + }, + { + "epoch": 0.4431260147120723, + "grad_norm": 1.9251725673675537, + "learning_rate": 1.8136471380790023e-05, + "loss": 3.22, + "step": 5527 + }, + { + "epoch": 0.4432061894930949, + "grad_norm": 1.9145666360855103, + "learning_rate": 1.8135716448267315e-05, + "loss": 3.2129, + "step": 5528 + }, + { + "epoch": 0.44328636427411755, + "grad_norm": 1.947363018989563, + "learning_rate": 1.8134961378579294e-05, + "loss": 3.2706, + "step": 5529 + }, + { + "epoch": 0.44336653905514023, + "grad_norm": 1.8011623620986938, + "learning_rate": 1.8134206171738694e-05, + "loss": 3.2499, + "step": 5530 + }, + { + "epoch": 0.44344671383616285, + "grad_norm": 1.846068024635315, + "learning_rate": 1.813345082775824e-05, + "loss": 3.3375, + "step": 5531 + }, + { + "epoch": 0.4435268886171855, + "grad_norm": 2.011549472808838, + "learning_rate": 1.8132695346650676e-05, + "loss": 3.2892, + "step": 5532 + }, + { + "epoch": 0.4436070633982081, + "grad_norm": 1.7637356519699097, + "learning_rate": 1.8131939728428733e-05, + "loss": 3.2665, + "step": 5533 + }, + { + "epoch": 0.4436872381792307, + "grad_norm": 1.8566734790802002, + "learning_rate": 1.8131183973105152e-05, + "loss": 3.302, + "step": 5534 + }, + { + "epoch": 0.44376741296025335, + "grad_norm": 1.8969663381576538, + "learning_rate": 1.8130428080692673e-05, + "loss": 3.2346, + "step": 5535 + }, + { + "epoch": 0.443847587741276, + "grad_norm": 2.020261764526367, + "learning_rate": 1.8129672051204043e-05, + "loss": 3.2619, + "step": 5536 + }, + { + "epoch": 0.4439277625222986, + "grad_norm": 1.935840129852295, + "learning_rate": 1.812891588465201e-05, + "loss": 3.1785, + "step": 5537 + }, + { + "epoch": 0.4440079373033212, + "grad_norm": 1.8780750036239624, + "learning_rate": 1.812815958104932e-05, + "loss": 3.2878, + "step": 5538 + }, + { + "epoch": 0.44408811208434384, + "grad_norm": 1.9103413820266724, + "learning_rate": 1.812740314040872e-05, + "loss": 3.2255, + "step": 5539 + }, + { + "epoch": 0.4441682868653665, + "grad_norm": 1.8479365110397339, + "learning_rate": 1.812664656274297e-05, + "loss": 3.2866, + "step": 5540 + }, + { + "epoch": 0.44424846164638915, + "grad_norm": 1.9100086688995361, + "learning_rate": 1.8125889848064825e-05, + "loss": 3.1528, + "step": 5541 + }, + { + "epoch": 0.44432863642741177, + "grad_norm": 0.8422753810882568, + "learning_rate": 1.812513299638704e-05, + "loss": 2.5162, + "step": 5542 + }, + { + "epoch": 0.4444088112084344, + "grad_norm": 1.9373862743377686, + "learning_rate": 1.8124376007722372e-05, + "loss": 3.1607, + "step": 5543 + }, + { + "epoch": 0.444488985989457, + "grad_norm": 1.8174477815628052, + "learning_rate": 1.812361888208359e-05, + "loss": 3.2899, + "step": 5544 + }, + { + "epoch": 0.44456916077047964, + "grad_norm": 1.8782658576965332, + "learning_rate": 1.8122861619483457e-05, + "loss": 3.2689, + "step": 5545 + }, + { + "epoch": 0.44464933555150227, + "grad_norm": 1.7498527765274048, + "learning_rate": 1.8122104219934746e-05, + "loss": 3.2477, + "step": 5546 + }, + { + "epoch": 0.4447295103325249, + "grad_norm": 1.8257875442504883, + "learning_rate": 1.8121346683450212e-05, + "loss": 3.3418, + "step": 5547 + }, + { + "epoch": 0.4448096851135475, + "grad_norm": 2.0755202770233154, + "learning_rate": 1.8120589010042637e-05, + "loss": 3.3228, + "step": 5548 + }, + { + "epoch": 0.44488985989457014, + "grad_norm": 1.8582557439804077, + "learning_rate": 1.8119831199724796e-05, + "loss": 3.1593, + "step": 5549 + }, + { + "epoch": 0.4449700346755928, + "grad_norm": 0.8070271611213684, + "learning_rate": 1.811907325250946e-05, + "loss": 2.5045, + "step": 5550 + }, + { + "epoch": 0.44505020945661544, + "grad_norm": 2.1581435203552246, + "learning_rate": 1.811831516840941e-05, + "loss": 3.2725, + "step": 5551 + }, + { + "epoch": 0.44513038423763807, + "grad_norm": 1.911993145942688, + "learning_rate": 1.811755694743743e-05, + "loss": 3.2743, + "step": 5552 + }, + { + "epoch": 0.4452105590186607, + "grad_norm": 1.9279366731643677, + "learning_rate": 1.8116798589606303e-05, + "loss": 3.2864, + "step": 5553 + }, + { + "epoch": 0.4452907337996833, + "grad_norm": 1.8305845260620117, + "learning_rate": 1.811604009492881e-05, + "loss": 3.194, + "step": 5554 + }, + { + "epoch": 0.44537090858070594, + "grad_norm": 1.9797419309616089, + "learning_rate": 1.8115281463417737e-05, + "loss": 3.1843, + "step": 5555 + }, + { + "epoch": 0.44545108336172856, + "grad_norm": 1.8844189643859863, + "learning_rate": 1.811452269508588e-05, + "loss": 3.397, + "step": 5556 + }, + { + "epoch": 0.4455312581427512, + "grad_norm": 0.8048921227455139, + "learning_rate": 1.8113763789946035e-05, + "loss": 2.5105, + "step": 5557 + }, + { + "epoch": 0.4456114329237738, + "grad_norm": 1.94492506980896, + "learning_rate": 1.811300474801099e-05, + "loss": 3.3093, + "step": 5558 + }, + { + "epoch": 0.44569160770479643, + "grad_norm": 1.833385944366455, + "learning_rate": 1.8112245569293543e-05, + "loss": 3.2671, + "step": 5559 + }, + { + "epoch": 0.4457717824858191, + "grad_norm": 1.8342350721359253, + "learning_rate": 1.8111486253806493e-05, + "loss": 3.3013, + "step": 5560 + }, + { + "epoch": 0.44585195726684174, + "grad_norm": 1.9243413209915161, + "learning_rate": 1.8110726801562647e-05, + "loss": 3.179, + "step": 5561 + }, + { + "epoch": 0.44593213204786436, + "grad_norm": 1.8315976858139038, + "learning_rate": 1.8109967212574805e-05, + "loss": 3.2236, + "step": 5562 + }, + { + "epoch": 0.446012306828887, + "grad_norm": 1.9153428077697754, + "learning_rate": 1.8109207486855776e-05, + "loss": 3.1064, + "step": 5563 + }, + { + "epoch": 0.4460924816099096, + "grad_norm": 1.754465103149414, + "learning_rate": 1.8108447624418363e-05, + "loss": 3.2636, + "step": 5564 + }, + { + "epoch": 0.44617265639093223, + "grad_norm": 2.0120911598205566, + "learning_rate": 1.8107687625275382e-05, + "loss": 3.2759, + "step": 5565 + }, + { + "epoch": 0.44625283117195486, + "grad_norm": 1.9675979614257812, + "learning_rate": 1.8106927489439643e-05, + "loss": 3.2081, + "step": 5566 + }, + { + "epoch": 0.4463330059529775, + "grad_norm": 1.8454009294509888, + "learning_rate": 1.8106167216923968e-05, + "loss": 3.1969, + "step": 5567 + }, + { + "epoch": 0.4464131807340001, + "grad_norm": 2.045013189315796, + "learning_rate": 1.8105406807741167e-05, + "loss": 3.3807, + "step": 5568 + }, + { + "epoch": 0.4464933555150227, + "grad_norm": 1.854312539100647, + "learning_rate": 1.8104646261904063e-05, + "loss": 3.3163, + "step": 5569 + }, + { + "epoch": 0.44657353029604535, + "grad_norm": 1.8214372396469116, + "learning_rate": 1.8103885579425485e-05, + "loss": 3.1631, + "step": 5570 + }, + { + "epoch": 0.44665370507706803, + "grad_norm": 2.0409135818481445, + "learning_rate": 1.8103124760318246e-05, + "loss": 3.3093, + "step": 5571 + }, + { + "epoch": 0.44673387985809065, + "grad_norm": 2.0306544303894043, + "learning_rate": 1.8102363804595184e-05, + "loss": 3.3439, + "step": 5572 + }, + { + "epoch": 0.4468140546391133, + "grad_norm": 1.7732895612716675, + "learning_rate": 1.8101602712269123e-05, + "loss": 3.2499, + "step": 5573 + }, + { + "epoch": 0.4468942294201359, + "grad_norm": 2.0416789054870605, + "learning_rate": 1.8100841483352894e-05, + "loss": 3.2164, + "step": 5574 + }, + { + "epoch": 0.4469744042011585, + "grad_norm": 1.9657273292541504, + "learning_rate": 1.810008011785933e-05, + "loss": 3.2764, + "step": 5575 + }, + { + "epoch": 0.44705457898218115, + "grad_norm": 1.841283917427063, + "learning_rate": 1.8099318615801274e-05, + "loss": 3.3485, + "step": 5576 + }, + { + "epoch": 0.4471347537632038, + "grad_norm": 1.7833360433578491, + "learning_rate": 1.809855697719156e-05, + "loss": 3.3826, + "step": 5577 + }, + { + "epoch": 0.4472149285442264, + "grad_norm": 1.8453421592712402, + "learning_rate": 1.809779520204303e-05, + "loss": 3.4041, + "step": 5578 + }, + { + "epoch": 0.447295103325249, + "grad_norm": 1.874778151512146, + "learning_rate": 1.8097033290368525e-05, + "loss": 3.3687, + "step": 5579 + }, + { + "epoch": 0.44737527810627165, + "grad_norm": 1.9754973649978638, + "learning_rate": 1.8096271242180898e-05, + "loss": 3.1963, + "step": 5580 + }, + { + "epoch": 0.4474554528872943, + "grad_norm": 1.9089136123657227, + "learning_rate": 1.8095509057492982e-05, + "loss": 3.3248, + "step": 5581 + }, + { + "epoch": 0.44753562766831695, + "grad_norm": 1.8350719213485718, + "learning_rate": 1.8094746736317644e-05, + "loss": 3.2285, + "step": 5582 + }, + { + "epoch": 0.4476158024493396, + "grad_norm": 1.8458119630813599, + "learning_rate": 1.8093984278667724e-05, + "loss": 3.2168, + "step": 5583 + }, + { + "epoch": 0.4476959772303622, + "grad_norm": 1.9641754627227783, + "learning_rate": 1.8093221684556082e-05, + "loss": 3.2007, + "step": 5584 + }, + { + "epoch": 0.4477761520113848, + "grad_norm": 1.9866291284561157, + "learning_rate": 1.809245895399558e-05, + "loss": 3.2614, + "step": 5585 + }, + { + "epoch": 0.44785632679240744, + "grad_norm": 1.8203097581863403, + "learning_rate": 1.809169608699907e-05, + "loss": 3.1729, + "step": 5586 + }, + { + "epoch": 0.44793650157343007, + "grad_norm": 0.7944501638412476, + "learning_rate": 1.809093308357941e-05, + "loss": 2.511, + "step": 5587 + }, + { + "epoch": 0.4480166763544527, + "grad_norm": 0.8019357323646545, + "learning_rate": 1.8090169943749477e-05, + "loss": 2.5063, + "step": 5588 + }, + { + "epoch": 0.4480968511354753, + "grad_norm": 1.8762009143829346, + "learning_rate": 1.8089406667522125e-05, + "loss": 3.1731, + "step": 5589 + }, + { + "epoch": 0.44817702591649794, + "grad_norm": 1.8570057153701782, + "learning_rate": 1.808864325491023e-05, + "loss": 3.1475, + "step": 5590 + }, + { + "epoch": 0.4482572006975206, + "grad_norm": 0.831332266330719, + "learning_rate": 1.808787970592666e-05, + "loss": 2.5208, + "step": 5591 + }, + { + "epoch": 0.44833737547854324, + "grad_norm": 0.8271859288215637, + "learning_rate": 1.808711602058429e-05, + "loss": 2.5117, + "step": 5592 + }, + { + "epoch": 0.44841755025956587, + "grad_norm": 1.8302350044250488, + "learning_rate": 1.8086352198895993e-05, + "loss": 3.1873, + "step": 5593 + }, + { + "epoch": 0.4484977250405885, + "grad_norm": 1.898112177848816, + "learning_rate": 1.8085588240874648e-05, + "loss": 3.2309, + "step": 5594 + }, + { + "epoch": 0.4485778998216111, + "grad_norm": 1.720138669013977, + "learning_rate": 1.8084824146533135e-05, + "loss": 3.2236, + "step": 5595 + }, + { + "epoch": 0.44865807460263374, + "grad_norm": 0.802154004573822, + "learning_rate": 1.8084059915884332e-05, + "loss": 2.5059, + "step": 5596 + }, + { + "epoch": 0.44873824938365636, + "grad_norm": 1.8711364269256592, + "learning_rate": 1.8083295548941135e-05, + "loss": 3.1537, + "step": 5597 + }, + { + "epoch": 0.448818424164679, + "grad_norm": 1.7904131412506104, + "learning_rate": 1.8082531045716418e-05, + "loss": 3.3342, + "step": 5598 + }, + { + "epoch": 0.4488985989457016, + "grad_norm": 1.8196126222610474, + "learning_rate": 1.8081766406223077e-05, + "loss": 3.2769, + "step": 5599 + }, + { + "epoch": 0.44897877372672423, + "grad_norm": 1.8341128826141357, + "learning_rate": 1.8081001630474007e-05, + "loss": 3.3195, + "step": 5600 + }, + { + "epoch": 0.4490589485077469, + "grad_norm": 1.8782315254211426, + "learning_rate": 1.8080236718482093e-05, + "loss": 3.1683, + "step": 5601 + }, + { + "epoch": 0.44913912328876954, + "grad_norm": 1.8728982210159302, + "learning_rate": 1.8079471670260237e-05, + "loss": 3.1824, + "step": 5602 + }, + { + "epoch": 0.44921929806979216, + "grad_norm": 1.8225603103637695, + "learning_rate": 1.8078706485821337e-05, + "loss": 3.1802, + "step": 5603 + }, + { + "epoch": 0.4492994728508148, + "grad_norm": 1.8721137046813965, + "learning_rate": 1.8077941165178293e-05, + "loss": 3.1975, + "step": 5604 + }, + { + "epoch": 0.4493796476318374, + "grad_norm": 1.8546512126922607, + "learning_rate": 1.807717570834401e-05, + "loss": 3.2745, + "step": 5605 + }, + { + "epoch": 0.44945982241286003, + "grad_norm": 1.95906400680542, + "learning_rate": 1.8076410115331385e-05, + "loss": 3.1526, + "step": 5606 + }, + { + "epoch": 0.44953999719388266, + "grad_norm": 1.97035551071167, + "learning_rate": 1.8075644386153336e-05, + "loss": 3.1903, + "step": 5607 + }, + { + "epoch": 0.4496201719749053, + "grad_norm": 1.8480461835861206, + "learning_rate": 1.807487852082277e-05, + "loss": 3.3218, + "step": 5608 + }, + { + "epoch": 0.4497003467559279, + "grad_norm": 0.8095079660415649, + "learning_rate": 1.8074112519352597e-05, + "loss": 2.5232, + "step": 5609 + }, + { + "epoch": 0.44978052153695053, + "grad_norm": 1.8665751218795776, + "learning_rate": 1.807334638175573e-05, + "loss": 3.3025, + "step": 5610 + }, + { + "epoch": 0.4498606963179732, + "grad_norm": 1.9871023893356323, + "learning_rate": 1.807258010804509e-05, + "loss": 3.2622, + "step": 5611 + }, + { + "epoch": 0.44994087109899583, + "grad_norm": 1.8730591535568237, + "learning_rate": 1.8071813698233596e-05, + "loss": 3.2636, + "step": 5612 + }, + { + "epoch": 0.45002104588001846, + "grad_norm": 1.795150637626648, + "learning_rate": 1.8071047152334165e-05, + "loss": 3.1628, + "step": 5613 + }, + { + "epoch": 0.4501012206610411, + "grad_norm": 1.9202680587768555, + "learning_rate": 1.8070280470359724e-05, + "loss": 3.2993, + "step": 5614 + }, + { + "epoch": 0.4501813954420637, + "grad_norm": 1.8444257974624634, + "learning_rate": 1.80695136523232e-05, + "loss": 3.308, + "step": 5615 + }, + { + "epoch": 0.4502615702230863, + "grad_norm": 1.8773322105407715, + "learning_rate": 1.806874669823752e-05, + "loss": 3.3854, + "step": 5616 + }, + { + "epoch": 0.45034174500410895, + "grad_norm": 1.9119595289230347, + "learning_rate": 1.8067979608115612e-05, + "loss": 3.1719, + "step": 5617 + }, + { + "epoch": 0.4504219197851316, + "grad_norm": 1.9180688858032227, + "learning_rate": 1.8067212381970414e-05, + "loss": 3.2767, + "step": 5618 + }, + { + "epoch": 0.4505020945661542, + "grad_norm": 1.9034441709518433, + "learning_rate": 1.806644501981486e-05, + "loss": 3.2513, + "step": 5619 + }, + { + "epoch": 0.4505822693471768, + "grad_norm": 1.9305505752563477, + "learning_rate": 1.806567752166188e-05, + "loss": 3.2749, + "step": 5620 + }, + { + "epoch": 0.4506624441281995, + "grad_norm": 2.0026538372039795, + "learning_rate": 1.8064909887524424e-05, + "loss": 3.2804, + "step": 5621 + }, + { + "epoch": 0.4507426189092221, + "grad_norm": 1.8666095733642578, + "learning_rate": 1.806414211741543e-05, + "loss": 3.2928, + "step": 5622 + }, + { + "epoch": 0.45082279369024475, + "grad_norm": 1.8664577007293701, + "learning_rate": 1.806337421134784e-05, + "loss": 3.2583, + "step": 5623 + }, + { + "epoch": 0.4509029684712674, + "grad_norm": 1.985636830329895, + "learning_rate": 1.80626061693346e-05, + "loss": 3.2037, + "step": 5624 + }, + { + "epoch": 0.45098314325229, + "grad_norm": 1.9806793928146362, + "learning_rate": 1.8061837991388668e-05, + "loss": 3.1715, + "step": 5625 + }, + { + "epoch": 0.4510633180333126, + "grad_norm": 2.0525729656219482, + "learning_rate": 1.8061069677522983e-05, + "loss": 3.3251, + "step": 5626 + }, + { + "epoch": 0.45114349281433525, + "grad_norm": 1.827348232269287, + "learning_rate": 1.8060301227750507e-05, + "loss": 3.3013, + "step": 5627 + }, + { + "epoch": 0.45122366759535787, + "grad_norm": 1.8485183715820312, + "learning_rate": 1.805953264208419e-05, + "loss": 3.2323, + "step": 5628 + }, + { + "epoch": 0.4513038423763805, + "grad_norm": 1.7748898267745972, + "learning_rate": 1.8058763920536995e-05, + "loss": 3.1879, + "step": 5629 + }, + { + "epoch": 0.4513840171574031, + "grad_norm": 1.9121983051300049, + "learning_rate": 1.8057995063121883e-05, + "loss": 3.2815, + "step": 5630 + }, + { + "epoch": 0.45146419193842574, + "grad_norm": 1.7607836723327637, + "learning_rate": 1.8057226069851814e-05, + "loss": 3.2498, + "step": 5631 + }, + { + "epoch": 0.4515443667194484, + "grad_norm": 1.9531304836273193, + "learning_rate": 1.805645694073975e-05, + "loss": 3.2942, + "step": 5632 + }, + { + "epoch": 0.45162454150047104, + "grad_norm": 1.8879656791687012, + "learning_rate": 1.8055687675798662e-05, + "loss": 3.2979, + "step": 5633 + }, + { + "epoch": 0.45170471628149367, + "grad_norm": 1.9141407012939453, + "learning_rate": 1.805491827504152e-05, + "loss": 3.2057, + "step": 5634 + }, + { + "epoch": 0.4517848910625163, + "grad_norm": 1.9739365577697754, + "learning_rate": 1.8054148738481293e-05, + "loss": 3.3111, + "step": 5635 + }, + { + "epoch": 0.4518650658435389, + "grad_norm": 2.086611032485962, + "learning_rate": 1.8053379066130957e-05, + "loss": 3.3334, + "step": 5636 + }, + { + "epoch": 0.45194524062456154, + "grad_norm": 1.7712913751602173, + "learning_rate": 1.805260925800349e-05, + "loss": 3.2228, + "step": 5637 + }, + { + "epoch": 0.45202541540558416, + "grad_norm": 2.96431303024292, + "learning_rate": 1.805183931411187e-05, + "loss": 3.2325, + "step": 5638 + }, + { + "epoch": 0.4521055901866068, + "grad_norm": 1.8879183530807495, + "learning_rate": 1.8051069234469073e-05, + "loss": 3.2594, + "step": 5639 + }, + { + "epoch": 0.4521857649676294, + "grad_norm": 0.8656778335571289, + "learning_rate": 1.805029901908809e-05, + "loss": 2.5179, + "step": 5640 + }, + { + "epoch": 0.45226593974865203, + "grad_norm": 1.9545248746871948, + "learning_rate": 1.80495286679819e-05, + "loss": 3.3049, + "step": 5641 + }, + { + "epoch": 0.4523461145296747, + "grad_norm": 1.8684816360473633, + "learning_rate": 1.8048758181163496e-05, + "loss": 3.2037, + "step": 5642 + }, + { + "epoch": 0.45242628931069734, + "grad_norm": 1.8672930002212524, + "learning_rate": 1.8047987558645864e-05, + "loss": 3.3001, + "step": 5643 + }, + { + "epoch": 0.45250646409171996, + "grad_norm": 1.7768495082855225, + "learning_rate": 1.8047216800441998e-05, + "loss": 3.1592, + "step": 5644 + }, + { + "epoch": 0.4525866388727426, + "grad_norm": 1.8919340372085571, + "learning_rate": 1.8046445906564893e-05, + "loss": 3.1591, + "step": 5645 + }, + { + "epoch": 0.4526668136537652, + "grad_norm": 1.9441012144088745, + "learning_rate": 1.8045674877027547e-05, + "loss": 3.2639, + "step": 5646 + }, + { + "epoch": 0.45274698843478783, + "grad_norm": 1.9705522060394287, + "learning_rate": 1.8044903711842957e-05, + "loss": 3.1793, + "step": 5647 + }, + { + "epoch": 0.45282716321581046, + "grad_norm": 1.7986950874328613, + "learning_rate": 1.8044132411024127e-05, + "loss": 3.1322, + "step": 5648 + }, + { + "epoch": 0.4529073379968331, + "grad_norm": 1.9281115531921387, + "learning_rate": 1.8043360974584055e-05, + "loss": 3.2427, + "step": 5649 + }, + { + "epoch": 0.4529875127778557, + "grad_norm": 2.0621325969696045, + "learning_rate": 1.8042589402535758e-05, + "loss": 3.1169, + "step": 5650 + }, + { + "epoch": 0.45306768755887833, + "grad_norm": 1.889514684677124, + "learning_rate": 1.8041817694892234e-05, + "loss": 3.2062, + "step": 5651 + }, + { + "epoch": 0.453147862339901, + "grad_norm": 1.8703795671463013, + "learning_rate": 1.80410458516665e-05, + "loss": 3.194, + "step": 5652 + }, + { + "epoch": 0.45322803712092363, + "grad_norm": 1.8274999856948853, + "learning_rate": 1.804027387287157e-05, + "loss": 3.1748, + "step": 5653 + }, + { + "epoch": 0.45330821190194626, + "grad_norm": 1.8340630531311035, + "learning_rate": 1.803950175852045e-05, + "loss": 3.295, + "step": 5654 + }, + { + "epoch": 0.4533883866829689, + "grad_norm": 1.8670849800109863, + "learning_rate": 1.8038729508626165e-05, + "loss": 3.1807, + "step": 5655 + }, + { + "epoch": 0.4534685614639915, + "grad_norm": 1.9173824787139893, + "learning_rate": 1.8037957123201736e-05, + "loss": 3.2816, + "step": 5656 + }, + { + "epoch": 0.45354873624501413, + "grad_norm": 1.9569629430770874, + "learning_rate": 1.8037184602260182e-05, + "loss": 3.3175, + "step": 5657 + }, + { + "epoch": 0.45362891102603675, + "grad_norm": 1.8266077041625977, + "learning_rate": 1.803641194581453e-05, + "loss": 3.2031, + "step": 5658 + }, + { + "epoch": 0.4537090858070594, + "grad_norm": 1.8303287029266357, + "learning_rate": 1.80356391538778e-05, + "loss": 3.2207, + "step": 5659 + }, + { + "epoch": 0.453789260588082, + "grad_norm": 1.8508225679397583, + "learning_rate": 1.803486622646303e-05, + "loss": 3.2953, + "step": 5660 + }, + { + "epoch": 0.4538694353691046, + "grad_norm": 1.8595874309539795, + "learning_rate": 1.8034093163583243e-05, + "loss": 3.1426, + "step": 5661 + }, + { + "epoch": 0.4539496101501273, + "grad_norm": 1.8658981323242188, + "learning_rate": 1.803331996525148e-05, + "loss": 3.2194, + "step": 5662 + }, + { + "epoch": 0.4540297849311499, + "grad_norm": 1.9572199583053589, + "learning_rate": 1.8032546631480775e-05, + "loss": 3.2682, + "step": 5663 + }, + { + "epoch": 0.45410995971217255, + "grad_norm": 1.9357997179031372, + "learning_rate": 1.8031773162284162e-05, + "loss": 3.2628, + "step": 5664 + }, + { + "epoch": 0.4541901344931952, + "grad_norm": 1.8223021030426025, + "learning_rate": 1.8030999557674683e-05, + "loss": 3.2531, + "step": 5665 + }, + { + "epoch": 0.4542703092742178, + "grad_norm": 1.8543496131896973, + "learning_rate": 1.8030225817665384e-05, + "loss": 3.2945, + "step": 5666 + }, + { + "epoch": 0.4543504840552404, + "grad_norm": 1.871235728263855, + "learning_rate": 1.802945194226931e-05, + "loss": 3.2365, + "step": 5667 + }, + { + "epoch": 0.45443065883626305, + "grad_norm": 1.851755976676941, + "learning_rate": 1.8028677931499502e-05, + "loss": 3.2823, + "step": 5668 + }, + { + "epoch": 0.45451083361728567, + "grad_norm": 1.9194673299789429, + "learning_rate": 1.8027903785369015e-05, + "loss": 3.2996, + "step": 5669 + }, + { + "epoch": 0.4545910083983083, + "grad_norm": 1.824845552444458, + "learning_rate": 1.80271295038909e-05, + "loss": 3.1898, + "step": 5670 + }, + { + "epoch": 0.4546711831793309, + "grad_norm": 1.8381950855255127, + "learning_rate": 1.802635508707821e-05, + "loss": 3.2435, + "step": 5671 + }, + { + "epoch": 0.4547513579603536, + "grad_norm": 1.9322742223739624, + "learning_rate": 1.8025580534944004e-05, + "loss": 3.4138, + "step": 5672 + }, + { + "epoch": 0.4548315327413762, + "grad_norm": 1.9366447925567627, + "learning_rate": 1.8024805847501337e-05, + "loss": 3.3063, + "step": 5673 + }, + { + "epoch": 0.45491170752239884, + "grad_norm": 0.8899372816085815, + "learning_rate": 1.802403102476327e-05, + "loss": 2.4935, + "step": 5674 + }, + { + "epoch": 0.45499188230342147, + "grad_norm": 1.9544323682785034, + "learning_rate": 1.8023256066742873e-05, + "loss": 3.1776, + "step": 5675 + }, + { + "epoch": 0.4550720570844441, + "grad_norm": 1.876394510269165, + "learning_rate": 1.8022480973453202e-05, + "loss": 3.253, + "step": 5676 + }, + { + "epoch": 0.4551522318654667, + "grad_norm": 1.8919918537139893, + "learning_rate": 1.802170574490733e-05, + "loss": 3.2249, + "step": 5677 + }, + { + "epoch": 0.45523240664648934, + "grad_norm": 1.7949600219726562, + "learning_rate": 1.8020930381118325e-05, + "loss": 3.2354, + "step": 5678 + }, + { + "epoch": 0.45531258142751196, + "grad_norm": 1.8753434419631958, + "learning_rate": 1.8020154882099262e-05, + "loss": 3.2012, + "step": 5679 + }, + { + "epoch": 0.4553927562085346, + "grad_norm": 1.8281482458114624, + "learning_rate": 1.8019379247863214e-05, + "loss": 3.196, + "step": 5680 + }, + { + "epoch": 0.4554729309895572, + "grad_norm": 1.875032901763916, + "learning_rate": 1.801860347842326e-05, + "loss": 3.2133, + "step": 5681 + }, + { + "epoch": 0.4555531057705799, + "grad_norm": 1.8496088981628418, + "learning_rate": 1.8017827573792478e-05, + "loss": 3.1024, + "step": 5682 + }, + { + "epoch": 0.4556332805516025, + "grad_norm": 1.8604849576950073, + "learning_rate": 1.8017051533983946e-05, + "loss": 3.4076, + "step": 5683 + }, + { + "epoch": 0.45571345533262514, + "grad_norm": 1.821485161781311, + "learning_rate": 1.801627535901075e-05, + "loss": 3.2002, + "step": 5684 + }, + { + "epoch": 0.45579363011364776, + "grad_norm": 1.8955553770065308, + "learning_rate": 1.801549904888598e-05, + "loss": 3.1833, + "step": 5685 + }, + { + "epoch": 0.4558738048946704, + "grad_norm": 2.3049588203430176, + "learning_rate": 1.8014722603622718e-05, + "loss": 3.441, + "step": 5686 + }, + { + "epoch": 0.455953979675693, + "grad_norm": 1.934578537940979, + "learning_rate": 1.8013946023234056e-05, + "loss": 3.1763, + "step": 5687 + }, + { + "epoch": 0.45603415445671563, + "grad_norm": 1.8550318479537964, + "learning_rate": 1.801316930773309e-05, + "loss": 3.2325, + "step": 5688 + }, + { + "epoch": 0.45611432923773826, + "grad_norm": 1.933969497680664, + "learning_rate": 1.801239245713292e-05, + "loss": 3.2567, + "step": 5689 + }, + { + "epoch": 0.4561945040187609, + "grad_norm": 1.835268497467041, + "learning_rate": 1.8011615471446632e-05, + "loss": 3.2506, + "step": 5690 + }, + { + "epoch": 0.4562746787997835, + "grad_norm": 1.9325203895568848, + "learning_rate": 1.8010838350687328e-05, + "loss": 3.3032, + "step": 5691 + }, + { + "epoch": 0.45635485358080613, + "grad_norm": 1.8760614395141602, + "learning_rate": 1.8010061094868116e-05, + "loss": 3.3167, + "step": 5692 + }, + { + "epoch": 0.4564350283618288, + "grad_norm": 2.0240371227264404, + "learning_rate": 1.8009283704002094e-05, + "loss": 3.2313, + "step": 5693 + }, + { + "epoch": 0.45651520314285143, + "grad_norm": 1.874626874923706, + "learning_rate": 1.8008506178102374e-05, + "loss": 3.2469, + "step": 5694 + }, + { + "epoch": 0.45659537792387406, + "grad_norm": 1.8259706497192383, + "learning_rate": 1.800772851718206e-05, + "loss": 3.2965, + "step": 5695 + }, + { + "epoch": 0.4566755527048967, + "grad_norm": 1.8883191347122192, + "learning_rate": 1.800695072125427e-05, + "loss": 3.2298, + "step": 5696 + }, + { + "epoch": 0.4567557274859193, + "grad_norm": 2.0112388134002686, + "learning_rate": 1.800617279033211e-05, + "loss": 3.302, + "step": 5697 + }, + { + "epoch": 0.45683590226694193, + "grad_norm": 1.876629114151001, + "learning_rate": 1.8005394724428698e-05, + "loss": 3.2161, + "step": 5698 + }, + { + "epoch": 0.45691607704796455, + "grad_norm": 1.8219066858291626, + "learning_rate": 1.8004616523557154e-05, + "loss": 3.2899, + "step": 5699 + }, + { + "epoch": 0.4569962518289872, + "grad_norm": 1.8448930978775024, + "learning_rate": 1.8003838187730595e-05, + "loss": 3.2294, + "step": 5700 + }, + { + "epoch": 0.4570764266100098, + "grad_norm": 1.8548310995101929, + "learning_rate": 1.8003059716962146e-05, + "loss": 3.3342, + "step": 5701 + }, + { + "epoch": 0.4571566013910324, + "grad_norm": 1.9661273956298828, + "learning_rate": 1.8002281111264927e-05, + "loss": 3.2666, + "step": 5702 + }, + { + "epoch": 0.4572367761720551, + "grad_norm": 1.971582055091858, + "learning_rate": 1.8001502370652072e-05, + "loss": 3.1748, + "step": 5703 + }, + { + "epoch": 0.4573169509530777, + "grad_norm": 1.9025681018829346, + "learning_rate": 1.8000723495136706e-05, + "loss": 3.287, + "step": 5704 + }, + { + "epoch": 0.45739712573410035, + "grad_norm": 1.9548941850662231, + "learning_rate": 1.7999944484731962e-05, + "loss": 3.2534, + "step": 5705 + }, + { + "epoch": 0.457477300515123, + "grad_norm": 2.0548200607299805, + "learning_rate": 1.7999165339450977e-05, + "loss": 3.1964, + "step": 5706 + }, + { + "epoch": 0.4575574752961456, + "grad_norm": 1.9080584049224854, + "learning_rate": 1.7998386059306877e-05, + "loss": 3.192, + "step": 5707 + }, + { + "epoch": 0.4576376500771682, + "grad_norm": 1.8694323301315308, + "learning_rate": 1.799760664431281e-05, + "loss": 3.1687, + "step": 5708 + }, + { + "epoch": 0.45771782485819085, + "grad_norm": 1.8810981512069702, + "learning_rate": 1.7996827094481918e-05, + "loss": 3.2896, + "step": 5709 + }, + { + "epoch": 0.45779799963921347, + "grad_norm": 1.9060122966766357, + "learning_rate": 1.7996047409827334e-05, + "loss": 3.2346, + "step": 5710 + }, + { + "epoch": 0.4578781744202361, + "grad_norm": 2.0272738933563232, + "learning_rate": 1.7995267590362208e-05, + "loss": 3.2977, + "step": 5711 + }, + { + "epoch": 0.4579583492012587, + "grad_norm": 1.9433327913284302, + "learning_rate": 1.799448763609969e-05, + "loss": 3.1482, + "step": 5712 + }, + { + "epoch": 0.4580385239822814, + "grad_norm": 1.9643678665161133, + "learning_rate": 1.7993707547052928e-05, + "loss": 3.2602, + "step": 5713 + }, + { + "epoch": 0.458118698763304, + "grad_norm": 2.055373191833496, + "learning_rate": 1.7992927323235076e-05, + "loss": 3.2152, + "step": 5714 + }, + { + "epoch": 0.45819887354432665, + "grad_norm": 1.8982781171798706, + "learning_rate": 1.7992146964659284e-05, + "loss": 3.3102, + "step": 5715 + }, + { + "epoch": 0.45827904832534927, + "grad_norm": 1.9292936325073242, + "learning_rate": 1.7991366471338713e-05, + "loss": 3.2955, + "step": 5716 + }, + { + "epoch": 0.4583592231063719, + "grad_norm": 1.836692214012146, + "learning_rate": 1.799058584328652e-05, + "loss": 3.201, + "step": 5717 + }, + { + "epoch": 0.4584393978873945, + "grad_norm": 0.8871569633483887, + "learning_rate": 1.798980508051586e-05, + "loss": 2.5127, + "step": 5718 + }, + { + "epoch": 0.45851957266841714, + "grad_norm": 2.0075504779815674, + "learning_rate": 1.798902418303991e-05, + "loss": 3.2151, + "step": 5719 + }, + { + "epoch": 0.45859974744943977, + "grad_norm": 1.9776941537857056, + "learning_rate": 1.7988243150871824e-05, + "loss": 3.2184, + "step": 5720 + }, + { + "epoch": 0.4586799222304624, + "grad_norm": 1.9564588069915771, + "learning_rate": 1.7987461984024772e-05, + "loss": 3.2113, + "step": 5721 + }, + { + "epoch": 0.458760097011485, + "grad_norm": 1.867756962776184, + "learning_rate": 1.798668068251193e-05, + "loss": 3.2698, + "step": 5722 + }, + { + "epoch": 0.4588402717925077, + "grad_norm": 1.758211612701416, + "learning_rate": 1.7985899246346462e-05, + "loss": 3.2992, + "step": 5723 + }, + { + "epoch": 0.4589204465735303, + "grad_norm": 1.7629488706588745, + "learning_rate": 1.798511767554155e-05, + "loss": 3.2858, + "step": 5724 + }, + { + "epoch": 0.45900062135455294, + "grad_norm": 2.032388687133789, + "learning_rate": 1.798433597011037e-05, + "loss": 3.3779, + "step": 5725 + }, + { + "epoch": 0.45908079613557556, + "grad_norm": 1.8047611713409424, + "learning_rate": 1.79835541300661e-05, + "loss": 3.2425, + "step": 5726 + }, + { + "epoch": 0.4591609709165982, + "grad_norm": 1.7636871337890625, + "learning_rate": 1.7982772155421916e-05, + "loss": 3.2152, + "step": 5727 + }, + { + "epoch": 0.4592411456976208, + "grad_norm": 1.8844984769821167, + "learning_rate": 1.7981990046191008e-05, + "loss": 3.3197, + "step": 5728 + }, + { + "epoch": 0.45932132047864344, + "grad_norm": 1.7636675834655762, + "learning_rate": 1.7981207802386565e-05, + "loss": 3.2507, + "step": 5729 + }, + { + "epoch": 0.45940149525966606, + "grad_norm": 1.7420639991760254, + "learning_rate": 1.798042542402177e-05, + "loss": 3.1913, + "step": 5730 + }, + { + "epoch": 0.4594816700406887, + "grad_norm": 1.9875198602676392, + "learning_rate": 1.7979642911109812e-05, + "loss": 3.2565, + "step": 5731 + }, + { + "epoch": 0.4595618448217113, + "grad_norm": 1.9573311805725098, + "learning_rate": 1.797886026366389e-05, + "loss": 3.2822, + "step": 5732 + }, + { + "epoch": 0.459642019602734, + "grad_norm": 1.8745979070663452, + "learning_rate": 1.797807748169719e-05, + "loss": 3.2473, + "step": 5733 + }, + { + "epoch": 0.4597221943837566, + "grad_norm": 1.8246597051620483, + "learning_rate": 1.797729456522292e-05, + "loss": 3.1754, + "step": 5734 + }, + { + "epoch": 0.45980236916477923, + "grad_norm": 0.8886393308639526, + "learning_rate": 1.7976511514254275e-05, + "loss": 2.5029, + "step": 5735 + }, + { + "epoch": 0.45988254394580186, + "grad_norm": 2.1247591972351074, + "learning_rate": 1.797572832880446e-05, + "loss": 3.2009, + "step": 5736 + }, + { + "epoch": 0.4599627187268245, + "grad_norm": 1.8292268514633179, + "learning_rate": 1.797494500888667e-05, + "loss": 3.2408, + "step": 5737 + }, + { + "epoch": 0.4600428935078471, + "grad_norm": 1.9384225606918335, + "learning_rate": 1.7974161554514122e-05, + "loss": 3.2821, + "step": 5738 + }, + { + "epoch": 0.46012306828886973, + "grad_norm": 1.8617831468582153, + "learning_rate": 1.7973377965700017e-05, + "loss": 3.2942, + "step": 5739 + }, + { + "epoch": 0.46020324306989235, + "grad_norm": 1.8367213010787964, + "learning_rate": 1.7972594242457575e-05, + "loss": 3.2262, + "step": 5740 + }, + { + "epoch": 0.460283417850915, + "grad_norm": 0.7826427221298218, + "learning_rate": 1.79718103848e-05, + "loss": 2.5055, + "step": 5741 + }, + { + "epoch": 0.4603635926319376, + "grad_norm": 1.771645188331604, + "learning_rate": 1.7971026392740513e-05, + "loss": 3.2905, + "step": 5742 + }, + { + "epoch": 0.4604437674129603, + "grad_norm": 1.8830469846725464, + "learning_rate": 1.797024226629233e-05, + "loss": 3.2588, + "step": 5743 + }, + { + "epoch": 0.4605239421939829, + "grad_norm": 1.8330051898956299, + "learning_rate": 1.796945800546867e-05, + "loss": 3.2769, + "step": 5744 + }, + { + "epoch": 0.46060411697500553, + "grad_norm": 1.9239028692245483, + "learning_rate": 1.796867361028276e-05, + "loss": 3.3206, + "step": 5745 + }, + { + "epoch": 0.46068429175602815, + "grad_norm": 1.8227298259735107, + "learning_rate": 1.7967889080747818e-05, + "loss": 3.1836, + "step": 5746 + }, + { + "epoch": 0.4607644665370508, + "grad_norm": 1.8116586208343506, + "learning_rate": 1.7967104416877074e-05, + "loss": 3.4064, + "step": 5747 + }, + { + "epoch": 0.4608446413180734, + "grad_norm": 2.0479087829589844, + "learning_rate": 1.796631961868376e-05, + "loss": 3.2244, + "step": 5748 + }, + { + "epoch": 0.460924816099096, + "grad_norm": 1.8702090978622437, + "learning_rate": 1.7965534686181105e-05, + "loss": 3.2896, + "step": 5749 + }, + { + "epoch": 0.46100499088011865, + "grad_norm": 2.0907039642333984, + "learning_rate": 1.796474961938234e-05, + "loss": 3.2303, + "step": 5750 + }, + { + "epoch": 0.46108516566114127, + "grad_norm": 1.9512686729431152, + "learning_rate": 1.7963964418300707e-05, + "loss": 3.3147, + "step": 5751 + }, + { + "epoch": 0.4611653404421639, + "grad_norm": 1.7767788171768188, + "learning_rate": 1.7963179082949435e-05, + "loss": 3.0919, + "step": 5752 + }, + { + "epoch": 0.4612455152231865, + "grad_norm": 1.8093342781066895, + "learning_rate": 1.7962393613341775e-05, + "loss": 3.2969, + "step": 5753 + }, + { + "epoch": 0.4613256900042092, + "grad_norm": 1.7951780557632446, + "learning_rate": 1.7961608009490966e-05, + "loss": 3.1608, + "step": 5754 + }, + { + "epoch": 0.4614058647852318, + "grad_norm": 1.8880492448806763, + "learning_rate": 1.7960822271410247e-05, + "loss": 3.1946, + "step": 5755 + }, + { + "epoch": 0.46148603956625445, + "grad_norm": 1.927444338798523, + "learning_rate": 1.796003639911287e-05, + "loss": 3.284, + "step": 5756 + }, + { + "epoch": 0.46156621434727707, + "grad_norm": 1.8529596328735352, + "learning_rate": 1.795925039261209e-05, + "loss": 3.3217, + "step": 5757 + }, + { + "epoch": 0.4616463891282997, + "grad_norm": 1.8043079376220703, + "learning_rate": 1.795846425192115e-05, + "loss": 3.2584, + "step": 5758 + }, + { + "epoch": 0.4617265639093223, + "grad_norm": 1.9471652507781982, + "learning_rate": 1.795767797705331e-05, + "loss": 3.1433, + "step": 5759 + }, + { + "epoch": 0.46180673869034494, + "grad_norm": 1.9238336086273193, + "learning_rate": 1.7956891568021822e-05, + "loss": 3.3422, + "step": 5760 + }, + { + "epoch": 0.46188691347136757, + "grad_norm": 1.8310552835464478, + "learning_rate": 1.7956105024839945e-05, + "loss": 3.1814, + "step": 5761 + }, + { + "epoch": 0.4619670882523902, + "grad_norm": 1.9775235652923584, + "learning_rate": 1.7955318347520942e-05, + "loss": 3.1486, + "step": 5762 + }, + { + "epoch": 0.4620472630334128, + "grad_norm": 1.827640175819397, + "learning_rate": 1.7954531536078076e-05, + "loss": 3.1554, + "step": 5763 + }, + { + "epoch": 0.4621274378144355, + "grad_norm": 1.867064356803894, + "learning_rate": 1.7953744590524612e-05, + "loss": 3.1813, + "step": 5764 + }, + { + "epoch": 0.4622076125954581, + "grad_norm": 1.8542536497116089, + "learning_rate": 1.7952957510873822e-05, + "loss": 3.2544, + "step": 5765 + }, + { + "epoch": 0.46228778737648074, + "grad_norm": 1.9705625772476196, + "learning_rate": 1.7952170297138963e-05, + "loss": 3.3696, + "step": 5766 + }, + { + "epoch": 0.46236796215750336, + "grad_norm": 1.8755028247833252, + "learning_rate": 1.795138294933332e-05, + "loss": 3.2493, + "step": 5767 + }, + { + "epoch": 0.462448136938526, + "grad_norm": 1.9113467931747437, + "learning_rate": 1.795059546747016e-05, + "loss": 3.2954, + "step": 5768 + }, + { + "epoch": 0.4625283117195486, + "grad_norm": 1.9388750791549683, + "learning_rate": 1.7949807851562766e-05, + "loss": 3.2722, + "step": 5769 + }, + { + "epoch": 0.46260848650057124, + "grad_norm": 1.8731608390808105, + "learning_rate": 1.7949020101624413e-05, + "loss": 3.1729, + "step": 5770 + }, + { + "epoch": 0.46268866128159386, + "grad_norm": 1.8697186708450317, + "learning_rate": 1.7948232217668378e-05, + "loss": 3.193, + "step": 5771 + }, + { + "epoch": 0.4627688360626165, + "grad_norm": 1.9198607206344604, + "learning_rate": 1.7947444199707954e-05, + "loss": 3.2938, + "step": 5772 + }, + { + "epoch": 0.4628490108436391, + "grad_norm": 1.8437200784683228, + "learning_rate": 1.7946656047756417e-05, + "loss": 3.2317, + "step": 5773 + }, + { + "epoch": 0.4629291856246618, + "grad_norm": 1.8836950063705444, + "learning_rate": 1.7945867761827064e-05, + "loss": 3.3773, + "step": 5774 + }, + { + "epoch": 0.4630093604056844, + "grad_norm": 2.0092501640319824, + "learning_rate": 1.7945079341933182e-05, + "loss": 3.2969, + "step": 5775 + }, + { + "epoch": 0.46308953518670704, + "grad_norm": 1.8611046075820923, + "learning_rate": 1.794429078808806e-05, + "loss": 3.1354, + "step": 5776 + }, + { + "epoch": 0.46316970996772966, + "grad_norm": 1.8348485231399536, + "learning_rate": 1.7943502100304995e-05, + "loss": 3.2323, + "step": 5777 + }, + { + "epoch": 0.4632498847487523, + "grad_norm": 1.8867754936218262, + "learning_rate": 1.7942713278597287e-05, + "loss": 3.1926, + "step": 5778 + }, + { + "epoch": 0.4633300595297749, + "grad_norm": 1.881575345993042, + "learning_rate": 1.7941924322978227e-05, + "loss": 3.2479, + "step": 5779 + }, + { + "epoch": 0.46341023431079753, + "grad_norm": 1.9847660064697266, + "learning_rate": 1.7941135233461128e-05, + "loss": 3.2413, + "step": 5780 + }, + { + "epoch": 0.46349040909182015, + "grad_norm": 1.9703807830810547, + "learning_rate": 1.7940346010059283e-05, + "loss": 3.2205, + "step": 5781 + }, + { + "epoch": 0.4635705838728428, + "grad_norm": 1.9434030055999756, + "learning_rate": 1.7939556652786002e-05, + "loss": 3.2846, + "step": 5782 + }, + { + "epoch": 0.4636507586538654, + "grad_norm": 2.0163919925689697, + "learning_rate": 1.7938767161654596e-05, + "loss": 3.2109, + "step": 5783 + }, + { + "epoch": 0.4637309334348881, + "grad_norm": 1.857640027999878, + "learning_rate": 1.793797753667838e-05, + "loss": 3.1244, + "step": 5784 + }, + { + "epoch": 0.4638111082159107, + "grad_norm": 2.046602249145508, + "learning_rate": 1.793718777787065e-05, + "loss": 3.2135, + "step": 5785 + }, + { + "epoch": 0.46389128299693333, + "grad_norm": 1.889825463294983, + "learning_rate": 1.7936397885244736e-05, + "loss": 3.4355, + "step": 5786 + }, + { + "epoch": 0.46397145777795595, + "grad_norm": 1.9447379112243652, + "learning_rate": 1.793560785881395e-05, + "loss": 3.2719, + "step": 5787 + }, + { + "epoch": 0.4640516325589786, + "grad_norm": 1.9023436307907104, + "learning_rate": 1.7934817698591612e-05, + "loss": 3.1578, + "step": 5788 + }, + { + "epoch": 0.4641318073400012, + "grad_norm": 1.95029616355896, + "learning_rate": 1.7934027404591042e-05, + "loss": 3.3054, + "step": 5789 + }, + { + "epoch": 0.4642119821210238, + "grad_norm": 1.9018362760543823, + "learning_rate": 1.7933236976825572e-05, + "loss": 3.2903, + "step": 5790 + }, + { + "epoch": 0.46429215690204645, + "grad_norm": 2.0361406803131104, + "learning_rate": 1.7932446415308518e-05, + "loss": 3.2357, + "step": 5791 + }, + { + "epoch": 0.4643723316830691, + "grad_norm": 1.8678927421569824, + "learning_rate": 1.7931655720053215e-05, + "loss": 3.2014, + "step": 5792 + }, + { + "epoch": 0.4644525064640917, + "grad_norm": 1.823154091835022, + "learning_rate": 1.793086489107299e-05, + "loss": 3.1356, + "step": 5793 + }, + { + "epoch": 0.4645326812451144, + "grad_norm": 1.8851069211959839, + "learning_rate": 1.793007392838118e-05, + "loss": 3.2434, + "step": 5794 + }, + { + "epoch": 0.464612856026137, + "grad_norm": 1.979417324066162, + "learning_rate": 1.7929282831991118e-05, + "loss": 3.2765, + "step": 5795 + }, + { + "epoch": 0.4646930308071596, + "grad_norm": 1.882797122001648, + "learning_rate": 1.7928491601916142e-05, + "loss": 3.259, + "step": 5796 + }, + { + "epoch": 0.46477320558818225, + "grad_norm": 0.9452863335609436, + "learning_rate": 1.7927700238169593e-05, + "loss": 2.5168, + "step": 5797 + }, + { + "epoch": 0.46485338036920487, + "grad_norm": 1.8244341611862183, + "learning_rate": 1.792690874076481e-05, + "loss": 3.172, + "step": 5798 + }, + { + "epoch": 0.4649335551502275, + "grad_norm": 1.9048020839691162, + "learning_rate": 1.792611710971514e-05, + "loss": 3.2793, + "step": 5799 + }, + { + "epoch": 0.4650137299312501, + "grad_norm": 1.9754692316055298, + "learning_rate": 1.792532534503393e-05, + "loss": 3.3417, + "step": 5800 + }, + { + "epoch": 0.46509390471227274, + "grad_norm": 2.0172390937805176, + "learning_rate": 1.7924533446734525e-05, + "loss": 3.2543, + "step": 5801 + }, + { + "epoch": 0.46517407949329537, + "grad_norm": 1.8078172206878662, + "learning_rate": 1.792374141483028e-05, + "loss": 3.1927, + "step": 5802 + }, + { + "epoch": 0.465254254274318, + "grad_norm": 1.9204963445663452, + "learning_rate": 1.7922949249334548e-05, + "loss": 3.1692, + "step": 5803 + }, + { + "epoch": 0.46533442905534067, + "grad_norm": 1.875643253326416, + "learning_rate": 1.7922156950260685e-05, + "loss": 3.1736, + "step": 5804 + }, + { + "epoch": 0.4654146038363633, + "grad_norm": 2.005570888519287, + "learning_rate": 1.792136451762205e-05, + "loss": 3.253, + "step": 5805 + }, + { + "epoch": 0.4654947786173859, + "grad_norm": 1.9009792804718018, + "learning_rate": 1.7920571951431994e-05, + "loss": 3.2189, + "step": 5806 + }, + { + "epoch": 0.46557495339840854, + "grad_norm": 1.9911047220230103, + "learning_rate": 1.791977925170389e-05, + "loss": 3.262, + "step": 5807 + }, + { + "epoch": 0.46565512817943117, + "grad_norm": 0.7952556014060974, + "learning_rate": 1.7918986418451102e-05, + "loss": 2.5121, + "step": 5808 + }, + { + "epoch": 0.4657353029604538, + "grad_norm": 0.7766630053520203, + "learning_rate": 1.791819345168699e-05, + "loss": 2.5155, + "step": 5809 + }, + { + "epoch": 0.4658154777414764, + "grad_norm": 1.8070929050445557, + "learning_rate": 1.791740035142493e-05, + "loss": 3.1836, + "step": 5810 + }, + { + "epoch": 0.46589565252249904, + "grad_norm": 1.967262625694275, + "learning_rate": 1.7916607117678287e-05, + "loss": 3.3387, + "step": 5811 + }, + { + "epoch": 0.46597582730352166, + "grad_norm": 1.7592132091522217, + "learning_rate": 1.7915813750460442e-05, + "loss": 3.1865, + "step": 5812 + }, + { + "epoch": 0.4660560020845443, + "grad_norm": 1.876189947128296, + "learning_rate": 1.7915020249784763e-05, + "loss": 3.261, + "step": 5813 + }, + { + "epoch": 0.4661361768655669, + "grad_norm": 1.7543524503707886, + "learning_rate": 1.7914226615664634e-05, + "loss": 3.1754, + "step": 5814 + }, + { + "epoch": 0.4662163516465896, + "grad_norm": 1.8358381986618042, + "learning_rate": 1.7913432848113434e-05, + "loss": 3.1853, + "step": 5815 + }, + { + "epoch": 0.4662965264276122, + "grad_norm": 1.850642204284668, + "learning_rate": 1.7912638947144546e-05, + "loss": 3.2311, + "step": 5816 + }, + { + "epoch": 0.46637670120863484, + "grad_norm": 1.935133934020996, + "learning_rate": 1.791184491277135e-05, + "loss": 3.22, + "step": 5817 + }, + { + "epoch": 0.46645687598965746, + "grad_norm": 1.8012348413467407, + "learning_rate": 1.7911050745007238e-05, + "loss": 3.1323, + "step": 5818 + }, + { + "epoch": 0.4665370507706801, + "grad_norm": 0.7753461003303528, + "learning_rate": 1.79102564438656e-05, + "loss": 2.4983, + "step": 5819 + }, + { + "epoch": 0.4666172255517027, + "grad_norm": 1.7715595960617065, + "learning_rate": 1.790946200935983e-05, + "loss": 3.1656, + "step": 5820 + }, + { + "epoch": 0.46669740033272533, + "grad_norm": 1.802634835243225, + "learning_rate": 1.7908667441503313e-05, + "loss": 3.3292, + "step": 5821 + }, + { + "epoch": 0.46677757511374796, + "grad_norm": 0.7657031416893005, + "learning_rate": 1.790787274030945e-05, + "loss": 2.4993, + "step": 5822 + }, + { + "epoch": 0.4668577498947706, + "grad_norm": 1.7658491134643555, + "learning_rate": 1.7907077905791644e-05, + "loss": 3.2189, + "step": 5823 + }, + { + "epoch": 0.4669379246757932, + "grad_norm": 1.940577507019043, + "learning_rate": 1.790628293796329e-05, + "loss": 3.2526, + "step": 5824 + }, + { + "epoch": 0.4670180994568159, + "grad_norm": 1.9008715152740479, + "learning_rate": 1.790548783683779e-05, + "loss": 3.2397, + "step": 5825 + }, + { + "epoch": 0.4670982742378385, + "grad_norm": 2.195861339569092, + "learning_rate": 1.790469260242855e-05, + "loss": 3.3496, + "step": 5826 + }, + { + "epoch": 0.46717844901886113, + "grad_norm": 1.882036566734314, + "learning_rate": 1.7903897234748983e-05, + "loss": 3.2349, + "step": 5827 + }, + { + "epoch": 0.46725862379988375, + "grad_norm": 1.8555898666381836, + "learning_rate": 1.790310173381249e-05, + "loss": 3.196, + "step": 5828 + }, + { + "epoch": 0.4673387985809064, + "grad_norm": 1.9168603420257568, + "learning_rate": 1.790230609963249e-05, + "loss": 3.2416, + "step": 5829 + }, + { + "epoch": 0.467418973361929, + "grad_norm": 1.7761741876602173, + "learning_rate": 1.790151033222239e-05, + "loss": 3.2261, + "step": 5830 + }, + { + "epoch": 0.4674991481429516, + "grad_norm": 1.990470051765442, + "learning_rate": 1.7900714431595617e-05, + "loss": 3.1589, + "step": 5831 + }, + { + "epoch": 0.46757932292397425, + "grad_norm": 1.838230848312378, + "learning_rate": 1.7899918397765578e-05, + "loss": 3.1971, + "step": 5832 + }, + { + "epoch": 0.4676594977049969, + "grad_norm": 2.0526957511901855, + "learning_rate": 1.7899122230745703e-05, + "loss": 3.1829, + "step": 5833 + }, + { + "epoch": 0.4677396724860195, + "grad_norm": 1.813183307647705, + "learning_rate": 1.789832593054941e-05, + "loss": 3.2535, + "step": 5834 + }, + { + "epoch": 0.4678198472670422, + "grad_norm": 0.8253101110458374, + "learning_rate": 1.7897529497190124e-05, + "loss": 2.5026, + "step": 5835 + }, + { + "epoch": 0.4679000220480648, + "grad_norm": 1.7507480382919312, + "learning_rate": 1.7896732930681274e-05, + "loss": 3.1315, + "step": 5836 + }, + { + "epoch": 0.4679801968290874, + "grad_norm": 1.7964860200881958, + "learning_rate": 1.7895936231036288e-05, + "loss": 3.1575, + "step": 5837 + }, + { + "epoch": 0.46806037161011005, + "grad_norm": 0.8648720979690552, + "learning_rate": 1.7895139398268604e-05, + "loss": 2.5209, + "step": 5838 + }, + { + "epoch": 0.4681405463911327, + "grad_norm": 1.8518495559692383, + "learning_rate": 1.7894342432391648e-05, + "loss": 3.2571, + "step": 5839 + }, + { + "epoch": 0.4682207211721553, + "grad_norm": 0.8241564035415649, + "learning_rate": 1.7893545333418867e-05, + "loss": 2.4922, + "step": 5840 + }, + { + "epoch": 0.4683008959531779, + "grad_norm": 1.7481497526168823, + "learning_rate": 1.7892748101363688e-05, + "loss": 3.1605, + "step": 5841 + }, + { + "epoch": 0.46838107073420054, + "grad_norm": 0.7727790474891663, + "learning_rate": 1.7891950736239558e-05, + "loss": 2.5024, + "step": 5842 + }, + { + "epoch": 0.46846124551522317, + "grad_norm": 1.9183197021484375, + "learning_rate": 1.7891153238059925e-05, + "loss": 3.295, + "step": 5843 + }, + { + "epoch": 0.4685414202962458, + "grad_norm": 1.7875161170959473, + "learning_rate": 1.7890355606838223e-05, + "loss": 3.3195, + "step": 5844 + }, + { + "epoch": 0.46862159507726847, + "grad_norm": 1.8121600151062012, + "learning_rate": 1.788955784258791e-05, + "loss": 3.2584, + "step": 5845 + }, + { + "epoch": 0.4687017698582911, + "grad_norm": 1.8476324081420898, + "learning_rate": 1.7888759945322432e-05, + "loss": 3.2916, + "step": 5846 + }, + { + "epoch": 0.4687819446393137, + "grad_norm": 0.8214660286903381, + "learning_rate": 1.788796191505524e-05, + "loss": 2.4955, + "step": 5847 + }, + { + "epoch": 0.46886211942033634, + "grad_norm": 1.7944029569625854, + "learning_rate": 1.7887163751799796e-05, + "loss": 3.343, + "step": 5848 + }, + { + "epoch": 0.46894229420135897, + "grad_norm": 1.930981993675232, + "learning_rate": 1.788636545556954e-05, + "loss": 3.3696, + "step": 5849 + }, + { + "epoch": 0.4690224689823816, + "grad_norm": 2.0787203311920166, + "learning_rate": 1.788556702637795e-05, + "loss": 3.3756, + "step": 5850 + }, + { + "epoch": 0.4691026437634042, + "grad_norm": 1.9449102878570557, + "learning_rate": 1.788476846423848e-05, + "loss": 3.287, + "step": 5851 + }, + { + "epoch": 0.46918281854442684, + "grad_norm": 0.805370032787323, + "learning_rate": 1.788396976916459e-05, + "loss": 2.5031, + "step": 5852 + }, + { + "epoch": 0.46926299332544946, + "grad_norm": 1.7734557390213013, + "learning_rate": 1.7883170941169746e-05, + "loss": 3.3037, + "step": 5853 + }, + { + "epoch": 0.4693431681064721, + "grad_norm": 1.7803019285202026, + "learning_rate": 1.7882371980267422e-05, + "loss": 3.2697, + "step": 5854 + }, + { + "epoch": 0.46942334288749477, + "grad_norm": 1.8646537065505981, + "learning_rate": 1.788157288647108e-05, + "loss": 3.2548, + "step": 5855 + }, + { + "epoch": 0.4695035176685174, + "grad_norm": 1.8919801712036133, + "learning_rate": 1.7880773659794203e-05, + "loss": 3.2173, + "step": 5856 + }, + { + "epoch": 0.46958369244954, + "grad_norm": 2.0065114498138428, + "learning_rate": 1.7879974300250256e-05, + "loss": 3.2303, + "step": 5857 + }, + { + "epoch": 0.46966386723056264, + "grad_norm": 1.9334741830825806, + "learning_rate": 1.787917480785272e-05, + "loss": 3.2776, + "step": 5858 + }, + { + "epoch": 0.46974404201158526, + "grad_norm": 1.9346776008605957, + "learning_rate": 1.7878375182615072e-05, + "loss": 3.3352, + "step": 5859 + }, + { + "epoch": 0.4698242167926079, + "grad_norm": 1.8782938718795776, + "learning_rate": 1.7877575424550797e-05, + "loss": 3.2814, + "step": 5860 + }, + { + "epoch": 0.4699043915736305, + "grad_norm": 1.868119716644287, + "learning_rate": 1.7876775533673376e-05, + "loss": 3.2162, + "step": 5861 + }, + { + "epoch": 0.46998456635465313, + "grad_norm": 1.9874557256698608, + "learning_rate": 1.7875975509996296e-05, + "loss": 3.2377, + "step": 5862 + }, + { + "epoch": 0.47006474113567576, + "grad_norm": 0.8026645183563232, + "learning_rate": 1.7875175353533042e-05, + "loss": 2.5259, + "step": 5863 + }, + { + "epoch": 0.4701449159166984, + "grad_norm": 0.7935627698898315, + "learning_rate": 1.787437506429711e-05, + "loss": 2.5028, + "step": 5864 + }, + { + "epoch": 0.470225090697721, + "grad_norm": 1.7772510051727295, + "learning_rate": 1.787357464230199e-05, + "loss": 3.1987, + "step": 5865 + }, + { + "epoch": 0.4703052654787437, + "grad_norm": 1.880112648010254, + "learning_rate": 1.7872774087561175e-05, + "loss": 3.2102, + "step": 5866 + }, + { + "epoch": 0.4703854402597663, + "grad_norm": 1.827715516090393, + "learning_rate": 1.7871973400088163e-05, + "loss": 3.2711, + "step": 5867 + }, + { + "epoch": 0.47046561504078893, + "grad_norm": 1.8324155807495117, + "learning_rate": 1.7871172579896458e-05, + "loss": 3.2773, + "step": 5868 + }, + { + "epoch": 0.47054578982181156, + "grad_norm": 1.8292280435562134, + "learning_rate": 1.7870371626999552e-05, + "loss": 3.2446, + "step": 5869 + }, + { + "epoch": 0.4706259646028342, + "grad_norm": 1.8922096490859985, + "learning_rate": 1.7869570541410955e-05, + "loss": 3.2094, + "step": 5870 + }, + { + "epoch": 0.4707061393838568, + "grad_norm": 1.8350907564163208, + "learning_rate": 1.7868769323144173e-05, + "loss": 3.3489, + "step": 5871 + }, + { + "epoch": 0.4707863141648794, + "grad_norm": 2.3462347984313965, + "learning_rate": 1.7867967972212713e-05, + "loss": 3.2112, + "step": 5872 + }, + { + "epoch": 0.47086648894590205, + "grad_norm": 1.818918228149414, + "learning_rate": 1.786716648863009e-05, + "loss": 3.2389, + "step": 5873 + }, + { + "epoch": 0.4709466637269247, + "grad_norm": 1.7777953147888184, + "learning_rate": 1.7866364872409807e-05, + "loss": 3.1581, + "step": 5874 + }, + { + "epoch": 0.4710268385079473, + "grad_norm": 1.7646863460540771, + "learning_rate": 1.7865563123565386e-05, + "loss": 3.1769, + "step": 5875 + }, + { + "epoch": 0.47110701328897, + "grad_norm": 1.7173995971679688, + "learning_rate": 1.7864761242110343e-05, + "loss": 3.2186, + "step": 5876 + }, + { + "epoch": 0.4711871880699926, + "grad_norm": 1.8233160972595215, + "learning_rate": 1.78639592280582e-05, + "loss": 3.2842, + "step": 5877 + }, + { + "epoch": 0.4712673628510152, + "grad_norm": 1.8646233081817627, + "learning_rate": 1.7863157081422473e-05, + "loss": 3.2085, + "step": 5878 + }, + { + "epoch": 0.47134753763203785, + "grad_norm": 1.931042194366455, + "learning_rate": 1.7862354802216693e-05, + "loss": 3.1465, + "step": 5879 + }, + { + "epoch": 0.4714277124130605, + "grad_norm": 1.9294850826263428, + "learning_rate": 1.7861552390454376e-05, + "loss": 3.1819, + "step": 5880 + }, + { + "epoch": 0.4715078871940831, + "grad_norm": 1.7817463874816895, + "learning_rate": 1.786074984614906e-05, + "loss": 3.2736, + "step": 5881 + }, + { + "epoch": 0.4715880619751057, + "grad_norm": 2.022536277770996, + "learning_rate": 1.785994716931427e-05, + "loss": 3.1758, + "step": 5882 + }, + { + "epoch": 0.47166823675612835, + "grad_norm": 1.8237193822860718, + "learning_rate": 1.7859144359963544e-05, + "loss": 3.3039, + "step": 5883 + }, + { + "epoch": 0.47174841153715097, + "grad_norm": 0.7991422414779663, + "learning_rate": 1.785834141811041e-05, + "loss": 2.4988, + "step": 5884 + }, + { + "epoch": 0.4718285863181736, + "grad_norm": 0.78987717628479, + "learning_rate": 1.7857538343768413e-05, + "loss": 2.4904, + "step": 5885 + }, + { + "epoch": 0.47190876109919627, + "grad_norm": 1.8061100244522095, + "learning_rate": 1.7856735136951086e-05, + "loss": 3.2971, + "step": 5886 + }, + { + "epoch": 0.4719889358802189, + "grad_norm": 1.8502954244613647, + "learning_rate": 1.7855931797671975e-05, + "loss": 3.1758, + "step": 5887 + }, + { + "epoch": 0.4720691106612415, + "grad_norm": 1.754137635231018, + "learning_rate": 1.7855128325944622e-05, + "loss": 3.2159, + "step": 5888 + }, + { + "epoch": 0.47214928544226414, + "grad_norm": 1.7989782094955444, + "learning_rate": 1.7854324721782573e-05, + "loss": 3.215, + "step": 5889 + }, + { + "epoch": 0.47222946022328677, + "grad_norm": 1.8358720541000366, + "learning_rate": 1.785352098519938e-05, + "loss": 3.3486, + "step": 5890 + }, + { + "epoch": 0.4723096350043094, + "grad_norm": 1.9112170934677124, + "learning_rate": 1.785271711620859e-05, + "loss": 3.2999, + "step": 5891 + }, + { + "epoch": 0.472389809785332, + "grad_norm": 1.8625602722167969, + "learning_rate": 1.7851913114823756e-05, + "loss": 3.2769, + "step": 5892 + }, + { + "epoch": 0.47246998456635464, + "grad_norm": 1.7764209508895874, + "learning_rate": 1.7851108981058433e-05, + "loss": 3.2413, + "step": 5893 + }, + { + "epoch": 0.47255015934737726, + "grad_norm": 2.3243532180786133, + "learning_rate": 1.7850304714926184e-05, + "loss": 3.3236, + "step": 5894 + }, + { + "epoch": 0.4726303341283999, + "grad_norm": 1.7813358306884766, + "learning_rate": 1.784950031644056e-05, + "loss": 3.2146, + "step": 5895 + }, + { + "epoch": 0.47271050890942257, + "grad_norm": 2.3095977306365967, + "learning_rate": 1.7848695785615127e-05, + "loss": 3.2202, + "step": 5896 + }, + { + "epoch": 0.4727906836904452, + "grad_norm": 1.9018293619155884, + "learning_rate": 1.7847891122463452e-05, + "loss": 3.2067, + "step": 5897 + }, + { + "epoch": 0.4728708584714678, + "grad_norm": 0.8700743317604065, + "learning_rate": 1.7847086326999098e-05, + "loss": 2.5176, + "step": 5898 + }, + { + "epoch": 0.47295103325249044, + "grad_norm": 1.9152894020080566, + "learning_rate": 1.784628139923563e-05, + "loss": 3.2484, + "step": 5899 + }, + { + "epoch": 0.47303120803351306, + "grad_norm": 1.8192250728607178, + "learning_rate": 1.784547633918663e-05, + "loss": 3.257, + "step": 5900 + }, + { + "epoch": 0.4731113828145357, + "grad_norm": 1.9391798973083496, + "learning_rate": 1.7844671146865656e-05, + "loss": 3.2457, + "step": 5901 + }, + { + "epoch": 0.4731915575955583, + "grad_norm": 1.9933212995529175, + "learning_rate": 1.7843865822286298e-05, + "loss": 3.2693, + "step": 5902 + }, + { + "epoch": 0.47327173237658093, + "grad_norm": 1.7950547933578491, + "learning_rate": 1.7843060365462122e-05, + "loss": 3.1991, + "step": 5903 + }, + { + "epoch": 0.47335190715760356, + "grad_norm": 1.8263312578201294, + "learning_rate": 1.7842254776406718e-05, + "loss": 3.2472, + "step": 5904 + }, + { + "epoch": 0.4734320819386262, + "grad_norm": 1.885432481765747, + "learning_rate": 1.7841449055133658e-05, + "loss": 3.1006, + "step": 5905 + }, + { + "epoch": 0.47351225671964886, + "grad_norm": 1.9435564279556274, + "learning_rate": 1.784064320165653e-05, + "loss": 3.2765, + "step": 5906 + }, + { + "epoch": 0.4735924315006715, + "grad_norm": 1.8741660118103027, + "learning_rate": 1.7839837215988922e-05, + "loss": 3.2319, + "step": 5907 + }, + { + "epoch": 0.4736726062816941, + "grad_norm": 1.7609493732452393, + "learning_rate": 1.783903109814442e-05, + "loss": 3.2122, + "step": 5908 + }, + { + "epoch": 0.47375278106271673, + "grad_norm": 0.8148191571235657, + "learning_rate": 1.783822484813662e-05, + "loss": 2.4965, + "step": 5909 + }, + { + "epoch": 0.47383295584373936, + "grad_norm": 1.8145264387130737, + "learning_rate": 1.7837418465979112e-05, + "loss": 3.1677, + "step": 5910 + }, + { + "epoch": 0.473913130624762, + "grad_norm": 1.8928674459457397, + "learning_rate": 1.7836611951685485e-05, + "loss": 3.3169, + "step": 5911 + }, + { + "epoch": 0.4739933054057846, + "grad_norm": 1.8644740581512451, + "learning_rate": 1.7835805305269346e-05, + "loss": 3.1746, + "step": 5912 + }, + { + "epoch": 0.47407348018680723, + "grad_norm": 1.7747927904129028, + "learning_rate": 1.7834998526744293e-05, + "loss": 3.2052, + "step": 5913 + }, + { + "epoch": 0.47415365496782985, + "grad_norm": 1.8062118291854858, + "learning_rate": 1.7834191616123924e-05, + "loss": 3.1959, + "step": 5914 + }, + { + "epoch": 0.4742338297488525, + "grad_norm": 1.823749303817749, + "learning_rate": 1.7833384573421844e-05, + "loss": 3.2627, + "step": 5915 + }, + { + "epoch": 0.47431400452987516, + "grad_norm": 2.6410105228424072, + "learning_rate": 1.7832577398651665e-05, + "loss": 3.1099, + "step": 5916 + }, + { + "epoch": 0.4743941793108978, + "grad_norm": 1.937142014503479, + "learning_rate": 1.7831770091826986e-05, + "loss": 3.3566, + "step": 5917 + }, + { + "epoch": 0.4744743540919204, + "grad_norm": 1.7368292808532715, + "learning_rate": 1.783096265296143e-05, + "loss": 3.2263, + "step": 5918 + }, + { + "epoch": 0.474554528872943, + "grad_norm": 2.0371057987213135, + "learning_rate": 1.78301550820686e-05, + "loss": 3.2833, + "step": 5919 + }, + { + "epoch": 0.47463470365396565, + "grad_norm": 1.8153257369995117, + "learning_rate": 1.7829347379162113e-05, + "loss": 3.2235, + "step": 5920 + }, + { + "epoch": 0.4747148784349883, + "grad_norm": 1.8273688554763794, + "learning_rate": 1.782853954425559e-05, + "loss": 3.2773, + "step": 5921 + }, + { + "epoch": 0.4747950532160109, + "grad_norm": 1.811120867729187, + "learning_rate": 1.782773157736265e-05, + "loss": 3.2184, + "step": 5922 + }, + { + "epoch": 0.4748752279970335, + "grad_norm": 2.020254373550415, + "learning_rate": 1.7826923478496915e-05, + "loss": 3.3561, + "step": 5923 + }, + { + "epoch": 0.47495540277805615, + "grad_norm": 1.877989411354065, + "learning_rate": 1.7826115247672007e-05, + "loss": 3.3397, + "step": 5924 + }, + { + "epoch": 0.47503557755907877, + "grad_norm": 1.749860405921936, + "learning_rate": 1.7825306884901558e-05, + "loss": 3.2397, + "step": 5925 + }, + { + "epoch": 0.4751157523401014, + "grad_norm": 1.9041650295257568, + "learning_rate": 1.7824498390199188e-05, + "loss": 3.1367, + "step": 5926 + }, + { + "epoch": 0.4751959271211241, + "grad_norm": 1.8299998044967651, + "learning_rate": 1.7823689763578535e-05, + "loss": 3.3097, + "step": 5927 + }, + { + "epoch": 0.4752761019021467, + "grad_norm": 1.9036173820495605, + "learning_rate": 1.782288100505323e-05, + "loss": 3.3187, + "step": 5928 + }, + { + "epoch": 0.4753562766831693, + "grad_norm": 1.9261962175369263, + "learning_rate": 1.7822072114636907e-05, + "loss": 3.2802, + "step": 5929 + }, + { + "epoch": 0.47543645146419194, + "grad_norm": 1.688616394996643, + "learning_rate": 1.7821263092343204e-05, + "loss": 3.1261, + "step": 5930 + }, + { + "epoch": 0.47551662624521457, + "grad_norm": 1.8990952968597412, + "learning_rate": 1.7820453938185766e-05, + "loss": 3.2139, + "step": 5931 + }, + { + "epoch": 0.4755968010262372, + "grad_norm": 1.8408013582229614, + "learning_rate": 1.7819644652178224e-05, + "loss": 3.2975, + "step": 5932 + }, + { + "epoch": 0.4756769758072598, + "grad_norm": 1.8534352779388428, + "learning_rate": 1.7818835234334235e-05, + "loss": 3.1348, + "step": 5933 + }, + { + "epoch": 0.47575715058828244, + "grad_norm": 1.9347124099731445, + "learning_rate": 1.781802568466744e-05, + "loss": 3.1014, + "step": 5934 + }, + { + "epoch": 0.47583732536930506, + "grad_norm": 1.9326460361480713, + "learning_rate": 1.7817216003191484e-05, + "loss": 3.2841, + "step": 5935 + }, + { + "epoch": 0.4759175001503277, + "grad_norm": 1.7479863166809082, + "learning_rate": 1.781640618992002e-05, + "loss": 3.2675, + "step": 5936 + }, + { + "epoch": 0.47599767493135037, + "grad_norm": 1.83118736743927, + "learning_rate": 1.7815596244866707e-05, + "loss": 3.1923, + "step": 5937 + }, + { + "epoch": 0.476077849712373, + "grad_norm": 1.9061088562011719, + "learning_rate": 1.7814786168045193e-05, + "loss": 3.2789, + "step": 5938 + }, + { + "epoch": 0.4761580244933956, + "grad_norm": 2.134460926055908, + "learning_rate": 1.7813975959469138e-05, + "loss": 3.2041, + "step": 5939 + }, + { + "epoch": 0.47623819927441824, + "grad_norm": 1.8612630367279053, + "learning_rate": 1.7813165619152206e-05, + "loss": 3.1908, + "step": 5940 + }, + { + "epoch": 0.47631837405544086, + "grad_norm": 1.9088386297225952, + "learning_rate": 1.781235514710805e-05, + "loss": 3.1886, + "step": 5941 + }, + { + "epoch": 0.4763985488364635, + "grad_norm": 1.8958827257156372, + "learning_rate": 1.7811544543350342e-05, + "loss": 3.2706, + "step": 5942 + }, + { + "epoch": 0.4764787236174861, + "grad_norm": 1.7827650308609009, + "learning_rate": 1.7810733807892746e-05, + "loss": 3.2438, + "step": 5943 + }, + { + "epoch": 0.47655889839850873, + "grad_norm": 1.9464927911758423, + "learning_rate": 1.7809922940748934e-05, + "loss": 3.2515, + "step": 5944 + }, + { + "epoch": 0.47663907317953136, + "grad_norm": 2.018784284591675, + "learning_rate": 1.780911194193257e-05, + "loss": 3.2766, + "step": 5945 + }, + { + "epoch": 0.476719247960554, + "grad_norm": 2.071622610092163, + "learning_rate": 1.780830081145733e-05, + "loss": 3.2742, + "step": 5946 + }, + { + "epoch": 0.47679942274157666, + "grad_norm": 0.8343549966812134, + "learning_rate": 1.780748954933689e-05, + "loss": 2.4957, + "step": 5947 + }, + { + "epoch": 0.4768795975225993, + "grad_norm": 1.889915108680725, + "learning_rate": 1.7806678155584933e-05, + "loss": 3.2757, + "step": 5948 + }, + { + "epoch": 0.4769597723036219, + "grad_norm": 1.9200702905654907, + "learning_rate": 1.780586663021513e-05, + "loss": 3.2406, + "step": 5949 + }, + { + "epoch": 0.47703994708464453, + "grad_norm": 2.1216893196105957, + "learning_rate": 1.7805054973241166e-05, + "loss": 3.304, + "step": 5950 + }, + { + "epoch": 0.47712012186566716, + "grad_norm": 2.0005383491516113, + "learning_rate": 1.7804243184676726e-05, + "loss": 3.247, + "step": 5951 + }, + { + "epoch": 0.4772002966466898, + "grad_norm": 0.8503440618515015, + "learning_rate": 1.78034312645355e-05, + "loss": 2.4962, + "step": 5952 + }, + { + "epoch": 0.4772804714277124, + "grad_norm": 1.871847152709961, + "learning_rate": 1.7802619212831167e-05, + "loss": 3.3317, + "step": 5953 + }, + { + "epoch": 0.47736064620873503, + "grad_norm": 1.9211143255233765, + "learning_rate": 1.7801807029577427e-05, + "loss": 3.2182, + "step": 5954 + }, + { + "epoch": 0.47744082098975765, + "grad_norm": 1.8727507591247559, + "learning_rate": 1.780099471478797e-05, + "loss": 3.2775, + "step": 5955 + }, + { + "epoch": 0.4775209957707803, + "grad_norm": 1.860800862312317, + "learning_rate": 1.7800182268476495e-05, + "loss": 3.3285, + "step": 5956 + }, + { + "epoch": 0.47760117055180296, + "grad_norm": 1.7835043668746948, + "learning_rate": 1.779936969065669e-05, + "loss": 3.1621, + "step": 5957 + }, + { + "epoch": 0.4776813453328256, + "grad_norm": 1.9050281047821045, + "learning_rate": 1.7798556981342265e-05, + "loss": 3.2262, + "step": 5958 + }, + { + "epoch": 0.4777615201138482, + "grad_norm": 1.8210399150848389, + "learning_rate": 1.7797744140546914e-05, + "loss": 3.2383, + "step": 5959 + }, + { + "epoch": 0.4778416948948708, + "grad_norm": 1.9738444089889526, + "learning_rate": 1.779693116828435e-05, + "loss": 3.2174, + "step": 5960 + }, + { + "epoch": 0.47792186967589345, + "grad_norm": 0.7961923480033875, + "learning_rate": 1.7796118064568268e-05, + "loss": 2.4996, + "step": 5961 + }, + { + "epoch": 0.4780020444569161, + "grad_norm": 1.8378500938415527, + "learning_rate": 1.779530482941239e-05, + "loss": 3.3088, + "step": 5962 + }, + { + "epoch": 0.4780822192379387, + "grad_norm": 1.9183204174041748, + "learning_rate": 1.7794491462830416e-05, + "loss": 3.2185, + "step": 5963 + }, + { + "epoch": 0.4781623940189613, + "grad_norm": 1.8972278833389282, + "learning_rate": 1.7793677964836063e-05, + "loss": 3.2267, + "step": 5964 + }, + { + "epoch": 0.47824256879998395, + "grad_norm": 1.9275457859039307, + "learning_rate": 1.7792864335443047e-05, + "loss": 3.3607, + "step": 5965 + }, + { + "epoch": 0.47832274358100657, + "grad_norm": 1.8467622995376587, + "learning_rate": 1.7792050574665083e-05, + "loss": 3.2363, + "step": 5966 + }, + { + "epoch": 0.47840291836202925, + "grad_norm": 2.0352392196655273, + "learning_rate": 1.7791236682515895e-05, + "loss": 3.2571, + "step": 5967 + }, + { + "epoch": 0.4784830931430519, + "grad_norm": 1.9368168115615845, + "learning_rate": 1.7790422659009202e-05, + "loss": 3.2377, + "step": 5968 + }, + { + "epoch": 0.4785632679240745, + "grad_norm": 1.9461842775344849, + "learning_rate": 1.7789608504158728e-05, + "loss": 3.3387, + "step": 5969 + }, + { + "epoch": 0.4786434427050971, + "grad_norm": 1.9763809442520142, + "learning_rate": 1.77887942179782e-05, + "loss": 3.2595, + "step": 5970 + }, + { + "epoch": 0.47872361748611975, + "grad_norm": 1.8121949434280396, + "learning_rate": 1.7787979800481347e-05, + "loss": 3.2803, + "step": 5971 + }, + { + "epoch": 0.47880379226714237, + "grad_norm": 1.9668264389038086, + "learning_rate": 1.7787165251681897e-05, + "loss": 3.1919, + "step": 5972 + }, + { + "epoch": 0.478883967048165, + "grad_norm": 1.7728506326675415, + "learning_rate": 1.778635057159359e-05, + "loss": 3.2581, + "step": 5973 + }, + { + "epoch": 0.4789641418291876, + "grad_norm": 1.8259495496749878, + "learning_rate": 1.7785535760230153e-05, + "loss": 3.2366, + "step": 5974 + }, + { + "epoch": 0.47904431661021024, + "grad_norm": 1.8978815078735352, + "learning_rate": 1.778472081760533e-05, + "loss": 3.2196, + "step": 5975 + }, + { + "epoch": 0.47912449139123287, + "grad_norm": 0.7875337600708008, + "learning_rate": 1.7783905743732856e-05, + "loss": 2.5106, + "step": 5976 + }, + { + "epoch": 0.47920466617225554, + "grad_norm": 1.882739543914795, + "learning_rate": 1.778309053862647e-05, + "loss": 3.2239, + "step": 5977 + }, + { + "epoch": 0.47928484095327817, + "grad_norm": 1.9444814920425415, + "learning_rate": 1.7782275202299927e-05, + "loss": 3.204, + "step": 5978 + }, + { + "epoch": 0.4793650157343008, + "grad_norm": 1.7974295616149902, + "learning_rate": 1.778145973476696e-05, + "loss": 3.2361, + "step": 5979 + }, + { + "epoch": 0.4794451905153234, + "grad_norm": 1.8880029916763306, + "learning_rate": 1.778064413604133e-05, + "loss": 3.3702, + "step": 5980 + }, + { + "epoch": 0.47952536529634604, + "grad_norm": 1.93864905834198, + "learning_rate": 1.7779828406136783e-05, + "loss": 3.2477, + "step": 5981 + }, + { + "epoch": 0.47960554007736866, + "grad_norm": 1.8635705709457397, + "learning_rate": 1.7779012545067067e-05, + "loss": 3.2266, + "step": 5982 + }, + { + "epoch": 0.4796857148583913, + "grad_norm": 1.8198034763336182, + "learning_rate": 1.7778196552845944e-05, + "loss": 3.2881, + "step": 5983 + }, + { + "epoch": 0.4797658896394139, + "grad_norm": 1.8677072525024414, + "learning_rate": 1.7777380429487167e-05, + "loss": 3.1946, + "step": 5984 + }, + { + "epoch": 0.47984606442043654, + "grad_norm": 1.8046541213989258, + "learning_rate": 1.7776564175004495e-05, + "loss": 3.1787, + "step": 5985 + }, + { + "epoch": 0.47992623920145916, + "grad_norm": 2.027663230895996, + "learning_rate": 1.7775747789411693e-05, + "loss": 3.0854, + "step": 5986 + }, + { + "epoch": 0.4800064139824818, + "grad_norm": 1.8149595260620117, + "learning_rate": 1.7774931272722525e-05, + "loss": 3.0779, + "step": 5987 + }, + { + "epoch": 0.48008658876350446, + "grad_norm": 1.8477286100387573, + "learning_rate": 1.7774114624950756e-05, + "loss": 3.2476, + "step": 5988 + }, + { + "epoch": 0.4801667635445271, + "grad_norm": 1.9533613920211792, + "learning_rate": 1.7773297846110152e-05, + "loss": 3.2587, + "step": 5989 + }, + { + "epoch": 0.4802469383255497, + "grad_norm": 1.8441067934036255, + "learning_rate": 1.777248093621449e-05, + "loss": 3.1948, + "step": 5990 + }, + { + "epoch": 0.48032711310657233, + "grad_norm": 1.910738229751587, + "learning_rate": 1.7771663895277535e-05, + "loss": 3.2584, + "step": 5991 + }, + { + "epoch": 0.48040728788759496, + "grad_norm": 1.9925662279129028, + "learning_rate": 1.7770846723313067e-05, + "loss": 3.2595, + "step": 5992 + }, + { + "epoch": 0.4804874626686176, + "grad_norm": 1.8816195726394653, + "learning_rate": 1.7770029420334866e-05, + "loss": 3.2141, + "step": 5993 + }, + { + "epoch": 0.4805676374496402, + "grad_norm": 1.8657240867614746, + "learning_rate": 1.77692119863567e-05, + "loss": 3.1711, + "step": 5994 + }, + { + "epoch": 0.48064781223066283, + "grad_norm": 1.8699977397918701, + "learning_rate": 1.776839442139236e-05, + "loss": 3.2656, + "step": 5995 + }, + { + "epoch": 0.48072798701168545, + "grad_norm": 1.9875015020370483, + "learning_rate": 1.7767576725455632e-05, + "loss": 3.2606, + "step": 5996 + }, + { + "epoch": 0.4808081617927081, + "grad_norm": 1.8819644451141357, + "learning_rate": 1.7766758898560295e-05, + "loss": 3.2194, + "step": 5997 + }, + { + "epoch": 0.48088833657373076, + "grad_norm": 1.900016188621521, + "learning_rate": 1.7765940940720143e-05, + "loss": 3.2969, + "step": 5998 + }, + { + "epoch": 0.4809685113547534, + "grad_norm": 1.9109488725662231, + "learning_rate": 1.7765122851948958e-05, + "loss": 3.2225, + "step": 5999 + }, + { + "epoch": 0.481048686135776, + "grad_norm": 1.8683944940567017, + "learning_rate": 1.776430463226054e-05, + "loss": 3.252, + "step": 6000 + }, + { + "epoch": 0.48112886091679863, + "grad_norm": 1.920392394065857, + "learning_rate": 1.7763486281668686e-05, + "loss": 3.3036, + "step": 6001 + }, + { + "epoch": 0.48120903569782125, + "grad_norm": 1.7656458616256714, + "learning_rate": 1.776266780018719e-05, + "loss": 3.0707, + "step": 6002 + }, + { + "epoch": 0.4812892104788439, + "grad_norm": 1.8100149631500244, + "learning_rate": 1.7761849187829845e-05, + "loss": 3.3235, + "step": 6003 + }, + { + "epoch": 0.4813693852598665, + "grad_norm": 1.8638139963150024, + "learning_rate": 1.776103044461046e-05, + "loss": 3.2357, + "step": 6004 + }, + { + "epoch": 0.4814495600408891, + "grad_norm": 1.9466181993484497, + "learning_rate": 1.776021157054284e-05, + "loss": 3.2349, + "step": 6005 + }, + { + "epoch": 0.48152973482191175, + "grad_norm": 1.8983964920043945, + "learning_rate": 1.7759392565640785e-05, + "loss": 3.2112, + "step": 6006 + }, + { + "epoch": 0.48160990960293437, + "grad_norm": 1.8008748292922974, + "learning_rate": 1.775857342991811e-05, + "loss": 3.2943, + "step": 6007 + }, + { + "epoch": 0.48169008438395705, + "grad_norm": 1.79539954662323, + "learning_rate": 1.7757754163388616e-05, + "loss": 3.2961, + "step": 6008 + }, + { + "epoch": 0.4817702591649797, + "grad_norm": 1.8237969875335693, + "learning_rate": 1.7756934766066124e-05, + "loss": 3.2062, + "step": 6009 + }, + { + "epoch": 0.4818504339460023, + "grad_norm": 1.7835592031478882, + "learning_rate": 1.7756115237964446e-05, + "loss": 3.1506, + "step": 6010 + }, + { + "epoch": 0.4819306087270249, + "grad_norm": 1.8986189365386963, + "learning_rate": 1.7755295579097397e-05, + "loss": 3.2024, + "step": 6011 + }, + { + "epoch": 0.48201078350804755, + "grad_norm": 1.7013158798217773, + "learning_rate": 1.77544757894788e-05, + "loss": 3.1035, + "step": 6012 + }, + { + "epoch": 0.48209095828907017, + "grad_norm": 1.8964643478393555, + "learning_rate": 1.775365586912247e-05, + "loss": 3.2295, + "step": 6013 + }, + { + "epoch": 0.4821711330700928, + "grad_norm": 1.7596796751022339, + "learning_rate": 1.7752835818042238e-05, + "loss": 3.2436, + "step": 6014 + }, + { + "epoch": 0.4822513078511154, + "grad_norm": 1.7979055643081665, + "learning_rate": 1.7752015636251925e-05, + "loss": 3.2056, + "step": 6015 + }, + { + "epoch": 0.48233148263213804, + "grad_norm": 1.7836599349975586, + "learning_rate": 1.775119532376536e-05, + "loss": 3.1566, + "step": 6016 + }, + { + "epoch": 0.48241165741316067, + "grad_norm": 1.8921589851379395, + "learning_rate": 1.7750374880596375e-05, + "loss": 3.1729, + "step": 6017 + }, + { + "epoch": 0.48249183219418335, + "grad_norm": 1.911673903465271, + "learning_rate": 1.7749554306758804e-05, + "loss": 3.2343, + "step": 6018 + }, + { + "epoch": 0.48257200697520597, + "grad_norm": 1.810977578163147, + "learning_rate": 1.7748733602266472e-05, + "loss": 3.1672, + "step": 6019 + }, + { + "epoch": 0.4826521817562286, + "grad_norm": 1.8911148309707642, + "learning_rate": 1.7747912767133227e-05, + "loss": 3.2079, + "step": 6020 + }, + { + "epoch": 0.4827323565372512, + "grad_norm": 1.8927667140960693, + "learning_rate": 1.77470918013729e-05, + "loss": 3.3112, + "step": 6021 + }, + { + "epoch": 0.48281253131827384, + "grad_norm": 1.8788436651229858, + "learning_rate": 1.774627070499934e-05, + "loss": 3.1674, + "step": 6022 + }, + { + "epoch": 0.48289270609929646, + "grad_norm": 1.8937842845916748, + "learning_rate": 1.7745449478026383e-05, + "loss": 3.3591, + "step": 6023 + }, + { + "epoch": 0.4829728808803191, + "grad_norm": 1.9840975999832153, + "learning_rate": 1.7744628120467878e-05, + "loss": 3.2865, + "step": 6024 + }, + { + "epoch": 0.4830530556613417, + "grad_norm": 1.9903719425201416, + "learning_rate": 1.7743806632337675e-05, + "loss": 3.2389, + "step": 6025 + }, + { + "epoch": 0.48313323044236434, + "grad_norm": 1.8427441120147705, + "learning_rate": 1.7742985013649616e-05, + "loss": 3.2587, + "step": 6026 + }, + { + "epoch": 0.48321340522338696, + "grad_norm": 1.893795132637024, + "learning_rate": 1.7742163264417563e-05, + "loss": 3.2734, + "step": 6027 + }, + { + "epoch": 0.48329358000440964, + "grad_norm": 1.817264437675476, + "learning_rate": 1.7741341384655366e-05, + "loss": 3.2455, + "step": 6028 + }, + { + "epoch": 0.48337375478543226, + "grad_norm": 1.82871413230896, + "learning_rate": 1.774051937437688e-05, + "loss": 3.2305, + "step": 6029 + }, + { + "epoch": 0.4834539295664549, + "grad_norm": 1.8384449481964111, + "learning_rate": 1.7739697233595967e-05, + "loss": 3.1592, + "step": 6030 + }, + { + "epoch": 0.4835341043474775, + "grad_norm": 1.9741265773773193, + "learning_rate": 1.7738874962326484e-05, + "loss": 3.2209, + "step": 6031 + }, + { + "epoch": 0.48361427912850014, + "grad_norm": 1.7175062894821167, + "learning_rate": 1.7738052560582296e-05, + "loss": 3.196, + "step": 6032 + }, + { + "epoch": 0.48369445390952276, + "grad_norm": 1.9056819677352905, + "learning_rate": 1.773723002837727e-05, + "loss": 3.1713, + "step": 6033 + }, + { + "epoch": 0.4837746286905454, + "grad_norm": 1.8798251152038574, + "learning_rate": 1.7736407365725273e-05, + "loss": 3.1714, + "step": 6034 + }, + { + "epoch": 0.483854803471568, + "grad_norm": 1.9472402334213257, + "learning_rate": 1.7735584572640174e-05, + "loss": 3.3295, + "step": 6035 + }, + { + "epoch": 0.48393497825259063, + "grad_norm": 1.959729790687561, + "learning_rate": 1.7734761649135844e-05, + "loss": 3.156, + "step": 6036 + }, + { + "epoch": 0.48401515303361325, + "grad_norm": 1.9951773881912231, + "learning_rate": 1.7733938595226163e-05, + "loss": 3.2326, + "step": 6037 + }, + { + "epoch": 0.48409532781463593, + "grad_norm": 0.8393548727035522, + "learning_rate": 1.7733115410925e-05, + "loss": 2.5134, + "step": 6038 + }, + { + "epoch": 0.48417550259565856, + "grad_norm": 1.8009568452835083, + "learning_rate": 1.7732292096246237e-05, + "loss": 3.1885, + "step": 6039 + }, + { + "epoch": 0.4842556773766812, + "grad_norm": 1.999441385269165, + "learning_rate": 1.7731468651203755e-05, + "loss": 3.368, + "step": 6040 + }, + { + "epoch": 0.4843358521577038, + "grad_norm": 1.8652693033218384, + "learning_rate": 1.7730645075811433e-05, + "loss": 3.1636, + "step": 6041 + }, + { + "epoch": 0.48441602693872643, + "grad_norm": 1.8802592754364014, + "learning_rate": 1.7729821370083164e-05, + "loss": 3.3446, + "step": 6042 + }, + { + "epoch": 0.48449620171974905, + "grad_norm": 1.9463741779327393, + "learning_rate": 1.7728997534032827e-05, + "loss": 3.3254, + "step": 6043 + }, + { + "epoch": 0.4845763765007717, + "grad_norm": 1.8235770463943481, + "learning_rate": 1.772817356767432e-05, + "loss": 3.2986, + "step": 6044 + }, + { + "epoch": 0.4846565512817943, + "grad_norm": 2.009855031967163, + "learning_rate": 1.7727349471021527e-05, + "loss": 3.1975, + "step": 6045 + }, + { + "epoch": 0.4847367260628169, + "grad_norm": 1.9304307699203491, + "learning_rate": 1.7726525244088345e-05, + "loss": 3.2192, + "step": 6046 + }, + { + "epoch": 0.48481690084383955, + "grad_norm": 1.7948294878005981, + "learning_rate": 1.7725700886888675e-05, + "loss": 3.2147, + "step": 6047 + }, + { + "epoch": 0.4848970756248622, + "grad_norm": 1.869417667388916, + "learning_rate": 1.772487639943641e-05, + "loss": 3.2214, + "step": 6048 + }, + { + "epoch": 0.48497725040588485, + "grad_norm": 1.895386815071106, + "learning_rate": 1.7724051781745445e-05, + "loss": 3.2273, + "step": 6049 + }, + { + "epoch": 0.4850574251869075, + "grad_norm": 1.8141621351242065, + "learning_rate": 1.7723227033829693e-05, + "loss": 3.1858, + "step": 6050 + }, + { + "epoch": 0.4851375999679301, + "grad_norm": 1.972041130065918, + "learning_rate": 1.7722402155703057e-05, + "loss": 3.2236, + "step": 6051 + }, + { + "epoch": 0.4852177747489527, + "grad_norm": 1.9183696508407593, + "learning_rate": 1.772157714737944e-05, + "loss": 3.2969, + "step": 6052 + }, + { + "epoch": 0.48529794952997535, + "grad_norm": 1.8831413984298706, + "learning_rate": 1.7720752008872755e-05, + "loss": 3.2516, + "step": 6053 + }, + { + "epoch": 0.48537812431099797, + "grad_norm": 1.862597107887268, + "learning_rate": 1.7719926740196913e-05, + "loss": 3.292, + "step": 6054 + }, + { + "epoch": 0.4854582990920206, + "grad_norm": 1.877394437789917, + "learning_rate": 1.771910134136582e-05, + "loss": 3.205, + "step": 6055 + }, + { + "epoch": 0.4855384738730432, + "grad_norm": 1.902071237564087, + "learning_rate": 1.7718275812393407e-05, + "loss": 3.1721, + "step": 6056 + }, + { + "epoch": 0.48561864865406584, + "grad_norm": 1.8461697101593018, + "learning_rate": 1.7717450153293584e-05, + "loss": 3.1461, + "step": 6057 + }, + { + "epoch": 0.48569882343508847, + "grad_norm": 1.9032061100006104, + "learning_rate": 1.7716624364080268e-05, + "loss": 3.2214, + "step": 6058 + }, + { + "epoch": 0.48577899821611115, + "grad_norm": 1.8242884874343872, + "learning_rate": 1.7715798444767384e-05, + "loss": 3.1152, + "step": 6059 + }, + { + "epoch": 0.48585917299713377, + "grad_norm": 1.871734619140625, + "learning_rate": 1.771497239536886e-05, + "loss": 3.2457, + "step": 6060 + }, + { + "epoch": 0.4859393477781564, + "grad_norm": 1.9656027555465698, + "learning_rate": 1.771414621589862e-05, + "loss": 3.1594, + "step": 6061 + }, + { + "epoch": 0.486019522559179, + "grad_norm": 1.8353657722473145, + "learning_rate": 1.771331990637059e-05, + "loss": 3.31, + "step": 6062 + }, + { + "epoch": 0.48609969734020164, + "grad_norm": 1.9035840034484863, + "learning_rate": 1.7712493466798706e-05, + "loss": 3.2374, + "step": 6063 + }, + { + "epoch": 0.48617987212122427, + "grad_norm": 1.8938381671905518, + "learning_rate": 1.7711666897196907e-05, + "loss": 3.2639, + "step": 6064 + }, + { + "epoch": 0.4862600469022469, + "grad_norm": 1.8049430847167969, + "learning_rate": 1.7710840197579115e-05, + "loss": 3.2339, + "step": 6065 + }, + { + "epoch": 0.4863402216832695, + "grad_norm": 1.8458470106124878, + "learning_rate": 1.7710013367959278e-05, + "loss": 3.2833, + "step": 6066 + }, + { + "epoch": 0.48642039646429214, + "grad_norm": 1.892777442932129, + "learning_rate": 1.770918640835133e-05, + "loss": 3.2194, + "step": 6067 + }, + { + "epoch": 0.48650057124531476, + "grad_norm": 1.8349148035049438, + "learning_rate": 1.770835931876922e-05, + "loss": 3.3064, + "step": 6068 + }, + { + "epoch": 0.48658074602633744, + "grad_norm": 1.7794500589370728, + "learning_rate": 1.7707532099226884e-05, + "loss": 3.2134, + "step": 6069 + }, + { + "epoch": 0.48666092080736006, + "grad_norm": 1.7249685525894165, + "learning_rate": 1.7706704749738276e-05, + "loss": 3.1812, + "step": 6070 + }, + { + "epoch": 0.4867410955883827, + "grad_norm": 1.8166251182556152, + "learning_rate": 1.7705877270317344e-05, + "loss": 3.1329, + "step": 6071 + }, + { + "epoch": 0.4868212703694053, + "grad_norm": 1.8933557271957397, + "learning_rate": 1.7705049660978035e-05, + "loss": 3.238, + "step": 6072 + }, + { + "epoch": 0.48690144515042794, + "grad_norm": 1.7828545570373535, + "learning_rate": 1.77042219217343e-05, + "loss": 3.2371, + "step": 6073 + }, + { + "epoch": 0.48698161993145056, + "grad_norm": 1.839938759803772, + "learning_rate": 1.7703394052600103e-05, + "loss": 3.1741, + "step": 6074 + }, + { + "epoch": 0.4870617947124732, + "grad_norm": 1.9818029403686523, + "learning_rate": 1.77025660535894e-05, + "loss": 3.245, + "step": 6075 + }, + { + "epoch": 0.4871419694934958, + "grad_norm": 1.8924124240875244, + "learning_rate": 1.7701737924716146e-05, + "loss": 3.1477, + "step": 6076 + }, + { + "epoch": 0.48722214427451843, + "grad_norm": 1.9156183004379272, + "learning_rate": 1.77009096659943e-05, + "loss": 3.2844, + "step": 6077 + }, + { + "epoch": 0.48730231905554106, + "grad_norm": 1.8570785522460938, + "learning_rate": 1.7700081277437834e-05, + "loss": 3.3416, + "step": 6078 + }, + { + "epoch": 0.48738249383656373, + "grad_norm": 1.794290542602539, + "learning_rate": 1.7699252759060716e-05, + "loss": 3.3106, + "step": 6079 + }, + { + "epoch": 0.48746266861758636, + "grad_norm": 1.8340922594070435, + "learning_rate": 1.7698424110876903e-05, + "loss": 3.1408, + "step": 6080 + }, + { + "epoch": 0.487542843398609, + "grad_norm": 1.9063684940338135, + "learning_rate": 1.769759533290038e-05, + "loss": 3.3243, + "step": 6081 + }, + { + "epoch": 0.4876230181796316, + "grad_norm": 1.84088933467865, + "learning_rate": 1.7696766425145108e-05, + "loss": 3.2, + "step": 6082 + }, + { + "epoch": 0.48770319296065423, + "grad_norm": 1.8814276456832886, + "learning_rate": 1.7695937387625068e-05, + "loss": 3.2745, + "step": 6083 + }, + { + "epoch": 0.48778336774167685, + "grad_norm": 1.8771580457687378, + "learning_rate": 1.7695108220354233e-05, + "loss": 3.2261, + "step": 6084 + }, + { + "epoch": 0.4878635425226995, + "grad_norm": 1.8501390218734741, + "learning_rate": 1.7694278923346585e-05, + "loss": 3.278, + "step": 6085 + }, + { + "epoch": 0.4879437173037221, + "grad_norm": 1.9510751962661743, + "learning_rate": 1.769344949661611e-05, + "loss": 3.2483, + "step": 6086 + }, + { + "epoch": 0.4880238920847447, + "grad_norm": 1.8954306840896606, + "learning_rate": 1.7692619940176786e-05, + "loss": 3.3391, + "step": 6087 + }, + { + "epoch": 0.48810406686576735, + "grad_norm": 1.909571647644043, + "learning_rate": 1.7691790254042604e-05, + "loss": 3.3285, + "step": 6088 + }, + { + "epoch": 0.48818424164679003, + "grad_norm": 1.9408479928970337, + "learning_rate": 1.7690960438227543e-05, + "loss": 3.3256, + "step": 6089 + }, + { + "epoch": 0.48826441642781265, + "grad_norm": 2.0707054138183594, + "learning_rate": 1.7690130492745605e-05, + "loss": 3.3019, + "step": 6090 + }, + { + "epoch": 0.4883445912088353, + "grad_norm": 1.9170162677764893, + "learning_rate": 1.7689300417610774e-05, + "loss": 3.287, + "step": 6091 + }, + { + "epoch": 0.4884247659898579, + "grad_norm": 1.9017153978347778, + "learning_rate": 1.768847021283705e-05, + "loss": 3.1125, + "step": 6092 + }, + { + "epoch": 0.4885049407708805, + "grad_norm": 1.8959689140319824, + "learning_rate": 1.768763987843843e-05, + "loss": 3.1985, + "step": 6093 + }, + { + "epoch": 0.48858511555190315, + "grad_norm": 1.8062092065811157, + "learning_rate": 1.7686809414428907e-05, + "loss": 3.275, + "step": 6094 + }, + { + "epoch": 0.4886652903329258, + "grad_norm": 1.8589149713516235, + "learning_rate": 1.7685978820822488e-05, + "loss": 3.1931, + "step": 6095 + }, + { + "epoch": 0.4887454651139484, + "grad_norm": 1.8535021543502808, + "learning_rate": 1.7685148097633174e-05, + "loss": 3.192, + "step": 6096 + }, + { + "epoch": 0.488825639894971, + "grad_norm": 1.936773419380188, + "learning_rate": 1.7684317244874975e-05, + "loss": 3.2601, + "step": 6097 + }, + { + "epoch": 0.48890581467599364, + "grad_norm": 1.7691407203674316, + "learning_rate": 1.768348626256189e-05, + "loss": 3.2544, + "step": 6098 + }, + { + "epoch": 0.48898598945701627, + "grad_norm": 1.841892957687378, + "learning_rate": 1.7682655150707937e-05, + "loss": 3.1219, + "step": 6099 + }, + { + "epoch": 0.48906616423803895, + "grad_norm": 1.764243721961975, + "learning_rate": 1.7681823909327125e-05, + "loss": 3.2393, + "step": 6100 + }, + { + "epoch": 0.48914633901906157, + "grad_norm": 1.8722467422485352, + "learning_rate": 1.768099253843347e-05, + "loss": 3.2006, + "step": 6101 + }, + { + "epoch": 0.4892265138000842, + "grad_norm": 1.9340263605117798, + "learning_rate": 1.768016103804099e-05, + "loss": 3.2278, + "step": 6102 + }, + { + "epoch": 0.4893066885811068, + "grad_norm": 1.791957139968872, + "learning_rate": 1.7679329408163698e-05, + "loss": 3.2692, + "step": 6103 + }, + { + "epoch": 0.48938686336212944, + "grad_norm": 0.813798189163208, + "learning_rate": 1.7678497648815622e-05, + "loss": 2.502, + "step": 6104 + }, + { + "epoch": 0.48946703814315207, + "grad_norm": 1.8542685508728027, + "learning_rate": 1.7677665760010777e-05, + "loss": 3.241, + "step": 6105 + }, + { + "epoch": 0.4895472129241747, + "grad_norm": 1.9794667959213257, + "learning_rate": 1.7676833741763198e-05, + "loss": 3.2041, + "step": 6106 + }, + { + "epoch": 0.4896273877051973, + "grad_norm": 1.878257393836975, + "learning_rate": 1.7676001594086908e-05, + "loss": 3.1866, + "step": 6107 + }, + { + "epoch": 0.48970756248621994, + "grad_norm": 1.7631529569625854, + "learning_rate": 1.7675169316995933e-05, + "loss": 3.1875, + "step": 6108 + }, + { + "epoch": 0.48978773726724256, + "grad_norm": 1.7973321676254272, + "learning_rate": 1.7674336910504307e-05, + "loss": 3.2516, + "step": 6109 + }, + { + "epoch": 0.48986791204826524, + "grad_norm": 1.834953784942627, + "learning_rate": 1.7673504374626068e-05, + "loss": 3.1549, + "step": 6110 + }, + { + "epoch": 0.48994808682928787, + "grad_norm": 1.7832679748535156, + "learning_rate": 1.767267170937525e-05, + "loss": 3.2682, + "step": 6111 + }, + { + "epoch": 0.4900282616103105, + "grad_norm": 1.8376542329788208, + "learning_rate": 1.7671838914765893e-05, + "loss": 3.2264, + "step": 6112 + }, + { + "epoch": 0.4901084363913331, + "grad_norm": 1.7589201927185059, + "learning_rate": 1.767100599081203e-05, + "loss": 3.2388, + "step": 6113 + }, + { + "epoch": 0.49018861117235574, + "grad_norm": 1.864669919013977, + "learning_rate": 1.7670172937527717e-05, + "loss": 3.1917, + "step": 6114 + }, + { + "epoch": 0.49026878595337836, + "grad_norm": 1.8582359552383423, + "learning_rate": 1.7669339754926985e-05, + "loss": 3.2634, + "step": 6115 + }, + { + "epoch": 0.490348960734401, + "grad_norm": 1.9393279552459717, + "learning_rate": 1.766850644302389e-05, + "loss": 3.3098, + "step": 6116 + }, + { + "epoch": 0.4904291355154236, + "grad_norm": 1.7656793594360352, + "learning_rate": 1.766767300183248e-05, + "loss": 3.2979, + "step": 6117 + }, + { + "epoch": 0.49050931029644623, + "grad_norm": 1.8165154457092285, + "learning_rate": 1.7666839431366804e-05, + "loss": 3.1699, + "step": 6118 + }, + { + "epoch": 0.49058948507746886, + "grad_norm": 1.814145565032959, + "learning_rate": 1.7666005731640917e-05, + "loss": 3.1967, + "step": 6119 + }, + { + "epoch": 0.49066965985849154, + "grad_norm": 1.8274173736572266, + "learning_rate": 1.7665171902668875e-05, + "loss": 3.1715, + "step": 6120 + }, + { + "epoch": 0.49074983463951416, + "grad_norm": 1.8280936479568481, + "learning_rate": 1.766433794446474e-05, + "loss": 3.2, + "step": 6121 + }, + { + "epoch": 0.4908300094205368, + "grad_norm": 2.083451509475708, + "learning_rate": 1.766350385704257e-05, + "loss": 3.2165, + "step": 6122 + }, + { + "epoch": 0.4909101842015594, + "grad_norm": 1.8346319198608398, + "learning_rate": 1.7662669640416415e-05, + "loss": 3.1177, + "step": 6123 + }, + { + "epoch": 0.49099035898258203, + "grad_norm": 2.004185676574707, + "learning_rate": 1.766183529460036e-05, + "loss": 3.2924, + "step": 6124 + }, + { + "epoch": 0.49107053376360466, + "grad_norm": 1.8690454959869385, + "learning_rate": 1.766100081960846e-05, + "loss": 3.2936, + "step": 6125 + }, + { + "epoch": 0.4911507085446273, + "grad_norm": 1.7995775938034058, + "learning_rate": 1.7660166215454785e-05, + "loss": 3.1313, + "step": 6126 + }, + { + "epoch": 0.4912308833256499, + "grad_norm": 1.9352962970733643, + "learning_rate": 1.7659331482153407e-05, + "loss": 3.2907, + "step": 6127 + }, + { + "epoch": 0.4913110581066725, + "grad_norm": 1.8105241060256958, + "learning_rate": 1.76584966197184e-05, + "loss": 3.2407, + "step": 6128 + }, + { + "epoch": 0.49139123288769515, + "grad_norm": 1.8674226999282837, + "learning_rate": 1.7657661628163837e-05, + "loss": 3.2633, + "step": 6129 + }, + { + "epoch": 0.49147140766871783, + "grad_norm": 1.9579076766967773, + "learning_rate": 1.7656826507503803e-05, + "loss": 3.1958, + "step": 6130 + }, + { + "epoch": 0.49155158244974045, + "grad_norm": 1.9048281908035278, + "learning_rate": 1.765599125775237e-05, + "loss": 3.2844, + "step": 6131 + }, + { + "epoch": 0.4916317572307631, + "grad_norm": 1.9036844968795776, + "learning_rate": 1.7655155878923623e-05, + "loss": 3.1769, + "step": 6132 + }, + { + "epoch": 0.4917119320117857, + "grad_norm": 2.4210145473480225, + "learning_rate": 1.7654320371031645e-05, + "loss": 3.1783, + "step": 6133 + }, + { + "epoch": 0.4917921067928083, + "grad_norm": 0.8261605501174927, + "learning_rate": 1.7653484734090522e-05, + "loss": 2.5057, + "step": 6134 + }, + { + "epoch": 0.49187228157383095, + "grad_norm": 1.9307430982589722, + "learning_rate": 1.7652648968114344e-05, + "loss": 3.2664, + "step": 6135 + }, + { + "epoch": 0.4919524563548536, + "grad_norm": 1.9068117141723633, + "learning_rate": 1.7651813073117204e-05, + "loss": 3.29, + "step": 6136 + }, + { + "epoch": 0.4920326311358762, + "grad_norm": 2.085538625717163, + "learning_rate": 1.765097704911319e-05, + "loss": 3.1569, + "step": 6137 + }, + { + "epoch": 0.4921128059168988, + "grad_norm": 1.8845455646514893, + "learning_rate": 1.7650140896116404e-05, + "loss": 3.2337, + "step": 6138 + }, + { + "epoch": 0.49219298069792145, + "grad_norm": 1.9153900146484375, + "learning_rate": 1.7649304614140935e-05, + "loss": 3.3512, + "step": 6139 + }, + { + "epoch": 0.4922731554789441, + "grad_norm": 2.0041112899780273, + "learning_rate": 1.7648468203200885e-05, + "loss": 3.3459, + "step": 6140 + }, + { + "epoch": 0.49235333025996675, + "grad_norm": 1.9557602405548096, + "learning_rate": 1.7647631663310358e-05, + "loss": 3.3124, + "step": 6141 + }, + { + "epoch": 0.49243350504098937, + "grad_norm": 1.9156596660614014, + "learning_rate": 1.7646794994483456e-05, + "loss": 3.1838, + "step": 6142 + }, + { + "epoch": 0.492513679822012, + "grad_norm": 2.056159019470215, + "learning_rate": 1.7645958196734287e-05, + "loss": 3.2163, + "step": 6143 + }, + { + "epoch": 0.4925938546030346, + "grad_norm": 1.8404178619384766, + "learning_rate": 1.7645121270076956e-05, + "loss": 3.1849, + "step": 6144 + }, + { + "epoch": 0.49267402938405724, + "grad_norm": 1.9356284141540527, + "learning_rate": 1.7644284214525574e-05, + "loss": 3.2205, + "step": 6145 + }, + { + "epoch": 0.49275420416507987, + "grad_norm": 1.8490846157073975, + "learning_rate": 1.7643447030094254e-05, + "loss": 3.1417, + "step": 6146 + }, + { + "epoch": 0.4928343789461025, + "grad_norm": 2.047694683074951, + "learning_rate": 1.7642609716797113e-05, + "loss": 3.2871, + "step": 6147 + }, + { + "epoch": 0.4929145537271251, + "grad_norm": 1.823763132095337, + "learning_rate": 1.7641772274648263e-05, + "loss": 3.1966, + "step": 6148 + }, + { + "epoch": 0.49299472850814774, + "grad_norm": 1.9890949726104736, + "learning_rate": 1.764093470366183e-05, + "loss": 3.2721, + "step": 6149 + }, + { + "epoch": 0.4930749032891704, + "grad_norm": 1.9950141906738281, + "learning_rate": 1.764009700385193e-05, + "loss": 3.2001, + "step": 6150 + }, + { + "epoch": 0.49315507807019304, + "grad_norm": 1.95452880859375, + "learning_rate": 1.7639259175232682e-05, + "loss": 3.1997, + "step": 6151 + }, + { + "epoch": 0.49323525285121567, + "grad_norm": 0.8155184388160706, + "learning_rate": 1.763842121781822e-05, + "loss": 2.5371, + "step": 6152 + }, + { + "epoch": 0.4933154276322383, + "grad_norm": 1.7123827934265137, + "learning_rate": 1.7637583131622668e-05, + "loss": 3.0853, + "step": 6153 + }, + { + "epoch": 0.4933956024132609, + "grad_norm": 1.762109398841858, + "learning_rate": 1.7636744916660157e-05, + "loss": 3.2097, + "step": 6154 + }, + { + "epoch": 0.49347577719428354, + "grad_norm": 1.8798104524612427, + "learning_rate": 1.763590657294482e-05, + "loss": 3.1863, + "step": 6155 + }, + { + "epoch": 0.49355595197530616, + "grad_norm": 1.9504879713058472, + "learning_rate": 1.7635068100490785e-05, + "loss": 3.1955, + "step": 6156 + }, + { + "epoch": 0.4936361267563288, + "grad_norm": 1.745255947113037, + "learning_rate": 1.7634229499312194e-05, + "loss": 3.1467, + "step": 6157 + }, + { + "epoch": 0.4937163015373514, + "grad_norm": 0.8215908408164978, + "learning_rate": 1.7633390769423186e-05, + "loss": 2.5319, + "step": 6158 + }, + { + "epoch": 0.49379647631837403, + "grad_norm": 2.0653796195983887, + "learning_rate": 1.76325519108379e-05, + "loss": 3.2791, + "step": 6159 + }, + { + "epoch": 0.49387665109939666, + "grad_norm": 2.0048625469207764, + "learning_rate": 1.7631712923570474e-05, + "loss": 3.285, + "step": 6160 + }, + { + "epoch": 0.49395682588041934, + "grad_norm": 1.9628703594207764, + "learning_rate": 1.7630873807635063e-05, + "loss": 3.3409, + "step": 6161 + }, + { + "epoch": 0.49403700066144196, + "grad_norm": 0.8183247447013855, + "learning_rate": 1.7630034563045807e-05, + "loss": 2.4989, + "step": 6162 + }, + { + "epoch": 0.4941171754424646, + "grad_norm": 1.8334507942199707, + "learning_rate": 1.7629195189816858e-05, + "loss": 3.1785, + "step": 6163 + }, + { + "epoch": 0.4941973502234872, + "grad_norm": 1.7640714645385742, + "learning_rate": 1.7628355687962368e-05, + "loss": 3.1736, + "step": 6164 + }, + { + "epoch": 0.49427752500450983, + "grad_norm": 1.7630290985107422, + "learning_rate": 1.7627516057496485e-05, + "loss": 3.0981, + "step": 6165 + }, + { + "epoch": 0.49435769978553246, + "grad_norm": 1.8656009435653687, + "learning_rate": 1.7626676298433373e-05, + "loss": 3.2067, + "step": 6166 + }, + { + "epoch": 0.4944378745665551, + "grad_norm": 1.8504692316055298, + "learning_rate": 1.7625836410787186e-05, + "loss": 3.1994, + "step": 6167 + }, + { + "epoch": 0.4945180493475777, + "grad_norm": 0.8135337233543396, + "learning_rate": 1.7624996394572088e-05, + "loss": 2.5234, + "step": 6168 + }, + { + "epoch": 0.49459822412860033, + "grad_norm": 1.9570012092590332, + "learning_rate": 1.7624156249802233e-05, + "loss": 3.2207, + "step": 6169 + }, + { + "epoch": 0.49467839890962295, + "grad_norm": 2.046417713165283, + "learning_rate": 1.762331597649179e-05, + "loss": 3.386, + "step": 6170 + }, + { + "epoch": 0.49475857369064563, + "grad_norm": 1.9270182847976685, + "learning_rate": 1.7622475574654932e-05, + "loss": 3.2686, + "step": 6171 + }, + { + "epoch": 0.49483874847166825, + "grad_norm": 1.8153226375579834, + "learning_rate": 1.762163504430582e-05, + "loss": 3.1786, + "step": 6172 + }, + { + "epoch": 0.4949189232526909, + "grad_norm": 1.943493127822876, + "learning_rate": 1.7620794385458625e-05, + "loss": 3.2823, + "step": 6173 + }, + { + "epoch": 0.4949990980337135, + "grad_norm": 1.81804358959198, + "learning_rate": 1.7619953598127524e-05, + "loss": 3.1166, + "step": 6174 + }, + { + "epoch": 0.4950792728147361, + "grad_norm": 1.8238818645477295, + "learning_rate": 1.761911268232669e-05, + "loss": 3.2828, + "step": 6175 + }, + { + "epoch": 0.49515944759575875, + "grad_norm": 1.7981834411621094, + "learning_rate": 1.7618271638070303e-05, + "loss": 3.2081, + "step": 6176 + }, + { + "epoch": 0.4952396223767814, + "grad_norm": 1.9126721620559692, + "learning_rate": 1.761743046537254e-05, + "loss": 3.2525, + "step": 6177 + }, + { + "epoch": 0.495319797157804, + "grad_norm": 1.794223427772522, + "learning_rate": 1.7616589164247586e-05, + "loss": 3.3018, + "step": 6178 + }, + { + "epoch": 0.4953999719388266, + "grad_norm": 2.0652992725372314, + "learning_rate": 1.761574773470962e-05, + "loss": 3.2915, + "step": 6179 + }, + { + "epoch": 0.49548014671984925, + "grad_norm": 1.9176214933395386, + "learning_rate": 1.761490617677283e-05, + "loss": 3.2267, + "step": 6180 + }, + { + "epoch": 0.4955603215008719, + "grad_norm": 1.7733747959136963, + "learning_rate": 1.7614064490451412e-05, + "loss": 3.2625, + "step": 6181 + }, + { + "epoch": 0.49564049628189455, + "grad_norm": 1.7342448234558105, + "learning_rate": 1.7613222675759548e-05, + "loss": 3.2057, + "step": 6182 + }, + { + "epoch": 0.4957206710629172, + "grad_norm": 2.014650344848633, + "learning_rate": 1.7612380732711427e-05, + "loss": 3.241, + "step": 6183 + }, + { + "epoch": 0.4958008458439398, + "grad_norm": 1.81796133518219, + "learning_rate": 1.7611538661321254e-05, + "loss": 3.1894, + "step": 6184 + }, + { + "epoch": 0.4958810206249624, + "grad_norm": 1.9349935054779053, + "learning_rate": 1.7610696461603222e-05, + "loss": 3.2173, + "step": 6185 + }, + { + "epoch": 0.49596119540598504, + "grad_norm": 2.1980183124542236, + "learning_rate": 1.7609854133571532e-05, + "loss": 3.1864, + "step": 6186 + }, + { + "epoch": 0.49604137018700767, + "grad_norm": 1.9599971771240234, + "learning_rate": 1.760901167724038e-05, + "loss": 3.1631, + "step": 6187 + }, + { + "epoch": 0.4961215449680303, + "grad_norm": 1.8245455026626587, + "learning_rate": 1.760816909262397e-05, + "loss": 3.2838, + "step": 6188 + }, + { + "epoch": 0.4962017197490529, + "grad_norm": 1.790819525718689, + "learning_rate": 1.7607326379736516e-05, + "loss": 3.3048, + "step": 6189 + }, + { + "epoch": 0.49628189453007554, + "grad_norm": 1.8952213525772095, + "learning_rate": 1.7606483538592217e-05, + "loss": 3.3112, + "step": 6190 + }, + { + "epoch": 0.4963620693110982, + "grad_norm": 1.8653572797775269, + "learning_rate": 1.7605640569205293e-05, + "loss": 3.3356, + "step": 6191 + }, + { + "epoch": 0.49644224409212084, + "grad_norm": 1.8944478034973145, + "learning_rate": 1.7604797471589943e-05, + "loss": 3.1623, + "step": 6192 + }, + { + "epoch": 0.49652241887314347, + "grad_norm": 1.8653534650802612, + "learning_rate": 1.760395424576039e-05, + "loss": 3.2453, + "step": 6193 + }, + { + "epoch": 0.4966025936541661, + "grad_norm": 1.859745979309082, + "learning_rate": 1.7603110891730847e-05, + "loss": 3.2472, + "step": 6194 + }, + { + "epoch": 0.4966827684351887, + "grad_norm": 1.8893697261810303, + "learning_rate": 1.7602267409515533e-05, + "loss": 3.2819, + "step": 6195 + }, + { + "epoch": 0.49676294321621134, + "grad_norm": 1.9264861345291138, + "learning_rate": 1.760142379912867e-05, + "loss": 3.2258, + "step": 6196 + }, + { + "epoch": 0.49684311799723396, + "grad_norm": 1.7752354145050049, + "learning_rate": 1.7600580060584483e-05, + "loss": 3.2964, + "step": 6197 + }, + { + "epoch": 0.4969232927782566, + "grad_norm": 1.815602421760559, + "learning_rate": 1.759973619389719e-05, + "loss": 3.2446, + "step": 6198 + }, + { + "epoch": 0.4970034675592792, + "grad_norm": 1.9462851285934448, + "learning_rate": 1.759889219908103e-05, + "loss": 3.2045, + "step": 6199 + }, + { + "epoch": 0.49708364234030183, + "grad_norm": 1.9139795303344727, + "learning_rate": 1.7598048076150225e-05, + "loss": 3.3003, + "step": 6200 + }, + { + "epoch": 0.4971638171213245, + "grad_norm": 1.910021185874939, + "learning_rate": 1.7597203825119002e-05, + "loss": 3.2544, + "step": 6201 + }, + { + "epoch": 0.49724399190234714, + "grad_norm": 1.940714716911316, + "learning_rate": 1.7596359446001605e-05, + "loss": 3.2379, + "step": 6202 + }, + { + "epoch": 0.49732416668336976, + "grad_norm": 1.8398996591567993, + "learning_rate": 1.759551493881226e-05, + "loss": 3.2487, + "step": 6203 + }, + { + "epoch": 0.4974043414643924, + "grad_norm": 1.9215893745422363, + "learning_rate": 1.7594670303565215e-05, + "loss": 3.2072, + "step": 6204 + }, + { + "epoch": 0.497484516245415, + "grad_norm": 1.8341318368911743, + "learning_rate": 1.7593825540274703e-05, + "loss": 3.2396, + "step": 6205 + }, + { + "epoch": 0.49756469102643763, + "grad_norm": 1.8562010526657104, + "learning_rate": 1.7592980648954968e-05, + "loss": 3.2569, + "step": 6206 + }, + { + "epoch": 0.49764486580746026, + "grad_norm": 1.8377203941345215, + "learning_rate": 1.7592135629620257e-05, + "loss": 3.2615, + "step": 6207 + }, + { + "epoch": 0.4977250405884829, + "grad_norm": 1.9303791522979736, + "learning_rate": 1.7591290482284814e-05, + "loss": 3.267, + "step": 6208 + }, + { + "epoch": 0.4978052153695055, + "grad_norm": 1.8823472261428833, + "learning_rate": 1.759044520696289e-05, + "loss": 3.2225, + "step": 6209 + }, + { + "epoch": 0.49788539015052813, + "grad_norm": 1.9588285684585571, + "learning_rate": 1.7589599803668732e-05, + "loss": 3.1963, + "step": 6210 + }, + { + "epoch": 0.4979655649315508, + "grad_norm": 1.8151074647903442, + "learning_rate": 1.7588754272416598e-05, + "loss": 3.2202, + "step": 6211 + }, + { + "epoch": 0.49804573971257343, + "grad_norm": 1.9096713066101074, + "learning_rate": 1.7587908613220744e-05, + "loss": 3.3842, + "step": 6212 + }, + { + "epoch": 0.49812591449359606, + "grad_norm": 1.9170621633529663, + "learning_rate": 1.7587062826095424e-05, + "loss": 3.2201, + "step": 6213 + }, + { + "epoch": 0.4982060892746187, + "grad_norm": 2.035311698913574, + "learning_rate": 1.7586216911054897e-05, + "loss": 3.3539, + "step": 6214 + }, + { + "epoch": 0.4982862640556413, + "grad_norm": 1.9673733711242676, + "learning_rate": 1.7585370868113426e-05, + "loss": 3.3804, + "step": 6215 + }, + { + "epoch": 0.4983664388366639, + "grad_norm": 1.791223406791687, + "learning_rate": 1.7584524697285274e-05, + "loss": 3.2015, + "step": 6216 + }, + { + "epoch": 0.49844661361768655, + "grad_norm": 0.8556339144706726, + "learning_rate": 1.758367839858471e-05, + "loss": 2.5009, + "step": 6217 + }, + { + "epoch": 0.4985267883987092, + "grad_norm": 1.9334893226623535, + "learning_rate": 1.7582831972026003e-05, + "loss": 3.3296, + "step": 6218 + }, + { + "epoch": 0.4986069631797318, + "grad_norm": 0.7699762582778931, + "learning_rate": 1.7581985417623423e-05, + "loss": 2.5011, + "step": 6219 + }, + { + "epoch": 0.4986871379607544, + "grad_norm": 1.9418472051620483, + "learning_rate": 1.7581138735391237e-05, + "loss": 3.3193, + "step": 6220 + }, + { + "epoch": 0.49876731274177705, + "grad_norm": 2.006131172180176, + "learning_rate": 1.7580291925343725e-05, + "loss": 3.2437, + "step": 6221 + }, + { + "epoch": 0.4988474875227997, + "grad_norm": 2.0325920581817627, + "learning_rate": 1.757944498749516e-05, + "loss": 3.2236, + "step": 6222 + }, + { + "epoch": 0.49892766230382235, + "grad_norm": 1.9088214635849, + "learning_rate": 1.757859792185983e-05, + "loss": 3.2462, + "step": 6223 + }, + { + "epoch": 0.499007837084845, + "grad_norm": 1.8771028518676758, + "learning_rate": 1.7577750728452007e-05, + "loss": 3.2082, + "step": 6224 + }, + { + "epoch": 0.4990880118658676, + "grad_norm": 1.7855406999588013, + "learning_rate": 1.7576903407285977e-05, + "loss": 3.2187, + "step": 6225 + }, + { + "epoch": 0.4991681866468902, + "grad_norm": 1.9201288223266602, + "learning_rate": 1.7576055958376027e-05, + "loss": 3.1792, + "step": 6226 + }, + { + "epoch": 0.49924836142791285, + "grad_norm": 1.8289251327514648, + "learning_rate": 1.757520838173644e-05, + "loss": 3.2986, + "step": 6227 + }, + { + "epoch": 0.49932853620893547, + "grad_norm": 1.9236067533493042, + "learning_rate": 1.7574360677381514e-05, + "loss": 3.1714, + "step": 6228 + }, + { + "epoch": 0.4994087109899581, + "grad_norm": 1.851603388786316, + "learning_rate": 1.7573512845325532e-05, + "loss": 3.291, + "step": 6229 + }, + { + "epoch": 0.4994888857709807, + "grad_norm": 0.9313539862632751, + "learning_rate": 1.7572664885582796e-05, + "loss": 2.5254, + "step": 6230 + }, + { + "epoch": 0.49956906055200334, + "grad_norm": 1.8818678855895996, + "learning_rate": 1.75718167981676e-05, + "loss": 3.1836, + "step": 6231 + }, + { + "epoch": 0.499649235333026, + "grad_norm": 1.9171043634414673, + "learning_rate": 1.757096858309424e-05, + "loss": 3.3136, + "step": 6232 + }, + { + "epoch": 0.49972941011404864, + "grad_norm": 2.071608781814575, + "learning_rate": 1.7570120240377015e-05, + "loss": 3.1697, + "step": 6233 + }, + { + "epoch": 0.49980958489507127, + "grad_norm": 1.8557053804397583, + "learning_rate": 1.756927177003023e-05, + "loss": 3.2549, + "step": 6234 + }, + { + "epoch": 0.4998897596760939, + "grad_norm": 1.8270926475524902, + "learning_rate": 1.7568423172068196e-05, + "loss": 3.2581, + "step": 6235 + }, + { + "epoch": 0.4999699344571165, + "grad_norm": 2.005675792694092, + "learning_rate": 1.7567574446505212e-05, + "loss": 3.2186, + "step": 6236 + }, + { + "epoch": 0.5000501092381392, + "grad_norm": 1.7599395513534546, + "learning_rate": 1.7566725593355593e-05, + "loss": 3.2441, + "step": 6237 + }, + { + "epoch": 0.5001302840191618, + "grad_norm": 0.8062405586242676, + "learning_rate": 1.7565876612633643e-05, + "loss": 2.5125, + "step": 6238 + }, + { + "epoch": 0.5002104588001844, + "grad_norm": 1.9672623872756958, + "learning_rate": 1.756502750435368e-05, + "loss": 3.238, + "step": 6239 + }, + { + "epoch": 0.5002906335812071, + "grad_norm": 1.9656050205230713, + "learning_rate": 1.756417826853002e-05, + "loss": 3.2361, + "step": 6240 + }, + { + "epoch": 0.5003708083622297, + "grad_norm": 1.931512713432312, + "learning_rate": 1.7563328905176983e-05, + "loss": 3.3778, + "step": 6241 + }, + { + "epoch": 0.5004509831432523, + "grad_norm": 1.913759708404541, + "learning_rate": 1.7562479414308882e-05, + "loss": 3.213, + "step": 6242 + }, + { + "epoch": 0.5005311579242749, + "grad_norm": 1.8765555620193481, + "learning_rate": 1.7561629795940046e-05, + "loss": 3.1985, + "step": 6243 + }, + { + "epoch": 0.5006113327052976, + "grad_norm": 1.9493069648742676, + "learning_rate": 1.7560780050084796e-05, + "loss": 3.2848, + "step": 6244 + }, + { + "epoch": 0.5006915074863202, + "grad_norm": 1.8900833129882812, + "learning_rate": 1.755993017675746e-05, + "loss": 3.2101, + "step": 6245 + }, + { + "epoch": 0.5007716822673428, + "grad_norm": 1.9734801054000854, + "learning_rate": 1.7559080175972367e-05, + "loss": 3.2573, + "step": 6246 + }, + { + "epoch": 0.5008518570483654, + "grad_norm": 1.8849647045135498, + "learning_rate": 1.755823004774384e-05, + "loss": 3.1372, + "step": 6247 + }, + { + "epoch": 0.5009320318293881, + "grad_norm": 1.9099117517471313, + "learning_rate": 1.7557379792086223e-05, + "loss": 3.2936, + "step": 6248 + }, + { + "epoch": 0.5010122066104107, + "grad_norm": 2.0168709754943848, + "learning_rate": 1.7556529409013846e-05, + "loss": 3.3193, + "step": 6249 + }, + { + "epoch": 0.5010923813914333, + "grad_norm": 1.9353407621383667, + "learning_rate": 1.7555678898541046e-05, + "loss": 3.1611, + "step": 6250 + }, + { + "epoch": 0.5011725561724559, + "grad_norm": 2.105048656463623, + "learning_rate": 1.7554828260682165e-05, + "loss": 3.2557, + "step": 6251 + }, + { + "epoch": 0.5012527309534786, + "grad_norm": 1.89267897605896, + "learning_rate": 1.7553977495451538e-05, + "loss": 3.2933, + "step": 6252 + }, + { + "epoch": 0.5013329057345012, + "grad_norm": 1.757783055305481, + "learning_rate": 1.7553126602863514e-05, + "loss": 3.2282, + "step": 6253 + }, + { + "epoch": 0.5014130805155238, + "grad_norm": 1.7983520030975342, + "learning_rate": 1.7552275582932438e-05, + "loss": 3.3259, + "step": 6254 + }, + { + "epoch": 0.5014932552965464, + "grad_norm": 1.8569130897521973, + "learning_rate": 1.755142443567266e-05, + "loss": 3.1865, + "step": 6255 + }, + { + "epoch": 0.501573430077569, + "grad_norm": 2.0143463611602783, + "learning_rate": 1.755057316109852e-05, + "loss": 3.3543, + "step": 6256 + }, + { + "epoch": 0.5016536048585918, + "grad_norm": 1.7783458232879639, + "learning_rate": 1.7549721759224382e-05, + "loss": 3.2665, + "step": 6257 + }, + { + "epoch": 0.5017337796396144, + "grad_norm": 1.8628875017166138, + "learning_rate": 1.7548870230064596e-05, + "loss": 3.2401, + "step": 6258 + }, + { + "epoch": 0.501813954420637, + "grad_norm": 1.828325629234314, + "learning_rate": 1.7548018573633522e-05, + "loss": 3.2529, + "step": 6259 + }, + { + "epoch": 0.5018941292016597, + "grad_norm": 1.9708861112594604, + "learning_rate": 1.754716678994551e-05, + "loss": 3.2102, + "step": 6260 + }, + { + "epoch": 0.5019743039826823, + "grad_norm": 1.763461709022522, + "learning_rate": 1.7546314879014926e-05, + "loss": 3.2186, + "step": 6261 + }, + { + "epoch": 0.5020544787637049, + "grad_norm": 1.8921024799346924, + "learning_rate": 1.7545462840856135e-05, + "loss": 3.178, + "step": 6262 + }, + { + "epoch": 0.5021346535447275, + "grad_norm": 1.8325169086456299, + "learning_rate": 1.75446106754835e-05, + "loss": 3.1814, + "step": 6263 + }, + { + "epoch": 0.5022148283257502, + "grad_norm": 1.8855836391448975, + "learning_rate": 1.7543758382911386e-05, + "loss": 3.2624, + "step": 6264 + }, + { + "epoch": 0.5022950031067728, + "grad_norm": 1.8658117055892944, + "learning_rate": 1.7542905963154166e-05, + "loss": 3.181, + "step": 6265 + }, + { + "epoch": 0.5023751778877954, + "grad_norm": 1.893561840057373, + "learning_rate": 1.7542053416226207e-05, + "loss": 3.2179, + "step": 6266 + }, + { + "epoch": 0.502455352668818, + "grad_norm": 1.9968291521072388, + "learning_rate": 1.7541200742141887e-05, + "loss": 3.2323, + "step": 6267 + }, + { + "epoch": 0.5025355274498406, + "grad_norm": 0.8630470037460327, + "learning_rate": 1.754034794091558e-05, + "loss": 2.5021, + "step": 6268 + }, + { + "epoch": 0.5026157022308633, + "grad_norm": 1.8800861835479736, + "learning_rate": 1.7539495012561665e-05, + "loss": 3.2499, + "step": 6269 + }, + { + "epoch": 0.5026958770118859, + "grad_norm": 1.8496849536895752, + "learning_rate": 1.753864195709452e-05, + "loss": 3.2674, + "step": 6270 + }, + { + "epoch": 0.5027760517929085, + "grad_norm": 2.047450065612793, + "learning_rate": 1.753778877452853e-05, + "loss": 3.2786, + "step": 6271 + }, + { + "epoch": 0.5028562265739311, + "grad_norm": 1.8032245635986328, + "learning_rate": 1.753693546487808e-05, + "loss": 3.3171, + "step": 6272 + }, + { + "epoch": 0.5029364013549538, + "grad_norm": 0.832652747631073, + "learning_rate": 1.7536082028157547e-05, + "loss": 2.5032, + "step": 6273 + }, + { + "epoch": 0.5030165761359764, + "grad_norm": 1.803690791130066, + "learning_rate": 1.7535228464381332e-05, + "loss": 3.3149, + "step": 6274 + }, + { + "epoch": 0.503096750916999, + "grad_norm": 1.9771933555603027, + "learning_rate": 1.753437477356382e-05, + "loss": 3.3192, + "step": 6275 + }, + { + "epoch": 0.5031769256980216, + "grad_norm": 0.8245745897293091, + "learning_rate": 1.7533520955719406e-05, + "loss": 2.5082, + "step": 6276 + }, + { + "epoch": 0.5032571004790443, + "grad_norm": 1.823217511177063, + "learning_rate": 1.753266701086248e-05, + "loss": 3.2408, + "step": 6277 + }, + { + "epoch": 0.503337275260067, + "grad_norm": 1.9165204763412476, + "learning_rate": 1.7531812939007447e-05, + "loss": 3.3217, + "step": 6278 + }, + { + "epoch": 0.5034174500410896, + "grad_norm": 1.880742073059082, + "learning_rate": 1.7530958740168703e-05, + "loss": 3.2556, + "step": 6279 + }, + { + "epoch": 0.5034976248221122, + "grad_norm": 1.8521873950958252, + "learning_rate": 1.7530104414360648e-05, + "loss": 3.1865, + "step": 6280 + }, + { + "epoch": 0.5035777996031349, + "grad_norm": 1.8627125024795532, + "learning_rate": 1.7529249961597686e-05, + "loss": 3.1683, + "step": 6281 + }, + { + "epoch": 0.5036579743841575, + "grad_norm": 1.8096182346343994, + "learning_rate": 1.7528395381894223e-05, + "loss": 3.2694, + "step": 6282 + }, + { + "epoch": 0.5037381491651801, + "grad_norm": 1.9198033809661865, + "learning_rate": 1.7527540675264668e-05, + "loss": 3.2953, + "step": 6283 + }, + { + "epoch": 0.5038183239462027, + "grad_norm": 1.8799413442611694, + "learning_rate": 1.7526685841723428e-05, + "loss": 3.1238, + "step": 6284 + }, + { + "epoch": 0.5038984987272254, + "grad_norm": 1.7823777198791504, + "learning_rate": 1.7525830881284916e-05, + "loss": 3.2383, + "step": 6285 + }, + { + "epoch": 0.503978673508248, + "grad_norm": 1.9116601943969727, + "learning_rate": 1.7524975793963552e-05, + "loss": 3.3126, + "step": 6286 + }, + { + "epoch": 0.5040588482892706, + "grad_norm": 1.8213603496551514, + "learning_rate": 1.7524120579773746e-05, + "loss": 3.24, + "step": 6287 + }, + { + "epoch": 0.5041390230702932, + "grad_norm": 1.8166697025299072, + "learning_rate": 1.7523265238729917e-05, + "loss": 3.3466, + "step": 6288 + }, + { + "epoch": 0.5042191978513159, + "grad_norm": 0.8295219540596008, + "learning_rate": 1.7522409770846492e-05, + "loss": 2.4958, + "step": 6289 + }, + { + "epoch": 0.5042993726323385, + "grad_norm": 1.9864071607589722, + "learning_rate": 1.752155417613789e-05, + "loss": 3.1377, + "step": 6290 + }, + { + "epoch": 0.5043795474133611, + "grad_norm": 1.7639368772506714, + "learning_rate": 1.752069845461853e-05, + "loss": 3.2812, + "step": 6291 + }, + { + "epoch": 0.5044597221943837, + "grad_norm": 1.8857852220535278, + "learning_rate": 1.7519842606302842e-05, + "loss": 3.3121, + "step": 6292 + }, + { + "epoch": 0.5045398969754064, + "grad_norm": 0.8088487386703491, + "learning_rate": 1.7518986631205267e-05, + "loss": 2.5167, + "step": 6293 + }, + { + "epoch": 0.504620071756429, + "grad_norm": 1.857999324798584, + "learning_rate": 1.7518130529340223e-05, + "loss": 3.3462, + "step": 6294 + }, + { + "epoch": 0.5047002465374516, + "grad_norm": 1.842932939529419, + "learning_rate": 1.7517274300722146e-05, + "loss": 3.2562, + "step": 6295 + }, + { + "epoch": 0.5047804213184742, + "grad_norm": 1.7781565189361572, + "learning_rate": 1.7516417945365477e-05, + "loss": 3.2751, + "step": 6296 + }, + { + "epoch": 0.5048605960994969, + "grad_norm": 1.9282777309417725, + "learning_rate": 1.751556146328465e-05, + "loss": 3.2407, + "step": 6297 + }, + { + "epoch": 0.5049407708805196, + "grad_norm": 1.9615166187286377, + "learning_rate": 1.7514704854494106e-05, + "loss": 3.2451, + "step": 6298 + }, + { + "epoch": 0.5050209456615422, + "grad_norm": 1.786456823348999, + "learning_rate": 1.7513848119008285e-05, + "loss": 3.2451, + "step": 6299 + }, + { + "epoch": 0.5051011204425648, + "grad_norm": 1.872410774230957, + "learning_rate": 1.7512991256841632e-05, + "loss": 3.1841, + "step": 6300 + }, + { + "epoch": 0.5051812952235875, + "grad_norm": 0.8884310126304626, + "learning_rate": 1.7512134268008595e-05, + "loss": 2.5153, + "step": 6301 + }, + { + "epoch": 0.5052614700046101, + "grad_norm": 1.940035343170166, + "learning_rate": 1.7511277152523624e-05, + "loss": 3.2943, + "step": 6302 + }, + { + "epoch": 0.5053416447856327, + "grad_norm": 1.7485692501068115, + "learning_rate": 1.7510419910401166e-05, + "loss": 3.2604, + "step": 6303 + }, + { + "epoch": 0.5054218195666553, + "grad_norm": 1.9978002309799194, + "learning_rate": 1.7509562541655674e-05, + "loss": 3.276, + "step": 6304 + }, + { + "epoch": 0.505501994347678, + "grad_norm": 1.8459959030151367, + "learning_rate": 1.7508705046301604e-05, + "loss": 3.1345, + "step": 6305 + }, + { + "epoch": 0.5055821691287006, + "grad_norm": 1.9094563722610474, + "learning_rate": 1.7507847424353414e-05, + "loss": 3.3499, + "step": 6306 + }, + { + "epoch": 0.5056623439097232, + "grad_norm": 2.0459625720977783, + "learning_rate": 1.7506989675825563e-05, + "loss": 3.2968, + "step": 6307 + }, + { + "epoch": 0.5057425186907458, + "grad_norm": 2.0567808151245117, + "learning_rate": 1.750613180073251e-05, + "loss": 3.1987, + "step": 6308 + }, + { + "epoch": 0.5058226934717684, + "grad_norm": 1.9081714153289795, + "learning_rate": 1.7505273799088724e-05, + "loss": 3.2584, + "step": 6309 + }, + { + "epoch": 0.5059028682527911, + "grad_norm": 1.8042389154434204, + "learning_rate": 1.7504415670908664e-05, + "loss": 3.252, + "step": 6310 + }, + { + "epoch": 0.5059830430338137, + "grad_norm": 1.7693456411361694, + "learning_rate": 1.75035574162068e-05, + "loss": 3.2375, + "step": 6311 + }, + { + "epoch": 0.5060632178148363, + "grad_norm": 1.8058252334594727, + "learning_rate": 1.7502699034997604e-05, + "loss": 3.3176, + "step": 6312 + }, + { + "epoch": 0.5061433925958589, + "grad_norm": 1.794380784034729, + "learning_rate": 1.7501840527295545e-05, + "loss": 3.2316, + "step": 6313 + }, + { + "epoch": 0.5062235673768816, + "grad_norm": 1.9709324836730957, + "learning_rate": 1.7500981893115103e-05, + "loss": 3.3387, + "step": 6314 + }, + { + "epoch": 0.5063037421579042, + "grad_norm": 1.7906922101974487, + "learning_rate": 1.7500123132470747e-05, + "loss": 3.1951, + "step": 6315 + }, + { + "epoch": 0.5063839169389268, + "grad_norm": 1.834766149520874, + "learning_rate": 1.7499264245376956e-05, + "loss": 3.2373, + "step": 6316 + }, + { + "epoch": 0.5064640917199494, + "grad_norm": 1.9679237604141235, + "learning_rate": 1.7498405231848217e-05, + "loss": 3.2384, + "step": 6317 + }, + { + "epoch": 0.5065442665009722, + "grad_norm": 2.0485968589782715, + "learning_rate": 1.7497546091899005e-05, + "loss": 3.3504, + "step": 6318 + }, + { + "epoch": 0.5066244412819948, + "grad_norm": 1.836122989654541, + "learning_rate": 1.749668682554381e-05, + "loss": 3.2213, + "step": 6319 + }, + { + "epoch": 0.5067046160630174, + "grad_norm": 1.8396549224853516, + "learning_rate": 1.7495827432797115e-05, + "loss": 3.1909, + "step": 6320 + }, + { + "epoch": 0.50678479084404, + "grad_norm": 0.7904005646705627, + "learning_rate": 1.7494967913673414e-05, + "loss": 2.5027, + "step": 6321 + }, + { + "epoch": 0.5068649656250627, + "grad_norm": 1.7570130825042725, + "learning_rate": 1.7494108268187197e-05, + "loss": 3.1808, + "step": 6322 + }, + { + "epoch": 0.5069451404060853, + "grad_norm": 1.8009612560272217, + "learning_rate": 1.749324849635295e-05, + "loss": 3.2605, + "step": 6323 + }, + { + "epoch": 0.5070253151871079, + "grad_norm": 2.0053932666778564, + "learning_rate": 1.749238859818518e-05, + "loss": 3.1937, + "step": 6324 + }, + { + "epoch": 0.5071054899681305, + "grad_norm": 1.8361058235168457, + "learning_rate": 1.7491528573698376e-05, + "loss": 3.2883, + "step": 6325 + }, + { + "epoch": 0.5071856647491532, + "grad_norm": 1.8906619548797607, + "learning_rate": 1.749066842290704e-05, + "loss": 3.1703, + "step": 6326 + }, + { + "epoch": 0.5072658395301758, + "grad_norm": 2.1914446353912354, + "learning_rate": 1.748980814582567e-05, + "loss": 3.3549, + "step": 6327 + }, + { + "epoch": 0.5073460143111984, + "grad_norm": 1.9743212461471558, + "learning_rate": 1.7488947742468783e-05, + "loss": 3.2364, + "step": 6328 + }, + { + "epoch": 0.507426189092221, + "grad_norm": 2.032705545425415, + "learning_rate": 1.7488087212850873e-05, + "loss": 3.158, + "step": 6329 + }, + { + "epoch": 0.5075063638732437, + "grad_norm": 1.8885126113891602, + "learning_rate": 1.7487226556986452e-05, + "loss": 3.2986, + "step": 6330 + }, + { + "epoch": 0.5075865386542663, + "grad_norm": 1.7443692684173584, + "learning_rate": 1.748636577489003e-05, + "loss": 3.1443, + "step": 6331 + }, + { + "epoch": 0.5076667134352889, + "grad_norm": 1.7898887395858765, + "learning_rate": 1.748550486657612e-05, + "loss": 3.2844, + "step": 6332 + }, + { + "epoch": 0.5077468882163115, + "grad_norm": 0.8203292489051819, + "learning_rate": 1.7484643832059234e-05, + "loss": 2.4971, + "step": 6333 + }, + { + "epoch": 0.5078270629973342, + "grad_norm": 1.8156673908233643, + "learning_rate": 1.7483782671353893e-05, + "loss": 3.2439, + "step": 6334 + }, + { + "epoch": 0.5079072377783568, + "grad_norm": 1.7692488431930542, + "learning_rate": 1.7482921384474613e-05, + "loss": 3.2563, + "step": 6335 + }, + { + "epoch": 0.5079874125593794, + "grad_norm": 1.895326018333435, + "learning_rate": 1.7482059971435915e-05, + "loss": 3.2876, + "step": 6336 + }, + { + "epoch": 0.508067587340402, + "grad_norm": 1.8507047891616821, + "learning_rate": 1.7481198432252323e-05, + "loss": 3.3165, + "step": 6337 + }, + { + "epoch": 0.5081477621214247, + "grad_norm": 1.950644612312317, + "learning_rate": 1.7480336766938363e-05, + "loss": 3.1653, + "step": 6338 + }, + { + "epoch": 0.5082279369024474, + "grad_norm": 1.8173744678497314, + "learning_rate": 1.747947497550856e-05, + "loss": 3.207, + "step": 6339 + }, + { + "epoch": 0.50830811168347, + "grad_norm": 1.9937430620193481, + "learning_rate": 1.7478613057977445e-05, + "loss": 3.224, + "step": 6340 + }, + { + "epoch": 0.5083882864644926, + "grad_norm": 1.7827863693237305, + "learning_rate": 1.7477751014359553e-05, + "loss": 3.2939, + "step": 6341 + }, + { + "epoch": 0.5084684612455153, + "grad_norm": 2.0235729217529297, + "learning_rate": 1.747688884466941e-05, + "loss": 3.2961, + "step": 6342 + }, + { + "epoch": 0.5085486360265379, + "grad_norm": 0.8590136766433716, + "learning_rate": 1.747602654892156e-05, + "loss": 2.5029, + "step": 6343 + }, + { + "epoch": 0.5086288108075605, + "grad_norm": 1.8581393957138062, + "learning_rate": 1.747516412713053e-05, + "loss": 3.2224, + "step": 6344 + }, + { + "epoch": 0.5087089855885831, + "grad_norm": 2.060577630996704, + "learning_rate": 1.7474301579310873e-05, + "loss": 3.1388, + "step": 6345 + }, + { + "epoch": 0.5087891603696058, + "grad_norm": 1.8753044605255127, + "learning_rate": 1.7473438905477128e-05, + "loss": 3.1978, + "step": 6346 + }, + { + "epoch": 0.5088693351506284, + "grad_norm": 1.8740264177322388, + "learning_rate": 1.747257610564383e-05, + "loss": 3.2535, + "step": 6347 + }, + { + "epoch": 0.508949509931651, + "grad_norm": 2.1214044094085693, + "learning_rate": 1.7471713179825538e-05, + "loss": 3.3738, + "step": 6348 + }, + { + "epoch": 0.5090296847126736, + "grad_norm": 1.9150819778442383, + "learning_rate": 1.7470850128036794e-05, + "loss": 3.2802, + "step": 6349 + }, + { + "epoch": 0.5091098594936962, + "grad_norm": 1.9442251920700073, + "learning_rate": 1.7469986950292148e-05, + "loss": 3.199, + "step": 6350 + }, + { + "epoch": 0.5091900342747189, + "grad_norm": 1.835118293762207, + "learning_rate": 1.7469123646606156e-05, + "loss": 3.1727, + "step": 6351 + }, + { + "epoch": 0.5092702090557415, + "grad_norm": 1.7628315687179565, + "learning_rate": 1.7468260216993372e-05, + "loss": 3.2909, + "step": 6352 + }, + { + "epoch": 0.5093503838367641, + "grad_norm": 1.9355155229568481, + "learning_rate": 1.746739666146835e-05, + "loss": 3.3099, + "step": 6353 + }, + { + "epoch": 0.5094305586177867, + "grad_norm": 1.9230923652648926, + "learning_rate": 1.7466532980045655e-05, + "loss": 3.2539, + "step": 6354 + }, + { + "epoch": 0.5095107333988094, + "grad_norm": 1.9404960870742798, + "learning_rate": 1.7465669172739847e-05, + "loss": 3.271, + "step": 6355 + }, + { + "epoch": 0.509590908179832, + "grad_norm": 1.9439923763275146, + "learning_rate": 1.7464805239565484e-05, + "loss": 3.1823, + "step": 6356 + }, + { + "epoch": 0.5096710829608546, + "grad_norm": 1.955227255821228, + "learning_rate": 1.7463941180537137e-05, + "loss": 3.2186, + "step": 6357 + }, + { + "epoch": 0.5097512577418772, + "grad_norm": 1.7939947843551636, + "learning_rate": 1.7463076995669374e-05, + "loss": 3.1942, + "step": 6358 + }, + { + "epoch": 0.5098314325229, + "grad_norm": 1.8223862648010254, + "learning_rate": 1.746221268497676e-05, + "loss": 3.2815, + "step": 6359 + }, + { + "epoch": 0.5099116073039226, + "grad_norm": 2.2710554599761963, + "learning_rate": 1.7461348248473872e-05, + "loss": 3.2324, + "step": 6360 + }, + { + "epoch": 0.5099917820849452, + "grad_norm": 1.9359359741210938, + "learning_rate": 1.7460483686175282e-05, + "loss": 3.3216, + "step": 6361 + }, + { + "epoch": 0.5100719568659678, + "grad_norm": 1.8770948648452759, + "learning_rate": 1.7459618998095565e-05, + "loss": 3.2285, + "step": 6362 + }, + { + "epoch": 0.5101521316469905, + "grad_norm": 1.8252859115600586, + "learning_rate": 1.7458754184249304e-05, + "loss": 3.2081, + "step": 6363 + }, + { + "epoch": 0.5102323064280131, + "grad_norm": 1.8443176746368408, + "learning_rate": 1.7457889244651075e-05, + "loss": 3.3402, + "step": 6364 + }, + { + "epoch": 0.5103124812090357, + "grad_norm": 2.004537343978882, + "learning_rate": 1.745702417931546e-05, + "loss": 3.1886, + "step": 6365 + }, + { + "epoch": 0.5103926559900583, + "grad_norm": 1.935875654220581, + "learning_rate": 1.7456158988257045e-05, + "loss": 3.2659, + "step": 6366 + }, + { + "epoch": 0.510472830771081, + "grad_norm": 1.8199962377548218, + "learning_rate": 1.745529367149042e-05, + "loss": 3.2321, + "step": 6367 + }, + { + "epoch": 0.5105530055521036, + "grad_norm": 1.8619617223739624, + "learning_rate": 1.7454428229030167e-05, + "loss": 3.2188, + "step": 6368 + }, + { + "epoch": 0.5106331803331262, + "grad_norm": 1.9819985628128052, + "learning_rate": 1.7453562660890885e-05, + "loss": 3.0806, + "step": 6369 + }, + { + "epoch": 0.5107133551141488, + "grad_norm": 1.7742887735366821, + "learning_rate": 1.7452696967087164e-05, + "loss": 3.1274, + "step": 6370 + }, + { + "epoch": 0.5107935298951715, + "grad_norm": 1.8935383558273315, + "learning_rate": 1.7451831147633597e-05, + "loss": 3.2576, + "step": 6371 + }, + { + "epoch": 0.5108737046761941, + "grad_norm": 0.836359977722168, + "learning_rate": 1.7450965202544782e-05, + "loss": 2.5034, + "step": 6372 + }, + { + "epoch": 0.5109538794572167, + "grad_norm": 1.9155675172805786, + "learning_rate": 1.745009913183532e-05, + "loss": 3.3531, + "step": 6373 + }, + { + "epoch": 0.5110340542382393, + "grad_norm": 1.8811596632003784, + "learning_rate": 1.7449232935519816e-05, + "loss": 3.2396, + "step": 6374 + }, + { + "epoch": 0.511114229019262, + "grad_norm": 1.9329478740692139, + "learning_rate": 1.7448366613612867e-05, + "loss": 3.3597, + "step": 6375 + }, + { + "epoch": 0.5111944038002846, + "grad_norm": 1.856134057044983, + "learning_rate": 1.7447500166129082e-05, + "loss": 3.1974, + "step": 6376 + }, + { + "epoch": 0.5112745785813072, + "grad_norm": 1.909629464149475, + "learning_rate": 1.7446633593083068e-05, + "loss": 3.1659, + "step": 6377 + }, + { + "epoch": 0.5113547533623298, + "grad_norm": 1.8379487991333008, + "learning_rate": 1.7445766894489437e-05, + "loss": 3.2083, + "step": 6378 + }, + { + "epoch": 0.5114349281433526, + "grad_norm": 1.8145389556884766, + "learning_rate": 1.74449000703628e-05, + "loss": 3.2716, + "step": 6379 + }, + { + "epoch": 0.5115151029243752, + "grad_norm": 1.8265864849090576, + "learning_rate": 1.7444033120717775e-05, + "loss": 3.2123, + "step": 6380 + }, + { + "epoch": 0.5115952777053978, + "grad_norm": 1.8538854122161865, + "learning_rate": 1.744316604556897e-05, + "loss": 3.1728, + "step": 6381 + }, + { + "epoch": 0.5116754524864204, + "grad_norm": 1.7943209409713745, + "learning_rate": 1.7442298844931013e-05, + "loss": 3.1707, + "step": 6382 + }, + { + "epoch": 0.5117556272674431, + "grad_norm": 1.8301604986190796, + "learning_rate": 1.7441431518818516e-05, + "loss": 3.2352, + "step": 6383 + }, + { + "epoch": 0.5118358020484657, + "grad_norm": 1.7233070135116577, + "learning_rate": 1.744056406724611e-05, + "loss": 3.1862, + "step": 6384 + }, + { + "epoch": 0.5119159768294883, + "grad_norm": 1.9313150644302368, + "learning_rate": 1.7439696490228413e-05, + "loss": 3.2128, + "step": 6385 + }, + { + "epoch": 0.5119961516105109, + "grad_norm": 1.7817045450210571, + "learning_rate": 1.7438828787780058e-05, + "loss": 3.2022, + "step": 6386 + }, + { + "epoch": 0.5120763263915336, + "grad_norm": 1.852297306060791, + "learning_rate": 1.7437960959915668e-05, + "loss": 3.3112, + "step": 6387 + }, + { + "epoch": 0.5121565011725562, + "grad_norm": 0.846350908279419, + "learning_rate": 1.743709300664988e-05, + "loss": 2.49, + "step": 6388 + }, + { + "epoch": 0.5122366759535788, + "grad_norm": 1.775861382484436, + "learning_rate": 1.7436224927997327e-05, + "loss": 3.2132, + "step": 6389 + }, + { + "epoch": 0.5123168507346014, + "grad_norm": 1.7570888996124268, + "learning_rate": 1.7435356723972642e-05, + "loss": 3.2692, + "step": 6390 + }, + { + "epoch": 0.512397025515624, + "grad_norm": 1.8811575174331665, + "learning_rate": 1.743448839459046e-05, + "loss": 3.2381, + "step": 6391 + }, + { + "epoch": 0.5124772002966467, + "grad_norm": 1.8621864318847656, + "learning_rate": 1.7433619939865424e-05, + "loss": 3.3397, + "step": 6392 + }, + { + "epoch": 0.5125573750776693, + "grad_norm": 1.9431030750274658, + "learning_rate": 1.7432751359812175e-05, + "loss": 3.2605, + "step": 6393 + }, + { + "epoch": 0.5126375498586919, + "grad_norm": 0.7755051255226135, + "learning_rate": 1.7431882654445362e-05, + "loss": 2.5032, + "step": 6394 + }, + { + "epoch": 0.5127177246397145, + "grad_norm": 1.7433106899261475, + "learning_rate": 1.7431013823779625e-05, + "loss": 3.2773, + "step": 6395 + }, + { + "epoch": 0.5127978994207372, + "grad_norm": 1.813740611076355, + "learning_rate": 1.7430144867829613e-05, + "loss": 3.306, + "step": 6396 + }, + { + "epoch": 0.5128780742017598, + "grad_norm": 1.8208866119384766, + "learning_rate": 1.7429275786609978e-05, + "loss": 3.2184, + "step": 6397 + }, + { + "epoch": 0.5129582489827824, + "grad_norm": 1.8863407373428345, + "learning_rate": 1.742840658013537e-05, + "loss": 3.2865, + "step": 6398 + }, + { + "epoch": 0.513038423763805, + "grad_norm": 1.9435381889343262, + "learning_rate": 1.742753724842045e-05, + "loss": 3.1917, + "step": 6399 + }, + { + "epoch": 0.5131185985448278, + "grad_norm": 1.86409330368042, + "learning_rate": 1.742666779147986e-05, + "loss": 3.1321, + "step": 6400 + }, + { + "epoch": 0.5131987733258504, + "grad_norm": 1.7622212171554565, + "learning_rate": 1.742579820932828e-05, + "loss": 3.1953, + "step": 6401 + }, + { + "epoch": 0.513278948106873, + "grad_norm": 1.7538117170333862, + "learning_rate": 1.742492850198035e-05, + "loss": 3.2098, + "step": 6402 + }, + { + "epoch": 0.5133591228878956, + "grad_norm": 1.7537471055984497, + "learning_rate": 1.7424058669450745e-05, + "loss": 3.2277, + "step": 6403 + }, + { + "epoch": 0.5134392976689183, + "grad_norm": 1.7957022190093994, + "learning_rate": 1.742318871175413e-05, + "loss": 3.2648, + "step": 6404 + }, + { + "epoch": 0.5135194724499409, + "grad_norm": 1.8198277950286865, + "learning_rate": 1.7422318628905165e-05, + "loss": 3.1991, + "step": 6405 + }, + { + "epoch": 0.5135996472309635, + "grad_norm": 0.7953027486801147, + "learning_rate": 1.742144842091853e-05, + "loss": 2.5165, + "step": 6406 + }, + { + "epoch": 0.5136798220119861, + "grad_norm": 1.8700172901153564, + "learning_rate": 1.7420578087808884e-05, + "loss": 3.2158, + "step": 6407 + }, + { + "epoch": 0.5137599967930088, + "grad_norm": 1.9957594871520996, + "learning_rate": 1.741970762959091e-05, + "loss": 3.3625, + "step": 6408 + }, + { + "epoch": 0.5138401715740314, + "grad_norm": 1.769639015197754, + "learning_rate": 1.741883704627928e-05, + "loss": 3.2424, + "step": 6409 + }, + { + "epoch": 0.513920346355054, + "grad_norm": 1.8660271167755127, + "learning_rate": 1.741796633788867e-05, + "loss": 3.2198, + "step": 6410 + }, + { + "epoch": 0.5140005211360766, + "grad_norm": 1.89810049533844, + "learning_rate": 1.7417095504433764e-05, + "loss": 3.1248, + "step": 6411 + }, + { + "epoch": 0.5140806959170993, + "grad_norm": 0.7873974442481995, + "learning_rate": 1.741622454592924e-05, + "loss": 2.501, + "step": 6412 + }, + { + "epoch": 0.5141608706981219, + "grad_norm": 1.8545639514923096, + "learning_rate": 1.7415353462389788e-05, + "loss": 3.213, + "step": 6413 + }, + { + "epoch": 0.5142410454791445, + "grad_norm": 2.1315770149230957, + "learning_rate": 1.7414482253830084e-05, + "loss": 3.2384, + "step": 6414 + }, + { + "epoch": 0.5143212202601671, + "grad_norm": 0.7960556149482727, + "learning_rate": 1.7413610920264827e-05, + "loss": 2.5162, + "step": 6415 + }, + { + "epoch": 0.5144013950411898, + "grad_norm": 1.9557931423187256, + "learning_rate": 1.74127394617087e-05, + "loss": 3.2252, + "step": 6416 + }, + { + "epoch": 0.5144815698222124, + "grad_norm": 1.8029160499572754, + "learning_rate": 1.7411867878176402e-05, + "loss": 3.1838, + "step": 6417 + }, + { + "epoch": 0.514561744603235, + "grad_norm": 1.9298032522201538, + "learning_rate": 1.7410996169682623e-05, + "loss": 3.2791, + "step": 6418 + }, + { + "epoch": 0.5146419193842576, + "grad_norm": 1.9603694677352905, + "learning_rate": 1.7410124336242055e-05, + "loss": 3.2787, + "step": 6419 + }, + { + "epoch": 0.5147220941652804, + "grad_norm": 1.827584981918335, + "learning_rate": 1.740925237786941e-05, + "loss": 3.1072, + "step": 6420 + }, + { + "epoch": 0.514802268946303, + "grad_norm": 1.7844436168670654, + "learning_rate": 1.7408380294579373e-05, + "loss": 3.1761, + "step": 6421 + }, + { + "epoch": 0.5148824437273256, + "grad_norm": 1.9156897068023682, + "learning_rate": 1.7407508086386662e-05, + "loss": 3.2867, + "step": 6422 + }, + { + "epoch": 0.5149626185083482, + "grad_norm": 0.817760169506073, + "learning_rate": 1.7406635753305973e-05, + "loss": 2.5013, + "step": 6423 + }, + { + "epoch": 0.5150427932893709, + "grad_norm": 1.8829964399337769, + "learning_rate": 1.7405763295352014e-05, + "loss": 3.1914, + "step": 6424 + }, + { + "epoch": 0.5151229680703935, + "grad_norm": 1.7670111656188965, + "learning_rate": 1.7404890712539497e-05, + "loss": 3.1944, + "step": 6425 + }, + { + "epoch": 0.5152031428514161, + "grad_norm": 1.8129123449325562, + "learning_rate": 1.7404018004883132e-05, + "loss": 3.2011, + "step": 6426 + }, + { + "epoch": 0.5152833176324387, + "grad_norm": 1.8490099906921387, + "learning_rate": 1.740314517239763e-05, + "loss": 3.2131, + "step": 6427 + }, + { + "epoch": 0.5153634924134614, + "grad_norm": 1.9975968599319458, + "learning_rate": 1.7402272215097716e-05, + "loss": 3.1921, + "step": 6428 + }, + { + "epoch": 0.515443667194484, + "grad_norm": 1.8064870834350586, + "learning_rate": 1.7401399132998098e-05, + "loss": 3.1948, + "step": 6429 + }, + { + "epoch": 0.5155238419755066, + "grad_norm": 0.7521665692329407, + "learning_rate": 1.7400525926113497e-05, + "loss": 2.4984, + "step": 6430 + }, + { + "epoch": 0.5156040167565292, + "grad_norm": 2.0013952255249023, + "learning_rate": 1.7399652594458643e-05, + "loss": 3.2397, + "step": 6431 + }, + { + "epoch": 0.5156841915375519, + "grad_norm": 1.7506784200668335, + "learning_rate": 1.7398779138048247e-05, + "loss": 3.2087, + "step": 6432 + }, + { + "epoch": 0.5157643663185745, + "grad_norm": 1.9642421007156372, + "learning_rate": 1.739790555689705e-05, + "loss": 3.3477, + "step": 6433 + }, + { + "epoch": 0.5158445410995971, + "grad_norm": 1.8096038103103638, + "learning_rate": 1.7397031851019766e-05, + "loss": 3.2424, + "step": 6434 + }, + { + "epoch": 0.5159247158806197, + "grad_norm": 1.8596477508544922, + "learning_rate": 1.7396158020431136e-05, + "loss": 3.2299, + "step": 6435 + }, + { + "epoch": 0.5160048906616423, + "grad_norm": 1.8140789270401, + "learning_rate": 1.7395284065145882e-05, + "loss": 3.2217, + "step": 6436 + }, + { + "epoch": 0.516085065442665, + "grad_norm": 1.8465077877044678, + "learning_rate": 1.7394409985178754e-05, + "loss": 3.2032, + "step": 6437 + }, + { + "epoch": 0.5161652402236876, + "grad_norm": 1.9379003047943115, + "learning_rate": 1.7393535780544474e-05, + "loss": 3.2913, + "step": 6438 + }, + { + "epoch": 0.5162454150047102, + "grad_norm": 1.9550907611846924, + "learning_rate": 1.739266145125779e-05, + "loss": 3.2495, + "step": 6439 + }, + { + "epoch": 0.516325589785733, + "grad_norm": 1.901697039604187, + "learning_rate": 1.7391786997333435e-05, + "loss": 3.2657, + "step": 6440 + }, + { + "epoch": 0.5164057645667556, + "grad_norm": 1.8199350833892822, + "learning_rate": 1.739091241878616e-05, + "loss": 3.3015, + "step": 6441 + }, + { + "epoch": 0.5164859393477782, + "grad_norm": 1.853655219078064, + "learning_rate": 1.7390037715630702e-05, + "loss": 3.2591, + "step": 6442 + }, + { + "epoch": 0.5165661141288008, + "grad_norm": 1.8920981884002686, + "learning_rate": 1.7389162887881815e-05, + "loss": 3.2511, + "step": 6443 + }, + { + "epoch": 0.5166462889098234, + "grad_norm": 1.9464902877807617, + "learning_rate": 1.738828793555425e-05, + "loss": 3.1924, + "step": 6444 + }, + { + "epoch": 0.5167264636908461, + "grad_norm": 1.8693032264709473, + "learning_rate": 1.738741285866275e-05, + "loss": 3.2839, + "step": 6445 + }, + { + "epoch": 0.5168066384718687, + "grad_norm": 1.7128584384918213, + "learning_rate": 1.7386537657222074e-05, + "loss": 3.2828, + "step": 6446 + }, + { + "epoch": 0.5168868132528913, + "grad_norm": 1.8265279531478882, + "learning_rate": 1.7385662331246974e-05, + "loss": 3.2687, + "step": 6447 + }, + { + "epoch": 0.5169669880339139, + "grad_norm": 1.9222315549850464, + "learning_rate": 1.7384786880752213e-05, + "loss": 3.1198, + "step": 6448 + }, + { + "epoch": 0.5170471628149366, + "grad_norm": 1.8569566011428833, + "learning_rate": 1.738391130575255e-05, + "loss": 3.163, + "step": 6449 + }, + { + "epoch": 0.5171273375959592, + "grad_norm": 1.7990738153457642, + "learning_rate": 1.738303560626274e-05, + "loss": 3.2363, + "step": 6450 + }, + { + "epoch": 0.5172075123769818, + "grad_norm": 1.8541796207427979, + "learning_rate": 1.7382159782297552e-05, + "loss": 3.1688, + "step": 6451 + }, + { + "epoch": 0.5172876871580044, + "grad_norm": 1.9598963260650635, + "learning_rate": 1.7381283833871755e-05, + "loss": 3.192, + "step": 6452 + }, + { + "epoch": 0.5173678619390271, + "grad_norm": 2.0421323776245117, + "learning_rate": 1.7380407761000112e-05, + "loss": 3.1901, + "step": 6453 + }, + { + "epoch": 0.5174480367200497, + "grad_norm": 1.7491822242736816, + "learning_rate": 1.7379531563697397e-05, + "loss": 3.1552, + "step": 6454 + }, + { + "epoch": 0.5175282115010723, + "grad_norm": 1.7281725406646729, + "learning_rate": 1.7378655241978374e-05, + "loss": 3.3076, + "step": 6455 + }, + { + "epoch": 0.5176083862820949, + "grad_norm": 1.8192460536956787, + "learning_rate": 1.737777879585783e-05, + "loss": 3.1877, + "step": 6456 + }, + { + "epoch": 0.5176885610631176, + "grad_norm": 1.8269238471984863, + "learning_rate": 1.7376902225350536e-05, + "loss": 3.2446, + "step": 6457 + }, + { + "epoch": 0.5177687358441402, + "grad_norm": 1.8279281854629517, + "learning_rate": 1.737602553047127e-05, + "loss": 3.2335, + "step": 6458 + }, + { + "epoch": 0.5178489106251628, + "grad_norm": 1.858662724494934, + "learning_rate": 1.737514871123481e-05, + "loss": 3.2155, + "step": 6459 + }, + { + "epoch": 0.5179290854061854, + "grad_norm": 1.876671314239502, + "learning_rate": 1.7374271767655946e-05, + "loss": 3.283, + "step": 6460 + }, + { + "epoch": 0.5180092601872082, + "grad_norm": 2.0166969299316406, + "learning_rate": 1.7373394699749458e-05, + "loss": 3.2646, + "step": 6461 + }, + { + "epoch": 0.5180894349682308, + "grad_norm": 1.7830729484558105, + "learning_rate": 1.7372517507530134e-05, + "loss": 3.1865, + "step": 6462 + }, + { + "epoch": 0.5181696097492534, + "grad_norm": 1.8068605661392212, + "learning_rate": 1.737164019101276e-05, + "loss": 3.2666, + "step": 6463 + }, + { + "epoch": 0.518249784530276, + "grad_norm": 1.8820604085922241, + "learning_rate": 1.7370762750212134e-05, + "loss": 3.2758, + "step": 6464 + }, + { + "epoch": 0.5183299593112987, + "grad_norm": 1.8541797399520874, + "learning_rate": 1.7369885185143042e-05, + "loss": 3.1724, + "step": 6465 + }, + { + "epoch": 0.5184101340923213, + "grad_norm": 1.8705922365188599, + "learning_rate": 1.736900749582029e-05, + "loss": 3.2558, + "step": 6466 + }, + { + "epoch": 0.5184903088733439, + "grad_norm": 1.9424620866775513, + "learning_rate": 1.7368129682258662e-05, + "loss": 3.3046, + "step": 6467 + }, + { + "epoch": 0.5185704836543665, + "grad_norm": 1.87410569190979, + "learning_rate": 1.7367251744472966e-05, + "loss": 3.2827, + "step": 6468 + }, + { + "epoch": 0.5186506584353892, + "grad_norm": 1.9772515296936035, + "learning_rate": 1.7366373682478003e-05, + "loss": 3.3545, + "step": 6469 + }, + { + "epoch": 0.5187308332164118, + "grad_norm": 1.7559434175491333, + "learning_rate": 1.7365495496288575e-05, + "loss": 3.2844, + "step": 6470 + }, + { + "epoch": 0.5188110079974344, + "grad_norm": 1.789790153503418, + "learning_rate": 1.736461718591949e-05, + "loss": 3.1522, + "step": 6471 + }, + { + "epoch": 0.518891182778457, + "grad_norm": 1.8114606142044067, + "learning_rate": 1.7363738751385555e-05, + "loss": 3.2289, + "step": 6472 + }, + { + "epoch": 0.5189713575594797, + "grad_norm": 0.7712846398353577, + "learning_rate": 1.7362860192701578e-05, + "loss": 2.5144, + "step": 6473 + }, + { + "epoch": 0.5190515323405023, + "grad_norm": 1.84572434425354, + "learning_rate": 1.7361981509882374e-05, + "loss": 3.2147, + "step": 6474 + }, + { + "epoch": 0.5191317071215249, + "grad_norm": 1.7689977884292603, + "learning_rate": 1.7361102702942757e-05, + "loss": 3.1505, + "step": 6475 + }, + { + "epoch": 0.5192118819025475, + "grad_norm": 1.8463261127471924, + "learning_rate": 1.7360223771897543e-05, + "loss": 3.2977, + "step": 6476 + }, + { + "epoch": 0.5192920566835701, + "grad_norm": 1.8566639423370361, + "learning_rate": 1.735934471676155e-05, + "loss": 3.1894, + "step": 6477 + }, + { + "epoch": 0.5193722314645928, + "grad_norm": 1.869457483291626, + "learning_rate": 1.73584655375496e-05, + "loss": 3.1367, + "step": 6478 + }, + { + "epoch": 0.5194524062456154, + "grad_norm": 1.8166086673736572, + "learning_rate": 1.7357586234276513e-05, + "loss": 3.2083, + "step": 6479 + }, + { + "epoch": 0.519532581026638, + "grad_norm": 1.7484002113342285, + "learning_rate": 1.735670680695711e-05, + "loss": 3.2408, + "step": 6480 + }, + { + "epoch": 0.5196127558076608, + "grad_norm": 1.8715457916259766, + "learning_rate": 1.7355827255606227e-05, + "loss": 3.2213, + "step": 6481 + }, + { + "epoch": 0.5196929305886834, + "grad_norm": 1.8202437162399292, + "learning_rate": 1.735494758023869e-05, + "loss": 3.2314, + "step": 6482 + }, + { + "epoch": 0.519773105369706, + "grad_norm": 1.8082261085510254, + "learning_rate": 1.735406778086933e-05, + "loss": 3.2767, + "step": 6483 + }, + { + "epoch": 0.5198532801507286, + "grad_norm": 1.7888091802597046, + "learning_rate": 1.7353187857512977e-05, + "loss": 3.1371, + "step": 6484 + }, + { + "epoch": 0.5199334549317512, + "grad_norm": 1.8199495077133179, + "learning_rate": 1.735230781018447e-05, + "loss": 3.1665, + "step": 6485 + }, + { + "epoch": 0.5200136297127739, + "grad_norm": 1.8937186002731323, + "learning_rate": 1.7351427638898644e-05, + "loss": 3.2941, + "step": 6486 + }, + { + "epoch": 0.5200938044937965, + "grad_norm": 2.044194221496582, + "learning_rate": 1.7350547343670338e-05, + "loss": 3.2507, + "step": 6487 + }, + { + "epoch": 0.5201739792748191, + "grad_norm": 1.8036752939224243, + "learning_rate": 1.7349666924514393e-05, + "loss": 3.288, + "step": 6488 + }, + { + "epoch": 0.5202541540558417, + "grad_norm": 1.810109257698059, + "learning_rate": 1.734878638144566e-05, + "loss": 3.1747, + "step": 6489 + }, + { + "epoch": 0.5203343288368644, + "grad_norm": 0.785599410533905, + "learning_rate": 1.734790571447897e-05, + "loss": 2.5026, + "step": 6490 + }, + { + "epoch": 0.520414503617887, + "grad_norm": 1.8435521125793457, + "learning_rate": 1.7347024923629185e-05, + "loss": 3.1958, + "step": 6491 + }, + { + "epoch": 0.5204946783989096, + "grad_norm": 1.85382878780365, + "learning_rate": 1.734614400891115e-05, + "loss": 3.1858, + "step": 6492 + }, + { + "epoch": 0.5205748531799322, + "grad_norm": 0.8010438084602356, + "learning_rate": 1.7345262970339712e-05, + "loss": 2.499, + "step": 6493 + }, + { + "epoch": 0.5206550279609549, + "grad_norm": 1.8340201377868652, + "learning_rate": 1.734438180792973e-05, + "loss": 3.1736, + "step": 6494 + }, + { + "epoch": 0.5207352027419775, + "grad_norm": 1.8435509204864502, + "learning_rate": 1.734350052169606e-05, + "loss": 3.2701, + "step": 6495 + }, + { + "epoch": 0.5208153775230001, + "grad_norm": 1.7921018600463867, + "learning_rate": 1.7342619111653556e-05, + "loss": 3.2503, + "step": 6496 + }, + { + "epoch": 0.5208955523040227, + "grad_norm": 1.8347837924957275, + "learning_rate": 1.734173757781709e-05, + "loss": 3.2438, + "step": 6497 + }, + { + "epoch": 0.5209757270850454, + "grad_norm": 1.980628252029419, + "learning_rate": 1.734085592020151e-05, + "loss": 3.3203, + "step": 6498 + }, + { + "epoch": 0.521055901866068, + "grad_norm": 1.8486071825027466, + "learning_rate": 1.7339974138821685e-05, + "loss": 3.1632, + "step": 6499 + }, + { + "epoch": 0.5211360766470906, + "grad_norm": 1.9538027048110962, + "learning_rate": 1.7339092233692487e-05, + "loss": 3.2885, + "step": 6500 + }, + { + "epoch": 0.5212162514281133, + "grad_norm": 1.8540873527526855, + "learning_rate": 1.7338210204828776e-05, + "loss": 3.2204, + "step": 6501 + }, + { + "epoch": 0.521296426209136, + "grad_norm": 1.8597146272659302, + "learning_rate": 1.733732805224543e-05, + "loss": 3.1347, + "step": 6502 + }, + { + "epoch": 0.5213766009901586, + "grad_norm": 1.8355692625045776, + "learning_rate": 1.7336445775957317e-05, + "loss": 3.2234, + "step": 6503 + }, + { + "epoch": 0.5214567757711812, + "grad_norm": 1.960200309753418, + "learning_rate": 1.7335563375979317e-05, + "loss": 3.2809, + "step": 6504 + }, + { + "epoch": 0.5215369505522038, + "grad_norm": 1.7835005521774292, + "learning_rate": 1.73346808523263e-05, + "loss": 3.1477, + "step": 6505 + }, + { + "epoch": 0.5216171253332265, + "grad_norm": 1.7143819332122803, + "learning_rate": 1.7333798205013152e-05, + "loss": 3.1105, + "step": 6506 + }, + { + "epoch": 0.5216973001142491, + "grad_norm": 1.7485212087631226, + "learning_rate": 1.7332915434054752e-05, + "loss": 3.3374, + "step": 6507 + }, + { + "epoch": 0.5217774748952717, + "grad_norm": 1.898672103881836, + "learning_rate": 1.733203253946598e-05, + "loss": 3.2068, + "step": 6508 + }, + { + "epoch": 0.5218576496762943, + "grad_norm": 1.874251127243042, + "learning_rate": 1.7331149521261723e-05, + "loss": 3.3096, + "step": 6509 + }, + { + "epoch": 0.521937824457317, + "grad_norm": 1.8099919557571411, + "learning_rate": 1.733026637945687e-05, + "loss": 3.1975, + "step": 6510 + }, + { + "epoch": 0.5220179992383396, + "grad_norm": 0.8211321234703064, + "learning_rate": 1.7329383114066306e-05, + "loss": 2.4965, + "step": 6511 + }, + { + "epoch": 0.5220981740193622, + "grad_norm": 1.9226194620132446, + "learning_rate": 1.732849972510493e-05, + "loss": 3.3607, + "step": 6512 + }, + { + "epoch": 0.5221783488003848, + "grad_norm": 2.4360930919647217, + "learning_rate": 1.732761621258763e-05, + "loss": 3.2816, + "step": 6513 + }, + { + "epoch": 0.5222585235814075, + "grad_norm": 1.896445631980896, + "learning_rate": 1.7326732576529305e-05, + "loss": 3.2301, + "step": 6514 + }, + { + "epoch": 0.5223386983624301, + "grad_norm": 1.8127049207687378, + "learning_rate": 1.7325848816944847e-05, + "loss": 3.23, + "step": 6515 + }, + { + "epoch": 0.5224188731434527, + "grad_norm": 2.0256497859954834, + "learning_rate": 1.732496493384916e-05, + "loss": 3.2783, + "step": 6516 + }, + { + "epoch": 0.5224990479244753, + "grad_norm": 1.9175536632537842, + "learning_rate": 1.7324080927257146e-05, + "loss": 3.3103, + "step": 6517 + }, + { + "epoch": 0.522579222705498, + "grad_norm": 1.8160229921340942, + "learning_rate": 1.7323196797183716e-05, + "loss": 3.2169, + "step": 6518 + }, + { + "epoch": 0.5226593974865206, + "grad_norm": 1.7718455791473389, + "learning_rate": 1.732231254364376e-05, + "loss": 3.1967, + "step": 6519 + }, + { + "epoch": 0.5227395722675432, + "grad_norm": 1.8364425897598267, + "learning_rate": 1.7321428166652197e-05, + "loss": 3.1993, + "step": 6520 + }, + { + "epoch": 0.5228197470485658, + "grad_norm": 1.8447479009628296, + "learning_rate": 1.7320543666223938e-05, + "loss": 3.303, + "step": 6521 + }, + { + "epoch": 0.5228999218295886, + "grad_norm": 1.9432600736618042, + "learning_rate": 1.7319659042373895e-05, + "loss": 3.2084, + "step": 6522 + }, + { + "epoch": 0.5229800966106112, + "grad_norm": 1.738051176071167, + "learning_rate": 1.7318774295116975e-05, + "loss": 3.1821, + "step": 6523 + }, + { + "epoch": 0.5230602713916338, + "grad_norm": 1.6995469331741333, + "learning_rate": 1.73178894244681e-05, + "loss": 3.2844, + "step": 6524 + }, + { + "epoch": 0.5231404461726564, + "grad_norm": 1.8618048429489136, + "learning_rate": 1.7317004430442192e-05, + "loss": 3.1149, + "step": 6525 + }, + { + "epoch": 0.523220620953679, + "grad_norm": 1.7856498956680298, + "learning_rate": 1.7316119313054165e-05, + "loss": 3.3152, + "step": 6526 + }, + { + "epoch": 0.5233007957347017, + "grad_norm": 1.8950668573379517, + "learning_rate": 1.7315234072318947e-05, + "loss": 3.2615, + "step": 6527 + }, + { + "epoch": 0.5233809705157243, + "grad_norm": 1.818502426147461, + "learning_rate": 1.731434870825146e-05, + "loss": 3.3616, + "step": 6528 + }, + { + "epoch": 0.5234611452967469, + "grad_norm": 1.8933573961257935, + "learning_rate": 1.731346322086663e-05, + "loss": 3.3301, + "step": 6529 + }, + { + "epoch": 0.5235413200777695, + "grad_norm": 1.7953753471374512, + "learning_rate": 1.731257761017939e-05, + "loss": 3.1902, + "step": 6530 + }, + { + "epoch": 0.5236214948587922, + "grad_norm": 1.75816810131073, + "learning_rate": 1.7311691876204668e-05, + "loss": 3.2283, + "step": 6531 + }, + { + "epoch": 0.5237016696398148, + "grad_norm": 1.8917759656906128, + "learning_rate": 1.7310806018957397e-05, + "loss": 3.2379, + "step": 6532 + }, + { + "epoch": 0.5237818444208374, + "grad_norm": 1.8299427032470703, + "learning_rate": 1.7309920038452514e-05, + "loss": 3.1738, + "step": 6533 + }, + { + "epoch": 0.52386201920186, + "grad_norm": 1.8162617683410645, + "learning_rate": 1.7309033934704954e-05, + "loss": 3.2015, + "step": 6534 + }, + { + "epoch": 0.5239421939828827, + "grad_norm": 1.9372798204421997, + "learning_rate": 1.7308147707729658e-05, + "loss": 3.0986, + "step": 6535 + }, + { + "epoch": 0.5240223687639053, + "grad_norm": 1.768688440322876, + "learning_rate": 1.7307261357541568e-05, + "loss": 3.1841, + "step": 6536 + }, + { + "epoch": 0.5241025435449279, + "grad_norm": 1.817578911781311, + "learning_rate": 1.7306374884155626e-05, + "loss": 3.2384, + "step": 6537 + }, + { + "epoch": 0.5241827183259505, + "grad_norm": 1.8394553661346436, + "learning_rate": 1.730548828758678e-05, + "loss": 3.191, + "step": 6538 + }, + { + "epoch": 0.5242628931069732, + "grad_norm": 1.819886326789856, + "learning_rate": 1.730460156784998e-05, + "loss": 3.1763, + "step": 6539 + }, + { + "epoch": 0.5243430678879958, + "grad_norm": 1.9987760782241821, + "learning_rate": 1.7303714724960165e-05, + "loss": 3.2986, + "step": 6540 + }, + { + "epoch": 0.5244232426690184, + "grad_norm": 1.8051680326461792, + "learning_rate": 1.7302827758932295e-05, + "loss": 3.1783, + "step": 6541 + }, + { + "epoch": 0.5245034174500411, + "grad_norm": 1.857459306716919, + "learning_rate": 1.7301940669781326e-05, + "loss": 3.2593, + "step": 6542 + }, + { + "epoch": 0.5245835922310638, + "grad_norm": 2.0213584899902344, + "learning_rate": 1.7301053457522208e-05, + "loss": 3.2039, + "step": 6543 + }, + { + "epoch": 0.5246637670120864, + "grad_norm": 1.8542096614837646, + "learning_rate": 1.73001661221699e-05, + "loss": 3.3356, + "step": 6544 + }, + { + "epoch": 0.524743941793109, + "grad_norm": 0.8328651785850525, + "learning_rate": 1.729927866373937e-05, + "loss": 2.5278, + "step": 6545 + }, + { + "epoch": 0.5248241165741316, + "grad_norm": 1.9042223691940308, + "learning_rate": 1.729839108224557e-05, + "loss": 3.2153, + "step": 6546 + }, + { + "epoch": 0.5249042913551543, + "grad_norm": 1.8602784872055054, + "learning_rate": 1.729750337770347e-05, + "loss": 3.1598, + "step": 6547 + }, + { + "epoch": 0.5249844661361769, + "grad_norm": 1.8289012908935547, + "learning_rate": 1.7296615550128035e-05, + "loss": 3.2333, + "step": 6548 + }, + { + "epoch": 0.5250646409171995, + "grad_norm": 1.8991221189498901, + "learning_rate": 1.729572759953423e-05, + "loss": 3.2369, + "step": 6549 + }, + { + "epoch": 0.5251448156982221, + "grad_norm": 1.8584189414978027, + "learning_rate": 1.7294839525937034e-05, + "loss": 3.2185, + "step": 6550 + }, + { + "epoch": 0.5252249904792448, + "grad_norm": 2.1050221920013428, + "learning_rate": 1.729395132935141e-05, + "loss": 3.2357, + "step": 6551 + }, + { + "epoch": 0.5253051652602674, + "grad_norm": 1.9592013359069824, + "learning_rate": 1.7293063009792342e-05, + "loss": 3.2229, + "step": 6552 + }, + { + "epoch": 0.52538534004129, + "grad_norm": 0.8686491847038269, + "learning_rate": 1.72921745672748e-05, + "loss": 2.526, + "step": 6553 + }, + { + "epoch": 0.5254655148223126, + "grad_norm": 1.8534080982208252, + "learning_rate": 1.7291286001813765e-05, + "loss": 3.2185, + "step": 6554 + }, + { + "epoch": 0.5255456896033353, + "grad_norm": 1.8130396604537964, + "learning_rate": 1.7290397313424215e-05, + "loss": 3.2353, + "step": 6555 + }, + { + "epoch": 0.5256258643843579, + "grad_norm": 1.7966413497924805, + "learning_rate": 1.728950850212114e-05, + "loss": 3.1902, + "step": 6556 + }, + { + "epoch": 0.5257060391653805, + "grad_norm": 1.770229458808899, + "learning_rate": 1.728861956791952e-05, + "loss": 3.2249, + "step": 6557 + }, + { + "epoch": 0.5257862139464031, + "grad_norm": 1.7967711687088013, + "learning_rate": 1.7287730510834338e-05, + "loss": 3.1906, + "step": 6558 + }, + { + "epoch": 0.5258663887274257, + "grad_norm": 1.8760088682174683, + "learning_rate": 1.728684133088059e-05, + "loss": 3.1703, + "step": 6559 + }, + { + "epoch": 0.5259465635084484, + "grad_norm": 2.1153013706207275, + "learning_rate": 1.7285952028073266e-05, + "loss": 3.2798, + "step": 6560 + }, + { + "epoch": 0.526026738289471, + "grad_norm": 1.7978527545928955, + "learning_rate": 1.7285062602427363e-05, + "loss": 3.29, + "step": 6561 + }, + { + "epoch": 0.5261069130704937, + "grad_norm": 1.795015811920166, + "learning_rate": 1.7284173053957868e-05, + "loss": 3.1088, + "step": 6562 + }, + { + "epoch": 0.5261870878515164, + "grad_norm": 1.8164101839065552, + "learning_rate": 1.728328338267978e-05, + "loss": 3.2435, + "step": 6563 + }, + { + "epoch": 0.526267262632539, + "grad_norm": 1.8803412914276123, + "learning_rate": 1.7282393588608104e-05, + "loss": 3.1577, + "step": 6564 + }, + { + "epoch": 0.5263474374135616, + "grad_norm": 1.7549214363098145, + "learning_rate": 1.728150367175784e-05, + "loss": 3.1473, + "step": 6565 + }, + { + "epoch": 0.5264276121945842, + "grad_norm": 1.7628660202026367, + "learning_rate": 1.728061363214399e-05, + "loss": 3.2397, + "step": 6566 + }, + { + "epoch": 0.5265077869756069, + "grad_norm": 1.7632660865783691, + "learning_rate": 1.727972346978156e-05, + "loss": 3.1999, + "step": 6567 + }, + { + "epoch": 0.5265879617566295, + "grad_norm": 1.8460218906402588, + "learning_rate": 1.7278833184685554e-05, + "loss": 3.2335, + "step": 6568 + }, + { + "epoch": 0.5266681365376521, + "grad_norm": 1.924728274345398, + "learning_rate": 1.727794277687099e-05, + "loss": 3.2628, + "step": 6569 + }, + { + "epoch": 0.5267483113186747, + "grad_norm": 2.035496234893799, + "learning_rate": 1.727705224635287e-05, + "loss": 3.2971, + "step": 6570 + }, + { + "epoch": 0.5268284860996973, + "grad_norm": 0.7941402196884155, + "learning_rate": 1.727616159314622e-05, + "loss": 2.5015, + "step": 6571 + }, + { + "epoch": 0.52690866088072, + "grad_norm": 1.8906441926956177, + "learning_rate": 1.727527081726605e-05, + "loss": 3.3192, + "step": 6572 + }, + { + "epoch": 0.5269888356617426, + "grad_norm": 1.8569493293762207, + "learning_rate": 1.7274379918727376e-05, + "loss": 3.2645, + "step": 6573 + }, + { + "epoch": 0.5270690104427652, + "grad_norm": 1.7701549530029297, + "learning_rate": 1.727348889754522e-05, + "loss": 3.2229, + "step": 6574 + }, + { + "epoch": 0.5271491852237878, + "grad_norm": 1.9122000932693481, + "learning_rate": 1.7272597753734606e-05, + "loss": 3.2225, + "step": 6575 + }, + { + "epoch": 0.5272293600048105, + "grad_norm": 1.800526738166809, + "learning_rate": 1.7271706487310556e-05, + "loss": 3.2063, + "step": 6576 + }, + { + "epoch": 0.5273095347858331, + "grad_norm": 1.687292218208313, + "learning_rate": 1.7270815098288096e-05, + "loss": 3.135, + "step": 6577 + }, + { + "epoch": 0.5273897095668557, + "grad_norm": 1.784808874130249, + "learning_rate": 1.726992358668226e-05, + "loss": 3.2258, + "step": 6578 + }, + { + "epoch": 0.5274698843478783, + "grad_norm": 1.7236946821212769, + "learning_rate": 1.726903195250807e-05, + "loss": 3.1558, + "step": 6579 + }, + { + "epoch": 0.527550059128901, + "grad_norm": 1.7662742137908936, + "learning_rate": 1.7268140195780567e-05, + "loss": 3.2317, + "step": 6580 + }, + { + "epoch": 0.5276302339099236, + "grad_norm": 1.7548034191131592, + "learning_rate": 1.726724831651478e-05, + "loss": 3.1494, + "step": 6581 + }, + { + "epoch": 0.5277104086909462, + "grad_norm": 1.8349283933639526, + "learning_rate": 1.7266356314725748e-05, + "loss": 3.2727, + "step": 6582 + }, + { + "epoch": 0.5277905834719689, + "grad_norm": 1.7735671997070312, + "learning_rate": 1.7265464190428508e-05, + "loss": 3.3207, + "step": 6583 + }, + { + "epoch": 0.5278707582529916, + "grad_norm": 1.8081605434417725, + "learning_rate": 1.7264571943638103e-05, + "loss": 3.2407, + "step": 6584 + }, + { + "epoch": 0.5279509330340142, + "grad_norm": 1.8461620807647705, + "learning_rate": 1.7263679574369577e-05, + "loss": 3.319, + "step": 6585 + }, + { + "epoch": 0.5280311078150368, + "grad_norm": 1.9127012491226196, + "learning_rate": 1.7262787082637972e-05, + "loss": 3.1471, + "step": 6586 + }, + { + "epoch": 0.5281112825960594, + "grad_norm": 1.8161436319351196, + "learning_rate": 1.7261894468458336e-05, + "loss": 3.2264, + "step": 6587 + }, + { + "epoch": 0.5281914573770821, + "grad_norm": 1.9148967266082764, + "learning_rate": 1.726100173184572e-05, + "loss": 3.2642, + "step": 6588 + }, + { + "epoch": 0.5282716321581047, + "grad_norm": 0.8195347189903259, + "learning_rate": 1.726010887281517e-05, + "loss": 2.5, + "step": 6589 + }, + { + "epoch": 0.5283518069391273, + "grad_norm": 1.8918368816375732, + "learning_rate": 1.7259215891381746e-05, + "loss": 3.1366, + "step": 6590 + }, + { + "epoch": 0.5284319817201499, + "grad_norm": 1.8130089044570923, + "learning_rate": 1.72583227875605e-05, + "loss": 3.2932, + "step": 6591 + }, + { + "epoch": 0.5285121565011726, + "grad_norm": 1.8400545120239258, + "learning_rate": 1.7257429561366493e-05, + "loss": 3.2534, + "step": 6592 + }, + { + "epoch": 0.5285923312821952, + "grad_norm": 1.7715517282485962, + "learning_rate": 1.725653621281478e-05, + "loss": 3.2981, + "step": 6593 + }, + { + "epoch": 0.5286725060632178, + "grad_norm": 0.7830035090446472, + "learning_rate": 1.725564274192042e-05, + "loss": 2.4913, + "step": 6594 + }, + { + "epoch": 0.5287526808442404, + "grad_norm": 1.8592246770858765, + "learning_rate": 1.725474914869848e-05, + "loss": 3.254, + "step": 6595 + }, + { + "epoch": 0.528832855625263, + "grad_norm": 1.7169458866119385, + "learning_rate": 1.7253855433164034e-05, + "loss": 3.0979, + "step": 6596 + }, + { + "epoch": 0.5289130304062857, + "grad_norm": 1.8917235136032104, + "learning_rate": 1.7252961595332136e-05, + "loss": 3.2465, + "step": 6597 + }, + { + "epoch": 0.5289932051873083, + "grad_norm": 1.8587111234664917, + "learning_rate": 1.7252067635217864e-05, + "loss": 3.2291, + "step": 6598 + }, + { + "epoch": 0.5290733799683309, + "grad_norm": 1.8159525394439697, + "learning_rate": 1.7251173552836285e-05, + "loss": 3.17, + "step": 6599 + }, + { + "epoch": 0.5291535547493535, + "grad_norm": 1.785967230796814, + "learning_rate": 1.7250279348202477e-05, + "loss": 3.193, + "step": 6600 + }, + { + "epoch": 0.5292337295303762, + "grad_norm": 1.9047852754592896, + "learning_rate": 1.7249385021331515e-05, + "loss": 3.1626, + "step": 6601 + }, + { + "epoch": 0.5293139043113988, + "grad_norm": 1.7914760112762451, + "learning_rate": 1.7248490572238473e-05, + "loss": 3.2549, + "step": 6602 + }, + { + "epoch": 0.5293940790924215, + "grad_norm": 1.8096973896026611, + "learning_rate": 1.724759600093844e-05, + "loss": 3.1592, + "step": 6603 + }, + { + "epoch": 0.5294742538734442, + "grad_norm": 1.8707430362701416, + "learning_rate": 1.724670130744649e-05, + "loss": 3.1554, + "step": 6604 + }, + { + "epoch": 0.5295544286544668, + "grad_norm": 1.8051965236663818, + "learning_rate": 1.7245806491777707e-05, + "loss": 3.1444, + "step": 6605 + }, + { + "epoch": 0.5296346034354894, + "grad_norm": 1.8315149545669556, + "learning_rate": 1.7244911553947186e-05, + "loss": 3.2755, + "step": 6606 + }, + { + "epoch": 0.529714778216512, + "grad_norm": 1.7213249206542969, + "learning_rate": 1.7244016493970008e-05, + "loss": 3.1896, + "step": 6607 + }, + { + "epoch": 0.5297949529975347, + "grad_norm": 1.907504677772522, + "learning_rate": 1.7243121311861263e-05, + "loss": 3.2079, + "step": 6608 + }, + { + "epoch": 0.5298751277785573, + "grad_norm": 2.0456900596618652, + "learning_rate": 1.7242226007636046e-05, + "loss": 3.3704, + "step": 6609 + }, + { + "epoch": 0.5299553025595799, + "grad_norm": 1.80085027217865, + "learning_rate": 1.724133058130945e-05, + "loss": 3.356, + "step": 6610 + }, + { + "epoch": 0.5300354773406025, + "grad_norm": 1.818281650543213, + "learning_rate": 1.7240435032896575e-05, + "loss": 3.2618, + "step": 6611 + }, + { + "epoch": 0.5301156521216251, + "grad_norm": 1.86196768283844, + "learning_rate": 1.7239539362412516e-05, + "loss": 3.2887, + "step": 6612 + }, + { + "epoch": 0.5301958269026478, + "grad_norm": 1.866455078125, + "learning_rate": 1.7238643569872375e-05, + "loss": 3.2039, + "step": 6613 + }, + { + "epoch": 0.5302760016836704, + "grad_norm": 1.9888958930969238, + "learning_rate": 1.7237747655291252e-05, + "loss": 3.1675, + "step": 6614 + }, + { + "epoch": 0.530356176464693, + "grad_norm": 1.8108881711959839, + "learning_rate": 1.7236851618684256e-05, + "loss": 3.3164, + "step": 6615 + }, + { + "epoch": 0.5304363512457156, + "grad_norm": 1.9292728900909424, + "learning_rate": 1.7235955460066493e-05, + "loss": 3.3772, + "step": 6616 + }, + { + "epoch": 0.5305165260267383, + "grad_norm": 1.8339779376983643, + "learning_rate": 1.7235059179453073e-05, + "loss": 3.0934, + "step": 6617 + }, + { + "epoch": 0.5305967008077609, + "grad_norm": 1.9968022108078003, + "learning_rate": 1.72341627768591e-05, + "loss": 3.2805, + "step": 6618 + }, + { + "epoch": 0.5306768755887835, + "grad_norm": 1.8422935009002686, + "learning_rate": 1.7233266252299697e-05, + "loss": 3.2507, + "step": 6619 + }, + { + "epoch": 0.5307570503698061, + "grad_norm": 1.7525216341018677, + "learning_rate": 1.723236960578997e-05, + "loss": 3.27, + "step": 6620 + }, + { + "epoch": 0.5308372251508288, + "grad_norm": 2.0232226848602295, + "learning_rate": 1.7231472837345045e-05, + "loss": 3.1853, + "step": 6621 + }, + { + "epoch": 0.5309173999318514, + "grad_norm": 1.8525773286819458, + "learning_rate": 1.7230575946980032e-05, + "loss": 3.1742, + "step": 6622 + }, + { + "epoch": 0.5309975747128741, + "grad_norm": 2.052762985229492, + "learning_rate": 1.722967893471006e-05, + "loss": 3.3344, + "step": 6623 + }, + { + "epoch": 0.5310777494938967, + "grad_norm": 2.034287691116333, + "learning_rate": 1.722878180055025e-05, + "loss": 3.2531, + "step": 6624 + }, + { + "epoch": 0.5311579242749194, + "grad_norm": 1.9618349075317383, + "learning_rate": 1.7227884544515724e-05, + "loss": 3.2327, + "step": 6625 + }, + { + "epoch": 0.531238099055942, + "grad_norm": 1.9076247215270996, + "learning_rate": 1.722698716662161e-05, + "loss": 3.2062, + "step": 6626 + }, + { + "epoch": 0.5313182738369646, + "grad_norm": 1.6772996187210083, + "learning_rate": 1.7226089666883043e-05, + "loss": 3.2666, + "step": 6627 + }, + { + "epoch": 0.5313984486179872, + "grad_norm": 1.872066617012024, + "learning_rate": 1.7225192045315152e-05, + "loss": 3.3323, + "step": 6628 + }, + { + "epoch": 0.5314786233990099, + "grad_norm": 1.8159056901931763, + "learning_rate": 1.722429430193307e-05, + "loss": 3.2111, + "step": 6629 + }, + { + "epoch": 0.5315587981800325, + "grad_norm": 1.8445805311203003, + "learning_rate": 1.7223396436751927e-05, + "loss": 3.1881, + "step": 6630 + }, + { + "epoch": 0.5316389729610551, + "grad_norm": 1.8696088790893555, + "learning_rate": 1.722249844978687e-05, + "loss": 3.2499, + "step": 6631 + }, + { + "epoch": 0.5317191477420777, + "grad_norm": 1.9070707559585571, + "learning_rate": 1.7221600341053033e-05, + "loss": 3.201, + "step": 6632 + }, + { + "epoch": 0.5317993225231004, + "grad_norm": 1.8498826026916504, + "learning_rate": 1.722070211056556e-05, + "loss": 3.351, + "step": 6633 + }, + { + "epoch": 0.531879497304123, + "grad_norm": 1.8085545301437378, + "learning_rate": 1.721980375833959e-05, + "loss": 3.1754, + "step": 6634 + }, + { + "epoch": 0.5319596720851456, + "grad_norm": 1.8156777620315552, + "learning_rate": 1.7218905284390278e-05, + "loss": 3.1207, + "step": 6635 + }, + { + "epoch": 0.5320398468661682, + "grad_norm": 1.9381440877914429, + "learning_rate": 1.7218006688732766e-05, + "loss": 3.2793, + "step": 6636 + }, + { + "epoch": 0.5321200216471909, + "grad_norm": 0.7729352116584778, + "learning_rate": 1.7217107971382206e-05, + "loss": 2.5004, + "step": 6637 + }, + { + "epoch": 0.5322001964282135, + "grad_norm": 1.8001229763031006, + "learning_rate": 1.721620913235375e-05, + "loss": 3.2881, + "step": 6638 + }, + { + "epoch": 0.5322803712092361, + "grad_norm": 1.8151578903198242, + "learning_rate": 1.7215310171662548e-05, + "loss": 3.3068, + "step": 6639 + }, + { + "epoch": 0.5323605459902587, + "grad_norm": 1.7908227443695068, + "learning_rate": 1.721441108932376e-05, + "loss": 3.1184, + "step": 6640 + }, + { + "epoch": 0.5324407207712814, + "grad_norm": 2.004595994949341, + "learning_rate": 1.7213511885352546e-05, + "loss": 3.2977, + "step": 6641 + }, + { + "epoch": 0.532520895552304, + "grad_norm": 1.7518917322158813, + "learning_rate": 1.7212612559764062e-05, + "loss": 3.2936, + "step": 6642 + }, + { + "epoch": 0.5326010703333266, + "grad_norm": 1.9988876581192017, + "learning_rate": 1.7211713112573473e-05, + "loss": 3.2134, + "step": 6643 + }, + { + "epoch": 0.5326812451143493, + "grad_norm": 1.9567981958389282, + "learning_rate": 1.7210813543795942e-05, + "loss": 3.2709, + "step": 6644 + }, + { + "epoch": 0.532761419895372, + "grad_norm": 0.9513174891471863, + "learning_rate": 1.7209913853446635e-05, + "loss": 2.5258, + "step": 6645 + }, + { + "epoch": 0.5328415946763946, + "grad_norm": 0.8418288230895996, + "learning_rate": 1.7209014041540728e-05, + "loss": 2.5225, + "step": 6646 + }, + { + "epoch": 0.5329217694574172, + "grad_norm": 1.8781802654266357, + "learning_rate": 1.7208114108093378e-05, + "loss": 3.2654, + "step": 6647 + }, + { + "epoch": 0.5330019442384398, + "grad_norm": 0.7468239665031433, + "learning_rate": 1.7207214053119766e-05, + "loss": 2.5034, + "step": 6648 + }, + { + "epoch": 0.5330821190194625, + "grad_norm": 1.981019377708435, + "learning_rate": 1.7206313876635062e-05, + "loss": 3.3234, + "step": 6649 + }, + { + "epoch": 0.5331622938004851, + "grad_norm": 1.8651516437530518, + "learning_rate": 1.720541357865445e-05, + "loss": 3.2675, + "step": 6650 + }, + { + "epoch": 0.5332424685815077, + "grad_norm": 1.7246214151382446, + "learning_rate": 1.7204513159193107e-05, + "loss": 3.1963, + "step": 6651 + }, + { + "epoch": 0.5333226433625303, + "grad_norm": 1.9144222736358643, + "learning_rate": 1.7203612618266206e-05, + "loss": 3.1964, + "step": 6652 + }, + { + "epoch": 0.533402818143553, + "grad_norm": 1.9734269380569458, + "learning_rate": 1.7202711955888936e-05, + "loss": 3.3469, + "step": 6653 + }, + { + "epoch": 0.5334829929245756, + "grad_norm": 1.7888976335525513, + "learning_rate": 1.7201811172076483e-05, + "loss": 3.2759, + "step": 6654 + }, + { + "epoch": 0.5335631677055982, + "grad_norm": 1.911947250366211, + "learning_rate": 1.720091026684403e-05, + "loss": 3.2509, + "step": 6655 + }, + { + "epoch": 0.5336433424866208, + "grad_norm": 1.8396501541137695, + "learning_rate": 1.7200009240206766e-05, + "loss": 3.2238, + "step": 6656 + }, + { + "epoch": 0.5337235172676434, + "grad_norm": 1.9131141901016235, + "learning_rate": 1.7199108092179886e-05, + "loss": 3.0983, + "step": 6657 + }, + { + "epoch": 0.5338036920486661, + "grad_norm": 1.8609386682510376, + "learning_rate": 1.7198206822778583e-05, + "loss": 3.1548, + "step": 6658 + }, + { + "epoch": 0.5338838668296887, + "grad_norm": 1.887492299079895, + "learning_rate": 1.7197305432018048e-05, + "loss": 3.2708, + "step": 6659 + }, + { + "epoch": 0.5339640416107113, + "grad_norm": 1.77082359790802, + "learning_rate": 1.719640391991348e-05, + "loss": 3.1799, + "step": 6660 + }, + { + "epoch": 0.5340442163917339, + "grad_norm": 1.7891862392425537, + "learning_rate": 1.7195502286480075e-05, + "loss": 3.2474, + "step": 6661 + }, + { + "epoch": 0.5341243911727566, + "grad_norm": 1.92422354221344, + "learning_rate": 1.7194600531733043e-05, + "loss": 3.1805, + "step": 6662 + }, + { + "epoch": 0.5342045659537792, + "grad_norm": 1.8954358100891113, + "learning_rate": 1.719369865568758e-05, + "loss": 3.1854, + "step": 6663 + }, + { + "epoch": 0.5342847407348019, + "grad_norm": 1.815285563468933, + "learning_rate": 1.719279665835889e-05, + "loss": 3.2086, + "step": 6664 + }, + { + "epoch": 0.5343649155158245, + "grad_norm": 1.7033201456069946, + "learning_rate": 1.7191894539762185e-05, + "loss": 3.2162, + "step": 6665 + }, + { + "epoch": 0.5344450902968472, + "grad_norm": 1.889507532119751, + "learning_rate": 1.7190992299912676e-05, + "loss": 3.3041, + "step": 6666 + }, + { + "epoch": 0.5345252650778698, + "grad_norm": 1.8844563961029053, + "learning_rate": 1.7190089938825565e-05, + "loss": 3.2666, + "step": 6667 + }, + { + "epoch": 0.5346054398588924, + "grad_norm": 1.9060853719711304, + "learning_rate": 1.7189187456516077e-05, + "loss": 3.2371, + "step": 6668 + }, + { + "epoch": 0.534685614639915, + "grad_norm": 0.8348785638809204, + "learning_rate": 1.718828485299942e-05, + "loss": 2.4984, + "step": 6669 + }, + { + "epoch": 0.5347657894209377, + "grad_norm": 1.953799843788147, + "learning_rate": 1.7187382128290812e-05, + "loss": 3.3302, + "step": 6670 + }, + { + "epoch": 0.5348459642019603, + "grad_norm": 1.827182412147522, + "learning_rate": 1.718647928240548e-05, + "loss": 3.196, + "step": 6671 + }, + { + "epoch": 0.5349261389829829, + "grad_norm": 1.8486264944076538, + "learning_rate": 1.7185576315358635e-05, + "loss": 3.2023, + "step": 6672 + }, + { + "epoch": 0.5350063137640055, + "grad_norm": 1.884146809577942, + "learning_rate": 1.7184673227165506e-05, + "loss": 3.219, + "step": 6673 + }, + { + "epoch": 0.5350864885450282, + "grad_norm": 1.8138923645019531, + "learning_rate": 1.718377001784132e-05, + "loss": 3.1143, + "step": 6674 + }, + { + "epoch": 0.5351666633260508, + "grad_norm": 1.882790446281433, + "learning_rate": 1.7182866687401302e-05, + "loss": 3.1545, + "step": 6675 + }, + { + "epoch": 0.5352468381070734, + "grad_norm": 1.7853835821151733, + "learning_rate": 1.7181963235860686e-05, + "loss": 3.2321, + "step": 6676 + }, + { + "epoch": 0.535327012888096, + "grad_norm": 1.8982617855072021, + "learning_rate": 1.7181059663234697e-05, + "loss": 3.2628, + "step": 6677 + }, + { + "epoch": 0.5354071876691187, + "grad_norm": 1.9637703895568848, + "learning_rate": 1.7180155969538576e-05, + "loss": 3.2572, + "step": 6678 + }, + { + "epoch": 0.5354873624501413, + "grad_norm": 1.8173432350158691, + "learning_rate": 1.7179252154787555e-05, + "loss": 3.1537, + "step": 6679 + }, + { + "epoch": 0.5355675372311639, + "grad_norm": 1.811964988708496, + "learning_rate": 1.7178348218996874e-05, + "loss": 3.2147, + "step": 6680 + }, + { + "epoch": 0.5356477120121865, + "grad_norm": 1.8570908308029175, + "learning_rate": 1.7177444162181767e-05, + "loss": 3.0726, + "step": 6681 + }, + { + "epoch": 0.5357278867932092, + "grad_norm": 1.8362491130828857, + "learning_rate": 1.7176539984357487e-05, + "loss": 3.2027, + "step": 6682 + }, + { + "epoch": 0.5358080615742318, + "grad_norm": 2.056182384490967, + "learning_rate": 1.717563568553927e-05, + "loss": 3.2357, + "step": 6683 + }, + { + "epoch": 0.5358882363552544, + "grad_norm": 0.8517770171165466, + "learning_rate": 1.7174731265742365e-05, + "loss": 2.515, + "step": 6684 + }, + { + "epoch": 0.5359684111362771, + "grad_norm": 1.920892357826233, + "learning_rate": 1.7173826724982018e-05, + "loss": 3.2792, + "step": 6685 + }, + { + "epoch": 0.5360485859172998, + "grad_norm": 2.1741855144500732, + "learning_rate": 1.717292206327348e-05, + "loss": 3.3574, + "step": 6686 + }, + { + "epoch": 0.5361287606983224, + "grad_norm": 1.900213360786438, + "learning_rate": 1.7172017280632008e-05, + "loss": 3.2924, + "step": 6687 + }, + { + "epoch": 0.536208935479345, + "grad_norm": 1.7967324256896973, + "learning_rate": 1.7171112377072846e-05, + "loss": 3.1224, + "step": 6688 + }, + { + "epoch": 0.5362891102603676, + "grad_norm": 2.0320820808410645, + "learning_rate": 1.717020735261126e-05, + "loss": 3.316, + "step": 6689 + }, + { + "epoch": 0.5363692850413903, + "grad_norm": 1.8998050689697266, + "learning_rate": 1.716930220726251e-05, + "loss": 3.1909, + "step": 6690 + }, + { + "epoch": 0.5364494598224129, + "grad_norm": 2.11956524848938, + "learning_rate": 1.7168396941041843e-05, + "loss": 3.265, + "step": 6691 + }, + { + "epoch": 0.5365296346034355, + "grad_norm": 1.8057780265808105, + "learning_rate": 1.7167491553964535e-05, + "loss": 3.1836, + "step": 6692 + }, + { + "epoch": 0.5366098093844581, + "grad_norm": 2.04879093170166, + "learning_rate": 1.7166586046045844e-05, + "loss": 3.2496, + "step": 6693 + }, + { + "epoch": 0.5366899841654807, + "grad_norm": 1.8238945007324219, + "learning_rate": 1.7165680417301037e-05, + "loss": 3.2169, + "step": 6694 + }, + { + "epoch": 0.5367701589465034, + "grad_norm": 1.733415961265564, + "learning_rate": 1.7164774667745384e-05, + "loss": 3.236, + "step": 6695 + }, + { + "epoch": 0.536850333727526, + "grad_norm": 1.7996704578399658, + "learning_rate": 1.716386879739416e-05, + "loss": 3.224, + "step": 6696 + }, + { + "epoch": 0.5369305085085486, + "grad_norm": 1.9382789134979248, + "learning_rate": 1.7162962806262627e-05, + "loss": 3.3094, + "step": 6697 + }, + { + "epoch": 0.5370106832895712, + "grad_norm": 1.8503767251968384, + "learning_rate": 1.7162056694366066e-05, + "loss": 3.3411, + "step": 6698 + }, + { + "epoch": 0.5370908580705939, + "grad_norm": 1.9060754776000977, + "learning_rate": 1.7161150461719754e-05, + "loss": 3.2212, + "step": 6699 + }, + { + "epoch": 0.5371710328516165, + "grad_norm": 1.9546608924865723, + "learning_rate": 1.7160244108338972e-05, + "loss": 3.2699, + "step": 6700 + }, + { + "epoch": 0.5372512076326391, + "grad_norm": 1.8033764362335205, + "learning_rate": 1.7159337634238994e-05, + "loss": 3.2128, + "step": 6701 + }, + { + "epoch": 0.5373313824136617, + "grad_norm": 1.9103788137435913, + "learning_rate": 1.715843103943511e-05, + "loss": 3.2273, + "step": 6702 + }, + { + "epoch": 0.5374115571946844, + "grad_norm": 1.7881715297698975, + "learning_rate": 1.71575243239426e-05, + "loss": 3.1278, + "step": 6703 + }, + { + "epoch": 0.537491731975707, + "grad_norm": 1.9969278573989868, + "learning_rate": 1.715661748777675e-05, + "loss": 3.2901, + "step": 6704 + }, + { + "epoch": 0.5375719067567297, + "grad_norm": 1.8737844228744507, + "learning_rate": 1.7155710530952855e-05, + "loss": 3.2312, + "step": 6705 + }, + { + "epoch": 0.5376520815377523, + "grad_norm": 1.7789441347122192, + "learning_rate": 1.7154803453486202e-05, + "loss": 3.2057, + "step": 6706 + }, + { + "epoch": 0.537732256318775, + "grad_norm": 1.8253363370895386, + "learning_rate": 1.7153896255392083e-05, + "loss": 3.207, + "step": 6707 + }, + { + "epoch": 0.5378124310997976, + "grad_norm": 1.785622477531433, + "learning_rate": 1.7152988936685795e-05, + "loss": 3.1257, + "step": 6708 + }, + { + "epoch": 0.5378926058808202, + "grad_norm": 1.8722797632217407, + "learning_rate": 1.715208149738263e-05, + "loss": 3.2131, + "step": 6709 + }, + { + "epoch": 0.5379727806618428, + "grad_norm": 1.8109441995620728, + "learning_rate": 1.71511739374979e-05, + "loss": 3.2144, + "step": 6710 + }, + { + "epoch": 0.5380529554428655, + "grad_norm": 1.8770201206207275, + "learning_rate": 1.7150266257046894e-05, + "loss": 3.2861, + "step": 6711 + }, + { + "epoch": 0.5381331302238881, + "grad_norm": 1.8642761707305908, + "learning_rate": 1.714935845604492e-05, + "loss": 3.207, + "step": 6712 + }, + { + "epoch": 0.5382133050049107, + "grad_norm": 1.8457750082015991, + "learning_rate": 1.714845053450728e-05, + "loss": 3.2742, + "step": 6713 + }, + { + "epoch": 0.5382934797859333, + "grad_norm": 1.9671550989151, + "learning_rate": 1.7147542492449285e-05, + "loss": 3.1766, + "step": 6714 + }, + { + "epoch": 0.538373654566956, + "grad_norm": 1.8147141933441162, + "learning_rate": 1.714663432988624e-05, + "loss": 3.2906, + "step": 6715 + }, + { + "epoch": 0.5384538293479786, + "grad_norm": 1.9992696046829224, + "learning_rate": 1.7145726046833463e-05, + "loss": 3.3424, + "step": 6716 + }, + { + "epoch": 0.5385340041290012, + "grad_norm": 1.7752604484558105, + "learning_rate": 1.714481764330626e-05, + "loss": 3.235, + "step": 6717 + }, + { + "epoch": 0.5386141789100238, + "grad_norm": 0.7560766935348511, + "learning_rate": 1.7143909119319947e-05, + "loss": 2.5182, + "step": 6718 + }, + { + "epoch": 0.5386943536910465, + "grad_norm": 1.766253113746643, + "learning_rate": 1.714300047488985e-05, + "loss": 3.2491, + "step": 6719 + }, + { + "epoch": 0.5387745284720691, + "grad_norm": 1.8443303108215332, + "learning_rate": 1.7142091710031282e-05, + "loss": 3.3458, + "step": 6720 + }, + { + "epoch": 0.5388547032530917, + "grad_norm": 1.8319391012191772, + "learning_rate": 1.714118282475956e-05, + "loss": 3.2565, + "step": 6721 + }, + { + "epoch": 0.5389348780341143, + "grad_norm": 1.8318250179290771, + "learning_rate": 1.7140273819090017e-05, + "loss": 3.1543, + "step": 6722 + }, + { + "epoch": 0.539015052815137, + "grad_norm": 1.856069564819336, + "learning_rate": 1.7139364693037972e-05, + "loss": 3.2209, + "step": 6723 + }, + { + "epoch": 0.5390952275961596, + "grad_norm": 1.8372219800949097, + "learning_rate": 1.7138455446618754e-05, + "loss": 3.2095, + "step": 6724 + }, + { + "epoch": 0.5391754023771823, + "grad_norm": 1.7677117586135864, + "learning_rate": 1.7137546079847694e-05, + "loss": 3.1536, + "step": 6725 + }, + { + "epoch": 0.5392555771582049, + "grad_norm": 1.900915265083313, + "learning_rate": 1.713663659274012e-05, + "loss": 3.254, + "step": 6726 + }, + { + "epoch": 0.5393357519392276, + "grad_norm": 1.8038910627365112, + "learning_rate": 1.713572698531137e-05, + "loss": 3.2817, + "step": 6727 + }, + { + "epoch": 0.5394159267202502, + "grad_norm": 1.8515342473983765, + "learning_rate": 1.713481725757678e-05, + "loss": 3.1986, + "step": 6728 + }, + { + "epoch": 0.5394961015012728, + "grad_norm": 1.891726016998291, + "learning_rate": 1.7133907409551683e-05, + "loss": 3.1578, + "step": 6729 + }, + { + "epoch": 0.5395762762822954, + "grad_norm": 1.802622675895691, + "learning_rate": 1.713299744125142e-05, + "loss": 3.1608, + "step": 6730 + }, + { + "epoch": 0.539656451063318, + "grad_norm": 0.8785640597343445, + "learning_rate": 1.713208735269134e-05, + "loss": 2.5132, + "step": 6731 + }, + { + "epoch": 0.5397366258443407, + "grad_norm": 1.8204680681228638, + "learning_rate": 1.7131177143886778e-05, + "loss": 3.1288, + "step": 6732 + }, + { + "epoch": 0.5398168006253633, + "grad_norm": 1.8628517389297485, + "learning_rate": 1.713026681485308e-05, + "loss": 3.1967, + "step": 6733 + }, + { + "epoch": 0.5398969754063859, + "grad_norm": 1.8909305334091187, + "learning_rate": 1.71293563656056e-05, + "loss": 3.2985, + "step": 6734 + }, + { + "epoch": 0.5399771501874085, + "grad_norm": 1.8321946859359741, + "learning_rate": 1.7128445796159677e-05, + "loss": 3.2017, + "step": 6735 + }, + { + "epoch": 0.5400573249684312, + "grad_norm": 1.8411444425582886, + "learning_rate": 1.7127535106530677e-05, + "loss": 3.2759, + "step": 6736 + }, + { + "epoch": 0.5401374997494538, + "grad_norm": 1.7827767133712769, + "learning_rate": 1.7126624296733946e-05, + "loss": 3.2111, + "step": 6737 + }, + { + "epoch": 0.5402176745304764, + "grad_norm": 1.7976094484329224, + "learning_rate": 1.7125713366784843e-05, + "loss": 3.3221, + "step": 6738 + }, + { + "epoch": 0.540297849311499, + "grad_norm": 1.9848434925079346, + "learning_rate": 1.7124802316698722e-05, + "loss": 3.2038, + "step": 6739 + }, + { + "epoch": 0.5403780240925217, + "grad_norm": 1.873957872390747, + "learning_rate": 1.7123891146490942e-05, + "loss": 3.2098, + "step": 6740 + }, + { + "epoch": 0.5404581988735443, + "grad_norm": 2.001692056655884, + "learning_rate": 1.7122979856176872e-05, + "loss": 3.2538, + "step": 6741 + }, + { + "epoch": 0.5405383736545669, + "grad_norm": 1.821621298789978, + "learning_rate": 1.712206844577187e-05, + "loss": 3.2627, + "step": 6742 + }, + { + "epoch": 0.5406185484355895, + "grad_norm": 1.8453264236450195, + "learning_rate": 1.7121156915291305e-05, + "loss": 3.2043, + "step": 6743 + }, + { + "epoch": 0.5406987232166122, + "grad_norm": 1.7768030166625977, + "learning_rate": 1.7120245264750545e-05, + "loss": 3.1801, + "step": 6744 + }, + { + "epoch": 0.5407788979976348, + "grad_norm": 1.8169126510620117, + "learning_rate": 1.7119333494164955e-05, + "loss": 3.228, + "step": 6745 + }, + { + "epoch": 0.5408590727786575, + "grad_norm": 1.756981372833252, + "learning_rate": 1.7118421603549914e-05, + "loss": 3.2716, + "step": 6746 + }, + { + "epoch": 0.5409392475596801, + "grad_norm": 1.8509386777877808, + "learning_rate": 1.7117509592920794e-05, + "loss": 3.2729, + "step": 6747 + }, + { + "epoch": 0.5410194223407028, + "grad_norm": 1.8668954372406006, + "learning_rate": 1.711659746229297e-05, + "loss": 3.2499, + "step": 6748 + }, + { + "epoch": 0.5410995971217254, + "grad_norm": 1.948418378829956, + "learning_rate": 1.7115685211681826e-05, + "loss": 3.3448, + "step": 6749 + }, + { + "epoch": 0.541179771902748, + "grad_norm": 1.789795160293579, + "learning_rate": 1.711477284110273e-05, + "loss": 3.3228, + "step": 6750 + }, + { + "epoch": 0.5412599466837706, + "grad_norm": 2.3741939067840576, + "learning_rate": 1.7113860350571077e-05, + "loss": 3.3121, + "step": 6751 + }, + { + "epoch": 0.5413401214647933, + "grad_norm": 1.869458556175232, + "learning_rate": 1.7112947740102243e-05, + "loss": 3.1819, + "step": 6752 + }, + { + "epoch": 0.5414202962458159, + "grad_norm": 1.9438459873199463, + "learning_rate": 1.7112035009711613e-05, + "loss": 3.2615, + "step": 6753 + }, + { + "epoch": 0.5415004710268385, + "grad_norm": 1.783921480178833, + "learning_rate": 1.7111122159414585e-05, + "loss": 3.2369, + "step": 6754 + }, + { + "epoch": 0.5415806458078611, + "grad_norm": 1.9883991479873657, + "learning_rate": 1.7110209189226543e-05, + "loss": 3.3579, + "step": 6755 + }, + { + "epoch": 0.5416608205888838, + "grad_norm": 1.870570182800293, + "learning_rate": 1.7109296099162877e-05, + "loss": 3.2327, + "step": 6756 + }, + { + "epoch": 0.5417409953699064, + "grad_norm": 1.8737410306930542, + "learning_rate": 1.7108382889238987e-05, + "loss": 3.202, + "step": 6757 + }, + { + "epoch": 0.541821170150929, + "grad_norm": 1.9509252309799194, + "learning_rate": 1.7107469559470263e-05, + "loss": 3.2565, + "step": 6758 + }, + { + "epoch": 0.5419013449319516, + "grad_norm": 2.0027904510498047, + "learning_rate": 1.7106556109872112e-05, + "loss": 3.2194, + "step": 6759 + }, + { + "epoch": 0.5419815197129743, + "grad_norm": 1.7383174896240234, + "learning_rate": 1.7105642540459925e-05, + "loss": 3.1859, + "step": 6760 + }, + { + "epoch": 0.5420616944939969, + "grad_norm": 0.8725689053535461, + "learning_rate": 1.7104728851249114e-05, + "loss": 2.5012, + "step": 6761 + }, + { + "epoch": 0.5421418692750195, + "grad_norm": 1.9418587684631348, + "learning_rate": 1.7103815042255077e-05, + "loss": 3.2317, + "step": 6762 + }, + { + "epoch": 0.5422220440560421, + "grad_norm": 1.976606011390686, + "learning_rate": 1.710290111349322e-05, + "loss": 3.2511, + "step": 6763 + }, + { + "epoch": 0.5423022188370648, + "grad_norm": 0.8012340664863586, + "learning_rate": 1.7101987064978955e-05, + "loss": 2.5051, + "step": 6764 + }, + { + "epoch": 0.5423823936180874, + "grad_norm": 1.8474762439727783, + "learning_rate": 1.7101072896727687e-05, + "loss": 3.2653, + "step": 6765 + }, + { + "epoch": 0.5424625683991101, + "grad_norm": 1.8904953002929688, + "learning_rate": 1.710015860875484e-05, + "loss": 3.1652, + "step": 6766 + }, + { + "epoch": 0.5425427431801327, + "grad_norm": 1.905781865119934, + "learning_rate": 1.7099244201075814e-05, + "loss": 3.2814, + "step": 6767 + }, + { + "epoch": 0.5426229179611554, + "grad_norm": 1.8799872398376465, + "learning_rate": 1.7098329673706036e-05, + "loss": 3.3367, + "step": 6768 + }, + { + "epoch": 0.542703092742178, + "grad_norm": 1.9129775762557983, + "learning_rate": 1.7097415026660922e-05, + "loss": 3.223, + "step": 6769 + }, + { + "epoch": 0.5427832675232006, + "grad_norm": 1.8475123643875122, + "learning_rate": 1.7096500259955894e-05, + "loss": 3.2618, + "step": 6770 + }, + { + "epoch": 0.5428634423042232, + "grad_norm": 1.8398292064666748, + "learning_rate": 1.709558537360637e-05, + "loss": 3.161, + "step": 6771 + }, + { + "epoch": 0.5429436170852459, + "grad_norm": 1.9201968908309937, + "learning_rate": 1.7094670367627775e-05, + "loss": 3.2302, + "step": 6772 + }, + { + "epoch": 0.5430237918662685, + "grad_norm": 1.877308964729309, + "learning_rate": 1.709375524203554e-05, + "loss": 3.2991, + "step": 6773 + }, + { + "epoch": 0.5431039666472911, + "grad_norm": 1.7856684923171997, + "learning_rate": 1.7092839996845093e-05, + "loss": 3.1711, + "step": 6774 + }, + { + "epoch": 0.5431841414283137, + "grad_norm": 1.925715684890747, + "learning_rate": 1.7091924632071862e-05, + "loss": 3.194, + "step": 6775 + }, + { + "epoch": 0.5432643162093364, + "grad_norm": 0.8193202018737793, + "learning_rate": 1.7091009147731285e-05, + "loss": 2.4883, + "step": 6776 + }, + { + "epoch": 0.543344490990359, + "grad_norm": 1.7544078826904297, + "learning_rate": 1.709009354383879e-05, + "loss": 3.2468, + "step": 6777 + }, + { + "epoch": 0.5434246657713816, + "grad_norm": 0.7952337861061096, + "learning_rate": 1.7089177820409813e-05, + "loss": 2.4976, + "step": 6778 + }, + { + "epoch": 0.5435048405524042, + "grad_norm": 1.7651183605194092, + "learning_rate": 1.7088261977459802e-05, + "loss": 3.21, + "step": 6779 + }, + { + "epoch": 0.5435850153334268, + "grad_norm": 1.9890013933181763, + "learning_rate": 1.708734601500419e-05, + "loss": 3.2507, + "step": 6780 + }, + { + "epoch": 0.5436651901144495, + "grad_norm": 1.9048185348510742, + "learning_rate": 1.7086429933058422e-05, + "loss": 3.2438, + "step": 6781 + }, + { + "epoch": 0.5437453648954721, + "grad_norm": 1.9746825695037842, + "learning_rate": 1.7085513731637942e-05, + "loss": 3.1708, + "step": 6782 + }, + { + "epoch": 0.5438255396764947, + "grad_norm": 1.8724645376205444, + "learning_rate": 1.70845974107582e-05, + "loss": 3.3453, + "step": 6783 + }, + { + "epoch": 0.5439057144575173, + "grad_norm": 1.862249732017517, + "learning_rate": 1.708368097043464e-05, + "loss": 3.2574, + "step": 6784 + }, + { + "epoch": 0.54398588923854, + "grad_norm": 1.7960779666900635, + "learning_rate": 1.7082764410682715e-05, + "loss": 3.2844, + "step": 6785 + }, + { + "epoch": 0.5440660640195627, + "grad_norm": 1.8921480178833008, + "learning_rate": 1.7081847731517882e-05, + "loss": 3.2686, + "step": 6786 + }, + { + "epoch": 0.5441462388005853, + "grad_norm": 1.84501051902771, + "learning_rate": 1.708093093295559e-05, + "loss": 3.1978, + "step": 6787 + }, + { + "epoch": 0.544226413581608, + "grad_norm": 1.8736077547073364, + "learning_rate": 1.7080014015011298e-05, + "loss": 3.1848, + "step": 6788 + }, + { + "epoch": 0.5443065883626306, + "grad_norm": 1.79838228225708, + "learning_rate": 1.707909697770046e-05, + "loss": 3.2294, + "step": 6789 + }, + { + "epoch": 0.5443867631436532, + "grad_norm": 1.7988094091415405, + "learning_rate": 1.707817982103855e-05, + "loss": 3.198, + "step": 6790 + }, + { + "epoch": 0.5444669379246758, + "grad_norm": 1.903623342514038, + "learning_rate": 1.7077262545041022e-05, + "loss": 3.2715, + "step": 6791 + }, + { + "epoch": 0.5445471127056984, + "grad_norm": 1.9077024459838867, + "learning_rate": 1.7076345149723334e-05, + "loss": 3.2832, + "step": 6792 + }, + { + "epoch": 0.5446272874867211, + "grad_norm": 1.8551291227340698, + "learning_rate": 1.7075427635100967e-05, + "loss": 3.1578, + "step": 6793 + }, + { + "epoch": 0.5447074622677437, + "grad_norm": 1.8876993656158447, + "learning_rate": 1.707451000118938e-05, + "loss": 3.314, + "step": 6794 + }, + { + "epoch": 0.5447876370487663, + "grad_norm": 2.376990556716919, + "learning_rate": 1.7073592248004054e-05, + "loss": 3.2202, + "step": 6795 + }, + { + "epoch": 0.5448678118297889, + "grad_norm": 1.8861336708068848, + "learning_rate": 1.7072674375560452e-05, + "loss": 3.2459, + "step": 6796 + }, + { + "epoch": 0.5449479866108116, + "grad_norm": 1.7948251962661743, + "learning_rate": 1.7071756383874052e-05, + "loss": 3.1567, + "step": 6797 + }, + { + "epoch": 0.5450281613918342, + "grad_norm": 1.8841508626937866, + "learning_rate": 1.707083827296033e-05, + "loss": 3.34, + "step": 6798 + }, + { + "epoch": 0.5451083361728568, + "grad_norm": 1.819350242614746, + "learning_rate": 1.706992004283477e-05, + "loss": 3.2249, + "step": 6799 + }, + { + "epoch": 0.5451885109538794, + "grad_norm": 1.890663743019104, + "learning_rate": 1.706900169351285e-05, + "loss": 3.3147, + "step": 6800 + }, + { + "epoch": 0.5452686857349021, + "grad_norm": 1.9304442405700684, + "learning_rate": 1.7068083225010054e-05, + "loss": 3.2985, + "step": 6801 + }, + { + "epoch": 0.5453488605159247, + "grad_norm": 1.732188105583191, + "learning_rate": 1.706716463734186e-05, + "loss": 3.2157, + "step": 6802 + }, + { + "epoch": 0.5454290352969473, + "grad_norm": 1.8506264686584473, + "learning_rate": 1.7066245930523764e-05, + "loss": 3.2573, + "step": 6803 + }, + { + "epoch": 0.5455092100779699, + "grad_norm": 1.8403366804122925, + "learning_rate": 1.7065327104571253e-05, + "loss": 3.2661, + "step": 6804 + }, + { + "epoch": 0.5455893848589926, + "grad_norm": 1.797363519668579, + "learning_rate": 1.7064408159499818e-05, + "loss": 3.1666, + "step": 6805 + }, + { + "epoch": 0.5456695596400152, + "grad_norm": 1.7347732782363892, + "learning_rate": 1.706348909532495e-05, + "loss": 3.212, + "step": 6806 + }, + { + "epoch": 0.5457497344210379, + "grad_norm": 1.8554213047027588, + "learning_rate": 1.7062569912062143e-05, + "loss": 3.2355, + "step": 6807 + }, + { + "epoch": 0.5458299092020605, + "grad_norm": 1.9390041828155518, + "learning_rate": 1.70616506097269e-05, + "loss": 3.2639, + "step": 6808 + }, + { + "epoch": 0.5459100839830832, + "grad_norm": 1.8260700702667236, + "learning_rate": 1.7060731188334715e-05, + "loss": 3.2415, + "step": 6809 + }, + { + "epoch": 0.5459902587641058, + "grad_norm": 1.7511159181594849, + "learning_rate": 1.7059811647901087e-05, + "loss": 3.2803, + "step": 6810 + }, + { + "epoch": 0.5460704335451284, + "grad_norm": 1.951966643333435, + "learning_rate": 1.705889198844153e-05, + "loss": 3.1763, + "step": 6811 + }, + { + "epoch": 0.546150608326151, + "grad_norm": 1.9104610681533813, + "learning_rate": 1.7057972209971538e-05, + "loss": 3.1815, + "step": 6812 + }, + { + "epoch": 0.5462307831071737, + "grad_norm": 1.8539928197860718, + "learning_rate": 1.7057052312506618e-05, + "loss": 3.2522, + "step": 6813 + }, + { + "epoch": 0.5463109578881963, + "grad_norm": 1.7646645307540894, + "learning_rate": 1.705613229606229e-05, + "loss": 3.1974, + "step": 6814 + }, + { + "epoch": 0.5463911326692189, + "grad_norm": 1.8589253425598145, + "learning_rate": 1.7055212160654054e-05, + "loss": 3.3483, + "step": 6815 + }, + { + "epoch": 0.5464713074502415, + "grad_norm": 1.816651701927185, + "learning_rate": 1.705429190629743e-05, + "loss": 3.3008, + "step": 6816 + }, + { + "epoch": 0.5465514822312642, + "grad_norm": 1.8159832954406738, + "learning_rate": 1.705337153300793e-05, + "loss": 3.2566, + "step": 6817 + }, + { + "epoch": 0.5466316570122868, + "grad_norm": 1.827050805091858, + "learning_rate": 1.705245104080107e-05, + "loss": 3.2657, + "step": 6818 + }, + { + "epoch": 0.5467118317933094, + "grad_norm": 1.8010053634643555, + "learning_rate": 1.7051530429692373e-05, + "loss": 3.2406, + "step": 6819 + }, + { + "epoch": 0.546792006574332, + "grad_norm": 1.9690899848937988, + "learning_rate": 1.7050609699697357e-05, + "loss": 3.2361, + "step": 6820 + }, + { + "epoch": 0.5468721813553546, + "grad_norm": 1.7159191370010376, + "learning_rate": 1.7049688850831545e-05, + "loss": 3.2991, + "step": 6821 + }, + { + "epoch": 0.5469523561363773, + "grad_norm": 0.8423767685890198, + "learning_rate": 1.7048767883110467e-05, + "loss": 2.4988, + "step": 6822 + }, + { + "epoch": 0.5470325309173999, + "grad_norm": 1.8554614782333374, + "learning_rate": 1.7047846796549647e-05, + "loss": 3.1737, + "step": 6823 + }, + { + "epoch": 0.5471127056984225, + "grad_norm": 1.84256911277771, + "learning_rate": 1.704692559116461e-05, + "loss": 3.2195, + "step": 6824 + }, + { + "epoch": 0.5471928804794451, + "grad_norm": 1.8848272562026978, + "learning_rate": 1.7046004266970894e-05, + "loss": 3.2721, + "step": 6825 + }, + { + "epoch": 0.5472730552604678, + "grad_norm": 1.747497320175171, + "learning_rate": 1.7045082823984027e-05, + "loss": 3.2104, + "step": 6826 + }, + { + "epoch": 0.5473532300414905, + "grad_norm": 2.0661580562591553, + "learning_rate": 1.704416126221955e-05, + "loss": 3.1816, + "step": 6827 + }, + { + "epoch": 0.5474334048225131, + "grad_norm": 1.8715580701828003, + "learning_rate": 1.704323958169299e-05, + "loss": 3.2271, + "step": 6828 + }, + { + "epoch": 0.5475135796035357, + "grad_norm": 1.8341730833053589, + "learning_rate": 1.7042317782419903e-05, + "loss": 3.1385, + "step": 6829 + }, + { + "epoch": 0.5475937543845584, + "grad_norm": 1.8840606212615967, + "learning_rate": 1.7041395864415814e-05, + "loss": 3.2535, + "step": 6830 + }, + { + "epoch": 0.547673929165581, + "grad_norm": 1.830082893371582, + "learning_rate": 1.7040473827696273e-05, + "loss": 3.1973, + "step": 6831 + }, + { + "epoch": 0.5477541039466036, + "grad_norm": 1.7371796369552612, + "learning_rate": 1.703955167227682e-05, + "loss": 3.1902, + "step": 6832 + }, + { + "epoch": 0.5478342787276262, + "grad_norm": 1.8953112363815308, + "learning_rate": 1.7038629398173013e-05, + "loss": 3.2146, + "step": 6833 + }, + { + "epoch": 0.5479144535086489, + "grad_norm": 1.733491063117981, + "learning_rate": 1.7037707005400394e-05, + "loss": 3.1082, + "step": 6834 + }, + { + "epoch": 0.5479946282896715, + "grad_norm": 1.7729171514511108, + "learning_rate": 1.7036784493974513e-05, + "loss": 3.1908, + "step": 6835 + }, + { + "epoch": 0.5480748030706941, + "grad_norm": 1.7648779153823853, + "learning_rate": 1.703586186391093e-05, + "loss": 3.186, + "step": 6836 + }, + { + "epoch": 0.5481549778517167, + "grad_norm": 1.8576672077178955, + "learning_rate": 1.703493911522519e-05, + "loss": 3.2529, + "step": 6837 + }, + { + "epoch": 0.5482351526327394, + "grad_norm": 1.7404751777648926, + "learning_rate": 1.703401624793286e-05, + "loss": 3.1365, + "step": 6838 + }, + { + "epoch": 0.548315327413762, + "grad_norm": 1.829351782798767, + "learning_rate": 1.7033093262049492e-05, + "loss": 3.2183, + "step": 6839 + }, + { + "epoch": 0.5483955021947846, + "grad_norm": 1.9218883514404297, + "learning_rate": 1.7032170157590653e-05, + "loss": 3.137, + "step": 6840 + }, + { + "epoch": 0.5484756769758072, + "grad_norm": 1.9745185375213623, + "learning_rate": 1.7031246934571905e-05, + "loss": 3.2156, + "step": 6841 + }, + { + "epoch": 0.5485558517568299, + "grad_norm": 1.8615854978561401, + "learning_rate": 1.7030323593008805e-05, + "loss": 3.1813, + "step": 6842 + }, + { + "epoch": 0.5486360265378525, + "grad_norm": 1.8834788799285889, + "learning_rate": 1.7029400132916932e-05, + "loss": 3.3015, + "step": 6843 + }, + { + "epoch": 0.5487162013188751, + "grad_norm": 1.7841273546218872, + "learning_rate": 1.702847655431185e-05, + "loss": 3.1987, + "step": 6844 + }, + { + "epoch": 0.5487963760998977, + "grad_norm": 1.9641361236572266, + "learning_rate": 1.702755285720913e-05, + "loss": 3.1981, + "step": 6845 + }, + { + "epoch": 0.5488765508809204, + "grad_norm": 1.8017714023590088, + "learning_rate": 1.7026629041624344e-05, + "loss": 3.1407, + "step": 6846 + }, + { + "epoch": 0.5489567256619431, + "grad_norm": 1.728115200996399, + "learning_rate": 1.702570510757307e-05, + "loss": 3.1141, + "step": 6847 + }, + { + "epoch": 0.5490369004429657, + "grad_norm": 1.7928287982940674, + "learning_rate": 1.702478105507089e-05, + "loss": 3.1964, + "step": 6848 + }, + { + "epoch": 0.5491170752239883, + "grad_norm": 0.8473199009895325, + "learning_rate": 1.7023856884133368e-05, + "loss": 2.5122, + "step": 6849 + }, + { + "epoch": 0.549197250005011, + "grad_norm": 2.0307013988494873, + "learning_rate": 1.7022932594776097e-05, + "loss": 3.2053, + "step": 6850 + }, + { + "epoch": 0.5492774247860336, + "grad_norm": 1.9788661003112793, + "learning_rate": 1.702200818701466e-05, + "loss": 3.1661, + "step": 6851 + }, + { + "epoch": 0.5493575995670562, + "grad_norm": 1.9544118642807007, + "learning_rate": 1.7021083660864635e-05, + "loss": 3.222, + "step": 6852 + }, + { + "epoch": 0.5494377743480788, + "grad_norm": 1.9086577892303467, + "learning_rate": 1.7020159016341618e-05, + "loss": 3.3042, + "step": 6853 + }, + { + "epoch": 0.5495179491291015, + "grad_norm": 1.8407490253448486, + "learning_rate": 1.7019234253461193e-05, + "loss": 3.2618, + "step": 6854 + }, + { + "epoch": 0.5495981239101241, + "grad_norm": 1.886410117149353, + "learning_rate": 1.7018309372238953e-05, + "loss": 3.1747, + "step": 6855 + }, + { + "epoch": 0.5496782986911467, + "grad_norm": 1.760804295539856, + "learning_rate": 1.701738437269049e-05, + "loss": 3.1953, + "step": 6856 + }, + { + "epoch": 0.5497584734721693, + "grad_norm": 1.7842044830322266, + "learning_rate": 1.7016459254831397e-05, + "loss": 3.244, + "step": 6857 + }, + { + "epoch": 0.549838648253192, + "grad_norm": 1.9168505668640137, + "learning_rate": 1.701553401867728e-05, + "loss": 3.2648, + "step": 6858 + }, + { + "epoch": 0.5499188230342146, + "grad_norm": 1.8619599342346191, + "learning_rate": 1.7014608664243725e-05, + "loss": 3.2472, + "step": 6859 + }, + { + "epoch": 0.5499989978152372, + "grad_norm": 1.8337359428405762, + "learning_rate": 1.7013683191546343e-05, + "loss": 3.2979, + "step": 6860 + }, + { + "epoch": 0.5500791725962598, + "grad_norm": 0.7801663875579834, + "learning_rate": 1.7012757600600736e-05, + "loss": 2.5044, + "step": 6861 + }, + { + "epoch": 0.5501593473772824, + "grad_norm": 1.856777310371399, + "learning_rate": 1.7011831891422505e-05, + "loss": 3.2559, + "step": 6862 + }, + { + "epoch": 0.5502395221583051, + "grad_norm": 1.9211403131484985, + "learning_rate": 1.701090606402726e-05, + "loss": 3.3567, + "step": 6863 + }, + { + "epoch": 0.5503196969393277, + "grad_norm": 1.8629915714263916, + "learning_rate": 1.7009980118430612e-05, + "loss": 3.3098, + "step": 6864 + }, + { + "epoch": 0.5503998717203503, + "grad_norm": 0.8144739270210266, + "learning_rate": 1.7009054054648167e-05, + "loss": 2.4998, + "step": 6865 + }, + { + "epoch": 0.5504800465013729, + "grad_norm": 1.7962050437927246, + "learning_rate": 1.700812787269554e-05, + "loss": 3.2965, + "step": 6866 + }, + { + "epoch": 0.5505602212823956, + "grad_norm": 1.8489400148391724, + "learning_rate": 1.700720157258835e-05, + "loss": 3.2018, + "step": 6867 + }, + { + "epoch": 0.5506403960634183, + "grad_norm": 1.846073865890503, + "learning_rate": 1.7006275154342212e-05, + "loss": 3.1851, + "step": 6868 + }, + { + "epoch": 0.5507205708444409, + "grad_norm": 2.029036521911621, + "learning_rate": 1.700534861797274e-05, + "loss": 3.2257, + "step": 6869 + }, + { + "epoch": 0.5508007456254636, + "grad_norm": 2.016366481781006, + "learning_rate": 1.700442196349556e-05, + "loss": 3.2742, + "step": 6870 + }, + { + "epoch": 0.5508809204064862, + "grad_norm": 1.7749274969100952, + "learning_rate": 1.7003495190926295e-05, + "loss": 3.1382, + "step": 6871 + }, + { + "epoch": 0.5509610951875088, + "grad_norm": 1.8308039903640747, + "learning_rate": 1.700256830028057e-05, + "loss": 3.1125, + "step": 6872 + }, + { + "epoch": 0.5510412699685314, + "grad_norm": 0.8197072148323059, + "learning_rate": 1.7001641291574015e-05, + "loss": 2.4913, + "step": 6873 + }, + { + "epoch": 0.551121444749554, + "grad_norm": 1.8011733293533325, + "learning_rate": 1.700071416482225e-05, + "loss": 3.1216, + "step": 6874 + }, + { + "epoch": 0.5512016195305767, + "grad_norm": 1.9689373970031738, + "learning_rate": 1.6999786920040914e-05, + "loss": 3.1849, + "step": 6875 + }, + { + "epoch": 0.5512817943115993, + "grad_norm": 1.9463427066802979, + "learning_rate": 1.6998859557245634e-05, + "loss": 3.2654, + "step": 6876 + }, + { + "epoch": 0.5513619690926219, + "grad_norm": 0.7674247026443481, + "learning_rate": 1.699793207645205e-05, + "loss": 2.4954, + "step": 6877 + }, + { + "epoch": 0.5514421438736445, + "grad_norm": 1.7816452980041504, + "learning_rate": 1.6997004477675802e-05, + "loss": 3.191, + "step": 6878 + }, + { + "epoch": 0.5515223186546672, + "grad_norm": 1.8977514505386353, + "learning_rate": 1.699607676093252e-05, + "loss": 3.323, + "step": 6879 + }, + { + "epoch": 0.5516024934356898, + "grad_norm": 0.785310685634613, + "learning_rate": 1.699514892623785e-05, + "loss": 2.4885, + "step": 6880 + }, + { + "epoch": 0.5516826682167124, + "grad_norm": 1.7531161308288574, + "learning_rate": 1.6994220973607434e-05, + "loss": 3.2439, + "step": 6881 + }, + { + "epoch": 0.551762842997735, + "grad_norm": 1.914648175239563, + "learning_rate": 1.699329290305692e-05, + "loss": 3.2081, + "step": 6882 + }, + { + "epoch": 0.5518430177787577, + "grad_norm": 1.7944549322128296, + "learning_rate": 1.699236471460195e-05, + "loss": 3.2614, + "step": 6883 + }, + { + "epoch": 0.5519231925597803, + "grad_norm": 1.7482075691223145, + "learning_rate": 1.6991436408258178e-05, + "loss": 3.1339, + "step": 6884 + }, + { + "epoch": 0.5520033673408029, + "grad_norm": 1.9468857049942017, + "learning_rate": 1.699050798404125e-05, + "loss": 3.2115, + "step": 6885 + }, + { + "epoch": 0.5520835421218255, + "grad_norm": 1.950496792793274, + "learning_rate": 1.698957944196682e-05, + "loss": 3.323, + "step": 6886 + }, + { + "epoch": 0.5521637169028482, + "grad_norm": 1.7830625772476196, + "learning_rate": 1.6988650782050543e-05, + "loss": 3.261, + "step": 6887 + }, + { + "epoch": 0.5522438916838709, + "grad_norm": 1.9882112741470337, + "learning_rate": 1.698772200430808e-05, + "loss": 3.3355, + "step": 6888 + }, + { + "epoch": 0.5523240664648935, + "grad_norm": 1.7214303016662598, + "learning_rate": 1.6986793108755086e-05, + "loss": 3.1311, + "step": 6889 + }, + { + "epoch": 0.5524042412459161, + "grad_norm": 1.7864408493041992, + "learning_rate": 1.6985864095407224e-05, + "loss": 3.2548, + "step": 6890 + }, + { + "epoch": 0.5524844160269388, + "grad_norm": 1.7620933055877686, + "learning_rate": 1.6984934964280153e-05, + "loss": 3.1011, + "step": 6891 + }, + { + "epoch": 0.5525645908079614, + "grad_norm": 1.722398042678833, + "learning_rate": 1.698400571538954e-05, + "loss": 3.3079, + "step": 6892 + }, + { + "epoch": 0.552644765588984, + "grad_norm": 1.7276490926742554, + "learning_rate": 1.698307634875105e-05, + "loss": 3.1398, + "step": 6893 + }, + { + "epoch": 0.5527249403700066, + "grad_norm": 1.8487271070480347, + "learning_rate": 1.698214686438036e-05, + "loss": 3.3043, + "step": 6894 + }, + { + "epoch": 0.5528051151510293, + "grad_norm": 1.9444066286087036, + "learning_rate": 1.6981217262293133e-05, + "loss": 3.3422, + "step": 6895 + }, + { + "epoch": 0.5528852899320519, + "grad_norm": 1.766228437423706, + "learning_rate": 1.698028754250504e-05, + "loss": 3.2714, + "step": 6896 + }, + { + "epoch": 0.5529654647130745, + "grad_norm": 2.022465467453003, + "learning_rate": 1.697935770503176e-05, + "loss": 3.3107, + "step": 6897 + }, + { + "epoch": 0.5530456394940971, + "grad_norm": 1.8667752742767334, + "learning_rate": 1.6978427749888972e-05, + "loss": 3.1923, + "step": 6898 + }, + { + "epoch": 0.5531258142751198, + "grad_norm": 1.8032746315002441, + "learning_rate": 1.697749767709235e-05, + "loss": 3.2748, + "step": 6899 + }, + { + "epoch": 0.5532059890561424, + "grad_norm": 1.877306580543518, + "learning_rate": 1.6976567486657575e-05, + "loss": 3.244, + "step": 6900 + }, + { + "epoch": 0.553286163837165, + "grad_norm": 0.7687788009643555, + "learning_rate": 1.697563717860033e-05, + "loss": 2.5063, + "step": 6901 + }, + { + "epoch": 0.5533663386181876, + "grad_norm": 1.8055287599563599, + "learning_rate": 1.6974706752936306e-05, + "loss": 3.1566, + "step": 6902 + }, + { + "epoch": 0.5534465133992102, + "grad_norm": 1.8452084064483643, + "learning_rate": 1.6973776209681178e-05, + "loss": 3.1904, + "step": 6903 + }, + { + "epoch": 0.5535266881802329, + "grad_norm": 1.8730942010879517, + "learning_rate": 1.6972845548850646e-05, + "loss": 3.2241, + "step": 6904 + }, + { + "epoch": 0.5536068629612555, + "grad_norm": 1.8809537887573242, + "learning_rate": 1.697191477046039e-05, + "loss": 3.2301, + "step": 6905 + }, + { + "epoch": 0.5536870377422781, + "grad_norm": 2.1940102577209473, + "learning_rate": 1.697098387452611e-05, + "loss": 3.2241, + "step": 6906 + }, + { + "epoch": 0.5537672125233007, + "grad_norm": 1.746050238609314, + "learning_rate": 1.6970052861063503e-05, + "loss": 3.1829, + "step": 6907 + }, + { + "epoch": 0.5538473873043235, + "grad_norm": 1.8758655786514282, + "learning_rate": 1.6969121730088255e-05, + "loss": 3.2297, + "step": 6908 + }, + { + "epoch": 0.5539275620853461, + "grad_norm": 1.8557474613189697, + "learning_rate": 1.6968190481616077e-05, + "loss": 3.2037, + "step": 6909 + }, + { + "epoch": 0.5540077368663687, + "grad_norm": 1.743957757949829, + "learning_rate": 1.696725911566266e-05, + "loss": 3.2262, + "step": 6910 + }, + { + "epoch": 0.5540879116473914, + "grad_norm": 1.7973136901855469, + "learning_rate": 1.6966327632243706e-05, + "loss": 3.3428, + "step": 6911 + }, + { + "epoch": 0.554168086428414, + "grad_norm": 0.7561852335929871, + "learning_rate": 1.696539603137493e-05, + "loss": 2.4792, + "step": 6912 + }, + { + "epoch": 0.5542482612094366, + "grad_norm": 1.7833856344223022, + "learning_rate": 1.696446431307203e-05, + "loss": 3.1964, + "step": 6913 + }, + { + "epoch": 0.5543284359904592, + "grad_norm": 1.9045681953430176, + "learning_rate": 1.6963532477350715e-05, + "loss": 3.2837, + "step": 6914 + }, + { + "epoch": 0.5544086107714818, + "grad_norm": 1.8617548942565918, + "learning_rate": 1.6962600524226695e-05, + "loss": 3.2793, + "step": 6915 + }, + { + "epoch": 0.5544887855525045, + "grad_norm": 1.9610611200332642, + "learning_rate": 1.6961668453715688e-05, + "loss": 3.2106, + "step": 6916 + }, + { + "epoch": 0.5545689603335271, + "grad_norm": 0.7579757571220398, + "learning_rate": 1.69607362658334e-05, + "loss": 2.5051, + "step": 6917 + }, + { + "epoch": 0.5546491351145497, + "grad_norm": 1.8232883214950562, + "learning_rate": 1.6959803960595557e-05, + "loss": 3.1722, + "step": 6918 + }, + { + "epoch": 0.5547293098955723, + "grad_norm": 1.7975620031356812, + "learning_rate": 1.6958871538017868e-05, + "loss": 3.1923, + "step": 6919 + }, + { + "epoch": 0.554809484676595, + "grad_norm": 1.8524686098098755, + "learning_rate": 1.6957938998116055e-05, + "loss": 3.2196, + "step": 6920 + }, + { + "epoch": 0.5548896594576176, + "grad_norm": 1.7987604141235352, + "learning_rate": 1.6957006340905845e-05, + "loss": 3.2135, + "step": 6921 + }, + { + "epoch": 0.5549698342386402, + "grad_norm": 1.8627516031265259, + "learning_rate": 1.6956073566402964e-05, + "loss": 3.3398, + "step": 6922 + }, + { + "epoch": 0.5550500090196628, + "grad_norm": 2.054405927658081, + "learning_rate": 1.695514067462313e-05, + "loss": 3.1168, + "step": 6923 + }, + { + "epoch": 0.5551301838006855, + "grad_norm": 1.8394943475723267, + "learning_rate": 1.6954207665582078e-05, + "loss": 3.2244, + "step": 6924 + }, + { + "epoch": 0.5552103585817081, + "grad_norm": 1.8346319198608398, + "learning_rate": 1.6953274539295535e-05, + "loss": 3.2745, + "step": 6925 + }, + { + "epoch": 0.5552905333627307, + "grad_norm": 2.0843026638031006, + "learning_rate": 1.6952341295779232e-05, + "loss": 3.2742, + "step": 6926 + }, + { + "epoch": 0.5553707081437533, + "grad_norm": 1.9104080200195312, + "learning_rate": 1.6951407935048907e-05, + "loss": 3.2542, + "step": 6927 + }, + { + "epoch": 0.555450882924776, + "grad_norm": 1.8436030149459839, + "learning_rate": 1.6950474457120295e-05, + "loss": 3.1819, + "step": 6928 + }, + { + "epoch": 0.5555310577057987, + "grad_norm": 1.7701832056045532, + "learning_rate": 1.694954086200913e-05, + "loss": 3.1752, + "step": 6929 + }, + { + "epoch": 0.5556112324868213, + "grad_norm": 1.8664482831954956, + "learning_rate": 1.6948607149731157e-05, + "loss": 3.1669, + "step": 6930 + }, + { + "epoch": 0.5556914072678439, + "grad_norm": 1.8210532665252686, + "learning_rate": 1.6947673320302117e-05, + "loss": 3.2462, + "step": 6931 + }, + { + "epoch": 0.5557715820488666, + "grad_norm": 1.8233435153961182, + "learning_rate": 1.694673937373775e-05, + "loss": 3.2155, + "step": 6932 + }, + { + "epoch": 0.5558517568298892, + "grad_norm": 1.8810392618179321, + "learning_rate": 1.694580531005381e-05, + "loss": 3.2019, + "step": 6933 + }, + { + "epoch": 0.5559319316109118, + "grad_norm": 0.7880833148956299, + "learning_rate": 1.694487112926604e-05, + "loss": 2.4753, + "step": 6934 + }, + { + "epoch": 0.5560121063919344, + "grad_norm": 1.7759768962860107, + "learning_rate": 1.6943936831390187e-05, + "loss": 3.2098, + "step": 6935 + }, + { + "epoch": 0.5560922811729571, + "grad_norm": 1.7484638690948486, + "learning_rate": 1.694300241644201e-05, + "loss": 3.2148, + "step": 6936 + }, + { + "epoch": 0.5561724559539797, + "grad_norm": 0.7419636845588684, + "learning_rate": 1.694206788443726e-05, + "loss": 2.4854, + "step": 6937 + }, + { + "epoch": 0.5562526307350023, + "grad_norm": 0.7511178851127625, + "learning_rate": 1.6941133235391686e-05, + "loss": 2.502, + "step": 6938 + }, + { + "epoch": 0.5563328055160249, + "grad_norm": 1.8431406021118164, + "learning_rate": 1.6940198469321056e-05, + "loss": 3.1908, + "step": 6939 + }, + { + "epoch": 0.5564129802970476, + "grad_norm": 1.7337027788162231, + "learning_rate": 1.6939263586241126e-05, + "loss": 3.2419, + "step": 6940 + }, + { + "epoch": 0.5564931550780702, + "grad_norm": 1.8694974184036255, + "learning_rate": 1.6938328586167657e-05, + "loss": 3.2114, + "step": 6941 + }, + { + "epoch": 0.5565733298590928, + "grad_norm": 1.8014321327209473, + "learning_rate": 1.6937393469116413e-05, + "loss": 3.1962, + "step": 6942 + }, + { + "epoch": 0.5566535046401154, + "grad_norm": 1.7792195081710815, + "learning_rate": 1.693645823510316e-05, + "loss": 3.1937, + "step": 6943 + }, + { + "epoch": 0.556733679421138, + "grad_norm": 1.810038447380066, + "learning_rate": 1.6935522884143667e-05, + "loss": 3.2277, + "step": 6944 + }, + { + "epoch": 0.5568138542021607, + "grad_norm": 1.7493493556976318, + "learning_rate": 1.69345874162537e-05, + "loss": 3.1874, + "step": 6945 + }, + { + "epoch": 0.5568940289831833, + "grad_norm": 1.7686090469360352, + "learning_rate": 1.6933651831449035e-05, + "loss": 3.1129, + "step": 6946 + }, + { + "epoch": 0.5569742037642059, + "grad_norm": 1.7864596843719482, + "learning_rate": 1.6932716129745442e-05, + "loss": 3.2075, + "step": 6947 + }, + { + "epoch": 0.5570543785452285, + "grad_norm": 1.8984228372573853, + "learning_rate": 1.6931780311158703e-05, + "loss": 3.2616, + "step": 6948 + }, + { + "epoch": 0.5571345533262513, + "grad_norm": 1.87833571434021, + "learning_rate": 1.693084437570459e-05, + "loss": 3.3468, + "step": 6949 + }, + { + "epoch": 0.5572147281072739, + "grad_norm": 1.8050847053527832, + "learning_rate": 1.6929908323398877e-05, + "loss": 3.2405, + "step": 6950 + }, + { + "epoch": 0.5572949028882965, + "grad_norm": 1.8032971620559692, + "learning_rate": 1.6928972154257358e-05, + "loss": 3.1199, + "step": 6951 + }, + { + "epoch": 0.5573750776693192, + "grad_norm": 1.8865876197814941, + "learning_rate": 1.692803586829581e-05, + "loss": 3.2108, + "step": 6952 + }, + { + "epoch": 0.5574552524503418, + "grad_norm": 1.9059191942214966, + "learning_rate": 1.6927099465530016e-05, + "loss": 3.2604, + "step": 6953 + }, + { + "epoch": 0.5575354272313644, + "grad_norm": 1.726750135421753, + "learning_rate": 1.6926162945975767e-05, + "loss": 3.1527, + "step": 6954 + }, + { + "epoch": 0.557615602012387, + "grad_norm": 1.7929813861846924, + "learning_rate": 1.6925226309648855e-05, + "loss": 3.2072, + "step": 6955 + }, + { + "epoch": 0.5576957767934096, + "grad_norm": 2.0389151573181152, + "learning_rate": 1.6924289556565065e-05, + "loss": 3.2394, + "step": 6956 + }, + { + "epoch": 0.5577759515744323, + "grad_norm": 1.9319838285446167, + "learning_rate": 1.6923352686740195e-05, + "loss": 3.2491, + "step": 6957 + }, + { + "epoch": 0.5578561263554549, + "grad_norm": 2.0087926387786865, + "learning_rate": 1.692241570019004e-05, + "loss": 3.1438, + "step": 6958 + }, + { + "epoch": 0.5579363011364775, + "grad_norm": 1.8602724075317383, + "learning_rate": 1.6921478596930393e-05, + "loss": 3.2803, + "step": 6959 + }, + { + "epoch": 0.5580164759175001, + "grad_norm": 1.9804513454437256, + "learning_rate": 1.6920541376977055e-05, + "loss": 3.2328, + "step": 6960 + }, + { + "epoch": 0.5580966506985228, + "grad_norm": 1.792373776435852, + "learning_rate": 1.6919604040345833e-05, + "loss": 3.2194, + "step": 6961 + }, + { + "epoch": 0.5581768254795454, + "grad_norm": 1.9654756784439087, + "learning_rate": 1.6918666587052523e-05, + "loss": 3.2869, + "step": 6962 + }, + { + "epoch": 0.558257000260568, + "grad_norm": 0.8275611400604248, + "learning_rate": 1.6917729017112934e-05, + "loss": 2.5026, + "step": 6963 + }, + { + "epoch": 0.5583371750415906, + "grad_norm": 1.7989457845687866, + "learning_rate": 1.691679133054287e-05, + "loss": 3.1869, + "step": 6964 + }, + { + "epoch": 0.5584173498226133, + "grad_norm": 1.7445348501205444, + "learning_rate": 1.6915853527358142e-05, + "loss": 3.1694, + "step": 6965 + }, + { + "epoch": 0.5584975246036359, + "grad_norm": 1.8083211183547974, + "learning_rate": 1.6914915607574562e-05, + "loss": 3.1993, + "step": 6966 + }, + { + "epoch": 0.5585776993846585, + "grad_norm": 0.7549558281898499, + "learning_rate": 1.6913977571207942e-05, + "loss": 2.4982, + "step": 6967 + }, + { + "epoch": 0.5586578741656811, + "grad_norm": 1.771886944770813, + "learning_rate": 1.6913039418274097e-05, + "loss": 3.099, + "step": 6968 + }, + { + "epoch": 0.5587380489467039, + "grad_norm": 2.038719415664673, + "learning_rate": 1.691210114878884e-05, + "loss": 3.2288, + "step": 6969 + }, + { + "epoch": 0.5588182237277265, + "grad_norm": 1.7941914796829224, + "learning_rate": 1.6911162762768e-05, + "loss": 3.1454, + "step": 6970 + }, + { + "epoch": 0.5588983985087491, + "grad_norm": 1.8224005699157715, + "learning_rate": 1.691022426022739e-05, + "loss": 3.2461, + "step": 6971 + }, + { + "epoch": 0.5589785732897717, + "grad_norm": 1.8633097410202026, + "learning_rate": 1.6909285641182833e-05, + "loss": 3.3777, + "step": 6972 + }, + { + "epoch": 0.5590587480707944, + "grad_norm": 1.8182861804962158, + "learning_rate": 1.6908346905650153e-05, + "loss": 3.2728, + "step": 6973 + }, + { + "epoch": 0.559138922851817, + "grad_norm": 1.8876533508300781, + "learning_rate": 1.6907408053645184e-05, + "loss": 3.2523, + "step": 6974 + }, + { + "epoch": 0.5592190976328396, + "grad_norm": 1.7833795547485352, + "learning_rate": 1.6906469085183745e-05, + "loss": 3.265, + "step": 6975 + }, + { + "epoch": 0.5592992724138622, + "grad_norm": 1.8956794738769531, + "learning_rate": 1.6905530000281674e-05, + "loss": 3.231, + "step": 6976 + }, + { + "epoch": 0.5593794471948849, + "grad_norm": 1.9500494003295898, + "learning_rate": 1.6904590798954798e-05, + "loss": 3.177, + "step": 6977 + }, + { + "epoch": 0.5594596219759075, + "grad_norm": 1.871625542640686, + "learning_rate": 1.6903651481218958e-05, + "loss": 3.2488, + "step": 6978 + }, + { + "epoch": 0.5595397967569301, + "grad_norm": 0.7737019658088684, + "learning_rate": 1.6902712047089983e-05, + "loss": 2.4944, + "step": 6979 + }, + { + "epoch": 0.5596199715379527, + "grad_norm": 1.8950670957565308, + "learning_rate": 1.6901772496583716e-05, + "loss": 3.2821, + "step": 6980 + }, + { + "epoch": 0.5597001463189754, + "grad_norm": 1.9641225337982178, + "learning_rate": 1.6900832829716e-05, + "loss": 3.3184, + "step": 6981 + }, + { + "epoch": 0.559780321099998, + "grad_norm": 0.8467822670936584, + "learning_rate": 1.6899893046502676e-05, + "loss": 2.517, + "step": 6982 + }, + { + "epoch": 0.5598604958810206, + "grad_norm": 1.918899655342102, + "learning_rate": 1.6898953146959586e-05, + "loss": 3.2207, + "step": 6983 + }, + { + "epoch": 0.5599406706620432, + "grad_norm": 1.9531593322753906, + "learning_rate": 1.6898013131102576e-05, + "loss": 3.1802, + "step": 6984 + }, + { + "epoch": 0.5600208454430659, + "grad_norm": 1.9056415557861328, + "learning_rate": 1.6897072998947495e-05, + "loss": 3.1949, + "step": 6985 + }, + { + "epoch": 0.5601010202240885, + "grad_norm": 1.9253973960876465, + "learning_rate": 1.6896132750510195e-05, + "loss": 3.2336, + "step": 6986 + }, + { + "epoch": 0.5601811950051111, + "grad_norm": 1.6953880786895752, + "learning_rate": 1.6895192385806528e-05, + "loss": 3.2252, + "step": 6987 + }, + { + "epoch": 0.5602613697861337, + "grad_norm": 1.8340981006622314, + "learning_rate": 1.6894251904852346e-05, + "loss": 3.0923, + "step": 6988 + }, + { + "epoch": 0.5603415445671563, + "grad_norm": 2.0338523387908936, + "learning_rate": 1.6893311307663508e-05, + "loss": 3.2039, + "step": 6989 + }, + { + "epoch": 0.5604217193481791, + "grad_norm": 1.8382045030593872, + "learning_rate": 1.689237059425587e-05, + "loss": 3.1409, + "step": 6990 + }, + { + "epoch": 0.5605018941292017, + "grad_norm": 1.8390122652053833, + "learning_rate": 1.6891429764645293e-05, + "loss": 3.2815, + "step": 6991 + }, + { + "epoch": 0.5605820689102243, + "grad_norm": 2.08298659324646, + "learning_rate": 1.6890488818847635e-05, + "loss": 3.2236, + "step": 6992 + }, + { + "epoch": 0.560662243691247, + "grad_norm": 1.8564801216125488, + "learning_rate": 1.688954775687877e-05, + "loss": 3.1479, + "step": 6993 + }, + { + "epoch": 0.5607424184722696, + "grad_norm": 1.701566457748413, + "learning_rate": 1.6888606578754556e-05, + "loss": 3.1541, + "step": 6994 + }, + { + "epoch": 0.5608225932532922, + "grad_norm": 1.8276375532150269, + "learning_rate": 1.688766528449086e-05, + "loss": 3.2819, + "step": 6995 + }, + { + "epoch": 0.5609027680343148, + "grad_norm": 1.8142427206039429, + "learning_rate": 1.688672387410356e-05, + "loss": 3.2207, + "step": 6996 + }, + { + "epoch": 0.5609829428153374, + "grad_norm": 1.8266521692276, + "learning_rate": 1.688578234760852e-05, + "loss": 3.2506, + "step": 6997 + }, + { + "epoch": 0.5610631175963601, + "grad_norm": 2.010287284851074, + "learning_rate": 1.688484070502162e-05, + "loss": 3.1867, + "step": 6998 + }, + { + "epoch": 0.5611432923773827, + "grad_norm": 1.8535327911376953, + "learning_rate": 1.6883898946358725e-05, + "loss": 3.2358, + "step": 6999 + }, + { + "epoch": 0.5612234671584053, + "grad_norm": 1.8041069507598877, + "learning_rate": 1.6882957071635724e-05, + "loss": 3.1747, + "step": 7000 + }, + { + "epoch": 0.5613036419394279, + "grad_norm": 1.977617621421814, + "learning_rate": 1.688201508086849e-05, + "loss": 3.2954, + "step": 7001 + }, + { + "epoch": 0.5613838167204506, + "grad_norm": 1.9670203924179077, + "learning_rate": 1.688107297407291e-05, + "loss": 3.223, + "step": 7002 + }, + { + "epoch": 0.5614639915014732, + "grad_norm": 1.9862757921218872, + "learning_rate": 1.6880130751264868e-05, + "loss": 3.2222, + "step": 7003 + }, + { + "epoch": 0.5615441662824958, + "grad_norm": 1.887589693069458, + "learning_rate": 1.687918841246024e-05, + "loss": 3.1705, + "step": 7004 + }, + { + "epoch": 0.5616243410635184, + "grad_norm": 0.7922408580780029, + "learning_rate": 1.6878245957674922e-05, + "loss": 2.4883, + "step": 7005 + }, + { + "epoch": 0.5617045158445411, + "grad_norm": 1.799497365951538, + "learning_rate": 1.6877303386924803e-05, + "loss": 3.2306, + "step": 7006 + }, + { + "epoch": 0.5617846906255637, + "grad_norm": 1.805954098701477, + "learning_rate": 1.687636070022577e-05, + "loss": 3.2323, + "step": 7007 + }, + { + "epoch": 0.5618648654065863, + "grad_norm": 1.8613693714141846, + "learning_rate": 1.687541789759372e-05, + "loss": 3.2438, + "step": 7008 + }, + { + "epoch": 0.5619450401876089, + "grad_norm": 1.962853193283081, + "learning_rate": 1.6874474979044548e-05, + "loss": 3.1483, + "step": 7009 + }, + { + "epoch": 0.5620252149686317, + "grad_norm": 1.8011313676834106, + "learning_rate": 1.687353194459415e-05, + "loss": 3.2495, + "step": 7010 + }, + { + "epoch": 0.5621053897496543, + "grad_norm": 1.8661367893218994, + "learning_rate": 1.6872588794258423e-05, + "loss": 3.228, + "step": 7011 + }, + { + "epoch": 0.5621855645306769, + "grad_norm": 1.786414623260498, + "learning_rate": 1.6871645528053275e-05, + "loss": 3.2437, + "step": 7012 + }, + { + "epoch": 0.5622657393116995, + "grad_norm": 2.122657537460327, + "learning_rate": 1.6870702145994605e-05, + "loss": 3.1277, + "step": 7013 + }, + { + "epoch": 0.5623459140927222, + "grad_norm": 1.833551049232483, + "learning_rate": 1.686975864809832e-05, + "loss": 3.1511, + "step": 7014 + }, + { + "epoch": 0.5624260888737448, + "grad_norm": 1.940111756324768, + "learning_rate": 1.686881503438032e-05, + "loss": 3.2417, + "step": 7015 + }, + { + "epoch": 0.5625062636547674, + "grad_norm": 1.8090745210647583, + "learning_rate": 1.6867871304856525e-05, + "loss": 3.2847, + "step": 7016 + }, + { + "epoch": 0.56258643843579, + "grad_norm": 1.8061871528625488, + "learning_rate": 1.6866927459542838e-05, + "loss": 3.2976, + "step": 7017 + }, + { + "epoch": 0.5626666132168127, + "grad_norm": 1.832078456878662, + "learning_rate": 1.686598349845517e-05, + "loss": 3.2542, + "step": 7018 + }, + { + "epoch": 0.5627467879978353, + "grad_norm": 1.9298158884048462, + "learning_rate": 1.6865039421609445e-05, + "loss": 3.2192, + "step": 7019 + }, + { + "epoch": 0.5628269627788579, + "grad_norm": 1.737235426902771, + "learning_rate": 1.6864095229021574e-05, + "loss": 3.1595, + "step": 7020 + }, + { + "epoch": 0.5629071375598805, + "grad_norm": 1.7789164781570435, + "learning_rate": 1.6863150920707475e-05, + "loss": 3.177, + "step": 7021 + }, + { + "epoch": 0.5629873123409032, + "grad_norm": 1.8414841890335083, + "learning_rate": 1.6862206496683074e-05, + "loss": 3.27, + "step": 7022 + }, + { + "epoch": 0.5630674871219258, + "grad_norm": 1.822208285331726, + "learning_rate": 1.686126195696429e-05, + "loss": 3.1914, + "step": 7023 + }, + { + "epoch": 0.5631476619029484, + "grad_norm": 0.8335476517677307, + "learning_rate": 1.6860317301567045e-05, + "loss": 2.5049, + "step": 7024 + }, + { + "epoch": 0.563227836683971, + "grad_norm": 1.7478488683700562, + "learning_rate": 1.6859372530507265e-05, + "loss": 3.1211, + "step": 7025 + }, + { + "epoch": 0.5633080114649937, + "grad_norm": 0.7903159856796265, + "learning_rate": 1.6858427643800885e-05, + "loss": 2.511, + "step": 7026 + }, + { + "epoch": 0.5633881862460163, + "grad_norm": 1.8893952369689941, + "learning_rate": 1.685748264146383e-05, + "loss": 3.2374, + "step": 7027 + }, + { + "epoch": 0.5634683610270389, + "grad_norm": 2.0728635787963867, + "learning_rate": 1.685653752351204e-05, + "loss": 3.2169, + "step": 7028 + }, + { + "epoch": 0.5635485358080615, + "grad_norm": 1.9004534482955933, + "learning_rate": 1.685559228996144e-05, + "loss": 3.2134, + "step": 7029 + }, + { + "epoch": 0.5636287105890843, + "grad_norm": 1.878994107246399, + "learning_rate": 1.6854646940827967e-05, + "loss": 3.298, + "step": 7030 + }, + { + "epoch": 0.5637088853701069, + "grad_norm": 1.8505244255065918, + "learning_rate": 1.6853701476127567e-05, + "loss": 3.2443, + "step": 7031 + }, + { + "epoch": 0.5637890601511295, + "grad_norm": 1.8844587802886963, + "learning_rate": 1.6852755895876173e-05, + "loss": 3.2943, + "step": 7032 + }, + { + "epoch": 0.5638692349321521, + "grad_norm": 1.8892521858215332, + "learning_rate": 1.6851810200089732e-05, + "loss": 3.2969, + "step": 7033 + }, + { + "epoch": 0.5639494097131748, + "grad_norm": 1.9480936527252197, + "learning_rate": 1.6850864388784186e-05, + "loss": 3.2505, + "step": 7034 + }, + { + "epoch": 0.5640295844941974, + "grad_norm": 1.8504753112792969, + "learning_rate": 1.6849918461975478e-05, + "loss": 3.3253, + "step": 7035 + }, + { + "epoch": 0.56410975927522, + "grad_norm": 1.7843624353408813, + "learning_rate": 1.684897241967956e-05, + "loss": 3.1502, + "step": 7036 + }, + { + "epoch": 0.5641899340562426, + "grad_norm": 1.8255465030670166, + "learning_rate": 1.6848026261912382e-05, + "loss": 3.1965, + "step": 7037 + }, + { + "epoch": 0.5642701088372652, + "grad_norm": 1.8276728391647339, + "learning_rate": 1.6847079988689893e-05, + "loss": 3.2121, + "step": 7038 + }, + { + "epoch": 0.5643502836182879, + "grad_norm": 1.8913120031356812, + "learning_rate": 1.684613360002805e-05, + "loss": 3.2457, + "step": 7039 + }, + { + "epoch": 0.5644304583993105, + "grad_norm": 1.7961286306381226, + "learning_rate": 1.684518709594281e-05, + "loss": 3.2328, + "step": 7040 + }, + { + "epoch": 0.5645106331803331, + "grad_norm": 1.9466438293457031, + "learning_rate": 1.6844240476450124e-05, + "loss": 3.2746, + "step": 7041 + }, + { + "epoch": 0.5645908079613557, + "grad_norm": 0.9042731523513794, + "learning_rate": 1.6843293741565957e-05, + "loss": 2.5123, + "step": 7042 + }, + { + "epoch": 0.5646709827423784, + "grad_norm": 1.7830085754394531, + "learning_rate": 1.684234689130627e-05, + "loss": 3.2543, + "step": 7043 + }, + { + "epoch": 0.564751157523401, + "grad_norm": 1.9345322847366333, + "learning_rate": 1.6841399925687022e-05, + "loss": 3.4226, + "step": 7044 + }, + { + "epoch": 0.5648313323044236, + "grad_norm": 1.898034691810608, + "learning_rate": 1.6840452844724184e-05, + "loss": 3.14, + "step": 7045 + }, + { + "epoch": 0.5649115070854462, + "grad_norm": 1.730863332748413, + "learning_rate": 1.6839505648433724e-05, + "loss": 3.239, + "step": 7046 + }, + { + "epoch": 0.5649916818664689, + "grad_norm": 1.6768256425857544, + "learning_rate": 1.683855833683161e-05, + "loss": 3.0674, + "step": 7047 + }, + { + "epoch": 0.5650718566474915, + "grad_norm": 1.8400098085403442, + "learning_rate": 1.683761090993381e-05, + "loss": 3.2258, + "step": 7048 + }, + { + "epoch": 0.5651520314285141, + "grad_norm": 1.7858366966247559, + "learning_rate": 1.68366633677563e-05, + "loss": 3.2599, + "step": 7049 + }, + { + "epoch": 0.5652322062095367, + "grad_norm": 1.8217804431915283, + "learning_rate": 1.6835715710315056e-05, + "loss": 3.2607, + "step": 7050 + }, + { + "epoch": 0.5653123809905595, + "grad_norm": 1.7368149757385254, + "learning_rate": 1.6834767937626055e-05, + "loss": 3.2262, + "step": 7051 + }, + { + "epoch": 0.5653925557715821, + "grad_norm": 1.777037262916565, + "learning_rate": 1.6833820049705275e-05, + "loss": 3.2177, + "step": 7052 + }, + { + "epoch": 0.5654727305526047, + "grad_norm": 1.77266263961792, + "learning_rate": 1.68328720465687e-05, + "loss": 3.2327, + "step": 7053 + }, + { + "epoch": 0.5655529053336273, + "grad_norm": 1.8266699314117432, + "learning_rate": 1.6831923928232304e-05, + "loss": 3.3833, + "step": 7054 + }, + { + "epoch": 0.56563308011465, + "grad_norm": 1.9606478214263916, + "learning_rate": 1.6830975694712083e-05, + "loss": 3.225, + "step": 7055 + }, + { + "epoch": 0.5657132548956726, + "grad_norm": 1.7301254272460938, + "learning_rate": 1.6830027346024018e-05, + "loss": 3.2321, + "step": 7056 + }, + { + "epoch": 0.5657934296766952, + "grad_norm": 1.754276990890503, + "learning_rate": 1.6829078882184097e-05, + "loss": 3.2488, + "step": 7057 + }, + { + "epoch": 0.5658736044577178, + "grad_norm": 1.7742202281951904, + "learning_rate": 1.6828130303208316e-05, + "loss": 3.2244, + "step": 7058 + }, + { + "epoch": 0.5659537792387405, + "grad_norm": 2.649834632873535, + "learning_rate": 1.6827181609112663e-05, + "loss": 3.1827, + "step": 7059 + }, + { + "epoch": 0.5660339540197631, + "grad_norm": 1.9012846946716309, + "learning_rate": 1.6826232799913134e-05, + "loss": 3.2521, + "step": 7060 + }, + { + "epoch": 0.5661141288007857, + "grad_norm": 1.7830971479415894, + "learning_rate": 1.682528387562573e-05, + "loss": 3.1377, + "step": 7061 + }, + { + "epoch": 0.5661943035818083, + "grad_norm": 1.9021128416061401, + "learning_rate": 1.6824334836266437e-05, + "loss": 3.1942, + "step": 7062 + }, + { + "epoch": 0.566274478362831, + "grad_norm": 1.892673134803772, + "learning_rate": 1.682338568185127e-05, + "loss": 3.2338, + "step": 7063 + }, + { + "epoch": 0.5663546531438536, + "grad_norm": 0.8876674771308899, + "learning_rate": 1.6822436412396227e-05, + "loss": 2.5128, + "step": 7064 + }, + { + "epoch": 0.5664348279248762, + "grad_norm": 2.024904251098633, + "learning_rate": 1.6821487027917304e-05, + "loss": 3.334, + "step": 7065 + }, + { + "epoch": 0.5665150027058988, + "grad_norm": 1.7725067138671875, + "learning_rate": 1.6820537528430517e-05, + "loss": 3.2719, + "step": 7066 + }, + { + "epoch": 0.5665951774869215, + "grad_norm": 1.9352651834487915, + "learning_rate": 1.6819587913951872e-05, + "loss": 3.1832, + "step": 7067 + }, + { + "epoch": 0.5666753522679441, + "grad_norm": 1.9198306798934937, + "learning_rate": 1.681863818449738e-05, + "loss": 3.2866, + "step": 7068 + }, + { + "epoch": 0.5667555270489667, + "grad_norm": 1.8514231443405151, + "learning_rate": 1.6817688340083048e-05, + "loss": 3.2546, + "step": 7069 + }, + { + "epoch": 0.5668357018299893, + "grad_norm": 1.8653167486190796, + "learning_rate": 1.6816738380724893e-05, + "loss": 3.2577, + "step": 7070 + }, + { + "epoch": 0.5669158766110121, + "grad_norm": 1.8126600980758667, + "learning_rate": 1.6815788306438932e-05, + "loss": 3.2225, + "step": 7071 + }, + { + "epoch": 0.5669960513920347, + "grad_norm": 1.9638526439666748, + "learning_rate": 1.6814838117241182e-05, + "loss": 3.2219, + "step": 7072 + }, + { + "epoch": 0.5670762261730573, + "grad_norm": 1.834121823310852, + "learning_rate": 1.6813887813147665e-05, + "loss": 3.2451, + "step": 7073 + }, + { + "epoch": 0.5671564009540799, + "grad_norm": 1.858994722366333, + "learning_rate": 1.68129373941744e-05, + "loss": 3.2445, + "step": 7074 + }, + { + "epoch": 0.5672365757351026, + "grad_norm": 1.9203466176986694, + "learning_rate": 1.681198686033741e-05, + "loss": 3.1711, + "step": 7075 + }, + { + "epoch": 0.5673167505161252, + "grad_norm": 1.731879472732544, + "learning_rate": 1.6811036211652724e-05, + "loss": 3.2468, + "step": 7076 + }, + { + "epoch": 0.5673969252971478, + "grad_norm": 1.8418618440628052, + "learning_rate": 1.681008544813637e-05, + "loss": 3.2991, + "step": 7077 + }, + { + "epoch": 0.5674771000781704, + "grad_norm": 1.7301576137542725, + "learning_rate": 1.6809134569804374e-05, + "loss": 3.2435, + "step": 7078 + }, + { + "epoch": 0.567557274859193, + "grad_norm": 1.8110466003417969, + "learning_rate": 1.6808183576672766e-05, + "loss": 3.2634, + "step": 7079 + }, + { + "epoch": 0.5676374496402157, + "grad_norm": 1.9407051801681519, + "learning_rate": 1.6807232468757586e-05, + "loss": 3.2663, + "step": 7080 + }, + { + "epoch": 0.5677176244212383, + "grad_norm": 1.817623257637024, + "learning_rate": 1.6806281246074862e-05, + "loss": 3.2701, + "step": 7081 + }, + { + "epoch": 0.5677977992022609, + "grad_norm": 1.9184272289276123, + "learning_rate": 1.6805329908640638e-05, + "loss": 3.1451, + "step": 7082 + }, + { + "epoch": 0.5678779739832835, + "grad_norm": 0.7975907921791077, + "learning_rate": 1.6804378456470952e-05, + "loss": 2.4907, + "step": 7083 + }, + { + "epoch": 0.5679581487643062, + "grad_norm": 1.8021893501281738, + "learning_rate": 1.680342688958184e-05, + "loss": 3.1688, + "step": 7084 + }, + { + "epoch": 0.5680383235453288, + "grad_norm": 1.9307061433792114, + "learning_rate": 1.680247520798935e-05, + "loss": 3.173, + "step": 7085 + }, + { + "epoch": 0.5681184983263514, + "grad_norm": 1.822426199913025, + "learning_rate": 1.680152341170952e-05, + "loss": 3.083, + "step": 7086 + }, + { + "epoch": 0.568198673107374, + "grad_norm": 1.9476429224014282, + "learning_rate": 1.6800571500758408e-05, + "loss": 3.2189, + "step": 7087 + }, + { + "epoch": 0.5682788478883967, + "grad_norm": 1.9260380268096924, + "learning_rate": 1.6799619475152056e-05, + "loss": 3.2095, + "step": 7088 + }, + { + "epoch": 0.5683590226694193, + "grad_norm": 1.783136010169983, + "learning_rate": 1.6798667334906516e-05, + "loss": 3.1931, + "step": 7089 + }, + { + "epoch": 0.5684391974504419, + "grad_norm": 1.9368252754211426, + "learning_rate": 1.6797715080037842e-05, + "loss": 3.2627, + "step": 7090 + }, + { + "epoch": 0.5685193722314646, + "grad_norm": 1.8155200481414795, + "learning_rate": 1.6796762710562083e-05, + "loss": 3.223, + "step": 7091 + }, + { + "epoch": 0.5685995470124873, + "grad_norm": 1.8691413402557373, + "learning_rate": 1.6795810226495306e-05, + "loss": 3.2506, + "step": 7092 + }, + { + "epoch": 0.5686797217935099, + "grad_norm": 1.7739447355270386, + "learning_rate": 1.679485762785356e-05, + "loss": 3.2439, + "step": 7093 + }, + { + "epoch": 0.5687598965745325, + "grad_norm": 1.888231873512268, + "learning_rate": 1.679390491465291e-05, + "loss": 3.2678, + "step": 7094 + }, + { + "epoch": 0.5688400713555551, + "grad_norm": 1.915033221244812, + "learning_rate": 1.679295208690942e-05, + "loss": 3.2024, + "step": 7095 + }, + { + "epoch": 0.5689202461365778, + "grad_norm": 1.8221625089645386, + "learning_rate": 1.679199914463915e-05, + "loss": 3.1997, + "step": 7096 + }, + { + "epoch": 0.5690004209176004, + "grad_norm": 2.017594575881958, + "learning_rate": 1.679104608785817e-05, + "loss": 3.2483, + "step": 7097 + }, + { + "epoch": 0.569080595698623, + "grad_norm": 1.7631467580795288, + "learning_rate": 1.6790092916582543e-05, + "loss": 3.2484, + "step": 7098 + }, + { + "epoch": 0.5691607704796456, + "grad_norm": 1.7455928325653076, + "learning_rate": 1.678913963082834e-05, + "loss": 3.225, + "step": 7099 + }, + { + "epoch": 0.5692409452606683, + "grad_norm": 1.727454662322998, + "learning_rate": 1.6788186230611643e-05, + "loss": 3.142, + "step": 7100 + }, + { + "epoch": 0.5693211200416909, + "grad_norm": 1.81209135055542, + "learning_rate": 1.6787232715948515e-05, + "loss": 3.1581, + "step": 7101 + }, + { + "epoch": 0.5694012948227135, + "grad_norm": 1.8863857984542847, + "learning_rate": 1.6786279086855034e-05, + "loss": 3.2827, + "step": 7102 + }, + { + "epoch": 0.5694814696037361, + "grad_norm": 1.8311420679092407, + "learning_rate": 1.678532534334728e-05, + "loss": 3.2742, + "step": 7103 + }, + { + "epoch": 0.5695616443847588, + "grad_norm": 1.813031792640686, + "learning_rate": 1.6784371485441333e-05, + "loss": 3.2794, + "step": 7104 + }, + { + "epoch": 0.5696418191657814, + "grad_norm": 1.8456788063049316, + "learning_rate": 1.6783417513153274e-05, + "loss": 3.1702, + "step": 7105 + }, + { + "epoch": 0.569721993946804, + "grad_norm": 1.8773738145828247, + "learning_rate": 1.6782463426499184e-05, + "loss": 3.2038, + "step": 7106 + }, + { + "epoch": 0.5698021687278266, + "grad_norm": 1.8990181684494019, + "learning_rate": 1.6781509225495153e-05, + "loss": 3.1283, + "step": 7107 + }, + { + "epoch": 0.5698823435088493, + "grad_norm": 1.8348056077957153, + "learning_rate": 1.678055491015727e-05, + "loss": 3.2283, + "step": 7108 + }, + { + "epoch": 0.5699625182898719, + "grad_norm": 1.819048523902893, + "learning_rate": 1.6779600480501616e-05, + "loss": 3.2449, + "step": 7109 + }, + { + "epoch": 0.5700426930708945, + "grad_norm": 1.7535197734832764, + "learning_rate": 1.6778645936544288e-05, + "loss": 3.1956, + "step": 7110 + }, + { + "epoch": 0.5701228678519171, + "grad_norm": 1.8009318113327026, + "learning_rate": 1.6777691278301378e-05, + "loss": 3.1898, + "step": 7111 + }, + { + "epoch": 0.5702030426329399, + "grad_norm": 1.922308325767517, + "learning_rate": 1.6776736505788984e-05, + "loss": 3.2087, + "step": 7112 + }, + { + "epoch": 0.5702832174139625, + "grad_norm": 1.8575646877288818, + "learning_rate": 1.67757816190232e-05, + "loss": 3.2643, + "step": 7113 + }, + { + "epoch": 0.5703633921949851, + "grad_norm": 1.8434607982635498, + "learning_rate": 1.6774826618020123e-05, + "loss": 3.2034, + "step": 7114 + }, + { + "epoch": 0.5704435669760077, + "grad_norm": 1.8345786333084106, + "learning_rate": 1.6773871502795863e-05, + "loss": 3.1916, + "step": 7115 + }, + { + "epoch": 0.5705237417570304, + "grad_norm": 1.8336119651794434, + "learning_rate": 1.677291627336651e-05, + "loss": 3.1496, + "step": 7116 + }, + { + "epoch": 0.570603916538053, + "grad_norm": 2.005535840988159, + "learning_rate": 1.677196092974818e-05, + "loss": 3.2151, + "step": 7117 + }, + { + "epoch": 0.5706840913190756, + "grad_norm": 1.8940171003341675, + "learning_rate": 1.6771005471956976e-05, + "loss": 3.2026, + "step": 7118 + }, + { + "epoch": 0.5707642661000982, + "grad_norm": 1.7441630363464355, + "learning_rate": 1.6770049900009e-05, + "loss": 3.1765, + "step": 7119 + }, + { + "epoch": 0.5708444408811209, + "grad_norm": 1.8829669952392578, + "learning_rate": 1.6769094213920375e-05, + "loss": 3.2372, + "step": 7120 + }, + { + "epoch": 0.5709246156621435, + "grad_norm": 1.831254243850708, + "learning_rate": 1.6768138413707206e-05, + "loss": 3.3296, + "step": 7121 + }, + { + "epoch": 0.5710047904431661, + "grad_norm": 1.8862415552139282, + "learning_rate": 1.676718249938561e-05, + "loss": 3.2292, + "step": 7122 + }, + { + "epoch": 0.5710849652241887, + "grad_norm": 1.783220648765564, + "learning_rate": 1.6766226470971698e-05, + "loss": 3.248, + "step": 7123 + }, + { + "epoch": 0.5711651400052113, + "grad_norm": 1.8708840608596802, + "learning_rate": 1.6765270328481594e-05, + "loss": 3.207, + "step": 7124 + }, + { + "epoch": 0.571245314786234, + "grad_norm": 1.8354170322418213, + "learning_rate": 1.6764314071931417e-05, + "loss": 3.2106, + "step": 7125 + }, + { + "epoch": 0.5713254895672566, + "grad_norm": 1.9507509469985962, + "learning_rate": 1.6763357701337292e-05, + "loss": 3.2334, + "step": 7126 + }, + { + "epoch": 0.5714056643482792, + "grad_norm": 1.8972970247268677, + "learning_rate": 1.6762401216715335e-05, + "loss": 3.2336, + "step": 7127 + }, + { + "epoch": 0.5714858391293018, + "grad_norm": 1.8606773614883423, + "learning_rate": 1.676144461808168e-05, + "loss": 3.2043, + "step": 7128 + }, + { + "epoch": 0.5715660139103245, + "grad_norm": 1.810133934020996, + "learning_rate": 1.6760487905452448e-05, + "loss": 3.1094, + "step": 7129 + }, + { + "epoch": 0.5716461886913471, + "grad_norm": 1.924286961555481, + "learning_rate": 1.6759531078843774e-05, + "loss": 3.1839, + "step": 7130 + }, + { + "epoch": 0.5717263634723697, + "grad_norm": 1.9521030187606812, + "learning_rate": 1.675857413827179e-05, + "loss": 3.3026, + "step": 7131 + }, + { + "epoch": 0.5718065382533924, + "grad_norm": 1.982653021812439, + "learning_rate": 1.6757617083752626e-05, + "loss": 3.2202, + "step": 7132 + }, + { + "epoch": 0.5718867130344151, + "grad_norm": 1.8895556926727295, + "learning_rate": 1.675665991530242e-05, + "loss": 3.1569, + "step": 7133 + }, + { + "epoch": 0.5719668878154377, + "grad_norm": 1.7687652111053467, + "learning_rate": 1.675570263293731e-05, + "loss": 3.2316, + "step": 7134 + }, + { + "epoch": 0.5720470625964603, + "grad_norm": 1.798351764678955, + "learning_rate": 1.6754745236673434e-05, + "loss": 3.2288, + "step": 7135 + }, + { + "epoch": 0.572127237377483, + "grad_norm": 1.922884225845337, + "learning_rate": 1.6753787726526932e-05, + "loss": 3.261, + "step": 7136 + }, + { + "epoch": 0.5722074121585056, + "grad_norm": 1.752333641052246, + "learning_rate": 1.6752830102513953e-05, + "loss": 3.2846, + "step": 7137 + }, + { + "epoch": 0.5722875869395282, + "grad_norm": 1.866832971572876, + "learning_rate": 1.675187236465064e-05, + "loss": 3.1462, + "step": 7138 + }, + { + "epoch": 0.5723677617205508, + "grad_norm": 1.9169021844863892, + "learning_rate": 1.675091451295313e-05, + "loss": 3.201, + "step": 7139 + }, + { + "epoch": 0.5724479365015734, + "grad_norm": 0.7432359457015991, + "learning_rate": 1.6749956547437587e-05, + "loss": 2.4977, + "step": 7140 + }, + { + "epoch": 0.5725281112825961, + "grad_norm": 1.8871192932128906, + "learning_rate": 1.6748998468120154e-05, + "loss": 3.3176, + "step": 7141 + }, + { + "epoch": 0.5726082860636187, + "grad_norm": 1.9156168699264526, + "learning_rate": 1.6748040275016987e-05, + "loss": 3.246, + "step": 7142 + }, + { + "epoch": 0.5726884608446413, + "grad_norm": 1.9015408754348755, + "learning_rate": 1.6747081968144238e-05, + "loss": 3.2252, + "step": 7143 + }, + { + "epoch": 0.5727686356256639, + "grad_norm": 1.8613231182098389, + "learning_rate": 1.6746123547518064e-05, + "loss": 3.1481, + "step": 7144 + }, + { + "epoch": 0.5728488104066866, + "grad_norm": 1.7897427082061768, + "learning_rate": 1.6745165013154626e-05, + "loss": 3.2826, + "step": 7145 + }, + { + "epoch": 0.5729289851877092, + "grad_norm": 1.8303627967834473, + "learning_rate": 1.674420636507008e-05, + "loss": 3.2916, + "step": 7146 + }, + { + "epoch": 0.5730091599687318, + "grad_norm": 1.8611879348754883, + "learning_rate": 1.6743247603280597e-05, + "loss": 3.3424, + "step": 7147 + }, + { + "epoch": 0.5730893347497544, + "grad_norm": 1.8999415636062622, + "learning_rate": 1.674228872780233e-05, + "loss": 3.2419, + "step": 7148 + }, + { + "epoch": 0.573169509530777, + "grad_norm": 1.8253146409988403, + "learning_rate": 1.6741329738651454e-05, + "loss": 3.2437, + "step": 7149 + }, + { + "epoch": 0.5732496843117997, + "grad_norm": 1.735087513923645, + "learning_rate": 1.6740370635844138e-05, + "loss": 3.2021, + "step": 7150 + }, + { + "epoch": 0.5733298590928223, + "grad_norm": 1.8034276962280273, + "learning_rate": 1.6739411419396544e-05, + "loss": 3.2294, + "step": 7151 + }, + { + "epoch": 0.573410033873845, + "grad_norm": 1.8820997476577759, + "learning_rate": 1.6738452089324847e-05, + "loss": 3.2495, + "step": 7152 + }, + { + "epoch": 0.5734902086548677, + "grad_norm": 1.835387110710144, + "learning_rate": 1.6737492645645228e-05, + "loss": 3.2473, + "step": 7153 + }, + { + "epoch": 0.5735703834358903, + "grad_norm": 0.8236435055732727, + "learning_rate": 1.6736533088373853e-05, + "loss": 2.5165, + "step": 7154 + }, + { + "epoch": 0.5736505582169129, + "grad_norm": 1.852637529373169, + "learning_rate": 1.673557341752691e-05, + "loss": 3.1156, + "step": 7155 + }, + { + "epoch": 0.5737307329979355, + "grad_norm": 1.9461809396743774, + "learning_rate": 1.6734613633120566e-05, + "loss": 3.2208, + "step": 7156 + }, + { + "epoch": 0.5738109077789582, + "grad_norm": 1.9006342887878418, + "learning_rate": 1.6733653735171015e-05, + "loss": 3.2054, + "step": 7157 + }, + { + "epoch": 0.5738910825599808, + "grad_norm": 1.8912451267242432, + "learning_rate": 1.6732693723694434e-05, + "loss": 3.1762, + "step": 7158 + }, + { + "epoch": 0.5739712573410034, + "grad_norm": 1.832892656326294, + "learning_rate": 1.673173359870701e-05, + "loss": 3.2964, + "step": 7159 + }, + { + "epoch": 0.574051432122026, + "grad_norm": 1.8555423021316528, + "learning_rate": 1.6730773360224927e-05, + "loss": 3.2653, + "step": 7160 + }, + { + "epoch": 0.5741316069030487, + "grad_norm": 1.8015309572219849, + "learning_rate": 1.672981300826438e-05, + "loss": 3.3051, + "step": 7161 + }, + { + "epoch": 0.5742117816840713, + "grad_norm": 1.850486159324646, + "learning_rate": 1.6728852542841556e-05, + "loss": 3.2902, + "step": 7162 + }, + { + "epoch": 0.5742919564650939, + "grad_norm": 1.8245949745178223, + "learning_rate": 1.672789196397265e-05, + "loss": 3.3286, + "step": 7163 + }, + { + "epoch": 0.5743721312461165, + "grad_norm": 1.8800184726715088, + "learning_rate": 1.672693127167386e-05, + "loss": 3.2749, + "step": 7164 + }, + { + "epoch": 0.5744523060271391, + "grad_norm": 1.847109079360962, + "learning_rate": 1.6725970465961373e-05, + "loss": 3.1888, + "step": 7165 + }, + { + "epoch": 0.5745324808081618, + "grad_norm": 1.7706117630004883, + "learning_rate": 1.6725009546851397e-05, + "loss": 3.2094, + "step": 7166 + }, + { + "epoch": 0.5746126555891844, + "grad_norm": 1.8128411769866943, + "learning_rate": 1.6724048514360133e-05, + "loss": 3.1819, + "step": 7167 + }, + { + "epoch": 0.574692830370207, + "grad_norm": 1.719184160232544, + "learning_rate": 1.6723087368503778e-05, + "loss": 3.1625, + "step": 7168 + }, + { + "epoch": 0.5747730051512296, + "grad_norm": 1.687174916267395, + "learning_rate": 1.672212610929854e-05, + "loss": 3.163, + "step": 7169 + }, + { + "epoch": 0.5748531799322523, + "grad_norm": 1.9825432300567627, + "learning_rate": 1.6721164736760628e-05, + "loss": 3.2209, + "step": 7170 + }, + { + "epoch": 0.5749333547132749, + "grad_norm": 1.7261747121810913, + "learning_rate": 1.672020325090624e-05, + "loss": 3.1331, + "step": 7171 + }, + { + "epoch": 0.5750135294942975, + "grad_norm": 1.7087430953979492, + "learning_rate": 1.67192416517516e-05, + "loss": 3.2242, + "step": 7172 + }, + { + "epoch": 0.5750937042753202, + "grad_norm": 1.747888207435608, + "learning_rate": 1.6718279939312912e-05, + "loss": 3.1536, + "step": 7173 + }, + { + "epoch": 0.5751738790563429, + "grad_norm": 1.7674312591552734, + "learning_rate": 1.6717318113606393e-05, + "loss": 3.283, + "step": 7174 + }, + { + "epoch": 0.5752540538373655, + "grad_norm": 1.9164717197418213, + "learning_rate": 1.6716356174648257e-05, + "loss": 3.2323, + "step": 7175 + }, + { + "epoch": 0.5753342286183881, + "grad_norm": 1.8708603382110596, + "learning_rate": 1.6715394122454723e-05, + "loss": 3.2568, + "step": 7176 + }, + { + "epoch": 0.5754144033994107, + "grad_norm": 0.7481711506843567, + "learning_rate": 1.671443195704201e-05, + "loss": 2.5006, + "step": 7177 + }, + { + "epoch": 0.5754945781804334, + "grad_norm": 1.905430555343628, + "learning_rate": 1.671346967842634e-05, + "loss": 3.2318, + "step": 7178 + }, + { + "epoch": 0.575574752961456, + "grad_norm": 2.1937873363494873, + "learning_rate": 1.6712507286623944e-05, + "loss": 3.1669, + "step": 7179 + }, + { + "epoch": 0.5756549277424786, + "grad_norm": 0.745698869228363, + "learning_rate": 1.671154478165103e-05, + "loss": 2.5068, + "step": 7180 + }, + { + "epoch": 0.5757351025235012, + "grad_norm": 1.8570743799209595, + "learning_rate": 1.6710582163523844e-05, + "loss": 3.196, + "step": 7181 + }, + { + "epoch": 0.5758152773045239, + "grad_norm": 0.7409843802452087, + "learning_rate": 1.6709619432258608e-05, + "loss": 2.4913, + "step": 7182 + }, + { + "epoch": 0.5758954520855465, + "grad_norm": 1.92213773727417, + "learning_rate": 1.670865658787155e-05, + "loss": 3.2862, + "step": 7183 + }, + { + "epoch": 0.5759756268665691, + "grad_norm": 1.87057363986969, + "learning_rate": 1.6707693630378904e-05, + "loss": 3.3245, + "step": 7184 + }, + { + "epoch": 0.5760558016475917, + "grad_norm": 1.8467116355895996, + "learning_rate": 1.670673055979691e-05, + "loss": 3.3431, + "step": 7185 + }, + { + "epoch": 0.5761359764286144, + "grad_norm": 1.8122566938400269, + "learning_rate": 1.6705767376141807e-05, + "loss": 3.2009, + "step": 7186 + }, + { + "epoch": 0.576216151209637, + "grad_norm": 1.8459880352020264, + "learning_rate": 1.6704804079429824e-05, + "loss": 3.1999, + "step": 7187 + }, + { + "epoch": 0.5762963259906596, + "grad_norm": 1.8274425268173218, + "learning_rate": 1.6703840669677208e-05, + "loss": 3.2245, + "step": 7188 + }, + { + "epoch": 0.5763765007716822, + "grad_norm": 1.9097634553909302, + "learning_rate": 1.6702877146900202e-05, + "loss": 3.3022, + "step": 7189 + }, + { + "epoch": 0.5764566755527049, + "grad_norm": 1.7966991662979126, + "learning_rate": 1.6701913511115048e-05, + "loss": 3.1528, + "step": 7190 + }, + { + "epoch": 0.5765368503337275, + "grad_norm": 1.8005834817886353, + "learning_rate": 1.6700949762337996e-05, + "loss": 3.2018, + "step": 7191 + }, + { + "epoch": 0.5766170251147501, + "grad_norm": 1.9558370113372803, + "learning_rate": 1.6699985900585292e-05, + "loss": 3.2956, + "step": 7192 + }, + { + "epoch": 0.5766971998957728, + "grad_norm": 1.7193182706832886, + "learning_rate": 1.6699021925873186e-05, + "loss": 3.2136, + "step": 7193 + }, + { + "epoch": 0.5767773746767955, + "grad_norm": 1.8985854387283325, + "learning_rate": 1.669805783821793e-05, + "loss": 3.1994, + "step": 7194 + }, + { + "epoch": 0.5768575494578181, + "grad_norm": 1.7173817157745361, + "learning_rate": 1.6697093637635785e-05, + "loss": 3.251, + "step": 7195 + }, + { + "epoch": 0.5769377242388407, + "grad_norm": 1.8814830780029297, + "learning_rate": 1.6696129324143e-05, + "loss": 3.2662, + "step": 7196 + }, + { + "epoch": 0.5770178990198633, + "grad_norm": 1.9101420640945435, + "learning_rate": 1.6695164897755832e-05, + "loss": 3.2363, + "step": 7197 + }, + { + "epoch": 0.577098073800886, + "grad_norm": 1.7885411977767944, + "learning_rate": 1.669420035849054e-05, + "loss": 3.1674, + "step": 7198 + }, + { + "epoch": 0.5771782485819086, + "grad_norm": 1.7191358804702759, + "learning_rate": 1.6693235706363394e-05, + "loss": 3.2179, + "step": 7199 + }, + { + "epoch": 0.5772584233629312, + "grad_norm": 0.7735863327980042, + "learning_rate": 1.6692270941390653e-05, + "loss": 2.5103, + "step": 7200 + }, + { + "epoch": 0.5773385981439538, + "grad_norm": 0.7337850332260132, + "learning_rate": 1.6691306063588583e-05, + "loss": 2.4979, + "step": 7201 + }, + { + "epoch": 0.5774187729249765, + "grad_norm": 0.775115430355072, + "learning_rate": 1.6690341072973454e-05, + "loss": 2.4962, + "step": 7202 + }, + { + "epoch": 0.5774989477059991, + "grad_norm": 1.7876232862472534, + "learning_rate": 1.6689375969561523e-05, + "loss": 3.2303, + "step": 7203 + }, + { + "epoch": 0.5775791224870217, + "grad_norm": 1.7550135850906372, + "learning_rate": 1.668841075336908e-05, + "loss": 3.2936, + "step": 7204 + }, + { + "epoch": 0.5776592972680443, + "grad_norm": 0.7363992929458618, + "learning_rate": 1.6687445424412382e-05, + "loss": 2.4801, + "step": 7205 + }, + { + "epoch": 0.577739472049067, + "grad_norm": 1.839927315711975, + "learning_rate": 1.6686479982707716e-05, + "loss": 3.2092, + "step": 7206 + }, + { + "epoch": 0.5778196468300896, + "grad_norm": 1.9946227073669434, + "learning_rate": 1.668551442827135e-05, + "loss": 3.2323, + "step": 7207 + }, + { + "epoch": 0.5778998216111122, + "grad_norm": 1.899902582168579, + "learning_rate": 1.668454876111957e-05, + "loss": 3.1841, + "step": 7208 + }, + { + "epoch": 0.5779799963921348, + "grad_norm": 1.8488022089004517, + "learning_rate": 1.668358298126865e-05, + "loss": 3.1628, + "step": 7209 + }, + { + "epoch": 0.5780601711731574, + "grad_norm": 1.7939274311065674, + "learning_rate": 1.6682617088734882e-05, + "loss": 3.319, + "step": 7210 + }, + { + "epoch": 0.5781403459541801, + "grad_norm": 1.7867319583892822, + "learning_rate": 1.668165108353454e-05, + "loss": 3.3118, + "step": 7211 + }, + { + "epoch": 0.5782205207352027, + "grad_norm": 2.0034615993499756, + "learning_rate": 1.668068496568392e-05, + "loss": 3.3462, + "step": 7212 + }, + { + "epoch": 0.5783006955162253, + "grad_norm": 1.7796905040740967, + "learning_rate": 1.6679718735199304e-05, + "loss": 3.2085, + "step": 7213 + }, + { + "epoch": 0.578380870297248, + "grad_norm": 0.7494759559631348, + "learning_rate": 1.6678752392096985e-05, + "loss": 2.5114, + "step": 7214 + }, + { + "epoch": 0.5784610450782707, + "grad_norm": 1.8393563032150269, + "learning_rate": 1.667778593639325e-05, + "loss": 3.2274, + "step": 7215 + }, + { + "epoch": 0.5785412198592933, + "grad_norm": 1.7462103366851807, + "learning_rate": 1.6676819368104403e-05, + "loss": 3.2432, + "step": 7216 + }, + { + "epoch": 0.5786213946403159, + "grad_norm": 1.7267694473266602, + "learning_rate": 1.6675852687246734e-05, + "loss": 3.2391, + "step": 7217 + }, + { + "epoch": 0.5787015694213385, + "grad_norm": 0.7466923594474792, + "learning_rate": 1.667488589383654e-05, + "loss": 2.5043, + "step": 7218 + }, + { + "epoch": 0.5787817442023612, + "grad_norm": 1.7959554195404053, + "learning_rate": 1.667391898789012e-05, + "loss": 3.068, + "step": 7219 + }, + { + "epoch": 0.5788619189833838, + "grad_norm": 1.8315014839172363, + "learning_rate": 1.6672951969423784e-05, + "loss": 3.1482, + "step": 7220 + }, + { + "epoch": 0.5789420937644064, + "grad_norm": 1.8771958351135254, + "learning_rate": 1.6671984838453827e-05, + "loss": 3.288, + "step": 7221 + }, + { + "epoch": 0.579022268545429, + "grad_norm": 1.6656277179718018, + "learning_rate": 1.6671017594996556e-05, + "loss": 3.2393, + "step": 7222 + }, + { + "epoch": 0.5791024433264517, + "grad_norm": 1.8248307704925537, + "learning_rate": 1.6670050239068278e-05, + "loss": 3.197, + "step": 7223 + }, + { + "epoch": 0.5791826181074743, + "grad_norm": 1.8725610971450806, + "learning_rate": 1.6669082770685304e-05, + "loss": 3.2429, + "step": 7224 + }, + { + "epoch": 0.5792627928884969, + "grad_norm": 1.843722939491272, + "learning_rate": 1.6668115189863948e-05, + "loss": 3.2543, + "step": 7225 + }, + { + "epoch": 0.5793429676695195, + "grad_norm": 1.8760312795639038, + "learning_rate": 1.6667147496620516e-05, + "loss": 3.1579, + "step": 7226 + }, + { + "epoch": 0.5794231424505422, + "grad_norm": 1.9519786834716797, + "learning_rate": 1.666617969097133e-05, + "loss": 3.2631, + "step": 7227 + }, + { + "epoch": 0.5795033172315648, + "grad_norm": 0.7865971326828003, + "learning_rate": 1.6665211772932704e-05, + "loss": 2.5089, + "step": 7228 + }, + { + "epoch": 0.5795834920125874, + "grad_norm": 1.7947853803634644, + "learning_rate": 1.6664243742520953e-05, + "loss": 3.2333, + "step": 7229 + }, + { + "epoch": 0.57966366679361, + "grad_norm": 1.8576958179473877, + "learning_rate": 1.6663275599752404e-05, + "loss": 3.1286, + "step": 7230 + }, + { + "epoch": 0.5797438415746327, + "grad_norm": 0.7519245743751526, + "learning_rate": 1.666230734464338e-05, + "loss": 2.5037, + "step": 7231 + }, + { + "epoch": 0.5798240163556553, + "grad_norm": 1.7730722427368164, + "learning_rate": 1.6661338977210198e-05, + "loss": 3.2197, + "step": 7232 + }, + { + "epoch": 0.5799041911366779, + "grad_norm": 1.7700914144515991, + "learning_rate": 1.6660370497469187e-05, + "loss": 3.1368, + "step": 7233 + }, + { + "epoch": 0.5799843659177006, + "grad_norm": 1.7869311571121216, + "learning_rate": 1.665940190543668e-05, + "loss": 3.2533, + "step": 7234 + }, + { + "epoch": 0.5800645406987233, + "grad_norm": 1.8175866603851318, + "learning_rate": 1.6658433201129003e-05, + "loss": 3.1857, + "step": 7235 + }, + { + "epoch": 0.5801447154797459, + "grad_norm": 1.8405985832214355, + "learning_rate": 1.6657464384562488e-05, + "loss": 3.1344, + "step": 7236 + }, + { + "epoch": 0.5802248902607685, + "grad_norm": 1.7316304445266724, + "learning_rate": 1.665649545575347e-05, + "loss": 3.1949, + "step": 7237 + }, + { + "epoch": 0.5803050650417911, + "grad_norm": 1.8226735591888428, + "learning_rate": 1.6655526414718286e-05, + "loss": 3.1891, + "step": 7238 + }, + { + "epoch": 0.5803852398228138, + "grad_norm": 1.7749849557876587, + "learning_rate": 1.665455726147327e-05, + "loss": 3.3289, + "step": 7239 + }, + { + "epoch": 0.5804654146038364, + "grad_norm": 1.9211407899856567, + "learning_rate": 1.6653587996034764e-05, + "loss": 3.2368, + "step": 7240 + }, + { + "epoch": 0.580545589384859, + "grad_norm": 2.0028719902038574, + "learning_rate": 1.6652618618419114e-05, + "loss": 3.2009, + "step": 7241 + }, + { + "epoch": 0.5806257641658816, + "grad_norm": 1.783750295639038, + "learning_rate": 1.6651649128642653e-05, + "loss": 3.1417, + "step": 7242 + }, + { + "epoch": 0.5807059389469043, + "grad_norm": 1.8349543809890747, + "learning_rate": 1.6650679526721735e-05, + "loss": 3.2618, + "step": 7243 + }, + { + "epoch": 0.5807861137279269, + "grad_norm": 1.8666003942489624, + "learning_rate": 1.6649709812672702e-05, + "loss": 3.2098, + "step": 7244 + }, + { + "epoch": 0.5808662885089495, + "grad_norm": 1.8345575332641602, + "learning_rate": 1.664873998651191e-05, + "loss": 3.2753, + "step": 7245 + }, + { + "epoch": 0.5809464632899721, + "grad_norm": 1.7776999473571777, + "learning_rate": 1.66477700482557e-05, + "loss": 3.2101, + "step": 7246 + }, + { + "epoch": 0.5810266380709947, + "grad_norm": 1.782478928565979, + "learning_rate": 1.664679999792043e-05, + "loss": 3.2452, + "step": 7247 + }, + { + "epoch": 0.5811068128520174, + "grad_norm": 1.6994842290878296, + "learning_rate": 1.664582983552246e-05, + "loss": 3.1058, + "step": 7248 + }, + { + "epoch": 0.58118698763304, + "grad_norm": 1.7890009880065918, + "learning_rate": 1.6644859561078136e-05, + "loss": 3.1941, + "step": 7249 + }, + { + "epoch": 0.5812671624140626, + "grad_norm": 1.8650161027908325, + "learning_rate": 1.6643889174603823e-05, + "loss": 3.1859, + "step": 7250 + }, + { + "epoch": 0.5813473371950852, + "grad_norm": 1.777890920639038, + "learning_rate": 1.6642918676115877e-05, + "loss": 3.3345, + "step": 7251 + }, + { + "epoch": 0.5814275119761079, + "grad_norm": 1.7580747604370117, + "learning_rate": 1.6641948065630666e-05, + "loss": 3.2273, + "step": 7252 + }, + { + "epoch": 0.5815076867571305, + "grad_norm": 1.7851930856704712, + "learning_rate": 1.6640977343164552e-05, + "loss": 3.2939, + "step": 7253 + }, + { + "epoch": 0.5815878615381532, + "grad_norm": 1.8776582479476929, + "learning_rate": 1.66400065087339e-05, + "loss": 3.321, + "step": 7254 + }, + { + "epoch": 0.5816680363191759, + "grad_norm": 1.811182975769043, + "learning_rate": 1.663903556235508e-05, + "loss": 3.2102, + "step": 7255 + }, + { + "epoch": 0.5817482111001985, + "grad_norm": 1.902241587638855, + "learning_rate": 1.663806450404446e-05, + "loss": 3.2402, + "step": 7256 + }, + { + "epoch": 0.5818283858812211, + "grad_norm": 1.8140531778335571, + "learning_rate": 1.663709333381841e-05, + "loss": 3.2159, + "step": 7257 + }, + { + "epoch": 0.5819085606622437, + "grad_norm": 1.7942384481430054, + "learning_rate": 1.6636122051693308e-05, + "loss": 3.1137, + "step": 7258 + }, + { + "epoch": 0.5819887354432663, + "grad_norm": 1.782517671585083, + "learning_rate": 1.6635150657685525e-05, + "loss": 3.2309, + "step": 7259 + }, + { + "epoch": 0.582068910224289, + "grad_norm": 1.7589819431304932, + "learning_rate": 1.6634179151811438e-05, + "loss": 3.2844, + "step": 7260 + }, + { + "epoch": 0.5821490850053116, + "grad_norm": 0.8405020833015442, + "learning_rate": 1.663320753408743e-05, + "loss": 2.5095, + "step": 7261 + }, + { + "epoch": 0.5822292597863342, + "grad_norm": 1.8014570474624634, + "learning_rate": 1.6632235804529883e-05, + "loss": 3.1924, + "step": 7262 + }, + { + "epoch": 0.5823094345673568, + "grad_norm": 1.87270987033844, + "learning_rate": 1.6631263963155177e-05, + "loss": 3.2732, + "step": 7263 + }, + { + "epoch": 0.5823896093483795, + "grad_norm": 1.9873865842819214, + "learning_rate": 1.66302920099797e-05, + "loss": 3.2262, + "step": 7264 + }, + { + "epoch": 0.5824697841294021, + "grad_norm": 1.894483208656311, + "learning_rate": 1.662931994501983e-05, + "loss": 3.1965, + "step": 7265 + }, + { + "epoch": 0.5825499589104247, + "grad_norm": 1.920066237449646, + "learning_rate": 1.6628347768291968e-05, + "loss": 3.1157, + "step": 7266 + }, + { + "epoch": 0.5826301336914473, + "grad_norm": 1.855563759803772, + "learning_rate": 1.6627375479812497e-05, + "loss": 3.0955, + "step": 7267 + }, + { + "epoch": 0.58271030847247, + "grad_norm": 1.7717537879943848, + "learning_rate": 1.662640307959781e-05, + "loss": 3.314, + "step": 7268 + }, + { + "epoch": 0.5827904832534926, + "grad_norm": 1.825767993927002, + "learning_rate": 1.6625430567664304e-05, + "loss": 3.2085, + "step": 7269 + }, + { + "epoch": 0.5828706580345152, + "grad_norm": 1.7614929676055908, + "learning_rate": 1.6624457944028373e-05, + "loss": 3.2856, + "step": 7270 + }, + { + "epoch": 0.5829508328155378, + "grad_norm": 1.843618392944336, + "learning_rate": 1.6623485208706414e-05, + "loss": 3.3431, + "step": 7271 + }, + { + "epoch": 0.5830310075965605, + "grad_norm": 1.8245885372161865, + "learning_rate": 1.6622512361714834e-05, + "loss": 3.2443, + "step": 7272 + }, + { + "epoch": 0.5831111823775831, + "grad_norm": 1.8003671169281006, + "learning_rate": 1.6621539403070027e-05, + "loss": 3.2911, + "step": 7273 + }, + { + "epoch": 0.5831913571586057, + "grad_norm": 1.8063162565231323, + "learning_rate": 1.6620566332788395e-05, + "loss": 3.1455, + "step": 7274 + }, + { + "epoch": 0.5832715319396284, + "grad_norm": 2.013627767562866, + "learning_rate": 1.6619593150886353e-05, + "loss": 3.2645, + "step": 7275 + }, + { + "epoch": 0.5833517067206511, + "grad_norm": 1.9604840278625488, + "learning_rate": 1.66186198573803e-05, + "loss": 3.2538, + "step": 7276 + }, + { + "epoch": 0.5834318815016737, + "grad_norm": 1.789867639541626, + "learning_rate": 1.661764645228665e-05, + "loss": 3.2651, + "step": 7277 + }, + { + "epoch": 0.5835120562826963, + "grad_norm": 1.7785584926605225, + "learning_rate": 1.6616672935621816e-05, + "loss": 3.2111, + "step": 7278 + }, + { + "epoch": 0.5835922310637189, + "grad_norm": 1.750127911567688, + "learning_rate": 1.6615699307402207e-05, + "loss": 3.1492, + "step": 7279 + }, + { + "epoch": 0.5836724058447416, + "grad_norm": 1.8736799955368042, + "learning_rate": 1.661472556764424e-05, + "loss": 3.2226, + "step": 7280 + }, + { + "epoch": 0.5837525806257642, + "grad_norm": 1.8471291065216064, + "learning_rate": 1.6613751716364328e-05, + "loss": 3.1704, + "step": 7281 + }, + { + "epoch": 0.5838327554067868, + "grad_norm": 1.8922970294952393, + "learning_rate": 1.6612777753578892e-05, + "loss": 3.2579, + "step": 7282 + }, + { + "epoch": 0.5839129301878094, + "grad_norm": 2.0346217155456543, + "learning_rate": 1.6611803679304358e-05, + "loss": 3.4028, + "step": 7283 + }, + { + "epoch": 0.583993104968832, + "grad_norm": 1.824500560760498, + "learning_rate": 1.661082949355714e-05, + "loss": 3.1774, + "step": 7284 + }, + { + "epoch": 0.5840732797498547, + "grad_norm": 1.9179357290267944, + "learning_rate": 1.660985519635367e-05, + "loss": 3.1495, + "step": 7285 + }, + { + "epoch": 0.5841534545308773, + "grad_norm": 1.72761869430542, + "learning_rate": 1.660888078771037e-05, + "loss": 3.2777, + "step": 7286 + }, + { + "epoch": 0.5842336293118999, + "grad_norm": 1.9698474407196045, + "learning_rate": 1.660790626764367e-05, + "loss": 3.2408, + "step": 7287 + }, + { + "epoch": 0.5843138040929226, + "grad_norm": 1.8718162775039673, + "learning_rate": 1.6606931636169997e-05, + "loss": 3.2212, + "step": 7288 + }, + { + "epoch": 0.5843939788739452, + "grad_norm": 1.8856139183044434, + "learning_rate": 1.6605956893305784e-05, + "loss": 3.2653, + "step": 7289 + }, + { + "epoch": 0.5844741536549678, + "grad_norm": 1.7889761924743652, + "learning_rate": 1.660498203906747e-05, + "loss": 3.189, + "step": 7290 + }, + { + "epoch": 0.5845543284359904, + "grad_norm": 1.7685949802398682, + "learning_rate": 1.6604007073471487e-05, + "loss": 3.2053, + "step": 7291 + }, + { + "epoch": 0.584634503217013, + "grad_norm": 1.9610722064971924, + "learning_rate": 1.6603031996534267e-05, + "loss": 3.1873, + "step": 7292 + }, + { + "epoch": 0.5847146779980357, + "grad_norm": 1.8239977359771729, + "learning_rate": 1.6602056808272257e-05, + "loss": 3.2355, + "step": 7293 + }, + { + "epoch": 0.5847948527790583, + "grad_norm": 1.7867672443389893, + "learning_rate": 1.66010815087019e-05, + "loss": 3.2977, + "step": 7294 + }, + { + "epoch": 0.584875027560081, + "grad_norm": 1.7146821022033691, + "learning_rate": 1.660010609783963e-05, + "loss": 3.2222, + "step": 7295 + }, + { + "epoch": 0.5849552023411037, + "grad_norm": 1.9508579969406128, + "learning_rate": 1.65991305757019e-05, + "loss": 3.2875, + "step": 7296 + }, + { + "epoch": 0.5850353771221263, + "grad_norm": 1.7701284885406494, + "learning_rate": 1.6598154942305154e-05, + "loss": 3.2892, + "step": 7297 + }, + { + "epoch": 0.5851155519031489, + "grad_norm": 1.8411985635757446, + "learning_rate": 1.659717919766584e-05, + "loss": 3.1849, + "step": 7298 + }, + { + "epoch": 0.5851957266841715, + "grad_norm": 1.8269867897033691, + "learning_rate": 1.659620334180041e-05, + "loss": 3.1552, + "step": 7299 + }, + { + "epoch": 0.5852759014651941, + "grad_norm": 1.7674378156661987, + "learning_rate": 1.6595227374725316e-05, + "loss": 3.1694, + "step": 7300 + }, + { + "epoch": 0.5853560762462168, + "grad_norm": 1.7754881381988525, + "learning_rate": 1.6594251296457012e-05, + "loss": 3.3379, + "step": 7301 + }, + { + "epoch": 0.5854362510272394, + "grad_norm": 1.79558265209198, + "learning_rate": 1.6593275107011958e-05, + "loss": 3.1632, + "step": 7302 + }, + { + "epoch": 0.585516425808262, + "grad_norm": 1.8949440717697144, + "learning_rate": 1.659229880640661e-05, + "loss": 3.3206, + "step": 7303 + }, + { + "epoch": 0.5855966005892846, + "grad_norm": 1.7729274034500122, + "learning_rate": 1.6591322394657426e-05, + "loss": 3.0711, + "step": 7304 + }, + { + "epoch": 0.5856767753703073, + "grad_norm": 1.7540035247802734, + "learning_rate": 1.6590345871780867e-05, + "loss": 3.1375, + "step": 7305 + }, + { + "epoch": 0.5857569501513299, + "grad_norm": 1.9131513833999634, + "learning_rate": 1.65893692377934e-05, + "loss": 3.2332, + "step": 7306 + }, + { + "epoch": 0.5858371249323525, + "grad_norm": 1.828710913658142, + "learning_rate": 1.6588392492711493e-05, + "loss": 3.2507, + "step": 7307 + }, + { + "epoch": 0.5859172997133751, + "grad_norm": 1.8423545360565186, + "learning_rate": 1.658741563655161e-05, + "loss": 3.2365, + "step": 7308 + }, + { + "epoch": 0.5859974744943978, + "grad_norm": 1.79787015914917, + "learning_rate": 1.6586438669330217e-05, + "loss": 3.2196, + "step": 7309 + }, + { + "epoch": 0.5860776492754204, + "grad_norm": 1.7527847290039062, + "learning_rate": 1.6585461591063794e-05, + "loss": 3.2587, + "step": 7310 + }, + { + "epoch": 0.586157824056443, + "grad_norm": 1.9085123538970947, + "learning_rate": 1.6584484401768808e-05, + "loss": 3.1863, + "step": 7311 + }, + { + "epoch": 0.5862379988374656, + "grad_norm": 1.8156101703643799, + "learning_rate": 1.6583507101461734e-05, + "loss": 3.1875, + "step": 7312 + }, + { + "epoch": 0.5863181736184883, + "grad_norm": 1.9019029140472412, + "learning_rate": 1.658252969015905e-05, + "loss": 3.2241, + "step": 7313 + }, + { + "epoch": 0.5863983483995109, + "grad_norm": 2.098085641860962, + "learning_rate": 1.6581552167877237e-05, + "loss": 3.2528, + "step": 7314 + }, + { + "epoch": 0.5864785231805336, + "grad_norm": 1.8422596454620361, + "learning_rate": 1.6580574534632777e-05, + "loss": 3.2795, + "step": 7315 + }, + { + "epoch": 0.5865586979615562, + "grad_norm": 1.8984174728393555, + "learning_rate": 1.6579596790442146e-05, + "loss": 3.218, + "step": 7316 + }, + { + "epoch": 0.5866388727425789, + "grad_norm": 1.8956506252288818, + "learning_rate": 1.657861893532183e-05, + "loss": 3.234, + "step": 7317 + }, + { + "epoch": 0.5867190475236015, + "grad_norm": 1.7803430557250977, + "learning_rate": 1.657764096928832e-05, + "loss": 3.265, + "step": 7318 + }, + { + "epoch": 0.5867992223046241, + "grad_norm": 1.9297832250595093, + "learning_rate": 1.6576662892358103e-05, + "loss": 3.2581, + "step": 7319 + }, + { + "epoch": 0.5868793970856467, + "grad_norm": 1.88937509059906, + "learning_rate": 1.6575684704547668e-05, + "loss": 3.2632, + "step": 7320 + }, + { + "epoch": 0.5869595718666694, + "grad_norm": 1.8274571895599365, + "learning_rate": 1.6574706405873505e-05, + "loss": 3.2213, + "step": 7321 + }, + { + "epoch": 0.587039746647692, + "grad_norm": 1.9798822402954102, + "learning_rate": 1.6573727996352107e-05, + "loss": 3.2655, + "step": 7322 + }, + { + "epoch": 0.5871199214287146, + "grad_norm": 1.8553858995437622, + "learning_rate": 1.6572749475999972e-05, + "loss": 3.2328, + "step": 7323 + }, + { + "epoch": 0.5872000962097372, + "grad_norm": 1.7203600406646729, + "learning_rate": 1.65717708448336e-05, + "loss": 3.1918, + "step": 7324 + }, + { + "epoch": 0.5872802709907599, + "grad_norm": 1.8068211078643799, + "learning_rate": 1.6570792102869484e-05, + "loss": 3.2029, + "step": 7325 + }, + { + "epoch": 0.5873604457717825, + "grad_norm": 1.8001165390014648, + "learning_rate": 1.656981325012413e-05, + "loss": 3.248, + "step": 7326 + }, + { + "epoch": 0.5874406205528051, + "grad_norm": 1.8673189878463745, + "learning_rate": 1.6568834286614044e-05, + "loss": 3.212, + "step": 7327 + }, + { + "epoch": 0.5875207953338277, + "grad_norm": 1.9309699535369873, + "learning_rate": 1.6567855212355725e-05, + "loss": 3.3313, + "step": 7328 + }, + { + "epoch": 0.5876009701148504, + "grad_norm": 1.8846638202667236, + "learning_rate": 1.656687602736568e-05, + "loss": 3.275, + "step": 7329 + }, + { + "epoch": 0.587681144895873, + "grad_norm": 1.9558804035186768, + "learning_rate": 1.656589673166042e-05, + "loss": 3.2927, + "step": 7330 + }, + { + "epoch": 0.5877613196768956, + "grad_norm": 1.877423644065857, + "learning_rate": 1.6564917325256455e-05, + "loss": 3.1965, + "step": 7331 + }, + { + "epoch": 0.5878414944579182, + "grad_norm": 1.886664628982544, + "learning_rate": 1.65639378081703e-05, + "loss": 3.2322, + "step": 7332 + }, + { + "epoch": 0.5879216692389408, + "grad_norm": 1.9625649452209473, + "learning_rate": 1.6562958180418464e-05, + "loss": 3.2348, + "step": 7333 + }, + { + "epoch": 0.5880018440199635, + "grad_norm": 1.8106400966644287, + "learning_rate": 1.6561978442017467e-05, + "loss": 3.188, + "step": 7334 + }, + { + "epoch": 0.5880820188009861, + "grad_norm": 1.7969485521316528, + "learning_rate": 1.6560998592983827e-05, + "loss": 3.1488, + "step": 7335 + }, + { + "epoch": 0.5881621935820088, + "grad_norm": 1.9813979864120483, + "learning_rate": 1.656001863333406e-05, + "loss": 3.2242, + "step": 7336 + }, + { + "epoch": 0.5882423683630315, + "grad_norm": 1.8764700889587402, + "learning_rate": 1.6559038563084694e-05, + "loss": 3.1886, + "step": 7337 + }, + { + "epoch": 0.5883225431440541, + "grad_norm": 1.8334757089614868, + "learning_rate": 1.655805838225225e-05, + "loss": 3.1752, + "step": 7338 + }, + { + "epoch": 0.5884027179250767, + "grad_norm": 1.9417552947998047, + "learning_rate": 1.6557078090853253e-05, + "loss": 3.3381, + "step": 7339 + }, + { + "epoch": 0.5884828927060993, + "grad_norm": 1.7721130847930908, + "learning_rate": 1.6556097688904226e-05, + "loss": 3.2537, + "step": 7340 + }, + { + "epoch": 0.588563067487122, + "grad_norm": 2.1959924697875977, + "learning_rate": 1.6555117176421706e-05, + "loss": 3.2415, + "step": 7341 + }, + { + "epoch": 0.5886432422681446, + "grad_norm": 1.7432760000228882, + "learning_rate": 1.6554136553422223e-05, + "loss": 3.1537, + "step": 7342 + }, + { + "epoch": 0.5887234170491672, + "grad_norm": 1.8004485368728638, + "learning_rate": 1.6553155819922305e-05, + "loss": 3.2006, + "step": 7343 + }, + { + "epoch": 0.5888035918301898, + "grad_norm": 1.90912926197052, + "learning_rate": 1.655217497593849e-05, + "loss": 3.2604, + "step": 7344 + }, + { + "epoch": 0.5888837666112124, + "grad_norm": 1.946137547492981, + "learning_rate": 1.655119402148731e-05, + "loss": 3.1942, + "step": 7345 + }, + { + "epoch": 0.5889639413922351, + "grad_norm": 1.8458590507507324, + "learning_rate": 1.655021295658531e-05, + "loss": 3.1232, + "step": 7346 + }, + { + "epoch": 0.5890441161732577, + "grad_norm": 1.7648138999938965, + "learning_rate": 1.654923178124903e-05, + "loss": 3.1995, + "step": 7347 + }, + { + "epoch": 0.5891242909542803, + "grad_norm": 1.9223707914352417, + "learning_rate": 1.6548250495495008e-05, + "loss": 3.1976, + "step": 7348 + }, + { + "epoch": 0.5892044657353029, + "grad_norm": 1.7513668537139893, + "learning_rate": 1.6547269099339796e-05, + "loss": 3.1959, + "step": 7349 + }, + { + "epoch": 0.5892846405163256, + "grad_norm": 1.9866482019424438, + "learning_rate": 1.654628759279993e-05, + "loss": 3.2162, + "step": 7350 + }, + { + "epoch": 0.5893648152973482, + "grad_norm": 1.7614229917526245, + "learning_rate": 1.654530597589196e-05, + "loss": 3.1641, + "step": 7351 + }, + { + "epoch": 0.5894449900783708, + "grad_norm": 1.7661364078521729, + "learning_rate": 1.6544324248632443e-05, + "loss": 3.1705, + "step": 7352 + }, + { + "epoch": 0.5895251648593934, + "grad_norm": 1.9757822751998901, + "learning_rate": 1.6543342411037923e-05, + "loss": 3.1392, + "step": 7353 + }, + { + "epoch": 0.5896053396404161, + "grad_norm": 1.7675868272781372, + "learning_rate": 1.6542360463124957e-05, + "loss": 3.1958, + "step": 7354 + }, + { + "epoch": 0.5896855144214387, + "grad_norm": 1.8800305128097534, + "learning_rate": 1.6541378404910098e-05, + "loss": 3.2537, + "step": 7355 + }, + { + "epoch": 0.5897656892024614, + "grad_norm": 1.9757980108261108, + "learning_rate": 1.6540396236409905e-05, + "loss": 3.3378, + "step": 7356 + }, + { + "epoch": 0.589845863983484, + "grad_norm": 1.875893473625183, + "learning_rate": 1.6539413957640937e-05, + "loss": 3.2242, + "step": 7357 + }, + { + "epoch": 0.5899260387645067, + "grad_norm": 2.1113929748535156, + "learning_rate": 1.6538431568619752e-05, + "loss": 3.2134, + "step": 7358 + }, + { + "epoch": 0.5900062135455293, + "grad_norm": 2.0404117107391357, + "learning_rate": 1.6537449069362917e-05, + "loss": 3.2784, + "step": 7359 + }, + { + "epoch": 0.5900863883265519, + "grad_norm": 1.935545802116394, + "learning_rate": 1.6536466459886998e-05, + "loss": 3.2806, + "step": 7360 + }, + { + "epoch": 0.5901665631075745, + "grad_norm": 1.9453433752059937, + "learning_rate": 1.6535483740208555e-05, + "loss": 3.2023, + "step": 7361 + }, + { + "epoch": 0.5902467378885972, + "grad_norm": 1.746748447418213, + "learning_rate": 1.6534500910344156e-05, + "loss": 3.1197, + "step": 7362 + }, + { + "epoch": 0.5903269126696198, + "grad_norm": 1.8262805938720703, + "learning_rate": 1.653351797031038e-05, + "loss": 3.266, + "step": 7363 + }, + { + "epoch": 0.5904070874506424, + "grad_norm": 0.7910563349723816, + "learning_rate": 1.6532534920123794e-05, + "loss": 2.489, + "step": 7364 + }, + { + "epoch": 0.590487262231665, + "grad_norm": 1.7602227926254272, + "learning_rate": 1.6531551759800966e-05, + "loss": 3.1794, + "step": 7365 + }, + { + "epoch": 0.5905674370126877, + "grad_norm": 1.8220967054367065, + "learning_rate": 1.6530568489358482e-05, + "loss": 3.1374, + "step": 7366 + }, + { + "epoch": 0.5906476117937103, + "grad_norm": 1.9011262655258179, + "learning_rate": 1.6529585108812913e-05, + "loss": 3.2494, + "step": 7367 + }, + { + "epoch": 0.5907277865747329, + "grad_norm": 1.9378535747528076, + "learning_rate": 1.6528601618180843e-05, + "loss": 3.2107, + "step": 7368 + }, + { + "epoch": 0.5908079613557555, + "grad_norm": 2.025658369064331, + "learning_rate": 1.6527618017478847e-05, + "loss": 3.3003, + "step": 7369 + }, + { + "epoch": 0.5908881361367782, + "grad_norm": 1.9887356758117676, + "learning_rate": 1.6526634306723514e-05, + "loss": 3.2264, + "step": 7370 + }, + { + "epoch": 0.5909683109178008, + "grad_norm": 1.7834473848342896, + "learning_rate": 1.6525650485931424e-05, + "loss": 3.3837, + "step": 7371 + }, + { + "epoch": 0.5910484856988234, + "grad_norm": 2.0278689861297607, + "learning_rate": 1.6524666555119168e-05, + "loss": 3.2168, + "step": 7372 + }, + { + "epoch": 0.591128660479846, + "grad_norm": 1.8660074472427368, + "learning_rate": 1.6523682514303335e-05, + "loss": 3.2988, + "step": 7373 + }, + { + "epoch": 0.5912088352608686, + "grad_norm": 1.7694718837738037, + "learning_rate": 1.6522698363500515e-05, + "loss": 3.2726, + "step": 7374 + }, + { + "epoch": 0.5912890100418913, + "grad_norm": 1.9289965629577637, + "learning_rate": 1.65217141027273e-05, + "loss": 3.3691, + "step": 7375 + }, + { + "epoch": 0.591369184822914, + "grad_norm": 1.9042694568634033, + "learning_rate": 1.6520729732000278e-05, + "loss": 3.3163, + "step": 7376 + }, + { + "epoch": 0.5914493596039366, + "grad_norm": 1.8092259168624878, + "learning_rate": 1.6519745251336055e-05, + "loss": 3.2162, + "step": 7377 + }, + { + "epoch": 0.5915295343849593, + "grad_norm": 1.8855562210083008, + "learning_rate": 1.6518760660751224e-05, + "loss": 3.2583, + "step": 7378 + }, + { + "epoch": 0.5916097091659819, + "grad_norm": 1.8558878898620605, + "learning_rate": 1.6517775960262385e-05, + "loss": 3.2367, + "step": 7379 + }, + { + "epoch": 0.5916898839470045, + "grad_norm": 2.0486319065093994, + "learning_rate": 1.651679114988614e-05, + "loss": 3.141, + "step": 7380 + }, + { + "epoch": 0.5917700587280271, + "grad_norm": 1.9404608011245728, + "learning_rate": 1.6515806229639094e-05, + "loss": 3.3308, + "step": 7381 + }, + { + "epoch": 0.5918502335090498, + "grad_norm": 1.8565386533737183, + "learning_rate": 1.6514821199537853e-05, + "loss": 3.1561, + "step": 7382 + }, + { + "epoch": 0.5919304082900724, + "grad_norm": 1.8029009103775024, + "learning_rate": 1.651383605959902e-05, + "loss": 3.2813, + "step": 7383 + }, + { + "epoch": 0.592010583071095, + "grad_norm": 1.8335599899291992, + "learning_rate": 1.6512850809839208e-05, + "loss": 3.2283, + "step": 7384 + }, + { + "epoch": 0.5920907578521176, + "grad_norm": 1.8667794466018677, + "learning_rate": 1.6511865450275025e-05, + "loss": 3.3392, + "step": 7385 + }, + { + "epoch": 0.5921709326331402, + "grad_norm": 1.9280171394348145, + "learning_rate": 1.651087998092309e-05, + "loss": 3.3072, + "step": 7386 + }, + { + "epoch": 0.5922511074141629, + "grad_norm": 1.8518272638320923, + "learning_rate": 1.650989440180001e-05, + "loss": 3.2543, + "step": 7387 + }, + { + "epoch": 0.5923312821951855, + "grad_norm": 1.9485769271850586, + "learning_rate": 1.6508908712922403e-05, + "loss": 3.1622, + "step": 7388 + }, + { + "epoch": 0.5924114569762081, + "grad_norm": 1.876475214958191, + "learning_rate": 1.650792291430689e-05, + "loss": 3.2042, + "step": 7389 + }, + { + "epoch": 0.5924916317572307, + "grad_norm": 1.8402376174926758, + "learning_rate": 1.6506937005970096e-05, + "loss": 3.2965, + "step": 7390 + }, + { + "epoch": 0.5925718065382534, + "grad_norm": 1.8794456720352173, + "learning_rate": 1.650595098792863e-05, + "loss": 3.2626, + "step": 7391 + }, + { + "epoch": 0.592651981319276, + "grad_norm": 1.8903751373291016, + "learning_rate": 1.6504964860199123e-05, + "loss": 3.1309, + "step": 7392 + }, + { + "epoch": 0.5927321561002986, + "grad_norm": 1.736854076385498, + "learning_rate": 1.6503978622798206e-05, + "loss": 3.2905, + "step": 7393 + }, + { + "epoch": 0.5928123308813212, + "grad_norm": 1.9253185987472534, + "learning_rate": 1.65029922757425e-05, + "loss": 3.3258, + "step": 7394 + }, + { + "epoch": 0.5928925056623439, + "grad_norm": 1.9499070644378662, + "learning_rate": 1.6502005819048635e-05, + "loss": 3.2791, + "step": 7395 + }, + { + "epoch": 0.5929726804433665, + "grad_norm": 1.9894908666610718, + "learning_rate": 1.650101925273324e-05, + "loss": 3.2053, + "step": 7396 + }, + { + "epoch": 0.5930528552243892, + "grad_norm": 1.7569383382797241, + "learning_rate": 1.650003257681296e-05, + "loss": 3.2593, + "step": 7397 + }, + { + "epoch": 0.5931330300054118, + "grad_norm": 1.7314387559890747, + "learning_rate": 1.6499045791304416e-05, + "loss": 3.3743, + "step": 7398 + }, + { + "epoch": 0.5932132047864345, + "grad_norm": 1.9717143774032593, + "learning_rate": 1.6498058896224252e-05, + "loss": 3.2811, + "step": 7399 + }, + { + "epoch": 0.5932933795674571, + "grad_norm": 1.814996361732483, + "learning_rate": 1.6497071891589102e-05, + "loss": 3.2684, + "step": 7400 + }, + { + "epoch": 0.5933735543484797, + "grad_norm": 2.2219269275665283, + "learning_rate": 1.6496084777415613e-05, + "loss": 3.3361, + "step": 7401 + }, + { + "epoch": 0.5934537291295023, + "grad_norm": 1.8631203174591064, + "learning_rate": 1.6495097553720426e-05, + "loss": 3.2677, + "step": 7402 + }, + { + "epoch": 0.593533903910525, + "grad_norm": 1.8861852884292603, + "learning_rate": 1.649411022052018e-05, + "loss": 3.2673, + "step": 7403 + }, + { + "epoch": 0.5936140786915476, + "grad_norm": 1.815930724143982, + "learning_rate": 1.6493122777831524e-05, + "loss": 3.2825, + "step": 7404 + }, + { + "epoch": 0.5936942534725702, + "grad_norm": 1.8628586530685425, + "learning_rate": 1.649213522567111e-05, + "loss": 3.2008, + "step": 7405 + }, + { + "epoch": 0.5937744282535928, + "grad_norm": 1.8027600049972534, + "learning_rate": 1.6491147564055578e-05, + "loss": 3.1957, + "step": 7406 + }, + { + "epoch": 0.5938546030346155, + "grad_norm": 1.8452582359313965, + "learning_rate": 1.649015979300159e-05, + "loss": 3.2823, + "step": 7407 + }, + { + "epoch": 0.5939347778156381, + "grad_norm": 1.8757120370864868, + "learning_rate": 1.6489171912525798e-05, + "loss": 3.2494, + "step": 7408 + }, + { + "epoch": 0.5940149525966607, + "grad_norm": 1.891700029373169, + "learning_rate": 1.648818392264485e-05, + "loss": 3.3571, + "step": 7409 + }, + { + "epoch": 0.5940951273776833, + "grad_norm": 1.8430650234222412, + "learning_rate": 1.648719582337541e-05, + "loss": 3.2833, + "step": 7410 + }, + { + "epoch": 0.594175302158706, + "grad_norm": 0.765129029750824, + "learning_rate": 1.648620761473413e-05, + "loss": 2.5088, + "step": 7411 + }, + { + "epoch": 0.5942554769397286, + "grad_norm": 1.7609386444091797, + "learning_rate": 1.648521929673768e-05, + "loss": 3.3105, + "step": 7412 + }, + { + "epoch": 0.5943356517207512, + "grad_norm": 1.8237136602401733, + "learning_rate": 1.648423086940272e-05, + "loss": 3.2466, + "step": 7413 + }, + { + "epoch": 0.5944158265017738, + "grad_norm": 1.6759742498397827, + "learning_rate": 1.648324233274591e-05, + "loss": 3.1987, + "step": 7414 + }, + { + "epoch": 0.5944960012827964, + "grad_norm": 1.907923936843872, + "learning_rate": 1.648225368678392e-05, + "loss": 3.3101, + "step": 7415 + }, + { + "epoch": 0.5945761760638191, + "grad_norm": 1.8793177604675293, + "learning_rate": 1.6481264931533417e-05, + "loss": 3.2738, + "step": 7416 + }, + { + "epoch": 0.5946563508448418, + "grad_norm": 1.795082926750183, + "learning_rate": 1.6480276067011073e-05, + "loss": 3.2255, + "step": 7417 + }, + { + "epoch": 0.5947365256258644, + "grad_norm": 1.7823947668075562, + "learning_rate": 1.6479287093233555e-05, + "loss": 3.2162, + "step": 7418 + }, + { + "epoch": 0.5948167004068871, + "grad_norm": 1.8034801483154297, + "learning_rate": 1.647829801021754e-05, + "loss": 3.176, + "step": 7419 + }, + { + "epoch": 0.5948968751879097, + "grad_norm": 1.740786075592041, + "learning_rate": 1.6477308817979708e-05, + "loss": 3.1469, + "step": 7420 + }, + { + "epoch": 0.5949770499689323, + "grad_norm": 1.8314311504364014, + "learning_rate": 1.647631951653673e-05, + "loss": 3.2516, + "step": 7421 + }, + { + "epoch": 0.5950572247499549, + "grad_norm": 1.796170711517334, + "learning_rate": 1.647533010590529e-05, + "loss": 3.199, + "step": 7422 + }, + { + "epoch": 0.5951373995309776, + "grad_norm": 1.9598388671875, + "learning_rate": 1.647434058610206e-05, + "loss": 3.165, + "step": 7423 + }, + { + "epoch": 0.5952175743120002, + "grad_norm": 1.828658103942871, + "learning_rate": 1.6473350957143734e-05, + "loss": 3.2924, + "step": 7424 + }, + { + "epoch": 0.5952977490930228, + "grad_norm": 1.8419369459152222, + "learning_rate": 1.6472361219046994e-05, + "loss": 3.3329, + "step": 7425 + }, + { + "epoch": 0.5953779238740454, + "grad_norm": 1.919709324836731, + "learning_rate": 1.647137137182852e-05, + "loss": 3.1976, + "step": 7426 + }, + { + "epoch": 0.595458098655068, + "grad_norm": 1.8849421739578247, + "learning_rate": 1.6470381415505008e-05, + "loss": 3.2577, + "step": 7427 + }, + { + "epoch": 0.5955382734360907, + "grad_norm": 1.7911432981491089, + "learning_rate": 1.6469391350093148e-05, + "loss": 3.2408, + "step": 7428 + }, + { + "epoch": 0.5956184482171133, + "grad_norm": 0.8248193264007568, + "learning_rate": 1.6468401175609624e-05, + "loss": 2.5073, + "step": 7429 + }, + { + "epoch": 0.5956986229981359, + "grad_norm": 1.784507155418396, + "learning_rate": 1.646741089207114e-05, + "loss": 3.2061, + "step": 7430 + }, + { + "epoch": 0.5957787977791585, + "grad_norm": 1.8588286638259888, + "learning_rate": 1.6466420499494387e-05, + "loss": 3.3104, + "step": 7431 + }, + { + "epoch": 0.5958589725601812, + "grad_norm": 1.8360635042190552, + "learning_rate": 1.6465429997896064e-05, + "loss": 3.099, + "step": 7432 + }, + { + "epoch": 0.5959391473412038, + "grad_norm": 1.89199960231781, + "learning_rate": 1.646443938729287e-05, + "loss": 3.2762, + "step": 7433 + }, + { + "epoch": 0.5960193221222264, + "grad_norm": 1.8360085487365723, + "learning_rate": 1.6463448667701504e-05, + "loss": 3.1783, + "step": 7434 + }, + { + "epoch": 0.596099496903249, + "grad_norm": 1.854379653930664, + "learning_rate": 1.6462457839138674e-05, + "loss": 3.1551, + "step": 7435 + }, + { + "epoch": 0.5961796716842717, + "grad_norm": 0.7593839168548584, + "learning_rate": 1.6461466901621082e-05, + "loss": 2.5015, + "step": 7436 + }, + { + "epoch": 0.5962598464652944, + "grad_norm": 1.783795952796936, + "learning_rate": 1.6460475855165434e-05, + "loss": 3.3119, + "step": 7437 + }, + { + "epoch": 0.596340021246317, + "grad_norm": 1.8537237644195557, + "learning_rate": 1.6459484699788443e-05, + "loss": 3.2399, + "step": 7438 + }, + { + "epoch": 0.5964201960273396, + "grad_norm": 1.8545418977737427, + "learning_rate": 1.6458493435506814e-05, + "loss": 3.102, + "step": 7439 + }, + { + "epoch": 0.5965003708083623, + "grad_norm": 1.8082324266433716, + "learning_rate": 1.645750206233726e-05, + "loss": 3.196, + "step": 7440 + }, + { + "epoch": 0.5965805455893849, + "grad_norm": 1.8373875617980957, + "learning_rate": 1.64565105802965e-05, + "loss": 3.1265, + "step": 7441 + }, + { + "epoch": 0.5966607203704075, + "grad_norm": 1.857001543045044, + "learning_rate": 1.6455518989401243e-05, + "loss": 3.2841, + "step": 7442 + }, + { + "epoch": 0.5967408951514301, + "grad_norm": 1.7550851106643677, + "learning_rate": 1.645452728966822e-05, + "loss": 3.2423, + "step": 7443 + }, + { + "epoch": 0.5968210699324528, + "grad_norm": 0.8089597225189209, + "learning_rate": 1.6453535481114135e-05, + "loss": 2.5215, + "step": 7444 + }, + { + "epoch": 0.5969012447134754, + "grad_norm": 1.8813238143920898, + "learning_rate": 1.6452543563755714e-05, + "loss": 3.3037, + "step": 7445 + }, + { + "epoch": 0.596981419494498, + "grad_norm": 1.8256267309188843, + "learning_rate": 1.6451551537609687e-05, + "loss": 3.1956, + "step": 7446 + }, + { + "epoch": 0.5970615942755206, + "grad_norm": 1.7740134000778198, + "learning_rate": 1.6450559402692774e-05, + "loss": 3.1007, + "step": 7447 + }, + { + "epoch": 0.5971417690565433, + "grad_norm": 1.7511061429977417, + "learning_rate": 1.6449567159021704e-05, + "loss": 3.2701, + "step": 7448 + }, + { + "epoch": 0.5972219438375659, + "grad_norm": 1.8263574838638306, + "learning_rate": 1.6448574806613202e-05, + "loss": 3.2883, + "step": 7449 + }, + { + "epoch": 0.5973021186185885, + "grad_norm": 1.844450831413269, + "learning_rate": 1.6447582345484003e-05, + "loss": 3.2499, + "step": 7450 + }, + { + "epoch": 0.5973822933996111, + "grad_norm": 1.8311299085617065, + "learning_rate": 1.6446589775650837e-05, + "loss": 3.2251, + "step": 7451 + }, + { + "epoch": 0.5974624681806338, + "grad_norm": 1.810072422027588, + "learning_rate": 1.644559709713044e-05, + "loss": 3.3087, + "step": 7452 + }, + { + "epoch": 0.5975426429616564, + "grad_norm": 1.7510439157485962, + "learning_rate": 1.6444604309939546e-05, + "loss": 3.1763, + "step": 7453 + }, + { + "epoch": 0.597622817742679, + "grad_norm": 1.7466790676116943, + "learning_rate": 1.64436114140949e-05, + "loss": 3.223, + "step": 7454 + }, + { + "epoch": 0.5977029925237016, + "grad_norm": 1.9660077095031738, + "learning_rate": 1.644261840961323e-05, + "loss": 3.2217, + "step": 7455 + }, + { + "epoch": 0.5977831673047243, + "grad_norm": 1.950912356376648, + "learning_rate": 1.6441625296511287e-05, + "loss": 3.2065, + "step": 7456 + }, + { + "epoch": 0.5978633420857469, + "grad_norm": 1.847819447517395, + "learning_rate": 1.644063207480581e-05, + "loss": 3.1407, + "step": 7457 + }, + { + "epoch": 0.5979435168667696, + "grad_norm": 1.6522715091705322, + "learning_rate": 1.643963874451355e-05, + "loss": 3.1937, + "step": 7458 + }, + { + "epoch": 0.5980236916477922, + "grad_norm": 0.7596213221549988, + "learning_rate": 1.6438645305651245e-05, + "loss": 2.4883, + "step": 7459 + }, + { + "epoch": 0.5981038664288149, + "grad_norm": 1.9150242805480957, + "learning_rate": 1.6437651758235654e-05, + "loss": 3.1979, + "step": 7460 + }, + { + "epoch": 0.5981840412098375, + "grad_norm": 1.882312536239624, + "learning_rate": 1.643665810228352e-05, + "loss": 3.2484, + "step": 7461 + }, + { + "epoch": 0.5982642159908601, + "grad_norm": 1.728355884552002, + "learning_rate": 1.6435664337811604e-05, + "loss": 3.2367, + "step": 7462 + }, + { + "epoch": 0.5983443907718827, + "grad_norm": 1.7743088006973267, + "learning_rate": 1.6434670464836653e-05, + "loss": 3.2747, + "step": 7463 + }, + { + "epoch": 0.5984245655529054, + "grad_norm": 1.7850667238235474, + "learning_rate": 1.6433676483375424e-05, + "loss": 3.3199, + "step": 7464 + }, + { + "epoch": 0.598504740333928, + "grad_norm": 1.7388780117034912, + "learning_rate": 1.6432682393444676e-05, + "loss": 3.2429, + "step": 7465 + }, + { + "epoch": 0.5985849151149506, + "grad_norm": 0.7718990445137024, + "learning_rate": 1.6431688195061172e-05, + "loss": 2.4933, + "step": 7466 + }, + { + "epoch": 0.5986650898959732, + "grad_norm": 1.8264182806015015, + "learning_rate": 1.6430693888241672e-05, + "loss": 3.22, + "step": 7467 + }, + { + "epoch": 0.5987452646769958, + "grad_norm": 1.8594974279403687, + "learning_rate": 1.6429699473002937e-05, + "loss": 3.1165, + "step": 7468 + }, + { + "epoch": 0.5988254394580185, + "grad_norm": 1.935067057609558, + "learning_rate": 1.642870494936174e-05, + "loss": 3.2526, + "step": 7469 + }, + { + "epoch": 0.5989056142390411, + "grad_norm": 1.9440703392028809, + "learning_rate": 1.642771031733484e-05, + "loss": 3.229, + "step": 7470 + }, + { + "epoch": 0.5989857890200637, + "grad_norm": 1.8551279306411743, + "learning_rate": 1.6426715576939007e-05, + "loss": 3.1683, + "step": 7471 + }, + { + "epoch": 0.5990659638010863, + "grad_norm": 1.8517533540725708, + "learning_rate": 1.642572072819102e-05, + "loss": 3.1549, + "step": 7472 + }, + { + "epoch": 0.599146138582109, + "grad_norm": 1.7323414087295532, + "learning_rate": 1.6424725771107642e-05, + "loss": 3.222, + "step": 7473 + }, + { + "epoch": 0.5992263133631316, + "grad_norm": 1.7771168947219849, + "learning_rate": 1.642373070570565e-05, + "loss": 3.1802, + "step": 7474 + }, + { + "epoch": 0.5993064881441542, + "grad_norm": 1.7670423984527588, + "learning_rate": 1.6422735532001825e-05, + "loss": 3.1147, + "step": 7475 + }, + { + "epoch": 0.5993866629251768, + "grad_norm": 1.8847827911376953, + "learning_rate": 1.6421740250012943e-05, + "loss": 3.2977, + "step": 7476 + }, + { + "epoch": 0.5994668377061995, + "grad_norm": 2.1504909992218018, + "learning_rate": 1.6420744859755782e-05, + "loss": 3.1609, + "step": 7477 + }, + { + "epoch": 0.5995470124872222, + "grad_norm": 0.7929425239562988, + "learning_rate": 1.641974936124713e-05, + "loss": 2.4879, + "step": 7478 + }, + { + "epoch": 0.5996271872682448, + "grad_norm": 1.8380274772644043, + "learning_rate": 1.6418753754503764e-05, + "loss": 3.2767, + "step": 7479 + }, + { + "epoch": 0.5997073620492674, + "grad_norm": 1.7003157138824463, + "learning_rate": 1.641775803954247e-05, + "loss": 3.1973, + "step": 7480 + }, + { + "epoch": 0.5997875368302901, + "grad_norm": 1.8375051021575928, + "learning_rate": 1.6416762216380036e-05, + "loss": 3.2244, + "step": 7481 + }, + { + "epoch": 0.5998677116113127, + "grad_norm": 1.839215636253357, + "learning_rate": 1.6415766285033258e-05, + "loss": 3.3033, + "step": 7482 + }, + { + "epoch": 0.5999478863923353, + "grad_norm": 2.0440127849578857, + "learning_rate": 1.6414770245518916e-05, + "loss": 3.3013, + "step": 7483 + }, + { + "epoch": 0.6000280611733579, + "grad_norm": 1.7097609043121338, + "learning_rate": 1.641377409785381e-05, + "loss": 3.2029, + "step": 7484 + }, + { + "epoch": 0.6001082359543806, + "grad_norm": 1.8901008367538452, + "learning_rate": 1.6412777842054735e-05, + "loss": 3.1799, + "step": 7485 + }, + { + "epoch": 0.6001884107354032, + "grad_norm": 1.803040623664856, + "learning_rate": 1.6411781478138483e-05, + "loss": 3.1682, + "step": 7486 + }, + { + "epoch": 0.6002685855164258, + "grad_norm": 1.8096733093261719, + "learning_rate": 1.6410785006121857e-05, + "loss": 3.1962, + "step": 7487 + }, + { + "epoch": 0.6003487602974484, + "grad_norm": 1.8705347776412964, + "learning_rate": 1.6409788426021655e-05, + "loss": 3.168, + "step": 7488 + }, + { + "epoch": 0.6004289350784711, + "grad_norm": 1.8819819688796997, + "learning_rate": 1.640879173785468e-05, + "loss": 3.2738, + "step": 7489 + }, + { + "epoch": 0.6005091098594937, + "grad_norm": 1.7798391580581665, + "learning_rate": 1.6407794941637735e-05, + "loss": 3.2039, + "step": 7490 + }, + { + "epoch": 0.6005892846405163, + "grad_norm": 1.9024790525436401, + "learning_rate": 1.6406798037387622e-05, + "loss": 3.3286, + "step": 7491 + }, + { + "epoch": 0.6006694594215389, + "grad_norm": 1.8979130983352661, + "learning_rate": 1.6405801025121155e-05, + "loss": 3.1851, + "step": 7492 + }, + { + "epoch": 0.6007496342025616, + "grad_norm": 1.9555219411849976, + "learning_rate": 1.640480390485514e-05, + "loss": 3.2455, + "step": 7493 + }, + { + "epoch": 0.6008298089835842, + "grad_norm": 1.8674567937850952, + "learning_rate": 1.6403806676606388e-05, + "loss": 3.0931, + "step": 7494 + }, + { + "epoch": 0.6009099837646068, + "grad_norm": 1.9359363317489624, + "learning_rate": 1.6402809340391712e-05, + "loss": 3.2709, + "step": 7495 + }, + { + "epoch": 0.6009901585456294, + "grad_norm": 1.8311628103256226, + "learning_rate": 1.640181189622793e-05, + "loss": 3.1994, + "step": 7496 + }, + { + "epoch": 0.601070333326652, + "grad_norm": 1.9473377466201782, + "learning_rate": 1.6400814344131853e-05, + "loss": 3.2969, + "step": 7497 + }, + { + "epoch": 0.6011505081076748, + "grad_norm": 1.7993464469909668, + "learning_rate": 1.63998166841203e-05, + "loss": 3.2594, + "step": 7498 + }, + { + "epoch": 0.6012306828886974, + "grad_norm": 1.8685302734375, + "learning_rate": 1.63988189162101e-05, + "loss": 3.1676, + "step": 7499 + }, + { + "epoch": 0.60131085766972, + "grad_norm": 1.892378568649292, + "learning_rate": 1.6397821040418062e-05, + "loss": 3.2628, + "step": 7500 + }, + { + "epoch": 0.6013910324507427, + "grad_norm": 1.802841305732727, + "learning_rate": 1.6396823056761018e-05, + "loss": 3.223, + "step": 7501 + }, + { + "epoch": 0.6014712072317653, + "grad_norm": 1.887242078781128, + "learning_rate": 1.639582496525579e-05, + "loss": 3.2137, + "step": 7502 + }, + { + "epoch": 0.6015513820127879, + "grad_norm": 1.9180864095687866, + "learning_rate": 1.639482676591921e-05, + "loss": 3.2625, + "step": 7503 + }, + { + "epoch": 0.6016315567938105, + "grad_norm": 1.6884340047836304, + "learning_rate": 1.6393828458768103e-05, + "loss": 3.2046, + "step": 7504 + }, + { + "epoch": 0.6017117315748332, + "grad_norm": 1.964465856552124, + "learning_rate": 1.6392830043819303e-05, + "loss": 3.3528, + "step": 7505 + }, + { + "epoch": 0.6017919063558558, + "grad_norm": 1.8272628784179688, + "learning_rate": 1.6391831521089637e-05, + "loss": 3.2624, + "step": 7506 + }, + { + "epoch": 0.6018720811368784, + "grad_norm": 0.8325141668319702, + "learning_rate": 1.639083289059595e-05, + "loss": 2.5243, + "step": 7507 + }, + { + "epoch": 0.601952255917901, + "grad_norm": 1.8036843538284302, + "learning_rate": 1.638983415235507e-05, + "loss": 3.2382, + "step": 7508 + }, + { + "epoch": 0.6020324306989236, + "grad_norm": 1.8752498626708984, + "learning_rate": 1.638883530638384e-05, + "loss": 3.2701, + "step": 7509 + }, + { + "epoch": 0.6021126054799463, + "grad_norm": 1.8343528509140015, + "learning_rate": 1.63878363526991e-05, + "loss": 3.3245, + "step": 7510 + }, + { + "epoch": 0.6021927802609689, + "grad_norm": 1.9691801071166992, + "learning_rate": 1.638683729131768e-05, + "loss": 3.3026, + "step": 7511 + }, + { + "epoch": 0.6022729550419915, + "grad_norm": 1.8121371269226074, + "learning_rate": 1.638583812225645e-05, + "loss": 3.2315, + "step": 7512 + }, + { + "epoch": 0.6023531298230141, + "grad_norm": 1.7812563180923462, + "learning_rate": 1.638483884553223e-05, + "loss": 3.237, + "step": 7513 + }, + { + "epoch": 0.6024333046040368, + "grad_norm": 1.9630824327468872, + "learning_rate": 1.6383839461161878e-05, + "loss": 3.2257, + "step": 7514 + }, + { + "epoch": 0.6025134793850594, + "grad_norm": 1.7775285243988037, + "learning_rate": 1.6382839969162242e-05, + "loss": 3.2239, + "step": 7515 + }, + { + "epoch": 0.602593654166082, + "grad_norm": 1.8327176570892334, + "learning_rate": 1.6381840369550174e-05, + "loss": 3.2188, + "step": 7516 + }, + { + "epoch": 0.6026738289471046, + "grad_norm": 1.8591612577438354, + "learning_rate": 1.638084066234253e-05, + "loss": 3.1486, + "step": 7517 + }, + { + "epoch": 0.6027540037281273, + "grad_norm": 0.7735044360160828, + "learning_rate": 1.6379840847556158e-05, + "loss": 2.5435, + "step": 7518 + }, + { + "epoch": 0.60283417850915, + "grad_norm": 2.171210527420044, + "learning_rate": 1.6378840925207918e-05, + "loss": 3.18, + "step": 7519 + }, + { + "epoch": 0.6029143532901726, + "grad_norm": 1.7700986862182617, + "learning_rate": 1.6377840895314666e-05, + "loss": 3.2122, + "step": 7520 + }, + { + "epoch": 0.6029945280711952, + "grad_norm": 1.7183643579483032, + "learning_rate": 1.637684075789327e-05, + "loss": 3.228, + "step": 7521 + }, + { + "epoch": 0.6030747028522179, + "grad_norm": 1.747893214225769, + "learning_rate": 1.6375840512960584e-05, + "loss": 3.126, + "step": 7522 + }, + { + "epoch": 0.6031548776332405, + "grad_norm": 1.9419615268707275, + "learning_rate": 1.6374840160533473e-05, + "loss": 3.246, + "step": 7523 + }, + { + "epoch": 0.6032350524142631, + "grad_norm": 1.8518073558807373, + "learning_rate": 1.6373839700628804e-05, + "loss": 3.2375, + "step": 7524 + }, + { + "epoch": 0.6033152271952857, + "grad_norm": 1.9585212469100952, + "learning_rate": 1.6372839133263443e-05, + "loss": 3.2831, + "step": 7525 + }, + { + "epoch": 0.6033954019763084, + "grad_norm": 1.851931095123291, + "learning_rate": 1.637183845845426e-05, + "loss": 3.2119, + "step": 7526 + }, + { + "epoch": 0.603475576757331, + "grad_norm": 1.7478411197662354, + "learning_rate": 1.637083767621813e-05, + "loss": 3.1362, + "step": 7527 + }, + { + "epoch": 0.6035557515383536, + "grad_norm": 1.8098430633544922, + "learning_rate": 1.6369836786571917e-05, + "loss": 3.1957, + "step": 7528 + }, + { + "epoch": 0.6036359263193762, + "grad_norm": 1.88503897190094, + "learning_rate": 1.6368835789532506e-05, + "loss": 3.2793, + "step": 7529 + }, + { + "epoch": 0.6037161011003989, + "grad_norm": 1.8161723613739014, + "learning_rate": 1.6367834685116764e-05, + "loss": 3.1938, + "step": 7530 + }, + { + "epoch": 0.6037962758814215, + "grad_norm": 1.9427374601364136, + "learning_rate": 1.6366833473341575e-05, + "loss": 3.1333, + "step": 7531 + }, + { + "epoch": 0.6038764506624441, + "grad_norm": 1.821117639541626, + "learning_rate": 1.6365832154223818e-05, + "loss": 3.1686, + "step": 7532 + }, + { + "epoch": 0.6039566254434667, + "grad_norm": 1.7790356874465942, + "learning_rate": 1.6364830727780373e-05, + "loss": 3.1736, + "step": 7533 + }, + { + "epoch": 0.6040368002244894, + "grad_norm": 1.792697548866272, + "learning_rate": 1.6363829194028126e-05, + "loss": 3.1566, + "step": 7534 + }, + { + "epoch": 0.604116975005512, + "grad_norm": 1.7769849300384521, + "learning_rate": 1.636282755298396e-05, + "loss": 3.1648, + "step": 7535 + }, + { + "epoch": 0.6041971497865346, + "grad_norm": 1.930041790008545, + "learning_rate": 1.6361825804664767e-05, + "loss": 3.2706, + "step": 7536 + }, + { + "epoch": 0.6042773245675572, + "grad_norm": 1.8518003225326538, + "learning_rate": 1.636082394908743e-05, + "loss": 3.2489, + "step": 7537 + }, + { + "epoch": 0.6043574993485799, + "grad_norm": 1.7848459482192993, + "learning_rate": 1.6359821986268846e-05, + "loss": 3.2057, + "step": 7538 + }, + { + "epoch": 0.6044376741296026, + "grad_norm": 1.7775485515594482, + "learning_rate": 1.6358819916225902e-05, + "loss": 3.2711, + "step": 7539 + }, + { + "epoch": 0.6045178489106252, + "grad_norm": 0.841618537902832, + "learning_rate": 1.6357817738975498e-05, + "loss": 2.4962, + "step": 7540 + }, + { + "epoch": 0.6045980236916478, + "grad_norm": 1.7682527303695679, + "learning_rate": 1.6356815454534525e-05, + "loss": 3.1783, + "step": 7541 + }, + { + "epoch": 0.6046781984726705, + "grad_norm": 1.8024849891662598, + "learning_rate": 1.6355813062919887e-05, + "loss": 3.1726, + "step": 7542 + }, + { + "epoch": 0.6047583732536931, + "grad_norm": 1.9040886163711548, + "learning_rate": 1.635481056414848e-05, + "loss": 3.2593, + "step": 7543 + }, + { + "epoch": 0.6048385480347157, + "grad_norm": 1.870457410812378, + "learning_rate": 1.63538079582372e-05, + "loss": 3.1873, + "step": 7544 + }, + { + "epoch": 0.6049187228157383, + "grad_norm": 1.7682136297225952, + "learning_rate": 1.6352805245202964e-05, + "loss": 3.2003, + "step": 7545 + }, + { + "epoch": 0.604998897596761, + "grad_norm": 1.8524073362350464, + "learning_rate": 1.6351802425062668e-05, + "loss": 3.26, + "step": 7546 + }, + { + "epoch": 0.6050790723777836, + "grad_norm": 1.8143091201782227, + "learning_rate": 1.6350799497833227e-05, + "loss": 3.2214, + "step": 7547 + }, + { + "epoch": 0.6051592471588062, + "grad_norm": 1.9781289100646973, + "learning_rate": 1.6349796463531536e-05, + "loss": 3.2275, + "step": 7548 + }, + { + "epoch": 0.6052394219398288, + "grad_norm": 2.0076382160186768, + "learning_rate": 1.634879332217452e-05, + "loss": 3.2694, + "step": 7549 + }, + { + "epoch": 0.6053195967208514, + "grad_norm": 1.8261717557907104, + "learning_rate": 1.6347790073779086e-05, + "loss": 3.1489, + "step": 7550 + }, + { + "epoch": 0.6053997715018741, + "grad_norm": 1.7761170864105225, + "learning_rate": 1.6346786718362147e-05, + "loss": 3.2278, + "step": 7551 + }, + { + "epoch": 0.6054799462828967, + "grad_norm": 1.8720577955245972, + "learning_rate": 1.6345783255940623e-05, + "loss": 3.2861, + "step": 7552 + }, + { + "epoch": 0.6055601210639193, + "grad_norm": 1.873124361038208, + "learning_rate": 1.6344779686531428e-05, + "loss": 3.2529, + "step": 7553 + }, + { + "epoch": 0.605640295844942, + "grad_norm": 1.7643424272537231, + "learning_rate": 1.6343776010151482e-05, + "loss": 3.1475, + "step": 7554 + }, + { + "epoch": 0.6057204706259646, + "grad_norm": 1.8591279983520508, + "learning_rate": 1.634277222681771e-05, + "loss": 3.2747, + "step": 7555 + }, + { + "epoch": 0.6058006454069872, + "grad_norm": 1.8720366954803467, + "learning_rate": 1.6341768336547033e-05, + "loss": 3.1265, + "step": 7556 + }, + { + "epoch": 0.6058808201880098, + "grad_norm": 0.7444832921028137, + "learning_rate": 1.6340764339356377e-05, + "loss": 2.5155, + "step": 7557 + }, + { + "epoch": 0.6059609949690324, + "grad_norm": 1.7336993217468262, + "learning_rate": 1.633976023526267e-05, + "loss": 3.1357, + "step": 7558 + }, + { + "epoch": 0.6060411697500552, + "grad_norm": 1.9241639375686646, + "learning_rate": 1.633875602428284e-05, + "loss": 3.2221, + "step": 7559 + }, + { + "epoch": 0.6061213445310778, + "grad_norm": 1.8564345836639404, + "learning_rate": 1.6337751706433813e-05, + "loss": 3.251, + "step": 7560 + }, + { + "epoch": 0.6062015193121004, + "grad_norm": 1.825809359550476, + "learning_rate": 1.633674728173253e-05, + "loss": 3.1982, + "step": 7561 + }, + { + "epoch": 0.606281694093123, + "grad_norm": 1.78408682346344, + "learning_rate": 1.633574275019592e-05, + "loss": 3.153, + "step": 7562 + }, + { + "epoch": 0.6063618688741457, + "grad_norm": 1.86508309841156, + "learning_rate": 1.6334738111840925e-05, + "loss": 3.2314, + "step": 7563 + }, + { + "epoch": 0.6064420436551683, + "grad_norm": 0.816183865070343, + "learning_rate": 1.6333733366684472e-05, + "loss": 2.4938, + "step": 7564 + }, + { + "epoch": 0.6065222184361909, + "grad_norm": 1.8081949949264526, + "learning_rate": 1.6332728514743508e-05, + "loss": 3.2496, + "step": 7565 + }, + { + "epoch": 0.6066023932172135, + "grad_norm": 1.8685235977172852, + "learning_rate": 1.6331723556034972e-05, + "loss": 3.2103, + "step": 7566 + }, + { + "epoch": 0.6066825679982362, + "grad_norm": 1.8272720575332642, + "learning_rate": 1.633071849057581e-05, + "loss": 3.2955, + "step": 7567 + }, + { + "epoch": 0.6067627427792588, + "grad_norm": 1.7291805744171143, + "learning_rate": 1.6329713318382964e-05, + "loss": 3.1767, + "step": 7568 + }, + { + "epoch": 0.6068429175602814, + "grad_norm": 1.8996095657348633, + "learning_rate": 1.6328708039473386e-05, + "loss": 3.3213, + "step": 7569 + }, + { + "epoch": 0.606923092341304, + "grad_norm": 1.8115580081939697, + "learning_rate": 1.6327702653864016e-05, + "loss": 3.1824, + "step": 7570 + }, + { + "epoch": 0.6070032671223267, + "grad_norm": 2.325981855392456, + "learning_rate": 1.6326697161571815e-05, + "loss": 3.3288, + "step": 7571 + }, + { + "epoch": 0.6070834419033493, + "grad_norm": 1.8051984310150146, + "learning_rate": 1.6325691562613724e-05, + "loss": 3.2465, + "step": 7572 + }, + { + "epoch": 0.6071636166843719, + "grad_norm": 0.8113737106323242, + "learning_rate": 1.6324685857006705e-05, + "loss": 2.4967, + "step": 7573 + }, + { + "epoch": 0.6072437914653945, + "grad_norm": 1.8697795867919922, + "learning_rate": 1.632368004476771e-05, + "loss": 3.1472, + "step": 7574 + }, + { + "epoch": 0.6073239662464172, + "grad_norm": 1.9612358808517456, + "learning_rate": 1.6322674125913702e-05, + "loss": 3.1866, + "step": 7575 + }, + { + "epoch": 0.6074041410274398, + "grad_norm": 1.9159125089645386, + "learning_rate": 1.6321668100461634e-05, + "loss": 3.2075, + "step": 7576 + }, + { + "epoch": 0.6074843158084624, + "grad_norm": 0.8058015704154968, + "learning_rate": 1.632066196842847e-05, + "loss": 2.5048, + "step": 7577 + }, + { + "epoch": 0.607564490589485, + "grad_norm": 1.7799328565597534, + "learning_rate": 1.6319655729831173e-05, + "loss": 3.1895, + "step": 7578 + }, + { + "epoch": 0.6076446653705077, + "grad_norm": 1.9276833534240723, + "learning_rate": 1.6318649384686705e-05, + "loss": 3.3259, + "step": 7579 + }, + { + "epoch": 0.6077248401515304, + "grad_norm": 1.7533013820648193, + "learning_rate": 1.631764293301204e-05, + "loss": 3.2603, + "step": 7580 + }, + { + "epoch": 0.607805014932553, + "grad_norm": 2.0861239433288574, + "learning_rate": 1.6316636374824138e-05, + "loss": 3.2396, + "step": 7581 + }, + { + "epoch": 0.6078851897135756, + "grad_norm": 1.8200020790100098, + "learning_rate": 1.6315629710139974e-05, + "loss": 3.1937, + "step": 7582 + }, + { + "epoch": 0.6079653644945983, + "grad_norm": 1.7678812742233276, + "learning_rate": 1.6314622938976522e-05, + "loss": 3.1721, + "step": 7583 + }, + { + "epoch": 0.6080455392756209, + "grad_norm": 1.7329235076904297, + "learning_rate": 1.631361606135075e-05, + "loss": 3.2333, + "step": 7584 + }, + { + "epoch": 0.6081257140566435, + "grad_norm": 1.8845829963684082, + "learning_rate": 1.6312609077279635e-05, + "loss": 3.3315, + "step": 7585 + }, + { + "epoch": 0.6082058888376661, + "grad_norm": 1.8532882928848267, + "learning_rate": 1.631160198678016e-05, + "loss": 3.3016, + "step": 7586 + }, + { + "epoch": 0.6082860636186888, + "grad_norm": 1.73226797580719, + "learning_rate": 1.6310594789869295e-05, + "loss": 3.1848, + "step": 7587 + }, + { + "epoch": 0.6083662383997114, + "grad_norm": 1.9214637279510498, + "learning_rate": 1.6309587486564028e-05, + "loss": 3.2557, + "step": 7588 + }, + { + "epoch": 0.608446413180734, + "grad_norm": 1.7157038450241089, + "learning_rate": 1.630858007688134e-05, + "loss": 3.298, + "step": 7589 + }, + { + "epoch": 0.6085265879617566, + "grad_norm": 1.829483985900879, + "learning_rate": 1.6307572560838214e-05, + "loss": 3.1857, + "step": 7590 + }, + { + "epoch": 0.6086067627427793, + "grad_norm": 1.761509895324707, + "learning_rate": 1.630656493845164e-05, + "loss": 3.1719, + "step": 7591 + }, + { + "epoch": 0.6086869375238019, + "grad_norm": 1.9353934526443481, + "learning_rate": 1.63055572097386e-05, + "loss": 3.1445, + "step": 7592 + }, + { + "epoch": 0.6087671123048245, + "grad_norm": 1.7676464319229126, + "learning_rate": 1.6304549374716094e-05, + "loss": 3.165, + "step": 7593 + }, + { + "epoch": 0.6088472870858471, + "grad_norm": 1.896835446357727, + "learning_rate": 1.63035414334011e-05, + "loss": 3.3605, + "step": 7594 + }, + { + "epoch": 0.6089274618668697, + "grad_norm": 1.950514554977417, + "learning_rate": 1.630253338581062e-05, + "loss": 3.2385, + "step": 7595 + }, + { + "epoch": 0.6090076366478924, + "grad_norm": 1.8196215629577637, + "learning_rate": 1.6301525231961655e-05, + "loss": 3.199, + "step": 7596 + }, + { + "epoch": 0.609087811428915, + "grad_norm": 1.822407603263855, + "learning_rate": 1.6300516971871192e-05, + "loss": 3.2279, + "step": 7597 + }, + { + "epoch": 0.6091679862099376, + "grad_norm": 1.8671317100524902, + "learning_rate": 1.629950860555623e-05, + "loss": 3.209, + "step": 7598 + }, + { + "epoch": 0.6092481609909602, + "grad_norm": 2.076141119003296, + "learning_rate": 1.6298500133033778e-05, + "loss": 3.2073, + "step": 7599 + }, + { + "epoch": 0.609328335771983, + "grad_norm": 1.879074215888977, + "learning_rate": 1.629749155432083e-05, + "loss": 3.2676, + "step": 7600 + }, + { + "epoch": 0.6094085105530056, + "grad_norm": 1.7771075963974, + "learning_rate": 1.6296482869434397e-05, + "loss": 3.3293, + "step": 7601 + }, + { + "epoch": 0.6094886853340282, + "grad_norm": 0.7857135534286499, + "learning_rate": 1.629547407839148e-05, + "loss": 2.5104, + "step": 7602 + }, + { + "epoch": 0.6095688601150508, + "grad_norm": 1.9830636978149414, + "learning_rate": 1.6294465181209088e-05, + "loss": 3.1271, + "step": 7603 + }, + { + "epoch": 0.6096490348960735, + "grad_norm": 1.7953625917434692, + "learning_rate": 1.6293456177904233e-05, + "loss": 3.2386, + "step": 7604 + }, + { + "epoch": 0.6097292096770961, + "grad_norm": 1.7954078912734985, + "learning_rate": 1.6292447068493925e-05, + "loss": 3.1578, + "step": 7605 + }, + { + "epoch": 0.6098093844581187, + "grad_norm": 1.875624656677246, + "learning_rate": 1.6291437852995174e-05, + "loss": 3.2513, + "step": 7606 + }, + { + "epoch": 0.6098895592391413, + "grad_norm": 1.8116638660430908, + "learning_rate": 1.6290428531425e-05, + "loss": 3.2147, + "step": 7607 + }, + { + "epoch": 0.609969734020164, + "grad_norm": 1.871699571609497, + "learning_rate": 1.628941910380042e-05, + "loss": 3.2723, + "step": 7608 + }, + { + "epoch": 0.6100499088011866, + "grad_norm": 1.8389250040054321, + "learning_rate": 1.628840957013845e-05, + "loss": 3.2291, + "step": 7609 + }, + { + "epoch": 0.6101300835822092, + "grad_norm": 1.9380265474319458, + "learning_rate": 1.6287399930456107e-05, + "loss": 3.2457, + "step": 7610 + }, + { + "epoch": 0.6102102583632318, + "grad_norm": 1.7535827159881592, + "learning_rate": 1.628639018477042e-05, + "loss": 3.1765, + "step": 7611 + }, + { + "epoch": 0.6102904331442545, + "grad_norm": 1.767364263534546, + "learning_rate": 1.628538033309841e-05, + "loss": 3.1367, + "step": 7612 + }, + { + "epoch": 0.6103706079252771, + "grad_norm": 1.7795966863632202, + "learning_rate": 1.6284370375457102e-05, + "loss": 3.1757, + "step": 7613 + }, + { + "epoch": 0.6104507827062997, + "grad_norm": 1.725055456161499, + "learning_rate": 1.6283360311863526e-05, + "loss": 3.2629, + "step": 7614 + }, + { + "epoch": 0.6105309574873223, + "grad_norm": 1.7638691663742065, + "learning_rate": 1.628235014233471e-05, + "loss": 3.2674, + "step": 7615 + }, + { + "epoch": 0.610611132268345, + "grad_norm": 1.736254334449768, + "learning_rate": 1.6281339866887684e-05, + "loss": 3.2071, + "step": 7616 + }, + { + "epoch": 0.6106913070493676, + "grad_norm": 1.7350990772247314, + "learning_rate": 1.628032948553948e-05, + "loss": 3.2311, + "step": 7617 + }, + { + "epoch": 0.6107714818303902, + "grad_norm": 1.8775476217269897, + "learning_rate": 1.6279318998307134e-05, + "loss": 3.27, + "step": 7618 + }, + { + "epoch": 0.6108516566114128, + "grad_norm": 0.7804147601127625, + "learning_rate": 1.6278308405207682e-05, + "loss": 2.4983, + "step": 7619 + }, + { + "epoch": 0.6109318313924356, + "grad_norm": 1.9082670211791992, + "learning_rate": 1.6277297706258165e-05, + "loss": 3.1854, + "step": 7620 + }, + { + "epoch": 0.6110120061734582, + "grad_norm": 1.7772958278656006, + "learning_rate": 1.627628690147562e-05, + "loss": 3.1528, + "step": 7621 + }, + { + "epoch": 0.6110921809544808, + "grad_norm": 1.8452497720718384, + "learning_rate": 1.6275275990877094e-05, + "loss": 3.209, + "step": 7622 + }, + { + "epoch": 0.6111723557355034, + "grad_norm": 1.8072404861450195, + "learning_rate": 1.6274264974479624e-05, + "loss": 3.2131, + "step": 7623 + }, + { + "epoch": 0.6112525305165261, + "grad_norm": 1.7276560068130493, + "learning_rate": 1.6273253852300256e-05, + "loss": 3.2505, + "step": 7624 + }, + { + "epoch": 0.6113327052975487, + "grad_norm": 1.8241994380950928, + "learning_rate": 1.6272242624356037e-05, + "loss": 3.2982, + "step": 7625 + }, + { + "epoch": 0.6114128800785713, + "grad_norm": 1.8062292337417603, + "learning_rate": 1.6271231290664024e-05, + "loss": 3.2661, + "step": 7626 + }, + { + "epoch": 0.6114930548595939, + "grad_norm": 0.7225414514541626, + "learning_rate": 1.6270219851241255e-05, + "loss": 2.4862, + "step": 7627 + }, + { + "epoch": 0.6115732296406166, + "grad_norm": 1.8248167037963867, + "learning_rate": 1.6269208306104794e-05, + "loss": 3.1956, + "step": 7628 + }, + { + "epoch": 0.6116534044216392, + "grad_norm": 2.009342908859253, + "learning_rate": 1.6268196655271688e-05, + "loss": 3.2871, + "step": 7629 + }, + { + "epoch": 0.6117335792026618, + "grad_norm": 0.738488495349884, + "learning_rate": 1.6267184898758998e-05, + "loss": 2.4901, + "step": 7630 + }, + { + "epoch": 0.6118137539836844, + "grad_norm": 1.6415834426879883, + "learning_rate": 1.6266173036583776e-05, + "loss": 3.1819, + "step": 7631 + }, + { + "epoch": 0.611893928764707, + "grad_norm": 1.8473683595657349, + "learning_rate": 1.6265161068763088e-05, + "loss": 3.2722, + "step": 7632 + }, + { + "epoch": 0.6119741035457297, + "grad_norm": 1.725093960762024, + "learning_rate": 1.626414899531399e-05, + "loss": 3.2303, + "step": 7633 + }, + { + "epoch": 0.6120542783267523, + "grad_norm": 1.8958678245544434, + "learning_rate": 1.626313681625355e-05, + "loss": 3.2517, + "step": 7634 + }, + { + "epoch": 0.6121344531077749, + "grad_norm": 1.783219337463379, + "learning_rate": 1.626212453159883e-05, + "loss": 3.1392, + "step": 7635 + }, + { + "epoch": 0.6122146278887975, + "grad_norm": 1.8534739017486572, + "learning_rate": 1.6261112141366897e-05, + "loss": 3.2394, + "step": 7636 + }, + { + "epoch": 0.6122948026698202, + "grad_norm": 1.8225102424621582, + "learning_rate": 1.6260099645574822e-05, + "loss": 3.1803, + "step": 7637 + }, + { + "epoch": 0.6123749774508428, + "grad_norm": 1.8049756288528442, + "learning_rate": 1.6259087044239666e-05, + "loss": 3.2556, + "step": 7638 + }, + { + "epoch": 0.6124551522318654, + "grad_norm": 1.7366775274276733, + "learning_rate": 1.6258074337378517e-05, + "loss": 3.2257, + "step": 7639 + }, + { + "epoch": 0.612535327012888, + "grad_norm": 1.793531894683838, + "learning_rate": 1.6257061525008436e-05, + "loss": 3.1913, + "step": 7640 + }, + { + "epoch": 0.6126155017939108, + "grad_norm": 1.7557481527328491, + "learning_rate": 1.6256048607146504e-05, + "loss": 3.2358, + "step": 7641 + }, + { + "epoch": 0.6126956765749334, + "grad_norm": 1.8285253047943115, + "learning_rate": 1.6255035583809797e-05, + "loss": 3.3188, + "step": 7642 + }, + { + "epoch": 0.612775851355956, + "grad_norm": 1.9581854343414307, + "learning_rate": 1.6254022455015394e-05, + "loss": 3.2671, + "step": 7643 + }, + { + "epoch": 0.6128560261369786, + "grad_norm": 1.749055027961731, + "learning_rate": 1.6253009220780377e-05, + "loss": 3.1926, + "step": 7644 + }, + { + "epoch": 0.6129362009180013, + "grad_norm": 0.7875562906265259, + "learning_rate": 1.6251995881121828e-05, + "loss": 2.4976, + "step": 7645 + }, + { + "epoch": 0.6130163756990239, + "grad_norm": 1.8638108968734741, + "learning_rate": 1.6250982436056832e-05, + "loss": 3.3221, + "step": 7646 + }, + { + "epoch": 0.6130965504800465, + "grad_norm": 1.8362725973129272, + "learning_rate": 1.6249968885602475e-05, + "loss": 3.2385, + "step": 7647 + }, + { + "epoch": 0.6131767252610691, + "grad_norm": 0.7482000589370728, + "learning_rate": 1.6248955229775847e-05, + "loss": 2.5217, + "step": 7648 + }, + { + "epoch": 0.6132569000420918, + "grad_norm": 1.80222749710083, + "learning_rate": 1.6247941468594035e-05, + "loss": 3.1984, + "step": 7649 + }, + { + "epoch": 0.6133370748231144, + "grad_norm": 0.7502118945121765, + "learning_rate": 1.624692760207413e-05, + "loss": 2.5142, + "step": 7650 + }, + { + "epoch": 0.613417249604137, + "grad_norm": 1.8086174726486206, + "learning_rate": 1.624591363023323e-05, + "loss": 3.2389, + "step": 7651 + }, + { + "epoch": 0.6134974243851596, + "grad_norm": 1.7489244937896729, + "learning_rate": 1.6244899553088426e-05, + "loss": 3.1873, + "step": 7652 + }, + { + "epoch": 0.6135775991661823, + "grad_norm": 1.8163915872573853, + "learning_rate": 1.6243885370656814e-05, + "loss": 3.351, + "step": 7653 + }, + { + "epoch": 0.6136577739472049, + "grad_norm": 1.8884989023208618, + "learning_rate": 1.6242871082955504e-05, + "loss": 3.2928, + "step": 7654 + }, + { + "epoch": 0.6137379487282275, + "grad_norm": 1.8518022298812866, + "learning_rate": 1.6241856690001578e-05, + "loss": 3.2159, + "step": 7655 + }, + { + "epoch": 0.6138181235092501, + "grad_norm": 1.8402575254440308, + "learning_rate": 1.6240842191812153e-05, + "loss": 3.2726, + "step": 7656 + }, + { + "epoch": 0.6138982982902728, + "grad_norm": 1.824514627456665, + "learning_rate": 1.6239827588404326e-05, + "loss": 3.2942, + "step": 7657 + }, + { + "epoch": 0.6139784730712954, + "grad_norm": 1.7550849914550781, + "learning_rate": 1.6238812879795206e-05, + "loss": 3.1177, + "step": 7658 + }, + { + "epoch": 0.614058647852318, + "grad_norm": 1.7605103254318237, + "learning_rate": 1.6237798066001902e-05, + "loss": 3.1345, + "step": 7659 + }, + { + "epoch": 0.6141388226333406, + "grad_norm": 1.7698001861572266, + "learning_rate": 1.6236783147041518e-05, + "loss": 3.1731, + "step": 7660 + }, + { + "epoch": 0.6142189974143634, + "grad_norm": 1.8250961303710938, + "learning_rate": 1.623576812293117e-05, + "loss": 3.2593, + "step": 7661 + }, + { + "epoch": 0.614299172195386, + "grad_norm": 1.8109896183013916, + "learning_rate": 1.6234752993687965e-05, + "loss": 3.2251, + "step": 7662 + }, + { + "epoch": 0.6143793469764086, + "grad_norm": 1.765188217163086, + "learning_rate": 1.6233737759329026e-05, + "loss": 3.1146, + "step": 7663 + }, + { + "epoch": 0.6144595217574312, + "grad_norm": 1.9339309930801392, + "learning_rate": 1.6232722419871465e-05, + "loss": 3.2251, + "step": 7664 + }, + { + "epoch": 0.6145396965384539, + "grad_norm": 1.816155195236206, + "learning_rate": 1.6231706975332396e-05, + "loss": 3.179, + "step": 7665 + }, + { + "epoch": 0.6146198713194765, + "grad_norm": 1.7415549755096436, + "learning_rate": 1.623069142572895e-05, + "loss": 3.2341, + "step": 7666 + }, + { + "epoch": 0.6147000461004991, + "grad_norm": 1.7732791900634766, + "learning_rate": 1.622967577107824e-05, + "loss": 3.2249, + "step": 7667 + }, + { + "epoch": 0.6147802208815217, + "grad_norm": 1.8655468225479126, + "learning_rate": 1.6228660011397394e-05, + "loss": 3.3038, + "step": 7668 + }, + { + "epoch": 0.6148603956625444, + "grad_norm": 1.6682217121124268, + "learning_rate": 1.6227644146703532e-05, + "loss": 3.126, + "step": 7669 + }, + { + "epoch": 0.614940570443567, + "grad_norm": 1.9821282625198364, + "learning_rate": 1.6226628177013787e-05, + "loss": 3.266, + "step": 7670 + }, + { + "epoch": 0.6150207452245896, + "grad_norm": 1.7527556419372559, + "learning_rate": 1.6225612102345282e-05, + "loss": 3.285, + "step": 7671 + }, + { + "epoch": 0.6151009200056122, + "grad_norm": 1.8625948429107666, + "learning_rate": 1.6224595922715155e-05, + "loss": 3.3281, + "step": 7672 + }, + { + "epoch": 0.6151810947866349, + "grad_norm": 1.7805331945419312, + "learning_rate": 1.6223579638140532e-05, + "loss": 3.12, + "step": 7673 + }, + { + "epoch": 0.6152612695676575, + "grad_norm": 2.2030045986175537, + "learning_rate": 1.622256324863855e-05, + "loss": 3.2517, + "step": 7674 + }, + { + "epoch": 0.6153414443486801, + "grad_norm": 1.8778287172317505, + "learning_rate": 1.6221546754226345e-05, + "loss": 3.0975, + "step": 7675 + }, + { + "epoch": 0.6154216191297027, + "grad_norm": 1.8958344459533691, + "learning_rate": 1.6220530154921056e-05, + "loss": 3.2189, + "step": 7676 + }, + { + "epoch": 0.6155017939107253, + "grad_norm": 1.806431770324707, + "learning_rate": 1.621951345073982e-05, + "loss": 3.2189, + "step": 7677 + }, + { + "epoch": 0.615581968691748, + "grad_norm": 1.8405619859695435, + "learning_rate": 1.621849664169978e-05, + "loss": 3.3391, + "step": 7678 + }, + { + "epoch": 0.6156621434727706, + "grad_norm": 1.8690166473388672, + "learning_rate": 1.6217479727818075e-05, + "loss": 3.1273, + "step": 7679 + }, + { + "epoch": 0.6157423182537932, + "grad_norm": 1.906927227973938, + "learning_rate": 1.6216462709111856e-05, + "loss": 3.2014, + "step": 7680 + }, + { + "epoch": 0.6158224930348158, + "grad_norm": 1.7904173135757446, + "learning_rate": 1.6215445585598265e-05, + "loss": 3.2263, + "step": 7681 + }, + { + "epoch": 0.6159026678158386, + "grad_norm": 1.7558826208114624, + "learning_rate": 1.6214428357294454e-05, + "loss": 3.1915, + "step": 7682 + }, + { + "epoch": 0.6159828425968612, + "grad_norm": 1.8778225183486938, + "learning_rate": 1.6213411024217566e-05, + "loss": 3.2291, + "step": 7683 + }, + { + "epoch": 0.6160630173778838, + "grad_norm": 1.8432310819625854, + "learning_rate": 1.6212393586384763e-05, + "loss": 3.1865, + "step": 7684 + }, + { + "epoch": 0.6161431921589064, + "grad_norm": 1.8939836025238037, + "learning_rate": 1.6211376043813196e-05, + "loss": 3.1934, + "step": 7685 + }, + { + "epoch": 0.6162233669399291, + "grad_norm": 1.860787272453308, + "learning_rate": 1.621035839652001e-05, + "loss": 3.2321, + "step": 7686 + }, + { + "epoch": 0.6163035417209517, + "grad_norm": 1.9360034465789795, + "learning_rate": 1.6209340644522376e-05, + "loss": 3.1993, + "step": 7687 + }, + { + "epoch": 0.6163837165019743, + "grad_norm": 2.012697458267212, + "learning_rate": 1.6208322787837445e-05, + "loss": 3.2593, + "step": 7688 + }, + { + "epoch": 0.616463891282997, + "grad_norm": 1.850142478942871, + "learning_rate": 1.620730482648238e-05, + "loss": 3.2024, + "step": 7689 + }, + { + "epoch": 0.6165440660640196, + "grad_norm": 1.8544621467590332, + "learning_rate": 1.6206286760474345e-05, + "loss": 3.2279, + "step": 7690 + }, + { + "epoch": 0.6166242408450422, + "grad_norm": 1.8595852851867676, + "learning_rate": 1.62052685898305e-05, + "loss": 3.1776, + "step": 7691 + }, + { + "epoch": 0.6167044156260648, + "grad_norm": 1.9389960765838623, + "learning_rate": 1.6204250314568016e-05, + "loss": 3.2675, + "step": 7692 + }, + { + "epoch": 0.6167845904070874, + "grad_norm": 1.7419606447219849, + "learning_rate": 1.6203231934704055e-05, + "loss": 3.2243, + "step": 7693 + }, + { + "epoch": 0.6168647651881101, + "grad_norm": 1.8906103372573853, + "learning_rate": 1.6202213450255794e-05, + "loss": 3.2647, + "step": 7694 + }, + { + "epoch": 0.6169449399691327, + "grad_norm": 1.7934377193450928, + "learning_rate": 1.6201194861240397e-05, + "loss": 3.1524, + "step": 7695 + }, + { + "epoch": 0.6170251147501553, + "grad_norm": 1.7545297145843506, + "learning_rate": 1.6200176167675043e-05, + "loss": 3.3362, + "step": 7696 + }, + { + "epoch": 0.6171052895311779, + "grad_norm": 1.8355813026428223, + "learning_rate": 1.61991573695769e-05, + "loss": 3.1433, + "step": 7697 + }, + { + "epoch": 0.6171854643122006, + "grad_norm": 1.8846724033355713, + "learning_rate": 1.6198138466963152e-05, + "loss": 3.255, + "step": 7698 + }, + { + "epoch": 0.6172656390932232, + "grad_norm": 1.830029845237732, + "learning_rate": 1.619711945985097e-05, + "loss": 3.2415, + "step": 7699 + }, + { + "epoch": 0.6173458138742458, + "grad_norm": 1.7089861631393433, + "learning_rate": 1.619610034825754e-05, + "loss": 3.1549, + "step": 7700 + }, + { + "epoch": 0.6174259886552684, + "grad_norm": 1.7859872579574585, + "learning_rate": 1.6195081132200042e-05, + "loss": 3.1885, + "step": 7701 + }, + { + "epoch": 0.6175061634362912, + "grad_norm": 1.7943885326385498, + "learning_rate": 1.6194061811695662e-05, + "loss": 3.2256, + "step": 7702 + }, + { + "epoch": 0.6175863382173138, + "grad_norm": 1.7216655015945435, + "learning_rate": 1.619304238676158e-05, + "loss": 3.173, + "step": 7703 + }, + { + "epoch": 0.6176665129983364, + "grad_norm": 1.858001470565796, + "learning_rate": 1.6192022857414986e-05, + "loss": 3.2576, + "step": 7704 + }, + { + "epoch": 0.617746687779359, + "grad_norm": 1.7665693759918213, + "learning_rate": 1.619100322367307e-05, + "loss": 3.2683, + "step": 7705 + }, + { + "epoch": 0.6178268625603817, + "grad_norm": 0.799502968788147, + "learning_rate": 1.618998348555302e-05, + "loss": 2.4965, + "step": 7706 + }, + { + "epoch": 0.6179070373414043, + "grad_norm": 1.9059908390045166, + "learning_rate": 1.618896364307203e-05, + "loss": 3.1715, + "step": 7707 + }, + { + "epoch": 0.6179872121224269, + "grad_norm": 1.7594841718673706, + "learning_rate": 1.6187943696247295e-05, + "loss": 3.1798, + "step": 7708 + }, + { + "epoch": 0.6180673869034495, + "grad_norm": 1.78549063205719, + "learning_rate": 1.6186923645096012e-05, + "loss": 3.2274, + "step": 7709 + }, + { + "epoch": 0.6181475616844722, + "grad_norm": 1.6692947149276733, + "learning_rate": 1.6185903489635376e-05, + "loss": 3.1943, + "step": 7710 + }, + { + "epoch": 0.6182277364654948, + "grad_norm": 1.8622874021530151, + "learning_rate": 1.6184883229882585e-05, + "loss": 3.1948, + "step": 7711 + }, + { + "epoch": 0.6183079112465174, + "grad_norm": 2.0834245681762695, + "learning_rate": 1.6183862865854845e-05, + "loss": 3.3257, + "step": 7712 + }, + { + "epoch": 0.61838808602754, + "grad_norm": 1.8025615215301514, + "learning_rate": 1.6182842397569356e-05, + "loss": 3.2076, + "step": 7713 + }, + { + "epoch": 0.6184682608085627, + "grad_norm": 1.8491615056991577, + "learning_rate": 1.6181821825043324e-05, + "loss": 3.3073, + "step": 7714 + }, + { + "epoch": 0.6185484355895853, + "grad_norm": 1.8558341264724731, + "learning_rate": 1.6180801148293953e-05, + "loss": 3.2948, + "step": 7715 + }, + { + "epoch": 0.6186286103706079, + "grad_norm": 1.9827044010162354, + "learning_rate": 1.6179780367338452e-05, + "loss": 3.1147, + "step": 7716 + }, + { + "epoch": 0.6187087851516305, + "grad_norm": 0.7855793833732605, + "learning_rate": 1.6178759482194033e-05, + "loss": 2.4917, + "step": 7717 + }, + { + "epoch": 0.6187889599326531, + "grad_norm": 1.9240491390228271, + "learning_rate": 1.6177738492877908e-05, + "loss": 3.2783, + "step": 7718 + }, + { + "epoch": 0.6188691347136758, + "grad_norm": 1.7862284183502197, + "learning_rate": 1.617671739940729e-05, + "loss": 3.326, + "step": 7719 + }, + { + "epoch": 0.6189493094946984, + "grad_norm": 1.7923150062561035, + "learning_rate": 1.617569620179939e-05, + "loss": 3.2472, + "step": 7720 + }, + { + "epoch": 0.619029484275721, + "grad_norm": 1.7755465507507324, + "learning_rate": 1.617467490007143e-05, + "loss": 3.2436, + "step": 7721 + }, + { + "epoch": 0.6191096590567438, + "grad_norm": 1.9480754137039185, + "learning_rate": 1.617365349424063e-05, + "loss": 3.3625, + "step": 7722 + }, + { + "epoch": 0.6191898338377664, + "grad_norm": 1.8771424293518066, + "learning_rate": 1.6172631984324204e-05, + "loss": 3.2411, + "step": 7723 + }, + { + "epoch": 0.619270008618789, + "grad_norm": 1.8382395505905151, + "learning_rate": 1.6171610370339383e-05, + "loss": 3.2263, + "step": 7724 + }, + { + "epoch": 0.6193501833998116, + "grad_norm": 1.759478211402893, + "learning_rate": 1.617058865230338e-05, + "loss": 3.1019, + "step": 7725 + }, + { + "epoch": 0.6194303581808343, + "grad_norm": 1.9810410737991333, + "learning_rate": 1.616956683023343e-05, + "loss": 3.2833, + "step": 7726 + }, + { + "epoch": 0.6195105329618569, + "grad_norm": 1.995888352394104, + "learning_rate": 1.6168544904146756e-05, + "loss": 3.2408, + "step": 7727 + }, + { + "epoch": 0.6195907077428795, + "grad_norm": 1.8328182697296143, + "learning_rate": 1.6167522874060594e-05, + "loss": 3.2638, + "step": 7728 + }, + { + "epoch": 0.6196708825239021, + "grad_norm": 1.7381848096847534, + "learning_rate": 1.6166500739992166e-05, + "loss": 3.1872, + "step": 7729 + }, + { + "epoch": 0.6197510573049247, + "grad_norm": 1.813892126083374, + "learning_rate": 1.616547850195871e-05, + "loss": 3.2155, + "step": 7730 + }, + { + "epoch": 0.6198312320859474, + "grad_norm": 2.704730749130249, + "learning_rate": 1.616445615997746e-05, + "loss": 3.1417, + "step": 7731 + }, + { + "epoch": 0.61991140686697, + "grad_norm": 1.9243764877319336, + "learning_rate": 1.616343371406565e-05, + "loss": 3.2496, + "step": 7732 + }, + { + "epoch": 0.6199915816479926, + "grad_norm": 1.7758058309555054, + "learning_rate": 1.6162411164240518e-05, + "loss": 3.2732, + "step": 7733 + }, + { + "epoch": 0.6200717564290152, + "grad_norm": 1.8555330038070679, + "learning_rate": 1.6161388510519308e-05, + "loss": 3.2516, + "step": 7734 + }, + { + "epoch": 0.6201519312100379, + "grad_norm": 0.7662234902381897, + "learning_rate": 1.6160365752919262e-05, + "loss": 2.5101, + "step": 7735 + }, + { + "epoch": 0.6202321059910605, + "grad_norm": 1.9319725036621094, + "learning_rate": 1.6159342891457617e-05, + "loss": 3.2493, + "step": 7736 + }, + { + "epoch": 0.6203122807720831, + "grad_norm": 1.8014600276947021, + "learning_rate": 1.615831992615162e-05, + "loss": 3.269, + "step": 7737 + }, + { + "epoch": 0.6203924555531057, + "grad_norm": 1.926025390625, + "learning_rate": 1.615729685701852e-05, + "loss": 3.2838, + "step": 7738 + }, + { + "epoch": 0.6204726303341284, + "grad_norm": 1.826198697090149, + "learning_rate": 1.615627368407557e-05, + "loss": 3.2253, + "step": 7739 + }, + { + "epoch": 0.620552805115151, + "grad_norm": 1.8586846590042114, + "learning_rate": 1.6155250407340012e-05, + "loss": 3.1815, + "step": 7740 + }, + { + "epoch": 0.6206329798961736, + "grad_norm": 0.8458767533302307, + "learning_rate": 1.6154227026829104e-05, + "loss": 2.4951, + "step": 7741 + }, + { + "epoch": 0.6207131546771962, + "grad_norm": 1.8630417585372925, + "learning_rate": 1.615320354256009e-05, + "loss": 3.2482, + "step": 7742 + }, + { + "epoch": 0.620793329458219, + "grad_norm": 1.7341742515563965, + "learning_rate": 1.615217995455024e-05, + "loss": 3.1466, + "step": 7743 + }, + { + "epoch": 0.6208735042392416, + "grad_norm": 1.9901912212371826, + "learning_rate": 1.6151156262816802e-05, + "loss": 3.2815, + "step": 7744 + }, + { + "epoch": 0.6209536790202642, + "grad_norm": 1.8597421646118164, + "learning_rate": 1.6150132467377036e-05, + "loss": 3.2919, + "step": 7745 + }, + { + "epoch": 0.6210338538012868, + "grad_norm": 1.8477364778518677, + "learning_rate": 1.6149108568248207e-05, + "loss": 3.2448, + "step": 7746 + }, + { + "epoch": 0.6211140285823095, + "grad_norm": 1.9390352964401245, + "learning_rate": 1.6148084565447572e-05, + "loss": 3.2394, + "step": 7747 + }, + { + "epoch": 0.6211942033633321, + "grad_norm": 0.8241510391235352, + "learning_rate": 1.61470604589924e-05, + "loss": 2.5088, + "step": 7748 + }, + { + "epoch": 0.6212743781443547, + "grad_norm": 1.8331845998764038, + "learning_rate": 1.6146036248899953e-05, + "loss": 3.3622, + "step": 7749 + }, + { + "epoch": 0.6213545529253773, + "grad_norm": 1.8183164596557617, + "learning_rate": 1.6145011935187504e-05, + "loss": 3.2604, + "step": 7750 + }, + { + "epoch": 0.6214347277064, + "grad_norm": 1.816925048828125, + "learning_rate": 1.6143987517872316e-05, + "loss": 3.1165, + "step": 7751 + }, + { + "epoch": 0.6215149024874226, + "grad_norm": 2.0536675453186035, + "learning_rate": 1.6142962996971667e-05, + "loss": 3.2518, + "step": 7752 + }, + { + "epoch": 0.6215950772684452, + "grad_norm": 1.7928744554519653, + "learning_rate": 1.6141938372502826e-05, + "loss": 3.1965, + "step": 7753 + }, + { + "epoch": 0.6216752520494678, + "grad_norm": 1.8768035173416138, + "learning_rate": 1.614091364448307e-05, + "loss": 3.2284, + "step": 7754 + }, + { + "epoch": 0.6217554268304905, + "grad_norm": 1.8346303701400757, + "learning_rate": 1.6139888812929668e-05, + "loss": 3.1895, + "step": 7755 + }, + { + "epoch": 0.6218356016115131, + "grad_norm": 1.729901671409607, + "learning_rate": 1.613886387785991e-05, + "loss": 3.1875, + "step": 7756 + }, + { + "epoch": 0.6219157763925357, + "grad_norm": 1.794780969619751, + "learning_rate": 1.613783883929107e-05, + "loss": 3.1821, + "step": 7757 + }, + { + "epoch": 0.6219959511735583, + "grad_norm": 1.8901740312576294, + "learning_rate": 1.613681369724043e-05, + "loss": 3.1823, + "step": 7758 + }, + { + "epoch": 0.622076125954581, + "grad_norm": 1.8047786951065063, + "learning_rate": 1.6135788451725272e-05, + "loss": 3.1792, + "step": 7759 + }, + { + "epoch": 0.6221563007356036, + "grad_norm": 1.8886393308639526, + "learning_rate": 1.6134763102762886e-05, + "loss": 3.3232, + "step": 7760 + }, + { + "epoch": 0.6222364755166262, + "grad_norm": 0.7262952327728271, + "learning_rate": 1.6133737650370554e-05, + "loss": 2.4894, + "step": 7761 + }, + { + "epoch": 0.6223166502976488, + "grad_norm": 1.9619038105010986, + "learning_rate": 1.613271209456557e-05, + "loss": 3.1759, + "step": 7762 + }, + { + "epoch": 0.6223968250786716, + "grad_norm": 1.826802372932434, + "learning_rate": 1.6131686435365216e-05, + "loss": 3.2579, + "step": 7763 + }, + { + "epoch": 0.6224769998596942, + "grad_norm": 1.7753385305404663, + "learning_rate": 1.6130660672786796e-05, + "loss": 3.3665, + "step": 7764 + }, + { + "epoch": 0.6225571746407168, + "grad_norm": 1.75595223903656, + "learning_rate": 1.6129634806847595e-05, + "loss": 3.2334, + "step": 7765 + }, + { + "epoch": 0.6226373494217394, + "grad_norm": 1.8005825281143188, + "learning_rate": 1.612860883756491e-05, + "loss": 3.1562, + "step": 7766 + }, + { + "epoch": 0.622717524202762, + "grad_norm": 1.8670482635498047, + "learning_rate": 1.612758276495604e-05, + "loss": 3.0858, + "step": 7767 + }, + { + "epoch": 0.6227976989837847, + "grad_norm": 1.8279167413711548, + "learning_rate": 1.6126556589038284e-05, + "loss": 3.2139, + "step": 7768 + }, + { + "epoch": 0.6228778737648073, + "grad_norm": 1.7475323677062988, + "learning_rate": 1.6125530309828947e-05, + "loss": 3.146, + "step": 7769 + }, + { + "epoch": 0.6229580485458299, + "grad_norm": 1.7570288181304932, + "learning_rate": 1.6124503927345324e-05, + "loss": 3.2236, + "step": 7770 + }, + { + "epoch": 0.6230382233268525, + "grad_norm": 1.783515453338623, + "learning_rate": 1.6123477441604726e-05, + "loss": 3.1606, + "step": 7771 + }, + { + "epoch": 0.6231183981078752, + "grad_norm": 1.7719324827194214, + "learning_rate": 1.6122450852624452e-05, + "loss": 3.1786, + "step": 7772 + }, + { + "epoch": 0.6231985728888978, + "grad_norm": 1.7794376611709595, + "learning_rate": 1.6121424160421816e-05, + "loss": 3.2236, + "step": 7773 + }, + { + "epoch": 0.6232787476699204, + "grad_norm": 0.7975783348083496, + "learning_rate": 1.612039736501413e-05, + "loss": 2.5115, + "step": 7774 + }, + { + "epoch": 0.623358922450943, + "grad_norm": 1.744873046875, + "learning_rate": 1.6119370466418697e-05, + "loss": 3.1427, + "step": 7775 + }, + { + "epoch": 0.6234390972319657, + "grad_norm": 1.7824044227600098, + "learning_rate": 1.6118343464652838e-05, + "loss": 3.2361, + "step": 7776 + }, + { + "epoch": 0.6235192720129883, + "grad_norm": 1.8344303369522095, + "learning_rate": 1.6117316359733862e-05, + "loss": 3.274, + "step": 7777 + }, + { + "epoch": 0.6235994467940109, + "grad_norm": 1.7186359167099, + "learning_rate": 1.611628915167909e-05, + "loss": 3.2865, + "step": 7778 + }, + { + "epoch": 0.6236796215750335, + "grad_norm": 1.8268405199050903, + "learning_rate": 1.6115261840505838e-05, + "loss": 3.1926, + "step": 7779 + }, + { + "epoch": 0.6237597963560562, + "grad_norm": 1.841933012008667, + "learning_rate": 1.6114234426231426e-05, + "loss": 3.241, + "step": 7780 + }, + { + "epoch": 0.6238399711370788, + "grad_norm": 1.918966293334961, + "learning_rate": 1.611320690887318e-05, + "loss": 3.3212, + "step": 7781 + }, + { + "epoch": 0.6239201459181014, + "grad_norm": 1.8844798803329468, + "learning_rate": 1.6112179288448414e-05, + "loss": 3.2448, + "step": 7782 + }, + { + "epoch": 0.6240003206991241, + "grad_norm": 1.7506122589111328, + "learning_rate": 1.611115156497446e-05, + "loss": 3.2323, + "step": 7783 + }, + { + "epoch": 0.6240804954801468, + "grad_norm": 1.848402738571167, + "learning_rate": 1.611012373846865e-05, + "loss": 3.2998, + "step": 7784 + }, + { + "epoch": 0.6241606702611694, + "grad_norm": 1.863008975982666, + "learning_rate": 1.61090958089483e-05, + "loss": 3.1061, + "step": 7785 + }, + { + "epoch": 0.624240845042192, + "grad_norm": 1.9336068630218506, + "learning_rate": 1.610806777643075e-05, + "loss": 3.2513, + "step": 7786 + }, + { + "epoch": 0.6243210198232146, + "grad_norm": 1.8272851705551147, + "learning_rate": 1.6107039640933335e-05, + "loss": 3.2998, + "step": 7787 + }, + { + "epoch": 0.6244011946042373, + "grad_norm": 0.7573847770690918, + "learning_rate": 1.6106011402473378e-05, + "loss": 2.4871, + "step": 7788 + }, + { + "epoch": 0.6244813693852599, + "grad_norm": 1.838168978691101, + "learning_rate": 1.6104983061068225e-05, + "loss": 3.2799, + "step": 7789 + }, + { + "epoch": 0.6245615441662825, + "grad_norm": 1.7352269887924194, + "learning_rate": 1.6103954616735203e-05, + "loss": 3.1119, + "step": 7790 + }, + { + "epoch": 0.6246417189473051, + "grad_norm": 1.8965654373168945, + "learning_rate": 1.6102926069491663e-05, + "loss": 3.188, + "step": 7791 + }, + { + "epoch": 0.6247218937283278, + "grad_norm": 1.8861621618270874, + "learning_rate": 1.610189741935494e-05, + "loss": 3.2042, + "step": 7792 + }, + { + "epoch": 0.6248020685093504, + "grad_norm": 0.7386541962623596, + "learning_rate": 1.6100868666342374e-05, + "loss": 2.5143, + "step": 7793 + }, + { + "epoch": 0.624882243290373, + "grad_norm": 1.8107999563217163, + "learning_rate": 1.6099839810471317e-05, + "loss": 3.186, + "step": 7794 + }, + { + "epoch": 0.6249624180713956, + "grad_norm": 1.7725387811660767, + "learning_rate": 1.6098810851759106e-05, + "loss": 3.3019, + "step": 7795 + }, + { + "epoch": 0.6250425928524183, + "grad_norm": 1.8036226034164429, + "learning_rate": 1.6097781790223096e-05, + "loss": 3.2509, + "step": 7796 + }, + { + "epoch": 0.6251227676334409, + "grad_norm": 0.8067800402641296, + "learning_rate": 1.609675262588063e-05, + "loss": 2.4991, + "step": 7797 + }, + { + "epoch": 0.6252029424144635, + "grad_norm": 1.8001283407211304, + "learning_rate": 1.609572335874907e-05, + "loss": 3.2825, + "step": 7798 + }, + { + "epoch": 0.6252831171954861, + "grad_norm": 1.713966965675354, + "learning_rate": 1.609469398884576e-05, + "loss": 3.2065, + "step": 7799 + }, + { + "epoch": 0.6253632919765088, + "grad_norm": 1.8979910612106323, + "learning_rate": 1.6093664516188058e-05, + "loss": 3.1181, + "step": 7800 + }, + { + "epoch": 0.6254434667575314, + "grad_norm": 1.9038587808609009, + "learning_rate": 1.6092634940793317e-05, + "loss": 3.2623, + "step": 7801 + }, + { + "epoch": 0.625523641538554, + "grad_norm": 1.8416576385498047, + "learning_rate": 1.60916052626789e-05, + "loss": 3.318, + "step": 7802 + }, + { + "epoch": 0.6256038163195766, + "grad_norm": 1.7929840087890625, + "learning_rate": 1.6090575481862166e-05, + "loss": 3.2498, + "step": 7803 + }, + { + "epoch": 0.6256839911005994, + "grad_norm": 1.8161765336990356, + "learning_rate": 1.608954559836047e-05, + "loss": 3.213, + "step": 7804 + }, + { + "epoch": 0.625764165881622, + "grad_norm": 1.7308286428451538, + "learning_rate": 1.608851561219119e-05, + "loss": 3.2152, + "step": 7805 + }, + { + "epoch": 0.6258443406626446, + "grad_norm": 1.733273983001709, + "learning_rate": 1.608748552337168e-05, + "loss": 3.1171, + "step": 7806 + }, + { + "epoch": 0.6259245154436672, + "grad_norm": 0.7641175985336304, + "learning_rate": 1.6086455331919306e-05, + "loss": 2.4938, + "step": 7807 + }, + { + "epoch": 0.6260046902246899, + "grad_norm": 1.7485188245773315, + "learning_rate": 1.6085425037851445e-05, + "loss": 3.2464, + "step": 7808 + }, + { + "epoch": 0.6260848650057125, + "grad_norm": 2.0113439559936523, + "learning_rate": 1.608439464118546e-05, + "loss": 3.3224, + "step": 7809 + }, + { + "epoch": 0.6261650397867351, + "grad_norm": 1.7507506608963013, + "learning_rate": 1.6083364141938725e-05, + "loss": 3.2067, + "step": 7810 + }, + { + "epoch": 0.6262452145677577, + "grad_norm": 1.8047572374343872, + "learning_rate": 1.6082333540128615e-05, + "loss": 3.207, + "step": 7811 + }, + { + "epoch": 0.6263253893487803, + "grad_norm": 1.7999311685562134, + "learning_rate": 1.60813028357725e-05, + "loss": 3.2979, + "step": 7812 + }, + { + "epoch": 0.626405564129803, + "grad_norm": 1.8189752101898193, + "learning_rate": 1.6080272028887773e-05, + "loss": 3.1649, + "step": 7813 + }, + { + "epoch": 0.6264857389108256, + "grad_norm": 1.8507561683654785, + "learning_rate": 1.6079241119491797e-05, + "loss": 3.2267, + "step": 7814 + }, + { + "epoch": 0.6265659136918482, + "grad_norm": 1.8920751810073853, + "learning_rate": 1.607821010760196e-05, + "loss": 3.2722, + "step": 7815 + }, + { + "epoch": 0.6266460884728708, + "grad_norm": 1.8278700113296509, + "learning_rate": 1.607717899323564e-05, + "loss": 3.1693, + "step": 7816 + }, + { + "epoch": 0.6267262632538935, + "grad_norm": 1.9649869203567505, + "learning_rate": 1.6076147776410225e-05, + "loss": 3.1177, + "step": 7817 + }, + { + "epoch": 0.6268064380349161, + "grad_norm": 1.7779408693313599, + "learning_rate": 1.60751164571431e-05, + "loss": 3.1597, + "step": 7818 + }, + { + "epoch": 0.6268866128159387, + "grad_norm": 1.7755498886108398, + "learning_rate": 1.6074085035451654e-05, + "loss": 3.2905, + "step": 7819 + }, + { + "epoch": 0.6269667875969613, + "grad_norm": 1.8859336376190186, + "learning_rate": 1.6073053511353274e-05, + "loss": 3.2521, + "step": 7820 + }, + { + "epoch": 0.627046962377984, + "grad_norm": 1.851550817489624, + "learning_rate": 1.607202188486535e-05, + "loss": 3.1967, + "step": 7821 + }, + { + "epoch": 0.6271271371590066, + "grad_norm": 1.8469520807266235, + "learning_rate": 1.607099015600528e-05, + "loss": 3.1622, + "step": 7822 + }, + { + "epoch": 0.6272073119400292, + "grad_norm": 1.8023337125778198, + "learning_rate": 1.6069958324790456e-05, + "loss": 3.3263, + "step": 7823 + }, + { + "epoch": 0.627287486721052, + "grad_norm": 1.8423908948898315, + "learning_rate": 1.606892639123827e-05, + "loss": 3.21, + "step": 7824 + }, + { + "epoch": 0.6273676615020746, + "grad_norm": 1.8612539768218994, + "learning_rate": 1.606789435536613e-05, + "loss": 3.244, + "step": 7825 + }, + { + "epoch": 0.6274478362830972, + "grad_norm": 1.8916282653808594, + "learning_rate": 1.6066862217191424e-05, + "loss": 3.2508, + "step": 7826 + }, + { + "epoch": 0.6275280110641198, + "grad_norm": 1.9517351388931274, + "learning_rate": 1.606582997673156e-05, + "loss": 3.2216, + "step": 7827 + }, + { + "epoch": 0.6276081858451424, + "grad_norm": 1.8228542804718018, + "learning_rate": 1.6064797634003945e-05, + "loss": 3.2291, + "step": 7828 + }, + { + "epoch": 0.6276883606261651, + "grad_norm": 1.735293984413147, + "learning_rate": 1.6063765189025974e-05, + "loss": 3.2551, + "step": 7829 + }, + { + "epoch": 0.6277685354071877, + "grad_norm": 1.6814501285552979, + "learning_rate": 1.606273264181506e-05, + "loss": 3.2443, + "step": 7830 + }, + { + "epoch": 0.6278487101882103, + "grad_norm": 1.8626974821090698, + "learning_rate": 1.6061699992388612e-05, + "loss": 3.2108, + "step": 7831 + }, + { + "epoch": 0.6279288849692329, + "grad_norm": 1.9152839183807373, + "learning_rate": 1.606066724076404e-05, + "loss": 3.3481, + "step": 7832 + }, + { + "epoch": 0.6280090597502556, + "grad_norm": 1.7375894784927368, + "learning_rate": 1.6059634386958753e-05, + "loss": 3.1052, + "step": 7833 + }, + { + "epoch": 0.6280892345312782, + "grad_norm": 1.7555402517318726, + "learning_rate": 1.6058601430990162e-05, + "loss": 3.2756, + "step": 7834 + }, + { + "epoch": 0.6281694093123008, + "grad_norm": 1.741943120956421, + "learning_rate": 1.605756837287569e-05, + "loss": 3.2615, + "step": 7835 + }, + { + "epoch": 0.6282495840933234, + "grad_norm": 2.2147955894470215, + "learning_rate": 1.605653521263275e-05, + "loss": 3.3004, + "step": 7836 + }, + { + "epoch": 0.6283297588743461, + "grad_norm": 1.9223783016204834, + "learning_rate": 1.6055501950278764e-05, + "loss": 3.2612, + "step": 7837 + }, + { + "epoch": 0.6284099336553687, + "grad_norm": 1.914040446281433, + "learning_rate": 1.6054468585831144e-05, + "loss": 3.3688, + "step": 7838 + }, + { + "epoch": 0.6284901084363913, + "grad_norm": 1.7888976335525513, + "learning_rate": 1.6053435119307318e-05, + "loss": 3.1733, + "step": 7839 + }, + { + "epoch": 0.6285702832174139, + "grad_norm": 1.7411504983901978, + "learning_rate": 1.605240155072471e-05, + "loss": 3.1752, + "step": 7840 + }, + { + "epoch": 0.6286504579984366, + "grad_norm": 0.7597728967666626, + "learning_rate": 1.6051367880100744e-05, + "loss": 2.4811, + "step": 7841 + }, + { + "epoch": 0.6287306327794592, + "grad_norm": 1.9151681661605835, + "learning_rate": 1.6050334107452852e-05, + "loss": 3.3333, + "step": 7842 + }, + { + "epoch": 0.6288108075604818, + "grad_norm": 1.9472404718399048, + "learning_rate": 1.6049300232798456e-05, + "loss": 3.2446, + "step": 7843 + }, + { + "epoch": 0.6288909823415045, + "grad_norm": 1.8471354246139526, + "learning_rate": 1.604826625615499e-05, + "loss": 3.1879, + "step": 7844 + }, + { + "epoch": 0.6289711571225272, + "grad_norm": 1.784880518913269, + "learning_rate": 1.6047232177539887e-05, + "loss": 3.2371, + "step": 7845 + }, + { + "epoch": 0.6290513319035498, + "grad_norm": 1.7543774843215942, + "learning_rate": 1.604619799697058e-05, + "loss": 3.2339, + "step": 7846 + }, + { + "epoch": 0.6291315066845724, + "grad_norm": 1.822634220123291, + "learning_rate": 1.6045163714464507e-05, + "loss": 3.1716, + "step": 7847 + }, + { + "epoch": 0.629211681465595, + "grad_norm": 0.7579230666160583, + "learning_rate": 1.6044129330039107e-05, + "loss": 2.4996, + "step": 7848 + }, + { + "epoch": 0.6292918562466177, + "grad_norm": 1.8162448406219482, + "learning_rate": 1.604309484371181e-05, + "loss": 3.2136, + "step": 7849 + }, + { + "epoch": 0.6293720310276403, + "grad_norm": 1.9194128513336182, + "learning_rate": 1.6042060255500073e-05, + "loss": 3.2543, + "step": 7850 + }, + { + "epoch": 0.6294522058086629, + "grad_norm": 1.826322078704834, + "learning_rate": 1.604102556542132e-05, + "loss": 3.1483, + "step": 7851 + }, + { + "epoch": 0.6295323805896855, + "grad_norm": 1.7889223098754883, + "learning_rate": 1.603999077349301e-05, + "loss": 3.197, + "step": 7852 + }, + { + "epoch": 0.6296125553707081, + "grad_norm": 1.7344173192977905, + "learning_rate": 1.6038955879732583e-05, + "loss": 3.2428, + "step": 7853 + }, + { + "epoch": 0.6296927301517308, + "grad_norm": 1.9560904502868652, + "learning_rate": 1.603792088415749e-05, + "loss": 3.2781, + "step": 7854 + }, + { + "epoch": 0.6297729049327534, + "grad_norm": 0.8308697938919067, + "learning_rate": 1.6036885786785175e-05, + "loss": 2.505, + "step": 7855 + }, + { + "epoch": 0.629853079713776, + "grad_norm": 1.836305022239685, + "learning_rate": 1.60358505876331e-05, + "loss": 3.3141, + "step": 7856 + }, + { + "epoch": 0.6299332544947986, + "grad_norm": 1.8938210010528564, + "learning_rate": 1.6034815286718703e-05, + "loss": 3.2325, + "step": 7857 + }, + { + "epoch": 0.6300134292758213, + "grad_norm": 1.7423375844955444, + "learning_rate": 1.6033779884059452e-05, + "loss": 3.153, + "step": 7858 + }, + { + "epoch": 0.6300936040568439, + "grad_norm": 2.116009473800659, + "learning_rate": 1.6032744379672798e-05, + "loss": 3.2925, + "step": 7859 + }, + { + "epoch": 0.6301737788378665, + "grad_norm": 1.9308950901031494, + "learning_rate": 1.60317087735762e-05, + "loss": 3.2722, + "step": 7860 + }, + { + "epoch": 0.6302539536188891, + "grad_norm": 1.8811111450195312, + "learning_rate": 1.6030673065787116e-05, + "loss": 3.2018, + "step": 7861 + }, + { + "epoch": 0.6303341283999118, + "grad_norm": 1.8337769508361816, + "learning_rate": 1.602963725632301e-05, + "loss": 3.2485, + "step": 7862 + }, + { + "epoch": 0.6304143031809344, + "grad_norm": 1.7810704708099365, + "learning_rate": 1.6028601345201346e-05, + "loss": 3.1058, + "step": 7863 + }, + { + "epoch": 0.630494477961957, + "grad_norm": 1.944313883781433, + "learning_rate": 1.6027565332439588e-05, + "loss": 3.231, + "step": 7864 + }, + { + "epoch": 0.6305746527429797, + "grad_norm": 1.9005612134933472, + "learning_rate": 1.6026529218055204e-05, + "loss": 3.3126, + "step": 7865 + }, + { + "epoch": 0.6306548275240024, + "grad_norm": 1.8506711721420288, + "learning_rate": 1.602549300206566e-05, + "loss": 3.3142, + "step": 7866 + }, + { + "epoch": 0.630735002305025, + "grad_norm": 1.8990122079849243, + "learning_rate": 1.6024456684488424e-05, + "loss": 3.2854, + "step": 7867 + }, + { + "epoch": 0.6308151770860476, + "grad_norm": 1.8869963884353638, + "learning_rate": 1.602342026534098e-05, + "loss": 3.2371, + "step": 7868 + }, + { + "epoch": 0.6308953518670702, + "grad_norm": 1.8617815971374512, + "learning_rate": 1.6022383744640787e-05, + "loss": 3.2153, + "step": 7869 + }, + { + "epoch": 0.6309755266480929, + "grad_norm": 1.7485698461532593, + "learning_rate": 1.6021347122405328e-05, + "loss": 3.16, + "step": 7870 + }, + { + "epoch": 0.6310557014291155, + "grad_norm": 1.8522173166275024, + "learning_rate": 1.602031039865208e-05, + "loss": 3.3001, + "step": 7871 + }, + { + "epoch": 0.6311358762101381, + "grad_norm": 1.8549535274505615, + "learning_rate": 1.6019273573398517e-05, + "loss": 3.2108, + "step": 7872 + }, + { + "epoch": 0.6312160509911607, + "grad_norm": 1.7351189851760864, + "learning_rate": 1.6018236646662123e-05, + "loss": 3.1846, + "step": 7873 + }, + { + "epoch": 0.6312962257721834, + "grad_norm": 1.8768858909606934, + "learning_rate": 1.6017199618460385e-05, + "loss": 3.2568, + "step": 7874 + }, + { + "epoch": 0.631376400553206, + "grad_norm": 1.9230021238327026, + "learning_rate": 1.6016162488810778e-05, + "loss": 3.2129, + "step": 7875 + }, + { + "epoch": 0.6314565753342286, + "grad_norm": 1.887070655822754, + "learning_rate": 1.601512525773079e-05, + "loss": 3.2559, + "step": 7876 + }, + { + "epoch": 0.6315367501152512, + "grad_norm": 1.8037022352218628, + "learning_rate": 1.6014087925237914e-05, + "loss": 3.223, + "step": 7877 + }, + { + "epoch": 0.6316169248962739, + "grad_norm": 1.7404848337173462, + "learning_rate": 1.6013050491349638e-05, + "loss": 3.2311, + "step": 7878 + }, + { + "epoch": 0.6316970996772965, + "grad_norm": 1.7956838607788086, + "learning_rate": 1.6012012956083443e-05, + "loss": 3.2607, + "step": 7879 + }, + { + "epoch": 0.6317772744583191, + "grad_norm": 1.8228012323379517, + "learning_rate": 1.6010975319456834e-05, + "loss": 3.172, + "step": 7880 + }, + { + "epoch": 0.6318574492393417, + "grad_norm": 1.8247517347335815, + "learning_rate": 1.6009937581487297e-05, + "loss": 3.197, + "step": 7881 + }, + { + "epoch": 0.6319376240203644, + "grad_norm": 0.791472852230072, + "learning_rate": 1.600889974219233e-05, + "loss": 2.4862, + "step": 7882 + }, + { + "epoch": 0.632017798801387, + "grad_norm": 1.7910414934158325, + "learning_rate": 1.6007861801589433e-05, + "loss": 3.2137, + "step": 7883 + }, + { + "epoch": 0.6320979735824096, + "grad_norm": 1.7802438735961914, + "learning_rate": 1.6006823759696102e-05, + "loss": 3.1047, + "step": 7884 + }, + { + "epoch": 0.6321781483634323, + "grad_norm": 1.844443440437317, + "learning_rate": 1.6005785616529838e-05, + "loss": 3.1963, + "step": 7885 + }, + { + "epoch": 0.632258323144455, + "grad_norm": 1.8149547576904297, + "learning_rate": 1.6004747372108145e-05, + "loss": 3.2604, + "step": 7886 + }, + { + "epoch": 0.6323384979254776, + "grad_norm": 1.8097026348114014, + "learning_rate": 1.600370902644853e-05, + "loss": 3.2252, + "step": 7887 + }, + { + "epoch": 0.6324186727065002, + "grad_norm": 1.8863829374313354, + "learning_rate": 1.6002670579568497e-05, + "loss": 3.2871, + "step": 7888 + }, + { + "epoch": 0.6324988474875228, + "grad_norm": 1.8399144411087036, + "learning_rate": 1.6001632031485553e-05, + "loss": 3.2135, + "step": 7889 + }, + { + "epoch": 0.6325790222685455, + "grad_norm": 1.8181384801864624, + "learning_rate": 1.600059338221721e-05, + "loss": 3.1479, + "step": 7890 + }, + { + "epoch": 0.6326591970495681, + "grad_norm": 1.9147218465805054, + "learning_rate": 1.599955463178097e-05, + "loss": 3.3355, + "step": 7891 + }, + { + "epoch": 0.6327393718305907, + "grad_norm": 1.799965262413025, + "learning_rate": 1.599851578019436e-05, + "loss": 3.1868, + "step": 7892 + }, + { + "epoch": 0.6328195466116133, + "grad_norm": 1.7833858728408813, + "learning_rate": 1.5997476827474887e-05, + "loss": 3.1547, + "step": 7893 + }, + { + "epoch": 0.632899721392636, + "grad_norm": 1.762845754623413, + "learning_rate": 1.599643777364007e-05, + "loss": 3.2341, + "step": 7894 + }, + { + "epoch": 0.6329798961736586, + "grad_norm": 1.7912182807922363, + "learning_rate": 1.599539861870742e-05, + "loss": 3.1133, + "step": 7895 + }, + { + "epoch": 0.6330600709546812, + "grad_norm": 1.853482961654663, + "learning_rate": 1.5994359362694468e-05, + "loss": 3.1936, + "step": 7896 + }, + { + "epoch": 0.6331402457357038, + "grad_norm": 0.736937403678894, + "learning_rate": 1.5993320005618725e-05, + "loss": 2.5116, + "step": 7897 + }, + { + "epoch": 0.6332204205167264, + "grad_norm": 1.7812939882278442, + "learning_rate": 1.599228054749772e-05, + "loss": 3.2058, + "step": 7898 + }, + { + "epoch": 0.6333005952977491, + "grad_norm": 1.8474767208099365, + "learning_rate": 1.599124098834898e-05, + "loss": 3.1934, + "step": 7899 + }, + { + "epoch": 0.6333807700787717, + "grad_norm": 1.7075726985931396, + "learning_rate": 1.5990201328190028e-05, + "loss": 3.2037, + "step": 7900 + }, + { + "epoch": 0.6334609448597943, + "grad_norm": 1.9096121788024902, + "learning_rate": 1.598916156703839e-05, + "loss": 3.1729, + "step": 7901 + }, + { + "epoch": 0.6335411196408169, + "grad_norm": 0.7229123711585999, + "learning_rate": 1.59881217049116e-05, + "loss": 2.4952, + "step": 7902 + }, + { + "epoch": 0.6336212944218396, + "grad_norm": 1.8031229972839355, + "learning_rate": 1.598708174182719e-05, + "loss": 3.2122, + "step": 7903 + }, + { + "epoch": 0.6337014692028622, + "grad_norm": 1.7713429927825928, + "learning_rate": 1.5986041677802688e-05, + "loss": 3.2102, + "step": 7904 + }, + { + "epoch": 0.6337816439838849, + "grad_norm": 1.8171123266220093, + "learning_rate": 1.5985001512855637e-05, + "loss": 3.1694, + "step": 7905 + }, + { + "epoch": 0.6338618187649075, + "grad_norm": 1.6872180700302124, + "learning_rate": 1.5983961247003567e-05, + "loss": 3.2028, + "step": 7906 + }, + { + "epoch": 0.6339419935459302, + "grad_norm": 2.0216541290283203, + "learning_rate": 1.5982920880264022e-05, + "loss": 3.2094, + "step": 7907 + }, + { + "epoch": 0.6340221683269528, + "grad_norm": 1.863769292831421, + "learning_rate": 1.598188041265454e-05, + "loss": 3.1718, + "step": 7908 + }, + { + "epoch": 0.6341023431079754, + "grad_norm": 1.8307377099990845, + "learning_rate": 1.598083984419266e-05, + "loss": 3.2355, + "step": 7909 + }, + { + "epoch": 0.634182517888998, + "grad_norm": 2.1401407718658447, + "learning_rate": 1.597979917489593e-05, + "loss": 3.3572, + "step": 7910 + }, + { + "epoch": 0.6342626926700207, + "grad_norm": 1.7421821355819702, + "learning_rate": 1.597875840478189e-05, + "loss": 3.298, + "step": 7911 + }, + { + "epoch": 0.6343428674510433, + "grad_norm": 2.020530939102173, + "learning_rate": 1.5977717533868093e-05, + "loss": 3.1417, + "step": 7912 + }, + { + "epoch": 0.6344230422320659, + "grad_norm": 1.8711507320404053, + "learning_rate": 1.5976676562172084e-05, + "loss": 3.1776, + "step": 7913 + }, + { + "epoch": 0.6345032170130885, + "grad_norm": 1.8180173635482788, + "learning_rate": 1.5975635489711416e-05, + "loss": 3.3314, + "step": 7914 + }, + { + "epoch": 0.6345833917941112, + "grad_norm": 1.6679250001907349, + "learning_rate": 1.5974594316503637e-05, + "loss": 3.266, + "step": 7915 + }, + { + "epoch": 0.6346635665751338, + "grad_norm": 1.7636810541152954, + "learning_rate": 1.5973553042566306e-05, + "loss": 3.1246, + "step": 7916 + }, + { + "epoch": 0.6347437413561564, + "grad_norm": 1.882491111755371, + "learning_rate": 1.5972511667916975e-05, + "loss": 3.1678, + "step": 7917 + }, + { + "epoch": 0.634823916137179, + "grad_norm": 2.0840227603912354, + "learning_rate": 1.5971470192573204e-05, + "loss": 3.233, + "step": 7918 + }, + { + "epoch": 0.6349040909182017, + "grad_norm": 1.835152268409729, + "learning_rate": 1.597042861655255e-05, + "loss": 3.1604, + "step": 7919 + }, + { + "epoch": 0.6349842656992243, + "grad_norm": 1.8300204277038574, + "learning_rate": 1.5969386939872568e-05, + "loss": 3.2484, + "step": 7920 + }, + { + "epoch": 0.6350644404802469, + "grad_norm": 1.8106340169906616, + "learning_rate": 1.596834516255083e-05, + "loss": 3.2968, + "step": 7921 + }, + { + "epoch": 0.6351446152612695, + "grad_norm": 1.783894419670105, + "learning_rate": 1.5967303284604897e-05, + "loss": 3.2707, + "step": 7922 + }, + { + "epoch": 0.6352247900422922, + "grad_norm": 1.7760660648345947, + "learning_rate": 1.5966261306052336e-05, + "loss": 3.3066, + "step": 7923 + }, + { + "epoch": 0.6353049648233148, + "grad_norm": 1.8062745332717896, + "learning_rate": 1.5965219226910707e-05, + "loss": 3.2717, + "step": 7924 + }, + { + "epoch": 0.6353851396043374, + "grad_norm": 0.8264740109443665, + "learning_rate": 1.5964177047197586e-05, + "loss": 2.51, + "step": 7925 + }, + { + "epoch": 0.6354653143853601, + "grad_norm": 1.7179820537567139, + "learning_rate": 1.5963134766930545e-05, + "loss": 3.196, + "step": 7926 + }, + { + "epoch": 0.6355454891663828, + "grad_norm": 1.8253682851791382, + "learning_rate": 1.596209238612715e-05, + "loss": 3.1364, + "step": 7927 + }, + { + "epoch": 0.6356256639474054, + "grad_norm": 1.8328944444656372, + "learning_rate": 1.5961049904804978e-05, + "loss": 3.2661, + "step": 7928 + }, + { + "epoch": 0.635705838728428, + "grad_norm": 1.7051968574523926, + "learning_rate": 1.5960007322981607e-05, + "loss": 3.2229, + "step": 7929 + }, + { + "epoch": 0.6357860135094506, + "grad_norm": 1.7420791387557983, + "learning_rate": 1.5958964640674612e-05, + "loss": 3.2256, + "step": 7930 + }, + { + "epoch": 0.6358661882904733, + "grad_norm": 1.8594348430633545, + "learning_rate": 1.5957921857901575e-05, + "loss": 3.1697, + "step": 7931 + }, + { + "epoch": 0.6359463630714959, + "grad_norm": 1.8020628690719604, + "learning_rate": 1.5956878974680073e-05, + "loss": 3.1636, + "step": 7932 + }, + { + "epoch": 0.6360265378525185, + "grad_norm": 1.7892324924468994, + "learning_rate": 1.5955835991027694e-05, + "loss": 3.2332, + "step": 7933 + }, + { + "epoch": 0.6361067126335411, + "grad_norm": 0.7603084444999695, + "learning_rate": 1.595479290696202e-05, + "loss": 2.4802, + "step": 7934 + }, + { + "epoch": 0.6361868874145638, + "grad_norm": 1.712235927581787, + "learning_rate": 1.5953749722500632e-05, + "loss": 3.1575, + "step": 7935 + }, + { + "epoch": 0.6362670621955864, + "grad_norm": 0.7415627241134644, + "learning_rate": 1.5952706437661125e-05, + "loss": 2.5034, + "step": 7936 + }, + { + "epoch": 0.636347236976609, + "grad_norm": 1.6765427589416504, + "learning_rate": 1.5951663052461084e-05, + "loss": 3.1392, + "step": 7937 + }, + { + "epoch": 0.6364274117576316, + "grad_norm": 1.8950397968292236, + "learning_rate": 1.5950619566918103e-05, + "loss": 3.215, + "step": 7938 + }, + { + "epoch": 0.6365075865386542, + "grad_norm": 1.8978785276412964, + "learning_rate": 1.594957598104977e-05, + "loss": 3.3198, + "step": 7939 + }, + { + "epoch": 0.6365877613196769, + "grad_norm": 1.784791111946106, + "learning_rate": 1.5948532294873683e-05, + "loss": 3.3259, + "step": 7940 + }, + { + "epoch": 0.6366679361006995, + "grad_norm": 1.8698546886444092, + "learning_rate": 1.594748850840744e-05, + "loss": 3.2786, + "step": 7941 + }, + { + "epoch": 0.6367481108817221, + "grad_norm": 1.7862285375595093, + "learning_rate": 1.5946444621668638e-05, + "loss": 3.3029, + "step": 7942 + }, + { + "epoch": 0.6368282856627447, + "grad_norm": 1.77959406375885, + "learning_rate": 1.5945400634674874e-05, + "loss": 3.1791, + "step": 7943 + }, + { + "epoch": 0.6369084604437674, + "grad_norm": 1.8359613418579102, + "learning_rate": 1.5944356547443752e-05, + "loss": 3.3032, + "step": 7944 + }, + { + "epoch": 0.63698863522479, + "grad_norm": 1.8613409996032715, + "learning_rate": 1.594331235999287e-05, + "loss": 3.1514, + "step": 7945 + }, + { + "epoch": 0.6370688100058127, + "grad_norm": 1.707476258277893, + "learning_rate": 1.594226807233984e-05, + "loss": 3.1755, + "step": 7946 + }, + { + "epoch": 0.6371489847868353, + "grad_norm": 2.0366218090057373, + "learning_rate": 1.594122368450226e-05, + "loss": 3.2372, + "step": 7947 + }, + { + "epoch": 0.637229159567858, + "grad_norm": 1.8924455642700195, + "learning_rate": 1.594017919649775e-05, + "loss": 3.2463, + "step": 7948 + }, + { + "epoch": 0.6373093343488806, + "grad_norm": 1.6891440153121948, + "learning_rate": 1.5939134608343905e-05, + "loss": 3.1588, + "step": 7949 + }, + { + "epoch": 0.6373895091299032, + "grad_norm": 1.8103573322296143, + "learning_rate": 1.5938089920058347e-05, + "loss": 3.1204, + "step": 7950 + }, + { + "epoch": 0.6374696839109258, + "grad_norm": 1.8507554531097412, + "learning_rate": 1.5937045131658685e-05, + "loss": 3.3365, + "step": 7951 + }, + { + "epoch": 0.6375498586919485, + "grad_norm": 1.9816501140594482, + "learning_rate": 1.5936000243162533e-05, + "loss": 3.1741, + "step": 7952 + }, + { + "epoch": 0.6376300334729711, + "grad_norm": 1.8242896795272827, + "learning_rate": 1.5934955254587515e-05, + "loss": 3.1176, + "step": 7953 + }, + { + "epoch": 0.6377102082539937, + "grad_norm": 1.6709434986114502, + "learning_rate": 1.5933910165951238e-05, + "loss": 3.1438, + "step": 7954 + }, + { + "epoch": 0.6377903830350163, + "grad_norm": 1.8210376501083374, + "learning_rate": 1.5932864977271327e-05, + "loss": 3.1661, + "step": 7955 + }, + { + "epoch": 0.637870557816039, + "grad_norm": 1.8200712203979492, + "learning_rate": 1.5931819688565403e-05, + "loss": 3.1917, + "step": 7956 + }, + { + "epoch": 0.6379507325970616, + "grad_norm": 1.8581517934799194, + "learning_rate": 1.5930774299851092e-05, + "loss": 3.2478, + "step": 7957 + }, + { + "epoch": 0.6380309073780842, + "grad_norm": 1.8574750423431396, + "learning_rate": 1.5929728811146015e-05, + "loss": 3.2385, + "step": 7958 + }, + { + "epoch": 0.6381110821591068, + "grad_norm": 1.9022489786148071, + "learning_rate": 1.5928683222467802e-05, + "loss": 3.2617, + "step": 7959 + }, + { + "epoch": 0.6381912569401295, + "grad_norm": 1.909616470336914, + "learning_rate": 1.5927637533834076e-05, + "loss": 3.2552, + "step": 7960 + }, + { + "epoch": 0.6382714317211521, + "grad_norm": 1.8673791885375977, + "learning_rate": 1.592659174526247e-05, + "loss": 3.1751, + "step": 7961 + }, + { + "epoch": 0.6383516065021747, + "grad_norm": 1.8762239217758179, + "learning_rate": 1.5925545856770623e-05, + "loss": 3.1976, + "step": 7962 + }, + { + "epoch": 0.6384317812831973, + "grad_norm": 1.7670670747756958, + "learning_rate": 1.5924499868376155e-05, + "loss": 3.3231, + "step": 7963 + }, + { + "epoch": 0.63851195606422, + "grad_norm": 1.8085863590240479, + "learning_rate": 1.5923453780096706e-05, + "loss": 3.2679, + "step": 7964 + }, + { + "epoch": 0.6385921308452426, + "grad_norm": 1.7683900594711304, + "learning_rate": 1.5922407591949918e-05, + "loss": 3.2384, + "step": 7965 + }, + { + "epoch": 0.6386723056262653, + "grad_norm": 1.7600502967834473, + "learning_rate": 1.5921361303953422e-05, + "loss": 3.1188, + "step": 7966 + }, + { + "epoch": 0.6387524804072879, + "grad_norm": 1.780377984046936, + "learning_rate": 1.5920314916124865e-05, + "loss": 3.1902, + "step": 7967 + }, + { + "epoch": 0.6388326551883106, + "grad_norm": 1.8314114809036255, + "learning_rate": 1.5919268428481883e-05, + "loss": 3.1686, + "step": 7968 + }, + { + "epoch": 0.6389128299693332, + "grad_norm": 1.8483755588531494, + "learning_rate": 1.591822184104212e-05, + "loss": 3.1936, + "step": 7969 + }, + { + "epoch": 0.6389930047503558, + "grad_norm": 1.879501223564148, + "learning_rate": 1.5917175153823222e-05, + "loss": 3.2893, + "step": 7970 + }, + { + "epoch": 0.6390731795313784, + "grad_norm": 1.796298623085022, + "learning_rate": 1.591612836684284e-05, + "loss": 3.2236, + "step": 7971 + }, + { + "epoch": 0.6391533543124011, + "grad_norm": 1.9140945672988892, + "learning_rate": 1.5915081480118618e-05, + "loss": 3.097, + "step": 7972 + }, + { + "epoch": 0.6392335290934237, + "grad_norm": 1.8150279521942139, + "learning_rate": 1.5914034493668203e-05, + "loss": 3.2646, + "step": 7973 + }, + { + "epoch": 0.6393137038744463, + "grad_norm": 1.7455848455429077, + "learning_rate": 1.5912987407509258e-05, + "loss": 3.2901, + "step": 7974 + }, + { + "epoch": 0.6393938786554689, + "grad_norm": 1.8332366943359375, + "learning_rate": 1.5911940221659424e-05, + "loss": 3.2332, + "step": 7975 + }, + { + "epoch": 0.6394740534364916, + "grad_norm": 1.845507264137268, + "learning_rate": 1.591089293613636e-05, + "loss": 3.1869, + "step": 7976 + }, + { + "epoch": 0.6395542282175142, + "grad_norm": 1.8128705024719238, + "learning_rate": 1.5909845550957728e-05, + "loss": 3.2248, + "step": 7977 + }, + { + "epoch": 0.6396344029985368, + "grad_norm": 1.6530077457427979, + "learning_rate": 1.590879806614118e-05, + "loss": 3.1443, + "step": 7978 + }, + { + "epoch": 0.6397145777795594, + "grad_norm": 1.8009581565856934, + "learning_rate": 1.590775048170438e-05, + "loss": 3.1893, + "step": 7979 + }, + { + "epoch": 0.639794752560582, + "grad_norm": 1.7694511413574219, + "learning_rate": 1.590670279766499e-05, + "loss": 3.1615, + "step": 7980 + }, + { + "epoch": 0.6398749273416047, + "grad_norm": 1.779989242553711, + "learning_rate": 1.590565501404067e-05, + "loss": 3.2347, + "step": 7981 + }, + { + "epoch": 0.6399551021226273, + "grad_norm": 0.7950932383537292, + "learning_rate": 1.5904607130849093e-05, + "loss": 2.4974, + "step": 7982 + }, + { + "epoch": 0.6400352769036499, + "grad_norm": 1.8406586647033691, + "learning_rate": 1.5903559148107918e-05, + "loss": 3.2385, + "step": 7983 + }, + { + "epoch": 0.6401154516846725, + "grad_norm": 1.9125889539718628, + "learning_rate": 1.5902511065834815e-05, + "loss": 3.2344, + "step": 7984 + }, + { + "epoch": 0.6401956264656952, + "grad_norm": 0.7570878267288208, + "learning_rate": 1.590146288404746e-05, + "loss": 2.5137, + "step": 7985 + }, + { + "epoch": 0.6402758012467178, + "grad_norm": 1.9201810359954834, + "learning_rate": 1.5900414602763515e-05, + "loss": 3.2749, + "step": 7986 + }, + { + "epoch": 0.6403559760277405, + "grad_norm": 1.751546859741211, + "learning_rate": 1.589936622200066e-05, + "loss": 3.1542, + "step": 7987 + }, + { + "epoch": 0.6404361508087631, + "grad_norm": 1.7485769987106323, + "learning_rate": 1.5898317741776574e-05, + "loss": 3.1862, + "step": 7988 + }, + { + "epoch": 0.6405163255897858, + "grad_norm": 1.787948489189148, + "learning_rate": 1.5897269162108926e-05, + "loss": 3.1488, + "step": 7989 + }, + { + "epoch": 0.6405965003708084, + "grad_norm": 1.8337455987930298, + "learning_rate": 1.58962204830154e-05, + "loss": 3.2009, + "step": 7990 + }, + { + "epoch": 0.640676675151831, + "grad_norm": 1.839849829673767, + "learning_rate": 1.589517170451367e-05, + "loss": 3.2179, + "step": 7991 + }, + { + "epoch": 0.6407568499328536, + "grad_norm": 1.93160879611969, + "learning_rate": 1.589412282662143e-05, + "loss": 3.304, + "step": 7992 + }, + { + "epoch": 0.6408370247138763, + "grad_norm": 1.9246855974197388, + "learning_rate": 1.5893073849356353e-05, + "loss": 3.1213, + "step": 7993 + }, + { + "epoch": 0.6409171994948989, + "grad_norm": 0.7655735611915588, + "learning_rate": 1.589202477273613e-05, + "loss": 2.4945, + "step": 7994 + }, + { + "epoch": 0.6409973742759215, + "grad_norm": 1.7732621431350708, + "learning_rate": 1.589097559677844e-05, + "loss": 3.1882, + "step": 7995 + }, + { + "epoch": 0.6410775490569441, + "grad_norm": 1.935181736946106, + "learning_rate": 1.5889926321500984e-05, + "loss": 3.1963, + "step": 7996 + }, + { + "epoch": 0.6411577238379668, + "grad_norm": 1.824042797088623, + "learning_rate": 1.5888876946921443e-05, + "loss": 3.2372, + "step": 7997 + }, + { + "epoch": 0.6412378986189894, + "grad_norm": 1.9532585144042969, + "learning_rate": 1.5887827473057514e-05, + "loss": 3.1282, + "step": 7998 + }, + { + "epoch": 0.641318073400012, + "grad_norm": 1.8029789924621582, + "learning_rate": 1.5886777899926887e-05, + "loss": 3.2382, + "step": 7999 + }, + { + "epoch": 0.6413982481810346, + "grad_norm": 1.8060437440872192, + "learning_rate": 1.588572822754726e-05, + "loss": 3.1951, + "step": 8000 + }, + { + "epoch": 0.6414784229620573, + "grad_norm": 1.8334084749221802, + "learning_rate": 1.5884678455936328e-05, + "loss": 3.1658, + "step": 8001 + }, + { + "epoch": 0.6415585977430799, + "grad_norm": 1.786691427230835, + "learning_rate": 1.588362858511179e-05, + "loss": 3.2205, + "step": 8002 + }, + { + "epoch": 0.6416387725241025, + "grad_norm": 1.8210340738296509, + "learning_rate": 1.5882578615091354e-05, + "loss": 3.2134, + "step": 8003 + }, + { + "epoch": 0.6417189473051251, + "grad_norm": 1.7981988191604614, + "learning_rate": 1.588152854589271e-05, + "loss": 3.2197, + "step": 8004 + }, + { + "epoch": 0.6417991220861478, + "grad_norm": 1.7805860042572021, + "learning_rate": 1.588047837753357e-05, + "loss": 3.2241, + "step": 8005 + }, + { + "epoch": 0.6418792968671704, + "grad_norm": 1.8989471197128296, + "learning_rate": 1.5879428110031635e-05, + "loss": 3.276, + "step": 8006 + }, + { + "epoch": 0.6419594716481931, + "grad_norm": 1.963097333908081, + "learning_rate": 1.5878377743404616e-05, + "loss": 3.1958, + "step": 8007 + }, + { + "epoch": 0.6420396464292157, + "grad_norm": 0.8176882863044739, + "learning_rate": 1.587732727767022e-05, + "loss": 2.5023, + "step": 8008 + }, + { + "epoch": 0.6421198212102384, + "grad_norm": 1.9036405086517334, + "learning_rate": 1.5876276712846157e-05, + "loss": 3.2986, + "step": 8009 + }, + { + "epoch": 0.642199995991261, + "grad_norm": 1.8661949634552002, + "learning_rate": 1.587522604895014e-05, + "loss": 3.3464, + "step": 8010 + }, + { + "epoch": 0.6422801707722836, + "grad_norm": 1.7522214651107788, + "learning_rate": 1.587417528599988e-05, + "loss": 3.0922, + "step": 8011 + }, + { + "epoch": 0.6423603455533062, + "grad_norm": 1.7317097187042236, + "learning_rate": 1.5873124424013097e-05, + "loss": 3.2407, + "step": 8012 + }, + { + "epoch": 0.6424405203343289, + "grad_norm": 1.9004970788955688, + "learning_rate": 1.5872073463007507e-05, + "loss": 3.2559, + "step": 8013 + }, + { + "epoch": 0.6425206951153515, + "grad_norm": 1.8506364822387695, + "learning_rate": 1.587102240300083e-05, + "loss": 3.1901, + "step": 8014 + }, + { + "epoch": 0.6426008698963741, + "grad_norm": 1.8269405364990234, + "learning_rate": 1.5869971244010778e-05, + "loss": 3.1889, + "step": 8015 + }, + { + "epoch": 0.6426810446773967, + "grad_norm": 1.7675933837890625, + "learning_rate": 1.5868919986055084e-05, + "loss": 3.2086, + "step": 8016 + }, + { + "epoch": 0.6427612194584194, + "grad_norm": 1.8663299083709717, + "learning_rate": 1.5867868629151465e-05, + "loss": 3.1436, + "step": 8017 + }, + { + "epoch": 0.642841394239442, + "grad_norm": 1.852636694908142, + "learning_rate": 1.586681717331765e-05, + "loss": 3.1582, + "step": 8018 + }, + { + "epoch": 0.6429215690204646, + "grad_norm": 1.8314112424850464, + "learning_rate": 1.5865765618571367e-05, + "loss": 3.1404, + "step": 8019 + }, + { + "epoch": 0.6430017438014872, + "grad_norm": 1.9287805557250977, + "learning_rate": 1.586471396493034e-05, + "loss": 3.2634, + "step": 8020 + }, + { + "epoch": 0.6430819185825098, + "grad_norm": 1.642054796218872, + "learning_rate": 1.5863662212412307e-05, + "loss": 3.124, + "step": 8021 + }, + { + "epoch": 0.6431620933635325, + "grad_norm": 0.7719558477401733, + "learning_rate": 1.5862610361034994e-05, + "loss": 2.4855, + "step": 8022 + }, + { + "epoch": 0.6432422681445551, + "grad_norm": 0.7656789422035217, + "learning_rate": 1.5861558410816132e-05, + "loss": 2.4984, + "step": 8023 + }, + { + "epoch": 0.6433224429255777, + "grad_norm": 1.8160796165466309, + "learning_rate": 1.5860506361773465e-05, + "loss": 3.2336, + "step": 8024 + }, + { + "epoch": 0.6434026177066003, + "grad_norm": 2.0008656978607178, + "learning_rate": 1.5859454213924727e-05, + "loss": 3.1821, + "step": 8025 + }, + { + "epoch": 0.643482792487623, + "grad_norm": 1.8431103229522705, + "learning_rate": 1.5858401967287655e-05, + "loss": 3.2619, + "step": 8026 + }, + { + "epoch": 0.6435629672686457, + "grad_norm": 1.9730502367019653, + "learning_rate": 1.5857349621879988e-05, + "loss": 3.1809, + "step": 8027 + }, + { + "epoch": 0.6436431420496683, + "grad_norm": 1.720018744468689, + "learning_rate": 1.5856297177719474e-05, + "loss": 3.1559, + "step": 8028 + }, + { + "epoch": 0.643723316830691, + "grad_norm": 1.7474602460861206, + "learning_rate": 1.585524463482385e-05, + "loss": 3.161, + "step": 8029 + }, + { + "epoch": 0.6438034916117136, + "grad_norm": 1.8768503665924072, + "learning_rate": 1.5854191993210868e-05, + "loss": 3.1055, + "step": 8030 + }, + { + "epoch": 0.6438836663927362, + "grad_norm": 1.7083232402801514, + "learning_rate": 1.5853139252898274e-05, + "loss": 3.2069, + "step": 8031 + }, + { + "epoch": 0.6439638411737588, + "grad_norm": 1.800879716873169, + "learning_rate": 1.585208641390381e-05, + "loss": 3.2206, + "step": 8032 + }, + { + "epoch": 0.6440440159547814, + "grad_norm": 1.9317699670791626, + "learning_rate": 1.585103347624524e-05, + "loss": 3.2552, + "step": 8033 + }, + { + "epoch": 0.6441241907358041, + "grad_norm": 1.9086920022964478, + "learning_rate": 1.5849980439940294e-05, + "loss": 3.2108, + "step": 8034 + }, + { + "epoch": 0.6442043655168267, + "grad_norm": 1.8491936922073364, + "learning_rate": 1.584892730500675e-05, + "loss": 3.1511, + "step": 8035 + }, + { + "epoch": 0.6442845402978493, + "grad_norm": 1.8412114381790161, + "learning_rate": 1.584787407146235e-05, + "loss": 3.2444, + "step": 8036 + }, + { + "epoch": 0.6443647150788719, + "grad_norm": 1.7627614736557007, + "learning_rate": 1.5846820739324854e-05, + "loss": 3.2341, + "step": 8037 + }, + { + "epoch": 0.6444448898598946, + "grad_norm": 1.8337677717208862, + "learning_rate": 1.5845767308612022e-05, + "loss": 3.1857, + "step": 8038 + }, + { + "epoch": 0.6445250646409172, + "grad_norm": 1.816846489906311, + "learning_rate": 1.584471377934161e-05, + "loss": 3.2491, + "step": 8039 + }, + { + "epoch": 0.6446052394219398, + "grad_norm": 1.7307108640670776, + "learning_rate": 1.584366015153139e-05, + "loss": 3.0943, + "step": 8040 + }, + { + "epoch": 0.6446854142029624, + "grad_norm": 1.8763874769210815, + "learning_rate": 1.5842606425199118e-05, + "loss": 3.2563, + "step": 8041 + }, + { + "epoch": 0.6447655889839851, + "grad_norm": 1.8248950242996216, + "learning_rate": 1.5841552600362556e-05, + "loss": 3.214, + "step": 8042 + }, + { + "epoch": 0.6448457637650077, + "grad_norm": 1.7996444702148438, + "learning_rate": 1.584049867703948e-05, + "loss": 3.2849, + "step": 8043 + }, + { + "epoch": 0.6449259385460303, + "grad_norm": 1.913309931755066, + "learning_rate": 1.583944465524765e-05, + "loss": 3.1831, + "step": 8044 + }, + { + "epoch": 0.6450061133270529, + "grad_norm": 1.7438664436340332, + "learning_rate": 1.5838390535004844e-05, + "loss": 3.1416, + "step": 8045 + }, + { + "epoch": 0.6450862881080756, + "grad_norm": 1.8982123136520386, + "learning_rate": 1.5837336316328832e-05, + "loss": 3.2024, + "step": 8046 + }, + { + "epoch": 0.6451664628890982, + "grad_norm": 1.656898021697998, + "learning_rate": 1.5836281999237386e-05, + "loss": 3.1886, + "step": 8047 + }, + { + "epoch": 0.6452466376701209, + "grad_norm": 1.6659754514694214, + "learning_rate": 1.5835227583748284e-05, + "loss": 3.2196, + "step": 8048 + }, + { + "epoch": 0.6453268124511435, + "grad_norm": 1.7491902112960815, + "learning_rate": 1.58341730698793e-05, + "loss": 3.2261, + "step": 8049 + }, + { + "epoch": 0.6454069872321662, + "grad_norm": 1.7399789094924927, + "learning_rate": 1.5833118457648217e-05, + "loss": 3.1952, + "step": 8050 + }, + { + "epoch": 0.6454871620131888, + "grad_norm": 1.7744264602661133, + "learning_rate": 1.5832063747072806e-05, + "loss": 3.1797, + "step": 8051 + }, + { + "epoch": 0.6455673367942114, + "grad_norm": 2.0085606575012207, + "learning_rate": 1.5831008938170863e-05, + "loss": 3.3098, + "step": 8052 + }, + { + "epoch": 0.645647511575234, + "grad_norm": 1.8111424446105957, + "learning_rate": 1.582995403096016e-05, + "loss": 3.287, + "step": 8053 + }, + { + "epoch": 0.6457276863562567, + "grad_norm": 1.8347924947738647, + "learning_rate": 1.5828899025458487e-05, + "loss": 3.2153, + "step": 8054 + }, + { + "epoch": 0.6458078611372793, + "grad_norm": 1.8361074924468994, + "learning_rate": 1.5827843921683633e-05, + "loss": 3.1067, + "step": 8055 + }, + { + "epoch": 0.6458880359183019, + "grad_norm": 2.5434982776641846, + "learning_rate": 1.5826788719653384e-05, + "loss": 3.2561, + "step": 8056 + }, + { + "epoch": 0.6459682106993245, + "grad_norm": 1.7037842273712158, + "learning_rate": 1.5825733419385528e-05, + "loss": 3.3083, + "step": 8057 + }, + { + "epoch": 0.6460483854803472, + "grad_norm": 0.789949893951416, + "learning_rate": 1.5824678020897862e-05, + "loss": 2.4941, + "step": 8058 + }, + { + "epoch": 0.6461285602613698, + "grad_norm": 1.8093022108078003, + "learning_rate": 1.5823622524208175e-05, + "loss": 3.1869, + "step": 8059 + }, + { + "epoch": 0.6462087350423924, + "grad_norm": 0.772899866104126, + "learning_rate": 1.582256692933427e-05, + "loss": 2.485, + "step": 8060 + }, + { + "epoch": 0.646288909823415, + "grad_norm": 0.7400366067886353, + "learning_rate": 1.5821511236293933e-05, + "loss": 2.4818, + "step": 8061 + }, + { + "epoch": 0.6463690846044376, + "grad_norm": 1.7812678813934326, + "learning_rate": 1.582045544510497e-05, + "loss": 3.1484, + "step": 8062 + }, + { + "epoch": 0.6464492593854603, + "grad_norm": 0.7114151120185852, + "learning_rate": 1.5819399555785178e-05, + "loss": 2.4736, + "step": 8063 + }, + { + "epoch": 0.6465294341664829, + "grad_norm": 1.8795839548110962, + "learning_rate": 1.5818343568352363e-05, + "loss": 3.2146, + "step": 8064 + }, + { + "epoch": 0.6466096089475055, + "grad_norm": 1.9802212715148926, + "learning_rate": 1.5817287482824328e-05, + "loss": 3.275, + "step": 8065 + }, + { + "epoch": 0.6466897837285281, + "grad_norm": 1.8436840772628784, + "learning_rate": 1.5816231299218872e-05, + "loss": 3.1963, + "step": 8066 + }, + { + "epoch": 0.6467699585095508, + "grad_norm": 1.7912681102752686, + "learning_rate": 1.5815175017553808e-05, + "loss": 3.2813, + "step": 8067 + }, + { + "epoch": 0.6468501332905735, + "grad_norm": 1.8031774759292603, + "learning_rate": 1.5814118637846944e-05, + "loss": 3.1836, + "step": 8068 + }, + { + "epoch": 0.6469303080715961, + "grad_norm": 1.8898437023162842, + "learning_rate": 1.5813062160116087e-05, + "loss": 3.2283, + "step": 8069 + }, + { + "epoch": 0.6470104828526188, + "grad_norm": 1.8199193477630615, + "learning_rate": 1.581200558437905e-05, + "loss": 3.1832, + "step": 8070 + }, + { + "epoch": 0.6470906576336414, + "grad_norm": 1.7539905309677124, + "learning_rate": 1.581094891065365e-05, + "loss": 3.2192, + "step": 8071 + }, + { + "epoch": 0.647170832414664, + "grad_norm": 1.8599061965942383, + "learning_rate": 1.5809892138957697e-05, + "loss": 3.2168, + "step": 8072 + }, + { + "epoch": 0.6472510071956866, + "grad_norm": 1.7515143156051636, + "learning_rate": 1.5808835269309012e-05, + "loss": 3.2671, + "step": 8073 + }, + { + "epoch": 0.6473311819767092, + "grad_norm": 1.8257739543914795, + "learning_rate": 1.580777830172541e-05, + "loss": 3.1566, + "step": 8074 + }, + { + "epoch": 0.6474113567577319, + "grad_norm": 1.7871509790420532, + "learning_rate": 1.5806721236224712e-05, + "loss": 3.1931, + "step": 8075 + }, + { + "epoch": 0.6474915315387545, + "grad_norm": 1.8197823762893677, + "learning_rate": 1.5805664072824744e-05, + "loss": 3.1409, + "step": 8076 + }, + { + "epoch": 0.6475717063197771, + "grad_norm": 1.8059115409851074, + "learning_rate": 1.580460681154332e-05, + "loss": 3.1089, + "step": 8077 + }, + { + "epoch": 0.6476518811007997, + "grad_norm": 1.8553682565689087, + "learning_rate": 1.5803549452398274e-05, + "loss": 3.1703, + "step": 8078 + }, + { + "epoch": 0.6477320558818224, + "grad_norm": 1.8373521566390991, + "learning_rate": 1.580249199540743e-05, + "loss": 3.2072, + "step": 8079 + }, + { + "epoch": 0.647812230662845, + "grad_norm": 1.9011600017547607, + "learning_rate": 1.5801434440588613e-05, + "loss": 3.2039, + "step": 8080 + }, + { + "epoch": 0.6478924054438676, + "grad_norm": 1.868597149848938, + "learning_rate": 1.5800376787959663e-05, + "loss": 3.204, + "step": 8081 + }, + { + "epoch": 0.6479725802248902, + "grad_norm": 1.7457331418991089, + "learning_rate": 1.5799319037538398e-05, + "loss": 3.1726, + "step": 8082 + }, + { + "epoch": 0.6480527550059129, + "grad_norm": 1.800216794013977, + "learning_rate": 1.5798261189342663e-05, + "loss": 3.1617, + "step": 8083 + }, + { + "epoch": 0.6481329297869355, + "grad_norm": 1.866369366645813, + "learning_rate": 1.5797203243390282e-05, + "loss": 3.333, + "step": 8084 + }, + { + "epoch": 0.6482131045679581, + "grad_norm": 1.8223470449447632, + "learning_rate": 1.5796145199699102e-05, + "loss": 3.1626, + "step": 8085 + }, + { + "epoch": 0.6482932793489807, + "grad_norm": 1.815803050994873, + "learning_rate": 1.5795087058286954e-05, + "loss": 3.257, + "step": 8086 + }, + { + "epoch": 0.6483734541300034, + "grad_norm": 1.8727415800094604, + "learning_rate": 1.5794028819171684e-05, + "loss": 3.2629, + "step": 8087 + }, + { + "epoch": 0.6484536289110261, + "grad_norm": 1.953526258468628, + "learning_rate": 1.5792970482371125e-05, + "loss": 3.3466, + "step": 8088 + }, + { + "epoch": 0.6485338036920487, + "grad_norm": 1.7932696342468262, + "learning_rate": 1.579191204790313e-05, + "loss": 3.1712, + "step": 8089 + }, + { + "epoch": 0.6486139784730713, + "grad_norm": 1.8562465906143188, + "learning_rate": 1.5790853515785537e-05, + "loss": 3.3905, + "step": 8090 + }, + { + "epoch": 0.648694153254094, + "grad_norm": 1.726957082748413, + "learning_rate": 1.5789794886036195e-05, + "loss": 3.1464, + "step": 8091 + }, + { + "epoch": 0.6487743280351166, + "grad_norm": 1.7683521509170532, + "learning_rate": 1.5788736158672953e-05, + "loss": 3.2897, + "step": 8092 + }, + { + "epoch": 0.6488545028161392, + "grad_norm": 1.8144584894180298, + "learning_rate": 1.578767733371366e-05, + "loss": 3.3206, + "step": 8093 + }, + { + "epoch": 0.6489346775971618, + "grad_norm": 2.008693218231201, + "learning_rate": 1.578661841117616e-05, + "loss": 3.3737, + "step": 8094 + }, + { + "epoch": 0.6490148523781845, + "grad_norm": 1.8287951946258545, + "learning_rate": 1.578555939107832e-05, + "loss": 3.2584, + "step": 8095 + }, + { + "epoch": 0.6490950271592071, + "grad_norm": 0.7746150493621826, + "learning_rate": 1.5784500273437984e-05, + "loss": 2.4963, + "step": 8096 + }, + { + "epoch": 0.6491752019402297, + "grad_norm": 1.8032208681106567, + "learning_rate": 1.5783441058273018e-05, + "loss": 3.136, + "step": 8097 + }, + { + "epoch": 0.6492553767212523, + "grad_norm": 1.8959317207336426, + "learning_rate": 1.5782381745601268e-05, + "loss": 3.2126, + "step": 8098 + }, + { + "epoch": 0.649335551502275, + "grad_norm": 1.9078925848007202, + "learning_rate": 1.57813223354406e-05, + "loss": 3.2252, + "step": 8099 + }, + { + "epoch": 0.6494157262832976, + "grad_norm": 0.7407214641571045, + "learning_rate": 1.5780262827808875e-05, + "loss": 2.4822, + "step": 8100 + }, + { + "epoch": 0.6494959010643202, + "grad_norm": 0.7841601967811584, + "learning_rate": 1.5779203222723964e-05, + "loss": 2.505, + "step": 8101 + }, + { + "epoch": 0.6495760758453428, + "grad_norm": 1.7460376024246216, + "learning_rate": 1.5778143520203714e-05, + "loss": 3.2672, + "step": 8102 + }, + { + "epoch": 0.6496562506263655, + "grad_norm": 1.8679611682891846, + "learning_rate": 1.5777083720266002e-05, + "loss": 3.1652, + "step": 8103 + }, + { + "epoch": 0.6497364254073881, + "grad_norm": 1.8134125471115112, + "learning_rate": 1.5776023822928696e-05, + "loss": 3.2422, + "step": 8104 + }, + { + "epoch": 0.6498166001884107, + "grad_norm": 1.8197026252746582, + "learning_rate": 1.5774963828209665e-05, + "loss": 3.2865, + "step": 8105 + }, + { + "epoch": 0.6498967749694333, + "grad_norm": 1.7485511302947998, + "learning_rate": 1.577390373612678e-05, + "loss": 3.322, + "step": 8106 + }, + { + "epoch": 0.649976949750456, + "grad_norm": 1.7713119983673096, + "learning_rate": 1.577284354669791e-05, + "loss": 3.1936, + "step": 8107 + }, + { + "epoch": 0.6500571245314786, + "grad_norm": 0.7823375463485718, + "learning_rate": 1.5771783259940934e-05, + "loss": 2.4995, + "step": 8108 + }, + { + "epoch": 0.6501372993125013, + "grad_norm": 0.7662980556488037, + "learning_rate": 1.5770722875873726e-05, + "loss": 2.4937, + "step": 8109 + }, + { + "epoch": 0.6502174740935239, + "grad_norm": 1.7535614967346191, + "learning_rate": 1.5769662394514162e-05, + "loss": 3.1617, + "step": 8110 + }, + { + "epoch": 0.6502976488745466, + "grad_norm": 1.8338192701339722, + "learning_rate": 1.5768601815880125e-05, + "loss": 3.1354, + "step": 8111 + }, + { + "epoch": 0.6503778236555692, + "grad_norm": 1.6899230480194092, + "learning_rate": 1.5767541139989498e-05, + "loss": 3.1746, + "step": 8112 + }, + { + "epoch": 0.6504579984365918, + "grad_norm": 1.779868245124817, + "learning_rate": 1.5766480366860155e-05, + "loss": 3.1633, + "step": 8113 + }, + { + "epoch": 0.6505381732176144, + "grad_norm": 2.2254786491394043, + "learning_rate": 1.5765419496509987e-05, + "loss": 3.1727, + "step": 8114 + }, + { + "epoch": 0.650618347998637, + "grad_norm": 1.7897804975509644, + "learning_rate": 1.576435852895688e-05, + "loss": 3.2036, + "step": 8115 + }, + { + "epoch": 0.6506985227796597, + "grad_norm": 0.770637571811676, + "learning_rate": 1.5763297464218718e-05, + "loss": 2.4968, + "step": 8116 + }, + { + "epoch": 0.6507786975606823, + "grad_norm": 1.7789905071258545, + "learning_rate": 1.5762236302313396e-05, + "loss": 3.1397, + "step": 8117 + }, + { + "epoch": 0.6508588723417049, + "grad_norm": 1.8234288692474365, + "learning_rate": 1.5761175043258795e-05, + "loss": 3.285, + "step": 8118 + }, + { + "epoch": 0.6509390471227275, + "grad_norm": 1.6965726613998413, + "learning_rate": 1.5760113687072812e-05, + "loss": 3.1488, + "step": 8119 + }, + { + "epoch": 0.6510192219037502, + "grad_norm": 1.9665061235427856, + "learning_rate": 1.5759052233773346e-05, + "loss": 3.3041, + "step": 8120 + }, + { + "epoch": 0.6510993966847728, + "grad_norm": 1.8504559993743896, + "learning_rate": 1.575799068337829e-05, + "loss": 3.3362, + "step": 8121 + }, + { + "epoch": 0.6511795714657954, + "grad_norm": 0.7356554865837097, + "learning_rate": 1.5756929035905538e-05, + "loss": 2.4963, + "step": 8122 + }, + { + "epoch": 0.651259746246818, + "grad_norm": 1.7680293321609497, + "learning_rate": 1.5755867291372994e-05, + "loss": 3.2222, + "step": 8123 + }, + { + "epoch": 0.6513399210278407, + "grad_norm": 1.8151938915252686, + "learning_rate": 1.5754805449798555e-05, + "loss": 3.1808, + "step": 8124 + }, + { + "epoch": 0.6514200958088633, + "grad_norm": 1.861488699913025, + "learning_rate": 1.5753743511200124e-05, + "loss": 3.2679, + "step": 8125 + }, + { + "epoch": 0.6515002705898859, + "grad_norm": 1.982914686203003, + "learning_rate": 1.57526814755956e-05, + "loss": 3.2841, + "step": 8126 + }, + { + "epoch": 0.6515804453709085, + "grad_norm": 1.798306941986084, + "learning_rate": 1.57516193430029e-05, + "loss": 3.2031, + "step": 8127 + }, + { + "epoch": 0.6516606201519312, + "grad_norm": 1.846084475517273, + "learning_rate": 1.5750557113439925e-05, + "loss": 3.3418, + "step": 8128 + }, + { + "epoch": 0.6517407949329539, + "grad_norm": 1.864256739616394, + "learning_rate": 1.574949478692458e-05, + "loss": 3.2216, + "step": 8129 + }, + { + "epoch": 0.6518209697139765, + "grad_norm": 1.7428326606750488, + "learning_rate": 1.5748432363474784e-05, + "loss": 3.2501, + "step": 8130 + }, + { + "epoch": 0.6519011444949991, + "grad_norm": 1.7674314975738525, + "learning_rate": 1.5747369843108443e-05, + "loss": 3.17, + "step": 8131 + }, + { + "epoch": 0.6519813192760218, + "grad_norm": 1.8296817541122437, + "learning_rate": 1.574630722584347e-05, + "loss": 3.249, + "step": 8132 + }, + { + "epoch": 0.6520614940570444, + "grad_norm": 1.8202404975891113, + "learning_rate": 1.5745244511697784e-05, + "loss": 3.2336, + "step": 8133 + }, + { + "epoch": 0.652141668838067, + "grad_norm": 1.9239716529846191, + "learning_rate": 1.57441817006893e-05, + "loss": 3.3237, + "step": 8134 + }, + { + "epoch": 0.6522218436190896, + "grad_norm": 1.8805304765701294, + "learning_rate": 1.5743118792835935e-05, + "loss": 3.2097, + "step": 8135 + }, + { + "epoch": 0.6523020184001123, + "grad_norm": 1.9629652500152588, + "learning_rate": 1.5742055788155616e-05, + "loss": 3.2219, + "step": 8136 + }, + { + "epoch": 0.6523821931811349, + "grad_norm": 1.798224687576294, + "learning_rate": 1.5740992686666255e-05, + "loss": 3.1726, + "step": 8137 + }, + { + "epoch": 0.6524623679621575, + "grad_norm": 1.8467059135437012, + "learning_rate": 1.573992948838578e-05, + "loss": 3.1671, + "step": 8138 + }, + { + "epoch": 0.6525425427431801, + "grad_norm": 1.7362573146820068, + "learning_rate": 1.5738866193332122e-05, + "loss": 3.2391, + "step": 8139 + }, + { + "epoch": 0.6526227175242028, + "grad_norm": 1.7567121982574463, + "learning_rate": 1.5737802801523198e-05, + "loss": 3.182, + "step": 8140 + }, + { + "epoch": 0.6527028923052254, + "grad_norm": 1.9867980480194092, + "learning_rate": 1.5736739312976944e-05, + "loss": 3.2141, + "step": 8141 + }, + { + "epoch": 0.652783067086248, + "grad_norm": 1.852699637413025, + "learning_rate": 1.5735675727711285e-05, + "loss": 3.2534, + "step": 8142 + }, + { + "epoch": 0.6528632418672706, + "grad_norm": 1.7860150337219238, + "learning_rate": 1.5734612045744156e-05, + "loss": 3.3433, + "step": 8143 + }, + { + "epoch": 0.6529434166482933, + "grad_norm": 1.735473871231079, + "learning_rate": 1.573354826709349e-05, + "loss": 3.2291, + "step": 8144 + }, + { + "epoch": 0.6530235914293159, + "grad_norm": 1.8567098379135132, + "learning_rate": 1.573248439177722e-05, + "loss": 3.241, + "step": 8145 + }, + { + "epoch": 0.6531037662103385, + "grad_norm": 1.7735153436660767, + "learning_rate": 1.573142041981328e-05, + "loss": 3.2561, + "step": 8146 + }, + { + "epoch": 0.6531839409913611, + "grad_norm": 1.9183663129806519, + "learning_rate": 1.5730356351219617e-05, + "loss": 3.3033, + "step": 8147 + }, + { + "epoch": 0.6532641157723837, + "grad_norm": 1.8386797904968262, + "learning_rate": 1.572929218601416e-05, + "loss": 3.1908, + "step": 8148 + }, + { + "epoch": 0.6533442905534065, + "grad_norm": 1.8293848037719727, + "learning_rate": 1.5728227924214857e-05, + "loss": 3.2554, + "step": 8149 + }, + { + "epoch": 0.6534244653344291, + "grad_norm": 1.843459963798523, + "learning_rate": 1.5727163565839655e-05, + "loss": 3.228, + "step": 8150 + }, + { + "epoch": 0.6535046401154517, + "grad_norm": 1.8275936841964722, + "learning_rate": 1.5726099110906492e-05, + "loss": 3.1902, + "step": 8151 + }, + { + "epoch": 0.6535848148964744, + "grad_norm": 1.8855377435684204, + "learning_rate": 1.5725034559433314e-05, + "loss": 3.3016, + "step": 8152 + }, + { + "epoch": 0.653664989677497, + "grad_norm": 0.777254581451416, + "learning_rate": 1.5723969911438072e-05, + "loss": 2.4933, + "step": 8153 + }, + { + "epoch": 0.6537451644585196, + "grad_norm": 1.7730475664138794, + "learning_rate": 1.5722905166938713e-05, + "loss": 3.1912, + "step": 8154 + }, + { + "epoch": 0.6538253392395422, + "grad_norm": 1.7941175699234009, + "learning_rate": 1.572184032595319e-05, + "loss": 3.3032, + "step": 8155 + }, + { + "epoch": 0.6539055140205648, + "grad_norm": 1.8602887392044067, + "learning_rate": 1.572077538849946e-05, + "loss": 3.1302, + "step": 8156 + }, + { + "epoch": 0.6539856888015875, + "grad_norm": 1.7711973190307617, + "learning_rate": 1.571971035459547e-05, + "loss": 3.1427, + "step": 8157 + }, + { + "epoch": 0.6540658635826101, + "grad_norm": 1.7458387613296509, + "learning_rate": 1.5718645224259178e-05, + "loss": 3.1584, + "step": 8158 + }, + { + "epoch": 0.6541460383636327, + "grad_norm": 1.7659413814544678, + "learning_rate": 1.5717579997508546e-05, + "loss": 3.2473, + "step": 8159 + }, + { + "epoch": 0.6542262131446553, + "grad_norm": 1.7096812725067139, + "learning_rate": 1.5716514674361526e-05, + "loss": 3.2444, + "step": 8160 + }, + { + "epoch": 0.654306387925678, + "grad_norm": 1.8002547025680542, + "learning_rate": 1.571544925483609e-05, + "loss": 3.2574, + "step": 8161 + }, + { + "epoch": 0.6543865627067006, + "grad_norm": 1.8027546405792236, + "learning_rate": 1.5714383738950188e-05, + "loss": 3.1483, + "step": 8162 + }, + { + "epoch": 0.6544667374877232, + "grad_norm": 0.7718815803527832, + "learning_rate": 1.571331812672179e-05, + "loss": 2.498, + "step": 8163 + }, + { + "epoch": 0.6545469122687458, + "grad_norm": 1.7860126495361328, + "learning_rate": 1.5712252418168864e-05, + "loss": 3.2451, + "step": 8164 + }, + { + "epoch": 0.6546270870497685, + "grad_norm": 1.7982118129730225, + "learning_rate": 1.5711186613309375e-05, + "loss": 3.2158, + "step": 8165 + }, + { + "epoch": 0.6547072618307911, + "grad_norm": 1.7865475416183472, + "learning_rate": 1.5710120712161292e-05, + "loss": 3.1719, + "step": 8166 + }, + { + "epoch": 0.6547874366118137, + "grad_norm": 1.766509771347046, + "learning_rate": 1.570905471474259e-05, + "loss": 3.1626, + "step": 8167 + }, + { + "epoch": 0.6548676113928363, + "grad_norm": 1.799805998802185, + "learning_rate": 1.570798862107123e-05, + "loss": 3.1285, + "step": 8168 + }, + { + "epoch": 0.654947786173859, + "grad_norm": 0.7188494205474854, + "learning_rate": 1.5706922431165202e-05, + "loss": 2.495, + "step": 8169 + }, + { + "epoch": 0.6550279609548817, + "grad_norm": 1.7627849578857422, + "learning_rate": 1.5705856145042467e-05, + "loss": 3.2327, + "step": 8170 + }, + { + "epoch": 0.6551081357359043, + "grad_norm": 1.7820159196853638, + "learning_rate": 1.570478976272101e-05, + "loss": 3.2218, + "step": 8171 + }, + { + "epoch": 0.6551883105169269, + "grad_norm": 1.8101047277450562, + "learning_rate": 1.5703723284218812e-05, + "loss": 3.2479, + "step": 8172 + }, + { + "epoch": 0.6552684852979496, + "grad_norm": 1.736089825630188, + "learning_rate": 1.5702656709553846e-05, + "loss": 3.295, + "step": 8173 + }, + { + "epoch": 0.6553486600789722, + "grad_norm": 1.6925790309906006, + "learning_rate": 1.57015900387441e-05, + "loss": 3.1133, + "step": 8174 + }, + { + "epoch": 0.6554288348599948, + "grad_norm": 1.7659984827041626, + "learning_rate": 1.570052327180755e-05, + "loss": 3.2229, + "step": 8175 + }, + { + "epoch": 0.6555090096410174, + "grad_norm": 1.732958197593689, + "learning_rate": 1.569945640876219e-05, + "loss": 3.2643, + "step": 8176 + }, + { + "epoch": 0.6555891844220401, + "grad_norm": 1.8070334196090698, + "learning_rate": 1.5698389449626005e-05, + "loss": 3.25, + "step": 8177 + }, + { + "epoch": 0.6556693592030627, + "grad_norm": 1.8374804258346558, + "learning_rate": 1.569732239441698e-05, + "loss": 3.215, + "step": 8178 + }, + { + "epoch": 0.6557495339840853, + "grad_norm": 1.7133088111877441, + "learning_rate": 1.5696255243153107e-05, + "loss": 3.153, + "step": 8179 + }, + { + "epoch": 0.6558297087651079, + "grad_norm": 2.0159547328948975, + "learning_rate": 1.569518799585238e-05, + "loss": 3.2596, + "step": 8180 + }, + { + "epoch": 0.6559098835461306, + "grad_norm": 1.8480021953582764, + "learning_rate": 1.569412065253279e-05, + "loss": 3.1815, + "step": 8181 + }, + { + "epoch": 0.6559900583271532, + "grad_norm": 1.8662381172180176, + "learning_rate": 1.5693053213212334e-05, + "loss": 3.1987, + "step": 8182 + }, + { + "epoch": 0.6560702331081758, + "grad_norm": 1.7674177885055542, + "learning_rate": 1.5691985677909006e-05, + "loss": 3.4053, + "step": 8183 + }, + { + "epoch": 0.6561504078891984, + "grad_norm": 1.9069386720657349, + "learning_rate": 1.5690918046640807e-05, + "loss": 3.303, + "step": 8184 + }, + { + "epoch": 0.656230582670221, + "grad_norm": 1.9408360719680786, + "learning_rate": 1.5689850319425733e-05, + "loss": 3.2889, + "step": 8185 + }, + { + "epoch": 0.6563107574512437, + "grad_norm": 1.7867780923843384, + "learning_rate": 1.568878249628179e-05, + "loss": 3.2122, + "step": 8186 + }, + { + "epoch": 0.6563909322322663, + "grad_norm": 1.7087430953979492, + "learning_rate": 1.5687714577226977e-05, + "loss": 3.196, + "step": 8187 + }, + { + "epoch": 0.6564711070132889, + "grad_norm": 1.8297617435455322, + "learning_rate": 1.56866465622793e-05, + "loss": 3.2801, + "step": 8188 + }, + { + "epoch": 0.6565512817943115, + "grad_norm": 1.7632417678833008, + "learning_rate": 1.5685578451456768e-05, + "loss": 3.0988, + "step": 8189 + }, + { + "epoch": 0.6566314565753343, + "grad_norm": 0.7340662479400635, + "learning_rate": 1.5684510244777386e-05, + "loss": 2.4823, + "step": 8190 + }, + { + "epoch": 0.6567116313563569, + "grad_norm": 1.7772210836410522, + "learning_rate": 1.5683441942259164e-05, + "loss": 3.2042, + "step": 8191 + }, + { + "epoch": 0.6567918061373795, + "grad_norm": 1.8880796432495117, + "learning_rate": 1.5682373543920117e-05, + "loss": 3.2139, + "step": 8192 + }, + { + "epoch": 0.6568719809184022, + "grad_norm": 1.8339123725891113, + "learning_rate": 1.568130504977825e-05, + "loss": 3.1575, + "step": 8193 + }, + { + "epoch": 0.6569521556994248, + "grad_norm": 1.7236649990081787, + "learning_rate": 1.5680236459851586e-05, + "loss": 3.2894, + "step": 8194 + }, + { + "epoch": 0.6570323304804474, + "grad_norm": 1.8405566215515137, + "learning_rate": 1.567916777415814e-05, + "loss": 3.118, + "step": 8195 + }, + { + "epoch": 0.65711250526147, + "grad_norm": 2.1503303050994873, + "learning_rate": 1.5678098992715922e-05, + "loss": 3.3892, + "step": 8196 + }, + { + "epoch": 0.6571926800424926, + "grad_norm": 1.7335580587387085, + "learning_rate": 1.5677030115542956e-05, + "loss": 3.2191, + "step": 8197 + }, + { + "epoch": 0.6572728548235153, + "grad_norm": 1.8144110441207886, + "learning_rate": 1.5675961142657266e-05, + "loss": 3.2394, + "step": 8198 + }, + { + "epoch": 0.6573530296045379, + "grad_norm": 1.8392940759658813, + "learning_rate": 1.567489207407687e-05, + "loss": 3.2091, + "step": 8199 + }, + { + "epoch": 0.6574332043855605, + "grad_norm": 1.8519551753997803, + "learning_rate": 1.5673822909819792e-05, + "loss": 3.2392, + "step": 8200 + }, + { + "epoch": 0.6575133791665831, + "grad_norm": 1.8427062034606934, + "learning_rate": 1.5672753649904062e-05, + "loss": 3.2566, + "step": 8201 + }, + { + "epoch": 0.6575935539476058, + "grad_norm": 1.746370792388916, + "learning_rate": 1.5671684294347704e-05, + "loss": 3.1768, + "step": 8202 + }, + { + "epoch": 0.6576737287286284, + "grad_norm": 1.819201111793518, + "learning_rate": 1.567061484316875e-05, + "loss": 3.3525, + "step": 8203 + }, + { + "epoch": 0.657753903509651, + "grad_norm": 1.8268004655838013, + "learning_rate": 1.5669545296385223e-05, + "loss": 3.2239, + "step": 8204 + }, + { + "epoch": 0.6578340782906736, + "grad_norm": 1.879231572151184, + "learning_rate": 1.5668475654015164e-05, + "loss": 3.2268, + "step": 8205 + }, + { + "epoch": 0.6579142530716963, + "grad_norm": 1.857922077178955, + "learning_rate": 1.56674059160766e-05, + "loss": 3.2987, + "step": 8206 + }, + { + "epoch": 0.6579944278527189, + "grad_norm": 1.775478720664978, + "learning_rate": 1.5666336082587574e-05, + "loss": 3.1383, + "step": 8207 + }, + { + "epoch": 0.6580746026337415, + "grad_norm": 1.9438259601593018, + "learning_rate": 1.5665266153566115e-05, + "loss": 3.2064, + "step": 8208 + }, + { + "epoch": 0.6581547774147641, + "grad_norm": 1.9142769575119019, + "learning_rate": 1.5664196129030268e-05, + "loss": 3.2589, + "step": 8209 + }, + { + "epoch": 0.6582349521957868, + "grad_norm": 1.8209744691848755, + "learning_rate": 1.566312600899807e-05, + "loss": 3.2368, + "step": 8210 + }, + { + "epoch": 0.6583151269768095, + "grad_norm": 1.883424997329712, + "learning_rate": 1.5662055793487564e-05, + "loss": 3.119, + "step": 8211 + }, + { + "epoch": 0.6583953017578321, + "grad_norm": 1.877163052558899, + "learning_rate": 1.566098548251679e-05, + "loss": 3.2618, + "step": 8212 + }, + { + "epoch": 0.6584754765388547, + "grad_norm": 1.7807602882385254, + "learning_rate": 1.5659915076103796e-05, + "loss": 3.2632, + "step": 8213 + }, + { + "epoch": 0.6585556513198774, + "grad_norm": 1.9286565780639648, + "learning_rate": 1.565884457426663e-05, + "loss": 3.3794, + "step": 8214 + }, + { + "epoch": 0.6586358261009, + "grad_norm": 1.8025354146957397, + "learning_rate": 1.565777397702334e-05, + "loss": 3.2574, + "step": 8215 + }, + { + "epoch": 0.6587160008819226, + "grad_norm": 1.829738736152649, + "learning_rate": 1.5656703284391976e-05, + "loss": 3.1641, + "step": 8216 + }, + { + "epoch": 0.6587961756629452, + "grad_norm": 1.64226233959198, + "learning_rate": 1.5655632496390584e-05, + "loss": 3.15, + "step": 8217 + }, + { + "epoch": 0.6588763504439679, + "grad_norm": 1.7817202806472778, + "learning_rate": 1.5654561613037223e-05, + "loss": 3.1036, + "step": 8218 + }, + { + "epoch": 0.6589565252249905, + "grad_norm": 1.7273569107055664, + "learning_rate": 1.5653490634349945e-05, + "loss": 3.2166, + "step": 8219 + }, + { + "epoch": 0.6590367000060131, + "grad_norm": 1.7243812084197998, + "learning_rate": 1.565241956034681e-05, + "loss": 3.3043, + "step": 8220 + }, + { + "epoch": 0.6591168747870357, + "grad_norm": 2.0075035095214844, + "learning_rate": 1.5651348391045874e-05, + "loss": 3.2826, + "step": 8221 + }, + { + "epoch": 0.6591970495680584, + "grad_norm": 1.8190652132034302, + "learning_rate": 1.5650277126465196e-05, + "loss": 3.184, + "step": 8222 + }, + { + "epoch": 0.659277224349081, + "grad_norm": 1.9722871780395508, + "learning_rate": 1.5649205766622832e-05, + "loss": 3.2524, + "step": 8223 + }, + { + "epoch": 0.6593573991301036, + "grad_norm": 1.8581962585449219, + "learning_rate": 1.5648134311536854e-05, + "loss": 3.1605, + "step": 8224 + }, + { + "epoch": 0.6594375739111262, + "grad_norm": 1.8413556814193726, + "learning_rate": 1.564706276122532e-05, + "loss": 3.2603, + "step": 8225 + }, + { + "epoch": 0.6595177486921489, + "grad_norm": 1.812500238418579, + "learning_rate": 1.5645991115706297e-05, + "loss": 3.1807, + "step": 8226 + }, + { + "epoch": 0.6595979234731715, + "grad_norm": 1.7936906814575195, + "learning_rate": 1.564491937499786e-05, + "loss": 3.218, + "step": 8227 + }, + { + "epoch": 0.6596780982541941, + "grad_norm": 1.74465012550354, + "learning_rate": 1.5643847539118066e-05, + "loss": 3.2058, + "step": 8228 + }, + { + "epoch": 0.6597582730352167, + "grad_norm": 1.8676003217697144, + "learning_rate": 1.564277560808499e-05, + "loss": 3.3463, + "step": 8229 + }, + { + "epoch": 0.6598384478162393, + "grad_norm": 1.8433747291564941, + "learning_rate": 1.5641703581916708e-05, + "loss": 3.275, + "step": 8230 + }, + { + "epoch": 0.6599186225972621, + "grad_norm": 1.7572563886642456, + "learning_rate": 1.5640631460631295e-05, + "loss": 3.1667, + "step": 8231 + }, + { + "epoch": 0.6599987973782847, + "grad_norm": 1.8688136339187622, + "learning_rate": 1.563955924424682e-05, + "loss": 3.1437, + "step": 8232 + }, + { + "epoch": 0.6600789721593073, + "grad_norm": 1.8020275831222534, + "learning_rate": 1.5638486932781363e-05, + "loss": 3.2007, + "step": 8233 + }, + { + "epoch": 0.66015914694033, + "grad_norm": 1.9604320526123047, + "learning_rate": 1.5637414526253007e-05, + "loss": 3.2842, + "step": 8234 + }, + { + "epoch": 0.6602393217213526, + "grad_norm": 1.8807612657546997, + "learning_rate": 1.563634202467983e-05, + "loss": 3.2997, + "step": 8235 + }, + { + "epoch": 0.6603194965023752, + "grad_norm": 1.823578953742981, + "learning_rate": 1.563526942807991e-05, + "loss": 3.2715, + "step": 8236 + }, + { + "epoch": 0.6603996712833978, + "grad_norm": 1.784705638885498, + "learning_rate": 1.563419673647133e-05, + "loss": 3.207, + "step": 8237 + }, + { + "epoch": 0.6604798460644205, + "grad_norm": 1.8706669807434082, + "learning_rate": 1.5633123949872184e-05, + "loss": 3.2063, + "step": 8238 + }, + { + "epoch": 0.6605600208454431, + "grad_norm": 1.8918477296829224, + "learning_rate": 1.5632051068300554e-05, + "loss": 3.1795, + "step": 8239 + }, + { + "epoch": 0.6606401956264657, + "grad_norm": 2.056108236312866, + "learning_rate": 1.5630978091774525e-05, + "loss": 3.2635, + "step": 8240 + }, + { + "epoch": 0.6607203704074883, + "grad_norm": 1.7332546710968018, + "learning_rate": 1.562990502031219e-05, + "loss": 3.0967, + "step": 8241 + }, + { + "epoch": 0.660800545188511, + "grad_norm": 1.7628371715545654, + "learning_rate": 1.5628831853931644e-05, + "loss": 3.2043, + "step": 8242 + }, + { + "epoch": 0.6608807199695336, + "grad_norm": 1.7298483848571777, + "learning_rate": 1.5627758592650974e-05, + "loss": 3.3033, + "step": 8243 + }, + { + "epoch": 0.6609608947505562, + "grad_norm": 1.8183939456939697, + "learning_rate": 1.5626685236488275e-05, + "loss": 3.1799, + "step": 8244 + }, + { + "epoch": 0.6610410695315788, + "grad_norm": 2.142810344696045, + "learning_rate": 1.5625611785461653e-05, + "loss": 3.2852, + "step": 8245 + }, + { + "epoch": 0.6611212443126014, + "grad_norm": 1.70719575881958, + "learning_rate": 1.562453823958919e-05, + "loss": 3.3354, + "step": 8246 + }, + { + "epoch": 0.6612014190936241, + "grad_norm": 1.8846538066864014, + "learning_rate": 1.5623464598889e-05, + "loss": 3.2296, + "step": 8247 + }, + { + "epoch": 0.6612815938746467, + "grad_norm": 1.7733186483383179, + "learning_rate": 1.5622390863379176e-05, + "loss": 3.2679, + "step": 8248 + }, + { + "epoch": 0.6613617686556693, + "grad_norm": 1.9490023851394653, + "learning_rate": 1.5621317033077828e-05, + "loss": 3.3115, + "step": 8249 + }, + { + "epoch": 0.6614419434366919, + "grad_norm": 1.772400975227356, + "learning_rate": 1.562024310800305e-05, + "loss": 3.2188, + "step": 8250 + }, + { + "epoch": 0.6615221182177147, + "grad_norm": 1.766957402229309, + "learning_rate": 1.561916908817296e-05, + "loss": 3.2079, + "step": 8251 + }, + { + "epoch": 0.6616022929987373, + "grad_norm": 1.7872588634490967, + "learning_rate": 1.561809497360566e-05, + "loss": 3.2631, + "step": 8252 + }, + { + "epoch": 0.6616824677797599, + "grad_norm": 1.8929784297943115, + "learning_rate": 1.5617020764319253e-05, + "loss": 3.2598, + "step": 8253 + }, + { + "epoch": 0.6617626425607825, + "grad_norm": 1.900347352027893, + "learning_rate": 1.561594646033186e-05, + "loss": 3.3947, + "step": 8254 + }, + { + "epoch": 0.6618428173418052, + "grad_norm": 1.8120168447494507, + "learning_rate": 1.5614872061661585e-05, + "loss": 3.1414, + "step": 8255 + }, + { + "epoch": 0.6619229921228278, + "grad_norm": 1.8029084205627441, + "learning_rate": 1.561379756832655e-05, + "loss": 3.265, + "step": 8256 + }, + { + "epoch": 0.6620031669038504, + "grad_norm": 1.8611953258514404, + "learning_rate": 1.5612722980344865e-05, + "loss": 3.2574, + "step": 8257 + }, + { + "epoch": 0.662083341684873, + "grad_norm": 1.8836195468902588, + "learning_rate": 1.561164829773465e-05, + "loss": 3.3204, + "step": 8258 + }, + { + "epoch": 0.6621635164658957, + "grad_norm": 1.861435890197754, + "learning_rate": 1.561057352051402e-05, + "loss": 3.2841, + "step": 8259 + }, + { + "epoch": 0.6622436912469183, + "grad_norm": 1.7182362079620361, + "learning_rate": 1.56094986487011e-05, + "loss": 3.1799, + "step": 8260 + }, + { + "epoch": 0.6623238660279409, + "grad_norm": 1.7898589372634888, + "learning_rate": 1.5608423682314005e-05, + "loss": 3.2048, + "step": 8261 + }, + { + "epoch": 0.6624040408089635, + "grad_norm": 1.851170539855957, + "learning_rate": 1.5607348621370865e-05, + "loss": 3.2638, + "step": 8262 + }, + { + "epoch": 0.6624842155899862, + "grad_norm": 1.9134953022003174, + "learning_rate": 1.5606273465889804e-05, + "loss": 3.3179, + "step": 8263 + }, + { + "epoch": 0.6625643903710088, + "grad_norm": 1.7591012716293335, + "learning_rate": 1.560519821588895e-05, + "loss": 3.4244, + "step": 8264 + }, + { + "epoch": 0.6626445651520314, + "grad_norm": 1.780640959739685, + "learning_rate": 1.560412287138643e-05, + "loss": 3.1769, + "step": 8265 + }, + { + "epoch": 0.662724739933054, + "grad_norm": 1.7835112810134888, + "learning_rate": 1.5603047432400373e-05, + "loss": 3.1345, + "step": 8266 + }, + { + "epoch": 0.6628049147140767, + "grad_norm": 1.8537834882736206, + "learning_rate": 1.560197189894891e-05, + "loss": 3.3491, + "step": 8267 + }, + { + "epoch": 0.6628850894950993, + "grad_norm": 1.7982150316238403, + "learning_rate": 1.5600896271050175e-05, + "loss": 3.2053, + "step": 8268 + }, + { + "epoch": 0.6629652642761219, + "grad_norm": 1.7517775297164917, + "learning_rate": 1.5599820548722303e-05, + "loss": 3.0474, + "step": 8269 + }, + { + "epoch": 0.6630454390571445, + "grad_norm": 1.786166787147522, + "learning_rate": 1.559874473198343e-05, + "loss": 3.1632, + "step": 8270 + }, + { + "epoch": 0.6631256138381671, + "grad_norm": 0.7543928623199463, + "learning_rate": 1.5597668820851698e-05, + "loss": 2.5109, + "step": 8271 + }, + { + "epoch": 0.6632057886191899, + "grad_norm": 1.8107064962387085, + "learning_rate": 1.5596592815345235e-05, + "loss": 3.2886, + "step": 8272 + }, + { + "epoch": 0.6632859634002125, + "grad_norm": 1.8846251964569092, + "learning_rate": 1.5595516715482192e-05, + "loss": 3.267, + "step": 8273 + }, + { + "epoch": 0.6633661381812351, + "grad_norm": 1.7485113143920898, + "learning_rate": 1.5594440521280715e-05, + "loss": 3.2158, + "step": 8274 + }, + { + "epoch": 0.6634463129622578, + "grad_norm": 1.8402653932571411, + "learning_rate": 1.5593364232758935e-05, + "loss": 3.2491, + "step": 8275 + }, + { + "epoch": 0.6635264877432804, + "grad_norm": 1.8421056270599365, + "learning_rate": 1.5592287849935012e-05, + "loss": 3.2916, + "step": 8276 + }, + { + "epoch": 0.663606662524303, + "grad_norm": 1.8847451210021973, + "learning_rate": 1.559121137282708e-05, + "loss": 3.2346, + "step": 8277 + }, + { + "epoch": 0.6636868373053256, + "grad_norm": 1.872086763381958, + "learning_rate": 1.55901348014533e-05, + "loss": 3.1765, + "step": 8278 + }, + { + "epoch": 0.6637670120863483, + "grad_norm": 1.8592722415924072, + "learning_rate": 1.5589058135831814e-05, + "loss": 3.2623, + "step": 8279 + }, + { + "epoch": 0.6638471868673709, + "grad_norm": 1.9224902391433716, + "learning_rate": 1.558798137598078e-05, + "loss": 3.2191, + "step": 8280 + }, + { + "epoch": 0.6639273616483935, + "grad_norm": 1.7984699010849, + "learning_rate": 1.5586904521918347e-05, + "loss": 3.2488, + "step": 8281 + }, + { + "epoch": 0.6640075364294161, + "grad_norm": 1.7849516868591309, + "learning_rate": 1.5585827573662675e-05, + "loss": 3.2253, + "step": 8282 + }, + { + "epoch": 0.6640877112104387, + "grad_norm": 0.7824373245239258, + "learning_rate": 1.558475053123192e-05, + "loss": 2.5087, + "step": 8283 + }, + { + "epoch": 0.6641678859914614, + "grad_norm": 1.9685901403427124, + "learning_rate": 1.5583673394644235e-05, + "loss": 3.3962, + "step": 8284 + }, + { + "epoch": 0.664248060772484, + "grad_norm": 1.783015251159668, + "learning_rate": 1.5582596163917787e-05, + "loss": 3.1813, + "step": 8285 + }, + { + "epoch": 0.6643282355535066, + "grad_norm": 0.7182527780532837, + "learning_rate": 1.5581518839070735e-05, + "loss": 2.4924, + "step": 8286 + }, + { + "epoch": 0.6644084103345292, + "grad_norm": 0.7198792099952698, + "learning_rate": 1.5580441420121243e-05, + "loss": 2.4874, + "step": 8287 + }, + { + "epoch": 0.6644885851155519, + "grad_norm": 1.8470851182937622, + "learning_rate": 1.5579363907087472e-05, + "loss": 3.2357, + "step": 8288 + }, + { + "epoch": 0.6645687598965745, + "grad_norm": 1.8280943632125854, + "learning_rate": 1.5578286299987595e-05, + "loss": 3.1902, + "step": 8289 + }, + { + "epoch": 0.6646489346775971, + "grad_norm": 0.742549479007721, + "learning_rate": 1.5577208598839777e-05, + "loss": 2.5027, + "step": 8290 + }, + { + "epoch": 0.6647291094586197, + "grad_norm": 1.8953968286514282, + "learning_rate": 1.5576130803662187e-05, + "loss": 3.2269, + "step": 8291 + }, + { + "epoch": 0.6648092842396425, + "grad_norm": 1.8497296571731567, + "learning_rate": 1.5575052914472997e-05, + "loss": 3.1591, + "step": 8292 + }, + { + "epoch": 0.6648894590206651, + "grad_norm": 1.7791454792022705, + "learning_rate": 1.557397493129038e-05, + "loss": 3.351, + "step": 8293 + }, + { + "epoch": 0.6649696338016877, + "grad_norm": 2.150688409805298, + "learning_rate": 1.557289685413251e-05, + "loss": 3.1587, + "step": 8294 + }, + { + "epoch": 0.6650498085827103, + "grad_norm": 1.6726353168487549, + "learning_rate": 1.5571818683017565e-05, + "loss": 3.2528, + "step": 8295 + }, + { + "epoch": 0.665129983363733, + "grad_norm": 1.746837854385376, + "learning_rate": 1.5570740417963718e-05, + "loss": 3.1213, + "step": 8296 + }, + { + "epoch": 0.6652101581447556, + "grad_norm": 1.83894681930542, + "learning_rate": 1.556966205898915e-05, + "loss": 3.1956, + "step": 8297 + }, + { + "epoch": 0.6652903329257782, + "grad_norm": 1.7474830150604248, + "learning_rate": 1.5568583606112042e-05, + "loss": 3.302, + "step": 8298 + }, + { + "epoch": 0.6653705077068008, + "grad_norm": 1.7587672472000122, + "learning_rate": 1.5567505059350582e-05, + "loss": 3.1706, + "step": 8299 + }, + { + "epoch": 0.6654506824878235, + "grad_norm": 0.7738673686981201, + "learning_rate": 1.5566426418722946e-05, + "loss": 2.5024, + "step": 8300 + }, + { + "epoch": 0.6655308572688461, + "grad_norm": 1.7665536403656006, + "learning_rate": 1.556534768424732e-05, + "loss": 3.1782, + "step": 8301 + }, + { + "epoch": 0.6656110320498687, + "grad_norm": 1.7668802738189697, + "learning_rate": 1.5564268855941896e-05, + "loss": 3.2064, + "step": 8302 + }, + { + "epoch": 0.6656912068308913, + "grad_norm": 1.8578987121582031, + "learning_rate": 1.5563189933824863e-05, + "loss": 3.2617, + "step": 8303 + }, + { + "epoch": 0.665771381611914, + "grad_norm": 1.758855938911438, + "learning_rate": 1.5562110917914404e-05, + "loss": 3.2418, + "step": 8304 + }, + { + "epoch": 0.6658515563929366, + "grad_norm": 1.8165980577468872, + "learning_rate": 1.556103180822872e-05, + "loss": 3.2624, + "step": 8305 + }, + { + "epoch": 0.6659317311739592, + "grad_norm": 1.7356659173965454, + "learning_rate": 1.5559952604785992e-05, + "loss": 3.3781, + "step": 8306 + }, + { + "epoch": 0.6660119059549818, + "grad_norm": 1.7202144861221313, + "learning_rate": 1.5558873307604427e-05, + "loss": 3.1617, + "step": 8307 + }, + { + "epoch": 0.6660920807360045, + "grad_norm": 2.011929750442505, + "learning_rate": 1.5557793916702216e-05, + "loss": 3.24, + "step": 8308 + }, + { + "epoch": 0.6661722555170271, + "grad_norm": 1.9283589124679565, + "learning_rate": 1.555671443209756e-05, + "loss": 3.2428, + "step": 8309 + }, + { + "epoch": 0.6662524302980497, + "grad_norm": 1.775497317314148, + "learning_rate": 1.5555634853808655e-05, + "loss": 3.1758, + "step": 8310 + }, + { + "epoch": 0.6663326050790723, + "grad_norm": 1.8588613271713257, + "learning_rate": 1.5554555181853707e-05, + "loss": 3.3026, + "step": 8311 + }, + { + "epoch": 0.6664127798600951, + "grad_norm": 1.7822773456573486, + "learning_rate": 1.5553475416250914e-05, + "loss": 3.2405, + "step": 8312 + }, + { + "epoch": 0.6664929546411177, + "grad_norm": 1.814028263092041, + "learning_rate": 1.5552395557018485e-05, + "loss": 3.1863, + "step": 8313 + }, + { + "epoch": 0.6665731294221403, + "grad_norm": 1.8636608123779297, + "learning_rate": 1.555131560417462e-05, + "loss": 3.1356, + "step": 8314 + }, + { + "epoch": 0.6666533042031629, + "grad_norm": 1.708481788635254, + "learning_rate": 1.5550235557737536e-05, + "loss": 3.2798, + "step": 8315 + }, + { + "epoch": 0.6667334789841856, + "grad_norm": 1.8644970655441284, + "learning_rate": 1.5549155417725432e-05, + "loss": 3.2305, + "step": 8316 + }, + { + "epoch": 0.6668136537652082, + "grad_norm": 1.9044164419174194, + "learning_rate": 1.5548075184156526e-05, + "loss": 3.1904, + "step": 8317 + }, + { + "epoch": 0.6668938285462308, + "grad_norm": 1.7749826908111572, + "learning_rate": 1.5546994857049026e-05, + "loss": 3.1168, + "step": 8318 + }, + { + "epoch": 0.6669740033272534, + "grad_norm": 1.8443958759307861, + "learning_rate": 1.5545914436421147e-05, + "loss": 3.1909, + "step": 8319 + }, + { + "epoch": 0.667054178108276, + "grad_norm": 1.9654279947280884, + "learning_rate": 1.5544833922291107e-05, + "loss": 3.1737, + "step": 8320 + }, + { + "epoch": 0.6671343528892987, + "grad_norm": 1.904217004776001, + "learning_rate": 1.554375331467712e-05, + "loss": 3.3069, + "step": 8321 + }, + { + "epoch": 0.6672145276703213, + "grad_norm": 1.7477272748947144, + "learning_rate": 1.5542672613597405e-05, + "loss": 3.2164, + "step": 8322 + }, + { + "epoch": 0.6672947024513439, + "grad_norm": 1.8232319355010986, + "learning_rate": 1.5541591819070186e-05, + "loss": 3.1693, + "step": 8323 + }, + { + "epoch": 0.6673748772323665, + "grad_norm": 1.774769902229309, + "learning_rate": 1.554051093111368e-05, + "loss": 3.2549, + "step": 8324 + }, + { + "epoch": 0.6674550520133892, + "grad_norm": 1.8570119142532349, + "learning_rate": 1.5539429949746116e-05, + "loss": 3.1749, + "step": 8325 + }, + { + "epoch": 0.6675352267944118, + "grad_norm": 1.8271715641021729, + "learning_rate": 1.553834887498571e-05, + "loss": 3.1804, + "step": 8326 + }, + { + "epoch": 0.6676154015754344, + "grad_norm": 1.85965096950531, + "learning_rate": 1.5537267706850696e-05, + "loss": 3.271, + "step": 8327 + }, + { + "epoch": 0.667695576356457, + "grad_norm": 1.8728331327438354, + "learning_rate": 1.55361864453593e-05, + "loss": 3.1599, + "step": 8328 + }, + { + "epoch": 0.6677757511374797, + "grad_norm": 0.7572563290596008, + "learning_rate": 1.5535105090529753e-05, + "loss": 2.5016, + "step": 8329 + }, + { + "epoch": 0.6678559259185023, + "grad_norm": 1.797802448272705, + "learning_rate": 1.5534023642380284e-05, + "loss": 3.1666, + "step": 8330 + }, + { + "epoch": 0.6679361006995249, + "grad_norm": 1.7821143865585327, + "learning_rate": 1.5532942100929125e-05, + "loss": 3.266, + "step": 8331 + }, + { + "epoch": 0.6680162754805475, + "grad_norm": 1.9355474710464478, + "learning_rate": 1.5531860466194513e-05, + "loss": 3.2902, + "step": 8332 + }, + { + "epoch": 0.6680964502615703, + "grad_norm": 1.75625479221344, + "learning_rate": 1.5530778738194686e-05, + "loss": 3.288, + "step": 8333 + }, + { + "epoch": 0.6681766250425929, + "grad_norm": 2.1188504695892334, + "learning_rate": 1.5529696916947877e-05, + "loss": 3.2252, + "step": 8334 + }, + { + "epoch": 0.6682567998236155, + "grad_norm": 1.8182107210159302, + "learning_rate": 1.5528615002472327e-05, + "loss": 3.1979, + "step": 8335 + }, + { + "epoch": 0.6683369746046381, + "grad_norm": 1.842892050743103, + "learning_rate": 1.5527532994786275e-05, + "loss": 3.2466, + "step": 8336 + }, + { + "epoch": 0.6684171493856608, + "grad_norm": 0.771941602230072, + "learning_rate": 1.5526450893907965e-05, + "loss": 2.509, + "step": 8337 + }, + { + "epoch": 0.6684973241666834, + "grad_norm": 0.7476075291633606, + "learning_rate": 1.552536869985564e-05, + "loss": 2.499, + "step": 8338 + }, + { + "epoch": 0.668577498947706, + "grad_norm": 1.896388053894043, + "learning_rate": 1.5524286412647548e-05, + "loss": 3.2116, + "step": 8339 + }, + { + "epoch": 0.6686576737287286, + "grad_norm": 1.8450089693069458, + "learning_rate": 1.5523204032301935e-05, + "loss": 3.2361, + "step": 8340 + }, + { + "epoch": 0.6687378485097513, + "grad_norm": 2.0542845726013184, + "learning_rate": 1.5522121558837047e-05, + "loss": 3.2874, + "step": 8341 + }, + { + "epoch": 0.6688180232907739, + "grad_norm": 1.800679326057434, + "learning_rate": 1.5521038992271135e-05, + "loss": 3.1472, + "step": 8342 + }, + { + "epoch": 0.6688981980717965, + "grad_norm": 1.80559504032135, + "learning_rate": 1.5519956332622452e-05, + "loss": 3.2563, + "step": 8343 + }, + { + "epoch": 0.6689783728528191, + "grad_norm": 1.9404404163360596, + "learning_rate": 1.5518873579909253e-05, + "loss": 3.2191, + "step": 8344 + }, + { + "epoch": 0.6690585476338418, + "grad_norm": 2.0879452228546143, + "learning_rate": 1.5517790734149786e-05, + "loss": 3.2418, + "step": 8345 + }, + { + "epoch": 0.6691387224148644, + "grad_norm": 0.799551784992218, + "learning_rate": 1.5516707795362314e-05, + "loss": 2.4985, + "step": 8346 + }, + { + "epoch": 0.669218897195887, + "grad_norm": 1.778228998184204, + "learning_rate": 1.5515624763565094e-05, + "loss": 3.2407, + "step": 8347 + }, + { + "epoch": 0.6692990719769096, + "grad_norm": 1.7550376653671265, + "learning_rate": 1.5514541638776383e-05, + "loss": 3.3929, + "step": 8348 + }, + { + "epoch": 0.6693792467579323, + "grad_norm": 1.8601728677749634, + "learning_rate": 1.5513458421014445e-05, + "loss": 3.36, + "step": 8349 + }, + { + "epoch": 0.6694594215389549, + "grad_norm": 1.8510444164276123, + "learning_rate": 1.551237511029754e-05, + "loss": 3.2744, + "step": 8350 + }, + { + "epoch": 0.6695395963199775, + "grad_norm": 1.8354542255401611, + "learning_rate": 1.551129170664393e-05, + "loss": 3.1953, + "step": 8351 + }, + { + "epoch": 0.6696197711010001, + "grad_norm": 1.888918399810791, + "learning_rate": 1.551020821007189e-05, + "loss": 3.3009, + "step": 8352 + }, + { + "epoch": 0.6696999458820229, + "grad_norm": 1.8044053316116333, + "learning_rate": 1.550912462059968e-05, + "loss": 3.2936, + "step": 8353 + }, + { + "epoch": 0.6697801206630455, + "grad_norm": 1.8353073596954346, + "learning_rate": 1.550804093824557e-05, + "loss": 3.1716, + "step": 8354 + }, + { + "epoch": 0.6698602954440681, + "grad_norm": 1.7106003761291504, + "learning_rate": 1.5506957163027833e-05, + "loss": 3.2867, + "step": 8355 + }, + { + "epoch": 0.6699404702250907, + "grad_norm": 1.7850395441055298, + "learning_rate": 1.5505873294964732e-05, + "loss": 3.2402, + "step": 8356 + }, + { + "epoch": 0.6700206450061134, + "grad_norm": 1.972266674041748, + "learning_rate": 1.5504789334074556e-05, + "loss": 3.3676, + "step": 8357 + }, + { + "epoch": 0.670100819787136, + "grad_norm": 1.7186355590820312, + "learning_rate": 1.5503705280375565e-05, + "loss": 3.176, + "step": 8358 + }, + { + "epoch": 0.6701809945681586, + "grad_norm": 1.8737858533859253, + "learning_rate": 1.550262113388605e-05, + "loss": 3.1998, + "step": 8359 + }, + { + "epoch": 0.6702611693491812, + "grad_norm": 2.0300581455230713, + "learning_rate": 1.5501536894624275e-05, + "loss": 3.2673, + "step": 8360 + }, + { + "epoch": 0.6703413441302039, + "grad_norm": 1.7719842195510864, + "learning_rate": 1.550045256260853e-05, + "loss": 3.2904, + "step": 8361 + }, + { + "epoch": 0.6704215189112265, + "grad_norm": 1.7949280738830566, + "learning_rate": 1.5499368137857094e-05, + "loss": 3.2811, + "step": 8362 + }, + { + "epoch": 0.6705016936922491, + "grad_norm": 1.8662084341049194, + "learning_rate": 1.549828362038825e-05, + "loss": 3.3579, + "step": 8363 + }, + { + "epoch": 0.6705818684732717, + "grad_norm": 1.876846432685852, + "learning_rate": 1.5497199010220278e-05, + "loss": 3.18, + "step": 8364 + }, + { + "epoch": 0.6706620432542943, + "grad_norm": 1.770174264907837, + "learning_rate": 1.549611430737147e-05, + "loss": 3.1581, + "step": 8365 + }, + { + "epoch": 0.670742218035317, + "grad_norm": 1.7107014656066895, + "learning_rate": 1.5495029511860112e-05, + "loss": 3.2172, + "step": 8366 + }, + { + "epoch": 0.6708223928163396, + "grad_norm": 1.7270656824111938, + "learning_rate": 1.5493944623704496e-05, + "loss": 3.2433, + "step": 8367 + }, + { + "epoch": 0.6709025675973622, + "grad_norm": 1.7826179265975952, + "learning_rate": 1.5492859642922906e-05, + "loss": 3.1375, + "step": 8368 + }, + { + "epoch": 0.6709827423783848, + "grad_norm": 1.7467751502990723, + "learning_rate": 1.549177456953364e-05, + "loss": 3.1798, + "step": 8369 + }, + { + "epoch": 0.6710629171594075, + "grad_norm": 1.7611663341522217, + "learning_rate": 1.549068940355499e-05, + "loss": 3.1042, + "step": 8370 + }, + { + "epoch": 0.6711430919404301, + "grad_norm": 2.177570104598999, + "learning_rate": 1.5489604145005253e-05, + "loss": 3.1509, + "step": 8371 + }, + { + "epoch": 0.6712232667214527, + "grad_norm": 1.8512725830078125, + "learning_rate": 1.5488518793902723e-05, + "loss": 3.2423, + "step": 8372 + }, + { + "epoch": 0.6713034415024755, + "grad_norm": 1.7880455255508423, + "learning_rate": 1.5487433350265702e-05, + "loss": 3.1929, + "step": 8373 + }, + { + "epoch": 0.6713836162834981, + "grad_norm": 1.7803977727890015, + "learning_rate": 1.5486347814112486e-05, + "loss": 3.2928, + "step": 8374 + }, + { + "epoch": 0.6714637910645207, + "grad_norm": 1.8639663457870483, + "learning_rate": 1.548526218546138e-05, + "loss": 3.1818, + "step": 8375 + }, + { + "epoch": 0.6715439658455433, + "grad_norm": 2.1291344165802, + "learning_rate": 1.548417646433069e-05, + "loss": 3.3613, + "step": 8376 + }, + { + "epoch": 0.671624140626566, + "grad_norm": 1.7890528440475464, + "learning_rate": 1.5483090650738713e-05, + "loss": 3.2061, + "step": 8377 + }, + { + "epoch": 0.6717043154075886, + "grad_norm": 1.7371052503585815, + "learning_rate": 1.5482004744703764e-05, + "loss": 3.293, + "step": 8378 + }, + { + "epoch": 0.6717844901886112, + "grad_norm": 1.7641195058822632, + "learning_rate": 1.5480918746244147e-05, + "loss": 3.1858, + "step": 8379 + }, + { + "epoch": 0.6718646649696338, + "grad_norm": 1.920333981513977, + "learning_rate": 1.547983265537817e-05, + "loss": 3.285, + "step": 8380 + }, + { + "epoch": 0.6719448397506564, + "grad_norm": 1.8291778564453125, + "learning_rate": 1.5478746472124144e-05, + "loss": 3.1584, + "step": 8381 + }, + { + "epoch": 0.6720250145316791, + "grad_norm": 1.8823456764221191, + "learning_rate": 1.5477660196500385e-05, + "loss": 3.2372, + "step": 8382 + }, + { + "epoch": 0.6721051893127017, + "grad_norm": 1.6702806949615479, + "learning_rate": 1.547657382852521e-05, + "loss": 3.086, + "step": 8383 + }, + { + "epoch": 0.6721853640937243, + "grad_norm": 1.8791813850402832, + "learning_rate": 1.5475487368216924e-05, + "loss": 3.1483, + "step": 8384 + }, + { + "epoch": 0.6722655388747469, + "grad_norm": 1.7663909196853638, + "learning_rate": 1.5474400815593854e-05, + "loss": 3.2672, + "step": 8385 + }, + { + "epoch": 0.6723457136557696, + "grad_norm": 1.8398666381835938, + "learning_rate": 1.547331417067431e-05, + "loss": 3.2443, + "step": 8386 + }, + { + "epoch": 0.6724258884367922, + "grad_norm": 1.6977382898330688, + "learning_rate": 1.5472227433476627e-05, + "loss": 3.1904, + "step": 8387 + }, + { + "epoch": 0.6725060632178148, + "grad_norm": 2.1581811904907227, + "learning_rate": 1.547114060401911e-05, + "loss": 3.3174, + "step": 8388 + }, + { + "epoch": 0.6725862379988374, + "grad_norm": 2.3000566959381104, + "learning_rate": 1.5470053682320097e-05, + "loss": 3.1777, + "step": 8389 + }, + { + "epoch": 0.6726664127798601, + "grad_norm": 1.8043286800384521, + "learning_rate": 1.5468966668397904e-05, + "loss": 3.1902, + "step": 8390 + }, + { + "epoch": 0.6727465875608827, + "grad_norm": 1.898166537284851, + "learning_rate": 1.546787956227086e-05, + "loss": 3.2924, + "step": 8391 + }, + { + "epoch": 0.6728267623419053, + "grad_norm": 1.8739479780197144, + "learning_rate": 1.5466792363957295e-05, + "loss": 3.2561, + "step": 8392 + }, + { + "epoch": 0.6729069371229279, + "grad_norm": 2.122342348098755, + "learning_rate": 1.5465705073475538e-05, + "loss": 3.1948, + "step": 8393 + }, + { + "epoch": 0.6729871119039507, + "grad_norm": 1.792770504951477, + "learning_rate": 1.5464617690843918e-05, + "loss": 3.2717, + "step": 8394 + }, + { + "epoch": 0.6730672866849733, + "grad_norm": 1.8083219528198242, + "learning_rate": 1.546353021608077e-05, + "loss": 3.2352, + "step": 8395 + }, + { + "epoch": 0.6731474614659959, + "grad_norm": 1.939896583557129, + "learning_rate": 1.546244264920443e-05, + "loss": 3.2695, + "step": 8396 + }, + { + "epoch": 0.6732276362470185, + "grad_norm": 1.7819291353225708, + "learning_rate": 1.5461354990233227e-05, + "loss": 3.2635, + "step": 8397 + }, + { + "epoch": 0.6733078110280412, + "grad_norm": 1.8577041625976562, + "learning_rate": 1.5460267239185507e-05, + "loss": 3.2618, + "step": 8398 + }, + { + "epoch": 0.6733879858090638, + "grad_norm": 1.8647047281265259, + "learning_rate": 1.5459179396079608e-05, + "loss": 3.2609, + "step": 8399 + }, + { + "epoch": 0.6734681605900864, + "grad_norm": 1.7562845945358276, + "learning_rate": 1.5458091460933865e-05, + "loss": 3.1853, + "step": 8400 + }, + { + "epoch": 0.673548335371109, + "grad_norm": 1.7955223321914673, + "learning_rate": 1.5457003433766625e-05, + "loss": 3.1064, + "step": 8401 + }, + { + "epoch": 0.6736285101521317, + "grad_norm": 1.9467939138412476, + "learning_rate": 1.545591531459623e-05, + "loss": 3.1148, + "step": 8402 + }, + { + "epoch": 0.6737086849331543, + "grad_norm": 1.7838585376739502, + "learning_rate": 1.5454827103441023e-05, + "loss": 3.2775, + "step": 8403 + }, + { + "epoch": 0.6737888597141769, + "grad_norm": 1.9022856950759888, + "learning_rate": 1.5453738800319356e-05, + "loss": 3.1798, + "step": 8404 + }, + { + "epoch": 0.6738690344951995, + "grad_norm": 1.91537606716156, + "learning_rate": 1.5452650405249573e-05, + "loss": 3.2814, + "step": 8405 + }, + { + "epoch": 0.6739492092762221, + "grad_norm": 1.767474889755249, + "learning_rate": 1.545156191825003e-05, + "loss": 3.259, + "step": 8406 + }, + { + "epoch": 0.6740293840572448, + "grad_norm": 1.8035229444503784, + "learning_rate": 1.545047333933907e-05, + "loss": 3.142, + "step": 8407 + }, + { + "epoch": 0.6741095588382674, + "grad_norm": 1.8485791683197021, + "learning_rate": 1.5449384668535056e-05, + "loss": 3.1635, + "step": 8408 + }, + { + "epoch": 0.67418973361929, + "grad_norm": 1.8341463804244995, + "learning_rate": 1.5448295905856335e-05, + "loss": 3.2209, + "step": 8409 + }, + { + "epoch": 0.6742699084003126, + "grad_norm": 2.0110952854156494, + "learning_rate": 1.544720705132126e-05, + "loss": 3.311, + "step": 8410 + }, + { + "epoch": 0.6743500831813353, + "grad_norm": 0.7608327865600586, + "learning_rate": 1.54461181049482e-05, + "loss": 2.4883, + "step": 8411 + }, + { + "epoch": 0.6744302579623579, + "grad_norm": 1.8056007623672485, + "learning_rate": 1.5445029066755505e-05, + "loss": 3.3341, + "step": 8412 + }, + { + "epoch": 0.6745104327433805, + "grad_norm": 1.8202465772628784, + "learning_rate": 1.544393993676154e-05, + "loss": 3.2829, + "step": 8413 + }, + { + "epoch": 0.6745906075244033, + "grad_norm": 2.0734643936157227, + "learning_rate": 1.544285071498467e-05, + "loss": 3.2449, + "step": 8414 + }, + { + "epoch": 0.6746707823054259, + "grad_norm": 1.8579071760177612, + "learning_rate": 1.5441761401443247e-05, + "loss": 3.2473, + "step": 8415 + }, + { + "epoch": 0.6747509570864485, + "grad_norm": 1.9364598989486694, + "learning_rate": 1.544067199615565e-05, + "loss": 3.2509, + "step": 8416 + }, + { + "epoch": 0.6748311318674711, + "grad_norm": 1.865004062652588, + "learning_rate": 1.543958249914024e-05, + "loss": 3.2336, + "step": 8417 + }, + { + "epoch": 0.6749113066484937, + "grad_norm": 1.8872027397155762, + "learning_rate": 1.5438492910415385e-05, + "loss": 3.2358, + "step": 8418 + }, + { + "epoch": 0.6749914814295164, + "grad_norm": 1.9347999095916748, + "learning_rate": 1.5437403229999458e-05, + "loss": 3.3754, + "step": 8419 + }, + { + "epoch": 0.675071656210539, + "grad_norm": 1.913774013519287, + "learning_rate": 1.5436313457910825e-05, + "loss": 3.204, + "step": 8420 + }, + { + "epoch": 0.6751518309915616, + "grad_norm": 2.0003015995025635, + "learning_rate": 1.5435223594167867e-05, + "loss": 3.237, + "step": 8421 + }, + { + "epoch": 0.6752320057725842, + "grad_norm": 0.7497463822364807, + "learning_rate": 1.5434133638788954e-05, + "loss": 2.4879, + "step": 8422 + }, + { + "epoch": 0.6753121805536069, + "grad_norm": 1.9389166831970215, + "learning_rate": 1.5433043591792464e-05, + "loss": 3.0782, + "step": 8423 + }, + { + "epoch": 0.6753923553346295, + "grad_norm": 1.840843677520752, + "learning_rate": 1.5431953453196774e-05, + "loss": 3.1997, + "step": 8424 + }, + { + "epoch": 0.6754725301156521, + "grad_norm": 0.7299753427505493, + "learning_rate": 1.543086322302026e-05, + "loss": 2.4862, + "step": 8425 + }, + { + "epoch": 0.6755527048966747, + "grad_norm": 1.7724049091339111, + "learning_rate": 1.5429772901281307e-05, + "loss": 3.1402, + "step": 8426 + }, + { + "epoch": 0.6756328796776974, + "grad_norm": 1.8454030752182007, + "learning_rate": 1.5428682487998298e-05, + "loss": 3.1777, + "step": 8427 + }, + { + "epoch": 0.67571305445872, + "grad_norm": 1.725353717803955, + "learning_rate": 1.5427591983189617e-05, + "loss": 3.0726, + "step": 8428 + }, + { + "epoch": 0.6757932292397426, + "grad_norm": 1.8215528726577759, + "learning_rate": 1.5426501386873644e-05, + "loss": 3.1881, + "step": 8429 + }, + { + "epoch": 0.6758734040207652, + "grad_norm": 1.781578540802002, + "learning_rate": 1.5425410699068774e-05, + "loss": 3.2526, + "step": 8430 + }, + { + "epoch": 0.6759535788017879, + "grad_norm": 1.8566380739212036, + "learning_rate": 1.542431991979339e-05, + "loss": 3.2271, + "step": 8431 + }, + { + "epoch": 0.6760337535828105, + "grad_norm": 1.7597837448120117, + "learning_rate": 1.542322904906588e-05, + "loss": 3.179, + "step": 8432 + }, + { + "epoch": 0.6761139283638331, + "grad_norm": 1.7165336608886719, + "learning_rate": 1.5422138086904642e-05, + "loss": 3.166, + "step": 8433 + }, + { + "epoch": 0.6761941031448558, + "grad_norm": 1.7679939270019531, + "learning_rate": 1.542104703332807e-05, + "loss": 3.2451, + "step": 8434 + }, + { + "epoch": 0.6762742779258785, + "grad_norm": 1.8310315608978271, + "learning_rate": 1.5419955888354553e-05, + "loss": 3.3597, + "step": 8435 + }, + { + "epoch": 0.6763544527069011, + "grad_norm": 1.920375943183899, + "learning_rate": 1.541886465200249e-05, + "loss": 3.187, + "step": 8436 + }, + { + "epoch": 0.6764346274879237, + "grad_norm": 1.8245631456375122, + "learning_rate": 1.541777332429028e-05, + "loss": 3.1935, + "step": 8437 + }, + { + "epoch": 0.6765148022689463, + "grad_norm": 1.7421451807022095, + "learning_rate": 1.541668190523632e-05, + "loss": 3.2375, + "step": 8438 + }, + { + "epoch": 0.676594977049969, + "grad_norm": 1.847930669784546, + "learning_rate": 1.5415590394859012e-05, + "loss": 3.2054, + "step": 8439 + }, + { + "epoch": 0.6766751518309916, + "grad_norm": 1.88258957862854, + "learning_rate": 1.5414498793176757e-05, + "loss": 3.2064, + "step": 8440 + }, + { + "epoch": 0.6767553266120142, + "grad_norm": 1.9024723768234253, + "learning_rate": 1.5413407100207965e-05, + "loss": 3.1243, + "step": 8441 + }, + { + "epoch": 0.6768355013930368, + "grad_norm": 1.757949709892273, + "learning_rate": 1.5412315315971037e-05, + "loss": 3.1776, + "step": 8442 + }, + { + "epoch": 0.6769156761740595, + "grad_norm": 1.855979323387146, + "learning_rate": 1.5411223440484375e-05, + "loss": 3.1179, + "step": 8443 + }, + { + "epoch": 0.6769958509550821, + "grad_norm": 1.9253426790237427, + "learning_rate": 1.54101314737664e-05, + "loss": 3.2382, + "step": 8444 + }, + { + "epoch": 0.6770760257361047, + "grad_norm": 1.7852964401245117, + "learning_rate": 1.540903941583551e-05, + "loss": 3.254, + "step": 8445 + }, + { + "epoch": 0.6771562005171273, + "grad_norm": 1.8793525695800781, + "learning_rate": 1.540794726671012e-05, + "loss": 3.2279, + "step": 8446 + }, + { + "epoch": 0.67723637529815, + "grad_norm": 1.909580945968628, + "learning_rate": 1.5406855026408648e-05, + "loss": 3.2551, + "step": 8447 + }, + { + "epoch": 0.6773165500791726, + "grad_norm": 1.7587027549743652, + "learning_rate": 1.5405762694949507e-05, + "loss": 3.2859, + "step": 8448 + }, + { + "epoch": 0.6773967248601952, + "grad_norm": 2.0132663249969482, + "learning_rate": 1.540467027235111e-05, + "loss": 3.1956, + "step": 8449 + }, + { + "epoch": 0.6774768996412178, + "grad_norm": 1.8971399068832397, + "learning_rate": 1.540357775863188e-05, + "loss": 3.1331, + "step": 8450 + }, + { + "epoch": 0.6775570744222404, + "grad_norm": 1.8235046863555908, + "learning_rate": 1.540248515381023e-05, + "loss": 3.2291, + "step": 8451 + }, + { + "epoch": 0.6776372492032631, + "grad_norm": 1.781803011894226, + "learning_rate": 1.5401392457904584e-05, + "loss": 3.2349, + "step": 8452 + }, + { + "epoch": 0.6777174239842857, + "grad_norm": 1.7945059537887573, + "learning_rate": 1.5400299670933367e-05, + "loss": 3.1598, + "step": 8453 + }, + { + "epoch": 0.6777975987653083, + "grad_norm": 1.852609634399414, + "learning_rate": 1.5399206792915e-05, + "loss": 3.2708, + "step": 8454 + }, + { + "epoch": 0.677877773546331, + "grad_norm": 1.806335210800171, + "learning_rate": 1.539811382386791e-05, + "loss": 3.2736, + "step": 8455 + }, + { + "epoch": 0.6779579483273537, + "grad_norm": 1.8288788795471191, + "learning_rate": 1.5397020763810523e-05, + "loss": 3.3798, + "step": 8456 + }, + { + "epoch": 0.6780381231083763, + "grad_norm": 1.7223097085952759, + "learning_rate": 1.5395927612761266e-05, + "loss": 3.2231, + "step": 8457 + }, + { + "epoch": 0.6781182978893989, + "grad_norm": 2.541534185409546, + "learning_rate": 1.5394834370738572e-05, + "loss": 3.2651, + "step": 8458 + }, + { + "epoch": 0.6781984726704215, + "grad_norm": 1.7278904914855957, + "learning_rate": 1.5393741037760875e-05, + "loss": 3.1545, + "step": 8459 + }, + { + "epoch": 0.6782786474514442, + "grad_norm": 1.9466370344161987, + "learning_rate": 1.5392647613846604e-05, + "loss": 3.2449, + "step": 8460 + }, + { + "epoch": 0.6783588222324668, + "grad_norm": 1.760095477104187, + "learning_rate": 1.5391554099014192e-05, + "loss": 3.183, + "step": 8461 + }, + { + "epoch": 0.6784389970134894, + "grad_norm": 1.8523492813110352, + "learning_rate": 1.5390460493282076e-05, + "loss": 3.2076, + "step": 8462 + }, + { + "epoch": 0.678519171794512, + "grad_norm": 1.7882977724075317, + "learning_rate": 1.5389366796668703e-05, + "loss": 3.0961, + "step": 8463 + }, + { + "epoch": 0.6785993465755347, + "grad_norm": 0.9701706767082214, + "learning_rate": 1.53882730091925e-05, + "loss": 2.4996, + "step": 8464 + }, + { + "epoch": 0.6786795213565573, + "grad_norm": 1.8497432470321655, + "learning_rate": 1.5387179130871916e-05, + "loss": 3.2104, + "step": 8465 + }, + { + "epoch": 0.6787596961375799, + "grad_norm": 1.8836908340454102, + "learning_rate": 1.538608516172539e-05, + "loss": 3.2054, + "step": 8466 + }, + { + "epoch": 0.6788398709186025, + "grad_norm": 1.7588016986846924, + "learning_rate": 1.5384991101771366e-05, + "loss": 3.1842, + "step": 8467 + }, + { + "epoch": 0.6789200456996252, + "grad_norm": 0.7332153916358948, + "learning_rate": 1.538389695102829e-05, + "loss": 2.4883, + "step": 8468 + }, + { + "epoch": 0.6790002204806478, + "grad_norm": 2.1154637336730957, + "learning_rate": 1.5382802709514607e-05, + "loss": 3.2256, + "step": 8469 + }, + { + "epoch": 0.6790803952616704, + "grad_norm": 1.7692713737487793, + "learning_rate": 1.538170837724877e-05, + "loss": 3.2388, + "step": 8470 + }, + { + "epoch": 0.679160570042693, + "grad_norm": 2.0182058811187744, + "learning_rate": 1.5380613954249228e-05, + "loss": 3.3304, + "step": 8471 + }, + { + "epoch": 0.6792407448237157, + "grad_norm": 1.896222472190857, + "learning_rate": 1.5379519440534422e-05, + "loss": 3.2153, + "step": 8472 + }, + { + "epoch": 0.6793209196047383, + "grad_norm": 1.7802444696426392, + "learning_rate": 1.537842483612282e-05, + "loss": 3.1445, + "step": 8473 + }, + { + "epoch": 0.6794010943857609, + "grad_norm": 1.92279851436615, + "learning_rate": 1.5377330141032874e-05, + "loss": 3.1254, + "step": 8474 + }, + { + "epoch": 0.6794812691667836, + "grad_norm": 1.8740618228912354, + "learning_rate": 1.5376235355283036e-05, + "loss": 3.2548, + "step": 8475 + }, + { + "epoch": 0.6795614439478063, + "grad_norm": 1.8999912738800049, + "learning_rate": 1.537514047889176e-05, + "loss": 3.2689, + "step": 8476 + }, + { + "epoch": 0.6796416187288289, + "grad_norm": 0.7495184540748596, + "learning_rate": 1.5374045511877516e-05, + "loss": 2.5043, + "step": 8477 + }, + { + "epoch": 0.6797217935098515, + "grad_norm": 1.8450379371643066, + "learning_rate": 1.5372950454258755e-05, + "loss": 3.1693, + "step": 8478 + }, + { + "epoch": 0.6798019682908741, + "grad_norm": 1.9793380498886108, + "learning_rate": 1.5371855306053944e-05, + "loss": 3.2391, + "step": 8479 + }, + { + "epoch": 0.6798821430718968, + "grad_norm": 1.8008863925933838, + "learning_rate": 1.5370760067281544e-05, + "loss": 3.1309, + "step": 8480 + }, + { + "epoch": 0.6799623178529194, + "grad_norm": 2.110865592956543, + "learning_rate": 1.5369664737960024e-05, + "loss": 3.2166, + "step": 8481 + }, + { + "epoch": 0.680042492633942, + "grad_norm": 1.8063126802444458, + "learning_rate": 1.5368569318107847e-05, + "loss": 3.3074, + "step": 8482 + }, + { + "epoch": 0.6801226674149646, + "grad_norm": 1.7566508054733276, + "learning_rate": 1.5367473807743488e-05, + "loss": 3.3019, + "step": 8483 + }, + { + "epoch": 0.6802028421959873, + "grad_norm": 1.8250254392623901, + "learning_rate": 1.5366378206885408e-05, + "loss": 3.2238, + "step": 8484 + }, + { + "epoch": 0.6802830169770099, + "grad_norm": 1.8801740407943726, + "learning_rate": 1.5365282515552086e-05, + "loss": 3.1772, + "step": 8485 + }, + { + "epoch": 0.6803631917580325, + "grad_norm": 1.8980231285095215, + "learning_rate": 1.536418673376199e-05, + "loss": 3.2506, + "step": 8486 + }, + { + "epoch": 0.6804433665390551, + "grad_norm": 1.9113308191299438, + "learning_rate": 1.536309086153359e-05, + "loss": 3.1595, + "step": 8487 + }, + { + "epoch": 0.6805235413200778, + "grad_norm": 1.8825029134750366, + "learning_rate": 1.5361994898885377e-05, + "loss": 3.2549, + "step": 8488 + }, + { + "epoch": 0.6806037161011004, + "grad_norm": 1.8280596733093262, + "learning_rate": 1.5360898845835817e-05, + "loss": 3.2437, + "step": 8489 + }, + { + "epoch": 0.680683890882123, + "grad_norm": 1.7853041887283325, + "learning_rate": 1.535980270240339e-05, + "loss": 3.3052, + "step": 8490 + }, + { + "epoch": 0.6807640656631456, + "grad_norm": 1.7854113578796387, + "learning_rate": 1.535870646860658e-05, + "loss": 3.2228, + "step": 8491 + }, + { + "epoch": 0.6808442404441682, + "grad_norm": 1.8220934867858887, + "learning_rate": 1.535761014446387e-05, + "loss": 3.1688, + "step": 8492 + }, + { + "epoch": 0.6809244152251909, + "grad_norm": 1.7563250064849854, + "learning_rate": 1.5356513729993736e-05, + "loss": 3.2208, + "step": 8493 + }, + { + "epoch": 0.6810045900062135, + "grad_norm": 1.8319261074066162, + "learning_rate": 1.5355417225214672e-05, + "loss": 3.1962, + "step": 8494 + }, + { + "epoch": 0.6810847647872362, + "grad_norm": 1.9874736070632935, + "learning_rate": 1.535432063014516e-05, + "loss": 3.1777, + "step": 8495 + }, + { + "epoch": 0.6811649395682589, + "grad_norm": 1.786865472793579, + "learning_rate": 1.5353223944803687e-05, + "loss": 3.2737, + "step": 8496 + }, + { + "epoch": 0.6812451143492815, + "grad_norm": 1.7248226404190063, + "learning_rate": 1.535212716920875e-05, + "loss": 3.2831, + "step": 8497 + }, + { + "epoch": 0.6813252891303041, + "grad_norm": 1.9857133626937866, + "learning_rate": 1.5351030303378827e-05, + "loss": 3.1431, + "step": 8498 + }, + { + "epoch": 0.6814054639113267, + "grad_norm": 1.9597384929656982, + "learning_rate": 1.5349933347332427e-05, + "loss": 3.2502, + "step": 8499 + }, + { + "epoch": 0.6814856386923493, + "grad_norm": 1.9208579063415527, + "learning_rate": 1.5348836301088035e-05, + "loss": 3.176, + "step": 8500 + }, + { + "epoch": 0.681565813473372, + "grad_norm": 1.8181345462799072, + "learning_rate": 1.5347739164664142e-05, + "loss": 3.2644, + "step": 8501 + }, + { + "epoch": 0.6816459882543946, + "grad_norm": 1.821931004524231, + "learning_rate": 1.534664193807926e-05, + "loss": 3.182, + "step": 8502 + }, + { + "epoch": 0.6817261630354172, + "grad_norm": 1.7936612367630005, + "learning_rate": 1.5345544621351875e-05, + "loss": 3.2451, + "step": 8503 + }, + { + "epoch": 0.6818063378164398, + "grad_norm": 0.8009708523750305, + "learning_rate": 1.534444721450049e-05, + "loss": 2.5015, + "step": 8504 + }, + { + "epoch": 0.6818865125974625, + "grad_norm": 1.9101872444152832, + "learning_rate": 1.5343349717543616e-05, + "loss": 3.2178, + "step": 8505 + }, + { + "epoch": 0.6819666873784851, + "grad_norm": 1.7294456958770752, + "learning_rate": 1.534225213049974e-05, + "loss": 3.2, + "step": 8506 + }, + { + "epoch": 0.6820468621595077, + "grad_norm": 1.8967348337173462, + "learning_rate": 1.5341154453387377e-05, + "loss": 3.2333, + "step": 8507 + }, + { + "epoch": 0.6821270369405303, + "grad_norm": 1.8481980562210083, + "learning_rate": 1.534005668622504e-05, + "loss": 3.1636, + "step": 8508 + }, + { + "epoch": 0.682207211721553, + "grad_norm": 1.7948036193847656, + "learning_rate": 1.5338958829031222e-05, + "loss": 3.2462, + "step": 8509 + }, + { + "epoch": 0.6822873865025756, + "grad_norm": 1.8706140518188477, + "learning_rate": 1.533786088182444e-05, + "loss": 3.198, + "step": 8510 + }, + { + "epoch": 0.6823675612835982, + "grad_norm": 1.9945948123931885, + "learning_rate": 1.5336762844623208e-05, + "loss": 3.161, + "step": 8511 + }, + { + "epoch": 0.6824477360646208, + "grad_norm": 1.809840202331543, + "learning_rate": 1.533566471744603e-05, + "loss": 3.2041, + "step": 8512 + }, + { + "epoch": 0.6825279108456435, + "grad_norm": 1.9249975681304932, + "learning_rate": 1.533456650031143e-05, + "loss": 3.252, + "step": 8513 + }, + { + "epoch": 0.6826080856266661, + "grad_norm": 1.7767739295959473, + "learning_rate": 1.533346819323792e-05, + "loss": 3.2109, + "step": 8514 + }, + { + "epoch": 0.6826882604076887, + "grad_norm": 1.8569698333740234, + "learning_rate": 1.5332369796244012e-05, + "loss": 3.2749, + "step": 8515 + }, + { + "epoch": 0.6827684351887114, + "grad_norm": 1.7572259902954102, + "learning_rate": 1.533127130934823e-05, + "loss": 3.1035, + "step": 8516 + }, + { + "epoch": 0.6828486099697341, + "grad_norm": 1.7273764610290527, + "learning_rate": 1.5330172732569093e-05, + "loss": 3.2864, + "step": 8517 + }, + { + "epoch": 0.6829287847507567, + "grad_norm": 2.0083303451538086, + "learning_rate": 1.5329074065925124e-05, + "loss": 3.2853, + "step": 8518 + }, + { + "epoch": 0.6830089595317793, + "grad_norm": 1.7773029804229736, + "learning_rate": 1.5327975309434842e-05, + "loss": 3.2666, + "step": 8519 + }, + { + "epoch": 0.6830891343128019, + "grad_norm": 1.8681484460830688, + "learning_rate": 1.5326876463116777e-05, + "loss": 3.152, + "step": 8520 + }, + { + "epoch": 0.6831693090938246, + "grad_norm": 1.8139489889144897, + "learning_rate": 1.5325777526989448e-05, + "loss": 3.2529, + "step": 8521 + }, + { + "epoch": 0.6832494838748472, + "grad_norm": 2.0417356491088867, + "learning_rate": 1.5324678501071385e-05, + "loss": 3.2645, + "step": 8522 + }, + { + "epoch": 0.6833296586558698, + "grad_norm": 1.7842994928359985, + "learning_rate": 1.5323579385381125e-05, + "loss": 3.1448, + "step": 8523 + }, + { + "epoch": 0.6834098334368924, + "grad_norm": 2.131061315536499, + "learning_rate": 1.532248017993719e-05, + "loss": 3.2469, + "step": 8524 + }, + { + "epoch": 0.6834900082179151, + "grad_norm": 1.701701045036316, + "learning_rate": 1.5321380884758118e-05, + "loss": 3.161, + "step": 8525 + }, + { + "epoch": 0.6835701829989377, + "grad_norm": 2.006826877593994, + "learning_rate": 1.5320281499862438e-05, + "loss": 3.1895, + "step": 8526 + }, + { + "epoch": 0.6836503577799603, + "grad_norm": 0.7554133534431458, + "learning_rate": 1.5319182025268685e-05, + "loss": 2.4786, + "step": 8527 + }, + { + "epoch": 0.6837305325609829, + "grad_norm": 1.8480987548828125, + "learning_rate": 1.53180824609954e-05, + "loss": 3.2963, + "step": 8528 + }, + { + "epoch": 0.6838107073420056, + "grad_norm": 0.7231072783470154, + "learning_rate": 1.5316982807061117e-05, + "loss": 2.4796, + "step": 8529 + }, + { + "epoch": 0.6838908821230282, + "grad_norm": 0.7236433029174805, + "learning_rate": 1.531588306348438e-05, + "loss": 2.4954, + "step": 8530 + }, + { + "epoch": 0.6839710569040508, + "grad_norm": 1.8128780126571655, + "learning_rate": 1.5314783230283726e-05, + "loss": 3.1943, + "step": 8531 + }, + { + "epoch": 0.6840512316850734, + "grad_norm": 1.8542218208312988, + "learning_rate": 1.53136833074777e-05, + "loss": 3.2496, + "step": 8532 + }, + { + "epoch": 0.684131406466096, + "grad_norm": 1.7691508531570435, + "learning_rate": 1.531258329508485e-05, + "loss": 3.1409, + "step": 8533 + }, + { + "epoch": 0.6842115812471187, + "grad_norm": 1.8496081829071045, + "learning_rate": 1.5311483193123713e-05, + "loss": 3.2286, + "step": 8534 + }, + { + "epoch": 0.6842917560281413, + "grad_norm": 1.8129518032073975, + "learning_rate": 1.5310383001612844e-05, + "loss": 3.3265, + "step": 8535 + }, + { + "epoch": 0.684371930809164, + "grad_norm": 0.7889212369918823, + "learning_rate": 1.5309282720570786e-05, + "loss": 2.4773, + "step": 8536 + }, + { + "epoch": 0.6844521055901867, + "grad_norm": 1.9329651594161987, + "learning_rate": 1.5308182350016095e-05, + "loss": 3.2655, + "step": 8537 + }, + { + "epoch": 0.6845322803712093, + "grad_norm": 1.7568799257278442, + "learning_rate": 1.5307081889967322e-05, + "loss": 3.1029, + "step": 8538 + }, + { + "epoch": 0.6846124551522319, + "grad_norm": 1.8065587282180786, + "learning_rate": 1.5305981340443016e-05, + "loss": 3.2668, + "step": 8539 + }, + { + "epoch": 0.6846926299332545, + "grad_norm": 1.7538012266159058, + "learning_rate": 1.5304880701461735e-05, + "loss": 3.2703, + "step": 8540 + }, + { + "epoch": 0.6847728047142772, + "grad_norm": 1.8928667306900024, + "learning_rate": 1.5303779973042037e-05, + "loss": 3.2603, + "step": 8541 + }, + { + "epoch": 0.6848529794952998, + "grad_norm": 1.8298406600952148, + "learning_rate": 1.5302679155202475e-05, + "loss": 3.3047, + "step": 8542 + }, + { + "epoch": 0.6849331542763224, + "grad_norm": 1.722741723060608, + "learning_rate": 1.5301578247961613e-05, + "loss": 3.1968, + "step": 8543 + }, + { + "epoch": 0.685013329057345, + "grad_norm": 1.8482476472854614, + "learning_rate": 1.530047725133801e-05, + "loss": 3.327, + "step": 8544 + }, + { + "epoch": 0.6850935038383676, + "grad_norm": 1.8574012517929077, + "learning_rate": 1.529937616535023e-05, + "loss": 3.1613, + "step": 8545 + }, + { + "epoch": 0.6851736786193903, + "grad_norm": 1.8075997829437256, + "learning_rate": 1.5298274990016833e-05, + "loss": 3.1764, + "step": 8546 + }, + { + "epoch": 0.6852538534004129, + "grad_norm": 1.7939614057540894, + "learning_rate": 1.5297173725356386e-05, + "loss": 3.2188, + "step": 8547 + }, + { + "epoch": 0.6853340281814355, + "grad_norm": 1.8674396276474, + "learning_rate": 1.529607237138746e-05, + "loss": 3.1651, + "step": 8548 + }, + { + "epoch": 0.6854142029624581, + "grad_norm": 1.7019813060760498, + "learning_rate": 1.529497092812862e-05, + "loss": 3.2404, + "step": 8549 + }, + { + "epoch": 0.6854943777434808, + "grad_norm": 1.8564881086349487, + "learning_rate": 1.5293869395598435e-05, + "loss": 3.134, + "step": 8550 + }, + { + "epoch": 0.6855745525245034, + "grad_norm": 1.8025226593017578, + "learning_rate": 1.529276777381548e-05, + "loss": 3.205, + "step": 8551 + }, + { + "epoch": 0.685654727305526, + "grad_norm": 1.7919292449951172, + "learning_rate": 1.5291666062798323e-05, + "loss": 3.1004, + "step": 8552 + }, + { + "epoch": 0.6857349020865486, + "grad_norm": 1.786090612411499, + "learning_rate": 1.5290564262565544e-05, + "loss": 3.2324, + "step": 8553 + }, + { + "epoch": 0.6858150768675713, + "grad_norm": 1.7821954488754272, + "learning_rate": 1.5289462373135712e-05, + "loss": 3.246, + "step": 8554 + }, + { + "epoch": 0.6858952516485939, + "grad_norm": 1.807618260383606, + "learning_rate": 1.5288360394527413e-05, + "loss": 3.2387, + "step": 8555 + }, + { + "epoch": 0.6859754264296166, + "grad_norm": 1.8954769372940063, + "learning_rate": 1.528725832675922e-05, + "loss": 3.3465, + "step": 8556 + }, + { + "epoch": 0.6860556012106392, + "grad_norm": 1.853709101676941, + "learning_rate": 1.5286156169849712e-05, + "loss": 3.2947, + "step": 8557 + }, + { + "epoch": 0.6861357759916619, + "grad_norm": 1.8258304595947266, + "learning_rate": 1.5285053923817475e-05, + "loss": 3.2709, + "step": 8558 + }, + { + "epoch": 0.6862159507726845, + "grad_norm": 1.80416738986969, + "learning_rate": 1.5283951588681092e-05, + "loss": 3.2037, + "step": 8559 + }, + { + "epoch": 0.6862961255537071, + "grad_norm": 1.7734204530715942, + "learning_rate": 1.528284916445915e-05, + "loss": 3.2295, + "step": 8560 + }, + { + "epoch": 0.6863763003347297, + "grad_norm": 1.949150562286377, + "learning_rate": 1.528174665117023e-05, + "loss": 3.3109, + "step": 8561 + }, + { + "epoch": 0.6864564751157524, + "grad_norm": 0.7498766779899597, + "learning_rate": 1.5280644048832924e-05, + "loss": 2.5026, + "step": 8562 + }, + { + "epoch": 0.686536649896775, + "grad_norm": 1.7326226234436035, + "learning_rate": 1.5279541357465824e-05, + "loss": 3.1254, + "step": 8563 + }, + { + "epoch": 0.6866168246777976, + "grad_norm": 1.8957191705703735, + "learning_rate": 1.527843857708751e-05, + "loss": 3.208, + "step": 8564 + }, + { + "epoch": 0.6866969994588202, + "grad_norm": 1.7052714824676514, + "learning_rate": 1.527733570771659e-05, + "loss": 3.2489, + "step": 8565 + }, + { + "epoch": 0.6867771742398429, + "grad_norm": 1.8880481719970703, + "learning_rate": 1.5276232749371643e-05, + "loss": 3.306, + "step": 8566 + }, + { + "epoch": 0.6868573490208655, + "grad_norm": 1.779893159866333, + "learning_rate": 1.5275129702071273e-05, + "loss": 3.1816, + "step": 8567 + }, + { + "epoch": 0.6869375238018881, + "grad_norm": 1.7859362363815308, + "learning_rate": 1.5274026565834076e-05, + "loss": 3.2402, + "step": 8568 + }, + { + "epoch": 0.6870176985829107, + "grad_norm": 1.8376396894454956, + "learning_rate": 1.527292334067865e-05, + "loss": 3.2059, + "step": 8569 + }, + { + "epoch": 0.6870978733639334, + "grad_norm": 1.8349710702896118, + "learning_rate": 1.5271820026623597e-05, + "loss": 3.2086, + "step": 8570 + }, + { + "epoch": 0.687178048144956, + "grad_norm": 1.939855694770813, + "learning_rate": 1.5270716623687513e-05, + "loss": 3.3431, + "step": 8571 + }, + { + "epoch": 0.6872582229259786, + "grad_norm": 1.7399142980575562, + "learning_rate": 1.5269613131889006e-05, + "loss": 3.1785, + "step": 8572 + }, + { + "epoch": 0.6873383977070012, + "grad_norm": 1.8299826383590698, + "learning_rate": 1.526850955124668e-05, + "loss": 3.2169, + "step": 8573 + }, + { + "epoch": 0.6874185724880238, + "grad_norm": 1.784509539604187, + "learning_rate": 1.526740588177914e-05, + "loss": 3.3109, + "step": 8574 + }, + { + "epoch": 0.6874987472690465, + "grad_norm": 1.8409767150878906, + "learning_rate": 1.5266302123504992e-05, + "loss": 3.2584, + "step": 8575 + }, + { + "epoch": 0.6875789220500691, + "grad_norm": 1.828791618347168, + "learning_rate": 1.526519827644284e-05, + "loss": 3.1062, + "step": 8576 + }, + { + "epoch": 0.6876590968310918, + "grad_norm": 1.7645199298858643, + "learning_rate": 1.526409434061131e-05, + "loss": 3.1501, + "step": 8577 + }, + { + "epoch": 0.6877392716121145, + "grad_norm": 1.7885069847106934, + "learning_rate": 1.5262990316029004e-05, + "loss": 3.2507, + "step": 8578 + }, + { + "epoch": 0.6878194463931371, + "grad_norm": 1.9696059226989746, + "learning_rate": 1.5261886202714533e-05, + "loss": 3.3138, + "step": 8579 + }, + { + "epoch": 0.6878996211741597, + "grad_norm": 1.794478178024292, + "learning_rate": 1.5260782000686516e-05, + "loss": 3.217, + "step": 8580 + }, + { + "epoch": 0.6879797959551823, + "grad_norm": 1.770113229751587, + "learning_rate": 1.5259677709963564e-05, + "loss": 3.1538, + "step": 8581 + }, + { + "epoch": 0.688059970736205, + "grad_norm": 1.8346145153045654, + "learning_rate": 1.5258573330564304e-05, + "loss": 3.1831, + "step": 8582 + }, + { + "epoch": 0.6881401455172276, + "grad_norm": 1.7833492755889893, + "learning_rate": 1.5257468862507352e-05, + "loss": 3.2407, + "step": 8583 + }, + { + "epoch": 0.6882203202982502, + "grad_norm": 1.7746936082839966, + "learning_rate": 1.5256364305811327e-05, + "loss": 3.2585, + "step": 8584 + }, + { + "epoch": 0.6883004950792728, + "grad_norm": 1.8516751527786255, + "learning_rate": 1.525525966049485e-05, + "loss": 3.2506, + "step": 8585 + }, + { + "epoch": 0.6883806698602954, + "grad_norm": 1.8923194408416748, + "learning_rate": 1.525415492657655e-05, + "loss": 3.207, + "step": 8586 + }, + { + "epoch": 0.6884608446413181, + "grad_norm": 1.739174485206604, + "learning_rate": 1.5253050104075046e-05, + "loss": 3.2439, + "step": 8587 + }, + { + "epoch": 0.6885410194223407, + "grad_norm": 1.7715731859207153, + "learning_rate": 1.5251945193008975e-05, + "loss": 3.3406, + "step": 8588 + }, + { + "epoch": 0.6886211942033633, + "grad_norm": 1.7026807069778442, + "learning_rate": 1.5250840193396954e-05, + "loss": 3.2635, + "step": 8589 + }, + { + "epoch": 0.6887013689843859, + "grad_norm": 1.8098686933517456, + "learning_rate": 1.524973510525762e-05, + "loss": 3.1569, + "step": 8590 + }, + { + "epoch": 0.6887815437654086, + "grad_norm": 1.8371068239212036, + "learning_rate": 1.5248629928609599e-05, + "loss": 3.2201, + "step": 8591 + }, + { + "epoch": 0.6888617185464312, + "grad_norm": 1.8055462837219238, + "learning_rate": 1.524752466347153e-05, + "loss": 3.1833, + "step": 8592 + }, + { + "epoch": 0.6889418933274538, + "grad_norm": 1.7484029531478882, + "learning_rate": 1.5246419309862046e-05, + "loss": 3.1598, + "step": 8593 + }, + { + "epoch": 0.6890220681084764, + "grad_norm": 1.7853893041610718, + "learning_rate": 1.5245313867799776e-05, + "loss": 3.2793, + "step": 8594 + }, + { + "epoch": 0.6891022428894991, + "grad_norm": 1.8350929021835327, + "learning_rate": 1.524420833730337e-05, + "loss": 3.3252, + "step": 8595 + }, + { + "epoch": 0.6891824176705217, + "grad_norm": 0.7945629954338074, + "learning_rate": 1.5243102718391453e-05, + "loss": 2.4897, + "step": 8596 + }, + { + "epoch": 0.6892625924515444, + "grad_norm": 2.0238687992095947, + "learning_rate": 1.5241997011082675e-05, + "loss": 3.3109, + "step": 8597 + }, + { + "epoch": 0.689342767232567, + "grad_norm": 0.7389070987701416, + "learning_rate": 1.5240891215395676e-05, + "loss": 2.4963, + "step": 8598 + }, + { + "epoch": 0.6894229420135897, + "grad_norm": 1.7250932455062866, + "learning_rate": 1.5239785331349097e-05, + "loss": 3.2076, + "step": 8599 + }, + { + "epoch": 0.6895031167946123, + "grad_norm": 1.669585943222046, + "learning_rate": 1.5238679358961585e-05, + "loss": 3.0525, + "step": 8600 + }, + { + "epoch": 0.6895832915756349, + "grad_norm": 1.7244696617126465, + "learning_rate": 1.5237573298251784e-05, + "loss": 3.2206, + "step": 8601 + }, + { + "epoch": 0.6896634663566575, + "grad_norm": 1.711678385734558, + "learning_rate": 1.5236467149238344e-05, + "loss": 3.123, + "step": 8602 + }, + { + "epoch": 0.6897436411376802, + "grad_norm": 1.8058183193206787, + "learning_rate": 1.5235360911939912e-05, + "loss": 3.2626, + "step": 8603 + }, + { + "epoch": 0.6898238159187028, + "grad_norm": 1.761074423789978, + "learning_rate": 1.5234254586375141e-05, + "loss": 3.2717, + "step": 8604 + }, + { + "epoch": 0.6899039906997254, + "grad_norm": 1.6710312366485596, + "learning_rate": 1.5233148172562685e-05, + "loss": 3.166, + "step": 8605 + }, + { + "epoch": 0.689984165480748, + "grad_norm": 1.8478935956954956, + "learning_rate": 1.5232041670521191e-05, + "loss": 3.2465, + "step": 8606 + }, + { + "epoch": 0.6900643402617707, + "grad_norm": 1.8095495700836182, + "learning_rate": 1.523093508026932e-05, + "loss": 3.2963, + "step": 8607 + }, + { + "epoch": 0.6901445150427933, + "grad_norm": 1.7714972496032715, + "learning_rate": 1.5229828401825728e-05, + "loss": 3.2188, + "step": 8608 + }, + { + "epoch": 0.6902246898238159, + "grad_norm": 1.7529401779174805, + "learning_rate": 1.522872163520907e-05, + "loss": 3.2064, + "step": 8609 + }, + { + "epoch": 0.6903048646048385, + "grad_norm": 1.9392789602279663, + "learning_rate": 1.5227614780438012e-05, + "loss": 3.1584, + "step": 8610 + }, + { + "epoch": 0.6903850393858612, + "grad_norm": 1.839648962020874, + "learning_rate": 1.5226507837531206e-05, + "loss": 3.1348, + "step": 8611 + }, + { + "epoch": 0.6904652141668838, + "grad_norm": 1.839462399482727, + "learning_rate": 1.5225400806507322e-05, + "loss": 3.3209, + "step": 8612 + }, + { + "epoch": 0.6905453889479064, + "grad_norm": 1.7547903060913086, + "learning_rate": 1.5224293687385023e-05, + "loss": 3.1632, + "step": 8613 + }, + { + "epoch": 0.690625563728929, + "grad_norm": 0.8229368925094604, + "learning_rate": 1.5223186480182971e-05, + "loss": 2.5022, + "step": 8614 + }, + { + "epoch": 0.6907057385099516, + "grad_norm": 1.8665770292282104, + "learning_rate": 1.5222079184919842e-05, + "loss": 3.2583, + "step": 8615 + }, + { + "epoch": 0.6907859132909743, + "grad_norm": 1.8354249000549316, + "learning_rate": 1.5220971801614293e-05, + "loss": 3.2671, + "step": 8616 + }, + { + "epoch": 0.690866088071997, + "grad_norm": 1.897084355354309, + "learning_rate": 1.5219864330284998e-05, + "loss": 3.2648, + "step": 8617 + }, + { + "epoch": 0.6909462628530196, + "grad_norm": 1.6826082468032837, + "learning_rate": 1.5218756770950636e-05, + "loss": 3.2375, + "step": 8618 + }, + { + "epoch": 0.6910264376340423, + "grad_norm": 1.7664849758148193, + "learning_rate": 1.521764912362987e-05, + "loss": 3.1895, + "step": 8619 + }, + { + "epoch": 0.6911066124150649, + "grad_norm": 1.703192949295044, + "learning_rate": 1.5216541388341379e-05, + "loss": 3.2642, + "step": 8620 + }, + { + "epoch": 0.6911867871960875, + "grad_norm": 1.9776908159255981, + "learning_rate": 1.5215433565103838e-05, + "loss": 3.2485, + "step": 8621 + }, + { + "epoch": 0.6912669619771101, + "grad_norm": 1.7813730239868164, + "learning_rate": 1.5214325653935928e-05, + "loss": 3.2195, + "step": 8622 + }, + { + "epoch": 0.6913471367581328, + "grad_norm": 1.7375915050506592, + "learning_rate": 1.5213217654856323e-05, + "loss": 3.1285, + "step": 8623 + }, + { + "epoch": 0.6914273115391554, + "grad_norm": 1.8347973823547363, + "learning_rate": 1.5212109567883706e-05, + "loss": 3.167, + "step": 8624 + }, + { + "epoch": 0.691507486320178, + "grad_norm": 1.8034344911575317, + "learning_rate": 1.5211001393036759e-05, + "loss": 3.2064, + "step": 8625 + }, + { + "epoch": 0.6915876611012006, + "grad_norm": 1.9085975885391235, + "learning_rate": 1.5209893130334166e-05, + "loss": 3.2678, + "step": 8626 + }, + { + "epoch": 0.6916678358822232, + "grad_norm": 1.802018404006958, + "learning_rate": 1.5208784779794608e-05, + "loss": 3.2495, + "step": 8627 + }, + { + "epoch": 0.6917480106632459, + "grad_norm": 1.8484023809432983, + "learning_rate": 1.5207676341436778e-05, + "loss": 3.2413, + "step": 8628 + }, + { + "epoch": 0.6918281854442685, + "grad_norm": 1.8095616102218628, + "learning_rate": 1.5206567815279358e-05, + "loss": 3.1893, + "step": 8629 + }, + { + "epoch": 0.6919083602252911, + "grad_norm": 1.8253436088562012, + "learning_rate": 1.5205459201341043e-05, + "loss": 3.2415, + "step": 8630 + }, + { + "epoch": 0.6919885350063137, + "grad_norm": 1.7771188020706177, + "learning_rate": 1.5204350499640517e-05, + "loss": 3.2141, + "step": 8631 + }, + { + "epoch": 0.6920687097873364, + "grad_norm": 1.7516907453536987, + "learning_rate": 1.5203241710196475e-05, + "loss": 3.149, + "step": 8632 + }, + { + "epoch": 0.692148884568359, + "grad_norm": 1.9177038669586182, + "learning_rate": 1.5202132833027614e-05, + "loss": 3.284, + "step": 8633 + }, + { + "epoch": 0.6922290593493816, + "grad_norm": 1.783122181892395, + "learning_rate": 1.5201023868152623e-05, + "loss": 3.1892, + "step": 8634 + }, + { + "epoch": 0.6923092341304042, + "grad_norm": 1.811926245689392, + "learning_rate": 1.5199914815590208e-05, + "loss": 3.0833, + "step": 8635 + }, + { + "epoch": 0.6923894089114269, + "grad_norm": 1.8503470420837402, + "learning_rate": 1.5198805675359057e-05, + "loss": 3.2354, + "step": 8636 + }, + { + "epoch": 0.6924695836924495, + "grad_norm": 1.7371139526367188, + "learning_rate": 1.5197696447477876e-05, + "loss": 3.2246, + "step": 8637 + }, + { + "epoch": 0.6925497584734722, + "grad_norm": 1.7897330522537231, + "learning_rate": 1.5196587131965367e-05, + "loss": 3.1091, + "step": 8638 + }, + { + "epoch": 0.6926299332544948, + "grad_norm": 1.7765750885009766, + "learning_rate": 1.5195477728840229e-05, + "loss": 3.1604, + "step": 8639 + }, + { + "epoch": 0.6927101080355175, + "grad_norm": 0.7248010635375977, + "learning_rate": 1.5194368238121167e-05, + "loss": 2.4757, + "step": 8640 + }, + { + "epoch": 0.6927902828165401, + "grad_norm": 1.7824550867080688, + "learning_rate": 1.5193258659826888e-05, + "loss": 3.2489, + "step": 8641 + }, + { + "epoch": 0.6928704575975627, + "grad_norm": 1.8102200031280518, + "learning_rate": 1.5192148993976097e-05, + "loss": 3.2394, + "step": 8642 + }, + { + "epoch": 0.6929506323785853, + "grad_norm": 1.7306690216064453, + "learning_rate": 1.5191039240587505e-05, + "loss": 3.1507, + "step": 8643 + }, + { + "epoch": 0.693030807159608, + "grad_norm": 1.8214430809020996, + "learning_rate": 1.518992939967982e-05, + "loss": 3.311, + "step": 8644 + }, + { + "epoch": 0.6931109819406306, + "grad_norm": 1.7034621238708496, + "learning_rate": 1.5188819471271759e-05, + "loss": 3.2404, + "step": 8645 + }, + { + "epoch": 0.6931911567216532, + "grad_norm": 1.726853370666504, + "learning_rate": 1.5187709455382024e-05, + "loss": 3.3127, + "step": 8646 + }, + { + "epoch": 0.6932713315026758, + "grad_norm": 0.7281148433685303, + "learning_rate": 1.518659935202934e-05, + "loss": 2.5131, + "step": 8647 + }, + { + "epoch": 0.6933515062836985, + "grad_norm": 1.8663991689682007, + "learning_rate": 1.5185489161232417e-05, + "loss": 3.2508, + "step": 8648 + }, + { + "epoch": 0.6934316810647211, + "grad_norm": 1.827917218208313, + "learning_rate": 1.5184378883009973e-05, + "loss": 3.1207, + "step": 8649 + }, + { + "epoch": 0.6935118558457437, + "grad_norm": 1.816182017326355, + "learning_rate": 1.518326851738073e-05, + "loss": 3.3134, + "step": 8650 + }, + { + "epoch": 0.6935920306267663, + "grad_norm": 1.9011036157608032, + "learning_rate": 1.5182158064363406e-05, + "loss": 3.1359, + "step": 8651 + }, + { + "epoch": 0.693672205407789, + "grad_norm": 1.856218695640564, + "learning_rate": 1.5181047523976724e-05, + "loss": 3.1777, + "step": 8652 + }, + { + "epoch": 0.6937523801888116, + "grad_norm": 1.9037150144577026, + "learning_rate": 1.5179936896239407e-05, + "loss": 3.2551, + "step": 8653 + }, + { + "epoch": 0.6938325549698342, + "grad_norm": 1.743664264678955, + "learning_rate": 1.5178826181170178e-05, + "loss": 3.2784, + "step": 8654 + }, + { + "epoch": 0.6939127297508568, + "grad_norm": 1.7853480577468872, + "learning_rate": 1.5177715378787765e-05, + "loss": 3.2146, + "step": 8655 + }, + { + "epoch": 0.6939929045318795, + "grad_norm": 1.7040692567825317, + "learning_rate": 1.5176604489110897e-05, + "loss": 3.2369, + "step": 8656 + }, + { + "epoch": 0.6940730793129021, + "grad_norm": 1.725192904472351, + "learning_rate": 1.5175493512158302e-05, + "loss": 3.2153, + "step": 8657 + }, + { + "epoch": 0.6941532540939248, + "grad_norm": 1.7396891117095947, + "learning_rate": 1.517438244794871e-05, + "loss": 3.2487, + "step": 8658 + }, + { + "epoch": 0.6942334288749474, + "grad_norm": 1.7459207773208618, + "learning_rate": 1.517327129650085e-05, + "loss": 3.1968, + "step": 8659 + }, + { + "epoch": 0.6943136036559701, + "grad_norm": 1.739617943763733, + "learning_rate": 1.5172160057833462e-05, + "loss": 3.1434, + "step": 8660 + }, + { + "epoch": 0.6943937784369927, + "grad_norm": 1.7627116441726685, + "learning_rate": 1.5171048731965279e-05, + "loss": 3.1471, + "step": 8661 + }, + { + "epoch": 0.6944739532180153, + "grad_norm": 1.8555065393447876, + "learning_rate": 1.5169937318915038e-05, + "loss": 3.2294, + "step": 8662 + }, + { + "epoch": 0.6945541279990379, + "grad_norm": 1.6623167991638184, + "learning_rate": 1.5168825818701474e-05, + "loss": 3.1646, + "step": 8663 + }, + { + "epoch": 0.6946343027800606, + "grad_norm": 1.7156383991241455, + "learning_rate": 1.516771423134333e-05, + "loss": 3.1907, + "step": 8664 + }, + { + "epoch": 0.6947144775610832, + "grad_norm": 1.804311752319336, + "learning_rate": 1.5166602556859343e-05, + "loss": 3.1841, + "step": 8665 + }, + { + "epoch": 0.6947946523421058, + "grad_norm": 1.7151552438735962, + "learning_rate": 1.5165490795268257e-05, + "loss": 3.2189, + "step": 8666 + }, + { + "epoch": 0.6948748271231284, + "grad_norm": 1.7922394275665283, + "learning_rate": 1.5164378946588819e-05, + "loss": 3.3491, + "step": 8667 + }, + { + "epoch": 0.694955001904151, + "grad_norm": 1.8177019357681274, + "learning_rate": 1.5163267010839771e-05, + "loss": 3.2673, + "step": 8668 + }, + { + "epoch": 0.6950351766851737, + "grad_norm": 1.8138720989227295, + "learning_rate": 1.5162154988039861e-05, + "loss": 3.2532, + "step": 8669 + }, + { + "epoch": 0.6951153514661963, + "grad_norm": 1.697016954421997, + "learning_rate": 1.5161042878207837e-05, + "loss": 3.2412, + "step": 8670 + }, + { + "epoch": 0.6951955262472189, + "grad_norm": 0.7930327653884888, + "learning_rate": 1.5159930681362451e-05, + "loss": 2.485, + "step": 8671 + }, + { + "epoch": 0.6952757010282415, + "grad_norm": 1.7842575311660767, + "learning_rate": 1.5158818397522451e-05, + "loss": 3.2097, + "step": 8672 + }, + { + "epoch": 0.6953558758092642, + "grad_norm": 1.8360053300857544, + "learning_rate": 1.5157706026706592e-05, + "loss": 3.2094, + "step": 8673 + }, + { + "epoch": 0.6954360505902868, + "grad_norm": 0.7118562459945679, + "learning_rate": 1.5156593568933624e-05, + "loss": 2.4934, + "step": 8674 + }, + { + "epoch": 0.6955162253713094, + "grad_norm": 1.8505030870437622, + "learning_rate": 1.5155481024222307e-05, + "loss": 3.3057, + "step": 8675 + }, + { + "epoch": 0.695596400152332, + "grad_norm": 1.7090497016906738, + "learning_rate": 1.5154368392591398e-05, + "loss": 3.2042, + "step": 8676 + }, + { + "epoch": 0.6956765749333547, + "grad_norm": 1.8870453834533691, + "learning_rate": 1.5153255674059652e-05, + "loss": 3.1783, + "step": 8677 + }, + { + "epoch": 0.6957567497143773, + "grad_norm": 1.756421685218811, + "learning_rate": 1.5152142868645837e-05, + "loss": 3.2637, + "step": 8678 + }, + { + "epoch": 0.6958369244954, + "grad_norm": 1.85467529296875, + "learning_rate": 1.5151029976368702e-05, + "loss": 3.2472, + "step": 8679 + }, + { + "epoch": 0.6959170992764226, + "grad_norm": 1.7816146612167358, + "learning_rate": 1.5149916997247021e-05, + "loss": 3.1079, + "step": 8680 + }, + { + "epoch": 0.6959972740574453, + "grad_norm": 1.7891438007354736, + "learning_rate": 1.5148803931299555e-05, + "loss": 3.2761, + "step": 8681 + }, + { + "epoch": 0.6960774488384679, + "grad_norm": 1.9847239255905151, + "learning_rate": 1.5147690778545068e-05, + "loss": 3.1876, + "step": 8682 + }, + { + "epoch": 0.6961576236194905, + "grad_norm": 1.8000226020812988, + "learning_rate": 1.514657753900233e-05, + "loss": 3.1438, + "step": 8683 + }, + { + "epoch": 0.6962377984005131, + "grad_norm": 1.8651515245437622, + "learning_rate": 1.5145464212690106e-05, + "loss": 3.36, + "step": 8684 + }, + { + "epoch": 0.6963179731815358, + "grad_norm": 1.9722250699996948, + "learning_rate": 1.514435079962717e-05, + "loss": 3.249, + "step": 8685 + }, + { + "epoch": 0.6963981479625584, + "grad_norm": 1.917115330696106, + "learning_rate": 1.5143237299832295e-05, + "loss": 3.2181, + "step": 8686 + }, + { + "epoch": 0.696478322743581, + "grad_norm": 1.7477940320968628, + "learning_rate": 1.5142123713324248e-05, + "loss": 3.1593, + "step": 8687 + }, + { + "epoch": 0.6965584975246036, + "grad_norm": 1.8049389123916626, + "learning_rate": 1.5141010040121811e-05, + "loss": 3.3081, + "step": 8688 + }, + { + "epoch": 0.6966386723056263, + "grad_norm": 1.7524296045303345, + "learning_rate": 1.5139896280243754e-05, + "loss": 3.1367, + "step": 8689 + }, + { + "epoch": 0.6967188470866489, + "grad_norm": 1.7010968923568726, + "learning_rate": 1.5138782433708856e-05, + "loss": 3.2144, + "step": 8690 + }, + { + "epoch": 0.6967990218676715, + "grad_norm": 1.7458733320236206, + "learning_rate": 1.51376685005359e-05, + "loss": 3.11, + "step": 8691 + }, + { + "epoch": 0.6968791966486941, + "grad_norm": 1.79578697681427, + "learning_rate": 1.5136554480743662e-05, + "loss": 3.2157, + "step": 8692 + }, + { + "epoch": 0.6969593714297168, + "grad_norm": 1.7257288694381714, + "learning_rate": 1.513544037435093e-05, + "loss": 3.2131, + "step": 8693 + }, + { + "epoch": 0.6970395462107394, + "grad_norm": 1.8034963607788086, + "learning_rate": 1.5134326181376477e-05, + "loss": 3.2823, + "step": 8694 + }, + { + "epoch": 0.697119720991762, + "grad_norm": 1.9443234205245972, + "learning_rate": 1.5133211901839095e-05, + "loss": 3.4081, + "step": 8695 + }, + { + "epoch": 0.6971998957727846, + "grad_norm": 1.9099781513214111, + "learning_rate": 1.5132097535757572e-05, + "loss": 3.1929, + "step": 8696 + }, + { + "epoch": 0.6972800705538073, + "grad_norm": 1.835378885269165, + "learning_rate": 1.5130983083150695e-05, + "loss": 3.235, + "step": 8697 + }, + { + "epoch": 0.6973602453348299, + "grad_norm": 1.7837377786636353, + "learning_rate": 1.5129868544037249e-05, + "loss": 3.2214, + "step": 8698 + }, + { + "epoch": 0.6974404201158526, + "grad_norm": 1.857717752456665, + "learning_rate": 1.5128753918436027e-05, + "loss": 3.2697, + "step": 8699 + }, + { + "epoch": 0.6975205948968752, + "grad_norm": 1.8943535089492798, + "learning_rate": 1.5127639206365822e-05, + "loss": 3.2159, + "step": 8700 + }, + { + "epoch": 0.6976007696778979, + "grad_norm": 1.8954572677612305, + "learning_rate": 1.512652440784543e-05, + "loss": 3.2306, + "step": 8701 + }, + { + "epoch": 0.6976809444589205, + "grad_norm": 1.7755769491195679, + "learning_rate": 1.5125409522893639e-05, + "loss": 3.2207, + "step": 8702 + }, + { + "epoch": 0.6977611192399431, + "grad_norm": 2.0364668369293213, + "learning_rate": 1.5124294551529251e-05, + "loss": 3.2378, + "step": 8703 + }, + { + "epoch": 0.6978412940209657, + "grad_norm": 1.7747818231582642, + "learning_rate": 1.5123179493771064e-05, + "loss": 3.2111, + "step": 8704 + }, + { + "epoch": 0.6979214688019884, + "grad_norm": 1.8382002115249634, + "learning_rate": 1.5122064349637872e-05, + "loss": 3.2972, + "step": 8705 + }, + { + "epoch": 0.698001643583011, + "grad_norm": 0.831903338432312, + "learning_rate": 1.5120949119148484e-05, + "loss": 2.502, + "step": 8706 + }, + { + "epoch": 0.6980818183640336, + "grad_norm": 1.857642650604248, + "learning_rate": 1.5119833802321699e-05, + "loss": 3.1818, + "step": 8707 + }, + { + "epoch": 0.6981619931450562, + "grad_norm": 1.8267297744750977, + "learning_rate": 1.5118718399176322e-05, + "loss": 3.1534, + "step": 8708 + }, + { + "epoch": 0.6982421679260788, + "grad_norm": 1.8119441270828247, + "learning_rate": 1.5117602909731152e-05, + "loss": 3.2268, + "step": 8709 + }, + { + "epoch": 0.6983223427071015, + "grad_norm": 1.8403990268707275, + "learning_rate": 1.5116487334005003e-05, + "loss": 3.1996, + "step": 8710 + }, + { + "epoch": 0.6984025174881241, + "grad_norm": 0.7459506988525391, + "learning_rate": 1.5115371672016684e-05, + "loss": 2.5116, + "step": 8711 + }, + { + "epoch": 0.6984826922691467, + "grad_norm": 1.8469135761260986, + "learning_rate": 1.5114255923784997e-05, + "loss": 3.1973, + "step": 8712 + }, + { + "epoch": 0.6985628670501693, + "grad_norm": 1.7945572137832642, + "learning_rate": 1.5113140089328759e-05, + "loss": 3.1636, + "step": 8713 + }, + { + "epoch": 0.698643041831192, + "grad_norm": 1.945163607597351, + "learning_rate": 1.5112024168666782e-05, + "loss": 3.2201, + "step": 8714 + }, + { + "epoch": 0.6987232166122146, + "grad_norm": 1.7994799613952637, + "learning_rate": 1.5110908161817874e-05, + "loss": 3.2932, + "step": 8715 + }, + { + "epoch": 0.6988033913932372, + "grad_norm": 1.8749675750732422, + "learning_rate": 1.5109792068800863e-05, + "loss": 3.255, + "step": 8716 + }, + { + "epoch": 0.6988835661742598, + "grad_norm": 1.7708882093429565, + "learning_rate": 1.5108675889634554e-05, + "loss": 3.1915, + "step": 8717 + }, + { + "epoch": 0.6989637409552825, + "grad_norm": 1.8267483711242676, + "learning_rate": 1.5107559624337774e-05, + "loss": 3.2045, + "step": 8718 + }, + { + "epoch": 0.6990439157363052, + "grad_norm": 1.9547154903411865, + "learning_rate": 1.5106443272929334e-05, + "loss": 3.279, + "step": 8719 + }, + { + "epoch": 0.6991240905173278, + "grad_norm": 1.8444410562515259, + "learning_rate": 1.5105326835428063e-05, + "loss": 3.1934, + "step": 8720 + }, + { + "epoch": 0.6992042652983504, + "grad_norm": 1.8075222969055176, + "learning_rate": 1.510421031185278e-05, + "loss": 3.2415, + "step": 8721 + }, + { + "epoch": 0.6992844400793731, + "grad_norm": 1.9345711469650269, + "learning_rate": 1.5103093702222311e-05, + "loss": 3.2282, + "step": 8722 + }, + { + "epoch": 0.6993646148603957, + "grad_norm": 1.9311046600341797, + "learning_rate": 1.510197700655548e-05, + "loss": 3.2153, + "step": 8723 + }, + { + "epoch": 0.6994447896414183, + "grad_norm": 1.8519846200942993, + "learning_rate": 1.5100860224871112e-05, + "loss": 3.1289, + "step": 8724 + }, + { + "epoch": 0.6995249644224409, + "grad_norm": 1.8075916767120361, + "learning_rate": 1.5099743357188042e-05, + "loss": 3.1436, + "step": 8725 + }, + { + "epoch": 0.6996051392034636, + "grad_norm": 1.9434274435043335, + "learning_rate": 1.5098626403525096e-05, + "loss": 3.264, + "step": 8726 + }, + { + "epoch": 0.6996853139844862, + "grad_norm": 1.9341449737548828, + "learning_rate": 1.5097509363901104e-05, + "loss": 3.2372, + "step": 8727 + }, + { + "epoch": 0.6997654887655088, + "grad_norm": 1.7730430364608765, + "learning_rate": 1.5096392238334905e-05, + "loss": 3.2834, + "step": 8728 + }, + { + "epoch": 0.6998456635465314, + "grad_norm": 1.7746800184249878, + "learning_rate": 1.5095275026845324e-05, + "loss": 3.2911, + "step": 8729 + }, + { + "epoch": 0.6999258383275541, + "grad_norm": 1.932450532913208, + "learning_rate": 1.5094157729451202e-05, + "loss": 3.1558, + "step": 8730 + }, + { + "epoch": 0.7000060131085767, + "grad_norm": 1.8070951700210571, + "learning_rate": 1.509304034617138e-05, + "loss": 3.2717, + "step": 8731 + }, + { + "epoch": 0.7000861878895993, + "grad_norm": 1.8833706378936768, + "learning_rate": 1.5091922877024687e-05, + "loss": 3.2596, + "step": 8732 + }, + { + "epoch": 0.7001663626706219, + "grad_norm": 1.911564826965332, + "learning_rate": 1.5090805322029975e-05, + "loss": 3.2854, + "step": 8733 + }, + { + "epoch": 0.7002465374516446, + "grad_norm": 1.8702999353408813, + "learning_rate": 1.5089687681206075e-05, + "loss": 3.2159, + "step": 8734 + }, + { + "epoch": 0.7003267122326672, + "grad_norm": 1.825535774230957, + "learning_rate": 1.5088569954571837e-05, + "loss": 3.2535, + "step": 8735 + }, + { + "epoch": 0.7004068870136898, + "grad_norm": 1.9323559999465942, + "learning_rate": 1.5087452142146104e-05, + "loss": 3.2427, + "step": 8736 + }, + { + "epoch": 0.7004870617947124, + "grad_norm": 1.7623306512832642, + "learning_rate": 1.5086334243947716e-05, + "loss": 3.1595, + "step": 8737 + }, + { + "epoch": 0.700567236575735, + "grad_norm": 1.844417929649353, + "learning_rate": 1.508521625999553e-05, + "loss": 3.158, + "step": 8738 + }, + { + "epoch": 0.7006474113567577, + "grad_norm": 1.7520018815994263, + "learning_rate": 1.5084098190308387e-05, + "loss": 3.2302, + "step": 8739 + }, + { + "epoch": 0.7007275861377804, + "grad_norm": 1.9278385639190674, + "learning_rate": 1.5082980034905142e-05, + "loss": 3.21, + "step": 8740 + }, + { + "epoch": 0.700807760918803, + "grad_norm": 1.8256100416183472, + "learning_rate": 1.5081861793804647e-05, + "loss": 3.2094, + "step": 8741 + }, + { + "epoch": 0.7008879356998257, + "grad_norm": 2.1046249866485596, + "learning_rate": 1.508074346702575e-05, + "loss": 3.196, + "step": 8742 + }, + { + "epoch": 0.7009681104808483, + "grad_norm": 1.8103997707366943, + "learning_rate": 1.507962505458731e-05, + "loss": 3.2598, + "step": 8743 + }, + { + "epoch": 0.7010482852618709, + "grad_norm": 1.8239256143569946, + "learning_rate": 1.507850655650818e-05, + "loss": 3.2965, + "step": 8744 + }, + { + "epoch": 0.7011284600428935, + "grad_norm": 1.783400535583496, + "learning_rate": 1.5077387972807218e-05, + "loss": 3.3069, + "step": 8745 + }, + { + "epoch": 0.7012086348239162, + "grad_norm": 1.679831624031067, + "learning_rate": 1.5076269303503291e-05, + "loss": 3.3021, + "step": 8746 + }, + { + "epoch": 0.7012888096049388, + "grad_norm": 1.9228206872940063, + "learning_rate": 1.5075150548615245e-05, + "loss": 3.2581, + "step": 8747 + }, + { + "epoch": 0.7013689843859614, + "grad_norm": 1.7762246131896973, + "learning_rate": 1.5074031708161956e-05, + "loss": 3.2336, + "step": 8748 + }, + { + "epoch": 0.701449159166984, + "grad_norm": 1.7298415899276733, + "learning_rate": 1.5072912782162274e-05, + "loss": 3.1544, + "step": 8749 + }, + { + "epoch": 0.7015293339480067, + "grad_norm": 1.7929481267929077, + "learning_rate": 1.5071793770635074e-05, + "loss": 3.1394, + "step": 8750 + }, + { + "epoch": 0.7016095087290293, + "grad_norm": 1.8499685525894165, + "learning_rate": 1.5070674673599221e-05, + "loss": 3.1958, + "step": 8751 + }, + { + "epoch": 0.7016896835100519, + "grad_norm": 1.7852489948272705, + "learning_rate": 1.5069555491073576e-05, + "loss": 3.1976, + "step": 8752 + }, + { + "epoch": 0.7017698582910745, + "grad_norm": 1.8746761083602905, + "learning_rate": 1.5068436223077014e-05, + "loss": 3.2345, + "step": 8753 + }, + { + "epoch": 0.7018500330720971, + "grad_norm": 1.7338159084320068, + "learning_rate": 1.5067316869628403e-05, + "loss": 3.2417, + "step": 8754 + }, + { + "epoch": 0.7019302078531198, + "grad_norm": 0.7862181663513184, + "learning_rate": 1.5066197430746616e-05, + "loss": 2.4982, + "step": 8755 + }, + { + "epoch": 0.7020103826341424, + "grad_norm": 1.9671249389648438, + "learning_rate": 1.5065077906450528e-05, + "loss": 3.3251, + "step": 8756 + }, + { + "epoch": 0.702090557415165, + "grad_norm": 1.8797612190246582, + "learning_rate": 1.5063958296759005e-05, + "loss": 3.2061, + "step": 8757 + }, + { + "epoch": 0.7021707321961876, + "grad_norm": 1.7558348178863525, + "learning_rate": 1.5062838601690936e-05, + "loss": 3.2297, + "step": 8758 + }, + { + "epoch": 0.7022509069772103, + "grad_norm": 1.741884469985962, + "learning_rate": 1.506171882126519e-05, + "loss": 3.1495, + "step": 8759 + }, + { + "epoch": 0.702331081758233, + "grad_norm": 1.7694575786590576, + "learning_rate": 1.506059895550065e-05, + "loss": 3.1318, + "step": 8760 + }, + { + "epoch": 0.7024112565392556, + "grad_norm": 1.9392284154891968, + "learning_rate": 1.5059479004416194e-05, + "loss": 3.2144, + "step": 8761 + }, + { + "epoch": 0.7024914313202782, + "grad_norm": 1.9055020809173584, + "learning_rate": 1.5058358968030704e-05, + "loss": 3.2478, + "step": 8762 + }, + { + "epoch": 0.7025716061013009, + "grad_norm": 1.7290879487991333, + "learning_rate": 1.5057238846363067e-05, + "loss": 3.201, + "step": 8763 + }, + { + "epoch": 0.7026517808823235, + "grad_norm": 1.8419101238250732, + "learning_rate": 1.5056118639432162e-05, + "loss": 3.2751, + "step": 8764 + }, + { + "epoch": 0.7027319556633461, + "grad_norm": 1.821912169456482, + "learning_rate": 1.5054998347256881e-05, + "loss": 3.2623, + "step": 8765 + }, + { + "epoch": 0.7028121304443687, + "grad_norm": 1.8527313470840454, + "learning_rate": 1.505387796985611e-05, + "loss": 3.1644, + "step": 8766 + }, + { + "epoch": 0.7028923052253914, + "grad_norm": 2.3741633892059326, + "learning_rate": 1.5052757507248737e-05, + "loss": 3.1649, + "step": 8767 + }, + { + "epoch": 0.702972480006414, + "grad_norm": 1.7717113494873047, + "learning_rate": 1.5051636959453656e-05, + "loss": 3.1851, + "step": 8768 + }, + { + "epoch": 0.7030526547874366, + "grad_norm": 1.8118476867675781, + "learning_rate": 1.505051632648975e-05, + "loss": 3.2126, + "step": 8769 + }, + { + "epoch": 0.7031328295684592, + "grad_norm": 1.8607488870620728, + "learning_rate": 1.5049395608375922e-05, + "loss": 3.3616, + "step": 8770 + }, + { + "epoch": 0.7032130043494819, + "grad_norm": 1.8593465089797974, + "learning_rate": 1.5048274805131064e-05, + "loss": 3.1792, + "step": 8771 + }, + { + "epoch": 0.7032931791305045, + "grad_norm": 1.8735432624816895, + "learning_rate": 1.5047153916774071e-05, + "loss": 3.1507, + "step": 8772 + }, + { + "epoch": 0.7033733539115271, + "grad_norm": 1.7470544576644897, + "learning_rate": 1.5046032943323843e-05, + "loss": 3.2689, + "step": 8773 + }, + { + "epoch": 0.7034535286925497, + "grad_norm": 1.8331992626190186, + "learning_rate": 1.5044911884799276e-05, + "loss": 3.2859, + "step": 8774 + }, + { + "epoch": 0.7035337034735724, + "grad_norm": 1.8085761070251465, + "learning_rate": 1.5043790741219275e-05, + "loss": 3.1405, + "step": 8775 + }, + { + "epoch": 0.703613878254595, + "grad_norm": 1.7940747737884521, + "learning_rate": 1.504266951260274e-05, + "loss": 3.3245, + "step": 8776 + }, + { + "epoch": 0.7036940530356176, + "grad_norm": 1.9099823236465454, + "learning_rate": 1.5041548198968569e-05, + "loss": 3.3464, + "step": 8777 + }, + { + "epoch": 0.7037742278166402, + "grad_norm": 1.83168625831604, + "learning_rate": 1.5040426800335679e-05, + "loss": 3.2731, + "step": 8778 + }, + { + "epoch": 0.7038544025976629, + "grad_norm": 1.7815159559249878, + "learning_rate": 1.5039305316722965e-05, + "loss": 3.2242, + "step": 8779 + }, + { + "epoch": 0.7039345773786856, + "grad_norm": 1.8755439519882202, + "learning_rate": 1.5038183748149341e-05, + "loss": 3.2936, + "step": 8780 + }, + { + "epoch": 0.7040147521597082, + "grad_norm": 1.815305471420288, + "learning_rate": 1.5037062094633716e-05, + "loss": 3.124, + "step": 8781 + }, + { + "epoch": 0.7040949269407308, + "grad_norm": 1.836374282836914, + "learning_rate": 1.5035940356194995e-05, + "loss": 3.2356, + "step": 8782 + }, + { + "epoch": 0.7041751017217535, + "grad_norm": 1.8328919410705566, + "learning_rate": 1.50348185328521e-05, + "loss": 3.2664, + "step": 8783 + }, + { + "epoch": 0.7042552765027761, + "grad_norm": 1.8402881622314453, + "learning_rate": 1.5033696624623937e-05, + "loss": 3.1498, + "step": 8784 + }, + { + "epoch": 0.7043354512837987, + "grad_norm": 1.7258223295211792, + "learning_rate": 1.503257463152942e-05, + "loss": 3.135, + "step": 8785 + }, + { + "epoch": 0.7044156260648213, + "grad_norm": 1.8773598670959473, + "learning_rate": 1.5031452553587472e-05, + "loss": 3.2461, + "step": 8786 + }, + { + "epoch": 0.704495800845844, + "grad_norm": 1.8555608987808228, + "learning_rate": 1.5030330390817005e-05, + "loss": 3.3589, + "step": 8787 + }, + { + "epoch": 0.7045759756268666, + "grad_norm": 1.8051235675811768, + "learning_rate": 1.5029208143236944e-05, + "loss": 3.1971, + "step": 8788 + }, + { + "epoch": 0.7046561504078892, + "grad_norm": 1.8045059442520142, + "learning_rate": 1.5028085810866202e-05, + "loss": 3.1471, + "step": 8789 + }, + { + "epoch": 0.7047363251889118, + "grad_norm": 1.7167550325393677, + "learning_rate": 1.5026963393723707e-05, + "loss": 3.1691, + "step": 8790 + }, + { + "epoch": 0.7048164999699345, + "grad_norm": 1.948744535446167, + "learning_rate": 1.502584089182838e-05, + "loss": 3.2412, + "step": 8791 + }, + { + "epoch": 0.7048966747509571, + "grad_norm": 1.82679283618927, + "learning_rate": 1.5024718305199147e-05, + "loss": 3.1315, + "step": 8792 + }, + { + "epoch": 0.7049768495319797, + "grad_norm": 1.6903564929962158, + "learning_rate": 1.5023595633854937e-05, + "loss": 3.2892, + "step": 8793 + }, + { + "epoch": 0.7050570243130023, + "grad_norm": 1.8279774188995361, + "learning_rate": 1.502247287781467e-05, + "loss": 3.1666, + "step": 8794 + }, + { + "epoch": 0.705137199094025, + "grad_norm": 1.7131781578063965, + "learning_rate": 1.5021350037097285e-05, + "loss": 3.2631, + "step": 8795 + }, + { + "epoch": 0.7052173738750476, + "grad_norm": 1.8823115825653076, + "learning_rate": 1.5020227111721707e-05, + "loss": 3.2666, + "step": 8796 + }, + { + "epoch": 0.7052975486560702, + "grad_norm": 1.7885605096817017, + "learning_rate": 1.5019104101706867e-05, + "loss": 3.1989, + "step": 8797 + }, + { + "epoch": 0.7053777234370928, + "grad_norm": 0.7393261790275574, + "learning_rate": 1.5017981007071702e-05, + "loss": 2.4884, + "step": 8798 + }, + { + "epoch": 0.7054578982181154, + "grad_norm": 1.929572582244873, + "learning_rate": 1.5016857827835148e-05, + "loss": 3.1836, + "step": 8799 + }, + { + "epoch": 0.7055380729991381, + "grad_norm": 0.7028825283050537, + "learning_rate": 1.5015734564016135e-05, + "loss": 2.502, + "step": 8800 + }, + { + "epoch": 0.7056182477801608, + "grad_norm": 1.7693402767181396, + "learning_rate": 1.5014611215633608e-05, + "loss": 3.1909, + "step": 8801 + }, + { + "epoch": 0.7056984225611834, + "grad_norm": 1.7242897748947144, + "learning_rate": 1.5013487782706501e-05, + "loss": 3.2059, + "step": 8802 + }, + { + "epoch": 0.705778597342206, + "grad_norm": 1.769113302230835, + "learning_rate": 1.501236426525376e-05, + "loss": 3.2218, + "step": 8803 + }, + { + "epoch": 0.7058587721232287, + "grad_norm": 1.7612683773040771, + "learning_rate": 1.5011240663294322e-05, + "loss": 3.1293, + "step": 8804 + }, + { + "epoch": 0.7059389469042513, + "grad_norm": 1.820295810699463, + "learning_rate": 1.5010116976847134e-05, + "loss": 3.2422, + "step": 8805 + }, + { + "epoch": 0.7060191216852739, + "grad_norm": 1.83512544631958, + "learning_rate": 1.5008993205931137e-05, + "loss": 3.199, + "step": 8806 + }, + { + "epoch": 0.7060992964662965, + "grad_norm": 2.046693801879883, + "learning_rate": 1.5007869350565281e-05, + "loss": 3.084, + "step": 8807 + }, + { + "epoch": 0.7061794712473192, + "grad_norm": 1.8079664707183838, + "learning_rate": 1.5006745410768516e-05, + "loss": 3.3351, + "step": 8808 + }, + { + "epoch": 0.7062596460283418, + "grad_norm": 1.836106538772583, + "learning_rate": 1.5005621386559784e-05, + "loss": 3.2634, + "step": 8809 + }, + { + "epoch": 0.7063398208093644, + "grad_norm": 1.8152587413787842, + "learning_rate": 1.500449727795804e-05, + "loss": 3.1849, + "step": 8810 + }, + { + "epoch": 0.706419995590387, + "grad_norm": 1.8874683380126953, + "learning_rate": 1.5003373084982235e-05, + "loss": 3.2878, + "step": 8811 + }, + { + "epoch": 0.7065001703714097, + "grad_norm": 0.7467862367630005, + "learning_rate": 1.5002248807651322e-05, + "loss": 2.5005, + "step": 8812 + }, + { + "epoch": 0.7065803451524323, + "grad_norm": 1.7554727792739868, + "learning_rate": 1.5001124445984261e-05, + "loss": 3.2786, + "step": 8813 + }, + { + "epoch": 0.7066605199334549, + "grad_norm": 2.0148234367370605, + "learning_rate": 1.5000000000000002e-05, + "loss": 3.1654, + "step": 8814 + }, + { + "epoch": 0.7067406947144775, + "grad_norm": 1.726805329322815, + "learning_rate": 1.4998875469717504e-05, + "loss": 3.1984, + "step": 8815 + }, + { + "epoch": 0.7068208694955002, + "grad_norm": 1.867340087890625, + "learning_rate": 1.4997750855155731e-05, + "loss": 3.1447, + "step": 8816 + }, + { + "epoch": 0.7069010442765228, + "grad_norm": 1.7632032632827759, + "learning_rate": 1.4996626156333638e-05, + "loss": 3.1693, + "step": 8817 + }, + { + "epoch": 0.7069812190575454, + "grad_norm": 1.7281004190444946, + "learning_rate": 1.499550137327019e-05, + "loss": 3.2877, + "step": 8818 + }, + { + "epoch": 0.707061393838568, + "grad_norm": 1.9229462146759033, + "learning_rate": 1.4994376505984348e-05, + "loss": 3.2222, + "step": 8819 + }, + { + "epoch": 0.7071415686195907, + "grad_norm": 1.845027208328247, + "learning_rate": 1.499325155449508e-05, + "loss": 3.2455, + "step": 8820 + }, + { + "epoch": 0.7072217434006134, + "grad_norm": 1.8730368614196777, + "learning_rate": 1.4992126518821349e-05, + "loss": 3.3003, + "step": 8821 + }, + { + "epoch": 0.707301918181636, + "grad_norm": 1.7394980192184448, + "learning_rate": 1.4991001398982124e-05, + "loss": 3.1622, + "step": 8822 + }, + { + "epoch": 0.7073820929626586, + "grad_norm": 2.0449633598327637, + "learning_rate": 1.4989876194996376e-05, + "loss": 3.1757, + "step": 8823 + }, + { + "epoch": 0.7074622677436813, + "grad_norm": 1.75945246219635, + "learning_rate": 1.4988750906883072e-05, + "loss": 3.2887, + "step": 8824 + }, + { + "epoch": 0.7075424425247039, + "grad_norm": 1.6912424564361572, + "learning_rate": 1.4987625534661186e-05, + "loss": 3.2098, + "step": 8825 + }, + { + "epoch": 0.7076226173057265, + "grad_norm": 1.741137146949768, + "learning_rate": 1.4986500078349693e-05, + "loss": 3.1851, + "step": 8826 + }, + { + "epoch": 0.7077027920867491, + "grad_norm": 1.7448902130126953, + "learning_rate": 1.4985374537967565e-05, + "loss": 3.132, + "step": 8827 + }, + { + "epoch": 0.7077829668677718, + "grad_norm": 1.8420089483261108, + "learning_rate": 1.4984248913533781e-05, + "loss": 3.2222, + "step": 8828 + }, + { + "epoch": 0.7078631416487944, + "grad_norm": 1.822798728942871, + "learning_rate": 1.4983123205067315e-05, + "loss": 3.2617, + "step": 8829 + }, + { + "epoch": 0.707943316429817, + "grad_norm": 1.835280179977417, + "learning_rate": 1.4981997412587148e-05, + "loss": 3.2366, + "step": 8830 + }, + { + "epoch": 0.7080234912108396, + "grad_norm": 1.890356421470642, + "learning_rate": 1.4980871536112263e-05, + "loss": 3.2555, + "step": 8831 + }, + { + "epoch": 0.7081036659918623, + "grad_norm": 1.9470984935760498, + "learning_rate": 1.4979745575661638e-05, + "loss": 3.172, + "step": 8832 + }, + { + "epoch": 0.7081838407728849, + "grad_norm": 1.8156487941741943, + "learning_rate": 1.4978619531254256e-05, + "loss": 3.1936, + "step": 8833 + }, + { + "epoch": 0.7082640155539075, + "grad_norm": 1.7807574272155762, + "learning_rate": 1.4977493402909105e-05, + "loss": 3.1351, + "step": 8834 + }, + { + "epoch": 0.7083441903349301, + "grad_norm": 1.7551413774490356, + "learning_rate": 1.4976367190645167e-05, + "loss": 3.1514, + "step": 8835 + }, + { + "epoch": 0.7084243651159527, + "grad_norm": 1.8252060413360596, + "learning_rate": 1.4975240894481435e-05, + "loss": 3.2354, + "step": 8836 + }, + { + "epoch": 0.7085045398969754, + "grad_norm": 1.7947399616241455, + "learning_rate": 1.4974114514436893e-05, + "loss": 3.2029, + "step": 8837 + }, + { + "epoch": 0.708584714677998, + "grad_norm": 1.7965468168258667, + "learning_rate": 1.4972988050530534e-05, + "loss": 3.1993, + "step": 8838 + }, + { + "epoch": 0.7086648894590206, + "grad_norm": 1.78781259059906, + "learning_rate": 1.497186150278135e-05, + "loss": 3.1887, + "step": 8839 + }, + { + "epoch": 0.7087450642400432, + "grad_norm": 1.828873872756958, + "learning_rate": 1.497073487120833e-05, + "loss": 3.2611, + "step": 8840 + }, + { + "epoch": 0.708825239021066, + "grad_norm": 1.8206735849380493, + "learning_rate": 1.4969608155830478e-05, + "loss": 3.1415, + "step": 8841 + }, + { + "epoch": 0.7089054138020886, + "grad_norm": 1.723302960395813, + "learning_rate": 1.4968481356666779e-05, + "loss": 3.1559, + "step": 8842 + }, + { + "epoch": 0.7089855885831112, + "grad_norm": 1.7467597723007202, + "learning_rate": 1.4967354473736235e-05, + "loss": 3.1697, + "step": 8843 + }, + { + "epoch": 0.7090657633641338, + "grad_norm": 1.8296494483947754, + "learning_rate": 1.4966227507057849e-05, + "loss": 3.2028, + "step": 8844 + }, + { + "epoch": 0.7091459381451565, + "grad_norm": 1.7620019912719727, + "learning_rate": 1.4965100456650614e-05, + "loss": 3.257, + "step": 8845 + }, + { + "epoch": 0.7092261129261791, + "grad_norm": 1.9001485109329224, + "learning_rate": 1.4963973322533538e-05, + "loss": 3.157, + "step": 8846 + }, + { + "epoch": 0.7093062877072017, + "grad_norm": 0.7310785055160522, + "learning_rate": 1.4962846104725621e-05, + "loss": 2.4935, + "step": 8847 + }, + { + "epoch": 0.7093864624882243, + "grad_norm": 1.8547015190124512, + "learning_rate": 1.4961718803245866e-05, + "loss": 3.198, + "step": 8848 + }, + { + "epoch": 0.709466637269247, + "grad_norm": 2.099357843399048, + "learning_rate": 1.4960591418113282e-05, + "loss": 3.1903, + "step": 8849 + }, + { + "epoch": 0.7095468120502696, + "grad_norm": 1.758715033531189, + "learning_rate": 1.4959463949346874e-05, + "loss": 3.3588, + "step": 8850 + }, + { + "epoch": 0.7096269868312922, + "grad_norm": 1.8705257177352905, + "learning_rate": 1.4958336396965654e-05, + "loss": 3.274, + "step": 8851 + }, + { + "epoch": 0.7097071616123148, + "grad_norm": 2.0934948921203613, + "learning_rate": 1.4957208760988625e-05, + "loss": 3.205, + "step": 8852 + }, + { + "epoch": 0.7097873363933375, + "grad_norm": 1.8604884147644043, + "learning_rate": 1.4956081041434807e-05, + "loss": 3.2455, + "step": 8853 + }, + { + "epoch": 0.7098675111743601, + "grad_norm": 1.7643438577651978, + "learning_rate": 1.4954953238323209e-05, + "loss": 3.1868, + "step": 8854 + }, + { + "epoch": 0.7099476859553827, + "grad_norm": 2.0861928462982178, + "learning_rate": 1.4953825351672844e-05, + "loss": 3.1782, + "step": 8855 + }, + { + "epoch": 0.7100278607364053, + "grad_norm": 1.7534383535385132, + "learning_rate": 1.4952697381502732e-05, + "loss": 3.2953, + "step": 8856 + }, + { + "epoch": 0.710108035517428, + "grad_norm": 1.786978006362915, + "learning_rate": 1.4951569327831886e-05, + "loss": 3.1976, + "step": 8857 + }, + { + "epoch": 0.7101882102984506, + "grad_norm": 1.7925658226013184, + "learning_rate": 1.4950441190679326e-05, + "loss": 3.1055, + "step": 8858 + }, + { + "epoch": 0.7102683850794732, + "grad_norm": 1.802922010421753, + "learning_rate": 1.4949312970064073e-05, + "loss": 3.2106, + "step": 8859 + }, + { + "epoch": 0.7103485598604958, + "grad_norm": 1.847282886505127, + "learning_rate": 1.4948184666005147e-05, + "loss": 3.1729, + "step": 8860 + }, + { + "epoch": 0.7104287346415185, + "grad_norm": 1.7624698877334595, + "learning_rate": 1.4947056278521575e-05, + "loss": 3.1727, + "step": 8861 + }, + { + "epoch": 0.7105089094225412, + "grad_norm": 1.9283418655395508, + "learning_rate": 1.4945927807632372e-05, + "loss": 3.2396, + "step": 8862 + }, + { + "epoch": 0.7105890842035638, + "grad_norm": 1.846015214920044, + "learning_rate": 1.4944799253356572e-05, + "loss": 3.2343, + "step": 8863 + }, + { + "epoch": 0.7106692589845864, + "grad_norm": 1.6710867881774902, + "learning_rate": 1.49436706157132e-05, + "loss": 3.1138, + "step": 8864 + }, + { + "epoch": 0.7107494337656091, + "grad_norm": 1.7775417566299438, + "learning_rate": 1.4942541894721284e-05, + "loss": 3.28, + "step": 8865 + }, + { + "epoch": 0.7108296085466317, + "grad_norm": 1.7290180921554565, + "learning_rate": 1.4941413090399852e-05, + "loss": 3.2217, + "step": 8866 + }, + { + "epoch": 0.7109097833276543, + "grad_norm": 1.7942761182785034, + "learning_rate": 1.4940284202767937e-05, + "loss": 3.2169, + "step": 8867 + }, + { + "epoch": 0.7109899581086769, + "grad_norm": 1.8679392337799072, + "learning_rate": 1.493915523184457e-05, + "loss": 3.2396, + "step": 8868 + }, + { + "epoch": 0.7110701328896996, + "grad_norm": 0.7388980388641357, + "learning_rate": 1.493802617764879e-05, + "loss": 2.4817, + "step": 8869 + }, + { + "epoch": 0.7111503076707222, + "grad_norm": 1.7894458770751953, + "learning_rate": 1.4936897040199627e-05, + "loss": 3.2141, + "step": 8870 + }, + { + "epoch": 0.7112304824517448, + "grad_norm": 1.967785120010376, + "learning_rate": 1.493576781951612e-05, + "loss": 3.2753, + "step": 8871 + }, + { + "epoch": 0.7113106572327674, + "grad_norm": 1.923857569694519, + "learning_rate": 1.4934638515617308e-05, + "loss": 3.1979, + "step": 8872 + }, + { + "epoch": 0.71139083201379, + "grad_norm": 1.8234124183654785, + "learning_rate": 1.4933509128522228e-05, + "loss": 3.1626, + "step": 8873 + }, + { + "epoch": 0.7114710067948127, + "grad_norm": 1.7579795122146606, + "learning_rate": 1.4932379658249924e-05, + "loss": 3.2542, + "step": 8874 + }, + { + "epoch": 0.7115511815758353, + "grad_norm": 1.757015347480774, + "learning_rate": 1.4931250104819437e-05, + "loss": 3.1848, + "step": 8875 + }, + { + "epoch": 0.7116313563568579, + "grad_norm": 1.8469038009643555, + "learning_rate": 1.4930120468249811e-05, + "loss": 3.246, + "step": 8876 + }, + { + "epoch": 0.7117115311378805, + "grad_norm": 1.7283295392990112, + "learning_rate": 1.4928990748560093e-05, + "loss": 3.17, + "step": 8877 + }, + { + "epoch": 0.7117917059189032, + "grad_norm": 1.8215560913085938, + "learning_rate": 1.4927860945769325e-05, + "loss": 3.2803, + "step": 8878 + }, + { + "epoch": 0.7118718806999258, + "grad_norm": 1.82686185836792, + "learning_rate": 1.4926731059896561e-05, + "loss": 3.2175, + "step": 8879 + }, + { + "epoch": 0.7119520554809484, + "grad_norm": 2.0064797401428223, + "learning_rate": 1.4925601090960846e-05, + "loss": 3.2955, + "step": 8880 + }, + { + "epoch": 0.712032230261971, + "grad_norm": 1.8228882551193237, + "learning_rate": 1.4924471038981234e-05, + "loss": 3.3036, + "step": 8881 + }, + { + "epoch": 0.7121124050429938, + "grad_norm": 1.8458317518234253, + "learning_rate": 1.4923340903976774e-05, + "loss": 3.1904, + "step": 8882 + }, + { + "epoch": 0.7121925798240164, + "grad_norm": 1.8074731826782227, + "learning_rate": 1.4922210685966524e-05, + "loss": 3.1281, + "step": 8883 + }, + { + "epoch": 0.712272754605039, + "grad_norm": 1.8398929834365845, + "learning_rate": 1.4921080384969538e-05, + "loss": 3.1885, + "step": 8884 + }, + { + "epoch": 0.7123529293860617, + "grad_norm": 1.8428677320480347, + "learning_rate": 1.4919950001004867e-05, + "loss": 3.2786, + "step": 8885 + }, + { + "epoch": 0.7124331041670843, + "grad_norm": 1.8525303602218628, + "learning_rate": 1.4918819534091574e-05, + "loss": 3.2036, + "step": 8886 + }, + { + "epoch": 0.7125132789481069, + "grad_norm": 1.8359652757644653, + "learning_rate": 1.4917688984248717e-05, + "loss": 3.2397, + "step": 8887 + }, + { + "epoch": 0.7125934537291295, + "grad_norm": 0.7499322295188904, + "learning_rate": 1.4916558351495357e-05, + "loss": 2.5194, + "step": 8888 + }, + { + "epoch": 0.7126736285101521, + "grad_norm": 1.7939683198928833, + "learning_rate": 1.4915427635850557e-05, + "loss": 3.1998, + "step": 8889 + }, + { + "epoch": 0.7127538032911748, + "grad_norm": 1.8028258085250854, + "learning_rate": 1.4914296837333378e-05, + "loss": 3.2649, + "step": 8890 + }, + { + "epoch": 0.7128339780721974, + "grad_norm": 1.786086082458496, + "learning_rate": 1.4913165955962888e-05, + "loss": 3.2041, + "step": 8891 + }, + { + "epoch": 0.71291415285322, + "grad_norm": 1.689049482345581, + "learning_rate": 1.491203499175815e-05, + "loss": 3.1965, + "step": 8892 + }, + { + "epoch": 0.7129943276342426, + "grad_norm": 0.7372407913208008, + "learning_rate": 1.4910903944738233e-05, + "loss": 2.4977, + "step": 8893 + }, + { + "epoch": 0.7130745024152653, + "grad_norm": 1.8046525716781616, + "learning_rate": 1.4909772814922207e-05, + "loss": 3.2666, + "step": 8894 + }, + { + "epoch": 0.7131546771962879, + "grad_norm": 1.8591201305389404, + "learning_rate": 1.490864160232914e-05, + "loss": 3.2916, + "step": 8895 + }, + { + "epoch": 0.7132348519773105, + "grad_norm": 1.877073049545288, + "learning_rate": 1.490751030697811e-05, + "loss": 3.2751, + "step": 8896 + }, + { + "epoch": 0.7133150267583331, + "grad_norm": 1.8092495203018188, + "learning_rate": 1.4906378928888183e-05, + "loss": 3.1984, + "step": 8897 + }, + { + "epoch": 0.7133952015393558, + "grad_norm": 1.8676321506500244, + "learning_rate": 1.4905247468078436e-05, + "loss": 3.2048, + "step": 8898 + }, + { + "epoch": 0.7134753763203784, + "grad_norm": 1.7314554452896118, + "learning_rate": 1.4904115924567947e-05, + "loss": 3.2965, + "step": 8899 + }, + { + "epoch": 0.713555551101401, + "grad_norm": 1.811231017112732, + "learning_rate": 1.4902984298375791e-05, + "loss": 3.1528, + "step": 8900 + }, + { + "epoch": 0.7136357258824236, + "grad_norm": 1.7316838502883911, + "learning_rate": 1.4901852589521046e-05, + "loss": 3.2445, + "step": 8901 + }, + { + "epoch": 0.7137159006634464, + "grad_norm": 1.7444102764129639, + "learning_rate": 1.4900720798022797e-05, + "loss": 3.137, + "step": 8902 + }, + { + "epoch": 0.713796075444469, + "grad_norm": 1.787721037864685, + "learning_rate": 1.489958892390012e-05, + "loss": 3.252, + "step": 8903 + }, + { + "epoch": 0.7138762502254916, + "grad_norm": 1.6844282150268555, + "learning_rate": 1.4898456967172102e-05, + "loss": 3.1534, + "step": 8904 + }, + { + "epoch": 0.7139564250065142, + "grad_norm": 1.7968813180923462, + "learning_rate": 1.4897324927857825e-05, + "loss": 3.1797, + "step": 8905 + }, + { + "epoch": 0.7140365997875369, + "grad_norm": 0.7294797897338867, + "learning_rate": 1.4896192805976377e-05, + "loss": 2.5001, + "step": 8906 + }, + { + "epoch": 0.7141167745685595, + "grad_norm": 1.7951525449752808, + "learning_rate": 1.4895060601546842e-05, + "loss": 3.275, + "step": 8907 + }, + { + "epoch": 0.7141969493495821, + "grad_norm": 1.8781185150146484, + "learning_rate": 1.4893928314588312e-05, + "loss": 3.3423, + "step": 8908 + }, + { + "epoch": 0.7142771241306047, + "grad_norm": 1.710699200630188, + "learning_rate": 1.4892795945119876e-05, + "loss": 3.2112, + "step": 8909 + }, + { + "epoch": 0.7143572989116274, + "grad_norm": 1.7007228136062622, + "learning_rate": 1.4891663493160624e-05, + "loss": 3.2202, + "step": 8910 + }, + { + "epoch": 0.71443747369265, + "grad_norm": 1.7760852575302124, + "learning_rate": 1.489053095872965e-05, + "loss": 3.217, + "step": 8911 + }, + { + "epoch": 0.7145176484736726, + "grad_norm": 1.8344427347183228, + "learning_rate": 1.4889398341846044e-05, + "loss": 3.2504, + "step": 8912 + }, + { + "epoch": 0.7145978232546952, + "grad_norm": 0.6951600909233093, + "learning_rate": 1.4888265642528908e-05, + "loss": 2.4994, + "step": 8913 + }, + { + "epoch": 0.7146779980357179, + "grad_norm": 1.6472761631011963, + "learning_rate": 1.4887132860797337e-05, + "loss": 3.1375, + "step": 8914 + }, + { + "epoch": 0.7147581728167405, + "grad_norm": 1.7249462604522705, + "learning_rate": 1.4885999996670428e-05, + "loss": 3.2588, + "step": 8915 + }, + { + "epoch": 0.7148383475977631, + "grad_norm": 1.8480302095413208, + "learning_rate": 1.4884867050167281e-05, + "loss": 3.1804, + "step": 8916 + }, + { + "epoch": 0.7149185223787857, + "grad_norm": 1.9301749467849731, + "learning_rate": 1.4883734021306998e-05, + "loss": 3.1428, + "step": 8917 + }, + { + "epoch": 0.7149986971598083, + "grad_norm": 1.8700834512710571, + "learning_rate": 1.4882600910108676e-05, + "loss": 3.3497, + "step": 8918 + }, + { + "epoch": 0.715078871940831, + "grad_norm": 1.9154154062271118, + "learning_rate": 1.488146771659143e-05, + "loss": 3.191, + "step": 8919 + }, + { + "epoch": 0.7151590467218536, + "grad_norm": 1.781874656677246, + "learning_rate": 1.4880334440774352e-05, + "loss": 3.2507, + "step": 8920 + }, + { + "epoch": 0.7152392215028762, + "grad_norm": 1.690849781036377, + "learning_rate": 1.487920108267656e-05, + "loss": 3.1949, + "step": 8921 + }, + { + "epoch": 0.7153193962838988, + "grad_norm": 1.8247079849243164, + "learning_rate": 1.4878067642317156e-05, + "loss": 3.1454, + "step": 8922 + }, + { + "epoch": 0.7153995710649216, + "grad_norm": 1.8078826665878296, + "learning_rate": 1.487693411971525e-05, + "loss": 3.1877, + "step": 8923 + }, + { + "epoch": 0.7154797458459442, + "grad_norm": 1.813211441040039, + "learning_rate": 1.4875800514889955e-05, + "loss": 3.1729, + "step": 8924 + }, + { + "epoch": 0.7155599206269668, + "grad_norm": 1.7073079347610474, + "learning_rate": 1.487466682786038e-05, + "loss": 3.103, + "step": 8925 + }, + { + "epoch": 0.7156400954079895, + "grad_norm": 2.040823459625244, + "learning_rate": 1.4873533058645642e-05, + "loss": 3.2469, + "step": 8926 + }, + { + "epoch": 0.7157202701890121, + "grad_norm": 1.8074488639831543, + "learning_rate": 1.4872399207264851e-05, + "loss": 3.3008, + "step": 8927 + }, + { + "epoch": 0.7158004449700347, + "grad_norm": 1.7307028770446777, + "learning_rate": 1.487126527373713e-05, + "loss": 3.1349, + "step": 8928 + }, + { + "epoch": 0.7158806197510573, + "grad_norm": 1.8633843660354614, + "learning_rate": 1.4870131258081593e-05, + "loss": 3.2773, + "step": 8929 + }, + { + "epoch": 0.71596079453208, + "grad_norm": 1.7682201862335205, + "learning_rate": 1.486899716031736e-05, + "loss": 3.1082, + "step": 8930 + }, + { + "epoch": 0.7160409693131026, + "grad_norm": 1.8640398979187012, + "learning_rate": 1.486786298046355e-05, + "loss": 3.2351, + "step": 8931 + }, + { + "epoch": 0.7161211440941252, + "grad_norm": 1.7564868927001953, + "learning_rate": 1.4866728718539284e-05, + "loss": 3.2199, + "step": 8932 + }, + { + "epoch": 0.7162013188751478, + "grad_norm": 1.9025079011917114, + "learning_rate": 1.4865594374563687e-05, + "loss": 3.2577, + "step": 8933 + }, + { + "epoch": 0.7162814936561704, + "grad_norm": 1.96880304813385, + "learning_rate": 1.4864459948555888e-05, + "loss": 3.1709, + "step": 8934 + }, + { + "epoch": 0.7163616684371931, + "grad_norm": 1.8336150646209717, + "learning_rate": 1.4863325440535004e-05, + "loss": 3.296, + "step": 8935 + }, + { + "epoch": 0.7164418432182157, + "grad_norm": 1.6531033515930176, + "learning_rate": 1.4862190850520169e-05, + "loss": 3.1816, + "step": 8936 + }, + { + "epoch": 0.7165220179992383, + "grad_norm": 1.869227647781372, + "learning_rate": 1.4861056178530509e-05, + "loss": 3.2938, + "step": 8937 + }, + { + "epoch": 0.7166021927802609, + "grad_norm": 0.7536813616752625, + "learning_rate": 1.4859921424585155e-05, + "loss": 2.4908, + "step": 8938 + }, + { + "epoch": 0.7166823675612836, + "grad_norm": 1.918819546699524, + "learning_rate": 1.485878658870324e-05, + "loss": 3.1659, + "step": 8939 + }, + { + "epoch": 0.7167625423423062, + "grad_norm": 1.8554073572158813, + "learning_rate": 1.4857651670903894e-05, + "loss": 3.1791, + "step": 8940 + }, + { + "epoch": 0.7168427171233288, + "grad_norm": 1.756299376487732, + "learning_rate": 1.4856516671206256e-05, + "loss": 3.2666, + "step": 8941 + }, + { + "epoch": 0.7169228919043514, + "grad_norm": 1.7497849464416504, + "learning_rate": 1.4855381589629453e-05, + "loss": 3.173, + "step": 8942 + }, + { + "epoch": 0.7170030666853742, + "grad_norm": 1.8018606901168823, + "learning_rate": 1.4854246426192632e-05, + "loss": 3.1281, + "step": 8943 + }, + { + "epoch": 0.7170832414663968, + "grad_norm": 0.7905453443527222, + "learning_rate": 1.4853111180914925e-05, + "loss": 2.4824, + "step": 8944 + }, + { + "epoch": 0.7171634162474194, + "grad_norm": 1.958345651626587, + "learning_rate": 1.4851975853815477e-05, + "loss": 3.2473, + "step": 8945 + }, + { + "epoch": 0.717243591028442, + "grad_norm": 1.8826767206192017, + "learning_rate": 1.4850840444913423e-05, + "loss": 3.1541, + "step": 8946 + }, + { + "epoch": 0.7173237658094647, + "grad_norm": 0.7272562980651855, + "learning_rate": 1.484970495422791e-05, + "loss": 2.5031, + "step": 8947 + }, + { + "epoch": 0.7174039405904873, + "grad_norm": 1.8173342943191528, + "learning_rate": 1.4848569381778077e-05, + "loss": 3.2301, + "step": 8948 + }, + { + "epoch": 0.7174841153715099, + "grad_norm": 1.6711472272872925, + "learning_rate": 1.4847433727583078e-05, + "loss": 3.2432, + "step": 8949 + }, + { + "epoch": 0.7175642901525325, + "grad_norm": 1.7669073343276978, + "learning_rate": 1.4846297991662053e-05, + "loss": 3.2682, + "step": 8950 + }, + { + "epoch": 0.7176444649335552, + "grad_norm": 0.7531901597976685, + "learning_rate": 1.4845162174034153e-05, + "loss": 2.4707, + "step": 8951 + }, + { + "epoch": 0.7177246397145778, + "grad_norm": 1.8650873899459839, + "learning_rate": 1.4844026274718524e-05, + "loss": 3.3051, + "step": 8952 + }, + { + "epoch": 0.7178048144956004, + "grad_norm": 1.7818944454193115, + "learning_rate": 1.484289029373432e-05, + "loss": 3.2234, + "step": 8953 + }, + { + "epoch": 0.717884989276623, + "grad_norm": 1.784804105758667, + "learning_rate": 1.4841754231100693e-05, + "loss": 3.1495, + "step": 8954 + }, + { + "epoch": 0.7179651640576457, + "grad_norm": 1.9547330141067505, + "learning_rate": 1.4840618086836795e-05, + "loss": 3.2903, + "step": 8955 + }, + { + "epoch": 0.7180453388386683, + "grad_norm": 2.3657593727111816, + "learning_rate": 1.4839481860961785e-05, + "loss": 3.2304, + "step": 8956 + }, + { + "epoch": 0.7181255136196909, + "grad_norm": 1.7073521614074707, + "learning_rate": 1.4838345553494812e-05, + "loss": 3.3434, + "step": 8957 + }, + { + "epoch": 0.7182056884007135, + "grad_norm": 1.844106912612915, + "learning_rate": 1.483720916445504e-05, + "loss": 3.1863, + "step": 8958 + }, + { + "epoch": 0.7182858631817362, + "grad_norm": 1.7290937900543213, + "learning_rate": 1.483607269386163e-05, + "loss": 3.3314, + "step": 8959 + }, + { + "epoch": 0.7183660379627588, + "grad_norm": 1.8982388973236084, + "learning_rate": 1.4834936141733733e-05, + "loss": 3.2292, + "step": 8960 + }, + { + "epoch": 0.7184462127437814, + "grad_norm": 1.7454384565353394, + "learning_rate": 1.483379950809052e-05, + "loss": 3.2075, + "step": 8961 + }, + { + "epoch": 0.718526387524804, + "grad_norm": 1.7934376001358032, + "learning_rate": 1.4832662792951151e-05, + "loss": 3.1367, + "step": 8962 + }, + { + "epoch": 0.7186065623058268, + "grad_norm": 1.8973203897476196, + "learning_rate": 1.4831525996334788e-05, + "loss": 3.1891, + "step": 8963 + }, + { + "epoch": 0.7186867370868494, + "grad_norm": 1.9755357503890991, + "learning_rate": 1.4830389118260602e-05, + "loss": 3.1975, + "step": 8964 + }, + { + "epoch": 0.718766911867872, + "grad_norm": 1.7734776735305786, + "learning_rate": 1.4829252158747758e-05, + "loss": 3.1626, + "step": 8965 + }, + { + "epoch": 0.7188470866488946, + "grad_norm": 1.757359504699707, + "learning_rate": 1.4828115117815424e-05, + "loss": 3.1595, + "step": 8966 + }, + { + "epoch": 0.7189272614299173, + "grad_norm": 1.7788859605789185, + "learning_rate": 1.482697799548277e-05, + "loss": 3.1527, + "step": 8967 + }, + { + "epoch": 0.7190074362109399, + "grad_norm": 1.7629895210266113, + "learning_rate": 1.4825840791768966e-05, + "loss": 3.2031, + "step": 8968 + }, + { + "epoch": 0.7190876109919625, + "grad_norm": 1.8325048685073853, + "learning_rate": 1.4824703506693192e-05, + "loss": 3.2245, + "step": 8969 + }, + { + "epoch": 0.7191677857729851, + "grad_norm": 1.8016953468322754, + "learning_rate": 1.4823566140274613e-05, + "loss": 3.2562, + "step": 8970 + }, + { + "epoch": 0.7192479605540077, + "grad_norm": 1.903732180595398, + "learning_rate": 1.4822428692532412e-05, + "loss": 3.2234, + "step": 8971 + }, + { + "epoch": 0.7193281353350304, + "grad_norm": 0.7280983328819275, + "learning_rate": 1.482129116348576e-05, + "loss": 2.4845, + "step": 8972 + }, + { + "epoch": 0.719408310116053, + "grad_norm": 1.8113375902175903, + "learning_rate": 1.482015355315384e-05, + "loss": 3.3965, + "step": 8973 + }, + { + "epoch": 0.7194884848970756, + "grad_norm": 1.7849076986312866, + "learning_rate": 1.481901586155583e-05, + "loss": 3.1458, + "step": 8974 + }, + { + "epoch": 0.7195686596780982, + "grad_norm": 1.8563178777694702, + "learning_rate": 1.4817878088710908e-05, + "loss": 3.1649, + "step": 8975 + }, + { + "epoch": 0.7196488344591209, + "grad_norm": 1.78612220287323, + "learning_rate": 1.4816740234638265e-05, + "loss": 3.2233, + "step": 8976 + }, + { + "epoch": 0.7197290092401435, + "grad_norm": 1.7452192306518555, + "learning_rate": 1.4815602299357074e-05, + "loss": 3.2736, + "step": 8977 + }, + { + "epoch": 0.7198091840211661, + "grad_norm": 1.8546524047851562, + "learning_rate": 1.4814464282886528e-05, + "loss": 3.2428, + "step": 8978 + }, + { + "epoch": 0.7198893588021887, + "grad_norm": 1.949039340019226, + "learning_rate": 1.4813326185245811e-05, + "loss": 3.3362, + "step": 8979 + }, + { + "epoch": 0.7199695335832114, + "grad_norm": 0.7560471892356873, + "learning_rate": 1.481218800645411e-05, + "loss": 2.5004, + "step": 8980 + }, + { + "epoch": 0.720049708364234, + "grad_norm": 1.947050929069519, + "learning_rate": 1.4811049746530616e-05, + "loss": 3.1665, + "step": 8981 + }, + { + "epoch": 0.7201298831452566, + "grad_norm": 1.836410403251648, + "learning_rate": 1.480991140549452e-05, + "loss": 3.2657, + "step": 8982 + }, + { + "epoch": 0.7202100579262792, + "grad_norm": 1.7534350156784058, + "learning_rate": 1.4808772983365009e-05, + "loss": 3.2085, + "step": 8983 + }, + { + "epoch": 0.720290232707302, + "grad_norm": 1.8570513725280762, + "learning_rate": 1.4807634480161284e-05, + "loss": 3.318, + "step": 8984 + }, + { + "epoch": 0.7203704074883246, + "grad_norm": 1.80226469039917, + "learning_rate": 1.4806495895902533e-05, + "loss": 3.2077, + "step": 8985 + }, + { + "epoch": 0.7204505822693472, + "grad_norm": 1.7579189538955688, + "learning_rate": 1.480535723060796e-05, + "loss": 3.1868, + "step": 8986 + }, + { + "epoch": 0.7205307570503698, + "grad_norm": 1.7709709405899048, + "learning_rate": 1.4804218484296753e-05, + "loss": 3.2632, + "step": 8987 + }, + { + "epoch": 0.7206109318313925, + "grad_norm": 0.77532559633255, + "learning_rate": 1.4803079656988115e-05, + "loss": 2.5061, + "step": 8988 + }, + { + "epoch": 0.7206911066124151, + "grad_norm": 1.7359956502914429, + "learning_rate": 1.480194074870125e-05, + "loss": 3.2449, + "step": 8989 + }, + { + "epoch": 0.7207712813934377, + "grad_norm": 1.7047072649002075, + "learning_rate": 1.4800801759455355e-05, + "loss": 3.1885, + "step": 8990 + }, + { + "epoch": 0.7208514561744603, + "grad_norm": 1.6944485902786255, + "learning_rate": 1.4799662689269636e-05, + "loss": 3.0881, + "step": 8991 + }, + { + "epoch": 0.720931630955483, + "grad_norm": 1.6737905740737915, + "learning_rate": 1.4798523538163291e-05, + "loss": 3.278, + "step": 8992 + }, + { + "epoch": 0.7210118057365056, + "grad_norm": 1.7423663139343262, + "learning_rate": 1.4797384306155533e-05, + "loss": 3.308, + "step": 8993 + }, + { + "epoch": 0.7210919805175282, + "grad_norm": 1.8353971242904663, + "learning_rate": 1.479624499326557e-05, + "loss": 3.3314, + "step": 8994 + }, + { + "epoch": 0.7211721552985508, + "grad_norm": 1.8536851406097412, + "learning_rate": 1.4795105599512602e-05, + "loss": 3.1869, + "step": 8995 + }, + { + "epoch": 0.7212523300795735, + "grad_norm": 1.7742500305175781, + "learning_rate": 1.4793966124915846e-05, + "loss": 3.223, + "step": 8996 + }, + { + "epoch": 0.7213325048605961, + "grad_norm": 1.781046986579895, + "learning_rate": 1.4792826569494509e-05, + "loss": 3.1536, + "step": 8997 + }, + { + "epoch": 0.7214126796416187, + "grad_norm": 0.7567793726921082, + "learning_rate": 1.4791686933267806e-05, + "loss": 2.5009, + "step": 8998 + }, + { + "epoch": 0.7214928544226413, + "grad_norm": 1.77732253074646, + "learning_rate": 1.4790547216254952e-05, + "loss": 3.2433, + "step": 8999 + }, + { + "epoch": 0.721573029203664, + "grad_norm": 1.812817931175232, + "learning_rate": 1.4789407418475157e-05, + "loss": 3.1506, + "step": 9000 + }, + { + "epoch": 0.7216532039846866, + "grad_norm": 1.7334203720092773, + "learning_rate": 1.4788267539947642e-05, + "loss": 3.237, + "step": 9001 + }, + { + "epoch": 0.7217333787657092, + "grad_norm": 1.7076523303985596, + "learning_rate": 1.4787127580691623e-05, + "loss": 3.2585, + "step": 9002 + }, + { + "epoch": 0.7218135535467318, + "grad_norm": 1.9257248640060425, + "learning_rate": 1.4785987540726321e-05, + "loss": 3.3357, + "step": 9003 + }, + { + "epoch": 0.7218937283277546, + "grad_norm": 1.6982282400131226, + "learning_rate": 1.4784847420070955e-05, + "loss": 3.2353, + "step": 9004 + }, + { + "epoch": 0.7219739031087772, + "grad_norm": 1.86944580078125, + "learning_rate": 1.4783707218744749e-05, + "loss": 3.1624, + "step": 9005 + }, + { + "epoch": 0.7220540778897998, + "grad_norm": 1.6708873510360718, + "learning_rate": 1.4782566936766927e-05, + "loss": 3.1593, + "step": 9006 + }, + { + "epoch": 0.7221342526708224, + "grad_norm": 1.8967093229293823, + "learning_rate": 1.478142657415671e-05, + "loss": 3.2365, + "step": 9007 + }, + { + "epoch": 0.722214427451845, + "grad_norm": 1.7642441987991333, + "learning_rate": 1.4780286130933323e-05, + "loss": 3.1895, + "step": 9008 + }, + { + "epoch": 0.7222946022328677, + "grad_norm": 2.017218589782715, + "learning_rate": 1.4779145607116002e-05, + "loss": 3.2178, + "step": 9009 + }, + { + "epoch": 0.7223747770138903, + "grad_norm": 1.845698595046997, + "learning_rate": 1.4778005002723967e-05, + "loss": 3.3286, + "step": 9010 + }, + { + "epoch": 0.7224549517949129, + "grad_norm": 1.7700459957122803, + "learning_rate": 1.4776864317776454e-05, + "loss": 3.1344, + "step": 9011 + }, + { + "epoch": 0.7225351265759355, + "grad_norm": 1.7666722536087036, + "learning_rate": 1.477572355229269e-05, + "loss": 3.1837, + "step": 9012 + }, + { + "epoch": 0.7226153013569582, + "grad_norm": 1.826788306236267, + "learning_rate": 1.477458270629191e-05, + "loss": 3.2498, + "step": 9013 + }, + { + "epoch": 0.7226954761379808, + "grad_norm": 1.833862543106079, + "learning_rate": 1.477344177979335e-05, + "loss": 3.2264, + "step": 9014 + }, + { + "epoch": 0.7227756509190034, + "grad_norm": 1.8430882692337036, + "learning_rate": 1.4772300772816241e-05, + "loss": 3.1484, + "step": 9015 + }, + { + "epoch": 0.722855825700026, + "grad_norm": 1.7367337942123413, + "learning_rate": 1.4771159685379828e-05, + "loss": 3.2328, + "step": 9016 + }, + { + "epoch": 0.7229360004810487, + "grad_norm": 1.7463966608047485, + "learning_rate": 1.477001851750334e-05, + "loss": 3.1348, + "step": 9017 + }, + { + "epoch": 0.7230161752620713, + "grad_norm": 1.8887146711349487, + "learning_rate": 1.4768877269206021e-05, + "loss": 3.2584, + "step": 9018 + }, + { + "epoch": 0.7230963500430939, + "grad_norm": 1.8530901670455933, + "learning_rate": 1.4767735940507112e-05, + "loss": 3.2896, + "step": 9019 + }, + { + "epoch": 0.7231765248241165, + "grad_norm": 1.851801872253418, + "learning_rate": 1.4766594531425858e-05, + "loss": 3.229, + "step": 9020 + }, + { + "epoch": 0.7232566996051392, + "grad_norm": 1.7750244140625, + "learning_rate": 1.4765453041981498e-05, + "loss": 3.2278, + "step": 9021 + }, + { + "epoch": 0.7233368743861618, + "grad_norm": 1.8054572343826294, + "learning_rate": 1.476431147219328e-05, + "loss": 3.2404, + "step": 9022 + }, + { + "epoch": 0.7234170491671844, + "grad_norm": 1.7743014097213745, + "learning_rate": 1.4763169822080447e-05, + "loss": 3.29, + "step": 9023 + }, + { + "epoch": 0.7234972239482071, + "grad_norm": 2.0265390872955322, + "learning_rate": 1.4762028091662255e-05, + "loss": 3.1304, + "step": 9024 + }, + { + "epoch": 0.7235773987292298, + "grad_norm": 1.8603770732879639, + "learning_rate": 1.4760886280957942e-05, + "loss": 3.1993, + "step": 9025 + }, + { + "epoch": 0.7236575735102524, + "grad_norm": 1.8874610662460327, + "learning_rate": 1.475974438998677e-05, + "loss": 3.2581, + "step": 9026 + }, + { + "epoch": 0.723737748291275, + "grad_norm": 1.6929912567138672, + "learning_rate": 1.475860241876798e-05, + "loss": 3.2394, + "step": 9027 + }, + { + "epoch": 0.7238179230722976, + "grad_norm": 1.824115514755249, + "learning_rate": 1.4757460367320831e-05, + "loss": 3.1483, + "step": 9028 + }, + { + "epoch": 0.7238980978533203, + "grad_norm": 1.8863292932510376, + "learning_rate": 1.475631823566458e-05, + "loss": 3.1465, + "step": 9029 + }, + { + "epoch": 0.7239782726343429, + "grad_norm": 1.7002719640731812, + "learning_rate": 1.4755176023818477e-05, + "loss": 3.2351, + "step": 9030 + }, + { + "epoch": 0.7240584474153655, + "grad_norm": 1.8850864171981812, + "learning_rate": 1.4754033731801784e-05, + "loss": 3.2585, + "step": 9031 + }, + { + "epoch": 0.7241386221963881, + "grad_norm": 1.7102932929992676, + "learning_rate": 1.4752891359633756e-05, + "loss": 3.1796, + "step": 9032 + }, + { + "epoch": 0.7242187969774108, + "grad_norm": 1.7968418598175049, + "learning_rate": 1.4751748907333654e-05, + "loss": 3.2834, + "step": 9033 + }, + { + "epoch": 0.7242989717584334, + "grad_norm": 1.8300981521606445, + "learning_rate": 1.4750606374920744e-05, + "loss": 3.285, + "step": 9034 + }, + { + "epoch": 0.724379146539456, + "grad_norm": 1.7590646743774414, + "learning_rate": 1.474946376241428e-05, + "loss": 3.1351, + "step": 9035 + }, + { + "epoch": 0.7244593213204786, + "grad_norm": 1.9060511589050293, + "learning_rate": 1.4748321069833536e-05, + "loss": 3.2865, + "step": 9036 + }, + { + "epoch": 0.7245394961015013, + "grad_norm": 1.9176324605941772, + "learning_rate": 1.4747178297197769e-05, + "loss": 3.3109, + "step": 9037 + }, + { + "epoch": 0.7246196708825239, + "grad_norm": 1.9984533786773682, + "learning_rate": 1.4746035444526248e-05, + "loss": 3.1655, + "step": 9038 + }, + { + "epoch": 0.7246998456635465, + "grad_norm": 0.7692698836326599, + "learning_rate": 1.4744892511838244e-05, + "loss": 2.4908, + "step": 9039 + }, + { + "epoch": 0.7247800204445691, + "grad_norm": 1.6608259677886963, + "learning_rate": 1.4743749499153024e-05, + "loss": 3.1223, + "step": 9040 + }, + { + "epoch": 0.7248601952255918, + "grad_norm": 0.7300740480422974, + "learning_rate": 1.474260640648986e-05, + "loss": 2.506, + "step": 9041 + }, + { + "epoch": 0.7249403700066144, + "grad_norm": 1.7744673490524292, + "learning_rate": 1.4741463233868021e-05, + "loss": 3.2794, + "step": 9042 + }, + { + "epoch": 0.725020544787637, + "grad_norm": 1.7425211668014526, + "learning_rate": 1.4740319981306786e-05, + "loss": 3.2018, + "step": 9043 + }, + { + "epoch": 0.7251007195686596, + "grad_norm": 2.062213659286499, + "learning_rate": 1.4739176648825425e-05, + "loss": 3.2478, + "step": 9044 + }, + { + "epoch": 0.7251808943496824, + "grad_norm": 1.78277587890625, + "learning_rate": 1.4738033236443217e-05, + "loss": 3.132, + "step": 9045 + }, + { + "epoch": 0.725261069130705, + "grad_norm": 1.9079645872116089, + "learning_rate": 1.4736889744179437e-05, + "loss": 3.2842, + "step": 9046 + }, + { + "epoch": 0.7253412439117276, + "grad_norm": 1.8484952449798584, + "learning_rate": 1.4735746172053367e-05, + "loss": 3.2498, + "step": 9047 + }, + { + "epoch": 0.7254214186927502, + "grad_norm": 1.8921047449111938, + "learning_rate": 1.4734602520084284e-05, + "loss": 3.3181, + "step": 9048 + }, + { + "epoch": 0.7255015934737729, + "grad_norm": 1.992827296257019, + "learning_rate": 1.4733458788291472e-05, + "loss": 3.164, + "step": 9049 + }, + { + "epoch": 0.7255817682547955, + "grad_norm": 1.6588728427886963, + "learning_rate": 1.4732314976694213e-05, + "loss": 3.2463, + "step": 9050 + }, + { + "epoch": 0.7256619430358181, + "grad_norm": 1.8606537580490112, + "learning_rate": 1.473117108531179e-05, + "loss": 3.2366, + "step": 9051 + }, + { + "epoch": 0.7257421178168407, + "grad_norm": 1.8146448135375977, + "learning_rate": 1.4730027114163492e-05, + "loss": 3.2643, + "step": 9052 + }, + { + "epoch": 0.7258222925978633, + "grad_norm": 2.001427173614502, + "learning_rate": 1.4728883063268605e-05, + "loss": 3.3384, + "step": 9053 + }, + { + "epoch": 0.725902467378886, + "grad_norm": 1.9184463024139404, + "learning_rate": 1.4727738932646413e-05, + "loss": 3.3086, + "step": 9054 + }, + { + "epoch": 0.7259826421599086, + "grad_norm": 1.9810972213745117, + "learning_rate": 1.472659472231621e-05, + "loss": 3.3058, + "step": 9055 + }, + { + "epoch": 0.7260628169409312, + "grad_norm": 1.8060566186904907, + "learning_rate": 1.4725450432297286e-05, + "loss": 3.1955, + "step": 9056 + }, + { + "epoch": 0.7261429917219538, + "grad_norm": 1.7861803770065308, + "learning_rate": 1.4724306062608934e-05, + "loss": 3.1046, + "step": 9057 + }, + { + "epoch": 0.7262231665029765, + "grad_norm": 1.8855764865875244, + "learning_rate": 1.4723161613270444e-05, + "loss": 3.2292, + "step": 9058 + }, + { + "epoch": 0.7263033412839991, + "grad_norm": 1.836202621459961, + "learning_rate": 1.4722017084301117e-05, + "loss": 3.228, + "step": 9059 + }, + { + "epoch": 0.7263835160650217, + "grad_norm": 1.7350982427597046, + "learning_rate": 1.4720872475720244e-05, + "loss": 3.2065, + "step": 9060 + }, + { + "epoch": 0.7264636908460443, + "grad_norm": 1.7375110387802124, + "learning_rate": 1.4719727787547124e-05, + "loss": 3.2474, + "step": 9061 + }, + { + "epoch": 0.726543865627067, + "grad_norm": 2.1155495643615723, + "learning_rate": 1.471858301980106e-05, + "loss": 3.2557, + "step": 9062 + }, + { + "epoch": 0.7266240404080896, + "grad_norm": 1.7397962808609009, + "learning_rate": 1.4717438172501348e-05, + "loss": 3.1509, + "step": 9063 + }, + { + "epoch": 0.7267042151891122, + "grad_norm": 1.918853759765625, + "learning_rate": 1.4716293245667293e-05, + "loss": 3.2795, + "step": 9064 + }, + { + "epoch": 0.726784389970135, + "grad_norm": 1.9037498235702515, + "learning_rate": 1.4715148239318194e-05, + "loss": 3.1213, + "step": 9065 + }, + { + "epoch": 0.7268645647511576, + "grad_norm": 1.8258299827575684, + "learning_rate": 1.4714003153473356e-05, + "loss": 3.2581, + "step": 9066 + }, + { + "epoch": 0.7269447395321802, + "grad_norm": 1.9933305978775024, + "learning_rate": 1.4712857988152087e-05, + "loss": 3.1842, + "step": 9067 + }, + { + "epoch": 0.7270249143132028, + "grad_norm": 1.881076455116272, + "learning_rate": 1.4711712743373693e-05, + "loss": 3.2876, + "step": 9068 + }, + { + "epoch": 0.7271050890942254, + "grad_norm": 1.8584604263305664, + "learning_rate": 1.4710567419157487e-05, + "loss": 3.1742, + "step": 9069 + }, + { + "epoch": 0.7271852638752481, + "grad_norm": 2.023881673812866, + "learning_rate": 1.470942201552277e-05, + "loss": 3.1935, + "step": 9070 + }, + { + "epoch": 0.7272654386562707, + "grad_norm": 1.9765480756759644, + "learning_rate": 1.4708276532488857e-05, + "loss": 3.2926, + "step": 9071 + }, + { + "epoch": 0.7273456134372933, + "grad_norm": 1.7380033731460571, + "learning_rate": 1.4707130970075065e-05, + "loss": 3.1227, + "step": 9072 + }, + { + "epoch": 0.7274257882183159, + "grad_norm": 1.7671735286712646, + "learning_rate": 1.47059853283007e-05, + "loss": 3.2395, + "step": 9073 + }, + { + "epoch": 0.7275059629993386, + "grad_norm": 2.223219633102417, + "learning_rate": 1.4704839607185084e-05, + "loss": 3.2365, + "step": 9074 + }, + { + "epoch": 0.7275861377803612, + "grad_norm": 1.8779579401016235, + "learning_rate": 1.4703693806747527e-05, + "loss": 3.2646, + "step": 9075 + }, + { + "epoch": 0.7276663125613838, + "grad_norm": 1.808550477027893, + "learning_rate": 1.470254792700735e-05, + "loss": 3.3047, + "step": 9076 + }, + { + "epoch": 0.7277464873424064, + "grad_norm": 1.7717534303665161, + "learning_rate": 1.4701401967983877e-05, + "loss": 3.3358, + "step": 9077 + }, + { + "epoch": 0.7278266621234291, + "grad_norm": 1.7693586349487305, + "learning_rate": 1.4700255929696421e-05, + "loss": 3.1553, + "step": 9078 + }, + { + "epoch": 0.7279068369044517, + "grad_norm": 1.8965165615081787, + "learning_rate": 1.4699109812164307e-05, + "loss": 3.1991, + "step": 9079 + }, + { + "epoch": 0.7279870116854743, + "grad_norm": 1.897377848625183, + "learning_rate": 1.4697963615406856e-05, + "loss": 3.2599, + "step": 9080 + }, + { + "epoch": 0.7280671864664969, + "grad_norm": 1.765762448310852, + "learning_rate": 1.4696817339443395e-05, + "loss": 3.0816, + "step": 9081 + }, + { + "epoch": 0.7281473612475196, + "grad_norm": 1.9700134992599487, + "learning_rate": 1.469567098429325e-05, + "loss": 3.2637, + "step": 9082 + }, + { + "epoch": 0.7282275360285422, + "grad_norm": 1.7872092723846436, + "learning_rate": 1.4694524549975747e-05, + "loss": 3.2601, + "step": 9083 + }, + { + "epoch": 0.7283077108095648, + "grad_norm": 1.9024877548217773, + "learning_rate": 1.4693378036510212e-05, + "loss": 3.2865, + "step": 9084 + }, + { + "epoch": 0.7283878855905875, + "grad_norm": 1.8602373600006104, + "learning_rate": 1.4692231443915981e-05, + "loss": 3.1757, + "step": 9085 + }, + { + "epoch": 0.7284680603716102, + "grad_norm": 0.7664164304733276, + "learning_rate": 1.469108477221238e-05, + "loss": 2.4987, + "step": 9086 + }, + { + "epoch": 0.7285482351526328, + "grad_norm": 1.8608617782592773, + "learning_rate": 1.4689938021418743e-05, + "loss": 3.2401, + "step": 9087 + }, + { + "epoch": 0.7286284099336554, + "grad_norm": 1.9030511379241943, + "learning_rate": 1.4688791191554403e-05, + "loss": 3.192, + "step": 9088 + }, + { + "epoch": 0.728708584714678, + "grad_norm": 1.8867467641830444, + "learning_rate": 1.4687644282638698e-05, + "loss": 3.1721, + "step": 9089 + }, + { + "epoch": 0.7287887594957007, + "grad_norm": 1.9380091428756714, + "learning_rate": 1.468649729469096e-05, + "loss": 3.3611, + "step": 9090 + }, + { + "epoch": 0.7288689342767233, + "grad_norm": 1.7908966541290283, + "learning_rate": 1.468535022773053e-05, + "loss": 3.3601, + "step": 9091 + }, + { + "epoch": 0.7289491090577459, + "grad_norm": 1.7236188650131226, + "learning_rate": 1.4684203081776749e-05, + "loss": 3.1052, + "step": 9092 + }, + { + "epoch": 0.7290292838387685, + "grad_norm": 1.7038919925689697, + "learning_rate": 1.4683055856848953e-05, + "loss": 3.1718, + "step": 9093 + }, + { + "epoch": 0.7291094586197912, + "grad_norm": 1.8209257125854492, + "learning_rate": 1.4681908552966485e-05, + "loss": 3.1997, + "step": 9094 + }, + { + "epoch": 0.7291896334008138, + "grad_norm": 1.8668720722198486, + "learning_rate": 1.468076117014869e-05, + "loss": 3.2073, + "step": 9095 + }, + { + "epoch": 0.7292698081818364, + "grad_norm": 1.6922403573989868, + "learning_rate": 1.467961370841491e-05, + "loss": 3.1801, + "step": 9096 + }, + { + "epoch": 0.729349982962859, + "grad_norm": 1.832251787185669, + "learning_rate": 1.4678466167784495e-05, + "loss": 3.1519, + "step": 9097 + }, + { + "epoch": 0.7294301577438816, + "grad_norm": 1.8010810613632202, + "learning_rate": 1.4677318548276787e-05, + "loss": 3.2269, + "step": 9098 + }, + { + "epoch": 0.7295103325249043, + "grad_norm": 1.936252474784851, + "learning_rate": 1.4676170849911138e-05, + "loss": 3.3223, + "step": 9099 + }, + { + "epoch": 0.7295905073059269, + "grad_norm": 1.7114561796188354, + "learning_rate": 1.4675023072706895e-05, + "loss": 3.1444, + "step": 9100 + }, + { + "epoch": 0.7296706820869495, + "grad_norm": 0.7767108082771301, + "learning_rate": 1.4673875216683412e-05, + "loss": 2.4824, + "step": 9101 + }, + { + "epoch": 0.7297508568679721, + "grad_norm": 1.7356443405151367, + "learning_rate": 1.467272728186004e-05, + "loss": 3.2096, + "step": 9102 + }, + { + "epoch": 0.7298310316489948, + "grad_norm": 1.800538182258606, + "learning_rate": 1.467157926825613e-05, + "loss": 3.3315, + "step": 9103 + }, + { + "epoch": 0.7299112064300174, + "grad_norm": 1.781248688697815, + "learning_rate": 1.4670431175891043e-05, + "loss": 3.1035, + "step": 9104 + }, + { + "epoch": 0.72999138121104, + "grad_norm": 1.6669715642929077, + "learning_rate": 1.4669283004784132e-05, + "loss": 3.2217, + "step": 9105 + }, + { + "epoch": 0.7300715559920627, + "grad_norm": 1.702516794204712, + "learning_rate": 1.4668134754954753e-05, + "loss": 3.2475, + "step": 9106 + }, + { + "epoch": 0.7301517307730854, + "grad_norm": 1.840153455734253, + "learning_rate": 1.466698642642227e-05, + "loss": 3.1339, + "step": 9107 + }, + { + "epoch": 0.730231905554108, + "grad_norm": 0.7530494928359985, + "learning_rate": 1.4665838019206038e-05, + "loss": 2.4929, + "step": 9108 + }, + { + "epoch": 0.7303120803351306, + "grad_norm": 1.9389560222625732, + "learning_rate": 1.4664689533325425e-05, + "loss": 3.2126, + "step": 9109 + }, + { + "epoch": 0.7303922551161532, + "grad_norm": 1.9338592290878296, + "learning_rate": 1.4663540968799787e-05, + "loss": 3.2952, + "step": 9110 + }, + { + "epoch": 0.7304724298971759, + "grad_norm": 2.064993143081665, + "learning_rate": 1.4662392325648493e-05, + "loss": 3.2153, + "step": 9111 + }, + { + "epoch": 0.7305526046781985, + "grad_norm": 1.952509880065918, + "learning_rate": 1.4661243603890909e-05, + "loss": 3.2178, + "step": 9112 + }, + { + "epoch": 0.7306327794592211, + "grad_norm": 2.243603467941284, + "learning_rate": 1.4660094803546398e-05, + "loss": 3.385, + "step": 9113 + }, + { + "epoch": 0.7307129542402437, + "grad_norm": 1.8647149801254272, + "learning_rate": 1.4658945924634336e-05, + "loss": 3.1167, + "step": 9114 + }, + { + "epoch": 0.7307931290212664, + "grad_norm": 1.8482329845428467, + "learning_rate": 1.4657796967174082e-05, + "loss": 3.1456, + "step": 9115 + }, + { + "epoch": 0.730873303802289, + "grad_norm": 1.7492469549179077, + "learning_rate": 1.4656647931185015e-05, + "loss": 3.2625, + "step": 9116 + }, + { + "epoch": 0.7309534785833116, + "grad_norm": 1.978061556816101, + "learning_rate": 1.4655498816686505e-05, + "loss": 3.1484, + "step": 9117 + }, + { + "epoch": 0.7310336533643342, + "grad_norm": 1.8617942333221436, + "learning_rate": 1.4654349623697925e-05, + "loss": 3.2159, + "step": 9118 + }, + { + "epoch": 0.7311138281453569, + "grad_norm": 1.8897732496261597, + "learning_rate": 1.4653200352238652e-05, + "loss": 3.339, + "step": 9119 + }, + { + "epoch": 0.7311940029263795, + "grad_norm": 1.6965982913970947, + "learning_rate": 1.465205100232806e-05, + "loss": 3.2405, + "step": 9120 + }, + { + "epoch": 0.7312741777074021, + "grad_norm": 2.1429696083068848, + "learning_rate": 1.4650901573985527e-05, + "loss": 3.106, + "step": 9121 + }, + { + "epoch": 0.7313543524884247, + "grad_norm": 1.7214232683181763, + "learning_rate": 1.4649752067230437e-05, + "loss": 3.1395, + "step": 9122 + }, + { + "epoch": 0.7314345272694474, + "grad_norm": 1.7787011861801147, + "learning_rate": 1.4648602482082159e-05, + "loss": 3.1905, + "step": 9123 + }, + { + "epoch": 0.73151470205047, + "grad_norm": 1.8916594982147217, + "learning_rate": 1.4647452818560086e-05, + "loss": 3.2404, + "step": 9124 + }, + { + "epoch": 0.7315948768314926, + "grad_norm": 1.7980475425720215, + "learning_rate": 1.4646303076683594e-05, + "loss": 3.2421, + "step": 9125 + }, + { + "epoch": 0.7316750516125153, + "grad_norm": 1.830019474029541, + "learning_rate": 1.464515325647207e-05, + "loss": 3.2208, + "step": 9126 + }, + { + "epoch": 0.731755226393538, + "grad_norm": 1.8398945331573486, + "learning_rate": 1.4644003357944903e-05, + "loss": 3.2622, + "step": 9127 + }, + { + "epoch": 0.7318354011745606, + "grad_norm": 1.7732075452804565, + "learning_rate": 1.4642853381121472e-05, + "loss": 3.3059, + "step": 9128 + }, + { + "epoch": 0.7319155759555832, + "grad_norm": 1.733600378036499, + "learning_rate": 1.4641703326021173e-05, + "loss": 3.2455, + "step": 9129 + }, + { + "epoch": 0.7319957507366058, + "grad_norm": 1.8300769329071045, + "learning_rate": 1.4640553192663387e-05, + "loss": 3.247, + "step": 9130 + }, + { + "epoch": 0.7320759255176285, + "grad_norm": 1.7412047386169434, + "learning_rate": 1.4639402981067512e-05, + "loss": 3.2368, + "step": 9131 + }, + { + "epoch": 0.7321561002986511, + "grad_norm": 1.7586320638656616, + "learning_rate": 1.4638252691252938e-05, + "loss": 3.1864, + "step": 9132 + }, + { + "epoch": 0.7322362750796737, + "grad_norm": 1.83353590965271, + "learning_rate": 1.4637102323239058e-05, + "loss": 3.0895, + "step": 9133 + }, + { + "epoch": 0.7323164498606963, + "grad_norm": 1.9235745668411255, + "learning_rate": 1.4635951877045269e-05, + "loss": 3.1964, + "step": 9134 + }, + { + "epoch": 0.732396624641719, + "grad_norm": 0.7454771399497986, + "learning_rate": 1.4634801352690962e-05, + "loss": 2.4846, + "step": 9135 + }, + { + "epoch": 0.7324767994227416, + "grad_norm": 1.8303146362304688, + "learning_rate": 1.4633650750195538e-05, + "loss": 3.0936, + "step": 9136 + }, + { + "epoch": 0.7325569742037642, + "grad_norm": 1.7686394453048706, + "learning_rate": 1.46325000695784e-05, + "loss": 3.1787, + "step": 9137 + }, + { + "epoch": 0.7326371489847868, + "grad_norm": 2.027472972869873, + "learning_rate": 1.4631349310858938e-05, + "loss": 3.2647, + "step": 9138 + }, + { + "epoch": 0.7327173237658094, + "grad_norm": 1.670059323310852, + "learning_rate": 1.4630198474056563e-05, + "loss": 3.1077, + "step": 9139 + }, + { + "epoch": 0.7327974985468321, + "grad_norm": 1.8923529386520386, + "learning_rate": 1.462904755919067e-05, + "loss": 3.2749, + "step": 9140 + }, + { + "epoch": 0.7328776733278547, + "grad_norm": 1.861138105392456, + "learning_rate": 1.4627896566280667e-05, + "loss": 3.2706, + "step": 9141 + }, + { + "epoch": 0.7329578481088773, + "grad_norm": 1.8974143266677856, + "learning_rate": 1.462674549534596e-05, + "loss": 3.1404, + "step": 9142 + }, + { + "epoch": 0.7330380228898999, + "grad_norm": 1.7904826402664185, + "learning_rate": 1.4625594346405953e-05, + "loss": 3.2699, + "step": 9143 + }, + { + "epoch": 0.7331181976709226, + "grad_norm": 1.8105649948120117, + "learning_rate": 1.4624443119480061e-05, + "loss": 3.2074, + "step": 9144 + }, + { + "epoch": 0.7331983724519452, + "grad_norm": 1.9573026895523071, + "learning_rate": 1.4623291814587682e-05, + "loss": 3.2921, + "step": 9145 + }, + { + "epoch": 0.7332785472329678, + "grad_norm": 1.8374017477035522, + "learning_rate": 1.4622140431748232e-05, + "loss": 3.1641, + "step": 9146 + }, + { + "epoch": 0.7333587220139905, + "grad_norm": 1.860245704650879, + "learning_rate": 1.462098897098113e-05, + "loss": 3.1852, + "step": 9147 + }, + { + "epoch": 0.7334388967950132, + "grad_norm": 1.8720858097076416, + "learning_rate": 1.4619837432305777e-05, + "loss": 3.1973, + "step": 9148 + }, + { + "epoch": 0.7335190715760358, + "grad_norm": 1.8039758205413818, + "learning_rate": 1.4618685815741595e-05, + "loss": 3.2126, + "step": 9149 + }, + { + "epoch": 0.7335992463570584, + "grad_norm": 1.8086092472076416, + "learning_rate": 1.4617534121307996e-05, + "loss": 3.2559, + "step": 9150 + }, + { + "epoch": 0.733679421138081, + "grad_norm": 1.8814061880111694, + "learning_rate": 1.46163823490244e-05, + "loss": 3.1723, + "step": 9151 + }, + { + "epoch": 0.7337595959191037, + "grad_norm": 1.8055630922317505, + "learning_rate": 1.4615230498910225e-05, + "loss": 3.2512, + "step": 9152 + }, + { + "epoch": 0.7338397707001263, + "grad_norm": 1.8141933679580688, + "learning_rate": 1.461407857098489e-05, + "loss": 3.192, + "step": 9153 + }, + { + "epoch": 0.7339199454811489, + "grad_norm": 1.8660621643066406, + "learning_rate": 1.4612926565267818e-05, + "loss": 3.1623, + "step": 9154 + }, + { + "epoch": 0.7340001202621715, + "grad_norm": 1.7700337171554565, + "learning_rate": 1.4611774481778427e-05, + "loss": 3.2794, + "step": 9155 + }, + { + "epoch": 0.7340802950431942, + "grad_norm": 1.9154692888259888, + "learning_rate": 1.4610622320536145e-05, + "loss": 3.1593, + "step": 9156 + }, + { + "epoch": 0.7341604698242168, + "grad_norm": 1.8110896348953247, + "learning_rate": 1.4609470081560395e-05, + "loss": 3.2758, + "step": 9157 + }, + { + "epoch": 0.7342406446052394, + "grad_norm": 1.8226916790008545, + "learning_rate": 1.4608317764870606e-05, + "loss": 3.319, + "step": 9158 + }, + { + "epoch": 0.734320819386262, + "grad_norm": 1.767863392829895, + "learning_rate": 1.4607165370486203e-05, + "loss": 3.2736, + "step": 9159 + }, + { + "epoch": 0.7344009941672847, + "grad_norm": 1.7991684675216675, + "learning_rate": 1.4606012898426614e-05, + "loss": 3.1522, + "step": 9160 + }, + { + "epoch": 0.7344811689483073, + "grad_norm": 1.80435049533844, + "learning_rate": 1.4604860348711269e-05, + "loss": 3.2464, + "step": 9161 + }, + { + "epoch": 0.7345613437293299, + "grad_norm": 1.685530662536621, + "learning_rate": 1.4603707721359609e-05, + "loss": 3.1388, + "step": 9162 + }, + { + "epoch": 0.7346415185103525, + "grad_norm": 1.8178800344467163, + "learning_rate": 1.4602555016391051e-05, + "loss": 3.1331, + "step": 9163 + }, + { + "epoch": 0.7347216932913752, + "grad_norm": 0.7757521867752075, + "learning_rate": 1.4601402233825042e-05, + "loss": 2.4894, + "step": 9164 + }, + { + "epoch": 0.7348018680723978, + "grad_norm": 1.7353521585464478, + "learning_rate": 1.4600249373681013e-05, + "loss": 3.3138, + "step": 9165 + }, + { + "epoch": 0.7348820428534204, + "grad_norm": 1.7444217205047607, + "learning_rate": 1.4599096435978396e-05, + "loss": 3.2024, + "step": 9166 + }, + { + "epoch": 0.7349622176344431, + "grad_norm": 1.8446863889694214, + "learning_rate": 1.4597943420736643e-05, + "loss": 3.3101, + "step": 9167 + }, + { + "epoch": 0.7350423924154658, + "grad_norm": 1.8891104459762573, + "learning_rate": 1.4596790327975177e-05, + "loss": 3.1858, + "step": 9168 + }, + { + "epoch": 0.7351225671964884, + "grad_norm": 1.9239251613616943, + "learning_rate": 1.459563715771345e-05, + "loss": 3.2648, + "step": 9169 + }, + { + "epoch": 0.735202741977511, + "grad_norm": 1.7374367713928223, + "learning_rate": 1.4594483909970896e-05, + "loss": 3.2426, + "step": 9170 + }, + { + "epoch": 0.7352829167585336, + "grad_norm": 1.7740451097488403, + "learning_rate": 1.4593330584766964e-05, + "loss": 3.2071, + "step": 9171 + }, + { + "epoch": 0.7353630915395563, + "grad_norm": 1.857225775718689, + "learning_rate": 1.45921771821211e-05, + "loss": 3.209, + "step": 9172 + }, + { + "epoch": 0.7354432663205789, + "grad_norm": 1.706465721130371, + "learning_rate": 1.4591023702052746e-05, + "loss": 3.2271, + "step": 9173 + }, + { + "epoch": 0.7355234411016015, + "grad_norm": 1.6888412237167358, + "learning_rate": 1.458987014458135e-05, + "loss": 3.3051, + "step": 9174 + }, + { + "epoch": 0.7356036158826241, + "grad_norm": 1.8363208770751953, + "learning_rate": 1.4588716509726362e-05, + "loss": 3.3066, + "step": 9175 + }, + { + "epoch": 0.7356837906636468, + "grad_norm": 1.874128818511963, + "learning_rate": 1.4587562797507229e-05, + "loss": 3.2166, + "step": 9176 + }, + { + "epoch": 0.7357639654446694, + "grad_norm": 1.8275771141052246, + "learning_rate": 1.4586409007943405e-05, + "loss": 3.1315, + "step": 9177 + }, + { + "epoch": 0.735844140225692, + "grad_norm": 1.646694540977478, + "learning_rate": 1.458525514105434e-05, + "loss": 3.173, + "step": 9178 + }, + { + "epoch": 0.7359243150067146, + "grad_norm": 1.9515299797058105, + "learning_rate": 1.458410119685949e-05, + "loss": 3.1612, + "step": 9179 + }, + { + "epoch": 0.7360044897877372, + "grad_norm": 1.9282076358795166, + "learning_rate": 1.458294717537831e-05, + "loss": 3.2374, + "step": 9180 + }, + { + "epoch": 0.7360846645687599, + "grad_norm": 1.854421615600586, + "learning_rate": 1.4581793076630255e-05, + "loss": 3.2497, + "step": 9181 + }, + { + "epoch": 0.7361648393497825, + "grad_norm": 1.723299503326416, + "learning_rate": 1.4580638900634788e-05, + "loss": 3.1863, + "step": 9182 + }, + { + "epoch": 0.7362450141308051, + "grad_norm": 1.8522225618362427, + "learning_rate": 1.4579484647411357e-05, + "loss": 3.2335, + "step": 9183 + }, + { + "epoch": 0.7363251889118277, + "grad_norm": 2.0279829502105713, + "learning_rate": 1.4578330316979434e-05, + "loss": 3.2922, + "step": 9184 + }, + { + "epoch": 0.7364053636928504, + "grad_norm": 1.7826929092407227, + "learning_rate": 1.4577175909358472e-05, + "loss": 3.1495, + "step": 9185 + }, + { + "epoch": 0.736485538473873, + "grad_norm": 1.797467589378357, + "learning_rate": 1.4576021424567936e-05, + "loss": 3.2711, + "step": 9186 + }, + { + "epoch": 0.7365657132548957, + "grad_norm": 1.8586432933807373, + "learning_rate": 1.4574866862627297e-05, + "loss": 3.2625, + "step": 9187 + }, + { + "epoch": 0.7366458880359184, + "grad_norm": 1.7316235303878784, + "learning_rate": 1.4573712223556008e-05, + "loss": 3.1318, + "step": 9188 + }, + { + "epoch": 0.736726062816941, + "grad_norm": 1.7589548826217651, + "learning_rate": 1.4572557507373551e-05, + "loss": 3.1642, + "step": 9189 + }, + { + "epoch": 0.7368062375979636, + "grad_norm": 1.827805757522583, + "learning_rate": 1.4571402714099381e-05, + "loss": 3.2103, + "step": 9190 + }, + { + "epoch": 0.7368864123789862, + "grad_norm": 0.8023554086685181, + "learning_rate": 1.4570247843752972e-05, + "loss": 2.4814, + "step": 9191 + }, + { + "epoch": 0.7369665871600088, + "grad_norm": 1.7537869215011597, + "learning_rate": 1.4569092896353798e-05, + "loss": 3.1648, + "step": 9192 + }, + { + "epoch": 0.7370467619410315, + "grad_norm": 1.7313936948776245, + "learning_rate": 1.4567937871921324e-05, + "loss": 3.1578, + "step": 9193 + }, + { + "epoch": 0.7371269367220541, + "grad_norm": 1.8768866062164307, + "learning_rate": 1.456678277047503e-05, + "loss": 3.3036, + "step": 9194 + }, + { + "epoch": 0.7372071115030767, + "grad_norm": 1.8684059381484985, + "learning_rate": 1.4565627592034389e-05, + "loss": 3.2375, + "step": 9195 + }, + { + "epoch": 0.7372872862840993, + "grad_norm": 1.7742910385131836, + "learning_rate": 1.4564472336618872e-05, + "loss": 3.1542, + "step": 9196 + }, + { + "epoch": 0.737367461065122, + "grad_norm": 2.065640926361084, + "learning_rate": 1.4563317004247965e-05, + "loss": 3.2581, + "step": 9197 + }, + { + "epoch": 0.7374476358461446, + "grad_norm": 1.830556869506836, + "learning_rate": 1.4562161594941137e-05, + "loss": 3.325, + "step": 9198 + }, + { + "epoch": 0.7375278106271672, + "grad_norm": 1.8575564622879028, + "learning_rate": 1.4561006108717876e-05, + "loss": 3.241, + "step": 9199 + }, + { + "epoch": 0.7376079854081898, + "grad_norm": 1.7530608177185059, + "learning_rate": 1.4559850545597658e-05, + "loss": 3.2465, + "step": 9200 + }, + { + "epoch": 0.7376881601892125, + "grad_norm": 1.8737683296203613, + "learning_rate": 1.4558694905599967e-05, + "loss": 3.1622, + "step": 9201 + }, + { + "epoch": 0.7377683349702351, + "grad_norm": 1.9555960893630981, + "learning_rate": 1.4557539188744288e-05, + "loss": 3.1888, + "step": 9202 + }, + { + "epoch": 0.7378485097512577, + "grad_norm": 2.025742292404175, + "learning_rate": 1.45563833950501e-05, + "loss": 3.1771, + "step": 9203 + }, + { + "epoch": 0.7379286845322803, + "grad_norm": 1.8171409368515015, + "learning_rate": 1.4555227524536899e-05, + "loss": 3.1893, + "step": 9204 + }, + { + "epoch": 0.738008859313303, + "grad_norm": 1.6936335563659668, + "learning_rate": 1.4554071577224162e-05, + "loss": 3.1766, + "step": 9205 + }, + { + "epoch": 0.7380890340943256, + "grad_norm": 1.7962357997894287, + "learning_rate": 1.4552915553131387e-05, + "loss": 3.2051, + "step": 9206 + }, + { + "epoch": 0.7381692088753482, + "grad_norm": 1.8073503971099854, + "learning_rate": 1.455175945227806e-05, + "loss": 3.2931, + "step": 9207 + }, + { + "epoch": 0.7382493836563709, + "grad_norm": 1.8262550830841064, + "learning_rate": 1.4550603274683672e-05, + "loss": 3.2042, + "step": 9208 + }, + { + "epoch": 0.7383295584373936, + "grad_norm": 1.8133095502853394, + "learning_rate": 1.4549447020367717e-05, + "loss": 3.1492, + "step": 9209 + }, + { + "epoch": 0.7384097332184162, + "grad_norm": 1.819575548171997, + "learning_rate": 1.4548290689349689e-05, + "loss": 3.2203, + "step": 9210 + }, + { + "epoch": 0.7384899079994388, + "grad_norm": 1.8214240074157715, + "learning_rate": 1.4547134281649082e-05, + "loss": 3.1993, + "step": 9211 + }, + { + "epoch": 0.7385700827804614, + "grad_norm": 1.8382847309112549, + "learning_rate": 1.4545977797285396e-05, + "loss": 3.2245, + "step": 9212 + }, + { + "epoch": 0.7386502575614841, + "grad_norm": 1.8862720727920532, + "learning_rate": 1.4544821236278122e-05, + "loss": 3.1703, + "step": 9213 + }, + { + "epoch": 0.7387304323425067, + "grad_norm": 1.9731101989746094, + "learning_rate": 1.4543664598646769e-05, + "loss": 3.225, + "step": 9214 + }, + { + "epoch": 0.7388106071235293, + "grad_norm": 1.7878830432891846, + "learning_rate": 1.4542507884410828e-05, + "loss": 3.2935, + "step": 9215 + }, + { + "epoch": 0.7388907819045519, + "grad_norm": 1.8879368305206299, + "learning_rate": 1.4541351093589806e-05, + "loss": 3.1985, + "step": 9216 + }, + { + "epoch": 0.7389709566855746, + "grad_norm": 1.7950555086135864, + "learning_rate": 1.4540194226203208e-05, + "loss": 3.1624, + "step": 9217 + }, + { + "epoch": 0.7390511314665972, + "grad_norm": 2.069633722305298, + "learning_rate": 1.4539037282270532e-05, + "loss": 3.2388, + "step": 9218 + }, + { + "epoch": 0.7391313062476198, + "grad_norm": 1.856677532196045, + "learning_rate": 1.4537880261811288e-05, + "loss": 3.3541, + "step": 9219 + }, + { + "epoch": 0.7392114810286424, + "grad_norm": 1.713296890258789, + "learning_rate": 1.4536723164844982e-05, + "loss": 3.088, + "step": 9220 + }, + { + "epoch": 0.739291655809665, + "grad_norm": 1.8050227165222168, + "learning_rate": 1.453556599139112e-05, + "loss": 3.2146, + "step": 9221 + }, + { + "epoch": 0.7393718305906877, + "grad_norm": 1.7648677825927734, + "learning_rate": 1.453440874146922e-05, + "loss": 3.1568, + "step": 9222 + }, + { + "epoch": 0.7394520053717103, + "grad_norm": 1.6770141124725342, + "learning_rate": 1.4533251415098782e-05, + "loss": 3.1346, + "step": 9223 + }, + { + "epoch": 0.7395321801527329, + "grad_norm": 1.9090429544448853, + "learning_rate": 1.4532094012299325e-05, + "loss": 3.183, + "step": 9224 + }, + { + "epoch": 0.7396123549337555, + "grad_norm": 1.656292200088501, + "learning_rate": 1.4530936533090358e-05, + "loss": 3.1635, + "step": 9225 + }, + { + "epoch": 0.7396925297147782, + "grad_norm": 1.8201676607131958, + "learning_rate": 1.45297789774914e-05, + "loss": 3.1984, + "step": 9226 + }, + { + "epoch": 0.7397727044958008, + "grad_norm": 1.790810465812683, + "learning_rate": 1.4528621345521965e-05, + "loss": 3.1711, + "step": 9227 + }, + { + "epoch": 0.7398528792768235, + "grad_norm": 1.8556807041168213, + "learning_rate": 1.4527463637201568e-05, + "loss": 3.0944, + "step": 9228 + }, + { + "epoch": 0.7399330540578462, + "grad_norm": 1.7921007871627808, + "learning_rate": 1.4526305852549731e-05, + "loss": 3.2478, + "step": 9229 + }, + { + "epoch": 0.7400132288388688, + "grad_norm": 0.7559865117073059, + "learning_rate": 1.4525147991585973e-05, + "loss": 2.4872, + "step": 9230 + }, + { + "epoch": 0.7400934036198914, + "grad_norm": 1.893297553062439, + "learning_rate": 1.4523990054329813e-05, + "loss": 3.1373, + "step": 9231 + }, + { + "epoch": 0.740173578400914, + "grad_norm": 1.9157458543777466, + "learning_rate": 1.4522832040800778e-05, + "loss": 3.22, + "step": 9232 + }, + { + "epoch": 0.7402537531819366, + "grad_norm": 1.7686234712600708, + "learning_rate": 1.4521673951018388e-05, + "loss": 3.1545, + "step": 9233 + }, + { + "epoch": 0.7403339279629593, + "grad_norm": 1.7195990085601807, + "learning_rate": 1.4520515785002165e-05, + "loss": 3.195, + "step": 9234 + }, + { + "epoch": 0.7404141027439819, + "grad_norm": 1.7600574493408203, + "learning_rate": 1.4519357542771644e-05, + "loss": 3.2124, + "step": 9235 + }, + { + "epoch": 0.7404942775250045, + "grad_norm": 1.8720874786376953, + "learning_rate": 1.4518199224346344e-05, + "loss": 3.2469, + "step": 9236 + }, + { + "epoch": 0.7405744523060271, + "grad_norm": 1.8394687175750732, + "learning_rate": 1.45170408297458e-05, + "loss": 3.1683, + "step": 9237 + }, + { + "epoch": 0.7406546270870498, + "grad_norm": 1.8216854333877563, + "learning_rate": 1.4515882358989538e-05, + "loss": 3.2579, + "step": 9238 + }, + { + "epoch": 0.7407348018680724, + "grad_norm": 1.8963992595672607, + "learning_rate": 1.4514723812097092e-05, + "loss": 3.3153, + "step": 9239 + }, + { + "epoch": 0.740814976649095, + "grad_norm": 1.7821927070617676, + "learning_rate": 1.4513565189087993e-05, + "loss": 3.2399, + "step": 9240 + }, + { + "epoch": 0.7408951514301176, + "grad_norm": 2.0910229682922363, + "learning_rate": 1.4512406489981773e-05, + "loss": 3.2831, + "step": 9241 + }, + { + "epoch": 0.7409753262111403, + "grad_norm": 1.8556302785873413, + "learning_rate": 1.4511247714797976e-05, + "loss": 3.3529, + "step": 9242 + }, + { + "epoch": 0.7410555009921629, + "grad_norm": 1.7705072164535522, + "learning_rate": 1.4510088863556128e-05, + "loss": 3.1295, + "step": 9243 + }, + { + "epoch": 0.7411356757731855, + "grad_norm": 1.7545064687728882, + "learning_rate": 1.4508929936275771e-05, + "loss": 3.209, + "step": 9244 + }, + { + "epoch": 0.7412158505542081, + "grad_norm": 1.738964319229126, + "learning_rate": 1.4507770932976449e-05, + "loss": 3.0817, + "step": 9245 + }, + { + "epoch": 0.7412960253352308, + "grad_norm": 1.6750737428665161, + "learning_rate": 1.4506611853677693e-05, + "loss": 3.1576, + "step": 9246 + }, + { + "epoch": 0.7413762001162534, + "grad_norm": 1.8033725023269653, + "learning_rate": 1.4505452698399052e-05, + "loss": 3.262, + "step": 9247 + }, + { + "epoch": 0.7414563748972761, + "grad_norm": 1.7761980295181274, + "learning_rate": 1.4504293467160065e-05, + "loss": 3.2069, + "step": 9248 + }, + { + "epoch": 0.7415365496782987, + "grad_norm": 1.8649803400039673, + "learning_rate": 1.4503134159980276e-05, + "loss": 3.2582, + "step": 9249 + }, + { + "epoch": 0.7416167244593214, + "grad_norm": 1.7667897939682007, + "learning_rate": 1.4501974776879236e-05, + "loss": 3.0837, + "step": 9250 + }, + { + "epoch": 0.741696899240344, + "grad_norm": 1.7639540433883667, + "learning_rate": 1.4500815317876484e-05, + "loss": 3.1561, + "step": 9251 + }, + { + "epoch": 0.7417770740213666, + "grad_norm": 1.8415327072143555, + "learning_rate": 1.4499655782991576e-05, + "loss": 3.2224, + "step": 9252 + }, + { + "epoch": 0.7418572488023892, + "grad_norm": 1.7546409368515015, + "learning_rate": 1.4498496172244054e-05, + "loss": 3.1625, + "step": 9253 + }, + { + "epoch": 0.7419374235834119, + "grad_norm": 1.6799144744873047, + "learning_rate": 1.4497336485653473e-05, + "loss": 3.11, + "step": 9254 + }, + { + "epoch": 0.7420175983644345, + "grad_norm": 2.522529125213623, + "learning_rate": 1.4496176723239387e-05, + "loss": 3.243, + "step": 9255 + }, + { + "epoch": 0.7420977731454571, + "grad_norm": 1.71079683303833, + "learning_rate": 1.449501688502134e-05, + "loss": 3.1434, + "step": 9256 + }, + { + "epoch": 0.7421779479264797, + "grad_norm": 1.6980392932891846, + "learning_rate": 1.4493856971018897e-05, + "loss": 3.1636, + "step": 9257 + }, + { + "epoch": 0.7422581227075024, + "grad_norm": 1.7142919301986694, + "learning_rate": 1.4492696981251608e-05, + "loss": 3.257, + "step": 9258 + }, + { + "epoch": 0.742338297488525, + "grad_norm": 1.7828155755996704, + "learning_rate": 1.449153691573903e-05, + "loss": 3.2535, + "step": 9259 + }, + { + "epoch": 0.7424184722695476, + "grad_norm": 1.9482475519180298, + "learning_rate": 1.4490376774500725e-05, + "loss": 3.2318, + "step": 9260 + }, + { + "epoch": 0.7424986470505702, + "grad_norm": 1.7554059028625488, + "learning_rate": 1.448921655755625e-05, + "loss": 3.1965, + "step": 9261 + }, + { + "epoch": 0.7425788218315929, + "grad_norm": 1.7870184183120728, + "learning_rate": 1.4488056264925166e-05, + "loss": 3.238, + "step": 9262 + }, + { + "epoch": 0.7426589966126155, + "grad_norm": 1.7559452056884766, + "learning_rate": 1.4486895896627031e-05, + "loss": 3.1635, + "step": 9263 + }, + { + "epoch": 0.7427391713936381, + "grad_norm": 1.897892951965332, + "learning_rate": 1.4485735452681415e-05, + "loss": 3.1483, + "step": 9264 + }, + { + "epoch": 0.7428193461746607, + "grad_norm": 1.7513922452926636, + "learning_rate": 1.4484574933107882e-05, + "loss": 3.2736, + "step": 9265 + }, + { + "epoch": 0.7428995209556833, + "grad_norm": 1.8494739532470703, + "learning_rate": 1.4483414337925994e-05, + "loss": 3.2365, + "step": 9266 + }, + { + "epoch": 0.742979695736706, + "grad_norm": 1.8972214460372925, + "learning_rate": 1.4482253667155323e-05, + "loss": 3.1346, + "step": 9267 + }, + { + "epoch": 0.7430598705177286, + "grad_norm": 1.81423020362854, + "learning_rate": 1.4481092920815432e-05, + "loss": 3.2886, + "step": 9268 + }, + { + "epoch": 0.7431400452987513, + "grad_norm": 1.8213980197906494, + "learning_rate": 1.4479932098925895e-05, + "loss": 3.291, + "step": 9269 + }, + { + "epoch": 0.743220220079774, + "grad_norm": 1.757699966430664, + "learning_rate": 1.4478771201506284e-05, + "loss": 3.1255, + "step": 9270 + }, + { + "epoch": 0.7433003948607966, + "grad_norm": 1.8036733865737915, + "learning_rate": 1.4477610228576165e-05, + "loss": 3.2389, + "step": 9271 + }, + { + "epoch": 0.7433805696418192, + "grad_norm": 1.8072682619094849, + "learning_rate": 1.4476449180155117e-05, + "loss": 3.2682, + "step": 9272 + }, + { + "epoch": 0.7434607444228418, + "grad_norm": 1.696129322052002, + "learning_rate": 1.4475288056262712e-05, + "loss": 3.1764, + "step": 9273 + }, + { + "epoch": 0.7435409192038644, + "grad_norm": 0.7440284490585327, + "learning_rate": 1.4474126856918532e-05, + "loss": 2.4733, + "step": 9274 + }, + { + "epoch": 0.7436210939848871, + "grad_norm": 1.7547516822814941, + "learning_rate": 1.4472965582142147e-05, + "loss": 3.1428, + "step": 9275 + }, + { + "epoch": 0.7437012687659097, + "grad_norm": 1.8197021484375, + "learning_rate": 1.4471804231953137e-05, + "loss": 3.2319, + "step": 9276 + }, + { + "epoch": 0.7437814435469323, + "grad_norm": 1.6932203769683838, + "learning_rate": 1.4470642806371086e-05, + "loss": 3.1801, + "step": 9277 + }, + { + "epoch": 0.7438616183279549, + "grad_norm": 1.8318222761154175, + "learning_rate": 1.4469481305415573e-05, + "loss": 3.1747, + "step": 9278 + }, + { + "epoch": 0.7439417931089776, + "grad_norm": 1.8374429941177368, + "learning_rate": 1.446831972910618e-05, + "loss": 3.1984, + "step": 9279 + }, + { + "epoch": 0.7440219678900002, + "grad_norm": 1.712438941001892, + "learning_rate": 1.446715807746249e-05, + "loss": 3.2574, + "step": 9280 + }, + { + "epoch": 0.7441021426710228, + "grad_norm": 1.6526418924331665, + "learning_rate": 1.4465996350504093e-05, + "loss": 3.1607, + "step": 9281 + }, + { + "epoch": 0.7441823174520454, + "grad_norm": 1.8040857315063477, + "learning_rate": 1.4464834548250568e-05, + "loss": 3.2869, + "step": 9282 + }, + { + "epoch": 0.7442624922330681, + "grad_norm": 1.7453322410583496, + "learning_rate": 1.4463672670721506e-05, + "loss": 3.236, + "step": 9283 + }, + { + "epoch": 0.7443426670140907, + "grad_norm": 1.8705261945724487, + "learning_rate": 1.4462510717936495e-05, + "loss": 3.1953, + "step": 9284 + }, + { + "epoch": 0.7444228417951133, + "grad_norm": 2.2379634380340576, + "learning_rate": 1.4461348689915128e-05, + "loss": 3.1505, + "step": 9285 + }, + { + "epoch": 0.7445030165761359, + "grad_norm": 1.886315107345581, + "learning_rate": 1.4460186586676996e-05, + "loss": 3.3294, + "step": 9286 + }, + { + "epoch": 0.7445831913571586, + "grad_norm": 1.7698631286621094, + "learning_rate": 1.4459024408241689e-05, + "loss": 3.386, + "step": 9287 + }, + { + "epoch": 0.7446633661381812, + "grad_norm": 1.8275922536849976, + "learning_rate": 1.44578621546288e-05, + "loss": 3.1911, + "step": 9288 + }, + { + "epoch": 0.7447435409192039, + "grad_norm": 1.9086511135101318, + "learning_rate": 1.4456699825857927e-05, + "loss": 3.1392, + "step": 9289 + }, + { + "epoch": 0.7448237157002265, + "grad_norm": 1.96347177028656, + "learning_rate": 1.4455537421948667e-05, + "loss": 3.2071, + "step": 9290 + }, + { + "epoch": 0.7449038904812492, + "grad_norm": 1.7818949222564697, + "learning_rate": 1.4454374942920615e-05, + "loss": 3.2057, + "step": 9291 + }, + { + "epoch": 0.7449840652622718, + "grad_norm": 1.7592544555664062, + "learning_rate": 1.4453212388793371e-05, + "loss": 3.1948, + "step": 9292 + }, + { + "epoch": 0.7450642400432944, + "grad_norm": 1.7445857524871826, + "learning_rate": 1.4452049759586537e-05, + "loss": 3.2158, + "step": 9293 + }, + { + "epoch": 0.745144414824317, + "grad_norm": 1.7980475425720215, + "learning_rate": 1.4450887055319714e-05, + "loss": 3.2122, + "step": 9294 + }, + { + "epoch": 0.7452245896053397, + "grad_norm": 1.7956271171569824, + "learning_rate": 1.4449724276012502e-05, + "loss": 3.3163, + "step": 9295 + }, + { + "epoch": 0.7453047643863623, + "grad_norm": 0.7536507248878479, + "learning_rate": 1.4448561421684507e-05, + "loss": 2.5129, + "step": 9296 + }, + { + "epoch": 0.7453849391673849, + "grad_norm": 1.8060628175735474, + "learning_rate": 1.4447398492355335e-05, + "loss": 3.1801, + "step": 9297 + }, + { + "epoch": 0.7454651139484075, + "grad_norm": 1.8344320058822632, + "learning_rate": 1.444623548804459e-05, + "loss": 3.2517, + "step": 9298 + }, + { + "epoch": 0.7455452887294302, + "grad_norm": 1.8154196739196777, + "learning_rate": 1.4445072408771883e-05, + "loss": 3.2136, + "step": 9299 + }, + { + "epoch": 0.7456254635104528, + "grad_norm": 1.958062767982483, + "learning_rate": 1.4443909254556823e-05, + "loss": 3.3066, + "step": 9300 + }, + { + "epoch": 0.7457056382914754, + "grad_norm": 1.7357168197631836, + "learning_rate": 1.4442746025419017e-05, + "loss": 3.0734, + "step": 9301 + }, + { + "epoch": 0.745785813072498, + "grad_norm": 1.820576548576355, + "learning_rate": 1.4441582721378083e-05, + "loss": 3.3259, + "step": 9302 + }, + { + "epoch": 0.7458659878535207, + "grad_norm": 1.9073905944824219, + "learning_rate": 1.4440419342453625e-05, + "loss": 3.261, + "step": 9303 + }, + { + "epoch": 0.7459461626345433, + "grad_norm": 1.832346796989441, + "learning_rate": 1.4439255888665263e-05, + "loss": 3.3069, + "step": 9304 + }, + { + "epoch": 0.7460263374155659, + "grad_norm": 1.7438442707061768, + "learning_rate": 1.4438092360032611e-05, + "loss": 3.2211, + "step": 9305 + }, + { + "epoch": 0.7461065121965885, + "grad_norm": 1.7617748975753784, + "learning_rate": 1.4436928756575286e-05, + "loss": 3.2507, + "step": 9306 + }, + { + "epoch": 0.7461866869776111, + "grad_norm": 1.885648488998413, + "learning_rate": 1.4435765078312906e-05, + "loss": 3.2544, + "step": 9307 + }, + { + "epoch": 0.7462668617586338, + "grad_norm": 1.774315357208252, + "learning_rate": 1.4434601325265092e-05, + "loss": 3.1432, + "step": 9308 + }, + { + "epoch": 0.7463470365396565, + "grad_norm": 1.7125779390335083, + "learning_rate": 1.443343749745146e-05, + "loss": 3.121, + "step": 9309 + }, + { + "epoch": 0.7464272113206791, + "grad_norm": 1.8937698602676392, + "learning_rate": 1.4432273594891637e-05, + "loss": 3.2587, + "step": 9310 + }, + { + "epoch": 0.7465073861017018, + "grad_norm": 1.7239699363708496, + "learning_rate": 1.4431109617605242e-05, + "loss": 3.2464, + "step": 9311 + }, + { + "epoch": 0.7465875608827244, + "grad_norm": 1.7119531631469727, + "learning_rate": 1.44299455656119e-05, + "loss": 3.2371, + "step": 9312 + }, + { + "epoch": 0.746667735663747, + "grad_norm": 1.857801914215088, + "learning_rate": 1.4428781438931236e-05, + "loss": 3.1507, + "step": 9313 + }, + { + "epoch": 0.7467479104447696, + "grad_norm": 1.7740118503570557, + "learning_rate": 1.4427617237582878e-05, + "loss": 3.185, + "step": 9314 + }, + { + "epoch": 0.7468280852257922, + "grad_norm": 1.830542802810669, + "learning_rate": 1.4426452961586457e-05, + "loss": 3.2665, + "step": 9315 + }, + { + "epoch": 0.7469082600068149, + "grad_norm": 1.8541922569274902, + "learning_rate": 1.4425288610961596e-05, + "loss": 3.1959, + "step": 9316 + }, + { + "epoch": 0.7469884347878375, + "grad_norm": 1.8449618816375732, + "learning_rate": 1.442412418572793e-05, + "loss": 3.1848, + "step": 9317 + }, + { + "epoch": 0.7470686095688601, + "grad_norm": 1.7939034700393677, + "learning_rate": 1.4422959685905089e-05, + "loss": 3.2028, + "step": 9318 + }, + { + "epoch": 0.7471487843498827, + "grad_norm": 1.8289682865142822, + "learning_rate": 1.4421795111512703e-05, + "loss": 3.348, + "step": 9319 + }, + { + "epoch": 0.7472289591309054, + "grad_norm": 1.711434006690979, + "learning_rate": 1.4420630462570416e-05, + "loss": 3.229, + "step": 9320 + }, + { + "epoch": 0.747309133911928, + "grad_norm": 1.76106858253479, + "learning_rate": 1.4419465739097854e-05, + "loss": 3.1705, + "step": 9321 + }, + { + "epoch": 0.7473893086929506, + "grad_norm": 1.737480640411377, + "learning_rate": 1.4418300941114657e-05, + "loss": 3.1427, + "step": 9322 + }, + { + "epoch": 0.7474694834739732, + "grad_norm": 1.8796710968017578, + "learning_rate": 1.4417136068640465e-05, + "loss": 3.1896, + "step": 9323 + }, + { + "epoch": 0.7475496582549959, + "grad_norm": 1.8579816818237305, + "learning_rate": 1.4415971121694914e-05, + "loss": 3.3691, + "step": 9324 + }, + { + "epoch": 0.7476298330360185, + "grad_norm": 1.7728573083877563, + "learning_rate": 1.4414806100297648e-05, + "loss": 3.2902, + "step": 9325 + }, + { + "epoch": 0.7477100078170411, + "grad_norm": 1.7986047267913818, + "learning_rate": 1.4413641004468306e-05, + "loss": 3.1591, + "step": 9326 + }, + { + "epoch": 0.7477901825980637, + "grad_norm": 1.711042046546936, + "learning_rate": 1.4412475834226534e-05, + "loss": 3.1587, + "step": 9327 + }, + { + "epoch": 0.7478703573790864, + "grad_norm": 1.8689830303192139, + "learning_rate": 1.4411310589591972e-05, + "loss": 3.2454, + "step": 9328 + }, + { + "epoch": 0.747950532160109, + "grad_norm": 1.8129558563232422, + "learning_rate": 1.4410145270584268e-05, + "loss": 3.2443, + "step": 9329 + }, + { + "epoch": 0.7480307069411317, + "grad_norm": 1.7289084196090698, + "learning_rate": 1.440897987722307e-05, + "loss": 3.1801, + "step": 9330 + }, + { + "epoch": 0.7481108817221543, + "grad_norm": 0.8370267748832703, + "learning_rate": 1.4407814409528026e-05, + "loss": 2.4847, + "step": 9331 + }, + { + "epoch": 0.748191056503177, + "grad_norm": 1.7360607385635376, + "learning_rate": 1.4406648867518785e-05, + "loss": 3.0933, + "step": 9332 + }, + { + "epoch": 0.7482712312841996, + "grad_norm": 1.8870999813079834, + "learning_rate": 1.4405483251214993e-05, + "loss": 3.2105, + "step": 9333 + }, + { + "epoch": 0.7483514060652222, + "grad_norm": 1.7999933958053589, + "learning_rate": 1.440431756063631e-05, + "loss": 3.296, + "step": 9334 + }, + { + "epoch": 0.7484315808462448, + "grad_norm": 1.723710060119629, + "learning_rate": 1.4403151795802384e-05, + "loss": 3.2213, + "step": 9335 + }, + { + "epoch": 0.7485117556272675, + "grad_norm": 1.8279091119766235, + "learning_rate": 1.4401985956732871e-05, + "loss": 3.1931, + "step": 9336 + }, + { + "epoch": 0.7485919304082901, + "grad_norm": 1.9473727941513062, + "learning_rate": 1.4400820043447428e-05, + "loss": 3.2664, + "step": 9337 + }, + { + "epoch": 0.7486721051893127, + "grad_norm": 1.9340211153030396, + "learning_rate": 1.4399654055965706e-05, + "loss": 3.2551, + "step": 9338 + }, + { + "epoch": 0.7487522799703353, + "grad_norm": 1.788063645362854, + "learning_rate": 1.439848799430737e-05, + "loss": 3.3124, + "step": 9339 + }, + { + "epoch": 0.748832454751358, + "grad_norm": 1.658360481262207, + "learning_rate": 1.4397321858492074e-05, + "loss": 3.1946, + "step": 9340 + }, + { + "epoch": 0.7489126295323806, + "grad_norm": 1.777328372001648, + "learning_rate": 1.4396155648539483e-05, + "loss": 3.329, + "step": 9341 + }, + { + "epoch": 0.7489928043134032, + "grad_norm": 1.8417381048202515, + "learning_rate": 1.4394989364469258e-05, + "loss": 3.2632, + "step": 9342 + }, + { + "epoch": 0.7490729790944258, + "grad_norm": 1.794407844543457, + "learning_rate": 1.439382300630106e-05, + "loss": 3.2412, + "step": 9343 + }, + { + "epoch": 0.7491531538754485, + "grad_norm": 1.8288424015045166, + "learning_rate": 1.4392656574054554e-05, + "loss": 3.1594, + "step": 9344 + }, + { + "epoch": 0.7492333286564711, + "grad_norm": 1.767754316329956, + "learning_rate": 1.4391490067749407e-05, + "loss": 3.2567, + "step": 9345 + }, + { + "epoch": 0.7493135034374937, + "grad_norm": 2.086735248565674, + "learning_rate": 1.4390323487405287e-05, + "loss": 3.2655, + "step": 9346 + }, + { + "epoch": 0.7493936782185163, + "grad_norm": 1.7348648309707642, + "learning_rate": 1.4389156833041858e-05, + "loss": 3.0581, + "step": 9347 + }, + { + "epoch": 0.749473852999539, + "grad_norm": 0.7394681572914124, + "learning_rate": 1.438799010467879e-05, + "loss": 2.4834, + "step": 9348 + }, + { + "epoch": 0.7495540277805616, + "grad_norm": 1.678036093711853, + "learning_rate": 1.4386823302335758e-05, + "loss": 3.1445, + "step": 9349 + }, + { + "epoch": 0.7496342025615843, + "grad_norm": 1.765848994255066, + "learning_rate": 1.4385656426032432e-05, + "loss": 3.1312, + "step": 9350 + }, + { + "epoch": 0.7497143773426069, + "grad_norm": 0.7132164835929871, + "learning_rate": 1.4384489475788484e-05, + "loss": 2.5089, + "step": 9351 + }, + { + "epoch": 0.7497945521236296, + "grad_norm": 1.860290288925171, + "learning_rate": 1.4383322451623587e-05, + "loss": 3.2604, + "step": 9352 + }, + { + "epoch": 0.7498747269046522, + "grad_norm": 1.754860520362854, + "learning_rate": 1.4382155353557418e-05, + "loss": 3.1414, + "step": 9353 + }, + { + "epoch": 0.7499549016856748, + "grad_norm": 1.7211476564407349, + "learning_rate": 1.4380988181609656e-05, + "loss": 3.116, + "step": 9354 + }, + { + "epoch": 0.7500350764666974, + "grad_norm": 1.882403016090393, + "learning_rate": 1.4379820935799978e-05, + "loss": 3.1562, + "step": 9355 + }, + { + "epoch": 0.75011525124772, + "grad_norm": 1.808089017868042, + "learning_rate": 1.437865361614806e-05, + "loss": 3.2965, + "step": 9356 + }, + { + "epoch": 0.7501954260287427, + "grad_norm": 1.7728943824768066, + "learning_rate": 1.4377486222673587e-05, + "loss": 3.2289, + "step": 9357 + }, + { + "epoch": 0.7502756008097653, + "grad_norm": 1.7567366361618042, + "learning_rate": 1.4376318755396236e-05, + "loss": 3.1521, + "step": 9358 + }, + { + "epoch": 0.7503557755907879, + "grad_norm": 1.8006477355957031, + "learning_rate": 1.4375151214335696e-05, + "loss": 3.1245, + "step": 9359 + }, + { + "epoch": 0.7504359503718105, + "grad_norm": 2.2946536540985107, + "learning_rate": 1.437398359951165e-05, + "loss": 3.2715, + "step": 9360 + }, + { + "epoch": 0.7505161251528332, + "grad_norm": 1.7839305400848389, + "learning_rate": 1.437281591094378e-05, + "loss": 3.214, + "step": 9361 + }, + { + "epoch": 0.7505962999338558, + "grad_norm": 0.7956647872924805, + "learning_rate": 1.4371648148651776e-05, + "loss": 2.4982, + "step": 9362 + }, + { + "epoch": 0.7506764747148784, + "grad_norm": 1.7021031379699707, + "learning_rate": 1.4370480312655324e-05, + "loss": 3.2078, + "step": 9363 + }, + { + "epoch": 0.750756649495901, + "grad_norm": 1.887162446975708, + "learning_rate": 1.4369312402974111e-05, + "loss": 3.228, + "step": 9364 + }, + { + "epoch": 0.7508368242769237, + "grad_norm": 1.7843095064163208, + "learning_rate": 1.4368144419627836e-05, + "loss": 3.181, + "step": 9365 + }, + { + "epoch": 0.7509169990579463, + "grad_norm": 1.8293931484222412, + "learning_rate": 1.4366976362636183e-05, + "loss": 3.1796, + "step": 9366 + }, + { + "epoch": 0.7509971738389689, + "grad_norm": 1.6623594760894775, + "learning_rate": 1.4365808232018849e-05, + "loss": 3.2048, + "step": 9367 + }, + { + "epoch": 0.7510773486199915, + "grad_norm": 1.9454411268234253, + "learning_rate": 1.4364640027795524e-05, + "loss": 3.3274, + "step": 9368 + }, + { + "epoch": 0.7511575234010142, + "grad_norm": 1.7964893579483032, + "learning_rate": 1.4363471749985909e-05, + "loss": 3.1461, + "step": 9369 + }, + { + "epoch": 0.7512376981820369, + "grad_norm": 1.8323023319244385, + "learning_rate": 1.4362303398609698e-05, + "loss": 3.1805, + "step": 9370 + }, + { + "epoch": 0.7513178729630595, + "grad_norm": 1.8217427730560303, + "learning_rate": 1.4361134973686587e-05, + "loss": 3.2444, + "step": 9371 + }, + { + "epoch": 0.7513980477440821, + "grad_norm": 0.7913638949394226, + "learning_rate": 1.4359966475236282e-05, + "loss": 2.5061, + "step": 9372 + }, + { + "epoch": 0.7514782225251048, + "grad_norm": 0.789340078830719, + "learning_rate": 1.4358797903278473e-05, + "loss": 2.4846, + "step": 9373 + }, + { + "epoch": 0.7515583973061274, + "grad_norm": 1.8023518323898315, + "learning_rate": 1.435762925783287e-05, + "loss": 3.1766, + "step": 9374 + }, + { + "epoch": 0.75163857208715, + "grad_norm": 1.7238801717758179, + "learning_rate": 1.4356460538919178e-05, + "loss": 3.0637, + "step": 9375 + }, + { + "epoch": 0.7517187468681726, + "grad_norm": 1.7670623064041138, + "learning_rate": 1.435529174655709e-05, + "loss": 3.2439, + "step": 9376 + }, + { + "epoch": 0.7517989216491953, + "grad_norm": 1.8818097114562988, + "learning_rate": 1.4354122880766321e-05, + "loss": 3.1802, + "step": 9377 + }, + { + "epoch": 0.7518790964302179, + "grad_norm": 1.8596913814544678, + "learning_rate": 1.4352953941566574e-05, + "loss": 3.1724, + "step": 9378 + }, + { + "epoch": 0.7519592712112405, + "grad_norm": 1.7771052122116089, + "learning_rate": 1.4351784928977557e-05, + "loss": 3.2967, + "step": 9379 + }, + { + "epoch": 0.7520394459922631, + "grad_norm": 1.746806025505066, + "learning_rate": 1.4350615843018982e-05, + "loss": 3.1296, + "step": 9380 + }, + { + "epoch": 0.7521196207732858, + "grad_norm": 1.808809518814087, + "learning_rate": 1.4349446683710556e-05, + "loss": 3.1547, + "step": 9381 + }, + { + "epoch": 0.7521997955543084, + "grad_norm": 1.8704239130020142, + "learning_rate": 1.4348277451071992e-05, + "loss": 3.1307, + "step": 9382 + }, + { + "epoch": 0.752279970335331, + "grad_norm": 1.8146413564682007, + "learning_rate": 1.4347108145123004e-05, + "loss": 3.2229, + "step": 9383 + }, + { + "epoch": 0.7523601451163536, + "grad_norm": 0.7213684320449829, + "learning_rate": 1.4345938765883301e-05, + "loss": 2.4874, + "step": 9384 + }, + { + "epoch": 0.7524403198973763, + "grad_norm": 1.806256651878357, + "learning_rate": 1.4344769313372606e-05, + "loss": 3.1816, + "step": 9385 + }, + { + "epoch": 0.7525204946783989, + "grad_norm": 0.7360286116600037, + "learning_rate": 1.434359978761063e-05, + "loss": 2.4948, + "step": 9386 + }, + { + "epoch": 0.7526006694594215, + "grad_norm": 1.8673723936080933, + "learning_rate": 1.434243018861709e-05, + "loss": 3.1636, + "step": 9387 + }, + { + "epoch": 0.7526808442404441, + "grad_norm": 1.7020001411437988, + "learning_rate": 1.434126051641171e-05, + "loss": 3.2585, + "step": 9388 + }, + { + "epoch": 0.7527610190214667, + "grad_norm": 1.8063207864761353, + "learning_rate": 1.4340090771014208e-05, + "loss": 3.2858, + "step": 9389 + }, + { + "epoch": 0.7528411938024894, + "grad_norm": 1.732170820236206, + "learning_rate": 1.4338920952444306e-05, + "loss": 3.1788, + "step": 9390 + }, + { + "epoch": 0.7529213685835121, + "grad_norm": 1.752240538597107, + "learning_rate": 1.4337751060721723e-05, + "loss": 3.1608, + "step": 9391 + }, + { + "epoch": 0.7530015433645347, + "grad_norm": 1.8209788799285889, + "learning_rate": 1.4336581095866188e-05, + "loss": 3.2405, + "step": 9392 + }, + { + "epoch": 0.7530817181455574, + "grad_norm": 1.8327829837799072, + "learning_rate": 1.4335411057897424e-05, + "loss": 3.1539, + "step": 9393 + }, + { + "epoch": 0.75316189292658, + "grad_norm": 1.7954508066177368, + "learning_rate": 1.4334240946835157e-05, + "loss": 3.2042, + "step": 9394 + }, + { + "epoch": 0.7532420677076026, + "grad_norm": 1.7975741624832153, + "learning_rate": 1.4333070762699114e-05, + "loss": 3.2109, + "step": 9395 + }, + { + "epoch": 0.7533222424886252, + "grad_norm": 1.7115609645843506, + "learning_rate": 1.4331900505509028e-05, + "loss": 3.1888, + "step": 9396 + }, + { + "epoch": 0.7534024172696479, + "grad_norm": 1.7238131761550903, + "learning_rate": 1.4330730175284626e-05, + "loss": 3.2715, + "step": 9397 + }, + { + "epoch": 0.7534825920506705, + "grad_norm": 1.8313677310943604, + "learning_rate": 1.4329559772045637e-05, + "loss": 3.1519, + "step": 9398 + }, + { + "epoch": 0.7535627668316931, + "grad_norm": 1.8960144519805908, + "learning_rate": 1.4328389295811796e-05, + "loss": 3.2256, + "step": 9399 + }, + { + "epoch": 0.7536429416127157, + "grad_norm": 0.7140706181526184, + "learning_rate": 1.432721874660284e-05, + "loss": 2.5005, + "step": 9400 + }, + { + "epoch": 0.7537231163937383, + "grad_norm": 1.8825714588165283, + "learning_rate": 1.4326048124438499e-05, + "loss": 3.2012, + "step": 9401 + }, + { + "epoch": 0.753803291174761, + "grad_norm": 1.8005987405776978, + "learning_rate": 1.432487742933851e-05, + "loss": 3.2311, + "step": 9402 + }, + { + "epoch": 0.7538834659557836, + "grad_norm": 1.7105251550674438, + "learning_rate": 1.4323706661322615e-05, + "loss": 3.1044, + "step": 9403 + }, + { + "epoch": 0.7539636407368062, + "grad_norm": 1.7001384496688843, + "learning_rate": 1.4322535820410545e-05, + "loss": 3.1438, + "step": 9404 + }, + { + "epoch": 0.7540438155178288, + "grad_norm": 1.7959721088409424, + "learning_rate": 1.432136490662205e-05, + "loss": 3.1809, + "step": 9405 + }, + { + "epoch": 0.7541239902988515, + "grad_norm": 1.8097202777862549, + "learning_rate": 1.4320193919976861e-05, + "loss": 3.2283, + "step": 9406 + }, + { + "epoch": 0.7542041650798741, + "grad_norm": 1.7692878246307373, + "learning_rate": 1.4319022860494728e-05, + "loss": 3.2523, + "step": 9407 + }, + { + "epoch": 0.7542843398608967, + "grad_norm": 1.9403865337371826, + "learning_rate": 1.4317851728195392e-05, + "loss": 3.2635, + "step": 9408 + }, + { + "epoch": 0.7543645146419193, + "grad_norm": 1.8204296827316284, + "learning_rate": 1.4316680523098593e-05, + "loss": 3.2895, + "step": 9409 + }, + { + "epoch": 0.754444689422942, + "grad_norm": 1.871273159980774, + "learning_rate": 1.4315509245224087e-05, + "loss": 3.224, + "step": 9410 + }, + { + "epoch": 0.7545248642039647, + "grad_norm": 1.9239041805267334, + "learning_rate": 1.4314337894591615e-05, + "loss": 3.1558, + "step": 9411 + }, + { + "epoch": 0.7546050389849873, + "grad_norm": 1.8549673557281494, + "learning_rate": 1.4313166471220924e-05, + "loss": 3.1537, + "step": 9412 + }, + { + "epoch": 0.7546852137660099, + "grad_norm": 1.7118840217590332, + "learning_rate": 1.4311994975131769e-05, + "loss": 3.213, + "step": 9413 + }, + { + "epoch": 0.7547653885470326, + "grad_norm": 1.8029282093048096, + "learning_rate": 1.4310823406343897e-05, + "loss": 3.2536, + "step": 9414 + }, + { + "epoch": 0.7548455633280552, + "grad_norm": 1.7266942262649536, + "learning_rate": 1.4309651764877064e-05, + "loss": 3.2064, + "step": 9415 + }, + { + "epoch": 0.7549257381090778, + "grad_norm": 1.7634451389312744, + "learning_rate": 1.4308480050751016e-05, + "loss": 3.1488, + "step": 9416 + }, + { + "epoch": 0.7550059128901004, + "grad_norm": 1.8451341390609741, + "learning_rate": 1.430730826398552e-05, + "loss": 3.2236, + "step": 9417 + }, + { + "epoch": 0.7550860876711231, + "grad_norm": 1.787042498588562, + "learning_rate": 1.4306136404600319e-05, + "loss": 3.174, + "step": 9418 + }, + { + "epoch": 0.7551662624521457, + "grad_norm": 1.740871548652649, + "learning_rate": 1.4304964472615176e-05, + "loss": 3.1377, + "step": 9419 + }, + { + "epoch": 0.7552464372331683, + "grad_norm": 1.735097050666809, + "learning_rate": 1.4303792468049852e-05, + "loss": 3.2362, + "step": 9420 + }, + { + "epoch": 0.7553266120141909, + "grad_norm": 1.8376909494400024, + "learning_rate": 1.4302620390924102e-05, + "loss": 3.319, + "step": 9421 + }, + { + "epoch": 0.7554067867952136, + "grad_norm": 1.7096748352050781, + "learning_rate": 1.4301448241257692e-05, + "loss": 3.2538, + "step": 9422 + }, + { + "epoch": 0.7554869615762362, + "grad_norm": 1.8196969032287598, + "learning_rate": 1.4300276019070376e-05, + "loss": 3.2855, + "step": 9423 + }, + { + "epoch": 0.7555671363572588, + "grad_norm": 1.848101258277893, + "learning_rate": 1.4299103724381924e-05, + "loss": 3.2937, + "step": 9424 + }, + { + "epoch": 0.7556473111382814, + "grad_norm": 1.7730993032455444, + "learning_rate": 1.4297931357212099e-05, + "loss": 3.1936, + "step": 9425 + }, + { + "epoch": 0.755727485919304, + "grad_norm": 1.7736454010009766, + "learning_rate": 1.4296758917580665e-05, + "loss": 3.1625, + "step": 9426 + }, + { + "epoch": 0.7558076607003267, + "grad_norm": 1.8828072547912598, + "learning_rate": 1.4295586405507389e-05, + "loss": 3.1749, + "step": 9427 + }, + { + "epoch": 0.7558878354813493, + "grad_norm": 1.874184250831604, + "learning_rate": 1.4294413821012041e-05, + "loss": 3.1885, + "step": 9428 + }, + { + "epoch": 0.7559680102623719, + "grad_norm": 1.8535473346710205, + "learning_rate": 1.4293241164114388e-05, + "loss": 3.1887, + "step": 9429 + }, + { + "epoch": 0.7560481850433945, + "grad_norm": 1.840206265449524, + "learning_rate": 1.4292068434834208e-05, + "loss": 3.1073, + "step": 9430 + }, + { + "epoch": 0.7561283598244173, + "grad_norm": 1.7774945497512817, + "learning_rate": 1.429089563319126e-05, + "loss": 3.2249, + "step": 9431 + }, + { + "epoch": 0.7562085346054399, + "grad_norm": 2.0330028533935547, + "learning_rate": 1.4289722759205327e-05, + "loss": 3.2753, + "step": 9432 + }, + { + "epoch": 0.7562887093864625, + "grad_norm": 1.7754929065704346, + "learning_rate": 1.4288549812896179e-05, + "loss": 3.2576, + "step": 9433 + }, + { + "epoch": 0.7563688841674852, + "grad_norm": 1.8265596628189087, + "learning_rate": 1.4287376794283594e-05, + "loss": 3.194, + "step": 9434 + }, + { + "epoch": 0.7564490589485078, + "grad_norm": 1.772193193435669, + "learning_rate": 1.4286203703387347e-05, + "loss": 3.2347, + "step": 9435 + }, + { + "epoch": 0.7565292337295304, + "grad_norm": 1.8738274574279785, + "learning_rate": 1.4285030540227216e-05, + "loss": 3.2421, + "step": 9436 + }, + { + "epoch": 0.756609408510553, + "grad_norm": 1.8380295038223267, + "learning_rate": 1.428385730482298e-05, + "loss": 3.1043, + "step": 9437 + }, + { + "epoch": 0.7566895832915757, + "grad_norm": 1.809800386428833, + "learning_rate": 1.4282683997194418e-05, + "loss": 3.1835, + "step": 9438 + }, + { + "epoch": 0.7567697580725983, + "grad_norm": 1.8669072389602661, + "learning_rate": 1.4281510617361317e-05, + "loss": 3.3009, + "step": 9439 + }, + { + "epoch": 0.7568499328536209, + "grad_norm": 1.8280682563781738, + "learning_rate": 1.4280337165343453e-05, + "loss": 3.2446, + "step": 9440 + }, + { + "epoch": 0.7569301076346435, + "grad_norm": 1.7637280225753784, + "learning_rate": 1.4279163641160611e-05, + "loss": 3.2055, + "step": 9441 + }, + { + "epoch": 0.7570102824156661, + "grad_norm": 1.7752649784088135, + "learning_rate": 1.4277990044832582e-05, + "loss": 3.2996, + "step": 9442 + }, + { + "epoch": 0.7570904571966888, + "grad_norm": 1.8114954233169556, + "learning_rate": 1.4276816376379148e-05, + "loss": 3.3008, + "step": 9443 + }, + { + "epoch": 0.7571706319777114, + "grad_norm": 1.798179268836975, + "learning_rate": 1.4275642635820096e-05, + "loss": 3.3527, + "step": 9444 + }, + { + "epoch": 0.757250806758734, + "grad_norm": 1.9399224519729614, + "learning_rate": 1.4274468823175218e-05, + "loss": 3.1923, + "step": 9445 + }, + { + "epoch": 0.7573309815397566, + "grad_norm": 1.6919468641281128, + "learning_rate": 1.4273294938464299e-05, + "loss": 3.2361, + "step": 9446 + }, + { + "epoch": 0.7574111563207793, + "grad_norm": 1.8115648031234741, + "learning_rate": 1.4272120981707133e-05, + "loss": 3.2734, + "step": 9447 + }, + { + "epoch": 0.7574913311018019, + "grad_norm": 1.8925789594650269, + "learning_rate": 1.4270946952923518e-05, + "loss": 3.1641, + "step": 9448 + }, + { + "epoch": 0.7575715058828245, + "grad_norm": 1.7972413301467896, + "learning_rate": 1.426977285213324e-05, + "loss": 3.3107, + "step": 9449 + }, + { + "epoch": 0.7576516806638471, + "grad_norm": 1.8258310556411743, + "learning_rate": 1.4268598679356096e-05, + "loss": 3.172, + "step": 9450 + }, + { + "epoch": 0.7577318554448698, + "grad_norm": 1.8275865316390991, + "learning_rate": 1.4267424434611883e-05, + "loss": 3.3986, + "step": 9451 + }, + { + "epoch": 0.7578120302258925, + "grad_norm": 1.6955006122589111, + "learning_rate": 1.4266250117920396e-05, + "loss": 3.159, + "step": 9452 + }, + { + "epoch": 0.7578922050069151, + "grad_norm": 1.7473069429397583, + "learning_rate": 1.4265075729301439e-05, + "loss": 3.3173, + "step": 9453 + }, + { + "epoch": 0.7579723797879377, + "grad_norm": 0.7360371351242065, + "learning_rate": 1.4263901268774805e-05, + "loss": 2.4799, + "step": 9454 + }, + { + "epoch": 0.7580525545689604, + "grad_norm": 1.7289153337478638, + "learning_rate": 1.4262726736360301e-05, + "loss": 3.126, + "step": 9455 + }, + { + "epoch": 0.758132729349983, + "grad_norm": 1.763871669769287, + "learning_rate": 1.4261552132077725e-05, + "loss": 3.1113, + "step": 9456 + }, + { + "epoch": 0.7582129041310056, + "grad_norm": 1.815438985824585, + "learning_rate": 1.4260377455946884e-05, + "loss": 3.2371, + "step": 9457 + }, + { + "epoch": 0.7582930789120282, + "grad_norm": 1.768264651298523, + "learning_rate": 1.4259202707987579e-05, + "loss": 3.1967, + "step": 9458 + }, + { + "epoch": 0.7583732536930509, + "grad_norm": 1.8735255002975464, + "learning_rate": 1.4258027888219616e-05, + "loss": 3.2932, + "step": 9459 + }, + { + "epoch": 0.7584534284740735, + "grad_norm": 1.7643721103668213, + "learning_rate": 1.4256852996662806e-05, + "loss": 3.2264, + "step": 9460 + }, + { + "epoch": 0.7585336032550961, + "grad_norm": 1.710517406463623, + "learning_rate": 1.4255678033336953e-05, + "loss": 3.1523, + "step": 9461 + }, + { + "epoch": 0.7586137780361187, + "grad_norm": 1.8460198640823364, + "learning_rate": 1.4254502998261868e-05, + "loss": 3.2861, + "step": 9462 + }, + { + "epoch": 0.7586939528171414, + "grad_norm": 1.7602949142456055, + "learning_rate": 1.4253327891457364e-05, + "loss": 3.2717, + "step": 9463 + }, + { + "epoch": 0.758774127598164, + "grad_norm": 1.8643674850463867, + "learning_rate": 1.425215271294325e-05, + "loss": 3.251, + "step": 9464 + }, + { + "epoch": 0.7588543023791866, + "grad_norm": 1.8996959924697876, + "learning_rate": 1.4250977462739339e-05, + "loss": 3.2195, + "step": 9465 + }, + { + "epoch": 0.7589344771602092, + "grad_norm": 2.009155750274658, + "learning_rate": 1.4249802140865448e-05, + "loss": 3.2491, + "step": 9466 + }, + { + "epoch": 0.7590146519412319, + "grad_norm": 1.7601845264434814, + "learning_rate": 1.4248626747341388e-05, + "loss": 3.2253, + "step": 9467 + }, + { + "epoch": 0.7590948267222545, + "grad_norm": 1.6683248281478882, + "learning_rate": 1.4247451282186981e-05, + "loss": 3.0652, + "step": 9468 + }, + { + "epoch": 0.7591750015032771, + "grad_norm": 1.868208885192871, + "learning_rate": 1.424627574542204e-05, + "loss": 3.2031, + "step": 9469 + }, + { + "epoch": 0.7592551762842997, + "grad_norm": 1.8147941827774048, + "learning_rate": 1.424510013706639e-05, + "loss": 3.243, + "step": 9470 + }, + { + "epoch": 0.7593353510653224, + "grad_norm": 1.9126447439193726, + "learning_rate": 1.4243924457139845e-05, + "loss": 3.2003, + "step": 9471 + }, + { + "epoch": 0.7594155258463451, + "grad_norm": 1.8217359781265259, + "learning_rate": 1.4242748705662228e-05, + "loss": 3.2019, + "step": 9472 + }, + { + "epoch": 0.7594957006273677, + "grad_norm": 1.797921895980835, + "learning_rate": 1.4241572882653367e-05, + "loss": 3.2299, + "step": 9473 + }, + { + "epoch": 0.7595758754083903, + "grad_norm": 1.7546381950378418, + "learning_rate": 1.4240396988133079e-05, + "loss": 3.2814, + "step": 9474 + }, + { + "epoch": 0.759656050189413, + "grad_norm": 1.7639518976211548, + "learning_rate": 1.4239221022121195e-05, + "loss": 3.2609, + "step": 9475 + }, + { + "epoch": 0.7597362249704356, + "grad_norm": 1.7283539772033691, + "learning_rate": 1.4238044984637536e-05, + "loss": 3.3013, + "step": 9476 + }, + { + "epoch": 0.7598163997514582, + "grad_norm": 1.8053463697433472, + "learning_rate": 1.4236868875701931e-05, + "loss": 3.2755, + "step": 9477 + }, + { + "epoch": 0.7598965745324808, + "grad_norm": 1.762336254119873, + "learning_rate": 1.4235692695334213e-05, + "loss": 3.2716, + "step": 9478 + }, + { + "epoch": 0.7599767493135035, + "grad_norm": 1.8412829637527466, + "learning_rate": 1.4234516443554206e-05, + "loss": 3.2392, + "step": 9479 + }, + { + "epoch": 0.7600569240945261, + "grad_norm": 1.7497724294662476, + "learning_rate": 1.423334012038175e-05, + "loss": 3.2194, + "step": 9480 + }, + { + "epoch": 0.7601370988755487, + "grad_norm": 1.7548551559448242, + "learning_rate": 1.4232163725836666e-05, + "loss": 3.0557, + "step": 9481 + }, + { + "epoch": 0.7602172736565713, + "grad_norm": 1.855183482170105, + "learning_rate": 1.4230987259938793e-05, + "loss": 3.0906, + "step": 9482 + }, + { + "epoch": 0.760297448437594, + "grad_norm": 1.8748023509979248, + "learning_rate": 1.4229810722707968e-05, + "loss": 3.3605, + "step": 9483 + }, + { + "epoch": 0.7603776232186166, + "grad_norm": 1.7074553966522217, + "learning_rate": 1.4228634114164023e-05, + "loss": 3.121, + "step": 9484 + }, + { + "epoch": 0.7604577979996392, + "grad_norm": 1.8219643831253052, + "learning_rate": 1.42274574343268e-05, + "loss": 3.2964, + "step": 9485 + }, + { + "epoch": 0.7605379727806618, + "grad_norm": 1.7967473268508911, + "learning_rate": 1.4226280683216133e-05, + "loss": 3.2244, + "step": 9486 + }, + { + "epoch": 0.7606181475616844, + "grad_norm": 1.7522488832473755, + "learning_rate": 1.4225103860851863e-05, + "loss": 3.2301, + "step": 9487 + }, + { + "epoch": 0.7606983223427071, + "grad_norm": 1.7686339616775513, + "learning_rate": 1.4223926967253833e-05, + "loss": 3.2187, + "step": 9488 + }, + { + "epoch": 0.7607784971237297, + "grad_norm": 1.862191081047058, + "learning_rate": 1.4222750002441881e-05, + "loss": 3.2894, + "step": 9489 + }, + { + "epoch": 0.7608586719047523, + "grad_norm": 1.7262300252914429, + "learning_rate": 1.4221572966435854e-05, + "loss": 3.1659, + "step": 9490 + }, + { + "epoch": 0.7609388466857749, + "grad_norm": 1.9997464418411255, + "learning_rate": 1.4220395859255594e-05, + "loss": 3.3273, + "step": 9491 + }, + { + "epoch": 0.7610190214667977, + "grad_norm": 1.8459988832473755, + "learning_rate": 1.4219218680920945e-05, + "loss": 3.1244, + "step": 9492 + }, + { + "epoch": 0.7610991962478203, + "grad_norm": 1.9164659976959229, + "learning_rate": 1.4218041431451758e-05, + "loss": 3.293, + "step": 9493 + }, + { + "epoch": 0.7611793710288429, + "grad_norm": 1.9094958305358887, + "learning_rate": 1.4216864110867879e-05, + "loss": 3.2836, + "step": 9494 + }, + { + "epoch": 0.7612595458098655, + "grad_norm": 1.8486281633377075, + "learning_rate": 1.4215686719189157e-05, + "loss": 3.2254, + "step": 9495 + }, + { + "epoch": 0.7613397205908882, + "grad_norm": 1.8323960304260254, + "learning_rate": 1.4214509256435442e-05, + "loss": 3.3175, + "step": 9496 + }, + { + "epoch": 0.7614198953719108, + "grad_norm": 1.810973882675171, + "learning_rate": 1.4213331722626587e-05, + "loss": 3.1827, + "step": 9497 + }, + { + "epoch": 0.7615000701529334, + "grad_norm": 1.7812284231185913, + "learning_rate": 1.4212154117782444e-05, + "loss": 3.1252, + "step": 9498 + }, + { + "epoch": 0.761580244933956, + "grad_norm": 1.7580026388168335, + "learning_rate": 1.4210976441922866e-05, + "loss": 3.2254, + "step": 9499 + }, + { + "epoch": 0.7616604197149787, + "grad_norm": 1.9047298431396484, + "learning_rate": 1.4209798695067712e-05, + "loss": 3.1285, + "step": 9500 + }, + { + "epoch": 0.7617405944960013, + "grad_norm": 1.8474221229553223, + "learning_rate": 1.4208620877236832e-05, + "loss": 3.221, + "step": 9501 + }, + { + "epoch": 0.7618207692770239, + "grad_norm": 1.783278226852417, + "learning_rate": 1.4207442988450089e-05, + "loss": 3.1071, + "step": 9502 + }, + { + "epoch": 0.7619009440580465, + "grad_norm": 1.8186224699020386, + "learning_rate": 1.4206265028727342e-05, + "loss": 3.1599, + "step": 9503 + }, + { + "epoch": 0.7619811188390692, + "grad_norm": 1.7565646171569824, + "learning_rate": 1.4205086998088447e-05, + "loss": 3.1691, + "step": 9504 + }, + { + "epoch": 0.7620612936200918, + "grad_norm": 0.7340672612190247, + "learning_rate": 1.4203908896553266e-05, + "loss": 2.4626, + "step": 9505 + }, + { + "epoch": 0.7621414684011144, + "grad_norm": 1.7689262628555298, + "learning_rate": 1.4202730724141664e-05, + "loss": 3.1488, + "step": 9506 + }, + { + "epoch": 0.762221643182137, + "grad_norm": 1.793953776359558, + "learning_rate": 1.4201552480873505e-05, + "loss": 3.1472, + "step": 9507 + }, + { + "epoch": 0.7623018179631597, + "grad_norm": 1.7400684356689453, + "learning_rate": 1.4200374166768649e-05, + "loss": 3.1244, + "step": 9508 + }, + { + "epoch": 0.7623819927441823, + "grad_norm": 1.8310014009475708, + "learning_rate": 1.4199195781846966e-05, + "loss": 3.292, + "step": 9509 + }, + { + "epoch": 0.7624621675252049, + "grad_norm": 1.6181929111480713, + "learning_rate": 1.4198017326128324e-05, + "loss": 3.1736, + "step": 9510 + }, + { + "epoch": 0.7625423423062275, + "grad_norm": 1.8507007360458374, + "learning_rate": 1.4196838799632587e-05, + "loss": 3.3059, + "step": 9511 + }, + { + "epoch": 0.7626225170872502, + "grad_norm": 1.8227944374084473, + "learning_rate": 1.4195660202379626e-05, + "loss": 3.1558, + "step": 9512 + }, + { + "epoch": 0.7627026918682729, + "grad_norm": 1.7865558862686157, + "learning_rate": 1.4194481534389314e-05, + "loss": 3.2084, + "step": 9513 + }, + { + "epoch": 0.7627828666492955, + "grad_norm": 1.7689192295074463, + "learning_rate": 1.4193302795681523e-05, + "loss": 3.2734, + "step": 9514 + }, + { + "epoch": 0.7628630414303181, + "grad_norm": 1.816599726676941, + "learning_rate": 1.4192123986276125e-05, + "loss": 3.2577, + "step": 9515 + }, + { + "epoch": 0.7629432162113408, + "grad_norm": 1.8429162502288818, + "learning_rate": 1.4190945106192993e-05, + "loss": 3.2218, + "step": 9516 + }, + { + "epoch": 0.7630233909923634, + "grad_norm": 1.861351728439331, + "learning_rate": 1.4189766155452e-05, + "loss": 3.282, + "step": 9517 + }, + { + "epoch": 0.763103565773386, + "grad_norm": 0.7349299788475037, + "learning_rate": 1.4188587134073031e-05, + "loss": 2.4833, + "step": 9518 + }, + { + "epoch": 0.7631837405544086, + "grad_norm": 1.7802518606185913, + "learning_rate": 1.4187408042075958e-05, + "loss": 3.2534, + "step": 9519 + }, + { + "epoch": 0.7632639153354313, + "grad_norm": 1.881116271018982, + "learning_rate": 1.4186228879480662e-05, + "loss": 3.277, + "step": 9520 + }, + { + "epoch": 0.7633440901164539, + "grad_norm": 1.828372597694397, + "learning_rate": 1.4185049646307021e-05, + "loss": 3.2245, + "step": 9521 + }, + { + "epoch": 0.7634242648974765, + "grad_norm": 2.57749080657959, + "learning_rate": 1.4183870342574918e-05, + "loss": 3.1997, + "step": 9522 + }, + { + "epoch": 0.7635044396784991, + "grad_norm": 1.8500679731369019, + "learning_rate": 1.4182690968304233e-05, + "loss": 3.2469, + "step": 9523 + }, + { + "epoch": 0.7635846144595217, + "grad_norm": 1.8063786029815674, + "learning_rate": 1.4181511523514857e-05, + "loss": 3.2241, + "step": 9524 + }, + { + "epoch": 0.7636647892405444, + "grad_norm": 0.756881594657898, + "learning_rate": 1.418033200822667e-05, + "loss": 2.4795, + "step": 9525 + }, + { + "epoch": 0.763744964021567, + "grad_norm": 1.7617501020431519, + "learning_rate": 1.4179152422459554e-05, + "loss": 3.1971, + "step": 9526 + }, + { + "epoch": 0.7638251388025896, + "grad_norm": 1.9254515171051025, + "learning_rate": 1.4177972766233406e-05, + "loss": 3.2354, + "step": 9527 + }, + { + "epoch": 0.7639053135836122, + "grad_norm": 1.8981010913848877, + "learning_rate": 1.4176793039568106e-05, + "loss": 3.2099, + "step": 9528 + }, + { + "epoch": 0.7639854883646349, + "grad_norm": 1.902539849281311, + "learning_rate": 1.417561324248355e-05, + "loss": 3.2469, + "step": 9529 + }, + { + "epoch": 0.7640656631456575, + "grad_norm": 1.8857924938201904, + "learning_rate": 1.4174433374999627e-05, + "loss": 3.2736, + "step": 9530 + }, + { + "epoch": 0.7641458379266801, + "grad_norm": 1.9554170370101929, + "learning_rate": 1.4173253437136225e-05, + "loss": 3.2693, + "step": 9531 + }, + { + "epoch": 0.7642260127077027, + "grad_norm": 1.7391940355300903, + "learning_rate": 1.4172073428913243e-05, + "loss": 3.1996, + "step": 9532 + }, + { + "epoch": 0.7643061874887255, + "grad_norm": 2.0568385124206543, + "learning_rate": 1.4170893350350573e-05, + "loss": 3.2415, + "step": 9533 + }, + { + "epoch": 0.7643863622697481, + "grad_norm": 1.681932806968689, + "learning_rate": 1.4169713201468111e-05, + "loss": 3.2119, + "step": 9534 + }, + { + "epoch": 0.7644665370507707, + "grad_norm": 1.7887464761734009, + "learning_rate": 1.4168532982285756e-05, + "loss": 3.1972, + "step": 9535 + }, + { + "epoch": 0.7645467118317933, + "grad_norm": 1.9855132102966309, + "learning_rate": 1.4167352692823402e-05, + "loss": 3.2677, + "step": 9536 + }, + { + "epoch": 0.764626886612816, + "grad_norm": 1.7859936952590942, + "learning_rate": 1.416617233310095e-05, + "loss": 3.1768, + "step": 9537 + }, + { + "epoch": 0.7647070613938386, + "grad_norm": 1.7976912260055542, + "learning_rate": 1.4164991903138303e-05, + "loss": 3.2574, + "step": 9538 + }, + { + "epoch": 0.7647872361748612, + "grad_norm": 2.1374073028564453, + "learning_rate": 1.4163811402955357e-05, + "loss": 3.2783, + "step": 9539 + }, + { + "epoch": 0.7648674109558838, + "grad_norm": 1.8273683786392212, + "learning_rate": 1.4162630832572022e-05, + "loss": 3.173, + "step": 9540 + }, + { + "epoch": 0.7649475857369065, + "grad_norm": 1.8164293766021729, + "learning_rate": 1.4161450192008195e-05, + "loss": 3.2434, + "step": 9541 + }, + { + "epoch": 0.7650277605179291, + "grad_norm": 1.8741800785064697, + "learning_rate": 1.4160269481283786e-05, + "loss": 3.2784, + "step": 9542 + }, + { + "epoch": 0.7651079352989517, + "grad_norm": 1.807839035987854, + "learning_rate": 1.41590887004187e-05, + "loss": 3.1852, + "step": 9543 + }, + { + "epoch": 0.7651881100799743, + "grad_norm": 1.8318264484405518, + "learning_rate": 1.4157907849432842e-05, + "loss": 3.2935, + "step": 9544 + }, + { + "epoch": 0.765268284860997, + "grad_norm": 1.8265676498413086, + "learning_rate": 1.4156726928346127e-05, + "loss": 3.1398, + "step": 9545 + }, + { + "epoch": 0.7653484596420196, + "grad_norm": 1.8415497541427612, + "learning_rate": 1.415554593717846e-05, + "loss": 3.1843, + "step": 9546 + }, + { + "epoch": 0.7654286344230422, + "grad_norm": 1.7703860998153687, + "learning_rate": 1.4154364875949749e-05, + "loss": 3.1968, + "step": 9547 + }, + { + "epoch": 0.7655088092040648, + "grad_norm": 2.0129306316375732, + "learning_rate": 1.4153183744679914e-05, + "loss": 3.3274, + "step": 9548 + }, + { + "epoch": 0.7655889839850875, + "grad_norm": 1.8682836294174194, + "learning_rate": 1.4152002543388862e-05, + "loss": 3.1058, + "step": 9549 + }, + { + "epoch": 0.7656691587661101, + "grad_norm": 1.7704204320907593, + "learning_rate": 1.4150821272096516e-05, + "loss": 3.2216, + "step": 9550 + }, + { + "epoch": 0.7657493335471327, + "grad_norm": 1.8350920677185059, + "learning_rate": 1.414963993082278e-05, + "loss": 3.2172, + "step": 9551 + }, + { + "epoch": 0.7658295083281553, + "grad_norm": 1.7608978748321533, + "learning_rate": 1.414845851958758e-05, + "loss": 3.1321, + "step": 9552 + }, + { + "epoch": 0.7659096831091781, + "grad_norm": 0.7449271082878113, + "learning_rate": 1.4147277038410831e-05, + "loss": 2.49, + "step": 9553 + }, + { + "epoch": 0.7659898578902007, + "grad_norm": 1.8147577047348022, + "learning_rate": 1.4146095487312454e-05, + "loss": 3.1979, + "step": 9554 + }, + { + "epoch": 0.7660700326712233, + "grad_norm": 1.9434350728988647, + "learning_rate": 1.4144913866312369e-05, + "loss": 3.2418, + "step": 9555 + }, + { + "epoch": 0.7661502074522459, + "grad_norm": 1.790204644203186, + "learning_rate": 1.4143732175430494e-05, + "loss": 3.1492, + "step": 9556 + }, + { + "epoch": 0.7662303822332686, + "grad_norm": 1.775905966758728, + "learning_rate": 1.4142550414686756e-05, + "loss": 3.1612, + "step": 9557 + }, + { + "epoch": 0.7663105570142912, + "grad_norm": 1.8110647201538086, + "learning_rate": 1.4141368584101078e-05, + "loss": 3.431, + "step": 9558 + }, + { + "epoch": 0.7663907317953138, + "grad_norm": 1.819069504737854, + "learning_rate": 1.4140186683693386e-05, + "loss": 3.1979, + "step": 9559 + }, + { + "epoch": 0.7664709065763364, + "grad_norm": 1.8399510383605957, + "learning_rate": 1.4139004713483606e-05, + "loss": 3.1871, + "step": 9560 + }, + { + "epoch": 0.766551081357359, + "grad_norm": 1.837546944618225, + "learning_rate": 1.4137822673491664e-05, + "loss": 3.2669, + "step": 9561 + }, + { + "epoch": 0.7666312561383817, + "grad_norm": 1.8415855169296265, + "learning_rate": 1.413664056373749e-05, + "loss": 3.2548, + "step": 9562 + }, + { + "epoch": 0.7667114309194043, + "grad_norm": 1.7933984994888306, + "learning_rate": 1.4135458384241015e-05, + "loss": 3.1603, + "step": 9563 + }, + { + "epoch": 0.7667916057004269, + "grad_norm": 1.7239822149276733, + "learning_rate": 1.4134276135022167e-05, + "loss": 3.1856, + "step": 9564 + }, + { + "epoch": 0.7668717804814495, + "grad_norm": 2.0140445232391357, + "learning_rate": 1.4133093816100884e-05, + "loss": 3.2113, + "step": 9565 + }, + { + "epoch": 0.7669519552624722, + "grad_norm": 1.9072662591934204, + "learning_rate": 1.4131911427497094e-05, + "loss": 3.1536, + "step": 9566 + }, + { + "epoch": 0.7670321300434948, + "grad_norm": 1.8065882921218872, + "learning_rate": 1.4130728969230731e-05, + "loss": 3.1777, + "step": 9567 + }, + { + "epoch": 0.7671123048245174, + "grad_norm": 1.811501383781433, + "learning_rate": 1.4129546441321738e-05, + "loss": 3.2402, + "step": 9568 + }, + { + "epoch": 0.76719247960554, + "grad_norm": 1.8155797719955444, + "learning_rate": 1.4128363843790043e-05, + "loss": 3.1372, + "step": 9569 + }, + { + "epoch": 0.7672726543865627, + "grad_norm": 1.7459537982940674, + "learning_rate": 1.4127181176655593e-05, + "loss": 3.1872, + "step": 9570 + }, + { + "epoch": 0.7673528291675853, + "grad_norm": 1.7575422525405884, + "learning_rate": 1.412599843993832e-05, + "loss": 3.1535, + "step": 9571 + }, + { + "epoch": 0.7674330039486079, + "grad_norm": 1.758505940437317, + "learning_rate": 1.4124815633658167e-05, + "loss": 3.1719, + "step": 9572 + }, + { + "epoch": 0.7675131787296305, + "grad_norm": 1.8895821571350098, + "learning_rate": 1.4123632757835078e-05, + "loss": 3.1917, + "step": 9573 + }, + { + "epoch": 0.7675933535106533, + "grad_norm": 1.76833975315094, + "learning_rate": 1.4122449812488993e-05, + "loss": 3.1576, + "step": 9574 + }, + { + "epoch": 0.7676735282916759, + "grad_norm": 1.7879263162612915, + "learning_rate": 1.412126679763986e-05, + "loss": 3.0864, + "step": 9575 + }, + { + "epoch": 0.7677537030726985, + "grad_norm": 1.7474614381790161, + "learning_rate": 1.4120083713307618e-05, + "loss": 3.2269, + "step": 9576 + }, + { + "epoch": 0.7678338778537211, + "grad_norm": 1.7934406995773315, + "learning_rate": 1.4118900559512216e-05, + "loss": 3.1312, + "step": 9577 + }, + { + "epoch": 0.7679140526347438, + "grad_norm": 1.7826294898986816, + "learning_rate": 1.4117717336273605e-05, + "loss": 3.1298, + "step": 9578 + }, + { + "epoch": 0.7679942274157664, + "grad_norm": 1.7020374536514282, + "learning_rate": 1.4116534043611727e-05, + "loss": 3.1208, + "step": 9579 + }, + { + "epoch": 0.768074402196789, + "grad_norm": 1.7370777130126953, + "learning_rate": 1.411535068154654e-05, + "loss": 3.1458, + "step": 9580 + }, + { + "epoch": 0.7681545769778116, + "grad_norm": 1.7693103551864624, + "learning_rate": 1.4114167250097988e-05, + "loss": 3.1781, + "step": 9581 + }, + { + "epoch": 0.7682347517588343, + "grad_norm": 1.840341567993164, + "learning_rate": 1.4112983749286027e-05, + "loss": 3.1956, + "step": 9582 + }, + { + "epoch": 0.7683149265398569, + "grad_norm": 1.752777338027954, + "learning_rate": 1.411180017913061e-05, + "loss": 3.1956, + "step": 9583 + }, + { + "epoch": 0.7683951013208795, + "grad_norm": 1.7295517921447754, + "learning_rate": 1.4110616539651691e-05, + "loss": 3.1626, + "step": 9584 + }, + { + "epoch": 0.7684752761019021, + "grad_norm": 1.7967212200164795, + "learning_rate": 1.4109432830869228e-05, + "loss": 3.1993, + "step": 9585 + }, + { + "epoch": 0.7685554508829248, + "grad_norm": 1.659613013267517, + "learning_rate": 1.410824905280317e-05, + "loss": 3.1218, + "step": 9586 + }, + { + "epoch": 0.7686356256639474, + "grad_norm": 1.7671403884887695, + "learning_rate": 1.4107065205473486e-05, + "loss": 3.2456, + "step": 9587 + }, + { + "epoch": 0.76871580044497, + "grad_norm": 1.8133624792099, + "learning_rate": 1.4105881288900129e-05, + "loss": 3.2961, + "step": 9588 + }, + { + "epoch": 0.7687959752259926, + "grad_norm": 0.8274117708206177, + "learning_rate": 1.4104697303103062e-05, + "loss": 2.5167, + "step": 9589 + }, + { + "epoch": 0.7688761500070153, + "grad_norm": 1.7991710901260376, + "learning_rate": 1.4103513248102242e-05, + "loss": 3.2429, + "step": 9590 + }, + { + "epoch": 0.7689563247880379, + "grad_norm": 1.8500385284423828, + "learning_rate": 1.4102329123917635e-05, + "loss": 3.2198, + "step": 9591 + }, + { + "epoch": 0.7690364995690605, + "grad_norm": 0.7101055979728699, + "learning_rate": 1.4101144930569207e-05, + "loss": 2.4822, + "step": 9592 + }, + { + "epoch": 0.7691166743500831, + "grad_norm": 1.9272159337997437, + "learning_rate": 1.409996066807692e-05, + "loss": 3.158, + "step": 9593 + }, + { + "epoch": 0.7691968491311059, + "grad_norm": 1.8703898191452026, + "learning_rate": 1.409877633646074e-05, + "loss": 3.2025, + "step": 9594 + }, + { + "epoch": 0.7692770239121285, + "grad_norm": 1.7016077041625977, + "learning_rate": 1.4097591935740638e-05, + "loss": 3.1514, + "step": 9595 + }, + { + "epoch": 0.7693571986931511, + "grad_norm": 1.8552738428115845, + "learning_rate": 1.4096407465936577e-05, + "loss": 3.1937, + "step": 9596 + }, + { + "epoch": 0.7694373734741737, + "grad_norm": 1.8194670677185059, + "learning_rate": 1.409522292706853e-05, + "loss": 3.1832, + "step": 9597 + }, + { + "epoch": 0.7695175482551964, + "grad_norm": 1.7436037063598633, + "learning_rate": 1.409403831915647e-05, + "loss": 3.1473, + "step": 9598 + }, + { + "epoch": 0.769597723036219, + "grad_norm": 1.8250364065170288, + "learning_rate": 1.4092853642220365e-05, + "loss": 3.2271, + "step": 9599 + }, + { + "epoch": 0.7696778978172416, + "grad_norm": 1.7006239891052246, + "learning_rate": 1.4091668896280193e-05, + "loss": 3.0651, + "step": 9600 + }, + { + "epoch": 0.7697580725982642, + "grad_norm": 1.7243398427963257, + "learning_rate": 1.4090484081355922e-05, + "loss": 3.1005, + "step": 9601 + }, + { + "epoch": 0.7698382473792869, + "grad_norm": 1.789017677307129, + "learning_rate": 1.4089299197467532e-05, + "loss": 3.1229, + "step": 9602 + }, + { + "epoch": 0.7699184221603095, + "grad_norm": 1.6563583612442017, + "learning_rate": 1.4088114244635e-05, + "loss": 3.2717, + "step": 9603 + }, + { + "epoch": 0.7699985969413321, + "grad_norm": 1.7781459093093872, + "learning_rate": 1.40869292228783e-05, + "loss": 3.2056, + "step": 9604 + }, + { + "epoch": 0.7700787717223547, + "grad_norm": 1.9075838327407837, + "learning_rate": 1.4085744132217417e-05, + "loss": 3.2418, + "step": 9605 + }, + { + "epoch": 0.7701589465033774, + "grad_norm": 0.7269071340560913, + "learning_rate": 1.4084558972672327e-05, + "loss": 2.4938, + "step": 9606 + }, + { + "epoch": 0.7702391212844, + "grad_norm": 1.7613412141799927, + "learning_rate": 1.408337374426301e-05, + "loss": 3.1208, + "step": 9607 + }, + { + "epoch": 0.7703192960654226, + "grad_norm": 1.9279122352600098, + "learning_rate": 1.4082188447009453e-05, + "loss": 3.219, + "step": 9608 + }, + { + "epoch": 0.7703994708464452, + "grad_norm": 1.7958003282546997, + "learning_rate": 1.4081003080931636e-05, + "loss": 3.1097, + "step": 9609 + }, + { + "epoch": 0.7704796456274678, + "grad_norm": 1.776288628578186, + "learning_rate": 1.4079817646049547e-05, + "loss": 3.1288, + "step": 9610 + }, + { + "epoch": 0.7705598204084905, + "grad_norm": 1.765720248222351, + "learning_rate": 1.4078632142383169e-05, + "loss": 3.1412, + "step": 9611 + }, + { + "epoch": 0.7706399951895131, + "grad_norm": 1.6293082237243652, + "learning_rate": 1.4077446569952491e-05, + "loss": 3.1184, + "step": 9612 + }, + { + "epoch": 0.7707201699705357, + "grad_norm": 1.8032253980636597, + "learning_rate": 1.4076260928777502e-05, + "loss": 3.2643, + "step": 9613 + }, + { + "epoch": 0.7708003447515585, + "grad_norm": 1.922435998916626, + "learning_rate": 1.4075075218878187e-05, + "loss": 3.1224, + "step": 9614 + }, + { + "epoch": 0.7708805195325811, + "grad_norm": 1.8107832670211792, + "learning_rate": 1.4073889440274544e-05, + "loss": 3.1973, + "step": 9615 + }, + { + "epoch": 0.7709606943136037, + "grad_norm": 1.8132891654968262, + "learning_rate": 1.4072703592986558e-05, + "loss": 3.2551, + "step": 9616 + }, + { + "epoch": 0.7710408690946263, + "grad_norm": 1.790281057357788, + "learning_rate": 1.4071517677034225e-05, + "loss": 3.2833, + "step": 9617 + }, + { + "epoch": 0.771121043875649, + "grad_norm": 1.8345528841018677, + "learning_rate": 1.407033169243754e-05, + "loss": 3.1477, + "step": 9618 + }, + { + "epoch": 0.7712012186566716, + "grad_norm": 1.7696856260299683, + "learning_rate": 1.4069145639216497e-05, + "loss": 3.214, + "step": 9619 + }, + { + "epoch": 0.7712813934376942, + "grad_norm": 1.8369581699371338, + "learning_rate": 1.4067959517391095e-05, + "loss": 3.2871, + "step": 9620 + }, + { + "epoch": 0.7713615682187168, + "grad_norm": 0.7607106566429138, + "learning_rate": 1.4066773326981325e-05, + "loss": 2.5049, + "step": 9621 + }, + { + "epoch": 0.7714417429997394, + "grad_norm": 1.75460684299469, + "learning_rate": 1.4065587068007192e-05, + "loss": 3.2034, + "step": 9622 + }, + { + "epoch": 0.7715219177807621, + "grad_norm": 0.754281222820282, + "learning_rate": 1.4064400740488693e-05, + "loss": 2.4954, + "step": 9623 + }, + { + "epoch": 0.7716020925617847, + "grad_norm": 1.700997233390808, + "learning_rate": 1.406321434444583e-05, + "loss": 3.1938, + "step": 9624 + }, + { + "epoch": 0.7716822673428073, + "grad_norm": 1.8254926204681396, + "learning_rate": 1.4062027879898606e-05, + "loss": 3.2617, + "step": 9625 + }, + { + "epoch": 0.7717624421238299, + "grad_norm": 1.7853167057037354, + "learning_rate": 1.4060841346867025e-05, + "loss": 3.2357, + "step": 9626 + }, + { + "epoch": 0.7718426169048526, + "grad_norm": 1.7751402854919434, + "learning_rate": 1.4059654745371085e-05, + "loss": 3.1544, + "step": 9627 + }, + { + "epoch": 0.7719227916858752, + "grad_norm": 1.9719398021697998, + "learning_rate": 1.4058468075430801e-05, + "loss": 3.0958, + "step": 9628 + }, + { + "epoch": 0.7720029664668978, + "grad_norm": 1.8929308652877808, + "learning_rate": 1.4057281337066175e-05, + "loss": 3.2282, + "step": 9629 + }, + { + "epoch": 0.7720831412479204, + "grad_norm": 1.98374342918396, + "learning_rate": 1.4056094530297216e-05, + "loss": 3.1705, + "step": 9630 + }, + { + "epoch": 0.7721633160289431, + "grad_norm": 1.8089945316314697, + "learning_rate": 1.4054907655143932e-05, + "loss": 3.2501, + "step": 9631 + }, + { + "epoch": 0.7722434908099657, + "grad_norm": 1.8423535823822021, + "learning_rate": 1.4053720711626334e-05, + "loss": 3.1591, + "step": 9632 + }, + { + "epoch": 0.7723236655909883, + "grad_norm": 1.785601258277893, + "learning_rate": 1.4052533699764435e-05, + "loss": 3.2152, + "step": 9633 + }, + { + "epoch": 0.7724038403720109, + "grad_norm": 0.7134716510772705, + "learning_rate": 1.4051346619578243e-05, + "loss": 2.4902, + "step": 9634 + }, + { + "epoch": 0.7724840151530337, + "grad_norm": 1.8450417518615723, + "learning_rate": 1.4050159471087776e-05, + "loss": 3.3206, + "step": 9635 + }, + { + "epoch": 0.7725641899340563, + "grad_norm": 1.7960506677627563, + "learning_rate": 1.4048972254313049e-05, + "loss": 3.2878, + "step": 9636 + }, + { + "epoch": 0.7726443647150789, + "grad_norm": 1.7198896408081055, + "learning_rate": 1.4047784969274077e-05, + "loss": 3.2812, + "step": 9637 + }, + { + "epoch": 0.7727245394961015, + "grad_norm": 1.7949849367141724, + "learning_rate": 1.4046597615990876e-05, + "loss": 3.2153, + "step": 9638 + }, + { + "epoch": 0.7728047142771242, + "grad_norm": 1.7923003435134888, + "learning_rate": 1.4045410194483464e-05, + "loss": 3.2748, + "step": 9639 + }, + { + "epoch": 0.7728848890581468, + "grad_norm": 1.797872543334961, + "learning_rate": 1.4044222704771863e-05, + "loss": 3.2168, + "step": 9640 + }, + { + "epoch": 0.7729650638391694, + "grad_norm": 1.9245104789733887, + "learning_rate": 1.4043035146876094e-05, + "loss": 3.1873, + "step": 9641 + }, + { + "epoch": 0.773045238620192, + "grad_norm": 1.890134334564209, + "learning_rate": 1.4041847520816173e-05, + "loss": 3.2431, + "step": 9642 + }, + { + "epoch": 0.7731254134012147, + "grad_norm": 1.7881661653518677, + "learning_rate": 1.4040659826612134e-05, + "loss": 3.1273, + "step": 9643 + }, + { + "epoch": 0.7732055881822373, + "grad_norm": 1.750365138053894, + "learning_rate": 1.403947206428399e-05, + "loss": 3.2534, + "step": 9644 + }, + { + "epoch": 0.7732857629632599, + "grad_norm": 1.914942979812622, + "learning_rate": 1.4038284233851768e-05, + "loss": 3.2665, + "step": 9645 + }, + { + "epoch": 0.7733659377442825, + "grad_norm": 1.7562332153320312, + "learning_rate": 1.4037096335335503e-05, + "loss": 3.1553, + "step": 9646 + }, + { + "epoch": 0.7734461125253052, + "grad_norm": 1.8243558406829834, + "learning_rate": 1.4035908368755213e-05, + "loss": 3.181, + "step": 9647 + }, + { + "epoch": 0.7735262873063278, + "grad_norm": 1.7682243585586548, + "learning_rate": 1.4034720334130934e-05, + "loss": 3.1252, + "step": 9648 + }, + { + "epoch": 0.7736064620873504, + "grad_norm": 1.7333588600158691, + "learning_rate": 1.4033532231482688e-05, + "loss": 3.215, + "step": 9649 + }, + { + "epoch": 0.773686636868373, + "grad_norm": 1.7881778478622437, + "learning_rate": 1.403234406083051e-05, + "loss": 3.1642, + "step": 9650 + }, + { + "epoch": 0.7737668116493956, + "grad_norm": 0.8355886936187744, + "learning_rate": 1.4031155822194435e-05, + "loss": 2.4792, + "step": 9651 + }, + { + "epoch": 0.7738469864304183, + "grad_norm": 1.7498494386672974, + "learning_rate": 1.4029967515594492e-05, + "loss": 3.1806, + "step": 9652 + }, + { + "epoch": 0.7739271612114409, + "grad_norm": 1.920203447341919, + "learning_rate": 1.402877914105072e-05, + "loss": 3.1589, + "step": 9653 + }, + { + "epoch": 0.7740073359924635, + "grad_norm": 1.9082170724868774, + "learning_rate": 1.4027590698583146e-05, + "loss": 3.2394, + "step": 9654 + }, + { + "epoch": 0.7740875107734863, + "grad_norm": 1.8217945098876953, + "learning_rate": 1.4026402188211814e-05, + "loss": 3.1836, + "step": 9655 + }, + { + "epoch": 0.7741676855545089, + "grad_norm": 1.7587898969650269, + "learning_rate": 1.4025213609956763e-05, + "loss": 3.218, + "step": 9656 + }, + { + "epoch": 0.7742478603355315, + "grad_norm": 1.8336431980133057, + "learning_rate": 1.4024024963838026e-05, + "loss": 3.1945, + "step": 9657 + }, + { + "epoch": 0.7743280351165541, + "grad_norm": 2.1556131839752197, + "learning_rate": 1.402283624987565e-05, + "loss": 3.2003, + "step": 9658 + }, + { + "epoch": 0.7744082098975767, + "grad_norm": 1.7899452447891235, + "learning_rate": 1.402164746808967e-05, + "loss": 3.2155, + "step": 9659 + }, + { + "epoch": 0.7744883846785994, + "grad_norm": 0.7778723239898682, + "learning_rate": 1.402045861850013e-05, + "loss": 2.4941, + "step": 9660 + }, + { + "epoch": 0.774568559459622, + "grad_norm": 1.7337327003479004, + "learning_rate": 1.4019269701127077e-05, + "loss": 3.114, + "step": 9661 + }, + { + "epoch": 0.7746487342406446, + "grad_norm": 1.74927818775177, + "learning_rate": 1.4018080715990553e-05, + "loss": 3.1516, + "step": 9662 + }, + { + "epoch": 0.7747289090216672, + "grad_norm": 1.774917721748352, + "learning_rate": 1.4016891663110603e-05, + "loss": 3.193, + "step": 9663 + }, + { + "epoch": 0.7748090838026899, + "grad_norm": 1.7403614521026611, + "learning_rate": 1.4015702542507277e-05, + "loss": 3.1611, + "step": 9664 + }, + { + "epoch": 0.7748892585837125, + "grad_norm": 1.6775404214859009, + "learning_rate": 1.4014513354200618e-05, + "loss": 3.2769, + "step": 9665 + }, + { + "epoch": 0.7749694333647351, + "grad_norm": 1.8964953422546387, + "learning_rate": 1.4013324098210683e-05, + "loss": 3.2586, + "step": 9666 + }, + { + "epoch": 0.7750496081457577, + "grad_norm": 1.9078806638717651, + "learning_rate": 1.4012134774557516e-05, + "loss": 3.1992, + "step": 9667 + }, + { + "epoch": 0.7751297829267804, + "grad_norm": 1.7868530750274658, + "learning_rate": 1.4010945383261172e-05, + "loss": 3.1863, + "step": 9668 + }, + { + "epoch": 0.775209957707803, + "grad_norm": 1.8146756887435913, + "learning_rate": 1.4009755924341701e-05, + "loss": 3.2184, + "step": 9669 + }, + { + "epoch": 0.7752901324888256, + "grad_norm": 1.778231143951416, + "learning_rate": 1.4008566397819156e-05, + "loss": 3.1575, + "step": 9670 + }, + { + "epoch": 0.7753703072698482, + "grad_norm": 1.8294384479522705, + "learning_rate": 1.4007376803713599e-05, + "loss": 3.2542, + "step": 9671 + }, + { + "epoch": 0.7754504820508709, + "grad_norm": 1.7787259817123413, + "learning_rate": 1.4006187142045078e-05, + "loss": 3.2695, + "step": 9672 + }, + { + "epoch": 0.7755306568318935, + "grad_norm": 1.8022807836532593, + "learning_rate": 1.4004997412833656e-05, + "loss": 3.2977, + "step": 9673 + }, + { + "epoch": 0.7756108316129161, + "grad_norm": 1.7821439504623413, + "learning_rate": 1.400380761609939e-05, + "loss": 3.217, + "step": 9674 + }, + { + "epoch": 0.7756910063939387, + "grad_norm": 1.7592799663543701, + "learning_rate": 1.4002617751862335e-05, + "loss": 3.3849, + "step": 9675 + }, + { + "epoch": 0.7757711811749615, + "grad_norm": 1.8292202949523926, + "learning_rate": 1.400142782014256e-05, + "loss": 3.1389, + "step": 9676 + }, + { + "epoch": 0.7758513559559841, + "grad_norm": 1.755743145942688, + "learning_rate": 1.4000237820960116e-05, + "loss": 3.1217, + "step": 9677 + }, + { + "epoch": 0.7759315307370067, + "grad_norm": 1.701352834701538, + "learning_rate": 1.3999047754335077e-05, + "loss": 3.3754, + "step": 9678 + }, + { + "epoch": 0.7760117055180293, + "grad_norm": 1.7192213535308838, + "learning_rate": 1.39978576202875e-05, + "loss": 3.2888, + "step": 9679 + }, + { + "epoch": 0.776091880299052, + "grad_norm": 1.7913384437561035, + "learning_rate": 1.3996667418837453e-05, + "loss": 3.1262, + "step": 9680 + }, + { + "epoch": 0.7761720550800746, + "grad_norm": 1.864614486694336, + "learning_rate": 1.3995477150005004e-05, + "loss": 3.1524, + "step": 9681 + }, + { + "epoch": 0.7762522298610972, + "grad_norm": 0.7291737794876099, + "learning_rate": 1.3994286813810214e-05, + "loss": 2.4832, + "step": 9682 + }, + { + "epoch": 0.7763324046421198, + "grad_norm": 1.863694429397583, + "learning_rate": 1.3993096410273162e-05, + "loss": 3.2252, + "step": 9683 + }, + { + "epoch": 0.7764125794231425, + "grad_norm": 1.8201225996017456, + "learning_rate": 1.3991905939413907e-05, + "loss": 3.2222, + "step": 9684 + }, + { + "epoch": 0.7764927542041651, + "grad_norm": 1.8001959323883057, + "learning_rate": 1.3990715401252523e-05, + "loss": 3.13, + "step": 9685 + }, + { + "epoch": 0.7765729289851877, + "grad_norm": 1.695218563079834, + "learning_rate": 1.398952479580909e-05, + "loss": 3.1061, + "step": 9686 + }, + { + "epoch": 0.7766531037662103, + "grad_norm": 1.6472891569137573, + "learning_rate": 1.3988334123103672e-05, + "loss": 3.2748, + "step": 9687 + }, + { + "epoch": 0.776733278547233, + "grad_norm": 1.7889161109924316, + "learning_rate": 1.3987143383156348e-05, + "loss": 3.1554, + "step": 9688 + }, + { + "epoch": 0.7768134533282556, + "grad_norm": 1.8549110889434814, + "learning_rate": 1.3985952575987186e-05, + "loss": 3.3022, + "step": 9689 + }, + { + "epoch": 0.7768936281092782, + "grad_norm": 1.8165003061294556, + "learning_rate": 1.3984761701616274e-05, + "loss": 3.2275, + "step": 9690 + }, + { + "epoch": 0.7769738028903008, + "grad_norm": 1.8348159790039062, + "learning_rate": 1.3983570760063683e-05, + "loss": 3.1565, + "step": 9691 + }, + { + "epoch": 0.7770539776713234, + "grad_norm": 1.8555960655212402, + "learning_rate": 1.3982379751349493e-05, + "loss": 3.1526, + "step": 9692 + }, + { + "epoch": 0.7771341524523461, + "grad_norm": 1.740315318107605, + "learning_rate": 1.3981188675493784e-05, + "loss": 3.2646, + "step": 9693 + }, + { + "epoch": 0.7772143272333687, + "grad_norm": 1.8213481903076172, + "learning_rate": 1.3979997532516636e-05, + "loss": 3.1798, + "step": 9694 + }, + { + "epoch": 0.7772945020143913, + "grad_norm": 1.8099534511566162, + "learning_rate": 1.3978806322438132e-05, + "loss": 3.4374, + "step": 9695 + }, + { + "epoch": 0.777374676795414, + "grad_norm": 1.9109305143356323, + "learning_rate": 1.3977615045278357e-05, + "loss": 3.309, + "step": 9696 + }, + { + "epoch": 0.7774548515764367, + "grad_norm": 1.7650221586227417, + "learning_rate": 1.3976423701057393e-05, + "loss": 3.1226, + "step": 9697 + }, + { + "epoch": 0.7775350263574593, + "grad_norm": 1.8775901794433594, + "learning_rate": 1.3975232289795329e-05, + "loss": 3.3567, + "step": 9698 + }, + { + "epoch": 0.7776152011384819, + "grad_norm": 0.7663770914077759, + "learning_rate": 1.3974040811512246e-05, + "loss": 2.4912, + "step": 9699 + }, + { + "epoch": 0.7776953759195046, + "grad_norm": 1.8244024515151978, + "learning_rate": 1.3972849266228237e-05, + "loss": 3.2088, + "step": 9700 + }, + { + "epoch": 0.7777755507005272, + "grad_norm": 1.7030513286590576, + "learning_rate": 1.397165765396339e-05, + "loss": 3.2126, + "step": 9701 + }, + { + "epoch": 0.7778557254815498, + "grad_norm": 1.8669586181640625, + "learning_rate": 1.3970465974737794e-05, + "loss": 3.2849, + "step": 9702 + }, + { + "epoch": 0.7779359002625724, + "grad_norm": 1.8386220932006836, + "learning_rate": 1.3969274228571545e-05, + "loss": 3.2285, + "step": 9703 + }, + { + "epoch": 0.778016075043595, + "grad_norm": 1.7672314643859863, + "learning_rate": 1.3968082415484728e-05, + "loss": 3.1299, + "step": 9704 + }, + { + "epoch": 0.7780962498246177, + "grad_norm": 1.8627763986587524, + "learning_rate": 1.3966890535497437e-05, + "loss": 3.1924, + "step": 9705 + }, + { + "epoch": 0.7781764246056403, + "grad_norm": 1.7779381275177002, + "learning_rate": 1.3965698588629775e-05, + "loss": 3.2189, + "step": 9706 + }, + { + "epoch": 0.7782565993866629, + "grad_norm": 1.832678198814392, + "learning_rate": 1.396450657490183e-05, + "loss": 3.1484, + "step": 9707 + }, + { + "epoch": 0.7783367741676855, + "grad_norm": 1.8340399265289307, + "learning_rate": 1.3963314494333702e-05, + "loss": 3.2812, + "step": 9708 + }, + { + "epoch": 0.7784169489487082, + "grad_norm": 1.825879693031311, + "learning_rate": 1.3962122346945486e-05, + "loss": 3.3052, + "step": 9709 + }, + { + "epoch": 0.7784971237297308, + "grad_norm": 1.7117496728897095, + "learning_rate": 1.3960930132757287e-05, + "loss": 3.248, + "step": 9710 + }, + { + "epoch": 0.7785772985107534, + "grad_norm": 1.7994651794433594, + "learning_rate": 1.39597378517892e-05, + "loss": 3.2859, + "step": 9711 + }, + { + "epoch": 0.778657473291776, + "grad_norm": 1.779293179512024, + "learning_rate": 1.395854550406133e-05, + "loss": 3.2897, + "step": 9712 + }, + { + "epoch": 0.7787376480727987, + "grad_norm": 1.824639081954956, + "learning_rate": 1.3957353089593778e-05, + "loss": 3.1773, + "step": 9713 + }, + { + "epoch": 0.7788178228538213, + "grad_norm": 1.6627216339111328, + "learning_rate": 1.3956160608406643e-05, + "loss": 3.189, + "step": 9714 + }, + { + "epoch": 0.7788979976348439, + "grad_norm": 1.7450778484344482, + "learning_rate": 1.3954968060520038e-05, + "loss": 3.1457, + "step": 9715 + }, + { + "epoch": 0.7789781724158666, + "grad_norm": 1.7588974237442017, + "learning_rate": 1.3953775445954068e-05, + "loss": 3.2106, + "step": 9716 + }, + { + "epoch": 0.7790583471968893, + "grad_norm": 1.7073898315429688, + "learning_rate": 1.3952582764728833e-05, + "loss": 3.1311, + "step": 9717 + }, + { + "epoch": 0.7791385219779119, + "grad_norm": 1.7816052436828613, + "learning_rate": 1.3951390016864447e-05, + "loss": 3.2511, + "step": 9718 + }, + { + "epoch": 0.7792186967589345, + "grad_norm": 1.7672683000564575, + "learning_rate": 1.3950197202381019e-05, + "loss": 3.3031, + "step": 9719 + }, + { + "epoch": 0.7792988715399571, + "grad_norm": 1.7841812372207642, + "learning_rate": 1.3949004321298656e-05, + "loss": 3.1848, + "step": 9720 + }, + { + "epoch": 0.7793790463209798, + "grad_norm": 1.6342921257019043, + "learning_rate": 1.3947811373637477e-05, + "loss": 3.1124, + "step": 9721 + }, + { + "epoch": 0.7794592211020024, + "grad_norm": 1.9377162456512451, + "learning_rate": 1.3946618359417583e-05, + "loss": 3.1834, + "step": 9722 + }, + { + "epoch": 0.779539395883025, + "grad_norm": 1.6825429201126099, + "learning_rate": 1.39454252786591e-05, + "loss": 3.2346, + "step": 9723 + }, + { + "epoch": 0.7796195706640476, + "grad_norm": 1.7760393619537354, + "learning_rate": 1.3944232131382132e-05, + "loss": 3.2063, + "step": 9724 + }, + { + "epoch": 0.7796997454450703, + "grad_norm": 1.8552170991897583, + "learning_rate": 1.3943038917606803e-05, + "loss": 3.2073, + "step": 9725 + }, + { + "epoch": 0.7797799202260929, + "grad_norm": 1.810238003730774, + "learning_rate": 1.3941845637353228e-05, + "loss": 3.1929, + "step": 9726 + }, + { + "epoch": 0.7798600950071155, + "grad_norm": 1.7666586637496948, + "learning_rate": 1.3940652290641525e-05, + "loss": 3.2219, + "step": 9727 + }, + { + "epoch": 0.7799402697881381, + "grad_norm": 1.8127970695495605, + "learning_rate": 1.3939458877491814e-05, + "loss": 3.2407, + "step": 9728 + }, + { + "epoch": 0.7800204445691608, + "grad_norm": 1.7903292179107666, + "learning_rate": 1.3938265397924213e-05, + "loss": 3.2073, + "step": 9729 + }, + { + "epoch": 0.7801006193501834, + "grad_norm": 1.7231172323226929, + "learning_rate": 1.3937071851958843e-05, + "loss": 3.1198, + "step": 9730 + }, + { + "epoch": 0.780180794131206, + "grad_norm": 1.8123126029968262, + "learning_rate": 1.3935878239615834e-05, + "loss": 3.1687, + "step": 9731 + }, + { + "epoch": 0.7802609689122286, + "grad_norm": 1.7866551876068115, + "learning_rate": 1.3934684560915301e-05, + "loss": 3.164, + "step": 9732 + }, + { + "epoch": 0.7803411436932512, + "grad_norm": 1.8030292987823486, + "learning_rate": 1.3933490815877377e-05, + "loss": 3.2496, + "step": 9733 + }, + { + "epoch": 0.7804213184742739, + "grad_norm": 1.730064868927002, + "learning_rate": 1.3932297004522182e-05, + "loss": 3.1613, + "step": 9734 + }, + { + "epoch": 0.7805014932552965, + "grad_norm": 1.8443686962127686, + "learning_rate": 1.3931103126869844e-05, + "loss": 3.2657, + "step": 9735 + }, + { + "epoch": 0.7805816680363191, + "grad_norm": 1.702126383781433, + "learning_rate": 1.3929909182940495e-05, + "loss": 3.2983, + "step": 9736 + }, + { + "epoch": 0.7806618428173419, + "grad_norm": 1.806159496307373, + "learning_rate": 1.392871517275426e-05, + "loss": 3.1489, + "step": 9737 + }, + { + "epoch": 0.7807420175983645, + "grad_norm": 1.7788752317428589, + "learning_rate": 1.3927521096331276e-05, + "loss": 3.1937, + "step": 9738 + }, + { + "epoch": 0.7808221923793871, + "grad_norm": 1.7984042167663574, + "learning_rate": 1.3926326953691667e-05, + "loss": 3.3171, + "step": 9739 + }, + { + "epoch": 0.7809023671604097, + "grad_norm": 1.7333781719207764, + "learning_rate": 1.3925132744855567e-05, + "loss": 3.2628, + "step": 9740 + }, + { + "epoch": 0.7809825419414324, + "grad_norm": 1.7614219188690186, + "learning_rate": 1.3923938469843117e-05, + "loss": 3.1829, + "step": 9741 + }, + { + "epoch": 0.781062716722455, + "grad_norm": 1.8019999265670776, + "learning_rate": 1.3922744128674445e-05, + "loss": 3.2912, + "step": 9742 + }, + { + "epoch": 0.7811428915034776, + "grad_norm": 1.82937490940094, + "learning_rate": 1.3921549721369692e-05, + "loss": 3.2043, + "step": 9743 + }, + { + "epoch": 0.7812230662845002, + "grad_norm": 1.8236027956008911, + "learning_rate": 1.3920355247948988e-05, + "loss": 3.0966, + "step": 9744 + }, + { + "epoch": 0.7813032410655228, + "grad_norm": 1.8199610710144043, + "learning_rate": 1.391916070843248e-05, + "loss": 3.1677, + "step": 9745 + }, + { + "epoch": 0.7813834158465455, + "grad_norm": 1.8315004110336304, + "learning_rate": 1.3917966102840303e-05, + "loss": 3.2335, + "step": 9746 + }, + { + "epoch": 0.7814635906275681, + "grad_norm": 1.7489889860153198, + "learning_rate": 1.3916771431192597e-05, + "loss": 3.2054, + "step": 9747 + }, + { + "epoch": 0.7815437654085907, + "grad_norm": 1.7172904014587402, + "learning_rate": 1.391557669350951e-05, + "loss": 3.189, + "step": 9748 + }, + { + "epoch": 0.7816239401896133, + "grad_norm": 1.7801278829574585, + "learning_rate": 1.3914381889811172e-05, + "loss": 3.1821, + "step": 9749 + }, + { + "epoch": 0.781704114970636, + "grad_norm": 1.7996364831924438, + "learning_rate": 1.391318702011774e-05, + "loss": 3.2118, + "step": 9750 + }, + { + "epoch": 0.7817842897516586, + "grad_norm": 1.7681293487548828, + "learning_rate": 1.3911992084449356e-05, + "loss": 3.0865, + "step": 9751 + }, + { + "epoch": 0.7818644645326812, + "grad_norm": 0.7825413346290588, + "learning_rate": 1.3910797082826157e-05, + "loss": 2.4941, + "step": 9752 + }, + { + "epoch": 0.7819446393137038, + "grad_norm": 1.9317806959152222, + "learning_rate": 1.3909602015268304e-05, + "loss": 3.2052, + "step": 9753 + }, + { + "epoch": 0.7820248140947265, + "grad_norm": 1.9479056596755981, + "learning_rate": 1.3908406881795935e-05, + "loss": 3.2358, + "step": 9754 + }, + { + "epoch": 0.7821049888757491, + "grad_norm": 1.751094102859497, + "learning_rate": 1.3907211682429205e-05, + "loss": 3.2155, + "step": 9755 + }, + { + "epoch": 0.7821851636567717, + "grad_norm": 1.8882560729980469, + "learning_rate": 1.3906016417188264e-05, + "loss": 3.1196, + "step": 9756 + }, + { + "epoch": 0.7822653384377944, + "grad_norm": 1.8123010396957397, + "learning_rate": 1.390482108609326e-05, + "loss": 3.2473, + "step": 9757 + }, + { + "epoch": 0.7823455132188171, + "grad_norm": 1.8358410596847534, + "learning_rate": 1.3903625689164353e-05, + "loss": 3.3137, + "step": 9758 + }, + { + "epoch": 0.7824256879998397, + "grad_norm": 1.7814302444458008, + "learning_rate": 1.390243022642169e-05, + "loss": 3.229, + "step": 9759 + }, + { + "epoch": 0.7825058627808623, + "grad_norm": 1.7792649269104004, + "learning_rate": 1.3901234697885429e-05, + "loss": 3.2665, + "step": 9760 + }, + { + "epoch": 0.7825860375618849, + "grad_norm": 1.7714569568634033, + "learning_rate": 1.3900039103575727e-05, + "loss": 3.2356, + "step": 9761 + }, + { + "epoch": 0.7826662123429076, + "grad_norm": 1.8082740306854248, + "learning_rate": 1.3898843443512737e-05, + "loss": 3.3281, + "step": 9762 + }, + { + "epoch": 0.7827463871239302, + "grad_norm": 1.931356430053711, + "learning_rate": 1.3897647717716623e-05, + "loss": 3.2928, + "step": 9763 + }, + { + "epoch": 0.7828265619049528, + "grad_norm": 1.75770103931427, + "learning_rate": 1.3896451926207542e-05, + "loss": 3.2551, + "step": 9764 + }, + { + "epoch": 0.7829067366859754, + "grad_norm": 1.7719391584396362, + "learning_rate": 1.3895256069005653e-05, + "loss": 3.1686, + "step": 9765 + }, + { + "epoch": 0.7829869114669981, + "grad_norm": 1.7547266483306885, + "learning_rate": 1.3894060146131124e-05, + "loss": 3.2118, + "step": 9766 + }, + { + "epoch": 0.7830670862480207, + "grad_norm": 1.7973649501800537, + "learning_rate": 1.3892864157604109e-05, + "loss": 3.2067, + "step": 9767 + }, + { + "epoch": 0.7831472610290433, + "grad_norm": 1.9151560068130493, + "learning_rate": 1.3891668103444778e-05, + "loss": 3.2497, + "step": 9768 + }, + { + "epoch": 0.7832274358100659, + "grad_norm": 0.7756634950637817, + "learning_rate": 1.3890471983673293e-05, + "loss": 2.4912, + "step": 9769 + }, + { + "epoch": 0.7833076105910886, + "grad_norm": 1.7884150743484497, + "learning_rate": 1.3889275798309822e-05, + "loss": 3.293, + "step": 9770 + }, + { + "epoch": 0.7833877853721112, + "grad_norm": 2.0809664726257324, + "learning_rate": 1.3888079547374533e-05, + "loss": 3.1848, + "step": 9771 + }, + { + "epoch": 0.7834679601531338, + "grad_norm": 1.7439221143722534, + "learning_rate": 1.3886883230887591e-05, + "loss": 3.1974, + "step": 9772 + }, + { + "epoch": 0.7835481349341564, + "grad_norm": 1.9017901420593262, + "learning_rate": 1.388568684886917e-05, + "loss": 3.1906, + "step": 9773 + }, + { + "epoch": 0.783628309715179, + "grad_norm": 1.7437093257904053, + "learning_rate": 1.3884490401339438e-05, + "loss": 3.2678, + "step": 9774 + }, + { + "epoch": 0.7837084844962017, + "grad_norm": 1.6893459558486938, + "learning_rate": 1.3883293888318565e-05, + "loss": 3.2773, + "step": 9775 + }, + { + "epoch": 0.7837886592772243, + "grad_norm": 1.748213768005371, + "learning_rate": 1.3882097309826728e-05, + "loss": 3.2986, + "step": 9776 + }, + { + "epoch": 0.783868834058247, + "grad_norm": 1.891621470451355, + "learning_rate": 1.3880900665884095e-05, + "loss": 3.2821, + "step": 9777 + }, + { + "epoch": 0.7839490088392697, + "grad_norm": 1.6767499446868896, + "learning_rate": 1.3879703956510848e-05, + "loss": 3.1371, + "step": 9778 + }, + { + "epoch": 0.7840291836202923, + "grad_norm": 1.7961057424545288, + "learning_rate": 1.3878507181727158e-05, + "loss": 3.1355, + "step": 9779 + }, + { + "epoch": 0.7841093584013149, + "grad_norm": 1.953253984451294, + "learning_rate": 1.3877310341553206e-05, + "loss": 3.142, + "step": 9780 + }, + { + "epoch": 0.7841895331823375, + "grad_norm": 1.6911283731460571, + "learning_rate": 1.3876113436009164e-05, + "loss": 3.1458, + "step": 9781 + }, + { + "epoch": 0.7842697079633602, + "grad_norm": 1.687516689300537, + "learning_rate": 1.3874916465115217e-05, + "loss": 3.0649, + "step": 9782 + }, + { + "epoch": 0.7843498827443828, + "grad_norm": 1.7655086517333984, + "learning_rate": 1.3873719428891547e-05, + "loss": 3.0939, + "step": 9783 + }, + { + "epoch": 0.7844300575254054, + "grad_norm": 0.7357281446456909, + "learning_rate": 1.387252232735833e-05, + "loss": 2.4978, + "step": 9784 + }, + { + "epoch": 0.784510232306428, + "grad_norm": 1.77824068069458, + "learning_rate": 1.3871325160535751e-05, + "loss": 3.0155, + "step": 9785 + }, + { + "epoch": 0.7845904070874506, + "grad_norm": 1.7921795845031738, + "learning_rate": 1.3870127928443998e-05, + "loss": 3.167, + "step": 9786 + }, + { + "epoch": 0.7846705818684733, + "grad_norm": 1.712807059288025, + "learning_rate": 1.3868930631103248e-05, + "loss": 3.0858, + "step": 9787 + }, + { + "epoch": 0.7847507566494959, + "grad_norm": 2.0232489109039307, + "learning_rate": 1.3867733268533694e-05, + "loss": 3.3117, + "step": 9788 + }, + { + "epoch": 0.7848309314305185, + "grad_norm": 1.793498158454895, + "learning_rate": 1.3866535840755518e-05, + "loss": 3.2334, + "step": 9789 + }, + { + "epoch": 0.7849111062115411, + "grad_norm": 1.704016089439392, + "learning_rate": 1.386533834778891e-05, + "loss": 3.1299, + "step": 9790 + }, + { + "epoch": 0.7849912809925638, + "grad_norm": 1.8916964530944824, + "learning_rate": 1.386414078965406e-05, + "loss": 3.2903, + "step": 9791 + }, + { + "epoch": 0.7850714557735864, + "grad_norm": 1.7650394439697266, + "learning_rate": 1.3862943166371159e-05, + "loss": 3.3132, + "step": 9792 + }, + { + "epoch": 0.785151630554609, + "grad_norm": 1.7464452981948853, + "learning_rate": 1.3861745477960398e-05, + "loss": 3.1393, + "step": 9793 + }, + { + "epoch": 0.7852318053356316, + "grad_norm": 1.8049557209014893, + "learning_rate": 1.3860547724441968e-05, + "loss": 3.3264, + "step": 9794 + }, + { + "epoch": 0.7853119801166543, + "grad_norm": 0.8095198273658752, + "learning_rate": 1.3859349905836065e-05, + "loss": 2.4948, + "step": 9795 + }, + { + "epoch": 0.7853921548976769, + "grad_norm": 1.766573429107666, + "learning_rate": 1.3858152022162882e-05, + "loss": 3.2745, + "step": 9796 + }, + { + "epoch": 0.7854723296786995, + "grad_norm": 1.70841383934021, + "learning_rate": 1.3856954073442617e-05, + "loss": 3.2051, + "step": 9797 + }, + { + "epoch": 0.7855525044597222, + "grad_norm": 1.8957467079162598, + "learning_rate": 1.3855756059695467e-05, + "loss": 3.2002, + "step": 9798 + }, + { + "epoch": 0.7856326792407449, + "grad_norm": 1.7658389806747437, + "learning_rate": 1.3854557980941626e-05, + "loss": 3.237, + "step": 9799 + }, + { + "epoch": 0.7857128540217675, + "grad_norm": 1.782652497291565, + "learning_rate": 1.3853359837201295e-05, + "loss": 3.1465, + "step": 9800 + }, + { + "epoch": 0.7857930288027901, + "grad_norm": 1.817663550376892, + "learning_rate": 1.3852161628494675e-05, + "loss": 3.1962, + "step": 9801 + }, + { + "epoch": 0.7858732035838127, + "grad_norm": 1.759389042854309, + "learning_rate": 1.385096335484197e-05, + "loss": 3.2054, + "step": 9802 + }, + { + "epoch": 0.7859533783648354, + "grad_norm": 1.8249616622924805, + "learning_rate": 1.3849765016263382e-05, + "loss": 3.1992, + "step": 9803 + }, + { + "epoch": 0.786033553145858, + "grad_norm": 1.7036184072494507, + "learning_rate": 1.3848566612779108e-05, + "loss": 3.1851, + "step": 9804 + }, + { + "epoch": 0.7861137279268806, + "grad_norm": 1.8958245515823364, + "learning_rate": 1.384736814440936e-05, + "loss": 3.2499, + "step": 9805 + }, + { + "epoch": 0.7861939027079032, + "grad_norm": 1.8861244916915894, + "learning_rate": 1.3846169611174341e-05, + "loss": 3.0842, + "step": 9806 + }, + { + "epoch": 0.7862740774889259, + "grad_norm": 1.7353007793426514, + "learning_rate": 1.3844971013094255e-05, + "loss": 3.161, + "step": 9807 + }, + { + "epoch": 0.7863542522699485, + "grad_norm": 1.8060604333877563, + "learning_rate": 1.3843772350189317e-05, + "loss": 3.2523, + "step": 9808 + }, + { + "epoch": 0.7864344270509711, + "grad_norm": 1.711040735244751, + "learning_rate": 1.3842573622479729e-05, + "loss": 3.1633, + "step": 9809 + }, + { + "epoch": 0.7865146018319937, + "grad_norm": 1.6531105041503906, + "learning_rate": 1.3841374829985703e-05, + "loss": 3.2229, + "step": 9810 + }, + { + "epoch": 0.7865947766130164, + "grad_norm": 1.7059961557388306, + "learning_rate": 1.3840175972727454e-05, + "loss": 3.1736, + "step": 9811 + }, + { + "epoch": 0.786674951394039, + "grad_norm": 2.0567991733551025, + "learning_rate": 1.383897705072519e-05, + "loss": 3.1357, + "step": 9812 + }, + { + "epoch": 0.7867551261750616, + "grad_norm": 1.746943712234497, + "learning_rate": 1.3837778063999127e-05, + "loss": 3.1948, + "step": 9813 + }, + { + "epoch": 0.7868353009560842, + "grad_norm": 1.739902377128601, + "learning_rate": 1.3836579012569478e-05, + "loss": 3.191, + "step": 9814 + }, + { + "epoch": 0.7869154757371069, + "grad_norm": 1.736441969871521, + "learning_rate": 1.3835379896456458e-05, + "loss": 3.0859, + "step": 9815 + }, + { + "epoch": 0.7869956505181295, + "grad_norm": 1.8555123805999756, + "learning_rate": 1.3834180715680286e-05, + "loss": 3.2119, + "step": 9816 + }, + { + "epoch": 0.7870758252991521, + "grad_norm": 1.7619494199752808, + "learning_rate": 1.3832981470261175e-05, + "loss": 3.1822, + "step": 9817 + }, + { + "epoch": 0.7871560000801748, + "grad_norm": 0.7186050415039062, + "learning_rate": 1.3831782160219351e-05, + "loss": 2.4642, + "step": 9818 + }, + { + "epoch": 0.7872361748611975, + "grad_norm": 1.7688465118408203, + "learning_rate": 1.383058278557503e-05, + "loss": 3.2568, + "step": 9819 + }, + { + "epoch": 0.7873163496422201, + "grad_norm": 1.7354347705841064, + "learning_rate": 1.3829383346348433e-05, + "loss": 3.1382, + "step": 9820 + }, + { + "epoch": 0.7873965244232427, + "grad_norm": 1.8372257947921753, + "learning_rate": 1.382818384255978e-05, + "loss": 3.2639, + "step": 9821 + }, + { + "epoch": 0.7874766992042653, + "grad_norm": 1.7938830852508545, + "learning_rate": 1.38269842742293e-05, + "loss": 3.2079, + "step": 9822 + }, + { + "epoch": 0.787556873985288, + "grad_norm": 1.8869750499725342, + "learning_rate": 1.382578464137721e-05, + "loss": 3.2629, + "step": 9823 + }, + { + "epoch": 0.7876370487663106, + "grad_norm": 1.7631856203079224, + "learning_rate": 1.3824584944023742e-05, + "loss": 3.2644, + "step": 9824 + }, + { + "epoch": 0.7877172235473332, + "grad_norm": 1.773496389389038, + "learning_rate": 1.3823385182189122e-05, + "loss": 3.1941, + "step": 9825 + }, + { + "epoch": 0.7877973983283558, + "grad_norm": 1.8609365224838257, + "learning_rate": 1.3822185355893571e-05, + "loss": 3.1938, + "step": 9826 + }, + { + "epoch": 0.7878775731093784, + "grad_norm": 1.7249784469604492, + "learning_rate": 1.3820985465157324e-05, + "loss": 3.2584, + "step": 9827 + }, + { + "epoch": 0.7879577478904011, + "grad_norm": 1.8028671741485596, + "learning_rate": 1.381978551000061e-05, + "loss": 3.1649, + "step": 9828 + }, + { + "epoch": 0.7880379226714237, + "grad_norm": 0.7963016629219055, + "learning_rate": 1.3818585490443654e-05, + "loss": 2.5198, + "step": 9829 + }, + { + "epoch": 0.7881180974524463, + "grad_norm": 1.875158667564392, + "learning_rate": 1.3817385406506697e-05, + "loss": 3.2014, + "step": 9830 + }, + { + "epoch": 0.7881982722334689, + "grad_norm": 1.9734443426132202, + "learning_rate": 1.3816185258209966e-05, + "loss": 3.1714, + "step": 9831 + }, + { + "epoch": 0.7882784470144916, + "grad_norm": 1.7727055549621582, + "learning_rate": 1.3814985045573696e-05, + "loss": 3.1523, + "step": 9832 + }, + { + "epoch": 0.7883586217955142, + "grad_norm": 2.0224218368530273, + "learning_rate": 1.3813784768618124e-05, + "loss": 3.202, + "step": 9833 + }, + { + "epoch": 0.7884387965765368, + "grad_norm": 1.8261874914169312, + "learning_rate": 1.3812584427363484e-05, + "loss": 3.2246, + "step": 9834 + }, + { + "epoch": 0.7885189713575594, + "grad_norm": 1.840253472328186, + "learning_rate": 1.3811384021830012e-05, + "loss": 3.137, + "step": 9835 + }, + { + "epoch": 0.7885991461385821, + "grad_norm": 1.7539043426513672, + "learning_rate": 1.3810183552037952e-05, + "loss": 3.2081, + "step": 9836 + }, + { + "epoch": 0.7886793209196047, + "grad_norm": 1.757790446281433, + "learning_rate": 1.380898301800754e-05, + "loss": 3.1327, + "step": 9837 + }, + { + "epoch": 0.7887594957006274, + "grad_norm": 1.6873741149902344, + "learning_rate": 1.3807782419759014e-05, + "loss": 3.1772, + "step": 9838 + }, + { + "epoch": 0.78883967048165, + "grad_norm": 1.7763397693634033, + "learning_rate": 1.3806581757312621e-05, + "loss": 3.1453, + "step": 9839 + }, + { + "epoch": 0.7889198452626727, + "grad_norm": 1.7880421876907349, + "learning_rate": 1.38053810306886e-05, + "loss": 3.2543, + "step": 9840 + }, + { + "epoch": 0.7890000200436953, + "grad_norm": 1.836137056350708, + "learning_rate": 1.3804180239907196e-05, + "loss": 3.3026, + "step": 9841 + }, + { + "epoch": 0.7890801948247179, + "grad_norm": 1.7716385126113892, + "learning_rate": 1.3802979384988656e-05, + "loss": 3.2062, + "step": 9842 + }, + { + "epoch": 0.7891603696057405, + "grad_norm": 1.7762595415115356, + "learning_rate": 1.3801778465953221e-05, + "loss": 3.1381, + "step": 9843 + }, + { + "epoch": 0.7892405443867632, + "grad_norm": 1.7632414102554321, + "learning_rate": 1.3800577482821143e-05, + "loss": 3.2784, + "step": 9844 + }, + { + "epoch": 0.7893207191677858, + "grad_norm": 1.8707711696624756, + "learning_rate": 1.3799376435612666e-05, + "loss": 3.1182, + "step": 9845 + }, + { + "epoch": 0.7894008939488084, + "grad_norm": 0.7497985363006592, + "learning_rate": 1.3798175324348043e-05, + "loss": 2.4939, + "step": 9846 + }, + { + "epoch": 0.789481068729831, + "grad_norm": 1.7912380695343018, + "learning_rate": 1.3796974149047521e-05, + "loss": 3.1846, + "step": 9847 + }, + { + "epoch": 0.7895612435108537, + "grad_norm": 1.8352855443954468, + "learning_rate": 1.3795772909731355e-05, + "loss": 3.2538, + "step": 9848 + }, + { + "epoch": 0.7896414182918763, + "grad_norm": 1.8221794366836548, + "learning_rate": 1.3794571606419793e-05, + "loss": 3.3009, + "step": 9849 + }, + { + "epoch": 0.7897215930728989, + "grad_norm": 1.7518857717514038, + "learning_rate": 1.3793370239133092e-05, + "loss": 3.2312, + "step": 9850 + }, + { + "epoch": 0.7898017678539215, + "grad_norm": 1.75957190990448, + "learning_rate": 1.3792168807891508e-05, + "loss": 3.2375, + "step": 9851 + }, + { + "epoch": 0.7898819426349442, + "grad_norm": 1.8034422397613525, + "learning_rate": 1.3790967312715291e-05, + "loss": 3.2324, + "step": 9852 + }, + { + "epoch": 0.7899621174159668, + "grad_norm": 1.7177777290344238, + "learning_rate": 1.3789765753624704e-05, + "loss": 3.2395, + "step": 9853 + }, + { + "epoch": 0.7900422921969894, + "grad_norm": 1.7805893421173096, + "learning_rate": 1.378856413064e-05, + "loss": 3.2811, + "step": 9854 + }, + { + "epoch": 0.790122466978012, + "grad_norm": 1.8138482570648193, + "learning_rate": 1.3787362443781441e-05, + "loss": 3.215, + "step": 9855 + }, + { + "epoch": 0.7902026417590347, + "grad_norm": 0.7323717474937439, + "learning_rate": 1.3786160693069289e-05, + "loss": 2.5048, + "step": 9856 + }, + { + "epoch": 0.7902828165400573, + "grad_norm": 1.7178462743759155, + "learning_rate": 1.3784958878523795e-05, + "loss": 3.1082, + "step": 9857 + }, + { + "epoch": 0.7903629913210799, + "grad_norm": 1.9508212804794312, + "learning_rate": 1.3783757000165235e-05, + "loss": 3.0995, + "step": 9858 + }, + { + "epoch": 0.7904431661021026, + "grad_norm": 1.8186559677124023, + "learning_rate": 1.378255505801386e-05, + "loss": 3.3141, + "step": 9859 + }, + { + "epoch": 0.7905233408831253, + "grad_norm": 1.867119550704956, + "learning_rate": 1.3781353052089942e-05, + "loss": 3.1995, + "step": 9860 + }, + { + "epoch": 0.7906035156641479, + "grad_norm": 1.833509087562561, + "learning_rate": 1.3780150982413748e-05, + "loss": 3.2975, + "step": 9861 + }, + { + "epoch": 0.7906836904451705, + "grad_norm": 1.7640653848648071, + "learning_rate": 1.3778948849005535e-05, + "loss": 3.0914, + "step": 9862 + }, + { + "epoch": 0.7907638652261931, + "grad_norm": 1.82669997215271, + "learning_rate": 1.377774665188558e-05, + "loss": 3.2109, + "step": 9863 + }, + { + "epoch": 0.7908440400072158, + "grad_norm": 1.8585789203643799, + "learning_rate": 1.3776544391074147e-05, + "loss": 3.2538, + "step": 9864 + }, + { + "epoch": 0.7909242147882384, + "grad_norm": 1.769208550453186, + "learning_rate": 1.3775342066591504e-05, + "loss": 3.249, + "step": 9865 + }, + { + "epoch": 0.791004389569261, + "grad_norm": 1.8186144828796387, + "learning_rate": 1.3774139678457928e-05, + "loss": 3.2068, + "step": 9866 + }, + { + "epoch": 0.7910845643502836, + "grad_norm": 1.678239107131958, + "learning_rate": 1.3772937226693684e-05, + "loss": 3.1842, + "step": 9867 + }, + { + "epoch": 0.7911647391313062, + "grad_norm": 1.7356873750686646, + "learning_rate": 1.3771734711319052e-05, + "loss": 3.2487, + "step": 9868 + }, + { + "epoch": 0.7912449139123289, + "grad_norm": 1.7905993461608887, + "learning_rate": 1.37705321323543e-05, + "loss": 3.2228, + "step": 9869 + }, + { + "epoch": 0.7913250886933515, + "grad_norm": 1.751645565032959, + "learning_rate": 1.3769329489819702e-05, + "loss": 3.2298, + "step": 9870 + }, + { + "epoch": 0.7914052634743741, + "grad_norm": 1.7271887063980103, + "learning_rate": 1.3768126783735541e-05, + "loss": 3.1322, + "step": 9871 + }, + { + "epoch": 0.7914854382553967, + "grad_norm": 1.6893837451934814, + "learning_rate": 1.376692401412209e-05, + "loss": 3.223, + "step": 9872 + }, + { + "epoch": 0.7915656130364194, + "grad_norm": 1.9700987339019775, + "learning_rate": 1.3765721180999628e-05, + "loss": 3.1472, + "step": 9873 + }, + { + "epoch": 0.791645787817442, + "grad_norm": 1.7716673612594604, + "learning_rate": 1.3764518284388433e-05, + "loss": 3.23, + "step": 9874 + }, + { + "epoch": 0.7917259625984646, + "grad_norm": 1.7909824848175049, + "learning_rate": 1.3763315324308785e-05, + "loss": 3.1351, + "step": 9875 + }, + { + "epoch": 0.7918061373794872, + "grad_norm": 1.7869293689727783, + "learning_rate": 1.376211230078097e-05, + "loss": 3.2761, + "step": 9876 + }, + { + "epoch": 0.7918863121605099, + "grad_norm": 1.7383886575698853, + "learning_rate": 1.3760909213825263e-05, + "loss": 3.1053, + "step": 9877 + }, + { + "epoch": 0.7919664869415325, + "grad_norm": 1.8186992406845093, + "learning_rate": 1.3759706063461958e-05, + "loss": 3.3577, + "step": 9878 + }, + { + "epoch": 0.7920466617225552, + "grad_norm": 1.8910733461380005, + "learning_rate": 1.375850284971133e-05, + "loss": 3.2029, + "step": 9879 + }, + { + "epoch": 0.7921268365035778, + "grad_norm": 1.7593539953231812, + "learning_rate": 1.375729957259367e-05, + "loss": 3.1274, + "step": 9880 + }, + { + "epoch": 0.7922070112846005, + "grad_norm": 1.7455497980117798, + "learning_rate": 1.3756096232129263e-05, + "loss": 3.2118, + "step": 9881 + }, + { + "epoch": 0.7922871860656231, + "grad_norm": 1.8340774774551392, + "learning_rate": 1.3754892828338397e-05, + "loss": 3.2278, + "step": 9882 + }, + { + "epoch": 0.7923673608466457, + "grad_norm": 1.7605172395706177, + "learning_rate": 1.3753689361241365e-05, + "loss": 3.2086, + "step": 9883 + }, + { + "epoch": 0.7924475356276683, + "grad_norm": 1.920061469078064, + "learning_rate": 1.3752485830858446e-05, + "loss": 3.2341, + "step": 9884 + }, + { + "epoch": 0.792527710408691, + "grad_norm": 1.8340198993682861, + "learning_rate": 1.3751282237209945e-05, + "loss": 3.2694, + "step": 9885 + }, + { + "epoch": 0.7926078851897136, + "grad_norm": 1.826625943183899, + "learning_rate": 1.3750078580316144e-05, + "loss": 3.2804, + "step": 9886 + }, + { + "epoch": 0.7926880599707362, + "grad_norm": 1.9209401607513428, + "learning_rate": 1.374887486019734e-05, + "loss": 3.2219, + "step": 9887 + }, + { + "epoch": 0.7927682347517588, + "grad_norm": 1.750499963760376, + "learning_rate": 1.374767107687383e-05, + "loss": 3.2629, + "step": 9888 + }, + { + "epoch": 0.7928484095327815, + "grad_norm": 1.7475683689117432, + "learning_rate": 1.3746467230365903e-05, + "loss": 3.1972, + "step": 9889 + }, + { + "epoch": 0.7929285843138041, + "grad_norm": 1.8264219760894775, + "learning_rate": 1.3745263320693857e-05, + "loss": 3.2141, + "step": 9890 + }, + { + "epoch": 0.7930087590948267, + "grad_norm": 0.740870475769043, + "learning_rate": 1.3744059347877999e-05, + "loss": 2.4773, + "step": 9891 + }, + { + "epoch": 0.7930889338758493, + "grad_norm": 0.7956336736679077, + "learning_rate": 1.3742855311938615e-05, + "loss": 2.4733, + "step": 9892 + }, + { + "epoch": 0.793169108656872, + "grad_norm": 1.8639343976974487, + "learning_rate": 1.374165121289601e-05, + "loss": 3.2156, + "step": 9893 + }, + { + "epoch": 0.7932492834378946, + "grad_norm": 1.8529736995697021, + "learning_rate": 1.3740447050770482e-05, + "loss": 3.1391, + "step": 9894 + }, + { + "epoch": 0.7933294582189172, + "grad_norm": 1.6451971530914307, + "learning_rate": 1.3739242825582338e-05, + "loss": 3.2985, + "step": 9895 + }, + { + "epoch": 0.7934096329999398, + "grad_norm": 1.8393861055374146, + "learning_rate": 1.3738038537351876e-05, + "loss": 3.1977, + "step": 9896 + }, + { + "epoch": 0.7934898077809625, + "grad_norm": 1.8121050596237183, + "learning_rate": 1.3736834186099403e-05, + "loss": 3.1163, + "step": 9897 + }, + { + "epoch": 0.7935699825619851, + "grad_norm": 1.9143255949020386, + "learning_rate": 1.3735629771845224e-05, + "loss": 3.1455, + "step": 9898 + }, + { + "epoch": 0.7936501573430078, + "grad_norm": 2.0916731357574463, + "learning_rate": 1.3734425294609642e-05, + "loss": 3.182, + "step": 9899 + }, + { + "epoch": 0.7937303321240304, + "grad_norm": 1.8265352249145508, + "learning_rate": 1.3733220754412963e-05, + "loss": 3.1932, + "step": 9900 + }, + { + "epoch": 0.7938105069050531, + "grad_norm": 1.7708762884140015, + "learning_rate": 1.3732016151275502e-05, + "loss": 3.1937, + "step": 9901 + }, + { + "epoch": 0.7938906816860757, + "grad_norm": 1.7093430757522583, + "learning_rate": 1.3730811485217561e-05, + "loss": 3.2538, + "step": 9902 + }, + { + "epoch": 0.7939708564670983, + "grad_norm": 1.9788105487823486, + "learning_rate": 1.3729606756259457e-05, + "loss": 3.1881, + "step": 9903 + }, + { + "epoch": 0.7940510312481209, + "grad_norm": 1.75067937374115, + "learning_rate": 1.3728401964421493e-05, + "loss": 3.1722, + "step": 9904 + }, + { + "epoch": 0.7941312060291436, + "grad_norm": 1.7955889701843262, + "learning_rate": 1.3727197109723989e-05, + "loss": 3.1403, + "step": 9905 + }, + { + "epoch": 0.7942113808101662, + "grad_norm": 1.8570032119750977, + "learning_rate": 1.3725992192187254e-05, + "loss": 3.1195, + "step": 9906 + }, + { + "epoch": 0.7942915555911888, + "grad_norm": 1.7702381610870361, + "learning_rate": 1.3724787211831602e-05, + "loss": 3.2657, + "step": 9907 + }, + { + "epoch": 0.7943717303722114, + "grad_norm": 1.7356771230697632, + "learning_rate": 1.3723582168677356e-05, + "loss": 3.1813, + "step": 9908 + }, + { + "epoch": 0.794451905153234, + "grad_norm": 1.8418453931808472, + "learning_rate": 1.3722377062744822e-05, + "loss": 3.2416, + "step": 9909 + }, + { + "epoch": 0.7945320799342567, + "grad_norm": 1.7786935567855835, + "learning_rate": 1.3721171894054324e-05, + "loss": 3.1421, + "step": 9910 + }, + { + "epoch": 0.7946122547152793, + "grad_norm": 1.8395720720291138, + "learning_rate": 1.3719966662626178e-05, + "loss": 3.2146, + "step": 9911 + }, + { + "epoch": 0.7946924294963019, + "grad_norm": 1.6463301181793213, + "learning_rate": 1.3718761368480706e-05, + "loss": 3.2227, + "step": 9912 + }, + { + "epoch": 0.7947726042773245, + "grad_norm": 1.8368912935256958, + "learning_rate": 1.3717556011638231e-05, + "loss": 3.1156, + "step": 9913 + }, + { + "epoch": 0.7948527790583472, + "grad_norm": 1.8419575691223145, + "learning_rate": 1.3716350592119067e-05, + "loss": 3.1976, + "step": 9914 + }, + { + "epoch": 0.7949329538393698, + "grad_norm": 1.7789788246154785, + "learning_rate": 1.3715145109943543e-05, + "loss": 3.2688, + "step": 9915 + }, + { + "epoch": 0.7950131286203924, + "grad_norm": 1.7932569980621338, + "learning_rate": 1.3713939565131986e-05, + "loss": 3.249, + "step": 9916 + }, + { + "epoch": 0.795093303401415, + "grad_norm": 1.8441410064697266, + "learning_rate": 1.3712733957704712e-05, + "loss": 3.1942, + "step": 9917 + }, + { + "epoch": 0.7951734781824377, + "grad_norm": 1.7826740741729736, + "learning_rate": 1.3711528287682056e-05, + "loss": 3.0966, + "step": 9918 + }, + { + "epoch": 0.7952536529634603, + "grad_norm": 1.8424937725067139, + "learning_rate": 1.3710322555084336e-05, + "loss": 3.2059, + "step": 9919 + }, + { + "epoch": 0.795333827744483, + "grad_norm": 1.8813059329986572, + "learning_rate": 1.3709116759931888e-05, + "loss": 3.1642, + "step": 9920 + }, + { + "epoch": 0.7954140025255056, + "grad_norm": 1.8174898624420166, + "learning_rate": 1.370791090224504e-05, + "loss": 3.1437, + "step": 9921 + }, + { + "epoch": 0.7954941773065283, + "grad_norm": 0.7572021484375, + "learning_rate": 1.3706704982044119e-05, + "loss": 2.5205, + "step": 9922 + }, + { + "epoch": 0.7955743520875509, + "grad_norm": 1.8112918138504028, + "learning_rate": 1.370549899934946e-05, + "loss": 3.2101, + "step": 9923 + }, + { + "epoch": 0.7956545268685735, + "grad_norm": 1.7721878290176392, + "learning_rate": 1.3704292954181394e-05, + "loss": 3.2134, + "step": 9924 + }, + { + "epoch": 0.7957347016495961, + "grad_norm": 1.8321388959884644, + "learning_rate": 1.3703086846560251e-05, + "loss": 3.2799, + "step": 9925 + }, + { + "epoch": 0.7958148764306188, + "grad_norm": 1.8580782413482666, + "learning_rate": 1.3701880676506374e-05, + "loss": 3.239, + "step": 9926 + }, + { + "epoch": 0.7958950512116414, + "grad_norm": 1.768011212348938, + "learning_rate": 1.370067444404009e-05, + "loss": 3.2378, + "step": 9927 + }, + { + "epoch": 0.795975225992664, + "grad_norm": 1.8199604749679565, + "learning_rate": 1.369946814918174e-05, + "loss": 3.2048, + "step": 9928 + }, + { + "epoch": 0.7960554007736866, + "grad_norm": 1.7240848541259766, + "learning_rate": 1.3698261791951659e-05, + "loss": 3.1859, + "step": 9929 + }, + { + "epoch": 0.7961355755547093, + "grad_norm": 1.9096522331237793, + "learning_rate": 1.369705537237019e-05, + "loss": 3.2175, + "step": 9930 + }, + { + "epoch": 0.7962157503357319, + "grad_norm": 1.7834731340408325, + "learning_rate": 1.369584889045767e-05, + "loss": 3.2484, + "step": 9931 + }, + { + "epoch": 0.7962959251167545, + "grad_norm": 1.6720894575119019, + "learning_rate": 1.3694642346234439e-05, + "loss": 3.1539, + "step": 9932 + }, + { + "epoch": 0.7963760998977771, + "grad_norm": 1.788570523262024, + "learning_rate": 1.3693435739720841e-05, + "loss": 3.105, + "step": 9933 + }, + { + "epoch": 0.7964562746787998, + "grad_norm": 1.8116344213485718, + "learning_rate": 1.3692229070937217e-05, + "loss": 3.1924, + "step": 9934 + }, + { + "epoch": 0.7965364494598224, + "grad_norm": 1.7420929670333862, + "learning_rate": 1.3691022339903915e-05, + "loss": 3.3035, + "step": 9935 + }, + { + "epoch": 0.796616624240845, + "grad_norm": 0.7243966460227966, + "learning_rate": 1.3689815546641274e-05, + "loss": 2.5081, + "step": 9936 + }, + { + "epoch": 0.7966967990218676, + "grad_norm": 1.7245224714279175, + "learning_rate": 1.3688608691169645e-05, + "loss": 3.2419, + "step": 9937 + }, + { + "epoch": 0.7967769738028903, + "grad_norm": 1.7817822694778442, + "learning_rate": 1.3687401773509374e-05, + "loss": 3.261, + "step": 9938 + }, + { + "epoch": 0.7968571485839129, + "grad_norm": 1.7808153629302979, + "learning_rate": 1.3686194793680807e-05, + "loss": 3.2204, + "step": 9939 + }, + { + "epoch": 0.7969373233649356, + "grad_norm": 1.839053750038147, + "learning_rate": 1.3684987751704295e-05, + "loss": 3.248, + "step": 9940 + }, + { + "epoch": 0.7970174981459582, + "grad_norm": 1.6749635934829712, + "learning_rate": 1.3683780647600191e-05, + "loss": 3.2407, + "step": 9941 + }, + { + "epoch": 0.7970976729269809, + "grad_norm": 1.6955373287200928, + "learning_rate": 1.3682573481388839e-05, + "loss": 3.1843, + "step": 9942 + }, + { + "epoch": 0.7971778477080035, + "grad_norm": 1.718650221824646, + "learning_rate": 1.36813662530906e-05, + "loss": 3.2261, + "step": 9943 + }, + { + "epoch": 0.7972580224890261, + "grad_norm": 1.8398398160934448, + "learning_rate": 1.3680158962725823e-05, + "loss": 3.3067, + "step": 9944 + }, + { + "epoch": 0.7973381972700487, + "grad_norm": 1.7815649509429932, + "learning_rate": 1.3678951610314858e-05, + "loss": 3.2119, + "step": 9945 + }, + { + "epoch": 0.7974183720510714, + "grad_norm": 1.7958476543426514, + "learning_rate": 1.3677744195878071e-05, + "loss": 3.2797, + "step": 9946 + }, + { + "epoch": 0.797498546832094, + "grad_norm": 0.7128342390060425, + "learning_rate": 1.3676536719435811e-05, + "loss": 2.4777, + "step": 9947 + }, + { + "epoch": 0.7975787216131166, + "grad_norm": 1.796079158782959, + "learning_rate": 1.367532918100844e-05, + "loss": 3.2482, + "step": 9948 + }, + { + "epoch": 0.7976588963941392, + "grad_norm": 1.702368140220642, + "learning_rate": 1.367412158061631e-05, + "loss": 3.2098, + "step": 9949 + }, + { + "epoch": 0.7977390711751619, + "grad_norm": 1.8204089403152466, + "learning_rate": 1.3672913918279788e-05, + "loss": 3.1958, + "step": 9950 + }, + { + "epoch": 0.7978192459561845, + "grad_norm": 1.7514163255691528, + "learning_rate": 1.3671706194019232e-05, + "loss": 3.194, + "step": 9951 + }, + { + "epoch": 0.7978994207372071, + "grad_norm": 1.8867038488388062, + "learning_rate": 1.3670498407855002e-05, + "loss": 3.224, + "step": 9952 + }, + { + "epoch": 0.7979795955182297, + "grad_norm": 1.863568663597107, + "learning_rate": 1.3669290559807464e-05, + "loss": 3.2333, + "step": 9953 + }, + { + "epoch": 0.7980597702992523, + "grad_norm": 1.8354628086090088, + "learning_rate": 1.3668082649896979e-05, + "loss": 3.2606, + "step": 9954 + }, + { + "epoch": 0.798139945080275, + "grad_norm": 1.7884058952331543, + "learning_rate": 1.3666874678143914e-05, + "loss": 3.1515, + "step": 9955 + }, + { + "epoch": 0.7982201198612976, + "grad_norm": 1.8245468139648438, + "learning_rate": 1.3665666644568634e-05, + "loss": 3.2711, + "step": 9956 + }, + { + "epoch": 0.7983002946423202, + "grad_norm": 1.731195092201233, + "learning_rate": 1.3664458549191505e-05, + "loss": 3.1818, + "step": 9957 + }, + { + "epoch": 0.7983804694233428, + "grad_norm": 1.8633061647415161, + "learning_rate": 1.3663250392032898e-05, + "loss": 3.2811, + "step": 9958 + }, + { + "epoch": 0.7984606442043655, + "grad_norm": 1.7596057653427124, + "learning_rate": 1.366204217311318e-05, + "loss": 3.155, + "step": 9959 + }, + { + "epoch": 0.7985408189853882, + "grad_norm": 1.7012064456939697, + "learning_rate": 1.366083389245272e-05, + "loss": 3.1974, + "step": 9960 + }, + { + "epoch": 0.7986209937664108, + "grad_norm": 1.75335693359375, + "learning_rate": 1.3659625550071896e-05, + "loss": 3.2708, + "step": 9961 + }, + { + "epoch": 0.7987011685474334, + "grad_norm": 1.6913005113601685, + "learning_rate": 1.365841714599107e-05, + "loss": 3.1246, + "step": 9962 + }, + { + "epoch": 0.7987813433284561, + "grad_norm": 1.760654330253601, + "learning_rate": 1.3657208680230622e-05, + "loss": 3.2894, + "step": 9963 + }, + { + "epoch": 0.7988615181094787, + "grad_norm": 1.8180738687515259, + "learning_rate": 1.3656000152810925e-05, + "loss": 3.1875, + "step": 9964 + }, + { + "epoch": 0.7989416928905013, + "grad_norm": 1.746026873588562, + "learning_rate": 1.3654791563752352e-05, + "loss": 3.2562, + "step": 9965 + }, + { + "epoch": 0.7990218676715239, + "grad_norm": 1.8179279565811157, + "learning_rate": 1.3653582913075283e-05, + "loss": 3.2406, + "step": 9966 + }, + { + "epoch": 0.7991020424525466, + "grad_norm": 1.878574013710022, + "learning_rate": 1.365237420080009e-05, + "loss": 3.2542, + "step": 9967 + }, + { + "epoch": 0.7991822172335692, + "grad_norm": 1.6365838050842285, + "learning_rate": 1.3651165426947162e-05, + "loss": 3.0304, + "step": 9968 + }, + { + "epoch": 0.7992623920145918, + "grad_norm": 1.7236577272415161, + "learning_rate": 1.3649956591536868e-05, + "loss": 3.2129, + "step": 9969 + }, + { + "epoch": 0.7993425667956144, + "grad_norm": 1.7794103622436523, + "learning_rate": 1.3648747694589592e-05, + "loss": 3.184, + "step": 9970 + }, + { + "epoch": 0.7994227415766371, + "grad_norm": 1.8302838802337646, + "learning_rate": 1.3647538736125716e-05, + "loss": 3.2042, + "step": 9971 + }, + { + "epoch": 0.7995029163576597, + "grad_norm": 1.7129262685775757, + "learning_rate": 1.3646329716165623e-05, + "loss": 3.3124, + "step": 9972 + }, + { + "epoch": 0.7995830911386823, + "grad_norm": 1.702531099319458, + "learning_rate": 1.3645120634729696e-05, + "loss": 3.137, + "step": 9973 + }, + { + "epoch": 0.7996632659197049, + "grad_norm": 1.688299298286438, + "learning_rate": 1.364391149183832e-05, + "loss": 3.1941, + "step": 9974 + }, + { + "epoch": 0.7997434407007276, + "grad_norm": 1.8609890937805176, + "learning_rate": 1.3642702287511877e-05, + "loss": 3.2889, + "step": 9975 + }, + { + "epoch": 0.7998236154817502, + "grad_norm": 1.781893253326416, + "learning_rate": 1.3641493021770762e-05, + "loss": 3.1133, + "step": 9976 + }, + { + "epoch": 0.7999037902627728, + "grad_norm": 1.779363989830017, + "learning_rate": 1.3640283694635356e-05, + "loss": 3.2793, + "step": 9977 + }, + { + "epoch": 0.7999839650437954, + "grad_norm": 1.7161306142807007, + "learning_rate": 1.363907430612605e-05, + "loss": 3.1998, + "step": 9978 + }, + { + "epoch": 0.800064139824818, + "grad_norm": 1.7856377363204956, + "learning_rate": 1.3637864856263235e-05, + "loss": 3.2124, + "step": 9979 + }, + { + "epoch": 0.8001443146058407, + "grad_norm": 1.710957407951355, + "learning_rate": 1.3636655345067297e-05, + "loss": 3.1747, + "step": 9980 + }, + { + "epoch": 0.8002244893868634, + "grad_norm": 1.9071587324142456, + "learning_rate": 1.3635445772558637e-05, + "loss": 3.2225, + "step": 9981 + }, + { + "epoch": 0.800304664167886, + "grad_norm": 1.9157320261001587, + "learning_rate": 1.3634236138757637e-05, + "loss": 3.1762, + "step": 9982 + }, + { + "epoch": 0.8003848389489087, + "grad_norm": 1.724145531654358, + "learning_rate": 1.36330264436847e-05, + "loss": 3.1998, + "step": 9983 + }, + { + "epoch": 0.8004650137299313, + "grad_norm": 1.834612250328064, + "learning_rate": 1.3631816687360217e-05, + "loss": 3.2119, + "step": 9984 + }, + { + "epoch": 0.8005451885109539, + "grad_norm": 1.8160113096237183, + "learning_rate": 1.3630606869804584e-05, + "loss": 3.2238, + "step": 9985 + }, + { + "epoch": 0.8006253632919765, + "grad_norm": 1.8429337739944458, + "learning_rate": 1.3629396991038199e-05, + "loss": 3.1352, + "step": 9986 + }, + { + "epoch": 0.8007055380729992, + "grad_norm": 1.7681050300598145, + "learning_rate": 1.362818705108146e-05, + "loss": 3.1567, + "step": 9987 + }, + { + "epoch": 0.8007857128540218, + "grad_norm": 1.7185770273208618, + "learning_rate": 1.3626977049954766e-05, + "loss": 3.2423, + "step": 9988 + }, + { + "epoch": 0.8008658876350444, + "grad_norm": 1.8939473628997803, + "learning_rate": 1.3625766987678517e-05, + "loss": 3.1011, + "step": 9989 + }, + { + "epoch": 0.800946062416067, + "grad_norm": 1.9175193309783936, + "learning_rate": 1.3624556864273113e-05, + "loss": 3.1735, + "step": 9990 + }, + { + "epoch": 0.8010262371970897, + "grad_norm": 1.7847034931182861, + "learning_rate": 1.3623346679758963e-05, + "loss": 3.166, + "step": 9991 + }, + { + "epoch": 0.8011064119781123, + "grad_norm": 1.7746375799179077, + "learning_rate": 1.3622136434156459e-05, + "loss": 3.1401, + "step": 9992 + }, + { + "epoch": 0.8011865867591349, + "grad_norm": 1.9384559392929077, + "learning_rate": 1.3620926127486014e-05, + "loss": 3.2566, + "step": 9993 + }, + { + "epoch": 0.8012667615401575, + "grad_norm": 1.7227600812911987, + "learning_rate": 1.3619715759768029e-05, + "loss": 3.1805, + "step": 9994 + }, + { + "epoch": 0.8013469363211801, + "grad_norm": 1.7840948104858398, + "learning_rate": 1.3618505331022912e-05, + "loss": 3.2902, + "step": 9995 + }, + { + "epoch": 0.8014271111022028, + "grad_norm": 1.7719383239746094, + "learning_rate": 1.3617294841271075e-05, + "loss": 3.223, + "step": 9996 + }, + { + "epoch": 0.8015072858832254, + "grad_norm": 1.7881172895431519, + "learning_rate": 1.3616084290532915e-05, + "loss": 3.1771, + "step": 9997 + }, + { + "epoch": 0.801587460664248, + "grad_norm": 1.8091416358947754, + "learning_rate": 1.3614873678828855e-05, + "loss": 3.2305, + "step": 9998 + }, + { + "epoch": 0.8016676354452706, + "grad_norm": 1.7023241519927979, + "learning_rate": 1.3613663006179294e-05, + "loss": 3.246, + "step": 9999 + }, + { + "epoch": 0.8017478102262933, + "grad_norm": 1.9113491773605347, + "learning_rate": 1.3612452272604648e-05, + "loss": 3.2205, + "step": 10000 + }, + { + "epoch": 0.801827985007316, + "grad_norm": 1.7493245601654053, + "learning_rate": 1.3611241478125331e-05, + "loss": 3.1782, + "step": 10001 + }, + { + "epoch": 0.8019081597883386, + "grad_norm": 1.6841626167297363, + "learning_rate": 1.3610030622761756e-05, + "loss": 3.2009, + "step": 10002 + }, + { + "epoch": 0.8019883345693612, + "grad_norm": 0.7597658634185791, + "learning_rate": 1.360881970653434e-05, + "loss": 2.4711, + "step": 10003 + }, + { + "epoch": 0.8020685093503839, + "grad_norm": 1.8722165822982788, + "learning_rate": 1.360760872946349e-05, + "loss": 3.2076, + "step": 10004 + }, + { + "epoch": 0.8021486841314065, + "grad_norm": 1.8484526872634888, + "learning_rate": 1.360639769156963e-05, + "loss": 3.389, + "step": 10005 + }, + { + "epoch": 0.8022288589124291, + "grad_norm": 0.7348129153251648, + "learning_rate": 1.3605186592873178e-05, + "loss": 2.4847, + "step": 10006 + }, + { + "epoch": 0.8023090336934517, + "grad_norm": 1.8738219738006592, + "learning_rate": 1.360397543339455e-05, + "loss": 3.1779, + "step": 10007 + }, + { + "epoch": 0.8023892084744744, + "grad_norm": 1.7504035234451294, + "learning_rate": 1.3602764213154167e-05, + "loss": 3.2104, + "step": 10008 + }, + { + "epoch": 0.802469383255497, + "grad_norm": 1.6885792016983032, + "learning_rate": 1.3601552932172446e-05, + "loss": 3.1685, + "step": 10009 + }, + { + "epoch": 0.8025495580365196, + "grad_norm": 1.8163816928863525, + "learning_rate": 1.3600341590469811e-05, + "loss": 3.1654, + "step": 10010 + }, + { + "epoch": 0.8026297328175422, + "grad_norm": 1.8143013715744019, + "learning_rate": 1.3599130188066692e-05, + "loss": 3.1771, + "step": 10011 + }, + { + "epoch": 0.8027099075985649, + "grad_norm": 1.8612184524536133, + "learning_rate": 1.35979187249835e-05, + "loss": 3.2866, + "step": 10012 + }, + { + "epoch": 0.8027900823795875, + "grad_norm": 1.8228683471679688, + "learning_rate": 1.3596707201240673e-05, + "loss": 3.2369, + "step": 10013 + }, + { + "epoch": 0.8028702571606101, + "grad_norm": 1.7624870538711548, + "learning_rate": 1.3595495616858625e-05, + "loss": 3.2227, + "step": 10014 + }, + { + "epoch": 0.8029504319416327, + "grad_norm": 1.7667394876480103, + "learning_rate": 1.3594283971857789e-05, + "loss": 3.2766, + "step": 10015 + }, + { + "epoch": 0.8030306067226554, + "grad_norm": 1.8526684045791626, + "learning_rate": 1.3593072266258594e-05, + "loss": 3.2269, + "step": 10016 + }, + { + "epoch": 0.803110781503678, + "grad_norm": 1.7440193891525269, + "learning_rate": 1.3591860500081464e-05, + "loss": 3.1493, + "step": 10017 + }, + { + "epoch": 0.8031909562847006, + "grad_norm": 1.794917345046997, + "learning_rate": 1.3590648673346834e-05, + "loss": 3.1591, + "step": 10018 + }, + { + "epoch": 0.8032711310657232, + "grad_norm": 1.761635184288025, + "learning_rate": 1.3589436786075133e-05, + "loss": 2.9971, + "step": 10019 + }, + { + "epoch": 0.8033513058467459, + "grad_norm": 1.8253605365753174, + "learning_rate": 1.358822483828679e-05, + "loss": 3.2295, + "step": 10020 + }, + { + "epoch": 0.8034314806277686, + "grad_norm": 1.8423775434494019, + "learning_rate": 1.3587012830002244e-05, + "loss": 3.2308, + "step": 10021 + }, + { + "epoch": 0.8035116554087912, + "grad_norm": 2.0130019187927246, + "learning_rate": 1.3585800761241923e-05, + "loss": 3.1513, + "step": 10022 + }, + { + "epoch": 0.8035918301898138, + "grad_norm": 1.7966452836990356, + "learning_rate": 1.3584588632026267e-05, + "loss": 3.2111, + "step": 10023 + }, + { + "epoch": 0.8036720049708365, + "grad_norm": 1.716203212738037, + "learning_rate": 1.358337644237571e-05, + "loss": 3.183, + "step": 10024 + }, + { + "epoch": 0.8037521797518591, + "grad_norm": 1.7385215759277344, + "learning_rate": 1.3582164192310687e-05, + "loss": 3.2256, + "step": 10025 + }, + { + "epoch": 0.8038323545328817, + "grad_norm": 1.8540252447128296, + "learning_rate": 1.3580951881851641e-05, + "loss": 3.295, + "step": 10026 + }, + { + "epoch": 0.8039125293139043, + "grad_norm": 1.8058935403823853, + "learning_rate": 1.3579739511019007e-05, + "loss": 3.2061, + "step": 10027 + }, + { + "epoch": 0.803992704094927, + "grad_norm": 1.8635612726211548, + "learning_rate": 1.3578527079833226e-05, + "loss": 3.2178, + "step": 10028 + }, + { + "epoch": 0.8040728788759496, + "grad_norm": 1.827377200126648, + "learning_rate": 1.3577314588314741e-05, + "loss": 3.2093, + "step": 10029 + }, + { + "epoch": 0.8041530536569722, + "grad_norm": 1.8394562005996704, + "learning_rate": 1.3576102036483992e-05, + "loss": 3.2522, + "step": 10030 + }, + { + "epoch": 0.8042332284379948, + "grad_norm": 0.7681650519371033, + "learning_rate": 1.3574889424361424e-05, + "loss": 2.4985, + "step": 10031 + }, + { + "epoch": 0.8043134032190175, + "grad_norm": 1.777753472328186, + "learning_rate": 1.357367675196748e-05, + "loss": 3.1915, + "step": 10032 + }, + { + "epoch": 0.8043935780000401, + "grad_norm": 0.7200958728790283, + "learning_rate": 1.3572464019322603e-05, + "loss": 2.474, + "step": 10033 + }, + { + "epoch": 0.8044737527810627, + "grad_norm": 1.8926340341567993, + "learning_rate": 1.3571251226447244e-05, + "loss": 3.2557, + "step": 10034 + }, + { + "epoch": 0.8045539275620853, + "grad_norm": 1.7538321018218994, + "learning_rate": 1.3570038373361846e-05, + "loss": 3.257, + "step": 10035 + }, + { + "epoch": 0.804634102343108, + "grad_norm": 1.7793537378311157, + "learning_rate": 1.3568825460086862e-05, + "loss": 3.2388, + "step": 10036 + }, + { + "epoch": 0.8047142771241306, + "grad_norm": 1.7695454359054565, + "learning_rate": 1.3567612486642735e-05, + "loss": 3.2463, + "step": 10037 + }, + { + "epoch": 0.8047944519051532, + "grad_norm": 1.7784677743911743, + "learning_rate": 1.356639945304992e-05, + "loss": 3.1814, + "step": 10038 + }, + { + "epoch": 0.8048746266861758, + "grad_norm": 1.6985571384429932, + "learning_rate": 1.356518635932887e-05, + "loss": 3.1628, + "step": 10039 + }, + { + "epoch": 0.8049548014671984, + "grad_norm": 1.744262933731079, + "learning_rate": 1.3563973205500033e-05, + "loss": 3.1394, + "step": 10040 + }, + { + "epoch": 0.8050349762482211, + "grad_norm": 1.7453666925430298, + "learning_rate": 1.3562759991583864e-05, + "loss": 3.1937, + "step": 10041 + }, + { + "epoch": 0.8051151510292438, + "grad_norm": 1.8398244380950928, + "learning_rate": 1.3561546717600815e-05, + "loss": 3.2189, + "step": 10042 + }, + { + "epoch": 0.8051953258102664, + "grad_norm": 2.1136996746063232, + "learning_rate": 1.3560333383571345e-05, + "loss": 3.3108, + "step": 10043 + }, + { + "epoch": 0.805275500591289, + "grad_norm": 1.6941697597503662, + "learning_rate": 1.3559119989515909e-05, + "loss": 3.1204, + "step": 10044 + }, + { + "epoch": 0.8053556753723117, + "grad_norm": 1.8492963314056396, + "learning_rate": 1.3557906535454965e-05, + "loss": 3.205, + "step": 10045 + }, + { + "epoch": 0.8054358501533343, + "grad_norm": 1.8485738039016724, + "learning_rate": 1.355669302140897e-05, + "loss": 3.1931, + "step": 10046 + }, + { + "epoch": 0.8055160249343569, + "grad_norm": 1.7372249364852905, + "learning_rate": 1.3555479447398385e-05, + "loss": 3.2994, + "step": 10047 + }, + { + "epoch": 0.8055961997153795, + "grad_norm": 0.7375771999359131, + "learning_rate": 1.3554265813443669e-05, + "loss": 2.4984, + "step": 10048 + }, + { + "epoch": 0.8056763744964022, + "grad_norm": 1.8165901899337769, + "learning_rate": 1.3553052119565285e-05, + "loss": 3.2308, + "step": 10049 + }, + { + "epoch": 0.8057565492774248, + "grad_norm": 1.7800780534744263, + "learning_rate": 1.3551838365783692e-05, + "loss": 3.167, + "step": 10050 + }, + { + "epoch": 0.8058367240584474, + "grad_norm": 2.230630874633789, + "learning_rate": 1.3550624552119358e-05, + "loss": 3.2232, + "step": 10051 + }, + { + "epoch": 0.80591689883947, + "grad_norm": 1.8248581886291504, + "learning_rate": 1.3549410678592745e-05, + "loss": 3.1997, + "step": 10052 + }, + { + "epoch": 0.8059970736204927, + "grad_norm": 1.873143196105957, + "learning_rate": 1.3548196745224316e-05, + "loss": 3.2134, + "step": 10053 + }, + { + "epoch": 0.8060772484015153, + "grad_norm": 1.8732863664627075, + "learning_rate": 1.3546982752034545e-05, + "loss": 3.3651, + "step": 10054 + }, + { + "epoch": 0.8061574231825379, + "grad_norm": 0.7771098613739014, + "learning_rate": 1.3545768699043893e-05, + "loss": 2.4768, + "step": 10055 + }, + { + "epoch": 0.8062375979635605, + "grad_norm": 1.7617989778518677, + "learning_rate": 1.3544554586272831e-05, + "loss": 3.1487, + "step": 10056 + }, + { + "epoch": 0.8063177727445832, + "grad_norm": 1.8027703762054443, + "learning_rate": 1.3543340413741826e-05, + "loss": 3.2145, + "step": 10057 + }, + { + "epoch": 0.8063979475256058, + "grad_norm": 1.8758152723312378, + "learning_rate": 1.3542126181471349e-05, + "loss": 3.3195, + "step": 10058 + }, + { + "epoch": 0.8064781223066284, + "grad_norm": 1.818005084991455, + "learning_rate": 1.3540911889481878e-05, + "loss": 3.2261, + "step": 10059 + }, + { + "epoch": 0.806558297087651, + "grad_norm": 1.89998197555542, + "learning_rate": 1.3539697537793877e-05, + "loss": 3.2942, + "step": 10060 + }, + { + "epoch": 0.8066384718686737, + "grad_norm": 0.6934544444084167, + "learning_rate": 1.3538483126427825e-05, + "loss": 2.4698, + "step": 10061 + }, + { + "epoch": 0.8067186466496964, + "grad_norm": 1.774613618850708, + "learning_rate": 1.3537268655404195e-05, + "loss": 3.2217, + "step": 10062 + }, + { + "epoch": 0.806798821430719, + "grad_norm": 1.8031673431396484, + "learning_rate": 1.3536054124743459e-05, + "loss": 3.1626, + "step": 10063 + }, + { + "epoch": 0.8068789962117416, + "grad_norm": 1.7162657976150513, + "learning_rate": 1.35348395344661e-05, + "loss": 3.207, + "step": 10064 + }, + { + "epoch": 0.8069591709927643, + "grad_norm": 1.7338532209396362, + "learning_rate": 1.3533624884592588e-05, + "loss": 3.2191, + "step": 10065 + }, + { + "epoch": 0.8070393457737869, + "grad_norm": 1.8160288333892822, + "learning_rate": 1.3532410175143408e-05, + "loss": 3.1135, + "step": 10066 + }, + { + "epoch": 0.8071195205548095, + "grad_norm": 1.6865720748901367, + "learning_rate": 1.3531195406139039e-05, + "loss": 3.2145, + "step": 10067 + }, + { + "epoch": 0.8071996953358321, + "grad_norm": 1.7541251182556152, + "learning_rate": 1.3529980577599959e-05, + "loss": 3.2336, + "step": 10068 + }, + { + "epoch": 0.8072798701168548, + "grad_norm": 1.7019528150558472, + "learning_rate": 1.352876568954665e-05, + "loss": 3.1306, + "step": 10069 + }, + { + "epoch": 0.8073600448978774, + "grad_norm": 0.7595229744911194, + "learning_rate": 1.3527550741999595e-05, + "loss": 2.4847, + "step": 10070 + }, + { + "epoch": 0.8074402196789, + "grad_norm": 1.7754219770431519, + "learning_rate": 1.352633573497928e-05, + "loss": 3.1918, + "step": 10071 + }, + { + "epoch": 0.8075203944599226, + "grad_norm": 1.7815691232681274, + "learning_rate": 1.3525120668506185e-05, + "loss": 3.226, + "step": 10072 + }, + { + "epoch": 0.8076005692409453, + "grad_norm": 1.8421543836593628, + "learning_rate": 1.3523905542600796e-05, + "loss": 3.2551, + "step": 10073 + }, + { + "epoch": 0.8076807440219679, + "grad_norm": 1.7994946241378784, + "learning_rate": 1.3522690357283606e-05, + "loss": 3.1732, + "step": 10074 + }, + { + "epoch": 0.8077609188029905, + "grad_norm": 1.791129469871521, + "learning_rate": 1.3521475112575096e-05, + "loss": 3.2294, + "step": 10075 + }, + { + "epoch": 0.8078410935840131, + "grad_norm": 1.8085697889328003, + "learning_rate": 1.3520259808495756e-05, + "loss": 3.1978, + "step": 10076 + }, + { + "epoch": 0.8079212683650357, + "grad_norm": 1.930603265762329, + "learning_rate": 1.3519044445066079e-05, + "loss": 3.209, + "step": 10077 + }, + { + "epoch": 0.8080014431460584, + "grad_norm": 1.7172049283981323, + "learning_rate": 1.3517829022306547e-05, + "loss": 3.1565, + "step": 10078 + }, + { + "epoch": 0.808081617927081, + "grad_norm": 0.7240562438964844, + "learning_rate": 1.3516613540237664e-05, + "loss": 2.4911, + "step": 10079 + }, + { + "epoch": 0.8081617927081036, + "grad_norm": 1.8048651218414307, + "learning_rate": 1.3515397998879913e-05, + "loss": 3.1517, + "step": 10080 + }, + { + "epoch": 0.8082419674891262, + "grad_norm": 0.742724597454071, + "learning_rate": 1.3514182398253793e-05, + "loss": 2.4847, + "step": 10081 + }, + { + "epoch": 0.808322142270149, + "grad_norm": 1.7202993631362915, + "learning_rate": 1.3512966738379794e-05, + "loss": 3.0681, + "step": 10082 + }, + { + "epoch": 0.8084023170511716, + "grad_norm": 1.6808289289474487, + "learning_rate": 1.3511751019278412e-05, + "loss": 3.3195, + "step": 10083 + }, + { + "epoch": 0.8084824918321942, + "grad_norm": 1.7614631652832031, + "learning_rate": 1.3510535240970151e-05, + "loss": 3.2671, + "step": 10084 + }, + { + "epoch": 0.8085626666132169, + "grad_norm": 1.8239365816116333, + "learning_rate": 1.3509319403475498e-05, + "loss": 3.2877, + "step": 10085 + }, + { + "epoch": 0.8086428413942395, + "grad_norm": 1.733921766281128, + "learning_rate": 1.3508103506814961e-05, + "loss": 3.3026, + "step": 10086 + }, + { + "epoch": 0.8087230161752621, + "grad_norm": 1.8507907390594482, + "learning_rate": 1.350688755100903e-05, + "loss": 3.1036, + "step": 10087 + }, + { + "epoch": 0.8088031909562847, + "grad_norm": 1.7753101587295532, + "learning_rate": 1.3505671536078214e-05, + "loss": 3.1556, + "step": 10088 + }, + { + "epoch": 0.8088833657373073, + "grad_norm": 1.833944320678711, + "learning_rate": 1.3504455462043011e-05, + "loss": 3.2522, + "step": 10089 + }, + { + "epoch": 0.80896354051833, + "grad_norm": 1.72703218460083, + "learning_rate": 1.3503239328923923e-05, + "loss": 3.2906, + "step": 10090 + }, + { + "epoch": 0.8090437152993526, + "grad_norm": 1.8322248458862305, + "learning_rate": 1.3502023136741458e-05, + "loss": 3.183, + "step": 10091 + }, + { + "epoch": 0.8091238900803752, + "grad_norm": 1.9549182653427124, + "learning_rate": 1.3500806885516113e-05, + "loss": 3.2484, + "step": 10092 + }, + { + "epoch": 0.8092040648613978, + "grad_norm": 1.7237091064453125, + "learning_rate": 1.3499590575268399e-05, + "loss": 3.1764, + "step": 10093 + }, + { + "epoch": 0.8092842396424205, + "grad_norm": 1.8345311880111694, + "learning_rate": 1.3498374206018823e-05, + "loss": 3.2236, + "step": 10094 + }, + { + "epoch": 0.8093644144234431, + "grad_norm": 1.803427815437317, + "learning_rate": 1.3497157777787888e-05, + "loss": 3.165, + "step": 10095 + }, + { + "epoch": 0.8094445892044657, + "grad_norm": 1.7950210571289062, + "learning_rate": 1.3495941290596109e-05, + "loss": 3.2669, + "step": 10096 + }, + { + "epoch": 0.8095247639854883, + "grad_norm": 1.7351491451263428, + "learning_rate": 1.3494724744463987e-05, + "loss": 3.2696, + "step": 10097 + }, + { + "epoch": 0.809604938766511, + "grad_norm": 0.7536272406578064, + "learning_rate": 1.3493508139412042e-05, + "loss": 2.5006, + "step": 10098 + }, + { + "epoch": 0.8096851135475336, + "grad_norm": 1.7746025323867798, + "learning_rate": 1.349229147546078e-05, + "loss": 3.2122, + "step": 10099 + }, + { + "epoch": 0.8097652883285562, + "grad_norm": 1.7933112382888794, + "learning_rate": 1.3491074752630713e-05, + "loss": 3.165, + "step": 10100 + }, + { + "epoch": 0.8098454631095788, + "grad_norm": 1.7771036624908447, + "learning_rate": 1.3489857970942356e-05, + "loss": 3.0527, + "step": 10101 + }, + { + "epoch": 0.8099256378906015, + "grad_norm": 1.8654297590255737, + "learning_rate": 1.3488641130416227e-05, + "loss": 3.2174, + "step": 10102 + }, + { + "epoch": 0.8100058126716242, + "grad_norm": 1.7075873613357544, + "learning_rate": 1.3487424231072832e-05, + "loss": 3.2026, + "step": 10103 + }, + { + "epoch": 0.8100859874526468, + "grad_norm": 1.8880064487457275, + "learning_rate": 1.3486207272932702e-05, + "loss": 3.2427, + "step": 10104 + }, + { + "epoch": 0.8101661622336694, + "grad_norm": 1.7121340036392212, + "learning_rate": 1.3484990256016337e-05, + "loss": 3.1188, + "step": 10105 + }, + { + "epoch": 0.8102463370146921, + "grad_norm": 1.875848412513733, + "learning_rate": 1.348377318034427e-05, + "loss": 3.1843, + "step": 10106 + }, + { + "epoch": 0.8103265117957147, + "grad_norm": 1.8255847692489624, + "learning_rate": 1.3482556045937014e-05, + "loss": 3.1453, + "step": 10107 + }, + { + "epoch": 0.8104066865767373, + "grad_norm": 1.925748586654663, + "learning_rate": 1.3481338852815089e-05, + "loss": 3.2648, + "step": 10108 + }, + { + "epoch": 0.8104868613577599, + "grad_norm": 1.7069387435913086, + "learning_rate": 1.3480121600999021e-05, + "loss": 3.155, + "step": 10109 + }, + { + "epoch": 0.8105670361387826, + "grad_norm": 1.7169060707092285, + "learning_rate": 1.3478904290509324e-05, + "loss": 3.1873, + "step": 10110 + }, + { + "epoch": 0.8106472109198052, + "grad_norm": 1.8903672695159912, + "learning_rate": 1.3477686921366532e-05, + "loss": 3.1763, + "step": 10111 + }, + { + "epoch": 0.8107273857008278, + "grad_norm": 1.7472953796386719, + "learning_rate": 1.3476469493591163e-05, + "loss": 3.2391, + "step": 10112 + }, + { + "epoch": 0.8108075604818504, + "grad_norm": 1.775978922843933, + "learning_rate": 1.347525200720374e-05, + "loss": 3.1077, + "step": 10113 + }, + { + "epoch": 0.810887735262873, + "grad_norm": 1.7996971607208252, + "learning_rate": 1.3474034462224798e-05, + "loss": 3.2487, + "step": 10114 + }, + { + "epoch": 0.8109679100438957, + "grad_norm": 1.7830960750579834, + "learning_rate": 1.3472816858674856e-05, + "loss": 3.1458, + "step": 10115 + }, + { + "epoch": 0.8110480848249183, + "grad_norm": 1.7112349271774292, + "learning_rate": 1.3471599196574447e-05, + "loss": 3.1604, + "step": 10116 + }, + { + "epoch": 0.8111282596059409, + "grad_norm": 1.8063774108886719, + "learning_rate": 1.3470381475944098e-05, + "loss": 3.3394, + "step": 10117 + }, + { + "epoch": 0.8112084343869636, + "grad_norm": 1.7725672721862793, + "learning_rate": 1.3469163696804341e-05, + "loss": 3.2678, + "step": 10118 + }, + { + "epoch": 0.8112886091679862, + "grad_norm": 1.7547216415405273, + "learning_rate": 1.3467945859175708e-05, + "loss": 3.2545, + "step": 10119 + }, + { + "epoch": 0.8113687839490088, + "grad_norm": 1.6736193895339966, + "learning_rate": 1.346672796307873e-05, + "loss": 3.1958, + "step": 10120 + }, + { + "epoch": 0.8114489587300314, + "grad_norm": 1.7814557552337646, + "learning_rate": 1.3465510008533939e-05, + "loss": 3.1781, + "step": 10121 + }, + { + "epoch": 0.811529133511054, + "grad_norm": 1.7609057426452637, + "learning_rate": 1.3464291995561868e-05, + "loss": 3.1832, + "step": 10122 + }, + { + "epoch": 0.8116093082920768, + "grad_norm": 1.8250775337219238, + "learning_rate": 1.3463073924183059e-05, + "loss": 3.1937, + "step": 10123 + }, + { + "epoch": 0.8116894830730994, + "grad_norm": 1.8594506978988647, + "learning_rate": 1.3461855794418043e-05, + "loss": 3.1959, + "step": 10124 + }, + { + "epoch": 0.811769657854122, + "grad_norm": 1.7875744104385376, + "learning_rate": 1.3460637606287358e-05, + "loss": 3.2585, + "step": 10125 + }, + { + "epoch": 0.8118498326351447, + "grad_norm": 1.7396574020385742, + "learning_rate": 1.3459419359811545e-05, + "loss": 3.2123, + "step": 10126 + }, + { + "epoch": 0.8119300074161673, + "grad_norm": 1.6890206336975098, + "learning_rate": 1.3458201055011138e-05, + "loss": 3.2703, + "step": 10127 + }, + { + "epoch": 0.8120101821971899, + "grad_norm": 1.8402271270751953, + "learning_rate": 1.3456982691906679e-05, + "loss": 3.2536, + "step": 10128 + }, + { + "epoch": 0.8120903569782125, + "grad_norm": 1.754299521446228, + "learning_rate": 1.3455764270518714e-05, + "loss": 3.1756, + "step": 10129 + }, + { + "epoch": 0.8121705317592351, + "grad_norm": 1.7390289306640625, + "learning_rate": 1.345454579086778e-05, + "loss": 3.2021, + "step": 10130 + }, + { + "epoch": 0.8122507065402578, + "grad_norm": 1.8119659423828125, + "learning_rate": 1.3453327252974419e-05, + "loss": 3.2172, + "step": 10131 + }, + { + "epoch": 0.8123308813212804, + "grad_norm": 1.7755753993988037, + "learning_rate": 1.345210865685918e-05, + "loss": 3.2401, + "step": 10132 + }, + { + "epoch": 0.812411056102303, + "grad_norm": 0.7494348883628845, + "learning_rate": 1.3450890002542604e-05, + "loss": 2.4763, + "step": 10133 + }, + { + "epoch": 0.8124912308833256, + "grad_norm": 1.6860578060150146, + "learning_rate": 1.344967129004524e-05, + "loss": 3.114, + "step": 10134 + }, + { + "epoch": 0.8125714056643483, + "grad_norm": 1.842086911201477, + "learning_rate": 1.3448452519387634e-05, + "loss": 3.2155, + "step": 10135 + }, + { + "epoch": 0.8126515804453709, + "grad_norm": 1.898508906364441, + "learning_rate": 1.3447233690590333e-05, + "loss": 3.2357, + "step": 10136 + }, + { + "epoch": 0.8127317552263935, + "grad_norm": 1.747911810874939, + "learning_rate": 1.3446014803673888e-05, + "loss": 3.23, + "step": 10137 + }, + { + "epoch": 0.8128119300074161, + "grad_norm": 1.7995147705078125, + "learning_rate": 1.3444795858658846e-05, + "loss": 3.2002, + "step": 10138 + }, + { + "epoch": 0.8128921047884388, + "grad_norm": 1.7720519304275513, + "learning_rate": 1.3443576855565763e-05, + "loss": 3.3655, + "step": 10139 + }, + { + "epoch": 0.8129722795694614, + "grad_norm": 1.798173427581787, + "learning_rate": 1.3442357794415184e-05, + "loss": 3.2598, + "step": 10140 + }, + { + "epoch": 0.813052454350484, + "grad_norm": 1.8280041217803955, + "learning_rate": 1.344113867522767e-05, + "loss": 3.1814, + "step": 10141 + }, + { + "epoch": 0.8131326291315066, + "grad_norm": 1.8132972717285156, + "learning_rate": 1.3439919498023769e-05, + "loss": 3.2102, + "step": 10142 + }, + { + "epoch": 0.8132128039125293, + "grad_norm": 1.7070730924606323, + "learning_rate": 1.3438700262824034e-05, + "loss": 3.2395, + "step": 10143 + }, + { + "epoch": 0.813292978693552, + "grad_norm": 1.7451478242874146, + "learning_rate": 1.343748096964903e-05, + "loss": 3.3098, + "step": 10144 + }, + { + "epoch": 0.8133731534745746, + "grad_norm": 1.7539925575256348, + "learning_rate": 1.3436261618519305e-05, + "loss": 3.1479, + "step": 10145 + }, + { + "epoch": 0.8134533282555972, + "grad_norm": 1.853788137435913, + "learning_rate": 1.3435042209455423e-05, + "loss": 3.2628, + "step": 10146 + }, + { + "epoch": 0.8135335030366199, + "grad_norm": 1.748972773551941, + "learning_rate": 1.3433822742477935e-05, + "loss": 3.0944, + "step": 10147 + }, + { + "epoch": 0.8136136778176425, + "grad_norm": 1.8671345710754395, + "learning_rate": 1.343260321760741e-05, + "loss": 3.2147, + "step": 10148 + }, + { + "epoch": 0.8136938525986651, + "grad_norm": 1.7570585012435913, + "learning_rate": 1.3431383634864403e-05, + "loss": 3.1665, + "step": 10149 + }, + { + "epoch": 0.8137740273796877, + "grad_norm": 1.8773999214172363, + "learning_rate": 1.3430163994269477e-05, + "loss": 3.2515, + "step": 10150 + }, + { + "epoch": 0.8138542021607104, + "grad_norm": 1.7436987161636353, + "learning_rate": 1.3428944295843196e-05, + "loss": 3.1836, + "step": 10151 + }, + { + "epoch": 0.813934376941733, + "grad_norm": 1.6992844343185425, + "learning_rate": 1.3427724539606121e-05, + "loss": 3.1897, + "step": 10152 + }, + { + "epoch": 0.8140145517227556, + "grad_norm": 1.8070933818817139, + "learning_rate": 1.342650472557882e-05, + "loss": 3.2112, + "step": 10153 + }, + { + "epoch": 0.8140947265037782, + "grad_norm": 1.7341716289520264, + "learning_rate": 1.342528485378186e-05, + "loss": 3.1519, + "step": 10154 + }, + { + "epoch": 0.8141749012848009, + "grad_norm": 1.7550402879714966, + "learning_rate": 1.3424064924235798e-05, + "loss": 3.1865, + "step": 10155 + }, + { + "epoch": 0.8142550760658235, + "grad_norm": 1.6950693130493164, + "learning_rate": 1.3422844936961212e-05, + "loss": 3.1973, + "step": 10156 + }, + { + "epoch": 0.8143352508468461, + "grad_norm": 1.7189197540283203, + "learning_rate": 1.3421624891978665e-05, + "loss": 3.1913, + "step": 10157 + }, + { + "epoch": 0.8144154256278687, + "grad_norm": 1.7795825004577637, + "learning_rate": 1.3420404789308728e-05, + "loss": 3.1555, + "step": 10158 + }, + { + "epoch": 0.8144956004088914, + "grad_norm": 1.8210971355438232, + "learning_rate": 1.3419184628971975e-05, + "loss": 3.3182, + "step": 10159 + }, + { + "epoch": 0.814575775189914, + "grad_norm": 1.8309643268585205, + "learning_rate": 1.341796441098897e-05, + "loss": 3.3937, + "step": 10160 + }, + { + "epoch": 0.8146559499709366, + "grad_norm": 1.7195672988891602, + "learning_rate": 1.3416744135380294e-05, + "loss": 3.1722, + "step": 10161 + }, + { + "epoch": 0.8147361247519592, + "grad_norm": 1.7695811986923218, + "learning_rate": 1.3415523802166512e-05, + "loss": 3.0935, + "step": 10162 + }, + { + "epoch": 0.8148162995329818, + "grad_norm": 1.7370495796203613, + "learning_rate": 1.3414303411368203e-05, + "loss": 3.1996, + "step": 10163 + }, + { + "epoch": 0.8148964743140046, + "grad_norm": 1.8078030347824097, + "learning_rate": 1.3413082963005945e-05, + "loss": 3.2175, + "step": 10164 + }, + { + "epoch": 0.8149766490950272, + "grad_norm": 0.7037444114685059, + "learning_rate": 1.3411862457100309e-05, + "loss": 2.4807, + "step": 10165 + }, + { + "epoch": 0.8150568238760498, + "grad_norm": 1.7550320625305176, + "learning_rate": 1.3410641893671874e-05, + "loss": 3.1884, + "step": 10166 + }, + { + "epoch": 0.8151369986570725, + "grad_norm": 1.8434890508651733, + "learning_rate": 1.3409421272741219e-05, + "loss": 3.2361, + "step": 10167 + }, + { + "epoch": 0.8152171734380951, + "grad_norm": 1.7364685535430908, + "learning_rate": 1.3408200594328922e-05, + "loss": 3.1379, + "step": 10168 + }, + { + "epoch": 0.8152973482191177, + "grad_norm": 1.871605396270752, + "learning_rate": 1.3406979858455566e-05, + "loss": 3.1719, + "step": 10169 + }, + { + "epoch": 0.8153775230001403, + "grad_norm": 1.7618968486785889, + "learning_rate": 1.340575906514173e-05, + "loss": 3.1643, + "step": 10170 + }, + { + "epoch": 0.815457697781163, + "grad_norm": 2.120836019515991, + "learning_rate": 1.3404538214407996e-05, + "loss": 3.2811, + "step": 10171 + }, + { + "epoch": 0.8155378725621856, + "grad_norm": 1.7233328819274902, + "learning_rate": 1.3403317306274948e-05, + "loss": 3.1378, + "step": 10172 + }, + { + "epoch": 0.8156180473432082, + "grad_norm": 1.8361948728561401, + "learning_rate": 1.3402096340763168e-05, + "loss": 3.2439, + "step": 10173 + }, + { + "epoch": 0.8156982221242308, + "grad_norm": 1.8867746591567993, + "learning_rate": 1.3400875317893245e-05, + "loss": 3.2384, + "step": 10174 + }, + { + "epoch": 0.8157783969052534, + "grad_norm": 1.847115397453308, + "learning_rate": 1.3399654237685762e-05, + "loss": 3.2313, + "step": 10175 + }, + { + "epoch": 0.8158585716862761, + "grad_norm": 0.7265251874923706, + "learning_rate": 1.3398433100161308e-05, + "loss": 2.4946, + "step": 10176 + }, + { + "epoch": 0.8159387464672987, + "grad_norm": 0.754084587097168, + "learning_rate": 1.3397211905340465e-05, + "loss": 2.4869, + "step": 10177 + }, + { + "epoch": 0.8160189212483213, + "grad_norm": 1.8547544479370117, + "learning_rate": 1.3395990653243831e-05, + "loss": 3.1984, + "step": 10178 + }, + { + "epoch": 0.8160990960293439, + "grad_norm": 1.7173173427581787, + "learning_rate": 1.339476934389199e-05, + "loss": 3.1552, + "step": 10179 + }, + { + "epoch": 0.8161792708103666, + "grad_norm": 1.776187539100647, + "learning_rate": 1.3393547977305534e-05, + "loss": 3.2214, + "step": 10180 + }, + { + "epoch": 0.8162594455913892, + "grad_norm": 1.778928518295288, + "learning_rate": 1.3392326553505057e-05, + "loss": 3.3445, + "step": 10181 + }, + { + "epoch": 0.8163396203724118, + "grad_norm": 1.8740342855453491, + "learning_rate": 1.3391105072511148e-05, + "loss": 3.2717, + "step": 10182 + }, + { + "epoch": 0.8164197951534344, + "grad_norm": 1.8326044082641602, + "learning_rate": 1.3389883534344401e-05, + "loss": 3.2297, + "step": 10183 + }, + { + "epoch": 0.8164999699344572, + "grad_norm": 1.7048970460891724, + "learning_rate": 1.3388661939025416e-05, + "loss": 3.2081, + "step": 10184 + }, + { + "epoch": 0.8165801447154798, + "grad_norm": 0.7572863698005676, + "learning_rate": 1.3387440286574783e-05, + "loss": 2.4792, + "step": 10185 + }, + { + "epoch": 0.8166603194965024, + "grad_norm": 1.7996472120285034, + "learning_rate": 1.33862185770131e-05, + "loss": 3.2175, + "step": 10186 + }, + { + "epoch": 0.816740494277525, + "grad_norm": 1.763621211051941, + "learning_rate": 1.3384996810360966e-05, + "loss": 3.0898, + "step": 10187 + }, + { + "epoch": 0.8168206690585477, + "grad_norm": 1.8029283285140991, + "learning_rate": 1.3383774986638977e-05, + "loss": 3.1549, + "step": 10188 + }, + { + "epoch": 0.8169008438395703, + "grad_norm": 1.7641565799713135, + "learning_rate": 1.3382553105867739e-05, + "loss": 3.1934, + "step": 10189 + }, + { + "epoch": 0.8169810186205929, + "grad_norm": 1.6525938510894775, + "learning_rate": 1.3381331168067844e-05, + "loss": 3.1356, + "step": 10190 + }, + { + "epoch": 0.8170611934016155, + "grad_norm": 1.821141242980957, + "learning_rate": 1.33801091732599e-05, + "loss": 3.1636, + "step": 10191 + }, + { + "epoch": 0.8171413681826382, + "grad_norm": 1.764623761177063, + "learning_rate": 1.3378887121464504e-05, + "loss": 3.1574, + "step": 10192 + }, + { + "epoch": 0.8172215429636608, + "grad_norm": 1.832663893699646, + "learning_rate": 1.3377665012702262e-05, + "loss": 3.23, + "step": 10193 + }, + { + "epoch": 0.8173017177446834, + "grad_norm": 1.8269984722137451, + "learning_rate": 1.337644284699378e-05, + "loss": 3.1825, + "step": 10194 + }, + { + "epoch": 0.817381892525706, + "grad_norm": 1.7062084674835205, + "learning_rate": 1.337522062435966e-05, + "loss": 3.1549, + "step": 10195 + }, + { + "epoch": 0.8174620673067287, + "grad_norm": 2.0818567276000977, + "learning_rate": 1.3373998344820512e-05, + "loss": 3.1429, + "step": 10196 + }, + { + "epoch": 0.8175422420877513, + "grad_norm": 1.7994292974472046, + "learning_rate": 1.3372776008396937e-05, + "loss": 3.244, + "step": 10197 + }, + { + "epoch": 0.8176224168687739, + "grad_norm": 1.7753522396087646, + "learning_rate": 1.3371553615109548e-05, + "loss": 3.2163, + "step": 10198 + }, + { + "epoch": 0.8177025916497965, + "grad_norm": 1.7388899326324463, + "learning_rate": 1.3370331164978958e-05, + "loss": 3.3165, + "step": 10199 + }, + { + "epoch": 0.8177827664308192, + "grad_norm": 1.9156957864761353, + "learning_rate": 1.3369108658025769e-05, + "loss": 3.2488, + "step": 10200 + }, + { + "epoch": 0.8178629412118418, + "grad_norm": 1.8123646974563599, + "learning_rate": 1.3367886094270597e-05, + "loss": 3.1676, + "step": 10201 + }, + { + "epoch": 0.8179431159928644, + "grad_norm": 1.8538092374801636, + "learning_rate": 1.3366663473734049e-05, + "loss": 3.1855, + "step": 10202 + }, + { + "epoch": 0.818023290773887, + "grad_norm": 1.7813217639923096, + "learning_rate": 1.3365440796436744e-05, + "loss": 3.1597, + "step": 10203 + }, + { + "epoch": 0.8181034655549096, + "grad_norm": 1.8031023740768433, + "learning_rate": 1.3364218062399292e-05, + "loss": 3.3014, + "step": 10204 + }, + { + "epoch": 0.8181836403359324, + "grad_norm": 1.8219823837280273, + "learning_rate": 1.336299527164231e-05, + "loss": 3.2653, + "step": 10205 + }, + { + "epoch": 0.818263815116955, + "grad_norm": 1.7144066095352173, + "learning_rate": 1.3361772424186414e-05, + "loss": 3.175, + "step": 10206 + }, + { + "epoch": 0.8183439898979776, + "grad_norm": 1.7151434421539307, + "learning_rate": 1.3360549520052216e-05, + "loss": 3.125, + "step": 10207 + }, + { + "epoch": 0.8184241646790003, + "grad_norm": 1.882439374923706, + "learning_rate": 1.3359326559260341e-05, + "loss": 3.2162, + "step": 10208 + }, + { + "epoch": 0.8185043394600229, + "grad_norm": 1.6773308515548706, + "learning_rate": 1.3358103541831404e-05, + "loss": 3.2313, + "step": 10209 + }, + { + "epoch": 0.8185845142410455, + "grad_norm": 1.7812048196792603, + "learning_rate": 1.3356880467786025e-05, + "loss": 3.2653, + "step": 10210 + }, + { + "epoch": 0.8186646890220681, + "grad_norm": 1.8345773220062256, + "learning_rate": 1.3355657337144823e-05, + "loss": 3.1317, + "step": 10211 + }, + { + "epoch": 0.8187448638030907, + "grad_norm": 1.6603572368621826, + "learning_rate": 1.3354434149928421e-05, + "loss": 3.2494, + "step": 10212 + }, + { + "epoch": 0.8188250385841134, + "grad_norm": 1.8495359420776367, + "learning_rate": 1.3353210906157441e-05, + "loss": 3.1655, + "step": 10213 + }, + { + "epoch": 0.818905213365136, + "grad_norm": 0.7988433837890625, + "learning_rate": 1.3351987605852507e-05, + "loss": 2.4993, + "step": 10214 + }, + { + "epoch": 0.8189853881461586, + "grad_norm": 1.899737000465393, + "learning_rate": 1.3350764249034245e-05, + "loss": 3.2932, + "step": 10215 + }, + { + "epoch": 0.8190655629271812, + "grad_norm": 1.8424761295318604, + "learning_rate": 1.3349540835723278e-05, + "loss": 3.1611, + "step": 10216 + }, + { + "epoch": 0.8191457377082039, + "grad_norm": 1.7253458499908447, + "learning_rate": 1.3348317365940231e-05, + "loss": 3.1589, + "step": 10217 + }, + { + "epoch": 0.8192259124892265, + "grad_norm": 1.8481751680374146, + "learning_rate": 1.3347093839705737e-05, + "loss": 3.1801, + "step": 10218 + }, + { + "epoch": 0.8193060872702491, + "grad_norm": 1.7472981214523315, + "learning_rate": 1.3345870257040418e-05, + "loss": 3.2548, + "step": 10219 + }, + { + "epoch": 0.8193862620512717, + "grad_norm": 1.6963233947753906, + "learning_rate": 1.3344646617964904e-05, + "loss": 3.1536, + "step": 10220 + }, + { + "epoch": 0.8194664368322944, + "grad_norm": 1.6893596649169922, + "learning_rate": 1.3343422922499829e-05, + "loss": 3.1839, + "step": 10221 + }, + { + "epoch": 0.819546611613317, + "grad_norm": 1.817196011543274, + "learning_rate": 1.3342199170665824e-05, + "loss": 3.3225, + "step": 10222 + }, + { + "epoch": 0.8196267863943396, + "grad_norm": 1.7343894243240356, + "learning_rate": 1.3340975362483514e-05, + "loss": 3.1602, + "step": 10223 + }, + { + "epoch": 0.8197069611753622, + "grad_norm": 1.7950729131698608, + "learning_rate": 1.3339751497973543e-05, + "loss": 3.2228, + "step": 10224 + }, + { + "epoch": 0.819787135956385, + "grad_norm": 0.753385066986084, + "learning_rate": 1.3338527577156532e-05, + "loss": 2.488, + "step": 10225 + }, + { + "epoch": 0.8198673107374076, + "grad_norm": 1.9701321125030518, + "learning_rate": 1.3337303600053126e-05, + "loss": 3.1326, + "step": 10226 + }, + { + "epoch": 0.8199474855184302, + "grad_norm": 1.7693605422973633, + "learning_rate": 1.333607956668396e-05, + "loss": 3.1078, + "step": 10227 + }, + { + "epoch": 0.8200276602994528, + "grad_norm": 1.7981750965118408, + "learning_rate": 1.3334855477069664e-05, + "loss": 3.2614, + "step": 10228 + }, + { + "epoch": 0.8201078350804755, + "grad_norm": 1.7598093748092651, + "learning_rate": 1.3333631331230883e-05, + "loss": 3.1715, + "step": 10229 + }, + { + "epoch": 0.8201880098614981, + "grad_norm": 1.6684386730194092, + "learning_rate": 1.333240712918825e-05, + "loss": 3.1172, + "step": 10230 + }, + { + "epoch": 0.8202681846425207, + "grad_norm": 1.844230055809021, + "learning_rate": 1.3331182870962408e-05, + "loss": 3.1794, + "step": 10231 + }, + { + "epoch": 0.8203483594235433, + "grad_norm": 1.8677088022232056, + "learning_rate": 1.3329958556573999e-05, + "loss": 3.341, + "step": 10232 + }, + { + "epoch": 0.820428534204566, + "grad_norm": 1.922563910484314, + "learning_rate": 1.332873418604366e-05, + "loss": 3.1674, + "step": 10233 + }, + { + "epoch": 0.8205087089855886, + "grad_norm": 1.8615524768829346, + "learning_rate": 1.3327509759392036e-05, + "loss": 3.1795, + "step": 10234 + }, + { + "epoch": 0.8205888837666112, + "grad_norm": 2.0941596031188965, + "learning_rate": 1.3326285276639772e-05, + "loss": 3.2802, + "step": 10235 + }, + { + "epoch": 0.8206690585476338, + "grad_norm": 1.7633039951324463, + "learning_rate": 1.3325060737807508e-05, + "loss": 3.2606, + "step": 10236 + }, + { + "epoch": 0.8207492333286565, + "grad_norm": 1.8523337841033936, + "learning_rate": 1.3323836142915893e-05, + "loss": 3.1442, + "step": 10237 + }, + { + "epoch": 0.8208294081096791, + "grad_norm": 1.8024851083755493, + "learning_rate": 1.332261149198557e-05, + "loss": 3.3422, + "step": 10238 + }, + { + "epoch": 0.8209095828907017, + "grad_norm": 1.8408147096633911, + "learning_rate": 1.3321386785037194e-05, + "loss": 3.2147, + "step": 10239 + }, + { + "epoch": 0.8209897576717243, + "grad_norm": 1.7155036926269531, + "learning_rate": 1.3320162022091403e-05, + "loss": 3.1588, + "step": 10240 + }, + { + "epoch": 0.821069932452747, + "grad_norm": 1.9827301502227783, + "learning_rate": 1.3318937203168848e-05, + "loss": 3.2998, + "step": 10241 + }, + { + "epoch": 0.8211501072337696, + "grad_norm": 1.8447496891021729, + "learning_rate": 1.3317712328290186e-05, + "loss": 3.2563, + "step": 10242 + }, + { + "epoch": 0.8212302820147922, + "grad_norm": 1.8757522106170654, + "learning_rate": 1.3316487397476063e-05, + "loss": 3.2723, + "step": 10243 + }, + { + "epoch": 0.8213104567958148, + "grad_norm": 1.675149917602539, + "learning_rate": 1.3315262410747132e-05, + "loss": 3.1345, + "step": 10244 + }, + { + "epoch": 0.8213906315768376, + "grad_norm": 0.7152847647666931, + "learning_rate": 1.3314037368124042e-05, + "loss": 2.4985, + "step": 10245 + }, + { + "epoch": 0.8214708063578602, + "grad_norm": 1.7359586954116821, + "learning_rate": 1.3312812269627452e-05, + "loss": 3.1619, + "step": 10246 + }, + { + "epoch": 0.8215509811388828, + "grad_norm": 1.8891887664794922, + "learning_rate": 1.3311587115278015e-05, + "loss": 3.2268, + "step": 10247 + }, + { + "epoch": 0.8216311559199054, + "grad_norm": 1.6962138414382935, + "learning_rate": 1.3310361905096385e-05, + "loss": 3.1733, + "step": 10248 + }, + { + "epoch": 0.821711330700928, + "grad_norm": 1.7197378873825073, + "learning_rate": 1.3309136639103221e-05, + "loss": 3.2922, + "step": 10249 + }, + { + "epoch": 0.8217915054819507, + "grad_norm": 1.8183979988098145, + "learning_rate": 1.3307911317319178e-05, + "loss": 3.2234, + "step": 10250 + }, + { + "epoch": 0.8218716802629733, + "grad_norm": 1.9053181409835815, + "learning_rate": 1.3306685939764916e-05, + "loss": 3.1731, + "step": 10251 + }, + { + "epoch": 0.8219518550439959, + "grad_norm": 1.7123134136199951, + "learning_rate": 1.3305460506461096e-05, + "loss": 3.2927, + "step": 10252 + }, + { + "epoch": 0.8220320298250186, + "grad_norm": 1.8063180446624756, + "learning_rate": 1.3304235017428375e-05, + "loss": 3.1894, + "step": 10253 + }, + { + "epoch": 0.8221122046060412, + "grad_norm": 1.7369683980941772, + "learning_rate": 1.330300947268742e-05, + "loss": 3.1567, + "step": 10254 + }, + { + "epoch": 0.8221923793870638, + "grad_norm": 1.7758381366729736, + "learning_rate": 1.3301783872258885e-05, + "loss": 3.2743, + "step": 10255 + }, + { + "epoch": 0.8222725541680864, + "grad_norm": 1.793933629989624, + "learning_rate": 1.330055821616344e-05, + "loss": 3.1749, + "step": 10256 + }, + { + "epoch": 0.822352728949109, + "grad_norm": 1.706440806388855, + "learning_rate": 1.3299332504421746e-05, + "loss": 3.1652, + "step": 10257 + }, + { + "epoch": 0.8224329037301317, + "grad_norm": 1.7085479497909546, + "learning_rate": 1.3298106737054468e-05, + "loss": 3.1958, + "step": 10258 + }, + { + "epoch": 0.8225130785111543, + "grad_norm": 1.8368818759918213, + "learning_rate": 1.3296880914082275e-05, + "loss": 3.3146, + "step": 10259 + }, + { + "epoch": 0.8225932532921769, + "grad_norm": 1.8103153705596924, + "learning_rate": 1.329565503552583e-05, + "loss": 3.2189, + "step": 10260 + }, + { + "epoch": 0.8226734280731995, + "grad_norm": 2.1826462745666504, + "learning_rate": 1.32944291014058e-05, + "loss": 3.1493, + "step": 10261 + }, + { + "epoch": 0.8227536028542222, + "grad_norm": 1.733607530593872, + "learning_rate": 1.3293203111742863e-05, + "loss": 3.1647, + "step": 10262 + }, + { + "epoch": 0.8228337776352448, + "grad_norm": 1.8638025522232056, + "learning_rate": 1.3291977066557677e-05, + "loss": 3.251, + "step": 10263 + }, + { + "epoch": 0.8229139524162674, + "grad_norm": 1.674248456954956, + "learning_rate": 1.3290750965870922e-05, + "loss": 3.0849, + "step": 10264 + }, + { + "epoch": 0.82299412719729, + "grad_norm": 1.7979726791381836, + "learning_rate": 1.3289524809703261e-05, + "loss": 3.2108, + "step": 10265 + }, + { + "epoch": 0.8230743019783128, + "grad_norm": 0.8405409455299377, + "learning_rate": 1.328829859807537e-05, + "loss": 2.4963, + "step": 10266 + }, + { + "epoch": 0.8231544767593354, + "grad_norm": 1.8235888481140137, + "learning_rate": 1.3287072331007932e-05, + "loss": 3.3047, + "step": 10267 + }, + { + "epoch": 0.823234651540358, + "grad_norm": 1.7390385866165161, + "learning_rate": 1.3285846008521604e-05, + "loss": 3.139, + "step": 10268 + }, + { + "epoch": 0.8233148263213806, + "grad_norm": 1.7916408777236938, + "learning_rate": 1.3284619630637075e-05, + "loss": 3.291, + "step": 10269 + }, + { + "epoch": 0.8233950011024033, + "grad_norm": 1.8446781635284424, + "learning_rate": 1.3283393197375016e-05, + "loss": 3.2419, + "step": 10270 + }, + { + "epoch": 0.8234751758834259, + "grad_norm": 1.8397616147994995, + "learning_rate": 1.3282166708756101e-05, + "loss": 3.1801, + "step": 10271 + }, + { + "epoch": 0.8235553506644485, + "grad_norm": 1.7628333568572998, + "learning_rate": 1.3280940164801017e-05, + "loss": 3.282, + "step": 10272 + }, + { + "epoch": 0.8236355254454711, + "grad_norm": 1.7092751264572144, + "learning_rate": 1.3279713565530436e-05, + "loss": 3.2743, + "step": 10273 + }, + { + "epoch": 0.8237157002264938, + "grad_norm": 1.8668639659881592, + "learning_rate": 1.3278486910965043e-05, + "loss": 3.1816, + "step": 10274 + }, + { + "epoch": 0.8237958750075164, + "grad_norm": 1.7550023794174194, + "learning_rate": 1.3277260201125513e-05, + "loss": 3.2213, + "step": 10275 + }, + { + "epoch": 0.823876049788539, + "grad_norm": 1.7484792470932007, + "learning_rate": 1.3276033436032529e-05, + "loss": 3.2128, + "step": 10276 + }, + { + "epoch": 0.8239562245695616, + "grad_norm": 1.886321783065796, + "learning_rate": 1.327480661570678e-05, + "loss": 3.2465, + "step": 10277 + }, + { + "epoch": 0.8240363993505843, + "grad_norm": 1.7306157350540161, + "learning_rate": 1.3273579740168945e-05, + "loss": 3.1513, + "step": 10278 + }, + { + "epoch": 0.8241165741316069, + "grad_norm": 1.7456315755844116, + "learning_rate": 1.327235280943971e-05, + "loss": 3.1644, + "step": 10279 + }, + { + "epoch": 0.8241967489126295, + "grad_norm": 0.7587887644767761, + "learning_rate": 1.3271125823539757e-05, + "loss": 2.4751, + "step": 10280 + }, + { + "epoch": 0.8242769236936521, + "grad_norm": 1.9147156476974487, + "learning_rate": 1.3269898782489776e-05, + "loss": 3.1975, + "step": 10281 + }, + { + "epoch": 0.8243570984746748, + "grad_norm": 1.7052381038665771, + "learning_rate": 1.3268671686310458e-05, + "loss": 3.0897, + "step": 10282 + }, + { + "epoch": 0.8244372732556974, + "grad_norm": 1.7548375129699707, + "learning_rate": 1.3267444535022482e-05, + "loss": 3.2446, + "step": 10283 + }, + { + "epoch": 0.82451744803672, + "grad_norm": 1.7369067668914795, + "learning_rate": 1.3266217328646547e-05, + "loss": 3.259, + "step": 10284 + }, + { + "epoch": 0.8245976228177426, + "grad_norm": 1.760448932647705, + "learning_rate": 1.3264990067203337e-05, + "loss": 3.2045, + "step": 10285 + }, + { + "epoch": 0.8246777975987654, + "grad_norm": 1.7461278438568115, + "learning_rate": 1.3263762750713542e-05, + "loss": 3.1366, + "step": 10286 + }, + { + "epoch": 0.824757972379788, + "grad_norm": 1.6854841709136963, + "learning_rate": 1.3262535379197863e-05, + "loss": 3.1537, + "step": 10287 + }, + { + "epoch": 0.8248381471608106, + "grad_norm": 0.791925847530365, + "learning_rate": 1.3261307952676982e-05, + "loss": 2.4825, + "step": 10288 + }, + { + "epoch": 0.8249183219418332, + "grad_norm": 1.7974135875701904, + "learning_rate": 1.3260080471171603e-05, + "loss": 3.2705, + "step": 10289 + }, + { + "epoch": 0.8249984967228559, + "grad_norm": 1.7616828680038452, + "learning_rate": 1.3258852934702416e-05, + "loss": 3.1433, + "step": 10290 + }, + { + "epoch": 0.8250786715038785, + "grad_norm": 2.17671537399292, + "learning_rate": 1.3257625343290112e-05, + "loss": 3.2057, + "step": 10291 + }, + { + "epoch": 0.8251588462849011, + "grad_norm": 1.8021888732910156, + "learning_rate": 1.3256397696955398e-05, + "loss": 3.1749, + "step": 10292 + }, + { + "epoch": 0.8252390210659237, + "grad_norm": 1.7849087715148926, + "learning_rate": 1.3255169995718964e-05, + "loss": 3.2878, + "step": 10293 + }, + { + "epoch": 0.8253191958469464, + "grad_norm": 1.8409487009048462, + "learning_rate": 1.3253942239601511e-05, + "loss": 3.2592, + "step": 10294 + }, + { + "epoch": 0.825399370627969, + "grad_norm": 1.7179268598556519, + "learning_rate": 1.3252714428623741e-05, + "loss": 3.1564, + "step": 10295 + }, + { + "epoch": 0.8254795454089916, + "grad_norm": 1.8089423179626465, + "learning_rate": 1.3251486562806348e-05, + "loss": 3.2786, + "step": 10296 + }, + { + "epoch": 0.8255597201900142, + "grad_norm": 1.8323432207107544, + "learning_rate": 1.3250258642170042e-05, + "loss": 3.1823, + "step": 10297 + }, + { + "epoch": 0.8256398949710368, + "grad_norm": 1.7882057428359985, + "learning_rate": 1.3249030666735518e-05, + "loss": 3.3362, + "step": 10298 + }, + { + "epoch": 0.8257200697520595, + "grad_norm": 1.833498477935791, + "learning_rate": 1.3247802636523484e-05, + "loss": 3.1778, + "step": 10299 + }, + { + "epoch": 0.8258002445330821, + "grad_norm": 1.8899059295654297, + "learning_rate": 1.3246574551554641e-05, + "loss": 3.2349, + "step": 10300 + }, + { + "epoch": 0.8258804193141047, + "grad_norm": 0.7202012538909912, + "learning_rate": 1.3245346411849693e-05, + "loss": 2.496, + "step": 10301 + }, + { + "epoch": 0.8259605940951273, + "grad_norm": 1.8273143768310547, + "learning_rate": 1.3244118217429351e-05, + "loss": 3.2462, + "step": 10302 + }, + { + "epoch": 0.82604076887615, + "grad_norm": 1.816552996635437, + "learning_rate": 1.324288996831432e-05, + "loss": 3.2511, + "step": 10303 + }, + { + "epoch": 0.8261209436571726, + "grad_norm": 1.776590347290039, + "learning_rate": 1.3241661664525305e-05, + "loss": 3.1865, + "step": 10304 + }, + { + "epoch": 0.8262011184381952, + "grad_norm": 1.7364685535430908, + "learning_rate": 1.3240433306083017e-05, + "loss": 3.3081, + "step": 10305 + }, + { + "epoch": 0.826281293219218, + "grad_norm": 1.8206878900527954, + "learning_rate": 1.3239204893008165e-05, + "loss": 3.1303, + "step": 10306 + }, + { + "epoch": 0.8263614680002406, + "grad_norm": 1.754115343093872, + "learning_rate": 1.3237976425321464e-05, + "loss": 3.2437, + "step": 10307 + }, + { + "epoch": 0.8264416427812632, + "grad_norm": 1.7559198141098022, + "learning_rate": 1.3236747903043616e-05, + "loss": 3.2383, + "step": 10308 + }, + { + "epoch": 0.8265218175622858, + "grad_norm": 1.7283610105514526, + "learning_rate": 1.3235519326195345e-05, + "loss": 3.1789, + "step": 10309 + }, + { + "epoch": 0.8266019923433084, + "grad_norm": 1.7712600231170654, + "learning_rate": 1.3234290694797354e-05, + "loss": 3.2509, + "step": 10310 + }, + { + "epoch": 0.8266821671243311, + "grad_norm": 1.812238097190857, + "learning_rate": 1.3233062008870361e-05, + "loss": 3.1358, + "step": 10311 + }, + { + "epoch": 0.8267623419053537, + "grad_norm": 1.8417974710464478, + "learning_rate": 1.3231833268435086e-05, + "loss": 3.2706, + "step": 10312 + }, + { + "epoch": 0.8268425166863763, + "grad_norm": 1.7120153903961182, + "learning_rate": 1.3230604473512239e-05, + "loss": 3.2341, + "step": 10313 + }, + { + "epoch": 0.8269226914673989, + "grad_norm": 1.8841991424560547, + "learning_rate": 1.322937562412254e-05, + "loss": 3.2533, + "step": 10314 + }, + { + "epoch": 0.8270028662484216, + "grad_norm": 1.8443421125411987, + "learning_rate": 1.3228146720286705e-05, + "loss": 3.2679, + "step": 10315 + }, + { + "epoch": 0.8270830410294442, + "grad_norm": 0.673017680644989, + "learning_rate": 1.3226917762025453e-05, + "loss": 2.4745, + "step": 10316 + }, + { + "epoch": 0.8271632158104668, + "grad_norm": 1.7156426906585693, + "learning_rate": 1.322568874935951e-05, + "loss": 3.2985, + "step": 10317 + }, + { + "epoch": 0.8272433905914894, + "grad_norm": 1.7067418098449707, + "learning_rate": 1.3224459682309586e-05, + "loss": 3.2037, + "step": 10318 + }, + { + "epoch": 0.8273235653725121, + "grad_norm": 1.6980587244033813, + "learning_rate": 1.322323056089641e-05, + "loss": 3.1908, + "step": 10319 + }, + { + "epoch": 0.8274037401535347, + "grad_norm": 1.7271226644515991, + "learning_rate": 1.3222001385140702e-05, + "loss": 3.2332, + "step": 10320 + }, + { + "epoch": 0.8274839149345573, + "grad_norm": 1.9008880853652954, + "learning_rate": 1.3220772155063187e-05, + "loss": 3.1315, + "step": 10321 + }, + { + "epoch": 0.8275640897155799, + "grad_norm": 1.681239128112793, + "learning_rate": 1.321954287068459e-05, + "loss": 3.1403, + "step": 10322 + }, + { + "epoch": 0.8276442644966026, + "grad_norm": 1.795330286026001, + "learning_rate": 1.3218313532025633e-05, + "loss": 3.2201, + "step": 10323 + }, + { + "epoch": 0.8277244392776252, + "grad_norm": 1.8210068941116333, + "learning_rate": 1.3217084139107044e-05, + "loss": 3.2537, + "step": 10324 + }, + { + "epoch": 0.8278046140586478, + "grad_norm": 1.7808783054351807, + "learning_rate": 1.321585469194955e-05, + "loss": 3.1981, + "step": 10325 + }, + { + "epoch": 0.8278847888396704, + "grad_norm": 1.8350038528442383, + "learning_rate": 1.3214625190573879e-05, + "loss": 3.2396, + "step": 10326 + }, + { + "epoch": 0.8279649636206932, + "grad_norm": 1.8622945547103882, + "learning_rate": 1.3213395635000762e-05, + "loss": 3.1735, + "step": 10327 + }, + { + "epoch": 0.8280451384017158, + "grad_norm": 1.8001269102096558, + "learning_rate": 1.3212166025250925e-05, + "loss": 3.116, + "step": 10328 + }, + { + "epoch": 0.8281253131827384, + "grad_norm": 1.7964941263198853, + "learning_rate": 1.3210936361345103e-05, + "loss": 3.1332, + "step": 10329 + }, + { + "epoch": 0.828205487963761, + "grad_norm": 1.8200181722640991, + "learning_rate": 1.3209706643304025e-05, + "loss": 3.1445, + "step": 10330 + }, + { + "epoch": 0.8282856627447837, + "grad_norm": 1.9540799856185913, + "learning_rate": 1.3208476871148424e-05, + "loss": 3.2248, + "step": 10331 + }, + { + "epoch": 0.8283658375258063, + "grad_norm": 1.8183395862579346, + "learning_rate": 1.3207247044899034e-05, + "loss": 3.2355, + "step": 10332 + }, + { + "epoch": 0.8284460123068289, + "grad_norm": 1.9065266847610474, + "learning_rate": 1.3206017164576587e-05, + "loss": 3.3016, + "step": 10333 + }, + { + "epoch": 0.8285261870878515, + "grad_norm": 1.742263913154602, + "learning_rate": 1.3204787230201826e-05, + "loss": 3.1277, + "step": 10334 + }, + { + "epoch": 0.8286063618688742, + "grad_norm": 1.8092083930969238, + "learning_rate": 1.3203557241795478e-05, + "loss": 3.1765, + "step": 10335 + }, + { + "epoch": 0.8286865366498968, + "grad_norm": 1.710861325263977, + "learning_rate": 1.3202327199378282e-05, + "loss": 3.2025, + "step": 10336 + }, + { + "epoch": 0.8287667114309194, + "grad_norm": 1.738517165184021, + "learning_rate": 1.3201097102970981e-05, + "loss": 3.2803, + "step": 10337 + }, + { + "epoch": 0.828846886211942, + "grad_norm": 1.8509840965270996, + "learning_rate": 1.3199866952594309e-05, + "loss": 3.2395, + "step": 10338 + }, + { + "epoch": 0.8289270609929646, + "grad_norm": 1.7868818044662476, + "learning_rate": 1.3198636748269008e-05, + "loss": 3.2241, + "step": 10339 + }, + { + "epoch": 0.8290072357739873, + "grad_norm": 1.771043062210083, + "learning_rate": 1.3197406490015822e-05, + "loss": 3.2838, + "step": 10340 + }, + { + "epoch": 0.8290874105550099, + "grad_norm": 1.6649659872055054, + "learning_rate": 1.3196176177855486e-05, + "loss": 3.0347, + "step": 10341 + }, + { + "epoch": 0.8291675853360325, + "grad_norm": 1.8366813659667969, + "learning_rate": 1.3194945811808748e-05, + "loss": 3.1583, + "step": 10342 + }, + { + "epoch": 0.8292477601170551, + "grad_norm": 1.9611656665802002, + "learning_rate": 1.3193715391896349e-05, + "loss": 3.2378, + "step": 10343 + }, + { + "epoch": 0.8293279348980778, + "grad_norm": 1.8356150388717651, + "learning_rate": 1.3192484918139035e-05, + "loss": 3.1284, + "step": 10344 + }, + { + "epoch": 0.8294081096791004, + "grad_norm": 1.7739567756652832, + "learning_rate": 1.3191254390557549e-05, + "loss": 3.2621, + "step": 10345 + }, + { + "epoch": 0.829488284460123, + "grad_norm": 1.8154524564743042, + "learning_rate": 1.3190023809172638e-05, + "loss": 3.2146, + "step": 10346 + }, + { + "epoch": 0.8295684592411458, + "grad_norm": 1.767926573753357, + "learning_rate": 1.3188793174005052e-05, + "loss": 3.2328, + "step": 10347 + }, + { + "epoch": 0.8296486340221684, + "grad_norm": 1.8143131732940674, + "learning_rate": 1.3187562485075535e-05, + "loss": 3.1949, + "step": 10348 + }, + { + "epoch": 0.829728808803191, + "grad_norm": 1.9343863725662231, + "learning_rate": 1.3186331742404842e-05, + "loss": 3.0165, + "step": 10349 + }, + { + "epoch": 0.8298089835842136, + "grad_norm": 1.9208911657333374, + "learning_rate": 1.3185100946013711e-05, + "loss": 3.2823, + "step": 10350 + }, + { + "epoch": 0.8298891583652362, + "grad_norm": 1.7399245500564575, + "learning_rate": 1.3183870095922906e-05, + "loss": 3.2224, + "step": 10351 + }, + { + "epoch": 0.8299693331462589, + "grad_norm": 1.911638855934143, + "learning_rate": 1.3182639192153173e-05, + "loss": 3.2057, + "step": 10352 + }, + { + "epoch": 0.8300495079272815, + "grad_norm": 1.9498826265335083, + "learning_rate": 1.3181408234725263e-05, + "loss": 3.1551, + "step": 10353 + }, + { + "epoch": 0.8301296827083041, + "grad_norm": 1.7654355764389038, + "learning_rate": 1.3180177223659936e-05, + "loss": 3.2006, + "step": 10354 + }, + { + "epoch": 0.8302098574893267, + "grad_norm": 1.8524508476257324, + "learning_rate": 1.3178946158977936e-05, + "loss": 3.3278, + "step": 10355 + }, + { + "epoch": 0.8302900322703494, + "grad_norm": 1.775060772895813, + "learning_rate": 1.3177715040700028e-05, + "loss": 3.1619, + "step": 10356 + }, + { + "epoch": 0.830370207051372, + "grad_norm": 1.7487515211105347, + "learning_rate": 1.3176483868846963e-05, + "loss": 3.2607, + "step": 10357 + }, + { + "epoch": 0.8304503818323946, + "grad_norm": 1.7933717966079712, + "learning_rate": 1.3175252643439498e-05, + "loss": 3.1903, + "step": 10358 + }, + { + "epoch": 0.8305305566134172, + "grad_norm": 1.780483603477478, + "learning_rate": 1.3174021364498395e-05, + "loss": 3.2307, + "step": 10359 + }, + { + "epoch": 0.8306107313944399, + "grad_norm": 1.7153079509735107, + "learning_rate": 1.317279003204441e-05, + "loss": 3.0177, + "step": 10360 + }, + { + "epoch": 0.8306909061754625, + "grad_norm": 1.846078634262085, + "learning_rate": 1.3171558646098302e-05, + "loss": 3.1599, + "step": 10361 + }, + { + "epoch": 0.8307710809564851, + "grad_norm": 1.6366581916809082, + "learning_rate": 1.3170327206680835e-05, + "loss": 3.1918, + "step": 10362 + }, + { + "epoch": 0.8308512557375077, + "grad_norm": 1.681844711303711, + "learning_rate": 1.3169095713812764e-05, + "loss": 3.1328, + "step": 10363 + }, + { + "epoch": 0.8309314305185304, + "grad_norm": 1.803837776184082, + "learning_rate": 1.3167864167514863e-05, + "loss": 3.1825, + "step": 10364 + }, + { + "epoch": 0.831011605299553, + "grad_norm": 1.8113783597946167, + "learning_rate": 1.3166632567807885e-05, + "loss": 3.1598, + "step": 10365 + }, + { + "epoch": 0.8310917800805756, + "grad_norm": 1.8518810272216797, + "learning_rate": 1.3165400914712597e-05, + "loss": 3.2646, + "step": 10366 + }, + { + "epoch": 0.8311719548615983, + "grad_norm": 1.767613410949707, + "learning_rate": 1.3164169208249766e-05, + "loss": 3.3193, + "step": 10367 + }, + { + "epoch": 0.831252129642621, + "grad_norm": 1.8062437772750854, + "learning_rate": 1.3162937448440159e-05, + "loss": 3.1622, + "step": 10368 + }, + { + "epoch": 0.8313323044236436, + "grad_norm": 1.7898967266082764, + "learning_rate": 1.316170563530454e-05, + "loss": 3.2272, + "step": 10369 + }, + { + "epoch": 0.8314124792046662, + "grad_norm": 1.8387733697891235, + "learning_rate": 1.3160473768863678e-05, + "loss": 3.1577, + "step": 10370 + }, + { + "epoch": 0.8314926539856888, + "grad_norm": 1.825901746749878, + "learning_rate": 1.315924184913834e-05, + "loss": 3.2347, + "step": 10371 + }, + { + "epoch": 0.8315728287667115, + "grad_norm": 1.746853232383728, + "learning_rate": 1.3158009876149303e-05, + "loss": 3.187, + "step": 10372 + }, + { + "epoch": 0.8316530035477341, + "grad_norm": 1.8842921257019043, + "learning_rate": 1.3156777849917326e-05, + "loss": 3.2815, + "step": 10373 + }, + { + "epoch": 0.8317331783287567, + "grad_norm": 1.8233444690704346, + "learning_rate": 1.3155545770463194e-05, + "loss": 3.2422, + "step": 10374 + }, + { + "epoch": 0.8318133531097793, + "grad_norm": 1.8718984127044678, + "learning_rate": 1.3154313637807667e-05, + "loss": 3.1749, + "step": 10375 + }, + { + "epoch": 0.831893527890802, + "grad_norm": 1.6727577447891235, + "learning_rate": 1.3153081451971525e-05, + "loss": 3.188, + "step": 10376 + }, + { + "epoch": 0.8319737026718246, + "grad_norm": 1.928215742111206, + "learning_rate": 1.3151849212975542e-05, + "loss": 3.3754, + "step": 10377 + }, + { + "epoch": 0.8320538774528472, + "grad_norm": 1.740201473236084, + "learning_rate": 1.3150616920840492e-05, + "loss": 3.2595, + "step": 10378 + }, + { + "epoch": 0.8321340522338698, + "grad_norm": 1.7360577583312988, + "learning_rate": 1.3149384575587153e-05, + "loss": 3.16, + "step": 10379 + }, + { + "epoch": 0.8322142270148924, + "grad_norm": 1.6965996026992798, + "learning_rate": 1.3148152177236297e-05, + "loss": 3.1952, + "step": 10380 + }, + { + "epoch": 0.8322944017959151, + "grad_norm": 1.7093913555145264, + "learning_rate": 1.3146919725808705e-05, + "loss": 3.1246, + "step": 10381 + }, + { + "epoch": 0.8323745765769377, + "grad_norm": 0.752907931804657, + "learning_rate": 1.3145687221325157e-05, + "loss": 2.4882, + "step": 10382 + }, + { + "epoch": 0.8324547513579603, + "grad_norm": 1.8364105224609375, + "learning_rate": 1.314445466380643e-05, + "loss": 3.2549, + "step": 10383 + }, + { + "epoch": 0.832534926138983, + "grad_norm": 1.7921116352081299, + "learning_rate": 1.3143222053273305e-05, + "loss": 3.2122, + "step": 10384 + }, + { + "epoch": 0.8326151009200056, + "grad_norm": 1.817313551902771, + "learning_rate": 1.3141989389746565e-05, + "loss": 3.1012, + "step": 10385 + }, + { + "epoch": 0.8326952757010282, + "grad_norm": 1.8046072721481323, + "learning_rate": 1.314075667324699e-05, + "loss": 3.147, + "step": 10386 + }, + { + "epoch": 0.8327754504820508, + "grad_norm": 1.7776014804840088, + "learning_rate": 1.3139523903795367e-05, + "loss": 3.2174, + "step": 10387 + }, + { + "epoch": 0.8328556252630736, + "grad_norm": 1.7515208721160889, + "learning_rate": 1.3138291081412475e-05, + "loss": 3.2791, + "step": 10388 + }, + { + "epoch": 0.8329358000440962, + "grad_norm": 1.8314770460128784, + "learning_rate": 1.3137058206119105e-05, + "loss": 3.1207, + "step": 10389 + }, + { + "epoch": 0.8330159748251188, + "grad_norm": 1.7973418235778809, + "learning_rate": 1.3135825277936034e-05, + "loss": 3.1852, + "step": 10390 + }, + { + "epoch": 0.8330961496061414, + "grad_norm": 1.9134995937347412, + "learning_rate": 1.3134592296884056e-05, + "loss": 3.2472, + "step": 10391 + }, + { + "epoch": 0.833176324387164, + "grad_norm": 1.8208755254745483, + "learning_rate": 1.313335926298396e-05, + "loss": 3.3275, + "step": 10392 + }, + { + "epoch": 0.8332564991681867, + "grad_norm": 1.8358323574066162, + "learning_rate": 1.3132126176256528e-05, + "loss": 3.2563, + "step": 10393 + }, + { + "epoch": 0.8333366739492093, + "grad_norm": 1.709758996963501, + "learning_rate": 1.3130893036722553e-05, + "loss": 3.271, + "step": 10394 + }, + { + "epoch": 0.8334168487302319, + "grad_norm": 1.8139888048171997, + "learning_rate": 1.3129659844402824e-05, + "loss": 3.1957, + "step": 10395 + }, + { + "epoch": 0.8334970235112545, + "grad_norm": 1.8766331672668457, + "learning_rate": 1.3128426599318134e-05, + "loss": 3.2749, + "step": 10396 + }, + { + "epoch": 0.8335771982922772, + "grad_norm": 1.7968614101409912, + "learning_rate": 1.3127193301489275e-05, + "loss": 3.3459, + "step": 10397 + }, + { + "epoch": 0.8336573730732998, + "grad_norm": 1.8461560010910034, + "learning_rate": 1.3125959950937034e-05, + "loss": 3.217, + "step": 10398 + }, + { + "epoch": 0.8337375478543224, + "grad_norm": 1.8679425716400146, + "learning_rate": 1.3124726547682216e-05, + "loss": 3.2297, + "step": 10399 + }, + { + "epoch": 0.833817722635345, + "grad_norm": 1.7574182748794556, + "learning_rate": 1.312349309174561e-05, + "loss": 3.1409, + "step": 10400 + }, + { + "epoch": 0.8338978974163677, + "grad_norm": 1.8337301015853882, + "learning_rate": 1.3122259583148007e-05, + "loss": 3.2007, + "step": 10401 + }, + { + "epoch": 0.8339780721973903, + "grad_norm": 0.7462592720985413, + "learning_rate": 1.3121026021910211e-05, + "loss": 2.4746, + "step": 10402 + }, + { + "epoch": 0.8340582469784129, + "grad_norm": 1.7591936588287354, + "learning_rate": 1.3119792408053014e-05, + "loss": 3.2448, + "step": 10403 + }, + { + "epoch": 0.8341384217594355, + "grad_norm": 1.8236805200576782, + "learning_rate": 1.3118558741597219e-05, + "loss": 3.1959, + "step": 10404 + }, + { + "epoch": 0.8342185965404582, + "grad_norm": 1.7833260297775269, + "learning_rate": 1.3117325022563618e-05, + "loss": 3.1819, + "step": 10405 + }, + { + "epoch": 0.8342987713214808, + "grad_norm": 1.7234477996826172, + "learning_rate": 1.3116091250973018e-05, + "loss": 3.2019, + "step": 10406 + }, + { + "epoch": 0.8343789461025034, + "grad_norm": 1.9783190488815308, + "learning_rate": 1.3114857426846222e-05, + "loss": 3.1254, + "step": 10407 + }, + { + "epoch": 0.8344591208835261, + "grad_norm": 1.7625666856765747, + "learning_rate": 1.3113623550204022e-05, + "loss": 3.2074, + "step": 10408 + }, + { + "epoch": 0.8345392956645488, + "grad_norm": 0.6992250680923462, + "learning_rate": 1.311238962106723e-05, + "loss": 2.4829, + "step": 10409 + }, + { + "epoch": 0.8346194704455714, + "grad_norm": 1.8455530405044556, + "learning_rate": 1.3111155639456644e-05, + "loss": 3.2744, + "step": 10410 + }, + { + "epoch": 0.834699645226594, + "grad_norm": 1.8214621543884277, + "learning_rate": 1.310992160539307e-05, + "loss": 3.1317, + "step": 10411 + }, + { + "epoch": 0.8347798200076166, + "grad_norm": 1.8627252578735352, + "learning_rate": 1.3108687518897316e-05, + "loss": 3.1923, + "step": 10412 + }, + { + "epoch": 0.8348599947886393, + "grad_norm": 1.7686409950256348, + "learning_rate": 1.3107453379990182e-05, + "loss": 3.2075, + "step": 10413 + }, + { + "epoch": 0.8349401695696619, + "grad_norm": 1.8587814569473267, + "learning_rate": 1.3106219188692481e-05, + "loss": 3.17, + "step": 10414 + }, + { + "epoch": 0.8350203443506845, + "grad_norm": 0.7084640264511108, + "learning_rate": 1.310498494502502e-05, + "loss": 2.4943, + "step": 10415 + }, + { + "epoch": 0.8351005191317071, + "grad_norm": 1.8419774770736694, + "learning_rate": 1.3103750649008609e-05, + "loss": 3.2573, + "step": 10416 + }, + { + "epoch": 0.8351806939127298, + "grad_norm": 1.8376809358596802, + "learning_rate": 1.3102516300664054e-05, + "loss": 3.1534, + "step": 10417 + }, + { + "epoch": 0.8352608686937524, + "grad_norm": 1.7808048725128174, + "learning_rate": 1.3101281900012169e-05, + "loss": 3.201, + "step": 10418 + }, + { + "epoch": 0.835341043474775, + "grad_norm": 1.7473224401474, + "learning_rate": 1.3100047447073761e-05, + "loss": 3.146, + "step": 10419 + }, + { + "epoch": 0.8354212182557976, + "grad_norm": 1.7692426443099976, + "learning_rate": 1.3098812941869648e-05, + "loss": 3.2429, + "step": 10420 + }, + { + "epoch": 0.8355013930368203, + "grad_norm": 1.7941954135894775, + "learning_rate": 1.309757838442064e-05, + "loss": 3.2778, + "step": 10421 + }, + { + "epoch": 0.8355815678178429, + "grad_norm": 1.8363292217254639, + "learning_rate": 1.3096343774747553e-05, + "loss": 3.1846, + "step": 10422 + }, + { + "epoch": 0.8356617425988655, + "grad_norm": 1.7188448905944824, + "learning_rate": 1.30951091128712e-05, + "loss": 3.1884, + "step": 10423 + }, + { + "epoch": 0.8357419173798881, + "grad_norm": 1.758681297302246, + "learning_rate": 1.30938743988124e-05, + "loss": 3.1519, + "step": 10424 + }, + { + "epoch": 0.8358220921609107, + "grad_norm": 1.713270902633667, + "learning_rate": 1.3092639632591966e-05, + "loss": 3.0062, + "step": 10425 + }, + { + "epoch": 0.8359022669419334, + "grad_norm": 1.7744159698486328, + "learning_rate": 1.3091404814230718e-05, + "loss": 3.1523, + "step": 10426 + }, + { + "epoch": 0.835982441722956, + "grad_norm": 1.7024645805358887, + "learning_rate": 1.3090169943749475e-05, + "loss": 3.2372, + "step": 10427 + }, + { + "epoch": 0.8360626165039787, + "grad_norm": 1.7403267621994019, + "learning_rate": 1.3088935021169056e-05, + "loss": 3.2025, + "step": 10428 + }, + { + "epoch": 0.8361427912850014, + "grad_norm": 1.7709890604019165, + "learning_rate": 1.3087700046510281e-05, + "loss": 3.1771, + "step": 10429 + }, + { + "epoch": 0.836222966066024, + "grad_norm": 1.8793175220489502, + "learning_rate": 1.3086465019793973e-05, + "loss": 3.3102, + "step": 10430 + }, + { + "epoch": 0.8363031408470466, + "grad_norm": 5.787628650665283, + "learning_rate": 1.3085229941040949e-05, + "loss": 3.2482, + "step": 10431 + }, + { + "epoch": 0.8363833156280692, + "grad_norm": 1.7589634656906128, + "learning_rate": 1.3083994810272038e-05, + "loss": 3.2927, + "step": 10432 + }, + { + "epoch": 0.8364634904090918, + "grad_norm": 1.8509204387664795, + "learning_rate": 1.308275962750806e-05, + "loss": 3.2172, + "step": 10433 + }, + { + "epoch": 0.8365436651901145, + "grad_norm": 2.0233709812164307, + "learning_rate": 1.3081524392769842e-05, + "loss": 3.1941, + "step": 10434 + }, + { + "epoch": 0.8366238399711371, + "grad_norm": 1.892857313156128, + "learning_rate": 1.308028910607821e-05, + "loss": 3.2563, + "step": 10435 + }, + { + "epoch": 0.8367040147521597, + "grad_norm": 1.824358582496643, + "learning_rate": 1.3079053767453986e-05, + "loss": 3.2357, + "step": 10436 + }, + { + "epoch": 0.8367841895331823, + "grad_norm": 1.7380530834197998, + "learning_rate": 1.3077818376918006e-05, + "loss": 3.1443, + "step": 10437 + }, + { + "epoch": 0.836864364314205, + "grad_norm": 0.7217404842376709, + "learning_rate": 1.3076582934491086e-05, + "loss": 2.5002, + "step": 10438 + }, + { + "epoch": 0.8369445390952276, + "grad_norm": 1.7667793035507202, + "learning_rate": 1.3075347440194066e-05, + "loss": 3.1538, + "step": 10439 + }, + { + "epoch": 0.8370247138762502, + "grad_norm": 1.8524260520935059, + "learning_rate": 1.3074111894047774e-05, + "loss": 3.2734, + "step": 10440 + }, + { + "epoch": 0.8371048886572728, + "grad_norm": 1.825446367263794, + "learning_rate": 1.3072876296073034e-05, + "loss": 3.2296, + "step": 10441 + }, + { + "epoch": 0.8371850634382955, + "grad_norm": 1.7851219177246094, + "learning_rate": 1.3071640646290687e-05, + "loss": 3.1486, + "step": 10442 + }, + { + "epoch": 0.8372652382193181, + "grad_norm": 1.7236931324005127, + "learning_rate": 1.3070404944721559e-05, + "loss": 3.1525, + "step": 10443 + }, + { + "epoch": 0.8373454130003407, + "grad_norm": 1.9713691473007202, + "learning_rate": 1.3069169191386487e-05, + "loss": 3.2272, + "step": 10444 + }, + { + "epoch": 0.8374255877813633, + "grad_norm": 1.7588571310043335, + "learning_rate": 1.3067933386306302e-05, + "loss": 3.2604, + "step": 10445 + }, + { + "epoch": 0.837505762562386, + "grad_norm": 0.7060651779174805, + "learning_rate": 1.3066697529501844e-05, + "loss": 2.4884, + "step": 10446 + }, + { + "epoch": 0.8375859373434086, + "grad_norm": 1.6576956510543823, + "learning_rate": 1.3065461620993944e-05, + "loss": 3.14, + "step": 10447 + }, + { + "epoch": 0.8376661121244312, + "grad_norm": 1.7525347471237183, + "learning_rate": 1.3064225660803443e-05, + "loss": 3.2238, + "step": 10448 + }, + { + "epoch": 0.8377462869054539, + "grad_norm": 1.9459811449050903, + "learning_rate": 1.3062989648951177e-05, + "loss": 3.353, + "step": 10449 + }, + { + "epoch": 0.8378264616864766, + "grad_norm": 0.7567959427833557, + "learning_rate": 1.3061753585457986e-05, + "loss": 2.5, + "step": 10450 + }, + { + "epoch": 0.8379066364674992, + "grad_norm": 1.9717158079147339, + "learning_rate": 1.3060517470344705e-05, + "loss": 3.1444, + "step": 10451 + }, + { + "epoch": 0.8379868112485218, + "grad_norm": 1.7572646141052246, + "learning_rate": 1.3059281303632182e-05, + "loss": 3.2227, + "step": 10452 + }, + { + "epoch": 0.8380669860295444, + "grad_norm": 0.7082312107086182, + "learning_rate": 1.3058045085341254e-05, + "loss": 2.4867, + "step": 10453 + }, + { + "epoch": 0.8381471608105671, + "grad_norm": 1.9143122434616089, + "learning_rate": 1.3056808815492761e-05, + "loss": 3.176, + "step": 10454 + }, + { + "epoch": 0.8382273355915897, + "grad_norm": 1.866562843322754, + "learning_rate": 1.3055572494107552e-05, + "loss": 3.3188, + "step": 10455 + }, + { + "epoch": 0.8383075103726123, + "grad_norm": 1.9045464992523193, + "learning_rate": 1.3054336121206464e-05, + "loss": 3.1992, + "step": 10456 + }, + { + "epoch": 0.8383876851536349, + "grad_norm": 3.0341849327087402, + "learning_rate": 1.3053099696810349e-05, + "loss": 3.3289, + "step": 10457 + }, + { + "epoch": 0.8384678599346576, + "grad_norm": 1.8560584783554077, + "learning_rate": 1.3051863220940047e-05, + "loss": 3.2257, + "step": 10458 + }, + { + "epoch": 0.8385480347156802, + "grad_norm": 1.852615475654602, + "learning_rate": 1.3050626693616406e-05, + "loss": 3.2979, + "step": 10459 + }, + { + "epoch": 0.8386282094967028, + "grad_norm": 1.812638282775879, + "learning_rate": 1.3049390114860277e-05, + "loss": 3.231, + "step": 10460 + }, + { + "epoch": 0.8387083842777254, + "grad_norm": 1.785561203956604, + "learning_rate": 1.3048153484692501e-05, + "loss": 3.2181, + "step": 10461 + }, + { + "epoch": 0.838788559058748, + "grad_norm": 1.9576036930084229, + "learning_rate": 1.3046916803133936e-05, + "loss": 3.3073, + "step": 10462 + }, + { + "epoch": 0.8388687338397707, + "grad_norm": 1.8307238817214966, + "learning_rate": 1.3045680070205426e-05, + "loss": 3.0654, + "step": 10463 + }, + { + "epoch": 0.8389489086207933, + "grad_norm": 1.8636128902435303, + "learning_rate": 1.304444328592782e-05, + "loss": 3.1832, + "step": 10464 + }, + { + "epoch": 0.8390290834018159, + "grad_norm": 1.8849740028381348, + "learning_rate": 1.3043206450321979e-05, + "loss": 3.1304, + "step": 10465 + }, + { + "epoch": 0.8391092581828385, + "grad_norm": 0.7826863527297974, + "learning_rate": 1.3041969563408747e-05, + "loss": 2.4839, + "step": 10466 + }, + { + "epoch": 0.8391894329638612, + "grad_norm": 1.7531564235687256, + "learning_rate": 1.304073262520898e-05, + "loss": 3.2184, + "step": 10467 + }, + { + "epoch": 0.8392696077448838, + "grad_norm": 1.8516578674316406, + "learning_rate": 1.3039495635743534e-05, + "loss": 3.261, + "step": 10468 + }, + { + "epoch": 0.8393497825259065, + "grad_norm": 1.7822000980377197, + "learning_rate": 1.303825859503326e-05, + "loss": 3.2714, + "step": 10469 + }, + { + "epoch": 0.8394299573069292, + "grad_norm": 1.7131344079971313, + "learning_rate": 1.3037021503099019e-05, + "loss": 3.2284, + "step": 10470 + }, + { + "epoch": 0.8395101320879518, + "grad_norm": 1.7654024362564087, + "learning_rate": 1.3035784359961665e-05, + "loss": 3.1707, + "step": 10471 + }, + { + "epoch": 0.8395903068689744, + "grad_norm": 1.7836706638336182, + "learning_rate": 1.3034547165642057e-05, + "loss": 3.1641, + "step": 10472 + }, + { + "epoch": 0.839670481649997, + "grad_norm": 1.9261374473571777, + "learning_rate": 1.3033309920161055e-05, + "loss": 3.1425, + "step": 10473 + }, + { + "epoch": 0.8397506564310196, + "grad_norm": 1.7403525114059448, + "learning_rate": 1.3032072623539513e-05, + "loss": 3.1329, + "step": 10474 + }, + { + "epoch": 0.8398308312120423, + "grad_norm": 1.8187299966812134, + "learning_rate": 1.30308352757983e-05, + "loss": 3.1956, + "step": 10475 + }, + { + "epoch": 0.8399110059930649, + "grad_norm": 1.7951568365097046, + "learning_rate": 1.3029597876958269e-05, + "loss": 3.1761, + "step": 10476 + }, + { + "epoch": 0.8399911807740875, + "grad_norm": 1.9012417793273926, + "learning_rate": 1.3028360427040286e-05, + "loss": 3.2396, + "step": 10477 + }, + { + "epoch": 0.8400713555551101, + "grad_norm": 2.403045892715454, + "learning_rate": 1.3027122926065212e-05, + "loss": 3.25, + "step": 10478 + }, + { + "epoch": 0.8401515303361328, + "grad_norm": 1.8216818571090698, + "learning_rate": 1.3025885374053914e-05, + "loss": 3.2641, + "step": 10479 + }, + { + "epoch": 0.8402317051171554, + "grad_norm": 1.7333124876022339, + "learning_rate": 1.3024647771027257e-05, + "loss": 3.2644, + "step": 10480 + }, + { + "epoch": 0.840311879898178, + "grad_norm": 1.669063687324524, + "learning_rate": 1.3023410117006101e-05, + "loss": 3.1589, + "step": 10481 + }, + { + "epoch": 0.8403920546792006, + "grad_norm": 1.7665398120880127, + "learning_rate": 1.3022172412011319e-05, + "loss": 3.1669, + "step": 10482 + }, + { + "epoch": 0.8404722294602233, + "grad_norm": 1.8209892511367798, + "learning_rate": 1.3020934656063773e-05, + "loss": 3.1736, + "step": 10483 + }, + { + "epoch": 0.8405524042412459, + "grad_norm": 1.9122750759124756, + "learning_rate": 1.3019696849184334e-05, + "loss": 3.2518, + "step": 10484 + }, + { + "epoch": 0.8406325790222685, + "grad_norm": 1.718280553817749, + "learning_rate": 1.3018458991393871e-05, + "loss": 3.2299, + "step": 10485 + }, + { + "epoch": 0.8407127538032911, + "grad_norm": 1.9070210456848145, + "learning_rate": 1.3017221082713252e-05, + "loss": 3.1991, + "step": 10486 + }, + { + "epoch": 0.8407929285843138, + "grad_norm": 1.8001424074172974, + "learning_rate": 1.3015983123163351e-05, + "loss": 3.2795, + "step": 10487 + }, + { + "epoch": 0.8408731033653364, + "grad_norm": 1.732063889503479, + "learning_rate": 1.3014745112765034e-05, + "loss": 3.1929, + "step": 10488 + }, + { + "epoch": 0.8409532781463591, + "grad_norm": 1.830886721611023, + "learning_rate": 1.3013507051539177e-05, + "loss": 3.2474, + "step": 10489 + }, + { + "epoch": 0.8410334529273817, + "grad_norm": 1.8896180391311646, + "learning_rate": 1.3012268939506658e-05, + "loss": 3.2424, + "step": 10490 + }, + { + "epoch": 0.8411136277084044, + "grad_norm": 1.7524259090423584, + "learning_rate": 1.3011030776688345e-05, + "loss": 3.1495, + "step": 10491 + }, + { + "epoch": 0.841193802489427, + "grad_norm": 1.8054521083831787, + "learning_rate": 1.3009792563105113e-05, + "loss": 3.2625, + "step": 10492 + }, + { + "epoch": 0.8412739772704496, + "grad_norm": 1.827473759651184, + "learning_rate": 1.3008554298777838e-05, + "loss": 3.1377, + "step": 10493 + }, + { + "epoch": 0.8413541520514722, + "grad_norm": 1.8379480838775635, + "learning_rate": 1.3007315983727398e-05, + "loss": 3.2467, + "step": 10494 + }, + { + "epoch": 0.8414343268324949, + "grad_norm": 1.7046478986740112, + "learning_rate": 1.3006077617974673e-05, + "loss": 3.1777, + "step": 10495 + }, + { + "epoch": 0.8415145016135175, + "grad_norm": 1.7460700273513794, + "learning_rate": 1.3004839201540536e-05, + "loss": 3.213, + "step": 10496 + }, + { + "epoch": 0.8415946763945401, + "grad_norm": 1.9493759870529175, + "learning_rate": 1.3003600734445872e-05, + "loss": 3.2103, + "step": 10497 + }, + { + "epoch": 0.8416748511755627, + "grad_norm": 0.7454043626785278, + "learning_rate": 1.3002362216711555e-05, + "loss": 2.5104, + "step": 10498 + }, + { + "epoch": 0.8417550259565854, + "grad_norm": 1.8769772052764893, + "learning_rate": 1.3001123648358469e-05, + "loss": 3.1174, + "step": 10499 + }, + { + "epoch": 0.841835200737608, + "grad_norm": 1.8161323070526123, + "learning_rate": 1.29998850294075e-05, + "loss": 3.1962, + "step": 10500 + }, + { + "epoch": 0.8419153755186306, + "grad_norm": 1.727658748626709, + "learning_rate": 1.299864635987952e-05, + "loss": 3.1925, + "step": 10501 + }, + { + "epoch": 0.8419955502996532, + "grad_norm": 1.940073013305664, + "learning_rate": 1.2997407639795428e-05, + "loss": 3.2676, + "step": 10502 + }, + { + "epoch": 0.8420757250806759, + "grad_norm": 1.8801796436309814, + "learning_rate": 1.2996168869176091e-05, + "loss": 3.1385, + "step": 10503 + }, + { + "epoch": 0.8421558998616985, + "grad_norm": 0.7081619501113892, + "learning_rate": 1.2994930048042406e-05, + "loss": 2.4932, + "step": 10504 + }, + { + "epoch": 0.8422360746427211, + "grad_norm": 1.8530257940292358, + "learning_rate": 1.299369117641526e-05, + "loss": 3.2311, + "step": 10505 + }, + { + "epoch": 0.8423162494237437, + "grad_norm": 1.633365273475647, + "learning_rate": 1.299245225431553e-05, + "loss": 3.1897, + "step": 10506 + }, + { + "epoch": 0.8423964242047663, + "grad_norm": 1.786308765411377, + "learning_rate": 1.2991213281764115e-05, + "loss": 3.1686, + "step": 10507 + }, + { + "epoch": 0.842476598985789, + "grad_norm": 0.7184181809425354, + "learning_rate": 1.2989974258781894e-05, + "loss": 2.4744, + "step": 10508 + }, + { + "epoch": 0.8425567737668116, + "grad_norm": 1.8941755294799805, + "learning_rate": 1.298873518538976e-05, + "loss": 3.0948, + "step": 10509 + }, + { + "epoch": 0.8426369485478343, + "grad_norm": 1.8203691244125366, + "learning_rate": 1.2987496061608609e-05, + "loss": 3.2257, + "step": 10510 + }, + { + "epoch": 0.842717123328857, + "grad_norm": 0.7171919941902161, + "learning_rate": 1.2986256887459324e-05, + "loss": 2.4908, + "step": 10511 + }, + { + "epoch": 0.8427972981098796, + "grad_norm": 1.7887076139450073, + "learning_rate": 1.2985017662962803e-05, + "loss": 3.2783, + "step": 10512 + }, + { + "epoch": 0.8428774728909022, + "grad_norm": 1.7062811851501465, + "learning_rate": 1.2983778388139932e-05, + "loss": 3.1755, + "step": 10513 + }, + { + "epoch": 0.8429576476719248, + "grad_norm": 1.718727707862854, + "learning_rate": 1.298253906301161e-05, + "loss": 3.1155, + "step": 10514 + }, + { + "epoch": 0.8430378224529474, + "grad_norm": 1.822821021080017, + "learning_rate": 1.2981299687598732e-05, + "loss": 3.2047, + "step": 10515 + }, + { + "epoch": 0.8431179972339701, + "grad_norm": 1.8643004894256592, + "learning_rate": 1.2980060261922191e-05, + "loss": 3.2209, + "step": 10516 + }, + { + "epoch": 0.8431981720149927, + "grad_norm": 1.9049218893051147, + "learning_rate": 1.2978820786002887e-05, + "loss": 3.1311, + "step": 10517 + }, + { + "epoch": 0.8432783467960153, + "grad_norm": 1.7534847259521484, + "learning_rate": 1.2977581259861711e-05, + "loss": 3.153, + "step": 10518 + }, + { + "epoch": 0.843358521577038, + "grad_norm": 1.7577064037322998, + "learning_rate": 1.2976341683519564e-05, + "loss": 3.3402, + "step": 10519 + }, + { + "epoch": 0.8434386963580606, + "grad_norm": 1.8191683292388916, + "learning_rate": 1.2975102056997348e-05, + "loss": 3.1088, + "step": 10520 + }, + { + "epoch": 0.8435188711390832, + "grad_norm": 1.8746147155761719, + "learning_rate": 1.2973862380315958e-05, + "loss": 3.2275, + "step": 10521 + }, + { + "epoch": 0.8435990459201058, + "grad_norm": 1.8209025859832764, + "learning_rate": 1.2972622653496299e-05, + "loss": 3.2021, + "step": 10522 + }, + { + "epoch": 0.8436792207011284, + "grad_norm": 1.9493573904037476, + "learning_rate": 1.2971382876559267e-05, + "loss": 3.2877, + "step": 10523 + }, + { + "epoch": 0.8437593954821511, + "grad_norm": 1.7128221988677979, + "learning_rate": 1.2970143049525765e-05, + "loss": 3.1578, + "step": 10524 + }, + { + "epoch": 0.8438395702631737, + "grad_norm": 1.7982580661773682, + "learning_rate": 1.2968903172416703e-05, + "loss": 3.1493, + "step": 10525 + }, + { + "epoch": 0.8439197450441963, + "grad_norm": 1.872865915298462, + "learning_rate": 1.2967663245252977e-05, + "loss": 3.0897, + "step": 10526 + }, + { + "epoch": 0.8439999198252189, + "grad_norm": 1.8118840456008911, + "learning_rate": 1.2966423268055497e-05, + "loss": 3.2347, + "step": 10527 + }, + { + "epoch": 0.8440800946062416, + "grad_norm": 2.0813663005828857, + "learning_rate": 1.2965183240845163e-05, + "loss": 3.1912, + "step": 10528 + }, + { + "epoch": 0.8441602693872642, + "grad_norm": 1.713901162147522, + "learning_rate": 1.2963943163642885e-05, + "loss": 3.2152, + "step": 10529 + }, + { + "epoch": 0.8442404441682869, + "grad_norm": 1.9111201763153076, + "learning_rate": 1.2962703036469573e-05, + "loss": 3.1991, + "step": 10530 + }, + { + "epoch": 0.8443206189493095, + "grad_norm": 1.848634958267212, + "learning_rate": 1.2961462859346129e-05, + "loss": 3.2831, + "step": 10531 + }, + { + "epoch": 0.8444007937303322, + "grad_norm": 1.7585012912750244, + "learning_rate": 1.2960222632293465e-05, + "loss": 3.2432, + "step": 10532 + }, + { + "epoch": 0.8444809685113548, + "grad_norm": 0.7317865490913391, + "learning_rate": 1.295898235533249e-05, + "loss": 2.5034, + "step": 10533 + }, + { + "epoch": 0.8445611432923774, + "grad_norm": 1.8254058361053467, + "learning_rate": 1.2957742028484117e-05, + "loss": 3.1239, + "step": 10534 + }, + { + "epoch": 0.8446413180734, + "grad_norm": 1.8331950902938843, + "learning_rate": 1.2956501651769256e-05, + "loss": 3.2274, + "step": 10535 + }, + { + "epoch": 0.8447214928544227, + "grad_norm": 1.9414864778518677, + "learning_rate": 1.2955261225208818e-05, + "loss": 3.1886, + "step": 10536 + }, + { + "epoch": 0.8448016676354453, + "grad_norm": 1.849496603012085, + "learning_rate": 1.2954020748823719e-05, + "loss": 3.1951, + "step": 10537 + }, + { + "epoch": 0.8448818424164679, + "grad_norm": 1.7707651853561401, + "learning_rate": 1.295278022263487e-05, + "loss": 3.2013, + "step": 10538 + }, + { + "epoch": 0.8449620171974905, + "grad_norm": 2.027780055999756, + "learning_rate": 1.2951539646663185e-05, + "loss": 3.1815, + "step": 10539 + }, + { + "epoch": 0.8450421919785132, + "grad_norm": 1.9011365175247192, + "learning_rate": 1.2950299020929584e-05, + "loss": 3.3326, + "step": 10540 + }, + { + "epoch": 0.8451223667595358, + "grad_norm": 1.7811436653137207, + "learning_rate": 1.294905834545498e-05, + "loss": 3.1992, + "step": 10541 + }, + { + "epoch": 0.8452025415405584, + "grad_norm": 1.7690136432647705, + "learning_rate": 1.2947817620260294e-05, + "loss": 3.1651, + "step": 10542 + }, + { + "epoch": 0.845282716321581, + "grad_norm": 1.6896518468856812, + "learning_rate": 1.294657684536644e-05, + "loss": 3.1469, + "step": 10543 + }, + { + "epoch": 0.8453628911026037, + "grad_norm": 1.9552698135375977, + "learning_rate": 1.2945336020794338e-05, + "loss": 3.3032, + "step": 10544 + }, + { + "epoch": 0.8454430658836263, + "grad_norm": 0.7392926216125488, + "learning_rate": 1.2944095146564913e-05, + "loss": 2.5046, + "step": 10545 + }, + { + "epoch": 0.8455232406646489, + "grad_norm": 1.7311780452728271, + "learning_rate": 1.2942854222699079e-05, + "loss": 3.2686, + "step": 10546 + }, + { + "epoch": 0.8456034154456715, + "grad_norm": 1.7157158851623535, + "learning_rate": 1.2941613249217761e-05, + "loss": 3.192, + "step": 10547 + }, + { + "epoch": 0.8456835902266941, + "grad_norm": 1.6281309127807617, + "learning_rate": 1.2940372226141877e-05, + "loss": 3.1621, + "step": 10548 + }, + { + "epoch": 0.8457637650077168, + "grad_norm": 1.7281498908996582, + "learning_rate": 1.2939131153492354e-05, + "loss": 3.2976, + "step": 10549 + }, + { + "epoch": 0.8458439397887395, + "grad_norm": 1.8555357456207275, + "learning_rate": 1.2937890031290122e-05, + "loss": 3.2723, + "step": 10550 + }, + { + "epoch": 0.8459241145697621, + "grad_norm": 1.7305965423583984, + "learning_rate": 1.2936648859556091e-05, + "loss": 3.1408, + "step": 10551 + }, + { + "epoch": 0.8460042893507848, + "grad_norm": 1.698639154434204, + "learning_rate": 1.2935407638311202e-05, + "loss": 3.2574, + "step": 10552 + }, + { + "epoch": 0.8460844641318074, + "grad_norm": 2.123249053955078, + "learning_rate": 1.293416636757637e-05, + "loss": 3.2629, + "step": 10553 + }, + { + "epoch": 0.84616463891283, + "grad_norm": 1.9107102155685425, + "learning_rate": 1.2932925047372526e-05, + "loss": 3.2384, + "step": 10554 + }, + { + "epoch": 0.8462448136938526, + "grad_norm": 1.8534648418426514, + "learning_rate": 1.2931683677720606e-05, + "loss": 3.3099, + "step": 10555 + }, + { + "epoch": 0.8463249884748753, + "grad_norm": 1.897691249847412, + "learning_rate": 1.2930442258641526e-05, + "loss": 3.2186, + "step": 10556 + }, + { + "epoch": 0.8464051632558979, + "grad_norm": 1.7411861419677734, + "learning_rate": 1.2929200790156226e-05, + "loss": 3.2081, + "step": 10557 + }, + { + "epoch": 0.8464853380369205, + "grad_norm": 1.833192229270935, + "learning_rate": 1.2927959272285632e-05, + "loss": 3.1935, + "step": 10558 + }, + { + "epoch": 0.8465655128179431, + "grad_norm": 1.7877330780029297, + "learning_rate": 1.2926717705050675e-05, + "loss": 3.3473, + "step": 10559 + }, + { + "epoch": 0.8466456875989657, + "grad_norm": 1.823371171951294, + "learning_rate": 1.2925476088472293e-05, + "loss": 3.3503, + "step": 10560 + }, + { + "epoch": 0.8467258623799884, + "grad_norm": 1.7736670970916748, + "learning_rate": 1.292423442257141e-05, + "loss": 3.2002, + "step": 10561 + }, + { + "epoch": 0.846806037161011, + "grad_norm": 1.6431782245635986, + "learning_rate": 1.2922992707368969e-05, + "loss": 3.088, + "step": 10562 + }, + { + "epoch": 0.8468862119420336, + "grad_norm": 1.8417996168136597, + "learning_rate": 1.29217509428859e-05, + "loss": 3.2865, + "step": 10563 + }, + { + "epoch": 0.8469663867230562, + "grad_norm": 1.8123340606689453, + "learning_rate": 1.292050912914314e-05, + "loss": 3.1593, + "step": 10564 + }, + { + "epoch": 0.8470465615040789, + "grad_norm": 1.6710424423217773, + "learning_rate": 1.2919267266161626e-05, + "loss": 3.2529, + "step": 10565 + }, + { + "epoch": 0.8471267362851015, + "grad_norm": 1.7231724262237549, + "learning_rate": 1.2918025353962293e-05, + "loss": 3.1792, + "step": 10566 + }, + { + "epoch": 0.8472069110661241, + "grad_norm": 1.8860127925872803, + "learning_rate": 1.2916783392566083e-05, + "loss": 3.1765, + "step": 10567 + }, + { + "epoch": 0.8472870858471467, + "grad_norm": 1.7293853759765625, + "learning_rate": 1.291554138199393e-05, + "loss": 3.1558, + "step": 10568 + }, + { + "epoch": 0.8473672606281694, + "grad_norm": 1.845777988433838, + "learning_rate": 1.2914299322266778e-05, + "loss": 3.2561, + "step": 10569 + }, + { + "epoch": 0.847447435409192, + "grad_norm": 1.716354250907898, + "learning_rate": 1.291305721340557e-05, + "loss": 3.1886, + "step": 10570 + }, + { + "epoch": 0.8475276101902147, + "grad_norm": 1.8496252298355103, + "learning_rate": 1.2911815055431239e-05, + "loss": 3.203, + "step": 10571 + }, + { + "epoch": 0.8476077849712373, + "grad_norm": 1.8511568307876587, + "learning_rate": 1.2910572848364735e-05, + "loss": 3.2705, + "step": 10572 + }, + { + "epoch": 0.84768795975226, + "grad_norm": 0.7813490033149719, + "learning_rate": 1.2909330592226999e-05, + "loss": 2.4716, + "step": 10573 + }, + { + "epoch": 0.8477681345332826, + "grad_norm": 1.6863970756530762, + "learning_rate": 1.2908088287038974e-05, + "loss": 3.172, + "step": 10574 + }, + { + "epoch": 0.8478483093143052, + "grad_norm": 1.846093773841858, + "learning_rate": 1.2906845932821606e-05, + "loss": 3.271, + "step": 10575 + }, + { + "epoch": 0.8479284840953278, + "grad_norm": 1.763641595840454, + "learning_rate": 1.2905603529595838e-05, + "loss": 3.1871, + "step": 10576 + }, + { + "epoch": 0.8480086588763505, + "grad_norm": 1.7633086442947388, + "learning_rate": 1.290436107738262e-05, + "loss": 3.2383, + "step": 10577 + }, + { + "epoch": 0.8480888336573731, + "grad_norm": 1.917998194694519, + "learning_rate": 1.2903118576202898e-05, + "loss": 3.1962, + "step": 10578 + }, + { + "epoch": 0.8481690084383957, + "grad_norm": 1.7481745481491089, + "learning_rate": 1.2901876026077617e-05, + "loss": 3.2045, + "step": 10579 + }, + { + "epoch": 0.8482491832194183, + "grad_norm": 1.8316683769226074, + "learning_rate": 1.2900633427027734e-05, + "loss": 3.2379, + "step": 10580 + }, + { + "epoch": 0.848329358000441, + "grad_norm": 1.7754689455032349, + "learning_rate": 1.289939077907419e-05, + "loss": 3.2776, + "step": 10581 + }, + { + "epoch": 0.8484095327814636, + "grad_norm": 1.7780835628509521, + "learning_rate": 1.289814808223794e-05, + "loss": 3.1556, + "step": 10582 + }, + { + "epoch": 0.8484897075624862, + "grad_norm": 1.8168517351150513, + "learning_rate": 1.2896905336539936e-05, + "loss": 3.1814, + "step": 10583 + }, + { + "epoch": 0.8485698823435088, + "grad_norm": 1.908409833908081, + "learning_rate": 1.2895662542001128e-05, + "loss": 3.2154, + "step": 10584 + }, + { + "epoch": 0.8486500571245315, + "grad_norm": 0.7885242104530334, + "learning_rate": 1.289441969864247e-05, + "loss": 2.5046, + "step": 10585 + }, + { + "epoch": 0.8487302319055541, + "grad_norm": 1.9046902656555176, + "learning_rate": 1.2893176806484916e-05, + "loss": 3.1785, + "step": 10586 + }, + { + "epoch": 0.8488104066865767, + "grad_norm": 1.6820099353790283, + "learning_rate": 1.2891933865549423e-05, + "loss": 3.1626, + "step": 10587 + }, + { + "epoch": 0.8488905814675993, + "grad_norm": 1.819991111755371, + "learning_rate": 1.2890690875856939e-05, + "loss": 3.1367, + "step": 10588 + }, + { + "epoch": 0.848970756248622, + "grad_norm": 1.7892426252365112, + "learning_rate": 1.288944783742843e-05, + "loss": 3.2534, + "step": 10589 + }, + { + "epoch": 0.8490509310296446, + "grad_norm": 1.8179054260253906, + "learning_rate": 1.288820475028485e-05, + "loss": 3.2172, + "step": 10590 + }, + { + "epoch": 0.8491311058106673, + "grad_norm": 1.7729991674423218, + "learning_rate": 1.2886961614447152e-05, + "loss": 3.2, + "step": 10591 + }, + { + "epoch": 0.8492112805916899, + "grad_norm": 1.6733402013778687, + "learning_rate": 1.2885718429936301e-05, + "loss": 3.1355, + "step": 10592 + }, + { + "epoch": 0.8492914553727126, + "grad_norm": 1.659045934677124, + "learning_rate": 1.2884475196773252e-05, + "loss": 3.1376, + "step": 10593 + }, + { + "epoch": 0.8493716301537352, + "grad_norm": 1.7856792211532593, + "learning_rate": 1.288323191497897e-05, + "loss": 3.1859, + "step": 10594 + }, + { + "epoch": 0.8494518049347578, + "grad_norm": 1.7949239015579224, + "learning_rate": 1.2881988584574412e-05, + "loss": 3.283, + "step": 10595 + }, + { + "epoch": 0.8495319797157804, + "grad_norm": 1.8898886442184448, + "learning_rate": 1.2880745205580543e-05, + "loss": 3.1957, + "step": 10596 + }, + { + "epoch": 0.849612154496803, + "grad_norm": 1.731954574584961, + "learning_rate": 1.2879501778018326e-05, + "loss": 3.2633, + "step": 10597 + }, + { + "epoch": 0.8496923292778257, + "grad_norm": 1.8650171756744385, + "learning_rate": 1.2878258301908724e-05, + "loss": 3.2156, + "step": 10598 + }, + { + "epoch": 0.8497725040588483, + "grad_norm": 1.9491254091262817, + "learning_rate": 1.2877014777272698e-05, + "loss": 3.2317, + "step": 10599 + }, + { + "epoch": 0.8498526788398709, + "grad_norm": 1.766871452331543, + "learning_rate": 1.2875771204131222e-05, + "loss": 3.2473, + "step": 10600 + }, + { + "epoch": 0.8499328536208935, + "grad_norm": 1.7959398031234741, + "learning_rate": 1.2874527582505255e-05, + "loss": 3.1849, + "step": 10601 + }, + { + "epoch": 0.8500130284019162, + "grad_norm": 1.717610239982605, + "learning_rate": 1.2873283912415765e-05, + "loss": 3.2844, + "step": 10602 + }, + { + "epoch": 0.8500932031829388, + "grad_norm": 0.7434692978858948, + "learning_rate": 1.2872040193883722e-05, + "loss": 2.4885, + "step": 10603 + }, + { + "epoch": 0.8501733779639614, + "grad_norm": 1.6727759838104248, + "learning_rate": 1.287079642693009e-05, + "loss": 3.2623, + "step": 10604 + }, + { + "epoch": 0.850253552744984, + "grad_norm": 1.8089869022369385, + "learning_rate": 1.2869552611575848e-05, + "loss": 3.2028, + "step": 10605 + }, + { + "epoch": 0.8503337275260067, + "grad_norm": 1.7176836729049683, + "learning_rate": 1.2868308747841956e-05, + "loss": 3.1644, + "step": 10606 + }, + { + "epoch": 0.8504139023070293, + "grad_norm": 1.8469318151474, + "learning_rate": 1.2867064835749392e-05, + "loss": 3.2629, + "step": 10607 + }, + { + "epoch": 0.8504940770880519, + "grad_norm": 1.8332422971725464, + "learning_rate": 1.2865820875319124e-05, + "loss": 3.1707, + "step": 10608 + }, + { + "epoch": 0.8505742518690745, + "grad_norm": 1.885216236114502, + "learning_rate": 1.2864576866572125e-05, + "loss": 3.2346, + "step": 10609 + }, + { + "epoch": 0.8506544266500972, + "grad_norm": 1.8801987171173096, + "learning_rate": 1.2863332809529371e-05, + "loss": 3.2178, + "step": 10610 + }, + { + "epoch": 0.8507346014311199, + "grad_norm": 1.7781858444213867, + "learning_rate": 1.2862088704211834e-05, + "loss": 3.1661, + "step": 10611 + }, + { + "epoch": 0.8508147762121425, + "grad_norm": 1.8601276874542236, + "learning_rate": 1.286084455064049e-05, + "loss": 3.1853, + "step": 10612 + }, + { + "epoch": 0.8508949509931651, + "grad_norm": 1.8809717893600464, + "learning_rate": 1.285960034883632e-05, + "loss": 3.2196, + "step": 10613 + }, + { + "epoch": 0.8509751257741878, + "grad_norm": 1.8144296407699585, + "learning_rate": 1.2858356098820288e-05, + "loss": 3.2004, + "step": 10614 + }, + { + "epoch": 0.8510553005552104, + "grad_norm": 1.8299994468688965, + "learning_rate": 1.2857111800613385e-05, + "loss": 3.2057, + "step": 10615 + }, + { + "epoch": 0.851135475336233, + "grad_norm": 1.8460853099822998, + "learning_rate": 1.2855867454236581e-05, + "loss": 3.2043, + "step": 10616 + }, + { + "epoch": 0.8512156501172556, + "grad_norm": 1.975612998008728, + "learning_rate": 1.285462305971086e-05, + "loss": 3.2667, + "step": 10617 + }, + { + "epoch": 0.8512958248982783, + "grad_norm": 1.8482475280761719, + "learning_rate": 1.28533786170572e-05, + "loss": 3.3037, + "step": 10618 + }, + { + "epoch": 0.8513759996793009, + "grad_norm": 0.8078134655952454, + "learning_rate": 1.2852134126296583e-05, + "loss": 2.5054, + "step": 10619 + }, + { + "epoch": 0.8514561744603235, + "grad_norm": 1.8403078317642212, + "learning_rate": 1.2850889587449988e-05, + "loss": 3.2139, + "step": 10620 + }, + { + "epoch": 0.8515363492413461, + "grad_norm": 1.819905161857605, + "learning_rate": 1.2849645000538399e-05, + "loss": 3.1315, + "step": 10621 + }, + { + "epoch": 0.8516165240223688, + "grad_norm": 1.785058617591858, + "learning_rate": 1.28484003655828e-05, + "loss": 3.2174, + "step": 10622 + }, + { + "epoch": 0.8516966988033914, + "grad_norm": 1.7592657804489136, + "learning_rate": 1.2847155682604177e-05, + "loss": 3.2877, + "step": 10623 + }, + { + "epoch": 0.851776873584414, + "grad_norm": 1.9109559059143066, + "learning_rate": 1.2845910951623513e-05, + "loss": 3.1837, + "step": 10624 + }, + { + "epoch": 0.8518570483654366, + "grad_norm": 1.7589404582977295, + "learning_rate": 1.284466617266179e-05, + "loss": 3.0536, + "step": 10625 + }, + { + "epoch": 0.8519372231464593, + "grad_norm": 1.7505866289138794, + "learning_rate": 1.284342134574e-05, + "loss": 3.1215, + "step": 10626 + }, + { + "epoch": 0.8520173979274819, + "grad_norm": 1.8429516553878784, + "learning_rate": 1.2842176470879127e-05, + "loss": 3.2113, + "step": 10627 + }, + { + "epoch": 0.8520975727085045, + "grad_norm": 1.8913179636001587, + "learning_rate": 1.2840931548100162e-05, + "loss": 3.2199, + "step": 10628 + }, + { + "epoch": 0.8521777474895271, + "grad_norm": 1.8635520935058594, + "learning_rate": 1.2839686577424092e-05, + "loss": 3.116, + "step": 10629 + }, + { + "epoch": 0.8522579222705498, + "grad_norm": 1.7688833475112915, + "learning_rate": 1.2838441558871906e-05, + "loss": 3.1575, + "step": 10630 + }, + { + "epoch": 0.8523380970515724, + "grad_norm": 1.7919833660125732, + "learning_rate": 1.2837196492464593e-05, + "loss": 3.181, + "step": 10631 + }, + { + "epoch": 0.8524182718325951, + "grad_norm": 1.6951316595077515, + "learning_rate": 1.283595137822315e-05, + "loss": 3.1815, + "step": 10632 + }, + { + "epoch": 0.8524984466136177, + "grad_norm": 1.7336077690124512, + "learning_rate": 1.2834706216168569e-05, + "loss": 3.2033, + "step": 10633 + }, + { + "epoch": 0.8525786213946404, + "grad_norm": 1.7426414489746094, + "learning_rate": 1.2833461006321837e-05, + "loss": 3.1676, + "step": 10634 + }, + { + "epoch": 0.852658796175663, + "grad_norm": 1.8259201049804688, + "learning_rate": 1.2832215748703954e-05, + "loss": 3.2499, + "step": 10635 + }, + { + "epoch": 0.8527389709566856, + "grad_norm": 1.7170315980911255, + "learning_rate": 1.2830970443335909e-05, + "loss": 3.2601, + "step": 10636 + }, + { + "epoch": 0.8528191457377082, + "grad_norm": 1.7491116523742676, + "learning_rate": 1.2829725090238698e-05, + "loss": 3.142, + "step": 10637 + }, + { + "epoch": 0.8528993205187309, + "grad_norm": 1.724236249923706, + "learning_rate": 1.2828479689433325e-05, + "loss": 3.2872, + "step": 10638 + }, + { + "epoch": 0.8529794952997535, + "grad_norm": 1.7625951766967773, + "learning_rate": 1.2827234240940776e-05, + "loss": 3.2274, + "step": 10639 + }, + { + "epoch": 0.8530596700807761, + "grad_norm": 1.7592341899871826, + "learning_rate": 1.2825988744782057e-05, + "loss": 3.2238, + "step": 10640 + }, + { + "epoch": 0.8531398448617987, + "grad_norm": 0.709554135799408, + "learning_rate": 1.2824743200978162e-05, + "loss": 2.4663, + "step": 10641 + }, + { + "epoch": 0.8532200196428213, + "grad_norm": 1.7758010625839233, + "learning_rate": 1.2823497609550093e-05, + "loss": 3.255, + "step": 10642 + }, + { + "epoch": 0.853300194423844, + "grad_norm": 1.6790568828582764, + "learning_rate": 1.282225197051885e-05, + "loss": 3.3044, + "step": 10643 + }, + { + "epoch": 0.8533803692048666, + "grad_norm": 1.769207239151001, + "learning_rate": 1.2821006283905431e-05, + "loss": 3.1275, + "step": 10644 + }, + { + "epoch": 0.8534605439858892, + "grad_norm": 2.189488410949707, + "learning_rate": 1.2819760549730843e-05, + "loss": 3.1567, + "step": 10645 + }, + { + "epoch": 0.8535407187669118, + "grad_norm": 1.7149593830108643, + "learning_rate": 1.2818514768016086e-05, + "loss": 3.1971, + "step": 10646 + }, + { + "epoch": 0.8536208935479345, + "grad_norm": 1.763654112815857, + "learning_rate": 1.2817268938782161e-05, + "loss": 3.2373, + "step": 10647 + }, + { + "epoch": 0.8537010683289571, + "grad_norm": 1.6512292623519897, + "learning_rate": 1.2816023062050076e-05, + "loss": 3.2314, + "step": 10648 + }, + { + "epoch": 0.8537812431099797, + "grad_norm": 1.8872522115707397, + "learning_rate": 1.2814777137840834e-05, + "loss": 3.2513, + "step": 10649 + }, + { + "epoch": 0.8538614178910023, + "grad_norm": 1.7497056722640991, + "learning_rate": 1.2813531166175443e-05, + "loss": 3.3224, + "step": 10650 + }, + { + "epoch": 0.853941592672025, + "grad_norm": 1.8274199962615967, + "learning_rate": 1.2812285147074908e-05, + "loss": 3.2022, + "step": 10651 + }, + { + "epoch": 0.8540217674530477, + "grad_norm": 1.7002332210540771, + "learning_rate": 1.2811039080560234e-05, + "loss": 3.1823, + "step": 10652 + }, + { + "epoch": 0.8541019422340703, + "grad_norm": 1.686826229095459, + "learning_rate": 1.2809792966652435e-05, + "loss": 3.1946, + "step": 10653 + }, + { + "epoch": 0.854182117015093, + "grad_norm": 1.6847214698791504, + "learning_rate": 1.2808546805372515e-05, + "loss": 3.1408, + "step": 10654 + }, + { + "epoch": 0.8542622917961156, + "grad_norm": 1.750215768814087, + "learning_rate": 1.2807300596741488e-05, + "loss": 3.1285, + "step": 10655 + }, + { + "epoch": 0.8543424665771382, + "grad_norm": 1.9388880729675293, + "learning_rate": 1.2806054340780359e-05, + "loss": 3.205, + "step": 10656 + }, + { + "epoch": 0.8544226413581608, + "grad_norm": 1.9729681015014648, + "learning_rate": 1.2804808037510144e-05, + "loss": 3.2349, + "step": 10657 + }, + { + "epoch": 0.8545028161391834, + "grad_norm": 1.7586485147476196, + "learning_rate": 1.2803561686951856e-05, + "loss": 3.1326, + "step": 10658 + }, + { + "epoch": 0.8545829909202061, + "grad_norm": 1.730610966682434, + "learning_rate": 1.2802315289126503e-05, + "loss": 3.1765, + "step": 10659 + }, + { + "epoch": 0.8546631657012287, + "grad_norm": 1.7828902006149292, + "learning_rate": 1.2801068844055102e-05, + "loss": 3.2205, + "step": 10660 + }, + { + "epoch": 0.8547433404822513, + "grad_norm": 1.8114609718322754, + "learning_rate": 1.2799822351758669e-05, + "loss": 3.2003, + "step": 10661 + }, + { + "epoch": 0.8548235152632739, + "grad_norm": 1.8225308656692505, + "learning_rate": 1.2798575812258216e-05, + "loss": 3.2299, + "step": 10662 + }, + { + "epoch": 0.8549036900442966, + "grad_norm": 1.7172393798828125, + "learning_rate": 1.2797329225574763e-05, + "loss": 3.2078, + "step": 10663 + }, + { + "epoch": 0.8549838648253192, + "grad_norm": 1.8307806253433228, + "learning_rate": 1.2796082591729326e-05, + "loss": 3.2797, + "step": 10664 + }, + { + "epoch": 0.8550640396063418, + "grad_norm": 1.8517546653747559, + "learning_rate": 1.2794835910742922e-05, + "loss": 3.3142, + "step": 10665 + }, + { + "epoch": 0.8551442143873644, + "grad_norm": 1.790877342224121, + "learning_rate": 1.2793589182636567e-05, + "loss": 3.1599, + "step": 10666 + }, + { + "epoch": 0.8552243891683871, + "grad_norm": 1.9574605226516724, + "learning_rate": 1.2792342407431283e-05, + "loss": 3.2062, + "step": 10667 + }, + { + "epoch": 0.8553045639494097, + "grad_norm": 0.6983086466789246, + "learning_rate": 1.2791095585148092e-05, + "loss": 2.4816, + "step": 10668 + }, + { + "epoch": 0.8553847387304323, + "grad_norm": 1.7173445224761963, + "learning_rate": 1.2789848715808013e-05, + "loss": 3.1764, + "step": 10669 + }, + { + "epoch": 0.8554649135114549, + "grad_norm": 1.8380826711654663, + "learning_rate": 1.2788601799432068e-05, + "loss": 3.3071, + "step": 10670 + }, + { + "epoch": 0.8555450882924776, + "grad_norm": 1.8035857677459717, + "learning_rate": 1.2787354836041277e-05, + "loss": 3.1074, + "step": 10671 + }, + { + "epoch": 0.8556252630735002, + "grad_norm": 1.739168405532837, + "learning_rate": 1.2786107825656667e-05, + "loss": 3.2335, + "step": 10672 + }, + { + "epoch": 0.8557054378545229, + "grad_norm": 1.7777115106582642, + "learning_rate": 1.2784860768299264e-05, + "loss": 3.121, + "step": 10673 + }, + { + "epoch": 0.8557856126355455, + "grad_norm": 1.9140719175338745, + "learning_rate": 1.2783613663990086e-05, + "loss": 3.236, + "step": 10674 + }, + { + "epoch": 0.8558657874165682, + "grad_norm": 1.6882866621017456, + "learning_rate": 1.2782366512750167e-05, + "loss": 3.1334, + "step": 10675 + }, + { + "epoch": 0.8559459621975908, + "grad_norm": 1.6715214252471924, + "learning_rate": 1.2781119314600523e-05, + "loss": 3.1793, + "step": 10676 + }, + { + "epoch": 0.8560261369786134, + "grad_norm": 1.717750072479248, + "learning_rate": 1.277987206956219e-05, + "loss": 3.1763, + "step": 10677 + }, + { + "epoch": 0.856106311759636, + "grad_norm": 1.945451021194458, + "learning_rate": 1.2778624777656196e-05, + "loss": 3.2518, + "step": 10678 + }, + { + "epoch": 0.8561864865406587, + "grad_norm": 1.7998268604278564, + "learning_rate": 1.2777377438903567e-05, + "loss": 3.1472, + "step": 10679 + }, + { + "epoch": 0.8562666613216813, + "grad_norm": 1.7183723449707031, + "learning_rate": 1.2776130053325334e-05, + "loss": 3.1774, + "step": 10680 + }, + { + "epoch": 0.8563468361027039, + "grad_norm": 1.7789490222930908, + "learning_rate": 1.2774882620942521e-05, + "loss": 3.1028, + "step": 10681 + }, + { + "epoch": 0.8564270108837265, + "grad_norm": 1.8565491437911987, + "learning_rate": 1.2773635141776168e-05, + "loss": 3.3323, + "step": 10682 + }, + { + "epoch": 0.8565071856647491, + "grad_norm": 1.7804903984069824, + "learning_rate": 1.2772387615847308e-05, + "loss": 3.2515, + "step": 10683 + }, + { + "epoch": 0.8565873604457718, + "grad_norm": 1.7455756664276123, + "learning_rate": 1.2771140043176966e-05, + "loss": 3.1626, + "step": 10684 + }, + { + "epoch": 0.8566675352267944, + "grad_norm": 1.866541862487793, + "learning_rate": 1.2769892423786182e-05, + "loss": 3.2125, + "step": 10685 + }, + { + "epoch": 0.856747710007817, + "grad_norm": 1.8648061752319336, + "learning_rate": 1.2768644757695985e-05, + "loss": 3.2202, + "step": 10686 + }, + { + "epoch": 0.8568278847888396, + "grad_norm": 1.859273910522461, + "learning_rate": 1.2767397044927413e-05, + "loss": 3.2535, + "step": 10687 + }, + { + "epoch": 0.8569080595698623, + "grad_norm": 1.7186626195907593, + "learning_rate": 1.2766149285501505e-05, + "loss": 3.1062, + "step": 10688 + }, + { + "epoch": 0.8569882343508849, + "grad_norm": 1.8279659748077393, + "learning_rate": 1.2764901479439293e-05, + "loss": 3.2432, + "step": 10689 + }, + { + "epoch": 0.8570684091319075, + "grad_norm": 1.7186083793640137, + "learning_rate": 1.2763653626761818e-05, + "loss": 3.2343, + "step": 10690 + }, + { + "epoch": 0.8571485839129301, + "grad_norm": 0.7405577301979065, + "learning_rate": 1.2762405727490115e-05, + "loss": 2.4814, + "step": 10691 + }, + { + "epoch": 0.8572287586939528, + "grad_norm": 1.8033840656280518, + "learning_rate": 1.2761157781645226e-05, + "loss": 3.2865, + "step": 10692 + }, + { + "epoch": 0.8573089334749755, + "grad_norm": 1.7818074226379395, + "learning_rate": 1.2759909789248191e-05, + "loss": 3.1681, + "step": 10693 + }, + { + "epoch": 0.8573891082559981, + "grad_norm": 2.0735795497894287, + "learning_rate": 1.2758661750320047e-05, + "loss": 3.3633, + "step": 10694 + }, + { + "epoch": 0.8574692830370207, + "grad_norm": 1.958341121673584, + "learning_rate": 1.2757413664881842e-05, + "loss": 3.2502, + "step": 10695 + }, + { + "epoch": 0.8575494578180434, + "grad_norm": 1.863466501235962, + "learning_rate": 1.275616553295461e-05, + "loss": 3.1726, + "step": 10696 + }, + { + "epoch": 0.857629632599066, + "grad_norm": 1.7863266468048096, + "learning_rate": 1.27549173545594e-05, + "loss": 3.3286, + "step": 10697 + }, + { + "epoch": 0.8577098073800886, + "grad_norm": 1.840267300605774, + "learning_rate": 1.2753669129717254e-05, + "loss": 3.3282, + "step": 10698 + }, + { + "epoch": 0.8577899821611112, + "grad_norm": 1.897278070449829, + "learning_rate": 1.275242085844922e-05, + "loss": 3.1851, + "step": 10699 + }, + { + "epoch": 0.8578701569421339, + "grad_norm": 1.951453447341919, + "learning_rate": 1.2751172540776338e-05, + "loss": 3.1438, + "step": 10700 + }, + { + "epoch": 0.8579503317231565, + "grad_norm": 1.7443854808807373, + "learning_rate": 1.2749924176719657e-05, + "loss": 3.1863, + "step": 10701 + }, + { + "epoch": 0.8580305065041791, + "grad_norm": 1.8601171970367432, + "learning_rate": 1.274867576630022e-05, + "loss": 3.3044, + "step": 10702 + }, + { + "epoch": 0.8581106812852017, + "grad_norm": 1.7421269416809082, + "learning_rate": 1.2747427309539085e-05, + "loss": 3.2209, + "step": 10703 + }, + { + "epoch": 0.8581908560662244, + "grad_norm": 1.758620023727417, + "learning_rate": 1.2746178806457287e-05, + "loss": 3.1196, + "step": 10704 + }, + { + "epoch": 0.858271030847247, + "grad_norm": 1.8524292707443237, + "learning_rate": 1.2744930257075886e-05, + "loss": 3.1913, + "step": 10705 + }, + { + "epoch": 0.8583512056282696, + "grad_norm": 0.7308001518249512, + "learning_rate": 1.274368166141593e-05, + "loss": 2.4722, + "step": 10706 + }, + { + "epoch": 0.8584313804092922, + "grad_norm": 1.725433349609375, + "learning_rate": 1.2742433019498463e-05, + "loss": 3.2614, + "step": 10707 + }, + { + "epoch": 0.8585115551903149, + "grad_norm": 0.7156065106391907, + "learning_rate": 1.2741184331344548e-05, + "loss": 2.4834, + "step": 10708 + }, + { + "epoch": 0.8585917299713375, + "grad_norm": 1.76617431640625, + "learning_rate": 1.2739935596975225e-05, + "loss": 3.2947, + "step": 10709 + }, + { + "epoch": 0.8586719047523601, + "grad_norm": 1.70106041431427, + "learning_rate": 1.2738686816411558e-05, + "loss": 3.1971, + "step": 10710 + }, + { + "epoch": 0.8587520795333827, + "grad_norm": 1.7684389352798462, + "learning_rate": 1.2737437989674596e-05, + "loss": 3.263, + "step": 10711 + }, + { + "epoch": 0.8588322543144054, + "grad_norm": 1.7985426187515259, + "learning_rate": 1.2736189116785393e-05, + "loss": 3.2499, + "step": 10712 + }, + { + "epoch": 0.8589124290954281, + "grad_norm": 1.849769949913025, + "learning_rate": 1.2734940197765007e-05, + "loss": 3.1783, + "step": 10713 + }, + { + "epoch": 0.8589926038764507, + "grad_norm": 1.8439075946807861, + "learning_rate": 1.273369123263449e-05, + "loss": 3.1935, + "step": 10714 + }, + { + "epoch": 0.8590727786574733, + "grad_norm": 1.7263778448104858, + "learning_rate": 1.2732442221414907e-05, + "loss": 3.202, + "step": 10715 + }, + { + "epoch": 0.859152953438496, + "grad_norm": 1.8851699829101562, + "learning_rate": 1.2731193164127306e-05, + "loss": 3.3301, + "step": 10716 + }, + { + "epoch": 0.8592331282195186, + "grad_norm": 1.895884394645691, + "learning_rate": 1.2729944060792754e-05, + "loss": 3.3143, + "step": 10717 + }, + { + "epoch": 0.8593133030005412, + "grad_norm": 0.7911091446876526, + "learning_rate": 1.272869491143231e-05, + "loss": 2.5123, + "step": 10718 + }, + { + "epoch": 0.8593934777815638, + "grad_norm": 1.7949423789978027, + "learning_rate": 1.2727445716067026e-05, + "loss": 3.1936, + "step": 10719 + }, + { + "epoch": 0.8594736525625865, + "grad_norm": 1.8033273220062256, + "learning_rate": 1.2726196474717973e-05, + "loss": 3.1888, + "step": 10720 + }, + { + "epoch": 0.8595538273436091, + "grad_norm": 1.750725507736206, + "learning_rate": 1.2724947187406205e-05, + "loss": 3.1958, + "step": 10721 + }, + { + "epoch": 0.8596340021246317, + "grad_norm": 1.9748668670654297, + "learning_rate": 1.2723697854152788e-05, + "loss": 3.1041, + "step": 10722 + }, + { + "epoch": 0.8597141769056543, + "grad_norm": 1.7395392656326294, + "learning_rate": 1.2722448474978785e-05, + "loss": 3.1783, + "step": 10723 + }, + { + "epoch": 0.859794351686677, + "grad_norm": 1.9873861074447632, + "learning_rate": 1.2721199049905262e-05, + "loss": 3.1436, + "step": 10724 + }, + { + "epoch": 0.8598745264676996, + "grad_norm": 1.7835631370544434, + "learning_rate": 1.2719949578953281e-05, + "loss": 3.1017, + "step": 10725 + }, + { + "epoch": 0.8599547012487222, + "grad_norm": 1.8250148296356201, + "learning_rate": 1.271870006214391e-05, + "loss": 3.2079, + "step": 10726 + }, + { + "epoch": 0.8600348760297448, + "grad_norm": 1.7246153354644775, + "learning_rate": 1.271745049949821e-05, + "loss": 3.1608, + "step": 10727 + }, + { + "epoch": 0.8601150508107674, + "grad_norm": 1.8953522443771362, + "learning_rate": 1.2716200891037258e-05, + "loss": 3.2225, + "step": 10728 + }, + { + "epoch": 0.8601952255917901, + "grad_norm": 1.7584248781204224, + "learning_rate": 1.2714951236782111e-05, + "loss": 3.2609, + "step": 10729 + }, + { + "epoch": 0.8602754003728127, + "grad_norm": 1.7142390012741089, + "learning_rate": 1.2713701536753848e-05, + "loss": 3.1757, + "step": 10730 + }, + { + "epoch": 0.8603555751538353, + "grad_norm": 0.7506260871887207, + "learning_rate": 1.271245179097353e-05, + "loss": 2.4974, + "step": 10731 + }, + { + "epoch": 0.8604357499348579, + "grad_norm": 1.8503062725067139, + "learning_rate": 1.271120199946223e-05, + "loss": 3.3506, + "step": 10732 + }, + { + "epoch": 0.8605159247158806, + "grad_norm": 1.8053520917892456, + "learning_rate": 1.2709952162241023e-05, + "loss": 3.1926, + "step": 10733 + }, + { + "epoch": 0.8605960994969033, + "grad_norm": 1.6750446557998657, + "learning_rate": 1.2708702279330976e-05, + "loss": 3.1313, + "step": 10734 + }, + { + "epoch": 0.8606762742779259, + "grad_norm": 1.9184455871582031, + "learning_rate": 1.2707452350753164e-05, + "loss": 3.1569, + "step": 10735 + }, + { + "epoch": 0.8607564490589485, + "grad_norm": 0.7291615009307861, + "learning_rate": 1.2706202376528658e-05, + "loss": 2.491, + "step": 10736 + }, + { + "epoch": 0.8608366238399712, + "grad_norm": 1.8423436880111694, + "learning_rate": 1.2704952356678535e-05, + "loss": 3.199, + "step": 10737 + }, + { + "epoch": 0.8609167986209938, + "grad_norm": 1.6878387928009033, + "learning_rate": 1.270370229122387e-05, + "loss": 3.3521, + "step": 10738 + }, + { + "epoch": 0.8609969734020164, + "grad_norm": 1.7939529418945312, + "learning_rate": 1.2702452180185734e-05, + "loss": 3.2977, + "step": 10739 + }, + { + "epoch": 0.861077148183039, + "grad_norm": 1.8879039287567139, + "learning_rate": 1.2701202023585209e-05, + "loss": 3.1849, + "step": 10740 + }, + { + "epoch": 0.8611573229640617, + "grad_norm": 1.7907934188842773, + "learning_rate": 1.2699951821443368e-05, + "loss": 3.2455, + "step": 10741 + }, + { + "epoch": 0.8612374977450843, + "grad_norm": 1.8267040252685547, + "learning_rate": 1.2698701573781291e-05, + "loss": 3.1721, + "step": 10742 + }, + { + "epoch": 0.8613176725261069, + "grad_norm": 1.7820180654525757, + "learning_rate": 1.269745128062006e-05, + "loss": 3.2321, + "step": 10743 + }, + { + "epoch": 0.8613978473071295, + "grad_norm": 1.7370247840881348, + "learning_rate": 1.2696200941980747e-05, + "loss": 3.0493, + "step": 10744 + }, + { + "epoch": 0.8614780220881522, + "grad_norm": 0.760126531124115, + "learning_rate": 1.2694950557884438e-05, + "loss": 2.4828, + "step": 10745 + }, + { + "epoch": 0.8615581968691748, + "grad_norm": 0.8038840889930725, + "learning_rate": 1.2693700128352214e-05, + "loss": 2.4764, + "step": 10746 + }, + { + "epoch": 0.8616383716501974, + "grad_norm": 2.1341025829315186, + "learning_rate": 1.2692449653405152e-05, + "loss": 3.2504, + "step": 10747 + }, + { + "epoch": 0.86171854643122, + "grad_norm": 1.8320749998092651, + "learning_rate": 1.269119913306434e-05, + "loss": 3.2494, + "step": 10748 + }, + { + "epoch": 0.8617987212122427, + "grad_norm": 1.7101120948791504, + "learning_rate": 1.2689948567350855e-05, + "loss": 3.1614, + "step": 10749 + }, + { + "epoch": 0.8618788959932653, + "grad_norm": 1.9137850999832153, + "learning_rate": 1.2688697956285792e-05, + "loss": 3.322, + "step": 10750 + }, + { + "epoch": 0.8619590707742879, + "grad_norm": 1.8674005270004272, + "learning_rate": 1.2687447299890226e-05, + "loss": 3.2683, + "step": 10751 + }, + { + "epoch": 0.8620392455553105, + "grad_norm": 1.7664979696273804, + "learning_rate": 1.2686196598185243e-05, + "loss": 3.2711, + "step": 10752 + }, + { + "epoch": 0.8621194203363332, + "grad_norm": 1.735873818397522, + "learning_rate": 1.2684945851191936e-05, + "loss": 3.1447, + "step": 10753 + }, + { + "epoch": 0.8621995951173559, + "grad_norm": 1.7869863510131836, + "learning_rate": 1.2683695058931384e-05, + "loss": 3.1669, + "step": 10754 + }, + { + "epoch": 0.8622797698983785, + "grad_norm": 1.9754217863082886, + "learning_rate": 1.2682444221424683e-05, + "loss": 3.1292, + "step": 10755 + }, + { + "epoch": 0.8623599446794011, + "grad_norm": 1.7854702472686768, + "learning_rate": 1.2681193338692916e-05, + "loss": 3.3326, + "step": 10756 + }, + { + "epoch": 0.8624401194604238, + "grad_norm": 1.7567896842956543, + "learning_rate": 1.2679942410757173e-05, + "loss": 3.2408, + "step": 10757 + }, + { + "epoch": 0.8625202942414464, + "grad_norm": 0.8258539438247681, + "learning_rate": 1.2678691437638549e-05, + "loss": 2.4786, + "step": 10758 + }, + { + "epoch": 0.862600469022469, + "grad_norm": 2.3619956970214844, + "learning_rate": 1.2677440419358129e-05, + "loss": 3.1953, + "step": 10759 + }, + { + "epoch": 0.8626806438034916, + "grad_norm": 1.8338327407836914, + "learning_rate": 1.267618935593701e-05, + "loss": 3.1396, + "step": 10760 + }, + { + "epoch": 0.8627608185845143, + "grad_norm": 1.750479817390442, + "learning_rate": 1.2674938247396277e-05, + "loss": 3.2215, + "step": 10761 + }, + { + "epoch": 0.8628409933655369, + "grad_norm": 0.7398033142089844, + "learning_rate": 1.267368709375703e-05, + "loss": 2.488, + "step": 10762 + }, + { + "epoch": 0.8629211681465595, + "grad_norm": 1.8883121013641357, + "learning_rate": 1.2672435895040362e-05, + "loss": 3.1426, + "step": 10763 + }, + { + "epoch": 0.8630013429275821, + "grad_norm": 1.8924822807312012, + "learning_rate": 1.2671184651267364e-05, + "loss": 3.1445, + "step": 10764 + }, + { + "epoch": 0.8630815177086048, + "grad_norm": 1.7416635751724243, + "learning_rate": 1.266993336245914e-05, + "loss": 3.2083, + "step": 10765 + }, + { + "epoch": 0.8631616924896274, + "grad_norm": 1.8498265743255615, + "learning_rate": 1.2668682028636775e-05, + "loss": 3.2179, + "step": 10766 + }, + { + "epoch": 0.86324186727065, + "grad_norm": 1.763372540473938, + "learning_rate": 1.266743064982137e-05, + "loss": 3.2356, + "step": 10767 + }, + { + "epoch": 0.8633220420516726, + "grad_norm": 1.8395787477493286, + "learning_rate": 1.266617922603403e-05, + "loss": 3.1943, + "step": 10768 + }, + { + "epoch": 0.8634022168326952, + "grad_norm": 1.8375643491744995, + "learning_rate": 1.2664927757295844e-05, + "loss": 3.2058, + "step": 10769 + }, + { + "epoch": 0.8634823916137179, + "grad_norm": 1.7913649082183838, + "learning_rate": 1.2663676243627919e-05, + "loss": 3.2517, + "step": 10770 + }, + { + "epoch": 0.8635625663947405, + "grad_norm": 1.8305598497390747, + "learning_rate": 1.2662424685051347e-05, + "loss": 3.161, + "step": 10771 + }, + { + "epoch": 0.8636427411757631, + "grad_norm": 1.7197697162628174, + "learning_rate": 1.2661173081587234e-05, + "loss": 3.1761, + "step": 10772 + }, + { + "epoch": 0.8637229159567857, + "grad_norm": 1.9017372131347656, + "learning_rate": 1.2659921433256683e-05, + "loss": 3.1376, + "step": 10773 + }, + { + "epoch": 0.8638030907378085, + "grad_norm": 1.6891931295394897, + "learning_rate": 1.265866974008079e-05, + "loss": 3.1758, + "step": 10774 + }, + { + "epoch": 0.8638832655188311, + "grad_norm": 1.819990634918213, + "learning_rate": 1.2657418002080668e-05, + "loss": 3.2126, + "step": 10775 + }, + { + "epoch": 0.8639634402998537, + "grad_norm": 1.832053780555725, + "learning_rate": 1.265616621927741e-05, + "loss": 3.2673, + "step": 10776 + }, + { + "epoch": 0.8640436150808763, + "grad_norm": 1.8031214475631714, + "learning_rate": 1.2654914391692128e-05, + "loss": 3.3057, + "step": 10777 + }, + { + "epoch": 0.864123789861899, + "grad_norm": 1.8429418802261353, + "learning_rate": 1.2653662519345925e-05, + "loss": 3.2212, + "step": 10778 + }, + { + "epoch": 0.8642039646429216, + "grad_norm": 1.7262955904006958, + "learning_rate": 1.2652410602259907e-05, + "loss": 3.2035, + "step": 10779 + }, + { + "epoch": 0.8642841394239442, + "grad_norm": 1.7936469316482544, + "learning_rate": 1.2651158640455183e-05, + "loss": 3.1435, + "step": 10780 + }, + { + "epoch": 0.8643643142049668, + "grad_norm": 1.8393782377243042, + "learning_rate": 1.2649906633952855e-05, + "loss": 3.2866, + "step": 10781 + }, + { + "epoch": 0.8644444889859895, + "grad_norm": 1.8674259185791016, + "learning_rate": 1.2648654582774035e-05, + "loss": 3.2755, + "step": 10782 + }, + { + "epoch": 0.8645246637670121, + "grad_norm": 1.7532312870025635, + "learning_rate": 1.2647402486939837e-05, + "loss": 3.1984, + "step": 10783 + }, + { + "epoch": 0.8646048385480347, + "grad_norm": 1.7217342853546143, + "learning_rate": 1.2646150346471362e-05, + "loss": 3.2391, + "step": 10784 + }, + { + "epoch": 0.8646850133290573, + "grad_norm": 1.8314144611358643, + "learning_rate": 1.2644898161389727e-05, + "loss": 3.0996, + "step": 10785 + }, + { + "epoch": 0.86476518811008, + "grad_norm": 1.7897841930389404, + "learning_rate": 1.2643645931716037e-05, + "loss": 3.1265, + "step": 10786 + }, + { + "epoch": 0.8648453628911026, + "grad_norm": 1.8829904794692993, + "learning_rate": 1.2642393657471412e-05, + "loss": 3.1938, + "step": 10787 + }, + { + "epoch": 0.8649255376721252, + "grad_norm": 0.769835889339447, + "learning_rate": 1.2641141338676961e-05, + "loss": 2.5023, + "step": 10788 + }, + { + "epoch": 0.8650057124531478, + "grad_norm": 1.7941898107528687, + "learning_rate": 1.2639888975353798e-05, + "loss": 3.2183, + "step": 10789 + }, + { + "epoch": 0.8650858872341705, + "grad_norm": 1.9364652633666992, + "learning_rate": 1.2638636567523037e-05, + "loss": 3.1842, + "step": 10790 + }, + { + "epoch": 0.8651660620151931, + "grad_norm": 1.8437000513076782, + "learning_rate": 1.2637384115205791e-05, + "loss": 3.2319, + "step": 10791 + }, + { + "epoch": 0.8652462367962157, + "grad_norm": 1.8743752241134644, + "learning_rate": 1.263613161842318e-05, + "loss": 3.2329, + "step": 10792 + }, + { + "epoch": 0.8653264115772383, + "grad_norm": 1.8159712553024292, + "learning_rate": 1.263487907719632e-05, + "loss": 3.2891, + "step": 10793 + }, + { + "epoch": 0.865406586358261, + "grad_norm": 1.8118607997894287, + "learning_rate": 1.2633626491546326e-05, + "loss": 3.1856, + "step": 10794 + }, + { + "epoch": 0.8654867611392837, + "grad_norm": 1.8683760166168213, + "learning_rate": 1.2632373861494321e-05, + "loss": 3.2581, + "step": 10795 + }, + { + "epoch": 0.8655669359203063, + "grad_norm": 1.8238052129745483, + "learning_rate": 1.2631121187061415e-05, + "loss": 3.1061, + "step": 10796 + }, + { + "epoch": 0.8656471107013289, + "grad_norm": 1.9227373600006104, + "learning_rate": 1.2629868468268736e-05, + "loss": 3.243, + "step": 10797 + }, + { + "epoch": 0.8657272854823516, + "grad_norm": 1.7834546566009521, + "learning_rate": 1.26286157051374e-05, + "loss": 3.2695, + "step": 10798 + }, + { + "epoch": 0.8658074602633742, + "grad_norm": 1.7348086833953857, + "learning_rate": 1.2627362897688531e-05, + "loss": 3.2181, + "step": 10799 + }, + { + "epoch": 0.8658876350443968, + "grad_norm": 1.7884528636932373, + "learning_rate": 1.2626110045943252e-05, + "loss": 3.2845, + "step": 10800 + }, + { + "epoch": 0.8659678098254194, + "grad_norm": 1.7198907136917114, + "learning_rate": 1.2624857149922678e-05, + "loss": 3.2297, + "step": 10801 + }, + { + "epoch": 0.8660479846064421, + "grad_norm": 1.7844641208648682, + "learning_rate": 1.2623604209647939e-05, + "loss": 3.2716, + "step": 10802 + }, + { + "epoch": 0.8661281593874647, + "grad_norm": 1.7840971946716309, + "learning_rate": 1.2622351225140159e-05, + "loss": 3.1381, + "step": 10803 + }, + { + "epoch": 0.8662083341684873, + "grad_norm": 1.7731059789657593, + "learning_rate": 1.2621098196420462e-05, + "loss": 3.1627, + "step": 10804 + }, + { + "epoch": 0.8662885089495099, + "grad_norm": 1.7820802927017212, + "learning_rate": 1.261984512350997e-05, + "loss": 3.2549, + "step": 10805 + }, + { + "epoch": 0.8663686837305326, + "grad_norm": 1.713563323020935, + "learning_rate": 1.2618592006429816e-05, + "loss": 3.2407, + "step": 10806 + }, + { + "epoch": 0.8664488585115552, + "grad_norm": 1.7670572996139526, + "learning_rate": 1.2617338845201124e-05, + "loss": 3.2672, + "step": 10807 + }, + { + "epoch": 0.8665290332925778, + "grad_norm": 1.9074361324310303, + "learning_rate": 1.261608563984502e-05, + "loss": 3.2004, + "step": 10808 + }, + { + "epoch": 0.8666092080736004, + "grad_norm": 1.919299602508545, + "learning_rate": 1.2614832390382634e-05, + "loss": 3.3118, + "step": 10809 + }, + { + "epoch": 0.866689382854623, + "grad_norm": 1.7623894214630127, + "learning_rate": 1.2613579096835093e-05, + "loss": 3.1638, + "step": 10810 + }, + { + "epoch": 0.8667695576356457, + "grad_norm": 1.7182587385177612, + "learning_rate": 1.2612325759223537e-05, + "loss": 3.1323, + "step": 10811 + }, + { + "epoch": 0.8668497324166683, + "grad_norm": 1.844378113746643, + "learning_rate": 1.2611072377569082e-05, + "loss": 3.1429, + "step": 10812 + }, + { + "epoch": 0.8669299071976909, + "grad_norm": 1.8085832595825195, + "learning_rate": 1.260981895189287e-05, + "loss": 3.1335, + "step": 10813 + }, + { + "epoch": 0.8670100819787135, + "grad_norm": 1.8481460809707642, + "learning_rate": 1.2608565482216029e-05, + "loss": 3.2097, + "step": 10814 + }, + { + "epoch": 0.8670902567597363, + "grad_norm": 1.7401431798934937, + "learning_rate": 1.2607311968559694e-05, + "loss": 3.3107, + "step": 10815 + }, + { + "epoch": 0.8671704315407589, + "grad_norm": 1.8877785205841064, + "learning_rate": 1.2606058410944998e-05, + "loss": 3.2435, + "step": 10816 + }, + { + "epoch": 0.8672506063217815, + "grad_norm": 1.819374918937683, + "learning_rate": 1.2604804809393076e-05, + "loss": 3.2473, + "step": 10817 + }, + { + "epoch": 0.8673307811028041, + "grad_norm": 1.8267152309417725, + "learning_rate": 1.2603551163925065e-05, + "loss": 3.2388, + "step": 10818 + }, + { + "epoch": 0.8674109558838268, + "grad_norm": 1.7296855449676514, + "learning_rate": 1.2602297474562096e-05, + "loss": 3.2688, + "step": 10819 + }, + { + "epoch": 0.8674911306648494, + "grad_norm": 1.7758817672729492, + "learning_rate": 1.2601043741325307e-05, + "loss": 3.2277, + "step": 10820 + }, + { + "epoch": 0.867571305445872, + "grad_norm": 1.6886563301086426, + "learning_rate": 1.2599789964235842e-05, + "loss": 3.2697, + "step": 10821 + }, + { + "epoch": 0.8676514802268946, + "grad_norm": 1.8083105087280273, + "learning_rate": 1.2598536143314832e-05, + "loss": 3.0676, + "step": 10822 + }, + { + "epoch": 0.8677316550079173, + "grad_norm": 1.8690197467803955, + "learning_rate": 1.2597282278583418e-05, + "loss": 3.2742, + "step": 10823 + }, + { + "epoch": 0.8678118297889399, + "grad_norm": 1.8439043760299683, + "learning_rate": 1.2596028370062741e-05, + "loss": 3.2107, + "step": 10824 + }, + { + "epoch": 0.8678920045699625, + "grad_norm": 1.8073760271072388, + "learning_rate": 1.2594774417773938e-05, + "loss": 3.2704, + "step": 10825 + }, + { + "epoch": 0.8679721793509851, + "grad_norm": 1.8173904418945312, + "learning_rate": 1.2593520421738157e-05, + "loss": 3.2137, + "step": 10826 + }, + { + "epoch": 0.8680523541320078, + "grad_norm": 1.7296096086502075, + "learning_rate": 1.2592266381976535e-05, + "loss": 3.2996, + "step": 10827 + }, + { + "epoch": 0.8681325289130304, + "grad_norm": 1.809976577758789, + "learning_rate": 1.2591012298510214e-05, + "loss": 3.1168, + "step": 10828 + }, + { + "epoch": 0.868212703694053, + "grad_norm": 1.6791527271270752, + "learning_rate": 1.2589758171360338e-05, + "loss": 3.2224, + "step": 10829 + }, + { + "epoch": 0.8682928784750756, + "grad_norm": 1.8717913627624512, + "learning_rate": 1.2588504000548051e-05, + "loss": 3.2339, + "step": 10830 + }, + { + "epoch": 0.8683730532560983, + "grad_norm": 1.7212326526641846, + "learning_rate": 1.2587249786094505e-05, + "loss": 3.2021, + "step": 10831 + }, + { + "epoch": 0.8684532280371209, + "grad_norm": 0.785858154296875, + "learning_rate": 1.2585995528020833e-05, + "loss": 2.4863, + "step": 10832 + }, + { + "epoch": 0.8685334028181435, + "grad_norm": 1.8134844303131104, + "learning_rate": 1.2584741226348192e-05, + "loss": 3.198, + "step": 10833 + }, + { + "epoch": 0.8686135775991661, + "grad_norm": 1.6957952976226807, + "learning_rate": 1.2583486881097721e-05, + "loss": 3.2407, + "step": 10834 + }, + { + "epoch": 0.8686937523801889, + "grad_norm": 1.739548921585083, + "learning_rate": 1.2582232492290574e-05, + "loss": 3.2263, + "step": 10835 + }, + { + "epoch": 0.8687739271612115, + "grad_norm": 1.6771924495697021, + "learning_rate": 1.2580978059947896e-05, + "loss": 3.1752, + "step": 10836 + }, + { + "epoch": 0.8688541019422341, + "grad_norm": 1.7394088506698608, + "learning_rate": 1.257972358409084e-05, + "loss": 3.1769, + "step": 10837 + }, + { + "epoch": 0.8689342767232567, + "grad_norm": 1.7025947570800781, + "learning_rate": 1.2578469064740553e-05, + "loss": 3.1593, + "step": 10838 + }, + { + "epoch": 0.8690144515042794, + "grad_norm": 1.84097421169281, + "learning_rate": 1.2577214501918184e-05, + "loss": 3.2371, + "step": 10839 + }, + { + "epoch": 0.869094626285302, + "grad_norm": 1.7337496280670166, + "learning_rate": 1.2575959895644887e-05, + "loss": 3.1174, + "step": 10840 + }, + { + "epoch": 0.8691748010663246, + "grad_norm": 1.8072476387023926, + "learning_rate": 1.2574705245941815e-05, + "loss": 3.278, + "step": 10841 + }, + { + "epoch": 0.8692549758473472, + "grad_norm": 1.8143091201782227, + "learning_rate": 1.257345055283012e-05, + "loss": 3.294, + "step": 10842 + }, + { + "epoch": 0.8693351506283699, + "grad_norm": 1.7344552278518677, + "learning_rate": 1.2572195816330957e-05, + "loss": 3.2021, + "step": 10843 + }, + { + "epoch": 0.8694153254093925, + "grad_norm": 1.7122440338134766, + "learning_rate": 1.2570941036465479e-05, + "loss": 3.2098, + "step": 10844 + }, + { + "epoch": 0.8694955001904151, + "grad_norm": 1.7340333461761475, + "learning_rate": 1.2569686213254837e-05, + "loss": 3.1393, + "step": 10845 + }, + { + "epoch": 0.8695756749714377, + "grad_norm": 1.7759065628051758, + "learning_rate": 1.2568431346720195e-05, + "loss": 3.0657, + "step": 10846 + }, + { + "epoch": 0.8696558497524604, + "grad_norm": 1.797177791595459, + "learning_rate": 1.2567176436882704e-05, + "loss": 3.1221, + "step": 10847 + }, + { + "epoch": 0.869736024533483, + "grad_norm": 1.742689609527588, + "learning_rate": 1.2565921483763525e-05, + "loss": 3.1834, + "step": 10848 + }, + { + "epoch": 0.8698161993145056, + "grad_norm": 1.787771463394165, + "learning_rate": 1.2564666487383814e-05, + "loss": 3.1475, + "step": 10849 + }, + { + "epoch": 0.8698963740955282, + "grad_norm": 0.7573091387748718, + "learning_rate": 1.2563411447764725e-05, + "loss": 2.4941, + "step": 10850 + }, + { + "epoch": 0.8699765488765508, + "grad_norm": 1.7932777404785156, + "learning_rate": 1.256215636492743e-05, + "loss": 3.2233, + "step": 10851 + }, + { + "epoch": 0.8700567236575735, + "grad_norm": 1.7725218534469604, + "learning_rate": 1.2560901238893078e-05, + "loss": 3.1323, + "step": 10852 + }, + { + "epoch": 0.8701368984385961, + "grad_norm": 1.7908310890197754, + "learning_rate": 1.2559646069682834e-05, + "loss": 3.2863, + "step": 10853 + }, + { + "epoch": 0.8702170732196187, + "grad_norm": 1.713352918624878, + "learning_rate": 1.2558390857317858e-05, + "loss": 3.1251, + "step": 10854 + }, + { + "epoch": 0.8702972480006413, + "grad_norm": 1.7189089059829712, + "learning_rate": 1.2557135601819312e-05, + "loss": 3.2684, + "step": 10855 + }, + { + "epoch": 0.8703774227816641, + "grad_norm": 1.8037679195404053, + "learning_rate": 1.2555880303208367e-05, + "loss": 3.2011, + "step": 10856 + }, + { + "epoch": 0.8704575975626867, + "grad_norm": 1.890142560005188, + "learning_rate": 1.2554624961506176e-05, + "loss": 3.2487, + "step": 10857 + }, + { + "epoch": 0.8705377723437093, + "grad_norm": 1.7492128610610962, + "learning_rate": 1.2553369576733911e-05, + "loss": 3.2154, + "step": 10858 + }, + { + "epoch": 0.870617947124732, + "grad_norm": 1.7953636646270752, + "learning_rate": 1.2552114148912735e-05, + "loss": 3.1979, + "step": 10859 + }, + { + "epoch": 0.8706981219057546, + "grad_norm": 1.810553789138794, + "learning_rate": 1.255085867806381e-05, + "loss": 3.1611, + "step": 10860 + }, + { + "epoch": 0.8707782966867772, + "grad_norm": 1.9986134767532349, + "learning_rate": 1.2549603164208313e-05, + "loss": 3.2522, + "step": 10861 + }, + { + "epoch": 0.8708584714677998, + "grad_norm": 1.7992579936981201, + "learning_rate": 1.2548347607367399e-05, + "loss": 3.1082, + "step": 10862 + }, + { + "epoch": 0.8709386462488224, + "grad_norm": 1.8260366916656494, + "learning_rate": 1.2547092007562248e-05, + "loss": 3.15, + "step": 10863 + }, + { + "epoch": 0.8710188210298451, + "grad_norm": 1.8295189142227173, + "learning_rate": 1.2545836364814019e-05, + "loss": 3.2001, + "step": 10864 + }, + { + "epoch": 0.8710989958108677, + "grad_norm": 1.7863280773162842, + "learning_rate": 1.2544580679143887e-05, + "loss": 3.1613, + "step": 10865 + }, + { + "epoch": 0.8711791705918903, + "grad_norm": 1.72101891040802, + "learning_rate": 1.2543324950573024e-05, + "loss": 3.1321, + "step": 10866 + }, + { + "epoch": 0.8712593453729129, + "grad_norm": 1.7424346208572388, + "learning_rate": 1.2542069179122596e-05, + "loss": 3.207, + "step": 10867 + }, + { + "epoch": 0.8713395201539356, + "grad_norm": 1.767008900642395, + "learning_rate": 1.254081336481378e-05, + "loss": 3.0979, + "step": 10868 + }, + { + "epoch": 0.8714196949349582, + "grad_norm": 1.8393800258636475, + "learning_rate": 1.2539557507667743e-05, + "loss": 3.1753, + "step": 10869 + }, + { + "epoch": 0.8714998697159808, + "grad_norm": 1.770048975944519, + "learning_rate": 1.2538301607705663e-05, + "loss": 3.2117, + "step": 10870 + }, + { + "epoch": 0.8715800444970034, + "grad_norm": 1.8098068237304688, + "learning_rate": 1.2537045664948714e-05, + "loss": 3.2476, + "step": 10871 + }, + { + "epoch": 0.8716602192780261, + "grad_norm": 1.7701212167739868, + "learning_rate": 1.2535789679418067e-05, + "loss": 3.2283, + "step": 10872 + }, + { + "epoch": 0.8717403940590487, + "grad_norm": 1.7469563484191895, + "learning_rate": 1.2534533651134901e-05, + "loss": 3.1436, + "step": 10873 + }, + { + "epoch": 0.8718205688400713, + "grad_norm": 1.917400598526001, + "learning_rate": 1.2533277580120392e-05, + "loss": 3.2936, + "step": 10874 + }, + { + "epoch": 0.8719007436210939, + "grad_norm": 1.8058195114135742, + "learning_rate": 1.2532021466395711e-05, + "loss": 3.0846, + "step": 10875 + }, + { + "epoch": 0.8719809184021167, + "grad_norm": 1.755465030670166, + "learning_rate": 1.2530765309982047e-05, + "loss": 3.1975, + "step": 10876 + }, + { + "epoch": 0.8720610931831393, + "grad_norm": 1.8636828660964966, + "learning_rate": 1.2529509110900566e-05, + "loss": 3.2889, + "step": 10877 + }, + { + "epoch": 0.8721412679641619, + "grad_norm": 1.7883461713790894, + "learning_rate": 1.2528252869172458e-05, + "loss": 3.1349, + "step": 10878 + }, + { + "epoch": 0.8722214427451845, + "grad_norm": 1.8714253902435303, + "learning_rate": 1.2526996584818894e-05, + "loss": 3.2137, + "step": 10879 + }, + { + "epoch": 0.8723016175262072, + "grad_norm": 1.655578851699829, + "learning_rate": 1.2525740257861058e-05, + "loss": 3.0423, + "step": 10880 + }, + { + "epoch": 0.8723817923072298, + "grad_norm": 1.8314509391784668, + "learning_rate": 1.2524483888320134e-05, + "loss": 3.101, + "step": 10881 + }, + { + "epoch": 0.8724619670882524, + "grad_norm": 1.8836842775344849, + "learning_rate": 1.25232274762173e-05, + "loss": 3.3238, + "step": 10882 + }, + { + "epoch": 0.872542141869275, + "grad_norm": 0.697581946849823, + "learning_rate": 1.252197102157374e-05, + "loss": 2.4883, + "step": 10883 + }, + { + "epoch": 0.8726223166502977, + "grad_norm": 2.1577882766723633, + "learning_rate": 1.2520714524410637e-05, + "loss": 3.2699, + "step": 10884 + }, + { + "epoch": 0.8727024914313203, + "grad_norm": 1.8891717195510864, + "learning_rate": 1.2519457984749175e-05, + "loss": 3.1799, + "step": 10885 + }, + { + "epoch": 0.8727826662123429, + "grad_norm": 1.789390206336975, + "learning_rate": 1.2518201402610541e-05, + "loss": 3.1155, + "step": 10886 + }, + { + "epoch": 0.8728628409933655, + "grad_norm": 1.8045953512191772, + "learning_rate": 1.2516944778015917e-05, + "loss": 3.1893, + "step": 10887 + }, + { + "epoch": 0.8729430157743882, + "grad_norm": 1.6978545188903809, + "learning_rate": 1.2515688110986495e-05, + "loss": 3.1901, + "step": 10888 + }, + { + "epoch": 0.8730231905554108, + "grad_norm": 1.7554430961608887, + "learning_rate": 1.2514431401543455e-05, + "loss": 3.1109, + "step": 10889 + }, + { + "epoch": 0.8731033653364334, + "grad_norm": 1.8439265489578247, + "learning_rate": 1.2513174649707987e-05, + "loss": 3.1948, + "step": 10890 + }, + { + "epoch": 0.873183540117456, + "grad_norm": 1.696990728378296, + "learning_rate": 1.251191785550128e-05, + "loss": 3.194, + "step": 10891 + }, + { + "epoch": 0.8732637148984786, + "grad_norm": 1.7217565774917603, + "learning_rate": 1.2510661018944526e-05, + "loss": 3.185, + "step": 10892 + }, + { + "epoch": 0.8733438896795013, + "grad_norm": 1.7792118787765503, + "learning_rate": 1.2509404140058909e-05, + "loss": 3.1714, + "step": 10893 + }, + { + "epoch": 0.8734240644605239, + "grad_norm": 1.8205863237380981, + "learning_rate": 1.2508147218865624e-05, + "loss": 3.1244, + "step": 10894 + }, + { + "epoch": 0.8735042392415465, + "grad_norm": 1.7656700611114502, + "learning_rate": 1.250689025538586e-05, + "loss": 3.1982, + "step": 10895 + }, + { + "epoch": 0.8735844140225693, + "grad_norm": 1.8136931657791138, + "learning_rate": 1.2505633249640811e-05, + "loss": 3.2493, + "step": 10896 + }, + { + "epoch": 0.8736645888035919, + "grad_norm": 1.7555347681045532, + "learning_rate": 1.250437620165167e-05, + "loss": 3.1462, + "step": 10897 + }, + { + "epoch": 0.8737447635846145, + "grad_norm": 1.7362518310546875, + "learning_rate": 1.2503119111439628e-05, + "loss": 3.2133, + "step": 10898 + }, + { + "epoch": 0.8738249383656371, + "grad_norm": 1.743996262550354, + "learning_rate": 1.2501861979025878e-05, + "loss": 3.2317, + "step": 10899 + }, + { + "epoch": 0.8739051131466598, + "grad_norm": 1.8772228956222534, + "learning_rate": 1.2500604804431618e-05, + "loss": 3.1957, + "step": 10900 + }, + { + "epoch": 0.8739852879276824, + "grad_norm": 1.7321012020111084, + "learning_rate": 1.2499347587678041e-05, + "loss": 3.2098, + "step": 10901 + }, + { + "epoch": 0.874065462708705, + "grad_norm": 1.792399287223816, + "learning_rate": 1.2498090328786346e-05, + "loss": 3.2204, + "step": 10902 + }, + { + "epoch": 0.8741456374897276, + "grad_norm": 1.7880470752716064, + "learning_rate": 1.2496833027777728e-05, + "loss": 3.309, + "step": 10903 + }, + { + "epoch": 0.8742258122707502, + "grad_norm": 1.8965644836425781, + "learning_rate": 1.2495575684673385e-05, + "loss": 3.1525, + "step": 10904 + }, + { + "epoch": 0.8743059870517729, + "grad_norm": 1.7632478475570679, + "learning_rate": 1.2494318299494515e-05, + "loss": 3.0858, + "step": 10905 + }, + { + "epoch": 0.8743861618327955, + "grad_norm": 1.8133025169372559, + "learning_rate": 1.249306087226232e-05, + "loss": 3.2377, + "step": 10906 + }, + { + "epoch": 0.8744663366138181, + "grad_norm": 1.8711086511611938, + "learning_rate": 1.2491803402997994e-05, + "loss": 3.1699, + "step": 10907 + }, + { + "epoch": 0.8745465113948407, + "grad_norm": 1.7820037603378296, + "learning_rate": 1.2490545891722746e-05, + "loss": 3.2103, + "step": 10908 + }, + { + "epoch": 0.8746266861758634, + "grad_norm": 1.9174413681030273, + "learning_rate": 1.2489288338457765e-05, + "loss": 3.226, + "step": 10909 + }, + { + "epoch": 0.874706860956886, + "grad_norm": 0.7084676027297974, + "learning_rate": 1.2488030743224264e-05, + "loss": 2.4869, + "step": 10910 + }, + { + "epoch": 0.8747870357379086, + "grad_norm": 0.6937839388847351, + "learning_rate": 1.2486773106043441e-05, + "loss": 2.4833, + "step": 10911 + }, + { + "epoch": 0.8748672105189312, + "grad_norm": 1.76421058177948, + "learning_rate": 1.2485515426936502e-05, + "loss": 3.0908, + "step": 10912 + }, + { + "epoch": 0.8749473852999539, + "grad_norm": 1.7394753694534302, + "learning_rate": 1.2484257705924645e-05, + "loss": 3.1652, + "step": 10913 + }, + { + "epoch": 0.8750275600809765, + "grad_norm": 1.748540997505188, + "learning_rate": 1.248299994302908e-05, + "loss": 3.1361, + "step": 10914 + }, + { + "epoch": 0.8751077348619991, + "grad_norm": 1.8216428756713867, + "learning_rate": 1.2481742138271011e-05, + "loss": 3.2775, + "step": 10915 + }, + { + "epoch": 0.8751879096430217, + "grad_norm": 1.739892840385437, + "learning_rate": 1.2480484291671646e-05, + "loss": 3.2152, + "step": 10916 + }, + { + "epoch": 0.8752680844240445, + "grad_norm": 1.660156011581421, + "learning_rate": 1.2479226403252186e-05, + "loss": 3.2106, + "step": 10917 + }, + { + "epoch": 0.8753482592050671, + "grad_norm": 1.7540524005889893, + "learning_rate": 1.2477968473033848e-05, + "loss": 3.2235, + "step": 10918 + }, + { + "epoch": 0.8754284339860897, + "grad_norm": 1.8261919021606445, + "learning_rate": 1.247671050103783e-05, + "loss": 3.2985, + "step": 10919 + }, + { + "epoch": 0.8755086087671123, + "grad_norm": 1.9071648120880127, + "learning_rate": 1.2475452487285349e-05, + "loss": 3.2524, + "step": 10920 + }, + { + "epoch": 0.875588783548135, + "grad_norm": 1.8190969228744507, + "learning_rate": 1.247419443179761e-05, + "loss": 3.193, + "step": 10921 + }, + { + "epoch": 0.8756689583291576, + "grad_norm": 2.494943380355835, + "learning_rate": 1.2472936334595824e-05, + "loss": 3.2062, + "step": 10922 + }, + { + "epoch": 0.8757491331101802, + "grad_norm": 1.724643588066101, + "learning_rate": 1.2471678195701207e-05, + "loss": 3.2749, + "step": 10923 + }, + { + "epoch": 0.8758293078912028, + "grad_norm": 1.7813457250595093, + "learning_rate": 1.2470420015134962e-05, + "loss": 3.1241, + "step": 10924 + }, + { + "epoch": 0.8759094826722255, + "grad_norm": 1.8972069025039673, + "learning_rate": 1.2469161792918307e-05, + "loss": 3.1892, + "step": 10925 + }, + { + "epoch": 0.8759896574532481, + "grad_norm": 1.7354626655578613, + "learning_rate": 1.2467903529072459e-05, + "loss": 3.2424, + "step": 10926 + }, + { + "epoch": 0.8760698322342707, + "grad_norm": 1.6873797178268433, + "learning_rate": 1.246664522361862e-05, + "loss": 3.1615, + "step": 10927 + }, + { + "epoch": 0.8761500070152933, + "grad_norm": 1.7298097610473633, + "learning_rate": 1.2465386876578018e-05, + "loss": 3.2482, + "step": 10928 + }, + { + "epoch": 0.876230181796316, + "grad_norm": 1.7852706909179688, + "learning_rate": 1.2464128487971859e-05, + "loss": 3.2484, + "step": 10929 + }, + { + "epoch": 0.8763103565773386, + "grad_norm": 1.7870274782180786, + "learning_rate": 1.2462870057821361e-05, + "loss": 3.1312, + "step": 10930 + }, + { + "epoch": 0.8763905313583612, + "grad_norm": 1.8025513887405396, + "learning_rate": 1.2461611586147746e-05, + "loss": 3.201, + "step": 10931 + }, + { + "epoch": 0.8764707061393838, + "grad_norm": 1.7379965782165527, + "learning_rate": 1.2460353072972224e-05, + "loss": 3.0891, + "step": 10932 + }, + { + "epoch": 0.8765508809204064, + "grad_norm": 1.7617244720458984, + "learning_rate": 1.245909451831602e-05, + "loss": 3.1478, + "step": 10933 + }, + { + "epoch": 0.8766310557014291, + "grad_norm": 1.8538763523101807, + "learning_rate": 1.2457835922200344e-05, + "loss": 3.2159, + "step": 10934 + }, + { + "epoch": 0.8767112304824517, + "grad_norm": 1.6671115159988403, + "learning_rate": 1.2456577284646422e-05, + "loss": 3.09, + "step": 10935 + }, + { + "epoch": 0.8767914052634743, + "grad_norm": 1.7674977779388428, + "learning_rate": 1.2455318605675475e-05, + "loss": 3.1421, + "step": 10936 + }, + { + "epoch": 0.8768715800444971, + "grad_norm": 1.7828408479690552, + "learning_rate": 1.245405988530872e-05, + "loss": 3.1584, + "step": 10937 + }, + { + "epoch": 0.8769517548255197, + "grad_norm": 1.7324072122573853, + "learning_rate": 1.2452801123567382e-05, + "loss": 3.1362, + "step": 10938 + }, + { + "epoch": 0.8770319296065423, + "grad_norm": 1.8578660488128662, + "learning_rate": 1.2451542320472679e-05, + "loss": 3.204, + "step": 10939 + }, + { + "epoch": 0.8771121043875649, + "grad_norm": 1.774056077003479, + "learning_rate": 1.2450283476045837e-05, + "loss": 3.1375, + "step": 10940 + }, + { + "epoch": 0.8771922791685876, + "grad_norm": 1.7836967706680298, + "learning_rate": 1.244902459030808e-05, + "loss": 3.1917, + "step": 10941 + }, + { + "epoch": 0.8772724539496102, + "grad_norm": 1.8793237209320068, + "learning_rate": 1.244776566328063e-05, + "loss": 3.3268, + "step": 10942 + }, + { + "epoch": 0.8773526287306328, + "grad_norm": 1.7341307401657104, + "learning_rate": 1.2446506694984717e-05, + "loss": 3.1118, + "step": 10943 + }, + { + "epoch": 0.8774328035116554, + "grad_norm": 1.8584367036819458, + "learning_rate": 1.2445247685441559e-05, + "loss": 3.2059, + "step": 10944 + }, + { + "epoch": 0.877512978292678, + "grad_norm": 1.7482342720031738, + "learning_rate": 1.244398863467239e-05, + "loss": 3.1252, + "step": 10945 + }, + { + "epoch": 0.8775931530737007, + "grad_norm": 1.6826924085617065, + "learning_rate": 1.2442729542698432e-05, + "loss": 3.2469, + "step": 10946 + }, + { + "epoch": 0.8776733278547233, + "grad_norm": 1.8757680654525757, + "learning_rate": 1.2441470409540917e-05, + "loss": 3.2168, + "step": 10947 + }, + { + "epoch": 0.8777535026357459, + "grad_norm": 1.7720307111740112, + "learning_rate": 1.2440211235221071e-05, + "loss": 3.2167, + "step": 10948 + }, + { + "epoch": 0.8778336774167685, + "grad_norm": 1.7748582363128662, + "learning_rate": 1.2438952019760123e-05, + "loss": 3.2206, + "step": 10949 + }, + { + "epoch": 0.8779138521977912, + "grad_norm": 1.7844849824905396, + "learning_rate": 1.24376927631793e-05, + "loss": 3.1743, + "step": 10950 + }, + { + "epoch": 0.8779940269788138, + "grad_norm": 1.8175992965698242, + "learning_rate": 1.2436433465499842e-05, + "loss": 3.2593, + "step": 10951 + }, + { + "epoch": 0.8780742017598364, + "grad_norm": 0.712280809879303, + "learning_rate": 1.2435174126742971e-05, + "loss": 2.4808, + "step": 10952 + }, + { + "epoch": 0.878154376540859, + "grad_norm": 1.8004494905471802, + "learning_rate": 1.2433914746929924e-05, + "loss": 3.1822, + "step": 10953 + }, + { + "epoch": 0.8782345513218817, + "grad_norm": 1.8456670045852661, + "learning_rate": 1.243265532608193e-05, + "loss": 3.2147, + "step": 10954 + }, + { + "epoch": 0.8783147261029043, + "grad_norm": 0.7115821838378906, + "learning_rate": 1.2431395864220226e-05, + "loss": 2.4865, + "step": 10955 + }, + { + "epoch": 0.8783949008839269, + "grad_norm": 1.868863582611084, + "learning_rate": 1.2430136361366046e-05, + "loss": 3.2751, + "step": 10956 + }, + { + "epoch": 0.8784750756649496, + "grad_norm": 1.8691964149475098, + "learning_rate": 1.242887681754062e-05, + "loss": 3.2168, + "step": 10957 + }, + { + "epoch": 0.8785552504459723, + "grad_norm": 1.6822839975357056, + "learning_rate": 1.2427617232765192e-05, + "loss": 3.115, + "step": 10958 + }, + { + "epoch": 0.8786354252269949, + "grad_norm": 1.8703073263168335, + "learning_rate": 1.2426357607060987e-05, + "loss": 3.2201, + "step": 10959 + }, + { + "epoch": 0.8787156000080175, + "grad_norm": 1.8737928867340088, + "learning_rate": 1.242509794044925e-05, + "loss": 3.2002, + "step": 10960 + }, + { + "epoch": 0.8787957747890401, + "grad_norm": 1.963731288909912, + "learning_rate": 1.242383823295122e-05, + "loss": 3.1897, + "step": 10961 + }, + { + "epoch": 0.8788759495700628, + "grad_norm": 1.6963863372802734, + "learning_rate": 1.2422578484588129e-05, + "loss": 3.1504, + "step": 10962 + }, + { + "epoch": 0.8789561243510854, + "grad_norm": 1.734310507774353, + "learning_rate": 1.2421318695381219e-05, + "loss": 3.1925, + "step": 10963 + }, + { + "epoch": 0.879036299132108, + "grad_norm": 1.7763938903808594, + "learning_rate": 1.2420058865351727e-05, + "loss": 3.2423, + "step": 10964 + }, + { + "epoch": 0.8791164739131306, + "grad_norm": 1.82472562789917, + "learning_rate": 1.2418798994520897e-05, + "loss": 3.1689, + "step": 10965 + }, + { + "epoch": 0.8791966486941533, + "grad_norm": 1.8690351247787476, + "learning_rate": 1.2417539082909968e-05, + "loss": 3.1682, + "step": 10966 + }, + { + "epoch": 0.8792768234751759, + "grad_norm": 1.8368207216262817, + "learning_rate": 1.2416279130540184e-05, + "loss": 3.1807, + "step": 10967 + }, + { + "epoch": 0.8793569982561985, + "grad_norm": 1.8654046058654785, + "learning_rate": 1.2415019137432784e-05, + "loss": 3.2272, + "step": 10968 + }, + { + "epoch": 0.8794371730372211, + "grad_norm": 1.9377108812332153, + "learning_rate": 1.2413759103609012e-05, + "loss": 3.2461, + "step": 10969 + }, + { + "epoch": 0.8795173478182438, + "grad_norm": 1.7705305814743042, + "learning_rate": 1.241249902909011e-05, + "loss": 3.224, + "step": 10970 + }, + { + "epoch": 0.8795975225992664, + "grad_norm": 1.8908649682998657, + "learning_rate": 1.241123891389733e-05, + "loss": 3.2874, + "step": 10971 + }, + { + "epoch": 0.879677697380289, + "grad_norm": 0.7517582178115845, + "learning_rate": 1.2409978758051909e-05, + "loss": 2.4924, + "step": 10972 + }, + { + "epoch": 0.8797578721613116, + "grad_norm": 1.7892441749572754, + "learning_rate": 1.2408718561575097e-05, + "loss": 3.1815, + "step": 10973 + }, + { + "epoch": 0.8798380469423343, + "grad_norm": 1.6563163995742798, + "learning_rate": 1.2407458324488135e-05, + "loss": 3.1426, + "step": 10974 + }, + { + "epoch": 0.8799182217233569, + "grad_norm": 1.752623438835144, + "learning_rate": 1.2406198046812276e-05, + "loss": 3.1547, + "step": 10975 + }, + { + "epoch": 0.8799983965043795, + "grad_norm": 1.6704765558242798, + "learning_rate": 1.2404937728568769e-05, + "loss": 3.1808, + "step": 10976 + }, + { + "epoch": 0.8800785712854021, + "grad_norm": 1.8671479225158691, + "learning_rate": 1.2403677369778855e-05, + "loss": 3.1365, + "step": 10977 + }, + { + "epoch": 0.8801587460664249, + "grad_norm": 1.748812198638916, + "learning_rate": 1.2402416970463792e-05, + "loss": 3.1237, + "step": 10978 + }, + { + "epoch": 0.8802389208474475, + "grad_norm": 1.6131870746612549, + "learning_rate": 1.2401156530644822e-05, + "loss": 3.1421, + "step": 10979 + }, + { + "epoch": 0.8803190956284701, + "grad_norm": 1.8325213193893433, + "learning_rate": 1.2399896050343199e-05, + "loss": 3.2081, + "step": 10980 + }, + { + "epoch": 0.8803992704094927, + "grad_norm": 0.7347011566162109, + "learning_rate": 1.2398635529580177e-05, + "loss": 2.4703, + "step": 10981 + }, + { + "epoch": 0.8804794451905154, + "grad_norm": 1.8703184127807617, + "learning_rate": 1.2397374968377003e-05, + "loss": 3.174, + "step": 10982 + }, + { + "epoch": 0.880559619971538, + "grad_norm": 1.7459710836410522, + "learning_rate": 1.2396114366754934e-05, + "loss": 3.1077, + "step": 10983 + }, + { + "epoch": 0.8806397947525606, + "grad_norm": 1.812933087348938, + "learning_rate": 1.239485372473522e-05, + "loss": 3.3036, + "step": 10984 + }, + { + "epoch": 0.8807199695335832, + "grad_norm": 1.8027886152267456, + "learning_rate": 1.2393593042339117e-05, + "loss": 3.1871, + "step": 10985 + }, + { + "epoch": 0.8808001443146058, + "grad_norm": 1.9909754991531372, + "learning_rate": 1.2392332319587878e-05, + "loss": 3.2557, + "step": 10986 + }, + { + "epoch": 0.8808803190956285, + "grad_norm": 1.8327577114105225, + "learning_rate": 1.239107155650276e-05, + "loss": 3.2595, + "step": 10987 + }, + { + "epoch": 0.8809604938766511, + "grad_norm": 1.705098032951355, + "learning_rate": 1.2389810753105018e-05, + "loss": 3.1876, + "step": 10988 + }, + { + "epoch": 0.8810406686576737, + "grad_norm": 1.809954047203064, + "learning_rate": 1.2388549909415908e-05, + "loss": 3.2064, + "step": 10989 + }, + { + "epoch": 0.8811208434386963, + "grad_norm": 1.8038287162780762, + "learning_rate": 1.238728902545669e-05, + "loss": 3.2363, + "step": 10990 + }, + { + "epoch": 0.881201018219719, + "grad_norm": 1.7667202949523926, + "learning_rate": 1.238602810124862e-05, + "loss": 3.282, + "step": 10991 + }, + { + "epoch": 0.8812811930007416, + "grad_norm": 1.7801740169525146, + "learning_rate": 1.2384767136812957e-05, + "loss": 3.2251, + "step": 10992 + }, + { + "epoch": 0.8813613677817642, + "grad_norm": 1.8318108320236206, + "learning_rate": 1.2383506132170959e-05, + "loss": 3.2659, + "step": 10993 + }, + { + "epoch": 0.8814415425627868, + "grad_norm": 1.833338975906372, + "learning_rate": 1.2382245087343889e-05, + "loss": 3.2657, + "step": 10994 + }, + { + "epoch": 0.8815217173438095, + "grad_norm": 1.7731176614761353, + "learning_rate": 1.2380984002353005e-05, + "loss": 3.2394, + "step": 10995 + }, + { + "epoch": 0.8816018921248321, + "grad_norm": 1.8004735708236694, + "learning_rate": 1.2379722877219573e-05, + "loss": 3.251, + "step": 10996 + }, + { + "epoch": 0.8816820669058547, + "grad_norm": 1.6819885969161987, + "learning_rate": 1.2378461711964848e-05, + "loss": 3.1622, + "step": 10997 + }, + { + "epoch": 0.8817622416868774, + "grad_norm": 0.7371557354927063, + "learning_rate": 1.2377200506610102e-05, + "loss": 2.4836, + "step": 10998 + }, + { + "epoch": 0.8818424164679001, + "grad_norm": 1.759204387664795, + "learning_rate": 1.237593926117659e-05, + "loss": 3.1268, + "step": 10999 + }, + { + "epoch": 0.8819225912489227, + "grad_norm": 1.757783055305481, + "learning_rate": 1.2374677975685579e-05, + "loss": 3.1759, + "step": 11000 + }, + { + "epoch": 0.8820027660299453, + "grad_norm": 1.7470451593399048, + "learning_rate": 1.2373416650158334e-05, + "loss": 3.1663, + "step": 11001 + }, + { + "epoch": 0.8820829408109679, + "grad_norm": 1.8541926145553589, + "learning_rate": 1.2372155284616123e-05, + "loss": 3.2666, + "step": 11002 + }, + { + "epoch": 0.8821631155919906, + "grad_norm": 1.7785508632659912, + "learning_rate": 1.2370893879080206e-05, + "loss": 3.2705, + "step": 11003 + }, + { + "epoch": 0.8822432903730132, + "grad_norm": 1.8574658632278442, + "learning_rate": 1.236963243357186e-05, + "loss": 3.2064, + "step": 11004 + }, + { + "epoch": 0.8823234651540358, + "grad_norm": 1.8173147439956665, + "learning_rate": 1.2368370948112341e-05, + "loss": 3.1432, + "step": 11005 + }, + { + "epoch": 0.8824036399350584, + "grad_norm": 1.7589945793151855, + "learning_rate": 1.2367109422722924e-05, + "loss": 3.2081, + "step": 11006 + }, + { + "epoch": 0.8824838147160811, + "grad_norm": 1.6976631879806519, + "learning_rate": 1.2365847857424877e-05, + "loss": 3.1995, + "step": 11007 + }, + { + "epoch": 0.8825639894971037, + "grad_norm": 1.7827813625335693, + "learning_rate": 1.2364586252239468e-05, + "loss": 3.2189, + "step": 11008 + }, + { + "epoch": 0.8826441642781263, + "grad_norm": 1.823121428489685, + "learning_rate": 1.236332460718797e-05, + "loss": 3.1967, + "step": 11009 + }, + { + "epoch": 0.8827243390591489, + "grad_norm": 1.6441112756729126, + "learning_rate": 1.236206292229165e-05, + "loss": 3.151, + "step": 11010 + }, + { + "epoch": 0.8828045138401716, + "grad_norm": 1.970677375793457, + "learning_rate": 1.2360801197571783e-05, + "loss": 3.2251, + "step": 11011 + }, + { + "epoch": 0.8828846886211942, + "grad_norm": 1.8025363683700562, + "learning_rate": 1.2359539433049639e-05, + "loss": 3.2263, + "step": 11012 + }, + { + "epoch": 0.8829648634022168, + "grad_norm": 1.7971566915512085, + "learning_rate": 1.2358277628746489e-05, + "loss": 3.1959, + "step": 11013 + }, + { + "epoch": 0.8830450381832394, + "grad_norm": 1.8359415531158447, + "learning_rate": 1.2357015784683614e-05, + "loss": 3.2977, + "step": 11014 + }, + { + "epoch": 0.883125212964262, + "grad_norm": 1.8288097381591797, + "learning_rate": 1.2355753900882282e-05, + "loss": 3.2398, + "step": 11015 + }, + { + "epoch": 0.8832053877452847, + "grad_norm": 1.894431710243225, + "learning_rate": 1.235449197736377e-05, + "loss": 3.1742, + "step": 11016 + }, + { + "epoch": 0.8832855625263073, + "grad_norm": 1.797176718711853, + "learning_rate": 1.2353230014149351e-05, + "loss": 3.2098, + "step": 11017 + }, + { + "epoch": 0.88336573730733, + "grad_norm": 1.8486075401306152, + "learning_rate": 1.2351968011260304e-05, + "loss": 3.2183, + "step": 11018 + }, + { + "epoch": 0.8834459120883527, + "grad_norm": 1.6686822175979614, + "learning_rate": 1.2350705968717906e-05, + "loss": 3.1182, + "step": 11019 + }, + { + "epoch": 0.8835260868693753, + "grad_norm": 1.787685513496399, + "learning_rate": 1.2349443886543433e-05, + "loss": 3.1505, + "step": 11020 + }, + { + "epoch": 0.8836062616503979, + "grad_norm": 1.8773627281188965, + "learning_rate": 1.2348181764758167e-05, + "loss": 3.2103, + "step": 11021 + }, + { + "epoch": 0.8836864364314205, + "grad_norm": 1.7828813791275024, + "learning_rate": 1.234691960338338e-05, + "loss": 3.2155, + "step": 11022 + }, + { + "epoch": 0.8837666112124432, + "grad_norm": 1.7862961292266846, + "learning_rate": 1.2345657402440356e-05, + "loss": 3.1299, + "step": 11023 + }, + { + "epoch": 0.8838467859934658, + "grad_norm": 1.8678228855133057, + "learning_rate": 1.234439516195038e-05, + "loss": 3.2417, + "step": 11024 + }, + { + "epoch": 0.8839269607744884, + "grad_norm": 1.727401614189148, + "learning_rate": 1.2343132881934723e-05, + "loss": 3.2113, + "step": 11025 + }, + { + "epoch": 0.884007135555511, + "grad_norm": 1.735214352607727, + "learning_rate": 1.2341870562414674e-05, + "loss": 3.2148, + "step": 11026 + }, + { + "epoch": 0.8840873103365336, + "grad_norm": 1.7381868362426758, + "learning_rate": 1.234060820341151e-05, + "loss": 3.2846, + "step": 11027 + }, + { + "epoch": 0.8841674851175563, + "grad_norm": 1.8102750778198242, + "learning_rate": 1.2339345804946516e-05, + "loss": 3.3699, + "step": 11028 + }, + { + "epoch": 0.8842476598985789, + "grad_norm": 1.6958037614822388, + "learning_rate": 1.233808336704098e-05, + "loss": 3.1802, + "step": 11029 + }, + { + "epoch": 0.8843278346796015, + "grad_norm": 1.857238531112671, + "learning_rate": 1.2336820889716179e-05, + "loss": 3.243, + "step": 11030 + }, + { + "epoch": 0.8844080094606241, + "grad_norm": 1.7585314512252808, + "learning_rate": 1.2335558372993404e-05, + "loss": 3.1326, + "step": 11031 + }, + { + "epoch": 0.8844881842416468, + "grad_norm": 1.8085440397262573, + "learning_rate": 1.2334295816893937e-05, + "loss": 3.2322, + "step": 11032 + }, + { + "epoch": 0.8845683590226694, + "grad_norm": 1.6562626361846924, + "learning_rate": 1.2333033221439063e-05, + "loss": 3.0941, + "step": 11033 + }, + { + "epoch": 0.884648533803692, + "grad_norm": 1.676115870475769, + "learning_rate": 1.2331770586650077e-05, + "loss": 3.2448, + "step": 11034 + }, + { + "epoch": 0.8847287085847146, + "grad_norm": 1.8915321826934814, + "learning_rate": 1.2330507912548253e-05, + "loss": 3.175, + "step": 11035 + }, + { + "epoch": 0.8848088833657373, + "grad_norm": 1.7277333736419678, + "learning_rate": 1.2329245199154893e-05, + "loss": 3.2454, + "step": 11036 + }, + { + "epoch": 0.8848890581467599, + "grad_norm": 1.7953773736953735, + "learning_rate": 1.2327982446491275e-05, + "loss": 3.1575, + "step": 11037 + }, + { + "epoch": 0.8849692329277825, + "grad_norm": 1.730568528175354, + "learning_rate": 1.2326719654578697e-05, + "loss": 3.2023, + "step": 11038 + }, + { + "epoch": 0.8850494077088052, + "grad_norm": 1.7504585981369019, + "learning_rate": 1.2325456823438447e-05, + "loss": 3.2239, + "step": 11039 + }, + { + "epoch": 0.8851295824898279, + "grad_norm": 1.7623887062072754, + "learning_rate": 1.232419395309181e-05, + "loss": 3.1823, + "step": 11040 + }, + { + "epoch": 0.8852097572708505, + "grad_norm": 1.8547141551971436, + "learning_rate": 1.2322931043560087e-05, + "loss": 3.1825, + "step": 11041 + }, + { + "epoch": 0.8852899320518731, + "grad_norm": 1.7970011234283447, + "learning_rate": 1.2321668094864561e-05, + "loss": 3.3481, + "step": 11042 + }, + { + "epoch": 0.8853701068328957, + "grad_norm": 0.7852107882499695, + "learning_rate": 1.232040510702653e-05, + "loss": 2.4946, + "step": 11043 + }, + { + "epoch": 0.8854502816139184, + "grad_norm": 1.7771142721176147, + "learning_rate": 1.2319142080067287e-05, + "loss": 3.186, + "step": 11044 + }, + { + "epoch": 0.885530456394941, + "grad_norm": 1.7493929862976074, + "learning_rate": 1.2317879014008128e-05, + "loss": 3.2751, + "step": 11045 + }, + { + "epoch": 0.8856106311759636, + "grad_norm": 1.819213628768921, + "learning_rate": 1.2316615908870345e-05, + "loss": 3.3202, + "step": 11046 + }, + { + "epoch": 0.8856908059569862, + "grad_norm": 1.8342663049697876, + "learning_rate": 1.2315352764675232e-05, + "loss": 3.2661, + "step": 11047 + }, + { + "epoch": 0.8857709807380089, + "grad_norm": 0.7182245254516602, + "learning_rate": 1.2314089581444086e-05, + "loss": 2.476, + "step": 11048 + }, + { + "epoch": 0.8858511555190315, + "grad_norm": 1.8497288227081299, + "learning_rate": 1.231282635919821e-05, + "loss": 3.2545, + "step": 11049 + }, + { + "epoch": 0.8859313303000541, + "grad_norm": 1.7924814224243164, + "learning_rate": 1.2311563097958894e-05, + "loss": 3.2634, + "step": 11050 + }, + { + "epoch": 0.8860115050810767, + "grad_norm": 0.7010107636451721, + "learning_rate": 1.2310299797747441e-05, + "loss": 2.4913, + "step": 11051 + }, + { + "epoch": 0.8860916798620994, + "grad_norm": 1.7153644561767578, + "learning_rate": 1.2309036458585144e-05, + "loss": 3.3097, + "step": 11052 + }, + { + "epoch": 0.886171854643122, + "grad_norm": 1.7080644369125366, + "learning_rate": 1.2307773080493308e-05, + "loss": 3.0568, + "step": 11053 + }, + { + "epoch": 0.8862520294241446, + "grad_norm": 1.6935797929763794, + "learning_rate": 1.2306509663493233e-05, + "loss": 3.2008, + "step": 11054 + }, + { + "epoch": 0.8863322042051672, + "grad_norm": 0.7098868489265442, + "learning_rate": 1.2305246207606216e-05, + "loss": 2.4854, + "step": 11055 + }, + { + "epoch": 0.8864123789861899, + "grad_norm": 1.655839443206787, + "learning_rate": 1.2303982712853562e-05, + "loss": 3.2752, + "step": 11056 + }, + { + "epoch": 0.8864925537672125, + "grad_norm": 0.7171908617019653, + "learning_rate": 1.2302719179256568e-05, + "loss": 2.4891, + "step": 11057 + }, + { + "epoch": 0.8865727285482351, + "grad_norm": 1.8761367797851562, + "learning_rate": 1.230145560683654e-05, + "loss": 3.203, + "step": 11058 + }, + { + "epoch": 0.8866529033292578, + "grad_norm": 1.8103859424591064, + "learning_rate": 1.2300191995614788e-05, + "loss": 3.2893, + "step": 11059 + }, + { + "epoch": 0.8867330781102805, + "grad_norm": 1.8214197158813477, + "learning_rate": 1.2298928345612604e-05, + "loss": 3.1631, + "step": 11060 + }, + { + "epoch": 0.8868132528913031, + "grad_norm": 1.755295991897583, + "learning_rate": 1.22976646568513e-05, + "loss": 3.1377, + "step": 11061 + }, + { + "epoch": 0.8868934276723257, + "grad_norm": 1.7188578844070435, + "learning_rate": 1.229640092935218e-05, + "loss": 3.1795, + "step": 11062 + }, + { + "epoch": 0.8869736024533483, + "grad_norm": 0.7239171266555786, + "learning_rate": 1.2295137163136548e-05, + "loss": 2.4656, + "step": 11063 + }, + { + "epoch": 0.887053777234371, + "grad_norm": 1.7323907613754272, + "learning_rate": 1.2293873358225714e-05, + "loss": 3.1735, + "step": 11064 + }, + { + "epoch": 0.8871339520153936, + "grad_norm": 1.6958774328231812, + "learning_rate": 1.2292609514640982e-05, + "loss": 3.1824, + "step": 11065 + }, + { + "epoch": 0.8872141267964162, + "grad_norm": 1.8908393383026123, + "learning_rate": 1.2291345632403667e-05, + "loss": 3.2264, + "step": 11066 + }, + { + "epoch": 0.8872943015774388, + "grad_norm": 1.7234035730361938, + "learning_rate": 1.2290081711535064e-05, + "loss": 3.2546, + "step": 11067 + }, + { + "epoch": 0.8873744763584615, + "grad_norm": 1.797681450843811, + "learning_rate": 1.2288817752056495e-05, + "loss": 3.297, + "step": 11068 + }, + { + "epoch": 0.8874546511394841, + "grad_norm": 1.9249751567840576, + "learning_rate": 1.2287553753989265e-05, + "loss": 3.2262, + "step": 11069 + }, + { + "epoch": 0.8875348259205067, + "grad_norm": 1.6372793912887573, + "learning_rate": 1.2286289717354684e-05, + "loss": 3.0685, + "step": 11070 + }, + { + "epoch": 0.8876150007015293, + "grad_norm": 1.79289972782135, + "learning_rate": 1.2285025642174067e-05, + "loss": 3.246, + "step": 11071 + }, + { + "epoch": 0.887695175482552, + "grad_norm": 1.8239065408706665, + "learning_rate": 1.228376152846872e-05, + "loss": 3.2306, + "step": 11072 + }, + { + "epoch": 0.8877753502635746, + "grad_norm": 0.7073574066162109, + "learning_rate": 1.2282497376259958e-05, + "loss": 2.4896, + "step": 11073 + }, + { + "epoch": 0.8878555250445972, + "grad_norm": 1.8371375799179077, + "learning_rate": 1.2281233185569097e-05, + "loss": 3.2554, + "step": 11074 + }, + { + "epoch": 0.8879356998256198, + "grad_norm": 1.6727315187454224, + "learning_rate": 1.2279968956417447e-05, + "loss": 3.157, + "step": 11075 + }, + { + "epoch": 0.8880158746066424, + "grad_norm": 1.7764853239059448, + "learning_rate": 1.2278704688826326e-05, + "loss": 3.2451, + "step": 11076 + }, + { + "epoch": 0.8880960493876651, + "grad_norm": 1.8528138399124146, + "learning_rate": 1.2277440382817043e-05, + "loss": 3.2896, + "step": 11077 + }, + { + "epoch": 0.8881762241686877, + "grad_norm": 1.7271554470062256, + "learning_rate": 1.227617603841092e-05, + "loss": 3.1521, + "step": 11078 + }, + { + "epoch": 0.8882563989497104, + "grad_norm": 1.7063223123550415, + "learning_rate": 1.2274911655629271e-05, + "loss": 3.1913, + "step": 11079 + }, + { + "epoch": 0.888336573730733, + "grad_norm": 1.870985984802246, + "learning_rate": 1.2273647234493415e-05, + "loss": 3.1832, + "step": 11080 + }, + { + "epoch": 0.8884167485117557, + "grad_norm": 1.6643624305725098, + "learning_rate": 1.2272382775024668e-05, + "loss": 3.1871, + "step": 11081 + }, + { + "epoch": 0.8884969232927783, + "grad_norm": 1.954408884048462, + "learning_rate": 1.2271118277244346e-05, + "loss": 3.1936, + "step": 11082 + }, + { + "epoch": 0.8885770980738009, + "grad_norm": 1.7842603921890259, + "learning_rate": 1.2269853741173769e-05, + "loss": 3.1978, + "step": 11083 + }, + { + "epoch": 0.8886572728548235, + "grad_norm": 1.8108292818069458, + "learning_rate": 1.2268589166834262e-05, + "loss": 3.2805, + "step": 11084 + }, + { + "epoch": 0.8887374476358462, + "grad_norm": 1.7576431035995483, + "learning_rate": 1.2267324554247138e-05, + "loss": 3.197, + "step": 11085 + }, + { + "epoch": 0.8888176224168688, + "grad_norm": 1.8811779022216797, + "learning_rate": 1.2266059903433722e-05, + "loss": 3.1301, + "step": 11086 + }, + { + "epoch": 0.8888977971978914, + "grad_norm": 1.819898247718811, + "learning_rate": 1.2264795214415335e-05, + "loss": 3.2589, + "step": 11087 + }, + { + "epoch": 0.888977971978914, + "grad_norm": 1.708117127418518, + "learning_rate": 1.2263530487213298e-05, + "loss": 3.2197, + "step": 11088 + }, + { + "epoch": 0.8890581467599367, + "grad_norm": 1.6543152332305908, + "learning_rate": 1.2262265721848939e-05, + "loss": 3.1493, + "step": 11089 + }, + { + "epoch": 0.8891383215409593, + "grad_norm": 1.716843843460083, + "learning_rate": 1.226100091834357e-05, + "loss": 3.1675, + "step": 11090 + }, + { + "epoch": 0.8892184963219819, + "grad_norm": 1.8502235412597656, + "learning_rate": 1.225973607671853e-05, + "loss": 3.3597, + "step": 11091 + }, + { + "epoch": 0.8892986711030045, + "grad_norm": 1.767016887664795, + "learning_rate": 1.225847119699513e-05, + "loss": 3.1256, + "step": 11092 + }, + { + "epoch": 0.8893788458840272, + "grad_norm": 1.8832083940505981, + "learning_rate": 1.2257206279194702e-05, + "loss": 3.2119, + "step": 11093 + }, + { + "epoch": 0.8894590206650498, + "grad_norm": 1.7981019020080566, + "learning_rate": 1.2255941323338576e-05, + "loss": 3.2252, + "step": 11094 + }, + { + "epoch": 0.8895391954460724, + "grad_norm": 1.7473262548446655, + "learning_rate": 1.225467632944807e-05, + "loss": 3.1922, + "step": 11095 + }, + { + "epoch": 0.889619370227095, + "grad_norm": 0.7013995051383972, + "learning_rate": 1.225341129754452e-05, + "loss": 2.465, + "step": 11096 + }, + { + "epoch": 0.8896995450081177, + "grad_norm": 0.72369384765625, + "learning_rate": 1.2252146227649246e-05, + "loss": 2.4817, + "step": 11097 + }, + { + "epoch": 0.8897797197891403, + "grad_norm": 0.7466555833816528, + "learning_rate": 1.2250881119783582e-05, + "loss": 2.4885, + "step": 11098 + }, + { + "epoch": 0.8898598945701629, + "grad_norm": 1.863944172859192, + "learning_rate": 1.2249615973968856e-05, + "loss": 3.2059, + "step": 11099 + }, + { + "epoch": 0.8899400693511856, + "grad_norm": 1.7543303966522217, + "learning_rate": 1.2248350790226396e-05, + "loss": 3.1494, + "step": 11100 + }, + { + "epoch": 0.8900202441322083, + "grad_norm": 1.6696398258209229, + "learning_rate": 1.2247085568577534e-05, + "loss": 3.2245, + "step": 11101 + }, + { + "epoch": 0.8901004189132309, + "grad_norm": 1.7707561254501343, + "learning_rate": 1.2245820309043603e-05, + "loss": 3.1308, + "step": 11102 + }, + { + "epoch": 0.8901805936942535, + "grad_norm": 1.8392072916030884, + "learning_rate": 1.224455501164593e-05, + "loss": 3.1952, + "step": 11103 + }, + { + "epoch": 0.8902607684752761, + "grad_norm": 1.8027305603027344, + "learning_rate": 1.2243289676405855e-05, + "loss": 3.2471, + "step": 11104 + }, + { + "epoch": 0.8903409432562988, + "grad_norm": 1.826070785522461, + "learning_rate": 1.2242024303344704e-05, + "loss": 3.2319, + "step": 11105 + }, + { + "epoch": 0.8904211180373214, + "grad_norm": 1.6371238231658936, + "learning_rate": 1.2240758892483815e-05, + "loss": 3.2327, + "step": 11106 + }, + { + "epoch": 0.890501292818344, + "grad_norm": 1.7989182472229004, + "learning_rate": 1.2239493443844518e-05, + "loss": 3.175, + "step": 11107 + }, + { + "epoch": 0.8905814675993666, + "grad_norm": 1.7152355909347534, + "learning_rate": 1.2238227957448152e-05, + "loss": 3.21, + "step": 11108 + }, + { + "epoch": 0.8906616423803893, + "grad_norm": 1.8898392915725708, + "learning_rate": 1.2236962433316053e-05, + "loss": 3.2395, + "step": 11109 + }, + { + "epoch": 0.8907418171614119, + "grad_norm": 1.8820816278457642, + "learning_rate": 1.2235696871469554e-05, + "loss": 3.2322, + "step": 11110 + }, + { + "epoch": 0.8908219919424345, + "grad_norm": 1.7074185609817505, + "learning_rate": 1.2234431271929996e-05, + "loss": 3.233, + "step": 11111 + }, + { + "epoch": 0.8909021667234571, + "grad_norm": 1.9313719272613525, + "learning_rate": 1.223316563471871e-05, + "loss": 3.2109, + "step": 11112 + }, + { + "epoch": 0.8909823415044797, + "grad_norm": 1.8655213117599487, + "learning_rate": 1.223189995985704e-05, + "loss": 3.2808, + "step": 11113 + }, + { + "epoch": 0.8910625162855024, + "grad_norm": 1.7674274444580078, + "learning_rate": 1.2230634247366327e-05, + "loss": 3.2413, + "step": 11114 + }, + { + "epoch": 0.891142691066525, + "grad_norm": 1.864802360534668, + "learning_rate": 1.2229368497267901e-05, + "loss": 3.1865, + "step": 11115 + }, + { + "epoch": 0.8912228658475476, + "grad_norm": 0.7406673431396484, + "learning_rate": 1.2228102709583115e-05, + "loss": 2.4904, + "step": 11116 + }, + { + "epoch": 0.8913030406285702, + "grad_norm": 1.8872944116592407, + "learning_rate": 1.2226836884333298e-05, + "loss": 3.2251, + "step": 11117 + }, + { + "epoch": 0.8913832154095929, + "grad_norm": 1.7119932174682617, + "learning_rate": 1.2225571021539795e-05, + "loss": 3.1829, + "step": 11118 + }, + { + "epoch": 0.8914633901906155, + "grad_norm": 1.8241294622421265, + "learning_rate": 1.2224305121223952e-05, + "loss": 3.1171, + "step": 11119 + }, + { + "epoch": 0.8915435649716382, + "grad_norm": 1.7452247142791748, + "learning_rate": 1.2223039183407104e-05, + "loss": 3.1846, + "step": 11120 + }, + { + "epoch": 0.8916237397526608, + "grad_norm": 1.691328525543213, + "learning_rate": 1.2221773208110603e-05, + "loss": 3.113, + "step": 11121 + }, + { + "epoch": 0.8917039145336835, + "grad_norm": 1.763016939163208, + "learning_rate": 1.2220507195355786e-05, + "loss": 3.1701, + "step": 11122 + }, + { + "epoch": 0.8917840893147061, + "grad_norm": 1.7451496124267578, + "learning_rate": 1.2219241145164002e-05, + "loss": 3.3115, + "step": 11123 + }, + { + "epoch": 0.8918642640957287, + "grad_norm": 1.745928168296814, + "learning_rate": 1.2217975057556597e-05, + "loss": 3.2233, + "step": 11124 + }, + { + "epoch": 0.8919444388767513, + "grad_norm": 1.728412389755249, + "learning_rate": 1.221670893255491e-05, + "loss": 3.1077, + "step": 11125 + }, + { + "epoch": 0.892024613657774, + "grad_norm": 1.7627371549606323, + "learning_rate": 1.2215442770180294e-05, + "loss": 3.267, + "step": 11126 + }, + { + "epoch": 0.8921047884387966, + "grad_norm": 1.7051903009414673, + "learning_rate": 1.2214176570454091e-05, + "loss": 3.0733, + "step": 11127 + }, + { + "epoch": 0.8921849632198192, + "grad_norm": 1.7194626331329346, + "learning_rate": 1.2212910333397652e-05, + "loss": 3.2023, + "step": 11128 + }, + { + "epoch": 0.8922651380008418, + "grad_norm": 1.9288289546966553, + "learning_rate": 1.2211644059032329e-05, + "loss": 3.2507, + "step": 11129 + }, + { + "epoch": 0.8923453127818645, + "grad_norm": 1.734114408493042, + "learning_rate": 1.2210377747379463e-05, + "loss": 3.2598, + "step": 11130 + }, + { + "epoch": 0.8924254875628871, + "grad_norm": 1.7864919900894165, + "learning_rate": 1.2209111398460407e-05, + "loss": 3.1687, + "step": 11131 + }, + { + "epoch": 0.8925056623439097, + "grad_norm": 1.814734935760498, + "learning_rate": 1.2207845012296511e-05, + "loss": 3.1757, + "step": 11132 + }, + { + "epoch": 0.8925858371249323, + "grad_norm": 1.8587383031845093, + "learning_rate": 1.2206578588909129e-05, + "loss": 3.2593, + "step": 11133 + }, + { + "epoch": 0.892666011905955, + "grad_norm": 1.843232274055481, + "learning_rate": 1.2205312128319606e-05, + "loss": 3.1157, + "step": 11134 + }, + { + "epoch": 0.8927461866869776, + "grad_norm": 1.7180461883544922, + "learning_rate": 1.22040456305493e-05, + "loss": 3.231, + "step": 11135 + }, + { + "epoch": 0.8928263614680002, + "grad_norm": 1.7076547145843506, + "learning_rate": 1.2202779095619561e-05, + "loss": 3.0763, + "step": 11136 + }, + { + "epoch": 0.8929065362490228, + "grad_norm": 1.821183443069458, + "learning_rate": 1.2201512523551743e-05, + "loss": 3.2452, + "step": 11137 + }, + { + "epoch": 0.8929867110300455, + "grad_norm": 2.162154197692871, + "learning_rate": 1.22002459143672e-05, + "loss": 3.2618, + "step": 11138 + }, + { + "epoch": 0.8930668858110681, + "grad_norm": 1.7289451360702515, + "learning_rate": 1.2198979268087285e-05, + "loss": 3.2186, + "step": 11139 + }, + { + "epoch": 0.8931470605920907, + "grad_norm": 1.7282445430755615, + "learning_rate": 1.2197712584733357e-05, + "loss": 3.2008, + "step": 11140 + }, + { + "epoch": 0.8932272353731134, + "grad_norm": 1.8889449834823608, + "learning_rate": 1.2196445864326769e-05, + "loss": 3.2141, + "step": 11141 + }, + { + "epoch": 0.8933074101541361, + "grad_norm": 1.867682695388794, + "learning_rate": 1.2195179106888875e-05, + "loss": 3.1616, + "step": 11142 + }, + { + "epoch": 0.8933875849351587, + "grad_norm": 1.7812293767929077, + "learning_rate": 1.2193912312441037e-05, + "loss": 3.2161, + "step": 11143 + }, + { + "epoch": 0.8934677597161813, + "grad_norm": 1.7959953546524048, + "learning_rate": 1.219264548100461e-05, + "loss": 3.1967, + "step": 11144 + }, + { + "epoch": 0.8935479344972039, + "grad_norm": 1.66875159740448, + "learning_rate": 1.2191378612600953e-05, + "loss": 3.1484, + "step": 11145 + }, + { + "epoch": 0.8936281092782266, + "grad_norm": 1.8114049434661865, + "learning_rate": 1.2190111707251427e-05, + "loss": 3.1603, + "step": 11146 + }, + { + "epoch": 0.8937082840592492, + "grad_norm": 1.6880065202713013, + "learning_rate": 1.2188844764977386e-05, + "loss": 3.1878, + "step": 11147 + }, + { + "epoch": 0.8937884588402718, + "grad_norm": 1.8785653114318848, + "learning_rate": 1.2187577785800194e-05, + "loss": 3.199, + "step": 11148 + }, + { + "epoch": 0.8938686336212944, + "grad_norm": 1.7710437774658203, + "learning_rate": 1.2186310769741214e-05, + "loss": 3.2019, + "step": 11149 + }, + { + "epoch": 0.893948808402317, + "grad_norm": 1.9416548013687134, + "learning_rate": 1.2185043716821803e-05, + "loss": 3.1846, + "step": 11150 + }, + { + "epoch": 0.8940289831833397, + "grad_norm": 1.8242135047912598, + "learning_rate": 1.2183776627063327e-05, + "loss": 3.2418, + "step": 11151 + }, + { + "epoch": 0.8941091579643623, + "grad_norm": 1.8322112560272217, + "learning_rate": 1.2182509500487142e-05, + "loss": 3.174, + "step": 11152 + }, + { + "epoch": 0.8941893327453849, + "grad_norm": 1.8358628749847412, + "learning_rate": 1.218124233711462e-05, + "loss": 3.1335, + "step": 11153 + }, + { + "epoch": 0.8942695075264075, + "grad_norm": 1.8237413167953491, + "learning_rate": 1.2179975136967122e-05, + "loss": 3.2229, + "step": 11154 + }, + { + "epoch": 0.8943496823074302, + "grad_norm": 0.7337849140167236, + "learning_rate": 1.2178707900066009e-05, + "loss": 2.4976, + "step": 11155 + }, + { + "epoch": 0.8944298570884528, + "grad_norm": 1.9730193614959717, + "learning_rate": 1.2177440626432651e-05, + "loss": 3.1782, + "step": 11156 + }, + { + "epoch": 0.8945100318694754, + "grad_norm": 1.7887629270553589, + "learning_rate": 1.2176173316088408e-05, + "loss": 3.2023, + "step": 11157 + }, + { + "epoch": 0.894590206650498, + "grad_norm": 1.9204471111297607, + "learning_rate": 1.2174905969054651e-05, + "loss": 3.1633, + "step": 11158 + }, + { + "epoch": 0.8946703814315207, + "grad_norm": 1.8265576362609863, + "learning_rate": 1.2173638585352749e-05, + "loss": 3.2539, + "step": 11159 + }, + { + "epoch": 0.8947505562125433, + "grad_norm": 1.8259530067443848, + "learning_rate": 1.2172371165004062e-05, + "loss": 3.2667, + "step": 11160 + }, + { + "epoch": 0.894830730993566, + "grad_norm": 1.8332356214523315, + "learning_rate": 1.2171103708029968e-05, + "loss": 3.157, + "step": 11161 + }, + { + "epoch": 0.8949109057745886, + "grad_norm": 1.6789861917495728, + "learning_rate": 1.2169836214451827e-05, + "loss": 3.1073, + "step": 11162 + }, + { + "epoch": 0.8949910805556113, + "grad_norm": 1.8717652559280396, + "learning_rate": 1.2168568684291011e-05, + "loss": 3.2405, + "step": 11163 + }, + { + "epoch": 0.8950712553366339, + "grad_norm": 1.785069227218628, + "learning_rate": 1.2167301117568895e-05, + "loss": 3.0942, + "step": 11164 + }, + { + "epoch": 0.8951514301176565, + "grad_norm": 1.736145257949829, + "learning_rate": 1.2166033514306844e-05, + "loss": 3.231, + "step": 11165 + }, + { + "epoch": 0.8952316048986791, + "grad_norm": 1.8766475915908813, + "learning_rate": 1.2164765874526232e-05, + "loss": 3.2026, + "step": 11166 + }, + { + "epoch": 0.8953117796797018, + "grad_norm": 0.7104892134666443, + "learning_rate": 1.216349819824843e-05, + "loss": 2.4689, + "step": 11167 + }, + { + "epoch": 0.8953919544607244, + "grad_norm": 1.824425220489502, + "learning_rate": 1.2162230485494808e-05, + "loss": 3.1833, + "step": 11168 + }, + { + "epoch": 0.895472129241747, + "grad_norm": 1.7520960569381714, + "learning_rate": 1.2160962736286746e-05, + "loss": 3.1515, + "step": 11169 + }, + { + "epoch": 0.8955523040227696, + "grad_norm": 1.8841745853424072, + "learning_rate": 1.215969495064561e-05, + "loss": 3.2433, + "step": 11170 + }, + { + "epoch": 0.8956324788037923, + "grad_norm": 1.7890515327453613, + "learning_rate": 1.2158427128592782e-05, + "loss": 3.1617, + "step": 11171 + }, + { + "epoch": 0.8957126535848149, + "grad_norm": 1.8120454549789429, + "learning_rate": 1.2157159270149633e-05, + "loss": 3.1234, + "step": 11172 + }, + { + "epoch": 0.8957928283658375, + "grad_norm": 1.8386421203613281, + "learning_rate": 1.2155891375337536e-05, + "loss": 3.2139, + "step": 11173 + }, + { + "epoch": 0.8958730031468601, + "grad_norm": 1.7148582935333252, + "learning_rate": 1.2154623444177874e-05, + "loss": 3.119, + "step": 11174 + }, + { + "epoch": 0.8959531779278828, + "grad_norm": 1.7342866659164429, + "learning_rate": 1.2153355476692018e-05, + "loss": 3.0784, + "step": 11175 + }, + { + "epoch": 0.8960333527089054, + "grad_norm": 1.7551275491714478, + "learning_rate": 1.2152087472901348e-05, + "loss": 3.1844, + "step": 11176 + }, + { + "epoch": 0.896113527489928, + "grad_norm": 1.7443604469299316, + "learning_rate": 1.215081943282724e-05, + "loss": 3.204, + "step": 11177 + }, + { + "epoch": 0.8961937022709506, + "grad_norm": 3.8903682231903076, + "learning_rate": 1.2149551356491076e-05, + "loss": 3.2284, + "step": 11178 + }, + { + "epoch": 0.8962738770519733, + "grad_norm": 1.8150675296783447, + "learning_rate": 1.2148283243914235e-05, + "loss": 3.288, + "step": 11179 + }, + { + "epoch": 0.8963540518329959, + "grad_norm": 1.7478861808776855, + "learning_rate": 1.2147015095118093e-05, + "loss": 3.2593, + "step": 11180 + }, + { + "epoch": 0.8964342266140186, + "grad_norm": 1.9235883951187134, + "learning_rate": 1.2145746910124036e-05, + "loss": 3.1718, + "step": 11181 + }, + { + "epoch": 0.8965144013950412, + "grad_norm": 1.7812273502349854, + "learning_rate": 1.214447868895344e-05, + "loss": 3.256, + "step": 11182 + }, + { + "epoch": 0.8965945761760639, + "grad_norm": 1.7520928382873535, + "learning_rate": 1.214321043162769e-05, + "loss": 3.1484, + "step": 11183 + }, + { + "epoch": 0.8966747509570865, + "grad_norm": 1.7697601318359375, + "learning_rate": 1.2141942138168168e-05, + "loss": 3.194, + "step": 11184 + }, + { + "epoch": 0.8967549257381091, + "grad_norm": 1.8188457489013672, + "learning_rate": 1.2140673808596256e-05, + "loss": 3.1257, + "step": 11185 + }, + { + "epoch": 0.8968351005191317, + "grad_norm": 1.9328279495239258, + "learning_rate": 1.2139405442933338e-05, + "loss": 3.1595, + "step": 11186 + }, + { + "epoch": 0.8969152753001544, + "grad_norm": 1.6871994733810425, + "learning_rate": 1.2138137041200797e-05, + "loss": 3.2734, + "step": 11187 + }, + { + "epoch": 0.896995450081177, + "grad_norm": 2.124218463897705, + "learning_rate": 1.213686860342002e-05, + "loss": 3.1735, + "step": 11188 + }, + { + "epoch": 0.8970756248621996, + "grad_norm": 1.8240042924880981, + "learning_rate": 1.2135600129612393e-05, + "loss": 3.1393, + "step": 11189 + }, + { + "epoch": 0.8971557996432222, + "grad_norm": 1.840036153793335, + "learning_rate": 1.2134331619799298e-05, + "loss": 3.2441, + "step": 11190 + }, + { + "epoch": 0.8972359744242449, + "grad_norm": 1.8197468519210815, + "learning_rate": 1.2133063074002127e-05, + "loss": 3.2197, + "step": 11191 + }, + { + "epoch": 0.8973161492052675, + "grad_norm": 1.7569481134414673, + "learning_rate": 1.2131794492242262e-05, + "loss": 3.2139, + "step": 11192 + }, + { + "epoch": 0.8973963239862901, + "grad_norm": 1.8482823371887207, + "learning_rate": 1.2130525874541093e-05, + "loss": 3.2788, + "step": 11193 + }, + { + "epoch": 0.8974764987673127, + "grad_norm": 1.781375765800476, + "learning_rate": 1.212925722092001e-05, + "loss": 3.2507, + "step": 11194 + }, + { + "epoch": 0.8975566735483353, + "grad_norm": 1.7382428646087646, + "learning_rate": 1.2127988531400399e-05, + "loss": 3.3382, + "step": 11195 + }, + { + "epoch": 0.897636848329358, + "grad_norm": 1.7723000049591064, + "learning_rate": 1.2126719806003651e-05, + "loss": 3.2053, + "step": 11196 + }, + { + "epoch": 0.8977170231103806, + "grad_norm": 1.7124965190887451, + "learning_rate": 1.212545104475116e-05, + "loss": 3.2948, + "step": 11197 + }, + { + "epoch": 0.8977971978914032, + "grad_norm": 1.7414171695709229, + "learning_rate": 1.212418224766431e-05, + "loss": 3.1618, + "step": 11198 + }, + { + "epoch": 0.8978773726724258, + "grad_norm": 1.8205821514129639, + "learning_rate": 1.2122913414764498e-05, + "loss": 3.1729, + "step": 11199 + }, + { + "epoch": 0.8979575474534485, + "grad_norm": 0.7516794204711914, + "learning_rate": 1.2121644546073113e-05, + "loss": 2.4659, + "step": 11200 + }, + { + "epoch": 0.8980377222344711, + "grad_norm": 1.922263264656067, + "learning_rate": 1.2120375641611549e-05, + "loss": 3.2385, + "step": 11201 + }, + { + "epoch": 0.8981178970154938, + "grad_norm": 2.285935401916504, + "learning_rate": 1.2119106701401199e-05, + "loss": 3.183, + "step": 11202 + }, + { + "epoch": 0.8981980717965165, + "grad_norm": 1.8369312286376953, + "learning_rate": 1.2117837725463456e-05, + "loss": 3.1881, + "step": 11203 + }, + { + "epoch": 0.8982782465775391, + "grad_norm": 1.774364709854126, + "learning_rate": 1.2116568713819716e-05, + "loss": 3.2523, + "step": 11204 + }, + { + "epoch": 0.8983584213585617, + "grad_norm": 1.7464673519134521, + "learning_rate": 1.2115299666491374e-05, + "loss": 3.304, + "step": 11205 + }, + { + "epoch": 0.8984385961395843, + "grad_norm": 1.758976697921753, + "learning_rate": 1.2114030583499825e-05, + "loss": 3.1877, + "step": 11206 + }, + { + "epoch": 0.898518770920607, + "grad_norm": 1.6930124759674072, + "learning_rate": 1.2112761464866464e-05, + "loss": 3.0091, + "step": 11207 + }, + { + "epoch": 0.8985989457016296, + "grad_norm": 1.8259475231170654, + "learning_rate": 1.2111492310612691e-05, + "loss": 3.2494, + "step": 11208 + }, + { + "epoch": 0.8986791204826522, + "grad_norm": 2.1563785076141357, + "learning_rate": 1.2110223120759903e-05, + "loss": 3.1497, + "step": 11209 + }, + { + "epoch": 0.8987592952636748, + "grad_norm": 1.6975808143615723, + "learning_rate": 1.2108953895329494e-05, + "loss": 3.2184, + "step": 11210 + }, + { + "epoch": 0.8988394700446974, + "grad_norm": 1.7235023975372314, + "learning_rate": 1.2107684634342866e-05, + "loss": 3.1552, + "step": 11211 + }, + { + "epoch": 0.8989196448257201, + "grad_norm": 1.7344154119491577, + "learning_rate": 1.210641533782142e-05, + "loss": 3.2453, + "step": 11212 + }, + { + "epoch": 0.8989998196067427, + "grad_norm": 1.6448136568069458, + "learning_rate": 1.2105146005786553e-05, + "loss": 3.1659, + "step": 11213 + }, + { + "epoch": 0.8990799943877653, + "grad_norm": 1.703035593032837, + "learning_rate": 1.2103876638259667e-05, + "loss": 3.283, + "step": 11214 + }, + { + "epoch": 0.8991601691687879, + "grad_norm": 1.7125592231750488, + "learning_rate": 1.2102607235262158e-05, + "loss": 3.2079, + "step": 11215 + }, + { + "epoch": 0.8992403439498106, + "grad_norm": 1.745205044746399, + "learning_rate": 1.2101337796815435e-05, + "loss": 3.3046, + "step": 11216 + }, + { + "epoch": 0.8993205187308332, + "grad_norm": 1.8166041374206543, + "learning_rate": 1.2100068322940898e-05, + "loss": 3.1821, + "step": 11217 + }, + { + "epoch": 0.8994006935118558, + "grad_norm": 1.784804105758667, + "learning_rate": 1.2098798813659949e-05, + "loss": 3.3088, + "step": 11218 + }, + { + "epoch": 0.8994808682928784, + "grad_norm": 1.78764009475708, + "learning_rate": 1.2097529268993993e-05, + "loss": 3.1769, + "step": 11219 + }, + { + "epoch": 0.8995610430739011, + "grad_norm": 1.7859388589859009, + "learning_rate": 1.2096259688964428e-05, + "loss": 3.2712, + "step": 11220 + }, + { + "epoch": 0.8996412178549237, + "grad_norm": 1.7851532697677612, + "learning_rate": 1.2094990073592667e-05, + "loss": 3.3511, + "step": 11221 + }, + { + "epoch": 0.8997213926359464, + "grad_norm": 1.7767717838287354, + "learning_rate": 1.2093720422900112e-05, + "loss": 3.3245, + "step": 11222 + }, + { + "epoch": 0.899801567416969, + "grad_norm": 1.7036212682724, + "learning_rate": 1.2092450736908167e-05, + "loss": 3.1837, + "step": 11223 + }, + { + "epoch": 0.8998817421979917, + "grad_norm": 1.711964726448059, + "learning_rate": 1.2091181015638243e-05, + "loss": 3.1589, + "step": 11224 + }, + { + "epoch": 0.8999619169790143, + "grad_norm": 1.7857930660247803, + "learning_rate": 1.208991125911174e-05, + "loss": 3.2414, + "step": 11225 + }, + { + "epoch": 0.9000420917600369, + "grad_norm": 0.7040237188339233, + "learning_rate": 1.2088641467350071e-05, + "loss": 2.4886, + "step": 11226 + }, + { + "epoch": 0.9001222665410595, + "grad_norm": 1.7890446186065674, + "learning_rate": 1.2087371640374646e-05, + "loss": 3.1713, + "step": 11227 + }, + { + "epoch": 0.9002024413220822, + "grad_norm": 1.8744416236877441, + "learning_rate": 1.2086101778206867e-05, + "loss": 3.2687, + "step": 11228 + }, + { + "epoch": 0.9002826161031048, + "grad_norm": 0.7141697406768799, + "learning_rate": 1.2084831880868149e-05, + "loss": 2.4972, + "step": 11229 + }, + { + "epoch": 0.9003627908841274, + "grad_norm": 1.7004565000534058, + "learning_rate": 1.2083561948379901e-05, + "loss": 3.1418, + "step": 11230 + }, + { + "epoch": 0.90044296566515, + "grad_norm": 1.959155559539795, + "learning_rate": 1.208229198076353e-05, + "loss": 3.1274, + "step": 11231 + }, + { + "epoch": 0.9005231404461727, + "grad_norm": 1.7988884449005127, + "learning_rate": 1.2081021978040453e-05, + "loss": 3.1775, + "step": 11232 + }, + { + "epoch": 0.9006033152271953, + "grad_norm": 1.8063913583755493, + "learning_rate": 1.2079751940232076e-05, + "loss": 3.2082, + "step": 11233 + }, + { + "epoch": 0.9006834900082179, + "grad_norm": 1.7422528266906738, + "learning_rate": 1.2078481867359816e-05, + "loss": 3.2797, + "step": 11234 + }, + { + "epoch": 0.9007636647892405, + "grad_norm": 1.751466989517212, + "learning_rate": 1.2077211759445084e-05, + "loss": 3.1553, + "step": 11235 + }, + { + "epoch": 0.9008438395702631, + "grad_norm": 1.7951360940933228, + "learning_rate": 1.2075941616509295e-05, + "loss": 3.2346, + "step": 11236 + }, + { + "epoch": 0.9009240143512858, + "grad_norm": 1.8054499626159668, + "learning_rate": 1.2074671438573862e-05, + "loss": 3.2067, + "step": 11237 + }, + { + "epoch": 0.9010041891323084, + "grad_norm": 1.8325014114379883, + "learning_rate": 1.2073401225660197e-05, + "loss": 3.2532, + "step": 11238 + }, + { + "epoch": 0.901084363913331, + "grad_norm": 1.8403878211975098, + "learning_rate": 1.2072130977789723e-05, + "loss": 3.188, + "step": 11239 + }, + { + "epoch": 0.9011645386943536, + "grad_norm": 1.7885397672653198, + "learning_rate": 1.2070860694983846e-05, + "loss": 3.2384, + "step": 11240 + }, + { + "epoch": 0.9012447134753763, + "grad_norm": 1.7532923221588135, + "learning_rate": 1.2069590377263989e-05, + "loss": 3.2595, + "step": 11241 + }, + { + "epoch": 0.901324888256399, + "grad_norm": 1.7801282405853271, + "learning_rate": 1.206832002465157e-05, + "loss": 3.1926, + "step": 11242 + }, + { + "epoch": 0.9014050630374216, + "grad_norm": 1.8436120748519897, + "learning_rate": 1.2067049637168e-05, + "loss": 3.3521, + "step": 11243 + }, + { + "epoch": 0.9014852378184443, + "grad_norm": 1.8815408945083618, + "learning_rate": 1.2065779214834706e-05, + "loss": 3.2673, + "step": 11244 + }, + { + "epoch": 0.9015654125994669, + "grad_norm": 1.8315067291259766, + "learning_rate": 1.2064508757673099e-05, + "loss": 3.1936, + "step": 11245 + }, + { + "epoch": 0.9016455873804895, + "grad_norm": 1.9005070924758911, + "learning_rate": 1.2063238265704602e-05, + "loss": 3.2027, + "step": 11246 + }, + { + "epoch": 0.9017257621615121, + "grad_norm": 1.8908615112304688, + "learning_rate": 1.2061967738950637e-05, + "loss": 3.1376, + "step": 11247 + }, + { + "epoch": 0.9018059369425347, + "grad_norm": 1.7874780893325806, + "learning_rate": 1.2060697177432621e-05, + "loss": 3.085, + "step": 11248 + }, + { + "epoch": 0.9018861117235574, + "grad_norm": 1.7669914960861206, + "learning_rate": 1.205942658117198e-05, + "loss": 3.089, + "step": 11249 + }, + { + "epoch": 0.90196628650458, + "grad_norm": 1.723983645439148, + "learning_rate": 1.205815595019013e-05, + "loss": 3.194, + "step": 11250 + }, + { + "epoch": 0.9020464612856026, + "grad_norm": 1.7552063465118408, + "learning_rate": 1.2056885284508494e-05, + "loss": 3.1309, + "step": 11251 + }, + { + "epoch": 0.9021266360666252, + "grad_norm": 1.7222976684570312, + "learning_rate": 1.2055614584148503e-05, + "loss": 3.2014, + "step": 11252 + }, + { + "epoch": 0.9022068108476479, + "grad_norm": 1.7247272729873657, + "learning_rate": 1.205434384913157e-05, + "loss": 3.2095, + "step": 11253 + }, + { + "epoch": 0.9022869856286705, + "grad_norm": 1.7867798805236816, + "learning_rate": 1.2053073079479125e-05, + "loss": 3.1992, + "step": 11254 + }, + { + "epoch": 0.9023671604096931, + "grad_norm": 1.7901772260665894, + "learning_rate": 1.2051802275212591e-05, + "loss": 3.249, + "step": 11255 + }, + { + "epoch": 0.9024473351907157, + "grad_norm": 1.9613080024719238, + "learning_rate": 1.2050531436353394e-05, + "loss": 3.2946, + "step": 11256 + }, + { + "epoch": 0.9025275099717384, + "grad_norm": 1.7465819120407104, + "learning_rate": 1.2049260562922959e-05, + "loss": 3.2413, + "step": 11257 + }, + { + "epoch": 0.902607684752761, + "grad_norm": 1.9952880144119263, + "learning_rate": 1.2047989654942716e-05, + "loss": 3.094, + "step": 11258 + }, + { + "epoch": 0.9026878595337836, + "grad_norm": 1.8729654550552368, + "learning_rate": 1.2046718712434086e-05, + "loss": 3.2911, + "step": 11259 + }, + { + "epoch": 0.9027680343148062, + "grad_norm": 1.7397516965866089, + "learning_rate": 1.2045447735418501e-05, + "loss": 3.2224, + "step": 11260 + }, + { + "epoch": 0.9028482090958289, + "grad_norm": 1.8464266061782837, + "learning_rate": 1.2044176723917386e-05, + "loss": 3.1603, + "step": 11261 + }, + { + "epoch": 0.9029283838768515, + "grad_norm": 1.8639159202575684, + "learning_rate": 1.2042905677952177e-05, + "loss": 3.2803, + "step": 11262 + }, + { + "epoch": 0.9030085586578742, + "grad_norm": 1.7285804748535156, + "learning_rate": 1.2041634597544293e-05, + "loss": 3.2144, + "step": 11263 + }, + { + "epoch": 0.9030887334388968, + "grad_norm": 1.8630883693695068, + "learning_rate": 1.2040363482715174e-05, + "loss": 3.2268, + "step": 11264 + }, + { + "epoch": 0.9031689082199195, + "grad_norm": 1.7753748893737793, + "learning_rate": 1.2039092333486242e-05, + "loss": 3.202, + "step": 11265 + }, + { + "epoch": 0.9032490830009421, + "grad_norm": 1.7792339324951172, + "learning_rate": 1.2037821149878932e-05, + "loss": 3.2145, + "step": 11266 + }, + { + "epoch": 0.9033292577819647, + "grad_norm": 1.7487947940826416, + "learning_rate": 1.2036549931914678e-05, + "loss": 3.2394, + "step": 11267 + }, + { + "epoch": 0.9034094325629873, + "grad_norm": 1.7503159046173096, + "learning_rate": 1.2035278679614908e-05, + "loss": 3.1434, + "step": 11268 + }, + { + "epoch": 0.90348960734401, + "grad_norm": 1.8614180088043213, + "learning_rate": 1.2034007393001057e-05, + "loss": 3.1432, + "step": 11269 + }, + { + "epoch": 0.9035697821250326, + "grad_norm": 1.7851617336273193, + "learning_rate": 1.203273607209456e-05, + "loss": 3.1904, + "step": 11270 + }, + { + "epoch": 0.9036499569060552, + "grad_norm": 1.7374004125595093, + "learning_rate": 1.2031464716916845e-05, + "loss": 3.2107, + "step": 11271 + }, + { + "epoch": 0.9037301316870778, + "grad_norm": 1.7031419277191162, + "learning_rate": 1.2030193327489354e-05, + "loss": 3.2277, + "step": 11272 + }, + { + "epoch": 0.9038103064681005, + "grad_norm": 0.7089508175849915, + "learning_rate": 1.202892190383352e-05, + "loss": 2.4854, + "step": 11273 + }, + { + "epoch": 0.9038904812491231, + "grad_norm": 1.768657922744751, + "learning_rate": 1.2027650445970778e-05, + "loss": 3.224, + "step": 11274 + }, + { + "epoch": 0.9039706560301457, + "grad_norm": 1.8103611469268799, + "learning_rate": 1.202637895392256e-05, + "loss": 3.1584, + "step": 11275 + }, + { + "epoch": 0.9040508308111683, + "grad_norm": 1.732710838317871, + "learning_rate": 1.202510742771031e-05, + "loss": 3.1893, + "step": 11276 + }, + { + "epoch": 0.904131005592191, + "grad_norm": 1.92520010471344, + "learning_rate": 1.2023835867355462e-05, + "loss": 3.1876, + "step": 11277 + }, + { + "epoch": 0.9042111803732136, + "grad_norm": 1.7640917301177979, + "learning_rate": 1.2022564272879454e-05, + "loss": 3.1837, + "step": 11278 + }, + { + "epoch": 0.9042913551542362, + "grad_norm": 1.7741084098815918, + "learning_rate": 1.2021292644303725e-05, + "loss": 3.2968, + "step": 11279 + }, + { + "epoch": 0.9043715299352588, + "grad_norm": 1.7977863550186157, + "learning_rate": 1.2020020981649717e-05, + "loss": 3.1954, + "step": 11280 + }, + { + "epoch": 0.9044517047162814, + "grad_norm": 1.7539730072021484, + "learning_rate": 1.2018749284938864e-05, + "loss": 3.1578, + "step": 11281 + }, + { + "epoch": 0.9045318794973041, + "grad_norm": 1.794014811515808, + "learning_rate": 1.2017477554192612e-05, + "loss": 3.2159, + "step": 11282 + }, + { + "epoch": 0.9046120542783268, + "grad_norm": 1.7860394716262817, + "learning_rate": 1.2016205789432395e-05, + "loss": 3.1749, + "step": 11283 + }, + { + "epoch": 0.9046922290593494, + "grad_norm": 1.7810008525848389, + "learning_rate": 1.2014933990679665e-05, + "loss": 3.2343, + "step": 11284 + }, + { + "epoch": 0.904772403840372, + "grad_norm": 2.0093557834625244, + "learning_rate": 1.2013662157955854e-05, + "loss": 3.2272, + "step": 11285 + }, + { + "epoch": 0.9048525786213947, + "grad_norm": 1.7231804132461548, + "learning_rate": 1.201239029128241e-05, + "loss": 3.1342, + "step": 11286 + }, + { + "epoch": 0.9049327534024173, + "grad_norm": 1.7084945440292358, + "learning_rate": 1.2011118390680776e-05, + "loss": 3.2576, + "step": 11287 + }, + { + "epoch": 0.9050129281834399, + "grad_norm": 1.7503818273544312, + "learning_rate": 1.2009846456172391e-05, + "loss": 3.3433, + "step": 11288 + }, + { + "epoch": 0.9050931029644625, + "grad_norm": 1.7318873405456543, + "learning_rate": 1.2008574487778708e-05, + "loss": 3.1726, + "step": 11289 + }, + { + "epoch": 0.9051732777454852, + "grad_norm": 1.6513012647628784, + "learning_rate": 1.2007302485521164e-05, + "loss": 3.1141, + "step": 11290 + }, + { + "epoch": 0.9052534525265078, + "grad_norm": 1.738145351409912, + "learning_rate": 1.2006030449421208e-05, + "loss": 3.2466, + "step": 11291 + }, + { + "epoch": 0.9053336273075304, + "grad_norm": 1.9247993230819702, + "learning_rate": 1.2004758379500287e-05, + "loss": 3.1508, + "step": 11292 + }, + { + "epoch": 0.905413802088553, + "grad_norm": 1.7499481439590454, + "learning_rate": 1.2003486275779843e-05, + "loss": 3.2186, + "step": 11293 + }, + { + "epoch": 0.9054939768695757, + "grad_norm": 1.8030879497528076, + "learning_rate": 1.2002214138281332e-05, + "loss": 3.115, + "step": 11294 + }, + { + "epoch": 0.9055741516505983, + "grad_norm": 2.014732837677002, + "learning_rate": 1.2000941967026192e-05, + "loss": 3.2853, + "step": 11295 + }, + { + "epoch": 0.9056543264316209, + "grad_norm": 1.8512563705444336, + "learning_rate": 1.1999669762035874e-05, + "loss": 3.1305, + "step": 11296 + }, + { + "epoch": 0.9057345012126435, + "grad_norm": 1.8252556324005127, + "learning_rate": 1.1998397523331833e-05, + "loss": 3.2139, + "step": 11297 + }, + { + "epoch": 0.9058146759936662, + "grad_norm": 1.7694865465164185, + "learning_rate": 1.199712525093551e-05, + "loss": 3.121, + "step": 11298 + }, + { + "epoch": 0.9058948507746888, + "grad_norm": 1.7581934928894043, + "learning_rate": 1.1995852944868363e-05, + "loss": 3.142, + "step": 11299 + }, + { + "epoch": 0.9059750255557114, + "grad_norm": 1.8206738233566284, + "learning_rate": 1.1994580605151838e-05, + "loss": 3.2443, + "step": 11300 + }, + { + "epoch": 0.906055200336734, + "grad_norm": 2.005526304244995, + "learning_rate": 1.1993308231807384e-05, + "loss": 3.1801, + "step": 11301 + }, + { + "epoch": 0.9061353751177567, + "grad_norm": 1.8539444208145142, + "learning_rate": 1.1992035824856458e-05, + "loss": 3.1693, + "step": 11302 + }, + { + "epoch": 0.9062155498987794, + "grad_norm": 0.7659796476364136, + "learning_rate": 1.199076338432051e-05, + "loss": 2.4774, + "step": 11303 + }, + { + "epoch": 0.906295724679802, + "grad_norm": 1.656874656677246, + "learning_rate": 1.1989490910220994e-05, + "loss": 3.1255, + "step": 11304 + }, + { + "epoch": 0.9063758994608246, + "grad_norm": 1.7132385969161987, + "learning_rate": 1.198821840257936e-05, + "loss": 3.1359, + "step": 11305 + }, + { + "epoch": 0.9064560742418473, + "grad_norm": 1.7708998918533325, + "learning_rate": 1.1986945861417066e-05, + "loss": 3.1542, + "step": 11306 + }, + { + "epoch": 0.9065362490228699, + "grad_norm": 1.7744652032852173, + "learning_rate": 1.1985673286755563e-05, + "loss": 3.2421, + "step": 11307 + }, + { + "epoch": 0.9066164238038925, + "grad_norm": 1.8113675117492676, + "learning_rate": 1.198440067861631e-05, + "loss": 3.1729, + "step": 11308 + }, + { + "epoch": 0.9066965985849151, + "grad_norm": 1.8489186763763428, + "learning_rate": 1.1983128037020763e-05, + "loss": 3.2799, + "step": 11309 + }, + { + "epoch": 0.9067767733659378, + "grad_norm": 1.7709531784057617, + "learning_rate": 1.1981855361990374e-05, + "loss": 3.2013, + "step": 11310 + }, + { + "epoch": 0.9068569481469604, + "grad_norm": 0.6973528861999512, + "learning_rate": 1.19805826535466e-05, + "loss": 2.479, + "step": 11311 + }, + { + "epoch": 0.906937122927983, + "grad_norm": 1.7684273719787598, + "learning_rate": 1.1979309911710903e-05, + "loss": 3.1898, + "step": 11312 + }, + { + "epoch": 0.9070172977090056, + "grad_norm": 1.8832982778549194, + "learning_rate": 1.1978037136504739e-05, + "loss": 3.1595, + "step": 11313 + }, + { + "epoch": 0.9070974724900283, + "grad_norm": 1.782135009765625, + "learning_rate": 1.1976764327949566e-05, + "loss": 3.3137, + "step": 11314 + }, + { + "epoch": 0.9071776472710509, + "grad_norm": 1.8244422674179077, + "learning_rate": 1.197549148606684e-05, + "loss": 3.2701, + "step": 11315 + }, + { + "epoch": 0.9072578220520735, + "grad_norm": 1.8508986234664917, + "learning_rate": 1.1974218610878026e-05, + "loss": 3.2003, + "step": 11316 + }, + { + "epoch": 0.9073379968330961, + "grad_norm": 1.6120527982711792, + "learning_rate": 1.1972945702404584e-05, + "loss": 3.1913, + "step": 11317 + }, + { + "epoch": 0.9074181716141188, + "grad_norm": 1.7173196077346802, + "learning_rate": 1.197167276066797e-05, + "loss": 3.1793, + "step": 11318 + }, + { + "epoch": 0.9074983463951414, + "grad_norm": 1.7370033264160156, + "learning_rate": 1.1970399785689648e-05, + "loss": 3.1926, + "step": 11319 + }, + { + "epoch": 0.907578521176164, + "grad_norm": 1.9259735345840454, + "learning_rate": 1.196912677749108e-05, + "loss": 3.1356, + "step": 11320 + }, + { + "epoch": 0.9076586959571866, + "grad_norm": 1.8271300792694092, + "learning_rate": 1.196785373609373e-05, + "loss": 3.2305, + "step": 11321 + }, + { + "epoch": 0.9077388707382092, + "grad_norm": 0.7013381123542786, + "learning_rate": 1.1966580661519061e-05, + "loss": 2.4685, + "step": 11322 + }, + { + "epoch": 0.9078190455192319, + "grad_norm": 1.712317943572998, + "learning_rate": 1.196530755378853e-05, + "loss": 3.1703, + "step": 11323 + }, + { + "epoch": 0.9078992203002546, + "grad_norm": 1.8930699825286865, + "learning_rate": 1.196403441292361e-05, + "loss": 3.2504, + "step": 11324 + }, + { + "epoch": 0.9079793950812772, + "grad_norm": 1.8455473184585571, + "learning_rate": 1.1962761238945762e-05, + "loss": 3.1102, + "step": 11325 + }, + { + "epoch": 0.9080595698622999, + "grad_norm": 1.7810935974121094, + "learning_rate": 1.196148803187645e-05, + "loss": 3.1107, + "step": 11326 + }, + { + "epoch": 0.9081397446433225, + "grad_norm": 1.845035195350647, + "learning_rate": 1.1960214791737142e-05, + "loss": 3.0947, + "step": 11327 + }, + { + "epoch": 0.9082199194243451, + "grad_norm": 1.810875415802002, + "learning_rate": 1.1958941518549303e-05, + "loss": 3.215, + "step": 11328 + }, + { + "epoch": 0.9083000942053677, + "grad_norm": 1.8425407409667969, + "learning_rate": 1.1957668212334402e-05, + "loss": 3.2595, + "step": 11329 + }, + { + "epoch": 0.9083802689863903, + "grad_norm": 1.805322527885437, + "learning_rate": 1.1956394873113901e-05, + "loss": 3.2153, + "step": 11330 + }, + { + "epoch": 0.908460443767413, + "grad_norm": 1.705790638923645, + "learning_rate": 1.1955121500909275e-05, + "loss": 3.1849, + "step": 11331 + }, + { + "epoch": 0.9085406185484356, + "grad_norm": 1.822769284248352, + "learning_rate": 1.195384809574199e-05, + "loss": 3.2113, + "step": 11332 + }, + { + "epoch": 0.9086207933294582, + "grad_norm": 1.8142558336257935, + "learning_rate": 1.1952574657633514e-05, + "loss": 3.3024, + "step": 11333 + }, + { + "epoch": 0.9087009681104808, + "grad_norm": 1.869588017463684, + "learning_rate": 1.1951301186605317e-05, + "loss": 3.2598, + "step": 11334 + }, + { + "epoch": 0.9087811428915035, + "grad_norm": 1.8273600339889526, + "learning_rate": 1.1950027682678868e-05, + "loss": 3.2584, + "step": 11335 + }, + { + "epoch": 0.9088613176725261, + "grad_norm": 1.7274234294891357, + "learning_rate": 1.194875414587564e-05, + "loss": 3.2609, + "step": 11336 + }, + { + "epoch": 0.9089414924535487, + "grad_norm": 1.8588240146636963, + "learning_rate": 1.1947480576217107e-05, + "loss": 3.2643, + "step": 11337 + }, + { + "epoch": 0.9090216672345713, + "grad_norm": 1.7938711643218994, + "learning_rate": 1.1946206973724736e-05, + "loss": 3.2059, + "step": 11338 + }, + { + "epoch": 0.909101842015594, + "grad_norm": 2.000718116760254, + "learning_rate": 1.1944933338420001e-05, + "loss": 3.3096, + "step": 11339 + }, + { + "epoch": 0.9091820167966166, + "grad_norm": 1.8602958917617798, + "learning_rate": 1.1943659670324376e-05, + "loss": 3.2598, + "step": 11340 + }, + { + "epoch": 0.9092621915776392, + "grad_norm": 0.7526415586471558, + "learning_rate": 1.1942385969459334e-05, + "loss": 2.5018, + "step": 11341 + }, + { + "epoch": 0.9093423663586618, + "grad_norm": 1.6947585344314575, + "learning_rate": 1.1941112235846349e-05, + "loss": 3.1454, + "step": 11342 + }, + { + "epoch": 0.9094225411396845, + "grad_norm": 1.8265323638916016, + "learning_rate": 1.1939838469506894e-05, + "loss": 3.121, + "step": 11343 + }, + { + "epoch": 0.9095027159207072, + "grad_norm": 1.8201448917388916, + "learning_rate": 1.1938564670462449e-05, + "loss": 3.3064, + "step": 11344 + }, + { + "epoch": 0.9095828907017298, + "grad_norm": 1.8244749307632446, + "learning_rate": 1.1937290838734485e-05, + "loss": 3.2977, + "step": 11345 + }, + { + "epoch": 0.9096630654827524, + "grad_norm": 1.882888913154602, + "learning_rate": 1.193601697434448e-05, + "loss": 3.244, + "step": 11346 + }, + { + "epoch": 0.9097432402637751, + "grad_norm": 1.7742111682891846, + "learning_rate": 1.1934743077313914e-05, + "loss": 3.2302, + "step": 11347 + }, + { + "epoch": 0.9098234150447977, + "grad_norm": 1.816994547843933, + "learning_rate": 1.1933469147664259e-05, + "loss": 3.1003, + "step": 11348 + }, + { + "epoch": 0.9099035898258203, + "grad_norm": 1.7171082496643066, + "learning_rate": 1.1932195185416997e-05, + "loss": 3.1019, + "step": 11349 + }, + { + "epoch": 0.9099837646068429, + "grad_norm": 1.8497300148010254, + "learning_rate": 1.1930921190593603e-05, + "loss": 3.2457, + "step": 11350 + }, + { + "epoch": 0.9100639393878656, + "grad_norm": 2.023244857788086, + "learning_rate": 1.1929647163215558e-05, + "loss": 3.1887, + "step": 11351 + }, + { + "epoch": 0.9101441141688882, + "grad_norm": 1.8553439378738403, + "learning_rate": 1.1928373103304345e-05, + "loss": 3.2316, + "step": 11352 + }, + { + "epoch": 0.9102242889499108, + "grad_norm": 1.7587668895721436, + "learning_rate": 1.192709901088144e-05, + "loss": 3.2567, + "step": 11353 + }, + { + "epoch": 0.9103044637309334, + "grad_norm": 1.758891224861145, + "learning_rate": 1.1925824885968325e-05, + "loss": 3.2527, + "step": 11354 + }, + { + "epoch": 0.9103846385119561, + "grad_norm": 1.7532877922058105, + "learning_rate": 1.192455072858648e-05, + "loss": 3.21, + "step": 11355 + }, + { + "epoch": 0.9104648132929787, + "grad_norm": 1.8007375001907349, + "learning_rate": 1.1923276538757387e-05, + "loss": 3.1736, + "step": 11356 + }, + { + "epoch": 0.9105449880740013, + "grad_norm": 1.796142816543579, + "learning_rate": 1.1922002316502532e-05, + "loss": 3.2396, + "step": 11357 + }, + { + "epoch": 0.9106251628550239, + "grad_norm": 1.5419809818267822, + "learning_rate": 1.1920728061843394e-05, + "loss": 3.144, + "step": 11358 + }, + { + "epoch": 0.9107053376360466, + "grad_norm": 1.7695153951644897, + "learning_rate": 1.1919453774801459e-05, + "loss": 3.0664, + "step": 11359 + }, + { + "epoch": 0.9107855124170692, + "grad_norm": 1.721488118171692, + "learning_rate": 1.1918179455398209e-05, + "loss": 3.2178, + "step": 11360 + }, + { + "epoch": 0.9108656871980918, + "grad_norm": 1.8809317350387573, + "learning_rate": 1.1916905103655128e-05, + "loss": 3.1021, + "step": 11361 + }, + { + "epoch": 0.9109458619791144, + "grad_norm": 1.7464954853057861, + "learning_rate": 1.1915630719593705e-05, + "loss": 3.198, + "step": 11362 + }, + { + "epoch": 0.911026036760137, + "grad_norm": 1.7884024381637573, + "learning_rate": 1.1914356303235422e-05, + "loss": 3.2086, + "step": 11363 + }, + { + "epoch": 0.9111062115411598, + "grad_norm": 1.8278558254241943, + "learning_rate": 1.1913081854601767e-05, + "loss": 3.1283, + "step": 11364 + }, + { + "epoch": 0.9111863863221824, + "grad_norm": 1.7750301361083984, + "learning_rate": 1.1911807373714224e-05, + "loss": 3.1231, + "step": 11365 + }, + { + "epoch": 0.911266561103205, + "grad_norm": 1.8743807077407837, + "learning_rate": 1.1910532860594283e-05, + "loss": 3.208, + "step": 11366 + }, + { + "epoch": 0.9113467358842277, + "grad_norm": 1.7342034578323364, + "learning_rate": 1.1909258315263435e-05, + "loss": 3.2233, + "step": 11367 + }, + { + "epoch": 0.9114269106652503, + "grad_norm": 1.8975117206573486, + "learning_rate": 1.1907983737743161e-05, + "loss": 3.2477, + "step": 11368 + }, + { + "epoch": 0.9115070854462729, + "grad_norm": 1.8670668601989746, + "learning_rate": 1.1906709128054956e-05, + "loss": 3.2147, + "step": 11369 + }, + { + "epoch": 0.9115872602272955, + "grad_norm": 1.9185923337936401, + "learning_rate": 1.1905434486220304e-05, + "loss": 3.2172, + "step": 11370 + }, + { + "epoch": 0.9116674350083181, + "grad_norm": 1.7967694997787476, + "learning_rate": 1.1904159812260698e-05, + "loss": 3.1511, + "step": 11371 + }, + { + "epoch": 0.9117476097893408, + "grad_norm": 1.8144959211349487, + "learning_rate": 1.1902885106197633e-05, + "loss": 3.2097, + "step": 11372 + }, + { + "epoch": 0.9118277845703634, + "grad_norm": 1.742833137512207, + "learning_rate": 1.1901610368052592e-05, + "loss": 3.224, + "step": 11373 + }, + { + "epoch": 0.911907959351386, + "grad_norm": 1.7394683361053467, + "learning_rate": 1.1900335597847073e-05, + "loss": 3.2576, + "step": 11374 + }, + { + "epoch": 0.9119881341324086, + "grad_norm": 1.8370304107666016, + "learning_rate": 1.1899060795602563e-05, + "loss": 3.2184, + "step": 11375 + }, + { + "epoch": 0.9120683089134313, + "grad_norm": 2.0371272563934326, + "learning_rate": 1.1897785961340557e-05, + "loss": 3.1454, + "step": 11376 + }, + { + "epoch": 0.9121484836944539, + "grad_norm": 1.783182144165039, + "learning_rate": 1.1896511095082552e-05, + "loss": 3.2485, + "step": 11377 + }, + { + "epoch": 0.9122286584754765, + "grad_norm": 1.9038770198822021, + "learning_rate": 1.1895236196850034e-05, + "loss": 3.275, + "step": 11378 + }, + { + "epoch": 0.9123088332564991, + "grad_norm": 1.8315014839172363, + "learning_rate": 1.1893961266664507e-05, + "loss": 3.2322, + "step": 11379 + }, + { + "epoch": 0.9123890080375218, + "grad_norm": 1.78927481174469, + "learning_rate": 1.1892686304547456e-05, + "loss": 3.1381, + "step": 11380 + }, + { + "epoch": 0.9124691828185444, + "grad_norm": 1.8239078521728516, + "learning_rate": 1.1891411310520382e-05, + "loss": 3.17, + "step": 11381 + }, + { + "epoch": 0.912549357599567, + "grad_norm": 1.7722892761230469, + "learning_rate": 1.1890136284604783e-05, + "loss": 3.203, + "step": 11382 + }, + { + "epoch": 0.9126295323805896, + "grad_norm": 1.8230899572372437, + "learning_rate": 1.1888861226822148e-05, + "loss": 3.1718, + "step": 11383 + }, + { + "epoch": 0.9127097071616123, + "grad_norm": 1.7029941082000732, + "learning_rate": 1.1887586137193982e-05, + "loss": 3.1409, + "step": 11384 + }, + { + "epoch": 0.912789881942635, + "grad_norm": 1.711194396018982, + "learning_rate": 1.1886311015741775e-05, + "loss": 3.164, + "step": 11385 + }, + { + "epoch": 0.9128700567236576, + "grad_norm": 1.774704098701477, + "learning_rate": 1.188503586248703e-05, + "loss": 3.2728, + "step": 11386 + }, + { + "epoch": 0.9129502315046802, + "grad_norm": 1.8665187358856201, + "learning_rate": 1.1883760677451248e-05, + "loss": 3.1069, + "step": 11387 + }, + { + "epoch": 0.9130304062857029, + "grad_norm": 1.8201066255569458, + "learning_rate": 1.1882485460655923e-05, + "loss": 3.1713, + "step": 11388 + }, + { + "epoch": 0.9131105810667255, + "grad_norm": 0.7149801254272461, + "learning_rate": 1.1881210212122556e-05, + "loss": 2.4789, + "step": 11389 + }, + { + "epoch": 0.9131907558477481, + "grad_norm": 1.8417142629623413, + "learning_rate": 1.1879934931872649e-05, + "loss": 3.2115, + "step": 11390 + }, + { + "epoch": 0.9132709306287707, + "grad_norm": 1.838252305984497, + "learning_rate": 1.18786596199277e-05, + "loss": 3.1198, + "step": 11391 + }, + { + "epoch": 0.9133511054097934, + "grad_norm": 1.9486966133117676, + "learning_rate": 1.1877384276309214e-05, + "loss": 3.1779, + "step": 11392 + }, + { + "epoch": 0.913431280190816, + "grad_norm": 1.7037506103515625, + "learning_rate": 1.187610890103869e-05, + "loss": 3.2587, + "step": 11393 + }, + { + "epoch": 0.9135114549718386, + "grad_norm": 1.6924846172332764, + "learning_rate": 1.187483349413763e-05, + "loss": 3.2195, + "step": 11394 + }, + { + "epoch": 0.9135916297528612, + "grad_norm": 1.7379133701324463, + "learning_rate": 1.1873558055627538e-05, + "loss": 3.0833, + "step": 11395 + }, + { + "epoch": 0.9136718045338839, + "grad_norm": 1.8275583982467651, + "learning_rate": 1.1872282585529918e-05, + "loss": 3.2748, + "step": 11396 + }, + { + "epoch": 0.9137519793149065, + "grad_norm": 1.7716758251190186, + "learning_rate": 1.1871007083866275e-05, + "loss": 3.271, + "step": 11397 + }, + { + "epoch": 0.9138321540959291, + "grad_norm": 0.7364860773086548, + "learning_rate": 1.186973155065811e-05, + "loss": 2.4839, + "step": 11398 + }, + { + "epoch": 0.9139123288769517, + "grad_norm": 1.876430630683899, + "learning_rate": 1.186845598592693e-05, + "loss": 3.1825, + "step": 11399 + }, + { + "epoch": 0.9139925036579744, + "grad_norm": 1.9457285404205322, + "learning_rate": 1.1867180389694242e-05, + "loss": 3.2782, + "step": 11400 + }, + { + "epoch": 0.914072678438997, + "grad_norm": 1.8418588638305664, + "learning_rate": 1.1865904761981551e-05, + "loss": 3.1868, + "step": 11401 + }, + { + "epoch": 0.9141528532200196, + "grad_norm": 2.0887229442596436, + "learning_rate": 1.1864629102810361e-05, + "loss": 3.1938, + "step": 11402 + }, + { + "epoch": 0.9142330280010422, + "grad_norm": 1.8476545810699463, + "learning_rate": 1.1863353412202184e-05, + "loss": 3.178, + "step": 11403 + }, + { + "epoch": 0.9143132027820648, + "grad_norm": 1.8060860633850098, + "learning_rate": 1.1862077690178522e-05, + "loss": 3.3135, + "step": 11404 + }, + { + "epoch": 0.9143933775630876, + "grad_norm": 0.7065815329551697, + "learning_rate": 1.186080193676089e-05, + "loss": 2.4784, + "step": 11405 + }, + { + "epoch": 0.9144735523441102, + "grad_norm": 1.8716671466827393, + "learning_rate": 1.185952615197079e-05, + "loss": 3.186, + "step": 11406 + }, + { + "epoch": 0.9145537271251328, + "grad_norm": 1.7359411716461182, + "learning_rate": 1.1858250335829737e-05, + "loss": 3.1749, + "step": 11407 + }, + { + "epoch": 0.9146339019061555, + "grad_norm": 1.8918763399124146, + "learning_rate": 1.1856974488359237e-05, + "loss": 3.2416, + "step": 11408 + }, + { + "epoch": 0.9147140766871781, + "grad_norm": 1.8120070695877075, + "learning_rate": 1.1855698609580801e-05, + "loss": 3.1608, + "step": 11409 + }, + { + "epoch": 0.9147942514682007, + "grad_norm": 1.777513861656189, + "learning_rate": 1.1854422699515942e-05, + "loss": 3.3093, + "step": 11410 + }, + { + "epoch": 0.9148744262492233, + "grad_norm": 1.7610989809036255, + "learning_rate": 1.185314675818617e-05, + "loss": 3.188, + "step": 11411 + }, + { + "epoch": 0.914954601030246, + "grad_norm": 1.751014232635498, + "learning_rate": 1.1851870785612996e-05, + "loss": 3.1581, + "step": 11412 + }, + { + "epoch": 0.9150347758112686, + "grad_norm": 1.7425717115402222, + "learning_rate": 1.1850594781817932e-05, + "loss": 3.2705, + "step": 11413 + }, + { + "epoch": 0.9151149505922912, + "grad_norm": 1.831465244293213, + "learning_rate": 1.1849318746822491e-05, + "loss": 3.2129, + "step": 11414 + }, + { + "epoch": 0.9151951253733138, + "grad_norm": 1.8180605173110962, + "learning_rate": 1.1848042680648193e-05, + "loss": 3.1909, + "step": 11415 + }, + { + "epoch": 0.9152753001543364, + "grad_norm": 1.7902337312698364, + "learning_rate": 1.1846766583316542e-05, + "loss": 3.1693, + "step": 11416 + }, + { + "epoch": 0.9153554749353591, + "grad_norm": 1.8213961124420166, + "learning_rate": 1.184549045484906e-05, + "loss": 3.241, + "step": 11417 + }, + { + "epoch": 0.9154356497163817, + "grad_norm": 1.6669328212738037, + "learning_rate": 1.1844214295267258e-05, + "loss": 3.0942, + "step": 11418 + }, + { + "epoch": 0.9155158244974043, + "grad_norm": 1.7214373350143433, + "learning_rate": 1.1842938104592652e-05, + "loss": 3.1591, + "step": 11419 + }, + { + "epoch": 0.9155959992784269, + "grad_norm": 1.7216261625289917, + "learning_rate": 1.1841661882846763e-05, + "loss": 3.1699, + "step": 11420 + }, + { + "epoch": 0.9156761740594496, + "grad_norm": 1.795363426208496, + "learning_rate": 1.1840385630051099e-05, + "loss": 3.2764, + "step": 11421 + }, + { + "epoch": 0.9157563488404722, + "grad_norm": 1.8323206901550293, + "learning_rate": 1.1839109346227185e-05, + "loss": 3.1932, + "step": 11422 + }, + { + "epoch": 0.9158365236214948, + "grad_norm": 1.807249665260315, + "learning_rate": 1.1837833031396536e-05, + "loss": 3.2698, + "step": 11423 + }, + { + "epoch": 0.9159166984025174, + "grad_norm": 1.7436317205429077, + "learning_rate": 1.1836556685580665e-05, + "loss": 3.1813, + "step": 11424 + }, + { + "epoch": 0.9159968731835402, + "grad_norm": 1.8833178281784058, + "learning_rate": 1.1835280308801101e-05, + "loss": 3.3843, + "step": 11425 + }, + { + "epoch": 0.9160770479645628, + "grad_norm": 1.752273678779602, + "learning_rate": 1.1834003901079354e-05, + "loss": 3.183, + "step": 11426 + }, + { + "epoch": 0.9161572227455854, + "grad_norm": 1.7246865034103394, + "learning_rate": 1.183272746243695e-05, + "loss": 3.1927, + "step": 11427 + }, + { + "epoch": 0.916237397526608, + "grad_norm": 0.7217278480529785, + "learning_rate": 1.1831450992895406e-05, + "loss": 2.4852, + "step": 11428 + }, + { + "epoch": 0.9163175723076307, + "grad_norm": 1.884339690208435, + "learning_rate": 1.1830174492476243e-05, + "loss": 3.2593, + "step": 11429 + }, + { + "epoch": 0.9163977470886533, + "grad_norm": 1.8082642555236816, + "learning_rate": 1.1828897961200984e-05, + "loss": 3.1693, + "step": 11430 + }, + { + "epoch": 0.9164779218696759, + "grad_norm": 1.8462051153182983, + "learning_rate": 1.182762139909115e-05, + "loss": 3.1205, + "step": 11431 + }, + { + "epoch": 0.9165580966506985, + "grad_norm": 1.7279314994812012, + "learning_rate": 1.1826344806168265e-05, + "loss": 3.1384, + "step": 11432 + }, + { + "epoch": 0.9166382714317212, + "grad_norm": 1.7313896417617798, + "learning_rate": 1.1825068182453847e-05, + "loss": 3.3033, + "step": 11433 + }, + { + "epoch": 0.9167184462127438, + "grad_norm": 1.7339503765106201, + "learning_rate": 1.1823791527969423e-05, + "loss": 3.2394, + "step": 11434 + }, + { + "epoch": 0.9167986209937664, + "grad_norm": 1.801002860069275, + "learning_rate": 1.1822514842736519e-05, + "loss": 3.2577, + "step": 11435 + }, + { + "epoch": 0.916878795774789, + "grad_norm": 1.8157292604446411, + "learning_rate": 1.1821238126776655e-05, + "loss": 3.1293, + "step": 11436 + }, + { + "epoch": 0.9169589705558117, + "grad_norm": 1.8129236698150635, + "learning_rate": 1.181996138011136e-05, + "loss": 3.2253, + "step": 11437 + }, + { + "epoch": 0.9170391453368343, + "grad_norm": 1.7799628973007202, + "learning_rate": 1.1818684602762156e-05, + "loss": 3.0695, + "step": 11438 + }, + { + "epoch": 0.9171193201178569, + "grad_norm": 1.7073240280151367, + "learning_rate": 1.1817407794750571e-05, + "loss": 3.2899, + "step": 11439 + }, + { + "epoch": 0.9171994948988795, + "grad_norm": 1.8115801811218262, + "learning_rate": 1.181613095609813e-05, + "loss": 3.1641, + "step": 11440 + }, + { + "epoch": 0.9172796696799022, + "grad_norm": 1.8581219911575317, + "learning_rate": 1.1814854086826364e-05, + "loss": 3.2849, + "step": 11441 + }, + { + "epoch": 0.9173598444609248, + "grad_norm": 0.699475109577179, + "learning_rate": 1.1813577186956797e-05, + "loss": 2.4729, + "step": 11442 + }, + { + "epoch": 0.9174400192419474, + "grad_norm": 1.7689698934555054, + "learning_rate": 1.1812300256510955e-05, + "loss": 3.209, + "step": 11443 + }, + { + "epoch": 0.91752019402297, + "grad_norm": 1.7671469449996948, + "learning_rate": 1.1811023295510372e-05, + "loss": 3.2695, + "step": 11444 + }, + { + "epoch": 0.9176003688039926, + "grad_norm": 1.8150253295898438, + "learning_rate": 1.1809746303976573e-05, + "loss": 3.2357, + "step": 11445 + }, + { + "epoch": 0.9176805435850154, + "grad_norm": 1.7632176876068115, + "learning_rate": 1.180846928193109e-05, + "loss": 3.2477, + "step": 11446 + }, + { + "epoch": 0.917760718366038, + "grad_norm": 1.813054084777832, + "learning_rate": 1.1807192229395452e-05, + "loss": 3.1675, + "step": 11447 + }, + { + "epoch": 0.9178408931470606, + "grad_norm": 0.6973825097084045, + "learning_rate": 1.1805915146391189e-05, + "loss": 2.4968, + "step": 11448 + }, + { + "epoch": 0.9179210679280833, + "grad_norm": 1.830973744392395, + "learning_rate": 1.1804638032939835e-05, + "loss": 3.205, + "step": 11449 + }, + { + "epoch": 0.9180012427091059, + "grad_norm": 1.800181269645691, + "learning_rate": 1.1803360889062919e-05, + "loss": 3.2034, + "step": 11450 + }, + { + "epoch": 0.9180814174901285, + "grad_norm": 1.7227931022644043, + "learning_rate": 1.1802083714781973e-05, + "loss": 3.1741, + "step": 11451 + }, + { + "epoch": 0.9181615922711511, + "grad_norm": 1.762218713760376, + "learning_rate": 1.1800806510118532e-05, + "loss": 3.171, + "step": 11452 + }, + { + "epoch": 0.9182417670521738, + "grad_norm": 1.8887745141983032, + "learning_rate": 1.1799529275094126e-05, + "loss": 3.224, + "step": 11453 + }, + { + "epoch": 0.9183219418331964, + "grad_norm": 1.8094112873077393, + "learning_rate": 1.1798252009730292e-05, + "loss": 3.2554, + "step": 11454 + }, + { + "epoch": 0.918402116614219, + "grad_norm": 1.7559362649917603, + "learning_rate": 1.1796974714048564e-05, + "loss": 3.2144, + "step": 11455 + }, + { + "epoch": 0.9184822913952416, + "grad_norm": 1.8024733066558838, + "learning_rate": 1.1795697388070476e-05, + "loss": 3.2226, + "step": 11456 + }, + { + "epoch": 0.9185624661762642, + "grad_norm": 1.708678126335144, + "learning_rate": 1.1794420031817561e-05, + "loss": 3.2417, + "step": 11457 + }, + { + "epoch": 0.9186426409572869, + "grad_norm": 1.791016697883606, + "learning_rate": 1.1793142645311355e-05, + "loss": 3.2071, + "step": 11458 + }, + { + "epoch": 0.9187228157383095, + "grad_norm": 1.7623440027236938, + "learning_rate": 1.1791865228573398e-05, + "loss": 3.3908, + "step": 11459 + }, + { + "epoch": 0.9188029905193321, + "grad_norm": 1.738843321800232, + "learning_rate": 1.1790587781625227e-05, + "loss": 3.1683, + "step": 11460 + }, + { + "epoch": 0.9188831653003547, + "grad_norm": 1.7581545114517212, + "learning_rate": 1.1789310304488373e-05, + "loss": 3.1639, + "step": 11461 + }, + { + "epoch": 0.9189633400813774, + "grad_norm": 1.725192904472351, + "learning_rate": 1.1788032797184381e-05, + "loss": 3.3528, + "step": 11462 + }, + { + "epoch": 0.9190435148624, + "grad_norm": 1.8388313055038452, + "learning_rate": 1.1786755259734783e-05, + "loss": 3.199, + "step": 11463 + }, + { + "epoch": 0.9191236896434226, + "grad_norm": 0.7231155037879944, + "learning_rate": 1.1785477692161123e-05, + "loss": 2.4733, + "step": 11464 + }, + { + "epoch": 0.9192038644244452, + "grad_norm": 1.7509760856628418, + "learning_rate": 1.1784200094484941e-05, + "loss": 3.2336, + "step": 11465 + }, + { + "epoch": 0.919284039205468, + "grad_norm": 1.8172500133514404, + "learning_rate": 1.178292246672777e-05, + "loss": 3.231, + "step": 11466 + }, + { + "epoch": 0.9193642139864906, + "grad_norm": 1.918960690498352, + "learning_rate": 1.1781644808911157e-05, + "loss": 3.1914, + "step": 11467 + }, + { + "epoch": 0.9194443887675132, + "grad_norm": 1.7668380737304688, + "learning_rate": 1.1780367121056639e-05, + "loss": 3.1851, + "step": 11468 + }, + { + "epoch": 0.9195245635485358, + "grad_norm": 1.7150969505310059, + "learning_rate": 1.1779089403185759e-05, + "loss": 3.1371, + "step": 11469 + }, + { + "epoch": 0.9196047383295585, + "grad_norm": 1.9403114318847656, + "learning_rate": 1.177781165532006e-05, + "loss": 3.3317, + "step": 11470 + }, + { + "epoch": 0.9196849131105811, + "grad_norm": 1.7723972797393799, + "learning_rate": 1.1776533877481081e-05, + "loss": 3.1898, + "step": 11471 + }, + { + "epoch": 0.9197650878916037, + "grad_norm": 1.8916361331939697, + "learning_rate": 1.1775256069690371e-05, + "loss": 3.2007, + "step": 11472 + }, + { + "epoch": 0.9198452626726263, + "grad_norm": 1.676382064819336, + "learning_rate": 1.1773978231969465e-05, + "loss": 3.2734, + "step": 11473 + }, + { + "epoch": 0.919925437453649, + "grad_norm": 1.8321868181228638, + "learning_rate": 1.1772700364339911e-05, + "loss": 3.1724, + "step": 11474 + }, + { + "epoch": 0.9200056122346716, + "grad_norm": 1.874668836593628, + "learning_rate": 1.1771422466823256e-05, + "loss": 3.3718, + "step": 11475 + }, + { + "epoch": 0.9200857870156942, + "grad_norm": 1.778359055519104, + "learning_rate": 1.1770144539441042e-05, + "loss": 3.1357, + "step": 11476 + }, + { + "epoch": 0.9201659617967168, + "grad_norm": 0.6815299987792969, + "learning_rate": 1.1768866582214815e-05, + "loss": 2.4885, + "step": 11477 + }, + { + "epoch": 0.9202461365777395, + "grad_norm": 1.8160094022750854, + "learning_rate": 1.176758859516612e-05, + "loss": 3.2307, + "step": 11478 + }, + { + "epoch": 0.9203263113587621, + "grad_norm": 1.681005597114563, + "learning_rate": 1.1766310578316504e-05, + "loss": 3.2064, + "step": 11479 + }, + { + "epoch": 0.9204064861397847, + "grad_norm": 1.7867448329925537, + "learning_rate": 1.1765032531687517e-05, + "loss": 3.1281, + "step": 11480 + }, + { + "epoch": 0.9204866609208073, + "grad_norm": 1.7374333143234253, + "learning_rate": 1.17637544553007e-05, + "loss": 3.1464, + "step": 11481 + }, + { + "epoch": 0.92056683570183, + "grad_norm": 1.6487220525741577, + "learning_rate": 1.1762476349177607e-05, + "loss": 3.1998, + "step": 11482 + }, + { + "epoch": 0.9206470104828526, + "grad_norm": 1.9013420343399048, + "learning_rate": 1.1761198213339783e-05, + "loss": 3.3033, + "step": 11483 + }, + { + "epoch": 0.9207271852638752, + "grad_norm": 1.7581757307052612, + "learning_rate": 1.1759920047808774e-05, + "loss": 3.2723, + "step": 11484 + }, + { + "epoch": 0.9208073600448978, + "grad_norm": 1.7809545993804932, + "learning_rate": 1.1758641852606141e-05, + "loss": 3.2596, + "step": 11485 + }, + { + "epoch": 0.9208875348259206, + "grad_norm": 1.7101136445999146, + "learning_rate": 1.175736362775342e-05, + "loss": 3.1898, + "step": 11486 + }, + { + "epoch": 0.9209677096069432, + "grad_norm": 1.7477248907089233, + "learning_rate": 1.1756085373272173e-05, + "loss": 3.1245, + "step": 11487 + }, + { + "epoch": 0.9210478843879658, + "grad_norm": 0.7234163880348206, + "learning_rate": 1.1754807089183939e-05, + "loss": 2.4758, + "step": 11488 + }, + { + "epoch": 0.9211280591689884, + "grad_norm": 1.7761086225509644, + "learning_rate": 1.1753528775510278e-05, + "loss": 3.2755, + "step": 11489 + }, + { + "epoch": 0.9212082339500111, + "grad_norm": 1.7651418447494507, + "learning_rate": 1.1752250432272745e-05, + "loss": 3.2712, + "step": 11490 + }, + { + "epoch": 0.9212884087310337, + "grad_norm": 1.82453453540802, + "learning_rate": 1.175097205949288e-05, + "loss": 3.1376, + "step": 11491 + }, + { + "epoch": 0.9213685835120563, + "grad_norm": 1.807139277458191, + "learning_rate": 1.174969365719225e-05, + "loss": 3.1888, + "step": 11492 + }, + { + "epoch": 0.9214487582930789, + "grad_norm": 1.783159613609314, + "learning_rate": 1.1748415225392396e-05, + "loss": 3.1702, + "step": 11493 + }, + { + "epoch": 0.9215289330741016, + "grad_norm": 1.7764755487442017, + "learning_rate": 1.174713676411488e-05, + "loss": 3.166, + "step": 11494 + }, + { + "epoch": 0.9216091078551242, + "grad_norm": 1.7263617515563965, + "learning_rate": 1.1745858273381255e-05, + "loss": 3.1654, + "step": 11495 + }, + { + "epoch": 0.9216892826361468, + "grad_norm": 1.8481953144073486, + "learning_rate": 1.1744579753213073e-05, + "loss": 3.3292, + "step": 11496 + }, + { + "epoch": 0.9217694574171694, + "grad_norm": 1.8472890853881836, + "learning_rate": 1.1743301203631892e-05, + "loss": 3.1965, + "step": 11497 + }, + { + "epoch": 0.921849632198192, + "grad_norm": 1.736473560333252, + "learning_rate": 1.1742022624659264e-05, + "loss": 3.1809, + "step": 11498 + }, + { + "epoch": 0.9219298069792147, + "grad_norm": 1.758235216140747, + "learning_rate": 1.1740744016316751e-05, + "loss": 3.2474, + "step": 11499 + }, + { + "epoch": 0.9220099817602373, + "grad_norm": 1.788291573524475, + "learning_rate": 1.1739465378625909e-05, + "loss": 3.3312, + "step": 11500 + }, + { + "epoch": 0.9220901565412599, + "grad_norm": 1.6791337728500366, + "learning_rate": 1.173818671160829e-05, + "loss": 3.1282, + "step": 11501 + }, + { + "epoch": 0.9221703313222825, + "grad_norm": 1.7464617490768433, + "learning_rate": 1.173690801528546e-05, + "loss": 3.2103, + "step": 11502 + }, + { + "epoch": 0.9222505061033052, + "grad_norm": 1.8879432678222656, + "learning_rate": 1.1735629289678969e-05, + "loss": 3.1204, + "step": 11503 + }, + { + "epoch": 0.9223306808843278, + "grad_norm": 1.7934650182724, + "learning_rate": 1.1734350534810381e-05, + "loss": 3.2499, + "step": 11504 + }, + { + "epoch": 0.9224108556653504, + "grad_norm": 1.8459309339523315, + "learning_rate": 1.1733071750701257e-05, + "loss": 3.1315, + "step": 11505 + }, + { + "epoch": 0.922491030446373, + "grad_norm": 1.9439069032669067, + "learning_rate": 1.1731792937373151e-05, + "loss": 3.1973, + "step": 11506 + }, + { + "epoch": 0.9225712052273958, + "grad_norm": 1.7848732471466064, + "learning_rate": 1.173051409484763e-05, + "loss": 3.2007, + "step": 11507 + }, + { + "epoch": 0.9226513800084184, + "grad_norm": 1.7919604778289795, + "learning_rate": 1.1729235223146248e-05, + "loss": 3.2016, + "step": 11508 + }, + { + "epoch": 0.922731554789441, + "grad_norm": 1.7048914432525635, + "learning_rate": 1.1727956322290568e-05, + "loss": 3.1327, + "step": 11509 + }, + { + "epoch": 0.9228117295704636, + "grad_norm": 1.8658809661865234, + "learning_rate": 1.1726677392302157e-05, + "loss": 3.3318, + "step": 11510 + }, + { + "epoch": 0.9228919043514863, + "grad_norm": 1.8110729455947876, + "learning_rate": 1.1725398433202572e-05, + "loss": 3.2183, + "step": 11511 + }, + { + "epoch": 0.9229720791325089, + "grad_norm": 1.80364990234375, + "learning_rate": 1.1724119445013379e-05, + "loss": 3.1767, + "step": 11512 + }, + { + "epoch": 0.9230522539135315, + "grad_norm": 1.7591701745986938, + "learning_rate": 1.172284042775614e-05, + "loss": 3.1235, + "step": 11513 + }, + { + "epoch": 0.9231324286945541, + "grad_norm": 1.7857890129089355, + "learning_rate": 1.1721561381452416e-05, + "loss": 3.1219, + "step": 11514 + }, + { + "epoch": 0.9232126034755768, + "grad_norm": 1.8788700103759766, + "learning_rate": 1.1720282306123776e-05, + "loss": 3.1174, + "step": 11515 + }, + { + "epoch": 0.9232927782565994, + "grad_norm": 1.8076565265655518, + "learning_rate": 1.171900320179178e-05, + "loss": 3.3746, + "step": 11516 + }, + { + "epoch": 0.923372953037622, + "grad_norm": 1.806984782218933, + "learning_rate": 1.1717724068477998e-05, + "loss": 3.1755, + "step": 11517 + }, + { + "epoch": 0.9234531278186446, + "grad_norm": 1.718265414237976, + "learning_rate": 1.1716444906203995e-05, + "loss": 3.1671, + "step": 11518 + }, + { + "epoch": 0.9235333025996673, + "grad_norm": 1.7454946041107178, + "learning_rate": 1.1715165714991333e-05, + "loss": 3.2424, + "step": 11519 + }, + { + "epoch": 0.9236134773806899, + "grad_norm": 1.6770243644714355, + "learning_rate": 1.1713886494861581e-05, + "loss": 3.1846, + "step": 11520 + }, + { + "epoch": 0.9236936521617125, + "grad_norm": 1.6977506875991821, + "learning_rate": 1.171260724583631e-05, + "loss": 3.2565, + "step": 11521 + }, + { + "epoch": 0.9237738269427351, + "grad_norm": 1.8051031827926636, + "learning_rate": 1.1711327967937083e-05, + "loss": 3.2062, + "step": 11522 + }, + { + "epoch": 0.9238540017237578, + "grad_norm": 1.7045859098434448, + "learning_rate": 1.1710048661185466e-05, + "loss": 3.1509, + "step": 11523 + }, + { + "epoch": 0.9239341765047804, + "grad_norm": 1.8386478424072266, + "learning_rate": 1.1708769325603035e-05, + "loss": 3.1683, + "step": 11524 + }, + { + "epoch": 0.924014351285803, + "grad_norm": 1.832967758178711, + "learning_rate": 1.1707489961211356e-05, + "loss": 3.2444, + "step": 11525 + }, + { + "epoch": 0.9240945260668256, + "grad_norm": 1.7113286256790161, + "learning_rate": 1.1706210568031998e-05, + "loss": 3.0962, + "step": 11526 + }, + { + "epoch": 0.9241747008478484, + "grad_norm": 1.8022456169128418, + "learning_rate": 1.170493114608653e-05, + "loss": 3.2441, + "step": 11527 + }, + { + "epoch": 0.924254875628871, + "grad_norm": 1.8577632904052734, + "learning_rate": 1.1703651695396523e-05, + "loss": 3.2026, + "step": 11528 + }, + { + "epoch": 0.9243350504098936, + "grad_norm": 1.901499629020691, + "learning_rate": 1.1702372215983549e-05, + "loss": 3.2854, + "step": 11529 + }, + { + "epoch": 0.9244152251909162, + "grad_norm": 1.8007863759994507, + "learning_rate": 1.1701092707869182e-05, + "loss": 3.1422, + "step": 11530 + }, + { + "epoch": 0.9244953999719389, + "grad_norm": 1.906842589378357, + "learning_rate": 1.169981317107499e-05, + "loss": 3.2174, + "step": 11531 + }, + { + "epoch": 0.9245755747529615, + "grad_norm": 1.7538970708847046, + "learning_rate": 1.1698533605622547e-05, + "loss": 3.2012, + "step": 11532 + }, + { + "epoch": 0.9246557495339841, + "grad_norm": 1.7922523021697998, + "learning_rate": 1.1697254011533424e-05, + "loss": 3.1938, + "step": 11533 + }, + { + "epoch": 0.9247359243150067, + "grad_norm": 0.7359314560890198, + "learning_rate": 1.1695974388829197e-05, + "loss": 2.4871, + "step": 11534 + }, + { + "epoch": 0.9248160990960294, + "grad_norm": 1.7393302917480469, + "learning_rate": 1.1694694737531442e-05, + "loss": 3.2146, + "step": 11535 + }, + { + "epoch": 0.924896273877052, + "grad_norm": 1.8642264604568481, + "learning_rate": 1.169341505766173e-05, + "loss": 3.1945, + "step": 11536 + }, + { + "epoch": 0.9249764486580746, + "grad_norm": 1.749790072441101, + "learning_rate": 1.1692135349241639e-05, + "loss": 3.292, + "step": 11537 + }, + { + "epoch": 0.9250566234390972, + "grad_norm": 1.7298674583435059, + "learning_rate": 1.1690855612292737e-05, + "loss": 3.1542, + "step": 11538 + }, + { + "epoch": 0.9251367982201198, + "grad_norm": 1.8040565252304077, + "learning_rate": 1.1689575846836607e-05, + "loss": 3.1575, + "step": 11539 + }, + { + "epoch": 0.9252169730011425, + "grad_norm": 1.7888880968093872, + "learning_rate": 1.1688296052894828e-05, + "loss": 3.0923, + "step": 11540 + }, + { + "epoch": 0.9252971477821651, + "grad_norm": 1.8346723318099976, + "learning_rate": 1.1687016230488968e-05, + "loss": 3.191, + "step": 11541 + }, + { + "epoch": 0.9253773225631877, + "grad_norm": 1.7992501258850098, + "learning_rate": 1.1685736379640612e-05, + "loss": 3.2377, + "step": 11542 + }, + { + "epoch": 0.9254574973442103, + "grad_norm": 1.8310868740081787, + "learning_rate": 1.1684456500371333e-05, + "loss": 3.322, + "step": 11543 + }, + { + "epoch": 0.925537672125233, + "grad_norm": 1.7840425968170166, + "learning_rate": 1.168317659270271e-05, + "loss": 3.116, + "step": 11544 + }, + { + "epoch": 0.9256178469062556, + "grad_norm": 1.818141222000122, + "learning_rate": 1.1681896656656324e-05, + "loss": 3.2125, + "step": 11545 + }, + { + "epoch": 0.9256980216872782, + "grad_norm": 1.7856196165084839, + "learning_rate": 1.1680616692253753e-05, + "loss": 3.1782, + "step": 11546 + }, + { + "epoch": 0.925778196468301, + "grad_norm": 0.7081677913665771, + "learning_rate": 1.167933669951658e-05, + "loss": 2.479, + "step": 11547 + }, + { + "epoch": 0.9258583712493236, + "grad_norm": 1.7862322330474854, + "learning_rate": 1.1678056678466376e-05, + "loss": 3.236, + "step": 11548 + }, + { + "epoch": 0.9259385460303462, + "grad_norm": 1.8255561590194702, + "learning_rate": 1.167677662912473e-05, + "loss": 3.2228, + "step": 11549 + }, + { + "epoch": 0.9260187208113688, + "grad_norm": 1.831946849822998, + "learning_rate": 1.1675496551513224e-05, + "loss": 3.2072, + "step": 11550 + }, + { + "epoch": 0.9260988955923914, + "grad_norm": 1.7246079444885254, + "learning_rate": 1.1674216445653435e-05, + "loss": 3.2185, + "step": 11551 + }, + { + "epoch": 0.9261790703734141, + "grad_norm": 1.7533811330795288, + "learning_rate": 1.1672936311566948e-05, + "loss": 3.2373, + "step": 11552 + }, + { + "epoch": 0.9262592451544367, + "grad_norm": 1.7503514289855957, + "learning_rate": 1.167165614927534e-05, + "loss": 3.1822, + "step": 11553 + }, + { + "epoch": 0.9263394199354593, + "grad_norm": 1.84312903881073, + "learning_rate": 1.1670375958800203e-05, + "loss": 3.183, + "step": 11554 + }, + { + "epoch": 0.9264195947164819, + "grad_norm": 1.8261382579803467, + "learning_rate": 1.1669095740163115e-05, + "loss": 3.2824, + "step": 11555 + }, + { + "epoch": 0.9264997694975046, + "grad_norm": 1.8490568399429321, + "learning_rate": 1.166781549338566e-05, + "loss": 3.2877, + "step": 11556 + }, + { + "epoch": 0.9265799442785272, + "grad_norm": 1.6427431106567383, + "learning_rate": 1.1666535218489423e-05, + "loss": 3.2924, + "step": 11557 + }, + { + "epoch": 0.9266601190595498, + "grad_norm": 1.9857040643692017, + "learning_rate": 1.1665254915495989e-05, + "loss": 3.1937, + "step": 11558 + }, + { + "epoch": 0.9267402938405724, + "grad_norm": 1.7434512376785278, + "learning_rate": 1.1663974584426944e-05, + "loss": 3.1493, + "step": 11559 + }, + { + "epoch": 0.9268204686215951, + "grad_norm": 0.7532596588134766, + "learning_rate": 1.1662694225303879e-05, + "loss": 2.4862, + "step": 11560 + }, + { + "epoch": 0.9269006434026177, + "grad_norm": 1.7943838834762573, + "learning_rate": 1.166141383814837e-05, + "loss": 3.2574, + "step": 11561 + }, + { + "epoch": 0.9269808181836403, + "grad_norm": 1.7562024593353271, + "learning_rate": 1.1660133422982013e-05, + "loss": 3.1101, + "step": 11562 + }, + { + "epoch": 0.9270609929646629, + "grad_norm": 1.8316055536270142, + "learning_rate": 1.165885297982639e-05, + "loss": 3.1822, + "step": 11563 + }, + { + "epoch": 0.9271411677456856, + "grad_norm": 1.8166054487228394, + "learning_rate": 1.1657572508703087e-05, + "loss": 3.2422, + "step": 11564 + }, + { + "epoch": 0.9272213425267082, + "grad_norm": 1.8391894102096558, + "learning_rate": 1.16562920096337e-05, + "loss": 3.2825, + "step": 11565 + }, + { + "epoch": 0.9273015173077308, + "grad_norm": 1.966701626777649, + "learning_rate": 1.1655011482639813e-05, + "loss": 3.177, + "step": 11566 + }, + { + "epoch": 0.9273816920887534, + "grad_norm": 1.738802194595337, + "learning_rate": 1.1653730927743014e-05, + "loss": 3.203, + "step": 11567 + }, + { + "epoch": 0.9274618668697762, + "grad_norm": 1.7493582963943481, + "learning_rate": 1.1652450344964897e-05, + "loss": 3.2073, + "step": 11568 + }, + { + "epoch": 0.9275420416507988, + "grad_norm": 1.7409504652023315, + "learning_rate": 1.1651169734327046e-05, + "loss": 3.167, + "step": 11569 + }, + { + "epoch": 0.9276222164318214, + "grad_norm": 1.7612192630767822, + "learning_rate": 1.164988909585106e-05, + "loss": 3.2399, + "step": 11570 + }, + { + "epoch": 0.927702391212844, + "grad_norm": 1.7705552577972412, + "learning_rate": 1.164860842955852e-05, + "loss": 3.1604, + "step": 11571 + }, + { + "epoch": 0.9277825659938667, + "grad_norm": 0.7311280369758606, + "learning_rate": 1.1647327735471025e-05, + "loss": 2.4906, + "step": 11572 + }, + { + "epoch": 0.9278627407748893, + "grad_norm": 1.7624155282974243, + "learning_rate": 1.1646047013610166e-05, + "loss": 3.2547, + "step": 11573 + }, + { + "epoch": 0.9279429155559119, + "grad_norm": 1.7825331687927246, + "learning_rate": 1.1644766263997531e-05, + "loss": 3.207, + "step": 11574 + }, + { + "epoch": 0.9280230903369345, + "grad_norm": 1.8588950634002686, + "learning_rate": 1.1643485486654721e-05, + "loss": 3.2631, + "step": 11575 + }, + { + "epoch": 0.9281032651179572, + "grad_norm": 1.7367178201675415, + "learning_rate": 1.1642204681603323e-05, + "loss": 3.1589, + "step": 11576 + }, + { + "epoch": 0.9281834398989798, + "grad_norm": 1.869034767150879, + "learning_rate": 1.1640923848864934e-05, + "loss": 3.2719, + "step": 11577 + }, + { + "epoch": 0.9282636146800024, + "grad_norm": 1.7621041536331177, + "learning_rate": 1.1639642988461147e-05, + "loss": 3.113, + "step": 11578 + }, + { + "epoch": 0.928343789461025, + "grad_norm": 1.9300843477249146, + "learning_rate": 1.1638362100413553e-05, + "loss": 3.2526, + "step": 11579 + }, + { + "epoch": 0.9284239642420477, + "grad_norm": 1.7564282417297363, + "learning_rate": 1.1637081184743756e-05, + "loss": 3.2053, + "step": 11580 + }, + { + "epoch": 0.9285041390230703, + "grad_norm": 1.75326669216156, + "learning_rate": 1.1635800241473345e-05, + "loss": 3.1344, + "step": 11581 + }, + { + "epoch": 0.9285843138040929, + "grad_norm": 1.923758625984192, + "learning_rate": 1.1634519270623919e-05, + "loss": 3.1653, + "step": 11582 + }, + { + "epoch": 0.9286644885851155, + "grad_norm": 1.7322314977645874, + "learning_rate": 1.1633238272217075e-05, + "loss": 3.0494, + "step": 11583 + }, + { + "epoch": 0.9287446633661381, + "grad_norm": 1.7291650772094727, + "learning_rate": 1.1631957246274408e-05, + "loss": 3.2573, + "step": 11584 + }, + { + "epoch": 0.9288248381471608, + "grad_norm": 1.7131595611572266, + "learning_rate": 1.1630676192817521e-05, + "loss": 3.228, + "step": 11585 + }, + { + "epoch": 0.9289050129281834, + "grad_norm": 1.8028953075408936, + "learning_rate": 1.1629395111868003e-05, + "loss": 3.205, + "step": 11586 + }, + { + "epoch": 0.928985187709206, + "grad_norm": 1.7746144533157349, + "learning_rate": 1.162811400344746e-05, + "loss": 3.2701, + "step": 11587 + }, + { + "epoch": 0.9290653624902288, + "grad_norm": 0.7671176791191101, + "learning_rate": 1.1626832867577487e-05, + "loss": 2.4879, + "step": 11588 + }, + { + "epoch": 0.9291455372712514, + "grad_norm": 1.7376827001571655, + "learning_rate": 1.1625551704279685e-05, + "loss": 3.2178, + "step": 11589 + }, + { + "epoch": 0.929225712052274, + "grad_norm": 1.7037272453308105, + "learning_rate": 1.1624270513575657e-05, + "loss": 3.2021, + "step": 11590 + }, + { + "epoch": 0.9293058868332966, + "grad_norm": 1.7337592840194702, + "learning_rate": 1.1622989295487e-05, + "loss": 3.1742, + "step": 11591 + }, + { + "epoch": 0.9293860616143192, + "grad_norm": 1.7405849695205688, + "learning_rate": 1.1621708050035315e-05, + "loss": 3.1713, + "step": 11592 + }, + { + "epoch": 0.9294662363953419, + "grad_norm": 1.6913779973983765, + "learning_rate": 1.1620426777242204e-05, + "loss": 3.1819, + "step": 11593 + }, + { + "epoch": 0.9295464111763645, + "grad_norm": 1.7955435514450073, + "learning_rate": 1.1619145477129265e-05, + "loss": 3.1993, + "step": 11594 + }, + { + "epoch": 0.9296265859573871, + "grad_norm": 1.7687503099441528, + "learning_rate": 1.161786414971811e-05, + "loss": 3.1635, + "step": 11595 + }, + { + "epoch": 0.9297067607384097, + "grad_norm": 1.8187737464904785, + "learning_rate": 1.161658279503033e-05, + "loss": 3.2729, + "step": 11596 + }, + { + "epoch": 0.9297869355194324, + "grad_norm": 1.7221158742904663, + "learning_rate": 1.1615301413087536e-05, + "loss": 3.1863, + "step": 11597 + }, + { + "epoch": 0.929867110300455, + "grad_norm": 1.7267659902572632, + "learning_rate": 1.1614020003911333e-05, + "loss": 3.2472, + "step": 11598 + }, + { + "epoch": 0.9299472850814776, + "grad_norm": 1.7403596639633179, + "learning_rate": 1.1612738567523317e-05, + "loss": 3.2242, + "step": 11599 + }, + { + "epoch": 0.9300274598625002, + "grad_norm": 1.6871565580368042, + "learning_rate": 1.1611457103945098e-05, + "loss": 3.0437, + "step": 11600 + }, + { + "epoch": 0.9301076346435229, + "grad_norm": 1.7790422439575195, + "learning_rate": 1.1610175613198282e-05, + "loss": 3.1923, + "step": 11601 + }, + { + "epoch": 0.9301878094245455, + "grad_norm": 1.6772013902664185, + "learning_rate": 1.160889409530447e-05, + "loss": 3.154, + "step": 11602 + }, + { + "epoch": 0.9302679842055681, + "grad_norm": 1.9833179712295532, + "learning_rate": 1.1607612550285273e-05, + "loss": 3.2295, + "step": 11603 + }, + { + "epoch": 0.9303481589865907, + "grad_norm": 1.9264132976531982, + "learning_rate": 1.1606330978162292e-05, + "loss": 3.1532, + "step": 11604 + }, + { + "epoch": 0.9304283337676134, + "grad_norm": 1.801839828491211, + "learning_rate": 1.160504937895714e-05, + "loss": 3.1819, + "step": 11605 + }, + { + "epoch": 0.930508508548636, + "grad_norm": 0.7047182321548462, + "learning_rate": 1.1603767752691419e-05, + "loss": 2.4763, + "step": 11606 + }, + { + "epoch": 0.9305886833296586, + "grad_norm": 1.8576565980911255, + "learning_rate": 1.160248609938674e-05, + "loss": 3.203, + "step": 11607 + }, + { + "epoch": 0.9306688581106813, + "grad_norm": 1.8588788509368896, + "learning_rate": 1.1601204419064712e-05, + "loss": 3.1591, + "step": 11608 + }, + { + "epoch": 0.930749032891704, + "grad_norm": 1.7610799074172974, + "learning_rate": 1.1599922711746938e-05, + "loss": 3.1346, + "step": 11609 + }, + { + "epoch": 0.9308292076727266, + "grad_norm": 1.7013911008834839, + "learning_rate": 1.1598640977455036e-05, + "loss": 3.2158, + "step": 11610 + }, + { + "epoch": 0.9309093824537492, + "grad_norm": 1.7274351119995117, + "learning_rate": 1.1597359216210604e-05, + "loss": 3.1341, + "step": 11611 + }, + { + "epoch": 0.9309895572347718, + "grad_norm": 1.728446364402771, + "learning_rate": 1.1596077428035262e-05, + "loss": 3.1511, + "step": 11612 + }, + { + "epoch": 0.9310697320157945, + "grad_norm": 1.8873517513275146, + "learning_rate": 1.1594795612950618e-05, + "loss": 3.2643, + "step": 11613 + }, + { + "epoch": 0.9311499067968171, + "grad_norm": 1.6718040704727173, + "learning_rate": 1.159351377097828e-05, + "loss": 3.1863, + "step": 11614 + }, + { + "epoch": 0.9312300815778397, + "grad_norm": 1.7241754531860352, + "learning_rate": 1.1592231902139863e-05, + "loss": 3.218, + "step": 11615 + }, + { + "epoch": 0.9313102563588623, + "grad_norm": 1.7816237211227417, + "learning_rate": 1.1590950006456976e-05, + "loss": 3.1465, + "step": 11616 + }, + { + "epoch": 0.931390431139885, + "grad_norm": 1.7466307878494263, + "learning_rate": 1.1589668083951234e-05, + "loss": 3.2089, + "step": 11617 + }, + { + "epoch": 0.9314706059209076, + "grad_norm": 1.7220313549041748, + "learning_rate": 1.1588386134644248e-05, + "loss": 3.0568, + "step": 11618 + }, + { + "epoch": 0.9315507807019302, + "grad_norm": 1.6984692811965942, + "learning_rate": 1.158710415855763e-05, + "loss": 3.2041, + "step": 11619 + }, + { + "epoch": 0.9316309554829528, + "grad_norm": 0.730307400226593, + "learning_rate": 1.1585822155712998e-05, + "loss": 2.4725, + "step": 11620 + }, + { + "epoch": 0.9317111302639755, + "grad_norm": 1.7857177257537842, + "learning_rate": 1.158454012613196e-05, + "loss": 3.1697, + "step": 11621 + }, + { + "epoch": 0.9317913050449981, + "grad_norm": 1.8058513402938843, + "learning_rate": 1.1583258069836137e-05, + "loss": 3.2502, + "step": 11622 + }, + { + "epoch": 0.9318714798260207, + "grad_norm": 1.8379589319229126, + "learning_rate": 1.158197598684714e-05, + "loss": 3.0554, + "step": 11623 + }, + { + "epoch": 0.9319516546070433, + "grad_norm": 1.806965708732605, + "learning_rate": 1.1580693877186583e-05, + "loss": 3.2965, + "step": 11624 + }, + { + "epoch": 0.932031829388066, + "grad_norm": 1.7995665073394775, + "learning_rate": 1.157941174087609e-05, + "loss": 3.1335, + "step": 11625 + }, + { + "epoch": 0.9321120041690886, + "grad_norm": 1.725684642791748, + "learning_rate": 1.1578129577937264e-05, + "loss": 3.1975, + "step": 11626 + }, + { + "epoch": 0.9321921789501112, + "grad_norm": 1.909946322441101, + "learning_rate": 1.1576847388391734e-05, + "loss": 3.1197, + "step": 11627 + }, + { + "epoch": 0.9322723537311338, + "grad_norm": 1.7529935836791992, + "learning_rate": 1.1575565172261113e-05, + "loss": 3.1918, + "step": 11628 + }, + { + "epoch": 0.9323525285121566, + "grad_norm": 1.8122034072875977, + "learning_rate": 1.1574282929567017e-05, + "loss": 3.2031, + "step": 11629 + }, + { + "epoch": 0.9324327032931792, + "grad_norm": 1.8065096139907837, + "learning_rate": 1.1573000660331069e-05, + "loss": 3.2324, + "step": 11630 + }, + { + "epoch": 0.9325128780742018, + "grad_norm": 1.798995852470398, + "learning_rate": 1.157171836457488e-05, + "loss": 3.0781, + "step": 11631 + }, + { + "epoch": 0.9325930528552244, + "grad_norm": 1.6826707124710083, + "learning_rate": 1.1570436042320074e-05, + "loss": 3.0897, + "step": 11632 + }, + { + "epoch": 0.932673227636247, + "grad_norm": 1.8861364126205444, + "learning_rate": 1.1569153693588273e-05, + "loss": 3.2573, + "step": 11633 + }, + { + "epoch": 0.9327534024172697, + "grad_norm": 1.6999481916427612, + "learning_rate": 1.1567871318401091e-05, + "loss": 3.3257, + "step": 11634 + }, + { + "epoch": 0.9328335771982923, + "grad_norm": 1.8562819957733154, + "learning_rate": 1.1566588916780153e-05, + "loss": 3.1797, + "step": 11635 + }, + { + "epoch": 0.9329137519793149, + "grad_norm": 1.8153514862060547, + "learning_rate": 1.1565306488747074e-05, + "loss": 3.3346, + "step": 11636 + }, + { + "epoch": 0.9329939267603375, + "grad_norm": 1.7775675058364868, + "learning_rate": 1.1564024034323483e-05, + "loss": 3.2202, + "step": 11637 + }, + { + "epoch": 0.9330741015413602, + "grad_norm": 1.81055748462677, + "learning_rate": 1.1562741553530999e-05, + "loss": 3.1978, + "step": 11638 + }, + { + "epoch": 0.9331542763223828, + "grad_norm": 1.9058271646499634, + "learning_rate": 1.156145904639124e-05, + "loss": 3.2607, + "step": 11639 + }, + { + "epoch": 0.9332344511034054, + "grad_norm": 1.9320528507232666, + "learning_rate": 1.1560176512925833e-05, + "loss": 3.1967, + "step": 11640 + }, + { + "epoch": 0.933314625884428, + "grad_norm": 1.8056602478027344, + "learning_rate": 1.1558893953156399e-05, + "loss": 3.1379, + "step": 11641 + }, + { + "epoch": 0.9333948006654507, + "grad_norm": 1.8794058561325073, + "learning_rate": 1.1557611367104561e-05, + "loss": 3.1893, + "step": 11642 + }, + { + "epoch": 0.9334749754464733, + "grad_norm": 1.8641432523727417, + "learning_rate": 1.1556328754791948e-05, + "loss": 3.1938, + "step": 11643 + }, + { + "epoch": 0.9335551502274959, + "grad_norm": 1.7620195150375366, + "learning_rate": 1.1555046116240179e-05, + "loss": 3.2003, + "step": 11644 + }, + { + "epoch": 0.9336353250085185, + "grad_norm": 1.632824420928955, + "learning_rate": 1.1553763451470883e-05, + "loss": 3.1666, + "step": 11645 + }, + { + "epoch": 0.9337154997895412, + "grad_norm": 1.8925400972366333, + "learning_rate": 1.1552480760505679e-05, + "loss": 3.2758, + "step": 11646 + }, + { + "epoch": 0.9337956745705638, + "grad_norm": 1.7485839128494263, + "learning_rate": 1.1551198043366196e-05, + "loss": 3.1818, + "step": 11647 + }, + { + "epoch": 0.9338758493515864, + "grad_norm": 1.8391839265823364, + "learning_rate": 1.1549915300074063e-05, + "loss": 3.1799, + "step": 11648 + }, + { + "epoch": 0.9339560241326091, + "grad_norm": 1.7456163167953491, + "learning_rate": 1.1548632530650906e-05, + "loss": 3.3123, + "step": 11649 + }, + { + "epoch": 0.9340361989136318, + "grad_norm": 0.7161794900894165, + "learning_rate": 1.1547349735118348e-05, + "loss": 2.4976, + "step": 11650 + }, + { + "epoch": 0.9341163736946544, + "grad_norm": 1.8848704099655151, + "learning_rate": 1.1546066913498018e-05, + "loss": 3.2861, + "step": 11651 + }, + { + "epoch": 0.934196548475677, + "grad_norm": 1.8147121667861938, + "learning_rate": 1.1544784065811546e-05, + "loss": 3.2053, + "step": 11652 + }, + { + "epoch": 0.9342767232566996, + "grad_norm": 1.7481764554977417, + "learning_rate": 1.154350119208056e-05, + "loss": 3.1625, + "step": 11653 + }, + { + "epoch": 0.9343568980377223, + "grad_norm": 1.6536587476730347, + "learning_rate": 1.1542218292326686e-05, + "loss": 3.1573, + "step": 11654 + }, + { + "epoch": 0.9344370728187449, + "grad_norm": 1.7523109912872314, + "learning_rate": 1.1540935366571557e-05, + "loss": 3.183, + "step": 11655 + }, + { + "epoch": 0.9345172475997675, + "grad_norm": 1.792255163192749, + "learning_rate": 1.15396524148368e-05, + "loss": 3.1645, + "step": 11656 + }, + { + "epoch": 0.9345974223807901, + "grad_norm": 1.8120530843734741, + "learning_rate": 1.1538369437144044e-05, + "loss": 3.3096, + "step": 11657 + }, + { + "epoch": 0.9346775971618128, + "grad_norm": 1.8615844249725342, + "learning_rate": 1.1537086433514923e-05, + "loss": 3.2881, + "step": 11658 + }, + { + "epoch": 0.9347577719428354, + "grad_norm": 1.7278993129730225, + "learning_rate": 1.1535803403971067e-05, + "loss": 3.1709, + "step": 11659 + }, + { + "epoch": 0.934837946723858, + "grad_norm": 1.7268868684768677, + "learning_rate": 1.1534520348534106e-05, + "loss": 3.2111, + "step": 11660 + }, + { + "epoch": 0.9349181215048806, + "grad_norm": 1.896903157234192, + "learning_rate": 1.1533237267225674e-05, + "loss": 3.2642, + "step": 11661 + }, + { + "epoch": 0.9349982962859033, + "grad_norm": 1.693868637084961, + "learning_rate": 1.15319541600674e-05, + "loss": 3.0745, + "step": 11662 + }, + { + "epoch": 0.9350784710669259, + "grad_norm": 1.8265273571014404, + "learning_rate": 1.153067102708092e-05, + "loss": 3.274, + "step": 11663 + }, + { + "epoch": 0.9351586458479485, + "grad_norm": 1.7079731225967407, + "learning_rate": 1.1529387868287865e-05, + "loss": 3.3259, + "step": 11664 + }, + { + "epoch": 0.9352388206289711, + "grad_norm": 1.805381178855896, + "learning_rate": 1.1528104683709874e-05, + "loss": 3.192, + "step": 11665 + }, + { + "epoch": 0.9353189954099937, + "grad_norm": 1.695341944694519, + "learning_rate": 1.1526821473368571e-05, + "loss": 3.1629, + "step": 11666 + }, + { + "epoch": 0.9353991701910164, + "grad_norm": 1.6440571546554565, + "learning_rate": 1.1525538237285596e-05, + "loss": 3.0603, + "step": 11667 + }, + { + "epoch": 0.935479344972039, + "grad_norm": 1.6955890655517578, + "learning_rate": 1.1524254975482587e-05, + "loss": 3.257, + "step": 11668 + }, + { + "epoch": 0.9355595197530616, + "grad_norm": 1.7409158945083618, + "learning_rate": 1.1522971687981175e-05, + "loss": 3.2161, + "step": 11669 + }, + { + "epoch": 0.9356396945340844, + "grad_norm": 1.7863327264785767, + "learning_rate": 1.1521688374803e-05, + "loss": 3.286, + "step": 11670 + }, + { + "epoch": 0.935719869315107, + "grad_norm": 1.7792975902557373, + "learning_rate": 1.152040503596969e-05, + "loss": 3.2548, + "step": 11671 + }, + { + "epoch": 0.9358000440961296, + "grad_norm": 1.832943320274353, + "learning_rate": 1.1519121671502891e-05, + "loss": 3.1295, + "step": 11672 + }, + { + "epoch": 0.9358802188771522, + "grad_norm": 1.8106309175491333, + "learning_rate": 1.1517838281424236e-05, + "loss": 3.2114, + "step": 11673 + }, + { + "epoch": 0.9359603936581748, + "grad_norm": 1.7979015111923218, + "learning_rate": 1.151655486575536e-05, + "loss": 3.1498, + "step": 11674 + }, + { + "epoch": 0.9360405684391975, + "grad_norm": 1.8716075420379639, + "learning_rate": 1.1515271424517906e-05, + "loss": 3.2275, + "step": 11675 + }, + { + "epoch": 0.9361207432202201, + "grad_norm": 1.7987381219863892, + "learning_rate": 1.1513987957733507e-05, + "loss": 3.1747, + "step": 11676 + }, + { + "epoch": 0.9362009180012427, + "grad_norm": 0.7030860185623169, + "learning_rate": 1.1512704465423805e-05, + "loss": 2.4814, + "step": 11677 + }, + { + "epoch": 0.9362810927822653, + "grad_norm": 0.7289131879806519, + "learning_rate": 1.1511420947610442e-05, + "loss": 2.5089, + "step": 11678 + }, + { + "epoch": 0.936361267563288, + "grad_norm": 1.8213319778442383, + "learning_rate": 1.1510137404315051e-05, + "loss": 3.2897, + "step": 11679 + }, + { + "epoch": 0.9364414423443106, + "grad_norm": 1.7999070882797241, + "learning_rate": 1.150885383555928e-05, + "loss": 3.1704, + "step": 11680 + }, + { + "epoch": 0.9365216171253332, + "grad_norm": 1.6611769199371338, + "learning_rate": 1.150757024136476e-05, + "loss": 3.1105, + "step": 11681 + }, + { + "epoch": 0.9366017919063558, + "grad_norm": 0.688940167427063, + "learning_rate": 1.1506286621753138e-05, + "loss": 2.4856, + "step": 11682 + }, + { + "epoch": 0.9366819666873785, + "grad_norm": 1.7439205646514893, + "learning_rate": 1.1505002976746056e-05, + "loss": 3.2683, + "step": 11683 + }, + { + "epoch": 0.9367621414684011, + "grad_norm": 1.8627493381500244, + "learning_rate": 1.1503719306365153e-05, + "loss": 3.1073, + "step": 11684 + }, + { + "epoch": 0.9368423162494237, + "grad_norm": 1.7409340143203735, + "learning_rate": 1.1502435610632077e-05, + "loss": 3.2198, + "step": 11685 + }, + { + "epoch": 0.9369224910304463, + "grad_norm": 1.7527579069137573, + "learning_rate": 1.150115188956846e-05, + "loss": 3.1702, + "step": 11686 + }, + { + "epoch": 0.937002665811469, + "grad_norm": 1.7168760299682617, + "learning_rate": 1.1499868143195953e-05, + "loss": 3.0736, + "step": 11687 + }, + { + "epoch": 0.9370828405924916, + "grad_norm": 1.7239177227020264, + "learning_rate": 1.14985843715362e-05, + "loss": 3.2529, + "step": 11688 + }, + { + "epoch": 0.9371630153735142, + "grad_norm": 1.6926171779632568, + "learning_rate": 1.1497300574610838e-05, + "loss": 3.137, + "step": 11689 + }, + { + "epoch": 0.9372431901545369, + "grad_norm": 1.8321205377578735, + "learning_rate": 1.149601675244152e-05, + "loss": 3.2172, + "step": 11690 + }, + { + "epoch": 0.9373233649355596, + "grad_norm": 1.805580973625183, + "learning_rate": 1.1494732905049886e-05, + "loss": 3.1198, + "step": 11691 + }, + { + "epoch": 0.9374035397165822, + "grad_norm": 1.7049040794372559, + "learning_rate": 1.1493449032457582e-05, + "loss": 3.1094, + "step": 11692 + }, + { + "epoch": 0.9374837144976048, + "grad_norm": 1.8341755867004395, + "learning_rate": 1.1492165134686256e-05, + "loss": 3.2316, + "step": 11693 + }, + { + "epoch": 0.9375638892786274, + "grad_norm": 1.7537963390350342, + "learning_rate": 1.1490881211757547e-05, + "loss": 3.1056, + "step": 11694 + }, + { + "epoch": 0.9376440640596501, + "grad_norm": 1.8110710382461548, + "learning_rate": 1.148959726369311e-05, + "loss": 3.203, + "step": 11695 + }, + { + "epoch": 0.9377242388406727, + "grad_norm": 1.8444952964782715, + "learning_rate": 1.1488313290514585e-05, + "loss": 3.2795, + "step": 11696 + }, + { + "epoch": 0.9378044136216953, + "grad_norm": 1.85777747631073, + "learning_rate": 1.1487029292243624e-05, + "loss": 3.2942, + "step": 11697 + }, + { + "epoch": 0.9378845884027179, + "grad_norm": 1.8671401739120483, + "learning_rate": 1.1485745268901874e-05, + "loss": 3.1451, + "step": 11698 + }, + { + "epoch": 0.9379647631837406, + "grad_norm": 1.9191014766693115, + "learning_rate": 1.148446122051098e-05, + "loss": 3.2975, + "step": 11699 + }, + { + "epoch": 0.9380449379647632, + "grad_norm": 1.7071869373321533, + "learning_rate": 1.1483177147092597e-05, + "loss": 3.0979, + "step": 11700 + }, + { + "epoch": 0.9381251127457858, + "grad_norm": 1.9654717445373535, + "learning_rate": 1.1481893048668367e-05, + "loss": 3.1804, + "step": 11701 + }, + { + "epoch": 0.9382052875268084, + "grad_norm": 1.7459365129470825, + "learning_rate": 1.1480608925259942e-05, + "loss": 3.1956, + "step": 11702 + }, + { + "epoch": 0.938285462307831, + "grad_norm": 1.761008620262146, + "learning_rate": 1.1479324776888974e-05, + "loss": 3.2122, + "step": 11703 + }, + { + "epoch": 0.9383656370888537, + "grad_norm": 1.8191657066345215, + "learning_rate": 1.147804060357711e-05, + "loss": 3.209, + "step": 11704 + }, + { + "epoch": 0.9384458118698763, + "grad_norm": 1.7658721208572388, + "learning_rate": 1.1476756405346004e-05, + "loss": 3.1282, + "step": 11705 + }, + { + "epoch": 0.9385259866508989, + "grad_norm": 1.7201533317565918, + "learning_rate": 1.1475472182217304e-05, + "loss": 3.2358, + "step": 11706 + }, + { + "epoch": 0.9386061614319215, + "grad_norm": 1.7791619300842285, + "learning_rate": 1.1474187934212664e-05, + "loss": 3.2062, + "step": 11707 + }, + { + "epoch": 0.9386863362129442, + "grad_norm": 1.769585371017456, + "learning_rate": 1.1472903661353737e-05, + "loss": 3.1284, + "step": 11708 + }, + { + "epoch": 0.9387665109939668, + "grad_norm": 1.6985630989074707, + "learning_rate": 1.1471619363662167e-05, + "loss": 3.1741, + "step": 11709 + }, + { + "epoch": 0.9388466857749895, + "grad_norm": 1.774154543876648, + "learning_rate": 1.147033504115962e-05, + "loss": 3.1014, + "step": 11710 + }, + { + "epoch": 0.9389268605560122, + "grad_norm": 1.7359040975570679, + "learning_rate": 1.1469050693867738e-05, + "loss": 3.2074, + "step": 11711 + }, + { + "epoch": 0.9390070353370348, + "grad_norm": 1.7955256700515747, + "learning_rate": 1.146776632180818e-05, + "loss": 3.064, + "step": 11712 + }, + { + "epoch": 0.9390872101180574, + "grad_norm": 1.856076955795288, + "learning_rate": 1.1466481925002602e-05, + "loss": 3.2582, + "step": 11713 + }, + { + "epoch": 0.93916738489908, + "grad_norm": 1.8525564670562744, + "learning_rate": 1.1465197503472654e-05, + "loss": 3.1737, + "step": 11714 + }, + { + "epoch": 0.9392475596801027, + "grad_norm": 1.7236554622650146, + "learning_rate": 1.1463913057239992e-05, + "loss": 3.2169, + "step": 11715 + }, + { + "epoch": 0.9393277344611253, + "grad_norm": 1.7898515462875366, + "learning_rate": 1.146262858632627e-05, + "loss": 3.1671, + "step": 11716 + }, + { + "epoch": 0.9394079092421479, + "grad_norm": 1.8464516401290894, + "learning_rate": 1.1461344090753148e-05, + "loss": 3.1931, + "step": 11717 + }, + { + "epoch": 0.9394880840231705, + "grad_norm": 1.7119364738464355, + "learning_rate": 1.146005957054228e-05, + "loss": 3.1661, + "step": 11718 + }, + { + "epoch": 0.9395682588041931, + "grad_norm": 1.708498239517212, + "learning_rate": 1.145877502571532e-05, + "loss": 3.1693, + "step": 11719 + }, + { + "epoch": 0.9396484335852158, + "grad_norm": 1.7288295030593872, + "learning_rate": 1.1457490456293929e-05, + "loss": 3.1157, + "step": 11720 + }, + { + "epoch": 0.9397286083662384, + "grad_norm": 1.7140307426452637, + "learning_rate": 1.1456205862299762e-05, + "loss": 3.1622, + "step": 11721 + }, + { + "epoch": 0.939808783147261, + "grad_norm": 1.8028950691223145, + "learning_rate": 1.1454921243754476e-05, + "loss": 3.2629, + "step": 11722 + }, + { + "epoch": 0.9398889579282836, + "grad_norm": 1.734307885169983, + "learning_rate": 1.1453636600679733e-05, + "loss": 3.1661, + "step": 11723 + }, + { + "epoch": 0.9399691327093063, + "grad_norm": 1.7916828393936157, + "learning_rate": 1.1452351933097188e-05, + "loss": 3.2637, + "step": 11724 + }, + { + "epoch": 0.9400493074903289, + "grad_norm": 1.633715033531189, + "learning_rate": 1.1451067241028501e-05, + "loss": 3.1594, + "step": 11725 + }, + { + "epoch": 0.9401294822713515, + "grad_norm": 1.8608452081680298, + "learning_rate": 1.1449782524495333e-05, + "loss": 3.2057, + "step": 11726 + }, + { + "epoch": 0.9402096570523741, + "grad_norm": 1.6963402032852173, + "learning_rate": 1.144849778351934e-05, + "loss": 3.2144, + "step": 11727 + }, + { + "epoch": 0.9402898318333968, + "grad_norm": 1.7426457405090332, + "learning_rate": 1.1447213018122187e-05, + "loss": 3.2631, + "step": 11728 + }, + { + "epoch": 0.9403700066144194, + "grad_norm": 1.809790015220642, + "learning_rate": 1.1445928228325532e-05, + "loss": 3.1782, + "step": 11729 + }, + { + "epoch": 0.940450181395442, + "grad_norm": 1.6875756978988647, + "learning_rate": 1.1444643414151037e-05, + "loss": 3.2339, + "step": 11730 + }, + { + "epoch": 0.9405303561764647, + "grad_norm": 1.7566770315170288, + "learning_rate": 1.1443358575620363e-05, + "loss": 3.2576, + "step": 11731 + }, + { + "epoch": 0.9406105309574874, + "grad_norm": 1.7217316627502441, + "learning_rate": 1.144207371275517e-05, + "loss": 3.1423, + "step": 11732 + }, + { + "epoch": 0.94069070573851, + "grad_norm": 1.7529246807098389, + "learning_rate": 1.1440788825577128e-05, + "loss": 3.2762, + "step": 11733 + }, + { + "epoch": 0.9407708805195326, + "grad_norm": 1.6624877452850342, + "learning_rate": 1.143950391410789e-05, + "loss": 3.2865, + "step": 11734 + }, + { + "epoch": 0.9408510553005552, + "grad_norm": 1.7029621601104736, + "learning_rate": 1.1438218978369125e-05, + "loss": 3.2739, + "step": 11735 + }, + { + "epoch": 0.9409312300815779, + "grad_norm": 1.7609963417053223, + "learning_rate": 1.1436934018382492e-05, + "loss": 3.1672, + "step": 11736 + }, + { + "epoch": 0.9410114048626005, + "grad_norm": 1.7935121059417725, + "learning_rate": 1.143564903416966e-05, + "loss": 3.2997, + "step": 11737 + }, + { + "epoch": 0.9410915796436231, + "grad_norm": 1.7244229316711426, + "learning_rate": 1.1434364025752292e-05, + "loss": 3.0729, + "step": 11738 + }, + { + "epoch": 0.9411717544246457, + "grad_norm": 1.7496521472930908, + "learning_rate": 1.1433078993152049e-05, + "loss": 3.1743, + "step": 11739 + }, + { + "epoch": 0.9412519292056684, + "grad_norm": 1.8406896591186523, + "learning_rate": 1.1431793936390602e-05, + "loss": 3.2298, + "step": 11740 + }, + { + "epoch": 0.941332103986691, + "grad_norm": 1.6674985885620117, + "learning_rate": 1.1430508855489614e-05, + "loss": 3.1561, + "step": 11741 + }, + { + "epoch": 0.9414122787677136, + "grad_norm": 1.6883479356765747, + "learning_rate": 1.1429223750470748e-05, + "loss": 3.2322, + "step": 11742 + }, + { + "epoch": 0.9414924535487362, + "grad_norm": 0.7337451577186584, + "learning_rate": 1.1427938621355676e-05, + "loss": 2.4888, + "step": 11743 + }, + { + "epoch": 0.9415726283297589, + "grad_norm": 1.8161516189575195, + "learning_rate": 1.1426653468166062e-05, + "loss": 3.2604, + "step": 11744 + }, + { + "epoch": 0.9416528031107815, + "grad_norm": 1.7597746849060059, + "learning_rate": 1.1425368290923572e-05, + "loss": 3.2211, + "step": 11745 + }, + { + "epoch": 0.9417329778918041, + "grad_norm": 1.8184679746627808, + "learning_rate": 1.1424083089649877e-05, + "loss": 3.1584, + "step": 11746 + }, + { + "epoch": 0.9418131526728267, + "grad_norm": 1.8020029067993164, + "learning_rate": 1.1422797864366642e-05, + "loss": 3.2291, + "step": 11747 + }, + { + "epoch": 0.9418933274538493, + "grad_norm": 1.805213212966919, + "learning_rate": 1.1421512615095536e-05, + "loss": 3.2265, + "step": 11748 + }, + { + "epoch": 0.941973502234872, + "grad_norm": 1.7923328876495361, + "learning_rate": 1.1420227341858229e-05, + "loss": 3.2507, + "step": 11749 + }, + { + "epoch": 0.9420536770158946, + "grad_norm": 1.764694094657898, + "learning_rate": 1.1418942044676389e-05, + "loss": 3.1638, + "step": 11750 + }, + { + "epoch": 0.9421338517969173, + "grad_norm": 1.8012810945510864, + "learning_rate": 1.1417656723571687e-05, + "loss": 3.2941, + "step": 11751 + }, + { + "epoch": 0.94221402657794, + "grad_norm": 1.772992730140686, + "learning_rate": 1.1416371378565789e-05, + "loss": 3.1094, + "step": 11752 + }, + { + "epoch": 0.9422942013589626, + "grad_norm": 1.7983372211456299, + "learning_rate": 1.1415086009680372e-05, + "loss": 3.3208, + "step": 11753 + }, + { + "epoch": 0.9423743761399852, + "grad_norm": 1.7780225276947021, + "learning_rate": 1.1413800616937103e-05, + "loss": 3.1985, + "step": 11754 + }, + { + "epoch": 0.9424545509210078, + "grad_norm": 1.7641942501068115, + "learning_rate": 1.1412515200357655e-05, + "loss": 3.2076, + "step": 11755 + }, + { + "epoch": 0.9425347257020305, + "grad_norm": 1.8392527103424072, + "learning_rate": 1.1411229759963696e-05, + "loss": 3.2925, + "step": 11756 + }, + { + "epoch": 0.9426149004830531, + "grad_norm": 1.7178783416748047, + "learning_rate": 1.14099442957769e-05, + "loss": 3.187, + "step": 11757 + }, + { + "epoch": 0.9426950752640757, + "grad_norm": 1.7650445699691772, + "learning_rate": 1.1408658807818944e-05, + "loss": 3.136, + "step": 11758 + }, + { + "epoch": 0.9427752500450983, + "grad_norm": 1.8151386976242065, + "learning_rate": 1.1407373296111495e-05, + "loss": 3.2249, + "step": 11759 + }, + { + "epoch": 0.942855424826121, + "grad_norm": 1.7972365617752075, + "learning_rate": 1.140608776067623e-05, + "loss": 3.2346, + "step": 11760 + }, + { + "epoch": 0.9429355996071436, + "grad_norm": 1.790120005607605, + "learning_rate": 1.1404802201534817e-05, + "loss": 3.165, + "step": 11761 + }, + { + "epoch": 0.9430157743881662, + "grad_norm": 1.7665003538131714, + "learning_rate": 1.1403516618708934e-05, + "loss": 3.1388, + "step": 11762 + }, + { + "epoch": 0.9430959491691888, + "grad_norm": 1.7787227630615234, + "learning_rate": 1.140223101222026e-05, + "loss": 3.1827, + "step": 11763 + }, + { + "epoch": 0.9431761239502114, + "grad_norm": 1.753642201423645, + "learning_rate": 1.1400945382090463e-05, + "loss": 3.1846, + "step": 11764 + }, + { + "epoch": 0.9432562987312341, + "grad_norm": 1.8053560256958008, + "learning_rate": 1.1399659728341222e-05, + "loss": 3.2494, + "step": 11765 + }, + { + "epoch": 0.9433364735122567, + "grad_norm": 1.9807138442993164, + "learning_rate": 1.1398374050994207e-05, + "loss": 3.2614, + "step": 11766 + }, + { + "epoch": 0.9434166482932793, + "grad_norm": 1.771647572517395, + "learning_rate": 1.1397088350071099e-05, + "loss": 3.1838, + "step": 11767 + }, + { + "epoch": 0.9434968230743019, + "grad_norm": 1.935170292854309, + "learning_rate": 1.1395802625593579e-05, + "loss": 3.3389, + "step": 11768 + }, + { + "epoch": 0.9435769978553246, + "grad_norm": 1.8646011352539062, + "learning_rate": 1.1394516877583314e-05, + "loss": 3.2229, + "step": 11769 + }, + { + "epoch": 0.9436571726363472, + "grad_norm": 1.7802929878234863, + "learning_rate": 1.1393231106061989e-05, + "loss": 3.1746, + "step": 11770 + }, + { + "epoch": 0.9437373474173699, + "grad_norm": 1.8025132417678833, + "learning_rate": 1.1391945311051276e-05, + "loss": 3.2439, + "step": 11771 + }, + { + "epoch": 0.9438175221983925, + "grad_norm": 1.8357149362564087, + "learning_rate": 1.1390659492572856e-05, + "loss": 3.1421, + "step": 11772 + }, + { + "epoch": 0.9438976969794152, + "grad_norm": 1.792451024055481, + "learning_rate": 1.138937365064841e-05, + "loss": 3.299, + "step": 11773 + }, + { + "epoch": 0.9439778717604378, + "grad_norm": 0.7864938974380493, + "learning_rate": 1.1388087785299609e-05, + "loss": 2.4761, + "step": 11774 + }, + { + "epoch": 0.9440580465414604, + "grad_norm": 1.7691960334777832, + "learning_rate": 1.138680189654814e-05, + "loss": 3.1914, + "step": 11775 + }, + { + "epoch": 0.944138221322483, + "grad_norm": 1.7517261505126953, + "learning_rate": 1.138551598441568e-05, + "loss": 3.1656, + "step": 11776 + }, + { + "epoch": 0.9442183961035057, + "grad_norm": 1.706384301185608, + "learning_rate": 1.1384230048923907e-05, + "loss": 3.1627, + "step": 11777 + }, + { + "epoch": 0.9442985708845283, + "grad_norm": 0.6940578818321228, + "learning_rate": 1.1382944090094505e-05, + "loss": 2.4805, + "step": 11778 + }, + { + "epoch": 0.9443787456655509, + "grad_norm": 1.7805185317993164, + "learning_rate": 1.1381658107949152e-05, + "loss": 3.1683, + "step": 11779 + }, + { + "epoch": 0.9444589204465735, + "grad_norm": 1.929057240486145, + "learning_rate": 1.1380372102509532e-05, + "loss": 3.2393, + "step": 11780 + }, + { + "epoch": 0.9445390952275962, + "grad_norm": 1.9719128608703613, + "learning_rate": 1.1379086073797323e-05, + "loss": 3.1632, + "step": 11781 + }, + { + "epoch": 0.9446192700086188, + "grad_norm": 1.7962489128112793, + "learning_rate": 1.1377800021834209e-05, + "loss": 3.2148, + "step": 11782 + }, + { + "epoch": 0.9446994447896414, + "grad_norm": 1.7086262702941895, + "learning_rate": 1.1376513946641875e-05, + "loss": 3.1471, + "step": 11783 + }, + { + "epoch": 0.944779619570664, + "grad_norm": 1.7628161907196045, + "learning_rate": 1.1375227848241997e-05, + "loss": 3.1442, + "step": 11784 + }, + { + "epoch": 0.9448597943516867, + "grad_norm": 1.6877044439315796, + "learning_rate": 1.1373941726656264e-05, + "loss": 3.0747, + "step": 11785 + }, + { + "epoch": 0.9449399691327093, + "grad_norm": 1.7011979818344116, + "learning_rate": 1.1372655581906358e-05, + "loss": 3.1987, + "step": 11786 + }, + { + "epoch": 0.9450201439137319, + "grad_norm": 1.5747402906417847, + "learning_rate": 1.1371369414013961e-05, + "loss": 3.1204, + "step": 11787 + }, + { + "epoch": 0.9451003186947545, + "grad_norm": 1.7930546998977661, + "learning_rate": 1.1370083223000763e-05, + "loss": 3.1455, + "step": 11788 + }, + { + "epoch": 0.9451804934757772, + "grad_norm": 1.818841814994812, + "learning_rate": 1.1368797008888439e-05, + "loss": 3.2048, + "step": 11789 + }, + { + "epoch": 0.9452606682567998, + "grad_norm": 1.861642837524414, + "learning_rate": 1.1367510771698682e-05, + "loss": 3.1652, + "step": 11790 + }, + { + "epoch": 0.9453408430378224, + "grad_norm": 1.814864158630371, + "learning_rate": 1.1366224511453178e-05, + "loss": 3.1756, + "step": 11791 + }, + { + "epoch": 0.9454210178188451, + "grad_norm": 1.9149024486541748, + "learning_rate": 1.1364938228173607e-05, + "loss": 3.2789, + "step": 11792 + }, + { + "epoch": 0.9455011925998678, + "grad_norm": 1.856696605682373, + "learning_rate": 1.1363651921881663e-05, + "loss": 3.1812, + "step": 11793 + }, + { + "epoch": 0.9455813673808904, + "grad_norm": 1.7383289337158203, + "learning_rate": 1.1362365592599023e-05, + "loss": 3.2235, + "step": 11794 + }, + { + "epoch": 0.945661542161913, + "grad_norm": 1.8263753652572632, + "learning_rate": 1.136107924034738e-05, + "loss": 3.179, + "step": 11795 + }, + { + "epoch": 0.9457417169429356, + "grad_norm": 1.7135019302368164, + "learning_rate": 1.1359792865148425e-05, + "loss": 3.1511, + "step": 11796 + }, + { + "epoch": 0.9458218917239583, + "grad_norm": 1.7518346309661865, + "learning_rate": 1.1358506467023838e-05, + "loss": 3.1348, + "step": 11797 + }, + { + "epoch": 0.9459020665049809, + "grad_norm": 1.687208890914917, + "learning_rate": 1.1357220045995312e-05, + "loss": 3.3014, + "step": 11798 + }, + { + "epoch": 0.9459822412860035, + "grad_norm": 1.7196002006530762, + "learning_rate": 1.1355933602084532e-05, + "loss": 3.1387, + "step": 11799 + }, + { + "epoch": 0.9460624160670261, + "grad_norm": 1.720665454864502, + "learning_rate": 1.135464713531319e-05, + "loss": 3.1157, + "step": 11800 + }, + { + "epoch": 0.9461425908480487, + "grad_norm": 1.752012014389038, + "learning_rate": 1.1353360645702978e-05, + "loss": 3.2974, + "step": 11801 + }, + { + "epoch": 0.9462227656290714, + "grad_norm": 1.7795212268829346, + "learning_rate": 1.1352074133275579e-05, + "loss": 3.262, + "step": 11802 + }, + { + "epoch": 0.946302940410094, + "grad_norm": 1.7521809339523315, + "learning_rate": 1.1350787598052689e-05, + "loss": 3.189, + "step": 11803 + }, + { + "epoch": 0.9463831151911166, + "grad_norm": 1.745632529258728, + "learning_rate": 1.1349501040055995e-05, + "loss": 3.1887, + "step": 11804 + }, + { + "epoch": 0.9464632899721392, + "grad_norm": 1.7758233547210693, + "learning_rate": 1.1348214459307188e-05, + "loss": 3.1843, + "step": 11805 + }, + { + "epoch": 0.9465434647531619, + "grad_norm": 1.922239899635315, + "learning_rate": 1.1346927855827963e-05, + "loss": 3.2056, + "step": 11806 + }, + { + "epoch": 0.9466236395341845, + "grad_norm": 1.710329294204712, + "learning_rate": 1.1345641229640006e-05, + "loss": 3.2056, + "step": 11807 + }, + { + "epoch": 0.9467038143152071, + "grad_norm": 1.7743637561798096, + "learning_rate": 1.1344354580765015e-05, + "loss": 3.2373, + "step": 11808 + }, + { + "epoch": 0.9467839890962297, + "grad_norm": 1.7788875102996826, + "learning_rate": 1.1343067909224676e-05, + "loss": 3.1868, + "step": 11809 + }, + { + "epoch": 0.9468641638772524, + "grad_norm": 1.766478180885315, + "learning_rate": 1.1341781215040688e-05, + "loss": 3.2017, + "step": 11810 + }, + { + "epoch": 0.946944338658275, + "grad_norm": 1.7327184677124023, + "learning_rate": 1.1340494498234742e-05, + "loss": 3.1339, + "step": 11811 + }, + { + "epoch": 0.9470245134392977, + "grad_norm": 1.838536024093628, + "learning_rate": 1.133920775882853e-05, + "loss": 3.2209, + "step": 11812 + }, + { + "epoch": 0.9471046882203203, + "grad_norm": 1.7430447340011597, + "learning_rate": 1.1337920996843748e-05, + "loss": 3.2035, + "step": 11813 + }, + { + "epoch": 0.947184863001343, + "grad_norm": 1.7423574924468994, + "learning_rate": 1.1336634212302088e-05, + "loss": 3.1228, + "step": 11814 + }, + { + "epoch": 0.9472650377823656, + "grad_norm": 1.7204844951629639, + "learning_rate": 1.1335347405225248e-05, + "loss": 3.1912, + "step": 11815 + }, + { + "epoch": 0.9473452125633882, + "grad_norm": 1.8793729543685913, + "learning_rate": 1.1334060575634922e-05, + "loss": 3.2385, + "step": 11816 + }, + { + "epoch": 0.9474253873444108, + "grad_norm": 1.9594062566757202, + "learning_rate": 1.1332773723552801e-05, + "loss": 3.1422, + "step": 11817 + }, + { + "epoch": 0.9475055621254335, + "grad_norm": 1.760186791419983, + "learning_rate": 1.133148684900059e-05, + "loss": 3.2102, + "step": 11818 + }, + { + "epoch": 0.9475857369064561, + "grad_norm": 1.856919765472412, + "learning_rate": 1.1330199951999976e-05, + "loss": 3.2145, + "step": 11819 + }, + { + "epoch": 0.9476659116874787, + "grad_norm": 1.764927864074707, + "learning_rate": 1.1328913032572659e-05, + "loss": 3.2648, + "step": 11820 + }, + { + "epoch": 0.9477460864685013, + "grad_norm": 1.799768090248108, + "learning_rate": 1.132762609074034e-05, + "loss": 3.1588, + "step": 11821 + }, + { + "epoch": 0.947826261249524, + "grad_norm": 1.7981998920440674, + "learning_rate": 1.1326339126524712e-05, + "loss": 3.0996, + "step": 11822 + }, + { + "epoch": 0.9479064360305466, + "grad_norm": 1.7857766151428223, + "learning_rate": 1.1325052139947474e-05, + "loss": 3.2022, + "step": 11823 + }, + { + "epoch": 0.9479866108115692, + "grad_norm": 1.7935775518417358, + "learning_rate": 1.1323765131030323e-05, + "loss": 3.2965, + "step": 11824 + }, + { + "epoch": 0.9480667855925918, + "grad_norm": 1.762718677520752, + "learning_rate": 1.1322478099794957e-05, + "loss": 3.153, + "step": 11825 + }, + { + "epoch": 0.9481469603736145, + "grad_norm": 1.746104121208191, + "learning_rate": 1.1321191046263081e-05, + "loss": 3.2311, + "step": 11826 + }, + { + "epoch": 0.9482271351546371, + "grad_norm": 1.8550548553466797, + "learning_rate": 1.1319903970456384e-05, + "loss": 3.1917, + "step": 11827 + }, + { + "epoch": 0.9483073099356597, + "grad_norm": 1.895452618598938, + "learning_rate": 1.1318616872396575e-05, + "loss": 3.2838, + "step": 11828 + }, + { + "epoch": 0.9483874847166823, + "grad_norm": 1.7902209758758545, + "learning_rate": 1.1317329752105348e-05, + "loss": 3.1273, + "step": 11829 + }, + { + "epoch": 0.948467659497705, + "grad_norm": 1.803135633468628, + "learning_rate": 1.1316042609604408e-05, + "loss": 3.2281, + "step": 11830 + }, + { + "epoch": 0.9485478342787276, + "grad_norm": 1.778801441192627, + "learning_rate": 1.1314755444915452e-05, + "loss": 3.252, + "step": 11831 + }, + { + "epoch": 0.9486280090597503, + "grad_norm": 1.9098914861679077, + "learning_rate": 1.1313468258060183e-05, + "loss": 3.1625, + "step": 11832 + }, + { + "epoch": 0.9487081838407729, + "grad_norm": 1.8066550493240356, + "learning_rate": 1.1312181049060302e-05, + "loss": 3.1415, + "step": 11833 + }, + { + "epoch": 0.9487883586217956, + "grad_norm": 1.7812343835830688, + "learning_rate": 1.1310893817937509e-05, + "loss": 3.1571, + "step": 11834 + }, + { + "epoch": 0.9488685334028182, + "grad_norm": 1.8150758743286133, + "learning_rate": 1.130960656471351e-05, + "loss": 3.2657, + "step": 11835 + }, + { + "epoch": 0.9489487081838408, + "grad_norm": 1.7416472434997559, + "learning_rate": 1.1308319289410008e-05, + "loss": 3.1154, + "step": 11836 + }, + { + "epoch": 0.9490288829648634, + "grad_norm": 1.811334252357483, + "learning_rate": 1.13070319920487e-05, + "loss": 3.2391, + "step": 11837 + }, + { + "epoch": 0.949109057745886, + "grad_norm": 2.006134033203125, + "learning_rate": 1.1305744672651297e-05, + "loss": 3.312, + "step": 11838 + }, + { + "epoch": 0.9491892325269087, + "grad_norm": 1.8181341886520386, + "learning_rate": 1.1304457331239495e-05, + "loss": 3.2423, + "step": 11839 + }, + { + "epoch": 0.9492694073079313, + "grad_norm": 1.733216404914856, + "learning_rate": 1.1303169967835003e-05, + "loss": 3.2413, + "step": 11840 + }, + { + "epoch": 0.9493495820889539, + "grad_norm": 0.7287856340408325, + "learning_rate": 1.130188258245953e-05, + "loss": 2.4863, + "step": 11841 + }, + { + "epoch": 0.9494297568699765, + "grad_norm": 1.9776911735534668, + "learning_rate": 1.1300595175134768e-05, + "loss": 3.1964, + "step": 11842 + }, + { + "epoch": 0.9495099316509992, + "grad_norm": 1.9425504207611084, + "learning_rate": 1.1299307745882436e-05, + "loss": 3.084, + "step": 11843 + }, + { + "epoch": 0.9495901064320218, + "grad_norm": 1.835341215133667, + "learning_rate": 1.1298020294724229e-05, + "loss": 3.1854, + "step": 11844 + }, + { + "epoch": 0.9496702812130444, + "grad_norm": 1.7643605470657349, + "learning_rate": 1.1296732821681857e-05, + "loss": 3.187, + "step": 11845 + }, + { + "epoch": 0.949750455994067, + "grad_norm": 1.7368043661117554, + "learning_rate": 1.129544532677703e-05, + "loss": 3.1249, + "step": 11846 + }, + { + "epoch": 0.9498306307750897, + "grad_norm": 1.7803272008895874, + "learning_rate": 1.1294157810031448e-05, + "loss": 3.1693, + "step": 11847 + }, + { + "epoch": 0.9499108055561123, + "grad_norm": 1.7496428489685059, + "learning_rate": 1.1292870271466825e-05, + "loss": 3.2053, + "step": 11848 + }, + { + "epoch": 0.9499909803371349, + "grad_norm": 1.7674658298492432, + "learning_rate": 1.1291582711104862e-05, + "loss": 3.3027, + "step": 11849 + }, + { + "epoch": 0.9500711551181575, + "grad_norm": 1.7045708894729614, + "learning_rate": 1.129029512896727e-05, + "loss": 3.1967, + "step": 11850 + }, + { + "epoch": 0.9501513298991802, + "grad_norm": 1.8085249662399292, + "learning_rate": 1.1289007525075757e-05, + "loss": 3.2819, + "step": 11851 + }, + { + "epoch": 0.9502315046802028, + "grad_norm": 1.706737995147705, + "learning_rate": 1.1287719899452032e-05, + "loss": 3.1738, + "step": 11852 + }, + { + "epoch": 0.9503116794612255, + "grad_norm": 1.7199143171310425, + "learning_rate": 1.1286432252117805e-05, + "loss": 3.2422, + "step": 11853 + }, + { + "epoch": 0.9503918542422481, + "grad_norm": 1.7767030000686646, + "learning_rate": 1.128514458309478e-05, + "loss": 3.2338, + "step": 11854 + }, + { + "epoch": 0.9504720290232708, + "grad_norm": 1.8012428283691406, + "learning_rate": 1.1283856892404672e-05, + "loss": 3.2173, + "step": 11855 + }, + { + "epoch": 0.9505522038042934, + "grad_norm": 1.7234336137771606, + "learning_rate": 1.1282569180069189e-05, + "loss": 3.2109, + "step": 11856 + }, + { + "epoch": 0.950632378585316, + "grad_norm": 1.7059242725372314, + "learning_rate": 1.1281281446110042e-05, + "loss": 3.1896, + "step": 11857 + }, + { + "epoch": 0.9507125533663386, + "grad_norm": 1.9166260957717896, + "learning_rate": 1.1279993690548943e-05, + "loss": 3.0724, + "step": 11858 + }, + { + "epoch": 0.9507927281473613, + "grad_norm": 1.7371970415115356, + "learning_rate": 1.1278705913407599e-05, + "loss": 3.2103, + "step": 11859 + }, + { + "epoch": 0.9508729029283839, + "grad_norm": 1.8107831478118896, + "learning_rate": 1.1277418114707723e-05, + "loss": 3.1882, + "step": 11860 + }, + { + "epoch": 0.9509530777094065, + "grad_norm": 1.797429084777832, + "learning_rate": 1.1276130294471032e-05, + "loss": 3.2135, + "step": 11861 + }, + { + "epoch": 0.9510332524904291, + "grad_norm": 1.7743011713027954, + "learning_rate": 1.127484245271923e-05, + "loss": 3.1974, + "step": 11862 + }, + { + "epoch": 0.9511134272714518, + "grad_norm": 1.662779450416565, + "learning_rate": 1.1273554589474039e-05, + "loss": 3.2125, + "step": 11863 + }, + { + "epoch": 0.9511936020524744, + "grad_norm": 1.6993423700332642, + "learning_rate": 1.127226670475716e-05, + "loss": 3.2072, + "step": 11864 + }, + { + "epoch": 0.951273776833497, + "grad_norm": 1.7948582172393799, + "learning_rate": 1.1270978798590316e-05, + "loss": 3.238, + "step": 11865 + }, + { + "epoch": 0.9513539516145196, + "grad_norm": 1.7503224611282349, + "learning_rate": 1.1269690870995215e-05, + "loss": 3.1541, + "step": 11866 + }, + { + "epoch": 0.9514341263955423, + "grad_norm": 1.733971118927002, + "learning_rate": 1.1268402921993575e-05, + "loss": 3.1906, + "step": 11867 + }, + { + "epoch": 0.9515143011765649, + "grad_norm": 0.7436437010765076, + "learning_rate": 1.1267114951607112e-05, + "loss": 2.4669, + "step": 11868 + }, + { + "epoch": 0.9515944759575875, + "grad_norm": 1.7459245920181274, + "learning_rate": 1.1265826959857533e-05, + "loss": 3.1816, + "step": 11869 + }, + { + "epoch": 0.9516746507386101, + "grad_norm": 1.8002060651779175, + "learning_rate": 1.1264538946766556e-05, + "loss": 3.2183, + "step": 11870 + }, + { + "epoch": 0.9517548255196328, + "grad_norm": 1.7486735582351685, + "learning_rate": 1.1263250912355904e-05, + "loss": 3.1873, + "step": 11871 + }, + { + "epoch": 0.9518350003006554, + "grad_norm": 1.7171630859375, + "learning_rate": 1.1261962856647284e-05, + "loss": 3.1957, + "step": 11872 + }, + { + "epoch": 0.9519151750816781, + "grad_norm": 1.6736559867858887, + "learning_rate": 1.1260674779662414e-05, + "loss": 3.1591, + "step": 11873 + }, + { + "epoch": 0.9519953498627007, + "grad_norm": 1.7434720993041992, + "learning_rate": 1.1259386681423013e-05, + "loss": 3.1842, + "step": 11874 + }, + { + "epoch": 0.9520755246437234, + "grad_norm": 1.90286123752594, + "learning_rate": 1.1258098561950794e-05, + "loss": 3.3365, + "step": 11875 + }, + { + "epoch": 0.952155699424746, + "grad_norm": 1.8443838357925415, + "learning_rate": 1.1256810421267482e-05, + "loss": 3.1409, + "step": 11876 + }, + { + "epoch": 0.9522358742057686, + "grad_norm": 1.8443481922149658, + "learning_rate": 1.1255522259394781e-05, + "loss": 3.1654, + "step": 11877 + }, + { + "epoch": 0.9523160489867912, + "grad_norm": 1.7692983150482178, + "learning_rate": 1.1254234076354424e-05, + "loss": 3.1605, + "step": 11878 + }, + { + "epoch": 0.9523962237678139, + "grad_norm": 1.907503604888916, + "learning_rate": 1.125294587216812e-05, + "loss": 3.287, + "step": 11879 + }, + { + "epoch": 0.9524763985488365, + "grad_norm": 1.5886727571487427, + "learning_rate": 1.125165764685759e-05, + "loss": 3.2742, + "step": 11880 + }, + { + "epoch": 0.9525565733298591, + "grad_norm": 0.7210434675216675, + "learning_rate": 1.1250369400444557e-05, + "loss": 2.4767, + "step": 11881 + }, + { + "epoch": 0.9526367481108817, + "grad_norm": 1.9169797897338867, + "learning_rate": 1.124908113295073e-05, + "loss": 3.2061, + "step": 11882 + }, + { + "epoch": 0.9527169228919043, + "grad_norm": 1.8105663061141968, + "learning_rate": 1.124779284439784e-05, + "loss": 3.1768, + "step": 11883 + }, + { + "epoch": 0.952797097672927, + "grad_norm": 1.6741909980773926, + "learning_rate": 1.1246504534807601e-05, + "loss": 3.1617, + "step": 11884 + }, + { + "epoch": 0.9528772724539496, + "grad_norm": 1.9076589345932007, + "learning_rate": 1.1245216204201735e-05, + "loss": 3.1775, + "step": 11885 + }, + { + "epoch": 0.9529574472349722, + "grad_norm": 1.8279788494110107, + "learning_rate": 1.1243927852601963e-05, + "loss": 3.1294, + "step": 11886 + }, + { + "epoch": 0.9530376220159948, + "grad_norm": 1.7243820428848267, + "learning_rate": 1.1242639480030005e-05, + "loss": 3.1415, + "step": 11887 + }, + { + "epoch": 0.9531177967970175, + "grad_norm": 1.7402998208999634, + "learning_rate": 1.1241351086507583e-05, + "loss": 3.1077, + "step": 11888 + }, + { + "epoch": 0.9531979715780401, + "grad_norm": 1.8814990520477295, + "learning_rate": 1.124006267205642e-05, + "loss": 3.2036, + "step": 11889 + }, + { + "epoch": 0.9532781463590627, + "grad_norm": 1.7498903274536133, + "learning_rate": 1.1238774236698236e-05, + "loss": 3.1567, + "step": 11890 + }, + { + "epoch": 0.9533583211400853, + "grad_norm": 1.7527159452438354, + "learning_rate": 1.1237485780454758e-05, + "loss": 3.2434, + "step": 11891 + }, + { + "epoch": 0.953438495921108, + "grad_norm": 1.7462306022644043, + "learning_rate": 1.1236197303347702e-05, + "loss": 3.23, + "step": 11892 + }, + { + "epoch": 0.9535186707021307, + "grad_norm": 1.7311598062515259, + "learning_rate": 1.12349088053988e-05, + "loss": 3.1685, + "step": 11893 + }, + { + "epoch": 0.9535988454831533, + "grad_norm": 1.8148865699768066, + "learning_rate": 1.1233620286629768e-05, + "loss": 3.1852, + "step": 11894 + }, + { + "epoch": 0.953679020264176, + "grad_norm": 1.7392280101776123, + "learning_rate": 1.123233174706233e-05, + "loss": 3.1779, + "step": 11895 + }, + { + "epoch": 0.9537591950451986, + "grad_norm": 1.7550756931304932, + "learning_rate": 1.1231043186718217e-05, + "loss": 3.2243, + "step": 11896 + }, + { + "epoch": 0.9538393698262212, + "grad_norm": 1.7999194860458374, + "learning_rate": 1.1229754605619148e-05, + "loss": 3.3337, + "step": 11897 + }, + { + "epoch": 0.9539195446072438, + "grad_norm": 1.7374720573425293, + "learning_rate": 1.122846600378685e-05, + "loss": 3.2536, + "step": 11898 + }, + { + "epoch": 0.9539997193882664, + "grad_norm": 1.796999216079712, + "learning_rate": 1.1227177381243048e-05, + "loss": 3.3029, + "step": 11899 + }, + { + "epoch": 0.9540798941692891, + "grad_norm": 1.970456600189209, + "learning_rate": 1.1225888738009469e-05, + "loss": 3.1321, + "step": 11900 + }, + { + "epoch": 0.9541600689503117, + "grad_norm": 1.8521889448165894, + "learning_rate": 1.1224600074107838e-05, + "loss": 3.2139, + "step": 11901 + }, + { + "epoch": 0.9542402437313343, + "grad_norm": 1.7303074598312378, + "learning_rate": 1.1223311389559878e-05, + "loss": 3.1786, + "step": 11902 + }, + { + "epoch": 0.9543204185123569, + "grad_norm": 1.8072168827056885, + "learning_rate": 1.1222022684387324e-05, + "loss": 3.0536, + "step": 11903 + }, + { + "epoch": 0.9544005932933796, + "grad_norm": 1.9025208950042725, + "learning_rate": 1.1220733958611895e-05, + "loss": 3.2808, + "step": 11904 + }, + { + "epoch": 0.9544807680744022, + "grad_norm": 1.7063006162643433, + "learning_rate": 1.1219445212255323e-05, + "loss": 3.2186, + "step": 11905 + }, + { + "epoch": 0.9545609428554248, + "grad_norm": 1.8427996635437012, + "learning_rate": 1.1218156445339334e-05, + "loss": 3.2185, + "step": 11906 + }, + { + "epoch": 0.9546411176364474, + "grad_norm": 1.6727348566055298, + "learning_rate": 1.1216867657885658e-05, + "loss": 3.0686, + "step": 11907 + }, + { + "epoch": 0.9547212924174701, + "grad_norm": 1.7754462957382202, + "learning_rate": 1.121557884991602e-05, + "loss": 3.1804, + "step": 11908 + }, + { + "epoch": 0.9548014671984927, + "grad_norm": 1.9170341491699219, + "learning_rate": 1.1214290021452153e-05, + "loss": 3.1467, + "step": 11909 + }, + { + "epoch": 0.9548816419795153, + "grad_norm": 1.7717152833938599, + "learning_rate": 1.121300117251578e-05, + "loss": 3.2279, + "step": 11910 + }, + { + "epoch": 0.9549618167605379, + "grad_norm": 1.782990574836731, + "learning_rate": 1.1211712303128639e-05, + "loss": 3.2781, + "step": 11911 + }, + { + "epoch": 0.9550419915415606, + "grad_norm": 1.9230009317398071, + "learning_rate": 1.1210423413312454e-05, + "loss": 3.1268, + "step": 11912 + }, + { + "epoch": 0.9551221663225832, + "grad_norm": 1.8122833967208862, + "learning_rate": 1.1209134503088958e-05, + "loss": 3.1641, + "step": 11913 + }, + { + "epoch": 0.9552023411036059, + "grad_norm": 1.7413579225540161, + "learning_rate": 1.1207845572479878e-05, + "loss": 3.2039, + "step": 11914 + }, + { + "epoch": 0.9552825158846285, + "grad_norm": 1.8063710927963257, + "learning_rate": 1.120655662150695e-05, + "loss": 3.0803, + "step": 11915 + }, + { + "epoch": 0.9553626906656512, + "grad_norm": 1.7844302654266357, + "learning_rate": 1.1205267650191902e-05, + "loss": 3.1792, + "step": 11916 + }, + { + "epoch": 0.9554428654466738, + "grad_norm": 1.7309337854385376, + "learning_rate": 1.1203978658556464e-05, + "loss": 3.1707, + "step": 11917 + }, + { + "epoch": 0.9555230402276964, + "grad_norm": 1.7999447584152222, + "learning_rate": 1.1202689646622375e-05, + "loss": 3.1488, + "step": 11918 + }, + { + "epoch": 0.955603215008719, + "grad_norm": 1.8651633262634277, + "learning_rate": 1.1201400614411358e-05, + "loss": 3.1648, + "step": 11919 + }, + { + "epoch": 0.9556833897897417, + "grad_norm": 1.761123776435852, + "learning_rate": 1.120011156194515e-05, + "loss": 3.1122, + "step": 11920 + }, + { + "epoch": 0.9557635645707643, + "grad_norm": 1.8000158071517944, + "learning_rate": 1.1198822489245487e-05, + "loss": 3.3154, + "step": 11921 + }, + { + "epoch": 0.9558437393517869, + "grad_norm": 1.783551573753357, + "learning_rate": 1.1197533396334096e-05, + "loss": 3.1691, + "step": 11922 + }, + { + "epoch": 0.9559239141328095, + "grad_norm": 1.7868499755859375, + "learning_rate": 1.1196244283232718e-05, + "loss": 3.2573, + "step": 11923 + }, + { + "epoch": 0.9560040889138322, + "grad_norm": 1.7325748205184937, + "learning_rate": 1.1194955149963081e-05, + "loss": 3.1974, + "step": 11924 + }, + { + "epoch": 0.9560842636948548, + "grad_norm": 1.6593888998031616, + "learning_rate": 1.1193665996546919e-05, + "loss": 3.1686, + "step": 11925 + }, + { + "epoch": 0.9561644384758774, + "grad_norm": 1.736779808998108, + "learning_rate": 1.1192376823005973e-05, + "loss": 3.2521, + "step": 11926 + }, + { + "epoch": 0.9562446132569, + "grad_norm": 1.972767949104309, + "learning_rate": 1.119108762936197e-05, + "loss": 3.2449, + "step": 11927 + }, + { + "epoch": 0.9563247880379226, + "grad_norm": 1.7403672933578491, + "learning_rate": 1.1189798415636652e-05, + "loss": 3.1531, + "step": 11928 + }, + { + "epoch": 0.9564049628189453, + "grad_norm": 1.783225417137146, + "learning_rate": 1.1188509181851752e-05, + "loss": 3.1627, + "step": 11929 + }, + { + "epoch": 0.9564851375999679, + "grad_norm": 1.8687299489974976, + "learning_rate": 1.1187219928029005e-05, + "loss": 3.262, + "step": 11930 + }, + { + "epoch": 0.9565653123809905, + "grad_norm": 1.8142147064208984, + "learning_rate": 1.118593065419015e-05, + "loss": 3.1172, + "step": 11931 + }, + { + "epoch": 0.9566454871620131, + "grad_norm": 0.7302458882331848, + "learning_rate": 1.118464136035692e-05, + "loss": 2.4786, + "step": 11932 + }, + { + "epoch": 0.9567256619430358, + "grad_norm": 1.7892093658447266, + "learning_rate": 1.1183352046551056e-05, + "loss": 3.2003, + "step": 11933 + }, + { + "epoch": 0.9568058367240585, + "grad_norm": 1.698355793952942, + "learning_rate": 1.1182062712794292e-05, + "loss": 3.209, + "step": 11934 + }, + { + "epoch": 0.9568860115050811, + "grad_norm": 1.7523741722106934, + "learning_rate": 1.1180773359108367e-05, + "loss": 3.1703, + "step": 11935 + }, + { + "epoch": 0.9569661862861037, + "grad_norm": 1.8056291341781616, + "learning_rate": 1.1179483985515023e-05, + "loss": 3.1578, + "step": 11936 + }, + { + "epoch": 0.9570463610671264, + "grad_norm": 1.8174399137496948, + "learning_rate": 1.1178194592035989e-05, + "loss": 3.244, + "step": 11937 + }, + { + "epoch": 0.957126535848149, + "grad_norm": 1.7360038757324219, + "learning_rate": 1.1176905178693013e-05, + "loss": 3.1746, + "step": 11938 + }, + { + "epoch": 0.9572067106291716, + "grad_norm": 1.7269960641860962, + "learning_rate": 1.117561574550783e-05, + "loss": 3.1309, + "step": 11939 + }, + { + "epoch": 0.9572868854101942, + "grad_norm": 1.8301924467086792, + "learning_rate": 1.117432629250218e-05, + "loss": 3.1357, + "step": 11940 + }, + { + "epoch": 0.9573670601912169, + "grad_norm": 1.7354825735092163, + "learning_rate": 1.1173036819697803e-05, + "loss": 3.113, + "step": 11941 + }, + { + "epoch": 0.9574472349722395, + "grad_norm": 1.8971461057662964, + "learning_rate": 1.1171747327116439e-05, + "loss": 3.2424, + "step": 11942 + }, + { + "epoch": 0.9575274097532621, + "grad_norm": 1.7534300088882446, + "learning_rate": 1.1170457814779826e-05, + "loss": 3.0516, + "step": 11943 + }, + { + "epoch": 0.9576075845342847, + "grad_norm": 1.8278529644012451, + "learning_rate": 1.1169168282709708e-05, + "loss": 3.2772, + "step": 11944 + }, + { + "epoch": 0.9576877593153074, + "grad_norm": 1.7059024572372437, + "learning_rate": 1.1167878730927824e-05, + "loss": 3.175, + "step": 11945 + }, + { + "epoch": 0.95776793409633, + "grad_norm": 1.7255305051803589, + "learning_rate": 1.1166589159455916e-05, + "loss": 3.1447, + "step": 11946 + }, + { + "epoch": 0.9578481088773526, + "grad_norm": 1.7256931066513062, + "learning_rate": 1.1165299568315726e-05, + "loss": 3.1366, + "step": 11947 + }, + { + "epoch": 0.9579282836583752, + "grad_norm": 1.6927791833877563, + "learning_rate": 1.1164009957528998e-05, + "loss": 3.2036, + "step": 11948 + }, + { + "epoch": 0.9580084584393979, + "grad_norm": 1.8298265933990479, + "learning_rate": 1.116272032711747e-05, + "loss": 3.2787, + "step": 11949 + }, + { + "epoch": 0.9580886332204205, + "grad_norm": 1.7931911945343018, + "learning_rate": 1.116143067710289e-05, + "loss": 3.2231, + "step": 11950 + }, + { + "epoch": 0.9581688080014431, + "grad_norm": 1.8547405004501343, + "learning_rate": 1.1160141007506994e-05, + "loss": 3.1929, + "step": 11951 + }, + { + "epoch": 0.9582489827824657, + "grad_norm": 1.8257063627243042, + "learning_rate": 1.1158851318351533e-05, + "loss": 3.2828, + "step": 11952 + }, + { + "epoch": 0.9583291575634884, + "grad_norm": 1.6936668157577515, + "learning_rate": 1.1157561609658247e-05, + "loss": 3.1553, + "step": 11953 + }, + { + "epoch": 0.9584093323445111, + "grad_norm": 1.7699849605560303, + "learning_rate": 1.1156271881448878e-05, + "loss": 3.2225, + "step": 11954 + }, + { + "epoch": 0.9584895071255337, + "grad_norm": 1.7431211471557617, + "learning_rate": 1.1154982133745174e-05, + "loss": 3.1868, + "step": 11955 + }, + { + "epoch": 0.9585696819065563, + "grad_norm": 1.9193968772888184, + "learning_rate": 1.1153692366568878e-05, + "loss": 3.2496, + "step": 11956 + }, + { + "epoch": 0.958649856687579, + "grad_norm": 1.823581576347351, + "learning_rate": 1.1152402579941735e-05, + "loss": 3.1494, + "step": 11957 + }, + { + "epoch": 0.9587300314686016, + "grad_norm": 1.6835778951644897, + "learning_rate": 1.1151112773885491e-05, + "loss": 3.1798, + "step": 11958 + }, + { + "epoch": 0.9588102062496242, + "grad_norm": 1.8581818342208862, + "learning_rate": 1.1149822948421891e-05, + "loss": 3.2264, + "step": 11959 + }, + { + "epoch": 0.9588903810306468, + "grad_norm": 1.7627184391021729, + "learning_rate": 1.114853310357268e-05, + "loss": 3.1818, + "step": 11960 + }, + { + "epoch": 0.9589705558116695, + "grad_norm": 1.810839295387268, + "learning_rate": 1.1147243239359608e-05, + "loss": 3.1684, + "step": 11961 + }, + { + "epoch": 0.9590507305926921, + "grad_norm": 1.7330787181854248, + "learning_rate": 1.1145953355804417e-05, + "loss": 3.2443, + "step": 11962 + }, + { + "epoch": 0.9591309053737147, + "grad_norm": 1.8008450269699097, + "learning_rate": 1.1144663452928862e-05, + "loss": 3.1652, + "step": 11963 + }, + { + "epoch": 0.9592110801547373, + "grad_norm": 1.7148473262786865, + "learning_rate": 1.1143373530754677e-05, + "loss": 3.2357, + "step": 11964 + }, + { + "epoch": 0.95929125493576, + "grad_norm": 0.7092136740684509, + "learning_rate": 1.114208358930362e-05, + "loss": 2.4762, + "step": 11965 + }, + { + "epoch": 0.9593714297167826, + "grad_norm": 1.9020425081253052, + "learning_rate": 1.1140793628597439e-05, + "loss": 3.2383, + "step": 11966 + }, + { + "epoch": 0.9594516044978052, + "grad_norm": 1.9105775356292725, + "learning_rate": 1.1139503648657877e-05, + "loss": 3.1735, + "step": 11967 + }, + { + "epoch": 0.9595317792788278, + "grad_norm": 1.8153777122497559, + "learning_rate": 1.1138213649506687e-05, + "loss": 3.2743, + "step": 11968 + }, + { + "epoch": 0.9596119540598504, + "grad_norm": 1.6721833944320679, + "learning_rate": 1.1136923631165613e-05, + "loss": 3.213, + "step": 11969 + }, + { + "epoch": 0.9596921288408731, + "grad_norm": 2.0366060733795166, + "learning_rate": 1.1135633593656408e-05, + "loss": 3.217, + "step": 11970 + }, + { + "epoch": 0.9597723036218957, + "grad_norm": 1.72697114944458, + "learning_rate": 1.1134343537000822e-05, + "loss": 3.2192, + "step": 11971 + }, + { + "epoch": 0.9598524784029183, + "grad_norm": 0.700076699256897, + "learning_rate": 1.1133053461220602e-05, + "loss": 2.4775, + "step": 11972 + }, + { + "epoch": 0.9599326531839409, + "grad_norm": 1.6623927354812622, + "learning_rate": 1.1131763366337502e-05, + "loss": 3.1964, + "step": 11973 + }, + { + "epoch": 0.9600128279649636, + "grad_norm": 1.82235848903656, + "learning_rate": 1.1130473252373271e-05, + "loss": 3.2468, + "step": 11974 + }, + { + "epoch": 0.9600930027459863, + "grad_norm": 1.6972005367279053, + "learning_rate": 1.1129183119349656e-05, + "loss": 3.1868, + "step": 11975 + }, + { + "epoch": 0.9601731775270089, + "grad_norm": 1.790350079536438, + "learning_rate": 1.1127892967288416e-05, + "loss": 3.2253, + "step": 11976 + }, + { + "epoch": 0.9602533523080315, + "grad_norm": 1.7608057260513306, + "learning_rate": 1.1126602796211294e-05, + "loss": 3.2174, + "step": 11977 + }, + { + "epoch": 0.9603335270890542, + "grad_norm": 1.728015661239624, + "learning_rate": 1.1125312606140048e-05, + "loss": 3.0663, + "step": 11978 + }, + { + "epoch": 0.9604137018700768, + "grad_norm": 1.7482151985168457, + "learning_rate": 1.1124022397096427e-05, + "loss": 3.1882, + "step": 11979 + }, + { + "epoch": 0.9604938766510994, + "grad_norm": 1.7296905517578125, + "learning_rate": 1.1122732169102186e-05, + "loss": 3.2289, + "step": 11980 + }, + { + "epoch": 0.960574051432122, + "grad_norm": 1.759853720664978, + "learning_rate": 1.1121441922179076e-05, + "loss": 3.2485, + "step": 11981 + }, + { + "epoch": 0.9606542262131447, + "grad_norm": 1.6914832592010498, + "learning_rate": 1.112015165634885e-05, + "loss": 3.1102, + "step": 11982 + }, + { + "epoch": 0.9607344009941673, + "grad_norm": 1.9140480756759644, + "learning_rate": 1.111886137163326e-05, + "loss": 3.2313, + "step": 11983 + }, + { + "epoch": 0.9608145757751899, + "grad_norm": 1.7964262962341309, + "learning_rate": 1.1117571068054067e-05, + "loss": 3.248, + "step": 11984 + }, + { + "epoch": 0.9608947505562125, + "grad_norm": 1.7379804849624634, + "learning_rate": 1.1116280745633015e-05, + "loss": 3.1418, + "step": 11985 + }, + { + "epoch": 0.9609749253372352, + "grad_norm": 1.7822659015655518, + "learning_rate": 1.1114990404391866e-05, + "loss": 3.1689, + "step": 11986 + }, + { + "epoch": 0.9610551001182578, + "grad_norm": 1.7958673238754272, + "learning_rate": 1.1113700044352366e-05, + "loss": 3.2096, + "step": 11987 + }, + { + "epoch": 0.9611352748992804, + "grad_norm": 1.7745084762573242, + "learning_rate": 1.111240966553628e-05, + "loss": 3.2428, + "step": 11988 + }, + { + "epoch": 0.961215449680303, + "grad_norm": 1.8373045921325684, + "learning_rate": 1.111111926796536e-05, + "loss": 3.2175, + "step": 11989 + }, + { + "epoch": 0.9612956244613257, + "grad_norm": 0.7209222316741943, + "learning_rate": 1.1109828851661359e-05, + "loss": 2.4724, + "step": 11990 + }, + { + "epoch": 0.9613757992423483, + "grad_norm": 0.686406135559082, + "learning_rate": 1.1108538416646038e-05, + "loss": 2.4745, + "step": 11991 + }, + { + "epoch": 0.9614559740233709, + "grad_norm": 1.8074431419372559, + "learning_rate": 1.1107247962941145e-05, + "loss": 3.2314, + "step": 11992 + }, + { + "epoch": 0.9615361488043935, + "grad_norm": 1.8770076036453247, + "learning_rate": 1.1105957490568444e-05, + "loss": 3.162, + "step": 11993 + }, + { + "epoch": 0.9616163235854162, + "grad_norm": 0.7081609964370728, + "learning_rate": 1.1104666999549689e-05, + "loss": 2.4806, + "step": 11994 + }, + { + "epoch": 0.9616964983664389, + "grad_norm": 1.7046561241149902, + "learning_rate": 1.1103376489906639e-05, + "loss": 3.2086, + "step": 11995 + }, + { + "epoch": 0.9617766731474615, + "grad_norm": 1.705203652381897, + "learning_rate": 1.110208596166105e-05, + "loss": 3.1182, + "step": 11996 + }, + { + "epoch": 0.9618568479284841, + "grad_norm": 1.7968045473098755, + "learning_rate": 1.1100795414834677e-05, + "loss": 3.1729, + "step": 11997 + }, + { + "epoch": 0.9619370227095068, + "grad_norm": 1.6611201763153076, + "learning_rate": 1.1099504849449284e-05, + "loss": 3.2004, + "step": 11998 + }, + { + "epoch": 0.9620171974905294, + "grad_norm": 1.7707406282424927, + "learning_rate": 1.1098214265526628e-05, + "loss": 3.2944, + "step": 11999 + }, + { + "epoch": 0.962097372271552, + "grad_norm": 1.8124183416366577, + "learning_rate": 1.1096923663088464e-05, + "loss": 3.2808, + "step": 12000 + }, + { + "epoch": 0.9621775470525746, + "grad_norm": 1.7057002782821655, + "learning_rate": 1.1095633042156555e-05, + "loss": 3.2154, + "step": 12001 + }, + { + "epoch": 0.9622577218335973, + "grad_norm": 1.7906298637390137, + "learning_rate": 1.1094342402752658e-05, + "loss": 3.1101, + "step": 12002 + }, + { + "epoch": 0.9623378966146199, + "grad_norm": 1.7649776935577393, + "learning_rate": 1.1093051744898533e-05, + "loss": 3.2852, + "step": 12003 + }, + { + "epoch": 0.9624180713956425, + "grad_norm": 1.7770494222640991, + "learning_rate": 1.1091761068615945e-05, + "loss": 3.1029, + "step": 12004 + }, + { + "epoch": 0.9624982461766651, + "grad_norm": 1.9329231977462769, + "learning_rate": 1.1090470373926645e-05, + "loss": 3.1234, + "step": 12005 + }, + { + "epoch": 0.9625784209576878, + "grad_norm": 1.7366374731063843, + "learning_rate": 1.1089179660852403e-05, + "loss": 3.1797, + "step": 12006 + }, + { + "epoch": 0.9626585957387104, + "grad_norm": 1.9747591018676758, + "learning_rate": 1.1087888929414972e-05, + "loss": 3.3243, + "step": 12007 + }, + { + "epoch": 0.962738770519733, + "grad_norm": 1.8486647605895996, + "learning_rate": 1.108659817963612e-05, + "loss": 3.162, + "step": 12008 + }, + { + "epoch": 0.9628189453007556, + "grad_norm": 1.8013229370117188, + "learning_rate": 1.1085307411537604e-05, + "loss": 3.1798, + "step": 12009 + }, + { + "epoch": 0.9628991200817782, + "grad_norm": 1.7334648370742798, + "learning_rate": 1.1084016625141187e-05, + "loss": 3.2588, + "step": 12010 + }, + { + "epoch": 0.9629792948628009, + "grad_norm": 1.7744044065475464, + "learning_rate": 1.1082725820468637e-05, + "loss": 3.2556, + "step": 12011 + }, + { + "epoch": 0.9630594696438235, + "grad_norm": 1.8923734426498413, + "learning_rate": 1.1081434997541705e-05, + "loss": 3.2468, + "step": 12012 + }, + { + "epoch": 0.9631396444248461, + "grad_norm": 1.806052327156067, + "learning_rate": 1.108014415638216e-05, + "loss": 3.095, + "step": 12013 + }, + { + "epoch": 0.9632198192058687, + "grad_norm": 1.8230043649673462, + "learning_rate": 1.1078853297011769e-05, + "loss": 3.1596, + "step": 12014 + }, + { + "epoch": 0.9632999939868915, + "grad_norm": 1.7953875064849854, + "learning_rate": 1.107756241945229e-05, + "loss": 3.1184, + "step": 12015 + }, + { + "epoch": 0.9633801687679141, + "grad_norm": 2.0802652835845947, + "learning_rate": 1.107627152372549e-05, + "loss": 3.288, + "step": 12016 + }, + { + "epoch": 0.9634603435489367, + "grad_norm": 1.9947601556777954, + "learning_rate": 1.1074980609853127e-05, + "loss": 3.2186, + "step": 12017 + }, + { + "epoch": 0.9635405183299594, + "grad_norm": 1.799763560295105, + "learning_rate": 1.1073689677856973e-05, + "loss": 3.2102, + "step": 12018 + }, + { + "epoch": 0.963620693110982, + "grad_norm": 1.772409439086914, + "learning_rate": 1.1072398727758791e-05, + "loss": 3.2303, + "step": 12019 + }, + { + "epoch": 0.9637008678920046, + "grad_norm": 1.821935772895813, + "learning_rate": 1.1071107759580342e-05, + "loss": 3.2272, + "step": 12020 + }, + { + "epoch": 0.9637810426730272, + "grad_norm": 1.777698040008545, + "learning_rate": 1.1069816773343394e-05, + "loss": 3.2004, + "step": 12021 + }, + { + "epoch": 0.9638612174540498, + "grad_norm": 1.8157247304916382, + "learning_rate": 1.1068525769069711e-05, + "loss": 3.2621, + "step": 12022 + }, + { + "epoch": 0.9639413922350725, + "grad_norm": 1.7761248350143433, + "learning_rate": 1.1067234746781061e-05, + "loss": 3.1898, + "step": 12023 + }, + { + "epoch": 0.9640215670160951, + "grad_norm": 1.8307421207427979, + "learning_rate": 1.1065943706499212e-05, + "loss": 3.2382, + "step": 12024 + }, + { + "epoch": 0.9641017417971177, + "grad_norm": 1.6991206407546997, + "learning_rate": 1.1064652648245924e-05, + "loss": 3.1519, + "step": 12025 + }, + { + "epoch": 0.9641819165781403, + "grad_norm": 1.7894046306610107, + "learning_rate": 1.106336157204297e-05, + "loss": 3.249, + "step": 12026 + }, + { + "epoch": 0.964262091359163, + "grad_norm": 1.818941354751587, + "learning_rate": 1.1062070477912115e-05, + "loss": 3.2484, + "step": 12027 + }, + { + "epoch": 0.9643422661401856, + "grad_norm": 1.784920334815979, + "learning_rate": 1.1060779365875123e-05, + "loss": 3.2153, + "step": 12028 + }, + { + "epoch": 0.9644224409212082, + "grad_norm": 1.7910559177398682, + "learning_rate": 1.1059488235953769e-05, + "loss": 3.1126, + "step": 12029 + }, + { + "epoch": 0.9645026157022308, + "grad_norm": 1.8086224794387817, + "learning_rate": 1.1058197088169813e-05, + "loss": 3.2011, + "step": 12030 + }, + { + "epoch": 0.9645827904832535, + "grad_norm": 1.7419235706329346, + "learning_rate": 1.105690592254503e-05, + "loss": 3.2812, + "step": 12031 + }, + { + "epoch": 0.9646629652642761, + "grad_norm": 1.7807716131210327, + "learning_rate": 1.1055614739101181e-05, + "loss": 3.3335, + "step": 12032 + }, + { + "epoch": 0.9647431400452987, + "grad_norm": 1.8160977363586426, + "learning_rate": 1.1054323537860043e-05, + "loss": 3.2274, + "step": 12033 + }, + { + "epoch": 0.9648233148263213, + "grad_norm": 1.7182340621948242, + "learning_rate": 1.1053032318843383e-05, + "loss": 3.1528, + "step": 12034 + }, + { + "epoch": 0.964903489607344, + "grad_norm": 1.6703686714172363, + "learning_rate": 1.1051741082072966e-05, + "loss": 3.2673, + "step": 12035 + }, + { + "epoch": 0.9649836643883667, + "grad_norm": 1.7730755805969238, + "learning_rate": 1.1050449827570569e-05, + "loss": 3.2589, + "step": 12036 + }, + { + "epoch": 0.9650638391693893, + "grad_norm": 1.7094544172286987, + "learning_rate": 1.1049158555357952e-05, + "loss": 3.1863, + "step": 12037 + }, + { + "epoch": 0.9651440139504119, + "grad_norm": 1.8650258779525757, + "learning_rate": 1.1047867265456895e-05, + "loss": 3.2109, + "step": 12038 + }, + { + "epoch": 0.9652241887314346, + "grad_norm": 1.9347189664840698, + "learning_rate": 1.1046575957889167e-05, + "loss": 3.2141, + "step": 12039 + }, + { + "epoch": 0.9653043635124572, + "grad_norm": 1.7818840742111206, + "learning_rate": 1.1045284632676535e-05, + "loss": 3.2256, + "step": 12040 + }, + { + "epoch": 0.9653845382934798, + "grad_norm": 1.715016484260559, + "learning_rate": 1.1043993289840775e-05, + "loss": 3.1219, + "step": 12041 + }, + { + "epoch": 0.9654647130745024, + "grad_norm": 1.7945398092269897, + "learning_rate": 1.1042701929403653e-05, + "loss": 3.2936, + "step": 12042 + }, + { + "epoch": 0.9655448878555251, + "grad_norm": 1.8543009757995605, + "learning_rate": 1.1041410551386944e-05, + "loss": 3.1532, + "step": 12043 + }, + { + "epoch": 0.9656250626365477, + "grad_norm": 1.793410062789917, + "learning_rate": 1.1040119155812424e-05, + "loss": 3.2585, + "step": 12044 + }, + { + "epoch": 0.9657052374175703, + "grad_norm": 1.7098032236099243, + "learning_rate": 1.1038827742701857e-05, + "loss": 3.0704, + "step": 12045 + }, + { + "epoch": 0.9657854121985929, + "grad_norm": 1.8002365827560425, + "learning_rate": 1.1037536312077023e-05, + "loss": 3.203, + "step": 12046 + }, + { + "epoch": 0.9658655869796156, + "grad_norm": 1.7547495365142822, + "learning_rate": 1.1036244863959691e-05, + "loss": 3.1718, + "step": 12047 + }, + { + "epoch": 0.9659457617606382, + "grad_norm": 1.776275634765625, + "learning_rate": 1.1034953398371635e-05, + "loss": 3.1732, + "step": 12048 + }, + { + "epoch": 0.9660259365416608, + "grad_norm": 1.7599674463272095, + "learning_rate": 1.1033661915334632e-05, + "loss": 3.2981, + "step": 12049 + }, + { + "epoch": 0.9661061113226834, + "grad_norm": 1.928133487701416, + "learning_rate": 1.1032370414870452e-05, + "loss": 3.2317, + "step": 12050 + }, + { + "epoch": 0.966186286103706, + "grad_norm": 1.88199782371521, + "learning_rate": 1.1031078897000873e-05, + "loss": 3.1635, + "step": 12051 + }, + { + "epoch": 0.9662664608847287, + "grad_norm": 1.7414518594741821, + "learning_rate": 1.1029787361747663e-05, + "loss": 3.207, + "step": 12052 + }, + { + "epoch": 0.9663466356657513, + "grad_norm": 1.7845115661621094, + "learning_rate": 1.1028495809132603e-05, + "loss": 3.0682, + "step": 12053 + }, + { + "epoch": 0.9664268104467739, + "grad_norm": 1.732946515083313, + "learning_rate": 1.1027204239177467e-05, + "loss": 3.2729, + "step": 12054 + }, + { + "epoch": 0.9665069852277965, + "grad_norm": 1.829667329788208, + "learning_rate": 1.1025912651904029e-05, + "loss": 3.2067, + "step": 12055 + }, + { + "epoch": 0.9665871600088193, + "grad_norm": 1.6739470958709717, + "learning_rate": 1.1024621047334065e-05, + "loss": 3.2153, + "step": 12056 + }, + { + "epoch": 0.9666673347898419, + "grad_norm": 1.7078016996383667, + "learning_rate": 1.1023329425489352e-05, + "loss": 3.1442, + "step": 12057 + }, + { + "epoch": 0.9667475095708645, + "grad_norm": 1.9027856588363647, + "learning_rate": 1.1022037786391663e-05, + "loss": 3.3063, + "step": 12058 + }, + { + "epoch": 0.9668276843518872, + "grad_norm": 1.7586870193481445, + "learning_rate": 1.102074613006278e-05, + "loss": 3.2326, + "step": 12059 + }, + { + "epoch": 0.9669078591329098, + "grad_norm": 1.7823405265808105, + "learning_rate": 1.1019454456524476e-05, + "loss": 3.2356, + "step": 12060 + }, + { + "epoch": 0.9669880339139324, + "grad_norm": 1.7906858921051025, + "learning_rate": 1.101816276579853e-05, + "loss": 3.2332, + "step": 12061 + }, + { + "epoch": 0.967068208694955, + "grad_norm": 1.772606611251831, + "learning_rate": 1.1016871057906718e-05, + "loss": 3.1892, + "step": 12062 + }, + { + "epoch": 0.9671483834759776, + "grad_norm": 1.9908159971237183, + "learning_rate": 1.1015579332870817e-05, + "loss": 3.1822, + "step": 12063 + }, + { + "epoch": 0.9672285582570003, + "grad_norm": 1.7740100622177124, + "learning_rate": 1.1014287590712609e-05, + "loss": 3.2076, + "step": 12064 + }, + { + "epoch": 0.9673087330380229, + "grad_norm": 1.770736813545227, + "learning_rate": 1.1012995831453867e-05, + "loss": 3.2385, + "step": 12065 + }, + { + "epoch": 0.9673889078190455, + "grad_norm": 1.770451545715332, + "learning_rate": 1.1011704055116374e-05, + "loss": 3.186, + "step": 12066 + }, + { + "epoch": 0.9674690826000681, + "grad_norm": 1.8907756805419922, + "learning_rate": 1.1010412261721908e-05, + "loss": 3.1296, + "step": 12067 + }, + { + "epoch": 0.9675492573810908, + "grad_norm": 1.6526529788970947, + "learning_rate": 1.1009120451292243e-05, + "loss": 3.1693, + "step": 12068 + }, + { + "epoch": 0.9676294321621134, + "grad_norm": 1.7726047039031982, + "learning_rate": 1.1007828623849168e-05, + "loss": 3.1006, + "step": 12069 + }, + { + "epoch": 0.967709606943136, + "grad_norm": 1.7794189453125, + "learning_rate": 1.1006536779414453e-05, + "loss": 3.2996, + "step": 12070 + }, + { + "epoch": 0.9677897817241586, + "grad_norm": 1.7445776462554932, + "learning_rate": 1.1005244918009888e-05, + "loss": 3.2587, + "step": 12071 + }, + { + "epoch": 0.9678699565051813, + "grad_norm": 1.717583179473877, + "learning_rate": 1.1003953039657245e-05, + "loss": 3.1522, + "step": 12072 + }, + { + "epoch": 0.9679501312862039, + "grad_norm": 1.7815231084823608, + "learning_rate": 1.1002661144378306e-05, + "loss": 3.0564, + "step": 12073 + }, + { + "epoch": 0.9680303060672265, + "grad_norm": 1.7630195617675781, + "learning_rate": 1.1001369232194856e-05, + "loss": 3.156, + "step": 12074 + }, + { + "epoch": 0.9681104808482491, + "grad_norm": 1.781253457069397, + "learning_rate": 1.1000077303128674e-05, + "loss": 3.1563, + "step": 12075 + }, + { + "epoch": 0.9681906556292719, + "grad_norm": 1.8305565118789673, + "learning_rate": 1.099878535720154e-05, + "loss": 3.1928, + "step": 12076 + }, + { + "epoch": 0.9682708304102945, + "grad_norm": 0.6895906329154968, + "learning_rate": 1.0997493394435234e-05, + "loss": 2.472, + "step": 12077 + }, + { + "epoch": 0.9683510051913171, + "grad_norm": 1.8179936408996582, + "learning_rate": 1.0996201414851544e-05, + "loss": 3.2554, + "step": 12078 + }, + { + "epoch": 0.9684311799723397, + "grad_norm": 1.8583978414535522, + "learning_rate": 1.099490941847225e-05, + "loss": 3.2629, + "step": 12079 + }, + { + "epoch": 0.9685113547533624, + "grad_norm": 1.8608311414718628, + "learning_rate": 1.0993617405319132e-05, + "loss": 3.2246, + "step": 12080 + }, + { + "epoch": 0.968591529534385, + "grad_norm": 1.7827986478805542, + "learning_rate": 1.0992325375413977e-05, + "loss": 3.2126, + "step": 12081 + }, + { + "epoch": 0.9686717043154076, + "grad_norm": 1.750524640083313, + "learning_rate": 1.0991033328778565e-05, + "loss": 3.2537, + "step": 12082 + }, + { + "epoch": 0.9687518790964302, + "grad_norm": 1.720159888267517, + "learning_rate": 1.0989741265434679e-05, + "loss": 3.1898, + "step": 12083 + }, + { + "epoch": 0.9688320538774529, + "grad_norm": 1.7718161344528198, + "learning_rate": 1.0988449185404106e-05, + "loss": 3.0651, + "step": 12084 + }, + { + "epoch": 0.9689122286584755, + "grad_norm": 1.7337710857391357, + "learning_rate": 1.0987157088708625e-05, + "loss": 3.1295, + "step": 12085 + }, + { + "epoch": 0.9689924034394981, + "grad_norm": 1.823130488395691, + "learning_rate": 1.0985864975370027e-05, + "loss": 3.3763, + "step": 12086 + }, + { + "epoch": 0.9690725782205207, + "grad_norm": 1.7505749464035034, + "learning_rate": 1.0984572845410091e-05, + "loss": 3.1796, + "step": 12087 + }, + { + "epoch": 0.9691527530015434, + "grad_norm": 1.7515308856964111, + "learning_rate": 1.0983280698850602e-05, + "loss": 3.1835, + "step": 12088 + }, + { + "epoch": 0.969232927782566, + "grad_norm": 1.7123550176620483, + "learning_rate": 1.098198853571335e-05, + "loss": 3.135, + "step": 12089 + }, + { + "epoch": 0.9693131025635886, + "grad_norm": 1.7325975894927979, + "learning_rate": 1.0980696356020117e-05, + "loss": 3.1598, + "step": 12090 + }, + { + "epoch": 0.9693932773446112, + "grad_norm": 1.7097413539886475, + "learning_rate": 1.097940415979269e-05, + "loss": 3.0496, + "step": 12091 + }, + { + "epoch": 0.9694734521256338, + "grad_norm": 1.905653476715088, + "learning_rate": 1.0978111947052852e-05, + "loss": 3.1475, + "step": 12092 + }, + { + "epoch": 0.9695536269066565, + "grad_norm": 1.7262343168258667, + "learning_rate": 1.0976819717822391e-05, + "loss": 3.1478, + "step": 12093 + }, + { + "epoch": 0.9696338016876791, + "grad_norm": 1.7803058624267578, + "learning_rate": 1.0975527472123096e-05, + "loss": 3.1626, + "step": 12094 + }, + { + "epoch": 0.9697139764687017, + "grad_norm": 1.814408302307129, + "learning_rate": 1.0974235209976748e-05, + "loss": 3.2088, + "step": 12095 + }, + { + "epoch": 0.9697941512497243, + "grad_norm": 1.8556517362594604, + "learning_rate": 1.0972942931405141e-05, + "loss": 3.2292, + "step": 12096 + }, + { + "epoch": 0.9698743260307471, + "grad_norm": 1.8135030269622803, + "learning_rate": 1.0971650636430059e-05, + "loss": 3.1681, + "step": 12097 + }, + { + "epoch": 0.9699545008117697, + "grad_norm": 1.7919161319732666, + "learning_rate": 1.0970358325073287e-05, + "loss": 3.1875, + "step": 12098 + }, + { + "epoch": 0.9700346755927923, + "grad_norm": 1.82545006275177, + "learning_rate": 1.096906599735662e-05, + "loss": 3.252, + "step": 12099 + }, + { + "epoch": 0.970114850373815, + "grad_norm": 2.0822103023529053, + "learning_rate": 1.0967773653301838e-05, + "loss": 3.3069, + "step": 12100 + }, + { + "epoch": 0.9701950251548376, + "grad_norm": 1.9085773229599, + "learning_rate": 1.0966481292930734e-05, + "loss": 3.1681, + "step": 12101 + }, + { + "epoch": 0.9702751999358602, + "grad_norm": 1.8686527013778687, + "learning_rate": 1.0965188916265097e-05, + "loss": 3.1881, + "step": 12102 + }, + { + "epoch": 0.9703553747168828, + "grad_norm": 1.7964290380477905, + "learning_rate": 1.0963896523326714e-05, + "loss": 3.249, + "step": 12103 + }, + { + "epoch": 0.9704355494979054, + "grad_norm": 0.6975285410881042, + "learning_rate": 1.0962604114137378e-05, + "loss": 2.4667, + "step": 12104 + }, + { + "epoch": 0.9705157242789281, + "grad_norm": 1.8582812547683716, + "learning_rate": 1.0961311688718872e-05, + "loss": 3.2412, + "step": 12105 + }, + { + "epoch": 0.9705958990599507, + "grad_norm": 1.8008358478546143, + "learning_rate": 1.0960019247092994e-05, + "loss": 3.1891, + "step": 12106 + }, + { + "epoch": 0.9706760738409733, + "grad_norm": 1.7954264879226685, + "learning_rate": 1.0958726789281526e-05, + "loss": 3.1521, + "step": 12107 + }, + { + "epoch": 0.9707562486219959, + "grad_norm": 1.7948145866394043, + "learning_rate": 1.0957434315306264e-05, + "loss": 3.2454, + "step": 12108 + }, + { + "epoch": 0.9708364234030186, + "grad_norm": 1.7196378707885742, + "learning_rate": 1.0956141825188998e-05, + "loss": 3.129, + "step": 12109 + }, + { + "epoch": 0.9709165981840412, + "grad_norm": 1.752724051475525, + "learning_rate": 1.0954849318951515e-05, + "loss": 3.1819, + "step": 12110 + }, + { + "epoch": 0.9709967729650638, + "grad_norm": 1.825840950012207, + "learning_rate": 1.095355679661561e-05, + "loss": 3.1464, + "step": 12111 + }, + { + "epoch": 0.9710769477460864, + "grad_norm": 1.8868244886398315, + "learning_rate": 1.0952264258203074e-05, + "loss": 3.2091, + "step": 12112 + }, + { + "epoch": 0.9711571225271091, + "grad_norm": 1.7931607961654663, + "learning_rate": 1.0950971703735697e-05, + "loss": 3.2074, + "step": 12113 + }, + { + "epoch": 0.9712372973081317, + "grad_norm": 1.827751636505127, + "learning_rate": 1.0949679133235274e-05, + "loss": 3.2029, + "step": 12114 + }, + { + "epoch": 0.9713174720891543, + "grad_norm": 1.734702229499817, + "learning_rate": 1.0948386546723596e-05, + "loss": 3.2189, + "step": 12115 + }, + { + "epoch": 0.9713976468701769, + "grad_norm": 1.896180510520935, + "learning_rate": 1.0947093944222453e-05, + "loss": 3.2363, + "step": 12116 + }, + { + "epoch": 0.9714778216511997, + "grad_norm": 1.8746522665023804, + "learning_rate": 1.094580132575364e-05, + "loss": 3.1046, + "step": 12117 + }, + { + "epoch": 0.9715579964322223, + "grad_norm": 1.9617459774017334, + "learning_rate": 1.094450869133895e-05, + "loss": 3.232, + "step": 12118 + }, + { + "epoch": 0.9716381712132449, + "grad_norm": 1.8558375835418701, + "learning_rate": 1.0943216041000177e-05, + "loss": 3.2302, + "step": 12119 + }, + { + "epoch": 0.9717183459942675, + "grad_norm": 1.7847905158996582, + "learning_rate": 1.0941923374759113e-05, + "loss": 3.2085, + "step": 12120 + }, + { + "epoch": 0.9717985207752902, + "grad_norm": 1.76227605342865, + "learning_rate": 1.0940630692637557e-05, + "loss": 3.2269, + "step": 12121 + }, + { + "epoch": 0.9718786955563128, + "grad_norm": 1.768020749092102, + "learning_rate": 1.0939337994657295e-05, + "loss": 3.1772, + "step": 12122 + }, + { + "epoch": 0.9719588703373354, + "grad_norm": 1.8422350883483887, + "learning_rate": 1.0938045280840125e-05, + "loss": 3.2024, + "step": 12123 + }, + { + "epoch": 0.972039045118358, + "grad_norm": 1.7042790651321411, + "learning_rate": 1.0936752551207846e-05, + "loss": 3.1005, + "step": 12124 + }, + { + "epoch": 0.9721192198993807, + "grad_norm": 1.7542486190795898, + "learning_rate": 1.0935459805782245e-05, + "loss": 3.1788, + "step": 12125 + }, + { + "epoch": 0.9721993946804033, + "grad_norm": 1.7225548028945923, + "learning_rate": 1.0934167044585125e-05, + "loss": 3.0829, + "step": 12126 + }, + { + "epoch": 0.9722795694614259, + "grad_norm": 1.7064805030822754, + "learning_rate": 1.0932874267638273e-05, + "loss": 3.1221, + "step": 12127 + }, + { + "epoch": 0.9723597442424485, + "grad_norm": 1.8564931154251099, + "learning_rate": 1.0931581474963493e-05, + "loss": 3.2343, + "step": 12128 + }, + { + "epoch": 0.9724399190234712, + "grad_norm": 0.7274773120880127, + "learning_rate": 1.0930288666582578e-05, + "loss": 2.4907, + "step": 12129 + }, + { + "epoch": 0.9725200938044938, + "grad_norm": 1.6983146667480469, + "learning_rate": 1.0928995842517323e-05, + "loss": 3.2085, + "step": 12130 + }, + { + "epoch": 0.9726002685855164, + "grad_norm": 1.8423751592636108, + "learning_rate": 1.0927703002789527e-05, + "loss": 3.2992, + "step": 12131 + }, + { + "epoch": 0.972680443366539, + "grad_norm": 1.799834966659546, + "learning_rate": 1.0926410147420982e-05, + "loss": 3.1157, + "step": 12132 + }, + { + "epoch": 0.9727606181475617, + "grad_norm": 1.8092379570007324, + "learning_rate": 1.092511727643349e-05, + "loss": 3.1299, + "step": 12133 + }, + { + "epoch": 0.9728407929285843, + "grad_norm": 1.8439054489135742, + "learning_rate": 1.092382438984885e-05, + "loss": 3.2565, + "step": 12134 + }, + { + "epoch": 0.9729209677096069, + "grad_norm": 1.7873541116714478, + "learning_rate": 1.0922531487688853e-05, + "loss": 3.1245, + "step": 12135 + }, + { + "epoch": 0.9730011424906295, + "grad_norm": 1.75563645362854, + "learning_rate": 1.0921238569975301e-05, + "loss": 3.1393, + "step": 12136 + }, + { + "epoch": 0.9730813172716521, + "grad_norm": 1.7264466285705566, + "learning_rate": 1.091994563672999e-05, + "loss": 3.1595, + "step": 12137 + }, + { + "epoch": 0.9731614920526749, + "grad_norm": 1.9934747219085693, + "learning_rate": 1.091865268797472e-05, + "loss": 3.188, + "step": 12138 + }, + { + "epoch": 0.9732416668336975, + "grad_norm": 1.7843483686447144, + "learning_rate": 1.091735972373129e-05, + "loss": 3.1494, + "step": 12139 + }, + { + "epoch": 0.9733218416147201, + "grad_norm": 1.816285252571106, + "learning_rate": 1.0916066744021499e-05, + "loss": 3.2688, + "step": 12140 + }, + { + "epoch": 0.9734020163957428, + "grad_norm": 0.7136083841323853, + "learning_rate": 1.0914773748867147e-05, + "loss": 2.4938, + "step": 12141 + }, + { + "epoch": 0.9734821911767654, + "grad_norm": 1.9324815273284912, + "learning_rate": 1.091348073829003e-05, + "loss": 3.1996, + "step": 12142 + }, + { + "epoch": 0.973562365957788, + "grad_norm": 1.7467926740646362, + "learning_rate": 1.0912187712311948e-05, + "loss": 3.2631, + "step": 12143 + }, + { + "epoch": 0.9736425407388106, + "grad_norm": 1.9523015022277832, + "learning_rate": 1.0910894670954706e-05, + "loss": 3.2516, + "step": 12144 + }, + { + "epoch": 0.9737227155198332, + "grad_norm": 0.7146855592727661, + "learning_rate": 1.09096016142401e-05, + "loss": 2.4822, + "step": 12145 + }, + { + "epoch": 0.9738028903008559, + "grad_norm": 0.7356863021850586, + "learning_rate": 1.0908308542189932e-05, + "loss": 2.4844, + "step": 12146 + }, + { + "epoch": 0.9738830650818785, + "grad_norm": 0.7214996814727783, + "learning_rate": 1.0907015454826e-05, + "loss": 2.4648, + "step": 12147 + }, + { + "epoch": 0.9739632398629011, + "grad_norm": 1.8855831623077393, + "learning_rate": 1.0905722352170107e-05, + "loss": 3.0718, + "step": 12148 + }, + { + "epoch": 0.9740434146439237, + "grad_norm": 1.718043327331543, + "learning_rate": 1.0904429234244056e-05, + "loss": 3.1219, + "step": 12149 + }, + { + "epoch": 0.9741235894249464, + "grad_norm": 1.750968337059021, + "learning_rate": 1.0903136101069644e-05, + "loss": 3.1342, + "step": 12150 + }, + { + "epoch": 0.974203764205969, + "grad_norm": 1.8104884624481201, + "learning_rate": 1.090184295266868e-05, + "loss": 3.24, + "step": 12151 + }, + { + "epoch": 0.9742839389869916, + "grad_norm": 1.7433758974075317, + "learning_rate": 1.090054978906296e-05, + "loss": 3.1649, + "step": 12152 + }, + { + "epoch": 0.9743641137680142, + "grad_norm": 1.9801419973373413, + "learning_rate": 1.0899256610274284e-05, + "loss": 3.1986, + "step": 12153 + }, + { + "epoch": 0.9744442885490369, + "grad_norm": 1.7207247018814087, + "learning_rate": 1.0897963416324466e-05, + "loss": 3.1957, + "step": 12154 + }, + { + "epoch": 0.9745244633300595, + "grad_norm": 1.818915843963623, + "learning_rate": 1.0896670207235296e-05, + "loss": 3.0605, + "step": 12155 + }, + { + "epoch": 0.9746046381110821, + "grad_norm": 1.867674469947815, + "learning_rate": 1.0895376983028585e-05, + "loss": 3.2235, + "step": 12156 + }, + { + "epoch": 0.9746848128921047, + "grad_norm": 1.6439920663833618, + "learning_rate": 1.0894083743726132e-05, + "loss": 3.2736, + "step": 12157 + }, + { + "epoch": 0.9747649876731275, + "grad_norm": 1.8267441987991333, + "learning_rate": 1.089279048934974e-05, + "loss": 3.1158, + "step": 12158 + }, + { + "epoch": 0.9748451624541501, + "grad_norm": 1.7514792680740356, + "learning_rate": 1.089149721992122e-05, + "loss": 3.2007, + "step": 12159 + }, + { + "epoch": 0.9749253372351727, + "grad_norm": 1.79424250125885, + "learning_rate": 1.0890203935462366e-05, + "loss": 3.1436, + "step": 12160 + }, + { + "epoch": 0.9750055120161953, + "grad_norm": 1.8774250745773315, + "learning_rate": 1.088891063599499e-05, + "loss": 3.2073, + "step": 12161 + }, + { + "epoch": 0.975085686797218, + "grad_norm": 1.7779278755187988, + "learning_rate": 1.0887617321540895e-05, + "loss": 3.1414, + "step": 12162 + }, + { + "epoch": 0.9751658615782406, + "grad_norm": 1.7619723081588745, + "learning_rate": 1.0886323992121884e-05, + "loss": 3.0612, + "step": 12163 + }, + { + "epoch": 0.9752460363592632, + "grad_norm": 1.852473497390747, + "learning_rate": 1.0885030647759763e-05, + "loss": 3.2516, + "step": 12164 + }, + { + "epoch": 0.9753262111402858, + "grad_norm": 1.723407506942749, + "learning_rate": 1.0883737288476335e-05, + "loss": 3.2571, + "step": 12165 + }, + { + "epoch": 0.9754063859213085, + "grad_norm": 1.9211162328720093, + "learning_rate": 1.0882443914293408e-05, + "loss": 3.2048, + "step": 12166 + }, + { + "epoch": 0.9754865607023311, + "grad_norm": 0.7267510294914246, + "learning_rate": 1.0881150525232791e-05, + "loss": 2.4722, + "step": 12167 + }, + { + "epoch": 0.9755667354833537, + "grad_norm": 1.7425575256347656, + "learning_rate": 1.0879857121316283e-05, + "loss": 3.165, + "step": 12168 + }, + { + "epoch": 0.9756469102643763, + "grad_norm": 1.7534748315811157, + "learning_rate": 1.0878563702565698e-05, + "loss": 3.1743, + "step": 12169 + }, + { + "epoch": 0.975727085045399, + "grad_norm": 1.8041408061981201, + "learning_rate": 1.0877270269002836e-05, + "loss": 3.2181, + "step": 12170 + }, + { + "epoch": 0.9758072598264216, + "grad_norm": 1.7474318742752075, + "learning_rate": 1.0875976820649506e-05, + "loss": 3.1286, + "step": 12171 + }, + { + "epoch": 0.9758874346074442, + "grad_norm": 1.8442184925079346, + "learning_rate": 1.0874683357527516e-05, + "loss": 3.1758, + "step": 12172 + }, + { + "epoch": 0.9759676093884668, + "grad_norm": 1.8132153749465942, + "learning_rate": 1.0873389879658674e-05, + "loss": 3.1131, + "step": 12173 + }, + { + "epoch": 0.9760477841694895, + "grad_norm": 1.7666858434677124, + "learning_rate": 1.0872096387064786e-05, + "loss": 3.308, + "step": 12174 + }, + { + "epoch": 0.9761279589505121, + "grad_norm": 1.7564947605133057, + "learning_rate": 1.0870802879767658e-05, + "loss": 3.213, + "step": 12175 + }, + { + "epoch": 0.9762081337315347, + "grad_norm": 1.7423557043075562, + "learning_rate": 1.0869509357789103e-05, + "loss": 3.1617, + "step": 12176 + }, + { + "epoch": 0.9762883085125573, + "grad_norm": 1.7187482118606567, + "learning_rate": 1.0868215821150927e-05, + "loss": 3.2543, + "step": 12177 + }, + { + "epoch": 0.9763684832935801, + "grad_norm": 1.776244878768921, + "learning_rate": 1.0866922269874935e-05, + "loss": 3.1224, + "step": 12178 + }, + { + "epoch": 0.9764486580746027, + "grad_norm": 1.7524058818817139, + "learning_rate": 1.0865628703982945e-05, + "loss": 3.3015, + "step": 12179 + }, + { + "epoch": 0.9765288328556253, + "grad_norm": 1.7251029014587402, + "learning_rate": 1.0864335123496754e-05, + "loss": 3.2007, + "step": 12180 + }, + { + "epoch": 0.9766090076366479, + "grad_norm": 1.86038076877594, + "learning_rate": 1.0863041528438178e-05, + "loss": 3.1782, + "step": 12181 + }, + { + "epoch": 0.9766891824176706, + "grad_norm": 1.705561637878418, + "learning_rate": 1.086174791882903e-05, + "loss": 3.1974, + "step": 12182 + }, + { + "epoch": 0.9767693571986932, + "grad_norm": 1.8423190116882324, + "learning_rate": 1.0860454294691112e-05, + "loss": 3.1812, + "step": 12183 + }, + { + "epoch": 0.9768495319797158, + "grad_norm": 1.698744297027588, + "learning_rate": 1.0859160656046242e-05, + "loss": 3.213, + "step": 12184 + }, + { + "epoch": 0.9769297067607384, + "grad_norm": 1.7188563346862793, + "learning_rate": 1.0857867002916222e-05, + "loss": 3.2339, + "step": 12185 + }, + { + "epoch": 0.977009881541761, + "grad_norm": 1.8084672689437866, + "learning_rate": 1.0856573335322866e-05, + "loss": 3.2642, + "step": 12186 + }, + { + "epoch": 0.9770900563227837, + "grad_norm": 1.744375467300415, + "learning_rate": 1.0855279653287991e-05, + "loss": 3.1797, + "step": 12187 + }, + { + "epoch": 0.9771702311038063, + "grad_norm": 1.9991228580474854, + "learning_rate": 1.0853985956833397e-05, + "loss": 3.255, + "step": 12188 + }, + { + "epoch": 0.9772504058848289, + "grad_norm": 1.7679591178894043, + "learning_rate": 1.0852692245980904e-05, + "loss": 3.1113, + "step": 12189 + }, + { + "epoch": 0.9773305806658515, + "grad_norm": 0.6945270895957947, + "learning_rate": 1.0851398520752319e-05, + "loss": 2.5037, + "step": 12190 + }, + { + "epoch": 0.9774107554468742, + "grad_norm": 1.7580111026763916, + "learning_rate": 1.0850104781169455e-05, + "loss": 3.2955, + "step": 12191 + }, + { + "epoch": 0.9774909302278968, + "grad_norm": 1.7751694917678833, + "learning_rate": 1.0848811027254124e-05, + "loss": 3.1233, + "step": 12192 + }, + { + "epoch": 0.9775711050089194, + "grad_norm": 1.723312258720398, + "learning_rate": 1.0847517259028138e-05, + "loss": 3.1697, + "step": 12193 + }, + { + "epoch": 0.977651279789942, + "grad_norm": 1.8454749584197998, + "learning_rate": 1.0846223476513312e-05, + "loss": 3.2375, + "step": 12194 + }, + { + "epoch": 0.9777314545709647, + "grad_norm": 1.8111352920532227, + "learning_rate": 1.0844929679731453e-05, + "loss": 3.2968, + "step": 12195 + }, + { + "epoch": 0.9778116293519873, + "grad_norm": 1.8777806758880615, + "learning_rate": 1.0843635868704378e-05, + "loss": 3.1914, + "step": 12196 + }, + { + "epoch": 0.9778918041330099, + "grad_norm": 1.7763954401016235, + "learning_rate": 1.0842342043453901e-05, + "loss": 3.3477, + "step": 12197 + }, + { + "epoch": 0.9779719789140325, + "grad_norm": 1.7794151306152344, + "learning_rate": 1.0841048204001834e-05, + "loss": 3.1538, + "step": 12198 + }, + { + "epoch": 0.9780521536950553, + "grad_norm": 1.722772479057312, + "learning_rate": 1.0839754350369991e-05, + "loss": 3.2297, + "step": 12199 + }, + { + "epoch": 0.9781323284760779, + "grad_norm": 1.7282756567001343, + "learning_rate": 1.0838460482580185e-05, + "loss": 3.1744, + "step": 12200 + }, + { + "epoch": 0.9782125032571005, + "grad_norm": 1.703254222869873, + "learning_rate": 1.083716660065423e-05, + "loss": 3.2032, + "step": 12201 + }, + { + "epoch": 0.9782926780381231, + "grad_norm": 0.7122766375541687, + "learning_rate": 1.0835872704613943e-05, + "loss": 2.4891, + "step": 12202 + }, + { + "epoch": 0.9783728528191458, + "grad_norm": 1.7593512535095215, + "learning_rate": 1.0834578794481134e-05, + "loss": 3.1608, + "step": 12203 + }, + { + "epoch": 0.9784530276001684, + "grad_norm": 1.6647785902023315, + "learning_rate": 1.0833284870277623e-05, + "loss": 3.1321, + "step": 12204 + }, + { + "epoch": 0.978533202381191, + "grad_norm": 1.7243201732635498, + "learning_rate": 1.0831990932025222e-05, + "loss": 3.2076, + "step": 12205 + }, + { + "epoch": 0.9786133771622136, + "grad_norm": 1.8172658681869507, + "learning_rate": 1.0830696979745744e-05, + "loss": 3.2006, + "step": 12206 + }, + { + "epoch": 0.9786935519432363, + "grad_norm": 1.752950668334961, + "learning_rate": 1.0829403013461014e-05, + "loss": 3.1975, + "step": 12207 + }, + { + "epoch": 0.9787737267242589, + "grad_norm": 1.8493324518203735, + "learning_rate": 1.0828109033192838e-05, + "loss": 3.2886, + "step": 12208 + }, + { + "epoch": 0.9788539015052815, + "grad_norm": 1.6311484575271606, + "learning_rate": 1.0826815038963037e-05, + "loss": 3.1084, + "step": 12209 + }, + { + "epoch": 0.9789340762863041, + "grad_norm": 1.761417031288147, + "learning_rate": 1.0825521030793423e-05, + "loss": 3.251, + "step": 12210 + }, + { + "epoch": 0.9790142510673268, + "grad_norm": 1.757306456565857, + "learning_rate": 1.0824227008705816e-05, + "loss": 3.2377, + "step": 12211 + }, + { + "epoch": 0.9790944258483494, + "grad_norm": 1.6674439907073975, + "learning_rate": 1.0822932972722034e-05, + "loss": 3.1758, + "step": 12212 + }, + { + "epoch": 0.979174600629372, + "grad_norm": 1.9580448865890503, + "learning_rate": 1.082163892286389e-05, + "loss": 3.2607, + "step": 12213 + }, + { + "epoch": 0.9792547754103946, + "grad_norm": 1.668577790260315, + "learning_rate": 1.0820344859153207e-05, + "loss": 3.0795, + "step": 12214 + }, + { + "epoch": 0.9793349501914173, + "grad_norm": 1.7496510744094849, + "learning_rate": 1.0819050781611794e-05, + "loss": 3.1929, + "step": 12215 + }, + { + "epoch": 0.9794151249724399, + "grad_norm": 1.8517097234725952, + "learning_rate": 1.0817756690261477e-05, + "loss": 3.1771, + "step": 12216 + }, + { + "epoch": 0.9794952997534625, + "grad_norm": 1.907670021057129, + "learning_rate": 1.081646258512407e-05, + "loss": 3.2187, + "step": 12217 + }, + { + "epoch": 0.9795754745344851, + "grad_norm": 1.8000141382217407, + "learning_rate": 1.0815168466221388e-05, + "loss": 3.239, + "step": 12218 + }, + { + "epoch": 0.9796556493155079, + "grad_norm": 1.7742458581924438, + "learning_rate": 1.0813874333575258e-05, + "loss": 3.2443, + "step": 12219 + }, + { + "epoch": 0.9797358240965305, + "grad_norm": 1.7669907808303833, + "learning_rate": 1.081258018720749e-05, + "loss": 3.1979, + "step": 12220 + }, + { + "epoch": 0.9798159988775531, + "grad_norm": 1.700301170349121, + "learning_rate": 1.0811286027139909e-05, + "loss": 3.175, + "step": 12221 + }, + { + "epoch": 0.9798961736585757, + "grad_norm": 1.7816468477249146, + "learning_rate": 1.0809991853394331e-05, + "loss": 3.084, + "step": 12222 + }, + { + "epoch": 0.9799763484395984, + "grad_norm": 1.7246943712234497, + "learning_rate": 1.0808697665992574e-05, + "loss": 3.2531, + "step": 12223 + }, + { + "epoch": 0.980056523220621, + "grad_norm": 1.7509477138519287, + "learning_rate": 1.080740346495646e-05, + "loss": 3.1769, + "step": 12224 + }, + { + "epoch": 0.9801366980016436, + "grad_norm": 1.78950035572052, + "learning_rate": 1.080610925030781e-05, + "loss": 3.1893, + "step": 12225 + }, + { + "epoch": 0.9802168727826662, + "grad_norm": 1.8032442331314087, + "learning_rate": 1.080481502206844e-05, + "loss": 3.2606, + "step": 12226 + }, + { + "epoch": 0.9802970475636889, + "grad_norm": 1.7509700059890747, + "learning_rate": 1.0803520780260173e-05, + "loss": 3.129, + "step": 12227 + }, + { + "epoch": 0.9803772223447115, + "grad_norm": 1.7761386632919312, + "learning_rate": 1.0802226524904829e-05, + "loss": 3.2123, + "step": 12228 + }, + { + "epoch": 0.9804573971257341, + "grad_norm": 1.723331332206726, + "learning_rate": 1.080093225602423e-05, + "loss": 3.1106, + "step": 12229 + }, + { + "epoch": 0.9805375719067567, + "grad_norm": 1.7358787059783936, + "learning_rate": 1.0799637973640194e-05, + "loss": 3.1437, + "step": 12230 + }, + { + "epoch": 0.9806177466877793, + "grad_norm": 1.8113417625427246, + "learning_rate": 1.0798343677774542e-05, + "loss": 3.1458, + "step": 12231 + }, + { + "epoch": 0.980697921468802, + "grad_norm": 1.7536107301712036, + "learning_rate": 1.0797049368449099e-05, + "loss": 3.2061, + "step": 12232 + }, + { + "epoch": 0.9807780962498246, + "grad_norm": 1.7600300312042236, + "learning_rate": 1.0795755045685684e-05, + "loss": 3.0905, + "step": 12233 + }, + { + "epoch": 0.9808582710308472, + "grad_norm": 1.8032495975494385, + "learning_rate": 1.079446070950612e-05, + "loss": 3.1409, + "step": 12234 + }, + { + "epoch": 0.9809384458118698, + "grad_norm": 1.726945400238037, + "learning_rate": 1.0793166359932225e-05, + "loss": 3.14, + "step": 12235 + }, + { + "epoch": 0.9810186205928925, + "grad_norm": 0.7235603332519531, + "learning_rate": 1.0791871996985827e-05, + "loss": 2.5026, + "step": 12236 + }, + { + "epoch": 0.9810987953739151, + "grad_norm": 1.7819130420684814, + "learning_rate": 1.0790577620688749e-05, + "loss": 3.2603, + "step": 12237 + }, + { + "epoch": 0.9811789701549377, + "grad_norm": 1.900199055671692, + "learning_rate": 1.0789283231062808e-05, + "loss": 3.1232, + "step": 12238 + }, + { + "epoch": 0.9812591449359604, + "grad_norm": 1.8459550142288208, + "learning_rate": 1.0787988828129831e-05, + "loss": 3.054, + "step": 12239 + }, + { + "epoch": 0.9813393197169831, + "grad_norm": 1.7818905115127563, + "learning_rate": 1.0786694411911641e-05, + "loss": 3.2374, + "step": 12240 + }, + { + "epoch": 0.9814194944980057, + "grad_norm": 1.7693582773208618, + "learning_rate": 1.0785399982430058e-05, + "loss": 3.1877, + "step": 12241 + }, + { + "epoch": 0.9814996692790283, + "grad_norm": 1.9081944227218628, + "learning_rate": 1.0784105539706911e-05, + "loss": 3.2352, + "step": 12242 + }, + { + "epoch": 0.9815798440600509, + "grad_norm": 1.7583807706832886, + "learning_rate": 1.0782811083764018e-05, + "loss": 3.2147, + "step": 12243 + }, + { + "epoch": 0.9816600188410736, + "grad_norm": 1.7979384660720825, + "learning_rate": 1.0781516614623206e-05, + "loss": 3.248, + "step": 12244 + }, + { + "epoch": 0.9817401936220962, + "grad_norm": 1.695688009262085, + "learning_rate": 1.0780222132306301e-05, + "loss": 3.1092, + "step": 12245 + }, + { + "epoch": 0.9818203684031188, + "grad_norm": 1.7109042406082153, + "learning_rate": 1.0778927636835123e-05, + "loss": 3.1889, + "step": 12246 + }, + { + "epoch": 0.9819005431841414, + "grad_norm": 1.7911713123321533, + "learning_rate": 1.0777633128231503e-05, + "loss": 3.2167, + "step": 12247 + }, + { + "epoch": 0.9819807179651641, + "grad_norm": 1.7894071340560913, + "learning_rate": 1.0776338606517262e-05, + "loss": 3.1969, + "step": 12248 + }, + { + "epoch": 0.9820608927461867, + "grad_norm": 1.737453818321228, + "learning_rate": 1.0775044071714227e-05, + "loss": 3.2283, + "step": 12249 + }, + { + "epoch": 0.9821410675272093, + "grad_norm": 1.8206865787506104, + "learning_rate": 1.0773749523844218e-05, + "loss": 3.1721, + "step": 12250 + }, + { + "epoch": 0.9822212423082319, + "grad_norm": 1.8115694522857666, + "learning_rate": 1.0772454962929067e-05, + "loss": 3.3079, + "step": 12251 + }, + { + "epoch": 0.9823014170892546, + "grad_norm": 1.7628942728042603, + "learning_rate": 1.0771160388990598e-05, + "loss": 3.1807, + "step": 12252 + }, + { + "epoch": 0.9823815918702772, + "grad_norm": 1.5953835248947144, + "learning_rate": 1.0769865802050637e-05, + "loss": 3.1214, + "step": 12253 + }, + { + "epoch": 0.9824617666512998, + "grad_norm": 1.81913423538208, + "learning_rate": 1.0768571202131009e-05, + "loss": 3.1615, + "step": 12254 + }, + { + "epoch": 0.9825419414323224, + "grad_norm": 1.7361493110656738, + "learning_rate": 1.076727658925354e-05, + "loss": 3.1816, + "step": 12255 + }, + { + "epoch": 0.982622116213345, + "grad_norm": 1.7142977714538574, + "learning_rate": 1.076598196344006e-05, + "loss": 3.2433, + "step": 12256 + }, + { + "epoch": 0.9827022909943677, + "grad_norm": 1.7513933181762695, + "learning_rate": 1.0764687324712394e-05, + "loss": 3.156, + "step": 12257 + }, + { + "epoch": 0.9827824657753903, + "grad_norm": 1.8028819561004639, + "learning_rate": 1.0763392673092367e-05, + "loss": 3.1428, + "step": 12258 + }, + { + "epoch": 0.9828626405564129, + "grad_norm": 1.6956349611282349, + "learning_rate": 1.0762098008601814e-05, + "loss": 3.1846, + "step": 12259 + }, + { + "epoch": 0.9829428153374357, + "grad_norm": 1.8795273303985596, + "learning_rate": 1.0760803331262553e-05, + "loss": 3.1765, + "step": 12260 + }, + { + "epoch": 0.9830229901184583, + "grad_norm": 0.7327881455421448, + "learning_rate": 1.0759508641096415e-05, + "loss": 2.4943, + "step": 12261 + }, + { + "epoch": 0.9831031648994809, + "grad_norm": 1.691830039024353, + "learning_rate": 1.075821393812523e-05, + "loss": 3.1019, + "step": 12262 + }, + { + "epoch": 0.9831833396805035, + "grad_norm": 0.7089197039604187, + "learning_rate": 1.0756919222370825e-05, + "loss": 2.4855, + "step": 12263 + }, + { + "epoch": 0.9832635144615262, + "grad_norm": 0.6846185326576233, + "learning_rate": 1.0755624493855033e-05, + "loss": 2.4736, + "step": 12264 + }, + { + "epoch": 0.9833436892425488, + "grad_norm": 1.7792359590530396, + "learning_rate": 1.0754329752599672e-05, + "loss": 3.1691, + "step": 12265 + }, + { + "epoch": 0.9834238640235714, + "grad_norm": 1.7527098655700684, + "learning_rate": 1.075303499862658e-05, + "loss": 3.0778, + "step": 12266 + }, + { + "epoch": 0.983504038804594, + "grad_norm": 1.8139500617980957, + "learning_rate": 1.0751740231957584e-05, + "loss": 3.2128, + "step": 12267 + }, + { + "epoch": 0.9835842135856167, + "grad_norm": 1.7297520637512207, + "learning_rate": 1.0750445452614511e-05, + "loss": 3.1435, + "step": 12268 + }, + { + "epoch": 0.9836643883666393, + "grad_norm": 1.8056625127792358, + "learning_rate": 1.0749150660619193e-05, + "loss": 3.2371, + "step": 12269 + }, + { + "epoch": 0.9837445631476619, + "grad_norm": 1.7976739406585693, + "learning_rate": 1.074785585599346e-05, + "loss": 3.317, + "step": 12270 + }, + { + "epoch": 0.9838247379286845, + "grad_norm": 1.823625922203064, + "learning_rate": 1.0746561038759138e-05, + "loss": 3.1206, + "step": 12271 + }, + { + "epoch": 0.9839049127097071, + "grad_norm": 1.7643029689788818, + "learning_rate": 1.0745266208938063e-05, + "loss": 3.3481, + "step": 12272 + }, + { + "epoch": 0.9839850874907298, + "grad_norm": 1.7743574380874634, + "learning_rate": 1.0743971366552061e-05, + "loss": 3.0707, + "step": 12273 + }, + { + "epoch": 0.9840652622717524, + "grad_norm": 1.7907569408416748, + "learning_rate": 1.0742676511622964e-05, + "loss": 3.182, + "step": 12274 + }, + { + "epoch": 0.984145437052775, + "grad_norm": 1.8008109331130981, + "learning_rate": 1.0741381644172603e-05, + "loss": 3.1397, + "step": 12275 + }, + { + "epoch": 0.9842256118337976, + "grad_norm": 1.7283660173416138, + "learning_rate": 1.0740086764222808e-05, + "loss": 3.163, + "step": 12276 + }, + { + "epoch": 0.9843057866148203, + "grad_norm": 1.7484512329101562, + "learning_rate": 1.0738791871795413e-05, + "loss": 3.1787, + "step": 12277 + }, + { + "epoch": 0.9843859613958429, + "grad_norm": 1.9319121837615967, + "learning_rate": 1.0737496966912247e-05, + "loss": 3.2028, + "step": 12278 + }, + { + "epoch": 0.9844661361768655, + "grad_norm": 1.6793878078460693, + "learning_rate": 1.073620204959514e-05, + "loss": 3.1915, + "step": 12279 + }, + { + "epoch": 0.9845463109578882, + "grad_norm": 1.894957184791565, + "learning_rate": 1.0734907119865926e-05, + "loss": 3.2561, + "step": 12280 + }, + { + "epoch": 0.9846264857389109, + "grad_norm": 1.739600419998169, + "learning_rate": 1.0733612177746438e-05, + "loss": 3.2687, + "step": 12281 + }, + { + "epoch": 0.9847066605199335, + "grad_norm": 0.7192650437355042, + "learning_rate": 1.0732317223258509e-05, + "loss": 2.4694, + "step": 12282 + }, + { + "epoch": 0.9847868353009561, + "grad_norm": 1.6850091218948364, + "learning_rate": 1.0731022256423966e-05, + "loss": 3.0786, + "step": 12283 + }, + { + "epoch": 0.9848670100819787, + "grad_norm": 1.8192194700241089, + "learning_rate": 1.072972727726465e-05, + "loss": 3.2672, + "step": 12284 + }, + { + "epoch": 0.9849471848630014, + "grad_norm": 1.7387326955795288, + "learning_rate": 1.0728432285802386e-05, + "loss": 3.1604, + "step": 12285 + }, + { + "epoch": 0.985027359644024, + "grad_norm": 1.7054595947265625, + "learning_rate": 1.0727137282059007e-05, + "loss": 3.1432, + "step": 12286 + }, + { + "epoch": 0.9851075344250466, + "grad_norm": 2.0528953075408936, + "learning_rate": 1.0725842266056357e-05, + "loss": 3.2759, + "step": 12287 + }, + { + "epoch": 0.9851877092060692, + "grad_norm": 1.7024376392364502, + "learning_rate": 1.0724547237816259e-05, + "loss": 3.2109, + "step": 12288 + }, + { + "epoch": 0.9852678839870919, + "grad_norm": 0.7265363335609436, + "learning_rate": 1.072325219736055e-05, + "loss": 2.4901, + "step": 12289 + }, + { + "epoch": 0.9853480587681145, + "grad_norm": 1.8520561456680298, + "learning_rate": 1.0721957144711063e-05, + "loss": 3.0852, + "step": 12290 + }, + { + "epoch": 0.9854282335491371, + "grad_norm": 1.779965877532959, + "learning_rate": 1.0720662079889632e-05, + "loss": 3.179, + "step": 12291 + }, + { + "epoch": 0.9855084083301597, + "grad_norm": 1.7970777750015259, + "learning_rate": 1.0719367002918095e-05, + "loss": 3.1963, + "step": 12292 + }, + { + "epoch": 0.9855885831111824, + "grad_norm": 1.8303873538970947, + "learning_rate": 1.0718071913818284e-05, + "loss": 3.2726, + "step": 12293 + }, + { + "epoch": 0.985668757892205, + "grad_norm": 1.7319893836975098, + "learning_rate": 1.0716776812612034e-05, + "loss": 3.1549, + "step": 12294 + }, + { + "epoch": 0.9857489326732276, + "grad_norm": 1.8601046800613403, + "learning_rate": 1.0715481699321175e-05, + "loss": 3.1786, + "step": 12295 + }, + { + "epoch": 0.9858291074542502, + "grad_norm": 1.7867108583450317, + "learning_rate": 1.0714186573967548e-05, + "loss": 3.2213, + "step": 12296 + }, + { + "epoch": 0.9859092822352729, + "grad_norm": 1.814540982246399, + "learning_rate": 1.0712891436572993e-05, + "loss": 3.2327, + "step": 12297 + }, + { + "epoch": 0.9859894570162955, + "grad_norm": 1.7975056171417236, + "learning_rate": 1.0711596287159334e-05, + "loss": 3.1559, + "step": 12298 + }, + { + "epoch": 0.9860696317973181, + "grad_norm": 1.7326886653900146, + "learning_rate": 1.0710301125748415e-05, + "loss": 3.2662, + "step": 12299 + }, + { + "epoch": 0.9861498065783408, + "grad_norm": 1.7233790159225464, + "learning_rate": 1.0709005952362066e-05, + "loss": 3.186, + "step": 12300 + }, + { + "epoch": 0.9862299813593635, + "grad_norm": 1.7271792888641357, + "learning_rate": 1.0707710767022129e-05, + "loss": 3.3174, + "step": 12301 + }, + { + "epoch": 0.9863101561403861, + "grad_norm": 1.796683430671692, + "learning_rate": 1.070641556975044e-05, + "loss": 3.2522, + "step": 12302 + }, + { + "epoch": 0.9863903309214087, + "grad_norm": 1.9197691679000854, + "learning_rate": 1.0705120360568829e-05, + "loss": 3.2179, + "step": 12303 + }, + { + "epoch": 0.9864705057024313, + "grad_norm": 1.880028247833252, + "learning_rate": 1.070382513949914e-05, + "loss": 3.2341, + "step": 12304 + }, + { + "epoch": 0.986550680483454, + "grad_norm": 1.6693634986877441, + "learning_rate": 1.0702529906563206e-05, + "loss": 3.2025, + "step": 12305 + }, + { + "epoch": 0.9866308552644766, + "grad_norm": 1.6992087364196777, + "learning_rate": 1.0701234661782864e-05, + "loss": 3.1243, + "step": 12306 + }, + { + "epoch": 0.9867110300454992, + "grad_norm": 1.8362319469451904, + "learning_rate": 1.0699939405179958e-05, + "loss": 3.197, + "step": 12307 + }, + { + "epoch": 0.9867912048265218, + "grad_norm": 1.83077871799469, + "learning_rate": 1.0698644136776315e-05, + "loss": 3.245, + "step": 12308 + }, + { + "epoch": 0.9868713796075445, + "grad_norm": 1.79458749294281, + "learning_rate": 1.069734885659378e-05, + "loss": 3.3392, + "step": 12309 + }, + { + "epoch": 0.9869515543885671, + "grad_norm": 1.862210750579834, + "learning_rate": 1.069605356465419e-05, + "loss": 3.272, + "step": 12310 + }, + { + "epoch": 0.9870317291695897, + "grad_norm": 1.834409236907959, + "learning_rate": 1.0694758260979379e-05, + "loss": 3.153, + "step": 12311 + }, + { + "epoch": 0.9871119039506123, + "grad_norm": 1.7680473327636719, + "learning_rate": 1.0693462945591193e-05, + "loss": 3.1754, + "step": 12312 + }, + { + "epoch": 0.987192078731635, + "grad_norm": 1.7389191389083862, + "learning_rate": 1.0692167618511461e-05, + "loss": 3.165, + "step": 12313 + }, + { + "epoch": 0.9872722535126576, + "grad_norm": 1.8299918174743652, + "learning_rate": 1.0690872279762033e-05, + "loss": 3.1814, + "step": 12314 + }, + { + "epoch": 0.9873524282936802, + "grad_norm": 1.905542016029358, + "learning_rate": 1.0689576929364737e-05, + "loss": 3.2601, + "step": 12315 + }, + { + "epoch": 0.9874326030747028, + "grad_norm": 1.7342264652252197, + "learning_rate": 1.0688281567341418e-05, + "loss": 3.1869, + "step": 12316 + }, + { + "epoch": 0.9875127778557254, + "grad_norm": 1.8241307735443115, + "learning_rate": 1.0686986193713917e-05, + "loss": 3.1731, + "step": 12317 + }, + { + "epoch": 0.9875929526367481, + "grad_norm": 1.9246410131454468, + "learning_rate": 1.0685690808504067e-05, + "loss": 3.2028, + "step": 12318 + }, + { + "epoch": 0.9876731274177707, + "grad_norm": 1.70782470703125, + "learning_rate": 1.0684395411733715e-05, + "loss": 3.2028, + "step": 12319 + }, + { + "epoch": 0.9877533021987933, + "grad_norm": 1.7683080434799194, + "learning_rate": 1.0683100003424696e-05, + "loss": 3.2704, + "step": 12320 + }, + { + "epoch": 0.987833476979816, + "grad_norm": 1.8196415901184082, + "learning_rate": 1.068180458359885e-05, + "loss": 3.3656, + "step": 12321 + }, + { + "epoch": 0.9879136517608387, + "grad_norm": 1.8263970613479614, + "learning_rate": 1.0680509152278023e-05, + "loss": 3.2489, + "step": 12322 + }, + { + "epoch": 0.9879938265418613, + "grad_norm": 1.7809396982192993, + "learning_rate": 1.067921370948405e-05, + "loss": 3.2279, + "step": 12323 + }, + { + "epoch": 0.9880740013228839, + "grad_norm": 1.719454288482666, + "learning_rate": 1.0677918255238773e-05, + "loss": 3.1609, + "step": 12324 + }, + { + "epoch": 0.9881541761039065, + "grad_norm": 1.7978264093399048, + "learning_rate": 1.067662278956403e-05, + "loss": 3.1834, + "step": 12325 + }, + { + "epoch": 0.9882343508849292, + "grad_norm": 1.7889679670333862, + "learning_rate": 1.0675327312481667e-05, + "loss": 3.0998, + "step": 12326 + }, + { + "epoch": 0.9883145256659518, + "grad_norm": 1.7790417671203613, + "learning_rate": 1.0674031824013527e-05, + "loss": 3.1364, + "step": 12327 + }, + { + "epoch": 0.9883947004469744, + "grad_norm": 1.8237037658691406, + "learning_rate": 1.0672736324181442e-05, + "loss": 3.2202, + "step": 12328 + }, + { + "epoch": 0.988474875227997, + "grad_norm": 1.7643910646438599, + "learning_rate": 1.0671440813007265e-05, + "loss": 3.2957, + "step": 12329 + }, + { + "epoch": 0.9885550500090197, + "grad_norm": 1.7408345937728882, + "learning_rate": 1.0670145290512832e-05, + "loss": 3.1508, + "step": 12330 + }, + { + "epoch": 0.9886352247900423, + "grad_norm": 1.7431285381317139, + "learning_rate": 1.0668849756719983e-05, + "loss": 3.2176, + "step": 12331 + }, + { + "epoch": 0.9887153995710649, + "grad_norm": 1.7028888463974, + "learning_rate": 1.0667554211650566e-05, + "loss": 3.1942, + "step": 12332 + }, + { + "epoch": 0.9887955743520875, + "grad_norm": 1.7936289310455322, + "learning_rate": 1.0666258655326416e-05, + "loss": 3.2383, + "step": 12333 + }, + { + "epoch": 0.9888757491331102, + "grad_norm": 1.6685830354690552, + "learning_rate": 1.0664963087769384e-05, + "loss": 3.141, + "step": 12334 + }, + { + "epoch": 0.9889559239141328, + "grad_norm": 1.8009227514266968, + "learning_rate": 1.0663667509001306e-05, + "loss": 3.1899, + "step": 12335 + }, + { + "epoch": 0.9890360986951554, + "grad_norm": 1.696234941482544, + "learning_rate": 1.0662371919044028e-05, + "loss": 3.1726, + "step": 12336 + }, + { + "epoch": 0.989116273476178, + "grad_norm": 1.6310230493545532, + "learning_rate": 1.0661076317919395e-05, + "loss": 3.1649, + "step": 12337 + }, + { + "epoch": 0.9891964482572007, + "grad_norm": 1.7752230167388916, + "learning_rate": 1.0659780705649245e-05, + "loss": 3.1682, + "step": 12338 + }, + { + "epoch": 0.9892766230382233, + "grad_norm": 1.7647260427474976, + "learning_rate": 1.0658485082255427e-05, + "loss": 3.1184, + "step": 12339 + }, + { + "epoch": 0.9893567978192459, + "grad_norm": 1.9874358177185059, + "learning_rate": 1.0657189447759783e-05, + "loss": 3.2825, + "step": 12340 + }, + { + "epoch": 0.9894369726002686, + "grad_norm": 1.8327046632766724, + "learning_rate": 1.0655893802184154e-05, + "loss": 3.1278, + "step": 12341 + }, + { + "epoch": 0.9895171473812913, + "grad_norm": 1.8004186153411865, + "learning_rate": 1.065459814555039e-05, + "loss": 3.1827, + "step": 12342 + }, + { + "epoch": 0.9895973221623139, + "grad_norm": 1.9427320957183838, + "learning_rate": 1.065330247788033e-05, + "loss": 3.2594, + "step": 12343 + }, + { + "epoch": 0.9896774969433365, + "grad_norm": 1.8029834032058716, + "learning_rate": 1.0652006799195822e-05, + "loss": 3.191, + "step": 12344 + }, + { + "epoch": 0.9897576717243591, + "grad_norm": 1.891775369644165, + "learning_rate": 1.0650711109518708e-05, + "loss": 3.2142, + "step": 12345 + }, + { + "epoch": 0.9898378465053818, + "grad_norm": 1.8333412408828735, + "learning_rate": 1.0649415408870831e-05, + "loss": 3.0985, + "step": 12346 + }, + { + "epoch": 0.9899180212864044, + "grad_norm": 1.777686357498169, + "learning_rate": 1.0648119697274043e-05, + "loss": 3.0776, + "step": 12347 + }, + { + "epoch": 0.989998196067427, + "grad_norm": 1.7597092390060425, + "learning_rate": 1.0646823974750185e-05, + "loss": 3.224, + "step": 12348 + }, + { + "epoch": 0.9900783708484496, + "grad_norm": 1.7006267309188843, + "learning_rate": 1.0645528241321103e-05, + "loss": 3.1699, + "step": 12349 + }, + { + "epoch": 0.9901585456294723, + "grad_norm": 1.7543140649795532, + "learning_rate": 1.0644232497008642e-05, + "loss": 3.2638, + "step": 12350 + }, + { + "epoch": 0.9902387204104949, + "grad_norm": 1.740384578704834, + "learning_rate": 1.0642936741834646e-05, + "loss": 3.2781, + "step": 12351 + }, + { + "epoch": 0.9903188951915175, + "grad_norm": 1.6973915100097656, + "learning_rate": 1.0641640975820965e-05, + "loss": 3.1588, + "step": 12352 + }, + { + "epoch": 0.9903990699725401, + "grad_norm": 1.8284413814544678, + "learning_rate": 1.0640345198989444e-05, + "loss": 3.3232, + "step": 12353 + }, + { + "epoch": 0.9904792447535627, + "grad_norm": 1.8314522504806519, + "learning_rate": 1.0639049411361929e-05, + "loss": 3.1697, + "step": 12354 + }, + { + "epoch": 0.9905594195345854, + "grad_norm": 1.8672317266464233, + "learning_rate": 1.0637753612960263e-05, + "loss": 3.1687, + "step": 12355 + }, + { + "epoch": 0.990639594315608, + "grad_norm": 1.8188748359680176, + "learning_rate": 1.0636457803806298e-05, + "loss": 3.2291, + "step": 12356 + }, + { + "epoch": 0.9907197690966306, + "grad_norm": 1.8538230657577515, + "learning_rate": 1.0635161983921879e-05, + "loss": 3.1923, + "step": 12357 + }, + { + "epoch": 0.9907999438776532, + "grad_norm": 1.7187788486480713, + "learning_rate": 1.063386615332885e-05, + "loss": 3.1877, + "step": 12358 + }, + { + "epoch": 0.9908801186586759, + "grad_norm": 1.82969069480896, + "learning_rate": 1.0632570312049064e-05, + "loss": 3.2736, + "step": 12359 + }, + { + "epoch": 0.9909602934396985, + "grad_norm": 1.8439801931381226, + "learning_rate": 1.0631274460104367e-05, + "loss": 3.2207, + "step": 12360 + }, + { + "epoch": 0.9910404682207212, + "grad_norm": 1.7286072969436646, + "learning_rate": 1.0629978597516599e-05, + "loss": 3.2893, + "step": 12361 + }, + { + "epoch": 0.9911206430017439, + "grad_norm": 1.8081616163253784, + "learning_rate": 1.062868272430762e-05, + "loss": 3.1808, + "step": 12362 + }, + { + "epoch": 0.9912008177827665, + "grad_norm": 1.8570605516433716, + "learning_rate": 1.062738684049927e-05, + "loss": 3.0745, + "step": 12363 + }, + { + "epoch": 0.9912809925637891, + "grad_norm": 1.8469470739364624, + "learning_rate": 1.0626090946113397e-05, + "loss": 3.2598, + "step": 12364 + }, + { + "epoch": 0.9913611673448117, + "grad_norm": 1.867645502090454, + "learning_rate": 1.0624795041171854e-05, + "loss": 3.2374, + "step": 12365 + }, + { + "epoch": 0.9914413421258343, + "grad_norm": 1.8159774541854858, + "learning_rate": 1.0623499125696484e-05, + "loss": 3.255, + "step": 12366 + }, + { + "epoch": 0.991521516906857, + "grad_norm": 2.007291555404663, + "learning_rate": 1.0622203199709142e-05, + "loss": 3.1146, + "step": 12367 + }, + { + "epoch": 0.9916016916878796, + "grad_norm": 0.7269201874732971, + "learning_rate": 1.062090726323167e-05, + "loss": 2.4884, + "step": 12368 + }, + { + "epoch": 0.9916818664689022, + "grad_norm": 1.6640946865081787, + "learning_rate": 1.061961131628592e-05, + "loss": 3.1356, + "step": 12369 + }, + { + "epoch": 0.9917620412499248, + "grad_norm": 1.7547824382781982, + "learning_rate": 1.0618315358893746e-05, + "loss": 3.1329, + "step": 12370 + }, + { + "epoch": 0.9918422160309475, + "grad_norm": 1.6914869546890259, + "learning_rate": 1.0617019391076991e-05, + "loss": 3.0796, + "step": 12371 + }, + { + "epoch": 0.9919223908119701, + "grad_norm": 1.7321118116378784, + "learning_rate": 1.0615723412857506e-05, + "loss": 3.1008, + "step": 12372 + }, + { + "epoch": 0.9920025655929927, + "grad_norm": 1.7608650922775269, + "learning_rate": 1.0614427424257138e-05, + "loss": 3.109, + "step": 12373 + }, + { + "epoch": 0.9920827403740153, + "grad_norm": 1.8016927242279053, + "learning_rate": 1.0613131425297745e-05, + "loss": 3.249, + "step": 12374 + }, + { + "epoch": 0.992162915155038, + "grad_norm": 1.6994390487670898, + "learning_rate": 1.061183541600117e-05, + "loss": 3.1802, + "step": 12375 + }, + { + "epoch": 0.9922430899360606, + "grad_norm": 1.842885136604309, + "learning_rate": 1.0610539396389265e-05, + "loss": 3.2001, + "step": 12376 + }, + { + "epoch": 0.9923232647170832, + "grad_norm": 1.759879231452942, + "learning_rate": 1.0609243366483886e-05, + "loss": 3.1679, + "step": 12377 + }, + { + "epoch": 0.9924034394981058, + "grad_norm": 1.759568214416504, + "learning_rate": 1.0607947326306874e-05, + "loss": 3.2455, + "step": 12378 + }, + { + "epoch": 0.9924836142791285, + "grad_norm": 1.8023556470870972, + "learning_rate": 1.0606651275880082e-05, + "loss": 3.1874, + "step": 12379 + }, + { + "epoch": 0.9925637890601511, + "grad_norm": 1.7750381231307983, + "learning_rate": 1.0605355215225368e-05, + "loss": 3.1817, + "step": 12380 + }, + { + "epoch": 0.9926439638411737, + "grad_norm": 1.745141625404358, + "learning_rate": 1.0604059144364576e-05, + "loss": 3.1396, + "step": 12381 + }, + { + "epoch": 0.9927241386221964, + "grad_norm": 1.8433656692504883, + "learning_rate": 1.0602763063319562e-05, + "loss": 3.2275, + "step": 12382 + }, + { + "epoch": 0.9928043134032191, + "grad_norm": 1.7576229572296143, + "learning_rate": 1.0601466972112174e-05, + "loss": 3.2744, + "step": 12383 + }, + { + "epoch": 0.9928844881842417, + "grad_norm": 1.6808013916015625, + "learning_rate": 1.0600170870764261e-05, + "loss": 3.1838, + "step": 12384 + }, + { + "epoch": 0.9929646629652643, + "grad_norm": 1.8636096715927124, + "learning_rate": 1.0598874759297685e-05, + "loss": 3.1743, + "step": 12385 + }, + { + "epoch": 0.9930448377462869, + "grad_norm": 0.7158401608467102, + "learning_rate": 1.0597578637734287e-05, + "loss": 2.4825, + "step": 12386 + }, + { + "epoch": 0.9931250125273096, + "grad_norm": 0.6986439824104309, + "learning_rate": 1.0596282506095928e-05, + "loss": 2.4828, + "step": 12387 + }, + { + "epoch": 0.9932051873083322, + "grad_norm": 1.7896631956100464, + "learning_rate": 1.0594986364404451e-05, + "loss": 3.0825, + "step": 12388 + }, + { + "epoch": 0.9932853620893548, + "grad_norm": 1.7995209693908691, + "learning_rate": 1.0593690212681715e-05, + "loss": 3.214, + "step": 12389 + }, + { + "epoch": 0.9933655368703774, + "grad_norm": 1.9084054231643677, + "learning_rate": 1.0592394050949575e-05, + "loss": 3.1738, + "step": 12390 + }, + { + "epoch": 0.9934457116514, + "grad_norm": 1.7449871301651, + "learning_rate": 1.0591097879229877e-05, + "loss": 3.134, + "step": 12391 + }, + { + "epoch": 0.9935258864324227, + "grad_norm": 1.9972119331359863, + "learning_rate": 1.0589801697544477e-05, + "loss": 3.194, + "step": 12392 + }, + { + "epoch": 0.9936060612134453, + "grad_norm": 1.7508636713027954, + "learning_rate": 1.0588505505915228e-05, + "loss": 3.1743, + "step": 12393 + }, + { + "epoch": 0.9936862359944679, + "grad_norm": 1.8618322610855103, + "learning_rate": 1.0587209304363984e-05, + "loss": 3.1849, + "step": 12394 + }, + { + "epoch": 0.9937664107754905, + "grad_norm": 1.6904529333114624, + "learning_rate": 1.0585913092912601e-05, + "loss": 3.1, + "step": 12395 + }, + { + "epoch": 0.9938465855565132, + "grad_norm": 1.689945936203003, + "learning_rate": 1.0584616871582924e-05, + "loss": 3.1567, + "step": 12396 + }, + { + "epoch": 0.9939267603375358, + "grad_norm": 1.6861660480499268, + "learning_rate": 1.0583320640396819e-05, + "loss": 3.1951, + "step": 12397 + }, + { + "epoch": 0.9940069351185584, + "grad_norm": 1.7612229585647583, + "learning_rate": 1.058202439937613e-05, + "loss": 3.1155, + "step": 12398 + }, + { + "epoch": 0.994087109899581, + "grad_norm": 1.7413759231567383, + "learning_rate": 1.0580728148542715e-05, + "loss": 3.2256, + "step": 12399 + }, + { + "epoch": 0.9941672846806037, + "grad_norm": 1.787842035293579, + "learning_rate": 1.0579431887918427e-05, + "loss": 3.1831, + "step": 12400 + }, + { + "epoch": 0.9942474594616263, + "grad_norm": 1.7791814804077148, + "learning_rate": 1.0578135617525122e-05, + "loss": 3.0964, + "step": 12401 + }, + { + "epoch": 0.994327634242649, + "grad_norm": 1.8296152353286743, + "learning_rate": 1.0576839337384656e-05, + "loss": 3.2017, + "step": 12402 + }, + { + "epoch": 0.9944078090236717, + "grad_norm": 1.769852638244629, + "learning_rate": 1.0575543047518882e-05, + "loss": 3.2181, + "step": 12403 + }, + { + "epoch": 0.9944879838046943, + "grad_norm": 1.999640703201294, + "learning_rate": 1.0574246747949652e-05, + "loss": 3.1137, + "step": 12404 + }, + { + "epoch": 0.9945681585857169, + "grad_norm": 2.033087968826294, + "learning_rate": 1.0572950438698828e-05, + "loss": 3.2239, + "step": 12405 + }, + { + "epoch": 0.9946483333667395, + "grad_norm": 1.7217803001403809, + "learning_rate": 1.057165411978826e-05, + "loss": 3.2001, + "step": 12406 + }, + { + "epoch": 0.9947285081477621, + "grad_norm": 1.8107707500457764, + "learning_rate": 1.0570357791239806e-05, + "loss": 3.1072, + "step": 12407 + }, + { + "epoch": 0.9948086829287848, + "grad_norm": 1.83070969581604, + "learning_rate": 1.0569061453075318e-05, + "loss": 3.1489, + "step": 12408 + }, + { + "epoch": 0.9948888577098074, + "grad_norm": 1.6823102235794067, + "learning_rate": 1.0567765105316655e-05, + "loss": 3.2369, + "step": 12409 + }, + { + "epoch": 0.99496903249083, + "grad_norm": 0.6981968283653259, + "learning_rate": 1.0566468747985673e-05, + "loss": 2.4829, + "step": 12410 + }, + { + "epoch": 0.9950492072718526, + "grad_norm": 1.8487554788589478, + "learning_rate": 1.056517238110423e-05, + "loss": 3.1893, + "step": 12411 + }, + { + "epoch": 0.9951293820528753, + "grad_norm": 1.8582813739776611, + "learning_rate": 1.0563876004694179e-05, + "loss": 3.2185, + "step": 12412 + }, + { + "epoch": 0.9952095568338979, + "grad_norm": 1.7661012411117554, + "learning_rate": 1.0562579618777374e-05, + "loss": 3.131, + "step": 12413 + }, + { + "epoch": 0.9952897316149205, + "grad_norm": 1.7225631475448608, + "learning_rate": 1.0561283223375675e-05, + "loss": 3.1656, + "step": 12414 + }, + { + "epoch": 0.9953699063959431, + "grad_norm": 1.9102274179458618, + "learning_rate": 1.0559986818510943e-05, + "loss": 3.2275, + "step": 12415 + }, + { + "epoch": 0.9954500811769658, + "grad_norm": 1.7379114627838135, + "learning_rate": 1.0558690404205027e-05, + "loss": 3.2549, + "step": 12416 + }, + { + "epoch": 0.9955302559579884, + "grad_norm": 1.7288522720336914, + "learning_rate": 1.055739398047979e-05, + "loss": 3.2422, + "step": 12417 + }, + { + "epoch": 0.995610430739011, + "grad_norm": 1.7367607355117798, + "learning_rate": 1.0556097547357086e-05, + "loss": 3.24, + "step": 12418 + }, + { + "epoch": 0.9956906055200336, + "grad_norm": 1.720834493637085, + "learning_rate": 1.0554801104858771e-05, + "loss": 3.0604, + "step": 12419 + }, + { + "epoch": 0.9957707803010563, + "grad_norm": 1.81935715675354, + "learning_rate": 1.0553504653006709e-05, + "loss": 3.2705, + "step": 12420 + }, + { + "epoch": 0.9958509550820789, + "grad_norm": 1.6835507154464722, + "learning_rate": 1.0552208191822753e-05, + "loss": 3.0887, + "step": 12421 + }, + { + "epoch": 0.9959311298631016, + "grad_norm": 1.808325171470642, + "learning_rate": 1.0550911721328761e-05, + "loss": 3.2493, + "step": 12422 + }, + { + "epoch": 0.9960113046441242, + "grad_norm": 1.7374801635742188, + "learning_rate": 1.0549615241546589e-05, + "loss": 3.1377, + "step": 12423 + }, + { + "epoch": 0.9960914794251469, + "grad_norm": 0.7065247893333435, + "learning_rate": 1.0548318752498098e-05, + "loss": 2.4731, + "step": 12424 + }, + { + "epoch": 0.9961716542061695, + "grad_norm": 1.7601014375686646, + "learning_rate": 1.0547022254205151e-05, + "loss": 3.2678, + "step": 12425 + }, + { + "epoch": 0.9962518289871921, + "grad_norm": 1.783005952835083, + "learning_rate": 1.0545725746689597e-05, + "loss": 3.2073, + "step": 12426 + }, + { + "epoch": 0.9963320037682147, + "grad_norm": 0.6837118268013, + "learning_rate": 1.05444292299733e-05, + "loss": 2.4889, + "step": 12427 + }, + { + "epoch": 0.9964121785492374, + "grad_norm": 1.7931140661239624, + "learning_rate": 1.0543132704078118e-05, + "loss": 3.2398, + "step": 12428 + }, + { + "epoch": 0.99649235333026, + "grad_norm": 1.8010483980178833, + "learning_rate": 1.0541836169025909e-05, + "loss": 3.1287, + "step": 12429 + }, + { + "epoch": 0.9965725281112826, + "grad_norm": 2.039259433746338, + "learning_rate": 1.0540539624838538e-05, + "loss": 3.1715, + "step": 12430 + }, + { + "epoch": 0.9966527028923052, + "grad_norm": 1.8083986043930054, + "learning_rate": 1.0539243071537855e-05, + "loss": 3.2213, + "step": 12431 + }, + { + "epoch": 0.9967328776733279, + "grad_norm": 1.8127005100250244, + "learning_rate": 1.0537946509145725e-05, + "loss": 3.2449, + "step": 12432 + }, + { + "epoch": 0.9968130524543505, + "grad_norm": 1.71242356300354, + "learning_rate": 1.0536649937684006e-05, + "loss": 3.2771, + "step": 12433 + }, + { + "epoch": 0.9968932272353731, + "grad_norm": 1.7533289194107056, + "learning_rate": 1.0535353357174558e-05, + "loss": 3.1786, + "step": 12434 + }, + { + "epoch": 0.9969734020163957, + "grad_norm": 1.7106854915618896, + "learning_rate": 1.0534056767639242e-05, + "loss": 3.0989, + "step": 12435 + }, + { + "epoch": 0.9970535767974184, + "grad_norm": 0.6987037062644958, + "learning_rate": 1.0532760169099915e-05, + "loss": 2.4901, + "step": 12436 + }, + { + "epoch": 0.997133751578441, + "grad_norm": 1.816336750984192, + "learning_rate": 1.0531463561578444e-05, + "loss": 3.1874, + "step": 12437 + }, + { + "epoch": 0.9972139263594636, + "grad_norm": 1.7305517196655273, + "learning_rate": 1.053016694509668e-05, + "loss": 3.1571, + "step": 12438 + }, + { + "epoch": 0.9972941011404862, + "grad_norm": 1.7336708307266235, + "learning_rate": 1.052887031967649e-05, + "loss": 3.1271, + "step": 12439 + }, + { + "epoch": 0.9973742759215088, + "grad_norm": 1.8112751245498657, + "learning_rate": 1.0527573685339733e-05, + "loss": 3.2149, + "step": 12440 + }, + { + "epoch": 0.9974544507025315, + "grad_norm": 1.8334697484970093, + "learning_rate": 1.0526277042108271e-05, + "loss": 3.3334, + "step": 12441 + }, + { + "epoch": 0.9975346254835541, + "grad_norm": 1.8904776573181152, + "learning_rate": 1.0524980390003963e-05, + "loss": 3.3329, + "step": 12442 + }, + { + "epoch": 0.9976148002645768, + "grad_norm": 1.7530930042266846, + "learning_rate": 1.0523683729048668e-05, + "loss": 3.1697, + "step": 12443 + }, + { + "epoch": 0.9976949750455995, + "grad_norm": 1.6943416595458984, + "learning_rate": 1.052238705926425e-05, + "loss": 3.1934, + "step": 12444 + }, + { + "epoch": 0.9977751498266221, + "grad_norm": 1.7956091165542603, + "learning_rate": 1.0521090380672574e-05, + "loss": 3.1579, + "step": 12445 + }, + { + "epoch": 0.9978553246076447, + "grad_norm": 1.7810722589492798, + "learning_rate": 1.0519793693295495e-05, + "loss": 3.3715, + "step": 12446 + }, + { + "epoch": 0.9979354993886673, + "grad_norm": 1.8674724102020264, + "learning_rate": 1.0518496997154878e-05, + "loss": 3.2396, + "step": 12447 + }, + { + "epoch": 0.99801567416969, + "grad_norm": 0.7306225895881653, + "learning_rate": 1.0517200292272584e-05, + "loss": 2.4767, + "step": 12448 + }, + { + "epoch": 0.9980958489507126, + "grad_norm": 0.6989032626152039, + "learning_rate": 1.0515903578670474e-05, + "loss": 2.4836, + "step": 12449 + }, + { + "epoch": 0.9981760237317352, + "grad_norm": 1.7762740850448608, + "learning_rate": 1.0514606856370416e-05, + "loss": 3.2681, + "step": 12450 + }, + { + "epoch": 0.9982561985127578, + "grad_norm": 1.8233660459518433, + "learning_rate": 1.0513310125394263e-05, + "loss": 3.2294, + "step": 12451 + }, + { + "epoch": 0.9983363732937804, + "grad_norm": 1.7623597383499146, + "learning_rate": 1.0512013385763886e-05, + "loss": 3.1661, + "step": 12452 + }, + { + "epoch": 0.9984165480748031, + "grad_norm": 1.7556639909744263, + "learning_rate": 1.0510716637501141e-05, + "loss": 3.2038, + "step": 12453 + }, + { + "epoch": 0.9984967228558257, + "grad_norm": 1.8253438472747803, + "learning_rate": 1.0509419880627893e-05, + "loss": 3.2524, + "step": 12454 + }, + { + "epoch": 0.9985768976368483, + "grad_norm": 1.9120967388153076, + "learning_rate": 1.050812311516601e-05, + "loss": 3.3483, + "step": 12455 + }, + { + "epoch": 0.9986570724178709, + "grad_norm": 0.7317367196083069, + "learning_rate": 1.0506826341137344e-05, + "loss": 2.4768, + "step": 12456 + }, + { + "epoch": 0.9987372471988936, + "grad_norm": 1.7693111896514893, + "learning_rate": 1.050552955856377e-05, + "loss": 3.1938, + "step": 12457 + }, + { + "epoch": 0.9988174219799162, + "grad_norm": 1.7288175821304321, + "learning_rate": 1.0504232767467142e-05, + "loss": 3.1791, + "step": 12458 + }, + { + "epoch": 0.9988975967609388, + "grad_norm": 0.7146478891372681, + "learning_rate": 1.0502935967869327e-05, + "loss": 2.4921, + "step": 12459 + }, + { + "epoch": 0.9989777715419614, + "grad_norm": 1.6270725727081299, + "learning_rate": 1.0501639159792194e-05, + "loss": 3.1263, + "step": 12460 + }, + { + "epoch": 0.9990579463229841, + "grad_norm": 1.8185640573501587, + "learning_rate": 1.0500342343257596e-05, + "loss": 3.2151, + "step": 12461 + }, + { + "epoch": 0.9991381211040067, + "grad_norm": 1.8948475122451782, + "learning_rate": 1.0499045518287408e-05, + "loss": 3.2172, + "step": 12462 + }, + { + "epoch": 0.9992182958850294, + "grad_norm": 1.644782304763794, + "learning_rate": 1.0497748684903481e-05, + "loss": 3.1722, + "step": 12463 + }, + { + "epoch": 0.999298470666052, + "grad_norm": 1.7881858348846436, + "learning_rate": 1.0496451843127691e-05, + "loss": 3.219, + "step": 12464 + }, + { + "epoch": 0.9993786454470747, + "grad_norm": 1.7484654188156128, + "learning_rate": 1.04951549929819e-05, + "loss": 3.1921, + "step": 12465 + }, + { + "epoch": 0.9994588202280973, + "grad_norm": 1.7942190170288086, + "learning_rate": 1.0493858134487966e-05, + "loss": 3.14, + "step": 12466 + }, + { + "epoch": 0.9995389950091199, + "grad_norm": 1.9363155364990234, + "learning_rate": 1.049256126766776e-05, + "loss": 3.2272, + "step": 12467 + }, + { + "epoch": 0.9996191697901425, + "grad_norm": 1.830392837524414, + "learning_rate": 1.0491264392543146e-05, + "loss": 3.1748, + "step": 12468 + }, + { + "epoch": 0.9996993445711652, + "grad_norm": 1.7398252487182617, + "learning_rate": 1.0489967509135983e-05, + "loss": 3.2557, + "step": 12469 + }, + { + "epoch": 0.9997795193521878, + "grad_norm": 1.814813256263733, + "learning_rate": 1.0488670617468145e-05, + "loss": 3.2587, + "step": 12470 + }, + { + "epoch": 0.9998596941332104, + "grad_norm": 1.7986011505126953, + "learning_rate": 1.0487373717561491e-05, + "loss": 3.1814, + "step": 12471 + }, + { + "epoch": 0.999939868914233, + "grad_norm": 1.7048025131225586, + "learning_rate": 1.0486076809437888e-05, + "loss": 3.239, + "step": 12472 + }, + { + "epoch": 1.0000200436952558, + "grad_norm": 1.9114865064620972, + "learning_rate": 1.0484779893119201e-05, + "loss": 3.1116, + "step": 12473 + }, + { + "epoch": 1.0001002184762784, + "grad_norm": 1.8224760293960571, + "learning_rate": 1.0483482968627295e-05, + "loss": 3.3044, + "step": 12474 + }, + { + "epoch": 1.0000801747810226, + "grad_norm": 1.8182132244110107, + "learning_rate": 1.048218603598404e-05, + "loss": 3.278, + "step": 12475 + }, + { + "epoch": 1.0001603495620452, + "grad_norm": 1.9010648727416992, + "learning_rate": 1.0480889095211292e-05, + "loss": 3.3246, + "step": 12476 + }, + { + "epoch": 1.0002405243430679, + "grad_norm": 1.7343391180038452, + "learning_rate": 1.047959214633093e-05, + "loss": 3.1443, + "step": 12477 + }, + { + "epoch": 1.0003206991240905, + "grad_norm": 1.7775626182556152, + "learning_rate": 1.0478295189364809e-05, + "loss": 3.2136, + "step": 12478 + }, + { + "epoch": 1.0004008739051131, + "grad_norm": 1.7631635665893555, + "learning_rate": 1.0476998224334802e-05, + "loss": 3.1678, + "step": 12479 + }, + { + "epoch": 1.0004810486861357, + "grad_norm": 1.9767341613769531, + "learning_rate": 1.0475701251262772e-05, + "loss": 3.1612, + "step": 12480 + }, + { + "epoch": 1.0005612234671584, + "grad_norm": 2.1474788188934326, + "learning_rate": 1.0474404270170586e-05, + "loss": 3.1478, + "step": 12481 + }, + { + "epoch": 1.000641398248181, + "grad_norm": 2.0377237796783447, + "learning_rate": 1.0473107281080115e-05, + "loss": 3.2957, + "step": 12482 + }, + { + "epoch": 1.0007215730292036, + "grad_norm": 1.7347646951675415, + "learning_rate": 1.0471810284013218e-05, + "loss": 3.1906, + "step": 12483 + }, + { + "epoch": 1.0008017478102262, + "grad_norm": 1.6702401638031006, + "learning_rate": 1.0470513278991765e-05, + "loss": 3.1544, + "step": 12484 + }, + { + "epoch": 1.0008819225912489, + "grad_norm": 1.7414228916168213, + "learning_rate": 1.0469216266037628e-05, + "loss": 3.0876, + "step": 12485 + }, + { + "epoch": 1.0009620973722715, + "grad_norm": 1.758704423904419, + "learning_rate": 1.0467919245172669e-05, + "loss": 3.2832, + "step": 12486 + }, + { + "epoch": 1.001042272153294, + "grad_norm": 1.9253311157226562, + "learning_rate": 1.0466622216418755e-05, + "loss": 3.2189, + "step": 12487 + }, + { + "epoch": 1.0011224469343167, + "grad_norm": 1.6666065454483032, + "learning_rate": 1.0465325179797754e-05, + "loss": 3.1328, + "step": 12488 + }, + { + "epoch": 1.0012026217153394, + "grad_norm": 1.8582875728607178, + "learning_rate": 1.0464028135331537e-05, + "loss": 3.2885, + "step": 12489 + }, + { + "epoch": 1.001282796496362, + "grad_norm": 1.8267686367034912, + "learning_rate": 1.0462731083041969e-05, + "loss": 3.1777, + "step": 12490 + }, + { + "epoch": 1.0013629712773846, + "grad_norm": 1.7660361528396606, + "learning_rate": 1.0461434022950916e-05, + "loss": 3.1447, + "step": 12491 + }, + { + "epoch": 1.0014431460584072, + "grad_norm": 1.847853422164917, + "learning_rate": 1.0460136955080248e-05, + "loss": 3.255, + "step": 12492 + }, + { + "epoch": 1.0015233208394299, + "grad_norm": 1.8379909992218018, + "learning_rate": 1.0458839879451835e-05, + "loss": 3.2092, + "step": 12493 + }, + { + "epoch": 1.0016034956204525, + "grad_norm": 1.752030611038208, + "learning_rate": 1.045754279608754e-05, + "loss": 3.1658, + "step": 12494 + }, + { + "epoch": 1.0016836704014753, + "grad_norm": 1.8345717191696167, + "learning_rate": 1.045624570500924e-05, + "loss": 3.1323, + "step": 12495 + }, + { + "epoch": 1.001763845182498, + "grad_norm": 2.057305335998535, + "learning_rate": 1.0454948606238796e-05, + "loss": 3.2228, + "step": 12496 + }, + { + "epoch": 1.0018440199635206, + "grad_norm": 1.767529845237732, + "learning_rate": 1.045365149979808e-05, + "loss": 3.2112, + "step": 12497 + }, + { + "epoch": 1.0019241947445432, + "grad_norm": 1.8351975679397583, + "learning_rate": 1.0452354385708956e-05, + "loss": 3.1508, + "step": 12498 + }, + { + "epoch": 1.0020043695255658, + "grad_norm": 1.7448976039886475, + "learning_rate": 1.0451057263993296e-05, + "loss": 3.1613, + "step": 12499 + }, + { + "epoch": 1.0020845443065884, + "grad_norm": 1.8625253438949585, + "learning_rate": 1.0449760134672975e-05, + "loss": 3.2008, + "step": 12500 + }, + { + "epoch": 1.002164719087611, + "grad_norm": 1.7377034425735474, + "learning_rate": 1.0448462997769852e-05, + "loss": 3.1253, + "step": 12501 + }, + { + "epoch": 1.0022448938686337, + "grad_norm": 1.7417750358581543, + "learning_rate": 1.0447165853305804e-05, + "loss": 3.1757, + "step": 12502 + }, + { + "epoch": 1.0023250686496563, + "grad_norm": 1.747412085533142, + "learning_rate": 1.0445868701302694e-05, + "loss": 3.1518, + "step": 12503 + }, + { + "epoch": 1.002405243430679, + "grad_norm": 1.8048629760742188, + "learning_rate": 1.0444571541782396e-05, + "loss": 3.2885, + "step": 12504 + }, + { + "epoch": 1.0024854182117016, + "grad_norm": 1.7586249113082886, + "learning_rate": 1.044327437476678e-05, + "loss": 3.2017, + "step": 12505 + }, + { + "epoch": 1.0025655929927242, + "grad_norm": 1.8924928903579712, + "learning_rate": 1.0441977200277713e-05, + "loss": 3.1819, + "step": 12506 + }, + { + "epoch": 1.0026457677737468, + "grad_norm": 1.8298521041870117, + "learning_rate": 1.0440680018337067e-05, + "loss": 3.1721, + "step": 12507 + }, + { + "epoch": 1.0027259425547694, + "grad_norm": 1.764767050743103, + "learning_rate": 1.0439382828966711e-05, + "loss": 3.236, + "step": 12508 + }, + { + "epoch": 1.002806117335792, + "grad_norm": 1.6499290466308594, + "learning_rate": 1.0438085632188512e-05, + "loss": 3.0977, + "step": 12509 + }, + { + "epoch": 1.0028862921168147, + "grad_norm": 1.7336560487747192, + "learning_rate": 1.043678842802435e-05, + "loss": 3.2579, + "step": 12510 + }, + { + "epoch": 1.0029664668978373, + "grad_norm": 1.9266525506973267, + "learning_rate": 1.0435491216496086e-05, + "loss": 3.183, + "step": 12511 + }, + { + "epoch": 1.00304664167886, + "grad_norm": 1.7110095024108887, + "learning_rate": 1.0434193997625592e-05, + "loss": 3.1609, + "step": 12512 + }, + { + "epoch": 1.0031268164598826, + "grad_norm": 1.8046454191207886, + "learning_rate": 1.0432896771434742e-05, + "loss": 3.2496, + "step": 12513 + }, + { + "epoch": 1.0032069912409052, + "grad_norm": 1.702429175376892, + "learning_rate": 1.0431599537945403e-05, + "loss": 3.2448, + "step": 12514 + }, + { + "epoch": 1.0032871660219278, + "grad_norm": 1.7968190908432007, + "learning_rate": 1.043030229717945e-05, + "loss": 3.2332, + "step": 12515 + }, + { + "epoch": 1.0033673408029504, + "grad_norm": 1.8404830694198608, + "learning_rate": 1.0429005049158751e-05, + "loss": 3.1265, + "step": 12516 + }, + { + "epoch": 1.003447515583973, + "grad_norm": 1.7470808029174805, + "learning_rate": 1.042770779390518e-05, + "loss": 3.1299, + "step": 12517 + }, + { + "epoch": 1.0035276903649957, + "grad_norm": 1.6800777912139893, + "learning_rate": 1.0426410531440605e-05, + "loss": 3.0798, + "step": 12518 + }, + { + "epoch": 1.0036078651460183, + "grad_norm": 1.7882721424102783, + "learning_rate": 1.0425113261786896e-05, + "loss": 3.2288, + "step": 12519 + }, + { + "epoch": 1.003688039927041, + "grad_norm": 1.7733615636825562, + "learning_rate": 1.0423815984965933e-05, + "loss": 3.2195, + "step": 12520 + }, + { + "epoch": 1.0037682147080635, + "grad_norm": 1.9078044891357422, + "learning_rate": 1.0422518700999576e-05, + "loss": 3.2051, + "step": 12521 + }, + { + "epoch": 1.0038483894890862, + "grad_norm": 1.7217599153518677, + "learning_rate": 1.0421221409909708e-05, + "loss": 3.1229, + "step": 12522 + }, + { + "epoch": 1.0039285642701088, + "grad_norm": 1.7499758005142212, + "learning_rate": 1.0419924111718191e-05, + "loss": 3.1769, + "step": 12523 + }, + { + "epoch": 1.0040087390511314, + "grad_norm": 1.7536604404449463, + "learning_rate": 1.0418626806446901e-05, + "loss": 3.2432, + "step": 12524 + }, + { + "epoch": 1.004088913832154, + "grad_norm": 1.7653660774230957, + "learning_rate": 1.0417329494117714e-05, + "loss": 3.223, + "step": 12525 + }, + { + "epoch": 1.0041690886131767, + "grad_norm": 1.9280340671539307, + "learning_rate": 1.0416032174752498e-05, + "loss": 3.1117, + "step": 12526 + }, + { + "epoch": 1.0042492633941993, + "grad_norm": 1.8526867628097534, + "learning_rate": 1.0414734848373126e-05, + "loss": 3.2141, + "step": 12527 + }, + { + "epoch": 1.004329438175222, + "grad_norm": 1.8274801969528198, + "learning_rate": 1.041343751500147e-05, + "loss": 3.1716, + "step": 12528 + }, + { + "epoch": 1.0044096129562445, + "grad_norm": 1.8239448070526123, + "learning_rate": 1.0412140174659403e-05, + "loss": 3.2622, + "step": 12529 + }, + { + "epoch": 1.0044897877372672, + "grad_norm": 1.7716734409332275, + "learning_rate": 1.0410842827368798e-05, + "loss": 3.2367, + "step": 12530 + }, + { + "epoch": 1.0045699625182898, + "grad_norm": 1.7923011779785156, + "learning_rate": 1.0409545473151528e-05, + "loss": 3.2487, + "step": 12531 + }, + { + "epoch": 1.0046501372993124, + "grad_norm": 1.9005894660949707, + "learning_rate": 1.0408248112029466e-05, + "loss": 3.1826, + "step": 12532 + }, + { + "epoch": 1.004730312080335, + "grad_norm": 1.7003124952316284, + "learning_rate": 1.0406950744024485e-05, + "loss": 3.0692, + "step": 12533 + }, + { + "epoch": 1.0048104868613577, + "grad_norm": 1.767276644706726, + "learning_rate": 1.0405653369158457e-05, + "loss": 3.2139, + "step": 12534 + }, + { + "epoch": 1.0048906616423805, + "grad_norm": 1.8324452638626099, + "learning_rate": 1.0404355987453258e-05, + "loss": 3.1341, + "step": 12535 + }, + { + "epoch": 1.0049708364234031, + "grad_norm": 1.8670212030410767, + "learning_rate": 1.0403058598930758e-05, + "loss": 3.1847, + "step": 12536 + }, + { + "epoch": 1.0050510112044257, + "grad_norm": 1.901282787322998, + "learning_rate": 1.0401761203612832e-05, + "loss": 3.2056, + "step": 12537 + }, + { + "epoch": 1.0051311859854484, + "grad_norm": 0.7453895807266235, + "learning_rate": 1.0400463801521352e-05, + "loss": 2.5045, + "step": 12538 + }, + { + "epoch": 1.005211360766471, + "grad_norm": 1.7975267171859741, + "learning_rate": 1.0399166392678195e-05, + "loss": 3.1812, + "step": 12539 + }, + { + "epoch": 1.0052915355474936, + "grad_norm": 1.672215461730957, + "learning_rate": 1.0397868977105236e-05, + "loss": 3.2524, + "step": 12540 + }, + { + "epoch": 1.0053717103285162, + "grad_norm": 1.7426388263702393, + "learning_rate": 1.0396571554824341e-05, + "loss": 3.1425, + "step": 12541 + }, + { + "epoch": 1.0054518851095389, + "grad_norm": 1.8141798973083496, + "learning_rate": 1.0395274125857396e-05, + "loss": 3.2796, + "step": 12542 + }, + { + "epoch": 1.0055320598905615, + "grad_norm": 1.737428069114685, + "learning_rate": 1.0393976690226264e-05, + "loss": 3.2067, + "step": 12543 + }, + { + "epoch": 1.0056122346715841, + "grad_norm": 0.735256016254425, + "learning_rate": 1.0392679247952824e-05, + "loss": 2.4642, + "step": 12544 + }, + { + "epoch": 1.0056924094526067, + "grad_norm": 1.8916715383529663, + "learning_rate": 1.039138179905895e-05, + "loss": 3.2543, + "step": 12545 + }, + { + "epoch": 1.0057725842336294, + "grad_norm": 0.7225174307823181, + "learning_rate": 1.0390084343566518e-05, + "loss": 2.4861, + "step": 12546 + }, + { + "epoch": 1.005852759014652, + "grad_norm": 1.8390824794769287, + "learning_rate": 1.0388786881497401e-05, + "loss": 3.3458, + "step": 12547 + }, + { + "epoch": 1.0059329337956746, + "grad_norm": 1.6538976430892944, + "learning_rate": 1.0387489412873474e-05, + "loss": 3.1276, + "step": 12548 + }, + { + "epoch": 1.0060131085766972, + "grad_norm": 1.7820181846618652, + "learning_rate": 1.0386191937716612e-05, + "loss": 3.1505, + "step": 12549 + }, + { + "epoch": 1.0060932833577199, + "grad_norm": 1.990462303161621, + "learning_rate": 1.0384894456048693e-05, + "loss": 3.243, + "step": 12550 + }, + { + "epoch": 1.0061734581387425, + "grad_norm": 1.8014541864395142, + "learning_rate": 1.0383596967891585e-05, + "loss": 3.2821, + "step": 12551 + }, + { + "epoch": 1.006253632919765, + "grad_norm": 1.7780804634094238, + "learning_rate": 1.038229947326717e-05, + "loss": 3.2098, + "step": 12552 + }, + { + "epoch": 1.0063338077007877, + "grad_norm": 1.7106760740280151, + "learning_rate": 1.0381001972197317e-05, + "loss": 3.0907, + "step": 12553 + }, + { + "epoch": 1.0064139824818104, + "grad_norm": 1.7511082887649536, + "learning_rate": 1.0379704464703907e-05, + "loss": 3.2964, + "step": 12554 + }, + { + "epoch": 1.006494157262833, + "grad_norm": 1.7838072776794434, + "learning_rate": 1.0378406950808817e-05, + "loss": 3.1342, + "step": 12555 + }, + { + "epoch": 1.0065743320438556, + "grad_norm": 1.7840560674667358, + "learning_rate": 1.0377109430533915e-05, + "loss": 3.2328, + "step": 12556 + }, + { + "epoch": 1.0066545068248782, + "grad_norm": 1.743385672569275, + "learning_rate": 1.037581190390108e-05, + "loss": 3.1478, + "step": 12557 + }, + { + "epoch": 1.0067346816059008, + "grad_norm": 1.8400650024414062, + "learning_rate": 1.0374514370932194e-05, + "loss": 3.1016, + "step": 12558 + }, + { + "epoch": 1.0068148563869235, + "grad_norm": 1.8032532930374146, + "learning_rate": 1.0373216831649124e-05, + "loss": 3.2123, + "step": 12559 + }, + { + "epoch": 1.006895031167946, + "grad_norm": 1.7718991041183472, + "learning_rate": 1.037191928607375e-05, + "loss": 3.2319, + "step": 12560 + }, + { + "epoch": 1.0069752059489687, + "grad_norm": 1.844977617263794, + "learning_rate": 1.0370621734227948e-05, + "loss": 3.245, + "step": 12561 + }, + { + "epoch": 1.0070553807299913, + "grad_norm": 1.847583293914795, + "learning_rate": 1.0369324176133594e-05, + "loss": 3.2107, + "step": 12562 + }, + { + "epoch": 1.007135555511014, + "grad_norm": 1.71328604221344, + "learning_rate": 1.0368026611812568e-05, + "loss": 3.1235, + "step": 12563 + }, + { + "epoch": 1.0072157302920366, + "grad_norm": 1.8192447423934937, + "learning_rate": 1.0366729041286738e-05, + "loss": 3.1661, + "step": 12564 + }, + { + "epoch": 1.0072959050730592, + "grad_norm": 1.770677924156189, + "learning_rate": 1.0365431464577989e-05, + "loss": 3.2915, + "step": 12565 + }, + { + "epoch": 1.0073760798540818, + "grad_norm": 1.7801872491836548, + "learning_rate": 1.0364133881708192e-05, + "loss": 3.2677, + "step": 12566 + }, + { + "epoch": 1.0074562546351045, + "grad_norm": 1.9541254043579102, + "learning_rate": 1.0362836292699225e-05, + "loss": 3.3787, + "step": 12567 + }, + { + "epoch": 1.007536429416127, + "grad_norm": 1.816299557685852, + "learning_rate": 1.036153869757297e-05, + "loss": 3.1562, + "step": 12568 + }, + { + "epoch": 1.0076166041971497, + "grad_norm": 1.6899347305297852, + "learning_rate": 1.0360241096351297e-05, + "loss": 3.1863, + "step": 12569 + }, + { + "epoch": 1.0076967789781723, + "grad_norm": 1.8838238716125488, + "learning_rate": 1.0358943489056088e-05, + "loss": 3.189, + "step": 12570 + }, + { + "epoch": 1.007776953759195, + "grad_norm": 1.7487462759017944, + "learning_rate": 1.0357645875709218e-05, + "loss": 3.252, + "step": 12571 + }, + { + "epoch": 1.0078571285402176, + "grad_norm": 1.7958704233169556, + "learning_rate": 1.0356348256332562e-05, + "loss": 3.2323, + "step": 12572 + }, + { + "epoch": 1.0079373033212402, + "grad_norm": 1.7571048736572266, + "learning_rate": 1.0355050630948002e-05, + "loss": 3.1356, + "step": 12573 + }, + { + "epoch": 1.0080174781022628, + "grad_norm": 1.9905561208724976, + "learning_rate": 1.0353752999577414e-05, + "loss": 3.1148, + "step": 12574 + }, + { + "epoch": 1.0080976528832855, + "grad_norm": 1.6920154094696045, + "learning_rate": 1.0352455362242674e-05, + "loss": 3.1349, + "step": 12575 + }, + { + "epoch": 1.0081778276643083, + "grad_norm": 1.803236961364746, + "learning_rate": 1.0351157718965663e-05, + "loss": 3.1113, + "step": 12576 + }, + { + "epoch": 1.008258002445331, + "grad_norm": 1.6411314010620117, + "learning_rate": 1.0349860069768253e-05, + "loss": 3.1678, + "step": 12577 + }, + { + "epoch": 1.0083381772263535, + "grad_norm": 1.7309709787368774, + "learning_rate": 1.034856241467233e-05, + "loss": 3.1348, + "step": 12578 + }, + { + "epoch": 1.0084183520073762, + "grad_norm": 1.722014307975769, + "learning_rate": 1.0347264753699764e-05, + "loss": 3.1279, + "step": 12579 + }, + { + "epoch": 1.0084985267883988, + "grad_norm": 1.893266201019287, + "learning_rate": 1.034596708687244e-05, + "loss": 3.3281, + "step": 12580 + }, + { + "epoch": 1.0085787015694214, + "grad_norm": 1.7328823804855347, + "learning_rate": 1.0344669414212228e-05, + "loss": 3.2247, + "step": 12581 + }, + { + "epoch": 1.008658876350444, + "grad_norm": 1.8482812643051147, + "learning_rate": 1.0343371735741013e-05, + "loss": 3.2175, + "step": 12582 + }, + { + "epoch": 1.0087390511314667, + "grad_norm": 1.74052095413208, + "learning_rate": 1.0342074051480673e-05, + "loss": 3.1277, + "step": 12583 + }, + { + "epoch": 1.0088192259124893, + "grad_norm": 1.6954476833343506, + "learning_rate": 1.0340776361453084e-05, + "loss": 3.2556, + "step": 12584 + }, + { + "epoch": 1.008899400693512, + "grad_norm": 1.848266363143921, + "learning_rate": 1.0339478665680127e-05, + "loss": 3.3146, + "step": 12585 + }, + { + "epoch": 1.0089795754745345, + "grad_norm": 1.9926650524139404, + "learning_rate": 1.0338180964183678e-05, + "loss": 3.1619, + "step": 12586 + }, + { + "epoch": 1.0090597502555572, + "grad_norm": 1.7734721899032593, + "learning_rate": 1.0336883256985615e-05, + "loss": 3.2211, + "step": 12587 + }, + { + "epoch": 1.0091399250365798, + "grad_norm": 2.098689079284668, + "learning_rate": 1.0335585544107823e-05, + "loss": 3.1582, + "step": 12588 + }, + { + "epoch": 1.0092200998176024, + "grad_norm": 1.8561121225357056, + "learning_rate": 1.0334287825572176e-05, + "loss": 3.2141, + "step": 12589 + }, + { + "epoch": 1.009300274598625, + "grad_norm": 0.7229263782501221, + "learning_rate": 1.0332990101400552e-05, + "loss": 2.5005, + "step": 12590 + }, + { + "epoch": 1.0093804493796477, + "grad_norm": 1.714587688446045, + "learning_rate": 1.0331692371614834e-05, + "loss": 3.3125, + "step": 12591 + }, + { + "epoch": 1.0094606241606703, + "grad_norm": 1.9946181774139404, + "learning_rate": 1.0330394636236898e-05, + "loss": 3.2999, + "step": 12592 + }, + { + "epoch": 1.009540798941693, + "grad_norm": 1.7728625535964966, + "learning_rate": 1.0329096895288628e-05, + "loss": 3.1315, + "step": 12593 + }, + { + "epoch": 1.0096209737227155, + "grad_norm": 1.8095710277557373, + "learning_rate": 1.0327799148791897e-05, + "loss": 3.2334, + "step": 12594 + }, + { + "epoch": 1.0097011485037382, + "grad_norm": 0.7376206517219543, + "learning_rate": 1.032650139676859e-05, + "loss": 2.462, + "step": 12595 + }, + { + "epoch": 1.0097813232847608, + "grad_norm": 1.7292389869689941, + "learning_rate": 1.0325203639240584e-05, + "loss": 3.2427, + "step": 12596 + }, + { + "epoch": 1.0098614980657834, + "grad_norm": 0.7662213444709778, + "learning_rate": 1.0323905876229758e-05, + "loss": 2.4871, + "step": 12597 + }, + { + "epoch": 1.009941672846806, + "grad_norm": 1.66849684715271, + "learning_rate": 1.0322608107757997e-05, + "loss": 3.1478, + "step": 12598 + }, + { + "epoch": 1.0100218476278287, + "grad_norm": 1.719295620918274, + "learning_rate": 1.0321310333847172e-05, + "loss": 3.2094, + "step": 12599 + }, + { + "epoch": 1.0101020224088513, + "grad_norm": 1.7662895917892456, + "learning_rate": 1.0320012554519173e-05, + "loss": 3.1151, + "step": 12600 + }, + { + "epoch": 1.010182197189874, + "grad_norm": 1.8516592979431152, + "learning_rate": 1.031871476979587e-05, + "loss": 3.2266, + "step": 12601 + }, + { + "epoch": 1.0102623719708965, + "grad_norm": 1.752966284751892, + "learning_rate": 1.0317416979699152e-05, + "loss": 3.0517, + "step": 12602 + }, + { + "epoch": 1.0103425467519191, + "grad_norm": 0.686181366443634, + "learning_rate": 1.0316119184250892e-05, + "loss": 2.4813, + "step": 12603 + }, + { + "epoch": 1.0104227215329418, + "grad_norm": 1.7081040143966675, + "learning_rate": 1.0314821383472976e-05, + "loss": 3.0959, + "step": 12604 + }, + { + "epoch": 1.0105028963139644, + "grad_norm": 1.6469193696975708, + "learning_rate": 1.0313523577387286e-05, + "loss": 3.0956, + "step": 12605 + }, + { + "epoch": 1.010583071094987, + "grad_norm": 1.8009546995162964, + "learning_rate": 1.0312225766015694e-05, + "loss": 3.1985, + "step": 12606 + }, + { + "epoch": 1.0106632458760096, + "grad_norm": 0.6938043832778931, + "learning_rate": 1.0310927949380084e-05, + "loss": 2.483, + "step": 12607 + }, + { + "epoch": 1.0107434206570323, + "grad_norm": 0.7002843618392944, + "learning_rate": 1.0309630127502344e-05, + "loss": 2.4915, + "step": 12608 + }, + { + "epoch": 1.010823595438055, + "grad_norm": 1.7133104801177979, + "learning_rate": 1.0308332300404346e-05, + "loss": 3.2253, + "step": 12609 + }, + { + "epoch": 1.0109037702190775, + "grad_norm": 1.7491178512573242, + "learning_rate": 1.0307034468107977e-05, + "loss": 3.0899, + "step": 12610 + }, + { + "epoch": 1.0109839450001001, + "grad_norm": 1.8062111139297485, + "learning_rate": 1.030573663063511e-05, + "loss": 3.2281, + "step": 12611 + }, + { + "epoch": 1.0110641197811228, + "grad_norm": 1.764572262763977, + "learning_rate": 1.0304438788007634e-05, + "loss": 3.2293, + "step": 12612 + }, + { + "epoch": 1.0111442945621454, + "grad_norm": 1.7453128099441528, + "learning_rate": 1.030314094024743e-05, + "loss": 3.2391, + "step": 12613 + }, + { + "epoch": 1.011224469343168, + "grad_norm": 1.8179714679718018, + "learning_rate": 1.0301843087376373e-05, + "loss": 3.2047, + "step": 12614 + }, + { + "epoch": 1.0113046441241906, + "grad_norm": 1.7235089540481567, + "learning_rate": 1.030054522941635e-05, + "loss": 3.2472, + "step": 12615 + }, + { + "epoch": 1.0113848189052133, + "grad_norm": 1.8350993394851685, + "learning_rate": 1.0299247366389238e-05, + "loss": 3.308, + "step": 12616 + }, + { + "epoch": 1.011464993686236, + "grad_norm": 1.7103838920593262, + "learning_rate": 1.029794949831692e-05, + "loss": 3.2075, + "step": 12617 + }, + { + "epoch": 1.0115451684672587, + "grad_norm": 1.8896526098251343, + "learning_rate": 1.0296651625221283e-05, + "loss": 3.23, + "step": 12618 + }, + { + "epoch": 1.0116253432482814, + "grad_norm": 0.77511066198349, + "learning_rate": 1.0295353747124201e-05, + "loss": 2.5002, + "step": 12619 + }, + { + "epoch": 1.011705518029304, + "grad_norm": 1.8570095300674438, + "learning_rate": 1.0294055864047561e-05, + "loss": 3.1493, + "step": 12620 + }, + { + "epoch": 1.0117856928103266, + "grad_norm": 1.762868046760559, + "learning_rate": 1.029275797601324e-05, + "loss": 3.1685, + "step": 12621 + }, + { + "epoch": 1.0118658675913492, + "grad_norm": 1.8463716506958008, + "learning_rate": 1.0291460083043124e-05, + "loss": 3.1472, + "step": 12622 + }, + { + "epoch": 1.0119460423723718, + "grad_norm": 1.8705192804336548, + "learning_rate": 1.0290162185159095e-05, + "loss": 3.2451, + "step": 12623 + }, + { + "epoch": 1.0120262171533945, + "grad_norm": 1.8636248111724854, + "learning_rate": 1.0288864282383031e-05, + "loss": 3.204, + "step": 12624 + }, + { + "epoch": 1.012106391934417, + "grad_norm": 1.7940022945404053, + "learning_rate": 1.0287566374736818e-05, + "loss": 3.2515, + "step": 12625 + }, + { + "epoch": 1.0121865667154397, + "grad_norm": 0.7064555287361145, + "learning_rate": 1.028626846224234e-05, + "loss": 2.4738, + "step": 12626 + }, + { + "epoch": 1.0122667414964623, + "grad_norm": 1.8314052820205688, + "learning_rate": 1.0284970544921472e-05, + "loss": 3.1584, + "step": 12627 + }, + { + "epoch": 1.012346916277485, + "grad_norm": 1.6834849119186401, + "learning_rate": 1.0283672622796105e-05, + "loss": 3.1224, + "step": 12628 + }, + { + "epoch": 1.0124270910585076, + "grad_norm": 1.7953269481658936, + "learning_rate": 1.0282374695888116e-05, + "loss": 3.0869, + "step": 12629 + }, + { + "epoch": 1.0125072658395302, + "grad_norm": 1.7001572847366333, + "learning_rate": 1.0281076764219391e-05, + "loss": 3.2429, + "step": 12630 + }, + { + "epoch": 1.0125874406205528, + "grad_norm": 1.8022767305374146, + "learning_rate": 1.0279778827811809e-05, + "loss": 3.1175, + "step": 12631 + }, + { + "epoch": 1.0126676154015755, + "grad_norm": 1.9152302742004395, + "learning_rate": 1.0278480886687252e-05, + "loss": 3.124, + "step": 12632 + }, + { + "epoch": 1.012747790182598, + "grad_norm": 1.8029712438583374, + "learning_rate": 1.0277182940867609e-05, + "loss": 3.153, + "step": 12633 + }, + { + "epoch": 1.0128279649636207, + "grad_norm": 1.753636360168457, + "learning_rate": 1.0275884990374758e-05, + "loss": 3.2168, + "step": 12634 + }, + { + "epoch": 1.0129081397446433, + "grad_norm": 1.7655807733535767, + "learning_rate": 1.0274587035230585e-05, + "loss": 3.26, + "step": 12635 + }, + { + "epoch": 1.012988314525666, + "grad_norm": 1.766352653503418, + "learning_rate": 1.027328907545697e-05, + "loss": 3.2072, + "step": 12636 + }, + { + "epoch": 1.0130684893066886, + "grad_norm": 1.8301901817321777, + "learning_rate": 1.0271991111075796e-05, + "loss": 3.0848, + "step": 12637 + }, + { + "epoch": 1.0131486640877112, + "grad_norm": 1.8362301588058472, + "learning_rate": 1.0270693142108951e-05, + "loss": 3.2881, + "step": 12638 + }, + { + "epoch": 1.0132288388687338, + "grad_norm": 1.7133928537368774, + "learning_rate": 1.0269395168578313e-05, + "loss": 3.2704, + "step": 12639 + }, + { + "epoch": 1.0133090136497565, + "grad_norm": 1.8441754579544067, + "learning_rate": 1.0268097190505768e-05, + "loss": 3.2025, + "step": 12640 + }, + { + "epoch": 1.013389188430779, + "grad_norm": 1.6769791841506958, + "learning_rate": 1.0266799207913199e-05, + "loss": 3.2527, + "step": 12641 + }, + { + "epoch": 1.0134693632118017, + "grad_norm": 1.923485279083252, + "learning_rate": 1.0265501220822487e-05, + "loss": 3.2224, + "step": 12642 + }, + { + "epoch": 1.0135495379928243, + "grad_norm": 1.766540765762329, + "learning_rate": 1.026420322925552e-05, + "loss": 3.2891, + "step": 12643 + }, + { + "epoch": 1.013629712773847, + "grad_norm": 1.7701292037963867, + "learning_rate": 1.026290523323418e-05, + "loss": 3.2305, + "step": 12644 + }, + { + "epoch": 1.0137098875548696, + "grad_norm": 1.7326843738555908, + "learning_rate": 1.026160723278035e-05, + "loss": 3.1544, + "step": 12645 + }, + { + "epoch": 1.0137900623358922, + "grad_norm": 1.8943713903427124, + "learning_rate": 1.0260309227915914e-05, + "loss": 3.334, + "step": 12646 + }, + { + "epoch": 1.0138702371169148, + "grad_norm": 1.7869504690170288, + "learning_rate": 1.0259011218662756e-05, + "loss": 3.2328, + "step": 12647 + }, + { + "epoch": 1.0139504118979374, + "grad_norm": 1.79265296459198, + "learning_rate": 1.0257713205042764e-05, + "loss": 3.0514, + "step": 12648 + }, + { + "epoch": 1.01403058667896, + "grad_norm": 1.787325382232666, + "learning_rate": 1.0256415187077813e-05, + "loss": 3.0865, + "step": 12649 + }, + { + "epoch": 1.0141107614599827, + "grad_norm": 1.7559722661972046, + "learning_rate": 1.0255117164789793e-05, + "loss": 3.0677, + "step": 12650 + }, + { + "epoch": 1.0141909362410053, + "grad_norm": 1.7465111017227173, + "learning_rate": 1.025381913820059e-05, + "loss": 3.2836, + "step": 12651 + }, + { + "epoch": 1.014271111022028, + "grad_norm": 1.7432682514190674, + "learning_rate": 1.0252521107332083e-05, + "loss": 3.2082, + "step": 12652 + }, + { + "epoch": 1.0143512858030506, + "grad_norm": 1.8119982481002808, + "learning_rate": 1.0251223072206163e-05, + "loss": 3.2189, + "step": 12653 + }, + { + "epoch": 1.0144314605840732, + "grad_norm": 1.7546520233154297, + "learning_rate": 1.0249925032844705e-05, + "loss": 3.2295, + "step": 12654 + }, + { + "epoch": 1.0145116353650958, + "grad_norm": 2.0331270694732666, + "learning_rate": 1.0248626989269603e-05, + "loss": 3.139, + "step": 12655 + }, + { + "epoch": 1.0145918101461184, + "grad_norm": 1.8138161897659302, + "learning_rate": 1.0247328941502736e-05, + "loss": 3.1255, + "step": 12656 + }, + { + "epoch": 1.014671984927141, + "grad_norm": 0.7026218175888062, + "learning_rate": 1.024603088956599e-05, + "loss": 2.4823, + "step": 12657 + }, + { + "epoch": 1.014752159708164, + "grad_norm": 1.6554795503616333, + "learning_rate": 1.0244732833481253e-05, + "loss": 3.1407, + "step": 12658 + }, + { + "epoch": 1.0148323344891865, + "grad_norm": 1.7313514947891235, + "learning_rate": 1.0243434773270402e-05, + "loss": 3.138, + "step": 12659 + }, + { + "epoch": 1.0149125092702092, + "grad_norm": 1.886391520500183, + "learning_rate": 1.024213670895533e-05, + "loss": 3.266, + "step": 12660 + }, + { + "epoch": 1.0149926840512318, + "grad_norm": 1.8328566551208496, + "learning_rate": 1.0240838640557917e-05, + "loss": 3.1777, + "step": 12661 + }, + { + "epoch": 1.0150728588322544, + "grad_norm": 1.8081395626068115, + "learning_rate": 1.023954056810005e-05, + "loss": 3.1696, + "step": 12662 + }, + { + "epoch": 1.015153033613277, + "grad_norm": 1.7931517362594604, + "learning_rate": 1.0238242491603614e-05, + "loss": 3.0649, + "step": 12663 + }, + { + "epoch": 1.0152332083942996, + "grad_norm": 1.6853686571121216, + "learning_rate": 1.0236944411090492e-05, + "loss": 3.1083, + "step": 12664 + }, + { + "epoch": 1.0153133831753223, + "grad_norm": 1.7275766134262085, + "learning_rate": 1.0235646326582574e-05, + "loss": 3.1747, + "step": 12665 + }, + { + "epoch": 1.015393557956345, + "grad_norm": 1.7330973148345947, + "learning_rate": 1.0234348238101737e-05, + "loss": 3.2351, + "step": 12666 + }, + { + "epoch": 1.0154737327373675, + "grad_norm": 1.9470105171203613, + "learning_rate": 1.0233050145669874e-05, + "loss": 3.2037, + "step": 12667 + }, + { + "epoch": 1.0155539075183901, + "grad_norm": 1.7220903635025024, + "learning_rate": 1.0231752049308867e-05, + "loss": 3.1207, + "step": 12668 + }, + { + "epoch": 1.0156340822994128, + "grad_norm": 1.755859375, + "learning_rate": 1.0230453949040604e-05, + "loss": 3.145, + "step": 12669 + }, + { + "epoch": 1.0157142570804354, + "grad_norm": 1.6258476972579956, + "learning_rate": 1.022915584488697e-05, + "loss": 3.1537, + "step": 12670 + }, + { + "epoch": 1.015794431861458, + "grad_norm": 1.7179533243179321, + "learning_rate": 1.0227857736869844e-05, + "loss": 3.2668, + "step": 12671 + }, + { + "epoch": 1.0158746066424806, + "grad_norm": 0.7191528081893921, + "learning_rate": 1.0226559625011118e-05, + "loss": 2.4871, + "step": 12672 + }, + { + "epoch": 1.0159547814235033, + "grad_norm": 1.8505932092666626, + "learning_rate": 1.0225261509332678e-05, + "loss": 3.2535, + "step": 12673 + }, + { + "epoch": 1.0160349562045259, + "grad_norm": 1.697982668876648, + "learning_rate": 1.0223963389856409e-05, + "loss": 3.2744, + "step": 12674 + }, + { + "epoch": 1.0161151309855485, + "grad_norm": 1.7829657793045044, + "learning_rate": 1.0222665266604197e-05, + "loss": 3.0886, + "step": 12675 + }, + { + "epoch": 1.0161953057665711, + "grad_norm": 1.749829888343811, + "learning_rate": 1.0221367139597924e-05, + "loss": 3.0665, + "step": 12676 + }, + { + "epoch": 1.0162754805475938, + "grad_norm": 1.8067113161087036, + "learning_rate": 1.022006900885948e-05, + "loss": 3.1875, + "step": 12677 + }, + { + "epoch": 1.0163556553286164, + "grad_norm": 0.6927756071090698, + "learning_rate": 1.021877087441075e-05, + "loss": 2.481, + "step": 12678 + }, + { + "epoch": 1.016435830109639, + "grad_norm": 1.8320136070251465, + "learning_rate": 1.0217472736273621e-05, + "loss": 3.2874, + "step": 12679 + }, + { + "epoch": 1.0165160048906616, + "grad_norm": 1.7488994598388672, + "learning_rate": 1.0216174594469978e-05, + "loss": 3.1261, + "step": 12680 + }, + { + "epoch": 1.0165961796716843, + "grad_norm": 1.9885450601577759, + "learning_rate": 1.0214876449021708e-05, + "loss": 3.0258, + "step": 12681 + }, + { + "epoch": 1.0166763544527069, + "grad_norm": 1.9221869707107544, + "learning_rate": 1.0213578299950695e-05, + "loss": 3.2161, + "step": 12682 + }, + { + "epoch": 1.0167565292337295, + "grad_norm": 1.8213261365890503, + "learning_rate": 1.021228014727883e-05, + "loss": 3.1715, + "step": 12683 + }, + { + "epoch": 1.0168367040147521, + "grad_norm": 1.7310277223587036, + "learning_rate": 1.0210981991027995e-05, + "loss": 3.2036, + "step": 12684 + }, + { + "epoch": 1.0169168787957747, + "grad_norm": 1.7821755409240723, + "learning_rate": 1.0209683831220079e-05, + "loss": 3.2202, + "step": 12685 + }, + { + "epoch": 1.0169970535767974, + "grad_norm": 1.7517082691192627, + "learning_rate": 1.0208385667876966e-05, + "loss": 3.1718, + "step": 12686 + }, + { + "epoch": 1.01707722835782, + "grad_norm": 1.6732735633850098, + "learning_rate": 1.0207087501020543e-05, + "loss": 3.1579, + "step": 12687 + }, + { + "epoch": 1.0171574031388426, + "grad_norm": 1.808660864830017, + "learning_rate": 1.0205789330672702e-05, + "loss": 3.2222, + "step": 12688 + }, + { + "epoch": 1.0172375779198652, + "grad_norm": 2.017169952392578, + "learning_rate": 1.020449115685532e-05, + "loss": 3.2477, + "step": 12689 + }, + { + "epoch": 1.0173177527008879, + "grad_norm": 1.6464385986328125, + "learning_rate": 1.0203192979590296e-05, + "loss": 3.1321, + "step": 12690 + }, + { + "epoch": 1.0173979274819105, + "grad_norm": 1.8644887208938599, + "learning_rate": 1.0201894798899504e-05, + "loss": 3.2358, + "step": 12691 + }, + { + "epoch": 1.0174781022629331, + "grad_norm": 1.860089659690857, + "learning_rate": 1.0200596614804837e-05, + "loss": 3.1592, + "step": 12692 + }, + { + "epoch": 1.0175582770439557, + "grad_norm": 1.8428430557250977, + "learning_rate": 1.0199298427328187e-05, + "loss": 3.1916, + "step": 12693 + }, + { + "epoch": 1.0176384518249784, + "grad_norm": 0.7477014064788818, + "learning_rate": 1.0198000236491429e-05, + "loss": 2.4752, + "step": 12694 + }, + { + "epoch": 1.017718626606001, + "grad_norm": 1.703230857849121, + "learning_rate": 1.0196702042316462e-05, + "loss": 3.2334, + "step": 12695 + }, + { + "epoch": 1.0177988013870236, + "grad_norm": 1.7578680515289307, + "learning_rate": 1.0195403844825165e-05, + "loss": 3.1546, + "step": 12696 + }, + { + "epoch": 1.0178789761680462, + "grad_norm": 1.791935920715332, + "learning_rate": 1.019410564403943e-05, + "loss": 3.2203, + "step": 12697 + }, + { + "epoch": 1.017959150949069, + "grad_norm": 1.8768748044967651, + "learning_rate": 1.019280743998114e-05, + "loss": 3.3301, + "step": 12698 + }, + { + "epoch": 1.0180393257300917, + "grad_norm": 1.789212703704834, + "learning_rate": 1.0191509232672185e-05, + "loss": 3.2098, + "step": 12699 + }, + { + "epoch": 1.0181195005111143, + "grad_norm": 1.8407084941864014, + "learning_rate": 1.0190211022134454e-05, + "loss": 3.2671, + "step": 12700 + }, + { + "epoch": 1.018199675292137, + "grad_norm": 1.700170874595642, + "learning_rate": 1.018891280838983e-05, + "loss": 3.2682, + "step": 12701 + }, + { + "epoch": 1.0182798500731596, + "grad_norm": 1.6657938957214355, + "learning_rate": 1.0187614591460204e-05, + "loss": 3.1343, + "step": 12702 + }, + { + "epoch": 1.0183600248541822, + "grad_norm": 1.7970854043960571, + "learning_rate": 1.0186316371367462e-05, + "loss": 3.1796, + "step": 12703 + }, + { + "epoch": 1.0184401996352048, + "grad_norm": 1.7596572637557983, + "learning_rate": 1.018501814813349e-05, + "loss": 3.2161, + "step": 12704 + }, + { + "epoch": 1.0185203744162274, + "grad_norm": 2.0336599349975586, + "learning_rate": 1.018371992178018e-05, + "loss": 3.2079, + "step": 12705 + }, + { + "epoch": 1.01860054919725, + "grad_norm": 1.7719670534133911, + "learning_rate": 1.0182421692329413e-05, + "loss": 3.1893, + "step": 12706 + }, + { + "epoch": 1.0186807239782727, + "grad_norm": 1.8284437656402588, + "learning_rate": 1.0181123459803083e-05, + "loss": 3.213, + "step": 12707 + }, + { + "epoch": 1.0187608987592953, + "grad_norm": 1.926045536994934, + "learning_rate": 1.0179825224223076e-05, + "loss": 3.1879, + "step": 12708 + }, + { + "epoch": 1.018841073540318, + "grad_norm": 1.7327181100845337, + "learning_rate": 1.0178526985611278e-05, + "loss": 3.1976, + "step": 12709 + }, + { + "epoch": 1.0189212483213406, + "grad_norm": 1.7642595767974854, + "learning_rate": 1.017722874398958e-05, + "loss": 3.2475, + "step": 12710 + }, + { + "epoch": 1.0190014231023632, + "grad_norm": 1.813133955001831, + "learning_rate": 1.0175930499379864e-05, + "loss": 3.1591, + "step": 12711 + }, + { + "epoch": 1.0190815978833858, + "grad_norm": 1.7221348285675049, + "learning_rate": 1.0174632251804023e-05, + "loss": 3.1398, + "step": 12712 + }, + { + "epoch": 1.0191617726644084, + "grad_norm": 1.9214439392089844, + "learning_rate": 1.0173334001283947e-05, + "loss": 3.1594, + "step": 12713 + }, + { + "epoch": 1.019241947445431, + "grad_norm": 1.7351359128952026, + "learning_rate": 1.0172035747841517e-05, + "loss": 3.1381, + "step": 12714 + }, + { + "epoch": 1.0193221222264537, + "grad_norm": 1.7916538715362549, + "learning_rate": 1.0170737491498628e-05, + "loss": 3.2025, + "step": 12715 + }, + { + "epoch": 1.0194022970074763, + "grad_norm": 1.7348779439926147, + "learning_rate": 1.0169439232277162e-05, + "loss": 3.2208, + "step": 12716 + }, + { + "epoch": 1.019482471788499, + "grad_norm": 1.8072450160980225, + "learning_rate": 1.0168140970199012e-05, + "loss": 3.174, + "step": 12717 + }, + { + "epoch": 1.0195626465695216, + "grad_norm": 1.7310298681259155, + "learning_rate": 1.0166842705286063e-05, + "loss": 3.1951, + "step": 12718 + }, + { + "epoch": 1.0196428213505442, + "grad_norm": 1.7723009586334229, + "learning_rate": 1.0165544437560207e-05, + "loss": 3.215, + "step": 12719 + }, + { + "epoch": 1.0197229961315668, + "grad_norm": 1.7557379007339478, + "learning_rate": 1.016424616704333e-05, + "loss": 3.144, + "step": 12720 + }, + { + "epoch": 1.0198031709125894, + "grad_norm": 1.7297106981277466, + "learning_rate": 1.016294789375732e-05, + "loss": 3.1083, + "step": 12721 + }, + { + "epoch": 1.019883345693612, + "grad_norm": 1.8196561336517334, + "learning_rate": 1.0161649617724063e-05, + "loss": 3.2726, + "step": 12722 + }, + { + "epoch": 1.0199635204746347, + "grad_norm": 1.640204668045044, + "learning_rate": 1.0160351338965455e-05, + "loss": 3.1941, + "step": 12723 + }, + { + "epoch": 1.0200436952556573, + "grad_norm": 1.832187533378601, + "learning_rate": 1.0159053057503377e-05, + "loss": 3.1704, + "step": 12724 + }, + { + "epoch": 1.02012387003668, + "grad_norm": 1.7199311256408691, + "learning_rate": 1.0157754773359723e-05, + "loss": 3.1597, + "step": 12725 + }, + { + "epoch": 1.0202040448177025, + "grad_norm": 1.745633840560913, + "learning_rate": 1.0156456486556377e-05, + "loss": 3.2034, + "step": 12726 + }, + { + "epoch": 1.0202842195987252, + "grad_norm": 1.7526036500930786, + "learning_rate": 1.0155158197115228e-05, + "loss": 3.2367, + "step": 12727 + }, + { + "epoch": 1.0203643943797478, + "grad_norm": 1.797454833984375, + "learning_rate": 1.0153859905058173e-05, + "loss": 3.2325, + "step": 12728 + }, + { + "epoch": 1.0204445691607704, + "grad_norm": 1.7563103437423706, + "learning_rate": 1.0152561610407089e-05, + "loss": 3.1149, + "step": 12729 + }, + { + "epoch": 1.020524743941793, + "grad_norm": 1.8079217672348022, + "learning_rate": 1.015126331318387e-05, + "loss": 3.2159, + "step": 12730 + }, + { + "epoch": 1.0206049187228157, + "grad_norm": 1.8828351497650146, + "learning_rate": 1.0149965013410406e-05, + "loss": 3.334, + "step": 12731 + }, + { + "epoch": 1.0206850935038383, + "grad_norm": 1.7775633335113525, + "learning_rate": 1.0148666711108583e-05, + "loss": 3.2219, + "step": 12732 + }, + { + "epoch": 1.020765268284861, + "grad_norm": 1.87154221534729, + "learning_rate": 1.0147368406300295e-05, + "loss": 3.1157, + "step": 12733 + }, + { + "epoch": 1.0208454430658835, + "grad_norm": 0.7349250316619873, + "learning_rate": 1.0146070099007425e-05, + "loss": 2.4575, + "step": 12734 + }, + { + "epoch": 1.0209256178469062, + "grad_norm": 1.7524114847183228, + "learning_rate": 1.0144771789251867e-05, + "loss": 3.2292, + "step": 12735 + }, + { + "epoch": 1.0210057926279288, + "grad_norm": 1.7400556802749634, + "learning_rate": 1.0143473477055505e-05, + "loss": 3.194, + "step": 12736 + }, + { + "epoch": 1.0210859674089514, + "grad_norm": 1.7406266927719116, + "learning_rate": 1.014217516244023e-05, + "loss": 3.1631, + "step": 12737 + }, + { + "epoch": 1.021166142189974, + "grad_norm": 1.753684639930725, + "learning_rate": 1.0140876845427936e-05, + "loss": 3.198, + "step": 12738 + }, + { + "epoch": 1.0212463169709969, + "grad_norm": 1.7767045497894287, + "learning_rate": 1.0139578526040505e-05, + "loss": 3.1594, + "step": 12739 + }, + { + "epoch": 1.0213264917520195, + "grad_norm": 1.7334157228469849, + "learning_rate": 1.013828020429983e-05, + "loss": 3.1673, + "step": 12740 + }, + { + "epoch": 1.0214066665330421, + "grad_norm": 1.7362983226776123, + "learning_rate": 1.0136981880227799e-05, + "loss": 3.2014, + "step": 12741 + }, + { + "epoch": 1.0214868413140648, + "grad_norm": 1.8405851125717163, + "learning_rate": 1.0135683553846303e-05, + "loss": 3.27, + "step": 12742 + }, + { + "epoch": 1.0215670160950874, + "grad_norm": 2.2840399742126465, + "learning_rate": 1.013438522517723e-05, + "loss": 3.1766, + "step": 12743 + }, + { + "epoch": 1.02164719087611, + "grad_norm": 1.7590386867523193, + "learning_rate": 1.0133086894242467e-05, + "loss": 3.1942, + "step": 12744 + }, + { + "epoch": 1.0217273656571326, + "grad_norm": 1.7859314680099487, + "learning_rate": 1.0131788561063908e-05, + "loss": 3.1606, + "step": 12745 + }, + { + "epoch": 1.0218075404381552, + "grad_norm": 1.7065458297729492, + "learning_rate": 1.0130490225663441e-05, + "loss": 3.2096, + "step": 12746 + }, + { + "epoch": 1.0218877152191779, + "grad_norm": 1.7561639547348022, + "learning_rate": 1.0129191888062952e-05, + "loss": 3.1655, + "step": 12747 + }, + { + "epoch": 1.0219678900002005, + "grad_norm": 1.6918177604675293, + "learning_rate": 1.0127893548284338e-05, + "loss": 3.1398, + "step": 12748 + }, + { + "epoch": 1.0220480647812231, + "grad_norm": 1.8037421703338623, + "learning_rate": 1.012659520634948e-05, + "loss": 3.2443, + "step": 12749 + }, + { + "epoch": 1.0221282395622457, + "grad_norm": 1.8797590732574463, + "learning_rate": 1.0125296862280272e-05, + "loss": 3.1956, + "step": 12750 + }, + { + "epoch": 1.0222084143432684, + "grad_norm": 1.8146768808364868, + "learning_rate": 1.0123998516098602e-05, + "loss": 3.0535, + "step": 12751 + }, + { + "epoch": 1.022288589124291, + "grad_norm": 1.8655632734298706, + "learning_rate": 1.0122700167826362e-05, + "loss": 3.2159, + "step": 12752 + }, + { + "epoch": 1.0223687639053136, + "grad_norm": 1.8189356327056885, + "learning_rate": 1.0121401817485443e-05, + "loss": 3.1477, + "step": 12753 + }, + { + "epoch": 1.0224489386863362, + "grad_norm": 1.7799956798553467, + "learning_rate": 1.0120103465097727e-05, + "loss": 3.2083, + "step": 12754 + }, + { + "epoch": 1.0225291134673589, + "grad_norm": 1.742227554321289, + "learning_rate": 1.0118805110685113e-05, + "loss": 3.2094, + "step": 12755 + }, + { + "epoch": 1.0226092882483815, + "grad_norm": 1.662978172302246, + "learning_rate": 1.0117506754269486e-05, + "loss": 3.2463, + "step": 12756 + }, + { + "epoch": 1.0226894630294041, + "grad_norm": 1.8596590757369995, + "learning_rate": 1.0116208395872735e-05, + "loss": 3.1444, + "step": 12757 + }, + { + "epoch": 1.0227696378104267, + "grad_norm": 1.7490618228912354, + "learning_rate": 1.0114910035516753e-05, + "loss": 3.1375, + "step": 12758 + }, + { + "epoch": 1.0228498125914494, + "grad_norm": 1.7417062520980835, + "learning_rate": 1.0113611673223427e-05, + "loss": 3.2292, + "step": 12759 + }, + { + "epoch": 1.022929987372472, + "grad_norm": 1.6877566576004028, + "learning_rate": 1.0112313309014648e-05, + "loss": 3.2385, + "step": 12760 + }, + { + "epoch": 1.0230101621534946, + "grad_norm": 0.735986053943634, + "learning_rate": 1.011101494291231e-05, + "loss": 2.4805, + "step": 12761 + }, + { + "epoch": 1.0230903369345172, + "grad_norm": 1.8109238147735596, + "learning_rate": 1.0109716574938293e-05, + "loss": 3.2692, + "step": 12762 + }, + { + "epoch": 1.0231705117155399, + "grad_norm": 0.7283927798271179, + "learning_rate": 1.0108418205114499e-05, + "loss": 2.4871, + "step": 12763 + }, + { + "epoch": 1.0232506864965625, + "grad_norm": 1.7632434368133545, + "learning_rate": 1.0107119833462809e-05, + "loss": 3.1616, + "step": 12764 + }, + { + "epoch": 1.023330861277585, + "grad_norm": 1.7285269498825073, + "learning_rate": 1.0105821460005116e-05, + "loss": 3.2171, + "step": 12765 + }, + { + "epoch": 1.0234110360586077, + "grad_norm": 1.8147919178009033, + "learning_rate": 1.0104523084763311e-05, + "loss": 3.2128, + "step": 12766 + }, + { + "epoch": 1.0234912108396303, + "grad_norm": 1.7947262525558472, + "learning_rate": 1.0103224707759285e-05, + "loss": 3.2466, + "step": 12767 + }, + { + "epoch": 1.023571385620653, + "grad_norm": 1.8185882568359375, + "learning_rate": 1.0101926329014925e-05, + "loss": 3.2249, + "step": 12768 + }, + { + "epoch": 1.0236515604016756, + "grad_norm": 1.779045581817627, + "learning_rate": 1.0100627948552125e-05, + "loss": 3.2615, + "step": 12769 + }, + { + "epoch": 1.0237317351826982, + "grad_norm": 1.728365421295166, + "learning_rate": 1.009932956639277e-05, + "loss": 3.1766, + "step": 12770 + }, + { + "epoch": 1.0238119099637208, + "grad_norm": 1.8389931917190552, + "learning_rate": 1.0098031182558758e-05, + "loss": 3.1807, + "step": 12771 + }, + { + "epoch": 1.0238920847447435, + "grad_norm": 1.817368984222412, + "learning_rate": 1.0096732797071973e-05, + "loss": 3.1465, + "step": 12772 + }, + { + "epoch": 1.023972259525766, + "grad_norm": 1.744098424911499, + "learning_rate": 1.0095434409954309e-05, + "loss": 3.079, + "step": 12773 + }, + { + "epoch": 1.0240524343067887, + "grad_norm": 1.7106279134750366, + "learning_rate": 1.009413602122765e-05, + "loss": 3.142, + "step": 12774 + }, + { + "epoch": 1.0241326090878113, + "grad_norm": 1.796257734298706, + "learning_rate": 1.0092837630913892e-05, + "loss": 3.21, + "step": 12775 + }, + { + "epoch": 1.024212783868834, + "grad_norm": 1.7278473377227783, + "learning_rate": 1.0091539239034927e-05, + "loss": 3.1785, + "step": 12776 + }, + { + "epoch": 1.0242929586498566, + "grad_norm": 1.7879817485809326, + "learning_rate": 1.009024084561264e-05, + "loss": 3.1746, + "step": 12777 + }, + { + "epoch": 1.0243731334308792, + "grad_norm": 1.8559807538986206, + "learning_rate": 1.0088942450668925e-05, + "loss": 3.2354, + "step": 12778 + }, + { + "epoch": 1.024453308211902, + "grad_norm": 1.9732896089553833, + "learning_rate": 1.0087644054225672e-05, + "loss": 3.1764, + "step": 12779 + }, + { + "epoch": 1.0245334829929247, + "grad_norm": 1.8301920890808105, + "learning_rate": 1.008634565630477e-05, + "loss": 3.207, + "step": 12780 + }, + { + "epoch": 1.0246136577739473, + "grad_norm": 1.7752593755722046, + "learning_rate": 1.0085047256928113e-05, + "loss": 3.2495, + "step": 12781 + }, + { + "epoch": 1.02469383255497, + "grad_norm": 1.684991717338562, + "learning_rate": 1.0083748856117585e-05, + "loss": 3.2356, + "step": 12782 + }, + { + "epoch": 1.0247740073359926, + "grad_norm": 1.7752646207809448, + "learning_rate": 1.0082450453895085e-05, + "loss": 3.1832, + "step": 12783 + }, + { + "epoch": 1.0248541821170152, + "grad_norm": 1.7708408832550049, + "learning_rate": 1.00811520502825e-05, + "loss": 3.1719, + "step": 12784 + }, + { + "epoch": 1.0249343568980378, + "grad_norm": 1.7451856136322021, + "learning_rate": 1.0079853645301717e-05, + "loss": 3.2609, + "step": 12785 + }, + { + "epoch": 1.0250145316790604, + "grad_norm": 1.9136399030685425, + "learning_rate": 1.0078555238974629e-05, + "loss": 3.2312, + "step": 12786 + }, + { + "epoch": 1.025094706460083, + "grad_norm": 1.8077226877212524, + "learning_rate": 1.007725683132313e-05, + "loss": 3.2772, + "step": 12787 + }, + { + "epoch": 1.0251748812411057, + "grad_norm": 1.7383934259414673, + "learning_rate": 1.0075958422369108e-05, + "loss": 3.1661, + "step": 12788 + }, + { + "epoch": 1.0252550560221283, + "grad_norm": 1.7822415828704834, + "learning_rate": 1.0074660012134453e-05, + "loss": 3.115, + "step": 12789 + }, + { + "epoch": 1.025335230803151, + "grad_norm": 1.8544375896453857, + "learning_rate": 1.0073361600641054e-05, + "loss": 3.0726, + "step": 12790 + }, + { + "epoch": 1.0254154055841735, + "grad_norm": 1.7083585262298584, + "learning_rate": 1.0072063187910807e-05, + "loss": 3.1617, + "step": 12791 + }, + { + "epoch": 1.0254955803651962, + "grad_norm": 1.7926554679870605, + "learning_rate": 1.00707647739656e-05, + "loss": 3.1211, + "step": 12792 + }, + { + "epoch": 1.0255757551462188, + "grad_norm": 1.7750849723815918, + "learning_rate": 1.0069466358827325e-05, + "loss": 3.1533, + "step": 12793 + }, + { + "epoch": 1.0256559299272414, + "grad_norm": 1.7019829750061035, + "learning_rate": 1.006816794251787e-05, + "loss": 3.2195, + "step": 12794 + }, + { + "epoch": 1.025736104708264, + "grad_norm": 0.7345090508460999, + "learning_rate": 1.0066869525059126e-05, + "loss": 2.4932, + "step": 12795 + }, + { + "epoch": 1.0258162794892867, + "grad_norm": 1.6742136478424072, + "learning_rate": 1.0065571106472988e-05, + "loss": 3.1554, + "step": 12796 + }, + { + "epoch": 1.0258964542703093, + "grad_norm": 0.6706973314285278, + "learning_rate": 1.0064272686781342e-05, + "loss": 2.4738, + "step": 12797 + }, + { + "epoch": 1.025976629051332, + "grad_norm": 1.8821864128112793, + "learning_rate": 1.0062974266006085e-05, + "loss": 3.2088, + "step": 12798 + }, + { + "epoch": 1.0260568038323545, + "grad_norm": 1.636700987815857, + "learning_rate": 1.00616758441691e-05, + "loss": 3.1095, + "step": 12799 + }, + { + "epoch": 1.0261369786133772, + "grad_norm": 2.0312249660491943, + "learning_rate": 1.0060377421292285e-05, + "loss": 3.1924, + "step": 12800 + }, + { + "epoch": 1.0262171533943998, + "grad_norm": 1.7741661071777344, + "learning_rate": 1.0059078997397526e-05, + "loss": 3.2335, + "step": 12801 + }, + { + "epoch": 1.0262973281754224, + "grad_norm": 1.8048195838928223, + "learning_rate": 1.0057780572506717e-05, + "loss": 3.2887, + "step": 12802 + }, + { + "epoch": 1.026377502956445, + "grad_norm": 1.7345702648162842, + "learning_rate": 1.0056482146641749e-05, + "loss": 3.3108, + "step": 12803 + }, + { + "epoch": 1.0264576777374677, + "grad_norm": 1.7814695835113525, + "learning_rate": 1.0055183719824508e-05, + "loss": 3.2285, + "step": 12804 + }, + { + "epoch": 1.0265378525184903, + "grad_norm": 0.7487102150917053, + "learning_rate": 1.005388529207689e-05, + "loss": 2.4884, + "step": 12805 + }, + { + "epoch": 1.026618027299513, + "grad_norm": 0.7503553032875061, + "learning_rate": 1.0052586863420788e-05, + "loss": 2.4764, + "step": 12806 + }, + { + "epoch": 1.0266982020805355, + "grad_norm": 1.8162870407104492, + "learning_rate": 1.0051288433878087e-05, + "loss": 3.2679, + "step": 12807 + }, + { + "epoch": 1.0267783768615582, + "grad_norm": 1.7453745603561401, + "learning_rate": 1.0049990003470685e-05, + "loss": 3.2329, + "step": 12808 + }, + { + "epoch": 1.0268585516425808, + "grad_norm": 1.681419014930725, + "learning_rate": 1.0048691572220464e-05, + "loss": 3.0845, + "step": 12809 + }, + { + "epoch": 1.0269387264236034, + "grad_norm": 1.764737844467163, + "learning_rate": 1.0047393140149322e-05, + "loss": 3.2114, + "step": 12810 + }, + { + "epoch": 1.027018901204626, + "grad_norm": 1.6516462564468384, + "learning_rate": 1.004609470727915e-05, + "loss": 3.2268, + "step": 12811 + }, + { + "epoch": 1.0270990759856486, + "grad_norm": 1.8608853816986084, + "learning_rate": 1.0044796273631837e-05, + "loss": 3.186, + "step": 12812 + }, + { + "epoch": 1.0271792507666713, + "grad_norm": 1.8873510360717773, + "learning_rate": 1.0043497839229275e-05, + "loss": 3.3657, + "step": 12813 + }, + { + "epoch": 1.027259425547694, + "grad_norm": 1.7610676288604736, + "learning_rate": 1.0042199404093353e-05, + "loss": 3.1915, + "step": 12814 + }, + { + "epoch": 1.0273396003287165, + "grad_norm": 1.745078682899475, + "learning_rate": 1.0040900968245963e-05, + "loss": 3.185, + "step": 12815 + }, + { + "epoch": 1.0274197751097391, + "grad_norm": 1.6082333326339722, + "learning_rate": 1.0039602531709e-05, + "loss": 3.2234, + "step": 12816 + }, + { + "epoch": 1.0274999498907618, + "grad_norm": 1.722447156906128, + "learning_rate": 1.0038304094504351e-05, + "loss": 3.2297, + "step": 12817 + }, + { + "epoch": 1.0275801246717844, + "grad_norm": 1.867624282836914, + "learning_rate": 1.0037005656653911e-05, + "loss": 3.2336, + "step": 12818 + }, + { + "epoch": 1.027660299452807, + "grad_norm": 1.797627329826355, + "learning_rate": 1.0035707218179565e-05, + "loss": 3.2802, + "step": 12819 + }, + { + "epoch": 1.0277404742338296, + "grad_norm": 1.8783832788467407, + "learning_rate": 1.0034408779103209e-05, + "loss": 3.2327, + "step": 12820 + }, + { + "epoch": 1.0278206490148525, + "grad_norm": 1.8337033987045288, + "learning_rate": 1.0033110339446734e-05, + "loss": 3.1149, + "step": 12821 + }, + { + "epoch": 1.027900823795875, + "grad_norm": 0.7397005558013916, + "learning_rate": 1.003181189923203e-05, + "loss": 2.4775, + "step": 12822 + }, + { + "epoch": 1.0279809985768977, + "grad_norm": 1.691192865371704, + "learning_rate": 1.0030513458480988e-05, + "loss": 3.1061, + "step": 12823 + }, + { + "epoch": 1.0280611733579204, + "grad_norm": 1.724416732788086, + "learning_rate": 1.0029215017215499e-05, + "loss": 3.2955, + "step": 12824 + }, + { + "epoch": 1.028141348138943, + "grad_norm": 1.7992972135543823, + "learning_rate": 1.0027916575457455e-05, + "loss": 3.1137, + "step": 12825 + }, + { + "epoch": 1.0282215229199656, + "grad_norm": 1.7134110927581787, + "learning_rate": 1.0026618133228751e-05, + "loss": 3.0928, + "step": 12826 + }, + { + "epoch": 1.0283016977009882, + "grad_norm": 1.761818766593933, + "learning_rate": 1.0025319690551272e-05, + "loss": 3.1891, + "step": 12827 + }, + { + "epoch": 1.0283818724820109, + "grad_norm": 1.6546812057495117, + "learning_rate": 1.0024021247446913e-05, + "loss": 3.1943, + "step": 12828 + }, + { + "epoch": 1.0284620472630335, + "grad_norm": 1.600722312927246, + "learning_rate": 1.0022722803937561e-05, + "loss": 3.2392, + "step": 12829 + }, + { + "epoch": 1.028542222044056, + "grad_norm": 0.7135890126228333, + "learning_rate": 1.0021424360045113e-05, + "loss": 2.4775, + "step": 12830 + }, + { + "epoch": 1.0286223968250787, + "grad_norm": 1.7593059539794922, + "learning_rate": 1.002012591579146e-05, + "loss": 3.2491, + "step": 12831 + }, + { + "epoch": 1.0287025716061013, + "grad_norm": 1.7411532402038574, + "learning_rate": 1.0018827471198486e-05, + "loss": 3.2166, + "step": 12832 + }, + { + "epoch": 1.028782746387124, + "grad_norm": 1.8419691324234009, + "learning_rate": 1.0017529026288092e-05, + "loss": 3.2709, + "step": 12833 + }, + { + "epoch": 1.0288629211681466, + "grad_norm": 1.79531729221344, + "learning_rate": 1.0016230581082162e-05, + "loss": 3.2772, + "step": 12834 + }, + { + "epoch": 1.0289430959491692, + "grad_norm": 0.696980893611908, + "learning_rate": 1.0014932135602591e-05, + "loss": 2.4664, + "step": 12835 + }, + { + "epoch": 1.0290232707301918, + "grad_norm": 1.7652106285095215, + "learning_rate": 1.0013633689871271e-05, + "loss": 3.1151, + "step": 12836 + }, + { + "epoch": 1.0291034455112145, + "grad_norm": 1.7434059381484985, + "learning_rate": 1.0012335243910091e-05, + "loss": 3.2127, + "step": 12837 + }, + { + "epoch": 1.029183620292237, + "grad_norm": 2.1724460124969482, + "learning_rate": 1.0011036797740944e-05, + "loss": 3.1591, + "step": 12838 + }, + { + "epoch": 1.0292637950732597, + "grad_norm": 1.8321019411087036, + "learning_rate": 1.0009738351385718e-05, + "loss": 3.1777, + "step": 12839 + }, + { + "epoch": 1.0293439698542823, + "grad_norm": 1.7102028131484985, + "learning_rate": 1.0008439904866308e-05, + "loss": 3.1393, + "step": 12840 + }, + { + "epoch": 1.029424144635305, + "grad_norm": 1.8465917110443115, + "learning_rate": 1.0007141458204606e-05, + "loss": 3.1953, + "step": 12841 + }, + { + "epoch": 1.0295043194163276, + "grad_norm": 1.8249261379241943, + "learning_rate": 1.0005843011422498e-05, + "loss": 3.1579, + "step": 12842 + }, + { + "epoch": 1.0295844941973502, + "grad_norm": 1.7609925270080566, + "learning_rate": 1.0004544564541883e-05, + "loss": 3.2075, + "step": 12843 + }, + { + "epoch": 1.0296646689783728, + "grad_norm": 1.873964786529541, + "learning_rate": 1.0003246117584645e-05, + "loss": 3.2422, + "step": 12844 + }, + { + "epoch": 1.0297448437593955, + "grad_norm": 0.709227442741394, + "learning_rate": 1.0001947670572678e-05, + "loss": 2.4885, + "step": 12845 + }, + { + "epoch": 1.029825018540418, + "grad_norm": 1.8581148386001587, + "learning_rate": 1.0000649223527876e-05, + "loss": 3.1085, + "step": 12846 + }, + { + "epoch": 1.0299051933214407, + "grad_norm": 1.7242796421051025, + "learning_rate": 9.999350776472126e-06, + "loss": 3.3167, + "step": 12847 + }, + { + "epoch": 1.0299853681024633, + "grad_norm": 1.7855229377746582, + "learning_rate": 9.998052329427326e-06, + "loss": 3.3036, + "step": 12848 + }, + { + "epoch": 1.030065542883486, + "grad_norm": 1.8911027908325195, + "learning_rate": 9.996753882415359e-06, + "loss": 3.2264, + "step": 12849 + }, + { + "epoch": 1.0301457176645086, + "grad_norm": 1.7991931438446045, + "learning_rate": 9.995455435458122e-06, + "loss": 3.1848, + "step": 12850 + }, + { + "epoch": 1.0302258924455312, + "grad_norm": 1.8148878812789917, + "learning_rate": 9.994156988577504e-06, + "loss": 3.2976, + "step": 12851 + }, + { + "epoch": 1.0303060672265538, + "grad_norm": 1.9079073667526245, + "learning_rate": 9.992858541795398e-06, + "loss": 3.2022, + "step": 12852 + }, + { + "epoch": 1.0303862420075764, + "grad_norm": 1.8488863706588745, + "learning_rate": 9.991560095133696e-06, + "loss": 3.2346, + "step": 12853 + }, + { + "epoch": 1.030466416788599, + "grad_norm": 1.8300222158432007, + "learning_rate": 9.990261648614285e-06, + "loss": 3.2504, + "step": 12854 + }, + { + "epoch": 1.0305465915696217, + "grad_norm": 1.7826595306396484, + "learning_rate": 9.988963202259058e-06, + "loss": 3.2262, + "step": 12855 + }, + { + "epoch": 1.0306267663506443, + "grad_norm": 0.7143580317497253, + "learning_rate": 9.987664756089912e-06, + "loss": 2.4841, + "step": 12856 + }, + { + "epoch": 1.030706941131667, + "grad_norm": 1.822771668434143, + "learning_rate": 9.986366310128732e-06, + "loss": 3.307, + "step": 12857 + }, + { + "epoch": 1.0307871159126896, + "grad_norm": 1.7382947206497192, + "learning_rate": 9.98506786439741e-06, + "loss": 3.1904, + "step": 12858 + }, + { + "epoch": 1.0308672906937122, + "grad_norm": 1.6823968887329102, + "learning_rate": 9.98376941891784e-06, + "loss": 3.1205, + "step": 12859 + }, + { + "epoch": 1.0309474654747348, + "grad_norm": 1.7573200464248657, + "learning_rate": 9.98247097371191e-06, + "loss": 3.2327, + "step": 12860 + }, + { + "epoch": 1.0310276402557577, + "grad_norm": 1.7280986309051514, + "learning_rate": 9.981172528801516e-06, + "loss": 3.1917, + "step": 12861 + }, + { + "epoch": 1.0311078150367803, + "grad_norm": 1.681454062461853, + "learning_rate": 9.979874084208544e-06, + "loss": 3.113, + "step": 12862 + }, + { + "epoch": 1.031187989817803, + "grad_norm": 1.6914150714874268, + "learning_rate": 9.978575639954889e-06, + "loss": 3.1926, + "step": 12863 + }, + { + "epoch": 1.0312681645988255, + "grad_norm": 1.7591031789779663, + "learning_rate": 9.97727719606244e-06, + "loss": 3.1558, + "step": 12864 + }, + { + "epoch": 1.0313483393798482, + "grad_norm": 1.813920497894287, + "learning_rate": 9.97597875255309e-06, + "loss": 3.2288, + "step": 12865 + }, + { + "epoch": 1.0314285141608708, + "grad_norm": 1.7641154527664185, + "learning_rate": 9.974680309448732e-06, + "loss": 3.1749, + "step": 12866 + }, + { + "epoch": 1.0315086889418934, + "grad_norm": 1.7303709983825684, + "learning_rate": 9.973381866771252e-06, + "loss": 3.1642, + "step": 12867 + }, + { + "epoch": 1.031588863722916, + "grad_norm": 1.8486546277999878, + "learning_rate": 9.972083424542546e-06, + "loss": 3.2288, + "step": 12868 + }, + { + "epoch": 1.0316690385039387, + "grad_norm": 1.809801459312439, + "learning_rate": 9.970784982784504e-06, + "loss": 3.2926, + "step": 12869 + }, + { + "epoch": 1.0317492132849613, + "grad_norm": 1.7598960399627686, + "learning_rate": 9.969486541519014e-06, + "loss": 3.0899, + "step": 12870 + }, + { + "epoch": 1.031829388065984, + "grad_norm": 1.7650872468948364, + "learning_rate": 9.968188100767975e-06, + "loss": 3.1409, + "step": 12871 + }, + { + "epoch": 1.0319095628470065, + "grad_norm": 2.106302499771118, + "learning_rate": 9.96688966055327e-06, + "loss": 3.1423, + "step": 12872 + }, + { + "epoch": 1.0319897376280291, + "grad_norm": 1.621870756149292, + "learning_rate": 9.965591220896796e-06, + "loss": 3.1433, + "step": 12873 + }, + { + "epoch": 1.0320699124090518, + "grad_norm": 1.8326647281646729, + "learning_rate": 9.964292781820438e-06, + "loss": 3.2346, + "step": 12874 + }, + { + "epoch": 1.0321500871900744, + "grad_norm": 1.8190549612045288, + "learning_rate": 9.962994343346092e-06, + "loss": 3.0321, + "step": 12875 + }, + { + "epoch": 1.032230261971097, + "grad_norm": 1.8351809978485107, + "learning_rate": 9.961695905495652e-06, + "loss": 3.1774, + "step": 12876 + }, + { + "epoch": 1.0323104367521196, + "grad_norm": 1.8337328433990479, + "learning_rate": 9.960397468291002e-06, + "loss": 3.2021, + "step": 12877 + }, + { + "epoch": 1.0323906115331423, + "grad_norm": 1.8060190677642822, + "learning_rate": 9.959099031754039e-06, + "loss": 3.1975, + "step": 12878 + }, + { + "epoch": 1.032470786314165, + "grad_norm": 1.8486480712890625, + "learning_rate": 9.95780059590665e-06, + "loss": 3.2107, + "step": 12879 + }, + { + "epoch": 1.0325509610951875, + "grad_norm": 1.8203400373458862, + "learning_rate": 9.956502160770728e-06, + "loss": 3.1144, + "step": 12880 + }, + { + "epoch": 1.0326311358762101, + "grad_norm": 1.731322169303894, + "learning_rate": 9.955203726368166e-06, + "loss": 3.1429, + "step": 12881 + }, + { + "epoch": 1.0327113106572328, + "grad_norm": 1.7500277757644653, + "learning_rate": 9.953905292720853e-06, + "loss": 3.2699, + "step": 12882 + }, + { + "epoch": 1.0327914854382554, + "grad_norm": 1.8570224046707153, + "learning_rate": 9.95260685985068e-06, + "loss": 3.2725, + "step": 12883 + }, + { + "epoch": 1.032871660219278, + "grad_norm": 1.818881630897522, + "learning_rate": 9.951308427779537e-06, + "loss": 3.1481, + "step": 12884 + }, + { + "epoch": 1.0329518350003006, + "grad_norm": 1.8222460746765137, + "learning_rate": 9.95000999652932e-06, + "loss": 3.228, + "step": 12885 + }, + { + "epoch": 1.0330320097813233, + "grad_norm": 1.942894697189331, + "learning_rate": 9.948711566121916e-06, + "loss": 3.2635, + "step": 12886 + }, + { + "epoch": 1.0331121845623459, + "grad_norm": 1.8040746450424194, + "learning_rate": 9.947413136579216e-06, + "loss": 3.2505, + "step": 12887 + }, + { + "epoch": 1.0331923593433685, + "grad_norm": 1.746986985206604, + "learning_rate": 9.946114707923113e-06, + "loss": 3.1274, + "step": 12888 + }, + { + "epoch": 1.0332725341243911, + "grad_norm": 1.9345126152038574, + "learning_rate": 9.944816280175494e-06, + "loss": 3.1295, + "step": 12889 + }, + { + "epoch": 1.0333527089054138, + "grad_norm": 1.7406859397888184, + "learning_rate": 9.943517853358256e-06, + "loss": 3.0726, + "step": 12890 + }, + { + "epoch": 1.0334328836864364, + "grad_norm": 1.7439409494400024, + "learning_rate": 9.942219427493286e-06, + "loss": 3.2065, + "step": 12891 + }, + { + "epoch": 1.033513058467459, + "grad_norm": 1.8024176359176636, + "learning_rate": 9.940921002602477e-06, + "loss": 3.2265, + "step": 12892 + }, + { + "epoch": 1.0335932332484816, + "grad_norm": 1.919291615486145, + "learning_rate": 9.93962257870772e-06, + "loss": 3.2225, + "step": 12893 + }, + { + "epoch": 1.0336734080295042, + "grad_norm": 1.711981177330017, + "learning_rate": 9.938324155830901e-06, + "loss": 3.1075, + "step": 12894 + }, + { + "epoch": 1.0337535828105269, + "grad_norm": 1.864655613899231, + "learning_rate": 9.937025733993919e-06, + "loss": 3.1121, + "step": 12895 + }, + { + "epoch": 1.0338337575915495, + "grad_norm": 1.9219728708267212, + "learning_rate": 9.93572731321866e-06, + "loss": 3.0492, + "step": 12896 + }, + { + "epoch": 1.0339139323725721, + "grad_norm": 1.750619888305664, + "learning_rate": 9.934428893527014e-06, + "loss": 3.1959, + "step": 12897 + }, + { + "epoch": 1.0339941071535947, + "grad_norm": 1.748288631439209, + "learning_rate": 9.933130474940876e-06, + "loss": 3.1636, + "step": 12898 + }, + { + "epoch": 1.0340742819346174, + "grad_norm": 1.7816983461380005, + "learning_rate": 9.931832057482134e-06, + "loss": 3.1121, + "step": 12899 + }, + { + "epoch": 1.03415445671564, + "grad_norm": 1.7516900300979614, + "learning_rate": 9.930533641172678e-06, + "loss": 3.0297, + "step": 12900 + }, + { + "epoch": 1.0342346314966626, + "grad_norm": 1.767250418663025, + "learning_rate": 9.929235226034403e-06, + "loss": 3.2626, + "step": 12901 + }, + { + "epoch": 1.0343148062776855, + "grad_norm": 1.721075415611267, + "learning_rate": 9.927936812089195e-06, + "loss": 3.192, + "step": 12902 + }, + { + "epoch": 1.034394981058708, + "grad_norm": 1.7213914394378662, + "learning_rate": 9.926638399358948e-06, + "loss": 3.2061, + "step": 12903 + }, + { + "epoch": 1.0344751558397307, + "grad_norm": 1.7263107299804688, + "learning_rate": 9.92533998786555e-06, + "loss": 3.1518, + "step": 12904 + }, + { + "epoch": 1.0345553306207533, + "grad_norm": 1.7705724239349365, + "learning_rate": 9.924041577630895e-06, + "loss": 3.1414, + "step": 12905 + }, + { + "epoch": 1.034635505401776, + "grad_norm": 1.7435821294784546, + "learning_rate": 9.922743168676874e-06, + "loss": 3.2195, + "step": 12906 + }, + { + "epoch": 1.0347156801827986, + "grad_norm": 1.9825865030288696, + "learning_rate": 9.921444761025373e-06, + "loss": 3.1382, + "step": 12907 + }, + { + "epoch": 1.0347958549638212, + "grad_norm": 1.7902902364730835, + "learning_rate": 9.920146354698288e-06, + "loss": 3.2263, + "step": 12908 + }, + { + "epoch": 1.0348760297448438, + "grad_norm": 1.8285329341888428, + "learning_rate": 9.918847949717506e-06, + "loss": 3.1455, + "step": 12909 + }, + { + "epoch": 1.0349562045258665, + "grad_norm": 1.7085692882537842, + "learning_rate": 9.917549546104916e-06, + "loss": 3.2174, + "step": 12910 + }, + { + "epoch": 1.035036379306889, + "grad_norm": 1.686715841293335, + "learning_rate": 9.916251143882417e-06, + "loss": 3.1773, + "step": 12911 + }, + { + "epoch": 1.0351165540879117, + "grad_norm": 1.8632770776748657, + "learning_rate": 9.91495274307189e-06, + "loss": 3.2529, + "step": 12912 + }, + { + "epoch": 1.0351967288689343, + "grad_norm": 1.9077719449996948, + "learning_rate": 9.913654343695233e-06, + "loss": 3.2256, + "step": 12913 + }, + { + "epoch": 1.035276903649957, + "grad_norm": 2.185330629348755, + "learning_rate": 9.912355945774331e-06, + "loss": 3.2355, + "step": 12914 + }, + { + "epoch": 1.0353570784309796, + "grad_norm": 1.872327208518982, + "learning_rate": 9.911057549331078e-06, + "loss": 3.2203, + "step": 12915 + }, + { + "epoch": 1.0354372532120022, + "grad_norm": 0.7338137030601501, + "learning_rate": 9.909759154387365e-06, + "loss": 2.4673, + "step": 12916 + }, + { + "epoch": 1.0355174279930248, + "grad_norm": 1.8146847486495972, + "learning_rate": 9.908460760965078e-06, + "loss": 3.2264, + "step": 12917 + }, + { + "epoch": 1.0355976027740474, + "grad_norm": 1.7415872812271118, + "learning_rate": 9.907162369086111e-06, + "loss": 3.3281, + "step": 12918 + }, + { + "epoch": 1.03567777755507, + "grad_norm": 1.7693530321121216, + "learning_rate": 9.905863978772352e-06, + "loss": 3.2474, + "step": 12919 + }, + { + "epoch": 1.0357579523360927, + "grad_norm": 1.9728986024856567, + "learning_rate": 9.904565590045696e-06, + "loss": 3.2051, + "step": 12920 + }, + { + "epoch": 1.0358381271171153, + "grad_norm": 1.857140064239502, + "learning_rate": 9.90326720292803e-06, + "loss": 3.1852, + "step": 12921 + }, + { + "epoch": 1.035918301898138, + "grad_norm": 1.8611501455307007, + "learning_rate": 9.901968817441245e-06, + "loss": 3.2457, + "step": 12922 + }, + { + "epoch": 1.0359984766791606, + "grad_norm": 1.7866441011428833, + "learning_rate": 9.900670433607231e-06, + "loss": 3.1465, + "step": 12923 + }, + { + "epoch": 1.0360786514601832, + "grad_norm": 1.717468500137329, + "learning_rate": 9.899372051447877e-06, + "loss": 3.1476, + "step": 12924 + }, + { + "epoch": 1.0361588262412058, + "grad_norm": 1.8079771995544434, + "learning_rate": 9.898073670985076e-06, + "loss": 3.1914, + "step": 12925 + }, + { + "epoch": 1.0362390010222284, + "grad_norm": 1.7720801830291748, + "learning_rate": 9.896775292240719e-06, + "loss": 3.1251, + "step": 12926 + }, + { + "epoch": 1.036319175803251, + "grad_norm": 1.7226355075836182, + "learning_rate": 9.89547691523669e-06, + "loss": 3.2204, + "step": 12927 + }, + { + "epoch": 1.0363993505842737, + "grad_norm": 0.7000604867935181, + "learning_rate": 9.894178539994887e-06, + "loss": 2.4929, + "step": 12928 + }, + { + "epoch": 1.0364795253652963, + "grad_norm": 1.7913267612457275, + "learning_rate": 9.892880166537194e-06, + "loss": 3.183, + "step": 12929 + }, + { + "epoch": 1.036559700146319, + "grad_norm": 1.8525819778442383, + "learning_rate": 9.891581794885504e-06, + "loss": 3.3037, + "step": 12930 + }, + { + "epoch": 1.0366398749273416, + "grad_norm": 1.8793340921401978, + "learning_rate": 9.890283425061708e-06, + "loss": 3.1422, + "step": 12931 + }, + { + "epoch": 1.0367200497083642, + "grad_norm": 1.7349778413772583, + "learning_rate": 9.888985057087695e-06, + "loss": 3.1523, + "step": 12932 + }, + { + "epoch": 1.0368002244893868, + "grad_norm": 1.7297004461288452, + "learning_rate": 9.887686690985353e-06, + "loss": 3.2007, + "step": 12933 + }, + { + "epoch": 1.0368803992704094, + "grad_norm": 1.7061560153961182, + "learning_rate": 9.886388326776576e-06, + "loss": 3.1514, + "step": 12934 + }, + { + "epoch": 1.036960574051432, + "grad_norm": 1.680816888809204, + "learning_rate": 9.885089964483248e-06, + "loss": 3.1654, + "step": 12935 + }, + { + "epoch": 1.0370407488324547, + "grad_norm": 1.8210910558700562, + "learning_rate": 9.883791604127268e-06, + "loss": 3.1985, + "step": 12936 + }, + { + "epoch": 1.0371209236134773, + "grad_norm": 1.8155173063278198, + "learning_rate": 9.882493245730517e-06, + "loss": 3.1812, + "step": 12937 + }, + { + "epoch": 1.0372010983945, + "grad_norm": 1.8376903533935547, + "learning_rate": 9.88119488931489e-06, + "loss": 3.118, + "step": 12938 + }, + { + "epoch": 1.0372812731755225, + "grad_norm": 1.8083014488220215, + "learning_rate": 9.879896534902274e-06, + "loss": 3.3044, + "step": 12939 + }, + { + "epoch": 1.0373614479565452, + "grad_norm": 1.7773135900497437, + "learning_rate": 9.87859818251456e-06, + "loss": 3.103, + "step": 12940 + }, + { + "epoch": 1.0374416227375678, + "grad_norm": 1.9809812307357788, + "learning_rate": 9.87729983217364e-06, + "loss": 3.1613, + "step": 12941 + }, + { + "epoch": 1.0375217975185906, + "grad_norm": 1.6426256895065308, + "learning_rate": 9.8760014839014e-06, + "loss": 3.1654, + "step": 12942 + }, + { + "epoch": 1.0376019722996133, + "grad_norm": 1.8786234855651855, + "learning_rate": 9.874703137719733e-06, + "loss": 3.2336, + "step": 12943 + }, + { + "epoch": 1.0376821470806359, + "grad_norm": 1.6638463735580444, + "learning_rate": 9.873404793650524e-06, + "loss": 3.1942, + "step": 12944 + }, + { + "epoch": 1.0377623218616585, + "grad_norm": 1.803704023361206, + "learning_rate": 9.872106451715665e-06, + "loss": 3.1915, + "step": 12945 + }, + { + "epoch": 1.0378424966426811, + "grad_norm": 1.8638867139816284, + "learning_rate": 9.87080811193705e-06, + "loss": 3.2637, + "step": 12946 + }, + { + "epoch": 1.0379226714237038, + "grad_norm": 0.7058343291282654, + "learning_rate": 9.869509774336562e-06, + "loss": 2.4845, + "step": 12947 + }, + { + "epoch": 1.0380028462047264, + "grad_norm": 1.8056083917617798, + "learning_rate": 9.868211438936093e-06, + "loss": 3.3275, + "step": 12948 + }, + { + "epoch": 1.038083020985749, + "grad_norm": 1.7703427076339722, + "learning_rate": 9.866913105757535e-06, + "loss": 3.1986, + "step": 12949 + }, + { + "epoch": 1.0381631957667716, + "grad_norm": 1.8374987840652466, + "learning_rate": 9.865614774822773e-06, + "loss": 3.3281, + "step": 12950 + }, + { + "epoch": 1.0382433705477943, + "grad_norm": 0.6841731071472168, + "learning_rate": 9.8643164461537e-06, + "loss": 2.4854, + "step": 12951 + }, + { + "epoch": 1.0383235453288169, + "grad_norm": 1.9572710990905762, + "learning_rate": 9.863018119772203e-06, + "loss": 3.2212, + "step": 12952 + }, + { + "epoch": 1.0384037201098395, + "grad_norm": 1.7659331560134888, + "learning_rate": 9.861719795700172e-06, + "loss": 3.2136, + "step": 12953 + }, + { + "epoch": 1.0384838948908621, + "grad_norm": 0.6781712770462036, + "learning_rate": 9.860421473959499e-06, + "loss": 2.4749, + "step": 12954 + }, + { + "epoch": 1.0385640696718847, + "grad_norm": 1.763221263885498, + "learning_rate": 9.859123154572068e-06, + "loss": 3.2872, + "step": 12955 + }, + { + "epoch": 1.0386442444529074, + "grad_norm": 1.6480625867843628, + "learning_rate": 9.857824837559771e-06, + "loss": 3.0817, + "step": 12956 + }, + { + "epoch": 1.03872441923393, + "grad_norm": 1.7072488069534302, + "learning_rate": 9.856526522944498e-06, + "loss": 3.131, + "step": 12957 + }, + { + "epoch": 1.0388045940149526, + "grad_norm": 1.866123080253601, + "learning_rate": 9.855228210748137e-06, + "loss": 3.1676, + "step": 12958 + }, + { + "epoch": 1.0388847687959752, + "grad_norm": 1.8187241554260254, + "learning_rate": 9.853929900992578e-06, + "loss": 3.2421, + "step": 12959 + }, + { + "epoch": 1.0389649435769979, + "grad_norm": 1.7514142990112305, + "learning_rate": 9.852631593699709e-06, + "loss": 3.1182, + "step": 12960 + }, + { + "epoch": 1.0390451183580205, + "grad_norm": 1.7967437505722046, + "learning_rate": 9.851333288891419e-06, + "loss": 3.1385, + "step": 12961 + }, + { + "epoch": 1.0391252931390431, + "grad_norm": 1.7815192937850952, + "learning_rate": 9.850034986589598e-06, + "loss": 3.3284, + "step": 12962 + }, + { + "epoch": 1.0392054679200657, + "grad_norm": 1.7302850484848022, + "learning_rate": 9.848736686816133e-06, + "loss": 3.1736, + "step": 12963 + }, + { + "epoch": 1.0392856427010884, + "grad_norm": 1.785388708114624, + "learning_rate": 9.847438389592916e-06, + "loss": 3.2255, + "step": 12964 + }, + { + "epoch": 1.039365817482111, + "grad_norm": 1.8160721063613892, + "learning_rate": 9.846140094941832e-06, + "loss": 3.1927, + "step": 12965 + }, + { + "epoch": 1.0394459922631336, + "grad_norm": 1.774479866027832, + "learning_rate": 9.844841802884773e-06, + "loss": 3.1942, + "step": 12966 + }, + { + "epoch": 1.0395261670441562, + "grad_norm": 0.6921418309211731, + "learning_rate": 9.843543513443626e-06, + "loss": 2.4733, + "step": 12967 + }, + { + "epoch": 1.0396063418251789, + "grad_norm": 1.752601146697998, + "learning_rate": 9.84224522664028e-06, + "loss": 3.0995, + "step": 12968 + }, + { + "epoch": 1.0396865166062015, + "grad_norm": 1.6521803140640259, + "learning_rate": 9.840946942496625e-06, + "loss": 3.2283, + "step": 12969 + }, + { + "epoch": 1.039766691387224, + "grad_norm": 1.7741682529449463, + "learning_rate": 9.839648661034547e-06, + "loss": 3.206, + "step": 12970 + }, + { + "epoch": 1.0398468661682467, + "grad_norm": 0.7511415481567383, + "learning_rate": 9.838350382275939e-06, + "loss": 2.5077, + "step": 12971 + }, + { + "epoch": 1.0399270409492694, + "grad_norm": 1.8718528747558594, + "learning_rate": 9.837052106242683e-06, + "loss": 3.2129, + "step": 12972 + }, + { + "epoch": 1.040007215730292, + "grad_norm": 1.6551949977874756, + "learning_rate": 9.835753832956672e-06, + "loss": 3.1774, + "step": 12973 + }, + { + "epoch": 1.0400873905113146, + "grad_norm": 1.8913154602050781, + "learning_rate": 9.834455562439795e-06, + "loss": 3.238, + "step": 12974 + }, + { + "epoch": 1.0401675652923372, + "grad_norm": 1.8425284624099731, + "learning_rate": 9.833157294713938e-06, + "loss": 3.1888, + "step": 12975 + }, + { + "epoch": 1.0402477400733599, + "grad_norm": 1.7417923212051392, + "learning_rate": 9.831859029800992e-06, + "loss": 3.1163, + "step": 12976 + }, + { + "epoch": 1.0403279148543825, + "grad_norm": 1.805067777633667, + "learning_rate": 9.83056076772284e-06, + "loss": 3.131, + "step": 12977 + }, + { + "epoch": 1.040408089635405, + "grad_norm": 1.7330889701843262, + "learning_rate": 9.829262508501375e-06, + "loss": 3.1864, + "step": 12978 + }, + { + "epoch": 1.0404882644164277, + "grad_norm": 0.700705885887146, + "learning_rate": 9.827964252158484e-06, + "loss": 2.4803, + "step": 12979 + }, + { + "epoch": 1.0405684391974503, + "grad_norm": 1.7507494688034058, + "learning_rate": 9.826665998716056e-06, + "loss": 3.2074, + "step": 12980 + }, + { + "epoch": 1.040648613978473, + "grad_norm": 1.804667592048645, + "learning_rate": 9.825367748195978e-06, + "loss": 3.1606, + "step": 12981 + }, + { + "epoch": 1.0407287887594956, + "grad_norm": 1.7048033475875854, + "learning_rate": 9.824069500620138e-06, + "loss": 3.1753, + "step": 12982 + }, + { + "epoch": 1.0408089635405182, + "grad_norm": 1.729651689529419, + "learning_rate": 9.822771256010422e-06, + "loss": 3.1286, + "step": 12983 + }, + { + "epoch": 1.040889138321541, + "grad_norm": 2.0471482276916504, + "learning_rate": 9.821473014388725e-06, + "loss": 3.1991, + "step": 12984 + }, + { + "epoch": 1.0409693131025637, + "grad_norm": 1.7238913774490356, + "learning_rate": 9.820174775776926e-06, + "loss": 3.2307, + "step": 12985 + }, + { + "epoch": 1.0410494878835863, + "grad_norm": 1.7371585369110107, + "learning_rate": 9.818876540196918e-06, + "loss": 3.2147, + "step": 12986 + }, + { + "epoch": 1.041129662664609, + "grad_norm": 1.7501723766326904, + "learning_rate": 9.817578307670588e-06, + "loss": 3.1837, + "step": 12987 + }, + { + "epoch": 1.0412098374456316, + "grad_norm": 1.9073684215545654, + "learning_rate": 9.816280078219823e-06, + "loss": 3.1396, + "step": 12988 + }, + { + "epoch": 1.0412900122266542, + "grad_norm": 1.7332255840301514, + "learning_rate": 9.814981851866513e-06, + "loss": 3.2122, + "step": 12989 + }, + { + "epoch": 1.0413701870076768, + "grad_norm": 1.7579351663589478, + "learning_rate": 9.813683628632541e-06, + "loss": 3.182, + "step": 12990 + }, + { + "epoch": 1.0414503617886994, + "grad_norm": 1.7609301805496216, + "learning_rate": 9.8123854085398e-06, + "loss": 3.1505, + "step": 12991 + }, + { + "epoch": 1.041530536569722, + "grad_norm": 0.7052568197250366, + "learning_rate": 9.811087191610172e-06, + "loss": 2.4982, + "step": 12992 + }, + { + "epoch": 1.0416107113507447, + "grad_norm": 1.7467106580734253, + "learning_rate": 9.809788977865549e-06, + "loss": 3.067, + "step": 12993 + }, + { + "epoch": 1.0416908861317673, + "grad_norm": 1.6812293529510498, + "learning_rate": 9.808490767327818e-06, + "loss": 3.2368, + "step": 12994 + }, + { + "epoch": 1.04177106091279, + "grad_norm": 1.7313804626464844, + "learning_rate": 9.807192560018862e-06, + "loss": 3.1154, + "step": 12995 + }, + { + "epoch": 1.0418512356938125, + "grad_norm": 1.7973183393478394, + "learning_rate": 9.805894355960574e-06, + "loss": 3.2175, + "step": 12996 + }, + { + "epoch": 1.0419314104748352, + "grad_norm": 1.7949399948120117, + "learning_rate": 9.804596155174838e-06, + "loss": 3.1575, + "step": 12997 + }, + { + "epoch": 1.0420115852558578, + "grad_norm": 1.7687841653823853, + "learning_rate": 9.803297957683541e-06, + "loss": 3.1954, + "step": 12998 + }, + { + "epoch": 1.0420917600368804, + "grad_norm": 1.846674919128418, + "learning_rate": 9.801999763508573e-06, + "loss": 3.3507, + "step": 12999 + }, + { + "epoch": 1.042171934817903, + "grad_norm": 1.6943942308425903, + "learning_rate": 9.800701572671818e-06, + "loss": 3.1442, + "step": 13000 + }, + { + "epoch": 1.0422521095989257, + "grad_norm": 1.780956745147705, + "learning_rate": 9.799403385195164e-06, + "loss": 3.2817, + "step": 13001 + }, + { + "epoch": 1.0423322843799483, + "grad_norm": 0.6995135545730591, + "learning_rate": 9.7981052011005e-06, + "loss": 2.4757, + "step": 13002 + }, + { + "epoch": 1.042412459160971, + "grad_norm": 1.6532222032546997, + "learning_rate": 9.79680702040971e-06, + "loss": 3.1301, + "step": 13003 + }, + { + "epoch": 1.0424926339419935, + "grad_norm": 0.7057437300682068, + "learning_rate": 9.795508843144681e-06, + "loss": 2.4745, + "step": 13004 + }, + { + "epoch": 1.0425728087230162, + "grad_norm": 0.7188941240310669, + "learning_rate": 9.794210669327303e-06, + "loss": 2.46, + "step": 13005 + }, + { + "epoch": 1.0426529835040388, + "grad_norm": 1.6633994579315186, + "learning_rate": 9.792912498979459e-06, + "loss": 3.2306, + "step": 13006 + }, + { + "epoch": 1.0427331582850614, + "grad_norm": 1.7355544567108154, + "learning_rate": 9.791614332123038e-06, + "loss": 3.1826, + "step": 13007 + }, + { + "epoch": 1.042813333066084, + "grad_norm": 1.825364351272583, + "learning_rate": 9.790316168779925e-06, + "loss": 3.1823, + "step": 13008 + }, + { + "epoch": 1.0428935078471067, + "grad_norm": 1.8766473531723022, + "learning_rate": 9.789018008972008e-06, + "loss": 3.2377, + "step": 13009 + }, + { + "epoch": 1.0429736826281293, + "grad_norm": 1.6480211019515991, + "learning_rate": 9.787719852721173e-06, + "loss": 3.1443, + "step": 13010 + }, + { + "epoch": 1.043053857409152, + "grad_norm": 1.7543810606002808, + "learning_rate": 9.786421700049306e-06, + "loss": 3.1504, + "step": 13011 + }, + { + "epoch": 1.0431340321901745, + "grad_norm": 1.778861165046692, + "learning_rate": 9.785123550978295e-06, + "loss": 3.3417, + "step": 13012 + }, + { + "epoch": 1.0432142069711972, + "grad_norm": 1.7729589939117432, + "learning_rate": 9.783825405530023e-06, + "loss": 3.1597, + "step": 13013 + }, + { + "epoch": 1.0432943817522198, + "grad_norm": 1.7029950618743896, + "learning_rate": 9.782527263726382e-06, + "loss": 3.1332, + "step": 13014 + }, + { + "epoch": 1.0433745565332424, + "grad_norm": 1.904170274734497, + "learning_rate": 9.781229125589253e-06, + "loss": 3.2332, + "step": 13015 + }, + { + "epoch": 1.043454731314265, + "grad_norm": 1.862021803855896, + "learning_rate": 9.779930991140524e-06, + "loss": 3.1911, + "step": 13016 + }, + { + "epoch": 1.0435349060952877, + "grad_norm": 1.7040537595748901, + "learning_rate": 9.778632860402079e-06, + "loss": 3.1742, + "step": 13017 + }, + { + "epoch": 1.0436150808763103, + "grad_norm": 1.69086492061615, + "learning_rate": 9.777334733395807e-06, + "loss": 3.2125, + "step": 13018 + }, + { + "epoch": 1.043695255657333, + "grad_norm": 1.7997257709503174, + "learning_rate": 9.776036610143595e-06, + "loss": 3.2422, + "step": 13019 + }, + { + "epoch": 1.0437754304383555, + "grad_norm": 1.6976583003997803, + "learning_rate": 9.774738490667323e-06, + "loss": 3.1003, + "step": 13020 + }, + { + "epoch": 1.0438556052193781, + "grad_norm": 1.9295991659164429, + "learning_rate": 9.773440374988885e-06, + "loss": 3.1998, + "step": 13021 + }, + { + "epoch": 1.0439357800004008, + "grad_norm": 1.8139809370040894, + "learning_rate": 9.77214226313016e-06, + "loss": 3.2092, + "step": 13022 + }, + { + "epoch": 1.0440159547814236, + "grad_norm": 0.7266306281089783, + "learning_rate": 9.770844155113036e-06, + "loss": 2.5009, + "step": 13023 + }, + { + "epoch": 1.0440961295624462, + "grad_norm": 0.7023396492004395, + "learning_rate": 9.7695460509594e-06, + "loss": 2.4906, + "step": 13024 + }, + { + "epoch": 1.0441763043434689, + "grad_norm": 1.7071529626846313, + "learning_rate": 9.768247950691134e-06, + "loss": 3.2041, + "step": 13025 + }, + { + "epoch": 1.0442564791244915, + "grad_norm": 1.9108607769012451, + "learning_rate": 9.76694985433013e-06, + "loss": 3.2867, + "step": 13026 + }, + { + "epoch": 1.0443366539055141, + "grad_norm": 1.773645043373108, + "learning_rate": 9.765651761898265e-06, + "loss": 3.2316, + "step": 13027 + }, + { + "epoch": 1.0444168286865367, + "grad_norm": 1.8260771036148071, + "learning_rate": 9.76435367341743e-06, + "loss": 3.2838, + "step": 13028 + }, + { + "epoch": 1.0444970034675594, + "grad_norm": 1.815300703048706, + "learning_rate": 9.763055588909511e-06, + "loss": 3.1762, + "step": 13029 + }, + { + "epoch": 1.044577178248582, + "grad_norm": 1.7508869171142578, + "learning_rate": 9.76175750839639e-06, + "loss": 3.179, + "step": 13030 + }, + { + "epoch": 1.0446573530296046, + "grad_norm": 1.771498680114746, + "learning_rate": 9.760459431899954e-06, + "loss": 3.2145, + "step": 13031 + }, + { + "epoch": 1.0447375278106272, + "grad_norm": 1.8145288228988647, + "learning_rate": 9.759161359442085e-06, + "loss": 3.1487, + "step": 13032 + }, + { + "epoch": 1.0448177025916499, + "grad_norm": 1.7650505304336548, + "learning_rate": 9.757863291044672e-06, + "loss": 3.194, + "step": 13033 + }, + { + "epoch": 1.0448978773726725, + "grad_norm": 1.7158187627792358, + "learning_rate": 9.7565652267296e-06, + "loss": 3.1652, + "step": 13034 + }, + { + "epoch": 1.044978052153695, + "grad_norm": 0.6877793073654175, + "learning_rate": 9.75526716651875e-06, + "loss": 2.4613, + "step": 13035 + }, + { + "epoch": 1.0450582269347177, + "grad_norm": 0.68150395154953, + "learning_rate": 9.753969110434013e-06, + "loss": 2.4683, + "step": 13036 + }, + { + "epoch": 1.0451384017157404, + "grad_norm": 1.7709426879882812, + "learning_rate": 9.752671058497266e-06, + "loss": 3.1679, + "step": 13037 + }, + { + "epoch": 1.045218576496763, + "grad_norm": 0.7115656137466431, + "learning_rate": 9.751373010730399e-06, + "loss": 2.4838, + "step": 13038 + }, + { + "epoch": 1.0452987512777856, + "grad_norm": 1.7534410953521729, + "learning_rate": 9.750074967155298e-06, + "loss": 3.2063, + "step": 13039 + }, + { + "epoch": 1.0453789260588082, + "grad_norm": 1.783423662185669, + "learning_rate": 9.748776927793842e-06, + "loss": 3.1653, + "step": 13040 + }, + { + "epoch": 1.0454591008398308, + "grad_norm": 1.7673944234848022, + "learning_rate": 9.74747889266792e-06, + "loss": 3.1529, + "step": 13041 + }, + { + "epoch": 1.0455392756208535, + "grad_norm": 1.8072662353515625, + "learning_rate": 9.746180861799413e-06, + "loss": 3.2931, + "step": 13042 + }, + { + "epoch": 1.045619450401876, + "grad_norm": 1.7798268795013428, + "learning_rate": 9.744882835210208e-06, + "loss": 3.2515, + "step": 13043 + }, + { + "epoch": 1.0456996251828987, + "grad_norm": 1.8186053037643433, + "learning_rate": 9.74358481292219e-06, + "loss": 3.1092, + "step": 13044 + }, + { + "epoch": 1.0457797999639213, + "grad_norm": 1.866614818572998, + "learning_rate": 9.742286794957241e-06, + "loss": 3.1947, + "step": 13045 + }, + { + "epoch": 1.045859974744944, + "grad_norm": 1.8181202411651611, + "learning_rate": 9.740988781337247e-06, + "loss": 3.1436, + "step": 13046 + }, + { + "epoch": 1.0459401495259666, + "grad_norm": 1.8635165691375732, + "learning_rate": 9.739690772084088e-06, + "loss": 3.2695, + "step": 13047 + }, + { + "epoch": 1.0460203243069892, + "grad_norm": 1.807141900062561, + "learning_rate": 9.738392767219653e-06, + "loss": 3.1665, + "step": 13048 + }, + { + "epoch": 1.0461004990880118, + "grad_norm": 1.7726491689682007, + "learning_rate": 9.737094766765824e-06, + "loss": 3.2198, + "step": 13049 + }, + { + "epoch": 1.0461806738690345, + "grad_norm": 1.8011133670806885, + "learning_rate": 9.735796770744481e-06, + "loss": 3.1049, + "step": 13050 + }, + { + "epoch": 1.046260848650057, + "grad_norm": 1.7704622745513916, + "learning_rate": 9.734498779177517e-06, + "loss": 3.1705, + "step": 13051 + }, + { + "epoch": 1.0463410234310797, + "grad_norm": 1.7074896097183228, + "learning_rate": 9.733200792086806e-06, + "loss": 3.3634, + "step": 13052 + }, + { + "epoch": 1.0464211982121023, + "grad_norm": 1.7212944030761719, + "learning_rate": 9.731902809494236e-06, + "loss": 3.2325, + "step": 13053 + }, + { + "epoch": 1.046501372993125, + "grad_norm": 0.7831724286079407, + "learning_rate": 9.73060483142169e-06, + "loss": 2.4558, + "step": 13054 + }, + { + "epoch": 1.0465815477741476, + "grad_norm": 1.7605297565460205, + "learning_rate": 9.729306857891052e-06, + "loss": 3.0919, + "step": 13055 + }, + { + "epoch": 1.0466617225551702, + "grad_norm": 1.771437168121338, + "learning_rate": 9.728008888924206e-06, + "loss": 3.14, + "step": 13056 + }, + { + "epoch": 1.0467418973361928, + "grad_norm": 1.6789470911026, + "learning_rate": 9.726710924543033e-06, + "loss": 3.2191, + "step": 13057 + }, + { + "epoch": 1.0468220721172155, + "grad_norm": 1.7572660446166992, + "learning_rate": 9.725412964769417e-06, + "loss": 3.2288, + "step": 13058 + }, + { + "epoch": 1.046902246898238, + "grad_norm": 0.6877861618995667, + "learning_rate": 9.724115009625245e-06, + "loss": 2.4655, + "step": 13059 + }, + { + "epoch": 1.0469824216792607, + "grad_norm": 1.8334990739822388, + "learning_rate": 9.722817059132393e-06, + "loss": 3.1421, + "step": 13060 + }, + { + "epoch": 1.0470625964602833, + "grad_norm": 1.8793920278549194, + "learning_rate": 9.721519113312751e-06, + "loss": 3.1781, + "step": 13061 + }, + { + "epoch": 1.047142771241306, + "grad_norm": 1.7267308235168457, + "learning_rate": 9.720221172188194e-06, + "loss": 3.1578, + "step": 13062 + }, + { + "epoch": 1.0472229460223286, + "grad_norm": 1.778897762298584, + "learning_rate": 9.718923235780612e-06, + "loss": 3.2032, + "step": 13063 + }, + { + "epoch": 1.0473031208033512, + "grad_norm": 1.7485707998275757, + "learning_rate": 9.717625304111886e-06, + "loss": 3.175, + "step": 13064 + }, + { + "epoch": 1.047383295584374, + "grad_norm": 1.6792019605636597, + "learning_rate": 9.716327377203896e-06, + "loss": 3.1173, + "step": 13065 + }, + { + "epoch": 1.0474634703653967, + "grad_norm": 1.7529568672180176, + "learning_rate": 9.71502945507853e-06, + "loss": 3.3143, + "step": 13066 + }, + { + "epoch": 1.0475436451464193, + "grad_norm": 1.8036069869995117, + "learning_rate": 9.713731537757664e-06, + "loss": 3.1997, + "step": 13067 + }, + { + "epoch": 1.047623819927442, + "grad_norm": 1.8934950828552246, + "learning_rate": 9.712433625263183e-06, + "loss": 3.1477, + "step": 13068 + }, + { + "epoch": 1.0477039947084645, + "grad_norm": 1.824259638786316, + "learning_rate": 9.711135717616972e-06, + "loss": 3.2086, + "step": 13069 + }, + { + "epoch": 1.0477841694894872, + "grad_norm": 1.826390027999878, + "learning_rate": 9.709837814840907e-06, + "loss": 3.2268, + "step": 13070 + }, + { + "epoch": 1.0478643442705098, + "grad_norm": 1.7781057357788086, + "learning_rate": 9.70853991695688e-06, + "loss": 3.3127, + "step": 13071 + }, + { + "epoch": 1.0479445190515324, + "grad_norm": 1.7340619564056396, + "learning_rate": 9.707242023986763e-06, + "loss": 3.1903, + "step": 13072 + }, + { + "epoch": 1.048024693832555, + "grad_norm": 1.8287562131881714, + "learning_rate": 9.705944135952442e-06, + "loss": 3.2087, + "step": 13073 + }, + { + "epoch": 1.0481048686135777, + "grad_norm": 1.7681541442871094, + "learning_rate": 9.704646252875802e-06, + "loss": 3.103, + "step": 13074 + }, + { + "epoch": 1.0481850433946003, + "grad_norm": 1.7155975103378296, + "learning_rate": 9.70334837477872e-06, + "loss": 3.2524, + "step": 13075 + }, + { + "epoch": 1.048265218175623, + "grad_norm": 0.7173834443092346, + "learning_rate": 9.702050501683083e-06, + "loss": 2.4874, + "step": 13076 + }, + { + "epoch": 1.0483453929566455, + "grad_norm": 1.7267515659332275, + "learning_rate": 9.700752633610764e-06, + "loss": 3.2241, + "step": 13077 + }, + { + "epoch": 1.0484255677376682, + "grad_norm": 1.9460206031799316, + "learning_rate": 9.699454770583653e-06, + "loss": 3.2899, + "step": 13078 + }, + { + "epoch": 1.0485057425186908, + "grad_norm": 0.6781555414199829, + "learning_rate": 9.69815691262363e-06, + "loss": 2.4696, + "step": 13079 + }, + { + "epoch": 1.0485859172997134, + "grad_norm": 1.7815897464752197, + "learning_rate": 9.696859059752574e-06, + "loss": 3.1894, + "step": 13080 + }, + { + "epoch": 1.048666092080736, + "grad_norm": 1.700476050376892, + "learning_rate": 9.695561211992368e-06, + "loss": 3.1646, + "step": 13081 + }, + { + "epoch": 1.0487462668617586, + "grad_norm": 1.7146542072296143, + "learning_rate": 9.694263369364892e-06, + "loss": 3.1492, + "step": 13082 + }, + { + "epoch": 1.0488264416427813, + "grad_norm": 1.8340387344360352, + "learning_rate": 9.692965531892026e-06, + "loss": 3.2206, + "step": 13083 + }, + { + "epoch": 1.048906616423804, + "grad_norm": 1.7470537424087524, + "learning_rate": 9.691667699595657e-06, + "loss": 3.1884, + "step": 13084 + }, + { + "epoch": 1.0489867912048265, + "grad_norm": 1.7578502893447876, + "learning_rate": 9.690369872497658e-06, + "loss": 3.1177, + "step": 13085 + }, + { + "epoch": 1.0490669659858491, + "grad_norm": 1.7412067651748657, + "learning_rate": 9.689072050619917e-06, + "loss": 3.2515, + "step": 13086 + }, + { + "epoch": 1.0491471407668718, + "grad_norm": 1.8823583126068115, + "learning_rate": 9.68777423398431e-06, + "loss": 3.1495, + "step": 13087 + }, + { + "epoch": 1.0492273155478944, + "grad_norm": 1.604285717010498, + "learning_rate": 9.686476422612719e-06, + "loss": 3.1333, + "step": 13088 + }, + { + "epoch": 1.049307490328917, + "grad_norm": 1.7856913805007935, + "learning_rate": 9.685178616527027e-06, + "loss": 3.3104, + "step": 13089 + }, + { + "epoch": 1.0493876651099396, + "grad_norm": 1.7630000114440918, + "learning_rate": 9.68388081574911e-06, + "loss": 3.1794, + "step": 13090 + }, + { + "epoch": 1.0494678398909623, + "grad_norm": 1.8459997177124023, + "learning_rate": 9.682583020300853e-06, + "loss": 3.2393, + "step": 13091 + }, + { + "epoch": 1.0495480146719849, + "grad_norm": 1.782740831375122, + "learning_rate": 9.681285230204133e-06, + "loss": 3.2917, + "step": 13092 + }, + { + "epoch": 1.0496281894530075, + "grad_norm": 1.7353419065475464, + "learning_rate": 9.67998744548083e-06, + "loss": 3.2938, + "step": 13093 + }, + { + "epoch": 1.0497083642340301, + "grad_norm": 1.8318650722503662, + "learning_rate": 9.67868966615283e-06, + "loss": 3.225, + "step": 13094 + }, + { + "epoch": 1.0497885390150528, + "grad_norm": 1.7558889389038086, + "learning_rate": 9.677391892242008e-06, + "loss": 3.1055, + "step": 13095 + }, + { + "epoch": 1.0498687137960754, + "grad_norm": 1.772949457168579, + "learning_rate": 9.676094123770244e-06, + "loss": 3.2676, + "step": 13096 + }, + { + "epoch": 1.049948888577098, + "grad_norm": 1.8792575597763062, + "learning_rate": 9.674796360759418e-06, + "loss": 3.1578, + "step": 13097 + }, + { + "epoch": 1.0500290633581206, + "grad_norm": 1.7714484930038452, + "learning_rate": 9.673498603231411e-06, + "loss": 3.2025, + "step": 13098 + }, + { + "epoch": 1.0501092381391433, + "grad_norm": 1.649172306060791, + "learning_rate": 9.672200851208105e-06, + "loss": 3.1587, + "step": 13099 + }, + { + "epoch": 1.0501894129201659, + "grad_norm": 1.724624514579773, + "learning_rate": 9.670903104711374e-06, + "loss": 3.1665, + "step": 13100 + }, + { + "epoch": 1.0502695877011885, + "grad_norm": 1.763881802558899, + "learning_rate": 9.669605363763104e-06, + "loss": 3.1458, + "step": 13101 + }, + { + "epoch": 1.0503497624822111, + "grad_norm": 1.7385543584823608, + "learning_rate": 9.668307628385169e-06, + "loss": 3.1036, + "step": 13102 + }, + { + "epoch": 1.0504299372632337, + "grad_norm": 0.68593430519104, + "learning_rate": 9.66700989859945e-06, + "loss": 2.4762, + "step": 13103 + }, + { + "epoch": 1.0505101120442564, + "grad_norm": 1.8516522645950317, + "learning_rate": 9.665712174427829e-06, + "loss": 3.1794, + "step": 13104 + }, + { + "epoch": 1.0505902868252792, + "grad_norm": 0.6879744529724121, + "learning_rate": 9.664414455892179e-06, + "loss": 2.4706, + "step": 13105 + }, + { + "epoch": 1.0506704616063018, + "grad_norm": 1.7404465675354004, + "learning_rate": 9.663116743014387e-06, + "loss": 3.1255, + "step": 13106 + }, + { + "epoch": 1.0507506363873245, + "grad_norm": 1.730364203453064, + "learning_rate": 9.661819035816326e-06, + "loss": 3.2522, + "step": 13107 + }, + { + "epoch": 1.050830811168347, + "grad_norm": 1.7572110891342163, + "learning_rate": 9.660521334319875e-06, + "loss": 3.2787, + "step": 13108 + }, + { + "epoch": 1.0509109859493697, + "grad_norm": 1.845833659172058, + "learning_rate": 9.65922363854692e-06, + "loss": 3.1774, + "step": 13109 + }, + { + "epoch": 1.0509911607303923, + "grad_norm": 1.7969410419464111, + "learning_rate": 9.65792594851933e-06, + "loss": 3.2776, + "step": 13110 + }, + { + "epoch": 1.051071335511415, + "grad_norm": 1.743038535118103, + "learning_rate": 9.65662826425899e-06, + "loss": 3.1578, + "step": 13111 + }, + { + "epoch": 1.0511515102924376, + "grad_norm": 1.6867579221725464, + "learning_rate": 9.655330585787774e-06, + "loss": 3.2114, + "step": 13112 + }, + { + "epoch": 1.0512316850734602, + "grad_norm": 1.7495654821395874, + "learning_rate": 9.654032913127564e-06, + "loss": 3.1672, + "step": 13113 + }, + { + "epoch": 1.0513118598544828, + "grad_norm": 1.7732893228530884, + "learning_rate": 9.65273524630024e-06, + "loss": 3.2472, + "step": 13114 + }, + { + "epoch": 1.0513920346355055, + "grad_norm": 1.6544146537780762, + "learning_rate": 9.651437585327673e-06, + "loss": 3.1692, + "step": 13115 + }, + { + "epoch": 1.051472209416528, + "grad_norm": 1.7629854679107666, + "learning_rate": 9.65013993023175e-06, + "loss": 3.2261, + "step": 13116 + }, + { + "epoch": 1.0515523841975507, + "grad_norm": 1.725721001625061, + "learning_rate": 9.64884228103434e-06, + "loss": 3.3103, + "step": 13117 + }, + { + "epoch": 1.0516325589785733, + "grad_norm": 1.694885492324829, + "learning_rate": 9.647544637757327e-06, + "loss": 3.1868, + "step": 13118 + }, + { + "epoch": 1.051712733759596, + "grad_norm": 1.7527470588684082, + "learning_rate": 9.646247000422588e-06, + "loss": 3.1877, + "step": 13119 + }, + { + "epoch": 1.0517929085406186, + "grad_norm": 1.754543662071228, + "learning_rate": 9.644949369052e-06, + "loss": 3.1549, + "step": 13120 + }, + { + "epoch": 1.0518730833216412, + "grad_norm": 1.7951260805130005, + "learning_rate": 9.64365174366744e-06, + "loss": 3.3347, + "step": 13121 + }, + { + "epoch": 1.0519532581026638, + "grad_norm": 1.8249866962432861, + "learning_rate": 9.642354124290786e-06, + "loss": 3.1466, + "step": 13122 + }, + { + "epoch": 1.0520334328836864, + "grad_norm": 1.7813293933868408, + "learning_rate": 9.641056510943914e-06, + "loss": 3.2634, + "step": 13123 + }, + { + "epoch": 1.052113607664709, + "grad_norm": 1.9084218740463257, + "learning_rate": 9.639758903648706e-06, + "loss": 3.1385, + "step": 13124 + }, + { + "epoch": 1.0521937824457317, + "grad_norm": 0.6780456900596619, + "learning_rate": 9.638461302427033e-06, + "loss": 2.4634, + "step": 13125 + }, + { + "epoch": 1.0522739572267543, + "grad_norm": 1.8299020528793335, + "learning_rate": 9.637163707300777e-06, + "loss": 3.2094, + "step": 13126 + }, + { + "epoch": 1.052354132007777, + "grad_norm": 1.760427713394165, + "learning_rate": 9.635866118291811e-06, + "loss": 3.1882, + "step": 13127 + }, + { + "epoch": 1.0524343067887996, + "grad_norm": 1.815726637840271, + "learning_rate": 9.634568535422015e-06, + "loss": 3.1332, + "step": 13128 + }, + { + "epoch": 1.0525144815698222, + "grad_norm": 1.8573756217956543, + "learning_rate": 9.633270958713265e-06, + "loss": 3.254, + "step": 13129 + }, + { + "epoch": 1.0525946563508448, + "grad_norm": 1.8303388357162476, + "learning_rate": 9.631973388187437e-06, + "loss": 3.1411, + "step": 13130 + }, + { + "epoch": 1.0526748311318674, + "grad_norm": 1.8922841548919678, + "learning_rate": 9.630675823866409e-06, + "loss": 3.2384, + "step": 13131 + }, + { + "epoch": 1.05275500591289, + "grad_norm": 1.6888457536697388, + "learning_rate": 9.629378265772055e-06, + "loss": 3.1482, + "step": 13132 + }, + { + "epoch": 1.0528351806939127, + "grad_norm": 1.846588134765625, + "learning_rate": 9.628080713926252e-06, + "loss": 3.1642, + "step": 13133 + }, + { + "epoch": 1.0529153554749353, + "grad_norm": 0.7333573698997498, + "learning_rate": 9.626783168350879e-06, + "loss": 2.4849, + "step": 13134 + }, + { + "epoch": 1.052995530255958, + "grad_norm": 1.7093373537063599, + "learning_rate": 9.625485629067809e-06, + "loss": 3.1741, + "step": 13135 + }, + { + "epoch": 1.0530757050369806, + "grad_norm": 1.8199108839035034, + "learning_rate": 9.624188096098921e-06, + "loss": 3.1861, + "step": 13136 + }, + { + "epoch": 1.0531558798180032, + "grad_norm": 1.7663358449935913, + "learning_rate": 9.622890569466088e-06, + "loss": 3.1359, + "step": 13137 + }, + { + "epoch": 1.0532360545990258, + "grad_norm": 1.8390617370605469, + "learning_rate": 9.621593049191186e-06, + "loss": 3.2191, + "step": 13138 + }, + { + "epoch": 1.0533162293800484, + "grad_norm": 1.7141557931900024, + "learning_rate": 9.620295535296094e-06, + "loss": 3.16, + "step": 13139 + }, + { + "epoch": 1.053396404161071, + "grad_norm": 1.9479165077209473, + "learning_rate": 9.618998027802684e-06, + "loss": 3.3206, + "step": 13140 + }, + { + "epoch": 1.0534765789420937, + "grad_norm": 1.616315484046936, + "learning_rate": 9.617700526732833e-06, + "loss": 3.0984, + "step": 13141 + }, + { + "epoch": 1.0535567537231163, + "grad_norm": 1.7160546779632568, + "learning_rate": 9.616403032108418e-06, + "loss": 3.1129, + "step": 13142 + }, + { + "epoch": 1.053636928504139, + "grad_norm": 1.715087652206421, + "learning_rate": 9.61510554395131e-06, + "loss": 3.1244, + "step": 13143 + }, + { + "epoch": 1.0537171032851615, + "grad_norm": 1.7459373474121094, + "learning_rate": 9.613808062283391e-06, + "loss": 3.1829, + "step": 13144 + }, + { + "epoch": 1.0537972780661842, + "grad_norm": 1.7294871807098389, + "learning_rate": 9.612510587126528e-06, + "loss": 3.2019, + "step": 13145 + }, + { + "epoch": 1.0538774528472068, + "grad_norm": 1.796575665473938, + "learning_rate": 9.6112131185026e-06, + "loss": 3.177, + "step": 13146 + }, + { + "epoch": 1.0539576276282296, + "grad_norm": 1.6511796712875366, + "learning_rate": 9.609915656433485e-06, + "loss": 3.1628, + "step": 13147 + }, + { + "epoch": 1.0540378024092523, + "grad_norm": 1.770370364189148, + "learning_rate": 9.608618200941052e-06, + "loss": 3.169, + "step": 13148 + }, + { + "epoch": 1.054117977190275, + "grad_norm": 0.7120949029922485, + "learning_rate": 9.60732075204718e-06, + "loss": 2.4907, + "step": 13149 + }, + { + "epoch": 1.0541981519712975, + "grad_norm": 1.7354955673217773, + "learning_rate": 9.60602330977374e-06, + "loss": 3.2477, + "step": 13150 + }, + { + "epoch": 1.0542783267523201, + "grad_norm": 1.7938125133514404, + "learning_rate": 9.604725874142607e-06, + "loss": 3.1464, + "step": 13151 + }, + { + "epoch": 1.0543585015333428, + "grad_norm": 1.6576119661331177, + "learning_rate": 9.60342844517566e-06, + "loss": 3.1374, + "step": 13152 + }, + { + "epoch": 1.0544386763143654, + "grad_norm": 1.804767370223999, + "learning_rate": 9.602131022894768e-06, + "loss": 3.1333, + "step": 13153 + }, + { + "epoch": 1.054518851095388, + "grad_norm": 1.819273829460144, + "learning_rate": 9.600833607321807e-06, + "loss": 3.1864, + "step": 13154 + }, + { + "epoch": 1.0545990258764106, + "grad_norm": 1.790151596069336, + "learning_rate": 9.59953619847865e-06, + "loss": 3.2433, + "step": 13155 + }, + { + "epoch": 1.0546792006574333, + "grad_norm": 1.7306534051895142, + "learning_rate": 9.59823879638717e-06, + "loss": 3.1177, + "step": 13156 + }, + { + "epoch": 1.0547593754384559, + "grad_norm": 1.9851956367492676, + "learning_rate": 9.596941401069247e-06, + "loss": 3.2181, + "step": 13157 + }, + { + "epoch": 1.0548395502194785, + "grad_norm": 1.738970160484314, + "learning_rate": 9.595644012546745e-06, + "loss": 3.1095, + "step": 13158 + }, + { + "epoch": 1.0549197250005011, + "grad_norm": 1.8210400342941284, + "learning_rate": 9.594346630841546e-06, + "loss": 3.2483, + "step": 13159 + }, + { + "epoch": 1.0549998997815238, + "grad_norm": 1.8369308710098267, + "learning_rate": 9.593049255975517e-06, + "loss": 3.1965, + "step": 13160 + }, + { + "epoch": 1.0550800745625464, + "grad_norm": 1.7692179679870605, + "learning_rate": 9.591751887970535e-06, + "loss": 3.1762, + "step": 13161 + }, + { + "epoch": 1.055160249343569, + "grad_norm": 0.7293958067893982, + "learning_rate": 9.590454526848476e-06, + "loss": 2.4838, + "step": 13162 + }, + { + "epoch": 1.0552404241245916, + "grad_norm": 1.843619704246521, + "learning_rate": 9.589157172631204e-06, + "loss": 3.3043, + "step": 13163 + }, + { + "epoch": 1.0553205989056142, + "grad_norm": 1.780203104019165, + "learning_rate": 9.5878598253406e-06, + "loss": 3.2684, + "step": 13164 + }, + { + "epoch": 1.0554007736866369, + "grad_norm": 1.7842483520507812, + "learning_rate": 9.586562484998533e-06, + "loss": 3.1632, + "step": 13165 + }, + { + "epoch": 1.0554809484676595, + "grad_norm": 1.732185959815979, + "learning_rate": 9.585265151626877e-06, + "loss": 3.2375, + "step": 13166 + }, + { + "epoch": 1.0555611232486821, + "grad_norm": 1.8741745948791504, + "learning_rate": 9.583967825247505e-06, + "loss": 3.22, + "step": 13167 + }, + { + "epoch": 1.0556412980297047, + "grad_norm": 1.8410731554031372, + "learning_rate": 9.582670505882288e-06, + "loss": 3.1461, + "step": 13168 + }, + { + "epoch": 1.0557214728107274, + "grad_norm": 0.7186578512191772, + "learning_rate": 9.5813731935531e-06, + "loss": 2.4569, + "step": 13169 + }, + { + "epoch": 1.05580164759175, + "grad_norm": 1.8095500469207764, + "learning_rate": 9.580075888281812e-06, + "loss": 3.2694, + "step": 13170 + }, + { + "epoch": 1.0558818223727726, + "grad_norm": 1.8371654748916626, + "learning_rate": 9.578778590090296e-06, + "loss": 3.1868, + "step": 13171 + }, + { + "epoch": 1.0559619971537952, + "grad_norm": 1.7512686252593994, + "learning_rate": 9.577481299000425e-06, + "loss": 3.1672, + "step": 13172 + }, + { + "epoch": 1.0560421719348179, + "grad_norm": 1.8536676168441772, + "learning_rate": 9.576184015034072e-06, + "loss": 3.2045, + "step": 13173 + }, + { + "epoch": 1.0561223467158405, + "grad_norm": 1.8483961820602417, + "learning_rate": 9.574886738213105e-06, + "loss": 3.2185, + "step": 13174 + }, + { + "epoch": 1.0562025214968631, + "grad_norm": 1.8230611085891724, + "learning_rate": 9.5735894685594e-06, + "loss": 3.2003, + "step": 13175 + }, + { + "epoch": 1.0562826962778857, + "grad_norm": 1.7681653499603271, + "learning_rate": 9.572292206094823e-06, + "loss": 3.2444, + "step": 13176 + }, + { + "epoch": 1.0563628710589084, + "grad_norm": 1.775174617767334, + "learning_rate": 9.570994950841252e-06, + "loss": 3.2849, + "step": 13177 + }, + { + "epoch": 1.056443045839931, + "grad_norm": 1.8788901567459106, + "learning_rate": 9.569697702820553e-06, + "loss": 3.2751, + "step": 13178 + }, + { + "epoch": 1.0565232206209536, + "grad_norm": 1.648598313331604, + "learning_rate": 9.5684004620546e-06, + "loss": 3.183, + "step": 13179 + }, + { + "epoch": 1.0566033954019762, + "grad_norm": 1.7704404592514038, + "learning_rate": 9.567103228565261e-06, + "loss": 3.2492, + "step": 13180 + }, + { + "epoch": 1.0566835701829989, + "grad_norm": 1.723527193069458, + "learning_rate": 9.56580600237441e-06, + "loss": 3.1427, + "step": 13181 + }, + { + "epoch": 1.0567637449640215, + "grad_norm": 1.6184650659561157, + "learning_rate": 9.56450878350392e-06, + "loss": 3.2837, + "step": 13182 + }, + { + "epoch": 1.056843919745044, + "grad_norm": 1.6877388954162598, + "learning_rate": 9.563211571975654e-06, + "loss": 3.0087, + "step": 13183 + }, + { + "epoch": 1.0569240945260667, + "grad_norm": 1.7980201244354248, + "learning_rate": 9.56191436781149e-06, + "loss": 3.1932, + "step": 13184 + }, + { + "epoch": 1.0570042693070894, + "grad_norm": 1.7687405347824097, + "learning_rate": 9.560617171033294e-06, + "loss": 3.1404, + "step": 13185 + }, + { + "epoch": 1.0570844440881122, + "grad_norm": 1.7478580474853516, + "learning_rate": 9.559319981662936e-06, + "loss": 3.1637, + "step": 13186 + }, + { + "epoch": 1.0571646188691348, + "grad_norm": 1.7821358442306519, + "learning_rate": 9.55802279972229e-06, + "loss": 3.2554, + "step": 13187 + }, + { + "epoch": 1.0572447936501574, + "grad_norm": 1.7311038970947266, + "learning_rate": 9.556725625233222e-06, + "loss": 3.1791, + "step": 13188 + }, + { + "epoch": 1.05732496843118, + "grad_norm": 1.9242336750030518, + "learning_rate": 9.555428458217608e-06, + "loss": 3.2768, + "step": 13189 + }, + { + "epoch": 1.0574051432122027, + "grad_norm": 1.8337926864624023, + "learning_rate": 9.55413129869731e-06, + "loss": 3.1147, + "step": 13190 + }, + { + "epoch": 1.0574853179932253, + "grad_norm": 1.8386354446411133, + "learning_rate": 9.5528341466942e-06, + "loss": 3.1605, + "step": 13191 + }, + { + "epoch": 1.057565492774248, + "grad_norm": 1.8244913816452026, + "learning_rate": 9.55153700223015e-06, + "loss": 3.3243, + "step": 13192 + }, + { + "epoch": 1.0576456675552706, + "grad_norm": 1.8801254034042358, + "learning_rate": 9.550239865327028e-06, + "loss": 3.1958, + "step": 13193 + }, + { + "epoch": 1.0577258423362932, + "grad_norm": 1.8066279888153076, + "learning_rate": 9.548942736006705e-06, + "loss": 3.1779, + "step": 13194 + }, + { + "epoch": 1.0578060171173158, + "grad_norm": 1.7037075757980347, + "learning_rate": 9.547645614291048e-06, + "loss": 3.2004, + "step": 13195 + }, + { + "epoch": 1.0578861918983384, + "grad_norm": 1.768683910369873, + "learning_rate": 9.546348500201924e-06, + "loss": 3.1681, + "step": 13196 + }, + { + "epoch": 1.057966366679361, + "grad_norm": 1.8293168544769287, + "learning_rate": 9.545051393761208e-06, + "loss": 3.1865, + "step": 13197 + }, + { + "epoch": 1.0580465414603837, + "grad_norm": 1.7841894626617432, + "learning_rate": 9.543754294990763e-06, + "loss": 3.1244, + "step": 13198 + }, + { + "epoch": 1.0581267162414063, + "grad_norm": 1.8173197507858276, + "learning_rate": 9.542457203912462e-06, + "loss": 3.207, + "step": 13199 + }, + { + "epoch": 1.058206891022429, + "grad_norm": 1.8738785982131958, + "learning_rate": 9.541160120548167e-06, + "loss": 3.1766, + "step": 13200 + }, + { + "epoch": 1.0582870658034516, + "grad_norm": 1.7749053239822388, + "learning_rate": 9.539863044919753e-06, + "loss": 3.2233, + "step": 13201 + }, + { + "epoch": 1.0583672405844742, + "grad_norm": 1.6560848951339722, + "learning_rate": 9.538565977049087e-06, + "loss": 3.1705, + "step": 13202 + }, + { + "epoch": 1.0584474153654968, + "grad_norm": 1.811525821685791, + "learning_rate": 9.537268916958035e-06, + "loss": 3.243, + "step": 13203 + }, + { + "epoch": 1.0585275901465194, + "grad_norm": 1.8076910972595215, + "learning_rate": 9.535971864668467e-06, + "loss": 3.1923, + "step": 13204 + }, + { + "epoch": 1.058607764927542, + "grad_norm": 1.8623923063278198, + "learning_rate": 9.534674820202248e-06, + "loss": 3.3354, + "step": 13205 + }, + { + "epoch": 1.0586879397085647, + "grad_norm": 1.761508584022522, + "learning_rate": 9.533377783581248e-06, + "loss": 3.1733, + "step": 13206 + }, + { + "epoch": 1.0587681144895873, + "grad_norm": 1.7977385520935059, + "learning_rate": 9.532080754827336e-06, + "loss": 3.1666, + "step": 13207 + }, + { + "epoch": 1.05884828927061, + "grad_norm": 1.9223474264144897, + "learning_rate": 9.530783733962374e-06, + "loss": 3.2527, + "step": 13208 + }, + { + "epoch": 1.0589284640516325, + "grad_norm": 1.813454508781433, + "learning_rate": 9.529486721008237e-06, + "loss": 3.1765, + "step": 13209 + }, + { + "epoch": 1.0590086388326552, + "grad_norm": 1.8243865966796875, + "learning_rate": 9.528189715986786e-06, + "loss": 3.1825, + "step": 13210 + }, + { + "epoch": 1.0590888136136778, + "grad_norm": 2.0709807872772217, + "learning_rate": 9.526892718919888e-06, + "loss": 3.0974, + "step": 13211 + }, + { + "epoch": 1.0591689883947004, + "grad_norm": 1.8219627141952515, + "learning_rate": 9.525595729829415e-06, + "loss": 3.2728, + "step": 13212 + }, + { + "epoch": 1.059249163175723, + "grad_norm": 1.7800519466400146, + "learning_rate": 9.52429874873723e-06, + "loss": 3.1583, + "step": 13213 + }, + { + "epoch": 1.0593293379567457, + "grad_norm": 1.7453582286834717, + "learning_rate": 9.523001775665202e-06, + "loss": 3.1278, + "step": 13214 + }, + { + "epoch": 1.0594095127377683, + "grad_norm": 1.8508367538452148, + "learning_rate": 9.521704810635193e-06, + "loss": 3.1582, + "step": 13215 + }, + { + "epoch": 1.059489687518791, + "grad_norm": 1.6713085174560547, + "learning_rate": 9.520407853669073e-06, + "loss": 3.2148, + "step": 13216 + }, + { + "epoch": 1.0595698622998135, + "grad_norm": 1.7986222505569458, + "learning_rate": 9.51911090478871e-06, + "loss": 3.1392, + "step": 13217 + }, + { + "epoch": 1.0596500370808362, + "grad_norm": 1.6888056993484497, + "learning_rate": 9.517813964015964e-06, + "loss": 3.1346, + "step": 13218 + }, + { + "epoch": 1.0597302118618588, + "grad_norm": 1.7268495559692383, + "learning_rate": 9.516517031372709e-06, + "loss": 3.2094, + "step": 13219 + }, + { + "epoch": 1.0598103866428814, + "grad_norm": 1.7717833518981934, + "learning_rate": 9.515220106880802e-06, + "loss": 3.1576, + "step": 13220 + }, + { + "epoch": 1.059890561423904, + "grad_norm": 1.8266260623931885, + "learning_rate": 9.513923190562115e-06, + "loss": 3.2919, + "step": 13221 + }, + { + "epoch": 1.0599707362049267, + "grad_norm": 1.8894011974334717, + "learning_rate": 9.512626282438514e-06, + "loss": 3.1653, + "step": 13222 + }, + { + "epoch": 1.0600509109859493, + "grad_norm": 1.9554275274276733, + "learning_rate": 9.511329382531857e-06, + "loss": 3.2509, + "step": 13223 + }, + { + "epoch": 1.060131085766972, + "grad_norm": 1.744078278541565, + "learning_rate": 9.510032490864019e-06, + "loss": 3.1105, + "step": 13224 + }, + { + "epoch": 1.0602112605479945, + "grad_norm": 1.786887288093567, + "learning_rate": 9.508735607456858e-06, + "loss": 3.1127, + "step": 13225 + }, + { + "epoch": 1.0602914353290172, + "grad_norm": 1.7952675819396973, + "learning_rate": 9.507438732332242e-06, + "loss": 3.2076, + "step": 13226 + }, + { + "epoch": 1.0603716101100398, + "grad_norm": 1.7256674766540527, + "learning_rate": 9.506141865512036e-06, + "loss": 3.1935, + "step": 13227 + }, + { + "epoch": 1.0604517848910626, + "grad_norm": 1.762213110923767, + "learning_rate": 9.504845007018104e-06, + "loss": 3.1491, + "step": 13228 + }, + { + "epoch": 1.0605319596720852, + "grad_norm": 1.7650091648101807, + "learning_rate": 9.503548156872312e-06, + "loss": 3.1728, + "step": 13229 + }, + { + "epoch": 1.0606121344531079, + "grad_norm": 1.6735461950302124, + "learning_rate": 9.50225131509652e-06, + "loss": 3.1783, + "step": 13230 + }, + { + "epoch": 1.0606923092341305, + "grad_norm": 1.7799417972564697, + "learning_rate": 9.500954481712596e-06, + "loss": 3.2213, + "step": 13231 + }, + { + "epoch": 1.0607724840151531, + "grad_norm": 1.8092432022094727, + "learning_rate": 9.499657656742405e-06, + "loss": 3.1981, + "step": 13232 + }, + { + "epoch": 1.0608526587961757, + "grad_norm": 1.8122990131378174, + "learning_rate": 9.49836084020781e-06, + "loss": 3.203, + "step": 13233 + }, + { + "epoch": 1.0609328335771984, + "grad_norm": 1.7747222185134888, + "learning_rate": 9.497064032130674e-06, + "loss": 3.1467, + "step": 13234 + }, + { + "epoch": 1.061013008358221, + "grad_norm": 1.7216931581497192, + "learning_rate": 9.49576723253286e-06, + "loss": 3.2222, + "step": 13235 + }, + { + "epoch": 1.0610931831392436, + "grad_norm": 1.9606356620788574, + "learning_rate": 9.494470441436233e-06, + "loss": 3.3024, + "step": 13236 + }, + { + "epoch": 1.0611733579202662, + "grad_norm": 2.060098648071289, + "learning_rate": 9.493173658862657e-06, + "loss": 3.2906, + "step": 13237 + }, + { + "epoch": 1.0612535327012889, + "grad_norm": 1.8144570589065552, + "learning_rate": 9.491876884833996e-06, + "loss": 3.1471, + "step": 13238 + }, + { + "epoch": 1.0613337074823115, + "grad_norm": 1.8043054342269897, + "learning_rate": 9.49058011937211e-06, + "loss": 3.1644, + "step": 13239 + }, + { + "epoch": 1.061413882263334, + "grad_norm": 1.7292912006378174, + "learning_rate": 9.489283362498862e-06, + "loss": 3.1136, + "step": 13240 + }, + { + "epoch": 1.0614940570443567, + "grad_norm": 1.7165288925170898, + "learning_rate": 9.487986614236118e-06, + "loss": 3.2113, + "step": 13241 + }, + { + "epoch": 1.0615742318253794, + "grad_norm": 1.7652788162231445, + "learning_rate": 9.48668987460574e-06, + "loss": 3.2636, + "step": 13242 + }, + { + "epoch": 1.061654406606402, + "grad_norm": 1.775343418121338, + "learning_rate": 9.485393143629588e-06, + "loss": 3.2487, + "step": 13243 + }, + { + "epoch": 1.0617345813874246, + "grad_norm": 1.7591357231140137, + "learning_rate": 9.484096421329528e-06, + "loss": 3.2626, + "step": 13244 + }, + { + "epoch": 1.0618147561684472, + "grad_norm": 1.7337613105773926, + "learning_rate": 9.48279970772742e-06, + "loss": 3.1672, + "step": 13245 + }, + { + "epoch": 1.0618949309494699, + "grad_norm": 1.8418914079666138, + "learning_rate": 9.481503002845125e-06, + "loss": 3.1145, + "step": 13246 + }, + { + "epoch": 1.0619751057304925, + "grad_norm": 1.8048955202102661, + "learning_rate": 9.480206306704509e-06, + "loss": 3.1094, + "step": 13247 + }, + { + "epoch": 1.062055280511515, + "grad_norm": 1.7665547132492065, + "learning_rate": 9.47890961932743e-06, + "loss": 3.0686, + "step": 13248 + }, + { + "epoch": 1.0621354552925377, + "grad_norm": 0.7263717651367188, + "learning_rate": 9.477612940735754e-06, + "loss": 2.5031, + "step": 13249 + }, + { + "epoch": 1.0622156300735603, + "grad_norm": 1.8960627317428589, + "learning_rate": 9.476316270951335e-06, + "loss": 3.2023, + "step": 13250 + }, + { + "epoch": 1.062295804854583, + "grad_norm": 1.9144915342330933, + "learning_rate": 9.47501960999604e-06, + "loss": 3.1892, + "step": 13251 + }, + { + "epoch": 1.0623759796356056, + "grad_norm": 1.8139562606811523, + "learning_rate": 9.473722957891734e-06, + "loss": 3.2302, + "step": 13252 + }, + { + "epoch": 1.0624561544166282, + "grad_norm": 0.7153328061103821, + "learning_rate": 9.472426314660269e-06, + "loss": 2.4824, + "step": 13253 + }, + { + "epoch": 1.0625363291976508, + "grad_norm": 1.7783159017562866, + "learning_rate": 9.471129680323513e-06, + "loss": 3.1543, + "step": 13254 + }, + { + "epoch": 1.0626165039786735, + "grad_norm": 1.9536749124526978, + "learning_rate": 9.469833054903321e-06, + "loss": 3.1049, + "step": 13255 + }, + { + "epoch": 1.062696678759696, + "grad_norm": 0.703855574131012, + "learning_rate": 9.468536438421558e-06, + "loss": 2.4657, + "step": 13256 + }, + { + "epoch": 1.0627768535407187, + "grad_norm": 1.8564485311508179, + "learning_rate": 9.467239830900086e-06, + "loss": 3.2994, + "step": 13257 + }, + { + "epoch": 1.0628570283217413, + "grad_norm": 1.599754810333252, + "learning_rate": 9.465943232360761e-06, + "loss": 3.0972, + "step": 13258 + }, + { + "epoch": 1.062937203102764, + "grad_norm": 0.7178443670272827, + "learning_rate": 9.464646642825445e-06, + "loss": 2.4799, + "step": 13259 + }, + { + "epoch": 1.0630173778837866, + "grad_norm": 1.819743275642395, + "learning_rate": 9.463350062315997e-06, + "loss": 3.3422, + "step": 13260 + }, + { + "epoch": 1.0630975526648092, + "grad_norm": 1.7993476390838623, + "learning_rate": 9.462053490854277e-06, + "loss": 3.1278, + "step": 13261 + }, + { + "epoch": 1.0631777274458318, + "grad_norm": 0.7223025560379028, + "learning_rate": 9.460756928462149e-06, + "loss": 2.4792, + "step": 13262 + }, + { + "epoch": 1.0632579022268545, + "grad_norm": 1.85601806640625, + "learning_rate": 9.459460375161465e-06, + "loss": 3.33, + "step": 13263 + }, + { + "epoch": 1.063338077007877, + "grad_norm": 1.7711771726608276, + "learning_rate": 9.458163830974093e-06, + "loss": 3.1954, + "step": 13264 + }, + { + "epoch": 1.0634182517888997, + "grad_norm": 1.708638310432434, + "learning_rate": 9.456867295921884e-06, + "loss": 3.1504, + "step": 13265 + }, + { + "epoch": 1.0634984265699223, + "grad_norm": 1.8470839262008667, + "learning_rate": 9.455570770026701e-06, + "loss": 3.1899, + "step": 13266 + }, + { + "epoch": 1.0635786013509452, + "grad_norm": 1.7283494472503662, + "learning_rate": 9.454274253310408e-06, + "loss": 3.1909, + "step": 13267 + }, + { + "epoch": 1.0636587761319678, + "grad_norm": 1.7739698886871338, + "learning_rate": 9.452977745794854e-06, + "loss": 3.2097, + "step": 13268 + }, + { + "epoch": 1.0637389509129904, + "grad_norm": 1.8080452680587769, + "learning_rate": 9.451681247501903e-06, + "loss": 3.2231, + "step": 13269 + }, + { + "epoch": 1.063819125694013, + "grad_norm": 1.805156946182251, + "learning_rate": 9.450384758453414e-06, + "loss": 3.2062, + "step": 13270 + }, + { + "epoch": 1.0638993004750357, + "grad_norm": 1.884727954864502, + "learning_rate": 9.449088278671242e-06, + "loss": 3.2007, + "step": 13271 + }, + { + "epoch": 1.0639794752560583, + "grad_norm": 1.7352882623672485, + "learning_rate": 9.44779180817725e-06, + "loss": 3.232, + "step": 13272 + }, + { + "epoch": 1.064059650037081, + "grad_norm": 1.8166773319244385, + "learning_rate": 9.446495346993293e-06, + "loss": 3.2096, + "step": 13273 + }, + { + "epoch": 1.0641398248181035, + "grad_norm": 1.8541170358657837, + "learning_rate": 9.44519889514123e-06, + "loss": 3.2339, + "step": 13274 + }, + { + "epoch": 1.0642199995991262, + "grad_norm": 0.7781304121017456, + "learning_rate": 9.443902452642916e-06, + "loss": 2.4803, + "step": 13275 + }, + { + "epoch": 1.0643001743801488, + "grad_norm": 1.7358850240707397, + "learning_rate": 9.442606019520211e-06, + "loss": 3.1479, + "step": 13276 + }, + { + "epoch": 1.0643803491611714, + "grad_norm": 1.7562813758850098, + "learning_rate": 9.441309595794976e-06, + "loss": 3.1436, + "step": 13277 + }, + { + "epoch": 1.064460523942194, + "grad_norm": 1.9056408405303955, + "learning_rate": 9.440013181489059e-06, + "loss": 3.2632, + "step": 13278 + }, + { + "epoch": 1.0645406987232167, + "grad_norm": 1.742840051651001, + "learning_rate": 9.438716776624327e-06, + "loss": 3.1388, + "step": 13279 + }, + { + "epoch": 1.0646208735042393, + "grad_norm": 1.7746042013168335, + "learning_rate": 9.437420381222628e-06, + "loss": 3.1927, + "step": 13280 + }, + { + "epoch": 1.064701048285262, + "grad_norm": 1.7176614999771118, + "learning_rate": 9.436123995305825e-06, + "loss": 3.1645, + "step": 13281 + }, + { + "epoch": 1.0647812230662845, + "grad_norm": 1.8238636255264282, + "learning_rate": 9.434827618895774e-06, + "loss": 3.1326, + "step": 13282 + }, + { + "epoch": 1.0648613978473072, + "grad_norm": 1.7337310314178467, + "learning_rate": 9.433531252014328e-06, + "loss": 3.1582, + "step": 13283 + }, + { + "epoch": 1.0649415726283298, + "grad_norm": 1.755285620689392, + "learning_rate": 9.432234894683349e-06, + "loss": 3.2071, + "step": 13284 + }, + { + "epoch": 1.0650217474093524, + "grad_norm": 1.7647514343261719, + "learning_rate": 9.430938546924685e-06, + "loss": 3.098, + "step": 13285 + }, + { + "epoch": 1.065101922190375, + "grad_norm": 1.7581584453582764, + "learning_rate": 9.429642208760198e-06, + "loss": 3.1478, + "step": 13286 + }, + { + "epoch": 1.0651820969713977, + "grad_norm": 1.7304657697677612, + "learning_rate": 9.428345880211745e-06, + "loss": 3.2061, + "step": 13287 + }, + { + "epoch": 1.0652622717524203, + "grad_norm": 1.7876368761062622, + "learning_rate": 9.427049561301175e-06, + "loss": 3.2336, + "step": 13288 + }, + { + "epoch": 1.065342446533443, + "grad_norm": 1.8839740753173828, + "learning_rate": 9.425753252050351e-06, + "loss": 3.2706, + "step": 13289 + }, + { + "epoch": 1.0654226213144655, + "grad_norm": 1.813515067100525, + "learning_rate": 9.424456952481123e-06, + "loss": 3.2506, + "step": 13290 + }, + { + "epoch": 1.0655027960954881, + "grad_norm": 1.6874548196792603, + "learning_rate": 9.423160662615345e-06, + "loss": 3.1327, + "step": 13291 + }, + { + "epoch": 1.0655829708765108, + "grad_norm": 1.8327354192733765, + "learning_rate": 9.421864382474882e-06, + "loss": 3.1646, + "step": 13292 + }, + { + "epoch": 1.0656631456575334, + "grad_norm": 1.7468881607055664, + "learning_rate": 9.420568112081575e-06, + "loss": 3.2456, + "step": 13293 + }, + { + "epoch": 1.065743320438556, + "grad_norm": 1.7177228927612305, + "learning_rate": 9.41927185145729e-06, + "loss": 3.1947, + "step": 13294 + }, + { + "epoch": 1.0658234952195786, + "grad_norm": 1.7157573699951172, + "learning_rate": 9.417975600623874e-06, + "loss": 3.1635, + "step": 13295 + }, + { + "epoch": 1.0659036700006013, + "grad_norm": 1.7814767360687256, + "learning_rate": 9.416679359603184e-06, + "loss": 3.2422, + "step": 13296 + }, + { + "epoch": 1.065983844781624, + "grad_norm": 1.6375612020492554, + "learning_rate": 9.415383128417078e-06, + "loss": 3.2086, + "step": 13297 + }, + { + "epoch": 1.0660640195626465, + "grad_norm": 1.7841273546218872, + "learning_rate": 9.414086907087402e-06, + "loss": 3.1794, + "step": 13298 + }, + { + "epoch": 1.0661441943436691, + "grad_norm": 1.966489315032959, + "learning_rate": 9.41279069563602e-06, + "loss": 3.2141, + "step": 13299 + }, + { + "epoch": 1.0662243691246918, + "grad_norm": 1.760496735572815, + "learning_rate": 9.411494494084774e-06, + "loss": 3.2343, + "step": 13300 + }, + { + "epoch": 1.0663045439057144, + "grad_norm": 1.758844256401062, + "learning_rate": 9.410198302455526e-06, + "loss": 3.2617, + "step": 13301 + }, + { + "epoch": 1.066384718686737, + "grad_norm": 1.7177839279174805, + "learning_rate": 9.408902120770126e-06, + "loss": 3.1328, + "step": 13302 + }, + { + "epoch": 1.0664648934677596, + "grad_norm": 1.7228859663009644, + "learning_rate": 9.407605949050429e-06, + "loss": 3.2069, + "step": 13303 + }, + { + "epoch": 1.0665450682487823, + "grad_norm": 1.8764444589614868, + "learning_rate": 9.406309787318288e-06, + "loss": 3.208, + "step": 13304 + }, + { + "epoch": 1.0666252430298049, + "grad_norm": 1.844085693359375, + "learning_rate": 9.40501363559555e-06, + "loss": 3.1913, + "step": 13305 + }, + { + "epoch": 1.0667054178108275, + "grad_norm": 1.7382572889328003, + "learning_rate": 9.403717493904077e-06, + "loss": 3.1714, + "step": 13306 + }, + { + "epoch": 1.0667855925918501, + "grad_norm": 1.817572832107544, + "learning_rate": 9.402421362265716e-06, + "loss": 3.2232, + "step": 13307 + }, + { + "epoch": 1.0668657673728728, + "grad_norm": 1.7564620971679688, + "learning_rate": 9.401125240702319e-06, + "loss": 3.1166, + "step": 13308 + }, + { + "epoch": 1.0669459421538954, + "grad_norm": 1.6986531019210815, + "learning_rate": 9.39982912923574e-06, + "loss": 3.1237, + "step": 13309 + }, + { + "epoch": 1.0670261169349182, + "grad_norm": 1.8959356546401978, + "learning_rate": 9.39853302788783e-06, + "loss": 3.2858, + "step": 13310 + }, + { + "epoch": 1.0671062917159408, + "grad_norm": 1.7097444534301758, + "learning_rate": 9.397236936680442e-06, + "loss": 3.1826, + "step": 13311 + }, + { + "epoch": 1.0671864664969635, + "grad_norm": 1.7904448509216309, + "learning_rate": 9.395940855635428e-06, + "loss": 3.264, + "step": 13312 + }, + { + "epoch": 1.067266641277986, + "grad_norm": 1.716436743736267, + "learning_rate": 9.394644784774635e-06, + "loss": 3.1526, + "step": 13313 + }, + { + "epoch": 1.0673468160590087, + "grad_norm": 1.8253505229949951, + "learning_rate": 9.39334872411992e-06, + "loss": 3.1567, + "step": 13314 + }, + { + "epoch": 1.0674269908400313, + "grad_norm": 1.7240275144577026, + "learning_rate": 9.392052673693131e-06, + "loss": 3.1794, + "step": 13315 + }, + { + "epoch": 1.067507165621054, + "grad_norm": 1.760960578918457, + "learning_rate": 9.390756633516119e-06, + "loss": 3.1263, + "step": 13316 + }, + { + "epoch": 1.0675873404020766, + "grad_norm": 1.8030438423156738, + "learning_rate": 9.389460603610736e-06, + "loss": 3.2166, + "step": 13317 + }, + { + "epoch": 1.0676675151830992, + "grad_norm": 1.7645268440246582, + "learning_rate": 9.388164583998832e-06, + "loss": 3.3539, + "step": 13318 + }, + { + "epoch": 1.0677476899641218, + "grad_norm": 1.6790592670440674, + "learning_rate": 9.386868574702259e-06, + "loss": 3.1335, + "step": 13319 + }, + { + "epoch": 1.0678278647451445, + "grad_norm": 1.7220741510391235, + "learning_rate": 9.385572575742864e-06, + "loss": 3.1846, + "step": 13320 + }, + { + "epoch": 1.067908039526167, + "grad_norm": 1.9882456064224243, + "learning_rate": 9.384276587142497e-06, + "loss": 3.2457, + "step": 13321 + }, + { + "epoch": 1.0679882143071897, + "grad_norm": 1.7829903364181519, + "learning_rate": 9.382980608923014e-06, + "loss": 3.1121, + "step": 13322 + }, + { + "epoch": 1.0680683890882123, + "grad_norm": 1.7324237823486328, + "learning_rate": 9.381684641106256e-06, + "loss": 3.1758, + "step": 13323 + }, + { + "epoch": 1.068148563869235, + "grad_norm": 1.7170889377593994, + "learning_rate": 9.380388683714082e-06, + "loss": 3.2379, + "step": 13324 + }, + { + "epoch": 1.0682287386502576, + "grad_norm": 1.7690685987472534, + "learning_rate": 9.379092736768332e-06, + "loss": 3.1577, + "step": 13325 + }, + { + "epoch": 1.0683089134312802, + "grad_norm": 1.8329240083694458, + "learning_rate": 9.377796800290861e-06, + "loss": 3.1979, + "step": 13326 + }, + { + "epoch": 1.0683890882123028, + "grad_norm": 1.7789143323898315, + "learning_rate": 9.37650087430352e-06, + "loss": 3.1334, + "step": 13327 + }, + { + "epoch": 1.0684692629933255, + "grad_norm": 1.6953909397125244, + "learning_rate": 9.37520495882815e-06, + "loss": 3.1707, + "step": 13328 + }, + { + "epoch": 1.068549437774348, + "grad_norm": 2.037386894226074, + "learning_rate": 9.373909053886606e-06, + "loss": 3.3285, + "step": 13329 + }, + { + "epoch": 1.0686296125553707, + "grad_norm": 1.820974588394165, + "learning_rate": 9.372613159500734e-06, + "loss": 3.2321, + "step": 13330 + }, + { + "epoch": 1.0687097873363933, + "grad_norm": 1.7652249336242676, + "learning_rate": 9.371317275692383e-06, + "loss": 3.2501, + "step": 13331 + }, + { + "epoch": 1.068789962117416, + "grad_norm": 1.8221368789672852, + "learning_rate": 9.370021402483403e-06, + "loss": 3.1572, + "step": 13332 + }, + { + "epoch": 1.0688701368984386, + "grad_norm": 1.8628031015396118, + "learning_rate": 9.368725539895638e-06, + "loss": 3.1832, + "step": 13333 + }, + { + "epoch": 1.0689503116794612, + "grad_norm": 0.6805848479270935, + "learning_rate": 9.367429687950938e-06, + "loss": 2.4768, + "step": 13334 + }, + { + "epoch": 1.0690304864604838, + "grad_norm": 1.7155159711837769, + "learning_rate": 9.366133846671153e-06, + "loss": 3.1955, + "step": 13335 + }, + { + "epoch": 1.0691106612415064, + "grad_norm": 1.8325011730194092, + "learning_rate": 9.364838016078125e-06, + "loss": 3.1835, + "step": 13336 + }, + { + "epoch": 1.069190836022529, + "grad_norm": 1.7773901224136353, + "learning_rate": 9.363542196193708e-06, + "loss": 3.2568, + "step": 13337 + }, + { + "epoch": 1.0692710108035517, + "grad_norm": 1.9070029258728027, + "learning_rate": 9.362246387039739e-06, + "loss": 3.1483, + "step": 13338 + }, + { + "epoch": 1.0693511855845743, + "grad_norm": 1.725193738937378, + "learning_rate": 9.360950588638075e-06, + "loss": 3.189, + "step": 13339 + }, + { + "epoch": 1.069431360365597, + "grad_norm": 1.8285621404647827, + "learning_rate": 9.359654801010561e-06, + "loss": 3.1667, + "step": 13340 + }, + { + "epoch": 1.0695115351466196, + "grad_norm": 0.6984085440635681, + "learning_rate": 9.358359024179036e-06, + "loss": 2.4694, + "step": 13341 + }, + { + "epoch": 1.0695917099276422, + "grad_norm": 1.793546438217163, + "learning_rate": 9.357063258165357e-06, + "loss": 3.1978, + "step": 13342 + }, + { + "epoch": 1.0696718847086648, + "grad_norm": 1.8299018144607544, + "learning_rate": 9.355767502991362e-06, + "loss": 3.2072, + "step": 13343 + }, + { + "epoch": 1.0697520594896874, + "grad_norm": 1.7934575080871582, + "learning_rate": 9.3544717586789e-06, + "loss": 3.2565, + "step": 13344 + }, + { + "epoch": 1.06983223427071, + "grad_norm": 1.8957111835479736, + "learning_rate": 9.353176025249818e-06, + "loss": 3.184, + "step": 13345 + }, + { + "epoch": 1.0699124090517327, + "grad_norm": 1.7440940141677856, + "learning_rate": 9.351880302725958e-06, + "loss": 3.2134, + "step": 13346 + }, + { + "epoch": 1.0699925838327553, + "grad_norm": 1.7881927490234375, + "learning_rate": 9.35058459112917e-06, + "loss": 3.2253, + "step": 13347 + }, + { + "epoch": 1.0700727586137782, + "grad_norm": 1.8027671575546265, + "learning_rate": 9.349288890481297e-06, + "loss": 3.214, + "step": 13348 + }, + { + "epoch": 1.0701529333948008, + "grad_norm": 1.7874218225479126, + "learning_rate": 9.34799320080418e-06, + "loss": 3.114, + "step": 13349 + }, + { + "epoch": 1.0702331081758234, + "grad_norm": 0.7279508113861084, + "learning_rate": 9.346697522119674e-06, + "loss": 2.4893, + "step": 13350 + }, + { + "epoch": 1.070313282956846, + "grad_norm": 1.773855209350586, + "learning_rate": 9.345401854449613e-06, + "loss": 3.1984, + "step": 13351 + }, + { + "epoch": 1.0703934577378686, + "grad_norm": 1.7786122560501099, + "learning_rate": 9.344106197815848e-06, + "loss": 3.1521, + "step": 13352 + }, + { + "epoch": 1.0704736325188913, + "grad_norm": 1.9484492540359497, + "learning_rate": 9.34281055224022e-06, + "loss": 3.1355, + "step": 13353 + }, + { + "epoch": 1.070553807299914, + "grad_norm": 1.90231454372406, + "learning_rate": 9.341514917744574e-06, + "loss": 3.2548, + "step": 13354 + }, + { + "epoch": 1.0706339820809365, + "grad_norm": 1.8346846103668213, + "learning_rate": 9.340219294350757e-06, + "loss": 3.1646, + "step": 13355 + }, + { + "epoch": 1.0707141568619591, + "grad_norm": 1.697798490524292, + "learning_rate": 9.338923682080609e-06, + "loss": 3.1425, + "step": 13356 + }, + { + "epoch": 1.0707943316429818, + "grad_norm": 1.7052382230758667, + "learning_rate": 9.337628080955975e-06, + "loss": 3.175, + "step": 13357 + }, + { + "epoch": 1.0708745064240044, + "grad_norm": 1.717303991317749, + "learning_rate": 9.336332490998696e-06, + "loss": 3.1372, + "step": 13358 + }, + { + "epoch": 1.070954681205027, + "grad_norm": 1.8429614305496216, + "learning_rate": 9.335036912230618e-06, + "loss": 3.2648, + "step": 13359 + }, + { + "epoch": 1.0710348559860496, + "grad_norm": 1.645580530166626, + "learning_rate": 9.333741344673585e-06, + "loss": 3.2006, + "step": 13360 + }, + { + "epoch": 1.0711150307670723, + "grad_norm": 1.7456773519515991, + "learning_rate": 9.332445788349438e-06, + "loss": 3.2726, + "step": 13361 + }, + { + "epoch": 1.0711952055480949, + "grad_norm": 0.7690947651863098, + "learning_rate": 9.33115024328002e-06, + "loss": 2.4901, + "step": 13362 + }, + { + "epoch": 1.0712753803291175, + "grad_norm": 1.6092445850372314, + "learning_rate": 9.329854709487171e-06, + "loss": 3.1752, + "step": 13363 + }, + { + "epoch": 1.0713555551101401, + "grad_norm": 1.8546003103256226, + "learning_rate": 9.328559186992737e-06, + "loss": 3.1801, + "step": 13364 + }, + { + "epoch": 1.0714357298911628, + "grad_norm": 1.7624510526657104, + "learning_rate": 9.32726367581856e-06, + "loss": 3.1741, + "step": 13365 + }, + { + "epoch": 1.0715159046721854, + "grad_norm": 1.7346832752227783, + "learning_rate": 9.325968175986476e-06, + "loss": 3.2104, + "step": 13366 + }, + { + "epoch": 1.071596079453208, + "grad_norm": 1.8267145156860352, + "learning_rate": 9.324672687518335e-06, + "loss": 3.1469, + "step": 13367 + }, + { + "epoch": 1.0716762542342306, + "grad_norm": 1.803281307220459, + "learning_rate": 9.323377210435973e-06, + "loss": 3.2701, + "step": 13368 + }, + { + "epoch": 1.0717564290152533, + "grad_norm": 1.7959649562835693, + "learning_rate": 9.322081744761232e-06, + "loss": 3.1887, + "step": 13369 + }, + { + "epoch": 1.0718366037962759, + "grad_norm": 1.9583827257156372, + "learning_rate": 9.320786290515956e-06, + "loss": 3.1256, + "step": 13370 + }, + { + "epoch": 1.0719167785772985, + "grad_norm": 1.761080026626587, + "learning_rate": 9.31949084772198e-06, + "loss": 3.2128, + "step": 13371 + }, + { + "epoch": 1.0719969533583211, + "grad_norm": 2.6153528690338135, + "learning_rate": 9.318195416401153e-06, + "loss": 3.1614, + "step": 13372 + }, + { + "epoch": 1.0720771281393437, + "grad_norm": 1.7751257419586182, + "learning_rate": 9.316899996575307e-06, + "loss": 3.1987, + "step": 13373 + }, + { + "epoch": 1.0721573029203664, + "grad_norm": 1.723353385925293, + "learning_rate": 9.315604588266286e-06, + "loss": 3.1317, + "step": 13374 + }, + { + "epoch": 1.072237477701389, + "grad_norm": 1.8055092096328735, + "learning_rate": 9.314309191495934e-06, + "loss": 3.2083, + "step": 13375 + }, + { + "epoch": 1.0723176524824116, + "grad_norm": 1.7767890691757202, + "learning_rate": 9.313013806286086e-06, + "loss": 3.1696, + "step": 13376 + }, + { + "epoch": 1.0723978272634342, + "grad_norm": 1.7558965682983398, + "learning_rate": 9.311718432658585e-06, + "loss": 3.1479, + "step": 13377 + }, + { + "epoch": 1.0724780020444569, + "grad_norm": 0.7009807825088501, + "learning_rate": 9.310423070635266e-06, + "loss": 2.4791, + "step": 13378 + }, + { + "epoch": 1.0725581768254795, + "grad_norm": 1.766813039779663, + "learning_rate": 9.30912772023797e-06, + "loss": 3.3983, + "step": 13379 + }, + { + "epoch": 1.0726383516065021, + "grad_norm": 1.832014799118042, + "learning_rate": 9.30783238148854e-06, + "loss": 3.0813, + "step": 13380 + }, + { + "epoch": 1.0727185263875247, + "grad_norm": 1.8186311721801758, + "learning_rate": 9.30653705440881e-06, + "loss": 3.1558, + "step": 13381 + }, + { + "epoch": 1.0727987011685474, + "grad_norm": 1.781646966934204, + "learning_rate": 9.305241739020625e-06, + "loss": 3.2184, + "step": 13382 + }, + { + "epoch": 1.07287887594957, + "grad_norm": 1.7487667798995972, + "learning_rate": 9.303946435345814e-06, + "loss": 3.2134, + "step": 13383 + }, + { + "epoch": 1.0729590507305926, + "grad_norm": 1.771023154258728, + "learning_rate": 9.302651143406221e-06, + "loss": 3.077, + "step": 13384 + }, + { + "epoch": 1.0730392255116152, + "grad_norm": 1.7438712120056152, + "learning_rate": 9.30135586322369e-06, + "loss": 3.1638, + "step": 13385 + }, + { + "epoch": 1.0731194002926379, + "grad_norm": 0.7377460598945618, + "learning_rate": 9.300060594820046e-06, + "loss": 2.4719, + "step": 13386 + }, + { + "epoch": 1.0731995750736605, + "grad_norm": 1.8909255266189575, + "learning_rate": 9.298765338217138e-06, + "loss": 3.2578, + "step": 13387 + }, + { + "epoch": 1.073279749854683, + "grad_norm": 1.8405420780181885, + "learning_rate": 9.297470093436797e-06, + "loss": 3.2164, + "step": 13388 + }, + { + "epoch": 1.0733599246357057, + "grad_norm": 1.7813656330108643, + "learning_rate": 9.296174860500862e-06, + "loss": 3.1372, + "step": 13389 + }, + { + "epoch": 1.0734400994167284, + "grad_norm": 1.683559775352478, + "learning_rate": 9.294879639431175e-06, + "loss": 3.1255, + "step": 13390 + }, + { + "epoch": 1.0735202741977512, + "grad_norm": 1.7778830528259277, + "learning_rate": 9.293584430249565e-06, + "loss": 3.2155, + "step": 13391 + }, + { + "epoch": 1.0736004489787738, + "grad_norm": 2.1780238151550293, + "learning_rate": 9.292289232977875e-06, + "loss": 3.2634, + "step": 13392 + }, + { + "epoch": 1.0736806237597964, + "grad_norm": 1.7698962688446045, + "learning_rate": 9.290994047637936e-06, + "loss": 3.1398, + "step": 13393 + }, + { + "epoch": 1.073760798540819, + "grad_norm": 1.8332022428512573, + "learning_rate": 9.289698874251588e-06, + "loss": 3.1987, + "step": 13394 + }, + { + "epoch": 1.0738409733218417, + "grad_norm": 1.7494666576385498, + "learning_rate": 9.288403712840669e-06, + "loss": 3.1735, + "step": 13395 + }, + { + "epoch": 1.0739211481028643, + "grad_norm": 1.7462879419326782, + "learning_rate": 9.28710856342701e-06, + "loss": 3.1812, + "step": 13396 + }, + { + "epoch": 1.074001322883887, + "grad_norm": 1.7992993593215942, + "learning_rate": 9.285813426032454e-06, + "loss": 3.2601, + "step": 13397 + }, + { + "epoch": 1.0740814976649096, + "grad_norm": 1.898829698562622, + "learning_rate": 9.284518300678827e-06, + "loss": 3.1154, + "step": 13398 + }, + { + "epoch": 1.0741616724459322, + "grad_norm": 0.698987603187561, + "learning_rate": 9.28322318738797e-06, + "loss": 2.4656, + "step": 13399 + }, + { + "epoch": 1.0742418472269548, + "grad_norm": 1.853621482849121, + "learning_rate": 9.28192808618172e-06, + "loss": 3.245, + "step": 13400 + }, + { + "epoch": 1.0743220220079774, + "grad_norm": 1.7136561870574951, + "learning_rate": 9.280632997081906e-06, + "loss": 3.302, + "step": 13401 + }, + { + "epoch": 1.074402196789, + "grad_norm": 0.6876980066299438, + "learning_rate": 9.27933792011037e-06, + "loss": 2.4805, + "step": 13402 + }, + { + "epoch": 1.0744823715700227, + "grad_norm": 1.783246636390686, + "learning_rate": 9.27804285528894e-06, + "loss": 3.1654, + "step": 13403 + }, + { + "epoch": 1.0745625463510453, + "grad_norm": 0.7000048756599426, + "learning_rate": 9.276747802639451e-06, + "loss": 2.4784, + "step": 13404 + }, + { + "epoch": 1.074642721132068, + "grad_norm": 1.6564075946807861, + "learning_rate": 9.275452762183744e-06, + "loss": 3.1378, + "step": 13405 + }, + { + "epoch": 1.0747228959130906, + "grad_norm": 1.8538388013839722, + "learning_rate": 9.274157733943645e-06, + "loss": 3.179, + "step": 13406 + }, + { + "epoch": 1.0748030706941132, + "grad_norm": 1.9026150703430176, + "learning_rate": 9.272862717940994e-06, + "loss": 3.2307, + "step": 13407 + }, + { + "epoch": 1.0748832454751358, + "grad_norm": 1.8453541994094849, + "learning_rate": 9.271567714197618e-06, + "loss": 3.3257, + "step": 13408 + }, + { + "epoch": 1.0749634202561584, + "grad_norm": 1.90166437625885, + "learning_rate": 9.270272722735353e-06, + "loss": 3.2509, + "step": 13409 + }, + { + "epoch": 1.075043595037181, + "grad_norm": 1.7481862306594849, + "learning_rate": 9.268977743576036e-06, + "loss": 3.251, + "step": 13410 + }, + { + "epoch": 1.0751237698182037, + "grad_norm": 1.7467976808547974, + "learning_rate": 9.267682776741495e-06, + "loss": 3.1743, + "step": 13411 + }, + { + "epoch": 1.0752039445992263, + "grad_norm": 1.9241523742675781, + "learning_rate": 9.266387822253564e-06, + "loss": 3.1591, + "step": 13412 + }, + { + "epoch": 1.075284119380249, + "grad_norm": 1.8247257471084595, + "learning_rate": 9.265092880134075e-06, + "loss": 3.1552, + "step": 13413 + }, + { + "epoch": 1.0753642941612716, + "grad_norm": 1.7421605587005615, + "learning_rate": 9.263797950404863e-06, + "loss": 3.1952, + "step": 13414 + }, + { + "epoch": 1.0754444689422942, + "grad_norm": 1.767098307609558, + "learning_rate": 9.262503033087758e-06, + "loss": 3.2551, + "step": 13415 + }, + { + "epoch": 1.0755246437233168, + "grad_norm": 0.7099665999412537, + "learning_rate": 9.26120812820459e-06, + "loss": 2.48, + "step": 13416 + }, + { + "epoch": 1.0756048185043394, + "grad_norm": 1.7696244716644287, + "learning_rate": 9.259913235777194e-06, + "loss": 3.196, + "step": 13417 + }, + { + "epoch": 1.075684993285362, + "grad_norm": 1.7794089317321777, + "learning_rate": 9.258618355827398e-06, + "loss": 3.1818, + "step": 13418 + }, + { + "epoch": 1.0757651680663847, + "grad_norm": 1.6786837577819824, + "learning_rate": 9.257323488377037e-06, + "loss": 3.0363, + "step": 13419 + }, + { + "epoch": 1.0758453428474073, + "grad_norm": 1.836976408958435, + "learning_rate": 9.256028633447942e-06, + "loss": 3.1953, + "step": 13420 + }, + { + "epoch": 1.07592551762843, + "grad_norm": 1.9014309644699097, + "learning_rate": 9.25473379106194e-06, + "loss": 3.1361, + "step": 13421 + }, + { + "epoch": 1.0760056924094525, + "grad_norm": 1.7362359762191772, + "learning_rate": 9.253438961240864e-06, + "loss": 3.1623, + "step": 13422 + }, + { + "epoch": 1.0760858671904752, + "grad_norm": 1.7033957242965698, + "learning_rate": 9.252144144006542e-06, + "loss": 3.2274, + "step": 13423 + }, + { + "epoch": 1.0761660419714978, + "grad_norm": 1.8920726776123047, + "learning_rate": 9.250849339380808e-06, + "loss": 3.2825, + "step": 13424 + }, + { + "epoch": 1.0762462167525204, + "grad_norm": 1.7398618459701538, + "learning_rate": 9.249554547385492e-06, + "loss": 3.1927, + "step": 13425 + }, + { + "epoch": 1.076326391533543, + "grad_norm": 1.6656379699707031, + "learning_rate": 9.248259768042418e-06, + "loss": 3.1526, + "step": 13426 + }, + { + "epoch": 1.0764065663145657, + "grad_norm": 1.7653082609176636, + "learning_rate": 9.246965001373424e-06, + "loss": 3.2011, + "step": 13427 + }, + { + "epoch": 1.0764867410955883, + "grad_norm": 1.6925737857818604, + "learning_rate": 9.24567024740033e-06, + "loss": 3.1771, + "step": 13428 + }, + { + "epoch": 1.076566915876611, + "grad_norm": 1.8484992980957031, + "learning_rate": 9.244375506144972e-06, + "loss": 3.1184, + "step": 13429 + }, + { + "epoch": 1.0766470906576338, + "grad_norm": 1.7511359453201294, + "learning_rate": 9.243080777629177e-06, + "loss": 3.2083, + "step": 13430 + }, + { + "epoch": 1.0767272654386564, + "grad_norm": 0.7124728560447693, + "learning_rate": 9.241786061874771e-06, + "loss": 2.4681, + "step": 13431 + }, + { + "epoch": 1.076807440219679, + "grad_norm": 1.7205017805099487, + "learning_rate": 9.240491358903589e-06, + "loss": 3.1819, + "step": 13432 + }, + { + "epoch": 1.0768876150007016, + "grad_norm": 1.7397562265396118, + "learning_rate": 9.239196668737452e-06, + "loss": 3.1997, + "step": 13433 + }, + { + "epoch": 1.0769677897817242, + "grad_norm": 1.7862542867660522, + "learning_rate": 9.237901991398191e-06, + "loss": 3.1912, + "step": 13434 + }, + { + "epoch": 1.0770479645627469, + "grad_norm": 1.6561187505722046, + "learning_rate": 9.236607326907634e-06, + "loss": 3.2563, + "step": 13435 + }, + { + "epoch": 1.0771281393437695, + "grad_norm": 1.9821258783340454, + "learning_rate": 9.235312675287609e-06, + "loss": 3.1716, + "step": 13436 + }, + { + "epoch": 1.0772083141247921, + "grad_norm": 1.7493940591812134, + "learning_rate": 9.234018036559944e-06, + "loss": 3.2443, + "step": 13437 + }, + { + "epoch": 1.0772884889058147, + "grad_norm": 0.6796020865440369, + "learning_rate": 9.232723410746463e-06, + "loss": 2.4913, + "step": 13438 + }, + { + "epoch": 1.0773686636868374, + "grad_norm": 1.925896406173706, + "learning_rate": 9.231428797868993e-06, + "loss": 3.2038, + "step": 13439 + }, + { + "epoch": 1.07744883846786, + "grad_norm": 1.8317772150039673, + "learning_rate": 9.230134197949368e-06, + "loss": 3.3026, + "step": 13440 + }, + { + "epoch": 1.0775290132488826, + "grad_norm": 1.7776983976364136, + "learning_rate": 9.228839611009404e-06, + "loss": 3.2394, + "step": 13441 + }, + { + "epoch": 1.0776091880299052, + "grad_norm": 1.6907061338424683, + "learning_rate": 9.227545037070936e-06, + "loss": 3.1705, + "step": 13442 + }, + { + "epoch": 1.0776893628109279, + "grad_norm": 1.7567020654678345, + "learning_rate": 9.226250476155783e-06, + "loss": 3.1921, + "step": 13443 + }, + { + "epoch": 1.0777695375919505, + "grad_norm": 1.8133838176727295, + "learning_rate": 9.224955928285776e-06, + "loss": 3.1437, + "step": 13444 + }, + { + "epoch": 1.0778497123729731, + "grad_norm": 1.7410252094268799, + "learning_rate": 9.223661393482741e-06, + "loss": 3.1367, + "step": 13445 + }, + { + "epoch": 1.0779298871539957, + "grad_norm": 0.69667649269104, + "learning_rate": 9.222366871768498e-06, + "loss": 2.468, + "step": 13446 + }, + { + "epoch": 1.0780100619350184, + "grad_norm": 1.7501072883605957, + "learning_rate": 9.221072363164879e-06, + "loss": 3.2454, + "step": 13447 + }, + { + "epoch": 1.078090236716041, + "grad_norm": 1.737234354019165, + "learning_rate": 9.219777867693702e-06, + "loss": 3.2948, + "step": 13448 + }, + { + "epoch": 1.0781704114970636, + "grad_norm": 1.8027753829956055, + "learning_rate": 9.218483385376796e-06, + "loss": 3.2697, + "step": 13449 + }, + { + "epoch": 1.0782505862780862, + "grad_norm": 1.8668891191482544, + "learning_rate": 9.217188916235986e-06, + "loss": 3.1847, + "step": 13450 + }, + { + "epoch": 1.0783307610591089, + "grad_norm": 1.6723278760910034, + "learning_rate": 9.215894460293094e-06, + "loss": 3.1978, + "step": 13451 + }, + { + "epoch": 1.0784109358401315, + "grad_norm": 1.7522544860839844, + "learning_rate": 9.214600017569945e-06, + "loss": 3.2404, + "step": 13452 + }, + { + "epoch": 1.078491110621154, + "grad_norm": 1.807838797569275, + "learning_rate": 9.213305588088364e-06, + "loss": 3.1758, + "step": 13453 + }, + { + "epoch": 1.0785712854021767, + "grad_norm": 1.745100975036621, + "learning_rate": 9.21201117187017e-06, + "loss": 3.1725, + "step": 13454 + }, + { + "epoch": 1.0786514601831994, + "grad_norm": 1.7479736804962158, + "learning_rate": 9.210716768937194e-06, + "loss": 3.16, + "step": 13455 + }, + { + "epoch": 1.078731634964222, + "grad_norm": 1.8087204694747925, + "learning_rate": 9.209422379311254e-06, + "loss": 3.2775, + "step": 13456 + }, + { + "epoch": 1.0788118097452446, + "grad_norm": 1.9183242321014404, + "learning_rate": 9.208128003014174e-06, + "loss": 3.2416, + "step": 13457 + }, + { + "epoch": 1.0788919845262672, + "grad_norm": 1.802106499671936, + "learning_rate": 9.206833640067777e-06, + "loss": 3.2377, + "step": 13458 + }, + { + "epoch": 1.0789721593072898, + "grad_norm": 0.6776262521743774, + "learning_rate": 9.205539290493883e-06, + "loss": 2.4672, + "step": 13459 + }, + { + "epoch": 1.0790523340883125, + "grad_norm": 1.8637109994888306, + "learning_rate": 9.20424495431432e-06, + "loss": 3.1157, + "step": 13460 + }, + { + "epoch": 1.079132508869335, + "grad_norm": 1.7849472761154175, + "learning_rate": 9.202950631550904e-06, + "loss": 3.2419, + "step": 13461 + }, + { + "epoch": 1.0792126836503577, + "grad_norm": 1.7809193134307861, + "learning_rate": 9.201656322225462e-06, + "loss": 3.1418, + "step": 13462 + }, + { + "epoch": 1.0792928584313803, + "grad_norm": 1.6677169799804688, + "learning_rate": 9.20036202635981e-06, + "loss": 3.23, + "step": 13463 + }, + { + "epoch": 1.079373033212403, + "grad_norm": 1.9117326736450195, + "learning_rate": 9.199067743975774e-06, + "loss": 3.2285, + "step": 13464 + }, + { + "epoch": 1.0794532079934256, + "grad_norm": 2.8218555450439453, + "learning_rate": 9.197773475095174e-06, + "loss": 3.256, + "step": 13465 + }, + { + "epoch": 1.0795333827744482, + "grad_norm": 1.7375966310501099, + "learning_rate": 9.196479219739828e-06, + "loss": 3.1773, + "step": 13466 + }, + { + "epoch": 1.0796135575554708, + "grad_norm": 1.8105906248092651, + "learning_rate": 9.195184977931563e-06, + "loss": 3.1591, + "step": 13467 + }, + { + "epoch": 1.0796937323364935, + "grad_norm": 1.7093825340270996, + "learning_rate": 9.193890749692193e-06, + "loss": 3.1724, + "step": 13468 + }, + { + "epoch": 1.079773907117516, + "grad_norm": 1.7330409288406372, + "learning_rate": 9.192596535043541e-06, + "loss": 3.1687, + "step": 13469 + }, + { + "epoch": 1.0798540818985387, + "grad_norm": 1.8264002799987793, + "learning_rate": 9.191302334007429e-06, + "loss": 3.1834, + "step": 13470 + }, + { + "epoch": 1.0799342566795613, + "grad_norm": 1.7828283309936523, + "learning_rate": 9.190008146605672e-06, + "loss": 3.127, + "step": 13471 + }, + { + "epoch": 1.080014431460584, + "grad_norm": 1.728583812713623, + "learning_rate": 9.188713972860094e-06, + "loss": 3.2993, + "step": 13472 + }, + { + "epoch": 1.0800946062416068, + "grad_norm": 1.8043651580810547, + "learning_rate": 9.187419812792513e-06, + "loss": 3.2164, + "step": 13473 + }, + { + "epoch": 1.0801747810226294, + "grad_norm": 1.8607534170150757, + "learning_rate": 9.186125666424745e-06, + "loss": 3.2109, + "step": 13474 + }, + { + "epoch": 1.080254955803652, + "grad_norm": 1.7342301607131958, + "learning_rate": 9.184831533778614e-06, + "loss": 3.249, + "step": 13475 + }, + { + "epoch": 1.0803351305846747, + "grad_norm": 0.7001641392707825, + "learning_rate": 9.183537414875934e-06, + "loss": 2.4812, + "step": 13476 + }, + { + "epoch": 1.0804153053656973, + "grad_norm": 1.7078697681427002, + "learning_rate": 9.182243309738527e-06, + "loss": 3.1902, + "step": 13477 + }, + { + "epoch": 1.08049548014672, + "grad_norm": 1.7148407697677612, + "learning_rate": 9.180949218388207e-06, + "loss": 3.1605, + "step": 13478 + }, + { + "epoch": 1.0805756549277425, + "grad_norm": 0.7126491665840149, + "learning_rate": 9.179655140846796e-06, + "loss": 2.4707, + "step": 13479 + }, + { + "epoch": 1.0806558297087652, + "grad_norm": 1.6903876066207886, + "learning_rate": 9.178361077136112e-06, + "loss": 3.1723, + "step": 13480 + }, + { + "epoch": 1.0807360044897878, + "grad_norm": 1.7251111268997192, + "learning_rate": 9.177067027277968e-06, + "loss": 3.3259, + "step": 13481 + }, + { + "epoch": 1.0808161792708104, + "grad_norm": 1.7414231300354004, + "learning_rate": 9.175772991294188e-06, + "loss": 3.1997, + "step": 13482 + }, + { + "epoch": 1.080896354051833, + "grad_norm": 1.798479676246643, + "learning_rate": 9.174478969206579e-06, + "loss": 3.1618, + "step": 13483 + }, + { + "epoch": 1.0809765288328557, + "grad_norm": 1.794956088066101, + "learning_rate": 9.173184961036967e-06, + "loss": 3.2183, + "step": 13484 + }, + { + "epoch": 1.0810567036138783, + "grad_norm": 0.6803146004676819, + "learning_rate": 9.171890966807166e-06, + "loss": 2.4609, + "step": 13485 + }, + { + "epoch": 1.081136878394901, + "grad_norm": 1.7575998306274414, + "learning_rate": 9.170596986538988e-06, + "loss": 3.3245, + "step": 13486 + }, + { + "epoch": 1.0812170531759235, + "grad_norm": 1.9612433910369873, + "learning_rate": 9.169303020254257e-06, + "loss": 3.1395, + "step": 13487 + }, + { + "epoch": 1.0812972279569462, + "grad_norm": 1.7440645694732666, + "learning_rate": 9.168009067974781e-06, + "loss": 3.2276, + "step": 13488 + }, + { + "epoch": 1.0813774027379688, + "grad_norm": 1.704079508781433, + "learning_rate": 9.16671512972238e-06, + "loss": 3.348, + "step": 13489 + }, + { + "epoch": 1.0814575775189914, + "grad_norm": 1.8411041498184204, + "learning_rate": 9.16542120551887e-06, + "loss": 3.1423, + "step": 13490 + }, + { + "epoch": 1.081537752300014, + "grad_norm": 1.8602718114852905, + "learning_rate": 9.16412729538606e-06, + "loss": 3.3102, + "step": 13491 + }, + { + "epoch": 1.0816179270810367, + "grad_norm": 1.6399861574172974, + "learning_rate": 9.162833399345774e-06, + "loss": 3.1239, + "step": 13492 + }, + { + "epoch": 1.0816981018620593, + "grad_norm": 1.7032737731933594, + "learning_rate": 9.161539517419819e-06, + "loss": 3.2034, + "step": 13493 + }, + { + "epoch": 1.081778276643082, + "grad_norm": 1.7131588459014893, + "learning_rate": 9.16024564963001e-06, + "loss": 3.0326, + "step": 13494 + }, + { + "epoch": 1.0818584514241045, + "grad_norm": 1.7517858743667603, + "learning_rate": 9.15895179599817e-06, + "loss": 3.2028, + "step": 13495 + }, + { + "epoch": 1.0819386262051272, + "grad_norm": 1.7264127731323242, + "learning_rate": 9.1576579565461e-06, + "loss": 3.2472, + "step": 13496 + }, + { + "epoch": 1.0820188009861498, + "grad_norm": 1.757979154586792, + "learning_rate": 9.156364131295624e-06, + "loss": 3.2618, + "step": 13497 + }, + { + "epoch": 1.0820989757671724, + "grad_norm": 1.7006767988204956, + "learning_rate": 9.155070320268549e-06, + "loss": 3.1949, + "step": 13498 + }, + { + "epoch": 1.082179150548195, + "grad_norm": 1.773091197013855, + "learning_rate": 9.15377652348669e-06, + "loss": 3.1718, + "step": 13499 + }, + { + "epoch": 1.0822593253292176, + "grad_norm": 1.7471214532852173, + "learning_rate": 9.152482740971864e-06, + "loss": 3.1057, + "step": 13500 + }, + { + "epoch": 1.0823395001102403, + "grad_norm": 1.8032357692718506, + "learning_rate": 9.151188972745878e-06, + "loss": 3.2217, + "step": 13501 + }, + { + "epoch": 1.082419674891263, + "grad_norm": 1.802015781402588, + "learning_rate": 9.149895218830548e-06, + "loss": 3.2826, + "step": 13502 + }, + { + "epoch": 1.0824998496722855, + "grad_norm": 1.8195548057556152, + "learning_rate": 9.148601479247684e-06, + "loss": 3.2361, + "step": 13503 + }, + { + "epoch": 1.0825800244533081, + "grad_norm": 1.9659385681152344, + "learning_rate": 9.147307754019098e-06, + "loss": 3.1924, + "step": 13504 + }, + { + "epoch": 1.0826601992343308, + "grad_norm": 1.8921095132827759, + "learning_rate": 9.146014043166605e-06, + "loss": 3.2062, + "step": 13505 + }, + { + "epoch": 1.0827403740153534, + "grad_norm": 1.8455190658569336, + "learning_rate": 9.144720346712012e-06, + "loss": 3.1578, + "step": 13506 + }, + { + "epoch": 1.082820548796376, + "grad_norm": 1.7329670190811157, + "learning_rate": 9.143426664677136e-06, + "loss": 3.1199, + "step": 13507 + }, + { + "epoch": 1.0829007235773986, + "grad_norm": 1.8145819902420044, + "learning_rate": 9.142132997083781e-06, + "loss": 3.182, + "step": 13508 + }, + { + "epoch": 1.0829808983584213, + "grad_norm": 1.7973963022232056, + "learning_rate": 9.140839343953762e-06, + "loss": 3.1931, + "step": 13509 + }, + { + "epoch": 1.0830610731394439, + "grad_norm": 0.7316559553146362, + "learning_rate": 9.13954570530889e-06, + "loss": 2.4868, + "step": 13510 + }, + { + "epoch": 1.0831412479204667, + "grad_norm": 1.7572894096374512, + "learning_rate": 9.138252081170973e-06, + "loss": 3.1749, + "step": 13511 + }, + { + "epoch": 1.0832214227014894, + "grad_norm": 1.7617186307907104, + "learning_rate": 9.136958471561825e-06, + "loss": 3.2651, + "step": 13512 + }, + { + "epoch": 1.083301597482512, + "grad_norm": 1.8413177728652954, + "learning_rate": 9.135664876503249e-06, + "loss": 3.2319, + "step": 13513 + }, + { + "epoch": 1.0833817722635346, + "grad_norm": 1.7933881282806396, + "learning_rate": 9.134371296017059e-06, + "loss": 3.2291, + "step": 13514 + }, + { + "epoch": 1.0834619470445572, + "grad_norm": 1.7059510946273804, + "learning_rate": 9.133077730125068e-06, + "loss": 3.139, + "step": 13515 + }, + { + "epoch": 1.0835421218255799, + "grad_norm": 0.6941426992416382, + "learning_rate": 9.131784178849076e-06, + "loss": 2.4747, + "step": 13516 + }, + { + "epoch": 1.0836222966066025, + "grad_norm": 1.8203685283660889, + "learning_rate": 9.1304906422109e-06, + "loss": 3.0299, + "step": 13517 + }, + { + "epoch": 1.083702471387625, + "grad_norm": 1.8012969493865967, + "learning_rate": 9.129197120232343e-06, + "loss": 3.2343, + "step": 13518 + }, + { + "epoch": 1.0837826461686477, + "grad_norm": 1.7730059623718262, + "learning_rate": 9.127903612935217e-06, + "loss": 3.205, + "step": 13519 + }, + { + "epoch": 1.0838628209496703, + "grad_norm": 1.7174394130706787, + "learning_rate": 9.126610120341331e-06, + "loss": 3.2693, + "step": 13520 + }, + { + "epoch": 1.083942995730693, + "grad_norm": 1.9128501415252686, + "learning_rate": 9.125316642472486e-06, + "loss": 3.2367, + "step": 13521 + }, + { + "epoch": 1.0840231705117156, + "grad_norm": 1.746086597442627, + "learning_rate": 9.124023179350498e-06, + "loss": 3.1412, + "step": 13522 + }, + { + "epoch": 1.0841033452927382, + "grad_norm": 1.7170764207839966, + "learning_rate": 9.122729730997168e-06, + "loss": 3.1703, + "step": 13523 + }, + { + "epoch": 1.0841835200737608, + "grad_norm": 1.8781256675720215, + "learning_rate": 9.121436297434305e-06, + "loss": 3.2952, + "step": 13524 + }, + { + "epoch": 1.0842636948547835, + "grad_norm": 1.821229338645935, + "learning_rate": 9.12014287868372e-06, + "loss": 3.1681, + "step": 13525 + }, + { + "epoch": 1.084343869635806, + "grad_norm": 1.7335233688354492, + "learning_rate": 9.118849474767212e-06, + "loss": 3.1456, + "step": 13526 + }, + { + "epoch": 1.0844240444168287, + "grad_norm": 1.7797006368637085, + "learning_rate": 9.117556085706595e-06, + "loss": 3.2336, + "step": 13527 + }, + { + "epoch": 1.0845042191978513, + "grad_norm": 0.7190922498703003, + "learning_rate": 9.116262711523668e-06, + "loss": 2.4685, + "step": 13528 + }, + { + "epoch": 1.084584393978874, + "grad_norm": 1.8217859268188477, + "learning_rate": 9.11496935224024e-06, + "loss": 3.1617, + "step": 13529 + }, + { + "epoch": 1.0846645687598966, + "grad_norm": 1.6819875240325928, + "learning_rate": 9.113676007878121e-06, + "loss": 3.0911, + "step": 13530 + }, + { + "epoch": 1.0847447435409192, + "grad_norm": 0.6851963996887207, + "learning_rate": 9.112382678459108e-06, + "loss": 2.483, + "step": 13531 + }, + { + "epoch": 1.0848249183219418, + "grad_norm": 1.7787078619003296, + "learning_rate": 9.111089364005011e-06, + "loss": 3.183, + "step": 13532 + }, + { + "epoch": 1.0849050931029645, + "grad_norm": 1.7613621950149536, + "learning_rate": 9.109796064537635e-06, + "loss": 3.2057, + "step": 13533 + }, + { + "epoch": 1.084985267883987, + "grad_norm": 1.8527394533157349, + "learning_rate": 9.108502780078784e-06, + "loss": 3.1423, + "step": 13534 + }, + { + "epoch": 1.0850654426650097, + "grad_norm": 1.7038817405700684, + "learning_rate": 9.107209510650263e-06, + "loss": 3.2494, + "step": 13535 + }, + { + "epoch": 1.0851456174460323, + "grad_norm": 1.760169267654419, + "learning_rate": 9.105916256273873e-06, + "loss": 3.3425, + "step": 13536 + }, + { + "epoch": 1.085225792227055, + "grad_norm": 1.7521024942398071, + "learning_rate": 9.104623016971419e-06, + "loss": 3.0149, + "step": 13537 + }, + { + "epoch": 1.0853059670080776, + "grad_norm": 1.8129603862762451, + "learning_rate": 9.103329792764708e-06, + "loss": 3.2177, + "step": 13538 + }, + { + "epoch": 1.0853861417891002, + "grad_norm": 1.7240029573440552, + "learning_rate": 9.102036583675537e-06, + "loss": 3.1311, + "step": 13539 + }, + { + "epoch": 1.0854663165701228, + "grad_norm": 1.7013987302780151, + "learning_rate": 9.100743389725717e-06, + "loss": 3.1825, + "step": 13540 + }, + { + "epoch": 1.0855464913511454, + "grad_norm": 1.8485486507415771, + "learning_rate": 9.099450210937045e-06, + "loss": 3.1667, + "step": 13541 + }, + { + "epoch": 1.085626666132168, + "grad_norm": 1.6882203817367554, + "learning_rate": 9.098157047331321e-06, + "loss": 3.1348, + "step": 13542 + }, + { + "epoch": 1.0857068409131907, + "grad_norm": 1.7602752447128296, + "learning_rate": 9.096863898930357e-06, + "loss": 3.1132, + "step": 13543 + }, + { + "epoch": 1.0857870156942133, + "grad_norm": 1.821099877357483, + "learning_rate": 9.095570765755946e-06, + "loss": 3.2101, + "step": 13544 + }, + { + "epoch": 1.085867190475236, + "grad_norm": 1.7116649150848389, + "learning_rate": 9.094277647829896e-06, + "loss": 3.1755, + "step": 13545 + }, + { + "epoch": 1.0859473652562586, + "grad_norm": 1.8553824424743652, + "learning_rate": 9.092984545174004e-06, + "loss": 3.2543, + "step": 13546 + }, + { + "epoch": 1.0860275400372812, + "grad_norm": 1.769902229309082, + "learning_rate": 9.091691457810073e-06, + "loss": 3.2904, + "step": 13547 + }, + { + "epoch": 1.0861077148183038, + "grad_norm": 1.7794840335845947, + "learning_rate": 9.090398385759904e-06, + "loss": 3.1287, + "step": 13548 + }, + { + "epoch": 1.0861878895993264, + "grad_norm": 1.766463279724121, + "learning_rate": 9.089105329045296e-06, + "loss": 3.1338, + "step": 13549 + }, + { + "epoch": 1.086268064380349, + "grad_norm": 1.744482159614563, + "learning_rate": 9.087812287688055e-06, + "loss": 3.1956, + "step": 13550 + }, + { + "epoch": 1.0863482391613717, + "grad_norm": 1.8454197645187378, + "learning_rate": 9.086519261709974e-06, + "loss": 3.2809, + "step": 13551 + }, + { + "epoch": 1.0864284139423943, + "grad_norm": 1.7401612997055054, + "learning_rate": 9.085226251132857e-06, + "loss": 3.2004, + "step": 13552 + }, + { + "epoch": 1.086508588723417, + "grad_norm": 1.8881474733352661, + "learning_rate": 9.083933255978503e-06, + "loss": 3.1616, + "step": 13553 + }, + { + "epoch": 1.0865887635044398, + "grad_norm": 1.8354177474975586, + "learning_rate": 9.082640276268711e-06, + "loss": 3.2028, + "step": 13554 + }, + { + "epoch": 1.0866689382854624, + "grad_norm": 1.7205348014831543, + "learning_rate": 9.081347312025283e-06, + "loss": 3.2354, + "step": 13555 + }, + { + "epoch": 1.086749113066485, + "grad_norm": 1.7696925401687622, + "learning_rate": 9.080054363270014e-06, + "loss": 3.1083, + "step": 13556 + }, + { + "epoch": 1.0868292878475077, + "grad_norm": 1.7705098390579224, + "learning_rate": 9.078761430024702e-06, + "loss": 3.1226, + "step": 13557 + }, + { + "epoch": 1.0869094626285303, + "grad_norm": 1.6564046144485474, + "learning_rate": 9.077468512311153e-06, + "loss": 3.1928, + "step": 13558 + }, + { + "epoch": 1.086989637409553, + "grad_norm": 1.8213599920272827, + "learning_rate": 9.076175610151154e-06, + "loss": 3.2078, + "step": 13559 + }, + { + "epoch": 1.0870698121905755, + "grad_norm": 1.7037734985351562, + "learning_rate": 9.074882723566512e-06, + "loss": 3.2373, + "step": 13560 + }, + { + "epoch": 1.0871499869715981, + "grad_norm": 0.7620607018470764, + "learning_rate": 9.07358985257902e-06, + "loss": 2.4712, + "step": 13561 + }, + { + "epoch": 1.0872301617526208, + "grad_norm": 1.846777319908142, + "learning_rate": 9.072296997210476e-06, + "loss": 3.1494, + "step": 13562 + }, + { + "epoch": 1.0873103365336434, + "grad_norm": 1.702903151512146, + "learning_rate": 9.07100415748268e-06, + "loss": 3.1945, + "step": 13563 + }, + { + "epoch": 1.087390511314666, + "grad_norm": 1.808139681816101, + "learning_rate": 9.069711333417426e-06, + "loss": 3.1179, + "step": 13564 + }, + { + "epoch": 1.0874706860956886, + "grad_norm": 0.7148334980010986, + "learning_rate": 9.06841852503651e-06, + "loss": 2.4662, + "step": 13565 + }, + { + "epoch": 1.0875508608767113, + "grad_norm": 1.6978400945663452, + "learning_rate": 9.067125732361729e-06, + "loss": 3.1405, + "step": 13566 + }, + { + "epoch": 1.087631035657734, + "grad_norm": 1.9775967597961426, + "learning_rate": 9.065832955414878e-06, + "loss": 3.1593, + "step": 13567 + }, + { + "epoch": 1.0877112104387565, + "grad_norm": 1.7427034378051758, + "learning_rate": 9.064540194217759e-06, + "loss": 3.1362, + "step": 13568 + }, + { + "epoch": 1.0877913852197791, + "grad_norm": 1.7175984382629395, + "learning_rate": 9.063247448792157e-06, + "loss": 3.3257, + "step": 13569 + }, + { + "epoch": 1.0878715600008018, + "grad_norm": 1.858812928199768, + "learning_rate": 9.061954719159878e-06, + "loss": 3.0874, + "step": 13570 + }, + { + "epoch": 1.0879517347818244, + "grad_norm": 1.8562276363372803, + "learning_rate": 9.060662005342708e-06, + "loss": 3.2645, + "step": 13571 + }, + { + "epoch": 1.088031909562847, + "grad_norm": 1.8615550994873047, + "learning_rate": 9.059369307362446e-06, + "loss": 3.2329, + "step": 13572 + }, + { + "epoch": 1.0881120843438696, + "grad_norm": 1.7938952445983887, + "learning_rate": 9.058076625240888e-06, + "loss": 3.2679, + "step": 13573 + }, + { + "epoch": 1.0881922591248923, + "grad_norm": 1.7294307947158813, + "learning_rate": 9.056783958999825e-06, + "loss": 3.1735, + "step": 13574 + }, + { + "epoch": 1.0882724339059149, + "grad_norm": 1.7606921195983887, + "learning_rate": 9.055491308661053e-06, + "loss": 3.175, + "step": 13575 + }, + { + "epoch": 1.0883526086869375, + "grad_norm": 1.8105274438858032, + "learning_rate": 9.054198674246363e-06, + "loss": 3.2026, + "step": 13576 + }, + { + "epoch": 1.0884327834679601, + "grad_norm": 1.8824774026870728, + "learning_rate": 9.052906055777549e-06, + "loss": 3.179, + "step": 13577 + }, + { + "epoch": 1.0885129582489828, + "grad_norm": 1.7819823026657104, + "learning_rate": 9.05161345327641e-06, + "loss": 3.1377, + "step": 13578 + }, + { + "epoch": 1.0885931330300054, + "grad_norm": 1.84552800655365, + "learning_rate": 9.050320866764727e-06, + "loss": 3.2054, + "step": 13579 + }, + { + "epoch": 1.088673307811028, + "grad_norm": 1.7409697771072388, + "learning_rate": 9.049028296264306e-06, + "loss": 3.1512, + "step": 13580 + }, + { + "epoch": 1.0887534825920506, + "grad_norm": 1.7788182497024536, + "learning_rate": 9.047735741796929e-06, + "loss": 3.1735, + "step": 13581 + }, + { + "epoch": 1.0888336573730732, + "grad_norm": 1.8088042736053467, + "learning_rate": 9.046443203384392e-06, + "loss": 3.1825, + "step": 13582 + }, + { + "epoch": 1.0889138321540959, + "grad_norm": 1.810666799545288, + "learning_rate": 9.045150681048489e-06, + "loss": 3.1747, + "step": 13583 + }, + { + "epoch": 1.0889940069351185, + "grad_norm": 1.9892805814743042, + "learning_rate": 9.043858174811007e-06, + "loss": 3.1846, + "step": 13584 + }, + { + "epoch": 1.0890741817161411, + "grad_norm": 1.7522674798965454, + "learning_rate": 9.042565684693741e-06, + "loss": 3.0762, + "step": 13585 + }, + { + "epoch": 1.0891543564971637, + "grad_norm": 1.7659587860107422, + "learning_rate": 9.041273210718477e-06, + "loss": 3.1641, + "step": 13586 + }, + { + "epoch": 1.0892345312781864, + "grad_norm": 1.75253164768219, + "learning_rate": 9.039980752907009e-06, + "loss": 3.2828, + "step": 13587 + }, + { + "epoch": 1.089314706059209, + "grad_norm": 1.7571858167648315, + "learning_rate": 9.03868831128113e-06, + "loss": 3.2379, + "step": 13588 + }, + { + "epoch": 1.0893948808402316, + "grad_norm": 1.6940807104110718, + "learning_rate": 9.037395885862625e-06, + "loss": 3.2854, + "step": 13589 + }, + { + "epoch": 1.0894750556212542, + "grad_norm": 1.8604222536087036, + "learning_rate": 9.036103476673288e-06, + "loss": 3.1017, + "step": 13590 + }, + { + "epoch": 1.0895552304022769, + "grad_norm": 1.800038456916809, + "learning_rate": 9.034811083734906e-06, + "loss": 3.1343, + "step": 13591 + }, + { + "epoch": 1.0896354051832997, + "grad_norm": 1.7896448373794556, + "learning_rate": 9.033518707069267e-06, + "loss": 3.2489, + "step": 13592 + }, + { + "epoch": 1.0897155799643223, + "grad_norm": 1.7863942384719849, + "learning_rate": 9.032226346698165e-06, + "loss": 3.1307, + "step": 13593 + }, + { + "epoch": 1.089795754745345, + "grad_norm": 0.6992242932319641, + "learning_rate": 9.030934002643383e-06, + "loss": 2.468, + "step": 13594 + }, + { + "epoch": 1.0898759295263676, + "grad_norm": 1.8979629278182983, + "learning_rate": 9.029641674926714e-06, + "loss": 3.306, + "step": 13595 + }, + { + "epoch": 1.0899561043073902, + "grad_norm": 1.7715944051742554, + "learning_rate": 9.028349363569943e-06, + "loss": 3.1754, + "step": 13596 + }, + { + "epoch": 1.0900362790884128, + "grad_norm": 1.7904788255691528, + "learning_rate": 9.02705706859486e-06, + "loss": 3.1715, + "step": 13597 + }, + { + "epoch": 1.0901164538694355, + "grad_norm": 1.729202151298523, + "learning_rate": 9.025764790023253e-06, + "loss": 3.1392, + "step": 13598 + }, + { + "epoch": 1.090196628650458, + "grad_norm": 1.6477642059326172, + "learning_rate": 9.024472527876908e-06, + "loss": 3.1205, + "step": 13599 + }, + { + "epoch": 1.0902768034314807, + "grad_norm": 1.8734934329986572, + "learning_rate": 9.023180282177612e-06, + "loss": 3.2057, + "step": 13600 + }, + { + "epoch": 1.0903569782125033, + "grad_norm": 1.8439620733261108, + "learning_rate": 9.021888052947152e-06, + "loss": 3.2301, + "step": 13601 + }, + { + "epoch": 1.090437152993526, + "grad_norm": 1.7636098861694336, + "learning_rate": 9.020595840207313e-06, + "loss": 3.2526, + "step": 13602 + }, + { + "epoch": 1.0905173277745486, + "grad_norm": 0.707268476486206, + "learning_rate": 9.019303643979887e-06, + "loss": 2.4774, + "step": 13603 + }, + { + "epoch": 1.0905975025555712, + "grad_norm": 1.8646552562713623, + "learning_rate": 9.018011464286652e-06, + "loss": 3.2267, + "step": 13604 + }, + { + "epoch": 1.0906776773365938, + "grad_norm": 1.7587840557098389, + "learning_rate": 9.0167193011494e-06, + "loss": 3.318, + "step": 13605 + }, + { + "epoch": 1.0907578521176164, + "grad_norm": 1.8617875576019287, + "learning_rate": 9.015427154589912e-06, + "loss": 3.2261, + "step": 13606 + }, + { + "epoch": 1.090838026898639, + "grad_norm": 1.9225589036941528, + "learning_rate": 9.014135024629975e-06, + "loss": 3.1769, + "step": 13607 + }, + { + "epoch": 1.0909182016796617, + "grad_norm": 1.7070341110229492, + "learning_rate": 9.012842911291377e-06, + "loss": 3.158, + "step": 13608 + }, + { + "epoch": 1.0909983764606843, + "grad_norm": 1.7739503383636475, + "learning_rate": 9.011550814595897e-06, + "loss": 3.1471, + "step": 13609 + }, + { + "epoch": 1.091078551241707, + "grad_norm": 1.7894247770309448, + "learning_rate": 9.010258734565325e-06, + "loss": 3.1122, + "step": 13610 + }, + { + "epoch": 1.0911587260227296, + "grad_norm": 0.7158881425857544, + "learning_rate": 9.008966671221439e-06, + "loss": 2.4884, + "step": 13611 + }, + { + "epoch": 1.0912389008037522, + "grad_norm": 1.8207298517227173, + "learning_rate": 9.007674624586024e-06, + "loss": 3.1593, + "step": 13612 + }, + { + "epoch": 1.0913190755847748, + "grad_norm": 1.7980666160583496, + "learning_rate": 9.00638259468087e-06, + "loss": 3.1989, + "step": 13613 + }, + { + "epoch": 1.0913992503657974, + "grad_norm": 1.743593692779541, + "learning_rate": 9.005090581527752e-06, + "loss": 3.2157, + "step": 13614 + }, + { + "epoch": 1.09147942514682, + "grad_norm": 1.7472114562988281, + "learning_rate": 9.003798585148457e-06, + "loss": 3.1019, + "step": 13615 + }, + { + "epoch": 1.0915595999278427, + "grad_norm": 3.103492259979248, + "learning_rate": 9.002506605564768e-06, + "loss": 3.2119, + "step": 13616 + }, + { + "epoch": 1.0916397747088653, + "grad_norm": 1.8251513242721558, + "learning_rate": 9.001214642798463e-06, + "loss": 3.1337, + "step": 13617 + }, + { + "epoch": 1.091719949489888, + "grad_norm": 1.9357218742370605, + "learning_rate": 8.999922696871331e-06, + "loss": 3.1866, + "step": 13618 + }, + { + "epoch": 1.0918001242709106, + "grad_norm": 1.7268226146697998, + "learning_rate": 8.998630767805147e-06, + "loss": 3.2086, + "step": 13619 + }, + { + "epoch": 1.0918802990519332, + "grad_norm": 1.746710181236267, + "learning_rate": 8.997338855621697e-06, + "loss": 3.1504, + "step": 13620 + }, + { + "epoch": 1.0919604738329558, + "grad_norm": 1.7285654544830322, + "learning_rate": 8.996046960342758e-06, + "loss": 3.1426, + "step": 13621 + }, + { + "epoch": 1.0920406486139784, + "grad_norm": 1.8191494941711426, + "learning_rate": 8.994755081990114e-06, + "loss": 3.1958, + "step": 13622 + }, + { + "epoch": 1.092120823395001, + "grad_norm": 1.7784372568130493, + "learning_rate": 8.993463220585548e-06, + "loss": 3.0847, + "step": 13623 + }, + { + "epoch": 1.0922009981760237, + "grad_norm": 1.7037501335144043, + "learning_rate": 8.992171376150834e-06, + "loss": 3.1056, + "step": 13624 + }, + { + "epoch": 1.0922811729570463, + "grad_norm": 1.8028534650802612, + "learning_rate": 8.99087954870776e-06, + "loss": 3.2173, + "step": 13625 + }, + { + "epoch": 1.092361347738069, + "grad_norm": 1.8290212154388428, + "learning_rate": 8.989587738278097e-06, + "loss": 3.1771, + "step": 13626 + }, + { + "epoch": 1.0924415225190915, + "grad_norm": 1.9680571556091309, + "learning_rate": 8.988295944883628e-06, + "loss": 3.2692, + "step": 13627 + }, + { + "epoch": 1.0925216973001142, + "grad_norm": 1.8008332252502441, + "learning_rate": 8.987004168546136e-06, + "loss": 3.3054, + "step": 13628 + }, + { + "epoch": 1.0926018720811368, + "grad_norm": 1.8975768089294434, + "learning_rate": 8.985712409287393e-06, + "loss": 3.1843, + "step": 13629 + }, + { + "epoch": 1.0926820468621594, + "grad_norm": 1.7447435855865479, + "learning_rate": 8.984420667129185e-06, + "loss": 3.1906, + "step": 13630 + }, + { + "epoch": 1.092762221643182, + "grad_norm": 1.769925832748413, + "learning_rate": 8.983128942093285e-06, + "loss": 3.1665, + "step": 13631 + }, + { + "epoch": 1.0928423964242047, + "grad_norm": 1.7504440546035767, + "learning_rate": 8.981837234201471e-06, + "loss": 3.1208, + "step": 13632 + }, + { + "epoch": 1.0929225712052273, + "grad_norm": 1.730454444885254, + "learning_rate": 8.980545543475526e-06, + "loss": 3.2275, + "step": 13633 + }, + { + "epoch": 1.09300274598625, + "grad_norm": 1.8793468475341797, + "learning_rate": 8.979253869937222e-06, + "loss": 3.1694, + "step": 13634 + }, + { + "epoch": 1.0930829207672725, + "grad_norm": 1.787010669708252, + "learning_rate": 8.977962213608339e-06, + "loss": 3.0691, + "step": 13635 + }, + { + "epoch": 1.0931630955482954, + "grad_norm": 1.893538236618042, + "learning_rate": 8.97667057451065e-06, + "loss": 3.201, + "step": 13636 + }, + { + "epoch": 1.093243270329318, + "grad_norm": 1.8102092742919922, + "learning_rate": 8.975378952665937e-06, + "loss": 3.3044, + "step": 13637 + }, + { + "epoch": 1.0933234451103406, + "grad_norm": 1.852125883102417, + "learning_rate": 8.974087348095974e-06, + "loss": 3.3096, + "step": 13638 + }, + { + "epoch": 1.0934036198913633, + "grad_norm": 1.8497899770736694, + "learning_rate": 8.972795760822535e-06, + "loss": 3.2975, + "step": 13639 + }, + { + "epoch": 1.0934837946723859, + "grad_norm": 1.757462501525879, + "learning_rate": 8.971504190867399e-06, + "loss": 3.2123, + "step": 13640 + }, + { + "epoch": 1.0935639694534085, + "grad_norm": 1.7469632625579834, + "learning_rate": 8.970212638252338e-06, + "loss": 3.1257, + "step": 13641 + }, + { + "epoch": 1.0936441442344311, + "grad_norm": 1.8066197633743286, + "learning_rate": 8.968921102999129e-06, + "loss": 3.1699, + "step": 13642 + }, + { + "epoch": 1.0937243190154537, + "grad_norm": 1.9003100395202637, + "learning_rate": 8.96762958512955e-06, + "loss": 3.2157, + "step": 13643 + }, + { + "epoch": 1.0938044937964764, + "grad_norm": 0.6833188533782959, + "learning_rate": 8.96633808466537e-06, + "loss": 2.4657, + "step": 13644 + }, + { + "epoch": 1.093884668577499, + "grad_norm": 1.8654558658599854, + "learning_rate": 8.965046601628367e-06, + "loss": 3.1961, + "step": 13645 + }, + { + "epoch": 1.0939648433585216, + "grad_norm": 1.9103784561157227, + "learning_rate": 8.963755136040312e-06, + "loss": 3.2433, + "step": 13646 + }, + { + "epoch": 1.0940450181395442, + "grad_norm": 0.7105717658996582, + "learning_rate": 8.96246368792298e-06, + "loss": 2.4846, + "step": 13647 + }, + { + "epoch": 1.0941251929205669, + "grad_norm": 1.9258112907409668, + "learning_rate": 8.961172257298147e-06, + "loss": 3.3969, + "step": 13648 + }, + { + "epoch": 1.0942053677015895, + "grad_norm": 1.714090347290039, + "learning_rate": 8.95988084418758e-06, + "loss": 3.1566, + "step": 13649 + }, + { + "epoch": 1.0942855424826121, + "grad_norm": 1.7495973110198975, + "learning_rate": 8.95858944861306e-06, + "loss": 3.3023, + "step": 13650 + }, + { + "epoch": 1.0943657172636347, + "grad_norm": 1.8241828680038452, + "learning_rate": 8.95729807059635e-06, + "loss": 3.1166, + "step": 13651 + }, + { + "epoch": 1.0944458920446574, + "grad_norm": 1.7544718980789185, + "learning_rate": 8.956006710159228e-06, + "loss": 3.1431, + "step": 13652 + }, + { + "epoch": 1.09452606682568, + "grad_norm": 1.726240873336792, + "learning_rate": 8.954715367323468e-06, + "loss": 3.1742, + "step": 13653 + }, + { + "epoch": 1.0946062416067026, + "grad_norm": 1.7822868824005127, + "learning_rate": 8.953424042110835e-06, + "loss": 3.145, + "step": 13654 + }, + { + "epoch": 1.0946864163877252, + "grad_norm": 1.8328578472137451, + "learning_rate": 8.952132734543107e-06, + "loss": 3.2313, + "step": 13655 + }, + { + "epoch": 1.0947665911687479, + "grad_norm": 1.7104902267456055, + "learning_rate": 8.95084144464205e-06, + "loss": 3.2068, + "step": 13656 + }, + { + "epoch": 1.0948467659497705, + "grad_norm": 1.808422327041626, + "learning_rate": 8.949550172429435e-06, + "loss": 3.1774, + "step": 13657 + }, + { + "epoch": 1.094926940730793, + "grad_norm": 1.8295384645462036, + "learning_rate": 8.948258917927037e-06, + "loss": 3.1785, + "step": 13658 + }, + { + "epoch": 1.0950071155118157, + "grad_norm": 1.805770754814148, + "learning_rate": 8.94696768115662e-06, + "loss": 3.2963, + "step": 13659 + }, + { + "epoch": 1.0950872902928384, + "grad_norm": 1.7073566913604736, + "learning_rate": 8.94567646213996e-06, + "loss": 3.2405, + "step": 13660 + }, + { + "epoch": 1.095167465073861, + "grad_norm": 1.8847672939300537, + "learning_rate": 8.94438526089882e-06, + "loss": 3.1552, + "step": 13661 + }, + { + "epoch": 1.0952476398548836, + "grad_norm": 1.8414926528930664, + "learning_rate": 8.943094077454974e-06, + "loss": 3.1901, + "step": 13662 + }, + { + "epoch": 1.0953278146359062, + "grad_norm": 1.8295507431030273, + "learning_rate": 8.94180291183019e-06, + "loss": 3.1341, + "step": 13663 + }, + { + "epoch": 1.0954079894169289, + "grad_norm": 1.7456040382385254, + "learning_rate": 8.940511764046235e-06, + "loss": 3.2235, + "step": 13664 + }, + { + "epoch": 1.0954881641979515, + "grad_norm": 1.7387701272964478, + "learning_rate": 8.93922063412488e-06, + "loss": 3.2172, + "step": 13665 + }, + { + "epoch": 1.095568338978974, + "grad_norm": 1.8173497915267944, + "learning_rate": 8.937929522087889e-06, + "loss": 3.1384, + "step": 13666 + }, + { + "epoch": 1.0956485137599967, + "grad_norm": 1.8534435033798218, + "learning_rate": 8.936638427957031e-06, + "loss": 3.1324, + "step": 13667 + }, + { + "epoch": 1.0957286885410193, + "grad_norm": 1.6956344842910767, + "learning_rate": 8.935347351754078e-06, + "loss": 3.2173, + "step": 13668 + }, + { + "epoch": 1.095808863322042, + "grad_norm": 1.7825886011123657, + "learning_rate": 8.934056293500792e-06, + "loss": 3.1975, + "step": 13669 + }, + { + "epoch": 1.0958890381030646, + "grad_norm": 0.6977701187133789, + "learning_rate": 8.93276525321894e-06, + "loss": 2.4917, + "step": 13670 + }, + { + "epoch": 1.0959692128840872, + "grad_norm": 1.7822492122650146, + "learning_rate": 8.93147423093029e-06, + "loss": 3.2007, + "step": 13671 + }, + { + "epoch": 1.0960493876651098, + "grad_norm": 1.7000882625579834, + "learning_rate": 8.930183226656608e-06, + "loss": 3.1455, + "step": 13672 + }, + { + "epoch": 1.0961295624461325, + "grad_norm": 1.6831814050674438, + "learning_rate": 8.928892240419662e-06, + "loss": 3.0457, + "step": 13673 + }, + { + "epoch": 1.0962097372271553, + "grad_norm": 1.8066697120666504, + "learning_rate": 8.927601272241212e-06, + "loss": 3.2489, + "step": 13674 + }, + { + "epoch": 1.096289912008178, + "grad_norm": 1.881090521812439, + "learning_rate": 8.92631032214303e-06, + "loss": 3.2112, + "step": 13675 + }, + { + "epoch": 1.0963700867892006, + "grad_norm": 1.8219658136367798, + "learning_rate": 8.925019390146874e-06, + "loss": 3.1593, + "step": 13676 + }, + { + "epoch": 1.0964502615702232, + "grad_norm": 1.738038420677185, + "learning_rate": 8.923728476274514e-06, + "loss": 3.1045, + "step": 13677 + }, + { + "epoch": 1.0965304363512458, + "grad_norm": 1.7682688236236572, + "learning_rate": 8.922437580547714e-06, + "loss": 3.2354, + "step": 13678 + }, + { + "epoch": 1.0966106111322684, + "grad_norm": 1.7341511249542236, + "learning_rate": 8.921146702988234e-06, + "loss": 3.1527, + "step": 13679 + }, + { + "epoch": 1.096690785913291, + "grad_norm": 1.9261833429336548, + "learning_rate": 8.919855843617842e-06, + "loss": 3.1416, + "step": 13680 + }, + { + "epoch": 1.0967709606943137, + "grad_norm": 1.7253741025924683, + "learning_rate": 8.918565002458299e-06, + "loss": 3.2673, + "step": 13681 + }, + { + "epoch": 1.0968511354753363, + "grad_norm": 1.9082919359207153, + "learning_rate": 8.917274179531368e-06, + "loss": 3.1542, + "step": 13682 + }, + { + "epoch": 1.096931310256359, + "grad_norm": 1.769060492515564, + "learning_rate": 8.915983374858816e-06, + "loss": 3.1278, + "step": 13683 + }, + { + "epoch": 1.0970114850373816, + "grad_norm": 1.7915189266204834, + "learning_rate": 8.914692588462399e-06, + "loss": 3.1585, + "step": 13684 + }, + { + "epoch": 1.0970916598184042, + "grad_norm": 1.8570648431777954, + "learning_rate": 8.913401820363884e-06, + "loss": 3.1863, + "step": 13685 + }, + { + "epoch": 1.0971718345994268, + "grad_norm": 0.6822360754013062, + "learning_rate": 8.91211107058503e-06, + "loss": 2.4686, + "step": 13686 + }, + { + "epoch": 1.0972520093804494, + "grad_norm": 1.645209789276123, + "learning_rate": 8.9108203391476e-06, + "loss": 3.1237, + "step": 13687 + }, + { + "epoch": 1.097332184161472, + "grad_norm": 1.8211164474487305, + "learning_rate": 8.909529626073357e-06, + "loss": 3.1533, + "step": 13688 + }, + { + "epoch": 1.0974123589424947, + "grad_norm": 0.718407392501831, + "learning_rate": 8.908238931384059e-06, + "loss": 2.4877, + "step": 13689 + }, + { + "epoch": 1.0974925337235173, + "grad_norm": 1.8631560802459717, + "learning_rate": 8.906948255101469e-06, + "loss": 3.2889, + "step": 13690 + }, + { + "epoch": 1.09757270850454, + "grad_norm": 0.6865147352218628, + "learning_rate": 8.905657597247345e-06, + "loss": 2.4833, + "step": 13691 + }, + { + "epoch": 1.0976528832855625, + "grad_norm": 1.7755804061889648, + "learning_rate": 8.904366957843448e-06, + "loss": 3.1743, + "step": 13692 + }, + { + "epoch": 1.0977330580665852, + "grad_norm": 1.7943902015686035, + "learning_rate": 8.90307633691154e-06, + "loss": 3.1523, + "step": 13693 + }, + { + "epoch": 1.0978132328476078, + "grad_norm": 1.8102179765701294, + "learning_rate": 8.901785734473376e-06, + "loss": 3.2612, + "step": 13694 + }, + { + "epoch": 1.0978934076286304, + "grad_norm": 0.6959167122840881, + "learning_rate": 8.900495150550718e-06, + "loss": 2.458, + "step": 13695 + }, + { + "epoch": 1.097973582409653, + "grad_norm": 1.6788140535354614, + "learning_rate": 8.899204585165326e-06, + "loss": 3.1127, + "step": 13696 + }, + { + "epoch": 1.0980537571906757, + "grad_norm": 1.6751558780670166, + "learning_rate": 8.897914038338952e-06, + "loss": 3.2839, + "step": 13697 + }, + { + "epoch": 1.0981339319716983, + "grad_norm": 1.8596428632736206, + "learning_rate": 8.896623510093365e-06, + "loss": 3.2552, + "step": 13698 + }, + { + "epoch": 1.098214106752721, + "grad_norm": 0.7053965926170349, + "learning_rate": 8.895333000450313e-06, + "loss": 2.4745, + "step": 13699 + }, + { + "epoch": 1.0982942815337435, + "grad_norm": 1.7920265197753906, + "learning_rate": 8.894042509431559e-06, + "loss": 3.1879, + "step": 13700 + }, + { + "epoch": 1.0983744563147662, + "grad_norm": 1.7168256044387817, + "learning_rate": 8.892752037058858e-06, + "loss": 3.1707, + "step": 13701 + }, + { + "epoch": 1.0984546310957888, + "grad_norm": 1.9195321798324585, + "learning_rate": 8.891461583353967e-06, + "loss": 3.1819, + "step": 13702 + }, + { + "epoch": 1.0985348058768114, + "grad_norm": 1.8147200345993042, + "learning_rate": 8.890171148338643e-06, + "loss": 3.1636, + "step": 13703 + }, + { + "epoch": 1.098614980657834, + "grad_norm": 1.8672088384628296, + "learning_rate": 8.88888073203464e-06, + "loss": 3.232, + "step": 13704 + }, + { + "epoch": 1.0986951554388567, + "grad_norm": 1.8130850791931152, + "learning_rate": 8.887590334463723e-06, + "loss": 3.1714, + "step": 13705 + }, + { + "epoch": 1.0987753302198793, + "grad_norm": 1.8719075918197632, + "learning_rate": 8.886299955647635e-06, + "loss": 3.2518, + "step": 13706 + }, + { + "epoch": 1.098855505000902, + "grad_norm": 1.781306505203247, + "learning_rate": 8.885009595608139e-06, + "loss": 3.1154, + "step": 13707 + }, + { + "epoch": 1.0989356797819245, + "grad_norm": 1.8030816316604614, + "learning_rate": 8.883719254366989e-06, + "loss": 3.1982, + "step": 13708 + }, + { + "epoch": 1.0990158545629471, + "grad_norm": 1.8228737115859985, + "learning_rate": 8.882428931945937e-06, + "loss": 3.2899, + "step": 13709 + }, + { + "epoch": 1.0990960293439698, + "grad_norm": 1.7766833305358887, + "learning_rate": 8.881138628366741e-06, + "loss": 3.1362, + "step": 13710 + }, + { + "epoch": 1.0991762041249924, + "grad_norm": 1.689712405204773, + "learning_rate": 8.879848343651152e-06, + "loss": 3.1976, + "step": 13711 + }, + { + "epoch": 1.099256378906015, + "grad_norm": 1.7413654327392578, + "learning_rate": 8.878558077820925e-06, + "loss": 3.0882, + "step": 13712 + }, + { + "epoch": 1.0993365536870376, + "grad_norm": 1.7191450595855713, + "learning_rate": 8.877267830897817e-06, + "loss": 3.2706, + "step": 13713 + }, + { + "epoch": 1.0994167284680603, + "grad_norm": 1.693852424621582, + "learning_rate": 8.875977602903574e-06, + "loss": 3.2615, + "step": 13714 + }, + { + "epoch": 1.099496903249083, + "grad_norm": 1.8160486221313477, + "learning_rate": 8.874687393859956e-06, + "loss": 3.188, + "step": 13715 + }, + { + "epoch": 1.0995770780301055, + "grad_norm": 1.876349687576294, + "learning_rate": 8.873397203788708e-06, + "loss": 3.1176, + "step": 13716 + }, + { + "epoch": 1.0996572528111284, + "grad_norm": 1.729657769203186, + "learning_rate": 8.872107032711586e-06, + "loss": 3.1538, + "step": 13717 + }, + { + "epoch": 1.099737427592151, + "grad_norm": 1.844961166381836, + "learning_rate": 8.870816880650345e-06, + "loss": 3.277, + "step": 13718 + }, + { + "epoch": 1.0998176023731736, + "grad_norm": 1.7841945886611938, + "learning_rate": 8.869526747626732e-06, + "loss": 3.2797, + "step": 13719 + }, + { + "epoch": 1.0998977771541962, + "grad_norm": 1.6943469047546387, + "learning_rate": 8.868236633662501e-06, + "loss": 3.2561, + "step": 13720 + }, + { + "epoch": 1.0999779519352189, + "grad_norm": 1.8870642185211182, + "learning_rate": 8.8669465387794e-06, + "loss": 3.2203, + "step": 13721 + }, + { + "epoch": 1.1000581267162415, + "grad_norm": 0.6999526619911194, + "learning_rate": 8.86565646299918e-06, + "loss": 2.4922, + "step": 13722 + }, + { + "epoch": 1.100138301497264, + "grad_norm": 1.713034987449646, + "learning_rate": 8.864366406343595e-06, + "loss": 3.221, + "step": 13723 + }, + { + "epoch": 1.1002184762782867, + "grad_norm": 1.7340203523635864, + "learning_rate": 8.86307636883439e-06, + "loss": 3.1747, + "step": 13724 + }, + { + "epoch": 1.1002986510593094, + "grad_norm": 1.8357537984848022, + "learning_rate": 8.861786350493316e-06, + "loss": 3.1662, + "step": 13725 + }, + { + "epoch": 1.100378825840332, + "grad_norm": 1.6778333187103271, + "learning_rate": 8.860496351342126e-06, + "loss": 3.0154, + "step": 13726 + }, + { + "epoch": 1.1004590006213546, + "grad_norm": 1.7089451551437378, + "learning_rate": 8.859206371402563e-06, + "loss": 3.1282, + "step": 13727 + }, + { + "epoch": 1.1005391754023772, + "grad_norm": 2.0637452602386475, + "learning_rate": 8.857916410696382e-06, + "loss": 3.3433, + "step": 13728 + }, + { + "epoch": 1.1006193501833998, + "grad_norm": 1.791580319404602, + "learning_rate": 8.856626469245325e-06, + "loss": 3.2632, + "step": 13729 + }, + { + "epoch": 1.1006995249644225, + "grad_norm": 1.7687368392944336, + "learning_rate": 8.855336547071142e-06, + "loss": 3.2111, + "step": 13730 + }, + { + "epoch": 1.100779699745445, + "grad_norm": 1.7888842821121216, + "learning_rate": 8.854046644195585e-06, + "loss": 3.1995, + "step": 13731 + }, + { + "epoch": 1.1008598745264677, + "grad_norm": 1.8390575647354126, + "learning_rate": 8.852756760640393e-06, + "loss": 3.2549, + "step": 13732 + }, + { + "epoch": 1.1009400493074903, + "grad_norm": 1.8731855154037476, + "learning_rate": 8.851466896427321e-06, + "loss": 3.2136, + "step": 13733 + }, + { + "epoch": 1.101020224088513, + "grad_norm": 1.7466590404510498, + "learning_rate": 8.850177051578112e-06, + "loss": 3.2162, + "step": 13734 + }, + { + "epoch": 1.1011003988695356, + "grad_norm": 1.951692819595337, + "learning_rate": 8.84888722611451e-06, + "loss": 3.1498, + "step": 13735 + }, + { + "epoch": 1.1011805736505582, + "grad_norm": 1.8359670639038086, + "learning_rate": 8.847597420058268e-06, + "loss": 3.2346, + "step": 13736 + }, + { + "epoch": 1.1012607484315808, + "grad_norm": 1.6921478509902954, + "learning_rate": 8.846307633431124e-06, + "loss": 3.1401, + "step": 13737 + }, + { + "epoch": 1.1013409232126035, + "grad_norm": 1.779658317565918, + "learning_rate": 8.845017866254831e-06, + "loss": 3.178, + "step": 13738 + }, + { + "epoch": 1.101421097993626, + "grad_norm": 1.845450758934021, + "learning_rate": 8.843728118551124e-06, + "loss": 3.1646, + "step": 13739 + }, + { + "epoch": 1.1015012727746487, + "grad_norm": 1.7923957109451294, + "learning_rate": 8.842438390341757e-06, + "loss": 3.1835, + "step": 13740 + }, + { + "epoch": 1.1015814475556713, + "grad_norm": 1.8192795515060425, + "learning_rate": 8.84114868164847e-06, + "loss": 3.192, + "step": 13741 + }, + { + "epoch": 1.101661622336694, + "grad_norm": 1.9024081230163574, + "learning_rate": 8.839858992493007e-06, + "loss": 3.2433, + "step": 13742 + }, + { + "epoch": 1.1017417971177166, + "grad_norm": 1.7501065731048584, + "learning_rate": 8.838569322897115e-06, + "loss": 3.2223, + "step": 13743 + }, + { + "epoch": 1.1018219718987392, + "grad_norm": 1.7632672786712646, + "learning_rate": 8.837279672882531e-06, + "loss": 3.2695, + "step": 13744 + }, + { + "epoch": 1.1019021466797618, + "grad_norm": 1.815469741821289, + "learning_rate": 8.835990042471004e-06, + "loss": 3.2555, + "step": 13745 + }, + { + "epoch": 1.1019823214607845, + "grad_norm": 1.7102407217025757, + "learning_rate": 8.834700431684275e-06, + "loss": 3.1796, + "step": 13746 + }, + { + "epoch": 1.102062496241807, + "grad_norm": 1.8357411623001099, + "learning_rate": 8.833410840544085e-06, + "loss": 3.1996, + "step": 13747 + }, + { + "epoch": 1.1021426710228297, + "grad_norm": 1.7253737449645996, + "learning_rate": 8.83212126907218e-06, + "loss": 3.1661, + "step": 13748 + }, + { + "epoch": 1.1022228458038523, + "grad_norm": 1.8005956411361694, + "learning_rate": 8.830831717290295e-06, + "loss": 3.1086, + "step": 13749 + }, + { + "epoch": 1.102303020584875, + "grad_norm": 1.9178553819656372, + "learning_rate": 8.829542185220177e-06, + "loss": 3.2033, + "step": 13750 + }, + { + "epoch": 1.1023831953658976, + "grad_norm": 1.7385590076446533, + "learning_rate": 8.828252672883566e-06, + "loss": 3.2086, + "step": 13751 + }, + { + "epoch": 1.1024633701469202, + "grad_norm": 1.763909935951233, + "learning_rate": 8.826963180302199e-06, + "loss": 3.2311, + "step": 13752 + }, + { + "epoch": 1.1025435449279428, + "grad_norm": 1.7930916547775269, + "learning_rate": 8.825673707497824e-06, + "loss": 3.2887, + "step": 13753 + }, + { + "epoch": 1.1026237197089654, + "grad_norm": 1.7199965715408325, + "learning_rate": 8.824384254492171e-06, + "loss": 3.2593, + "step": 13754 + }, + { + "epoch": 1.1027038944899883, + "grad_norm": 3.354304313659668, + "learning_rate": 8.823094821306988e-06, + "loss": 3.1053, + "step": 13755 + }, + { + "epoch": 1.102784069271011, + "grad_norm": 1.8271409273147583, + "learning_rate": 8.821805407964013e-06, + "loss": 3.105, + "step": 13756 + }, + { + "epoch": 1.1028642440520335, + "grad_norm": 1.762630820274353, + "learning_rate": 8.820516014484982e-06, + "loss": 3.162, + "step": 13757 + }, + { + "epoch": 1.1029444188330562, + "grad_norm": 1.7359696626663208, + "learning_rate": 8.819226640891636e-06, + "loss": 3.2042, + "step": 13758 + }, + { + "epoch": 1.1030245936140788, + "grad_norm": 1.7491767406463623, + "learning_rate": 8.817937287205711e-06, + "loss": 3.1384, + "step": 13759 + }, + { + "epoch": 1.1031047683951014, + "grad_norm": 1.7619637250900269, + "learning_rate": 8.816647953448947e-06, + "loss": 3.2115, + "step": 13760 + }, + { + "epoch": 1.103184943176124, + "grad_norm": 1.6788837909698486, + "learning_rate": 8.815358639643083e-06, + "loss": 3.1534, + "step": 13761 + }, + { + "epoch": 1.1032651179571467, + "grad_norm": 1.8600674867630005, + "learning_rate": 8.814069345809853e-06, + "loss": 3.1264, + "step": 13762 + }, + { + "epoch": 1.1033452927381693, + "grad_norm": 1.7996023893356323, + "learning_rate": 8.812780071970999e-06, + "loss": 3.3177, + "step": 13763 + }, + { + "epoch": 1.103425467519192, + "grad_norm": 1.8362536430358887, + "learning_rate": 8.811490818148251e-06, + "loss": 3.15, + "step": 13764 + }, + { + "epoch": 1.1035056423002145, + "grad_norm": 1.6823856830596924, + "learning_rate": 8.81020158436335e-06, + "loss": 3.3135, + "step": 13765 + }, + { + "epoch": 1.1035858170812372, + "grad_norm": 1.7333869934082031, + "learning_rate": 8.808912370638032e-06, + "loss": 3.1381, + "step": 13766 + }, + { + "epoch": 1.1036659918622598, + "grad_norm": 1.8727585077285767, + "learning_rate": 8.80762317699403e-06, + "loss": 3.1701, + "step": 13767 + }, + { + "epoch": 1.1037461666432824, + "grad_norm": 1.8859540224075317, + "learning_rate": 8.806334003453084e-06, + "loss": 3.2129, + "step": 13768 + }, + { + "epoch": 1.103826341424305, + "grad_norm": 1.6967005729675293, + "learning_rate": 8.805044850036924e-06, + "loss": 3.2362, + "step": 13769 + }, + { + "epoch": 1.1039065162053276, + "grad_norm": 1.8443137407302856, + "learning_rate": 8.803755716767285e-06, + "loss": 3.0906, + "step": 13770 + }, + { + "epoch": 1.1039866909863503, + "grad_norm": 1.7565898895263672, + "learning_rate": 8.802466603665905e-06, + "loss": 3.2033, + "step": 13771 + }, + { + "epoch": 1.104066865767373, + "grad_norm": 1.771102786064148, + "learning_rate": 8.801177510754516e-06, + "loss": 3.1685, + "step": 13772 + }, + { + "epoch": 1.1041470405483955, + "grad_norm": 1.7815886735916138, + "learning_rate": 8.799888438054851e-06, + "loss": 3.2179, + "step": 13773 + }, + { + "epoch": 1.1042272153294181, + "grad_norm": 1.676356554031372, + "learning_rate": 8.798599385588645e-06, + "loss": 3.113, + "step": 13774 + }, + { + "epoch": 1.1043073901104408, + "grad_norm": 1.8138728141784668, + "learning_rate": 8.797310353377628e-06, + "loss": 3.1927, + "step": 13775 + }, + { + "epoch": 1.1043875648914634, + "grad_norm": 1.7280123233795166, + "learning_rate": 8.796021341443537e-06, + "loss": 3.2241, + "step": 13776 + }, + { + "epoch": 1.104467739672486, + "grad_norm": 1.6392052173614502, + "learning_rate": 8.7947323498081e-06, + "loss": 3.1816, + "step": 13777 + }, + { + "epoch": 1.1045479144535086, + "grad_norm": 1.9006494283676147, + "learning_rate": 8.793443378493054e-06, + "loss": 3.2172, + "step": 13778 + }, + { + "epoch": 1.1046280892345313, + "grad_norm": 1.7110052108764648, + "learning_rate": 8.792154427520123e-06, + "loss": 3.1307, + "step": 13779 + }, + { + "epoch": 1.1047082640155539, + "grad_norm": 1.739441156387329, + "learning_rate": 8.790865496911043e-06, + "loss": 3.2933, + "step": 13780 + }, + { + "epoch": 1.1047884387965765, + "grad_norm": 0.7208360433578491, + "learning_rate": 8.78957658668755e-06, + "loss": 2.4751, + "step": 13781 + }, + { + "epoch": 1.1048686135775991, + "grad_norm": 1.6603333950042725, + "learning_rate": 8.788287696871363e-06, + "loss": 3.2004, + "step": 13782 + }, + { + "epoch": 1.1049487883586218, + "grad_norm": 1.7818318605422974, + "learning_rate": 8.786998827484222e-06, + "loss": 3.2204, + "step": 13783 + }, + { + "epoch": 1.1050289631396444, + "grad_norm": 1.7434128522872925, + "learning_rate": 8.785709978547852e-06, + "loss": 3.2094, + "step": 13784 + }, + { + "epoch": 1.105109137920667, + "grad_norm": 1.7888362407684326, + "learning_rate": 8.784421150083983e-06, + "loss": 3.3063, + "step": 13785 + }, + { + "epoch": 1.1051893127016896, + "grad_norm": 1.7516216039657593, + "learning_rate": 8.783132342114347e-06, + "loss": 3.2003, + "step": 13786 + }, + { + "epoch": 1.1052694874827123, + "grad_norm": 1.7674939632415771, + "learning_rate": 8.781843554660668e-06, + "loss": 3.2319, + "step": 13787 + }, + { + "epoch": 1.1053496622637349, + "grad_norm": 1.8521479368209839, + "learning_rate": 8.78055478774468e-06, + "loss": 3.1804, + "step": 13788 + }, + { + "epoch": 1.1054298370447575, + "grad_norm": 1.7817683219909668, + "learning_rate": 8.779266041388106e-06, + "loss": 3.3179, + "step": 13789 + }, + { + "epoch": 1.1055100118257801, + "grad_norm": 1.9219433069229126, + "learning_rate": 8.777977315612677e-06, + "loss": 3.208, + "step": 13790 + }, + { + "epoch": 1.1055901866068027, + "grad_norm": 1.7768590450286865, + "learning_rate": 8.776688610440124e-06, + "loss": 3.1768, + "step": 13791 + }, + { + "epoch": 1.1056703613878254, + "grad_norm": 1.8031508922576904, + "learning_rate": 8.775399925892165e-06, + "loss": 3.1862, + "step": 13792 + }, + { + "epoch": 1.105750536168848, + "grad_norm": 1.8124547004699707, + "learning_rate": 8.774111261990534e-06, + "loss": 3.1664, + "step": 13793 + }, + { + "epoch": 1.1058307109498706, + "grad_norm": 2.0613503456115723, + "learning_rate": 8.772822618756953e-06, + "loss": 3.1318, + "step": 13794 + }, + { + "epoch": 1.1059108857308932, + "grad_norm": 0.7440069913864136, + "learning_rate": 8.771533996213151e-06, + "loss": 2.4661, + "step": 13795 + }, + { + "epoch": 1.1059910605119159, + "grad_norm": 1.7774484157562256, + "learning_rate": 8.770245394380854e-06, + "loss": 3.1941, + "step": 13796 + }, + { + "epoch": 1.1060712352929385, + "grad_norm": 1.7442244291305542, + "learning_rate": 8.768956813281784e-06, + "loss": 3.2079, + "step": 13797 + }, + { + "epoch": 1.1061514100739613, + "grad_norm": 1.737817406654358, + "learning_rate": 8.767668252937672e-06, + "loss": 3.1943, + "step": 13798 + }, + { + "epoch": 1.106231584854984, + "grad_norm": 1.6542212963104248, + "learning_rate": 8.766379713370237e-06, + "loss": 3.2154, + "step": 13799 + }, + { + "epoch": 1.1063117596360066, + "grad_norm": 1.9245002269744873, + "learning_rate": 8.765091194601202e-06, + "loss": 3.1566, + "step": 13800 + }, + { + "epoch": 1.1063919344170292, + "grad_norm": 1.7803122997283936, + "learning_rate": 8.7638026966523e-06, + "loss": 3.0789, + "step": 13801 + }, + { + "epoch": 1.1064721091980518, + "grad_norm": 0.6907920837402344, + "learning_rate": 8.762514219545246e-06, + "loss": 2.4786, + "step": 13802 + }, + { + "epoch": 1.1065522839790745, + "grad_norm": 1.8140780925750732, + "learning_rate": 8.761225763301765e-06, + "loss": 3.186, + "step": 13803 + }, + { + "epoch": 1.106632458760097, + "grad_norm": 1.7272205352783203, + "learning_rate": 8.759937327943583e-06, + "loss": 3.2003, + "step": 13804 + }, + { + "epoch": 1.1067126335411197, + "grad_norm": 1.6886537075042725, + "learning_rate": 8.758648913492419e-06, + "loss": 3.1572, + "step": 13805 + }, + { + "epoch": 1.1067928083221423, + "grad_norm": 1.7970309257507324, + "learning_rate": 8.757360519969998e-06, + "loss": 3.2635, + "step": 13806 + }, + { + "epoch": 1.106872983103165, + "grad_norm": 1.819384217262268, + "learning_rate": 8.75607214739804e-06, + "loss": 3.1367, + "step": 13807 + }, + { + "epoch": 1.1069531578841876, + "grad_norm": 1.864802360534668, + "learning_rate": 8.75478379579827e-06, + "loss": 3.1799, + "step": 13808 + }, + { + "epoch": 1.1070333326652102, + "grad_norm": 1.7015961408615112, + "learning_rate": 8.753495465192402e-06, + "loss": 3.0901, + "step": 13809 + }, + { + "epoch": 1.1071135074462328, + "grad_norm": 1.8096282482147217, + "learning_rate": 8.752207155602163e-06, + "loss": 3.1867, + "step": 13810 + }, + { + "epoch": 1.1071936822272554, + "grad_norm": 1.7555674314498901, + "learning_rate": 8.750918867049271e-06, + "loss": 3.1654, + "step": 13811 + }, + { + "epoch": 1.107273857008278, + "grad_norm": 1.7839053869247437, + "learning_rate": 8.749630599555448e-06, + "loss": 3.153, + "step": 13812 + }, + { + "epoch": 1.1073540317893007, + "grad_norm": 1.7843400239944458, + "learning_rate": 8.748342353142412e-06, + "loss": 3.082, + "step": 13813 + }, + { + "epoch": 1.1074342065703233, + "grad_norm": 0.7033902406692505, + "learning_rate": 8.747054127831882e-06, + "loss": 2.4805, + "step": 13814 + }, + { + "epoch": 1.107514381351346, + "grad_norm": 1.7833287715911865, + "learning_rate": 8.745765923645578e-06, + "loss": 3.1593, + "step": 13815 + }, + { + "epoch": 1.1075945561323686, + "grad_norm": 1.7439703941345215, + "learning_rate": 8.74447774060522e-06, + "loss": 3.1862, + "step": 13816 + }, + { + "epoch": 1.1076747309133912, + "grad_norm": 1.7446128129959106, + "learning_rate": 8.743189578732523e-06, + "loss": 3.2148, + "step": 13817 + }, + { + "epoch": 1.1077549056944138, + "grad_norm": 1.7346582412719727, + "learning_rate": 8.741901438049209e-06, + "loss": 3.3483, + "step": 13818 + }, + { + "epoch": 1.1078350804754364, + "grad_norm": 1.791758418083191, + "learning_rate": 8.74061331857699e-06, + "loss": 3.1489, + "step": 13819 + }, + { + "epoch": 1.107915255256459, + "grad_norm": 1.9147977828979492, + "learning_rate": 8.739325220337587e-06, + "loss": 3.3514, + "step": 13820 + }, + { + "epoch": 1.1079954300374817, + "grad_norm": 1.8642717599868774, + "learning_rate": 8.73803714335272e-06, + "loss": 3.1326, + "step": 13821 + }, + { + "epoch": 1.1080756048185043, + "grad_norm": 0.7550550699234009, + "learning_rate": 8.736749087644098e-06, + "loss": 2.481, + "step": 13822 + }, + { + "epoch": 1.108155779599527, + "grad_norm": 1.7026400566101074, + "learning_rate": 8.735461053233445e-06, + "loss": 3.1669, + "step": 13823 + }, + { + "epoch": 1.1082359543805496, + "grad_norm": 1.801632046699524, + "learning_rate": 8.73417304014247e-06, + "loss": 3.1427, + "step": 13824 + }, + { + "epoch": 1.1083161291615722, + "grad_norm": 1.6890183687210083, + "learning_rate": 8.732885048392891e-06, + "loss": 3.184, + "step": 13825 + }, + { + "epoch": 1.1083963039425948, + "grad_norm": 1.8189241886138916, + "learning_rate": 8.731597078006427e-06, + "loss": 3.2892, + "step": 13826 + }, + { + "epoch": 1.1084764787236174, + "grad_norm": 1.8247219324111938, + "learning_rate": 8.730309129004786e-06, + "loss": 3.1599, + "step": 13827 + }, + { + "epoch": 1.10855665350464, + "grad_norm": 1.6944783926010132, + "learning_rate": 8.72902120140969e-06, + "loss": 3.185, + "step": 13828 + }, + { + "epoch": 1.1086368282856627, + "grad_norm": 1.7116551399230957, + "learning_rate": 8.727733295242842e-06, + "loss": 3.269, + "step": 13829 + }, + { + "epoch": 1.1087170030666853, + "grad_norm": 1.7274630069732666, + "learning_rate": 8.726445410525966e-06, + "loss": 3.2119, + "step": 13830 + }, + { + "epoch": 1.108797177847708, + "grad_norm": 1.7820018529891968, + "learning_rate": 8.725157547280772e-06, + "loss": 3.1674, + "step": 13831 + }, + { + "epoch": 1.1088773526287306, + "grad_norm": 1.8252193927764893, + "learning_rate": 8.723869705528971e-06, + "loss": 3.2077, + "step": 13832 + }, + { + "epoch": 1.1089575274097532, + "grad_norm": 1.777978539466858, + "learning_rate": 8.722581885292278e-06, + "loss": 3.2264, + "step": 13833 + }, + { + "epoch": 1.1090377021907758, + "grad_norm": 1.9524201154708862, + "learning_rate": 8.721294086592403e-06, + "loss": 3.2566, + "step": 13834 + }, + { + "epoch": 1.1091178769717984, + "grad_norm": 1.7456706762313843, + "learning_rate": 8.720006309451059e-06, + "loss": 3.1841, + "step": 13835 + }, + { + "epoch": 1.1091980517528213, + "grad_norm": 1.6879833936691284, + "learning_rate": 8.71871855388996e-06, + "loss": 3.1837, + "step": 13836 + }, + { + "epoch": 1.109278226533844, + "grad_norm": 1.7832926511764526, + "learning_rate": 8.717430819930813e-06, + "loss": 3.1111, + "step": 13837 + }, + { + "epoch": 1.1093584013148665, + "grad_norm": 1.8057805299758911, + "learning_rate": 8.716143107595331e-06, + "loss": 3.2833, + "step": 13838 + }, + { + "epoch": 1.1094385760958891, + "grad_norm": 1.816840648651123, + "learning_rate": 8.714855416905223e-06, + "loss": 3.2504, + "step": 13839 + }, + { + "epoch": 1.1095187508769118, + "grad_norm": 0.7434176802635193, + "learning_rate": 8.713567747882198e-06, + "loss": 2.4774, + "step": 13840 + }, + { + "epoch": 1.1095989256579344, + "grad_norm": 1.7386202812194824, + "learning_rate": 8.71228010054797e-06, + "loss": 3.1445, + "step": 13841 + }, + { + "epoch": 1.109679100438957, + "grad_norm": 1.7642722129821777, + "learning_rate": 8.710992474924245e-06, + "loss": 3.1149, + "step": 13842 + }, + { + "epoch": 1.1097592752199796, + "grad_norm": 1.7619017362594604, + "learning_rate": 8.709704871032733e-06, + "loss": 3.1583, + "step": 13843 + }, + { + "epoch": 1.1098394500010023, + "grad_norm": 1.681395173072815, + "learning_rate": 8.708417288895141e-06, + "loss": 3.1506, + "step": 13844 + }, + { + "epoch": 1.1099196247820249, + "grad_norm": 0.6574442386627197, + "learning_rate": 8.707129728533178e-06, + "loss": 2.4578, + "step": 13845 + }, + { + "epoch": 1.1099997995630475, + "grad_norm": 1.7615019083023071, + "learning_rate": 8.705842189968554e-06, + "loss": 3.1837, + "step": 13846 + }, + { + "epoch": 1.1100799743440701, + "grad_norm": 1.738021969795227, + "learning_rate": 8.704554673222973e-06, + "loss": 3.1313, + "step": 13847 + }, + { + "epoch": 1.1101601491250928, + "grad_norm": 1.723865270614624, + "learning_rate": 8.703267178318145e-06, + "loss": 3.0987, + "step": 13848 + }, + { + "epoch": 1.1102403239061154, + "grad_norm": 0.7008875012397766, + "learning_rate": 8.701979705275775e-06, + "loss": 2.4786, + "step": 13849 + }, + { + "epoch": 1.110320498687138, + "grad_norm": 1.7516601085662842, + "learning_rate": 8.700692254117568e-06, + "loss": 3.1608, + "step": 13850 + }, + { + "epoch": 1.1104006734681606, + "grad_norm": 1.723752498626709, + "learning_rate": 8.699404824865234e-06, + "loss": 3.1926, + "step": 13851 + }, + { + "epoch": 1.1104808482491832, + "grad_norm": 1.7283151149749756, + "learning_rate": 8.698117417540475e-06, + "loss": 3.1458, + "step": 13852 + }, + { + "epoch": 1.1105610230302059, + "grad_norm": 1.7244441509246826, + "learning_rate": 8.696830032164999e-06, + "loss": 3.19, + "step": 13853 + }, + { + "epoch": 1.1106411978112285, + "grad_norm": 1.7600122690200806, + "learning_rate": 8.695542668760508e-06, + "loss": 3.1909, + "step": 13854 + }, + { + "epoch": 1.1107213725922511, + "grad_norm": 1.8276739120483398, + "learning_rate": 8.694255327348706e-06, + "loss": 3.1266, + "step": 13855 + }, + { + "epoch": 1.1108015473732737, + "grad_norm": 1.7561546564102173, + "learning_rate": 8.692968007951302e-06, + "loss": 3.1326, + "step": 13856 + }, + { + "epoch": 1.1108817221542964, + "grad_norm": 1.725127935409546, + "learning_rate": 8.691680710589996e-06, + "loss": 3.1146, + "step": 13857 + }, + { + "epoch": 1.110961896935319, + "grad_norm": 1.8123003244400024, + "learning_rate": 8.690393435286493e-06, + "loss": 3.1423, + "step": 13858 + }, + { + "epoch": 1.1110420717163416, + "grad_norm": 1.8234542608261108, + "learning_rate": 8.689106182062493e-06, + "loss": 3.0904, + "step": 13859 + }, + { + "epoch": 1.1111222464973642, + "grad_norm": 1.7371746301651, + "learning_rate": 8.6878189509397e-06, + "loss": 3.0983, + "step": 13860 + }, + { + "epoch": 1.1112024212783869, + "grad_norm": 0.6967493295669556, + "learning_rate": 8.686531741939822e-06, + "loss": 2.4855, + "step": 13861 + }, + { + "epoch": 1.1112825960594095, + "grad_norm": 1.9099063873291016, + "learning_rate": 8.685244555084551e-06, + "loss": 3.2883, + "step": 13862 + }, + { + "epoch": 1.1113627708404321, + "grad_norm": 1.726590871810913, + "learning_rate": 8.683957390395596e-06, + "loss": 3.207, + "step": 13863 + }, + { + "epoch": 1.1114429456214547, + "grad_norm": 1.947176456451416, + "learning_rate": 8.682670247894653e-06, + "loss": 3.287, + "step": 13864 + }, + { + "epoch": 1.1115231204024774, + "grad_norm": 1.8234493732452393, + "learning_rate": 8.681383127603426e-06, + "loss": 3.1658, + "step": 13865 + }, + { + "epoch": 1.1116032951835, + "grad_norm": 1.761555552482605, + "learning_rate": 8.680096029543619e-06, + "loss": 3.2522, + "step": 13866 + }, + { + "epoch": 1.1116834699645226, + "grad_norm": 1.985520362854004, + "learning_rate": 8.678808953736924e-06, + "loss": 3.2095, + "step": 13867 + }, + { + "epoch": 1.1117636447455452, + "grad_norm": 1.8142551183700562, + "learning_rate": 8.677521900205044e-06, + "loss": 3.1927, + "step": 13868 + }, + { + "epoch": 1.1118438195265679, + "grad_norm": 1.7612535953521729, + "learning_rate": 8.676234868969679e-06, + "loss": 3.2501, + "step": 13869 + }, + { + "epoch": 1.1119239943075905, + "grad_norm": 1.812324047088623, + "learning_rate": 8.674947860052529e-06, + "loss": 3.128, + "step": 13870 + }, + { + "epoch": 1.112004169088613, + "grad_norm": 1.8500847816467285, + "learning_rate": 8.673660873475292e-06, + "loss": 3.1829, + "step": 13871 + }, + { + "epoch": 1.1120843438696357, + "grad_norm": 1.732446312904358, + "learning_rate": 8.672373909259662e-06, + "loss": 3.2207, + "step": 13872 + }, + { + "epoch": 1.1121645186506584, + "grad_norm": 1.999912142753601, + "learning_rate": 8.671086967427343e-06, + "loss": 3.2341, + "step": 13873 + }, + { + "epoch": 1.112244693431681, + "grad_norm": 1.853090763092041, + "learning_rate": 8.669800048000027e-06, + "loss": 3.0452, + "step": 13874 + }, + { + "epoch": 1.1123248682127036, + "grad_norm": 1.753201961517334, + "learning_rate": 8.668513150999414e-06, + "loss": 3.1749, + "step": 13875 + }, + { + "epoch": 1.1124050429937262, + "grad_norm": 1.7823808193206787, + "learning_rate": 8.6672262764472e-06, + "loss": 3.1891, + "step": 13876 + }, + { + "epoch": 1.1124852177747488, + "grad_norm": 1.7544364929199219, + "learning_rate": 8.665939424365082e-06, + "loss": 3.1724, + "step": 13877 + }, + { + "epoch": 1.1125653925557715, + "grad_norm": 1.7281479835510254, + "learning_rate": 8.664652594774756e-06, + "loss": 3.1781, + "step": 13878 + }, + { + "epoch": 1.112645567336794, + "grad_norm": 1.8715109825134277, + "learning_rate": 8.663365787697913e-06, + "loss": 3.1066, + "step": 13879 + }, + { + "epoch": 1.112725742117817, + "grad_norm": 1.6844687461853027, + "learning_rate": 8.662079003156254e-06, + "loss": 3.1903, + "step": 13880 + }, + { + "epoch": 1.1128059168988396, + "grad_norm": 1.825583577156067, + "learning_rate": 8.660792241171473e-06, + "loss": 3.2526, + "step": 13881 + }, + { + "epoch": 1.1128860916798622, + "grad_norm": 1.726723551750183, + "learning_rate": 8.659505501765261e-06, + "loss": 3.1941, + "step": 13882 + }, + { + "epoch": 1.1129662664608848, + "grad_norm": 1.78196382522583, + "learning_rate": 8.658218784959314e-06, + "loss": 3.225, + "step": 13883 + }, + { + "epoch": 1.1130464412419074, + "grad_norm": 1.8745670318603516, + "learning_rate": 8.656932090775326e-06, + "loss": 3.1875, + "step": 13884 + }, + { + "epoch": 1.11312661602293, + "grad_norm": 1.8129079341888428, + "learning_rate": 8.655645419234988e-06, + "loss": 3.1143, + "step": 13885 + }, + { + "epoch": 1.1132067908039527, + "grad_norm": 1.997122883796692, + "learning_rate": 8.654358770359997e-06, + "loss": 3.1699, + "step": 13886 + }, + { + "epoch": 1.1132869655849753, + "grad_norm": 1.6836354732513428, + "learning_rate": 8.653072144172042e-06, + "loss": 3.1832, + "step": 13887 + }, + { + "epoch": 1.113367140365998, + "grad_norm": 1.7681821584701538, + "learning_rate": 8.651785540692816e-06, + "loss": 3.2335, + "step": 13888 + }, + { + "epoch": 1.1134473151470206, + "grad_norm": 1.8063987493515015, + "learning_rate": 8.650498959944009e-06, + "loss": 3.085, + "step": 13889 + }, + { + "epoch": 1.1135274899280432, + "grad_norm": 0.7051234245300293, + "learning_rate": 8.649212401947313e-06, + "loss": 2.4795, + "step": 13890 + }, + { + "epoch": 1.1136076647090658, + "grad_norm": 1.8017805814743042, + "learning_rate": 8.647925866724424e-06, + "loss": 3.1106, + "step": 13891 + }, + { + "epoch": 1.1136878394900884, + "grad_norm": 1.8733361959457397, + "learning_rate": 8.646639354297025e-06, + "loss": 3.2377, + "step": 13892 + }, + { + "epoch": 1.113768014271111, + "grad_norm": 1.8422080278396606, + "learning_rate": 8.645352864686812e-06, + "loss": 3.1441, + "step": 13893 + }, + { + "epoch": 1.1138481890521337, + "grad_norm": 1.7824112176895142, + "learning_rate": 8.64406639791547e-06, + "loss": 3.2007, + "step": 13894 + }, + { + "epoch": 1.1139283638331563, + "grad_norm": 1.7833530902862549, + "learning_rate": 8.642779954004691e-06, + "loss": 3.2767, + "step": 13895 + }, + { + "epoch": 1.114008538614179, + "grad_norm": 1.852990984916687, + "learning_rate": 8.641493532976166e-06, + "loss": 3.1967, + "step": 13896 + }, + { + "epoch": 1.1140887133952015, + "grad_norm": 1.7970600128173828, + "learning_rate": 8.640207134851579e-06, + "loss": 3.2321, + "step": 13897 + }, + { + "epoch": 1.1141688881762242, + "grad_norm": 0.7124155163764954, + "learning_rate": 8.638920759652623e-06, + "loss": 2.4657, + "step": 13898 + }, + { + "epoch": 1.1142490629572468, + "grad_norm": 1.7070033550262451, + "learning_rate": 8.63763440740098e-06, + "loss": 3.2032, + "step": 13899 + }, + { + "epoch": 1.1143292377382694, + "grad_norm": 1.8124339580535889, + "learning_rate": 8.63634807811834e-06, + "loss": 3.1472, + "step": 13900 + }, + { + "epoch": 1.114409412519292, + "grad_norm": 1.7270344495773315, + "learning_rate": 8.635061771826396e-06, + "loss": 3.2456, + "step": 13901 + }, + { + "epoch": 1.1144895873003147, + "grad_norm": 1.6702558994293213, + "learning_rate": 8.633775488546825e-06, + "loss": 3.2351, + "step": 13902 + }, + { + "epoch": 1.1145697620813373, + "grad_norm": 1.8838683366775513, + "learning_rate": 8.63248922830132e-06, + "loss": 3.1888, + "step": 13903 + }, + { + "epoch": 1.11464993686236, + "grad_norm": 1.8215588331222534, + "learning_rate": 8.631202991111563e-06, + "loss": 3.2938, + "step": 13904 + }, + { + "epoch": 1.1147301116433825, + "grad_norm": 1.7206302881240845, + "learning_rate": 8.62991677699924e-06, + "loss": 3.2175, + "step": 13905 + }, + { + "epoch": 1.1148102864244052, + "grad_norm": 1.650711178779602, + "learning_rate": 8.628630585986042e-06, + "loss": 3.0577, + "step": 13906 + }, + { + "epoch": 1.1148904612054278, + "grad_norm": 1.7590103149414062, + "learning_rate": 8.627344418093645e-06, + "loss": 3.2079, + "step": 13907 + }, + { + "epoch": 1.1149706359864504, + "grad_norm": 1.7460136413574219, + "learning_rate": 8.62605827334374e-06, + "loss": 3.2499, + "step": 13908 + }, + { + "epoch": 1.115050810767473, + "grad_norm": 1.8054946660995483, + "learning_rate": 8.624772151758005e-06, + "loss": 3.2445, + "step": 13909 + }, + { + "epoch": 1.1151309855484957, + "grad_norm": 1.770986795425415, + "learning_rate": 8.623486053358127e-06, + "loss": 3.1754, + "step": 13910 + }, + { + "epoch": 1.1152111603295183, + "grad_norm": 1.8380284309387207, + "learning_rate": 8.622199978165793e-06, + "loss": 3.1177, + "step": 13911 + }, + { + "epoch": 1.115291335110541, + "grad_norm": 1.781999111175537, + "learning_rate": 8.620913926202679e-06, + "loss": 3.1782, + "step": 13912 + }, + { + "epoch": 1.1153715098915635, + "grad_norm": 0.7269862294197083, + "learning_rate": 8.619627897490473e-06, + "loss": 2.483, + "step": 13913 + }, + { + "epoch": 1.1154516846725862, + "grad_norm": 1.8490748405456543, + "learning_rate": 8.61834189205085e-06, + "loss": 3.2082, + "step": 13914 + }, + { + "epoch": 1.1155318594536088, + "grad_norm": 1.7769222259521484, + "learning_rate": 8.617055909905497e-06, + "loss": 3.1942, + "step": 13915 + }, + { + "epoch": 1.1156120342346314, + "grad_norm": 1.7542757987976074, + "learning_rate": 8.615769951076095e-06, + "loss": 3.149, + "step": 13916 + }, + { + "epoch": 1.115692209015654, + "grad_norm": 1.7598216533660889, + "learning_rate": 8.614484015584324e-06, + "loss": 3.2644, + "step": 13917 + }, + { + "epoch": 1.1157723837966769, + "grad_norm": 1.7468745708465576, + "learning_rate": 8.613198103451861e-06, + "loss": 3.1976, + "step": 13918 + }, + { + "epoch": 1.1158525585776995, + "grad_norm": 1.861390471458435, + "learning_rate": 8.611912214700393e-06, + "loss": 3.2074, + "step": 13919 + }, + { + "epoch": 1.1159327333587221, + "grad_norm": 1.920710563659668, + "learning_rate": 8.610626349351595e-06, + "loss": 3.1497, + "step": 13920 + }, + { + "epoch": 1.1160129081397447, + "grad_norm": 1.734208106994629, + "learning_rate": 8.609340507427147e-06, + "loss": 3.1444, + "step": 13921 + }, + { + "epoch": 1.1160930829207674, + "grad_norm": 1.7053043842315674, + "learning_rate": 8.608054688948727e-06, + "loss": 3.2183, + "step": 13922 + }, + { + "epoch": 1.11617325770179, + "grad_norm": 1.8051851987838745, + "learning_rate": 8.606768893938013e-06, + "loss": 3.245, + "step": 13923 + }, + { + "epoch": 1.1162534324828126, + "grad_norm": 1.7889894247055054, + "learning_rate": 8.605483122416689e-06, + "loss": 3.1849, + "step": 13924 + }, + { + "epoch": 1.1163336072638352, + "grad_norm": 0.6880777478218079, + "learning_rate": 8.604197374406425e-06, + "loss": 2.4586, + "step": 13925 + }, + { + "epoch": 1.1164137820448579, + "grad_norm": 1.8444894552230835, + "learning_rate": 8.602911649928903e-06, + "loss": 3.2125, + "step": 13926 + }, + { + "epoch": 1.1164939568258805, + "grad_norm": 1.9476557970046997, + "learning_rate": 8.601625949005794e-06, + "loss": 3.2296, + "step": 13927 + }, + { + "epoch": 1.116574131606903, + "grad_norm": 1.9513392448425293, + "learning_rate": 8.600340271658783e-06, + "loss": 3.265, + "step": 13928 + }, + { + "epoch": 1.1166543063879257, + "grad_norm": 1.731115698814392, + "learning_rate": 8.599054617909542e-06, + "loss": 3.0686, + "step": 13929 + }, + { + "epoch": 1.1167344811689484, + "grad_norm": 1.7766449451446533, + "learning_rate": 8.597768987779743e-06, + "loss": 3.2242, + "step": 13930 + }, + { + "epoch": 1.116814655949971, + "grad_norm": 1.7213108539581299, + "learning_rate": 8.596483381291068e-06, + "loss": 3.185, + "step": 13931 + }, + { + "epoch": 1.1168948307309936, + "grad_norm": 1.6743627786636353, + "learning_rate": 8.595197798465186e-06, + "loss": 3.0945, + "step": 13932 + }, + { + "epoch": 1.1169750055120162, + "grad_norm": 1.750139594078064, + "learning_rate": 8.593912239323774e-06, + "loss": 3.2324, + "step": 13933 + }, + { + "epoch": 1.1170551802930389, + "grad_norm": 1.7834525108337402, + "learning_rate": 8.592626703888508e-06, + "loss": 3.1001, + "step": 13934 + }, + { + "epoch": 1.1171353550740615, + "grad_norm": 1.827636480331421, + "learning_rate": 8.591341192181058e-06, + "loss": 3.2073, + "step": 13935 + }, + { + "epoch": 1.117215529855084, + "grad_norm": 1.7703208923339844, + "learning_rate": 8.590055704223102e-06, + "loss": 3.1707, + "step": 13936 + }, + { + "epoch": 1.1172957046361067, + "grad_norm": 1.8110183477401733, + "learning_rate": 8.588770240036308e-06, + "loss": 3.2173, + "step": 13937 + }, + { + "epoch": 1.1173758794171293, + "grad_norm": 1.7887769937515259, + "learning_rate": 8.587484799642348e-06, + "loss": 3.2015, + "step": 13938 + }, + { + "epoch": 1.117456054198152, + "grad_norm": 1.8532962799072266, + "learning_rate": 8.5861993830629e-06, + "loss": 3.274, + "step": 13939 + }, + { + "epoch": 1.1175362289791746, + "grad_norm": 1.7921438217163086, + "learning_rate": 8.58491399031963e-06, + "loss": 3.2018, + "step": 13940 + }, + { + "epoch": 1.1176164037601972, + "grad_norm": 1.911414623260498, + "learning_rate": 8.583628621434214e-06, + "loss": 3.2096, + "step": 13941 + }, + { + "epoch": 1.1176965785412198, + "grad_norm": 0.6881842613220215, + "learning_rate": 8.582343276428318e-06, + "loss": 2.4805, + "step": 13942 + }, + { + "epoch": 1.1177767533222425, + "grad_norm": 1.8425114154815674, + "learning_rate": 8.581057955323615e-06, + "loss": 3.1168, + "step": 13943 + }, + { + "epoch": 1.117856928103265, + "grad_norm": 1.7463903427124023, + "learning_rate": 8.579772658141775e-06, + "loss": 3.2064, + "step": 13944 + }, + { + "epoch": 1.1179371028842877, + "grad_norm": 1.7262502908706665, + "learning_rate": 8.578487384904465e-06, + "loss": 3.304, + "step": 13945 + }, + { + "epoch": 1.1180172776653103, + "grad_norm": 1.7840465307235718, + "learning_rate": 8.577202135633362e-06, + "loss": 3.0388, + "step": 13946 + }, + { + "epoch": 1.118097452446333, + "grad_norm": 1.7164201736450195, + "learning_rate": 8.575916910350126e-06, + "loss": 3.1768, + "step": 13947 + }, + { + "epoch": 1.1181776272273556, + "grad_norm": 1.8433233499526978, + "learning_rate": 8.57463170907643e-06, + "loss": 3.1848, + "step": 13948 + }, + { + "epoch": 1.1182578020083782, + "grad_norm": 0.6836932301521301, + "learning_rate": 8.57334653183394e-06, + "loss": 2.4878, + "step": 13949 + }, + { + "epoch": 1.1183379767894008, + "grad_norm": 1.7811305522918701, + "learning_rate": 8.572061378644326e-06, + "loss": 3.1636, + "step": 13950 + }, + { + "epoch": 1.1184181515704235, + "grad_norm": 1.6728664636611938, + "learning_rate": 8.570776249529254e-06, + "loss": 3.1285, + "step": 13951 + }, + { + "epoch": 1.118498326351446, + "grad_norm": 1.950695514678955, + "learning_rate": 8.569491144510389e-06, + "loss": 3.2028, + "step": 13952 + }, + { + "epoch": 1.1185785011324687, + "grad_norm": 1.8261678218841553, + "learning_rate": 8.5682060636094e-06, + "loss": 3.304, + "step": 13953 + }, + { + "epoch": 1.1186586759134913, + "grad_norm": 1.69900643825531, + "learning_rate": 8.566921006847953e-06, + "loss": 2.9847, + "step": 13954 + }, + { + "epoch": 1.118738850694514, + "grad_norm": 1.7524877786636353, + "learning_rate": 8.565635974247711e-06, + "loss": 3.1975, + "step": 13955 + }, + { + "epoch": 1.1188190254755366, + "grad_norm": 1.7163264751434326, + "learning_rate": 8.564350965830343e-06, + "loss": 3.1967, + "step": 13956 + }, + { + "epoch": 1.1188992002565592, + "grad_norm": 1.8295378684997559, + "learning_rate": 8.563065981617511e-06, + "loss": 3.1827, + "step": 13957 + }, + { + "epoch": 1.1189793750375818, + "grad_norm": 1.724525809288025, + "learning_rate": 8.561781021630878e-06, + "loss": 3.1781, + "step": 13958 + }, + { + "epoch": 1.1190595498186044, + "grad_norm": 1.7686091661453247, + "learning_rate": 8.560496085892113e-06, + "loss": 3.2353, + "step": 13959 + }, + { + "epoch": 1.119139724599627, + "grad_norm": 1.7066121101379395, + "learning_rate": 8.559211174422876e-06, + "loss": 3.2228, + "step": 13960 + }, + { + "epoch": 1.11921989938065, + "grad_norm": 1.7376431226730347, + "learning_rate": 8.557926287244831e-06, + "loss": 3.1898, + "step": 13961 + }, + { + "epoch": 1.1193000741616725, + "grad_norm": 1.7926509380340576, + "learning_rate": 8.556641424379638e-06, + "loss": 3.246, + "step": 13962 + }, + { + "epoch": 1.1193802489426952, + "grad_norm": 1.7515259981155396, + "learning_rate": 8.555356585848965e-06, + "loss": 3.2195, + "step": 13963 + }, + { + "epoch": 1.1194604237237178, + "grad_norm": 1.7848567962646484, + "learning_rate": 8.55407177167447e-06, + "loss": 3.1569, + "step": 13964 + }, + { + "epoch": 1.1195405985047404, + "grad_norm": 1.6487748622894287, + "learning_rate": 8.552786981877815e-06, + "loss": 3.0995, + "step": 13965 + }, + { + "epoch": 1.119620773285763, + "grad_norm": 1.7807344198226929, + "learning_rate": 8.551502216480662e-06, + "loss": 3.1508, + "step": 13966 + }, + { + "epoch": 1.1197009480667857, + "grad_norm": 1.71510910987854, + "learning_rate": 8.55021747550467e-06, + "loss": 3.1532, + "step": 13967 + }, + { + "epoch": 1.1197811228478083, + "grad_norm": 1.6667240858078003, + "learning_rate": 8.5489327589715e-06, + "loss": 3.1726, + "step": 13968 + }, + { + "epoch": 1.119861297628831, + "grad_norm": 0.7109718322753906, + "learning_rate": 8.547648066902817e-06, + "loss": 2.4836, + "step": 13969 + }, + { + "epoch": 1.1199414724098535, + "grad_norm": 1.7795015573501587, + "learning_rate": 8.54636339932027e-06, + "loss": 3.2578, + "step": 13970 + }, + { + "epoch": 1.1200216471908762, + "grad_norm": 1.7868974208831787, + "learning_rate": 8.545078756245527e-06, + "loss": 3.2554, + "step": 13971 + }, + { + "epoch": 1.1201018219718988, + "grad_norm": 1.815710186958313, + "learning_rate": 8.54379413770024e-06, + "loss": 3.1692, + "step": 13972 + }, + { + "epoch": 1.1201819967529214, + "grad_norm": 1.8104453086853027, + "learning_rate": 8.542509543706073e-06, + "loss": 3.1826, + "step": 13973 + }, + { + "epoch": 1.120262171533944, + "grad_norm": 0.6621876358985901, + "learning_rate": 8.541224974284683e-06, + "loss": 2.4827, + "step": 13974 + }, + { + "epoch": 1.1203423463149667, + "grad_norm": 1.9171923398971558, + "learning_rate": 8.539940429457722e-06, + "loss": 3.3447, + "step": 13975 + }, + { + "epoch": 1.1204225210959893, + "grad_norm": 1.75220787525177, + "learning_rate": 8.538655909246855e-06, + "loss": 3.2333, + "step": 13976 + }, + { + "epoch": 1.120502695877012, + "grad_norm": 1.7596074342727661, + "learning_rate": 8.537371413673732e-06, + "loss": 3.1819, + "step": 13977 + }, + { + "epoch": 1.1205828706580345, + "grad_norm": 1.7711033821105957, + "learning_rate": 8.536086942760011e-06, + "loss": 3.1636, + "step": 13978 + }, + { + "epoch": 1.1206630454390571, + "grad_norm": 1.7507224082946777, + "learning_rate": 8.53480249652735e-06, + "loss": 3.1693, + "step": 13979 + }, + { + "epoch": 1.1207432202200798, + "grad_norm": 1.804988980293274, + "learning_rate": 8.5335180749974e-06, + "loss": 3.2089, + "step": 13980 + }, + { + "epoch": 1.1208233950011024, + "grad_norm": 1.847320556640625, + "learning_rate": 8.532233678191821e-06, + "loss": 3.2015, + "step": 13981 + }, + { + "epoch": 1.120903569782125, + "grad_norm": 1.690429449081421, + "learning_rate": 8.530949306132263e-06, + "loss": 3.2579, + "step": 13982 + }, + { + "epoch": 1.1209837445631476, + "grad_norm": 1.989478588104248, + "learning_rate": 8.529664958840382e-06, + "loss": 3.1606, + "step": 13983 + }, + { + "epoch": 1.1210639193441703, + "grad_norm": 1.633665919303894, + "learning_rate": 8.528380636337835e-06, + "loss": 3.1435, + "step": 13984 + }, + { + "epoch": 1.121144094125193, + "grad_norm": 1.8497281074523926, + "learning_rate": 8.527096338646268e-06, + "loss": 3.2966, + "step": 13985 + }, + { + "epoch": 1.1212242689062155, + "grad_norm": 1.713119626045227, + "learning_rate": 8.52581206578734e-06, + "loss": 3.217, + "step": 13986 + }, + { + "epoch": 1.1213044436872381, + "grad_norm": 1.6938080787658691, + "learning_rate": 8.524527817782699e-06, + "loss": 3.1389, + "step": 13987 + }, + { + "epoch": 1.1213846184682608, + "grad_norm": 1.6251471042633057, + "learning_rate": 8.523243594653998e-06, + "loss": 3.1414, + "step": 13988 + }, + { + "epoch": 1.1214647932492834, + "grad_norm": 1.7780150175094604, + "learning_rate": 8.521959396422893e-06, + "loss": 3.1869, + "step": 13989 + }, + { + "epoch": 1.121544968030306, + "grad_norm": 1.7383911609649658, + "learning_rate": 8.520675223111029e-06, + "loss": 3.2189, + "step": 13990 + }, + { + "epoch": 1.1216251428113286, + "grad_norm": 1.7418766021728516, + "learning_rate": 8.519391074740062e-06, + "loss": 3.269, + "step": 13991 + }, + { + "epoch": 1.1217053175923513, + "grad_norm": 1.7452601194381714, + "learning_rate": 8.518106951331636e-06, + "loss": 3.1649, + "step": 13992 + }, + { + "epoch": 1.1217854923733739, + "grad_norm": 1.8280080556869507, + "learning_rate": 8.516822852907407e-06, + "loss": 3.1691, + "step": 13993 + }, + { + "epoch": 1.1218656671543965, + "grad_norm": 1.791637897491455, + "learning_rate": 8.515538779489023e-06, + "loss": 3.2298, + "step": 13994 + }, + { + "epoch": 1.1219458419354191, + "grad_norm": 1.7248836755752563, + "learning_rate": 8.51425473109813e-06, + "loss": 3.2493, + "step": 13995 + }, + { + "epoch": 1.1220260167164418, + "grad_norm": 1.7629510164260864, + "learning_rate": 8.512970707756381e-06, + "loss": 3.2205, + "step": 13996 + }, + { + "epoch": 1.1221061914974644, + "grad_norm": 1.8037490844726562, + "learning_rate": 8.511686709485419e-06, + "loss": 3.203, + "step": 13997 + }, + { + "epoch": 1.122186366278487, + "grad_norm": 0.7038210034370422, + "learning_rate": 8.510402736306892e-06, + "loss": 2.4677, + "step": 13998 + }, + { + "epoch": 1.1222665410595098, + "grad_norm": 1.795810341835022, + "learning_rate": 8.509118788242456e-06, + "loss": 3.202, + "step": 13999 + }, + { + "epoch": 1.1223467158405325, + "grad_norm": 1.833135962486267, + "learning_rate": 8.50783486531375e-06, + "loss": 3.1494, + "step": 14000 + }, + { + "epoch": 1.122426890621555, + "grad_norm": 0.710902750492096, + "learning_rate": 8.506550967542423e-06, + "loss": 2.4609, + "step": 14001 + }, + { + "epoch": 1.1225070654025777, + "grad_norm": 1.7931965589523315, + "learning_rate": 8.505267094950116e-06, + "loss": 3.1138, + "step": 14002 + }, + { + "epoch": 1.1225872401836003, + "grad_norm": 1.7507109642028809, + "learning_rate": 8.503983247558481e-06, + "loss": 3.137, + "step": 14003 + }, + { + "epoch": 1.122667414964623, + "grad_norm": 1.7549079656600952, + "learning_rate": 8.502699425389163e-06, + "loss": 3.2334, + "step": 14004 + }, + { + "epoch": 1.1227475897456456, + "grad_norm": 1.8914328813552856, + "learning_rate": 8.501415628463802e-06, + "loss": 3.1693, + "step": 14005 + }, + { + "epoch": 1.1228277645266682, + "grad_norm": 1.8982341289520264, + "learning_rate": 8.50013185680405e-06, + "loss": 3.1642, + "step": 14006 + }, + { + "epoch": 1.1229079393076908, + "grad_norm": 1.7924516201019287, + "learning_rate": 8.498848110431543e-06, + "loss": 3.2441, + "step": 14007 + }, + { + "epoch": 1.1229881140887135, + "grad_norm": 0.701586902141571, + "learning_rate": 8.497564389367927e-06, + "loss": 2.4814, + "step": 14008 + }, + { + "epoch": 1.123068288869736, + "grad_norm": 1.7640438079833984, + "learning_rate": 8.496280693634849e-06, + "loss": 3.1396, + "step": 14009 + }, + { + "epoch": 1.1231484636507587, + "grad_norm": 1.8207013607025146, + "learning_rate": 8.494997023253945e-06, + "loss": 3.2162, + "step": 14010 + }, + { + "epoch": 1.1232286384317813, + "grad_norm": 1.8145782947540283, + "learning_rate": 8.493713378246866e-06, + "loss": 3.152, + "step": 14011 + }, + { + "epoch": 1.123308813212804, + "grad_norm": 1.756278157234192, + "learning_rate": 8.492429758635243e-06, + "loss": 3.1931, + "step": 14012 + }, + { + "epoch": 1.1233889879938266, + "grad_norm": 1.7814171314239502, + "learning_rate": 8.491146164440725e-06, + "loss": 3.1464, + "step": 14013 + }, + { + "epoch": 1.1234691627748492, + "grad_norm": 1.7808198928833008, + "learning_rate": 8.489862595684952e-06, + "loss": 3.1698, + "step": 14014 + }, + { + "epoch": 1.1235493375558718, + "grad_norm": 1.8431016206741333, + "learning_rate": 8.48857905238956e-06, + "loss": 3.2755, + "step": 14015 + }, + { + "epoch": 1.1236295123368945, + "grad_norm": 1.7670395374298096, + "learning_rate": 8.487295534576197e-06, + "loss": 3.1733, + "step": 14016 + }, + { + "epoch": 1.123709687117917, + "grad_norm": 1.9286494255065918, + "learning_rate": 8.486012042266496e-06, + "loss": 3.2667, + "step": 14017 + }, + { + "epoch": 1.1237898618989397, + "grad_norm": 1.7121132612228394, + "learning_rate": 8.484728575482098e-06, + "loss": 3.1813, + "step": 14018 + }, + { + "epoch": 1.1238700366799623, + "grad_norm": 0.6815562844276428, + "learning_rate": 8.483445134244643e-06, + "loss": 2.4675, + "step": 14019 + }, + { + "epoch": 1.123950211460985, + "grad_norm": 0.6916275024414062, + "learning_rate": 8.482161718575767e-06, + "loss": 2.4716, + "step": 14020 + }, + { + "epoch": 1.1240303862420076, + "grad_norm": 1.7527709007263184, + "learning_rate": 8.480878328497112e-06, + "loss": 3.1653, + "step": 14021 + }, + { + "epoch": 1.1241105610230302, + "grad_norm": 1.7459526062011719, + "learning_rate": 8.479594964030311e-06, + "loss": 3.1252, + "step": 14022 + }, + { + "epoch": 1.1241907358040528, + "grad_norm": 1.7917577028274536, + "learning_rate": 8.478311625197004e-06, + "loss": 3.2912, + "step": 14023 + }, + { + "epoch": 1.1242709105850754, + "grad_norm": 1.7833329439163208, + "learning_rate": 8.477028312018827e-06, + "loss": 3.0973, + "step": 14024 + }, + { + "epoch": 1.124351085366098, + "grad_norm": 1.7592747211456299, + "learning_rate": 8.475745024517415e-06, + "loss": 3.1981, + "step": 14025 + }, + { + "epoch": 1.1244312601471207, + "grad_norm": 1.8716046810150146, + "learning_rate": 8.474461762714405e-06, + "loss": 3.0723, + "step": 14026 + }, + { + "epoch": 1.1245114349281433, + "grad_norm": 1.786121129989624, + "learning_rate": 8.47317852663143e-06, + "loss": 3.1682, + "step": 14027 + }, + { + "epoch": 1.124591609709166, + "grad_norm": 1.7692089080810547, + "learning_rate": 8.471895316290131e-06, + "loss": 3.0813, + "step": 14028 + }, + { + "epoch": 1.1246717844901886, + "grad_norm": 1.9914064407348633, + "learning_rate": 8.470612131712137e-06, + "loss": 3.1812, + "step": 14029 + }, + { + "epoch": 1.1247519592712112, + "grad_norm": 1.8139342069625854, + "learning_rate": 8.469328972919082e-06, + "loss": 3.2378, + "step": 14030 + }, + { + "epoch": 1.1248321340522338, + "grad_norm": 1.767323613166809, + "learning_rate": 8.468045839932604e-06, + "loss": 3.1457, + "step": 14031 + }, + { + "epoch": 1.1249123088332564, + "grad_norm": 1.7943658828735352, + "learning_rate": 8.46676273277433e-06, + "loss": 3.2542, + "step": 14032 + }, + { + "epoch": 1.124992483614279, + "grad_norm": 1.8165433406829834, + "learning_rate": 8.465479651465896e-06, + "loss": 3.1958, + "step": 14033 + }, + { + "epoch": 1.1250726583953017, + "grad_norm": 1.773780345916748, + "learning_rate": 8.464196596028936e-06, + "loss": 3.2082, + "step": 14034 + }, + { + "epoch": 1.1251528331763243, + "grad_norm": 1.793300986289978, + "learning_rate": 8.46291356648508e-06, + "loss": 3.1703, + "step": 14035 + }, + { + "epoch": 1.125233007957347, + "grad_norm": 1.731691598892212, + "learning_rate": 8.461630562855959e-06, + "loss": 3.1887, + "step": 14036 + }, + { + "epoch": 1.1253131827383696, + "grad_norm": 1.8187313079833984, + "learning_rate": 8.460347585163204e-06, + "loss": 3.3113, + "step": 14037 + }, + { + "epoch": 1.1253933575193922, + "grad_norm": 1.9052128791809082, + "learning_rate": 8.459064633428446e-06, + "loss": 3.1407, + "step": 14038 + }, + { + "epoch": 1.1254735323004148, + "grad_norm": 1.7100368738174438, + "learning_rate": 8.457781707673317e-06, + "loss": 3.1832, + "step": 14039 + }, + { + "epoch": 1.1255537070814374, + "grad_norm": 1.8575721979141235, + "learning_rate": 8.456498807919443e-06, + "loss": 3.3658, + "step": 14040 + }, + { + "epoch": 1.12563388186246, + "grad_norm": 1.770408034324646, + "learning_rate": 8.455215934188458e-06, + "loss": 3.2186, + "step": 14041 + }, + { + "epoch": 1.1257140566434827, + "grad_norm": 1.778342366218567, + "learning_rate": 8.453933086501984e-06, + "loss": 3.0758, + "step": 14042 + }, + { + "epoch": 1.1257942314245053, + "grad_norm": 1.6536601781845093, + "learning_rate": 8.452650264881654e-06, + "loss": 3.1814, + "step": 14043 + }, + { + "epoch": 1.1258744062055281, + "grad_norm": 1.7989754676818848, + "learning_rate": 8.451367469349098e-06, + "loss": 3.1398, + "step": 14044 + }, + { + "epoch": 1.1259545809865508, + "grad_norm": 1.7272847890853882, + "learning_rate": 8.450084699925939e-06, + "loss": 3.1709, + "step": 14045 + }, + { + "epoch": 1.1260347557675734, + "grad_norm": 1.825188398361206, + "learning_rate": 8.448801956633807e-06, + "loss": 3.1568, + "step": 14046 + }, + { + "epoch": 1.126114930548596, + "grad_norm": 1.810390591621399, + "learning_rate": 8.447519239494324e-06, + "loss": 3.2311, + "step": 14047 + }, + { + "epoch": 1.1261951053296186, + "grad_norm": 1.794504165649414, + "learning_rate": 8.446236548529122e-06, + "loss": 3.2375, + "step": 14048 + }, + { + "epoch": 1.1262752801106413, + "grad_norm": 1.7594454288482666, + "learning_rate": 8.444953883759823e-06, + "loss": 3.2225, + "step": 14049 + }, + { + "epoch": 1.1263554548916639, + "grad_norm": 2.0199615955352783, + "learning_rate": 8.443671245208053e-06, + "loss": 3.2318, + "step": 14050 + }, + { + "epoch": 1.1264356296726865, + "grad_norm": 1.7719762325286865, + "learning_rate": 8.44238863289544e-06, + "loss": 3.3153, + "step": 14051 + }, + { + "epoch": 1.1265158044537091, + "grad_norm": 1.8696041107177734, + "learning_rate": 8.441106046843603e-06, + "loss": 3.2701, + "step": 14052 + }, + { + "epoch": 1.1265959792347318, + "grad_norm": 1.8073253631591797, + "learning_rate": 8.43982348707417e-06, + "loss": 3.1382, + "step": 14053 + }, + { + "epoch": 1.1266761540157544, + "grad_norm": 1.6936030387878418, + "learning_rate": 8.438540953608764e-06, + "loss": 3.1121, + "step": 14054 + }, + { + "epoch": 1.126756328796777, + "grad_norm": 1.7719348669052124, + "learning_rate": 8.437258446469006e-06, + "loss": 3.2505, + "step": 14055 + }, + { + "epoch": 1.1268365035777996, + "grad_norm": 1.7490336894989014, + "learning_rate": 8.43597596567652e-06, + "loss": 3.2001, + "step": 14056 + }, + { + "epoch": 1.1269166783588223, + "grad_norm": 0.695473313331604, + "learning_rate": 8.434693511252927e-06, + "loss": 2.4812, + "step": 14057 + }, + { + "epoch": 1.1269968531398449, + "grad_norm": 1.8060529232025146, + "learning_rate": 8.43341108321985e-06, + "loss": 3.1565, + "step": 14058 + }, + { + "epoch": 1.1270770279208675, + "grad_norm": 1.7368285655975342, + "learning_rate": 8.432128681598912e-06, + "loss": 3.2158, + "step": 14059 + }, + { + "epoch": 1.1271572027018901, + "grad_norm": 1.681334376335144, + "learning_rate": 8.430846306411729e-06, + "loss": 3.171, + "step": 14060 + }, + { + "epoch": 1.1272373774829128, + "grad_norm": 1.7714489698410034, + "learning_rate": 8.429563957679928e-06, + "loss": 3.1844, + "step": 14061 + }, + { + "epoch": 1.1273175522639354, + "grad_norm": 1.8596901893615723, + "learning_rate": 8.428281635425121e-06, + "loss": 3.2248, + "step": 14062 + }, + { + "epoch": 1.127397727044958, + "grad_norm": 1.8264532089233398, + "learning_rate": 8.426999339668934e-06, + "loss": 3.1271, + "step": 14063 + }, + { + "epoch": 1.1274779018259806, + "grad_norm": 1.7286880016326904, + "learning_rate": 8.425717070432985e-06, + "loss": 3.1566, + "step": 14064 + }, + { + "epoch": 1.1275580766070032, + "grad_norm": 1.7964673042297363, + "learning_rate": 8.424434827738888e-06, + "loss": 3.2876, + "step": 14065 + }, + { + "epoch": 1.1276382513880259, + "grad_norm": 1.8738043308258057, + "learning_rate": 8.423152611608269e-06, + "loss": 3.2416, + "step": 14066 + }, + { + "epoch": 1.1277184261690485, + "grad_norm": 1.700365424156189, + "learning_rate": 8.421870422062737e-06, + "loss": 3.245, + "step": 14067 + }, + { + "epoch": 1.1277986009500711, + "grad_norm": 0.6865878701210022, + "learning_rate": 8.420588259123913e-06, + "loss": 2.4769, + "step": 14068 + }, + { + "epoch": 1.1278787757310937, + "grad_norm": 1.7974133491516113, + "learning_rate": 8.419306122813418e-06, + "loss": 3.1221, + "step": 14069 + }, + { + "epoch": 1.1279589505121164, + "grad_norm": 1.936587929725647, + "learning_rate": 8.418024013152863e-06, + "loss": 3.2906, + "step": 14070 + }, + { + "epoch": 1.128039125293139, + "grad_norm": 1.7470088005065918, + "learning_rate": 8.416741930163868e-06, + "loss": 3.1503, + "step": 14071 + }, + { + "epoch": 1.1281193000741616, + "grad_norm": 1.6367576122283936, + "learning_rate": 8.41545987386804e-06, + "loss": 3.1602, + "step": 14072 + }, + { + "epoch": 1.1281994748551842, + "grad_norm": 1.7885136604309082, + "learning_rate": 8.414177844287006e-06, + "loss": 3.1871, + "step": 14073 + }, + { + "epoch": 1.1282796496362069, + "grad_norm": 1.7497949600219727, + "learning_rate": 8.412895841442372e-06, + "loss": 3.1528, + "step": 14074 + }, + { + "epoch": 1.1283598244172295, + "grad_norm": 1.9668583869934082, + "learning_rate": 8.411613865355754e-06, + "loss": 3.0696, + "step": 14075 + }, + { + "epoch": 1.128439999198252, + "grad_norm": 1.8296891450881958, + "learning_rate": 8.41033191604877e-06, + "loss": 3.2495, + "step": 14076 + }, + { + "epoch": 1.1285201739792747, + "grad_norm": 1.7857544422149658, + "learning_rate": 8.409049993543026e-06, + "loss": 3.2385, + "step": 14077 + }, + { + "epoch": 1.1286003487602974, + "grad_norm": 0.7159144282341003, + "learning_rate": 8.407768097860139e-06, + "loss": 2.4888, + "step": 14078 + }, + { + "epoch": 1.1286805235413202, + "grad_norm": 1.8017040491104126, + "learning_rate": 8.406486229021722e-06, + "loss": 3.2507, + "step": 14079 + }, + { + "epoch": 1.1287606983223428, + "grad_norm": 1.7988957166671753, + "learning_rate": 8.405204387049384e-06, + "loss": 3.1591, + "step": 14080 + }, + { + "epoch": 1.1288408731033654, + "grad_norm": 1.9046826362609863, + "learning_rate": 8.403922571964742e-06, + "loss": 3.256, + "step": 14081 + }, + { + "epoch": 1.128921047884388, + "grad_norm": 1.773594856262207, + "learning_rate": 8.402640783789397e-06, + "loss": 3.1827, + "step": 14082 + }, + { + "epoch": 1.1290012226654107, + "grad_norm": 0.6960132718086243, + "learning_rate": 8.401359022544968e-06, + "loss": 2.4658, + "step": 14083 + }, + { + "epoch": 1.1290813974464333, + "grad_norm": 1.734902024269104, + "learning_rate": 8.400077288253064e-06, + "loss": 3.1591, + "step": 14084 + }, + { + "epoch": 1.129161572227456, + "grad_norm": 1.922974705696106, + "learning_rate": 8.398795580935292e-06, + "loss": 3.199, + "step": 14085 + }, + { + "epoch": 1.1292417470084786, + "grad_norm": 1.8013505935668945, + "learning_rate": 8.397513900613263e-06, + "loss": 3.2331, + "step": 14086 + }, + { + "epoch": 1.1293219217895012, + "grad_norm": 1.8071030378341675, + "learning_rate": 8.396232247308583e-06, + "loss": 3.2697, + "step": 14087 + }, + { + "epoch": 1.1294020965705238, + "grad_norm": 1.6463165283203125, + "learning_rate": 8.394950621042861e-06, + "loss": 3.1761, + "step": 14088 + }, + { + "epoch": 1.1294822713515464, + "grad_norm": 1.917726993560791, + "learning_rate": 8.39366902183771e-06, + "loss": 3.2104, + "step": 14089 + }, + { + "epoch": 1.129562446132569, + "grad_norm": 1.7768669128417969, + "learning_rate": 8.392387449714729e-06, + "loss": 3.1646, + "step": 14090 + }, + { + "epoch": 1.1296426209135917, + "grad_norm": 1.759178876876831, + "learning_rate": 8.391105904695534e-06, + "loss": 3.2366, + "step": 14091 + }, + { + "epoch": 1.1297227956946143, + "grad_norm": 1.7436717748641968, + "learning_rate": 8.389824386801722e-06, + "loss": 3.136, + "step": 14092 + }, + { + "epoch": 1.129802970475637, + "grad_norm": 1.7458449602127075, + "learning_rate": 8.388542896054904e-06, + "loss": 3.2672, + "step": 14093 + }, + { + "epoch": 1.1298831452566596, + "grad_norm": 1.7109448909759521, + "learning_rate": 8.387261432476686e-06, + "loss": 3.2716, + "step": 14094 + }, + { + "epoch": 1.1299633200376822, + "grad_norm": 1.8609789609909058, + "learning_rate": 8.38597999608867e-06, + "loss": 3.1416, + "step": 14095 + }, + { + "epoch": 1.1300434948187048, + "grad_norm": 1.7535839080810547, + "learning_rate": 8.384698586912467e-06, + "loss": 3.2289, + "step": 14096 + }, + { + "epoch": 1.1301236695997274, + "grad_norm": 1.7867857217788696, + "learning_rate": 8.38341720496967e-06, + "loss": 3.2166, + "step": 14097 + }, + { + "epoch": 1.13020384438075, + "grad_norm": 1.9237085580825806, + "learning_rate": 8.382135850281894e-06, + "loss": 3.1322, + "step": 14098 + }, + { + "epoch": 1.1302840191617727, + "grad_norm": 1.9784857034683228, + "learning_rate": 8.380854522870737e-06, + "loss": 3.1514, + "step": 14099 + }, + { + "epoch": 1.1303641939427953, + "grad_norm": 1.7275282144546509, + "learning_rate": 8.3795732227578e-06, + "loss": 3.1041, + "step": 14100 + }, + { + "epoch": 1.130444368723818, + "grad_norm": 1.6909635066986084, + "learning_rate": 8.37829194996469e-06, + "loss": 3.1385, + "step": 14101 + }, + { + "epoch": 1.1305245435048406, + "grad_norm": 1.82931649684906, + "learning_rate": 8.377010704513003e-06, + "loss": 3.1946, + "step": 14102 + }, + { + "epoch": 1.1306047182858632, + "grad_norm": 1.8167665004730225, + "learning_rate": 8.375729486424345e-06, + "loss": 3.2079, + "step": 14103 + }, + { + "epoch": 1.1306848930668858, + "grad_norm": 1.8012479543685913, + "learning_rate": 8.374448295720317e-06, + "loss": 3.1658, + "step": 14104 + }, + { + "epoch": 1.1307650678479084, + "grad_norm": 1.7903213500976562, + "learning_rate": 8.373167132422515e-06, + "loss": 3.2405, + "step": 14105 + }, + { + "epoch": 1.130845242628931, + "grad_norm": 1.8374435901641846, + "learning_rate": 8.371885996552544e-06, + "loss": 3.1519, + "step": 14106 + }, + { + "epoch": 1.1309254174099537, + "grad_norm": 0.700518786907196, + "learning_rate": 8.370604888132e-06, + "loss": 2.4867, + "step": 14107 + }, + { + "epoch": 1.1310055921909763, + "grad_norm": 1.834307312965393, + "learning_rate": 8.369323807182484e-06, + "loss": 3.2732, + "step": 14108 + }, + { + "epoch": 1.131085766971999, + "grad_norm": 1.7231864929199219, + "learning_rate": 8.368042753725594e-06, + "loss": 3.1779, + "step": 14109 + }, + { + "epoch": 1.1311659417530215, + "grad_norm": 1.713077425956726, + "learning_rate": 8.366761727782927e-06, + "loss": 3.1754, + "step": 14110 + }, + { + "epoch": 1.1312461165340442, + "grad_norm": 1.779721736907959, + "learning_rate": 8.365480729376083e-06, + "loss": 3.2163, + "step": 14111 + }, + { + "epoch": 1.1313262913150668, + "grad_norm": 1.9017479419708252, + "learning_rate": 8.364199758526657e-06, + "loss": 3.2087, + "step": 14112 + }, + { + "epoch": 1.1314064660960894, + "grad_norm": 1.8443504571914673, + "learning_rate": 8.362918815256245e-06, + "loss": 3.1859, + "step": 14113 + }, + { + "epoch": 1.131486640877112, + "grad_norm": 2.0552425384521484, + "learning_rate": 8.361637899586448e-06, + "loss": 3.2026, + "step": 14114 + }, + { + "epoch": 1.1315668156581347, + "grad_norm": 1.7599653005599976, + "learning_rate": 8.360357011538858e-06, + "loss": 3.3834, + "step": 14115 + }, + { + "epoch": 1.1316469904391573, + "grad_norm": 1.8159457445144653, + "learning_rate": 8.359076151135068e-06, + "loss": 3.2631, + "step": 14116 + }, + { + "epoch": 1.13172716522018, + "grad_norm": 0.6904811263084412, + "learning_rate": 8.357795318396679e-06, + "loss": 2.4939, + "step": 14117 + }, + { + "epoch": 1.1318073400012025, + "grad_norm": 1.7003858089447021, + "learning_rate": 8.35651451334528e-06, + "loss": 3.1958, + "step": 14118 + }, + { + "epoch": 1.1318875147822252, + "grad_norm": 1.7434544563293457, + "learning_rate": 8.35523373600247e-06, + "loss": 3.1502, + "step": 14119 + }, + { + "epoch": 1.1319676895632478, + "grad_norm": 1.7844619750976562, + "learning_rate": 8.353952986389837e-06, + "loss": 3.1088, + "step": 14120 + }, + { + "epoch": 1.1320478643442704, + "grad_norm": 0.719818651676178, + "learning_rate": 8.352672264528976e-06, + "loss": 2.4768, + "step": 14121 + }, + { + "epoch": 1.132128039125293, + "grad_norm": 1.7928211688995361, + "learning_rate": 8.351391570441482e-06, + "loss": 3.1203, + "step": 14122 + }, + { + "epoch": 1.1322082139063157, + "grad_norm": 1.7777589559555054, + "learning_rate": 8.350110904148945e-06, + "loss": 3.1698, + "step": 14123 + }, + { + "epoch": 1.1322883886873383, + "grad_norm": 1.7991451025009155, + "learning_rate": 8.348830265672957e-06, + "loss": 3.1393, + "step": 14124 + }, + { + "epoch": 1.1323685634683611, + "grad_norm": 0.6676585674285889, + "learning_rate": 8.347549655035108e-06, + "loss": 2.4818, + "step": 14125 + }, + { + "epoch": 1.1324487382493837, + "grad_norm": 1.666838526725769, + "learning_rate": 8.346269072256988e-06, + "loss": 3.0953, + "step": 14126 + }, + { + "epoch": 1.1325289130304064, + "grad_norm": 1.7866543531417847, + "learning_rate": 8.34498851736019e-06, + "loss": 3.1816, + "step": 14127 + }, + { + "epoch": 1.132609087811429, + "grad_norm": 1.7644133567810059, + "learning_rate": 8.343707990366303e-06, + "loss": 3.188, + "step": 14128 + }, + { + "epoch": 1.1326892625924516, + "grad_norm": 1.72189199924469, + "learning_rate": 8.342427491296915e-06, + "loss": 3.3254, + "step": 14129 + }, + { + "epoch": 1.1327694373734742, + "grad_norm": 1.8623558282852173, + "learning_rate": 8.341147020173615e-06, + "loss": 3.2555, + "step": 14130 + }, + { + "epoch": 1.1328496121544969, + "grad_norm": 0.6858062744140625, + "learning_rate": 8.33986657701799e-06, + "loss": 2.4688, + "step": 14131 + }, + { + "epoch": 1.1329297869355195, + "grad_norm": 2.519625663757324, + "learning_rate": 8.338586161851632e-06, + "loss": 3.2232, + "step": 14132 + }, + { + "epoch": 1.1330099617165421, + "grad_norm": 1.8252084255218506, + "learning_rate": 8.337305774696124e-06, + "loss": 3.2161, + "step": 14133 + }, + { + "epoch": 1.1330901364975647, + "grad_norm": 1.8566884994506836, + "learning_rate": 8.336025415573057e-06, + "loss": 3.1038, + "step": 14134 + }, + { + "epoch": 1.1331703112785874, + "grad_norm": 1.7212475538253784, + "learning_rate": 8.334745084504013e-06, + "loss": 3.1329, + "step": 14135 + }, + { + "epoch": 1.13325048605961, + "grad_norm": 1.8147820234298706, + "learning_rate": 8.33346478151058e-06, + "loss": 3.1144, + "step": 14136 + }, + { + "epoch": 1.1333306608406326, + "grad_norm": 1.775847315788269, + "learning_rate": 8.332184506614345e-06, + "loss": 3.236, + "step": 14137 + }, + { + "epoch": 1.1334108356216552, + "grad_norm": 1.8414888381958008, + "learning_rate": 8.33090425983689e-06, + "loss": 3.2615, + "step": 14138 + }, + { + "epoch": 1.1334910104026779, + "grad_norm": 1.7602123022079468, + "learning_rate": 8.329624041199802e-06, + "loss": 3.1921, + "step": 14139 + }, + { + "epoch": 1.1335711851837005, + "grad_norm": 1.8808382749557495, + "learning_rate": 8.328343850724663e-06, + "loss": 3.2322, + "step": 14140 + }, + { + "epoch": 1.133651359964723, + "grad_norm": 1.858198642730713, + "learning_rate": 8.327063688433057e-06, + "loss": 3.1313, + "step": 14141 + }, + { + "epoch": 1.1337315347457457, + "grad_norm": 1.8826881647109985, + "learning_rate": 8.32578355434657e-06, + "loss": 3.1855, + "step": 14142 + }, + { + "epoch": 1.1338117095267684, + "grad_norm": 1.807042121887207, + "learning_rate": 8.324503448486777e-06, + "loss": 3.2032, + "step": 14143 + }, + { + "epoch": 1.133891884307791, + "grad_norm": 1.8972727060317993, + "learning_rate": 8.323223370875271e-06, + "loss": 3.1375, + "step": 14144 + }, + { + "epoch": 1.1339720590888136, + "grad_norm": 1.7258199453353882, + "learning_rate": 8.321943321533627e-06, + "loss": 3.1296, + "step": 14145 + }, + { + "epoch": 1.1340522338698362, + "grad_norm": 1.71628999710083, + "learning_rate": 8.320663300483426e-06, + "loss": 3.2021, + "step": 14146 + }, + { + "epoch": 1.1341324086508588, + "grad_norm": 1.7275516986846924, + "learning_rate": 8.319383307746249e-06, + "loss": 3.131, + "step": 14147 + }, + { + "epoch": 1.1342125834318815, + "grad_norm": 1.8093068599700928, + "learning_rate": 8.31810334334368e-06, + "loss": 3.2358, + "step": 14148 + }, + { + "epoch": 1.134292758212904, + "grad_norm": 1.769159197807312, + "learning_rate": 8.316823407297295e-06, + "loss": 3.2058, + "step": 14149 + }, + { + "epoch": 1.1343729329939267, + "grad_norm": 1.8839564323425293, + "learning_rate": 8.315543499628672e-06, + "loss": 3.2135, + "step": 14150 + }, + { + "epoch": 1.1344531077749493, + "grad_norm": 1.71065092086792, + "learning_rate": 8.314263620359392e-06, + "loss": 3.0819, + "step": 14151 + }, + { + "epoch": 1.134533282555972, + "grad_norm": 1.8357429504394531, + "learning_rate": 8.312983769511035e-06, + "loss": 3.3052, + "step": 14152 + }, + { + "epoch": 1.1346134573369946, + "grad_norm": 1.8753234148025513, + "learning_rate": 8.311703947105176e-06, + "loss": 3.1998, + "step": 14153 + }, + { + "epoch": 1.1346936321180172, + "grad_norm": 1.6881253719329834, + "learning_rate": 8.310424153163396e-06, + "loss": 3.2072, + "step": 14154 + }, + { + "epoch": 1.1347738068990398, + "grad_norm": 1.8158209323883057, + "learning_rate": 8.309144387707265e-06, + "loss": 3.2794, + "step": 14155 + }, + { + "epoch": 1.1348539816800625, + "grad_norm": 0.7489500045776367, + "learning_rate": 8.307864650758366e-06, + "loss": 2.4719, + "step": 14156 + }, + { + "epoch": 1.134934156461085, + "grad_norm": 1.8026652336120605, + "learning_rate": 8.306584942338274e-06, + "loss": 3.1165, + "step": 14157 + }, + { + "epoch": 1.1350143312421077, + "grad_norm": 1.786382794380188, + "learning_rate": 8.30530526246856e-06, + "loss": 3.137, + "step": 14158 + }, + { + "epoch": 1.1350945060231303, + "grad_norm": 2.0295774936676025, + "learning_rate": 8.304025611170804e-06, + "loss": 3.1137, + "step": 14159 + }, + { + "epoch": 1.135174680804153, + "grad_norm": 1.8264743089675903, + "learning_rate": 8.302745988466578e-06, + "loss": 3.1918, + "step": 14160 + }, + { + "epoch": 1.1352548555851758, + "grad_norm": 1.7712639570236206, + "learning_rate": 8.301466394377456e-06, + "loss": 3.1804, + "step": 14161 + }, + { + "epoch": 1.1353350303661984, + "grad_norm": 1.8331718444824219, + "learning_rate": 8.300186828925015e-06, + "loss": 3.1978, + "step": 14162 + }, + { + "epoch": 1.135415205147221, + "grad_norm": 1.8429063558578491, + "learning_rate": 8.298907292130822e-06, + "loss": 3.1448, + "step": 14163 + }, + { + "epoch": 1.1354953799282437, + "grad_norm": 1.7318657636642456, + "learning_rate": 8.297627784016453e-06, + "loss": 3.1354, + "step": 14164 + }, + { + "epoch": 1.1355755547092663, + "grad_norm": 1.8615630865097046, + "learning_rate": 8.296348304603478e-06, + "loss": 3.2313, + "step": 14165 + }, + { + "epoch": 1.135655729490289, + "grad_norm": 1.6761589050292969, + "learning_rate": 8.295068853913473e-06, + "loss": 3.189, + "step": 14166 + }, + { + "epoch": 1.1357359042713115, + "grad_norm": 1.7721576690673828, + "learning_rate": 8.293789431968006e-06, + "loss": 3.1945, + "step": 14167 + }, + { + "epoch": 1.1358160790523342, + "grad_norm": 1.9110369682312012, + "learning_rate": 8.292510038788645e-06, + "loss": 3.2005, + "step": 14168 + }, + { + "epoch": 1.1358962538333568, + "grad_norm": 1.825944423675537, + "learning_rate": 8.291230674396967e-06, + "loss": 3.297, + "step": 14169 + }, + { + "epoch": 1.1359764286143794, + "grad_norm": 1.931525707244873, + "learning_rate": 8.289951338814535e-06, + "loss": 3.2286, + "step": 14170 + }, + { + "epoch": 1.136056603395402, + "grad_norm": 1.8088678121566772, + "learning_rate": 8.28867203206292e-06, + "loss": 3.2703, + "step": 14171 + }, + { + "epoch": 1.1361367781764247, + "grad_norm": 1.7759413719177246, + "learning_rate": 8.287392754163695e-06, + "loss": 3.2374, + "step": 14172 + }, + { + "epoch": 1.1362169529574473, + "grad_norm": 1.818644642829895, + "learning_rate": 8.28611350513842e-06, + "loss": 3.2402, + "step": 14173 + }, + { + "epoch": 1.13629712773847, + "grad_norm": 1.8618780374526978, + "learning_rate": 8.284834285008672e-06, + "loss": 3.1398, + "step": 14174 + }, + { + "epoch": 1.1363773025194925, + "grad_norm": 1.914506196975708, + "learning_rate": 8.28355509379601e-06, + "loss": 3.2406, + "step": 14175 + }, + { + "epoch": 1.1364574773005152, + "grad_norm": 1.7365808486938477, + "learning_rate": 8.282275931522004e-06, + "loss": 3.1724, + "step": 14176 + }, + { + "epoch": 1.1365376520815378, + "grad_norm": 1.7521589994430542, + "learning_rate": 8.280996798208223e-06, + "loss": 3.1589, + "step": 14177 + }, + { + "epoch": 1.1366178268625604, + "grad_norm": 0.7642972469329834, + "learning_rate": 8.279717693876229e-06, + "loss": 2.4657, + "step": 14178 + }, + { + "epoch": 1.136698001643583, + "grad_norm": 1.7754759788513184, + "learning_rate": 8.278438618547588e-06, + "loss": 3.0686, + "step": 14179 + }, + { + "epoch": 1.1367781764246057, + "grad_norm": 1.858612060546875, + "learning_rate": 8.277159572243865e-06, + "loss": 3.2776, + "step": 14180 + }, + { + "epoch": 1.1368583512056283, + "grad_norm": 1.7358877658843994, + "learning_rate": 8.275880554986624e-06, + "loss": 3.1807, + "step": 14181 + }, + { + "epoch": 1.136938525986651, + "grad_norm": 1.826965570449829, + "learning_rate": 8.274601566797431e-06, + "loss": 3.254, + "step": 14182 + }, + { + "epoch": 1.1370187007676735, + "grad_norm": 1.7487629652023315, + "learning_rate": 8.273322607697845e-06, + "loss": 3.202, + "step": 14183 + }, + { + "epoch": 1.1370988755486962, + "grad_norm": 1.6790223121643066, + "learning_rate": 8.272043677709434e-06, + "loss": 3.1558, + "step": 14184 + }, + { + "epoch": 1.1371790503297188, + "grad_norm": 1.7915140390396118, + "learning_rate": 8.270764776853756e-06, + "loss": 3.1871, + "step": 14185 + }, + { + "epoch": 1.1372592251107414, + "grad_norm": 1.8224647045135498, + "learning_rate": 8.269485905152374e-06, + "loss": 3.2601, + "step": 14186 + }, + { + "epoch": 1.137339399891764, + "grad_norm": 1.7665787935256958, + "learning_rate": 8.268207062626852e-06, + "loss": 3.2313, + "step": 14187 + }, + { + "epoch": 1.1374195746727866, + "grad_norm": 1.7799723148345947, + "learning_rate": 8.266928249298746e-06, + "loss": 3.0996, + "step": 14188 + }, + { + "epoch": 1.1374997494538093, + "grad_norm": 1.8131492137908936, + "learning_rate": 8.26564946518962e-06, + "loss": 3.3481, + "step": 14189 + }, + { + "epoch": 1.137579924234832, + "grad_norm": 1.8257113695144653, + "learning_rate": 8.264370710321033e-06, + "loss": 3.1732, + "step": 14190 + }, + { + "epoch": 1.1376600990158545, + "grad_norm": 1.9177896976470947, + "learning_rate": 8.263091984714544e-06, + "loss": 3.3, + "step": 14191 + }, + { + "epoch": 1.1377402737968771, + "grad_norm": 1.7796188592910767, + "learning_rate": 8.261813288391711e-06, + "loss": 3.2023, + "step": 14192 + }, + { + "epoch": 1.1378204485778998, + "grad_norm": 1.8932222127914429, + "learning_rate": 8.260534621374095e-06, + "loss": 3.2435, + "step": 14193 + }, + { + "epoch": 1.1379006233589224, + "grad_norm": 1.6689801216125488, + "learning_rate": 8.259255983683252e-06, + "loss": 3.1837, + "step": 14194 + }, + { + "epoch": 1.137980798139945, + "grad_norm": 1.856794834136963, + "learning_rate": 8.257977375340738e-06, + "loss": 3.2859, + "step": 14195 + }, + { + "epoch": 1.1380609729209676, + "grad_norm": 1.7389805316925049, + "learning_rate": 8.256698796368112e-06, + "loss": 3.189, + "step": 14196 + }, + { + "epoch": 1.1381411477019903, + "grad_norm": 1.7762913703918457, + "learning_rate": 8.255420246786932e-06, + "loss": 3.1667, + "step": 14197 + }, + { + "epoch": 1.1382213224830129, + "grad_norm": 1.704142689704895, + "learning_rate": 8.25414172661875e-06, + "loss": 3.2145, + "step": 14198 + }, + { + "epoch": 1.1383014972640355, + "grad_norm": 1.7353999614715576, + "learning_rate": 8.252863235885124e-06, + "loss": 3.0779, + "step": 14199 + }, + { + "epoch": 1.1383816720450581, + "grad_norm": 1.7973824739456177, + "learning_rate": 8.251584774607605e-06, + "loss": 3.1978, + "step": 14200 + }, + { + "epoch": 1.1384618468260808, + "grad_norm": 1.8234820365905762, + "learning_rate": 8.250306342807753e-06, + "loss": 3.1462, + "step": 14201 + }, + { + "epoch": 1.1385420216071034, + "grad_norm": 1.762487769126892, + "learning_rate": 8.249027940507121e-06, + "loss": 3.2136, + "step": 14202 + }, + { + "epoch": 1.138622196388126, + "grad_norm": 1.7309809923171997, + "learning_rate": 8.24774956772726e-06, + "loss": 3.1804, + "step": 14203 + }, + { + "epoch": 1.1387023711691486, + "grad_norm": 1.6893200874328613, + "learning_rate": 8.246471224489723e-06, + "loss": 3.142, + "step": 14204 + }, + { + "epoch": 1.1387825459501713, + "grad_norm": 1.8764172792434692, + "learning_rate": 8.245192910816063e-06, + "loss": 3.0988, + "step": 14205 + }, + { + "epoch": 1.1388627207311939, + "grad_norm": 1.7584747076034546, + "learning_rate": 8.243914626727832e-06, + "loss": 3.135, + "step": 14206 + }, + { + "epoch": 1.1389428955122167, + "grad_norm": 1.845731258392334, + "learning_rate": 8.242636372246582e-06, + "loss": 3.224, + "step": 14207 + }, + { + "epoch": 1.1390230702932393, + "grad_norm": 1.766864538192749, + "learning_rate": 8.241358147393864e-06, + "loss": 3.2099, + "step": 14208 + }, + { + "epoch": 1.139103245074262, + "grad_norm": 1.6702394485473633, + "learning_rate": 8.240079952191227e-06, + "loss": 3.2556, + "step": 14209 + }, + { + "epoch": 1.1391834198552846, + "grad_norm": 1.695544719696045, + "learning_rate": 8.238801786660219e-06, + "loss": 3.1135, + "step": 14210 + }, + { + "epoch": 1.1392635946363072, + "grad_norm": 1.8932064771652222, + "learning_rate": 8.237523650822395e-06, + "loss": 3.1963, + "step": 14211 + }, + { + "epoch": 1.1393437694173298, + "grad_norm": 1.7453800439834595, + "learning_rate": 8.236245544699303e-06, + "loss": 3.2297, + "step": 14212 + }, + { + "epoch": 1.1394239441983525, + "grad_norm": 1.804714560508728, + "learning_rate": 8.234967468312487e-06, + "loss": 3.1202, + "step": 14213 + }, + { + "epoch": 1.139504118979375, + "grad_norm": 1.8078104257583618, + "learning_rate": 8.233689421683498e-06, + "loss": 3.1667, + "step": 14214 + }, + { + "epoch": 1.1395842937603977, + "grad_norm": 1.8084725141525269, + "learning_rate": 8.232411404833882e-06, + "loss": 3.2256, + "step": 14215 + }, + { + "epoch": 1.1396644685414203, + "grad_norm": 1.916723370552063, + "learning_rate": 8.231133417785188e-06, + "loss": 3.2333, + "step": 14216 + }, + { + "epoch": 1.139744643322443, + "grad_norm": 1.801010251045227, + "learning_rate": 8.229855460558962e-06, + "loss": 3.2, + "step": 14217 + }, + { + "epoch": 1.1398248181034656, + "grad_norm": 1.7268182039260864, + "learning_rate": 8.228577533176745e-06, + "loss": 3.1814, + "step": 14218 + }, + { + "epoch": 1.1399049928844882, + "grad_norm": 1.747105598449707, + "learning_rate": 8.227299635660092e-06, + "loss": 3.1584, + "step": 14219 + }, + { + "epoch": 1.1399851676655108, + "grad_norm": 1.7932474613189697, + "learning_rate": 8.226021768030538e-06, + "loss": 3.2029, + "step": 14220 + }, + { + "epoch": 1.1400653424465335, + "grad_norm": 1.712846040725708, + "learning_rate": 8.224743930309634e-06, + "loss": 3.1667, + "step": 14221 + }, + { + "epoch": 1.140145517227556, + "grad_norm": 1.8394732475280762, + "learning_rate": 8.223466122518922e-06, + "loss": 3.1515, + "step": 14222 + }, + { + "epoch": 1.1402256920085787, + "grad_norm": 1.6384270191192627, + "learning_rate": 8.222188344679942e-06, + "loss": 3.1782, + "step": 14223 + }, + { + "epoch": 1.1403058667896013, + "grad_norm": 1.7931541204452515, + "learning_rate": 8.220910596814245e-06, + "loss": 3.1383, + "step": 14224 + }, + { + "epoch": 1.140386041570624, + "grad_norm": 1.7114379405975342, + "learning_rate": 8.219632878943363e-06, + "loss": 3.2284, + "step": 14225 + }, + { + "epoch": 1.1404662163516466, + "grad_norm": 1.7101718187332153, + "learning_rate": 8.218355191088845e-06, + "loss": 3.1105, + "step": 14226 + }, + { + "epoch": 1.1405463911326692, + "grad_norm": 0.6662381291389465, + "learning_rate": 8.217077533272233e-06, + "loss": 2.4734, + "step": 14227 + }, + { + "epoch": 1.1406265659136918, + "grad_norm": 1.7677979469299316, + "learning_rate": 8.215799905515062e-06, + "loss": 3.1175, + "step": 14228 + }, + { + "epoch": 1.1407067406947144, + "grad_norm": 1.8567965030670166, + "learning_rate": 8.214522307838879e-06, + "loss": 3.2755, + "step": 14229 + }, + { + "epoch": 1.140786915475737, + "grad_norm": 1.7929469347000122, + "learning_rate": 8.213244740265218e-06, + "loss": 3.1498, + "step": 14230 + }, + { + "epoch": 1.1408670902567597, + "grad_norm": 1.8824615478515625, + "learning_rate": 8.211967202815622e-06, + "loss": 3.2643, + "step": 14231 + }, + { + "epoch": 1.1409472650377823, + "grad_norm": 1.8470771312713623, + "learning_rate": 8.21068969551163e-06, + "loss": 3.261, + "step": 14232 + }, + { + "epoch": 1.141027439818805, + "grad_norm": 1.7926386594772339, + "learning_rate": 8.209412218374776e-06, + "loss": 3.149, + "step": 14233 + }, + { + "epoch": 1.1411076145998276, + "grad_norm": 1.848510503768921, + "learning_rate": 8.208134771426605e-06, + "loss": 3.1701, + "step": 14234 + }, + { + "epoch": 1.1411877893808502, + "grad_norm": 1.9054155349731445, + "learning_rate": 8.206857354688647e-06, + "loss": 3.2418, + "step": 14235 + }, + { + "epoch": 1.1412679641618728, + "grad_norm": 1.7940654754638672, + "learning_rate": 8.205579968182442e-06, + "loss": 3.2143, + "step": 14236 + }, + { + "epoch": 1.1413481389428954, + "grad_norm": 1.7289152145385742, + "learning_rate": 8.20430261192953e-06, + "loss": 3.1924, + "step": 14237 + }, + { + "epoch": 1.141428313723918, + "grad_norm": 2.078235626220703, + "learning_rate": 8.203025285951438e-06, + "loss": 3.1027, + "step": 14238 + }, + { + "epoch": 1.1415084885049407, + "grad_norm": 1.68754243850708, + "learning_rate": 8.20174799026971e-06, + "loss": 3.1709, + "step": 14239 + }, + { + "epoch": 1.1415886632859633, + "grad_norm": 1.6988484859466553, + "learning_rate": 8.200470724905875e-06, + "loss": 3.1656, + "step": 14240 + }, + { + "epoch": 1.141668838066986, + "grad_norm": 1.766884207725525, + "learning_rate": 8.199193489881471e-06, + "loss": 3.1136, + "step": 14241 + }, + { + "epoch": 1.1417490128480088, + "grad_norm": 1.8077988624572754, + "learning_rate": 8.19791628521803e-06, + "loss": 3.1937, + "step": 14242 + }, + { + "epoch": 1.1418291876290314, + "grad_norm": 1.7810554504394531, + "learning_rate": 8.196639110937083e-06, + "loss": 3.0912, + "step": 14243 + }, + { + "epoch": 1.141909362410054, + "grad_norm": 1.6964256763458252, + "learning_rate": 8.19536196706017e-06, + "loss": 3.2061, + "step": 14244 + }, + { + "epoch": 1.1419895371910767, + "grad_norm": 1.9050081968307495, + "learning_rate": 8.194084853608814e-06, + "loss": 3.1418, + "step": 14245 + }, + { + "epoch": 1.1420697119720993, + "grad_norm": 1.8133299350738525, + "learning_rate": 8.192807770604551e-06, + "loss": 3.2164, + "step": 14246 + }, + { + "epoch": 1.142149886753122, + "grad_norm": 1.8705178499221802, + "learning_rate": 8.191530718068915e-06, + "loss": 3.0756, + "step": 14247 + }, + { + "epoch": 1.1422300615341445, + "grad_norm": 1.6939342021942139, + "learning_rate": 8.190253696023429e-06, + "loss": 3.0438, + "step": 14248 + }, + { + "epoch": 1.1423102363151671, + "grad_norm": 1.5866405963897705, + "learning_rate": 8.188976704489633e-06, + "loss": 3.1147, + "step": 14249 + }, + { + "epoch": 1.1423904110961898, + "grad_norm": 1.7844163179397583, + "learning_rate": 8.187699743489048e-06, + "loss": 3.2721, + "step": 14250 + }, + { + "epoch": 1.1424705858772124, + "grad_norm": 1.7281912565231323, + "learning_rate": 8.186422813043206e-06, + "loss": 3.1792, + "step": 14251 + }, + { + "epoch": 1.142550760658235, + "grad_norm": 1.6031314134597778, + "learning_rate": 8.18514591317364e-06, + "loss": 3.0599, + "step": 14252 + }, + { + "epoch": 1.1426309354392576, + "grad_norm": 1.8141614198684692, + "learning_rate": 8.183869043901872e-06, + "loss": 3.1735, + "step": 14253 + }, + { + "epoch": 1.1427111102202803, + "grad_norm": 1.7570741176605225, + "learning_rate": 8.182592205249432e-06, + "loss": 3.1431, + "step": 14254 + }, + { + "epoch": 1.142791285001303, + "grad_norm": 1.727321982383728, + "learning_rate": 8.181315397237847e-06, + "loss": 3.2085, + "step": 14255 + }, + { + "epoch": 1.1428714597823255, + "grad_norm": 1.8106820583343506, + "learning_rate": 8.180038619888642e-06, + "loss": 3.184, + "step": 14256 + }, + { + "epoch": 1.1429516345633481, + "grad_norm": 1.7886065244674683, + "learning_rate": 8.178761873223347e-06, + "loss": 3.1228, + "step": 14257 + }, + { + "epoch": 1.1430318093443708, + "grad_norm": 1.790745496749878, + "learning_rate": 8.177485157263483e-06, + "loss": 3.1617, + "step": 14258 + }, + { + "epoch": 1.1431119841253934, + "grad_norm": 1.7697136402130127, + "learning_rate": 8.176208472030579e-06, + "loss": 3.1317, + "step": 14259 + }, + { + "epoch": 1.143192158906416, + "grad_norm": 1.8349610567092896, + "learning_rate": 8.174931817546155e-06, + "loss": 3.1817, + "step": 14260 + }, + { + "epoch": 1.1432723336874386, + "grad_norm": 1.8562997579574585, + "learning_rate": 8.173655193831738e-06, + "loss": 3.3174, + "step": 14261 + }, + { + "epoch": 1.1433525084684613, + "grad_norm": 1.8575845956802368, + "learning_rate": 8.172378600908852e-06, + "loss": 3.2411, + "step": 14262 + }, + { + "epoch": 1.1434326832494839, + "grad_norm": 1.8077620267868042, + "learning_rate": 8.171102038799018e-06, + "loss": 3.255, + "step": 14263 + }, + { + "epoch": 1.1435128580305065, + "grad_norm": 1.7655996084213257, + "learning_rate": 8.169825507523759e-06, + "loss": 3.2784, + "step": 14264 + }, + { + "epoch": 1.1435930328115291, + "grad_norm": 1.811947226524353, + "learning_rate": 8.168549007104598e-06, + "loss": 3.1361, + "step": 14265 + }, + { + "epoch": 1.1436732075925518, + "grad_norm": 1.7935352325439453, + "learning_rate": 8.167272537563052e-06, + "loss": 3.225, + "step": 14266 + }, + { + "epoch": 1.1437533823735744, + "grad_norm": 1.7510826587677002, + "learning_rate": 8.165996098920648e-06, + "loss": 3.2352, + "step": 14267 + }, + { + "epoch": 1.143833557154597, + "grad_norm": 1.920200228691101, + "learning_rate": 8.164719691198902e-06, + "loss": 3.218, + "step": 14268 + }, + { + "epoch": 1.1439137319356196, + "grad_norm": 1.8732489347457886, + "learning_rate": 8.163443314419336e-06, + "loss": 3.1437, + "step": 14269 + }, + { + "epoch": 1.1439939067166423, + "grad_norm": 1.7465111017227173, + "learning_rate": 8.16216696860347e-06, + "loss": 3.1935, + "step": 14270 + }, + { + "epoch": 1.1440740814976649, + "grad_norm": 1.7738617658615112, + "learning_rate": 8.160890653772817e-06, + "loss": 3.1453, + "step": 14271 + }, + { + "epoch": 1.1441542562786875, + "grad_norm": 0.698672890663147, + "learning_rate": 8.159614369948904e-06, + "loss": 2.4744, + "step": 14272 + }, + { + "epoch": 1.1442344310597101, + "grad_norm": 1.7846883535385132, + "learning_rate": 8.15833811715324e-06, + "loss": 3.1263, + "step": 14273 + }, + { + "epoch": 1.1443146058407327, + "grad_norm": 1.7438870668411255, + "learning_rate": 8.15706189540735e-06, + "loss": 3.106, + "step": 14274 + }, + { + "epoch": 1.1443947806217554, + "grad_norm": 1.7109707593917847, + "learning_rate": 8.155785704732745e-06, + "loss": 3.2465, + "step": 14275 + }, + { + "epoch": 1.144474955402778, + "grad_norm": 1.7178643941879272, + "learning_rate": 8.154509545150943e-06, + "loss": 3.0983, + "step": 14276 + }, + { + "epoch": 1.1445551301838006, + "grad_norm": 1.7401479482650757, + "learning_rate": 8.153233416683461e-06, + "loss": 3.1672, + "step": 14277 + }, + { + "epoch": 1.1446353049648232, + "grad_norm": 1.8565192222595215, + "learning_rate": 8.15195731935181e-06, + "loss": 3.346, + "step": 14278 + }, + { + "epoch": 1.1447154797458459, + "grad_norm": 1.7124249935150146, + "learning_rate": 8.15068125317751e-06, + "loss": 3.0859, + "step": 14279 + }, + { + "epoch": 1.1447956545268685, + "grad_norm": 1.7675609588623047, + "learning_rate": 8.149405218182071e-06, + "loss": 3.1899, + "step": 14280 + }, + { + "epoch": 1.1448758293078911, + "grad_norm": 1.7516238689422607, + "learning_rate": 8.148129214387007e-06, + "loss": 3.2054, + "step": 14281 + }, + { + "epoch": 1.1449560040889137, + "grad_norm": 1.788377046585083, + "learning_rate": 8.146853241813835e-06, + "loss": 3.173, + "step": 14282 + }, + { + "epoch": 1.1450361788699364, + "grad_norm": 1.7562007904052734, + "learning_rate": 8.14557730048406e-06, + "loss": 3.2583, + "step": 14283 + }, + { + "epoch": 1.145116353650959, + "grad_norm": 1.7409179210662842, + "learning_rate": 8.144301390419202e-06, + "loss": 3.2372, + "step": 14284 + }, + { + "epoch": 1.1451965284319816, + "grad_norm": 1.8040980100631714, + "learning_rate": 8.143025511640765e-06, + "loss": 3.24, + "step": 14285 + }, + { + "epoch": 1.1452767032130042, + "grad_norm": 1.6819474697113037, + "learning_rate": 8.141749664170265e-06, + "loss": 3.0836, + "step": 14286 + }, + { + "epoch": 1.1453568779940269, + "grad_norm": 1.8226820230484009, + "learning_rate": 8.140473848029213e-06, + "loss": 3.0962, + "step": 14287 + }, + { + "epoch": 1.1454370527750497, + "grad_norm": 1.7753772735595703, + "learning_rate": 8.139198063239113e-06, + "loss": 3.2263, + "step": 14288 + }, + { + "epoch": 1.1455172275560723, + "grad_norm": 1.879775881767273, + "learning_rate": 8.137922309821481e-06, + "loss": 3.4135, + "step": 14289 + }, + { + "epoch": 1.145597402337095, + "grad_norm": 1.8044954538345337, + "learning_rate": 8.13664658779782e-06, + "loss": 3.1594, + "step": 14290 + }, + { + "epoch": 1.1456775771181176, + "grad_norm": 1.8152108192443848, + "learning_rate": 8.13537089718964e-06, + "loss": 3.2284, + "step": 14291 + }, + { + "epoch": 1.1457577518991402, + "grad_norm": 1.7775541543960571, + "learning_rate": 8.134095238018454e-06, + "loss": 3.2924, + "step": 14292 + }, + { + "epoch": 1.1458379266801628, + "grad_norm": 0.6888429522514343, + "learning_rate": 8.13281961030576e-06, + "loss": 2.4868, + "step": 14293 + }, + { + "epoch": 1.1459181014611854, + "grad_norm": 1.759177565574646, + "learning_rate": 8.131544014073073e-06, + "loss": 3.1431, + "step": 14294 + }, + { + "epoch": 1.145998276242208, + "grad_norm": 1.7414605617523193, + "learning_rate": 8.130268449341893e-06, + "loss": 3.1138, + "step": 14295 + }, + { + "epoch": 1.1460784510232307, + "grad_norm": 1.8132766485214233, + "learning_rate": 8.128992916133728e-06, + "loss": 3.1388, + "step": 14296 + }, + { + "epoch": 1.1461586258042533, + "grad_norm": 1.8800731897354126, + "learning_rate": 8.127717414470084e-06, + "loss": 3.2245, + "step": 14297 + }, + { + "epoch": 1.146238800585276, + "grad_norm": 1.8462539911270142, + "learning_rate": 8.126441944372464e-06, + "loss": 3.2159, + "step": 14298 + }, + { + "epoch": 1.1463189753662986, + "grad_norm": 1.7739545106887817, + "learning_rate": 8.125166505862374e-06, + "loss": 3.2174, + "step": 14299 + }, + { + "epoch": 1.1463991501473212, + "grad_norm": 1.8157984018325806, + "learning_rate": 8.123891098961315e-06, + "loss": 3.2617, + "step": 14300 + }, + { + "epoch": 1.1464793249283438, + "grad_norm": 1.7938060760498047, + "learning_rate": 8.12261572369079e-06, + "loss": 3.1516, + "step": 14301 + }, + { + "epoch": 1.1465594997093664, + "grad_norm": 1.7499631643295288, + "learning_rate": 8.121340380072302e-06, + "loss": 3.1823, + "step": 14302 + }, + { + "epoch": 1.146639674490389, + "grad_norm": 1.740365982055664, + "learning_rate": 8.120065068127354e-06, + "loss": 3.2057, + "step": 14303 + }, + { + "epoch": 1.1467198492714117, + "grad_norm": 1.856270670890808, + "learning_rate": 8.118789787877446e-06, + "loss": 3.1638, + "step": 14304 + }, + { + "epoch": 1.1468000240524343, + "grad_norm": 1.8643680810928345, + "learning_rate": 8.11751453934408e-06, + "loss": 3.2261, + "step": 14305 + }, + { + "epoch": 1.146880198833457, + "grad_norm": 1.7622153759002686, + "learning_rate": 8.116239322548754e-06, + "loss": 3.2355, + "step": 14306 + }, + { + "epoch": 1.1469603736144796, + "grad_norm": 1.7211376428604126, + "learning_rate": 8.11496413751297e-06, + "loss": 3.1605, + "step": 14307 + }, + { + "epoch": 1.1470405483955022, + "grad_norm": 1.7098554372787476, + "learning_rate": 8.113688984258226e-06, + "loss": 3.1317, + "step": 14308 + }, + { + "epoch": 1.1471207231765248, + "grad_norm": 1.7617114782333374, + "learning_rate": 8.112413862806021e-06, + "loss": 3.1161, + "step": 14309 + }, + { + "epoch": 1.1472008979575474, + "grad_norm": 1.7714884281158447, + "learning_rate": 8.111138773177854e-06, + "loss": 3.2758, + "step": 14310 + }, + { + "epoch": 1.14728107273857, + "grad_norm": 1.7582788467407227, + "learning_rate": 8.10986371539522e-06, + "loss": 3.1266, + "step": 14311 + }, + { + "epoch": 1.1473612475195927, + "grad_norm": 1.7843983173370361, + "learning_rate": 8.10858868947962e-06, + "loss": 3.0983, + "step": 14312 + }, + { + "epoch": 1.1474414223006153, + "grad_norm": 1.7441834211349487, + "learning_rate": 8.107313695452546e-06, + "loss": 3.2705, + "step": 14313 + }, + { + "epoch": 1.147521597081638, + "grad_norm": 1.841353416442871, + "learning_rate": 8.106038733335495e-06, + "loss": 3.0655, + "step": 14314 + }, + { + "epoch": 1.1476017718626605, + "grad_norm": 1.789922833442688, + "learning_rate": 8.104763803149967e-06, + "loss": 3.2803, + "step": 14315 + }, + { + "epoch": 1.1476819466436832, + "grad_norm": 1.8345357179641724, + "learning_rate": 8.103488904917451e-06, + "loss": 3.1228, + "step": 14316 + }, + { + "epoch": 1.1477621214247058, + "grad_norm": 1.6834588050842285, + "learning_rate": 8.102214038659444e-06, + "loss": 3.0495, + "step": 14317 + }, + { + "epoch": 1.1478422962057284, + "grad_norm": 1.7470028400421143, + "learning_rate": 8.10093920439744e-06, + "loss": 3.1677, + "step": 14318 + }, + { + "epoch": 1.147922470986751, + "grad_norm": 1.7435197830200195, + "learning_rate": 8.099664402152928e-06, + "loss": 3.2125, + "step": 14319 + }, + { + "epoch": 1.1480026457677737, + "grad_norm": 1.8377782106399536, + "learning_rate": 8.098389631947411e-06, + "loss": 3.2773, + "step": 14320 + }, + { + "epoch": 1.1480828205487963, + "grad_norm": 0.7039839625358582, + "learning_rate": 8.09711489380237e-06, + "loss": 2.4564, + "step": 14321 + }, + { + "epoch": 1.148162995329819, + "grad_norm": 1.822511911392212, + "learning_rate": 8.095840187739304e-06, + "loss": 3.1721, + "step": 14322 + }, + { + "epoch": 1.1482431701108415, + "grad_norm": 1.756613850593567, + "learning_rate": 8.094565513779699e-06, + "loss": 3.2526, + "step": 14323 + }, + { + "epoch": 1.1483233448918644, + "grad_norm": 1.8201541900634766, + "learning_rate": 8.093290871945047e-06, + "loss": 3.2427, + "step": 14324 + }, + { + "epoch": 1.148403519672887, + "grad_norm": 1.7699130773544312, + "learning_rate": 8.092016262256844e-06, + "loss": 3.1138, + "step": 14325 + }, + { + "epoch": 1.1484836944539096, + "grad_norm": 1.7687350511550903, + "learning_rate": 8.09074168473657e-06, + "loss": 3.0564, + "step": 14326 + }, + { + "epoch": 1.1485638692349323, + "grad_norm": 1.7753723859786987, + "learning_rate": 8.08946713940572e-06, + "loss": 3.1349, + "step": 14327 + }, + { + "epoch": 1.1486440440159549, + "grad_norm": 1.831197738647461, + "learning_rate": 8.08819262628578e-06, + "loss": 3.2019, + "step": 14328 + }, + { + "epoch": 1.1487242187969775, + "grad_norm": 0.6882591247558594, + "learning_rate": 8.086918145398236e-06, + "loss": 2.4711, + "step": 14329 + }, + { + "epoch": 1.1488043935780001, + "grad_norm": 1.6596567630767822, + "learning_rate": 8.085643696764583e-06, + "loss": 3.1394, + "step": 14330 + }, + { + "epoch": 1.1488845683590228, + "grad_norm": 1.7517707347869873, + "learning_rate": 8.084369280406298e-06, + "loss": 3.1268, + "step": 14331 + }, + { + "epoch": 1.1489647431400454, + "grad_norm": 0.6643960475921631, + "learning_rate": 8.083094896344875e-06, + "loss": 2.4645, + "step": 14332 + }, + { + "epoch": 1.149044917921068, + "grad_norm": 1.7745038270950317, + "learning_rate": 8.081820544601794e-06, + "loss": 3.1548, + "step": 14333 + }, + { + "epoch": 1.1491250927020906, + "grad_norm": 1.7515405416488647, + "learning_rate": 8.080546225198545e-06, + "loss": 3.125, + "step": 14334 + }, + { + "epoch": 1.1492052674831132, + "grad_norm": 2.035973310470581, + "learning_rate": 8.079271938156609e-06, + "loss": 3.3611, + "step": 14335 + }, + { + "epoch": 1.1492854422641359, + "grad_norm": 1.7832881212234497, + "learning_rate": 8.07799768349747e-06, + "loss": 3.0995, + "step": 14336 + }, + { + "epoch": 1.1493656170451585, + "grad_norm": 1.7675552368164062, + "learning_rate": 8.076723461242616e-06, + "loss": 3.1514, + "step": 14337 + }, + { + "epoch": 1.1494457918261811, + "grad_norm": 1.7319382429122925, + "learning_rate": 8.075449271413524e-06, + "loss": 3.1895, + "step": 14338 + }, + { + "epoch": 1.1495259666072037, + "grad_norm": 1.7578389644622803, + "learning_rate": 8.074175114031678e-06, + "loss": 3.1613, + "step": 14339 + }, + { + "epoch": 1.1496061413882264, + "grad_norm": 1.9041048288345337, + "learning_rate": 8.072900989118564e-06, + "loss": 3.1573, + "step": 14340 + }, + { + "epoch": 1.149686316169249, + "grad_norm": 1.7647312879562378, + "learning_rate": 8.071626896695656e-06, + "loss": 3.1486, + "step": 14341 + }, + { + "epoch": 1.1497664909502716, + "grad_norm": 1.9464043378829956, + "learning_rate": 8.070352836784443e-06, + "loss": 3.192, + "step": 14342 + }, + { + "epoch": 1.1498466657312942, + "grad_norm": 1.7234523296356201, + "learning_rate": 8.0690788094064e-06, + "loss": 3.1914, + "step": 14343 + }, + { + "epoch": 1.1499268405123169, + "grad_norm": 1.7571765184402466, + "learning_rate": 8.067804814583005e-06, + "loss": 3.1286, + "step": 14344 + }, + { + "epoch": 1.1500070152933395, + "grad_norm": 1.8890734910964966, + "learning_rate": 8.066530852335745e-06, + "loss": 3.1549, + "step": 14345 + }, + { + "epoch": 1.150087190074362, + "grad_norm": 1.766879916191101, + "learning_rate": 8.065256922686088e-06, + "loss": 3.2385, + "step": 14346 + }, + { + "epoch": 1.1501673648553847, + "grad_norm": 1.8872637748718262, + "learning_rate": 8.063983025655521e-06, + "loss": 3.2133, + "step": 14347 + }, + { + "epoch": 1.1502475396364074, + "grad_norm": 1.9040837287902832, + "learning_rate": 8.062709161265516e-06, + "loss": 3.2891, + "step": 14348 + }, + { + "epoch": 1.15032771441743, + "grad_norm": 1.8062090873718262, + "learning_rate": 8.061435329537553e-06, + "loss": 3.1567, + "step": 14349 + }, + { + "epoch": 1.1504078891984526, + "grad_norm": 1.8326622247695923, + "learning_rate": 8.060161530493107e-06, + "loss": 3.2442, + "step": 14350 + }, + { + "epoch": 1.1504880639794752, + "grad_norm": 1.8706930875778198, + "learning_rate": 8.058887764153653e-06, + "loss": 3.3, + "step": 14351 + }, + { + "epoch": 1.1505682387604979, + "grad_norm": 1.8342581987380981, + "learning_rate": 8.05761403054067e-06, + "loss": 3.2748, + "step": 14352 + }, + { + "epoch": 1.1506484135415205, + "grad_norm": 2.4987101554870605, + "learning_rate": 8.056340329675627e-06, + "loss": 3.1514, + "step": 14353 + }, + { + "epoch": 1.150728588322543, + "grad_norm": 1.8877301216125488, + "learning_rate": 8.05506666158e-06, + "loss": 3.2089, + "step": 14354 + }, + { + "epoch": 1.1508087631035657, + "grad_norm": 1.7321709394454956, + "learning_rate": 8.053793026275267e-06, + "loss": 3.1068, + "step": 14355 + }, + { + "epoch": 1.1508889378845883, + "grad_norm": 1.7373783588409424, + "learning_rate": 8.052519423782896e-06, + "loss": 3.2149, + "step": 14356 + }, + { + "epoch": 1.150969112665611, + "grad_norm": 1.727044939994812, + "learning_rate": 8.051245854124363e-06, + "loss": 3.1765, + "step": 14357 + }, + { + "epoch": 1.1510492874466336, + "grad_norm": 1.808598518371582, + "learning_rate": 8.049972317321133e-06, + "loss": 3.2428, + "step": 14358 + }, + { + "epoch": 1.1511294622276562, + "grad_norm": 1.7870889902114868, + "learning_rate": 8.048698813394686e-06, + "loss": 3.2103, + "step": 14359 + }, + { + "epoch": 1.1512096370086788, + "grad_norm": 1.7427716255187988, + "learning_rate": 8.047425342366491e-06, + "loss": 3.2325, + "step": 14360 + }, + { + "epoch": 1.1512898117897015, + "grad_norm": 1.7976666688919067, + "learning_rate": 8.046151904258012e-06, + "loss": 3.2047, + "step": 14361 + }, + { + "epoch": 1.151369986570724, + "grad_norm": 1.8278793096542358, + "learning_rate": 8.044878499090728e-06, + "loss": 3.2228, + "step": 14362 + }, + { + "epoch": 1.1514501613517467, + "grad_norm": 1.851715326309204, + "learning_rate": 8.0436051268861e-06, + "loss": 3.1659, + "step": 14363 + }, + { + "epoch": 1.1515303361327693, + "grad_norm": 1.852653980255127, + "learning_rate": 8.042331787665601e-06, + "loss": 3.1774, + "step": 14364 + }, + { + "epoch": 1.151610510913792, + "grad_norm": 1.796020269393921, + "learning_rate": 8.0410584814507e-06, + "loss": 3.1663, + "step": 14365 + }, + { + "epoch": 1.1516906856948146, + "grad_norm": 1.722334384918213, + "learning_rate": 8.03978520826286e-06, + "loss": 3.2135, + "step": 14366 + }, + { + "epoch": 1.1517708604758372, + "grad_norm": 1.7844164371490479, + "learning_rate": 8.038511968123554e-06, + "loss": 3.2332, + "step": 14367 + }, + { + "epoch": 1.1518510352568598, + "grad_norm": 1.8634308576583862, + "learning_rate": 8.037238761054242e-06, + "loss": 3.296, + "step": 14368 + }, + { + "epoch": 1.1519312100378827, + "grad_norm": 1.7542030811309814, + "learning_rate": 8.035965587076393e-06, + "loss": 3.2, + "step": 14369 + }, + { + "epoch": 1.1520113848189053, + "grad_norm": 1.6536234617233276, + "learning_rate": 8.034692446211472e-06, + "loss": 3.1057, + "step": 14370 + }, + { + "epoch": 1.152091559599928, + "grad_norm": 1.7392750978469849, + "learning_rate": 8.033419338480944e-06, + "loss": 3.176, + "step": 14371 + }, + { + "epoch": 1.1521717343809506, + "grad_norm": 1.819169282913208, + "learning_rate": 8.032146263906272e-06, + "loss": 3.1304, + "step": 14372 + }, + { + "epoch": 1.1522519091619732, + "grad_norm": 1.7537939548492432, + "learning_rate": 8.030873222508922e-06, + "loss": 3.1422, + "step": 14373 + }, + { + "epoch": 1.1523320839429958, + "grad_norm": 1.6704692840576172, + "learning_rate": 8.029600214310354e-06, + "loss": 3.2821, + "step": 14374 + }, + { + "epoch": 1.1524122587240184, + "grad_norm": 2.0041754245758057, + "learning_rate": 8.028327239332034e-06, + "loss": 3.1598, + "step": 14375 + }, + { + "epoch": 1.152492433505041, + "grad_norm": 1.7895959615707397, + "learning_rate": 8.02705429759542e-06, + "loss": 3.1561, + "step": 14376 + }, + { + "epoch": 1.1525726082860637, + "grad_norm": 1.7327560186386108, + "learning_rate": 8.025781389121977e-06, + "loss": 3.1772, + "step": 14377 + }, + { + "epoch": 1.1526527830670863, + "grad_norm": 0.7266795039176941, + "learning_rate": 8.024508513933163e-06, + "loss": 2.4978, + "step": 14378 + }, + { + "epoch": 1.152732957848109, + "grad_norm": 0.7248619794845581, + "learning_rate": 8.023235672050439e-06, + "loss": 2.4832, + "step": 14379 + }, + { + "epoch": 1.1528131326291315, + "grad_norm": 1.8044159412384033, + "learning_rate": 8.021962863495264e-06, + "loss": 3.1863, + "step": 14380 + }, + { + "epoch": 1.1528933074101542, + "grad_norm": 1.7888511419296265, + "learning_rate": 8.020690088289099e-06, + "loss": 3.1688, + "step": 14381 + }, + { + "epoch": 1.1529734821911768, + "grad_norm": 1.7628997564315796, + "learning_rate": 8.019417346453404e-06, + "loss": 3.1229, + "step": 14382 + }, + { + "epoch": 1.1530536569721994, + "grad_norm": 1.8672829866409302, + "learning_rate": 8.01814463800963e-06, + "loss": 3.2275, + "step": 14383 + }, + { + "epoch": 1.153133831753222, + "grad_norm": 1.7923812866210938, + "learning_rate": 8.01687196297924e-06, + "loss": 3.3339, + "step": 14384 + }, + { + "epoch": 1.1532140065342447, + "grad_norm": 1.7639405727386475, + "learning_rate": 8.015599321383693e-06, + "loss": 3.1393, + "step": 14385 + }, + { + "epoch": 1.1532941813152673, + "grad_norm": 1.8911203145980835, + "learning_rate": 8.014326713244439e-06, + "loss": 3.2224, + "step": 14386 + }, + { + "epoch": 1.15337435609629, + "grad_norm": 0.6942532062530518, + "learning_rate": 8.013054138582939e-06, + "loss": 2.4844, + "step": 14387 + }, + { + "epoch": 1.1534545308773125, + "grad_norm": 1.8654216527938843, + "learning_rate": 8.011781597420642e-06, + "loss": 3.2264, + "step": 14388 + }, + { + "epoch": 1.1535347056583352, + "grad_norm": 1.7505016326904297, + "learning_rate": 8.010509089779008e-06, + "loss": 3.0989, + "step": 14389 + }, + { + "epoch": 1.1536148804393578, + "grad_norm": 1.7651294469833374, + "learning_rate": 8.009236615679494e-06, + "loss": 3.2351, + "step": 14390 + }, + { + "epoch": 1.1536950552203804, + "grad_norm": 1.878178596496582, + "learning_rate": 8.007964175143544e-06, + "loss": 3.3214, + "step": 14391 + }, + { + "epoch": 1.153775230001403, + "grad_norm": 1.733783483505249, + "learning_rate": 8.006691768192618e-06, + "loss": 3.124, + "step": 14392 + }, + { + "epoch": 1.1538554047824257, + "grad_norm": 1.7956565618515015, + "learning_rate": 8.005419394848166e-06, + "loss": 3.2162, + "step": 14393 + }, + { + "epoch": 1.1539355795634483, + "grad_norm": 1.75580894947052, + "learning_rate": 8.004147055131639e-06, + "loss": 3.2369, + "step": 14394 + }, + { + "epoch": 1.154015754344471, + "grad_norm": 1.7164803743362427, + "learning_rate": 8.002874749064491e-06, + "loss": 3.2252, + "step": 14395 + }, + { + "epoch": 1.1540959291254935, + "grad_norm": 1.841597318649292, + "learning_rate": 8.001602476668169e-06, + "loss": 3.1144, + "step": 14396 + }, + { + "epoch": 1.1541761039065161, + "grad_norm": 1.6571139097213745, + "learning_rate": 8.000330237964129e-06, + "loss": 3.1931, + "step": 14397 + }, + { + "epoch": 1.1542562786875388, + "grad_norm": 1.8417115211486816, + "learning_rate": 7.999058032973811e-06, + "loss": 3.2373, + "step": 14398 + }, + { + "epoch": 1.1543364534685614, + "grad_norm": 1.854029893875122, + "learning_rate": 7.997785861718673e-06, + "loss": 3.1334, + "step": 14399 + }, + { + "epoch": 1.154416628249584, + "grad_norm": 1.7873249053955078, + "learning_rate": 7.996513724220159e-06, + "loss": 3.1021, + "step": 14400 + }, + { + "epoch": 1.1544968030306066, + "grad_norm": 1.7820823192596436, + "learning_rate": 7.995241620499716e-06, + "loss": 3.1656, + "step": 14401 + }, + { + "epoch": 1.1545769778116293, + "grad_norm": 1.7393521070480347, + "learning_rate": 7.993969550578795e-06, + "loss": 3.1153, + "step": 14402 + }, + { + "epoch": 1.154657152592652, + "grad_norm": 1.6849921941757202, + "learning_rate": 7.99269751447884e-06, + "loss": 3.2189, + "step": 14403 + }, + { + "epoch": 1.1547373273736745, + "grad_norm": 1.742214322090149, + "learning_rate": 7.991425512221295e-06, + "loss": 3.2303, + "step": 14404 + }, + { + "epoch": 1.1548175021546974, + "grad_norm": 1.7936739921569824, + "learning_rate": 7.99015354382761e-06, + "loss": 3.2394, + "step": 14405 + }, + { + "epoch": 1.15489767693572, + "grad_norm": 1.8046109676361084, + "learning_rate": 7.988881609319226e-06, + "loss": 3.205, + "step": 14406 + }, + { + "epoch": 1.1549778517167426, + "grad_norm": 1.8437014818191528, + "learning_rate": 7.987609708717592e-06, + "loss": 3.0966, + "step": 14407 + }, + { + "epoch": 1.1550580264977652, + "grad_norm": 1.6124267578125, + "learning_rate": 7.986337842044148e-06, + "loss": 3.1459, + "step": 14408 + }, + { + "epoch": 1.1551382012787879, + "grad_norm": 1.896569848060608, + "learning_rate": 7.985066009320338e-06, + "loss": 3.2332, + "step": 14409 + }, + { + "epoch": 1.1552183760598105, + "grad_norm": 0.7697740197181702, + "learning_rate": 7.983794210567607e-06, + "loss": 2.4955, + "step": 14410 + }, + { + "epoch": 1.155298550840833, + "grad_norm": 1.7813681364059448, + "learning_rate": 7.982522445807391e-06, + "loss": 3.1494, + "step": 14411 + }, + { + "epoch": 1.1553787256218557, + "grad_norm": 1.8332031965255737, + "learning_rate": 7.981250715061139e-06, + "loss": 3.0731, + "step": 14412 + }, + { + "epoch": 1.1554589004028784, + "grad_norm": 1.7715872526168823, + "learning_rate": 7.979979018350287e-06, + "loss": 3.1608, + "step": 14413 + }, + { + "epoch": 1.155539075183901, + "grad_norm": 1.9218147993087769, + "learning_rate": 7.978707355696276e-06, + "loss": 3.2029, + "step": 14414 + }, + { + "epoch": 1.1556192499649236, + "grad_norm": 1.803027629852295, + "learning_rate": 7.97743572712055e-06, + "loss": 3.2646, + "step": 14415 + }, + { + "epoch": 1.1556994247459462, + "grad_norm": 1.8611925840377808, + "learning_rate": 7.976164132644541e-06, + "loss": 3.265, + "step": 14416 + }, + { + "epoch": 1.1557795995269688, + "grad_norm": 2.021479606628418, + "learning_rate": 7.974892572289695e-06, + "loss": 3.1639, + "step": 14417 + }, + { + "epoch": 1.1558597743079915, + "grad_norm": 1.899489402770996, + "learning_rate": 7.973621046077442e-06, + "loss": 3.1866, + "step": 14418 + }, + { + "epoch": 1.155939949089014, + "grad_norm": 1.862202763557434, + "learning_rate": 7.972349554029227e-06, + "loss": 3.1932, + "step": 14419 + }, + { + "epoch": 1.1560201238700367, + "grad_norm": 1.7197799682617188, + "learning_rate": 7.971078096166483e-06, + "loss": 3.147, + "step": 14420 + }, + { + "epoch": 1.1561002986510593, + "grad_norm": 1.7605868577957153, + "learning_rate": 7.969806672510647e-06, + "loss": 3.1758, + "step": 14421 + }, + { + "epoch": 1.156180473432082, + "grad_norm": 1.7384185791015625, + "learning_rate": 7.968535283083157e-06, + "loss": 3.11, + "step": 14422 + }, + { + "epoch": 1.1562606482131046, + "grad_norm": 1.880338191986084, + "learning_rate": 7.967263927905444e-06, + "loss": 3.2382, + "step": 14423 + }, + { + "epoch": 1.1563408229941272, + "grad_norm": 1.8954185247421265, + "learning_rate": 7.965992606998944e-06, + "loss": 3.2485, + "step": 14424 + }, + { + "epoch": 1.1564209977751498, + "grad_norm": 1.8601346015930176, + "learning_rate": 7.964721320385095e-06, + "loss": 3.2229, + "step": 14425 + }, + { + "epoch": 1.1565011725561725, + "grad_norm": 1.7148720026016235, + "learning_rate": 7.963450068085326e-06, + "loss": 3.1851, + "step": 14426 + }, + { + "epoch": 1.156581347337195, + "grad_norm": 1.799584984779358, + "learning_rate": 7.962178850121071e-06, + "loss": 3.1893, + "step": 14427 + }, + { + "epoch": 1.1566615221182177, + "grad_norm": 1.7743421792984009, + "learning_rate": 7.960907666513763e-06, + "loss": 3.1465, + "step": 14428 + }, + { + "epoch": 1.1567416968992403, + "grad_norm": 1.778228998184204, + "learning_rate": 7.95963651728483e-06, + "loss": 3.2559, + "step": 14429 + }, + { + "epoch": 1.156821871680263, + "grad_norm": 1.9110981225967407, + "learning_rate": 7.958365402455708e-06, + "loss": 3.2067, + "step": 14430 + }, + { + "epoch": 1.1569020464612856, + "grad_norm": 1.7365055084228516, + "learning_rate": 7.957094322047828e-06, + "loss": 3.1882, + "step": 14431 + }, + { + "epoch": 1.1569822212423082, + "grad_norm": 1.7486108541488647, + "learning_rate": 7.955823276082616e-06, + "loss": 3.0795, + "step": 14432 + }, + { + "epoch": 1.1570623960233308, + "grad_norm": 1.7736603021621704, + "learning_rate": 7.954552264581502e-06, + "loss": 3.1765, + "step": 14433 + }, + { + "epoch": 1.1571425708043535, + "grad_norm": 0.7016701102256775, + "learning_rate": 7.953281287565916e-06, + "loss": 2.4808, + "step": 14434 + }, + { + "epoch": 1.157222745585376, + "grad_norm": 1.7827035188674927, + "learning_rate": 7.952010345057289e-06, + "loss": 3.0972, + "step": 14435 + }, + { + "epoch": 1.1573029203663987, + "grad_norm": 1.7805638313293457, + "learning_rate": 7.950739437077042e-06, + "loss": 3.2507, + "step": 14436 + }, + { + "epoch": 1.1573830951474213, + "grad_norm": 1.809086799621582, + "learning_rate": 7.949468563646609e-06, + "loss": 3.233, + "step": 14437 + }, + { + "epoch": 1.157463269928444, + "grad_norm": 1.9231789112091064, + "learning_rate": 7.948197724787412e-06, + "loss": 3.2065, + "step": 14438 + }, + { + "epoch": 1.1575434447094666, + "grad_norm": 1.750206470489502, + "learning_rate": 7.946926920520876e-06, + "loss": 3.2241, + "step": 14439 + }, + { + "epoch": 1.1576236194904892, + "grad_norm": 1.81095290184021, + "learning_rate": 7.945656150868432e-06, + "loss": 3.129, + "step": 14440 + }, + { + "epoch": 1.1577037942715118, + "grad_norm": 1.7478036880493164, + "learning_rate": 7.944385415851502e-06, + "loss": 3.1147, + "step": 14441 + }, + { + "epoch": 1.1577839690525344, + "grad_norm": 1.8756356239318848, + "learning_rate": 7.943114715491507e-06, + "loss": 3.2497, + "step": 14442 + }, + { + "epoch": 1.157864143833557, + "grad_norm": 0.7135317325592041, + "learning_rate": 7.941844049809872e-06, + "loss": 2.4706, + "step": 14443 + }, + { + "epoch": 1.1579443186145797, + "grad_norm": 1.8055717945098877, + "learning_rate": 7.940573418828022e-06, + "loss": 3.3435, + "step": 14444 + }, + { + "epoch": 1.1580244933956023, + "grad_norm": 1.7041066884994507, + "learning_rate": 7.93930282256738e-06, + "loss": 3.2246, + "step": 14445 + }, + { + "epoch": 1.158104668176625, + "grad_norm": 1.9272267818450928, + "learning_rate": 7.938032261049364e-06, + "loss": 3.2062, + "step": 14446 + }, + { + "epoch": 1.1581848429576476, + "grad_norm": 1.7081146240234375, + "learning_rate": 7.936761734295401e-06, + "loss": 3.1441, + "step": 14447 + }, + { + "epoch": 1.1582650177386702, + "grad_norm": 1.771190881729126, + "learning_rate": 7.935491242326903e-06, + "loss": 3.2284, + "step": 14448 + }, + { + "epoch": 1.1583451925196928, + "grad_norm": 1.7678242921829224, + "learning_rate": 7.934220785165297e-06, + "loss": 3.1632, + "step": 14449 + }, + { + "epoch": 1.1584253673007154, + "grad_norm": 1.8670789003372192, + "learning_rate": 7.932950362832003e-06, + "loss": 3.1666, + "step": 14450 + }, + { + "epoch": 1.1585055420817383, + "grad_norm": 0.6734486818313599, + "learning_rate": 7.931679975348433e-06, + "loss": 2.4608, + "step": 14451 + }, + { + "epoch": 1.158585716862761, + "grad_norm": 1.8346818685531616, + "learning_rate": 7.930409622736015e-06, + "loss": 3.2418, + "step": 14452 + }, + { + "epoch": 1.1586658916437835, + "grad_norm": 1.758680820465088, + "learning_rate": 7.929139305016156e-06, + "loss": 3.1175, + "step": 14453 + }, + { + "epoch": 1.1587460664248062, + "grad_norm": 1.7872954607009888, + "learning_rate": 7.927869022210282e-06, + "loss": 3.1324, + "step": 14454 + }, + { + "epoch": 1.1588262412058288, + "grad_norm": 1.826357364654541, + "learning_rate": 7.926598774339805e-06, + "loss": 3.2642, + "step": 14455 + }, + { + "epoch": 1.1589064159868514, + "grad_norm": 1.7848529815673828, + "learning_rate": 7.925328561426141e-06, + "loss": 3.2107, + "step": 14456 + }, + { + "epoch": 1.158986590767874, + "grad_norm": 1.8876749277114868, + "learning_rate": 7.924058383490709e-06, + "loss": 3.2654, + "step": 14457 + }, + { + "epoch": 1.1590667655488966, + "grad_norm": 1.7430291175842285, + "learning_rate": 7.922788240554918e-06, + "loss": 3.1769, + "step": 14458 + }, + { + "epoch": 1.1591469403299193, + "grad_norm": 1.8841792345046997, + "learning_rate": 7.921518132640185e-06, + "loss": 3.3004, + "step": 14459 + }, + { + "epoch": 1.159227115110942, + "grad_norm": 1.7687650918960571, + "learning_rate": 7.920248059767926e-06, + "loss": 3.1564, + "step": 14460 + }, + { + "epoch": 1.1593072898919645, + "grad_norm": 1.7807860374450684, + "learning_rate": 7.91897802195955e-06, + "loss": 3.2227, + "step": 14461 + }, + { + "epoch": 1.1593874646729871, + "grad_norm": 0.7586305737495422, + "learning_rate": 7.917708019236473e-06, + "loss": 2.4707, + "step": 14462 + }, + { + "epoch": 1.1594676394540098, + "grad_norm": 1.708522915840149, + "learning_rate": 7.916438051620104e-06, + "loss": 3.2151, + "step": 14463 + }, + { + "epoch": 1.1595478142350324, + "grad_norm": 1.8869857788085938, + "learning_rate": 7.915168119131853e-06, + "loss": 3.2351, + "step": 14464 + }, + { + "epoch": 1.159627989016055, + "grad_norm": 1.7812881469726562, + "learning_rate": 7.913898221793136e-06, + "loss": 3.1694, + "step": 14465 + }, + { + "epoch": 1.1597081637970776, + "grad_norm": 1.7464500665664673, + "learning_rate": 7.912628359625359e-06, + "loss": 3.2479, + "step": 14466 + }, + { + "epoch": 1.1597883385781003, + "grad_norm": 1.7896801233291626, + "learning_rate": 7.911358532649932e-06, + "loss": 3.1583, + "step": 14467 + }, + { + "epoch": 1.1598685133591229, + "grad_norm": 1.7050689458847046, + "learning_rate": 7.910088740888262e-06, + "loss": 3.0783, + "step": 14468 + }, + { + "epoch": 1.1599486881401455, + "grad_norm": 1.7068501710891724, + "learning_rate": 7.90881898436176e-06, + "loss": 3.1826, + "step": 14469 + }, + { + "epoch": 1.1600288629211681, + "grad_norm": 0.7015902996063232, + "learning_rate": 7.907549263091836e-06, + "loss": 2.4772, + "step": 14470 + }, + { + "epoch": 1.1601090377021908, + "grad_norm": 1.7111783027648926, + "learning_rate": 7.90627957709989e-06, + "loss": 3.1317, + "step": 14471 + }, + { + "epoch": 1.1601892124832134, + "grad_norm": 1.68667471408844, + "learning_rate": 7.905009926407337e-06, + "loss": 3.1167, + "step": 14472 + }, + { + "epoch": 1.160269387264236, + "grad_norm": 1.8957972526550293, + "learning_rate": 7.903740311035573e-06, + "loss": 3.1385, + "step": 14473 + }, + { + "epoch": 1.1603495620452586, + "grad_norm": 1.8662832975387573, + "learning_rate": 7.90247073100601e-06, + "loss": 3.2537, + "step": 14474 + }, + { + "epoch": 1.1604297368262813, + "grad_norm": 2.0804057121276855, + "learning_rate": 7.901201186340055e-06, + "loss": 3.2474, + "step": 14475 + }, + { + "epoch": 1.1605099116073039, + "grad_norm": 1.7962785959243774, + "learning_rate": 7.899931677059105e-06, + "loss": 3.139, + "step": 14476 + }, + { + "epoch": 1.1605900863883265, + "grad_norm": 2.207329750061035, + "learning_rate": 7.898662203184568e-06, + "loss": 3.2704, + "step": 14477 + }, + { + "epoch": 1.1606702611693491, + "grad_norm": 1.8014429807662964, + "learning_rate": 7.897392764737845e-06, + "loss": 3.1854, + "step": 14478 + }, + { + "epoch": 1.1607504359503718, + "grad_norm": 1.7542377710342407, + "learning_rate": 7.896123361740338e-06, + "loss": 3.1548, + "step": 14479 + }, + { + "epoch": 1.1608306107313944, + "grad_norm": 1.6818550825119019, + "learning_rate": 7.894853994213452e-06, + "loss": 3.0583, + "step": 14480 + }, + { + "epoch": 1.160910785512417, + "grad_norm": 1.797285556793213, + "learning_rate": 7.893584662178582e-06, + "loss": 3.1464, + "step": 14481 + }, + { + "epoch": 1.1609909602934396, + "grad_norm": 1.8081905841827393, + "learning_rate": 7.892315365657137e-06, + "loss": 3.2053, + "step": 14482 + }, + { + "epoch": 1.1610711350744622, + "grad_norm": 1.74184250831604, + "learning_rate": 7.89104610467051e-06, + "loss": 3.172, + "step": 14483 + }, + { + "epoch": 1.1611513098554849, + "grad_norm": 1.9153779745101929, + "learning_rate": 7.8897768792401e-06, + "loss": 3.1205, + "step": 14484 + }, + { + "epoch": 1.1612314846365075, + "grad_norm": 1.6711437702178955, + "learning_rate": 7.888507689387312e-06, + "loss": 3.1248, + "step": 14485 + }, + { + "epoch": 1.1613116594175303, + "grad_norm": 1.751611590385437, + "learning_rate": 7.887238535133537e-06, + "loss": 3.1445, + "step": 14486 + }, + { + "epoch": 1.161391834198553, + "grad_norm": 1.8303985595703125, + "learning_rate": 7.885969416500178e-06, + "loss": 3.1744, + "step": 14487 + }, + { + "epoch": 1.1614720089795756, + "grad_norm": 1.8010766506195068, + "learning_rate": 7.88470033350863e-06, + "loss": 3.1919, + "step": 14488 + }, + { + "epoch": 1.1615521837605982, + "grad_norm": 1.7999259233474731, + "learning_rate": 7.883431286180286e-06, + "loss": 3.2294, + "step": 14489 + }, + { + "epoch": 1.1616323585416208, + "grad_norm": 1.7432405948638916, + "learning_rate": 7.882162274536546e-06, + "loss": 3.1144, + "step": 14490 + }, + { + "epoch": 1.1617125333226435, + "grad_norm": 1.8560197353363037, + "learning_rate": 7.880893298598804e-06, + "loss": 3.1467, + "step": 14491 + }, + { + "epoch": 1.161792708103666, + "grad_norm": 1.7703272104263306, + "learning_rate": 7.879624358388456e-06, + "loss": 3.2244, + "step": 14492 + }, + { + "epoch": 1.1618728828846887, + "grad_norm": 1.7894244194030762, + "learning_rate": 7.87835545392689e-06, + "loss": 3.0966, + "step": 14493 + }, + { + "epoch": 1.1619530576657113, + "grad_norm": 1.867174744606018, + "learning_rate": 7.877086585235505e-06, + "loss": 3.2728, + "step": 14494 + }, + { + "epoch": 1.162033232446734, + "grad_norm": 1.8458008766174316, + "learning_rate": 7.875817752335694e-06, + "loss": 3.0995, + "step": 14495 + }, + { + "epoch": 1.1621134072277566, + "grad_norm": 1.7795026302337646, + "learning_rate": 7.874548955248843e-06, + "loss": 3.1697, + "step": 14496 + }, + { + "epoch": 1.1621935820087792, + "grad_norm": 1.830120325088501, + "learning_rate": 7.873280193996352e-06, + "loss": 3.1632, + "step": 14497 + }, + { + "epoch": 1.1622737567898018, + "grad_norm": 1.686043381690979, + "learning_rate": 7.872011468599606e-06, + "loss": 3.2847, + "step": 14498 + }, + { + "epoch": 1.1623539315708245, + "grad_norm": 1.7470568418502808, + "learning_rate": 7.870742779079994e-06, + "loss": 3.1459, + "step": 14499 + }, + { + "epoch": 1.162434106351847, + "grad_norm": 1.7585597038269043, + "learning_rate": 7.869474125458912e-06, + "loss": 3.2355, + "step": 14500 + }, + { + "epoch": 1.1625142811328697, + "grad_norm": 1.8163208961486816, + "learning_rate": 7.868205507757743e-06, + "loss": 3.2357, + "step": 14501 + }, + { + "epoch": 1.1625944559138923, + "grad_norm": 1.7863538265228271, + "learning_rate": 7.866936925997876e-06, + "loss": 3.1927, + "step": 14502 + }, + { + "epoch": 1.162674630694915, + "grad_norm": 1.6988173723220825, + "learning_rate": 7.865668380200705e-06, + "loss": 3.2017, + "step": 14503 + }, + { + "epoch": 1.1627548054759376, + "grad_norm": 1.7117929458618164, + "learning_rate": 7.86439987038761e-06, + "loss": 3.2787, + "step": 14504 + }, + { + "epoch": 1.1628349802569602, + "grad_norm": 1.807127594947815, + "learning_rate": 7.863131396579983e-06, + "loss": 3.2031, + "step": 14505 + }, + { + "epoch": 1.1629151550379828, + "grad_norm": 1.8414124250411987, + "learning_rate": 7.861862958799205e-06, + "loss": 3.2772, + "step": 14506 + }, + { + "epoch": 1.1629953298190054, + "grad_norm": 1.8518483638763428, + "learning_rate": 7.860594557066663e-06, + "loss": 3.1526, + "step": 14507 + }, + { + "epoch": 1.163075504600028, + "grad_norm": 1.8527050018310547, + "learning_rate": 7.859326191403748e-06, + "loss": 3.197, + "step": 14508 + }, + { + "epoch": 1.1631556793810507, + "grad_norm": 1.8844175338745117, + "learning_rate": 7.858057861831835e-06, + "loss": 3.1559, + "step": 14509 + }, + { + "epoch": 1.1632358541620733, + "grad_norm": 0.6779905557632446, + "learning_rate": 7.856789568372313e-06, + "loss": 2.4708, + "step": 14510 + }, + { + "epoch": 1.163316028943096, + "grad_norm": 1.7979702949523926, + "learning_rate": 7.855521311046562e-06, + "loss": 3.1304, + "step": 14511 + }, + { + "epoch": 1.1633962037241186, + "grad_norm": 1.7505874633789062, + "learning_rate": 7.854253089875966e-06, + "loss": 3.2687, + "step": 14512 + }, + { + "epoch": 1.1634763785051412, + "grad_norm": 0.7155869007110596, + "learning_rate": 7.852984904881908e-06, + "loss": 2.4904, + "step": 14513 + }, + { + "epoch": 1.1635565532861638, + "grad_norm": 1.768883228302002, + "learning_rate": 7.851716756085768e-06, + "loss": 3.1906, + "step": 14514 + }, + { + "epoch": 1.1636367280671864, + "grad_norm": 1.836230754852295, + "learning_rate": 7.850448643508926e-06, + "loss": 3.2364, + "step": 14515 + }, + { + "epoch": 1.163716902848209, + "grad_norm": 1.7438106536865234, + "learning_rate": 7.849180567172761e-06, + "loss": 3.2329, + "step": 14516 + }, + { + "epoch": 1.1637970776292317, + "grad_norm": 1.9115959405899048, + "learning_rate": 7.847912527098655e-06, + "loss": 3.1841, + "step": 14517 + }, + { + "epoch": 1.1638772524102543, + "grad_norm": 1.6705124378204346, + "learning_rate": 7.846644523307984e-06, + "loss": 3.1637, + "step": 14518 + }, + { + "epoch": 1.163957427191277, + "grad_norm": 1.7865698337554932, + "learning_rate": 7.845376555822128e-06, + "loss": 3.167, + "step": 14519 + }, + { + "epoch": 1.1640376019722996, + "grad_norm": 0.6765446066856384, + "learning_rate": 7.844108624662467e-06, + "loss": 2.474, + "step": 14520 + }, + { + "epoch": 1.1641177767533222, + "grad_norm": 1.8202918767929077, + "learning_rate": 7.84284072985037e-06, + "loss": 3.1421, + "step": 14521 + }, + { + "epoch": 1.1641979515343448, + "grad_norm": 1.716404914855957, + "learning_rate": 7.84157287140722e-06, + "loss": 3.1219, + "step": 14522 + }, + { + "epoch": 1.1642781263153674, + "grad_norm": 1.7207597494125366, + "learning_rate": 7.840305049354391e-06, + "loss": 3.1069, + "step": 14523 + }, + { + "epoch": 1.16435830109639, + "grad_norm": 1.7529840469360352, + "learning_rate": 7.839037263713258e-06, + "loss": 3.199, + "step": 14524 + }, + { + "epoch": 1.1644384758774127, + "grad_norm": 1.8804563283920288, + "learning_rate": 7.837769514505195e-06, + "loss": 3.1981, + "step": 14525 + }, + { + "epoch": 1.1645186506584353, + "grad_norm": 1.8024052381515503, + "learning_rate": 7.836501801751575e-06, + "loss": 3.192, + "step": 14526 + }, + { + "epoch": 1.164598825439458, + "grad_norm": 1.7823243141174316, + "learning_rate": 7.835234125473773e-06, + "loss": 3.1473, + "step": 14527 + }, + { + "epoch": 1.1646790002204805, + "grad_norm": 0.6823776364326477, + "learning_rate": 7.83396648569316e-06, + "loss": 2.4839, + "step": 14528 + }, + { + "epoch": 1.1647591750015032, + "grad_norm": 0.6934969425201416, + "learning_rate": 7.832698882431108e-06, + "loss": 2.4917, + "step": 14529 + }, + { + "epoch": 1.1648393497825258, + "grad_norm": 1.72166907787323, + "learning_rate": 7.831431315708992e-06, + "loss": 3.1921, + "step": 14530 + }, + { + "epoch": 1.1649195245635484, + "grad_norm": 1.7686916589736938, + "learning_rate": 7.830163785548178e-06, + "loss": 3.1345, + "step": 14531 + }, + { + "epoch": 1.1649996993445713, + "grad_norm": 1.8512990474700928, + "learning_rate": 7.828896291970036e-06, + "loss": 3.2566, + "step": 14532 + }, + { + "epoch": 1.1650798741255939, + "grad_norm": 1.9203920364379883, + "learning_rate": 7.82762883499594e-06, + "loss": 3.2067, + "step": 14533 + }, + { + "epoch": 1.1651600489066165, + "grad_norm": 1.769990086555481, + "learning_rate": 7.826361414647254e-06, + "loss": 3.181, + "step": 14534 + }, + { + "epoch": 1.1652402236876391, + "grad_norm": 1.743865966796875, + "learning_rate": 7.82509403094535e-06, + "loss": 3.075, + "step": 14535 + }, + { + "epoch": 1.1653203984686618, + "grad_norm": 1.8792188167572021, + "learning_rate": 7.823826683911594e-06, + "loss": 3.151, + "step": 14536 + }, + { + "epoch": 1.1654005732496844, + "grad_norm": 1.713590145111084, + "learning_rate": 7.822559373567352e-06, + "loss": 3.2651, + "step": 14537 + }, + { + "epoch": 1.165480748030707, + "grad_norm": 1.768495798110962, + "learning_rate": 7.821292099933993e-06, + "loss": 3.2072, + "step": 14538 + }, + { + "epoch": 1.1655609228117296, + "grad_norm": 1.7615501880645752, + "learning_rate": 7.82002486303288e-06, + "loss": 3.1251, + "step": 14539 + }, + { + "epoch": 1.1656410975927523, + "grad_norm": 1.8586243391036987, + "learning_rate": 7.818757662885382e-06, + "loss": 3.1498, + "step": 14540 + }, + { + "epoch": 1.1657212723737749, + "grad_norm": 1.7665692567825317, + "learning_rate": 7.81749049951286e-06, + "loss": 3.2708, + "step": 14541 + }, + { + "epoch": 1.1658014471547975, + "grad_norm": 1.7384052276611328, + "learning_rate": 7.816223372936676e-06, + "loss": 3.2527, + "step": 14542 + }, + { + "epoch": 1.1658816219358201, + "grad_norm": 1.783200979232788, + "learning_rate": 7.8149562831782e-06, + "loss": 3.2531, + "step": 14543 + }, + { + "epoch": 1.1659617967168427, + "grad_norm": 1.7079648971557617, + "learning_rate": 7.81368923025879e-06, + "loss": 3.1539, + "step": 14544 + }, + { + "epoch": 1.1660419714978654, + "grad_norm": 1.7640646696090698, + "learning_rate": 7.81242221419981e-06, + "loss": 3.1496, + "step": 14545 + }, + { + "epoch": 1.166122146278888, + "grad_norm": 1.8755700588226318, + "learning_rate": 7.811155235022618e-06, + "loss": 3.225, + "step": 14546 + }, + { + "epoch": 1.1662023210599106, + "grad_norm": 1.7124991416931152, + "learning_rate": 7.809888292748578e-06, + "loss": 3.1191, + "step": 14547 + }, + { + "epoch": 1.1662824958409332, + "grad_norm": 1.7524667978286743, + "learning_rate": 7.80862138739905e-06, + "loss": 3.1635, + "step": 14548 + }, + { + "epoch": 1.1663626706219559, + "grad_norm": 1.915185570716858, + "learning_rate": 7.807354518995393e-06, + "loss": 3.2522, + "step": 14549 + }, + { + "epoch": 1.1664428454029785, + "grad_norm": 1.7108826637268066, + "learning_rate": 7.806087687558967e-06, + "loss": 3.1469, + "step": 14550 + }, + { + "epoch": 1.1665230201840011, + "grad_norm": 1.745369553565979, + "learning_rate": 7.804820893111128e-06, + "loss": 3.2056, + "step": 14551 + }, + { + "epoch": 1.1666031949650237, + "grad_norm": 1.7612271308898926, + "learning_rate": 7.803554135673235e-06, + "loss": 3.1066, + "step": 14552 + }, + { + "epoch": 1.1666833697460464, + "grad_norm": 1.7925364971160889, + "learning_rate": 7.802287415266646e-06, + "loss": 3.2228, + "step": 14553 + }, + { + "epoch": 1.166763544527069, + "grad_norm": 1.8352051973342896, + "learning_rate": 7.801020731912716e-06, + "loss": 3.245, + "step": 14554 + }, + { + "epoch": 1.1668437193080916, + "grad_norm": 1.7648297548294067, + "learning_rate": 7.799754085632804e-06, + "loss": 3.1658, + "step": 14555 + }, + { + "epoch": 1.1669238940891142, + "grad_norm": 1.7199279069900513, + "learning_rate": 7.79848747644826e-06, + "loss": 3.1451, + "step": 14556 + }, + { + "epoch": 1.1670040688701369, + "grad_norm": 1.7549467086791992, + "learning_rate": 7.79722090438044e-06, + "loss": 3.1732, + "step": 14557 + }, + { + "epoch": 1.1670842436511595, + "grad_norm": 1.8488476276397705, + "learning_rate": 7.795954369450703e-06, + "loss": 3.3364, + "step": 14558 + }, + { + "epoch": 1.167164418432182, + "grad_norm": 1.6851810216903687, + "learning_rate": 7.794687871680396e-06, + "loss": 3.2738, + "step": 14559 + }, + { + "epoch": 1.1672445932132047, + "grad_norm": 1.7826869487762451, + "learning_rate": 7.793421411090876e-06, + "loss": 3.2273, + "step": 14560 + }, + { + "epoch": 1.1673247679942274, + "grad_norm": 1.7752636671066284, + "learning_rate": 7.79215498770349e-06, + "loss": 3.1287, + "step": 14561 + }, + { + "epoch": 1.16740494277525, + "grad_norm": 1.7707798480987549, + "learning_rate": 7.790888601539596e-06, + "loss": 3.2123, + "step": 14562 + }, + { + "epoch": 1.1674851175562726, + "grad_norm": 1.7214418649673462, + "learning_rate": 7.78962225262054e-06, + "loss": 3.2523, + "step": 14563 + }, + { + "epoch": 1.1675652923372952, + "grad_norm": 1.7091891765594482, + "learning_rate": 7.788355940967675e-06, + "loss": 3.2134, + "step": 14564 + }, + { + "epoch": 1.1676454671183178, + "grad_norm": 1.6880062818527222, + "learning_rate": 7.78708966660235e-06, + "loss": 3.196, + "step": 14565 + }, + { + "epoch": 1.1677256418993405, + "grad_norm": 1.8059697151184082, + "learning_rate": 7.78582342954591e-06, + "loss": 3.0939, + "step": 14566 + }, + { + "epoch": 1.167805816680363, + "grad_norm": 1.583611249923706, + "learning_rate": 7.784557229819709e-06, + "loss": 3.1158, + "step": 14567 + }, + { + "epoch": 1.167885991461386, + "grad_norm": 1.910445213317871, + "learning_rate": 7.783291067445094e-06, + "loss": 3.1247, + "step": 14568 + }, + { + "epoch": 1.1679661662424086, + "grad_norm": 1.644443154335022, + "learning_rate": 7.782024942443408e-06, + "loss": 3.1012, + "step": 14569 + }, + { + "epoch": 1.1680463410234312, + "grad_norm": 1.7116400003433228, + "learning_rate": 7.780758854836e-06, + "loss": 3.2216, + "step": 14570 + }, + { + "epoch": 1.1681265158044538, + "grad_norm": 1.8646091222763062, + "learning_rate": 7.779492804644216e-06, + "loss": 3.1483, + "step": 14571 + }, + { + "epoch": 1.1682066905854764, + "grad_norm": 1.8204160928726196, + "learning_rate": 7.778226791889399e-06, + "loss": 3.0781, + "step": 14572 + }, + { + "epoch": 1.168286865366499, + "grad_norm": 1.7070187330245972, + "learning_rate": 7.776960816592898e-06, + "loss": 3.1485, + "step": 14573 + }, + { + "epoch": 1.1683670401475217, + "grad_norm": 1.7589662075042725, + "learning_rate": 7.775694878776052e-06, + "loss": 3.1785, + "step": 14574 + }, + { + "epoch": 1.1684472149285443, + "grad_norm": 1.7946168184280396, + "learning_rate": 7.774428978460209e-06, + "loss": 3.2111, + "step": 14575 + }, + { + "epoch": 1.168527389709567, + "grad_norm": 1.7168529033660889, + "learning_rate": 7.773163115666707e-06, + "loss": 3.147, + "step": 14576 + }, + { + "epoch": 1.1686075644905896, + "grad_norm": 1.6996785402297974, + "learning_rate": 7.771897290416889e-06, + "loss": 3.0745, + "step": 14577 + }, + { + "epoch": 1.1686877392716122, + "grad_norm": 1.7949880361557007, + "learning_rate": 7.7706315027321e-06, + "loss": 3.2928, + "step": 14578 + }, + { + "epoch": 1.1687679140526348, + "grad_norm": 1.7274318933486938, + "learning_rate": 7.769365752633676e-06, + "loss": 3.3152, + "step": 14579 + }, + { + "epoch": 1.1688480888336574, + "grad_norm": 1.759432077407837, + "learning_rate": 7.768100040142963e-06, + "loss": 3.2612, + "step": 14580 + }, + { + "epoch": 1.16892826361468, + "grad_norm": 0.6991264224052429, + "learning_rate": 7.766834365281291e-06, + "loss": 2.4609, + "step": 14581 + }, + { + "epoch": 1.1690084383957027, + "grad_norm": 1.776473045349121, + "learning_rate": 7.765568728070007e-06, + "loss": 3.2144, + "step": 14582 + }, + { + "epoch": 1.1690886131767253, + "grad_norm": 0.6713080406188965, + "learning_rate": 7.764303128530449e-06, + "loss": 2.494, + "step": 14583 + }, + { + "epoch": 1.169168787957748, + "grad_norm": 1.750412106513977, + "learning_rate": 7.76303756668395e-06, + "loss": 3.1599, + "step": 14584 + }, + { + "epoch": 1.1692489627387705, + "grad_norm": 0.6895837783813477, + "learning_rate": 7.761772042551851e-06, + "loss": 2.4729, + "step": 14585 + }, + { + "epoch": 1.1693291375197932, + "grad_norm": 1.7866052389144897, + "learning_rate": 7.760506556155485e-06, + "loss": 3.1506, + "step": 14586 + }, + { + "epoch": 1.1694093123008158, + "grad_norm": 1.8361164331436157, + "learning_rate": 7.759241107516188e-06, + "loss": 3.2339, + "step": 14587 + }, + { + "epoch": 1.1694894870818384, + "grad_norm": 0.6707380414009094, + "learning_rate": 7.7579756966553e-06, + "loss": 2.4847, + "step": 14588 + }, + { + "epoch": 1.169569661862861, + "grad_norm": 1.7445205450057983, + "learning_rate": 7.756710323594147e-06, + "loss": 3.1136, + "step": 14589 + }, + { + "epoch": 1.1696498366438837, + "grad_norm": 1.81741201877594, + "learning_rate": 7.755444988354071e-06, + "loss": 3.0951, + "step": 14590 + }, + { + "epoch": 1.1697300114249063, + "grad_norm": 1.8267604112625122, + "learning_rate": 7.7541796909564e-06, + "loss": 3.1604, + "step": 14591 + }, + { + "epoch": 1.169810186205929, + "grad_norm": 1.83096444606781, + "learning_rate": 7.752914431422467e-06, + "loss": 3.2368, + "step": 14592 + }, + { + "epoch": 1.1698903609869515, + "grad_norm": 1.7736986875534058, + "learning_rate": 7.751649209773608e-06, + "loss": 3.0556, + "step": 14593 + }, + { + "epoch": 1.1699705357679742, + "grad_norm": 1.7873855829238892, + "learning_rate": 7.750384026031148e-06, + "loss": 3.198, + "step": 14594 + }, + { + "epoch": 1.1700507105489968, + "grad_norm": 1.706712245941162, + "learning_rate": 7.749118880216421e-06, + "loss": 3.1692, + "step": 14595 + }, + { + "epoch": 1.1701308853300194, + "grad_norm": 1.7785283327102661, + "learning_rate": 7.747853772350758e-06, + "loss": 3.2581, + "step": 14596 + }, + { + "epoch": 1.170211060111042, + "grad_norm": 1.719313144683838, + "learning_rate": 7.746588702455484e-06, + "loss": 3.2527, + "step": 14597 + }, + { + "epoch": 1.1702912348920647, + "grad_norm": 1.6919859647750854, + "learning_rate": 7.745323670551932e-06, + "loss": 3.1833, + "step": 14598 + }, + { + "epoch": 1.1703714096730873, + "grad_norm": 1.7237313985824585, + "learning_rate": 7.744058676661427e-06, + "loss": 3.0968, + "step": 14599 + }, + { + "epoch": 1.17045158445411, + "grad_norm": 1.7495739459991455, + "learning_rate": 7.7427937208053e-06, + "loss": 3.1647, + "step": 14600 + }, + { + "epoch": 1.1705317592351325, + "grad_norm": 1.8724212646484375, + "learning_rate": 7.741528803004873e-06, + "loss": 3.2755, + "step": 14601 + }, + { + "epoch": 1.1706119340161552, + "grad_norm": 1.7769793272018433, + "learning_rate": 7.740263923281474e-06, + "loss": 3.2281, + "step": 14602 + }, + { + "epoch": 1.1706921087971778, + "grad_norm": 1.6368486881256104, + "learning_rate": 7.738999081656431e-06, + "loss": 3.1758, + "step": 14603 + }, + { + "epoch": 1.1707722835782004, + "grad_norm": 1.7699588537216187, + "learning_rate": 7.737734278151066e-06, + "loss": 3.1549, + "step": 14604 + }, + { + "epoch": 1.170852458359223, + "grad_norm": 1.769576072692871, + "learning_rate": 7.736469512786704e-06, + "loss": 3.179, + "step": 14605 + }, + { + "epoch": 1.1709326331402456, + "grad_norm": 1.7747708559036255, + "learning_rate": 7.735204785584667e-06, + "loss": 3.1714, + "step": 14606 + }, + { + "epoch": 1.1710128079212683, + "grad_norm": 1.7762296199798584, + "learning_rate": 7.73394009656628e-06, + "loss": 3.1146, + "step": 14607 + }, + { + "epoch": 1.171092982702291, + "grad_norm": 1.8956780433654785, + "learning_rate": 7.732675445752865e-06, + "loss": 3.1129, + "step": 14608 + }, + { + "epoch": 1.1711731574833135, + "grad_norm": 1.9153321981430054, + "learning_rate": 7.731410833165741e-06, + "loss": 3.1694, + "step": 14609 + }, + { + "epoch": 1.1712533322643361, + "grad_norm": 1.7525445222854614, + "learning_rate": 7.730146258826233e-06, + "loss": 3.0934, + "step": 14610 + }, + { + "epoch": 1.1713335070453588, + "grad_norm": 1.8566434383392334, + "learning_rate": 7.728881722755658e-06, + "loss": 3.2302, + "step": 14611 + }, + { + "epoch": 1.1714136818263814, + "grad_norm": 0.7066161632537842, + "learning_rate": 7.727617224975336e-06, + "loss": 2.4856, + "step": 14612 + }, + { + "epoch": 1.1714938566074042, + "grad_norm": 1.8415313959121704, + "learning_rate": 7.726352765506588e-06, + "loss": 3.2869, + "step": 14613 + }, + { + "epoch": 1.1715740313884269, + "grad_norm": 1.8183265924453735, + "learning_rate": 7.72508834437073e-06, + "loss": 3.2422, + "step": 14614 + }, + { + "epoch": 1.1716542061694495, + "grad_norm": 1.8348983526229858, + "learning_rate": 7.723823961589083e-06, + "loss": 3.2148, + "step": 14615 + }, + { + "epoch": 1.171734380950472, + "grad_norm": 1.7694659233093262, + "learning_rate": 7.722559617182958e-06, + "loss": 3.2412, + "step": 14616 + }, + { + "epoch": 1.1718145557314947, + "grad_norm": 1.829514741897583, + "learning_rate": 7.721295311173677e-06, + "loss": 3.2781, + "step": 14617 + }, + { + "epoch": 1.1718947305125174, + "grad_norm": 1.7519605159759521, + "learning_rate": 7.720031043582556e-06, + "loss": 3.1387, + "step": 14618 + }, + { + "epoch": 1.17197490529354, + "grad_norm": 1.7713247537612915, + "learning_rate": 7.718766814430907e-06, + "loss": 3.0948, + "step": 14619 + }, + { + "epoch": 1.1720550800745626, + "grad_norm": 1.9060845375061035, + "learning_rate": 7.717502623740045e-06, + "loss": 3.2971, + "step": 14620 + }, + { + "epoch": 1.1721352548555852, + "grad_norm": 1.750178575515747, + "learning_rate": 7.716238471531283e-06, + "loss": 3.2051, + "step": 14621 + }, + { + "epoch": 1.1722154296366079, + "grad_norm": 1.7868783473968506, + "learning_rate": 7.714974357825937e-06, + "loss": 3.1918, + "step": 14622 + }, + { + "epoch": 1.1722956044176305, + "grad_norm": 1.7502319812774658, + "learning_rate": 7.713710282645317e-06, + "loss": 3.1084, + "step": 14623 + }, + { + "epoch": 1.172375779198653, + "grad_norm": 1.765078067779541, + "learning_rate": 7.712446246010738e-06, + "loss": 3.2771, + "step": 14624 + }, + { + "epoch": 1.1724559539796757, + "grad_norm": 1.60875403881073, + "learning_rate": 7.711182247943509e-06, + "loss": 3.1743, + "step": 14625 + }, + { + "epoch": 1.1725361287606983, + "grad_norm": 1.7971727848052979, + "learning_rate": 7.709918288464937e-06, + "loss": 3.3163, + "step": 14626 + }, + { + "epoch": 1.172616303541721, + "grad_norm": 1.8942393064498901, + "learning_rate": 7.708654367596338e-06, + "loss": 3.0922, + "step": 14627 + }, + { + "epoch": 1.1726964783227436, + "grad_norm": 1.8699334859848022, + "learning_rate": 7.70739048535902e-06, + "loss": 3.173, + "step": 14628 + }, + { + "epoch": 1.1727766531037662, + "grad_norm": 1.7389750480651855, + "learning_rate": 7.706126641774287e-06, + "loss": 3.1572, + "step": 14629 + }, + { + "epoch": 1.1728568278847888, + "grad_norm": 1.8427382707595825, + "learning_rate": 7.704862836863455e-06, + "loss": 3.1339, + "step": 14630 + }, + { + "epoch": 1.1729370026658115, + "grad_norm": 1.807469129562378, + "learning_rate": 7.703599070647822e-06, + "loss": 3.1617, + "step": 14631 + }, + { + "epoch": 1.173017177446834, + "grad_norm": 1.7977215051651, + "learning_rate": 7.702335343148702e-06, + "loss": 3.2958, + "step": 14632 + }, + { + "epoch": 1.1730973522278567, + "grad_norm": 1.7635878324508667, + "learning_rate": 7.701071654387398e-06, + "loss": 3.1932, + "step": 14633 + }, + { + "epoch": 1.1731775270088793, + "grad_norm": 1.77853524684906, + "learning_rate": 7.699808004385216e-06, + "loss": 3.0602, + "step": 14634 + }, + { + "epoch": 1.173257701789902, + "grad_norm": 1.9940171241760254, + "learning_rate": 7.698544393163461e-06, + "loss": 3.2951, + "step": 14635 + }, + { + "epoch": 1.1733378765709246, + "grad_norm": 1.8316069841384888, + "learning_rate": 7.697280820743435e-06, + "loss": 3.2903, + "step": 14636 + }, + { + "epoch": 1.1734180513519472, + "grad_norm": 0.695598840713501, + "learning_rate": 7.696017287146443e-06, + "loss": 2.4711, + "step": 14637 + }, + { + "epoch": 1.1734982261329698, + "grad_norm": 1.7926193475723267, + "learning_rate": 7.694753792393787e-06, + "loss": 3.2666, + "step": 14638 + }, + { + "epoch": 1.1735784009139925, + "grad_norm": 1.6959246397018433, + "learning_rate": 7.69349033650677e-06, + "loss": 3.0725, + "step": 14639 + }, + { + "epoch": 1.173658575695015, + "grad_norm": 1.7244927883148193, + "learning_rate": 7.692226919506694e-06, + "loss": 3.1415, + "step": 14640 + }, + { + "epoch": 1.1737387504760377, + "grad_norm": 1.8014696836471558, + "learning_rate": 7.690963541414858e-06, + "loss": 3.3135, + "step": 14641 + }, + { + "epoch": 1.1738189252570603, + "grad_norm": 1.675959587097168, + "learning_rate": 7.689700202252562e-06, + "loss": 3.1208, + "step": 14642 + }, + { + "epoch": 1.173899100038083, + "grad_norm": 1.8072447776794434, + "learning_rate": 7.68843690204111e-06, + "loss": 3.1168, + "step": 14643 + }, + { + "epoch": 1.1739792748191056, + "grad_norm": 1.807136058807373, + "learning_rate": 7.687173640801793e-06, + "loss": 3.2113, + "step": 14644 + }, + { + "epoch": 1.1740594496001282, + "grad_norm": 1.7795082330703735, + "learning_rate": 7.685910418555916e-06, + "loss": 3.1425, + "step": 14645 + }, + { + "epoch": 1.1741396243811508, + "grad_norm": 1.791656494140625, + "learning_rate": 7.684647235324771e-06, + "loss": 3.2731, + "step": 14646 + }, + { + "epoch": 1.1742197991621734, + "grad_norm": 1.7348006963729858, + "learning_rate": 7.683384091129659e-06, + "loss": 3.124, + "step": 14647 + }, + { + "epoch": 1.174299973943196, + "grad_norm": 1.969125509262085, + "learning_rate": 7.682120985991877e-06, + "loss": 3.2256, + "step": 14648 + }, + { + "epoch": 1.174380148724219, + "grad_norm": 1.804140329360962, + "learning_rate": 7.680857919932715e-06, + "loss": 3.1181, + "step": 14649 + }, + { + "epoch": 1.1744603235052415, + "grad_norm": 1.8352998495101929, + "learning_rate": 7.679594892973474e-06, + "loss": 3.1957, + "step": 14650 + }, + { + "epoch": 1.1745404982862642, + "grad_norm": 1.8265573978424072, + "learning_rate": 7.678331905135442e-06, + "loss": 3.2139, + "step": 14651 + }, + { + "epoch": 1.1746206730672868, + "grad_norm": 1.7605984210968018, + "learning_rate": 7.677068956439918e-06, + "loss": 3.1995, + "step": 14652 + }, + { + "epoch": 1.1747008478483094, + "grad_norm": 1.785187005996704, + "learning_rate": 7.675806046908192e-06, + "loss": 3.1805, + "step": 14653 + }, + { + "epoch": 1.174781022629332, + "grad_norm": 1.8168885707855225, + "learning_rate": 7.674543176561558e-06, + "loss": 3.1999, + "step": 14654 + }, + { + "epoch": 1.1748611974103547, + "grad_norm": 1.8228307962417603, + "learning_rate": 7.673280345421305e-06, + "loss": 3.0939, + "step": 14655 + }, + { + "epoch": 1.1749413721913773, + "grad_norm": 1.8456403017044067, + "learning_rate": 7.672017553508726e-06, + "loss": 3.2173, + "step": 14656 + }, + { + "epoch": 1.1750215469724, + "grad_norm": 1.7422362565994263, + "learning_rate": 7.67075480084511e-06, + "loss": 3.1416, + "step": 14657 + }, + { + "epoch": 1.1751017217534225, + "grad_norm": 1.8148337602615356, + "learning_rate": 7.669492087451749e-06, + "loss": 3.2199, + "step": 14658 + }, + { + "epoch": 1.1751818965344452, + "grad_norm": 1.959113597869873, + "learning_rate": 7.668229413349928e-06, + "loss": 3.231, + "step": 14659 + }, + { + "epoch": 1.1752620713154678, + "grad_norm": 1.916817545890808, + "learning_rate": 7.66696677856094e-06, + "loss": 3.1692, + "step": 14660 + }, + { + "epoch": 1.1753422460964904, + "grad_norm": 1.746691107749939, + "learning_rate": 7.665704183106066e-06, + "loss": 3.1448, + "step": 14661 + }, + { + "epoch": 1.175422420877513, + "grad_norm": 1.7548226118087769, + "learning_rate": 7.664441627006598e-06, + "loss": 3.1154, + "step": 14662 + }, + { + "epoch": 1.1755025956585357, + "grad_norm": 1.7160823345184326, + "learning_rate": 7.663179110283824e-06, + "loss": 3.2117, + "step": 14663 + }, + { + "epoch": 1.1755827704395583, + "grad_norm": 1.7321455478668213, + "learning_rate": 7.661916632959023e-06, + "loss": 3.4347, + "step": 14664 + }, + { + "epoch": 1.175662945220581, + "grad_norm": 1.7331572771072388, + "learning_rate": 7.660654195053487e-06, + "loss": 3.174, + "step": 14665 + }, + { + "epoch": 1.1757431200016035, + "grad_norm": 1.7275625467300415, + "learning_rate": 7.659391796588493e-06, + "loss": 3.2396, + "step": 14666 + }, + { + "epoch": 1.1758232947826261, + "grad_norm": 1.753425121307373, + "learning_rate": 7.65812943758533e-06, + "loss": 3.1838, + "step": 14667 + }, + { + "epoch": 1.1759034695636488, + "grad_norm": 0.689216136932373, + "learning_rate": 7.65686711806528e-06, + "loss": 2.4603, + "step": 14668 + }, + { + "epoch": 1.1759836443446714, + "grad_norm": 1.8457443714141846, + "learning_rate": 7.655604838049623e-06, + "loss": 3.3571, + "step": 14669 + }, + { + "epoch": 1.176063819125694, + "grad_norm": 1.7820253372192383, + "learning_rate": 7.654342597559646e-06, + "loss": 3.1393, + "step": 14670 + }, + { + "epoch": 1.1761439939067166, + "grad_norm": 1.688090205192566, + "learning_rate": 7.653080396616621e-06, + "loss": 3.3152, + "step": 14671 + }, + { + "epoch": 1.1762241686877393, + "grad_norm": 2.0718178749084473, + "learning_rate": 7.651818235241835e-06, + "loss": 3.0966, + "step": 14672 + }, + { + "epoch": 1.176304343468762, + "grad_norm": 1.7499300241470337, + "learning_rate": 7.650556113456569e-06, + "loss": 3.0852, + "step": 14673 + }, + { + "epoch": 1.1763845182497845, + "grad_norm": 1.687147617340088, + "learning_rate": 7.649294031282096e-06, + "loss": 3.1442, + "step": 14674 + }, + { + "epoch": 1.1764646930308071, + "grad_norm": 1.78769850730896, + "learning_rate": 7.6480319887397e-06, + "loss": 3.0662, + "step": 14675 + }, + { + "epoch": 1.1765448678118298, + "grad_norm": 1.8281962871551514, + "learning_rate": 7.646769985850652e-06, + "loss": 3.2175, + "step": 14676 + }, + { + "epoch": 1.1766250425928524, + "grad_norm": 1.927778959274292, + "learning_rate": 7.645508022636234e-06, + "loss": 3.1893, + "step": 14677 + }, + { + "epoch": 1.176705217373875, + "grad_norm": 1.9642318487167358, + "learning_rate": 7.644246099117722e-06, + "loss": 3.1836, + "step": 14678 + }, + { + "epoch": 1.1767853921548976, + "grad_norm": 1.871134877204895, + "learning_rate": 7.64298421531639e-06, + "loss": 3.2663, + "step": 14679 + }, + { + "epoch": 1.1768655669359203, + "grad_norm": 1.7854050397872925, + "learning_rate": 7.641722371253513e-06, + "loss": 3.1565, + "step": 14680 + }, + { + "epoch": 1.1769457417169429, + "grad_norm": 1.8672959804534912, + "learning_rate": 7.640460566950364e-06, + "loss": 3.2871, + "step": 14681 + }, + { + "epoch": 1.1770259164979655, + "grad_norm": 1.7628287076950073, + "learning_rate": 7.63919880242822e-06, + "loss": 3.1278, + "step": 14682 + }, + { + "epoch": 1.1771060912789881, + "grad_norm": 1.9211275577545166, + "learning_rate": 7.637937077708354e-06, + "loss": 3.3455, + "step": 14683 + }, + { + "epoch": 1.1771862660600108, + "grad_norm": 0.7036932110786438, + "learning_rate": 7.636675392812033e-06, + "loss": 2.4686, + "step": 14684 + }, + { + "epoch": 1.1772664408410334, + "grad_norm": 1.8420345783233643, + "learning_rate": 7.635413747760535e-06, + "loss": 3.1819, + "step": 14685 + }, + { + "epoch": 1.177346615622056, + "grad_norm": 1.7587366104125977, + "learning_rate": 7.634152142575126e-06, + "loss": 3.2174, + "step": 14686 + }, + { + "epoch": 1.1774267904030786, + "grad_norm": 1.7628315687179565, + "learning_rate": 7.632890577277077e-06, + "loss": 3.2075, + "step": 14687 + }, + { + "epoch": 1.1775069651841013, + "grad_norm": 0.6931021213531494, + "learning_rate": 7.631629051887662e-06, + "loss": 2.4725, + "step": 14688 + }, + { + "epoch": 1.1775871399651239, + "grad_norm": 1.8304308652877808, + "learning_rate": 7.630367566428144e-06, + "loss": 3.1722, + "step": 14689 + }, + { + "epoch": 1.1776673147461465, + "grad_norm": 1.6881606578826904, + "learning_rate": 7.629106120919796e-06, + "loss": 3.1628, + "step": 14690 + }, + { + "epoch": 1.1777474895271691, + "grad_norm": 1.7538583278656006, + "learning_rate": 7.627844715383881e-06, + "loss": 3.192, + "step": 14691 + }, + { + "epoch": 1.1778276643081917, + "grad_norm": 1.8627216815948486, + "learning_rate": 7.6265833498416674e-06, + "loss": 3.2078, + "step": 14692 + }, + { + "epoch": 1.1779078390892144, + "grad_norm": 2.1591618061065674, + "learning_rate": 7.6253220243144245e-06, + "loss": 3.2042, + "step": 14693 + }, + { + "epoch": 1.177988013870237, + "grad_norm": 1.8473109006881714, + "learning_rate": 7.6240607388234135e-06, + "loss": 3.2771, + "step": 14694 + }, + { + "epoch": 1.1780681886512598, + "grad_norm": 1.7102770805358887, + "learning_rate": 7.622799493389903e-06, + "loss": 3.1703, + "step": 14695 + }, + { + "epoch": 1.1781483634322825, + "grad_norm": 1.7756435871124268, + "learning_rate": 7.621538288035153e-06, + "loss": 3.1384, + "step": 14696 + }, + { + "epoch": 1.178228538213305, + "grad_norm": 1.75860595703125, + "learning_rate": 7.620277122780429e-06, + "loss": 3.1615, + "step": 14697 + }, + { + "epoch": 1.1783087129943277, + "grad_norm": 1.8563505411148071, + "learning_rate": 7.6190159976469965e-06, + "loss": 3.1868, + "step": 14698 + }, + { + "epoch": 1.1783888877753503, + "grad_norm": 0.7091949582099915, + "learning_rate": 7.6177549126561135e-06, + "loss": 2.4554, + "step": 14699 + }, + { + "epoch": 1.178469062556373, + "grad_norm": 1.7678544521331787, + "learning_rate": 7.6164938678290425e-06, + "loss": 3.1578, + "step": 14700 + }, + { + "epoch": 1.1785492373373956, + "grad_norm": 1.7925711870193481, + "learning_rate": 7.615232863187048e-06, + "loss": 3.1192, + "step": 14701 + }, + { + "epoch": 1.1786294121184182, + "grad_norm": 1.6983331441879272, + "learning_rate": 7.613971898751383e-06, + "loss": 3.2128, + "step": 14702 + }, + { + "epoch": 1.1787095868994408, + "grad_norm": 1.855505108833313, + "learning_rate": 7.612710974543313e-06, + "loss": 3.2133, + "step": 14703 + }, + { + "epoch": 1.1787897616804635, + "grad_norm": 1.809731364250183, + "learning_rate": 7.6114500905840935e-06, + "loss": 3.2102, + "step": 14704 + }, + { + "epoch": 1.178869936461486, + "grad_norm": 1.70709228515625, + "learning_rate": 7.610189246894985e-06, + "loss": 3.1799, + "step": 14705 + }, + { + "epoch": 1.1789501112425087, + "grad_norm": 0.6675413250923157, + "learning_rate": 7.608928443497243e-06, + "loss": 2.4589, + "step": 14706 + }, + { + "epoch": 1.1790302860235313, + "grad_norm": 1.7900760173797607, + "learning_rate": 7.607667680412124e-06, + "loss": 3.188, + "step": 14707 + }, + { + "epoch": 1.179110460804554, + "grad_norm": 0.681403636932373, + "learning_rate": 7.606406957660887e-06, + "loss": 2.4855, + "step": 14708 + }, + { + "epoch": 1.1791906355855766, + "grad_norm": 1.7394747734069824, + "learning_rate": 7.605146275264783e-06, + "loss": 3.1653, + "step": 14709 + }, + { + "epoch": 1.1792708103665992, + "grad_norm": 1.694541096687317, + "learning_rate": 7.603885633245069e-06, + "loss": 3.1161, + "step": 14710 + }, + { + "epoch": 1.1793509851476218, + "grad_norm": 1.8045984506607056, + "learning_rate": 7.602625031622999e-06, + "loss": 3.2542, + "step": 14711 + }, + { + "epoch": 1.1794311599286444, + "grad_norm": 1.8839765787124634, + "learning_rate": 7.601364470419825e-06, + "loss": 3.0955, + "step": 14712 + }, + { + "epoch": 1.179511334709667, + "grad_norm": 1.7108608484268188, + "learning_rate": 7.600103949656804e-06, + "loss": 3.2001, + "step": 14713 + }, + { + "epoch": 1.1795915094906897, + "grad_norm": 1.7393587827682495, + "learning_rate": 7.598843469355181e-06, + "loss": 3.2819, + "step": 14714 + }, + { + "epoch": 1.1796716842717123, + "grad_norm": 1.823846697807312, + "learning_rate": 7.597583029536212e-06, + "loss": 3.1959, + "step": 14715 + }, + { + "epoch": 1.179751859052735, + "grad_norm": 0.6984516382217407, + "learning_rate": 7.596322630221148e-06, + "loss": 2.4761, + "step": 14716 + }, + { + "epoch": 1.1798320338337576, + "grad_norm": 1.7233595848083496, + "learning_rate": 7.595062271431235e-06, + "loss": 3.2096, + "step": 14717 + }, + { + "epoch": 1.1799122086147802, + "grad_norm": 1.8787678480148315, + "learning_rate": 7.593801953187727e-06, + "loss": 3.2026, + "step": 14718 + }, + { + "epoch": 1.1799923833958028, + "grad_norm": 1.8306777477264404, + "learning_rate": 7.592541675511867e-06, + "loss": 3.213, + "step": 14719 + }, + { + "epoch": 1.1800725581768254, + "grad_norm": 1.7205824851989746, + "learning_rate": 7.591281438424907e-06, + "loss": 3.2555, + "step": 14720 + }, + { + "epoch": 1.180152732957848, + "grad_norm": 1.8925362825393677, + "learning_rate": 7.590021241948095e-06, + "loss": 3.2401, + "step": 14721 + }, + { + "epoch": 1.1802329077388707, + "grad_norm": 1.789484977722168, + "learning_rate": 7.588761086102673e-06, + "loss": 3.25, + "step": 14722 + }, + { + "epoch": 1.1803130825198933, + "grad_norm": 2.006889581680298, + "learning_rate": 7.587500970909891e-06, + "loss": 3.252, + "step": 14723 + }, + { + "epoch": 1.180393257300916, + "grad_norm": 1.8895989656448364, + "learning_rate": 7.586240896390991e-06, + "loss": 3.3082, + "step": 14724 + }, + { + "epoch": 1.1804734320819386, + "grad_norm": 1.7590689659118652, + "learning_rate": 7.584980862567219e-06, + "loss": 3.1511, + "step": 14725 + }, + { + "epoch": 1.1805536068629612, + "grad_norm": 1.7269926071166992, + "learning_rate": 7.58372086945982e-06, + "loss": 3.1642, + "step": 14726 + }, + { + "epoch": 1.1806337816439838, + "grad_norm": 1.706602931022644, + "learning_rate": 7.582460917090033e-06, + "loss": 3.0895, + "step": 14727 + }, + { + "epoch": 1.1807139564250064, + "grad_norm": 0.6914767026901245, + "learning_rate": 7.581201005479106e-06, + "loss": 2.4579, + "step": 14728 + }, + { + "epoch": 1.180794131206029, + "grad_norm": 1.809622049331665, + "learning_rate": 7.579941134648275e-06, + "loss": 3.3107, + "step": 14729 + }, + { + "epoch": 1.180874305987052, + "grad_norm": 1.7087628841400146, + "learning_rate": 7.578681304618785e-06, + "loss": 3.1409, + "step": 14730 + }, + { + "epoch": 1.1809544807680745, + "grad_norm": 1.7757445573806763, + "learning_rate": 7.577421515411875e-06, + "loss": 3.3142, + "step": 14731 + }, + { + "epoch": 1.1810346555490971, + "grad_norm": 1.8790297508239746, + "learning_rate": 7.576161767048783e-06, + "loss": 3.1571, + "step": 14732 + }, + { + "epoch": 1.1811148303301198, + "grad_norm": 1.6797372102737427, + "learning_rate": 7.574902059550752e-06, + "loss": 3.2332, + "step": 14733 + }, + { + "epoch": 1.1811950051111424, + "grad_norm": 1.6723202466964722, + "learning_rate": 7.573642392939014e-06, + "loss": 3.2277, + "step": 14734 + }, + { + "epoch": 1.181275179892165, + "grad_norm": 1.7708957195281982, + "learning_rate": 7.5723827672348115e-06, + "loss": 3.2532, + "step": 14735 + }, + { + "epoch": 1.1813553546731876, + "grad_norm": 1.8165837526321411, + "learning_rate": 7.571123182459381e-06, + "loss": 3.2021, + "step": 14736 + }, + { + "epoch": 1.1814355294542103, + "grad_norm": 0.7111621499061584, + "learning_rate": 7.5698636386339565e-06, + "loss": 2.4735, + "step": 14737 + }, + { + "epoch": 1.1815157042352329, + "grad_norm": 1.7564438581466675, + "learning_rate": 7.568604135779777e-06, + "loss": 3.137, + "step": 14738 + }, + { + "epoch": 1.1815958790162555, + "grad_norm": 1.8187298774719238, + "learning_rate": 7.567344673918072e-06, + "loss": 3.2166, + "step": 14739 + }, + { + "epoch": 1.1816760537972781, + "grad_norm": 1.8813656568527222, + "learning_rate": 7.566085253070079e-06, + "loss": 3.1671, + "step": 14740 + }, + { + "epoch": 1.1817562285783008, + "grad_norm": 1.8372818231582642, + "learning_rate": 7.564825873257032e-06, + "loss": 3.1733, + "step": 14741 + }, + { + "epoch": 1.1818364033593234, + "grad_norm": 1.8000694513320923, + "learning_rate": 7.563566534500161e-06, + "loss": 3.221, + "step": 14742 + }, + { + "epoch": 1.181916578140346, + "grad_norm": 1.7081489562988281, + "learning_rate": 7.562307236820702e-06, + "loss": 3.1533, + "step": 14743 + }, + { + "epoch": 1.1819967529213686, + "grad_norm": 1.7819080352783203, + "learning_rate": 7.5610479802398815e-06, + "loss": 3.1203, + "step": 14744 + }, + { + "epoch": 1.1820769277023913, + "grad_norm": 1.7287232875823975, + "learning_rate": 7.559788764778933e-06, + "loss": 3.1732, + "step": 14745 + }, + { + "epoch": 1.1821571024834139, + "grad_norm": 1.8165923357009888, + "learning_rate": 7.5585295904590865e-06, + "loss": 3.1473, + "step": 14746 + }, + { + "epoch": 1.1822372772644365, + "grad_norm": 1.8117519617080688, + "learning_rate": 7.557270457301569e-06, + "loss": 3.203, + "step": 14747 + }, + { + "epoch": 1.1823174520454591, + "grad_norm": 1.8434207439422607, + "learning_rate": 7.5560113653276135e-06, + "loss": 3.1693, + "step": 14748 + }, + { + "epoch": 1.1823976268264818, + "grad_norm": 1.799757480621338, + "learning_rate": 7.5547523145584426e-06, + "loss": 3.0535, + "step": 14749 + }, + { + "epoch": 1.1824778016075044, + "grad_norm": 0.7021704316139221, + "learning_rate": 7.553493305015286e-06, + "loss": 2.4598, + "step": 14750 + }, + { + "epoch": 1.182557976388527, + "grad_norm": 1.7917258739471436, + "learning_rate": 7.552234336719372e-06, + "loss": 3.2831, + "step": 14751 + }, + { + "epoch": 1.1826381511695496, + "grad_norm": 0.7230285406112671, + "learning_rate": 7.550975409691923e-06, + "loss": 2.4882, + "step": 14752 + }, + { + "epoch": 1.1827183259505722, + "grad_norm": 1.9726696014404297, + "learning_rate": 7.549716523954167e-06, + "loss": 3.3323, + "step": 14753 + }, + { + "epoch": 1.1827985007315949, + "grad_norm": 1.7768077850341797, + "learning_rate": 7.548457679527325e-06, + "loss": 3.0717, + "step": 14754 + }, + { + "epoch": 1.1828786755126175, + "grad_norm": 1.7348575592041016, + "learning_rate": 7.547198876432622e-06, + "loss": 3.2341, + "step": 14755 + }, + { + "epoch": 1.1829588502936401, + "grad_norm": 1.8864600658416748, + "learning_rate": 7.5459401146912836e-06, + "loss": 3.1024, + "step": 14756 + }, + { + "epoch": 1.1830390250746627, + "grad_norm": 1.8306703567504883, + "learning_rate": 7.544681394324528e-06, + "loss": 3.1707, + "step": 14757 + }, + { + "epoch": 1.1831191998556854, + "grad_norm": 1.7540868520736694, + "learning_rate": 7.543422715353581e-06, + "loss": 3.2188, + "step": 14758 + }, + { + "epoch": 1.183199374636708, + "grad_norm": 1.7844237089157104, + "learning_rate": 7.542164077799659e-06, + "loss": 3.2087, + "step": 14759 + }, + { + "epoch": 1.1832795494177306, + "grad_norm": 1.8306119441986084, + "learning_rate": 7.540905481683984e-06, + "loss": 3.261, + "step": 14760 + }, + { + "epoch": 1.1833597241987532, + "grad_norm": 1.809523105621338, + "learning_rate": 7.539646927027779e-06, + "loss": 3.1244, + "step": 14761 + }, + { + "epoch": 1.1834398989797759, + "grad_norm": 1.7665477991104126, + "learning_rate": 7.5383884138522565e-06, + "loss": 3.1544, + "step": 14762 + }, + { + "epoch": 1.1835200737607985, + "grad_norm": 1.7688831090927124, + "learning_rate": 7.537129942178641e-06, + "loss": 3.1281, + "step": 14763 + }, + { + "epoch": 1.183600248541821, + "grad_norm": 1.8483595848083496, + "learning_rate": 7.535871512028144e-06, + "loss": 3.2471, + "step": 14764 + }, + { + "epoch": 1.1836804233228437, + "grad_norm": 1.727489709854126, + "learning_rate": 7.534613123421985e-06, + "loss": 3.1213, + "step": 14765 + }, + { + "epoch": 1.1837605981038664, + "grad_norm": 1.823297142982483, + "learning_rate": 7.533354776381382e-06, + "loss": 3.1376, + "step": 14766 + }, + { + "epoch": 1.183840772884889, + "grad_norm": 1.8874776363372803, + "learning_rate": 7.5320964709275455e-06, + "loss": 3.2975, + "step": 14767 + }, + { + "epoch": 1.1839209476659116, + "grad_norm": 1.726840853691101, + "learning_rate": 7.530838207081696e-06, + "loss": 3.1369, + "step": 14768 + }, + { + "epoch": 1.1840011224469342, + "grad_norm": 1.718154788017273, + "learning_rate": 7.529579984865041e-06, + "loss": 3.1156, + "step": 14769 + }, + { + "epoch": 1.1840812972279569, + "grad_norm": 1.729727029800415, + "learning_rate": 7.528321804298797e-06, + "loss": 3.1731, + "step": 14770 + }, + { + "epoch": 1.1841614720089795, + "grad_norm": 1.7876152992248535, + "learning_rate": 7.527063665404179e-06, + "loss": 3.215, + "step": 14771 + }, + { + "epoch": 1.184241646790002, + "grad_norm": 1.8283576965332031, + "learning_rate": 7.5258055682023925e-06, + "loss": 3.2597, + "step": 14772 + }, + { + "epoch": 1.1843218215710247, + "grad_norm": 1.765533685684204, + "learning_rate": 7.524547512714655e-06, + "loss": 3.0936, + "step": 14773 + }, + { + "epoch": 1.1844019963520473, + "grad_norm": 1.8994008302688599, + "learning_rate": 7.523289498962172e-06, + "loss": 3.189, + "step": 14774 + }, + { + "epoch": 1.18448217113307, + "grad_norm": 1.7948025465011597, + "learning_rate": 7.522031526966156e-06, + "loss": 3.018, + "step": 14775 + }, + { + "epoch": 1.1845623459140928, + "grad_norm": 1.843306303024292, + "learning_rate": 7.520773596747815e-06, + "loss": 3.1534, + "step": 14776 + }, + { + "epoch": 1.1846425206951154, + "grad_norm": 1.8292136192321777, + "learning_rate": 7.519515708328358e-06, + "loss": 3.191, + "step": 14777 + }, + { + "epoch": 1.184722695476138, + "grad_norm": 1.780312418937683, + "learning_rate": 7.518257861728992e-06, + "loss": 3.2502, + "step": 14778 + }, + { + "epoch": 1.1848028702571607, + "grad_norm": 1.8414034843444824, + "learning_rate": 7.517000056970922e-06, + "loss": 3.2537, + "step": 14779 + }, + { + "epoch": 1.1848830450381833, + "grad_norm": 0.7083969712257385, + "learning_rate": 7.5157422940753565e-06, + "loss": 2.4755, + "step": 14780 + }, + { + "epoch": 1.184963219819206, + "grad_norm": 1.7825857400894165, + "learning_rate": 7.514484573063503e-06, + "loss": 3.0924, + "step": 14781 + }, + { + "epoch": 1.1850433946002286, + "grad_norm": 1.8122159242630005, + "learning_rate": 7.5132268939565615e-06, + "loss": 3.21, + "step": 14782 + }, + { + "epoch": 1.1851235693812512, + "grad_norm": 1.702077031135559, + "learning_rate": 7.5119692567757396e-06, + "loss": 3.2354, + "step": 14783 + }, + { + "epoch": 1.1852037441622738, + "grad_norm": 1.690229058265686, + "learning_rate": 7.510711661542237e-06, + "loss": 3.268, + "step": 14784 + }, + { + "epoch": 1.1852839189432964, + "grad_norm": 1.8612767457962036, + "learning_rate": 7.509454108277259e-06, + "loss": 3.1485, + "step": 14785 + }, + { + "epoch": 1.185364093724319, + "grad_norm": 1.8067822456359863, + "learning_rate": 7.508196597002007e-06, + "loss": 3.1543, + "step": 14786 + }, + { + "epoch": 1.1854442685053417, + "grad_norm": 1.780983805656433, + "learning_rate": 7.506939127737682e-06, + "loss": 3.2948, + "step": 14787 + }, + { + "epoch": 1.1855244432863643, + "grad_norm": 1.8025460243225098, + "learning_rate": 7.505681700505487e-06, + "loss": 3.2303, + "step": 14788 + }, + { + "epoch": 1.185604618067387, + "grad_norm": 1.7833021879196167, + "learning_rate": 7.5044243153266175e-06, + "loss": 3.1801, + "step": 14789 + }, + { + "epoch": 1.1856847928484096, + "grad_norm": 1.7694814205169678, + "learning_rate": 7.503166972222273e-06, + "loss": 3.1733, + "step": 14790 + }, + { + "epoch": 1.1857649676294322, + "grad_norm": 1.70577073097229, + "learning_rate": 7.501909671213657e-06, + "loss": 3.0732, + "step": 14791 + }, + { + "epoch": 1.1858451424104548, + "grad_norm": 1.7970211505889893, + "learning_rate": 7.50065241232196e-06, + "loss": 3.3452, + "step": 14792 + }, + { + "epoch": 1.1859253171914774, + "grad_norm": 1.748983383178711, + "learning_rate": 7.499395195568386e-06, + "loss": 3.2142, + "step": 14793 + }, + { + "epoch": 1.1860054919725, + "grad_norm": 1.7128989696502686, + "learning_rate": 7.498138020974125e-06, + "loss": 3.1958, + "step": 14794 + }, + { + "epoch": 1.1860856667535227, + "grad_norm": 1.7071954011917114, + "learning_rate": 7.496880888560375e-06, + "loss": 3.1254, + "step": 14795 + }, + { + "epoch": 1.1861658415345453, + "grad_norm": 1.7494566440582275, + "learning_rate": 7.495623798348334e-06, + "loss": 3.0712, + "step": 14796 + }, + { + "epoch": 1.186246016315568, + "grad_norm": 1.6881530284881592, + "learning_rate": 7.49436675035919e-06, + "loss": 3.1659, + "step": 14797 + }, + { + "epoch": 1.1863261910965905, + "grad_norm": 1.6805843114852905, + "learning_rate": 7.493109744614142e-06, + "loss": 3.1533, + "step": 14798 + }, + { + "epoch": 1.1864063658776132, + "grad_norm": 1.779820442199707, + "learning_rate": 7.4918527811343775e-06, + "loss": 3.1998, + "step": 14799 + }, + { + "epoch": 1.1864865406586358, + "grad_norm": 1.7361702919006348, + "learning_rate": 7.490595859941092e-06, + "loss": 3.2853, + "step": 14800 + }, + { + "epoch": 1.1865667154396584, + "grad_norm": 1.9891514778137207, + "learning_rate": 7.489338981055478e-06, + "loss": 3.2231, + "step": 14801 + }, + { + "epoch": 1.186646890220681, + "grad_norm": 1.7355459928512573, + "learning_rate": 7.488082144498722e-06, + "loss": 3.2076, + "step": 14802 + }, + { + "epoch": 1.1867270650017037, + "grad_norm": 1.7548980712890625, + "learning_rate": 7.486825350292017e-06, + "loss": 3.2656, + "step": 14803 + }, + { + "epoch": 1.1868072397827263, + "grad_norm": 1.9240225553512573, + "learning_rate": 7.48556859845655e-06, + "loss": 3.2218, + "step": 14804 + }, + { + "epoch": 1.186887414563749, + "grad_norm": 1.7641029357910156, + "learning_rate": 7.484311889013509e-06, + "loss": 3.1315, + "step": 14805 + }, + { + "epoch": 1.1869675893447715, + "grad_norm": 0.7246750593185425, + "learning_rate": 7.483055221984086e-06, + "loss": 2.4832, + "step": 14806 + }, + { + "epoch": 1.1870477641257942, + "grad_norm": 1.8403170108795166, + "learning_rate": 7.4817985973894615e-06, + "loss": 3.1526, + "step": 14807 + }, + { + "epoch": 1.1871279389068168, + "grad_norm": 1.927213191986084, + "learning_rate": 7.480542015250828e-06, + "loss": 3.2419, + "step": 14808 + }, + { + "epoch": 1.1872081136878394, + "grad_norm": 1.7784072160720825, + "learning_rate": 7.479285475589366e-06, + "loss": 3.2206, + "step": 14809 + }, + { + "epoch": 1.187288288468862, + "grad_norm": 1.7279051542282104, + "learning_rate": 7.478028978426263e-06, + "loss": 3.2068, + "step": 14810 + }, + { + "epoch": 1.1873684632498847, + "grad_norm": 1.647801160812378, + "learning_rate": 7.476772523782704e-06, + "loss": 3.2477, + "step": 14811 + }, + { + "epoch": 1.1874486380309075, + "grad_norm": 1.7159126996994019, + "learning_rate": 7.475516111679869e-06, + "loss": 3.1476, + "step": 14812 + }, + { + "epoch": 1.1875288128119301, + "grad_norm": 1.6720093488693237, + "learning_rate": 7.474259742138945e-06, + "loss": 3.2277, + "step": 14813 + }, + { + "epoch": 1.1876089875929527, + "grad_norm": 1.7913455963134766, + "learning_rate": 7.47300341518111e-06, + "loss": 3.2314, + "step": 14814 + }, + { + "epoch": 1.1876891623739754, + "grad_norm": 1.754334568977356, + "learning_rate": 7.471747130827546e-06, + "loss": 3.1111, + "step": 14815 + }, + { + "epoch": 1.187769337154998, + "grad_norm": 1.784456729888916, + "learning_rate": 7.470490889099437e-06, + "loss": 3.2537, + "step": 14816 + }, + { + "epoch": 1.1878495119360206, + "grad_norm": 1.6770530939102173, + "learning_rate": 7.469234690017957e-06, + "loss": 3.1823, + "step": 14817 + }, + { + "epoch": 1.1879296867170432, + "grad_norm": 1.8035016059875488, + "learning_rate": 7.4679785336042905e-06, + "loss": 3.1627, + "step": 14818 + }, + { + "epoch": 1.1880098614980659, + "grad_norm": 1.7460068464279175, + "learning_rate": 7.4667224198796125e-06, + "loss": 3.291, + "step": 14819 + }, + { + "epoch": 1.1880900362790885, + "grad_norm": 1.7589524984359741, + "learning_rate": 7.465466348865101e-06, + "loss": 3.2109, + "step": 14820 + }, + { + "epoch": 1.1881702110601111, + "grad_norm": 1.7995715141296387, + "learning_rate": 7.464210320581936e-06, + "loss": 3.2839, + "step": 14821 + }, + { + "epoch": 1.1882503858411337, + "grad_norm": 1.7907806634902954, + "learning_rate": 7.462954335051289e-06, + "loss": 3.2278, + "step": 14822 + }, + { + "epoch": 1.1883305606221564, + "grad_norm": 1.787484884262085, + "learning_rate": 7.46169839229434e-06, + "loss": 3.1616, + "step": 14823 + }, + { + "epoch": 1.188410735403179, + "grad_norm": 0.6770790219306946, + "learning_rate": 7.460442492332259e-06, + "loss": 2.4647, + "step": 14824 + }, + { + "epoch": 1.1884909101842016, + "grad_norm": 1.7261128425598145, + "learning_rate": 7.459186635186223e-06, + "loss": 3.1306, + "step": 14825 + }, + { + "epoch": 1.1885710849652242, + "grad_norm": 1.7497313022613525, + "learning_rate": 7.457930820877407e-06, + "loss": 3.2086, + "step": 14826 + }, + { + "epoch": 1.1886512597462469, + "grad_norm": 1.8264384269714355, + "learning_rate": 7.456675049426979e-06, + "loss": 3.1135, + "step": 14827 + }, + { + "epoch": 1.1887314345272695, + "grad_norm": 1.7836107015609741, + "learning_rate": 7.455419320856116e-06, + "loss": 3.1772, + "step": 14828 + }, + { + "epoch": 1.188811609308292, + "grad_norm": 1.7764227390289307, + "learning_rate": 7.454163635185984e-06, + "loss": 3.1672, + "step": 14829 + }, + { + "epoch": 1.1888917840893147, + "grad_norm": 1.750975489616394, + "learning_rate": 7.4529079924377555e-06, + "loss": 3.1199, + "step": 14830 + }, + { + "epoch": 1.1889719588703374, + "grad_norm": 1.6930724382400513, + "learning_rate": 7.451652392632603e-06, + "loss": 3.1903, + "step": 14831 + }, + { + "epoch": 1.18905213365136, + "grad_norm": 1.765515685081482, + "learning_rate": 7.4503968357916915e-06, + "loss": 3.1562, + "step": 14832 + }, + { + "epoch": 1.1891323084323826, + "grad_norm": 1.8671575784683228, + "learning_rate": 7.4491413219361915e-06, + "loss": 3.1736, + "step": 14833 + }, + { + "epoch": 1.1892124832134052, + "grad_norm": 1.7768217325210571, + "learning_rate": 7.447885851087269e-06, + "loss": 3.2743, + "step": 14834 + }, + { + "epoch": 1.1892926579944278, + "grad_norm": 1.8560340404510498, + "learning_rate": 7.44663042326609e-06, + "loss": 3.1636, + "step": 14835 + }, + { + "epoch": 1.1893728327754505, + "grad_norm": 1.708817481994629, + "learning_rate": 7.445375038493826e-06, + "loss": 3.269, + "step": 14836 + }, + { + "epoch": 1.189453007556473, + "grad_norm": 1.7750240564346313, + "learning_rate": 7.444119696791636e-06, + "loss": 3.1399, + "step": 14837 + }, + { + "epoch": 1.1895331823374957, + "grad_norm": 1.7474583387374878, + "learning_rate": 7.442864398180689e-06, + "loss": 3.1546, + "step": 14838 + }, + { + "epoch": 1.1896133571185183, + "grad_norm": 1.775041937828064, + "learning_rate": 7.441609142682145e-06, + "loss": 3.149, + "step": 14839 + }, + { + "epoch": 1.189693531899541, + "grad_norm": 1.8187135457992554, + "learning_rate": 7.440353930317169e-06, + "loss": 3.1379, + "step": 14840 + }, + { + "epoch": 1.1897737066805636, + "grad_norm": 1.761813759803772, + "learning_rate": 7.439098761106927e-06, + "loss": 3.1857, + "step": 14841 + }, + { + "epoch": 1.1898538814615862, + "grad_norm": 1.6899158954620361, + "learning_rate": 7.437843635072572e-06, + "loss": 3.1548, + "step": 14842 + }, + { + "epoch": 1.1899340562426088, + "grad_norm": 1.7719494104385376, + "learning_rate": 7.436588552235275e-06, + "loss": 3.1465, + "step": 14843 + }, + { + "epoch": 1.1900142310236315, + "grad_norm": 1.7454514503479004, + "learning_rate": 7.43533351261619e-06, + "loss": 3.0981, + "step": 14844 + }, + { + "epoch": 1.190094405804654, + "grad_norm": 1.770227074623108, + "learning_rate": 7.434078516236476e-06, + "loss": 3.1999, + "step": 14845 + }, + { + "epoch": 1.1901745805856767, + "grad_norm": 1.760598063468933, + "learning_rate": 7.432823563117297e-06, + "loss": 3.1723, + "step": 14846 + }, + { + "epoch": 1.1902547553666993, + "grad_norm": 1.8079681396484375, + "learning_rate": 7.431568653279806e-06, + "loss": 3.1785, + "step": 14847 + }, + { + "epoch": 1.190334930147722, + "grad_norm": 1.836591362953186, + "learning_rate": 7.430313786745165e-06, + "loss": 3.3328, + "step": 14848 + }, + { + "epoch": 1.1904151049287446, + "grad_norm": 1.7817319631576538, + "learning_rate": 7.429058963534525e-06, + "loss": 3.2866, + "step": 14849 + }, + { + "epoch": 1.1904952797097672, + "grad_norm": 1.8456532955169678, + "learning_rate": 7.427804183669045e-06, + "loss": 3.0992, + "step": 14850 + }, + { + "epoch": 1.1905754544907898, + "grad_norm": 1.656556487083435, + "learning_rate": 7.426549447169882e-06, + "loss": 3.1481, + "step": 14851 + }, + { + "epoch": 1.1906556292718125, + "grad_norm": 1.8003178834915161, + "learning_rate": 7.425294754058186e-06, + "loss": 3.1625, + "step": 14852 + }, + { + "epoch": 1.190735804052835, + "grad_norm": 1.6579383611679077, + "learning_rate": 7.424040104355116e-06, + "loss": 3.1759, + "step": 14853 + }, + { + "epoch": 1.1908159788338577, + "grad_norm": 1.9479008913040161, + "learning_rate": 7.422785498081819e-06, + "loss": 3.1638, + "step": 14854 + }, + { + "epoch": 1.1908961536148803, + "grad_norm": 1.787747859954834, + "learning_rate": 7.421530935259451e-06, + "loss": 3.1417, + "step": 14855 + }, + { + "epoch": 1.190976328395903, + "grad_norm": 1.8685475587844849, + "learning_rate": 7.420276415909164e-06, + "loss": 3.3289, + "step": 14856 + }, + { + "epoch": 1.1910565031769256, + "grad_norm": 1.8587439060211182, + "learning_rate": 7.419021940052104e-06, + "loss": 3.1469, + "step": 14857 + }, + { + "epoch": 1.1911366779579484, + "grad_norm": 1.6527825593948364, + "learning_rate": 7.41776750770943e-06, + "loss": 3.1044, + "step": 14858 + }, + { + "epoch": 1.191216852738971, + "grad_norm": 1.7705492973327637, + "learning_rate": 7.416513118902282e-06, + "loss": 3.1399, + "step": 14859 + }, + { + "epoch": 1.1912970275199937, + "grad_norm": 0.7149985432624817, + "learning_rate": 7.415258773651811e-06, + "loss": 2.4747, + "step": 14860 + }, + { + "epoch": 1.1913772023010163, + "grad_norm": 1.8624099493026733, + "learning_rate": 7.41400447197917e-06, + "loss": 3.2021, + "step": 14861 + }, + { + "epoch": 1.191457377082039, + "grad_norm": 0.7115876078605652, + "learning_rate": 7.4127502139054995e-06, + "loss": 2.4684, + "step": 14862 + }, + { + "epoch": 1.1915375518630615, + "grad_norm": 1.8913378715515137, + "learning_rate": 7.411495999451949e-06, + "loss": 3.2057, + "step": 14863 + }, + { + "epoch": 1.1916177266440842, + "grad_norm": 1.6629128456115723, + "learning_rate": 7.4102418286396635e-06, + "loss": 3.1363, + "step": 14864 + }, + { + "epoch": 1.1916979014251068, + "grad_norm": 1.7711766958236694, + "learning_rate": 7.408987701489789e-06, + "loss": 3.2306, + "step": 14865 + }, + { + "epoch": 1.1917780762061294, + "grad_norm": 1.8341835737228394, + "learning_rate": 7.40773361802347e-06, + "loss": 3.1281, + "step": 14866 + }, + { + "epoch": 1.191858250987152, + "grad_norm": 1.7320162057876587, + "learning_rate": 7.4064795782618445e-06, + "loss": 3.1979, + "step": 14867 + }, + { + "epoch": 1.1919384257681747, + "grad_norm": 1.843732237815857, + "learning_rate": 7.4052255822260635e-06, + "loss": 3.1713, + "step": 14868 + }, + { + "epoch": 1.1920186005491973, + "grad_norm": 1.82068932056427, + "learning_rate": 7.403971629937262e-06, + "loss": 3.1859, + "step": 14869 + }, + { + "epoch": 1.19209877533022, + "grad_norm": 1.8210794925689697, + "learning_rate": 7.402717721416583e-06, + "loss": 3.236, + "step": 14870 + }, + { + "epoch": 1.1921789501112425, + "grad_norm": 1.7884243726730347, + "learning_rate": 7.401463856685171e-06, + "loss": 3.2557, + "step": 14871 + }, + { + "epoch": 1.1922591248922652, + "grad_norm": 1.7447634935379028, + "learning_rate": 7.400210035764162e-06, + "loss": 3.1506, + "step": 14872 + }, + { + "epoch": 1.1923392996732878, + "grad_norm": 0.7245191335678101, + "learning_rate": 7.398956258674694e-06, + "loss": 2.4918, + "step": 14873 + }, + { + "epoch": 1.1924194744543104, + "grad_norm": 1.9023692607879639, + "learning_rate": 7.397702525437908e-06, + "loss": 3.2307, + "step": 14874 + }, + { + "epoch": 1.192499649235333, + "grad_norm": 1.806283712387085, + "learning_rate": 7.3964488360749384e-06, + "loss": 3.1186, + "step": 14875 + }, + { + "epoch": 1.1925798240163556, + "grad_norm": 1.7575249671936035, + "learning_rate": 7.395195190606927e-06, + "loss": 3.256, + "step": 14876 + }, + { + "epoch": 1.1926599987973783, + "grad_norm": 1.836775779724121, + "learning_rate": 7.393941589055003e-06, + "loss": 3.1296, + "step": 14877 + }, + { + "epoch": 1.192740173578401, + "grad_norm": 1.6960805654525757, + "learning_rate": 7.39268803144031e-06, + "loss": 3.1549, + "step": 14878 + }, + { + "epoch": 1.1928203483594235, + "grad_norm": 1.7234783172607422, + "learning_rate": 7.391434517783973e-06, + "loss": 3.1658, + "step": 14879 + }, + { + "epoch": 1.1929005231404461, + "grad_norm": 1.892206072807312, + "learning_rate": 7.3901810481071326e-06, + "loss": 3.1978, + "step": 14880 + }, + { + "epoch": 1.1929806979214688, + "grad_norm": 1.7092481851577759, + "learning_rate": 7.388927622430921e-06, + "loss": 3.0911, + "step": 14881 + }, + { + "epoch": 1.1930608727024914, + "grad_norm": 1.751921534538269, + "learning_rate": 7.387674240776468e-06, + "loss": 3.2335, + "step": 14882 + }, + { + "epoch": 1.193141047483514, + "grad_norm": 1.8815945386886597, + "learning_rate": 7.386420903164908e-06, + "loss": 3.2969, + "step": 14883 + }, + { + "epoch": 1.1932212222645366, + "grad_norm": 1.8039494752883911, + "learning_rate": 7.38516760961737e-06, + "loss": 3.1752, + "step": 14884 + }, + { + "epoch": 1.1933013970455593, + "grad_norm": 1.6965420246124268, + "learning_rate": 7.383914360154983e-06, + "loss": 3.2009, + "step": 14885 + }, + { + "epoch": 1.1933815718265819, + "grad_norm": 1.7275018692016602, + "learning_rate": 7.38266115479888e-06, + "loss": 3.109, + "step": 14886 + }, + { + "epoch": 1.1934617466076045, + "grad_norm": 1.7051634788513184, + "learning_rate": 7.381407993570185e-06, + "loss": 3.1626, + "step": 14887 + }, + { + "epoch": 1.1935419213886271, + "grad_norm": 1.9410909414291382, + "learning_rate": 7.380154876490031e-06, + "loss": 3.1734, + "step": 14888 + }, + { + "epoch": 1.1936220961696498, + "grad_norm": 1.8631222248077393, + "learning_rate": 7.378901803579541e-06, + "loss": 3.1862, + "step": 14889 + }, + { + "epoch": 1.1937022709506724, + "grad_norm": 1.750876545906067, + "learning_rate": 7.377648774859842e-06, + "loss": 3.1978, + "step": 14890 + }, + { + "epoch": 1.193782445731695, + "grad_norm": 1.739639163017273, + "learning_rate": 7.376395790352064e-06, + "loss": 3.1743, + "step": 14891 + }, + { + "epoch": 1.1938626205127176, + "grad_norm": 1.9679453372955322, + "learning_rate": 7.375142850077324e-06, + "loss": 3.1716, + "step": 14892 + }, + { + "epoch": 1.1939427952937405, + "grad_norm": 1.7854596376419067, + "learning_rate": 7.3738899540567525e-06, + "loss": 3.1451, + "step": 14893 + }, + { + "epoch": 1.194022970074763, + "grad_norm": 1.8615328073501587, + "learning_rate": 7.372637102311471e-06, + "loss": 3.1597, + "step": 14894 + }, + { + "epoch": 1.1941031448557857, + "grad_norm": 1.8664727210998535, + "learning_rate": 7.371384294862602e-06, + "loss": 3.1862, + "step": 14895 + }, + { + "epoch": 1.1941833196368083, + "grad_norm": 1.9038161039352417, + "learning_rate": 7.370131531731268e-06, + "loss": 3.1863, + "step": 14896 + }, + { + "epoch": 1.194263494417831, + "grad_norm": 1.8078722953796387, + "learning_rate": 7.368878812938587e-06, + "loss": 3.1632, + "step": 14897 + }, + { + "epoch": 1.1943436691988536, + "grad_norm": 1.8339486122131348, + "learning_rate": 7.367626138505684e-06, + "loss": 3.1757, + "step": 14898 + }, + { + "epoch": 1.1944238439798762, + "grad_norm": 1.8356695175170898, + "learning_rate": 7.366373508453677e-06, + "loss": 3.1237, + "step": 14899 + }, + { + "epoch": 1.1945040187608988, + "grad_norm": 1.9001684188842773, + "learning_rate": 7.365120922803682e-06, + "loss": 3.1329, + "step": 14900 + }, + { + "epoch": 1.1945841935419215, + "grad_norm": 1.7709708213806152, + "learning_rate": 7.363868381576823e-06, + "loss": 3.1218, + "step": 14901 + }, + { + "epoch": 1.194664368322944, + "grad_norm": 1.7908906936645508, + "learning_rate": 7.362615884794211e-06, + "loss": 3.2128, + "step": 14902 + }, + { + "epoch": 1.1947445431039667, + "grad_norm": 1.7529795169830322, + "learning_rate": 7.361363432476967e-06, + "loss": 3.1716, + "step": 14903 + }, + { + "epoch": 1.1948247178849893, + "grad_norm": 1.8071768283843994, + "learning_rate": 7.360111024646206e-06, + "loss": 3.2635, + "step": 14904 + }, + { + "epoch": 1.194904892666012, + "grad_norm": 1.798239827156067, + "learning_rate": 7.358858661323042e-06, + "loss": 3.2694, + "step": 14905 + }, + { + "epoch": 1.1949850674470346, + "grad_norm": 1.821570634841919, + "learning_rate": 7.357606342528592e-06, + "loss": 3.1257, + "step": 14906 + }, + { + "epoch": 1.1950652422280572, + "grad_norm": 1.7502259016036987, + "learning_rate": 7.356354068283964e-06, + "loss": 3.1647, + "step": 14907 + }, + { + "epoch": 1.1951454170090798, + "grad_norm": 1.9200479984283447, + "learning_rate": 7.3551018386102764e-06, + "loss": 3.2373, + "step": 14908 + }, + { + "epoch": 1.1952255917901025, + "grad_norm": 1.8006112575531006, + "learning_rate": 7.353849653528642e-06, + "loss": 3.1885, + "step": 14909 + }, + { + "epoch": 1.195305766571125, + "grad_norm": 1.7131379842758179, + "learning_rate": 7.352597513060167e-06, + "loss": 3.1815, + "step": 14910 + }, + { + "epoch": 1.1953859413521477, + "grad_norm": 1.7191917896270752, + "learning_rate": 7.3513454172259665e-06, + "loss": 3.2019, + "step": 14911 + }, + { + "epoch": 1.1954661161331703, + "grad_norm": 1.750205397605896, + "learning_rate": 7.3500933660471476e-06, + "loss": 3.0632, + "step": 14912 + }, + { + "epoch": 1.195546290914193, + "grad_norm": 1.715198278427124, + "learning_rate": 7.3488413595448205e-06, + "loss": 3.1427, + "step": 14913 + }, + { + "epoch": 1.1956264656952156, + "grad_norm": 1.7759580612182617, + "learning_rate": 7.347589397740096e-06, + "loss": 3.1507, + "step": 14914 + }, + { + "epoch": 1.1957066404762382, + "grad_norm": 2.067124843597412, + "learning_rate": 7.346337480654076e-06, + "loss": 3.1787, + "step": 14915 + }, + { + "epoch": 1.1957868152572608, + "grad_norm": 1.7917312383651733, + "learning_rate": 7.345085608307873e-06, + "loss": 3.1588, + "step": 14916 + }, + { + "epoch": 1.1958669900382835, + "grad_norm": 1.6901674270629883, + "learning_rate": 7.343833780722591e-06, + "loss": 3.2078, + "step": 14917 + }, + { + "epoch": 1.195947164819306, + "grad_norm": 1.6909219026565552, + "learning_rate": 7.342581997919334e-06, + "loss": 3.295, + "step": 14918 + }, + { + "epoch": 1.1960273396003287, + "grad_norm": 1.7882983684539795, + "learning_rate": 7.34133025991921e-06, + "loss": 3.2115, + "step": 14919 + }, + { + "epoch": 1.1961075143813513, + "grad_norm": 1.795937418937683, + "learning_rate": 7.3400785667433195e-06, + "loss": 3.2382, + "step": 14920 + }, + { + "epoch": 1.196187689162374, + "grad_norm": 1.7376161813735962, + "learning_rate": 7.338826918412769e-06, + "loss": 3.0995, + "step": 14921 + }, + { + "epoch": 1.1962678639433966, + "grad_norm": 1.7434865236282349, + "learning_rate": 7.337575314948655e-06, + "loss": 3.189, + "step": 14922 + }, + { + "epoch": 1.1963480387244192, + "grad_norm": 1.7743213176727295, + "learning_rate": 7.336323756372084e-06, + "loss": 3.0895, + "step": 14923 + }, + { + "epoch": 1.1964282135054418, + "grad_norm": 1.6812381744384766, + "learning_rate": 7.335072242704157e-06, + "loss": 3.1479, + "step": 14924 + }, + { + "epoch": 1.1965083882864644, + "grad_norm": 1.7491528987884521, + "learning_rate": 7.333820773965973e-06, + "loss": 3.1259, + "step": 14925 + }, + { + "epoch": 1.196588563067487, + "grad_norm": 1.7438057661056519, + "learning_rate": 7.332569350178631e-06, + "loss": 3.1332, + "step": 14926 + }, + { + "epoch": 1.1966687378485097, + "grad_norm": 1.8447487354278564, + "learning_rate": 7.331317971363228e-06, + "loss": 3.1661, + "step": 14927 + }, + { + "epoch": 1.1967489126295323, + "grad_norm": 1.7608510255813599, + "learning_rate": 7.330066637540865e-06, + "loss": 3.1565, + "step": 14928 + }, + { + "epoch": 1.196829087410555, + "grad_norm": 1.867089033126831, + "learning_rate": 7.328815348732637e-06, + "loss": 3.1891, + "step": 14929 + }, + { + "epoch": 1.1969092621915776, + "grad_norm": 1.8661712408065796, + "learning_rate": 7.32756410495964e-06, + "loss": 3.1918, + "step": 14930 + }, + { + "epoch": 1.1969894369726002, + "grad_norm": 1.8259146213531494, + "learning_rate": 7.326312906242973e-06, + "loss": 3.2833, + "step": 14931 + }, + { + "epoch": 1.1970696117536228, + "grad_norm": 1.7528352737426758, + "learning_rate": 7.325061752603725e-06, + "loss": 3.1544, + "step": 14932 + }, + { + "epoch": 1.1971497865346454, + "grad_norm": 1.7587388753890991, + "learning_rate": 7.323810644062994e-06, + "loss": 3.1931, + "step": 14933 + }, + { + "epoch": 1.197229961315668, + "grad_norm": 1.784918189048767, + "learning_rate": 7.3225595806418734e-06, + "loss": 3.1146, + "step": 14934 + }, + { + "epoch": 1.1973101360966907, + "grad_norm": 1.7361680269241333, + "learning_rate": 7.3213085623614536e-06, + "loss": 3.1475, + "step": 14935 + }, + { + "epoch": 1.1973903108777133, + "grad_norm": 1.7356942892074585, + "learning_rate": 7.320057589242828e-06, + "loss": 3.0994, + "step": 14936 + }, + { + "epoch": 1.197470485658736, + "grad_norm": 1.8421097993850708, + "learning_rate": 7.318806661307086e-06, + "loss": 3.1981, + "step": 14937 + }, + { + "epoch": 1.1975506604397586, + "grad_norm": 1.6388834714889526, + "learning_rate": 7.3175557785753195e-06, + "loss": 3.1436, + "step": 14938 + }, + { + "epoch": 1.1976308352207814, + "grad_norm": 1.7557029724121094, + "learning_rate": 7.316304941068618e-06, + "loss": 3.1318, + "step": 14939 + }, + { + "epoch": 1.197711010001804, + "grad_norm": 1.656128168106079, + "learning_rate": 7.3150541488080675e-06, + "loss": 3.1599, + "step": 14940 + }, + { + "epoch": 1.1977911847828266, + "grad_norm": 1.7358219623565674, + "learning_rate": 7.313803401814761e-06, + "loss": 3.1328, + "step": 14941 + }, + { + "epoch": 1.1978713595638493, + "grad_norm": 1.8472020626068115, + "learning_rate": 7.312552700109779e-06, + "loss": 3.2018, + "step": 14942 + }, + { + "epoch": 1.197951534344872, + "grad_norm": 1.7099463939666748, + "learning_rate": 7.311302043714211e-06, + "loss": 3.0887, + "step": 14943 + }, + { + "epoch": 1.1980317091258945, + "grad_norm": 1.7776433229446411, + "learning_rate": 7.310051432649145e-06, + "loss": 3.238, + "step": 14944 + }, + { + "epoch": 1.1981118839069171, + "grad_norm": 2.001375913619995, + "learning_rate": 7.308800866935664e-06, + "loss": 3.2396, + "step": 14945 + }, + { + "epoch": 1.1981920586879398, + "grad_norm": 1.806295394897461, + "learning_rate": 7.307550346594852e-06, + "loss": 3.1484, + "step": 14946 + }, + { + "epoch": 1.1982722334689624, + "grad_norm": 1.8795089721679688, + "learning_rate": 7.30629987164779e-06, + "loss": 3.2291, + "step": 14947 + }, + { + "epoch": 1.198352408249985, + "grad_norm": 1.835487723350525, + "learning_rate": 7.305049442115564e-06, + "loss": 3.2108, + "step": 14948 + }, + { + "epoch": 1.1984325830310076, + "grad_norm": 0.7622061967849731, + "learning_rate": 7.3037990580192565e-06, + "loss": 2.47, + "step": 14949 + }, + { + "epoch": 1.1985127578120303, + "grad_norm": 1.78109872341156, + "learning_rate": 7.302548719379944e-06, + "loss": 3.166, + "step": 14950 + }, + { + "epoch": 1.1985929325930529, + "grad_norm": 1.845467209815979, + "learning_rate": 7.301298426218711e-06, + "loss": 3.2208, + "step": 14951 + }, + { + "epoch": 1.1986731073740755, + "grad_norm": 1.80046546459198, + "learning_rate": 7.300048178556635e-06, + "loss": 3.2579, + "step": 14952 + }, + { + "epoch": 1.1987532821550981, + "grad_norm": 1.8183003664016724, + "learning_rate": 7.2987979764147935e-06, + "loss": 3.1799, + "step": 14953 + }, + { + "epoch": 1.1988334569361208, + "grad_norm": 1.7826988697052002, + "learning_rate": 7.297547819814269e-06, + "loss": 3.197, + "step": 14954 + }, + { + "epoch": 1.1989136317171434, + "grad_norm": 1.7994636297225952, + "learning_rate": 7.296297708776134e-06, + "loss": 3.1012, + "step": 14955 + }, + { + "epoch": 1.198993806498166, + "grad_norm": 1.8133951425552368, + "learning_rate": 7.295047643321469e-06, + "loss": 3.186, + "step": 14956 + }, + { + "epoch": 1.1990739812791886, + "grad_norm": 1.7502613067626953, + "learning_rate": 7.293797623471344e-06, + "loss": 3.192, + "step": 14957 + }, + { + "epoch": 1.1991541560602113, + "grad_norm": 1.7698118686676025, + "learning_rate": 7.2925476492468385e-06, + "loss": 3.1461, + "step": 14958 + }, + { + "epoch": 1.1992343308412339, + "grad_norm": 1.7657524347305298, + "learning_rate": 7.291297720669026e-06, + "loss": 3.2288, + "step": 14959 + }, + { + "epoch": 1.1993145056222565, + "grad_norm": 1.8396432399749756, + "learning_rate": 7.290047837758979e-06, + "loss": 3.1526, + "step": 14960 + }, + { + "epoch": 1.1993946804032791, + "grad_norm": 1.7395970821380615, + "learning_rate": 7.288798000537773e-06, + "loss": 3.0637, + "step": 14961 + }, + { + "epoch": 1.1994748551843017, + "grad_norm": 1.8225736618041992, + "learning_rate": 7.287548209026474e-06, + "loss": 3.1179, + "step": 14962 + }, + { + "epoch": 1.1995550299653244, + "grad_norm": 1.8487178087234497, + "learning_rate": 7.2862984632461555e-06, + "loss": 3.3474, + "step": 14963 + }, + { + "epoch": 1.199635204746347, + "grad_norm": 1.860506296157837, + "learning_rate": 7.2850487632178904e-06, + "loss": 3.2086, + "step": 14964 + }, + { + "epoch": 1.1997153795273696, + "grad_norm": 1.7999814748764038, + "learning_rate": 7.2837991089627455e-06, + "loss": 3.2787, + "step": 14965 + }, + { + "epoch": 1.1997955543083922, + "grad_norm": 1.7761292457580566, + "learning_rate": 7.282549500501792e-06, + "loss": 3.1445, + "step": 14966 + }, + { + "epoch": 1.1998757290894149, + "grad_norm": 1.7451516389846802, + "learning_rate": 7.281299937856095e-06, + "loss": 3.1627, + "step": 14967 + }, + { + "epoch": 1.1999559038704375, + "grad_norm": 1.775966763496399, + "learning_rate": 7.28005042104672e-06, + "loss": 3.2927, + "step": 14968 + }, + { + "epoch": 1.2000360786514601, + "grad_norm": 1.7716538906097412, + "learning_rate": 7.278800950094742e-06, + "loss": 3.144, + "step": 14969 + }, + { + "epoch": 1.2001162534324827, + "grad_norm": 1.846742033958435, + "learning_rate": 7.277551525021216e-06, + "loss": 3.2156, + "step": 14970 + }, + { + "epoch": 1.2001964282135054, + "grad_norm": 1.8921984434127808, + "learning_rate": 7.276302145847216e-06, + "loss": 3.1714, + "step": 14971 + }, + { + "epoch": 1.200276602994528, + "grad_norm": 0.6882405281066895, + "learning_rate": 7.275052812593799e-06, + "loss": 2.4812, + "step": 14972 + }, + { + "epoch": 1.2003567777755506, + "grad_norm": 1.7138993740081787, + "learning_rate": 7.27380352528203e-06, + "loss": 3.0795, + "step": 14973 + }, + { + "epoch": 1.2004369525565732, + "grad_norm": 1.9042696952819824, + "learning_rate": 7.272554283932976e-06, + "loss": 3.1302, + "step": 14974 + }, + { + "epoch": 1.200517127337596, + "grad_norm": 1.7310391664505005, + "learning_rate": 7.271305088567695e-06, + "loss": 3.229, + "step": 14975 + }, + { + "epoch": 1.2005973021186187, + "grad_norm": 1.7259433269500732, + "learning_rate": 7.270055939207248e-06, + "loss": 3.1179, + "step": 14976 + }, + { + "epoch": 1.2006774768996413, + "grad_norm": 0.6674362421035767, + "learning_rate": 7.268806835872695e-06, + "loss": 2.479, + "step": 14977 + }, + { + "epoch": 1.200757651680664, + "grad_norm": 1.759141445159912, + "learning_rate": 7.267557778585096e-06, + "loss": 3.1041, + "step": 14978 + }, + { + "epoch": 1.2008378264616866, + "grad_norm": 1.8495614528656006, + "learning_rate": 7.266308767365511e-06, + "loss": 3.2367, + "step": 14979 + }, + { + "epoch": 1.2009180012427092, + "grad_norm": 1.7370591163635254, + "learning_rate": 7.265059802234996e-06, + "loss": 3.2145, + "step": 14980 + }, + { + "epoch": 1.2009981760237318, + "grad_norm": 1.7228913307189941, + "learning_rate": 7.263810883214611e-06, + "loss": 3.1724, + "step": 14981 + }, + { + "epoch": 1.2010783508047544, + "grad_norm": 1.614403247833252, + "learning_rate": 7.262562010325407e-06, + "loss": 3.2645, + "step": 14982 + }, + { + "epoch": 1.201158525585777, + "grad_norm": 1.7664430141448975, + "learning_rate": 7.2613131835884434e-06, + "loss": 3.1479, + "step": 14983 + }, + { + "epoch": 1.2012387003667997, + "grad_norm": 1.796491265296936, + "learning_rate": 7.260064403024776e-06, + "loss": 3.167, + "step": 14984 + }, + { + "epoch": 1.2013188751478223, + "grad_norm": 1.8165453672409058, + "learning_rate": 7.258815668655455e-06, + "loss": 3.1885, + "step": 14985 + }, + { + "epoch": 1.201399049928845, + "grad_norm": 1.7710886001586914, + "learning_rate": 7.257566980501538e-06, + "loss": 3.1979, + "step": 14986 + }, + { + "epoch": 1.2014792247098676, + "grad_norm": 1.8062477111816406, + "learning_rate": 7.256318338584074e-06, + "loss": 3.0748, + "step": 14987 + }, + { + "epoch": 1.2015593994908902, + "grad_norm": 1.846814751625061, + "learning_rate": 7.2550697429241146e-06, + "loss": 3.1406, + "step": 14988 + }, + { + "epoch": 1.2016395742719128, + "grad_norm": 1.8180768489837646, + "learning_rate": 7.2538211935427135e-06, + "loss": 3.2416, + "step": 14989 + }, + { + "epoch": 1.2017197490529354, + "grad_norm": 1.7822489738464355, + "learning_rate": 7.252572690460919e-06, + "loss": 3.1935, + "step": 14990 + }, + { + "epoch": 1.201799923833958, + "grad_norm": 1.7417598962783813, + "learning_rate": 7.251324233699781e-06, + "loss": 3.191, + "step": 14991 + }, + { + "epoch": 1.2018800986149807, + "grad_norm": 1.804019808769226, + "learning_rate": 7.250075823280347e-06, + "loss": 3.2159, + "step": 14992 + }, + { + "epoch": 1.2019602733960033, + "grad_norm": 1.784031629562378, + "learning_rate": 7.2488274592236655e-06, + "loss": 3.1548, + "step": 14993 + }, + { + "epoch": 1.202040448177026, + "grad_norm": 0.681205153465271, + "learning_rate": 7.247579141550784e-06, + "loss": 2.4686, + "step": 14994 + }, + { + "epoch": 1.2021206229580486, + "grad_norm": 1.8466334342956543, + "learning_rate": 7.246330870282747e-06, + "loss": 3.1494, + "step": 14995 + }, + { + "epoch": 1.2022007977390712, + "grad_norm": 0.6847318410873413, + "learning_rate": 7.2450826454406044e-06, + "loss": 2.4639, + "step": 14996 + }, + { + "epoch": 1.2022809725200938, + "grad_norm": 1.7429814338684082, + "learning_rate": 7.243834467045393e-06, + "loss": 3.1709, + "step": 14997 + }, + { + "epoch": 1.2023611473011164, + "grad_norm": 1.722590684890747, + "learning_rate": 7.242586335118163e-06, + "loss": 3.1072, + "step": 14998 + }, + { + "epoch": 1.202441322082139, + "grad_norm": 0.6606228351593018, + "learning_rate": 7.241338249679956e-06, + "loss": 2.4818, + "step": 14999 + }, + { + "epoch": 1.2025214968631617, + "grad_norm": 1.7401454448699951, + "learning_rate": 7.240090210751813e-06, + "loss": 3.1952, + "step": 15000 + }, + { + "epoch": 1.2026016716441843, + "grad_norm": 1.656770944595337, + "learning_rate": 7.238842218354778e-06, + "loss": 3.1401, + "step": 15001 + }, + { + "epoch": 1.202681846425207, + "grad_norm": 1.958642601966858, + "learning_rate": 7.237594272509888e-06, + "loss": 3.2354, + "step": 15002 + }, + { + "epoch": 1.2027620212062295, + "grad_norm": 1.7268413305282593, + "learning_rate": 7.236346373238186e-06, + "loss": 3.2156, + "step": 15003 + }, + { + "epoch": 1.2028421959872522, + "grad_norm": 1.7105162143707275, + "learning_rate": 7.2350985205607105e-06, + "loss": 3.1234, + "step": 15004 + }, + { + "epoch": 1.2029223707682748, + "grad_norm": 1.788933515548706, + "learning_rate": 7.233850714498498e-06, + "loss": 3.2494, + "step": 15005 + }, + { + "epoch": 1.2030025455492974, + "grad_norm": 0.654455304145813, + "learning_rate": 7.23260295507259e-06, + "loss": 2.4702, + "step": 15006 + }, + { + "epoch": 1.20308272033032, + "grad_norm": 1.7748916149139404, + "learning_rate": 7.231355242304018e-06, + "loss": 3.1954, + "step": 15007 + }, + { + "epoch": 1.2031628951113427, + "grad_norm": 1.9931650161743164, + "learning_rate": 7.230107576213822e-06, + "loss": 3.236, + "step": 15008 + }, + { + "epoch": 1.2032430698923653, + "grad_norm": 1.7577502727508545, + "learning_rate": 7.228859956823038e-06, + "loss": 3.2273, + "step": 15009 + }, + { + "epoch": 1.203323244673388, + "grad_norm": 1.7605153322219849, + "learning_rate": 7.2276123841526956e-06, + "loss": 3.2489, + "step": 15010 + }, + { + "epoch": 1.2034034194544105, + "grad_norm": 1.6756432056427002, + "learning_rate": 7.226364858223833e-06, + "loss": 3.1022, + "step": 15011 + }, + { + "epoch": 1.2034835942354332, + "grad_norm": 1.9669122695922852, + "learning_rate": 7.225117379057481e-06, + "loss": 3.2523, + "step": 15012 + }, + { + "epoch": 1.2035637690164558, + "grad_norm": 1.712528109550476, + "learning_rate": 7.223869946674671e-06, + "loss": 3.1177, + "step": 15013 + }, + { + "epoch": 1.2036439437974784, + "grad_norm": 1.702194333076477, + "learning_rate": 7.222622561096437e-06, + "loss": 3.1852, + "step": 15014 + }, + { + "epoch": 1.203724118578501, + "grad_norm": 1.740496277809143, + "learning_rate": 7.221375222343806e-06, + "loss": 3.1865, + "step": 15015 + }, + { + "epoch": 1.2038042933595237, + "grad_norm": 1.730543613433838, + "learning_rate": 7.220127930437813e-06, + "loss": 3.197, + "step": 15016 + }, + { + "epoch": 1.2038844681405463, + "grad_norm": 1.7519185543060303, + "learning_rate": 7.218880685399478e-06, + "loss": 3.0766, + "step": 15017 + }, + { + "epoch": 1.203964642921569, + "grad_norm": 1.7710765600204468, + "learning_rate": 7.217633487249838e-06, + "loss": 3.1201, + "step": 15018 + }, + { + "epoch": 1.2040448177025915, + "grad_norm": 1.7563316822052002, + "learning_rate": 7.2163863360099175e-06, + "loss": 3.152, + "step": 15019 + }, + { + "epoch": 1.2041249924836144, + "grad_norm": 1.8479911088943481, + "learning_rate": 7.21513923170074e-06, + "loss": 3.2243, + "step": 15020 + }, + { + "epoch": 1.204205167264637, + "grad_norm": 1.797463297843933, + "learning_rate": 7.213892174343336e-06, + "loss": 3.2702, + "step": 15021 + }, + { + "epoch": 1.2042853420456596, + "grad_norm": 1.649598479270935, + "learning_rate": 7.212645163958725e-06, + "loss": 3.1831, + "step": 15022 + }, + { + "epoch": 1.2043655168266822, + "grad_norm": 1.7845042943954468, + "learning_rate": 7.211398200567935e-06, + "loss": 3.1618, + "step": 15023 + }, + { + "epoch": 1.2044456916077049, + "grad_norm": 1.7022440433502197, + "learning_rate": 7.210151284191991e-06, + "loss": 3.1389, + "step": 15024 + }, + { + "epoch": 1.2045258663887275, + "grad_norm": 1.7756333351135254, + "learning_rate": 7.208904414851909e-06, + "loss": 3.2192, + "step": 15025 + }, + { + "epoch": 1.2046060411697501, + "grad_norm": 1.7151494026184082, + "learning_rate": 7.20765759256872e-06, + "loss": 3.1991, + "step": 15026 + }, + { + "epoch": 1.2046862159507727, + "grad_norm": 1.7119406461715698, + "learning_rate": 7.206410817363436e-06, + "loss": 3.1368, + "step": 15027 + }, + { + "epoch": 1.2047663907317954, + "grad_norm": 0.7182565927505493, + "learning_rate": 7.205164089257082e-06, + "loss": 2.4808, + "step": 15028 + }, + { + "epoch": 1.204846565512818, + "grad_norm": 1.8019863367080688, + "learning_rate": 7.203917408270677e-06, + "loss": 3.1595, + "step": 15029 + }, + { + "epoch": 1.2049267402938406, + "grad_norm": 1.9014610052108765, + "learning_rate": 7.202670774425237e-06, + "loss": 3.2002, + "step": 15030 + }, + { + "epoch": 1.2050069150748632, + "grad_norm": 1.7055532932281494, + "learning_rate": 7.201424187741786e-06, + "loss": 3.1875, + "step": 15031 + }, + { + "epoch": 1.2050870898558859, + "grad_norm": 1.828752040863037, + "learning_rate": 7.200177648241334e-06, + "loss": 3.1828, + "step": 15032 + }, + { + "epoch": 1.2051672646369085, + "grad_norm": 1.9567396640777588, + "learning_rate": 7.198931155944899e-06, + "loss": 3.1763, + "step": 15033 + }, + { + "epoch": 1.205247439417931, + "grad_norm": 1.6819833517074585, + "learning_rate": 7.1976847108735006e-06, + "loss": 3.2177, + "step": 15034 + }, + { + "epoch": 1.2053276141989537, + "grad_norm": 1.759716272354126, + "learning_rate": 7.1964383130481485e-06, + "loss": 3.1699, + "step": 15035 + }, + { + "epoch": 1.2054077889799764, + "grad_norm": 1.799759864807129, + "learning_rate": 7.19519196248986e-06, + "loss": 3.192, + "step": 15036 + }, + { + "epoch": 1.205487963760999, + "grad_norm": 1.7764267921447754, + "learning_rate": 7.193945659219644e-06, + "loss": 3.148, + "step": 15037 + }, + { + "epoch": 1.2055681385420216, + "grad_norm": 1.782222032546997, + "learning_rate": 7.192699403258516e-06, + "loss": 3.179, + "step": 15038 + }, + { + "epoch": 1.2056483133230442, + "grad_norm": 1.7614959478378296, + "learning_rate": 7.191453194627488e-06, + "loss": 3.1458, + "step": 15039 + }, + { + "epoch": 1.2057284881040669, + "grad_norm": 1.849298119544983, + "learning_rate": 7.190207033347567e-06, + "loss": 3.2431, + "step": 15040 + }, + { + "epoch": 1.2058086628850895, + "grad_norm": 1.8506088256835938, + "learning_rate": 7.188960919439769e-06, + "loss": 3.2786, + "step": 15041 + }, + { + "epoch": 1.205888837666112, + "grad_norm": 1.802404522895813, + "learning_rate": 7.187714852925095e-06, + "loss": 3.153, + "step": 15042 + }, + { + "epoch": 1.2059690124471347, + "grad_norm": 1.7700475454330444, + "learning_rate": 7.186468833824558e-06, + "loss": 3.2061, + "step": 15043 + }, + { + "epoch": 1.2060491872281573, + "grad_norm": 0.7089138627052307, + "learning_rate": 7.185222862159168e-06, + "loss": 2.4678, + "step": 15044 + }, + { + "epoch": 1.20612936200918, + "grad_norm": 1.7672868967056274, + "learning_rate": 7.183976937949926e-06, + "loss": 3.1421, + "step": 15045 + }, + { + "epoch": 1.2062095367902026, + "grad_norm": 1.8349374532699585, + "learning_rate": 7.182731061217842e-06, + "loss": 3.2231, + "step": 15046 + }, + { + "epoch": 1.2062897115712252, + "grad_norm": 1.730258822441101, + "learning_rate": 7.181485231983918e-06, + "loss": 3.2293, + "step": 15047 + }, + { + "epoch": 1.2063698863522478, + "grad_norm": 1.7224541902542114, + "learning_rate": 7.180239450269158e-06, + "loss": 3.1334, + "step": 15048 + }, + { + "epoch": 1.2064500611332705, + "grad_norm": 0.7080786824226379, + "learning_rate": 7.178993716094571e-06, + "loss": 2.4656, + "step": 15049 + }, + { + "epoch": 1.206530235914293, + "grad_norm": 1.6441279649734497, + "learning_rate": 7.177748029481153e-06, + "loss": 3.1862, + "step": 15050 + }, + { + "epoch": 1.2066104106953157, + "grad_norm": 1.7739146947860718, + "learning_rate": 7.176502390449911e-06, + "loss": 3.2338, + "step": 15051 + }, + { + "epoch": 1.2066905854763383, + "grad_norm": 0.6933513879776001, + "learning_rate": 7.175256799021841e-06, + "loss": 2.4827, + "step": 15052 + }, + { + "epoch": 1.206770760257361, + "grad_norm": 1.7343231439590454, + "learning_rate": 7.174011255217945e-06, + "loss": 3.2642, + "step": 15053 + }, + { + "epoch": 1.2068509350383836, + "grad_norm": 0.6589087843894958, + "learning_rate": 7.172765759059228e-06, + "loss": 2.4675, + "step": 15054 + }, + { + "epoch": 1.2069311098194062, + "grad_norm": 1.8045637607574463, + "learning_rate": 7.1715203105666795e-06, + "loss": 3.2939, + "step": 15055 + }, + { + "epoch": 1.207011284600429, + "grad_norm": 1.7728612422943115, + "learning_rate": 7.170274909761305e-06, + "loss": 3.1694, + "step": 15056 + }, + { + "epoch": 1.2070914593814517, + "grad_norm": 1.8262120485305786, + "learning_rate": 7.169029556664095e-06, + "loss": 3.165, + "step": 15057 + }, + { + "epoch": 1.2071716341624743, + "grad_norm": 1.7597023248672485, + "learning_rate": 7.167784251296049e-06, + "loss": 3.1843, + "step": 15058 + }, + { + "epoch": 1.207251808943497, + "grad_norm": 0.6989243626594543, + "learning_rate": 7.1665389936781644e-06, + "loss": 2.4872, + "step": 15059 + }, + { + "epoch": 1.2073319837245196, + "grad_norm": 1.7705233097076416, + "learning_rate": 7.165293783831433e-06, + "loss": 3.1765, + "step": 15060 + }, + { + "epoch": 1.2074121585055422, + "grad_norm": 1.9405368566513062, + "learning_rate": 7.164048621776852e-06, + "loss": 3.2982, + "step": 15061 + }, + { + "epoch": 1.2074923332865648, + "grad_norm": 1.6954408884048462, + "learning_rate": 7.162803507535407e-06, + "loss": 3.1503, + "step": 15062 + }, + { + "epoch": 1.2075725080675874, + "grad_norm": 1.7709287405014038, + "learning_rate": 7.161558441128097e-06, + "loss": 3.1794, + "step": 15063 + }, + { + "epoch": 1.20765268284861, + "grad_norm": 1.6994428634643555, + "learning_rate": 7.160313422575913e-06, + "loss": 3.1953, + "step": 15064 + }, + { + "epoch": 1.2077328576296327, + "grad_norm": 1.7822438478469849, + "learning_rate": 7.159068451899841e-06, + "loss": 3.2242, + "step": 15065 + }, + { + "epoch": 1.2078130324106553, + "grad_norm": 0.6884745359420776, + "learning_rate": 7.157823529120876e-06, + "loss": 2.4886, + "step": 15066 + }, + { + "epoch": 1.207893207191678, + "grad_norm": 1.6602486371994019, + "learning_rate": 7.156578654260003e-06, + "loss": 3.0834, + "step": 15067 + }, + { + "epoch": 1.2079733819727005, + "grad_norm": 1.7696894407272339, + "learning_rate": 7.155333827338211e-06, + "loss": 3.1656, + "step": 15068 + }, + { + "epoch": 1.2080535567537232, + "grad_norm": 1.7878904342651367, + "learning_rate": 7.154089048376492e-06, + "loss": 3.157, + "step": 15069 + }, + { + "epoch": 1.2081337315347458, + "grad_norm": 1.8477082252502441, + "learning_rate": 7.152844317395824e-06, + "loss": 3.2092, + "step": 15070 + }, + { + "epoch": 1.2082139063157684, + "grad_norm": 1.7399369478225708, + "learning_rate": 7.1515996344172014e-06, + "loss": 3.0713, + "step": 15071 + }, + { + "epoch": 1.208294081096791, + "grad_norm": 1.775194764137268, + "learning_rate": 7.150354999461602e-06, + "loss": 3.1901, + "step": 15072 + }, + { + "epoch": 1.2083742558778137, + "grad_norm": 1.7208480834960938, + "learning_rate": 7.149110412550013e-06, + "loss": 3.1461, + "step": 15073 + }, + { + "epoch": 1.2084544306588363, + "grad_norm": 1.9282554388046265, + "learning_rate": 7.147865873703421e-06, + "loss": 3.1468, + "step": 15074 + }, + { + "epoch": 1.208534605439859, + "grad_norm": 1.9718929529190063, + "learning_rate": 7.146621382942802e-06, + "loss": 3.1421, + "step": 15075 + }, + { + "epoch": 1.2086147802208815, + "grad_norm": 0.7092657685279846, + "learning_rate": 7.145376940289144e-06, + "loss": 2.4935, + "step": 15076 + }, + { + "epoch": 1.2086949550019042, + "grad_norm": 1.7959917783737183, + "learning_rate": 7.144132545763421e-06, + "loss": 3.1315, + "step": 15077 + }, + { + "epoch": 1.2087751297829268, + "grad_norm": 1.7600467205047607, + "learning_rate": 7.142888199386617e-06, + "loss": 3.1824, + "step": 15078 + }, + { + "epoch": 1.2088553045639494, + "grad_norm": 1.8429551124572754, + "learning_rate": 7.141643901179714e-06, + "loss": 3.3304, + "step": 15079 + }, + { + "epoch": 1.208935479344972, + "grad_norm": 1.8740503787994385, + "learning_rate": 7.140399651163686e-06, + "loss": 3.2407, + "step": 15080 + }, + { + "epoch": 1.2090156541259947, + "grad_norm": 1.861710548400879, + "learning_rate": 7.139155449359513e-06, + "loss": 3.2637, + "step": 15081 + }, + { + "epoch": 1.2090958289070173, + "grad_norm": 1.921811580657959, + "learning_rate": 7.137911295788169e-06, + "loss": 3.2399, + "step": 15082 + }, + { + "epoch": 1.20917600368804, + "grad_norm": 1.6961115598678589, + "learning_rate": 7.136667190470632e-06, + "loss": 3.1573, + "step": 15083 + }, + { + "epoch": 1.2092561784690625, + "grad_norm": 1.8039358854293823, + "learning_rate": 7.135423133427879e-06, + "loss": 3.1549, + "step": 15084 + }, + { + "epoch": 1.2093363532500851, + "grad_norm": 1.7677195072174072, + "learning_rate": 7.13417912468088e-06, + "loss": 3.3113, + "step": 15085 + }, + { + "epoch": 1.2094165280311078, + "grad_norm": 1.744174599647522, + "learning_rate": 7.132935164250612e-06, + "loss": 3.2605, + "step": 15086 + }, + { + "epoch": 1.2094967028121304, + "grad_norm": 1.7754180431365967, + "learning_rate": 7.131691252158047e-06, + "loss": 3.205, + "step": 15087 + }, + { + "epoch": 1.209576877593153, + "grad_norm": 1.8480596542358398, + "learning_rate": 7.130447388424155e-06, + "loss": 3.1902, + "step": 15088 + }, + { + "epoch": 1.2096570523741756, + "grad_norm": 1.7262780666351318, + "learning_rate": 7.129203573069912e-06, + "loss": 3.1594, + "step": 15089 + }, + { + "epoch": 1.2097372271551983, + "grad_norm": 1.7730379104614258, + "learning_rate": 7.127959806116283e-06, + "loss": 3.1313, + "step": 15090 + }, + { + "epoch": 1.209817401936221, + "grad_norm": 1.8659400939941406, + "learning_rate": 7.126716087584238e-06, + "loss": 3.2532, + "step": 15091 + }, + { + "epoch": 1.2098975767172435, + "grad_norm": 1.7003039121627808, + "learning_rate": 7.125472417494749e-06, + "loss": 3.2183, + "step": 15092 + }, + { + "epoch": 1.2099777514982661, + "grad_norm": 0.701518177986145, + "learning_rate": 7.124228795868782e-06, + "loss": 2.478, + "step": 15093 + }, + { + "epoch": 1.2100579262792888, + "grad_norm": 1.7539234161376953, + "learning_rate": 7.122985222727303e-06, + "loss": 3.2464, + "step": 15094 + }, + { + "epoch": 1.2101381010603114, + "grad_norm": 1.8467849493026733, + "learning_rate": 7.12174169809128e-06, + "loss": 3.148, + "step": 15095 + }, + { + "epoch": 1.210218275841334, + "grad_norm": 1.7619749307632446, + "learning_rate": 7.1204982219816765e-06, + "loss": 3.1462, + "step": 15096 + }, + { + "epoch": 1.2102984506223566, + "grad_norm": 1.6368472576141357, + "learning_rate": 7.11925479441946e-06, + "loss": 3.2081, + "step": 15097 + }, + { + "epoch": 1.2103786254033793, + "grad_norm": 1.7850171327590942, + "learning_rate": 7.11801141542559e-06, + "loss": 3.1381, + "step": 15098 + }, + { + "epoch": 1.2104588001844019, + "grad_norm": 1.7611804008483887, + "learning_rate": 7.116768085021036e-06, + "loss": 3.1342, + "step": 15099 + }, + { + "epoch": 1.2105389749654245, + "grad_norm": 1.7145617008209229, + "learning_rate": 7.115524803226751e-06, + "loss": 3.102, + "step": 15100 + }, + { + "epoch": 1.2106191497464471, + "grad_norm": 1.646519660949707, + "learning_rate": 7.1142815700637025e-06, + "loss": 3.1366, + "step": 15101 + }, + { + "epoch": 1.21069932452747, + "grad_norm": 1.7791856527328491, + "learning_rate": 7.113038385552852e-06, + "loss": 3.1838, + "step": 15102 + }, + { + "epoch": 1.2107794993084926, + "grad_norm": 1.9411585330963135, + "learning_rate": 7.1117952497151544e-06, + "loss": 3.1845, + "step": 15103 + }, + { + "epoch": 1.2108596740895152, + "grad_norm": 1.6153478622436523, + "learning_rate": 7.110552162571574e-06, + "loss": 3.0862, + "step": 15104 + }, + { + "epoch": 1.2109398488705378, + "grad_norm": 1.741989016532898, + "learning_rate": 7.1093091241430625e-06, + "loss": 3.1503, + "step": 15105 + }, + { + "epoch": 1.2110200236515605, + "grad_norm": 1.9839656352996826, + "learning_rate": 7.108066134450581e-06, + "loss": 3.2158, + "step": 15106 + }, + { + "epoch": 1.211100198432583, + "grad_norm": 1.8234946727752686, + "learning_rate": 7.106823193515087e-06, + "loss": 3.2308, + "step": 15107 + }, + { + "epoch": 1.2111803732136057, + "grad_norm": 1.7332158088684082, + "learning_rate": 7.1055803013575316e-06, + "loss": 3.1214, + "step": 15108 + }, + { + "epoch": 1.2112605479946283, + "grad_norm": 1.7105000019073486, + "learning_rate": 7.104337457998875e-06, + "loss": 3.1419, + "step": 15109 + }, + { + "epoch": 1.211340722775651, + "grad_norm": 1.785689353942871, + "learning_rate": 7.103094663460067e-06, + "loss": 3.2555, + "step": 15110 + }, + { + "epoch": 1.2114208975566736, + "grad_norm": 1.8044604063034058, + "learning_rate": 7.101851917762062e-06, + "loss": 3.2493, + "step": 15111 + }, + { + "epoch": 1.2115010723376962, + "grad_norm": 1.733654499053955, + "learning_rate": 7.100609220925814e-06, + "loss": 3.2532, + "step": 15112 + }, + { + "epoch": 1.2115812471187188, + "grad_norm": 1.7187368869781494, + "learning_rate": 7.09936657297227e-06, + "loss": 3.1689, + "step": 15113 + }, + { + "epoch": 1.2116614218997415, + "grad_norm": 1.8440186977386475, + "learning_rate": 7.098123973922385e-06, + "loss": 3.2145, + "step": 15114 + }, + { + "epoch": 1.211741596680764, + "grad_norm": 1.805328607559204, + "learning_rate": 7.096881423797106e-06, + "loss": 3.0496, + "step": 15115 + }, + { + "epoch": 1.2118217714617867, + "grad_norm": 1.826218605041504, + "learning_rate": 7.095638922617383e-06, + "loss": 3.2085, + "step": 15116 + }, + { + "epoch": 1.2119019462428093, + "grad_norm": 1.7074863910675049, + "learning_rate": 7.0943964704041655e-06, + "loss": 3.1501, + "step": 15117 + }, + { + "epoch": 1.211982121023832, + "grad_norm": 1.8334355354309082, + "learning_rate": 7.093154067178398e-06, + "loss": 3.0848, + "step": 15118 + }, + { + "epoch": 1.2120622958048546, + "grad_norm": 1.7944669723510742, + "learning_rate": 7.091911712961031e-06, + "loss": 3.1559, + "step": 15119 + }, + { + "epoch": 1.2121424705858772, + "grad_norm": 1.7445485591888428, + "learning_rate": 7.090669407773004e-06, + "loss": 3.1732, + "step": 15120 + }, + { + "epoch": 1.2122226453668998, + "grad_norm": 1.7081345319747925, + "learning_rate": 7.089427151635266e-06, + "loss": 3.233, + "step": 15121 + }, + { + "epoch": 1.2123028201479225, + "grad_norm": 1.9184504747390747, + "learning_rate": 7.088184944568764e-06, + "loss": 3.2834, + "step": 15122 + }, + { + "epoch": 1.212382994928945, + "grad_norm": 1.9590587615966797, + "learning_rate": 7.086942786594433e-06, + "loss": 3.209, + "step": 15123 + }, + { + "epoch": 1.2124631697099677, + "grad_norm": 1.8031835556030273, + "learning_rate": 7.085700677733225e-06, + "loss": 3.2087, + "step": 15124 + }, + { + "epoch": 1.2125433444909903, + "grad_norm": 1.8769961595535278, + "learning_rate": 7.084458618006072e-06, + "loss": 3.1427, + "step": 15125 + }, + { + "epoch": 1.212623519272013, + "grad_norm": 1.9134972095489502, + "learning_rate": 7.083216607433921e-06, + "loss": 3.2292, + "step": 15126 + }, + { + "epoch": 1.2127036940530356, + "grad_norm": 1.850965976715088, + "learning_rate": 7.081974646037711e-06, + "loss": 3.2223, + "step": 15127 + }, + { + "epoch": 1.2127838688340582, + "grad_norm": 1.8029751777648926, + "learning_rate": 7.080732733838378e-06, + "loss": 3.1544, + "step": 15128 + }, + { + "epoch": 1.2128640436150808, + "grad_norm": 1.8093806505203247, + "learning_rate": 7.079490870856864e-06, + "loss": 3.1501, + "step": 15129 + }, + { + "epoch": 1.2129442183961034, + "grad_norm": 1.7080209255218506, + "learning_rate": 7.078249057114103e-06, + "loss": 3.192, + "step": 15130 + }, + { + "epoch": 1.213024393177126, + "grad_norm": 1.794993281364441, + "learning_rate": 7.077007292631033e-06, + "loss": 3.1419, + "step": 15131 + }, + { + "epoch": 1.2131045679581487, + "grad_norm": 1.7689223289489746, + "learning_rate": 7.075765577428593e-06, + "loss": 3.1782, + "step": 15132 + }, + { + "epoch": 1.2131847427391713, + "grad_norm": 1.893142580986023, + "learning_rate": 7.074523911527711e-06, + "loss": 3.2516, + "step": 15133 + }, + { + "epoch": 1.213264917520194, + "grad_norm": 1.890133023262024, + "learning_rate": 7.0732822949493275e-06, + "loss": 3.2203, + "step": 15134 + }, + { + "epoch": 1.2133450923012166, + "grad_norm": 1.7182917594909668, + "learning_rate": 7.072040727714371e-06, + "loss": 3.0748, + "step": 15135 + }, + { + "epoch": 1.2134252670822392, + "grad_norm": 1.7643355131149292, + "learning_rate": 7.070799209843776e-06, + "loss": 3.0601, + "step": 15136 + }, + { + "epoch": 1.213505441863262, + "grad_norm": 1.8321044445037842, + "learning_rate": 7.069557741358476e-06, + "loss": 3.2675, + "step": 15137 + }, + { + "epoch": 1.2135856166442847, + "grad_norm": 1.675946831703186, + "learning_rate": 7.068316322279398e-06, + "loss": 3.1849, + "step": 15138 + }, + { + "epoch": 1.2136657914253073, + "grad_norm": 1.7703206539154053, + "learning_rate": 7.0670749526274755e-06, + "loss": 3.0542, + "step": 15139 + }, + { + "epoch": 1.21374596620633, + "grad_norm": 1.900104284286499, + "learning_rate": 7.0658336324236336e-06, + "loss": 3.1735, + "step": 15140 + }, + { + "epoch": 1.2138261409873525, + "grad_norm": 1.7044388055801392, + "learning_rate": 7.064592361688803e-06, + "loss": 3.1855, + "step": 15141 + }, + { + "epoch": 1.2139063157683752, + "grad_norm": 1.8578664064407349, + "learning_rate": 7.063351140443911e-06, + "loss": 3.2581, + "step": 15142 + }, + { + "epoch": 1.2139864905493978, + "grad_norm": 1.8116827011108398, + "learning_rate": 7.0621099687098825e-06, + "loss": 3.212, + "step": 15143 + }, + { + "epoch": 1.2140666653304204, + "grad_norm": 0.694546103477478, + "learning_rate": 7.060868846507647e-06, + "loss": 2.4683, + "step": 15144 + }, + { + "epoch": 1.214146840111443, + "grad_norm": 1.8164836168289185, + "learning_rate": 7.059627773858126e-06, + "loss": 3.1534, + "step": 15145 + }, + { + "epoch": 1.2142270148924657, + "grad_norm": 1.7601377964019775, + "learning_rate": 7.058386750782242e-06, + "loss": 3.2544, + "step": 15146 + }, + { + "epoch": 1.2143071896734883, + "grad_norm": 0.7028876543045044, + "learning_rate": 7.057145777300925e-06, + "loss": 2.4817, + "step": 15147 + }, + { + "epoch": 1.214387364454511, + "grad_norm": 1.803077220916748, + "learning_rate": 7.05590485343509e-06, + "loss": 3.1459, + "step": 15148 + }, + { + "epoch": 1.2144675392355335, + "grad_norm": 1.7183541059494019, + "learning_rate": 7.0546639792056624e-06, + "loss": 3.1957, + "step": 15149 + }, + { + "epoch": 1.2145477140165561, + "grad_norm": 1.8446043729782104, + "learning_rate": 7.05342315463356e-06, + "loss": 3.2203, + "step": 15150 + }, + { + "epoch": 1.2146278887975788, + "grad_norm": 1.8144114017486572, + "learning_rate": 7.052182379739707e-06, + "loss": 3.2227, + "step": 15151 + }, + { + "epoch": 1.2147080635786014, + "grad_norm": 1.7877817153930664, + "learning_rate": 7.050941654545022e-06, + "loss": 3.1856, + "step": 15152 + }, + { + "epoch": 1.214788238359624, + "grad_norm": 1.7843183279037476, + "learning_rate": 7.049700979070417e-06, + "loss": 3.1716, + "step": 15153 + }, + { + "epoch": 1.2148684131406466, + "grad_norm": 1.8016878366470337, + "learning_rate": 7.048460353336818e-06, + "loss": 3.0911, + "step": 15154 + }, + { + "epoch": 1.2149485879216693, + "grad_norm": 0.6855641603469849, + "learning_rate": 7.047219777365134e-06, + "loss": 2.4763, + "step": 15155 + }, + { + "epoch": 1.215028762702692, + "grad_norm": 1.714621663093567, + "learning_rate": 7.045979251176285e-06, + "loss": 3.0536, + "step": 15156 + }, + { + "epoch": 1.2151089374837145, + "grad_norm": 1.8365756273269653, + "learning_rate": 7.0447387747911854e-06, + "loss": 3.1268, + "step": 15157 + }, + { + "epoch": 1.2151891122647371, + "grad_norm": 1.7008248567581177, + "learning_rate": 7.043498348230747e-06, + "loss": 3.1609, + "step": 15158 + }, + { + "epoch": 1.2152692870457598, + "grad_norm": 1.9499753713607788, + "learning_rate": 7.042257971515887e-06, + "loss": 3.2417, + "step": 15159 + }, + { + "epoch": 1.2153494618267824, + "grad_norm": 1.7672245502471924, + "learning_rate": 7.041017644667513e-06, + "loss": 3.1097, + "step": 15160 + }, + { + "epoch": 1.215429636607805, + "grad_norm": 1.755936622619629, + "learning_rate": 7.039777367706537e-06, + "loss": 3.14, + "step": 15161 + }, + { + "epoch": 1.2155098113888276, + "grad_norm": 1.8912346363067627, + "learning_rate": 7.038537140653876e-06, + "loss": 3.1877, + "step": 15162 + }, + { + "epoch": 1.2155899861698503, + "grad_norm": 1.819057583808899, + "learning_rate": 7.037296963530431e-06, + "loss": 3.1888, + "step": 15163 + }, + { + "epoch": 1.2156701609508729, + "grad_norm": 1.789819359779358, + "learning_rate": 7.036056836357119e-06, + "loss": 3.2635, + "step": 15164 + }, + { + "epoch": 1.2157503357318955, + "grad_norm": 1.828565239906311, + "learning_rate": 7.03481675915484e-06, + "loss": 3.2611, + "step": 15165 + }, + { + "epoch": 1.2158305105129181, + "grad_norm": 1.8965743780136108, + "learning_rate": 7.033576731944507e-06, + "loss": 3.2012, + "step": 15166 + }, + { + "epoch": 1.2159106852939408, + "grad_norm": 1.7562506198883057, + "learning_rate": 7.032336754747027e-06, + "loss": 3.2161, + "step": 15167 + }, + { + "epoch": 1.2159908600749634, + "grad_norm": 1.787300944328308, + "learning_rate": 7.031096827583299e-06, + "loss": 3.1848, + "step": 15168 + }, + { + "epoch": 1.216071034855986, + "grad_norm": 0.7022716403007507, + "learning_rate": 7.029856950474236e-06, + "loss": 2.4894, + "step": 15169 + }, + { + "epoch": 1.2161512096370086, + "grad_norm": 1.778220534324646, + "learning_rate": 7.028617123440736e-06, + "loss": 3.2131, + "step": 15170 + }, + { + "epoch": 1.2162313844180312, + "grad_norm": 1.776108980178833, + "learning_rate": 7.027377346503703e-06, + "loss": 3.1357, + "step": 15171 + }, + { + "epoch": 1.2163115591990539, + "grad_norm": 0.7073774337768555, + "learning_rate": 7.026137619684043e-06, + "loss": 2.4808, + "step": 15172 + }, + { + "epoch": 1.2163917339800765, + "grad_norm": 1.925522804260254, + "learning_rate": 7.024897943002653e-06, + "loss": 3.2319, + "step": 15173 + }, + { + "epoch": 1.2164719087610991, + "grad_norm": 1.7758625745773315, + "learning_rate": 7.023658316480437e-06, + "loss": 3.1564, + "step": 15174 + }, + { + "epoch": 1.2165520835421217, + "grad_norm": 1.695308804512024, + "learning_rate": 7.022418740138291e-06, + "loss": 3.2473, + "step": 15175 + }, + { + "epoch": 1.2166322583231444, + "grad_norm": 1.7959955930709839, + "learning_rate": 7.021179213997116e-06, + "loss": 3.1871, + "step": 15176 + }, + { + "epoch": 1.216712433104167, + "grad_norm": 1.8940272331237793, + "learning_rate": 7.019939738077811e-06, + "loss": 3.3356, + "step": 15177 + }, + { + "epoch": 1.2167926078851896, + "grad_norm": 1.766338586807251, + "learning_rate": 7.01870031240127e-06, + "loss": 3.1594, + "step": 15178 + }, + { + "epoch": 1.2168727826662122, + "grad_norm": 1.7840943336486816, + "learning_rate": 7.0174609369883925e-06, + "loss": 3.24, + "step": 15179 + }, + { + "epoch": 1.2169529574472349, + "grad_norm": 1.8365486860275269, + "learning_rate": 7.016221611860071e-06, + "loss": 3.0718, + "step": 15180 + }, + { + "epoch": 1.2170331322282575, + "grad_norm": 1.7650662660598755, + "learning_rate": 7.014982337037202e-06, + "loss": 3.1197, + "step": 15181 + }, + { + "epoch": 1.21711330700928, + "grad_norm": 1.6816980838775635, + "learning_rate": 7.01374311254068e-06, + "loss": 3.2104, + "step": 15182 + }, + { + "epoch": 1.217193481790303, + "grad_norm": 1.773519515991211, + "learning_rate": 7.012503938391394e-06, + "loss": 3.2764, + "step": 15183 + }, + { + "epoch": 1.2172736565713256, + "grad_norm": 1.6634269952774048, + "learning_rate": 7.011264814610241e-06, + "loss": 3.1227, + "step": 15184 + }, + { + "epoch": 1.2173538313523482, + "grad_norm": 1.8462332487106323, + "learning_rate": 7.010025741218109e-06, + "loss": 3.2072, + "step": 15185 + }, + { + "epoch": 1.2174340061333708, + "grad_norm": 1.7717121839523315, + "learning_rate": 7.008786718235889e-06, + "loss": 3.1755, + "step": 15186 + }, + { + "epoch": 1.2175141809143935, + "grad_norm": 1.8052713871002197, + "learning_rate": 7.007547745684472e-06, + "loss": 3.1578, + "step": 15187 + }, + { + "epoch": 1.217594355695416, + "grad_norm": 1.9376953840255737, + "learning_rate": 7.006308823584745e-06, + "loss": 3.3279, + "step": 15188 + }, + { + "epoch": 1.2176745304764387, + "grad_norm": 1.767209529876709, + "learning_rate": 7.0050699519575945e-06, + "loss": 3.1824, + "step": 15189 + }, + { + "epoch": 1.2177547052574613, + "grad_norm": 1.6513605117797852, + "learning_rate": 7.0038311308239105e-06, + "loss": 3.1516, + "step": 15190 + }, + { + "epoch": 1.217834880038484, + "grad_norm": 1.7498639822006226, + "learning_rate": 7.002592360204576e-06, + "loss": 3.168, + "step": 15191 + }, + { + "epoch": 1.2179150548195066, + "grad_norm": 1.819982647895813, + "learning_rate": 7.001353640120481e-06, + "loss": 3.2548, + "step": 15192 + }, + { + "epoch": 1.2179952296005292, + "grad_norm": 0.66236811876297, + "learning_rate": 7.000114970592505e-06, + "loss": 2.4628, + "step": 15193 + }, + { + "epoch": 1.2180754043815518, + "grad_norm": 1.7471905946731567, + "learning_rate": 6.998876351641534e-06, + "loss": 3.187, + "step": 15194 + }, + { + "epoch": 1.2181555791625744, + "grad_norm": 1.779995083808899, + "learning_rate": 6.997637783288449e-06, + "loss": 3.2393, + "step": 15195 + }, + { + "epoch": 1.218235753943597, + "grad_norm": 1.8722693920135498, + "learning_rate": 6.996399265554132e-06, + "loss": 3.1186, + "step": 15196 + }, + { + "epoch": 1.2183159287246197, + "grad_norm": 0.6921184659004211, + "learning_rate": 6.995160798459467e-06, + "loss": 2.479, + "step": 15197 + }, + { + "epoch": 1.2183961035056423, + "grad_norm": 1.8684026002883911, + "learning_rate": 6.993922382025331e-06, + "loss": 3.1581, + "step": 15198 + }, + { + "epoch": 1.218476278286665, + "grad_norm": 1.82537043094635, + "learning_rate": 6.992684016272605e-06, + "loss": 3.1732, + "step": 15199 + }, + { + "epoch": 1.2185564530676876, + "grad_norm": 1.6793149709701538, + "learning_rate": 6.991445701222165e-06, + "loss": 3.1115, + "step": 15200 + }, + { + "epoch": 1.2186366278487102, + "grad_norm": 1.795183777809143, + "learning_rate": 6.99020743689489e-06, + "loss": 3.296, + "step": 15201 + }, + { + "epoch": 1.2187168026297328, + "grad_norm": 1.707329273223877, + "learning_rate": 6.988969223311659e-06, + "loss": 3.2191, + "step": 15202 + }, + { + "epoch": 1.2187969774107554, + "grad_norm": 1.7354531288146973, + "learning_rate": 6.987731060493344e-06, + "loss": 3.272, + "step": 15203 + }, + { + "epoch": 1.218877152191778, + "grad_norm": 1.7817895412445068, + "learning_rate": 6.986492948460823e-06, + "loss": 3.1998, + "step": 15204 + }, + { + "epoch": 1.2189573269728007, + "grad_norm": 1.828933835029602, + "learning_rate": 6.985254887234968e-06, + "loss": 3.2292, + "step": 15205 + }, + { + "epoch": 1.2190375017538233, + "grad_norm": 1.709392786026001, + "learning_rate": 6.984016876836652e-06, + "loss": 3.187, + "step": 15206 + }, + { + "epoch": 1.219117676534846, + "grad_norm": 1.774924397468567, + "learning_rate": 6.982778917286751e-06, + "loss": 3.1616, + "step": 15207 + }, + { + "epoch": 1.2191978513158686, + "grad_norm": 1.714674711227417, + "learning_rate": 6.981541008606132e-06, + "loss": 3.2681, + "step": 15208 + }, + { + "epoch": 1.2192780260968912, + "grad_norm": 0.7024875283241272, + "learning_rate": 6.98030315081567e-06, + "loss": 2.467, + "step": 15209 + }, + { + "epoch": 1.2193582008779138, + "grad_norm": 1.7799307107925415, + "learning_rate": 6.97906534393623e-06, + "loss": 3.1848, + "step": 15210 + }, + { + "epoch": 1.2194383756589364, + "grad_norm": 1.5898487567901611, + "learning_rate": 6.977827587988684e-06, + "loss": 3.1271, + "step": 15211 + }, + { + "epoch": 1.219518550439959, + "grad_norm": 1.7394862174987793, + "learning_rate": 6.976589882993903e-06, + "loss": 3.1024, + "step": 15212 + }, + { + "epoch": 1.2195987252209817, + "grad_norm": 1.7311452627182007, + "learning_rate": 6.975352228972747e-06, + "loss": 3.1835, + "step": 15213 + }, + { + "epoch": 1.2196789000020043, + "grad_norm": 1.8172895908355713, + "learning_rate": 6.974114625946089e-06, + "loss": 3.257, + "step": 15214 + }, + { + "epoch": 1.219759074783027, + "grad_norm": 1.7933591604232788, + "learning_rate": 6.97287707393479e-06, + "loss": 3.1931, + "step": 15215 + }, + { + "epoch": 1.2198392495640495, + "grad_norm": 1.816853642463684, + "learning_rate": 6.9716395729597166e-06, + "loss": 3.1687, + "step": 15216 + }, + { + "epoch": 1.2199194243450722, + "grad_norm": 1.9062000513076782, + "learning_rate": 6.970402123041735e-06, + "loss": 3.2316, + "step": 15217 + }, + { + "epoch": 1.2199995991260948, + "grad_norm": 1.7653498649597168, + "learning_rate": 6.9691647242017044e-06, + "loss": 3.155, + "step": 15218 + }, + { + "epoch": 1.2200797739071176, + "grad_norm": 1.7395228147506714, + "learning_rate": 6.96792737646049e-06, + "loss": 3.2278, + "step": 15219 + }, + { + "epoch": 1.2201599486881403, + "grad_norm": 1.7440342903137207, + "learning_rate": 6.9666900798389496e-06, + "loss": 3.2513, + "step": 15220 + }, + { + "epoch": 1.2202401234691629, + "grad_norm": 1.75737464427948, + "learning_rate": 6.9654528343579445e-06, + "loss": 3.0975, + "step": 15221 + }, + { + "epoch": 1.2203202982501855, + "grad_norm": 1.8396015167236328, + "learning_rate": 6.964215640038338e-06, + "loss": 3.2433, + "step": 15222 + }, + { + "epoch": 1.2204004730312081, + "grad_norm": 1.720739483833313, + "learning_rate": 6.962978496900983e-06, + "loss": 3.1909, + "step": 15223 + }, + { + "epoch": 1.2204806478122308, + "grad_norm": 1.8723241090774536, + "learning_rate": 6.961741404966743e-06, + "loss": 3.2657, + "step": 15224 + }, + { + "epoch": 1.2205608225932534, + "grad_norm": 1.7329589128494263, + "learning_rate": 6.96050436425647e-06, + "loss": 3.1086, + "step": 15225 + }, + { + "epoch": 1.220640997374276, + "grad_norm": 1.6890509128570557, + "learning_rate": 6.959267374791022e-06, + "loss": 3.1759, + "step": 15226 + }, + { + "epoch": 1.2207211721552986, + "grad_norm": 1.9184857606887817, + "learning_rate": 6.958030436591257e-06, + "loss": 3.1263, + "step": 15227 + }, + { + "epoch": 1.2208013469363213, + "grad_norm": 1.8675198554992676, + "learning_rate": 6.956793549678024e-06, + "loss": 3.1666, + "step": 15228 + }, + { + "epoch": 1.2208815217173439, + "grad_norm": 1.721860647201538, + "learning_rate": 6.955556714072181e-06, + "loss": 3.2133, + "step": 15229 + }, + { + "epoch": 1.2209616964983665, + "grad_norm": 1.7410026788711548, + "learning_rate": 6.954319929794577e-06, + "loss": 3.0778, + "step": 15230 + }, + { + "epoch": 1.2210418712793891, + "grad_norm": 1.802600622177124, + "learning_rate": 6.953083196866066e-06, + "loss": 3.1825, + "step": 15231 + }, + { + "epoch": 1.2211220460604117, + "grad_norm": 0.7096085548400879, + "learning_rate": 6.951846515307501e-06, + "loss": 2.4808, + "step": 15232 + }, + { + "epoch": 1.2212022208414344, + "grad_norm": 1.7383867502212524, + "learning_rate": 6.9506098851397276e-06, + "loss": 3.2228, + "step": 15233 + }, + { + "epoch": 1.221282395622457, + "grad_norm": 1.7156776189804077, + "learning_rate": 6.949373306383598e-06, + "loss": 3.189, + "step": 15234 + }, + { + "epoch": 1.2213625704034796, + "grad_norm": 1.7834973335266113, + "learning_rate": 6.948136779059957e-06, + "loss": 3.2407, + "step": 15235 + }, + { + "epoch": 1.2214427451845022, + "grad_norm": 1.797642707824707, + "learning_rate": 6.946900303189654e-06, + "loss": 3.2354, + "step": 15236 + }, + { + "epoch": 1.2215229199655249, + "grad_norm": 1.8758671283721924, + "learning_rate": 6.9456638787935385e-06, + "loss": 3.0712, + "step": 15237 + }, + { + "epoch": 1.2216030947465475, + "grad_norm": 1.7203342914581299, + "learning_rate": 6.944427505892453e-06, + "loss": 3.2288, + "step": 15238 + }, + { + "epoch": 1.2216832695275701, + "grad_norm": 1.8544918298721313, + "learning_rate": 6.943191184507243e-06, + "loss": 3.2339, + "step": 15239 + }, + { + "epoch": 1.2217634443085927, + "grad_norm": 1.8399230241775513, + "learning_rate": 6.9419549146587506e-06, + "loss": 3.2148, + "step": 15240 + }, + { + "epoch": 1.2218436190896154, + "grad_norm": 1.755428671836853, + "learning_rate": 6.940718696367821e-06, + "loss": 3.0875, + "step": 15241 + }, + { + "epoch": 1.221923793870638, + "grad_norm": 1.837821364402771, + "learning_rate": 6.939482529655297e-06, + "loss": 3.1525, + "step": 15242 + }, + { + "epoch": 1.2220039686516606, + "grad_norm": 1.8649719953536987, + "learning_rate": 6.938246414542018e-06, + "loss": 3.2377, + "step": 15243 + }, + { + "epoch": 1.2220841434326832, + "grad_norm": 1.704236388206482, + "learning_rate": 6.937010351048827e-06, + "loss": 3.1128, + "step": 15244 + }, + { + "epoch": 1.2221643182137059, + "grad_norm": 1.8560385704040527, + "learning_rate": 6.935774339196559e-06, + "loss": 3.2102, + "step": 15245 + }, + { + "epoch": 1.2222444929947285, + "grad_norm": 1.7513412237167358, + "learning_rate": 6.934538379006057e-06, + "loss": 3.2845, + "step": 15246 + }, + { + "epoch": 1.222324667775751, + "grad_norm": 1.8017315864562988, + "learning_rate": 6.93330247049816e-06, + "loss": 3.1255, + "step": 15247 + }, + { + "epoch": 1.2224048425567737, + "grad_norm": 1.8160266876220703, + "learning_rate": 6.9320666136937e-06, + "loss": 3.202, + "step": 15248 + }, + { + "epoch": 1.2224850173377964, + "grad_norm": 1.8505163192749023, + "learning_rate": 6.9308308086135175e-06, + "loss": 3.2045, + "step": 15249 + }, + { + "epoch": 1.222565192118819, + "grad_norm": 1.8022137880325317, + "learning_rate": 6.929595055278444e-06, + "loss": 3.3016, + "step": 15250 + }, + { + "epoch": 1.2226453668998416, + "grad_norm": 1.8975635766983032, + "learning_rate": 6.928359353709315e-06, + "loss": 3.235, + "step": 15251 + }, + { + "epoch": 1.2227255416808642, + "grad_norm": 1.8628151416778564, + "learning_rate": 6.927123703926968e-06, + "loss": 3.1884, + "step": 15252 + }, + { + "epoch": 1.2228057164618868, + "grad_norm": 1.7917368412017822, + "learning_rate": 6.92588810595223e-06, + "loss": 3.2028, + "step": 15253 + }, + { + "epoch": 1.2228858912429095, + "grad_norm": 1.905699610710144, + "learning_rate": 6.924652559805935e-06, + "loss": 3.2266, + "step": 15254 + }, + { + "epoch": 1.222966066023932, + "grad_norm": 1.824876308441162, + "learning_rate": 6.923417065508915e-06, + "loss": 3.0286, + "step": 15255 + }, + { + "epoch": 1.2230462408049547, + "grad_norm": 1.7743028402328491, + "learning_rate": 6.922181623081998e-06, + "loss": 3.1115, + "step": 15256 + }, + { + "epoch": 1.2231264155859773, + "grad_norm": 0.698601245880127, + "learning_rate": 6.920946232546015e-06, + "loss": 2.4872, + "step": 15257 + }, + { + "epoch": 1.223206590367, + "grad_norm": 1.8645801544189453, + "learning_rate": 6.919710893921794e-06, + "loss": 3.1035, + "step": 15258 + }, + { + "epoch": 1.2232867651480226, + "grad_norm": 1.7872710227966309, + "learning_rate": 6.918475607230162e-06, + "loss": 3.2306, + "step": 15259 + }, + { + "epoch": 1.2233669399290452, + "grad_norm": 1.685733675956726, + "learning_rate": 6.917240372491942e-06, + "loss": 3.2279, + "step": 15260 + }, + { + "epoch": 1.2234471147100678, + "grad_norm": 1.8510609865188599, + "learning_rate": 6.916005189727964e-06, + "loss": 3.0975, + "step": 15261 + }, + { + "epoch": 1.2235272894910905, + "grad_norm": 1.7897993326187134, + "learning_rate": 6.914770058959055e-06, + "loss": 3.1315, + "step": 15262 + }, + { + "epoch": 1.223607464272113, + "grad_norm": 1.7886837720870972, + "learning_rate": 6.913534980206032e-06, + "loss": 3.0734, + "step": 15263 + }, + { + "epoch": 1.2236876390531357, + "grad_norm": 1.739720106124878, + "learning_rate": 6.912299953489723e-06, + "loss": 3.091, + "step": 15264 + }, + { + "epoch": 1.2237678138341586, + "grad_norm": 1.6875454187393188, + "learning_rate": 6.911064978830947e-06, + "loss": 3.1422, + "step": 15265 + }, + { + "epoch": 1.2238479886151812, + "grad_norm": 1.8191595077514648, + "learning_rate": 6.909830056250527e-06, + "loss": 3.1909, + "step": 15266 + }, + { + "epoch": 1.2239281633962038, + "grad_norm": 1.8074089288711548, + "learning_rate": 6.908595185769285e-06, + "loss": 3.1257, + "step": 15267 + }, + { + "epoch": 1.2240083381772264, + "grad_norm": 1.7846466302871704, + "learning_rate": 6.907360367408037e-06, + "loss": 3.1933, + "step": 15268 + }, + { + "epoch": 1.224088512958249, + "grad_norm": 1.7623757123947144, + "learning_rate": 6.906125601187606e-06, + "loss": 3.2146, + "step": 15269 + }, + { + "epoch": 1.2241686877392717, + "grad_norm": 1.7522754669189453, + "learning_rate": 6.9048908871288035e-06, + "loss": 3.2483, + "step": 15270 + }, + { + "epoch": 1.2242488625202943, + "grad_norm": 1.8009520769119263, + "learning_rate": 6.90365622525245e-06, + "loss": 3.1997, + "step": 15271 + }, + { + "epoch": 1.224329037301317, + "grad_norm": 1.7651585340499878, + "learning_rate": 6.902421615579364e-06, + "loss": 3.1274, + "step": 15272 + }, + { + "epoch": 1.2244092120823395, + "grad_norm": 1.750123143196106, + "learning_rate": 6.901187058130355e-06, + "loss": 3.1132, + "step": 15273 + }, + { + "epoch": 1.2244893868633622, + "grad_norm": 1.9000657796859741, + "learning_rate": 6.899952552926243e-06, + "loss": 3.0875, + "step": 15274 + }, + { + "epoch": 1.2245695616443848, + "grad_norm": 0.7010677456855774, + "learning_rate": 6.8987180999878355e-06, + "loss": 2.4869, + "step": 15275 + }, + { + "epoch": 1.2246497364254074, + "grad_norm": 1.839300274848938, + "learning_rate": 6.897483699335948e-06, + "loss": 3.0722, + "step": 15276 + }, + { + "epoch": 1.22472991120643, + "grad_norm": 1.7514923810958862, + "learning_rate": 6.896249350991395e-06, + "loss": 3.2683, + "step": 15277 + }, + { + "epoch": 1.2248100859874527, + "grad_norm": 1.7945300340652466, + "learning_rate": 6.8950150549749805e-06, + "loss": 3.2098, + "step": 15278 + }, + { + "epoch": 1.2248902607684753, + "grad_norm": 1.705533504486084, + "learning_rate": 6.893780811307521e-06, + "loss": 3.1967, + "step": 15279 + }, + { + "epoch": 1.224970435549498, + "grad_norm": 0.7023494839668274, + "learning_rate": 6.89254662000982e-06, + "loss": 2.473, + "step": 15280 + }, + { + "epoch": 1.2250506103305205, + "grad_norm": 1.7923263311386108, + "learning_rate": 6.891312481102688e-06, + "loss": 3.243, + "step": 15281 + }, + { + "epoch": 1.2251307851115432, + "grad_norm": 1.7331925630569458, + "learning_rate": 6.8900783946069335e-06, + "loss": 3.1897, + "step": 15282 + }, + { + "epoch": 1.2252109598925658, + "grad_norm": 1.724989652633667, + "learning_rate": 6.88884436054336e-06, + "loss": 3.0355, + "step": 15283 + }, + { + "epoch": 1.2252911346735884, + "grad_norm": 1.8858757019042969, + "learning_rate": 6.887610378932773e-06, + "loss": 3.1264, + "step": 15284 + }, + { + "epoch": 1.225371309454611, + "grad_norm": 1.7369483709335327, + "learning_rate": 6.886376449795981e-06, + "loss": 3.2748, + "step": 15285 + }, + { + "epoch": 1.2254514842356337, + "grad_norm": 1.9240610599517822, + "learning_rate": 6.885142573153783e-06, + "loss": 3.2487, + "step": 15286 + }, + { + "epoch": 1.2255316590166563, + "grad_norm": 1.6850591897964478, + "learning_rate": 6.883908749026984e-06, + "loss": 3.2295, + "step": 15287 + }, + { + "epoch": 1.225611833797679, + "grad_norm": 0.6865147352218628, + "learning_rate": 6.882674977436383e-06, + "loss": 2.4869, + "step": 15288 + }, + { + "epoch": 1.2256920085787015, + "grad_norm": 1.8416264057159424, + "learning_rate": 6.881441258402786e-06, + "loss": 3.2621, + "step": 15289 + }, + { + "epoch": 1.2257721833597242, + "grad_norm": 1.7395343780517578, + "learning_rate": 6.88020759194699e-06, + "loss": 3.1829, + "step": 15290 + }, + { + "epoch": 1.2258523581407468, + "grad_norm": 1.7496973276138306, + "learning_rate": 6.878973978089794e-06, + "loss": 3.2281, + "step": 15291 + }, + { + "epoch": 1.2259325329217694, + "grad_norm": 1.6919646263122559, + "learning_rate": 6.877740416851997e-06, + "loss": 3.0995, + "step": 15292 + }, + { + "epoch": 1.226012707702792, + "grad_norm": 1.8236514329910278, + "learning_rate": 6.876506908254395e-06, + "loss": 3.2242, + "step": 15293 + }, + { + "epoch": 1.2260928824838147, + "grad_norm": 1.824864387512207, + "learning_rate": 6.8752734523177855e-06, + "loss": 3.2419, + "step": 15294 + }, + { + "epoch": 1.2261730572648373, + "grad_norm": 1.7678852081298828, + "learning_rate": 6.874040049062966e-06, + "loss": 3.1024, + "step": 15295 + }, + { + "epoch": 1.22625323204586, + "grad_norm": 1.7007008790969849, + "learning_rate": 6.872806698510728e-06, + "loss": 3.1898, + "step": 15296 + }, + { + "epoch": 1.2263334068268825, + "grad_norm": 1.7517881393432617, + "learning_rate": 6.8715734006818705e-06, + "loss": 3.1923, + "step": 15297 + }, + { + "epoch": 1.2264135816079051, + "grad_norm": 1.9224979877471924, + "learning_rate": 6.87034015559718e-06, + "loss": 3.3477, + "step": 15298 + }, + { + "epoch": 1.2264937563889278, + "grad_norm": 1.7608646154403687, + "learning_rate": 6.86910696327745e-06, + "loss": 3.195, + "step": 15299 + }, + { + "epoch": 1.2265739311699506, + "grad_norm": 1.8271785974502563, + "learning_rate": 6.867873823743476e-06, + "loss": 3.2499, + "step": 15300 + }, + { + "epoch": 1.2266541059509732, + "grad_norm": 1.6802281141281128, + "learning_rate": 6.866640737016044e-06, + "loss": 3.0972, + "step": 15301 + }, + { + "epoch": 1.2267342807319959, + "grad_norm": 1.7646201848983765, + "learning_rate": 6.865407703115945e-06, + "loss": 3.0451, + "step": 15302 + }, + { + "epoch": 1.2268144555130185, + "grad_norm": 1.764365792274475, + "learning_rate": 6.864174722063968e-06, + "loss": 3.0793, + "step": 15303 + }, + { + "epoch": 1.226894630294041, + "grad_norm": 1.709480881690979, + "learning_rate": 6.8629417938808984e-06, + "loss": 3.1187, + "step": 15304 + }, + { + "epoch": 1.2269748050750637, + "grad_norm": 1.8788199424743652, + "learning_rate": 6.861708918587526e-06, + "loss": 3.1426, + "step": 15305 + }, + { + "epoch": 1.2270549798560864, + "grad_norm": 1.659359335899353, + "learning_rate": 6.860476096204636e-06, + "loss": 3.151, + "step": 15306 + }, + { + "epoch": 1.227135154637109, + "grad_norm": 1.743019938468933, + "learning_rate": 6.859243326753012e-06, + "loss": 3.1349, + "step": 15307 + }, + { + "epoch": 1.2272153294181316, + "grad_norm": 1.902052879333496, + "learning_rate": 6.858010610253437e-06, + "loss": 3.2247, + "step": 15308 + }, + { + "epoch": 1.2272955041991542, + "grad_norm": 1.904779076576233, + "learning_rate": 6.856777946726696e-06, + "loss": 3.1705, + "step": 15309 + }, + { + "epoch": 1.2273756789801769, + "grad_norm": 1.8676683902740479, + "learning_rate": 6.855545336193573e-06, + "loss": 3.1515, + "step": 15310 + }, + { + "epoch": 1.2274558537611995, + "grad_norm": 1.833159327507019, + "learning_rate": 6.854312778674846e-06, + "loss": 3.2517, + "step": 15311 + }, + { + "epoch": 1.227536028542222, + "grad_norm": 1.7265172004699707, + "learning_rate": 6.853080274191298e-06, + "loss": 3.2037, + "step": 15312 + }, + { + "epoch": 1.2276162033232447, + "grad_norm": 1.8378276824951172, + "learning_rate": 6.851847822763706e-06, + "loss": 3.1617, + "step": 15313 + }, + { + "epoch": 1.2276963781042673, + "grad_norm": 1.7164734601974487, + "learning_rate": 6.85061542441285e-06, + "loss": 3.1665, + "step": 15314 + }, + { + "epoch": 1.22777655288529, + "grad_norm": 1.7524210214614868, + "learning_rate": 6.849383079159511e-06, + "loss": 3.1254, + "step": 15315 + }, + { + "epoch": 1.2278567276663126, + "grad_norm": 1.7878323793411255, + "learning_rate": 6.848150787024459e-06, + "loss": 3.2657, + "step": 15316 + }, + { + "epoch": 1.2279369024473352, + "grad_norm": 1.7087496519088745, + "learning_rate": 6.8469185480284785e-06, + "loss": 3.2032, + "step": 15317 + }, + { + "epoch": 1.2280170772283578, + "grad_norm": 1.737670660018921, + "learning_rate": 6.845686362192336e-06, + "loss": 3.2013, + "step": 15318 + }, + { + "epoch": 1.2280972520093805, + "grad_norm": 1.7806870937347412, + "learning_rate": 6.84445422953681e-06, + "loss": 3.2438, + "step": 15319 + }, + { + "epoch": 1.228177426790403, + "grad_norm": 1.698917031288147, + "learning_rate": 6.8432221500826756e-06, + "loss": 3.1785, + "step": 15320 + }, + { + "epoch": 1.2282576015714257, + "grad_norm": 1.801816463470459, + "learning_rate": 6.841990123850701e-06, + "loss": 3.1946, + "step": 15321 + }, + { + "epoch": 1.2283377763524483, + "grad_norm": 1.8648560047149658, + "learning_rate": 6.840758150861663e-06, + "loss": 3.307, + "step": 15322 + }, + { + "epoch": 1.228417951133471, + "grad_norm": 1.8088579177856445, + "learning_rate": 6.839526231136326e-06, + "loss": 3.1931, + "step": 15323 + }, + { + "epoch": 1.2284981259144936, + "grad_norm": 1.8430850505828857, + "learning_rate": 6.838294364695464e-06, + "loss": 3.2557, + "step": 15324 + }, + { + "epoch": 1.2285783006955162, + "grad_norm": 1.869422197341919, + "learning_rate": 6.837062551559845e-06, + "loss": 3.3595, + "step": 15325 + }, + { + "epoch": 1.2286584754765388, + "grad_norm": 1.7862567901611328, + "learning_rate": 6.835830791750235e-06, + "loss": 3.2225, + "step": 15326 + }, + { + "epoch": 1.2287386502575615, + "grad_norm": 1.8382971286773682, + "learning_rate": 6.8345990852874065e-06, + "loss": 3.2742, + "step": 15327 + }, + { + "epoch": 1.228818825038584, + "grad_norm": 1.8026245832443237, + "learning_rate": 6.83336743219212e-06, + "loss": 3.3566, + "step": 15328 + }, + { + "epoch": 1.2288989998196067, + "grad_norm": 1.8568612337112427, + "learning_rate": 6.83213583248514e-06, + "loss": 3.1726, + "step": 15329 + }, + { + "epoch": 1.2289791746006293, + "grad_norm": 1.8217189311981201, + "learning_rate": 6.830904286187236e-06, + "loss": 3.2216, + "step": 15330 + }, + { + "epoch": 1.229059349381652, + "grad_norm": 1.8441756963729858, + "learning_rate": 6.829672793319169e-06, + "loss": 3.1009, + "step": 15331 + }, + { + "epoch": 1.2291395241626746, + "grad_norm": 1.775905966758728, + "learning_rate": 6.828441353901701e-06, + "loss": 3.158, + "step": 15332 + }, + { + "epoch": 1.2292196989436972, + "grad_norm": 1.7241864204406738, + "learning_rate": 6.827209967955593e-06, + "loss": 3.224, + "step": 15333 + }, + { + "epoch": 1.2292998737247198, + "grad_norm": 1.8260759115219116, + "learning_rate": 6.825978635501607e-06, + "loss": 3.331, + "step": 15334 + }, + { + "epoch": 1.2293800485057425, + "grad_norm": 1.8068945407867432, + "learning_rate": 6.824747356560504e-06, + "loss": 3.183, + "step": 15335 + }, + { + "epoch": 1.229460223286765, + "grad_norm": 1.7481805086135864, + "learning_rate": 6.8235161311530395e-06, + "loss": 3.1113, + "step": 15336 + }, + { + "epoch": 1.2295403980677877, + "grad_norm": 1.637284755706787, + "learning_rate": 6.822284959299976e-06, + "loss": 3.1379, + "step": 15337 + }, + { + "epoch": 1.2296205728488103, + "grad_norm": 1.7253097295761108, + "learning_rate": 6.821053841022065e-06, + "loss": 3.1937, + "step": 15338 + }, + { + "epoch": 1.229700747629833, + "grad_norm": 1.791584849357605, + "learning_rate": 6.8198227763400684e-06, + "loss": 3.1926, + "step": 15339 + }, + { + "epoch": 1.2297809224108556, + "grad_norm": 1.8459473848342896, + "learning_rate": 6.818591765274738e-06, + "loss": 3.1517, + "step": 15340 + }, + { + "epoch": 1.2298610971918782, + "grad_norm": 1.819670557975769, + "learning_rate": 6.817360807846829e-06, + "loss": 3.1552, + "step": 15341 + }, + { + "epoch": 1.2299412719729008, + "grad_norm": 1.6688129901885986, + "learning_rate": 6.816129904077097e-06, + "loss": 3.1125, + "step": 15342 + }, + { + "epoch": 1.2300214467539234, + "grad_norm": 1.7429349422454834, + "learning_rate": 6.81489905398629e-06, + "loss": 3.2261, + "step": 15343 + }, + { + "epoch": 1.230101621534946, + "grad_norm": 1.7305197715759277, + "learning_rate": 6.813668257595163e-06, + "loss": 3.1963, + "step": 15344 + }, + { + "epoch": 1.2301817963159687, + "grad_norm": 1.7439839839935303, + "learning_rate": 6.812437514924468e-06, + "loss": 3.1637, + "step": 15345 + }, + { + "epoch": 1.2302619710969915, + "grad_norm": 1.8524901866912842, + "learning_rate": 6.811206825994952e-06, + "loss": 3.2019, + "step": 15346 + }, + { + "epoch": 1.2303421458780142, + "grad_norm": 1.819624900817871, + "learning_rate": 6.809976190827365e-06, + "loss": 3.1946, + "step": 15347 + }, + { + "epoch": 1.2304223206590368, + "grad_norm": 1.8365097045898438, + "learning_rate": 6.808745609442454e-06, + "loss": 3.2583, + "step": 15348 + }, + { + "epoch": 1.2305024954400594, + "grad_norm": 1.7840358018875122, + "learning_rate": 6.807515081860968e-06, + "loss": 3.2099, + "step": 15349 + }, + { + "epoch": 1.230582670221082, + "grad_norm": 1.727192997932434, + "learning_rate": 6.806284608103655e-06, + "loss": 3.0754, + "step": 15350 + }, + { + "epoch": 1.2306628450021047, + "grad_norm": 1.6602504253387451, + "learning_rate": 6.805054188191255e-06, + "loss": 3.082, + "step": 15351 + }, + { + "epoch": 1.2307430197831273, + "grad_norm": 1.6787317991256714, + "learning_rate": 6.803823822144518e-06, + "loss": 3.13, + "step": 15352 + }, + { + "epoch": 1.23082319456415, + "grad_norm": 0.6960234045982361, + "learning_rate": 6.802593509984181e-06, + "loss": 2.4664, + "step": 15353 + }, + { + "epoch": 1.2309033693451725, + "grad_norm": 1.7427659034729004, + "learning_rate": 6.801363251730992e-06, + "loss": 3.0457, + "step": 15354 + }, + { + "epoch": 1.2309835441261952, + "grad_norm": 1.793296456336975, + "learning_rate": 6.800133047405694e-06, + "loss": 3.2428, + "step": 15355 + }, + { + "epoch": 1.2310637189072178, + "grad_norm": 1.7758457660675049, + "learning_rate": 6.798902897029022e-06, + "loss": 3.1866, + "step": 15356 + }, + { + "epoch": 1.2311438936882404, + "grad_norm": 1.8068796396255493, + "learning_rate": 6.797672800621722e-06, + "loss": 3.0879, + "step": 15357 + }, + { + "epoch": 1.231224068469263, + "grad_norm": 1.7632081508636475, + "learning_rate": 6.7964427582045264e-06, + "loss": 3.2117, + "step": 15358 + }, + { + "epoch": 1.2313042432502856, + "grad_norm": 1.7944302558898926, + "learning_rate": 6.7952127697981784e-06, + "loss": 3.1599, + "step": 15359 + }, + { + "epoch": 1.2313844180313083, + "grad_norm": 1.806305170059204, + "learning_rate": 6.793982835423414e-06, + "loss": 3.2097, + "step": 15360 + }, + { + "epoch": 1.231464592812331, + "grad_norm": 1.8475043773651123, + "learning_rate": 6.792752955100968e-06, + "loss": 3.109, + "step": 15361 + }, + { + "epoch": 1.2315447675933535, + "grad_norm": 1.7285100221633911, + "learning_rate": 6.791523128851579e-06, + "loss": 3.1809, + "step": 15362 + }, + { + "epoch": 1.2316249423743761, + "grad_norm": 1.798733115196228, + "learning_rate": 6.790293356695977e-06, + "loss": 3.1473, + "step": 15363 + }, + { + "epoch": 1.2317051171553988, + "grad_norm": 1.7377943992614746, + "learning_rate": 6.789063638654898e-06, + "loss": 3.1824, + "step": 15364 + }, + { + "epoch": 1.2317852919364214, + "grad_norm": 1.7856932878494263, + "learning_rate": 6.787833974749076e-06, + "loss": 3.0145, + "step": 15365 + }, + { + "epoch": 1.231865466717444, + "grad_norm": 1.8491249084472656, + "learning_rate": 6.786604364999241e-06, + "loss": 3.1243, + "step": 15366 + }, + { + "epoch": 1.2319456414984666, + "grad_norm": 1.7037667036056519, + "learning_rate": 6.785374809426124e-06, + "loss": 3.2068, + "step": 15367 + }, + { + "epoch": 1.2320258162794893, + "grad_norm": 1.71318781375885, + "learning_rate": 6.784145308050453e-06, + "loss": 3.1522, + "step": 15368 + }, + { + "epoch": 1.2321059910605119, + "grad_norm": 1.7655055522918701, + "learning_rate": 6.782915860892958e-06, + "loss": 3.1288, + "step": 15369 + }, + { + "epoch": 1.2321861658415345, + "grad_norm": 1.7030236721038818, + "learning_rate": 6.781686467974371e-06, + "loss": 3.0538, + "step": 15370 + }, + { + "epoch": 1.2322663406225571, + "grad_norm": 1.8462766408920288, + "learning_rate": 6.780457129315414e-06, + "loss": 3.2012, + "step": 15371 + }, + { + "epoch": 1.2323465154035798, + "grad_norm": 1.7438831329345703, + "learning_rate": 6.779227844936817e-06, + "loss": 3.1392, + "step": 15372 + }, + { + "epoch": 1.2324266901846024, + "grad_norm": 1.788046956062317, + "learning_rate": 6.7779986148593e-06, + "loss": 3.173, + "step": 15373 + }, + { + "epoch": 1.232506864965625, + "grad_norm": 1.7241209745407104, + "learning_rate": 6.776769439103592e-06, + "loss": 3.0815, + "step": 15374 + }, + { + "epoch": 1.2325870397466476, + "grad_norm": 1.8036800622940063, + "learning_rate": 6.775540317690417e-06, + "loss": 3.1464, + "step": 15375 + }, + { + "epoch": 1.2326672145276703, + "grad_norm": 1.7975531816482544, + "learning_rate": 6.7743112506404954e-06, + "loss": 3.2054, + "step": 15376 + }, + { + "epoch": 1.2327473893086929, + "grad_norm": 1.712390422821045, + "learning_rate": 6.773082237974549e-06, + "loss": 3.0905, + "step": 15377 + }, + { + "epoch": 1.2328275640897155, + "grad_norm": 1.7788504362106323, + "learning_rate": 6.771853279713297e-06, + "loss": 3.1542, + "step": 15378 + }, + { + "epoch": 1.2329077388707381, + "grad_norm": 1.762308955192566, + "learning_rate": 6.770624375877462e-06, + "loss": 3.172, + "step": 15379 + }, + { + "epoch": 1.2329879136517607, + "grad_norm": 1.7540435791015625, + "learning_rate": 6.769395526487763e-06, + "loss": 3.1291, + "step": 15380 + }, + { + "epoch": 1.2330680884327834, + "grad_norm": 1.712486743927002, + "learning_rate": 6.7681667315649155e-06, + "loss": 3.0609, + "step": 15381 + }, + { + "epoch": 1.2331482632138062, + "grad_norm": 1.7268129587173462, + "learning_rate": 6.766937991129641e-06, + "loss": 3.2488, + "step": 15382 + }, + { + "epoch": 1.2332284379948288, + "grad_norm": 1.698900818824768, + "learning_rate": 6.765709305202649e-06, + "loss": 3.1896, + "step": 15383 + }, + { + "epoch": 1.2333086127758515, + "grad_norm": 1.7279409170150757, + "learning_rate": 6.764480673804658e-06, + "loss": 3.2327, + "step": 15384 + }, + { + "epoch": 1.233388787556874, + "grad_norm": 1.7896642684936523, + "learning_rate": 6.763252096956385e-06, + "loss": 3.22, + "step": 15385 + }, + { + "epoch": 1.2334689623378967, + "grad_norm": 1.804122805595398, + "learning_rate": 6.76202357467854e-06, + "loss": 3.0585, + "step": 15386 + }, + { + "epoch": 1.2335491371189193, + "grad_norm": 0.679203987121582, + "learning_rate": 6.760795106991838e-06, + "loss": 2.4797, + "step": 15387 + }, + { + "epoch": 1.233629311899942, + "grad_norm": 1.7190786600112915, + "learning_rate": 6.759566693916986e-06, + "loss": 3.0612, + "step": 15388 + }, + { + "epoch": 1.2337094866809646, + "grad_norm": 1.7645106315612793, + "learning_rate": 6.758338335474697e-06, + "loss": 3.1865, + "step": 15389 + }, + { + "epoch": 1.2337896614619872, + "grad_norm": 1.7286525964736938, + "learning_rate": 6.757110031685683e-06, + "loss": 3.2062, + "step": 15390 + }, + { + "epoch": 1.2338698362430098, + "grad_norm": 1.875504970550537, + "learning_rate": 6.7558817825706505e-06, + "loss": 3.1948, + "step": 15391 + }, + { + "epoch": 1.2339500110240325, + "grad_norm": 1.809956669807434, + "learning_rate": 6.754653588150309e-06, + "loss": 3.2872, + "step": 15392 + }, + { + "epoch": 1.234030185805055, + "grad_norm": 1.923771858215332, + "learning_rate": 6.7534254484453636e-06, + "loss": 3.2544, + "step": 15393 + }, + { + "epoch": 1.2341103605860777, + "grad_norm": 1.778164267539978, + "learning_rate": 6.752197363476518e-06, + "loss": 3.1105, + "step": 15394 + }, + { + "epoch": 1.2341905353671003, + "grad_norm": 0.6909009218215942, + "learning_rate": 6.750969333264485e-06, + "loss": 2.4697, + "step": 15395 + }, + { + "epoch": 1.234270710148123, + "grad_norm": 0.7003875374794006, + "learning_rate": 6.749741357829961e-06, + "loss": 2.4739, + "step": 15396 + }, + { + "epoch": 1.2343508849291456, + "grad_norm": 1.7501587867736816, + "learning_rate": 6.748513437193653e-06, + "loss": 3.2842, + "step": 15397 + }, + { + "epoch": 1.2344310597101682, + "grad_norm": 1.7895163297653198, + "learning_rate": 6.747285571376263e-06, + "loss": 3.1947, + "step": 15398 + }, + { + "epoch": 1.2345112344911908, + "grad_norm": 1.7445155382156372, + "learning_rate": 6.74605776039849e-06, + "loss": 3.2428, + "step": 15399 + }, + { + "epoch": 1.2345914092722134, + "grad_norm": 1.7375332117080688, + "learning_rate": 6.744830004281039e-06, + "loss": 3.1852, + "step": 15400 + }, + { + "epoch": 1.234671584053236, + "grad_norm": 1.7808235883712769, + "learning_rate": 6.743602303044605e-06, + "loss": 3.1741, + "step": 15401 + }, + { + "epoch": 1.2347517588342587, + "grad_norm": 1.7631802558898926, + "learning_rate": 6.74237465670989e-06, + "loss": 3.1844, + "step": 15402 + }, + { + "epoch": 1.2348319336152813, + "grad_norm": 1.7477675676345825, + "learning_rate": 6.741147065297589e-06, + "loss": 3.254, + "step": 15403 + }, + { + "epoch": 1.234912108396304, + "grad_norm": 1.7575808763504028, + "learning_rate": 6.739919528828399e-06, + "loss": 3.1539, + "step": 15404 + }, + { + "epoch": 1.2349922831773266, + "grad_norm": 1.8157418966293335, + "learning_rate": 6.7386920473230186e-06, + "loss": 3.1987, + "step": 15405 + }, + { + "epoch": 1.2350724579583492, + "grad_norm": 1.8293876647949219, + "learning_rate": 6.737464620802141e-06, + "loss": 3.0271, + "step": 15406 + }, + { + "epoch": 1.2351526327393718, + "grad_norm": 1.7751609086990356, + "learning_rate": 6.7362372492864605e-06, + "loss": 3.2535, + "step": 15407 + }, + { + "epoch": 1.2352328075203944, + "grad_norm": 1.787124752998352, + "learning_rate": 6.7350099327966674e-06, + "loss": 3.1998, + "step": 15408 + }, + { + "epoch": 1.235312982301417, + "grad_norm": 1.7224183082580566, + "learning_rate": 6.733782671353457e-06, + "loss": 3.0662, + "step": 15409 + }, + { + "epoch": 1.2353931570824397, + "grad_norm": 1.8018954992294312, + "learning_rate": 6.732555464977521e-06, + "loss": 3.2747, + "step": 15410 + }, + { + "epoch": 1.2354733318634623, + "grad_norm": 1.728179931640625, + "learning_rate": 6.731328313689547e-06, + "loss": 3.1081, + "step": 15411 + }, + { + "epoch": 1.235553506644485, + "grad_norm": 1.6921862363815308, + "learning_rate": 6.730101217510226e-06, + "loss": 3.0354, + "step": 15412 + }, + { + "epoch": 1.2356336814255076, + "grad_norm": 1.8887015581130981, + "learning_rate": 6.728874176460245e-06, + "loss": 3.1039, + "step": 15413 + }, + { + "epoch": 1.2357138562065302, + "grad_norm": 1.7838877439498901, + "learning_rate": 6.727647190560294e-06, + "loss": 3.1838, + "step": 15414 + }, + { + "epoch": 1.2357940309875528, + "grad_norm": 1.751196026802063, + "learning_rate": 6.726420259831058e-06, + "loss": 3.135, + "step": 15415 + }, + { + "epoch": 1.2358742057685754, + "grad_norm": 1.7680341005325317, + "learning_rate": 6.725193384293222e-06, + "loss": 3.0986, + "step": 15416 + }, + { + "epoch": 1.235954380549598, + "grad_norm": 0.6991583108901978, + "learning_rate": 6.723966563967472e-06, + "loss": 2.4753, + "step": 15417 + }, + { + "epoch": 1.2360345553306207, + "grad_norm": 1.8956842422485352, + "learning_rate": 6.72273979887449e-06, + "loss": 3.241, + "step": 15418 + }, + { + "epoch": 1.2361147301116433, + "grad_norm": 1.8955868482589722, + "learning_rate": 6.72151308903496e-06, + "loss": 3.2311, + "step": 15419 + }, + { + "epoch": 1.236194904892666, + "grad_norm": 1.7367862462997437, + "learning_rate": 6.720286434469566e-06, + "loss": 3.2157, + "step": 15420 + }, + { + "epoch": 1.2362750796736885, + "grad_norm": 1.9333583116531372, + "learning_rate": 6.719059835198985e-06, + "loss": 3.2454, + "step": 15421 + }, + { + "epoch": 1.2363552544547112, + "grad_norm": 1.7386428117752075, + "learning_rate": 6.717833291243901e-06, + "loss": 3.2124, + "step": 15422 + }, + { + "epoch": 1.2364354292357338, + "grad_norm": 1.8154743909835815, + "learning_rate": 6.716606802624988e-06, + "loss": 3.1292, + "step": 15423 + }, + { + "epoch": 1.2365156040167564, + "grad_norm": 1.767561674118042, + "learning_rate": 6.715380369362927e-06, + "loss": 3.1502, + "step": 15424 + }, + { + "epoch": 1.236595778797779, + "grad_norm": 1.727112889289856, + "learning_rate": 6.714153991478398e-06, + "loss": 3.1741, + "step": 15425 + }, + { + "epoch": 1.2366759535788017, + "grad_norm": 1.6979798078536987, + "learning_rate": 6.712927668992073e-06, + "loss": 3.1718, + "step": 15426 + }, + { + "epoch": 1.2367561283598245, + "grad_norm": 1.8157809972763062, + "learning_rate": 6.71170140192463e-06, + "loss": 3.2622, + "step": 15427 + }, + { + "epoch": 1.2368363031408471, + "grad_norm": 1.7575671672821045, + "learning_rate": 6.7104751902967414e-06, + "loss": 3.1682, + "step": 15428 + }, + { + "epoch": 1.2369164779218698, + "grad_norm": 1.7829340696334839, + "learning_rate": 6.709249034129082e-06, + "loss": 3.2184, + "step": 15429 + }, + { + "epoch": 1.2369966527028924, + "grad_norm": 1.9488400220870972, + "learning_rate": 6.708022933442327e-06, + "loss": 3.2003, + "step": 15430 + }, + { + "epoch": 1.237076827483915, + "grad_norm": 1.7667161226272583, + "learning_rate": 6.706796888257141e-06, + "loss": 3.2504, + "step": 15431 + }, + { + "epoch": 1.2371570022649376, + "grad_norm": 1.837302803993225, + "learning_rate": 6.7055708985942e-06, + "loss": 3.2381, + "step": 15432 + }, + { + "epoch": 1.2372371770459603, + "grad_norm": 1.7376734018325806, + "learning_rate": 6.704344964474173e-06, + "loss": 3.198, + "step": 15433 + }, + { + "epoch": 1.2373173518269829, + "grad_norm": 1.7714049816131592, + "learning_rate": 6.703119085917728e-06, + "loss": 3.3575, + "step": 15434 + }, + { + "epoch": 1.2373975266080055, + "grad_norm": 1.764140248298645, + "learning_rate": 6.701893262945535e-06, + "loss": 3.1729, + "step": 15435 + }, + { + "epoch": 1.2374777013890281, + "grad_norm": 1.912566900253296, + "learning_rate": 6.700667495578257e-06, + "loss": 3.2357, + "step": 15436 + }, + { + "epoch": 1.2375578761700508, + "grad_norm": 1.8872829675674438, + "learning_rate": 6.699441783836565e-06, + "loss": 3.3202, + "step": 15437 + }, + { + "epoch": 1.2376380509510734, + "grad_norm": 1.8626841306686401, + "learning_rate": 6.698216127741118e-06, + "loss": 3.1805, + "step": 15438 + }, + { + "epoch": 1.237718225732096, + "grad_norm": 1.7345311641693115, + "learning_rate": 6.696990527312584e-06, + "loss": 3.1219, + "step": 15439 + }, + { + "epoch": 1.2377984005131186, + "grad_norm": 1.732970952987671, + "learning_rate": 6.695764982571627e-06, + "loss": 3.1042, + "step": 15440 + }, + { + "epoch": 1.2378785752941412, + "grad_norm": 1.999303936958313, + "learning_rate": 6.694539493538906e-06, + "loss": 3.1744, + "step": 15441 + }, + { + "epoch": 1.2379587500751639, + "grad_norm": 1.8225173950195312, + "learning_rate": 6.693314060235088e-06, + "loss": 3.1738, + "step": 15442 + }, + { + "epoch": 1.2380389248561865, + "grad_norm": 1.8030815124511719, + "learning_rate": 6.692088682680825e-06, + "loss": 3.1558, + "step": 15443 + }, + { + "epoch": 1.2381190996372091, + "grad_norm": 1.8877490758895874, + "learning_rate": 6.690863360896782e-06, + "loss": 3.2817, + "step": 15444 + }, + { + "epoch": 1.2381992744182317, + "grad_norm": 1.6927013397216797, + "learning_rate": 6.689638094903619e-06, + "loss": 3.1439, + "step": 15445 + }, + { + "epoch": 1.2382794491992544, + "grad_norm": 1.7971012592315674, + "learning_rate": 6.688412884721989e-06, + "loss": 3.195, + "step": 15446 + }, + { + "epoch": 1.238359623980277, + "grad_norm": 1.8120075464248657, + "learning_rate": 6.687187730372553e-06, + "loss": 3.231, + "step": 15447 + }, + { + "epoch": 1.2384397987612996, + "grad_norm": 1.7542887926101685, + "learning_rate": 6.6859626318759605e-06, + "loss": 3.1257, + "step": 15448 + }, + { + "epoch": 1.2385199735423222, + "grad_norm": 0.6753889322280884, + "learning_rate": 6.684737589252872e-06, + "loss": 2.4624, + "step": 15449 + }, + { + "epoch": 1.2386001483233449, + "grad_norm": 1.6632091999053955, + "learning_rate": 6.683512602523941e-06, + "loss": 3.2675, + "step": 15450 + }, + { + "epoch": 1.2386803231043675, + "grad_norm": 1.7514371871948242, + "learning_rate": 6.682287671709815e-06, + "loss": 3.1639, + "step": 15451 + }, + { + "epoch": 1.23876049788539, + "grad_norm": 1.7912453413009644, + "learning_rate": 6.681062796831152e-06, + "loss": 3.2496, + "step": 15452 + }, + { + "epoch": 1.2388406726664127, + "grad_norm": 1.8340870141983032, + "learning_rate": 6.679837977908601e-06, + "loss": 3.2067, + "step": 15453 + }, + { + "epoch": 1.2389208474474354, + "grad_norm": 1.708410382270813, + "learning_rate": 6.678613214962809e-06, + "loss": 3.2355, + "step": 15454 + }, + { + "epoch": 1.239001022228458, + "grad_norm": 2.01282000541687, + "learning_rate": 6.677388508014431e-06, + "loss": 3.2559, + "step": 15455 + }, + { + "epoch": 1.2390811970094806, + "grad_norm": 1.7294785976409912, + "learning_rate": 6.676163857084109e-06, + "loss": 3.1948, + "step": 15456 + }, + { + "epoch": 1.2391613717905032, + "grad_norm": 1.7000799179077148, + "learning_rate": 6.674939262192495e-06, + "loss": 3.2479, + "step": 15457 + }, + { + "epoch": 1.2392415465715259, + "grad_norm": 1.7438431978225708, + "learning_rate": 6.6737147233602316e-06, + "loss": 3.2224, + "step": 15458 + }, + { + "epoch": 1.2393217213525485, + "grad_norm": 2.1562912464141846, + "learning_rate": 6.672490240607965e-06, + "loss": 3.2009, + "step": 15459 + }, + { + "epoch": 1.239401896133571, + "grad_norm": 1.8300347328186035, + "learning_rate": 6.671265813956344e-06, + "loss": 3.2224, + "step": 15460 + }, + { + "epoch": 1.2394820709145937, + "grad_norm": 1.7092803716659546, + "learning_rate": 6.670041443426004e-06, + "loss": 3.1683, + "step": 15461 + }, + { + "epoch": 1.2395622456956163, + "grad_norm": 1.7710756063461304, + "learning_rate": 6.668817129037595e-06, + "loss": 3.02, + "step": 15462 + }, + { + "epoch": 1.2396424204766392, + "grad_norm": 0.6729008555412292, + "learning_rate": 6.667592870811752e-06, + "loss": 2.4662, + "step": 15463 + }, + { + "epoch": 1.2397225952576618, + "grad_norm": 1.8159035444259644, + "learning_rate": 6.66636866876912e-06, + "loss": 3.1685, + "step": 15464 + }, + { + "epoch": 1.2398027700386844, + "grad_norm": 1.819459080696106, + "learning_rate": 6.665144522930339e-06, + "loss": 3.1838, + "step": 15465 + }, + { + "epoch": 1.239882944819707, + "grad_norm": 1.7644076347351074, + "learning_rate": 6.663920433316045e-06, + "loss": 3.2358, + "step": 15466 + }, + { + "epoch": 1.2399631196007297, + "grad_norm": 1.7353202104568481, + "learning_rate": 6.662696399946876e-06, + "loss": 3.1502, + "step": 15467 + }, + { + "epoch": 1.2400432943817523, + "grad_norm": 1.735926866531372, + "learning_rate": 6.661472422843469e-06, + "loss": 3.1497, + "step": 15468 + }, + { + "epoch": 1.240123469162775, + "grad_norm": 1.7332593202590942, + "learning_rate": 6.660248502026461e-06, + "loss": 3.2673, + "step": 15469 + }, + { + "epoch": 1.2402036439437976, + "grad_norm": 1.918316125869751, + "learning_rate": 6.659024637516488e-06, + "loss": 3.2234, + "step": 15470 + }, + { + "epoch": 1.2402838187248202, + "grad_norm": 1.768576979637146, + "learning_rate": 6.65780082933418e-06, + "loss": 3.1564, + "step": 15471 + }, + { + "epoch": 1.2403639935058428, + "grad_norm": 1.9667894840240479, + "learning_rate": 6.6565770775001735e-06, + "loss": 3.2367, + "step": 15472 + }, + { + "epoch": 1.2404441682868654, + "grad_norm": 1.777745246887207, + "learning_rate": 6.655353382035097e-06, + "loss": 3.0729, + "step": 15473 + }, + { + "epoch": 1.240524343067888, + "grad_norm": 1.7225446701049805, + "learning_rate": 6.654129742959585e-06, + "loss": 3.0891, + "step": 15474 + }, + { + "epoch": 1.2406045178489107, + "grad_norm": 1.7165613174438477, + "learning_rate": 6.652906160294268e-06, + "loss": 3.1855, + "step": 15475 + }, + { + "epoch": 1.2406846926299333, + "grad_norm": 1.7582008838653564, + "learning_rate": 6.6516826340597706e-06, + "loss": 3.2305, + "step": 15476 + }, + { + "epoch": 1.240764867410956, + "grad_norm": 1.7446341514587402, + "learning_rate": 6.650459164276725e-06, + "loss": 3.209, + "step": 15477 + }, + { + "epoch": 1.2408450421919786, + "grad_norm": 1.7626291513442993, + "learning_rate": 6.649235750965759e-06, + "loss": 3.1658, + "step": 15478 + }, + { + "epoch": 1.2409252169730012, + "grad_norm": 1.7851449251174927, + "learning_rate": 6.648012394147494e-06, + "loss": 3.2277, + "step": 15479 + }, + { + "epoch": 1.2410053917540238, + "grad_norm": 1.866366982460022, + "learning_rate": 6.646789093842563e-06, + "loss": 3.1814, + "step": 15480 + }, + { + "epoch": 1.2410855665350464, + "grad_norm": 1.7823338508605957, + "learning_rate": 6.645565850071583e-06, + "loss": 3.1456, + "step": 15481 + }, + { + "epoch": 1.241165741316069, + "grad_norm": 1.8257222175598145, + "learning_rate": 6.64434266285518e-06, + "loss": 3.1383, + "step": 15482 + }, + { + "epoch": 1.2412459160970917, + "grad_norm": 1.786060094833374, + "learning_rate": 6.643119532213979e-06, + "loss": 3.1488, + "step": 15483 + }, + { + "epoch": 1.2413260908781143, + "grad_norm": 1.7992959022521973, + "learning_rate": 6.641896458168599e-06, + "loss": 3.1263, + "step": 15484 + }, + { + "epoch": 1.241406265659137, + "grad_norm": 1.867470145225525, + "learning_rate": 6.640673440739662e-06, + "loss": 3.2726, + "step": 15485 + }, + { + "epoch": 1.2414864404401595, + "grad_norm": 1.7458125352859497, + "learning_rate": 6.639450479947784e-06, + "loss": 3.1462, + "step": 15486 + }, + { + "epoch": 1.2415666152211822, + "grad_norm": 1.7959953546524048, + "learning_rate": 6.6382275758135895e-06, + "loss": 3.0284, + "step": 15487 + }, + { + "epoch": 1.2416467900022048, + "grad_norm": 1.9043132066726685, + "learning_rate": 6.637004728357692e-06, + "loss": 3.3129, + "step": 15488 + }, + { + "epoch": 1.2417269647832274, + "grad_norm": 1.6498847007751465, + "learning_rate": 6.635781937600709e-06, + "loss": 3.0359, + "step": 15489 + }, + { + "epoch": 1.24180713956425, + "grad_norm": 1.717730164527893, + "learning_rate": 6.63455920356326e-06, + "loss": 3.0792, + "step": 15490 + }, + { + "epoch": 1.2418873143452727, + "grad_norm": 1.8195958137512207, + "learning_rate": 6.633336526265953e-06, + "loss": 3.2039, + "step": 15491 + }, + { + "epoch": 1.2419674891262953, + "grad_norm": 1.6899864673614502, + "learning_rate": 6.632113905729407e-06, + "loss": 3.1575, + "step": 15492 + }, + { + "epoch": 1.242047663907318, + "grad_norm": 1.8357300758361816, + "learning_rate": 6.630891341974235e-06, + "loss": 3.2089, + "step": 15493 + }, + { + "epoch": 1.2421278386883405, + "grad_norm": 1.7530959844589233, + "learning_rate": 6.629668835021046e-06, + "loss": 3.1785, + "step": 15494 + }, + { + "epoch": 1.2422080134693632, + "grad_norm": 1.8148995637893677, + "learning_rate": 6.628446384890453e-06, + "loss": 3.1791, + "step": 15495 + }, + { + "epoch": 1.2422881882503858, + "grad_norm": 1.7611607313156128, + "learning_rate": 6.627223991603065e-06, + "loss": 3.1508, + "step": 15496 + }, + { + "epoch": 1.2423683630314084, + "grad_norm": 1.7937393188476562, + "learning_rate": 6.626001655179492e-06, + "loss": 3.1544, + "step": 15497 + }, + { + "epoch": 1.242448537812431, + "grad_norm": 1.8057639598846436, + "learning_rate": 6.624779375640343e-06, + "loss": 3.1173, + "step": 15498 + }, + { + "epoch": 1.2425287125934537, + "grad_norm": 1.7445731163024902, + "learning_rate": 6.6235571530062235e-06, + "loss": 3.163, + "step": 15499 + }, + { + "epoch": 1.2426088873744763, + "grad_norm": 1.8446505069732666, + "learning_rate": 6.622334987297741e-06, + "loss": 3.1744, + "step": 15500 + }, + { + "epoch": 1.242689062155499, + "grad_norm": 1.8363028764724731, + "learning_rate": 6.621112878535498e-06, + "loss": 3.1692, + "step": 15501 + }, + { + "epoch": 1.2427692369365215, + "grad_norm": 1.9496256113052368, + "learning_rate": 6.619890826740103e-06, + "loss": 3.1387, + "step": 15502 + }, + { + "epoch": 1.2428494117175442, + "grad_norm": 1.8885282278060913, + "learning_rate": 6.618668831932158e-06, + "loss": 3.1969, + "step": 15503 + }, + { + "epoch": 1.2429295864985668, + "grad_norm": 1.6481252908706665, + "learning_rate": 6.617446894132264e-06, + "loss": 3.052, + "step": 15504 + }, + { + "epoch": 1.2430097612795894, + "grad_norm": 1.7930923700332642, + "learning_rate": 6.616225013361024e-06, + "loss": 3.1604, + "step": 15505 + }, + { + "epoch": 1.243089936060612, + "grad_norm": 1.7258025407791138, + "learning_rate": 6.615003189639035e-06, + "loss": 3.2369, + "step": 15506 + }, + { + "epoch": 1.2431701108416346, + "grad_norm": 1.8288414478302002, + "learning_rate": 6.613781422986902e-06, + "loss": 3.0752, + "step": 15507 + }, + { + "epoch": 1.2432502856226573, + "grad_norm": 1.7833003997802734, + "learning_rate": 6.612559713425221e-06, + "loss": 3.2381, + "step": 15508 + }, + { + "epoch": 1.2433304604036801, + "grad_norm": 1.6885651350021362, + "learning_rate": 6.611338060974587e-06, + "loss": 3.2026, + "step": 15509 + }, + { + "epoch": 1.2434106351847027, + "grad_norm": 1.8444533348083496, + "learning_rate": 6.610116465655601e-06, + "loss": 3.1483, + "step": 15510 + }, + { + "epoch": 1.2434908099657254, + "grad_norm": 1.7662299871444702, + "learning_rate": 6.608894927488856e-06, + "loss": 3.2145, + "step": 15511 + }, + { + "epoch": 1.243570984746748, + "grad_norm": 1.7017184495925903, + "learning_rate": 6.607673446494947e-06, + "loss": 3.0875, + "step": 15512 + }, + { + "epoch": 1.2436511595277706, + "grad_norm": 1.7837555408477783, + "learning_rate": 6.60645202269447e-06, + "loss": 3.2343, + "step": 15513 + }, + { + "epoch": 1.2437313343087932, + "grad_norm": 1.7397392988204956, + "learning_rate": 6.605230656108013e-06, + "loss": 3.1902, + "step": 15514 + }, + { + "epoch": 1.2438115090898159, + "grad_norm": 1.6762754917144775, + "learning_rate": 6.604009346756173e-06, + "loss": 3.0544, + "step": 15515 + }, + { + "epoch": 1.2438916838708385, + "grad_norm": 1.7429888248443604, + "learning_rate": 6.602788094659537e-06, + "loss": 3.1808, + "step": 15516 + }, + { + "epoch": 1.243971858651861, + "grad_norm": 1.7721463441848755, + "learning_rate": 6.601566899838696e-06, + "loss": 3.1039, + "step": 15517 + }, + { + "epoch": 1.2440520334328837, + "grad_norm": 1.6435109376907349, + "learning_rate": 6.600345762314242e-06, + "loss": 3.1257, + "step": 15518 + }, + { + "epoch": 1.2441322082139064, + "grad_norm": 1.86740243434906, + "learning_rate": 6.599124682106757e-06, + "loss": 3.3321, + "step": 15519 + }, + { + "epoch": 1.244212382994929, + "grad_norm": 1.9395177364349365, + "learning_rate": 6.597903659236835e-06, + "loss": 3.267, + "step": 15520 + }, + { + "epoch": 1.2442925577759516, + "grad_norm": 0.7180946469306946, + "learning_rate": 6.596682693725055e-06, + "loss": 2.5035, + "step": 15521 + }, + { + "epoch": 1.2443727325569742, + "grad_norm": 1.7243387699127197, + "learning_rate": 6.595461785592005e-06, + "loss": 3.1624, + "step": 15522 + }, + { + "epoch": 1.2444529073379968, + "grad_norm": 0.6936090588569641, + "learning_rate": 6.594240934858274e-06, + "loss": 2.4741, + "step": 15523 + }, + { + "epoch": 1.2445330821190195, + "grad_norm": 1.8062944412231445, + "learning_rate": 6.593020141544436e-06, + "loss": 3.2635, + "step": 15524 + }, + { + "epoch": 1.244613256900042, + "grad_norm": 1.7739068269729614, + "learning_rate": 6.59179940567108e-06, + "loss": 3.1083, + "step": 15525 + }, + { + "epoch": 1.2446934316810647, + "grad_norm": 1.8332831859588623, + "learning_rate": 6.590578727258784e-06, + "loss": 3.229, + "step": 15526 + }, + { + "epoch": 1.2447736064620873, + "grad_norm": 1.8143656253814697, + "learning_rate": 6.589358106328127e-06, + "loss": 3.2413, + "step": 15527 + }, + { + "epoch": 1.24485378124311, + "grad_norm": 1.7325325012207031, + "learning_rate": 6.588137542899695e-06, + "loss": 3.2413, + "step": 15528 + }, + { + "epoch": 1.2449339560241326, + "grad_norm": 1.8353387117385864, + "learning_rate": 6.586917036994058e-06, + "loss": 3.2014, + "step": 15529 + }, + { + "epoch": 1.2450141308051552, + "grad_norm": 1.765057921409607, + "learning_rate": 6.5856965886317996e-06, + "loss": 3.234, + "step": 15530 + }, + { + "epoch": 1.2450943055861778, + "grad_norm": 0.7107200026512146, + "learning_rate": 6.584476197833491e-06, + "loss": 2.4714, + "step": 15531 + }, + { + "epoch": 1.2451744803672005, + "grad_norm": 1.785469651222229, + "learning_rate": 6.583255864619709e-06, + "loss": 3.1764, + "step": 15532 + }, + { + "epoch": 1.245254655148223, + "grad_norm": 1.799031138420105, + "learning_rate": 6.582035589011032e-06, + "loss": 3.2173, + "step": 15533 + }, + { + "epoch": 1.2453348299292457, + "grad_norm": 1.9499763250350952, + "learning_rate": 6.580815371028029e-06, + "loss": 3.1253, + "step": 15534 + }, + { + "epoch": 1.2454150047102683, + "grad_norm": 1.837440848350525, + "learning_rate": 6.5795952106912745e-06, + "loss": 3.1495, + "step": 15535 + }, + { + "epoch": 1.245495179491291, + "grad_norm": 1.7538537979125977, + "learning_rate": 6.578375108021338e-06, + "loss": 3.262, + "step": 15536 + }, + { + "epoch": 1.2455753542723136, + "grad_norm": 1.8668195009231567, + "learning_rate": 6.57715506303879e-06, + "loss": 3.1307, + "step": 15537 + }, + { + "epoch": 1.2456555290533362, + "grad_norm": 1.8409898281097412, + "learning_rate": 6.575935075764205e-06, + "loss": 3.147, + "step": 15538 + }, + { + "epoch": 1.2457357038343588, + "grad_norm": 2.340430498123169, + "learning_rate": 6.5747151462181445e-06, + "loss": 3.2833, + "step": 15539 + }, + { + "epoch": 1.2458158786153815, + "grad_norm": 1.7761847972869873, + "learning_rate": 6.573495274421182e-06, + "loss": 3.235, + "step": 15540 + }, + { + "epoch": 1.245896053396404, + "grad_norm": 1.855570673942566, + "learning_rate": 6.57227546039388e-06, + "loss": 3.1597, + "step": 15541 + }, + { + "epoch": 1.2459762281774267, + "grad_norm": 1.8149279356002808, + "learning_rate": 6.571055704156806e-06, + "loss": 3.1178, + "step": 15542 + }, + { + "epoch": 1.2460564029584493, + "grad_norm": 1.776299238204956, + "learning_rate": 6.5698360057305255e-06, + "loss": 3.2054, + "step": 15543 + }, + { + "epoch": 1.2461365777394722, + "grad_norm": 1.790044903755188, + "learning_rate": 6.568616365135599e-06, + "loss": 3.267, + "step": 15544 + }, + { + "epoch": 1.2462167525204948, + "grad_norm": 1.7818373441696167, + "learning_rate": 6.567396782392594e-06, + "loss": 3.2202, + "step": 15545 + }, + { + "epoch": 1.2462969273015174, + "grad_norm": 0.694659411907196, + "learning_rate": 6.566177257522066e-06, + "loss": 2.4579, + "step": 15546 + }, + { + "epoch": 1.24637710208254, + "grad_norm": 1.8617817163467407, + "learning_rate": 6.5649577905445814e-06, + "loss": 3.1335, + "step": 15547 + }, + { + "epoch": 1.2464572768635627, + "grad_norm": 2.0201265811920166, + "learning_rate": 6.563738381480698e-06, + "loss": 3.2645, + "step": 15548 + }, + { + "epoch": 1.2465374516445853, + "grad_norm": 1.8079266548156738, + "learning_rate": 6.562519030350973e-06, + "loss": 3.1984, + "step": 15549 + }, + { + "epoch": 1.246617626425608, + "grad_norm": 1.779870629310608, + "learning_rate": 6.561299737175968e-06, + "loss": 3.1476, + "step": 15550 + }, + { + "epoch": 1.2466978012066305, + "grad_norm": 1.808668851852417, + "learning_rate": 6.5600805019762355e-06, + "loss": 3.2598, + "step": 15551 + }, + { + "epoch": 1.2467779759876532, + "grad_norm": 1.7266508340835571, + "learning_rate": 6.558861324772333e-06, + "loss": 3.1882, + "step": 15552 + }, + { + "epoch": 1.2468581507686758, + "grad_norm": 1.8542581796646118, + "learning_rate": 6.557642205584818e-06, + "loss": 3.1488, + "step": 15553 + }, + { + "epoch": 1.2469383255496984, + "grad_norm": 1.7416372299194336, + "learning_rate": 6.5564231444342405e-06, + "loss": 3.301, + "step": 15554 + }, + { + "epoch": 1.247018500330721, + "grad_norm": 1.7418023347854614, + "learning_rate": 6.555204141341156e-06, + "loss": 3.1129, + "step": 15555 + }, + { + "epoch": 1.2470986751117437, + "grad_norm": 1.815208077430725, + "learning_rate": 6.553985196326116e-06, + "loss": 3.0889, + "step": 15556 + }, + { + "epoch": 1.2471788498927663, + "grad_norm": 1.7688885927200317, + "learning_rate": 6.552766309409668e-06, + "loss": 3.2098, + "step": 15557 + }, + { + "epoch": 1.247259024673789, + "grad_norm": 1.796539306640625, + "learning_rate": 6.551547480612369e-06, + "loss": 3.1768, + "step": 15558 + }, + { + "epoch": 1.2473391994548115, + "grad_norm": 1.8174206018447876, + "learning_rate": 6.550328709954762e-06, + "loss": 3.3088, + "step": 15559 + }, + { + "epoch": 1.2474193742358342, + "grad_norm": 1.9525415897369385, + "learning_rate": 6.549109997457399e-06, + "loss": 3.1657, + "step": 15560 + }, + { + "epoch": 1.2474995490168568, + "grad_norm": 1.8386547565460205, + "learning_rate": 6.547891343140823e-06, + "loss": 3.1801, + "step": 15561 + }, + { + "epoch": 1.2475797237978794, + "grad_norm": 1.748464584350586, + "learning_rate": 6.546672747025582e-06, + "loss": 3.2064, + "step": 15562 + }, + { + "epoch": 1.247659898578902, + "grad_norm": 1.8006426095962524, + "learning_rate": 6.545454209132225e-06, + "loss": 3.0806, + "step": 15563 + }, + { + "epoch": 1.2477400733599247, + "grad_norm": 1.837697148323059, + "learning_rate": 6.5442357294812895e-06, + "loss": 3.1624, + "step": 15564 + }, + { + "epoch": 1.2478202481409473, + "grad_norm": 1.784207820892334, + "learning_rate": 6.543017308093322e-06, + "loss": 3.1464, + "step": 15565 + }, + { + "epoch": 1.24790042292197, + "grad_norm": 1.6975687742233276, + "learning_rate": 6.541798944988865e-06, + "loss": 3.2421, + "step": 15566 + }, + { + "epoch": 1.2479805977029925, + "grad_norm": 1.7426468133926392, + "learning_rate": 6.540580640188459e-06, + "loss": 3.1267, + "step": 15567 + }, + { + "epoch": 1.2480607724840151, + "grad_norm": 1.7536197900772095, + "learning_rate": 6.539362393712645e-06, + "loss": 3.0853, + "step": 15568 + }, + { + "epoch": 1.2481409472650378, + "grad_norm": 1.8355597257614136, + "learning_rate": 6.5381442055819585e-06, + "loss": 3.2626, + "step": 15569 + }, + { + "epoch": 1.2482211220460604, + "grad_norm": 1.8313900232315063, + "learning_rate": 6.536926075816944e-06, + "loss": 3.2843, + "step": 15570 + }, + { + "epoch": 1.248301296827083, + "grad_norm": 1.7558950185775757, + "learning_rate": 6.535708004438133e-06, + "loss": 3.2115, + "step": 15571 + }, + { + "epoch": 1.2483814716081056, + "grad_norm": 1.7793748378753662, + "learning_rate": 6.534489991466064e-06, + "loss": 3.2161, + "step": 15572 + }, + { + "epoch": 1.2484616463891283, + "grad_norm": 1.8067338466644287, + "learning_rate": 6.533272036921274e-06, + "loss": 3.1884, + "step": 15573 + }, + { + "epoch": 1.248541821170151, + "grad_norm": 1.7641526460647583, + "learning_rate": 6.532054140824294e-06, + "loss": 3.0717, + "step": 15574 + }, + { + "epoch": 1.2486219959511735, + "grad_norm": 1.8430315256118774, + "learning_rate": 6.530836303195662e-06, + "loss": 3.1165, + "step": 15575 + }, + { + "epoch": 1.2487021707321961, + "grad_norm": 1.7834584712982178, + "learning_rate": 6.529618524055904e-06, + "loss": 3.1808, + "step": 15576 + }, + { + "epoch": 1.2487823455132188, + "grad_norm": 1.8096401691436768, + "learning_rate": 6.528400803425554e-06, + "loss": 3.149, + "step": 15577 + }, + { + "epoch": 1.2488625202942414, + "grad_norm": 1.815833330154419, + "learning_rate": 6.527183141325146e-06, + "loss": 3.2593, + "step": 15578 + }, + { + "epoch": 1.248942695075264, + "grad_norm": 1.8360439538955688, + "learning_rate": 6.525965537775205e-06, + "loss": 3.2672, + "step": 15579 + }, + { + "epoch": 1.2490228698562866, + "grad_norm": 1.841271162033081, + "learning_rate": 6.524747992796262e-06, + "loss": 3.1708, + "step": 15580 + }, + { + "epoch": 1.2491030446373093, + "grad_norm": 1.795475721359253, + "learning_rate": 6.5235305064088406e-06, + "loss": 3.1543, + "step": 15581 + }, + { + "epoch": 1.2491832194183319, + "grad_norm": 1.7839494943618774, + "learning_rate": 6.522313078633471e-06, + "loss": 3.2281, + "step": 15582 + }, + { + "epoch": 1.2492633941993545, + "grad_norm": 1.706966757774353, + "learning_rate": 6.5210957094906766e-06, + "loss": 3.16, + "step": 15583 + }, + { + "epoch": 1.2493435689803771, + "grad_norm": 1.748000144958496, + "learning_rate": 6.519878399000983e-06, + "loss": 3.0203, + "step": 15584 + }, + { + "epoch": 1.2494237437613998, + "grad_norm": 1.6944913864135742, + "learning_rate": 6.518661147184915e-06, + "loss": 3.1331, + "step": 15585 + }, + { + "epoch": 1.2495039185424224, + "grad_norm": 1.7253179550170898, + "learning_rate": 6.51744395406299e-06, + "loss": 3.1881, + "step": 15586 + }, + { + "epoch": 1.249584093323445, + "grad_norm": 1.780490756034851, + "learning_rate": 6.516226819655732e-06, + "loss": 3.1209, + "step": 15587 + }, + { + "epoch": 1.2496642681044676, + "grad_norm": 1.8032537698745728, + "learning_rate": 6.515009743983664e-06, + "loss": 3.1196, + "step": 15588 + }, + { + "epoch": 1.2497444428854902, + "grad_norm": 1.7093734741210938, + "learning_rate": 6.513792727067303e-06, + "loss": 3.1348, + "step": 15589 + }, + { + "epoch": 1.249824617666513, + "grad_norm": 1.960822582244873, + "learning_rate": 6.512575768927169e-06, + "loss": 3.2285, + "step": 15590 + }, + { + "epoch": 1.2499047924475357, + "grad_norm": 1.8002091646194458, + "learning_rate": 6.5113588695837775e-06, + "loss": 3.2605, + "step": 15591 + }, + { + "epoch": 1.2499849672285583, + "grad_norm": 1.811944842338562, + "learning_rate": 6.5101420290576444e-06, + "loss": 3.2991, + "step": 15592 + }, + { + "epoch": 1.250065142009581, + "grad_norm": 1.6510541439056396, + "learning_rate": 6.50892524736929e-06, + "loss": 3.1194, + "step": 15593 + }, + { + "epoch": 1.2501453167906036, + "grad_norm": 1.834452748298645, + "learning_rate": 6.507708524539222e-06, + "loss": 3.1774, + "step": 15594 + }, + { + "epoch": 1.2502254915716262, + "grad_norm": 1.8255820274353027, + "learning_rate": 6.506491860587962e-06, + "loss": 3.1281, + "step": 15595 + }, + { + "epoch": 1.2503056663526488, + "grad_norm": 1.7635741233825684, + "learning_rate": 6.505275255536014e-06, + "loss": 3.2538, + "step": 15596 + }, + { + "epoch": 1.2503858411336715, + "grad_norm": 1.7937003374099731, + "learning_rate": 6.504058709403894e-06, + "loss": 3.2048, + "step": 15597 + }, + { + "epoch": 1.250466015914694, + "grad_norm": 1.6289576292037964, + "learning_rate": 6.502842222212115e-06, + "loss": 3.145, + "step": 15598 + }, + { + "epoch": 1.2505461906957167, + "grad_norm": 1.8915563821792603, + "learning_rate": 6.50162579398118e-06, + "loss": 3.2109, + "step": 15599 + }, + { + "epoch": 1.2506263654767393, + "grad_norm": 0.6834831833839417, + "learning_rate": 6.500409424731604e-06, + "loss": 2.4696, + "step": 15600 + }, + { + "epoch": 1.250706540257762, + "grad_norm": 1.7214176654815674, + "learning_rate": 6.49919311448389e-06, + "loss": 3.0959, + "step": 15601 + }, + { + "epoch": 1.2507867150387846, + "grad_norm": 1.798566222190857, + "learning_rate": 6.497976863258546e-06, + "loss": 3.1514, + "step": 15602 + }, + { + "epoch": 1.2508668898198072, + "grad_norm": 1.7982182502746582, + "learning_rate": 6.496760671076079e-06, + "loss": 3.1284, + "step": 15603 + }, + { + "epoch": 1.2509470646008298, + "grad_norm": 1.7944368124008179, + "learning_rate": 6.495544537956991e-06, + "loss": 3.1672, + "step": 15604 + }, + { + "epoch": 1.2510272393818525, + "grad_norm": 1.7314534187316895, + "learning_rate": 6.494328463921791e-06, + "loss": 3.1376, + "step": 15605 + }, + { + "epoch": 1.251107414162875, + "grad_norm": 1.8403750658035278, + "learning_rate": 6.493112448990973e-06, + "loss": 3.1854, + "step": 15606 + }, + { + "epoch": 1.2511875889438977, + "grad_norm": 1.7468342781066895, + "learning_rate": 6.491896493185044e-06, + "loss": 3.2039, + "step": 15607 + }, + { + "epoch": 1.2512677637249203, + "grad_norm": 0.6936105489730835, + "learning_rate": 6.4906805965245055e-06, + "loss": 2.4645, + "step": 15608 + }, + { + "epoch": 1.251347938505943, + "grad_norm": 1.6579461097717285, + "learning_rate": 6.489464759029853e-06, + "loss": 3.114, + "step": 15609 + }, + { + "epoch": 1.2514281132869656, + "grad_norm": 1.7333968877792358, + "learning_rate": 6.48824898072159e-06, + "loss": 3.2463, + "step": 15610 + }, + { + "epoch": 1.2515082880679882, + "grad_norm": 0.6800491213798523, + "learning_rate": 6.48703326162021e-06, + "loss": 2.4727, + "step": 15611 + }, + { + "epoch": 1.2515884628490108, + "grad_norm": 1.8475722074508667, + "learning_rate": 6.48581760174621e-06, + "loss": 3.1656, + "step": 15612 + }, + { + "epoch": 1.2516686376300334, + "grad_norm": 1.7487492561340332, + "learning_rate": 6.48460200112009e-06, + "loss": 3.114, + "step": 15613 + }, + { + "epoch": 1.251748812411056, + "grad_norm": 1.718653678894043, + "learning_rate": 6.483386459762339e-06, + "loss": 3.1046, + "step": 15614 + }, + { + "epoch": 1.2518289871920787, + "grad_norm": 1.7836307287216187, + "learning_rate": 6.482170977693454e-06, + "loss": 3.2539, + "step": 15615 + }, + { + "epoch": 1.2519091619731013, + "grad_norm": 1.8740756511688232, + "learning_rate": 6.480955554933926e-06, + "loss": 3.2898, + "step": 15616 + }, + { + "epoch": 1.251989336754124, + "grad_norm": 1.7448407411575317, + "learning_rate": 6.479740191504244e-06, + "loss": 3.1245, + "step": 15617 + }, + { + "epoch": 1.2520695115351466, + "grad_norm": 1.8454822301864624, + "learning_rate": 6.478524887424907e-06, + "loss": 3.2247, + "step": 15618 + }, + { + "epoch": 1.2521496863161692, + "grad_norm": 1.8671005964279175, + "learning_rate": 6.477309642716397e-06, + "loss": 3.1829, + "step": 15619 + }, + { + "epoch": 1.2522298610971918, + "grad_norm": 0.6758650541305542, + "learning_rate": 6.476094457399206e-06, + "loss": 2.4581, + "step": 15620 + }, + { + "epoch": 1.2523100358782144, + "grad_norm": 1.8232120275497437, + "learning_rate": 6.474879331493818e-06, + "loss": 3.2255, + "step": 15621 + }, + { + "epoch": 1.252390210659237, + "grad_norm": 1.7602653503417969, + "learning_rate": 6.4736642650207225e-06, + "loss": 3.2114, + "step": 15622 + }, + { + "epoch": 1.25247038544026, + "grad_norm": 1.723705530166626, + "learning_rate": 6.472449258000407e-06, + "loss": 3.2456, + "step": 15623 + }, + { + "epoch": 1.2525505602212825, + "grad_norm": 1.952025294303894, + "learning_rate": 6.471234310453352e-06, + "loss": 3.205, + "step": 15624 + }, + { + "epoch": 1.2526307350023052, + "grad_norm": 1.8317986726760864, + "learning_rate": 6.470019422400045e-06, + "loss": 3.1595, + "step": 15625 + }, + { + "epoch": 1.2527109097833278, + "grad_norm": 1.81025230884552, + "learning_rate": 6.468804593860963e-06, + "loss": 3.2289, + "step": 15626 + }, + { + "epoch": 1.2527910845643504, + "grad_norm": 1.7831615209579468, + "learning_rate": 6.467589824856593e-06, + "loss": 3.166, + "step": 15627 + }, + { + "epoch": 1.252871259345373, + "grad_norm": 1.7676317691802979, + "learning_rate": 6.466375115407415e-06, + "loss": 3.1143, + "step": 15628 + }, + { + "epoch": 1.2529514341263956, + "grad_norm": 1.7087775468826294, + "learning_rate": 6.465160465533904e-06, + "loss": 3.1987, + "step": 15629 + }, + { + "epoch": 1.2530316089074183, + "grad_norm": 0.7013884782791138, + "learning_rate": 6.463945875256546e-06, + "loss": 2.4781, + "step": 15630 + }, + { + "epoch": 1.253111783688441, + "grad_norm": 1.7127821445465088, + "learning_rate": 6.46273134459581e-06, + "loss": 3.2415, + "step": 15631 + }, + { + "epoch": 1.2531919584694635, + "grad_norm": 1.8282785415649414, + "learning_rate": 6.461516873572178e-06, + "loss": 3.1962, + "step": 15632 + }, + { + "epoch": 1.2532721332504861, + "grad_norm": 1.7615728378295898, + "learning_rate": 6.460302462206125e-06, + "loss": 3.2502, + "step": 15633 + }, + { + "epoch": 1.2533523080315088, + "grad_norm": 1.7973487377166748, + "learning_rate": 6.459088110518125e-06, + "loss": 3.2109, + "step": 15634 + }, + { + "epoch": 1.2534324828125314, + "grad_norm": 1.8142050504684448, + "learning_rate": 6.457873818528653e-06, + "loss": 3.1633, + "step": 15635 + }, + { + "epoch": 1.253512657593554, + "grad_norm": 1.7536497116088867, + "learning_rate": 6.4566595862581774e-06, + "loss": 3.0968, + "step": 15636 + }, + { + "epoch": 1.2535928323745766, + "grad_norm": 1.7559326887130737, + "learning_rate": 6.455445413727173e-06, + "loss": 3.1679, + "step": 15637 + }, + { + "epoch": 1.2536730071555993, + "grad_norm": 1.7468678951263428, + "learning_rate": 6.454231300956111e-06, + "loss": 3.1796, + "step": 15638 + }, + { + "epoch": 1.2537531819366219, + "grad_norm": 1.7276729345321655, + "learning_rate": 6.453017247965459e-06, + "loss": 3.168, + "step": 15639 + }, + { + "epoch": 1.2538333567176445, + "grad_norm": 1.7320111989974976, + "learning_rate": 6.451803254775686e-06, + "loss": 3.1914, + "step": 15640 + }, + { + "epoch": 1.2539135314986671, + "grad_norm": 1.7845149040222168, + "learning_rate": 6.450589321407259e-06, + "loss": 3.194, + "step": 15641 + }, + { + "epoch": 1.2539937062796898, + "grad_norm": 1.9544974565505981, + "learning_rate": 6.449375447880646e-06, + "loss": 3.2241, + "step": 15642 + }, + { + "epoch": 1.2540738810607124, + "grad_norm": 1.7666763067245483, + "learning_rate": 6.448161634216313e-06, + "loss": 3.1718, + "step": 15643 + }, + { + "epoch": 1.254154055841735, + "grad_norm": 1.8435806035995483, + "learning_rate": 6.4469478804347194e-06, + "loss": 3.1743, + "step": 15644 + }, + { + "epoch": 1.2542342306227576, + "grad_norm": 0.6870036721229553, + "learning_rate": 6.445734186556335e-06, + "loss": 2.4712, + "step": 15645 + }, + { + "epoch": 1.2543144054037803, + "grad_norm": 1.775694727897644, + "learning_rate": 6.444520552601619e-06, + "loss": 3.2171, + "step": 15646 + }, + { + "epoch": 1.2543945801848029, + "grad_norm": 1.639920949935913, + "learning_rate": 6.443306978591032e-06, + "loss": 3.1708, + "step": 15647 + }, + { + "epoch": 1.2544747549658255, + "grad_norm": 1.7230782508850098, + "learning_rate": 6.442093464545038e-06, + "loss": 3.1945, + "step": 15648 + }, + { + "epoch": 1.2545549297468481, + "grad_norm": 1.9257546663284302, + "learning_rate": 6.440880010484092e-06, + "loss": 3.1361, + "step": 15649 + }, + { + "epoch": 1.2546351045278707, + "grad_norm": 1.6974588632583618, + "learning_rate": 6.439666616428657e-06, + "loss": 3.0744, + "step": 15650 + }, + { + "epoch": 1.2547152793088934, + "grad_norm": 1.7499983310699463, + "learning_rate": 6.438453282399186e-06, + "loss": 3.2017, + "step": 15651 + }, + { + "epoch": 1.254795454089916, + "grad_norm": 1.661141276359558, + "learning_rate": 6.437240008416139e-06, + "loss": 3.1437, + "step": 15652 + }, + { + "epoch": 1.2548756288709386, + "grad_norm": 1.952286720275879, + "learning_rate": 6.43602679449997e-06, + "loss": 3.1923, + "step": 15653 + }, + { + "epoch": 1.2549558036519612, + "grad_norm": 1.7220640182495117, + "learning_rate": 6.4348136406711314e-06, + "loss": 3.1526, + "step": 15654 + }, + { + "epoch": 1.2550359784329839, + "grad_norm": 0.6958615183830261, + "learning_rate": 6.433600546950081e-06, + "loss": 2.4689, + "step": 15655 + }, + { + "epoch": 1.2551161532140065, + "grad_norm": 1.8134773969650269, + "learning_rate": 6.432387513357267e-06, + "loss": 3.1221, + "step": 15656 + }, + { + "epoch": 1.2551963279950291, + "grad_norm": 1.885467290878296, + "learning_rate": 6.431174539913141e-06, + "loss": 3.2708, + "step": 15657 + }, + { + "epoch": 1.2552765027760517, + "grad_norm": 1.763952612876892, + "learning_rate": 6.429961626638157e-06, + "loss": 3.1267, + "step": 15658 + }, + { + "epoch": 1.2553566775570744, + "grad_norm": 1.783872127532959, + "learning_rate": 6.42874877355276e-06, + "loss": 3.1892, + "step": 15659 + }, + { + "epoch": 1.255436852338097, + "grad_norm": 1.862634301185608, + "learning_rate": 6.4275359806774006e-06, + "loss": 3.2379, + "step": 15660 + }, + { + "epoch": 1.2555170271191196, + "grad_norm": 1.7813282012939453, + "learning_rate": 6.426323248032525e-06, + "loss": 3.1296, + "step": 15661 + }, + { + "epoch": 1.2555972019001422, + "grad_norm": 1.6852446794509888, + "learning_rate": 6.425110575638579e-06, + "loss": 3.101, + "step": 15662 + }, + { + "epoch": 1.2556773766811649, + "grad_norm": 1.7593271732330322, + "learning_rate": 6.4238979635160116e-06, + "loss": 3.0789, + "step": 15663 + }, + { + "epoch": 1.2557575514621875, + "grad_norm": 1.8203673362731934, + "learning_rate": 6.422685411685261e-06, + "loss": 3.3695, + "step": 15664 + }, + { + "epoch": 1.25583772624321, + "grad_norm": 0.6700478196144104, + "learning_rate": 6.421472920166777e-06, + "loss": 2.4786, + "step": 15665 + }, + { + "epoch": 1.2559179010242327, + "grad_norm": 1.82180655002594, + "learning_rate": 6.420260488980996e-06, + "loss": 3.1296, + "step": 15666 + }, + { + "epoch": 1.2559980758052554, + "grad_norm": 1.8465194702148438, + "learning_rate": 6.419048118148361e-06, + "loss": 3.2371, + "step": 15667 + }, + { + "epoch": 1.256078250586278, + "grad_norm": 1.806882381439209, + "learning_rate": 6.417835807689314e-06, + "loss": 3.1978, + "step": 15668 + }, + { + "epoch": 1.2561584253673006, + "grad_norm": 1.773751139640808, + "learning_rate": 6.416623557624293e-06, + "loss": 3.1838, + "step": 15669 + }, + { + "epoch": 1.2562386001483232, + "grad_norm": 1.7420668601989746, + "learning_rate": 6.415411367973734e-06, + "loss": 3.1802, + "step": 15670 + }, + { + "epoch": 1.2563187749293458, + "grad_norm": 1.7252155542373657, + "learning_rate": 6.414199238758079e-06, + "loss": 3.1027, + "step": 15671 + }, + { + "epoch": 1.2563989497103685, + "grad_norm": 1.7384414672851562, + "learning_rate": 6.41298716999776e-06, + "loss": 3.1897, + "step": 15672 + }, + { + "epoch": 1.256479124491391, + "grad_norm": 1.7453744411468506, + "learning_rate": 6.411775161713213e-06, + "loss": 3.2643, + "step": 15673 + }, + { + "epoch": 1.256559299272414, + "grad_norm": 1.774928092956543, + "learning_rate": 6.4105632139248706e-06, + "loss": 3.3044, + "step": 15674 + }, + { + "epoch": 1.2566394740534366, + "grad_norm": 1.7491706609725952, + "learning_rate": 6.409351326653168e-06, + "loss": 3.2773, + "step": 15675 + }, + { + "epoch": 1.2567196488344592, + "grad_norm": 1.856870174407959, + "learning_rate": 6.4081394999185385e-06, + "loss": 3.1379, + "step": 15676 + }, + { + "epoch": 1.2567998236154818, + "grad_norm": 1.7254866361618042, + "learning_rate": 6.406927733741409e-06, + "loss": 3.2379, + "step": 15677 + }, + { + "epoch": 1.2568799983965044, + "grad_norm": 2.1347997188568115, + "learning_rate": 6.405716028142214e-06, + "loss": 3.2208, + "step": 15678 + }, + { + "epoch": 1.256960173177527, + "grad_norm": 1.6907647848129272, + "learning_rate": 6.404504383141378e-06, + "loss": 3.1055, + "step": 15679 + }, + { + "epoch": 1.2570403479585497, + "grad_norm": 1.7629445791244507, + "learning_rate": 6.403292798759331e-06, + "loss": 3.1379, + "step": 15680 + }, + { + "epoch": 1.2571205227395723, + "grad_norm": 1.8390251398086548, + "learning_rate": 6.402081275016501e-06, + "loss": 3.1341, + "step": 15681 + }, + { + "epoch": 1.257200697520595, + "grad_norm": 0.6761904358863831, + "learning_rate": 6.40086981193331e-06, + "loss": 2.4485, + "step": 15682 + }, + { + "epoch": 1.2572808723016176, + "grad_norm": 1.8459017276763916, + "learning_rate": 6.399658409530189e-06, + "loss": 3.2928, + "step": 15683 + }, + { + "epoch": 1.2573610470826402, + "grad_norm": 1.762610673904419, + "learning_rate": 6.398447067827556e-06, + "loss": 3.1842, + "step": 15684 + }, + { + "epoch": 1.2574412218636628, + "grad_norm": 1.788871169090271, + "learning_rate": 6.3972357868458364e-06, + "loss": 3.1624, + "step": 15685 + }, + { + "epoch": 1.2575213966446854, + "grad_norm": 1.8756877183914185, + "learning_rate": 6.396024566605453e-06, + "loss": 3.2919, + "step": 15686 + }, + { + "epoch": 1.257601571425708, + "grad_norm": 1.7520478963851929, + "learning_rate": 6.394813407126823e-06, + "loss": 3.2611, + "step": 15687 + }, + { + "epoch": 1.2576817462067307, + "grad_norm": 1.7072041034698486, + "learning_rate": 6.393602308430371e-06, + "loss": 3.2479, + "step": 15688 + }, + { + "epoch": 1.2577619209877533, + "grad_norm": 2.010051727294922, + "learning_rate": 6.392391270536511e-06, + "loss": 3.314, + "step": 15689 + }, + { + "epoch": 1.257842095768776, + "grad_norm": 0.713638186454773, + "learning_rate": 6.391180293465664e-06, + "loss": 2.4733, + "step": 15690 + }, + { + "epoch": 1.2579222705497985, + "grad_norm": 1.784719467163086, + "learning_rate": 6.389969377238246e-06, + "loss": 3.1076, + "step": 15691 + }, + { + "epoch": 1.2580024453308212, + "grad_norm": 1.8921440839767456, + "learning_rate": 6.3887585218746694e-06, + "loss": 3.1667, + "step": 15692 + }, + { + "epoch": 1.2580826201118438, + "grad_norm": 1.7881277799606323, + "learning_rate": 6.387547727395355e-06, + "loss": 3.1997, + "step": 15693 + }, + { + "epoch": 1.2581627948928664, + "grad_norm": 1.6175888776779175, + "learning_rate": 6.38633699382071e-06, + "loss": 3.2051, + "step": 15694 + }, + { + "epoch": 1.258242969673889, + "grad_norm": 1.8414584398269653, + "learning_rate": 6.385126321171149e-06, + "loss": 3.1614, + "step": 15695 + }, + { + "epoch": 1.2583231444549117, + "grad_norm": 1.7852873802185059, + "learning_rate": 6.383915709467087e-06, + "loss": 3.1854, + "step": 15696 + }, + { + "epoch": 1.2584033192359343, + "grad_norm": 1.8278952836990356, + "learning_rate": 6.382705158728929e-06, + "loss": 3.33, + "step": 15697 + }, + { + "epoch": 1.258483494016957, + "grad_norm": 1.7420616149902344, + "learning_rate": 6.381494668977091e-06, + "loss": 3.1213, + "step": 15698 + }, + { + "epoch": 1.2585636687979795, + "grad_norm": 1.7103750705718994, + "learning_rate": 6.380284240231973e-06, + "loss": 3.0609, + "step": 15699 + }, + { + "epoch": 1.2586438435790022, + "grad_norm": 1.6745564937591553, + "learning_rate": 6.379073872513988e-06, + "loss": 3.2261, + "step": 15700 + }, + { + "epoch": 1.2587240183600248, + "grad_norm": 1.6877734661102295, + "learning_rate": 6.377863565843544e-06, + "loss": 3.0669, + "step": 15701 + }, + { + "epoch": 1.2588041931410474, + "grad_norm": 1.7835791110992432, + "learning_rate": 6.376653320241042e-06, + "loss": 3.272, + "step": 15702 + }, + { + "epoch": 1.25888436792207, + "grad_norm": 1.8261587619781494, + "learning_rate": 6.375443135726888e-06, + "loss": 3.2274, + "step": 15703 + }, + { + "epoch": 1.2589645427030927, + "grad_norm": 1.678435206413269, + "learning_rate": 6.374233012321486e-06, + "loss": 3.1962, + "step": 15704 + }, + { + "epoch": 1.2590447174841155, + "grad_norm": 1.7814844846725464, + "learning_rate": 6.373022950045236e-06, + "loss": 3.1185, + "step": 15705 + }, + { + "epoch": 1.2591248922651381, + "grad_norm": 1.8084139823913574, + "learning_rate": 6.371812948918543e-06, + "loss": 3.2767, + "step": 15706 + }, + { + "epoch": 1.2592050670461608, + "grad_norm": 1.7063579559326172, + "learning_rate": 6.370603008961803e-06, + "loss": 3.3501, + "step": 15707 + }, + { + "epoch": 1.2592852418271834, + "grad_norm": 1.6809906959533691, + "learning_rate": 6.36939313019542e-06, + "loss": 3.168, + "step": 15708 + }, + { + "epoch": 1.259365416608206, + "grad_norm": 1.9356521368026733, + "learning_rate": 6.368183312639786e-06, + "loss": 3.2301, + "step": 15709 + }, + { + "epoch": 1.2594455913892286, + "grad_norm": 1.7339847087860107, + "learning_rate": 6.366973556315302e-06, + "loss": 3.2157, + "step": 15710 + }, + { + "epoch": 1.2595257661702512, + "grad_norm": 1.6782197952270508, + "learning_rate": 6.365763861242364e-06, + "loss": 3.1436, + "step": 15711 + }, + { + "epoch": 1.2596059409512739, + "grad_norm": 1.7268959283828735, + "learning_rate": 6.364554227441366e-06, + "loss": 3.2872, + "step": 15712 + }, + { + "epoch": 1.2596861157322965, + "grad_norm": 1.7536982297897339, + "learning_rate": 6.3633446549327045e-06, + "loss": 3.1917, + "step": 15713 + }, + { + "epoch": 1.2597662905133191, + "grad_norm": 1.751343846321106, + "learning_rate": 6.3621351437367685e-06, + "loss": 3.2593, + "step": 15714 + }, + { + "epoch": 1.2598464652943417, + "grad_norm": 0.6646305322647095, + "learning_rate": 6.360925693873952e-06, + "loss": 2.4664, + "step": 15715 + }, + { + "epoch": 1.2599266400753644, + "grad_norm": 1.8125907182693481, + "learning_rate": 6.359716305364647e-06, + "loss": 3.1481, + "step": 15716 + }, + { + "epoch": 1.260006814856387, + "grad_norm": 1.699989914894104, + "learning_rate": 6.35850697822924e-06, + "loss": 3.1205, + "step": 15717 + }, + { + "epoch": 1.2600869896374096, + "grad_norm": 1.796266794204712, + "learning_rate": 6.357297712488124e-06, + "loss": 3.22, + "step": 15718 + }, + { + "epoch": 1.2601671644184322, + "grad_norm": 1.7053520679473877, + "learning_rate": 6.3560885081616844e-06, + "loss": 3.1609, + "step": 15719 + }, + { + "epoch": 1.2602473391994549, + "grad_norm": 0.6820704936981201, + "learning_rate": 6.354879365270306e-06, + "loss": 2.4702, + "step": 15720 + }, + { + "epoch": 1.2603275139804775, + "grad_norm": 1.675653100013733, + "learning_rate": 6.353670283834381e-06, + "loss": 3.1959, + "step": 15721 + }, + { + "epoch": 1.2604076887615001, + "grad_norm": 1.8320809602737427, + "learning_rate": 6.352461263874286e-06, + "loss": 3.0927, + "step": 15722 + }, + { + "epoch": 1.2604878635425227, + "grad_norm": 1.7174116373062134, + "learning_rate": 6.351252305410412e-06, + "loss": 3.2538, + "step": 15723 + }, + { + "epoch": 1.2605680383235454, + "grad_norm": 1.9258103370666504, + "learning_rate": 6.350043408463136e-06, + "loss": 3.2517, + "step": 15724 + }, + { + "epoch": 1.260648213104568, + "grad_norm": 1.7054089307785034, + "learning_rate": 6.34883457305284e-06, + "loss": 3.2321, + "step": 15725 + }, + { + "epoch": 1.2607283878855906, + "grad_norm": 1.772225022315979, + "learning_rate": 6.34762579919991e-06, + "loss": 3.1788, + "step": 15726 + }, + { + "epoch": 1.2608085626666132, + "grad_norm": 0.6870915293693542, + "learning_rate": 6.346417086924719e-06, + "loss": 2.4652, + "step": 15727 + }, + { + "epoch": 1.2608887374476359, + "grad_norm": 1.7924827337265015, + "learning_rate": 6.345208436247652e-06, + "loss": 3.1035, + "step": 15728 + }, + { + "epoch": 1.2609689122286585, + "grad_norm": 1.808830976486206, + "learning_rate": 6.343999847189079e-06, + "loss": 3.2637, + "step": 15729 + }, + { + "epoch": 1.261049087009681, + "grad_norm": 1.6921583414077759, + "learning_rate": 6.34279131976938e-06, + "loss": 3.2502, + "step": 15730 + }, + { + "epoch": 1.2611292617907037, + "grad_norm": 0.6602006554603577, + "learning_rate": 6.341582854008933e-06, + "loss": 2.4968, + "step": 15731 + }, + { + "epoch": 1.2612094365717263, + "grad_norm": 1.6672015190124512, + "learning_rate": 6.340374449928109e-06, + "loss": 3.2521, + "step": 15732 + }, + { + "epoch": 1.261289611352749, + "grad_norm": 1.782954216003418, + "learning_rate": 6.339166107547281e-06, + "loss": 3.1091, + "step": 15733 + }, + { + "epoch": 1.2613697861337716, + "grad_norm": 1.8330695629119873, + "learning_rate": 6.3379578268868226e-06, + "loss": 3.1857, + "step": 15734 + }, + { + "epoch": 1.2614499609147942, + "grad_norm": 1.796952247619629, + "learning_rate": 6.336749607967103e-06, + "loss": 3.0558, + "step": 15735 + }, + { + "epoch": 1.2615301356958168, + "grad_norm": 1.7602030038833618, + "learning_rate": 6.335541450808498e-06, + "loss": 3.1726, + "step": 15736 + }, + { + "epoch": 1.2616103104768395, + "grad_norm": 1.737717628479004, + "learning_rate": 6.334333355431369e-06, + "loss": 3.1672, + "step": 15737 + }, + { + "epoch": 1.261690485257862, + "grad_norm": 1.7158515453338623, + "learning_rate": 6.3331253218560905e-06, + "loss": 3.1259, + "step": 15738 + }, + { + "epoch": 1.2617706600388847, + "grad_norm": 1.782691478729248, + "learning_rate": 6.331917350103024e-06, + "loss": 3.2848, + "step": 15739 + }, + { + "epoch": 1.2618508348199073, + "grad_norm": 1.8082166910171509, + "learning_rate": 6.330709440192539e-06, + "loss": 3.2655, + "step": 15740 + }, + { + "epoch": 1.26193100960093, + "grad_norm": 1.7455626726150513, + "learning_rate": 6.329501592145001e-06, + "loss": 3.0882, + "step": 15741 + }, + { + "epoch": 1.2620111843819526, + "grad_norm": 1.8610608577728271, + "learning_rate": 6.3282938059807705e-06, + "loss": 3.2968, + "step": 15742 + }, + { + "epoch": 1.2620913591629752, + "grad_norm": 1.8099554777145386, + "learning_rate": 6.327086081720215e-06, + "loss": 3.1063, + "step": 15743 + }, + { + "epoch": 1.2621715339439978, + "grad_norm": 1.694478988647461, + "learning_rate": 6.325878419383691e-06, + "loss": 3.2448, + "step": 15744 + }, + { + "epoch": 1.2622517087250205, + "grad_norm": 1.781238079071045, + "learning_rate": 6.324670818991563e-06, + "loss": 3.1815, + "step": 15745 + }, + { + "epoch": 1.262331883506043, + "grad_norm": 1.771481990814209, + "learning_rate": 6.32346328056419e-06, + "loss": 3.2575, + "step": 15746 + }, + { + "epoch": 1.2624120582870657, + "grad_norm": 1.6524460315704346, + "learning_rate": 6.322255804121931e-06, + "loss": 3.1786, + "step": 15747 + }, + { + "epoch": 1.2624922330680883, + "grad_norm": 1.7751020193099976, + "learning_rate": 6.321048389685142e-06, + "loss": 3.2212, + "step": 15748 + }, + { + "epoch": 1.262572407849111, + "grad_norm": 1.860501766204834, + "learning_rate": 6.319841037274182e-06, + "loss": 3.2704, + "step": 15749 + }, + { + "epoch": 1.2626525826301336, + "grad_norm": 1.7434192895889282, + "learning_rate": 6.318633746909402e-06, + "loss": 3.2308, + "step": 15750 + }, + { + "epoch": 1.2627327574111562, + "grad_norm": 1.7213770151138306, + "learning_rate": 6.317426518611163e-06, + "loss": 3.2234, + "step": 15751 + }, + { + "epoch": 1.2628129321921788, + "grad_norm": 1.838278889656067, + "learning_rate": 6.316219352399813e-06, + "loss": 3.0924, + "step": 15752 + }, + { + "epoch": 1.2628931069732015, + "grad_norm": 1.7394436597824097, + "learning_rate": 6.315012248295708e-06, + "loss": 3.2423, + "step": 15753 + }, + { + "epoch": 1.262973281754224, + "grad_norm": 1.7584666013717651, + "learning_rate": 6.313805206319196e-06, + "loss": 3.2343, + "step": 15754 + }, + { + "epoch": 1.263053456535247, + "grad_norm": 1.698904275894165, + "learning_rate": 6.312598226490629e-06, + "loss": 3.2116, + "step": 15755 + }, + { + "epoch": 1.2631336313162695, + "grad_norm": 1.8268812894821167, + "learning_rate": 6.311391308830358e-06, + "loss": 3.1693, + "step": 15756 + }, + { + "epoch": 1.2632138060972922, + "grad_norm": 1.8098044395446777, + "learning_rate": 6.310184453358727e-06, + "loss": 3.1713, + "step": 15757 + }, + { + "epoch": 1.2632939808783148, + "grad_norm": 1.9234216213226318, + "learning_rate": 6.30897766009609e-06, + "loss": 3.1697, + "step": 15758 + }, + { + "epoch": 1.2633741556593374, + "grad_norm": 1.7500370740890503, + "learning_rate": 6.3077709290627845e-06, + "loss": 3.2101, + "step": 15759 + }, + { + "epoch": 1.26345433044036, + "grad_norm": 1.7194212675094604, + "learning_rate": 6.306564260279161e-06, + "loss": 3.179, + "step": 15760 + }, + { + "epoch": 1.2635345052213827, + "grad_norm": 1.751304030418396, + "learning_rate": 6.305357653765565e-06, + "loss": 3.1167, + "step": 15761 + }, + { + "epoch": 1.2636146800024053, + "grad_norm": 1.7105273008346558, + "learning_rate": 6.304151109542333e-06, + "loss": 3.1439, + "step": 15762 + }, + { + "epoch": 1.263694854783428, + "grad_norm": 1.8622760772705078, + "learning_rate": 6.302944627629814e-06, + "loss": 3.2254, + "step": 15763 + }, + { + "epoch": 1.2637750295644505, + "grad_norm": 1.896178960800171, + "learning_rate": 6.3017382080483426e-06, + "loss": 3.2156, + "step": 15764 + }, + { + "epoch": 1.2638552043454732, + "grad_norm": 1.8009147644042969, + "learning_rate": 6.300531850818264e-06, + "loss": 3.129, + "step": 15765 + }, + { + "epoch": 1.2639353791264958, + "grad_norm": 1.7494691610336304, + "learning_rate": 6.299325555959914e-06, + "loss": 3.1735, + "step": 15766 + }, + { + "epoch": 1.2640155539075184, + "grad_norm": 1.8224847316741943, + "learning_rate": 6.29811932349363e-06, + "loss": 3.2589, + "step": 15767 + }, + { + "epoch": 1.264095728688541, + "grad_norm": 1.7684950828552246, + "learning_rate": 6.296913153439751e-06, + "loss": 3.1905, + "step": 15768 + }, + { + "epoch": 1.2641759034695637, + "grad_norm": 1.8116806745529175, + "learning_rate": 6.29570704581861e-06, + "loss": 3.1832, + "step": 15769 + }, + { + "epoch": 1.2642560782505863, + "grad_norm": 1.8216272592544556, + "learning_rate": 6.2945010006505415e-06, + "loss": 3.1826, + "step": 15770 + }, + { + "epoch": 1.264336253031609, + "grad_norm": 0.6829999685287476, + "learning_rate": 6.293295017955883e-06, + "loss": 2.4709, + "step": 15771 + }, + { + "epoch": 1.2644164278126315, + "grad_norm": 1.7860338687896729, + "learning_rate": 6.292089097754962e-06, + "loss": 3.2837, + "step": 15772 + }, + { + "epoch": 1.2644966025936542, + "grad_norm": 1.879296064376831, + "learning_rate": 6.2908832400681145e-06, + "loss": 3.178, + "step": 15773 + }, + { + "epoch": 1.2645767773746768, + "grad_norm": 1.836796522140503, + "learning_rate": 6.289677444915665e-06, + "loss": 3.1351, + "step": 15774 + }, + { + "epoch": 1.2646569521556994, + "grad_norm": 1.7508906126022339, + "learning_rate": 6.288471712317947e-06, + "loss": 3.1949, + "step": 15775 + }, + { + "epoch": 1.264737126936722, + "grad_norm": 1.8521862030029297, + "learning_rate": 6.28726604229529e-06, + "loss": 3.2412, + "step": 15776 + }, + { + "epoch": 1.2648173017177446, + "grad_norm": 1.73625648021698, + "learning_rate": 6.2860604348680174e-06, + "loss": 3.2119, + "step": 15777 + }, + { + "epoch": 1.2648974764987673, + "grad_norm": 1.8093459606170654, + "learning_rate": 6.2848548900564575e-06, + "loss": 3.1097, + "step": 15778 + }, + { + "epoch": 1.26497765127979, + "grad_norm": 1.8196970224380493, + "learning_rate": 6.283649407880934e-06, + "loss": 3.1814, + "step": 15779 + }, + { + "epoch": 1.2650578260608125, + "grad_norm": 0.7303425669670105, + "learning_rate": 6.282443988361773e-06, + "loss": 2.4836, + "step": 15780 + }, + { + "epoch": 1.2651380008418351, + "grad_norm": 1.894888162612915, + "learning_rate": 6.281238631519295e-06, + "loss": 3.2552, + "step": 15781 + }, + { + "epoch": 1.2652181756228578, + "grad_norm": 1.7681978940963745, + "learning_rate": 6.280033337373823e-06, + "loss": 3.2549, + "step": 15782 + }, + { + "epoch": 1.2652983504038804, + "grad_norm": 1.837379813194275, + "learning_rate": 6.27882810594568e-06, + "loss": 3.1526, + "step": 15783 + }, + { + "epoch": 1.265378525184903, + "grad_norm": 1.8575751781463623, + "learning_rate": 6.277622937255181e-06, + "loss": 3.1821, + "step": 15784 + }, + { + "epoch": 1.2654586999659256, + "grad_norm": 1.824872612953186, + "learning_rate": 6.276417831322647e-06, + "loss": 3.2434, + "step": 15785 + }, + { + "epoch": 1.2655388747469485, + "grad_norm": 1.843770980834961, + "learning_rate": 6.2752127881683985e-06, + "loss": 3.1603, + "step": 15786 + }, + { + "epoch": 1.265619049527971, + "grad_norm": 1.8708443641662598, + "learning_rate": 6.274007807812749e-06, + "loss": 3.034, + "step": 15787 + }, + { + "epoch": 1.2656992243089937, + "grad_norm": 1.8957117795944214, + "learning_rate": 6.2728028902760154e-06, + "loss": 3.2026, + "step": 15788 + }, + { + "epoch": 1.2657793990900164, + "grad_norm": 0.6582483649253845, + "learning_rate": 6.27159803557851e-06, + "loss": 2.4503, + "step": 15789 + }, + { + "epoch": 1.265859573871039, + "grad_norm": 1.943435788154602, + "learning_rate": 6.270393243740547e-06, + "loss": 3.196, + "step": 15790 + }, + { + "epoch": 1.2659397486520616, + "grad_norm": 1.9056850671768188, + "learning_rate": 6.2691885147824404e-06, + "loss": 3.0742, + "step": 15791 + }, + { + "epoch": 1.2660199234330842, + "grad_norm": 1.7650338411331177, + "learning_rate": 6.267983848724502e-06, + "loss": 3.2028, + "step": 15792 + }, + { + "epoch": 1.2661000982141069, + "grad_norm": 1.9442548751831055, + "learning_rate": 6.26677924558704e-06, + "loss": 3.2245, + "step": 15793 + }, + { + "epoch": 1.2661802729951295, + "grad_norm": 1.83533775806427, + "learning_rate": 6.2655747053903626e-06, + "loss": 3.1616, + "step": 15794 + }, + { + "epoch": 1.266260447776152, + "grad_norm": 1.8212252855300903, + "learning_rate": 6.264370228154779e-06, + "loss": 3.1643, + "step": 15795 + }, + { + "epoch": 1.2663406225571747, + "grad_norm": 1.76069974899292, + "learning_rate": 6.2631658139006e-06, + "loss": 3.312, + "step": 15796 + }, + { + "epoch": 1.2664207973381973, + "grad_norm": 1.8770509958267212, + "learning_rate": 6.261961462648125e-06, + "loss": 3.1886, + "step": 15797 + }, + { + "epoch": 1.26650097211922, + "grad_norm": 1.7587138414382935, + "learning_rate": 6.260757174417666e-06, + "loss": 3.2186, + "step": 15798 + }, + { + "epoch": 1.2665811469002426, + "grad_norm": 1.8053346872329712, + "learning_rate": 6.25955294922952e-06, + "loss": 3.2571, + "step": 15799 + }, + { + "epoch": 1.2666613216812652, + "grad_norm": 1.7503900527954102, + "learning_rate": 6.2583487871039935e-06, + "loss": 3.0496, + "step": 15800 + }, + { + "epoch": 1.2667414964622878, + "grad_norm": 1.7676681280136108, + "learning_rate": 6.25714468806139e-06, + "loss": 3.1998, + "step": 15801 + }, + { + "epoch": 1.2668216712433105, + "grad_norm": 1.8360528945922852, + "learning_rate": 6.255940652122006e-06, + "loss": 3.1982, + "step": 15802 + }, + { + "epoch": 1.266901846024333, + "grad_norm": 1.895490050315857, + "learning_rate": 6.254736679306143e-06, + "loss": 3.172, + "step": 15803 + }, + { + "epoch": 1.2669820208053557, + "grad_norm": 1.795093297958374, + "learning_rate": 6.253532769634101e-06, + "loss": 3.2688, + "step": 15804 + }, + { + "epoch": 1.2670621955863783, + "grad_norm": 1.8247814178466797, + "learning_rate": 6.252328923126174e-06, + "loss": 3.1426, + "step": 15805 + }, + { + "epoch": 1.267142370367401, + "grad_norm": 2.088310718536377, + "learning_rate": 6.251125139802663e-06, + "loss": 3.157, + "step": 15806 + }, + { + "epoch": 1.2672225451484236, + "grad_norm": 0.6836390495300293, + "learning_rate": 6.249921419683859e-06, + "loss": 2.4612, + "step": 15807 + }, + { + "epoch": 1.2673027199294462, + "grad_norm": 1.7866368293762207, + "learning_rate": 6.24871776279006e-06, + "loss": 3.1204, + "step": 15808 + }, + { + "epoch": 1.2673828947104688, + "grad_norm": 1.7267677783966064, + "learning_rate": 6.247514169141555e-06, + "loss": 3.2351, + "step": 15809 + }, + { + "epoch": 1.2674630694914915, + "grad_norm": 1.8694562911987305, + "learning_rate": 6.246310638758639e-06, + "loss": 3.2386, + "step": 15810 + }, + { + "epoch": 1.267543244272514, + "grad_norm": 1.767568826675415, + "learning_rate": 6.245107171661605e-06, + "loss": 3.1204, + "step": 15811 + }, + { + "epoch": 1.2676234190535367, + "grad_norm": 1.8940906524658203, + "learning_rate": 6.243903767870738e-06, + "loss": 3.0856, + "step": 15812 + }, + { + "epoch": 1.2677035938345593, + "grad_norm": 1.8250523805618286, + "learning_rate": 6.242700427406333e-06, + "loss": 3.241, + "step": 15813 + }, + { + "epoch": 1.267783768615582, + "grad_norm": 1.8923068046569824, + "learning_rate": 6.2414971502886726e-06, + "loss": 3.1992, + "step": 15814 + }, + { + "epoch": 1.2678639433966046, + "grad_norm": 1.9272748231887817, + "learning_rate": 6.240293936538044e-06, + "loss": 3.2433, + "step": 15815 + }, + { + "epoch": 1.2679441181776272, + "grad_norm": 1.9078834056854248, + "learning_rate": 6.239090786174737e-06, + "loss": 3.1452, + "step": 15816 + }, + { + "epoch": 1.2680242929586498, + "grad_norm": 1.8163622617721558, + "learning_rate": 6.237887699219033e-06, + "loss": 3.2224, + "step": 15817 + }, + { + "epoch": 1.2681044677396724, + "grad_norm": 1.7532750368118286, + "learning_rate": 6.236684675691218e-06, + "loss": 3.1872, + "step": 15818 + }, + { + "epoch": 1.268184642520695, + "grad_norm": 1.7214076519012451, + "learning_rate": 6.235481715611571e-06, + "loss": 3.1851, + "step": 15819 + }, + { + "epoch": 1.2682648173017177, + "grad_norm": 1.9043184518814087, + "learning_rate": 6.234278819000375e-06, + "loss": 3.163, + "step": 15820 + }, + { + "epoch": 1.2683449920827403, + "grad_norm": 1.770204782485962, + "learning_rate": 6.233075985877914e-06, + "loss": 3.1624, + "step": 15821 + }, + { + "epoch": 1.268425166863763, + "grad_norm": 0.6656551361083984, + "learning_rate": 6.231873216264461e-06, + "loss": 2.4806, + "step": 15822 + }, + { + "epoch": 1.2685053416447856, + "grad_norm": 1.7949351072311401, + "learning_rate": 6.230670510180301e-06, + "loss": 3.1963, + "step": 15823 + }, + { + "epoch": 1.2685855164258082, + "grad_norm": 1.88198983669281, + "learning_rate": 6.229467867645704e-06, + "loss": 3.3643, + "step": 15824 + }, + { + "epoch": 1.2686656912068308, + "grad_norm": 1.7724621295928955, + "learning_rate": 6.228265288680951e-06, + "loss": 3.1752, + "step": 15825 + }, + { + "epoch": 1.2687458659878534, + "grad_norm": 1.7809579372406006, + "learning_rate": 6.227062773306318e-06, + "loss": 3.2114, + "step": 15826 + }, + { + "epoch": 1.268826040768876, + "grad_norm": 1.7993911504745483, + "learning_rate": 6.225860321542074e-06, + "loss": 3.169, + "step": 15827 + }, + { + "epoch": 1.2689062155498987, + "grad_norm": 1.8581198453903198, + "learning_rate": 6.224657933408497e-06, + "loss": 3.3251, + "step": 15828 + }, + { + "epoch": 1.2689863903309213, + "grad_norm": 1.8033944368362427, + "learning_rate": 6.223455608925857e-06, + "loss": 3.2836, + "step": 15829 + }, + { + "epoch": 1.269066565111944, + "grad_norm": 1.7935601472854614, + "learning_rate": 6.222253348114422e-06, + "loss": 3.2878, + "step": 15830 + }, + { + "epoch": 1.2691467398929666, + "grad_norm": 1.8633586168289185, + "learning_rate": 6.221051150994467e-06, + "loss": 3.1879, + "step": 15831 + }, + { + "epoch": 1.2692269146739892, + "grad_norm": 1.8918758630752563, + "learning_rate": 6.219849017586256e-06, + "loss": 3.234, + "step": 15832 + }, + { + "epoch": 1.2693070894550118, + "grad_norm": 1.7950152158737183, + "learning_rate": 6.21864694791006e-06, + "loss": 3.1348, + "step": 15833 + }, + { + "epoch": 1.2693872642360344, + "grad_norm": 1.7979432344436646, + "learning_rate": 6.217444941986141e-06, + "loss": 3.2086, + "step": 15834 + }, + { + "epoch": 1.269467439017057, + "grad_norm": 1.8590266704559326, + "learning_rate": 6.2162429998347684e-06, + "loss": 3.2091, + "step": 15835 + }, + { + "epoch": 1.2695476137980797, + "grad_norm": 1.8273584842681885, + "learning_rate": 6.215041121476206e-06, + "loss": 3.2314, + "step": 15836 + }, + { + "epoch": 1.2696277885791025, + "grad_norm": 1.769720435142517, + "learning_rate": 6.213839306930715e-06, + "loss": 3.0913, + "step": 15837 + }, + { + "epoch": 1.2697079633601251, + "grad_norm": 1.7700965404510498, + "learning_rate": 6.2126375562185615e-06, + "loss": 3.1831, + "step": 15838 + }, + { + "epoch": 1.2697881381411478, + "grad_norm": 1.7254432439804077, + "learning_rate": 6.211435869360002e-06, + "loss": 3.1917, + "step": 15839 + }, + { + "epoch": 1.2698683129221704, + "grad_norm": 1.9165534973144531, + "learning_rate": 6.210234246375298e-06, + "loss": 3.3616, + "step": 15840 + }, + { + "epoch": 1.269948487703193, + "grad_norm": 1.8282550573349, + "learning_rate": 6.2090326872847105e-06, + "loss": 3.2441, + "step": 15841 + }, + { + "epoch": 1.2700286624842156, + "grad_norm": 1.6490070819854736, + "learning_rate": 6.207831192108495e-06, + "loss": 3.0898, + "step": 15842 + }, + { + "epoch": 1.2701088372652383, + "grad_norm": 1.7671396732330322, + "learning_rate": 6.20662976086691e-06, + "loss": 3.2186, + "step": 15843 + }, + { + "epoch": 1.270189012046261, + "grad_norm": 1.8742057085037231, + "learning_rate": 6.205428393580208e-06, + "loss": 3.1323, + "step": 15844 + }, + { + "epoch": 1.2702691868272835, + "grad_norm": 1.792110562324524, + "learning_rate": 6.204227090268649e-06, + "loss": 3.0497, + "step": 15845 + }, + { + "epoch": 1.2703493616083061, + "grad_norm": 1.8077880144119263, + "learning_rate": 6.203025850952482e-06, + "loss": 3.2026, + "step": 15846 + }, + { + "epoch": 1.2704295363893288, + "grad_norm": 1.788104772567749, + "learning_rate": 6.20182467565196e-06, + "loss": 3.2574, + "step": 15847 + }, + { + "epoch": 1.2705097111703514, + "grad_norm": 1.8096214532852173, + "learning_rate": 6.200623564387338e-06, + "loss": 3.156, + "step": 15848 + }, + { + "epoch": 1.270589885951374, + "grad_norm": 1.7506362199783325, + "learning_rate": 6.199422517178862e-06, + "loss": 3.148, + "step": 15849 + }, + { + "epoch": 1.2706700607323966, + "grad_norm": 1.7526907920837402, + "learning_rate": 6.198221534046781e-06, + "loss": 3.2251, + "step": 15850 + }, + { + "epoch": 1.2707502355134193, + "grad_norm": 1.8472270965576172, + "learning_rate": 6.197020615011349e-06, + "loss": 3.1144, + "step": 15851 + }, + { + "epoch": 1.2708304102944419, + "grad_norm": 1.8327776193618774, + "learning_rate": 6.195819760092805e-06, + "loss": 3.0515, + "step": 15852 + }, + { + "epoch": 1.2709105850754645, + "grad_norm": 0.6934331059455872, + "learning_rate": 6.194618969311404e-06, + "loss": 2.4935, + "step": 15853 + }, + { + "epoch": 1.2709907598564871, + "grad_norm": 1.785004734992981, + "learning_rate": 6.193418242687381e-06, + "loss": 3.1829, + "step": 15854 + }, + { + "epoch": 1.2710709346375098, + "grad_norm": 1.8558809757232666, + "learning_rate": 6.192217580240986e-06, + "loss": 3.1659, + "step": 15855 + }, + { + "epoch": 1.2711511094185324, + "grad_norm": 1.7591744661331177, + "learning_rate": 6.191016981992464e-06, + "loss": 3.0806, + "step": 15856 + }, + { + "epoch": 1.271231284199555, + "grad_norm": 1.783322811126709, + "learning_rate": 6.1898164479620495e-06, + "loss": 3.2113, + "step": 15857 + }, + { + "epoch": 1.2713114589805776, + "grad_norm": 1.8831192255020142, + "learning_rate": 6.18861597816999e-06, + "loss": 3.3058, + "step": 15858 + }, + { + "epoch": 1.2713916337616002, + "grad_norm": 1.6729549169540405, + "learning_rate": 6.187415572636519e-06, + "loss": 3.0667, + "step": 15859 + }, + { + "epoch": 1.2714718085426229, + "grad_norm": 1.7687718868255615, + "learning_rate": 6.186215231381878e-06, + "loss": 3.2325, + "step": 15860 + }, + { + "epoch": 1.2715519833236455, + "grad_norm": 1.6408061981201172, + "learning_rate": 6.185014954426307e-06, + "loss": 3.1036, + "step": 15861 + }, + { + "epoch": 1.2716321581046681, + "grad_norm": 1.763139009475708, + "learning_rate": 6.183814741790036e-06, + "loss": 3.1177, + "step": 15862 + }, + { + "epoch": 1.2717123328856907, + "grad_norm": 1.7840253114700317, + "learning_rate": 6.182614593493307e-06, + "loss": 3.1024, + "step": 15863 + }, + { + "epoch": 1.2717925076667134, + "grad_norm": 1.731757640838623, + "learning_rate": 6.181414509556347e-06, + "loss": 3.223, + "step": 15864 + }, + { + "epoch": 1.271872682447736, + "grad_norm": 1.694939374923706, + "learning_rate": 6.180214489999393e-06, + "loss": 3.1779, + "step": 15865 + }, + { + "epoch": 1.2719528572287586, + "grad_norm": 1.760825276374817, + "learning_rate": 6.17901453484268e-06, + "loss": 3.2056, + "step": 15866 + }, + { + "epoch": 1.2720330320097815, + "grad_norm": 1.903800129890442, + "learning_rate": 6.177814644106431e-06, + "loss": 3.3348, + "step": 15867 + }, + { + "epoch": 1.272113206790804, + "grad_norm": 1.8438384532928467, + "learning_rate": 6.176614817810881e-06, + "loss": 3.2948, + "step": 15868 + }, + { + "epoch": 1.2721933815718267, + "grad_norm": 1.7596290111541748, + "learning_rate": 6.175415055976259e-06, + "loss": 3.1672, + "step": 15869 + }, + { + "epoch": 1.2722735563528493, + "grad_norm": 1.666330099105835, + "learning_rate": 6.17421535862279e-06, + "loss": 3.2141, + "step": 15870 + }, + { + "epoch": 1.272353731133872, + "grad_norm": 1.7696954011917114, + "learning_rate": 6.1730157257707045e-06, + "loss": 3.0951, + "step": 15871 + }, + { + "epoch": 1.2724339059148946, + "grad_norm": 1.856682538986206, + "learning_rate": 6.1718161574402225e-06, + "loss": 3.13, + "step": 15872 + }, + { + "epoch": 1.2725140806959172, + "grad_norm": 1.6338762044906616, + "learning_rate": 6.170616653651571e-06, + "loss": 3.1902, + "step": 15873 + }, + { + "epoch": 1.2725942554769398, + "grad_norm": 1.7567176818847656, + "learning_rate": 6.169417214424974e-06, + "loss": 3.1187, + "step": 15874 + }, + { + "epoch": 1.2726744302579625, + "grad_norm": 1.827436089515686, + "learning_rate": 6.168217839780652e-06, + "loss": 3.3723, + "step": 15875 + }, + { + "epoch": 1.272754605038985, + "grad_norm": 1.777644395828247, + "learning_rate": 6.167018529738829e-06, + "loss": 3.1916, + "step": 15876 + }, + { + "epoch": 1.2728347798200077, + "grad_norm": 1.7271684408187866, + "learning_rate": 6.165819284319718e-06, + "loss": 3.1005, + "step": 15877 + }, + { + "epoch": 1.2729149546010303, + "grad_norm": 2.1029539108276367, + "learning_rate": 6.164620103543546e-06, + "loss": 3.2219, + "step": 15878 + }, + { + "epoch": 1.272995129382053, + "grad_norm": 1.8341883420944214, + "learning_rate": 6.163420987430528e-06, + "loss": 3.3159, + "step": 15879 + }, + { + "epoch": 1.2730753041630756, + "grad_norm": 1.7753643989562988, + "learning_rate": 6.162221936000878e-06, + "loss": 3.1423, + "step": 15880 + }, + { + "epoch": 1.2731554789440982, + "grad_norm": 1.7675377130508423, + "learning_rate": 6.161022949274815e-06, + "loss": 3.2175, + "step": 15881 + }, + { + "epoch": 1.2732356537251208, + "grad_norm": 1.7688171863555908, + "learning_rate": 6.1598240272725505e-06, + "loss": 3.1857, + "step": 15882 + }, + { + "epoch": 1.2733158285061434, + "grad_norm": 1.6369982957839966, + "learning_rate": 6.1586251700143e-06, + "loss": 3.1559, + "step": 15883 + }, + { + "epoch": 1.273396003287166, + "grad_norm": 1.7839686870574951, + "learning_rate": 6.157426377520277e-06, + "loss": 3.1798, + "step": 15884 + }, + { + "epoch": 1.2734761780681887, + "grad_norm": 1.8247241973876953, + "learning_rate": 6.156227649810689e-06, + "loss": 3.1515, + "step": 15885 + }, + { + "epoch": 1.2735563528492113, + "grad_norm": 1.6792677640914917, + "learning_rate": 6.15502898690575e-06, + "loss": 3.1568, + "step": 15886 + }, + { + "epoch": 1.273636527630234, + "grad_norm": 1.7255983352661133, + "learning_rate": 6.153830388825666e-06, + "loss": 3.1455, + "step": 15887 + }, + { + "epoch": 1.2737167024112566, + "grad_norm": 1.7778434753417969, + "learning_rate": 6.152631855590645e-06, + "loss": 3.1778, + "step": 15888 + }, + { + "epoch": 1.2737968771922792, + "grad_norm": 0.684469997882843, + "learning_rate": 6.151433387220898e-06, + "loss": 2.4662, + "step": 15889 + }, + { + "epoch": 1.2738770519733018, + "grad_norm": 1.8021984100341797, + "learning_rate": 6.1502349837366234e-06, + "loss": 3.3108, + "step": 15890 + }, + { + "epoch": 1.2739572267543244, + "grad_norm": 1.7928227186203003, + "learning_rate": 6.149036645158034e-06, + "loss": 3.2482, + "step": 15891 + }, + { + "epoch": 1.274037401535347, + "grad_norm": 1.7592113018035889, + "learning_rate": 6.147838371505328e-06, + "loss": 3.1155, + "step": 15892 + }, + { + "epoch": 1.2741175763163697, + "grad_norm": 1.7542122602462769, + "learning_rate": 6.146640162798708e-06, + "loss": 3.1312, + "step": 15893 + }, + { + "epoch": 1.2741977510973923, + "grad_norm": 1.7990846633911133, + "learning_rate": 6.145442019058381e-06, + "loss": 3.0896, + "step": 15894 + }, + { + "epoch": 1.274277925878415, + "grad_norm": 1.6738190650939941, + "learning_rate": 6.14424394030454e-06, + "loss": 3.1478, + "step": 15895 + }, + { + "epoch": 1.2743581006594376, + "grad_norm": 1.765352725982666, + "learning_rate": 6.143045926557387e-06, + "loss": 3.0538, + "step": 15896 + }, + { + "epoch": 1.2744382754404602, + "grad_norm": 1.767293930053711, + "learning_rate": 6.141847977837121e-06, + "loss": 3.1293, + "step": 15897 + }, + { + "epoch": 1.2745184502214828, + "grad_norm": 1.7192296981811523, + "learning_rate": 6.140650094163938e-06, + "loss": 3.1353, + "step": 15898 + }, + { + "epoch": 1.2745986250025054, + "grad_norm": 1.831358790397644, + "learning_rate": 6.1394522755580355e-06, + "loss": 3.2638, + "step": 15899 + }, + { + "epoch": 1.274678799783528, + "grad_norm": 1.7863308191299438, + "learning_rate": 6.1382545220396066e-06, + "loss": 3.1687, + "step": 15900 + }, + { + "epoch": 1.2747589745645507, + "grad_norm": 1.762252926826477, + "learning_rate": 6.137056833628845e-06, + "loss": 3.2095, + "step": 15901 + }, + { + "epoch": 1.2748391493455733, + "grad_norm": 1.763641357421875, + "learning_rate": 6.1358592103459435e-06, + "loss": 3.2759, + "step": 15902 + }, + { + "epoch": 1.274919324126596, + "grad_norm": 1.643919587135315, + "learning_rate": 6.1346616522110934e-06, + "loss": 3.1124, + "step": 15903 + }, + { + "epoch": 1.2749994989076185, + "grad_norm": 1.7731610536575317, + "learning_rate": 6.1334641592444875e-06, + "loss": 3.1658, + "step": 15904 + }, + { + "epoch": 1.2750796736886412, + "grad_norm": 1.7240134477615356, + "learning_rate": 6.132266731466312e-06, + "loss": 3.1979, + "step": 15905 + }, + { + "epoch": 1.2751598484696638, + "grad_norm": 1.7309445142745972, + "learning_rate": 6.1310693688967575e-06, + "loss": 3.115, + "step": 15906 + }, + { + "epoch": 1.2752400232506864, + "grad_norm": 1.8416292667388916, + "learning_rate": 6.129872071556006e-06, + "loss": 3.1394, + "step": 15907 + }, + { + "epoch": 1.275320198031709, + "grad_norm": 1.7502646446228027, + "learning_rate": 6.1286748394642505e-06, + "loss": 3.1023, + "step": 15908 + }, + { + "epoch": 1.2754003728127317, + "grad_norm": 1.8119149208068848, + "learning_rate": 6.127477672641674e-06, + "loss": 3.3085, + "step": 15909 + }, + { + "epoch": 1.2754805475937543, + "grad_norm": 1.7456564903259277, + "learning_rate": 6.126280571108456e-06, + "loss": 3.1731, + "step": 15910 + }, + { + "epoch": 1.275560722374777, + "grad_norm": 1.8127001523971558, + "learning_rate": 6.125083534884786e-06, + "loss": 3.2494, + "step": 15911 + }, + { + "epoch": 1.2756408971557995, + "grad_norm": 1.8932347297668457, + "learning_rate": 6.123886563990839e-06, + "loss": 3.2457, + "step": 15912 + }, + { + "epoch": 1.2757210719368222, + "grad_norm": 1.7397198677062988, + "learning_rate": 6.122689658446799e-06, + "loss": 3.2304, + "step": 15913 + }, + { + "epoch": 1.2758012467178448, + "grad_norm": 1.8043031692504883, + "learning_rate": 6.121492818272846e-06, + "loss": 3.3436, + "step": 15914 + }, + { + "epoch": 1.2758814214988674, + "grad_norm": 1.7712374925613403, + "learning_rate": 6.120296043489155e-06, + "loss": 3.2487, + "step": 15915 + }, + { + "epoch": 1.27596159627989, + "grad_norm": 1.7344996929168701, + "learning_rate": 6.119099334115909e-06, + "loss": 3.1771, + "step": 15916 + }, + { + "epoch": 1.2760417710609127, + "grad_norm": 1.8278902769088745, + "learning_rate": 6.117902690173276e-06, + "loss": 3.1699, + "step": 15917 + }, + { + "epoch": 1.2761219458419355, + "grad_norm": 1.7174750566482544, + "learning_rate": 6.116706111681439e-06, + "loss": 3.2045, + "step": 15918 + }, + { + "epoch": 1.2762021206229581, + "grad_norm": 1.786382794380188, + "learning_rate": 6.1155095986605676e-06, + "loss": 3.2307, + "step": 15919 + }, + { + "epoch": 1.2762822954039807, + "grad_norm": 1.7039061784744263, + "learning_rate": 6.114313151130833e-06, + "loss": 3.158, + "step": 15920 + }, + { + "epoch": 1.2763624701850034, + "grad_norm": 1.8182785511016846, + "learning_rate": 6.113116769112412e-06, + "loss": 3.0289, + "step": 15921 + }, + { + "epoch": 1.276442644966026, + "grad_norm": 1.7477171421051025, + "learning_rate": 6.111920452625472e-06, + "loss": 3.2063, + "step": 15922 + }, + { + "epoch": 1.2765228197470486, + "grad_norm": 0.6605470180511475, + "learning_rate": 6.110724201690182e-06, + "loss": 2.4636, + "step": 15923 + }, + { + "epoch": 1.2766029945280712, + "grad_norm": 1.8517372608184814, + "learning_rate": 6.109528016326711e-06, + "loss": 3.1193, + "step": 15924 + }, + { + "epoch": 1.2766831693090939, + "grad_norm": 1.739073395729065, + "learning_rate": 6.108331896555226e-06, + "loss": 3.2185, + "step": 15925 + }, + { + "epoch": 1.2767633440901165, + "grad_norm": 1.8032342195510864, + "learning_rate": 6.107135842395897e-06, + "loss": 3.2358, + "step": 15926 + }, + { + "epoch": 1.2768435188711391, + "grad_norm": 1.7704639434814453, + "learning_rate": 6.105939853868883e-06, + "loss": 3.154, + "step": 15927 + }, + { + "epoch": 1.2769236936521617, + "grad_norm": 1.7806475162506104, + "learning_rate": 6.10474393099435e-06, + "loss": 3.0791, + "step": 15928 + }, + { + "epoch": 1.2770038684331844, + "grad_norm": 1.734777808189392, + "learning_rate": 6.103548073792463e-06, + "loss": 3.1819, + "step": 15929 + }, + { + "epoch": 1.277084043214207, + "grad_norm": 1.8469582796096802, + "learning_rate": 6.102352282283381e-06, + "loss": 3.2256, + "step": 15930 + }, + { + "epoch": 1.2771642179952296, + "grad_norm": 1.7588920593261719, + "learning_rate": 6.101156556487269e-06, + "loss": 3.0891, + "step": 15931 + }, + { + "epoch": 1.2772443927762522, + "grad_norm": 1.8477718830108643, + "learning_rate": 6.099960896424279e-06, + "loss": 3.1215, + "step": 15932 + }, + { + "epoch": 1.2773245675572749, + "grad_norm": 1.7638304233551025, + "learning_rate": 6.098765302114575e-06, + "loss": 3.1564, + "step": 15933 + }, + { + "epoch": 1.2774047423382975, + "grad_norm": 2.007498264312744, + "learning_rate": 6.097569773578315e-06, + "loss": 3.1818, + "step": 15934 + }, + { + "epoch": 1.27748491711932, + "grad_norm": 1.8096458911895752, + "learning_rate": 6.096374310835652e-06, + "loss": 3.2175, + "step": 15935 + }, + { + "epoch": 1.2775650919003427, + "grad_norm": 1.8103924989700317, + "learning_rate": 6.095178913906743e-06, + "loss": 3.0838, + "step": 15936 + }, + { + "epoch": 1.2776452666813654, + "grad_norm": 1.6526206731796265, + "learning_rate": 6.09398358281174e-06, + "loss": 3.0944, + "step": 15937 + }, + { + "epoch": 1.277725441462388, + "grad_norm": 1.7052525281906128, + "learning_rate": 6.092788317570798e-06, + "loss": 3.1519, + "step": 15938 + }, + { + "epoch": 1.2778056162434106, + "grad_norm": 1.7348341941833496, + "learning_rate": 6.091593118204069e-06, + "loss": 3.2542, + "step": 15939 + }, + { + "epoch": 1.2778857910244332, + "grad_norm": 1.8877601623535156, + "learning_rate": 6.0903979847317006e-06, + "loss": 3.1741, + "step": 15940 + }, + { + "epoch": 1.2779659658054559, + "grad_norm": 1.7986526489257812, + "learning_rate": 6.089202917173846e-06, + "loss": 3.2152, + "step": 15941 + }, + { + "epoch": 1.2780461405864785, + "grad_norm": 1.7556746006011963, + "learning_rate": 6.088007915550651e-06, + "loss": 3.2417, + "step": 15942 + }, + { + "epoch": 1.278126315367501, + "grad_norm": 1.7608776092529297, + "learning_rate": 6.086812979882264e-06, + "loss": 3.1549, + "step": 15943 + }, + { + "epoch": 1.2782064901485237, + "grad_norm": 1.8253997564315796, + "learning_rate": 6.085618110188831e-06, + "loss": 3.1838, + "step": 15944 + }, + { + "epoch": 1.2782866649295463, + "grad_norm": 1.8363220691680908, + "learning_rate": 6.084423306490497e-06, + "loss": 3.2523, + "step": 15945 + }, + { + "epoch": 1.278366839710569, + "grad_norm": 1.8076808452606201, + "learning_rate": 6.083228568807407e-06, + "loss": 3.3113, + "step": 15946 + }, + { + "epoch": 1.2784470144915916, + "grad_norm": 1.7230702638626099, + "learning_rate": 6.082033897159701e-06, + "loss": 3.212, + "step": 15947 + }, + { + "epoch": 1.2785271892726142, + "grad_norm": 1.7641451358795166, + "learning_rate": 6.080839291567523e-06, + "loss": 3.2697, + "step": 15948 + }, + { + "epoch": 1.278607364053637, + "grad_norm": 1.8473237752914429, + "learning_rate": 6.079644752051015e-06, + "loss": 3.1813, + "step": 15949 + }, + { + "epoch": 1.2786875388346597, + "grad_norm": 1.8018076419830322, + "learning_rate": 6.078450278630313e-06, + "loss": 3.0985, + "step": 15950 + }, + { + "epoch": 1.2787677136156823, + "grad_norm": 1.7573928833007812, + "learning_rate": 6.0772558713255604e-06, + "loss": 3.2767, + "step": 15951 + }, + { + "epoch": 1.278847888396705, + "grad_norm": 1.7213455438613892, + "learning_rate": 6.076061530156887e-06, + "loss": 3.1708, + "step": 15952 + }, + { + "epoch": 1.2789280631777276, + "grad_norm": 1.721649169921875, + "learning_rate": 6.074867255144434e-06, + "loss": 3.207, + "step": 15953 + }, + { + "epoch": 1.2790082379587502, + "grad_norm": 0.6776793599128723, + "learning_rate": 6.073673046308339e-06, + "loss": 2.455, + "step": 15954 + }, + { + "epoch": 1.2790884127397728, + "grad_norm": 1.8593624830245972, + "learning_rate": 6.07247890366873e-06, + "loss": 3.1127, + "step": 15955 + }, + { + "epoch": 1.2791685875207954, + "grad_norm": 1.7435317039489746, + "learning_rate": 6.0712848272457445e-06, + "loss": 3.2075, + "step": 15956 + }, + { + "epoch": 1.279248762301818, + "grad_norm": 0.6556874513626099, + "learning_rate": 6.07009081705951e-06, + "loss": 2.4766, + "step": 15957 + }, + { + "epoch": 1.2793289370828407, + "grad_norm": 1.7793601751327515, + "learning_rate": 6.068896873130159e-06, + "loss": 3.1282, + "step": 15958 + }, + { + "epoch": 1.2794091118638633, + "grad_norm": 1.7805982828140259, + "learning_rate": 6.067702995477824e-06, + "loss": 3.088, + "step": 15959 + }, + { + "epoch": 1.279489286644886, + "grad_norm": 1.796532154083252, + "learning_rate": 6.066509184122629e-06, + "loss": 3.0638, + "step": 15960 + }, + { + "epoch": 1.2795694614259085, + "grad_norm": 1.716636061668396, + "learning_rate": 6.065315439084702e-06, + "loss": 3.1632, + "step": 15961 + }, + { + "epoch": 1.2796496362069312, + "grad_norm": 1.7963330745697021, + "learning_rate": 6.064121760384172e-06, + "loss": 3.2067, + "step": 15962 + }, + { + "epoch": 1.2797298109879538, + "grad_norm": 1.7775726318359375, + "learning_rate": 6.062928148041158e-06, + "loss": 3.1289, + "step": 15963 + }, + { + "epoch": 1.2798099857689764, + "grad_norm": 1.7470694780349731, + "learning_rate": 6.061734602075792e-06, + "loss": 3.08, + "step": 15964 + }, + { + "epoch": 1.279890160549999, + "grad_norm": 1.8736798763275146, + "learning_rate": 6.0605411225081915e-06, + "loss": 3.2059, + "step": 15965 + }, + { + "epoch": 1.2799703353310217, + "grad_norm": 1.7993420362472534, + "learning_rate": 6.05934770935848e-06, + "loss": 3.2158, + "step": 15966 + }, + { + "epoch": 1.2800505101120443, + "grad_norm": 1.9025726318359375, + "learning_rate": 6.058154362646775e-06, + "loss": 3.1756, + "step": 15967 + }, + { + "epoch": 1.280130684893067, + "grad_norm": 1.8069525957107544, + "learning_rate": 6.0569610823932e-06, + "loss": 3.1824, + "step": 15968 + }, + { + "epoch": 1.2802108596740895, + "grad_norm": 1.7432862520217896, + "learning_rate": 6.055767868617872e-06, + "loss": 3.1597, + "step": 15969 + }, + { + "epoch": 1.2802910344551122, + "grad_norm": 1.7851781845092773, + "learning_rate": 6.054574721340906e-06, + "loss": 3.1735, + "step": 15970 + }, + { + "epoch": 1.2803712092361348, + "grad_norm": 1.783968210220337, + "learning_rate": 6.0533816405824205e-06, + "loss": 3.0734, + "step": 15971 + }, + { + "epoch": 1.2804513840171574, + "grad_norm": 1.6432329416275024, + "learning_rate": 6.0521886263625296e-06, + "loss": 3.1826, + "step": 15972 + }, + { + "epoch": 1.28053155879818, + "grad_norm": 1.6860153675079346, + "learning_rate": 6.050995678701346e-06, + "loss": 3.2687, + "step": 15973 + }, + { + "epoch": 1.2806117335792027, + "grad_norm": 1.7550677061080933, + "learning_rate": 6.049802797618986e-06, + "loss": 3.2018, + "step": 15974 + }, + { + "epoch": 1.2806919083602253, + "grad_norm": 1.8017762899398804, + "learning_rate": 6.048609983135556e-06, + "loss": 3.2288, + "step": 15975 + }, + { + "epoch": 1.280772083141248, + "grad_norm": 1.8799717426300049, + "learning_rate": 6.0474172352711715e-06, + "loss": 3.1267, + "step": 15976 + }, + { + "epoch": 1.2808522579222705, + "grad_norm": 1.8461512327194214, + "learning_rate": 6.0462245540459386e-06, + "loss": 3.1433, + "step": 15977 + }, + { + "epoch": 1.2809324327032932, + "grad_norm": 1.8856202363967896, + "learning_rate": 6.045031939479964e-06, + "loss": 3.1962, + "step": 15978 + }, + { + "epoch": 1.2810126074843158, + "grad_norm": 1.795889139175415, + "learning_rate": 6.04383939159336e-06, + "loss": 3.2809, + "step": 15979 + }, + { + "epoch": 1.2810927822653384, + "grad_norm": 1.8002406358718872, + "learning_rate": 6.042646910406228e-06, + "loss": 3.1952, + "step": 15980 + }, + { + "epoch": 1.281172957046361, + "grad_norm": 1.7966911792755127, + "learning_rate": 6.041454495938677e-06, + "loss": 3.1791, + "step": 15981 + }, + { + "epoch": 1.2812531318273837, + "grad_norm": 2.0246663093566895, + "learning_rate": 6.040262148210804e-06, + "loss": 3.1094, + "step": 15982 + }, + { + "epoch": 1.2813333066084063, + "grad_norm": 1.7088615894317627, + "learning_rate": 6.039069867242717e-06, + "loss": 3.1129, + "step": 15983 + }, + { + "epoch": 1.281413481389429, + "grad_norm": 1.7345579862594604, + "learning_rate": 6.037877653054518e-06, + "loss": 3.1935, + "step": 15984 + }, + { + "epoch": 1.2814936561704515, + "grad_norm": 1.7037752866744995, + "learning_rate": 6.036685505666302e-06, + "loss": 3.2228, + "step": 15985 + }, + { + "epoch": 1.2815738309514741, + "grad_norm": 1.91664457321167, + "learning_rate": 6.035493425098176e-06, + "loss": 3.219, + "step": 15986 + }, + { + "epoch": 1.2816540057324968, + "grad_norm": 1.7624897956848145, + "learning_rate": 6.034301411370229e-06, + "loss": 3.1444, + "step": 15987 + }, + { + "epoch": 1.2817341805135194, + "grad_norm": 1.8219783306121826, + "learning_rate": 6.0331094645025645e-06, + "loss": 3.1825, + "step": 15988 + }, + { + "epoch": 1.281814355294542, + "grad_norm": 1.7744685411453247, + "learning_rate": 6.031917584515279e-06, + "loss": 3.0304, + "step": 15989 + }, + { + "epoch": 1.2818945300755646, + "grad_norm": 1.7449144124984741, + "learning_rate": 6.030725771428461e-06, + "loss": 3.136, + "step": 15990 + }, + { + "epoch": 1.2819747048565873, + "grad_norm": 1.6973501443862915, + "learning_rate": 6.029534025262209e-06, + "loss": 3.1879, + "step": 15991 + }, + { + "epoch": 1.28205487963761, + "grad_norm": 1.718113899230957, + "learning_rate": 6.0283423460366135e-06, + "loss": 3.1429, + "step": 15992 + }, + { + "epoch": 1.2821350544186325, + "grad_norm": 1.977509617805481, + "learning_rate": 6.027150733771765e-06, + "loss": 3.2316, + "step": 15993 + }, + { + "epoch": 1.2822152291996551, + "grad_norm": 1.833250880241394, + "learning_rate": 6.025959188487757e-06, + "loss": 3.283, + "step": 15994 + }, + { + "epoch": 1.2822954039806778, + "grad_norm": 1.6853482723236084, + "learning_rate": 6.0247677102046765e-06, + "loss": 3.2009, + "step": 15995 + }, + { + "epoch": 1.2823755787617004, + "grad_norm": 1.7758241891860962, + "learning_rate": 6.023576298942613e-06, + "loss": 3.2192, + "step": 15996 + }, + { + "epoch": 1.282455753542723, + "grad_norm": 1.7589166164398193, + "learning_rate": 6.022384954721647e-06, + "loss": 3.1021, + "step": 15997 + }, + { + "epoch": 1.2825359283237456, + "grad_norm": 1.6866323947906494, + "learning_rate": 6.0211936775618716e-06, + "loss": 3.2176, + "step": 15998 + }, + { + "epoch": 1.2826161031047683, + "grad_norm": 1.8843952417373657, + "learning_rate": 6.02000246748337e-06, + "loss": 3.1963, + "step": 15999 + }, + { + "epoch": 1.282696277885791, + "grad_norm": 1.7176010608673096, + "learning_rate": 6.018811324506221e-06, + "loss": 3.1382, + "step": 16000 + }, + { + "epoch": 1.2827764526668137, + "grad_norm": 1.7852615118026733, + "learning_rate": 6.017620248650513e-06, + "loss": 3.1586, + "step": 16001 + }, + { + "epoch": 1.2828566274478364, + "grad_norm": 1.8207370042800903, + "learning_rate": 6.016429239936321e-06, + "loss": 3.1962, + "step": 16002 + }, + { + "epoch": 1.282936802228859, + "grad_norm": 1.7208281755447388, + "learning_rate": 6.015238298383729e-06, + "loss": 3.1223, + "step": 16003 + }, + { + "epoch": 1.2830169770098816, + "grad_norm": 1.8889625072479248, + "learning_rate": 6.014047424012817e-06, + "loss": 3.0737, + "step": 16004 + }, + { + "epoch": 1.2830971517909042, + "grad_norm": 1.7696304321289062, + "learning_rate": 6.012856616843657e-06, + "loss": 3.1896, + "step": 16005 + }, + { + "epoch": 1.2831773265719268, + "grad_norm": 1.8692947626113892, + "learning_rate": 6.0116658768963335e-06, + "loss": 3.3568, + "step": 16006 + }, + { + "epoch": 1.2832575013529495, + "grad_norm": 0.6628378033638, + "learning_rate": 6.010475204190915e-06, + "loss": 2.4618, + "step": 16007 + }, + { + "epoch": 1.283337676133972, + "grad_norm": 1.7778290510177612, + "learning_rate": 6.009284598747478e-06, + "loss": 3.1129, + "step": 16008 + }, + { + "epoch": 1.2834178509149947, + "grad_norm": 1.8481603860855103, + "learning_rate": 6.008094060586098e-06, + "loss": 3.1099, + "step": 16009 + }, + { + "epoch": 1.2834980256960173, + "grad_norm": 1.8230377435684204, + "learning_rate": 6.006903589726844e-06, + "loss": 3.2812, + "step": 16010 + }, + { + "epoch": 1.28357820047704, + "grad_norm": 1.744807481765747, + "learning_rate": 6.00571318618979e-06, + "loss": 3.0964, + "step": 16011 + }, + { + "epoch": 1.2836583752580626, + "grad_norm": 1.7701466083526611, + "learning_rate": 6.004522849995001e-06, + "loss": 3.1335, + "step": 16012 + }, + { + "epoch": 1.2837385500390852, + "grad_norm": 1.7620023488998413, + "learning_rate": 6.00333258116255e-06, + "loss": 3.1992, + "step": 16013 + }, + { + "epoch": 1.2838187248201078, + "grad_norm": 1.8118304014205933, + "learning_rate": 6.002142379712505e-06, + "loss": 3.1613, + "step": 16014 + }, + { + "epoch": 1.2838988996011305, + "grad_norm": 2.0823135375976562, + "learning_rate": 6.000952245664929e-06, + "loss": 3.1682, + "step": 16015 + }, + { + "epoch": 1.283979074382153, + "grad_norm": 1.8535794019699097, + "learning_rate": 5.999762179039888e-06, + "loss": 3.2268, + "step": 16016 + }, + { + "epoch": 1.2840592491631757, + "grad_norm": 1.6501489877700806, + "learning_rate": 5.998572179857447e-06, + "loss": 3.1317, + "step": 16017 + }, + { + "epoch": 1.2841394239441983, + "grad_norm": 1.8070333003997803, + "learning_rate": 5.997382248137669e-06, + "loss": 3.204, + "step": 16018 + }, + { + "epoch": 1.284219598725221, + "grad_norm": 1.7017253637313843, + "learning_rate": 5.9961923839006165e-06, + "loss": 3.1501, + "step": 16019 + }, + { + "epoch": 1.2842997735062436, + "grad_norm": 1.8424842357635498, + "learning_rate": 5.9950025871663475e-06, + "loss": 3.1617, + "step": 16020 + }, + { + "epoch": 1.2843799482872662, + "grad_norm": 0.6978600025177002, + "learning_rate": 5.993812857954926e-06, + "loss": 2.4738, + "step": 16021 + }, + { + "epoch": 1.2844601230682888, + "grad_norm": 1.9522842168807983, + "learning_rate": 5.9926231962864054e-06, + "loss": 3.1913, + "step": 16022 + }, + { + "epoch": 1.2845402978493115, + "grad_norm": 1.8565269708633423, + "learning_rate": 5.991433602180845e-06, + "loss": 3.1687, + "step": 16023 + }, + { + "epoch": 1.284620472630334, + "grad_norm": 1.8258373737335205, + "learning_rate": 5.9902440756583044e-06, + "loss": 3.2037, + "step": 16024 + }, + { + "epoch": 1.2847006474113567, + "grad_norm": 1.7723920345306396, + "learning_rate": 5.989054616738833e-06, + "loss": 3.1155, + "step": 16025 + }, + { + "epoch": 1.2847808221923793, + "grad_norm": 1.824004888534546, + "learning_rate": 5.9878652254424886e-06, + "loss": 3.2209, + "step": 16026 + }, + { + "epoch": 1.284860996973402, + "grad_norm": 1.8310267925262451, + "learning_rate": 5.9866759017893216e-06, + "loss": 3.1056, + "step": 16027 + }, + { + "epoch": 1.2849411717544246, + "grad_norm": 1.7543420791625977, + "learning_rate": 5.985486645799383e-06, + "loss": 3.0759, + "step": 16028 + }, + { + "epoch": 1.2850213465354472, + "grad_norm": 0.6899828314781189, + "learning_rate": 5.984297457492729e-06, + "loss": 2.462, + "step": 16029 + }, + { + "epoch": 1.28510152131647, + "grad_norm": 1.7159442901611328, + "learning_rate": 5.983108336889401e-06, + "loss": 3.1333, + "step": 16030 + }, + { + "epoch": 1.2851816960974927, + "grad_norm": 1.8197784423828125, + "learning_rate": 5.981919284009452e-06, + "loss": 3.1523, + "step": 16031 + }, + { + "epoch": 1.2852618708785153, + "grad_norm": 1.8478642702102661, + "learning_rate": 5.980730298872927e-06, + "loss": 3.2506, + "step": 16032 + }, + { + "epoch": 1.285342045659538, + "grad_norm": 1.790662169456482, + "learning_rate": 5.9795413814998724e-06, + "loss": 3.1924, + "step": 16033 + }, + { + "epoch": 1.2854222204405605, + "grad_norm": 1.8002426624298096, + "learning_rate": 5.978352531910335e-06, + "loss": 3.272, + "step": 16034 + }, + { + "epoch": 1.2855023952215832, + "grad_norm": 1.6986926794052124, + "learning_rate": 5.977163750124355e-06, + "loss": 3.2266, + "step": 16035 + }, + { + "epoch": 1.2855825700026058, + "grad_norm": 1.887243390083313, + "learning_rate": 5.975975036161979e-06, + "loss": 3.2327, + "step": 16036 + }, + { + "epoch": 1.2856627447836284, + "grad_norm": 1.762291669845581, + "learning_rate": 5.974786390043241e-06, + "loss": 3.2552, + "step": 16037 + }, + { + "epoch": 1.285742919564651, + "grad_norm": 1.7383779287338257, + "learning_rate": 5.9735978117881875e-06, + "loss": 3.0856, + "step": 16038 + }, + { + "epoch": 1.2858230943456737, + "grad_norm": 0.6796582937240601, + "learning_rate": 5.972409301416859e-06, + "loss": 2.4681, + "step": 16039 + }, + { + "epoch": 1.2859032691266963, + "grad_norm": 0.711889922618866, + "learning_rate": 5.971220858949287e-06, + "loss": 2.4912, + "step": 16040 + }, + { + "epoch": 1.285983443907719, + "grad_norm": 1.854989767074585, + "learning_rate": 5.970032484405512e-06, + "loss": 3.2139, + "step": 16041 + }, + { + "epoch": 1.2860636186887415, + "grad_norm": 1.742271900177002, + "learning_rate": 5.968844177805571e-06, + "loss": 3.2052, + "step": 16042 + }, + { + "epoch": 1.2861437934697642, + "grad_norm": 1.7739920616149902, + "learning_rate": 5.967655939169492e-06, + "loss": 3.1951, + "step": 16043 + }, + { + "epoch": 1.2862239682507868, + "grad_norm": 1.8143892288208008, + "learning_rate": 5.966467768517316e-06, + "loss": 3.2864, + "step": 16044 + }, + { + "epoch": 1.2863041430318094, + "grad_norm": 1.800706148147583, + "learning_rate": 5.9652796658690724e-06, + "loss": 3.1262, + "step": 16045 + }, + { + "epoch": 1.286384317812832, + "grad_norm": 1.8474012613296509, + "learning_rate": 5.964091631244792e-06, + "loss": 3.1365, + "step": 16046 + }, + { + "epoch": 1.2864644925938546, + "grad_norm": 1.9859325885772705, + "learning_rate": 5.9629036646645014e-06, + "loss": 3.3318, + "step": 16047 + }, + { + "epoch": 1.2865446673748773, + "grad_norm": 1.7844080924987793, + "learning_rate": 5.961715766148233e-06, + "loss": 3.179, + "step": 16048 + }, + { + "epoch": 1.2866248421559, + "grad_norm": 1.724473237991333, + "learning_rate": 5.960527935716017e-06, + "loss": 3.1769, + "step": 16049 + }, + { + "epoch": 1.2867050169369225, + "grad_norm": 1.7383146286010742, + "learning_rate": 5.959340173387872e-06, + "loss": 3.1506, + "step": 16050 + }, + { + "epoch": 1.2867851917179451, + "grad_norm": 1.8803684711456299, + "learning_rate": 5.95815247918383e-06, + "loss": 3.1602, + "step": 16051 + }, + { + "epoch": 1.2868653664989678, + "grad_norm": 1.752547025680542, + "learning_rate": 5.956964853123912e-06, + "loss": 3.2604, + "step": 16052 + }, + { + "epoch": 1.2869455412799904, + "grad_norm": 1.9467754364013672, + "learning_rate": 5.95577729522814e-06, + "loss": 3.267, + "step": 16053 + }, + { + "epoch": 1.287025716061013, + "grad_norm": 1.8354952335357666, + "learning_rate": 5.954589805516541e-06, + "loss": 3.1464, + "step": 16054 + }, + { + "epoch": 1.2871058908420356, + "grad_norm": 1.6986814737319946, + "learning_rate": 5.953402384009129e-06, + "loss": 3.1829, + "step": 16055 + }, + { + "epoch": 1.2871860656230583, + "grad_norm": 1.8260358572006226, + "learning_rate": 5.952215030725929e-06, + "loss": 3.2058, + "step": 16056 + }, + { + "epoch": 1.2872662404040809, + "grad_norm": 1.76815664768219, + "learning_rate": 5.951027745686954e-06, + "loss": 3.1964, + "step": 16057 + }, + { + "epoch": 1.2873464151851035, + "grad_norm": 1.7981793880462646, + "learning_rate": 5.949840528912227e-06, + "loss": 3.1339, + "step": 16058 + }, + { + "epoch": 1.2874265899661261, + "grad_norm": 1.8458881378173828, + "learning_rate": 5.948653380421762e-06, + "loss": 3.1862, + "step": 16059 + }, + { + "epoch": 1.2875067647471488, + "grad_norm": 1.9256805181503296, + "learning_rate": 5.94746630023557e-06, + "loss": 3.2282, + "step": 16060 + }, + { + "epoch": 1.2875869395281714, + "grad_norm": 1.7086070775985718, + "learning_rate": 5.946279288373669e-06, + "loss": 3.2117, + "step": 16061 + }, + { + "epoch": 1.287667114309194, + "grad_norm": 1.6793344020843506, + "learning_rate": 5.945092344856072e-06, + "loss": 3.0939, + "step": 16062 + }, + { + "epoch": 1.2877472890902166, + "grad_norm": 0.7091732025146484, + "learning_rate": 5.943905469702788e-06, + "loss": 2.4595, + "step": 16063 + }, + { + "epoch": 1.2878274638712393, + "grad_norm": 1.7618016004562378, + "learning_rate": 5.9427186629338284e-06, + "loss": 3.1862, + "step": 16064 + }, + { + "epoch": 1.2879076386522619, + "grad_norm": 1.7218437194824219, + "learning_rate": 5.941531924569202e-06, + "loss": 3.0952, + "step": 16065 + }, + { + "epoch": 1.2879878134332845, + "grad_norm": 1.6684579849243164, + "learning_rate": 5.940345254628917e-06, + "loss": 3.15, + "step": 16066 + }, + { + "epoch": 1.2880679882143071, + "grad_norm": 1.7545527219772339, + "learning_rate": 5.939158653132981e-06, + "loss": 3.1881, + "step": 16067 + }, + { + "epoch": 1.2881481629953297, + "grad_norm": 0.7031002640724182, + "learning_rate": 5.937972120101398e-06, + "loss": 2.4883, + "step": 16068 + }, + { + "epoch": 1.2882283377763524, + "grad_norm": 1.6701478958129883, + "learning_rate": 5.936785655554174e-06, + "loss": 3.2228, + "step": 16069 + }, + { + "epoch": 1.288308512557375, + "grad_norm": 1.7941571474075317, + "learning_rate": 5.935599259511311e-06, + "loss": 3.2235, + "step": 16070 + }, + { + "epoch": 1.2883886873383976, + "grad_norm": 1.8721567392349243, + "learning_rate": 5.934412931992812e-06, + "loss": 3.122, + "step": 16071 + }, + { + "epoch": 1.2884688621194202, + "grad_norm": 1.8172630071640015, + "learning_rate": 5.9332266730186795e-06, + "loss": 3.1556, + "step": 16072 + }, + { + "epoch": 1.2885490369004429, + "grad_norm": 1.6073215007781982, + "learning_rate": 5.932040482608911e-06, + "loss": 3.1786, + "step": 16073 + }, + { + "epoch": 1.2886292116814655, + "grad_norm": 1.7935280799865723, + "learning_rate": 5.930854360783508e-06, + "loss": 3.1769, + "step": 16074 + }, + { + "epoch": 1.2887093864624881, + "grad_norm": 1.7725850343704224, + "learning_rate": 5.929668307562464e-06, + "loss": 3.1064, + "step": 16075 + }, + { + "epoch": 1.2887895612435107, + "grad_norm": 1.696683645248413, + "learning_rate": 5.928482322965779e-06, + "loss": 3.2263, + "step": 16076 + }, + { + "epoch": 1.2888697360245334, + "grad_norm": 1.837411880493164, + "learning_rate": 5.927296407013448e-06, + "loss": 3.2996, + "step": 16077 + }, + { + "epoch": 1.288949910805556, + "grad_norm": 1.7828123569488525, + "learning_rate": 5.926110559725461e-06, + "loss": 3.1761, + "step": 16078 + }, + { + "epoch": 1.2890300855865786, + "grad_norm": 1.7517317533493042, + "learning_rate": 5.924924781121817e-06, + "loss": 3.1918, + "step": 16079 + }, + { + "epoch": 1.2891102603676012, + "grad_norm": 1.751993179321289, + "learning_rate": 5.923739071222504e-06, + "loss": 3.1868, + "step": 16080 + }, + { + "epoch": 1.289190435148624, + "grad_norm": 1.6564067602157593, + "learning_rate": 5.922553430047513e-06, + "loss": 3.1253, + "step": 16081 + }, + { + "epoch": 1.2892706099296467, + "grad_norm": 1.9584447145462036, + "learning_rate": 5.921367857616835e-06, + "loss": 3.2252, + "step": 16082 + }, + { + "epoch": 1.2893507847106693, + "grad_norm": 1.790414810180664, + "learning_rate": 5.920182353950458e-06, + "loss": 3.1664, + "step": 16083 + }, + { + "epoch": 1.289430959491692, + "grad_norm": 0.6780251264572144, + "learning_rate": 5.918996919068369e-06, + "loss": 2.4924, + "step": 16084 + }, + { + "epoch": 1.2895111342727146, + "grad_norm": 1.8267736434936523, + "learning_rate": 5.917811552990552e-06, + "loss": 3.1525, + "step": 16085 + }, + { + "epoch": 1.2895913090537372, + "grad_norm": 1.8158248662948608, + "learning_rate": 5.9166262557369935e-06, + "loss": 3.3078, + "step": 16086 + }, + { + "epoch": 1.2896714838347598, + "grad_norm": 1.8492828607559204, + "learning_rate": 5.91544102732768e-06, + "loss": 3.2008, + "step": 16087 + }, + { + "epoch": 1.2897516586157824, + "grad_norm": 1.7626498937606812, + "learning_rate": 5.914255867782588e-06, + "loss": 3.2283, + "step": 16088 + }, + { + "epoch": 1.289831833396805, + "grad_norm": 1.7191685438156128, + "learning_rate": 5.913070777121705e-06, + "loss": 3.0755, + "step": 16089 + }, + { + "epoch": 1.2899120081778277, + "grad_norm": 1.8552724123001099, + "learning_rate": 5.911885755365005e-06, + "loss": 3.1708, + "step": 16090 + }, + { + "epoch": 1.2899921829588503, + "grad_norm": 1.8956118822097778, + "learning_rate": 5.910700802532472e-06, + "loss": 3.0438, + "step": 16091 + }, + { + "epoch": 1.290072357739873, + "grad_norm": 1.8782994747161865, + "learning_rate": 5.909515918644083e-06, + "loss": 3.1191, + "step": 16092 + }, + { + "epoch": 1.2901525325208956, + "grad_norm": 1.855285882949829, + "learning_rate": 5.908331103719813e-06, + "loss": 3.2032, + "step": 16093 + }, + { + "epoch": 1.2902327073019182, + "grad_norm": 1.8760780096054077, + "learning_rate": 5.90714635777964e-06, + "loss": 3.1416, + "step": 16094 + }, + { + "epoch": 1.2903128820829408, + "grad_norm": 1.7287297248840332, + "learning_rate": 5.905961680843535e-06, + "loss": 3.0994, + "step": 16095 + }, + { + "epoch": 1.2903930568639634, + "grad_norm": 1.7933295965194702, + "learning_rate": 5.904777072931473e-06, + "loss": 3.1215, + "step": 16096 + }, + { + "epoch": 1.290473231644986, + "grad_norm": 1.744179368019104, + "learning_rate": 5.903592534063428e-06, + "loss": 3.2322, + "step": 16097 + }, + { + "epoch": 1.2905534064260087, + "grad_norm": 1.80337393283844, + "learning_rate": 5.902408064259368e-06, + "loss": 3.14, + "step": 16098 + }, + { + "epoch": 1.2906335812070313, + "grad_norm": 1.8771311044692993, + "learning_rate": 5.901223663539266e-06, + "loss": 3.1715, + "step": 16099 + }, + { + "epoch": 1.290713755988054, + "grad_norm": 1.7503528594970703, + "learning_rate": 5.900039331923085e-06, + "loss": 3.1792, + "step": 16100 + }, + { + "epoch": 1.2907939307690766, + "grad_norm": 1.8305381536483765, + "learning_rate": 5.8988550694307975e-06, + "loss": 3.1013, + "step": 16101 + }, + { + "epoch": 1.2908741055500992, + "grad_norm": 1.7836910486221313, + "learning_rate": 5.897670876082369e-06, + "loss": 3.1375, + "step": 16102 + }, + { + "epoch": 1.2909542803311218, + "grad_norm": 1.6875903606414795, + "learning_rate": 5.896486751897763e-06, + "loss": 3.1237, + "step": 16103 + }, + { + "epoch": 1.2910344551121444, + "grad_norm": 1.717793583869934, + "learning_rate": 5.895302696896945e-06, + "loss": 3.1625, + "step": 16104 + }, + { + "epoch": 1.291114629893167, + "grad_norm": 1.7502343654632568, + "learning_rate": 5.894118711099875e-06, + "loss": 3.1244, + "step": 16105 + }, + { + "epoch": 1.2911948046741897, + "grad_norm": 1.8562268018722534, + "learning_rate": 5.892934794526517e-06, + "loss": 3.2076, + "step": 16106 + }, + { + "epoch": 1.2912749794552123, + "grad_norm": 1.713365912437439, + "learning_rate": 5.891750947196833e-06, + "loss": 3.2096, + "step": 16107 + }, + { + "epoch": 1.291355154236235, + "grad_norm": 0.7018409371376038, + "learning_rate": 5.8905671691307774e-06, + "loss": 2.4706, + "step": 16108 + }, + { + "epoch": 1.2914353290172575, + "grad_norm": 1.7567296028137207, + "learning_rate": 5.889383460348314e-06, + "loss": 3.1838, + "step": 16109 + }, + { + "epoch": 1.2915155037982802, + "grad_norm": 1.7874013185501099, + "learning_rate": 5.888199820869393e-06, + "loss": 3.1838, + "step": 16110 + }, + { + "epoch": 1.2915956785793028, + "grad_norm": 1.8090543746948242, + "learning_rate": 5.887016250713976e-06, + "loss": 3.1472, + "step": 16111 + }, + { + "epoch": 1.2916758533603256, + "grad_norm": 2.1655948162078857, + "learning_rate": 5.885832749902016e-06, + "loss": 3.2031, + "step": 16112 + }, + { + "epoch": 1.2917560281413483, + "grad_norm": 1.8962087631225586, + "learning_rate": 5.884649318453465e-06, + "loss": 3.2056, + "step": 16113 + }, + { + "epoch": 1.291836202922371, + "grad_norm": 1.7828009128570557, + "learning_rate": 5.883465956388277e-06, + "loss": 3.2949, + "step": 16114 + }, + { + "epoch": 1.2919163777033935, + "grad_norm": 1.8777679204940796, + "learning_rate": 5.882282663726401e-06, + "loss": 3.2228, + "step": 16115 + }, + { + "epoch": 1.2919965524844161, + "grad_norm": 0.6544712781906128, + "learning_rate": 5.881099440487788e-06, + "loss": 2.4552, + "step": 16116 + }, + { + "epoch": 1.2920767272654388, + "grad_norm": 1.6714378595352173, + "learning_rate": 5.879916286692389e-06, + "loss": 3.1692, + "step": 16117 + }, + { + "epoch": 1.2921569020464614, + "grad_norm": 1.8115547895431519, + "learning_rate": 5.878733202360146e-06, + "loss": 3.1793, + "step": 16118 + }, + { + "epoch": 1.292237076827484, + "grad_norm": 1.8046518564224243, + "learning_rate": 5.877550187511012e-06, + "loss": 3.1467, + "step": 16119 + }, + { + "epoch": 1.2923172516085066, + "grad_norm": 1.871241569519043, + "learning_rate": 5.876367242164926e-06, + "loss": 3.2664, + "step": 16120 + }, + { + "epoch": 1.2923974263895293, + "grad_norm": 1.7903411388397217, + "learning_rate": 5.875184366341835e-06, + "loss": 3.2706, + "step": 16121 + }, + { + "epoch": 1.2924776011705519, + "grad_norm": 1.6452256441116333, + "learning_rate": 5.874001560061685e-06, + "loss": 3.1614, + "step": 16122 + }, + { + "epoch": 1.2925577759515745, + "grad_norm": 1.83003568649292, + "learning_rate": 5.872818823344411e-06, + "loss": 3.1803, + "step": 16123 + }, + { + "epoch": 1.2926379507325971, + "grad_norm": 1.8191118240356445, + "learning_rate": 5.871636156209961e-06, + "loss": 3.1606, + "step": 16124 + }, + { + "epoch": 1.2927181255136198, + "grad_norm": 2.0021018981933594, + "learning_rate": 5.870453558678267e-06, + "loss": 3.1344, + "step": 16125 + }, + { + "epoch": 1.2927983002946424, + "grad_norm": 1.7356226444244385, + "learning_rate": 5.869271030769271e-06, + "loss": 3.2974, + "step": 16126 + }, + { + "epoch": 1.292878475075665, + "grad_norm": 0.7169474959373474, + "learning_rate": 5.8680885725029125e-06, + "loss": 2.454, + "step": 16127 + }, + { + "epoch": 1.2929586498566876, + "grad_norm": 1.7922923564910889, + "learning_rate": 5.866906183899122e-06, + "loss": 3.1495, + "step": 16128 + }, + { + "epoch": 1.2930388246377102, + "grad_norm": 1.811221718788147, + "learning_rate": 5.865723864977837e-06, + "loss": 3.1979, + "step": 16129 + }, + { + "epoch": 1.2931189994187329, + "grad_norm": 0.6748844981193542, + "learning_rate": 5.86454161575899e-06, + "loss": 2.4696, + "step": 16130 + }, + { + "epoch": 1.2931991741997555, + "grad_norm": 1.7164298295974731, + "learning_rate": 5.8633594362625144e-06, + "loss": 3.2606, + "step": 16131 + }, + { + "epoch": 1.2932793489807781, + "grad_norm": 1.8416509628295898, + "learning_rate": 5.862177326508341e-06, + "loss": 3.2045, + "step": 16132 + }, + { + "epoch": 1.2933595237618007, + "grad_norm": 1.7873181104660034, + "learning_rate": 5.8609952865164e-06, + "loss": 3.2108, + "step": 16133 + }, + { + "epoch": 1.2934396985428234, + "grad_norm": 1.7888901233673096, + "learning_rate": 5.85981331630662e-06, + "loss": 3.2847, + "step": 16134 + }, + { + "epoch": 1.293519873323846, + "grad_norm": 1.8615041971206665, + "learning_rate": 5.858631415898926e-06, + "loss": 3.2323, + "step": 16135 + }, + { + "epoch": 1.2936000481048686, + "grad_norm": 1.8272325992584229, + "learning_rate": 5.857449585313249e-06, + "loss": 3.1837, + "step": 16136 + }, + { + "epoch": 1.2936802228858912, + "grad_norm": 1.735384464263916, + "learning_rate": 5.8562678245695106e-06, + "loss": 3.3383, + "step": 16137 + }, + { + "epoch": 1.2937603976669139, + "grad_norm": 1.8263503313064575, + "learning_rate": 5.855086133687637e-06, + "loss": 3.1331, + "step": 16138 + }, + { + "epoch": 1.2938405724479365, + "grad_norm": 1.9260828495025635, + "learning_rate": 5.853904512687551e-06, + "loss": 3.2208, + "step": 16139 + }, + { + "epoch": 1.2939207472289591, + "grad_norm": 1.8136444091796875, + "learning_rate": 5.852722961589172e-06, + "loss": 3.2812, + "step": 16140 + }, + { + "epoch": 1.2940009220099817, + "grad_norm": 1.8097635507583618, + "learning_rate": 5.851541480412424e-06, + "loss": 3.2426, + "step": 16141 + }, + { + "epoch": 1.2940810967910044, + "grad_norm": 1.7442008256912231, + "learning_rate": 5.850360069177224e-06, + "loss": 3.0934, + "step": 16142 + }, + { + "epoch": 1.294161271572027, + "grad_norm": 1.6988898515701294, + "learning_rate": 5.8491787279034905e-06, + "loss": 3.1405, + "step": 16143 + }, + { + "epoch": 1.2942414463530496, + "grad_norm": 1.8110665082931519, + "learning_rate": 5.847997456611141e-06, + "loss": 3.1792, + "step": 16144 + }, + { + "epoch": 1.2943216211340722, + "grad_norm": 1.8812811374664307, + "learning_rate": 5.846816255320091e-06, + "loss": 3.254, + "step": 16145 + }, + { + "epoch": 1.2944017959150949, + "grad_norm": 1.6828114986419678, + "learning_rate": 5.845635124050254e-06, + "loss": 3.1548, + "step": 16146 + }, + { + "epoch": 1.2944819706961175, + "grad_norm": 1.8648205995559692, + "learning_rate": 5.8444540628215476e-06, + "loss": 3.2394, + "step": 16147 + }, + { + "epoch": 1.29456214547714, + "grad_norm": 1.812935471534729, + "learning_rate": 5.843273071653879e-06, + "loss": 3.1588, + "step": 16148 + }, + { + "epoch": 1.2946423202581627, + "grad_norm": 1.7675532102584839, + "learning_rate": 5.842092150567162e-06, + "loss": 3.1114, + "step": 16149 + }, + { + "epoch": 1.2947224950391854, + "grad_norm": 1.6493604183197021, + "learning_rate": 5.840911299581304e-06, + "loss": 3.0789, + "step": 16150 + }, + { + "epoch": 1.294802669820208, + "grad_norm": 1.7671377658843994, + "learning_rate": 5.839730518716218e-06, + "loss": 3.2203, + "step": 16151 + }, + { + "epoch": 1.2948828446012306, + "grad_norm": 1.6619969606399536, + "learning_rate": 5.838549807991809e-06, + "loss": 3.2301, + "step": 16152 + }, + { + "epoch": 1.2949630193822532, + "grad_norm": 1.7982538938522339, + "learning_rate": 5.837369167427983e-06, + "loss": 3.215, + "step": 16153 + }, + { + "epoch": 1.2950431941632758, + "grad_norm": 1.844049096107483, + "learning_rate": 5.8361885970446456e-06, + "loss": 3.2313, + "step": 16154 + }, + { + "epoch": 1.2951233689442985, + "grad_norm": 1.9555606842041016, + "learning_rate": 5.835008096861705e-06, + "loss": 3.261, + "step": 16155 + }, + { + "epoch": 1.295203543725321, + "grad_norm": 1.7867361307144165, + "learning_rate": 5.8338276668990545e-06, + "loss": 3.194, + "step": 16156 + }, + { + "epoch": 1.2952837185063437, + "grad_norm": 1.7627232074737549, + "learning_rate": 5.832647307176602e-06, + "loss": 3.166, + "step": 16157 + }, + { + "epoch": 1.2953638932873663, + "grad_norm": 1.765693187713623, + "learning_rate": 5.831467017714249e-06, + "loss": 3.1353, + "step": 16158 + }, + { + "epoch": 1.295444068068389, + "grad_norm": 1.840993881225586, + "learning_rate": 5.8302867985318914e-06, + "loss": 3.1888, + "step": 16159 + }, + { + "epoch": 1.2955242428494116, + "grad_norm": 1.8208075761795044, + "learning_rate": 5.829106649649428e-06, + "loss": 3.2036, + "step": 16160 + }, + { + "epoch": 1.2956044176304342, + "grad_norm": 1.8455225229263306, + "learning_rate": 5.827926571086761e-06, + "loss": 3.0784, + "step": 16161 + }, + { + "epoch": 1.295684592411457, + "grad_norm": 1.7832558155059814, + "learning_rate": 5.8267465628637785e-06, + "loss": 3.103, + "step": 16162 + }, + { + "epoch": 1.2957647671924797, + "grad_norm": 1.7635687589645386, + "learning_rate": 5.825566625000379e-06, + "loss": 3.0688, + "step": 16163 + }, + { + "epoch": 1.2958449419735023, + "grad_norm": 1.7502763271331787, + "learning_rate": 5.824386757516454e-06, + "loss": 3.1117, + "step": 16164 + }, + { + "epoch": 1.295925116754525, + "grad_norm": 1.7721052169799805, + "learning_rate": 5.823206960431896e-06, + "loss": 3.2112, + "step": 16165 + }, + { + "epoch": 1.2960052915355476, + "grad_norm": 1.760055422782898, + "learning_rate": 5.822027233766601e-06, + "loss": 3.1537, + "step": 16166 + }, + { + "epoch": 1.2960854663165702, + "grad_norm": 1.7812702655792236, + "learning_rate": 5.82084757754045e-06, + "loss": 3.2216, + "step": 16167 + }, + { + "epoch": 1.2961656410975928, + "grad_norm": 1.7572617530822754, + "learning_rate": 5.819667991773337e-06, + "loss": 3.1653, + "step": 16168 + }, + { + "epoch": 1.2962458158786154, + "grad_norm": 1.7447491884231567, + "learning_rate": 5.818488476485147e-06, + "loss": 3.088, + "step": 16169 + }, + { + "epoch": 1.296325990659638, + "grad_norm": 1.8089526891708374, + "learning_rate": 5.817309031695768e-06, + "loss": 3.1599, + "step": 16170 + }, + { + "epoch": 1.2964061654406607, + "grad_norm": 1.7120240926742554, + "learning_rate": 5.816129657425089e-06, + "loss": 3.1274, + "step": 16171 + }, + { + "epoch": 1.2964863402216833, + "grad_norm": 1.8593165874481201, + "learning_rate": 5.814950353692986e-06, + "loss": 3.2798, + "step": 16172 + }, + { + "epoch": 1.296566515002706, + "grad_norm": 1.8736495971679688, + "learning_rate": 5.813771120519343e-06, + "loss": 3.1949, + "step": 16173 + }, + { + "epoch": 1.2966466897837285, + "grad_norm": 1.7927746772766113, + "learning_rate": 5.812591957924046e-06, + "loss": 3.1511, + "step": 16174 + }, + { + "epoch": 1.2967268645647512, + "grad_norm": 0.6844316124916077, + "learning_rate": 5.811412865926971e-06, + "loss": 2.4783, + "step": 16175 + }, + { + "epoch": 1.2968070393457738, + "grad_norm": 1.8182028532028198, + "learning_rate": 5.8102338445480035e-06, + "loss": 3.2256, + "step": 16176 + }, + { + "epoch": 1.2968872141267964, + "grad_norm": 1.7352266311645508, + "learning_rate": 5.809054893807014e-06, + "loss": 3.2061, + "step": 16177 + }, + { + "epoch": 1.296967388907819, + "grad_norm": 1.7898660898208618, + "learning_rate": 5.8078760137238795e-06, + "loss": 3.2309, + "step": 16178 + }, + { + "epoch": 1.2970475636888417, + "grad_norm": 1.8883222341537476, + "learning_rate": 5.8066972043184796e-06, + "loss": 3.127, + "step": 16179 + }, + { + "epoch": 1.2971277384698643, + "grad_norm": 1.7100154161453247, + "learning_rate": 5.805518465610686e-06, + "loss": 3.1547, + "step": 16180 + }, + { + "epoch": 1.297207913250887, + "grad_norm": 1.699418067932129, + "learning_rate": 5.804339797620379e-06, + "loss": 3.1444, + "step": 16181 + }, + { + "epoch": 1.2972880880319095, + "grad_norm": 1.8523972034454346, + "learning_rate": 5.803161200367418e-06, + "loss": 3.2758, + "step": 16182 + }, + { + "epoch": 1.2973682628129322, + "grad_norm": 0.7346298694610596, + "learning_rate": 5.8019826738716814e-06, + "loss": 2.4619, + "step": 16183 + }, + { + "epoch": 1.2974484375939548, + "grad_norm": 1.8495994806289673, + "learning_rate": 5.800804218153036e-06, + "loss": 3.2282, + "step": 16184 + }, + { + "epoch": 1.2975286123749774, + "grad_norm": 1.7734043598175049, + "learning_rate": 5.799625833231353e-06, + "loss": 3.1519, + "step": 16185 + }, + { + "epoch": 1.297608787156, + "grad_norm": 1.8379195928573608, + "learning_rate": 5.798447519126501e-06, + "loss": 3.3125, + "step": 16186 + }, + { + "epoch": 1.2976889619370227, + "grad_norm": 1.8173779249191284, + "learning_rate": 5.797269275858339e-06, + "loss": 3.1671, + "step": 16187 + }, + { + "epoch": 1.2977691367180453, + "grad_norm": 1.816608190536499, + "learning_rate": 5.796091103446737e-06, + "loss": 3.163, + "step": 16188 + }, + { + "epoch": 1.297849311499068, + "grad_norm": 1.8860015869140625, + "learning_rate": 5.794913001911558e-06, + "loss": 3.2781, + "step": 16189 + }, + { + "epoch": 1.2979294862800905, + "grad_norm": 1.8427790403366089, + "learning_rate": 5.7937349712726625e-06, + "loss": 3.2106, + "step": 16190 + }, + { + "epoch": 1.2980096610611132, + "grad_norm": 1.697521448135376, + "learning_rate": 5.792557011549915e-06, + "loss": 3.1577, + "step": 16191 + }, + { + "epoch": 1.2980898358421358, + "grad_norm": 1.8243221044540405, + "learning_rate": 5.791379122763172e-06, + "loss": 3.2109, + "step": 16192 + }, + { + "epoch": 1.2981700106231586, + "grad_norm": 1.8334590196609497, + "learning_rate": 5.7902013049322925e-06, + "loss": 3.2188, + "step": 16193 + }, + { + "epoch": 1.2982501854041812, + "grad_norm": 1.719855546951294, + "learning_rate": 5.789023558077136e-06, + "loss": 3.1443, + "step": 16194 + }, + { + "epoch": 1.2983303601852039, + "grad_norm": 1.7344938516616821, + "learning_rate": 5.787845882217559e-06, + "loss": 3.1447, + "step": 16195 + }, + { + "epoch": 1.2984105349662265, + "grad_norm": 1.8650308847427368, + "learning_rate": 5.78666827737342e-06, + "loss": 3.1149, + "step": 16196 + }, + { + "epoch": 1.2984907097472491, + "grad_norm": 1.7354594469070435, + "learning_rate": 5.7854907435645635e-06, + "loss": 3.0459, + "step": 16197 + }, + { + "epoch": 1.2985708845282717, + "grad_norm": 1.8903627395629883, + "learning_rate": 5.784313280810847e-06, + "loss": 3.1768, + "step": 16198 + }, + { + "epoch": 1.2986510593092944, + "grad_norm": 1.7188490629196167, + "learning_rate": 5.783135889132125e-06, + "loss": 3.2152, + "step": 16199 + }, + { + "epoch": 1.298731234090317, + "grad_norm": 1.8006091117858887, + "learning_rate": 5.781958568548245e-06, + "loss": 3.1759, + "step": 16200 + }, + { + "epoch": 1.2988114088713396, + "grad_norm": 1.7187381982803345, + "learning_rate": 5.780781319079061e-06, + "loss": 3.2205, + "step": 16201 + }, + { + "epoch": 1.2988915836523622, + "grad_norm": 1.8611971139907837, + "learning_rate": 5.779604140744412e-06, + "loss": 3.1625, + "step": 16202 + }, + { + "epoch": 1.2989717584333849, + "grad_norm": 0.6698269844055176, + "learning_rate": 5.778427033564152e-06, + "loss": 2.4922, + "step": 16203 + }, + { + "epoch": 1.2990519332144075, + "grad_norm": 1.7758190631866455, + "learning_rate": 5.7772499975581205e-06, + "loss": 3.2013, + "step": 16204 + }, + { + "epoch": 1.29913210799543, + "grad_norm": 1.8240656852722168, + "learning_rate": 5.776073032746169e-06, + "loss": 3.1757, + "step": 16205 + }, + { + "epoch": 1.2992122827764527, + "grad_norm": 1.6860451698303223, + "learning_rate": 5.774896139148141e-06, + "loss": 3.2012, + "step": 16206 + }, + { + "epoch": 1.2992924575574754, + "grad_norm": 1.7066259384155273, + "learning_rate": 5.77371931678387e-06, + "loss": 3.2392, + "step": 16207 + }, + { + "epoch": 1.299372632338498, + "grad_norm": 1.8184469938278198, + "learning_rate": 5.772542565673203e-06, + "loss": 3.2114, + "step": 16208 + }, + { + "epoch": 1.2994528071195206, + "grad_norm": 1.6996811628341675, + "learning_rate": 5.771365885835978e-06, + "loss": 3.1429, + "step": 16209 + }, + { + "epoch": 1.2995329819005432, + "grad_norm": 1.957244873046875, + "learning_rate": 5.7701892772920335e-06, + "loss": 3.1573, + "step": 16210 + }, + { + "epoch": 1.2996131566815659, + "grad_norm": 1.6962324380874634, + "learning_rate": 5.769012740061212e-06, + "loss": 3.225, + "step": 16211 + }, + { + "epoch": 1.2996933314625885, + "grad_norm": 1.6792657375335693, + "learning_rate": 5.7678362741633394e-06, + "loss": 3.183, + "step": 16212 + }, + { + "epoch": 1.299773506243611, + "grad_norm": 1.7033696174621582, + "learning_rate": 5.766659879618257e-06, + "loss": 3.0933, + "step": 16213 + }, + { + "epoch": 1.2998536810246337, + "grad_norm": 1.877482533454895, + "learning_rate": 5.765483556445795e-06, + "loss": 3.325, + "step": 16214 + }, + { + "epoch": 1.2999338558056563, + "grad_norm": 1.9871081113815308, + "learning_rate": 5.76430730466579e-06, + "loss": 3.1385, + "step": 16215 + }, + { + "epoch": 1.300014030586679, + "grad_norm": 1.7639586925506592, + "learning_rate": 5.763131124298074e-06, + "loss": 3.094, + "step": 16216 + }, + { + "epoch": 1.3000942053677016, + "grad_norm": 1.7682149410247803, + "learning_rate": 5.76195501536247e-06, + "loss": 3.1503, + "step": 16217 + }, + { + "epoch": 1.3001743801487242, + "grad_norm": 1.8602571487426758, + "learning_rate": 5.760778977878811e-06, + "loss": 3.1162, + "step": 16218 + }, + { + "epoch": 1.3002545549297468, + "grad_norm": 1.8731629848480225, + "learning_rate": 5.759603011866925e-06, + "loss": 3.201, + "step": 16219 + }, + { + "epoch": 1.3003347297107695, + "grad_norm": 1.7352330684661865, + "learning_rate": 5.758427117346637e-06, + "loss": 3.1561, + "step": 16220 + }, + { + "epoch": 1.300414904491792, + "grad_norm": 1.774212121963501, + "learning_rate": 5.757251294337777e-06, + "loss": 3.1556, + "step": 16221 + }, + { + "epoch": 1.3004950792728147, + "grad_norm": 1.6918965578079224, + "learning_rate": 5.756075542860161e-06, + "loss": 3.1749, + "step": 16222 + }, + { + "epoch": 1.3005752540538373, + "grad_norm": 1.7871835231781006, + "learning_rate": 5.754899862933616e-06, + "loss": 3.1952, + "step": 16223 + }, + { + "epoch": 1.30065542883486, + "grad_norm": 1.7501381635665894, + "learning_rate": 5.7537242545779625e-06, + "loss": 3.1691, + "step": 16224 + }, + { + "epoch": 1.3007356036158826, + "grad_norm": 2.027538299560547, + "learning_rate": 5.752548717813022e-06, + "loss": 3.1465, + "step": 16225 + }, + { + "epoch": 1.3008157783969052, + "grad_norm": 1.7413069009780884, + "learning_rate": 5.751373252658616e-06, + "loss": 3.2063, + "step": 16226 + }, + { + "epoch": 1.3008959531779278, + "grad_norm": 1.838286280632019, + "learning_rate": 5.750197859134557e-06, + "loss": 3.1083, + "step": 16227 + }, + { + "epoch": 1.3009761279589505, + "grad_norm": 1.8568720817565918, + "learning_rate": 5.7490225372606645e-06, + "loss": 3.1698, + "step": 16228 + }, + { + "epoch": 1.301056302739973, + "grad_norm": 1.8565189838409424, + "learning_rate": 5.747847287056753e-06, + "loss": 3.1852, + "step": 16229 + }, + { + "epoch": 1.3011364775209957, + "grad_norm": 1.7692762613296509, + "learning_rate": 5.746672108542638e-06, + "loss": 3.1044, + "step": 16230 + }, + { + "epoch": 1.3012166523020183, + "grad_norm": 1.9218618869781494, + "learning_rate": 5.745497001738137e-06, + "loss": 3.2121, + "step": 16231 + }, + { + "epoch": 1.301296827083041, + "grad_norm": 1.7105354070663452, + "learning_rate": 5.744321966663051e-06, + "loss": 3.0659, + "step": 16232 + }, + { + "epoch": 1.3013770018640636, + "grad_norm": 1.773646593093872, + "learning_rate": 5.743147003337198e-06, + "loss": 3.299, + "step": 16233 + }, + { + "epoch": 1.3014571766450862, + "grad_norm": 1.76064932346344, + "learning_rate": 5.741972111780387e-06, + "loss": 3.1798, + "step": 16234 + }, + { + "epoch": 1.3015373514261088, + "grad_norm": 0.6836373805999756, + "learning_rate": 5.740797292012424e-06, + "loss": 2.4812, + "step": 16235 + }, + { + "epoch": 1.3016175262071314, + "grad_norm": 1.7619082927703857, + "learning_rate": 5.739622544053123e-06, + "loss": 3.1989, + "step": 16236 + }, + { + "epoch": 1.301697700988154, + "grad_norm": 1.675439476966858, + "learning_rate": 5.738447867922279e-06, + "loss": 3.1517, + "step": 16237 + }, + { + "epoch": 1.3017778757691767, + "grad_norm": 1.992118239402771, + "learning_rate": 5.737273263639703e-06, + "loss": 3.0976, + "step": 16238 + }, + { + "epoch": 1.3018580505501993, + "grad_norm": 1.8139387369155884, + "learning_rate": 5.736098731225197e-06, + "loss": 3.2236, + "step": 16239 + }, + { + "epoch": 1.301938225331222, + "grad_norm": 1.874554991722107, + "learning_rate": 5.734924270698564e-06, + "loss": 3.162, + "step": 16240 + }, + { + "epoch": 1.3020184001122446, + "grad_norm": 0.6879628896713257, + "learning_rate": 5.7337498820796085e-06, + "loss": 2.4795, + "step": 16241 + }, + { + "epoch": 1.3020985748932672, + "grad_norm": 1.8146507740020752, + "learning_rate": 5.732575565388123e-06, + "loss": 3.1635, + "step": 16242 + }, + { + "epoch": 1.3021787496742898, + "grad_norm": 1.7724485397338867, + "learning_rate": 5.731401320643909e-06, + "loss": 3.1444, + "step": 16243 + }, + { + "epoch": 1.3022589244553127, + "grad_norm": 1.8062386512756348, + "learning_rate": 5.730227147866765e-06, + "loss": 3.1166, + "step": 16244 + }, + { + "epoch": 1.3023390992363353, + "grad_norm": 1.7152169942855835, + "learning_rate": 5.729053047076485e-06, + "loss": 3.1023, + "step": 16245 + }, + { + "epoch": 1.302419274017358, + "grad_norm": 1.67842698097229, + "learning_rate": 5.72787901829287e-06, + "loss": 3.2258, + "step": 16246 + }, + { + "epoch": 1.3024994487983805, + "grad_norm": 1.7062221765518188, + "learning_rate": 5.726705061535705e-06, + "loss": 3.1878, + "step": 16247 + }, + { + "epoch": 1.3025796235794032, + "grad_norm": 1.849585771560669, + "learning_rate": 5.725531176824788e-06, + "loss": 3.1096, + "step": 16248 + }, + { + "epoch": 1.3026597983604258, + "grad_norm": 1.740230917930603, + "learning_rate": 5.724357364179907e-06, + "loss": 3.3269, + "step": 16249 + }, + { + "epoch": 1.3027399731414484, + "grad_norm": 2.059338092803955, + "learning_rate": 5.723183623620855e-06, + "loss": 3.2578, + "step": 16250 + }, + { + "epoch": 1.302820147922471, + "grad_norm": 1.8067946434020996, + "learning_rate": 5.722009955167422e-06, + "loss": 3.203, + "step": 16251 + }, + { + "epoch": 1.3029003227034937, + "grad_norm": 1.7033039331436157, + "learning_rate": 5.720836358839392e-06, + "loss": 3.2256, + "step": 16252 + }, + { + "epoch": 1.3029804974845163, + "grad_norm": 0.6914034485816956, + "learning_rate": 5.719662834656552e-06, + "loss": 2.4792, + "step": 16253 + }, + { + "epoch": 1.303060672265539, + "grad_norm": 1.8946764469146729, + "learning_rate": 5.718489382638688e-06, + "loss": 3.2096, + "step": 16254 + }, + { + "epoch": 1.3031408470465615, + "grad_norm": 1.8111636638641357, + "learning_rate": 5.717316002805584e-06, + "loss": 3.2097, + "step": 16255 + }, + { + "epoch": 1.3032210218275841, + "grad_norm": 1.7635167837142944, + "learning_rate": 5.716142695177027e-06, + "loss": 3.0968, + "step": 16256 + }, + { + "epoch": 1.3033011966086068, + "grad_norm": 0.6781859397888184, + "learning_rate": 5.71496945977279e-06, + "loss": 2.4545, + "step": 16257 + }, + { + "epoch": 1.3033813713896294, + "grad_norm": 1.6876362562179565, + "learning_rate": 5.713796296612657e-06, + "loss": 3.1111, + "step": 16258 + }, + { + "epoch": 1.303461546170652, + "grad_norm": 1.8417880535125732, + "learning_rate": 5.71262320571641e-06, + "loss": 3.189, + "step": 16259 + }, + { + "epoch": 1.3035417209516746, + "grad_norm": 1.6683298349380493, + "learning_rate": 5.711450187103823e-06, + "loss": 3.1761, + "step": 16260 + }, + { + "epoch": 1.3036218957326973, + "grad_norm": 1.6496707201004028, + "learning_rate": 5.710277240794679e-06, + "loss": 3.0761, + "step": 16261 + }, + { + "epoch": 1.30370207051372, + "grad_norm": 1.7702504396438599, + "learning_rate": 5.709104366808745e-06, + "loss": 3.2017, + "step": 16262 + }, + { + "epoch": 1.3037822452947425, + "grad_norm": 0.6799158453941345, + "learning_rate": 5.707931565165798e-06, + "loss": 2.4471, + "step": 16263 + }, + { + "epoch": 1.3038624200757651, + "grad_norm": 1.8562214374542236, + "learning_rate": 5.706758835885613e-06, + "loss": 3.2924, + "step": 16264 + }, + { + "epoch": 1.3039425948567878, + "grad_norm": 1.7495957612991333, + "learning_rate": 5.705586178987961e-06, + "loss": 3.2054, + "step": 16265 + }, + { + "epoch": 1.3040227696378104, + "grad_norm": 1.8383166790008545, + "learning_rate": 5.704413594492617e-06, + "loss": 3.1771, + "step": 16266 + }, + { + "epoch": 1.304102944418833, + "grad_norm": 0.7396060228347778, + "learning_rate": 5.703241082419342e-06, + "loss": 2.4769, + "step": 16267 + }, + { + "epoch": 1.3041831191998556, + "grad_norm": 1.755924940109253, + "learning_rate": 5.702068642787907e-06, + "loss": 3.27, + "step": 16268 + }, + { + "epoch": 1.3042632939808783, + "grad_norm": 1.7455761432647705, + "learning_rate": 5.70089627561808e-06, + "loss": 3.1256, + "step": 16269 + }, + { + "epoch": 1.3043434687619009, + "grad_norm": 1.7360610961914062, + "learning_rate": 5.699723980929627e-06, + "loss": 3.0995, + "step": 16270 + }, + { + "epoch": 1.3044236435429235, + "grad_norm": 1.952213168144226, + "learning_rate": 5.6985517587423165e-06, + "loss": 3.304, + "step": 16271 + }, + { + "epoch": 1.3045038183239461, + "grad_norm": 1.6943129301071167, + "learning_rate": 5.6973796090759025e-06, + "loss": 3.1958, + "step": 16272 + }, + { + "epoch": 1.3045839931049688, + "grad_norm": 1.6594784259796143, + "learning_rate": 5.696207531950153e-06, + "loss": 3.2062, + "step": 16273 + }, + { + "epoch": 1.3046641678859916, + "grad_norm": 1.7886059284210205, + "learning_rate": 5.695035527384827e-06, + "loss": 3.1899, + "step": 16274 + }, + { + "epoch": 1.3047443426670142, + "grad_norm": 1.849769949913025, + "learning_rate": 5.693863595399684e-06, + "loss": 3.0304, + "step": 16275 + }, + { + "epoch": 1.3048245174480368, + "grad_norm": 1.939984679222107, + "learning_rate": 5.692691736014489e-06, + "loss": 3.271, + "step": 16276 + }, + { + "epoch": 1.3049046922290595, + "grad_norm": 1.7083832025527954, + "learning_rate": 5.691519949248987e-06, + "loss": 3.2083, + "step": 16277 + }, + { + "epoch": 1.304984867010082, + "grad_norm": 1.7931581735610962, + "learning_rate": 5.690348235122942e-06, + "loss": 3.1903, + "step": 16278 + }, + { + "epoch": 1.3050650417911047, + "grad_norm": 1.7754757404327393, + "learning_rate": 5.689176593656107e-06, + "loss": 3.1277, + "step": 16279 + }, + { + "epoch": 1.3051452165721273, + "grad_norm": 2.000378131866455, + "learning_rate": 5.688005024868234e-06, + "loss": 3.2314, + "step": 16280 + }, + { + "epoch": 1.30522539135315, + "grad_norm": 1.9091604948043823, + "learning_rate": 5.686833528779081e-06, + "loss": 3.2479, + "step": 16281 + }, + { + "epoch": 1.3053055661341726, + "grad_norm": 1.7866610288619995, + "learning_rate": 5.6856621054083915e-06, + "loss": 3.3435, + "step": 16282 + }, + { + "epoch": 1.3053857409151952, + "grad_norm": 1.7199254035949707, + "learning_rate": 5.684490754775917e-06, + "loss": 3.292, + "step": 16283 + }, + { + "epoch": 1.3054659156962178, + "grad_norm": 1.793105959892273, + "learning_rate": 5.6833194769014084e-06, + "loss": 3.1942, + "step": 16284 + }, + { + "epoch": 1.3055460904772405, + "grad_norm": 1.7385698556900024, + "learning_rate": 5.682148271804613e-06, + "loss": 3.1409, + "step": 16285 + }, + { + "epoch": 1.305626265258263, + "grad_norm": 1.6253581047058105, + "learning_rate": 5.6809771395052785e-06, + "loss": 3.1111, + "step": 16286 + }, + { + "epoch": 1.3057064400392857, + "grad_norm": 1.7135639190673828, + "learning_rate": 5.679806080023144e-06, + "loss": 3.0528, + "step": 16287 + }, + { + "epoch": 1.3057866148203083, + "grad_norm": 1.8845946788787842, + "learning_rate": 5.678635093377956e-06, + "loss": 3.2501, + "step": 16288 + }, + { + "epoch": 1.305866789601331, + "grad_norm": 1.7962442636489868, + "learning_rate": 5.677464179589457e-06, + "loss": 3.0868, + "step": 16289 + }, + { + "epoch": 1.3059469643823536, + "grad_norm": 1.7113556861877441, + "learning_rate": 5.676293338677389e-06, + "loss": 3.1137, + "step": 16290 + }, + { + "epoch": 1.3060271391633762, + "grad_norm": 0.6962838172912598, + "learning_rate": 5.675122570661495e-06, + "loss": 2.4774, + "step": 16291 + }, + { + "epoch": 1.3061073139443988, + "grad_norm": 1.7210279703140259, + "learning_rate": 5.673951875561508e-06, + "loss": 3.1527, + "step": 16292 + }, + { + "epoch": 1.3061874887254215, + "grad_norm": 1.7567867040634155, + "learning_rate": 5.672781253397165e-06, + "loss": 3.1589, + "step": 16293 + }, + { + "epoch": 1.306267663506444, + "grad_norm": 1.7747747898101807, + "learning_rate": 5.671610704188206e-06, + "loss": 3.3459, + "step": 16294 + }, + { + "epoch": 1.3063478382874667, + "grad_norm": 1.9120824337005615, + "learning_rate": 5.670440227954366e-06, + "loss": 3.1666, + "step": 16295 + }, + { + "epoch": 1.3064280130684893, + "grad_norm": 1.7554988861083984, + "learning_rate": 5.669269824715381e-06, + "loss": 3.1411, + "step": 16296 + }, + { + "epoch": 1.306508187849512, + "grad_norm": 0.6930145025253296, + "learning_rate": 5.668099494490978e-06, + "loss": 2.4574, + "step": 16297 + }, + { + "epoch": 1.3065883626305346, + "grad_norm": 1.8282437324523926, + "learning_rate": 5.666929237300889e-06, + "loss": 3.0942, + "step": 16298 + }, + { + "epoch": 1.3066685374115572, + "grad_norm": 1.8292149305343628, + "learning_rate": 5.665759053164846e-06, + "loss": 3.2502, + "step": 16299 + }, + { + "epoch": 1.3067487121925798, + "grad_norm": 0.7241186499595642, + "learning_rate": 5.664588942102578e-06, + "loss": 2.4893, + "step": 16300 + }, + { + "epoch": 1.3068288869736024, + "grad_norm": 1.997800588607788, + "learning_rate": 5.6634189041338175e-06, + "loss": 3.1646, + "step": 16301 + }, + { + "epoch": 1.306909061754625, + "grad_norm": 1.7705035209655762, + "learning_rate": 5.662248939278281e-06, + "loss": 3.2191, + "step": 16302 + }, + { + "epoch": 1.3069892365356477, + "grad_norm": 1.8114187717437744, + "learning_rate": 5.661079047555698e-06, + "loss": 3.2045, + "step": 16303 + }, + { + "epoch": 1.3070694113166703, + "grad_norm": 1.867576003074646, + "learning_rate": 5.6599092289857945e-06, + "loss": 3.1691, + "step": 16304 + }, + { + "epoch": 1.307149586097693, + "grad_norm": 1.8104499578475952, + "learning_rate": 5.658739483588291e-06, + "loss": 3.3813, + "step": 16305 + }, + { + "epoch": 1.3072297608787156, + "grad_norm": 2.0714099407196045, + "learning_rate": 5.6575698113829145e-06, + "loss": 3.2434, + "step": 16306 + }, + { + "epoch": 1.3073099356597382, + "grad_norm": 1.7667498588562012, + "learning_rate": 5.656400212389376e-06, + "loss": 3.2155, + "step": 16307 + }, + { + "epoch": 1.3073901104407608, + "grad_norm": 1.802765130996704, + "learning_rate": 5.6552306866274e-06, + "loss": 3.288, + "step": 16308 + }, + { + "epoch": 1.3074702852217834, + "grad_norm": 1.830557107925415, + "learning_rate": 5.654061234116701e-06, + "loss": 3.097, + "step": 16309 + }, + { + "epoch": 1.307550460002806, + "grad_norm": 1.8042988777160645, + "learning_rate": 5.652891854877001e-06, + "loss": 3.2289, + "step": 16310 + }, + { + "epoch": 1.3076306347838287, + "grad_norm": 1.661730408668518, + "learning_rate": 5.651722548928014e-06, + "loss": 3.0854, + "step": 16311 + }, + { + "epoch": 1.3077108095648513, + "grad_norm": 1.7971876859664917, + "learning_rate": 5.650553316289449e-06, + "loss": 3.2592, + "step": 16312 + }, + { + "epoch": 1.307790984345874, + "grad_norm": 0.6975507140159607, + "learning_rate": 5.649384156981022e-06, + "loss": 2.4833, + "step": 16313 + }, + { + "epoch": 1.3078711591268966, + "grad_norm": 1.7486894130706787, + "learning_rate": 5.648215071022445e-06, + "loss": 3.111, + "step": 16314 + }, + { + "epoch": 1.3079513339079192, + "grad_norm": 1.7594269514083862, + "learning_rate": 5.647046058433429e-06, + "loss": 3.1934, + "step": 16315 + }, + { + "epoch": 1.3080315086889418, + "grad_norm": 1.7219678163528442, + "learning_rate": 5.645877119233685e-06, + "loss": 3.0947, + "step": 16316 + }, + { + "epoch": 1.3081116834699644, + "grad_norm": 1.7363471984863281, + "learning_rate": 5.644708253442914e-06, + "loss": 3.2542, + "step": 16317 + }, + { + "epoch": 1.308191858250987, + "grad_norm": 1.7913035154342651, + "learning_rate": 5.643539461080829e-06, + "loss": 3.2052, + "step": 16318 + }, + { + "epoch": 1.3082720330320097, + "grad_norm": 0.6695003509521484, + "learning_rate": 5.642370742167132e-06, + "loss": 2.4584, + "step": 16319 + }, + { + "epoch": 1.3083522078130323, + "grad_norm": 1.7898832559585571, + "learning_rate": 5.6412020967215275e-06, + "loss": 3.0745, + "step": 16320 + }, + { + "epoch": 1.308432382594055, + "grad_norm": 1.7708361148834229, + "learning_rate": 5.640033524763725e-06, + "loss": 3.0929, + "step": 16321 + }, + { + "epoch": 1.3085125573750775, + "grad_norm": 1.7569260597229004, + "learning_rate": 5.638865026313416e-06, + "loss": 3.1408, + "step": 16322 + }, + { + "epoch": 1.3085927321561002, + "grad_norm": 1.9320148229599, + "learning_rate": 5.637696601390305e-06, + "loss": 3.3013, + "step": 16323 + }, + { + "epoch": 1.3086729069371228, + "grad_norm": 1.857303261756897, + "learning_rate": 5.636528250014094e-06, + "loss": 3.1154, + "step": 16324 + }, + { + "epoch": 1.3087530817181456, + "grad_norm": 1.6835006475448608, + "learning_rate": 5.6353599722044775e-06, + "loss": 3.2386, + "step": 16325 + }, + { + "epoch": 1.3088332564991683, + "grad_norm": 1.8073662519454956, + "learning_rate": 5.634191767981158e-06, + "loss": 3.2585, + "step": 16326 + }, + { + "epoch": 1.3089134312801909, + "grad_norm": 1.6252154111862183, + "learning_rate": 5.633023637363821e-06, + "loss": 3.1138, + "step": 16327 + }, + { + "epoch": 1.3089936060612135, + "grad_norm": 1.7710708379745483, + "learning_rate": 5.631855580372169e-06, + "loss": 3.2766, + "step": 16328 + }, + { + "epoch": 1.3090737808422361, + "grad_norm": 1.757533311843872, + "learning_rate": 5.63068759702589e-06, + "loss": 3.2061, + "step": 16329 + }, + { + "epoch": 1.3091539556232588, + "grad_norm": 1.7144309282302856, + "learning_rate": 5.62951968734468e-06, + "loss": 3.1842, + "step": 16330 + }, + { + "epoch": 1.3092341304042814, + "grad_norm": 1.8351346254348755, + "learning_rate": 5.628351851348231e-06, + "loss": 3.2216, + "step": 16331 + }, + { + "epoch": 1.309314305185304, + "grad_norm": 1.8548630475997925, + "learning_rate": 5.6271840890562255e-06, + "loss": 3.1794, + "step": 16332 + }, + { + "epoch": 1.3093944799663266, + "grad_norm": 1.6714376211166382, + "learning_rate": 5.6260164004883545e-06, + "loss": 3.0687, + "step": 16333 + }, + { + "epoch": 1.3094746547473493, + "grad_norm": 1.7254348993301392, + "learning_rate": 5.624848785664305e-06, + "loss": 3.2465, + "step": 16334 + }, + { + "epoch": 1.3095548295283719, + "grad_norm": 1.7983733415603638, + "learning_rate": 5.6236812446037644e-06, + "loss": 3.2231, + "step": 16335 + }, + { + "epoch": 1.3096350043093945, + "grad_norm": 0.6830950975418091, + "learning_rate": 5.62251377732642e-06, + "loss": 2.4897, + "step": 16336 + }, + { + "epoch": 1.3097151790904171, + "grad_norm": 0.6739800572395325, + "learning_rate": 5.621346383851944e-06, + "loss": 2.4719, + "step": 16337 + }, + { + "epoch": 1.3097953538714397, + "grad_norm": 1.699245810508728, + "learning_rate": 5.620179064200028e-06, + "loss": 3.0441, + "step": 16338 + }, + { + "epoch": 1.3098755286524624, + "grad_norm": 1.8482600450515747, + "learning_rate": 5.619011818390348e-06, + "loss": 3.3173, + "step": 16339 + }, + { + "epoch": 1.309955703433485, + "grad_norm": 1.6929765939712524, + "learning_rate": 5.6178446464425825e-06, + "loss": 3.2581, + "step": 16340 + }, + { + "epoch": 1.3100358782145076, + "grad_norm": 1.7187095880508423, + "learning_rate": 5.616677548376419e-06, + "loss": 3.2636, + "step": 16341 + }, + { + "epoch": 1.3101160529955302, + "grad_norm": 1.6782341003417969, + "learning_rate": 5.615510524211522e-06, + "loss": 3.2672, + "step": 16342 + }, + { + "epoch": 1.3101962277765529, + "grad_norm": 1.7442781925201416, + "learning_rate": 5.614343573967572e-06, + "loss": 3.1614, + "step": 16343 + }, + { + "epoch": 1.3102764025575755, + "grad_norm": 1.7835841178894043, + "learning_rate": 5.6131766976642445e-06, + "loss": 3.1799, + "step": 16344 + }, + { + "epoch": 1.3103565773385981, + "grad_norm": 1.8256752490997314, + "learning_rate": 5.612009895321211e-06, + "loss": 3.1901, + "step": 16345 + }, + { + "epoch": 1.3104367521196207, + "grad_norm": 1.6895384788513184, + "learning_rate": 5.610843166958149e-06, + "loss": 3.0441, + "step": 16346 + }, + { + "epoch": 1.3105169269006434, + "grad_norm": 1.8000518083572388, + "learning_rate": 5.60967651259472e-06, + "loss": 3.1608, + "step": 16347 + }, + { + "epoch": 1.310597101681666, + "grad_norm": 0.6681274771690369, + "learning_rate": 5.608509932250596e-06, + "loss": 2.4838, + "step": 16348 + }, + { + "epoch": 1.3106772764626886, + "grad_norm": 1.7787084579467773, + "learning_rate": 5.60734342594545e-06, + "loss": 3.1923, + "step": 16349 + }, + { + "epoch": 1.3107574512437112, + "grad_norm": 1.828358769416809, + "learning_rate": 5.606176993698944e-06, + "loss": 3.1373, + "step": 16350 + }, + { + "epoch": 1.3108376260247339, + "grad_norm": 1.7046113014221191, + "learning_rate": 5.6050106355307486e-06, + "loss": 3.0795, + "step": 16351 + }, + { + "epoch": 1.3109178008057565, + "grad_norm": 1.7451817989349365, + "learning_rate": 5.603844351460523e-06, + "loss": 3.1488, + "step": 16352 + }, + { + "epoch": 1.310997975586779, + "grad_norm": 1.79727041721344, + "learning_rate": 5.60267814150793e-06, + "loss": 3.2606, + "step": 16353 + }, + { + "epoch": 1.3110781503678017, + "grad_norm": 1.6859519481658936, + "learning_rate": 5.601512005692636e-06, + "loss": 3.1598, + "step": 16354 + }, + { + "epoch": 1.3111583251488244, + "grad_norm": 1.9396790266036987, + "learning_rate": 5.600345944034297e-06, + "loss": 3.2131, + "step": 16355 + }, + { + "epoch": 1.3112384999298472, + "grad_norm": 1.836103081703186, + "learning_rate": 5.599179956552581e-06, + "loss": 3.121, + "step": 16356 + }, + { + "epoch": 1.3113186747108698, + "grad_norm": 1.629978060722351, + "learning_rate": 5.5980140432671345e-06, + "loss": 3.135, + "step": 16357 + }, + { + "epoch": 1.3113988494918924, + "grad_norm": 1.7856494188308716, + "learning_rate": 5.59684820419762e-06, + "loss": 3.1804, + "step": 16358 + }, + { + "epoch": 1.311479024272915, + "grad_norm": 1.6936825513839722, + "learning_rate": 5.595682439363693e-06, + "loss": 3.1967, + "step": 16359 + }, + { + "epoch": 1.3115591990539377, + "grad_norm": 1.6566647291183472, + "learning_rate": 5.594516748785009e-06, + "loss": 3.1252, + "step": 16360 + }, + { + "epoch": 1.3116393738349603, + "grad_norm": 1.870364785194397, + "learning_rate": 5.593351132481222e-06, + "loss": 3.2856, + "step": 16361 + }, + { + "epoch": 1.311719548615983, + "grad_norm": 1.7824779748916626, + "learning_rate": 5.592185590471979e-06, + "loss": 3.2903, + "step": 16362 + }, + { + "epoch": 1.3117997233970056, + "grad_norm": 1.7953218221664429, + "learning_rate": 5.591020122776934e-06, + "loss": 3.1216, + "step": 16363 + }, + { + "epoch": 1.3118798981780282, + "grad_norm": 1.8014196157455444, + "learning_rate": 5.589854729415735e-06, + "loss": 3.2101, + "step": 16364 + }, + { + "epoch": 1.3119600729590508, + "grad_norm": 1.8092966079711914, + "learning_rate": 5.58868941040803e-06, + "loss": 3.1329, + "step": 16365 + }, + { + "epoch": 1.3120402477400734, + "grad_norm": 1.8070534467697144, + "learning_rate": 5.587524165773472e-06, + "loss": 3.1974, + "step": 16366 + }, + { + "epoch": 1.312120422521096, + "grad_norm": 1.907076358795166, + "learning_rate": 5.586358995531698e-06, + "loss": 3.1683, + "step": 16367 + }, + { + "epoch": 1.3122005973021187, + "grad_norm": 1.8214682340621948, + "learning_rate": 5.585193899702356e-06, + "loss": 3.21, + "step": 16368 + }, + { + "epoch": 1.3122807720831413, + "grad_norm": 1.7463457584381104, + "learning_rate": 5.584028878305088e-06, + "loss": 3.1338, + "step": 16369 + }, + { + "epoch": 1.312360946864164, + "grad_norm": 1.743056297302246, + "learning_rate": 5.582863931359537e-06, + "loss": 3.1058, + "step": 16370 + }, + { + "epoch": 1.3124411216451866, + "grad_norm": 1.7712782621383667, + "learning_rate": 5.5816990588853485e-06, + "loss": 3.1517, + "step": 16371 + }, + { + "epoch": 1.3125212964262092, + "grad_norm": 1.7537167072296143, + "learning_rate": 5.580534260902151e-06, + "loss": 3.095, + "step": 16372 + }, + { + "epoch": 1.3126014712072318, + "grad_norm": 1.7170884609222412, + "learning_rate": 5.579369537429588e-06, + "loss": 3.1786, + "step": 16373 + }, + { + "epoch": 1.3126816459882544, + "grad_norm": 1.7269573211669922, + "learning_rate": 5.578204888487297e-06, + "loss": 3.1758, + "step": 16374 + }, + { + "epoch": 1.312761820769277, + "grad_norm": 1.8265506029129028, + "learning_rate": 5.577040314094915e-06, + "loss": 3.0685, + "step": 16375 + }, + { + "epoch": 1.3128419955502997, + "grad_norm": 1.6458684206008911, + "learning_rate": 5.575875814272077e-06, + "loss": 3.1387, + "step": 16376 + }, + { + "epoch": 1.3129221703313223, + "grad_norm": 1.9385789632797241, + "learning_rate": 5.5747113890384094e-06, + "loss": 3.1388, + "step": 16377 + }, + { + "epoch": 1.313002345112345, + "grad_norm": 1.7070765495300293, + "learning_rate": 5.573547038413549e-06, + "loss": 3.1072, + "step": 16378 + }, + { + "epoch": 1.3130825198933676, + "grad_norm": 1.9016227722167969, + "learning_rate": 5.572382762417125e-06, + "loss": 3.0505, + "step": 16379 + }, + { + "epoch": 1.3131626946743902, + "grad_norm": 1.7409662008285522, + "learning_rate": 5.571218561068767e-06, + "loss": 3.1984, + "step": 16380 + }, + { + "epoch": 1.3132428694554128, + "grad_norm": 1.858231782913208, + "learning_rate": 5.570054434388108e-06, + "loss": 3.1312, + "step": 16381 + }, + { + "epoch": 1.3133230442364354, + "grad_norm": 1.690989375114441, + "learning_rate": 5.568890382394765e-06, + "loss": 3.1493, + "step": 16382 + }, + { + "epoch": 1.313403219017458, + "grad_norm": 1.790109634399414, + "learning_rate": 5.5677264051083694e-06, + "loss": 3.1779, + "step": 16383 + }, + { + "epoch": 1.3134833937984807, + "grad_norm": 1.8204054832458496, + "learning_rate": 5.566562502548543e-06, + "loss": 3.1161, + "step": 16384 + }, + { + "epoch": 1.3135635685795033, + "grad_norm": 1.8549909591674805, + "learning_rate": 5.565398674734912e-06, + "loss": 3.2018, + "step": 16385 + }, + { + "epoch": 1.313643743360526, + "grad_norm": 1.6740857362747192, + "learning_rate": 5.5642349216870995e-06, + "loss": 3.1497, + "step": 16386 + }, + { + "epoch": 1.3137239181415485, + "grad_norm": 1.8854001760482788, + "learning_rate": 5.563071243424718e-06, + "loss": 3.1484, + "step": 16387 + }, + { + "epoch": 1.3138040929225712, + "grad_norm": 1.734507441520691, + "learning_rate": 5.561907639967393e-06, + "loss": 3.168, + "step": 16388 + }, + { + "epoch": 1.3138842677035938, + "grad_norm": 1.7524542808532715, + "learning_rate": 5.560744111334742e-06, + "loss": 3.0288, + "step": 16389 + }, + { + "epoch": 1.3139644424846164, + "grad_norm": 2.762721061706543, + "learning_rate": 5.559580657546379e-06, + "loss": 3.2017, + "step": 16390 + }, + { + "epoch": 1.314044617265639, + "grad_norm": 1.786885142326355, + "learning_rate": 5.558417278621926e-06, + "loss": 3.1896, + "step": 16391 + }, + { + "epoch": 1.3141247920466617, + "grad_norm": 1.7653613090515137, + "learning_rate": 5.557253974580987e-06, + "loss": 3.1435, + "step": 16392 + }, + { + "epoch": 1.3142049668276843, + "grad_norm": 1.7185652256011963, + "learning_rate": 5.556090745443181e-06, + "loss": 3.203, + "step": 16393 + }, + { + "epoch": 1.314285141608707, + "grad_norm": 1.729347586631775, + "learning_rate": 5.55492759122812e-06, + "loss": 3.1926, + "step": 16394 + }, + { + "epoch": 1.3143653163897295, + "grad_norm": 1.6640026569366455, + "learning_rate": 5.5537645119554114e-06, + "loss": 3.2148, + "step": 16395 + }, + { + "epoch": 1.3144454911707522, + "grad_norm": 1.7683982849121094, + "learning_rate": 5.552601507644671e-06, + "loss": 3.2241, + "step": 16396 + }, + { + "epoch": 1.3145256659517748, + "grad_norm": 1.735679030418396, + "learning_rate": 5.551438578315498e-06, + "loss": 3.1666, + "step": 16397 + }, + { + "epoch": 1.3146058407327974, + "grad_norm": 1.8888063430786133, + "learning_rate": 5.5502757239875025e-06, + "loss": 3.2166, + "step": 16398 + }, + { + "epoch": 1.31468601551382, + "grad_norm": 1.7284919023513794, + "learning_rate": 5.549112944680291e-06, + "loss": 3.1564, + "step": 16399 + }, + { + "epoch": 1.3147661902948427, + "grad_norm": 1.7840073108673096, + "learning_rate": 5.547950240413465e-06, + "loss": 3.1986, + "step": 16400 + }, + { + "epoch": 1.3148463650758653, + "grad_norm": 1.6885418891906738, + "learning_rate": 5.546787611206635e-06, + "loss": 3.0978, + "step": 16401 + }, + { + "epoch": 1.314926539856888, + "grad_norm": 1.8548603057861328, + "learning_rate": 5.545625057079389e-06, + "loss": 3.1851, + "step": 16402 + }, + { + "epoch": 1.3150067146379105, + "grad_norm": 1.7661242485046387, + "learning_rate": 5.544462578051337e-06, + "loss": 3.0594, + "step": 16403 + }, + { + "epoch": 1.3150868894189331, + "grad_norm": 1.732427716255188, + "learning_rate": 5.543300174142075e-06, + "loss": 3.2557, + "step": 16404 + }, + { + "epoch": 1.3151670641999558, + "grad_norm": 1.8032341003417969, + "learning_rate": 5.542137845371203e-06, + "loss": 3.2421, + "step": 16405 + }, + { + "epoch": 1.3152472389809784, + "grad_norm": 1.8143595457077026, + "learning_rate": 5.540975591758317e-06, + "loss": 3.1119, + "step": 16406 + }, + { + "epoch": 1.3153274137620012, + "grad_norm": 1.7800979614257812, + "learning_rate": 5.539813413323009e-06, + "loss": 3.1578, + "step": 16407 + }, + { + "epoch": 1.3154075885430239, + "grad_norm": 1.7510137557983398, + "learning_rate": 5.538651310084875e-06, + "loss": 3.1753, + "step": 16408 + }, + { + "epoch": 1.3154877633240465, + "grad_norm": 1.8137584924697876, + "learning_rate": 5.537489282063506e-06, + "loss": 3.2111, + "step": 16409 + }, + { + "epoch": 1.3155679381050691, + "grad_norm": 1.6718202829360962, + "learning_rate": 5.536327329278497e-06, + "loss": 3.0258, + "step": 16410 + }, + { + "epoch": 1.3156481128860917, + "grad_norm": 1.8635667562484741, + "learning_rate": 5.535165451749438e-06, + "loss": 3.2079, + "step": 16411 + }, + { + "epoch": 1.3157282876671144, + "grad_norm": 1.8300272226333618, + "learning_rate": 5.534003649495914e-06, + "loss": 3.2184, + "step": 16412 + }, + { + "epoch": 1.315808462448137, + "grad_norm": 1.7465876340866089, + "learning_rate": 5.532841922537513e-06, + "loss": 3.1066, + "step": 16413 + }, + { + "epoch": 1.3158886372291596, + "grad_norm": 1.753974437713623, + "learning_rate": 5.531680270893824e-06, + "loss": 3.2074, + "step": 16414 + }, + { + "epoch": 1.3159688120101822, + "grad_norm": 1.8180184364318848, + "learning_rate": 5.53051869458443e-06, + "loss": 3.2009, + "step": 16415 + }, + { + "epoch": 1.3160489867912049, + "grad_norm": 1.7479029893875122, + "learning_rate": 5.529357193628919e-06, + "loss": 3.1367, + "step": 16416 + }, + { + "epoch": 1.3161291615722275, + "grad_norm": 1.7713996171951294, + "learning_rate": 5.528195768046867e-06, + "loss": 3.1351, + "step": 16417 + }, + { + "epoch": 1.31620933635325, + "grad_norm": 1.824702501296997, + "learning_rate": 5.5270344178578584e-06, + "loss": 3.1852, + "step": 16418 + }, + { + "epoch": 1.3162895111342727, + "grad_norm": 1.7274847030639648, + "learning_rate": 5.525873143081473e-06, + "loss": 3.1733, + "step": 16419 + }, + { + "epoch": 1.3163696859152954, + "grad_norm": 1.7746868133544922, + "learning_rate": 5.5247119437372884e-06, + "loss": 3.2335, + "step": 16420 + }, + { + "epoch": 1.316449860696318, + "grad_norm": 1.6861249208450317, + "learning_rate": 5.5235508198448895e-06, + "loss": 3.2626, + "step": 16421 + }, + { + "epoch": 1.3165300354773406, + "grad_norm": 1.756917953491211, + "learning_rate": 5.522389771423841e-06, + "loss": 3.1317, + "step": 16422 + }, + { + "epoch": 1.3166102102583632, + "grad_norm": 1.837494969367981, + "learning_rate": 5.521228798493723e-06, + "loss": 3.196, + "step": 16423 + }, + { + "epoch": 1.3166903850393858, + "grad_norm": 1.7411627769470215, + "learning_rate": 5.520067901074109e-06, + "loss": 3.2145, + "step": 16424 + }, + { + "epoch": 1.3167705598204085, + "grad_norm": 1.7514451742172241, + "learning_rate": 5.51890707918457e-06, + "loss": 3.0593, + "step": 16425 + }, + { + "epoch": 1.316850734601431, + "grad_norm": 1.816048502922058, + "learning_rate": 5.517746332844684e-06, + "loss": 3.1912, + "step": 16426 + }, + { + "epoch": 1.3169309093824537, + "grad_norm": 1.726904273033142, + "learning_rate": 5.51658566207401e-06, + "loss": 3.239, + "step": 16427 + }, + { + "epoch": 1.3170110841634763, + "grad_norm": 1.7803375720977783, + "learning_rate": 5.515425066892122e-06, + "loss": 3.1967, + "step": 16428 + }, + { + "epoch": 1.317091258944499, + "grad_norm": 1.7806485891342163, + "learning_rate": 5.514264547318586e-06, + "loss": 3.0463, + "step": 16429 + }, + { + "epoch": 1.3171714337255216, + "grad_norm": 1.7043989896774292, + "learning_rate": 5.513104103372971e-06, + "loss": 3.1637, + "step": 16430 + }, + { + "epoch": 1.3172516085065442, + "grad_norm": 1.690182089805603, + "learning_rate": 5.511943735074843e-06, + "loss": 3.1661, + "step": 16431 + }, + { + "epoch": 1.3173317832875668, + "grad_norm": 1.7483925819396973, + "learning_rate": 5.510783442443757e-06, + "loss": 3.3058, + "step": 16432 + }, + { + "epoch": 1.3174119580685895, + "grad_norm": 1.741584062576294, + "learning_rate": 5.509623225499279e-06, + "loss": 3.1206, + "step": 16433 + }, + { + "epoch": 1.317492132849612, + "grad_norm": 1.7648136615753174, + "learning_rate": 5.508463084260974e-06, + "loss": 3.2358, + "step": 16434 + }, + { + "epoch": 1.3175723076306347, + "grad_norm": 1.7488470077514648, + "learning_rate": 5.5073030187483955e-06, + "loss": 3.1636, + "step": 16435 + }, + { + "epoch": 1.3176524824116573, + "grad_norm": 1.8482979536056519, + "learning_rate": 5.506143028981109e-06, + "loss": 3.281, + "step": 16436 + }, + { + "epoch": 1.3177326571926802, + "grad_norm": 1.8793352842330933, + "learning_rate": 5.504983114978664e-06, + "loss": 3.1752, + "step": 16437 + }, + { + "epoch": 1.3178128319737028, + "grad_norm": 1.7371892929077148, + "learning_rate": 5.503823276760619e-06, + "loss": 3.2307, + "step": 16438 + }, + { + "epoch": 1.3178930067547254, + "grad_norm": 1.7545714378356934, + "learning_rate": 5.502663514346529e-06, + "loss": 3.1284, + "step": 16439 + }, + { + "epoch": 1.317973181535748, + "grad_norm": 1.826758861541748, + "learning_rate": 5.501503827755948e-06, + "loss": 3.2673, + "step": 16440 + }, + { + "epoch": 1.3180533563167707, + "grad_norm": 1.8120927810668945, + "learning_rate": 5.500344217008431e-06, + "loss": 3.1957, + "step": 16441 + }, + { + "epoch": 1.3181335310977933, + "grad_norm": 1.7599033117294312, + "learning_rate": 5.499184682123519e-06, + "loss": 3.1928, + "step": 16442 + }, + { + "epoch": 1.318213705878816, + "grad_norm": 1.7424898147583008, + "learning_rate": 5.4980252231207685e-06, + "loss": 3.1775, + "step": 16443 + }, + { + "epoch": 1.3182938806598385, + "grad_norm": 1.7151250839233398, + "learning_rate": 5.496865840019727e-06, + "loss": 3.2402, + "step": 16444 + }, + { + "epoch": 1.3183740554408612, + "grad_norm": 1.7556341886520386, + "learning_rate": 5.495706532839939e-06, + "loss": 3.1584, + "step": 16445 + }, + { + "epoch": 1.3184542302218838, + "grad_norm": 1.8263378143310547, + "learning_rate": 5.4945473016009555e-06, + "loss": 3.2397, + "step": 16446 + }, + { + "epoch": 1.3185344050029064, + "grad_norm": 1.8380498886108398, + "learning_rate": 5.493388146322313e-06, + "loss": 3.2788, + "step": 16447 + }, + { + "epoch": 1.318614579783929, + "grad_norm": 1.7861868143081665, + "learning_rate": 5.492229067023558e-06, + "loss": 3.1875, + "step": 16448 + }, + { + "epoch": 1.3186947545649517, + "grad_norm": 1.855040192604065, + "learning_rate": 5.491070063724231e-06, + "loss": 3.2125, + "step": 16449 + }, + { + "epoch": 1.3187749293459743, + "grad_norm": 1.730499029159546, + "learning_rate": 5.489911136443875e-06, + "loss": 3.1554, + "step": 16450 + }, + { + "epoch": 1.318855104126997, + "grad_norm": 1.9252307415008545, + "learning_rate": 5.488752285202031e-06, + "loss": 3.1669, + "step": 16451 + }, + { + "epoch": 1.3189352789080195, + "grad_norm": 1.8281813859939575, + "learning_rate": 5.48759351001823e-06, + "loss": 3.2335, + "step": 16452 + }, + { + "epoch": 1.3190154536890422, + "grad_norm": 1.892866849899292, + "learning_rate": 5.486434810912012e-06, + "loss": 3.1591, + "step": 16453 + }, + { + "epoch": 1.3190956284700648, + "grad_norm": 1.8161125183105469, + "learning_rate": 5.4852761879029125e-06, + "loss": 3.1767, + "step": 16454 + }, + { + "epoch": 1.3191758032510874, + "grad_norm": 1.7696897983551025, + "learning_rate": 5.484117641010465e-06, + "loss": 3.2764, + "step": 16455 + }, + { + "epoch": 1.31925597803211, + "grad_norm": 1.7401713132858276, + "learning_rate": 5.482959170254207e-06, + "loss": 3.175, + "step": 16456 + }, + { + "epoch": 1.3193361528131327, + "grad_norm": 0.7022193670272827, + "learning_rate": 5.481800775653661e-06, + "loss": 2.4868, + "step": 16457 + }, + { + "epoch": 1.3194163275941553, + "grad_norm": 1.7960002422332764, + "learning_rate": 5.4806424572283625e-06, + "loss": 3.1902, + "step": 16458 + }, + { + "epoch": 1.319496502375178, + "grad_norm": 0.7163276672363281, + "learning_rate": 5.479484214997838e-06, + "loss": 2.4862, + "step": 16459 + }, + { + "epoch": 1.3195766771562005, + "grad_norm": 1.7377365827560425, + "learning_rate": 5.478326048981617e-06, + "loss": 3.1533, + "step": 16460 + }, + { + "epoch": 1.3196568519372232, + "grad_norm": 1.7694209814071655, + "learning_rate": 5.477167959199229e-06, + "loss": 3.1961, + "step": 16461 + }, + { + "epoch": 1.3197370267182458, + "grad_norm": 1.8253549337387085, + "learning_rate": 5.476009945670191e-06, + "loss": 3.2338, + "step": 16462 + }, + { + "epoch": 1.3198172014992684, + "grad_norm": 1.8446269035339355, + "learning_rate": 5.474852008414031e-06, + "loss": 3.1101, + "step": 16463 + }, + { + "epoch": 1.319897376280291, + "grad_norm": 1.7593523263931274, + "learning_rate": 5.473694147450272e-06, + "loss": 3.1102, + "step": 16464 + }, + { + "epoch": 1.3199775510613136, + "grad_norm": 1.7102078199386597, + "learning_rate": 5.472536362798435e-06, + "loss": 3.1826, + "step": 16465 + }, + { + "epoch": 1.3200577258423363, + "grad_norm": 1.7654865980148315, + "learning_rate": 5.471378654478042e-06, + "loss": 3.201, + "step": 16466 + }, + { + "epoch": 1.320137900623359, + "grad_norm": 1.7504163980484009, + "learning_rate": 5.470221022508606e-06, + "loss": 3.2094, + "step": 16467 + }, + { + "epoch": 1.3202180754043815, + "grad_norm": 1.7433016300201416, + "learning_rate": 5.469063466909646e-06, + "loss": 3.2358, + "step": 16468 + }, + { + "epoch": 1.3202982501854041, + "grad_norm": 1.834580898284912, + "learning_rate": 5.467905987700679e-06, + "loss": 3.1776, + "step": 16469 + }, + { + "epoch": 1.3203784249664268, + "grad_norm": 1.6875367164611816, + "learning_rate": 5.46674858490122e-06, + "loss": 3.1341, + "step": 16470 + }, + { + "epoch": 1.3204585997474494, + "grad_norm": 1.720045566558838, + "learning_rate": 5.4655912585307865e-06, + "loss": 3.2161, + "step": 16471 + }, + { + "epoch": 1.320538774528472, + "grad_norm": 1.819944977760315, + "learning_rate": 5.464434008608883e-06, + "loss": 3.1141, + "step": 16472 + }, + { + "epoch": 1.3206189493094946, + "grad_norm": 1.8773642778396606, + "learning_rate": 5.463276835155023e-06, + "loss": 3.13, + "step": 16473 + }, + { + "epoch": 1.3206991240905173, + "grad_norm": 1.7730423212051392, + "learning_rate": 5.4621197381887156e-06, + "loss": 3.1217, + "step": 16474 + }, + { + "epoch": 1.3207792988715399, + "grad_norm": 0.6743487119674683, + "learning_rate": 5.4609627177294725e-06, + "loss": 2.4838, + "step": 16475 + }, + { + "epoch": 1.3208594736525625, + "grad_norm": 1.8382089138031006, + "learning_rate": 5.4598057737968e-06, + "loss": 3.1492, + "step": 16476 + }, + { + "epoch": 1.3209396484335851, + "grad_norm": 1.779829740524292, + "learning_rate": 5.4586489064101985e-06, + "loss": 3.1876, + "step": 16477 + }, + { + "epoch": 1.3210198232146078, + "grad_norm": 1.7954169511795044, + "learning_rate": 5.457492115589176e-06, + "loss": 3.2065, + "step": 16478 + }, + { + "epoch": 1.3210999979956304, + "grad_norm": 0.6577844619750977, + "learning_rate": 5.456335401353236e-06, + "loss": 2.4855, + "step": 16479 + }, + { + "epoch": 1.321180172776653, + "grad_norm": 1.6587368249893188, + "learning_rate": 5.45517876372188e-06, + "loss": 3.0878, + "step": 16480 + }, + { + "epoch": 1.3212603475576756, + "grad_norm": 1.7636035680770874, + "learning_rate": 5.454022202714612e-06, + "loss": 3.1704, + "step": 16481 + }, + { + "epoch": 1.3213405223386983, + "grad_norm": 1.842811942100525, + "learning_rate": 5.4528657183509235e-06, + "loss": 3.1277, + "step": 16482 + }, + { + "epoch": 1.3214206971197209, + "grad_norm": 1.7417664527893066, + "learning_rate": 5.4517093106503146e-06, + "loss": 3.2084, + "step": 16483 + }, + { + "epoch": 1.3215008719007435, + "grad_norm": 1.8105636835098267, + "learning_rate": 5.450552979632286e-06, + "loss": 3.1471, + "step": 16484 + }, + { + "epoch": 1.3215810466817661, + "grad_norm": 1.7853341102600098, + "learning_rate": 5.449396725316332e-06, + "loss": 3.1978, + "step": 16485 + }, + { + "epoch": 1.3216612214627887, + "grad_norm": 1.7439415454864502, + "learning_rate": 5.448240547721946e-06, + "loss": 3.0924, + "step": 16486 + }, + { + "epoch": 1.3217413962438114, + "grad_norm": 1.7180827856063843, + "learning_rate": 5.447084446868618e-06, + "loss": 3.1663, + "step": 16487 + }, + { + "epoch": 1.3218215710248342, + "grad_norm": 1.6346033811569214, + "learning_rate": 5.445928422775841e-06, + "loss": 3.0573, + "step": 16488 + }, + { + "epoch": 1.3219017458058568, + "grad_norm": 1.7177702188491821, + "learning_rate": 5.444772475463107e-06, + "loss": 3.1483, + "step": 16489 + }, + { + "epoch": 1.3219819205868795, + "grad_norm": 1.777105689048767, + "learning_rate": 5.443616604949903e-06, + "loss": 3.1616, + "step": 16490 + }, + { + "epoch": 1.322062095367902, + "grad_norm": 1.8683310747146606, + "learning_rate": 5.442460811255721e-06, + "loss": 3.1307, + "step": 16491 + }, + { + "epoch": 1.3221422701489247, + "grad_norm": 1.671632170677185, + "learning_rate": 5.441305094400039e-06, + "loss": 3.1172, + "step": 16492 + }, + { + "epoch": 1.3222224449299473, + "grad_norm": 1.785331130027771, + "learning_rate": 5.440149454402347e-06, + "loss": 3.2197, + "step": 16493 + }, + { + "epoch": 1.32230261971097, + "grad_norm": 1.894896149635315, + "learning_rate": 5.438993891282128e-06, + "loss": 3.2186, + "step": 16494 + }, + { + "epoch": 1.3223827944919926, + "grad_norm": 1.75314462184906, + "learning_rate": 5.437838405058865e-06, + "loss": 3.1119, + "step": 16495 + }, + { + "epoch": 1.3224629692730152, + "grad_norm": 1.7928831577301025, + "learning_rate": 5.436682995752043e-06, + "loss": 3.1657, + "step": 16496 + }, + { + "epoch": 1.3225431440540378, + "grad_norm": 1.8391410112380981, + "learning_rate": 5.435527663381133e-06, + "loss": 3.1838, + "step": 16497 + }, + { + "epoch": 1.3226233188350605, + "grad_norm": 1.6748706102371216, + "learning_rate": 5.434372407965617e-06, + "loss": 3.0979, + "step": 16498 + }, + { + "epoch": 1.322703493616083, + "grad_norm": 1.7408671379089355, + "learning_rate": 5.433217229524974e-06, + "loss": 3.2799, + "step": 16499 + }, + { + "epoch": 1.3227836683971057, + "grad_norm": 1.7694858312606812, + "learning_rate": 5.432062128078679e-06, + "loss": 3.1855, + "step": 16500 + }, + { + "epoch": 1.3228638431781283, + "grad_norm": 1.9247499704360962, + "learning_rate": 5.43090710364621e-06, + "loss": 3.1652, + "step": 16501 + }, + { + "epoch": 1.322944017959151, + "grad_norm": 1.7323822975158691, + "learning_rate": 5.4297521562470335e-06, + "loss": 3.1501, + "step": 16502 + }, + { + "epoch": 1.3230241927401736, + "grad_norm": 1.7973755598068237, + "learning_rate": 5.428597285900625e-06, + "loss": 3.257, + "step": 16503 + }, + { + "epoch": 1.3231043675211962, + "grad_norm": 1.9928299188613892, + "learning_rate": 5.427442492626455e-06, + "loss": 3.218, + "step": 16504 + }, + { + "epoch": 1.3231845423022188, + "grad_norm": 1.8067429065704346, + "learning_rate": 5.426287776443992e-06, + "loss": 3.2614, + "step": 16505 + }, + { + "epoch": 1.3232647170832414, + "grad_norm": 1.741604208946228, + "learning_rate": 5.425133137372711e-06, + "loss": 3.1558, + "step": 16506 + }, + { + "epoch": 1.323344891864264, + "grad_norm": 1.860226035118103, + "learning_rate": 5.423978575432067e-06, + "loss": 3.2316, + "step": 16507 + }, + { + "epoch": 1.3234250666452867, + "grad_norm": 2.0346145629882812, + "learning_rate": 5.422824090641533e-06, + "loss": 3.2133, + "step": 16508 + }, + { + "epoch": 1.3235052414263093, + "grad_norm": 0.663159966468811, + "learning_rate": 5.421669683020571e-06, + "loss": 2.4672, + "step": 16509 + }, + { + "epoch": 1.323585416207332, + "grad_norm": 1.8505903482437134, + "learning_rate": 5.420515352588645e-06, + "loss": 3.1644, + "step": 16510 + }, + { + "epoch": 1.3236655909883546, + "grad_norm": 1.8800920248031616, + "learning_rate": 5.419361099365221e-06, + "loss": 3.2471, + "step": 16511 + }, + { + "epoch": 1.3237457657693772, + "grad_norm": 0.6619951725006104, + "learning_rate": 5.418206923369748e-06, + "loss": 2.4596, + "step": 16512 + }, + { + "epoch": 1.3238259405503998, + "grad_norm": 1.791002631187439, + "learning_rate": 5.4170528246216935e-06, + "loss": 3.1592, + "step": 16513 + }, + { + "epoch": 1.3239061153314224, + "grad_norm": 1.8448700904846191, + "learning_rate": 5.4158988031405136e-06, + "loss": 3.2241, + "step": 16514 + }, + { + "epoch": 1.323986290112445, + "grad_norm": 1.7410529851913452, + "learning_rate": 5.414744858945663e-06, + "loss": 3.2211, + "step": 16515 + }, + { + "epoch": 1.3240664648934677, + "grad_norm": 0.6920922994613647, + "learning_rate": 5.4135909920566035e-06, + "loss": 2.4768, + "step": 16516 + }, + { + "epoch": 1.3241466396744903, + "grad_norm": 1.8397992849349976, + "learning_rate": 5.412437202492778e-06, + "loss": 3.0375, + "step": 16517 + }, + { + "epoch": 1.3242268144555132, + "grad_norm": 1.7002657651901245, + "learning_rate": 5.411283490273644e-06, + "loss": 3.1764, + "step": 16518 + }, + { + "epoch": 1.3243069892365358, + "grad_norm": 1.7835590839385986, + "learning_rate": 5.410129855418654e-06, + "loss": 3.1881, + "step": 16519 + }, + { + "epoch": 1.3243871640175584, + "grad_norm": 1.7951467037200928, + "learning_rate": 5.408976297947257e-06, + "loss": 3.1114, + "step": 16520 + }, + { + "epoch": 1.324467338798581, + "grad_norm": 0.6648241281509399, + "learning_rate": 5.407822817878906e-06, + "loss": 2.4744, + "step": 16521 + }, + { + "epoch": 1.3245475135796037, + "grad_norm": 1.9056663513183594, + "learning_rate": 5.406669415233039e-06, + "loss": 3.1217, + "step": 16522 + }, + { + "epoch": 1.3246276883606263, + "grad_norm": 1.856740951538086, + "learning_rate": 5.405516090029108e-06, + "loss": 3.1676, + "step": 16523 + }, + { + "epoch": 1.324707863141649, + "grad_norm": 1.8369604349136353, + "learning_rate": 5.404362842286555e-06, + "loss": 3.2318, + "step": 16524 + }, + { + "epoch": 1.3247880379226715, + "grad_norm": 1.8732385635375977, + "learning_rate": 5.403209672024826e-06, + "loss": 3.2933, + "step": 16525 + }, + { + "epoch": 1.3248682127036941, + "grad_norm": 1.997860312461853, + "learning_rate": 5.402056579263366e-06, + "loss": 3.2751, + "step": 16526 + }, + { + "epoch": 1.3249483874847168, + "grad_norm": 1.8485331535339355, + "learning_rate": 5.400903564021607e-06, + "loss": 3.23, + "step": 16527 + }, + { + "epoch": 1.3250285622657394, + "grad_norm": 1.6855387687683105, + "learning_rate": 5.399750626318992e-06, + "loss": 3.2089, + "step": 16528 + }, + { + "epoch": 1.325108737046762, + "grad_norm": 1.850521206855774, + "learning_rate": 5.398597766174962e-06, + "loss": 3.2352, + "step": 16529 + }, + { + "epoch": 1.3251889118277846, + "grad_norm": 1.7745386362075806, + "learning_rate": 5.39744498360895e-06, + "loss": 3.2117, + "step": 16530 + }, + { + "epoch": 1.3252690866088073, + "grad_norm": 1.7840986251831055, + "learning_rate": 5.396292278640399e-06, + "loss": 3.2307, + "step": 16531 + }, + { + "epoch": 1.32534926138983, + "grad_norm": 1.9686675071716309, + "learning_rate": 5.395139651288734e-06, + "loss": 3.1705, + "step": 16532 + }, + { + "epoch": 1.3254294361708525, + "grad_norm": 1.722177505493164, + "learning_rate": 5.3939871015733905e-06, + "loss": 3.1678, + "step": 16533 + }, + { + "epoch": 1.3255096109518751, + "grad_norm": 1.7571614980697632, + "learning_rate": 5.392834629513801e-06, + "loss": 3.0818, + "step": 16534 + }, + { + "epoch": 1.3255897857328978, + "grad_norm": 2.234135866165161, + "learning_rate": 5.3916822351293974e-06, + "loss": 3.0596, + "step": 16535 + }, + { + "epoch": 1.3256699605139204, + "grad_norm": 1.7473771572113037, + "learning_rate": 5.390529918439611e-06, + "loss": 3.127, + "step": 16536 + }, + { + "epoch": 1.325750135294943, + "grad_norm": 1.7330138683319092, + "learning_rate": 5.389377679463861e-06, + "loss": 3.2327, + "step": 16537 + }, + { + "epoch": 1.3258303100759656, + "grad_norm": 1.729272723197937, + "learning_rate": 5.388225518221577e-06, + "loss": 3.2303, + "step": 16538 + }, + { + "epoch": 1.3259104848569883, + "grad_norm": 1.8408249616622925, + "learning_rate": 5.387073434732188e-06, + "loss": 3.1975, + "step": 16539 + }, + { + "epoch": 1.3259906596380109, + "grad_norm": 1.652229905128479, + "learning_rate": 5.385921429015113e-06, + "loss": 3.1257, + "step": 16540 + }, + { + "epoch": 1.3260708344190335, + "grad_norm": 1.8906689882278442, + "learning_rate": 5.384769501089782e-06, + "loss": 3.2023, + "step": 16541 + }, + { + "epoch": 1.3261510092000561, + "grad_norm": 1.8108075857162476, + "learning_rate": 5.383617650975606e-06, + "loss": 3.145, + "step": 16542 + }, + { + "epoch": 1.3262311839810788, + "grad_norm": 1.7738008499145508, + "learning_rate": 5.382465878692009e-06, + "loss": 3.1257, + "step": 16543 + }, + { + "epoch": 1.3263113587621014, + "grad_norm": 1.688184142112732, + "learning_rate": 5.3813141842584105e-06, + "loss": 3.1826, + "step": 16544 + }, + { + "epoch": 1.326391533543124, + "grad_norm": 1.720308780670166, + "learning_rate": 5.380162567694227e-06, + "loss": 3.1392, + "step": 16545 + }, + { + "epoch": 1.3264717083241466, + "grad_norm": 1.8068126440048218, + "learning_rate": 5.379011029018875e-06, + "loss": 3.1608, + "step": 16546 + }, + { + "epoch": 1.3265518831051692, + "grad_norm": 1.7418948411941528, + "learning_rate": 5.377859568251771e-06, + "loss": 3.2089, + "step": 16547 + }, + { + "epoch": 1.3266320578861919, + "grad_norm": 1.7786129713058472, + "learning_rate": 5.376708185412323e-06, + "loss": 3.1735, + "step": 16548 + }, + { + "epoch": 1.3267122326672145, + "grad_norm": 1.7558425664901733, + "learning_rate": 5.375556880519945e-06, + "loss": 3.1238, + "step": 16549 + }, + { + "epoch": 1.3267924074482371, + "grad_norm": 1.7952066659927368, + "learning_rate": 5.374405653594048e-06, + "loss": 3.1521, + "step": 16550 + }, + { + "epoch": 1.3268725822292597, + "grad_norm": 1.8177012205123901, + "learning_rate": 5.3732545046540416e-06, + "loss": 3.3001, + "step": 16551 + }, + { + "epoch": 1.3269527570102824, + "grad_norm": 1.9274382591247559, + "learning_rate": 5.3721034337193375e-06, + "loss": 3.1102, + "step": 16552 + }, + { + "epoch": 1.327032931791305, + "grad_norm": 1.7245962619781494, + "learning_rate": 5.370952440809335e-06, + "loss": 3.2159, + "step": 16553 + }, + { + "epoch": 1.3271131065723276, + "grad_norm": 1.7424640655517578, + "learning_rate": 5.369801525943443e-06, + "loss": 3.2128, + "step": 16554 + }, + { + "epoch": 1.3271932813533502, + "grad_norm": 1.836621642112732, + "learning_rate": 5.368650689141065e-06, + "loss": 3.1381, + "step": 16555 + }, + { + "epoch": 1.3272734561343729, + "grad_norm": 1.8474332094192505, + "learning_rate": 5.367499930421605e-06, + "loss": 3.1388, + "step": 16556 + }, + { + "epoch": 1.3273536309153955, + "grad_norm": 1.7415498495101929, + "learning_rate": 5.366349249804465e-06, + "loss": 3.0874, + "step": 16557 + }, + { + "epoch": 1.3274338056964181, + "grad_norm": 1.7857303619384766, + "learning_rate": 5.365198647309042e-06, + "loss": 3.1371, + "step": 16558 + }, + { + "epoch": 1.3275139804774407, + "grad_norm": 1.8017076253890991, + "learning_rate": 5.364048122954736e-06, + "loss": 3.1287, + "step": 16559 + }, + { + "epoch": 1.3275941552584634, + "grad_norm": 1.7172507047653198, + "learning_rate": 5.362897676760944e-06, + "loss": 3.2136, + "step": 16560 + }, + { + "epoch": 1.327674330039486, + "grad_norm": 1.6834280490875244, + "learning_rate": 5.361747308747064e-06, + "loss": 3.2479, + "step": 16561 + }, + { + "epoch": 1.3277545048205086, + "grad_norm": 1.6617592573165894, + "learning_rate": 5.360597018932493e-06, + "loss": 3.1597, + "step": 16562 + }, + { + "epoch": 1.3278346796015312, + "grad_norm": 2.009676694869995, + "learning_rate": 5.359446807336617e-06, + "loss": 3.2558, + "step": 16563 + }, + { + "epoch": 1.3279148543825539, + "grad_norm": 1.7842110395431519, + "learning_rate": 5.358296673978833e-06, + "loss": 3.1236, + "step": 16564 + }, + { + "epoch": 1.3279950291635765, + "grad_norm": 1.739343285560608, + "learning_rate": 5.35714661887853e-06, + "loss": 3.1321, + "step": 16565 + }, + { + "epoch": 1.328075203944599, + "grad_norm": 1.7667473554611206, + "learning_rate": 5.355996642055099e-06, + "loss": 3.0748, + "step": 16566 + }, + { + "epoch": 1.3281553787256217, + "grad_norm": 1.7631385326385498, + "learning_rate": 5.3548467435279326e-06, + "loss": 3.0499, + "step": 16567 + }, + { + "epoch": 1.3282355535066444, + "grad_norm": 2.131483793258667, + "learning_rate": 5.353696923316409e-06, + "loss": 3.1077, + "step": 16568 + }, + { + "epoch": 1.3283157282876672, + "grad_norm": 1.9835115671157837, + "learning_rate": 5.3525471814399175e-06, + "loss": 3.2179, + "step": 16569 + }, + { + "epoch": 1.3283959030686898, + "grad_norm": 1.7502721548080444, + "learning_rate": 5.351397517917843e-06, + "loss": 3.0381, + "step": 16570 + }, + { + "epoch": 1.3284760778497124, + "grad_norm": 1.8163448572158813, + "learning_rate": 5.350247932769567e-06, + "loss": 3.1386, + "step": 16571 + }, + { + "epoch": 1.328556252630735, + "grad_norm": 1.8574894666671753, + "learning_rate": 5.349098426014478e-06, + "loss": 3.267, + "step": 16572 + }, + { + "epoch": 1.3286364274117577, + "grad_norm": 1.9384007453918457, + "learning_rate": 5.347948997671944e-06, + "loss": 3.3577, + "step": 16573 + }, + { + "epoch": 1.3287166021927803, + "grad_norm": 1.7429949045181274, + "learning_rate": 5.346799647761352e-06, + "loss": 3.2368, + "step": 16574 + }, + { + "epoch": 1.328796776973803, + "grad_norm": 1.7298048734664917, + "learning_rate": 5.345650376302078e-06, + "loss": 3.2183, + "step": 16575 + }, + { + "epoch": 1.3288769517548256, + "grad_norm": 1.7339487075805664, + "learning_rate": 5.344501183313497e-06, + "loss": 3.2678, + "step": 16576 + }, + { + "epoch": 1.3289571265358482, + "grad_norm": 1.9516384601593018, + "learning_rate": 5.343352068814991e-06, + "loss": 3.121, + "step": 16577 + }, + { + "epoch": 1.3290373013168708, + "grad_norm": 1.8856871128082275, + "learning_rate": 5.3422030328259235e-06, + "loss": 3.2684, + "step": 16578 + }, + { + "epoch": 1.3291174760978934, + "grad_norm": 1.769913673400879, + "learning_rate": 5.34105407536567e-06, + "loss": 3.127, + "step": 16579 + }, + { + "epoch": 1.329197650878916, + "grad_norm": 1.7746740579605103, + "learning_rate": 5.339905196453603e-06, + "loss": 3.2319, + "step": 16580 + }, + { + "epoch": 1.3292778256599387, + "grad_norm": 1.7973828315734863, + "learning_rate": 5.338756396109093e-06, + "loss": 3.2137, + "step": 16581 + }, + { + "epoch": 1.3293580004409613, + "grad_norm": 1.7859944105148315, + "learning_rate": 5.337607674351511e-06, + "loss": 3.2573, + "step": 16582 + }, + { + "epoch": 1.329438175221984, + "grad_norm": 1.734453558921814, + "learning_rate": 5.336459031200216e-06, + "loss": 3.1862, + "step": 16583 + }, + { + "epoch": 1.3295183500030066, + "grad_norm": 1.8572940826416016, + "learning_rate": 5.335310466674579e-06, + "loss": 3.1782, + "step": 16584 + }, + { + "epoch": 1.3295985247840292, + "grad_norm": 1.7798552513122559, + "learning_rate": 5.334161980793964e-06, + "loss": 3.1886, + "step": 16585 + }, + { + "epoch": 1.3296786995650518, + "grad_norm": 1.7541084289550781, + "learning_rate": 5.333013573577732e-06, + "loss": 3.3215, + "step": 16586 + }, + { + "epoch": 1.3297588743460744, + "grad_norm": 1.6460236310958862, + "learning_rate": 5.331865245045252e-06, + "loss": 3.1351, + "step": 16587 + }, + { + "epoch": 1.329839049127097, + "grad_norm": 1.7184921503067017, + "learning_rate": 5.3307169952158735e-06, + "loss": 3.1824, + "step": 16588 + }, + { + "epoch": 1.3299192239081197, + "grad_norm": 1.7804194688796997, + "learning_rate": 5.329568824108962e-06, + "loss": 3.1973, + "step": 16589 + }, + { + "epoch": 1.3299993986891423, + "grad_norm": 0.6966257691383362, + "learning_rate": 5.328420731743873e-06, + "loss": 2.4679, + "step": 16590 + }, + { + "epoch": 1.330079573470165, + "grad_norm": 1.930728554725647, + "learning_rate": 5.327272718139964e-06, + "loss": 3.1461, + "step": 16591 + }, + { + "epoch": 1.3301597482511875, + "grad_norm": 1.8730907440185547, + "learning_rate": 5.326124783316595e-06, + "loss": 3.2504, + "step": 16592 + }, + { + "epoch": 1.3302399230322102, + "grad_norm": 0.663902223110199, + "learning_rate": 5.32497692729311e-06, + "loss": 2.4447, + "step": 16593 + }, + { + "epoch": 1.3303200978132328, + "grad_norm": 1.6642072200775146, + "learning_rate": 5.323829150088866e-06, + "loss": 3.1744, + "step": 16594 + }, + { + "epoch": 1.3304002725942554, + "grad_norm": 1.7346022129058838, + "learning_rate": 5.322681451723216e-06, + "loss": 3.1538, + "step": 16595 + }, + { + "epoch": 1.330480447375278, + "grad_norm": 1.8347969055175781, + "learning_rate": 5.321533832215508e-06, + "loss": 3.2296, + "step": 16596 + }, + { + "epoch": 1.3305606221563007, + "grad_norm": 1.8295296430587769, + "learning_rate": 5.320386291585093e-06, + "loss": 3.2335, + "step": 16597 + }, + { + "epoch": 1.3306407969373233, + "grad_norm": 1.817649006843567, + "learning_rate": 5.319238829851314e-06, + "loss": 3.3182, + "step": 16598 + }, + { + "epoch": 1.330720971718346, + "grad_norm": 1.6678818464279175, + "learning_rate": 5.318091447033518e-06, + "loss": 3.175, + "step": 16599 + }, + { + "epoch": 1.3308011464993688, + "grad_norm": 1.7527776956558228, + "learning_rate": 5.31694414315105e-06, + "loss": 3.1296, + "step": 16600 + }, + { + "epoch": 1.3308813212803914, + "grad_norm": 1.9887285232543945, + "learning_rate": 5.3157969182232525e-06, + "loss": 3.1802, + "step": 16601 + }, + { + "epoch": 1.330961496061414, + "grad_norm": 1.7925491333007812, + "learning_rate": 5.314649772269474e-06, + "loss": 3.1589, + "step": 16602 + }, + { + "epoch": 1.3310416708424366, + "grad_norm": 1.7912253141403198, + "learning_rate": 5.313502705309043e-06, + "loss": 3.1753, + "step": 16603 + }, + { + "epoch": 1.3311218456234593, + "grad_norm": 1.7153406143188477, + "learning_rate": 5.312355717361306e-06, + "loss": 3.1678, + "step": 16604 + }, + { + "epoch": 1.3312020204044819, + "grad_norm": 1.8988224267959595, + "learning_rate": 5.311208808445599e-06, + "loss": 3.1957, + "step": 16605 + }, + { + "epoch": 1.3312821951855045, + "grad_norm": 1.758042573928833, + "learning_rate": 5.31006197858126e-06, + "loss": 3.1907, + "step": 16606 + }, + { + "epoch": 1.3313623699665271, + "grad_norm": 1.7432712316513062, + "learning_rate": 5.308915227787626e-06, + "loss": 3.1619, + "step": 16607 + }, + { + "epoch": 1.3314425447475497, + "grad_norm": 1.9301525354385376, + "learning_rate": 5.307768556084023e-06, + "loss": 3.3669, + "step": 16608 + }, + { + "epoch": 1.3315227195285724, + "grad_norm": 1.6583573818206787, + "learning_rate": 5.306621963489791e-06, + "loss": 3.0293, + "step": 16609 + }, + { + "epoch": 1.331602894309595, + "grad_norm": 1.8086695671081543, + "learning_rate": 5.3054754500242564e-06, + "loss": 3.2056, + "step": 16610 + }, + { + "epoch": 1.3316830690906176, + "grad_norm": 1.7628529071807861, + "learning_rate": 5.3043290157067525e-06, + "loss": 3.1721, + "step": 16611 + }, + { + "epoch": 1.3317632438716402, + "grad_norm": 0.6860686540603638, + "learning_rate": 5.30318266055661e-06, + "loss": 2.4772, + "step": 16612 + }, + { + "epoch": 1.3318434186526629, + "grad_norm": 1.7494516372680664, + "learning_rate": 5.3020363845931484e-06, + "loss": 3.2047, + "step": 16613 + }, + { + "epoch": 1.3319235934336855, + "grad_norm": 1.8030588626861572, + "learning_rate": 5.3008901878356975e-06, + "loss": 3.2573, + "step": 16614 + }, + { + "epoch": 1.3320037682147081, + "grad_norm": 1.7008055448532104, + "learning_rate": 5.299744070303583e-06, + "loss": 3.0763, + "step": 16615 + }, + { + "epoch": 1.3320839429957307, + "grad_norm": 1.8027582168579102, + "learning_rate": 5.298598032016126e-06, + "loss": 3.1878, + "step": 16616 + }, + { + "epoch": 1.3321641177767534, + "grad_norm": 1.7393364906311035, + "learning_rate": 5.297452072992653e-06, + "loss": 3.2299, + "step": 16617 + }, + { + "epoch": 1.332244292557776, + "grad_norm": 1.7529531717300415, + "learning_rate": 5.296306193252477e-06, + "loss": 3.1597, + "step": 16618 + }, + { + "epoch": 1.3323244673387986, + "grad_norm": 1.7415622472763062, + "learning_rate": 5.295160392814922e-06, + "loss": 3.1718, + "step": 16619 + }, + { + "epoch": 1.3324046421198212, + "grad_norm": 1.7211425304412842, + "learning_rate": 5.294014671699304e-06, + "loss": 3.1676, + "step": 16620 + }, + { + "epoch": 1.3324848169008439, + "grad_norm": 1.7680306434631348, + "learning_rate": 5.292869029924939e-06, + "loss": 3.3146, + "step": 16621 + }, + { + "epoch": 1.3325649916818665, + "grad_norm": 1.7050435543060303, + "learning_rate": 5.291723467511147e-06, + "loss": 3.0753, + "step": 16622 + }, + { + "epoch": 1.332645166462889, + "grad_norm": 1.7770617008209229, + "learning_rate": 5.290577984477236e-06, + "loss": 3.2449, + "step": 16623 + }, + { + "epoch": 1.3327253412439117, + "grad_norm": 1.7992969751358032, + "learning_rate": 5.289432580842519e-06, + "loss": 3.2432, + "step": 16624 + }, + { + "epoch": 1.3328055160249344, + "grad_norm": 1.8527699708938599, + "learning_rate": 5.288287256626307e-06, + "loss": 3.2526, + "step": 16625 + }, + { + "epoch": 1.332885690805957, + "grad_norm": 1.8499640226364136, + "learning_rate": 5.2871420118479145e-06, + "loss": 3.245, + "step": 16626 + }, + { + "epoch": 1.3329658655869796, + "grad_norm": 1.7997890710830688, + "learning_rate": 5.285996846526648e-06, + "loss": 3.1306, + "step": 16627 + }, + { + "epoch": 1.3330460403680022, + "grad_norm": 1.746569037437439, + "learning_rate": 5.284851760681812e-06, + "loss": 3.2766, + "step": 16628 + }, + { + "epoch": 1.3331262151490249, + "grad_norm": 1.8558341264724731, + "learning_rate": 5.283706754332712e-06, + "loss": 3.1764, + "step": 16629 + }, + { + "epoch": 1.3332063899300475, + "grad_norm": 1.7058757543563843, + "learning_rate": 5.282561827498653e-06, + "loss": 3.1133, + "step": 16630 + }, + { + "epoch": 1.33328656471107, + "grad_norm": 1.8378896713256836, + "learning_rate": 5.281416980198941e-06, + "loss": 3.2556, + "step": 16631 + }, + { + "epoch": 1.3333667394920927, + "grad_norm": 1.7844476699829102, + "learning_rate": 5.280272212452879e-06, + "loss": 3.1932, + "step": 16632 + }, + { + "epoch": 1.3334469142731153, + "grad_norm": 1.9969513416290283, + "learning_rate": 5.279127524279759e-06, + "loss": 3.3353, + "step": 16633 + }, + { + "epoch": 1.333527089054138, + "grad_norm": 1.7152262926101685, + "learning_rate": 5.277982915698887e-06, + "loss": 3.2183, + "step": 16634 + }, + { + "epoch": 1.3336072638351606, + "grad_norm": 1.8507636785507202, + "learning_rate": 5.276838386729557e-06, + "loss": 3.196, + "step": 16635 + }, + { + "epoch": 1.3336874386161832, + "grad_norm": 1.730666160583496, + "learning_rate": 5.275693937391069e-06, + "loss": 3.203, + "step": 16636 + }, + { + "epoch": 1.3337676133972058, + "grad_norm": 1.7559363842010498, + "learning_rate": 5.274549567702719e-06, + "loss": 3.1856, + "step": 16637 + }, + { + "epoch": 1.3338477881782285, + "grad_norm": 1.8393412828445435, + "learning_rate": 5.273405277683794e-06, + "loss": 3.1764, + "step": 16638 + }, + { + "epoch": 1.333927962959251, + "grad_norm": 0.6978288888931274, + "learning_rate": 5.27226106735359e-06, + "loss": 2.4625, + "step": 16639 + }, + { + "epoch": 1.3340081377402737, + "grad_norm": 1.7271593809127808, + "learning_rate": 5.2711169367314e-06, + "loss": 3.2532, + "step": 16640 + }, + { + "epoch": 1.3340883125212963, + "grad_norm": 1.820188283920288, + "learning_rate": 5.2699728858365096e-06, + "loss": 3.1435, + "step": 16641 + }, + { + "epoch": 1.334168487302319, + "grad_norm": 1.8259544372558594, + "learning_rate": 5.268828914688213e-06, + "loss": 3.1487, + "step": 16642 + }, + { + "epoch": 1.3342486620833416, + "grad_norm": 1.7561328411102295, + "learning_rate": 5.267685023305791e-06, + "loss": 3.121, + "step": 16643 + }, + { + "epoch": 1.3343288368643642, + "grad_norm": 1.7373815774917603, + "learning_rate": 5.266541211708531e-06, + "loss": 3.1381, + "step": 16644 + }, + { + "epoch": 1.3344090116453868, + "grad_norm": 1.7306876182556152, + "learning_rate": 5.265397479915719e-06, + "loss": 3.1007, + "step": 16645 + }, + { + "epoch": 1.3344891864264095, + "grad_norm": 1.7908220291137695, + "learning_rate": 5.264253827946636e-06, + "loss": 3.1781, + "step": 16646 + }, + { + "epoch": 1.334569361207432, + "grad_norm": 1.7517682313919067, + "learning_rate": 5.263110255820568e-06, + "loss": 3.2145, + "step": 16647 + }, + { + "epoch": 1.3346495359884547, + "grad_norm": 1.8093934059143066, + "learning_rate": 5.261966763556788e-06, + "loss": 3.2385, + "step": 16648 + }, + { + "epoch": 1.3347297107694773, + "grad_norm": 1.7693825960159302, + "learning_rate": 5.26082335117458e-06, + "loss": 3.0988, + "step": 16649 + }, + { + "epoch": 1.3348098855505, + "grad_norm": 1.8196240663528442, + "learning_rate": 5.2596800186932185e-06, + "loss": 3.1743, + "step": 16650 + }, + { + "epoch": 1.3348900603315228, + "grad_norm": 1.7267813682556152, + "learning_rate": 5.258536766131981e-06, + "loss": 3.1689, + "step": 16651 + }, + { + "epoch": 1.3349702351125454, + "grad_norm": 1.7530157566070557, + "learning_rate": 5.257393593510147e-06, + "loss": 3.1381, + "step": 16652 + }, + { + "epoch": 1.335050409893568, + "grad_norm": 1.7397024631500244, + "learning_rate": 5.256250500846981e-06, + "loss": 3.1801, + "step": 16653 + }, + { + "epoch": 1.3351305846745907, + "grad_norm": 1.837103009223938, + "learning_rate": 5.25510748816176e-06, + "loss": 3.299, + "step": 16654 + }, + { + "epoch": 1.3352107594556133, + "grad_norm": 1.7992451190948486, + "learning_rate": 5.253964555473755e-06, + "loss": 3.1528, + "step": 16655 + }, + { + "epoch": 1.335290934236636, + "grad_norm": 1.8072274923324585, + "learning_rate": 5.252821702802235e-06, + "loss": 3.1707, + "step": 16656 + }, + { + "epoch": 1.3353711090176585, + "grad_norm": 0.7293029427528381, + "learning_rate": 5.251678930166471e-06, + "loss": 2.4585, + "step": 16657 + }, + { + "epoch": 1.3354512837986812, + "grad_norm": 1.9082088470458984, + "learning_rate": 5.250536237585724e-06, + "loss": 3.1598, + "step": 16658 + }, + { + "epoch": 1.3355314585797038, + "grad_norm": 1.7710423469543457, + "learning_rate": 5.249393625079262e-06, + "loss": 3.2662, + "step": 16659 + }, + { + "epoch": 1.3356116333607264, + "grad_norm": 1.8704181909561157, + "learning_rate": 5.2482510926663475e-06, + "loss": 3.151, + "step": 16660 + }, + { + "epoch": 1.335691808141749, + "grad_norm": 1.714920997619629, + "learning_rate": 5.247108640366248e-06, + "loss": 3.1484, + "step": 16661 + }, + { + "epoch": 1.3357719829227717, + "grad_norm": 1.7921984195709229, + "learning_rate": 5.245966268198223e-06, + "loss": 3.17, + "step": 16662 + }, + { + "epoch": 1.3358521577037943, + "grad_norm": 1.868249535560608, + "learning_rate": 5.2448239761815275e-06, + "loss": 3.1425, + "step": 16663 + }, + { + "epoch": 1.335932332484817, + "grad_norm": 1.8383493423461914, + "learning_rate": 5.243681764335426e-06, + "loss": 3.2524, + "step": 16664 + }, + { + "epoch": 1.3360125072658395, + "grad_norm": 1.858465313911438, + "learning_rate": 5.242539632679171e-06, + "loss": 3.2469, + "step": 16665 + }, + { + "epoch": 1.3360926820468622, + "grad_norm": 1.8416234254837036, + "learning_rate": 5.241397581232023e-06, + "loss": 3.186, + "step": 16666 + }, + { + "epoch": 1.3361728568278848, + "grad_norm": 1.7900089025497437, + "learning_rate": 5.2402556100132384e-06, + "loss": 3.1807, + "step": 16667 + }, + { + "epoch": 1.3362530316089074, + "grad_norm": 0.6757766008377075, + "learning_rate": 5.239113719042062e-06, + "loss": 2.4897, + "step": 16668 + }, + { + "epoch": 1.33633320638993, + "grad_norm": 1.728776454925537, + "learning_rate": 5.237971908337751e-06, + "loss": 3.1843, + "step": 16669 + }, + { + "epoch": 1.3364133811709527, + "grad_norm": 1.7763019800186157, + "learning_rate": 5.236830177919554e-06, + "loss": 3.1121, + "step": 16670 + }, + { + "epoch": 1.3364935559519753, + "grad_norm": 1.75749671459198, + "learning_rate": 5.235688527806724e-06, + "loss": 3.1525, + "step": 16671 + }, + { + "epoch": 1.336573730732998, + "grad_norm": 1.7757353782653809, + "learning_rate": 5.234546958018507e-06, + "loss": 3.1979, + "step": 16672 + }, + { + "epoch": 1.3366539055140205, + "grad_norm": 1.781983733177185, + "learning_rate": 5.233405468574147e-06, + "loss": 3.1714, + "step": 16673 + }, + { + "epoch": 1.3367340802950431, + "grad_norm": 1.8862578868865967, + "learning_rate": 5.232264059492891e-06, + "loss": 3.1935, + "step": 16674 + }, + { + "epoch": 1.3368142550760658, + "grad_norm": 1.8157142400741577, + "learning_rate": 5.231122730793982e-06, + "loss": 3.1338, + "step": 16675 + }, + { + "epoch": 1.3368944298570884, + "grad_norm": 1.802474021911621, + "learning_rate": 5.229981482496663e-06, + "loss": 3.289, + "step": 16676 + }, + { + "epoch": 1.336974604638111, + "grad_norm": 1.8403180837631226, + "learning_rate": 5.228840314620179e-06, + "loss": 3.1601, + "step": 16677 + }, + { + "epoch": 1.3370547794191336, + "grad_norm": 1.8308923244476318, + "learning_rate": 5.227699227183763e-06, + "loss": 3.239, + "step": 16678 + }, + { + "epoch": 1.3371349542001563, + "grad_norm": 1.807723045349121, + "learning_rate": 5.2265582202066546e-06, + "loss": 3.1312, + "step": 16679 + }, + { + "epoch": 1.337215128981179, + "grad_norm": 1.6901742219924927, + "learning_rate": 5.225417293708092e-06, + "loss": 3.1971, + "step": 16680 + }, + { + "epoch": 1.3372953037622017, + "grad_norm": 1.8973910808563232, + "learning_rate": 5.224276447707313e-06, + "loss": 3.277, + "step": 16681 + }, + { + "epoch": 1.3373754785432244, + "grad_norm": 1.8268663883209229, + "learning_rate": 5.223135682223552e-06, + "loss": 3.2397, + "step": 16682 + }, + { + "epoch": 1.337455653324247, + "grad_norm": 1.8297549486160278, + "learning_rate": 5.2219949972760376e-06, + "loss": 3.1959, + "step": 16683 + }, + { + "epoch": 1.3375358281052696, + "grad_norm": 1.7556947469711304, + "learning_rate": 5.2208543928840026e-06, + "loss": 3.1064, + "step": 16684 + }, + { + "epoch": 1.3376160028862922, + "grad_norm": 1.8164726495742798, + "learning_rate": 5.219713869066678e-06, + "loss": 3.1398, + "step": 16685 + }, + { + "epoch": 1.3376961776673149, + "grad_norm": 0.6829904317855835, + "learning_rate": 5.218573425843294e-06, + "loss": 2.4771, + "step": 16686 + }, + { + "epoch": 1.3377763524483375, + "grad_norm": 1.7812745571136475, + "learning_rate": 5.217433063233079e-06, + "loss": 3.1954, + "step": 16687 + }, + { + "epoch": 1.33785652722936, + "grad_norm": 1.7524336576461792, + "learning_rate": 5.216292781255255e-06, + "loss": 3.1273, + "step": 16688 + }, + { + "epoch": 1.3379367020103827, + "grad_norm": 1.8005073070526123, + "learning_rate": 5.215152579929048e-06, + "loss": 3.0546, + "step": 16689 + }, + { + "epoch": 1.3380168767914054, + "grad_norm": 1.8676851987838745, + "learning_rate": 5.214012459273683e-06, + "loss": 3.101, + "step": 16690 + }, + { + "epoch": 1.338097051572428, + "grad_norm": 1.745246171951294, + "learning_rate": 5.212872419308379e-06, + "loss": 3.1799, + "step": 16691 + }, + { + "epoch": 1.3381772263534506, + "grad_norm": 1.712346076965332, + "learning_rate": 5.211732460052365e-06, + "loss": 3.2585, + "step": 16692 + }, + { + "epoch": 1.3382574011344732, + "grad_norm": 1.8404158353805542, + "learning_rate": 5.210592581524849e-06, + "loss": 3.1832, + "step": 16693 + }, + { + "epoch": 1.3383375759154958, + "grad_norm": 1.7339880466461182, + "learning_rate": 5.209452783745055e-06, + "loss": 3.1174, + "step": 16694 + }, + { + "epoch": 1.3384177506965185, + "grad_norm": 1.918200969696045, + "learning_rate": 5.208313066732198e-06, + "loss": 3.2208, + "step": 16695 + }, + { + "epoch": 1.338497925477541, + "grad_norm": 1.685592532157898, + "learning_rate": 5.207173430505494e-06, + "loss": 3.0487, + "step": 16696 + }, + { + "epoch": 1.3385781002585637, + "grad_norm": 1.7526836395263672, + "learning_rate": 5.20603387508416e-06, + "loss": 3.1644, + "step": 16697 + }, + { + "epoch": 1.3386582750395863, + "grad_norm": 1.8661166429519653, + "learning_rate": 5.204894400487404e-06, + "loss": 3.2855, + "step": 16698 + }, + { + "epoch": 1.338738449820609, + "grad_norm": 1.8244267702102661, + "learning_rate": 5.203755006734436e-06, + "loss": 3.1512, + "step": 16699 + }, + { + "epoch": 1.3388186246016316, + "grad_norm": 1.7988862991333008, + "learning_rate": 5.202615693844468e-06, + "loss": 3.1371, + "step": 16700 + }, + { + "epoch": 1.3388987993826542, + "grad_norm": 1.7725017070770264, + "learning_rate": 5.2014764618367095e-06, + "loss": 3.2755, + "step": 16701 + }, + { + "epoch": 1.3389789741636768, + "grad_norm": 1.8370130062103271, + "learning_rate": 5.200337310730371e-06, + "loss": 3.2887, + "step": 16702 + }, + { + "epoch": 1.3390591489446995, + "grad_norm": 1.8213080167770386, + "learning_rate": 5.1991982405446495e-06, + "loss": 3.1385, + "step": 16703 + }, + { + "epoch": 1.339139323725722, + "grad_norm": 1.770416021347046, + "learning_rate": 5.1980592512987535e-06, + "loss": 3.1971, + "step": 16704 + }, + { + "epoch": 1.3392194985067447, + "grad_norm": 1.8342021703720093, + "learning_rate": 5.196920343011886e-06, + "loss": 3.1848, + "step": 16705 + }, + { + "epoch": 1.3392996732877673, + "grad_norm": 1.729973554611206, + "learning_rate": 5.195781515703251e-06, + "loss": 3.1401, + "step": 16706 + }, + { + "epoch": 1.33937984806879, + "grad_norm": 1.777725338935852, + "learning_rate": 5.194642769392047e-06, + "loss": 3.1036, + "step": 16707 + }, + { + "epoch": 1.3394600228498126, + "grad_norm": 1.8709626197814941, + "learning_rate": 5.193504104097471e-06, + "loss": 3.2091, + "step": 16708 + }, + { + "epoch": 1.3395401976308352, + "grad_norm": 1.8135982751846313, + "learning_rate": 5.19236551983872e-06, + "loss": 3.1659, + "step": 16709 + }, + { + "epoch": 1.3396203724118578, + "grad_norm": 1.7764499187469482, + "learning_rate": 5.1912270166349945e-06, + "loss": 3.2305, + "step": 16710 + }, + { + "epoch": 1.3397005471928805, + "grad_norm": 1.7202534675598145, + "learning_rate": 5.190088594505484e-06, + "loss": 3.1421, + "step": 16711 + }, + { + "epoch": 1.339780721973903, + "grad_norm": 1.7831425666809082, + "learning_rate": 5.188950253469389e-06, + "loss": 3.1335, + "step": 16712 + }, + { + "epoch": 1.3398608967549257, + "grad_norm": 1.7806705236434937, + "learning_rate": 5.187811993545894e-06, + "loss": 3.1257, + "step": 16713 + }, + { + "epoch": 1.3399410715359483, + "grad_norm": 1.751440405845642, + "learning_rate": 5.186673814754193e-06, + "loss": 3.2196, + "step": 16714 + }, + { + "epoch": 1.340021246316971, + "grad_norm": 1.8179508447647095, + "learning_rate": 5.185535717113475e-06, + "loss": 3.2161, + "step": 16715 + }, + { + "epoch": 1.3401014210979936, + "grad_norm": 1.7316546440124512, + "learning_rate": 5.184397700642927e-06, + "loss": 3.1224, + "step": 16716 + }, + { + "epoch": 1.3401815958790162, + "grad_norm": 1.8161674737930298, + "learning_rate": 5.183259765361741e-06, + "loss": 3.2318, + "step": 16717 + }, + { + "epoch": 1.3402617706600388, + "grad_norm": 1.7396713495254517, + "learning_rate": 5.182121911289095e-06, + "loss": 3.2803, + "step": 16718 + }, + { + "epoch": 1.3403419454410614, + "grad_norm": 1.69658362865448, + "learning_rate": 5.180984138444174e-06, + "loss": 3.1163, + "step": 16719 + }, + { + "epoch": 1.340422120222084, + "grad_norm": 1.7856123447418213, + "learning_rate": 5.179846446846163e-06, + "loss": 3.0994, + "step": 16720 + }, + { + "epoch": 1.3405022950031067, + "grad_norm": 1.8317301273345947, + "learning_rate": 5.178708836514242e-06, + "loss": 3.149, + "step": 16721 + }, + { + "epoch": 1.3405824697841293, + "grad_norm": 1.7214208841323853, + "learning_rate": 5.177571307467595e-06, + "loss": 3.2386, + "step": 16722 + }, + { + "epoch": 1.340662644565152, + "grad_norm": 1.745710015296936, + "learning_rate": 5.176433859725392e-06, + "loss": 3.2184, + "step": 16723 + }, + { + "epoch": 1.3407428193461746, + "grad_norm": 0.6506456732749939, + "learning_rate": 5.175296493306814e-06, + "loss": 2.4557, + "step": 16724 + }, + { + "epoch": 1.3408229941271972, + "grad_norm": 1.8658980131149292, + "learning_rate": 5.174159208231036e-06, + "loss": 3.2832, + "step": 16725 + }, + { + "epoch": 1.3409031689082198, + "grad_norm": 1.871551513671875, + "learning_rate": 5.173022004517234e-06, + "loss": 3.1433, + "step": 16726 + }, + { + "epoch": 1.3409833436892424, + "grad_norm": 1.7278996706008911, + "learning_rate": 5.171884882184583e-06, + "loss": 3.1336, + "step": 16727 + }, + { + "epoch": 1.341063518470265, + "grad_norm": 1.7711973190307617, + "learning_rate": 5.170747841252249e-06, + "loss": 3.1171, + "step": 16728 + }, + { + "epoch": 1.3411436932512877, + "grad_norm": 1.786813497543335, + "learning_rate": 5.169610881739402e-06, + "loss": 3.2058, + "step": 16729 + }, + { + "epoch": 1.3412238680323103, + "grad_norm": 1.7560237646102905, + "learning_rate": 5.1684740036652135e-06, + "loss": 3.0283, + "step": 16730 + }, + { + "epoch": 1.341304042813333, + "grad_norm": 1.7921394109725952, + "learning_rate": 5.167337207048852e-06, + "loss": 3.2183, + "step": 16731 + }, + { + "epoch": 1.3413842175943558, + "grad_norm": 1.744627594947815, + "learning_rate": 5.1662004919094855e-06, + "loss": 3.2217, + "step": 16732 + }, + { + "epoch": 1.3414643923753784, + "grad_norm": 1.7375482320785522, + "learning_rate": 5.165063858266271e-06, + "loss": 3.2045, + "step": 16733 + }, + { + "epoch": 1.341544567156401, + "grad_norm": 1.761642336845398, + "learning_rate": 5.163927306138375e-06, + "loss": 3.1567, + "step": 16734 + }, + { + "epoch": 1.3416247419374236, + "grad_norm": 1.7223453521728516, + "learning_rate": 5.162790835544962e-06, + "loss": 3.1691, + "step": 16735 + }, + { + "epoch": 1.3417049167184463, + "grad_norm": 1.7664074897766113, + "learning_rate": 5.1616544465051884e-06, + "loss": 3.1156, + "step": 16736 + }, + { + "epoch": 1.341785091499469, + "grad_norm": 1.671774983406067, + "learning_rate": 5.160518139038222e-06, + "loss": 3.2567, + "step": 16737 + }, + { + "epoch": 1.3418652662804915, + "grad_norm": 1.8431771993637085, + "learning_rate": 5.159381913163208e-06, + "loss": 3.1998, + "step": 16738 + }, + { + "epoch": 1.3419454410615141, + "grad_norm": 1.7539976835250854, + "learning_rate": 5.15824576889931e-06, + "loss": 3.2047, + "step": 16739 + }, + { + "epoch": 1.3420256158425368, + "grad_norm": 1.7711948156356812, + "learning_rate": 5.1571097062656825e-06, + "loss": 3.1855, + "step": 16740 + }, + { + "epoch": 1.3421057906235594, + "grad_norm": 1.788763403892517, + "learning_rate": 5.155973725281478e-06, + "loss": 3.0454, + "step": 16741 + }, + { + "epoch": 1.342185965404582, + "grad_norm": 1.7182128429412842, + "learning_rate": 5.1548378259658535e-06, + "loss": 3.0605, + "step": 16742 + }, + { + "epoch": 1.3422661401856046, + "grad_norm": 0.6755055785179138, + "learning_rate": 5.153702008337952e-06, + "loss": 2.4637, + "step": 16743 + }, + { + "epoch": 1.3423463149666273, + "grad_norm": 1.7559093236923218, + "learning_rate": 5.152566272416926e-06, + "loss": 3.1846, + "step": 16744 + }, + { + "epoch": 1.3424264897476499, + "grad_norm": 1.7586045265197754, + "learning_rate": 5.1514306182219246e-06, + "loss": 3.1405, + "step": 16745 + }, + { + "epoch": 1.3425066645286725, + "grad_norm": 1.8754626512527466, + "learning_rate": 5.150295045772094e-06, + "loss": 3.075, + "step": 16746 + }, + { + "epoch": 1.3425868393096951, + "grad_norm": 1.856847882270813, + "learning_rate": 5.1491595550865825e-06, + "loss": 3.2614, + "step": 16747 + }, + { + "epoch": 1.3426670140907178, + "grad_norm": 1.7780961990356445, + "learning_rate": 5.148024146184529e-06, + "loss": 3.2102, + "step": 16748 + }, + { + "epoch": 1.3427471888717404, + "grad_norm": 1.7954108715057373, + "learning_rate": 5.1468888190850785e-06, + "loss": 3.1316, + "step": 16749 + }, + { + "epoch": 1.342827363652763, + "grad_norm": 1.7896201610565186, + "learning_rate": 5.145753573807371e-06, + "loss": 3.1124, + "step": 16750 + }, + { + "epoch": 1.3429075384337856, + "grad_norm": 1.6971179246902466, + "learning_rate": 5.144618410370548e-06, + "loss": 3.1373, + "step": 16751 + }, + { + "epoch": 1.3429877132148083, + "grad_norm": 1.8468369245529175, + "learning_rate": 5.143483328793752e-06, + "loss": 3.1404, + "step": 16752 + }, + { + "epoch": 1.3430678879958309, + "grad_norm": 1.6816554069519043, + "learning_rate": 5.14234832909611e-06, + "loss": 3.1305, + "step": 16753 + }, + { + "epoch": 1.3431480627768535, + "grad_norm": 1.827882170677185, + "learning_rate": 5.1412134112967636e-06, + "loss": 3.0848, + "step": 16754 + }, + { + "epoch": 1.3432282375578761, + "grad_norm": 1.783394455909729, + "learning_rate": 5.140078575414847e-06, + "loss": 3.2648, + "step": 16755 + }, + { + "epoch": 1.3433084123388987, + "grad_norm": 0.7057132720947266, + "learning_rate": 5.138943821469493e-06, + "loss": 2.4708, + "step": 16756 + }, + { + "epoch": 1.3433885871199214, + "grad_norm": 1.7595967054367065, + "learning_rate": 5.137809149479838e-06, + "loss": 3.0897, + "step": 16757 + }, + { + "epoch": 1.343468761900944, + "grad_norm": 1.7655112743377686, + "learning_rate": 5.136674559465e-06, + "loss": 3.2029, + "step": 16758 + }, + { + "epoch": 1.3435489366819666, + "grad_norm": 1.7356877326965332, + "learning_rate": 5.135540051444118e-06, + "loss": 3.0764, + "step": 16759 + }, + { + "epoch": 1.3436291114629892, + "grad_norm": 1.7271519899368286, + "learning_rate": 5.134405625436315e-06, + "loss": 3.1489, + "step": 16760 + }, + { + "epoch": 1.3437092862440119, + "grad_norm": 1.7829844951629639, + "learning_rate": 5.133271281460719e-06, + "loss": 3.1295, + "step": 16761 + }, + { + "epoch": 1.3437894610250345, + "grad_norm": 1.8342411518096924, + "learning_rate": 5.132137019536457e-06, + "loss": 3.1584, + "step": 16762 + }, + { + "epoch": 1.3438696358060573, + "grad_norm": 1.845415472984314, + "learning_rate": 5.131002839682646e-06, + "loss": 3.1801, + "step": 16763 + }, + { + "epoch": 1.34394981058708, + "grad_norm": 1.894866704940796, + "learning_rate": 5.129868741918411e-06, + "loss": 3.2659, + "step": 16764 + }, + { + "epoch": 1.3440299853681026, + "grad_norm": 1.7171789407730103, + "learning_rate": 5.128734726262872e-06, + "loss": 3.1343, + "step": 16765 + }, + { + "epoch": 1.3441101601491252, + "grad_norm": 1.986110806465149, + "learning_rate": 5.12760079273515e-06, + "loss": 3.1136, + "step": 16766 + }, + { + "epoch": 1.3441903349301478, + "grad_norm": 1.7776468992233276, + "learning_rate": 5.126466941354364e-06, + "loss": 3.1911, + "step": 16767 + }, + { + "epoch": 1.3442705097111705, + "grad_norm": 1.9183918237686157, + "learning_rate": 5.125333172139625e-06, + "loss": 3.1997, + "step": 16768 + }, + { + "epoch": 1.344350684492193, + "grad_norm": 1.7610546350479126, + "learning_rate": 5.1241994851100505e-06, + "loss": 3.0935, + "step": 16769 + }, + { + "epoch": 1.3444308592732157, + "grad_norm": 1.7818093299865723, + "learning_rate": 5.123065880284753e-06, + "loss": 3.2442, + "step": 16770 + }, + { + "epoch": 1.3445110340542383, + "grad_norm": 1.7194786071777344, + "learning_rate": 5.121932357682847e-06, + "loss": 3.1608, + "step": 16771 + }, + { + "epoch": 1.344591208835261, + "grad_norm": 0.6630292534828186, + "learning_rate": 5.120798917323446e-06, + "loss": 2.47, + "step": 16772 + }, + { + "epoch": 1.3446713836162836, + "grad_norm": 1.7144725322723389, + "learning_rate": 5.1196655592256525e-06, + "loss": 3.1409, + "step": 16773 + }, + { + "epoch": 1.3447515583973062, + "grad_norm": 1.8736807107925415, + "learning_rate": 5.118532283408577e-06, + "loss": 3.1969, + "step": 16774 + }, + { + "epoch": 1.3448317331783288, + "grad_norm": 1.7481259107589722, + "learning_rate": 5.117399089891326e-06, + "loss": 3.1561, + "step": 16775 + }, + { + "epoch": 1.3449119079593514, + "grad_norm": 1.9260385036468506, + "learning_rate": 5.116265978693007e-06, + "loss": 3.1705, + "step": 16776 + }, + { + "epoch": 1.344992082740374, + "grad_norm": 1.8376033306121826, + "learning_rate": 5.115132949832726e-06, + "loss": 3.1677, + "step": 16777 + }, + { + "epoch": 1.3450722575213967, + "grad_norm": 1.8098384141921997, + "learning_rate": 5.114000003329578e-06, + "loss": 3.2668, + "step": 16778 + }, + { + "epoch": 1.3451524323024193, + "grad_norm": 1.8146700859069824, + "learning_rate": 5.112867139202666e-06, + "loss": 3.2165, + "step": 16779 + }, + { + "epoch": 1.345232607083442, + "grad_norm": 1.7929291725158691, + "learning_rate": 5.111734357471094e-06, + "loss": 3.1145, + "step": 16780 + }, + { + "epoch": 1.3453127818644646, + "grad_norm": 1.777256727218628, + "learning_rate": 5.110601658153957e-06, + "loss": 3.2008, + "step": 16781 + }, + { + "epoch": 1.3453929566454872, + "grad_norm": 1.7679215669631958, + "learning_rate": 5.1094690412703574e-06, + "loss": 3.1431, + "step": 16782 + }, + { + "epoch": 1.3454731314265098, + "grad_norm": 1.8095531463623047, + "learning_rate": 5.108336506839382e-06, + "loss": 3.2367, + "step": 16783 + }, + { + "epoch": 1.3455533062075324, + "grad_norm": 1.7092450857162476, + "learning_rate": 5.107204054880128e-06, + "loss": 3.0766, + "step": 16784 + }, + { + "epoch": 1.345633480988555, + "grad_norm": 1.8724206686019897, + "learning_rate": 5.106071685411691e-06, + "loss": 3.217, + "step": 16785 + }, + { + "epoch": 1.3457136557695777, + "grad_norm": 1.7539464235305786, + "learning_rate": 5.104939398453159e-06, + "loss": 3.2013, + "step": 16786 + }, + { + "epoch": 1.3457938305506003, + "grad_norm": 1.7394605875015259, + "learning_rate": 5.103807194023629e-06, + "loss": 3.1099, + "step": 16787 + }, + { + "epoch": 1.345874005331623, + "grad_norm": 1.7666313648223877, + "learning_rate": 5.102675072142179e-06, + "loss": 3.1518, + "step": 16788 + }, + { + "epoch": 1.3459541801126456, + "grad_norm": 1.8232311010360718, + "learning_rate": 5.101543032827902e-06, + "loss": 3.2484, + "step": 16789 + }, + { + "epoch": 1.3460343548936682, + "grad_norm": 1.9084582328796387, + "learning_rate": 5.100411076099882e-06, + "loss": 3.1595, + "step": 16790 + }, + { + "epoch": 1.3461145296746908, + "grad_norm": 1.7482191324234009, + "learning_rate": 5.0992792019772065e-06, + "loss": 3.1852, + "step": 16791 + }, + { + "epoch": 1.3461947044557134, + "grad_norm": 1.746836543083191, + "learning_rate": 5.098147410478958e-06, + "loss": 3.1154, + "step": 16792 + }, + { + "epoch": 1.346274879236736, + "grad_norm": 1.7905173301696777, + "learning_rate": 5.097015701624215e-06, + "loss": 3.1442, + "step": 16793 + }, + { + "epoch": 1.3463550540177587, + "grad_norm": 1.8642226457595825, + "learning_rate": 5.0958840754320585e-06, + "loss": 3.1638, + "step": 16794 + }, + { + "epoch": 1.3464352287987813, + "grad_norm": 1.770708441734314, + "learning_rate": 5.094752531921567e-06, + "loss": 3.2054, + "step": 16795 + }, + { + "epoch": 1.346515403579804, + "grad_norm": 1.782800555229187, + "learning_rate": 5.093621071111821e-06, + "loss": 3.1818, + "step": 16796 + }, + { + "epoch": 1.3465955783608266, + "grad_norm": 1.8345168828964233, + "learning_rate": 5.0924896930218955e-06, + "loss": 3.212, + "step": 16797 + }, + { + "epoch": 1.3466757531418492, + "grad_norm": 1.927276372909546, + "learning_rate": 5.091358397670863e-06, + "loss": 3.0797, + "step": 16798 + }, + { + "epoch": 1.3467559279228718, + "grad_norm": 1.8152339458465576, + "learning_rate": 5.090227185077796e-06, + "loss": 3.1434, + "step": 16799 + }, + { + "epoch": 1.3468361027038944, + "grad_norm": 1.688149094581604, + "learning_rate": 5.0890960552617695e-06, + "loss": 3.214, + "step": 16800 + }, + { + "epoch": 1.346916277484917, + "grad_norm": 1.7015272378921509, + "learning_rate": 5.087965008241853e-06, + "loss": 3.1371, + "step": 16801 + }, + { + "epoch": 1.3469964522659397, + "grad_norm": 1.8242056369781494, + "learning_rate": 5.086834044037119e-06, + "loss": 3.2043, + "step": 16802 + }, + { + "epoch": 1.3470766270469623, + "grad_norm": 1.805618405342102, + "learning_rate": 5.0857031626666265e-06, + "loss": 3.1437, + "step": 16803 + }, + { + "epoch": 1.347156801827985, + "grad_norm": 1.8089770078659058, + "learning_rate": 5.084572364149447e-06, + "loss": 3.1504, + "step": 16804 + }, + { + "epoch": 1.3472369766090075, + "grad_norm": 1.774668574333191, + "learning_rate": 5.083441648504646e-06, + "loss": 3.2183, + "step": 16805 + }, + { + "epoch": 1.3473171513900302, + "grad_norm": 1.7303638458251953, + "learning_rate": 5.082311015751286e-06, + "loss": 3.1074, + "step": 16806 + }, + { + "epoch": 1.3473973261710528, + "grad_norm": 1.8067842721939087, + "learning_rate": 5.081180465908433e-06, + "loss": 3.2144, + "step": 16807 + }, + { + "epoch": 1.3474775009520754, + "grad_norm": 1.7211451530456543, + "learning_rate": 5.080049998995139e-06, + "loss": 3.2206, + "step": 16808 + }, + { + "epoch": 1.347557675733098, + "grad_norm": 1.7252970933914185, + "learning_rate": 5.078919615030469e-06, + "loss": 3.1581, + "step": 16809 + }, + { + "epoch": 1.3476378505141207, + "grad_norm": 1.8010203838348389, + "learning_rate": 5.077789314033479e-06, + "loss": 3.0675, + "step": 16810 + }, + { + "epoch": 1.3477180252951433, + "grad_norm": 1.7732837200164795, + "learning_rate": 5.076659096023226e-06, + "loss": 3.1941, + "step": 16811 + }, + { + "epoch": 1.347798200076166, + "grad_norm": 1.7800214290618896, + "learning_rate": 5.075528961018772e-06, + "loss": 3.1608, + "step": 16812 + }, + { + "epoch": 1.3478783748571885, + "grad_norm": 1.7408355474472046, + "learning_rate": 5.074398909039158e-06, + "loss": 3.0635, + "step": 16813 + }, + { + "epoch": 1.3479585496382114, + "grad_norm": 1.7683196067810059, + "learning_rate": 5.073268940103443e-06, + "loss": 3.1871, + "step": 16814 + }, + { + "epoch": 1.348038724419234, + "grad_norm": 1.7254096269607544, + "learning_rate": 5.072139054230677e-06, + "loss": 3.1008, + "step": 16815 + }, + { + "epoch": 1.3481188992002566, + "grad_norm": 0.6702049970626831, + "learning_rate": 5.07100925143991e-06, + "loss": 2.4648, + "step": 16816 + }, + { + "epoch": 1.3481990739812793, + "grad_norm": 1.7303861379623413, + "learning_rate": 5.069879531750195e-06, + "loss": 3.2385, + "step": 16817 + }, + { + "epoch": 1.3482792487623019, + "grad_norm": 1.8945116996765137, + "learning_rate": 5.068749895180567e-06, + "loss": 3.2721, + "step": 16818 + }, + { + "epoch": 1.3483594235433245, + "grad_norm": 1.7943673133850098, + "learning_rate": 5.067620341750079e-06, + "loss": 3.1288, + "step": 16819 + }, + { + "epoch": 1.3484395983243471, + "grad_norm": 1.7931270599365234, + "learning_rate": 5.066490871477775e-06, + "loss": 3.2222, + "step": 16820 + }, + { + "epoch": 1.3485197731053697, + "grad_norm": 1.7630866765975952, + "learning_rate": 5.065361484382696e-06, + "loss": 3.1488, + "step": 16821 + }, + { + "epoch": 1.3485999478863924, + "grad_norm": 1.7528594732284546, + "learning_rate": 5.064232180483885e-06, + "loss": 3.0911, + "step": 16822 + }, + { + "epoch": 1.348680122667415, + "grad_norm": 1.9787815809249878, + "learning_rate": 5.063102959800378e-06, + "loss": 3.1354, + "step": 16823 + }, + { + "epoch": 1.3487602974484376, + "grad_norm": 1.7432206869125366, + "learning_rate": 5.061973822351214e-06, + "loss": 3.2515, + "step": 16824 + }, + { + "epoch": 1.3488404722294602, + "grad_norm": 1.86196768283844, + "learning_rate": 5.060844768155432e-06, + "loss": 3.2103, + "step": 16825 + }, + { + "epoch": 1.3489206470104829, + "grad_norm": 1.8382103443145752, + "learning_rate": 5.0597157972320655e-06, + "loss": 3.2219, + "step": 16826 + }, + { + "epoch": 1.3490008217915055, + "grad_norm": 1.731642484664917, + "learning_rate": 5.058586909600154e-06, + "loss": 3.1045, + "step": 16827 + }, + { + "epoch": 1.3490809965725281, + "grad_norm": 1.8518871068954468, + "learning_rate": 5.0574581052787215e-06, + "loss": 3.1911, + "step": 16828 + }, + { + "epoch": 1.3491611713535507, + "grad_norm": 1.8439828157424927, + "learning_rate": 5.0563293842868046e-06, + "loss": 3.1953, + "step": 16829 + }, + { + "epoch": 1.3492413461345734, + "grad_norm": 1.7655296325683594, + "learning_rate": 5.055200746643432e-06, + "loss": 3.1882, + "step": 16830 + }, + { + "epoch": 1.349321520915596, + "grad_norm": 1.8060741424560547, + "learning_rate": 5.05407219236763e-06, + "loss": 3.1393, + "step": 16831 + }, + { + "epoch": 1.3494016956966186, + "grad_norm": 1.7814205884933472, + "learning_rate": 5.052943721478433e-06, + "loss": 3.1317, + "step": 16832 + }, + { + "epoch": 1.3494818704776412, + "grad_norm": 1.8357200622558594, + "learning_rate": 5.051815333994856e-06, + "loss": 3.272, + "step": 16833 + }, + { + "epoch": 1.3495620452586639, + "grad_norm": 1.8328908681869507, + "learning_rate": 5.05068702993593e-06, + "loss": 3.2658, + "step": 16834 + }, + { + "epoch": 1.3496422200396865, + "grad_norm": 1.7312994003295898, + "learning_rate": 5.049558809320677e-06, + "loss": 3.2542, + "step": 16835 + }, + { + "epoch": 1.349722394820709, + "grad_norm": 1.7389787435531616, + "learning_rate": 5.048430672168117e-06, + "loss": 3.2488, + "step": 16836 + }, + { + "epoch": 1.3498025696017317, + "grad_norm": 1.6932854652404785, + "learning_rate": 5.047302618497274e-06, + "loss": 3.1308, + "step": 16837 + }, + { + "epoch": 1.3498827443827544, + "grad_norm": 1.7266318798065186, + "learning_rate": 5.04617464832716e-06, + "loss": 3.1866, + "step": 16838 + }, + { + "epoch": 1.349962919163777, + "grad_norm": 1.657034158706665, + "learning_rate": 5.045046761676795e-06, + "loss": 3.1001, + "step": 16839 + }, + { + "epoch": 1.3500430939447996, + "grad_norm": 1.9355926513671875, + "learning_rate": 5.043918958565197e-06, + "loss": 3.1815, + "step": 16840 + }, + { + "epoch": 1.3501232687258222, + "grad_norm": 1.736646294593811, + "learning_rate": 5.042791239011378e-06, + "loss": 3.0878, + "step": 16841 + }, + { + "epoch": 1.3502034435068448, + "grad_norm": 1.7982317209243774, + "learning_rate": 5.0416636030343545e-06, + "loss": 3.2467, + "step": 16842 + }, + { + "epoch": 1.3502836182878675, + "grad_norm": 1.768871545791626, + "learning_rate": 5.0405360506531306e-06, + "loss": 3.244, + "step": 16843 + }, + { + "epoch": 1.3503637930688903, + "grad_norm": 1.7197552919387817, + "learning_rate": 5.039408581886723e-06, + "loss": 3.1509, + "step": 16844 + }, + { + "epoch": 1.350443967849913, + "grad_norm": 1.8982398509979248, + "learning_rate": 5.038281196754138e-06, + "loss": 3.1349, + "step": 16845 + }, + { + "epoch": 1.3505241426309356, + "grad_norm": 1.6474461555480957, + "learning_rate": 5.037153895274383e-06, + "loss": 3.1323, + "step": 16846 + }, + { + "epoch": 1.3506043174119582, + "grad_norm": 1.898580551147461, + "learning_rate": 5.036026677466468e-06, + "loss": 3.2326, + "step": 16847 + }, + { + "epoch": 1.3506844921929808, + "grad_norm": 1.8028180599212646, + "learning_rate": 5.03489954334939e-06, + "loss": 3.2051, + "step": 16848 + }, + { + "epoch": 1.3507646669740034, + "grad_norm": 1.8709607124328613, + "learning_rate": 5.033772492942155e-06, + "loss": 3.0627, + "step": 16849 + }, + { + "epoch": 1.350844841755026, + "grad_norm": 1.770125389099121, + "learning_rate": 5.032645526263766e-06, + "loss": 3.0485, + "step": 16850 + }, + { + "epoch": 1.3509250165360487, + "grad_norm": 1.7186676263809204, + "learning_rate": 5.0315186433332244e-06, + "loss": 3.1462, + "step": 16851 + }, + { + "epoch": 1.3510051913170713, + "grad_norm": 1.7741116285324097, + "learning_rate": 5.03039184416953e-06, + "loss": 3.1946, + "step": 16852 + }, + { + "epoch": 1.351085366098094, + "grad_norm": 1.8836493492126465, + "learning_rate": 5.029265128791674e-06, + "loss": 3.1429, + "step": 16853 + }, + { + "epoch": 1.3511655408791166, + "grad_norm": 1.7926716804504395, + "learning_rate": 5.028138497218655e-06, + "loss": 3.1469, + "step": 16854 + }, + { + "epoch": 1.3512457156601392, + "grad_norm": 1.8131682872772217, + "learning_rate": 5.027011949469469e-06, + "loss": 3.0793, + "step": 16855 + }, + { + "epoch": 1.3513258904411618, + "grad_norm": 1.7958128452301025, + "learning_rate": 5.02588548556311e-06, + "loss": 3.2271, + "step": 16856 + }, + { + "epoch": 1.3514060652221844, + "grad_norm": 1.7680954933166504, + "learning_rate": 5.024759105518572e-06, + "loss": 3.2234, + "step": 16857 + }, + { + "epoch": 1.351486240003207, + "grad_norm": 1.7881277799606323, + "learning_rate": 5.023632809354838e-06, + "loss": 3.0834, + "step": 16858 + }, + { + "epoch": 1.3515664147842297, + "grad_norm": 1.9543243646621704, + "learning_rate": 5.022506597090902e-06, + "loss": 3.2524, + "step": 16859 + }, + { + "epoch": 1.3516465895652523, + "grad_norm": 1.7842271327972412, + "learning_rate": 5.021380468745749e-06, + "loss": 3.1174, + "step": 16860 + }, + { + "epoch": 1.351726764346275, + "grad_norm": 1.769459843635559, + "learning_rate": 5.020254424338367e-06, + "loss": 3.2064, + "step": 16861 + }, + { + "epoch": 1.3518069391272975, + "grad_norm": 1.6260994672775269, + "learning_rate": 5.0191284638877434e-06, + "loss": 3.1644, + "step": 16862 + }, + { + "epoch": 1.3518871139083202, + "grad_norm": 1.814003825187683, + "learning_rate": 5.018002587412857e-06, + "loss": 3.2518, + "step": 16863 + }, + { + "epoch": 1.3519672886893428, + "grad_norm": 1.8035171031951904, + "learning_rate": 5.0168767949326905e-06, + "loss": 3.1714, + "step": 16864 + }, + { + "epoch": 1.3520474634703654, + "grad_norm": 1.8056477308273315, + "learning_rate": 5.015751086466223e-06, + "loss": 3.178, + "step": 16865 + }, + { + "epoch": 1.352127638251388, + "grad_norm": 1.7860524654388428, + "learning_rate": 5.014625462032438e-06, + "loss": 3.1639, + "step": 16866 + }, + { + "epoch": 1.3522078130324107, + "grad_norm": 1.9137665033340454, + "learning_rate": 5.013499921650313e-06, + "loss": 3.134, + "step": 16867 + }, + { + "epoch": 1.3522879878134333, + "grad_norm": 1.7503715753555298, + "learning_rate": 5.012374465338819e-06, + "loss": 3.2055, + "step": 16868 + }, + { + "epoch": 1.352368162594456, + "grad_norm": 1.795672059059143, + "learning_rate": 5.0112490931169326e-06, + "loss": 3.2327, + "step": 16869 + }, + { + "epoch": 1.3524483373754785, + "grad_norm": 1.8751142024993896, + "learning_rate": 5.010123805003628e-06, + "loss": 3.1566, + "step": 16870 + }, + { + "epoch": 1.3525285121565012, + "grad_norm": 1.7761495113372803, + "learning_rate": 5.008998601017879e-06, + "loss": 3.2495, + "step": 16871 + }, + { + "epoch": 1.3526086869375238, + "grad_norm": 1.8608952760696411, + "learning_rate": 5.007873481178658e-06, + "loss": 3.2321, + "step": 16872 + }, + { + "epoch": 1.3526888617185464, + "grad_norm": 1.7470850944519043, + "learning_rate": 5.006748445504927e-06, + "loss": 3.1029, + "step": 16873 + }, + { + "epoch": 1.352769036499569, + "grad_norm": 1.7384109497070312, + "learning_rate": 5.0056234940156564e-06, + "loss": 3.2054, + "step": 16874 + }, + { + "epoch": 1.3528492112805917, + "grad_norm": 1.6868518590927124, + "learning_rate": 5.004498626729815e-06, + "loss": 3.2367, + "step": 16875 + }, + { + "epoch": 1.3529293860616143, + "grad_norm": 1.7052674293518066, + "learning_rate": 5.0033738436663645e-06, + "loss": 3.1755, + "step": 16876 + }, + { + "epoch": 1.353009560842637, + "grad_norm": 0.6551171541213989, + "learning_rate": 5.0022491448442756e-06, + "loss": 2.462, + "step": 16877 + }, + { + "epoch": 1.3530897356236595, + "grad_norm": 1.9213778972625732, + "learning_rate": 5.001124530282499e-06, + "loss": 3.2545, + "step": 16878 + }, + { + "epoch": 1.3531699104046822, + "grad_norm": 1.718650221824646, + "learning_rate": 5.000000000000003e-06, + "loss": 3.1526, + "step": 16879 + }, + { + "epoch": 1.3532500851857048, + "grad_norm": 1.8098291158676147, + "learning_rate": 4.998875554015743e-06, + "loss": 3.1108, + "step": 16880 + }, + { + "epoch": 1.3533302599667274, + "grad_norm": 1.895645260810852, + "learning_rate": 4.99775119234868e-06, + "loss": 3.2002, + "step": 16881 + }, + { + "epoch": 1.35341043474775, + "grad_norm": 1.745499849319458, + "learning_rate": 4.996626915017772e-06, + "loss": 3.2134, + "step": 16882 + }, + { + "epoch": 1.3534906095287726, + "grad_norm": 1.7153327465057373, + "learning_rate": 4.995502722041967e-06, + "loss": 3.1754, + "step": 16883 + }, + { + "epoch": 1.3535707843097953, + "grad_norm": 1.7566906213760376, + "learning_rate": 4.994378613440223e-06, + "loss": 3.1174, + "step": 16884 + }, + { + "epoch": 1.353650959090818, + "grad_norm": 0.6730977296829224, + "learning_rate": 4.993254589231491e-06, + "loss": 2.4752, + "step": 16885 + }, + { + "epoch": 1.3537311338718405, + "grad_norm": 1.793390154838562, + "learning_rate": 4.9921306494347224e-06, + "loss": 3.2921, + "step": 16886 + }, + { + "epoch": 1.3538113086528631, + "grad_norm": 1.9459872245788574, + "learning_rate": 4.991006794068869e-06, + "loss": 3.3281, + "step": 16887 + }, + { + "epoch": 1.3538914834338858, + "grad_norm": 1.7790449857711792, + "learning_rate": 4.989883023152873e-06, + "loss": 3.1399, + "step": 16888 + }, + { + "epoch": 1.3539716582149084, + "grad_norm": 1.8024258613586426, + "learning_rate": 4.988759336705683e-06, + "loss": 3.1409, + "step": 16889 + }, + { + "epoch": 1.354051832995931, + "grad_norm": 1.719119668006897, + "learning_rate": 4.987635734746244e-06, + "loss": 3.221, + "step": 16890 + }, + { + "epoch": 1.3541320077769536, + "grad_norm": 0.6713903546333313, + "learning_rate": 4.986512217293501e-06, + "loss": 2.4758, + "step": 16891 + }, + { + "epoch": 1.3542121825579763, + "grad_norm": 1.7961541414260864, + "learning_rate": 4.985388784366398e-06, + "loss": 3.145, + "step": 16892 + }, + { + "epoch": 1.3542923573389989, + "grad_norm": 1.7588071823120117, + "learning_rate": 4.984265435983869e-06, + "loss": 3.1825, + "step": 16893 + }, + { + "epoch": 1.3543725321200215, + "grad_norm": 1.8367034196853638, + "learning_rate": 4.983142172164858e-06, + "loss": 3.2104, + "step": 16894 + }, + { + "epoch": 1.3544527069010444, + "grad_norm": 1.819021224975586, + "learning_rate": 4.982018992928302e-06, + "loss": 3.214, + "step": 16895 + }, + { + "epoch": 1.354532881682067, + "grad_norm": 1.7332053184509277, + "learning_rate": 4.980895898293135e-06, + "loss": 3.0836, + "step": 16896 + }, + { + "epoch": 1.3546130564630896, + "grad_norm": 1.7878608703613281, + "learning_rate": 4.9797728882783e-06, + "loss": 3.246, + "step": 16897 + }, + { + "epoch": 1.3546932312441122, + "grad_norm": 1.7019307613372803, + "learning_rate": 4.978649962902721e-06, + "loss": 3.0768, + "step": 16898 + }, + { + "epoch": 1.3547734060251349, + "grad_norm": 1.7063699960708618, + "learning_rate": 4.977527122185333e-06, + "loss": 3.142, + "step": 16899 + }, + { + "epoch": 1.3548535808061575, + "grad_norm": 1.7818583250045776, + "learning_rate": 4.976404366145068e-06, + "loss": 3.152, + "step": 16900 + }, + { + "epoch": 1.35493375558718, + "grad_norm": 1.8253904581069946, + "learning_rate": 4.975281694800856e-06, + "loss": 3.2515, + "step": 16901 + }, + { + "epoch": 1.3550139303682027, + "grad_norm": 1.7315863370895386, + "learning_rate": 4.974159108171626e-06, + "loss": 3.2021, + "step": 16902 + }, + { + "epoch": 1.3550941051492253, + "grad_norm": 1.7539010047912598, + "learning_rate": 4.9730366062762984e-06, + "loss": 3.2418, + "step": 16903 + }, + { + "epoch": 1.355174279930248, + "grad_norm": 1.7954734563827515, + "learning_rate": 4.971914189133803e-06, + "loss": 3.188, + "step": 16904 + }, + { + "epoch": 1.3552544547112706, + "grad_norm": 0.671323835849762, + "learning_rate": 4.970791856763061e-06, + "loss": 2.4802, + "step": 16905 + }, + { + "epoch": 1.3553346294922932, + "grad_norm": 1.8584157228469849, + "learning_rate": 4.969669609182997e-06, + "loss": 3.1749, + "step": 16906 + }, + { + "epoch": 1.3554148042733158, + "grad_norm": 1.8044555187225342, + "learning_rate": 4.968547446412534e-06, + "loss": 3.1632, + "step": 16907 + }, + { + "epoch": 1.3554949790543385, + "grad_norm": 1.764376163482666, + "learning_rate": 4.967425368470584e-06, + "loss": 3.1764, + "step": 16908 + }, + { + "epoch": 1.355575153835361, + "grad_norm": 1.8429572582244873, + "learning_rate": 4.9663033753760685e-06, + "loss": 3.2572, + "step": 16909 + }, + { + "epoch": 1.3556553286163837, + "grad_norm": 1.8592978715896606, + "learning_rate": 4.965181467147904e-06, + "loss": 3.1857, + "step": 16910 + }, + { + "epoch": 1.3557355033974063, + "grad_norm": 1.778172254562378, + "learning_rate": 4.964059643805006e-06, + "loss": 3.2922, + "step": 16911 + }, + { + "epoch": 1.355815678178429, + "grad_norm": 1.841459035873413, + "learning_rate": 4.962937905366291e-06, + "loss": 3.2593, + "step": 16912 + }, + { + "epoch": 1.3558958529594516, + "grad_norm": 1.8601031303405762, + "learning_rate": 4.961816251850663e-06, + "loss": 3.1849, + "step": 16913 + }, + { + "epoch": 1.3559760277404742, + "grad_norm": 1.8289332389831543, + "learning_rate": 4.960694683277038e-06, + "loss": 3.2404, + "step": 16914 + }, + { + "epoch": 1.3560562025214968, + "grad_norm": 1.8681334257125854, + "learning_rate": 4.959573199664326e-06, + "loss": 3.2526, + "step": 16915 + }, + { + "epoch": 1.3561363773025195, + "grad_norm": 1.8227605819702148, + "learning_rate": 4.958451801031432e-06, + "loss": 3.1271, + "step": 16916 + }, + { + "epoch": 1.356216552083542, + "grad_norm": 1.8142120838165283, + "learning_rate": 4.957330487397267e-06, + "loss": 3.0988, + "step": 16917 + }, + { + "epoch": 1.3562967268645647, + "grad_norm": 1.839270830154419, + "learning_rate": 4.956209258780731e-06, + "loss": 3.1741, + "step": 16918 + }, + { + "epoch": 1.3563769016455873, + "grad_norm": 1.8248625993728638, + "learning_rate": 4.955088115200728e-06, + "loss": 3.1937, + "step": 16919 + }, + { + "epoch": 1.35645707642661, + "grad_norm": 1.7289048433303833, + "learning_rate": 4.953967056676161e-06, + "loss": 3.1743, + "step": 16920 + }, + { + "epoch": 1.3565372512076326, + "grad_norm": 1.797906517982483, + "learning_rate": 4.9528460832259314e-06, + "loss": 3.1626, + "step": 16921 + }, + { + "epoch": 1.3566174259886552, + "grad_norm": 1.6937189102172852, + "learning_rate": 4.951725194868942e-06, + "loss": 3.2084, + "step": 16922 + }, + { + "epoch": 1.3566976007696778, + "grad_norm": 1.7124099731445312, + "learning_rate": 4.950604391624083e-06, + "loss": 3.2158, + "step": 16923 + }, + { + "epoch": 1.3567777755507004, + "grad_norm": 1.7371867895126343, + "learning_rate": 4.9494836735102545e-06, + "loss": 3.1796, + "step": 16924 + }, + { + "epoch": 1.3568579503317233, + "grad_norm": 1.7708767652511597, + "learning_rate": 4.948363040546351e-06, + "loss": 3.1639, + "step": 16925 + }, + { + "epoch": 1.356938125112746, + "grad_norm": 1.688214898109436, + "learning_rate": 4.947242492751266e-06, + "loss": 3.2211, + "step": 16926 + }, + { + "epoch": 1.3570182998937685, + "grad_norm": 1.8341243267059326, + "learning_rate": 4.946122030143896e-06, + "loss": 3.1075, + "step": 16927 + }, + { + "epoch": 1.3570984746747912, + "grad_norm": 1.783233404159546, + "learning_rate": 4.945001652743123e-06, + "loss": 3.0919, + "step": 16928 + }, + { + "epoch": 1.3571786494558138, + "grad_norm": 1.7611019611358643, + "learning_rate": 4.943881360567842e-06, + "loss": 3.2458, + "step": 16929 + }, + { + "epoch": 1.3572588242368364, + "grad_norm": 1.7824852466583252, + "learning_rate": 4.942761153636938e-06, + "loss": 3.2085, + "step": 16930 + }, + { + "epoch": 1.357338999017859, + "grad_norm": 1.6863592863082886, + "learning_rate": 4.9416410319692985e-06, + "loss": 3.132, + "step": 16931 + }, + { + "epoch": 1.3574191737988817, + "grad_norm": 1.7475115060806274, + "learning_rate": 4.940520995583813e-06, + "loss": 3.2041, + "step": 16932 + }, + { + "epoch": 1.3574993485799043, + "grad_norm": 1.726971983909607, + "learning_rate": 4.939401044499357e-06, + "loss": 3.1902, + "step": 16933 + }, + { + "epoch": 1.357579523360927, + "grad_norm": 1.7835108041763306, + "learning_rate": 4.938281178734815e-06, + "loss": 3.1932, + "step": 16934 + }, + { + "epoch": 1.3576596981419495, + "grad_norm": 1.7655795812606812, + "learning_rate": 4.937161398309069e-06, + "loss": 3.1808, + "step": 16935 + }, + { + "epoch": 1.3577398729229722, + "grad_norm": 1.628722071647644, + "learning_rate": 4.936041703240997e-06, + "loss": 3.1927, + "step": 16936 + }, + { + "epoch": 1.3578200477039948, + "grad_norm": 1.8583859205245972, + "learning_rate": 4.934922093549477e-06, + "loss": 3.1039, + "step": 16937 + }, + { + "epoch": 1.3579002224850174, + "grad_norm": 1.7056218385696411, + "learning_rate": 4.9338025692533895e-06, + "loss": 3.1132, + "step": 16938 + }, + { + "epoch": 1.35798039726604, + "grad_norm": 1.7551579475402832, + "learning_rate": 4.932683130371602e-06, + "loss": 3.1963, + "step": 16939 + }, + { + "epoch": 1.3580605720470627, + "grad_norm": 2.189577341079712, + "learning_rate": 4.93156377692299e-06, + "loss": 3.1456, + "step": 16940 + }, + { + "epoch": 1.3581407468280853, + "grad_norm": 1.9299391508102417, + "learning_rate": 4.9304445089264265e-06, + "loss": 3.1698, + "step": 16941 + }, + { + "epoch": 1.358220921609108, + "grad_norm": 1.8473421335220337, + "learning_rate": 4.929325326400782e-06, + "loss": 3.0974, + "step": 16942 + }, + { + "epoch": 1.3583010963901305, + "grad_norm": 1.6961921453475952, + "learning_rate": 4.92820622936493e-06, + "loss": 3.1029, + "step": 16943 + }, + { + "epoch": 1.3583812711711531, + "grad_norm": 0.6720011234283447, + "learning_rate": 4.9270872178377295e-06, + "loss": 2.4769, + "step": 16944 + }, + { + "epoch": 1.3584614459521758, + "grad_norm": 1.7854857444763184, + "learning_rate": 4.9259682918380495e-06, + "loss": 3.1633, + "step": 16945 + }, + { + "epoch": 1.3585416207331984, + "grad_norm": 0.6955296993255615, + "learning_rate": 4.924849451384757e-06, + "loss": 2.4763, + "step": 16946 + }, + { + "epoch": 1.358621795514221, + "grad_norm": 1.7421834468841553, + "learning_rate": 4.923730696496713e-06, + "loss": 3.1921, + "step": 16947 + }, + { + "epoch": 1.3587019702952436, + "grad_norm": 1.6409368515014648, + "learning_rate": 4.922612027192785e-06, + "loss": 3.1335, + "step": 16948 + }, + { + "epoch": 1.3587821450762663, + "grad_norm": 1.7749392986297607, + "learning_rate": 4.921493443491825e-06, + "loss": 3.2865, + "step": 16949 + }, + { + "epoch": 1.358862319857289, + "grad_norm": 1.855448842048645, + "learning_rate": 4.920374945412696e-06, + "loss": 3.1284, + "step": 16950 + }, + { + "epoch": 1.3589424946383115, + "grad_norm": 1.7807254791259766, + "learning_rate": 4.919256532974255e-06, + "loss": 3.1659, + "step": 16951 + }, + { + "epoch": 1.3590226694193341, + "grad_norm": 1.8159277439117432, + "learning_rate": 4.9181382061953574e-06, + "loss": 3.1335, + "step": 16952 + }, + { + "epoch": 1.3591028442003568, + "grad_norm": 1.8644258975982666, + "learning_rate": 4.9170199650948635e-06, + "loss": 3.1833, + "step": 16953 + }, + { + "epoch": 1.3591830189813794, + "grad_norm": 1.814341425895691, + "learning_rate": 4.915901809691617e-06, + "loss": 3.3303, + "step": 16954 + }, + { + "epoch": 1.359263193762402, + "grad_norm": 1.8746092319488525, + "learning_rate": 4.914783740004475e-06, + "loss": 3.1552, + "step": 16955 + }, + { + "epoch": 1.3593433685434246, + "grad_norm": 3.1524341106414795, + "learning_rate": 4.9136657560522864e-06, + "loss": 3.111, + "step": 16956 + }, + { + "epoch": 1.3594235433244473, + "grad_norm": 1.8659802675247192, + "learning_rate": 4.9125478578539e-06, + "loss": 3.2721, + "step": 16957 + }, + { + "epoch": 1.3595037181054699, + "grad_norm": 1.8532932996749878, + "learning_rate": 4.911430045428168e-06, + "loss": 3.3173, + "step": 16958 + }, + { + "epoch": 1.3595838928864925, + "grad_norm": 1.8000541925430298, + "learning_rate": 4.9103123187939285e-06, + "loss": 3.1425, + "step": 16959 + }, + { + "epoch": 1.3596640676675151, + "grad_norm": 1.7922577857971191, + "learning_rate": 4.909194677970029e-06, + "loss": 3.1876, + "step": 16960 + }, + { + "epoch": 1.3597442424485378, + "grad_norm": 1.794811487197876, + "learning_rate": 4.908077122975315e-06, + "loss": 3.1859, + "step": 16961 + }, + { + "epoch": 1.3598244172295604, + "grad_norm": 1.867215633392334, + "learning_rate": 4.906959653828623e-06, + "loss": 3.1556, + "step": 16962 + }, + { + "epoch": 1.359904592010583, + "grad_norm": 1.7444498538970947, + "learning_rate": 4.905842270548802e-06, + "loss": 3.2128, + "step": 16963 + }, + { + "epoch": 1.3599847667916056, + "grad_norm": 1.839004397392273, + "learning_rate": 4.904724973154681e-06, + "loss": 3.2155, + "step": 16964 + }, + { + "epoch": 1.3600649415726282, + "grad_norm": 1.7810709476470947, + "learning_rate": 4.903607761665101e-06, + "loss": 3.2703, + "step": 16965 + }, + { + "epoch": 1.3601451163536509, + "grad_norm": 1.8378586769104004, + "learning_rate": 4.9024906360988974e-06, + "loss": 3.0987, + "step": 16966 + }, + { + "epoch": 1.3602252911346735, + "grad_norm": 1.7534719705581665, + "learning_rate": 4.9013735964749065e-06, + "loss": 3.1868, + "step": 16967 + }, + { + "epoch": 1.3603054659156961, + "grad_norm": 1.7614139318466187, + "learning_rate": 4.900256642811963e-06, + "loss": 3.1736, + "step": 16968 + }, + { + "epoch": 1.3603856406967187, + "grad_norm": 1.699779748916626, + "learning_rate": 4.899139775128891e-06, + "loss": 3.2218, + "step": 16969 + }, + { + "epoch": 1.3604658154777414, + "grad_norm": 1.7335922718048096, + "learning_rate": 4.898022993444525e-06, + "loss": 3.1115, + "step": 16970 + }, + { + "epoch": 1.360545990258764, + "grad_norm": 1.8971213102340698, + "learning_rate": 4.896906297777693e-06, + "loss": 3.2, + "step": 16971 + }, + { + "epoch": 1.3606261650397866, + "grad_norm": 1.7485949993133545, + "learning_rate": 4.8957896881472224e-06, + "loss": 3.0934, + "step": 16972 + }, + { + "epoch": 1.3607063398208092, + "grad_norm": 1.7499148845672607, + "learning_rate": 4.8946731645719425e-06, + "loss": 3.1321, + "step": 16973 + }, + { + "epoch": 1.3607865146018319, + "grad_norm": 1.7453678846359253, + "learning_rate": 4.893556727070671e-06, + "loss": 3.212, + "step": 16974 + }, + { + "epoch": 1.3608666893828545, + "grad_norm": 1.7297331094741821, + "learning_rate": 4.892440375662232e-06, + "loss": 3.1792, + "step": 16975 + }, + { + "epoch": 1.3609468641638773, + "grad_norm": 1.794471263885498, + "learning_rate": 4.891324110365449e-06, + "loss": 3.181, + "step": 16976 + }, + { + "epoch": 1.3610270389449, + "grad_norm": 1.8925899267196655, + "learning_rate": 4.89020793119914e-06, + "loss": 3.2216, + "step": 16977 + }, + { + "epoch": 1.3611072137259226, + "grad_norm": 1.7178504467010498, + "learning_rate": 4.88909183818213e-06, + "loss": 3.1937, + "step": 16978 + }, + { + "epoch": 1.3611873885069452, + "grad_norm": 1.7912040948867798, + "learning_rate": 4.887975831333225e-06, + "loss": 3.1662, + "step": 16979 + }, + { + "epoch": 1.3612675632879678, + "grad_norm": 1.8129905462265015, + "learning_rate": 4.886859910671246e-06, + "loss": 3.2409, + "step": 16980 + }, + { + "epoch": 1.3613477380689905, + "grad_norm": 1.7846529483795166, + "learning_rate": 4.8857440762150065e-06, + "loss": 3.1013, + "step": 16981 + }, + { + "epoch": 1.361427912850013, + "grad_norm": 1.773324966430664, + "learning_rate": 4.88462832798332e-06, + "loss": 3.1942, + "step": 16982 + }, + { + "epoch": 1.3615080876310357, + "grad_norm": 1.8030880689620972, + "learning_rate": 4.8835126659950016e-06, + "loss": 3.2542, + "step": 16983 + }, + { + "epoch": 1.3615882624120583, + "grad_norm": 1.8589413166046143, + "learning_rate": 4.882397090268852e-06, + "loss": 3.1778, + "step": 16984 + }, + { + "epoch": 1.361668437193081, + "grad_norm": 1.7682501077651978, + "learning_rate": 4.881281600823684e-06, + "loss": 3.1854, + "step": 16985 + }, + { + "epoch": 1.3617486119741036, + "grad_norm": 1.8367453813552856, + "learning_rate": 4.8801661976783035e-06, + "loss": 3.194, + "step": 16986 + }, + { + "epoch": 1.3618287867551262, + "grad_norm": 1.9058445692062378, + "learning_rate": 4.879050880851517e-06, + "loss": 3.1813, + "step": 16987 + }, + { + "epoch": 1.3619089615361488, + "grad_norm": 1.770124077796936, + "learning_rate": 4.877935650362132e-06, + "loss": 3.2914, + "step": 16988 + }, + { + "epoch": 1.3619891363171714, + "grad_norm": 1.7363195419311523, + "learning_rate": 4.876820506228942e-06, + "loss": 3.1682, + "step": 16989 + }, + { + "epoch": 1.362069311098194, + "grad_norm": 1.7608627080917358, + "learning_rate": 4.875705448470753e-06, + "loss": 3.1857, + "step": 16990 + }, + { + "epoch": 1.3621494858792167, + "grad_norm": 1.7920582294464111, + "learning_rate": 4.8745904771063644e-06, + "loss": 3.2433, + "step": 16991 + }, + { + "epoch": 1.3622296606602393, + "grad_norm": 1.8222013711929321, + "learning_rate": 4.8734755921545755e-06, + "loss": 3.1165, + "step": 16992 + }, + { + "epoch": 1.362309835441262, + "grad_norm": 1.7731631994247437, + "learning_rate": 4.872360793634183e-06, + "loss": 3.2473, + "step": 16993 + }, + { + "epoch": 1.3623900102222846, + "grad_norm": 1.8747360706329346, + "learning_rate": 4.871246081563977e-06, + "loss": 3.1354, + "step": 16994 + }, + { + "epoch": 1.3624701850033072, + "grad_norm": 0.7007744312286377, + "learning_rate": 4.870131455962754e-06, + "loss": 2.498, + "step": 16995 + }, + { + "epoch": 1.3625503597843298, + "grad_norm": 1.7503437995910645, + "learning_rate": 4.869016916849308e-06, + "loss": 3.1307, + "step": 16996 + }, + { + "epoch": 1.3626305345653524, + "grad_norm": 1.7940545082092285, + "learning_rate": 4.867902464242428e-06, + "loss": 3.1358, + "step": 16997 + }, + { + "epoch": 1.362710709346375, + "grad_norm": 1.7574321031570435, + "learning_rate": 4.866788098160907e-06, + "loss": 3.2317, + "step": 16998 + }, + { + "epoch": 1.3627908841273977, + "grad_norm": 1.7541899681091309, + "learning_rate": 4.865673818623527e-06, + "loss": 3.2003, + "step": 16999 + }, + { + "epoch": 1.3628710589084203, + "grad_norm": 1.781706690788269, + "learning_rate": 4.864559625649076e-06, + "loss": 3.186, + "step": 17000 + }, + { + "epoch": 1.362951233689443, + "grad_norm": 0.6719933748245239, + "learning_rate": 4.8634455192563405e-06, + "loss": 2.4644, + "step": 17001 + }, + { + "epoch": 1.3630314084704656, + "grad_norm": 0.6636399030685425, + "learning_rate": 4.862331499464102e-06, + "loss": 2.4546, + "step": 17002 + }, + { + "epoch": 1.3631115832514882, + "grad_norm": 1.95955491065979, + "learning_rate": 4.861217566291149e-06, + "loss": 3.1721, + "step": 17003 + }, + { + "epoch": 1.3631917580325108, + "grad_norm": 0.6685060262680054, + "learning_rate": 4.860103719756252e-06, + "loss": 2.4749, + "step": 17004 + }, + { + "epoch": 1.3632719328135334, + "grad_norm": 1.7060056924819946, + "learning_rate": 4.8589899598781945e-06, + "loss": 3.2552, + "step": 17005 + }, + { + "epoch": 1.363352107594556, + "grad_norm": 1.7310166358947754, + "learning_rate": 4.857876286675755e-06, + "loss": 3.1542, + "step": 17006 + }, + { + "epoch": 1.363432282375579, + "grad_norm": 0.6807122826576233, + "learning_rate": 4.856762700167709e-06, + "loss": 2.4844, + "step": 17007 + }, + { + "epoch": 1.3635124571566015, + "grad_norm": 1.7942456007003784, + "learning_rate": 4.8556492003728354e-06, + "loss": 3.1384, + "step": 17008 + }, + { + "epoch": 1.3635926319376241, + "grad_norm": 1.7526792287826538, + "learning_rate": 4.854535787309898e-06, + "loss": 3.1863, + "step": 17009 + }, + { + "epoch": 1.3636728067186468, + "grad_norm": 1.7872977256774902, + "learning_rate": 4.853422460997674e-06, + "loss": 3.0866, + "step": 17010 + }, + { + "epoch": 1.3637529814996694, + "grad_norm": 1.7065212726593018, + "learning_rate": 4.852309221454935e-06, + "loss": 3.107, + "step": 17011 + }, + { + "epoch": 1.363833156280692, + "grad_norm": 1.7421211004257202, + "learning_rate": 4.851196068700448e-06, + "loss": 3.1787, + "step": 17012 + }, + { + "epoch": 1.3639133310617146, + "grad_norm": 1.7122814655303955, + "learning_rate": 4.850083002752983e-06, + "loss": 3.1626, + "step": 17013 + }, + { + "epoch": 1.3639935058427373, + "grad_norm": 1.8458154201507568, + "learning_rate": 4.848970023631301e-06, + "loss": 3.2291, + "step": 17014 + }, + { + "epoch": 1.3640736806237599, + "grad_norm": 1.7936981916427612, + "learning_rate": 4.847857131354169e-06, + "loss": 3.0824, + "step": 17015 + }, + { + "epoch": 1.3641538554047825, + "grad_norm": 1.8259564638137817, + "learning_rate": 4.84674432594035e-06, + "loss": 3.1906, + "step": 17016 + }, + { + "epoch": 1.3642340301858051, + "grad_norm": 1.7061201333999634, + "learning_rate": 4.845631607408605e-06, + "loss": 3.12, + "step": 17017 + }, + { + "epoch": 1.3643142049668278, + "grad_norm": 2.0008513927459717, + "learning_rate": 4.844518975777699e-06, + "loss": 3.2255, + "step": 17018 + }, + { + "epoch": 1.3643943797478504, + "grad_norm": 1.682834506034851, + "learning_rate": 4.843406431066382e-06, + "loss": 3.1461, + "step": 17019 + }, + { + "epoch": 1.364474554528873, + "grad_norm": 1.7519601583480835, + "learning_rate": 4.842293973293415e-06, + "loss": 3.1032, + "step": 17020 + }, + { + "epoch": 1.3645547293098956, + "grad_norm": 1.7529997825622559, + "learning_rate": 4.841181602477554e-06, + "loss": 3.1632, + "step": 17021 + }, + { + "epoch": 1.3646349040909183, + "grad_norm": 1.8087247610092163, + "learning_rate": 4.840069318637553e-06, + "loss": 3.2679, + "step": 17022 + }, + { + "epoch": 1.3647150788719409, + "grad_norm": 1.7582660913467407, + "learning_rate": 4.8389571217921685e-06, + "loss": 3.1897, + "step": 17023 + }, + { + "epoch": 1.3647952536529635, + "grad_norm": 1.6613963842391968, + "learning_rate": 4.837845011960144e-06, + "loss": 3.1338, + "step": 17024 + }, + { + "epoch": 1.3648754284339861, + "grad_norm": 0.6761514544487, + "learning_rate": 4.836732989160233e-06, + "loss": 2.457, + "step": 17025 + }, + { + "epoch": 1.3649556032150088, + "grad_norm": 1.720847249031067, + "learning_rate": 4.8356210534111845e-06, + "loss": 3.1334, + "step": 17026 + }, + { + "epoch": 1.3650357779960314, + "grad_norm": 1.7354788780212402, + "learning_rate": 4.834509204731746e-06, + "loss": 3.1968, + "step": 17027 + }, + { + "epoch": 1.365115952777054, + "grad_norm": 1.7263723611831665, + "learning_rate": 4.833397443140664e-06, + "loss": 3.223, + "step": 17028 + }, + { + "epoch": 1.3651961275580766, + "grad_norm": 1.7354437112808228, + "learning_rate": 4.832285768656677e-06, + "loss": 3.1399, + "step": 17029 + }, + { + "epoch": 1.3652763023390992, + "grad_norm": 0.693997323513031, + "learning_rate": 4.83117418129853e-06, + "loss": 2.4728, + "step": 17030 + }, + { + "epoch": 1.3653564771201219, + "grad_norm": 1.8552809953689575, + "learning_rate": 4.830062681084966e-06, + "loss": 3.2198, + "step": 17031 + }, + { + "epoch": 1.3654366519011445, + "grad_norm": 1.7519514560699463, + "learning_rate": 4.828951268034723e-06, + "loss": 3.2262, + "step": 17032 + }, + { + "epoch": 1.3655168266821671, + "grad_norm": 1.650308609008789, + "learning_rate": 4.827839942166542e-06, + "loss": 3.2052, + "step": 17033 + }, + { + "epoch": 1.3655970014631897, + "grad_norm": 1.773093581199646, + "learning_rate": 4.826728703499153e-06, + "loss": 3.1478, + "step": 17034 + }, + { + "epoch": 1.3656771762442124, + "grad_norm": 1.6935452222824097, + "learning_rate": 4.825617552051295e-06, + "loss": 3.192, + "step": 17035 + }, + { + "epoch": 1.365757351025235, + "grad_norm": 1.7644790410995483, + "learning_rate": 4.824506487841702e-06, + "loss": 3.1162, + "step": 17036 + }, + { + "epoch": 1.3658375258062576, + "grad_norm": 1.777620792388916, + "learning_rate": 4.823395510889105e-06, + "loss": 3.2161, + "step": 17037 + }, + { + "epoch": 1.3659177005872802, + "grad_norm": 1.8152087926864624, + "learning_rate": 4.82228462121224e-06, + "loss": 3.2679, + "step": 17038 + }, + { + "epoch": 1.3659978753683029, + "grad_norm": 1.843441367149353, + "learning_rate": 4.8211738188298265e-06, + "loss": 3.1295, + "step": 17039 + }, + { + "epoch": 1.3660780501493255, + "grad_norm": 1.7458546161651611, + "learning_rate": 4.820063103760599e-06, + "loss": 3.2, + "step": 17040 + }, + { + "epoch": 1.366158224930348, + "grad_norm": 0.6807805299758911, + "learning_rate": 4.81895247602328e-06, + "loss": 2.4639, + "step": 17041 + }, + { + "epoch": 1.3662383997113707, + "grad_norm": 1.7707324028015137, + "learning_rate": 4.817841935636597e-06, + "loss": 3.1385, + "step": 17042 + }, + { + "epoch": 1.3663185744923934, + "grad_norm": 1.7156413793563843, + "learning_rate": 4.8167314826192756e-06, + "loss": 3.1195, + "step": 17043 + }, + { + "epoch": 1.366398749273416, + "grad_norm": 1.8114354610443115, + "learning_rate": 4.815621116990032e-06, + "loss": 3.1312, + "step": 17044 + }, + { + "epoch": 1.3664789240544386, + "grad_norm": 1.6723281145095825, + "learning_rate": 4.814510838767589e-06, + "loss": 3.1445, + "step": 17045 + }, + { + "epoch": 1.3665590988354612, + "grad_norm": 1.8084800243377686, + "learning_rate": 4.813400647970665e-06, + "loss": 3.1522, + "step": 17046 + }, + { + "epoch": 1.3666392736164839, + "grad_norm": 1.7820786237716675, + "learning_rate": 4.812290544617979e-06, + "loss": 3.1331, + "step": 17047 + }, + { + "epoch": 1.3667194483975065, + "grad_norm": 1.789526343345642, + "learning_rate": 4.8111805287282496e-06, + "loss": 3.1904, + "step": 17048 + }, + { + "epoch": 1.366799623178529, + "grad_norm": 1.771315097808838, + "learning_rate": 4.8100706003201835e-06, + "loss": 3.1177, + "step": 17049 + }, + { + "epoch": 1.3668797979595517, + "grad_norm": 1.8112596273422241, + "learning_rate": 4.808960759412499e-06, + "loss": 3.2517, + "step": 17050 + }, + { + "epoch": 1.3669599727405743, + "grad_norm": 1.8225958347320557, + "learning_rate": 4.8078510060239055e-06, + "loss": 3.183, + "step": 17051 + }, + { + "epoch": 1.367040147521597, + "grad_norm": 1.8228914737701416, + "learning_rate": 4.806741340173115e-06, + "loss": 3.2118, + "step": 17052 + }, + { + "epoch": 1.3671203223026196, + "grad_norm": 1.8723397254943848, + "learning_rate": 4.805631761878839e-06, + "loss": 3.1462, + "step": 17053 + }, + { + "epoch": 1.3672004970836422, + "grad_norm": 1.7824169397354126, + "learning_rate": 4.804522271159776e-06, + "loss": 3.1802, + "step": 17054 + }, + { + "epoch": 1.3672806718646648, + "grad_norm": 1.7578816413879395, + "learning_rate": 4.803412868034637e-06, + "loss": 3.1354, + "step": 17055 + }, + { + "epoch": 1.3673608466456875, + "grad_norm": 1.703190565109253, + "learning_rate": 4.802303552522126e-06, + "loss": 3.1205, + "step": 17056 + }, + { + "epoch": 1.36744102142671, + "grad_norm": 1.8885939121246338, + "learning_rate": 4.801194324640946e-06, + "loss": 3.2155, + "step": 17057 + }, + { + "epoch": 1.367521196207733, + "grad_norm": 1.6992934942245483, + "learning_rate": 4.800085184409799e-06, + "loss": 3.1213, + "step": 17058 + }, + { + "epoch": 1.3676013709887556, + "grad_norm": 1.7324014902114868, + "learning_rate": 4.79897613184738e-06, + "loss": 3.1526, + "step": 17059 + }, + { + "epoch": 1.3676815457697782, + "grad_norm": 1.7671796083450317, + "learning_rate": 4.797867166972392e-06, + "loss": 3.184, + "step": 17060 + }, + { + "epoch": 1.3677617205508008, + "grad_norm": 1.7274259328842163, + "learning_rate": 4.7967582898035295e-06, + "loss": 3.0731, + "step": 17061 + }, + { + "epoch": 1.3678418953318234, + "grad_norm": 1.8249229192733765, + "learning_rate": 4.7956495003594874e-06, + "loss": 3.0807, + "step": 17062 + }, + { + "epoch": 1.367922070112846, + "grad_norm": 1.8190587759017944, + "learning_rate": 4.794540798658966e-06, + "loss": 3.1777, + "step": 17063 + }, + { + "epoch": 1.3680022448938687, + "grad_norm": 1.822535753250122, + "learning_rate": 4.793432184720647e-06, + "loss": 3.1256, + "step": 17064 + }, + { + "epoch": 1.3680824196748913, + "grad_norm": 1.916639804840088, + "learning_rate": 4.792323658563226e-06, + "loss": 3.1666, + "step": 17065 + }, + { + "epoch": 1.368162594455914, + "grad_norm": 1.6761668920516968, + "learning_rate": 4.791215220205394e-06, + "loss": 3.1763, + "step": 17066 + }, + { + "epoch": 1.3682427692369366, + "grad_norm": 1.8748058080673218, + "learning_rate": 4.790106869665837e-06, + "loss": 3.1835, + "step": 17067 + }, + { + "epoch": 1.3683229440179592, + "grad_norm": 1.827406644821167, + "learning_rate": 4.7889986069632455e-06, + "loss": 3.1378, + "step": 17068 + }, + { + "epoch": 1.3684031187989818, + "grad_norm": 1.7317631244659424, + "learning_rate": 4.787890432116298e-06, + "loss": 3.2167, + "step": 17069 + }, + { + "epoch": 1.3684832935800044, + "grad_norm": 1.8254597187042236, + "learning_rate": 4.78678234514368e-06, + "loss": 3.2294, + "step": 17070 + }, + { + "epoch": 1.368563468361027, + "grad_norm": 0.6871566772460938, + "learning_rate": 4.785674346064075e-06, + "loss": 2.447, + "step": 17071 + }, + { + "epoch": 1.3686436431420497, + "grad_norm": 1.9006636142730713, + "learning_rate": 4.784566434896163e-06, + "loss": 3.2085, + "step": 17072 + }, + { + "epoch": 1.3687238179230723, + "grad_norm": 1.8659228086471558, + "learning_rate": 4.7834586116586265e-06, + "loss": 3.2121, + "step": 17073 + }, + { + "epoch": 1.368803992704095, + "grad_norm": 1.775740385055542, + "learning_rate": 4.782350876370135e-06, + "loss": 3.16, + "step": 17074 + }, + { + "epoch": 1.3688841674851175, + "grad_norm": 1.6316266059875488, + "learning_rate": 4.781243229049369e-06, + "loss": 3.121, + "step": 17075 + }, + { + "epoch": 1.3689643422661402, + "grad_norm": 1.714156150817871, + "learning_rate": 4.780135669715003e-06, + "loss": 3.1312, + "step": 17076 + }, + { + "epoch": 1.3690445170471628, + "grad_norm": 1.7358450889587402, + "learning_rate": 4.77902819838571e-06, + "loss": 3.2159, + "step": 17077 + }, + { + "epoch": 1.3691246918281854, + "grad_norm": 1.782637596130371, + "learning_rate": 4.777920815080166e-06, + "loss": 3.1522, + "step": 17078 + }, + { + "epoch": 1.369204866609208, + "grad_norm": 1.7190816402435303, + "learning_rate": 4.776813519817032e-06, + "loss": 3.2247, + "step": 17079 + }, + { + "epoch": 1.3692850413902307, + "grad_norm": 1.7854101657867432, + "learning_rate": 4.775706312614981e-06, + "loss": 3.1899, + "step": 17080 + }, + { + "epoch": 1.3693652161712533, + "grad_norm": 1.7803733348846436, + "learning_rate": 4.774599193492681e-06, + "loss": 3.1825, + "step": 17081 + }, + { + "epoch": 1.369445390952276, + "grad_norm": 1.8022074699401855, + "learning_rate": 4.773492162468797e-06, + "loss": 3.1227, + "step": 17082 + }, + { + "epoch": 1.3695255657332985, + "grad_norm": 1.7839622497558594, + "learning_rate": 4.7723852195619955e-06, + "loss": 3.2327, + "step": 17083 + }, + { + "epoch": 1.3696057405143212, + "grad_norm": 1.9335641860961914, + "learning_rate": 4.771278364790934e-06, + "loss": 3.2401, + "step": 17084 + }, + { + "epoch": 1.3696859152953438, + "grad_norm": 1.8077826499938965, + "learning_rate": 4.770171598174278e-06, + "loss": 3.2057, + "step": 17085 + }, + { + "epoch": 1.3697660900763664, + "grad_norm": 1.6700156927108765, + "learning_rate": 4.769064919730684e-06, + "loss": 3.2676, + "step": 17086 + }, + { + "epoch": 1.369846264857389, + "grad_norm": 1.626179575920105, + "learning_rate": 4.767958329478812e-06, + "loss": 3.0462, + "step": 17087 + }, + { + "epoch": 1.3699264396384119, + "grad_norm": 1.8147380352020264, + "learning_rate": 4.766851827437323e-06, + "loss": 3.2565, + "step": 17088 + }, + { + "epoch": 1.3700066144194345, + "grad_norm": 1.7557258605957031, + "learning_rate": 4.765745413624864e-06, + "loss": 3.2107, + "step": 17089 + }, + { + "epoch": 1.3700867892004571, + "grad_norm": 1.7448588609695435, + "learning_rate": 4.7646390880600915e-06, + "loss": 3.1095, + "step": 17090 + }, + { + "epoch": 1.3701669639814797, + "grad_norm": 1.770959734916687, + "learning_rate": 4.763532850761659e-06, + "loss": 3.0967, + "step": 17091 + }, + { + "epoch": 1.3702471387625024, + "grad_norm": 1.7286763191223145, + "learning_rate": 4.762426701748219e-06, + "loss": 3.1539, + "step": 17092 + }, + { + "epoch": 1.370327313543525, + "grad_norm": 1.7226061820983887, + "learning_rate": 4.761320641038422e-06, + "loss": 3.1609, + "step": 17093 + }, + { + "epoch": 1.3704074883245476, + "grad_norm": 1.7395247220993042, + "learning_rate": 4.7602146686509076e-06, + "loss": 3.1542, + "step": 17094 + }, + { + "epoch": 1.3704876631055702, + "grad_norm": 1.7884366512298584, + "learning_rate": 4.759108784604328e-06, + "loss": 3.1993, + "step": 17095 + }, + { + "epoch": 1.3705678378865929, + "grad_norm": 1.7017314434051514, + "learning_rate": 4.758002988917328e-06, + "loss": 3.0775, + "step": 17096 + }, + { + "epoch": 1.3706480126676155, + "grad_norm": 1.7944447994232178, + "learning_rate": 4.756897281608549e-06, + "loss": 3.1738, + "step": 17097 + }, + { + "epoch": 1.3707281874486381, + "grad_norm": 1.8387234210968018, + "learning_rate": 4.755791662696637e-06, + "loss": 3.1776, + "step": 17098 + }, + { + "epoch": 1.3708083622296607, + "grad_norm": 1.691165566444397, + "learning_rate": 4.754686132200227e-06, + "loss": 3.1647, + "step": 17099 + }, + { + "epoch": 1.3708885370106834, + "grad_norm": 1.7932416200637817, + "learning_rate": 4.75358069013796e-06, + "loss": 3.1151, + "step": 17100 + }, + { + "epoch": 1.370968711791706, + "grad_norm": 1.7674567699432373, + "learning_rate": 4.752475336528473e-06, + "loss": 3.2643, + "step": 17101 + }, + { + "epoch": 1.3710488865727286, + "grad_norm": 1.8833856582641602, + "learning_rate": 4.751370071390403e-06, + "loss": 3.1893, + "step": 17102 + }, + { + "epoch": 1.3711290613537512, + "grad_norm": 1.858154296875, + "learning_rate": 4.750264894742388e-06, + "loss": 3.186, + "step": 17103 + }, + { + "epoch": 1.3712092361347739, + "grad_norm": 1.769945502281189, + "learning_rate": 4.749159806603052e-06, + "loss": 3.217, + "step": 17104 + }, + { + "epoch": 1.3712894109157965, + "grad_norm": 1.736750841140747, + "learning_rate": 4.748054806991031e-06, + "loss": 3.139, + "step": 17105 + }, + { + "epoch": 1.371369585696819, + "grad_norm": 1.7666584253311157, + "learning_rate": 4.746949895924956e-06, + "loss": 3.1875, + "step": 17106 + }, + { + "epoch": 1.3714497604778417, + "grad_norm": 1.8891931772232056, + "learning_rate": 4.745845073423454e-06, + "loss": 3.1338, + "step": 17107 + }, + { + "epoch": 1.3715299352588644, + "grad_norm": 1.8133056163787842, + "learning_rate": 4.744740339505155e-06, + "loss": 3.2578, + "step": 17108 + }, + { + "epoch": 1.371610110039887, + "grad_norm": 1.6869885921478271, + "learning_rate": 4.74363569418868e-06, + "loss": 3.1213, + "step": 17109 + }, + { + "epoch": 1.3716902848209096, + "grad_norm": 1.7848376035690308, + "learning_rate": 4.742531137492653e-06, + "loss": 3.0922, + "step": 17110 + }, + { + "epoch": 1.3717704596019322, + "grad_norm": 1.853726863861084, + "learning_rate": 4.741426669435698e-06, + "loss": 3.181, + "step": 17111 + }, + { + "epoch": 1.3718506343829548, + "grad_norm": 1.7441340684890747, + "learning_rate": 4.740322290036438e-06, + "loss": 3.2633, + "step": 17112 + }, + { + "epoch": 1.3719308091639775, + "grad_norm": 1.7984668016433716, + "learning_rate": 4.739217999313492e-06, + "loss": 3.1159, + "step": 17113 + }, + { + "epoch": 1.372010983945, + "grad_norm": 1.8837043046951294, + "learning_rate": 4.738113797285474e-06, + "loss": 3.2489, + "step": 17114 + }, + { + "epoch": 1.3720911587260227, + "grad_norm": 1.8937991857528687, + "learning_rate": 4.737009683971002e-06, + "loss": 3.2261, + "step": 17115 + }, + { + "epoch": 1.3721713335070453, + "grad_norm": 1.6734130382537842, + "learning_rate": 4.735905659388693e-06, + "loss": 3.2055, + "step": 17116 + }, + { + "epoch": 1.372251508288068, + "grad_norm": 1.7391031980514526, + "learning_rate": 4.734801723557158e-06, + "loss": 3.1612, + "step": 17117 + }, + { + "epoch": 1.3723316830690906, + "grad_norm": 1.733418345451355, + "learning_rate": 4.733697876495015e-06, + "loss": 3.1949, + "step": 17118 + }, + { + "epoch": 1.3724118578501132, + "grad_norm": 1.7774631977081299, + "learning_rate": 4.732594118220866e-06, + "loss": 3.1912, + "step": 17119 + }, + { + "epoch": 1.3724920326311358, + "grad_norm": 1.8174803256988525, + "learning_rate": 4.7314904487533244e-06, + "loss": 3.0807, + "step": 17120 + }, + { + "epoch": 1.3725722074121585, + "grad_norm": 0.6536170244216919, + "learning_rate": 4.730386868110995e-06, + "loss": 2.4722, + "step": 17121 + }, + { + "epoch": 1.372652382193181, + "grad_norm": 1.842818021774292, + "learning_rate": 4.729283376312488e-06, + "loss": 3.3047, + "step": 17122 + }, + { + "epoch": 1.3727325569742037, + "grad_norm": 1.6670798063278198, + "learning_rate": 4.728179973376408e-06, + "loss": 3.1161, + "step": 17123 + }, + { + "epoch": 1.3728127317552263, + "grad_norm": 1.6896601915359497, + "learning_rate": 4.727076659321353e-06, + "loss": 3.1335, + "step": 17124 + }, + { + "epoch": 1.372892906536249, + "grad_norm": 1.779837965965271, + "learning_rate": 4.725973434165925e-06, + "loss": 3.1859, + "step": 17125 + }, + { + "epoch": 1.3729730813172716, + "grad_norm": 1.793382167816162, + "learning_rate": 4.724870297928729e-06, + "loss": 3.2618, + "step": 17126 + }, + { + "epoch": 1.3730532560982942, + "grad_norm": 1.7205843925476074, + "learning_rate": 4.7237672506283595e-06, + "loss": 3.1369, + "step": 17127 + }, + { + "epoch": 1.3731334308793168, + "grad_norm": 1.7163726091384888, + "learning_rate": 4.722664292283417e-06, + "loss": 3.2422, + "step": 17128 + }, + { + "epoch": 1.3732136056603395, + "grad_norm": 1.8076528310775757, + "learning_rate": 4.721561422912493e-06, + "loss": 3.311, + "step": 17129 + }, + { + "epoch": 1.373293780441362, + "grad_norm": 1.7197054624557495, + "learning_rate": 4.720458642534183e-06, + "loss": 3.0858, + "step": 17130 + }, + { + "epoch": 1.3733739552223847, + "grad_norm": 1.8159385919570923, + "learning_rate": 4.719355951167078e-06, + "loss": 3.0612, + "step": 17131 + }, + { + "epoch": 1.3734541300034073, + "grad_norm": 1.7965210676193237, + "learning_rate": 4.718253348829771e-06, + "loss": 3.1565, + "step": 17132 + }, + { + "epoch": 1.37353430478443, + "grad_norm": 1.6948573589324951, + "learning_rate": 4.717150835540856e-06, + "loss": 3.1877, + "step": 17133 + }, + { + "epoch": 1.3736144795654526, + "grad_norm": 1.7332284450531006, + "learning_rate": 4.716048411318911e-06, + "loss": 3.0441, + "step": 17134 + }, + { + "epoch": 1.3736946543464752, + "grad_norm": 1.7448244094848633, + "learning_rate": 4.714946076182528e-06, + "loss": 3.1968, + "step": 17135 + }, + { + "epoch": 1.3737748291274978, + "grad_norm": 1.7188210487365723, + "learning_rate": 4.713843830150292e-06, + "loss": 3.1723, + "step": 17136 + }, + { + "epoch": 1.3738550039085204, + "grad_norm": 1.7991267442703247, + "learning_rate": 4.712741673240785e-06, + "loss": 3.2039, + "step": 17137 + }, + { + "epoch": 1.373935178689543, + "grad_norm": 1.7684268951416016, + "learning_rate": 4.711639605472595e-06, + "loss": 3.2684, + "step": 17138 + }, + { + "epoch": 1.374015353470566, + "grad_norm": 0.6783658862113953, + "learning_rate": 4.710537626864292e-06, + "loss": 2.4678, + "step": 17139 + }, + { + "epoch": 1.3740955282515885, + "grad_norm": 1.7398262023925781, + "learning_rate": 4.709435737434462e-06, + "loss": 3.2819, + "step": 17140 + }, + { + "epoch": 1.3741757030326112, + "grad_norm": 1.853499174118042, + "learning_rate": 4.70833393720168e-06, + "loss": 3.2105, + "step": 17141 + }, + { + "epoch": 1.3742558778136338, + "grad_norm": 1.8163189888000488, + "learning_rate": 4.707232226184524e-06, + "loss": 3.2872, + "step": 17142 + }, + { + "epoch": 1.3743360525946564, + "grad_norm": 1.9117807149887085, + "learning_rate": 4.70613060440157e-06, + "loss": 3.2423, + "step": 17143 + }, + { + "epoch": 1.374416227375679, + "grad_norm": 1.6841871738433838, + "learning_rate": 4.705029071871384e-06, + "loss": 3.2254, + "step": 17144 + }, + { + "epoch": 1.3744964021567017, + "grad_norm": 1.8030999898910522, + "learning_rate": 4.7039276286125436e-06, + "loss": 3.1262, + "step": 17145 + }, + { + "epoch": 1.3745765769377243, + "grad_norm": 1.7605353593826294, + "learning_rate": 4.702826274643615e-06, + "loss": 3.1857, + "step": 17146 + }, + { + "epoch": 1.374656751718747, + "grad_norm": 1.8321086168289185, + "learning_rate": 4.70172500998317e-06, + "loss": 3.1787, + "step": 17147 + }, + { + "epoch": 1.3747369264997695, + "grad_norm": 0.66685551404953, + "learning_rate": 4.700623834649777e-06, + "loss": 2.4617, + "step": 17148 + }, + { + "epoch": 1.3748171012807922, + "grad_norm": 1.7266497611999512, + "learning_rate": 4.699522748661994e-06, + "loss": 3.0934, + "step": 17149 + }, + { + "epoch": 1.3748972760618148, + "grad_norm": 1.7943634986877441, + "learning_rate": 4.69842175203839e-06, + "loss": 3.1969, + "step": 17150 + }, + { + "epoch": 1.3749774508428374, + "grad_norm": 1.7126659154891968, + "learning_rate": 4.697320844797527e-06, + "loss": 3.2043, + "step": 17151 + }, + { + "epoch": 1.37505762562386, + "grad_norm": 1.7974236011505127, + "learning_rate": 4.696220026957966e-06, + "loss": 3.2437, + "step": 17152 + }, + { + "epoch": 1.3751378004048826, + "grad_norm": 1.7386468648910522, + "learning_rate": 4.69511929853827e-06, + "loss": 3.1433, + "step": 17153 + }, + { + "epoch": 1.3752179751859053, + "grad_norm": 1.9614588022232056, + "learning_rate": 4.694018659556988e-06, + "loss": 3.2219, + "step": 17154 + }, + { + "epoch": 1.375298149966928, + "grad_norm": 1.7521696090698242, + "learning_rate": 4.692918110032683e-06, + "loss": 3.1172, + "step": 17155 + }, + { + "epoch": 1.3753783247479505, + "grad_norm": 1.8370431661605835, + "learning_rate": 4.691817649983907e-06, + "loss": 3.1966, + "step": 17156 + }, + { + "epoch": 1.3754584995289731, + "grad_norm": 1.912663459777832, + "learning_rate": 4.690717279429216e-06, + "loss": 3.2123, + "step": 17157 + }, + { + "epoch": 1.3755386743099958, + "grad_norm": 1.6866534948349, + "learning_rate": 4.6896169983871635e-06, + "loss": 3.0815, + "step": 17158 + }, + { + "epoch": 1.3756188490910184, + "grad_norm": 1.8317811489105225, + "learning_rate": 4.688516806876292e-06, + "loss": 3.0228, + "step": 17159 + }, + { + "epoch": 1.375699023872041, + "grad_norm": 1.7570096254348755, + "learning_rate": 4.687416704915156e-06, + "loss": 3.1136, + "step": 17160 + }, + { + "epoch": 1.3757791986530636, + "grad_norm": 1.7620030641555786, + "learning_rate": 4.686316692522303e-06, + "loss": 3.1176, + "step": 17161 + }, + { + "epoch": 1.3758593734340863, + "grad_norm": 1.8597145080566406, + "learning_rate": 4.685216769716276e-06, + "loss": 3.176, + "step": 17162 + }, + { + "epoch": 1.3759395482151089, + "grad_norm": 1.8154010772705078, + "learning_rate": 4.684116936515626e-06, + "loss": 3.2499, + "step": 17163 + }, + { + "epoch": 1.3760197229961315, + "grad_norm": 1.7700507640838623, + "learning_rate": 4.6830171929388865e-06, + "loss": 3.1991, + "step": 17164 + }, + { + "epoch": 1.3760998977771541, + "grad_norm": 1.8570008277893066, + "learning_rate": 4.681917539004605e-06, + "loss": 3.0466, + "step": 17165 + }, + { + "epoch": 1.3761800725581768, + "grad_norm": 1.8846650123596191, + "learning_rate": 4.6808179747313176e-06, + "loss": 3.223, + "step": 17166 + }, + { + "epoch": 1.3762602473391994, + "grad_norm": 1.7465190887451172, + "learning_rate": 4.679718500137566e-06, + "loss": 3.1656, + "step": 17167 + }, + { + "epoch": 1.376340422120222, + "grad_norm": 1.6452170610427856, + "learning_rate": 4.6786191152418884e-06, + "loss": 3.2379, + "step": 17168 + }, + { + "epoch": 1.3764205969012446, + "grad_norm": 1.8712902069091797, + "learning_rate": 4.677519820062813e-06, + "loss": 3.301, + "step": 17169 + }, + { + "epoch": 1.3765007716822675, + "grad_norm": 1.7781747579574585, + "learning_rate": 4.676420614618877e-06, + "loss": 3.1519, + "step": 17170 + }, + { + "epoch": 1.37658094646329, + "grad_norm": 1.8711086511611938, + "learning_rate": 4.675321498928615e-06, + "loss": 3.1765, + "step": 17171 + }, + { + "epoch": 1.3766611212443127, + "grad_norm": 0.6548333764076233, + "learning_rate": 4.674222473010555e-06, + "loss": 2.4539, + "step": 17172 + }, + { + "epoch": 1.3767412960253353, + "grad_norm": 1.8015241622924805, + "learning_rate": 4.673123536883231e-06, + "loss": 3.1603, + "step": 17173 + }, + { + "epoch": 1.376821470806358, + "grad_norm": 1.822595477104187, + "learning_rate": 4.672024690565163e-06, + "loss": 3.2823, + "step": 17174 + }, + { + "epoch": 1.3769016455873806, + "grad_norm": 1.7246493101119995, + "learning_rate": 4.670925934074881e-06, + "loss": 3.1748, + "step": 17175 + }, + { + "epoch": 1.3769818203684032, + "grad_norm": 1.8254722356796265, + "learning_rate": 4.66982726743091e-06, + "loss": 3.3055, + "step": 17176 + }, + { + "epoch": 1.3770619951494258, + "grad_norm": 1.7358710765838623, + "learning_rate": 4.668728690651771e-06, + "loss": 3.2084, + "step": 17177 + }, + { + "epoch": 1.3771421699304485, + "grad_norm": 1.7676738500595093, + "learning_rate": 4.6676302037559926e-06, + "loss": 3.2252, + "step": 17178 + }, + { + "epoch": 1.377222344711471, + "grad_norm": 1.8537571430206299, + "learning_rate": 4.666531806762087e-06, + "loss": 3.2177, + "step": 17179 + }, + { + "epoch": 1.3773025194924937, + "grad_norm": 1.830965518951416, + "learning_rate": 4.665433499688573e-06, + "loss": 3.2152, + "step": 17180 + }, + { + "epoch": 1.3773826942735163, + "grad_norm": 1.7022665739059448, + "learning_rate": 4.6643352825539725e-06, + "loss": 3.1293, + "step": 17181 + }, + { + "epoch": 1.377462869054539, + "grad_norm": 1.762953519821167, + "learning_rate": 4.6632371553767966e-06, + "loss": 3.1017, + "step": 17182 + }, + { + "epoch": 1.3775430438355616, + "grad_norm": 1.77438485622406, + "learning_rate": 4.662139118175566e-06, + "loss": 3.2498, + "step": 17183 + }, + { + "epoch": 1.3776232186165842, + "grad_norm": 1.8545175790786743, + "learning_rate": 4.6610411709687854e-06, + "loss": 3.1755, + "step": 17184 + }, + { + "epoch": 1.3777033933976068, + "grad_norm": 1.7412916421890259, + "learning_rate": 4.659943313774968e-06, + "loss": 3.2308, + "step": 17185 + }, + { + "epoch": 1.3777835681786295, + "grad_norm": 1.9300041198730469, + "learning_rate": 4.658845546612625e-06, + "loss": 3.1213, + "step": 17186 + }, + { + "epoch": 1.377863742959652, + "grad_norm": 1.8664451837539673, + "learning_rate": 4.657747869500263e-06, + "loss": 3.1927, + "step": 17187 + }, + { + "epoch": 1.3779439177406747, + "grad_norm": 1.7994129657745361, + "learning_rate": 4.6566502824563946e-06, + "loss": 3.2058, + "step": 17188 + }, + { + "epoch": 1.3780240925216973, + "grad_norm": 1.7166391611099243, + "learning_rate": 4.655552785499513e-06, + "loss": 3.1871, + "step": 17189 + }, + { + "epoch": 1.37810426730272, + "grad_norm": 1.8142050504684448, + "learning_rate": 4.654455378648129e-06, + "loss": 3.2551, + "step": 17190 + }, + { + "epoch": 1.3781844420837426, + "grad_norm": 1.8355679512023926, + "learning_rate": 4.653358061920744e-06, + "loss": 3.1136, + "step": 17191 + }, + { + "epoch": 1.3782646168647652, + "grad_norm": 1.808386206626892, + "learning_rate": 4.652260835335858e-06, + "loss": 3.2462, + "step": 17192 + }, + { + "epoch": 1.3783447916457878, + "grad_norm": 1.8300321102142334, + "learning_rate": 4.651163698911972e-06, + "loss": 3.2513, + "step": 17193 + }, + { + "epoch": 1.3784249664268104, + "grad_norm": 0.695613443851471, + "learning_rate": 4.6500666526675775e-06, + "loss": 2.455, + "step": 17194 + }, + { + "epoch": 1.378505141207833, + "grad_norm": 1.8147149085998535, + "learning_rate": 4.648969696621174e-06, + "loss": 3.2415, + "step": 17195 + }, + { + "epoch": 1.3785853159888557, + "grad_norm": 1.833855390548706, + "learning_rate": 4.647872830791256e-06, + "loss": 3.1861, + "step": 17196 + }, + { + "epoch": 1.3786654907698783, + "grad_norm": 1.6927379369735718, + "learning_rate": 4.6467760551963145e-06, + "loss": 3.1424, + "step": 17197 + }, + { + "epoch": 1.378745665550901, + "grad_norm": 1.86545991897583, + "learning_rate": 4.645679369854847e-06, + "loss": 3.1962, + "step": 17198 + }, + { + "epoch": 1.3788258403319236, + "grad_norm": 1.9095503091812134, + "learning_rate": 4.644582774785334e-06, + "loss": 3.3298, + "step": 17199 + }, + { + "epoch": 1.3789060151129462, + "grad_norm": 1.787072777748108, + "learning_rate": 4.643486270006269e-06, + "loss": 3.1716, + "step": 17200 + }, + { + "epoch": 1.3789861898939688, + "grad_norm": 1.7926056385040283, + "learning_rate": 4.642389855536135e-06, + "loss": 3.1313, + "step": 17201 + }, + { + "epoch": 1.3790663646749914, + "grad_norm": 1.7740172147750854, + "learning_rate": 4.6412935313934225e-06, + "loss": 3.2761, + "step": 17202 + }, + { + "epoch": 1.379146539456014, + "grad_norm": 1.7792190313339233, + "learning_rate": 4.640197297596616e-06, + "loss": 3.2093, + "step": 17203 + }, + { + "epoch": 1.3792267142370367, + "grad_norm": 1.8578860759735107, + "learning_rate": 4.63910115416419e-06, + "loss": 3.1835, + "step": 17204 + }, + { + "epoch": 1.3793068890180593, + "grad_norm": 1.789954662322998, + "learning_rate": 4.638005101114627e-06, + "loss": 3.1821, + "step": 17205 + }, + { + "epoch": 1.379387063799082, + "grad_norm": 1.6573808193206787, + "learning_rate": 4.636909138466411e-06, + "loss": 3.1908, + "step": 17206 + }, + { + "epoch": 1.3794672385801046, + "grad_norm": 1.718371868133545, + "learning_rate": 4.6358132662380165e-06, + "loss": 3.291, + "step": 17207 + }, + { + "epoch": 1.3795474133611272, + "grad_norm": 0.7236756086349487, + "learning_rate": 4.634717484447922e-06, + "loss": 2.4563, + "step": 17208 + }, + { + "epoch": 1.3796275881421498, + "grad_norm": 1.9272308349609375, + "learning_rate": 4.633621793114597e-06, + "loss": 3.1888, + "step": 17209 + }, + { + "epoch": 1.3797077629231724, + "grad_norm": 1.8629374504089355, + "learning_rate": 4.632526192256518e-06, + "loss": 3.1434, + "step": 17210 + }, + { + "epoch": 1.379787937704195, + "grad_norm": 1.602739691734314, + "learning_rate": 4.631430681892155e-06, + "loss": 3.1766, + "step": 17211 + }, + { + "epoch": 1.3798681124852177, + "grad_norm": 1.7496824264526367, + "learning_rate": 4.630335262039979e-06, + "loss": 3.182, + "step": 17212 + }, + { + "epoch": 1.3799482872662403, + "grad_norm": 1.774538516998291, + "learning_rate": 4.629239932718461e-06, + "loss": 3.0957, + "step": 17213 + }, + { + "epoch": 1.380028462047263, + "grad_norm": 1.8059340715408325, + "learning_rate": 4.628144693946061e-06, + "loss": 3.1661, + "step": 17214 + }, + { + "epoch": 1.3801086368282856, + "grad_norm": 1.8005605936050415, + "learning_rate": 4.627049545741249e-06, + "loss": 3.1542, + "step": 17215 + }, + { + "epoch": 1.3801888116093082, + "grad_norm": 1.7088080644607544, + "learning_rate": 4.625954488122488e-06, + "loss": 3.1821, + "step": 17216 + }, + { + "epoch": 1.3802689863903308, + "grad_norm": 1.7532999515533447, + "learning_rate": 4.62485952110824e-06, + "loss": 3.1853, + "step": 17217 + }, + { + "epoch": 1.3803491611713534, + "grad_norm": 1.869794249534607, + "learning_rate": 4.623764644716971e-06, + "loss": 3.1356, + "step": 17218 + }, + { + "epoch": 1.380429335952376, + "grad_norm": 0.6518619656562805, + "learning_rate": 4.622669858967131e-06, + "loss": 2.4657, + "step": 17219 + }, + { + "epoch": 1.380509510733399, + "grad_norm": 1.8215254545211792, + "learning_rate": 4.621575163877181e-06, + "loss": 3.1914, + "step": 17220 + }, + { + "epoch": 1.3805896855144215, + "grad_norm": 1.8164737224578857, + "learning_rate": 4.620480559465579e-06, + "loss": 3.185, + "step": 17221 + }, + { + "epoch": 1.3806698602954441, + "grad_norm": 1.7231591939926147, + "learning_rate": 4.619386045750778e-06, + "loss": 3.1155, + "step": 17222 + }, + { + "epoch": 1.3807500350764668, + "grad_norm": 1.7435508966445923, + "learning_rate": 4.618291622751237e-06, + "loss": 3.1876, + "step": 17223 + }, + { + "epoch": 1.3808302098574894, + "grad_norm": 1.7163910865783691, + "learning_rate": 4.617197290485398e-06, + "loss": 3.2191, + "step": 17224 + }, + { + "epoch": 1.380910384638512, + "grad_norm": 1.8033870458602905, + "learning_rate": 4.616103048971715e-06, + "loss": 3.1824, + "step": 17225 + }, + { + "epoch": 1.3809905594195346, + "grad_norm": 0.6852205395698547, + "learning_rate": 4.615008898228638e-06, + "loss": 2.4858, + "step": 17226 + }, + { + "epoch": 1.3810707342005573, + "grad_norm": 1.7408710718154907, + "learning_rate": 4.613914838274614e-06, + "loss": 3.1885, + "step": 17227 + }, + { + "epoch": 1.3811509089815799, + "grad_norm": 1.7570912837982178, + "learning_rate": 4.6128208691280895e-06, + "loss": 3.1307, + "step": 17228 + }, + { + "epoch": 1.3812310837626025, + "grad_norm": 1.8131271600723267, + "learning_rate": 4.611726990807504e-06, + "loss": 3.1443, + "step": 17229 + }, + { + "epoch": 1.3813112585436251, + "grad_norm": 1.848397970199585, + "learning_rate": 4.6106332033313015e-06, + "loss": 3.1714, + "step": 17230 + }, + { + "epoch": 1.3813914333246478, + "grad_norm": 1.839152216911316, + "learning_rate": 4.6095395067179236e-06, + "loss": 3.2036, + "step": 17231 + }, + { + "epoch": 1.3814716081056704, + "grad_norm": 1.7606545686721802, + "learning_rate": 4.6084459009858115e-06, + "loss": 3.1303, + "step": 17232 + }, + { + "epoch": 1.381551782886693, + "grad_norm": 1.8182247877120972, + "learning_rate": 4.6073523861534044e-06, + "loss": 3.2365, + "step": 17233 + }, + { + "epoch": 1.3816319576677156, + "grad_norm": 1.8003368377685547, + "learning_rate": 4.60625896223913e-06, + "loss": 3.1547, + "step": 17234 + }, + { + "epoch": 1.3817121324487383, + "grad_norm": 1.7884899377822876, + "learning_rate": 4.605165629261431e-06, + "loss": 3.2579, + "step": 17235 + }, + { + "epoch": 1.3817923072297609, + "grad_norm": 1.9059648513793945, + "learning_rate": 4.6040723872387375e-06, + "loss": 3.3601, + "step": 17236 + }, + { + "epoch": 1.3818724820107835, + "grad_norm": 1.7536540031433105, + "learning_rate": 4.602979236189481e-06, + "loss": 3.2367, + "step": 17237 + }, + { + "epoch": 1.3819526567918061, + "grad_norm": 1.8515666723251343, + "learning_rate": 4.601886176132098e-06, + "loss": 3.2284, + "step": 17238 + }, + { + "epoch": 1.3820328315728287, + "grad_norm": 1.7787079811096191, + "learning_rate": 4.600793207085006e-06, + "loss": 3.095, + "step": 17239 + }, + { + "epoch": 1.3821130063538514, + "grad_norm": 1.8356605768203735, + "learning_rate": 4.599700329066639e-06, + "loss": 3.1412, + "step": 17240 + }, + { + "epoch": 1.382193181134874, + "grad_norm": 1.8629777431488037, + "learning_rate": 4.59860754209542e-06, + "loss": 3.1339, + "step": 17241 + }, + { + "epoch": 1.3822733559158966, + "grad_norm": 1.7585053443908691, + "learning_rate": 4.597514846189775e-06, + "loss": 3.2771, + "step": 17242 + }, + { + "epoch": 1.3823535306969192, + "grad_norm": 1.74276602268219, + "learning_rate": 4.596422241368128e-06, + "loss": 3.1784, + "step": 17243 + }, + { + "epoch": 1.3824337054779419, + "grad_norm": 1.8915152549743652, + "learning_rate": 4.595329727648895e-06, + "loss": 3.233, + "step": 17244 + }, + { + "epoch": 1.3825138802589645, + "grad_norm": 1.736246109008789, + "learning_rate": 4.5942373050504985e-06, + "loss": 3.2781, + "step": 17245 + }, + { + "epoch": 1.3825940550399871, + "grad_norm": 1.7957298755645752, + "learning_rate": 4.593144973591356e-06, + "loss": 3.2593, + "step": 17246 + }, + { + "epoch": 1.3826742298210097, + "grad_norm": 1.661818504333496, + "learning_rate": 4.592052733289883e-06, + "loss": 3.1108, + "step": 17247 + }, + { + "epoch": 1.3827544046020324, + "grad_norm": 1.696067452430725, + "learning_rate": 4.590960584164498e-06, + "loss": 3.0179, + "step": 17248 + }, + { + "epoch": 1.382834579383055, + "grad_norm": 1.7212131023406982, + "learning_rate": 4.5898685262336086e-06, + "loss": 3.1102, + "step": 17249 + }, + { + "epoch": 1.3829147541640776, + "grad_norm": 1.8109208345413208, + "learning_rate": 4.588776559515629e-06, + "loss": 3.1399, + "step": 17250 + }, + { + "epoch": 1.3829949289451005, + "grad_norm": 0.7269354462623596, + "learning_rate": 4.5876846840289694e-06, + "loss": 2.485, + "step": 17251 + }, + { + "epoch": 1.383075103726123, + "grad_norm": 0.6856929063796997, + "learning_rate": 4.5865928997920385e-06, + "loss": 2.475, + "step": 17252 + }, + { + "epoch": 1.3831552785071457, + "grad_norm": 1.743263602256775, + "learning_rate": 4.5855012068232465e-06, + "loss": 3.137, + "step": 17253 + }, + { + "epoch": 1.3832354532881683, + "grad_norm": 1.7257870435714722, + "learning_rate": 4.5844096051409926e-06, + "loss": 3.1059, + "step": 17254 + }, + { + "epoch": 1.383315628069191, + "grad_norm": 1.7427500486373901, + "learning_rate": 4.5833180947636845e-06, + "loss": 3.1142, + "step": 17255 + }, + { + "epoch": 1.3833958028502136, + "grad_norm": 1.9273464679718018, + "learning_rate": 4.582226675709723e-06, + "loss": 3.2024, + "step": 17256 + }, + { + "epoch": 1.3834759776312362, + "grad_norm": 0.679093062877655, + "learning_rate": 4.581135347997512e-06, + "loss": 2.4774, + "step": 17257 + }, + { + "epoch": 1.3835561524122588, + "grad_norm": 1.7207905054092407, + "learning_rate": 4.580044111645453e-06, + "loss": 3.0582, + "step": 17258 + }, + { + "epoch": 1.3836363271932814, + "grad_norm": 1.8094509840011597, + "learning_rate": 4.5789529666719355e-06, + "loss": 3.1849, + "step": 17259 + }, + { + "epoch": 1.383716501974304, + "grad_norm": 0.6648291945457458, + "learning_rate": 4.577861913095361e-06, + "loss": 2.4725, + "step": 17260 + }, + { + "epoch": 1.3837966767553267, + "grad_norm": 1.8037792444229126, + "learning_rate": 4.576770950934123e-06, + "loss": 3.2485, + "step": 17261 + }, + { + "epoch": 1.3838768515363493, + "grad_norm": 1.7261319160461426, + "learning_rate": 4.575680080206616e-06, + "loss": 3.0591, + "step": 17262 + }, + { + "epoch": 1.383957026317372, + "grad_norm": 1.696745753288269, + "learning_rate": 4.574589300931235e-06, + "loss": 3.0717, + "step": 17263 + }, + { + "epoch": 1.3840372010983946, + "grad_norm": 1.7868565320968628, + "learning_rate": 4.5734986131263615e-06, + "loss": 3.2186, + "step": 17264 + }, + { + "epoch": 1.3841173758794172, + "grad_norm": 1.7248239517211914, + "learning_rate": 4.57240801681039e-06, + "loss": 3.1879, + "step": 17265 + }, + { + "epoch": 1.3841975506604398, + "grad_norm": 1.7689006328582764, + "learning_rate": 4.5713175120017065e-06, + "loss": 3.1533, + "step": 17266 + }, + { + "epoch": 1.3842777254414624, + "grad_norm": 1.7030961513519287, + "learning_rate": 4.570227098718696e-06, + "loss": 3.0569, + "step": 17267 + }, + { + "epoch": 1.384357900222485, + "grad_norm": 1.753186583518982, + "learning_rate": 4.5691367769797465e-06, + "loss": 3.1521, + "step": 17268 + }, + { + "epoch": 1.3844380750035077, + "grad_norm": 1.7527817487716675, + "learning_rate": 4.568046546803233e-06, + "loss": 3.1893, + "step": 17269 + }, + { + "epoch": 1.3845182497845303, + "grad_norm": 1.741212248802185, + "learning_rate": 4.566956408207542e-06, + "loss": 3.1329, + "step": 17270 + }, + { + "epoch": 1.384598424565553, + "grad_norm": 1.8816370964050293, + "learning_rate": 4.5658663612110485e-06, + "loss": 3.2018, + "step": 17271 + }, + { + "epoch": 1.3846785993465756, + "grad_norm": 1.9105597734451294, + "learning_rate": 4.564776405832135e-06, + "loss": 3.156, + "step": 17272 + }, + { + "epoch": 1.3847587741275982, + "grad_norm": 1.7955964803695679, + "learning_rate": 4.563686542089179e-06, + "loss": 3.2356, + "step": 17273 + }, + { + "epoch": 1.3848389489086208, + "grad_norm": 1.7385175228118896, + "learning_rate": 4.562596770000548e-06, + "loss": 3.1485, + "step": 17274 + }, + { + "epoch": 1.3849191236896434, + "grad_norm": 1.8542808294296265, + "learning_rate": 4.561507089584619e-06, + "loss": 3.2062, + "step": 17275 + }, + { + "epoch": 1.384999298470666, + "grad_norm": 1.8279951810836792, + "learning_rate": 4.560417500859764e-06, + "loss": 3.1894, + "step": 17276 + }, + { + "epoch": 1.3850794732516887, + "grad_norm": 1.8913402557373047, + "learning_rate": 4.5593280038443535e-06, + "loss": 3.1774, + "step": 17277 + }, + { + "epoch": 1.3851596480327113, + "grad_norm": 1.7527375221252441, + "learning_rate": 4.558238598556758e-06, + "loss": 3.1674, + "step": 17278 + }, + { + "epoch": 1.385239822813734, + "grad_norm": 1.7855474948883057, + "learning_rate": 4.557149285015339e-06, + "loss": 3.1729, + "step": 17279 + }, + { + "epoch": 1.3853199975947565, + "grad_norm": 1.805457592010498, + "learning_rate": 4.556060063238464e-06, + "loss": 3.2761, + "step": 17280 + }, + { + "epoch": 1.3854001723757792, + "grad_norm": 0.6900845170021057, + "learning_rate": 4.554970933244498e-06, + "loss": 2.4758, + "step": 17281 + }, + { + "epoch": 1.3854803471568018, + "grad_norm": 1.9030730724334717, + "learning_rate": 4.5538818950518035e-06, + "loss": 3.2206, + "step": 17282 + }, + { + "epoch": 1.3855605219378244, + "grad_norm": 1.8360172510147095, + "learning_rate": 4.552792948678744e-06, + "loss": 3.2239, + "step": 17283 + }, + { + "epoch": 1.385640696718847, + "grad_norm": 1.7029081583023071, + "learning_rate": 4.551704094143672e-06, + "loss": 3.0878, + "step": 17284 + }, + { + "epoch": 1.3857208714998697, + "grad_norm": 1.936667561531067, + "learning_rate": 4.550615331464949e-06, + "loss": 3.1768, + "step": 17285 + }, + { + "epoch": 1.3858010462808923, + "grad_norm": 1.800421118736267, + "learning_rate": 4.549526660660931e-06, + "loss": 3.1316, + "step": 17286 + }, + { + "epoch": 1.385881221061915, + "grad_norm": 1.756264567375183, + "learning_rate": 4.548438081749973e-06, + "loss": 3.2004, + "step": 17287 + }, + { + "epoch": 1.3859613958429375, + "grad_norm": 1.8233035802841187, + "learning_rate": 4.5473495947504295e-06, + "loss": 3.2059, + "step": 17288 + }, + { + "epoch": 1.3860415706239602, + "grad_norm": 1.7130731344223022, + "learning_rate": 4.546261199680648e-06, + "loss": 3.1775, + "step": 17289 + }, + { + "epoch": 1.3861217454049828, + "grad_norm": 1.7516711950302124, + "learning_rate": 4.54517289655898e-06, + "loss": 3.0816, + "step": 17290 + }, + { + "epoch": 1.3862019201860054, + "grad_norm": 0.7018876075744629, + "learning_rate": 4.544084685403776e-06, + "loss": 2.4724, + "step": 17291 + }, + { + "epoch": 1.386282094967028, + "grad_norm": 1.7356836795806885, + "learning_rate": 4.54299656623338e-06, + "loss": 3.1639, + "step": 17292 + }, + { + "epoch": 1.3863622697480507, + "grad_norm": 1.884448766708374, + "learning_rate": 4.5419085390661425e-06, + "loss": 3.21, + "step": 17293 + }, + { + "epoch": 1.3864424445290733, + "grad_norm": 1.8367427587509155, + "learning_rate": 4.5408206039204e-06, + "loss": 3.0803, + "step": 17294 + }, + { + "epoch": 1.386522619310096, + "grad_norm": 1.809329628944397, + "learning_rate": 4.539732760814497e-06, + "loss": 3.1785, + "step": 17295 + }, + { + "epoch": 1.3866027940911185, + "grad_norm": 1.8213812112808228, + "learning_rate": 4.538645009766777e-06, + "loss": 3.1527, + "step": 17296 + }, + { + "epoch": 1.3866829688721412, + "grad_norm": 1.7502732276916504, + "learning_rate": 4.537557350795576e-06, + "loss": 3.1998, + "step": 17297 + }, + { + "epoch": 1.3867631436531638, + "grad_norm": 1.8176038265228271, + "learning_rate": 4.536469783919237e-06, + "loss": 3.2202, + "step": 17298 + }, + { + "epoch": 1.3868433184341864, + "grad_norm": 0.6528580784797668, + "learning_rate": 4.535382309156089e-06, + "loss": 2.4663, + "step": 17299 + }, + { + "epoch": 1.386923493215209, + "grad_norm": 1.7816014289855957, + "learning_rate": 4.534294926524469e-06, + "loss": 3.2734, + "step": 17300 + }, + { + "epoch": 1.3870036679962316, + "grad_norm": 1.7549103498458862, + "learning_rate": 4.533207636042709e-06, + "loss": 3.0988, + "step": 17301 + }, + { + "epoch": 1.3870838427772545, + "grad_norm": 1.8008395433425903, + "learning_rate": 4.532120437729143e-06, + "loss": 3.1583, + "step": 17302 + }, + { + "epoch": 1.3871640175582771, + "grad_norm": 1.725796103477478, + "learning_rate": 4.531033331602102e-06, + "loss": 3.2202, + "step": 17303 + }, + { + "epoch": 1.3872441923392997, + "grad_norm": 1.6987125873565674, + "learning_rate": 4.529946317679909e-06, + "loss": 3.2122, + "step": 17304 + }, + { + "epoch": 1.3873243671203224, + "grad_norm": 1.754571557044983, + "learning_rate": 4.528859395980892e-06, + "loss": 3.1246, + "step": 17305 + }, + { + "epoch": 1.387404541901345, + "grad_norm": 0.6650561690330505, + "learning_rate": 4.527772566523378e-06, + "loss": 2.476, + "step": 17306 + }, + { + "epoch": 1.3874847166823676, + "grad_norm": 1.6908901929855347, + "learning_rate": 4.52668582932569e-06, + "loss": 3.2124, + "step": 17307 + }, + { + "epoch": 1.3875648914633902, + "grad_norm": 1.7864415645599365, + "learning_rate": 4.525599184406153e-06, + "loss": 3.1727, + "step": 17308 + }, + { + "epoch": 1.3876450662444129, + "grad_norm": 1.7018446922302246, + "learning_rate": 4.524512631783082e-06, + "loss": 3.2684, + "step": 17309 + }, + { + "epoch": 1.3877252410254355, + "grad_norm": 1.8813846111297607, + "learning_rate": 4.5234261714747975e-06, + "loss": 3.2141, + "step": 17310 + }, + { + "epoch": 1.387805415806458, + "grad_norm": 1.820214033126831, + "learning_rate": 4.522339803499618e-06, + "loss": 3.2926, + "step": 17311 + }, + { + "epoch": 1.3878855905874807, + "grad_norm": 1.7315738201141357, + "learning_rate": 4.521253527875858e-06, + "loss": 3.1277, + "step": 17312 + }, + { + "epoch": 1.3879657653685034, + "grad_norm": 1.7897907495498657, + "learning_rate": 4.520167344621836e-06, + "loss": 3.1859, + "step": 17313 + }, + { + "epoch": 1.388045940149526, + "grad_norm": 1.869749665260315, + "learning_rate": 4.519081253755859e-06, + "loss": 3.2316, + "step": 17314 + }, + { + "epoch": 1.3881261149305486, + "grad_norm": 1.80641770362854, + "learning_rate": 4.5179952552962405e-06, + "loss": 3.2101, + "step": 17315 + }, + { + "epoch": 1.3882062897115712, + "grad_norm": 1.752130389213562, + "learning_rate": 4.516909349261289e-06, + "loss": 3.2461, + "step": 17316 + }, + { + "epoch": 1.3882864644925939, + "grad_norm": 1.8389418125152588, + "learning_rate": 4.515823535669314e-06, + "loss": 3.1771, + "step": 17317 + }, + { + "epoch": 1.3883666392736165, + "grad_norm": 1.7637752294540405, + "learning_rate": 4.514737814538624e-06, + "loss": 3.1357, + "step": 17318 + }, + { + "epoch": 1.388446814054639, + "grad_norm": 1.6875126361846924, + "learning_rate": 4.51365218588752e-06, + "loss": 3.1697, + "step": 17319 + }, + { + "epoch": 1.3885269888356617, + "grad_norm": 1.7121385335922241, + "learning_rate": 4.512566649734303e-06, + "loss": 3.1265, + "step": 17320 + }, + { + "epoch": 1.3886071636166843, + "grad_norm": 1.8510863780975342, + "learning_rate": 4.5114812060972815e-06, + "loss": 3.23, + "step": 17321 + }, + { + "epoch": 1.388687338397707, + "grad_norm": 1.8426110744476318, + "learning_rate": 4.510395854994751e-06, + "loss": 3.1882, + "step": 17322 + }, + { + "epoch": 1.3887675131787296, + "grad_norm": 1.894199013710022, + "learning_rate": 4.509310596445016e-06, + "loss": 3.1981, + "step": 17323 + }, + { + "epoch": 1.3888476879597522, + "grad_norm": 1.7559449672698975, + "learning_rate": 4.508225430466365e-06, + "loss": 3.155, + "step": 17324 + }, + { + "epoch": 1.3889278627407748, + "grad_norm": 1.6938235759735107, + "learning_rate": 4.507140357077098e-06, + "loss": 3.1715, + "step": 17325 + }, + { + "epoch": 1.3890080375217975, + "grad_norm": 1.760545015335083, + "learning_rate": 4.5060553762955085e-06, + "loss": 3.1981, + "step": 17326 + }, + { + "epoch": 1.38908821230282, + "grad_norm": 1.7224940061569214, + "learning_rate": 4.504970488139889e-06, + "loss": 3.1901, + "step": 17327 + }, + { + "epoch": 1.3891683870838427, + "grad_norm": 1.8627756834030151, + "learning_rate": 4.503885692628531e-06, + "loss": 3.1628, + "step": 17328 + }, + { + "epoch": 1.3892485618648653, + "grad_norm": 1.67520272731781, + "learning_rate": 4.5028009897797264e-06, + "loss": 3.1884, + "step": 17329 + }, + { + "epoch": 1.389328736645888, + "grad_norm": 1.8098351955413818, + "learning_rate": 4.501716379611757e-06, + "loss": 3.0974, + "step": 17330 + }, + { + "epoch": 1.3894089114269106, + "grad_norm": 1.820755958557129, + "learning_rate": 4.50063186214291e-06, + "loss": 3.1784, + "step": 17331 + }, + { + "epoch": 1.3894890862079334, + "grad_norm": 1.6941841840744019, + "learning_rate": 4.499547437391473e-06, + "loss": 3.1887, + "step": 17332 + }, + { + "epoch": 1.389569260988956, + "grad_norm": 1.7306078672409058, + "learning_rate": 4.4984631053757266e-06, + "loss": 3.2351, + "step": 17333 + }, + { + "epoch": 1.3896494357699787, + "grad_norm": 1.823725938796997, + "learning_rate": 4.497378866113959e-06, + "loss": 3.0836, + "step": 17334 + }, + { + "epoch": 1.3897296105510013, + "grad_norm": 1.7858421802520752, + "learning_rate": 4.496294719624437e-06, + "loss": 3.1536, + "step": 17335 + }, + { + "epoch": 1.389809785332024, + "grad_norm": 1.7947704792022705, + "learning_rate": 4.49521066592545e-06, + "loss": 3.1968, + "step": 17336 + }, + { + "epoch": 1.3898899601130466, + "grad_norm": 1.6795481443405151, + "learning_rate": 4.49412670503527e-06, + "loss": 3.0944, + "step": 17337 + }, + { + "epoch": 1.3899701348940692, + "grad_norm": 1.6699845790863037, + "learning_rate": 4.493042836972173e-06, + "loss": 3.1342, + "step": 17338 + }, + { + "epoch": 1.3900503096750918, + "grad_norm": 1.8852719068527222, + "learning_rate": 4.491959061754436e-06, + "loss": 3.2281, + "step": 17339 + }, + { + "epoch": 1.3901304844561144, + "grad_norm": 1.8139700889587402, + "learning_rate": 4.490875379400326e-06, + "loss": 3.2131, + "step": 17340 + }, + { + "epoch": 1.390210659237137, + "grad_norm": 1.735774040222168, + "learning_rate": 4.489791789928114e-06, + "loss": 3.1432, + "step": 17341 + }, + { + "epoch": 1.3902908340181597, + "grad_norm": 1.8144071102142334, + "learning_rate": 4.488708293356071e-06, + "loss": 3.1797, + "step": 17342 + }, + { + "epoch": 1.3903710087991823, + "grad_norm": 1.8392975330352783, + "learning_rate": 4.487624889702463e-06, + "loss": 3.1364, + "step": 17343 + }, + { + "epoch": 1.390451183580205, + "grad_norm": 1.800695776939392, + "learning_rate": 4.486541578985562e-06, + "loss": 3.2103, + "step": 17344 + }, + { + "epoch": 1.3905313583612275, + "grad_norm": 1.752153754234314, + "learning_rate": 4.4854583612236215e-06, + "loss": 3.1768, + "step": 17345 + }, + { + "epoch": 1.3906115331422502, + "grad_norm": 1.6961114406585693, + "learning_rate": 4.48437523643491e-06, + "loss": 3.2725, + "step": 17346 + }, + { + "epoch": 1.3906917079232728, + "grad_norm": 1.971144437789917, + "learning_rate": 4.483292204637688e-06, + "loss": 3.1094, + "step": 17347 + }, + { + "epoch": 1.3907718827042954, + "grad_norm": 1.7897285223007202, + "learning_rate": 4.482209265850216e-06, + "loss": 3.2593, + "step": 17348 + }, + { + "epoch": 1.390852057485318, + "grad_norm": 1.7835510969161987, + "learning_rate": 4.481126420090754e-06, + "loss": 3.1697, + "step": 17349 + }, + { + "epoch": 1.3909322322663407, + "grad_norm": 1.802790880203247, + "learning_rate": 4.480043667377552e-06, + "loss": 3.2325, + "step": 17350 + }, + { + "epoch": 1.3910124070473633, + "grad_norm": 1.855720043182373, + "learning_rate": 4.478961007728868e-06, + "loss": 3.1991, + "step": 17351 + }, + { + "epoch": 1.391092581828386, + "grad_norm": 1.781294822692871, + "learning_rate": 4.477878441162956e-06, + "loss": 3.1763, + "step": 17352 + }, + { + "epoch": 1.3911727566094085, + "grad_norm": 1.7639695405960083, + "learning_rate": 4.476795967698067e-06, + "loss": 3.3318, + "step": 17353 + }, + { + "epoch": 1.3912529313904312, + "grad_norm": 1.6566588878631592, + "learning_rate": 4.475713587352456e-06, + "loss": 3.1406, + "step": 17354 + }, + { + "epoch": 1.3913331061714538, + "grad_norm": 1.822588562965393, + "learning_rate": 4.474631300144363e-06, + "loss": 3.1785, + "step": 17355 + }, + { + "epoch": 1.3914132809524764, + "grad_norm": 0.6926658153533936, + "learning_rate": 4.473549106092039e-06, + "loss": 2.4637, + "step": 17356 + }, + { + "epoch": 1.391493455733499, + "grad_norm": 1.799174427986145, + "learning_rate": 4.472467005213729e-06, + "loss": 3.2745, + "step": 17357 + }, + { + "epoch": 1.3915736305145217, + "grad_norm": 1.7519502639770508, + "learning_rate": 4.471384997527677e-06, + "loss": 3.1552, + "step": 17358 + }, + { + "epoch": 1.3916538052955443, + "grad_norm": 1.8623751401901245, + "learning_rate": 4.470303083052129e-06, + "loss": 3.0959, + "step": 17359 + }, + { + "epoch": 1.391733980076567, + "grad_norm": 1.7760207653045654, + "learning_rate": 4.4692212618053185e-06, + "loss": 3.2172, + "step": 17360 + }, + { + "epoch": 1.3918141548575895, + "grad_norm": 0.7243291139602661, + "learning_rate": 4.468139533805489e-06, + "loss": 2.495, + "step": 17361 + }, + { + "epoch": 1.3918943296386121, + "grad_norm": 1.8458895683288574, + "learning_rate": 4.467057899070877e-06, + "loss": 3.1694, + "step": 17362 + }, + { + "epoch": 1.3919745044196348, + "grad_norm": 1.7554166316986084, + "learning_rate": 4.465976357619719e-06, + "loss": 3.1716, + "step": 17363 + }, + { + "epoch": 1.3920546792006574, + "grad_norm": 0.6655102372169495, + "learning_rate": 4.4648949094702535e-06, + "loss": 2.4692, + "step": 17364 + }, + { + "epoch": 1.39213485398168, + "grad_norm": 0.6907227039337158, + "learning_rate": 4.4638135546407035e-06, + "loss": 2.4727, + "step": 17365 + }, + { + "epoch": 1.3922150287627026, + "grad_norm": 1.767608642578125, + "learning_rate": 4.462732293149308e-06, + "loss": 3.3174, + "step": 17366 + }, + { + "epoch": 1.3922952035437253, + "grad_norm": 1.783879280090332, + "learning_rate": 4.4616511250142925e-06, + "loss": 3.2726, + "step": 17367 + }, + { + "epoch": 1.392375378324748, + "grad_norm": 1.9351710081100464, + "learning_rate": 4.4605700502538885e-06, + "loss": 3.2107, + "step": 17368 + }, + { + "epoch": 1.3924555531057705, + "grad_norm": 0.6797356605529785, + "learning_rate": 4.459489068886324e-06, + "loss": 2.4821, + "step": 17369 + }, + { + "epoch": 1.3925357278867931, + "grad_norm": 1.7584751844406128, + "learning_rate": 4.4584081809298185e-06, + "loss": 3.1344, + "step": 17370 + }, + { + "epoch": 1.3926159026678158, + "grad_norm": 1.8077912330627441, + "learning_rate": 4.457327386402597e-06, + "loss": 3.1383, + "step": 17371 + }, + { + "epoch": 1.3926960774488384, + "grad_norm": 1.765062689781189, + "learning_rate": 4.456246685322883e-06, + "loss": 3.2379, + "step": 17372 + }, + { + "epoch": 1.392776252229861, + "grad_norm": 1.8084901571273804, + "learning_rate": 4.455166077708895e-06, + "loss": 3.1669, + "step": 17373 + }, + { + "epoch": 1.3928564270108836, + "grad_norm": 1.7396245002746582, + "learning_rate": 4.454085563578859e-06, + "loss": 3.2035, + "step": 17374 + }, + { + "epoch": 1.3929366017919063, + "grad_norm": 1.7615801095962524, + "learning_rate": 4.453005142950979e-06, + "loss": 3.2232, + "step": 17375 + }, + { + "epoch": 1.3930167765729289, + "grad_norm": 1.8096665143966675, + "learning_rate": 4.451924815843479e-06, + "loss": 3.2828, + "step": 17376 + }, + { + "epoch": 1.3930969513539515, + "grad_norm": 1.7297714948654175, + "learning_rate": 4.450844582274571e-06, + "loss": 3.1882, + "step": 17377 + }, + { + "epoch": 1.3931771261349741, + "grad_norm": 1.774558424949646, + "learning_rate": 4.449764442262467e-06, + "loss": 3.1453, + "step": 17378 + }, + { + "epoch": 1.3932573009159968, + "grad_norm": 1.8235771656036377, + "learning_rate": 4.448684395825383e-06, + "loss": 3.2502, + "step": 17379 + }, + { + "epoch": 1.3933374756970194, + "grad_norm": 1.8338431119918823, + "learning_rate": 4.44760444298152e-06, + "loss": 3.174, + "step": 17380 + }, + { + "epoch": 1.393417650478042, + "grad_norm": 1.777830958366394, + "learning_rate": 4.446524583749089e-06, + "loss": 3.1818, + "step": 17381 + }, + { + "epoch": 1.3934978252590646, + "grad_norm": 1.7497121095657349, + "learning_rate": 4.4454448181462965e-06, + "loss": 3.1949, + "step": 17382 + }, + { + "epoch": 1.3935780000400875, + "grad_norm": 1.7423062324523926, + "learning_rate": 4.444365146191346e-06, + "loss": 3.1906, + "step": 17383 + }, + { + "epoch": 1.39365817482111, + "grad_norm": 1.6952160596847534, + "learning_rate": 4.443285567902446e-06, + "loss": 3.1885, + "step": 17384 + }, + { + "epoch": 1.3937383496021327, + "grad_norm": 1.7041432857513428, + "learning_rate": 4.442206083297788e-06, + "loss": 3.0436, + "step": 17385 + }, + { + "epoch": 1.3938185243831553, + "grad_norm": 1.745648980140686, + "learning_rate": 4.441126692395577e-06, + "loss": 3.1927, + "step": 17386 + }, + { + "epoch": 1.393898699164178, + "grad_norm": 1.8512905836105347, + "learning_rate": 4.440047395214011e-06, + "loss": 3.2496, + "step": 17387 + }, + { + "epoch": 1.3939788739452006, + "grad_norm": 1.8750907182693481, + "learning_rate": 4.438968191771287e-06, + "loss": 3.1251, + "step": 17388 + }, + { + "epoch": 1.3940590487262232, + "grad_norm": 1.6662280559539795, + "learning_rate": 4.4378890820856015e-06, + "loss": 3.1277, + "step": 17389 + }, + { + "epoch": 1.3941392235072458, + "grad_norm": 1.6974481344223022, + "learning_rate": 4.436810066175143e-06, + "loss": 3.1056, + "step": 17390 + }, + { + "epoch": 1.3942193982882685, + "grad_norm": 1.7989630699157715, + "learning_rate": 4.4357311440581065e-06, + "loss": 3.2025, + "step": 17391 + }, + { + "epoch": 1.394299573069291, + "grad_norm": 1.783565640449524, + "learning_rate": 4.434652315752682e-06, + "loss": 3.2332, + "step": 17392 + }, + { + "epoch": 1.3943797478503137, + "grad_norm": 0.6727156639099121, + "learning_rate": 4.433573581277058e-06, + "loss": 2.4551, + "step": 17393 + }, + { + "epoch": 1.3944599226313363, + "grad_norm": 1.771106243133545, + "learning_rate": 4.432494940649425e-06, + "loss": 3.2703, + "step": 17394 + }, + { + "epoch": 1.394540097412359, + "grad_norm": 1.7987004518508911, + "learning_rate": 4.431416393887961e-06, + "loss": 3.2858, + "step": 17395 + }, + { + "epoch": 1.3946202721933816, + "grad_norm": 1.7521486282348633, + "learning_rate": 4.430337941010854e-06, + "loss": 3.1697, + "step": 17396 + }, + { + "epoch": 1.3947004469744042, + "grad_norm": 1.7653083801269531, + "learning_rate": 4.429259582036287e-06, + "loss": 3.2182, + "step": 17397 + }, + { + "epoch": 1.3947806217554268, + "grad_norm": 1.7247943878173828, + "learning_rate": 4.428181316982439e-06, + "loss": 3.1968, + "step": 17398 + }, + { + "epoch": 1.3948607965364495, + "grad_norm": 1.6842455863952637, + "learning_rate": 4.427103145867497e-06, + "loss": 3.1985, + "step": 17399 + }, + { + "epoch": 1.394940971317472, + "grad_norm": 1.6723475456237793, + "learning_rate": 4.426025068709625e-06, + "loss": 3.1381, + "step": 17400 + }, + { + "epoch": 1.3950211460984947, + "grad_norm": 1.782547950744629, + "learning_rate": 4.424947085527007e-06, + "loss": 3.3363, + "step": 17401 + }, + { + "epoch": 1.3951013208795173, + "grad_norm": 1.8099405765533447, + "learning_rate": 4.423869196337817e-06, + "loss": 3.2144, + "step": 17402 + }, + { + "epoch": 1.39518149566054, + "grad_norm": 1.7419012784957886, + "learning_rate": 4.4227914011602255e-06, + "loss": 3.1433, + "step": 17403 + }, + { + "epoch": 1.3952616704415626, + "grad_norm": 1.6820425987243652, + "learning_rate": 4.4217137000124096e-06, + "loss": 3.1763, + "step": 17404 + }, + { + "epoch": 1.3953418452225852, + "grad_norm": 1.7403206825256348, + "learning_rate": 4.420636092912533e-06, + "loss": 3.1647, + "step": 17405 + }, + { + "epoch": 1.3954220200036078, + "grad_norm": 1.913029432296753, + "learning_rate": 4.4195585798787634e-06, + "loss": 3.037, + "step": 17406 + }, + { + "epoch": 1.3955021947846304, + "grad_norm": 1.7660009860992432, + "learning_rate": 4.41848116092927e-06, + "loss": 3.0983, + "step": 17407 + }, + { + "epoch": 1.395582369565653, + "grad_norm": 1.9403679370880127, + "learning_rate": 4.417403836082217e-06, + "loss": 3.2615, + "step": 17408 + }, + { + "epoch": 1.3956625443466757, + "grad_norm": 1.818711519241333, + "learning_rate": 4.416326605355771e-06, + "loss": 3.0634, + "step": 17409 + }, + { + "epoch": 1.3957427191276983, + "grad_norm": 1.7222800254821777, + "learning_rate": 4.415249468768088e-06, + "loss": 3.1752, + "step": 17410 + }, + { + "epoch": 1.395822893908721, + "grad_norm": 1.7253081798553467, + "learning_rate": 4.41417242633733e-06, + "loss": 3.1473, + "step": 17411 + }, + { + "epoch": 1.3959030686897436, + "grad_norm": 1.7090407609939575, + "learning_rate": 4.413095478081656e-06, + "loss": 3.1373, + "step": 17412 + }, + { + "epoch": 1.3959832434707662, + "grad_norm": 1.7916276454925537, + "learning_rate": 4.412018624019224e-06, + "loss": 3.2366, + "step": 17413 + }, + { + "epoch": 1.396063418251789, + "grad_norm": 1.759209394454956, + "learning_rate": 4.410941864168193e-06, + "loss": 3.123, + "step": 17414 + }, + { + "epoch": 1.3961435930328117, + "grad_norm": 1.7092726230621338, + "learning_rate": 4.4098651985467065e-06, + "loss": 3.1698, + "step": 17415 + }, + { + "epoch": 1.3962237678138343, + "grad_norm": 1.7226452827453613, + "learning_rate": 4.408788627172924e-06, + "loss": 3.0787, + "step": 17416 + }, + { + "epoch": 1.396303942594857, + "grad_norm": 1.7341378927230835, + "learning_rate": 4.407712150064994e-06, + "loss": 3.2692, + "step": 17417 + }, + { + "epoch": 1.3963841173758795, + "grad_norm": 1.705728530883789, + "learning_rate": 4.4066357672410666e-06, + "loss": 3.11, + "step": 17418 + }, + { + "epoch": 1.3964642921569022, + "grad_norm": 1.7329413890838623, + "learning_rate": 4.405559478719292e-06, + "loss": 3.1593, + "step": 17419 + }, + { + "epoch": 1.3965444669379248, + "grad_norm": 1.8300154209136963, + "learning_rate": 4.4044832845178106e-06, + "loss": 3.2692, + "step": 17420 + }, + { + "epoch": 1.3966246417189474, + "grad_norm": 1.8414297103881836, + "learning_rate": 4.403407184654767e-06, + "loss": 3.1844, + "step": 17421 + }, + { + "epoch": 1.39670481649997, + "grad_norm": 1.7628331184387207, + "learning_rate": 4.402331179148308e-06, + "loss": 3.1297, + "step": 17422 + }, + { + "epoch": 1.3967849912809926, + "grad_norm": 1.6740782260894775, + "learning_rate": 4.401255268016572e-06, + "loss": 3.2502, + "step": 17423 + }, + { + "epoch": 1.3968651660620153, + "grad_norm": 2.022510290145874, + "learning_rate": 4.400179451277702e-06, + "loss": 3.2347, + "step": 17424 + }, + { + "epoch": 1.396945340843038, + "grad_norm": 0.6770249605178833, + "learning_rate": 4.39910372894983e-06, + "loss": 2.4755, + "step": 17425 + }, + { + "epoch": 1.3970255156240605, + "grad_norm": 1.7647309303283691, + "learning_rate": 4.398028101051094e-06, + "loss": 3.2005, + "step": 17426 + }, + { + "epoch": 1.3971056904050831, + "grad_norm": 1.7803521156311035, + "learning_rate": 4.396952567599631e-06, + "loss": 3.3201, + "step": 17427 + }, + { + "epoch": 1.3971858651861058, + "grad_norm": 1.7702999114990234, + "learning_rate": 4.395877128613573e-06, + "loss": 3.2314, + "step": 17428 + }, + { + "epoch": 1.3972660399671284, + "grad_norm": 1.791920781135559, + "learning_rate": 4.394801784111053e-06, + "loss": 3.0995, + "step": 17429 + }, + { + "epoch": 1.397346214748151, + "grad_norm": 1.7707016468048096, + "learning_rate": 4.393726534110199e-06, + "loss": 3.1633, + "step": 17430 + }, + { + "epoch": 1.3974263895291736, + "grad_norm": 1.7696653604507446, + "learning_rate": 4.392651378629138e-06, + "loss": 3.1708, + "step": 17431 + }, + { + "epoch": 1.3975065643101963, + "grad_norm": 1.6903283596038818, + "learning_rate": 4.391576317685998e-06, + "loss": 3.1246, + "step": 17432 + }, + { + "epoch": 1.3975867390912189, + "grad_norm": 1.744028091430664, + "learning_rate": 4.390501351298906e-06, + "loss": 3.0494, + "step": 17433 + }, + { + "epoch": 1.3976669138722415, + "grad_norm": 1.955450177192688, + "learning_rate": 4.3894264794859876e-06, + "loss": 3.1875, + "step": 17434 + }, + { + "epoch": 1.3977470886532641, + "grad_norm": 1.9241501092910767, + "learning_rate": 4.388351702265357e-06, + "loss": 3.0815, + "step": 17435 + }, + { + "epoch": 1.3978272634342868, + "grad_norm": 1.728027105331421, + "learning_rate": 4.38727701965514e-06, + "loss": 3.1742, + "step": 17436 + }, + { + "epoch": 1.3979074382153094, + "grad_norm": 1.7231606245040894, + "learning_rate": 4.386202431673454e-06, + "loss": 3.0776, + "step": 17437 + }, + { + "epoch": 1.397987612996332, + "grad_norm": 1.781603217124939, + "learning_rate": 4.385127938338416e-06, + "loss": 3.2456, + "step": 17438 + }, + { + "epoch": 1.3980677877773546, + "grad_norm": 1.799769401550293, + "learning_rate": 4.384053539668146e-06, + "loss": 3.2327, + "step": 17439 + }, + { + "epoch": 1.3981479625583773, + "grad_norm": 1.7855745553970337, + "learning_rate": 4.382979235680751e-06, + "loss": 3.1897, + "step": 17440 + }, + { + "epoch": 1.3982281373393999, + "grad_norm": 1.745679259300232, + "learning_rate": 4.381905026394346e-06, + "loss": 3.2136, + "step": 17441 + }, + { + "epoch": 1.3983083121204225, + "grad_norm": 1.7261749505996704, + "learning_rate": 4.380830911827042e-06, + "loss": 3.1814, + "step": 17442 + }, + { + "epoch": 1.3983884869014451, + "grad_norm": 2.0149405002593994, + "learning_rate": 4.379756891996949e-06, + "loss": 3.1508, + "step": 17443 + }, + { + "epoch": 1.3984686616824678, + "grad_norm": 1.7466336488723755, + "learning_rate": 4.378682966922178e-06, + "loss": 3.1442, + "step": 17444 + }, + { + "epoch": 1.3985488364634904, + "grad_norm": 1.822563886642456, + "learning_rate": 4.377609136620826e-06, + "loss": 3.1981, + "step": 17445 + }, + { + "epoch": 1.398629011244513, + "grad_norm": 1.7332344055175781, + "learning_rate": 4.376535401111003e-06, + "loss": 3.1835, + "step": 17446 + }, + { + "epoch": 1.3987091860255356, + "grad_norm": 1.8468588590621948, + "learning_rate": 4.375461760410812e-06, + "loss": 3.1118, + "step": 17447 + }, + { + "epoch": 1.3987893608065582, + "grad_norm": 1.7471306324005127, + "learning_rate": 4.374388214538352e-06, + "loss": 3.0904, + "step": 17448 + }, + { + "epoch": 1.3988695355875809, + "grad_norm": 1.8103387355804443, + "learning_rate": 4.373314763511729e-06, + "loss": 3.1174, + "step": 17449 + }, + { + "epoch": 1.3989497103686035, + "grad_norm": 1.8053539991378784, + "learning_rate": 4.372241407349031e-06, + "loss": 3.1813, + "step": 17450 + }, + { + "epoch": 1.3990298851496261, + "grad_norm": 1.7663747072219849, + "learning_rate": 4.371168146068361e-06, + "loss": 3.277, + "step": 17451 + }, + { + "epoch": 1.3991100599306487, + "grad_norm": 1.8380231857299805, + "learning_rate": 4.370094979687812e-06, + "loss": 3.1105, + "step": 17452 + }, + { + "epoch": 1.3991902347116714, + "grad_norm": 1.7671444416046143, + "learning_rate": 4.369021908225477e-06, + "loss": 3.2523, + "step": 17453 + }, + { + "epoch": 1.399270409492694, + "grad_norm": 1.9712034463882446, + "learning_rate": 4.367948931699452e-06, + "loss": 3.2083, + "step": 17454 + }, + { + "epoch": 1.3993505842737166, + "grad_norm": 1.7067097425460815, + "learning_rate": 4.36687605012782e-06, + "loss": 3.1984, + "step": 17455 + }, + { + "epoch": 1.3994307590547392, + "grad_norm": 1.7687742710113525, + "learning_rate": 4.365803263528671e-06, + "loss": 3.1313, + "step": 17456 + }, + { + "epoch": 1.3995109338357619, + "grad_norm": 1.7741011381149292, + "learning_rate": 4.364730571920094e-06, + "loss": 3.1166, + "step": 17457 + }, + { + "epoch": 1.3995911086167845, + "grad_norm": 1.7845097780227661, + "learning_rate": 4.363657975320173e-06, + "loss": 3.3271, + "step": 17458 + }, + { + "epoch": 1.399671283397807, + "grad_norm": 1.7838101387023926, + "learning_rate": 4.362585473746997e-06, + "loss": 3.1124, + "step": 17459 + }, + { + "epoch": 1.3997514581788297, + "grad_norm": 1.7399284839630127, + "learning_rate": 4.361513067218639e-06, + "loss": 3.1603, + "step": 17460 + }, + { + "epoch": 1.3998316329598524, + "grad_norm": 1.8321523666381836, + "learning_rate": 4.360440755753184e-06, + "loss": 3.2667, + "step": 17461 + }, + { + "epoch": 1.399911807740875, + "grad_norm": 1.7394548654556274, + "learning_rate": 4.359368539368709e-06, + "loss": 3.1725, + "step": 17462 + }, + { + "epoch": 1.3999919825218976, + "grad_norm": 1.7545639276504517, + "learning_rate": 4.358296418083292e-06, + "loss": 3.2032, + "step": 17463 + }, + { + "epoch": 1.4000721573029202, + "grad_norm": 1.757502555847168, + "learning_rate": 4.357224391915015e-06, + "loss": 3.1298, + "step": 17464 + }, + { + "epoch": 1.400152332083943, + "grad_norm": 1.7592428922653198, + "learning_rate": 4.356152460881941e-06, + "loss": 3.3282, + "step": 17465 + }, + { + "epoch": 1.4002325068649657, + "grad_norm": 1.9057211875915527, + "learning_rate": 4.355080625002147e-06, + "loss": 3.2896, + "step": 17466 + }, + { + "epoch": 1.4003126816459883, + "grad_norm": 1.8382929563522339, + "learning_rate": 4.354008884293704e-06, + "loss": 3.1335, + "step": 17467 + }, + { + "epoch": 1.400392856427011, + "grad_norm": 1.7710717916488647, + "learning_rate": 4.352937238774683e-06, + "loss": 3.3424, + "step": 17468 + }, + { + "epoch": 1.4004730312080336, + "grad_norm": 1.8255504369735718, + "learning_rate": 4.351865688463152e-06, + "loss": 3.1958, + "step": 17469 + }, + { + "epoch": 1.4005532059890562, + "grad_norm": 1.691360592842102, + "learning_rate": 4.350794233377172e-06, + "loss": 3.1439, + "step": 17470 + }, + { + "epoch": 1.4006333807700788, + "grad_norm": 1.7927906513214111, + "learning_rate": 4.349722873534811e-06, + "loss": 3.1865, + "step": 17471 + }, + { + "epoch": 1.4007135555511014, + "grad_norm": 1.7359907627105713, + "learning_rate": 4.3486516089541294e-06, + "loss": 3.2358, + "step": 17472 + }, + { + "epoch": 1.400793730332124, + "grad_norm": 1.7754117250442505, + "learning_rate": 4.3475804396531915e-06, + "loss": 3.1471, + "step": 17473 + }, + { + "epoch": 1.4008739051131467, + "grad_norm": 1.8700060844421387, + "learning_rate": 4.346509365650058e-06, + "loss": 3.2269, + "step": 17474 + }, + { + "epoch": 1.4009540798941693, + "grad_norm": 1.9014381170272827, + "learning_rate": 4.34543838696278e-06, + "loss": 3.1484, + "step": 17475 + }, + { + "epoch": 1.401034254675192, + "grad_norm": 1.797420859336853, + "learning_rate": 4.344367503609419e-06, + "loss": 3.0998, + "step": 17476 + }, + { + "epoch": 1.4011144294562146, + "grad_norm": 1.8542503118515015, + "learning_rate": 4.343296715608029e-06, + "loss": 3.2205, + "step": 17477 + }, + { + "epoch": 1.4011946042372372, + "grad_norm": 1.8513436317443848, + "learning_rate": 4.3422260229766615e-06, + "loss": 3.2191, + "step": 17478 + }, + { + "epoch": 1.4012747790182598, + "grad_norm": 1.7956353425979614, + "learning_rate": 4.3411554257333734e-06, + "loss": 3.191, + "step": 17479 + }, + { + "epoch": 1.4013549537992824, + "grad_norm": 1.7861753702163696, + "learning_rate": 4.340084923896207e-06, + "loss": 3.2894, + "step": 17480 + }, + { + "epoch": 1.401435128580305, + "grad_norm": 1.7718621492385864, + "learning_rate": 4.339014517483213e-06, + "loss": 3.1866, + "step": 17481 + }, + { + "epoch": 1.4015153033613277, + "grad_norm": 1.791843056678772, + "learning_rate": 4.3379442065124414e-06, + "loss": 3.1279, + "step": 17482 + }, + { + "epoch": 1.4015954781423503, + "grad_norm": 1.831520438194275, + "learning_rate": 4.336873991001933e-06, + "loss": 3.1821, + "step": 17483 + }, + { + "epoch": 1.401675652923373, + "grad_norm": 1.7543270587921143, + "learning_rate": 4.335803870969737e-06, + "loss": 3.2106, + "step": 17484 + }, + { + "epoch": 1.4017558277043956, + "grad_norm": 1.8217570781707764, + "learning_rate": 4.334733846433889e-06, + "loss": 3.1549, + "step": 17485 + }, + { + "epoch": 1.4018360024854182, + "grad_norm": 1.7973796129226685, + "learning_rate": 4.3336639174124305e-06, + "loss": 3.1611, + "step": 17486 + }, + { + "epoch": 1.4019161772664408, + "grad_norm": 1.80465567111969, + "learning_rate": 4.332594083923402e-06, + "loss": 3.1643, + "step": 17487 + }, + { + "epoch": 1.4019963520474634, + "grad_norm": 1.8520642518997192, + "learning_rate": 4.331524345984839e-06, + "loss": 3.1254, + "step": 17488 + }, + { + "epoch": 1.402076526828486, + "grad_norm": 1.7378196716308594, + "learning_rate": 4.3304547036147824e-06, + "loss": 3.1238, + "step": 17489 + }, + { + "epoch": 1.4021567016095087, + "grad_norm": 1.862382173538208, + "learning_rate": 4.329385156831257e-06, + "loss": 3.2402, + "step": 17490 + }, + { + "epoch": 1.4022368763905313, + "grad_norm": 1.770378828048706, + "learning_rate": 4.3283157056522995e-06, + "loss": 3.2361, + "step": 17491 + }, + { + "epoch": 1.402317051171554, + "grad_norm": 1.7862865924835205, + "learning_rate": 4.3272463500959405e-06, + "loss": 3.135, + "step": 17492 + }, + { + "epoch": 1.4023972259525765, + "grad_norm": 1.75892972946167, + "learning_rate": 4.3261770901802095e-06, + "loss": 3.1481, + "step": 17493 + }, + { + "epoch": 1.4024774007335992, + "grad_norm": 1.744174838066101, + "learning_rate": 4.325107925923136e-06, + "loss": 3.1404, + "step": 17494 + }, + { + "epoch": 1.402557575514622, + "grad_norm": 1.6876240968704224, + "learning_rate": 4.3240388573427394e-06, + "loss": 3.0669, + "step": 17495 + }, + { + "epoch": 1.4026377502956446, + "grad_norm": 1.8770768642425537, + "learning_rate": 4.322969884457047e-06, + "loss": 3.2647, + "step": 17496 + }, + { + "epoch": 1.4027179250766673, + "grad_norm": 1.6863397359848022, + "learning_rate": 4.321901007284081e-06, + "loss": 3.136, + "step": 17497 + }, + { + "epoch": 1.4027980998576899, + "grad_norm": 1.7732347249984741, + "learning_rate": 4.320832225841865e-06, + "loss": 3.1299, + "step": 17498 + }, + { + "epoch": 1.4028782746387125, + "grad_norm": 0.6795474886894226, + "learning_rate": 4.319763540148417e-06, + "loss": 2.4541, + "step": 17499 + }, + { + "epoch": 1.4029584494197351, + "grad_norm": 1.9070221185684204, + "learning_rate": 4.318694950221751e-06, + "loss": 3.1857, + "step": 17500 + }, + { + "epoch": 1.4030386242007578, + "grad_norm": 1.8213101625442505, + "learning_rate": 4.317626456079887e-06, + "loss": 3.0963, + "step": 17501 + }, + { + "epoch": 1.4031187989817804, + "grad_norm": 0.6641445755958557, + "learning_rate": 4.316558057740837e-06, + "loss": 2.4647, + "step": 17502 + }, + { + "epoch": 1.403198973762803, + "grad_norm": 2.0436923503875732, + "learning_rate": 4.315489755222616e-06, + "loss": 3.1877, + "step": 17503 + }, + { + "epoch": 1.4032791485438256, + "grad_norm": 1.7136598825454712, + "learning_rate": 4.314421548543238e-06, + "loss": 3.166, + "step": 17504 + }, + { + "epoch": 1.4033593233248483, + "grad_norm": 1.9720078706741333, + "learning_rate": 4.313353437720705e-06, + "loss": 3.208, + "step": 17505 + }, + { + "epoch": 1.4034394981058709, + "grad_norm": 1.784804344177246, + "learning_rate": 4.312285422773029e-06, + "loss": 3.1639, + "step": 17506 + }, + { + "epoch": 1.4035196728868935, + "grad_norm": 0.6656602621078491, + "learning_rate": 4.311217503718216e-06, + "loss": 2.483, + "step": 17507 + }, + { + "epoch": 1.4035998476679161, + "grad_norm": 0.6723634600639343, + "learning_rate": 4.310149680574271e-06, + "loss": 2.4479, + "step": 17508 + }, + { + "epoch": 1.4036800224489387, + "grad_norm": 1.769052267074585, + "learning_rate": 4.309081953359201e-06, + "loss": 3.2725, + "step": 17509 + }, + { + "epoch": 1.4037601972299614, + "grad_norm": 1.70967698097229, + "learning_rate": 4.308014322090999e-06, + "loss": 3.2388, + "step": 17510 + }, + { + "epoch": 1.403840372010984, + "grad_norm": 1.8568110466003418, + "learning_rate": 4.30694678678767e-06, + "loss": 3.203, + "step": 17511 + }, + { + "epoch": 1.4039205467920066, + "grad_norm": 1.8550788164138794, + "learning_rate": 4.305879347467213e-06, + "loss": 3.0774, + "step": 17512 + }, + { + "epoch": 1.4040007215730292, + "grad_norm": 1.7277545928955078, + "learning_rate": 4.3048120041476215e-06, + "loss": 3.192, + "step": 17513 + }, + { + "epoch": 1.4040808963540519, + "grad_norm": 1.7639198303222656, + "learning_rate": 4.303744756846897e-06, + "loss": 3.0928, + "step": 17514 + }, + { + "epoch": 1.4041610711350745, + "grad_norm": 1.907335877418518, + "learning_rate": 4.302677605583024e-06, + "loss": 3.1377, + "step": 17515 + }, + { + "epoch": 1.4042412459160971, + "grad_norm": 1.8062787055969238, + "learning_rate": 4.301610550374001e-06, + "loss": 3.1761, + "step": 17516 + }, + { + "epoch": 1.4043214206971197, + "grad_norm": 1.7426834106445312, + "learning_rate": 4.300543591237812e-06, + "loss": 3.0978, + "step": 17517 + }, + { + "epoch": 1.4044015954781424, + "grad_norm": 1.8911291360855103, + "learning_rate": 4.299476728192452e-06, + "loss": 3.1488, + "step": 17518 + }, + { + "epoch": 1.404481770259165, + "grad_norm": 1.7016016244888306, + "learning_rate": 4.298409961255908e-06, + "loss": 3.1875, + "step": 17519 + }, + { + "epoch": 1.4045619450401876, + "grad_norm": 1.7315903902053833, + "learning_rate": 4.29734329044616e-06, + "loss": 3.2152, + "step": 17520 + }, + { + "epoch": 1.4046421198212102, + "grad_norm": 1.862098217010498, + "learning_rate": 4.296276715781193e-06, + "loss": 3.1854, + "step": 17521 + }, + { + "epoch": 1.4047222946022329, + "grad_norm": 1.7942965030670166, + "learning_rate": 4.295210237278992e-06, + "loss": 3.077, + "step": 17522 + }, + { + "epoch": 1.4048024693832555, + "grad_norm": 1.7860641479492188, + "learning_rate": 4.2941438549575345e-06, + "loss": 3.0833, + "step": 17523 + }, + { + "epoch": 1.404882644164278, + "grad_norm": 1.8582994937896729, + "learning_rate": 4.293077568834805e-06, + "loss": 3.0532, + "step": 17524 + }, + { + "epoch": 1.4049628189453007, + "grad_norm": 1.8921279907226562, + "learning_rate": 4.292011378928772e-06, + "loss": 3.3229, + "step": 17525 + }, + { + "epoch": 1.4050429937263234, + "grad_norm": 1.7737021446228027, + "learning_rate": 4.2909452852574156e-06, + "loss": 3.1365, + "step": 17526 + }, + { + "epoch": 1.405123168507346, + "grad_norm": 1.7377115488052368, + "learning_rate": 4.289879287838709e-06, + "loss": 3.1765, + "step": 17527 + }, + { + "epoch": 1.4052033432883686, + "grad_norm": 1.7855068445205688, + "learning_rate": 4.288813386690627e-06, + "loss": 3.1641, + "step": 17528 + }, + { + "epoch": 1.4052835180693912, + "grad_norm": 0.7114686369895935, + "learning_rate": 4.287747581831141e-06, + "loss": 2.4851, + "step": 17529 + }, + { + "epoch": 1.4053636928504138, + "grad_norm": 1.7784062623977661, + "learning_rate": 4.2866818732782145e-06, + "loss": 3.1613, + "step": 17530 + }, + { + "epoch": 1.4054438676314365, + "grad_norm": 0.6947540044784546, + "learning_rate": 4.285616261049818e-06, + "loss": 2.4713, + "step": 17531 + }, + { + "epoch": 1.405524042412459, + "grad_norm": 1.663967490196228, + "learning_rate": 4.284550745163917e-06, + "loss": 3.079, + "step": 17532 + }, + { + "epoch": 1.4056042171934817, + "grad_norm": 1.7534162998199463, + "learning_rate": 4.283485325638475e-06, + "loss": 3.1631, + "step": 17533 + }, + { + "epoch": 1.4056843919745043, + "grad_norm": 1.729637622833252, + "learning_rate": 4.282420002491461e-06, + "loss": 3.1678, + "step": 17534 + }, + { + "epoch": 1.405764566755527, + "grad_norm": 0.6733207702636719, + "learning_rate": 4.281354775740827e-06, + "loss": 2.4828, + "step": 17535 + }, + { + "epoch": 1.4058447415365496, + "grad_norm": 1.8128516674041748, + "learning_rate": 4.280289645404535e-06, + "loss": 3.1952, + "step": 17536 + }, + { + "epoch": 1.4059249163175722, + "grad_norm": 1.6633987426757812, + "learning_rate": 4.279224611500544e-06, + "loss": 3.1394, + "step": 17537 + }, + { + "epoch": 1.4060050910985948, + "grad_norm": 1.8369321823120117, + "learning_rate": 4.278159674046811e-06, + "loss": 3.158, + "step": 17538 + }, + { + "epoch": 1.4060852658796175, + "grad_norm": 1.8085689544677734, + "learning_rate": 4.277094833061292e-06, + "loss": 3.1901, + "step": 17539 + }, + { + "epoch": 1.40616544066064, + "grad_norm": 0.675733745098114, + "learning_rate": 4.276030088561934e-06, + "loss": 2.4812, + "step": 17540 + }, + { + "epoch": 1.4062456154416627, + "grad_norm": 1.8011237382888794, + "learning_rate": 4.274965440566691e-06, + "loss": 3.1073, + "step": 17541 + }, + { + "epoch": 1.4063257902226853, + "grad_norm": 1.8085435628890991, + "learning_rate": 4.273900889093513e-06, + "loss": 3.144, + "step": 17542 + }, + { + "epoch": 1.406405965003708, + "grad_norm": 1.6451234817504883, + "learning_rate": 4.272836434160348e-06, + "loss": 3.2109, + "step": 17543 + }, + { + "epoch": 1.4064861397847306, + "grad_norm": 1.7270846366882324, + "learning_rate": 4.271772075785145e-06, + "loss": 3.1093, + "step": 17544 + }, + { + "epoch": 1.4065663145657532, + "grad_norm": 1.7864986658096313, + "learning_rate": 4.270707813985844e-06, + "loss": 3.1705, + "step": 17545 + }, + { + "epoch": 1.406646489346776, + "grad_norm": 1.7769707441329956, + "learning_rate": 4.26964364878039e-06, + "loss": 3.2149, + "step": 17546 + }, + { + "epoch": 1.4067266641277987, + "grad_norm": 1.8730634450912476, + "learning_rate": 4.2685795801867235e-06, + "loss": 3.1947, + "step": 17547 + }, + { + "epoch": 1.4068068389088213, + "grad_norm": 1.7794599533081055, + "learning_rate": 4.267515608222785e-06, + "loss": 3.2435, + "step": 17548 + }, + { + "epoch": 1.406887013689844, + "grad_norm": 1.7886135578155518, + "learning_rate": 4.266451732906518e-06, + "loss": 3.2419, + "step": 17549 + }, + { + "epoch": 1.4069671884708665, + "grad_norm": 1.6567078828811646, + "learning_rate": 4.26538795425585e-06, + "loss": 3.1375, + "step": 17550 + }, + { + "epoch": 1.4070473632518892, + "grad_norm": 1.9282695055007935, + "learning_rate": 4.264324272288719e-06, + "loss": 3.2471, + "step": 17551 + }, + { + "epoch": 1.4071275380329118, + "grad_norm": 1.6998955011367798, + "learning_rate": 4.263260687023059e-06, + "loss": 3.1192, + "step": 17552 + }, + { + "epoch": 1.4072077128139344, + "grad_norm": 1.7899783849716187, + "learning_rate": 4.262197198476804e-06, + "loss": 3.2444, + "step": 17553 + }, + { + "epoch": 1.407287887594957, + "grad_norm": 1.7850861549377441, + "learning_rate": 4.261133806667885e-06, + "loss": 3.2318, + "step": 17554 + }, + { + "epoch": 1.4073680623759797, + "grad_norm": 0.6649971604347229, + "learning_rate": 4.260070511614224e-06, + "loss": 2.4728, + "step": 17555 + }, + { + "epoch": 1.4074482371570023, + "grad_norm": 1.753136157989502, + "learning_rate": 4.25900731333375e-06, + "loss": 3.0677, + "step": 17556 + }, + { + "epoch": 1.407528411938025, + "grad_norm": 1.683396577835083, + "learning_rate": 4.25794421184439e-06, + "loss": 3.1003, + "step": 17557 + }, + { + "epoch": 1.4076085867190475, + "grad_norm": 1.8372294902801514, + "learning_rate": 4.256881207164068e-06, + "loss": 3.1445, + "step": 17558 + }, + { + "epoch": 1.4076887615000702, + "grad_norm": 1.852859377861023, + "learning_rate": 4.2558182993107065e-06, + "loss": 3.257, + "step": 17559 + }, + { + "epoch": 1.4077689362810928, + "grad_norm": 1.7184062004089355, + "learning_rate": 4.254755488302222e-06, + "loss": 3.1226, + "step": 17560 + }, + { + "epoch": 1.4078491110621154, + "grad_norm": 1.859539270401001, + "learning_rate": 4.253692774156535e-06, + "loss": 3.2865, + "step": 17561 + }, + { + "epoch": 1.407929285843138, + "grad_norm": 1.8497509956359863, + "learning_rate": 4.252630156891562e-06, + "loss": 3.2029, + "step": 17562 + }, + { + "epoch": 1.4080094606241607, + "grad_norm": 1.7383631467819214, + "learning_rate": 4.251567636525219e-06, + "loss": 3.0811, + "step": 17563 + }, + { + "epoch": 1.4080896354051833, + "grad_norm": 1.7513870000839233, + "learning_rate": 4.250505213075424e-06, + "loss": 3.1465, + "step": 17564 + }, + { + "epoch": 1.408169810186206, + "grad_norm": 1.7619398832321167, + "learning_rate": 4.2494428865600805e-06, + "loss": 3.134, + "step": 17565 + }, + { + "epoch": 1.4082499849672285, + "grad_norm": 0.6695141792297363, + "learning_rate": 4.2483806569971034e-06, + "loss": 2.4738, + "step": 17566 + }, + { + "epoch": 1.4083301597482512, + "grad_norm": 1.8235853910446167, + "learning_rate": 4.2473185244044005e-06, + "loss": 3.1652, + "step": 17567 + }, + { + "epoch": 1.4084103345292738, + "grad_norm": 1.8175413608551025, + "learning_rate": 4.24625648879988e-06, + "loss": 3.1809, + "step": 17568 + }, + { + "epoch": 1.4084905093102964, + "grad_norm": 1.7648593187332153, + "learning_rate": 4.245194550201453e-06, + "loss": 3.1656, + "step": 17569 + }, + { + "epoch": 1.408570684091319, + "grad_norm": 1.7498304843902588, + "learning_rate": 4.244132708627012e-06, + "loss": 3.1848, + "step": 17570 + }, + { + "epoch": 1.4086508588723416, + "grad_norm": 1.7843953371047974, + "learning_rate": 4.2430709640944656e-06, + "loss": 3.036, + "step": 17571 + }, + { + "epoch": 1.4087310336533643, + "grad_norm": 1.7873101234436035, + "learning_rate": 4.242009316621713e-06, + "loss": 3.1689, + "step": 17572 + }, + { + "epoch": 1.408811208434387, + "grad_norm": 1.757835865020752, + "learning_rate": 4.240947766226655e-06, + "loss": 3.1743, + "step": 17573 + }, + { + "epoch": 1.4088913832154095, + "grad_norm": 1.6927521228790283, + "learning_rate": 4.239886312927192e-06, + "loss": 3.1316, + "step": 17574 + }, + { + "epoch": 1.4089715579964321, + "grad_norm": 1.7616478204727173, + "learning_rate": 4.238824956741211e-06, + "loss": 3.208, + "step": 17575 + }, + { + "epoch": 1.4090517327774548, + "grad_norm": 1.7332850694656372, + "learning_rate": 4.237763697686612e-06, + "loss": 3.1702, + "step": 17576 + }, + { + "epoch": 1.4091319075584776, + "grad_norm": 2.0242557525634766, + "learning_rate": 4.236702535781285e-06, + "loss": 3.1933, + "step": 17577 + }, + { + "epoch": 1.4092120823395002, + "grad_norm": 1.771846055984497, + "learning_rate": 4.235641471043122e-06, + "loss": 3.1125, + "step": 17578 + }, + { + "epoch": 1.4092922571205229, + "grad_norm": 1.7697560787200928, + "learning_rate": 4.234580503490018e-06, + "loss": 3.17, + "step": 17579 + }, + { + "epoch": 1.4093724319015455, + "grad_norm": 1.6794792413711548, + "learning_rate": 4.233519633139849e-06, + "loss": 3.1554, + "step": 17580 + }, + { + "epoch": 1.409452606682568, + "grad_norm": 1.824928641319275, + "learning_rate": 4.232458860010507e-06, + "loss": 3.152, + "step": 17581 + }, + { + "epoch": 1.4095327814635907, + "grad_norm": 1.8314963579177856, + "learning_rate": 4.2313981841198755e-06, + "loss": 3.157, + "step": 17582 + }, + { + "epoch": 1.4096129562446134, + "grad_norm": 1.6988215446472168, + "learning_rate": 4.230337605485839e-06, + "loss": 3.1325, + "step": 17583 + }, + { + "epoch": 1.409693131025636, + "grad_norm": 1.6785472631454468, + "learning_rate": 4.22927712412628e-06, + "loss": 3.1914, + "step": 17584 + }, + { + "epoch": 1.4097733058066586, + "grad_norm": 1.922669768333435, + "learning_rate": 4.228216740059071e-06, + "loss": 3.1624, + "step": 17585 + }, + { + "epoch": 1.4098534805876812, + "grad_norm": 1.8805769681930542, + "learning_rate": 4.2271564533020946e-06, + "loss": 3.2194, + "step": 17586 + }, + { + "epoch": 1.4099336553687039, + "grad_norm": 1.8293273448944092, + "learning_rate": 4.226096263873225e-06, + "loss": 3.2755, + "step": 17587 + }, + { + "epoch": 1.4100138301497265, + "grad_norm": 0.6700746417045593, + "learning_rate": 4.225036171790338e-06, + "loss": 2.4714, + "step": 17588 + }, + { + "epoch": 1.410094004930749, + "grad_norm": 1.8152662515640259, + "learning_rate": 4.223976177071308e-06, + "loss": 3.2214, + "step": 17589 + }, + { + "epoch": 1.4101741797117717, + "grad_norm": 1.9063501358032227, + "learning_rate": 4.222916279734002e-06, + "loss": 3.123, + "step": 17590 + }, + { + "epoch": 1.4102543544927943, + "grad_norm": 1.78202486038208, + "learning_rate": 4.22185647979629e-06, + "loss": 3.0956, + "step": 17591 + }, + { + "epoch": 1.410334529273817, + "grad_norm": 1.7984024286270142, + "learning_rate": 4.220796777276043e-06, + "loss": 3.1499, + "step": 17592 + }, + { + "epoch": 1.4104147040548396, + "grad_norm": 0.6472750902175903, + "learning_rate": 4.219737172191124e-06, + "loss": 2.4728, + "step": 17593 + }, + { + "epoch": 1.4104948788358622, + "grad_norm": 1.7696863412857056, + "learning_rate": 4.218677664559404e-06, + "loss": 3.191, + "step": 17594 + }, + { + "epoch": 1.4105750536168848, + "grad_norm": 0.6713624000549316, + "learning_rate": 4.217618254398736e-06, + "loss": 2.4656, + "step": 17595 + }, + { + "epoch": 1.4106552283979075, + "grad_norm": 1.7135316133499146, + "learning_rate": 4.2165589417269885e-06, + "loss": 3.2495, + "step": 17596 + }, + { + "epoch": 1.41073540317893, + "grad_norm": 1.725771427154541, + "learning_rate": 4.215499726562017e-06, + "loss": 3.219, + "step": 17597 + }, + { + "epoch": 1.4108155779599527, + "grad_norm": 1.7173084020614624, + "learning_rate": 4.2144406089216826e-06, + "loss": 3.1246, + "step": 17598 + }, + { + "epoch": 1.4108957527409753, + "grad_norm": 1.7396492958068848, + "learning_rate": 4.213381588823844e-06, + "loss": 3.1975, + "step": 17599 + }, + { + "epoch": 1.410975927521998, + "grad_norm": 1.7371225357055664, + "learning_rate": 4.212322666286348e-06, + "loss": 3.137, + "step": 17600 + }, + { + "epoch": 1.4110561023030206, + "grad_norm": 1.7630633115768433, + "learning_rate": 4.211263841327052e-06, + "loss": 3.1042, + "step": 17601 + }, + { + "epoch": 1.4111362770840432, + "grad_norm": 1.8755912780761719, + "learning_rate": 4.210205113963809e-06, + "loss": 3.0902, + "step": 17602 + }, + { + "epoch": 1.4112164518650658, + "grad_norm": 1.8259985446929932, + "learning_rate": 4.209146484214466e-06, + "loss": 3.2344, + "step": 17603 + }, + { + "epoch": 1.4112966266460885, + "grad_norm": 1.724064588546753, + "learning_rate": 4.208087952096877e-06, + "loss": 3.0589, + "step": 17604 + }, + { + "epoch": 1.411376801427111, + "grad_norm": 1.9064081907272339, + "learning_rate": 4.207029517628879e-06, + "loss": 3.2235, + "step": 17605 + }, + { + "epoch": 1.4114569762081337, + "grad_norm": 1.7177578210830688, + "learning_rate": 4.205971180828322e-06, + "loss": 3.1552, + "step": 17606 + }, + { + "epoch": 1.4115371509891563, + "grad_norm": 0.6844996213912964, + "learning_rate": 4.204912941713049e-06, + "loss": 2.468, + "step": 17607 + }, + { + "epoch": 1.411617325770179, + "grad_norm": 1.7110474109649658, + "learning_rate": 4.203854800300902e-06, + "loss": 3.2416, + "step": 17608 + }, + { + "epoch": 1.4116975005512016, + "grad_norm": 1.7434319257736206, + "learning_rate": 4.202796756609723e-06, + "loss": 3.1427, + "step": 17609 + }, + { + "epoch": 1.4117776753322242, + "grad_norm": 1.7182042598724365, + "learning_rate": 4.201738810657344e-06, + "loss": 3.1497, + "step": 17610 + }, + { + "epoch": 1.4118578501132468, + "grad_norm": 1.833656668663025, + "learning_rate": 4.2006809624616055e-06, + "loss": 3.2251, + "step": 17611 + }, + { + "epoch": 1.4119380248942695, + "grad_norm": 1.7837417125701904, + "learning_rate": 4.199623212040343e-06, + "loss": 3.3032, + "step": 17612 + }, + { + "epoch": 1.412018199675292, + "grad_norm": 1.6606788635253906, + "learning_rate": 4.198565559411387e-06, + "loss": 3.2008, + "step": 17613 + }, + { + "epoch": 1.4120983744563147, + "grad_norm": 1.7445859909057617, + "learning_rate": 4.197508004592576e-06, + "loss": 3.1102, + "step": 17614 + }, + { + "epoch": 1.4121785492373373, + "grad_norm": 1.798875093460083, + "learning_rate": 4.19645054760173e-06, + "loss": 3.2374, + "step": 17615 + }, + { + "epoch": 1.41225872401836, + "grad_norm": 1.7852827310562134, + "learning_rate": 4.1953931884566845e-06, + "loss": 3.1882, + "step": 17616 + }, + { + "epoch": 1.4123388987993826, + "grad_norm": 0.6618704795837402, + "learning_rate": 4.194335927175262e-06, + "loss": 2.4802, + "step": 17617 + }, + { + "epoch": 1.4124190735804052, + "grad_norm": 1.7480971813201904, + "learning_rate": 4.193278763775291e-06, + "loss": 3.2059, + "step": 17618 + }, + { + "epoch": 1.4124992483614278, + "grad_norm": 1.8472449779510498, + "learning_rate": 4.192221698274598e-06, + "loss": 3.2581, + "step": 17619 + }, + { + "epoch": 1.4125794231424504, + "grad_norm": 1.8154280185699463, + "learning_rate": 4.1911647306909955e-06, + "loss": 3.2388, + "step": 17620 + }, + { + "epoch": 1.412659597923473, + "grad_norm": 1.700317144393921, + "learning_rate": 4.190107861042309e-06, + "loss": 3.2399, + "step": 17621 + }, + { + "epoch": 1.4127397727044957, + "grad_norm": 1.773987889289856, + "learning_rate": 4.189051089346355e-06, + "loss": 3.2109, + "step": 17622 + }, + { + "epoch": 1.4128199474855183, + "grad_norm": 1.832885503768921, + "learning_rate": 4.187994415620953e-06, + "loss": 3.0787, + "step": 17623 + }, + { + "epoch": 1.412900122266541, + "grad_norm": 1.7796136140823364, + "learning_rate": 4.186937839883919e-06, + "loss": 3.149, + "step": 17624 + }, + { + "epoch": 1.4129802970475636, + "grad_norm": 1.7629845142364502, + "learning_rate": 4.185881362153062e-06, + "loss": 3.2509, + "step": 17625 + }, + { + "epoch": 1.4130604718285862, + "grad_norm": 1.9267857074737549, + "learning_rate": 4.184824982446196e-06, + "loss": 3.2365, + "step": 17626 + }, + { + "epoch": 1.413140646609609, + "grad_norm": 1.7991902828216553, + "learning_rate": 4.183768700781131e-06, + "loss": 3.2457, + "step": 17627 + }, + { + "epoch": 1.4132208213906317, + "grad_norm": 1.763295292854309, + "learning_rate": 4.182712517175676e-06, + "loss": 3.2793, + "step": 17628 + }, + { + "epoch": 1.4133009961716543, + "grad_norm": 1.8017421960830688, + "learning_rate": 4.1816564316476416e-06, + "loss": 3.2244, + "step": 17629 + }, + { + "epoch": 1.413381170952677, + "grad_norm": 0.6942033171653748, + "learning_rate": 4.180600444214825e-06, + "loss": 2.4686, + "step": 17630 + }, + { + "epoch": 1.4134613457336995, + "grad_norm": 1.7313001155853271, + "learning_rate": 4.179544554895034e-06, + "loss": 3.2186, + "step": 17631 + }, + { + "epoch": 1.4135415205147221, + "grad_norm": 1.761088490486145, + "learning_rate": 4.178488763706071e-06, + "loss": 3.2242, + "step": 17632 + }, + { + "epoch": 1.4136216952957448, + "grad_norm": 1.820457935333252, + "learning_rate": 4.177433070665735e-06, + "loss": 3.2302, + "step": 17633 + }, + { + "epoch": 1.4137018700767674, + "grad_norm": 1.7472994327545166, + "learning_rate": 4.176377475791829e-06, + "loss": 3.1395, + "step": 17634 + }, + { + "epoch": 1.41378204485779, + "grad_norm": 1.8223614692687988, + "learning_rate": 4.1753219791021436e-06, + "loss": 3.1538, + "step": 17635 + }, + { + "epoch": 1.4138622196388126, + "grad_norm": 1.8328392505645752, + "learning_rate": 4.174266580614477e-06, + "loss": 3.1526, + "step": 17636 + }, + { + "epoch": 1.4139423944198353, + "grad_norm": 1.8206450939178467, + "learning_rate": 4.1732112803466215e-06, + "loss": 3.1492, + "step": 17637 + }, + { + "epoch": 1.414022569200858, + "grad_norm": 1.8109923601150513, + "learning_rate": 4.172156078316371e-06, + "loss": 3.1416, + "step": 17638 + }, + { + "epoch": 1.4141027439818805, + "grad_norm": 1.8207491636276245, + "learning_rate": 4.1711009745415184e-06, + "loss": 3.3138, + "step": 17639 + }, + { + "epoch": 1.4141829187629031, + "grad_norm": 1.8262721300125122, + "learning_rate": 4.170045969039845e-06, + "loss": 3.2401, + "step": 17640 + }, + { + "epoch": 1.4142630935439258, + "grad_norm": 1.9014694690704346, + "learning_rate": 4.168991061829143e-06, + "loss": 3.2394, + "step": 17641 + }, + { + "epoch": 1.4143432683249484, + "grad_norm": 1.769447922706604, + "learning_rate": 4.167936252927196e-06, + "loss": 3.2098, + "step": 17642 + }, + { + "epoch": 1.414423443105971, + "grad_norm": 1.6774883270263672, + "learning_rate": 4.166881542351788e-06, + "loss": 3.1121, + "step": 17643 + }, + { + "epoch": 1.4145036178869936, + "grad_norm": 1.7624547481536865, + "learning_rate": 4.165826930120706e-06, + "loss": 3.1691, + "step": 17644 + }, + { + "epoch": 1.4145837926680163, + "grad_norm": 1.7719221115112305, + "learning_rate": 4.1647724162517215e-06, + "loss": 3.1219, + "step": 17645 + }, + { + "epoch": 1.4146639674490389, + "grad_norm": 2.0960497856140137, + "learning_rate": 4.163718000762617e-06, + "loss": 3.1587, + "step": 17646 + }, + { + "epoch": 1.4147441422300615, + "grad_norm": 1.8252780437469482, + "learning_rate": 4.162663683671171e-06, + "loss": 3.3461, + "step": 17647 + }, + { + "epoch": 1.4148243170110841, + "grad_norm": 1.8500243425369263, + "learning_rate": 4.1616094649951575e-06, + "loss": 3.2708, + "step": 17648 + }, + { + "epoch": 1.4149044917921068, + "grad_norm": 1.7642803192138672, + "learning_rate": 4.160555344752355e-06, + "loss": 3.13, + "step": 17649 + }, + { + "epoch": 1.4149846665731294, + "grad_norm": 0.7012763023376465, + "learning_rate": 4.159501322960528e-06, + "loss": 2.4851, + "step": 17650 + }, + { + "epoch": 1.415064841354152, + "grad_norm": 1.8222614526748657, + "learning_rate": 4.158447399637449e-06, + "loss": 3.2428, + "step": 17651 + }, + { + "epoch": 1.4151450161351746, + "grad_norm": 1.7493633031845093, + "learning_rate": 4.157393574800888e-06, + "loss": 3.1654, + "step": 17652 + }, + { + "epoch": 1.4152251909161973, + "grad_norm": 0.6787571310997009, + "learning_rate": 4.156339848468613e-06, + "loss": 2.4786, + "step": 17653 + }, + { + "epoch": 1.4153053656972199, + "grad_norm": 1.7273424863815308, + "learning_rate": 4.155286220658393e-06, + "loss": 3.2719, + "step": 17654 + }, + { + "epoch": 1.4153855404782425, + "grad_norm": 1.7674834728240967, + "learning_rate": 4.154232691387983e-06, + "loss": 3.2106, + "step": 17655 + }, + { + "epoch": 1.4154657152592651, + "grad_norm": 1.836363434791565, + "learning_rate": 4.15317926067515e-06, + "loss": 3.1212, + "step": 17656 + }, + { + "epoch": 1.4155458900402877, + "grad_norm": 1.8904427289962769, + "learning_rate": 4.152125928537654e-06, + "loss": 3.2253, + "step": 17657 + }, + { + "epoch": 1.4156260648213106, + "grad_norm": 1.8729475736618042, + "learning_rate": 4.151072694993253e-06, + "loss": 3.3208, + "step": 17658 + }, + { + "epoch": 1.4157062396023332, + "grad_norm": 1.6944351196289062, + "learning_rate": 4.150019560059709e-06, + "loss": 3.2115, + "step": 17659 + }, + { + "epoch": 1.4157864143833558, + "grad_norm": 0.6645165085792542, + "learning_rate": 4.14896652375477e-06, + "loss": 2.4645, + "step": 17660 + }, + { + "epoch": 1.4158665891643785, + "grad_norm": 0.6528031826019287, + "learning_rate": 4.147913586096194e-06, + "loss": 2.4642, + "step": 17661 + }, + { + "epoch": 1.415946763945401, + "grad_norm": 1.748847484588623, + "learning_rate": 4.1468607471017304e-06, + "loss": 3.1826, + "step": 17662 + }, + { + "epoch": 1.4160269387264237, + "grad_norm": 1.8016715049743652, + "learning_rate": 4.145808006789134e-06, + "loss": 3.2179, + "step": 17663 + }, + { + "epoch": 1.4161071135074463, + "grad_norm": 1.8022314310073853, + "learning_rate": 4.144755365176154e-06, + "loss": 3.2849, + "step": 17664 + }, + { + "epoch": 1.416187288288469, + "grad_norm": 1.6943247318267822, + "learning_rate": 4.143702822280532e-06, + "loss": 3.1094, + "step": 17665 + }, + { + "epoch": 1.4162674630694916, + "grad_norm": 1.6928075551986694, + "learning_rate": 4.142650378120016e-06, + "loss": 3.078, + "step": 17666 + }, + { + "epoch": 1.4163476378505142, + "grad_norm": 1.8259538412094116, + "learning_rate": 4.1415980327123496e-06, + "loss": 3.1212, + "step": 17667 + }, + { + "epoch": 1.4164278126315368, + "grad_norm": 1.675811529159546, + "learning_rate": 4.140545786075276e-06, + "loss": 3.0835, + "step": 17668 + }, + { + "epoch": 1.4165079874125595, + "grad_norm": 1.7594568729400635, + "learning_rate": 4.139493638226541e-06, + "loss": 3.1758, + "step": 17669 + }, + { + "epoch": 1.416588162193582, + "grad_norm": 1.707206130027771, + "learning_rate": 4.138441589183871e-06, + "loss": 3.2088, + "step": 17670 + }, + { + "epoch": 1.4166683369746047, + "grad_norm": 1.8551722764968872, + "learning_rate": 4.137389638965011e-06, + "loss": 3.2922, + "step": 17671 + }, + { + "epoch": 1.4167485117556273, + "grad_norm": 1.8153221607208252, + "learning_rate": 4.1363377875876965e-06, + "loss": 3.3759, + "step": 17672 + }, + { + "epoch": 1.41682868653665, + "grad_norm": 1.6601266860961914, + "learning_rate": 4.13528603506966e-06, + "loss": 3.1741, + "step": 17673 + }, + { + "epoch": 1.4169088613176726, + "grad_norm": 0.6714270114898682, + "learning_rate": 4.134234381428638e-06, + "loss": 2.4752, + "step": 17674 + }, + { + "epoch": 1.4169890360986952, + "grad_norm": 1.9065693616867065, + "learning_rate": 4.133182826682353e-06, + "loss": 3.0816, + "step": 17675 + }, + { + "epoch": 1.4170692108797178, + "grad_norm": 1.829636812210083, + "learning_rate": 4.132131370848538e-06, + "loss": 3.2277, + "step": 17676 + }, + { + "epoch": 1.4171493856607404, + "grad_norm": 1.7189401388168335, + "learning_rate": 4.131080013944919e-06, + "loss": 3.1099, + "step": 17677 + }, + { + "epoch": 1.417229560441763, + "grad_norm": 1.755601167678833, + "learning_rate": 4.130028755989224e-06, + "loss": 3.1424, + "step": 17678 + }, + { + "epoch": 1.4173097352227857, + "grad_norm": 1.8034974336624146, + "learning_rate": 4.128977596999179e-06, + "loss": 3.1727, + "step": 17679 + }, + { + "epoch": 1.4173899100038083, + "grad_norm": 1.7761166095733643, + "learning_rate": 4.127926536992497e-06, + "loss": 3.1383, + "step": 17680 + }, + { + "epoch": 1.417470084784831, + "grad_norm": 1.7677593231201172, + "learning_rate": 4.126875575986906e-06, + "loss": 3.1802, + "step": 17681 + }, + { + "epoch": 1.4175502595658536, + "grad_norm": 1.7080671787261963, + "learning_rate": 4.125824714000122e-06, + "loss": 3.092, + "step": 17682 + }, + { + "epoch": 1.4176304343468762, + "grad_norm": 0.728672444820404, + "learning_rate": 4.124773951049863e-06, + "loss": 2.4711, + "step": 17683 + }, + { + "epoch": 1.4177106091278988, + "grad_norm": 1.82059907913208, + "learning_rate": 4.1237232871538494e-06, + "loss": 3.1298, + "step": 17684 + }, + { + "epoch": 1.4177907839089214, + "grad_norm": 1.9096909761428833, + "learning_rate": 4.122672722329786e-06, + "loss": 3.1647, + "step": 17685 + }, + { + "epoch": 1.417870958689944, + "grad_norm": 1.6731771230697632, + "learning_rate": 4.121622256595388e-06, + "loss": 3.1456, + "step": 17686 + }, + { + "epoch": 1.4179511334709667, + "grad_norm": 1.8103598356246948, + "learning_rate": 4.120571889968368e-06, + "loss": 3.1681, + "step": 17687 + }, + { + "epoch": 1.4180313082519893, + "grad_norm": 1.6754158735275269, + "learning_rate": 4.119521622466434e-06, + "loss": 3.1063, + "step": 17688 + }, + { + "epoch": 1.418111483033012, + "grad_norm": 1.798073172569275, + "learning_rate": 4.118471454107296e-06, + "loss": 3.1916, + "step": 17689 + }, + { + "epoch": 1.4181916578140346, + "grad_norm": 1.8477591276168823, + "learning_rate": 4.117421384908653e-06, + "loss": 3.1403, + "step": 17690 + }, + { + "epoch": 1.4182718325950572, + "grad_norm": 1.8118717670440674, + "learning_rate": 4.116371414888213e-06, + "loss": 3.0908, + "step": 17691 + }, + { + "epoch": 1.4183520073760798, + "grad_norm": 1.8131637573242188, + "learning_rate": 4.115321544063676e-06, + "loss": 3.1869, + "step": 17692 + }, + { + "epoch": 1.4184321821571024, + "grad_norm": 1.779558539390564, + "learning_rate": 4.1142717724527444e-06, + "loss": 3.2313, + "step": 17693 + }, + { + "epoch": 1.418512356938125, + "grad_norm": 1.7564723491668701, + "learning_rate": 4.11322210007312e-06, + "loss": 3.2981, + "step": 17694 + }, + { + "epoch": 1.4185925317191477, + "grad_norm": 1.7150065898895264, + "learning_rate": 4.1121725269424926e-06, + "loss": 3.1495, + "step": 17695 + }, + { + "epoch": 1.4186727065001703, + "grad_norm": 1.8030626773834229, + "learning_rate": 4.1111230530785615e-06, + "loss": 3.1002, + "step": 17696 + }, + { + "epoch": 1.418752881281193, + "grad_norm": 1.7934869527816772, + "learning_rate": 4.11007367849902e-06, + "loss": 3.2103, + "step": 17697 + }, + { + "epoch": 1.4188330560622155, + "grad_norm": 0.6755185127258301, + "learning_rate": 4.109024403221561e-06, + "loss": 2.4666, + "step": 17698 + }, + { + "epoch": 1.4189132308432382, + "grad_norm": 1.782353401184082, + "learning_rate": 4.107975227263877e-06, + "loss": 3.1916, + "step": 17699 + }, + { + "epoch": 1.4189934056242608, + "grad_norm": 1.7153520584106445, + "learning_rate": 4.106926150643652e-06, + "loss": 3.2142, + "step": 17700 + }, + { + "epoch": 1.4190735804052834, + "grad_norm": 1.903050422668457, + "learning_rate": 4.105877173378574e-06, + "loss": 3.1326, + "step": 17701 + }, + { + "epoch": 1.419153755186306, + "grad_norm": 1.710112452507019, + "learning_rate": 4.104828295486331e-06, + "loss": 3.147, + "step": 17702 + }, + { + "epoch": 1.4192339299673287, + "grad_norm": 1.8045316934585571, + "learning_rate": 4.103779516984604e-06, + "loss": 3.251, + "step": 17703 + }, + { + "epoch": 1.4193141047483513, + "grad_norm": 0.6970905065536499, + "learning_rate": 4.10273083789108e-06, + "loss": 2.4685, + "step": 17704 + }, + { + "epoch": 1.419394279529374, + "grad_norm": 1.8106769323349, + "learning_rate": 4.101682258223432e-06, + "loss": 3.2511, + "step": 17705 + }, + { + "epoch": 1.4194744543103965, + "grad_norm": 1.765572428703308, + "learning_rate": 4.100633777999342e-06, + "loss": 3.1519, + "step": 17706 + }, + { + "epoch": 1.4195546290914192, + "grad_norm": 1.8564846515655518, + "learning_rate": 4.099585397236488e-06, + "loss": 3.1667, + "step": 17707 + }, + { + "epoch": 1.4196348038724418, + "grad_norm": 1.8462141752243042, + "learning_rate": 4.098537115952545e-06, + "loss": 3.1283, + "step": 17708 + }, + { + "epoch": 1.4197149786534646, + "grad_norm": 1.8084964752197266, + "learning_rate": 4.09748893416519e-06, + "loss": 3.1016, + "step": 17709 + }, + { + "epoch": 1.4197951534344873, + "grad_norm": 1.8512948751449585, + "learning_rate": 4.096440851892087e-06, + "loss": 3.1793, + "step": 17710 + }, + { + "epoch": 1.4198753282155099, + "grad_norm": 1.7190011739730835, + "learning_rate": 4.095392869150912e-06, + "loss": 3.1447, + "step": 17711 + }, + { + "epoch": 1.4199555029965325, + "grad_norm": 1.7744860649108887, + "learning_rate": 4.09434498595933e-06, + "loss": 3.2371, + "step": 17712 + }, + { + "epoch": 1.4200356777775551, + "grad_norm": 1.7789993286132812, + "learning_rate": 4.093297202335012e-06, + "loss": 3.123, + "step": 17713 + }, + { + "epoch": 1.4201158525585778, + "grad_norm": 1.7651726007461548, + "learning_rate": 4.09224951829562e-06, + "loss": 3.2726, + "step": 17714 + }, + { + "epoch": 1.4201960273396004, + "grad_norm": 1.7416908740997314, + "learning_rate": 4.091201933858824e-06, + "loss": 3.1801, + "step": 17715 + }, + { + "epoch": 1.420276202120623, + "grad_norm": 1.7337579727172852, + "learning_rate": 4.090154449042276e-06, + "loss": 3.3234, + "step": 17716 + }, + { + "epoch": 1.4203563769016456, + "grad_norm": 1.7352572679519653, + "learning_rate": 4.089107063863642e-06, + "loss": 3.1855, + "step": 17717 + }, + { + "epoch": 1.4204365516826682, + "grad_norm": 1.8727916479110718, + "learning_rate": 4.0880597783405795e-06, + "loss": 3.2922, + "step": 17718 + }, + { + "epoch": 1.4205167264636909, + "grad_norm": 1.9086686372756958, + "learning_rate": 4.087012592490746e-06, + "loss": 3.1073, + "step": 17719 + }, + { + "epoch": 1.4205969012447135, + "grad_norm": 1.7389742136001587, + "learning_rate": 4.085965506331799e-06, + "loss": 3.1691, + "step": 17720 + }, + { + "epoch": 1.4206770760257361, + "grad_norm": 1.700467586517334, + "learning_rate": 4.0849185198813865e-06, + "loss": 3.0967, + "step": 17721 + }, + { + "epoch": 1.4207572508067587, + "grad_norm": 1.7100071907043457, + "learning_rate": 4.0838716331571635e-06, + "loss": 3.1577, + "step": 17722 + }, + { + "epoch": 1.4208374255877814, + "grad_norm": 1.701325535774231, + "learning_rate": 4.082824846176778e-06, + "loss": 3.1849, + "step": 17723 + }, + { + "epoch": 1.420917600368804, + "grad_norm": 1.7245103120803833, + "learning_rate": 4.081778158957882e-06, + "loss": 3.1183, + "step": 17724 + }, + { + "epoch": 1.4209977751498266, + "grad_norm": 1.745324730873108, + "learning_rate": 4.0807315715181225e-06, + "loss": 3.1713, + "step": 17725 + }, + { + "epoch": 1.4210779499308492, + "grad_norm": 1.817929744720459, + "learning_rate": 4.07968508387514e-06, + "loss": 3.2662, + "step": 17726 + }, + { + "epoch": 1.4211581247118719, + "grad_norm": 0.6856880784034729, + "learning_rate": 4.078638696046581e-06, + "loss": 2.4727, + "step": 17727 + }, + { + "epoch": 1.4212382994928945, + "grad_norm": 1.9095344543457031, + "learning_rate": 4.077592408050085e-06, + "loss": 3.166, + "step": 17728 + }, + { + "epoch": 1.421318474273917, + "grad_norm": 1.7293976545333862, + "learning_rate": 4.0765462199032945e-06, + "loss": 3.1951, + "step": 17729 + }, + { + "epoch": 1.4213986490549397, + "grad_norm": 1.8323029279708862, + "learning_rate": 4.075500131623851e-06, + "loss": 3.194, + "step": 17730 + }, + { + "epoch": 1.4214788238359624, + "grad_norm": 0.6773753762245178, + "learning_rate": 4.074454143229384e-06, + "loss": 2.4717, + "step": 17731 + }, + { + "epoch": 1.421558998616985, + "grad_norm": 1.7427091598510742, + "learning_rate": 4.0734082547375306e-06, + "loss": 3.0266, + "step": 17732 + }, + { + "epoch": 1.4216391733980076, + "grad_norm": 1.7763042449951172, + "learning_rate": 4.072362466165926e-06, + "loss": 3.1499, + "step": 17733 + }, + { + "epoch": 1.4217193481790302, + "grad_norm": 1.8453024625778198, + "learning_rate": 4.0713167775322015e-06, + "loss": 3.0973, + "step": 17734 + }, + { + "epoch": 1.4217995229600529, + "grad_norm": 1.7808468341827393, + "learning_rate": 4.0702711888539895e-06, + "loss": 3.2465, + "step": 17735 + }, + { + "epoch": 1.4218796977410755, + "grad_norm": 1.7921723127365112, + "learning_rate": 4.069225700148912e-06, + "loss": 3.2546, + "step": 17736 + }, + { + "epoch": 1.421959872522098, + "grad_norm": 1.8902034759521484, + "learning_rate": 4.0681803114346e-06, + "loss": 3.2471, + "step": 17737 + }, + { + "epoch": 1.4220400473031207, + "grad_norm": 1.7172335386276245, + "learning_rate": 4.067135022728675e-06, + "loss": 3.0728, + "step": 17738 + }, + { + "epoch": 1.4221202220841436, + "grad_norm": 1.7215726375579834, + "learning_rate": 4.0660898340487655e-06, + "loss": 3.1488, + "step": 17739 + }, + { + "epoch": 1.4222003968651662, + "grad_norm": 1.8022857904434204, + "learning_rate": 4.0650447454124915e-06, + "loss": 3.2177, + "step": 17740 + }, + { + "epoch": 1.4222805716461888, + "grad_norm": 1.7493120431900024, + "learning_rate": 4.063999756837469e-06, + "loss": 3.1221, + "step": 17741 + }, + { + "epoch": 1.4223607464272114, + "grad_norm": 1.7917375564575195, + "learning_rate": 4.062954868341318e-06, + "loss": 3.2127, + "step": 17742 + }, + { + "epoch": 1.422440921208234, + "grad_norm": 1.7823203802108765, + "learning_rate": 4.061910079941656e-06, + "loss": 3.1887, + "step": 17743 + }, + { + "epoch": 1.4225210959892567, + "grad_norm": 1.8055312633514404, + "learning_rate": 4.0608653916560955e-06, + "loss": 3.1971, + "step": 17744 + }, + { + "epoch": 1.4226012707702793, + "grad_norm": 1.7259398698806763, + "learning_rate": 4.0598208035022565e-06, + "loss": 3.0558, + "step": 17745 + }, + { + "epoch": 1.422681445551302, + "grad_norm": 1.7653239965438843, + "learning_rate": 4.0587763154977426e-06, + "loss": 3.155, + "step": 17746 + }, + { + "epoch": 1.4227616203323246, + "grad_norm": 0.6744408011436462, + "learning_rate": 4.057731927660164e-06, + "loss": 2.4572, + "step": 17747 + }, + { + "epoch": 1.4228417951133472, + "grad_norm": 1.9951955080032349, + "learning_rate": 4.056687640007132e-06, + "loss": 3.1599, + "step": 17748 + }, + { + "epoch": 1.4229219698943698, + "grad_norm": 1.8994590044021606, + "learning_rate": 4.055643452556252e-06, + "loss": 3.2031, + "step": 17749 + }, + { + "epoch": 1.4230021446753924, + "grad_norm": 1.6599770784378052, + "learning_rate": 4.054599365325131e-06, + "loss": 3.1385, + "step": 17750 + }, + { + "epoch": 1.423082319456415, + "grad_norm": 1.8051470518112183, + "learning_rate": 4.053555378331366e-06, + "loss": 3.0186, + "step": 17751 + }, + { + "epoch": 1.4231624942374377, + "grad_norm": 1.8526601791381836, + "learning_rate": 4.052511491592562e-06, + "loss": 3.1509, + "step": 17752 + }, + { + "epoch": 1.4232426690184603, + "grad_norm": 1.7435745000839233, + "learning_rate": 4.051467705126318e-06, + "loss": 3.2097, + "step": 17753 + }, + { + "epoch": 1.423322843799483, + "grad_norm": 1.7693442106246948, + "learning_rate": 4.050424018950233e-06, + "loss": 3.1352, + "step": 17754 + }, + { + "epoch": 1.4234030185805056, + "grad_norm": 1.842192530632019, + "learning_rate": 4.049380433081904e-06, + "loss": 3.1847, + "step": 17755 + }, + { + "epoch": 1.4234831933615282, + "grad_norm": 1.7727326154708862, + "learning_rate": 4.0483369475389215e-06, + "loss": 3.13, + "step": 17756 + }, + { + "epoch": 1.4235633681425508, + "grad_norm": 1.6539719104766846, + "learning_rate": 4.04729356233888e-06, + "loss": 3.0742, + "step": 17757 + }, + { + "epoch": 1.4236435429235734, + "grad_norm": 1.9022111892700195, + "learning_rate": 4.046250277499371e-06, + "loss": 3.2169, + "step": 17758 + }, + { + "epoch": 1.423723717704596, + "grad_norm": 1.8452138900756836, + "learning_rate": 4.0452070930379835e-06, + "loss": 3.1711, + "step": 17759 + }, + { + "epoch": 1.4238038924856187, + "grad_norm": 1.7964420318603516, + "learning_rate": 4.044164008972311e-06, + "loss": 3.172, + "step": 17760 + }, + { + "epoch": 1.4238840672666413, + "grad_norm": 1.7948516607284546, + "learning_rate": 4.043121025319929e-06, + "loss": 3.2018, + "step": 17761 + }, + { + "epoch": 1.423964242047664, + "grad_norm": 1.8277779817581177, + "learning_rate": 4.042078142098428e-06, + "loss": 3.2697, + "step": 17762 + }, + { + "epoch": 1.4240444168286865, + "grad_norm": 1.7776634693145752, + "learning_rate": 4.041035359325389e-06, + "loss": 3.1619, + "step": 17763 + }, + { + "epoch": 1.4241245916097092, + "grad_norm": 1.7777550220489502, + "learning_rate": 4.039992677018395e-06, + "loss": 3.1895, + "step": 17764 + }, + { + "epoch": 1.4242047663907318, + "grad_norm": 1.6349108219146729, + "learning_rate": 4.0389500951950265e-06, + "loss": 3.0935, + "step": 17765 + }, + { + "epoch": 1.4242849411717544, + "grad_norm": 1.789314866065979, + "learning_rate": 4.037907613872856e-06, + "loss": 3.2163, + "step": 17766 + }, + { + "epoch": 1.424365115952777, + "grad_norm": 1.8320263624191284, + "learning_rate": 4.03686523306946e-06, + "loss": 3.1844, + "step": 17767 + }, + { + "epoch": 1.4244452907337997, + "grad_norm": 1.6914159059524536, + "learning_rate": 4.035822952802416e-06, + "loss": 3.1069, + "step": 17768 + }, + { + "epoch": 1.4245254655148223, + "grad_norm": 1.7293387651443481, + "learning_rate": 4.034780773089294e-06, + "loss": 3.1296, + "step": 17769 + }, + { + "epoch": 1.424605640295845, + "grad_norm": 1.7070828676223755, + "learning_rate": 4.033738693947671e-06, + "loss": 3.1458, + "step": 17770 + }, + { + "epoch": 1.4246858150768675, + "grad_norm": 1.782771348953247, + "learning_rate": 4.032696715395106e-06, + "loss": 3.1064, + "step": 17771 + }, + { + "epoch": 1.4247659898578902, + "grad_norm": 1.9269874095916748, + "learning_rate": 4.031654837449172e-06, + "loss": 3.3064, + "step": 17772 + }, + { + "epoch": 1.4248461646389128, + "grad_norm": 1.7933741807937622, + "learning_rate": 4.030613060127433e-06, + "loss": 3.1206, + "step": 17773 + }, + { + "epoch": 1.4249263394199354, + "grad_norm": 1.8279869556427002, + "learning_rate": 4.029571383447455e-06, + "loss": 3.1249, + "step": 17774 + }, + { + "epoch": 1.425006514200958, + "grad_norm": 1.7102566957473755, + "learning_rate": 4.028529807426802e-06, + "loss": 3.1671, + "step": 17775 + }, + { + "epoch": 1.4250866889819807, + "grad_norm": 1.7484755516052246, + "learning_rate": 4.027488332083029e-06, + "loss": 3.2676, + "step": 17776 + }, + { + "epoch": 1.4251668637630033, + "grad_norm": 1.6736146211624146, + "learning_rate": 4.026446957433697e-06, + "loss": 3.2036, + "step": 17777 + }, + { + "epoch": 1.425247038544026, + "grad_norm": 1.7376513481140137, + "learning_rate": 4.025405683496365e-06, + "loss": 3.1587, + "step": 17778 + }, + { + "epoch": 1.4253272133250485, + "grad_norm": 1.8049991130828857, + "learning_rate": 4.024364510288588e-06, + "loss": 3.2458, + "step": 17779 + }, + { + "epoch": 1.4254073881060711, + "grad_norm": 1.723332166671753, + "learning_rate": 4.023323437827922e-06, + "loss": 3.1877, + "step": 17780 + }, + { + "epoch": 1.4254875628870938, + "grad_norm": 1.8083714246749878, + "learning_rate": 4.022282466131913e-06, + "loss": 3.1286, + "step": 17781 + }, + { + "epoch": 1.4255677376681164, + "grad_norm": 1.6926687955856323, + "learning_rate": 4.0212415952181135e-06, + "loss": 3.1902, + "step": 17782 + }, + { + "epoch": 1.425647912449139, + "grad_norm": 1.794205665588379, + "learning_rate": 4.020200825104076e-06, + "loss": 3.2057, + "step": 17783 + }, + { + "epoch": 1.4257280872301616, + "grad_norm": 1.8703441619873047, + "learning_rate": 4.0191601558073435e-06, + "loss": 3.1587, + "step": 17784 + }, + { + "epoch": 1.4258082620111843, + "grad_norm": 1.7824300527572632, + "learning_rate": 4.018119587345467e-06, + "loss": 3.1418, + "step": 17785 + }, + { + "epoch": 1.425888436792207, + "grad_norm": 1.7860559225082397, + "learning_rate": 4.017079119735983e-06, + "loss": 3.1664, + "step": 17786 + }, + { + "epoch": 1.4259686115732295, + "grad_norm": 1.8085269927978516, + "learning_rate": 4.016038752996435e-06, + "loss": 3.2517, + "step": 17787 + }, + { + "epoch": 1.4260487863542521, + "grad_norm": 1.7568283081054688, + "learning_rate": 4.0149984871443656e-06, + "loss": 3.1834, + "step": 17788 + }, + { + "epoch": 1.4261289611352748, + "grad_norm": 1.7564336061477661, + "learning_rate": 4.013958322197313e-06, + "loss": 3.1161, + "step": 17789 + }, + { + "epoch": 1.4262091359162976, + "grad_norm": 1.7511216402053833, + "learning_rate": 4.012918258172817e-06, + "loss": 3.1122, + "step": 17790 + }, + { + "epoch": 1.4262893106973202, + "grad_norm": 1.854215383529663, + "learning_rate": 4.011878295088404e-06, + "loss": 3.111, + "step": 17791 + }, + { + "epoch": 1.4263694854783429, + "grad_norm": 1.680510401725769, + "learning_rate": 4.010838432961613e-06, + "loss": 3.1133, + "step": 17792 + }, + { + "epoch": 1.4264496602593655, + "grad_norm": 1.788317322731018, + "learning_rate": 4.009798671809975e-06, + "loss": 3.2045, + "step": 17793 + }, + { + "epoch": 1.426529835040388, + "grad_norm": 1.742424726486206, + "learning_rate": 4.008759011651022e-06, + "loss": 3.1218, + "step": 17794 + }, + { + "epoch": 1.4266100098214107, + "grad_norm": 1.8434512615203857, + "learning_rate": 4.007719452502283e-06, + "loss": 3.2361, + "step": 17795 + }, + { + "epoch": 1.4266901846024334, + "grad_norm": 1.7504487037658691, + "learning_rate": 4.0066799943812785e-06, + "loss": 3.2168, + "step": 17796 + }, + { + "epoch": 1.426770359383456, + "grad_norm": 1.8397353887557983, + "learning_rate": 4.005640637305538e-06, + "loss": 3.1854, + "step": 17797 + }, + { + "epoch": 1.4268505341644786, + "grad_norm": 1.7566279172897339, + "learning_rate": 4.0046013812925825e-06, + "loss": 3.0955, + "step": 17798 + }, + { + "epoch": 1.4269307089455012, + "grad_norm": 1.7850524187088013, + "learning_rate": 4.003562226359935e-06, + "loss": 3.203, + "step": 17799 + }, + { + "epoch": 1.4270108837265238, + "grad_norm": 1.804450273513794, + "learning_rate": 4.002523172525119e-06, + "loss": 3.1485, + "step": 17800 + }, + { + "epoch": 1.4270910585075465, + "grad_norm": 1.8450210094451904, + "learning_rate": 4.0014842198056446e-06, + "loss": 3.2341, + "step": 17801 + }, + { + "epoch": 1.427171233288569, + "grad_norm": 1.7435132265090942, + "learning_rate": 4.000445368219033e-06, + "loss": 3.1902, + "step": 17802 + }, + { + "epoch": 1.4272514080695917, + "grad_norm": 1.7679567337036133, + "learning_rate": 3.9994066177827965e-06, + "loss": 3.3403, + "step": 17803 + }, + { + "epoch": 1.4273315828506143, + "grad_norm": 1.7145740985870361, + "learning_rate": 3.998367968514451e-06, + "loss": 3.1744, + "step": 17804 + }, + { + "epoch": 1.427411757631637, + "grad_norm": 1.7345154285430908, + "learning_rate": 3.9973294204315085e-06, + "loss": 3.0768, + "step": 17805 + }, + { + "epoch": 1.4274919324126596, + "grad_norm": 1.6726233959197998, + "learning_rate": 3.996290973551475e-06, + "loss": 3.1414, + "step": 17806 + }, + { + "epoch": 1.4275721071936822, + "grad_norm": 0.660466194152832, + "learning_rate": 3.995252627891857e-06, + "loss": 2.4572, + "step": 17807 + }, + { + "epoch": 1.4276522819747048, + "grad_norm": 1.7756283283233643, + "learning_rate": 3.994214383470165e-06, + "loss": 3.1049, + "step": 17808 + }, + { + "epoch": 1.4277324567557275, + "grad_norm": 1.7575912475585938, + "learning_rate": 3.993176240303902e-06, + "loss": 3.1453, + "step": 17809 + }, + { + "epoch": 1.42781263153675, + "grad_norm": 1.6842364072799683, + "learning_rate": 3.992138198410575e-06, + "loss": 3.0804, + "step": 17810 + }, + { + "epoch": 1.4278928063177727, + "grad_norm": 1.8360193967819214, + "learning_rate": 3.991100257807675e-06, + "loss": 3.1578, + "step": 17811 + }, + { + "epoch": 1.4279729810987953, + "grad_norm": 1.7186789512634277, + "learning_rate": 3.990062418512708e-06, + "loss": 3.1144, + "step": 17812 + }, + { + "epoch": 1.428053155879818, + "grad_norm": 0.6791062355041504, + "learning_rate": 3.98902468054317e-06, + "loss": 2.4436, + "step": 17813 + }, + { + "epoch": 1.4281333306608406, + "grad_norm": 1.7723814249038696, + "learning_rate": 3.987987043916558e-06, + "loss": 3.2618, + "step": 17814 + }, + { + "epoch": 1.4282135054418632, + "grad_norm": 1.8175897598266602, + "learning_rate": 3.986949508650368e-06, + "loss": 3.2426, + "step": 17815 + }, + { + "epoch": 1.4282936802228858, + "grad_norm": 1.8297940492630005, + "learning_rate": 3.985912074762088e-06, + "loss": 3.0786, + "step": 17816 + }, + { + "epoch": 1.4283738550039085, + "grad_norm": 1.7441086769104004, + "learning_rate": 3.984874742269211e-06, + "loss": 3.1942, + "step": 17817 + }, + { + "epoch": 1.428454029784931, + "grad_norm": 1.7917646169662476, + "learning_rate": 3.983837511189225e-06, + "loss": 3.1781, + "step": 17818 + }, + { + "epoch": 1.4285342045659537, + "grad_norm": 1.9440253973007202, + "learning_rate": 3.982800381539619e-06, + "loss": 3.2861, + "step": 17819 + }, + { + "epoch": 1.4286143793469763, + "grad_norm": 1.792297601699829, + "learning_rate": 3.981763353337881e-06, + "loss": 3.2515, + "step": 17820 + }, + { + "epoch": 1.4286945541279992, + "grad_norm": 1.7910408973693848, + "learning_rate": 3.980726426601488e-06, + "loss": 3.1759, + "step": 17821 + }, + { + "epoch": 1.4287747289090218, + "grad_norm": 1.8467389345169067, + "learning_rate": 3.979689601347926e-06, + "loss": 3.2627, + "step": 17822 + }, + { + "epoch": 1.4288549036900444, + "grad_norm": 1.778958797454834, + "learning_rate": 3.978652877594675e-06, + "loss": 3.1199, + "step": 17823 + }, + { + "epoch": 1.428935078471067, + "grad_norm": 1.731236219406128, + "learning_rate": 3.977616255359216e-06, + "loss": 3.1431, + "step": 17824 + }, + { + "epoch": 1.4290152532520897, + "grad_norm": 1.8181636333465576, + "learning_rate": 3.976579734659028e-06, + "loss": 3.0948, + "step": 17825 + }, + { + "epoch": 1.4290954280331123, + "grad_norm": 1.765636920928955, + "learning_rate": 3.975543315511577e-06, + "loss": 3.1523, + "step": 17826 + }, + { + "epoch": 1.429175602814135, + "grad_norm": 1.768401026725769, + "learning_rate": 3.974506997934344e-06, + "loss": 3.1614, + "step": 17827 + }, + { + "epoch": 1.4292557775951575, + "grad_norm": 1.7792271375656128, + "learning_rate": 3.973470781944799e-06, + "loss": 3.2941, + "step": 17828 + }, + { + "epoch": 1.4293359523761802, + "grad_norm": 1.9673467874526978, + "learning_rate": 3.972434667560414e-06, + "loss": 3.1505, + "step": 17829 + }, + { + "epoch": 1.4294161271572028, + "grad_norm": 1.8098852634429932, + "learning_rate": 3.971398654798658e-06, + "loss": 3.1266, + "step": 17830 + }, + { + "epoch": 1.4294963019382254, + "grad_norm": 1.6943591833114624, + "learning_rate": 3.970362743676995e-06, + "loss": 3.2405, + "step": 17831 + }, + { + "epoch": 1.429576476719248, + "grad_norm": 1.7998754978179932, + "learning_rate": 3.969326934212888e-06, + "loss": 3.2013, + "step": 17832 + }, + { + "epoch": 1.4296566515002707, + "grad_norm": 1.6894762516021729, + "learning_rate": 3.968291226423805e-06, + "loss": 3.1814, + "step": 17833 + }, + { + "epoch": 1.4297368262812933, + "grad_norm": 1.895581603050232, + "learning_rate": 3.967255620327206e-06, + "loss": 3.2642, + "step": 17834 + }, + { + "epoch": 1.429817001062316, + "grad_norm": 1.8516205549240112, + "learning_rate": 3.966220115940554e-06, + "loss": 3.2825, + "step": 17835 + }, + { + "epoch": 1.4298971758433385, + "grad_norm": 1.8284584283828735, + "learning_rate": 3.965184713281302e-06, + "loss": 3.1908, + "step": 17836 + }, + { + "epoch": 1.4299773506243612, + "grad_norm": 1.8041253089904785, + "learning_rate": 3.9641494123669075e-06, + "loss": 3.2196, + "step": 17837 + }, + { + "epoch": 1.4300575254053838, + "grad_norm": 1.7789539098739624, + "learning_rate": 3.963114213214827e-06, + "loss": 3.2108, + "step": 17838 + }, + { + "epoch": 1.4301377001864064, + "grad_norm": 0.7012564539909363, + "learning_rate": 3.962079115842514e-06, + "loss": 2.4808, + "step": 17839 + }, + { + "epoch": 1.430217874967429, + "grad_norm": 1.8424663543701172, + "learning_rate": 3.9610441202674225e-06, + "loss": 3.1916, + "step": 17840 + }, + { + "epoch": 1.4302980497484516, + "grad_norm": 1.7257320880889893, + "learning_rate": 3.9600092265069945e-06, + "loss": 3.1906, + "step": 17841 + }, + { + "epoch": 1.4303782245294743, + "grad_norm": 0.6816081404685974, + "learning_rate": 3.958974434578683e-06, + "loss": 2.4639, + "step": 17842 + }, + { + "epoch": 1.430458399310497, + "grad_norm": 1.7669025659561157, + "learning_rate": 3.957939744499933e-06, + "loss": 3.1742, + "step": 17843 + }, + { + "epoch": 1.4305385740915195, + "grad_norm": 2.019000768661499, + "learning_rate": 3.95690515628819e-06, + "loss": 3.1094, + "step": 17844 + }, + { + "epoch": 1.4306187488725421, + "grad_norm": 1.8292726278305054, + "learning_rate": 3.9558706699608995e-06, + "loss": 3.121, + "step": 17845 + }, + { + "epoch": 1.4306989236535648, + "grad_norm": 1.814527153968811, + "learning_rate": 3.954836285535496e-06, + "loss": 3.1246, + "step": 17846 + }, + { + "epoch": 1.4307790984345874, + "grad_norm": 1.8421123027801514, + "learning_rate": 3.9538020030294224e-06, + "loss": 3.1495, + "step": 17847 + }, + { + "epoch": 1.43085927321561, + "grad_norm": 1.965468168258667, + "learning_rate": 3.952767822460115e-06, + "loss": 3.1896, + "step": 17848 + }, + { + "epoch": 1.4309394479966326, + "grad_norm": 1.7873061895370483, + "learning_rate": 3.951733743845012e-06, + "loss": 3.0994, + "step": 17849 + }, + { + "epoch": 1.4310196227776553, + "grad_norm": 1.8862940073013306, + "learning_rate": 3.9506997672015505e-06, + "loss": 3.1637, + "step": 17850 + }, + { + "epoch": 1.4310997975586779, + "grad_norm": 1.7915846109390259, + "learning_rate": 3.9496658925471545e-06, + "loss": 3.0894, + "step": 17851 + }, + { + "epoch": 1.4311799723397005, + "grad_norm": 1.8767815828323364, + "learning_rate": 3.948632119899258e-06, + "loss": 3.158, + "step": 17852 + }, + { + "epoch": 1.4312601471207231, + "grad_norm": 1.79048490524292, + "learning_rate": 3.947598449275292e-06, + "loss": 3.0692, + "step": 17853 + }, + { + "epoch": 1.4313403219017458, + "grad_norm": 1.699923038482666, + "learning_rate": 3.946564880692685e-06, + "loss": 3.087, + "step": 17854 + }, + { + "epoch": 1.4314204966827684, + "grad_norm": 1.9145119190216064, + "learning_rate": 3.945531414168862e-06, + "loss": 3.1594, + "step": 17855 + }, + { + "epoch": 1.431500671463791, + "grad_norm": 1.6669948101043701, + "learning_rate": 3.944498049721244e-06, + "loss": 3.0878, + "step": 17856 + }, + { + "epoch": 1.4315808462448136, + "grad_norm": 1.779678225517273, + "learning_rate": 3.943464787367253e-06, + "loss": 3.184, + "step": 17857 + }, + { + "epoch": 1.4316610210258363, + "grad_norm": 1.869181513786316, + "learning_rate": 3.9424316271243125e-06, + "loss": 3.0859, + "step": 17858 + }, + { + "epoch": 1.4317411958068589, + "grad_norm": 1.860169529914856, + "learning_rate": 3.9413985690098386e-06, + "loss": 3.2471, + "step": 17859 + }, + { + "epoch": 1.4318213705878815, + "grad_norm": 0.6847892999649048, + "learning_rate": 3.940365613041254e-06, + "loss": 2.459, + "step": 17860 + }, + { + "epoch": 1.4319015453689041, + "grad_norm": 1.845994234085083, + "learning_rate": 3.939332759235965e-06, + "loss": 3.1359, + "step": 17861 + }, + { + "epoch": 1.4319817201499268, + "grad_norm": 1.9114916324615479, + "learning_rate": 3.93830000761139e-06, + "loss": 3.2285, + "step": 17862 + }, + { + "epoch": 1.4320618949309494, + "grad_norm": 1.796664834022522, + "learning_rate": 3.937267358184941e-06, + "loss": 3.1176, + "step": 17863 + }, + { + "epoch": 1.432142069711972, + "grad_norm": 1.7143323421478271, + "learning_rate": 3.936234810974028e-06, + "loss": 3.2418, + "step": 17864 + }, + { + "epoch": 1.4322222444929946, + "grad_norm": 1.750130295753479, + "learning_rate": 3.935202365996062e-06, + "loss": 3.1304, + "step": 17865 + }, + { + "epoch": 1.4323024192740172, + "grad_norm": 1.7381176948547363, + "learning_rate": 3.934170023268442e-06, + "loss": 3.0884, + "step": 17866 + }, + { + "epoch": 1.4323825940550399, + "grad_norm": 1.7093054056167603, + "learning_rate": 3.933137782808579e-06, + "loss": 3.2826, + "step": 17867 + }, + { + "epoch": 1.4324627688360625, + "grad_norm": 1.7281494140625, + "learning_rate": 3.932105644633874e-06, + "loss": 3.2382, + "step": 17868 + }, + { + "epoch": 1.4325429436170851, + "grad_norm": 1.7818703651428223, + "learning_rate": 3.931073608761731e-06, + "loss": 3.1449, + "step": 17869 + }, + { + "epoch": 1.4326231183981077, + "grad_norm": 1.8858227729797363, + "learning_rate": 3.93004167520955e-06, + "loss": 3.2284, + "step": 17870 + }, + { + "epoch": 1.4327032931791304, + "grad_norm": 1.8413082361221313, + "learning_rate": 3.929009843994724e-06, + "loss": 3.2631, + "step": 17871 + }, + { + "epoch": 1.4327834679601532, + "grad_norm": 1.7371747493743896, + "learning_rate": 3.927978115134652e-06, + "loss": 3.2211, + "step": 17872 + }, + { + "epoch": 1.4328636427411758, + "grad_norm": 1.7114267349243164, + "learning_rate": 3.9269464886467305e-06, + "loss": 3.1375, + "step": 17873 + }, + { + "epoch": 1.4329438175221985, + "grad_norm": 1.7093515396118164, + "learning_rate": 3.925914964548348e-06, + "loss": 3.2969, + "step": 17874 + }, + { + "epoch": 1.433023992303221, + "grad_norm": 0.659419596195221, + "learning_rate": 3.924883542856904e-06, + "loss": 2.464, + "step": 17875 + }, + { + "epoch": 1.4331041670842437, + "grad_norm": 1.7853394746780396, + "learning_rate": 3.923852223589779e-06, + "loss": 3.1621, + "step": 17876 + }, + { + "epoch": 1.4331843418652663, + "grad_norm": 1.7839934825897217, + "learning_rate": 3.9228210067643635e-06, + "loss": 3.1112, + "step": 17877 + }, + { + "epoch": 1.433264516646289, + "grad_norm": 1.733269453048706, + "learning_rate": 3.921789892398045e-06, + "loss": 3.2682, + "step": 17878 + }, + { + "epoch": 1.4333446914273116, + "grad_norm": 1.8009872436523438, + "learning_rate": 3.920758880508205e-06, + "loss": 3.1042, + "step": 17879 + }, + { + "epoch": 1.4334248662083342, + "grad_norm": 1.8544121980667114, + "learning_rate": 3.9197279711122326e-06, + "loss": 3.2596, + "step": 17880 + }, + { + "epoch": 1.4335050409893568, + "grad_norm": 1.7474852800369263, + "learning_rate": 3.918697164227499e-06, + "loss": 3.1436, + "step": 17881 + }, + { + "epoch": 1.4335852157703795, + "grad_norm": 1.7670904397964478, + "learning_rate": 3.9176664598713896e-06, + "loss": 3.2317, + "step": 17882 + }, + { + "epoch": 1.433665390551402, + "grad_norm": 1.8544107675552368, + "learning_rate": 3.91663585806128e-06, + "loss": 3.2723, + "step": 17883 + }, + { + "epoch": 1.4337455653324247, + "grad_norm": 1.846599817276001, + "learning_rate": 3.915605358814544e-06, + "loss": 3.2817, + "step": 17884 + }, + { + "epoch": 1.4338257401134473, + "grad_norm": 1.8469454050064087, + "learning_rate": 3.914574962148562e-06, + "loss": 3.39, + "step": 17885 + }, + { + "epoch": 1.43390591489447, + "grad_norm": 1.752487301826477, + "learning_rate": 3.913544668080697e-06, + "loss": 3.1234, + "step": 17886 + }, + { + "epoch": 1.4339860896754926, + "grad_norm": 1.6938834190368652, + "learning_rate": 3.9125144766283255e-06, + "loss": 3.1061, + "step": 17887 + }, + { + "epoch": 1.4340662644565152, + "grad_norm": 1.873250961303711, + "learning_rate": 3.911484387808814e-06, + "loss": 3.1246, + "step": 17888 + }, + { + "epoch": 1.4341464392375378, + "grad_norm": 1.7563191652297974, + "learning_rate": 3.910454401639529e-06, + "loss": 3.243, + "step": 17889 + }, + { + "epoch": 1.4342266140185604, + "grad_norm": 1.6856093406677246, + "learning_rate": 3.909424518137841e-06, + "loss": 3.2383, + "step": 17890 + }, + { + "epoch": 1.434306788799583, + "grad_norm": 1.7670481204986572, + "learning_rate": 3.908394737321105e-06, + "loss": 3.1685, + "step": 17891 + }, + { + "epoch": 1.4343869635806057, + "grad_norm": 1.69923734664917, + "learning_rate": 3.907365059206687e-06, + "loss": 3.1753, + "step": 17892 + }, + { + "epoch": 1.4344671383616283, + "grad_norm": 1.7427996397018433, + "learning_rate": 3.906335483811947e-06, + "loss": 3.1686, + "step": 17893 + }, + { + "epoch": 1.434547313142651, + "grad_norm": 1.7591904401779175, + "learning_rate": 3.905306011154243e-06, + "loss": 3.1424, + "step": 17894 + }, + { + "epoch": 1.4346274879236736, + "grad_norm": 1.8187886476516724, + "learning_rate": 3.904276641250935e-06, + "loss": 3.1619, + "step": 17895 + }, + { + "epoch": 1.4347076627046962, + "grad_norm": 1.8053264617919922, + "learning_rate": 3.903247374119371e-06, + "loss": 3.1301, + "step": 17896 + }, + { + "epoch": 1.4347878374857188, + "grad_norm": 1.8270875215530396, + "learning_rate": 3.902218209776908e-06, + "loss": 3.2261, + "step": 17897 + }, + { + "epoch": 1.4348680122667414, + "grad_norm": 1.8122272491455078, + "learning_rate": 3.901189148240897e-06, + "loss": 3.0986, + "step": 17898 + }, + { + "epoch": 1.434948187047764, + "grad_norm": 1.7418389320373535, + "learning_rate": 3.9001601895286865e-06, + "loss": 3.1506, + "step": 17899 + }, + { + "epoch": 1.4350283618287867, + "grad_norm": 1.822834849357605, + "learning_rate": 3.89913133365763e-06, + "loss": 3.1158, + "step": 17900 + }, + { + "epoch": 1.4351085366098093, + "grad_norm": 1.7586944103240967, + "learning_rate": 3.898102580645066e-06, + "loss": 3.1467, + "step": 17901 + }, + { + "epoch": 1.4351887113908322, + "grad_norm": 1.787416934967041, + "learning_rate": 3.8970739305083405e-06, + "loss": 3.1689, + "step": 17902 + }, + { + "epoch": 1.4352688861718548, + "grad_norm": 1.6943789720535278, + "learning_rate": 3.896045383264798e-06, + "loss": 3.2265, + "step": 17903 + }, + { + "epoch": 1.4353490609528774, + "grad_norm": 1.7429920434951782, + "learning_rate": 3.89501693893178e-06, + "loss": 3.1665, + "step": 17904 + }, + { + "epoch": 1.4354292357339, + "grad_norm": 1.7445636987686157, + "learning_rate": 3.893988597526627e-06, + "loss": 3.1313, + "step": 17905 + }, + { + "epoch": 1.4355094105149226, + "grad_norm": 1.778905987739563, + "learning_rate": 3.892960359066672e-06, + "loss": 3.1537, + "step": 17906 + }, + { + "epoch": 1.4355895852959453, + "grad_norm": 1.782265067100525, + "learning_rate": 3.891932223569252e-06, + "loss": 3.2072, + "step": 17907 + }, + { + "epoch": 1.435669760076968, + "grad_norm": 1.7971941232681274, + "learning_rate": 3.890904191051703e-06, + "loss": 3.0666, + "step": 17908 + }, + { + "epoch": 1.4357499348579905, + "grad_norm": 1.7184187173843384, + "learning_rate": 3.889876261531356e-06, + "loss": 3.1329, + "step": 17909 + }, + { + "epoch": 1.4358301096390131, + "grad_norm": 1.8485485315322876, + "learning_rate": 3.888848435025545e-06, + "loss": 3.0888, + "step": 17910 + }, + { + "epoch": 1.4359102844200358, + "grad_norm": 1.7860714197158813, + "learning_rate": 3.887820711551592e-06, + "loss": 3.2155, + "step": 17911 + }, + { + "epoch": 1.4359904592010584, + "grad_norm": 1.7722634077072144, + "learning_rate": 3.886793091126828e-06, + "loss": 3.1203, + "step": 17912 + }, + { + "epoch": 1.436070633982081, + "grad_norm": 1.7714767456054688, + "learning_rate": 3.885765573768577e-06, + "loss": 3.0758, + "step": 17913 + }, + { + "epoch": 1.4361508087631036, + "grad_norm": 1.751166820526123, + "learning_rate": 3.884738159494164e-06, + "loss": 3.1691, + "step": 17914 + }, + { + "epoch": 1.4362309835441263, + "grad_norm": 1.717484951019287, + "learning_rate": 3.8837108483209155e-06, + "loss": 3.0853, + "step": 17915 + }, + { + "epoch": 1.4363111583251489, + "grad_norm": 1.7376281023025513, + "learning_rate": 3.882683640266142e-06, + "loss": 3.2147, + "step": 17916 + }, + { + "epoch": 1.4363913331061715, + "grad_norm": 1.7628175020217896, + "learning_rate": 3.881656535347166e-06, + "loss": 3.2136, + "step": 17917 + }, + { + "epoch": 1.4364715078871941, + "grad_norm": 1.8553117513656616, + "learning_rate": 3.880629533581305e-06, + "loss": 3.1447, + "step": 17918 + }, + { + "epoch": 1.4365516826682168, + "grad_norm": 1.7978332042694092, + "learning_rate": 3.879602634985873e-06, + "loss": 3.1605, + "step": 17919 + }, + { + "epoch": 1.4366318574492394, + "grad_norm": 1.7321354150772095, + "learning_rate": 3.878575839578187e-06, + "loss": 3.1484, + "step": 17920 + }, + { + "epoch": 1.436712032230262, + "grad_norm": 1.6528009176254272, + "learning_rate": 3.877549147375551e-06, + "loss": 3.0629, + "step": 17921 + }, + { + "epoch": 1.4367922070112846, + "grad_norm": 1.8672478199005127, + "learning_rate": 3.876522558395279e-06, + "loss": 3.209, + "step": 17922 + }, + { + "epoch": 1.4368723817923073, + "grad_norm": 1.8076726198196411, + "learning_rate": 3.87549607265468e-06, + "loss": 3.2069, + "step": 17923 + }, + { + "epoch": 1.4369525565733299, + "grad_norm": 1.8182976245880127, + "learning_rate": 3.8744696901710565e-06, + "loss": 3.2349, + "step": 17924 + }, + { + "epoch": 1.4370327313543525, + "grad_norm": 1.796269178390503, + "learning_rate": 3.87344341096172e-06, + "loss": 3.1729, + "step": 17925 + }, + { + "epoch": 1.4371129061353751, + "grad_norm": 1.8292630910873413, + "learning_rate": 3.872417235043965e-06, + "loss": 3.0461, + "step": 17926 + }, + { + "epoch": 1.4371930809163977, + "grad_norm": 1.8739105463027954, + "learning_rate": 3.871391162435095e-06, + "loss": 3.2728, + "step": 17927 + }, + { + "epoch": 1.4372732556974204, + "grad_norm": 1.7613935470581055, + "learning_rate": 3.87036519315241e-06, + "loss": 3.2327, + "step": 17928 + }, + { + "epoch": 1.437353430478443, + "grad_norm": 1.818748950958252, + "learning_rate": 3.869339327213209e-06, + "loss": 3.3129, + "step": 17929 + }, + { + "epoch": 1.4374336052594656, + "grad_norm": 1.8501570224761963, + "learning_rate": 3.868313564634788e-06, + "loss": 3.1921, + "step": 17930 + }, + { + "epoch": 1.4375137800404882, + "grad_norm": 1.7387731075286865, + "learning_rate": 3.867287905434437e-06, + "loss": 3.1974, + "step": 17931 + }, + { + "epoch": 1.4375939548215109, + "grad_norm": 1.8474581241607666, + "learning_rate": 3.866262349629451e-06, + "loss": 3.1658, + "step": 17932 + }, + { + "epoch": 1.4376741296025335, + "grad_norm": 1.8455836772918701, + "learning_rate": 3.865236897237118e-06, + "loss": 3.2133, + "step": 17933 + }, + { + "epoch": 1.4377543043835561, + "grad_norm": 1.6457151174545288, + "learning_rate": 3.86421154827473e-06, + "loss": 3.1791, + "step": 17934 + }, + { + "epoch": 1.4378344791645787, + "grad_norm": 1.6740612983703613, + "learning_rate": 3.863186302759576e-06, + "loss": 3.199, + "step": 17935 + }, + { + "epoch": 1.4379146539456014, + "grad_norm": 0.7040858268737793, + "learning_rate": 3.862161160708935e-06, + "loss": 2.4647, + "step": 17936 + }, + { + "epoch": 1.437994828726624, + "grad_norm": 1.8580739498138428, + "learning_rate": 3.861136122140093e-06, + "loss": 3.2858, + "step": 17937 + }, + { + "epoch": 1.4380750035076466, + "grad_norm": 1.802791714668274, + "learning_rate": 3.860111187070333e-06, + "loss": 3.0882, + "step": 17938 + }, + { + "epoch": 1.4381551782886692, + "grad_norm": 1.8111742734909058, + "learning_rate": 3.8590863555169355e-06, + "loss": 3.2179, + "step": 17939 + }, + { + "epoch": 1.4382353530696919, + "grad_norm": 1.7681232690811157, + "learning_rate": 3.858061627497181e-06, + "loss": 3.0905, + "step": 17940 + }, + { + "epoch": 1.4383155278507145, + "grad_norm": 1.7534058094024658, + "learning_rate": 3.857037003028338e-06, + "loss": 3.0599, + "step": 17941 + }, + { + "epoch": 1.438395702631737, + "grad_norm": 1.7273585796356201, + "learning_rate": 3.856012482127687e-06, + "loss": 3.1808, + "step": 17942 + }, + { + "epoch": 1.4384758774127597, + "grad_norm": 1.8193795680999756, + "learning_rate": 3.8549880648125e-06, + "loss": 3.1793, + "step": 17943 + }, + { + "epoch": 1.4385560521937824, + "grad_norm": 1.7603495121002197, + "learning_rate": 3.853963751100048e-06, + "loss": 3.2391, + "step": 17944 + }, + { + "epoch": 1.438636226974805, + "grad_norm": 1.7988325357437134, + "learning_rate": 3.852939541007606e-06, + "loss": 3.1626, + "step": 17945 + }, + { + "epoch": 1.4387164017558276, + "grad_norm": 1.7527424097061157, + "learning_rate": 3.851915434552431e-06, + "loss": 3.199, + "step": 17946 + }, + { + "epoch": 1.4387965765368502, + "grad_norm": 1.7988927364349365, + "learning_rate": 3.850891431751797e-06, + "loss": 3.1942, + "step": 17947 + }, + { + "epoch": 1.4388767513178728, + "grad_norm": 1.7004464864730835, + "learning_rate": 3.849867532622965e-06, + "loss": 3.1562, + "step": 17948 + }, + { + "epoch": 1.4389569260988955, + "grad_norm": 1.8640531301498413, + "learning_rate": 3.8488437371832e-06, + "loss": 3.147, + "step": 17949 + }, + { + "epoch": 1.439037100879918, + "grad_norm": 1.7627196311950684, + "learning_rate": 3.847820045449765e-06, + "loss": 3.1335, + "step": 17950 + }, + { + "epoch": 1.4391172756609407, + "grad_norm": 1.7640668153762817, + "learning_rate": 3.846796457439912e-06, + "loss": 3.1368, + "step": 17951 + }, + { + "epoch": 1.4391974504419633, + "grad_norm": 1.670106291770935, + "learning_rate": 3.845772973170902e-06, + "loss": 3.1472, + "step": 17952 + }, + { + "epoch": 1.4392776252229862, + "grad_norm": 1.7385046482086182, + "learning_rate": 3.844749592659991e-06, + "loss": 3.1461, + "step": 17953 + }, + { + "epoch": 1.4393578000040088, + "grad_norm": 1.7684584856033325, + "learning_rate": 3.843726315924433e-06, + "loss": 3.1154, + "step": 17954 + }, + { + "epoch": 1.4394379747850314, + "grad_norm": 0.6655102968215942, + "learning_rate": 3.842703142981482e-06, + "loss": 2.481, + "step": 17955 + }, + { + "epoch": 1.439518149566054, + "grad_norm": 1.6688525676727295, + "learning_rate": 3.841680073848383e-06, + "loss": 3.2123, + "step": 17956 + }, + { + "epoch": 1.4395983243470767, + "grad_norm": 1.8426356315612793, + "learning_rate": 3.840657108542388e-06, + "loss": 3.1931, + "step": 17957 + }, + { + "epoch": 1.4396784991280993, + "grad_norm": 1.7719011306762695, + "learning_rate": 3.839634247080743e-06, + "loss": 3.1429, + "step": 17958 + }, + { + "epoch": 1.439758673909122, + "grad_norm": 1.7860407829284668, + "learning_rate": 3.838611489480693e-06, + "loss": 3.225, + "step": 17959 + }, + { + "epoch": 1.4398388486901446, + "grad_norm": 1.955544352531433, + "learning_rate": 3.837588835759486e-06, + "loss": 3.1896, + "step": 17960 + }, + { + "epoch": 1.4399190234711672, + "grad_norm": 1.850981593132019, + "learning_rate": 3.8365662859343555e-06, + "loss": 3.1421, + "step": 17961 + }, + { + "epoch": 1.4399991982521898, + "grad_norm": 1.7191030979156494, + "learning_rate": 3.835543840022546e-06, + "loss": 3.1033, + "step": 17962 + }, + { + "epoch": 1.4400793730332124, + "grad_norm": 1.7415375709533691, + "learning_rate": 3.834521498041294e-06, + "loss": 3.1068, + "step": 17963 + }, + { + "epoch": 1.440159547814235, + "grad_norm": 1.8296821117401123, + "learning_rate": 3.833499260007837e-06, + "loss": 3.1108, + "step": 17964 + }, + { + "epoch": 1.4402397225952577, + "grad_norm": 1.7237874269485474, + "learning_rate": 3.832477125939412e-06, + "loss": 3.079, + "step": 17965 + }, + { + "epoch": 1.4403198973762803, + "grad_norm": 1.77456533908844, + "learning_rate": 3.831455095853246e-06, + "loss": 3.1511, + "step": 17966 + }, + { + "epoch": 1.440400072157303, + "grad_norm": 1.8756630420684814, + "learning_rate": 3.8304331697665734e-06, + "loss": 3.2183, + "step": 17967 + }, + { + "epoch": 1.4404802469383255, + "grad_norm": 1.8365122079849243, + "learning_rate": 3.8294113476966235e-06, + "loss": 3.2046, + "step": 17968 + }, + { + "epoch": 1.4405604217193482, + "grad_norm": 1.658017873764038, + "learning_rate": 3.828389629660623e-06, + "loss": 3.2019, + "step": 17969 + }, + { + "epoch": 1.4406405965003708, + "grad_norm": 1.850067377090454, + "learning_rate": 3.827368015675801e-06, + "loss": 3.1973, + "step": 17970 + }, + { + "epoch": 1.4407207712813934, + "grad_norm": 1.80731999874115, + "learning_rate": 3.826346505759377e-06, + "loss": 3.2367, + "step": 17971 + }, + { + "epoch": 1.440800946062416, + "grad_norm": 1.7763137817382812, + "learning_rate": 3.825325099928573e-06, + "loss": 3.15, + "step": 17972 + }, + { + "epoch": 1.4408811208434387, + "grad_norm": 1.7579455375671387, + "learning_rate": 3.8243037982006125e-06, + "loss": 3.1759, + "step": 17973 + }, + { + "epoch": 1.4409612956244613, + "grad_norm": 1.7852838039398193, + "learning_rate": 3.8232826005927146e-06, + "loss": 3.3451, + "step": 17974 + }, + { + "epoch": 1.441041470405484, + "grad_norm": 1.955701470375061, + "learning_rate": 3.822261507122097e-06, + "loss": 3.1721, + "step": 17975 + }, + { + "epoch": 1.4411216451865065, + "grad_norm": 1.7887070178985596, + "learning_rate": 3.821240517805971e-06, + "loss": 3.3163, + "step": 17976 + }, + { + "epoch": 1.4412018199675292, + "grad_norm": 1.757728934288025, + "learning_rate": 3.820219632661552e-06, + "loss": 3.2629, + "step": 17977 + }, + { + "epoch": 1.4412819947485518, + "grad_norm": 1.7196897268295288, + "learning_rate": 3.819198851706051e-06, + "loss": 3.1888, + "step": 17978 + }, + { + "epoch": 1.4413621695295744, + "grad_norm": 1.7984445095062256, + "learning_rate": 3.8181781749566805e-06, + "loss": 3.2102, + "step": 17979 + }, + { + "epoch": 1.441442344310597, + "grad_norm": 1.8476614952087402, + "learning_rate": 3.817157602430649e-06, + "loss": 3.1726, + "step": 17980 + }, + { + "epoch": 1.4415225190916197, + "grad_norm": 1.832411527633667, + "learning_rate": 3.8161371341451594e-06, + "loss": 3.1622, + "step": 17981 + }, + { + "epoch": 1.4416026938726423, + "grad_norm": 1.6887602806091309, + "learning_rate": 3.815116770117417e-06, + "loss": 3.1026, + "step": 17982 + }, + { + "epoch": 1.4416828686536651, + "grad_norm": 1.8459396362304688, + "learning_rate": 3.8140965103646277e-06, + "loss": 2.9937, + "step": 17983 + }, + { + "epoch": 1.4417630434346878, + "grad_norm": 1.8229162693023682, + "learning_rate": 3.8130763549039895e-06, + "loss": 3.144, + "step": 17984 + }, + { + "epoch": 1.4418432182157104, + "grad_norm": 1.848362684249878, + "learning_rate": 3.812056303752708e-06, + "loss": 3.266, + "step": 17985 + }, + { + "epoch": 1.441923392996733, + "grad_norm": 1.8025281429290771, + "learning_rate": 3.811036356927973e-06, + "loss": 3.1987, + "step": 17986 + }, + { + "epoch": 1.4420035677777556, + "grad_norm": 1.7976661920547485, + "learning_rate": 3.8100165144469826e-06, + "loss": 3.2771, + "step": 17987 + }, + { + "epoch": 1.4420837425587782, + "grad_norm": 1.7458536624908447, + "learning_rate": 3.808996776326933e-06, + "loss": 3.1619, + "step": 17988 + }, + { + "epoch": 1.4421639173398009, + "grad_norm": 1.7395380735397339, + "learning_rate": 3.8079771425850165e-06, + "loss": 3.1871, + "step": 17989 + }, + { + "epoch": 1.4422440921208235, + "grad_norm": 1.7246125936508179, + "learning_rate": 3.806957613238426e-06, + "loss": 3.2222, + "step": 17990 + }, + { + "epoch": 1.4423242669018461, + "grad_norm": 1.7924913167953491, + "learning_rate": 3.8059381883043435e-06, + "loss": 3.2395, + "step": 17991 + }, + { + "epoch": 1.4424044416828687, + "grad_norm": 1.848413109779358, + "learning_rate": 3.8049188677999605e-06, + "loss": 3.2284, + "step": 17992 + }, + { + "epoch": 1.4424846164638914, + "grad_norm": 1.8374418020248413, + "learning_rate": 3.803899651742462e-06, + "loss": 3.1085, + "step": 17993 + }, + { + "epoch": 1.442564791244914, + "grad_norm": 1.7631250619888306, + "learning_rate": 3.8028805401490322e-06, + "loss": 3.131, + "step": 17994 + }, + { + "epoch": 1.4426449660259366, + "grad_norm": 1.7534306049346924, + "learning_rate": 3.8018615330368545e-06, + "loss": 3.2166, + "step": 17995 + }, + { + "epoch": 1.4427251408069592, + "grad_norm": 0.6974304914474487, + "learning_rate": 3.800842630423105e-06, + "loss": 2.4752, + "step": 17996 + }, + { + "epoch": 1.4428053155879819, + "grad_norm": 1.7519681453704834, + "learning_rate": 3.799823832324964e-06, + "loss": 3.3114, + "step": 17997 + }, + { + "epoch": 1.4428854903690045, + "grad_norm": 1.951180100440979, + "learning_rate": 3.798805138759607e-06, + "loss": 3.1772, + "step": 17998 + }, + { + "epoch": 1.442965665150027, + "grad_norm": 0.6728654503822327, + "learning_rate": 3.79778654974421e-06, + "loss": 2.4583, + "step": 17999 + }, + { + "epoch": 1.4430458399310497, + "grad_norm": 1.8149957656860352, + "learning_rate": 3.7967680652959494e-06, + "loss": 3.1683, + "step": 18000 + }, + { + "epoch": 1.4431260147120724, + "grad_norm": 1.8316069841384888, + "learning_rate": 3.79574968543199e-06, + "loss": 3.1752, + "step": 18001 + }, + { + "epoch": 1.443206189493095, + "grad_norm": 1.8260259628295898, + "learning_rate": 3.7947314101695045e-06, + "loss": 3.2657, + "step": 18002 + }, + { + "epoch": 1.4432863642741176, + "grad_norm": 1.7979607582092285, + "learning_rate": 3.7937132395256594e-06, + "loss": 3.2149, + "step": 18003 + }, + { + "epoch": 1.4433665390551402, + "grad_norm": 1.7312835454940796, + "learning_rate": 3.792695173517622e-06, + "loss": 3.1285, + "step": 18004 + }, + { + "epoch": 1.4434467138361629, + "grad_norm": 1.7536063194274902, + "learning_rate": 3.79167721216256e-06, + "loss": 3.2495, + "step": 18005 + }, + { + "epoch": 1.4435268886171855, + "grad_norm": 1.766109585762024, + "learning_rate": 3.7906593554776296e-06, + "loss": 3.2086, + "step": 18006 + }, + { + "epoch": 1.443607063398208, + "grad_norm": 1.9909613132476807, + "learning_rate": 3.789641603479993e-06, + "loss": 3.1809, + "step": 18007 + }, + { + "epoch": 1.4436872381792307, + "grad_norm": 1.7701407670974731, + "learning_rate": 3.788623956186811e-06, + "loss": 3.1006, + "step": 18008 + }, + { + "epoch": 1.4437674129602533, + "grad_norm": 1.855155110359192, + "learning_rate": 3.7876064136152392e-06, + "loss": 3.2429, + "step": 18009 + }, + { + "epoch": 1.443847587741276, + "grad_norm": 1.7492836713790894, + "learning_rate": 3.786588975782438e-06, + "loss": 3.1879, + "step": 18010 + }, + { + "epoch": 1.4439277625222986, + "grad_norm": 1.755600094795227, + "learning_rate": 3.7855716427055522e-06, + "loss": 3.1833, + "step": 18011 + }, + { + "epoch": 1.4440079373033212, + "grad_norm": 1.7741085290908813, + "learning_rate": 3.78455441440174e-06, + "loss": 3.3031, + "step": 18012 + }, + { + "epoch": 1.4440881120843438, + "grad_norm": 0.6574879884719849, + "learning_rate": 3.7835372908881486e-06, + "loss": 2.4545, + "step": 18013 + }, + { + "epoch": 1.4441682868653665, + "grad_norm": 1.717745304107666, + "learning_rate": 3.7825202721819275e-06, + "loss": 3.1071, + "step": 18014 + }, + { + "epoch": 1.444248461646389, + "grad_norm": 2.357496500015259, + "learning_rate": 3.781503358300228e-06, + "loss": 3.1879, + "step": 18015 + }, + { + "epoch": 1.4443286364274117, + "grad_norm": 0.6665770411491394, + "learning_rate": 3.780486549260186e-06, + "loss": 2.4576, + "step": 18016 + }, + { + "epoch": 1.4444088112084343, + "grad_norm": 1.8149415254592896, + "learning_rate": 3.779469845078949e-06, + "loss": 3.1215, + "step": 18017 + }, + { + "epoch": 1.444488985989457, + "grad_norm": 1.7365869283676147, + "learning_rate": 3.778453245773658e-06, + "loss": 3.0726, + "step": 18018 + }, + { + "epoch": 1.4445691607704796, + "grad_norm": 1.687648892402649, + "learning_rate": 3.777436751361452e-06, + "loss": 3.1174, + "step": 18019 + }, + { + "epoch": 1.4446493355515022, + "grad_norm": 1.773361325263977, + "learning_rate": 3.776420361859474e-06, + "loss": 3.1907, + "step": 18020 + }, + { + "epoch": 1.4447295103325248, + "grad_norm": 1.7434090375900269, + "learning_rate": 3.775404077284851e-06, + "loss": 3.192, + "step": 18021 + }, + { + "epoch": 1.4448096851135475, + "grad_norm": 1.7522318363189697, + "learning_rate": 3.774387897654721e-06, + "loss": 3.0705, + "step": 18022 + }, + { + "epoch": 1.44488985989457, + "grad_norm": 1.842658519744873, + "learning_rate": 3.773371822986218e-06, + "loss": 3.1076, + "step": 18023 + }, + { + "epoch": 1.4449700346755927, + "grad_norm": 1.7645888328552246, + "learning_rate": 3.772355853296471e-06, + "loss": 3.2273, + "step": 18024 + }, + { + "epoch": 1.4450502094566153, + "grad_norm": 1.7658205032348633, + "learning_rate": 3.7713399886026135e-06, + "loss": 3.108, + "step": 18025 + }, + { + "epoch": 1.445130384237638, + "grad_norm": 1.8105801343917847, + "learning_rate": 3.7703242289217645e-06, + "loss": 3.2035, + "step": 18026 + }, + { + "epoch": 1.4452105590186606, + "grad_norm": 1.864510178565979, + "learning_rate": 3.7693085742710544e-06, + "loss": 3.2455, + "step": 18027 + }, + { + "epoch": 1.4452907337996832, + "grad_norm": 1.8592778444290161, + "learning_rate": 3.768293024667605e-06, + "loss": 3.2158, + "step": 18028 + }, + { + "epoch": 1.4453709085807058, + "grad_norm": 1.8196828365325928, + "learning_rate": 3.767277580128539e-06, + "loss": 3.232, + "step": 18029 + }, + { + "epoch": 1.4454510833617285, + "grad_norm": 1.7630870342254639, + "learning_rate": 3.76626224067098e-06, + "loss": 3.18, + "step": 18030 + }, + { + "epoch": 1.445531258142751, + "grad_norm": 1.8214895725250244, + "learning_rate": 3.7652470063120393e-06, + "loss": 3.2202, + "step": 18031 + }, + { + "epoch": 1.4456114329237737, + "grad_norm": 1.881697416305542, + "learning_rate": 3.7642318770688356e-06, + "loss": 3.3033, + "step": 18032 + }, + { + "epoch": 1.4456916077047963, + "grad_norm": 1.6753438711166382, + "learning_rate": 3.7632168529584868e-06, + "loss": 3.1582, + "step": 18033 + }, + { + "epoch": 1.4457717824858192, + "grad_norm": 1.7912721633911133, + "learning_rate": 3.762201933998102e-06, + "loss": 3.1776, + "step": 18034 + }, + { + "epoch": 1.4458519572668418, + "grad_norm": 1.7997115850448608, + "learning_rate": 3.761187120204799e-06, + "loss": 3.1425, + "step": 18035 + }, + { + "epoch": 1.4459321320478644, + "grad_norm": 1.9187229871749878, + "learning_rate": 3.7601724115956785e-06, + "loss": 3.236, + "step": 18036 + }, + { + "epoch": 1.446012306828887, + "grad_norm": 1.8536224365234375, + "learning_rate": 3.759157808187851e-06, + "loss": 3.1956, + "step": 18037 + }, + { + "epoch": 1.4460924816099097, + "grad_norm": 1.8738837242126465, + "learning_rate": 3.7581433099984244e-06, + "loss": 3.2073, + "step": 18038 + }, + { + "epoch": 1.4461726563909323, + "grad_norm": 1.8134336471557617, + "learning_rate": 3.757128917044501e-06, + "loss": 3.162, + "step": 18039 + }, + { + "epoch": 1.446252831171955, + "grad_norm": 0.6630106568336487, + "learning_rate": 3.7561146293431885e-06, + "loss": 2.4934, + "step": 18040 + }, + { + "epoch": 1.4463330059529775, + "grad_norm": 1.7867285013198853, + "learning_rate": 3.755100446911578e-06, + "loss": 3.2345, + "step": 18041 + }, + { + "epoch": 1.4464131807340002, + "grad_norm": 0.6672760844230652, + "learning_rate": 3.754086369766774e-06, + "loss": 2.4788, + "step": 18042 + }, + { + "epoch": 1.4464933555150228, + "grad_norm": 0.671281635761261, + "learning_rate": 3.753072397925872e-06, + "loss": 2.468, + "step": 18043 + }, + { + "epoch": 1.4465735302960454, + "grad_norm": 1.7348220348358154, + "learning_rate": 3.7520585314059686e-06, + "loss": 3.1812, + "step": 18044 + }, + { + "epoch": 1.446653705077068, + "grad_norm": 1.8504951000213623, + "learning_rate": 3.7510447702241594e-06, + "loss": 3.0855, + "step": 18045 + }, + { + "epoch": 1.4467338798580907, + "grad_norm": 1.7769684791564941, + "learning_rate": 3.7500311143975287e-06, + "loss": 3.209, + "step": 18046 + }, + { + "epoch": 1.4468140546391133, + "grad_norm": 1.7875990867614746, + "learning_rate": 3.7490175639431715e-06, + "loss": 3.2646, + "step": 18047 + }, + { + "epoch": 1.446894229420136, + "grad_norm": 1.7964850664138794, + "learning_rate": 3.748004118878175e-06, + "loss": 3.167, + "step": 18048 + }, + { + "epoch": 1.4469744042011585, + "grad_norm": 1.7794828414916992, + "learning_rate": 3.746990779219626e-06, + "loss": 3.129, + "step": 18049 + }, + { + "epoch": 1.4470545789821811, + "grad_norm": 1.716513752937317, + "learning_rate": 3.7459775449846115e-06, + "loss": 3.1311, + "step": 18050 + }, + { + "epoch": 1.4471347537632038, + "grad_norm": 1.720763087272644, + "learning_rate": 3.7449644161902087e-06, + "loss": 3.156, + "step": 18051 + }, + { + "epoch": 1.4472149285442264, + "grad_norm": 1.7476223707199097, + "learning_rate": 3.743951392853501e-06, + "loss": 3.1141, + "step": 18052 + }, + { + "epoch": 1.447295103325249, + "grad_norm": 1.8166002035140991, + "learning_rate": 3.7429384749915675e-06, + "loss": 3.1979, + "step": 18053 + }, + { + "epoch": 1.4473752781062716, + "grad_norm": 1.659909963607788, + "learning_rate": 3.741925662621486e-06, + "loss": 3.1987, + "step": 18054 + }, + { + "epoch": 1.4474554528872943, + "grad_norm": 1.6184494495391846, + "learning_rate": 3.7409129557603363e-06, + "loss": 3.2205, + "step": 18055 + }, + { + "epoch": 1.447535627668317, + "grad_norm": 1.8016761541366577, + "learning_rate": 3.7399003544251853e-06, + "loss": 3.2469, + "step": 18056 + }, + { + "epoch": 1.4476158024493395, + "grad_norm": 1.7935537099838257, + "learning_rate": 3.738887858633107e-06, + "loss": 3.0419, + "step": 18057 + }, + { + "epoch": 1.4476959772303621, + "grad_norm": 1.703181505203247, + "learning_rate": 3.737875468401173e-06, + "loss": 3.178, + "step": 18058 + }, + { + "epoch": 1.4477761520113848, + "grad_norm": 1.7282054424285889, + "learning_rate": 3.736863183746452e-06, + "loss": 3.2167, + "step": 18059 + }, + { + "epoch": 1.4478563267924074, + "grad_norm": 1.7532031536102295, + "learning_rate": 3.7358510046860143e-06, + "loss": 3.0935, + "step": 18060 + }, + { + "epoch": 1.44793650157343, + "grad_norm": 1.8744641542434692, + "learning_rate": 3.734838931236917e-06, + "loss": 3.256, + "step": 18061 + }, + { + "epoch": 1.4480166763544526, + "grad_norm": 1.6916896104812622, + "learning_rate": 3.733826963416227e-06, + "loss": 3.0974, + "step": 18062 + }, + { + "epoch": 1.4480968511354753, + "grad_norm": 1.7272748947143555, + "learning_rate": 3.7328151012410064e-06, + "loss": 3.2998, + "step": 18063 + }, + { + "epoch": 1.4481770259164979, + "grad_norm": 1.734047770500183, + "learning_rate": 3.7318033447283144e-06, + "loss": 3.083, + "step": 18064 + }, + { + "epoch": 1.4482572006975207, + "grad_norm": 1.7013890743255615, + "learning_rate": 3.7307916938952115e-06, + "loss": 3.0864, + "step": 18065 + }, + { + "epoch": 1.4483373754785434, + "grad_norm": 0.6842887997627258, + "learning_rate": 3.7297801487587482e-06, + "loss": 2.4666, + "step": 18066 + }, + { + "epoch": 1.448417550259566, + "grad_norm": 1.7210896015167236, + "learning_rate": 3.7287687093359825e-06, + "loss": 3.2375, + "step": 18067 + }, + { + "epoch": 1.4484977250405886, + "grad_norm": 1.814693808555603, + "learning_rate": 3.7277573756439655e-06, + "loss": 3.1665, + "step": 18068 + }, + { + "epoch": 1.4485778998216112, + "grad_norm": 1.7691434621810913, + "learning_rate": 3.726746147699748e-06, + "loss": 3.2073, + "step": 18069 + }, + { + "epoch": 1.4486580746026338, + "grad_norm": 1.873582363128662, + "learning_rate": 3.725735025520384e-06, + "loss": 3.1742, + "step": 18070 + }, + { + "epoch": 1.4487382493836565, + "grad_norm": 1.750983715057373, + "learning_rate": 3.724724009122912e-06, + "loss": 3.1906, + "step": 18071 + }, + { + "epoch": 1.448818424164679, + "grad_norm": 1.7391284704208374, + "learning_rate": 3.723713098524382e-06, + "loss": 3.2365, + "step": 18072 + }, + { + "epoch": 1.4488985989457017, + "grad_norm": 1.7416431903839111, + "learning_rate": 3.722702293741838e-06, + "loss": 3.1619, + "step": 18073 + }, + { + "epoch": 1.4489787737267243, + "grad_norm": 1.700779914855957, + "learning_rate": 3.7216915947923194e-06, + "loss": 3.2181, + "step": 18074 + }, + { + "epoch": 1.449058948507747, + "grad_norm": 1.7520381212234497, + "learning_rate": 3.720681001692872e-06, + "loss": 3.2203, + "step": 18075 + }, + { + "epoch": 1.4491391232887696, + "grad_norm": 1.7169102430343628, + "learning_rate": 3.719670514460526e-06, + "loss": 3.0894, + "step": 18076 + }, + { + "epoch": 1.4492192980697922, + "grad_norm": 1.9354504346847534, + "learning_rate": 3.7186601331123227e-06, + "loss": 3.1884, + "step": 18077 + }, + { + "epoch": 1.4492994728508148, + "grad_norm": 1.7716255187988281, + "learning_rate": 3.7176498576652954e-06, + "loss": 3.1003, + "step": 18078 + }, + { + "epoch": 1.4493796476318375, + "grad_norm": 1.8151391744613647, + "learning_rate": 3.7166396881364774e-06, + "loss": 3.2418, + "step": 18079 + }, + { + "epoch": 1.44945982241286, + "grad_norm": 0.6659196019172668, + "learning_rate": 3.7156296245429023e-06, + "loss": 2.46, + "step": 18080 + }, + { + "epoch": 1.4495399971938827, + "grad_norm": 1.7881042957305908, + "learning_rate": 3.7146196669015945e-06, + "loss": 3.1844, + "step": 18081 + }, + { + "epoch": 1.4496201719749053, + "grad_norm": 1.7494041919708252, + "learning_rate": 3.713609815229584e-06, + "loss": 3.1473, + "step": 18082 + }, + { + "epoch": 1.449700346755928, + "grad_norm": 1.8397408723831177, + "learning_rate": 3.712600069543896e-06, + "loss": 3.1917, + "step": 18083 + }, + { + "epoch": 1.4497805215369506, + "grad_norm": 1.8290989398956299, + "learning_rate": 3.7115904298615546e-06, + "loss": 3.1257, + "step": 18084 + }, + { + "epoch": 1.4498606963179732, + "grad_norm": 1.728672981262207, + "learning_rate": 3.710580896199587e-06, + "loss": 3.1167, + "step": 18085 + }, + { + "epoch": 1.4499408710989958, + "grad_norm": 1.7973194122314453, + "learning_rate": 3.709571468575004e-06, + "loss": 3.1621, + "step": 18086 + }, + { + "epoch": 1.4500210458800185, + "grad_norm": 1.8778305053710938, + "learning_rate": 3.7085621470048293e-06, + "loss": 3.0806, + "step": 18087 + }, + { + "epoch": 1.450101220661041, + "grad_norm": 1.8397237062454224, + "learning_rate": 3.707552931506081e-06, + "loss": 3.0739, + "step": 18088 + }, + { + "epoch": 1.4501813954420637, + "grad_norm": 1.8313366174697876, + "learning_rate": 3.7065438220957706e-06, + "loss": 3.1185, + "step": 18089 + }, + { + "epoch": 1.4502615702230863, + "grad_norm": 1.7330983877182007, + "learning_rate": 3.705534818790918e-06, + "loss": 3.165, + "step": 18090 + }, + { + "epoch": 1.450341745004109, + "grad_norm": 1.7780210971832275, + "learning_rate": 3.704525921608526e-06, + "loss": 3.1899, + "step": 18091 + }, + { + "epoch": 1.4504219197851316, + "grad_norm": 1.8067429065704346, + "learning_rate": 3.703517130565609e-06, + "loss": 3.2507, + "step": 18092 + }, + { + "epoch": 1.4505020945661542, + "grad_norm": 1.824499487876892, + "learning_rate": 3.7025084456791736e-06, + "loss": 3.2355, + "step": 18093 + }, + { + "epoch": 1.4505822693471768, + "grad_norm": 1.6935206651687622, + "learning_rate": 3.701499866966226e-06, + "loss": 3.0785, + "step": 18094 + }, + { + "epoch": 1.4506624441281994, + "grad_norm": 1.7443903684616089, + "learning_rate": 3.700491394443775e-06, + "loss": 3.2112, + "step": 18095 + }, + { + "epoch": 1.450742618909222, + "grad_norm": 1.7540749311447144, + "learning_rate": 3.699483028128814e-06, + "loss": 3.1653, + "step": 18096 + }, + { + "epoch": 1.4508227936902447, + "grad_norm": 1.698996663093567, + "learning_rate": 3.6984747680383505e-06, + "loss": 3.1323, + "step": 18097 + }, + { + "epoch": 1.4509029684712673, + "grad_norm": 1.7141350507736206, + "learning_rate": 3.69746661418938e-06, + "loss": 3.2213, + "step": 18098 + }, + { + "epoch": 1.45098314325229, + "grad_norm": 1.7240604162216187, + "learning_rate": 3.6964585665989026e-06, + "loss": 3.084, + "step": 18099 + }, + { + "epoch": 1.4510633180333126, + "grad_norm": 1.796431541442871, + "learning_rate": 3.6954506252839152e-06, + "loss": 3.1322, + "step": 18100 + }, + { + "epoch": 1.4511434928143352, + "grad_norm": 1.7356146574020386, + "learning_rate": 3.6944427902614043e-06, + "loss": 3.1764, + "step": 18101 + }, + { + "epoch": 1.4512236675953578, + "grad_norm": 0.6701119542121887, + "learning_rate": 3.6934350615483663e-06, + "loss": 2.4709, + "step": 18102 + }, + { + "epoch": 1.4513038423763804, + "grad_norm": 1.9158318042755127, + "learning_rate": 3.6924274391617898e-06, + "loss": 3.2008, + "step": 18103 + }, + { + "epoch": 1.451384017157403, + "grad_norm": 1.8226324319839478, + "learning_rate": 3.6914199231186633e-06, + "loss": 3.3288, + "step": 18104 + }, + { + "epoch": 1.4514641919384257, + "grad_norm": 1.7609360218048096, + "learning_rate": 3.690412513435975e-06, + "loss": 3.16, + "step": 18105 + }, + { + "epoch": 1.4515443667194483, + "grad_norm": 1.932316541671753, + "learning_rate": 3.6894052101307097e-06, + "loss": 3.1413, + "step": 18106 + }, + { + "epoch": 1.451624541500471, + "grad_norm": 1.7132031917572021, + "learning_rate": 3.688398013219846e-06, + "loss": 3.1914, + "step": 18107 + }, + { + "epoch": 1.4517047162814936, + "grad_norm": 1.7144310474395752, + "learning_rate": 3.6873909227203676e-06, + "loss": 3.2052, + "step": 18108 + }, + { + "epoch": 1.4517848910625162, + "grad_norm": 1.7368061542510986, + "learning_rate": 3.686383938649254e-06, + "loss": 3.081, + "step": 18109 + }, + { + "epoch": 1.4518650658435388, + "grad_norm": 1.6296124458312988, + "learning_rate": 3.6853770610234806e-06, + "loss": 3.1523, + "step": 18110 + }, + { + "epoch": 1.4519452406245614, + "grad_norm": 1.7932462692260742, + "learning_rate": 3.6843702898600285e-06, + "loss": 3.1722, + "step": 18111 + }, + { + "epoch": 1.452025415405584, + "grad_norm": 1.766941785812378, + "learning_rate": 3.683363625175864e-06, + "loss": 3.2106, + "step": 18112 + }, + { + "epoch": 1.4521055901866067, + "grad_norm": 0.6630855798721313, + "learning_rate": 3.682357066987964e-06, + "loss": 2.45, + "step": 18113 + }, + { + "epoch": 1.4521857649676293, + "grad_norm": 1.8989838361740112, + "learning_rate": 3.6813506153132948e-06, + "loss": 3.1257, + "step": 18114 + }, + { + "epoch": 1.452265939748652, + "grad_norm": 0.6729923486709595, + "learning_rate": 3.6803442701688298e-06, + "loss": 2.4747, + "step": 18115 + }, + { + "epoch": 1.4523461145296748, + "grad_norm": 1.754933476448059, + "learning_rate": 3.6793380315715343e-06, + "loss": 3.1799, + "step": 18116 + }, + { + "epoch": 1.4524262893106974, + "grad_norm": 1.7730953693389893, + "learning_rate": 3.6783318995383697e-06, + "loss": 3.1694, + "step": 18117 + }, + { + "epoch": 1.45250646409172, + "grad_norm": 1.8084503412246704, + "learning_rate": 3.6773258740863016e-06, + "loss": 3.0534, + "step": 18118 + }, + { + "epoch": 1.4525866388727426, + "grad_norm": 1.7329349517822266, + "learning_rate": 3.6763199552322902e-06, + "loss": 3.1494, + "step": 18119 + }, + { + "epoch": 1.4526668136537653, + "grad_norm": 0.7023032903671265, + "learning_rate": 3.675314142993296e-06, + "loss": 2.475, + "step": 18120 + }, + { + "epoch": 1.452746988434788, + "grad_norm": 1.802670955657959, + "learning_rate": 3.67430843738628e-06, + "loss": 3.2113, + "step": 18121 + }, + { + "epoch": 1.4528271632158105, + "grad_norm": 0.6503587365150452, + "learning_rate": 3.6733028384281912e-06, + "loss": 2.465, + "step": 18122 + }, + { + "epoch": 1.4529073379968331, + "grad_norm": 1.6592719554901123, + "learning_rate": 3.6722973461359855e-06, + "loss": 3.2692, + "step": 18123 + }, + { + "epoch": 1.4529875127778558, + "grad_norm": 1.7382984161376953, + "learning_rate": 3.6712919605266173e-06, + "loss": 3.153, + "step": 18124 + }, + { + "epoch": 1.4530676875588784, + "grad_norm": 1.7445937395095825, + "learning_rate": 3.6702866816170357e-06, + "loss": 3.1146, + "step": 18125 + }, + { + "epoch": 1.453147862339901, + "grad_norm": 1.6623399257659912, + "learning_rate": 3.669281509424194e-06, + "loss": 3.2065, + "step": 18126 + }, + { + "epoch": 1.4532280371209236, + "grad_norm": 1.8176316022872925, + "learning_rate": 3.668276443965031e-06, + "loss": 3.0815, + "step": 18127 + }, + { + "epoch": 1.4533082119019463, + "grad_norm": 0.6677019000053406, + "learning_rate": 3.6672714852564962e-06, + "loss": 2.4659, + "step": 18128 + }, + { + "epoch": 1.4533883866829689, + "grad_norm": 1.767577886581421, + "learning_rate": 3.6662666333155316e-06, + "loss": 3.152, + "step": 18129 + }, + { + "epoch": 1.4534685614639915, + "grad_norm": 1.8306688070297241, + "learning_rate": 3.6652618881590797e-06, + "loss": 3.1942, + "step": 18130 + }, + { + "epoch": 1.4535487362450141, + "grad_norm": 1.813088059425354, + "learning_rate": 3.664257249804083e-06, + "loss": 3.093, + "step": 18131 + }, + { + "epoch": 1.4536289110260368, + "grad_norm": 1.7468699216842651, + "learning_rate": 3.6632527182674726e-06, + "loss": 3.1445, + "step": 18132 + }, + { + "epoch": 1.4537090858070594, + "grad_norm": 1.8103231191635132, + "learning_rate": 3.6622482935661886e-06, + "loss": 3.0916, + "step": 18133 + }, + { + "epoch": 1.453789260588082, + "grad_norm": 1.8249202966690063, + "learning_rate": 3.6612439757171647e-06, + "loss": 3.1875, + "step": 18134 + }, + { + "epoch": 1.4538694353691046, + "grad_norm": 1.767381191253662, + "learning_rate": 3.6602397647373332e-06, + "loss": 3.1424, + "step": 18135 + }, + { + "epoch": 1.4539496101501272, + "grad_norm": 1.8612651824951172, + "learning_rate": 3.6592356606436275e-06, + "loss": 3.2617, + "step": 18136 + }, + { + "epoch": 1.4540297849311499, + "grad_norm": 1.7422738075256348, + "learning_rate": 3.6582316634529714e-06, + "loss": 3.1783, + "step": 18137 + }, + { + "epoch": 1.4541099597121725, + "grad_norm": 1.661562442779541, + "learning_rate": 3.6572277731822948e-06, + "loss": 3.1956, + "step": 18138 + }, + { + "epoch": 1.4541901344931951, + "grad_norm": 1.850945234298706, + "learning_rate": 3.656223989848522e-06, + "loss": 3.186, + "step": 18139 + }, + { + "epoch": 1.4542703092742177, + "grad_norm": 1.7108091115951538, + "learning_rate": 3.655220313468576e-06, + "loss": 3.1593, + "step": 18140 + }, + { + "epoch": 1.4543504840552404, + "grad_norm": 0.6400585770606995, + "learning_rate": 3.6542167440593845e-06, + "loss": 2.4573, + "step": 18141 + }, + { + "epoch": 1.454430658836263, + "grad_norm": 1.7594280242919922, + "learning_rate": 3.6532132816378575e-06, + "loss": 3.1814, + "step": 18142 + }, + { + "epoch": 1.4545108336172856, + "grad_norm": 1.779771327972412, + "learning_rate": 3.6522099262209186e-06, + "loss": 3.0104, + "step": 18143 + }, + { + "epoch": 1.4545910083983082, + "grad_norm": 1.8639048337936401, + "learning_rate": 3.651206677825483e-06, + "loss": 3.2205, + "step": 18144 + }, + { + "epoch": 1.4546711831793309, + "grad_norm": 1.745184302330017, + "learning_rate": 3.650203536468465e-06, + "loss": 3.0989, + "step": 18145 + }, + { + "epoch": 1.4547513579603537, + "grad_norm": 1.7655296325683594, + "learning_rate": 3.649200502166781e-06, + "loss": 3.1876, + "step": 18146 + }, + { + "epoch": 1.4548315327413763, + "grad_norm": 1.74197518825531, + "learning_rate": 3.6481975749373342e-06, + "loss": 3.2153, + "step": 18147 + }, + { + "epoch": 1.454911707522399, + "grad_norm": 1.7440017461776733, + "learning_rate": 3.6471947547970388e-06, + "loss": 3.169, + "step": 18148 + }, + { + "epoch": 1.4549918823034216, + "grad_norm": 1.8115817308425903, + "learning_rate": 3.6461920417628003e-06, + "loss": 3.3011, + "step": 18149 + }, + { + "epoch": 1.4550720570844442, + "grad_norm": 1.7744741439819336, + "learning_rate": 3.645189435851525e-06, + "loss": 3.0756, + "step": 18150 + }, + { + "epoch": 1.4551522318654668, + "grad_norm": 1.8472425937652588, + "learning_rate": 3.6441869370801197e-06, + "loss": 3.1256, + "step": 18151 + }, + { + "epoch": 1.4552324066464895, + "grad_norm": 1.838636875152588, + "learning_rate": 3.643184545465479e-06, + "loss": 3.1238, + "step": 18152 + }, + { + "epoch": 1.455312581427512, + "grad_norm": 1.8503830432891846, + "learning_rate": 3.6421822610245063e-06, + "loss": 3.0931, + "step": 18153 + }, + { + "epoch": 1.4553927562085347, + "grad_norm": 1.703277349472046, + "learning_rate": 3.6411800837740994e-06, + "loss": 3.1859, + "step": 18154 + }, + { + "epoch": 1.4554729309895573, + "grad_norm": 1.7972851991653442, + "learning_rate": 3.640178013731156e-06, + "loss": 3.1158, + "step": 18155 + }, + { + "epoch": 1.45555310577058, + "grad_norm": 1.8215343952178955, + "learning_rate": 3.6391760509125738e-06, + "loss": 3.255, + "step": 18156 + }, + { + "epoch": 1.4556332805516026, + "grad_norm": 1.7402238845825195, + "learning_rate": 3.6381741953352377e-06, + "loss": 3.0938, + "step": 18157 + }, + { + "epoch": 1.4557134553326252, + "grad_norm": 1.764371633529663, + "learning_rate": 3.6371724470160417e-06, + "loss": 3.2017, + "step": 18158 + }, + { + "epoch": 1.4557936301136478, + "grad_norm": 1.8452738523483276, + "learning_rate": 3.636170805971877e-06, + "loss": 3.1641, + "step": 18159 + }, + { + "epoch": 1.4558738048946704, + "grad_norm": 1.854668140411377, + "learning_rate": 3.6351692722196296e-06, + "loss": 3.1663, + "step": 18160 + }, + { + "epoch": 1.455953979675693, + "grad_norm": 1.7446614503860474, + "learning_rate": 3.6341678457761875e-06, + "loss": 3.1575, + "step": 18161 + }, + { + "epoch": 1.4560341544567157, + "grad_norm": 1.7982112169265747, + "learning_rate": 3.6331665266584293e-06, + "loss": 3.2597, + "step": 18162 + }, + { + "epoch": 1.4561143292377383, + "grad_norm": 1.811445951461792, + "learning_rate": 3.6321653148832393e-06, + "loss": 3.1283, + "step": 18163 + }, + { + "epoch": 1.456194504018761, + "grad_norm": 1.7492237091064453, + "learning_rate": 3.6311642104674983e-06, + "loss": 3.0851, + "step": 18164 + }, + { + "epoch": 1.4562746787997836, + "grad_norm": 1.7873789072036743, + "learning_rate": 3.630163213428084e-06, + "loss": 3.2637, + "step": 18165 + }, + { + "epoch": 1.4563548535808062, + "grad_norm": 1.7210602760314941, + "learning_rate": 3.6291623237818764e-06, + "loss": 3.178, + "step": 18166 + }, + { + "epoch": 1.4564350283618288, + "grad_norm": 1.753440499305725, + "learning_rate": 3.628161541545743e-06, + "loss": 3.0632, + "step": 18167 + }, + { + "epoch": 1.4565152031428514, + "grad_norm": 1.719555139541626, + "learning_rate": 3.62716086673656e-06, + "loss": 3.1183, + "step": 18168 + }, + { + "epoch": 1.456595377923874, + "grad_norm": 1.829040288925171, + "learning_rate": 3.6261602993711997e-06, + "loss": 3.1938, + "step": 18169 + }, + { + "epoch": 1.4566755527048967, + "grad_norm": 1.8024049997329712, + "learning_rate": 3.62515983946653e-06, + "loss": 3.1674, + "step": 18170 + }, + { + "epoch": 1.4567557274859193, + "grad_norm": 1.7325901985168457, + "learning_rate": 3.6241594870394214e-06, + "loss": 3.2435, + "step": 18171 + }, + { + "epoch": 1.456835902266942, + "grad_norm": 1.8909778594970703, + "learning_rate": 3.6231592421067343e-06, + "loss": 3.2229, + "step": 18172 + }, + { + "epoch": 1.4569160770479646, + "grad_norm": 1.796104073524475, + "learning_rate": 3.6221591046853345e-06, + "loss": 3.1378, + "step": 18173 + }, + { + "epoch": 1.4569962518289872, + "grad_norm": 1.781376838684082, + "learning_rate": 3.6211590747920844e-06, + "loss": 3.0805, + "step": 18174 + }, + { + "epoch": 1.4570764266100098, + "grad_norm": 1.7886936664581299, + "learning_rate": 3.620159152443845e-06, + "loss": 3.2388, + "step": 18175 + }, + { + "epoch": 1.4571566013910324, + "grad_norm": 1.7896451950073242, + "learning_rate": 3.6191593376574765e-06, + "loss": 3.1151, + "step": 18176 + }, + { + "epoch": 1.457236776172055, + "grad_norm": 1.8403449058532715, + "learning_rate": 3.618159630449829e-06, + "loss": 3.1917, + "step": 18177 + }, + { + "epoch": 1.4573169509530777, + "grad_norm": 1.8021070957183838, + "learning_rate": 3.617160030837762e-06, + "loss": 3.3149, + "step": 18178 + }, + { + "epoch": 1.4573971257341003, + "grad_norm": 1.7361677885055542, + "learning_rate": 3.616160538838126e-06, + "loss": 3.1484, + "step": 18179 + }, + { + "epoch": 1.457477300515123, + "grad_norm": 1.7526298761367798, + "learning_rate": 3.6151611544677744e-06, + "loss": 3.3143, + "step": 18180 + }, + { + "epoch": 1.4575574752961455, + "grad_norm": 1.7527263164520264, + "learning_rate": 3.6141618777435583e-06, + "loss": 3.1655, + "step": 18181 + }, + { + "epoch": 1.4576376500771682, + "grad_norm": 1.7237374782562256, + "learning_rate": 3.6131627086823196e-06, + "loss": 3.1608, + "step": 18182 + }, + { + "epoch": 1.4577178248581908, + "grad_norm": 1.8322182893753052, + "learning_rate": 3.612163647300907e-06, + "loss": 3.2159, + "step": 18183 + }, + { + "epoch": 1.4577979996392134, + "grad_norm": 1.7646605968475342, + "learning_rate": 3.6111646936161627e-06, + "loss": 3.2061, + "step": 18184 + }, + { + "epoch": 1.457878174420236, + "grad_norm": 1.6414097547531128, + "learning_rate": 3.610165847644931e-06, + "loss": 3.0551, + "step": 18185 + }, + { + "epoch": 1.4579583492012587, + "grad_norm": 1.846090316772461, + "learning_rate": 3.6091671094040547e-06, + "loss": 3.1661, + "step": 18186 + }, + { + "epoch": 1.4580385239822813, + "grad_norm": 1.8193029165267944, + "learning_rate": 3.6081684789103646e-06, + "loss": 3.1886, + "step": 18187 + }, + { + "epoch": 1.458118698763304, + "grad_norm": 1.7471963167190552, + "learning_rate": 3.6071699561807007e-06, + "loss": 3.2021, + "step": 18188 + }, + { + "epoch": 1.4581988735443265, + "grad_norm": 1.7708758115768433, + "learning_rate": 3.6061715412318997e-06, + "loss": 3.198, + "step": 18189 + }, + { + "epoch": 1.4582790483253492, + "grad_norm": 1.7852368354797363, + "learning_rate": 3.605173234080792e-06, + "loss": 3.126, + "step": 18190 + }, + { + "epoch": 1.4583592231063718, + "grad_norm": 1.743914246559143, + "learning_rate": 3.6041750347442137e-06, + "loss": 3.2118, + "step": 18191 + }, + { + "epoch": 1.4584393978873944, + "grad_norm": 1.7004404067993164, + "learning_rate": 3.603176943238986e-06, + "loss": 3.2125, + "step": 18192 + }, + { + "epoch": 1.458519572668417, + "grad_norm": 1.7674105167388916, + "learning_rate": 3.602178959581942e-06, + "loss": 3.2406, + "step": 18193 + }, + { + "epoch": 1.4585997474494397, + "grad_norm": 1.7486774921417236, + "learning_rate": 3.6011810837899053e-06, + "loss": 3.1687, + "step": 18194 + }, + { + "epoch": 1.4586799222304623, + "grad_norm": 1.8141025304794312, + "learning_rate": 3.6001833158797005e-06, + "loss": 3.2886, + "step": 18195 + }, + { + "epoch": 1.458760097011485, + "grad_norm": 1.737933874130249, + "learning_rate": 3.599185655868153e-06, + "loss": 3.1819, + "step": 18196 + }, + { + "epoch": 1.4588402717925077, + "grad_norm": 1.855432391166687, + "learning_rate": 3.5981881037720755e-06, + "loss": 3.2678, + "step": 18197 + }, + { + "epoch": 1.4589204465735304, + "grad_norm": 1.8606336116790771, + "learning_rate": 3.5971906596082905e-06, + "loss": 3.1726, + "step": 18198 + }, + { + "epoch": 1.459000621354553, + "grad_norm": 1.92420494556427, + "learning_rate": 3.596193323393614e-06, + "loss": 3.3021, + "step": 18199 + }, + { + "epoch": 1.4590807961355756, + "grad_norm": 0.6877702474594116, + "learning_rate": 3.5951960951448615e-06, + "loss": 2.4746, + "step": 18200 + }, + { + "epoch": 1.4591609709165982, + "grad_norm": 1.694116234779358, + "learning_rate": 3.5941989748788497e-06, + "loss": 3.1513, + "step": 18201 + }, + { + "epoch": 1.4592411456976209, + "grad_norm": 1.7461742162704468, + "learning_rate": 3.5932019626123814e-06, + "loss": 3.1517, + "step": 18202 + }, + { + "epoch": 1.4593213204786435, + "grad_norm": 1.7567020654678345, + "learning_rate": 3.59220505836227e-06, + "loss": 3.2252, + "step": 18203 + }, + { + "epoch": 1.4594014952596661, + "grad_norm": 1.7438539266586304, + "learning_rate": 3.591208262145324e-06, + "loss": 3.0851, + "step": 18204 + }, + { + "epoch": 1.4594816700406887, + "grad_norm": 1.8249636888504028, + "learning_rate": 3.5902115739783462e-06, + "loss": 3.1483, + "step": 18205 + }, + { + "epoch": 1.4595618448217114, + "grad_norm": 1.7316800355911255, + "learning_rate": 3.5892149938781474e-06, + "loss": 3.2293, + "step": 18206 + }, + { + "epoch": 1.459642019602734, + "grad_norm": 1.752605676651001, + "learning_rate": 3.5882185218615197e-06, + "loss": 3.162, + "step": 18207 + }, + { + "epoch": 1.4597221943837566, + "grad_norm": 1.7393964529037476, + "learning_rate": 3.5872221579452692e-06, + "loss": 3.2374, + "step": 18208 + }, + { + "epoch": 1.4598023691647792, + "grad_norm": 1.7988700866699219, + "learning_rate": 3.586225902146192e-06, + "loss": 3.1851, + "step": 18209 + }, + { + "epoch": 1.4598825439458019, + "grad_norm": 1.8598766326904297, + "learning_rate": 3.5852297544810865e-06, + "loss": 3.2048, + "step": 18210 + }, + { + "epoch": 1.4599627187268245, + "grad_norm": 1.744278073310852, + "learning_rate": 3.5842337149667492e-06, + "loss": 3.157, + "step": 18211 + }, + { + "epoch": 1.460042893507847, + "grad_norm": 1.70891535282135, + "learning_rate": 3.5832377836199673e-06, + "loss": 3.1141, + "step": 18212 + }, + { + "epoch": 1.4601230682888697, + "grad_norm": 1.7784355878829956, + "learning_rate": 3.582241960457534e-06, + "loss": 3.1513, + "step": 18213 + }, + { + "epoch": 1.4602032430698924, + "grad_norm": 1.7825672626495361, + "learning_rate": 3.5812462454962407e-06, + "loss": 3.2928, + "step": 18214 + }, + { + "epoch": 1.460283417850915, + "grad_norm": 1.84080171585083, + "learning_rate": 3.5802506387528735e-06, + "loss": 3.1771, + "step": 18215 + }, + { + "epoch": 1.4603635926319376, + "grad_norm": 1.8379948139190674, + "learning_rate": 3.579255140244221e-06, + "loss": 3.0862, + "step": 18216 + }, + { + "epoch": 1.4604437674129602, + "grad_norm": 1.862571120262146, + "learning_rate": 3.57825974998706e-06, + "loss": 3.1636, + "step": 18217 + }, + { + "epoch": 1.4605239421939828, + "grad_norm": 1.7582497596740723, + "learning_rate": 3.577264467998177e-06, + "loss": 3.147, + "step": 18218 + }, + { + "epoch": 1.4606041169750055, + "grad_norm": 1.7569749355316162, + "learning_rate": 3.576269294294351e-06, + "loss": 3.0707, + "step": 18219 + }, + { + "epoch": 1.460684291756028, + "grad_norm": 1.8299976587295532, + "learning_rate": 3.5752742288923615e-06, + "loss": 3.3029, + "step": 18220 + }, + { + "epoch": 1.4607644665370507, + "grad_norm": 1.6787025928497314, + "learning_rate": 3.574279271808987e-06, + "loss": 3.1661, + "step": 18221 + }, + { + "epoch": 1.4608446413180733, + "grad_norm": 1.8625959157943726, + "learning_rate": 3.5732844230609966e-06, + "loss": 3.2479, + "step": 18222 + }, + { + "epoch": 1.460924816099096, + "grad_norm": 1.74349045753479, + "learning_rate": 3.5722896826651654e-06, + "loss": 3.1897, + "step": 18223 + }, + { + "epoch": 1.4610049908801186, + "grad_norm": 1.7641987800598145, + "learning_rate": 3.571295050638265e-06, + "loss": 3.0856, + "step": 18224 + }, + { + "epoch": 1.4610851656611412, + "grad_norm": 1.7969292402267456, + "learning_rate": 3.5703005269970636e-06, + "loss": 3.1956, + "step": 18225 + }, + { + "epoch": 1.4611653404421638, + "grad_norm": 1.9308173656463623, + "learning_rate": 3.5693061117583326e-06, + "loss": 3.2721, + "step": 18226 + }, + { + "epoch": 1.4612455152231865, + "grad_norm": 1.8031036853790283, + "learning_rate": 3.568311804938832e-06, + "loss": 3.2163, + "step": 18227 + }, + { + "epoch": 1.4613256900042093, + "grad_norm": 1.776121735572815, + "learning_rate": 3.5673176065553273e-06, + "loss": 3.0759, + "step": 18228 + }, + { + "epoch": 1.461405864785232, + "grad_norm": 1.7247426509857178, + "learning_rate": 3.5663235166245803e-06, + "loss": 3.1402, + "step": 18229 + }, + { + "epoch": 1.4614860395662546, + "grad_norm": 1.8307781219482422, + "learning_rate": 3.565329535163351e-06, + "loss": 3.1987, + "step": 18230 + }, + { + "epoch": 1.4615662143472772, + "grad_norm": 1.8458760976791382, + "learning_rate": 3.5643356621884016e-06, + "loss": 3.1156, + "step": 18231 + }, + { + "epoch": 1.4616463891282998, + "grad_norm": 1.766424298286438, + "learning_rate": 3.563341897716481e-06, + "loss": 3.126, + "step": 18232 + }, + { + "epoch": 1.4617265639093224, + "grad_norm": 1.8155916929244995, + "learning_rate": 3.5623482417643483e-06, + "loss": 3.1676, + "step": 18233 + }, + { + "epoch": 1.461806738690345, + "grad_norm": 1.772642731666565, + "learning_rate": 3.5613546943487553e-06, + "loss": 3.2144, + "step": 18234 + }, + { + "epoch": 1.4618869134713677, + "grad_norm": 1.7351380586624146, + "learning_rate": 3.560361255486453e-06, + "loss": 3.0316, + "step": 18235 + }, + { + "epoch": 1.4619670882523903, + "grad_norm": 1.7630000114440918, + "learning_rate": 3.559367925194194e-06, + "loss": 3.1893, + "step": 18236 + }, + { + "epoch": 1.462047263033413, + "grad_norm": 1.7419497966766357, + "learning_rate": 3.558374703488717e-06, + "loss": 3.2135, + "step": 18237 + }, + { + "epoch": 1.4621274378144355, + "grad_norm": 1.8685730695724487, + "learning_rate": 3.557381590386774e-06, + "loss": 3.1034, + "step": 18238 + }, + { + "epoch": 1.4622076125954582, + "grad_norm": 1.6800421476364136, + "learning_rate": 3.5563885859051063e-06, + "loss": 3.1234, + "step": 18239 + }, + { + "epoch": 1.4622877873764808, + "grad_norm": 1.762618899345398, + "learning_rate": 3.555395690060456e-06, + "loss": 3.2179, + "step": 18240 + }, + { + "epoch": 1.4623679621575034, + "grad_norm": 0.6863480806350708, + "learning_rate": 3.554402902869566e-06, + "loss": 2.4867, + "step": 18241 + }, + { + "epoch": 1.462448136938526, + "grad_norm": 1.7473266124725342, + "learning_rate": 3.5534102243491685e-06, + "loss": 3.1605, + "step": 18242 + }, + { + "epoch": 1.4625283117195487, + "grad_norm": 1.7695603370666504, + "learning_rate": 3.5524176545160027e-06, + "loss": 3.2148, + "step": 18243 + }, + { + "epoch": 1.4626084865005713, + "grad_norm": 1.6242315769195557, + "learning_rate": 3.551425193386803e-06, + "loss": 3.056, + "step": 18244 + }, + { + "epoch": 1.462688661281594, + "grad_norm": 1.8009296655654907, + "learning_rate": 3.5504328409783006e-06, + "loss": 3.1482, + "step": 18245 + }, + { + "epoch": 1.4627688360626165, + "grad_norm": 1.7995539903640747, + "learning_rate": 3.549440597307232e-06, + "loss": 3.2102, + "step": 18246 + }, + { + "epoch": 1.4628490108436392, + "grad_norm": 1.7950034141540527, + "learning_rate": 3.5484484623903183e-06, + "loss": 3.1833, + "step": 18247 + }, + { + "epoch": 1.4629291856246618, + "grad_norm": 1.740425944328308, + "learning_rate": 3.547456436244289e-06, + "loss": 3.0958, + "step": 18248 + }, + { + "epoch": 1.4630093604056844, + "grad_norm": 1.7227281332015991, + "learning_rate": 3.5464645188858703e-06, + "loss": 3.2015, + "step": 18249 + }, + { + "epoch": 1.463089535186707, + "grad_norm": 1.7997324466705322, + "learning_rate": 3.5454727103317854e-06, + "loss": 3.1698, + "step": 18250 + }, + { + "epoch": 1.4631697099677297, + "grad_norm": 1.6905804872512817, + "learning_rate": 3.544481010598759e-06, + "loss": 3.0729, + "step": 18251 + }, + { + "epoch": 1.4632498847487523, + "grad_norm": 1.8207006454467773, + "learning_rate": 3.543489419703505e-06, + "loss": 3.2372, + "step": 18252 + }, + { + "epoch": 1.463330059529775, + "grad_norm": 1.849777102470398, + "learning_rate": 3.542497937662743e-06, + "loss": 3.2357, + "step": 18253 + }, + { + "epoch": 1.4634102343107975, + "grad_norm": 1.7743785381317139, + "learning_rate": 3.5415065644931913e-06, + "loss": 3.1301, + "step": 18254 + }, + { + "epoch": 1.4634904090918202, + "grad_norm": 1.813185453414917, + "learning_rate": 3.540515300211561e-06, + "loss": 3.1323, + "step": 18255 + }, + { + "epoch": 1.4635705838728428, + "grad_norm": 1.8155277967453003, + "learning_rate": 3.5395241448345706e-06, + "loss": 3.1492, + "step": 18256 + }, + { + "epoch": 1.4636507586538654, + "grad_norm": 1.7666027545928955, + "learning_rate": 3.5385330983789224e-06, + "loss": 3.1799, + "step": 18257 + }, + { + "epoch": 1.463730933434888, + "grad_norm": 1.657492995262146, + "learning_rate": 3.5375421608613292e-06, + "loss": 3.2038, + "step": 18258 + }, + { + "epoch": 1.4638111082159107, + "grad_norm": 1.8095699548721313, + "learning_rate": 3.536551332298498e-06, + "loss": 3.1978, + "step": 18259 + }, + { + "epoch": 1.4638912829969333, + "grad_norm": 1.8185118436813354, + "learning_rate": 3.535560612707133e-06, + "loss": 3.2122, + "step": 18260 + }, + { + "epoch": 1.463971457777956, + "grad_norm": 1.7536345720291138, + "learning_rate": 3.534570002103941e-06, + "loss": 3.1733, + "step": 18261 + }, + { + "epoch": 1.4640516325589785, + "grad_norm": 1.7249106168746948, + "learning_rate": 3.5335795005056174e-06, + "loss": 3.1643, + "step": 18262 + }, + { + "epoch": 1.4641318073400011, + "grad_norm": 1.7053261995315552, + "learning_rate": 3.5325891079288634e-06, + "loss": 3.1291, + "step": 18263 + }, + { + "epoch": 1.4642119821210238, + "grad_norm": 1.8389321565628052, + "learning_rate": 3.5315988243903778e-06, + "loss": 3.1616, + "step": 18264 + }, + { + "epoch": 1.4642921569020464, + "grad_norm": 1.7457537651062012, + "learning_rate": 3.530608649906857e-06, + "loss": 3.2507, + "step": 18265 + }, + { + "epoch": 1.464372331683069, + "grad_norm": 1.8079783916473389, + "learning_rate": 3.529618584494997e-06, + "loss": 3.2684, + "step": 18266 + }, + { + "epoch": 1.4644525064640916, + "grad_norm": 1.7924861907958984, + "learning_rate": 3.5286286281714844e-06, + "loss": 3.1823, + "step": 18267 + }, + { + "epoch": 1.4645326812451143, + "grad_norm": 0.6887668371200562, + "learning_rate": 3.527638780953012e-06, + "loss": 2.473, + "step": 18268 + }, + { + "epoch": 1.4646128560261369, + "grad_norm": 1.8384146690368652, + "learning_rate": 3.5266490428562695e-06, + "loss": 3.2258, + "step": 18269 + }, + { + "epoch": 1.4646930308071595, + "grad_norm": 1.7134146690368652, + "learning_rate": 3.5256594138979418e-06, + "loss": 3.1837, + "step": 18270 + }, + { + "epoch": 1.4647732055881821, + "grad_norm": 0.6738194823265076, + "learning_rate": 3.5246698940947188e-06, + "loss": 2.4663, + "step": 18271 + }, + { + "epoch": 1.4648533803692048, + "grad_norm": 1.6685223579406738, + "learning_rate": 3.523680483463275e-06, + "loss": 3.1327, + "step": 18272 + }, + { + "epoch": 1.4649335551502274, + "grad_norm": 1.7886650562286377, + "learning_rate": 3.5226911820202968e-06, + "loss": 3.1449, + "step": 18273 + }, + { + "epoch": 1.46501372993125, + "grad_norm": 1.7263119220733643, + "learning_rate": 3.521701989782461e-06, + "loss": 3.1989, + "step": 18274 + }, + { + "epoch": 1.4650939047122726, + "grad_norm": 1.6429439783096313, + "learning_rate": 3.520712906766448e-06, + "loss": 3.1326, + "step": 18275 + }, + { + "epoch": 1.4651740794932953, + "grad_norm": 1.8765002489089966, + "learning_rate": 3.519723932988934e-06, + "loss": 3.1978, + "step": 18276 + }, + { + "epoch": 1.4652542542743179, + "grad_norm": 1.7581127882003784, + "learning_rate": 3.5187350684665877e-06, + "loss": 3.1425, + "step": 18277 + }, + { + "epoch": 1.4653344290553407, + "grad_norm": 1.8757028579711914, + "learning_rate": 3.517746313216085e-06, + "loss": 3.2236, + "step": 18278 + }, + { + "epoch": 1.4654146038363633, + "grad_norm": 1.7957725524902344, + "learning_rate": 3.5167576672540936e-06, + "loss": 3.1821, + "step": 18279 + }, + { + "epoch": 1.465494778617386, + "grad_norm": 1.8807365894317627, + "learning_rate": 3.515769130597283e-06, + "loss": 3.131, + "step": 18280 + }, + { + "epoch": 1.4655749533984086, + "grad_norm": 0.6684507727622986, + "learning_rate": 3.5147807032623238e-06, + "loss": 2.4634, + "step": 18281 + }, + { + "epoch": 1.4656551281794312, + "grad_norm": 1.7763359546661377, + "learning_rate": 3.513792385265873e-06, + "loss": 3.2722, + "step": 18282 + }, + { + "epoch": 1.4657353029604538, + "grad_norm": 1.8492534160614014, + "learning_rate": 3.5128041766245957e-06, + "loss": 3.1199, + "step": 18283 + }, + { + "epoch": 1.4658154777414765, + "grad_norm": 1.791764259338379, + "learning_rate": 3.5118160773551535e-06, + "loss": 3.1737, + "step": 18284 + }, + { + "epoch": 1.465895652522499, + "grad_norm": 1.7423179149627686, + "learning_rate": 3.510828087474206e-06, + "loss": 3.0674, + "step": 18285 + }, + { + "epoch": 1.4659758273035217, + "grad_norm": 1.7855496406555176, + "learning_rate": 3.509840206998414e-06, + "loss": 3.1163, + "step": 18286 + }, + { + "epoch": 1.4660560020845443, + "grad_norm": 2.554245948791504, + "learning_rate": 3.5088524359444243e-06, + "loss": 3.1968, + "step": 18287 + }, + { + "epoch": 1.466136176865567, + "grad_norm": 1.6514339447021484, + "learning_rate": 3.5078647743288964e-06, + "loss": 3.1313, + "step": 18288 + }, + { + "epoch": 1.4662163516465896, + "grad_norm": 1.8838272094726562, + "learning_rate": 3.5068772221684787e-06, + "loss": 3.1539, + "step": 18289 + }, + { + "epoch": 1.4662965264276122, + "grad_norm": 1.7757552862167358, + "learning_rate": 3.505889779479823e-06, + "loss": 3.0797, + "step": 18290 + }, + { + "epoch": 1.4663767012086348, + "grad_norm": 1.691172480583191, + "learning_rate": 3.504902446279581e-06, + "loss": 3.1805, + "step": 18291 + }, + { + "epoch": 1.4664568759896575, + "grad_norm": 1.7262870073318481, + "learning_rate": 3.503915222584391e-06, + "loss": 3.1499, + "step": 18292 + }, + { + "epoch": 1.46653705077068, + "grad_norm": 1.8166193962097168, + "learning_rate": 3.5029281084109e-06, + "loss": 3.2095, + "step": 18293 + }, + { + "epoch": 1.4666172255517027, + "grad_norm": 0.6674491167068481, + "learning_rate": 3.501941103775752e-06, + "loss": 2.4697, + "step": 18294 + }, + { + "epoch": 1.4666974003327253, + "grad_norm": 1.7406550645828247, + "learning_rate": 3.500954208695587e-06, + "loss": 3.135, + "step": 18295 + }, + { + "epoch": 1.466777575113748, + "grad_norm": 1.739467978477478, + "learning_rate": 3.4999674231870474e-06, + "loss": 3.1779, + "step": 18296 + }, + { + "epoch": 1.4668577498947706, + "grad_norm": 1.8423763513565063, + "learning_rate": 3.4989807472667616e-06, + "loss": 3.1796, + "step": 18297 + }, + { + "epoch": 1.4669379246757932, + "grad_norm": 1.8428798913955688, + "learning_rate": 3.4979941809513695e-06, + "loss": 3.1783, + "step": 18298 + }, + { + "epoch": 1.4670180994568158, + "grad_norm": 1.765600323677063, + "learning_rate": 3.497007724257505e-06, + "loss": 3.243, + "step": 18299 + }, + { + "epoch": 1.4670982742378385, + "grad_norm": 1.699336051940918, + "learning_rate": 3.496021377201797e-06, + "loss": 3.0778, + "step": 18300 + }, + { + "epoch": 1.467178449018861, + "grad_norm": 1.7745016813278198, + "learning_rate": 3.4950351398008808e-06, + "loss": 3.134, + "step": 18301 + }, + { + "epoch": 1.4672586237998837, + "grad_norm": 1.7609466314315796, + "learning_rate": 3.4940490120713768e-06, + "loss": 3.0374, + "step": 18302 + }, + { + "epoch": 1.4673387985809063, + "grad_norm": 1.727186679840088, + "learning_rate": 3.493062994029912e-06, + "loss": 3.1254, + "step": 18303 + }, + { + "epoch": 1.467418973361929, + "grad_norm": 1.755629539489746, + "learning_rate": 3.492077085693112e-06, + "loss": 3.1, + "step": 18304 + }, + { + "epoch": 1.4674991481429516, + "grad_norm": 1.8354965448379517, + "learning_rate": 3.4910912870775994e-06, + "loss": 3.2154, + "step": 18305 + }, + { + "epoch": 1.4675793229239742, + "grad_norm": 1.8917254209518433, + "learning_rate": 3.4901055981999964e-06, + "loss": 3.202, + "step": 18306 + }, + { + "epoch": 1.4676594977049968, + "grad_norm": 1.784292459487915, + "learning_rate": 3.4891200190769157e-06, + "loss": 3.2228, + "step": 18307 + }, + { + "epoch": 1.4677396724860194, + "grad_norm": 1.8224133253097534, + "learning_rate": 3.4881345497249776e-06, + "loss": 3.1171, + "step": 18308 + }, + { + "epoch": 1.4678198472670423, + "grad_norm": 1.6469076871871948, + "learning_rate": 3.4871491901607956e-06, + "loss": 3.1427, + "step": 18309 + }, + { + "epoch": 1.467900022048065, + "grad_norm": 1.7834217548370361, + "learning_rate": 3.486163940400982e-06, + "loss": 3.137, + "step": 18310 + }, + { + "epoch": 1.4679801968290875, + "grad_norm": 1.7836483716964722, + "learning_rate": 3.4851788004621525e-06, + "loss": 3.2468, + "step": 18311 + }, + { + "epoch": 1.4680603716101102, + "grad_norm": 1.709871530532837, + "learning_rate": 3.48419377036091e-06, + "loss": 3.0776, + "step": 18312 + }, + { + "epoch": 1.4681405463911328, + "grad_norm": 0.6774117350578308, + "learning_rate": 3.483208850113863e-06, + "loss": 2.4697, + "step": 18313 + }, + { + "epoch": 1.4682207211721554, + "grad_norm": 1.7715785503387451, + "learning_rate": 3.4822240397376183e-06, + "loss": 3.1804, + "step": 18314 + }, + { + "epoch": 1.468300895953178, + "grad_norm": 1.7642033100128174, + "learning_rate": 3.4812393392487785e-06, + "loss": 3.1532, + "step": 18315 + }, + { + "epoch": 1.4683810707342007, + "grad_norm": 1.8268312215805054, + "learning_rate": 3.4802547486639503e-06, + "loss": 3.1715, + "step": 18316 + }, + { + "epoch": 1.4684612455152233, + "grad_norm": 1.8527772426605225, + "learning_rate": 3.479270267999726e-06, + "loss": 3.1415, + "step": 18317 + }, + { + "epoch": 1.468541420296246, + "grad_norm": 1.861111044883728, + "learning_rate": 3.4782858972727063e-06, + "loss": 3.1482, + "step": 18318 + }, + { + "epoch": 1.4686215950772685, + "grad_norm": 1.6997215747833252, + "learning_rate": 3.4773016364994883e-06, + "loss": 3.2233, + "step": 18319 + }, + { + "epoch": 1.4687017698582912, + "grad_norm": 1.6901676654815674, + "learning_rate": 3.4763174856966663e-06, + "loss": 3.1682, + "step": 18320 + }, + { + "epoch": 1.4687819446393138, + "grad_norm": 1.677601933479309, + "learning_rate": 3.4753334448808353e-06, + "loss": 3.1249, + "step": 18321 + }, + { + "epoch": 1.4688621194203364, + "grad_norm": 1.8113315105438232, + "learning_rate": 3.474349514068579e-06, + "loss": 3.0825, + "step": 18322 + }, + { + "epoch": 1.468942294201359, + "grad_norm": 1.6911484003067017, + "learning_rate": 3.4733656932764902e-06, + "loss": 3.1672, + "step": 18323 + }, + { + "epoch": 1.4690224689823816, + "grad_norm": 1.7965714931488037, + "learning_rate": 3.4723819825211557e-06, + "loss": 3.244, + "step": 18324 + }, + { + "epoch": 1.4691026437634043, + "grad_norm": 1.8165603876113892, + "learning_rate": 3.471398381819161e-06, + "loss": 3.1725, + "step": 18325 + }, + { + "epoch": 1.469182818544427, + "grad_norm": 1.7717480659484863, + "learning_rate": 3.4704148911870915e-06, + "loss": 3.1516, + "step": 18326 + }, + { + "epoch": 1.4692629933254495, + "grad_norm": 1.7006139755249023, + "learning_rate": 3.4694315106415224e-06, + "loss": 3.1936, + "step": 18327 + }, + { + "epoch": 1.4693431681064721, + "grad_norm": 1.7845289707183838, + "learning_rate": 3.468448240199036e-06, + "loss": 3.1769, + "step": 18328 + }, + { + "epoch": 1.4694233428874948, + "grad_norm": 1.8936847448349, + "learning_rate": 3.467465079876211e-06, + "loss": 3.1934, + "step": 18329 + }, + { + "epoch": 1.4695035176685174, + "grad_norm": 1.855790615081787, + "learning_rate": 3.4664820296896217e-06, + "loss": 3.286, + "step": 18330 + }, + { + "epoch": 1.46958369244954, + "grad_norm": 1.7640430927276611, + "learning_rate": 3.465499089655847e-06, + "loss": 3.17, + "step": 18331 + }, + { + "epoch": 1.4696638672305626, + "grad_norm": 1.7277729511260986, + "learning_rate": 3.4645162597914505e-06, + "loss": 3.1296, + "step": 18332 + }, + { + "epoch": 1.4697440420115853, + "grad_norm": 1.8237072229385376, + "learning_rate": 3.463533540113008e-06, + "loss": 3.1642, + "step": 18333 + }, + { + "epoch": 1.4698242167926079, + "grad_norm": 1.9076656103134155, + "learning_rate": 3.4625509306370854e-06, + "loss": 3.3724, + "step": 18334 + }, + { + "epoch": 1.4699043915736305, + "grad_norm": 1.8198622465133667, + "learning_rate": 3.4615684313802487e-06, + "loss": 3.1903, + "step": 18335 + }, + { + "epoch": 1.4699845663546531, + "grad_norm": 1.7473232746124268, + "learning_rate": 3.4605860423590686e-06, + "loss": 3.2289, + "step": 18336 + }, + { + "epoch": 1.4700647411356758, + "grad_norm": 1.724198341369629, + "learning_rate": 3.4596037635901004e-06, + "loss": 3.1227, + "step": 18337 + }, + { + "epoch": 1.4701449159166984, + "grad_norm": 1.6509568691253662, + "learning_rate": 3.458621595089907e-06, + "loss": 3.1466, + "step": 18338 + }, + { + "epoch": 1.470225090697721, + "grad_norm": 1.7136815786361694, + "learning_rate": 3.457639536875048e-06, + "loss": 3.2162, + "step": 18339 + }, + { + "epoch": 1.4703052654787436, + "grad_norm": 1.7772306203842163, + "learning_rate": 3.45665758896208e-06, + "loss": 3.198, + "step": 18340 + }, + { + "epoch": 1.4703854402597663, + "grad_norm": 0.6688451766967773, + "learning_rate": 3.455675751367563e-06, + "loss": 2.468, + "step": 18341 + }, + { + "epoch": 1.4704656150407889, + "grad_norm": 1.7612837553024292, + "learning_rate": 3.454694024108044e-06, + "loss": 3.0988, + "step": 18342 + }, + { + "epoch": 1.4705457898218115, + "grad_norm": 0.6659294366836548, + "learning_rate": 3.4537124072000762e-06, + "loss": 2.4512, + "step": 18343 + }, + { + "epoch": 1.4706259646028341, + "grad_norm": 1.8117882013320923, + "learning_rate": 3.4527309006602095e-06, + "loss": 3.2369, + "step": 18344 + }, + { + "epoch": 1.4707061393838567, + "grad_norm": 1.7727699279785156, + "learning_rate": 3.4517495045049933e-06, + "loss": 3.2716, + "step": 18345 + }, + { + "epoch": 1.4707863141648794, + "grad_norm": 1.8413397073745728, + "learning_rate": 3.450768218750975e-06, + "loss": 3.219, + "step": 18346 + }, + { + "epoch": 1.470866488945902, + "grad_norm": 1.8888499736785889, + "learning_rate": 3.449787043414693e-06, + "loss": 3.1882, + "step": 18347 + }, + { + "epoch": 1.4709466637269246, + "grad_norm": 1.844314455986023, + "learning_rate": 3.4488059785126936e-06, + "loss": 3.1278, + "step": 18348 + }, + { + "epoch": 1.4710268385079472, + "grad_norm": 1.9133154153823853, + "learning_rate": 3.447825024061516e-06, + "loss": 3.1619, + "step": 18349 + }, + { + "epoch": 1.4711070132889699, + "grad_norm": 1.8051930665969849, + "learning_rate": 3.4468441800776995e-06, + "loss": 3.1402, + "step": 18350 + }, + { + "epoch": 1.4711871880699925, + "grad_norm": 1.7300138473510742, + "learning_rate": 3.445863446577784e-06, + "loss": 3.1573, + "step": 18351 + }, + { + "epoch": 1.4712673628510151, + "grad_norm": 1.7013936042785645, + "learning_rate": 3.444882823578298e-06, + "loss": 3.0956, + "step": 18352 + }, + { + "epoch": 1.4713475376320377, + "grad_norm": 1.896964192390442, + "learning_rate": 3.443902311095777e-06, + "loss": 3.329, + "step": 18353 + }, + { + "epoch": 1.4714277124130604, + "grad_norm": 1.7189658880233765, + "learning_rate": 3.4429219091467525e-06, + "loss": 3.1412, + "step": 18354 + }, + { + "epoch": 1.471507887194083, + "grad_norm": 1.7699311971664429, + "learning_rate": 3.4419416177477528e-06, + "loss": 3.1978, + "step": 18355 + }, + { + "epoch": 1.4715880619751056, + "grad_norm": 1.8144478797912598, + "learning_rate": 3.4409614369153098e-06, + "loss": 3.1305, + "step": 18356 + }, + { + "epoch": 1.4716682367561282, + "grad_norm": 1.6073293685913086, + "learning_rate": 3.4399813666659433e-06, + "loss": 3.1166, + "step": 18357 + }, + { + "epoch": 1.4717484115371509, + "grad_norm": 1.6341608762741089, + "learning_rate": 3.439001407016178e-06, + "loss": 3.0524, + "step": 18358 + }, + { + "epoch": 1.4718285863181735, + "grad_norm": 1.801857829093933, + "learning_rate": 3.438021557982536e-06, + "loss": 3.2154, + "step": 18359 + }, + { + "epoch": 1.4719087610991963, + "grad_norm": 1.749611735343933, + "learning_rate": 3.437041819581539e-06, + "loss": 3.1922, + "step": 18360 + }, + { + "epoch": 1.471988935880219, + "grad_norm": 1.6948180198669434, + "learning_rate": 3.4360621918297067e-06, + "loss": 3.1077, + "step": 18361 + }, + { + "epoch": 1.4720691106612416, + "grad_norm": 1.7396987676620483, + "learning_rate": 3.4350826747435496e-06, + "loss": 3.1661, + "step": 18362 + }, + { + "epoch": 1.4721492854422642, + "grad_norm": 0.6726493835449219, + "learning_rate": 3.4341032683395847e-06, + "loss": 2.4554, + "step": 18363 + }, + { + "epoch": 1.4722294602232868, + "grad_norm": 2.178483486175537, + "learning_rate": 3.4331239726343247e-06, + "loss": 3.1666, + "step": 18364 + }, + { + "epoch": 1.4723096350043094, + "grad_norm": 1.7904223203659058, + "learning_rate": 3.432144787644279e-06, + "loss": 3.1619, + "step": 18365 + }, + { + "epoch": 1.472389809785332, + "grad_norm": 1.7708052396774292, + "learning_rate": 3.4311657133859623e-06, + "loss": 3.2101, + "step": 18366 + }, + { + "epoch": 1.4724699845663547, + "grad_norm": 0.6605420708656311, + "learning_rate": 3.430186749875872e-06, + "loss": 2.4666, + "step": 18367 + }, + { + "epoch": 1.4725501593473773, + "grad_norm": 1.9479676485061646, + "learning_rate": 3.4292078971305188e-06, + "loss": 3.2078, + "step": 18368 + }, + { + "epoch": 1.4726303341284, + "grad_norm": 1.8209809064865112, + "learning_rate": 3.4282291551664047e-06, + "loss": 3.1186, + "step": 18369 + }, + { + "epoch": 1.4727105089094226, + "grad_norm": 1.9906761646270752, + "learning_rate": 3.427250524000031e-06, + "loss": 3.2567, + "step": 18370 + }, + { + "epoch": 1.4727906836904452, + "grad_norm": 0.6560079455375671, + "learning_rate": 3.426272003647899e-06, + "loss": 2.4575, + "step": 18371 + }, + { + "epoch": 1.4728708584714678, + "grad_norm": 1.8360518217086792, + "learning_rate": 3.425293594126502e-06, + "loss": 3.1711, + "step": 18372 + }, + { + "epoch": 1.4729510332524904, + "grad_norm": 1.7768216133117676, + "learning_rate": 3.4243152954523373e-06, + "loss": 3.145, + "step": 18373 + }, + { + "epoch": 1.473031208033513, + "grad_norm": 1.78954017162323, + "learning_rate": 3.4233371076418997e-06, + "loss": 3.1468, + "step": 18374 + }, + { + "epoch": 1.4731113828145357, + "grad_norm": 1.8336378335952759, + "learning_rate": 3.4223590307116803e-06, + "loss": 3.1655, + "step": 18375 + }, + { + "epoch": 1.4731915575955583, + "grad_norm": 1.759728193283081, + "learning_rate": 3.421381064678173e-06, + "loss": 3.0873, + "step": 18376 + }, + { + "epoch": 1.473271732376581, + "grad_norm": 1.7633883953094482, + "learning_rate": 3.420403209557859e-06, + "loss": 3.1581, + "step": 18377 + }, + { + "epoch": 1.4733519071576036, + "grad_norm": 1.7588738203048706, + "learning_rate": 3.4194254653672275e-06, + "loss": 3.2515, + "step": 18378 + }, + { + "epoch": 1.4734320819386262, + "grad_norm": 1.8617541790008545, + "learning_rate": 3.4184478321227643e-06, + "loss": 3.1519, + "step": 18379 + }, + { + "epoch": 1.4735122567196488, + "grad_norm": 1.8425663709640503, + "learning_rate": 3.4174703098409513e-06, + "loss": 3.1143, + "step": 18380 + }, + { + "epoch": 1.4735924315006714, + "grad_norm": 1.6932531595230103, + "learning_rate": 3.4164928985382717e-06, + "loss": 3.1652, + "step": 18381 + }, + { + "epoch": 1.473672606281694, + "grad_norm": 1.7965754270553589, + "learning_rate": 3.4155155982311983e-06, + "loss": 3.2529, + "step": 18382 + }, + { + "epoch": 1.4737527810627167, + "grad_norm": 1.914226770401001, + "learning_rate": 3.414538408936211e-06, + "loss": 3.1668, + "step": 18383 + }, + { + "epoch": 1.4738329558437393, + "grad_norm": 1.7576446533203125, + "learning_rate": 3.4135613306697857e-06, + "loss": 3.2568, + "step": 18384 + }, + { + "epoch": 1.473913130624762, + "grad_norm": 1.812378168106079, + "learning_rate": 3.4125843634483945e-06, + "loss": 3.1559, + "step": 18385 + }, + { + "epoch": 1.4739933054057845, + "grad_norm": 1.7234898805618286, + "learning_rate": 3.4116075072885124e-06, + "loss": 3.0921, + "step": 18386 + }, + { + "epoch": 1.4740734801868072, + "grad_norm": 1.8530470132827759, + "learning_rate": 3.4106307622066036e-06, + "loss": 3.1841, + "step": 18387 + }, + { + "epoch": 1.4741536549678298, + "grad_norm": 1.8071986436843872, + "learning_rate": 3.4096541282191374e-06, + "loss": 3.3226, + "step": 18388 + }, + { + "epoch": 1.4742338297488524, + "grad_norm": 1.7408857345581055, + "learning_rate": 3.4086776053425796e-06, + "loss": 3.1368, + "step": 18389 + }, + { + "epoch": 1.4743140045298753, + "grad_norm": 1.7668648958206177, + "learning_rate": 3.4077011935933944e-06, + "loss": 3.183, + "step": 18390 + }, + { + "epoch": 1.474394179310898, + "grad_norm": 1.807449460029602, + "learning_rate": 3.406724892988047e-06, + "loss": 3.1435, + "step": 18391 + }, + { + "epoch": 1.4744743540919205, + "grad_norm": 1.843005895614624, + "learning_rate": 3.4057487035429913e-06, + "loss": 3.1373, + "step": 18392 + }, + { + "epoch": 1.4745545288729431, + "grad_norm": 1.785938024520874, + "learning_rate": 3.4047726252746882e-06, + "loss": 3.2205, + "step": 18393 + }, + { + "epoch": 1.4746347036539658, + "grad_norm": 0.6868214011192322, + "learning_rate": 3.403796658199594e-06, + "loss": 2.4752, + "step": 18394 + }, + { + "epoch": 1.4747148784349884, + "grad_norm": 1.8082568645477295, + "learning_rate": 3.4028208023341635e-06, + "loss": 3.1913, + "step": 18395 + }, + { + "epoch": 1.474795053216011, + "grad_norm": 1.8402725458145142, + "learning_rate": 3.4018450576948525e-06, + "loss": 3.2198, + "step": 18396 + }, + { + "epoch": 1.4748752279970336, + "grad_norm": 1.8032091856002808, + "learning_rate": 3.4008694242981056e-06, + "loss": 3.1512, + "step": 18397 + }, + { + "epoch": 1.4749554027780563, + "grad_norm": 1.6902616024017334, + "learning_rate": 3.3998939021603738e-06, + "loss": 3.045, + "step": 18398 + }, + { + "epoch": 1.4750355775590789, + "grad_norm": 1.7517095804214478, + "learning_rate": 3.3989184912981057e-06, + "loss": 3.151, + "step": 18399 + }, + { + "epoch": 1.4751157523401015, + "grad_norm": 1.7883179187774658, + "learning_rate": 3.397943191727744e-06, + "loss": 3.0676, + "step": 18400 + }, + { + "epoch": 1.4751959271211241, + "grad_norm": 0.6690191030502319, + "learning_rate": 3.396968003465738e-06, + "loss": 2.4619, + "step": 18401 + }, + { + "epoch": 1.4752761019021468, + "grad_norm": 1.6789088249206543, + "learning_rate": 3.3959929265285195e-06, + "loss": 3.1053, + "step": 18402 + }, + { + "epoch": 1.4753562766831694, + "grad_norm": 1.7791926860809326, + "learning_rate": 3.395017960932534e-06, + "loss": 3.2891, + "step": 18403 + }, + { + "epoch": 1.475436451464192, + "grad_norm": 1.711646318435669, + "learning_rate": 3.394043106694218e-06, + "loss": 3.1625, + "step": 18404 + }, + { + "epoch": 1.4755166262452146, + "grad_norm": 1.7797006368637085, + "learning_rate": 3.3930683638300066e-06, + "loss": 3.1629, + "step": 18405 + }, + { + "epoch": 1.4755968010262372, + "grad_norm": 1.8170993328094482, + "learning_rate": 3.3920937323563373e-06, + "loss": 3.1892, + "step": 18406 + }, + { + "epoch": 1.4756769758072599, + "grad_norm": 1.6700235605239868, + "learning_rate": 3.391119212289635e-06, + "loss": 3.1559, + "step": 18407 + }, + { + "epoch": 1.4757571505882825, + "grad_norm": 1.755258321762085, + "learning_rate": 3.390144803646335e-06, + "loss": 3.1736, + "step": 18408 + }, + { + "epoch": 1.4758373253693051, + "grad_norm": 1.7433288097381592, + "learning_rate": 3.3891705064428627e-06, + "loss": 3.1702, + "step": 18409 + }, + { + "epoch": 1.4759175001503277, + "grad_norm": 1.7670620679855347, + "learning_rate": 3.388196320695646e-06, + "loss": 3.1783, + "step": 18410 + }, + { + "epoch": 1.4759976749313504, + "grad_norm": 1.8204008340835571, + "learning_rate": 3.3872222464211126e-06, + "loss": 3.1272, + "step": 18411 + }, + { + "epoch": 1.476077849712373, + "grad_norm": 1.7412008047103882, + "learning_rate": 3.386248283635678e-06, + "loss": 3.0638, + "step": 18412 + }, + { + "epoch": 1.4761580244933956, + "grad_norm": 1.745182991027832, + "learning_rate": 3.3852744323557674e-06, + "loss": 3.1213, + "step": 18413 + }, + { + "epoch": 1.4762381992744182, + "grad_norm": 1.8425025939941406, + "learning_rate": 3.3843006925977972e-06, + "loss": 3.1934, + "step": 18414 + }, + { + "epoch": 1.4763183740554409, + "grad_norm": 0.6869921684265137, + "learning_rate": 3.383327064378187e-06, + "loss": 2.4698, + "step": 18415 + }, + { + "epoch": 1.4763985488364635, + "grad_norm": 1.8045670986175537, + "learning_rate": 3.382353547713353e-06, + "loss": 3.2546, + "step": 18416 + }, + { + "epoch": 1.476478723617486, + "grad_norm": 1.7284334897994995, + "learning_rate": 3.3813801426197034e-06, + "loss": 3.0926, + "step": 18417 + }, + { + "epoch": 1.4765588983985087, + "grad_norm": 1.9405626058578491, + "learning_rate": 3.380406849113651e-06, + "loss": 3.1709, + "step": 18418 + }, + { + "epoch": 1.4766390731795314, + "grad_norm": 1.7564011812210083, + "learning_rate": 3.3794336672116067e-06, + "loss": 3.2098, + "step": 18419 + }, + { + "epoch": 1.476719247960554, + "grad_norm": 1.8367425203323364, + "learning_rate": 3.378460596929978e-06, + "loss": 3.2222, + "step": 18420 + }, + { + "epoch": 1.4767994227415766, + "grad_norm": 1.710524320602417, + "learning_rate": 3.377487638285173e-06, + "loss": 3.1886, + "step": 18421 + }, + { + "epoch": 1.4768795975225992, + "grad_norm": 1.7724581956863403, + "learning_rate": 3.376514791293588e-06, + "loss": 3.0825, + "step": 18422 + }, + { + "epoch": 1.4769597723036219, + "grad_norm": 1.7489160299301147, + "learning_rate": 3.37554205597163e-06, + "loss": 3.1078, + "step": 18423 + }, + { + "epoch": 1.4770399470846445, + "grad_norm": 1.8377439975738525, + "learning_rate": 3.3745694323356994e-06, + "loss": 3.2389, + "step": 18424 + }, + { + "epoch": 1.477120121865667, + "grad_norm": 1.7906852960586548, + "learning_rate": 3.373596920402191e-06, + "loss": 3.1603, + "step": 18425 + }, + { + "epoch": 1.4772002966466897, + "grad_norm": 1.7342535257339478, + "learning_rate": 3.3726245201875074e-06, + "loss": 3.1208, + "step": 18426 + }, + { + "epoch": 1.4772804714277123, + "grad_norm": 1.783484935760498, + "learning_rate": 3.371652231708036e-06, + "loss": 3.2348, + "step": 18427 + }, + { + "epoch": 1.477360646208735, + "grad_norm": 1.7916055917739868, + "learning_rate": 3.370680054980171e-06, + "loss": 3.2261, + "step": 18428 + }, + { + "epoch": 1.4774408209897576, + "grad_norm": 1.820356845855713, + "learning_rate": 3.3697079900203043e-06, + "loss": 3.2589, + "step": 18429 + }, + { + "epoch": 1.4775209957707802, + "grad_norm": 1.7346546649932861, + "learning_rate": 3.368736036844824e-06, + "loss": 3.1311, + "step": 18430 + }, + { + "epoch": 1.4776011705518028, + "grad_norm": 1.7768734693527222, + "learning_rate": 3.367764195470121e-06, + "loss": 3.1307, + "step": 18431 + }, + { + "epoch": 1.4776813453328255, + "grad_norm": 1.7126432657241821, + "learning_rate": 3.3667924659125727e-06, + "loss": 3.2307, + "step": 18432 + }, + { + "epoch": 1.477761520113848, + "grad_norm": 1.7885453701019287, + "learning_rate": 3.365820848188566e-06, + "loss": 3.1697, + "step": 18433 + }, + { + "epoch": 1.4778416948948707, + "grad_norm": 0.7098292708396912, + "learning_rate": 3.36484934231448e-06, + "loss": 2.4646, + "step": 18434 + }, + { + "epoch": 1.4779218696758933, + "grad_norm": 1.7283464670181274, + "learning_rate": 3.3638779483066977e-06, + "loss": 3.1601, + "step": 18435 + }, + { + "epoch": 1.478002044456916, + "grad_norm": 1.8508596420288086, + "learning_rate": 3.3629066661815957e-06, + "loss": 3.1842, + "step": 18436 + }, + { + "epoch": 1.4780822192379386, + "grad_norm": 1.7429004907608032, + "learning_rate": 3.361935495955546e-06, + "loss": 3.162, + "step": 18437 + }, + { + "epoch": 1.4781623940189612, + "grad_norm": 1.7537665367126465, + "learning_rate": 3.3609644376449247e-06, + "loss": 3.1729, + "step": 18438 + }, + { + "epoch": 1.4782425687999838, + "grad_norm": 1.750641942024231, + "learning_rate": 3.3599934912661024e-06, + "loss": 3.2352, + "step": 18439 + }, + { + "epoch": 1.4783227435810065, + "grad_norm": 1.8190375566482544, + "learning_rate": 3.35902265683545e-06, + "loss": 3.1398, + "step": 18440 + }, + { + "epoch": 1.4784029183620293, + "grad_norm": 1.7678561210632324, + "learning_rate": 3.3580519343693386e-06, + "loss": 3.2056, + "step": 18441 + }, + { + "epoch": 1.478483093143052, + "grad_norm": 1.729114055633545, + "learning_rate": 3.357081323884127e-06, + "loss": 3.0975, + "step": 18442 + }, + { + "epoch": 1.4785632679240746, + "grad_norm": 1.7776821851730347, + "learning_rate": 3.356110825396183e-06, + "loss": 3.0723, + "step": 18443 + }, + { + "epoch": 1.4786434427050972, + "grad_norm": 1.8738173246383667, + "learning_rate": 3.355140438921869e-06, + "loss": 3.2123, + "step": 18444 + }, + { + "epoch": 1.4787236174861198, + "grad_norm": 1.7502049207687378, + "learning_rate": 3.354170164477546e-06, + "loss": 3.0823, + "step": 18445 + }, + { + "epoch": 1.4788037922671424, + "grad_norm": 1.7843258380889893, + "learning_rate": 3.3532000020795742e-06, + "loss": 3.1893, + "step": 18446 + }, + { + "epoch": 1.478883967048165, + "grad_norm": 1.7564374208450317, + "learning_rate": 3.352229951744306e-06, + "loss": 3.1411, + "step": 18447 + }, + { + "epoch": 1.4789641418291877, + "grad_norm": 1.7755558490753174, + "learning_rate": 3.3512600134880956e-06, + "loss": 3.1534, + "step": 18448 + }, + { + "epoch": 1.4790443166102103, + "grad_norm": 1.8218960762023926, + "learning_rate": 3.3502901873273e-06, + "loss": 3.1189, + "step": 18449 + }, + { + "epoch": 1.479124491391233, + "grad_norm": 0.6615951657295227, + "learning_rate": 3.349320473278268e-06, + "loss": 2.463, + "step": 18450 + }, + { + "epoch": 1.4792046661722555, + "grad_norm": 1.8451027870178223, + "learning_rate": 3.348350871357352e-06, + "loss": 3.2267, + "step": 18451 + }, + { + "epoch": 1.4792848409532782, + "grad_norm": 1.8430354595184326, + "learning_rate": 3.347381381580892e-06, + "loss": 3.1475, + "step": 18452 + }, + { + "epoch": 1.4793650157343008, + "grad_norm": 1.814629316329956, + "learning_rate": 3.346412003965238e-06, + "loss": 3.2068, + "step": 18453 + }, + { + "epoch": 1.4794451905153234, + "grad_norm": 1.7529151439666748, + "learning_rate": 3.345442738526733e-06, + "loss": 3.1791, + "step": 18454 + }, + { + "epoch": 1.479525365296346, + "grad_norm": 1.7450084686279297, + "learning_rate": 3.3444735852817178e-06, + "loss": 3.2312, + "step": 18455 + }, + { + "epoch": 1.4796055400773687, + "grad_norm": 1.8210738897323608, + "learning_rate": 3.343504544246535e-06, + "loss": 3.2418, + "step": 18456 + }, + { + "epoch": 1.4796857148583913, + "grad_norm": 1.7974883317947388, + "learning_rate": 3.3425356154375165e-06, + "loss": 3.2458, + "step": 18457 + }, + { + "epoch": 1.479765889639414, + "grad_norm": 1.817978858947754, + "learning_rate": 3.3415667988710022e-06, + "loss": 3.1197, + "step": 18458 + }, + { + "epoch": 1.4798460644204365, + "grad_norm": 1.7915745973587036, + "learning_rate": 3.3405980945633242e-06, + "loss": 3.0984, + "step": 18459 + }, + { + "epoch": 1.4799262392014592, + "grad_norm": 1.8123923540115356, + "learning_rate": 3.3396295025308155e-06, + "loss": 3.2042, + "step": 18460 + }, + { + "epoch": 1.4800064139824818, + "grad_norm": 1.7404906749725342, + "learning_rate": 3.338661022789809e-06, + "loss": 3.2182, + "step": 18461 + }, + { + "epoch": 1.4800865887635044, + "grad_norm": 1.7883859872817993, + "learning_rate": 3.337692655356627e-06, + "loss": 3.1737, + "step": 18462 + }, + { + "epoch": 1.480166763544527, + "grad_norm": 1.8504277467727661, + "learning_rate": 3.3367244002475985e-06, + "loss": 3.1483, + "step": 18463 + }, + { + "epoch": 1.4802469383255497, + "grad_norm": 1.8144385814666748, + "learning_rate": 3.335756257479049e-06, + "loss": 3.1398, + "step": 18464 + }, + { + "epoch": 1.4803271131065723, + "grad_norm": 1.7276993989944458, + "learning_rate": 3.3347882270672993e-06, + "loss": 3.0783, + "step": 18465 + }, + { + "epoch": 1.480407287887595, + "grad_norm": 1.7560144662857056, + "learning_rate": 3.3338203090286746e-06, + "loss": 3.1703, + "step": 18466 + }, + { + "epoch": 1.4804874626686175, + "grad_norm": 1.8019930124282837, + "learning_rate": 3.3328525033794878e-06, + "loss": 3.1822, + "step": 18467 + }, + { + "epoch": 1.4805676374496402, + "grad_norm": 1.7913609743118286, + "learning_rate": 3.3318848101360568e-06, + "loss": 3.2452, + "step": 18468 + }, + { + "epoch": 1.4806478122306628, + "grad_norm": 1.9565248489379883, + "learning_rate": 3.3309172293146984e-06, + "loss": 3.1909, + "step": 18469 + }, + { + "epoch": 1.4807279870116854, + "grad_norm": 1.8342963457107544, + "learning_rate": 3.3299497609317254e-06, + "loss": 3.1828, + "step": 18470 + }, + { + "epoch": 1.480808161792708, + "grad_norm": 1.7212145328521729, + "learning_rate": 3.3289824050034515e-06, + "loss": 3.2715, + "step": 18471 + }, + { + "epoch": 1.4808883365737309, + "grad_norm": 1.792004942893982, + "learning_rate": 3.32801516154618e-06, + "loss": 3.2443, + "step": 18472 + }, + { + "epoch": 1.4809685113547535, + "grad_norm": 1.845473289489746, + "learning_rate": 3.327048030576222e-06, + "loss": 3.0892, + "step": 18473 + }, + { + "epoch": 1.4810486861357761, + "grad_norm": 1.7621687650680542, + "learning_rate": 3.3260810121098817e-06, + "loss": 3.0288, + "step": 18474 + }, + { + "epoch": 1.4811288609167987, + "grad_norm": 1.798252820968628, + "learning_rate": 3.325114106163463e-06, + "loss": 3.1827, + "step": 18475 + }, + { + "epoch": 1.4812090356978214, + "grad_norm": 1.885419487953186, + "learning_rate": 3.324147312753272e-06, + "loss": 3.1223, + "step": 18476 + }, + { + "epoch": 1.481289210478844, + "grad_norm": 1.8810908794403076, + "learning_rate": 3.3231806318956018e-06, + "loss": 3.2113, + "step": 18477 + }, + { + "epoch": 1.4813693852598666, + "grad_norm": 1.7205709218978882, + "learning_rate": 3.3222140636067513e-06, + "loss": 3.1113, + "step": 18478 + }, + { + "epoch": 1.4814495600408892, + "grad_norm": 1.8203604221343994, + "learning_rate": 3.3212476079030198e-06, + "loss": 3.1877, + "step": 18479 + }, + { + "epoch": 1.4815297348219119, + "grad_norm": 1.8501816987991333, + "learning_rate": 3.320281264800699e-06, + "loss": 3.2812, + "step": 18480 + }, + { + "epoch": 1.4816099096029345, + "grad_norm": 1.8577815294265747, + "learning_rate": 3.319315034316085e-06, + "loss": 3.2465, + "step": 18481 + }, + { + "epoch": 1.481690084383957, + "grad_norm": 0.6725439429283142, + "learning_rate": 3.3183489164654624e-06, + "loss": 2.4667, + "step": 18482 + }, + { + "epoch": 1.4817702591649797, + "grad_norm": 1.7451927661895752, + "learning_rate": 3.3173829112651223e-06, + "loss": 3.1273, + "step": 18483 + }, + { + "epoch": 1.4818504339460024, + "grad_norm": 1.8680881261825562, + "learning_rate": 3.3164170187313505e-06, + "loss": 3.0824, + "step": 18484 + }, + { + "epoch": 1.481930608727025, + "grad_norm": 1.6806081533432007, + "learning_rate": 3.315451238880433e-06, + "loss": 3.1394, + "step": 18485 + }, + { + "epoch": 1.4820107835080476, + "grad_norm": 1.7700461149215698, + "learning_rate": 3.314485571728654e-06, + "loss": 3.1301, + "step": 18486 + }, + { + "epoch": 1.4820909582890702, + "grad_norm": 0.6624699831008911, + "learning_rate": 3.31352001729229e-06, + "loss": 2.4481, + "step": 18487 + }, + { + "epoch": 1.4821711330700928, + "grad_norm": 0.6709625720977783, + "learning_rate": 3.312554575587621e-06, + "loss": 2.4761, + "step": 18488 + }, + { + "epoch": 1.4822513078511155, + "grad_norm": 1.7608245611190796, + "learning_rate": 3.311589246630925e-06, + "loss": 3.1721, + "step": 18489 + }, + { + "epoch": 1.482331482632138, + "grad_norm": 1.8144720792770386, + "learning_rate": 3.3106240304384774e-06, + "loss": 3.2119, + "step": 18490 + }, + { + "epoch": 1.4824116574131607, + "grad_norm": 1.8117892742156982, + "learning_rate": 3.3096589270265545e-06, + "loss": 3.1978, + "step": 18491 + }, + { + "epoch": 1.4824918321941833, + "grad_norm": 1.8095104694366455, + "learning_rate": 3.308693936411421e-06, + "loss": 3.0903, + "step": 18492 + }, + { + "epoch": 1.482572006975206, + "grad_norm": 1.720045804977417, + "learning_rate": 3.307729058609349e-06, + "loss": 3.2182, + "step": 18493 + }, + { + "epoch": 1.4826521817562286, + "grad_norm": 1.8382848501205444, + "learning_rate": 3.3067642936366073e-06, + "loss": 3.2122, + "step": 18494 + }, + { + "epoch": 1.4827323565372512, + "grad_norm": 1.701171875, + "learning_rate": 3.3057996415094594e-06, + "loss": 3.1425, + "step": 18495 + }, + { + "epoch": 1.4828125313182738, + "grad_norm": 1.7889717817306519, + "learning_rate": 3.3048351022441714e-06, + "loss": 3.2264, + "step": 18496 + }, + { + "epoch": 1.4828927060992965, + "grad_norm": 1.8207391500473022, + "learning_rate": 3.303870675857007e-06, + "loss": 3.1612, + "step": 18497 + }, + { + "epoch": 1.482972880880319, + "grad_norm": 0.6558117270469666, + "learning_rate": 3.30290636236422e-06, + "loss": 2.4606, + "step": 18498 + }, + { + "epoch": 1.4830530556613417, + "grad_norm": 1.799277901649475, + "learning_rate": 3.3019421617820702e-06, + "loss": 3.148, + "step": 18499 + }, + { + "epoch": 1.4831332304423643, + "grad_norm": 1.7633671760559082, + "learning_rate": 3.300978074126816e-06, + "loss": 3.2224, + "step": 18500 + }, + { + "epoch": 1.483213405223387, + "grad_norm": 1.7338972091674805, + "learning_rate": 3.30001409941471e-06, + "loss": 3.1417, + "step": 18501 + }, + { + "epoch": 1.4832935800044096, + "grad_norm": 1.8030275106430054, + "learning_rate": 3.2990502376620093e-06, + "loss": 3.2404, + "step": 18502 + }, + { + "epoch": 1.4833737547854322, + "grad_norm": 1.6797802448272705, + "learning_rate": 3.298086488884955e-06, + "loss": 3.1462, + "step": 18503 + }, + { + "epoch": 1.4834539295664548, + "grad_norm": 1.8015267848968506, + "learning_rate": 3.2971228530998025e-06, + "loss": 3.1902, + "step": 18504 + }, + { + "epoch": 1.4835341043474775, + "grad_norm": 1.7632089853286743, + "learning_rate": 3.2961593303227945e-06, + "loss": 3.1465, + "step": 18505 + }, + { + "epoch": 1.4836142791285, + "grad_norm": 1.7700660228729248, + "learning_rate": 3.295195920570179e-06, + "loss": 3.1275, + "step": 18506 + }, + { + "epoch": 1.4836944539095227, + "grad_norm": 1.6973243951797485, + "learning_rate": 3.294232623858199e-06, + "loss": 3.1286, + "step": 18507 + }, + { + "epoch": 1.4837746286905453, + "grad_norm": 1.7899543046951294, + "learning_rate": 3.293269440203092e-06, + "loss": 3.1814, + "step": 18508 + }, + { + "epoch": 1.483854803471568, + "grad_norm": 1.782127022743225, + "learning_rate": 3.292306369621098e-06, + "loss": 3.166, + "step": 18509 + }, + { + "epoch": 1.4839349782525906, + "grad_norm": 1.8374162912368774, + "learning_rate": 3.291343412128454e-06, + "loss": 3.1958, + "step": 18510 + }, + { + "epoch": 1.4840151530336132, + "grad_norm": 1.7368862628936768, + "learning_rate": 3.290380567741396e-06, + "loss": 3.1613, + "step": 18511 + }, + { + "epoch": 1.4840953278146358, + "grad_norm": 1.6901555061340332, + "learning_rate": 3.2894178364761597e-06, + "loss": 3.1125, + "step": 18512 + }, + { + "epoch": 1.4841755025956584, + "grad_norm": 1.860036849975586, + "learning_rate": 3.2884552183489716e-06, + "loss": 3.1519, + "step": 18513 + }, + { + "epoch": 1.484255677376681, + "grad_norm": 1.7282780408859253, + "learning_rate": 3.2874927133760627e-06, + "loss": 3.1575, + "step": 18514 + }, + { + "epoch": 1.4843358521577037, + "grad_norm": 1.8338948488235474, + "learning_rate": 3.28653032157366e-06, + "loss": 3.1348, + "step": 18515 + }, + { + "epoch": 1.4844160269387263, + "grad_norm": 1.716185212135315, + "learning_rate": 3.2855680429579916e-06, + "loss": 3.1927, + "step": 18516 + }, + { + "epoch": 1.484496201719749, + "grad_norm": 0.6760298609733582, + "learning_rate": 3.2846058775452816e-06, + "loss": 2.4825, + "step": 18517 + }, + { + "epoch": 1.4845763765007716, + "grad_norm": 1.7845481634140015, + "learning_rate": 3.2836438253517468e-06, + "loss": 3.1564, + "step": 18518 + }, + { + "epoch": 1.4846565512817942, + "grad_norm": 0.6922516822814941, + "learning_rate": 3.2826818863936104e-06, + "loss": 2.4633, + "step": 18519 + }, + { + "epoch": 1.4847367260628168, + "grad_norm": 1.7262282371520996, + "learning_rate": 3.281720060687089e-06, + "loss": 3.2399, + "step": 18520 + }, + { + "epoch": 1.4848169008438394, + "grad_norm": 1.7554348707199097, + "learning_rate": 3.2807583482484006e-06, + "loss": 3.0701, + "step": 18521 + }, + { + "epoch": 1.484897075624862, + "grad_norm": 0.6634209752082825, + "learning_rate": 3.279796749093762e-06, + "loss": 2.4855, + "step": 18522 + }, + { + "epoch": 1.484977250405885, + "grad_norm": 1.6947197914123535, + "learning_rate": 3.2788352632393783e-06, + "loss": 3.1475, + "step": 18523 + }, + { + "epoch": 1.4850574251869075, + "grad_norm": 1.676193356513977, + "learning_rate": 3.277873890701463e-06, + "loss": 3.1024, + "step": 18524 + }, + { + "epoch": 1.4851375999679302, + "grad_norm": 1.869987964630127, + "learning_rate": 3.276912631496224e-06, + "loss": 3.1017, + "step": 18525 + }, + { + "epoch": 1.4852177747489528, + "grad_norm": 1.7434524297714233, + "learning_rate": 3.2759514856398687e-06, + "loss": 3.1739, + "step": 18526 + }, + { + "epoch": 1.4852979495299754, + "grad_norm": 1.8308660984039307, + "learning_rate": 3.2749904531486064e-06, + "loss": 3.1427, + "step": 18527 + }, + { + "epoch": 1.485378124310998, + "grad_norm": 1.745116949081421, + "learning_rate": 3.2740295340386307e-06, + "loss": 3.1236, + "step": 18528 + }, + { + "epoch": 1.4854582990920207, + "grad_norm": 1.7051217555999756, + "learning_rate": 3.2730687283261463e-06, + "loss": 3.097, + "step": 18529 + }, + { + "epoch": 1.4855384738730433, + "grad_norm": 1.8147282600402832, + "learning_rate": 3.2721080360273527e-06, + "loss": 3.1395, + "step": 18530 + }, + { + "epoch": 1.485618648654066, + "grad_norm": 1.8043030500411987, + "learning_rate": 3.271147457158447e-06, + "loss": 3.2159, + "step": 18531 + }, + { + "epoch": 1.4856988234350885, + "grad_norm": 1.8621746301651, + "learning_rate": 3.2701869917356256e-06, + "loss": 3.2441, + "step": 18532 + }, + { + "epoch": 1.4857789982161111, + "grad_norm": 1.6801077127456665, + "learning_rate": 3.2692266397750772e-06, + "loss": 3.0984, + "step": 18533 + }, + { + "epoch": 1.4858591729971338, + "grad_norm": 1.7458628416061401, + "learning_rate": 3.2682664012929956e-06, + "loss": 3.1441, + "step": 18534 + }, + { + "epoch": 1.4859393477781564, + "grad_norm": 1.7981665134429932, + "learning_rate": 3.2673062763055696e-06, + "loss": 3.2776, + "step": 18535 + }, + { + "epoch": 1.486019522559179, + "grad_norm": 1.6944507360458374, + "learning_rate": 3.266346264828988e-06, + "loss": 3.0776, + "step": 18536 + }, + { + "epoch": 1.4860996973402016, + "grad_norm": 1.812760353088379, + "learning_rate": 3.265386366879437e-06, + "loss": 3.2046, + "step": 18537 + }, + { + "epoch": 1.4861798721212243, + "grad_norm": 1.7949562072753906, + "learning_rate": 3.264426582473096e-06, + "loss": 3.1663, + "step": 18538 + }, + { + "epoch": 1.486260046902247, + "grad_norm": 1.7488667964935303, + "learning_rate": 3.2634669116261485e-06, + "loss": 3.1674, + "step": 18539 + }, + { + "epoch": 1.4863402216832695, + "grad_norm": 1.7808325290679932, + "learning_rate": 3.2625073543547746e-06, + "loss": 3.1763, + "step": 18540 + }, + { + "epoch": 1.4864203964642921, + "grad_norm": 1.7427983283996582, + "learning_rate": 3.2615479106751526e-06, + "loss": 3.1086, + "step": 18541 + }, + { + "epoch": 1.4865005712453148, + "grad_norm": 1.7194228172302246, + "learning_rate": 3.260588580603461e-06, + "loss": 3.1813, + "step": 18542 + }, + { + "epoch": 1.4865807460263374, + "grad_norm": 1.7283003330230713, + "learning_rate": 3.2596293641558684e-06, + "loss": 3.1469, + "step": 18543 + }, + { + "epoch": 1.48666092080736, + "grad_norm": 1.8104381561279297, + "learning_rate": 3.2586702613485475e-06, + "loss": 3.2906, + "step": 18544 + }, + { + "epoch": 1.4867410955883826, + "grad_norm": 1.7686179876327515, + "learning_rate": 3.257711272197671e-06, + "loss": 3.136, + "step": 18545 + }, + { + "epoch": 1.4868212703694053, + "grad_norm": 1.7854124307632446, + "learning_rate": 3.256752396719407e-06, + "loss": 3.2041, + "step": 18546 + }, + { + "epoch": 1.4869014451504279, + "grad_norm": 1.8090286254882812, + "learning_rate": 3.2557936349299234e-06, + "loss": 3.1609, + "step": 18547 + }, + { + "epoch": 1.4869816199314505, + "grad_norm": 1.9040709733963013, + "learning_rate": 3.254834986845379e-06, + "loss": 3.3068, + "step": 18548 + }, + { + "epoch": 1.4870617947124731, + "grad_norm": 1.6746259927749634, + "learning_rate": 3.25387645248194e-06, + "loss": 3.2715, + "step": 18549 + }, + { + "epoch": 1.4871419694934958, + "grad_norm": 1.7757238149642944, + "learning_rate": 3.2529180318557664e-06, + "loss": 3.1871, + "step": 18550 + }, + { + "epoch": 1.4872221442745184, + "grad_norm": 1.726243019104004, + "learning_rate": 3.251959724983016e-06, + "loss": 3.1719, + "step": 18551 + }, + { + "epoch": 1.487302319055541, + "grad_norm": 1.835547685623169, + "learning_rate": 3.2510015318798504e-06, + "loss": 3.1617, + "step": 18552 + }, + { + "epoch": 1.4873824938365638, + "grad_norm": 1.700911283493042, + "learning_rate": 3.250043452562417e-06, + "loss": 3.2084, + "step": 18553 + }, + { + "epoch": 1.4874626686175865, + "grad_norm": 1.7486872673034668, + "learning_rate": 3.2490854870468713e-06, + "loss": 3.2123, + "step": 18554 + }, + { + "epoch": 1.487542843398609, + "grad_norm": 1.8438693284988403, + "learning_rate": 3.248127635349366e-06, + "loss": 3.3051, + "step": 18555 + }, + { + "epoch": 1.4876230181796317, + "grad_norm": 1.8178843259811401, + "learning_rate": 3.2471698974860487e-06, + "loss": 3.209, + "step": 18556 + }, + { + "epoch": 1.4877031929606543, + "grad_norm": 1.7941431999206543, + "learning_rate": 3.2462122734730707e-06, + "loss": 3.1572, + "step": 18557 + }, + { + "epoch": 1.487783367741677, + "grad_norm": 1.675688624382019, + "learning_rate": 3.2452547633265706e-06, + "loss": 3.2158, + "step": 18558 + }, + { + "epoch": 1.4878635425226996, + "grad_norm": 1.7601696252822876, + "learning_rate": 3.244297367062693e-06, + "loss": 3.088, + "step": 18559 + }, + { + "epoch": 1.4879437173037222, + "grad_norm": 0.6527132391929626, + "learning_rate": 3.2433400846975825e-06, + "loss": 2.4666, + "step": 18560 + }, + { + "epoch": 1.4880238920847448, + "grad_norm": 1.6839510202407837, + "learning_rate": 3.2423829162473763e-06, + "loss": 3.1174, + "step": 18561 + }, + { + "epoch": 1.4881040668657675, + "grad_norm": 1.662503957748413, + "learning_rate": 3.2414258617282155e-06, + "loss": 3.2362, + "step": 18562 + }, + { + "epoch": 1.48818424164679, + "grad_norm": 1.889715313911438, + "learning_rate": 3.24046892115623e-06, + "loss": 3.1265, + "step": 18563 + }, + { + "epoch": 1.4882644164278127, + "grad_norm": 1.791388750076294, + "learning_rate": 3.2395120945475557e-06, + "loss": 3.192, + "step": 18564 + }, + { + "epoch": 1.4883445912088353, + "grad_norm": 1.7882423400878906, + "learning_rate": 3.2385553819183248e-06, + "loss": 3.2059, + "step": 18565 + }, + { + "epoch": 1.488424765989858, + "grad_norm": 1.7537059783935547, + "learning_rate": 3.237598783284668e-06, + "loss": 3.172, + "step": 18566 + }, + { + "epoch": 1.4885049407708806, + "grad_norm": 1.870471477508545, + "learning_rate": 3.2366422986627143e-06, + "loss": 3.1841, + "step": 18567 + }, + { + "epoch": 1.4885851155519032, + "grad_norm": 1.7410988807678223, + "learning_rate": 3.235685928068586e-06, + "loss": 3.1645, + "step": 18568 + }, + { + "epoch": 1.4886652903329258, + "grad_norm": 1.8488529920578003, + "learning_rate": 3.2347296715184086e-06, + "loss": 3.2511, + "step": 18569 + }, + { + "epoch": 1.4887454651139485, + "grad_norm": 1.7620693445205688, + "learning_rate": 3.2337735290283047e-06, + "loss": 3.2964, + "step": 18570 + }, + { + "epoch": 1.488825639894971, + "grad_norm": 1.8433464765548706, + "learning_rate": 3.2328175006143934e-06, + "loss": 3.1498, + "step": 18571 + }, + { + "epoch": 1.4889058146759937, + "grad_norm": 1.720755696296692, + "learning_rate": 3.2318615862927982e-06, + "loss": 3.135, + "step": 18572 + }, + { + "epoch": 1.4889859894570163, + "grad_norm": 1.7953851222991943, + "learning_rate": 3.2309057860796277e-06, + "loss": 3.1854, + "step": 18573 + }, + { + "epoch": 1.489066164238039, + "grad_norm": 1.678005337715149, + "learning_rate": 3.2299500999910005e-06, + "loss": 3.1349, + "step": 18574 + }, + { + "epoch": 1.4891463390190616, + "grad_norm": 1.848921537399292, + "learning_rate": 3.228994528043028e-06, + "loss": 3.0959, + "step": 18575 + }, + { + "epoch": 1.4892265138000842, + "grad_norm": 1.7472407817840576, + "learning_rate": 3.2280390702518215e-06, + "loss": 3.0665, + "step": 18576 + }, + { + "epoch": 1.4893066885811068, + "grad_norm": 1.6877013444900513, + "learning_rate": 3.227083726633493e-06, + "loss": 3.0699, + "step": 18577 + }, + { + "epoch": 1.4893868633621294, + "grad_norm": 0.6884402632713318, + "learning_rate": 3.226128497204143e-06, + "loss": 2.4652, + "step": 18578 + }, + { + "epoch": 1.489467038143152, + "grad_norm": 1.70298171043396, + "learning_rate": 3.225173381979879e-06, + "loss": 3.1642, + "step": 18579 + }, + { + "epoch": 1.4895472129241747, + "grad_norm": 1.7899937629699707, + "learning_rate": 3.2242183809768036e-06, + "loss": 3.1432, + "step": 18580 + }, + { + "epoch": 1.4896273877051973, + "grad_norm": 1.7110731601715088, + "learning_rate": 3.2232634942110184e-06, + "loss": 3.1523, + "step": 18581 + }, + { + "epoch": 1.48970756248622, + "grad_norm": 1.8228249549865723, + "learning_rate": 3.222308721698627e-06, + "loss": 3.146, + "step": 18582 + }, + { + "epoch": 1.4897877372672426, + "grad_norm": 0.6607210636138916, + "learning_rate": 3.2213540634557174e-06, + "loss": 2.4613, + "step": 18583 + }, + { + "epoch": 1.4898679120482652, + "grad_norm": 1.6953221559524536, + "learning_rate": 3.2203995194983893e-06, + "loss": 3.2252, + "step": 18584 + }, + { + "epoch": 1.4899480868292878, + "grad_norm": 1.6535035371780396, + "learning_rate": 3.2194450898427355e-06, + "loss": 3.1487, + "step": 18585 + }, + { + "epoch": 1.4900282616103104, + "grad_norm": 1.8777711391448975, + "learning_rate": 3.218490774504849e-06, + "loss": 3.1415, + "step": 18586 + }, + { + "epoch": 1.490108436391333, + "grad_norm": 1.8484567403793335, + "learning_rate": 3.21753657350082e-06, + "loss": 3.2835, + "step": 18587 + }, + { + "epoch": 1.4901886111723557, + "grad_norm": 1.869797945022583, + "learning_rate": 3.216582486846731e-06, + "loss": 3.1065, + "step": 18588 + }, + { + "epoch": 1.4902687859533783, + "grad_norm": 1.8125108480453491, + "learning_rate": 3.215628514558671e-06, + "loss": 3.1825, + "step": 18589 + }, + { + "epoch": 1.490348960734401, + "grad_norm": 1.8929613828659058, + "learning_rate": 3.214674656652723e-06, + "loss": 3.1653, + "step": 18590 + }, + { + "epoch": 1.4904291355154236, + "grad_norm": 1.8312455415725708, + "learning_rate": 3.2137209131449686e-06, + "loss": 3.2976, + "step": 18591 + }, + { + "epoch": 1.4905093102964462, + "grad_norm": 1.8854185342788696, + "learning_rate": 3.2127672840514923e-06, + "loss": 3.1696, + "step": 18592 + }, + { + "epoch": 1.4905894850774688, + "grad_norm": 1.8643784523010254, + "learning_rate": 3.2118137693883635e-06, + "loss": 3.2997, + "step": 18593 + }, + { + "epoch": 1.4906696598584914, + "grad_norm": 1.7774670124053955, + "learning_rate": 3.210860369171661e-06, + "loss": 3.1467, + "step": 18594 + }, + { + "epoch": 1.490749834639514, + "grad_norm": 1.7246277332305908, + "learning_rate": 3.209907083417462e-06, + "loss": 3.3011, + "step": 18595 + }, + { + "epoch": 1.4908300094205367, + "grad_norm": 1.8073903322219849, + "learning_rate": 3.2089539121418355e-06, + "loss": 3.1344, + "step": 18596 + }, + { + "epoch": 1.4909101842015593, + "grad_norm": 1.728348731994629, + "learning_rate": 3.2080008553608554e-06, + "loss": 3.1282, + "step": 18597 + }, + { + "epoch": 1.490990358982582, + "grad_norm": 1.8210787773132324, + "learning_rate": 3.207047913090585e-06, + "loss": 3.2411, + "step": 18598 + }, + { + "epoch": 1.4910705337636045, + "grad_norm": 1.715754747390747, + "learning_rate": 3.2060950853470916e-06, + "loss": 3.0952, + "step": 18599 + }, + { + "epoch": 1.4911507085446272, + "grad_norm": 1.7907536029815674, + "learning_rate": 3.2051423721464414e-06, + "loss": 3.0638, + "step": 18600 + }, + { + "epoch": 1.4912308833256498, + "grad_norm": 1.8124431371688843, + "learning_rate": 3.2041897735046955e-06, + "loss": 3.1238, + "step": 18601 + }, + { + "epoch": 1.4913110581066724, + "grad_norm": 1.7837632894515991, + "learning_rate": 3.203237289437919e-06, + "loss": 3.1219, + "step": 18602 + }, + { + "epoch": 1.491391232887695, + "grad_norm": 1.7566919326782227, + "learning_rate": 3.202284919962163e-06, + "loss": 3.0366, + "step": 18603 + }, + { + "epoch": 1.4914714076687179, + "grad_norm": 1.7783825397491455, + "learning_rate": 3.2013326650934874e-06, + "loss": 3.1268, + "step": 18604 + }, + { + "epoch": 1.4915515824497405, + "grad_norm": 1.700212836265564, + "learning_rate": 3.200380524847947e-06, + "loss": 3.1043, + "step": 18605 + }, + { + "epoch": 1.4916317572307631, + "grad_norm": 1.7054270505905151, + "learning_rate": 3.1994284992415937e-06, + "loss": 3.2244, + "step": 18606 + }, + { + "epoch": 1.4917119320117858, + "grad_norm": 1.7940430641174316, + "learning_rate": 3.1984765882904833e-06, + "loss": 3.1752, + "step": 18607 + }, + { + "epoch": 1.4917921067928084, + "grad_norm": 1.7712496519088745, + "learning_rate": 3.1975247920106565e-06, + "loss": 3.1525, + "step": 18608 + }, + { + "epoch": 1.491872281573831, + "grad_norm": 1.7135889530181885, + "learning_rate": 3.196573110418165e-06, + "loss": 3.0992, + "step": 18609 + }, + { + "epoch": 1.4919524563548536, + "grad_norm": 1.7787877321243286, + "learning_rate": 3.1956215435290537e-06, + "loss": 3.101, + "step": 18610 + }, + { + "epoch": 1.4920326311358763, + "grad_norm": 1.849980115890503, + "learning_rate": 3.1946700913593644e-06, + "loss": 3.1977, + "step": 18611 + }, + { + "epoch": 1.4921128059168989, + "grad_norm": 1.8641997575759888, + "learning_rate": 3.1937187539251414e-06, + "loss": 3.2144, + "step": 18612 + }, + { + "epoch": 1.4921929806979215, + "grad_norm": 1.730129599571228, + "learning_rate": 3.192767531242419e-06, + "loss": 3.0805, + "step": 18613 + }, + { + "epoch": 1.4922731554789441, + "grad_norm": 1.74764084815979, + "learning_rate": 3.1918164233272364e-06, + "loss": 3.2455, + "step": 18614 + }, + { + "epoch": 1.4923533302599667, + "grad_norm": 1.8005764484405518, + "learning_rate": 3.19086543019563e-06, + "loss": 3.2272, + "step": 18615 + }, + { + "epoch": 1.4924335050409894, + "grad_norm": 1.7489326000213623, + "learning_rate": 3.189914551863633e-06, + "loss": 3.2084, + "step": 18616 + }, + { + "epoch": 1.492513679822012, + "grad_norm": 1.687670350074768, + "learning_rate": 3.188963788347279e-06, + "loss": 3.1837, + "step": 18617 + }, + { + "epoch": 1.4925938546030346, + "grad_norm": 0.6889973878860474, + "learning_rate": 3.1880131396625925e-06, + "loss": 2.4728, + "step": 18618 + }, + { + "epoch": 1.4926740293840572, + "grad_norm": 1.7377182245254517, + "learning_rate": 3.1870626058256036e-06, + "loss": 3.1697, + "step": 18619 + }, + { + "epoch": 1.4927542041650799, + "grad_norm": 1.8341385126113892, + "learning_rate": 3.186112186852337e-06, + "loss": 3.1575, + "step": 18620 + }, + { + "epoch": 1.4928343789461025, + "grad_norm": 1.7084473371505737, + "learning_rate": 3.185161882758819e-06, + "loss": 3.1371, + "step": 18621 + }, + { + "epoch": 1.4929145537271251, + "grad_norm": 1.8250151872634888, + "learning_rate": 3.1842116935610724e-06, + "loss": 3.1724, + "step": 18622 + }, + { + "epoch": 1.4929947285081477, + "grad_norm": 1.787773609161377, + "learning_rate": 3.183261619275111e-06, + "loss": 3.1519, + "step": 18623 + }, + { + "epoch": 1.4930749032891704, + "grad_norm": 1.750792145729065, + "learning_rate": 3.182311659916957e-06, + "loss": 3.2056, + "step": 18624 + }, + { + "epoch": 1.493155078070193, + "grad_norm": 1.7723854780197144, + "learning_rate": 3.181361815502625e-06, + "loss": 3.1414, + "step": 18625 + }, + { + "epoch": 1.4932352528512156, + "grad_norm": 1.84732985496521, + "learning_rate": 3.18041208604813e-06, + "loss": 3.403, + "step": 18626 + }, + { + "epoch": 1.4933154276322382, + "grad_norm": 1.7881802320480347, + "learning_rate": 3.1794624715694866e-06, + "loss": 3.2581, + "step": 18627 + }, + { + "epoch": 1.4933956024132609, + "grad_norm": 1.8240816593170166, + "learning_rate": 3.178512972082699e-06, + "loss": 3.2451, + "step": 18628 + }, + { + "epoch": 1.4934757771942835, + "grad_norm": 1.7904446125030518, + "learning_rate": 3.177563587603779e-06, + "loss": 3.1448, + "step": 18629 + }, + { + "epoch": 1.493555951975306, + "grad_norm": 1.8143848180770874, + "learning_rate": 3.176614318148732e-06, + "loss": 3.152, + "step": 18630 + }, + { + "epoch": 1.4936361267563287, + "grad_norm": 1.8129968643188477, + "learning_rate": 3.1756651637335634e-06, + "loss": 3.1906, + "step": 18631 + }, + { + "epoch": 1.4937163015373514, + "grad_norm": 1.676951289176941, + "learning_rate": 3.1747161243742775e-06, + "loss": 3.1226, + "step": 18632 + }, + { + "epoch": 1.493796476318374, + "grad_norm": 1.728763461112976, + "learning_rate": 3.173767200086869e-06, + "loss": 3.1146, + "step": 18633 + }, + { + "epoch": 1.4938766510993966, + "grad_norm": 1.822780728340149, + "learning_rate": 3.1728183908873402e-06, + "loss": 3.1608, + "step": 18634 + }, + { + "epoch": 1.4939568258804194, + "grad_norm": 1.7070188522338867, + "learning_rate": 3.1718696967916872e-06, + "loss": 3.1942, + "step": 18635 + }, + { + "epoch": 1.494037000661442, + "grad_norm": 1.8293561935424805, + "learning_rate": 3.1709211178159047e-06, + "loss": 3.1926, + "step": 18636 + }, + { + "epoch": 1.4941171754424647, + "grad_norm": 1.7516772747039795, + "learning_rate": 3.1699726539759877e-06, + "loss": 3.1415, + "step": 18637 + }, + { + "epoch": 1.4941973502234873, + "grad_norm": 1.835492491722107, + "learning_rate": 3.169024305287922e-06, + "loss": 3.1967, + "step": 18638 + }, + { + "epoch": 1.49427752500451, + "grad_norm": 1.7705363035202026, + "learning_rate": 3.1680760717676993e-06, + "loss": 3.2072, + "step": 18639 + }, + { + "epoch": 1.4943576997855326, + "grad_norm": 1.865151047706604, + "learning_rate": 3.1671279534313073e-06, + "loss": 3.2787, + "step": 18640 + }, + { + "epoch": 1.4944378745665552, + "grad_norm": 1.7804837226867676, + "learning_rate": 3.166179950294728e-06, + "loss": 3.2251, + "step": 18641 + }, + { + "epoch": 1.4945180493475778, + "grad_norm": 1.756123423576355, + "learning_rate": 3.1652320623739496e-06, + "loss": 3.2092, + "step": 18642 + }, + { + "epoch": 1.4945982241286004, + "grad_norm": 1.8301846981048584, + "learning_rate": 3.164284289684948e-06, + "loss": 3.193, + "step": 18643 + }, + { + "epoch": 1.494678398909623, + "grad_norm": 1.7874715328216553, + "learning_rate": 3.163336632243703e-06, + "loss": 3.1719, + "step": 18644 + }, + { + "epoch": 1.4947585736906457, + "grad_norm": 1.7027877569198608, + "learning_rate": 3.1623890900661935e-06, + "loss": 3.0955, + "step": 18645 + }, + { + "epoch": 1.4948387484716683, + "grad_norm": 1.7969422340393066, + "learning_rate": 3.161441663168393e-06, + "loss": 3.1641, + "step": 18646 + }, + { + "epoch": 1.494918923252691, + "grad_norm": 1.9174188375473022, + "learning_rate": 3.16049435156628e-06, + "loss": 3.135, + "step": 18647 + }, + { + "epoch": 1.4949990980337136, + "grad_norm": 1.797149419784546, + "learning_rate": 3.1595471552758185e-06, + "loss": 3.1403, + "step": 18648 + }, + { + "epoch": 1.4950792728147362, + "grad_norm": 1.80536949634552, + "learning_rate": 3.158600074312982e-06, + "loss": 3.1946, + "step": 18649 + }, + { + "epoch": 1.4951594475957588, + "grad_norm": 1.7834078073501587, + "learning_rate": 3.1576531086937357e-06, + "loss": 3.1105, + "step": 18650 + }, + { + "epoch": 1.4952396223767814, + "grad_norm": 1.7590346336364746, + "learning_rate": 3.156706258434047e-06, + "loss": 3.2267, + "step": 18651 + }, + { + "epoch": 1.495319797157804, + "grad_norm": 1.877214789390564, + "learning_rate": 3.1557595235498817e-06, + "loss": 3.215, + "step": 18652 + }, + { + "epoch": 1.4953999719388267, + "grad_norm": 1.908730149269104, + "learning_rate": 3.1548129040571964e-06, + "loss": 3.3152, + "step": 18653 + }, + { + "epoch": 1.4954801467198493, + "grad_norm": 1.7896815538406372, + "learning_rate": 3.153866399971953e-06, + "loss": 3.2278, + "step": 18654 + }, + { + "epoch": 1.495560321500872, + "grad_norm": 1.7011252641677856, + "learning_rate": 3.1529200113101087e-06, + "loss": 3.1741, + "step": 18655 + }, + { + "epoch": 1.4956404962818945, + "grad_norm": 1.7137290239334106, + "learning_rate": 3.1519737380876204e-06, + "loss": 3.139, + "step": 18656 + }, + { + "epoch": 1.4957206710629172, + "grad_norm": 1.7538385391235352, + "learning_rate": 3.1510275803204437e-06, + "loss": 3.2097, + "step": 18657 + }, + { + "epoch": 1.4958008458439398, + "grad_norm": 1.659754991531372, + "learning_rate": 3.1500815380245265e-06, + "loss": 3.0811, + "step": 18658 + }, + { + "epoch": 1.4958810206249624, + "grad_norm": 1.843287706375122, + "learning_rate": 3.149135611215819e-06, + "loss": 3.2897, + "step": 18659 + }, + { + "epoch": 1.495961195405985, + "grad_norm": 1.8307279348373413, + "learning_rate": 3.1481897999102705e-06, + "loss": 3.2856, + "step": 18660 + }, + { + "epoch": 1.4960413701870077, + "grad_norm": 0.6339858174324036, + "learning_rate": 3.147244104123829e-06, + "loss": 2.463, + "step": 18661 + }, + { + "epoch": 1.4961215449680303, + "grad_norm": 0.6554408669471741, + "learning_rate": 3.1462985238724385e-06, + "loss": 2.4406, + "step": 18662 + }, + { + "epoch": 1.496201719749053, + "grad_norm": 1.8498101234436035, + "learning_rate": 3.145353059172036e-06, + "loss": 3.1342, + "step": 18663 + }, + { + "epoch": 1.4962818945300755, + "grad_norm": 1.8112785816192627, + "learning_rate": 3.144407710038566e-06, + "loss": 3.1672, + "step": 18664 + }, + { + "epoch": 1.4963620693110982, + "grad_norm": 1.7689238786697388, + "learning_rate": 3.143462476487965e-06, + "loss": 3.1816, + "step": 18665 + }, + { + "epoch": 1.4964422440921208, + "grad_norm": 1.7916847467422485, + "learning_rate": 3.1425173585361702e-06, + "loss": 3.0673, + "step": 18666 + }, + { + "epoch": 1.4965224188731434, + "grad_norm": 1.8363608121871948, + "learning_rate": 3.1415723561991195e-06, + "loss": 3.162, + "step": 18667 + }, + { + "epoch": 1.496602593654166, + "grad_norm": 1.7502548694610596, + "learning_rate": 3.1406274694927386e-06, + "loss": 3.181, + "step": 18668 + }, + { + "epoch": 1.4966827684351887, + "grad_norm": 1.747755527496338, + "learning_rate": 3.139682698432962e-06, + "loss": 3.183, + "step": 18669 + }, + { + "epoch": 1.4967629432162113, + "grad_norm": 1.7772889137268066, + "learning_rate": 3.138738043035716e-06, + "loss": 3.1603, + "step": 18670 + }, + { + "epoch": 1.496843117997234, + "grad_norm": 1.678736686706543, + "learning_rate": 3.137793503316928e-06, + "loss": 3.1619, + "step": 18671 + }, + { + "epoch": 1.4969232927782565, + "grad_norm": 1.8319027423858643, + "learning_rate": 3.1368490792925275e-06, + "loss": 3.2932, + "step": 18672 + }, + { + "epoch": 1.4970034675592792, + "grad_norm": 1.6619764566421509, + "learning_rate": 3.1359047709784294e-06, + "loss": 3.1804, + "step": 18673 + }, + { + "epoch": 1.4970836423403018, + "grad_norm": 1.870923399925232, + "learning_rate": 3.1349605783905567e-06, + "loss": 3.1584, + "step": 18674 + }, + { + "epoch": 1.4971638171213244, + "grad_norm": 1.8274028301239014, + "learning_rate": 3.13401650154483e-06, + "loss": 3.1696, + "step": 18675 + }, + { + "epoch": 1.497243991902347, + "grad_norm": 1.7417542934417725, + "learning_rate": 3.133072540457166e-06, + "loss": 3.146, + "step": 18676 + }, + { + "epoch": 1.4973241666833697, + "grad_norm": 1.7521064281463623, + "learning_rate": 3.1321286951434802e-06, + "loss": 3.2406, + "step": 18677 + }, + { + "epoch": 1.4974043414643923, + "grad_norm": 1.86049222946167, + "learning_rate": 3.1311849656196823e-06, + "loss": 3.1904, + "step": 18678 + }, + { + "epoch": 1.497484516245415, + "grad_norm": 1.847929835319519, + "learning_rate": 3.130241351901685e-06, + "loss": 3.2539, + "step": 18679 + }, + { + "epoch": 1.4975646910264375, + "grad_norm": 1.7238057851791382, + "learning_rate": 3.1292978540053963e-06, + "loss": 3.0533, + "step": 18680 + }, + { + "epoch": 1.4976448658074601, + "grad_norm": 1.6687101125717163, + "learning_rate": 3.128354471946725e-06, + "loss": 3.0778, + "step": 18681 + }, + { + "epoch": 1.4977250405884828, + "grad_norm": 1.7501887083053589, + "learning_rate": 3.1274112057415797e-06, + "loss": 3.2159, + "step": 18682 + }, + { + "epoch": 1.4978052153695054, + "grad_norm": 1.7131116390228271, + "learning_rate": 3.1264680554058545e-06, + "loss": 3.2091, + "step": 18683 + }, + { + "epoch": 1.497885390150528, + "grad_norm": 1.8728973865509033, + "learning_rate": 3.1255250209554567e-06, + "loss": 3.1602, + "step": 18684 + }, + { + "epoch": 1.4979655649315509, + "grad_norm": 1.7970727682113647, + "learning_rate": 3.124582102406283e-06, + "loss": 3.1658, + "step": 18685 + }, + { + "epoch": 1.4980457397125735, + "grad_norm": 1.7449557781219482, + "learning_rate": 3.1236392997742326e-06, + "loss": 3.1881, + "step": 18686 + }, + { + "epoch": 1.4981259144935961, + "grad_norm": 1.7975419759750366, + "learning_rate": 3.1226966130752035e-06, + "loss": 3.1993, + "step": 18687 + }, + { + "epoch": 1.4982060892746187, + "grad_norm": 0.6646019816398621, + "learning_rate": 3.1217540423250826e-06, + "loss": 2.4751, + "step": 18688 + }, + { + "epoch": 1.4982862640556414, + "grad_norm": 1.7995469570159912, + "learning_rate": 3.1208115875397637e-06, + "loss": 3.1915, + "step": 18689 + }, + { + "epoch": 1.498366438836664, + "grad_norm": 1.6859134435653687, + "learning_rate": 3.119869248735138e-06, + "loss": 3.1001, + "step": 18690 + }, + { + "epoch": 1.4984466136176866, + "grad_norm": 1.7499282360076904, + "learning_rate": 3.118927025927091e-06, + "loss": 3.264, + "step": 18691 + }, + { + "epoch": 1.4985267883987092, + "grad_norm": 1.904105544090271, + "learning_rate": 3.1179849191315125e-06, + "loss": 3.1923, + "step": 18692 + }, + { + "epoch": 1.4986069631797319, + "grad_norm": 1.8714938163757324, + "learning_rate": 3.11704292836428e-06, + "loss": 3.0962, + "step": 18693 + }, + { + "epoch": 1.4986871379607545, + "grad_norm": 2.3810222148895264, + "learning_rate": 3.1161010536412785e-06, + "loss": 3.1354, + "step": 18694 + }, + { + "epoch": 1.498767312741777, + "grad_norm": 1.8386459350585938, + "learning_rate": 3.115159294978386e-06, + "loss": 3.2827, + "step": 18695 + }, + { + "epoch": 1.4988474875227997, + "grad_norm": 1.7570469379425049, + "learning_rate": 3.1142176523914815e-06, + "loss": 3.1769, + "step": 18696 + }, + { + "epoch": 1.4989276623038224, + "grad_norm": 1.7397191524505615, + "learning_rate": 3.1132761258964438e-06, + "loss": 3.1511, + "step": 18697 + }, + { + "epoch": 1.499007837084845, + "grad_norm": 1.7417216300964355, + "learning_rate": 3.112334715509141e-06, + "loss": 3.2058, + "step": 18698 + }, + { + "epoch": 1.4990880118658676, + "grad_norm": 1.7649178504943848, + "learning_rate": 3.1113934212454466e-06, + "loss": 3.1785, + "step": 18699 + }, + { + "epoch": 1.4991681866468902, + "grad_norm": 1.7396148443222046, + "learning_rate": 3.1104522431212326e-06, + "loss": 3.2216, + "step": 18700 + }, + { + "epoch": 1.4992483614279128, + "grad_norm": 1.7169278860092163, + "learning_rate": 3.109511181152365e-06, + "loss": 3.204, + "step": 18701 + }, + { + "epoch": 1.4993285362089355, + "grad_norm": 1.7359519004821777, + "learning_rate": 3.108570235354713e-06, + "loss": 3.1434, + "step": 18702 + }, + { + "epoch": 1.499408710989958, + "grad_norm": 1.709969401359558, + "learning_rate": 3.1076294057441358e-06, + "loss": 3.2202, + "step": 18703 + }, + { + "epoch": 1.4994888857709807, + "grad_norm": 0.672502875328064, + "learning_rate": 3.106688692336497e-06, + "loss": 2.4626, + "step": 18704 + }, + { + "epoch": 1.4995690605520033, + "grad_norm": 1.782407522201538, + "learning_rate": 3.105748095147657e-06, + "loss": 3.1583, + "step": 18705 + }, + { + "epoch": 1.499649235333026, + "grad_norm": 1.786872386932373, + "learning_rate": 3.1048076141934757e-06, + "loss": 3.2361, + "step": 18706 + }, + { + "epoch": 1.4997294101140486, + "grad_norm": 1.8550186157226562, + "learning_rate": 3.1038672494898105e-06, + "loss": 3.1816, + "step": 18707 + }, + { + "epoch": 1.4998095848950712, + "grad_norm": 1.8092728853225708, + "learning_rate": 3.1029270010525092e-06, + "loss": 3.2491, + "step": 18708 + }, + { + "epoch": 1.4998897596760938, + "grad_norm": 1.7394263744354248, + "learning_rate": 3.1019868688974288e-06, + "loss": 3.09, + "step": 18709 + }, + { + "epoch": 1.4999699344571165, + "grad_norm": 1.8134502172470093, + "learning_rate": 3.101046853040418e-06, + "loss": 3.1685, + "step": 18710 + }, + { + "epoch": 1.5000501092381393, + "grad_norm": 1.7028164863586426, + "learning_rate": 3.100106953497326e-06, + "loss": 3.2414, + "step": 18711 + }, + { + "epoch": 1.500130284019162, + "grad_norm": 1.7493090629577637, + "learning_rate": 3.099167170284003e-06, + "loss": 3.178, + "step": 18712 + }, + { + "epoch": 1.5002104588001846, + "grad_norm": 1.7353968620300293, + "learning_rate": 3.0982275034162855e-06, + "loss": 3.1801, + "step": 18713 + }, + { + "epoch": 1.5002906335812072, + "grad_norm": 1.6698733568191528, + "learning_rate": 3.09728795291002e-06, + "loss": 3.0847, + "step": 18714 + }, + { + "epoch": 1.5003708083622298, + "grad_norm": 1.783158302307129, + "learning_rate": 3.0963485187810473e-06, + "loss": 3.1407, + "step": 18715 + }, + { + "epoch": 1.5004509831432524, + "grad_norm": 1.808760643005371, + "learning_rate": 3.0954092010452043e-06, + "loss": 3.072, + "step": 18716 + }, + { + "epoch": 1.500531157924275, + "grad_norm": 1.708605170249939, + "learning_rate": 3.0944699997183325e-06, + "loss": 3.0616, + "step": 18717 + }, + { + "epoch": 1.5006113327052977, + "grad_norm": 1.9224109649658203, + "learning_rate": 3.0935309148162605e-06, + "loss": 3.1079, + "step": 18718 + }, + { + "epoch": 1.5006915074863203, + "grad_norm": 1.8448553085327148, + "learning_rate": 3.0925919463548226e-06, + "loss": 3.1673, + "step": 18719 + }, + { + "epoch": 1.500771682267343, + "grad_norm": 1.816473126411438, + "learning_rate": 3.091653094349849e-06, + "loss": 3.2156, + "step": 18720 + }, + { + "epoch": 1.5008518570483655, + "grad_norm": 1.8488125801086426, + "learning_rate": 3.0907143588171705e-06, + "loss": 3.1321, + "step": 18721 + }, + { + "epoch": 1.5009320318293882, + "grad_norm": 1.786213994026184, + "learning_rate": 3.089775739772617e-06, + "loss": 3.1685, + "step": 18722 + }, + { + "epoch": 1.5010122066104108, + "grad_norm": 1.727054476737976, + "learning_rate": 3.0888372372320043e-06, + "loss": 3.1173, + "step": 18723 + }, + { + "epoch": 1.5010923813914334, + "grad_norm": 0.669457733631134, + "learning_rate": 3.0878988512111605e-06, + "loss": 2.4672, + "step": 18724 + }, + { + "epoch": 1.501172556172456, + "grad_norm": 1.6595350503921509, + "learning_rate": 3.0869605817259063e-06, + "loss": 3.1445, + "step": 18725 + }, + { + "epoch": 1.5012527309534787, + "grad_norm": 0.6695218086242676, + "learning_rate": 3.0860224287920603e-06, + "loss": 2.47, + "step": 18726 + }, + { + "epoch": 1.5013329057345013, + "grad_norm": 1.7426108121871948, + "learning_rate": 3.0850843924254427e-06, + "loss": 3.2131, + "step": 18727 + }, + { + "epoch": 1.501413080515524, + "grad_norm": 1.872867226600647, + "learning_rate": 3.084146472641861e-06, + "loss": 3.1942, + "step": 18728 + }, + { + "epoch": 1.5014932552965465, + "grad_norm": 0.6691962480545044, + "learning_rate": 3.083208669457134e-06, + "loss": 2.4674, + "step": 18729 + }, + { + "epoch": 1.5015734300775692, + "grad_norm": 1.850388526916504, + "learning_rate": 3.08227098288707e-06, + "loss": 3.1281, + "step": 18730 + }, + { + "epoch": 1.5016536048585918, + "grad_norm": 1.6789565086364746, + "learning_rate": 3.0813334129474793e-06, + "loss": 3.1105, + "step": 18731 + }, + { + "epoch": 1.5017337796396144, + "grad_norm": 1.7835719585418701, + "learning_rate": 3.080395959654172e-06, + "loss": 3.241, + "step": 18732 + }, + { + "epoch": 1.501813954420637, + "grad_norm": 1.796297550201416, + "learning_rate": 3.0794586230229472e-06, + "loss": 3.2452, + "step": 18733 + }, + { + "epoch": 1.5018941292016597, + "grad_norm": 1.7770922183990479, + "learning_rate": 3.078521403069611e-06, + "loss": 3.2731, + "step": 18734 + }, + { + "epoch": 1.5019743039826823, + "grad_norm": 1.829342007637024, + "learning_rate": 3.077584299809965e-06, + "loss": 3.1455, + "step": 18735 + }, + { + "epoch": 1.502054478763705, + "grad_norm": 1.7132165431976318, + "learning_rate": 3.0766473132598074e-06, + "loss": 3.1284, + "step": 18736 + }, + { + "epoch": 1.5021346535447275, + "grad_norm": 1.8574599027633667, + "learning_rate": 3.0757104434349395e-06, + "loss": 3.1835, + "step": 18737 + }, + { + "epoch": 1.5022148283257502, + "grad_norm": 1.9438194036483765, + "learning_rate": 3.074773690351149e-06, + "loss": 3.2072, + "step": 18738 + }, + { + "epoch": 1.5022950031067728, + "grad_norm": 1.8096697330474854, + "learning_rate": 3.0738370540242345e-06, + "loss": 3.1934, + "step": 18739 + }, + { + "epoch": 1.5023751778877954, + "grad_norm": 1.7755142450332642, + "learning_rate": 3.072900534469987e-06, + "loss": 3.1556, + "step": 18740 + }, + { + "epoch": 1.502455352668818, + "grad_norm": 1.8066688776016235, + "learning_rate": 3.0719641317041935e-06, + "loss": 3.1873, + "step": 18741 + }, + { + "epoch": 1.5025355274498406, + "grad_norm": 1.7807830572128296, + "learning_rate": 3.0710278457426467e-06, + "loss": 3.1625, + "step": 18742 + }, + { + "epoch": 1.5026157022308633, + "grad_norm": 1.8498296737670898, + "learning_rate": 3.0700916766011257e-06, + "loss": 3.2183, + "step": 18743 + }, + { + "epoch": 1.502695877011886, + "grad_norm": 1.8513031005859375, + "learning_rate": 3.069155624295417e-06, + "loss": 3.1666, + "step": 18744 + }, + { + "epoch": 1.5027760517929085, + "grad_norm": 1.73488187789917, + "learning_rate": 3.0682196888413006e-06, + "loss": 3.1777, + "step": 18745 + }, + { + "epoch": 1.5028562265739311, + "grad_norm": 1.6845622062683105, + "learning_rate": 3.0672838702545584e-06, + "loss": 3.1643, + "step": 18746 + }, + { + "epoch": 1.5029364013549538, + "grad_norm": 1.7698882818222046, + "learning_rate": 3.0663481685509687e-06, + "loss": 3.212, + "step": 18747 + }, + { + "epoch": 1.5030165761359764, + "grad_norm": 1.7855192422866821, + "learning_rate": 3.0654125837463033e-06, + "loss": 3.2058, + "step": 18748 + }, + { + "epoch": 1.503096750916999, + "grad_norm": 1.7688332796096802, + "learning_rate": 3.064477115856337e-06, + "loss": 3.1898, + "step": 18749 + }, + { + "epoch": 1.5031769256980216, + "grad_norm": 1.8178526163101196, + "learning_rate": 3.0635417648968426e-06, + "loss": 3.1676, + "step": 18750 + }, + { + "epoch": 1.5032571004790443, + "grad_norm": 1.7041999101638794, + "learning_rate": 3.0626065308835895e-06, + "loss": 3.1545, + "step": 18751 + }, + { + "epoch": 1.5033372752600669, + "grad_norm": 1.7458010911941528, + "learning_rate": 3.0616714138323477e-06, + "loss": 3.2242, + "step": 18752 + }, + { + "epoch": 1.5034174500410895, + "grad_norm": 1.7229640483856201, + "learning_rate": 3.060736413758878e-06, + "loss": 3.1497, + "step": 18753 + }, + { + "epoch": 1.5034976248221121, + "grad_norm": 1.8313876390457153, + "learning_rate": 3.0598015306789475e-06, + "loss": 3.2273, + "step": 18754 + }, + { + "epoch": 1.5035777996031348, + "grad_norm": 0.6675519943237305, + "learning_rate": 3.058866764608316e-06, + "loss": 2.4714, + "step": 18755 + }, + { + "epoch": 1.5036579743841574, + "grad_norm": 1.7531301975250244, + "learning_rate": 3.057932115562745e-06, + "loss": 3.249, + "step": 18756 + }, + { + "epoch": 1.50373814916518, + "grad_norm": 1.7951552867889404, + "learning_rate": 3.0569975835579956e-06, + "loss": 3.1566, + "step": 18757 + }, + { + "epoch": 1.5038183239462026, + "grad_norm": 1.7021808624267578, + "learning_rate": 3.0560631686098164e-06, + "loss": 3.1017, + "step": 18758 + }, + { + "epoch": 1.5038984987272253, + "grad_norm": 1.8124557733535767, + "learning_rate": 3.0551288707339643e-06, + "loss": 3.1851, + "step": 18759 + }, + { + "epoch": 1.5039786735082479, + "grad_norm": 1.8208518028259277, + "learning_rate": 3.0541946899461926e-06, + "loss": 3.2603, + "step": 18760 + }, + { + "epoch": 1.5040588482892705, + "grad_norm": 1.719555377960205, + "learning_rate": 3.05326062626225e-06, + "loss": 3.2395, + "step": 18761 + }, + { + "epoch": 1.5041390230702931, + "grad_norm": 1.7299799919128418, + "learning_rate": 3.0523266796978877e-06, + "loss": 3.2234, + "step": 18762 + }, + { + "epoch": 1.5042191978513157, + "grad_norm": 1.805072546005249, + "learning_rate": 3.0513928502688472e-06, + "loss": 3.2069, + "step": 18763 + }, + { + "epoch": 1.5042993726323384, + "grad_norm": 1.7182722091674805, + "learning_rate": 3.0504591379908733e-06, + "loss": 3.0977, + "step": 18764 + }, + { + "epoch": 1.504379547413361, + "grad_norm": 1.835080862045288, + "learning_rate": 3.04952554287971e-06, + "loss": 3.184, + "step": 18765 + }, + { + "epoch": 1.5044597221943836, + "grad_norm": 1.797703742980957, + "learning_rate": 3.0485920649510947e-06, + "loss": 3.1512, + "step": 18766 + }, + { + "epoch": 1.5045398969754062, + "grad_norm": 1.800222635269165, + "learning_rate": 3.0476587042207717e-06, + "loss": 3.1896, + "step": 18767 + }, + { + "epoch": 1.5046200717564289, + "grad_norm": 1.694797396659851, + "learning_rate": 3.0467254607044693e-06, + "loss": 3.2085, + "step": 18768 + }, + { + "epoch": 1.5047002465374515, + "grad_norm": 1.67952299118042, + "learning_rate": 3.0457923344179253e-06, + "loss": 3.1696, + "step": 18769 + }, + { + "epoch": 1.5047804213184741, + "grad_norm": 1.7986465692520142, + "learning_rate": 3.044859325376872e-06, + "loss": 3.1087, + "step": 18770 + }, + { + "epoch": 1.5048605960994967, + "grad_norm": 1.7971279621124268, + "learning_rate": 3.0439264335970386e-06, + "loss": 3.1128, + "step": 18771 + }, + { + "epoch": 1.5049407708805196, + "grad_norm": 1.7040789127349854, + "learning_rate": 3.0429936590941578e-06, + "loss": 3.11, + "step": 18772 + }, + { + "epoch": 1.5050209456615422, + "grad_norm": 1.7147886753082275, + "learning_rate": 3.042061001883948e-06, + "loss": 3.1578, + "step": 18773 + }, + { + "epoch": 1.5051011204425648, + "grad_norm": 1.9089261293411255, + "learning_rate": 3.041128461982138e-06, + "loss": 3.2412, + "step": 18774 + }, + { + "epoch": 1.5051812952235875, + "grad_norm": 1.7234413623809814, + "learning_rate": 3.0401960394044496e-06, + "loss": 3.1539, + "step": 18775 + }, + { + "epoch": 1.50526147000461, + "grad_norm": 0.6620159149169922, + "learning_rate": 3.039263734166602e-06, + "loss": 2.4673, + "step": 18776 + }, + { + "epoch": 1.5053416447856327, + "grad_norm": 1.8542416095733643, + "learning_rate": 3.038331546284319e-06, + "loss": 3.1516, + "step": 18777 + }, + { + "epoch": 1.5054218195666553, + "grad_norm": 1.7373020648956299, + "learning_rate": 3.0373994757733096e-06, + "loss": 3.127, + "step": 18778 + }, + { + "epoch": 1.505501994347678, + "grad_norm": 1.7517733573913574, + "learning_rate": 3.03646752264929e-06, + "loss": 3.1014, + "step": 18779 + }, + { + "epoch": 1.5055821691287006, + "grad_norm": 1.813469409942627, + "learning_rate": 3.035535686927975e-06, + "loss": 3.0381, + "step": 18780 + }, + { + "epoch": 1.5056623439097232, + "grad_norm": 1.7607529163360596, + "learning_rate": 3.0346039686250728e-06, + "loss": 3.1769, + "step": 18781 + }, + { + "epoch": 1.5057425186907458, + "grad_norm": 1.839114785194397, + "learning_rate": 3.033672367756297e-06, + "loss": 3.1425, + "step": 18782 + }, + { + "epoch": 1.5058226934717684, + "grad_norm": 1.7435399293899536, + "learning_rate": 3.0327408843373464e-06, + "loss": 3.0867, + "step": 18783 + }, + { + "epoch": 1.505902868252791, + "grad_norm": 1.7452099323272705, + "learning_rate": 3.031809518383929e-06, + "loss": 3.1929, + "step": 18784 + }, + { + "epoch": 1.5059830430338137, + "grad_norm": 1.8415035009384155, + "learning_rate": 3.0308782699117466e-06, + "loss": 3.2054, + "step": 18785 + }, + { + "epoch": 1.5060632178148363, + "grad_norm": 1.853784441947937, + "learning_rate": 3.029947138936501e-06, + "loss": 3.265, + "step": 18786 + }, + { + "epoch": 1.506143392595859, + "grad_norm": 1.7349601984024048, + "learning_rate": 3.029016125473894e-06, + "loss": 3.1855, + "step": 18787 + }, + { + "epoch": 1.5062235673768816, + "grad_norm": 1.7569019794464111, + "learning_rate": 3.0280852295396133e-06, + "loss": 3.1371, + "step": 18788 + }, + { + "epoch": 1.5063037421579042, + "grad_norm": 1.8031938076019287, + "learning_rate": 3.02715445114936e-06, + "loss": 3.2517, + "step": 18789 + }, + { + "epoch": 1.5063839169389268, + "grad_norm": 1.7673985958099365, + "learning_rate": 3.0262237903188254e-06, + "loss": 3.1112, + "step": 18790 + }, + { + "epoch": 1.5064640917199494, + "grad_norm": 1.728667140007019, + "learning_rate": 3.025293247063699e-06, + "loss": 3.0987, + "step": 18791 + }, + { + "epoch": 1.5065442665009723, + "grad_norm": 1.8410261869430542, + "learning_rate": 3.0243628213996733e-06, + "loss": 3.14, + "step": 18792 + }, + { + "epoch": 1.506624441281995, + "grad_norm": 1.722174048423767, + "learning_rate": 3.0234325133424302e-06, + "loss": 3.2334, + "step": 18793 + }, + { + "epoch": 1.5067046160630175, + "grad_norm": 1.873092532157898, + "learning_rate": 3.0225023229076543e-06, + "loss": 3.2439, + "step": 18794 + }, + { + "epoch": 1.5067847908440402, + "grad_norm": 1.7839672565460205, + "learning_rate": 3.021572250111031e-06, + "loss": 3.1749, + "step": 18795 + }, + { + "epoch": 1.5068649656250628, + "grad_norm": 1.8087882995605469, + "learning_rate": 3.020642294968241e-06, + "loss": 3.145, + "step": 18796 + }, + { + "epoch": 1.5069451404060854, + "grad_norm": 1.8509242534637451, + "learning_rate": 3.0197124574949644e-06, + "loss": 3.1937, + "step": 18797 + }, + { + "epoch": 1.507025315187108, + "grad_norm": 1.8787267208099365, + "learning_rate": 3.0187827377068725e-06, + "loss": 3.2229, + "step": 18798 + }, + { + "epoch": 1.5071054899681307, + "grad_norm": 1.7803972959518433, + "learning_rate": 3.017853135619644e-06, + "loss": 3.1793, + "step": 18799 + }, + { + "epoch": 1.5071856647491533, + "grad_norm": 1.6883294582366943, + "learning_rate": 3.0169236512489495e-06, + "loss": 3.0928, + "step": 18800 + }, + { + "epoch": 1.507265839530176, + "grad_norm": 1.7048271894454956, + "learning_rate": 3.0159942846104618e-06, + "loss": 3.2816, + "step": 18801 + }, + { + "epoch": 1.5073460143111985, + "grad_norm": 2.041820526123047, + "learning_rate": 3.015065035719853e-06, + "loss": 3.1509, + "step": 18802 + }, + { + "epoch": 1.5074261890922211, + "grad_norm": 1.739761471748352, + "learning_rate": 3.014135904592782e-06, + "loss": 3.1636, + "step": 18803 + }, + { + "epoch": 1.5075063638732438, + "grad_norm": 1.8177130222320557, + "learning_rate": 3.0132068912449174e-06, + "loss": 3.1723, + "step": 18804 + }, + { + "epoch": 1.5075865386542664, + "grad_norm": 1.7096054553985596, + "learning_rate": 3.0122779956919224e-06, + "loss": 3.2065, + "step": 18805 + }, + { + "epoch": 1.507666713435289, + "grad_norm": 1.8042542934417725, + "learning_rate": 3.0113492179494574e-06, + "loss": 3.097, + "step": 18806 + }, + { + "epoch": 1.5077468882163116, + "grad_norm": 1.7726011276245117, + "learning_rate": 3.0104205580331857e-06, + "loss": 3.1885, + "step": 18807 + }, + { + "epoch": 1.5078270629973343, + "grad_norm": 1.8233301639556885, + "learning_rate": 3.009492015958756e-06, + "loss": 3.2153, + "step": 18808 + }, + { + "epoch": 1.507907237778357, + "grad_norm": 1.721633791923523, + "learning_rate": 3.0085635917418277e-06, + "loss": 3.2001, + "step": 18809 + }, + { + "epoch": 1.5079874125593795, + "grad_norm": 1.9125510454177856, + "learning_rate": 3.0076352853980528e-06, + "loss": 3.1404, + "step": 18810 + }, + { + "epoch": 1.5080675873404021, + "grad_norm": 1.7452536821365356, + "learning_rate": 3.0067070969430834e-06, + "loss": 3.1926, + "step": 18811 + }, + { + "epoch": 1.5081477621214248, + "grad_norm": 1.7662056684494019, + "learning_rate": 3.00577902639257e-06, + "loss": 3.1477, + "step": 18812 + }, + { + "epoch": 1.5082279369024474, + "grad_norm": 1.7902928590774536, + "learning_rate": 3.004851073762155e-06, + "loss": 3.1302, + "step": 18813 + }, + { + "epoch": 1.50830811168347, + "grad_norm": 1.6673376560211182, + "learning_rate": 3.0039232390674842e-06, + "loss": 3.1219, + "step": 18814 + }, + { + "epoch": 1.5083882864644926, + "grad_norm": 1.7168058156967163, + "learning_rate": 3.0029955223242026e-06, + "loss": 3.1385, + "step": 18815 + }, + { + "epoch": 1.5084684612455153, + "grad_norm": 1.760420560836792, + "learning_rate": 3.002067923547951e-06, + "loss": 3.054, + "step": 18816 + }, + { + "epoch": 1.5085486360265379, + "grad_norm": 1.845502257347107, + "learning_rate": 3.00114044275437e-06, + "loss": 3.1573, + "step": 18817 + }, + { + "epoch": 1.5086288108075605, + "grad_norm": 1.775830626487732, + "learning_rate": 3.0002130799590922e-06, + "loss": 3.1942, + "step": 18818 + }, + { + "epoch": 1.5087089855885831, + "grad_norm": 1.9059836864471436, + "learning_rate": 2.9992858351777556e-06, + "loss": 3.1718, + "step": 18819 + }, + { + "epoch": 1.5087891603696058, + "grad_norm": 1.7333122491836548, + "learning_rate": 2.998358708425991e-06, + "loss": 3.2727, + "step": 18820 + }, + { + "epoch": 1.5088693351506284, + "grad_norm": 1.76571786403656, + "learning_rate": 2.997431699719432e-06, + "loss": 3.2129, + "step": 18821 + }, + { + "epoch": 1.508949509931651, + "grad_norm": 1.8038400411605835, + "learning_rate": 2.9965048090737092e-06, + "loss": 3.2754, + "step": 18822 + }, + { + "epoch": 1.5090296847126736, + "grad_norm": 1.8021539449691772, + "learning_rate": 2.995578036504444e-06, + "loss": 3.1276, + "step": 18823 + }, + { + "epoch": 1.5091098594936962, + "grad_norm": 1.7675950527191162, + "learning_rate": 2.994651382027264e-06, + "loss": 3.2303, + "step": 18824 + }, + { + "epoch": 1.5091900342747189, + "grad_norm": 1.984700322151184, + "learning_rate": 2.993724845657794e-06, + "loss": 3.2106, + "step": 18825 + }, + { + "epoch": 1.5092702090557415, + "grad_norm": 1.864762783050537, + "learning_rate": 2.992798427411653e-06, + "loss": 3.109, + "step": 18826 + }, + { + "epoch": 1.5093503838367641, + "grad_norm": 1.9559067487716675, + "learning_rate": 2.9918721273044638e-06, + "loss": 3.2365, + "step": 18827 + }, + { + "epoch": 1.5094305586177867, + "grad_norm": 1.7936893701553345, + "learning_rate": 2.9909459453518373e-06, + "loss": 3.1511, + "step": 18828 + }, + { + "epoch": 1.5095107333988094, + "grad_norm": 1.6907339096069336, + "learning_rate": 2.9900198815693927e-06, + "loss": 3.059, + "step": 18829 + }, + { + "epoch": 1.509590908179832, + "grad_norm": 1.791675329208374, + "learning_rate": 2.9890939359727424e-06, + "loss": 3.2182, + "step": 18830 + }, + { + "epoch": 1.5096710829608546, + "grad_norm": 0.683844268321991, + "learning_rate": 2.9881681085774973e-06, + "loss": 2.4792, + "step": 18831 + }, + { + "epoch": 1.5097512577418772, + "grad_norm": 1.9411214590072632, + "learning_rate": 2.9872423993992694e-06, + "loss": 3.1542, + "step": 18832 + }, + { + "epoch": 1.5098314325228999, + "grad_norm": 1.7175970077514648, + "learning_rate": 2.9863168084536608e-06, + "loss": 3.1584, + "step": 18833 + }, + { + "epoch": 1.5099116073039225, + "grad_norm": 1.8167598247528076, + "learning_rate": 2.985391335756278e-06, + "loss": 3.2904, + "step": 18834 + }, + { + "epoch": 1.5099917820849451, + "grad_norm": 1.8357492685317993, + "learning_rate": 2.984465981322726e-06, + "loss": 3.1171, + "step": 18835 + }, + { + "epoch": 1.5100719568659677, + "grad_norm": 1.7358477115631104, + "learning_rate": 2.983540745168605e-06, + "loss": 3.1713, + "step": 18836 + }, + { + "epoch": 1.5101521316469904, + "grad_norm": 1.7767860889434814, + "learning_rate": 2.9826156273095165e-06, + "loss": 3.1873, + "step": 18837 + }, + { + "epoch": 1.510232306428013, + "grad_norm": 1.8254380226135254, + "learning_rate": 2.9816906277610525e-06, + "loss": 3.2338, + "step": 18838 + }, + { + "epoch": 1.5103124812090356, + "grad_norm": 1.813712477684021, + "learning_rate": 2.9807657465388116e-06, + "loss": 3.1649, + "step": 18839 + }, + { + "epoch": 1.5103926559900582, + "grad_norm": 1.883352279663086, + "learning_rate": 2.9798409836583864e-06, + "loss": 3.106, + "step": 18840 + }, + { + "epoch": 1.5104728307710809, + "grad_norm": 1.7602145671844482, + "learning_rate": 2.978916339135367e-06, + "loss": 3.1565, + "step": 18841 + }, + { + "epoch": 1.5105530055521035, + "grad_norm": 1.775095820426941, + "learning_rate": 2.977991812985348e-06, + "loss": 3.0902, + "step": 18842 + }, + { + "epoch": 1.510633180333126, + "grad_norm": 1.7169625759124756, + "learning_rate": 2.9770674052239077e-06, + "loss": 3.0916, + "step": 18843 + }, + { + "epoch": 1.5107133551141487, + "grad_norm": 1.7578327655792236, + "learning_rate": 2.9761431158666367e-06, + "loss": 3.1938, + "step": 18844 + }, + { + "epoch": 1.5107935298951713, + "grad_norm": 1.8013317584991455, + "learning_rate": 2.975218944929118e-06, + "loss": 3.1601, + "step": 18845 + }, + { + "epoch": 1.510873704676194, + "grad_norm": 1.8655792474746704, + "learning_rate": 2.974294892426931e-06, + "loss": 3.101, + "step": 18846 + }, + { + "epoch": 1.5109538794572166, + "grad_norm": 1.7016981840133667, + "learning_rate": 2.9733709583756596e-06, + "loss": 3.0737, + "step": 18847 + }, + { + "epoch": 1.5110340542382392, + "grad_norm": 1.7638041973114014, + "learning_rate": 2.972447142790874e-06, + "loss": 3.1954, + "step": 18848 + }, + { + "epoch": 1.5111142290192618, + "grad_norm": 1.751952886581421, + "learning_rate": 2.9715234456881538e-06, + "loss": 3.2961, + "step": 18849 + }, + { + "epoch": 1.5111944038002845, + "grad_norm": 1.8447058200836182, + "learning_rate": 2.9705998670830707e-06, + "loss": 3.2555, + "step": 18850 + }, + { + "epoch": 1.511274578581307, + "grad_norm": 1.7290233373641968, + "learning_rate": 2.9696764069911953e-06, + "loss": 3.129, + "step": 18851 + }, + { + "epoch": 1.5113547533623297, + "grad_norm": 1.7247259616851807, + "learning_rate": 2.968753065428103e-06, + "loss": 3.2706, + "step": 18852 + }, + { + "epoch": 1.5114349281433526, + "grad_norm": 1.7961301803588867, + "learning_rate": 2.967829842409351e-06, + "loss": 3.1545, + "step": 18853 + }, + { + "epoch": 1.5115151029243752, + "grad_norm": 1.7284109592437744, + "learning_rate": 2.96690673795051e-06, + "loss": 3.3297, + "step": 18854 + }, + { + "epoch": 1.5115952777053978, + "grad_norm": 1.860076904296875, + "learning_rate": 2.965983752067143e-06, + "loss": 3.1639, + "step": 18855 + }, + { + "epoch": 1.5116754524864204, + "grad_norm": 1.795695185661316, + "learning_rate": 2.965060884774811e-06, + "loss": 3.1671, + "step": 18856 + }, + { + "epoch": 1.511755627267443, + "grad_norm": 1.868190884590149, + "learning_rate": 2.9641381360890765e-06, + "loss": 3.1489, + "step": 18857 + }, + { + "epoch": 1.5118358020484657, + "grad_norm": 1.806876540184021, + "learning_rate": 2.9632155060254897e-06, + "loss": 3.1903, + "step": 18858 + }, + { + "epoch": 1.5119159768294883, + "grad_norm": 1.8370044231414795, + "learning_rate": 2.9622929945996106e-06, + "loss": 3.1746, + "step": 18859 + }, + { + "epoch": 1.511996151610511, + "grad_norm": 1.7034268379211426, + "learning_rate": 2.9613706018269893e-06, + "loss": 3.2079, + "step": 18860 + }, + { + "epoch": 1.5120763263915336, + "grad_norm": 1.8810975551605225, + "learning_rate": 2.96044832772318e-06, + "loss": 3.1285, + "step": 18861 + }, + { + "epoch": 1.5121565011725562, + "grad_norm": 1.7870979309082031, + "learning_rate": 2.959526172303735e-06, + "loss": 3.0412, + "step": 18862 + }, + { + "epoch": 1.5122366759535788, + "grad_norm": 1.854663372039795, + "learning_rate": 2.9586041355841933e-06, + "loss": 3.2728, + "step": 18863 + }, + { + "epoch": 1.5123168507346014, + "grad_norm": 1.8201483488082886, + "learning_rate": 2.9576822175801047e-06, + "loss": 3.1855, + "step": 18864 + }, + { + "epoch": 1.512397025515624, + "grad_norm": 1.6637969017028809, + "learning_rate": 2.956760418307011e-06, + "loss": 3.0646, + "step": 18865 + }, + { + "epoch": 1.5124772002966467, + "grad_norm": 1.7032561302185059, + "learning_rate": 2.9558387377804553e-06, + "loss": 3.1344, + "step": 18866 + }, + { + "epoch": 1.5125573750776693, + "grad_norm": 1.68036949634552, + "learning_rate": 2.954917176015978e-06, + "loss": 3.2418, + "step": 18867 + }, + { + "epoch": 1.512637549858692, + "grad_norm": 1.7167714834213257, + "learning_rate": 2.9539957330291126e-06, + "loss": 3.08, + "step": 18868 + }, + { + "epoch": 1.5127177246397145, + "grad_norm": 1.7447246313095093, + "learning_rate": 2.9530744088353948e-06, + "loss": 3.2258, + "step": 18869 + }, + { + "epoch": 1.5127978994207372, + "grad_norm": 1.842447280883789, + "learning_rate": 2.9521532034503586e-06, + "loss": 3.2242, + "step": 18870 + }, + { + "epoch": 1.5128780742017598, + "grad_norm": 1.7497162818908691, + "learning_rate": 2.9512321168895365e-06, + "loss": 3.1979, + "step": 18871 + }, + { + "epoch": 1.5129582489827824, + "grad_norm": 1.8504737615585327, + "learning_rate": 2.950311149168459e-06, + "loss": 3.1583, + "step": 18872 + }, + { + "epoch": 1.513038423763805, + "grad_norm": 0.6556927561759949, + "learning_rate": 2.949390300302648e-06, + "loss": 2.4688, + "step": 18873 + }, + { + "epoch": 1.5131185985448279, + "grad_norm": 1.769500970840454, + "learning_rate": 2.9484695703076315e-06, + "loss": 3.0215, + "step": 18874 + }, + { + "epoch": 1.5131987733258505, + "grad_norm": 1.788517713546753, + "learning_rate": 2.9475489591989337e-06, + "loss": 3.2463, + "step": 18875 + }, + { + "epoch": 1.5132789481068731, + "grad_norm": 0.654114305973053, + "learning_rate": 2.9466284669920742e-06, + "loss": 2.4702, + "step": 18876 + }, + { + "epoch": 1.5133591228878958, + "grad_norm": 1.7949738502502441, + "learning_rate": 2.945708093702576e-06, + "loss": 3.1053, + "step": 18877 + }, + { + "epoch": 1.5134392976689184, + "grad_norm": 1.893821120262146, + "learning_rate": 2.9447878393459505e-06, + "loss": 3.1664, + "step": 18878 + }, + { + "epoch": 1.513519472449941, + "grad_norm": 1.6319572925567627, + "learning_rate": 2.943867703937715e-06, + "loss": 3.0892, + "step": 18879 + }, + { + "epoch": 1.5135996472309636, + "grad_norm": 1.680864691734314, + "learning_rate": 2.9429476874933837e-06, + "loss": 3.1688, + "step": 18880 + }, + { + "epoch": 1.5136798220119863, + "grad_norm": 1.7775357961654663, + "learning_rate": 2.9420277900284664e-06, + "loss": 3.1814, + "step": 18881 + }, + { + "epoch": 1.5137599967930089, + "grad_norm": 0.6717460751533508, + "learning_rate": 2.9411080115584743e-06, + "loss": 2.481, + "step": 18882 + }, + { + "epoch": 1.5138401715740315, + "grad_norm": 1.743248701095581, + "learning_rate": 2.940188352098916e-06, + "loss": 3.2169, + "step": 18883 + }, + { + "epoch": 1.5139203463550541, + "grad_norm": 1.8104469776153564, + "learning_rate": 2.939268811665291e-06, + "loss": 3.3356, + "step": 18884 + }, + { + "epoch": 1.5140005211360767, + "grad_norm": 1.765794277191162, + "learning_rate": 2.938349390273105e-06, + "loss": 3.1541, + "step": 18885 + }, + { + "epoch": 1.5140806959170994, + "grad_norm": 1.7302567958831787, + "learning_rate": 2.937430087937859e-06, + "loss": 3.1857, + "step": 18886 + }, + { + "epoch": 1.514160870698122, + "grad_norm": 1.8150365352630615, + "learning_rate": 2.936510904675054e-06, + "loss": 3.1606, + "step": 18887 + }, + { + "epoch": 1.5142410454791446, + "grad_norm": 1.8414100408554077, + "learning_rate": 2.935591840500188e-06, + "loss": 3.151, + "step": 18888 + }, + { + "epoch": 1.5143212202601672, + "grad_norm": 0.6623860597610474, + "learning_rate": 2.934672895428751e-06, + "loss": 2.4662, + "step": 18889 + }, + { + "epoch": 1.5144013950411899, + "grad_norm": 1.8096044063568115, + "learning_rate": 2.933754069476238e-06, + "loss": 3.2446, + "step": 18890 + }, + { + "epoch": 1.5144815698222125, + "grad_norm": 1.7478687763214111, + "learning_rate": 2.932835362658142e-06, + "loss": 3.2232, + "step": 18891 + }, + { + "epoch": 1.5145617446032351, + "grad_norm": 1.798516869544983, + "learning_rate": 2.9319167749899514e-06, + "loss": 3.2531, + "step": 18892 + }, + { + "epoch": 1.5146419193842577, + "grad_norm": 1.7223044633865356, + "learning_rate": 2.930998306487155e-06, + "loss": 3.1976, + "step": 18893 + }, + { + "epoch": 1.5147220941652804, + "grad_norm": 1.8783965110778809, + "learning_rate": 2.9300799571652326e-06, + "loss": 3.2068, + "step": 18894 + }, + { + "epoch": 1.514802268946303, + "grad_norm": 1.9065570831298828, + "learning_rate": 2.9291617270396712e-06, + "loss": 3.1526, + "step": 18895 + }, + { + "epoch": 1.5148824437273256, + "grad_norm": 1.7654105424880981, + "learning_rate": 2.928243616125951e-06, + "loss": 3.1569, + "step": 18896 + }, + { + "epoch": 1.5149626185083482, + "grad_norm": 1.803619146347046, + "learning_rate": 2.927325624439551e-06, + "loss": 3.0623, + "step": 18897 + }, + { + "epoch": 1.5150427932893709, + "grad_norm": 0.6885570287704468, + "learning_rate": 2.9264077519959476e-06, + "loss": 2.4713, + "step": 18898 + }, + { + "epoch": 1.5151229680703935, + "grad_norm": 1.754218339920044, + "learning_rate": 2.925489998810618e-06, + "loss": 3.0964, + "step": 18899 + }, + { + "epoch": 1.515203142851416, + "grad_norm": 1.76352858543396, + "learning_rate": 2.924572364899033e-06, + "loss": 3.0535, + "step": 18900 + }, + { + "epoch": 1.5152833176324387, + "grad_norm": 1.784623622894287, + "learning_rate": 2.923654850276667e-06, + "loss": 3.1945, + "step": 18901 + }, + { + "epoch": 1.5153634924134614, + "grad_norm": 1.7396389245986938, + "learning_rate": 2.9227374549589837e-06, + "loss": 3.1903, + "step": 18902 + }, + { + "epoch": 1.515443667194484, + "grad_norm": 1.7813682556152344, + "learning_rate": 2.921820178961452e-06, + "loss": 3.3085, + "step": 18903 + }, + { + "epoch": 1.5155238419755066, + "grad_norm": 1.7848966121673584, + "learning_rate": 2.9209030222995384e-06, + "loss": 3.182, + "step": 18904 + }, + { + "epoch": 1.5156040167565292, + "grad_norm": 1.6886333227157593, + "learning_rate": 2.9199859849887047e-06, + "loss": 3.2461, + "step": 18905 + }, + { + "epoch": 1.5156841915375519, + "grad_norm": 1.6983342170715332, + "learning_rate": 2.9190690670444144e-06, + "loss": 3.2384, + "step": 18906 + }, + { + "epoch": 1.5157643663185745, + "grad_norm": 1.7124534845352173, + "learning_rate": 2.9181522684821218e-06, + "loss": 3.2388, + "step": 18907 + }, + { + "epoch": 1.515844541099597, + "grad_norm": 0.6931567192077637, + "learning_rate": 2.9172355893172854e-06, + "loss": 2.4734, + "step": 18908 + }, + { + "epoch": 1.5159247158806197, + "grad_norm": 1.8813862800598145, + "learning_rate": 2.916319029565361e-06, + "loss": 3.2172, + "step": 18909 + }, + { + "epoch": 1.5160048906616423, + "grad_norm": 1.7323883771896362, + "learning_rate": 2.9154025892418013e-06, + "loss": 3.1616, + "step": 18910 + }, + { + "epoch": 1.516085065442665, + "grad_norm": 1.7758703231811523, + "learning_rate": 2.914486268362058e-06, + "loss": 3.1505, + "step": 18911 + }, + { + "epoch": 1.5161652402236876, + "grad_norm": 1.8223291635513306, + "learning_rate": 2.9135700669415813e-06, + "loss": 3.2512, + "step": 18912 + }, + { + "epoch": 1.5162454150047102, + "grad_norm": 1.804876446723938, + "learning_rate": 2.9126539849958126e-06, + "loss": 3.194, + "step": 18913 + }, + { + "epoch": 1.5163255897857328, + "grad_norm": 1.7095950841903687, + "learning_rate": 2.911738022540199e-06, + "loss": 3.2159, + "step": 18914 + }, + { + "epoch": 1.5164057645667555, + "grad_norm": 1.7313119173049927, + "learning_rate": 2.9108221795901857e-06, + "loss": 3.2252, + "step": 18915 + }, + { + "epoch": 1.516485939347778, + "grad_norm": 1.858017921447754, + "learning_rate": 2.909906456161211e-06, + "loss": 3.1298, + "step": 18916 + }, + { + "epoch": 1.5165661141288007, + "grad_norm": 1.733625888824463, + "learning_rate": 2.908990852268718e-06, + "loss": 3.1227, + "step": 18917 + }, + { + "epoch": 1.5166462889098233, + "grad_norm": 1.8177121877670288, + "learning_rate": 2.908075367928137e-06, + "loss": 3.1615, + "step": 18918 + }, + { + "epoch": 1.516726463690846, + "grad_norm": 1.7834153175354004, + "learning_rate": 2.9071600031549064e-06, + "loss": 3.1952, + "step": 18919 + }, + { + "epoch": 1.5168066384718686, + "grad_norm": 1.8616726398468018, + "learning_rate": 2.9062447579644582e-06, + "loss": 3.3251, + "step": 18920 + }, + { + "epoch": 1.5168868132528912, + "grad_norm": 1.901088833808899, + "learning_rate": 2.9053296323722234e-06, + "loss": 3.1849, + "step": 18921 + }, + { + "epoch": 1.5169669880339138, + "grad_norm": 1.8536789417266846, + "learning_rate": 2.9044146263936336e-06, + "loss": 3.2891, + "step": 18922 + }, + { + "epoch": 1.5170471628149365, + "grad_norm": 1.8084328174591064, + "learning_rate": 2.903499740044109e-06, + "loss": 3.1252, + "step": 18923 + }, + { + "epoch": 1.517127337595959, + "grad_norm": 1.826582908630371, + "learning_rate": 2.9025849733390774e-06, + "loss": 3.1365, + "step": 18924 + }, + { + "epoch": 1.5172075123769817, + "grad_norm": 1.8523708581924438, + "learning_rate": 2.9016703262939626e-06, + "loss": 3.1686, + "step": 18925 + }, + { + "epoch": 1.5172876871580043, + "grad_norm": 1.6890753507614136, + "learning_rate": 2.9007557989241843e-06, + "loss": 3.1245, + "step": 18926 + }, + { + "epoch": 1.517367861939027, + "grad_norm": 1.819280982017517, + "learning_rate": 2.899841391245164e-06, + "loss": 3.2109, + "step": 18927 + }, + { + "epoch": 1.5174480367200496, + "grad_norm": 1.732539176940918, + "learning_rate": 2.8989271032723123e-06, + "loss": 3.2074, + "step": 18928 + }, + { + "epoch": 1.5175282115010722, + "grad_norm": 1.7428756952285767, + "learning_rate": 2.898012935021047e-06, + "loss": 3.0695, + "step": 18929 + }, + { + "epoch": 1.5176083862820948, + "grad_norm": 1.7501380443572998, + "learning_rate": 2.897098886506781e-06, + "loss": 3.1319, + "step": 18930 + }, + { + "epoch": 1.5176885610631174, + "grad_norm": 1.7802989482879639, + "learning_rate": 2.8961849577449243e-06, + "loss": 3.1611, + "step": 18931 + }, + { + "epoch": 1.51776873584414, + "grad_norm": 1.7156190872192383, + "learning_rate": 2.8952711487508888e-06, + "loss": 3.0709, + "step": 18932 + }, + { + "epoch": 1.5178489106251627, + "grad_norm": 1.7618306875228882, + "learning_rate": 2.8943574595400747e-06, + "loss": 3.1491, + "step": 18933 + }, + { + "epoch": 1.5179290854061853, + "grad_norm": 1.8014068603515625, + "learning_rate": 2.8934438901278896e-06, + "loss": 3.1338, + "step": 18934 + }, + { + "epoch": 1.5180092601872082, + "grad_norm": 1.779822826385498, + "learning_rate": 2.892530440529735e-06, + "loss": 3.2093, + "step": 18935 + }, + { + "epoch": 1.5180894349682308, + "grad_norm": 1.7001426219940186, + "learning_rate": 2.8916171107610138e-06, + "loss": 3.0451, + "step": 18936 + }, + { + "epoch": 1.5181696097492534, + "grad_norm": 1.8122657537460327, + "learning_rate": 2.890703900837125e-06, + "loss": 3.1565, + "step": 18937 + }, + { + "epoch": 1.518249784530276, + "grad_norm": 1.7916854619979858, + "learning_rate": 2.8897908107734594e-06, + "loss": 3.1349, + "step": 18938 + }, + { + "epoch": 1.5183299593112987, + "grad_norm": 1.8376638889312744, + "learning_rate": 2.8888778405854155e-06, + "loss": 3.1686, + "step": 18939 + }, + { + "epoch": 1.5184101340923213, + "grad_norm": 1.7673664093017578, + "learning_rate": 2.8879649902883855e-06, + "loss": 3.2299, + "step": 18940 + }, + { + "epoch": 1.518490308873344, + "grad_norm": 1.938106656074524, + "learning_rate": 2.887052259897758e-06, + "loss": 3.235, + "step": 18941 + }, + { + "epoch": 1.5185704836543665, + "grad_norm": 1.7297770977020264, + "learning_rate": 2.8861396494289264e-06, + "loss": 3.2186, + "step": 18942 + }, + { + "epoch": 1.5186506584353892, + "grad_norm": 1.7657076120376587, + "learning_rate": 2.8852271588972704e-06, + "loss": 3.1449, + "step": 18943 + }, + { + "epoch": 1.5187308332164118, + "grad_norm": 1.7149428129196167, + "learning_rate": 2.8843147883181764e-06, + "loss": 3.2415, + "step": 18944 + }, + { + "epoch": 1.5188110079974344, + "grad_norm": 1.6785902976989746, + "learning_rate": 2.8834025377070273e-06, + "loss": 3.1077, + "step": 18945 + }, + { + "epoch": 1.518891182778457, + "grad_norm": 1.8516865968704224, + "learning_rate": 2.882490407079204e-06, + "loss": 3.15, + "step": 18946 + }, + { + "epoch": 1.5189713575594797, + "grad_norm": 1.7835369110107422, + "learning_rate": 2.8815783964500867e-06, + "loss": 3.2308, + "step": 18947 + }, + { + "epoch": 1.5190515323405023, + "grad_norm": 1.7101404666900635, + "learning_rate": 2.8806665058350446e-06, + "loss": 3.145, + "step": 18948 + }, + { + "epoch": 1.519131707121525, + "grad_norm": 1.6613577604293823, + "learning_rate": 2.879754735249457e-06, + "loss": 3.0631, + "step": 18949 + }, + { + "epoch": 1.5192118819025475, + "grad_norm": 1.6998953819274902, + "learning_rate": 2.878843084708696e-06, + "loss": 3.1788, + "step": 18950 + }, + { + "epoch": 1.5192920566835701, + "grad_norm": 1.7714316844940186, + "learning_rate": 2.8779315542281296e-06, + "loss": 3.2167, + "step": 18951 + }, + { + "epoch": 1.5193722314645928, + "grad_norm": 1.8530104160308838, + "learning_rate": 2.8770201438231306e-06, + "loss": 3.1539, + "step": 18952 + }, + { + "epoch": 1.5194524062456154, + "grad_norm": 1.7458122968673706, + "learning_rate": 2.876108853509059e-06, + "loss": 3.1099, + "step": 18953 + }, + { + "epoch": 1.519532581026638, + "grad_norm": 1.7620418071746826, + "learning_rate": 2.8751976833012805e-06, + "loss": 3.1016, + "step": 18954 + }, + { + "epoch": 1.5196127558076609, + "grad_norm": 1.7688685655593872, + "learning_rate": 2.8742866332151585e-06, + "loss": 3.2794, + "step": 18955 + }, + { + "epoch": 1.5196929305886835, + "grad_norm": 1.7379997968673706, + "learning_rate": 2.8733757032660526e-06, + "loss": 3.2146, + "step": 18956 + }, + { + "epoch": 1.5197731053697061, + "grad_norm": 1.737676978111267, + "learning_rate": 2.8724648934693245e-06, + "loss": 3.1508, + "step": 18957 + }, + { + "epoch": 1.5198532801507287, + "grad_norm": 1.7325003147125244, + "learning_rate": 2.871554203840322e-06, + "loss": 3.3062, + "step": 18958 + }, + { + "epoch": 1.5199334549317514, + "grad_norm": 0.6591764092445374, + "learning_rate": 2.8706436343944033e-06, + "loss": 2.4673, + "step": 18959 + }, + { + "epoch": 1.520013629712774, + "grad_norm": 1.8329495191574097, + "learning_rate": 2.869733185146921e-06, + "loss": 3.1896, + "step": 18960 + }, + { + "epoch": 1.5200938044937966, + "grad_norm": 1.8373243808746338, + "learning_rate": 2.8688228561132236e-06, + "loss": 3.311, + "step": 18961 + }, + { + "epoch": 1.5201739792748192, + "grad_norm": 1.7454955577850342, + "learning_rate": 2.867912647308664e-06, + "loss": 3.1196, + "step": 18962 + }, + { + "epoch": 1.5202541540558419, + "grad_norm": 1.796905517578125, + "learning_rate": 2.8670025587485793e-06, + "loss": 3.2385, + "step": 18963 + }, + { + "epoch": 1.5203343288368645, + "grad_norm": 1.7771738767623901, + "learning_rate": 2.866092590448317e-06, + "loss": 3.2026, + "step": 18964 + }, + { + "epoch": 1.520414503617887, + "grad_norm": 1.6905485391616821, + "learning_rate": 2.8651827424232204e-06, + "loss": 3.1508, + "step": 18965 + }, + { + "epoch": 1.5204946783989097, + "grad_norm": 0.6790682673454285, + "learning_rate": 2.864273014688629e-06, + "loss": 2.4639, + "step": 18966 + }, + { + "epoch": 1.5205748531799324, + "grad_norm": 1.8110963106155396, + "learning_rate": 2.8633634072598815e-06, + "loss": 3.2247, + "step": 18967 + }, + { + "epoch": 1.520655027960955, + "grad_norm": 1.8136250972747803, + "learning_rate": 2.8624539201523082e-06, + "loss": 3.1412, + "step": 18968 + }, + { + "epoch": 1.5207352027419776, + "grad_norm": 1.8496026992797852, + "learning_rate": 2.861544553381248e-06, + "loss": 3.1957, + "step": 18969 + }, + { + "epoch": 1.5208153775230002, + "grad_norm": 1.8643568754196167, + "learning_rate": 2.8606353069620297e-06, + "loss": 3.2319, + "step": 18970 + }, + { + "epoch": 1.5208955523040228, + "grad_norm": 1.8778815269470215, + "learning_rate": 2.859726180909984e-06, + "loss": 3.2368, + "step": 18971 + }, + { + "epoch": 1.5209757270850455, + "grad_norm": 1.7174288034439087, + "learning_rate": 2.8588171752404415e-06, + "loss": 3.1014, + "step": 18972 + }, + { + "epoch": 1.521055901866068, + "grad_norm": 1.8170102834701538, + "learning_rate": 2.8579082899687216e-06, + "loss": 3.1699, + "step": 18973 + }, + { + "epoch": 1.5211360766470907, + "grad_norm": 1.7107951641082764, + "learning_rate": 2.8569995251101513e-06, + "loss": 3.176, + "step": 18974 + }, + { + "epoch": 1.5212162514281133, + "grad_norm": 1.7598522901535034, + "learning_rate": 2.8560908806800514e-06, + "loss": 3.2665, + "step": 18975 + }, + { + "epoch": 1.521296426209136, + "grad_norm": 1.759803056716919, + "learning_rate": 2.855182356693741e-06, + "loss": 3.0881, + "step": 18976 + }, + { + "epoch": 1.5213766009901586, + "grad_norm": 1.787809133529663, + "learning_rate": 2.854273953166541e-06, + "loss": 3.1842, + "step": 18977 + }, + { + "epoch": 1.5214567757711812, + "grad_norm": 1.7259244918823242, + "learning_rate": 2.8533656701137615e-06, + "loss": 3.1889, + "step": 18978 + }, + { + "epoch": 1.5215369505522038, + "grad_norm": 1.8734768629074097, + "learning_rate": 2.8524575075507166e-06, + "loss": 3.1329, + "step": 18979 + }, + { + "epoch": 1.5216171253332265, + "grad_norm": 1.7695562839508057, + "learning_rate": 2.8515494654927202e-06, + "loss": 3.1814, + "step": 18980 + }, + { + "epoch": 1.521697300114249, + "grad_norm": 1.7807179689407349, + "learning_rate": 2.8506415439550804e-06, + "loss": 3.175, + "step": 18981 + }, + { + "epoch": 1.5217774748952717, + "grad_norm": 1.7272979021072388, + "learning_rate": 2.849733742953108e-06, + "loss": 3.2305, + "step": 18982 + }, + { + "epoch": 1.5218576496762943, + "grad_norm": 1.9301663637161255, + "learning_rate": 2.8488260625021004e-06, + "loss": 3.1228, + "step": 18983 + }, + { + "epoch": 1.521937824457317, + "grad_norm": 1.765130639076233, + "learning_rate": 2.847918502617366e-06, + "loss": 3.1979, + "step": 18984 + }, + { + "epoch": 1.5220179992383396, + "grad_norm": 1.8780235052108765, + "learning_rate": 2.8470110633142043e-06, + "loss": 3.2188, + "step": 18985 + }, + { + "epoch": 1.5220981740193622, + "grad_norm": 1.6843181848526, + "learning_rate": 2.8461037446079166e-06, + "loss": 3.1339, + "step": 18986 + }, + { + "epoch": 1.5221783488003848, + "grad_norm": 1.7119901180267334, + "learning_rate": 2.8451965465138e-06, + "loss": 3.1598, + "step": 18987 + }, + { + "epoch": 1.5222585235814075, + "grad_norm": 1.7982349395751953, + "learning_rate": 2.8442894690471457e-06, + "loss": 3.1501, + "step": 18988 + }, + { + "epoch": 1.52233869836243, + "grad_norm": 1.6773016452789307, + "learning_rate": 2.8433825122232485e-06, + "loss": 3.1223, + "step": 18989 + }, + { + "epoch": 1.5224188731434527, + "grad_norm": 1.7325561046600342, + "learning_rate": 2.8424756760574e-06, + "loss": 3.1867, + "step": 18990 + }, + { + "epoch": 1.5224990479244753, + "grad_norm": 1.7655892372131348, + "learning_rate": 2.8415689605648898e-06, + "loss": 3.1548, + "step": 18991 + }, + { + "epoch": 1.522579222705498, + "grad_norm": 1.8068560361862183, + "learning_rate": 2.8406623657610065e-06, + "loss": 3.1426, + "step": 18992 + }, + { + "epoch": 1.5226593974865206, + "grad_norm": 1.7289788722991943, + "learning_rate": 2.83975589166103e-06, + "loss": 3.1891, + "step": 18993 + }, + { + "epoch": 1.5227395722675432, + "grad_norm": 1.909729242324829, + "learning_rate": 2.838849538280245e-06, + "loss": 3.1543, + "step": 18994 + }, + { + "epoch": 1.5228197470485658, + "grad_norm": 1.7310879230499268, + "learning_rate": 2.8379433056339333e-06, + "loss": 3.2074, + "step": 18995 + }, + { + "epoch": 1.5228999218295884, + "grad_norm": 1.7241508960723877, + "learning_rate": 2.837037193737373e-06, + "loss": 3.1367, + "step": 18996 + }, + { + "epoch": 1.522980096610611, + "grad_norm": 1.7286620140075684, + "learning_rate": 2.8361312026058443e-06, + "loss": 3.1478, + "step": 18997 + }, + { + "epoch": 1.5230602713916337, + "grad_norm": 1.8605881929397583, + "learning_rate": 2.835225332254615e-06, + "loss": 3.1234, + "step": 18998 + }, + { + "epoch": 1.5231404461726563, + "grad_norm": 1.755916953086853, + "learning_rate": 2.8343195826989633e-06, + "loss": 3.0986, + "step": 18999 + }, + { + "epoch": 1.523220620953679, + "grad_norm": 1.778427004814148, + "learning_rate": 2.833413953954156e-06, + "loss": 3.0631, + "step": 19000 + }, + { + "epoch": 1.5233007957347016, + "grad_norm": 1.8072112798690796, + "learning_rate": 2.8325084460354656e-06, + "loss": 3.1857, + "step": 19001 + }, + { + "epoch": 1.5233809705157242, + "grad_norm": 1.744641661643982, + "learning_rate": 2.831603058958159e-06, + "loss": 3.1318, + "step": 19002 + }, + { + "epoch": 1.5234611452967468, + "grad_norm": 1.6301497220993042, + "learning_rate": 2.8306977927374945e-06, + "loss": 3.1448, + "step": 19003 + }, + { + "epoch": 1.5235413200777694, + "grad_norm": 1.8465436697006226, + "learning_rate": 2.8297926473887395e-06, + "loss": 3.1693, + "step": 19004 + }, + { + "epoch": 1.523621494858792, + "grad_norm": 0.6669951677322388, + "learning_rate": 2.8288876229271533e-06, + "loss": 2.4646, + "step": 19005 + }, + { + "epoch": 1.5237016696398147, + "grad_norm": 1.7982821464538574, + "learning_rate": 2.8279827193679943e-06, + "loss": 3.1652, + "step": 19006 + }, + { + "epoch": 1.5237818444208373, + "grad_norm": 1.8742082118988037, + "learning_rate": 2.827077936726522e-06, + "loss": 3.11, + "step": 19007 + }, + { + "epoch": 1.52386201920186, + "grad_norm": 1.7623798847198486, + "learning_rate": 2.826173275017985e-06, + "loss": 3.2213, + "step": 19008 + }, + { + "epoch": 1.5239421939828826, + "grad_norm": 1.8142526149749756, + "learning_rate": 2.8252687342576367e-06, + "loss": 3.2111, + "step": 19009 + }, + { + "epoch": 1.5240223687639052, + "grad_norm": 1.6689977645874023, + "learning_rate": 2.8243643144607304e-06, + "loss": 3.1818, + "step": 19010 + }, + { + "epoch": 1.5241025435449278, + "grad_norm": 1.8145750761032104, + "learning_rate": 2.823460015642513e-06, + "loss": 3.1797, + "step": 19011 + }, + { + "epoch": 1.5241827183259504, + "grad_norm": 1.8190997838974, + "learning_rate": 2.8225558378182326e-06, + "loss": 3.1608, + "step": 19012 + }, + { + "epoch": 1.524262893106973, + "grad_norm": 1.8025842905044556, + "learning_rate": 2.8216517810031297e-06, + "loss": 3.2286, + "step": 19013 + }, + { + "epoch": 1.5243430678879957, + "grad_norm": 1.7283121347427368, + "learning_rate": 2.8207478452124458e-06, + "loss": 3.136, + "step": 19014 + }, + { + "epoch": 1.5244232426690183, + "grad_norm": 1.6423351764678955, + "learning_rate": 2.819844030461425e-06, + "loss": 3.027, + "step": 19015 + }, + { + "epoch": 1.5245034174500411, + "grad_norm": 1.7366191148757935, + "learning_rate": 2.8189403367653023e-06, + "loss": 3.1963, + "step": 19016 + }, + { + "epoch": 1.5245835922310638, + "grad_norm": 1.8331077098846436, + "learning_rate": 2.818036764139317e-06, + "loss": 3.2029, + "step": 19017 + }, + { + "epoch": 1.5246637670120864, + "grad_norm": 1.8005837202072144, + "learning_rate": 2.817133312598699e-06, + "loss": 3.1024, + "step": 19018 + }, + { + "epoch": 1.524743941793109, + "grad_norm": 1.9242775440216064, + "learning_rate": 2.8162299821586814e-06, + "loss": 3.2722, + "step": 19019 + }, + { + "epoch": 1.5248241165741316, + "grad_norm": 1.7109054327011108, + "learning_rate": 2.8153267728344936e-06, + "loss": 3.1277, + "step": 19020 + }, + { + "epoch": 1.5249042913551543, + "grad_norm": 1.7190628051757812, + "learning_rate": 2.8144236846413653e-06, + "loss": 3.1482, + "step": 19021 + }, + { + "epoch": 1.5249844661361769, + "grad_norm": 1.6200261116027832, + "learning_rate": 2.8135207175945243e-06, + "loss": 3.1528, + "step": 19022 + }, + { + "epoch": 1.5250646409171995, + "grad_norm": 1.7706108093261719, + "learning_rate": 2.8126178717091878e-06, + "loss": 3.1521, + "step": 19023 + }, + { + "epoch": 1.5251448156982221, + "grad_norm": 1.7737699747085571, + "learning_rate": 2.811715147000581e-06, + "loss": 3.2012, + "step": 19024 + }, + { + "epoch": 1.5252249904792448, + "grad_norm": 1.7180461883544922, + "learning_rate": 2.810812543483924e-06, + "loss": 3.0871, + "step": 19025 + }, + { + "epoch": 1.5253051652602674, + "grad_norm": 1.8604052066802979, + "learning_rate": 2.8099100611744333e-06, + "loss": 3.276, + "step": 19026 + }, + { + "epoch": 1.52538534004129, + "grad_norm": 1.8247358798980713, + "learning_rate": 2.809007700087327e-06, + "loss": 3.1329, + "step": 19027 + }, + { + "epoch": 1.5254655148223126, + "grad_norm": 1.767632246017456, + "learning_rate": 2.8081054602378157e-06, + "loss": 3.2391, + "step": 19028 + }, + { + "epoch": 1.5255456896033353, + "grad_norm": 1.6910111904144287, + "learning_rate": 2.8072033416411092e-06, + "loss": 3.1456, + "step": 19029 + }, + { + "epoch": 1.5256258643843579, + "grad_norm": 1.880305528640747, + "learning_rate": 2.806301344312421e-06, + "loss": 3.1887, + "step": 19030 + }, + { + "epoch": 1.5257060391653805, + "grad_norm": 1.9229881763458252, + "learning_rate": 2.805399468266956e-06, + "loss": 3.2533, + "step": 19031 + }, + { + "epoch": 1.5257862139464031, + "grad_norm": 1.7964847087860107, + "learning_rate": 2.8044977135199246e-06, + "loss": 3.0388, + "step": 19032 + }, + { + "epoch": 1.5258663887274257, + "grad_norm": 1.8037203550338745, + "learning_rate": 2.8035960800865214e-06, + "loss": 3.1843, + "step": 19033 + }, + { + "epoch": 1.5259465635084484, + "grad_norm": 1.919115424156189, + "learning_rate": 2.8026945679819528e-06, + "loss": 3.2744, + "step": 19034 + }, + { + "epoch": 1.526026738289471, + "grad_norm": 2.0255401134490967, + "learning_rate": 2.8017931772214167e-06, + "loss": 3.1564, + "step": 19035 + }, + { + "epoch": 1.5261069130704938, + "grad_norm": 1.7619320154190063, + "learning_rate": 2.8008919078201114e-06, + "loss": 3.1305, + "step": 19036 + }, + { + "epoch": 1.5261870878515165, + "grad_norm": 1.7003206014633179, + "learning_rate": 2.799990759793234e-06, + "loss": 3.1775, + "step": 19037 + }, + { + "epoch": 1.526267262632539, + "grad_norm": 1.762978196144104, + "learning_rate": 2.7990897331559717e-06, + "loss": 3.1541, + "step": 19038 + }, + { + "epoch": 1.5263474374135617, + "grad_norm": 1.7248635292053223, + "learning_rate": 2.798188827923519e-06, + "loss": 3.2562, + "step": 19039 + }, + { + "epoch": 1.5264276121945843, + "grad_norm": 1.728267788887024, + "learning_rate": 2.797288044111064e-06, + "loss": 3.1237, + "step": 19040 + }, + { + "epoch": 1.526507786975607, + "grad_norm": 1.755320429801941, + "learning_rate": 2.796387381733795e-06, + "loss": 3.1027, + "step": 19041 + }, + { + "epoch": 1.5265879617566296, + "grad_norm": 1.7436320781707764, + "learning_rate": 2.795486840806898e-06, + "loss": 3.1676, + "step": 19042 + }, + { + "epoch": 1.5266681365376522, + "grad_norm": 1.8104774951934814, + "learning_rate": 2.794586421345551e-06, + "loss": 3.1698, + "step": 19043 + }, + { + "epoch": 1.5267483113186748, + "grad_norm": 1.769838571548462, + "learning_rate": 2.793686123364937e-06, + "loss": 3.1118, + "step": 19044 + }, + { + "epoch": 1.5268284860996975, + "grad_norm": 0.750227153301239, + "learning_rate": 2.7927859468802364e-06, + "loss": 2.5059, + "step": 19045 + }, + { + "epoch": 1.52690866088072, + "grad_norm": 1.9358947277069092, + "learning_rate": 2.791885891906624e-06, + "loss": 3.2051, + "step": 19046 + }, + { + "epoch": 1.5269888356617427, + "grad_norm": 1.7253822088241577, + "learning_rate": 2.790985958459278e-06, + "loss": 3.1854, + "step": 19047 + }, + { + "epoch": 1.5270690104427653, + "grad_norm": 1.9008060693740845, + "learning_rate": 2.7900861465533646e-06, + "loss": 3.3031, + "step": 19048 + }, + { + "epoch": 1.527149185223788, + "grad_norm": 1.8488067388534546, + "learning_rate": 2.789186456204058e-06, + "loss": 3.1819, + "step": 19049 + }, + { + "epoch": 1.5272293600048106, + "grad_norm": 1.7570277452468872, + "learning_rate": 2.7882868874265277e-06, + "loss": 3.1456, + "step": 19050 + }, + { + "epoch": 1.5273095347858332, + "grad_norm": 1.783291220664978, + "learning_rate": 2.787387440235938e-06, + "loss": 3.2297, + "step": 19051 + }, + { + "epoch": 1.5273897095668558, + "grad_norm": 0.6539521217346191, + "learning_rate": 2.7864881146474565e-06, + "loss": 2.4688, + "step": 19052 + }, + { + "epoch": 1.5274698843478784, + "grad_norm": 1.7128709554672241, + "learning_rate": 2.78558891067624e-06, + "loss": 3.1583, + "step": 19053 + }, + { + "epoch": 1.527550059128901, + "grad_norm": 1.8170480728149414, + "learning_rate": 2.7846898283374534e-06, + "loss": 3.1338, + "step": 19054 + }, + { + "epoch": 1.5276302339099237, + "grad_norm": 1.773783802986145, + "learning_rate": 2.783790867646252e-06, + "loss": 3.1215, + "step": 19055 + }, + { + "epoch": 1.5277104086909463, + "grad_norm": 1.8330014944076538, + "learning_rate": 2.782892028617794e-06, + "loss": 3.0969, + "step": 19056 + }, + { + "epoch": 1.527790583471969, + "grad_norm": 1.8098771572113037, + "learning_rate": 2.781993311267235e-06, + "loss": 3.1504, + "step": 19057 + }, + { + "epoch": 1.5278707582529916, + "grad_norm": 1.7323211431503296, + "learning_rate": 2.781094715609722e-06, + "loss": 3.1123, + "step": 19058 + }, + { + "epoch": 1.5279509330340142, + "grad_norm": 1.7583106756210327, + "learning_rate": 2.780196241660409e-06, + "loss": 3.1864, + "step": 19059 + }, + { + "epoch": 1.5280311078150368, + "grad_norm": 1.7137912511825562, + "learning_rate": 2.7792978894344424e-06, + "loss": 3.1161, + "step": 19060 + }, + { + "epoch": 1.5281112825960594, + "grad_norm": 1.7964825630187988, + "learning_rate": 2.7783996589469674e-06, + "loss": 3.1658, + "step": 19061 + }, + { + "epoch": 1.528191457377082, + "grad_norm": 1.769634485244751, + "learning_rate": 2.7775015502131342e-06, + "loss": 3.2274, + "step": 19062 + }, + { + "epoch": 1.5282716321581047, + "grad_norm": 1.7278445959091187, + "learning_rate": 2.776603563248075e-06, + "loss": 3.1886, + "step": 19063 + }, + { + "epoch": 1.5283518069391273, + "grad_norm": 1.7648853063583374, + "learning_rate": 2.775705698066934e-06, + "loss": 3.1622, + "step": 19064 + }, + { + "epoch": 1.52843198172015, + "grad_norm": 1.8946645259857178, + "learning_rate": 2.7748079546848485e-06, + "loss": 3.2239, + "step": 19065 + }, + { + "epoch": 1.5285121565011726, + "grad_norm": 1.9026790857315063, + "learning_rate": 2.773910333116955e-06, + "loss": 3.1556, + "step": 19066 + }, + { + "epoch": 1.5285923312821952, + "grad_norm": 1.7539950609207153, + "learning_rate": 2.77301283337839e-06, + "loss": 3.2137, + "step": 19067 + }, + { + "epoch": 1.5286725060632178, + "grad_norm": 1.8469551801681519, + "learning_rate": 2.772115455484278e-06, + "loss": 3.2297, + "step": 19068 + }, + { + "epoch": 1.5287526808442404, + "grad_norm": 0.6791603565216064, + "learning_rate": 2.7712181994497524e-06, + "loss": 2.4802, + "step": 19069 + }, + { + "epoch": 1.528832855625263, + "grad_norm": 1.703080415725708, + "learning_rate": 2.7703210652899394e-06, + "loss": 3.2316, + "step": 19070 + }, + { + "epoch": 1.5289130304062857, + "grad_norm": 1.7600691318511963, + "learning_rate": 2.7694240530199654e-06, + "loss": 3.1696, + "step": 19071 + }, + { + "epoch": 1.5289932051873083, + "grad_norm": 1.8159505128860474, + "learning_rate": 2.768527162654958e-06, + "loss": 3.2089, + "step": 19072 + }, + { + "epoch": 1.529073379968331, + "grad_norm": 1.7585951089859009, + "learning_rate": 2.76763039421003e-06, + "loss": 3.1026, + "step": 19073 + }, + { + "epoch": 1.5291535547493535, + "grad_norm": 1.8615391254425049, + "learning_rate": 2.7667337477003042e-06, + "loss": 3.1798, + "step": 19074 + }, + { + "epoch": 1.5292337295303762, + "grad_norm": 1.790583610534668, + "learning_rate": 2.765837223140899e-06, + "loss": 3.1794, + "step": 19075 + }, + { + "epoch": 1.5293139043113988, + "grad_norm": 1.6803072690963745, + "learning_rate": 2.7649408205469286e-06, + "loss": 3.194, + "step": 19076 + }, + { + "epoch": 1.5293940790924214, + "grad_norm": 1.7271841764450073, + "learning_rate": 2.7640445399335083e-06, + "loss": 3.1578, + "step": 19077 + }, + { + "epoch": 1.529474253873444, + "grad_norm": 1.7186604738235474, + "learning_rate": 2.7631483813157446e-06, + "loss": 3.22, + "step": 19078 + }, + { + "epoch": 1.5295544286544667, + "grad_norm": 1.8136894702911377, + "learning_rate": 2.7622523447087478e-06, + "loss": 3.1279, + "step": 19079 + }, + { + "epoch": 1.5296346034354893, + "grad_norm": 1.826435923576355, + "learning_rate": 2.7613564301276276e-06, + "loss": 3.2924, + "step": 19080 + }, + { + "epoch": 1.529714778216512, + "grad_norm": 1.7773191928863525, + "learning_rate": 2.7604606375874854e-06, + "loss": 3.1453, + "step": 19081 + }, + { + "epoch": 1.5297949529975345, + "grad_norm": 1.6496738195419312, + "learning_rate": 2.7595649671034273e-06, + "loss": 3.1337, + "step": 19082 + }, + { + "epoch": 1.5298751277785572, + "grad_norm": 1.7975322008132935, + "learning_rate": 2.7586694186905506e-06, + "loss": 3.1421, + "step": 19083 + }, + { + "epoch": 1.5299553025595798, + "grad_norm": 1.7360953092575073, + "learning_rate": 2.757773992363956e-06, + "loss": 3.2205, + "step": 19084 + }, + { + "epoch": 1.5300354773406024, + "grad_norm": 1.9411725997924805, + "learning_rate": 2.7568786881387376e-06, + "loss": 3.1304, + "step": 19085 + }, + { + "epoch": 1.530115652121625, + "grad_norm": 1.7962098121643066, + "learning_rate": 2.7559835060299935e-06, + "loss": 3.1834, + "step": 19086 + }, + { + "epoch": 1.5301958269026477, + "grad_norm": 1.9473466873168945, + "learning_rate": 2.755088446052816e-06, + "loss": 3.1819, + "step": 19087 + }, + { + "epoch": 1.5302760016836703, + "grad_norm": 1.7472295761108398, + "learning_rate": 2.7541935082222928e-06, + "loss": 3.251, + "step": 19088 + }, + { + "epoch": 1.530356176464693, + "grad_norm": 1.8264912366867065, + "learning_rate": 2.753298692553512e-06, + "loss": 3.2219, + "step": 19089 + }, + { + "epoch": 1.5304363512457155, + "grad_norm": 1.8194174766540527, + "learning_rate": 2.7524039990615614e-06, + "loss": 3.2305, + "step": 19090 + }, + { + "epoch": 1.5305165260267382, + "grad_norm": 1.8164888620376587, + "learning_rate": 2.7515094277615252e-06, + "loss": 3.1993, + "step": 19091 + }, + { + "epoch": 1.5305967008077608, + "grad_norm": 1.764451503753662, + "learning_rate": 2.7506149786684876e-06, + "loss": 3.1725, + "step": 19092 + }, + { + "epoch": 1.5306768755887834, + "grad_norm": 0.6937307119369507, + "learning_rate": 2.749720651797525e-06, + "loss": 2.4565, + "step": 19093 + }, + { + "epoch": 1.530757050369806, + "grad_norm": 1.9176136255264282, + "learning_rate": 2.748826447163716e-06, + "loss": 3.2002, + "step": 19094 + }, + { + "epoch": 1.5308372251508287, + "grad_norm": 1.7886364459991455, + "learning_rate": 2.7479323647821376e-06, + "loss": 3.2428, + "step": 19095 + }, + { + "epoch": 1.5309173999318513, + "grad_norm": 1.7871177196502686, + "learning_rate": 2.7470384046678644e-06, + "loss": 3.1642, + "step": 19096 + }, + { + "epoch": 1.5309975747128741, + "grad_norm": 1.763266921043396, + "learning_rate": 2.74614456683597e-06, + "loss": 3.2124, + "step": 19097 + }, + { + "epoch": 1.5310777494938967, + "grad_norm": 1.7561054229736328, + "learning_rate": 2.7452508513015186e-06, + "loss": 3.1124, + "step": 19098 + }, + { + "epoch": 1.5311579242749194, + "grad_norm": 1.8391239643096924, + "learning_rate": 2.7443572580795805e-06, + "loss": 3.1623, + "step": 19099 + }, + { + "epoch": 1.531238099055942, + "grad_norm": 1.7708760499954224, + "learning_rate": 2.7434637871852233e-06, + "loss": 3.1378, + "step": 19100 + }, + { + "epoch": 1.5313182738369646, + "grad_norm": 1.7918108701705933, + "learning_rate": 2.7425704386335086e-06, + "loss": 3.1371, + "step": 19101 + }, + { + "epoch": 1.5313984486179872, + "grad_norm": 1.77410888671875, + "learning_rate": 2.7416772124395007e-06, + "loss": 3.1621, + "step": 19102 + }, + { + "epoch": 1.5314786233990099, + "grad_norm": 0.6858269572257996, + "learning_rate": 2.7407841086182553e-06, + "loss": 2.4602, + "step": 19103 + }, + { + "epoch": 1.5315587981800325, + "grad_norm": 1.839072346687317, + "learning_rate": 2.7398911271848305e-06, + "loss": 3.1562, + "step": 19104 + }, + { + "epoch": 1.5316389729610551, + "grad_norm": 1.7006372213363647, + "learning_rate": 2.7389982681542824e-06, + "loss": 3.2507, + "step": 19105 + }, + { + "epoch": 1.5317191477420777, + "grad_norm": 1.8713083267211914, + "learning_rate": 2.738105531541665e-06, + "loss": 3.1615, + "step": 19106 + }, + { + "epoch": 1.5317993225231004, + "grad_norm": 1.8054791688919067, + "learning_rate": 2.737212917362032e-06, + "loss": 3.3362, + "step": 19107 + }, + { + "epoch": 1.531879497304123, + "grad_norm": 1.6699752807617188, + "learning_rate": 2.7363204256304265e-06, + "loss": 3.1487, + "step": 19108 + }, + { + "epoch": 1.5319596720851456, + "grad_norm": 1.730455994606018, + "learning_rate": 2.7354280563618984e-06, + "loss": 3.2123, + "step": 19109 + }, + { + "epoch": 1.5320398468661682, + "grad_norm": 1.7733185291290283, + "learning_rate": 2.7345358095714936e-06, + "loss": 3.1696, + "step": 19110 + }, + { + "epoch": 1.5321200216471909, + "grad_norm": 0.7116353511810303, + "learning_rate": 2.733643685274253e-06, + "loss": 2.4641, + "step": 19111 + }, + { + "epoch": 1.5322001964282135, + "grad_norm": 1.9127247333526611, + "learning_rate": 2.732751683485224e-06, + "loss": 3.2106, + "step": 19112 + }, + { + "epoch": 1.532280371209236, + "grad_norm": 1.8535468578338623, + "learning_rate": 2.7318598042194357e-06, + "loss": 3.0564, + "step": 19113 + }, + { + "epoch": 1.5323605459902587, + "grad_norm": 1.7681794166564941, + "learning_rate": 2.730968047491931e-06, + "loss": 3.2566, + "step": 19114 + }, + { + "epoch": 1.5324407207712814, + "grad_norm": 1.726241111755371, + "learning_rate": 2.7300764133177413e-06, + "loss": 3.1245, + "step": 19115 + }, + { + "epoch": 1.532520895552304, + "grad_norm": 1.8021074533462524, + "learning_rate": 2.729184901711903e-06, + "loss": 3.002, + "step": 19116 + }, + { + "epoch": 1.5326010703333266, + "grad_norm": 1.7938008308410645, + "learning_rate": 2.728293512689447e-06, + "loss": 3.2153, + "step": 19117 + }, + { + "epoch": 1.5326812451143494, + "grad_norm": 1.8473312854766846, + "learning_rate": 2.727402246265396e-06, + "loss": 3.2207, + "step": 19118 + }, + { + "epoch": 1.532761419895372, + "grad_norm": 1.7830593585968018, + "learning_rate": 2.7265111024547808e-06, + "loss": 3.1463, + "step": 19119 + }, + { + "epoch": 1.5328415946763947, + "grad_norm": 1.6909147500991821, + "learning_rate": 2.7256200812726253e-06, + "loss": 3.1854, + "step": 19120 + }, + { + "epoch": 1.5329217694574173, + "grad_norm": 1.7162673473358154, + "learning_rate": 2.7247291827339508e-06, + "loss": 3.1385, + "step": 19121 + }, + { + "epoch": 1.53300194423844, + "grad_norm": 1.7798523902893066, + "learning_rate": 2.7238384068537815e-06, + "loss": 3.3011, + "step": 19122 + }, + { + "epoch": 1.5330821190194626, + "grad_norm": 1.7362957000732422, + "learning_rate": 2.722947753647128e-06, + "loss": 3.2568, + "step": 19123 + }, + { + "epoch": 1.5331622938004852, + "grad_norm": 1.7516047954559326, + "learning_rate": 2.722057223129012e-06, + "loss": 3.1332, + "step": 19124 + }, + { + "epoch": 1.5332424685815078, + "grad_norm": 1.8011242151260376, + "learning_rate": 2.7211668153144465e-06, + "loss": 3.2144, + "step": 19125 + }, + { + "epoch": 1.5333226433625304, + "grad_norm": 1.731365442276001, + "learning_rate": 2.7202765302184432e-06, + "loss": 3.1936, + "step": 19126 + }, + { + "epoch": 1.533402818143553, + "grad_norm": 1.705886721611023, + "learning_rate": 2.719386367856014e-06, + "loss": 3.1314, + "step": 19127 + }, + { + "epoch": 1.5334829929245757, + "grad_norm": 1.6517095565795898, + "learning_rate": 2.7184963282421626e-06, + "loss": 3.1216, + "step": 19128 + }, + { + "epoch": 1.5335631677055983, + "grad_norm": 1.7844529151916504, + "learning_rate": 2.7176064113918964e-06, + "loss": 3.2554, + "step": 19129 + }, + { + "epoch": 1.533643342486621, + "grad_norm": 1.7872743606567383, + "learning_rate": 2.7167166173202197e-06, + "loss": 3.1667, + "step": 19130 + }, + { + "epoch": 1.5337235172676436, + "grad_norm": 1.796498417854309, + "learning_rate": 2.715826946042134e-06, + "loss": 3.1774, + "step": 19131 + }, + { + "epoch": 1.5338036920486662, + "grad_norm": 1.7964797019958496, + "learning_rate": 2.7149373975726414e-06, + "loss": 3.2528, + "step": 19132 + }, + { + "epoch": 1.5338838668296888, + "grad_norm": 1.710261583328247, + "learning_rate": 2.7140479719267343e-06, + "loss": 3.1788, + "step": 19133 + }, + { + "epoch": 1.5339640416107114, + "grad_norm": 1.794896125793457, + "learning_rate": 2.71315866911941e-06, + "loss": 3.2935, + "step": 19134 + }, + { + "epoch": 1.534044216391734, + "grad_norm": 1.6793566942214966, + "learning_rate": 2.712269489165663e-06, + "loss": 3.2416, + "step": 19135 + }, + { + "epoch": 1.5341243911727567, + "grad_norm": 1.6318062543869019, + "learning_rate": 2.7113804320804835e-06, + "loss": 3.2106, + "step": 19136 + }, + { + "epoch": 1.5342045659537793, + "grad_norm": 1.7870196104049683, + "learning_rate": 2.7104914978788645e-06, + "loss": 3.2565, + "step": 19137 + }, + { + "epoch": 1.534284740734802, + "grad_norm": 1.752050518989563, + "learning_rate": 2.7096026865757864e-06, + "loss": 3.1628, + "step": 19138 + }, + { + "epoch": 1.5343649155158245, + "grad_norm": 1.78346586227417, + "learning_rate": 2.7087139981862377e-06, + "loss": 3.1519, + "step": 19139 + }, + { + "epoch": 1.5344450902968472, + "grad_norm": 1.8475720882415771, + "learning_rate": 2.7078254327252017e-06, + "loss": 3.1766, + "step": 19140 + }, + { + "epoch": 1.5345252650778698, + "grad_norm": 1.9805324077606201, + "learning_rate": 2.706936990207658e-06, + "loss": 3.1892, + "step": 19141 + }, + { + "epoch": 1.5346054398588924, + "grad_norm": 1.7238147258758545, + "learning_rate": 2.70604867064859e-06, + "loss": 3.0792, + "step": 19142 + }, + { + "epoch": 1.534685614639915, + "grad_norm": 1.7403676509857178, + "learning_rate": 2.705160474062968e-06, + "loss": 3.2372, + "step": 19143 + }, + { + "epoch": 1.5347657894209377, + "grad_norm": 1.84440279006958, + "learning_rate": 2.7042724004657694e-06, + "loss": 3.1587, + "step": 19144 + }, + { + "epoch": 1.5348459642019603, + "grad_norm": 1.8075648546218872, + "learning_rate": 2.7033844498719663e-06, + "loss": 3.1218, + "step": 19145 + }, + { + "epoch": 1.534926138982983, + "grad_norm": 1.7842011451721191, + "learning_rate": 2.70249662229653e-06, + "loss": 3.223, + "step": 19146 + }, + { + "epoch": 1.5350063137640055, + "grad_norm": 1.7726362943649292, + "learning_rate": 2.7016089177544325e-06, + "loss": 3.1444, + "step": 19147 + }, + { + "epoch": 1.5350864885450282, + "grad_norm": 1.6980410814285278, + "learning_rate": 2.700721336260632e-06, + "loss": 3.0985, + "step": 19148 + }, + { + "epoch": 1.5351666633260508, + "grad_norm": 1.793965458869934, + "learning_rate": 2.699833877830099e-06, + "loss": 3.2928, + "step": 19149 + }, + { + "epoch": 1.5352468381070734, + "grad_norm": 1.8128387928009033, + "learning_rate": 2.698946542477793e-06, + "loss": 3.2189, + "step": 19150 + }, + { + "epoch": 1.535327012888096, + "grad_norm": 1.733959436416626, + "learning_rate": 2.6980593302186752e-06, + "loss": 3.2247, + "step": 19151 + }, + { + "epoch": 1.5354071876691187, + "grad_norm": 1.6877901554107666, + "learning_rate": 2.697172241067707e-06, + "loss": 3.0748, + "step": 19152 + }, + { + "epoch": 1.5354873624501413, + "grad_norm": 1.761454463005066, + "learning_rate": 2.696285275039837e-06, + "loss": 3.2164, + "step": 19153 + }, + { + "epoch": 1.535567537231164, + "grad_norm": 1.7428257465362549, + "learning_rate": 2.695398432150024e-06, + "loss": 3.249, + "step": 19154 + }, + { + "epoch": 1.5356477120121865, + "grad_norm": 1.7437655925750732, + "learning_rate": 2.694511712413219e-06, + "loss": 3.1886, + "step": 19155 + }, + { + "epoch": 1.5357278867932092, + "grad_norm": 1.9139209985733032, + "learning_rate": 2.6936251158443717e-06, + "loss": 3.249, + "step": 19156 + }, + { + "epoch": 1.5358080615742318, + "grad_norm": 2.295198917388916, + "learning_rate": 2.6927386424584336e-06, + "loss": 3.1578, + "step": 19157 + }, + { + "epoch": 1.5358882363552544, + "grad_norm": 1.7832131385803223, + "learning_rate": 2.6918522922703426e-06, + "loss": 3.1858, + "step": 19158 + }, + { + "epoch": 1.535968411136277, + "grad_norm": 0.6611799597740173, + "learning_rate": 2.6909660652950463e-06, + "loss": 2.4661, + "step": 19159 + }, + { + "epoch": 1.5360485859172996, + "grad_norm": 1.821302056312561, + "learning_rate": 2.690079961547487e-06, + "loss": 3.1919, + "step": 19160 + }, + { + "epoch": 1.5361287606983223, + "grad_norm": 1.8127152919769287, + "learning_rate": 2.6891939810426037e-06, + "loss": 3.2047, + "step": 19161 + }, + { + "epoch": 1.536208935479345, + "grad_norm": 1.675450325012207, + "learning_rate": 2.6883081237953357e-06, + "loss": 3.1022, + "step": 19162 + }, + { + "epoch": 1.5362891102603675, + "grad_norm": 0.662182092666626, + "learning_rate": 2.687422389820612e-06, + "loss": 2.4651, + "step": 19163 + }, + { + "epoch": 1.5363692850413901, + "grad_norm": 1.7234724760055542, + "learning_rate": 2.6865367791333706e-06, + "loss": 3.1849, + "step": 19164 + }, + { + "epoch": 1.5364494598224128, + "grad_norm": 1.8523707389831543, + "learning_rate": 2.6856512917485412e-06, + "loss": 3.2161, + "step": 19165 + }, + { + "epoch": 1.5365296346034354, + "grad_norm": 1.791029453277588, + "learning_rate": 2.6847659276810533e-06, + "loss": 3.1729, + "step": 19166 + }, + { + "epoch": 1.536609809384458, + "grad_norm": 1.729347586631775, + "learning_rate": 2.6838806869458367e-06, + "loss": 3.0582, + "step": 19167 + }, + { + "epoch": 1.5366899841654806, + "grad_norm": 1.7730870246887207, + "learning_rate": 2.6829955695578105e-06, + "loss": 3.1104, + "step": 19168 + }, + { + "epoch": 1.5367701589465033, + "grad_norm": 1.7183893918991089, + "learning_rate": 2.6821105755319e-06, + "loss": 3.1712, + "step": 19169 + }, + { + "epoch": 1.5368503337275259, + "grad_norm": 1.9261995553970337, + "learning_rate": 2.681225704883026e-06, + "loss": 3.205, + "step": 19170 + }, + { + "epoch": 1.5369305085085485, + "grad_norm": 1.8092010021209717, + "learning_rate": 2.6803409576261075e-06, + "loss": 3.0986, + "step": 19171 + }, + { + "epoch": 1.5370106832895711, + "grad_norm": 1.7552109956741333, + "learning_rate": 2.6794563337760637e-06, + "loss": 3.2025, + "step": 19172 + }, + { + "epoch": 1.5370908580705938, + "grad_norm": 1.8012980222702026, + "learning_rate": 2.6785718333478026e-06, + "loss": 3.112, + "step": 19173 + }, + { + "epoch": 1.5371710328516164, + "grad_norm": 1.809616208076477, + "learning_rate": 2.6776874563562405e-06, + "loss": 3.2067, + "step": 19174 + }, + { + "epoch": 1.537251207632639, + "grad_norm": 1.8916864395141602, + "learning_rate": 2.6768032028162872e-06, + "loss": 3.1631, + "step": 19175 + }, + { + "epoch": 1.5373313824136616, + "grad_norm": 1.762332558631897, + "learning_rate": 2.6759190727428507e-06, + "loss": 3.1424, + "step": 19176 + }, + { + "epoch": 1.5374115571946843, + "grad_norm": 1.7379978895187378, + "learning_rate": 2.6750350661508406e-06, + "loss": 3.1443, + "step": 19177 + }, + { + "epoch": 1.5374917319757069, + "grad_norm": 1.8555659055709839, + "learning_rate": 2.6741511830551548e-06, + "loss": 3.1442, + "step": 19178 + }, + { + "epoch": 1.5375719067567297, + "grad_norm": 0.6567903757095337, + "learning_rate": 2.673267423470699e-06, + "loss": 2.4567, + "step": 19179 + }, + { + "epoch": 1.5376520815377523, + "grad_norm": 1.8124487400054932, + "learning_rate": 2.6723837874123714e-06, + "loss": 3.1359, + "step": 19180 + }, + { + "epoch": 1.537732256318775, + "grad_norm": 1.920332908630371, + "learning_rate": 2.6715002748950704e-06, + "loss": 3.3157, + "step": 19181 + }, + { + "epoch": 1.5378124310997976, + "grad_norm": 1.6908336877822876, + "learning_rate": 2.6706168859336955e-06, + "loss": 3.1547, + "step": 19182 + }, + { + "epoch": 1.5378926058808202, + "grad_norm": 1.6676753759384155, + "learning_rate": 2.6697336205431336e-06, + "loss": 3.1934, + "step": 19183 + }, + { + "epoch": 1.5379727806618428, + "grad_norm": 1.8300237655639648, + "learning_rate": 2.6688504787382787e-06, + "loss": 3.1658, + "step": 19184 + }, + { + "epoch": 1.5380529554428655, + "grad_norm": 1.7085448503494263, + "learning_rate": 2.667967460534021e-06, + "loss": 3.2357, + "step": 19185 + }, + { + "epoch": 1.538133130223888, + "grad_norm": 1.7592732906341553, + "learning_rate": 2.6670845659452495e-06, + "loss": 3.2457, + "step": 19186 + }, + { + "epoch": 1.5382133050049107, + "grad_norm": 1.9326612949371338, + "learning_rate": 2.6662017949868492e-06, + "loss": 3.1747, + "step": 19187 + }, + { + "epoch": 1.5382934797859333, + "grad_norm": 1.7190947532653809, + "learning_rate": 2.6653191476736993e-06, + "loss": 3.2096, + "step": 19188 + }, + { + "epoch": 1.538373654566956, + "grad_norm": 1.857477068901062, + "learning_rate": 2.6644366240206843e-06, + "loss": 3.2977, + "step": 19189 + }, + { + "epoch": 1.5384538293479786, + "grad_norm": 1.7859525680541992, + "learning_rate": 2.6635542240426813e-06, + "loss": 3.0361, + "step": 19190 + }, + { + "epoch": 1.5385340041290012, + "grad_norm": 1.7461769580841064, + "learning_rate": 2.6626719477545703e-06, + "loss": 3.1569, + "step": 19191 + }, + { + "epoch": 1.5386141789100238, + "grad_norm": 1.8179874420166016, + "learning_rate": 2.6617897951712256e-06, + "loss": 3.3446, + "step": 19192 + }, + { + "epoch": 1.5386943536910465, + "grad_norm": 1.9393166303634644, + "learning_rate": 2.660907766307517e-06, + "loss": 3.1592, + "step": 19193 + }, + { + "epoch": 1.538774528472069, + "grad_norm": 1.7649039030075073, + "learning_rate": 2.6600258611783157e-06, + "loss": 3.2676, + "step": 19194 + }, + { + "epoch": 1.5388547032530917, + "grad_norm": 1.9469430446624756, + "learning_rate": 2.6591440797984925e-06, + "loss": 3.1202, + "step": 19195 + }, + { + "epoch": 1.5389348780341143, + "grad_norm": 1.8006806373596191, + "learning_rate": 2.6582624221829124e-06, + "loss": 3.1404, + "step": 19196 + }, + { + "epoch": 1.539015052815137, + "grad_norm": 1.8204399347305298, + "learning_rate": 2.6573808883464436e-06, + "loss": 3.1478, + "step": 19197 + }, + { + "epoch": 1.5390952275961596, + "grad_norm": 1.7977938652038574, + "learning_rate": 2.656499478303942e-06, + "loss": 3.1141, + "step": 19198 + }, + { + "epoch": 1.5391754023771824, + "grad_norm": 1.7630493640899658, + "learning_rate": 2.6556181920702715e-06, + "loss": 3.2323, + "step": 19199 + }, + { + "epoch": 1.539255577158205, + "grad_norm": 1.8392431735992432, + "learning_rate": 2.6547370296602894e-06, + "loss": 3.1494, + "step": 19200 + }, + { + "epoch": 1.5393357519392277, + "grad_norm": 0.6856323480606079, + "learning_rate": 2.6538559910888527e-06, + "loss": 2.4928, + "step": 19201 + }, + { + "epoch": 1.5394159267202503, + "grad_norm": 1.9019372463226318, + "learning_rate": 2.652975076370817e-06, + "loss": 3.1592, + "step": 19202 + }, + { + "epoch": 1.539496101501273, + "grad_norm": 0.6982600092887878, + "learning_rate": 2.6520942855210306e-06, + "loss": 2.4691, + "step": 19203 + }, + { + "epoch": 1.5395762762822955, + "grad_norm": 1.7876298427581787, + "learning_rate": 2.6512136185543434e-06, + "loss": 3.1581, + "step": 19204 + }, + { + "epoch": 1.5396564510633182, + "grad_norm": 1.7625044584274292, + "learning_rate": 2.650333075485606e-06, + "loss": 3.1456, + "step": 19205 + }, + { + "epoch": 1.5397366258443408, + "grad_norm": 1.72360360622406, + "learning_rate": 2.649452656329662e-06, + "loss": 3.2116, + "step": 19206 + }, + { + "epoch": 1.5398168006253634, + "grad_norm": 1.7855216264724731, + "learning_rate": 2.6485723611013593e-06, + "loss": 3.1201, + "step": 19207 + }, + { + "epoch": 1.539896975406386, + "grad_norm": 1.6992613077163696, + "learning_rate": 2.647692189815532e-06, + "loss": 3.1753, + "step": 19208 + }, + { + "epoch": 1.5399771501874087, + "grad_norm": 1.7325491905212402, + "learning_rate": 2.646812142487023e-06, + "loss": 3.1827, + "step": 19209 + }, + { + "epoch": 1.5400573249684313, + "grad_norm": 1.7800687551498413, + "learning_rate": 2.64593221913067e-06, + "loss": 3.1793, + "step": 19210 + }, + { + "epoch": 1.540137499749454, + "grad_norm": 1.7030479907989502, + "learning_rate": 2.6450524197613083e-06, + "loss": 3.1854, + "step": 19211 + }, + { + "epoch": 1.5402176745304765, + "grad_norm": 1.7880667448043823, + "learning_rate": 2.6441727443937736e-06, + "loss": 3.1482, + "step": 19212 + }, + { + "epoch": 1.5402978493114992, + "grad_norm": 1.810328722000122, + "learning_rate": 2.6432931930428907e-06, + "loss": 3.1922, + "step": 19213 + }, + { + "epoch": 1.5403780240925218, + "grad_norm": 1.6791136264801025, + "learning_rate": 2.6424137657234917e-06, + "loss": 3.1583, + "step": 19214 + }, + { + "epoch": 1.5404581988735444, + "grad_norm": 1.715438723564148, + "learning_rate": 2.6415344624504037e-06, + "loss": 3.2035, + "step": 19215 + }, + { + "epoch": 1.540538373654567, + "grad_norm": 1.8121473789215088, + "learning_rate": 2.6406552832384513e-06, + "loss": 3.1364, + "step": 19216 + }, + { + "epoch": 1.5406185484355897, + "grad_norm": 0.6826191544532776, + "learning_rate": 2.6397762281024595e-06, + "loss": 2.4821, + "step": 19217 + }, + { + "epoch": 1.5406987232166123, + "grad_norm": 1.6760351657867432, + "learning_rate": 2.6388972970572447e-06, + "loss": 3.2026, + "step": 19218 + }, + { + "epoch": 1.540778897997635, + "grad_norm": 1.7232518196105957, + "learning_rate": 2.6380184901176265e-06, + "loss": 3.1642, + "step": 19219 + }, + { + "epoch": 1.5408590727786575, + "grad_norm": 1.7917667627334595, + "learning_rate": 2.637139807298422e-06, + "loss": 3.1982, + "step": 19220 + }, + { + "epoch": 1.5409392475596801, + "grad_norm": 0.6762243509292603, + "learning_rate": 2.636261248614446e-06, + "loss": 2.4757, + "step": 19221 + }, + { + "epoch": 1.5410194223407028, + "grad_norm": 1.7656612396240234, + "learning_rate": 2.6353828140805125e-06, + "loss": 3.1157, + "step": 19222 + }, + { + "epoch": 1.5410995971217254, + "grad_norm": 1.7656699419021606, + "learning_rate": 2.634504503711426e-06, + "loss": 3.1137, + "step": 19223 + }, + { + "epoch": 1.541179771902748, + "grad_norm": 1.9848006963729858, + "learning_rate": 2.6336263175219978e-06, + "loss": 3.2363, + "step": 19224 + }, + { + "epoch": 1.5412599466837706, + "grad_norm": 0.702192485332489, + "learning_rate": 2.6327482555270345e-06, + "loss": 2.4822, + "step": 19225 + }, + { + "epoch": 1.5413401214647933, + "grad_norm": 1.7768229246139526, + "learning_rate": 2.6318703177413384e-06, + "loss": 3.1653, + "step": 19226 + }, + { + "epoch": 1.541420296245816, + "grad_norm": 1.8785651922225952, + "learning_rate": 2.6309925041797156e-06, + "loss": 3.1952, + "step": 19227 + }, + { + "epoch": 1.5415004710268385, + "grad_norm": 1.8353500366210938, + "learning_rate": 2.6301148148569587e-06, + "loss": 3.2115, + "step": 19228 + }, + { + "epoch": 1.5415806458078611, + "grad_norm": 1.7554214000701904, + "learning_rate": 2.629237249787868e-06, + "loss": 3.1776, + "step": 19229 + }, + { + "epoch": 1.5416608205888838, + "grad_norm": 1.7969204187393188, + "learning_rate": 2.628359808987241e-06, + "loss": 3.2159, + "step": 19230 + }, + { + "epoch": 1.5417409953699064, + "grad_norm": 1.7727113962173462, + "learning_rate": 2.6274824924698685e-06, + "loss": 3.0892, + "step": 19231 + }, + { + "epoch": 1.541821170150929, + "grad_norm": 1.712609887123108, + "learning_rate": 2.6266053002505453e-06, + "loss": 3.0992, + "step": 19232 + }, + { + "epoch": 1.5419013449319516, + "grad_norm": 1.9220794439315796, + "learning_rate": 2.625728232344056e-06, + "loss": 3.2823, + "step": 19233 + }, + { + "epoch": 1.5419815197129743, + "grad_norm": 1.7847657203674316, + "learning_rate": 2.62485128876519e-06, + "loss": 3.1971, + "step": 19234 + }, + { + "epoch": 1.5420616944939969, + "grad_norm": 1.800065279006958, + "learning_rate": 2.6239744695287318e-06, + "loss": 3.0808, + "step": 19235 + }, + { + "epoch": 1.5421418692750195, + "grad_norm": 1.8356356620788574, + "learning_rate": 2.6230977746494647e-06, + "loss": 3.1646, + "step": 19236 + }, + { + "epoch": 1.5422220440560421, + "grad_norm": 1.8548479080200195, + "learning_rate": 2.6222212041421714e-06, + "loss": 3.1636, + "step": 19237 + }, + { + "epoch": 1.5423022188370648, + "grad_norm": 1.7330535650253296, + "learning_rate": 2.6213447580216258e-06, + "loss": 3.1626, + "step": 19238 + }, + { + "epoch": 1.5423823936180874, + "grad_norm": 1.7557111978530884, + "learning_rate": 2.620468436302608e-06, + "loss": 3.2186, + "step": 19239 + }, + { + "epoch": 1.54246256839911, + "grad_norm": 1.9268525838851929, + "learning_rate": 2.6195922389998897e-06, + "loss": 3.2164, + "step": 19240 + }, + { + "epoch": 1.5425427431801326, + "grad_norm": 1.6617698669433594, + "learning_rate": 2.6187161661282467e-06, + "loss": 3.1415, + "step": 19241 + }, + { + "epoch": 1.5426229179611552, + "grad_norm": 1.7689327001571655, + "learning_rate": 2.6178402177024498e-06, + "loss": 3.244, + "step": 19242 + }, + { + "epoch": 1.5427030927421779, + "grad_norm": 1.8043078184127808, + "learning_rate": 2.6169643937372626e-06, + "loss": 3.2206, + "step": 19243 + }, + { + "epoch": 1.5427832675232005, + "grad_norm": 1.7141520977020264, + "learning_rate": 2.616088694247454e-06, + "loss": 3.1868, + "step": 19244 + }, + { + "epoch": 1.5428634423042231, + "grad_norm": 1.789280652999878, + "learning_rate": 2.615213119247787e-06, + "loss": 3.0663, + "step": 19245 + }, + { + "epoch": 1.5429436170852457, + "grad_norm": 1.7830318212509155, + "learning_rate": 2.614337668753024e-06, + "loss": 3.1785, + "step": 19246 + }, + { + "epoch": 1.5430237918662684, + "grad_norm": 1.7266058921813965, + "learning_rate": 2.6134623427779295e-06, + "loss": 3.2047, + "step": 19247 + }, + { + "epoch": 1.543103966647291, + "grad_norm": 1.725460171699524, + "learning_rate": 2.6125871413372517e-06, + "loss": 3.1489, + "step": 19248 + }, + { + "epoch": 1.5431841414283136, + "grad_norm": 1.7820535898208618, + "learning_rate": 2.611712064445753e-06, + "loss": 3.222, + "step": 19249 + }, + { + "epoch": 1.5432643162093362, + "grad_norm": 1.7545340061187744, + "learning_rate": 2.610837112118183e-06, + "loss": 3.0529, + "step": 19250 + }, + { + "epoch": 1.5433444909903589, + "grad_norm": 1.8241922855377197, + "learning_rate": 2.609962284369296e-06, + "loss": 3.1482, + "step": 19251 + }, + { + "epoch": 1.5434246657713815, + "grad_norm": 1.8161702156066895, + "learning_rate": 2.609087581213844e-06, + "loss": 3.1803, + "step": 19252 + }, + { + "epoch": 1.5435048405524041, + "grad_norm": 1.7094184160232544, + "learning_rate": 2.6082130026665665e-06, + "loss": 3.1984, + "step": 19253 + }, + { + "epoch": 1.5435850153334267, + "grad_norm": 1.779143214225769, + "learning_rate": 2.607338548742213e-06, + "loss": 3.1733, + "step": 19254 + }, + { + "epoch": 1.5436651901144494, + "grad_norm": 1.7424930334091187, + "learning_rate": 2.6064642194555265e-06, + "loss": 3.1117, + "step": 19255 + }, + { + "epoch": 1.543745364895472, + "grad_norm": 1.7839807271957397, + "learning_rate": 2.6055900148212476e-06, + "loss": 3.1626, + "step": 19256 + }, + { + "epoch": 1.5438255396764946, + "grad_norm": 1.7594271898269653, + "learning_rate": 2.6047159348541173e-06, + "loss": 3.1198, + "step": 19257 + }, + { + "epoch": 1.5439057144575172, + "grad_norm": 1.7394120693206787, + "learning_rate": 2.6038419795688676e-06, + "loss": 3.1309, + "step": 19258 + }, + { + "epoch": 1.5439858892385399, + "grad_norm": 1.8273717164993286, + "learning_rate": 2.6029681489802348e-06, + "loss": 3.1121, + "step": 19259 + }, + { + "epoch": 1.5440660640195627, + "grad_norm": 1.7732388973236084, + "learning_rate": 2.6020944431029525e-06, + "loss": 3.298, + "step": 19260 + }, + { + "epoch": 1.5441462388005853, + "grad_norm": 1.7430096864700317, + "learning_rate": 2.601220861951751e-06, + "loss": 3.0667, + "step": 19261 + }, + { + "epoch": 1.544226413581608, + "grad_norm": 1.8621140718460083, + "learning_rate": 2.6003474055413614e-06, + "loss": 3.2457, + "step": 19262 + }, + { + "epoch": 1.5443065883626306, + "grad_norm": 1.866499900817871, + "learning_rate": 2.599474073886503e-06, + "loss": 3.2079, + "step": 19263 + }, + { + "epoch": 1.5443867631436532, + "grad_norm": 1.764992117881775, + "learning_rate": 2.5986008670019036e-06, + "loss": 3.1521, + "step": 19264 + }, + { + "epoch": 1.5444669379246758, + "grad_norm": 1.9409196376800537, + "learning_rate": 2.5977277849022855e-06, + "loss": 3.1738, + "step": 19265 + }, + { + "epoch": 1.5445471127056984, + "grad_norm": 1.7987868785858154, + "learning_rate": 2.5968548276023677e-06, + "loss": 3.1525, + "step": 19266 + }, + { + "epoch": 1.544627287486721, + "grad_norm": 1.7750986814498901, + "learning_rate": 2.595981995116871e-06, + "loss": 3.2007, + "step": 19267 + }, + { + "epoch": 1.5447074622677437, + "grad_norm": 1.8571101427078247, + "learning_rate": 2.5951092874605054e-06, + "loss": 3.1992, + "step": 19268 + }, + { + "epoch": 1.5447876370487663, + "grad_norm": 1.6830685138702393, + "learning_rate": 2.5942367046479876e-06, + "loss": 3.1427, + "step": 19269 + }, + { + "epoch": 1.544867811829789, + "grad_norm": 1.6481088399887085, + "learning_rate": 2.59336424669403e-06, + "loss": 3.0725, + "step": 19270 + }, + { + "epoch": 1.5449479866108116, + "grad_norm": 1.7411407232284546, + "learning_rate": 2.59249191361334e-06, + "loss": 3.2367, + "step": 19271 + }, + { + "epoch": 1.5450281613918342, + "grad_norm": 1.853875994682312, + "learning_rate": 2.591619705420628e-06, + "loss": 3.1602, + "step": 19272 + }, + { + "epoch": 1.5451083361728568, + "grad_norm": 1.7999236583709717, + "learning_rate": 2.590747622130595e-06, + "loss": 3.1549, + "step": 19273 + }, + { + "epoch": 1.5451885109538794, + "grad_norm": 1.8206794261932373, + "learning_rate": 2.5898756637579457e-06, + "loss": 3.2552, + "step": 19274 + }, + { + "epoch": 1.545268685734902, + "grad_norm": 1.6999285221099854, + "learning_rate": 2.5890038303173805e-06, + "loss": 3.0514, + "step": 19275 + }, + { + "epoch": 1.5453488605159247, + "grad_norm": 1.7501684427261353, + "learning_rate": 2.5881321218235987e-06, + "loss": 3.1161, + "step": 19276 + }, + { + "epoch": 1.5454290352969473, + "grad_norm": 1.8226555585861206, + "learning_rate": 2.5872605382913005e-06, + "loss": 3.1268, + "step": 19277 + }, + { + "epoch": 1.54550921007797, + "grad_norm": 1.8913135528564453, + "learning_rate": 2.5863890797351744e-06, + "loss": 3.2032, + "step": 19278 + }, + { + "epoch": 1.5455893848589926, + "grad_norm": 1.7212239503860474, + "learning_rate": 2.5855177461699155e-06, + "loss": 3.1691, + "step": 19279 + }, + { + "epoch": 1.5456695596400152, + "grad_norm": 1.7470667362213135, + "learning_rate": 2.5846465376102147e-06, + "loss": 3.1716, + "step": 19280 + }, + { + "epoch": 1.545749734421038, + "grad_norm": 1.7657928466796875, + "learning_rate": 2.5837754540707593e-06, + "loss": 3.1819, + "step": 19281 + }, + { + "epoch": 1.5458299092020606, + "grad_norm": 0.6735210418701172, + "learning_rate": 2.5829044955662387e-06, + "loss": 2.4656, + "step": 19282 + }, + { + "epoch": 1.5459100839830833, + "grad_norm": 1.9403200149536133, + "learning_rate": 2.582033662111332e-06, + "loss": 3.2845, + "step": 19283 + }, + { + "epoch": 1.545990258764106, + "grad_norm": 1.720978021621704, + "learning_rate": 2.5811629537207227e-06, + "loss": 3.1886, + "step": 19284 + }, + { + "epoch": 1.5460704335451285, + "grad_norm": 1.831833004951477, + "learning_rate": 2.580292370409092e-06, + "loss": 3.2161, + "step": 19285 + }, + { + "epoch": 1.5461506083261511, + "grad_norm": 1.7494019269943237, + "learning_rate": 2.579421912191117e-06, + "loss": 3.1431, + "step": 19286 + }, + { + "epoch": 1.5462307831071738, + "grad_norm": 1.7200508117675781, + "learning_rate": 2.5785515790814752e-06, + "loss": 3.0856, + "step": 19287 + }, + { + "epoch": 1.5463109578881964, + "grad_norm": 1.7748113870620728, + "learning_rate": 2.5776813710948357e-06, + "loss": 3.1571, + "step": 19288 + }, + { + "epoch": 1.546391132669219, + "grad_norm": 1.8110278844833374, + "learning_rate": 2.576811288245873e-06, + "loss": 3.1456, + "step": 19289 + }, + { + "epoch": 1.5464713074502416, + "grad_norm": 1.7254786491394043, + "learning_rate": 2.5759413305492555e-06, + "loss": 3.088, + "step": 19290 + }, + { + "epoch": 1.5465514822312643, + "grad_norm": 1.7292288541793823, + "learning_rate": 2.57507149801965e-06, + "loss": 3.0367, + "step": 19291 + }, + { + "epoch": 1.5466316570122869, + "grad_norm": 1.8580198287963867, + "learning_rate": 2.5742017906717266e-06, + "loss": 3.1851, + "step": 19292 + }, + { + "epoch": 1.5467118317933095, + "grad_norm": 1.767726182937622, + "learning_rate": 2.5733322085201396e-06, + "loss": 3.0648, + "step": 19293 + }, + { + "epoch": 1.5467920065743321, + "grad_norm": 0.6660855412483215, + "learning_rate": 2.572462751579555e-06, + "loss": 2.4707, + "step": 19294 + }, + { + "epoch": 1.5468721813553548, + "grad_norm": 1.7499250173568726, + "learning_rate": 2.571593419864631e-06, + "loss": 3.2146, + "step": 19295 + }, + { + "epoch": 1.5469523561363774, + "grad_norm": 0.667351245880127, + "learning_rate": 2.570724213390022e-06, + "loss": 2.4741, + "step": 19296 + }, + { + "epoch": 1.5470325309174, + "grad_norm": 1.7315654754638672, + "learning_rate": 2.569855132170389e-06, + "loss": 3.1626, + "step": 19297 + }, + { + "epoch": 1.5471127056984226, + "grad_norm": 1.8178799152374268, + "learning_rate": 2.5689861762203774e-06, + "loss": 3.179, + "step": 19298 + }, + { + "epoch": 1.5471928804794453, + "grad_norm": 1.7429144382476807, + "learning_rate": 2.568117345554638e-06, + "loss": 3.1493, + "step": 19299 + }, + { + "epoch": 1.5472730552604679, + "grad_norm": 1.7972270250320435, + "learning_rate": 2.5672486401878228e-06, + "loss": 3.1855, + "step": 19300 + }, + { + "epoch": 1.5473532300414905, + "grad_norm": 1.820507526397705, + "learning_rate": 2.5663800601345747e-06, + "loss": 3.2102, + "step": 19301 + }, + { + "epoch": 1.5474334048225131, + "grad_norm": 1.8599940538406372, + "learning_rate": 2.56551160540954e-06, + "loss": 3.2, + "step": 19302 + }, + { + "epoch": 1.5475135796035357, + "grad_norm": 1.8140389919281006, + "learning_rate": 2.564643276027362e-06, + "loss": 3.2128, + "step": 19303 + }, + { + "epoch": 1.5475937543845584, + "grad_norm": 1.8944381475448608, + "learning_rate": 2.563775072002674e-06, + "loss": 3.1233, + "step": 19304 + }, + { + "epoch": 1.547673929165581, + "grad_norm": 1.8514037132263184, + "learning_rate": 2.5629069933501184e-06, + "loss": 3.2654, + "step": 19305 + }, + { + "epoch": 1.5477541039466036, + "grad_norm": 1.8501888513565063, + "learning_rate": 2.5620390400843297e-06, + "loss": 3.1731, + "step": 19306 + }, + { + "epoch": 1.5478342787276262, + "grad_norm": 1.784020185470581, + "learning_rate": 2.5611712122199417e-06, + "loss": 3.1921, + "step": 19307 + }, + { + "epoch": 1.5479144535086489, + "grad_norm": 0.700358510017395, + "learning_rate": 2.560303509771588e-06, + "loss": 2.4634, + "step": 19308 + }, + { + "epoch": 1.5479946282896715, + "grad_norm": 1.7739818096160889, + "learning_rate": 2.559435932753892e-06, + "loss": 3.2346, + "step": 19309 + }, + { + "epoch": 1.5480748030706941, + "grad_norm": 1.76752769947052, + "learning_rate": 2.558568481181484e-06, + "loss": 3.1931, + "step": 19310 + }, + { + "epoch": 1.5481549778517167, + "grad_norm": 1.6769938468933105, + "learning_rate": 2.557701155068989e-06, + "loss": 3.0955, + "step": 19311 + }, + { + "epoch": 1.5482351526327394, + "grad_norm": 1.7339577674865723, + "learning_rate": 2.5568339544310296e-06, + "loss": 3.1678, + "step": 19312 + }, + { + "epoch": 1.548315327413762, + "grad_norm": 1.756468415260315, + "learning_rate": 2.555966879282229e-06, + "loss": 3.1495, + "step": 19313 + }, + { + "epoch": 1.5483955021947846, + "grad_norm": 1.8460031747817993, + "learning_rate": 2.5550999296372004e-06, + "loss": 3.1183, + "step": 19314 + }, + { + "epoch": 1.5484756769758072, + "grad_norm": 1.7799606323242188, + "learning_rate": 2.554233105510564e-06, + "loss": 3.1843, + "step": 19315 + }, + { + "epoch": 1.5485558517568299, + "grad_norm": 1.69962739944458, + "learning_rate": 2.553366406916933e-06, + "loss": 3.114, + "step": 19316 + }, + { + "epoch": 1.5486360265378525, + "grad_norm": 1.6784073114395142, + "learning_rate": 2.5524998338709183e-06, + "loss": 3.0416, + "step": 19317 + }, + { + "epoch": 1.548716201318875, + "grad_norm": 1.819944143295288, + "learning_rate": 2.551633386387137e-06, + "loss": 3.2369, + "step": 19318 + }, + { + "epoch": 1.5487963760998977, + "grad_norm": 1.6131627559661865, + "learning_rate": 2.5507670644801874e-06, + "loss": 3.1694, + "step": 19319 + }, + { + "epoch": 1.5488765508809204, + "grad_norm": 1.7927793264389038, + "learning_rate": 2.54990086816468e-06, + "loss": 3.1716, + "step": 19320 + }, + { + "epoch": 1.548956725661943, + "grad_norm": 1.8324238061904907, + "learning_rate": 2.5490347974552175e-06, + "loss": 3.0549, + "step": 19321 + }, + { + "epoch": 1.5490369004429656, + "grad_norm": 1.7438222169876099, + "learning_rate": 2.5481688523664037e-06, + "loss": 3.1675, + "step": 19322 + }, + { + "epoch": 1.5491170752239882, + "grad_norm": 1.829997181892395, + "learning_rate": 2.547303032912839e-06, + "loss": 3.172, + "step": 19323 + }, + { + "epoch": 1.5491972500050109, + "grad_norm": 0.6888315677642822, + "learning_rate": 2.5464373391091155e-06, + "loss": 2.4762, + "step": 19324 + }, + { + "epoch": 1.5492774247860335, + "grad_norm": 1.802694320678711, + "learning_rate": 2.5455717709698325e-06, + "loss": 3.1906, + "step": 19325 + }, + { + "epoch": 1.549357599567056, + "grad_norm": 1.7875193357467651, + "learning_rate": 2.544706328509582e-06, + "loss": 3.1031, + "step": 19326 + }, + { + "epoch": 1.5494377743480787, + "grad_norm": 1.772255301475525, + "learning_rate": 2.5438410117429545e-06, + "loss": 3.2046, + "step": 19327 + }, + { + "epoch": 1.5495179491291013, + "grad_norm": 0.6730871200561523, + "learning_rate": 2.5429758206845432e-06, + "loss": 2.4731, + "step": 19328 + }, + { + "epoch": 1.549598123910124, + "grad_norm": 1.8140369653701782, + "learning_rate": 2.5421107553489275e-06, + "loss": 3.1353, + "step": 19329 + }, + { + "epoch": 1.5496782986911466, + "grad_norm": 1.870579719543457, + "learning_rate": 2.541245815750698e-06, + "loss": 3.2435, + "step": 19330 + }, + { + "epoch": 1.5497584734721692, + "grad_norm": 1.8649927377700806, + "learning_rate": 2.540381001904434e-06, + "loss": 3.0865, + "step": 19331 + }, + { + "epoch": 1.5498386482531918, + "grad_norm": 1.7644027471542358, + "learning_rate": 2.539516313824717e-06, + "loss": 3.1173, + "step": 19332 + }, + { + "epoch": 1.5499188230342145, + "grad_norm": 0.6687283515930176, + "learning_rate": 2.5386517515261298e-06, + "loss": 2.4476, + "step": 19333 + }, + { + "epoch": 1.549998997815237, + "grad_norm": 1.773229956626892, + "learning_rate": 2.53778731502324e-06, + "loss": 3.0342, + "step": 19334 + }, + { + "epoch": 1.5500791725962597, + "grad_norm": 1.8251125812530518, + "learning_rate": 2.536923004330628e-06, + "loss": 3.1236, + "step": 19335 + }, + { + "epoch": 1.5501593473772823, + "grad_norm": 1.876987099647522, + "learning_rate": 2.536058819462862e-06, + "loss": 3.2141, + "step": 19336 + }, + { + "epoch": 1.550239522158305, + "grad_norm": 1.9426116943359375, + "learning_rate": 2.535194760434515e-06, + "loss": 3.1805, + "step": 19337 + }, + { + "epoch": 1.5503196969393276, + "grad_norm": 1.7948472499847412, + "learning_rate": 2.534330827260156e-06, + "loss": 3.104, + "step": 19338 + }, + { + "epoch": 1.5503998717203502, + "grad_norm": 1.732050895690918, + "learning_rate": 2.5334670199543455e-06, + "loss": 3.2045, + "step": 19339 + }, + { + "epoch": 1.5504800465013728, + "grad_norm": 1.724963903427124, + "learning_rate": 2.5326033385316497e-06, + "loss": 3.1417, + "step": 19340 + }, + { + "epoch": 1.5505602212823955, + "grad_norm": 1.9292001724243164, + "learning_rate": 2.531739783006629e-06, + "loss": 3.1957, + "step": 19341 + }, + { + "epoch": 1.5506403960634183, + "grad_norm": 1.7331287860870361, + "learning_rate": 2.530876353393844e-06, + "loss": 3.0748, + "step": 19342 + }, + { + "epoch": 1.550720570844441, + "grad_norm": 1.7221239805221558, + "learning_rate": 2.5300130497078533e-06, + "loss": 3.1252, + "step": 19343 + }, + { + "epoch": 1.5508007456254636, + "grad_norm": 1.8531781435012817, + "learning_rate": 2.529149871963208e-06, + "loss": 3.1031, + "step": 19344 + }, + { + "epoch": 1.5508809204064862, + "grad_norm": 1.6852443218231201, + "learning_rate": 2.5282868201744626e-06, + "loss": 3.1276, + "step": 19345 + }, + { + "epoch": 1.5509610951875088, + "grad_norm": 1.7583446502685547, + "learning_rate": 2.527423894356169e-06, + "loss": 3.1663, + "step": 19346 + }, + { + "epoch": 1.5510412699685314, + "grad_norm": 1.783617377281189, + "learning_rate": 2.5265610945228736e-06, + "loss": 3.1935, + "step": 19347 + }, + { + "epoch": 1.551121444749554, + "grad_norm": 1.677242636680603, + "learning_rate": 2.525698420689128e-06, + "loss": 3.1301, + "step": 19348 + }, + { + "epoch": 1.5512016195305767, + "grad_norm": 1.7701663970947266, + "learning_rate": 2.52483587286947e-06, + "loss": 3.2036, + "step": 19349 + }, + { + "epoch": 1.5512817943115993, + "grad_norm": 1.727623701095581, + "learning_rate": 2.523973451078444e-06, + "loss": 3.1135, + "step": 19350 + }, + { + "epoch": 1.551361969092622, + "grad_norm": 1.7231779098510742, + "learning_rate": 2.523111155330591e-06, + "loss": 3.0848, + "step": 19351 + }, + { + "epoch": 1.5514421438736445, + "grad_norm": 1.7792154550552368, + "learning_rate": 2.5222489856404487e-06, + "loss": 3.2634, + "step": 19352 + }, + { + "epoch": 1.5515223186546672, + "grad_norm": 1.7546056509017944, + "learning_rate": 2.521386942022557e-06, + "loss": 3.1812, + "step": 19353 + }, + { + "epoch": 1.5516024934356898, + "grad_norm": 1.8126388788223267, + "learning_rate": 2.5205250244914413e-06, + "loss": 3.2239, + "step": 19354 + }, + { + "epoch": 1.5516826682167124, + "grad_norm": 1.8695604801177979, + "learning_rate": 2.5196632330616387e-06, + "loss": 3.1715, + "step": 19355 + }, + { + "epoch": 1.551762842997735, + "grad_norm": 1.8263959884643555, + "learning_rate": 2.5188015677476775e-06, + "loss": 3.2244, + "step": 19356 + }, + { + "epoch": 1.5518430177787577, + "grad_norm": 1.714756965637207, + "learning_rate": 2.5179400285640854e-06, + "loss": 3.1487, + "step": 19357 + }, + { + "epoch": 1.5519231925597803, + "grad_norm": 1.7518614530563354, + "learning_rate": 2.51707861552539e-06, + "loss": 3.1886, + "step": 19358 + }, + { + "epoch": 1.552003367340803, + "grad_norm": 1.7484771013259888, + "learning_rate": 2.516217328646109e-06, + "loss": 3.1497, + "step": 19359 + }, + { + "epoch": 1.5520835421218255, + "grad_norm": 1.7781635522842407, + "learning_rate": 2.515356167940767e-06, + "loss": 3.173, + "step": 19360 + }, + { + "epoch": 1.5521637169028482, + "grad_norm": 1.756930947303772, + "learning_rate": 2.514495133423881e-06, + "loss": 3.1398, + "step": 19361 + }, + { + "epoch": 1.552243891683871, + "grad_norm": 1.8054685592651367, + "learning_rate": 2.5136342251099698e-06, + "loss": 3.163, + "step": 19362 + }, + { + "epoch": 1.5523240664648936, + "grad_norm": 1.8520689010620117, + "learning_rate": 2.51277344301355e-06, + "loss": 3.1166, + "step": 19363 + }, + { + "epoch": 1.5524042412459162, + "grad_norm": 1.7608157396316528, + "learning_rate": 2.5119127871491277e-06, + "loss": 3.1181, + "step": 19364 + }, + { + "epoch": 1.5524844160269389, + "grad_norm": 1.6749192476272583, + "learning_rate": 2.511052257531217e-06, + "loss": 3.2772, + "step": 19365 + }, + { + "epoch": 1.5525645908079615, + "grad_norm": 1.837923526763916, + "learning_rate": 2.5101918541743253e-06, + "loss": 3.097, + "step": 19366 + }, + { + "epoch": 1.5526447655889841, + "grad_norm": 1.8107919692993164, + "learning_rate": 2.5093315770929605e-06, + "loss": 3.207, + "step": 19367 + }, + { + "epoch": 1.5527249403700067, + "grad_norm": 1.7004835605621338, + "learning_rate": 2.5084714263016273e-06, + "loss": 3.1748, + "step": 19368 + }, + { + "epoch": 1.5528051151510294, + "grad_norm": 1.7705456018447876, + "learning_rate": 2.5076114018148233e-06, + "loss": 3.1084, + "step": 19369 + }, + { + "epoch": 1.552885289932052, + "grad_norm": 1.8087595701217651, + "learning_rate": 2.5067515036470504e-06, + "loss": 3.1176, + "step": 19370 + }, + { + "epoch": 1.5529654647130746, + "grad_norm": 1.854728102684021, + "learning_rate": 2.5058917318128062e-06, + "loss": 3.2308, + "step": 19371 + }, + { + "epoch": 1.5530456394940972, + "grad_norm": 1.6204477548599243, + "learning_rate": 2.5050320863265853e-06, + "loss": 3.1384, + "step": 19372 + }, + { + "epoch": 1.5531258142751199, + "grad_norm": 1.6954470872879028, + "learning_rate": 2.504172567202886e-06, + "loss": 3.1155, + "step": 19373 + }, + { + "epoch": 1.5532059890561425, + "grad_norm": 1.6809606552124023, + "learning_rate": 2.5033131744561924e-06, + "loss": 3.194, + "step": 19374 + }, + { + "epoch": 1.5532861638371651, + "grad_norm": 1.7692468166351318, + "learning_rate": 2.5024539081009967e-06, + "loss": 3.1321, + "step": 19375 + }, + { + "epoch": 1.5533663386181877, + "grad_norm": 2.082632303237915, + "learning_rate": 2.501594768151785e-06, + "loss": 3.2682, + "step": 19376 + }, + { + "epoch": 1.5534465133992104, + "grad_norm": 1.8714033365249634, + "learning_rate": 2.500735754623044e-06, + "loss": 3.1863, + "step": 19377 + }, + { + "epoch": 1.553526688180233, + "grad_norm": 1.7932519912719727, + "learning_rate": 2.4998768675292563e-06, + "loss": 3.1932, + "step": 19378 + }, + { + "epoch": 1.5536068629612556, + "grad_norm": 1.7634347677230835, + "learning_rate": 2.4990181068848995e-06, + "loss": 3.1081, + "step": 19379 + }, + { + "epoch": 1.5536870377422782, + "grad_norm": 1.795052409172058, + "learning_rate": 2.498159472704453e-06, + "loss": 3.1738, + "step": 19380 + }, + { + "epoch": 1.5537672125233009, + "grad_norm": 1.7600525617599487, + "learning_rate": 2.4973009650023948e-06, + "loss": 3.163, + "step": 19381 + }, + { + "epoch": 1.5538473873043235, + "grad_norm": 1.8007997274398804, + "learning_rate": 2.4964425837931984e-06, + "loss": 3.1628, + "step": 19382 + }, + { + "epoch": 1.553927562085346, + "grad_norm": 1.8252276182174683, + "learning_rate": 2.495584329091338e-06, + "loss": 3.2318, + "step": 19383 + }, + { + "epoch": 1.5540077368663687, + "grad_norm": 1.8145185708999634, + "learning_rate": 2.4947262009112785e-06, + "loss": 3.2575, + "step": 19384 + }, + { + "epoch": 1.5540879116473914, + "grad_norm": 1.795201063156128, + "learning_rate": 2.4938681992674895e-06, + "loss": 3.1527, + "step": 19385 + }, + { + "epoch": 1.554168086428414, + "grad_norm": 1.801194667816162, + "learning_rate": 2.4930103241744374e-06, + "loss": 3.174, + "step": 19386 + }, + { + "epoch": 1.5542482612094366, + "grad_norm": 1.8222520351409912, + "learning_rate": 2.492152575646586e-06, + "loss": 3.2674, + "step": 19387 + }, + { + "epoch": 1.5543284359904592, + "grad_norm": 0.6666812896728516, + "learning_rate": 2.4912949536983977e-06, + "loss": 2.4564, + "step": 19388 + }, + { + "epoch": 1.5544086107714818, + "grad_norm": 1.7582144737243652, + "learning_rate": 2.490437458344329e-06, + "loss": 3.2149, + "step": 19389 + }, + { + "epoch": 1.5544887855525045, + "grad_norm": 1.8786596059799194, + "learning_rate": 2.4895800895988366e-06, + "loss": 3.1267, + "step": 19390 + }, + { + "epoch": 1.554568960333527, + "grad_norm": 1.794861912727356, + "learning_rate": 2.488722847476378e-06, + "loss": 3.2362, + "step": 19391 + }, + { + "epoch": 1.5546491351145497, + "grad_norm": 1.7187579870224, + "learning_rate": 2.4878657319914044e-06, + "loss": 3.153, + "step": 19392 + }, + { + "epoch": 1.5547293098955723, + "grad_norm": 1.7810726165771484, + "learning_rate": 2.48700874315837e-06, + "loss": 3.1671, + "step": 19393 + }, + { + "epoch": 1.554809484676595, + "grad_norm": 1.795670509338379, + "learning_rate": 2.486151880991717e-06, + "loss": 3.0983, + "step": 19394 + }, + { + "epoch": 1.5548896594576176, + "grad_norm": 0.6933769583702087, + "learning_rate": 2.485295145505897e-06, + "loss": 2.49, + "step": 19395 + }, + { + "epoch": 1.5549698342386402, + "grad_norm": 1.9193801879882812, + "learning_rate": 2.484438536715351e-06, + "loss": 3.1864, + "step": 19396 + }, + { + "epoch": 1.5550500090196628, + "grad_norm": 1.75734543800354, + "learning_rate": 2.4835820546345223e-06, + "loss": 3.2014, + "step": 19397 + }, + { + "epoch": 1.5551301838006855, + "grad_norm": 1.8047912120819092, + "learning_rate": 2.482725699277855e-06, + "loss": 3.1671, + "step": 19398 + }, + { + "epoch": 1.555210358581708, + "grad_norm": 1.738386631011963, + "learning_rate": 2.481869470659779e-06, + "loss": 3.0985, + "step": 19399 + }, + { + "epoch": 1.5552905333627307, + "grad_norm": 1.8355822563171387, + "learning_rate": 2.481013368794735e-06, + "loss": 3.1539, + "step": 19400 + }, + { + "epoch": 1.5553707081437533, + "grad_norm": 1.8338921070098877, + "learning_rate": 2.4801573936971546e-06, + "loss": 3.1537, + "step": 19401 + }, + { + "epoch": 1.555450882924776, + "grad_norm": 1.805915117263794, + "learning_rate": 2.4793015453814716e-06, + "loss": 3.2057, + "step": 19402 + }, + { + "epoch": 1.5555310577057986, + "grad_norm": 1.8468822240829468, + "learning_rate": 2.478445823862116e-06, + "loss": 3.1624, + "step": 19403 + }, + { + "epoch": 1.5556112324868212, + "grad_norm": 1.7119684219360352, + "learning_rate": 2.47759022915351e-06, + "loss": 3.1547, + "step": 19404 + }, + { + "epoch": 1.5556914072678438, + "grad_norm": 1.8853436708450317, + "learning_rate": 2.476734761270082e-06, + "loss": 3.1805, + "step": 19405 + }, + { + "epoch": 1.5557715820488665, + "grad_norm": 1.710949182510376, + "learning_rate": 2.475879420226255e-06, + "loss": 3.1421, + "step": 19406 + }, + { + "epoch": 1.555851756829889, + "grad_norm": 1.863344430923462, + "learning_rate": 2.475024206036448e-06, + "loss": 3.1851, + "step": 19407 + }, + { + "epoch": 1.5559319316109117, + "grad_norm": 1.693835735321045, + "learning_rate": 2.474169118715085e-06, + "loss": 3.2186, + "step": 19408 + }, + { + "epoch": 1.5560121063919343, + "grad_norm": 1.6952961683273315, + "learning_rate": 2.473314158276574e-06, + "loss": 3.1849, + "step": 19409 + }, + { + "epoch": 1.556092281172957, + "grad_norm": 1.7485182285308838, + "learning_rate": 2.4724593247353346e-06, + "loss": 3.2766, + "step": 19410 + }, + { + "epoch": 1.5561724559539796, + "grad_norm": 1.8480491638183594, + "learning_rate": 2.4716046181057773e-06, + "loss": 3.0642, + "step": 19411 + }, + { + "epoch": 1.5562526307350022, + "grad_norm": 1.841082215309143, + "learning_rate": 2.4707500384023143e-06, + "loss": 3.1684, + "step": 19412 + }, + { + "epoch": 1.5563328055160248, + "grad_norm": 1.8136100769042969, + "learning_rate": 2.4698955856393534e-06, + "loss": 3.2029, + "step": 19413 + }, + { + "epoch": 1.5564129802970474, + "grad_norm": 0.6683050394058228, + "learning_rate": 2.4690412598312975e-06, + "loss": 2.4707, + "step": 19414 + }, + { + "epoch": 1.55649315507807, + "grad_norm": 0.6511310338973999, + "learning_rate": 2.4681870609925507e-06, + "loss": 2.4833, + "step": 19415 + }, + { + "epoch": 1.5565733298590927, + "grad_norm": 1.9298328161239624, + "learning_rate": 2.4673329891375164e-06, + "loss": 3.131, + "step": 19416 + }, + { + "epoch": 1.5566535046401153, + "grad_norm": 1.751255750656128, + "learning_rate": 2.4664790442805918e-06, + "loss": 3.1091, + "step": 19417 + }, + { + "epoch": 1.556733679421138, + "grad_norm": 1.7704638242721558, + "learning_rate": 2.46562522643618e-06, + "loss": 3.1604, + "step": 19418 + }, + { + "epoch": 1.5568138542021606, + "grad_norm": 1.8558250665664673, + "learning_rate": 2.4647715356186676e-06, + "loss": 3.1859, + "step": 19419 + }, + { + "epoch": 1.5568940289831832, + "grad_norm": 0.6565327644348145, + "learning_rate": 2.463917971842452e-06, + "loss": 2.4701, + "step": 19420 + }, + { + "epoch": 1.5569742037642058, + "grad_norm": 1.6879485845565796, + "learning_rate": 2.4630645351219227e-06, + "loss": 3.1093, + "step": 19421 + }, + { + "epoch": 1.5570543785452284, + "grad_norm": 0.6494991183280945, + "learning_rate": 2.462211225471469e-06, + "loss": 2.481, + "step": 19422 + }, + { + "epoch": 1.5571345533262513, + "grad_norm": 1.865850567817688, + "learning_rate": 2.461358042905482e-06, + "loss": 3.1851, + "step": 19423 + }, + { + "epoch": 1.557214728107274, + "grad_norm": 1.8364909887313843, + "learning_rate": 2.460504987438337e-06, + "loss": 3.1053, + "step": 19424 + }, + { + "epoch": 1.5572949028882965, + "grad_norm": 1.845938801765442, + "learning_rate": 2.459652059084421e-06, + "loss": 3.1277, + "step": 19425 + }, + { + "epoch": 1.5573750776693192, + "grad_norm": 1.726208209991455, + "learning_rate": 2.4587992578581145e-06, + "loss": 3.1603, + "step": 19426 + }, + { + "epoch": 1.5574552524503418, + "grad_norm": 1.6984219551086426, + "learning_rate": 2.4579465837737937e-06, + "loss": 3.0903, + "step": 19427 + }, + { + "epoch": 1.5575354272313644, + "grad_norm": 1.760035753250122, + "learning_rate": 2.4570940368458396e-06, + "loss": 3.1982, + "step": 19428 + }, + { + "epoch": 1.557615602012387, + "grad_norm": 1.712396264076233, + "learning_rate": 2.4562416170886174e-06, + "loss": 3.269, + "step": 19429 + }, + { + "epoch": 1.5576957767934096, + "grad_norm": 1.750298261642456, + "learning_rate": 2.455389324516504e-06, + "loss": 3.1979, + "step": 19430 + }, + { + "epoch": 1.5577759515744323, + "grad_norm": 1.8097236156463623, + "learning_rate": 2.454537159143867e-06, + "loss": 3.2882, + "step": 19431 + }, + { + "epoch": 1.557856126355455, + "grad_norm": 1.7945966720581055, + "learning_rate": 2.4536851209850745e-06, + "loss": 3.0942, + "step": 19432 + }, + { + "epoch": 1.5579363011364775, + "grad_norm": 1.723586916923523, + "learning_rate": 2.452833210054494e-06, + "loss": 3.0934, + "step": 19433 + }, + { + "epoch": 1.5580164759175001, + "grad_norm": 1.7066470384597778, + "learning_rate": 2.451981426366482e-06, + "loss": 3.1525, + "step": 19434 + }, + { + "epoch": 1.5580966506985228, + "grad_norm": 0.6746688485145569, + "learning_rate": 2.4511297699354043e-06, + "loss": 2.4725, + "step": 19435 + }, + { + "epoch": 1.5581768254795454, + "grad_norm": 0.6815391778945923, + "learning_rate": 2.4502782407756174e-06, + "loss": 2.4596, + "step": 19436 + }, + { + "epoch": 1.558257000260568, + "grad_norm": 1.6635524034500122, + "learning_rate": 2.4494268389014785e-06, + "loss": 3.1107, + "step": 19437 + }, + { + "epoch": 1.5583371750415906, + "grad_norm": 1.8558392524719238, + "learning_rate": 2.4485755643273457e-06, + "loss": 3.161, + "step": 19438 + }, + { + "epoch": 1.5584173498226133, + "grad_norm": 1.7915977239608765, + "learning_rate": 2.447724417067564e-06, + "loss": 3.1442, + "step": 19439 + }, + { + "epoch": 1.5584975246036359, + "grad_norm": 1.8405174016952515, + "learning_rate": 2.4468733971364877e-06, + "loss": 3.1696, + "step": 19440 + }, + { + "epoch": 1.5585776993846585, + "grad_norm": 1.7799123525619507, + "learning_rate": 2.446022504548463e-06, + "loss": 3.298, + "step": 19441 + }, + { + "epoch": 1.5586578741656811, + "grad_norm": 1.8300931453704834, + "learning_rate": 2.4451717393178364e-06, + "loss": 3.2433, + "step": 19442 + }, + { + "epoch": 1.558738048946704, + "grad_norm": 1.7000713348388672, + "learning_rate": 2.4443211014589553e-06, + "loss": 3.0422, + "step": 19443 + }, + { + "epoch": 1.5588182237277266, + "grad_norm": 1.6762137413024902, + "learning_rate": 2.4434705909861545e-06, + "loss": 3.0762, + "step": 19444 + }, + { + "epoch": 1.5588983985087492, + "grad_norm": 1.7624341249465942, + "learning_rate": 2.4426202079137773e-06, + "loss": 3.2113, + "step": 19445 + }, + { + "epoch": 1.5589785732897719, + "grad_norm": 0.6633417010307312, + "learning_rate": 2.4417699522561588e-06, + "loss": 2.4646, + "step": 19446 + }, + { + "epoch": 1.5590587480707945, + "grad_norm": 1.7826483249664307, + "learning_rate": 2.4409198240276346e-06, + "loss": 3.1973, + "step": 19447 + }, + { + "epoch": 1.559138922851817, + "grad_norm": 1.7826608419418335, + "learning_rate": 2.4400698232425424e-06, + "loss": 3.2043, + "step": 19448 + }, + { + "epoch": 1.5592190976328397, + "grad_norm": 1.7169828414916992, + "learning_rate": 2.439219949915205e-06, + "loss": 3.1497, + "step": 19449 + }, + { + "epoch": 1.5592992724138623, + "grad_norm": 1.7409594058990479, + "learning_rate": 2.4383702040599544e-06, + "loss": 3.1614, + "step": 19450 + }, + { + "epoch": 1.559379447194885, + "grad_norm": 1.7611796855926514, + "learning_rate": 2.4375205856911166e-06, + "loss": 3.243, + "step": 19451 + }, + { + "epoch": 1.5594596219759076, + "grad_norm": 1.9076411724090576, + "learning_rate": 2.4366710948230177e-06, + "loss": 3.236, + "step": 19452 + }, + { + "epoch": 1.5595397967569302, + "grad_norm": 1.7626097202301025, + "learning_rate": 2.4358217314699807e-06, + "loss": 3.1782, + "step": 19453 + }, + { + "epoch": 1.5596199715379528, + "grad_norm": 1.8145508766174316, + "learning_rate": 2.4349724956463207e-06, + "loss": 3.2019, + "step": 19454 + }, + { + "epoch": 1.5597001463189755, + "grad_norm": 1.6877201795578003, + "learning_rate": 2.434123387366358e-06, + "loss": 3.1443, + "step": 19455 + }, + { + "epoch": 1.559780321099998, + "grad_norm": 1.8228099346160889, + "learning_rate": 2.4332744066444093e-06, + "loss": 3.174, + "step": 19456 + }, + { + "epoch": 1.5598604958810207, + "grad_norm": 1.7007176876068115, + "learning_rate": 2.432425553494787e-06, + "loss": 3.1467, + "step": 19457 + }, + { + "epoch": 1.5599406706620433, + "grad_norm": 1.6606895923614502, + "learning_rate": 2.4315768279318053e-06, + "loss": 3.1448, + "step": 19458 + }, + { + "epoch": 1.560020845443066, + "grad_norm": 1.727469563484192, + "learning_rate": 2.430728229969769e-06, + "loss": 3.1217, + "step": 19459 + }, + { + "epoch": 1.5601010202240886, + "grad_norm": 1.7243950366973877, + "learning_rate": 2.429879759622986e-06, + "loss": 3.0919, + "step": 19460 + }, + { + "epoch": 1.5601811950051112, + "grad_norm": 1.7939879894256592, + "learning_rate": 2.429031416905763e-06, + "loss": 3.2559, + "step": 19461 + }, + { + "epoch": 1.5602613697861338, + "grad_norm": 1.819297432899475, + "learning_rate": 2.4281832018324016e-06, + "loss": 3.1822, + "step": 19462 + }, + { + "epoch": 1.5603415445671565, + "grad_norm": 1.7121251821517944, + "learning_rate": 2.4273351144172054e-06, + "loss": 3.1566, + "step": 19463 + }, + { + "epoch": 1.560421719348179, + "grad_norm": 1.7721006870269775, + "learning_rate": 2.4264871546744684e-06, + "loss": 3.2003, + "step": 19464 + }, + { + "epoch": 1.5605018941292017, + "grad_norm": 1.7159724235534668, + "learning_rate": 2.425639322618488e-06, + "loss": 3.2213, + "step": 19465 + }, + { + "epoch": 1.5605820689102243, + "grad_norm": 1.7003519535064697, + "learning_rate": 2.424791618263559e-06, + "loss": 3.1279, + "step": 19466 + }, + { + "epoch": 1.560662243691247, + "grad_norm": 1.689086675643921, + "learning_rate": 2.4239440416239746e-06, + "loss": 3.1947, + "step": 19467 + }, + { + "epoch": 1.5607424184722696, + "grad_norm": 1.7773807048797607, + "learning_rate": 2.423096592714026e-06, + "loss": 3.2949, + "step": 19468 + }, + { + "epoch": 1.5608225932532922, + "grad_norm": 1.9428400993347168, + "learning_rate": 2.422249271547995e-06, + "loss": 3.2251, + "step": 19469 + }, + { + "epoch": 1.5609027680343148, + "grad_norm": 1.7297905683517456, + "learning_rate": 2.4214020781401716e-06, + "loss": 3.0687, + "step": 19470 + }, + { + "epoch": 1.5609829428153374, + "grad_norm": 1.8726613521575928, + "learning_rate": 2.420555012504837e-06, + "loss": 3.2907, + "step": 19471 + }, + { + "epoch": 1.56106311759636, + "grad_norm": 1.7708446979522705, + "learning_rate": 2.419708074656274e-06, + "loss": 3.1465, + "step": 19472 + }, + { + "epoch": 1.5611432923773827, + "grad_norm": 1.7621417045593262, + "learning_rate": 2.418861264608765e-06, + "loss": 3.2258, + "step": 19473 + }, + { + "epoch": 1.5612234671584053, + "grad_norm": 1.724959135055542, + "learning_rate": 2.41801458237658e-06, + "loss": 3.1781, + "step": 19474 + }, + { + "epoch": 1.561303641939428, + "grad_norm": 1.709409475326538, + "learning_rate": 2.417168027973996e-06, + "loss": 3.1733, + "step": 19475 + }, + { + "epoch": 1.5613838167204506, + "grad_norm": 1.7996604442596436, + "learning_rate": 2.416321601415288e-06, + "loss": 3.2145, + "step": 19476 + }, + { + "epoch": 1.5614639915014732, + "grad_norm": 1.81399405002594, + "learning_rate": 2.4154753027147247e-06, + "loss": 3.1667, + "step": 19477 + }, + { + "epoch": 1.5615441662824958, + "grad_norm": 1.8855416774749756, + "learning_rate": 2.4146291318865776e-06, + "loss": 3.1893, + "step": 19478 + }, + { + "epoch": 1.5616243410635184, + "grad_norm": 1.7423774003982544, + "learning_rate": 2.4137830889451064e-06, + "loss": 3.0616, + "step": 19479 + }, + { + "epoch": 1.561704515844541, + "grad_norm": 1.6865737438201904, + "learning_rate": 2.4129371739045794e-06, + "loss": 3.1388, + "step": 19480 + }, + { + "epoch": 1.5617846906255637, + "grad_norm": 1.6789956092834473, + "learning_rate": 2.4120913867792572e-06, + "loss": 3.151, + "step": 19481 + }, + { + "epoch": 1.5618648654065863, + "grad_norm": 0.6821890473365784, + "learning_rate": 2.411245727583401e-06, + "loss": 2.4626, + "step": 19482 + }, + { + "epoch": 1.561945040187609, + "grad_norm": 1.8867902755737305, + "learning_rate": 2.4104001963312696e-06, + "loss": 3.1634, + "step": 19483 + }, + { + "epoch": 1.5620252149686316, + "grad_norm": 1.8386359214782715, + "learning_rate": 2.4095547930371123e-06, + "loss": 3.2341, + "step": 19484 + }, + { + "epoch": 1.5621053897496542, + "grad_norm": 1.7044095993041992, + "learning_rate": 2.4087095177151876e-06, + "loss": 3.1132, + "step": 19485 + }, + { + "epoch": 1.5621855645306768, + "grad_norm": 1.816142201423645, + "learning_rate": 2.4078643703797435e-06, + "loss": 3.2341, + "step": 19486 + }, + { + "epoch": 1.5622657393116994, + "grad_norm": 1.7789019346237183, + "learning_rate": 2.4070193510450312e-06, + "loss": 3.2815, + "step": 19487 + }, + { + "epoch": 1.562345914092722, + "grad_norm": 0.6677438616752625, + "learning_rate": 2.4061744597253e-06, + "loss": 2.4652, + "step": 19488 + }, + { + "epoch": 1.5624260888737447, + "grad_norm": 0.648307740688324, + "learning_rate": 2.4053296964347873e-06, + "loss": 2.4581, + "step": 19489 + }, + { + "epoch": 1.5625062636547673, + "grad_norm": 1.7740273475646973, + "learning_rate": 2.404485061187739e-06, + "loss": 3.1279, + "step": 19490 + }, + { + "epoch": 1.56258643843579, + "grad_norm": 1.7336426973342896, + "learning_rate": 2.4036405539983967e-06, + "loss": 3.0897, + "step": 19491 + }, + { + "epoch": 1.5626666132168126, + "grad_norm": 1.8128496408462524, + "learning_rate": 2.402796174880997e-06, + "loss": 3.1729, + "step": 19492 + }, + { + "epoch": 1.5627467879978352, + "grad_norm": 1.817703127861023, + "learning_rate": 2.401951923849779e-06, + "loss": 3.1274, + "step": 19493 + }, + { + "epoch": 1.5628269627788578, + "grad_norm": 1.7327176332473755, + "learning_rate": 2.4011078009189716e-06, + "loss": 3.2495, + "step": 19494 + }, + { + "epoch": 1.5629071375598804, + "grad_norm": 1.8387489318847656, + "learning_rate": 2.4002638061028073e-06, + "loss": 3.2234, + "step": 19495 + }, + { + "epoch": 1.562987312340903, + "grad_norm": 1.7666003704071045, + "learning_rate": 2.3994199394155182e-06, + "loss": 3.1715, + "step": 19496 + }, + { + "epoch": 1.5630674871219257, + "grad_norm": 1.7979713678359985, + "learning_rate": 2.3985762008713287e-06, + "loss": 3.2208, + "step": 19497 + }, + { + "epoch": 1.5631476619029483, + "grad_norm": 1.931948184967041, + "learning_rate": 2.3977325904844696e-06, + "loss": 3.2803, + "step": 19498 + }, + { + "epoch": 1.563227836683971, + "grad_norm": 1.7546197175979614, + "learning_rate": 2.396889108269156e-06, + "loss": 3.1775, + "step": 19499 + }, + { + "epoch": 1.5633080114649935, + "grad_norm": 1.7346558570861816, + "learning_rate": 2.396045754239613e-06, + "loss": 3.2031, + "step": 19500 + }, + { + "epoch": 1.5633881862460162, + "grad_norm": 1.8607215881347656, + "learning_rate": 2.395202528410059e-06, + "loss": 3.237, + "step": 19501 + }, + { + "epoch": 1.5634683610270388, + "grad_norm": 1.8485558032989502, + "learning_rate": 2.3943594307947095e-06, + "loss": 3.1758, + "step": 19502 + }, + { + "epoch": 1.5635485358080614, + "grad_norm": 1.7283681631088257, + "learning_rate": 2.3935164614077823e-06, + "loss": 3.1873, + "step": 19503 + }, + { + "epoch": 1.5636287105890843, + "grad_norm": 1.784993290901184, + "learning_rate": 2.392673620263485e-06, + "loss": 3.0753, + "step": 19504 + }, + { + "epoch": 1.5637088853701069, + "grad_norm": 1.8099794387817383, + "learning_rate": 2.391830907376028e-06, + "loss": 3.2126, + "step": 19505 + }, + { + "epoch": 1.5637890601511295, + "grad_norm": 1.8808034658432007, + "learning_rate": 2.390988322759621e-06, + "loss": 3.1953, + "step": 19506 + }, + { + "epoch": 1.5638692349321521, + "grad_norm": 1.7160205841064453, + "learning_rate": 2.3901458664284694e-06, + "loss": 3.2003, + "step": 19507 + }, + { + "epoch": 1.5639494097131748, + "grad_norm": 1.8643460273742676, + "learning_rate": 2.3893035383967787e-06, + "loss": 3.1468, + "step": 19508 + }, + { + "epoch": 1.5640295844941974, + "grad_norm": 1.8424656391143799, + "learning_rate": 2.388461338678746e-06, + "loss": 3.2462, + "step": 19509 + }, + { + "epoch": 1.56410975927522, + "grad_norm": 1.7435065507888794, + "learning_rate": 2.387619267288572e-06, + "loss": 3.1633, + "step": 19510 + }, + { + "epoch": 1.5641899340562426, + "grad_norm": 1.7885181903839111, + "learning_rate": 2.386777324240456e-06, + "loss": 3.1915, + "step": 19511 + }, + { + "epoch": 1.5642701088372652, + "grad_norm": 1.664812445640564, + "learning_rate": 2.385935509548589e-06, + "loss": 3.0835, + "step": 19512 + }, + { + "epoch": 1.5643502836182879, + "grad_norm": 1.696480393409729, + "learning_rate": 2.385093823227169e-06, + "loss": 3.2811, + "step": 19513 + }, + { + "epoch": 1.5644304583993105, + "grad_norm": 1.7335474491119385, + "learning_rate": 2.384252265290382e-06, + "loss": 3.1654, + "step": 19514 + }, + { + "epoch": 1.5645106331803331, + "grad_norm": 1.8755940198898315, + "learning_rate": 2.3834108357524168e-06, + "loss": 3.219, + "step": 19515 + }, + { + "epoch": 1.5645908079613557, + "grad_norm": 1.8702609539031982, + "learning_rate": 2.38256953462746e-06, + "loss": 3.1644, + "step": 19516 + }, + { + "epoch": 1.5646709827423784, + "grad_norm": 1.784963607788086, + "learning_rate": 2.3817283619296973e-06, + "loss": 3.1387, + "step": 19517 + }, + { + "epoch": 1.564751157523401, + "grad_norm": 1.7432438135147095, + "learning_rate": 2.3808873176733116e-06, + "loss": 3.1299, + "step": 19518 + }, + { + "epoch": 1.5648313323044236, + "grad_norm": 0.6591522097587585, + "learning_rate": 2.380046401872478e-06, + "loss": 2.4583, + "step": 19519 + }, + { + "epoch": 1.5649115070854462, + "grad_norm": 1.7943530082702637, + "learning_rate": 2.3792056145413754e-06, + "loss": 3.1594, + "step": 19520 + }, + { + "epoch": 1.5649916818664689, + "grad_norm": 1.7817957401275635, + "learning_rate": 2.3783649556941825e-06, + "loss": 3.193, + "step": 19521 + }, + { + "epoch": 1.5650718566474915, + "grad_norm": 1.7435729503631592, + "learning_rate": 2.377524425345068e-06, + "loss": 3.158, + "step": 19522 + }, + { + "epoch": 1.5651520314285141, + "grad_norm": 1.7214018106460571, + "learning_rate": 2.3766840235082102e-06, + "loss": 3.0661, + "step": 19523 + }, + { + "epoch": 1.5652322062095367, + "grad_norm": 1.7385953664779663, + "learning_rate": 2.375843750197768e-06, + "loss": 3.2245, + "step": 19524 + }, + { + "epoch": 1.5653123809905596, + "grad_norm": 1.800403118133545, + "learning_rate": 2.3750036054279147e-06, + "loss": 3.0812, + "step": 19525 + }, + { + "epoch": 1.5653925557715822, + "grad_norm": 1.8019412755966187, + "learning_rate": 2.3741635892128132e-06, + "loss": 3.1632, + "step": 19526 + }, + { + "epoch": 1.5654727305526048, + "grad_norm": 1.870190143585205, + "learning_rate": 2.3733237015666253e-06, + "loss": 3.2227, + "step": 19527 + }, + { + "epoch": 1.5655529053336275, + "grad_norm": 1.842079520225525, + "learning_rate": 2.3724839425035163e-06, + "loss": 3.2681, + "step": 19528 + }, + { + "epoch": 1.56563308011465, + "grad_norm": 1.7761839628219604, + "learning_rate": 2.3716443120376355e-06, + "loss": 3.2045, + "step": 19529 + }, + { + "epoch": 1.5657132548956727, + "grad_norm": 1.772778034210205, + "learning_rate": 2.370804810183144e-06, + "loss": 3.1769, + "step": 19530 + }, + { + "epoch": 1.5657934296766953, + "grad_norm": 1.7806200981140137, + "learning_rate": 2.3699654369541934e-06, + "loss": 3.1531, + "step": 19531 + }, + { + "epoch": 1.565873604457718, + "grad_norm": 1.799654483795166, + "learning_rate": 2.3691261923649366e-06, + "loss": 3.1395, + "step": 19532 + }, + { + "epoch": 1.5659537792387406, + "grad_norm": 1.802748441696167, + "learning_rate": 2.3682870764295274e-06, + "loss": 3.2068, + "step": 19533 + }, + { + "epoch": 1.5660339540197632, + "grad_norm": 1.7783112525939941, + "learning_rate": 2.367448089162104e-06, + "loss": 3.1796, + "step": 19534 + }, + { + "epoch": 1.5661141288007858, + "grad_norm": 1.8212717771530151, + "learning_rate": 2.366609230576816e-06, + "loss": 3.1484, + "step": 19535 + }, + { + "epoch": 1.5661943035818084, + "grad_norm": 1.739899754524231, + "learning_rate": 2.3657705006878064e-06, + "loss": 3.1579, + "step": 19536 + }, + { + "epoch": 1.566274478362831, + "grad_norm": 1.8666552305221558, + "learning_rate": 2.3649318995092152e-06, + "loss": 3.2024, + "step": 19537 + }, + { + "epoch": 1.5663546531438537, + "grad_norm": 1.6715707778930664, + "learning_rate": 2.3640934270551854e-06, + "loss": 3.1388, + "step": 19538 + }, + { + "epoch": 1.5664348279248763, + "grad_norm": 1.826458215713501, + "learning_rate": 2.363255083339845e-06, + "loss": 3.2683, + "step": 19539 + }, + { + "epoch": 1.566515002705899, + "grad_norm": 1.7534435987472534, + "learning_rate": 2.362416868377333e-06, + "loss": 3.0401, + "step": 19540 + }, + { + "epoch": 1.5665951774869216, + "grad_norm": 1.6915805339813232, + "learning_rate": 2.3615787821817803e-06, + "loss": 3.1494, + "step": 19541 + }, + { + "epoch": 1.5666753522679442, + "grad_norm": 1.7700002193450928, + "learning_rate": 2.3607408247673182e-06, + "loss": 3.1561, + "step": 19542 + }, + { + "epoch": 1.5667555270489668, + "grad_norm": 1.8015074729919434, + "learning_rate": 2.3599029961480757e-06, + "loss": 3.1013, + "step": 19543 + }, + { + "epoch": 1.5668357018299894, + "grad_norm": 1.7247042655944824, + "learning_rate": 2.359065296338173e-06, + "loss": 3.1572, + "step": 19544 + }, + { + "epoch": 1.566915876611012, + "grad_norm": 1.7942345142364502, + "learning_rate": 2.358227725351737e-06, + "loss": 3.2323, + "step": 19545 + }, + { + "epoch": 1.5669960513920347, + "grad_norm": 1.7478915452957153, + "learning_rate": 2.357390283202888e-06, + "loss": 3.2438, + "step": 19546 + }, + { + "epoch": 1.5670762261730573, + "grad_norm": 1.7509417533874512, + "learning_rate": 2.3565529699057467e-06, + "loss": 3.0925, + "step": 19547 + }, + { + "epoch": 1.56715640095408, + "grad_norm": 1.7006407976150513, + "learning_rate": 2.3557157854744296e-06, + "loss": 3.1714, + "step": 19548 + }, + { + "epoch": 1.5672365757351026, + "grad_norm": 1.7497429847717285, + "learning_rate": 2.354878729923048e-06, + "loss": 3.1651, + "step": 19549 + }, + { + "epoch": 1.5673167505161252, + "grad_norm": 1.807035207748413, + "learning_rate": 2.3540418032657163e-06, + "loss": 3.056, + "step": 19550 + }, + { + "epoch": 1.5673969252971478, + "grad_norm": 1.7400176525115967, + "learning_rate": 2.3532050055165457e-06, + "loss": 3.1108, + "step": 19551 + }, + { + "epoch": 1.5674771000781704, + "grad_norm": 1.7759417295455933, + "learning_rate": 2.3523683366896433e-06, + "loss": 3.1502, + "step": 19552 + }, + { + "epoch": 1.567557274859193, + "grad_norm": 1.778031349182129, + "learning_rate": 2.351531796799118e-06, + "loss": 3.2341, + "step": 19553 + }, + { + "epoch": 1.5676374496402157, + "grad_norm": 1.8258925676345825, + "learning_rate": 2.3506953858590687e-06, + "loss": 3.2205, + "step": 19554 + }, + { + "epoch": 1.5677176244212383, + "grad_norm": 1.7293386459350586, + "learning_rate": 2.3498591038835995e-06, + "loss": 3.1365, + "step": 19555 + }, + { + "epoch": 1.567797799202261, + "grad_norm": 0.6837702989578247, + "learning_rate": 2.349022950886809e-06, + "loss": 2.4694, + "step": 19556 + }, + { + "epoch": 1.5678779739832835, + "grad_norm": 1.8185646533966064, + "learning_rate": 2.348186926882796e-06, + "loss": 3.174, + "step": 19557 + }, + { + "epoch": 1.5679581487643062, + "grad_norm": 1.683327317237854, + "learning_rate": 2.3473510318856565e-06, + "loss": 3.1753, + "step": 19558 + }, + { + "epoch": 1.5680383235453288, + "grad_norm": 1.7210866212844849, + "learning_rate": 2.3465152659094793e-06, + "loss": 3.1285, + "step": 19559 + }, + { + "epoch": 1.5681184983263514, + "grad_norm": 1.8212686777114868, + "learning_rate": 2.345679628968357e-06, + "loss": 3.2324, + "step": 19560 + }, + { + "epoch": 1.568198673107374, + "grad_norm": 1.8501132726669312, + "learning_rate": 2.3448441210763785e-06, + "loss": 3.2286, + "step": 19561 + }, + { + "epoch": 1.5682788478883967, + "grad_norm": 1.8511322736740112, + "learning_rate": 2.344008742247631e-06, + "loss": 3.1075, + "step": 19562 + }, + { + "epoch": 1.5683590226694193, + "grad_norm": 1.7715016603469849, + "learning_rate": 2.3431734924962e-06, + "loss": 3.2371, + "step": 19563 + }, + { + "epoch": 1.568439197450442, + "grad_norm": 1.7897393703460693, + "learning_rate": 2.3423383718361625e-06, + "loss": 3.1698, + "step": 19564 + }, + { + "epoch": 1.5685193722314645, + "grad_norm": 1.705310583114624, + "learning_rate": 2.341503380281601e-06, + "loss": 3.2306, + "step": 19565 + }, + { + "epoch": 1.5685995470124872, + "grad_norm": 1.8506128787994385, + "learning_rate": 2.3406685178465937e-06, + "loss": 3.1204, + "step": 19566 + }, + { + "epoch": 1.5686797217935098, + "grad_norm": 1.82648503780365, + "learning_rate": 2.3398337845452167e-06, + "loss": 3.1865, + "step": 19567 + }, + { + "epoch": 1.5687598965745324, + "grad_norm": 1.6769227981567383, + "learning_rate": 2.3389991803915435e-06, + "loss": 3.1052, + "step": 19568 + }, + { + "epoch": 1.568840071355555, + "grad_norm": 1.664083480834961, + "learning_rate": 2.3381647053996425e-06, + "loss": 3.1978, + "step": 19569 + }, + { + "epoch": 1.5689202461365777, + "grad_norm": 1.7501674890518188, + "learning_rate": 2.337330359583584e-06, + "loss": 3.1358, + "step": 19570 + }, + { + "epoch": 1.5690004209176003, + "grad_norm": 2.07299542427063, + "learning_rate": 2.336496142957435e-06, + "loss": 3.2729, + "step": 19571 + }, + { + "epoch": 1.569080595698623, + "grad_norm": 1.8249708414077759, + "learning_rate": 2.3356620555352605e-06, + "loss": 3.1985, + "step": 19572 + }, + { + "epoch": 1.5691607704796455, + "grad_norm": 0.6738833785057068, + "learning_rate": 2.3348280973311245e-06, + "loss": 2.4739, + "step": 19573 + }, + { + "epoch": 1.5692409452606682, + "grad_norm": 1.7910958528518677, + "learning_rate": 2.333994268359083e-06, + "loss": 3.116, + "step": 19574 + }, + { + "epoch": 1.5693211200416908, + "grad_norm": 1.708121657371521, + "learning_rate": 2.333160568633197e-06, + "loss": 3.1681, + "step": 19575 + }, + { + "epoch": 1.5694012948227134, + "grad_norm": 1.8487240076065063, + "learning_rate": 2.332326998167521e-06, + "loss": 3.2827, + "step": 19576 + }, + { + "epoch": 1.569481469603736, + "grad_norm": 1.6980241537094116, + "learning_rate": 2.331493556976109e-06, + "loss": 3.1669, + "step": 19577 + }, + { + "epoch": 1.5695616443847586, + "grad_norm": 1.8732820749282837, + "learning_rate": 2.3306602450730164e-06, + "loss": 3.15, + "step": 19578 + }, + { + "epoch": 1.5696418191657813, + "grad_norm": 1.7905347347259521, + "learning_rate": 2.3298270624722863e-06, + "loss": 3.1049, + "step": 19579 + }, + { + "epoch": 1.569721993946804, + "grad_norm": 1.6657679080963135, + "learning_rate": 2.328994009187969e-06, + "loss": 3.0484, + "step": 19580 + }, + { + "epoch": 1.5698021687278265, + "grad_norm": 1.725214958190918, + "learning_rate": 2.3281610852341087e-06, + "loss": 3.1628, + "step": 19581 + }, + { + "epoch": 1.5698823435088491, + "grad_norm": 1.7953898906707764, + "learning_rate": 2.3273282906247495e-06, + "loss": 3.1721, + "step": 19582 + }, + { + "epoch": 1.5699625182898718, + "grad_norm": 1.701336145401001, + "learning_rate": 2.3264956253739335e-06, + "loss": 3.1398, + "step": 19583 + }, + { + "epoch": 1.5700426930708944, + "grad_norm": 1.80738365650177, + "learning_rate": 2.3256630894956933e-06, + "loss": 3.1288, + "step": 19584 + }, + { + "epoch": 1.570122867851917, + "grad_norm": 1.7800450325012207, + "learning_rate": 2.324830683004069e-06, + "loss": 3.2446, + "step": 19585 + }, + { + "epoch": 1.5702030426329399, + "grad_norm": 1.8715311288833618, + "learning_rate": 2.323998405913095e-06, + "loss": 3.1266, + "step": 19586 + }, + { + "epoch": 1.5702832174139625, + "grad_norm": 1.6785308122634888, + "learning_rate": 2.323166258236803e-06, + "loss": 3.1212, + "step": 19587 + }, + { + "epoch": 1.570363392194985, + "grad_norm": 0.6796497702598572, + "learning_rate": 2.3223342399892248e-06, + "loss": 2.4857, + "step": 19588 + }, + { + "epoch": 1.5704435669760077, + "grad_norm": 1.7489725351333618, + "learning_rate": 2.3215023511843816e-06, + "loss": 3.1702, + "step": 19589 + }, + { + "epoch": 1.5705237417570304, + "grad_norm": 1.8443775177001953, + "learning_rate": 2.320670591836304e-06, + "loss": 3.2076, + "step": 19590 + }, + { + "epoch": 1.570603916538053, + "grad_norm": 1.7234878540039062, + "learning_rate": 2.3198389619590124e-06, + "loss": 3.137, + "step": 19591 + }, + { + "epoch": 1.5706840913190756, + "grad_norm": 1.7954779863357544, + "learning_rate": 2.319007461566529e-06, + "loss": 3.2001, + "step": 19592 + }, + { + "epoch": 1.5707642661000982, + "grad_norm": 0.68422931432724, + "learning_rate": 2.3181760906728766e-06, + "loss": 2.4692, + "step": 19593 + }, + { + "epoch": 1.5708444408811209, + "grad_norm": 1.6917035579681396, + "learning_rate": 2.3173448492920656e-06, + "loss": 3.171, + "step": 19594 + }, + { + "epoch": 1.5709246156621435, + "grad_norm": 1.8253690004348755, + "learning_rate": 2.316513737438112e-06, + "loss": 3.1902, + "step": 19595 + }, + { + "epoch": 1.571004790443166, + "grad_norm": 1.823952078819275, + "learning_rate": 2.3156827551250283e-06, + "loss": 3.1311, + "step": 19596 + }, + { + "epoch": 1.5710849652241887, + "grad_norm": 1.7424135208129883, + "learning_rate": 2.3148519023668257e-06, + "loss": 3.1843, + "step": 19597 + }, + { + "epoch": 1.5711651400052113, + "grad_norm": 1.7776445150375366, + "learning_rate": 2.314021179177515e-06, + "loss": 3.0703, + "step": 19598 + }, + { + "epoch": 1.571245314786234, + "grad_norm": 1.6664490699768066, + "learning_rate": 2.3131905855710956e-06, + "loss": 3.1464, + "step": 19599 + }, + { + "epoch": 1.5713254895672566, + "grad_norm": 1.694115400314331, + "learning_rate": 2.3123601215615733e-06, + "loss": 3.2017, + "step": 19600 + }, + { + "epoch": 1.5714056643482792, + "grad_norm": 1.8524867296218872, + "learning_rate": 2.311529787162949e-06, + "loss": 3.078, + "step": 19601 + }, + { + "epoch": 1.5714858391293018, + "grad_norm": 1.7674864530563354, + "learning_rate": 2.310699582389224e-06, + "loss": 3.2037, + "step": 19602 + }, + { + "epoch": 1.5715660139103245, + "grad_norm": 1.867961049079895, + "learning_rate": 2.3098695072543963e-06, + "loss": 3.1794, + "step": 19603 + }, + { + "epoch": 1.571646188691347, + "grad_norm": 1.8237566947937012, + "learning_rate": 2.309039561772457e-06, + "loss": 3.1285, + "step": 19604 + }, + { + "epoch": 1.5717263634723697, + "grad_norm": 1.7557754516601562, + "learning_rate": 2.3082097459573982e-06, + "loss": 3.1906, + "step": 19605 + }, + { + "epoch": 1.5718065382533926, + "grad_norm": 1.7580420970916748, + "learning_rate": 2.3073800598232134e-06, + "loss": 3.1231, + "step": 19606 + }, + { + "epoch": 1.5718867130344152, + "grad_norm": 1.6854939460754395, + "learning_rate": 2.30655050338389e-06, + "loss": 3.0732, + "step": 19607 + }, + { + "epoch": 1.5719668878154378, + "grad_norm": 1.74349045753479, + "learning_rate": 2.3057210766534156e-06, + "loss": 3.2478, + "step": 19608 + }, + { + "epoch": 1.5720470625964604, + "grad_norm": 1.8718501329421997, + "learning_rate": 2.304891779645769e-06, + "loss": 3.1917, + "step": 19609 + }, + { + "epoch": 1.572127237377483, + "grad_norm": 1.7276335954666138, + "learning_rate": 2.3040626123749366e-06, + "loss": 3.2651, + "step": 19610 + }, + { + "epoch": 1.5722074121585057, + "grad_norm": 1.727246880531311, + "learning_rate": 2.303233574854895e-06, + "loss": 3.0262, + "step": 19611 + }, + { + "epoch": 1.5722875869395283, + "grad_norm": 1.7393587827682495, + "learning_rate": 2.3024046670996224e-06, + "loss": 3.1218, + "step": 19612 + }, + { + "epoch": 1.572367761720551, + "grad_norm": 1.7921310663223267, + "learning_rate": 2.301575889123098e-06, + "loss": 3.2064, + "step": 19613 + }, + { + "epoch": 1.5724479365015736, + "grad_norm": 1.686827301979065, + "learning_rate": 2.3007472409392872e-06, + "loss": 3.233, + "step": 19614 + }, + { + "epoch": 1.5725281112825962, + "grad_norm": 1.8127344846725464, + "learning_rate": 2.2999187225621653e-06, + "loss": 3.1578, + "step": 19615 + }, + { + "epoch": 1.5726082860636188, + "grad_norm": 1.8743896484375, + "learning_rate": 2.2990903340056993e-06, + "loss": 3.2019, + "step": 19616 + }, + { + "epoch": 1.5726884608446414, + "grad_norm": 1.6871585845947266, + "learning_rate": 2.298262075283857e-06, + "loss": 3.087, + "step": 19617 + }, + { + "epoch": 1.572768635625664, + "grad_norm": 1.7602167129516602, + "learning_rate": 2.2974339464106034e-06, + "loss": 3.2556, + "step": 19618 + }, + { + "epoch": 1.5728488104066867, + "grad_norm": 1.7035077810287476, + "learning_rate": 2.2966059473998968e-06, + "loss": 3.064, + "step": 19619 + }, + { + "epoch": 1.5729289851877093, + "grad_norm": 1.737166404724121, + "learning_rate": 2.295778078265699e-06, + "loss": 3.2034, + "step": 19620 + }, + { + "epoch": 1.573009159968732, + "grad_norm": 0.6520169973373413, + "learning_rate": 2.2949503390219674e-06, + "loss": 2.4619, + "step": 19621 + }, + { + "epoch": 1.5730893347497545, + "grad_norm": 1.8174538612365723, + "learning_rate": 2.2941227296826573e-06, + "loss": 3.158, + "step": 19622 + }, + { + "epoch": 1.5731695095307772, + "grad_norm": 1.6858937740325928, + "learning_rate": 2.2932952502617256e-06, + "loss": 3.1602, + "step": 19623 + }, + { + "epoch": 1.5732496843117998, + "grad_norm": 1.954964518547058, + "learning_rate": 2.2924679007731167e-06, + "loss": 3.2144, + "step": 19624 + }, + { + "epoch": 1.5733298590928224, + "grad_norm": 1.653261423110962, + "learning_rate": 2.2916406812307825e-06, + "loss": 3.1813, + "step": 19625 + }, + { + "epoch": 1.573410033873845, + "grad_norm": 1.771064043045044, + "learning_rate": 2.2908135916486697e-06, + "loss": 3.1973, + "step": 19626 + }, + { + "epoch": 1.5734902086548677, + "grad_norm": 1.7536238431930542, + "learning_rate": 2.2899866320407227e-06, + "loss": 3.1597, + "step": 19627 + }, + { + "epoch": 1.5735703834358903, + "grad_norm": 1.659989595413208, + "learning_rate": 2.2891598024208874e-06, + "loss": 3.1584, + "step": 19628 + }, + { + "epoch": 1.573650558216913, + "grad_norm": 1.7745466232299805, + "learning_rate": 2.2883331028030963e-06, + "loss": 3.2369, + "step": 19629 + }, + { + "epoch": 1.5737307329979355, + "grad_norm": 1.7963359355926514, + "learning_rate": 2.2875065332012923e-06, + "loss": 3.2312, + "step": 19630 + }, + { + "epoch": 1.5738109077789582, + "grad_norm": 1.694848895072937, + "learning_rate": 2.2866800936294087e-06, + "loss": 3.1768, + "step": 19631 + }, + { + "epoch": 1.5738910825599808, + "grad_norm": 1.81838059425354, + "learning_rate": 2.285853784101382e-06, + "loss": 3.1702, + "step": 19632 + }, + { + "epoch": 1.5739712573410034, + "grad_norm": 1.7299234867095947, + "learning_rate": 2.2850276046311436e-06, + "loss": 3.2269, + "step": 19633 + }, + { + "epoch": 1.574051432122026, + "grad_norm": 1.814146637916565, + "learning_rate": 2.284201555232618e-06, + "loss": 3.1329, + "step": 19634 + }, + { + "epoch": 1.5741316069030487, + "grad_norm": 1.8219372034072876, + "learning_rate": 2.2833756359197345e-06, + "loss": 3.1406, + "step": 19635 + }, + { + "epoch": 1.5742117816840713, + "grad_norm": 1.755476474761963, + "learning_rate": 2.282549846706419e-06, + "loss": 3.2636, + "step": 19636 + }, + { + "epoch": 1.574291956465094, + "grad_norm": 1.7660140991210938, + "learning_rate": 2.281724187606592e-06, + "loss": 3.1669, + "step": 19637 + }, + { + "epoch": 1.5743721312461165, + "grad_norm": 1.739522099494934, + "learning_rate": 2.2808986586341787e-06, + "loss": 3.2068, + "step": 19638 + }, + { + "epoch": 1.5744523060271391, + "grad_norm": 1.7786834239959717, + "learning_rate": 2.2800732598030906e-06, + "loss": 3.2388, + "step": 19639 + }, + { + "epoch": 1.5745324808081618, + "grad_norm": 1.7392240762710571, + "learning_rate": 2.279247991127247e-06, + "loss": 3.1512, + "step": 19640 + }, + { + "epoch": 1.5746126555891844, + "grad_norm": 1.8064664602279663, + "learning_rate": 2.278422852620561e-06, + "loss": 3.168, + "step": 19641 + }, + { + "epoch": 1.574692830370207, + "grad_norm": 1.7707184553146362, + "learning_rate": 2.2775978442969437e-06, + "loss": 3.1007, + "step": 19642 + }, + { + "epoch": 1.5747730051512296, + "grad_norm": 1.737797737121582, + "learning_rate": 2.2767729661703086e-06, + "loss": 3.0117, + "step": 19643 + }, + { + "epoch": 1.5748531799322523, + "grad_norm": 1.8704899549484253, + "learning_rate": 2.2759482182545566e-06, + "loss": 3.196, + "step": 19644 + }, + { + "epoch": 1.574933354713275, + "grad_norm": 1.698531985282898, + "learning_rate": 2.2751236005635946e-06, + "loss": 3.1023, + "step": 19645 + }, + { + "epoch": 1.5750135294942975, + "grad_norm": 1.7615249156951904, + "learning_rate": 2.274299113111327e-06, + "loss": 3.1879, + "step": 19646 + }, + { + "epoch": 1.5750937042753201, + "grad_norm": 1.7332260608673096, + "learning_rate": 2.2734747559116534e-06, + "loss": 3.2305, + "step": 19647 + }, + { + "epoch": 1.5751738790563428, + "grad_norm": 1.7321361303329468, + "learning_rate": 2.272650528978475e-06, + "loss": 3.1454, + "step": 19648 + }, + { + "epoch": 1.5752540538373654, + "grad_norm": 1.6928770542144775, + "learning_rate": 2.2718264323256822e-06, + "loss": 3.1371, + "step": 19649 + }, + { + "epoch": 1.575334228618388, + "grad_norm": 1.7165582180023193, + "learning_rate": 2.2710024659671726e-06, + "loss": 3.0129, + "step": 19650 + }, + { + "epoch": 1.5754144033994106, + "grad_norm": 1.7710702419281006, + "learning_rate": 2.270178629916837e-06, + "loss": 3.136, + "step": 19651 + }, + { + "epoch": 1.5754945781804333, + "grad_norm": 1.7381030321121216, + "learning_rate": 2.269354924188566e-06, + "loss": 3.2322, + "step": 19652 + }, + { + "epoch": 1.5755747529614559, + "grad_norm": 1.8938674926757812, + "learning_rate": 2.268531348796249e-06, + "loss": 3.1298, + "step": 19653 + }, + { + "epoch": 1.5756549277424785, + "grad_norm": 1.6521931886672974, + "learning_rate": 2.267707903753765e-06, + "loss": 3.168, + "step": 19654 + }, + { + "epoch": 1.5757351025235011, + "grad_norm": 1.883886694908142, + "learning_rate": 2.266884589075001e-06, + "loss": 3.2179, + "step": 19655 + }, + { + "epoch": 1.5758152773045238, + "grad_norm": 1.8493348360061646, + "learning_rate": 2.2660614047738383e-06, + "loss": 3.0924, + "step": 19656 + }, + { + "epoch": 1.5758954520855464, + "grad_norm": 1.7615679502487183, + "learning_rate": 2.265238350864154e-06, + "loss": 3.1283, + "step": 19657 + }, + { + "epoch": 1.575975626866569, + "grad_norm": 1.7623708248138428, + "learning_rate": 2.2644154273598285e-06, + "loss": 3.2052, + "step": 19658 + }, + { + "epoch": 1.5760558016475916, + "grad_norm": 1.792102575302124, + "learning_rate": 2.2635926342747296e-06, + "loss": 3.1678, + "step": 19659 + }, + { + "epoch": 1.5761359764286142, + "grad_norm": 1.766869306564331, + "learning_rate": 2.262769971622731e-06, + "loss": 3.0638, + "step": 19660 + }, + { + "epoch": 1.5762161512096369, + "grad_norm": 1.792609453201294, + "learning_rate": 2.2619474394177055e-06, + "loss": 3.2525, + "step": 19661 + }, + { + "epoch": 1.5762963259906595, + "grad_norm": 1.7712970972061157, + "learning_rate": 2.261125037673518e-06, + "loss": 3.209, + "step": 19662 + }, + { + "epoch": 1.5763765007716821, + "grad_norm": 1.737027645111084, + "learning_rate": 2.260302766404039e-06, + "loss": 3.2161, + "step": 19663 + }, + { + "epoch": 1.5764566755527047, + "grad_norm": 1.9170764684677124, + "learning_rate": 2.259480625623124e-06, + "loss": 3.1929, + "step": 19664 + }, + { + "epoch": 1.5765368503337274, + "grad_norm": 0.685429573059082, + "learning_rate": 2.2586586153446365e-06, + "loss": 2.4733, + "step": 19665 + }, + { + "epoch": 1.57661702511475, + "grad_norm": 1.7457499504089355, + "learning_rate": 2.2578367355824382e-06, + "loss": 3.245, + "step": 19666 + }, + { + "epoch": 1.5766971998957728, + "grad_norm": 1.677128791809082, + "learning_rate": 2.2570149863503833e-06, + "loss": 3.1995, + "step": 19667 + }, + { + "epoch": 1.5767773746767955, + "grad_norm": 1.8625996112823486, + "learning_rate": 2.2561933676623292e-06, + "loss": 3.1815, + "step": 19668 + }, + { + "epoch": 1.576857549457818, + "grad_norm": 1.7292039394378662, + "learning_rate": 2.2553718795321232e-06, + "loss": 3.1151, + "step": 19669 + }, + { + "epoch": 1.5769377242388407, + "grad_norm": 1.734766960144043, + "learning_rate": 2.2545505219736173e-06, + "loss": 3.1844, + "step": 19670 + }, + { + "epoch": 1.5770178990198633, + "grad_norm": 1.822314977645874, + "learning_rate": 2.2537292950006605e-06, + "loss": 3.2009, + "step": 19671 + }, + { + "epoch": 1.577098073800886, + "grad_norm": 1.7746864557266235, + "learning_rate": 2.2529081986270982e-06, + "loss": 3.1377, + "step": 19672 + }, + { + "epoch": 1.5771782485819086, + "grad_norm": 1.6524600982666016, + "learning_rate": 2.2520872328667754e-06, + "loss": 3.0261, + "step": 19673 + }, + { + "epoch": 1.5772584233629312, + "grad_norm": 0.6550518870353699, + "learning_rate": 2.251266397733529e-06, + "loss": 2.4639, + "step": 19674 + }, + { + "epoch": 1.5773385981439538, + "grad_norm": 1.7234007120132446, + "learning_rate": 2.2504456932412e-06, + "loss": 3.1559, + "step": 19675 + }, + { + "epoch": 1.5774187729249765, + "grad_norm": 1.8234834671020508, + "learning_rate": 2.2496251194036245e-06, + "loss": 3.061, + "step": 19676 + }, + { + "epoch": 1.577498947705999, + "grad_norm": 0.671403169631958, + "learning_rate": 2.2488046762346395e-06, + "loss": 2.4575, + "step": 19677 + }, + { + "epoch": 1.5775791224870217, + "grad_norm": 1.7740662097930908, + "learning_rate": 2.247984363748078e-06, + "loss": 3.1644, + "step": 19678 + }, + { + "epoch": 1.5776592972680443, + "grad_norm": 1.8255589008331299, + "learning_rate": 2.247164181957765e-06, + "loss": 3.1937, + "step": 19679 + }, + { + "epoch": 1.577739472049067, + "grad_norm": 1.766763687133789, + "learning_rate": 2.2463441308775314e-06, + "loss": 3.1556, + "step": 19680 + }, + { + "epoch": 1.5778196468300896, + "grad_norm": 1.7828727960586548, + "learning_rate": 2.245524210521204e-06, + "loss": 3.1299, + "step": 19681 + }, + { + "epoch": 1.5778998216111122, + "grad_norm": 1.7552144527435303, + "learning_rate": 2.244704420902605e-06, + "loss": 3.1321, + "step": 19682 + }, + { + "epoch": 1.5779799963921348, + "grad_norm": 1.86885666847229, + "learning_rate": 2.2438847620355587e-06, + "loss": 3.241, + "step": 19683 + }, + { + "epoch": 1.5780601711731574, + "grad_norm": 1.7790027856826782, + "learning_rate": 2.243065233933879e-06, + "loss": 3.2478, + "step": 19684 + }, + { + "epoch": 1.57814034595418, + "grad_norm": 1.7790104150772095, + "learning_rate": 2.242245836611385e-06, + "loss": 3.2565, + "step": 19685 + }, + { + "epoch": 1.5782205207352027, + "grad_norm": 1.8194788694381714, + "learning_rate": 2.241426570081894e-06, + "loss": 3.2118, + "step": 19686 + }, + { + "epoch": 1.5783006955162253, + "grad_norm": 1.706172227859497, + "learning_rate": 2.2406074343592155e-06, + "loss": 3.1084, + "step": 19687 + }, + { + "epoch": 1.5783808702972482, + "grad_norm": 1.7285685539245605, + "learning_rate": 2.2397884294571636e-06, + "loss": 3.1864, + "step": 19688 + }, + { + "epoch": 1.5784610450782708, + "grad_norm": 1.87336003780365, + "learning_rate": 2.2389695553895406e-06, + "loss": 3.198, + "step": 19689 + }, + { + "epoch": 1.5785412198592934, + "grad_norm": 1.7942783832550049, + "learning_rate": 2.2381508121701565e-06, + "loss": 3.1328, + "step": 19690 + }, + { + "epoch": 1.578621394640316, + "grad_norm": 1.718072533607483, + "learning_rate": 2.237332199812814e-06, + "loss": 3.1904, + "step": 19691 + }, + { + "epoch": 1.5787015694213387, + "grad_norm": 1.7428892850875854, + "learning_rate": 2.2365137183313158e-06, + "loss": 3.1883, + "step": 19692 + }, + { + "epoch": 1.5787817442023613, + "grad_norm": 0.6789404153823853, + "learning_rate": 2.2356953677394588e-06, + "loss": 2.4691, + "step": 19693 + }, + { + "epoch": 1.578861918983384, + "grad_norm": 1.8839647769927979, + "learning_rate": 2.2348771480510444e-06, + "loss": 3.2302, + "step": 19694 + }, + { + "epoch": 1.5789420937644065, + "grad_norm": 1.9271000623703003, + "learning_rate": 2.2340590592798615e-06, + "loss": 3.2693, + "step": 19695 + }, + { + "epoch": 1.5790222685454292, + "grad_norm": 1.8709383010864258, + "learning_rate": 2.233241101439707e-06, + "loss": 3.1392, + "step": 19696 + }, + { + "epoch": 1.5791024433264518, + "grad_norm": 1.8278632164001465, + "learning_rate": 2.23242327454437e-06, + "loss": 3.1039, + "step": 19697 + }, + { + "epoch": 1.5791826181074744, + "grad_norm": 1.799217939376831, + "learning_rate": 2.2316055786076386e-06, + "loss": 3.1727, + "step": 19698 + }, + { + "epoch": 1.579262792888497, + "grad_norm": 1.8114310503005981, + "learning_rate": 2.2307880136433024e-06, + "loss": 3.3757, + "step": 19699 + }, + { + "epoch": 1.5793429676695196, + "grad_norm": 1.768561601638794, + "learning_rate": 2.2299705796651385e-06, + "loss": 3.1013, + "step": 19700 + }, + { + "epoch": 1.5794231424505423, + "grad_norm": 1.6396621465682983, + "learning_rate": 2.229153276686933e-06, + "loss": 3.1958, + "step": 19701 + }, + { + "epoch": 1.579503317231565, + "grad_norm": 1.7633026838302612, + "learning_rate": 2.228336104722465e-06, + "loss": 3.1484, + "step": 19702 + }, + { + "epoch": 1.5795834920125875, + "grad_norm": 1.7953755855560303, + "learning_rate": 2.2275190637855105e-06, + "loss": 3.1548, + "step": 19703 + }, + { + "epoch": 1.5796636667936101, + "grad_norm": 1.7834587097167969, + "learning_rate": 2.2267021538898482e-06, + "loss": 3.1279, + "step": 19704 + }, + { + "epoch": 1.5797438415746328, + "grad_norm": 1.8619670867919922, + "learning_rate": 2.225885375049245e-06, + "loss": 3.1998, + "step": 19705 + }, + { + "epoch": 1.5798240163556554, + "grad_norm": 1.8854891061782837, + "learning_rate": 2.225068727277475e-06, + "loss": 3.3002, + "step": 19706 + }, + { + "epoch": 1.579904191136678, + "grad_norm": 1.8722114562988281, + "learning_rate": 2.2242522105883056e-06, + "loss": 3.2144, + "step": 19707 + }, + { + "epoch": 1.5799843659177006, + "grad_norm": 1.7534480094909668, + "learning_rate": 2.2234358249955035e-06, + "loss": 3.0793, + "step": 19708 + }, + { + "epoch": 1.5800645406987233, + "grad_norm": 1.7120534181594849, + "learning_rate": 2.2226195705128363e-06, + "loss": 3.0048, + "step": 19709 + }, + { + "epoch": 1.5801447154797459, + "grad_norm": 1.7818807363510132, + "learning_rate": 2.2218034471540583e-06, + "loss": 3.1991, + "step": 19710 + }, + { + "epoch": 1.5802248902607685, + "grad_norm": 1.8242626190185547, + "learning_rate": 2.2209874549329335e-06, + "loss": 3.2839, + "step": 19711 + }, + { + "epoch": 1.5803050650417911, + "grad_norm": 1.7274972200393677, + "learning_rate": 2.2201715938632186e-06, + "loss": 3.134, + "step": 19712 + }, + { + "epoch": 1.5803852398228138, + "grad_norm": 1.8263038396835327, + "learning_rate": 2.219355863958669e-06, + "loss": 3.1345, + "step": 19713 + }, + { + "epoch": 1.5804654146038364, + "grad_norm": 1.7702207565307617, + "learning_rate": 2.2185402652330402e-06, + "loss": 3.2139, + "step": 19714 + }, + { + "epoch": 1.580545589384859, + "grad_norm": 1.7722485065460205, + "learning_rate": 2.217724797700076e-06, + "loss": 3.089, + "step": 19715 + }, + { + "epoch": 1.5806257641658816, + "grad_norm": 0.6692527532577515, + "learning_rate": 2.21690946137353e-06, + "loss": 2.454, + "step": 19716 + }, + { + "epoch": 1.5807059389469043, + "grad_norm": 1.7036464214324951, + "learning_rate": 2.216094256267147e-06, + "loss": 3.1532, + "step": 19717 + }, + { + "epoch": 1.5807861137279269, + "grad_norm": 1.8092434406280518, + "learning_rate": 2.215279182394672e-06, + "loss": 3.2388, + "step": 19718 + }, + { + "epoch": 1.5808662885089495, + "grad_norm": 1.736416220664978, + "learning_rate": 2.214464239769849e-06, + "loss": 3.0983, + "step": 19719 + }, + { + "epoch": 1.5809464632899721, + "grad_norm": 1.7626540660858154, + "learning_rate": 2.2136494284064124e-06, + "loss": 3.2276, + "step": 19720 + }, + { + "epoch": 1.5810266380709947, + "grad_norm": 1.750066876411438, + "learning_rate": 2.2128347483181024e-06, + "loss": 3.1567, + "step": 19721 + }, + { + "epoch": 1.5811068128520174, + "grad_norm": 1.8474769592285156, + "learning_rate": 2.2120201995186542e-06, + "loss": 3.1187, + "step": 19722 + }, + { + "epoch": 1.58118698763304, + "grad_norm": 1.7287222146987915, + "learning_rate": 2.2112057820218e-06, + "loss": 3.1761, + "step": 19723 + }, + { + "epoch": 1.5812671624140626, + "grad_norm": 1.9161183834075928, + "learning_rate": 2.210391495841274e-06, + "loss": 3.0981, + "step": 19724 + }, + { + "epoch": 1.5813473371950852, + "grad_norm": 1.7342873811721802, + "learning_rate": 2.2095773409907994e-06, + "loss": 3.0749, + "step": 19725 + }, + { + "epoch": 1.5814275119761079, + "grad_norm": 1.7567578554153442, + "learning_rate": 2.208763317484106e-06, + "loss": 3.1315, + "step": 19726 + }, + { + "epoch": 1.5815076867571305, + "grad_norm": 1.7565609216690063, + "learning_rate": 2.2079494253349165e-06, + "loss": 3.1483, + "step": 19727 + }, + { + "epoch": 1.5815878615381531, + "grad_norm": 1.7011833190917969, + "learning_rate": 2.207135664556953e-06, + "loss": 3.1802, + "step": 19728 + }, + { + "epoch": 1.5816680363191757, + "grad_norm": 1.7700178623199463, + "learning_rate": 2.2063220351639394e-06, + "loss": 3.0947, + "step": 19729 + }, + { + "epoch": 1.5817482111001984, + "grad_norm": 1.8298487663269043, + "learning_rate": 2.205508537169586e-06, + "loss": 3.149, + "step": 19730 + }, + { + "epoch": 1.581828385881221, + "grad_norm": 1.7619695663452148, + "learning_rate": 2.204695170587613e-06, + "loss": 3.2319, + "step": 19731 + }, + { + "epoch": 1.5819085606622436, + "grad_norm": 1.8521373271942139, + "learning_rate": 2.203881935431731e-06, + "loss": 3.1165, + "step": 19732 + }, + { + "epoch": 1.5819887354432662, + "grad_norm": 1.740986943244934, + "learning_rate": 2.2030688317156525e-06, + "loss": 3.2095, + "step": 19733 + }, + { + "epoch": 1.5820689102242889, + "grad_norm": 1.6868557929992676, + "learning_rate": 2.2022558594530874e-06, + "loss": 3.1166, + "step": 19734 + }, + { + "epoch": 1.5821490850053115, + "grad_norm": 1.7215321063995361, + "learning_rate": 2.2014430186577386e-06, + "loss": 3.0912, + "step": 19735 + }, + { + "epoch": 1.582229259786334, + "grad_norm": 1.7334245443344116, + "learning_rate": 2.2006303093433112e-06, + "loss": 3.2015, + "step": 19736 + }, + { + "epoch": 1.5823094345673567, + "grad_norm": 1.777295470237732, + "learning_rate": 2.199817731523508e-06, + "loss": 3.2246, + "step": 19737 + }, + { + "epoch": 1.5823896093483794, + "grad_norm": 1.713173747062683, + "learning_rate": 2.1990052852120293e-06, + "loss": 3.1915, + "step": 19738 + }, + { + "epoch": 1.582469784129402, + "grad_norm": 1.757967472076416, + "learning_rate": 2.1981929704225747e-06, + "loss": 3.192, + "step": 19739 + }, + { + "epoch": 1.5825499589104246, + "grad_norm": 1.8794986009597778, + "learning_rate": 2.1973807871688346e-06, + "loss": 3.2437, + "step": 19740 + }, + { + "epoch": 1.5826301336914472, + "grad_norm": 1.7727479934692383, + "learning_rate": 2.1965687354645036e-06, + "loss": 3.1595, + "step": 19741 + }, + { + "epoch": 1.5827103084724699, + "grad_norm": 1.6855432987213135, + "learning_rate": 2.1957568153232743e-06, + "loss": 3.1706, + "step": 19742 + }, + { + "epoch": 1.5827904832534925, + "grad_norm": 1.7233803272247314, + "learning_rate": 2.1949450267588344e-06, + "loss": 3.2106, + "step": 19743 + }, + { + "epoch": 1.582870658034515, + "grad_norm": 1.778114676475525, + "learning_rate": 2.1941333697848744e-06, + "loss": 3.2103, + "step": 19744 + }, + { + "epoch": 1.5829508328155377, + "grad_norm": 1.7332807779312134, + "learning_rate": 2.1933218444150706e-06, + "loss": 3.1862, + "step": 19745 + }, + { + "epoch": 1.5830310075965603, + "grad_norm": 1.7999117374420166, + "learning_rate": 2.192510450663109e-06, + "loss": 3.2031, + "step": 19746 + }, + { + "epoch": 1.583111182377583, + "grad_norm": 1.8104979991912842, + "learning_rate": 2.191699188542671e-06, + "loss": 3.1809, + "step": 19747 + }, + { + "epoch": 1.5831913571586056, + "grad_norm": 1.6723064184188843, + "learning_rate": 2.190888058067432e-06, + "loss": 3.1586, + "step": 19748 + }, + { + "epoch": 1.5832715319396284, + "grad_norm": 1.7597218751907349, + "learning_rate": 2.19007705925107e-06, + "loss": 3.1074, + "step": 19749 + }, + { + "epoch": 1.583351706720651, + "grad_norm": 1.6743677854537964, + "learning_rate": 2.1892661921072544e-06, + "loss": 3.1184, + "step": 19750 + }, + { + "epoch": 1.5834318815016737, + "grad_norm": 1.8488781452178955, + "learning_rate": 2.1884554566496587e-06, + "loss": 3.2248, + "step": 19751 + }, + { + "epoch": 1.5835120562826963, + "grad_norm": 1.7766578197479248, + "learning_rate": 2.187644852891949e-06, + "loss": 3.157, + "step": 19752 + }, + { + "epoch": 1.583592231063719, + "grad_norm": 1.7964134216308594, + "learning_rate": 2.1868343808477954e-06, + "loss": 3.1321, + "step": 19753 + }, + { + "epoch": 1.5836724058447416, + "grad_norm": 1.7579773664474487, + "learning_rate": 2.186024040530863e-06, + "loss": 3.1036, + "step": 19754 + }, + { + "epoch": 1.5837525806257642, + "grad_norm": 1.9054219722747803, + "learning_rate": 2.1852138319548087e-06, + "loss": 3.1931, + "step": 19755 + }, + { + "epoch": 1.5838327554067868, + "grad_norm": 1.715468406677246, + "learning_rate": 2.184403755133294e-06, + "loss": 3.1465, + "step": 19756 + }, + { + "epoch": 1.5839129301878094, + "grad_norm": 1.775383472442627, + "learning_rate": 2.1835938100799793e-06, + "loss": 3.1376, + "step": 19757 + }, + { + "epoch": 1.583993104968832, + "grad_norm": 1.7481049299240112, + "learning_rate": 2.1827839968085163e-06, + "loss": 3.2194, + "step": 19758 + }, + { + "epoch": 1.5840732797498547, + "grad_norm": 1.7537546157836914, + "learning_rate": 2.1819743153325635e-06, + "loss": 3.1936, + "step": 19759 + }, + { + "epoch": 1.5841534545308773, + "grad_norm": 0.6506038904190063, + "learning_rate": 2.1811647656657666e-06, + "loss": 2.4678, + "step": 19760 + }, + { + "epoch": 1.5842336293119, + "grad_norm": 1.8337043523788452, + "learning_rate": 2.1803553478217742e-06, + "loss": 3.0943, + "step": 19761 + }, + { + "epoch": 1.5843138040929226, + "grad_norm": 1.6934045553207397, + "learning_rate": 2.179546061814236e-06, + "loss": 3.084, + "step": 19762 + }, + { + "epoch": 1.5843939788739452, + "grad_norm": 1.7353519201278687, + "learning_rate": 2.1787369076567954e-06, + "loss": 3.1105, + "step": 19763 + }, + { + "epoch": 1.5844741536549678, + "grad_norm": 1.81449556350708, + "learning_rate": 2.1779278853630957e-06, + "loss": 3.285, + "step": 19764 + }, + { + "epoch": 1.5845543284359904, + "grad_norm": 1.8381048440933228, + "learning_rate": 2.177118994946773e-06, + "loss": 3.2321, + "step": 19765 + }, + { + "epoch": 1.584634503217013, + "grad_norm": 1.745381236076355, + "learning_rate": 2.176310236421467e-06, + "loss": 3.1674, + "step": 19766 + }, + { + "epoch": 1.5847146779980357, + "grad_norm": 1.8137980699539185, + "learning_rate": 2.175501609800813e-06, + "loss": 3.3053, + "step": 19767 + }, + { + "epoch": 1.5847948527790583, + "grad_norm": 1.8069767951965332, + "learning_rate": 2.174693115098444e-06, + "loss": 3.2175, + "step": 19768 + }, + { + "epoch": 1.5848750275600811, + "grad_norm": 1.7580138444900513, + "learning_rate": 2.173884752327995e-06, + "loss": 3.1523, + "step": 19769 + }, + { + "epoch": 1.5849552023411038, + "grad_norm": 1.7215156555175781, + "learning_rate": 2.173076521503087e-06, + "loss": 3.0841, + "step": 19770 + }, + { + "epoch": 1.5850353771221264, + "grad_norm": 1.7263431549072266, + "learning_rate": 2.1722684226373503e-06, + "loss": 3.2451, + "step": 19771 + }, + { + "epoch": 1.585115551903149, + "grad_norm": 1.9188146591186523, + "learning_rate": 2.1714604557444095e-06, + "loss": 3.2421, + "step": 19772 + }, + { + "epoch": 1.5851957266841716, + "grad_norm": 1.904260516166687, + "learning_rate": 2.1706526208378866e-06, + "loss": 3.1265, + "step": 19773 + }, + { + "epoch": 1.5852759014651943, + "grad_norm": 1.760014295578003, + "learning_rate": 2.169844917931404e-06, + "loss": 3.0988, + "step": 19774 + }, + { + "epoch": 1.5853560762462169, + "grad_norm": 1.8600239753723145, + "learning_rate": 2.169037347038573e-06, + "loss": 3.2122, + "step": 19775 + }, + { + "epoch": 1.5854362510272395, + "grad_norm": 0.6632816791534424, + "learning_rate": 2.1682299081730127e-06, + "loss": 2.469, + "step": 19776 + }, + { + "epoch": 1.5855164258082621, + "grad_norm": 1.7632439136505127, + "learning_rate": 2.1674226013483367e-06, + "loss": 3.1981, + "step": 19777 + }, + { + "epoch": 1.5855966005892848, + "grad_norm": 1.7770991325378418, + "learning_rate": 2.1666154265781546e-06, + "loss": 3.1774, + "step": 19778 + }, + { + "epoch": 1.5856767753703074, + "grad_norm": 0.6806126236915588, + "learning_rate": 2.165808383876078e-06, + "loss": 2.4774, + "step": 19779 + }, + { + "epoch": 1.58575695015133, + "grad_norm": 1.7755779027938843, + "learning_rate": 2.165001473255709e-06, + "loss": 3.0748, + "step": 19780 + }, + { + "epoch": 1.5858371249323526, + "grad_norm": 1.6478646993637085, + "learning_rate": 2.1641946947306536e-06, + "loss": 3.1099, + "step": 19781 + }, + { + "epoch": 1.5859172997133753, + "grad_norm": 1.8783273696899414, + "learning_rate": 2.163388048314514e-06, + "loss": 3.1139, + "step": 19782 + }, + { + "epoch": 1.5859974744943979, + "grad_norm": 1.7943588495254517, + "learning_rate": 2.16258153402089e-06, + "loss": 3.1327, + "step": 19783 + }, + { + "epoch": 1.5860776492754205, + "grad_norm": 1.752082109451294, + "learning_rate": 2.1617751518633823e-06, + "loss": 3.1418, + "step": 19784 + }, + { + "epoch": 1.5861578240564431, + "grad_norm": 1.8167439699172974, + "learning_rate": 2.1609689018555802e-06, + "loss": 3.1552, + "step": 19785 + }, + { + "epoch": 1.5862379988374657, + "grad_norm": 1.691460132598877, + "learning_rate": 2.1601627840110785e-06, + "loss": 3.2139, + "step": 19786 + }, + { + "epoch": 1.5863181736184884, + "grad_norm": 1.8727105855941772, + "learning_rate": 2.1593567983434704e-06, + "loss": 3.3159, + "step": 19787 + }, + { + "epoch": 1.586398348399511, + "grad_norm": 1.8209559917449951, + "learning_rate": 2.1585509448663435e-06, + "loss": 3.131, + "step": 19788 + }, + { + "epoch": 1.5864785231805336, + "grad_norm": 1.7899699211120605, + "learning_rate": 2.157745223593286e-06, + "loss": 3.2242, + "step": 19789 + }, + { + "epoch": 1.5865586979615562, + "grad_norm": 1.7244693040847778, + "learning_rate": 2.156939634537878e-06, + "loss": 3.0695, + "step": 19790 + }, + { + "epoch": 1.5866388727425789, + "grad_norm": 1.6932246685028076, + "learning_rate": 2.1561341777137034e-06, + "loss": 3.1956, + "step": 19791 + }, + { + "epoch": 1.5867190475236015, + "grad_norm": 1.6690239906311035, + "learning_rate": 2.155328853134342e-06, + "loss": 3.02, + "step": 19792 + }, + { + "epoch": 1.5867992223046241, + "grad_norm": 0.6555830836296082, + "learning_rate": 2.1545236608133724e-06, + "loss": 2.4653, + "step": 19793 + }, + { + "epoch": 1.5868793970856467, + "grad_norm": 1.8479558229446411, + "learning_rate": 2.1537186007643705e-06, + "loss": 3.1737, + "step": 19794 + }, + { + "epoch": 1.5869595718666694, + "grad_norm": 1.805544376373291, + "learning_rate": 2.1529136730009047e-06, + "loss": 3.2507, + "step": 19795 + }, + { + "epoch": 1.587039746647692, + "grad_norm": 1.7372331619262695, + "learning_rate": 2.1521088775365497e-06, + "loss": 3.1924, + "step": 19796 + }, + { + "epoch": 1.5871199214287146, + "grad_norm": 1.7757827043533325, + "learning_rate": 2.1513042143848726e-06, + "loss": 3.2103, + "step": 19797 + }, + { + "epoch": 1.5872000962097372, + "grad_norm": 1.805241584777832, + "learning_rate": 2.1504996835594405e-06, + "loss": 3.1694, + "step": 19798 + }, + { + "epoch": 1.5872802709907599, + "grad_norm": 1.7972207069396973, + "learning_rate": 2.1496952850738207e-06, + "loss": 3.2374, + "step": 19799 + }, + { + "epoch": 1.5873604457717825, + "grad_norm": 1.8415082693099976, + "learning_rate": 2.148891018941568e-06, + "loss": 3.2056, + "step": 19800 + }, + { + "epoch": 1.587440620552805, + "grad_norm": 1.759898066520691, + "learning_rate": 2.148086885176246e-06, + "loss": 3.1959, + "step": 19801 + }, + { + "epoch": 1.5875207953338277, + "grad_norm": 1.8810657262802124, + "learning_rate": 2.1472828837914117e-06, + "loss": 3.1686, + "step": 19802 + }, + { + "epoch": 1.5876009701148504, + "grad_norm": 1.7272881269454956, + "learning_rate": 2.1464790148006198e-06, + "loss": 3.1572, + "step": 19803 + }, + { + "epoch": 1.587681144895873, + "grad_norm": 1.7486909627914429, + "learning_rate": 2.1456752782174282e-06, + "loss": 3.1857, + "step": 19804 + }, + { + "epoch": 1.5877613196768956, + "grad_norm": 1.6985893249511719, + "learning_rate": 2.14487167405538e-06, + "loss": 3.1359, + "step": 19805 + }, + { + "epoch": 1.5878414944579182, + "grad_norm": 1.7294715642929077, + "learning_rate": 2.1440682023280258e-06, + "loss": 3.1393, + "step": 19806 + }, + { + "epoch": 1.5879216692389408, + "grad_norm": 1.7336405515670776, + "learning_rate": 2.143264863048914e-06, + "loss": 3.1317, + "step": 19807 + }, + { + "epoch": 1.5880018440199635, + "grad_norm": 1.73603093624115, + "learning_rate": 2.142461656231587e-06, + "loss": 3.2502, + "step": 19808 + }, + { + "epoch": 1.588082018800986, + "grad_norm": 1.7460321187973022, + "learning_rate": 2.1416585818895906e-06, + "loss": 3.0908, + "step": 19809 + }, + { + "epoch": 1.5881621935820087, + "grad_norm": 1.8072305917739868, + "learning_rate": 2.140855640036459e-06, + "loss": 3.214, + "step": 19810 + }, + { + "epoch": 1.5882423683630313, + "grad_norm": 1.8718880414962769, + "learning_rate": 2.1400528306857306e-06, + "loss": 3.2183, + "step": 19811 + }, + { + "epoch": 1.588322543144054, + "grad_norm": 1.9917709827423096, + "learning_rate": 2.1392501538509402e-06, + "loss": 3.1399, + "step": 19812 + }, + { + "epoch": 1.5884027179250766, + "grad_norm": 1.7715343236923218, + "learning_rate": 2.138447609545623e-06, + "loss": 3.1754, + "step": 19813 + }, + { + "epoch": 1.5884828927060992, + "grad_norm": 1.7317156791687012, + "learning_rate": 2.137645197783311e-06, + "loss": 3.0955, + "step": 19814 + }, + { + "epoch": 1.5885630674871218, + "grad_norm": 1.6661320924758911, + "learning_rate": 2.1368429185775275e-06, + "loss": 3.1321, + "step": 19815 + }, + { + "epoch": 1.5886432422681445, + "grad_norm": 1.8020809888839722, + "learning_rate": 2.136040771941801e-06, + "loss": 3.187, + "step": 19816 + }, + { + "epoch": 1.588723417049167, + "grad_norm": 1.7136733531951904, + "learning_rate": 2.1352387578896563e-06, + "loss": 3.1866, + "step": 19817 + }, + { + "epoch": 1.5888035918301897, + "grad_norm": 1.8410673141479492, + "learning_rate": 2.1344368764346124e-06, + "loss": 3.187, + "step": 19818 + }, + { + "epoch": 1.5888837666112123, + "grad_norm": 0.6686134338378906, + "learning_rate": 2.1336351275901946e-06, + "loss": 2.4585, + "step": 19819 + }, + { + "epoch": 1.588963941392235, + "grad_norm": 1.7581428289413452, + "learning_rate": 2.132833511369914e-06, + "loss": 3.1322, + "step": 19820 + }, + { + "epoch": 1.5890441161732576, + "grad_norm": 1.8173526525497437, + "learning_rate": 2.1320320277872874e-06, + "loss": 3.2234, + "step": 19821 + }, + { + "epoch": 1.5891242909542802, + "grad_norm": 1.8198636770248413, + "learning_rate": 2.1312306768558268e-06, + "loss": 3.2138, + "step": 19822 + }, + { + "epoch": 1.5892044657353028, + "grad_norm": 1.7453479766845703, + "learning_rate": 2.1304294585890452e-06, + "loss": 3.183, + "step": 19823 + }, + { + "epoch": 1.5892846405163255, + "grad_norm": 1.7598729133605957, + "learning_rate": 2.129628373000451e-06, + "loss": 3.17, + "step": 19824 + }, + { + "epoch": 1.589364815297348, + "grad_norm": 1.7885265350341797, + "learning_rate": 2.128827420103546e-06, + "loss": 3.226, + "step": 19825 + }, + { + "epoch": 1.5894449900783707, + "grad_norm": 1.6906598806381226, + "learning_rate": 2.1280265999118376e-06, + "loss": 3.063, + "step": 19826 + }, + { + "epoch": 1.5895251648593933, + "grad_norm": 1.7636481523513794, + "learning_rate": 2.1272259124388263e-06, + "loss": 3.1448, + "step": 19827 + }, + { + "epoch": 1.589605339640416, + "grad_norm": 1.8189061880111694, + "learning_rate": 2.126425357698011e-06, + "loss": 3.2483, + "step": 19828 + }, + { + "epoch": 1.5896855144214386, + "grad_norm": 1.734904408454895, + "learning_rate": 2.125624935702891e-06, + "loss": 3.2622, + "step": 19829 + }, + { + "epoch": 1.5897656892024614, + "grad_norm": 0.6700955033302307, + "learning_rate": 2.1248246464669585e-06, + "loss": 2.4734, + "step": 19830 + }, + { + "epoch": 1.589845863983484, + "grad_norm": 1.9248778820037842, + "learning_rate": 2.1240244900037066e-06, + "loss": 3.2249, + "step": 19831 + }, + { + "epoch": 1.5899260387645067, + "grad_norm": 1.7788578271865845, + "learning_rate": 2.1232244663266255e-06, + "loss": 3.197, + "step": 19832 + }, + { + "epoch": 1.5900062135455293, + "grad_norm": 1.78165602684021, + "learning_rate": 2.1224245754492033e-06, + "loss": 3.0436, + "step": 19833 + }, + { + "epoch": 1.590086388326552, + "grad_norm": 1.7555116415023804, + "learning_rate": 2.1216248173849297e-06, + "loss": 3.2584, + "step": 19834 + }, + { + "epoch": 1.5901665631075745, + "grad_norm": 1.7333147525787354, + "learning_rate": 2.120825192147282e-06, + "loss": 3.2817, + "step": 19835 + }, + { + "epoch": 1.5902467378885972, + "grad_norm": 1.9869656562805176, + "learning_rate": 2.120025699749746e-06, + "loss": 3.1636, + "step": 19836 + }, + { + "epoch": 1.5903269126696198, + "grad_norm": 1.7141895294189453, + "learning_rate": 2.1192263402057987e-06, + "loss": 3.1503, + "step": 19837 + }, + { + "epoch": 1.5904070874506424, + "grad_norm": 1.7584877014160156, + "learning_rate": 2.118427113528917e-06, + "loss": 3.1935, + "step": 19838 + }, + { + "epoch": 1.590487262231665, + "grad_norm": 1.7426635026931763, + "learning_rate": 2.1176280197325805e-06, + "loss": 3.2079, + "step": 19839 + }, + { + "epoch": 1.5905674370126877, + "grad_norm": 1.6784216165542603, + "learning_rate": 2.1168290588302544e-06, + "loss": 3.1646, + "step": 19840 + }, + { + "epoch": 1.5906476117937103, + "grad_norm": 2.032632827758789, + "learning_rate": 2.116030230835412e-06, + "loss": 3.2961, + "step": 19841 + }, + { + "epoch": 1.590727786574733, + "grad_norm": 1.7786225080490112, + "learning_rate": 2.1152315357615206e-06, + "loss": 3.1586, + "step": 19842 + }, + { + "epoch": 1.5908079613557555, + "grad_norm": 1.8996745347976685, + "learning_rate": 2.114432973622048e-06, + "loss": 3.1772, + "step": 19843 + }, + { + "epoch": 1.5908881361367782, + "grad_norm": 0.6536622643470764, + "learning_rate": 2.1136345444304583e-06, + "loss": 2.4832, + "step": 19844 + }, + { + "epoch": 1.5909683109178008, + "grad_norm": 1.781455636024475, + "learning_rate": 2.1128362482002074e-06, + "loss": 3.1801, + "step": 19845 + }, + { + "epoch": 1.5910484856988234, + "grad_norm": 1.7869682312011719, + "learning_rate": 2.1120380849447585e-06, + "loss": 3.1134, + "step": 19846 + }, + { + "epoch": 1.591128660479846, + "grad_norm": 1.862168312072754, + "learning_rate": 2.111240054677568e-06, + "loss": 3.254, + "step": 19847 + }, + { + "epoch": 1.5912088352608686, + "grad_norm": 1.7583681344985962, + "learning_rate": 2.1104421574120892e-06, + "loss": 3.1171, + "step": 19848 + }, + { + "epoch": 1.5912890100418913, + "grad_norm": 1.7968535423278809, + "learning_rate": 2.109644393161777e-06, + "loss": 3.1989, + "step": 19849 + }, + { + "epoch": 1.5913691848229141, + "grad_norm": 1.75690758228302, + "learning_rate": 2.1088467619400775e-06, + "loss": 3.1287, + "step": 19850 + }, + { + "epoch": 1.5914493596039367, + "grad_norm": 1.6965221166610718, + "learning_rate": 2.1080492637604413e-06, + "loss": 3.1879, + "step": 19851 + }, + { + "epoch": 1.5915295343849594, + "grad_norm": 1.8099712133407593, + "learning_rate": 2.1072518986363122e-06, + "loss": 3.2554, + "step": 19852 + }, + { + "epoch": 1.591609709165982, + "grad_norm": 1.914223074913025, + "learning_rate": 2.1064546665811346e-06, + "loss": 3.1803, + "step": 19853 + }, + { + "epoch": 1.5916898839470046, + "grad_norm": 1.9761910438537598, + "learning_rate": 2.1056575676083525e-06, + "loss": 3.1526, + "step": 19854 + }, + { + "epoch": 1.5917700587280272, + "grad_norm": 1.8248101472854614, + "learning_rate": 2.1048606017313986e-06, + "loss": 3.1428, + "step": 19855 + }, + { + "epoch": 1.5918502335090499, + "grad_norm": 1.762742519378662, + "learning_rate": 2.1040637689637123e-06, + "loss": 3.2815, + "step": 19856 + }, + { + "epoch": 1.5919304082900725, + "grad_norm": 1.8281716108322144, + "learning_rate": 2.103267069318727e-06, + "loss": 3.2365, + "step": 19857 + }, + { + "epoch": 1.592010583071095, + "grad_norm": 1.9507793188095093, + "learning_rate": 2.102470502809878e-06, + "loss": 3.1311, + "step": 19858 + }, + { + "epoch": 1.5920907578521177, + "grad_norm": 1.8959765434265137, + "learning_rate": 2.1016740694505944e-06, + "loss": 3.3029, + "step": 19859 + }, + { + "epoch": 1.5921709326331404, + "grad_norm": 1.7765498161315918, + "learning_rate": 2.1008777692543003e-06, + "loss": 3.0939, + "step": 19860 + }, + { + "epoch": 1.592251107414163, + "grad_norm": 1.7634226083755493, + "learning_rate": 2.100081602234422e-06, + "loss": 3.2063, + "step": 19861 + }, + { + "epoch": 1.5923312821951856, + "grad_norm": 1.9895883798599243, + "learning_rate": 2.0992855684043847e-06, + "loss": 3.1537, + "step": 19862 + }, + { + "epoch": 1.5924114569762082, + "grad_norm": 0.6516119241714478, + "learning_rate": 2.098489667777608e-06, + "loss": 2.4618, + "step": 19863 + }, + { + "epoch": 1.5924916317572309, + "grad_norm": 1.7361868619918823, + "learning_rate": 2.0976939003675134e-06, + "loss": 3.2899, + "step": 19864 + }, + { + "epoch": 1.5925718065382535, + "grad_norm": 1.7849652767181396, + "learning_rate": 2.0968982661875115e-06, + "loss": 3.0699, + "step": 19865 + }, + { + "epoch": 1.592651981319276, + "grad_norm": 1.7050728797912598, + "learning_rate": 2.096102765251019e-06, + "loss": 3.2167, + "step": 19866 + }, + { + "epoch": 1.5927321561002987, + "grad_norm": 1.726564884185791, + "learning_rate": 2.09530739757145e-06, + "loss": 3.1459, + "step": 19867 + }, + { + "epoch": 1.5928123308813213, + "grad_norm": 1.7168635129928589, + "learning_rate": 2.0945121631622112e-06, + "loss": 3.176, + "step": 19868 + }, + { + "epoch": 1.592892505662344, + "grad_norm": 1.7665339708328247, + "learning_rate": 2.093717062036714e-06, + "loss": 3.1675, + "step": 19869 + }, + { + "epoch": 1.5929726804433666, + "grad_norm": 0.6723809838294983, + "learning_rate": 2.0929220942083585e-06, + "loss": 2.4536, + "step": 19870 + }, + { + "epoch": 1.5930528552243892, + "grad_norm": 1.7370023727416992, + "learning_rate": 2.0921272596905497e-06, + "loss": 3.1735, + "step": 19871 + }, + { + "epoch": 1.5931330300054118, + "grad_norm": 1.7744325399398804, + "learning_rate": 2.091332558496687e-06, + "loss": 3.2239, + "step": 19872 + }, + { + "epoch": 1.5932132047864345, + "grad_norm": 1.751694917678833, + "learning_rate": 2.090537990640171e-06, + "loss": 3.1448, + "step": 19873 + }, + { + "epoch": 1.593293379567457, + "grad_norm": 1.6801979541778564, + "learning_rate": 2.0897435561344005e-06, + "loss": 3.1312, + "step": 19874 + }, + { + "epoch": 1.5933735543484797, + "grad_norm": 1.9088574647903442, + "learning_rate": 2.0889492549927613e-06, + "loss": 3.2347, + "step": 19875 + }, + { + "epoch": 1.5934537291295023, + "grad_norm": 0.6733301877975464, + "learning_rate": 2.0881550872286506e-06, + "loss": 2.4786, + "step": 19876 + }, + { + "epoch": 1.593533903910525, + "grad_norm": 1.7917481660842896, + "learning_rate": 2.0873610528554566e-06, + "loss": 3.1557, + "step": 19877 + }, + { + "epoch": 1.5936140786915476, + "grad_norm": 1.7538460493087769, + "learning_rate": 2.086567151886567e-06, + "loss": 3.1343, + "step": 19878 + }, + { + "epoch": 1.5936942534725702, + "grad_norm": 0.6767707467079163, + "learning_rate": 2.0857733843353687e-06, + "loss": 2.4827, + "step": 19879 + }, + { + "epoch": 1.5937744282535928, + "grad_norm": 1.759766697883606, + "learning_rate": 2.084979750215239e-06, + "loss": 3.1887, + "step": 19880 + }, + { + "epoch": 1.5938546030346155, + "grad_norm": 1.9994418621063232, + "learning_rate": 2.0841862495395616e-06, + "loss": 3.1919, + "step": 19881 + }, + { + "epoch": 1.593934777815638, + "grad_norm": 1.7465013265609741, + "learning_rate": 2.083392882321713e-06, + "loss": 3.164, + "step": 19882 + }, + { + "epoch": 1.5940149525966607, + "grad_norm": 1.7038391828536987, + "learning_rate": 2.082599648575071e-06, + "loss": 3.1532, + "step": 19883 + }, + { + "epoch": 1.5940951273776833, + "grad_norm": 1.7734389305114746, + "learning_rate": 2.0818065483130123e-06, + "loss": 3.2328, + "step": 19884 + }, + { + "epoch": 1.594175302158706, + "grad_norm": 0.6776929497718811, + "learning_rate": 2.0810135815489007e-06, + "loss": 2.4552, + "step": 19885 + }, + { + "epoch": 1.5942554769397286, + "grad_norm": 1.833443522453308, + "learning_rate": 2.080220748296109e-06, + "loss": 3.1138, + "step": 19886 + }, + { + "epoch": 1.5943356517207512, + "grad_norm": 1.7845127582550049, + "learning_rate": 2.079428048568005e-06, + "loss": 3.1373, + "step": 19887 + }, + { + "epoch": 1.5944158265017738, + "grad_norm": 1.883650779724121, + "learning_rate": 2.078635482377953e-06, + "loss": 3.2587, + "step": 19888 + }, + { + "epoch": 1.5944960012827964, + "grad_norm": 1.8113077878952026, + "learning_rate": 2.0778430497393166e-06, + "loss": 3.1897, + "step": 19889 + }, + { + "epoch": 1.594576176063819, + "grad_norm": 1.8867579698562622, + "learning_rate": 2.0770507506654526e-06, + "loss": 3.2775, + "step": 19890 + }, + { + "epoch": 1.5946563508448417, + "grad_norm": 1.7553036212921143, + "learning_rate": 2.076258585169719e-06, + "loss": 3.1687, + "step": 19891 + }, + { + "epoch": 1.5947365256258643, + "grad_norm": 1.7120351791381836, + "learning_rate": 2.075466553265475e-06, + "loss": 3.19, + "step": 19892 + }, + { + "epoch": 1.594816700406887, + "grad_norm": 1.8006328344345093, + "learning_rate": 2.074674654966071e-06, + "loss": 3.1632, + "step": 19893 + }, + { + "epoch": 1.5948968751879096, + "grad_norm": 1.8023185729980469, + "learning_rate": 2.073882890284863e-06, + "loss": 3.1515, + "step": 19894 + }, + { + "epoch": 1.5949770499689322, + "grad_norm": 1.8258178234100342, + "learning_rate": 2.0730912592351917e-06, + "loss": 3.1498, + "step": 19895 + }, + { + "epoch": 1.5950572247499548, + "grad_norm": 1.902453064918518, + "learning_rate": 2.0722997618304086e-06, + "loss": 3.2175, + "step": 19896 + }, + { + "epoch": 1.5951373995309774, + "grad_norm": 1.8107489347457886, + "learning_rate": 2.071508398083858e-06, + "loss": 3.2415, + "step": 19897 + }, + { + "epoch": 1.595217574312, + "grad_norm": 1.7769619226455688, + "learning_rate": 2.070717168008881e-06, + "loss": 3.1505, + "step": 19898 + }, + { + "epoch": 1.5952977490930227, + "grad_norm": 1.7213791608810425, + "learning_rate": 2.069926071618822e-06, + "loss": 3.2101, + "step": 19899 + }, + { + "epoch": 1.5953779238740453, + "grad_norm": 1.676252841949463, + "learning_rate": 2.0691351089270108e-06, + "loss": 3.0741, + "step": 19900 + }, + { + "epoch": 1.595458098655068, + "grad_norm": 1.7094467878341675, + "learning_rate": 2.068344279946786e-06, + "loss": 3.115, + "step": 19901 + }, + { + "epoch": 1.5955382734360906, + "grad_norm": 1.8076777458190918, + "learning_rate": 2.067553584691482e-06, + "loss": 3.2105, + "step": 19902 + }, + { + "epoch": 1.5956184482171132, + "grad_norm": 1.7730954885482788, + "learning_rate": 2.0667630231744283e-06, + "loss": 3.1202, + "step": 19903 + }, + { + "epoch": 1.5956986229981358, + "grad_norm": 1.8202608823776245, + "learning_rate": 2.0659725954089573e-06, + "loss": 3.1238, + "step": 19904 + }, + { + "epoch": 1.5957787977791584, + "grad_norm": 1.8238122463226318, + "learning_rate": 2.0651823014083895e-06, + "loss": 3.218, + "step": 19905 + }, + { + "epoch": 1.595858972560181, + "grad_norm": 1.8246164321899414, + "learning_rate": 2.064392141186051e-06, + "loss": 3.1414, + "step": 19906 + }, + { + "epoch": 1.5959391473412037, + "grad_norm": 1.6854056119918823, + "learning_rate": 2.0636021147552653e-06, + "loss": 3.1085, + "step": 19907 + }, + { + "epoch": 1.5960193221222263, + "grad_norm": 1.8075114488601685, + "learning_rate": 2.06281222212935e-06, + "loss": 3.2667, + "step": 19908 + }, + { + "epoch": 1.596099496903249, + "grad_norm": 1.740742564201355, + "learning_rate": 2.0620224633216266e-06, + "loss": 3.051, + "step": 19909 + }, + { + "epoch": 1.5961796716842716, + "grad_norm": 1.7906768321990967, + "learning_rate": 2.0612328383454037e-06, + "loss": 3.1543, + "step": 19910 + }, + { + "epoch": 1.5962598464652944, + "grad_norm": 1.8805426359176636, + "learning_rate": 2.0604433472139974e-06, + "loss": 3.2865, + "step": 19911 + }, + { + "epoch": 1.596340021246317, + "grad_norm": 1.672142744064331, + "learning_rate": 2.059653989940719e-06, + "loss": 3.1553, + "step": 19912 + }, + { + "epoch": 1.5964201960273396, + "grad_norm": 1.8921983242034912, + "learning_rate": 2.058864766538874e-06, + "loss": 3.2395, + "step": 19913 + }, + { + "epoch": 1.5965003708083623, + "grad_norm": 1.8431026935577393, + "learning_rate": 2.0580756770217747e-06, + "loss": 3.1951, + "step": 19914 + }, + { + "epoch": 1.596580545589385, + "grad_norm": 1.8138114213943481, + "learning_rate": 2.057286721402717e-06, + "loss": 3.1395, + "step": 19915 + }, + { + "epoch": 1.5966607203704075, + "grad_norm": 1.893811583518982, + "learning_rate": 2.0564978996950056e-06, + "loss": 3.1216, + "step": 19916 + }, + { + "epoch": 1.5967408951514301, + "grad_norm": 1.8760689496994019, + "learning_rate": 2.0557092119119404e-06, + "loss": 3.2666, + "step": 19917 + }, + { + "epoch": 1.5968210699324528, + "grad_norm": 1.868324637413025, + "learning_rate": 2.054920658066818e-06, + "loss": 3.1417, + "step": 19918 + }, + { + "epoch": 1.5969012447134754, + "grad_norm": 1.8116086721420288, + "learning_rate": 2.0541322381729355e-06, + "loss": 3.1296, + "step": 19919 + }, + { + "epoch": 1.596981419494498, + "grad_norm": 1.8137601613998413, + "learning_rate": 2.0533439522435818e-06, + "loss": 3.1823, + "step": 19920 + }, + { + "epoch": 1.5970615942755206, + "grad_norm": 1.7301157712936401, + "learning_rate": 2.0525558002920466e-06, + "loss": 3.1353, + "step": 19921 + }, + { + "epoch": 1.5971417690565433, + "grad_norm": 1.7884128093719482, + "learning_rate": 2.051767782331621e-06, + "loss": 3.1753, + "step": 19922 + }, + { + "epoch": 1.5972219438375659, + "grad_norm": 1.8123376369476318, + "learning_rate": 2.050979898375588e-06, + "loss": 3.2153, + "step": 19923 + }, + { + "epoch": 1.5973021186185885, + "grad_norm": 1.744857668876648, + "learning_rate": 2.0501921484372365e-06, + "loss": 3.1891, + "step": 19924 + }, + { + "epoch": 1.5973822933996111, + "grad_norm": 1.7747732400894165, + "learning_rate": 2.0494045325298404e-06, + "loss": 3.1925, + "step": 19925 + }, + { + "epoch": 1.5974624681806338, + "grad_norm": 1.7726200819015503, + "learning_rate": 2.0486170506666814e-06, + "loss": 3.0468, + "step": 19926 + }, + { + "epoch": 1.5975426429616564, + "grad_norm": 0.6872404217720032, + "learning_rate": 2.047829702861037e-06, + "loss": 2.4835, + "step": 19927 + }, + { + "epoch": 1.597622817742679, + "grad_norm": 1.7070344686508179, + "learning_rate": 2.0470424891261807e-06, + "loss": 3.0902, + "step": 19928 + }, + { + "epoch": 1.5977029925237016, + "grad_norm": 0.6910985708236694, + "learning_rate": 2.0462554094753885e-06, + "loss": 2.4518, + "step": 19929 + }, + { + "epoch": 1.5977831673047243, + "grad_norm": 1.7007062435150146, + "learning_rate": 2.045468463921925e-06, + "loss": 3.067, + "step": 19930 + }, + { + "epoch": 1.5978633420857469, + "grad_norm": 1.7778445482254028, + "learning_rate": 2.044681652479058e-06, + "loss": 3.1502, + "step": 19931 + }, + { + "epoch": 1.5979435168667697, + "grad_norm": 1.8509328365325928, + "learning_rate": 2.0438949751600555e-06, + "loss": 3.2051, + "step": 19932 + }, + { + "epoch": 1.5980236916477923, + "grad_norm": 1.7727172374725342, + "learning_rate": 2.0431084319781792e-06, + "loss": 3.208, + "step": 19933 + }, + { + "epoch": 1.598103866428815, + "grad_norm": 1.7920408248901367, + "learning_rate": 2.0423220229466935e-06, + "loss": 3.2332, + "step": 19934 + }, + { + "epoch": 1.5981840412098376, + "grad_norm": 1.7450886964797974, + "learning_rate": 2.041535748078851e-06, + "loss": 3.1886, + "step": 19935 + }, + { + "epoch": 1.5982642159908602, + "grad_norm": 0.6749354600906372, + "learning_rate": 2.0407496073879117e-06, + "loss": 2.4609, + "step": 19936 + }, + { + "epoch": 1.5983443907718828, + "grad_norm": 1.7037044763565063, + "learning_rate": 2.039963600887128e-06, + "loss": 3.1543, + "step": 19937 + }, + { + "epoch": 1.5984245655529055, + "grad_norm": 1.785552978515625, + "learning_rate": 2.039177728589753e-06, + "loss": 3.2449, + "step": 19938 + }, + { + "epoch": 1.598504740333928, + "grad_norm": 1.7388561964035034, + "learning_rate": 2.0383919905090386e-06, + "loss": 3.1714, + "step": 19939 + }, + { + "epoch": 1.5985849151149507, + "grad_norm": 1.7144579887390137, + "learning_rate": 2.037606386658226e-06, + "loss": 3.1384, + "step": 19940 + }, + { + "epoch": 1.5986650898959733, + "grad_norm": 1.9147894382476807, + "learning_rate": 2.036820917050565e-06, + "loss": 3.2297, + "step": 19941 + }, + { + "epoch": 1.598745264676996, + "grad_norm": 1.6621514558792114, + "learning_rate": 2.036035581699296e-06, + "loss": 3.2321, + "step": 19942 + }, + { + "epoch": 1.5988254394580186, + "grad_norm": 1.7707898616790771, + "learning_rate": 2.035250380617659e-06, + "loss": 3.1588, + "step": 19943 + }, + { + "epoch": 1.5989056142390412, + "grad_norm": 1.842735767364502, + "learning_rate": 2.0344653138188983e-06, + "loss": 3.2073, + "step": 19944 + }, + { + "epoch": 1.5989857890200638, + "grad_norm": 1.9386459589004517, + "learning_rate": 2.033680381316241e-06, + "loss": 3.1537, + "step": 19945 + }, + { + "epoch": 1.5990659638010865, + "grad_norm": 1.7381511926651, + "learning_rate": 2.032895583122926e-06, + "loss": 3.2056, + "step": 19946 + }, + { + "epoch": 1.599146138582109, + "grad_norm": 1.8238352537155151, + "learning_rate": 2.0321109192521828e-06, + "loss": 3.1735, + "step": 19947 + }, + { + "epoch": 1.5992263133631317, + "grad_norm": 1.814798355102539, + "learning_rate": 2.031326389717242e-06, + "loss": 3.2203, + "step": 19948 + }, + { + "epoch": 1.5993064881441543, + "grad_norm": 1.8433594703674316, + "learning_rate": 2.030541994531331e-06, + "loss": 3.1546, + "step": 19949 + }, + { + "epoch": 1.599386662925177, + "grad_norm": 1.7814135551452637, + "learning_rate": 2.0297577337076722e-06, + "loss": 3.1431, + "step": 19950 + }, + { + "epoch": 1.5994668377061996, + "grad_norm": 1.7348723411560059, + "learning_rate": 2.0289736072594877e-06, + "loss": 3.2358, + "step": 19951 + }, + { + "epoch": 1.5995470124872222, + "grad_norm": 1.8203848600387573, + "learning_rate": 2.0281896152e-06, + "loss": 3.2906, + "step": 19952 + }, + { + "epoch": 1.5996271872682448, + "grad_norm": 1.673262357711792, + "learning_rate": 2.0274057575424254e-06, + "loss": 3.2697, + "step": 19953 + }, + { + "epoch": 1.5997073620492674, + "grad_norm": 1.7464042901992798, + "learning_rate": 2.026622034299983e-06, + "loss": 3.0878, + "step": 19954 + }, + { + "epoch": 1.59978753683029, + "grad_norm": 1.7292299270629883, + "learning_rate": 2.0258384454858803e-06, + "loss": 3.1376, + "step": 19955 + }, + { + "epoch": 1.5998677116113127, + "grad_norm": 1.7154204845428467, + "learning_rate": 2.025054991113331e-06, + "loss": 3.2412, + "step": 19956 + }, + { + "epoch": 1.5999478863923353, + "grad_norm": 1.8730056285858154, + "learning_rate": 2.0242716711955436e-06, + "loss": 3.2156, + "step": 19957 + }, + { + "epoch": 1.600028061173358, + "grad_norm": 0.6738923788070679, + "learning_rate": 2.0234884857457247e-06, + "loss": 2.4476, + "step": 19958 + }, + { + "epoch": 1.6001082359543806, + "grad_norm": 1.7575713396072388, + "learning_rate": 2.0227054347770813e-06, + "loss": 3.1763, + "step": 19959 + }, + { + "epoch": 1.6001884107354032, + "grad_norm": 1.7588021755218506, + "learning_rate": 2.021922518302811e-06, + "loss": 3.1759, + "step": 19960 + }, + { + "epoch": 1.6002685855164258, + "grad_norm": 1.8309217691421509, + "learning_rate": 2.021139736336115e-06, + "loss": 3.2013, + "step": 19961 + }, + { + "epoch": 1.6003487602974484, + "grad_norm": 1.9431490898132324, + "learning_rate": 2.02035708889019e-06, + "loss": 3.1243, + "step": 19962 + }, + { + "epoch": 1.600428935078471, + "grad_norm": 1.7759664058685303, + "learning_rate": 2.0195745759782336e-06, + "loss": 3.0891, + "step": 19963 + }, + { + "epoch": 1.6005091098594937, + "grad_norm": 1.839938759803772, + "learning_rate": 2.0187921976134395e-06, + "loss": 3.2188, + "step": 19964 + }, + { + "epoch": 1.6005892846405163, + "grad_norm": 1.7398114204406738, + "learning_rate": 2.0180099538089927e-06, + "loss": 3.2227, + "step": 19965 + }, + { + "epoch": 1.600669459421539, + "grad_norm": 1.8844202756881714, + "learning_rate": 2.0172278445780857e-06, + "loss": 3.2709, + "step": 19966 + }, + { + "epoch": 1.6007496342025616, + "grad_norm": 1.803614854812622, + "learning_rate": 2.016445869933904e-06, + "loss": 3.225, + "step": 19967 + }, + { + "epoch": 1.6008298089835842, + "grad_norm": 1.790298581123352, + "learning_rate": 2.0156640298896314e-06, + "loss": 3.1897, + "step": 19968 + }, + { + "epoch": 1.6009099837646068, + "grad_norm": 1.7323039770126343, + "learning_rate": 2.0148823244584513e-06, + "loss": 3.1964, + "step": 19969 + }, + { + "epoch": 1.6009901585456294, + "grad_norm": 1.7520047426223755, + "learning_rate": 2.0141007536535384e-06, + "loss": 3.2148, + "step": 19970 + }, + { + "epoch": 1.601070333326652, + "grad_norm": 1.7632091045379639, + "learning_rate": 2.0133193174880718e-06, + "loss": 3.2845, + "step": 19971 + }, + { + "epoch": 1.6011505081076747, + "grad_norm": 1.772118330001831, + "learning_rate": 2.0125380159752273e-06, + "loss": 3.1947, + "step": 19972 + }, + { + "epoch": 1.6012306828886973, + "grad_norm": 0.6759819388389587, + "learning_rate": 2.011756849128177e-06, + "loss": 2.4756, + "step": 19973 + }, + { + "epoch": 1.60131085766972, + "grad_norm": 1.7325032949447632, + "learning_rate": 2.0109758169600935e-06, + "loss": 3.0511, + "step": 19974 + }, + { + "epoch": 1.6013910324507425, + "grad_norm": 1.8834891319274902, + "learning_rate": 2.0101949194841396e-06, + "loss": 3.0985, + "step": 19975 + }, + { + "epoch": 1.6014712072317652, + "grad_norm": 1.8491448163986206, + "learning_rate": 2.009414156713483e-06, + "loss": 3.1696, + "step": 19976 + }, + { + "epoch": 1.6015513820127878, + "grad_norm": 1.7804380655288696, + "learning_rate": 2.0086335286612877e-06, + "loss": 3.1095, + "step": 19977 + }, + { + "epoch": 1.6016315567938104, + "grad_norm": 1.759927749633789, + "learning_rate": 2.007853035340714e-06, + "loss": 3.1984, + "step": 19978 + }, + { + "epoch": 1.601711731574833, + "grad_norm": 1.7539331912994385, + "learning_rate": 2.007072676764926e-06, + "loss": 3.2329, + "step": 19979 + }, + { + "epoch": 1.6017919063558557, + "grad_norm": 1.774968147277832, + "learning_rate": 2.0062924529470717e-06, + "loss": 3.2074, + "step": 19980 + }, + { + "epoch": 1.6018720811368783, + "grad_norm": 1.9049822092056274, + "learning_rate": 2.0055123639003094e-06, + "loss": 3.1262, + "step": 19981 + }, + { + "epoch": 1.601952255917901, + "grad_norm": 1.8027287721633911, + "learning_rate": 2.004732409637792e-06, + "loss": 3.139, + "step": 19982 + }, + { + "epoch": 1.6020324306989235, + "grad_norm": 1.737849235534668, + "learning_rate": 2.0039525901726664e-06, + "loss": 3.1542, + "step": 19983 + }, + { + "epoch": 1.6021126054799462, + "grad_norm": 1.6691724061965942, + "learning_rate": 2.003172905518087e-06, + "loss": 3.0632, + "step": 19984 + }, + { + "epoch": 1.6021927802609688, + "grad_norm": 1.7454488277435303, + "learning_rate": 2.00239335568719e-06, + "loss": 3.1498, + "step": 19985 + }, + { + "epoch": 1.6022729550419914, + "grad_norm": 1.7490431070327759, + "learning_rate": 2.001613940693122e-06, + "loss": 3.132, + "step": 19986 + }, + { + "epoch": 1.602353129823014, + "grad_norm": 1.7682489156723022, + "learning_rate": 2.000834660549026e-06, + "loss": 3.1763, + "step": 19987 + }, + { + "epoch": 1.6024333046040367, + "grad_norm": 2.09207820892334, + "learning_rate": 2.0000555152680366e-06, + "loss": 3.1663, + "step": 19988 + }, + { + "epoch": 1.6025134793850593, + "grad_norm": 1.7691797018051147, + "learning_rate": 1.9992765048632955e-06, + "loss": 2.9702, + "step": 19989 + }, + { + "epoch": 1.602593654166082, + "grad_norm": 1.8365614414215088, + "learning_rate": 1.9984976293479296e-06, + "loss": 3.2, + "step": 19990 + }, + { + "epoch": 1.6026738289471045, + "grad_norm": 1.8974674940109253, + "learning_rate": 1.997718888735073e-06, + "loss": 3.224, + "step": 19991 + }, + { + "epoch": 1.6027540037281272, + "grad_norm": 1.6747238636016846, + "learning_rate": 1.996940283037857e-06, + "loss": 3.1198, + "step": 19992 + }, + { + "epoch": 1.60283417850915, + "grad_norm": 1.8528670072555542, + "learning_rate": 1.996161812269407e-06, + "loss": 3.2188, + "step": 19993 + }, + { + "epoch": 1.6029143532901726, + "grad_norm": 1.7309324741363525, + "learning_rate": 1.99538347644285e-06, + "loss": 3.1664, + "step": 19994 + }, + { + "epoch": 1.6029945280711952, + "grad_norm": 1.8053176403045654, + "learning_rate": 1.994605275571304e-06, + "loss": 3.1651, + "step": 19995 + }, + { + "epoch": 1.6030747028522179, + "grad_norm": 1.753572940826416, + "learning_rate": 1.9938272096678925e-06, + "loss": 3.2242, + "step": 19996 + }, + { + "epoch": 1.6031548776332405, + "grad_norm": 1.8179967403411865, + "learning_rate": 1.9930492787457312e-06, + "loss": 3.1073, + "step": 19997 + }, + { + "epoch": 1.6032350524142631, + "grad_norm": 1.8389745950698853, + "learning_rate": 1.992271482817938e-06, + "loss": 3.1996, + "step": 19998 + }, + { + "epoch": 1.6033152271952857, + "grad_norm": 1.817502498626709, + "learning_rate": 1.9914938218976286e-06, + "loss": 3.1469, + "step": 19999 + }, + { + "epoch": 1.6033954019763084, + "grad_norm": 1.8266417980194092, + "learning_rate": 1.9907162959979075e-06, + "loss": 3.1477, + "step": 20000 + }, + { + "epoch": 1.603475576757331, + "grad_norm": 1.8040838241577148, + "learning_rate": 1.989938905131886e-06, + "loss": 3.1207, + "step": 20001 + }, + { + "epoch": 1.6035557515383536, + "grad_norm": 1.8774912357330322, + "learning_rate": 1.989161649312673e-06, + "loss": 3.2579, + "step": 20002 + }, + { + "epoch": 1.6036359263193762, + "grad_norm": 1.8961061239242554, + "learning_rate": 1.988384528553371e-06, + "loss": 3.298, + "step": 20003 + }, + { + "epoch": 1.6037161011003989, + "grad_norm": 1.801941156387329, + "learning_rate": 1.987607542867085e-06, + "loss": 3.1538, + "step": 20004 + }, + { + "epoch": 1.6037962758814215, + "grad_norm": 1.8103886842727661, + "learning_rate": 1.986830692266909e-06, + "loss": 3.1727, + "step": 20005 + }, + { + "epoch": 1.603876450662444, + "grad_norm": 1.7161526679992676, + "learning_rate": 1.9860539767659438e-06, + "loss": 3.0847, + "step": 20006 + }, + { + "epoch": 1.6039566254434667, + "grad_norm": 1.7660295963287354, + "learning_rate": 1.985277396377284e-06, + "loss": 3.2285, + "step": 20007 + }, + { + "epoch": 1.6040368002244894, + "grad_norm": 1.7166162729263306, + "learning_rate": 1.9845009511140225e-06, + "loss": 3.2101, + "step": 20008 + }, + { + "epoch": 1.604116975005512, + "grad_norm": 1.7099559307098389, + "learning_rate": 1.983724640989252e-06, + "loss": 3.1527, + "step": 20009 + }, + { + "epoch": 1.6041971497865346, + "grad_norm": 0.6564337611198425, + "learning_rate": 1.982948466016057e-06, + "loss": 2.481, + "step": 20010 + }, + { + "epoch": 1.6042773245675572, + "grad_norm": 1.8425451517105103, + "learning_rate": 1.982172426207526e-06, + "loss": 3.0942, + "step": 20011 + }, + { + "epoch": 1.6043574993485799, + "grad_norm": 1.8969680070877075, + "learning_rate": 1.9813965215767407e-06, + "loss": 3.1382, + "step": 20012 + }, + { + "epoch": 1.6044376741296027, + "grad_norm": 1.7084367275238037, + "learning_rate": 1.980620752136785e-06, + "loss": 3.1357, + "step": 20013 + }, + { + "epoch": 1.6045178489106253, + "grad_norm": 1.7426738739013672, + "learning_rate": 1.9798451179007392e-06, + "loss": 3.2311, + "step": 20014 + }, + { + "epoch": 1.604598023691648, + "grad_norm": 1.7134099006652832, + "learning_rate": 1.979069618881676e-06, + "loss": 3.0894, + "step": 20015 + }, + { + "epoch": 1.6046781984726706, + "grad_norm": 1.7966187000274658, + "learning_rate": 1.978294255092672e-06, + "loss": 3.1272, + "step": 20016 + }, + { + "epoch": 1.6047583732536932, + "grad_norm": 1.634950876235962, + "learning_rate": 1.9775190265468003e-06, + "loss": 3.0578, + "step": 20017 + }, + { + "epoch": 1.6048385480347158, + "grad_norm": 1.8339616060256958, + "learning_rate": 1.9767439332571292e-06, + "loss": 3.1964, + "step": 20018 + }, + { + "epoch": 1.6049187228157384, + "grad_norm": 1.7559552192687988, + "learning_rate": 1.9759689752367305e-06, + "loss": 3.3077, + "step": 20019 + }, + { + "epoch": 1.604998897596761, + "grad_norm": 1.7264808416366577, + "learning_rate": 1.975194152498665e-06, + "loss": 3.1924, + "step": 20020 + }, + { + "epoch": 1.6050790723777837, + "grad_norm": 0.6514266729354858, + "learning_rate": 1.9744194650559977e-06, + "loss": 2.476, + "step": 20021 + }, + { + "epoch": 1.6051592471588063, + "grad_norm": 1.8892226219177246, + "learning_rate": 1.9736449129217907e-06, + "loss": 3.1899, + "step": 20022 + }, + { + "epoch": 1.605239421939829, + "grad_norm": 1.8826019763946533, + "learning_rate": 1.9728704961091007e-06, + "loss": 3.2364, + "step": 20023 + }, + { + "epoch": 1.6053195967208516, + "grad_norm": 1.7226933240890503, + "learning_rate": 1.9720962146309876e-06, + "loss": 3.201, + "step": 20024 + }, + { + "epoch": 1.6053997715018742, + "grad_norm": 1.6754823923110962, + "learning_rate": 1.9713220685004996e-06, + "loss": 3.1034, + "step": 20025 + }, + { + "epoch": 1.6054799462828968, + "grad_norm": 1.670870065689087, + "learning_rate": 1.9705480577306936e-06, + "loss": 3.0886, + "step": 20026 + }, + { + "epoch": 1.6055601210639194, + "grad_norm": 1.8031553030014038, + "learning_rate": 1.969774182334616e-06, + "loss": 3.1659, + "step": 20027 + }, + { + "epoch": 1.605640295844942, + "grad_norm": 1.831255316734314, + "learning_rate": 1.9690004423253164e-06, + "loss": 3.2103, + "step": 20028 + }, + { + "epoch": 1.6057204706259647, + "grad_norm": 1.776520013809204, + "learning_rate": 1.968226837715842e-06, + "loss": 3.2017, + "step": 20029 + }, + { + "epoch": 1.6058006454069873, + "grad_norm": 0.6502957940101624, + "learning_rate": 1.9674533685192287e-06, + "loss": 2.4657, + "step": 20030 + }, + { + "epoch": 1.60588082018801, + "grad_norm": 1.7875151634216309, + "learning_rate": 1.9666800347485217e-06, + "loss": 3.1937, + "step": 20031 + }, + { + "epoch": 1.6059609949690326, + "grad_norm": 1.7516635656356812, + "learning_rate": 1.965906836416758e-06, + "loss": 3.1287, + "step": 20032 + }, + { + "epoch": 1.6060411697500552, + "grad_norm": 0.6606157422065735, + "learning_rate": 1.9651337735369724e-06, + "loss": 2.4716, + "step": 20033 + }, + { + "epoch": 1.6061213445310778, + "grad_norm": 1.7576746940612793, + "learning_rate": 1.964360846122203e-06, + "loss": 3.1411, + "step": 20034 + }, + { + "epoch": 1.6062015193121004, + "grad_norm": 1.8500028848648071, + "learning_rate": 1.963588054185476e-06, + "loss": 3.1562, + "step": 20035 + }, + { + "epoch": 1.606281694093123, + "grad_norm": 2.015305519104004, + "learning_rate": 1.96281539773982e-06, + "loss": 3.2056, + "step": 20036 + }, + { + "epoch": 1.6063618688741457, + "grad_norm": 1.7718429565429688, + "learning_rate": 1.9620428767982656e-06, + "loss": 3.0528, + "step": 20037 + }, + { + "epoch": 1.6064420436551683, + "grad_norm": 1.6427748203277588, + "learning_rate": 1.9612704913738346e-06, + "loss": 3.1377, + "step": 20038 + }, + { + "epoch": 1.606522218436191, + "grad_norm": 1.8310796022415161, + "learning_rate": 1.9604982414795528e-06, + "loss": 3.1716, + "step": 20039 + }, + { + "epoch": 1.6066023932172135, + "grad_norm": 1.9752243757247925, + "learning_rate": 1.9597261271284353e-06, + "loss": 3.2452, + "step": 20040 + }, + { + "epoch": 1.6066825679982362, + "grad_norm": 1.8011600971221924, + "learning_rate": 1.9589541483335005e-06, + "loss": 3.1296, + "step": 20041 + }, + { + "epoch": 1.6067627427792588, + "grad_norm": 1.7243789434432983, + "learning_rate": 1.9581823051077653e-06, + "loss": 3.036, + "step": 20042 + }, + { + "epoch": 1.6068429175602814, + "grad_norm": 0.6714457273483276, + "learning_rate": 1.957410597464242e-06, + "loss": 2.4789, + "step": 20043 + }, + { + "epoch": 1.606923092341304, + "grad_norm": 1.7269080877304077, + "learning_rate": 1.956639025415945e-06, + "loss": 3.2051, + "step": 20044 + }, + { + "epoch": 1.6070032671223267, + "grad_norm": 1.8796035051345825, + "learning_rate": 1.955867588975876e-06, + "loss": 3.2154, + "step": 20045 + }, + { + "epoch": 1.6070834419033493, + "grad_norm": 1.6887457370758057, + "learning_rate": 1.9550962881570444e-06, + "loss": 3.1339, + "step": 20046 + }, + { + "epoch": 1.607163616684372, + "grad_norm": 1.7004265785217285, + "learning_rate": 1.9543251229724547e-06, + "loss": 3.1406, + "step": 20047 + }, + { + "epoch": 1.6072437914653945, + "grad_norm": 1.7866876125335693, + "learning_rate": 1.9535540934351073e-06, + "loss": 3.1996, + "step": 20048 + }, + { + "epoch": 1.6073239662464172, + "grad_norm": 0.6590977907180786, + "learning_rate": 1.952783199558005e-06, + "loss": 2.4816, + "step": 20049 + }, + { + "epoch": 1.6074041410274398, + "grad_norm": 1.8544156551361084, + "learning_rate": 1.952012441354139e-06, + "loss": 3.2108, + "step": 20050 + }, + { + "epoch": 1.6074843158084624, + "grad_norm": 1.7231818437576294, + "learning_rate": 1.9512418188365068e-06, + "loss": 3.0925, + "step": 20051 + }, + { + "epoch": 1.607564490589485, + "grad_norm": 1.7289319038391113, + "learning_rate": 1.9504713320181002e-06, + "loss": 3.1489, + "step": 20052 + }, + { + "epoch": 1.6076446653705077, + "grad_norm": 1.8078899383544922, + "learning_rate": 1.9497009809119105e-06, + "loss": 3.3241, + "step": 20053 + }, + { + "epoch": 1.6077248401515303, + "grad_norm": 1.8539906740188599, + "learning_rate": 1.9489307655309285e-06, + "loss": 3.1919, + "step": 20054 + }, + { + "epoch": 1.607805014932553, + "grad_norm": 1.7068310976028442, + "learning_rate": 1.948160685888133e-06, + "loss": 3.1478, + "step": 20055 + }, + { + "epoch": 1.6078851897135755, + "grad_norm": 0.6926897764205933, + "learning_rate": 1.947390741996511e-06, + "loss": 2.4603, + "step": 20056 + }, + { + "epoch": 1.6079653644945981, + "grad_norm": 1.8469607830047607, + "learning_rate": 1.9466209338690424e-06, + "loss": 3.1405, + "step": 20057 + }, + { + "epoch": 1.6080455392756208, + "grad_norm": 1.914548635482788, + "learning_rate": 1.9458512615187065e-06, + "loss": 3.1194, + "step": 20058 + }, + { + "epoch": 1.6081257140566434, + "grad_norm": 1.7435731887817383, + "learning_rate": 1.945081724958483e-06, + "loss": 3.2842, + "step": 20059 + }, + { + "epoch": 1.608205888837666, + "grad_norm": 1.7427260875701904, + "learning_rate": 1.9443123242013406e-06, + "loss": 3.1434, + "step": 20060 + }, + { + "epoch": 1.6082860636186886, + "grad_norm": 1.8149186372756958, + "learning_rate": 1.9435430592602523e-06, + "loss": 3.1854, + "step": 20061 + }, + { + "epoch": 1.6083662383997113, + "grad_norm": 1.8447993993759155, + "learning_rate": 1.9427739301481883e-06, + "loss": 3.1805, + "step": 20062 + }, + { + "epoch": 1.608446413180734, + "grad_norm": 1.759732961654663, + "learning_rate": 1.942004936878118e-06, + "loss": 3.1884, + "step": 20063 + }, + { + "epoch": 1.6085265879617565, + "grad_norm": 1.6959989070892334, + "learning_rate": 1.941236079463006e-06, + "loss": 3.0584, + "step": 20064 + }, + { + "epoch": 1.6086067627427791, + "grad_norm": 0.6704512238502502, + "learning_rate": 1.9404673579158107e-06, + "loss": 2.4713, + "step": 20065 + }, + { + "epoch": 1.6086869375238018, + "grad_norm": 1.7247397899627686, + "learning_rate": 1.9396987722494953e-06, + "loss": 3.1897, + "step": 20066 + }, + { + "epoch": 1.6087671123048244, + "grad_norm": 1.6864005327224731, + "learning_rate": 1.9389303224770183e-06, + "loss": 3.1937, + "step": 20067 + }, + { + "epoch": 1.608847287085847, + "grad_norm": 1.740298867225647, + "learning_rate": 1.938162008611335e-06, + "loss": 3.056, + "step": 20068 + }, + { + "epoch": 1.6089274618668696, + "grad_norm": 1.8958203792572021, + "learning_rate": 1.9373938306654017e-06, + "loss": 3.2155, + "step": 20069 + }, + { + "epoch": 1.6090076366478923, + "grad_norm": 1.739476203918457, + "learning_rate": 1.9366257886521643e-06, + "loss": 3.1716, + "step": 20070 + }, + { + "epoch": 1.6090878114289149, + "grad_norm": 1.764927625656128, + "learning_rate": 1.935857882584574e-06, + "loss": 3.2018, + "step": 20071 + }, + { + "epoch": 1.6091679862099375, + "grad_norm": 1.7039551734924316, + "learning_rate": 1.9350901124755773e-06, + "loss": 3.191, + "step": 20072 + }, + { + "epoch": 1.6092481609909601, + "grad_norm": 1.9275199174880981, + "learning_rate": 1.9343224783381197e-06, + "loss": 3.2284, + "step": 20073 + }, + { + "epoch": 1.609328335771983, + "grad_norm": 1.7576618194580078, + "learning_rate": 1.933554980185145e-06, + "loss": 3.1703, + "step": 20074 + }, + { + "epoch": 1.6094085105530056, + "grad_norm": 1.9095810651779175, + "learning_rate": 1.932787618029588e-06, + "loss": 3.335, + "step": 20075 + }, + { + "epoch": 1.6094886853340282, + "grad_norm": 1.7504675388336182, + "learning_rate": 1.9320203918843882e-06, + "loss": 3.2419, + "step": 20076 + }, + { + "epoch": 1.6095688601150508, + "grad_norm": 1.769945502281189, + "learning_rate": 1.9312533017624812e-06, + "loss": 3.0402, + "step": 20077 + }, + { + "epoch": 1.6096490348960735, + "grad_norm": 1.8075275421142578, + "learning_rate": 1.9304863476767998e-06, + "loss": 3.3301, + "step": 20078 + }, + { + "epoch": 1.609729209677096, + "grad_norm": 0.6653004288673401, + "learning_rate": 1.9297195296402747e-06, + "loss": 2.4723, + "step": 20079 + }, + { + "epoch": 1.6098093844581187, + "grad_norm": 1.8038278818130493, + "learning_rate": 1.928952847665837e-06, + "loss": 3.1775, + "step": 20080 + }, + { + "epoch": 1.6098895592391413, + "grad_norm": 1.92758309841156, + "learning_rate": 1.9281863017664058e-06, + "loss": 3.1911, + "step": 20081 + }, + { + "epoch": 1.609969734020164, + "grad_norm": 1.813469409942627, + "learning_rate": 1.9274198919549104e-06, + "loss": 3.106, + "step": 20082 + }, + { + "epoch": 1.6100499088011866, + "grad_norm": 1.6910278797149658, + "learning_rate": 1.9266536182442697e-06, + "loss": 3.1807, + "step": 20083 + }, + { + "epoch": 1.6101300835822092, + "grad_norm": 1.7693922519683838, + "learning_rate": 1.9258874806474038e-06, + "loss": 3.1492, + "step": 20084 + }, + { + "epoch": 1.6102102583632318, + "grad_norm": 0.6530836820602417, + "learning_rate": 1.925121479177232e-06, + "loss": 2.4675, + "step": 20085 + }, + { + "epoch": 1.6102904331442545, + "grad_norm": 1.8281670808792114, + "learning_rate": 1.9243556138466645e-06, + "loss": 3.2233, + "step": 20086 + }, + { + "epoch": 1.610370607925277, + "grad_norm": 1.7939798831939697, + "learning_rate": 1.923589884668614e-06, + "loss": 3.0935, + "step": 20087 + }, + { + "epoch": 1.6104507827062997, + "grad_norm": 1.7098238468170166, + "learning_rate": 1.9228242916559925e-06, + "loss": 3.0635, + "step": 20088 + }, + { + "epoch": 1.6105309574873223, + "grad_norm": 1.8399418592453003, + "learning_rate": 1.9220588348217064e-06, + "loss": 3.2125, + "step": 20089 + }, + { + "epoch": 1.610611132268345, + "grad_norm": 1.8513299226760864, + "learning_rate": 1.921293514178664e-06, + "loss": 3.1561, + "step": 20090 + }, + { + "epoch": 1.6106913070493676, + "grad_norm": 1.6630641222000122, + "learning_rate": 1.9205283297397624e-06, + "loss": 3.0652, + "step": 20091 + }, + { + "epoch": 1.6107714818303902, + "grad_norm": 1.8098968267440796, + "learning_rate": 1.9197632815179068e-06, + "loss": 3.0848, + "step": 20092 + }, + { + "epoch": 1.6108516566114128, + "grad_norm": 1.737831473350525, + "learning_rate": 1.918998369525994e-06, + "loss": 3.1683, + "step": 20093 + }, + { + "epoch": 1.6109318313924357, + "grad_norm": 1.8785338401794434, + "learning_rate": 1.9182335937769214e-06, + "loss": 3.1337, + "step": 20094 + }, + { + "epoch": 1.6110120061734583, + "grad_norm": 1.7152597904205322, + "learning_rate": 1.9174689542835833e-06, + "loss": 3.2789, + "step": 20095 + }, + { + "epoch": 1.611092180954481, + "grad_norm": 1.8020246028900146, + "learning_rate": 1.9167044510588683e-06, + "loss": 3.17, + "step": 20096 + }, + { + "epoch": 1.6111723557355035, + "grad_norm": 1.6870403289794922, + "learning_rate": 1.9159400841156673e-06, + "loss": 3.1896, + "step": 20097 + }, + { + "epoch": 1.6112525305165262, + "grad_norm": 1.813904047012329, + "learning_rate": 1.9151758534668675e-06, + "loss": 3.1198, + "step": 20098 + }, + { + "epoch": 1.6113327052975488, + "grad_norm": 1.812747836112976, + "learning_rate": 1.9144117591253543e-06, + "loss": 3.2279, + "step": 20099 + }, + { + "epoch": 1.6114128800785714, + "grad_norm": 1.6835261583328247, + "learning_rate": 1.9136478011040104e-06, + "loss": 3.0643, + "step": 20100 + }, + { + "epoch": 1.611493054859594, + "grad_norm": 1.799548625946045, + "learning_rate": 1.9128839794157127e-06, + "loss": 3.1202, + "step": 20101 + }, + { + "epoch": 1.6115732296406167, + "grad_norm": 1.8465754985809326, + "learning_rate": 1.9121202940733408e-06, + "loss": 3.0445, + "step": 20102 + }, + { + "epoch": 1.6116534044216393, + "grad_norm": 1.727298378944397, + "learning_rate": 1.911356745089771e-06, + "loss": 3.1833, + "step": 20103 + }, + { + "epoch": 1.611733579202662, + "grad_norm": 1.7105913162231445, + "learning_rate": 1.9105933324778747e-06, + "loss": 3.0374, + "step": 20104 + }, + { + "epoch": 1.6118137539836845, + "grad_norm": 1.758589744567871, + "learning_rate": 1.9098300562505266e-06, + "loss": 3.2028, + "step": 20105 + }, + { + "epoch": 1.6118939287647072, + "grad_norm": 1.930385708808899, + "learning_rate": 1.90906691642059e-06, + "loss": 3.2449, + "step": 20106 + }, + { + "epoch": 1.6119741035457298, + "grad_norm": 1.7562962770462036, + "learning_rate": 1.908303913000934e-06, + "loss": 3.104, + "step": 20107 + }, + { + "epoch": 1.6120542783267524, + "grad_norm": 1.7536959648132324, + "learning_rate": 1.907541046004422e-06, + "loss": 3.1716, + "step": 20108 + }, + { + "epoch": 1.612134453107775, + "grad_norm": 1.8108147382736206, + "learning_rate": 1.9067783154439157e-06, + "loss": 3.1971, + "step": 20109 + }, + { + "epoch": 1.6122146278887977, + "grad_norm": 1.6804499626159668, + "learning_rate": 1.9060157213322773e-06, + "loss": 3.1567, + "step": 20110 + }, + { + "epoch": 1.6122948026698203, + "grad_norm": 1.7708412408828735, + "learning_rate": 1.9052532636823595e-06, + "loss": 3.163, + "step": 20111 + }, + { + "epoch": 1.612374977450843, + "grad_norm": 1.7036844491958618, + "learning_rate": 1.904490942507018e-06, + "loss": 3.2776, + "step": 20112 + }, + { + "epoch": 1.6124551522318655, + "grad_norm": 1.8396117687225342, + "learning_rate": 1.9037287578191055e-06, + "loss": 3.2136, + "step": 20113 + }, + { + "epoch": 1.6125353270128882, + "grad_norm": 1.753164291381836, + "learning_rate": 1.9029667096314742e-06, + "loss": 3.1694, + "step": 20114 + }, + { + "epoch": 1.6126155017939108, + "grad_norm": 1.7980307340621948, + "learning_rate": 1.9022047979569724e-06, + "loss": 3.1521, + "step": 20115 + }, + { + "epoch": 1.6126956765749334, + "grad_norm": 0.658984363079071, + "learning_rate": 1.9014430228084412e-06, + "loss": 2.4734, + "step": 20116 + }, + { + "epoch": 1.612775851355956, + "grad_norm": 1.7596274614334106, + "learning_rate": 1.9006813841987259e-06, + "loss": 3.1267, + "step": 20117 + }, + { + "epoch": 1.6128560261369786, + "grad_norm": 2.4671285152435303, + "learning_rate": 1.8999198821406684e-06, + "loss": 3.163, + "step": 20118 + }, + { + "epoch": 1.6129362009180013, + "grad_norm": 1.8305648565292358, + "learning_rate": 1.8991585166471072e-06, + "loss": 3.225, + "step": 20119 + }, + { + "epoch": 1.613016375699024, + "grad_norm": 1.7139511108398438, + "learning_rate": 1.898397287730881e-06, + "loss": 3.1545, + "step": 20120 + }, + { + "epoch": 1.6130965504800465, + "grad_norm": 1.7535037994384766, + "learning_rate": 1.8976361954048184e-06, + "loss": 3.1381, + "step": 20121 + }, + { + "epoch": 1.6131767252610691, + "grad_norm": 1.8055319786071777, + "learning_rate": 1.8968752396817536e-06, + "loss": 3.2126, + "step": 20122 + }, + { + "epoch": 1.6132569000420918, + "grad_norm": 1.86989426612854, + "learning_rate": 1.8961144205745174e-06, + "loss": 3.306, + "step": 20123 + }, + { + "epoch": 1.6133370748231144, + "grad_norm": 1.8918495178222656, + "learning_rate": 1.8953537380959352e-06, + "loss": 3.2198, + "step": 20124 + }, + { + "epoch": 1.613417249604137, + "grad_norm": 1.7596722841262817, + "learning_rate": 1.8945931922588356e-06, + "loss": 3.0513, + "step": 20125 + }, + { + "epoch": 1.6134974243851596, + "grad_norm": 1.801252841949463, + "learning_rate": 1.893832783076035e-06, + "loss": 3.1813, + "step": 20126 + }, + { + "epoch": 1.6135775991661823, + "grad_norm": 2.091252326965332, + "learning_rate": 1.8930725105603576e-06, + "loss": 3.2015, + "step": 20127 + }, + { + "epoch": 1.6136577739472049, + "grad_norm": 1.7304097414016724, + "learning_rate": 1.8923123747246196e-06, + "loss": 3.219, + "step": 20128 + }, + { + "epoch": 1.6137379487282275, + "grad_norm": 1.7892181873321533, + "learning_rate": 1.8915523755816378e-06, + "loss": 3.2242, + "step": 20129 + }, + { + "epoch": 1.6138181235092501, + "grad_norm": 1.7556253671646118, + "learning_rate": 1.8907925131442283e-06, + "loss": 3.1771, + "step": 20130 + }, + { + "epoch": 1.6138982982902728, + "grad_norm": 1.8589333295822144, + "learning_rate": 1.8900327874251967e-06, + "loss": 3.1873, + "step": 20131 + }, + { + "epoch": 1.6139784730712954, + "grad_norm": 1.6970572471618652, + "learning_rate": 1.8892731984373536e-06, + "loss": 3.1697, + "step": 20132 + }, + { + "epoch": 1.614058647852318, + "grad_norm": 1.7032684087753296, + "learning_rate": 1.8885137461935054e-06, + "loss": 3.1455, + "step": 20133 + }, + { + "epoch": 1.6141388226333406, + "grad_norm": 1.748234510421753, + "learning_rate": 1.8877544307064577e-06, + "loss": 3.1223, + "step": 20134 + }, + { + "epoch": 1.6142189974143633, + "grad_norm": 1.7242796421051025, + "learning_rate": 1.8869952519890134e-06, + "loss": 3.1032, + "step": 20135 + }, + { + "epoch": 1.6142991721953859, + "grad_norm": 1.7276530265808105, + "learning_rate": 1.8862362100539667e-06, + "loss": 3.2012, + "step": 20136 + }, + { + "epoch": 1.6143793469764085, + "grad_norm": 1.6441891193389893, + "learning_rate": 1.8854773049141184e-06, + "loss": 3.192, + "step": 20137 + }, + { + "epoch": 1.6144595217574311, + "grad_norm": 1.791203498840332, + "learning_rate": 1.884718536582263e-06, + "loss": 3.1039, + "step": 20138 + }, + { + "epoch": 1.6145396965384538, + "grad_norm": 1.7605068683624268, + "learning_rate": 1.8839599050711931e-06, + "loss": 3.1639, + "step": 20139 + }, + { + "epoch": 1.6146198713194764, + "grad_norm": 1.743007779121399, + "learning_rate": 1.8832014103937014e-06, + "loss": 3.1394, + "step": 20140 + }, + { + "epoch": 1.614700046100499, + "grad_norm": 1.7360254526138306, + "learning_rate": 1.8824430525625703e-06, + "loss": 3.1313, + "step": 20141 + }, + { + "epoch": 1.6147802208815216, + "grad_norm": 1.8025281429290771, + "learning_rate": 1.881684831590589e-06, + "loss": 3.1692, + "step": 20142 + }, + { + "epoch": 1.6148603956625442, + "grad_norm": 1.8496166467666626, + "learning_rate": 1.88092674749054e-06, + "loss": 3.1962, + "step": 20143 + }, + { + "epoch": 1.6149405704435669, + "grad_norm": 1.8623334169387817, + "learning_rate": 1.880168800275205e-06, + "loss": 3.167, + "step": 20144 + }, + { + "epoch": 1.6150207452245895, + "grad_norm": 1.757706880569458, + "learning_rate": 1.8794109899573642e-06, + "loss": 3.1757, + "step": 20145 + }, + { + "epoch": 1.6151009200056121, + "grad_norm": 1.7469608783721924, + "learning_rate": 1.87865331654979e-06, + "loss": 3.1702, + "step": 20146 + }, + { + "epoch": 1.6151810947866347, + "grad_norm": 1.7852669954299927, + "learning_rate": 1.8778957800652575e-06, + "loss": 3.3158, + "step": 20147 + }, + { + "epoch": 1.6152612695676574, + "grad_norm": 1.815637230873108, + "learning_rate": 1.8771383805165411e-06, + "loss": 3.1121, + "step": 20148 + }, + { + "epoch": 1.61534144434868, + "grad_norm": 1.745992660522461, + "learning_rate": 1.8763811179164083e-06, + "loss": 3.1443, + "step": 20149 + }, + { + "epoch": 1.6154216191297026, + "grad_norm": 1.734235167503357, + "learning_rate": 1.8756239922776298e-06, + "loss": 3.1578, + "step": 20150 + }, + { + "epoch": 1.6155017939107252, + "grad_norm": 0.6419763565063477, + "learning_rate": 1.8748670036129646e-06, + "loss": 2.4437, + "step": 20151 + }, + { + "epoch": 1.6155819686917479, + "grad_norm": 1.8623733520507812, + "learning_rate": 1.8741101519351778e-06, + "loss": 3.2556, + "step": 20152 + }, + { + "epoch": 1.6156621434727705, + "grad_norm": 1.7686301469802856, + "learning_rate": 1.8733534372570295e-06, + "loss": 3.2213, + "step": 20153 + }, + { + "epoch": 1.615742318253793, + "grad_norm": 1.7811455726623535, + "learning_rate": 1.8725968595912792e-06, + "loss": 3.1665, + "step": 20154 + }, + { + "epoch": 1.6158224930348157, + "grad_norm": 1.7416095733642578, + "learning_rate": 1.8718404189506834e-06, + "loss": 3.1858, + "step": 20155 + }, + { + "epoch": 1.6159026678158386, + "grad_norm": 1.8957315683364868, + "learning_rate": 1.8710841153479919e-06, + "loss": 3.1958, + "step": 20156 + }, + { + "epoch": 1.6159828425968612, + "grad_norm": 0.6581953167915344, + "learning_rate": 1.870327948795957e-06, + "loss": 2.4664, + "step": 20157 + }, + { + "epoch": 1.6160630173778838, + "grad_norm": 1.7441378831863403, + "learning_rate": 1.869571919307327e-06, + "loss": 3.1755, + "step": 20158 + }, + { + "epoch": 1.6161431921589064, + "grad_norm": 1.7162784337997437, + "learning_rate": 1.8688160268948496e-06, + "loss": 3.1912, + "step": 20159 + }, + { + "epoch": 1.616223366939929, + "grad_norm": 1.8271729946136475, + "learning_rate": 1.8680602715712703e-06, + "loss": 3.1143, + "step": 20160 + }, + { + "epoch": 1.6163035417209517, + "grad_norm": 0.6627394556999207, + "learning_rate": 1.8673046533493267e-06, + "loss": 2.4824, + "step": 20161 + }, + { + "epoch": 1.6163837165019743, + "grad_norm": 1.783199429512024, + "learning_rate": 1.8665491722417607e-06, + "loss": 3.1932, + "step": 20162 + }, + { + "epoch": 1.616463891282997, + "grad_norm": 1.6440778970718384, + "learning_rate": 1.8657938282613087e-06, + "loss": 3.1494, + "step": 20163 + }, + { + "epoch": 1.6165440660640196, + "grad_norm": 1.7574882507324219, + "learning_rate": 1.8650386214207051e-06, + "loss": 3.2959, + "step": 20164 + }, + { + "epoch": 1.6166242408450422, + "grad_norm": 1.8113914728164673, + "learning_rate": 1.8642835517326873e-06, + "loss": 3.2253, + "step": 20165 + }, + { + "epoch": 1.6167044156260648, + "grad_norm": 1.7783451080322266, + "learning_rate": 1.8635286192099788e-06, + "loss": 3.1691, + "step": 20166 + }, + { + "epoch": 1.6167845904070874, + "grad_norm": 1.7752783298492432, + "learning_rate": 1.8627738238653092e-06, + "loss": 3.1807, + "step": 20167 + }, + { + "epoch": 1.61686476518811, + "grad_norm": 0.6419260501861572, + "learning_rate": 1.8620191657114061e-06, + "loss": 2.4672, + "step": 20168 + }, + { + "epoch": 1.6169449399691327, + "grad_norm": 1.7400561571121216, + "learning_rate": 1.8612646447609917e-06, + "loss": 3.1503, + "step": 20169 + }, + { + "epoch": 1.6170251147501553, + "grad_norm": 1.7808732986450195, + "learning_rate": 1.860510261026789e-06, + "loss": 3.1492, + "step": 20170 + }, + { + "epoch": 1.617105289531178, + "grad_norm": 1.7158803939819336, + "learning_rate": 1.8597560145215121e-06, + "loss": 3.2094, + "step": 20171 + }, + { + "epoch": 1.6171854643122006, + "grad_norm": 1.7594550848007202, + "learning_rate": 1.85900190525788e-06, + "loss": 3.1353, + "step": 20172 + }, + { + "epoch": 1.6172656390932232, + "grad_norm": 1.7575438022613525, + "learning_rate": 1.858247933248607e-06, + "loss": 3.2683, + "step": 20173 + }, + { + "epoch": 1.6173458138742458, + "grad_norm": 1.721474289894104, + "learning_rate": 1.857494098506405e-06, + "loss": 3.1257, + "step": 20174 + }, + { + "epoch": 1.6174259886552684, + "grad_norm": 1.7453652620315552, + "learning_rate": 1.8567404010439837e-06, + "loss": 3.1617, + "step": 20175 + }, + { + "epoch": 1.6175061634362913, + "grad_norm": 1.7609783411026, + "learning_rate": 1.8559868408740478e-06, + "loss": 3.1351, + "step": 20176 + }, + { + "epoch": 1.617586338217314, + "grad_norm": 1.75919771194458, + "learning_rate": 1.8552334180093035e-06, + "loss": 3.1553, + "step": 20177 + }, + { + "epoch": 1.6176665129983365, + "grad_norm": 1.7640737295150757, + "learning_rate": 1.8544801324624528e-06, + "loss": 3.128, + "step": 20178 + }, + { + "epoch": 1.6177466877793591, + "grad_norm": 1.7881345748901367, + "learning_rate": 1.8537269842461969e-06, + "loss": 3.0867, + "step": 20179 + }, + { + "epoch": 1.6178268625603818, + "grad_norm": 1.9035518169403076, + "learning_rate": 1.8529739733732356e-06, + "loss": 3.1713, + "step": 20180 + }, + { + "epoch": 1.6179070373414044, + "grad_norm": 0.6357977390289307, + "learning_rate": 1.8522210998562584e-06, + "loss": 2.4591, + "step": 20181 + }, + { + "epoch": 1.617987212122427, + "grad_norm": 1.8033391237258911, + "learning_rate": 1.8514683637079634e-06, + "loss": 3.0668, + "step": 20182 + }, + { + "epoch": 1.6180673869034496, + "grad_norm": 1.7770394086837769, + "learning_rate": 1.850715764941039e-06, + "loss": 3.1575, + "step": 20183 + }, + { + "epoch": 1.6181475616844723, + "grad_norm": 1.8138030767440796, + "learning_rate": 1.849963303568174e-06, + "loss": 3.1635, + "step": 20184 + }, + { + "epoch": 1.618227736465495, + "grad_norm": 1.7434841394424438, + "learning_rate": 1.8492109796020584e-06, + "loss": 3.2377, + "step": 20185 + }, + { + "epoch": 1.6183079112465175, + "grad_norm": 1.8637638092041016, + "learning_rate": 1.848458793055371e-06, + "loss": 3.0966, + "step": 20186 + }, + { + "epoch": 1.6183880860275401, + "grad_norm": 1.7281429767608643, + "learning_rate": 1.8477067439407958e-06, + "loss": 3.0764, + "step": 20187 + }, + { + "epoch": 1.6184682608085628, + "grad_norm": 0.6361684799194336, + "learning_rate": 1.8469548322710106e-06, + "loss": 2.4858, + "step": 20188 + }, + { + "epoch": 1.6185484355895854, + "grad_norm": 1.7477998733520508, + "learning_rate": 1.8462030580586942e-06, + "loss": 3.2229, + "step": 20189 + }, + { + "epoch": 1.618628610370608, + "grad_norm": 1.7839477062225342, + "learning_rate": 1.845451421316523e-06, + "loss": 3.2076, + "step": 20190 + }, + { + "epoch": 1.6187087851516306, + "grad_norm": 1.809026837348938, + "learning_rate": 1.8446999220571648e-06, + "loss": 3.2223, + "step": 20191 + }, + { + "epoch": 1.6187889599326533, + "grad_norm": 1.7444254159927368, + "learning_rate": 1.8439485602932905e-06, + "loss": 3.1831, + "step": 20192 + }, + { + "epoch": 1.6188691347136759, + "grad_norm": 1.7514177560806274, + "learning_rate": 1.84319733603757e-06, + "loss": 3.1453, + "step": 20193 + }, + { + "epoch": 1.6189493094946985, + "grad_norm": 1.6126121282577515, + "learning_rate": 1.8424462493026674e-06, + "loss": 3.0337, + "step": 20194 + }, + { + "epoch": 1.6190294842757211, + "grad_norm": 1.7472364902496338, + "learning_rate": 1.8416953001012494e-06, + "loss": 3.0355, + "step": 20195 + }, + { + "epoch": 1.6191096590567438, + "grad_norm": 1.7395952939987183, + "learning_rate": 1.8409444884459692e-06, + "loss": 3.1001, + "step": 20196 + }, + { + "epoch": 1.6191898338377664, + "grad_norm": 1.7671481370925903, + "learning_rate": 1.8401938143494902e-06, + "loss": 3.1671, + "step": 20197 + }, + { + "epoch": 1.619270008618789, + "grad_norm": 1.7882914543151855, + "learning_rate": 1.8394432778244687e-06, + "loss": 3.181, + "step": 20198 + }, + { + "epoch": 1.6193501833998116, + "grad_norm": 1.8757588863372803, + "learning_rate": 1.8386928788835556e-06, + "loss": 3.0881, + "step": 20199 + }, + { + "epoch": 1.6194303581808343, + "grad_norm": 1.7264357805252075, + "learning_rate": 1.8379426175394076e-06, + "loss": 3.1744, + "step": 20200 + }, + { + "epoch": 1.6195105329618569, + "grad_norm": 1.7975869178771973, + "learning_rate": 1.8371924938046681e-06, + "loss": 3.1907, + "step": 20201 + }, + { + "epoch": 1.6195907077428795, + "grad_norm": 1.7187445163726807, + "learning_rate": 1.8364425076919868e-06, + "loss": 3.1959, + "step": 20202 + }, + { + "epoch": 1.6196708825239021, + "grad_norm": 1.7617992162704468, + "learning_rate": 1.8356926592140067e-06, + "loss": 3.1808, + "step": 20203 + }, + { + "epoch": 1.6197510573049247, + "grad_norm": 1.7820764780044556, + "learning_rate": 1.8349429483833713e-06, + "loss": 3.2036, + "step": 20204 + }, + { + "epoch": 1.6198312320859474, + "grad_norm": 1.7821550369262695, + "learning_rate": 1.8341933752127227e-06, + "loss": 3.2388, + "step": 20205 + }, + { + "epoch": 1.61991140686697, + "grad_norm": 1.760383129119873, + "learning_rate": 1.8334439397146941e-06, + "loss": 3.2517, + "step": 20206 + }, + { + "epoch": 1.6199915816479926, + "grad_norm": 1.8836979866027832, + "learning_rate": 1.832694641901922e-06, + "loss": 3.1904, + "step": 20207 + }, + { + "epoch": 1.6200717564290152, + "grad_norm": 1.7935127019882202, + "learning_rate": 1.8319454817870396e-06, + "loss": 3.1603, + "step": 20208 + }, + { + "epoch": 1.6201519312100379, + "grad_norm": 1.827683448791504, + "learning_rate": 1.8311964593826782e-06, + "loss": 3.1724, + "step": 20209 + }, + { + "epoch": 1.6202321059910605, + "grad_norm": 1.7586216926574707, + "learning_rate": 1.8304475747014672e-06, + "loss": 3.135, + "step": 20210 + }, + { + "epoch": 1.6203122807720831, + "grad_norm": 1.745787501335144, + "learning_rate": 1.8296988277560291e-06, + "loss": 3.3049, + "step": 20211 + }, + { + "epoch": 1.6203924555531057, + "grad_norm": 0.6970466375350952, + "learning_rate": 1.828950218558988e-06, + "loss": 2.4629, + "step": 20212 + }, + { + "epoch": 1.6204726303341284, + "grad_norm": 1.783798098564148, + "learning_rate": 1.8282017471229674e-06, + "loss": 3.1834, + "step": 20213 + }, + { + "epoch": 1.620552805115151, + "grad_norm": 1.9213773012161255, + "learning_rate": 1.827453413460586e-06, + "loss": 3.2753, + "step": 20214 + }, + { + "epoch": 1.6206329798961736, + "grad_norm": 1.7830874919891357, + "learning_rate": 1.8267052175844612e-06, + "loss": 3.1968, + "step": 20215 + }, + { + "epoch": 1.6207131546771962, + "grad_norm": 1.8849725723266602, + "learning_rate": 1.8259571595072034e-06, + "loss": 3.1444, + "step": 20216 + }, + { + "epoch": 1.6207933294582189, + "grad_norm": 1.8561338186264038, + "learning_rate": 1.8252092392414278e-06, + "loss": 3.1293, + "step": 20217 + }, + { + "epoch": 1.6208735042392415, + "grad_norm": 1.9329618215560913, + "learning_rate": 1.824461456799743e-06, + "loss": 3.2036, + "step": 20218 + }, + { + "epoch": 1.620953679020264, + "grad_norm": 1.8521101474761963, + "learning_rate": 1.823713812194756e-06, + "loss": 3.1573, + "step": 20219 + }, + { + "epoch": 1.6210338538012867, + "grad_norm": 1.8687307834625244, + "learning_rate": 1.8229663054390767e-06, + "loss": 3.1312, + "step": 20220 + }, + { + "epoch": 1.6211140285823094, + "grad_norm": 1.847031593322754, + "learning_rate": 1.8222189365453003e-06, + "loss": 3.1763, + "step": 20221 + }, + { + "epoch": 1.621194203363332, + "grad_norm": 1.942349910736084, + "learning_rate": 1.8214717055260301e-06, + "loss": 3.1791, + "step": 20222 + }, + { + "epoch": 1.6212743781443546, + "grad_norm": 1.9972610473632812, + "learning_rate": 1.820724612393865e-06, + "loss": 3.2764, + "step": 20223 + }, + { + "epoch": 1.6213545529253772, + "grad_norm": 0.6560233235359192, + "learning_rate": 1.8199776571613993e-06, + "loss": 2.4738, + "step": 20224 + }, + { + "epoch": 1.6214347277063998, + "grad_norm": 1.8873077630996704, + "learning_rate": 1.8192308398412306e-06, + "loss": 3.1237, + "step": 20225 + }, + { + "epoch": 1.6215149024874225, + "grad_norm": 1.7883189916610718, + "learning_rate": 1.8184841604459446e-06, + "loss": 3.2673, + "step": 20226 + }, + { + "epoch": 1.621595077268445, + "grad_norm": 1.7627640962600708, + "learning_rate": 1.8177376189881313e-06, + "loss": 3.1407, + "step": 20227 + }, + { + "epoch": 1.6216752520494677, + "grad_norm": 1.7129958868026733, + "learning_rate": 1.8169912154803792e-06, + "loss": 3.1946, + "step": 20228 + }, + { + "epoch": 1.6217554268304903, + "grad_norm": 1.7217668294906616, + "learning_rate": 1.8162449499352709e-06, + "loss": 3.1261, + "step": 20229 + }, + { + "epoch": 1.621835601611513, + "grad_norm": 1.9178431034088135, + "learning_rate": 1.8154988223653903e-06, + "loss": 3.2971, + "step": 20230 + }, + { + "epoch": 1.6219157763925356, + "grad_norm": 1.6804004907608032, + "learning_rate": 1.814752832783313e-06, + "loss": 3.0227, + "step": 20231 + }, + { + "epoch": 1.6219959511735582, + "grad_norm": 1.7252289056777954, + "learning_rate": 1.8140069812016181e-06, + "loss": 3.1049, + "step": 20232 + }, + { + "epoch": 1.6220761259545808, + "grad_norm": 1.737304449081421, + "learning_rate": 1.8132612676328809e-06, + "loss": 3.1274, + "step": 20233 + }, + { + "epoch": 1.6221563007356035, + "grad_norm": 1.7736760377883911, + "learning_rate": 1.8125156920896724e-06, + "loss": 3.22, + "step": 20234 + }, + { + "epoch": 1.622236475516626, + "grad_norm": 1.9627269506454468, + "learning_rate": 1.8117702545845671e-06, + "loss": 3.3145, + "step": 20235 + }, + { + "epoch": 1.6223166502976487, + "grad_norm": 1.7710169553756714, + "learning_rate": 1.8110249551301272e-06, + "loss": 3.2131, + "step": 20236 + }, + { + "epoch": 1.6223968250786716, + "grad_norm": 1.8325235843658447, + "learning_rate": 1.8102797937389205e-06, + "loss": 3.1669, + "step": 20237 + }, + { + "epoch": 1.6224769998596942, + "grad_norm": 1.8124810457229614, + "learning_rate": 1.80953477042351e-06, + "loss": 3.248, + "step": 20238 + }, + { + "epoch": 1.6225571746407168, + "grad_norm": 1.7377225160598755, + "learning_rate": 1.8087898851964557e-06, + "loss": 3.1637, + "step": 20239 + }, + { + "epoch": 1.6226373494217394, + "grad_norm": 1.689452886581421, + "learning_rate": 1.8080451380703212e-06, + "loss": 3.1041, + "step": 20240 + }, + { + "epoch": 1.622717524202762, + "grad_norm": 1.777884602546692, + "learning_rate": 1.8073005290576552e-06, + "loss": 3.2452, + "step": 20241 + }, + { + "epoch": 1.6227976989837847, + "grad_norm": 1.7363232374191284, + "learning_rate": 1.8065560581710152e-06, + "loss": 3.1328, + "step": 20242 + }, + { + "epoch": 1.6228778737648073, + "grad_norm": 1.7706674337387085, + "learning_rate": 1.8058117254229513e-06, + "loss": 3.1853, + "step": 20243 + }, + { + "epoch": 1.62295804854583, + "grad_norm": 1.7469345331192017, + "learning_rate": 1.8050675308260145e-06, + "loss": 3.1591, + "step": 20244 + }, + { + "epoch": 1.6230382233268525, + "grad_norm": 1.9073925018310547, + "learning_rate": 1.804323474392754e-06, + "loss": 3.2129, + "step": 20245 + }, + { + "epoch": 1.6231183981078752, + "grad_norm": 0.6515526175498962, + "learning_rate": 1.8035795561357084e-06, + "loss": 2.4626, + "step": 20246 + }, + { + "epoch": 1.6231985728888978, + "grad_norm": 1.676491141319275, + "learning_rate": 1.802835776067423e-06, + "loss": 3.1985, + "step": 20247 + }, + { + "epoch": 1.6232787476699204, + "grad_norm": 1.762820839881897, + "learning_rate": 1.8020921342004371e-06, + "loss": 3.2159, + "step": 20248 + }, + { + "epoch": 1.623358922450943, + "grad_norm": 1.8015046119689941, + "learning_rate": 1.8013486305472883e-06, + "loss": 3.1309, + "step": 20249 + }, + { + "epoch": 1.6234390972319657, + "grad_norm": 1.838219404220581, + "learning_rate": 1.8006052651205153e-06, + "loss": 3.2007, + "step": 20250 + }, + { + "epoch": 1.6235192720129883, + "grad_norm": 1.8260396718978882, + "learning_rate": 1.7998620379326448e-06, + "loss": 3.0923, + "step": 20251 + }, + { + "epoch": 1.623599446794011, + "grad_norm": 1.7371885776519775, + "learning_rate": 1.7991189489962102e-06, + "loss": 3.0771, + "step": 20252 + }, + { + "epoch": 1.6236796215750335, + "grad_norm": 1.709968090057373, + "learning_rate": 1.7983759983237404e-06, + "loss": 3.2283, + "step": 20253 + }, + { + "epoch": 1.6237597963560562, + "grad_norm": 1.7386211156845093, + "learning_rate": 1.7976331859277606e-06, + "loss": 3.132, + "step": 20254 + }, + { + "epoch": 1.6238399711370788, + "grad_norm": 1.8251805305480957, + "learning_rate": 1.7968905118207968e-06, + "loss": 3.0722, + "step": 20255 + }, + { + "epoch": 1.6239201459181014, + "grad_norm": 1.6529957056045532, + "learning_rate": 1.7961479760153654e-06, + "loss": 3.1215, + "step": 20256 + }, + { + "epoch": 1.6240003206991243, + "grad_norm": 1.8087224960327148, + "learning_rate": 1.7954055785239876e-06, + "loss": 3.1219, + "step": 20257 + }, + { + "epoch": 1.6240804954801469, + "grad_norm": 1.7692461013793945, + "learning_rate": 1.7946633193591811e-06, + "loss": 3.1079, + "step": 20258 + }, + { + "epoch": 1.6241606702611695, + "grad_norm": 0.6572635173797607, + "learning_rate": 1.793921198533458e-06, + "loss": 2.475, + "step": 20259 + }, + { + "epoch": 1.6242408450421921, + "grad_norm": 1.7486051321029663, + "learning_rate": 1.7931792160593353e-06, + "loss": 3.075, + "step": 20260 + }, + { + "epoch": 1.6243210198232148, + "grad_norm": 1.7866039276123047, + "learning_rate": 1.7924373719493148e-06, + "loss": 3.1243, + "step": 20261 + }, + { + "epoch": 1.6244011946042374, + "grad_norm": 1.8308039903640747, + "learning_rate": 1.791695666215908e-06, + "loss": 3.0636, + "step": 20262 + }, + { + "epoch": 1.62448136938526, + "grad_norm": 1.807878017425537, + "learning_rate": 1.790954098871619e-06, + "loss": 3.1461, + "step": 20263 + }, + { + "epoch": 1.6245615441662826, + "grad_norm": 1.772731065750122, + "learning_rate": 1.7902126699289513e-06, + "loss": 3.1508, + "step": 20264 + }, + { + "epoch": 1.6246417189473052, + "grad_norm": 1.8681150674819946, + "learning_rate": 1.7894713794004071e-06, + "loss": 3.1398, + "step": 20265 + }, + { + "epoch": 1.6247218937283279, + "grad_norm": 1.879340410232544, + "learning_rate": 1.7887302272984796e-06, + "loss": 3.1302, + "step": 20266 + }, + { + "epoch": 1.6248020685093505, + "grad_norm": 1.7627182006835938, + "learning_rate": 1.7879892136356658e-06, + "loss": 3.1601, + "step": 20267 + }, + { + "epoch": 1.6248822432903731, + "grad_norm": 1.742895245552063, + "learning_rate": 1.7872483384244599e-06, + "loss": 3.2292, + "step": 20268 + }, + { + "epoch": 1.6249624180713957, + "grad_norm": 1.7477835416793823, + "learning_rate": 1.786507601677353e-06, + "loss": 3.2318, + "step": 20269 + }, + { + "epoch": 1.6250425928524184, + "grad_norm": 1.7514255046844482, + "learning_rate": 1.7857670034068352e-06, + "loss": 3.1472, + "step": 20270 + }, + { + "epoch": 1.625122767633441, + "grad_norm": 1.7211252450942993, + "learning_rate": 1.7850265436253889e-06, + "loss": 3.151, + "step": 20271 + }, + { + "epoch": 1.6252029424144636, + "grad_norm": 1.8858073949813843, + "learning_rate": 1.7842862223454992e-06, + "loss": 3.1873, + "step": 20272 + }, + { + "epoch": 1.6252831171954862, + "grad_norm": 1.8761296272277832, + "learning_rate": 1.7835460395796477e-06, + "loss": 3.1984, + "step": 20273 + }, + { + "epoch": 1.6253632919765089, + "grad_norm": 1.8608686923980713, + "learning_rate": 1.7828059953403153e-06, + "loss": 3.1611, + "step": 20274 + }, + { + "epoch": 1.6254434667575315, + "grad_norm": 1.7517518997192383, + "learning_rate": 1.78206608963998e-06, + "loss": 3.1416, + "step": 20275 + }, + { + "epoch": 1.625523641538554, + "grad_norm": 1.8172374963760376, + "learning_rate": 1.7813263224911116e-06, + "loss": 3.0621, + "step": 20276 + }, + { + "epoch": 1.6256038163195767, + "grad_norm": 1.7091976404190063, + "learning_rate": 1.7805866939061845e-06, + "loss": 3.0306, + "step": 20277 + }, + { + "epoch": 1.6256839911005994, + "grad_norm": 1.7291100025177002, + "learning_rate": 1.77984720389767e-06, + "loss": 3.1641, + "step": 20278 + }, + { + "epoch": 1.625764165881622, + "grad_norm": 0.6555626392364502, + "learning_rate": 1.7791078524780336e-06, + "loss": 2.4786, + "step": 20279 + }, + { + "epoch": 1.6258443406626446, + "grad_norm": 1.813132405281067, + "learning_rate": 1.7783686396597445e-06, + "loss": 3.1913, + "step": 20280 + }, + { + "epoch": 1.6259245154436672, + "grad_norm": 1.761917233467102, + "learning_rate": 1.777629565455259e-06, + "loss": 3.0815, + "step": 20281 + }, + { + "epoch": 1.6260046902246899, + "grad_norm": 1.7229894399642944, + "learning_rate": 1.7768906298770428e-06, + "loss": 3.0322, + "step": 20282 + }, + { + "epoch": 1.6260848650057125, + "grad_norm": 1.683936595916748, + "learning_rate": 1.7761518329375516e-06, + "loss": 3.1982, + "step": 20283 + }, + { + "epoch": 1.626165039786735, + "grad_norm": 1.7087844610214233, + "learning_rate": 1.775413174649242e-06, + "loss": 3.2509, + "step": 20284 + }, + { + "epoch": 1.6262452145677577, + "grad_norm": 1.794538974761963, + "learning_rate": 1.7746746550245697e-06, + "loss": 3.229, + "step": 20285 + }, + { + "epoch": 1.6263253893487803, + "grad_norm": 0.6631638407707214, + "learning_rate": 1.7739362740759824e-06, + "loss": 2.4843, + "step": 20286 + }, + { + "epoch": 1.626405564129803, + "grad_norm": 1.8169777393341064, + "learning_rate": 1.7731980318159303e-06, + "loss": 3.1958, + "step": 20287 + }, + { + "epoch": 1.6264857389108256, + "grad_norm": 1.813495397567749, + "learning_rate": 1.7724599282568588e-06, + "loss": 3.117, + "step": 20288 + }, + { + "epoch": 1.6265659136918482, + "grad_norm": 0.6656306385993958, + "learning_rate": 1.7717219634112149e-06, + "loss": 2.4644, + "step": 20289 + }, + { + "epoch": 1.6266460884728708, + "grad_norm": 1.890923023223877, + "learning_rate": 1.7709841372914393e-06, + "loss": 3.1576, + "step": 20290 + }, + { + "epoch": 1.6267262632538935, + "grad_norm": 1.7686737775802612, + "learning_rate": 1.7702464499099703e-06, + "loss": 3.1413, + "step": 20291 + }, + { + "epoch": 1.626806438034916, + "grad_norm": 1.8243199586868286, + "learning_rate": 1.7695089012792443e-06, + "loss": 3.1552, + "step": 20292 + }, + { + "epoch": 1.6268866128159387, + "grad_norm": 1.8366236686706543, + "learning_rate": 1.7687714914116982e-06, + "loss": 3.2163, + "step": 20293 + }, + { + "epoch": 1.6269667875969613, + "grad_norm": 1.793362021446228, + "learning_rate": 1.7680342203197641e-06, + "loss": 3.0804, + "step": 20294 + }, + { + "epoch": 1.627046962377984, + "grad_norm": 1.78350031375885, + "learning_rate": 1.7672970880158747e-06, + "loss": 3.2017, + "step": 20295 + }, + { + "epoch": 1.6271271371590066, + "grad_norm": 1.7692996263504028, + "learning_rate": 1.7665600945124518e-06, + "loss": 3.1279, + "step": 20296 + }, + { + "epoch": 1.6272073119400292, + "grad_norm": 0.6531975269317627, + "learning_rate": 1.7658232398219233e-06, + "loss": 2.4728, + "step": 20297 + }, + { + "epoch": 1.6272874867210518, + "grad_norm": 1.7954466342926025, + "learning_rate": 1.765086523956714e-06, + "loss": 3.1598, + "step": 20298 + }, + { + "epoch": 1.6273676615020745, + "grad_norm": 1.7790377140045166, + "learning_rate": 1.7643499469292425e-06, + "loss": 3.2043, + "step": 20299 + }, + { + "epoch": 1.627447836283097, + "grad_norm": 1.825875163078308, + "learning_rate": 1.7636135087519323e-06, + "loss": 3.2054, + "step": 20300 + }, + { + "epoch": 1.6275280110641197, + "grad_norm": 1.8731831312179565, + "learning_rate": 1.7628772094371915e-06, + "loss": 3.1639, + "step": 20301 + }, + { + "epoch": 1.6276081858451423, + "grad_norm": 1.857702612876892, + "learning_rate": 1.762141048997439e-06, + "loss": 3.182, + "step": 20302 + }, + { + "epoch": 1.627688360626165, + "grad_norm": 1.7537046670913696, + "learning_rate": 1.7614050274450844e-06, + "loss": 3.1946, + "step": 20303 + }, + { + "epoch": 1.6277685354071876, + "grad_norm": 1.8468047380447388, + "learning_rate": 1.7606691447925385e-06, + "loss": 3.2375, + "step": 20304 + }, + { + "epoch": 1.6278487101882102, + "grad_norm": 0.6500931978225708, + "learning_rate": 1.7599334010522084e-06, + "loss": 2.4513, + "step": 20305 + }, + { + "epoch": 1.6279288849692328, + "grad_norm": 1.8829094171524048, + "learning_rate": 1.759197796236496e-06, + "loss": 3.2687, + "step": 20306 + }, + { + "epoch": 1.6280090597502554, + "grad_norm": 1.7122817039489746, + "learning_rate": 1.758462330357803e-06, + "loss": 3.1853, + "step": 20307 + }, + { + "epoch": 1.628089234531278, + "grad_norm": 0.6463294625282288, + "learning_rate": 1.757727003428531e-06, + "loss": 2.4634, + "step": 20308 + }, + { + "epoch": 1.6281694093123007, + "grad_norm": 1.7611538171768188, + "learning_rate": 1.7569918154610766e-06, + "loss": 3.1479, + "step": 20309 + }, + { + "epoch": 1.6282495840933233, + "grad_norm": 1.7437551021575928, + "learning_rate": 1.7562567664678386e-06, + "loss": 3.1908, + "step": 20310 + }, + { + "epoch": 1.628329758874346, + "grad_norm": 1.7370573282241821, + "learning_rate": 1.7555218564612031e-06, + "loss": 3.1622, + "step": 20311 + }, + { + "epoch": 1.6284099336553686, + "grad_norm": 1.6818726062774658, + "learning_rate": 1.754787085453563e-06, + "loss": 3.2013, + "step": 20312 + }, + { + "epoch": 1.6284901084363912, + "grad_norm": 1.7425535917282104, + "learning_rate": 1.7540524534573066e-06, + "loss": 3.1025, + "step": 20313 + }, + { + "epoch": 1.6285702832174138, + "grad_norm": 1.8018032312393188, + "learning_rate": 1.7533179604848204e-06, + "loss": 3.2892, + "step": 20314 + }, + { + "epoch": 1.6286504579984364, + "grad_norm": 1.7350280284881592, + "learning_rate": 1.7525836065484892e-06, + "loss": 3.0754, + "step": 20315 + }, + { + "epoch": 1.628730632779459, + "grad_norm": 1.8488956689834595, + "learning_rate": 1.7518493916606893e-06, + "loss": 3.218, + "step": 20316 + }, + { + "epoch": 1.6288108075604817, + "grad_norm": 1.78863525390625, + "learning_rate": 1.7511153158338013e-06, + "loss": 3.2, + "step": 20317 + }, + { + "epoch": 1.6288909823415045, + "grad_norm": 1.7561408281326294, + "learning_rate": 1.7503813790802027e-06, + "loss": 3.1852, + "step": 20318 + }, + { + "epoch": 1.6289711571225272, + "grad_norm": 1.7343946695327759, + "learning_rate": 1.749647581412266e-06, + "loss": 3.058, + "step": 20319 + }, + { + "epoch": 1.6290513319035498, + "grad_norm": 1.7920911312103271, + "learning_rate": 1.7489139228423658e-06, + "loss": 3.2487, + "step": 20320 + }, + { + "epoch": 1.6291315066845724, + "grad_norm": 1.6510893106460571, + "learning_rate": 1.7481804033828665e-06, + "loss": 3.1011, + "step": 20321 + }, + { + "epoch": 1.629211681465595, + "grad_norm": 1.8478775024414062, + "learning_rate": 1.747447023046137e-06, + "loss": 3.2146, + "step": 20322 + }, + { + "epoch": 1.6292918562466177, + "grad_norm": 1.8167638778686523, + "learning_rate": 1.7467137818445434e-06, + "loss": 3.205, + "step": 20323 + }, + { + "epoch": 1.6293720310276403, + "grad_norm": 1.8055574893951416, + "learning_rate": 1.7459806797904454e-06, + "loss": 3.1838, + "step": 20324 + }, + { + "epoch": 1.629452205808663, + "grad_norm": 1.6922931671142578, + "learning_rate": 1.7452477168962067e-06, + "loss": 3.2522, + "step": 20325 + }, + { + "epoch": 1.6295323805896855, + "grad_norm": 1.7415099143981934, + "learning_rate": 1.7445148931741806e-06, + "loss": 3.0782, + "step": 20326 + }, + { + "epoch": 1.6296125553707081, + "grad_norm": 1.8265810012817383, + "learning_rate": 1.743782208636724e-06, + "loss": 3.1163, + "step": 20327 + }, + { + "epoch": 1.6296927301517308, + "grad_norm": 1.7209041118621826, + "learning_rate": 1.7430496632961891e-06, + "loss": 3.1556, + "step": 20328 + }, + { + "epoch": 1.6297729049327534, + "grad_norm": 1.8550209999084473, + "learning_rate": 1.742317257164926e-06, + "loss": 3.1932, + "step": 20329 + }, + { + "epoch": 1.629853079713776, + "grad_norm": 0.7017265558242798, + "learning_rate": 1.7415849902552872e-06, + "loss": 2.4647, + "step": 20330 + }, + { + "epoch": 1.6299332544947986, + "grad_norm": 1.8160516023635864, + "learning_rate": 1.7408528625796128e-06, + "loss": 3.1107, + "step": 20331 + }, + { + "epoch": 1.6300134292758213, + "grad_norm": 1.7246406078338623, + "learning_rate": 1.7401208741502473e-06, + "loss": 3.2249, + "step": 20332 + }, + { + "epoch": 1.630093604056844, + "grad_norm": 1.746323823928833, + "learning_rate": 1.739389024979533e-06, + "loss": 3.0742, + "step": 20333 + }, + { + "epoch": 1.6301737788378665, + "grad_norm": 1.7835769653320312, + "learning_rate": 1.7386573150798092e-06, + "loss": 3.1569, + "step": 20334 + }, + { + "epoch": 1.6302539536188891, + "grad_norm": 1.7725915908813477, + "learning_rate": 1.7379257444634135e-06, + "loss": 3.1196, + "step": 20335 + }, + { + "epoch": 1.6303341283999118, + "grad_norm": 1.9017353057861328, + "learning_rate": 1.737194313142675e-06, + "loss": 3.2562, + "step": 20336 + }, + { + "epoch": 1.6304143031809344, + "grad_norm": 1.96170973777771, + "learning_rate": 1.7364630211299283e-06, + "loss": 3.3081, + "step": 20337 + }, + { + "epoch": 1.630494477961957, + "grad_norm": 1.8118841648101807, + "learning_rate": 1.7357318684375035e-06, + "loss": 3.1703, + "step": 20338 + }, + { + "epoch": 1.6305746527429799, + "grad_norm": 1.8055040836334229, + "learning_rate": 1.7350008550777264e-06, + "loss": 3.2292, + "step": 20339 + }, + { + "epoch": 1.6306548275240025, + "grad_norm": 1.7339818477630615, + "learning_rate": 1.734269981062925e-06, + "loss": 3.2005, + "step": 20340 + }, + { + "epoch": 1.630735002305025, + "grad_norm": 1.739088535308838, + "learning_rate": 1.7335392464054146e-06, + "loss": 3.2104, + "step": 20341 + }, + { + "epoch": 1.6308151770860477, + "grad_norm": 1.7740187644958496, + "learning_rate": 1.7328086511175202e-06, + "loss": 3.1846, + "step": 20342 + }, + { + "epoch": 1.6308953518670704, + "grad_norm": 1.7795937061309814, + "learning_rate": 1.7320781952115584e-06, + "loss": 3.2095, + "step": 20343 + }, + { + "epoch": 1.630975526648093, + "grad_norm": 1.804104208946228, + "learning_rate": 1.731347878699844e-06, + "loss": 3.2361, + "step": 20344 + }, + { + "epoch": 1.6310557014291156, + "grad_norm": 1.785221815109253, + "learning_rate": 1.7306177015946935e-06, + "loss": 3.2052, + "step": 20345 + }, + { + "epoch": 1.6311358762101382, + "grad_norm": 1.7335654497146606, + "learning_rate": 1.7298876639084106e-06, + "loss": 3.2429, + "step": 20346 + }, + { + "epoch": 1.6312160509911608, + "grad_norm": 1.7778366804122925, + "learning_rate": 1.7291577656533077e-06, + "loss": 3.2018, + "step": 20347 + }, + { + "epoch": 1.6312962257721835, + "grad_norm": 1.826845407485962, + "learning_rate": 1.7284280068416892e-06, + "loss": 3.1468, + "step": 20348 + }, + { + "epoch": 1.631376400553206, + "grad_norm": 1.7007051706314087, + "learning_rate": 1.72769838748586e-06, + "loss": 3.2156, + "step": 20349 + }, + { + "epoch": 1.6314565753342287, + "grad_norm": 1.6857954263687134, + "learning_rate": 1.7269689075981223e-06, + "loss": 3.1684, + "step": 20350 + }, + { + "epoch": 1.6315367501152513, + "grad_norm": 1.8169128894805908, + "learning_rate": 1.7262395671907717e-06, + "loss": 3.0826, + "step": 20351 + }, + { + "epoch": 1.631616924896274, + "grad_norm": 1.7615923881530762, + "learning_rate": 1.7255103662761052e-06, + "loss": 3.1361, + "step": 20352 + }, + { + "epoch": 1.6316970996772966, + "grad_norm": 1.8708395957946777, + "learning_rate": 1.7247813048664174e-06, + "loss": 3.1301, + "step": 20353 + }, + { + "epoch": 1.6317772744583192, + "grad_norm": 0.6708822846412659, + "learning_rate": 1.7240523829740009e-06, + "loss": 2.4829, + "step": 20354 + }, + { + "epoch": 1.6318574492393418, + "grad_norm": 1.8426772356033325, + "learning_rate": 1.7233236006111475e-06, + "loss": 3.3063, + "step": 20355 + }, + { + "epoch": 1.6319376240203645, + "grad_norm": 1.7931969165802002, + "learning_rate": 1.722594957790138e-06, + "loss": 3.1875, + "step": 20356 + }, + { + "epoch": 1.632017798801387, + "grad_norm": 1.8246245384216309, + "learning_rate": 1.72186645452326e-06, + "loss": 3.1104, + "step": 20357 + }, + { + "epoch": 1.6320979735824097, + "grad_norm": 1.843117117881775, + "learning_rate": 1.721138090822797e-06, + "loss": 3.1805, + "step": 20358 + }, + { + "epoch": 1.6321781483634323, + "grad_norm": 1.7398320436477661, + "learning_rate": 1.7204098667010272e-06, + "loss": 3.1814, + "step": 20359 + }, + { + "epoch": 1.632258323144455, + "grad_norm": 1.8663641214370728, + "learning_rate": 1.7196817821702316e-06, + "loss": 3.2163, + "step": 20360 + }, + { + "epoch": 1.6323384979254776, + "grad_norm": 0.6667073965072632, + "learning_rate": 1.7189538372426816e-06, + "loss": 2.4667, + "step": 20361 + }, + { + "epoch": 1.6324186727065002, + "grad_norm": 1.7880933284759521, + "learning_rate": 1.71822603193065e-06, + "loss": 3.163, + "step": 20362 + }, + { + "epoch": 1.6324988474875228, + "grad_norm": 1.7310233116149902, + "learning_rate": 1.7174983662464096e-06, + "loss": 3.3619, + "step": 20363 + }, + { + "epoch": 1.6325790222685455, + "grad_norm": 1.777239203453064, + "learning_rate": 1.7167708402022265e-06, + "loss": 3.1466, + "step": 20364 + }, + { + "epoch": 1.632659197049568, + "grad_norm": 1.756771445274353, + "learning_rate": 1.716043453810372e-06, + "loss": 3.2056, + "step": 20365 + }, + { + "epoch": 1.6327393718305907, + "grad_norm": 0.6608191728591919, + "learning_rate": 1.7153162070831009e-06, + "loss": 2.453, + "step": 20366 + }, + { + "epoch": 1.6328195466116133, + "grad_norm": 1.8036118745803833, + "learning_rate": 1.7145891000326797e-06, + "loss": 3.177, + "step": 20367 + }, + { + "epoch": 1.632899721392636, + "grad_norm": 1.687756896018982, + "learning_rate": 1.7138621326713667e-06, + "loss": 3.1194, + "step": 20368 + }, + { + "epoch": 1.6329798961736586, + "grad_norm": 1.8262726068496704, + "learning_rate": 1.7131353050114174e-06, + "loss": 3.225, + "step": 20369 + }, + { + "epoch": 1.6330600709546812, + "grad_norm": 1.7721507549285889, + "learning_rate": 1.712408617065089e-06, + "loss": 3.0706, + "step": 20370 + }, + { + "epoch": 1.6331402457357038, + "grad_norm": 1.8668338060379028, + "learning_rate": 1.7116820688446278e-06, + "loss": 3.2784, + "step": 20371 + }, + { + "epoch": 1.6332204205167264, + "grad_norm": 1.816280484199524, + "learning_rate": 1.7109556603622856e-06, + "loss": 3.2781, + "step": 20372 + }, + { + "epoch": 1.633300595297749, + "grad_norm": 1.787131667137146, + "learning_rate": 1.7102293916303091e-06, + "loss": 3.1818, + "step": 20373 + }, + { + "epoch": 1.6333807700787717, + "grad_norm": 1.7509888410568237, + "learning_rate": 1.7095032626609443e-06, + "loss": 3.181, + "step": 20374 + }, + { + "epoch": 1.6334609448597943, + "grad_norm": 1.7502764463424683, + "learning_rate": 1.7087772734664353e-06, + "loss": 3.1538, + "step": 20375 + }, + { + "epoch": 1.633541119640817, + "grad_norm": 1.7596077919006348, + "learning_rate": 1.7080514240590161e-06, + "loss": 3.2608, + "step": 20376 + }, + { + "epoch": 1.6336212944218396, + "grad_norm": 1.7182352542877197, + "learning_rate": 1.7073257144509282e-06, + "loss": 3.2344, + "step": 20377 + }, + { + "epoch": 1.6337014692028622, + "grad_norm": 1.7840226888656616, + "learning_rate": 1.7066001446544056e-06, + "loss": 3.1272, + "step": 20378 + }, + { + "epoch": 1.6337816439838848, + "grad_norm": 1.7698326110839844, + "learning_rate": 1.7058747146816824e-06, + "loss": 3.2288, + "step": 20379 + }, + { + "epoch": 1.6338618187649074, + "grad_norm": 1.736331582069397, + "learning_rate": 1.7051494245449907e-06, + "loss": 3.2767, + "step": 20380 + }, + { + "epoch": 1.63394199354593, + "grad_norm": 0.669063925743103, + "learning_rate": 1.704424274256553e-06, + "loss": 2.4713, + "step": 20381 + }, + { + "epoch": 1.6340221683269527, + "grad_norm": 1.8611414432525635, + "learning_rate": 1.7036992638285988e-06, + "loss": 3.1703, + "step": 20382 + }, + { + "epoch": 1.6341023431079753, + "grad_norm": 1.8009796142578125, + "learning_rate": 1.702974393273351e-06, + "loss": 3.3455, + "step": 20383 + }, + { + "epoch": 1.634182517888998, + "grad_norm": 1.756537675857544, + "learning_rate": 1.7022496626030317e-06, + "loss": 3.1851, + "step": 20384 + }, + { + "epoch": 1.6342626926700206, + "grad_norm": 1.7169322967529297, + "learning_rate": 1.7015250718298603e-06, + "loss": 3.1536, + "step": 20385 + }, + { + "epoch": 1.6343428674510432, + "grad_norm": 1.7232822179794312, + "learning_rate": 1.7008006209660487e-06, + "loss": 3.1695, + "step": 20386 + }, + { + "epoch": 1.6344230422320658, + "grad_norm": 2.0083553791046143, + "learning_rate": 1.7000763100238148e-06, + "loss": 3.2073, + "step": 20387 + }, + { + "epoch": 1.6345032170130884, + "grad_norm": 1.6978436708450317, + "learning_rate": 1.6993521390153689e-06, + "loss": 3.0854, + "step": 20388 + }, + { + "epoch": 1.634583391794111, + "grad_norm": 0.6515190005302429, + "learning_rate": 1.6986281079529198e-06, + "loss": 2.4509, + "step": 20389 + }, + { + "epoch": 1.6346635665751337, + "grad_norm": 1.782699465751648, + "learning_rate": 1.6979042168486782e-06, + "loss": 3.2097, + "step": 20390 + }, + { + "epoch": 1.6347437413561563, + "grad_norm": 1.877537488937378, + "learning_rate": 1.6971804657148427e-06, + "loss": 3.2522, + "step": 20391 + }, + { + "epoch": 1.634823916137179, + "grad_norm": 1.855285882949829, + "learning_rate": 1.6964568545636184e-06, + "loss": 3.2447, + "step": 20392 + }, + { + "epoch": 1.6349040909182015, + "grad_norm": 1.8236421346664429, + "learning_rate": 1.6957333834072054e-06, + "loss": 3.1095, + "step": 20393 + }, + { + "epoch": 1.6349842656992242, + "grad_norm": 0.6583083271980286, + "learning_rate": 1.6950100522578005e-06, + "loss": 2.467, + "step": 20394 + }, + { + "epoch": 1.6350644404802468, + "grad_norm": 1.747542142868042, + "learning_rate": 1.694286861127602e-06, + "loss": 3.1486, + "step": 20395 + }, + { + "epoch": 1.6351446152612694, + "grad_norm": 1.7295422554016113, + "learning_rate": 1.6935638100287966e-06, + "loss": 3.2264, + "step": 20396 + }, + { + "epoch": 1.635224790042292, + "grad_norm": 1.7038053274154663, + "learning_rate": 1.6928408989735789e-06, + "loss": 3.1368, + "step": 20397 + }, + { + "epoch": 1.6353049648233147, + "grad_norm": 1.8287506103515625, + "learning_rate": 1.692118127974135e-06, + "loss": 3.1582, + "step": 20398 + }, + { + "epoch": 1.6353851396043373, + "grad_norm": 1.7089917659759521, + "learning_rate": 1.6913954970426516e-06, + "loss": 3.1355, + "step": 20399 + }, + { + "epoch": 1.6354653143853601, + "grad_norm": 1.6589010953903198, + "learning_rate": 1.6906730061913156e-06, + "loss": 3.0563, + "step": 20400 + }, + { + "epoch": 1.6355454891663828, + "grad_norm": 1.7171649932861328, + "learning_rate": 1.6899506554323008e-06, + "loss": 3.1824, + "step": 20401 + }, + { + "epoch": 1.6356256639474054, + "grad_norm": 1.7810113430023193, + "learning_rate": 1.6892284447777885e-06, + "loss": 3.2025, + "step": 20402 + }, + { + "epoch": 1.635705838728428, + "grad_norm": 1.7525886297225952, + "learning_rate": 1.6885063742399567e-06, + "loss": 3.2375, + "step": 20403 + }, + { + "epoch": 1.6357860135094506, + "grad_norm": 1.690691351890564, + "learning_rate": 1.6877844438309776e-06, + "loss": 3.1113, + "step": 20404 + }, + { + "epoch": 1.6358661882904733, + "grad_norm": 1.6990993022918701, + "learning_rate": 1.6870626535630263e-06, + "loss": 3.0917, + "step": 20405 + }, + { + "epoch": 1.6359463630714959, + "grad_norm": 1.923722505569458, + "learning_rate": 1.6863410034482653e-06, + "loss": 3.119, + "step": 20406 + }, + { + "epoch": 1.6360265378525185, + "grad_norm": 1.7127759456634521, + "learning_rate": 1.6856194934988657e-06, + "loss": 3.1389, + "step": 20407 + }, + { + "epoch": 1.6361067126335411, + "grad_norm": 1.7777043581008911, + "learning_rate": 1.6848981237269913e-06, + "loss": 3.1259, + "step": 20408 + }, + { + "epoch": 1.6361868874145638, + "grad_norm": 1.8266104459762573, + "learning_rate": 1.6841768941448045e-06, + "loss": 3.1555, + "step": 20409 + }, + { + "epoch": 1.6362670621955864, + "grad_norm": 1.6767407655715942, + "learning_rate": 1.6834558047644655e-06, + "loss": 3.0324, + "step": 20410 + }, + { + "epoch": 1.636347236976609, + "grad_norm": 1.7397596836090088, + "learning_rate": 1.6827348555981293e-06, + "loss": 3.1882, + "step": 20411 + }, + { + "epoch": 1.6364274117576316, + "grad_norm": 1.7860080003738403, + "learning_rate": 1.6820140466579526e-06, + "loss": 3.1754, + "step": 20412 + }, + { + "epoch": 1.6365075865386542, + "grad_norm": 1.7618911266326904, + "learning_rate": 1.6812933779560858e-06, + "loss": 3.1139, + "step": 20413 + }, + { + "epoch": 1.6365877613196769, + "grad_norm": 1.71407949924469, + "learning_rate": 1.6805728495046825e-06, + "loss": 3.2377, + "step": 20414 + }, + { + "epoch": 1.6366679361006995, + "grad_norm": 1.782716155052185, + "learning_rate": 1.6798524613158906e-06, + "loss": 3.3251, + "step": 20415 + }, + { + "epoch": 1.6367481108817221, + "grad_norm": 1.8051759004592896, + "learning_rate": 1.6791322134018517e-06, + "loss": 3.0618, + "step": 20416 + }, + { + "epoch": 1.6368282856627447, + "grad_norm": 1.8963488340377808, + "learning_rate": 1.6784121057747104e-06, + "loss": 3.1048, + "step": 20417 + }, + { + "epoch": 1.6369084604437674, + "grad_norm": 1.788185715675354, + "learning_rate": 1.6776921384466095e-06, + "loss": 3.1469, + "step": 20418 + }, + { + "epoch": 1.63698863522479, + "grad_norm": 1.747887372970581, + "learning_rate": 1.6769723114296844e-06, + "loss": 3.0863, + "step": 20419 + }, + { + "epoch": 1.6370688100058128, + "grad_norm": 1.832816243171692, + "learning_rate": 1.6762526247360765e-06, + "loss": 3.1466, + "step": 20420 + }, + { + "epoch": 1.6371489847868355, + "grad_norm": 0.668327271938324, + "learning_rate": 1.6755330783779133e-06, + "loss": 2.4744, + "step": 20421 + }, + { + "epoch": 1.637229159567858, + "grad_norm": 1.8199869394302368, + "learning_rate": 1.674813672367328e-06, + "loss": 3.2097, + "step": 20422 + }, + { + "epoch": 1.6373093343488807, + "grad_norm": 1.8308753967285156, + "learning_rate": 1.67409440671645e-06, + "loss": 3.2394, + "step": 20423 + }, + { + "epoch": 1.6373895091299033, + "grad_norm": 1.7207670211791992, + "learning_rate": 1.6733752814374071e-06, + "loss": 3.077, + "step": 20424 + }, + { + "epoch": 1.637469683910926, + "grad_norm": 0.6493257880210876, + "learning_rate": 1.6726562965423232e-06, + "loss": 2.4567, + "step": 20425 + }, + { + "epoch": 1.6375498586919486, + "grad_norm": 1.8149824142456055, + "learning_rate": 1.6719374520433185e-06, + "loss": 3.1699, + "step": 20426 + }, + { + "epoch": 1.6376300334729712, + "grad_norm": 0.698051393032074, + "learning_rate": 1.6712187479525122e-06, + "loss": 2.4783, + "step": 20427 + }, + { + "epoch": 1.6377102082539938, + "grad_norm": 0.6543152928352356, + "learning_rate": 1.6705001842820223e-06, + "loss": 2.4656, + "step": 20428 + }, + { + "epoch": 1.6377903830350165, + "grad_norm": 1.8533638715744019, + "learning_rate": 1.669781761043965e-06, + "loss": 3.131, + "step": 20429 + }, + { + "epoch": 1.637870557816039, + "grad_norm": 1.8034926652908325, + "learning_rate": 1.6690634782504523e-06, + "loss": 3.0786, + "step": 20430 + }, + { + "epoch": 1.6379507325970617, + "grad_norm": 1.8137649297714233, + "learning_rate": 1.6683453359135915e-06, + "loss": 3.2038, + "step": 20431 + }, + { + "epoch": 1.6380309073780843, + "grad_norm": 1.7459211349487305, + "learning_rate": 1.6676273340454918e-06, + "loss": 3.1506, + "step": 20432 + }, + { + "epoch": 1.638111082159107, + "grad_norm": 1.8459999561309814, + "learning_rate": 1.666909472658259e-06, + "loss": 3.2473, + "step": 20433 + }, + { + "epoch": 1.6381912569401296, + "grad_norm": 1.6872642040252686, + "learning_rate": 1.6661917517639948e-06, + "loss": 3.1982, + "step": 20434 + }, + { + "epoch": 1.6382714317211522, + "grad_norm": 1.7108231782913208, + "learning_rate": 1.6654741713748036e-06, + "loss": 3.1304, + "step": 20435 + }, + { + "epoch": 1.6383516065021748, + "grad_norm": 1.7005681991577148, + "learning_rate": 1.664756731502778e-06, + "loss": 3.1164, + "step": 20436 + }, + { + "epoch": 1.6384317812831974, + "grad_norm": 1.818490743637085, + "learning_rate": 1.6640394321600161e-06, + "loss": 3.173, + "step": 20437 + }, + { + "epoch": 1.63851195606422, + "grad_norm": 1.715013861656189, + "learning_rate": 1.6633222733586118e-06, + "loss": 3.075, + "step": 20438 + }, + { + "epoch": 1.6385921308452427, + "grad_norm": 1.7755672931671143, + "learning_rate": 1.6626052551106565e-06, + "loss": 3.1317, + "step": 20439 + }, + { + "epoch": 1.6386723056262653, + "grad_norm": 1.82355797290802, + "learning_rate": 1.6618883774282391e-06, + "loss": 3.1849, + "step": 20440 + }, + { + "epoch": 1.638752480407288, + "grad_norm": 1.8468940258026123, + "learning_rate": 1.6611716403234435e-06, + "loss": 3.2086, + "step": 20441 + }, + { + "epoch": 1.6388326551883106, + "grad_norm": 1.7415729761123657, + "learning_rate": 1.6604550438083556e-06, + "loss": 3.065, + "step": 20442 + }, + { + "epoch": 1.6389128299693332, + "grad_norm": 1.9107040166854858, + "learning_rate": 1.6597385878950556e-06, + "loss": 3.2489, + "step": 20443 + }, + { + "epoch": 1.6389930047503558, + "grad_norm": 1.7350878715515137, + "learning_rate": 1.659022272595624e-06, + "loss": 3.2313, + "step": 20444 + }, + { + "epoch": 1.6390731795313784, + "grad_norm": 1.8035881519317627, + "learning_rate": 1.6583060979221398e-06, + "loss": 3.3452, + "step": 20445 + }, + { + "epoch": 1.639153354312401, + "grad_norm": 1.759131669998169, + "learning_rate": 1.6575900638866727e-06, + "loss": 3.1588, + "step": 20446 + }, + { + "epoch": 1.6392335290934237, + "grad_norm": 1.8664315938949585, + "learning_rate": 1.6568741705012958e-06, + "loss": 3.2092, + "step": 20447 + }, + { + "epoch": 1.6393137038744463, + "grad_norm": 1.7476050853729248, + "learning_rate": 1.656158417778081e-06, + "loss": 3.0978, + "step": 20448 + }, + { + "epoch": 1.639393878655469, + "grad_norm": 1.759488582611084, + "learning_rate": 1.6554428057290938e-06, + "loss": 3.197, + "step": 20449 + }, + { + "epoch": 1.6394740534364916, + "grad_norm": 1.7944713830947876, + "learning_rate": 1.6547273343664039e-06, + "loss": 3.1715, + "step": 20450 + }, + { + "epoch": 1.6395542282175142, + "grad_norm": 1.7494410276412964, + "learning_rate": 1.6540120037020658e-06, + "loss": 3.1548, + "step": 20451 + }, + { + "epoch": 1.6396344029985368, + "grad_norm": 0.6690473556518555, + "learning_rate": 1.6532968137481453e-06, + "loss": 2.4683, + "step": 20452 + }, + { + "epoch": 1.6397145777795594, + "grad_norm": 0.6825379729270935, + "learning_rate": 1.6525817645166986e-06, + "loss": 2.491, + "step": 20453 + }, + { + "epoch": 1.639794752560582, + "grad_norm": 1.7140599489212036, + "learning_rate": 1.6518668560197804e-06, + "loss": 3.1759, + "step": 20454 + }, + { + "epoch": 1.6398749273416047, + "grad_norm": 1.6929161548614502, + "learning_rate": 1.6511520882694488e-06, + "loss": 3.1841, + "step": 20455 + }, + { + "epoch": 1.6399551021226273, + "grad_norm": 1.671316146850586, + "learning_rate": 1.6504374612777473e-06, + "loss": 3.1182, + "step": 20456 + }, + { + "epoch": 1.64003527690365, + "grad_norm": 1.7803237438201904, + "learning_rate": 1.6497229750567267e-06, + "loss": 3.2458, + "step": 20457 + }, + { + "epoch": 1.6401154516846725, + "grad_norm": 1.8995369672775269, + "learning_rate": 1.649008629618435e-06, + "loss": 3.0683, + "step": 20458 + }, + { + "epoch": 1.6401956264656952, + "grad_norm": 1.7653260231018066, + "learning_rate": 1.6482944249749144e-06, + "loss": 3.2634, + "step": 20459 + }, + { + "epoch": 1.6402758012467178, + "grad_norm": 1.7462812662124634, + "learning_rate": 1.647580361138209e-06, + "loss": 3.2061, + "step": 20460 + }, + { + "epoch": 1.6403559760277404, + "grad_norm": 1.7458522319793701, + "learning_rate": 1.6468664381203524e-06, + "loss": 3.1165, + "step": 20461 + }, + { + "epoch": 1.640436150808763, + "grad_norm": 1.7328557968139648, + "learning_rate": 1.646152655933383e-06, + "loss": 3.2216, + "step": 20462 + }, + { + "epoch": 1.6405163255897857, + "grad_norm": 1.8674108982086182, + "learning_rate": 1.6454390145893361e-06, + "loss": 3.2386, + "step": 20463 + }, + { + "epoch": 1.6405965003708083, + "grad_norm": 1.818966269493103, + "learning_rate": 1.644725514100244e-06, + "loss": 3.1351, + "step": 20464 + }, + { + "epoch": 1.640676675151831, + "grad_norm": 1.70103919506073, + "learning_rate": 1.6440121544781362e-06, + "loss": 3.1521, + "step": 20465 + }, + { + "epoch": 1.6407568499328535, + "grad_norm": 1.7580461502075195, + "learning_rate": 1.643298935735037e-06, + "loss": 3.1317, + "step": 20466 + }, + { + "epoch": 1.6408370247138762, + "grad_norm": 1.8463765382766724, + "learning_rate": 1.6425858578829723e-06, + "loss": 3.2584, + "step": 20467 + }, + { + "epoch": 1.6409171994948988, + "grad_norm": 1.9823517799377441, + "learning_rate": 1.6418729209339634e-06, + "loss": 3.1787, + "step": 20468 + }, + { + "epoch": 1.6409973742759214, + "grad_norm": 1.808775544166565, + "learning_rate": 1.6411601249000331e-06, + "loss": 3.1101, + "step": 20469 + }, + { + "epoch": 1.641077549056944, + "grad_norm": 1.8073104619979858, + "learning_rate": 1.640447469793196e-06, + "loss": 3.2622, + "step": 20470 + }, + { + "epoch": 1.6411577238379667, + "grad_norm": 1.813984990119934, + "learning_rate": 1.6397349556254716e-06, + "loss": 3.1615, + "step": 20471 + }, + { + "epoch": 1.6412378986189893, + "grad_norm": 1.7491368055343628, + "learning_rate": 1.639022582408867e-06, + "loss": 3.1223, + "step": 20472 + }, + { + "epoch": 1.641318073400012, + "grad_norm": 1.8151681423187256, + "learning_rate": 1.6383103501553942e-06, + "loss": 3.1195, + "step": 20473 + }, + { + "epoch": 1.6413982481810345, + "grad_norm": 1.7913755178451538, + "learning_rate": 1.6375982588770622e-06, + "loss": 3.172, + "step": 20474 + }, + { + "epoch": 1.6414784229620571, + "grad_norm": 1.7697806358337402, + "learning_rate": 1.6368863085858766e-06, + "loss": 3.1671, + "step": 20475 + }, + { + "epoch": 1.6415585977430798, + "grad_norm": 1.751241683959961, + "learning_rate": 1.6361744992938421e-06, + "loss": 3.1636, + "step": 20476 + }, + { + "epoch": 1.6416387725241024, + "grad_norm": 1.7560595273971558, + "learning_rate": 1.635462831012956e-06, + "loss": 3.1326, + "step": 20477 + }, + { + "epoch": 1.641718947305125, + "grad_norm": 1.6704800128936768, + "learning_rate": 1.6347513037552187e-06, + "loss": 3.2709, + "step": 20478 + }, + { + "epoch": 1.6417991220861476, + "grad_norm": 1.8657463788986206, + "learning_rate": 1.634039917532626e-06, + "loss": 3.1285, + "step": 20479 + }, + { + "epoch": 1.6418792968671703, + "grad_norm": 1.7762539386749268, + "learning_rate": 1.6333286723571728e-06, + "loss": 3.1104, + "step": 20480 + }, + { + "epoch": 1.6419594716481931, + "grad_norm": 2.0093703269958496, + "learning_rate": 1.6326175682408507e-06, + "loss": 3.2109, + "step": 20481 + }, + { + "epoch": 1.6420396464292157, + "grad_norm": 1.8105404376983643, + "learning_rate": 1.6319066051956457e-06, + "loss": 3.2303, + "step": 20482 + }, + { + "epoch": 1.6421198212102384, + "grad_norm": 1.7125965356826782, + "learning_rate": 1.6311957832335456e-06, + "loss": 3.1229, + "step": 20483 + }, + { + "epoch": 1.642199995991261, + "grad_norm": 1.8375415802001953, + "learning_rate": 1.6304851023665358e-06, + "loss": 3.0922, + "step": 20484 + }, + { + "epoch": 1.6422801707722836, + "grad_norm": 1.7324535846710205, + "learning_rate": 1.6297745626065964e-06, + "loss": 3.1622, + "step": 20485 + }, + { + "epoch": 1.6423603455533062, + "grad_norm": 1.8131855726242065, + "learning_rate": 1.6290641639657113e-06, + "loss": 3.1415, + "step": 20486 + }, + { + "epoch": 1.6424405203343289, + "grad_norm": 1.6874523162841797, + "learning_rate": 1.6283539064558518e-06, + "loss": 3.1354, + "step": 20487 + }, + { + "epoch": 1.6425206951153515, + "grad_norm": 1.7438210248947144, + "learning_rate": 1.627643790088994e-06, + "loss": 3.0938, + "step": 20488 + }, + { + "epoch": 1.642600869896374, + "grad_norm": 1.8736861944198608, + "learning_rate": 1.6269338148771118e-06, + "loss": 3.1316, + "step": 20489 + }, + { + "epoch": 1.6426810446773967, + "grad_norm": 0.6813695430755615, + "learning_rate": 1.6262239808321755e-06, + "loss": 2.4587, + "step": 20490 + }, + { + "epoch": 1.6427612194584194, + "grad_norm": 1.7818864583969116, + "learning_rate": 1.6255142879661522e-06, + "loss": 3.1747, + "step": 20491 + }, + { + "epoch": 1.642841394239442, + "grad_norm": 1.715339183807373, + "learning_rate": 1.6248047362910057e-06, + "loss": 3.1788, + "step": 20492 + }, + { + "epoch": 1.6429215690204646, + "grad_norm": 1.7611223459243774, + "learning_rate": 1.6240953258186988e-06, + "loss": 3.1403, + "step": 20493 + }, + { + "epoch": 1.6430017438014872, + "grad_norm": 1.744408369064331, + "learning_rate": 1.6233860565611937e-06, + "loss": 3.1476, + "step": 20494 + }, + { + "epoch": 1.6430819185825098, + "grad_norm": 1.7930546998977661, + "learning_rate": 1.6226769285304466e-06, + "loss": 3.2075, + "step": 20495 + }, + { + "epoch": 1.6431620933635325, + "grad_norm": 1.888741135597229, + "learning_rate": 1.621967941738416e-06, + "loss": 3.1816, + "step": 20496 + }, + { + "epoch": 1.643242268144555, + "grad_norm": 1.7406448125839233, + "learning_rate": 1.621259096197052e-06, + "loss": 3.1025, + "step": 20497 + }, + { + "epoch": 1.6433224429255777, + "grad_norm": 1.7869882583618164, + "learning_rate": 1.6205503919183062e-06, + "loss": 3.1825, + "step": 20498 + }, + { + "epoch": 1.6434026177066003, + "grad_norm": 1.7763960361480713, + "learning_rate": 1.6198418289141271e-06, + "loss": 3.2833, + "step": 20499 + }, + { + "epoch": 1.643482792487623, + "grad_norm": 1.701165795326233, + "learning_rate": 1.6191334071964615e-06, + "loss": 3.1051, + "step": 20500 + }, + { + "epoch": 1.6435629672686458, + "grad_norm": 1.8253250122070312, + "learning_rate": 1.6184251267772556e-06, + "loss": 3.2623, + "step": 20501 + }, + { + "epoch": 1.6436431420496684, + "grad_norm": 1.7944890260696411, + "learning_rate": 1.6177169876684463e-06, + "loss": 3.0732, + "step": 20502 + }, + { + "epoch": 1.643723316830691, + "grad_norm": 1.789570689201355, + "learning_rate": 1.6170089898819729e-06, + "loss": 3.1892, + "step": 20503 + }, + { + "epoch": 1.6438034916117137, + "grad_norm": 1.7565803527832031, + "learning_rate": 1.616301133429775e-06, + "loss": 3.1571, + "step": 20504 + }, + { + "epoch": 1.6438836663927363, + "grad_norm": 1.706778883934021, + "learning_rate": 1.6155934183237842e-06, + "loss": 3.1339, + "step": 20505 + }, + { + "epoch": 1.643963841173759, + "grad_norm": 1.7708295583724976, + "learning_rate": 1.6148858445759363e-06, + "loss": 3.2915, + "step": 20506 + }, + { + "epoch": 1.6440440159547816, + "grad_norm": 1.9238386154174805, + "learning_rate": 1.6141784121981553e-06, + "loss": 3.186, + "step": 20507 + }, + { + "epoch": 1.6441241907358042, + "grad_norm": 1.800136685371399, + "learning_rate": 1.6134711212023713e-06, + "loss": 3.1816, + "step": 20508 + }, + { + "epoch": 1.6442043655168268, + "grad_norm": 1.744911551475525, + "learning_rate": 1.6127639716005084e-06, + "loss": 3.1516, + "step": 20509 + }, + { + "epoch": 1.6442845402978494, + "grad_norm": 1.8145232200622559, + "learning_rate": 1.612056963404488e-06, + "loss": 3.1869, + "step": 20510 + }, + { + "epoch": 1.644364715078872, + "grad_norm": 1.8661105632781982, + "learning_rate": 1.611350096626234e-06, + "loss": 3.187, + "step": 20511 + }, + { + "epoch": 1.6444448898598947, + "grad_norm": 1.8730194568634033, + "learning_rate": 1.6106433712776593e-06, + "loss": 3.1703, + "step": 20512 + }, + { + "epoch": 1.6445250646409173, + "grad_norm": 1.7887722253799438, + "learning_rate": 1.6099367873706796e-06, + "loss": 3.1643, + "step": 20513 + }, + { + "epoch": 1.64460523942194, + "grad_norm": 1.846357822418213, + "learning_rate": 1.6092303449172087e-06, + "loss": 3.1616, + "step": 20514 + }, + { + "epoch": 1.6446854142029625, + "grad_norm": 1.9217864274978638, + "learning_rate": 1.6085240439291572e-06, + "loss": 3.1269, + "step": 20515 + }, + { + "epoch": 1.6447655889839852, + "grad_norm": 1.805039882659912, + "learning_rate": 1.6078178844184355e-06, + "loss": 3.2235, + "step": 20516 + }, + { + "epoch": 1.6448457637650078, + "grad_norm": 1.860242247581482, + "learning_rate": 1.6071118663969443e-06, + "loss": 3.1833, + "step": 20517 + }, + { + "epoch": 1.6449259385460304, + "grad_norm": 1.9363566637039185, + "learning_rate": 1.6064059898765893e-06, + "loss": 3.2817, + "step": 20518 + }, + { + "epoch": 1.645006113327053, + "grad_norm": 1.7700669765472412, + "learning_rate": 1.6057002548692701e-06, + "loss": 3.2056, + "step": 20519 + }, + { + "epoch": 1.6450862881080757, + "grad_norm": 0.6790754795074463, + "learning_rate": 1.6049946613868873e-06, + "loss": 2.478, + "step": 20520 + }, + { + "epoch": 1.6451664628890983, + "grad_norm": 1.6799370050430298, + "learning_rate": 1.6042892094413376e-06, + "loss": 3.1605, + "step": 20521 + }, + { + "epoch": 1.645246637670121, + "grad_norm": 1.7611463069915771, + "learning_rate": 1.6035838990445107e-06, + "loss": 3.1722, + "step": 20522 + }, + { + "epoch": 1.6453268124511435, + "grad_norm": 1.8354218006134033, + "learning_rate": 1.6028787302083005e-06, + "loss": 3.1462, + "step": 20523 + }, + { + "epoch": 1.6454069872321662, + "grad_norm": 1.8530911207199097, + "learning_rate": 1.602173702944595e-06, + "loss": 3.2212, + "step": 20524 + }, + { + "epoch": 1.6454871620131888, + "grad_norm": 1.7491750717163086, + "learning_rate": 1.6014688172652805e-06, + "loss": 3.0299, + "step": 20525 + }, + { + "epoch": 1.6455673367942114, + "grad_norm": 1.8213162422180176, + "learning_rate": 1.6007640731822449e-06, + "loss": 3.1852, + "step": 20526 + }, + { + "epoch": 1.645647511575234, + "grad_norm": 1.788415551185608, + "learning_rate": 1.6000594707073647e-06, + "loss": 3.1427, + "step": 20527 + }, + { + "epoch": 1.6457276863562567, + "grad_norm": 1.8394310474395752, + "learning_rate": 1.5993550098525211e-06, + "loss": 3.1517, + "step": 20528 + }, + { + "epoch": 1.6458078611372793, + "grad_norm": 1.7761790752410889, + "learning_rate": 1.5986506906295919e-06, + "loss": 3.0675, + "step": 20529 + }, + { + "epoch": 1.645888035918302, + "grad_norm": 1.7263673543930054, + "learning_rate": 1.5979465130504502e-06, + "loss": 3.1946, + "step": 20530 + }, + { + "epoch": 1.6459682106993245, + "grad_norm": 1.7938514947891235, + "learning_rate": 1.5972424771269723e-06, + "loss": 3.2645, + "step": 20531 + }, + { + "epoch": 1.6460483854803472, + "grad_norm": 1.733643651008606, + "learning_rate": 1.5965385828710223e-06, + "loss": 3.1353, + "step": 20532 + }, + { + "epoch": 1.6461285602613698, + "grad_norm": 1.6519505977630615, + "learning_rate": 1.5958348302944693e-06, + "loss": 3.1196, + "step": 20533 + }, + { + "epoch": 1.6462087350423924, + "grad_norm": 1.7737483978271484, + "learning_rate": 1.5951312194091794e-06, + "loss": 3.0948, + "step": 20534 + }, + { + "epoch": 1.646288909823415, + "grad_norm": 1.6968767642974854, + "learning_rate": 1.5944277502270156e-06, + "loss": 3.1756, + "step": 20535 + }, + { + "epoch": 1.6463690846044376, + "grad_norm": 1.7291518449783325, + "learning_rate": 1.5937244227598404e-06, + "loss": 3.1016, + "step": 20536 + }, + { + "epoch": 1.6464492593854603, + "grad_norm": 1.7624380588531494, + "learning_rate": 1.5930212370195054e-06, + "loss": 3.1854, + "step": 20537 + }, + { + "epoch": 1.646529434166483, + "grad_norm": 1.7173503637313843, + "learning_rate": 1.5923181930178688e-06, + "loss": 3.2679, + "step": 20538 + }, + { + "epoch": 1.6466096089475055, + "grad_norm": 1.7824527025222778, + "learning_rate": 1.5916152907667858e-06, + "loss": 3.226, + "step": 20539 + }, + { + "epoch": 1.6466897837285281, + "grad_norm": 1.727605938911438, + "learning_rate": 1.5909125302781049e-06, + "loss": 3.1796, + "step": 20540 + }, + { + "epoch": 1.6467699585095508, + "grad_norm": 1.752695918083191, + "learning_rate": 1.590209911563677e-06, + "loss": 3.1124, + "step": 20541 + }, + { + "epoch": 1.6468501332905734, + "grad_norm": 1.799920916557312, + "learning_rate": 1.5895074346353444e-06, + "loss": 3.1146, + "step": 20542 + }, + { + "epoch": 1.646930308071596, + "grad_norm": 1.796128511428833, + "learning_rate": 1.588805099504952e-06, + "loss": 3.2352, + "step": 20543 + }, + { + "epoch": 1.6470104828526186, + "grad_norm": 0.6606810688972473, + "learning_rate": 1.5881029061843424e-06, + "loss": 2.4769, + "step": 20544 + }, + { + "epoch": 1.6470906576336413, + "grad_norm": 1.8600658178329468, + "learning_rate": 1.5874008546853515e-06, + "loss": 3.1862, + "step": 20545 + }, + { + "epoch": 1.6471708324146639, + "grad_norm": 1.7917004823684692, + "learning_rate": 1.58669894501982e-06, + "loss": 3.088, + "step": 20546 + }, + { + "epoch": 1.6472510071956865, + "grad_norm": 1.7409509420394897, + "learning_rate": 1.5859971771995774e-06, + "loss": 3.1256, + "step": 20547 + }, + { + "epoch": 1.6473311819767091, + "grad_norm": 1.892688274383545, + "learning_rate": 1.5852955512364566e-06, + "loss": 3.1667, + "step": 20548 + }, + { + "epoch": 1.6474113567577318, + "grad_norm": 1.8360708951950073, + "learning_rate": 1.5845940671422878e-06, + "loss": 3.1449, + "step": 20549 + }, + { + "epoch": 1.6474915315387544, + "grad_norm": 1.7050317525863647, + "learning_rate": 1.583892724928896e-06, + "loss": 3.1937, + "step": 20550 + }, + { + "epoch": 1.647571706319777, + "grad_norm": 1.7594228982925415, + "learning_rate": 1.5831915246081098e-06, + "loss": 3.0979, + "step": 20551 + }, + { + "epoch": 1.6476518811007996, + "grad_norm": 1.8086175918579102, + "learning_rate": 1.5824904661917463e-06, + "loss": 3.231, + "step": 20552 + }, + { + "epoch": 1.6477320558818223, + "grad_norm": 1.7579131126403809, + "learning_rate": 1.5817895496916258e-06, + "loss": 3.1701, + "step": 20553 + }, + { + "epoch": 1.6478122306628449, + "grad_norm": 1.8059947490692139, + "learning_rate": 1.581088775119567e-06, + "loss": 3.1373, + "step": 20554 + }, + { + "epoch": 1.6478924054438675, + "grad_norm": 1.7505276203155518, + "learning_rate": 1.5803881424873846e-06, + "loss": 3.1416, + "step": 20555 + }, + { + "epoch": 1.6479725802248901, + "grad_norm": 0.6359657049179077, + "learning_rate": 1.5796876518068926e-06, + "loss": 2.4563, + "step": 20556 + }, + { + "epoch": 1.6480527550059128, + "grad_norm": 1.834270715713501, + "learning_rate": 1.5789873030898971e-06, + "loss": 3.1523, + "step": 20557 + }, + { + "epoch": 1.6481329297869354, + "grad_norm": 1.792266607284546, + "learning_rate": 1.5782870963482089e-06, + "loss": 3.1794, + "step": 20558 + }, + { + "epoch": 1.648213104567958, + "grad_norm": 1.7784454822540283, + "learning_rate": 1.5775870315936303e-06, + "loss": 3.2844, + "step": 20559 + }, + { + "epoch": 1.6482932793489806, + "grad_norm": 1.733763337135315, + "learning_rate": 1.5768871088379677e-06, + "loss": 3.2237, + "step": 20560 + }, + { + "epoch": 1.6483734541300032, + "grad_norm": 1.8528650999069214, + "learning_rate": 1.5761873280930208e-06, + "loss": 3.1472, + "step": 20561 + }, + { + "epoch": 1.648453628911026, + "grad_norm": 1.790487289428711, + "learning_rate": 1.575487689370585e-06, + "loss": 3.1997, + "step": 20562 + }, + { + "epoch": 1.6485338036920487, + "grad_norm": 1.7444652318954468, + "learning_rate": 1.5747881926824581e-06, + "loss": 3.0753, + "step": 20563 + }, + { + "epoch": 1.6486139784730713, + "grad_norm": 1.753530502319336, + "learning_rate": 1.5740888380404317e-06, + "loss": 3.1625, + "step": 20564 + }, + { + "epoch": 1.648694153254094, + "grad_norm": 1.8607791662216187, + "learning_rate": 1.5733896254562996e-06, + "loss": 3.2338, + "step": 20565 + }, + { + "epoch": 1.6487743280351166, + "grad_norm": 1.8539029359817505, + "learning_rate": 1.5726905549418492e-06, + "loss": 3.2485, + "step": 20566 + }, + { + "epoch": 1.6488545028161392, + "grad_norm": 1.6783946752548218, + "learning_rate": 1.5719916265088653e-06, + "loss": 3.0781, + "step": 20567 + }, + { + "epoch": 1.6489346775971618, + "grad_norm": 1.8072398900985718, + "learning_rate": 1.5712928401691308e-06, + "loss": 3.2014, + "step": 20568 + }, + { + "epoch": 1.6490148523781845, + "grad_norm": 1.8424311876296997, + "learning_rate": 1.5705941959344294e-06, + "loss": 3.2158, + "step": 20569 + }, + { + "epoch": 1.649095027159207, + "grad_norm": 1.595767855644226, + "learning_rate": 1.5698956938165389e-06, + "loss": 3.1564, + "step": 20570 + }, + { + "epoch": 1.6491752019402297, + "grad_norm": 0.7053964734077454, + "learning_rate": 1.5691973338272381e-06, + "loss": 2.4655, + "step": 20571 + }, + { + "epoch": 1.6492553767212523, + "grad_norm": 1.7752124071121216, + "learning_rate": 1.568499115978296e-06, + "loss": 3.1148, + "step": 20572 + }, + { + "epoch": 1.649335551502275, + "grad_norm": 1.765553593635559, + "learning_rate": 1.5678010402814881e-06, + "loss": 3.1287, + "step": 20573 + }, + { + "epoch": 1.6494157262832976, + "grad_norm": 0.6649237275123596, + "learning_rate": 1.5671031067485832e-06, + "loss": 2.4619, + "step": 20574 + }, + { + "epoch": 1.6494959010643202, + "grad_norm": 1.7040461301803589, + "learning_rate": 1.5664053153913473e-06, + "loss": 3.053, + "step": 20575 + }, + { + "epoch": 1.6495760758453428, + "grad_norm": 1.7699381113052368, + "learning_rate": 1.5657076662215476e-06, + "loss": 3.2092, + "step": 20576 + }, + { + "epoch": 1.6496562506263655, + "grad_norm": 1.7028530836105347, + "learning_rate": 1.5650101592509415e-06, + "loss": 3.0943, + "step": 20577 + }, + { + "epoch": 1.649736425407388, + "grad_norm": 1.7972164154052734, + "learning_rate": 1.5643127944912916e-06, + "loss": 3.2306, + "step": 20578 + }, + { + "epoch": 1.6498166001884107, + "grad_norm": 1.7976186275482178, + "learning_rate": 1.5636155719543556e-06, + "loss": 3.2521, + "step": 20579 + }, + { + "epoch": 1.6498967749694333, + "grad_norm": 1.8939902782440186, + "learning_rate": 1.5629184916518869e-06, + "loss": 3.2606, + "step": 20580 + }, + { + "epoch": 1.649976949750456, + "grad_norm": 1.865561842918396, + "learning_rate": 1.562221553595642e-06, + "loss": 3.2893, + "step": 20581 + }, + { + "epoch": 1.6500571245314786, + "grad_norm": 1.785902500152588, + "learning_rate": 1.5615247577973658e-06, + "loss": 3.2196, + "step": 20582 + }, + { + "epoch": 1.6501372993125014, + "grad_norm": 1.7639700174331665, + "learning_rate": 1.5608281042688079e-06, + "loss": 3.1796, + "step": 20583 + }, + { + "epoch": 1.650217474093524, + "grad_norm": 1.8555632829666138, + "learning_rate": 1.560131593021713e-06, + "loss": 3.1414, + "step": 20584 + }, + { + "epoch": 1.6502976488745467, + "grad_norm": 1.7585315704345703, + "learning_rate": 1.5594352240678257e-06, + "loss": 3.0607, + "step": 20585 + }, + { + "epoch": 1.6503778236555693, + "grad_norm": 1.782500147819519, + "learning_rate": 1.5587389974188882e-06, + "loss": 3.2153, + "step": 20586 + }, + { + "epoch": 1.650457998436592, + "grad_norm": 1.7856370210647583, + "learning_rate": 1.5580429130866338e-06, + "loss": 3.0447, + "step": 20587 + }, + { + "epoch": 1.6505381732176145, + "grad_norm": 1.761043906211853, + "learning_rate": 1.5573469710828004e-06, + "loss": 3.2294, + "step": 20588 + }, + { + "epoch": 1.6506183479986372, + "grad_norm": 1.744983434677124, + "learning_rate": 1.5566511714191223e-06, + "loss": 3.1776, + "step": 20589 + }, + { + "epoch": 1.6506985227796598, + "grad_norm": 1.6891896724700928, + "learning_rate": 1.555955514107329e-06, + "loss": 3.1093, + "step": 20590 + }, + { + "epoch": 1.6507786975606824, + "grad_norm": 1.8178247213363647, + "learning_rate": 1.5552599991591522e-06, + "loss": 3.1761, + "step": 20591 + }, + { + "epoch": 1.650858872341705, + "grad_norm": 1.730825424194336, + "learning_rate": 1.5545646265863134e-06, + "loss": 3.1843, + "step": 20592 + }, + { + "epoch": 1.6509390471227277, + "grad_norm": 1.809814453125, + "learning_rate": 1.5538693964005392e-06, + "loss": 3.118, + "step": 20593 + }, + { + "epoch": 1.6510192219037503, + "grad_norm": 1.8746118545532227, + "learning_rate": 1.5531743086135487e-06, + "loss": 3.1112, + "step": 20594 + }, + { + "epoch": 1.651099396684773, + "grad_norm": 1.8414944410324097, + "learning_rate": 1.5524793632370637e-06, + "loss": 3.1941, + "step": 20595 + }, + { + "epoch": 1.6511795714657955, + "grad_norm": 0.6634572744369507, + "learning_rate": 1.5517845602828019e-06, + "loss": 2.4535, + "step": 20596 + }, + { + "epoch": 1.6512597462468181, + "grad_norm": 1.6499437093734741, + "learning_rate": 1.5510898997624713e-06, + "loss": 3.0768, + "step": 20597 + }, + { + "epoch": 1.6513399210278408, + "grad_norm": 1.7259538173675537, + "learning_rate": 1.550395381687788e-06, + "loss": 3.1209, + "step": 20598 + }, + { + "epoch": 1.6514200958088634, + "grad_norm": 1.7354016304016113, + "learning_rate": 1.5497010060704609e-06, + "loss": 3.0388, + "step": 20599 + }, + { + "epoch": 1.651500270589886, + "grad_norm": 1.6584148406982422, + "learning_rate": 1.5490067729221969e-06, + "loss": 3.1281, + "step": 20600 + }, + { + "epoch": 1.6515804453709086, + "grad_norm": 1.7268402576446533, + "learning_rate": 1.5483126822547023e-06, + "loss": 3.217, + "step": 20601 + }, + { + "epoch": 1.6516606201519313, + "grad_norm": 1.8160568475723267, + "learning_rate": 1.5476187340796745e-06, + "loss": 3.1581, + "step": 20602 + }, + { + "epoch": 1.651740794932954, + "grad_norm": 1.8273212909698486, + "learning_rate": 1.5469249284088162e-06, + "loss": 3.177, + "step": 20603 + }, + { + "epoch": 1.6518209697139765, + "grad_norm": 1.7625774145126343, + "learning_rate": 1.5462312652538248e-06, + "loss": 3.1587, + "step": 20604 + }, + { + "epoch": 1.6519011444949991, + "grad_norm": 1.7474792003631592, + "learning_rate": 1.5455377446263942e-06, + "loss": 3.2067, + "step": 20605 + }, + { + "epoch": 1.6519813192760218, + "grad_norm": 1.7741643190383911, + "learning_rate": 1.5448443665382206e-06, + "loss": 3.1587, + "step": 20606 + }, + { + "epoch": 1.6520614940570444, + "grad_norm": 1.8093611001968384, + "learning_rate": 1.5441511310009883e-06, + "loss": 3.1571, + "step": 20607 + }, + { + "epoch": 1.652141668838067, + "grad_norm": 1.7489765882492065, + "learning_rate": 1.5434580380263886e-06, + "loss": 3.1351, + "step": 20608 + }, + { + "epoch": 1.6522218436190896, + "grad_norm": 1.8570988178253174, + "learning_rate": 1.5427650876261047e-06, + "loss": 3.2194, + "step": 20609 + }, + { + "epoch": 1.6523020184001123, + "grad_norm": 1.7979190349578857, + "learning_rate": 1.5420722798118215e-06, + "loss": 3.1378, + "step": 20610 + }, + { + "epoch": 1.6523821931811349, + "grad_norm": 1.7223050594329834, + "learning_rate": 1.541379614595221e-06, + "loss": 3.1602, + "step": 20611 + }, + { + "epoch": 1.6524623679621575, + "grad_norm": 1.819815993309021, + "learning_rate": 1.540687091987977e-06, + "loss": 3.1561, + "step": 20612 + }, + { + "epoch": 1.6525425427431801, + "grad_norm": 1.7017936706542969, + "learning_rate": 1.5399947120017677e-06, + "loss": 3.2183, + "step": 20613 + }, + { + "epoch": 1.6526227175242028, + "grad_norm": 1.741452693939209, + "learning_rate": 1.5393024746482653e-06, + "loss": 3.2004, + "step": 20614 + }, + { + "epoch": 1.6527028923052254, + "grad_norm": 1.8821828365325928, + "learning_rate": 1.5386103799391417e-06, + "loss": 3.1489, + "step": 20615 + }, + { + "epoch": 1.652783067086248, + "grad_norm": 1.764622449874878, + "learning_rate": 1.5379184278860671e-06, + "loss": 3.209, + "step": 20616 + }, + { + "epoch": 1.6528632418672706, + "grad_norm": 1.8185701370239258, + "learning_rate": 1.5372266185007022e-06, + "loss": 3.3005, + "step": 20617 + }, + { + "epoch": 1.6529434166482933, + "grad_norm": 1.9204245805740356, + "learning_rate": 1.5365349517947158e-06, + "loss": 3.2195, + "step": 20618 + }, + { + "epoch": 1.6530235914293159, + "grad_norm": 1.8179054260253906, + "learning_rate": 1.5358434277797663e-06, + "loss": 3.1924, + "step": 20619 + }, + { + "epoch": 1.6531037662103385, + "grad_norm": 1.7884944677352905, + "learning_rate": 1.5351520464675129e-06, + "loss": 3.2053, + "step": 20620 + }, + { + "epoch": 1.6531839409913611, + "grad_norm": 1.8218673467636108, + "learning_rate": 1.5344608078696167e-06, + "loss": 3.3011, + "step": 20621 + }, + { + "epoch": 1.6532641157723837, + "grad_norm": 1.7149596214294434, + "learning_rate": 1.5337697119977235e-06, + "loss": 3.1863, + "step": 20622 + }, + { + "epoch": 1.6533442905534064, + "grad_norm": 1.8106756210327148, + "learning_rate": 1.533078758863491e-06, + "loss": 3.2565, + "step": 20623 + }, + { + "epoch": 1.653424465334429, + "grad_norm": 1.7734304666519165, + "learning_rate": 1.5323879484785654e-06, + "loss": 3.1146, + "step": 20624 + }, + { + "epoch": 1.6535046401154516, + "grad_norm": 1.7901157140731812, + "learning_rate": 1.5316972808545948e-06, + "loss": 3.0989, + "step": 20625 + }, + { + "epoch": 1.6535848148964742, + "grad_norm": 1.7418245077133179, + "learning_rate": 1.5310067560032261e-06, + "loss": 3.1396, + "step": 20626 + }, + { + "epoch": 1.6536649896774969, + "grad_norm": 1.7364016771316528, + "learning_rate": 1.530316373936096e-06, + "loss": 3.1508, + "step": 20627 + }, + { + "epoch": 1.6537451644585195, + "grad_norm": 1.8016682863235474, + "learning_rate": 1.5296261346648477e-06, + "loss": 3.2269, + "step": 20628 + }, + { + "epoch": 1.6538253392395421, + "grad_norm": 1.6360037326812744, + "learning_rate": 1.5289360382011165e-06, + "loss": 3.0849, + "step": 20629 + }, + { + "epoch": 1.6539055140205647, + "grad_norm": 1.7293530702590942, + "learning_rate": 1.5282460845565395e-06, + "loss": 3.1713, + "step": 20630 + }, + { + "epoch": 1.6539856888015874, + "grad_norm": 1.7756743431091309, + "learning_rate": 1.5275562737427485e-06, + "loss": 3.2738, + "step": 20631 + }, + { + "epoch": 1.65406586358261, + "grad_norm": 1.8393871784210205, + "learning_rate": 1.5268666057713722e-06, + "loss": 3.2192, + "step": 20632 + }, + { + "epoch": 1.6541460383636326, + "grad_norm": 1.9208000898361206, + "learning_rate": 1.5261770806540377e-06, + "loss": 3.1611, + "step": 20633 + }, + { + "epoch": 1.6542262131446552, + "grad_norm": 1.7925310134887695, + "learning_rate": 1.5254876984023715e-06, + "loss": 3.1941, + "step": 20634 + }, + { + "epoch": 1.6543063879256779, + "grad_norm": 1.7764075994491577, + "learning_rate": 1.524798459027995e-06, + "loss": 3.1705, + "step": 20635 + }, + { + "epoch": 1.6543865627067005, + "grad_norm": 1.7052156925201416, + "learning_rate": 1.5241093625425329e-06, + "loss": 3.1248, + "step": 20636 + }, + { + "epoch": 1.654466737487723, + "grad_norm": 1.8634058237075806, + "learning_rate": 1.5234204089575966e-06, + "loss": 3.1121, + "step": 20637 + }, + { + "epoch": 1.6545469122687457, + "grad_norm": 1.7636408805847168, + "learning_rate": 1.5227315982848056e-06, + "loss": 3.2137, + "step": 20638 + }, + { + "epoch": 1.6546270870497684, + "grad_norm": 1.8239270448684692, + "learning_rate": 1.522042930535772e-06, + "loss": 3.1682, + "step": 20639 + }, + { + "epoch": 1.654707261830791, + "grad_norm": 1.8009543418884277, + "learning_rate": 1.5213544057221064e-06, + "loss": 3.178, + "step": 20640 + }, + { + "epoch": 1.6547874366118136, + "grad_norm": 1.812605381011963, + "learning_rate": 1.5206660238554194e-06, + "loss": 3.1426, + "step": 20641 + }, + { + "epoch": 1.6548676113928362, + "grad_norm": 1.8818644285202026, + "learning_rate": 1.519977784947313e-06, + "loss": 3.2607, + "step": 20642 + }, + { + "epoch": 1.6549477861738588, + "grad_norm": 1.9595450162887573, + "learning_rate": 1.5192896890093923e-06, + "loss": 3.2035, + "step": 20643 + }, + { + "epoch": 1.6550279609548817, + "grad_norm": 1.880513310432434, + "learning_rate": 1.518601736053259e-06, + "loss": 3.2118, + "step": 20644 + }, + { + "epoch": 1.6551081357359043, + "grad_norm": 1.758360743522644, + "learning_rate": 1.5179139260905108e-06, + "loss": 3.1689, + "step": 20645 + }, + { + "epoch": 1.655188310516927, + "grad_norm": 1.7832667827606201, + "learning_rate": 1.517226259132747e-06, + "loss": 3.1177, + "step": 20646 + }, + { + "epoch": 1.6552684852979496, + "grad_norm": 1.6999338865280151, + "learning_rate": 1.516538735191556e-06, + "loss": 3.1495, + "step": 20647 + }, + { + "epoch": 1.6553486600789722, + "grad_norm": 1.7651020288467407, + "learning_rate": 1.5158513542785325e-06, + "loss": 3.0672, + "step": 20648 + }, + { + "epoch": 1.6554288348599948, + "grad_norm": 1.7259511947631836, + "learning_rate": 1.5151641164052654e-06, + "loss": 3.157, + "step": 20649 + }, + { + "epoch": 1.6555090096410174, + "grad_norm": 1.7114578485488892, + "learning_rate": 1.5144770215833404e-06, + "loss": 3.187, + "step": 20650 + }, + { + "epoch": 1.65558918442204, + "grad_norm": 1.7783973217010498, + "learning_rate": 1.5137900698243446e-06, + "loss": 3.2137, + "step": 20651 + }, + { + "epoch": 1.6556693592030627, + "grad_norm": 1.705014944076538, + "learning_rate": 1.5131032611398556e-06, + "loss": 3.0862, + "step": 20652 + }, + { + "epoch": 1.6557495339840853, + "grad_norm": 1.7531168460845947, + "learning_rate": 1.5124165955414538e-06, + "loss": 3.2052, + "step": 20653 + }, + { + "epoch": 1.655829708765108, + "grad_norm": 1.7293187379837036, + "learning_rate": 1.5117300730407181e-06, + "loss": 3.1361, + "step": 20654 + }, + { + "epoch": 1.6559098835461306, + "grad_norm": 1.7580978870391846, + "learning_rate": 1.51104369364922e-06, + "loss": 3.2394, + "step": 20655 + }, + { + "epoch": 1.6559900583271532, + "grad_norm": 1.765112042427063, + "learning_rate": 1.5103574573785372e-06, + "loss": 3.1515, + "step": 20656 + }, + { + "epoch": 1.6560702331081758, + "grad_norm": 1.7905582189559937, + "learning_rate": 1.5096713642402327e-06, + "loss": 3.1678, + "step": 20657 + }, + { + "epoch": 1.6561504078891984, + "grad_norm": 1.8061485290527344, + "learning_rate": 1.508985414245877e-06, + "loss": 3.059, + "step": 20658 + }, + { + "epoch": 1.656230582670221, + "grad_norm": 1.8348336219787598, + "learning_rate": 1.5082996074070345e-06, + "loss": 3.2052, + "step": 20659 + }, + { + "epoch": 1.6563107574512437, + "grad_norm": 1.7560150623321533, + "learning_rate": 1.5076139437352676e-06, + "loss": 3.2513, + "step": 20660 + }, + { + "epoch": 1.6563909322322663, + "grad_norm": 1.7126132249832153, + "learning_rate": 1.50692842324214e-06, + "loss": 3.1127, + "step": 20661 + }, + { + "epoch": 1.656471107013289, + "grad_norm": 1.767810583114624, + "learning_rate": 1.5062430459392018e-06, + "loss": 3.1546, + "step": 20662 + }, + { + "epoch": 1.6565512817943115, + "grad_norm": 1.715257167816162, + "learning_rate": 1.5055578118380144e-06, + "loss": 3.1327, + "step": 20663 + }, + { + "epoch": 1.6566314565753344, + "grad_norm": 1.7857733964920044, + "learning_rate": 1.5048727209501267e-06, + "loss": 3.2041, + "step": 20664 + }, + { + "epoch": 1.656711631356357, + "grad_norm": 1.8668485879898071, + "learning_rate": 1.5041877732870924e-06, + "loss": 3.1789, + "step": 20665 + }, + { + "epoch": 1.6567918061373796, + "grad_norm": 1.832200050354004, + "learning_rate": 1.5035029688604597e-06, + "loss": 3.194, + "step": 20666 + }, + { + "epoch": 1.6568719809184023, + "grad_norm": 1.7539633512496948, + "learning_rate": 1.5028183076817703e-06, + "loss": 3.0514, + "step": 20667 + }, + { + "epoch": 1.656952155699425, + "grad_norm": 0.6646159291267395, + "learning_rate": 1.5021337897625688e-06, + "loss": 2.458, + "step": 20668 + }, + { + "epoch": 1.6570323304804475, + "grad_norm": 1.7637234926223755, + "learning_rate": 1.5014494151143977e-06, + "loss": 3.1935, + "step": 20669 + }, + { + "epoch": 1.6571125052614701, + "grad_norm": 1.670335054397583, + "learning_rate": 1.5007651837487935e-06, + "loss": 3.1317, + "step": 20670 + }, + { + "epoch": 1.6571926800424928, + "grad_norm": 1.8836777210235596, + "learning_rate": 1.500081095677296e-06, + "loss": 3.1559, + "step": 20671 + }, + { + "epoch": 1.6572728548235154, + "grad_norm": 1.6399493217468262, + "learning_rate": 1.4993971509114325e-06, + "loss": 3.1427, + "step": 20672 + }, + { + "epoch": 1.657353029604538, + "grad_norm": 1.7904926538467407, + "learning_rate": 1.498713349462737e-06, + "loss": 3.1555, + "step": 20673 + }, + { + "epoch": 1.6574332043855606, + "grad_norm": 1.7727742195129395, + "learning_rate": 1.4980296913427395e-06, + "loss": 3.2599, + "step": 20674 + }, + { + "epoch": 1.6575133791665833, + "grad_norm": 1.7374615669250488, + "learning_rate": 1.4973461765629638e-06, + "loss": 3.1703, + "step": 20675 + }, + { + "epoch": 1.6575935539476059, + "grad_norm": 1.803145408630371, + "learning_rate": 1.4966628051349386e-06, + "loss": 3.1835, + "step": 20676 + }, + { + "epoch": 1.6576737287286285, + "grad_norm": 1.8214117288589478, + "learning_rate": 1.495979577070178e-06, + "loss": 3.1957, + "step": 20677 + }, + { + "epoch": 1.6577539035096511, + "grad_norm": 1.6978938579559326, + "learning_rate": 1.495296492380205e-06, + "loss": 3.0646, + "step": 20678 + }, + { + "epoch": 1.6578340782906738, + "grad_norm": 1.6947462558746338, + "learning_rate": 1.4946135510765358e-06, + "loss": 3.1188, + "step": 20679 + }, + { + "epoch": 1.6579142530716964, + "grad_norm": 1.8127853870391846, + "learning_rate": 1.4939307531706848e-06, + "loss": 3.0479, + "step": 20680 + }, + { + "epoch": 1.657994427852719, + "grad_norm": 1.8526030778884888, + "learning_rate": 1.493248098674165e-06, + "loss": 3.1639, + "step": 20681 + }, + { + "epoch": 1.6580746026337416, + "grad_norm": 0.6505349278450012, + "learning_rate": 1.4925655875984823e-06, + "loss": 2.4725, + "step": 20682 + }, + { + "epoch": 1.6581547774147642, + "grad_norm": 1.8351786136627197, + "learning_rate": 1.4918832199551458e-06, + "loss": 3.203, + "step": 20683 + }, + { + "epoch": 1.6582349521957869, + "grad_norm": 1.726070761680603, + "learning_rate": 1.4912009957556594e-06, + "loss": 3.2048, + "step": 20684 + }, + { + "epoch": 1.6583151269768095, + "grad_norm": 1.7722227573394775, + "learning_rate": 1.4905189150115263e-06, + "loss": 3.1814, + "step": 20685 + }, + { + "epoch": 1.6583953017578321, + "grad_norm": 1.8987176418304443, + "learning_rate": 1.489836977734247e-06, + "loss": 3.15, + "step": 20686 + }, + { + "epoch": 1.6584754765388547, + "grad_norm": 0.6754077076911926, + "learning_rate": 1.4891551839353135e-06, + "loss": 2.4613, + "step": 20687 + }, + { + "epoch": 1.6585556513198774, + "grad_norm": 1.7840873003005981, + "learning_rate": 1.4884735336262257e-06, + "loss": 3.2266, + "step": 20688 + }, + { + "epoch": 1.6586358261009, + "grad_norm": 0.6653422117233276, + "learning_rate": 1.4877920268184732e-06, + "loss": 2.4535, + "step": 20689 + }, + { + "epoch": 1.6587160008819226, + "grad_norm": 0.6592947840690613, + "learning_rate": 1.4871106635235476e-06, + "loss": 2.4626, + "step": 20690 + }, + { + "epoch": 1.6587961756629452, + "grad_norm": 1.9514493942260742, + "learning_rate": 1.4864294437529393e-06, + "loss": 3.1861, + "step": 20691 + }, + { + "epoch": 1.6588763504439679, + "grad_norm": 1.8010629415512085, + "learning_rate": 1.485748367518126e-06, + "loss": 3.21, + "step": 20692 + }, + { + "epoch": 1.6589565252249905, + "grad_norm": 1.7252992391586304, + "learning_rate": 1.485067434830596e-06, + "loss": 3.1412, + "step": 20693 + }, + { + "epoch": 1.659036700006013, + "grad_norm": 1.8854024410247803, + "learning_rate": 1.4843866457018274e-06, + "loss": 3.1137, + "step": 20694 + }, + { + "epoch": 1.6591168747870357, + "grad_norm": 1.714200735092163, + "learning_rate": 1.483706000143298e-06, + "loss": 3.0909, + "step": 20695 + }, + { + "epoch": 1.6591970495680584, + "grad_norm": 1.8204622268676758, + "learning_rate": 1.483025498166487e-06, + "loss": 3.1723, + "step": 20696 + }, + { + "epoch": 1.659277224349081, + "grad_norm": 1.7861251831054688, + "learning_rate": 1.4823451397828625e-06, + "loss": 3.1375, + "step": 20697 + }, + { + "epoch": 1.6593573991301036, + "grad_norm": 1.7628518342971802, + "learning_rate": 1.4816649250038973e-06, + "loss": 3.2471, + "step": 20698 + }, + { + "epoch": 1.6594375739111262, + "grad_norm": 1.755977988243103, + "learning_rate": 1.4809848538410587e-06, + "loss": 3.2139, + "step": 20699 + }, + { + "epoch": 1.6595177486921489, + "grad_norm": 1.8547546863555908, + "learning_rate": 1.4803049263058133e-06, + "loss": 3.1384, + "step": 20700 + }, + { + "epoch": 1.6595979234731715, + "grad_norm": 1.669791340827942, + "learning_rate": 1.4796251424096263e-06, + "loss": 3.1157, + "step": 20701 + }, + { + "epoch": 1.659678098254194, + "grad_norm": 1.7327152490615845, + "learning_rate": 1.4789455021639543e-06, + "loss": 3.1914, + "step": 20702 + }, + { + "epoch": 1.6597582730352167, + "grad_norm": 1.8572522401809692, + "learning_rate": 1.4782660055802577e-06, + "loss": 3.2224, + "step": 20703 + }, + { + "epoch": 1.6598384478162393, + "grad_norm": 1.8240702152252197, + "learning_rate": 1.4775866526699934e-06, + "loss": 3.057, + "step": 20704 + }, + { + "epoch": 1.659918622597262, + "grad_norm": 1.711771011352539, + "learning_rate": 1.476907443444614e-06, + "loss": 3.2086, + "step": 20705 + }, + { + "epoch": 1.6599987973782846, + "grad_norm": 1.720201849937439, + "learning_rate": 1.4762283779155739e-06, + "loss": 3.1403, + "step": 20706 + }, + { + "epoch": 1.6600789721593072, + "grad_norm": 1.8075456619262695, + "learning_rate": 1.4755494560943173e-06, + "loss": 3.2748, + "step": 20707 + }, + { + "epoch": 1.6601591469403298, + "grad_norm": 1.6996545791625977, + "learning_rate": 1.4748706779922916e-06, + "loss": 3.0942, + "step": 20708 + }, + { + "epoch": 1.6602393217213525, + "grad_norm": 1.7345412969589233, + "learning_rate": 1.4741920436209434e-06, + "loss": 3.3009, + "step": 20709 + }, + { + "epoch": 1.660319496502375, + "grad_norm": 1.7407811880111694, + "learning_rate": 1.473513552991711e-06, + "loss": 3.101, + "step": 20710 + }, + { + "epoch": 1.6603996712833977, + "grad_norm": 0.651770830154419, + "learning_rate": 1.4728352061160377e-06, + "loss": 2.4556, + "step": 20711 + }, + { + "epoch": 1.6604798460644203, + "grad_norm": 1.6778303384780884, + "learning_rate": 1.4721570030053556e-06, + "loss": 3.1176, + "step": 20712 + }, + { + "epoch": 1.660560020845443, + "grad_norm": 1.815857172012329, + "learning_rate": 1.4714789436711008e-06, + "loss": 3.1731, + "step": 20713 + }, + { + "epoch": 1.6606401956264656, + "grad_norm": 0.6984689831733704, + "learning_rate": 1.4708010281247054e-06, + "loss": 2.5071, + "step": 20714 + }, + { + "epoch": 1.6607203704074882, + "grad_norm": 1.8815864324569702, + "learning_rate": 1.470123256377599e-06, + "loss": 3.228, + "step": 20715 + }, + { + "epoch": 1.6608005451885108, + "grad_norm": 1.9825376272201538, + "learning_rate": 1.4694456284412095e-06, + "loss": 3.2611, + "step": 20716 + }, + { + "epoch": 1.6608807199695335, + "grad_norm": 0.6653544306755066, + "learning_rate": 1.4687681443269586e-06, + "loss": 2.4574, + "step": 20717 + }, + { + "epoch": 1.660960894750556, + "grad_norm": 1.751867413520813, + "learning_rate": 1.4680908040462705e-06, + "loss": 3.0972, + "step": 20718 + }, + { + "epoch": 1.6610410695315787, + "grad_norm": 1.7869964838027954, + "learning_rate": 1.4674136076105638e-06, + "loss": 3.1675, + "step": 20719 + }, + { + "epoch": 1.6611212443126013, + "grad_norm": 1.7372758388519287, + "learning_rate": 1.4667365550312574e-06, + "loss": 3.2352, + "step": 20720 + }, + { + "epoch": 1.661201419093624, + "grad_norm": 1.97640860080719, + "learning_rate": 1.4660596463197663e-06, + "loss": 3.2578, + "step": 20721 + }, + { + "epoch": 1.6612815938746466, + "grad_norm": 2.07155442237854, + "learning_rate": 1.4653828814875003e-06, + "loss": 3.1917, + "step": 20722 + }, + { + "epoch": 1.6613617686556692, + "grad_norm": 1.7706629037857056, + "learning_rate": 1.4647062605458706e-06, + "loss": 3.1589, + "step": 20723 + }, + { + "epoch": 1.6614419434366918, + "grad_norm": 1.7110273838043213, + "learning_rate": 1.4640297835062856e-06, + "loss": 3.2163, + "step": 20724 + }, + { + "epoch": 1.6615221182177147, + "grad_norm": 0.6649085283279419, + "learning_rate": 1.4633534503801505e-06, + "loss": 2.4686, + "step": 20725 + }, + { + "epoch": 1.6616022929987373, + "grad_norm": 1.8814094066619873, + "learning_rate": 1.4626772611788698e-06, + "loss": 3.2199, + "step": 20726 + }, + { + "epoch": 1.66168246777976, + "grad_norm": 1.8088825941085815, + "learning_rate": 1.4620012159138386e-06, + "loss": 3.2352, + "step": 20727 + }, + { + "epoch": 1.6617626425607825, + "grad_norm": 1.7749427556991577, + "learning_rate": 1.461325314596458e-06, + "loss": 3.1378, + "step": 20728 + }, + { + "epoch": 1.6618428173418052, + "grad_norm": 1.7659684419631958, + "learning_rate": 1.4606495572381242e-06, + "loss": 3.115, + "step": 20729 + }, + { + "epoch": 1.6619229921228278, + "grad_norm": 0.6633031368255615, + "learning_rate": 1.459973943850228e-06, + "loss": 2.4713, + "step": 20730 + }, + { + "epoch": 1.6620031669038504, + "grad_norm": 1.8637137413024902, + "learning_rate": 1.4592984744441651e-06, + "loss": 3.2078, + "step": 20731 + }, + { + "epoch": 1.662083341684873, + "grad_norm": 1.8304564952850342, + "learning_rate": 1.4586231490313162e-06, + "loss": 3.2469, + "step": 20732 + }, + { + "epoch": 1.6621635164658957, + "grad_norm": 1.8595633506774902, + "learning_rate": 1.4579479676230712e-06, + "loss": 3.1653, + "step": 20733 + }, + { + "epoch": 1.6622436912469183, + "grad_norm": 1.7900359630584717, + "learning_rate": 1.4572729302308132e-06, + "loss": 3.1264, + "step": 20734 + }, + { + "epoch": 1.662323866027941, + "grad_norm": 1.7606627941131592, + "learning_rate": 1.4565980368659227e-06, + "loss": 3.1522, + "step": 20735 + }, + { + "epoch": 1.6624040408089635, + "grad_norm": 1.7670598030090332, + "learning_rate": 1.45592328753978e-06, + "loss": 3.2059, + "step": 20736 + }, + { + "epoch": 1.6624842155899862, + "grad_norm": 1.6802116632461548, + "learning_rate": 1.4552486822637578e-06, + "loss": 3.0746, + "step": 20737 + }, + { + "epoch": 1.6625643903710088, + "grad_norm": 1.7852197885513306, + "learning_rate": 1.4545742210492319e-06, + "loss": 3.1202, + "step": 20738 + }, + { + "epoch": 1.6626445651520314, + "grad_norm": 1.7193670272827148, + "learning_rate": 1.4538999039075718e-06, + "loss": 3.0724, + "step": 20739 + }, + { + "epoch": 1.662724739933054, + "grad_norm": 1.7271308898925781, + "learning_rate": 1.4532257308501484e-06, + "loss": 3.1577, + "step": 20740 + }, + { + "epoch": 1.6628049147140767, + "grad_norm": 1.8543341159820557, + "learning_rate": 1.4525517018883284e-06, + "loss": 3.2198, + "step": 20741 + }, + { + "epoch": 1.6628850894950993, + "grad_norm": 1.8347973823547363, + "learning_rate": 1.4518778170334723e-06, + "loss": 3.184, + "step": 20742 + }, + { + "epoch": 1.662965264276122, + "grad_norm": 1.6933379173278809, + "learning_rate": 1.4512040762969436e-06, + "loss": 3.0811, + "step": 20743 + }, + { + "epoch": 1.6630454390571445, + "grad_norm": 1.7825983762741089, + "learning_rate": 1.4505304796901021e-06, + "loss": 3.0745, + "step": 20744 + }, + { + "epoch": 1.6631256138381671, + "grad_norm": 1.7863038778305054, + "learning_rate": 1.449857027224303e-06, + "loss": 3.1799, + "step": 20745 + }, + { + "epoch": 1.66320578861919, + "grad_norm": 1.7173466682434082, + "learning_rate": 1.4491837189109027e-06, + "loss": 3.2223, + "step": 20746 + }, + { + "epoch": 1.6632859634002126, + "grad_norm": 1.829524278640747, + "learning_rate": 1.4485105547612488e-06, + "loss": 3.1409, + "step": 20747 + }, + { + "epoch": 1.6633661381812352, + "grad_norm": 1.7308439016342163, + "learning_rate": 1.4478375347866947e-06, + "loss": 3.1471, + "step": 20748 + }, + { + "epoch": 1.6634463129622579, + "grad_norm": 1.7655483484268188, + "learning_rate": 1.4471646589985843e-06, + "loss": 3.1808, + "step": 20749 + }, + { + "epoch": 1.6635264877432805, + "grad_norm": 1.7758591175079346, + "learning_rate": 1.4464919274082645e-06, + "loss": 3.2607, + "step": 20750 + }, + { + "epoch": 1.6636066625243031, + "grad_norm": 1.8087959289550781, + "learning_rate": 1.445819340027077e-06, + "loss": 3.2755, + "step": 20751 + }, + { + "epoch": 1.6636868373053257, + "grad_norm": 1.7840709686279297, + "learning_rate": 1.4451468968663595e-06, + "loss": 3.0956, + "step": 20752 + }, + { + "epoch": 1.6637670120863484, + "grad_norm": 1.7287400960922241, + "learning_rate": 1.4444745979374507e-06, + "loss": 3.1028, + "step": 20753 + }, + { + "epoch": 1.663847186867371, + "grad_norm": 1.7732431888580322, + "learning_rate": 1.443802443251684e-06, + "loss": 3.0956, + "step": 20754 + }, + { + "epoch": 1.6639273616483936, + "grad_norm": 0.6581438779830933, + "learning_rate": 1.4431304328203932e-06, + "loss": 2.4481, + "step": 20755 + }, + { + "epoch": 1.6640075364294162, + "grad_norm": 1.826427936553955, + "learning_rate": 1.4424585666549096e-06, + "loss": 3.1178, + "step": 20756 + }, + { + "epoch": 1.6640877112104389, + "grad_norm": 1.7996301651000977, + "learning_rate": 1.4417868447665562e-06, + "loss": 3.1879, + "step": 20757 + }, + { + "epoch": 1.6641678859914615, + "grad_norm": 1.7377792596817017, + "learning_rate": 1.4411152671666606e-06, + "loss": 3.216, + "step": 20758 + }, + { + "epoch": 1.664248060772484, + "grad_norm": 1.845831274986267, + "learning_rate": 1.4404438338665449e-06, + "loss": 3.1717, + "step": 20759 + }, + { + "epoch": 1.6643282355535067, + "grad_norm": 1.8085930347442627, + "learning_rate": 1.439772544877529e-06, + "loss": 3.1376, + "step": 20760 + }, + { + "epoch": 1.6644084103345294, + "grad_norm": 1.6406461000442505, + "learning_rate": 1.4391014002109348e-06, + "loss": 3.077, + "step": 20761 + }, + { + "epoch": 1.664488585115552, + "grad_norm": 1.8621736764907837, + "learning_rate": 1.4384303998780703e-06, + "loss": 3.2842, + "step": 20762 + }, + { + "epoch": 1.6645687598965746, + "grad_norm": 1.6562615633010864, + "learning_rate": 1.4377595438902525e-06, + "loss": 3.1814, + "step": 20763 + }, + { + "epoch": 1.6646489346775972, + "grad_norm": 1.7306773662567139, + "learning_rate": 1.4370888322587905e-06, + "loss": 3.2396, + "step": 20764 + }, + { + "epoch": 1.6647291094586198, + "grad_norm": 1.8191488981246948, + "learning_rate": 1.4364182649949942e-06, + "loss": 3.2524, + "step": 20765 + }, + { + "epoch": 1.6648092842396425, + "grad_norm": 1.8754805326461792, + "learning_rate": 1.435747842110169e-06, + "loss": 3.1389, + "step": 20766 + }, + { + "epoch": 1.664889459020665, + "grad_norm": 0.656557559967041, + "learning_rate": 1.4350775636156145e-06, + "loss": 2.4877, + "step": 20767 + }, + { + "epoch": 1.6649696338016877, + "grad_norm": 1.8264769315719604, + "learning_rate": 1.434407429522634e-06, + "loss": 3.1808, + "step": 20768 + }, + { + "epoch": 1.6650498085827103, + "grad_norm": 1.7338300943374634, + "learning_rate": 1.4337374398425264e-06, + "loss": 3.06, + "step": 20769 + }, + { + "epoch": 1.665129983363733, + "grad_norm": 1.7911142110824585, + "learning_rate": 1.4330675945865856e-06, + "loss": 3.0842, + "step": 20770 + }, + { + "epoch": 1.6652101581447556, + "grad_norm": 2.0353424549102783, + "learning_rate": 1.4323978937661075e-06, + "loss": 3.1511, + "step": 20771 + }, + { + "epoch": 1.6652903329257782, + "grad_norm": 1.7947138547897339, + "learning_rate": 1.4317283373923808e-06, + "loss": 3.1949, + "step": 20772 + }, + { + "epoch": 1.6653705077068008, + "grad_norm": 1.7541494369506836, + "learning_rate": 1.4310589254766927e-06, + "loss": 3.2362, + "step": 20773 + }, + { + "epoch": 1.6654506824878235, + "grad_norm": 1.862466812133789, + "learning_rate": 1.4303896580303323e-06, + "loss": 3.2035, + "step": 20774 + }, + { + "epoch": 1.665530857268846, + "grad_norm": 2.021771192550659, + "learning_rate": 1.4297205350645815e-06, + "loss": 3.2726, + "step": 20775 + }, + { + "epoch": 1.6656110320498687, + "grad_norm": 1.8064885139465332, + "learning_rate": 1.4290515565907248e-06, + "loss": 3.1635, + "step": 20776 + }, + { + "epoch": 1.6656912068308913, + "grad_norm": 1.868042230606079, + "learning_rate": 1.4283827226200353e-06, + "loss": 3.1603, + "step": 20777 + }, + { + "epoch": 1.665771381611914, + "grad_norm": 0.665851354598999, + "learning_rate": 1.427714033163793e-06, + "loss": 2.4745, + "step": 20778 + }, + { + "epoch": 1.6658515563929366, + "grad_norm": 1.6662627458572388, + "learning_rate": 1.4270454882332707e-06, + "loss": 3.0968, + "step": 20779 + }, + { + "epoch": 1.6659317311739592, + "grad_norm": 1.720821499824524, + "learning_rate": 1.4263770878397398e-06, + "loss": 3.2144, + "step": 20780 + }, + { + "epoch": 1.6660119059549818, + "grad_norm": 1.8120110034942627, + "learning_rate": 1.4257088319944712e-06, + "loss": 3.0974, + "step": 20781 + }, + { + "epoch": 1.6660920807360045, + "grad_norm": 1.6471450328826904, + "learning_rate": 1.4250407207087292e-06, + "loss": 3.2279, + "step": 20782 + }, + { + "epoch": 1.666172255517027, + "grad_norm": 1.7852866649627686, + "learning_rate": 1.4243727539937768e-06, + "loss": 3.1831, + "step": 20783 + }, + { + "epoch": 1.6662524302980497, + "grad_norm": 1.8444823026657104, + "learning_rate": 1.4237049318608786e-06, + "loss": 3.1116, + "step": 20784 + }, + { + "epoch": 1.6663326050790723, + "grad_norm": 1.8526750802993774, + "learning_rate": 1.4230372543212923e-06, + "loss": 3.2191, + "step": 20785 + }, + { + "epoch": 1.666412779860095, + "grad_norm": 1.7473610639572144, + "learning_rate": 1.4223697213862774e-06, + "loss": 3.0718, + "step": 20786 + }, + { + "epoch": 1.6664929546411176, + "grad_norm": 1.8307489156723022, + "learning_rate": 1.421702333067083e-06, + "loss": 3.1296, + "step": 20787 + }, + { + "epoch": 1.6665731294221402, + "grad_norm": 1.7009795904159546, + "learning_rate": 1.4210350893749647e-06, + "loss": 3.157, + "step": 20788 + }, + { + "epoch": 1.6666533042031628, + "grad_norm": 1.809729814529419, + "learning_rate": 1.4203679903211709e-06, + "loss": 3.0729, + "step": 20789 + }, + { + "epoch": 1.6667334789841854, + "grad_norm": 1.7872605323791504, + "learning_rate": 1.4197010359169494e-06, + "loss": 3.2092, + "step": 20790 + }, + { + "epoch": 1.666813653765208, + "grad_norm": 1.8112627267837524, + "learning_rate": 1.4190342261735468e-06, + "loss": 3.1748, + "step": 20791 + }, + { + "epoch": 1.6668938285462307, + "grad_norm": 1.841533899307251, + "learning_rate": 1.4183675611022007e-06, + "loss": 3.2102, + "step": 20792 + }, + { + "epoch": 1.6669740033272533, + "grad_norm": 1.878257155418396, + "learning_rate": 1.4177010407141523e-06, + "loss": 3.2524, + "step": 20793 + }, + { + "epoch": 1.667054178108276, + "grad_norm": 1.809053659439087, + "learning_rate": 1.41703466502064e-06, + "loss": 3.2117, + "step": 20794 + }, + { + "epoch": 1.6671343528892986, + "grad_norm": 1.7911192178726196, + "learning_rate": 1.4163684340328987e-06, + "loss": 3.1588, + "step": 20795 + }, + { + "epoch": 1.6672145276703212, + "grad_norm": 1.698230504989624, + "learning_rate": 1.4157023477621624e-06, + "loss": 3.1929, + "step": 20796 + }, + { + "epoch": 1.6672947024513438, + "grad_norm": 1.769291639328003, + "learning_rate": 1.4150364062196565e-06, + "loss": 3.1525, + "step": 20797 + }, + { + "epoch": 1.6673748772323664, + "grad_norm": 1.70107102394104, + "learning_rate": 1.414370609416612e-06, + "loss": 3.1375, + "step": 20798 + }, + { + "epoch": 1.667455052013389, + "grad_norm": 1.7896071672439575, + "learning_rate": 1.4137049573642525e-06, + "loss": 3.2722, + "step": 20799 + }, + { + "epoch": 1.6675352267944117, + "grad_norm": 1.7106447219848633, + "learning_rate": 1.413039450073802e-06, + "loss": 3.1919, + "step": 20800 + }, + { + "epoch": 1.6676154015754343, + "grad_norm": 1.8847264051437378, + "learning_rate": 1.4123740875564817e-06, + "loss": 3.3469, + "step": 20801 + }, + { + "epoch": 1.667695576356457, + "grad_norm": 1.7522554397583008, + "learning_rate": 1.4117088698235048e-06, + "loss": 3.0978, + "step": 20802 + }, + { + "epoch": 1.6677757511374796, + "grad_norm": 1.7882193326950073, + "learning_rate": 1.4110437968860913e-06, + "loss": 3.1633, + "step": 20803 + }, + { + "epoch": 1.6678559259185022, + "grad_norm": 1.7943319082260132, + "learning_rate": 1.410378868755451e-06, + "loss": 3.117, + "step": 20804 + }, + { + "epoch": 1.6679361006995248, + "grad_norm": 1.7211761474609375, + "learning_rate": 1.409714085442796e-06, + "loss": 3.152, + "step": 20805 + }, + { + "epoch": 1.6680162754805474, + "grad_norm": 1.7322033643722534, + "learning_rate": 1.4090494469593364e-06, + "loss": 3.2205, + "step": 20806 + }, + { + "epoch": 1.6680964502615703, + "grad_norm": 1.7263952493667603, + "learning_rate": 1.408384953316273e-06, + "loss": 3.1978, + "step": 20807 + }, + { + "epoch": 1.668176625042593, + "grad_norm": 1.9064668416976929, + "learning_rate": 1.4077206045248115e-06, + "loss": 3.1392, + "step": 20808 + }, + { + "epoch": 1.6682567998236155, + "grad_norm": 1.7196588516235352, + "learning_rate": 1.4070564005961518e-06, + "loss": 3.1439, + "step": 20809 + }, + { + "epoch": 1.6683369746046381, + "grad_norm": 1.8610206842422485, + "learning_rate": 1.4063923415414938e-06, + "loss": 3.1399, + "step": 20810 + }, + { + "epoch": 1.6684171493856608, + "grad_norm": 1.6482800245285034, + "learning_rate": 1.405728427372034e-06, + "loss": 3.1367, + "step": 20811 + }, + { + "epoch": 1.6684973241666834, + "grad_norm": 1.9058226346969604, + "learning_rate": 1.4050646580989612e-06, + "loss": 3.2435, + "step": 20812 + }, + { + "epoch": 1.668577498947706, + "grad_norm": 1.739202618598938, + "learning_rate": 1.404401033733469e-06, + "loss": 3.1338, + "step": 20813 + }, + { + "epoch": 1.6686576737287286, + "grad_norm": 1.622628927230835, + "learning_rate": 1.403737554286747e-06, + "loss": 3.168, + "step": 20814 + }, + { + "epoch": 1.6687378485097513, + "grad_norm": 1.945619821548462, + "learning_rate": 1.4030742197699798e-06, + "loss": 3.068, + "step": 20815 + }, + { + "epoch": 1.6688180232907739, + "grad_norm": 1.8840495347976685, + "learning_rate": 1.402411030194354e-06, + "loss": 3.1877, + "step": 20816 + }, + { + "epoch": 1.6688981980717965, + "grad_norm": 1.7320533990859985, + "learning_rate": 1.4017479855710448e-06, + "loss": 3.1368, + "step": 20817 + }, + { + "epoch": 1.6689783728528191, + "grad_norm": 1.7739759683609009, + "learning_rate": 1.4010850859112347e-06, + "loss": 3.1862, + "step": 20818 + }, + { + "epoch": 1.6690585476338418, + "grad_norm": 1.7785444259643555, + "learning_rate": 1.4004223312261e-06, + "loss": 3.1426, + "step": 20819 + }, + { + "epoch": 1.6691387224148644, + "grad_norm": 0.6641748547554016, + "learning_rate": 1.3997597215268143e-06, + "loss": 2.4706, + "step": 20820 + }, + { + "epoch": 1.669218897195887, + "grad_norm": 1.7896376848220825, + "learning_rate": 1.3990972568245498e-06, + "loss": 3.2535, + "step": 20821 + }, + { + "epoch": 1.6692990719769096, + "grad_norm": 1.674290657043457, + "learning_rate": 1.3984349371304727e-06, + "loss": 3.1571, + "step": 20822 + }, + { + "epoch": 1.6693792467579323, + "grad_norm": 1.7419922351837158, + "learning_rate": 1.3977727624557512e-06, + "loss": 3.1287, + "step": 20823 + }, + { + "epoch": 1.6694594215389549, + "grad_norm": 1.7488152980804443, + "learning_rate": 1.3971107328115485e-06, + "loss": 3.1638, + "step": 20824 + }, + { + "epoch": 1.6695395963199775, + "grad_norm": 1.7471158504486084, + "learning_rate": 1.3964488482090277e-06, + "loss": 3.205, + "step": 20825 + }, + { + "epoch": 1.6696197711010001, + "grad_norm": 1.8828028440475464, + "learning_rate": 1.395787108659349e-06, + "loss": 3.1739, + "step": 20826 + }, + { + "epoch": 1.669699945882023, + "grad_norm": 1.7167555093765259, + "learning_rate": 1.3951255141736653e-06, + "loss": 3.2269, + "step": 20827 + }, + { + "epoch": 1.6697801206630456, + "grad_norm": 1.666283369064331, + "learning_rate": 1.3944640647631325e-06, + "loss": 3.1337, + "step": 20828 + }, + { + "epoch": 1.6698602954440682, + "grad_norm": 1.8088905811309814, + "learning_rate": 1.3938027604389026e-06, + "loss": 3.1821, + "step": 20829 + }, + { + "epoch": 1.6699404702250908, + "grad_norm": 1.872148036956787, + "learning_rate": 1.3931416012121257e-06, + "loss": 3.1886, + "step": 20830 + }, + { + "epoch": 1.6700206450061135, + "grad_norm": 1.7425273656845093, + "learning_rate": 1.3924805870939495e-06, + "loss": 3.2421, + "step": 20831 + }, + { + "epoch": 1.670100819787136, + "grad_norm": 1.834041714668274, + "learning_rate": 1.3918197180955152e-06, + "loss": 3.1004, + "step": 20832 + }, + { + "epoch": 1.6701809945681587, + "grad_norm": 1.8001635074615479, + "learning_rate": 1.3911589942279658e-06, + "loss": 3.0742, + "step": 20833 + }, + { + "epoch": 1.6702611693491813, + "grad_norm": 1.7550660371780396, + "learning_rate": 1.3904984155024425e-06, + "loss": 3.1464, + "step": 20834 + }, + { + "epoch": 1.670341344130204, + "grad_norm": 1.722107172012329, + "learning_rate": 1.3898379819300822e-06, + "loss": 3.0547, + "step": 20835 + }, + { + "epoch": 1.6704215189112266, + "grad_norm": 1.704276204109192, + "learning_rate": 1.3891776935220203e-06, + "loss": 3.1208, + "step": 20836 + }, + { + "epoch": 1.6705016936922492, + "grad_norm": 1.8158859014511108, + "learning_rate": 1.3885175502893855e-06, + "loss": 3.2373, + "step": 20837 + }, + { + "epoch": 1.6705818684732718, + "grad_norm": 0.6684439778327942, + "learning_rate": 1.3878575522433103e-06, + "loss": 2.4705, + "step": 20838 + }, + { + "epoch": 1.6706620432542945, + "grad_norm": 1.8204104900360107, + "learning_rate": 1.387197699394921e-06, + "loss": 3.1863, + "step": 20839 + }, + { + "epoch": 1.670742218035317, + "grad_norm": 1.7472293376922607, + "learning_rate": 1.3865379917553434e-06, + "loss": 3.1483, + "step": 20840 + }, + { + "epoch": 1.6708223928163397, + "grad_norm": 1.7821869850158691, + "learning_rate": 1.385878429335702e-06, + "loss": 3.1985, + "step": 20841 + }, + { + "epoch": 1.6709025675973623, + "grad_norm": 0.6663165092468262, + "learning_rate": 1.385219012147112e-06, + "loss": 2.4723, + "step": 20842 + }, + { + "epoch": 1.670982742378385, + "grad_norm": 1.815705418586731, + "learning_rate": 1.3845597402006938e-06, + "loss": 3.2095, + "step": 20843 + }, + { + "epoch": 1.6710629171594076, + "grad_norm": 1.6765763759613037, + "learning_rate": 1.3839006135075606e-06, + "loss": 3.1613, + "step": 20844 + }, + { + "epoch": 1.6711430919404302, + "grad_norm": 1.6881334781646729, + "learning_rate": 1.3832416320788278e-06, + "loss": 3.1831, + "step": 20845 + }, + { + "epoch": 1.6712232667214528, + "grad_norm": 1.7049050331115723, + "learning_rate": 1.3825827959256055e-06, + "loss": 3.2141, + "step": 20846 + }, + { + "epoch": 1.6713034415024755, + "grad_norm": 1.7956339120864868, + "learning_rate": 1.3819241050589981e-06, + "loss": 3.1124, + "step": 20847 + }, + { + "epoch": 1.671383616283498, + "grad_norm": 1.7027045488357544, + "learning_rate": 1.3812655594901137e-06, + "loss": 3.2175, + "step": 20848 + }, + { + "epoch": 1.6714637910645207, + "grad_norm": 1.8218259811401367, + "learning_rate": 1.3806071592300541e-06, + "loss": 3.1637, + "step": 20849 + }, + { + "epoch": 1.6715439658455433, + "grad_norm": 1.795326590538025, + "learning_rate": 1.3799489042899206e-06, + "loss": 3.1992, + "step": 20850 + }, + { + "epoch": 1.671624140626566, + "grad_norm": 1.7524224519729614, + "learning_rate": 1.379290794680812e-06, + "loss": 3.0878, + "step": 20851 + }, + { + "epoch": 1.6717043154075886, + "grad_norm": 1.7345004081726074, + "learning_rate": 1.3786328304138197e-06, + "loss": 3.1194, + "step": 20852 + }, + { + "epoch": 1.6717844901886112, + "grad_norm": 1.7944459915161133, + "learning_rate": 1.37797501150004e-06, + "loss": 3.2354, + "step": 20853 + }, + { + "epoch": 1.6718646649696338, + "grad_norm": 0.6650057435035706, + "learning_rate": 1.3773173379505633e-06, + "loss": 2.4672, + "step": 20854 + }, + { + "epoch": 1.6719448397506564, + "grad_norm": 1.7300910949707031, + "learning_rate": 1.376659809776477e-06, + "loss": 3.0759, + "step": 20855 + }, + { + "epoch": 1.672025014531679, + "grad_norm": 1.7847000360488892, + "learning_rate": 1.3760024269888694e-06, + "loss": 3.1738, + "step": 20856 + }, + { + "epoch": 1.6721051893127017, + "grad_norm": 1.7535173892974854, + "learning_rate": 1.375345189598819e-06, + "loss": 3.2244, + "step": 20857 + }, + { + "epoch": 1.6721853640937243, + "grad_norm": 1.7758824825286865, + "learning_rate": 1.3746880976174094e-06, + "loss": 3.0643, + "step": 20858 + }, + { + "epoch": 1.672265538874747, + "grad_norm": 1.7594356536865234, + "learning_rate": 1.3740311510557192e-06, + "loss": 3.2264, + "step": 20859 + }, + { + "epoch": 1.6723457136557696, + "grad_norm": 1.7773370742797852, + "learning_rate": 1.373374349924822e-06, + "loss": 3.1222, + "step": 20860 + }, + { + "epoch": 1.6724258884367922, + "grad_norm": 1.7828705310821533, + "learning_rate": 1.3727176942357945e-06, + "loss": 3.1791, + "step": 20861 + }, + { + "epoch": 1.6725060632178148, + "grad_norm": 1.776047945022583, + "learning_rate": 1.3720611839997077e-06, + "loss": 3.1512, + "step": 20862 + }, + { + "epoch": 1.6725862379988374, + "grad_norm": 1.8292014598846436, + "learning_rate": 1.3714048192276263e-06, + "loss": 3.1305, + "step": 20863 + }, + { + "epoch": 1.67266641277986, + "grad_norm": 1.8423060178756714, + "learning_rate": 1.370748599930619e-06, + "loss": 3.0992, + "step": 20864 + }, + { + "epoch": 1.6727465875608827, + "grad_norm": 1.6968703269958496, + "learning_rate": 1.3700925261197484e-06, + "loss": 3.1826, + "step": 20865 + }, + { + "epoch": 1.6728267623419053, + "grad_norm": 1.7957191467285156, + "learning_rate": 1.3694365978060775e-06, + "loss": 3.228, + "step": 20866 + }, + { + "epoch": 1.672906937122928, + "grad_norm": 1.917631983757019, + "learning_rate": 1.3687808150006654e-06, + "loss": 3.1921, + "step": 20867 + }, + { + "epoch": 1.6729871119039506, + "grad_norm": 1.876584529876709, + "learning_rate": 1.3681251777145642e-06, + "loss": 3.215, + "step": 20868 + }, + { + "epoch": 1.6730672866849732, + "grad_norm": 1.8959383964538574, + "learning_rate": 1.367469685958831e-06, + "loss": 3.1144, + "step": 20869 + }, + { + "epoch": 1.6731474614659958, + "grad_norm": 1.9213807582855225, + "learning_rate": 1.3668143397445166e-06, + "loss": 3.2206, + "step": 20870 + }, + { + "epoch": 1.6732276362470184, + "grad_norm": 1.8397119045257568, + "learning_rate": 1.366159139082669e-06, + "loss": 3.1996, + "step": 20871 + }, + { + "epoch": 1.673307811028041, + "grad_norm": 1.7878522872924805, + "learning_rate": 1.365504083984338e-06, + "loss": 3.181, + "step": 20872 + }, + { + "epoch": 1.6733879858090637, + "grad_norm": 1.8175456523895264, + "learning_rate": 1.364849174460563e-06, + "loss": 3.1455, + "step": 20873 + }, + { + "epoch": 1.6734681605900863, + "grad_norm": 1.8670711517333984, + "learning_rate": 1.364194410522387e-06, + "loss": 3.1597, + "step": 20874 + }, + { + "epoch": 1.673548335371109, + "grad_norm": 1.7514839172363281, + "learning_rate": 1.3635397921808503e-06, + "loss": 3.0807, + "step": 20875 + }, + { + "epoch": 1.6736285101521315, + "grad_norm": 1.8943294286727905, + "learning_rate": 1.3628853194469894e-06, + "loss": 3.1703, + "step": 20876 + }, + { + "epoch": 1.6737086849331542, + "grad_norm": 1.705987811088562, + "learning_rate": 1.362230992331839e-06, + "loss": 3.228, + "step": 20877 + }, + { + "epoch": 1.6737888597141768, + "grad_norm": 1.7716118097305298, + "learning_rate": 1.361576810846429e-06, + "loss": 3.221, + "step": 20878 + }, + { + "epoch": 1.6738690344951994, + "grad_norm": 1.7710813283920288, + "learning_rate": 1.3609227750017884e-06, + "loss": 3.1638, + "step": 20879 + }, + { + "epoch": 1.673949209276222, + "grad_norm": 1.9418375492095947, + "learning_rate": 1.3602688848089452e-06, + "loss": 3.2324, + "step": 20880 + }, + { + "epoch": 1.6740293840572447, + "grad_norm": 1.7366372346878052, + "learning_rate": 1.3596151402789237e-06, + "loss": 3.0898, + "step": 20881 + }, + { + "epoch": 1.6741095588382673, + "grad_norm": 1.7419986724853516, + "learning_rate": 1.3589615414227486e-06, + "loss": 3.1116, + "step": 20882 + }, + { + "epoch": 1.67418973361929, + "grad_norm": 1.7291444540023804, + "learning_rate": 1.3583080882514344e-06, + "loss": 3.2378, + "step": 20883 + }, + { + "epoch": 1.6742699084003125, + "grad_norm": 1.8052401542663574, + "learning_rate": 1.3576547807759989e-06, + "loss": 3.072, + "step": 20884 + }, + { + "epoch": 1.6743500831813352, + "grad_norm": 1.7009565830230713, + "learning_rate": 1.3570016190074587e-06, + "loss": 3.0796, + "step": 20885 + }, + { + "epoch": 1.6744302579623578, + "grad_norm": 1.8013750314712524, + "learning_rate": 1.356348602956825e-06, + "loss": 3.1776, + "step": 20886 + }, + { + "epoch": 1.6745104327433804, + "grad_norm": 1.710403561592102, + "learning_rate": 1.3556957326351106e-06, + "loss": 3.1725, + "step": 20887 + }, + { + "epoch": 1.6745906075244033, + "grad_norm": 1.7889784574508667, + "learning_rate": 1.3550430080533162e-06, + "loss": 3.0598, + "step": 20888 + }, + { + "epoch": 1.6746707823054259, + "grad_norm": 1.8399235010147095, + "learning_rate": 1.3543904292224508e-06, + "loss": 3.1587, + "step": 20889 + }, + { + "epoch": 1.6747509570864485, + "grad_norm": 1.6179473400115967, + "learning_rate": 1.3537379961535156e-06, + "loss": 3.1104, + "step": 20890 + }, + { + "epoch": 1.6748311318674711, + "grad_norm": 1.6888738870620728, + "learning_rate": 1.35308570885751e-06, + "loss": 3.1132, + "step": 20891 + }, + { + "epoch": 1.6749113066484937, + "grad_norm": 1.8143049478530884, + "learning_rate": 1.3524335673454347e-06, + "loss": 3.1019, + "step": 20892 + }, + { + "epoch": 1.6749914814295164, + "grad_norm": 1.78717041015625, + "learning_rate": 1.351781571628279e-06, + "loss": 3.1981, + "step": 20893 + }, + { + "epoch": 1.675071656210539, + "grad_norm": 1.6814199686050415, + "learning_rate": 1.3511297217170382e-06, + "loss": 3.1538, + "step": 20894 + }, + { + "epoch": 1.6751518309915616, + "grad_norm": 1.7688599824905396, + "learning_rate": 1.3504780176227018e-06, + "loss": 3.1046, + "step": 20895 + }, + { + "epoch": 1.6752320057725842, + "grad_norm": 1.7084083557128906, + "learning_rate": 1.3498264593562582e-06, + "loss": 3.1127, + "step": 20896 + }, + { + "epoch": 1.6753121805536069, + "grad_norm": 1.8210211992263794, + "learning_rate": 1.3491750469286946e-06, + "loss": 3.104, + "step": 20897 + }, + { + "epoch": 1.6753923553346295, + "grad_norm": 1.6661111116409302, + "learning_rate": 1.3485237803509875e-06, + "loss": 3.1244, + "step": 20898 + }, + { + "epoch": 1.6754725301156521, + "grad_norm": 1.7260005474090576, + "learning_rate": 1.3478726596341207e-06, + "loss": 3.1366, + "step": 20899 + }, + { + "epoch": 1.6755527048966747, + "grad_norm": 1.7543011903762817, + "learning_rate": 1.3472216847890718e-06, + "loss": 3.1388, + "step": 20900 + }, + { + "epoch": 1.6756328796776974, + "grad_norm": 1.8021414279937744, + "learning_rate": 1.3465708558268164e-06, + "loss": 3.1512, + "step": 20901 + }, + { + "epoch": 1.67571305445872, + "grad_norm": 1.8666638135910034, + "learning_rate": 1.3459201727583283e-06, + "loss": 3.1052, + "step": 20902 + }, + { + "epoch": 1.6757932292397426, + "grad_norm": 1.6960608959197998, + "learning_rate": 1.3452696355945726e-06, + "loss": 3.0809, + "step": 20903 + }, + { + "epoch": 1.6758734040207652, + "grad_norm": 1.7826727628707886, + "learning_rate": 1.3446192443465222e-06, + "loss": 3.3328, + "step": 20904 + }, + { + "epoch": 1.6759535788017879, + "grad_norm": 1.8449019193649292, + "learning_rate": 1.3439689990251404e-06, + "loss": 3.326, + "step": 20905 + }, + { + "epoch": 1.6760337535828105, + "grad_norm": 1.7796258926391602, + "learning_rate": 1.3433188996413905e-06, + "loss": 3.1926, + "step": 20906 + }, + { + "epoch": 1.676113928363833, + "grad_norm": 1.713598370552063, + "learning_rate": 1.3426689462062359e-06, + "loss": 3.1781, + "step": 20907 + }, + { + "epoch": 1.676194103144856, + "grad_norm": 1.734815001487732, + "learning_rate": 1.3420191387306292e-06, + "loss": 3.1493, + "step": 20908 + }, + { + "epoch": 1.6762742779258786, + "grad_norm": 1.6375010013580322, + "learning_rate": 1.3413694772255281e-06, + "loss": 3.233, + "step": 20909 + }, + { + "epoch": 1.6763544527069012, + "grad_norm": 1.7808908224105835, + "learning_rate": 1.3407199617018862e-06, + "loss": 3.2544, + "step": 20910 + }, + { + "epoch": 1.6764346274879238, + "grad_norm": 1.8720470666885376, + "learning_rate": 1.3400705921706536e-06, + "loss": 3.132, + "step": 20911 + }, + { + "epoch": 1.6765148022689464, + "grad_norm": 1.726609468460083, + "learning_rate": 1.3394213686427816e-06, + "loss": 3.1659, + "step": 20912 + }, + { + "epoch": 1.676594977049969, + "grad_norm": 1.8344188928604126, + "learning_rate": 1.33877229112921e-06, + "loss": 3.078, + "step": 20913 + }, + { + "epoch": 1.6766751518309917, + "grad_norm": 1.748769998550415, + "learning_rate": 1.338123359640886e-06, + "loss": 3.095, + "step": 20914 + }, + { + "epoch": 1.6767553266120143, + "grad_norm": 1.7816877365112305, + "learning_rate": 1.3374745741887486e-06, + "loss": 3.1712, + "step": 20915 + }, + { + "epoch": 1.676835501393037, + "grad_norm": 1.901626706123352, + "learning_rate": 1.3368259347837375e-06, + "loss": 3.1274, + "step": 20916 + }, + { + "epoch": 1.6769156761740596, + "grad_norm": 1.808854103088379, + "learning_rate": 1.3361774414367901e-06, + "loss": 3.1385, + "step": 20917 + }, + { + "epoch": 1.6769958509550822, + "grad_norm": 1.7797638177871704, + "learning_rate": 1.335529094158835e-06, + "loss": 3.1953, + "step": 20918 + }, + { + "epoch": 1.6770760257361048, + "grad_norm": 1.7591279745101929, + "learning_rate": 1.3348808929608058e-06, + "loss": 3.2008, + "step": 20919 + }, + { + "epoch": 1.6771562005171274, + "grad_norm": 1.8115178346633911, + "learning_rate": 1.3342328378536306e-06, + "loss": 3.2125, + "step": 20920 + }, + { + "epoch": 1.67723637529815, + "grad_norm": 1.7201800346374512, + "learning_rate": 1.3335849288482361e-06, + "loss": 3.15, + "step": 20921 + }, + { + "epoch": 1.6773165500791727, + "grad_norm": 1.8513206243515015, + "learning_rate": 1.3329371659555478e-06, + "loss": 3.1749, + "step": 20922 + }, + { + "epoch": 1.6773967248601953, + "grad_norm": 1.765563726425171, + "learning_rate": 1.3322895491864817e-06, + "loss": 3.111, + "step": 20923 + }, + { + "epoch": 1.677476899641218, + "grad_norm": 1.736392855644226, + "learning_rate": 1.3316420785519591e-06, + "loss": 3.0665, + "step": 20924 + }, + { + "epoch": 1.6775570744222406, + "grad_norm": 1.9152346849441528, + "learning_rate": 1.3309947540628965e-06, + "loss": 3.22, + "step": 20925 + }, + { + "epoch": 1.6776372492032632, + "grad_norm": 1.7116235494613647, + "learning_rate": 1.3303475757302064e-06, + "loss": 3.2041, + "step": 20926 + }, + { + "epoch": 1.6777174239842858, + "grad_norm": 1.7314897775650024, + "learning_rate": 1.3297005435648025e-06, + "loss": 3.1649, + "step": 20927 + }, + { + "epoch": 1.6777975987653084, + "grad_norm": 1.7688689231872559, + "learning_rate": 1.3290536575775903e-06, + "loss": 3.1884, + "step": 20928 + }, + { + "epoch": 1.677877773546331, + "grad_norm": 1.7565573453903198, + "learning_rate": 1.328406917779478e-06, + "loss": 3.1605, + "step": 20929 + }, + { + "epoch": 1.6779579483273537, + "grad_norm": 1.8979501724243164, + "learning_rate": 1.3277603241813674e-06, + "loss": 3.2358, + "step": 20930 + }, + { + "epoch": 1.6780381231083763, + "grad_norm": 1.7284492254257202, + "learning_rate": 1.3271138767941627e-06, + "loss": 3.2505, + "step": 20931 + }, + { + "epoch": 1.678118297889399, + "grad_norm": 1.7086474895477295, + "learning_rate": 1.3264675756287627e-06, + "loss": 3.2173, + "step": 20932 + }, + { + "epoch": 1.6781984726704215, + "grad_norm": 1.6704765558242798, + "learning_rate": 1.3258214206960606e-06, + "loss": 3.148, + "step": 20933 + }, + { + "epoch": 1.6782786474514442, + "grad_norm": 1.7814276218414307, + "learning_rate": 1.3251754120069527e-06, + "loss": 3.1885, + "step": 20934 + }, + { + "epoch": 1.6783588222324668, + "grad_norm": 1.9094783067703247, + "learning_rate": 1.32452954957233e-06, + "loss": 3.2197, + "step": 20935 + }, + { + "epoch": 1.6784389970134894, + "grad_norm": 1.7060993909835815, + "learning_rate": 1.3238838334030812e-06, + "loss": 3.2621, + "step": 20936 + }, + { + "epoch": 1.678519171794512, + "grad_norm": 1.7443779706954956, + "learning_rate": 1.323238263510095e-06, + "loss": 3.3104, + "step": 20937 + }, + { + "epoch": 1.6785993465755347, + "grad_norm": 1.6642014980316162, + "learning_rate": 1.322592839904252e-06, + "loss": 3.1525, + "step": 20938 + }, + { + "epoch": 1.6786795213565573, + "grad_norm": 1.6974014043807983, + "learning_rate": 1.3219475625964363e-06, + "loss": 3.1571, + "step": 20939 + }, + { + "epoch": 1.67875969613758, + "grad_norm": 1.9204248189926147, + "learning_rate": 1.3213024315975254e-06, + "loss": 3.2271, + "step": 20940 + }, + { + "epoch": 1.6788398709186025, + "grad_norm": 1.7015352249145508, + "learning_rate": 1.3206574469183964e-06, + "loss": 3.2273, + "step": 20941 + }, + { + "epoch": 1.6789200456996252, + "grad_norm": 1.8736741542816162, + "learning_rate": 1.3200126085699272e-06, + "loss": 3.1526, + "step": 20942 + }, + { + "epoch": 1.6790002204806478, + "grad_norm": 1.7975330352783203, + "learning_rate": 1.3193679165629835e-06, + "loss": 3.2756, + "step": 20943 + }, + { + "epoch": 1.6790803952616704, + "grad_norm": 1.8831804990768433, + "learning_rate": 1.3187233709084367e-06, + "loss": 3.213, + "step": 20944 + }, + { + "epoch": 1.679160570042693, + "grad_norm": 1.7149735689163208, + "learning_rate": 1.318078971617155e-06, + "loss": 3.1218, + "step": 20945 + }, + { + "epoch": 1.6792407448237157, + "grad_norm": 1.8031200170516968, + "learning_rate": 1.3174347187000025e-06, + "loss": 3.1329, + "step": 20946 + }, + { + "epoch": 1.6793209196047383, + "grad_norm": 1.7538719177246094, + "learning_rate": 1.316790612167842e-06, + "loss": 3.1558, + "step": 20947 + }, + { + "epoch": 1.679401094385761, + "grad_norm": 1.777742862701416, + "learning_rate": 1.3161466520315302e-06, + "loss": 3.0435, + "step": 20948 + }, + { + "epoch": 1.6794812691667835, + "grad_norm": 1.6569997072219849, + "learning_rate": 1.3155028383019241e-06, + "loss": 3.1721, + "step": 20949 + }, + { + "epoch": 1.6795614439478062, + "grad_norm": 1.769140362739563, + "learning_rate": 1.3148591709898806e-06, + "loss": 3.1437, + "step": 20950 + }, + { + "epoch": 1.6796416187288288, + "grad_norm": 1.8569384813308716, + "learning_rate": 1.3142156501062498e-06, + "loss": 3.0848, + "step": 20951 + }, + { + "epoch": 1.6797217935098514, + "grad_norm": 1.8268544673919678, + "learning_rate": 1.3135722756618852e-06, + "loss": 3.0834, + "step": 20952 + }, + { + "epoch": 1.679801968290874, + "grad_norm": 1.7222734689712524, + "learning_rate": 1.3129290476676271e-06, + "loss": 3.1178, + "step": 20953 + }, + { + "epoch": 1.6798821430718966, + "grad_norm": 1.7592507600784302, + "learning_rate": 1.3122859661343246e-06, + "loss": 3.0568, + "step": 20954 + }, + { + "epoch": 1.6799623178529193, + "grad_norm": 1.9166117906570435, + "learning_rate": 1.311643031072818e-06, + "loss": 3.1956, + "step": 20955 + }, + { + "epoch": 1.680042492633942, + "grad_norm": 1.7345826625823975, + "learning_rate": 1.3110002424939484e-06, + "loss": 3.1849, + "step": 20956 + }, + { + "epoch": 1.6801226674149645, + "grad_norm": 1.6624211072921753, + "learning_rate": 1.310357600408555e-06, + "loss": 3.1338, + "step": 20957 + }, + { + "epoch": 1.6802028421959871, + "grad_norm": 1.7517770528793335, + "learning_rate": 1.309715104827467e-06, + "loss": 3.0933, + "step": 20958 + }, + { + "epoch": 1.6802830169770098, + "grad_norm": 1.8048673868179321, + "learning_rate": 1.3090727557615202e-06, + "loss": 3.186, + "step": 20959 + }, + { + "epoch": 1.6803631917580324, + "grad_norm": 1.8465864658355713, + "learning_rate": 1.3084305532215436e-06, + "loss": 3.0931, + "step": 20960 + }, + { + "epoch": 1.680443366539055, + "grad_norm": 1.7197887897491455, + "learning_rate": 1.3077884972183641e-06, + "loss": 3.1264, + "step": 20961 + }, + { + "epoch": 1.6805235413200776, + "grad_norm": 1.7487998008728027, + "learning_rate": 1.3071465877628098e-06, + "loss": 3.1285, + "step": 20962 + }, + { + "epoch": 1.6806037161011003, + "grad_norm": 1.722076416015625, + "learning_rate": 1.3065048248656976e-06, + "loss": 3.1879, + "step": 20963 + }, + { + "epoch": 1.6806838908821229, + "grad_norm": 1.8048806190490723, + "learning_rate": 1.305863208537851e-06, + "loss": 3.2006, + "step": 20964 + }, + { + "epoch": 1.6807640656631455, + "grad_norm": 1.8349415063858032, + "learning_rate": 1.3052217387900855e-06, + "loss": 3.1866, + "step": 20965 + }, + { + "epoch": 1.6808442404441681, + "grad_norm": 1.8124821186065674, + "learning_rate": 1.3045804156332187e-06, + "loss": 3.1726, + "step": 20966 + }, + { + "epoch": 1.6809244152251908, + "grad_norm": 1.774894118309021, + "learning_rate": 1.3039392390780626e-06, + "loss": 3.1512, + "step": 20967 + }, + { + "epoch": 1.6810045900062134, + "grad_norm": 1.735059380531311, + "learning_rate": 1.303298209135424e-06, + "loss": 3.1183, + "step": 20968 + }, + { + "epoch": 1.6810847647872362, + "grad_norm": 1.7310760021209717, + "learning_rate": 1.3026573258161124e-06, + "loss": 3.1012, + "step": 20969 + }, + { + "epoch": 1.6811649395682589, + "grad_norm": 1.8101216554641724, + "learning_rate": 1.3020165891309345e-06, + "loss": 3.2278, + "step": 20970 + }, + { + "epoch": 1.6812451143492815, + "grad_norm": 1.87346613407135, + "learning_rate": 1.3013759990906893e-06, + "loss": 3.1286, + "step": 20971 + }, + { + "epoch": 1.681325289130304, + "grad_norm": 1.802565336227417, + "learning_rate": 1.300735555706183e-06, + "loss": 3.2196, + "step": 20972 + }, + { + "epoch": 1.6814054639113267, + "grad_norm": 1.7740579843521118, + "learning_rate": 1.3000952589882065e-06, + "loss": 3.1859, + "step": 20973 + }, + { + "epoch": 1.6814856386923493, + "grad_norm": 1.7656216621398926, + "learning_rate": 1.2994551089475571e-06, + "loss": 3.2506, + "step": 20974 + }, + { + "epoch": 1.681565813473372, + "grad_norm": 1.779714822769165, + "learning_rate": 1.2988151055950292e-06, + "loss": 3.1343, + "step": 20975 + }, + { + "epoch": 1.6816459882543946, + "grad_norm": 1.8579543828964233, + "learning_rate": 1.2981752489414122e-06, + "loss": 3.1699, + "step": 20976 + }, + { + "epoch": 1.6817261630354172, + "grad_norm": 1.7350136041641235, + "learning_rate": 1.2975355389974952e-06, + "loss": 3.1107, + "step": 20977 + }, + { + "epoch": 1.6818063378164398, + "grad_norm": 1.7632246017456055, + "learning_rate": 1.2968959757740595e-06, + "loss": 3.109, + "step": 20978 + }, + { + "epoch": 1.6818865125974625, + "grad_norm": 1.7692005634307861, + "learning_rate": 1.296256559281891e-06, + "loss": 3.1954, + "step": 20979 + }, + { + "epoch": 1.681966687378485, + "grad_norm": 1.8446747064590454, + "learning_rate": 1.2956172895317686e-06, + "loss": 3.2363, + "step": 20980 + }, + { + "epoch": 1.6820468621595077, + "grad_norm": 1.7322502136230469, + "learning_rate": 1.2949781665344717e-06, + "loss": 3.3216, + "step": 20981 + }, + { + "epoch": 1.6821270369405303, + "grad_norm": 1.8395895957946777, + "learning_rate": 1.2943391903007773e-06, + "loss": 3.2488, + "step": 20982 + }, + { + "epoch": 1.682207211721553, + "grad_norm": 1.837330937385559, + "learning_rate": 1.2937003608414534e-06, + "loss": 3.2893, + "step": 20983 + }, + { + "epoch": 1.6822873865025756, + "grad_norm": 1.7998390197753906, + "learning_rate": 1.2930616781672734e-06, + "loss": 3.0775, + "step": 20984 + }, + { + "epoch": 1.6823675612835982, + "grad_norm": 1.812223196029663, + "learning_rate": 1.2924231422890042e-06, + "loss": 3.0956, + "step": 20985 + }, + { + "epoch": 1.6824477360646208, + "grad_norm": 1.940659523010254, + "learning_rate": 1.2917847532174132e-06, + "loss": 3.2203, + "step": 20986 + }, + { + "epoch": 1.6825279108456435, + "grad_norm": 1.8582576513290405, + "learning_rate": 1.2911465109632636e-06, + "loss": 3.1724, + "step": 20987 + }, + { + "epoch": 1.682608085626666, + "grad_norm": 1.8079861402511597, + "learning_rate": 1.2905084155373116e-06, + "loss": 3.2212, + "step": 20988 + }, + { + "epoch": 1.6826882604076887, + "grad_norm": 1.7725794315338135, + "learning_rate": 1.2898704669503192e-06, + "loss": 3.2184, + "step": 20989 + }, + { + "epoch": 1.6827684351887116, + "grad_norm": 0.6661941409111023, + "learning_rate": 1.2892326652130405e-06, + "loss": 2.4747, + "step": 20990 + }, + { + "epoch": 1.6828486099697342, + "grad_norm": 1.7792952060699463, + "learning_rate": 1.2885950103362299e-06, + "loss": 3.1925, + "step": 20991 + }, + { + "epoch": 1.6829287847507568, + "grad_norm": 1.8485363721847534, + "learning_rate": 1.287957502330639e-06, + "loss": 3.1989, + "step": 20992 + }, + { + "epoch": 1.6830089595317794, + "grad_norm": 1.773497462272644, + "learning_rate": 1.2873201412070113e-06, + "loss": 3.1261, + "step": 20993 + }, + { + "epoch": 1.683089134312802, + "grad_norm": 1.8594905138015747, + "learning_rate": 1.286682926976096e-06, + "loss": 3.0638, + "step": 20994 + }, + { + "epoch": 1.6831693090938247, + "grad_norm": 1.7956866025924683, + "learning_rate": 1.2860458596486359e-06, + "loss": 3.1519, + "step": 20995 + }, + { + "epoch": 1.6832494838748473, + "grad_norm": 1.9003974199295044, + "learning_rate": 1.285408939235371e-06, + "loss": 3.2075, + "step": 20996 + }, + { + "epoch": 1.68332965865587, + "grad_norm": 1.7248889207839966, + "learning_rate": 1.2847721657470425e-06, + "loss": 3.1371, + "step": 20997 + }, + { + "epoch": 1.6834098334368925, + "grad_norm": 1.7850736379623413, + "learning_rate": 1.2841355391943811e-06, + "loss": 3.2672, + "step": 20998 + }, + { + "epoch": 1.6834900082179152, + "grad_norm": 1.8590816259384155, + "learning_rate": 1.2834990595881235e-06, + "loss": 3.1851, + "step": 20999 + }, + { + "epoch": 1.6835701829989378, + "grad_norm": 1.7641198635101318, + "learning_rate": 1.2828627269389993e-06, + "loss": 3.2834, + "step": 21000 + }, + { + "epoch": 1.6836503577799604, + "grad_norm": 1.788159728050232, + "learning_rate": 1.2822265412577373e-06, + "loss": 3.1166, + "step": 21001 + }, + { + "epoch": 1.683730532560983, + "grad_norm": 1.8054602146148682, + "learning_rate": 1.2815905025550658e-06, + "loss": 3.3141, + "step": 21002 + }, + { + "epoch": 1.6838107073420057, + "grad_norm": 1.7320269346237183, + "learning_rate": 1.2809546108417026e-06, + "loss": 3.2183, + "step": 21003 + }, + { + "epoch": 1.6838908821230283, + "grad_norm": 1.7906367778778076, + "learning_rate": 1.2803188661283728e-06, + "loss": 3.0627, + "step": 21004 + }, + { + "epoch": 1.683971056904051, + "grad_norm": 1.6759568452835083, + "learning_rate": 1.2796832684257942e-06, + "loss": 3.1357, + "step": 21005 + }, + { + "epoch": 1.6840512316850735, + "grad_norm": 1.7575961351394653, + "learning_rate": 1.2790478177446818e-06, + "loss": 3.1782, + "step": 21006 + }, + { + "epoch": 1.6841314064660962, + "grad_norm": 0.6407110095024109, + "learning_rate": 1.2784125140957516e-06, + "loss": 2.4673, + "step": 21007 + }, + { + "epoch": 1.6842115812471188, + "grad_norm": 1.6723159551620483, + "learning_rate": 1.2777773574897102e-06, + "loss": 3.193, + "step": 21008 + }, + { + "epoch": 1.6842917560281414, + "grad_norm": 1.7326406240463257, + "learning_rate": 1.277142347937269e-06, + "loss": 3.165, + "step": 21009 + }, + { + "epoch": 1.684371930809164, + "grad_norm": 1.8309404850006104, + "learning_rate": 1.2765074854491332e-06, + "loss": 3.3235, + "step": 21010 + }, + { + "epoch": 1.6844521055901867, + "grad_norm": 1.6898193359375, + "learning_rate": 1.2758727700360074e-06, + "loss": 3.1481, + "step": 21011 + }, + { + "epoch": 1.6845322803712093, + "grad_norm": 1.7536219358444214, + "learning_rate": 1.275238201708594e-06, + "loss": 3.1475, + "step": 21012 + }, + { + "epoch": 1.684612455152232, + "grad_norm": 1.8639918565750122, + "learning_rate": 1.2746037804775868e-06, + "loss": 3.2634, + "step": 21013 + }, + { + "epoch": 1.6846926299332545, + "grad_norm": 1.7070884704589844, + "learning_rate": 1.2739695063536862e-06, + "loss": 3.1685, + "step": 21014 + }, + { + "epoch": 1.6847728047142772, + "grad_norm": 1.8537609577178955, + "learning_rate": 1.2733353793475834e-06, + "loss": 3.0808, + "step": 21015 + }, + { + "epoch": 1.6848529794952998, + "grad_norm": 1.7125194072723389, + "learning_rate": 1.2727013994699699e-06, + "loss": 3.1877, + "step": 21016 + }, + { + "epoch": 1.6849331542763224, + "grad_norm": 1.5982662439346313, + "learning_rate": 1.2720675667315385e-06, + "loss": 3.0357, + "step": 21017 + }, + { + "epoch": 1.685013329057345, + "grad_norm": 1.7841894626617432, + "learning_rate": 1.2714338811429684e-06, + "loss": 3.1759, + "step": 21018 + }, + { + "epoch": 1.6850935038383676, + "grad_norm": 1.7883538007736206, + "learning_rate": 1.2708003427149474e-06, + "loss": 3.1078, + "step": 21019 + }, + { + "epoch": 1.6851736786193903, + "grad_norm": 1.9250006675720215, + "learning_rate": 1.270166951458156e-06, + "loss": 3.3119, + "step": 21020 + }, + { + "epoch": 1.685253853400413, + "grad_norm": 1.7153244018554688, + "learning_rate": 1.2695337073832736e-06, + "loss": 3.2341, + "step": 21021 + }, + { + "epoch": 1.6853340281814355, + "grad_norm": 1.8688480854034424, + "learning_rate": 1.2689006105009782e-06, + "loss": 3.2733, + "step": 21022 + }, + { + "epoch": 1.6854142029624581, + "grad_norm": 1.759761929512024, + "learning_rate": 1.2682676608219392e-06, + "loss": 3.1325, + "step": 21023 + }, + { + "epoch": 1.6854943777434808, + "grad_norm": 1.8210569620132446, + "learning_rate": 1.2676348583568287e-06, + "loss": 3.215, + "step": 21024 + }, + { + "epoch": 1.6855745525245034, + "grad_norm": 1.6943567991256714, + "learning_rate": 1.2670022031163187e-06, + "loss": 3.124, + "step": 21025 + }, + { + "epoch": 1.685654727305526, + "grad_norm": 1.7577604055404663, + "learning_rate": 1.2663696951110726e-06, + "loss": 3.178, + "step": 21026 + }, + { + "epoch": 1.6857349020865486, + "grad_norm": 1.9665781259536743, + "learning_rate": 1.2657373343517587e-06, + "loss": 3.198, + "step": 21027 + }, + { + "epoch": 1.6858150768675713, + "grad_norm": 1.729172945022583, + "learning_rate": 1.2651051208490316e-06, + "loss": 3.1349, + "step": 21028 + }, + { + "epoch": 1.6858952516485939, + "grad_norm": 1.7018176317214966, + "learning_rate": 1.2644730546135543e-06, + "loss": 3.2075, + "step": 21029 + }, + { + "epoch": 1.6859754264296165, + "grad_norm": 0.6622101068496704, + "learning_rate": 1.2638411356559833e-06, + "loss": 2.4668, + "step": 21030 + }, + { + "epoch": 1.6860556012106391, + "grad_norm": 1.8375301361083984, + "learning_rate": 1.2632093639869702e-06, + "loss": 3.1703, + "step": 21031 + }, + { + "epoch": 1.6861357759916618, + "grad_norm": 1.731371521949768, + "learning_rate": 1.2625777396171713e-06, + "loss": 3.0656, + "step": 21032 + }, + { + "epoch": 1.6862159507726844, + "grad_norm": 1.7750463485717773, + "learning_rate": 1.2619462625572298e-06, + "loss": 3.1956, + "step": 21033 + }, + { + "epoch": 1.686296125553707, + "grad_norm": 1.7010465860366821, + "learning_rate": 1.2613149328177953e-06, + "loss": 3.1824, + "step": 21034 + }, + { + "epoch": 1.6863763003347296, + "grad_norm": 1.68540358543396, + "learning_rate": 1.2606837504095116e-06, + "loss": 3.1516, + "step": 21035 + }, + { + "epoch": 1.6864564751157523, + "grad_norm": 1.7785604000091553, + "learning_rate": 1.2600527153430186e-06, + "loss": 3.1793, + "step": 21036 + }, + { + "epoch": 1.6865366498967749, + "grad_norm": 1.7900340557098389, + "learning_rate": 1.2594218276289593e-06, + "loss": 3.2595, + "step": 21037 + }, + { + "epoch": 1.6866168246777975, + "grad_norm": 1.7691024541854858, + "learning_rate": 1.2587910872779652e-06, + "loss": 3.2055, + "step": 21038 + }, + { + "epoch": 1.6866969994588201, + "grad_norm": 1.640712022781372, + "learning_rate": 1.2581604943006731e-06, + "loss": 3.1873, + "step": 21039 + }, + { + "epoch": 1.6867771742398427, + "grad_norm": 1.796615481376648, + "learning_rate": 1.2575300487077136e-06, + "loss": 3.2528, + "step": 21040 + }, + { + "epoch": 1.6868573490208654, + "grad_norm": 1.8788610696792603, + "learning_rate": 1.2568997505097168e-06, + "loss": 3.0858, + "step": 21041 + }, + { + "epoch": 1.686937523801888, + "grad_norm": 1.7098006010055542, + "learning_rate": 1.2562695997173102e-06, + "loss": 3.0856, + "step": 21042 + }, + { + "epoch": 1.6870176985829106, + "grad_norm": 1.7437788248062134, + "learning_rate": 1.255639596341115e-06, + "loss": 3.1989, + "step": 21043 + }, + { + "epoch": 1.6870978733639332, + "grad_norm": 1.6545594930648804, + "learning_rate": 1.2550097403917549e-06, + "loss": 3.1626, + "step": 21044 + }, + { + "epoch": 1.6871780481449559, + "grad_norm": 1.804848551750183, + "learning_rate": 1.254380031879847e-06, + "loss": 3.2605, + "step": 21045 + }, + { + "epoch": 1.6872582229259785, + "grad_norm": 1.7711176872253418, + "learning_rate": 1.2537504708160109e-06, + "loss": 3.2192, + "step": 21046 + }, + { + "epoch": 1.6873383977070011, + "grad_norm": 1.7268366813659668, + "learning_rate": 1.253121057210861e-06, + "loss": 3.1763, + "step": 21047 + }, + { + "epoch": 1.6874185724880237, + "grad_norm": 1.7496315240859985, + "learning_rate": 1.2524917910750057e-06, + "loss": 3.1459, + "step": 21048 + }, + { + "epoch": 1.6874987472690464, + "grad_norm": 1.7249327898025513, + "learning_rate": 1.2518626724190553e-06, + "loss": 3.2043, + "step": 21049 + }, + { + "epoch": 1.687578922050069, + "grad_norm": 1.8603757619857788, + "learning_rate": 1.251233701253618e-06, + "loss": 3.2104, + "step": 21050 + }, + { + "epoch": 1.6876590968310918, + "grad_norm": 1.7714641094207764, + "learning_rate": 1.2506048775892954e-06, + "loss": 3.1429, + "step": 21051 + }, + { + "epoch": 1.6877392716121145, + "grad_norm": 1.7444971799850464, + "learning_rate": 1.2499762014366945e-06, + "loss": 3.2516, + "step": 21052 + }, + { + "epoch": 1.687819446393137, + "grad_norm": 1.8597607612609863, + "learning_rate": 1.2493476728064092e-06, + "loss": 3.2711, + "step": 21053 + }, + { + "epoch": 1.6878996211741597, + "grad_norm": 1.8244577646255493, + "learning_rate": 1.2487192917090373e-06, + "loss": 3.0967, + "step": 21054 + }, + { + "epoch": 1.6879797959551823, + "grad_norm": 0.6908852458000183, + "learning_rate": 1.2480910581551742e-06, + "loss": 2.4866, + "step": 21055 + }, + { + "epoch": 1.688059970736205, + "grad_norm": 1.9184322357177734, + "learning_rate": 1.247462972155411e-06, + "loss": 3.2191, + "step": 21056 + }, + { + "epoch": 1.6881401455172276, + "grad_norm": 1.66104257106781, + "learning_rate": 1.2468350337203395e-06, + "loss": 3.0719, + "step": 21057 + }, + { + "epoch": 1.6882203202982502, + "grad_norm": 1.7928255796432495, + "learning_rate": 1.2462072428605432e-06, + "loss": 3.1262, + "step": 21058 + }, + { + "epoch": 1.6883004950792728, + "grad_norm": 1.7591487169265747, + "learning_rate": 1.2455795995866072e-06, + "loss": 3.1785, + "step": 21059 + }, + { + "epoch": 1.6883806698602954, + "grad_norm": 1.7265615463256836, + "learning_rate": 1.2449521039091128e-06, + "loss": 3.2744, + "step": 21060 + }, + { + "epoch": 1.688460844641318, + "grad_norm": 0.6447307467460632, + "learning_rate": 1.2443247558386418e-06, + "loss": 2.4622, + "step": 21061 + }, + { + "epoch": 1.6885410194223407, + "grad_norm": 1.7963312864303589, + "learning_rate": 1.2436975553857712e-06, + "loss": 3.1362, + "step": 21062 + }, + { + "epoch": 1.6886211942033633, + "grad_norm": 1.741267204284668, + "learning_rate": 1.2430705025610712e-06, + "loss": 3.1727, + "step": 21063 + }, + { + "epoch": 1.688701368984386, + "grad_norm": 1.7683659791946411, + "learning_rate": 1.242443597375117e-06, + "loss": 3.1707, + "step": 21064 + }, + { + "epoch": 1.6887815437654086, + "grad_norm": 1.7838387489318848, + "learning_rate": 1.2418168398384768e-06, + "loss": 3.1775, + "step": 21065 + }, + { + "epoch": 1.6888617185464312, + "grad_norm": 1.7859896421432495, + "learning_rate": 1.2411902299617186e-06, + "loss": 3.0746, + "step": 21066 + }, + { + "epoch": 1.6889418933274538, + "grad_norm": 1.7468702793121338, + "learning_rate": 1.2405637677554073e-06, + "loss": 3.0925, + "step": 21067 + }, + { + "epoch": 1.6890220681084764, + "grad_norm": 1.8877888917922974, + "learning_rate": 1.2399374532301022e-06, + "loss": 3.1794, + "step": 21068 + }, + { + "epoch": 1.689102242889499, + "grad_norm": 0.6619818210601807, + "learning_rate": 1.2393112863963642e-06, + "loss": 2.4486, + "step": 21069 + }, + { + "epoch": 1.6891824176705217, + "grad_norm": 1.8189080953598022, + "learning_rate": 1.2386852672647509e-06, + "loss": 3.2255, + "step": 21070 + }, + { + "epoch": 1.6892625924515445, + "grad_norm": 0.670813798904419, + "learning_rate": 1.2380593958458154e-06, + "loss": 2.462, + "step": 21071 + }, + { + "epoch": 1.6893427672325672, + "grad_norm": 1.816964030265808, + "learning_rate": 1.2374336721501135e-06, + "loss": 3.1599, + "step": 21072 + }, + { + "epoch": 1.6894229420135898, + "grad_norm": 1.8438557386398315, + "learning_rate": 1.236808096188189e-06, + "loss": 3.1541, + "step": 21073 + }, + { + "epoch": 1.6895031167946124, + "grad_norm": 1.8753033876419067, + "learning_rate": 1.2361826679705914e-06, + "loss": 3.2076, + "step": 21074 + }, + { + "epoch": 1.689583291575635, + "grad_norm": 1.8125083446502686, + "learning_rate": 1.2355573875078653e-06, + "loss": 3.1642, + "step": 21075 + }, + { + "epoch": 1.6896634663566577, + "grad_norm": 1.6585553884506226, + "learning_rate": 1.2349322548105524e-06, + "loss": 3.132, + "step": 21076 + }, + { + "epoch": 1.6897436411376803, + "grad_norm": 1.8104592561721802, + "learning_rate": 1.2343072698891955e-06, + "loss": 3.1615, + "step": 21077 + }, + { + "epoch": 1.689823815918703, + "grad_norm": 1.8120324611663818, + "learning_rate": 1.2336824327543262e-06, + "loss": 3.0614, + "step": 21078 + }, + { + "epoch": 1.6899039906997255, + "grad_norm": 1.8826935291290283, + "learning_rate": 1.2330577434164825e-06, + "loss": 3.234, + "step": 21079 + }, + { + "epoch": 1.6899841654807481, + "grad_norm": 1.9769401550292969, + "learning_rate": 1.232433201886194e-06, + "loss": 3.3488, + "step": 21080 + }, + { + "epoch": 1.6900643402617708, + "grad_norm": 1.7728520631790161, + "learning_rate": 1.231808808173992e-06, + "loss": 3.2094, + "step": 21081 + }, + { + "epoch": 1.6901445150427934, + "grad_norm": 1.7316445112228394, + "learning_rate": 1.2311845622904062e-06, + "loss": 3.1158, + "step": 21082 + }, + { + "epoch": 1.690224689823816, + "grad_norm": 1.7847568988800049, + "learning_rate": 1.2305604642459556e-06, + "loss": 3.2966, + "step": 21083 + }, + { + "epoch": 1.6903048646048386, + "grad_norm": 1.6270265579223633, + "learning_rate": 1.2299365140511654e-06, + "loss": 3.0392, + "step": 21084 + }, + { + "epoch": 1.6903850393858613, + "grad_norm": 2.0327036380767822, + "learning_rate": 1.2293127117165537e-06, + "loss": 3.1096, + "step": 21085 + }, + { + "epoch": 1.690465214166884, + "grad_norm": 1.874415397644043, + "learning_rate": 1.22868905725264e-06, + "loss": 3.1837, + "step": 21086 + }, + { + "epoch": 1.6905453889479065, + "grad_norm": 1.7679227590560913, + "learning_rate": 1.228065550669939e-06, + "loss": 3.1682, + "step": 21087 + }, + { + "epoch": 1.6906255637289291, + "grad_norm": 0.6676449775695801, + "learning_rate": 1.227442191978958e-06, + "loss": 2.4691, + "step": 21088 + }, + { + "epoch": 1.6907057385099518, + "grad_norm": 1.8298382759094238, + "learning_rate": 1.226818981190212e-06, + "loss": 3.166, + "step": 21089 + }, + { + "epoch": 1.6907859132909744, + "grad_norm": 1.7790422439575195, + "learning_rate": 1.2261959183142048e-06, + "loss": 3.1564, + "step": 21090 + }, + { + "epoch": 1.690866088071997, + "grad_norm": 1.7614479064941406, + "learning_rate": 1.225573003361442e-06, + "loss": 3.1717, + "step": 21091 + }, + { + "epoch": 1.6909462628530196, + "grad_norm": 1.701658010482788, + "learning_rate": 1.2249502363424281e-06, + "loss": 3.1392, + "step": 21092 + }, + { + "epoch": 1.6910264376340423, + "grad_norm": 0.6469077467918396, + "learning_rate": 1.2243276172676588e-06, + "loss": 2.4581, + "step": 21093 + }, + { + "epoch": 1.6911066124150649, + "grad_norm": 1.7789987325668335, + "learning_rate": 1.2237051461476335e-06, + "loss": 3.1219, + "step": 21094 + }, + { + "epoch": 1.6911867871960875, + "grad_norm": 1.8685272932052612, + "learning_rate": 1.223082822992846e-06, + "loss": 3.1185, + "step": 21095 + }, + { + "epoch": 1.6912669619771101, + "grad_norm": 1.851609230041504, + "learning_rate": 1.222460647813788e-06, + "loss": 3.1703, + "step": 21096 + }, + { + "epoch": 1.6913471367581328, + "grad_norm": 1.8514854907989502, + "learning_rate": 1.221838620620953e-06, + "loss": 3.1223, + "step": 21097 + }, + { + "epoch": 1.6914273115391554, + "grad_norm": 1.8486632108688354, + "learning_rate": 1.2212167414248233e-06, + "loss": 3.0571, + "step": 21098 + }, + { + "epoch": 1.691507486320178, + "grad_norm": 0.6678259968757629, + "learning_rate": 1.2205950102358844e-06, + "loss": 2.4745, + "step": 21099 + }, + { + "epoch": 1.6915876611012006, + "grad_norm": 1.6734132766723633, + "learning_rate": 1.2199734270646202e-06, + "loss": 3.1416, + "step": 21100 + }, + { + "epoch": 1.6916678358822232, + "grad_norm": 1.7070938348770142, + "learning_rate": 1.219351991921509e-06, + "loss": 3.1467, + "step": 21101 + }, + { + "epoch": 1.6917480106632459, + "grad_norm": 1.7436691522598267, + "learning_rate": 1.2187307048170315e-06, + "loss": 3.041, + "step": 21102 + }, + { + "epoch": 1.6918281854442685, + "grad_norm": 1.730668067932129, + "learning_rate": 1.218109565761657e-06, + "loss": 3.1589, + "step": 21103 + }, + { + "epoch": 1.6919083602252911, + "grad_norm": 1.6717270612716675, + "learning_rate": 1.2174885747658606e-06, + "loss": 3.1305, + "step": 21104 + }, + { + "epoch": 1.6919885350063137, + "grad_norm": 1.743713617324829, + "learning_rate": 1.21686773184011e-06, + "loss": 3.1585, + "step": 21105 + }, + { + "epoch": 1.6920687097873364, + "grad_norm": 1.8408712148666382, + "learning_rate": 1.2162470369948754e-06, + "loss": 3.174, + "step": 21106 + }, + { + "epoch": 1.692148884568359, + "grad_norm": 1.8315215110778809, + "learning_rate": 1.2156264902406224e-06, + "loss": 3.198, + "step": 21107 + }, + { + "epoch": 1.6922290593493816, + "grad_norm": 1.7826361656188965, + "learning_rate": 1.2150060915878082e-06, + "loss": 3.236, + "step": 21108 + }, + { + "epoch": 1.6923092341304042, + "grad_norm": 1.7368836402893066, + "learning_rate": 1.2143858410468945e-06, + "loss": 3.201, + "step": 21109 + }, + { + "epoch": 1.6923894089114269, + "grad_norm": 0.6557473540306091, + "learning_rate": 1.2137657386283408e-06, + "loss": 2.4615, + "step": 21110 + }, + { + "epoch": 1.6924695836924495, + "grad_norm": 1.7273914813995361, + "learning_rate": 1.2131457843425987e-06, + "loss": 3.1986, + "step": 21111 + }, + { + "epoch": 1.692549758473472, + "grad_norm": 1.771346092224121, + "learning_rate": 1.2125259782001243e-06, + "loss": 3.1448, + "step": 21112 + }, + { + "epoch": 1.6926299332544947, + "grad_norm": 1.670562744140625, + "learning_rate": 1.2119063202113634e-06, + "loss": 3.1361, + "step": 21113 + }, + { + "epoch": 1.6927101080355174, + "grad_norm": 1.6479382514953613, + "learning_rate": 1.2112868103867647e-06, + "loss": 3.1418, + "step": 21114 + }, + { + "epoch": 1.69279028281654, + "grad_norm": 1.8614552021026611, + "learning_rate": 1.210667448736772e-06, + "loss": 3.1894, + "step": 21115 + }, + { + "epoch": 1.6928704575975626, + "grad_norm": 0.6770567297935486, + "learning_rate": 1.210048235271829e-06, + "loss": 2.4487, + "step": 21116 + }, + { + "epoch": 1.6929506323785852, + "grad_norm": 1.7840596437454224, + "learning_rate": 1.2094291700023776e-06, + "loss": 3.1432, + "step": 21117 + }, + { + "epoch": 1.6930308071596079, + "grad_norm": 1.889509677886963, + "learning_rate": 1.208810252938849e-06, + "loss": 3.2889, + "step": 21118 + }, + { + "epoch": 1.6931109819406305, + "grad_norm": 1.6933513879776, + "learning_rate": 1.208191484091682e-06, + "loss": 3.1049, + "step": 21119 + }, + { + "epoch": 1.693191156721653, + "grad_norm": 1.8237667083740234, + "learning_rate": 1.207572863471308e-06, + "loss": 3.2336, + "step": 21120 + }, + { + "epoch": 1.6932713315026757, + "grad_norm": 1.7648427486419678, + "learning_rate": 1.2069543910881565e-06, + "loss": 3.1548, + "step": 21121 + }, + { + "epoch": 1.6933515062836983, + "grad_norm": 1.7268961668014526, + "learning_rate": 1.2063360669526581e-06, + "loss": 3.0359, + "step": 21122 + }, + { + "epoch": 1.693431681064721, + "grad_norm": 1.8371938467025757, + "learning_rate": 1.2057178910752309e-06, + "loss": 3.1622, + "step": 21123 + }, + { + "epoch": 1.6935118558457436, + "grad_norm": 1.795916199684143, + "learning_rate": 1.2050998634663013e-06, + "loss": 3.1144, + "step": 21124 + }, + { + "epoch": 1.6935920306267662, + "grad_norm": 1.6935614347457886, + "learning_rate": 1.2044819841362886e-06, + "loss": 3.0884, + "step": 21125 + }, + { + "epoch": 1.6936722054077888, + "grad_norm": 1.6734524965286255, + "learning_rate": 1.2038642530956091e-06, + "loss": 3.1711, + "step": 21126 + }, + { + "epoch": 1.6937523801888115, + "grad_norm": 0.6783542037010193, + "learning_rate": 1.2032466703546807e-06, + "loss": 2.4963, + "step": 21127 + }, + { + "epoch": 1.693832554969834, + "grad_norm": 1.7804286479949951, + "learning_rate": 1.2026292359239111e-06, + "loss": 3.1622, + "step": 21128 + }, + { + "epoch": 1.6939127297508567, + "grad_norm": 1.7871150970458984, + "learning_rate": 1.202011949813713e-06, + "loss": 3.2156, + "step": 21129 + }, + { + "epoch": 1.6939929045318793, + "grad_norm": 1.8770638704299927, + "learning_rate": 1.2013948120344908e-06, + "loss": 3.2665, + "step": 21130 + }, + { + "epoch": 1.694073079312902, + "grad_norm": 1.8206855058670044, + "learning_rate": 1.2007778225966526e-06, + "loss": 3.2, + "step": 21131 + }, + { + "epoch": 1.6941532540939248, + "grad_norm": 1.8213493824005127, + "learning_rate": 1.2001609815106008e-06, + "loss": 3.222, + "step": 21132 + }, + { + "epoch": 1.6942334288749474, + "grad_norm": 1.7659292221069336, + "learning_rate": 1.1995442887867315e-06, + "loss": 3.0793, + "step": 21133 + }, + { + "epoch": 1.69431360365597, + "grad_norm": 1.8343746662139893, + "learning_rate": 1.198927744435443e-06, + "loss": 3.1543, + "step": 21134 + }, + { + "epoch": 1.6943937784369927, + "grad_norm": 1.7817859649658203, + "learning_rate": 1.1983113484671326e-06, + "loss": 3.1252, + "step": 21135 + }, + { + "epoch": 1.6944739532180153, + "grad_norm": 1.846508502960205, + "learning_rate": 1.19769510089219e-06, + "loss": 3.1533, + "step": 21136 + }, + { + "epoch": 1.694554127999038, + "grad_norm": 1.7960612773895264, + "learning_rate": 1.1970790017210077e-06, + "loss": 3.1805, + "step": 21137 + }, + { + "epoch": 1.6946343027800606, + "grad_norm": 1.854714035987854, + "learning_rate": 1.1964630509639685e-06, + "loss": 3.3317, + "step": 21138 + }, + { + "epoch": 1.6947144775610832, + "grad_norm": 1.8086243867874146, + "learning_rate": 1.1958472486314609e-06, + "loss": 3.3234, + "step": 21139 + }, + { + "epoch": 1.6947946523421058, + "grad_norm": 1.7723314762115479, + "learning_rate": 1.1952315947338643e-06, + "loss": 3.1512, + "step": 21140 + }, + { + "epoch": 1.6948748271231284, + "grad_norm": 1.7565858364105225, + "learning_rate": 1.1946160892815605e-06, + "loss": 3.1004, + "step": 21141 + }, + { + "epoch": 1.694955001904151, + "grad_norm": 1.7008812427520752, + "learning_rate": 1.194000732284929e-06, + "loss": 3.0847, + "step": 21142 + }, + { + "epoch": 1.6950351766851737, + "grad_norm": 1.735295057296753, + "learning_rate": 1.193385523754339e-06, + "loss": 3.1085, + "step": 21143 + }, + { + "epoch": 1.6951153514661963, + "grad_norm": 1.8495503664016724, + "learning_rate": 1.1927704637001648e-06, + "loss": 3.1719, + "step": 21144 + }, + { + "epoch": 1.695195526247219, + "grad_norm": 1.7523550987243652, + "learning_rate": 1.1921555521327776e-06, + "loss": 3.2354, + "step": 21145 + }, + { + "epoch": 1.6952757010282415, + "grad_norm": 1.6711206436157227, + "learning_rate": 1.191540789062543e-06, + "loss": 3.1922, + "step": 21146 + }, + { + "epoch": 1.6953558758092642, + "grad_norm": 1.7848892211914062, + "learning_rate": 1.190926174499828e-06, + "loss": 3.2047, + "step": 21147 + }, + { + "epoch": 1.6954360505902868, + "grad_norm": 1.802539348602295, + "learning_rate": 1.190311708454991e-06, + "loss": 3.2588, + "step": 21148 + }, + { + "epoch": 1.6955162253713094, + "grad_norm": 1.7240033149719238, + "learning_rate": 1.1896973909383946e-06, + "loss": 3.1102, + "step": 21149 + }, + { + "epoch": 1.695596400152332, + "grad_norm": 1.8931244611740112, + "learning_rate": 1.1890832219603943e-06, + "loss": 3.1876, + "step": 21150 + }, + { + "epoch": 1.6956765749333547, + "grad_norm": 1.7810107469558716, + "learning_rate": 1.188469201531346e-06, + "loss": 3.1244, + "step": 21151 + }, + { + "epoch": 1.6957567497143773, + "grad_norm": 1.8560947179794312, + "learning_rate": 1.1878553296616036e-06, + "loss": 3.192, + "step": 21152 + }, + { + "epoch": 1.6958369244954001, + "grad_norm": 1.8174868822097778, + "learning_rate": 1.1872416063615132e-06, + "loss": 3.2073, + "step": 21153 + }, + { + "epoch": 1.6959170992764228, + "grad_norm": 1.7739243507385254, + "learning_rate": 1.1866280316414224e-06, + "loss": 3.1642, + "step": 21154 + }, + { + "epoch": 1.6959972740574454, + "grad_norm": 1.769359827041626, + "learning_rate": 1.1860146055116772e-06, + "loss": 3.1721, + "step": 21155 + }, + { + "epoch": 1.696077448838468, + "grad_norm": 1.7315882444381714, + "learning_rate": 1.1854013279826194e-06, + "loss": 3.1135, + "step": 21156 + }, + { + "epoch": 1.6961576236194906, + "grad_norm": 1.752746820449829, + "learning_rate": 1.1847881990645904e-06, + "loss": 3.1884, + "step": 21157 + }, + { + "epoch": 1.6962377984005133, + "grad_norm": 1.7482773065567017, + "learning_rate": 1.1841752187679244e-06, + "loss": 3.153, + "step": 21158 + }, + { + "epoch": 1.6963179731815359, + "grad_norm": 1.768797516822815, + "learning_rate": 1.1835623871029567e-06, + "loss": 3.1797, + "step": 21159 + }, + { + "epoch": 1.6963981479625585, + "grad_norm": 1.7573573589324951, + "learning_rate": 1.1829497040800197e-06, + "loss": 3.2209, + "step": 21160 + }, + { + "epoch": 1.6964783227435811, + "grad_norm": 1.747856855392456, + "learning_rate": 1.1823371697094432e-06, + "loss": 3.1204, + "step": 21161 + }, + { + "epoch": 1.6965584975246037, + "grad_norm": 1.7919464111328125, + "learning_rate": 1.1817247840015566e-06, + "loss": 3.0674, + "step": 21162 + }, + { + "epoch": 1.6966386723056264, + "grad_norm": 1.7455191612243652, + "learning_rate": 1.1811125469666806e-06, + "loss": 3.1475, + "step": 21163 + }, + { + "epoch": 1.696718847086649, + "grad_norm": 1.7031093835830688, + "learning_rate": 1.1805004586151381e-06, + "loss": 3.13, + "step": 21164 + }, + { + "epoch": 1.6967990218676716, + "grad_norm": 1.6943964958190918, + "learning_rate": 1.1798885189572506e-06, + "loss": 3.237, + "step": 21165 + }, + { + "epoch": 1.6968791966486942, + "grad_norm": 1.7527981996536255, + "learning_rate": 1.1792767280033334e-06, + "loss": 3.066, + "step": 21166 + }, + { + "epoch": 1.6969593714297169, + "grad_norm": 1.7674154043197632, + "learning_rate": 1.1786650857637039e-06, + "loss": 3.1034, + "step": 21167 + }, + { + "epoch": 1.6970395462107395, + "grad_norm": 1.7185742855072021, + "learning_rate": 1.1780535922486702e-06, + "loss": 3.1, + "step": 21168 + }, + { + "epoch": 1.6971197209917621, + "grad_norm": 1.7533683776855469, + "learning_rate": 1.1774422474685443e-06, + "loss": 3.1973, + "step": 21169 + }, + { + "epoch": 1.6971998957727847, + "grad_norm": 0.6648993492126465, + "learning_rate": 1.1768310514336324e-06, + "loss": 2.4718, + "step": 21170 + }, + { + "epoch": 1.6972800705538074, + "grad_norm": 1.600301742553711, + "learning_rate": 1.1762200041542394e-06, + "loss": 3.0776, + "step": 21171 + }, + { + "epoch": 1.69736024533483, + "grad_norm": 1.7578675746917725, + "learning_rate": 1.1756091056406694e-06, + "loss": 3.2632, + "step": 21172 + }, + { + "epoch": 1.6974404201158526, + "grad_norm": 1.8104486465454102, + "learning_rate": 1.1749983559032173e-06, + "loss": 3.1656, + "step": 21173 + }, + { + "epoch": 1.6975205948968752, + "grad_norm": 1.7425962686538696, + "learning_rate": 1.1743877549521832e-06, + "loss": 3.1619, + "step": 21174 + }, + { + "epoch": 1.6976007696778979, + "grad_norm": 1.6908522844314575, + "learning_rate": 1.1737773027978615e-06, + "loss": 3.1945, + "step": 21175 + }, + { + "epoch": 1.6976809444589205, + "grad_norm": 1.7135801315307617, + "learning_rate": 1.1731669994505435e-06, + "loss": 3.192, + "step": 21176 + }, + { + "epoch": 1.697761119239943, + "grad_norm": 1.7446510791778564, + "learning_rate": 1.1725568449205204e-06, + "loss": 3.1368, + "step": 21177 + }, + { + "epoch": 1.6978412940209657, + "grad_norm": 1.679102897644043, + "learning_rate": 1.1719468392180766e-06, + "loss": 3.0867, + "step": 21178 + }, + { + "epoch": 1.6979214688019884, + "grad_norm": 0.6449390053749084, + "learning_rate": 1.171336982353497e-06, + "loss": 2.4633, + "step": 21179 + }, + { + "epoch": 1.698001643583011, + "grad_norm": 1.7481107711791992, + "learning_rate": 1.1707272743370646e-06, + "loss": 3.1107, + "step": 21180 + }, + { + "epoch": 1.6980818183640336, + "grad_norm": 1.7769871950149536, + "learning_rate": 1.1701177151790588e-06, + "loss": 3.0484, + "step": 21181 + }, + { + "epoch": 1.6981619931450562, + "grad_norm": 1.8044283390045166, + "learning_rate": 1.1695083048897582e-06, + "loss": 3.1567, + "step": 21182 + }, + { + "epoch": 1.6982421679260788, + "grad_norm": 0.6506040692329407, + "learning_rate": 1.1688990434794346e-06, + "loss": 2.4622, + "step": 21183 + }, + { + "epoch": 1.6983223427071015, + "grad_norm": 1.7478301525115967, + "learning_rate": 1.1682899309583595e-06, + "loss": 3.1854, + "step": 21184 + }, + { + "epoch": 1.698402517488124, + "grad_norm": 1.84158194065094, + "learning_rate": 1.167680967336804e-06, + "loss": 3.2189, + "step": 21185 + }, + { + "epoch": 1.6984826922691467, + "grad_norm": 1.7872604131698608, + "learning_rate": 1.1670721526250339e-06, + "loss": 3.1382, + "step": 21186 + }, + { + "epoch": 1.6985628670501693, + "grad_norm": 1.8973370790481567, + "learning_rate": 1.1664634868333168e-06, + "loss": 3.1989, + "step": 21187 + }, + { + "epoch": 1.698643041831192, + "grad_norm": 1.8283206224441528, + "learning_rate": 1.1658549699719101e-06, + "loss": 3.1493, + "step": 21188 + }, + { + "epoch": 1.6987232166122146, + "grad_norm": 1.748993992805481, + "learning_rate": 1.1652466020510756e-06, + "loss": 3.1406, + "step": 21189 + }, + { + "epoch": 1.6988033913932372, + "grad_norm": 1.7568175792694092, + "learning_rate": 1.164638383081068e-06, + "loss": 3.1712, + "step": 21190 + }, + { + "epoch": 1.6988835661742598, + "grad_norm": 1.8127132654190063, + "learning_rate": 1.1640303130721453e-06, + "loss": 3.2678, + "step": 21191 + }, + { + "epoch": 1.6989637409552825, + "grad_norm": 0.6443927884101868, + "learning_rate": 1.1634223920345588e-06, + "loss": 2.4613, + "step": 21192 + }, + { + "epoch": 1.699043915736305, + "grad_norm": 1.8639379739761353, + "learning_rate": 1.162814619978554e-06, + "loss": 3.0907, + "step": 21193 + }, + { + "epoch": 1.6991240905173277, + "grad_norm": 1.8510876893997192, + "learning_rate": 1.1622069969143812e-06, + "loss": 3.196, + "step": 21194 + }, + { + "epoch": 1.6992042652983503, + "grad_norm": 1.6896017789840698, + "learning_rate": 1.1615995228522824e-06, + "loss": 3.084, + "step": 21195 + }, + { + "epoch": 1.699284440079373, + "grad_norm": 1.795606255531311, + "learning_rate": 1.1609921978025018e-06, + "loss": 3.2538, + "step": 21196 + }, + { + "epoch": 1.6993646148603956, + "grad_norm": 1.7223994731903076, + "learning_rate": 1.1603850217752788e-06, + "loss": 3.1796, + "step": 21197 + }, + { + "epoch": 1.6994447896414182, + "grad_norm": 1.6603177785873413, + "learning_rate": 1.1597779947808474e-06, + "loss": 3.1201, + "step": 21198 + }, + { + "epoch": 1.6995249644224408, + "grad_norm": 1.7176498174667358, + "learning_rate": 1.159171116829443e-06, + "loss": 3.1443, + "step": 21199 + }, + { + "epoch": 1.6996051392034635, + "grad_norm": 1.786073923110962, + "learning_rate": 1.158564387931298e-06, + "loss": 3.1857, + "step": 21200 + }, + { + "epoch": 1.699685313984486, + "grad_norm": 1.6750274896621704, + "learning_rate": 1.1579578080966414e-06, + "loss": 3.208, + "step": 21201 + }, + { + "epoch": 1.6997654887655087, + "grad_norm": 1.8549306392669678, + "learning_rate": 1.1573513773357025e-06, + "loss": 3.1913, + "step": 21202 + }, + { + "epoch": 1.6998456635465313, + "grad_norm": 1.7052403688430786, + "learning_rate": 1.1567450956587001e-06, + "loss": 3.0966, + "step": 21203 + }, + { + "epoch": 1.699925838327554, + "grad_norm": 0.6547728776931763, + "learning_rate": 1.156138963075859e-06, + "loss": 2.4626, + "step": 21204 + }, + { + "epoch": 1.7000060131085766, + "grad_norm": 1.7637132406234741, + "learning_rate": 1.155532979597398e-06, + "loss": 3.109, + "step": 21205 + }, + { + "epoch": 1.7000861878895992, + "grad_norm": 1.9537394046783447, + "learning_rate": 1.1549271452335342e-06, + "loss": 3.1614, + "step": 21206 + }, + { + "epoch": 1.7001663626706218, + "grad_norm": 1.9084969758987427, + "learning_rate": 1.1543214599944829e-06, + "loss": 3.1433, + "step": 21207 + }, + { + "epoch": 1.7002465374516444, + "grad_norm": 1.7868950366973877, + "learning_rate": 1.1537159238904528e-06, + "loss": 3.1953, + "step": 21208 + }, + { + "epoch": 1.700326712232667, + "grad_norm": 1.8865816593170166, + "learning_rate": 1.1531105369316542e-06, + "loss": 3.2352, + "step": 21209 + }, + { + "epoch": 1.7004068870136897, + "grad_norm": 1.8323187828063965, + "learning_rate": 1.1525052991282938e-06, + "loss": 3.2101, + "step": 21210 + }, + { + "epoch": 1.7004870617947123, + "grad_norm": 1.760688066482544, + "learning_rate": 1.1519002104905763e-06, + "loss": 3.2236, + "step": 21211 + }, + { + "epoch": 1.700567236575735, + "grad_norm": 1.7581300735473633, + "learning_rate": 1.151295271028705e-06, + "loss": 3.1504, + "step": 21212 + }, + { + "epoch": 1.7006474113567576, + "grad_norm": 1.6642998456954956, + "learning_rate": 1.150690480752874e-06, + "loss": 3.1368, + "step": 21213 + }, + { + "epoch": 1.7007275861377804, + "grad_norm": 1.761212706565857, + "learning_rate": 1.1500858396732838e-06, + "loss": 3.0796, + "step": 21214 + }, + { + "epoch": 1.700807760918803, + "grad_norm": 1.7284587621688843, + "learning_rate": 1.1494813478001265e-06, + "loss": 3.1627, + "step": 21215 + }, + { + "epoch": 1.7008879356998257, + "grad_norm": 1.730477213859558, + "learning_rate": 1.148877005143595e-06, + "loss": 3.1982, + "step": 21216 + }, + { + "epoch": 1.7009681104808483, + "grad_norm": 1.8830398321151733, + "learning_rate": 1.1482728117138798e-06, + "loss": 3.164, + "step": 21217 + }, + { + "epoch": 1.701048285261871, + "grad_norm": 1.75285005569458, + "learning_rate": 1.1476687675211628e-06, + "loss": 3.1296, + "step": 21218 + }, + { + "epoch": 1.7011284600428935, + "grad_norm": 1.6612272262573242, + "learning_rate": 1.1470648725756305e-06, + "loss": 3.116, + "step": 21219 + }, + { + "epoch": 1.7012086348239162, + "grad_norm": 1.845465064048767, + "learning_rate": 1.1464611268874648e-06, + "loss": 3.1635, + "step": 21220 + }, + { + "epoch": 1.7012888096049388, + "grad_norm": 1.829474925994873, + "learning_rate": 1.145857530466844e-06, + "loss": 3.174, + "step": 21221 + }, + { + "epoch": 1.7013689843859614, + "grad_norm": 1.7926008701324463, + "learning_rate": 1.1452540833239467e-06, + "loss": 3.2555, + "step": 21222 + }, + { + "epoch": 1.701449159166984, + "grad_norm": 1.735944390296936, + "learning_rate": 1.1446507854689425e-06, + "loss": 3.1802, + "step": 21223 + }, + { + "epoch": 1.7015293339480067, + "grad_norm": 1.8284584283828735, + "learning_rate": 1.1440476369120056e-06, + "loss": 3.1666, + "step": 21224 + }, + { + "epoch": 1.7016095087290293, + "grad_norm": 1.8289285898208618, + "learning_rate": 1.1434446376633047e-06, + "loss": 3.1496, + "step": 21225 + }, + { + "epoch": 1.701689683510052, + "grad_norm": 1.7611597776412964, + "learning_rate": 1.1428417877330055e-06, + "loss": 3.2025, + "step": 21226 + }, + { + "epoch": 1.7017698582910745, + "grad_norm": 1.825003743171692, + "learning_rate": 1.1422390871312739e-06, + "loss": 3.1766, + "step": 21227 + }, + { + "epoch": 1.7018500330720971, + "grad_norm": 1.7787827253341675, + "learning_rate": 1.141636535868268e-06, + "loss": 3.1549, + "step": 21228 + }, + { + "epoch": 1.7019302078531198, + "grad_norm": 0.67524653673172, + "learning_rate": 1.1410341339541487e-06, + "loss": 2.462, + "step": 21229 + }, + { + "epoch": 1.7020103826341424, + "grad_norm": 1.9541014432907104, + "learning_rate": 1.14043188139907e-06, + "loss": 3.2628, + "step": 21230 + }, + { + "epoch": 1.702090557415165, + "grad_norm": 1.7053152322769165, + "learning_rate": 1.1398297782131883e-06, + "loss": 3.1024, + "step": 21231 + }, + { + "epoch": 1.7021707321961876, + "grad_norm": 1.8463436365127563, + "learning_rate": 1.1392278244066569e-06, + "loss": 3.1584, + "step": 21232 + }, + { + "epoch": 1.7022509069772103, + "grad_norm": 1.7426539659500122, + "learning_rate": 1.1386260199896183e-06, + "loss": 3.1375, + "step": 21233 + }, + { + "epoch": 1.7023310817582331, + "grad_norm": 1.8028651475906372, + "learning_rate": 1.1380243649722223e-06, + "loss": 3.2043, + "step": 21234 + }, + { + "epoch": 1.7024112565392557, + "grad_norm": 1.8405921459197998, + "learning_rate": 1.137422859364612e-06, + "loss": 3.1072, + "step": 21235 + }, + { + "epoch": 1.7024914313202784, + "grad_norm": 0.6543944478034973, + "learning_rate": 1.1368215031769291e-06, + "loss": 2.4636, + "step": 21236 + }, + { + "epoch": 1.702571606101301, + "grad_norm": 1.831749439239502, + "learning_rate": 1.1362202964193137e-06, + "loss": 3.16, + "step": 21237 + }, + { + "epoch": 1.7026517808823236, + "grad_norm": 1.787299394607544, + "learning_rate": 1.1356192391018983e-06, + "loss": 3.2008, + "step": 21238 + }, + { + "epoch": 1.7027319556633462, + "grad_norm": 1.7270172834396362, + "learning_rate": 1.1350183312348183e-06, + "loss": 3.1563, + "step": 21239 + }, + { + "epoch": 1.7028121304443689, + "grad_norm": 1.8252745866775513, + "learning_rate": 1.1344175728282048e-06, + "loss": 3.2399, + "step": 21240 + }, + { + "epoch": 1.7028923052253915, + "grad_norm": 1.8233932256698608, + "learning_rate": 1.133816963892187e-06, + "loss": 3.1346, + "step": 21241 + }, + { + "epoch": 1.702972480006414, + "grad_norm": 1.7953861951828003, + "learning_rate": 1.1332165044368926e-06, + "loss": 3.1865, + "step": 21242 + }, + { + "epoch": 1.7030526547874367, + "grad_norm": 1.7043980360031128, + "learning_rate": 1.13261619447244e-06, + "loss": 3.1794, + "step": 21243 + }, + { + "epoch": 1.7031328295684593, + "grad_norm": 1.8013116121292114, + "learning_rate": 1.1320160340089536e-06, + "loss": 3.2054, + "step": 21244 + }, + { + "epoch": 1.703213004349482, + "grad_norm": 1.7906259298324585, + "learning_rate": 1.1314160230565518e-06, + "loss": 3.1973, + "step": 21245 + }, + { + "epoch": 1.7032931791305046, + "grad_norm": 1.816870927810669, + "learning_rate": 1.13081616162535e-06, + "loss": 3.173, + "step": 21246 + }, + { + "epoch": 1.7033733539115272, + "grad_norm": 1.7409098148345947, + "learning_rate": 1.130216449725462e-06, + "loss": 3.1663, + "step": 21247 + }, + { + "epoch": 1.7034535286925498, + "grad_norm": 1.7604016065597534, + "learning_rate": 1.129616887367001e-06, + "loss": 3.1792, + "step": 21248 + }, + { + "epoch": 1.7035337034735725, + "grad_norm": 1.7447524070739746, + "learning_rate": 1.1290174745600702e-06, + "loss": 3.1083, + "step": 21249 + }, + { + "epoch": 1.703613878254595, + "grad_norm": 1.7685091495513916, + "learning_rate": 1.128418211314779e-06, + "loss": 3.3102, + "step": 21250 + }, + { + "epoch": 1.7036940530356177, + "grad_norm": 1.8480302095413208, + "learning_rate": 1.1278190976412306e-06, + "loss": 3.2051, + "step": 21251 + }, + { + "epoch": 1.7037742278166403, + "grad_norm": 1.8087241649627686, + "learning_rate": 1.1272201335495248e-06, + "loss": 3.1596, + "step": 21252 + }, + { + "epoch": 1.703854402597663, + "grad_norm": 1.7075624465942383, + "learning_rate": 1.1266213190497622e-06, + "loss": 3.0884, + "step": 21253 + }, + { + "epoch": 1.7039345773786856, + "grad_norm": 1.7088918685913086, + "learning_rate": 1.1260226541520347e-06, + "loss": 3.1141, + "step": 21254 + }, + { + "epoch": 1.7040147521597082, + "grad_norm": 1.7553520202636719, + "learning_rate": 1.1254241388664378e-06, + "loss": 3.1582, + "step": 21255 + }, + { + "epoch": 1.7040949269407308, + "grad_norm": 1.8256540298461914, + "learning_rate": 1.1248257732030632e-06, + "loss": 3.2177, + "step": 21256 + }, + { + "epoch": 1.7041751017217535, + "grad_norm": 1.825499415397644, + "learning_rate": 1.1242275571719973e-06, + "loss": 3.1208, + "step": 21257 + }, + { + "epoch": 1.704255276502776, + "grad_norm": 1.7640682458877563, + "learning_rate": 1.123629490783329e-06, + "loss": 3.1246, + "step": 21258 + }, + { + "epoch": 1.7043354512837987, + "grad_norm": 1.8185571432113647, + "learning_rate": 1.1230315740471364e-06, + "loss": 3.0521, + "step": 21259 + }, + { + "epoch": 1.7044156260648213, + "grad_norm": 1.8224761486053467, + "learning_rate": 1.1224338069735041e-06, + "loss": 3.163, + "step": 21260 + }, + { + "epoch": 1.704495800845844, + "grad_norm": 1.7314162254333496, + "learning_rate": 1.1218361895725082e-06, + "loss": 3.1321, + "step": 21261 + }, + { + "epoch": 1.7045759756268666, + "grad_norm": 1.8591090440750122, + "learning_rate": 1.1212387218542252e-06, + "loss": 3.1704, + "step": 21262 + }, + { + "epoch": 1.7046561504078892, + "grad_norm": 1.7329248189926147, + "learning_rate": 1.1206414038287305e-06, + "loss": 3.2257, + "step": 21263 + }, + { + "epoch": 1.7047363251889118, + "grad_norm": 1.8416038751602173, + "learning_rate": 1.1200442355060902e-06, + "loss": 3.1823, + "step": 21264 + }, + { + "epoch": 1.7048164999699345, + "grad_norm": 1.7967231273651123, + "learning_rate": 1.1194472168963743e-06, + "loss": 3.2991, + "step": 21265 + }, + { + "epoch": 1.704896674750957, + "grad_norm": 1.7713572978973389, + "learning_rate": 1.118850348009649e-06, + "loss": 3.181, + "step": 21266 + }, + { + "epoch": 1.7049768495319797, + "grad_norm": 1.722978949546814, + "learning_rate": 1.1182536288559764e-06, + "loss": 3.1516, + "step": 21267 + }, + { + "epoch": 1.7050570243130023, + "grad_norm": 1.8391287326812744, + "learning_rate": 1.1176570594454206e-06, + "loss": 3.1247, + "step": 21268 + }, + { + "epoch": 1.705137199094025, + "grad_norm": 1.7390286922454834, + "learning_rate": 1.1170606397880324e-06, + "loss": 3.1016, + "step": 21269 + }, + { + "epoch": 1.7052173738750476, + "grad_norm": 1.773400068283081, + "learning_rate": 1.1164643698938727e-06, + "loss": 3.1905, + "step": 21270 + }, + { + "epoch": 1.7052975486560702, + "grad_norm": 1.7316370010375977, + "learning_rate": 1.1158682497729922e-06, + "loss": 3.115, + "step": 21271 + }, + { + "epoch": 1.7053777234370928, + "grad_norm": 1.776902198791504, + "learning_rate": 1.1152722794354421e-06, + "loss": 3.1794, + "step": 21272 + }, + { + "epoch": 1.7054578982181154, + "grad_norm": 1.8945850133895874, + "learning_rate": 1.1146764588912718e-06, + "loss": 3.1791, + "step": 21273 + }, + { + "epoch": 1.705538072999138, + "grad_norm": 1.7984864711761475, + "learning_rate": 1.1140807881505233e-06, + "loss": 3.0421, + "step": 21274 + }, + { + "epoch": 1.7056182477801607, + "grad_norm": 1.7667940855026245, + "learning_rate": 1.1134852672232398e-06, + "loss": 3.1271, + "step": 21275 + }, + { + "epoch": 1.7056984225611833, + "grad_norm": 1.702012538909912, + "learning_rate": 1.1128898961194644e-06, + "loss": 3.1623, + "step": 21276 + }, + { + "epoch": 1.705778597342206, + "grad_norm": 1.8268167972564697, + "learning_rate": 1.1122946748492326e-06, + "loss": 3.2542, + "step": 21277 + }, + { + "epoch": 1.7058587721232286, + "grad_norm": 0.6441702842712402, + "learning_rate": 1.1116996034225824e-06, + "loss": 2.4628, + "step": 21278 + }, + { + "epoch": 1.7059389469042512, + "grad_norm": 1.7265418767929077, + "learning_rate": 1.1111046818495418e-06, + "loss": 3.2223, + "step": 21279 + }, + { + "epoch": 1.7060191216852738, + "grad_norm": 1.839354395866394, + "learning_rate": 1.1105099101401439e-06, + "loss": 3.1907, + "step": 21280 + }, + { + "epoch": 1.7060992964662964, + "grad_norm": 1.7073839902877808, + "learning_rate": 1.1099152883044162e-06, + "loss": 3.1906, + "step": 21281 + }, + { + "epoch": 1.706179471247319, + "grad_norm": 1.6705942153930664, + "learning_rate": 1.1093208163523827e-06, + "loss": 3.0612, + "step": 21282 + }, + { + "epoch": 1.7062596460283417, + "grad_norm": 1.7712725400924683, + "learning_rate": 1.1087264942940702e-06, + "loss": 3.1191, + "step": 21283 + }, + { + "epoch": 1.7063398208093643, + "grad_norm": 1.7328095436096191, + "learning_rate": 1.1081323221394923e-06, + "loss": 3.0919, + "step": 21284 + }, + { + "epoch": 1.706419995590387, + "grad_norm": 1.7031971216201782, + "learning_rate": 1.1075382998986705e-06, + "loss": 3.122, + "step": 21285 + }, + { + "epoch": 1.7065001703714096, + "grad_norm": 1.815725326538086, + "learning_rate": 1.1069444275816178e-06, + "loss": 3.1725, + "step": 21286 + }, + { + "epoch": 1.7065803451524322, + "grad_norm": 1.6986451148986816, + "learning_rate": 1.106350705198349e-06, + "loss": 3.1086, + "step": 21287 + }, + { + "epoch": 1.7066605199334548, + "grad_norm": 1.9782218933105469, + "learning_rate": 1.1057571327588746e-06, + "loss": 3.0947, + "step": 21288 + }, + { + "epoch": 1.7067406947144774, + "grad_norm": 1.7644952535629272, + "learning_rate": 1.1051637102731983e-06, + "loss": 3.2752, + "step": 21289 + }, + { + "epoch": 1.7068208694955, + "grad_norm": 1.7825011014938354, + "learning_rate": 1.1045704377513267e-06, + "loss": 3.2122, + "step": 21290 + }, + { + "epoch": 1.7069010442765227, + "grad_norm": 1.7863010168075562, + "learning_rate": 1.1039773152032618e-06, + "loss": 3.1498, + "step": 21291 + }, + { + "epoch": 1.7069812190575453, + "grad_norm": 1.6577547788619995, + "learning_rate": 1.1033843426390055e-06, + "loss": 3.1082, + "step": 21292 + }, + { + "epoch": 1.707061393838568, + "grad_norm": 1.7718898057937622, + "learning_rate": 1.1027915200685546e-06, + "loss": 3.0532, + "step": 21293 + }, + { + "epoch": 1.7071415686195905, + "grad_norm": 1.7850754261016846, + "learning_rate": 1.102198847501902e-06, + "loss": 3.0903, + "step": 21294 + }, + { + "epoch": 1.7072217434006134, + "grad_norm": 1.7762519121170044, + "learning_rate": 1.1016063249490405e-06, + "loss": 3.1195, + "step": 21295 + }, + { + "epoch": 1.707301918181636, + "grad_norm": 1.7959011793136597, + "learning_rate": 1.1010139524199603e-06, + "loss": 3.2236, + "step": 21296 + }, + { + "epoch": 1.7073820929626586, + "grad_norm": 1.7932504415512085, + "learning_rate": 1.1004217299246479e-06, + "loss": 3.1741, + "step": 21297 + }, + { + "epoch": 1.7074622677436813, + "grad_norm": 1.8290581703186035, + "learning_rate": 1.0998296574730916e-06, + "loss": 3.2523, + "step": 21298 + }, + { + "epoch": 1.7075424425247039, + "grad_norm": 1.8746892213821411, + "learning_rate": 1.0992377350752682e-06, + "loss": 3.3356, + "step": 21299 + }, + { + "epoch": 1.7076226173057265, + "grad_norm": 1.7379395961761475, + "learning_rate": 1.0986459627411594e-06, + "loss": 3.147, + "step": 21300 + }, + { + "epoch": 1.7077027920867491, + "grad_norm": 1.8383313417434692, + "learning_rate": 1.098054340480742e-06, + "loss": 3.1479, + "step": 21301 + }, + { + "epoch": 1.7077829668677718, + "grad_norm": 1.7434701919555664, + "learning_rate": 1.0974628683039924e-06, + "loss": 3.1605, + "step": 21302 + }, + { + "epoch": 1.7078631416487944, + "grad_norm": 1.942796230316162, + "learning_rate": 1.0968715462208824e-06, + "loss": 3.216, + "step": 21303 + }, + { + "epoch": 1.707943316429817, + "grad_norm": 1.8178859949111938, + "learning_rate": 1.0962803742413785e-06, + "loss": 3.2815, + "step": 21304 + }, + { + "epoch": 1.7080234912108396, + "grad_norm": 1.683691143989563, + "learning_rate": 1.0956893523754498e-06, + "loss": 3.1299, + "step": 21305 + }, + { + "epoch": 1.7081036659918623, + "grad_norm": 1.797153353691101, + "learning_rate": 1.09509848063306e-06, + "loss": 3.1632, + "step": 21306 + }, + { + "epoch": 1.7081838407728849, + "grad_norm": 1.7457010746002197, + "learning_rate": 1.0945077590241726e-06, + "loss": 3.0807, + "step": 21307 + }, + { + "epoch": 1.7082640155539075, + "grad_norm": 1.7251436710357666, + "learning_rate": 1.0939171875587461e-06, + "loss": 3.2067, + "step": 21308 + }, + { + "epoch": 1.7083441903349301, + "grad_norm": 1.7545851469039917, + "learning_rate": 1.0933267662467361e-06, + "loss": 3.2322, + "step": 21309 + }, + { + "epoch": 1.7084243651159527, + "grad_norm": 1.7574574947357178, + "learning_rate": 1.092736495098098e-06, + "loss": 3.2354, + "step": 21310 + }, + { + "epoch": 1.7085045398969754, + "grad_norm": 1.7387334108352661, + "learning_rate": 1.0921463741227823e-06, + "loss": 3.0939, + "step": 21311 + }, + { + "epoch": 1.708584714677998, + "grad_norm": 1.7193106412887573, + "learning_rate": 1.09155640333074e-06, + "loss": 3.2146, + "step": 21312 + }, + { + "epoch": 1.7086648894590206, + "grad_norm": 1.9230787754058838, + "learning_rate": 1.0909665827319194e-06, + "loss": 3.3246, + "step": 21313 + }, + { + "epoch": 1.7087450642400432, + "grad_norm": 0.671306312084198, + "learning_rate": 1.0903769123362594e-06, + "loss": 2.4642, + "step": 21314 + }, + { + "epoch": 1.708825239021066, + "grad_norm": 1.6581577062606812, + "learning_rate": 1.089787392153705e-06, + "loss": 3.0615, + "step": 21315 + }, + { + "epoch": 1.7089054138020887, + "grad_norm": 1.7083699703216553, + "learning_rate": 1.0891980221941955e-06, + "loss": 3.099, + "step": 21316 + }, + { + "epoch": 1.7089855885831113, + "grad_norm": 1.723488450050354, + "learning_rate": 1.0886088024676655e-06, + "loss": 3.1479, + "step": 21317 + }, + { + "epoch": 1.709065763364134, + "grad_norm": 0.6504719853401184, + "learning_rate": 1.0880197329840537e-06, + "loss": 2.4443, + "step": 21318 + }, + { + "epoch": 1.7091459381451566, + "grad_norm": 1.735594630241394, + "learning_rate": 1.0874308137532863e-06, + "loss": 3.0797, + "step": 21319 + }, + { + "epoch": 1.7092261129261792, + "grad_norm": 1.7745277881622314, + "learning_rate": 1.0868420447852935e-06, + "loss": 3.1979, + "step": 21320 + }, + { + "epoch": 1.7093062877072018, + "grad_norm": 1.8133323192596436, + "learning_rate": 1.0862534260900027e-06, + "loss": 3.1804, + "step": 21321 + }, + { + "epoch": 1.7093864624882245, + "grad_norm": 1.7863765954971313, + "learning_rate": 1.0856649576773382e-06, + "loss": 3.2488, + "step": 21322 + }, + { + "epoch": 1.709466637269247, + "grad_norm": 1.7443222999572754, + "learning_rate": 1.0850766395572222e-06, + "loss": 3.1718, + "step": 21323 + }, + { + "epoch": 1.7095468120502697, + "grad_norm": 1.7626057863235474, + "learning_rate": 1.0844884717395709e-06, + "loss": 3.1992, + "step": 21324 + }, + { + "epoch": 1.7096269868312923, + "grad_norm": 1.876563549041748, + "learning_rate": 1.0839004542343012e-06, + "loss": 3.148, + "step": 21325 + }, + { + "epoch": 1.709707161612315, + "grad_norm": 0.6512997150421143, + "learning_rate": 1.0833125870513273e-06, + "loss": 2.4708, + "step": 21326 + }, + { + "epoch": 1.7097873363933376, + "grad_norm": 1.8729443550109863, + "learning_rate": 1.0827248702005611e-06, + "loss": 3.2307, + "step": 21327 + }, + { + "epoch": 1.7098675111743602, + "grad_norm": 1.733216643333435, + "learning_rate": 1.0821373036919125e-06, + "loss": 3.3322, + "step": 21328 + }, + { + "epoch": 1.7099476859553828, + "grad_norm": 1.6922428607940674, + "learning_rate": 1.0815498875352848e-06, + "loss": 3.09, + "step": 21329 + }, + { + "epoch": 1.7100278607364054, + "grad_norm": 1.8430302143096924, + "learning_rate": 1.0809626217405822e-06, + "loss": 3.139, + "step": 21330 + }, + { + "epoch": 1.710108035517428, + "grad_norm": 1.8103426694869995, + "learning_rate": 1.0803755063177068e-06, + "loss": 3.1816, + "step": 21331 + }, + { + "epoch": 1.7101882102984507, + "grad_norm": 1.7825528383255005, + "learning_rate": 1.0797885412765563e-06, + "loss": 3.2814, + "step": 21332 + }, + { + "epoch": 1.7102683850794733, + "grad_norm": 1.8070623874664307, + "learning_rate": 1.0792017266270294e-06, + "loss": 3.2947, + "step": 21333 + }, + { + "epoch": 1.710348559860496, + "grad_norm": 1.7837425470352173, + "learning_rate": 1.0786150623790148e-06, + "loss": 3.2185, + "step": 21334 + }, + { + "epoch": 1.7104287346415186, + "grad_norm": 1.8300825357437134, + "learning_rate": 1.078028548542407e-06, + "loss": 3.1124, + "step": 21335 + }, + { + "epoch": 1.7105089094225412, + "grad_norm": 1.8277140855789185, + "learning_rate": 1.0774421851270934e-06, + "loss": 3.1655, + "step": 21336 + }, + { + "epoch": 1.7105890842035638, + "grad_norm": 1.7977675199508667, + "learning_rate": 1.0768559721429594e-06, + "loss": 3.1204, + "step": 21337 + }, + { + "epoch": 1.7106692589845864, + "grad_norm": 1.8519142866134644, + "learning_rate": 1.0762699095998907e-06, + "loss": 3.1072, + "step": 21338 + }, + { + "epoch": 1.710749433765609, + "grad_norm": 1.7874304056167603, + "learning_rate": 1.075683997507765e-06, + "loss": 3.3068, + "step": 21339 + }, + { + "epoch": 1.7108296085466317, + "grad_norm": 0.6675302982330322, + "learning_rate": 1.0750982358764618e-06, + "loss": 2.4581, + "step": 21340 + }, + { + "epoch": 1.7109097833276543, + "grad_norm": 1.7293556928634644, + "learning_rate": 1.0745126247158567e-06, + "loss": 3.1372, + "step": 21341 + }, + { + "epoch": 1.710989958108677, + "grad_norm": 1.7826647758483887, + "learning_rate": 1.0739271640358228e-06, + "loss": 3.1997, + "step": 21342 + }, + { + "epoch": 1.7110701328896996, + "grad_norm": 1.815165638923645, + "learning_rate": 1.0733418538462326e-06, + "loss": 3.2214, + "step": 21343 + }, + { + "epoch": 1.7111503076707222, + "grad_norm": 1.788227915763855, + "learning_rate": 1.0727566941569522e-06, + "loss": 3.084, + "step": 21344 + }, + { + "epoch": 1.7112304824517448, + "grad_norm": 1.7461858987808228, + "learning_rate": 1.0721716849778462e-06, + "loss": 3.2756, + "step": 21345 + }, + { + "epoch": 1.7113106572327674, + "grad_norm": 1.8452104330062866, + "learning_rate": 1.0715868263187801e-06, + "loss": 3.2815, + "step": 21346 + }, + { + "epoch": 1.71139083201379, + "grad_norm": 1.7473020553588867, + "learning_rate": 1.0710021181896135e-06, + "loss": 3.1564, + "step": 21347 + }, + { + "epoch": 1.7114710067948127, + "grad_norm": 1.8454207181930542, + "learning_rate": 1.0704175606002053e-06, + "loss": 3.192, + "step": 21348 + }, + { + "epoch": 1.7115511815758353, + "grad_norm": 1.74248468875885, + "learning_rate": 1.0698331535604089e-06, + "loss": 3.1404, + "step": 21349 + }, + { + "epoch": 1.711631356356858, + "grad_norm": 1.7663358449935913, + "learning_rate": 1.0692488970800773e-06, + "loss": 3.1847, + "step": 21350 + }, + { + "epoch": 1.7117115311378805, + "grad_norm": 1.8526989221572876, + "learning_rate": 1.0686647911690628e-06, + "loss": 3.153, + "step": 21351 + }, + { + "epoch": 1.7117917059189032, + "grad_norm": 1.9265018701553345, + "learning_rate": 1.0680808358372118e-06, + "loss": 3.2493, + "step": 21352 + }, + { + "epoch": 1.7118718806999258, + "grad_norm": 1.7808350324630737, + "learning_rate": 1.0674970310943723e-06, + "loss": 3.1611, + "step": 21353 + }, + { + "epoch": 1.7119520554809484, + "grad_norm": 1.8768390417099, + "learning_rate": 1.0669133769503825e-06, + "loss": 3.2113, + "step": 21354 + }, + { + "epoch": 1.712032230261971, + "grad_norm": 1.792055368423462, + "learning_rate": 1.066329873415085e-06, + "loss": 3.1938, + "step": 21355 + }, + { + "epoch": 1.7121124050429937, + "grad_norm": 1.664118766784668, + "learning_rate": 1.0657465204983165e-06, + "loss": 3.2366, + "step": 21356 + }, + { + "epoch": 1.7121925798240163, + "grad_norm": 1.7285512685775757, + "learning_rate": 1.0651633182099141e-06, + "loss": 3.2321, + "step": 21357 + }, + { + "epoch": 1.712272754605039, + "grad_norm": 1.7627969980239868, + "learning_rate": 1.0645802665597104e-06, + "loss": 3.2167, + "step": 21358 + }, + { + "epoch": 1.7123529293860615, + "grad_norm": 1.9728269577026367, + "learning_rate": 1.0639973655575319e-06, + "loss": 3.175, + "step": 21359 + }, + { + "epoch": 1.7124331041670842, + "grad_norm": 1.7953500747680664, + "learning_rate": 1.0634146152132097e-06, + "loss": 3.1143, + "step": 21360 + }, + { + "epoch": 1.7125132789481068, + "grad_norm": 1.8563225269317627, + "learning_rate": 1.0628320155365668e-06, + "loss": 3.1904, + "step": 21361 + }, + { + "epoch": 1.7125934537291294, + "grad_norm": 1.7451411485671997, + "learning_rate": 1.0622495665374265e-06, + "loss": 3.1697, + "step": 21362 + }, + { + "epoch": 1.712673628510152, + "grad_norm": 1.7241345643997192, + "learning_rate": 1.0616672682256102e-06, + "loss": 3.2603, + "step": 21363 + }, + { + "epoch": 1.7127538032911747, + "grad_norm": 1.8563750982284546, + "learning_rate": 1.0610851206109317e-06, + "loss": 3.1907, + "step": 21364 + }, + { + "epoch": 1.7128339780721973, + "grad_norm": 1.8712434768676758, + "learning_rate": 1.060503123703208e-06, + "loss": 3.1837, + "step": 21365 + }, + { + "epoch": 1.71291415285322, + "grad_norm": 1.7639029026031494, + "learning_rate": 1.0599212775122514e-06, + "loss": 3.1436, + "step": 21366 + }, + { + "epoch": 1.7129943276342425, + "grad_norm": 1.8848910331726074, + "learning_rate": 1.0593395820478703e-06, + "loss": 3.2046, + "step": 21367 + }, + { + "epoch": 1.7130745024152652, + "grad_norm": 1.8968071937561035, + "learning_rate": 1.0587580373198757e-06, + "loss": 3.1707, + "step": 21368 + }, + { + "epoch": 1.7131546771962878, + "grad_norm": 1.7581086158752441, + "learning_rate": 1.058176643338067e-06, + "loss": 3.0896, + "step": 21369 + }, + { + "epoch": 1.7132348519773104, + "grad_norm": 1.7673667669296265, + "learning_rate": 1.0575954001122478e-06, + "loss": 3.1831, + "step": 21370 + }, + { + "epoch": 1.713315026758333, + "grad_norm": 1.8073325157165527, + "learning_rate": 1.05701430765222e-06, + "loss": 3.1428, + "step": 21371 + }, + { + "epoch": 1.7133952015393557, + "grad_norm": 1.7827333211898804, + "learning_rate": 1.0564333659677783e-06, + "loss": 3.1407, + "step": 21372 + }, + { + "epoch": 1.7134753763203783, + "grad_norm": 1.7648497819900513, + "learning_rate": 1.055852575068721e-06, + "loss": 3.145, + "step": 21373 + }, + { + "epoch": 1.713555551101401, + "grad_norm": 1.9040606021881104, + "learning_rate": 1.0552719349648344e-06, + "loss": 3.2238, + "step": 21374 + }, + { + "epoch": 1.7136357258824235, + "grad_norm": 1.856087565422058, + "learning_rate": 1.0546914456659096e-06, + "loss": 3.1962, + "step": 21375 + }, + { + "epoch": 1.7137159006634464, + "grad_norm": 1.905577301979065, + "learning_rate": 1.0541111071817355e-06, + "loss": 3.1338, + "step": 21376 + }, + { + "epoch": 1.713796075444469, + "grad_norm": 1.8517483472824097, + "learning_rate": 1.0535309195220954e-06, + "loss": 3.1116, + "step": 21377 + }, + { + "epoch": 1.7138762502254916, + "grad_norm": 1.7918822765350342, + "learning_rate": 1.0529508826967727e-06, + "loss": 3.2627, + "step": 21378 + }, + { + "epoch": 1.7139564250065142, + "grad_norm": 1.7979047298431396, + "learning_rate": 1.0523709967155426e-06, + "loss": 3.2162, + "step": 21379 + }, + { + "epoch": 1.7140365997875369, + "grad_norm": 1.7750496864318848, + "learning_rate": 1.0517912615881843e-06, + "loss": 3.1941, + "step": 21380 + }, + { + "epoch": 1.7141167745685595, + "grad_norm": 1.7266499996185303, + "learning_rate": 1.0512116773244717e-06, + "loss": 3.2106, + "step": 21381 + }, + { + "epoch": 1.714196949349582, + "grad_norm": 1.8470849990844727, + "learning_rate": 1.0506322439341765e-06, + "loss": 3.1219, + "step": 21382 + }, + { + "epoch": 1.7142771241306047, + "grad_norm": 1.8079488277435303, + "learning_rate": 1.0500529614270694e-06, + "loss": 3.1708, + "step": 21383 + }, + { + "epoch": 1.7143572989116274, + "grad_norm": 1.7979127168655396, + "learning_rate": 1.0494738298129137e-06, + "loss": 3.1377, + "step": 21384 + }, + { + "epoch": 1.71443747369265, + "grad_norm": 1.7292814254760742, + "learning_rate": 1.0488948491014738e-06, + "loss": 3.1752, + "step": 21385 + }, + { + "epoch": 1.7145176484736726, + "grad_norm": 1.7740720510482788, + "learning_rate": 1.048316019302512e-06, + "loss": 3.2192, + "step": 21386 + }, + { + "epoch": 1.7145978232546952, + "grad_norm": 1.7906208038330078, + "learning_rate": 1.0477373404257885e-06, + "loss": 3.1641, + "step": 21387 + }, + { + "epoch": 1.7146779980357179, + "grad_norm": 1.685189127922058, + "learning_rate": 1.0471588124810594e-06, + "loss": 3.063, + "step": 21388 + }, + { + "epoch": 1.7147581728167405, + "grad_norm": 1.9262747764587402, + "learning_rate": 1.046580435478075e-06, + "loss": 3.1542, + "step": 21389 + }, + { + "epoch": 1.714838347597763, + "grad_norm": 1.860188603401184, + "learning_rate": 1.0460022094265898e-06, + "loss": 3.1202, + "step": 21390 + }, + { + "epoch": 1.7149185223787857, + "grad_norm": 1.7911701202392578, + "learning_rate": 1.0454241343363513e-06, + "loss": 3.1459, + "step": 21391 + }, + { + "epoch": 1.7149986971598083, + "grad_norm": 1.7696608304977417, + "learning_rate": 1.0448462102171065e-06, + "loss": 3.1091, + "step": 21392 + }, + { + "epoch": 1.715078871940831, + "grad_norm": 1.7929178476333618, + "learning_rate": 1.0442684370786005e-06, + "loss": 3.1097, + "step": 21393 + }, + { + "epoch": 1.7151590467218536, + "grad_norm": 0.6607376337051392, + "learning_rate": 1.043690814930569e-06, + "loss": 2.4829, + "step": 21394 + }, + { + "epoch": 1.7152392215028762, + "grad_norm": 1.7340372800827026, + "learning_rate": 1.0431133437827557e-06, + "loss": 3.1301, + "step": 21395 + }, + { + "epoch": 1.7153193962838988, + "grad_norm": 1.837867021560669, + "learning_rate": 1.0425360236448933e-06, + "loss": 3.1018, + "step": 21396 + }, + { + "epoch": 1.7153995710649217, + "grad_norm": 1.716138243675232, + "learning_rate": 1.0419588545267179e-06, + "loss": 3.1816, + "step": 21397 + }, + { + "epoch": 1.7154797458459443, + "grad_norm": 1.732864260673523, + "learning_rate": 1.0413818364379601e-06, + "loss": 3.1466, + "step": 21398 + }, + { + "epoch": 1.715559920626967, + "grad_norm": 1.706671118736267, + "learning_rate": 1.0408049693883449e-06, + "loss": 3.0625, + "step": 21399 + }, + { + "epoch": 1.7156400954079896, + "grad_norm": 1.898957371711731, + "learning_rate": 1.0402282533876007e-06, + "loss": 3.1731, + "step": 21400 + }, + { + "epoch": 1.7157202701890122, + "grad_norm": 1.8337459564208984, + "learning_rate": 1.039651688445451e-06, + "loss": 3.2109, + "step": 21401 + }, + { + "epoch": 1.7158004449700348, + "grad_norm": 1.8471215963363647, + "learning_rate": 1.039075274571616e-06, + "loss": 3.2079, + "step": 21402 + }, + { + "epoch": 1.7158806197510574, + "grad_norm": 1.7617994546890259, + "learning_rate": 1.0384990117758153e-06, + "loss": 3.1622, + "step": 21403 + }, + { + "epoch": 1.71596079453208, + "grad_norm": 1.7467149496078491, + "learning_rate": 1.0379229000677604e-06, + "loss": 3.1805, + "step": 21404 + }, + { + "epoch": 1.7160409693131027, + "grad_norm": 1.7634036540985107, + "learning_rate": 1.0373469394571677e-06, + "loss": 3.1891, + "step": 21405 + }, + { + "epoch": 1.7161211440941253, + "grad_norm": 1.7481430768966675, + "learning_rate": 1.0367711299537463e-06, + "loss": 3.2522, + "step": 21406 + }, + { + "epoch": 1.716201318875148, + "grad_norm": 1.814159631729126, + "learning_rate": 1.0361954715672051e-06, + "loss": 3.1793, + "step": 21407 + }, + { + "epoch": 1.7162814936561706, + "grad_norm": 1.764147400856018, + "learning_rate": 1.0356199643072507e-06, + "loss": 3.1858, + "step": 21408 + }, + { + "epoch": 1.7163616684371932, + "grad_norm": 1.7963840961456299, + "learning_rate": 1.035044608183583e-06, + "loss": 3.2318, + "step": 21409 + }, + { + "epoch": 1.7164418432182158, + "grad_norm": 1.8275620937347412, + "learning_rate": 1.0344694032059032e-06, + "loss": 3.3458, + "step": 21410 + }, + { + "epoch": 1.7165220179992384, + "grad_norm": 1.7328428030014038, + "learning_rate": 1.0338943493839093e-06, + "loss": 3.2721, + "step": 21411 + }, + { + "epoch": 1.716602192780261, + "grad_norm": 1.7626160383224487, + "learning_rate": 1.0333194467272967e-06, + "loss": 3.1884, + "step": 21412 + }, + { + "epoch": 1.7166823675612837, + "grad_norm": 0.6451457738876343, + "learning_rate": 1.032744695245761e-06, + "loss": 2.4767, + "step": 21413 + }, + { + "epoch": 1.7167625423423063, + "grad_norm": 1.8183794021606445, + "learning_rate": 1.0321700949489865e-06, + "loss": 3.2051, + "step": 21414 + }, + { + "epoch": 1.716842717123329, + "grad_norm": 1.7766720056533813, + "learning_rate": 1.0315956458466636e-06, + "loss": 3.2027, + "step": 21415 + }, + { + "epoch": 1.7169228919043515, + "grad_norm": 1.7995679378509521, + "learning_rate": 1.0310213479484776e-06, + "loss": 3.1773, + "step": 21416 + }, + { + "epoch": 1.7170030666853742, + "grad_norm": 1.8297488689422607, + "learning_rate": 1.030447201264111e-06, + "loss": 3.2041, + "step": 21417 + }, + { + "epoch": 1.7170832414663968, + "grad_norm": 1.7758907079696655, + "learning_rate": 1.0298732058032445e-06, + "loss": 3.1406, + "step": 21418 + }, + { + "epoch": 1.7171634162474194, + "grad_norm": 1.7929593324661255, + "learning_rate": 1.0292993615755531e-06, + "loss": 3.2753, + "step": 21419 + }, + { + "epoch": 1.717243591028442, + "grad_norm": 1.8142433166503906, + "learning_rate": 1.028725668590712e-06, + "loss": 3.1551, + "step": 21420 + }, + { + "epoch": 1.7173237658094647, + "grad_norm": 1.7809381484985352, + "learning_rate": 1.028152126858396e-06, + "loss": 3.08, + "step": 21421 + }, + { + "epoch": 1.7174039405904873, + "grad_norm": 0.6676216721534729, + "learning_rate": 1.0275787363882717e-06, + "loss": 2.4669, + "step": 21422 + }, + { + "epoch": 1.71748411537151, + "grad_norm": 1.7866469621658325, + "learning_rate": 1.02700549719001e-06, + "loss": 3.2277, + "step": 21423 + }, + { + "epoch": 1.7175642901525325, + "grad_norm": 1.7753040790557861, + "learning_rate": 1.0264324092732714e-06, + "loss": 3.0922, + "step": 21424 + }, + { + "epoch": 1.7176444649335552, + "grad_norm": 0.7048854231834412, + "learning_rate": 1.02585947264772e-06, + "loss": 2.4573, + "step": 21425 + }, + { + "epoch": 1.7177246397145778, + "grad_norm": 1.651218295097351, + "learning_rate": 1.025286687323015e-06, + "loss": 3.0558, + "step": 21426 + }, + { + "epoch": 1.7178048144956004, + "grad_norm": 1.7326092720031738, + "learning_rate": 1.024714053308814e-06, + "loss": 3.1835, + "step": 21427 + }, + { + "epoch": 1.717884989276623, + "grad_norm": 1.8662995100021362, + "learning_rate": 1.0241415706147717e-06, + "loss": 3.2064, + "step": 21428 + }, + { + "epoch": 1.7179651640576457, + "grad_norm": 1.841524362564087, + "learning_rate": 1.023569239250538e-06, + "loss": 3.1171, + "step": 21429 + }, + { + "epoch": 1.7180453388386683, + "grad_norm": 1.7604080438613892, + "learning_rate": 1.0229970592257633e-06, + "loss": 3.2278, + "step": 21430 + }, + { + "epoch": 1.718125513619691, + "grad_norm": 0.6541135907173157, + "learning_rate": 1.0224250305500938e-06, + "loss": 2.4463, + "step": 21431 + }, + { + "epoch": 1.7182056884007135, + "grad_norm": 1.807847499847412, + "learning_rate": 1.0218531532331743e-06, + "loss": 3.1277, + "step": 21432 + }, + { + "epoch": 1.7182858631817362, + "grad_norm": 1.7815250158309937, + "learning_rate": 1.0212814272846483e-06, + "loss": 3.1505, + "step": 21433 + }, + { + "epoch": 1.7183660379627588, + "grad_norm": 1.7533546686172485, + "learning_rate": 1.0207098527141512e-06, + "loss": 3.1957, + "step": 21434 + }, + { + "epoch": 1.7184462127437814, + "grad_norm": 1.8526455163955688, + "learning_rate": 1.0201384295313221e-06, + "loss": 3.1923, + "step": 21435 + }, + { + "epoch": 1.718526387524804, + "grad_norm": 1.7185945510864258, + "learning_rate": 1.0195671577457934e-06, + "loss": 3.2257, + "step": 21436 + }, + { + "epoch": 1.7186065623058266, + "grad_norm": 1.745522141456604, + "learning_rate": 1.0189960373671981e-06, + "loss": 3.0701, + "step": 21437 + }, + { + "epoch": 1.7186867370868493, + "grad_norm": 1.7151368856430054, + "learning_rate": 1.0184250684051667e-06, + "loss": 3.1396, + "step": 21438 + }, + { + "epoch": 1.718766911867872, + "grad_norm": 1.841432809829712, + "learning_rate": 1.0178542508693202e-06, + "loss": 3.0968, + "step": 21439 + }, + { + "epoch": 1.7188470866488945, + "grad_norm": 1.8803585767745972, + "learning_rate": 1.0172835847692852e-06, + "loss": 3.2393, + "step": 21440 + }, + { + "epoch": 1.7189272614299171, + "grad_norm": 1.8800382614135742, + "learning_rate": 1.0167130701146844e-06, + "loss": 3.1437, + "step": 21441 + }, + { + "epoch": 1.7190074362109398, + "grad_norm": 1.892152190208435, + "learning_rate": 1.0161427069151342e-06, + "loss": 3.2144, + "step": 21442 + }, + { + "epoch": 1.7190876109919624, + "grad_norm": 1.8485051393508911, + "learning_rate": 1.0155724951802536e-06, + "loss": 3.2002, + "step": 21443 + }, + { + "epoch": 1.719167785772985, + "grad_norm": 1.7899366617202759, + "learning_rate": 1.015002434919653e-06, + "loss": 3.1787, + "step": 21444 + }, + { + "epoch": 1.7192479605540076, + "grad_norm": 1.8785361051559448, + "learning_rate": 1.0144325261429455e-06, + "loss": 3.1818, + "step": 21445 + }, + { + "epoch": 1.7193281353350303, + "grad_norm": 1.795266032218933, + "learning_rate": 1.013862768859737e-06, + "loss": 3.2349, + "step": 21446 + }, + { + "epoch": 1.7194083101160529, + "grad_norm": 1.9265246391296387, + "learning_rate": 1.0132931630796361e-06, + "loss": 3.2596, + "step": 21447 + }, + { + "epoch": 1.7194884848970755, + "grad_norm": 1.7579575777053833, + "learning_rate": 1.0127237088122477e-06, + "loss": 3.1332, + "step": 21448 + }, + { + "epoch": 1.7195686596780981, + "grad_norm": 1.7044267654418945, + "learning_rate": 1.0121544060671674e-06, + "loss": 3.1273, + "step": 21449 + }, + { + "epoch": 1.7196488344591208, + "grad_norm": 1.7079389095306396, + "learning_rate": 1.0115852548539963e-06, + "loss": 3.1315, + "step": 21450 + }, + { + "epoch": 1.7197290092401434, + "grad_norm": 1.7534266710281372, + "learning_rate": 1.01101625518233e-06, + "loss": 3.2156, + "step": 21451 + }, + { + "epoch": 1.719809184021166, + "grad_norm": 1.7545756101608276, + "learning_rate": 1.010447407061762e-06, + "loss": 3.0848, + "step": 21452 + }, + { + "epoch": 1.7198893588021886, + "grad_norm": 1.9204769134521484, + "learning_rate": 1.0098787105018837e-06, + "loss": 3.2725, + "step": 21453 + }, + { + "epoch": 1.7199695335832113, + "grad_norm": 1.74919855594635, + "learning_rate": 1.0093101655122806e-06, + "loss": 3.1386, + "step": 21454 + }, + { + "epoch": 1.7200497083642339, + "grad_norm": 0.6498160362243652, + "learning_rate": 1.0087417721025394e-06, + "loss": 2.4509, + "step": 21455 + }, + { + "epoch": 1.7201298831452565, + "grad_norm": 1.792492389678955, + "learning_rate": 1.0081735302822437e-06, + "loss": 3.1748, + "step": 21456 + }, + { + "epoch": 1.7202100579262791, + "grad_norm": 1.787266492843628, + "learning_rate": 1.0076054400609725e-06, + "loss": 3.1304, + "step": 21457 + }, + { + "epoch": 1.720290232707302, + "grad_norm": 1.8091789484024048, + "learning_rate": 1.0070375014483058e-06, + "loss": 3.2384, + "step": 21458 + }, + { + "epoch": 1.7203704074883246, + "grad_norm": 1.7781388759613037, + "learning_rate": 1.006469714453816e-06, + "loss": 3.1417, + "step": 21459 + }, + { + "epoch": 1.7204505822693472, + "grad_norm": 1.7627679109573364, + "learning_rate": 1.0059020790870777e-06, + "loss": 3.1579, + "step": 21460 + }, + { + "epoch": 1.7205307570503698, + "grad_norm": 1.7365514039993286, + "learning_rate": 1.0053345953576598e-06, + "loss": 3.1191, + "step": 21461 + }, + { + "epoch": 1.7206109318313925, + "grad_norm": 1.7213841676712036, + "learning_rate": 1.0047672632751303e-06, + "loss": 3.2302, + "step": 21462 + }, + { + "epoch": 1.720691106612415, + "grad_norm": 1.8328752517700195, + "learning_rate": 1.0042000828490573e-06, + "loss": 3.1777, + "step": 21463 + }, + { + "epoch": 1.7207712813934377, + "grad_norm": 1.7855006456375122, + "learning_rate": 1.0036330540889982e-06, + "loss": 3.1154, + "step": 21464 + }, + { + "epoch": 1.7208514561744603, + "grad_norm": 1.7140387296676636, + "learning_rate": 1.0030661770045147e-06, + "loss": 3.2287, + "step": 21465 + }, + { + "epoch": 1.720931630955483, + "grad_norm": 1.7103915214538574, + "learning_rate": 1.0024994516051656e-06, + "loss": 3.1867, + "step": 21466 + }, + { + "epoch": 1.7210118057365056, + "grad_norm": 1.857637882232666, + "learning_rate": 1.0019328779005045e-06, + "loss": 3.1952, + "step": 21467 + }, + { + "epoch": 1.7210919805175282, + "grad_norm": 1.7356712818145752, + "learning_rate": 1.0013664559000858e-06, + "loss": 3.2408, + "step": 21468 + }, + { + "epoch": 1.7211721552985508, + "grad_norm": 1.7179455757141113, + "learning_rate": 1.0008001856134542e-06, + "loss": 3.1388, + "step": 21469 + }, + { + "epoch": 1.7212523300795735, + "grad_norm": 1.9042457342147827, + "learning_rate": 1.0002340670501609e-06, + "loss": 3.1738, + "step": 21470 + }, + { + "epoch": 1.721332504860596, + "grad_norm": 1.653327465057373, + "learning_rate": 9.996681002197494e-07, + "loss": 3.18, + "step": 21471 + }, + { + "epoch": 1.7214126796416187, + "grad_norm": 1.7689497470855713, + "learning_rate": 9.99102285131761e-07, + "loss": 3.1503, + "step": 21472 + }, + { + "epoch": 1.7214928544226413, + "grad_norm": 1.7055715322494507, + "learning_rate": 9.985366217957381e-07, + "loss": 3.0906, + "step": 21473 + }, + { + "epoch": 1.721573029203664, + "grad_norm": 1.7618118524551392, + "learning_rate": 9.979711102212141e-07, + "loss": 3.1698, + "step": 21474 + }, + { + "epoch": 1.7216532039846866, + "grad_norm": 1.7318764925003052, + "learning_rate": 9.974057504177237e-07, + "loss": 3.1354, + "step": 21475 + }, + { + "epoch": 1.7217333787657092, + "grad_norm": 1.734344720840454, + "learning_rate": 9.968405423948002e-07, + "loss": 3.0538, + "step": 21476 + }, + { + "epoch": 1.7218135535467318, + "grad_norm": 0.6753585338592529, + "learning_rate": 9.962754861619717e-07, + "loss": 2.4755, + "step": 21477 + }, + { + "epoch": 1.7218937283277547, + "grad_norm": 1.6666665077209473, + "learning_rate": 9.95710581728767e-07, + "loss": 3.1851, + "step": 21478 + }, + { + "epoch": 1.7219739031087773, + "grad_norm": 1.7143592834472656, + "learning_rate": 9.951458291047067e-07, + "loss": 3.0571, + "step": 21479 + }, + { + "epoch": 1.7220540778898, + "grad_norm": 1.6797927618026733, + "learning_rate": 9.94581228299315e-07, + "loss": 3.2003, + "step": 21480 + }, + { + "epoch": 1.7221342526708225, + "grad_norm": 1.7722042798995972, + "learning_rate": 9.94016779322109e-07, + "loss": 3.3331, + "step": 21481 + }, + { + "epoch": 1.7222144274518452, + "grad_norm": 1.726578712463379, + "learning_rate": 9.934524821826064e-07, + "loss": 3.2074, + "step": 21482 + }, + { + "epoch": 1.7222946022328678, + "grad_norm": 1.7705687284469604, + "learning_rate": 9.92888336890322e-07, + "loss": 3.0865, + "step": 21483 + }, + { + "epoch": 1.7223747770138904, + "grad_norm": 1.7686245441436768, + "learning_rate": 9.923243434547647e-07, + "loss": 3.2059, + "step": 21484 + }, + { + "epoch": 1.722454951794913, + "grad_norm": 1.8339999914169312, + "learning_rate": 9.917605018854437e-07, + "loss": 3.1403, + "step": 21485 + }, + { + "epoch": 1.7225351265759357, + "grad_norm": 1.7619589567184448, + "learning_rate": 9.91196812191867e-07, + "loss": 3.1456, + "step": 21486 + }, + { + "epoch": 1.7226153013569583, + "grad_norm": 1.8275632858276367, + "learning_rate": 9.906332743835356e-07, + "loss": 3.2314, + "step": 21487 + }, + { + "epoch": 1.722695476137981, + "grad_norm": 1.8020570278167725, + "learning_rate": 9.900698884699545e-07, + "loss": 3.1667, + "step": 21488 + }, + { + "epoch": 1.7227756509190035, + "grad_norm": 1.757169246673584, + "learning_rate": 9.89506654460618e-07, + "loss": 3.2062, + "step": 21489 + }, + { + "epoch": 1.7228558257000262, + "grad_norm": 1.7921265363693237, + "learning_rate": 9.889435723650232e-07, + "loss": 3.1356, + "step": 21490 + }, + { + "epoch": 1.7229360004810488, + "grad_norm": 1.7915562391281128, + "learning_rate": 9.883806421926644e-07, + "loss": 3.0794, + "step": 21491 + }, + { + "epoch": 1.7230161752620714, + "grad_norm": 1.8333746194839478, + "learning_rate": 9.878178639530323e-07, + "loss": 3.0334, + "step": 21492 + }, + { + "epoch": 1.723096350043094, + "grad_norm": 1.698445439338684, + "learning_rate": 9.872552376556166e-07, + "loss": 3.1621, + "step": 21493 + }, + { + "epoch": 1.7231765248241167, + "grad_norm": 1.7305777072906494, + "learning_rate": 9.866927633098988e-07, + "loss": 3.1455, + "step": 21494 + }, + { + "epoch": 1.7232566996051393, + "grad_norm": 1.7249358892440796, + "learning_rate": 9.861304409253647e-07, + "loss": 3.2696, + "step": 21495 + }, + { + "epoch": 1.723336874386162, + "grad_norm": 1.6399468183517456, + "learning_rate": 9.855682705114944e-07, + "loss": 3.1413, + "step": 21496 + }, + { + "epoch": 1.7234170491671845, + "grad_norm": 1.8219306468963623, + "learning_rate": 9.85006252077766e-07, + "loss": 3.1486, + "step": 21497 + }, + { + "epoch": 1.7234972239482071, + "grad_norm": 1.8581236600875854, + "learning_rate": 9.844443856336571e-07, + "loss": 3.3417, + "step": 21498 + }, + { + "epoch": 1.7235773987292298, + "grad_norm": 1.7621687650680542, + "learning_rate": 9.838826711886363e-07, + "loss": 3.1315, + "step": 21499 + }, + { + "epoch": 1.7236575735102524, + "grad_norm": 1.8187505006790161, + "learning_rate": 9.833211087521766e-07, + "loss": 3.2012, + "step": 21500 + }, + { + "epoch": 1.723737748291275, + "grad_norm": 1.8562649488449097, + "learning_rate": 9.82759698333744e-07, + "loss": 3.2051, + "step": 21501 + }, + { + "epoch": 1.7238179230722976, + "grad_norm": 1.7525190114974976, + "learning_rate": 9.821984399428064e-07, + "loss": 3.1759, + "step": 21502 + }, + { + "epoch": 1.7238980978533203, + "grad_norm": 1.780909538269043, + "learning_rate": 9.816373335888262e-07, + "loss": 3.3098, + "step": 21503 + }, + { + "epoch": 1.723978272634343, + "grad_norm": 1.8048876523971558, + "learning_rate": 9.810763792812605e-07, + "loss": 3.0942, + "step": 21504 + }, + { + "epoch": 1.7240584474153655, + "grad_norm": 1.8505926132202148, + "learning_rate": 9.805155770295683e-07, + "loss": 3.2073, + "step": 21505 + }, + { + "epoch": 1.7241386221963881, + "grad_norm": 0.6573603749275208, + "learning_rate": 9.79954926843204e-07, + "loss": 2.4816, + "step": 21506 + }, + { + "epoch": 1.7242187969774108, + "grad_norm": 1.800008773803711, + "learning_rate": 9.793944287316215e-07, + "loss": 3.2411, + "step": 21507 + }, + { + "epoch": 1.7242989717584334, + "grad_norm": 1.8299860954284668, + "learning_rate": 9.78834082704272e-07, + "loss": 3.125, + "step": 21508 + }, + { + "epoch": 1.724379146539456, + "grad_norm": 1.780566930770874, + "learning_rate": 9.78273888770598e-07, + "loss": 3.1432, + "step": 21509 + }, + { + "epoch": 1.7244593213204786, + "grad_norm": 1.7394989728927612, + "learning_rate": 9.77713846940046e-07, + "loss": 3.231, + "step": 21510 + }, + { + "epoch": 1.7245394961015013, + "grad_norm": 0.64028400182724, + "learning_rate": 9.771539572220602e-07, + "loss": 2.4727, + "step": 21511 + }, + { + "epoch": 1.7246196708825239, + "grad_norm": 1.7928358316421509, + "learning_rate": 9.76594219626078e-07, + "loss": 3.1969, + "step": 21512 + }, + { + "epoch": 1.7246998456635465, + "grad_norm": 1.7294021844863892, + "learning_rate": 9.76034634161539e-07, + "loss": 3.2103, + "step": 21513 + }, + { + "epoch": 1.7247800204445691, + "grad_norm": 1.84128737449646, + "learning_rate": 9.754752008378732e-07, + "loss": 3.1686, + "step": 21514 + }, + { + "epoch": 1.7248601952255918, + "grad_norm": 1.7818853855133057, + "learning_rate": 9.749159196645164e-07, + "loss": 3.0773, + "step": 21515 + }, + { + "epoch": 1.7249403700066144, + "grad_norm": 1.8580741882324219, + "learning_rate": 9.743567906508956e-07, + "loss": 3.162, + "step": 21516 + }, + { + "epoch": 1.725020544787637, + "grad_norm": 1.703720211982727, + "learning_rate": 9.737978138064375e-07, + "loss": 3.1718, + "step": 21517 + }, + { + "epoch": 1.7251007195686596, + "grad_norm": 1.8593004941940308, + "learning_rate": 9.732389891405703e-07, + "loss": 3.1791, + "step": 21518 + }, + { + "epoch": 1.7251808943496822, + "grad_norm": 1.7118653059005737, + "learning_rate": 9.726803166627107e-07, + "loss": 3.0905, + "step": 21519 + }, + { + "epoch": 1.7252610691307049, + "grad_norm": 1.781356692314148, + "learning_rate": 9.721217963822783e-07, + "loss": 3.1964, + "step": 21520 + }, + { + "epoch": 1.7253412439117275, + "grad_norm": 1.7861437797546387, + "learning_rate": 9.715634283086917e-07, + "loss": 3.1171, + "step": 21521 + }, + { + "epoch": 1.7254214186927501, + "grad_norm": 1.6601752042770386, + "learning_rate": 9.710052124513624e-07, + "loss": 3.0165, + "step": 21522 + }, + { + "epoch": 1.7255015934737727, + "grad_norm": 1.7165826559066772, + "learning_rate": 9.704471488197064e-07, + "loss": 3.179, + "step": 21523 + }, + { + "epoch": 1.7255817682547954, + "grad_norm": 1.734153389930725, + "learning_rate": 9.698892374231273e-07, + "loss": 3.1414, + "step": 21524 + }, + { + "epoch": 1.725661943035818, + "grad_norm": 1.850860357284546, + "learning_rate": 9.693314782710317e-07, + "loss": 3.2151, + "step": 21525 + }, + { + "epoch": 1.7257421178168406, + "grad_norm": 1.7015577554702759, + "learning_rate": 9.687738713728257e-07, + "loss": 3.2511, + "step": 21526 + }, + { + "epoch": 1.7258222925978632, + "grad_norm": 1.9094924926757812, + "learning_rate": 9.682164167379094e-07, + "loss": 3.2766, + "step": 21527 + }, + { + "epoch": 1.7259024673788859, + "grad_norm": 1.826051950454712, + "learning_rate": 9.67659114375683e-07, + "loss": 3.1688, + "step": 21528 + }, + { + "epoch": 1.7259826421599085, + "grad_norm": 1.7485857009887695, + "learning_rate": 9.671019642955382e-07, + "loss": 3.1752, + "step": 21529 + }, + { + "epoch": 1.726062816940931, + "grad_norm": 1.8663426637649536, + "learning_rate": 9.665449665068715e-07, + "loss": 3.2115, + "step": 21530 + }, + { + "epoch": 1.7261429917219537, + "grad_norm": 1.7336146831512451, + "learning_rate": 9.659881210190724e-07, + "loss": 3.052, + "step": 21531 + }, + { + "epoch": 1.7262231665029764, + "grad_norm": 1.7168539762496948, + "learning_rate": 9.6543142784153e-07, + "loss": 3.157, + "step": 21532 + }, + { + "epoch": 1.726303341283999, + "grad_norm": 1.948968529701233, + "learning_rate": 9.648748869836323e-07, + "loss": 3.1954, + "step": 21533 + }, + { + "epoch": 1.7263835160650216, + "grad_norm": 1.7382407188415527, + "learning_rate": 9.643184984547561e-07, + "loss": 3.2103, + "step": 21534 + }, + { + "epoch": 1.7264636908460442, + "grad_norm": 1.7955940961837769, + "learning_rate": 9.637622622642873e-07, + "loss": 3.1447, + "step": 21535 + }, + { + "epoch": 1.7265438656270669, + "grad_norm": 1.7573354244232178, + "learning_rate": 9.632061784216006e-07, + "loss": 3.1104, + "step": 21536 + }, + { + "epoch": 1.7266240404080895, + "grad_norm": 1.7188646793365479, + "learning_rate": 9.626502469360733e-07, + "loss": 3.307, + "step": 21537 + }, + { + "epoch": 1.726704215189112, + "grad_norm": 1.751754879951477, + "learning_rate": 9.620944678170795e-07, + "loss": 3.0539, + "step": 21538 + }, + { + "epoch": 1.726784389970135, + "grad_norm": 0.6785702705383301, + "learning_rate": 9.615388410739868e-07, + "loss": 2.4926, + "step": 21539 + }, + { + "epoch": 1.7268645647511576, + "grad_norm": 1.7938119173049927, + "learning_rate": 9.609833667161628e-07, + "loss": 3.1091, + "step": 21540 + }, + { + "epoch": 1.7269447395321802, + "grad_norm": 1.7705274820327759, + "learning_rate": 9.604280447529736e-07, + "loss": 3.215, + "step": 21541 + }, + { + "epoch": 1.7270249143132028, + "grad_norm": 1.72230863571167, + "learning_rate": 9.598728751937824e-07, + "loss": 3.093, + "step": 21542 + }, + { + "epoch": 1.7271050890942254, + "grad_norm": 1.7037097215652466, + "learning_rate": 9.593178580479502e-07, + "loss": 3.1776, + "step": 21543 + }, + { + "epoch": 1.727185263875248, + "grad_norm": 1.8468152284622192, + "learning_rate": 9.587629933248299e-07, + "loss": 3.1473, + "step": 21544 + }, + { + "epoch": 1.7272654386562707, + "grad_norm": 1.7278629541397095, + "learning_rate": 9.5820828103378e-07, + "loss": 3.207, + "step": 21545 + }, + { + "epoch": 1.7273456134372933, + "grad_norm": 1.7693510055541992, + "learning_rate": 9.576537211841518e-07, + "loss": 3.1415, + "step": 21546 + }, + { + "epoch": 1.727425788218316, + "grad_norm": 1.78860604763031, + "learning_rate": 9.570993137852947e-07, + "loss": 3.2072, + "step": 21547 + }, + { + "epoch": 1.7275059629993386, + "grad_norm": 1.9185357093811035, + "learning_rate": 9.565450588465587e-07, + "loss": 3.2681, + "step": 21548 + }, + { + "epoch": 1.7275861377803612, + "grad_norm": 1.62388277053833, + "learning_rate": 9.55990956377283e-07, + "loss": 3.1864, + "step": 21549 + }, + { + "epoch": 1.7276663125613838, + "grad_norm": 1.7325408458709717, + "learning_rate": 9.554370063868124e-07, + "loss": 3.0865, + "step": 21550 + }, + { + "epoch": 1.7277464873424064, + "grad_norm": 1.6431576013565063, + "learning_rate": 9.548832088844862e-07, + "loss": 3.2173, + "step": 21551 + }, + { + "epoch": 1.727826662123429, + "grad_norm": 1.8095037937164307, + "learning_rate": 9.543295638796423e-07, + "loss": 3.1635, + "step": 21552 + }, + { + "epoch": 1.7279068369044517, + "grad_norm": 1.7055535316467285, + "learning_rate": 9.537760713816146e-07, + "loss": 3.278, + "step": 21553 + }, + { + "epoch": 1.7279870116854743, + "grad_norm": 1.799180269241333, + "learning_rate": 9.532227313997333e-07, + "loss": 3.1084, + "step": 21554 + }, + { + "epoch": 1.728067186466497, + "grad_norm": 1.7042795419692993, + "learning_rate": 9.526695439433275e-07, + "loss": 3.1628, + "step": 21555 + }, + { + "epoch": 1.7281473612475196, + "grad_norm": 1.8103916645050049, + "learning_rate": 9.521165090217255e-07, + "loss": 3.1658, + "step": 21556 + }, + { + "epoch": 1.7282275360285422, + "grad_norm": 1.6958949565887451, + "learning_rate": 9.515636266442497e-07, + "loss": 3.1324, + "step": 21557 + }, + { + "epoch": 1.7283077108095648, + "grad_norm": 1.682600975036621, + "learning_rate": 9.510108968202236e-07, + "loss": 3.1615, + "step": 21558 + }, + { + "epoch": 1.7283878855905876, + "grad_norm": 1.8554171323776245, + "learning_rate": 9.504583195589634e-07, + "loss": 3.1769, + "step": 21559 + }, + { + "epoch": 1.7284680603716103, + "grad_norm": 0.6426724791526794, + "learning_rate": 9.49905894869787e-07, + "loss": 2.47, + "step": 21560 + }, + { + "epoch": 1.728548235152633, + "grad_norm": 1.7174887657165527, + "learning_rate": 9.49353622762007e-07, + "loss": 3.1263, + "step": 21561 + }, + { + "epoch": 1.7286284099336555, + "grad_norm": 0.6547379493713379, + "learning_rate": 9.488015032449349e-07, + "loss": 2.4607, + "step": 21562 + }, + { + "epoch": 1.7287085847146781, + "grad_norm": 1.8802251815795898, + "learning_rate": 9.482495363278821e-07, + "loss": 3.1863, + "step": 21563 + }, + { + "epoch": 1.7287887594957008, + "grad_norm": 1.8405297994613647, + "learning_rate": 9.4769772202015e-07, + "loss": 3.1368, + "step": 21564 + }, + { + "epoch": 1.7288689342767234, + "grad_norm": 1.7204707860946655, + "learning_rate": 9.471460603310434e-07, + "loss": 3.1202, + "step": 21565 + }, + { + "epoch": 1.728949109057746, + "grad_norm": 1.6187318563461304, + "learning_rate": 9.465945512698649e-07, + "loss": 3.1303, + "step": 21566 + }, + { + "epoch": 1.7290292838387686, + "grad_norm": 1.8887263536453247, + "learning_rate": 9.460431948459103e-07, + "loss": 3.241, + "step": 21567 + }, + { + "epoch": 1.7291094586197913, + "grad_norm": 1.795363187789917, + "learning_rate": 9.45491991068479e-07, + "loss": 3.1589, + "step": 21568 + }, + { + "epoch": 1.7291896334008139, + "grad_norm": 1.8181512355804443, + "learning_rate": 9.44940939946859e-07, + "loss": 3.1267, + "step": 21569 + }, + { + "epoch": 1.7292698081818365, + "grad_norm": 1.7969028949737549, + "learning_rate": 9.44390041490344e-07, + "loss": 3.1232, + "step": 21570 + }, + { + "epoch": 1.7293499829628591, + "grad_norm": 1.7347121238708496, + "learning_rate": 9.43839295708221e-07, + "loss": 3.1082, + "step": 21571 + }, + { + "epoch": 1.7294301577438818, + "grad_norm": 1.7381120920181274, + "learning_rate": 9.43288702609777e-07, + "loss": 3.1414, + "step": 21572 + }, + { + "epoch": 1.7295103325249044, + "grad_norm": 1.773824691772461, + "learning_rate": 9.427382622042947e-07, + "loss": 3.2734, + "step": 21573 + }, + { + "epoch": 1.729590507305927, + "grad_norm": 1.7280147075653076, + "learning_rate": 9.42187974501051e-07, + "loss": 3.087, + "step": 21574 + }, + { + "epoch": 1.7296706820869496, + "grad_norm": 1.7810649871826172, + "learning_rate": 9.416378395093262e-07, + "loss": 3.2695, + "step": 21575 + }, + { + "epoch": 1.7297508568679723, + "grad_norm": 1.7998735904693604, + "learning_rate": 9.410878572383941e-07, + "loss": 3.2161, + "step": 21576 + }, + { + "epoch": 1.7298310316489949, + "grad_norm": 1.8380290269851685, + "learning_rate": 9.405380276975285e-07, + "loss": 3.1438, + "step": 21577 + }, + { + "epoch": 1.7299112064300175, + "grad_norm": 1.8567255735397339, + "learning_rate": 9.399883508960017e-07, + "loss": 3.1422, + "step": 21578 + }, + { + "epoch": 1.7299913812110401, + "grad_norm": 1.8525660037994385, + "learning_rate": 9.394388268430754e-07, + "loss": 3.2542, + "step": 21579 + }, + { + "epoch": 1.7300715559920627, + "grad_norm": 1.677146315574646, + "learning_rate": 9.388894555480177e-07, + "loss": 3.171, + "step": 21580 + }, + { + "epoch": 1.7301517307730854, + "grad_norm": 1.800408124923706, + "learning_rate": 9.383402370200901e-07, + "loss": 3.2036, + "step": 21581 + }, + { + "epoch": 1.730231905554108, + "grad_norm": 1.7672266960144043, + "learning_rate": 9.377911712685517e-07, + "loss": 3.171, + "step": 21582 + }, + { + "epoch": 1.7303120803351306, + "grad_norm": 1.7576311826705933, + "learning_rate": 9.372422583026631e-07, + "loss": 3.1052, + "step": 21583 + }, + { + "epoch": 1.7303922551161532, + "grad_norm": 1.698415994644165, + "learning_rate": 9.366934981316744e-07, + "loss": 3.1359, + "step": 21584 + }, + { + "epoch": 1.7304724298971759, + "grad_norm": 1.8460174798965454, + "learning_rate": 9.361448907648385e-07, + "loss": 3.1898, + "step": 21585 + }, + { + "epoch": 1.7305526046781985, + "grad_norm": 1.9142158031463623, + "learning_rate": 9.355964362114056e-07, + "loss": 3.1353, + "step": 21586 + }, + { + "epoch": 1.7306327794592211, + "grad_norm": 1.7843966484069824, + "learning_rate": 9.350481344806217e-07, + "loss": 3.2239, + "step": 21587 + }, + { + "epoch": 1.7307129542402437, + "grad_norm": 1.7825525999069214, + "learning_rate": 9.344999855817338e-07, + "loss": 3.1218, + "step": 21588 + }, + { + "epoch": 1.7307931290212664, + "grad_norm": 1.7904661893844604, + "learning_rate": 9.33951989523979e-07, + "loss": 3.0957, + "step": 21589 + }, + { + "epoch": 1.730873303802289, + "grad_norm": 0.6645745635032654, + "learning_rate": 9.334041463165988e-07, + "loss": 2.4686, + "step": 21590 + }, + { + "epoch": 1.7309534785833116, + "grad_norm": 1.8307632207870483, + "learning_rate": 9.328564559688291e-07, + "loss": 3.1741, + "step": 21591 + }, + { + "epoch": 1.7310336533643342, + "grad_norm": 0.6475135684013367, + "learning_rate": 9.323089184899036e-07, + "loss": 2.4675, + "step": 21592 + }, + { + "epoch": 1.7311138281453569, + "grad_norm": 1.8183687925338745, + "learning_rate": 9.317615338890562e-07, + "loss": 3.1452, + "step": 21593 + }, + { + "epoch": 1.7311940029263795, + "grad_norm": 1.8322676420211792, + "learning_rate": 9.312143021755127e-07, + "loss": 3.2412, + "step": 21594 + }, + { + "epoch": 1.731274177707402, + "grad_norm": 1.7646067142486572, + "learning_rate": 9.30667223358498e-07, + "loss": 3.1868, + "step": 21595 + }, + { + "epoch": 1.7313543524884247, + "grad_norm": 1.799018144607544, + "learning_rate": 9.301202974472389e-07, + "loss": 3.1848, + "step": 21596 + }, + { + "epoch": 1.7314345272694474, + "grad_norm": 1.811421513557434, + "learning_rate": 9.29573524450954e-07, + "loss": 3.2351, + "step": 21597 + }, + { + "epoch": 1.73151470205047, + "grad_norm": 1.8560882806777954, + "learning_rate": 9.290269043788658e-07, + "loss": 3.1948, + "step": 21598 + }, + { + "epoch": 1.7315948768314926, + "grad_norm": 1.8476619720458984, + "learning_rate": 9.284804372401845e-07, + "loss": 3.2474, + "step": 21599 + }, + { + "epoch": 1.7316750516125152, + "grad_norm": 1.704509973526001, + "learning_rate": 9.279341230441252e-07, + "loss": 3.1322, + "step": 21600 + }, + { + "epoch": 1.7317552263935378, + "grad_norm": 1.7728174924850464, + "learning_rate": 9.273879617999004e-07, + "loss": 3.1216, + "step": 21601 + }, + { + "epoch": 1.7318354011745605, + "grad_norm": 1.7959952354431152, + "learning_rate": 9.268419535167172e-07, + "loss": 3.1364, + "step": 21602 + }, + { + "epoch": 1.731915575955583, + "grad_norm": 1.8951109647750854, + "learning_rate": 9.262960982037817e-07, + "loss": 3.2826, + "step": 21603 + }, + { + "epoch": 1.7319957507366057, + "grad_norm": 1.7152016162872314, + "learning_rate": 9.257503958702952e-07, + "loss": 3.1835, + "step": 21604 + }, + { + "epoch": 1.7320759255176283, + "grad_norm": 0.6574459075927734, + "learning_rate": 9.252048465254582e-07, + "loss": 2.4611, + "step": 21605 + }, + { + "epoch": 1.732156100298651, + "grad_norm": 2.2197868824005127, + "learning_rate": 9.2465945017847e-07, + "loss": 3.1192, + "step": 21606 + }, + { + "epoch": 1.7322362750796736, + "grad_norm": 1.7020258903503418, + "learning_rate": 9.241142068385244e-07, + "loss": 3.1179, + "step": 21607 + }, + { + "epoch": 1.7323164498606962, + "grad_norm": 1.8118542432785034, + "learning_rate": 9.235691165148175e-07, + "loss": 3.2052, + "step": 21608 + }, + { + "epoch": 1.7323966246417188, + "grad_norm": 1.898510217666626, + "learning_rate": 9.230241792165329e-07, + "loss": 3.2393, + "step": 21609 + }, + { + "epoch": 1.7324767994227415, + "grad_norm": 1.780443787574768, + "learning_rate": 9.224793949528632e-07, + "loss": 3.238, + "step": 21610 + }, + { + "epoch": 1.732556974203764, + "grad_norm": 1.8345707654953003, + "learning_rate": 9.219347637329912e-07, + "loss": 3.2389, + "step": 21611 + }, + { + "epoch": 1.7326371489847867, + "grad_norm": 1.7185348272323608, + "learning_rate": 9.213902855660994e-07, + "loss": 3.1462, + "step": 21612 + }, + { + "epoch": 1.7327173237658093, + "grad_norm": 1.7455512285232544, + "learning_rate": 9.208459604613695e-07, + "loss": 3.105, + "step": 21613 + }, + { + "epoch": 1.732797498546832, + "grad_norm": 1.6612540483474731, + "learning_rate": 9.203017884279753e-07, + "loss": 3.09, + "step": 21614 + }, + { + "epoch": 1.7328776733278546, + "grad_norm": 1.6910760402679443, + "learning_rate": 9.197577694750925e-07, + "loss": 3.2312, + "step": 21615 + }, + { + "epoch": 1.7329578481088772, + "grad_norm": 1.7502849102020264, + "learning_rate": 9.192139036118941e-07, + "loss": 3.0555, + "step": 21616 + }, + { + "epoch": 1.7330380228898998, + "grad_norm": 1.8616416454315186, + "learning_rate": 9.186701908475493e-07, + "loss": 3.1932, + "step": 21617 + }, + { + "epoch": 1.7331181976709225, + "grad_norm": 1.76664137840271, + "learning_rate": 9.181266311912252e-07, + "loss": 3.1583, + "step": 21618 + }, + { + "epoch": 1.733198372451945, + "grad_norm": 1.9346544742584229, + "learning_rate": 9.175832246520844e-07, + "loss": 3.1636, + "step": 21619 + }, + { + "epoch": 1.7332785472329677, + "grad_norm": 1.8222241401672363, + "learning_rate": 9.170399712392885e-07, + "loss": 3.2373, + "step": 21620 + }, + { + "epoch": 1.7333587220139905, + "grad_norm": 1.794299840927124, + "learning_rate": 9.164968709619981e-07, + "loss": 3.2155, + "step": 21621 + }, + { + "epoch": 1.7334388967950132, + "grad_norm": 1.8537406921386719, + "learning_rate": 9.159539238293702e-07, + "loss": 3.2744, + "step": 21622 + }, + { + "epoch": 1.7335190715760358, + "grad_norm": 1.8721519708633423, + "learning_rate": 9.154111298505574e-07, + "loss": 3.1768, + "step": 21623 + }, + { + "epoch": 1.7335992463570584, + "grad_norm": 1.7785660028457642, + "learning_rate": 9.148684890347115e-07, + "loss": 3.0864, + "step": 21624 + }, + { + "epoch": 1.733679421138081, + "grad_norm": 1.8563098907470703, + "learning_rate": 9.143260013909794e-07, + "loss": 3.2412, + "step": 21625 + }, + { + "epoch": 1.7337595959191037, + "grad_norm": 1.7004590034484863, + "learning_rate": 9.137836669285093e-07, + "loss": 3.1172, + "step": 21626 + }, + { + "epoch": 1.7338397707001263, + "grad_norm": 1.9070155620574951, + "learning_rate": 9.132414856564442e-07, + "loss": 3.1896, + "step": 21627 + }, + { + "epoch": 1.733919945481149, + "grad_norm": 1.8096262216567993, + "learning_rate": 9.126994575839277e-07, + "loss": 3.2161, + "step": 21628 + }, + { + "epoch": 1.7340001202621715, + "grad_norm": 1.7219431400299072, + "learning_rate": 9.121575827200935e-07, + "loss": 3.2292, + "step": 21629 + }, + { + "epoch": 1.7340802950431942, + "grad_norm": 1.755564570426941, + "learning_rate": 9.116158610740789e-07, + "loss": 3.1335, + "step": 21630 + }, + { + "epoch": 1.7341604698242168, + "grad_norm": 0.6602195501327515, + "learning_rate": 9.110742926550187e-07, + "loss": 2.4561, + "step": 21631 + }, + { + "epoch": 1.7342406446052394, + "grad_norm": 1.76458740234375, + "learning_rate": 9.105328774720412e-07, + "loss": 3.1478, + "step": 21632 + }, + { + "epoch": 1.734320819386262, + "grad_norm": 1.8083782196044922, + "learning_rate": 9.099916155342792e-07, + "loss": 3.2148, + "step": 21633 + }, + { + "epoch": 1.7344009941672847, + "grad_norm": 0.660686194896698, + "learning_rate": 9.09450506850853e-07, + "loss": 2.4492, + "step": 21634 + }, + { + "epoch": 1.7344811689483073, + "grad_norm": 1.7688921689987183, + "learning_rate": 9.089095514308865e-07, + "loss": 3.1669, + "step": 21635 + }, + { + "epoch": 1.73456134372933, + "grad_norm": 1.7577112913131714, + "learning_rate": 9.083687492835024e-07, + "loss": 3.201, + "step": 21636 + }, + { + "epoch": 1.7346415185103525, + "grad_norm": 1.7506252527236938, + "learning_rate": 9.078281004178158e-07, + "loss": 3.1052, + "step": 21637 + }, + { + "epoch": 1.7347216932913752, + "grad_norm": 1.7139934301376343, + "learning_rate": 9.072876048429436e-07, + "loss": 3.1923, + "step": 21638 + }, + { + "epoch": 1.7348018680723978, + "grad_norm": 1.7823387384414673, + "learning_rate": 9.067472625679996e-07, + "loss": 3.0989, + "step": 21639 + }, + { + "epoch": 1.7348820428534204, + "grad_norm": 1.724469780921936, + "learning_rate": 9.0620707360209e-07, + "loss": 3.1069, + "step": 21640 + }, + { + "epoch": 1.7349622176344432, + "grad_norm": 1.8339893817901611, + "learning_rate": 9.056670379543242e-07, + "loss": 3.1713, + "step": 21641 + }, + { + "epoch": 1.7350423924154659, + "grad_norm": 1.7931045293807983, + "learning_rate": 9.051271556338071e-07, + "loss": 3.2494, + "step": 21642 + }, + { + "epoch": 1.7351225671964885, + "grad_norm": 1.719124436378479, + "learning_rate": 9.045874266496402e-07, + "loss": 3.1067, + "step": 21643 + }, + { + "epoch": 1.7352027419775111, + "grad_norm": 1.7624516487121582, + "learning_rate": 9.040478510109263e-07, + "loss": 3.0862, + "step": 21644 + }, + { + "epoch": 1.7352829167585337, + "grad_norm": 1.7524112462997437, + "learning_rate": 9.035084287267582e-07, + "loss": 3.2908, + "step": 21645 + }, + { + "epoch": 1.7353630915395564, + "grad_norm": 1.8828126192092896, + "learning_rate": 9.029691598062318e-07, + "loss": 3.286, + "step": 21646 + }, + { + "epoch": 1.735443266320579, + "grad_norm": 1.863551139831543, + "learning_rate": 9.024300442584388e-07, + "loss": 3.1831, + "step": 21647 + }, + { + "epoch": 1.7355234411016016, + "grad_norm": 1.689084529876709, + "learning_rate": 9.018910820924698e-07, + "loss": 3.1456, + "step": 21648 + }, + { + "epoch": 1.7356036158826242, + "grad_norm": 1.818200707435608, + "learning_rate": 9.013522733174118e-07, + "loss": 3.2382, + "step": 21649 + }, + { + "epoch": 1.7356837906636469, + "grad_norm": 1.819350004196167, + "learning_rate": 9.008136179423454e-07, + "loss": 3.201, + "step": 21650 + }, + { + "epoch": 1.7357639654446695, + "grad_norm": 1.9396730661392212, + "learning_rate": 9.002751159763556e-07, + "loss": 3.2629, + "step": 21651 + }, + { + "epoch": 1.7358441402256921, + "grad_norm": 0.6692875623703003, + "learning_rate": 8.997367674285206e-07, + "loss": 2.4631, + "step": 21652 + }, + { + "epoch": 1.7359243150067147, + "grad_norm": 1.7430421113967896, + "learning_rate": 8.991985723079155e-07, + "loss": 3.2275, + "step": 21653 + }, + { + "epoch": 1.7360044897877374, + "grad_norm": 1.8465197086334229, + "learning_rate": 8.986605306236174e-07, + "loss": 3.0941, + "step": 21654 + }, + { + "epoch": 1.73608466456876, + "grad_norm": 1.8263578414916992, + "learning_rate": 8.981226423846934e-07, + "loss": 3.1921, + "step": 21655 + }, + { + "epoch": 1.7361648393497826, + "grad_norm": 1.7782959938049316, + "learning_rate": 8.97584907600213e-07, + "loss": 3.1024, + "step": 21656 + }, + { + "epoch": 1.7362450141308052, + "grad_norm": 1.7388724088668823, + "learning_rate": 8.970473262792445e-07, + "loss": 3.163, + "step": 21657 + }, + { + "epoch": 1.7363251889118279, + "grad_norm": 1.768725872039795, + "learning_rate": 8.965098984308484e-07, + "loss": 3.1368, + "step": 21658 + }, + { + "epoch": 1.7364053636928505, + "grad_norm": 1.825448751449585, + "learning_rate": 8.959726240640898e-07, + "loss": 3.1593, + "step": 21659 + }, + { + "epoch": 1.736485538473873, + "grad_norm": 0.6537943482398987, + "learning_rate": 8.954355031880224e-07, + "loss": 2.4698, + "step": 21660 + }, + { + "epoch": 1.7365657132548957, + "grad_norm": 1.762514352798462, + "learning_rate": 8.948985358117023e-07, + "loss": 3.1968, + "step": 21661 + }, + { + "epoch": 1.7366458880359184, + "grad_norm": 1.717703938484192, + "learning_rate": 8.943617219441858e-07, + "loss": 3.1138, + "step": 21662 + }, + { + "epoch": 1.736726062816941, + "grad_norm": 1.6514029502868652, + "learning_rate": 8.938250615945199e-07, + "loss": 3.0914, + "step": 21663 + }, + { + "epoch": 1.7368062375979636, + "grad_norm": 1.9670792818069458, + "learning_rate": 8.932885547717573e-07, + "loss": 3.1377, + "step": 21664 + }, + { + "epoch": 1.7368864123789862, + "grad_norm": 1.646393060684204, + "learning_rate": 8.927522014849377e-07, + "loss": 3.1819, + "step": 21665 + }, + { + "epoch": 1.7369665871600088, + "grad_norm": 1.894281029701233, + "learning_rate": 8.92216001743107e-07, + "loss": 3.2274, + "step": 21666 + }, + { + "epoch": 1.7370467619410315, + "grad_norm": 1.816270112991333, + "learning_rate": 8.916799555553035e-07, + "loss": 3.2181, + "step": 21667 + }, + { + "epoch": 1.737126936722054, + "grad_norm": 1.7599483728408813, + "learning_rate": 8.911440629305656e-07, + "loss": 3.1451, + "step": 21668 + }, + { + "epoch": 1.7372071115030767, + "grad_norm": 1.9523499011993408, + "learning_rate": 8.906083238779318e-07, + "loss": 3.1688, + "step": 21669 + }, + { + "epoch": 1.7372872862840993, + "grad_norm": 1.6850287914276123, + "learning_rate": 8.900727384064289e-07, + "loss": 3.1129, + "step": 21670 + }, + { + "epoch": 1.737367461065122, + "grad_norm": 1.715410828590393, + "learning_rate": 8.895373065250878e-07, + "loss": 3.1948, + "step": 21671 + }, + { + "epoch": 1.7374476358461446, + "grad_norm": 1.70268976688385, + "learning_rate": 8.890020282429379e-07, + "loss": 3.0953, + "step": 21672 + }, + { + "epoch": 1.7375278106271672, + "grad_norm": 1.7294598817825317, + "learning_rate": 8.88466903569003e-07, + "loss": 3.0798, + "step": 21673 + }, + { + "epoch": 1.7376079854081898, + "grad_norm": 1.7816129922866821, + "learning_rate": 8.87931932512307e-07, + "loss": 3.0731, + "step": 21674 + }, + { + "epoch": 1.7376881601892125, + "grad_norm": 1.848605990409851, + "learning_rate": 8.873971150818639e-07, + "loss": 3.2052, + "step": 21675 + }, + { + "epoch": 1.737768334970235, + "grad_norm": 1.7535916566848755, + "learning_rate": 8.868624512866952e-07, + "loss": 3.2047, + "step": 21676 + }, + { + "epoch": 1.7378485097512577, + "grad_norm": 1.7844717502593994, + "learning_rate": 8.863279411358139e-07, + "loss": 3.176, + "step": 21677 + }, + { + "epoch": 1.7379286845322803, + "grad_norm": 1.8629392385482788, + "learning_rate": 8.857935846382304e-07, + "loss": 3.1962, + "step": 21678 + }, + { + "epoch": 1.738008859313303, + "grad_norm": 1.7787827253341675, + "learning_rate": 8.852593818029575e-07, + "loss": 3.2626, + "step": 21679 + }, + { + "epoch": 1.7380890340943256, + "grad_norm": 1.7811017036437988, + "learning_rate": 8.847253326389971e-07, + "loss": 3.1588, + "step": 21680 + }, + { + "epoch": 1.7381692088753482, + "grad_norm": 1.7568416595458984, + "learning_rate": 8.841914371553551e-07, + "loss": 3.2514, + "step": 21681 + }, + { + "epoch": 1.7382493836563708, + "grad_norm": 1.8067960739135742, + "learning_rate": 8.836576953610321e-07, + "loss": 3.2874, + "step": 21682 + }, + { + "epoch": 1.7383295584373935, + "grad_norm": 2.1368775367736816, + "learning_rate": 8.831241072650287e-07, + "loss": 3.1049, + "step": 21683 + }, + { + "epoch": 1.738409733218416, + "grad_norm": 1.828506588935852, + "learning_rate": 8.8259067287634e-07, + "loss": 3.2347, + "step": 21684 + }, + { + "epoch": 1.7384899079994387, + "grad_norm": 1.8781157732009888, + "learning_rate": 8.820573922039588e-07, + "loss": 3.138, + "step": 21685 + }, + { + "epoch": 1.7385700827804613, + "grad_norm": 1.7324682474136353, + "learning_rate": 8.815242652568756e-07, + "loss": 3.1013, + "step": 21686 + }, + { + "epoch": 1.738650257561484, + "grad_norm": 1.6764885187149048, + "learning_rate": 8.8099129204408e-07, + "loss": 3.1595, + "step": 21687 + }, + { + "epoch": 1.7387304323425066, + "grad_norm": 1.7375463247299194, + "learning_rate": 8.80458472574558e-07, + "loss": 3.1264, + "step": 21688 + }, + { + "epoch": 1.7388106071235292, + "grad_norm": 1.7714439630508423, + "learning_rate": 8.799258068572935e-07, + "loss": 3.1384, + "step": 21689 + }, + { + "epoch": 1.7388907819045518, + "grad_norm": 1.7483261823654175, + "learning_rate": 8.793932949012629e-07, + "loss": 3.045, + "step": 21690 + }, + { + "epoch": 1.7389709566855744, + "grad_norm": 1.8271043300628662, + "learning_rate": 8.788609367154477e-07, + "loss": 3.0913, + "step": 21691 + }, + { + "epoch": 1.739051131466597, + "grad_norm": 1.7540677785873413, + "learning_rate": 8.78328732308823e-07, + "loss": 3.1055, + "step": 21692 + }, + { + "epoch": 1.7391313062476197, + "grad_norm": 1.76109778881073, + "learning_rate": 8.777966816903605e-07, + "loss": 3.2195, + "step": 21693 + }, + { + "epoch": 1.7392114810286423, + "grad_norm": 1.7281053066253662, + "learning_rate": 8.77264784869033e-07, + "loss": 3.0885, + "step": 21694 + }, + { + "epoch": 1.739291655809665, + "grad_norm": 1.7977519035339355, + "learning_rate": 8.767330418538045e-07, + "loss": 3.2067, + "step": 21695 + }, + { + "epoch": 1.7393718305906876, + "grad_norm": 1.7461848258972168, + "learning_rate": 8.762014526536411e-07, + "loss": 3.203, + "step": 21696 + }, + { + "epoch": 1.7394520053717102, + "grad_norm": 1.7945739030838013, + "learning_rate": 8.756700172775057e-07, + "loss": 3.0922, + "step": 21697 + }, + { + "epoch": 1.7395321801527328, + "grad_norm": 1.791827917098999, + "learning_rate": 8.751387357343577e-07, + "loss": 3.1464, + "step": 21698 + }, + { + "epoch": 1.7396123549337554, + "grad_norm": 1.697690725326538, + "learning_rate": 8.746076080331578e-07, + "loss": 3.1732, + "step": 21699 + }, + { + "epoch": 1.739692529714778, + "grad_norm": 1.7951734066009521, + "learning_rate": 8.740766341828544e-07, + "loss": 3.182, + "step": 21700 + }, + { + "epoch": 1.7397727044958007, + "grad_norm": 1.81728196144104, + "learning_rate": 8.735458141924036e-07, + "loss": 3.1772, + "step": 21701 + }, + { + "epoch": 1.7398528792768235, + "grad_norm": 1.8666189908981323, + "learning_rate": 8.730151480707527e-07, + "loss": 3.0844, + "step": 21702 + }, + { + "epoch": 1.7399330540578462, + "grad_norm": 1.7494580745697021, + "learning_rate": 8.724846358268513e-07, + "loss": 3.1336, + "step": 21703 + }, + { + "epoch": 1.7400132288388688, + "grad_norm": 1.7177425622940063, + "learning_rate": 8.719542774696433e-07, + "loss": 3.1575, + "step": 21704 + }, + { + "epoch": 1.7400934036198914, + "grad_norm": 1.8157216310501099, + "learning_rate": 8.714240730080669e-07, + "loss": 3.1295, + "step": 21705 + }, + { + "epoch": 1.740173578400914, + "grad_norm": 1.7419692277908325, + "learning_rate": 8.708940224510632e-07, + "loss": 3.2446, + "step": 21706 + }, + { + "epoch": 1.7402537531819366, + "grad_norm": 1.7740710973739624, + "learning_rate": 8.70364125807569e-07, + "loss": 3.2035, + "step": 21707 + }, + { + "epoch": 1.7403339279629593, + "grad_norm": 1.6960632801055908, + "learning_rate": 8.698343830865186e-07, + "loss": 3.2032, + "step": 21708 + }, + { + "epoch": 1.740414102743982, + "grad_norm": 1.7710412740707397, + "learning_rate": 8.693047942968447e-07, + "loss": 3.2282, + "step": 21709 + }, + { + "epoch": 1.7404942775250045, + "grad_norm": 1.7168077230453491, + "learning_rate": 8.687753594474712e-07, + "loss": 3.0748, + "step": 21710 + }, + { + "epoch": 1.7405744523060271, + "grad_norm": 0.6611255407333374, + "learning_rate": 8.682460785473279e-07, + "loss": 2.4728, + "step": 21711 + }, + { + "epoch": 1.7406546270870498, + "grad_norm": 1.8364717960357666, + "learning_rate": 8.677169516053374e-07, + "loss": 3.2463, + "step": 21712 + }, + { + "epoch": 1.7407348018680724, + "grad_norm": 1.7869969606399536, + "learning_rate": 8.671879786304205e-07, + "loss": 3.1903, + "step": 21713 + }, + { + "epoch": 1.740814976649095, + "grad_norm": 1.6890922784805298, + "learning_rate": 8.666591596314966e-07, + "loss": 3.2455, + "step": 21714 + }, + { + "epoch": 1.7408951514301176, + "grad_norm": 1.8323347568511963, + "learning_rate": 8.661304946174787e-07, + "loss": 3.2571, + "step": 21715 + }, + { + "epoch": 1.7409753262111403, + "grad_norm": 1.8246241807937622, + "learning_rate": 8.656019835972818e-07, + "loss": 3.1859, + "step": 21716 + }, + { + "epoch": 1.7410555009921629, + "grad_norm": 1.8061102628707886, + "learning_rate": 8.650736265798165e-07, + "loss": 3.2377, + "step": 21717 + }, + { + "epoch": 1.7411356757731855, + "grad_norm": 1.7340540885925293, + "learning_rate": 8.645454235739903e-07, + "loss": 3.1118, + "step": 21718 + }, + { + "epoch": 1.7412158505542081, + "grad_norm": 1.7816721200942993, + "learning_rate": 8.640173745887104e-07, + "loss": 3.1804, + "step": 21719 + }, + { + "epoch": 1.7412960253352308, + "grad_norm": 1.7506756782531738, + "learning_rate": 8.634894796328764e-07, + "loss": 3.1533, + "step": 21720 + }, + { + "epoch": 1.7413762001162534, + "grad_norm": 1.7732927799224854, + "learning_rate": 8.629617387153888e-07, + "loss": 3.1309, + "step": 21721 + }, + { + "epoch": 1.7414563748972762, + "grad_norm": 1.6114118099212646, + "learning_rate": 8.624341518451462e-07, + "loss": 3.131, + "step": 21722 + }, + { + "epoch": 1.7415365496782989, + "grad_norm": 2.0690553188323975, + "learning_rate": 8.619067190310437e-07, + "loss": 3.2054, + "step": 21723 + }, + { + "epoch": 1.7416167244593215, + "grad_norm": 2.065420627593994, + "learning_rate": 8.613794402819752e-07, + "loss": 3.1875, + "step": 21724 + }, + { + "epoch": 1.741696899240344, + "grad_norm": 1.8775789737701416, + "learning_rate": 8.608523156068271e-07, + "loss": 3.1184, + "step": 21725 + }, + { + "epoch": 1.7417770740213667, + "grad_norm": 1.7764109373092651, + "learning_rate": 8.603253450144866e-07, + "loss": 3.2045, + "step": 21726 + }, + { + "epoch": 1.7418572488023893, + "grad_norm": 1.7274967432022095, + "learning_rate": 8.597985285138411e-07, + "loss": 3.1544, + "step": 21727 + }, + { + "epoch": 1.741937423583412, + "grad_norm": 1.8283153772354126, + "learning_rate": 8.592718661137711e-07, + "loss": 3.2207, + "step": 21728 + }, + { + "epoch": 1.7420175983644346, + "grad_norm": 1.7373358011245728, + "learning_rate": 8.587453578231564e-07, + "loss": 3.1023, + "step": 21729 + }, + { + "epoch": 1.7420977731454572, + "grad_norm": 1.756283164024353, + "learning_rate": 8.58219003650873e-07, + "loss": 3.1732, + "step": 21730 + }, + { + "epoch": 1.7421779479264798, + "grad_norm": 0.6583598852157593, + "learning_rate": 8.57692803605794e-07, + "loss": 2.4615, + "step": 21731 + }, + { + "epoch": 1.7422581227075025, + "grad_norm": 1.7993851900100708, + "learning_rate": 8.571667576967935e-07, + "loss": 3.1884, + "step": 21732 + }, + { + "epoch": 1.742338297488525, + "grad_norm": 1.7824649810791016, + "learning_rate": 8.566408659327386e-07, + "loss": 3.1506, + "step": 21733 + }, + { + "epoch": 1.7424184722695477, + "grad_norm": 1.7258001565933228, + "learning_rate": 8.561151283224978e-07, + "loss": 3.2351, + "step": 21734 + }, + { + "epoch": 1.7424986470505703, + "grad_norm": 1.8991010189056396, + "learning_rate": 8.555895448749319e-07, + "loss": 3.2247, + "step": 21735 + }, + { + "epoch": 1.742578821831593, + "grad_norm": 1.9124255180358887, + "learning_rate": 8.550641155989037e-07, + "loss": 3.1828, + "step": 21736 + }, + { + "epoch": 1.7426589966126156, + "grad_norm": 1.7920693159103394, + "learning_rate": 8.545388405032717e-07, + "loss": 3.1108, + "step": 21737 + }, + { + "epoch": 1.7427391713936382, + "grad_norm": 1.8164513111114502, + "learning_rate": 8.540137195968912e-07, + "loss": 3.1629, + "step": 21738 + }, + { + "epoch": 1.7428193461746608, + "grad_norm": 1.8026081323623657, + "learning_rate": 8.534887528886171e-07, + "loss": 3.2024, + "step": 21739 + }, + { + "epoch": 1.7428995209556835, + "grad_norm": 1.7798161506652832, + "learning_rate": 8.529639403872981e-07, + "loss": 3.1027, + "step": 21740 + }, + { + "epoch": 1.742979695736706, + "grad_norm": 1.7898882627487183, + "learning_rate": 8.524392821017835e-07, + "loss": 3.1372, + "step": 21741 + }, + { + "epoch": 1.7430598705177287, + "grad_norm": 1.7806740999221802, + "learning_rate": 8.519147780409188e-07, + "loss": 3.1409, + "step": 21742 + }, + { + "epoch": 1.7431400452987513, + "grad_norm": 1.7966889142990112, + "learning_rate": 8.513904282135466e-07, + "loss": 3.1852, + "step": 21743 + }, + { + "epoch": 1.743220220079774, + "grad_norm": 1.5921860933303833, + "learning_rate": 8.508662326285089e-07, + "loss": 3.1393, + "step": 21744 + }, + { + "epoch": 1.7433003948607966, + "grad_norm": 1.8107361793518066, + "learning_rate": 8.503421912946408e-07, + "loss": 3.2012, + "step": 21745 + }, + { + "epoch": 1.7433805696418192, + "grad_norm": 1.7331758737564087, + "learning_rate": 8.498183042207786e-07, + "loss": 3.1656, + "step": 21746 + }, + { + "epoch": 1.7434607444228418, + "grad_norm": 1.819563388824463, + "learning_rate": 8.492945714157541e-07, + "loss": 3.2656, + "step": 21747 + }, + { + "epoch": 1.7435409192038644, + "grad_norm": 1.8937945365905762, + "learning_rate": 8.487709928883991e-07, + "loss": 3.1799, + "step": 21748 + }, + { + "epoch": 1.743621093984887, + "grad_norm": 1.7679246664047241, + "learning_rate": 8.482475686475411e-07, + "loss": 3.1099, + "step": 21749 + }, + { + "epoch": 1.7437012687659097, + "grad_norm": 0.6881504654884338, + "learning_rate": 8.477242987020029e-07, + "loss": 2.4889, + "step": 21750 + }, + { + "epoch": 1.7437814435469323, + "grad_norm": 1.9330408573150635, + "learning_rate": 8.472011830606064e-07, + "loss": 3.1717, + "step": 21751 + }, + { + "epoch": 1.743861618327955, + "grad_norm": 1.886109471321106, + "learning_rate": 8.466782217321723e-07, + "loss": 3.1804, + "step": 21752 + }, + { + "epoch": 1.7439417931089776, + "grad_norm": 1.7257956266403198, + "learning_rate": 8.46155414725518e-07, + "loss": 3.1808, + "step": 21753 + }, + { + "epoch": 1.7440219678900002, + "grad_norm": 0.6686102151870728, + "learning_rate": 8.456327620494586e-07, + "loss": 2.4663, + "step": 21754 + }, + { + "epoch": 1.7441021426710228, + "grad_norm": 1.827889084815979, + "learning_rate": 8.451102637128039e-07, + "loss": 3.1041, + "step": 21755 + }, + { + "epoch": 1.7441823174520454, + "grad_norm": 0.6777212023735046, + "learning_rate": 8.445879197243623e-07, + "loss": 2.4717, + "step": 21756 + }, + { + "epoch": 1.744262492233068, + "grad_norm": 1.7456461191177368, + "learning_rate": 8.440657300929423e-07, + "loss": 3.2069, + "step": 21757 + }, + { + "epoch": 1.7443426670140907, + "grad_norm": 1.8717728853225708, + "learning_rate": 8.435436948273468e-07, + "loss": 3.1884, + "step": 21758 + }, + { + "epoch": 1.7444228417951133, + "grad_norm": 1.833472728729248, + "learning_rate": 8.430218139363799e-07, + "loss": 3.2731, + "step": 21759 + }, + { + "epoch": 1.744503016576136, + "grad_norm": 1.8410112857818604, + "learning_rate": 8.425000874288358e-07, + "loss": 3.1943, + "step": 21760 + }, + { + "epoch": 1.7445831913571586, + "grad_norm": 1.6809391975402832, + "learning_rate": 8.41978515313513e-07, + "loss": 3.0806, + "step": 21761 + }, + { + "epoch": 1.7446633661381812, + "grad_norm": 1.6685751676559448, + "learning_rate": 8.414570975992043e-07, + "loss": 3.1206, + "step": 21762 + }, + { + "epoch": 1.7447435409192038, + "grad_norm": 1.7846897840499878, + "learning_rate": 8.409358342947016e-07, + "loss": 3.1832, + "step": 21763 + }, + { + "epoch": 1.7448237157002264, + "grad_norm": 1.7447350025177002, + "learning_rate": 8.404147254087947e-07, + "loss": 3.1686, + "step": 21764 + }, + { + "epoch": 1.744903890481249, + "grad_norm": 0.6595573425292969, + "learning_rate": 8.398937709502652e-07, + "loss": 2.4528, + "step": 21765 + }, + { + "epoch": 1.7449840652622717, + "grad_norm": 1.8852211236953735, + "learning_rate": 8.393729709278986e-07, + "loss": 3.1876, + "step": 21766 + }, + { + "epoch": 1.7450642400432943, + "grad_norm": 1.6721917390823364, + "learning_rate": 8.388523253504754e-07, + "loss": 3.0573, + "step": 21767 + }, + { + "epoch": 1.745144414824317, + "grad_norm": 1.6490018367767334, + "learning_rate": 8.383318342267732e-07, + "loss": 3.1216, + "step": 21768 + }, + { + "epoch": 1.7452245896053395, + "grad_norm": 1.8106086254119873, + "learning_rate": 8.378114975655683e-07, + "loss": 3.0754, + "step": 21769 + }, + { + "epoch": 1.7453047643863622, + "grad_norm": 1.791067361831665, + "learning_rate": 8.372913153756324e-07, + "loss": 3.1095, + "step": 21770 + }, + { + "epoch": 1.7453849391673848, + "grad_norm": 2.168459177017212, + "learning_rate": 8.367712876657341e-07, + "loss": 3.205, + "step": 21771 + }, + { + "epoch": 1.7454651139484074, + "grad_norm": 1.853767991065979, + "learning_rate": 8.362514144446432e-07, + "loss": 3.3158, + "step": 21772 + }, + { + "epoch": 1.74554528872943, + "grad_norm": 1.760725498199463, + "learning_rate": 8.357316957211237e-07, + "loss": 3.2205, + "step": 21773 + }, + { + "epoch": 1.7456254635104527, + "grad_norm": 1.8124626874923706, + "learning_rate": 8.352121315039396e-07, + "loss": 3.1649, + "step": 21774 + }, + { + "epoch": 1.7457056382914753, + "grad_norm": 1.7592706680297852, + "learning_rate": 8.346927218018474e-07, + "loss": 3.1966, + "step": 21775 + }, + { + "epoch": 1.745785813072498, + "grad_norm": 1.693711280822754, + "learning_rate": 8.341734666236057e-07, + "loss": 3.1208, + "step": 21776 + }, + { + "epoch": 1.7458659878535205, + "grad_norm": 0.636367678642273, + "learning_rate": 8.336543659779694e-07, + "loss": 2.4646, + "step": 21777 + }, + { + "epoch": 1.7459461626345432, + "grad_norm": 1.7900352478027344, + "learning_rate": 8.331354198736897e-07, + "loss": 3.1993, + "step": 21778 + }, + { + "epoch": 1.7460263374155658, + "grad_norm": 1.6916522979736328, + "learning_rate": 8.32616628319518e-07, + "loss": 3.1454, + "step": 21779 + }, + { + "epoch": 1.7461065121965884, + "grad_norm": 1.7717705965042114, + "learning_rate": 8.320979913241966e-07, + "loss": 3.1858, + "step": 21780 + }, + { + "epoch": 1.746186686977611, + "grad_norm": 1.6999287605285645, + "learning_rate": 8.315795088964729e-07, + "loss": 3.0632, + "step": 21781 + }, + { + "epoch": 1.7462668617586337, + "grad_norm": 1.7096186876296997, + "learning_rate": 8.310611810450864e-07, + "loss": 3.1925, + "step": 21782 + }, + { + "epoch": 1.7463470365396565, + "grad_norm": 1.8148133754730225, + "learning_rate": 8.30543007778778e-07, + "loss": 3.2572, + "step": 21783 + }, + { + "epoch": 1.7464272113206791, + "grad_norm": 1.7718530893325806, + "learning_rate": 8.300249891062839e-07, + "loss": 3.2673, + "step": 21784 + }, + { + "epoch": 1.7465073861017018, + "grad_norm": 0.6687977910041809, + "learning_rate": 8.295071250363351e-07, + "loss": 2.4674, + "step": 21785 + }, + { + "epoch": 1.7465875608827244, + "grad_norm": 1.8856309652328491, + "learning_rate": 8.289894155776646e-07, + "loss": 3.1823, + "step": 21786 + }, + { + "epoch": 1.746667735663747, + "grad_norm": 0.7196207046508789, + "learning_rate": 8.284718607389997e-07, + "loss": 2.4927, + "step": 21787 + }, + { + "epoch": 1.7467479104447696, + "grad_norm": 1.856880784034729, + "learning_rate": 8.279544605290668e-07, + "loss": 3.0671, + "step": 21788 + }, + { + "epoch": 1.7468280852257922, + "grad_norm": 1.815016508102417, + "learning_rate": 8.274372149565913e-07, + "loss": 3.1594, + "step": 21789 + }, + { + "epoch": 1.7469082600068149, + "grad_norm": 1.7432981729507446, + "learning_rate": 8.269201240302893e-07, + "loss": 3.1404, + "step": 21790 + }, + { + "epoch": 1.7469884347878375, + "grad_norm": 1.817713737487793, + "learning_rate": 8.264031877588819e-07, + "loss": 3.1103, + "step": 21791 + }, + { + "epoch": 1.7470686095688601, + "grad_norm": 1.8781557083129883, + "learning_rate": 8.25886406151084e-07, + "loss": 3.1754, + "step": 21792 + }, + { + "epoch": 1.7471487843498827, + "grad_norm": 1.7613836526870728, + "learning_rate": 8.253697792156068e-07, + "loss": 3.1479, + "step": 21793 + }, + { + "epoch": 1.7472289591309054, + "grad_norm": 1.7736705541610718, + "learning_rate": 8.248533069611641e-07, + "loss": 3.1204, + "step": 21794 + }, + { + "epoch": 1.747309133911928, + "grad_norm": 1.6726926565170288, + "learning_rate": 8.243369893964592e-07, + "loss": 3.1923, + "step": 21795 + }, + { + "epoch": 1.7473893086929506, + "grad_norm": 2.1944479942321777, + "learning_rate": 8.238208265301994e-07, + "loss": 3.1335, + "step": 21796 + }, + { + "epoch": 1.7474694834739732, + "grad_norm": 1.7657383680343628, + "learning_rate": 8.233048183710857e-07, + "loss": 3.1046, + "step": 21797 + }, + { + "epoch": 1.7475496582549959, + "grad_norm": 1.8606436252593994, + "learning_rate": 8.227889649278187e-07, + "loss": 3.1861, + "step": 21798 + }, + { + "epoch": 1.7476298330360185, + "grad_norm": 1.777082920074463, + "learning_rate": 8.222732662090971e-07, + "loss": 3.2148, + "step": 21799 + }, + { + "epoch": 1.7477100078170411, + "grad_norm": 1.7772188186645508, + "learning_rate": 8.217577222236117e-07, + "loss": 3.1962, + "step": 21800 + }, + { + "epoch": 1.7477901825980637, + "grad_norm": 1.7405099868774414, + "learning_rate": 8.212423329800567e-07, + "loss": 3.1892, + "step": 21801 + }, + { + "epoch": 1.7478703573790864, + "grad_norm": 1.7167199850082397, + "learning_rate": 8.207270984871218e-07, + "loss": 3.1476, + "step": 21802 + }, + { + "epoch": 1.747950532160109, + "grad_norm": 1.7519829273223877, + "learning_rate": 8.202120187534923e-07, + "loss": 3.1881, + "step": 21803 + }, + { + "epoch": 1.7480307069411318, + "grad_norm": 1.7490061521530151, + "learning_rate": 8.196970937878546e-07, + "loss": 3.184, + "step": 21804 + }, + { + "epoch": 1.7481108817221545, + "grad_norm": 1.7449506521224976, + "learning_rate": 8.19182323598886e-07, + "loss": 3.1953, + "step": 21805 + }, + { + "epoch": 1.748191056503177, + "grad_norm": 1.708714246749878, + "learning_rate": 8.186677081952688e-07, + "loss": 3.1522, + "step": 21806 + }, + { + "epoch": 1.7482712312841997, + "grad_norm": 1.775897741317749, + "learning_rate": 8.181532475856779e-07, + "loss": 3.2416, + "step": 21807 + }, + { + "epoch": 1.7483514060652223, + "grad_norm": 1.788131594657898, + "learning_rate": 8.176389417787877e-07, + "loss": 3.1846, + "step": 21808 + }, + { + "epoch": 1.748431580846245, + "grad_norm": 1.7985326051712036, + "learning_rate": 8.171247907832692e-07, + "loss": 3.2498, + "step": 21809 + }, + { + "epoch": 1.7485117556272676, + "grad_norm": 1.794357419013977, + "learning_rate": 8.166107946077895e-07, + "loss": 3.1836, + "step": 21810 + }, + { + "epoch": 1.7485919304082902, + "grad_norm": 1.8403207063674927, + "learning_rate": 8.160969532610153e-07, + "loss": 3.2135, + "step": 21811 + }, + { + "epoch": 1.7486721051893128, + "grad_norm": 1.6896203756332397, + "learning_rate": 8.155832667516094e-07, + "loss": 3.123, + "step": 21812 + }, + { + "epoch": 1.7487522799703354, + "grad_norm": 1.784601092338562, + "learning_rate": 8.15069735088233e-07, + "loss": 3.2061, + "step": 21813 + }, + { + "epoch": 1.748832454751358, + "grad_norm": 1.738759160041809, + "learning_rate": 8.145563582795457e-07, + "loss": 3.0636, + "step": 21814 + }, + { + "epoch": 1.7489126295323807, + "grad_norm": 1.8816672563552856, + "learning_rate": 8.140431363341994e-07, + "loss": 3.2206, + "step": 21815 + }, + { + "epoch": 1.7489928043134033, + "grad_norm": 1.7476361989974976, + "learning_rate": 8.135300692608483e-07, + "loss": 3.1563, + "step": 21816 + }, + { + "epoch": 1.749072979094426, + "grad_norm": 1.7455122470855713, + "learning_rate": 8.130171570681422e-07, + "loss": 3.1931, + "step": 21817 + }, + { + "epoch": 1.7491531538754486, + "grad_norm": 1.752112865447998, + "learning_rate": 8.125043997647286e-07, + "loss": 3.2399, + "step": 21818 + }, + { + "epoch": 1.7492333286564712, + "grad_norm": 1.6664509773254395, + "learning_rate": 8.119917973592551e-07, + "loss": 3.0995, + "step": 21819 + }, + { + "epoch": 1.7493135034374938, + "grad_norm": 1.8036091327667236, + "learning_rate": 8.114793498603602e-07, + "loss": 3.2382, + "step": 21820 + }, + { + "epoch": 1.7493936782185164, + "grad_norm": 1.7552099227905273, + "learning_rate": 8.109670572766848e-07, + "loss": 3.2009, + "step": 21821 + }, + { + "epoch": 1.749473852999539, + "grad_norm": 1.7294033765792847, + "learning_rate": 8.104549196168665e-07, + "loss": 3.156, + "step": 21822 + }, + { + "epoch": 1.7495540277805617, + "grad_norm": 1.7195320129394531, + "learning_rate": 8.099429368895395e-07, + "loss": 3.0849, + "step": 21823 + }, + { + "epoch": 1.7496342025615843, + "grad_norm": 1.8347184658050537, + "learning_rate": 8.094311091033369e-07, + "loss": 3.0876, + "step": 21824 + }, + { + "epoch": 1.749714377342607, + "grad_norm": 1.7560334205627441, + "learning_rate": 8.089194362668862e-07, + "loss": 3.1497, + "step": 21825 + }, + { + "epoch": 1.7497945521236296, + "grad_norm": 1.7593411207199097, + "learning_rate": 8.084079183888138e-07, + "loss": 3.1294, + "step": 21826 + }, + { + "epoch": 1.7498747269046522, + "grad_norm": 1.9360177516937256, + "learning_rate": 8.078965554777441e-07, + "loss": 3.1721, + "step": 21827 + }, + { + "epoch": 1.7499549016856748, + "grad_norm": 1.807477593421936, + "learning_rate": 8.073853475422988e-07, + "loss": 3.0998, + "step": 21828 + }, + { + "epoch": 1.7500350764666974, + "grad_norm": 1.7301899194717407, + "learning_rate": 8.06874294591099e-07, + "loss": 3.1967, + "step": 21829 + }, + { + "epoch": 1.75011525124772, + "grad_norm": 1.8245563507080078, + "learning_rate": 8.063633966327578e-07, + "loss": 3.2562, + "step": 21830 + }, + { + "epoch": 1.7501954260287427, + "grad_norm": 1.7795928716659546, + "learning_rate": 8.058526536758882e-07, + "loss": 3.2326, + "step": 21831 + }, + { + "epoch": 1.7502756008097653, + "grad_norm": 1.7370061874389648, + "learning_rate": 8.053420657291022e-07, + "loss": 3.1184, + "step": 21832 + }, + { + "epoch": 1.750355775590788, + "grad_norm": 1.8396795988082886, + "learning_rate": 8.048316328010097e-07, + "loss": 3.1654, + "step": 21833 + }, + { + "epoch": 1.7504359503718105, + "grad_norm": 1.7831542491912842, + "learning_rate": 8.043213549002161e-07, + "loss": 3.2515, + "step": 21834 + }, + { + "epoch": 1.7505161251528332, + "grad_norm": 1.8189979791641235, + "learning_rate": 8.038112320353219e-07, + "loss": 3.2352, + "step": 21835 + }, + { + "epoch": 1.7505962999338558, + "grad_norm": 1.834807276725769, + "learning_rate": 8.033012642149296e-07, + "loss": 3.1165, + "step": 21836 + }, + { + "epoch": 1.7506764747148784, + "grad_norm": 0.660620927810669, + "learning_rate": 8.027914514476376e-07, + "loss": 2.4724, + "step": 21837 + }, + { + "epoch": 1.750756649495901, + "grad_norm": 1.7346798181533813, + "learning_rate": 8.02281793742039e-07, + "loss": 3.1527, + "step": 21838 + }, + { + "epoch": 1.7508368242769237, + "grad_norm": 1.7828468084335327, + "learning_rate": 8.017722911067294e-07, + "loss": 3.2009, + "step": 21839 + }, + { + "epoch": 1.7509169990579463, + "grad_norm": 1.807742953300476, + "learning_rate": 8.012629435502961e-07, + "loss": 3.1284, + "step": 21840 + }, + { + "epoch": 1.750997173838969, + "grad_norm": 1.6605740785598755, + "learning_rate": 8.007537510813279e-07, + "loss": 3.0942, + "step": 21841 + }, + { + "epoch": 1.7510773486199915, + "grad_norm": 1.7731726169586182, + "learning_rate": 8.002447137084091e-07, + "loss": 3.0695, + "step": 21842 + }, + { + "epoch": 1.7511575234010142, + "grad_norm": 1.8844901323318481, + "learning_rate": 7.997358314401216e-07, + "loss": 3.1024, + "step": 21843 + }, + { + "epoch": 1.7512376981820368, + "grad_norm": 1.847001075744629, + "learning_rate": 7.992271042850474e-07, + "loss": 3.2362, + "step": 21844 + }, + { + "epoch": 1.7513178729630594, + "grad_norm": 1.962169885635376, + "learning_rate": 7.987185322517599e-07, + "loss": 3.2428, + "step": 21845 + }, + { + "epoch": 1.751398047744082, + "grad_norm": 1.7836058139801025, + "learning_rate": 7.982101153488353e-07, + "loss": 3.1387, + "step": 21846 + }, + { + "epoch": 1.7514782225251047, + "grad_norm": 1.6649656295776367, + "learning_rate": 7.977018535848446e-07, + "loss": 3.0612, + "step": 21847 + }, + { + "epoch": 1.7515583973061273, + "grad_norm": 1.7595174312591553, + "learning_rate": 7.971937469683578e-07, + "loss": 3.1921, + "step": 21848 + }, + { + "epoch": 1.75163857208715, + "grad_norm": 1.7515074014663696, + "learning_rate": 7.966857955079432e-07, + "loss": 3.2343, + "step": 21849 + }, + { + "epoch": 1.7517187468681725, + "grad_norm": 0.6633621454238892, + "learning_rate": 7.96177999212161e-07, + "loss": 2.4636, + "step": 21850 + }, + { + "epoch": 1.7517989216491952, + "grad_norm": 1.8854316473007202, + "learning_rate": 7.95670358089573e-07, + "loss": 3.1807, + "step": 21851 + }, + { + "epoch": 1.7518790964302178, + "grad_norm": 1.7128572463989258, + "learning_rate": 7.951628721487392e-07, + "loss": 3.1591, + "step": 21852 + }, + { + "epoch": 1.7519592712112404, + "grad_norm": 1.8944720029830933, + "learning_rate": 7.946555413982149e-07, + "loss": 3.1665, + "step": 21853 + }, + { + "epoch": 1.752039445992263, + "grad_norm": 1.7732208967208862, + "learning_rate": 7.941483658465565e-07, + "loss": 3.223, + "step": 21854 + }, + { + "epoch": 1.7521196207732856, + "grad_norm": 1.65639066696167, + "learning_rate": 7.936413455023096e-07, + "loss": 3.1614, + "step": 21855 + }, + { + "epoch": 1.7521997955543083, + "grad_norm": 1.8228274583816528, + "learning_rate": 7.931344803740259e-07, + "loss": 3.1498, + "step": 21856 + }, + { + "epoch": 1.752279970335331, + "grad_norm": 0.6586769223213196, + "learning_rate": 7.92627770470249e-07, + "loss": 2.4958, + "step": 21857 + }, + { + "epoch": 1.7523601451163535, + "grad_norm": 1.7278460264205933, + "learning_rate": 7.921212157995239e-07, + "loss": 3.2104, + "step": 21858 + }, + { + "epoch": 1.7524403198973761, + "grad_norm": 1.7352094650268555, + "learning_rate": 7.916148163703907e-07, + "loss": 3.1823, + "step": 21859 + }, + { + "epoch": 1.7525204946783988, + "grad_norm": 1.7516870498657227, + "learning_rate": 7.911085721913858e-07, + "loss": 3.0884, + "step": 21860 + }, + { + "epoch": 1.7526006694594214, + "grad_norm": 1.7590601444244385, + "learning_rate": 7.906024832710435e-07, + "loss": 3.2563, + "step": 21861 + }, + { + "epoch": 1.752680844240444, + "grad_norm": 1.829797625541687, + "learning_rate": 7.900965496178981e-07, + "loss": 3.2151, + "step": 21862 + }, + { + "epoch": 1.7527610190214666, + "grad_norm": 0.6671677231788635, + "learning_rate": 7.895907712404793e-07, + "loss": 2.4623, + "step": 21863 + }, + { + "epoch": 1.7528411938024893, + "grad_norm": 1.8760818243026733, + "learning_rate": 7.890851481473161e-07, + "loss": 3.1752, + "step": 21864 + }, + { + "epoch": 1.752921368583512, + "grad_norm": 1.7595889568328857, + "learning_rate": 7.885796803469292e-07, + "loss": 3.1459, + "step": 21865 + }, + { + "epoch": 1.7530015433645347, + "grad_norm": 2.0814220905303955, + "learning_rate": 7.88074367847842e-07, + "loss": 3.2089, + "step": 21866 + }, + { + "epoch": 1.7530817181455574, + "grad_norm": 1.7764267921447754, + "learning_rate": 7.875692106585741e-07, + "loss": 3.1174, + "step": 21867 + }, + { + "epoch": 1.75316189292658, + "grad_norm": 1.7598737478256226, + "learning_rate": 7.870642087876435e-07, + "loss": 3.1, + "step": 21868 + }, + { + "epoch": 1.7532420677076026, + "grad_norm": 1.8368216753005981, + "learning_rate": 7.865593622435641e-07, + "loss": 3.234, + "step": 21869 + }, + { + "epoch": 1.7533222424886252, + "grad_norm": 1.7512288093566895, + "learning_rate": 7.860546710348449e-07, + "loss": 3.1493, + "step": 21870 + }, + { + "epoch": 1.7534024172696479, + "grad_norm": 1.720906138420105, + "learning_rate": 7.855501351699979e-07, + "loss": 3.1496, + "step": 21871 + }, + { + "epoch": 1.7534825920506705, + "grad_norm": 1.7830337285995483, + "learning_rate": 7.850457546575274e-07, + "loss": 3.2156, + "step": 21872 + }, + { + "epoch": 1.753562766831693, + "grad_norm": 1.7966060638427734, + "learning_rate": 7.845415295059378e-07, + "loss": 3.1533, + "step": 21873 + }, + { + "epoch": 1.7536429416127157, + "grad_norm": 1.7997301816940308, + "learning_rate": 7.840374597237321e-07, + "loss": 3.1849, + "step": 21874 + }, + { + "epoch": 1.7537231163937383, + "grad_norm": 1.7987810373306274, + "learning_rate": 7.835335453194048e-07, + "loss": 3.1385, + "step": 21875 + }, + { + "epoch": 1.753803291174761, + "grad_norm": 1.7833271026611328, + "learning_rate": 7.830297863014535e-07, + "loss": 3.2736, + "step": 21876 + }, + { + "epoch": 1.7538834659557836, + "grad_norm": 1.8036277294158936, + "learning_rate": 7.825261826783715e-07, + "loss": 3.1871, + "step": 21877 + }, + { + "epoch": 1.7539636407368062, + "grad_norm": 1.7931365966796875, + "learning_rate": 7.820227344586506e-07, + "loss": 3.1428, + "step": 21878 + }, + { + "epoch": 1.7540438155178288, + "grad_norm": 1.8000272512435913, + "learning_rate": 7.815194416507788e-07, + "loss": 3.1548, + "step": 21879 + }, + { + "epoch": 1.7541239902988515, + "grad_norm": 1.731300711631775, + "learning_rate": 7.810163042632379e-07, + "loss": 3.115, + "step": 21880 + }, + { + "epoch": 1.754204165079874, + "grad_norm": 1.6931567192077637, + "learning_rate": 7.805133223045136e-07, + "loss": 3.2115, + "step": 21881 + }, + { + "epoch": 1.7542843398608967, + "grad_norm": 1.8586206436157227, + "learning_rate": 7.800104957830857e-07, + "loss": 3.2932, + "step": 21882 + }, + { + "epoch": 1.7543645146419193, + "grad_norm": 1.7961866855621338, + "learning_rate": 7.795078247074317e-07, + "loss": 3.2878, + "step": 21883 + }, + { + "epoch": 1.754444689422942, + "grad_norm": 1.7669051885604858, + "learning_rate": 7.790053090860273e-07, + "loss": 3.2797, + "step": 21884 + }, + { + "epoch": 1.7545248642039648, + "grad_norm": 1.6882485151290894, + "learning_rate": 7.785029489273421e-07, + "loss": 3.1175, + "step": 21885 + }, + { + "epoch": 1.7546050389849874, + "grad_norm": 0.6585519313812256, + "learning_rate": 7.780007442398463e-07, + "loss": 2.469, + "step": 21886 + }, + { + "epoch": 1.75468521376601, + "grad_norm": 1.662749171257019, + "learning_rate": 7.774986950320085e-07, + "loss": 3.1632, + "step": 21887 + }, + { + "epoch": 1.7547653885470327, + "grad_norm": 1.768938660621643, + "learning_rate": 7.769968013122931e-07, + "loss": 3.1833, + "step": 21888 + }, + { + "epoch": 1.7548455633280553, + "grad_norm": 1.6917767524719238, + "learning_rate": 7.764950630891622e-07, + "loss": 3.1904, + "step": 21889 + }, + { + "epoch": 1.754925738109078, + "grad_norm": 0.6600466370582581, + "learning_rate": 7.759934803710722e-07, + "loss": 2.4729, + "step": 21890 + }, + { + "epoch": 1.7550059128901006, + "grad_norm": 1.780898094177246, + "learning_rate": 7.754920531664811e-07, + "loss": 3.1066, + "step": 21891 + }, + { + "epoch": 1.7550860876711232, + "grad_norm": 1.9066411256790161, + "learning_rate": 7.749907814838431e-07, + "loss": 3.3766, + "step": 21892 + }, + { + "epoch": 1.7551662624521458, + "grad_norm": 1.704911231994629, + "learning_rate": 7.744896653316092e-07, + "loss": 3.1241, + "step": 21893 + }, + { + "epoch": 1.7552464372331684, + "grad_norm": 1.8539600372314453, + "learning_rate": 7.739887047182293e-07, + "loss": 3.1658, + "step": 21894 + }, + { + "epoch": 1.755326612014191, + "grad_norm": 1.9861235618591309, + "learning_rate": 7.734878996521477e-07, + "loss": 3.2452, + "step": 21895 + }, + { + "epoch": 1.7554067867952137, + "grad_norm": 0.6711223125457764, + "learning_rate": 7.729872501418079e-07, + "loss": 2.4622, + "step": 21896 + }, + { + "epoch": 1.7554869615762363, + "grad_norm": 2.0580027103424072, + "learning_rate": 7.724867561956506e-07, + "loss": 3.243, + "step": 21897 + }, + { + "epoch": 1.755567136357259, + "grad_norm": 1.7061344385147095, + "learning_rate": 7.719864178221148e-07, + "loss": 3.1395, + "step": 21898 + }, + { + "epoch": 1.7556473111382815, + "grad_norm": 1.844253420829773, + "learning_rate": 7.714862350296382e-07, + "loss": 3.1736, + "step": 21899 + }, + { + "epoch": 1.7557274859193042, + "grad_norm": 1.8112363815307617, + "learning_rate": 7.709862078266483e-07, + "loss": 3.1307, + "step": 21900 + }, + { + "epoch": 1.7558076607003268, + "grad_norm": 1.708377718925476, + "learning_rate": 7.704863362215798e-07, + "loss": 3.1361, + "step": 21901 + }, + { + "epoch": 1.7558878354813494, + "grad_norm": 1.8437504768371582, + "learning_rate": 7.699866202228579e-07, + "loss": 3.2089, + "step": 21902 + }, + { + "epoch": 1.755968010262372, + "grad_norm": 1.824531078338623, + "learning_rate": 7.694870598389082e-07, + "loss": 3.1591, + "step": 21903 + }, + { + "epoch": 1.7560481850433947, + "grad_norm": 1.7046703100204468, + "learning_rate": 7.689876550781561e-07, + "loss": 3.1614, + "step": 21904 + }, + { + "epoch": 1.7561283598244173, + "grad_norm": 1.7655925750732422, + "learning_rate": 7.684884059490172e-07, + "loss": 3.1931, + "step": 21905 + }, + { + "epoch": 1.75620853460544, + "grad_norm": 1.8018451929092407, + "learning_rate": 7.679893124599091e-07, + "loss": 3.1199, + "step": 21906 + }, + { + "epoch": 1.7562887093864625, + "grad_norm": 1.660895586013794, + "learning_rate": 7.674903746192486e-07, + "loss": 3.1021, + "step": 21907 + }, + { + "epoch": 1.7563688841674852, + "grad_norm": 1.7865512371063232, + "learning_rate": 7.669915924354465e-07, + "loss": 3.2532, + "step": 21908 + }, + { + "epoch": 1.7564490589485078, + "grad_norm": 0.6611632108688354, + "learning_rate": 7.66492965916914e-07, + "loss": 2.4705, + "step": 21909 + }, + { + "epoch": 1.7565292337295304, + "grad_norm": 1.629569172859192, + "learning_rate": 7.659944950720533e-07, + "loss": 3.0566, + "step": 21910 + }, + { + "epoch": 1.756609408510553, + "grad_norm": 1.816208004951477, + "learning_rate": 7.654961799092719e-07, + "loss": 3.2762, + "step": 21911 + }, + { + "epoch": 1.7566895832915757, + "grad_norm": 1.8377938270568848, + "learning_rate": 7.6499802043697e-07, + "loss": 3.2128, + "step": 21912 + }, + { + "epoch": 1.7567697580725983, + "grad_norm": 1.8117600679397583, + "learning_rate": 7.645000166635463e-07, + "loss": 3.1179, + "step": 21913 + }, + { + "epoch": 1.756849932853621, + "grad_norm": 1.7566395998001099, + "learning_rate": 7.640021685973998e-07, + "loss": 3.1661, + "step": 21914 + }, + { + "epoch": 1.7569301076346435, + "grad_norm": 1.6829291582107544, + "learning_rate": 7.63504476246919e-07, + "loss": 3.096, + "step": 21915 + }, + { + "epoch": 1.7570102824156661, + "grad_norm": 1.7850788831710815, + "learning_rate": 7.630069396204987e-07, + "loss": 3.1858, + "step": 21916 + }, + { + "epoch": 1.7570904571966888, + "grad_norm": 1.7921055555343628, + "learning_rate": 7.625095587265252e-07, + "loss": 3.157, + "step": 21917 + }, + { + "epoch": 1.7571706319777114, + "grad_norm": 1.8198853731155396, + "learning_rate": 7.620123335733853e-07, + "loss": 3.2228, + "step": 21918 + }, + { + "epoch": 1.757250806758734, + "grad_norm": 0.6865971088409424, + "learning_rate": 7.615152641694623e-07, + "loss": 2.4647, + "step": 21919 + }, + { + "epoch": 1.7573309815397566, + "grad_norm": 1.7995766401290894, + "learning_rate": 7.610183505231361e-07, + "loss": 3.0613, + "step": 21920 + }, + { + "epoch": 1.7574111563207793, + "grad_norm": 1.8405711650848389, + "learning_rate": 7.605215926427834e-07, + "loss": 3.1143, + "step": 21921 + }, + { + "epoch": 1.757491331101802, + "grad_norm": 1.8605799674987793, + "learning_rate": 7.600249905367795e-07, + "loss": 3.1974, + "step": 21922 + }, + { + "epoch": 1.7575715058828245, + "grad_norm": 1.796703577041626, + "learning_rate": 7.595285442134992e-07, + "loss": 3.1855, + "step": 21923 + }, + { + "epoch": 1.7576516806638471, + "grad_norm": 1.6875169277191162, + "learning_rate": 7.590322536813122e-07, + "loss": 3.1122, + "step": 21924 + }, + { + "epoch": 1.7577318554448698, + "grad_norm": 1.9342257976531982, + "learning_rate": 7.58536118948583e-07, + "loss": 3.1349, + "step": 21925 + }, + { + "epoch": 1.7578120302258924, + "grad_norm": 1.7470109462738037, + "learning_rate": 7.58040140023677e-07, + "loss": 3.1763, + "step": 21926 + }, + { + "epoch": 1.757892205006915, + "grad_norm": 1.8396023511886597, + "learning_rate": 7.575443169149577e-07, + "loss": 3.219, + "step": 21927 + }, + { + "epoch": 1.7579723797879376, + "grad_norm": 1.7323089838027954, + "learning_rate": 7.570486496307838e-07, + "loss": 3.17, + "step": 21928 + }, + { + "epoch": 1.7580525545689603, + "grad_norm": 1.8059954643249512, + "learning_rate": 7.565531381795144e-07, + "loss": 3.1382, + "step": 21929 + }, + { + "epoch": 1.7581327293499829, + "grad_norm": 1.6970715522766113, + "learning_rate": 7.560577825694992e-07, + "loss": 3.0649, + "step": 21930 + }, + { + "epoch": 1.7582129041310055, + "grad_norm": 1.7025580406188965, + "learning_rate": 7.555625828090918e-07, + "loss": 3.1696, + "step": 21931 + }, + { + "epoch": 1.7582930789120281, + "grad_norm": 1.7344913482666016, + "learning_rate": 7.550675389066408e-07, + "loss": 3.1524, + "step": 21932 + }, + { + "epoch": 1.7583732536930508, + "grad_norm": 1.6911550760269165, + "learning_rate": 7.545726508704942e-07, + "loss": 3.1387, + "step": 21933 + }, + { + "epoch": 1.7584534284740734, + "grad_norm": 1.8382532596588135, + "learning_rate": 7.540779187089952e-07, + "loss": 3.2972, + "step": 21934 + }, + { + "epoch": 1.758533603255096, + "grad_norm": 1.907403826713562, + "learning_rate": 7.535833424304817e-07, + "loss": 3.2777, + "step": 21935 + }, + { + "epoch": 1.7586137780361186, + "grad_norm": 1.7886346578598022, + "learning_rate": 7.530889220432947e-07, + "loss": 3.2432, + "step": 21936 + }, + { + "epoch": 1.7586939528171412, + "grad_norm": 1.721859335899353, + "learning_rate": 7.525946575557697e-07, + "loss": 3.168, + "step": 21937 + }, + { + "epoch": 1.7587741275981639, + "grad_norm": 1.7556298971176147, + "learning_rate": 7.5210054897624e-07, + "loss": 3.1792, + "step": 21938 + }, + { + "epoch": 1.7588543023791865, + "grad_norm": 1.83879816532135, + "learning_rate": 7.51606596313037e-07, + "loss": 3.2163, + "step": 21939 + }, + { + "epoch": 1.7589344771602091, + "grad_norm": 1.694547176361084, + "learning_rate": 7.511127995744849e-07, + "loss": 3.1257, + "step": 21940 + }, + { + "epoch": 1.7590146519412317, + "grad_norm": 1.8578810691833496, + "learning_rate": 7.506191587689127e-07, + "loss": 3.2748, + "step": 21941 + }, + { + "epoch": 1.7590948267222544, + "grad_norm": 1.993481993675232, + "learning_rate": 7.501256739046414e-07, + "loss": 3.0546, + "step": 21942 + }, + { + "epoch": 1.759175001503277, + "grad_norm": 1.8093490600585938, + "learning_rate": 7.496323449899911e-07, + "loss": 3.1895, + "step": 21943 + }, + { + "epoch": 1.7592551762842996, + "grad_norm": 1.6928035020828247, + "learning_rate": 7.491391720332808e-07, + "loss": 3.1877, + "step": 21944 + }, + { + "epoch": 1.7593353510653222, + "grad_norm": 1.833083987236023, + "learning_rate": 7.486461550428215e-07, + "loss": 3.1107, + "step": 21945 + }, + { + "epoch": 1.759415525846345, + "grad_norm": 1.7352004051208496, + "learning_rate": 7.481532940269287e-07, + "loss": 3.1255, + "step": 21946 + }, + { + "epoch": 1.7594957006273677, + "grad_norm": 1.7561293840408325, + "learning_rate": 7.476605889939104e-07, + "loss": 3.2049, + "step": 21947 + }, + { + "epoch": 1.7595758754083903, + "grad_norm": 1.8627848625183105, + "learning_rate": 7.471680399520731e-07, + "loss": 3.224, + "step": 21948 + }, + { + "epoch": 1.759656050189413, + "grad_norm": 1.7388296127319336, + "learning_rate": 7.466756469097236e-07, + "loss": 3.2038, + "step": 21949 + }, + { + "epoch": 1.7597362249704356, + "grad_norm": 1.7547410726547241, + "learning_rate": 7.461834098751608e-07, + "loss": 3.2228, + "step": 21950 + }, + { + "epoch": 1.7598163997514582, + "grad_norm": 1.7531757354736328, + "learning_rate": 7.456913288566825e-07, + "loss": 3.1915, + "step": 21951 + }, + { + "epoch": 1.7598965745324808, + "grad_norm": 1.8591768741607666, + "learning_rate": 7.451994038625887e-07, + "loss": 3.134, + "step": 21952 + }, + { + "epoch": 1.7599767493135035, + "grad_norm": 1.6999688148498535, + "learning_rate": 7.447076349011706e-07, + "loss": 3.185, + "step": 21953 + }, + { + "epoch": 1.760056924094526, + "grad_norm": 1.7434049844741821, + "learning_rate": 7.442160219807204e-07, + "loss": 3.163, + "step": 21954 + }, + { + "epoch": 1.7601370988755487, + "grad_norm": 1.8570853471755981, + "learning_rate": 7.43724565109526e-07, + "loss": 3.2689, + "step": 21955 + }, + { + "epoch": 1.7602172736565713, + "grad_norm": 1.7159608602523804, + "learning_rate": 7.432332642958728e-07, + "loss": 3.1809, + "step": 21956 + }, + { + "epoch": 1.760297448437594, + "grad_norm": 1.8275245428085327, + "learning_rate": 7.427421195480433e-07, + "loss": 3.2125, + "step": 21957 + }, + { + "epoch": 1.7603776232186166, + "grad_norm": 1.7572863101959229, + "learning_rate": 7.422511308743196e-07, + "loss": 3.2267, + "step": 21958 + }, + { + "epoch": 1.7604577979996392, + "grad_norm": 1.7741119861602783, + "learning_rate": 7.417602982829819e-07, + "loss": 3.1731, + "step": 21959 + }, + { + "epoch": 1.7605379727806618, + "grad_norm": 1.876655101776123, + "learning_rate": 7.412696217823001e-07, + "loss": 3.2419, + "step": 21960 + }, + { + "epoch": 1.7606181475616844, + "grad_norm": 1.8410195112228394, + "learning_rate": 7.407791013805488e-07, + "loss": 3.2417, + "step": 21961 + }, + { + "epoch": 1.760698322342707, + "grad_norm": 1.8274500370025635, + "learning_rate": 7.40288737085999e-07, + "loss": 3.1686, + "step": 21962 + }, + { + "epoch": 1.7607784971237297, + "grad_norm": 1.7733864784240723, + "learning_rate": 7.397985289069188e-07, + "loss": 3.1238, + "step": 21963 + }, + { + "epoch": 1.7608586719047523, + "grad_norm": 1.6869162321090698, + "learning_rate": 7.393084768515724e-07, + "loss": 3.0983, + "step": 21964 + }, + { + "epoch": 1.760938846685775, + "grad_norm": 1.80897057056427, + "learning_rate": 7.388185809282189e-07, + "loss": 3.2017, + "step": 21965 + }, + { + "epoch": 1.7610190214667978, + "grad_norm": 1.7435239553451538, + "learning_rate": 7.383288411451217e-07, + "loss": 3.1677, + "step": 21966 + }, + { + "epoch": 1.7610991962478204, + "grad_norm": 1.6292035579681396, + "learning_rate": 7.378392575105353e-07, + "loss": 3.1008, + "step": 21967 + }, + { + "epoch": 1.761179371028843, + "grad_norm": 1.7410892248153687, + "learning_rate": 7.373498300327143e-07, + "loss": 3.1189, + "step": 21968 + }, + { + "epoch": 1.7612595458098657, + "grad_norm": 1.9377840757369995, + "learning_rate": 7.36860558719913e-07, + "loss": 3.2433, + "step": 21969 + }, + { + "epoch": 1.7613397205908883, + "grad_norm": 1.7161140441894531, + "learning_rate": 7.36371443580377e-07, + "loss": 3.1497, + "step": 21970 + }, + { + "epoch": 1.761419895371911, + "grad_norm": 1.6766581535339355, + "learning_rate": 7.358824846223523e-07, + "loss": 3.0565, + "step": 21971 + }, + { + "epoch": 1.7615000701529335, + "grad_norm": 1.7421138286590576, + "learning_rate": 7.353936818540852e-07, + "loss": 3.1087, + "step": 21972 + }, + { + "epoch": 1.7615802449339562, + "grad_norm": 1.7765791416168213, + "learning_rate": 7.349050352838138e-07, + "loss": 3.1571, + "step": 21973 + }, + { + "epoch": 1.7616604197149788, + "grad_norm": 1.8210082054138184, + "learning_rate": 7.344165449197815e-07, + "loss": 3.0859, + "step": 21974 + }, + { + "epoch": 1.7617405944960014, + "grad_norm": 1.915543556213379, + "learning_rate": 7.339282107702184e-07, + "loss": 3.2932, + "step": 21975 + }, + { + "epoch": 1.761820769277024, + "grad_norm": 1.8261879682540894, + "learning_rate": 7.3344003284336e-07, + "loss": 3.1555, + "step": 21976 + }, + { + "epoch": 1.7619009440580466, + "grad_norm": 1.9168717861175537, + "learning_rate": 7.329520111474365e-07, + "loss": 3.1559, + "step": 21977 + }, + { + "epoch": 1.7619811188390693, + "grad_norm": 1.7600088119506836, + "learning_rate": 7.324641456906767e-07, + "loss": 3.0919, + "step": 21978 + }, + { + "epoch": 1.762061293620092, + "grad_norm": 1.879930853843689, + "learning_rate": 7.319764364813065e-07, + "loss": 3.2727, + "step": 21979 + }, + { + "epoch": 1.7621414684011145, + "grad_norm": 1.772886037826538, + "learning_rate": 7.314888835275446e-07, + "loss": 3.2045, + "step": 21980 + }, + { + "epoch": 1.7622216431821371, + "grad_norm": 1.825398325920105, + "learning_rate": 7.310014868376147e-07, + "loss": 3.2432, + "step": 21981 + }, + { + "epoch": 1.7623018179631598, + "grad_norm": 1.6986792087554932, + "learning_rate": 7.305142464197335e-07, + "loss": 3.0333, + "step": 21982 + }, + { + "epoch": 1.7623819927441824, + "grad_norm": 1.8198041915893555, + "learning_rate": 7.300271622821142e-07, + "loss": 3.196, + "step": 21983 + }, + { + "epoch": 1.762462167525205, + "grad_norm": 1.6541147232055664, + "learning_rate": 7.295402344329716e-07, + "loss": 3.1372, + "step": 21984 + }, + { + "epoch": 1.7625423423062276, + "grad_norm": 1.70880925655365, + "learning_rate": 7.290534628805123e-07, + "loss": 3.212, + "step": 21985 + }, + { + "epoch": 1.7626225170872503, + "grad_norm": 1.7789092063903809, + "learning_rate": 7.285668476329444e-07, + "loss": 3.205, + "step": 21986 + }, + { + "epoch": 1.7627026918682729, + "grad_norm": 1.8345460891723633, + "learning_rate": 7.280803886984711e-07, + "loss": 3.2542, + "step": 21987 + }, + { + "epoch": 1.7627828666492955, + "grad_norm": 1.7703944444656372, + "learning_rate": 7.275940860852948e-07, + "loss": 3.1418, + "step": 21988 + }, + { + "epoch": 1.7628630414303181, + "grad_norm": 1.6934586763381958, + "learning_rate": 7.271079398016167e-07, + "loss": 3.1636, + "step": 21989 + }, + { + "epoch": 1.7629432162113408, + "grad_norm": 1.754010796546936, + "learning_rate": 7.266219498556282e-07, + "loss": 3.1733, + "step": 21990 + }, + { + "epoch": 1.7630233909923634, + "grad_norm": 1.833591103553772, + "learning_rate": 7.261361162555259e-07, + "loss": 3.0482, + "step": 21991 + }, + { + "epoch": 1.763103565773386, + "grad_norm": 1.7868249416351318, + "learning_rate": 7.256504390094998e-07, + "loss": 3.0955, + "step": 21992 + }, + { + "epoch": 1.7631837405544086, + "grad_norm": 0.6558399200439453, + "learning_rate": 7.251649181257391e-07, + "loss": 2.4657, + "step": 21993 + }, + { + "epoch": 1.7632639153354313, + "grad_norm": 1.8118187189102173, + "learning_rate": 7.246795536124307e-07, + "loss": 3.1808, + "step": 21994 + }, + { + "epoch": 1.7633440901164539, + "grad_norm": 1.8181864023208618, + "learning_rate": 7.241943454777544e-07, + "loss": 3.2178, + "step": 21995 + }, + { + "epoch": 1.7634242648974765, + "grad_norm": 1.7749983072280884, + "learning_rate": 7.237092937298928e-07, + "loss": 3.1317, + "step": 21996 + }, + { + "epoch": 1.7635044396784991, + "grad_norm": 1.6797804832458496, + "learning_rate": 7.232243983770226e-07, + "loss": 3.0515, + "step": 21997 + }, + { + "epoch": 1.7635846144595217, + "grad_norm": 1.7418557405471802, + "learning_rate": 7.227396594273195e-07, + "loss": 3.1893, + "step": 21998 + }, + { + "epoch": 1.7636647892405444, + "grad_norm": 1.8460122346878052, + "learning_rate": 7.22255076888958e-07, + "loss": 3.1749, + "step": 21999 + }, + { + "epoch": 1.763744964021567, + "grad_norm": 1.7654415369033813, + "learning_rate": 7.217706507701039e-07, + "loss": 3.1418, + "step": 22000 + }, + { + "epoch": 1.7638251388025896, + "grad_norm": 1.8609213829040527, + "learning_rate": 7.212863810789272e-07, + "loss": 3.1665, + "step": 22001 + }, + { + "epoch": 1.7639053135836122, + "grad_norm": 1.8016513586044312, + "learning_rate": 7.208022678235926e-07, + "loss": 3.1511, + "step": 22002 + }, + { + "epoch": 1.7639854883646349, + "grad_norm": 1.8132500648498535, + "learning_rate": 7.203183110122602e-07, + "loss": 3.1344, + "step": 22003 + }, + { + "epoch": 1.7640656631456575, + "grad_norm": 1.7239757776260376, + "learning_rate": 7.198345106530935e-07, + "loss": 3.1398, + "step": 22004 + }, + { + "epoch": 1.7641458379266801, + "grad_norm": 1.884891390800476, + "learning_rate": 7.193508667542437e-07, + "loss": 3.1696, + "step": 22005 + }, + { + "epoch": 1.7642260127077027, + "grad_norm": 1.9491976499557495, + "learning_rate": 7.188673793238677e-07, + "loss": 3.2421, + "step": 22006 + }, + { + "epoch": 1.7643061874887254, + "grad_norm": 1.8192678689956665, + "learning_rate": 7.183840483701166e-07, + "loss": 3.2294, + "step": 22007 + }, + { + "epoch": 1.764386362269748, + "grad_norm": 1.7560220956802368, + "learning_rate": 7.179008739011395e-07, + "loss": 3.1544, + "step": 22008 + }, + { + "epoch": 1.7644665370507706, + "grad_norm": 1.7402255535125732, + "learning_rate": 7.174178559250844e-07, + "loss": 3.0804, + "step": 22009 + }, + { + "epoch": 1.7645467118317932, + "grad_norm": 1.7573038339614868, + "learning_rate": 7.169349944500903e-07, + "loss": 3.2518, + "step": 22010 + }, + { + "epoch": 1.7646268866128159, + "grad_norm": 1.7804088592529297, + "learning_rate": 7.164522894843007e-07, + "loss": 3.0724, + "step": 22011 + }, + { + "epoch": 1.7647070613938385, + "grad_norm": 1.764187216758728, + "learning_rate": 7.159697410358546e-07, + "loss": 3.1768, + "step": 22012 + }, + { + "epoch": 1.764787236174861, + "grad_norm": 0.6545199155807495, + "learning_rate": 7.154873491128855e-07, + "loss": 2.4781, + "step": 22013 + }, + { + "epoch": 1.7648674109558837, + "grad_norm": 1.7370998859405518, + "learning_rate": 7.150051137235292e-07, + "loss": 3.1862, + "step": 22014 + }, + { + "epoch": 1.7649475857369064, + "grad_norm": 1.8280402421951294, + "learning_rate": 7.145230348759136e-07, + "loss": 3.199, + "step": 22015 + }, + { + "epoch": 1.765027760517929, + "grad_norm": 1.7800308465957642, + "learning_rate": 7.140411125781666e-07, + "loss": 3.1159, + "step": 22016 + }, + { + "epoch": 1.7651079352989516, + "grad_norm": 0.6711441874504089, + "learning_rate": 7.135593468384139e-07, + "loss": 2.4626, + "step": 22017 + }, + { + "epoch": 1.7651881100799742, + "grad_norm": 1.6834443807601929, + "learning_rate": 7.130777376647768e-07, + "loss": 3.1504, + "step": 22018 + }, + { + "epoch": 1.7652682848609969, + "grad_norm": 1.837615966796875, + "learning_rate": 7.125962850653789e-07, + "loss": 3.172, + "step": 22019 + }, + { + "epoch": 1.7653484596420195, + "grad_norm": 1.8096580505371094, + "learning_rate": 7.121149890483325e-07, + "loss": 3.0969, + "step": 22020 + }, + { + "epoch": 1.765428634423042, + "grad_norm": 1.720503807067871, + "learning_rate": 7.116338496217534e-07, + "loss": 3.1053, + "step": 22021 + }, + { + "epoch": 1.7655088092040647, + "grad_norm": 1.7582507133483887, + "learning_rate": 7.111528667937551e-07, + "loss": 3.2186, + "step": 22022 + }, + { + "epoch": 1.7655889839850873, + "grad_norm": 1.6930779218673706, + "learning_rate": 7.106720405724443e-07, + "loss": 3.0966, + "step": 22023 + }, + { + "epoch": 1.76566915876611, + "grad_norm": 1.7751445770263672, + "learning_rate": 7.101913709659325e-07, + "loss": 3.2424, + "step": 22024 + }, + { + "epoch": 1.7657493335471326, + "grad_norm": 1.7318538427352905, + "learning_rate": 7.097108579823175e-07, + "loss": 3.1437, + "step": 22025 + }, + { + "epoch": 1.7658295083281552, + "grad_norm": 1.6984074115753174, + "learning_rate": 7.09230501629703e-07, + "loss": 3.2103, + "step": 22026 + }, + { + "epoch": 1.765909683109178, + "grad_norm": 1.8912478685379028, + "learning_rate": 7.087503019161879e-07, + "loss": 3.2515, + "step": 22027 + }, + { + "epoch": 1.7659898578902007, + "grad_norm": 1.7899199724197388, + "learning_rate": 7.08270258849868e-07, + "loss": 3.1805, + "step": 22028 + }, + { + "epoch": 1.7660700326712233, + "grad_norm": 1.7936738729476929, + "learning_rate": 7.07790372438838e-07, + "loss": 3.2526, + "step": 22029 + }, + { + "epoch": 1.766150207452246, + "grad_norm": 1.7055379152297974, + "learning_rate": 7.073106426911879e-07, + "loss": 3.1884, + "step": 22030 + }, + { + "epoch": 1.7662303822332686, + "grad_norm": 1.7962348461151123, + "learning_rate": 7.068310696150038e-07, + "loss": 3.1469, + "step": 22031 + }, + { + "epoch": 1.7663105570142912, + "grad_norm": 1.7758336067199707, + "learning_rate": 7.063516532183723e-07, + "loss": 3.1311, + "step": 22032 + }, + { + "epoch": 1.7663907317953138, + "grad_norm": 1.7770873308181763, + "learning_rate": 7.058723935093781e-07, + "loss": 3.1495, + "step": 22033 + }, + { + "epoch": 1.7664709065763364, + "grad_norm": 1.7781785726547241, + "learning_rate": 7.053932904960981e-07, + "loss": 3.1174, + "step": 22034 + }, + { + "epoch": 1.766551081357359, + "grad_norm": 1.7819881439208984, + "learning_rate": 7.049143441866135e-07, + "loss": 3.1581, + "step": 22035 + }, + { + "epoch": 1.7666312561383817, + "grad_norm": 1.842015266418457, + "learning_rate": 7.044355545889958e-07, + "loss": 3.135, + "step": 22036 + }, + { + "epoch": 1.7667114309194043, + "grad_norm": 0.6410608291625977, + "learning_rate": 7.039569217113184e-07, + "loss": 2.4674, + "step": 22037 + }, + { + "epoch": 1.766791605700427, + "grad_norm": 1.9135236740112305, + "learning_rate": 7.034784455616516e-07, + "loss": 3.2266, + "step": 22038 + }, + { + "epoch": 1.7668717804814495, + "grad_norm": 1.7607556581497192, + "learning_rate": 7.030001261480613e-07, + "loss": 3.2072, + "step": 22039 + }, + { + "epoch": 1.7669519552624722, + "grad_norm": 1.795141339302063, + "learning_rate": 7.025219634786141e-07, + "loss": 3.2306, + "step": 22040 + }, + { + "epoch": 1.7670321300434948, + "grad_norm": 1.7215921878814697, + "learning_rate": 7.02043957561368e-07, + "loss": 3.2073, + "step": 22041 + }, + { + "epoch": 1.7671123048245174, + "grad_norm": 1.8160451650619507, + "learning_rate": 7.015661084043845e-07, + "loss": 3.1309, + "step": 22042 + }, + { + "epoch": 1.76719247960554, + "grad_norm": 0.6674776077270508, + "learning_rate": 7.010884160157183e-07, + "loss": 2.463, + "step": 22043 + }, + { + "epoch": 1.7672726543865627, + "grad_norm": 1.8805381059646606, + "learning_rate": 7.00610880403425e-07, + "loss": 3.2011, + "step": 22044 + }, + { + "epoch": 1.7673528291675853, + "grad_norm": 1.7487907409667969, + "learning_rate": 7.001335015755562e-07, + "loss": 3.1598, + "step": 22045 + }, + { + "epoch": 1.767433003948608, + "grad_norm": 1.7290761470794678, + "learning_rate": 6.996562795401573e-07, + "loss": 3.0612, + "step": 22046 + }, + { + "epoch": 1.7675131787296305, + "grad_norm": 1.7898027896881104, + "learning_rate": 6.991792143052755e-07, + "loss": 3.2427, + "step": 22047 + }, + { + "epoch": 1.7675933535106534, + "grad_norm": 1.780539870262146, + "learning_rate": 6.987023058789544e-07, + "loss": 3.1902, + "step": 22048 + }, + { + "epoch": 1.767673528291676, + "grad_norm": 1.7458137273788452, + "learning_rate": 6.98225554269234e-07, + "loss": 3.1835, + "step": 22049 + }, + { + "epoch": 1.7677537030726986, + "grad_norm": 1.7667450904846191, + "learning_rate": 6.977489594841547e-07, + "loss": 3.1893, + "step": 22050 + }, + { + "epoch": 1.7678338778537213, + "grad_norm": 1.8126003742218018, + "learning_rate": 6.972725215317477e-07, + "loss": 3.2573, + "step": 22051 + }, + { + "epoch": 1.7679140526347439, + "grad_norm": 1.7354955673217773, + "learning_rate": 6.967962404200479e-07, + "loss": 3.197, + "step": 22052 + }, + { + "epoch": 1.7679942274157665, + "grad_norm": 1.7846580743789673, + "learning_rate": 6.963201161570843e-07, + "loss": 3.1991, + "step": 22053 + }, + { + "epoch": 1.7680744021967891, + "grad_norm": 0.6701536178588867, + "learning_rate": 6.958441487508838e-07, + "loss": 2.4581, + "step": 22054 + }, + { + "epoch": 1.7681545769778118, + "grad_norm": 1.8122482299804688, + "learning_rate": 6.953683382094745e-07, + "loss": 3.1357, + "step": 22055 + }, + { + "epoch": 1.7682347517588344, + "grad_norm": 1.745315670967102, + "learning_rate": 6.948926845408743e-07, + "loss": 3.0941, + "step": 22056 + }, + { + "epoch": 1.768314926539857, + "grad_norm": 0.6569187641143799, + "learning_rate": 6.944171877531036e-07, + "loss": 2.4495, + "step": 22057 + }, + { + "epoch": 1.7683951013208796, + "grad_norm": 1.7721869945526123, + "learning_rate": 6.939418478541793e-07, + "loss": 3.2027, + "step": 22058 + }, + { + "epoch": 1.7684752761019022, + "grad_norm": 1.6738402843475342, + "learning_rate": 6.934666648521149e-07, + "loss": 3.0842, + "step": 22059 + }, + { + "epoch": 1.7685554508829249, + "grad_norm": 1.843762993812561, + "learning_rate": 6.929916387549251e-07, + "loss": 3.1672, + "step": 22060 + }, + { + "epoch": 1.7686356256639475, + "grad_norm": 1.7203662395477295, + "learning_rate": 6.925167695706147e-07, + "loss": 3.1255, + "step": 22061 + }, + { + "epoch": 1.7687158004449701, + "grad_norm": 1.7635540962219238, + "learning_rate": 6.920420573071917e-07, + "loss": 3.2435, + "step": 22062 + }, + { + "epoch": 1.7687959752259927, + "grad_norm": 1.7275242805480957, + "learning_rate": 6.915675019726587e-07, + "loss": 3.1378, + "step": 22063 + }, + { + "epoch": 1.7688761500070154, + "grad_norm": 1.8551456928253174, + "learning_rate": 6.910931035750168e-07, + "loss": 3.1473, + "step": 22064 + }, + { + "epoch": 1.768956324788038, + "grad_norm": 1.7159500122070312, + "learning_rate": 6.906188621222653e-07, + "loss": 3.2202, + "step": 22065 + }, + { + "epoch": 1.7690364995690606, + "grad_norm": 1.9394220113754272, + "learning_rate": 6.90144777622399e-07, + "loss": 3.2015, + "step": 22066 + }, + { + "epoch": 1.7691166743500832, + "grad_norm": 1.8314605951309204, + "learning_rate": 6.896708500834093e-07, + "loss": 3.1588, + "step": 22067 + }, + { + "epoch": 1.7691968491311059, + "grad_norm": 1.7186297178268433, + "learning_rate": 6.891970795132885e-07, + "loss": 3.2156, + "step": 22068 + }, + { + "epoch": 1.7692770239121285, + "grad_norm": 1.6420338153839111, + "learning_rate": 6.887234659200226e-07, + "loss": 3.075, + "step": 22069 + }, + { + "epoch": 1.7693571986931511, + "grad_norm": 1.6905320882797241, + "learning_rate": 6.882500093115996e-07, + "loss": 3.2295, + "step": 22070 + }, + { + "epoch": 1.7694373734741737, + "grad_norm": 1.740979552268982, + "learning_rate": 6.877767096959986e-07, + "loss": 3.0704, + "step": 22071 + }, + { + "epoch": 1.7695175482551964, + "grad_norm": 1.8522076606750488, + "learning_rate": 6.873035670812001e-07, + "loss": 3.201, + "step": 22072 + }, + { + "epoch": 1.769597723036219, + "grad_norm": 0.666064441204071, + "learning_rate": 6.868305814751808e-07, + "loss": 2.4739, + "step": 22073 + }, + { + "epoch": 1.7696778978172416, + "grad_norm": 1.841209053993225, + "learning_rate": 6.863577528859155e-07, + "loss": 3.248, + "step": 22074 + }, + { + "epoch": 1.7697580725982642, + "grad_norm": 1.7790391445159912, + "learning_rate": 6.85885081321378e-07, + "loss": 3.2849, + "step": 22075 + }, + { + "epoch": 1.7698382473792869, + "grad_norm": 1.7369121313095093, + "learning_rate": 6.854125667895339e-07, + "loss": 3.1713, + "step": 22076 + }, + { + "epoch": 1.7699184221603095, + "grad_norm": 1.6667287349700928, + "learning_rate": 6.849402092983514e-07, + "loss": 3.2471, + "step": 22077 + }, + { + "epoch": 1.769998596941332, + "grad_norm": 1.7920643091201782, + "learning_rate": 6.844680088557931e-07, + "loss": 3.1606, + "step": 22078 + }, + { + "epoch": 1.7700787717223547, + "grad_norm": 1.7615591287612915, + "learning_rate": 6.839959654698225e-07, + "loss": 3.1273, + "step": 22079 + }, + { + "epoch": 1.7701589465033774, + "grad_norm": 1.7440228462219238, + "learning_rate": 6.835240791483966e-07, + "loss": 3.152, + "step": 22080 + }, + { + "epoch": 1.7702391212844, + "grad_norm": 1.7810394763946533, + "learning_rate": 6.830523498994712e-07, + "loss": 3.1826, + "step": 22081 + }, + { + "epoch": 1.7703192960654226, + "grad_norm": 0.6501046419143677, + "learning_rate": 6.825807777309989e-07, + "loss": 2.4692, + "step": 22082 + }, + { + "epoch": 1.7703994708464452, + "grad_norm": 0.6717592477798462, + "learning_rate": 6.821093626509301e-07, + "loss": 2.4618, + "step": 22083 + }, + { + "epoch": 1.7704796456274678, + "grad_norm": 1.7400764226913452, + "learning_rate": 6.81638104667215e-07, + "loss": 3.1746, + "step": 22084 + }, + { + "epoch": 1.7705598204084905, + "grad_norm": 1.7613619565963745, + "learning_rate": 6.811670037877982e-07, + "loss": 3.1436, + "step": 22085 + }, + { + "epoch": 1.770639995189513, + "grad_norm": 1.8293389081954956, + "learning_rate": 6.806960600206191e-07, + "loss": 3.2175, + "step": 22086 + }, + { + "epoch": 1.7707201699705357, + "grad_norm": 0.6704058647155762, + "learning_rate": 6.802252733736214e-07, + "loss": 2.4686, + "step": 22087 + }, + { + "epoch": 1.7708003447515583, + "grad_norm": 1.8229223489761353, + "learning_rate": 6.797546438547397e-07, + "loss": 3.1173, + "step": 22088 + }, + { + "epoch": 1.770880519532581, + "grad_norm": 1.733001947402954, + "learning_rate": 6.792841714719111e-07, + "loss": 3.1523, + "step": 22089 + }, + { + "epoch": 1.7709606943136036, + "grad_norm": 0.6575987339019775, + "learning_rate": 6.788138562330671e-07, + "loss": 2.4507, + "step": 22090 + }, + { + "epoch": 1.7710408690946262, + "grad_norm": 1.7987539768218994, + "learning_rate": 6.783436981461356e-07, + "loss": 3.2045, + "step": 22091 + }, + { + "epoch": 1.7711210438756488, + "grad_norm": 1.682403326034546, + "learning_rate": 6.778736972190437e-07, + "loss": 3.0603, + "step": 22092 + }, + { + "epoch": 1.7712012186566715, + "grad_norm": 1.7083486318588257, + "learning_rate": 6.774038534597161e-07, + "loss": 3.1386, + "step": 22093 + }, + { + "epoch": 1.771281393437694, + "grad_norm": 1.8573360443115234, + "learning_rate": 6.769341668760731e-07, + "loss": 3.2216, + "step": 22094 + }, + { + "epoch": 1.7713615682187167, + "grad_norm": 1.7727218866348267, + "learning_rate": 6.764646374760364e-07, + "loss": 3.2028, + "step": 22095 + }, + { + "epoch": 1.7714417429997393, + "grad_norm": 1.8245655298233032, + "learning_rate": 6.759952652675184e-07, + "loss": 3.1392, + "step": 22096 + }, + { + "epoch": 1.771521917780762, + "grad_norm": 2.004513740539551, + "learning_rate": 6.755260502584338e-07, + "loss": 3.1372, + "step": 22097 + }, + { + "epoch": 1.7716020925617846, + "grad_norm": 1.837747573852539, + "learning_rate": 6.750569924566941e-07, + "loss": 3.1843, + "step": 22098 + }, + { + "epoch": 1.7716822673428072, + "grad_norm": 1.7629528045654297, + "learning_rate": 6.745880918702074e-07, + "loss": 3.2032, + "step": 22099 + }, + { + "epoch": 1.7717624421238298, + "grad_norm": 1.7319505214691162, + "learning_rate": 6.741193485068798e-07, + "loss": 3.2079, + "step": 22100 + }, + { + "epoch": 1.7718426169048525, + "grad_norm": 1.7111725807189941, + "learning_rate": 6.736507623746125e-07, + "loss": 3.0932, + "step": 22101 + }, + { + "epoch": 1.771922791685875, + "grad_norm": 1.6877361536026, + "learning_rate": 6.731823334813048e-07, + "loss": 3.1489, + "step": 22102 + }, + { + "epoch": 1.7720029664668977, + "grad_norm": 1.7878611087799072, + "learning_rate": 6.727140618348571e-07, + "loss": 3.2358, + "step": 22103 + }, + { + "epoch": 1.7720831412479203, + "grad_norm": 1.8171484470367432, + "learning_rate": 6.722459474431631e-07, + "loss": 3.2321, + "step": 22104 + }, + { + "epoch": 1.772163316028943, + "grad_norm": 1.7022508382797241, + "learning_rate": 6.717779903141164e-07, + "loss": 3.1785, + "step": 22105 + }, + { + "epoch": 1.7722434908099656, + "grad_norm": 0.6524787545204163, + "learning_rate": 6.71310190455604e-07, + "loss": 2.4606, + "step": 22106 + }, + { + "epoch": 1.7723236655909882, + "grad_norm": 1.7540229558944702, + "learning_rate": 6.708425478755131e-07, + "loss": 3.1357, + "step": 22107 + }, + { + "epoch": 1.7724038403720108, + "grad_norm": 1.7442442178726196, + "learning_rate": 6.703750625817296e-07, + "loss": 3.17, + "step": 22108 + }, + { + "epoch": 1.7724840151530337, + "grad_norm": 1.9440231323242188, + "learning_rate": 6.699077345821337e-07, + "loss": 3.1671, + "step": 22109 + }, + { + "epoch": 1.7725641899340563, + "grad_norm": 1.8039520978927612, + "learning_rate": 6.694405638846079e-07, + "loss": 3.1671, + "step": 22110 + }, + { + "epoch": 1.772644364715079, + "grad_norm": 1.8444277048110962, + "learning_rate": 6.689735504970241e-07, + "loss": 3.1809, + "step": 22111 + }, + { + "epoch": 1.7727245394961015, + "grad_norm": 1.7166094779968262, + "learning_rate": 6.685066944272567e-07, + "loss": 3.1423, + "step": 22112 + }, + { + "epoch": 1.7728047142771242, + "grad_norm": 1.8086612224578857, + "learning_rate": 6.680399956831785e-07, + "loss": 3.1782, + "step": 22113 + }, + { + "epoch": 1.7728848890581468, + "grad_norm": 1.7183197736740112, + "learning_rate": 6.675734542726564e-07, + "loss": 3.3091, + "step": 22114 + }, + { + "epoch": 1.7729650638391694, + "grad_norm": 1.8322142362594604, + "learning_rate": 6.671070702035587e-07, + "loss": 3.1388, + "step": 22115 + }, + { + "epoch": 1.773045238620192, + "grad_norm": 1.662564754486084, + "learning_rate": 6.666408434837457e-07, + "loss": 3.1403, + "step": 22116 + }, + { + "epoch": 1.7731254134012147, + "grad_norm": 1.7399818897247314, + "learning_rate": 6.661747741210778e-07, + "loss": 3.1333, + "step": 22117 + }, + { + "epoch": 1.7732055881822373, + "grad_norm": 1.768861174583435, + "learning_rate": 6.657088621234131e-07, + "loss": 3.1858, + "step": 22118 + }, + { + "epoch": 1.77328576296326, + "grad_norm": 0.6529321074485779, + "learning_rate": 6.652431074986088e-07, + "loss": 2.4462, + "step": 22119 + }, + { + "epoch": 1.7733659377442825, + "grad_norm": 1.7754795551300049, + "learning_rate": 6.647775102545162e-07, + "loss": 3.2054, + "step": 22120 + }, + { + "epoch": 1.7734461125253052, + "grad_norm": 0.6438304781913757, + "learning_rate": 6.643120703989836e-07, + "loss": 2.4635, + "step": 22121 + }, + { + "epoch": 1.7735262873063278, + "grad_norm": 1.8335140943527222, + "learning_rate": 6.638467879398592e-07, + "loss": 3.1443, + "step": 22122 + }, + { + "epoch": 1.7736064620873504, + "grad_norm": 1.9366941452026367, + "learning_rate": 6.633816628849876e-07, + "loss": 3.2439, + "step": 22123 + }, + { + "epoch": 1.773686636868373, + "grad_norm": 1.7645657062530518, + "learning_rate": 6.629166952422106e-07, + "loss": 3.197, + "step": 22124 + }, + { + "epoch": 1.7737668116493956, + "grad_norm": 1.584958791732788, + "learning_rate": 6.624518850193695e-07, + "loss": 3.2076, + "step": 22125 + }, + { + "epoch": 1.7738469864304183, + "grad_norm": 1.6660399436950684, + "learning_rate": 6.619872322242959e-07, + "loss": 3.1529, + "step": 22126 + }, + { + "epoch": 1.773927161211441, + "grad_norm": 1.757759690284729, + "learning_rate": 6.615227368648281e-07, + "loss": 3.1722, + "step": 22127 + }, + { + "epoch": 1.7740073359924635, + "grad_norm": 1.8445836305618286, + "learning_rate": 6.61058398948795e-07, + "loss": 3.0903, + "step": 22128 + }, + { + "epoch": 1.7740875107734864, + "grad_norm": 1.8290014266967773, + "learning_rate": 6.605942184840253e-07, + "loss": 3.2607, + "step": 22129 + }, + { + "epoch": 1.774167685554509, + "grad_norm": 1.7788876295089722, + "learning_rate": 6.601301954783479e-07, + "loss": 3.2338, + "step": 22130 + }, + { + "epoch": 1.7742478603355316, + "grad_norm": 1.670548915863037, + "learning_rate": 6.596663299395823e-07, + "loss": 3.1899, + "step": 22131 + }, + { + "epoch": 1.7743280351165542, + "grad_norm": 1.6796727180480957, + "learning_rate": 6.592026218755499e-07, + "loss": 3.0742, + "step": 22132 + }, + { + "epoch": 1.7744082098975769, + "grad_norm": 1.8065391778945923, + "learning_rate": 6.5873907129407e-07, + "loss": 3.1613, + "step": 22133 + }, + { + "epoch": 1.7744883846785995, + "grad_norm": 1.9412506818771362, + "learning_rate": 6.582756782029565e-07, + "loss": 3.2024, + "step": 22134 + }, + { + "epoch": 1.774568559459622, + "grad_norm": 0.636414647102356, + "learning_rate": 6.578124426100252e-07, + "loss": 2.4456, + "step": 22135 + }, + { + "epoch": 1.7746487342406447, + "grad_norm": 1.8351588249206543, + "learning_rate": 6.57349364523081e-07, + "loss": 3.1326, + "step": 22136 + }, + { + "epoch": 1.7747289090216674, + "grad_norm": 1.8165804147720337, + "learning_rate": 6.568864439499345e-07, + "loss": 3.1355, + "step": 22137 + }, + { + "epoch": 1.77480908380269, + "grad_norm": 1.790624976158142, + "learning_rate": 6.564236808983893e-07, + "loss": 3.2037, + "step": 22138 + }, + { + "epoch": 1.7748892585837126, + "grad_norm": 0.6537156701087952, + "learning_rate": 6.55961075376248e-07, + "loss": 2.468, + "step": 22139 + }, + { + "epoch": 1.7749694333647352, + "grad_norm": 0.6602160930633545, + "learning_rate": 6.554986273913111e-07, + "loss": 2.4756, + "step": 22140 + }, + { + "epoch": 1.7750496081457579, + "grad_norm": 1.8222993612289429, + "learning_rate": 6.550363369513723e-07, + "loss": 3.0758, + "step": 22141 + }, + { + "epoch": 1.7751297829267805, + "grad_norm": 1.7250690460205078, + "learning_rate": 6.545742040642289e-07, + "loss": 3.1077, + "step": 22142 + }, + { + "epoch": 1.775209957707803, + "grad_norm": 1.7943673133850098, + "learning_rate": 6.541122287376689e-07, + "loss": 3.2256, + "step": 22143 + }, + { + "epoch": 1.7752901324888257, + "grad_norm": 1.7813441753387451, + "learning_rate": 6.536504109794839e-07, + "loss": 3.2867, + "step": 22144 + }, + { + "epoch": 1.7753703072698483, + "grad_norm": 1.7617300748825073, + "learning_rate": 6.531887507974599e-07, + "loss": 3.1063, + "step": 22145 + }, + { + "epoch": 1.775450482050871, + "grad_norm": 1.6858553886413574, + "learning_rate": 6.527272481993785e-07, + "loss": 3.1952, + "step": 22146 + }, + { + "epoch": 1.7755306568318936, + "grad_norm": 0.6546177864074707, + "learning_rate": 6.522659031930212e-07, + "loss": 2.4662, + "step": 22147 + }, + { + "epoch": 1.7756108316129162, + "grad_norm": 1.9081604480743408, + "learning_rate": 6.518047157861662e-07, + "loss": 3.1934, + "step": 22148 + }, + { + "epoch": 1.7756910063939388, + "grad_norm": 0.6386083364486694, + "learning_rate": 6.513436859865895e-07, + "loss": 2.4678, + "step": 22149 + }, + { + "epoch": 1.7757711811749615, + "grad_norm": 1.7572453022003174, + "learning_rate": 6.508828138020651e-07, + "loss": 3.0989, + "step": 22150 + }, + { + "epoch": 1.775851355955984, + "grad_norm": 1.6520891189575195, + "learning_rate": 6.504220992403599e-07, + "loss": 3.2181, + "step": 22151 + }, + { + "epoch": 1.7759315307370067, + "grad_norm": 1.8691273927688599, + "learning_rate": 6.499615423092421e-07, + "loss": 3.109, + "step": 22152 + }, + { + "epoch": 1.7760117055180293, + "grad_norm": 1.9338064193725586, + "learning_rate": 6.495011430164778e-07, + "loss": 3.2225, + "step": 22153 + }, + { + "epoch": 1.776091880299052, + "grad_norm": 1.7475885152816772, + "learning_rate": 6.490409013698296e-07, + "loss": 3.187, + "step": 22154 + }, + { + "epoch": 1.7761720550800746, + "grad_norm": 1.74395751953125, + "learning_rate": 6.485808173770569e-07, + "loss": 3.1295, + "step": 22155 + }, + { + "epoch": 1.7762522298610972, + "grad_norm": 1.8070403337478638, + "learning_rate": 6.481208910459147e-07, + "loss": 3.1562, + "step": 22156 + }, + { + "epoch": 1.7763324046421198, + "grad_norm": 1.7498786449432373, + "learning_rate": 6.476611223841578e-07, + "loss": 3.1252, + "step": 22157 + }, + { + "epoch": 1.7764125794231425, + "grad_norm": 1.7224315404891968, + "learning_rate": 6.472015113995389e-07, + "loss": 3.205, + "step": 22158 + }, + { + "epoch": 1.776492754204165, + "grad_norm": 1.848778486251831, + "learning_rate": 6.467420580998063e-07, + "loss": 3.1287, + "step": 22159 + }, + { + "epoch": 1.7765729289851877, + "grad_norm": 0.6705501079559326, + "learning_rate": 6.462827624927071e-07, + "loss": 2.4597, + "step": 22160 + }, + { + "epoch": 1.7766531037662103, + "grad_norm": 1.702349066734314, + "learning_rate": 6.458236245859828e-07, + "loss": 3.1794, + "step": 22161 + }, + { + "epoch": 1.776733278547233, + "grad_norm": 1.7465555667877197, + "learning_rate": 6.453646443873762e-07, + "loss": 3.2449, + "step": 22162 + }, + { + "epoch": 1.7768134533282556, + "grad_norm": 1.8321588039398193, + "learning_rate": 6.449058219046234e-07, + "loss": 3.1398, + "step": 22163 + }, + { + "epoch": 1.7768936281092782, + "grad_norm": 1.9153316020965576, + "learning_rate": 6.444471571454624e-07, + "loss": 3.0927, + "step": 22164 + }, + { + "epoch": 1.7769738028903008, + "grad_norm": 1.719918131828308, + "learning_rate": 6.439886501176262e-07, + "loss": 3.1355, + "step": 22165 + }, + { + "epoch": 1.7770539776713234, + "grad_norm": 1.7323238849639893, + "learning_rate": 6.435303008288429e-07, + "loss": 3.1272, + "step": 22166 + }, + { + "epoch": 1.777134152452346, + "grad_norm": 1.706191897392273, + "learning_rate": 6.430721092868408e-07, + "loss": 3.2106, + "step": 22167 + }, + { + "epoch": 1.7772143272333687, + "grad_norm": 1.7284725904464722, + "learning_rate": 6.426140754993449e-07, + "loss": 3.091, + "step": 22168 + }, + { + "epoch": 1.7772945020143913, + "grad_norm": 1.7297908067703247, + "learning_rate": 6.421561994740788e-07, + "loss": 3.1316, + "step": 22169 + }, + { + "epoch": 1.777374676795414, + "grad_norm": 1.8625385761260986, + "learning_rate": 6.416984812187621e-07, + "loss": 3.2078, + "step": 22170 + }, + { + "epoch": 1.7774548515764366, + "grad_norm": 1.6974079608917236, + "learning_rate": 6.412409207411085e-07, + "loss": 3.0802, + "step": 22171 + }, + { + "epoch": 1.7775350263574592, + "grad_norm": 1.8103477954864502, + "learning_rate": 6.407835180488364e-07, + "loss": 3.179, + "step": 22172 + }, + { + "epoch": 1.7776152011384818, + "grad_norm": 1.7603759765625, + "learning_rate": 6.40326273149654e-07, + "loss": 3.2432, + "step": 22173 + }, + { + "epoch": 1.7776953759195044, + "grad_norm": 1.7249704599380493, + "learning_rate": 6.39869186051273e-07, + "loss": 3.172, + "step": 22174 + }, + { + "epoch": 1.777775550700527, + "grad_norm": 1.8406307697296143, + "learning_rate": 6.394122567614003e-07, + "loss": 3.1419, + "step": 22175 + }, + { + "epoch": 1.7778557254815497, + "grad_norm": 0.6612259745597839, + "learning_rate": 6.389554852877355e-07, + "loss": 2.4584, + "step": 22176 + }, + { + "epoch": 1.7779359002625723, + "grad_norm": 1.710536003112793, + "learning_rate": 6.384988716379825e-07, + "loss": 3.1738, + "step": 22177 + }, + { + "epoch": 1.778016075043595, + "grad_norm": 1.705268383026123, + "learning_rate": 6.380424158198395e-07, + "loss": 3.2411, + "step": 22178 + }, + { + "epoch": 1.7780962498246176, + "grad_norm": 1.822251558303833, + "learning_rate": 6.375861178410014e-07, + "loss": 3.1758, + "step": 22179 + }, + { + "epoch": 1.7781764246056402, + "grad_norm": 0.6460837125778198, + "learning_rate": 6.371299777091644e-07, + "loss": 2.4603, + "step": 22180 + }, + { + "epoch": 1.7782565993866628, + "grad_norm": 1.7516101598739624, + "learning_rate": 6.366739954320134e-07, + "loss": 3.1335, + "step": 22181 + }, + { + "epoch": 1.7783367741676854, + "grad_norm": 1.8556108474731445, + "learning_rate": 6.362181710172388e-07, + "loss": 3.1237, + "step": 22182 + }, + { + "epoch": 1.778416948948708, + "grad_norm": 1.6534937620162964, + "learning_rate": 6.357625044725268e-07, + "loss": 3.2313, + "step": 22183 + }, + { + "epoch": 1.7784971237297307, + "grad_norm": 1.7608650922775269, + "learning_rate": 6.35306995805558e-07, + "loss": 3.2374, + "step": 22184 + }, + { + "epoch": 1.7785772985107533, + "grad_norm": 1.8005942106246948, + "learning_rate": 6.348516450240139e-07, + "loss": 3.1256, + "step": 22185 + }, + { + "epoch": 1.778657473291776, + "grad_norm": 1.889927864074707, + "learning_rate": 6.343964521355695e-07, + "loss": 3.3531, + "step": 22186 + }, + { + "epoch": 1.7787376480727985, + "grad_norm": 1.8129119873046875, + "learning_rate": 6.339414171479008e-07, + "loss": 3.1177, + "step": 22187 + }, + { + "epoch": 1.7788178228538212, + "grad_norm": 1.748854637145996, + "learning_rate": 6.334865400686785e-07, + "loss": 3.2666, + "step": 22188 + }, + { + "epoch": 1.7788979976348438, + "grad_norm": 1.8443429470062256, + "learning_rate": 6.330318209055708e-07, + "loss": 3.1586, + "step": 22189 + }, + { + "epoch": 1.7789781724158666, + "grad_norm": 1.752550482749939, + "learning_rate": 6.325772596662482e-07, + "loss": 3.219, + "step": 22190 + }, + { + "epoch": 1.7790583471968893, + "grad_norm": 1.7298893928527832, + "learning_rate": 6.321228563583704e-07, + "loss": 3.1882, + "step": 22191 + }, + { + "epoch": 1.779138521977912, + "grad_norm": 0.6833195090293884, + "learning_rate": 6.316686109895986e-07, + "loss": 2.4811, + "step": 22192 + }, + { + "epoch": 1.7792186967589345, + "grad_norm": 1.6721422672271729, + "learning_rate": 6.312145235675926e-07, + "loss": 3.1653, + "step": 22193 + }, + { + "epoch": 1.7792988715399571, + "grad_norm": 1.785014033317566, + "learning_rate": 6.307605941000084e-07, + "loss": 3.2697, + "step": 22194 + }, + { + "epoch": 1.7793790463209798, + "grad_norm": 1.7758103609085083, + "learning_rate": 6.303068225944997e-07, + "loss": 3.1577, + "step": 22195 + }, + { + "epoch": 1.7794592211020024, + "grad_norm": 1.709715485572815, + "learning_rate": 6.29853209058714e-07, + "loss": 3.1431, + "step": 22196 + }, + { + "epoch": 1.779539395883025, + "grad_norm": 1.8232388496398926, + "learning_rate": 6.293997535003004e-07, + "loss": 3.1553, + "step": 22197 + }, + { + "epoch": 1.7796195706640476, + "grad_norm": 1.847147822380066, + "learning_rate": 6.289464559269054e-07, + "loss": 3.1404, + "step": 22198 + }, + { + "epoch": 1.7796997454450703, + "grad_norm": 1.8366519212722778, + "learning_rate": 6.284933163461704e-07, + "loss": 3.1823, + "step": 22199 + }, + { + "epoch": 1.7797799202260929, + "grad_norm": 0.6572603583335876, + "learning_rate": 6.28040334765736e-07, + "loss": 2.4696, + "step": 22200 + }, + { + "epoch": 1.7798600950071155, + "grad_norm": 1.826100468635559, + "learning_rate": 6.275875111932373e-07, + "loss": 3.2316, + "step": 22201 + }, + { + "epoch": 1.7799402697881381, + "grad_norm": 1.736412763595581, + "learning_rate": 6.271348456363102e-07, + "loss": 3.1761, + "step": 22202 + }, + { + "epoch": 1.7800204445691608, + "grad_norm": 1.7917444705963135, + "learning_rate": 6.266823381025867e-07, + "loss": 3.147, + "step": 22203 + }, + { + "epoch": 1.7801006193501834, + "grad_norm": 1.999140739440918, + "learning_rate": 6.262299885996947e-07, + "loss": 3.1743, + "step": 22204 + }, + { + "epoch": 1.780180794131206, + "grad_norm": 1.862986445426941, + "learning_rate": 6.257777971352641e-07, + "loss": 3.1086, + "step": 22205 + }, + { + "epoch": 1.7802609689122286, + "grad_norm": 1.8547930717468262, + "learning_rate": 6.253257637169141e-07, + "loss": 3.1339, + "step": 22206 + }, + { + "epoch": 1.7803411436932512, + "grad_norm": 1.8274935483932495, + "learning_rate": 6.248738883522676e-07, + "loss": 3.1978, + "step": 22207 + }, + { + "epoch": 1.7804213184742739, + "grad_norm": 1.6978747844696045, + "learning_rate": 6.244221710489429e-07, + "loss": 3.0762, + "step": 22208 + }, + { + "epoch": 1.7805014932552965, + "grad_norm": 1.8304647207260132, + "learning_rate": 6.239706118145561e-07, + "loss": 3.1859, + "step": 22209 + }, + { + "epoch": 1.7805816680363191, + "grad_norm": 1.7289423942565918, + "learning_rate": 6.235192106567223e-07, + "loss": 3.0977, + "step": 22210 + }, + { + "epoch": 1.780661842817342, + "grad_norm": 1.8195164203643799, + "learning_rate": 6.230679675830476e-07, + "loss": 3.1226, + "step": 22211 + }, + { + "epoch": 1.7807420175983646, + "grad_norm": 1.7509536743164062, + "learning_rate": 6.226168826011436e-07, + "loss": 3.2117, + "step": 22212 + }, + { + "epoch": 1.7808221923793872, + "grad_norm": 1.751790165901184, + "learning_rate": 6.221659557186132e-07, + "loss": 3.0916, + "step": 22213 + }, + { + "epoch": 1.7809023671604098, + "grad_norm": 1.84070885181427, + "learning_rate": 6.217151869430593e-07, + "loss": 3.2374, + "step": 22214 + }, + { + "epoch": 1.7809825419414325, + "grad_norm": 1.759037971496582, + "learning_rate": 6.212645762820845e-07, + "loss": 3.1819, + "step": 22215 + }, + { + "epoch": 1.781062716722455, + "grad_norm": 1.8590095043182373, + "learning_rate": 6.208141237432818e-07, + "loss": 3.2054, + "step": 22216 + }, + { + "epoch": 1.7811428915034777, + "grad_norm": 1.7029454708099365, + "learning_rate": 6.203638293342474e-07, + "loss": 3.1946, + "step": 22217 + }, + { + "epoch": 1.7812230662845003, + "grad_norm": 1.7964167594909668, + "learning_rate": 6.199136930625737e-07, + "loss": 3.1264, + "step": 22218 + }, + { + "epoch": 1.781303241065523, + "grad_norm": 1.8534488677978516, + "learning_rate": 6.194637149358484e-07, + "loss": 3.148, + "step": 22219 + }, + { + "epoch": 1.7813834158465456, + "grad_norm": 1.9137324094772339, + "learning_rate": 6.190138949616609e-07, + "loss": 3.2074, + "step": 22220 + }, + { + "epoch": 1.7814635906275682, + "grad_norm": 1.7897425889968872, + "learning_rate": 6.185642331475905e-07, + "loss": 3.1133, + "step": 22221 + }, + { + "epoch": 1.7815437654085908, + "grad_norm": 1.8328031301498413, + "learning_rate": 6.181147295012225e-07, + "loss": 3.1821, + "step": 22222 + }, + { + "epoch": 1.7816239401896135, + "grad_norm": 1.7906044721603394, + "learning_rate": 6.176653840301316e-07, + "loss": 3.0779, + "step": 22223 + }, + { + "epoch": 1.781704114970636, + "grad_norm": 1.7993121147155762, + "learning_rate": 6.172161967418977e-07, + "loss": 3.167, + "step": 22224 + }, + { + "epoch": 1.7817842897516587, + "grad_norm": 1.7477965354919434, + "learning_rate": 6.167671676440923e-07, + "loss": 3.136, + "step": 22225 + }, + { + "epoch": 1.7818644645326813, + "grad_norm": 1.807728886604309, + "learning_rate": 6.163182967442837e-07, + "loss": 3.166, + "step": 22226 + }, + { + "epoch": 1.781944639313704, + "grad_norm": 1.741926908493042, + "learning_rate": 6.158695840500417e-07, + "loss": 3.2228, + "step": 22227 + }, + { + "epoch": 1.7820248140947266, + "grad_norm": 1.6540828943252563, + "learning_rate": 6.154210295689311e-07, + "loss": 3.133, + "step": 22228 + }, + { + "epoch": 1.7821049888757492, + "grad_norm": 1.8248354196548462, + "learning_rate": 6.149726333085149e-07, + "loss": 3.2166, + "step": 22229 + }, + { + "epoch": 1.7821851636567718, + "grad_norm": 1.7881745100021362, + "learning_rate": 6.145243952763546e-07, + "loss": 3.2119, + "step": 22230 + }, + { + "epoch": 1.7822653384377944, + "grad_norm": 1.816094160079956, + "learning_rate": 6.140763154800033e-07, + "loss": 3.2573, + "step": 22231 + }, + { + "epoch": 1.782345513218817, + "grad_norm": 1.7069573402404785, + "learning_rate": 6.136283939270171e-07, + "loss": 3.1603, + "step": 22232 + }, + { + "epoch": 1.7824256879998397, + "grad_norm": 1.723034143447876, + "learning_rate": 6.131806306249477e-07, + "loss": 3.1947, + "step": 22233 + }, + { + "epoch": 1.7825058627808623, + "grad_norm": 1.788012981414795, + "learning_rate": 6.127330255813446e-07, + "loss": 3.2531, + "step": 22234 + }, + { + "epoch": 1.782586037561885, + "grad_norm": 1.6923648118972778, + "learning_rate": 6.122855788037574e-07, + "loss": 3.0369, + "step": 22235 + }, + { + "epoch": 1.7826662123429076, + "grad_norm": 1.7996783256530762, + "learning_rate": 6.118382902997233e-07, + "loss": 3.1803, + "step": 22236 + }, + { + "epoch": 1.7827463871239302, + "grad_norm": 1.7724188566207886, + "learning_rate": 6.113911600767885e-07, + "loss": 3.0482, + "step": 22237 + }, + { + "epoch": 1.7828265619049528, + "grad_norm": 0.6517345905303955, + "learning_rate": 6.109441881424882e-07, + "loss": 2.4684, + "step": 22238 + }, + { + "epoch": 1.7829067366859754, + "grad_norm": 1.8122133016586304, + "learning_rate": 6.104973745043619e-07, + "loss": 3.2228, + "step": 22239 + }, + { + "epoch": 1.782986911466998, + "grad_norm": 1.7571638822555542, + "learning_rate": 6.100507191699401e-07, + "loss": 3.1286, + "step": 22240 + }, + { + "epoch": 1.7830670862480207, + "grad_norm": 1.7954833507537842, + "learning_rate": 6.096042221467536e-07, + "loss": 3.2026, + "step": 22241 + }, + { + "epoch": 1.7831472610290433, + "grad_norm": 1.7816827297210693, + "learning_rate": 6.091578834423306e-07, + "loss": 3.0706, + "step": 22242 + }, + { + "epoch": 1.783227435810066, + "grad_norm": 1.7694424390792847, + "learning_rate": 6.087117030641953e-07, + "loss": 3.0518, + "step": 22243 + }, + { + "epoch": 1.7833076105910886, + "grad_norm": 1.7897911071777344, + "learning_rate": 6.082656810198717e-07, + "loss": 3.2247, + "step": 22244 + }, + { + "epoch": 1.7833877853721112, + "grad_norm": 1.7976409196853638, + "learning_rate": 6.078198173168792e-07, + "loss": 3.1772, + "step": 22245 + }, + { + "epoch": 1.7834679601531338, + "grad_norm": 0.6470176577568054, + "learning_rate": 6.07374111962733e-07, + "loss": 2.4681, + "step": 22246 + }, + { + "epoch": 1.7835481349341564, + "grad_norm": 1.8067771196365356, + "learning_rate": 6.069285649649503e-07, + "loss": 3.2012, + "step": 22247 + }, + { + "epoch": 1.783628309715179, + "grad_norm": 1.58943510055542, + "learning_rate": 6.064831763310408e-07, + "loss": 3.0023, + "step": 22248 + }, + { + "epoch": 1.7837084844962017, + "grad_norm": 1.705388069152832, + "learning_rate": 6.06037946068514e-07, + "loss": 3.2014, + "step": 22249 + }, + { + "epoch": 1.7837886592772243, + "grad_norm": 1.8604178428649902, + "learning_rate": 6.055928741848782e-07, + "loss": 3.2058, + "step": 22250 + }, + { + "epoch": 1.783868834058247, + "grad_norm": 1.7095489501953125, + "learning_rate": 6.051479606876342e-07, + "loss": 3.105, + "step": 22251 + }, + { + "epoch": 1.7839490088392695, + "grad_norm": 0.6449933052062988, + "learning_rate": 6.04703205584285e-07, + "loss": 2.4631, + "step": 22252 + }, + { + "epoch": 1.7840291836202922, + "grad_norm": 1.7874102592468262, + "learning_rate": 6.042586088823277e-07, + "loss": 3.1624, + "step": 22253 + }, + { + "epoch": 1.7841093584013148, + "grad_norm": 1.7152354717254639, + "learning_rate": 6.038141705892597e-07, + "loss": 3.2593, + "step": 22254 + }, + { + "epoch": 1.7841895331823374, + "grad_norm": 1.8111910820007324, + "learning_rate": 6.033698907125751e-07, + "loss": 3.1576, + "step": 22255 + }, + { + "epoch": 1.78426970796336, + "grad_norm": 1.822668433189392, + "learning_rate": 6.029257692597601e-07, + "loss": 3.2343, + "step": 22256 + }, + { + "epoch": 1.7843498827443827, + "grad_norm": 1.793035864830017, + "learning_rate": 6.024818062383042e-07, + "loss": 3.2752, + "step": 22257 + }, + { + "epoch": 1.7844300575254053, + "grad_norm": 1.7198104858398438, + "learning_rate": 6.020380016556948e-07, + "loss": 3.147, + "step": 22258 + }, + { + "epoch": 1.784510232306428, + "grad_norm": 1.773574948310852, + "learning_rate": 6.015943555194115e-07, + "loss": 3.1191, + "step": 22259 + }, + { + "epoch": 1.7845904070874505, + "grad_norm": 1.8037961721420288, + "learning_rate": 6.01150867836936e-07, + "loss": 3.0605, + "step": 22260 + }, + { + "epoch": 1.7846705818684732, + "grad_norm": 1.8378593921661377, + "learning_rate": 6.007075386157435e-07, + "loss": 3.2277, + "step": 22261 + }, + { + "epoch": 1.7847507566494958, + "grad_norm": 1.711582064628601, + "learning_rate": 6.002643678633091e-07, + "loss": 3.1318, + "step": 22262 + }, + { + "epoch": 1.7848309314305184, + "grad_norm": 1.7088749408721924, + "learning_rate": 5.998213555871046e-07, + "loss": 3.2273, + "step": 22263 + }, + { + "epoch": 1.784911106211541, + "grad_norm": 1.8880579471588135, + "learning_rate": 5.993785017945997e-07, + "loss": 3.1934, + "step": 22264 + }, + { + "epoch": 1.7849912809925637, + "grad_norm": 1.7716740369796753, + "learning_rate": 5.989358064932616e-07, + "loss": 3.0906, + "step": 22265 + }, + { + "epoch": 1.7850714557735863, + "grad_norm": 1.727966547012329, + "learning_rate": 5.98493269690551e-07, + "loss": 3.1197, + "step": 22266 + }, + { + "epoch": 1.785151630554609, + "grad_norm": 1.7544825077056885, + "learning_rate": 5.980508913939298e-07, + "loss": 3.2629, + "step": 22267 + }, + { + "epoch": 1.7852318053356315, + "grad_norm": 1.7765406370162964, + "learning_rate": 5.976086716108575e-07, + "loss": 3.2481, + "step": 22268 + }, + { + "epoch": 1.7853119801166542, + "grad_norm": 1.8089044094085693, + "learning_rate": 5.971666103487894e-07, + "loss": 3.1277, + "step": 22269 + }, + { + "epoch": 1.7853921548976768, + "grad_norm": 1.6965982913970947, + "learning_rate": 5.967247076151794e-07, + "loss": 3.1675, + "step": 22270 + }, + { + "epoch": 1.7854723296786994, + "grad_norm": 1.8088526725769043, + "learning_rate": 5.96282963417475e-07, + "loss": 3.2724, + "step": 22271 + }, + { + "epoch": 1.7855525044597222, + "grad_norm": 1.7296974658966064, + "learning_rate": 5.958413777631266e-07, + "loss": 3.1614, + "step": 22272 + }, + { + "epoch": 1.7856326792407449, + "grad_norm": 1.7459653615951538, + "learning_rate": 5.953999506595775e-07, + "loss": 3.118, + "step": 22273 + }, + { + "epoch": 1.7857128540217675, + "grad_norm": 1.7608028650283813, + "learning_rate": 5.949586821142717e-07, + "loss": 3.0833, + "step": 22274 + }, + { + "epoch": 1.7857930288027901, + "grad_norm": 1.77200186252594, + "learning_rate": 5.945175721346486e-07, + "loss": 3.1934, + "step": 22275 + }, + { + "epoch": 1.7858732035838127, + "grad_norm": 1.8088587522506714, + "learning_rate": 5.940766207281423e-07, + "loss": 3.2568, + "step": 22276 + }, + { + "epoch": 1.7859533783648354, + "grad_norm": 1.9008787870407104, + "learning_rate": 5.936358279021903e-07, + "loss": 3.1208, + "step": 22277 + }, + { + "epoch": 1.786033553145858, + "grad_norm": 1.7321646213531494, + "learning_rate": 5.931951936642222e-07, + "loss": 3.1865, + "step": 22278 + }, + { + "epoch": 1.7861137279268806, + "grad_norm": 1.7329028844833374, + "learning_rate": 5.927547180216686e-07, + "loss": 3.187, + "step": 22279 + }, + { + "epoch": 1.7861939027079032, + "grad_norm": 1.6483627557754517, + "learning_rate": 5.923144009819559e-07, + "loss": 3.1786, + "step": 22280 + }, + { + "epoch": 1.7862740774889259, + "grad_norm": 1.8087371587753296, + "learning_rate": 5.918742425525059e-07, + "loss": 3.2166, + "step": 22281 + }, + { + "epoch": 1.7863542522699485, + "grad_norm": 1.8328845500946045, + "learning_rate": 5.914342427407405e-07, + "loss": 3.1315, + "step": 22282 + }, + { + "epoch": 1.786434427050971, + "grad_norm": 1.758054494857788, + "learning_rate": 5.909944015540781e-07, + "loss": 3.1208, + "step": 22283 + }, + { + "epoch": 1.7865146018319937, + "grad_norm": 1.678252935409546, + "learning_rate": 5.905547189999327e-07, + "loss": 3.2281, + "step": 22284 + }, + { + "epoch": 1.7865947766130164, + "grad_norm": 1.7535253763198853, + "learning_rate": 5.901151950857209e-07, + "loss": 3.2137, + "step": 22285 + }, + { + "epoch": 1.786674951394039, + "grad_norm": 0.654075026512146, + "learning_rate": 5.896758298188499e-07, + "loss": 2.4658, + "step": 22286 + }, + { + "epoch": 1.7867551261750616, + "grad_norm": 1.7535299062728882, + "learning_rate": 5.89236623206727e-07, + "loss": 3.1483, + "step": 22287 + }, + { + "epoch": 1.7868353009560842, + "grad_norm": 1.7940922975540161, + "learning_rate": 5.887975752567576e-07, + "loss": 3.1497, + "step": 22288 + }, + { + "epoch": 1.7869154757371069, + "grad_norm": 1.7921751737594604, + "learning_rate": 5.883586859763446e-07, + "loss": 3.243, + "step": 22289 + }, + { + "epoch": 1.7869956505181295, + "grad_norm": 1.8424508571624756, + "learning_rate": 5.879199553728887e-07, + "loss": 3.1683, + "step": 22290 + }, + { + "epoch": 1.787075825299152, + "grad_norm": 1.7514328956604004, + "learning_rate": 5.87481383453784e-07, + "loss": 3.142, + "step": 22291 + }, + { + "epoch": 1.787156000080175, + "grad_norm": 1.8254045248031616, + "learning_rate": 5.870429702264247e-07, + "loss": 3.2284, + "step": 22292 + }, + { + "epoch": 1.7872361748611976, + "grad_norm": 1.8726155757904053, + "learning_rate": 5.866047156982047e-07, + "loss": 3.2933, + "step": 22293 + }, + { + "epoch": 1.7873163496422202, + "grad_norm": 2.00838303565979, + "learning_rate": 5.861666198765114e-07, + "loss": 3.3188, + "step": 22294 + }, + { + "epoch": 1.7873965244232428, + "grad_norm": 1.8309104442596436, + "learning_rate": 5.857286827687325e-07, + "loss": 3.1716, + "step": 22295 + }, + { + "epoch": 1.7874766992042654, + "grad_norm": 1.8079733848571777, + "learning_rate": 5.852909043822486e-07, + "loss": 3.1688, + "step": 22296 + }, + { + "epoch": 1.787556873985288, + "grad_norm": 1.8626917600631714, + "learning_rate": 5.848532847244415e-07, + "loss": 3.1259, + "step": 22297 + }, + { + "epoch": 1.7876370487663107, + "grad_norm": 1.7923234701156616, + "learning_rate": 5.844158238026909e-07, + "loss": 3.1964, + "step": 22298 + }, + { + "epoch": 1.7877172235473333, + "grad_norm": 1.6287699937820435, + "learning_rate": 5.839785216243698e-07, + "loss": 3.1643, + "step": 22299 + }, + { + "epoch": 1.787797398328356, + "grad_norm": 1.7340819835662842, + "learning_rate": 5.835413781968547e-07, + "loss": 3.1361, + "step": 22300 + }, + { + "epoch": 1.7878775731093786, + "grad_norm": 1.7799078226089478, + "learning_rate": 5.831043935275116e-07, + "loss": 3.1499, + "step": 22301 + }, + { + "epoch": 1.7879577478904012, + "grad_norm": 1.9083260297775269, + "learning_rate": 5.826675676237092e-07, + "loss": 3.2111, + "step": 22302 + }, + { + "epoch": 1.7880379226714238, + "grad_norm": 1.7658392190933228, + "learning_rate": 5.822309004928139e-07, + "loss": 3.3232, + "step": 22303 + }, + { + "epoch": 1.7881180974524464, + "grad_norm": 1.7032902240753174, + "learning_rate": 5.817943921421854e-07, + "loss": 3.2274, + "step": 22304 + }, + { + "epoch": 1.788198272233469, + "grad_norm": 1.8042712211608887, + "learning_rate": 5.813580425791854e-07, + "loss": 3.1466, + "step": 22305 + }, + { + "epoch": 1.7882784470144917, + "grad_norm": 1.76926589012146, + "learning_rate": 5.809218518111681e-07, + "loss": 3.3082, + "step": 22306 + }, + { + "epoch": 1.7883586217955143, + "grad_norm": 1.8160971403121948, + "learning_rate": 5.804858198454888e-07, + "loss": 3.2606, + "step": 22307 + }, + { + "epoch": 1.788438796576537, + "grad_norm": 1.782368540763855, + "learning_rate": 5.800499466894993e-07, + "loss": 3.0887, + "step": 22308 + }, + { + "epoch": 1.7885189713575596, + "grad_norm": 1.8124147653579712, + "learning_rate": 5.796142323505471e-07, + "loss": 3.2208, + "step": 22309 + }, + { + "epoch": 1.7885991461385822, + "grad_norm": 1.7919840812683105, + "learning_rate": 5.791786768359808e-07, + "loss": 3.1376, + "step": 22310 + }, + { + "epoch": 1.7886793209196048, + "grad_norm": 1.8387794494628906, + "learning_rate": 5.787432801531401e-07, + "loss": 3.2517, + "step": 22311 + }, + { + "epoch": 1.7887594957006274, + "grad_norm": 1.7001593112945557, + "learning_rate": 5.783080423093679e-07, + "loss": 3.2241, + "step": 22312 + }, + { + "epoch": 1.78883967048165, + "grad_norm": 1.6669049263000488, + "learning_rate": 5.778729633120006e-07, + "loss": 3.0911, + "step": 22313 + }, + { + "epoch": 1.7889198452626727, + "grad_norm": 1.730209469795227, + "learning_rate": 5.774380431683757e-07, + "loss": 3.1309, + "step": 22314 + }, + { + "epoch": 1.7890000200436953, + "grad_norm": 1.7641865015029907, + "learning_rate": 5.770032818858251e-07, + "loss": 3.1979, + "step": 22315 + }, + { + "epoch": 1.789080194824718, + "grad_norm": 1.8229472637176514, + "learning_rate": 5.765686794716762e-07, + "loss": 3.2779, + "step": 22316 + }, + { + "epoch": 1.7891603696057405, + "grad_norm": 1.7655507326126099, + "learning_rate": 5.7613423593326e-07, + "loss": 3.2361, + "step": 22317 + }, + { + "epoch": 1.7892405443867632, + "grad_norm": 1.8241422176361084, + "learning_rate": 5.756999512778971e-07, + "loss": 3.2054, + "step": 22318 + }, + { + "epoch": 1.7893207191677858, + "grad_norm": 0.6566358804702759, + "learning_rate": 5.752658255129129e-07, + "loss": 2.4594, + "step": 22319 + }, + { + "epoch": 1.7894008939488084, + "grad_norm": 2.0047574043273926, + "learning_rate": 5.74831858645627e-07, + "loss": 3.1477, + "step": 22320 + }, + { + "epoch": 1.789481068729831, + "grad_norm": 1.7003862857818604, + "learning_rate": 5.743980506833524e-07, + "loss": 3.1913, + "step": 22321 + }, + { + "epoch": 1.7895612435108537, + "grad_norm": 1.7959301471710205, + "learning_rate": 5.739644016334045e-07, + "loss": 3.145, + "step": 22322 + }, + { + "epoch": 1.7896414182918763, + "grad_norm": 1.8341161012649536, + "learning_rate": 5.735309115030951e-07, + "loss": 3.3, + "step": 22323 + }, + { + "epoch": 1.789721593072899, + "grad_norm": 1.7670410871505737, + "learning_rate": 5.730975802997318e-07, + "loss": 3.2015, + "step": 22324 + }, + { + "epoch": 1.7898017678539215, + "grad_norm": 1.6993550062179565, + "learning_rate": 5.726644080306232e-07, + "loss": 3.1044, + "step": 22325 + }, + { + "epoch": 1.7898819426349442, + "grad_norm": 0.6693583726882935, + "learning_rate": 5.722313947030678e-07, + "loss": 2.4785, + "step": 22326 + }, + { + "epoch": 1.7899621174159668, + "grad_norm": 1.8700416088104248, + "learning_rate": 5.717985403243687e-07, + "loss": 3.1569, + "step": 22327 + }, + { + "epoch": 1.7900422921969894, + "grad_norm": 1.6647098064422607, + "learning_rate": 5.713658449018233e-07, + "loss": 3.1274, + "step": 22328 + }, + { + "epoch": 1.790122466978012, + "grad_norm": 1.7485599517822266, + "learning_rate": 5.709333084427271e-07, + "loss": 3.1702, + "step": 22329 + }, + { + "epoch": 1.7902026417590347, + "grad_norm": 1.8204060792922974, + "learning_rate": 5.705009309543741e-07, + "loss": 3.2131, + "step": 22330 + }, + { + "epoch": 1.7902828165400573, + "grad_norm": 1.796363115310669, + "learning_rate": 5.700687124440496e-07, + "loss": 3.2283, + "step": 22331 + }, + { + "epoch": 1.79036299132108, + "grad_norm": 1.818429708480835, + "learning_rate": 5.696366529190433e-07, + "loss": 3.2063, + "step": 22332 + }, + { + "epoch": 1.7904431661021025, + "grad_norm": 1.8453867435455322, + "learning_rate": 5.692047523866407e-07, + "loss": 3.1599, + "step": 22333 + }, + { + "epoch": 1.7905233408831251, + "grad_norm": 1.7271939516067505, + "learning_rate": 5.687730108541212e-07, + "loss": 3.1437, + "step": 22334 + }, + { + "epoch": 1.7906035156641478, + "grad_norm": 1.7818737030029297, + "learning_rate": 5.683414283287669e-07, + "loss": 3.1831, + "step": 22335 + }, + { + "epoch": 1.7906836904451704, + "grad_norm": 1.7818081378936768, + "learning_rate": 5.679100048178499e-07, + "loss": 3.1085, + "step": 22336 + }, + { + "epoch": 1.790763865226193, + "grad_norm": 1.7445993423461914, + "learning_rate": 5.674787403286464e-07, + "loss": 3.1761, + "step": 22337 + }, + { + "epoch": 1.7908440400072156, + "grad_norm": 1.716515302658081, + "learning_rate": 5.670476348684262e-07, + "loss": 3.1859, + "step": 22338 + }, + { + "epoch": 1.7909242147882383, + "grad_norm": 1.7939754724502563, + "learning_rate": 5.666166884444602e-07, + "loss": 3.072, + "step": 22339 + }, + { + "epoch": 1.791004389569261, + "grad_norm": 1.6965793371200562, + "learning_rate": 5.661859010640125e-07, + "loss": 3.2181, + "step": 22340 + }, + { + "epoch": 1.7910845643502835, + "grad_norm": 1.788692593574524, + "learning_rate": 5.65755272734344e-07, + "loss": 3.1813, + "step": 22341 + }, + { + "epoch": 1.7911647391313061, + "grad_norm": 1.7781699895858765, + "learning_rate": 5.653248034627168e-07, + "loss": 3.2615, + "step": 22342 + }, + { + "epoch": 1.7912449139123288, + "grad_norm": 1.830407738685608, + "learning_rate": 5.648944932563882e-07, + "loss": 3.2295, + "step": 22343 + }, + { + "epoch": 1.7913250886933514, + "grad_norm": 1.7153074741363525, + "learning_rate": 5.644643421226137e-07, + "loss": 3.057, + "step": 22344 + }, + { + "epoch": 1.791405263474374, + "grad_norm": 1.7373638153076172, + "learning_rate": 5.640343500686463e-07, + "loss": 3.1697, + "step": 22345 + }, + { + "epoch": 1.7914854382553966, + "grad_norm": 1.762205719947815, + "learning_rate": 5.636045171017323e-07, + "loss": 3.1842, + "step": 22346 + }, + { + "epoch": 1.7915656130364193, + "grad_norm": 0.6706719398498535, + "learning_rate": 5.631748432291207e-07, + "loss": 2.47, + "step": 22347 + }, + { + "epoch": 1.7916457878174419, + "grad_norm": 1.7285679578781128, + "learning_rate": 5.627453284580564e-07, + "loss": 3.1766, + "step": 22348 + }, + { + "epoch": 1.7917259625984645, + "grad_norm": 1.6995677947998047, + "learning_rate": 5.623159727957784e-07, + "loss": 3.1363, + "step": 22349 + }, + { + "epoch": 1.7918061373794871, + "grad_norm": 1.8149256706237793, + "learning_rate": 5.618867762495284e-07, + "loss": 3.2183, + "step": 22350 + }, + { + "epoch": 1.7918863121605098, + "grad_norm": 1.7327061891555786, + "learning_rate": 5.614577388265407e-07, + "loss": 3.1887, + "step": 22351 + }, + { + "epoch": 1.7919664869415324, + "grad_norm": 1.8427194356918335, + "learning_rate": 5.610288605340486e-07, + "loss": 3.2444, + "step": 22352 + }, + { + "epoch": 1.7920466617225552, + "grad_norm": 1.7461074590682983, + "learning_rate": 5.606001413792828e-07, + "loss": 3.1967, + "step": 22353 + }, + { + "epoch": 1.7921268365035778, + "grad_norm": 1.8518201112747192, + "learning_rate": 5.601715813694719e-07, + "loss": 3.186, + "step": 22354 + }, + { + "epoch": 1.7922070112846005, + "grad_norm": 1.9402247667312622, + "learning_rate": 5.597431805118436e-07, + "loss": 3.2784, + "step": 22355 + }, + { + "epoch": 1.792287186065623, + "grad_norm": 1.8958630561828613, + "learning_rate": 5.593149388136154e-07, + "loss": 3.2028, + "step": 22356 + }, + { + "epoch": 1.7923673608466457, + "grad_norm": 1.766992211341858, + "learning_rate": 5.588868562820116e-07, + "loss": 3.1741, + "step": 22357 + }, + { + "epoch": 1.7924475356276683, + "grad_norm": 1.708827018737793, + "learning_rate": 5.584589329242473e-07, + "loss": 3.0969, + "step": 22358 + }, + { + "epoch": 1.792527710408691, + "grad_norm": 1.7119966745376587, + "learning_rate": 5.58031168747537e-07, + "loss": 3.2308, + "step": 22359 + }, + { + "epoch": 1.7926078851897136, + "grad_norm": 0.6569008827209473, + "learning_rate": 5.57603563759096e-07, + "loss": 2.4838, + "step": 22360 + }, + { + "epoch": 1.7926880599707362, + "grad_norm": 1.8320289850234985, + "learning_rate": 5.571761179661295e-07, + "loss": 3.2031, + "step": 22361 + }, + { + "epoch": 1.7927682347517588, + "grad_norm": 1.7543926239013672, + "learning_rate": 5.567488313758462e-07, + "loss": 3.21, + "step": 22362 + }, + { + "epoch": 1.7928484095327815, + "grad_norm": 1.6589045524597168, + "learning_rate": 5.563217039954483e-07, + "loss": 3.1597, + "step": 22363 + }, + { + "epoch": 1.792928584313804, + "grad_norm": 1.7237428426742554, + "learning_rate": 5.558947358321387e-07, + "loss": 3.1041, + "step": 22364 + }, + { + "epoch": 1.7930087590948267, + "grad_norm": 1.8490601778030396, + "learning_rate": 5.554679268931174e-07, + "loss": 3.2781, + "step": 22365 + }, + { + "epoch": 1.7930889338758493, + "grad_norm": 1.7817931175231934, + "learning_rate": 5.550412771855762e-07, + "loss": 2.9895, + "step": 22366 + }, + { + "epoch": 1.793169108656872, + "grad_norm": 1.7250210046768188, + "learning_rate": 5.546147867167107e-07, + "loss": 3.1368, + "step": 22367 + }, + { + "epoch": 1.7932492834378946, + "grad_norm": 1.6931488513946533, + "learning_rate": 5.541884554937105e-07, + "loss": 3.1859, + "step": 22368 + }, + { + "epoch": 1.7933294582189172, + "grad_norm": 1.7488223314285278, + "learning_rate": 5.537622835237644e-07, + "loss": 3.1483, + "step": 22369 + }, + { + "epoch": 1.7934096329999398, + "grad_norm": 1.8468077182769775, + "learning_rate": 5.533362708140588e-07, + "loss": 3.234, + "step": 22370 + }, + { + "epoch": 1.7934898077809625, + "grad_norm": 1.692030906677246, + "learning_rate": 5.529104173717725e-07, + "loss": 3.0137, + "step": 22371 + }, + { + "epoch": 1.793569982561985, + "grad_norm": 1.8191739320755005, + "learning_rate": 5.524847232040864e-07, + "loss": 3.0796, + "step": 22372 + }, + { + "epoch": 1.793650157343008, + "grad_norm": 1.7055715322494507, + "learning_rate": 5.520591883181792e-07, + "loss": 3.1068, + "step": 22373 + }, + { + "epoch": 1.7937303321240305, + "grad_norm": 1.6728581190109253, + "learning_rate": 5.51633812721224e-07, + "loss": 3.2017, + "step": 22374 + }, + { + "epoch": 1.7938105069050532, + "grad_norm": 1.797702670097351, + "learning_rate": 5.51208596420394e-07, + "loss": 3.116, + "step": 22375 + }, + { + "epoch": 1.7938906816860758, + "grad_norm": 1.7170079946517944, + "learning_rate": 5.507835394228566e-07, + "loss": 3.1108, + "step": 22376 + }, + { + "epoch": 1.7939708564670984, + "grad_norm": 1.8565070629119873, + "learning_rate": 5.503586417357776e-07, + "loss": 3.1031, + "step": 22377 + }, + { + "epoch": 1.794051031248121, + "grad_norm": 1.7816956043243408, + "learning_rate": 5.499339033663209e-07, + "loss": 3.187, + "step": 22378 + }, + { + "epoch": 1.7941312060291437, + "grad_norm": 1.8304505348205566, + "learning_rate": 5.495093243216487e-07, + "loss": 3.0533, + "step": 22379 + }, + { + "epoch": 1.7942113808101663, + "grad_norm": 1.6927624940872192, + "learning_rate": 5.490849046089208e-07, + "loss": 3.1282, + "step": 22380 + }, + { + "epoch": 1.794291555591189, + "grad_norm": 1.741969108581543, + "learning_rate": 5.486606442352882e-07, + "loss": 3.1034, + "step": 22381 + }, + { + "epoch": 1.7943717303722115, + "grad_norm": 1.7870696783065796, + "learning_rate": 5.482365432079062e-07, + "loss": 3.089, + "step": 22382 + }, + { + "epoch": 1.7944519051532342, + "grad_norm": 1.7903510332107544, + "learning_rate": 5.478126015339258e-07, + "loss": 3.2269, + "step": 22383 + }, + { + "epoch": 1.7945320799342568, + "grad_norm": 1.7966746091842651, + "learning_rate": 5.473888192204923e-07, + "loss": 3.2021, + "step": 22384 + }, + { + "epoch": 1.7946122547152794, + "grad_norm": 1.7506362199783325, + "learning_rate": 5.469651962747546e-07, + "loss": 3.1244, + "step": 22385 + }, + { + "epoch": 1.794692429496302, + "grad_norm": 1.7395482063293457, + "learning_rate": 5.465417327038502e-07, + "loss": 3.144, + "step": 22386 + }, + { + "epoch": 1.7947726042773247, + "grad_norm": 1.792020559310913, + "learning_rate": 5.461184285149201e-07, + "loss": 3.1479, + "step": 22387 + }, + { + "epoch": 1.7948527790583473, + "grad_norm": 1.851200819015503, + "learning_rate": 5.456952837151019e-07, + "loss": 3.1814, + "step": 22388 + }, + { + "epoch": 1.79493295383937, + "grad_norm": 1.6814202070236206, + "learning_rate": 5.452722983115289e-07, + "loss": 3.1509, + "step": 22389 + }, + { + "epoch": 1.7950131286203925, + "grad_norm": 1.8148764371871948, + "learning_rate": 5.44849472311334e-07, + "loss": 3.1873, + "step": 22390 + }, + { + "epoch": 1.7950933034014152, + "grad_norm": 1.750409722328186, + "learning_rate": 5.44426805721644e-07, + "loss": 3.1895, + "step": 22391 + }, + { + "epoch": 1.7951734781824378, + "grad_norm": 1.6329691410064697, + "learning_rate": 5.440042985495853e-07, + "loss": 3.2311, + "step": 22392 + }, + { + "epoch": 1.7952536529634604, + "grad_norm": 1.7052322626113892, + "learning_rate": 5.435819508022811e-07, + "loss": 3.1554, + "step": 22393 + }, + { + "epoch": 1.795333827744483, + "grad_norm": 1.8068877458572388, + "learning_rate": 5.431597624868523e-07, + "loss": 3.2064, + "step": 22394 + }, + { + "epoch": 1.7954140025255056, + "grad_norm": 1.7650226354599, + "learning_rate": 5.427377336104189e-07, + "loss": 3.2018, + "step": 22395 + }, + { + "epoch": 1.7954941773065283, + "grad_norm": 0.6628796458244324, + "learning_rate": 5.42315864180093e-07, + "loss": 2.4647, + "step": 22396 + }, + { + "epoch": 1.795574352087551, + "grad_norm": 1.7157241106033325, + "learning_rate": 5.418941542029887e-07, + "loss": 3.1769, + "step": 22397 + }, + { + "epoch": 1.7956545268685735, + "grad_norm": 1.7158406972885132, + "learning_rate": 5.41472603686215e-07, + "loss": 3.0988, + "step": 22398 + }, + { + "epoch": 1.7957347016495961, + "grad_norm": 1.6805788278579712, + "learning_rate": 5.410512126368795e-07, + "loss": 3.1055, + "step": 22399 + }, + { + "epoch": 1.7958148764306188, + "grad_norm": 1.7360951900482178, + "learning_rate": 5.406299810620885e-07, + "loss": 3.1193, + "step": 22400 + }, + { + "epoch": 1.7958950512116414, + "grad_norm": 1.653564453125, + "learning_rate": 5.402089089689411e-07, + "loss": 3.0911, + "step": 22401 + }, + { + "epoch": 1.795975225992664, + "grad_norm": 1.797966718673706, + "learning_rate": 5.397879963645369e-07, + "loss": 3.1926, + "step": 22402 + }, + { + "epoch": 1.7960554007736866, + "grad_norm": 1.7500971555709839, + "learning_rate": 5.393672432559738e-07, + "loss": 3.103, + "step": 22403 + }, + { + "epoch": 1.7961355755547093, + "grad_norm": 1.7098602056503296, + "learning_rate": 5.389466496503449e-07, + "loss": 3.0457, + "step": 22404 + }, + { + "epoch": 1.7962157503357319, + "grad_norm": 1.7865045070648193, + "learning_rate": 5.385262155547411e-07, + "loss": 3.1429, + "step": 22405 + }, + { + "epoch": 1.7962959251167545, + "grad_norm": 1.742056131362915, + "learning_rate": 5.381059409762501e-07, + "loss": 3.0882, + "step": 22406 + }, + { + "epoch": 1.7963760998977771, + "grad_norm": 1.7416431903839111, + "learning_rate": 5.376858259219586e-07, + "loss": 3.0977, + "step": 22407 + }, + { + "epoch": 1.7964562746787998, + "grad_norm": 1.9136240482330322, + "learning_rate": 5.372658703989486e-07, + "loss": 3.1238, + "step": 22408 + }, + { + "epoch": 1.7965364494598224, + "grad_norm": 1.833959698677063, + "learning_rate": 5.368460744143023e-07, + "loss": 3.1897, + "step": 22409 + }, + { + "epoch": 1.796616624240845, + "grad_norm": 1.9407116174697876, + "learning_rate": 5.364264379750961e-07, + "loss": 3.2655, + "step": 22410 + }, + { + "epoch": 1.7966967990218676, + "grad_norm": 1.8070216178894043, + "learning_rate": 5.360069610884033e-07, + "loss": 3.2908, + "step": 22411 + }, + { + "epoch": 1.7967769738028903, + "grad_norm": 1.7806504964828491, + "learning_rate": 5.355876437612984e-07, + "loss": 3.1273, + "step": 22412 + }, + { + "epoch": 1.7968571485839129, + "grad_norm": 1.7738279104232788, + "learning_rate": 5.351684860008511e-07, + "loss": 3.1864, + "step": 22413 + }, + { + "epoch": 1.7969373233649355, + "grad_norm": 1.8221633434295654, + "learning_rate": 5.347494878141257e-07, + "loss": 3.2147, + "step": 22414 + }, + { + "epoch": 1.7970174981459581, + "grad_norm": 1.7678385972976685, + "learning_rate": 5.34330649208189e-07, + "loss": 3.1353, + "step": 22415 + }, + { + "epoch": 1.7970976729269807, + "grad_norm": 1.7133524417877197, + "learning_rate": 5.33911970190103e-07, + "loss": 3.1588, + "step": 22416 + }, + { + "epoch": 1.7971778477080034, + "grad_norm": 1.773358702659607, + "learning_rate": 5.334934507669231e-07, + "loss": 3.2038, + "step": 22417 + }, + { + "epoch": 1.797258022489026, + "grad_norm": 1.776823878288269, + "learning_rate": 5.330750909457072e-07, + "loss": 3.1549, + "step": 22418 + }, + { + "epoch": 1.7973381972700486, + "grad_norm": 1.7926372289657593, + "learning_rate": 5.326568907335095e-07, + "loss": 3.1231, + "step": 22419 + }, + { + "epoch": 1.7974183720510712, + "grad_norm": 1.9293769598007202, + "learning_rate": 5.322388501373798e-07, + "loss": 3.0806, + "step": 22420 + }, + { + "epoch": 1.7974985468320939, + "grad_norm": 1.7573410272598267, + "learning_rate": 5.318209691643672e-07, + "loss": 3.1008, + "step": 22421 + }, + { + "epoch": 1.7975787216131165, + "grad_norm": 1.7732969522476196, + "learning_rate": 5.314032478215159e-07, + "loss": 3.1771, + "step": 22422 + }, + { + "epoch": 1.7976588963941391, + "grad_norm": 0.6621696949005127, + "learning_rate": 5.30985686115868e-07, + "loss": 2.4811, + "step": 22423 + }, + { + "epoch": 1.7977390711751617, + "grad_norm": 2.0292625427246094, + "learning_rate": 5.305682840544635e-07, + "loss": 3.2218, + "step": 22424 + }, + { + "epoch": 1.7978192459561844, + "grad_norm": 1.7903478145599365, + "learning_rate": 5.301510416443411e-07, + "loss": 3.2996, + "step": 22425 + }, + { + "epoch": 1.797899420737207, + "grad_norm": 1.7294466495513916, + "learning_rate": 5.297339588925365e-07, + "loss": 3.0933, + "step": 22426 + }, + { + "epoch": 1.7979795955182296, + "grad_norm": 1.7277830839157104, + "learning_rate": 5.293170358060773e-07, + "loss": 3.1465, + "step": 22427 + }, + { + "epoch": 1.7980597702992522, + "grad_norm": 1.766318678855896, + "learning_rate": 5.289002723919956e-07, + "loss": 3.2255, + "step": 22428 + }, + { + "epoch": 1.7981399450802749, + "grad_norm": 1.733080267906189, + "learning_rate": 5.28483668657318e-07, + "loss": 3.1556, + "step": 22429 + }, + { + "epoch": 1.7982201198612975, + "grad_norm": 1.7779101133346558, + "learning_rate": 5.280672246090667e-07, + "loss": 3.1221, + "step": 22430 + }, + { + "epoch": 1.79830029464232, + "grad_norm": 1.9570008516311646, + "learning_rate": 5.276509402542662e-07, + "loss": 3.2013, + "step": 22431 + }, + { + "epoch": 1.7983804694233427, + "grad_norm": 1.9108492136001587, + "learning_rate": 5.272348155999296e-07, + "loss": 3.2012, + "step": 22432 + }, + { + "epoch": 1.7984606442043654, + "grad_norm": 1.7996890544891357, + "learning_rate": 5.268188506530769e-07, + "loss": 3.0217, + "step": 22433 + }, + { + "epoch": 1.7985408189853882, + "grad_norm": 1.7796576023101807, + "learning_rate": 5.264030454207191e-07, + "loss": 3.2437, + "step": 22434 + }, + { + "epoch": 1.7986209937664108, + "grad_norm": 1.6859973669052124, + "learning_rate": 5.259873999098675e-07, + "loss": 3.1221, + "step": 22435 + }, + { + "epoch": 1.7987011685474334, + "grad_norm": 1.7699363231658936, + "learning_rate": 5.255719141275295e-07, + "loss": 3.1063, + "step": 22436 + }, + { + "epoch": 1.798781343328456, + "grad_norm": 1.7851123809814453, + "learning_rate": 5.251565880807097e-07, + "loss": 3.1764, + "step": 22437 + }, + { + "epoch": 1.7988615181094787, + "grad_norm": 1.834561824798584, + "learning_rate": 5.247414217764101e-07, + "loss": 3.0757, + "step": 22438 + }, + { + "epoch": 1.7989416928905013, + "grad_norm": 1.6744356155395508, + "learning_rate": 5.243264152216309e-07, + "loss": 3.1868, + "step": 22439 + }, + { + "epoch": 1.799021867671524, + "grad_norm": 1.6606858968734741, + "learning_rate": 5.239115684233687e-07, + "loss": 3.1599, + "step": 22440 + }, + { + "epoch": 1.7991020424525466, + "grad_norm": 1.759918212890625, + "learning_rate": 5.234968813886198e-07, + "loss": 3.2998, + "step": 22441 + }, + { + "epoch": 1.7991822172335692, + "grad_norm": 0.6504904627799988, + "learning_rate": 5.230823541243713e-07, + "loss": 2.4603, + "step": 22442 + }, + { + "epoch": 1.7992623920145918, + "grad_norm": 1.7685351371765137, + "learning_rate": 5.22667986637615e-07, + "loss": 3.1515, + "step": 22443 + }, + { + "epoch": 1.7993425667956144, + "grad_norm": 1.7636767625808716, + "learning_rate": 5.222537789353355e-07, + "loss": 3.1389, + "step": 22444 + }, + { + "epoch": 1.799422741576637, + "grad_norm": 1.6676397323608398, + "learning_rate": 5.218397310245183e-07, + "loss": 2.9833, + "step": 22445 + }, + { + "epoch": 1.7995029163576597, + "grad_norm": 1.7327263355255127, + "learning_rate": 5.214258429121432e-07, + "loss": 3.0964, + "step": 22446 + }, + { + "epoch": 1.7995830911386823, + "grad_norm": 1.80758535861969, + "learning_rate": 5.210121146051861e-07, + "loss": 3.1254, + "step": 22447 + }, + { + "epoch": 1.799663265919705, + "grad_norm": 1.770804762840271, + "learning_rate": 5.205985461106255e-07, + "loss": 3.2034, + "step": 22448 + }, + { + "epoch": 1.7997434407007276, + "grad_norm": 1.7155884504318237, + "learning_rate": 5.201851374354317e-07, + "loss": 3.1198, + "step": 22449 + }, + { + "epoch": 1.7998236154817502, + "grad_norm": 1.7581610679626465, + "learning_rate": 5.197718885865743e-07, + "loss": 3.1304, + "step": 22450 + }, + { + "epoch": 1.7999037902627728, + "grad_norm": 1.64752197265625, + "learning_rate": 5.193587995710247e-07, + "loss": 3.158, + "step": 22451 + }, + { + "epoch": 1.7999839650437954, + "grad_norm": 1.7968077659606934, + "learning_rate": 5.189458703957429e-07, + "loss": 3.1884, + "step": 22452 + }, + { + "epoch": 1.800064139824818, + "grad_norm": 1.8019261360168457, + "learning_rate": 5.185331010676919e-07, + "loss": 3.072, + "step": 22453 + }, + { + "epoch": 1.8001443146058407, + "grad_norm": 1.8020323514938354, + "learning_rate": 5.181204915938309e-07, + "loss": 3.196, + "step": 22454 + }, + { + "epoch": 1.8002244893868635, + "grad_norm": 1.7287917137145996, + "learning_rate": 5.177080419811176e-07, + "loss": 3.0637, + "step": 22455 + }, + { + "epoch": 1.8003046641678861, + "grad_norm": 1.9827580451965332, + "learning_rate": 5.172957522365052e-07, + "loss": 3.1243, + "step": 22456 + }, + { + "epoch": 1.8003848389489088, + "grad_norm": 1.741438388824463, + "learning_rate": 5.168836223669437e-07, + "loss": 3.19, + "step": 22457 + }, + { + "epoch": 1.8004650137299314, + "grad_norm": 1.7816718816757202, + "learning_rate": 5.164716523793811e-07, + "loss": 3.1286, + "step": 22458 + }, + { + "epoch": 1.800545188510954, + "grad_norm": 1.7911224365234375, + "learning_rate": 5.16059842280765e-07, + "loss": 3.1928, + "step": 22459 + }, + { + "epoch": 1.8006253632919766, + "grad_norm": 1.7661021947860718, + "learning_rate": 5.156481920780377e-07, + "loss": 3.1541, + "step": 22460 + }, + { + "epoch": 1.8007055380729993, + "grad_norm": 1.7468527555465698, + "learning_rate": 5.152367017781402e-07, + "loss": 3.2295, + "step": 22461 + }, + { + "epoch": 1.800785712854022, + "grad_norm": 1.8495562076568604, + "learning_rate": 5.148253713880081e-07, + "loss": 3.1459, + "step": 22462 + }, + { + "epoch": 1.8008658876350445, + "grad_norm": 1.7522481679916382, + "learning_rate": 5.14414200914577e-07, + "loss": 3.129, + "step": 22463 + }, + { + "epoch": 1.8009460624160671, + "grad_norm": 1.7887283563613892, + "learning_rate": 5.140031903647801e-07, + "loss": 3.1492, + "step": 22464 + }, + { + "epoch": 1.8010262371970898, + "grad_norm": 1.6997328996658325, + "learning_rate": 5.135923397455456e-07, + "loss": 3.1914, + "step": 22465 + }, + { + "epoch": 1.8011064119781124, + "grad_norm": 0.6601444482803345, + "learning_rate": 5.13181649063802e-07, + "loss": 2.4589, + "step": 22466 + }, + { + "epoch": 1.801186586759135, + "grad_norm": 1.8222979307174683, + "learning_rate": 5.127711183264716e-07, + "loss": 3.2958, + "step": 22467 + }, + { + "epoch": 1.8012667615401576, + "grad_norm": 1.799872875213623, + "learning_rate": 5.123607475404757e-07, + "loss": 3.1475, + "step": 22468 + }, + { + "epoch": 1.8013469363211803, + "grad_norm": 1.8130460977554321, + "learning_rate": 5.119505367127353e-07, + "loss": 3.1183, + "step": 22469 + }, + { + "epoch": 1.8014271111022029, + "grad_norm": 1.831749439239502, + "learning_rate": 5.115404858501638e-07, + "loss": 3.1577, + "step": 22470 + }, + { + "epoch": 1.8015072858832255, + "grad_norm": 1.7634583711624146, + "learning_rate": 5.111305949596768e-07, + "loss": 3.2416, + "step": 22471 + }, + { + "epoch": 1.8015874606642481, + "grad_norm": 1.7542085647583008, + "learning_rate": 5.107208640481831e-07, + "loss": 3.1239, + "step": 22472 + }, + { + "epoch": 1.8016676354452708, + "grad_norm": 1.7713457345962524, + "learning_rate": 5.103112931225907e-07, + "loss": 3.1964, + "step": 22473 + }, + { + "epoch": 1.8017478102262934, + "grad_norm": 1.8406708240509033, + "learning_rate": 5.09901882189806e-07, + "loss": 3.1112, + "step": 22474 + }, + { + "epoch": 1.801827985007316, + "grad_norm": 1.7757985591888428, + "learning_rate": 5.094926312567305e-07, + "loss": 3.1342, + "step": 22475 + }, + { + "epoch": 1.8019081597883386, + "grad_norm": 1.7404372692108154, + "learning_rate": 5.090835403302663e-07, + "loss": 3.1475, + "step": 22476 + }, + { + "epoch": 1.8019883345693612, + "grad_norm": 1.6933238506317139, + "learning_rate": 5.086746094173068e-07, + "loss": 3.1481, + "step": 22477 + }, + { + "epoch": 1.8020685093503839, + "grad_norm": 1.7924357652664185, + "learning_rate": 5.082658385247486e-07, + "loss": 3.1781, + "step": 22478 + }, + { + "epoch": 1.8021486841314065, + "grad_norm": 1.728014349937439, + "learning_rate": 5.078572276594829e-07, + "loss": 3.1617, + "step": 22479 + }, + { + "epoch": 1.8022288589124291, + "grad_norm": 1.7914923429489136, + "learning_rate": 5.074487768283997e-07, + "loss": 3.1566, + "step": 22480 + }, + { + "epoch": 1.8023090336934517, + "grad_norm": 1.6746203899383545, + "learning_rate": 5.070404860383849e-07, + "loss": 3.2246, + "step": 22481 + }, + { + "epoch": 1.8023892084744744, + "grad_norm": 1.814098596572876, + "learning_rate": 5.066323552963215e-07, + "loss": 3.2319, + "step": 22482 + }, + { + "epoch": 1.802469383255497, + "grad_norm": 0.669674813747406, + "learning_rate": 5.062243846090909e-07, + "loss": 2.4769, + "step": 22483 + }, + { + "epoch": 1.8025495580365196, + "grad_norm": 1.8044601678848267, + "learning_rate": 5.058165739835697e-07, + "loss": 3.209, + "step": 22484 + }, + { + "epoch": 1.8026297328175422, + "grad_norm": 1.7607269287109375, + "learning_rate": 5.054089234266369e-07, + "loss": 3.2268, + "step": 22485 + }, + { + "epoch": 1.8027099075985649, + "grad_norm": 1.9795527458190918, + "learning_rate": 5.050014329451636e-07, + "loss": 3.1721, + "step": 22486 + }, + { + "epoch": 1.8027900823795875, + "grad_norm": 1.8142693042755127, + "learning_rate": 5.045941025460188e-07, + "loss": 3.1543, + "step": 22487 + }, + { + "epoch": 1.8028702571606101, + "grad_norm": 2.0654492378234863, + "learning_rate": 5.041869322360704e-07, + "loss": 3.1887, + "step": 22488 + }, + { + "epoch": 1.8029504319416327, + "grad_norm": 1.743293046951294, + "learning_rate": 5.03779922022184e-07, + "loss": 3.109, + "step": 22489 + }, + { + "epoch": 1.8030306067226554, + "grad_norm": 1.9467220306396484, + "learning_rate": 5.033730719112218e-07, + "loss": 3.2363, + "step": 22490 + }, + { + "epoch": 1.803110781503678, + "grad_norm": 1.7201796770095825, + "learning_rate": 5.029663819100438e-07, + "loss": 3.168, + "step": 22491 + }, + { + "epoch": 1.8031909562847006, + "grad_norm": 1.7629289627075195, + "learning_rate": 5.025598520255038e-07, + "loss": 3.0873, + "step": 22492 + }, + { + "epoch": 1.8032711310657232, + "grad_norm": 1.8393265008926392, + "learning_rate": 5.02153482264458e-07, + "loss": 3.1704, + "step": 22493 + }, + { + "epoch": 1.8033513058467459, + "grad_norm": 1.764784812927246, + "learning_rate": 5.017472726337569e-07, + "loss": 3.2299, + "step": 22494 + }, + { + "epoch": 1.8034314806277685, + "grad_norm": 1.7715446949005127, + "learning_rate": 5.013412231402493e-07, + "loss": 3.248, + "step": 22495 + }, + { + "epoch": 1.803511655408791, + "grad_norm": 1.8856972455978394, + "learning_rate": 5.009353337907818e-07, + "loss": 3.2255, + "step": 22496 + }, + { + "epoch": 1.8035918301898137, + "grad_norm": 1.7863720655441284, + "learning_rate": 5.005296045921959e-07, + "loss": 3.1469, + "step": 22497 + }, + { + "epoch": 1.8036720049708364, + "grad_norm": 1.7809696197509766, + "learning_rate": 5.001240355513326e-07, + "loss": 3.2987, + "step": 22498 + }, + { + "epoch": 1.803752179751859, + "grad_norm": 1.79301917552948, + "learning_rate": 4.997186266750309e-07, + "loss": 3.1901, + "step": 22499 + }, + { + "epoch": 1.8038323545328816, + "grad_norm": 1.8145493268966675, + "learning_rate": 4.993133779701243e-07, + "loss": 3.2023, + "step": 22500 + }, + { + "epoch": 1.8039125293139042, + "grad_norm": 1.7623769044876099, + "learning_rate": 4.989082894434472e-07, + "loss": 3.1331, + "step": 22501 + }, + { + "epoch": 1.8039927040949268, + "grad_norm": 1.737986445426941, + "learning_rate": 4.985033611018264e-07, + "loss": 3.1858, + "step": 22502 + }, + { + "epoch": 1.8040728788759495, + "grad_norm": 1.7166210412979126, + "learning_rate": 4.980985929520898e-07, + "loss": 3.1851, + "step": 22503 + }, + { + "epoch": 1.804153053656972, + "grad_norm": 1.7500278949737549, + "learning_rate": 4.976939850010631e-07, + "loss": 3.1096, + "step": 22504 + }, + { + "epoch": 1.8042332284379947, + "grad_norm": 1.9199446439743042, + "learning_rate": 4.972895372555664e-07, + "loss": 3.1337, + "step": 22505 + }, + { + "epoch": 1.8043134032190173, + "grad_norm": 1.9191484451293945, + "learning_rate": 4.968852497224197e-07, + "loss": 3.2369, + "step": 22506 + }, + { + "epoch": 1.80439357800004, + "grad_norm": 1.7332338094711304, + "learning_rate": 4.964811224084387e-07, + "loss": 3.1949, + "step": 22507 + }, + { + "epoch": 1.8044737527810626, + "grad_norm": 1.7839463949203491, + "learning_rate": 4.960771553204358e-07, + "loss": 3.123, + "step": 22508 + }, + { + "epoch": 1.8045539275620852, + "grad_norm": 1.7306032180786133, + "learning_rate": 4.956733484652221e-07, + "loss": 3.2202, + "step": 22509 + }, + { + "epoch": 1.8046341023431078, + "grad_norm": 0.6693871021270752, + "learning_rate": 4.952697018496067e-07, + "loss": 2.4756, + "step": 22510 + }, + { + "epoch": 1.8047142771241305, + "grad_norm": 1.924179196357727, + "learning_rate": 4.948662154803951e-07, + "loss": 3.2117, + "step": 22511 + }, + { + "epoch": 1.804794451905153, + "grad_norm": 1.7746797800064087, + "learning_rate": 4.944628893643887e-07, + "loss": 3.2636, + "step": 22512 + }, + { + "epoch": 1.8048746266861757, + "grad_norm": 1.7551060914993286, + "learning_rate": 4.940597235083866e-07, + "loss": 3.2194, + "step": 22513 + }, + { + "epoch": 1.8049548014671983, + "grad_norm": 1.810368299484253, + "learning_rate": 4.936567179191887e-07, + "loss": 3.08, + "step": 22514 + }, + { + "epoch": 1.805034976248221, + "grad_norm": 1.8006333112716675, + "learning_rate": 4.932538726035874e-07, + "loss": 3.3295, + "step": 22515 + }, + { + "epoch": 1.8051151510292438, + "grad_norm": 0.6401997804641724, + "learning_rate": 4.928511875683773e-07, + "loss": 2.4719, + "step": 22516 + }, + { + "epoch": 1.8051953258102664, + "grad_norm": 1.867071270942688, + "learning_rate": 4.924486628203429e-07, + "loss": 3.2197, + "step": 22517 + }, + { + "epoch": 1.805275500591289, + "grad_norm": 1.7312663793563843, + "learning_rate": 4.920462983662743e-07, + "loss": 3.1781, + "step": 22518 + }, + { + "epoch": 1.8053556753723117, + "grad_norm": 1.895332932472229, + "learning_rate": 4.916440942129541e-07, + "loss": 3.1786, + "step": 22519 + }, + { + "epoch": 1.8054358501533343, + "grad_norm": 1.8021996021270752, + "learning_rate": 4.912420503671633e-07, + "loss": 3.0917, + "step": 22520 + }, + { + "epoch": 1.805516024934357, + "grad_norm": 1.7382493019104004, + "learning_rate": 4.908401668356821e-07, + "loss": 3.1993, + "step": 22521 + }, + { + "epoch": 1.8055961997153795, + "grad_norm": 1.7829170227050781, + "learning_rate": 4.904384436252818e-07, + "loss": 3.2788, + "step": 22522 + }, + { + "epoch": 1.8056763744964022, + "grad_norm": 1.812748908996582, + "learning_rate": 4.900368807427392e-07, + "loss": 3.1631, + "step": 22523 + }, + { + "epoch": 1.8057565492774248, + "grad_norm": 1.747999906539917, + "learning_rate": 4.89635478194822e-07, + "loss": 3.1263, + "step": 22524 + }, + { + "epoch": 1.8058367240584474, + "grad_norm": 1.7313274145126343, + "learning_rate": 4.892342359882995e-07, + "loss": 3.1132, + "step": 22525 + }, + { + "epoch": 1.80591689883947, + "grad_norm": 1.8033332824707031, + "learning_rate": 4.888331541299374e-07, + "loss": 3.2731, + "step": 22526 + }, + { + "epoch": 1.8059970736204927, + "grad_norm": 1.7967725992202759, + "learning_rate": 4.884322326264946e-07, + "loss": 3.1611, + "step": 22527 + }, + { + "epoch": 1.8060772484015153, + "grad_norm": 1.7190344333648682, + "learning_rate": 4.880314714847323e-07, + "loss": 3.2625, + "step": 22528 + }, + { + "epoch": 1.806157423182538, + "grad_norm": 1.8204782009124756, + "learning_rate": 4.876308707114063e-07, + "loss": 3.2046, + "step": 22529 + }, + { + "epoch": 1.8062375979635605, + "grad_norm": 1.781122088432312, + "learning_rate": 4.872304303132724e-07, + "loss": 3.1305, + "step": 22530 + }, + { + "epoch": 1.8063177727445832, + "grad_norm": 1.7285304069519043, + "learning_rate": 4.868301502970818e-07, + "loss": 3.2534, + "step": 22531 + }, + { + "epoch": 1.8063979475256058, + "grad_norm": 1.762603521347046, + "learning_rate": 4.864300306695812e-07, + "loss": 3.2258, + "step": 22532 + }, + { + "epoch": 1.8064781223066284, + "grad_norm": 1.8460874557495117, + "learning_rate": 4.860300714375166e-07, + "loss": 3.1986, + "step": 22533 + }, + { + "epoch": 1.806558297087651, + "grad_norm": 1.7437688112258911, + "learning_rate": 4.856302726076323e-07, + "loss": 3.1603, + "step": 22534 + }, + { + "epoch": 1.8066384718686737, + "grad_norm": 1.801153540611267, + "learning_rate": 4.852306341866686e-07, + "loss": 3.0888, + "step": 22535 + }, + { + "epoch": 1.8067186466496965, + "grad_norm": 1.72432279586792, + "learning_rate": 4.848311561813635e-07, + "loss": 3.2227, + "step": 22536 + }, + { + "epoch": 1.8067988214307191, + "grad_norm": 1.7914113998413086, + "learning_rate": 4.844318385984515e-07, + "loss": 3.2662, + "step": 22537 + }, + { + "epoch": 1.8068789962117418, + "grad_norm": 1.7211190462112427, + "learning_rate": 4.840326814446638e-07, + "loss": 3.0435, + "step": 22538 + }, + { + "epoch": 1.8069591709927644, + "grad_norm": 1.712868094444275, + "learning_rate": 4.83633684726732e-07, + "loss": 3.2059, + "step": 22539 + }, + { + "epoch": 1.807039345773787, + "grad_norm": 2.0328547954559326, + "learning_rate": 4.832348484513827e-07, + "loss": 3.3164, + "step": 22540 + }, + { + "epoch": 1.8071195205548096, + "grad_norm": 0.670761227607727, + "learning_rate": 4.828361726253406e-07, + "loss": 2.455, + "step": 22541 + }, + { + "epoch": 1.8071996953358322, + "grad_norm": 1.8218454122543335, + "learning_rate": 4.824376572553257e-07, + "loss": 3.158, + "step": 22542 + }, + { + "epoch": 1.8072798701168549, + "grad_norm": 0.6931374669075012, + "learning_rate": 4.820393023480574e-07, + "loss": 2.4724, + "step": 22543 + }, + { + "epoch": 1.8073600448978775, + "grad_norm": 1.7827951908111572, + "learning_rate": 4.816411079102523e-07, + "loss": 3.2323, + "step": 22544 + }, + { + "epoch": 1.8074402196789001, + "grad_norm": 1.9127684831619263, + "learning_rate": 4.812430739486229e-07, + "loss": 3.1245, + "step": 22545 + }, + { + "epoch": 1.8075203944599227, + "grad_norm": 1.9379551410675049, + "learning_rate": 4.808452004698827e-07, + "loss": 3.2012, + "step": 22546 + }, + { + "epoch": 1.8076005692409454, + "grad_norm": 1.6528141498565674, + "learning_rate": 4.804474874807352e-07, + "loss": 3.1304, + "step": 22547 + }, + { + "epoch": 1.807680744021968, + "grad_norm": 0.6602364778518677, + "learning_rate": 4.800499349878884e-07, + "loss": 2.4561, + "step": 22548 + }, + { + "epoch": 1.8077609188029906, + "grad_norm": 1.7428406476974487, + "learning_rate": 4.796525429980448e-07, + "loss": 3.136, + "step": 22549 + }, + { + "epoch": 1.8078410935840132, + "grad_norm": 1.9306327104568481, + "learning_rate": 4.792553115179033e-07, + "loss": 3.1768, + "step": 22550 + }, + { + "epoch": 1.8079212683650359, + "grad_norm": 1.763082504272461, + "learning_rate": 4.788582405541642e-07, + "loss": 3.2676, + "step": 22551 + }, + { + "epoch": 1.8080014431460585, + "grad_norm": 1.7539093494415283, + "learning_rate": 4.784613301135177e-07, + "loss": 3.1656, + "step": 22552 + }, + { + "epoch": 1.808081617927081, + "grad_norm": 1.7480318546295166, + "learning_rate": 4.780645802026574e-07, + "loss": 3.1254, + "step": 22553 + }, + { + "epoch": 1.8081617927081037, + "grad_norm": 1.8224430084228516, + "learning_rate": 4.776679908282721e-07, + "loss": 3.2062, + "step": 22554 + }, + { + "epoch": 1.8082419674891264, + "grad_norm": 1.7438230514526367, + "learning_rate": 4.772715619970492e-07, + "loss": 3.118, + "step": 22555 + }, + { + "epoch": 1.808322142270149, + "grad_norm": 1.8105974197387695, + "learning_rate": 4.768752937156728e-07, + "loss": 3.1948, + "step": 22556 + }, + { + "epoch": 1.8084023170511716, + "grad_norm": 1.7747005224227905, + "learning_rate": 4.7647918599082134e-07, + "loss": 3.2245, + "step": 22557 + }, + { + "epoch": 1.8084824918321942, + "grad_norm": 1.734218716621399, + "learning_rate": 4.7608323882917363e-07, + "loss": 3.0809, + "step": 22558 + }, + { + "epoch": 1.8085626666132169, + "grad_norm": 1.8686925172805786, + "learning_rate": 4.756874522374066e-07, + "loss": 3.2174, + "step": 22559 + }, + { + "epoch": 1.8086428413942395, + "grad_norm": 1.9063197374343872, + "learning_rate": 4.752918262221928e-07, + "loss": 3.372, + "step": 22560 + }, + { + "epoch": 1.808723016175262, + "grad_norm": 1.798863172531128, + "learning_rate": 4.748963607902035e-07, + "loss": 3.1445, + "step": 22561 + }, + { + "epoch": 1.8088031909562847, + "grad_norm": 1.8118904829025269, + "learning_rate": 4.745010559481023e-07, + "loss": 3.1733, + "step": 22562 + }, + { + "epoch": 1.8088833657373073, + "grad_norm": 1.7430005073547363, + "learning_rate": 4.74105911702557e-07, + "loss": 3.2157, + "step": 22563 + }, + { + "epoch": 1.80896354051833, + "grad_norm": 1.7332669496536255, + "learning_rate": 4.737109280602281e-07, + "loss": 3.0725, + "step": 22564 + }, + { + "epoch": 1.8090437152993526, + "grad_norm": 1.7941906452178955, + "learning_rate": 4.733161050277757e-07, + "loss": 3.2344, + "step": 22565 + }, + { + "epoch": 1.8091238900803752, + "grad_norm": 1.8207982778549194, + "learning_rate": 4.729214426118578e-07, + "loss": 3.2522, + "step": 22566 + }, + { + "epoch": 1.8092040648613978, + "grad_norm": 1.7964754104614258, + "learning_rate": 4.7252694081912467e-07, + "loss": 3.2491, + "step": 22567 + }, + { + "epoch": 1.8092842396424205, + "grad_norm": 1.7318601608276367, + "learning_rate": 4.721325996562309e-07, + "loss": 3.1746, + "step": 22568 + }, + { + "epoch": 1.809364414423443, + "grad_norm": 1.7953630685806274, + "learning_rate": 4.7173841912982246e-07, + "loss": 3.1604, + "step": 22569 + }, + { + "epoch": 1.8094445892044657, + "grad_norm": 1.7538481950759888, + "learning_rate": 4.7134439924654605e-07, + "loss": 3.2387, + "step": 22570 + }, + { + "epoch": 1.8095247639854883, + "grad_norm": 1.8292008638381958, + "learning_rate": 4.709505400130465e-07, + "loss": 3.1741, + "step": 22571 + }, + { + "epoch": 1.809604938766511, + "grad_norm": 0.6549168229103088, + "learning_rate": 4.705568414359618e-07, + "loss": 2.4689, + "step": 22572 + }, + { + "epoch": 1.8096851135475336, + "grad_norm": 1.8773096799850464, + "learning_rate": 4.7016330352192994e-07, + "loss": 3.2552, + "step": 22573 + }, + { + "epoch": 1.8097652883285562, + "grad_norm": 1.753811240196228, + "learning_rate": 4.6976992627758567e-07, + "loss": 3.0924, + "step": 22574 + }, + { + "epoch": 1.8098454631095788, + "grad_norm": 1.7214007377624512, + "learning_rate": 4.693767097095614e-07, + "loss": 3.1749, + "step": 22575 + }, + { + "epoch": 1.8099256378906015, + "grad_norm": 1.7298659086227417, + "learning_rate": 4.689836538244896e-07, + "loss": 3.2222, + "step": 22576 + }, + { + "epoch": 1.810005812671624, + "grad_norm": 1.8429114818572998, + "learning_rate": 4.685907586289917e-07, + "loss": 3.2345, + "step": 22577 + }, + { + "epoch": 1.8100859874526467, + "grad_norm": 1.7269799709320068, + "learning_rate": 4.681980241296946e-07, + "loss": 3.1169, + "step": 22578 + }, + { + "epoch": 1.8101661622336693, + "grad_norm": 1.8192297220230103, + "learning_rate": 4.6780545033321965e-07, + "loss": 3.3059, + "step": 22579 + }, + { + "epoch": 1.810246337014692, + "grad_norm": 0.6537273526191711, + "learning_rate": 4.6741303724618603e-07, + "loss": 2.467, + "step": 22580 + }, + { + "epoch": 1.8103265117957146, + "grad_norm": 1.755530595779419, + "learning_rate": 4.6702078487520955e-07, + "loss": 3.2024, + "step": 22581 + }, + { + "epoch": 1.8104066865767372, + "grad_norm": 1.8789560794830322, + "learning_rate": 4.666286932269015e-07, + "loss": 3.2141, + "step": 22582 + }, + { + "epoch": 1.8104868613577598, + "grad_norm": 1.8623981475830078, + "learning_rate": 4.6623676230787454e-07, + "loss": 3.1965, + "step": 22583 + }, + { + "epoch": 1.8105670361387824, + "grad_norm": 1.927795648574829, + "learning_rate": 4.6584499212473433e-07, + "loss": 3.2034, + "step": 22584 + }, + { + "epoch": 1.810647210919805, + "grad_norm": 1.7211941480636597, + "learning_rate": 4.65453382684089e-07, + "loss": 3.1091, + "step": 22585 + }, + { + "epoch": 1.8107273857008277, + "grad_norm": 1.7831511497497559, + "learning_rate": 4.6506193399253995e-07, + "loss": 3.2045, + "step": 22586 + }, + { + "epoch": 1.8108075604818503, + "grad_norm": 1.915683627128601, + "learning_rate": 4.646706460566841e-07, + "loss": 3.1388, + "step": 22587 + }, + { + "epoch": 1.810887735262873, + "grad_norm": 1.7255876064300537, + "learning_rate": 4.642795188831217e-07, + "loss": 3.1013, + "step": 22588 + }, + { + "epoch": 1.8109679100438956, + "grad_norm": 1.9368069171905518, + "learning_rate": 4.6388855247844534e-07, + "loss": 3.2035, + "step": 22589 + }, + { + "epoch": 1.8110480848249182, + "grad_norm": 1.7821309566497803, + "learning_rate": 4.634977468492474e-07, + "loss": 3.2211, + "step": 22590 + }, + { + "epoch": 1.8111282596059408, + "grad_norm": 1.800714135169983, + "learning_rate": 4.6310710200211826e-07, + "loss": 3.1581, + "step": 22591 + }, + { + "epoch": 1.8112084343869634, + "grad_norm": 1.753058910369873, + "learning_rate": 4.627166179436404e-07, + "loss": 3.1528, + "step": 22592 + }, + { + "epoch": 1.811288609167986, + "grad_norm": 1.7704750299453735, + "learning_rate": 4.623262946803997e-07, + "loss": 3.2034, + "step": 22593 + }, + { + "epoch": 1.8113687839490087, + "grad_norm": 1.7548961639404297, + "learning_rate": 4.6193613221897524e-07, + "loss": 3.1855, + "step": 22594 + }, + { + "epoch": 1.8114489587300313, + "grad_norm": 1.7587506771087646, + "learning_rate": 4.615461305659463e-07, + "loss": 3.2332, + "step": 22595 + }, + { + "epoch": 1.811529133511054, + "grad_norm": 1.7039259672164917, + "learning_rate": 4.6115628972788984e-07, + "loss": 3.1635, + "step": 22596 + }, + { + "epoch": 1.8116093082920768, + "grad_norm": 1.8157778978347778, + "learning_rate": 4.60766609711375e-07, + "loss": 3.2377, + "step": 22597 + }, + { + "epoch": 1.8116894830730994, + "grad_norm": 1.7369781732559204, + "learning_rate": 4.603770905229732e-07, + "loss": 3.1659, + "step": 22598 + }, + { + "epoch": 1.811769657854122, + "grad_norm": 1.7458888292312622, + "learning_rate": 4.599877321692503e-07, + "loss": 3.1481, + "step": 22599 + }, + { + "epoch": 1.8118498326351447, + "grad_norm": 1.78700852394104, + "learning_rate": 4.595985346567733e-07, + "loss": 3.2225, + "step": 22600 + }, + { + "epoch": 1.8119300074161673, + "grad_norm": 1.6748120784759521, + "learning_rate": 4.5920949799210356e-07, + "loss": 3.1099, + "step": 22601 + }, + { + "epoch": 1.81201018219719, + "grad_norm": 1.8129055500030518, + "learning_rate": 4.588206221817981e-07, + "loss": 3.1868, + "step": 22602 + }, + { + "epoch": 1.8120903569782125, + "grad_norm": 1.7835354804992676, + "learning_rate": 4.5843190723241503e-07, + "loss": 3.2244, + "step": 22603 + }, + { + "epoch": 1.8121705317592351, + "grad_norm": 1.7374539375305176, + "learning_rate": 4.5804335315050577e-07, + "loss": 3.1465, + "step": 22604 + }, + { + "epoch": 1.8122507065402578, + "grad_norm": 1.7468394041061401, + "learning_rate": 4.5765495994262386e-07, + "loss": 3.1639, + "step": 22605 + }, + { + "epoch": 1.8123308813212804, + "grad_norm": 1.8207899332046509, + "learning_rate": 4.5726672761531645e-07, + "loss": 3.1254, + "step": 22606 + }, + { + "epoch": 1.812411056102303, + "grad_norm": 1.7067115306854248, + "learning_rate": 4.568786561751282e-07, + "loss": 3.2359, + "step": 22607 + }, + { + "epoch": 1.8124912308833256, + "grad_norm": 1.8475003242492676, + "learning_rate": 4.564907456286027e-07, + "loss": 3.1627, + "step": 22608 + }, + { + "epoch": 1.8125714056643483, + "grad_norm": 1.8950088024139404, + "learning_rate": 4.5610299598227936e-07, + "loss": 3.0917, + "step": 22609 + }, + { + "epoch": 1.812651580445371, + "grad_norm": 1.8500100374221802, + "learning_rate": 4.5571540724269614e-07, + "loss": 3.0948, + "step": 22610 + }, + { + "epoch": 1.8127317552263935, + "grad_norm": 1.9942538738250732, + "learning_rate": 4.5532797941638897e-07, + "loss": 3.188, + "step": 22611 + }, + { + "epoch": 1.8128119300074161, + "grad_norm": 1.6997241973876953, + "learning_rate": 4.549407125098859e-07, + "loss": 3.0458, + "step": 22612 + }, + { + "epoch": 1.8128921047884388, + "grad_norm": 1.6749646663665771, + "learning_rate": 4.545536065297196e-07, + "loss": 3.1553, + "step": 22613 + }, + { + "epoch": 1.8129722795694614, + "grad_norm": 1.7967156171798706, + "learning_rate": 4.541666614824147e-07, + "loss": 3.2319, + "step": 22614 + }, + { + "epoch": 1.813052454350484, + "grad_norm": 1.8010611534118652, + "learning_rate": 4.5377987737449614e-07, + "loss": 3.125, + "step": 22615 + }, + { + "epoch": 1.8131326291315066, + "grad_norm": 1.8510156869888306, + "learning_rate": 4.5339325421248527e-07, + "loss": 3.1522, + "step": 22616 + }, + { + "epoch": 1.8132128039125293, + "grad_norm": 1.6695703268051147, + "learning_rate": 4.53006792002898e-07, + "loss": 3.2017, + "step": 22617 + }, + { + "epoch": 1.813292978693552, + "grad_norm": 1.8585211038589478, + "learning_rate": 4.5262049075225135e-07, + "loss": 3.1766, + "step": 22618 + }, + { + "epoch": 1.8133731534745747, + "grad_norm": 1.7852344512939453, + "learning_rate": 4.5223435046705897e-07, + "loss": 3.1093, + "step": 22619 + }, + { + "epoch": 1.8134533282555974, + "grad_norm": 1.7864751815795898, + "learning_rate": 4.518483711538313e-07, + "loss": 3.2593, + "step": 22620 + }, + { + "epoch": 1.81353350303662, + "grad_norm": 1.9431374073028564, + "learning_rate": 4.5146255281907524e-07, + "loss": 3.1291, + "step": 22621 + }, + { + "epoch": 1.8136136778176426, + "grad_norm": 1.8930686712265015, + "learning_rate": 4.5107689546929347e-07, + "loss": 3.2758, + "step": 22622 + }, + { + "epoch": 1.8136938525986652, + "grad_norm": 1.8546524047851562, + "learning_rate": 4.506913991109907e-07, + "loss": 3.1367, + "step": 22623 + }, + { + "epoch": 1.8137740273796878, + "grad_norm": 1.840215802192688, + "learning_rate": 4.503060637506662e-07, + "loss": 3.2478, + "step": 22624 + }, + { + "epoch": 1.8138542021607105, + "grad_norm": 1.7841436862945557, + "learning_rate": 4.499208893948148e-07, + "loss": 3.1337, + "step": 22625 + }, + { + "epoch": 1.813934376941733, + "grad_norm": 1.7716692686080933, + "learning_rate": 4.4953587604993244e-07, + "loss": 3.2103, + "step": 22626 + }, + { + "epoch": 1.8140145517227557, + "grad_norm": 1.740038275718689, + "learning_rate": 4.491510237225083e-07, + "loss": 3.181, + "step": 22627 + }, + { + "epoch": 1.8140947265037783, + "grad_norm": 1.9053974151611328, + "learning_rate": 4.4876633241903275e-07, + "loss": 3.1022, + "step": 22628 + }, + { + "epoch": 1.814174901284801, + "grad_norm": 0.6720634698867798, + "learning_rate": 4.483818021459896e-07, + "loss": 2.4579, + "step": 22629 + }, + { + "epoch": 1.8142550760658236, + "grad_norm": 1.7917234897613525, + "learning_rate": 4.479974329098635e-07, + "loss": 3.1757, + "step": 22630 + }, + { + "epoch": 1.8143352508468462, + "grad_norm": 1.7913845777511597, + "learning_rate": 4.4761322471713496e-07, + "loss": 3.1412, + "step": 22631 + }, + { + "epoch": 1.8144154256278688, + "grad_norm": 1.7971714735031128, + "learning_rate": 4.4722917757427986e-07, + "loss": 3.2063, + "step": 22632 + }, + { + "epoch": 1.8144956004088915, + "grad_norm": 1.6936490535736084, + "learning_rate": 4.468452914877741e-07, + "loss": 3.1754, + "step": 22633 + }, + { + "epoch": 1.814575775189914, + "grad_norm": 1.721946120262146, + "learning_rate": 4.464615664640903e-07, + "loss": 3.113, + "step": 22634 + }, + { + "epoch": 1.8146559499709367, + "grad_norm": 1.8482534885406494, + "learning_rate": 4.4607800250969666e-07, + "loss": 3.2352, + "step": 22635 + }, + { + "epoch": 1.8147361247519593, + "grad_norm": 1.8022500276565552, + "learning_rate": 4.4569459963106244e-07, + "loss": 3.2159, + "step": 22636 + }, + { + "epoch": 1.814816299532982, + "grad_norm": 1.7195969820022583, + "learning_rate": 4.4531135783464907e-07, + "loss": 3.1524, + "step": 22637 + }, + { + "epoch": 1.8148964743140046, + "grad_norm": 1.8647959232330322, + "learning_rate": 4.449282771269181e-07, + "loss": 3.2008, + "step": 22638 + }, + { + "epoch": 1.8149766490950272, + "grad_norm": 1.8030356168746948, + "learning_rate": 4.4454535751432993e-07, + "loss": 3.0809, + "step": 22639 + }, + { + "epoch": 1.8150568238760498, + "grad_norm": 1.8514142036437988, + "learning_rate": 4.4416259900333824e-07, + "loss": 3.2584, + "step": 22640 + }, + { + "epoch": 1.8151369986570725, + "grad_norm": 1.8177056312561035, + "learning_rate": 4.43780001600399e-07, + "loss": 3.1827, + "step": 22641 + }, + { + "epoch": 1.815217173438095, + "grad_norm": 1.8263592720031738, + "learning_rate": 4.4339756531196046e-07, + "loss": 3.2411, + "step": 22642 + }, + { + "epoch": 1.8152973482191177, + "grad_norm": 1.7536414861679077, + "learning_rate": 4.4301529014446954e-07, + "loss": 3.1534, + "step": 22643 + }, + { + "epoch": 1.8153775230001403, + "grad_norm": 1.7833930253982544, + "learning_rate": 4.426331761043734e-07, + "loss": 3.2627, + "step": 22644 + }, + { + "epoch": 1.815457697781163, + "grad_norm": 1.7496196031570435, + "learning_rate": 4.422512231981135e-07, + "loss": 3.1096, + "step": 22645 + }, + { + "epoch": 1.8155378725621856, + "grad_norm": 1.759178638458252, + "learning_rate": 4.4186943143213035e-07, + "loss": 3.2281, + "step": 22646 + }, + { + "epoch": 1.8156180473432082, + "grad_norm": 1.8602616786956787, + "learning_rate": 4.414878008128598e-07, + "loss": 3.1388, + "step": 22647 + }, + { + "epoch": 1.8156982221242308, + "grad_norm": 1.8326003551483154, + "learning_rate": 4.411063313467356e-07, + "loss": 3.1788, + "step": 22648 + }, + { + "epoch": 1.8157783969052534, + "grad_norm": 1.678971529006958, + "learning_rate": 4.407250230401894e-07, + "loss": 3.1399, + "step": 22649 + }, + { + "epoch": 1.815858571686276, + "grad_norm": 1.6662346124649048, + "learning_rate": 4.4034387589965033e-07, + "loss": 3.1344, + "step": 22650 + }, + { + "epoch": 1.8159387464672987, + "grad_norm": 1.8624166250228882, + "learning_rate": 4.399628899315456e-07, + "loss": 3.2296, + "step": 22651 + }, + { + "epoch": 1.8160189212483213, + "grad_norm": 1.8810813426971436, + "learning_rate": 4.395820651422966e-07, + "loss": 3.1113, + "step": 22652 + }, + { + "epoch": 1.816099096029344, + "grad_norm": 1.812574863433838, + "learning_rate": 4.392014015383239e-07, + "loss": 3.1274, + "step": 22653 + }, + { + "epoch": 1.8161792708103666, + "grad_norm": 1.8626515865325928, + "learning_rate": 4.3882089912604675e-07, + "loss": 3.2013, + "step": 22654 + }, + { + "epoch": 1.8162594455913892, + "grad_norm": 1.744403600692749, + "learning_rate": 4.384405579118789e-07, + "loss": 3.1592, + "step": 22655 + }, + { + "epoch": 1.8163396203724118, + "grad_norm": 1.8279961347579956, + "learning_rate": 4.3806037790223523e-07, + "loss": 3.1721, + "step": 22656 + }, + { + "epoch": 1.8164197951534344, + "grad_norm": 1.8787827491760254, + "learning_rate": 4.376803591035217e-07, + "loss": 3.2035, + "step": 22657 + }, + { + "epoch": 1.816499969934457, + "grad_norm": 1.8526244163513184, + "learning_rate": 4.3730050152214763e-07, + "loss": 3.1654, + "step": 22658 + }, + { + "epoch": 1.8165801447154797, + "grad_norm": 1.7570854425430298, + "learning_rate": 4.3692080516451686e-07, + "loss": 3.1005, + "step": 22659 + }, + { + "epoch": 1.8166603194965023, + "grad_norm": 1.759199619293213, + "learning_rate": 4.365412700370308e-07, + "loss": 3.1012, + "step": 22660 + }, + { + "epoch": 1.816740494277525, + "grad_norm": 1.708603024482727, + "learning_rate": 4.3616189614609005e-07, + "loss": 3.1833, + "step": 22661 + }, + { + "epoch": 1.8168206690585476, + "grad_norm": 1.8197683095932007, + "learning_rate": 4.357826834980872e-07, + "loss": 3.1102, + "step": 22662 + }, + { + "epoch": 1.8169008438395702, + "grad_norm": 0.6416076421737671, + "learning_rate": 4.3540363209941815e-07, + "loss": 2.4675, + "step": 22663 + }, + { + "epoch": 1.8169810186205928, + "grad_norm": 1.7602168321609497, + "learning_rate": 4.350247419564724e-07, + "loss": 3.2358, + "step": 22664 + }, + { + "epoch": 1.8170611934016154, + "grad_norm": 1.7611702680587769, + "learning_rate": 4.346460130756391e-07, + "loss": 3.1829, + "step": 22665 + }, + { + "epoch": 1.817141368182638, + "grad_norm": 1.8774582147598267, + "learning_rate": 4.3426744546330445e-07, + "loss": 3.1348, + "step": 22666 + }, + { + "epoch": 1.8172215429636607, + "grad_norm": 1.885610580444336, + "learning_rate": 4.338890391258477e-07, + "loss": 3.2635, + "step": 22667 + }, + { + "epoch": 1.8173017177446833, + "grad_norm": 1.8556013107299805, + "learning_rate": 4.3351079406964923e-07, + "loss": 3.2491, + "step": 22668 + }, + { + "epoch": 1.817381892525706, + "grad_norm": 1.7572638988494873, + "learning_rate": 4.331327103010885e-07, + "loss": 3.1725, + "step": 22669 + }, + { + "epoch": 1.8174620673067285, + "grad_norm": 1.699608325958252, + "learning_rate": 4.3275478782653815e-07, + "loss": 3.0909, + "step": 22670 + }, + { + "epoch": 1.8175422420877512, + "grad_norm": 1.8029032945632935, + "learning_rate": 4.3237702665237193e-07, + "loss": 3.1823, + "step": 22671 + }, + { + "epoch": 1.8176224168687738, + "grad_norm": 1.7983659505844116, + "learning_rate": 4.319994267849559e-07, + "loss": 3.1917, + "step": 22672 + }, + { + "epoch": 1.8177025916497964, + "grad_norm": 1.777422308921814, + "learning_rate": 4.3162198823065714e-07, + "loss": 3.1247, + "step": 22673 + }, + { + "epoch": 1.817782766430819, + "grad_norm": 1.8417645692825317, + "learning_rate": 4.312447109958384e-07, + "loss": 3.2343, + "step": 22674 + }, + { + "epoch": 1.8178629412118417, + "grad_norm": 0.6475798487663269, + "learning_rate": 4.308675950868624e-07, + "loss": 2.4644, + "step": 22675 + }, + { + "epoch": 1.8179431159928643, + "grad_norm": 1.6480811834335327, + "learning_rate": 4.3049064051008835e-07, + "loss": 3.0153, + "step": 22676 + }, + { + "epoch": 1.818023290773887, + "grad_norm": 1.674526333808899, + "learning_rate": 4.301138472718669e-07, + "loss": 3.1724, + "step": 22677 + }, + { + "epoch": 1.8181034655549095, + "grad_norm": 1.7906289100646973, + "learning_rate": 4.2973721537855394e-07, + "loss": 3.1652, + "step": 22678 + }, + { + "epoch": 1.8181836403359324, + "grad_norm": 1.747822880744934, + "learning_rate": 4.293607448364989e-07, + "loss": 3.1809, + "step": 22679 + }, + { + "epoch": 1.818263815116955, + "grad_norm": 1.700282335281372, + "learning_rate": 4.2898443565204783e-07, + "loss": 3.1442, + "step": 22680 + }, + { + "epoch": 1.8183439898979776, + "grad_norm": 1.8144475221633911, + "learning_rate": 4.2860828783154787e-07, + "loss": 3.1411, + "step": 22681 + }, + { + "epoch": 1.8184241646790003, + "grad_norm": 1.7077720165252686, + "learning_rate": 4.282323013813372e-07, + "loss": 3.2523, + "step": 22682 + }, + { + "epoch": 1.8185043394600229, + "grad_norm": 1.8296124935150146, + "learning_rate": 4.278564763077564e-07, + "loss": 3.1404, + "step": 22683 + }, + { + "epoch": 1.8185845142410455, + "grad_norm": 1.7517434358596802, + "learning_rate": 4.274808126171426e-07, + "loss": 3.2771, + "step": 22684 + }, + { + "epoch": 1.8186646890220681, + "grad_norm": 1.7741174697875977, + "learning_rate": 4.2710531031582846e-07, + "loss": 3.2189, + "step": 22685 + }, + { + "epoch": 1.8187448638030907, + "grad_norm": 1.9283726215362549, + "learning_rate": 4.2672996941014564e-07, + "loss": 3.2166, + "step": 22686 + }, + { + "epoch": 1.8188250385841134, + "grad_norm": 1.8024723529815674, + "learning_rate": 4.263547899064202e-07, + "loss": 3.1486, + "step": 22687 + }, + { + "epoch": 1.818905213365136, + "grad_norm": 1.8485522270202637, + "learning_rate": 4.259797718109804e-07, + "loss": 3.1112, + "step": 22688 + }, + { + "epoch": 1.8189853881461586, + "grad_norm": 1.7886054515838623, + "learning_rate": 4.256049151301456e-07, + "loss": 3.182, + "step": 22689 + }, + { + "epoch": 1.8190655629271812, + "grad_norm": 1.6831752061843872, + "learning_rate": 4.252302198702385e-07, + "loss": 3.1918, + "step": 22690 + }, + { + "epoch": 1.8191457377082039, + "grad_norm": 1.7687979936599731, + "learning_rate": 4.248556860375763e-07, + "loss": 3.1539, + "step": 22691 + }, + { + "epoch": 1.8192259124892265, + "grad_norm": 1.8459302186965942, + "learning_rate": 4.244813136384718e-07, + "loss": 3.1223, + "step": 22692 + }, + { + "epoch": 1.8193060872702491, + "grad_norm": 1.6713870763778687, + "learning_rate": 4.241071026792376e-07, + "loss": 3.1349, + "step": 22693 + }, + { + "epoch": 1.8193862620512717, + "grad_norm": 0.6608567833900452, + "learning_rate": 4.2373305316618316e-07, + "loss": 2.4548, + "step": 22694 + }, + { + "epoch": 1.8194664368322944, + "grad_norm": 1.7904716730117798, + "learning_rate": 4.2335916510561346e-07, + "loss": 3.1776, + "step": 22695 + }, + { + "epoch": 1.819546611613317, + "grad_norm": 1.7905628681182861, + "learning_rate": 4.2298543850383455e-07, + "loss": 3.1229, + "step": 22696 + }, + { + "epoch": 1.8196267863943396, + "grad_norm": 1.756814956665039, + "learning_rate": 4.226118733671458e-07, + "loss": 3.1362, + "step": 22697 + }, + { + "epoch": 1.8197069611753622, + "grad_norm": 1.7695001363754272, + "learning_rate": 4.2223846970184447e-07, + "loss": 3.1952, + "step": 22698 + }, + { + "epoch": 1.819787135956385, + "grad_norm": 1.697547197341919, + "learning_rate": 4.2186522751422656e-07, + "loss": 3.1982, + "step": 22699 + }, + { + "epoch": 1.8198673107374077, + "grad_norm": 1.7519922256469727, + "learning_rate": 4.214921468105859e-07, + "loss": 3.0807, + "step": 22700 + }, + { + "epoch": 1.8199474855184303, + "grad_norm": 1.6430872678756714, + "learning_rate": 4.2111922759721313e-07, + "loss": 3.033, + "step": 22701 + }, + { + "epoch": 1.820027660299453, + "grad_norm": 1.8977102041244507, + "learning_rate": 4.2074646988039314e-07, + "loss": 3.1719, + "step": 22702 + }, + { + "epoch": 1.8201078350804756, + "grad_norm": 1.7826765775680542, + "learning_rate": 4.203738736664109e-07, + "loss": 3.1021, + "step": 22703 + }, + { + "epoch": 1.8201880098614982, + "grad_norm": 1.8464452028274536, + "learning_rate": 4.2000143896154924e-07, + "loss": 3.1409, + "step": 22704 + }, + { + "epoch": 1.8202681846425208, + "grad_norm": 1.8265032768249512, + "learning_rate": 4.1962916577208747e-07, + "loss": 3.2557, + "step": 22705 + }, + { + "epoch": 1.8203483594235434, + "grad_norm": 1.8190652132034302, + "learning_rate": 4.1925705410430286e-07, + "loss": 3.1753, + "step": 22706 + }, + { + "epoch": 1.820428534204566, + "grad_norm": 1.714314341545105, + "learning_rate": 4.188851039644659e-07, + "loss": 3.1602, + "step": 22707 + }, + { + "epoch": 1.8205087089855887, + "grad_norm": 1.7662533521652222, + "learning_rate": 4.1851331535884944e-07, + "loss": 3.1918, + "step": 22708 + }, + { + "epoch": 1.8205888837666113, + "grad_norm": 1.7855920791625977, + "learning_rate": 4.181416882937217e-07, + "loss": 3.22, + "step": 22709 + }, + { + "epoch": 1.820669058547634, + "grad_norm": 1.770602822303772, + "learning_rate": 4.1777022277534884e-07, + "loss": 3.1035, + "step": 22710 + }, + { + "epoch": 1.8207492333286566, + "grad_norm": 1.7755056619644165, + "learning_rate": 4.1739891880999364e-07, + "loss": 3.1574, + "step": 22711 + }, + { + "epoch": 1.8208294081096792, + "grad_norm": 1.8080523014068604, + "learning_rate": 4.170277764039132e-07, + "loss": 3.2418, + "step": 22712 + }, + { + "epoch": 1.8209095828907018, + "grad_norm": 1.8182495832443237, + "learning_rate": 4.166567955633682e-07, + "loss": 3.183, + "step": 22713 + }, + { + "epoch": 1.8209897576717244, + "grad_norm": 1.6999530792236328, + "learning_rate": 4.1628597629461144e-07, + "loss": 3.191, + "step": 22714 + }, + { + "epoch": 1.821069932452747, + "grad_norm": 1.742425560951233, + "learning_rate": 4.159153186038967e-07, + "loss": 3.2266, + "step": 22715 + }, + { + "epoch": 1.8211501072337697, + "grad_norm": 1.7416722774505615, + "learning_rate": 4.155448224974723e-07, + "loss": 3.1375, + "step": 22716 + }, + { + "epoch": 1.8212302820147923, + "grad_norm": 1.7658216953277588, + "learning_rate": 4.1517448798158334e-07, + "loss": 3.2244, + "step": 22717 + }, + { + "epoch": 1.821310456795815, + "grad_norm": 1.7934527397155762, + "learning_rate": 4.148043150624737e-07, + "loss": 3.1907, + "step": 22718 + }, + { + "epoch": 1.8213906315768376, + "grad_norm": 1.7749451398849487, + "learning_rate": 4.1443430374638604e-07, + "loss": 3.1508, + "step": 22719 + }, + { + "epoch": 1.8214708063578602, + "grad_norm": 1.7426362037658691, + "learning_rate": 4.1406445403955776e-07, + "loss": 3.2369, + "step": 22720 + }, + { + "epoch": 1.8215509811388828, + "grad_norm": 1.7036257982254028, + "learning_rate": 4.1369476594822487e-07, + "loss": 3.1312, + "step": 22721 + }, + { + "epoch": 1.8216311559199054, + "grad_norm": 1.7563538551330566, + "learning_rate": 4.1332523947861914e-07, + "loss": 3.0503, + "step": 22722 + }, + { + "epoch": 1.821711330700928, + "grad_norm": 1.7863584756851196, + "learning_rate": 4.129558746369711e-07, + "loss": 3.2152, + "step": 22723 + }, + { + "epoch": 1.8217915054819507, + "grad_norm": 1.7527437210083008, + "learning_rate": 4.125866714295079e-07, + "loss": 3.1894, + "step": 22724 + }, + { + "epoch": 1.8218716802629733, + "grad_norm": 1.7539879083633423, + "learning_rate": 4.122176298624547e-07, + "loss": 3.1312, + "step": 22725 + }, + { + "epoch": 1.821951855043996, + "grad_norm": 1.7182434797286987, + "learning_rate": 4.118487499420343e-07, + "loss": 3.1022, + "step": 22726 + }, + { + "epoch": 1.8220320298250186, + "grad_norm": 1.8084105253219604, + "learning_rate": 4.1148003167446384e-07, + "loss": 3.1494, + "step": 22727 + }, + { + "epoch": 1.8221122046060412, + "grad_norm": 1.8068116903305054, + "learning_rate": 4.111114750659606e-07, + "loss": 3.1929, + "step": 22728 + }, + { + "epoch": 1.8221923793870638, + "grad_norm": 0.6718275547027588, + "learning_rate": 4.1074308012273855e-07, + "loss": 2.4613, + "step": 22729 + }, + { + "epoch": 1.8222725541680864, + "grad_norm": 1.81505286693573, + "learning_rate": 4.103748468510083e-07, + "loss": 3.1407, + "step": 22730 + }, + { + "epoch": 1.822352728949109, + "grad_norm": 1.7029187679290771, + "learning_rate": 4.1000677525697916e-07, + "loss": 3.1205, + "step": 22731 + }, + { + "epoch": 1.8224329037301317, + "grad_norm": 1.817428469657898, + "learning_rate": 4.0963886534685525e-07, + "loss": 3.202, + "step": 22732 + }, + { + "epoch": 1.8225130785111543, + "grad_norm": 1.829918622970581, + "learning_rate": 4.092711171268404e-07, + "loss": 3.156, + "step": 22733 + }, + { + "epoch": 1.822593253292177, + "grad_norm": 1.7322719097137451, + "learning_rate": 4.0890353060313416e-07, + "loss": 3.0984, + "step": 22734 + }, + { + "epoch": 1.8226734280731995, + "grad_norm": 1.7083625793457031, + "learning_rate": 4.085361057819348e-07, + "loss": 3.089, + "step": 22735 + }, + { + "epoch": 1.8227536028542222, + "grad_norm": 1.805732250213623, + "learning_rate": 4.0816884266943636e-07, + "loss": 3.1902, + "step": 22736 + }, + { + "epoch": 1.8228337776352448, + "grad_norm": 1.7740570306777954, + "learning_rate": 4.0780174127183045e-07, + "loss": 3.1709, + "step": 22737 + }, + { + "epoch": 1.8229139524162674, + "grad_norm": 1.829507827758789, + "learning_rate": 4.074348015953056e-07, + "loss": 3.2673, + "step": 22738 + }, + { + "epoch": 1.82299412719729, + "grad_norm": 1.8189167976379395, + "learning_rate": 4.0706802364605e-07, + "loss": 3.1366, + "step": 22739 + }, + { + "epoch": 1.8230743019783127, + "grad_norm": 1.8412681818008423, + "learning_rate": 4.0670140743024666e-07, + "loss": 3.1727, + "step": 22740 + }, + { + "epoch": 1.8231544767593353, + "grad_norm": 1.8444703817367554, + "learning_rate": 4.063349529540772e-07, + "loss": 3.1502, + "step": 22741 + }, + { + "epoch": 1.823234651540358, + "grad_norm": 1.8283237218856812, + "learning_rate": 4.0596866022371896e-07, + "loss": 3.1991, + "step": 22742 + }, + { + "epoch": 1.8233148263213805, + "grad_norm": 1.7840291261672974, + "learning_rate": 4.0560252924534695e-07, + "loss": 3.0891, + "step": 22743 + }, + { + "epoch": 1.8233950011024032, + "grad_norm": 1.7288389205932617, + "learning_rate": 4.0523656002513624e-07, + "loss": 3.1266, + "step": 22744 + }, + { + "epoch": 1.8234751758834258, + "grad_norm": 0.6605602502822876, + "learning_rate": 4.0487075256925413e-07, + "loss": 2.4602, + "step": 22745 + }, + { + "epoch": 1.8235553506644484, + "grad_norm": 1.833587408065796, + "learning_rate": 4.0450510688387125e-07, + "loss": 3.178, + "step": 22746 + }, + { + "epoch": 1.823635525445471, + "grad_norm": 1.8259261846542358, + "learning_rate": 4.041396229751493e-07, + "loss": 3.1149, + "step": 22747 + }, + { + "epoch": 1.8237157002264937, + "grad_norm": 1.7026454210281372, + "learning_rate": 4.0377430084925117e-07, + "loss": 3.0988, + "step": 22748 + }, + { + "epoch": 1.8237958750075163, + "grad_norm": 1.604237675666809, + "learning_rate": 4.034091405123375e-07, + "loss": 3.075, + "step": 22749 + }, + { + "epoch": 1.823876049788539, + "grad_norm": 1.7947996854782104, + "learning_rate": 4.030441419705622e-07, + "loss": 3.1407, + "step": 22750 + }, + { + "epoch": 1.8239562245695615, + "grad_norm": 1.7543784379959106, + "learning_rate": 4.026793052300815e-07, + "loss": 3.0563, + "step": 22751 + }, + { + "epoch": 1.8240363993505841, + "grad_norm": 1.7739458084106445, + "learning_rate": 4.023146302970449e-07, + "loss": 3.2095, + "step": 22752 + }, + { + "epoch": 1.8241165741316068, + "grad_norm": 1.7927751541137695, + "learning_rate": 4.019501171776008e-07, + "loss": 3.2218, + "step": 22753 + }, + { + "epoch": 1.8241967489126294, + "grad_norm": 1.8161269426345825, + "learning_rate": 4.015857658778954e-07, + "loss": 3.3038, + "step": 22754 + }, + { + "epoch": 1.824276923693652, + "grad_norm": 1.734946370124817, + "learning_rate": 4.0122157640407054e-07, + "loss": 3.1569, + "step": 22755 + }, + { + "epoch": 1.8243570984746746, + "grad_norm": 1.7737541198730469, + "learning_rate": 4.008575487622679e-07, + "loss": 3.1562, + "step": 22756 + }, + { + "epoch": 1.8244372732556973, + "grad_norm": 1.7796176671981812, + "learning_rate": 4.004936829586237e-07, + "loss": 3.09, + "step": 22757 + }, + { + "epoch": 1.82451744803672, + "grad_norm": 1.793141484260559, + "learning_rate": 4.0012997899927187e-07, + "loss": 3.1505, + "step": 22758 + }, + { + "epoch": 1.8245976228177425, + "grad_norm": 1.736147165298462, + "learning_rate": 3.9976643689034643e-07, + "loss": 3.1644, + "step": 22759 + }, + { + "epoch": 1.8246777975987654, + "grad_norm": 1.8012945652008057, + "learning_rate": 3.9940305663797474e-07, + "loss": 3.2203, + "step": 22760 + }, + { + "epoch": 1.824757972379788, + "grad_norm": 1.7064920663833618, + "learning_rate": 3.990398382482841e-07, + "loss": 3.2306, + "step": 22761 + }, + { + "epoch": 1.8248381471608106, + "grad_norm": 1.8493609428405762, + "learning_rate": 3.9867678172739843e-07, + "loss": 3.2524, + "step": 22762 + }, + { + "epoch": 1.8249183219418332, + "grad_norm": 1.8047175407409668, + "learning_rate": 3.983138870814374e-07, + "loss": 3.0741, + "step": 22763 + }, + { + "epoch": 1.8249984967228559, + "grad_norm": 1.8010156154632568, + "learning_rate": 3.9795115431652154e-07, + "loss": 3.1387, + "step": 22764 + }, + { + "epoch": 1.8250786715038785, + "grad_norm": 1.8099206686019897, + "learning_rate": 3.975885834387638e-07, + "loss": 3.0616, + "step": 22765 + }, + { + "epoch": 1.825158846284901, + "grad_norm": 0.6694516539573669, + "learning_rate": 3.9722617445428045e-07, + "loss": 2.4641, + "step": 22766 + }, + { + "epoch": 1.8252390210659237, + "grad_norm": 1.713819980621338, + "learning_rate": 3.968639273691777e-07, + "loss": 3.0936, + "step": 22767 + }, + { + "epoch": 1.8253191958469464, + "grad_norm": 1.784788966178894, + "learning_rate": 3.96501842189565e-07, + "loss": 3.2356, + "step": 22768 + }, + { + "epoch": 1.825399370627969, + "grad_norm": 1.8362876176834106, + "learning_rate": 3.9613991892154534e-07, + "loss": 3.2979, + "step": 22769 + }, + { + "epoch": 1.8254795454089916, + "grad_norm": 0.6595775485038757, + "learning_rate": 3.957781575712238e-07, + "loss": 2.4535, + "step": 22770 + }, + { + "epoch": 1.8255597201900142, + "grad_norm": 1.7550386190414429, + "learning_rate": 3.9541655814469783e-07, + "loss": 3.1276, + "step": 22771 + }, + { + "epoch": 1.8256398949710368, + "grad_norm": 1.745720386505127, + "learning_rate": 3.950551206480624e-07, + "loss": 3.1027, + "step": 22772 + }, + { + "epoch": 1.8257200697520595, + "grad_norm": 1.709770679473877, + "learning_rate": 3.946938450874127e-07, + "loss": 3.1089, + "step": 22773 + }, + { + "epoch": 1.825800244533082, + "grad_norm": 0.6860920190811157, + "learning_rate": 3.9433273146884054e-07, + "loss": 2.4672, + "step": 22774 + }, + { + "epoch": 1.8258804193141047, + "grad_norm": 1.7104291915893555, + "learning_rate": 3.939717797984322e-07, + "loss": 3.1262, + "step": 22775 + }, + { + "epoch": 1.8259605940951273, + "grad_norm": 1.6938862800598145, + "learning_rate": 3.936109900822749e-07, + "loss": 3.1329, + "step": 22776 + }, + { + "epoch": 1.82604076887615, + "grad_norm": 1.6679664850234985, + "learning_rate": 3.932503623264505e-07, + "loss": 3.1708, + "step": 22777 + }, + { + "epoch": 1.8261209436571726, + "grad_norm": 1.852232813835144, + "learning_rate": 3.928898965370387e-07, + "loss": 3.1815, + "step": 22778 + }, + { + "epoch": 1.8262011184381952, + "grad_norm": 1.6477185487747192, + "learning_rate": 3.925295927201178e-07, + "loss": 3.1196, + "step": 22779 + }, + { + "epoch": 1.826281293219218, + "grad_norm": 0.6792691349983215, + "learning_rate": 3.9216945088176084e-07, + "loss": 2.4647, + "step": 22780 + }, + { + "epoch": 1.8263614680002407, + "grad_norm": 1.774282455444336, + "learning_rate": 3.918094710280429e-07, + "loss": 3.186, + "step": 22781 + }, + { + "epoch": 1.8264416427812633, + "grad_norm": 1.8948818445205688, + "learning_rate": 3.914496531650303e-07, + "loss": 3.1616, + "step": 22782 + }, + { + "epoch": 1.826521817562286, + "grad_norm": 1.777587890625, + "learning_rate": 3.9108999729878916e-07, + "loss": 3.1806, + "step": 22783 + }, + { + "epoch": 1.8266019923433086, + "grad_norm": 1.7499310970306396, + "learning_rate": 3.9073050343538476e-07, + "loss": 3.2521, + "step": 22784 + }, + { + "epoch": 1.8266821671243312, + "grad_norm": 1.710891604423523, + "learning_rate": 3.9037117158087776e-07, + "loss": 3.0871, + "step": 22785 + }, + { + "epoch": 1.8267623419053538, + "grad_norm": 1.722467064857483, + "learning_rate": 3.900120017413267e-07, + "loss": 3.1694, + "step": 22786 + }, + { + "epoch": 1.8268425166863764, + "grad_norm": 0.6489289999008179, + "learning_rate": 3.8965299392278554e-07, + "loss": 2.4618, + "step": 22787 + }, + { + "epoch": 1.826922691467399, + "grad_norm": 1.7032599449157715, + "learning_rate": 3.892941481313073e-07, + "loss": 3.1259, + "step": 22788 + }, + { + "epoch": 1.8270028662484217, + "grad_norm": 1.7643858194351196, + "learning_rate": 3.889354643729426e-07, + "loss": 3.0692, + "step": 22789 + }, + { + "epoch": 1.8270830410294443, + "grad_norm": 1.7702982425689697, + "learning_rate": 3.885769426537389e-07, + "loss": 3.1396, + "step": 22790 + }, + { + "epoch": 1.827163215810467, + "grad_norm": 1.7574795484542847, + "learning_rate": 3.882185829797425e-07, + "loss": 3.1108, + "step": 22791 + }, + { + "epoch": 1.8272433905914895, + "grad_norm": 1.845291256904602, + "learning_rate": 3.8786038535699067e-07, + "loss": 3.1869, + "step": 22792 + }, + { + "epoch": 1.8273235653725122, + "grad_norm": 1.7438448667526245, + "learning_rate": 3.8750234979152645e-07, + "loss": 3.1556, + "step": 22793 + }, + { + "epoch": 1.8274037401535348, + "grad_norm": 1.8237799406051636, + "learning_rate": 3.871444762893839e-07, + "loss": 3.1617, + "step": 22794 + }, + { + "epoch": 1.8274839149345574, + "grad_norm": 1.7178831100463867, + "learning_rate": 3.8678676485659817e-07, + "loss": 3.1322, + "step": 22795 + }, + { + "epoch": 1.82756408971558, + "grad_norm": 1.7182121276855469, + "learning_rate": 3.864292154992e-07, + "loss": 3.2191, + "step": 22796 + }, + { + "epoch": 1.8276442644966027, + "grad_norm": 1.8035706281661987, + "learning_rate": 3.8607182822321677e-07, + "loss": 3.2245, + "step": 22797 + }, + { + "epoch": 1.8277244392776253, + "grad_norm": 1.707062840461731, + "learning_rate": 3.8571460303467366e-07, + "loss": 3.1046, + "step": 22798 + }, + { + "epoch": 1.827804614058648, + "grad_norm": 1.749570369720459, + "learning_rate": 3.8535753993959477e-07, + "loss": 3.0538, + "step": 22799 + }, + { + "epoch": 1.8278847888396705, + "grad_norm": 1.8620489835739136, + "learning_rate": 3.8500063894399864e-07, + "loss": 3.3076, + "step": 22800 + }, + { + "epoch": 1.8279649636206932, + "grad_norm": 1.8971458673477173, + "learning_rate": 3.8464390005390374e-07, + "loss": 3.111, + "step": 22801 + }, + { + "epoch": 1.8280451384017158, + "grad_norm": 1.8017314672470093, + "learning_rate": 3.84287323275323e-07, + "loss": 3.0882, + "step": 22802 + }, + { + "epoch": 1.8281253131827384, + "grad_norm": 1.6525390148162842, + "learning_rate": 3.839309086142695e-07, + "loss": 3.0806, + "step": 22803 + }, + { + "epoch": 1.828205487963761, + "grad_norm": 1.7138904333114624, + "learning_rate": 3.835746560767517e-07, + "loss": 3.283, + "step": 22804 + }, + { + "epoch": 1.8282856627447837, + "grad_norm": 1.7962533235549927, + "learning_rate": 3.832185656687759e-07, + "loss": 3.1053, + "step": 22805 + }, + { + "epoch": 1.8283658375258063, + "grad_norm": 1.8090331554412842, + "learning_rate": 3.828626373963462e-07, + "loss": 3.1301, + "step": 22806 + }, + { + "epoch": 1.828446012306829, + "grad_norm": 1.860929012298584, + "learning_rate": 3.825068712654634e-07, + "loss": 3.1295, + "step": 22807 + }, + { + "epoch": 1.8285261870878515, + "grad_norm": 1.771437644958496, + "learning_rate": 3.821512672821248e-07, + "loss": 3.1219, + "step": 22808 + }, + { + "epoch": 1.8286063618688742, + "grad_norm": 1.8458997011184692, + "learning_rate": 3.8179582545232575e-07, + "loss": 3.1346, + "step": 22809 + }, + { + "epoch": 1.8286865366498968, + "grad_norm": 1.78816819190979, + "learning_rate": 3.8144054578206025e-07, + "loss": 3.2083, + "step": 22810 + }, + { + "epoch": 1.8287667114309194, + "grad_norm": 1.7218103408813477, + "learning_rate": 3.8108542827731575e-07, + "loss": 3.146, + "step": 22811 + }, + { + "epoch": 1.828846886211942, + "grad_norm": 1.803756594657898, + "learning_rate": 3.8073047294408303e-07, + "loss": 3.2197, + "step": 22812 + }, + { + "epoch": 1.8289270609929646, + "grad_norm": 1.771649718284607, + "learning_rate": 3.803756797883429e-07, + "loss": 3.1151, + "step": 22813 + }, + { + "epoch": 1.8290072357739873, + "grad_norm": 1.7459319829940796, + "learning_rate": 3.8002104881607826e-07, + "loss": 3.1169, + "step": 22814 + }, + { + "epoch": 1.82908741055501, + "grad_norm": 1.7654865980148315, + "learning_rate": 3.7966658003326885e-07, + "loss": 3.1666, + "step": 22815 + }, + { + "epoch": 1.8291675853360325, + "grad_norm": 1.8079907894134521, + "learning_rate": 3.7931227344589097e-07, + "loss": 3.1514, + "step": 22816 + }, + { + "epoch": 1.8292477601170551, + "grad_norm": 1.7251437902450562, + "learning_rate": 3.7895812905991756e-07, + "loss": 3.1044, + "step": 22817 + }, + { + "epoch": 1.8293279348980778, + "grad_norm": 1.7648916244506836, + "learning_rate": 3.786041468813184e-07, + "loss": 3.0945, + "step": 22818 + }, + { + "epoch": 1.8294081096791004, + "grad_norm": 1.8604390621185303, + "learning_rate": 3.78250326916062e-07, + "loss": 3.2266, + "step": 22819 + }, + { + "epoch": 1.829488284460123, + "grad_norm": 1.8045192956924438, + "learning_rate": 3.778966691701147e-07, + "loss": 3.2025, + "step": 22820 + }, + { + "epoch": 1.8295684592411456, + "grad_norm": 1.7316604852676392, + "learning_rate": 3.7754317364943837e-07, + "loss": 3.1947, + "step": 22821 + }, + { + "epoch": 1.8296486340221683, + "grad_norm": 1.8494775295257568, + "learning_rate": 3.7718984035999383e-07, + "loss": 3.2274, + "step": 22822 + }, + { + "epoch": 1.8297288088031909, + "grad_norm": 1.755867838859558, + "learning_rate": 3.768366693077363e-07, + "loss": 3.2013, + "step": 22823 + }, + { + "epoch": 1.8298089835842135, + "grad_norm": 1.8022217750549316, + "learning_rate": 3.7648366049861996e-07, + "loss": 3.1469, + "step": 22824 + }, + { + "epoch": 1.8298891583652361, + "grad_norm": 1.7545831203460693, + "learning_rate": 3.7613081393859885e-07, + "loss": 3.1552, + "step": 22825 + }, + { + "epoch": 1.8299693331462588, + "grad_norm": 1.7925455570220947, + "learning_rate": 3.757781296336194e-07, + "loss": 3.1697, + "step": 22826 + }, + { + "epoch": 1.8300495079272814, + "grad_norm": 1.7801766395568848, + "learning_rate": 3.754256075896312e-07, + "loss": 3.1701, + "step": 22827 + }, + { + "epoch": 1.830129682708304, + "grad_norm": 1.7818984985351562, + "learning_rate": 3.750732478125729e-07, + "loss": 3.123, + "step": 22828 + }, + { + "epoch": 1.8302098574893266, + "grad_norm": 1.7757198810577393, + "learning_rate": 3.7472105030838755e-07, + "loss": 3.1653, + "step": 22829 + }, + { + "epoch": 1.8302900322703493, + "grad_norm": 1.8731015920639038, + "learning_rate": 3.743690150830137e-07, + "loss": 3.2154, + "step": 22830 + }, + { + "epoch": 1.8303702070513719, + "grad_norm": 1.7750760316848755, + "learning_rate": 3.740171421423855e-07, + "loss": 3.2812, + "step": 22831 + }, + { + "epoch": 1.8304503818323945, + "grad_norm": 1.8334262371063232, + "learning_rate": 3.736654314924359e-07, + "loss": 3.1716, + "step": 22832 + }, + { + "epoch": 1.8305305566134171, + "grad_norm": 1.7026844024658203, + "learning_rate": 3.733138831390948e-07, + "loss": 3.1328, + "step": 22833 + }, + { + "epoch": 1.8306107313944397, + "grad_norm": 0.6587993502616882, + "learning_rate": 3.7296249708828834e-07, + "loss": 2.4658, + "step": 22834 + }, + { + "epoch": 1.8306909061754624, + "grad_norm": 1.7795377969741821, + "learning_rate": 3.726112733459408e-07, + "loss": 3.2416, + "step": 22835 + }, + { + "epoch": 1.830771080956485, + "grad_norm": 1.929040551185608, + "learning_rate": 3.7226021191797524e-07, + "loss": 3.275, + "step": 22836 + }, + { + "epoch": 1.8308512557375076, + "grad_norm": 1.7095142602920532, + "learning_rate": 3.719093128103102e-07, + "loss": 3.1796, + "step": 22837 + }, + { + "epoch": 1.8309314305185302, + "grad_norm": 1.7203288078308105, + "learning_rate": 3.7155857602885867e-07, + "loss": 3.0544, + "step": 22838 + }, + { + "epoch": 1.8310116052995529, + "grad_norm": 1.7982354164123535, + "learning_rate": 3.712080015795372e-07, + "loss": 3.1741, + "step": 22839 + }, + { + "epoch": 1.8310917800805755, + "grad_norm": 1.8346364498138428, + "learning_rate": 3.7085758946825534e-07, + "loss": 3.1499, + "step": 22840 + }, + { + "epoch": 1.8311719548615983, + "grad_norm": 1.7381057739257812, + "learning_rate": 3.705073397009207e-07, + "loss": 3.132, + "step": 22841 + }, + { + "epoch": 1.831252129642621, + "grad_norm": 1.9054163694381714, + "learning_rate": 3.701572522834396e-07, + "loss": 3.196, + "step": 22842 + }, + { + "epoch": 1.8313323044236436, + "grad_norm": 1.7667858600616455, + "learning_rate": 3.6980732722171177e-07, + "loss": 3.2806, + "step": 22843 + }, + { + "epoch": 1.8314124792046662, + "grad_norm": 1.7646315097808838, + "learning_rate": 3.694575645216392e-07, + "loss": 3.0455, + "step": 22844 + }, + { + "epoch": 1.8314926539856888, + "grad_norm": 1.8815429210662842, + "learning_rate": 3.6910796418911707e-07, + "loss": 3.1633, + "step": 22845 + }, + { + "epoch": 1.8315728287667115, + "grad_norm": 1.7767301797866821, + "learning_rate": 3.6875852623004084e-07, + "loss": 3.1948, + "step": 22846 + }, + { + "epoch": 1.831653003547734, + "grad_norm": 1.7551591396331787, + "learning_rate": 3.684092506503023e-07, + "loss": 3.2086, + "step": 22847 + }, + { + "epoch": 1.8317331783287567, + "grad_norm": 1.7247507572174072, + "learning_rate": 3.6806013745578906e-07, + "loss": 3.1256, + "step": 22848 + }, + { + "epoch": 1.8318133531097793, + "grad_norm": 1.9147446155548096, + "learning_rate": 3.677111866523864e-07, + "loss": 3.1699, + "step": 22849 + }, + { + "epoch": 1.831893527890802, + "grad_norm": 1.7981104850769043, + "learning_rate": 3.673623982459784e-07, + "loss": 3.283, + "step": 22850 + }, + { + "epoch": 1.8319737026718246, + "grad_norm": 1.8079780340194702, + "learning_rate": 3.6701377224244606e-07, + "loss": 3.2437, + "step": 22851 + }, + { + "epoch": 1.8320538774528472, + "grad_norm": 1.7933106422424316, + "learning_rate": 3.6666530864766793e-07, + "loss": 3.111, + "step": 22852 + }, + { + "epoch": 1.8321340522338698, + "grad_norm": 1.7188963890075684, + "learning_rate": 3.66317007467516e-07, + "loss": 3.1366, + "step": 22853 + }, + { + "epoch": 1.8322142270148924, + "grad_norm": 1.7825337648391724, + "learning_rate": 3.6596886870786333e-07, + "loss": 3.1505, + "step": 22854 + }, + { + "epoch": 1.832294401795915, + "grad_norm": 1.7238235473632812, + "learning_rate": 3.6562089237458075e-07, + "loss": 3.1609, + "step": 22855 + }, + { + "epoch": 1.8323745765769377, + "grad_norm": 1.7798388004302979, + "learning_rate": 3.6527307847353476e-07, + "loss": 3.0844, + "step": 22856 + }, + { + "epoch": 1.8324547513579603, + "grad_norm": 1.8569692373275757, + "learning_rate": 3.649254270105906e-07, + "loss": 3.1729, + "step": 22857 + }, + { + "epoch": 1.832534926138983, + "grad_norm": 1.7801767587661743, + "learning_rate": 3.645779379916059e-07, + "loss": 3.1335, + "step": 22858 + }, + { + "epoch": 1.8326151009200056, + "grad_norm": 1.7435489892959595, + "learning_rate": 3.6423061142244145e-07, + "loss": 3.274, + "step": 22859 + }, + { + "epoch": 1.8326952757010282, + "grad_norm": 1.7197962999343872, + "learning_rate": 3.6388344730895364e-07, + "loss": 3.0329, + "step": 22860 + }, + { + "epoch": 1.8327754504820508, + "grad_norm": 1.7770522832870483, + "learning_rate": 3.6353644565699344e-07, + "loss": 3.1096, + "step": 22861 + }, + { + "epoch": 1.8328556252630737, + "grad_norm": 0.6450154185295105, + "learning_rate": 3.6318960647241496e-07, + "loss": 2.4509, + "step": 22862 + }, + { + "epoch": 1.8329358000440963, + "grad_norm": 1.7693475484848022, + "learning_rate": 3.628429297610614e-07, + "loss": 3.2155, + "step": 22863 + }, + { + "epoch": 1.833015974825119, + "grad_norm": 1.7204933166503906, + "learning_rate": 3.624964155287802e-07, + "loss": 3.2299, + "step": 22864 + }, + { + "epoch": 1.8330961496061415, + "grad_norm": 1.8617286682128906, + "learning_rate": 3.6215006378141237e-07, + "loss": 3.111, + "step": 22865 + }, + { + "epoch": 1.8331763243871642, + "grad_norm": 1.6418628692626953, + "learning_rate": 3.618038745247987e-07, + "loss": 3.1481, + "step": 22866 + }, + { + "epoch": 1.8332564991681868, + "grad_norm": 0.6540483832359314, + "learning_rate": 3.6145784776477565e-07, + "loss": 2.4548, + "step": 22867 + }, + { + "epoch": 1.8333366739492094, + "grad_norm": 1.7021667957305908, + "learning_rate": 3.6111198350717414e-07, + "loss": 3.1683, + "step": 22868 + }, + { + "epoch": 1.833416848730232, + "grad_norm": 1.826533317565918, + "learning_rate": 3.6076628175782834e-07, + "loss": 3.1774, + "step": 22869 + }, + { + "epoch": 1.8334970235112547, + "grad_norm": 1.9046108722686768, + "learning_rate": 3.6042074252256585e-07, + "loss": 3.2443, + "step": 22870 + }, + { + "epoch": 1.8335771982922773, + "grad_norm": 1.9296072721481323, + "learning_rate": 3.60075365807212e-07, + "loss": 3.2736, + "step": 22871 + }, + { + "epoch": 1.8336573730733, + "grad_norm": 1.7537927627563477, + "learning_rate": 3.597301516175922e-07, + "loss": 3.1996, + "step": 22872 + }, + { + "epoch": 1.8337375478543225, + "grad_norm": 1.7081161737442017, + "learning_rate": 3.5938509995952273e-07, + "loss": 3.1342, + "step": 22873 + }, + { + "epoch": 1.8338177226353451, + "grad_norm": 1.8499621152877808, + "learning_rate": 3.5904021083882243e-07, + "loss": 3.1962, + "step": 22874 + }, + { + "epoch": 1.8338978974163678, + "grad_norm": 1.7079592943191528, + "learning_rate": 3.586954842613066e-07, + "loss": 3.1478, + "step": 22875 + }, + { + "epoch": 1.8339780721973904, + "grad_norm": 1.8017592430114746, + "learning_rate": 3.583509202327862e-07, + "loss": 3.1522, + "step": 22876 + }, + { + "epoch": 1.834058246978413, + "grad_norm": 1.7102376222610474, + "learning_rate": 3.580065187590731e-07, + "loss": 3.0872, + "step": 22877 + }, + { + "epoch": 1.8341384217594356, + "grad_norm": 1.8665060997009277, + "learning_rate": 3.5766227984597166e-07, + "loss": 3.2386, + "step": 22878 + }, + { + "epoch": 1.8342185965404583, + "grad_norm": 1.800445795059204, + "learning_rate": 3.57318203499285e-07, + "loss": 3.1976, + "step": 22879 + }, + { + "epoch": 1.834298771321481, + "grad_norm": 1.9318935871124268, + "learning_rate": 3.569742897248152e-07, + "loss": 3.2434, + "step": 22880 + }, + { + "epoch": 1.8343789461025035, + "grad_norm": 0.6685748100280762, + "learning_rate": 3.566305385283597e-07, + "loss": 2.4592, + "step": 22881 + }, + { + "epoch": 1.8344591208835261, + "grad_norm": 1.7770648002624512, + "learning_rate": 3.562869499157162e-07, + "loss": 3.1897, + "step": 22882 + }, + { + "epoch": 1.8345392956645488, + "grad_norm": 1.8406399488449097, + "learning_rate": 3.559435238926756e-07, + "loss": 3.179, + "step": 22883 + }, + { + "epoch": 1.8346194704455714, + "grad_norm": 1.8060091733932495, + "learning_rate": 3.556002604650266e-07, + "loss": 3.2268, + "step": 22884 + }, + { + "epoch": 1.834699645226594, + "grad_norm": 1.7241486310958862, + "learning_rate": 3.552571596385601e-07, + "loss": 3.1483, + "step": 22885 + }, + { + "epoch": 1.8347798200076166, + "grad_norm": 1.7372602224349976, + "learning_rate": 3.5491422141905706e-07, + "loss": 3.2542, + "step": 22886 + }, + { + "epoch": 1.8348599947886393, + "grad_norm": 1.7941557168960571, + "learning_rate": 3.5457144581230394e-07, + "loss": 3.1722, + "step": 22887 + }, + { + "epoch": 1.8349401695696619, + "grad_norm": 1.7936646938323975, + "learning_rate": 3.542288328240739e-07, + "loss": 3.2006, + "step": 22888 + }, + { + "epoch": 1.8350203443506845, + "grad_norm": 0.6636542081832886, + "learning_rate": 3.5388638246014684e-07, + "loss": 2.4791, + "step": 22889 + }, + { + "epoch": 1.8351005191317071, + "grad_norm": 0.6582680940628052, + "learning_rate": 3.5354409472629693e-07, + "loss": 2.4629, + "step": 22890 + }, + { + "epoch": 1.8351806939127298, + "grad_norm": 1.8977049589157104, + "learning_rate": 3.532019696282929e-07, + "loss": 3.1352, + "step": 22891 + }, + { + "epoch": 1.8352608686937524, + "grad_norm": 1.701823115348816, + "learning_rate": 3.528600071719046e-07, + "loss": 3.2262, + "step": 22892 + }, + { + "epoch": 1.835341043474775, + "grad_norm": 0.6518415212631226, + "learning_rate": 3.525182073628952e-07, + "loss": 2.4774, + "step": 22893 + }, + { + "epoch": 1.8354212182557976, + "grad_norm": 1.7097634077072144, + "learning_rate": 3.521765702070301e-07, + "loss": 3.136, + "step": 22894 + }, + { + "epoch": 1.8355013930368203, + "grad_norm": 1.7418053150177002, + "learning_rate": 3.518350957100669e-07, + "loss": 3.1836, + "step": 22895 + }, + { + "epoch": 1.8355815678178429, + "grad_norm": 1.6985408067703247, + "learning_rate": 3.5149378387776324e-07, + "loss": 3.1783, + "step": 22896 + }, + { + "epoch": 1.8356617425988655, + "grad_norm": 1.8353627920150757, + "learning_rate": 3.5115263471587555e-07, + "loss": 3.1609, + "step": 22897 + }, + { + "epoch": 1.8357419173798881, + "grad_norm": 1.8061366081237793, + "learning_rate": 3.5081164823015267e-07, + "loss": 3.2399, + "step": 22898 + }, + { + "epoch": 1.8358220921609107, + "grad_norm": 1.739108681678772, + "learning_rate": 3.504708244263455e-07, + "loss": 3.1452, + "step": 22899 + }, + { + "epoch": 1.8359022669419334, + "grad_norm": 1.6841964721679688, + "learning_rate": 3.501301633101983e-07, + "loss": 3.1023, + "step": 22900 + }, + { + "epoch": 1.835982441722956, + "grad_norm": 1.9479832649230957, + "learning_rate": 3.497896648874555e-07, + "loss": 3.145, + "step": 22901 + }, + { + "epoch": 1.8360626165039786, + "grad_norm": 1.8794920444488525, + "learning_rate": 3.494493291638601e-07, + "loss": 3.1538, + "step": 22902 + }, + { + "epoch": 1.8361427912850012, + "grad_norm": 1.6675482988357544, + "learning_rate": 3.491091561451454e-07, + "loss": 3.3074, + "step": 22903 + }, + { + "epoch": 1.8362229660660239, + "grad_norm": 1.7208607196807861, + "learning_rate": 3.487691458370501e-07, + "loss": 3.1403, + "step": 22904 + }, + { + "epoch": 1.8363031408470465, + "grad_norm": 1.8362720012664795, + "learning_rate": 3.4842929824530524e-07, + "loss": 3.2893, + "step": 22905 + }, + { + "epoch": 1.8363833156280691, + "grad_norm": 0.641438901424408, + "learning_rate": 3.480896133756406e-07, + "loss": 2.4769, + "step": 22906 + }, + { + "epoch": 1.8364634904090917, + "grad_norm": 1.8608242273330688, + "learning_rate": 3.47750091233785e-07, + "loss": 3.148, + "step": 22907 + }, + { + "epoch": 1.8365436651901144, + "grad_norm": 0.657088041305542, + "learning_rate": 3.4741073182545935e-07, + "loss": 2.4459, + "step": 22908 + }, + { + "epoch": 1.836623839971137, + "grad_norm": 2.0547852516174316, + "learning_rate": 3.47071535156388e-07, + "loss": 3.1615, + "step": 22909 + }, + { + "epoch": 1.8367040147521596, + "grad_norm": 1.8019585609436035, + "learning_rate": 3.467325012322875e-07, + "loss": 3.2284, + "step": 22910 + }, + { + "epoch": 1.8367841895331822, + "grad_norm": 1.8395264148712158, + "learning_rate": 3.463936300588755e-07, + "loss": 3.1855, + "step": 22911 + }, + { + "epoch": 1.8368643643142049, + "grad_norm": 1.704276204109192, + "learning_rate": 3.4605492164186517e-07, + "loss": 3.1127, + "step": 22912 + }, + { + "epoch": 1.8369445390952275, + "grad_norm": 1.7556322813034058, + "learning_rate": 3.4571637598696526e-07, + "loss": 3.205, + "step": 22913 + }, + { + "epoch": 1.83702471387625, + "grad_norm": 1.7898048162460327, + "learning_rate": 3.453779930998846e-07, + "loss": 3.1533, + "step": 22914 + }, + { + "epoch": 1.8371048886572727, + "grad_norm": 1.757930874824524, + "learning_rate": 3.4503977298632973e-07, + "loss": 3.1329, + "step": 22915 + }, + { + "epoch": 1.8371850634382954, + "grad_norm": 1.973419427871704, + "learning_rate": 3.447017156520005e-07, + "loss": 3.2035, + "step": 22916 + }, + { + "epoch": 1.837265238219318, + "grad_norm": 1.8458523750305176, + "learning_rate": 3.44363821102599e-07, + "loss": 3.1865, + "step": 22917 + }, + { + "epoch": 1.8373454130003406, + "grad_norm": 1.8420348167419434, + "learning_rate": 3.4402608934381855e-07, + "loss": 3.2725, + "step": 22918 + }, + { + "epoch": 1.8374255877813632, + "grad_norm": 1.6700031757354736, + "learning_rate": 3.436885203813556e-07, + "loss": 3.1646, + "step": 22919 + }, + { + "epoch": 1.8375057625623858, + "grad_norm": 1.9123833179473877, + "learning_rate": 3.433511142209012e-07, + "loss": 3.1931, + "step": 22920 + }, + { + "epoch": 1.8375859373434085, + "grad_norm": 1.8529880046844482, + "learning_rate": 3.430138708681441e-07, + "loss": 3.1394, + "step": 22921 + }, + { + "epoch": 1.837666112124431, + "grad_norm": 1.8137850761413574, + "learning_rate": 3.4267679032876975e-07, + "loss": 3.1371, + "step": 22922 + }, + { + "epoch": 1.837746286905454, + "grad_norm": 1.779115915298462, + "learning_rate": 3.4233987260846145e-07, + "loss": 3.1685, + "step": 22923 + }, + { + "epoch": 1.8378264616864766, + "grad_norm": 1.7578343152999878, + "learning_rate": 3.4200311771289797e-07, + "loss": 3.1603, + "step": 22924 + }, + { + "epoch": 1.8379066364674992, + "grad_norm": 0.641437292098999, + "learning_rate": 3.41666525647758e-07, + "loss": 2.4534, + "step": 22925 + }, + { + "epoch": 1.8379868112485218, + "grad_norm": 1.779800295829773, + "learning_rate": 3.4133009641871716e-07, + "loss": 3.2546, + "step": 22926 + }, + { + "epoch": 1.8380669860295444, + "grad_norm": 1.7332143783569336, + "learning_rate": 3.409938300314475e-07, + "loss": 3.1685, + "step": 22927 + }, + { + "epoch": 1.838147160810567, + "grad_norm": 1.7596969604492188, + "learning_rate": 3.4065772649161775e-07, + "loss": 3.1476, + "step": 22928 + }, + { + "epoch": 1.8382273355915897, + "grad_norm": 1.862729549407959, + "learning_rate": 3.403217858048935e-07, + "loss": 3.1717, + "step": 22929 + }, + { + "epoch": 1.8383075103726123, + "grad_norm": 1.694676160812378, + "learning_rate": 3.3998600797693904e-07, + "loss": 3.2488, + "step": 22930 + }, + { + "epoch": 1.838387685153635, + "grad_norm": 1.7492907047271729, + "learning_rate": 3.396503930134165e-07, + "loss": 3.2625, + "step": 22931 + }, + { + "epoch": 1.8384678599346576, + "grad_norm": 1.9086360931396484, + "learning_rate": 3.3931494091998473e-07, + "loss": 3.1739, + "step": 22932 + }, + { + "epoch": 1.8385480347156802, + "grad_norm": 1.714734435081482, + "learning_rate": 3.389796517022981e-07, + "loss": 3.1483, + "step": 22933 + }, + { + "epoch": 1.8386282094967028, + "grad_norm": 1.7001347541809082, + "learning_rate": 3.386445253660087e-07, + "loss": 3.1017, + "step": 22934 + }, + { + "epoch": 1.8387083842777254, + "grad_norm": 1.8601566553115845, + "learning_rate": 3.3830956191676754e-07, + "loss": 3.2087, + "step": 22935 + }, + { + "epoch": 1.838788559058748, + "grad_norm": 1.8040660619735718, + "learning_rate": 3.3797476136022243e-07, + "loss": 3.2653, + "step": 22936 + }, + { + "epoch": 1.8388687338397707, + "grad_norm": 1.7384177446365356, + "learning_rate": 3.376401237020188e-07, + "loss": 3.1632, + "step": 22937 + }, + { + "epoch": 1.8389489086207933, + "grad_norm": 1.6914775371551514, + "learning_rate": 3.373056489477955e-07, + "loss": 3.1172, + "step": 22938 + }, + { + "epoch": 1.839029083401816, + "grad_norm": 1.8163187503814697, + "learning_rate": 3.369713371031935e-07, + "loss": 3.2054, + "step": 22939 + }, + { + "epoch": 1.8391092581828385, + "grad_norm": 1.7899634838104248, + "learning_rate": 3.366371881738506e-07, + "loss": 3.0523, + "step": 22940 + }, + { + "epoch": 1.8391894329638612, + "grad_norm": 1.9004766941070557, + "learning_rate": 3.363032021653978e-07, + "loss": 3.2434, + "step": 22941 + }, + { + "epoch": 1.8392696077448838, + "grad_norm": 1.741502046585083, + "learning_rate": 3.359693790834684e-07, + "loss": 3.1343, + "step": 22942 + }, + { + "epoch": 1.8393497825259066, + "grad_norm": 1.736700415611267, + "learning_rate": 3.3563571893368896e-07, + "loss": 3.0796, + "step": 22943 + }, + { + "epoch": 1.8394299573069293, + "grad_norm": 1.7692816257476807, + "learning_rate": 3.3530222172168393e-07, + "loss": 3.1775, + "step": 22944 + }, + { + "epoch": 1.8395101320879519, + "grad_norm": 1.7050739526748657, + "learning_rate": 3.3496888745307874e-07, + "loss": 3.2336, + "step": 22945 + }, + { + "epoch": 1.8395903068689745, + "grad_norm": 1.7503653764724731, + "learning_rate": 3.3463571613349124e-07, + "loss": 3.0709, + "step": 22946 + }, + { + "epoch": 1.8396704816499971, + "grad_norm": 1.709896206855774, + "learning_rate": 3.3430270776854015e-07, + "loss": 3.1908, + "step": 22947 + }, + { + "epoch": 1.8397506564310198, + "grad_norm": 1.751947045326233, + "learning_rate": 3.3396986236383765e-07, + "loss": 3.0696, + "step": 22948 + }, + { + "epoch": 1.8398308312120424, + "grad_norm": 1.9351609945297241, + "learning_rate": 3.33637179924996e-07, + "loss": 3.0978, + "step": 22949 + }, + { + "epoch": 1.839911005993065, + "grad_norm": 1.714863657951355, + "learning_rate": 3.3330466045762624e-07, + "loss": 3.09, + "step": 22950 + }, + { + "epoch": 1.8399911807740876, + "grad_norm": 1.643042802810669, + "learning_rate": 3.329723039673316e-07, + "loss": 3.1576, + "step": 22951 + }, + { + "epoch": 1.8400713555551103, + "grad_norm": 1.6606098413467407, + "learning_rate": 3.326401104597188e-07, + "loss": 3.1917, + "step": 22952 + }, + { + "epoch": 1.8401515303361329, + "grad_norm": 1.8460686206817627, + "learning_rate": 3.323080799403855e-07, + "loss": 3.2009, + "step": 22953 + }, + { + "epoch": 1.8402317051171555, + "grad_norm": 1.7101802825927734, + "learning_rate": 3.319762124149306e-07, + "loss": 3.199, + "step": 22954 + }, + { + "epoch": 1.8403118798981781, + "grad_norm": 1.7948371171951294, + "learning_rate": 3.3164450788894966e-07, + "loss": 3.1293, + "step": 22955 + }, + { + "epoch": 1.8403920546792008, + "grad_norm": 1.8025703430175781, + "learning_rate": 3.313129663680348e-07, + "loss": 3.1803, + "step": 22956 + }, + { + "epoch": 1.8404722294602234, + "grad_norm": 1.7935872077941895, + "learning_rate": 3.3098158785777714e-07, + "loss": 3.1663, + "step": 22957 + }, + { + "epoch": 1.840552404241246, + "grad_norm": 1.811438798904419, + "learning_rate": 3.306503723637611e-07, + "loss": 3.1619, + "step": 22958 + }, + { + "epoch": 1.8406325790222686, + "grad_norm": 1.8013640642166138, + "learning_rate": 3.303193198915711e-07, + "loss": 3.2559, + "step": 22959 + }, + { + "epoch": 1.8407127538032912, + "grad_norm": 1.800400972366333, + "learning_rate": 3.299884304467904e-07, + "loss": 3.1387, + "step": 22960 + }, + { + "epoch": 1.8407929285843139, + "grad_norm": 1.9761778116226196, + "learning_rate": 3.2965770403499573e-07, + "loss": 3.2687, + "step": 22961 + }, + { + "epoch": 1.8408731033653365, + "grad_norm": 1.7085723876953125, + "learning_rate": 3.29327140661766e-07, + "loss": 3.0918, + "step": 22962 + }, + { + "epoch": 1.8409532781463591, + "grad_norm": 1.7936747074127197, + "learning_rate": 3.2899674033267105e-07, + "loss": 3.1499, + "step": 22963 + }, + { + "epoch": 1.8410334529273817, + "grad_norm": 1.7988505363464355, + "learning_rate": 3.286665030532821e-07, + "loss": 3.0639, + "step": 22964 + }, + { + "epoch": 1.8411136277084044, + "grad_norm": 1.7626378536224365, + "learning_rate": 3.2833642882916796e-07, + "loss": 3.0951, + "step": 22965 + }, + { + "epoch": 1.841193802489427, + "grad_norm": 1.768955945968628, + "learning_rate": 3.280065176658931e-07, + "loss": 3.0457, + "step": 22966 + }, + { + "epoch": 1.8412739772704496, + "grad_norm": 1.7055344581604004, + "learning_rate": 3.276767695690208e-07, + "loss": 3.1071, + "step": 22967 + }, + { + "epoch": 1.8413541520514722, + "grad_norm": 1.7685678005218506, + "learning_rate": 3.2734718454410784e-07, + "loss": 3.2026, + "step": 22968 + }, + { + "epoch": 1.8414343268324949, + "grad_norm": 1.804646372795105, + "learning_rate": 3.270177625967119e-07, + "loss": 3.1488, + "step": 22969 + }, + { + "epoch": 1.8415145016135175, + "grad_norm": 1.755157709121704, + "learning_rate": 3.2668850373238856e-07, + "loss": 3.1162, + "step": 22970 + }, + { + "epoch": 1.84159467639454, + "grad_norm": 1.696108102798462, + "learning_rate": 3.2635940795668675e-07, + "loss": 3.1218, + "step": 22971 + }, + { + "epoch": 1.8416748511755627, + "grad_norm": 1.7960454225540161, + "learning_rate": 3.2603047527515753e-07, + "loss": 3.0962, + "step": 22972 + }, + { + "epoch": 1.8417550259565854, + "grad_norm": 1.7594548463821411, + "learning_rate": 3.257017056933443e-07, + "loss": 3.1438, + "step": 22973 + }, + { + "epoch": 1.841835200737608, + "grad_norm": 1.8799335956573486, + "learning_rate": 3.253730992167903e-07, + "loss": 3.125, + "step": 22974 + }, + { + "epoch": 1.8419153755186306, + "grad_norm": 1.6637403964996338, + "learning_rate": 3.250446558510367e-07, + "loss": 2.97, + "step": 22975 + }, + { + "epoch": 1.8419955502996532, + "grad_norm": 1.8092941045761108, + "learning_rate": 3.247163756016203e-07, + "loss": 3.1567, + "step": 22976 + }, + { + "epoch": 1.8420757250806759, + "grad_norm": 1.776240348815918, + "learning_rate": 3.243882584740765e-07, + "loss": 3.0321, + "step": 22977 + }, + { + "epoch": 1.8421558998616985, + "grad_norm": 1.781286597251892, + "learning_rate": 3.240603044739354e-07, + "loss": 3.1719, + "step": 22978 + }, + { + "epoch": 1.842236074642721, + "grad_norm": 1.7344646453857422, + "learning_rate": 3.237325136067282e-07, + "loss": 3.1799, + "step": 22979 + }, + { + "epoch": 1.8423162494237437, + "grad_norm": 1.6694345474243164, + "learning_rate": 3.2340488587798034e-07, + "loss": 3.0824, + "step": 22980 + }, + { + "epoch": 1.8423964242047663, + "grad_norm": 1.728760838508606, + "learning_rate": 3.230774212932153e-07, + "loss": 3.1267, + "step": 22981 + }, + { + "epoch": 1.842476598985789, + "grad_norm": 1.7716559171676636, + "learning_rate": 3.2275011985795635e-07, + "loss": 3.0989, + "step": 22982 + }, + { + "epoch": 1.8425567737668116, + "grad_norm": 1.725101113319397, + "learning_rate": 3.22422981577718e-07, + "loss": 3.1996, + "step": 22983 + }, + { + "epoch": 1.8426369485478342, + "grad_norm": 1.8378154039382935, + "learning_rate": 3.220960064580181e-07, + "loss": 3.2582, + "step": 22984 + }, + { + "epoch": 1.8427171233288568, + "grad_norm": 1.7796080112457275, + "learning_rate": 3.217691945043677e-07, + "loss": 3.1016, + "step": 22985 + }, + { + "epoch": 1.8427972981098795, + "grad_norm": 1.8339686393737793, + "learning_rate": 3.214425457222781e-07, + "loss": 3.2899, + "step": 22986 + }, + { + "epoch": 1.842877472890902, + "grad_norm": 0.6479327082633972, + "learning_rate": 3.2111606011725806e-07, + "loss": 2.4764, + "step": 22987 + }, + { + "epoch": 1.8429576476719247, + "grad_norm": 1.7713048458099365, + "learning_rate": 3.207897376948077e-07, + "loss": 3.1494, + "step": 22988 + }, + { + "epoch": 1.8430378224529473, + "grad_norm": 1.7387962341308594, + "learning_rate": 3.204635784604326e-07, + "loss": 3.3105, + "step": 22989 + }, + { + "epoch": 1.84311799723397, + "grad_norm": 1.7666003704071045, + "learning_rate": 3.2013758241962957e-07, + "loss": 3.2107, + "step": 22990 + }, + { + "epoch": 1.8431981720149926, + "grad_norm": 1.775466799736023, + "learning_rate": 3.1981174957789405e-07, + "loss": 3.1317, + "step": 22991 + }, + { + "epoch": 1.8432783467960152, + "grad_norm": 1.8116391897201538, + "learning_rate": 3.19486079940724e-07, + "loss": 3.2023, + "step": 22992 + }, + { + "epoch": 1.8433585215770378, + "grad_norm": 1.7639024257659912, + "learning_rate": 3.1916057351360385e-07, + "loss": 3.1166, + "step": 22993 + }, + { + "epoch": 1.8434386963580605, + "grad_norm": 1.7439186573028564, + "learning_rate": 3.188352303020259e-07, + "loss": 3.1748, + "step": 22994 + }, + { + "epoch": 1.843518871139083, + "grad_norm": 1.7296216487884521, + "learning_rate": 3.1851005031147353e-07, + "loss": 3.1664, + "step": 22995 + }, + { + "epoch": 1.8435990459201057, + "grad_norm": 1.7944287061691284, + "learning_rate": 3.1818503354743014e-07, + "loss": 3.1672, + "step": 22996 + }, + { + "epoch": 1.8436792207011283, + "grad_norm": 1.6718515157699585, + "learning_rate": 3.178601800153758e-07, + "loss": 3.1644, + "step": 22997 + }, + { + "epoch": 1.843759395482151, + "grad_norm": 1.8869436979293823, + "learning_rate": 3.17535489720785e-07, + "loss": 3.1706, + "step": 22998 + }, + { + "epoch": 1.8438395702631736, + "grad_norm": 1.746189832687378, + "learning_rate": 3.1721096266913333e-07, + "loss": 3.1937, + "step": 22999 + }, + { + "epoch": 1.8439197450441962, + "grad_norm": 1.7946547269821167, + "learning_rate": 3.16886598865892e-07, + "loss": 3.2063, + "step": 23000 + }, + { + "epoch": 1.8439999198252188, + "grad_norm": 1.725804328918457, + "learning_rate": 3.165623983165311e-07, + "loss": 3.0502, + "step": 23001 + }, + { + "epoch": 1.8440800946062414, + "grad_norm": 0.647245466709137, + "learning_rate": 3.162383610265163e-07, + "loss": 2.4689, + "step": 23002 + }, + { + "epoch": 1.844160269387264, + "grad_norm": 1.860460638999939, + "learning_rate": 3.1591448700130755e-07, + "loss": 3.225, + "step": 23003 + }, + { + "epoch": 1.844240444168287, + "grad_norm": 1.7511649131774902, + "learning_rate": 3.1559077624636837e-07, + "loss": 3.2161, + "step": 23004 + }, + { + "epoch": 1.8443206189493095, + "grad_norm": 1.7934104204177856, + "learning_rate": 3.1526722876715546e-07, + "loss": 3.1573, + "step": 23005 + }, + { + "epoch": 1.8444007937303322, + "grad_norm": 1.7449355125427246, + "learning_rate": 3.149438445691244e-07, + "loss": 3.1847, + "step": 23006 + }, + { + "epoch": 1.8444809685113548, + "grad_norm": 1.7206443548202515, + "learning_rate": 3.146206236577276e-07, + "loss": 3.1063, + "step": 23007 + }, + { + "epoch": 1.8445611432923774, + "grad_norm": 1.7705601453781128, + "learning_rate": 3.142975660384118e-07, + "loss": 3.2338, + "step": 23008 + }, + { + "epoch": 1.8446413180734, + "grad_norm": 1.6527462005615234, + "learning_rate": 3.139746717166259e-07, + "loss": 3.0977, + "step": 23009 + }, + { + "epoch": 1.8447214928544227, + "grad_norm": 1.7912203073501587, + "learning_rate": 3.136519406978133e-07, + "loss": 3.1525, + "step": 23010 + }, + { + "epoch": 1.8448016676354453, + "grad_norm": 1.8021048307418823, + "learning_rate": 3.133293729874154e-07, + "loss": 3.2554, + "step": 23011 + }, + { + "epoch": 1.844881842416468, + "grad_norm": 1.7493479251861572, + "learning_rate": 3.130069685908721e-07, + "loss": 3.1091, + "step": 23012 + }, + { + "epoch": 1.8449620171974905, + "grad_norm": 1.8395075798034668, + "learning_rate": 3.126847275136147e-07, + "loss": 3.2238, + "step": 23013 + }, + { + "epoch": 1.8450421919785132, + "grad_norm": 1.8258823156356812, + "learning_rate": 3.123626497610799e-07, + "loss": 3.1351, + "step": 23014 + }, + { + "epoch": 1.8451223667595358, + "grad_norm": 1.802221655845642, + "learning_rate": 3.120407353386956e-07, + "loss": 3.2563, + "step": 23015 + }, + { + "epoch": 1.8452025415405584, + "grad_norm": 1.7273720502853394, + "learning_rate": 3.117189842518908e-07, + "loss": 3.1503, + "step": 23016 + }, + { + "epoch": 1.845282716321581, + "grad_norm": 1.7902886867523193, + "learning_rate": 3.1139739650609014e-07, + "loss": 3.1013, + "step": 23017 + }, + { + "epoch": 1.8453628911026037, + "grad_norm": 1.785851001739502, + "learning_rate": 3.110759721067136e-07, + "loss": 3.2016, + "step": 23018 + }, + { + "epoch": 1.8454430658836263, + "grad_norm": 1.7531784772872925, + "learning_rate": 3.1075471105918245e-07, + "loss": 3.2925, + "step": 23019 + }, + { + "epoch": 1.845523240664649, + "grad_norm": 1.862399697303772, + "learning_rate": 3.104336133689112e-07, + "loss": 3.2368, + "step": 23020 + }, + { + "epoch": 1.8456034154456715, + "grad_norm": 1.8711137771606445, + "learning_rate": 3.101126790413145e-07, + "loss": 3.2789, + "step": 23021 + }, + { + "epoch": 1.8456835902266941, + "grad_norm": 1.7741901874542236, + "learning_rate": 3.097919080818035e-07, + "loss": 3.1871, + "step": 23022 + }, + { + "epoch": 1.8457637650077168, + "grad_norm": 1.85262131690979, + "learning_rate": 3.094713004957861e-07, + "loss": 3.2023, + "step": 23023 + }, + { + "epoch": 1.8458439397887396, + "grad_norm": 1.8386024236679077, + "learning_rate": 3.0915085628866583e-07, + "loss": 3.1449, + "step": 23024 + }, + { + "epoch": 1.8459241145697622, + "grad_norm": 1.8159021139144897, + "learning_rate": 3.0883057546584717e-07, + "loss": 3.2019, + "step": 23025 + }, + { + "epoch": 1.8460042893507849, + "grad_norm": 1.6336612701416016, + "learning_rate": 3.085104580327303e-07, + "loss": 3.1611, + "step": 23026 + }, + { + "epoch": 1.8460844641318075, + "grad_norm": 1.7334562540054321, + "learning_rate": 3.0819050399471196e-07, + "loss": 3.2096, + "step": 23027 + }, + { + "epoch": 1.8461646389128301, + "grad_norm": 1.9573397636413574, + "learning_rate": 3.0787071335718454e-07, + "loss": 3.1469, + "step": 23028 + }, + { + "epoch": 1.8462448136938527, + "grad_norm": 1.7744756937026978, + "learning_rate": 3.075510861255415e-07, + "loss": 3.2291, + "step": 23029 + }, + { + "epoch": 1.8463249884748754, + "grad_norm": 0.6497072577476501, + "learning_rate": 3.0723162230517076e-07, + "loss": 2.4626, + "step": 23030 + }, + { + "epoch": 1.846405163255898, + "grad_norm": 1.7638667821884155, + "learning_rate": 3.0691232190145914e-07, + "loss": 3.2242, + "step": 23031 + }, + { + "epoch": 1.8464853380369206, + "grad_norm": 1.753485918045044, + "learning_rate": 3.065931849197912e-07, + "loss": 3.1896, + "step": 23032 + }, + { + "epoch": 1.8465655128179432, + "grad_norm": 1.802337408065796, + "learning_rate": 3.062742113655437e-07, + "loss": 3.1951, + "step": 23033 + }, + { + "epoch": 1.8466456875989659, + "grad_norm": 1.611727237701416, + "learning_rate": 3.0595540124409686e-07, + "loss": 3.1071, + "step": 23034 + }, + { + "epoch": 1.8467258623799885, + "grad_norm": 1.7226779460906982, + "learning_rate": 3.056367545608263e-07, + "loss": 3.1558, + "step": 23035 + }, + { + "epoch": 1.846806037161011, + "grad_norm": 1.6817090511322021, + "learning_rate": 3.0531827132110226e-07, + "loss": 3.1891, + "step": 23036 + }, + { + "epoch": 1.8468862119420337, + "grad_norm": 1.7525347471237183, + "learning_rate": 3.0499995153029813e-07, + "loss": 3.1168, + "step": 23037 + }, + { + "epoch": 1.8469663867230564, + "grad_norm": 1.8946787118911743, + "learning_rate": 3.0468179519377525e-07, + "loss": 3.0952, + "step": 23038 + }, + { + "epoch": 1.847046561504079, + "grad_norm": 1.8118205070495605, + "learning_rate": 3.0436380231690153e-07, + "loss": 3.1407, + "step": 23039 + }, + { + "epoch": 1.8471267362851016, + "grad_norm": 1.7164711952209473, + "learning_rate": 3.04045972905036e-07, + "loss": 3.1252, + "step": 23040 + }, + { + "epoch": 1.8472069110661242, + "grad_norm": 1.8508957624435425, + "learning_rate": 3.037283069635388e-07, + "loss": 3.2319, + "step": 23041 + }, + { + "epoch": 1.8472870858471468, + "grad_norm": 1.7611600160598755, + "learning_rate": 3.0341080449776573e-07, + "loss": 3.1023, + "step": 23042 + }, + { + "epoch": 1.8473672606281695, + "grad_norm": 1.7860538959503174, + "learning_rate": 3.0309346551306794e-07, + "loss": 3.1743, + "step": 23043 + }, + { + "epoch": 1.847447435409192, + "grad_norm": 1.8566343784332275, + "learning_rate": 3.0277629001479793e-07, + "loss": 3.2102, + "step": 23044 + }, + { + "epoch": 1.8475276101902147, + "grad_norm": 1.832364797592163, + "learning_rate": 3.0245927800830134e-07, + "loss": 3.183, + "step": 23045 + }, + { + "epoch": 1.8476077849712373, + "grad_norm": 1.8245460987091064, + "learning_rate": 3.021424294989228e-07, + "loss": 3.2555, + "step": 23046 + }, + { + "epoch": 1.84768795975226, + "grad_norm": 1.7283414602279663, + "learning_rate": 3.01825744492007e-07, + "loss": 3.0824, + "step": 23047 + }, + { + "epoch": 1.8477681345332826, + "grad_norm": 1.7946244478225708, + "learning_rate": 3.015092229928895e-07, + "loss": 3.2036, + "step": 23048 + }, + { + "epoch": 1.8478483093143052, + "grad_norm": 1.7965542078018188, + "learning_rate": 3.011928650069096e-07, + "loss": 3.2019, + "step": 23049 + }, + { + "epoch": 1.8479284840953278, + "grad_norm": 1.7185050249099731, + "learning_rate": 3.008766705393984e-07, + "loss": 3.1436, + "step": 23050 + }, + { + "epoch": 1.8480086588763505, + "grad_norm": 0.6478806734085083, + "learning_rate": 3.005606395956895e-07, + "loss": 2.4664, + "step": 23051 + }, + { + "epoch": 1.848088833657373, + "grad_norm": 1.8896054029464722, + "learning_rate": 3.002447721811097e-07, + "loss": 3.1177, + "step": 23052 + }, + { + "epoch": 1.8481690084383957, + "grad_norm": 1.8256404399871826, + "learning_rate": 2.9992906830098365e-07, + "loss": 3.2238, + "step": 23053 + }, + { + "epoch": 1.8482491832194183, + "grad_norm": 1.7416056394577026, + "learning_rate": 2.9961352796063606e-07, + "loss": 3.1218, + "step": 23054 + }, + { + "epoch": 1.848329358000441, + "grad_norm": 1.7589600086212158, + "learning_rate": 2.992981511653847e-07, + "loss": 3.183, + "step": 23055 + }, + { + "epoch": 1.8484095327814636, + "grad_norm": 1.7773686647415161, + "learning_rate": 2.989829379205478e-07, + "loss": 3.1821, + "step": 23056 + }, + { + "epoch": 1.8484897075624862, + "grad_norm": 1.7781134843826294, + "learning_rate": 2.9866788823143976e-07, + "loss": 3.2325, + "step": 23057 + }, + { + "epoch": 1.8485698823435088, + "grad_norm": 1.8688174486160278, + "learning_rate": 2.98353002103372e-07, + "loss": 3.0887, + "step": 23058 + }, + { + "epoch": 1.8486500571245315, + "grad_norm": 1.7251347303390503, + "learning_rate": 2.9803827954165363e-07, + "loss": 3.1499, + "step": 23059 + }, + { + "epoch": 1.848730231905554, + "grad_norm": 1.821962594985962, + "learning_rate": 2.977237205515904e-07, + "loss": 3.1079, + "step": 23060 + }, + { + "epoch": 1.8488104066865767, + "grad_norm": 0.6613922715187073, + "learning_rate": 2.9740932513848465e-07, + "loss": 2.4808, + "step": 23061 + }, + { + "epoch": 1.8488905814675993, + "grad_norm": 1.8006315231323242, + "learning_rate": 2.9709509330764106e-07, + "loss": 3.0681, + "step": 23062 + }, + { + "epoch": 1.848970756248622, + "grad_norm": 1.7685760259628296, + "learning_rate": 2.967810250643521e-07, + "loss": 3.1953, + "step": 23063 + }, + { + "epoch": 1.8490509310296446, + "grad_norm": 1.7937482595443726, + "learning_rate": 2.9646712041391577e-07, + "loss": 3.1556, + "step": 23064 + }, + { + "epoch": 1.8491311058106672, + "grad_norm": 1.7676974534988403, + "learning_rate": 2.9615337936162445e-07, + "loss": 3.2227, + "step": 23065 + }, + { + "epoch": 1.8492112805916898, + "grad_norm": 1.7524137496948242, + "learning_rate": 2.958398019127662e-07, + "loss": 3.1713, + "step": 23066 + }, + { + "epoch": 1.8492914553727124, + "grad_norm": 1.8164844512939453, + "learning_rate": 2.9552638807263114e-07, + "loss": 3.1566, + "step": 23067 + }, + { + "epoch": 1.849371630153735, + "grad_norm": 1.7845098972320557, + "learning_rate": 2.952131378464984e-07, + "loss": 3.0375, + "step": 23068 + }, + { + "epoch": 1.8494518049347577, + "grad_norm": 1.8001859188079834, + "learning_rate": 2.949000512396527e-07, + "loss": 3.0497, + "step": 23069 + }, + { + "epoch": 1.8495319797157803, + "grad_norm": 1.7452164888381958, + "learning_rate": 2.94587128257372e-07, + "loss": 3.1598, + "step": 23070 + }, + { + "epoch": 1.849612154496803, + "grad_norm": 1.784812092781067, + "learning_rate": 2.94274368904931e-07, + "loss": 3.1096, + "step": 23071 + }, + { + "epoch": 1.8496923292778256, + "grad_norm": 0.6539276242256165, + "learning_rate": 2.9396177318760546e-07, + "loss": 2.4595, + "step": 23072 + }, + { + "epoch": 1.8497725040588482, + "grad_norm": 1.7632619142532349, + "learning_rate": 2.9364934111066224e-07, + "loss": 3.1658, + "step": 23073 + }, + { + "epoch": 1.8498526788398708, + "grad_norm": 1.8370896577835083, + "learning_rate": 2.9333707267937053e-07, + "loss": 3.2001, + "step": 23074 + }, + { + "epoch": 1.8499328536208934, + "grad_norm": 1.8009097576141357, + "learning_rate": 2.9302496789899384e-07, + "loss": 3.261, + "step": 23075 + }, + { + "epoch": 1.850013028401916, + "grad_norm": 1.8027167320251465, + "learning_rate": 2.9271302677479575e-07, + "loss": 3.1232, + "step": 23076 + }, + { + "epoch": 1.8500932031829387, + "grad_norm": 1.898378610610962, + "learning_rate": 2.924012493120365e-07, + "loss": 3.1588, + "step": 23077 + }, + { + "epoch": 1.8501733779639613, + "grad_norm": 1.7786765098571777, + "learning_rate": 2.920896355159686e-07, + "loss": 3.2227, + "step": 23078 + }, + { + "epoch": 1.850253552744984, + "grad_norm": 1.7864834070205688, + "learning_rate": 2.917781853918489e-07, + "loss": 3.071, + "step": 23079 + }, + { + "epoch": 1.8503337275260066, + "grad_norm": 1.7049012184143066, + "learning_rate": 2.9146689894492764e-07, + "loss": 3.1748, + "step": 23080 + }, + { + "epoch": 1.8504139023070292, + "grad_norm": 1.897465705871582, + "learning_rate": 2.9115577618045285e-07, + "loss": 3.1011, + "step": 23081 + }, + { + "epoch": 1.8504940770880518, + "grad_norm": 0.6647346019744873, + "learning_rate": 2.908448171036704e-07, + "loss": 2.4415, + "step": 23082 + }, + { + "epoch": 1.8505742518690744, + "grad_norm": 1.7954298257827759, + "learning_rate": 2.905340217198216e-07, + "loss": 3.1431, + "step": 23083 + }, + { + "epoch": 1.850654426650097, + "grad_norm": 1.708866834640503, + "learning_rate": 2.902233900341478e-07, + "loss": 3.1804, + "step": 23084 + }, + { + "epoch": 1.85073460143112, + "grad_norm": 1.843278169631958, + "learning_rate": 2.8991292205188484e-07, + "loss": 3.1967, + "step": 23085 + }, + { + "epoch": 1.8508147762121425, + "grad_norm": 0.6621363162994385, + "learning_rate": 2.896026177782685e-07, + "loss": 2.4462, + "step": 23086 + }, + { + "epoch": 1.8508949509931651, + "grad_norm": 1.7035789489746094, + "learning_rate": 2.8929247721853016e-07, + "loss": 3.1477, + "step": 23087 + }, + { + "epoch": 1.8509751257741878, + "grad_norm": 1.7124838829040527, + "learning_rate": 2.8898250037789676e-07, + "loss": 3.1355, + "step": 23088 + }, + { + "epoch": 1.8510553005552104, + "grad_norm": 1.7926733493804932, + "learning_rate": 2.8867268726159745e-07, + "loss": 3.1965, + "step": 23089 + }, + { + "epoch": 1.851135475336233, + "grad_norm": 1.8514760732650757, + "learning_rate": 2.8836303787485255e-07, + "loss": 3.1256, + "step": 23090 + }, + { + "epoch": 1.8512156501172556, + "grad_norm": 1.830434799194336, + "learning_rate": 2.8805355222288556e-07, + "loss": 3.1852, + "step": 23091 + }, + { + "epoch": 1.8512958248982783, + "grad_norm": 1.854285478591919, + "learning_rate": 2.877442303109124e-07, + "loss": 3.2039, + "step": 23092 + }, + { + "epoch": 1.8513759996793009, + "grad_norm": 1.8815475702285767, + "learning_rate": 2.874350721441488e-07, + "loss": 3.285, + "step": 23093 + }, + { + "epoch": 1.8514561744603235, + "grad_norm": 1.8749455213546753, + "learning_rate": 2.871260777278062e-07, + "loss": 3.2878, + "step": 23094 + }, + { + "epoch": 1.8515363492413461, + "grad_norm": 1.7759345769882202, + "learning_rate": 2.868172470670949e-07, + "loss": 3.0867, + "step": 23095 + }, + { + "epoch": 1.8516165240223688, + "grad_norm": 1.763675332069397, + "learning_rate": 2.8650858016722185e-07, + "loss": 3.1248, + "step": 23096 + }, + { + "epoch": 1.8516966988033914, + "grad_norm": 1.7443257570266724, + "learning_rate": 2.862000770333917e-07, + "loss": 3.1207, + "step": 23097 + }, + { + "epoch": 1.851776873584414, + "grad_norm": 1.7926660776138306, + "learning_rate": 2.858917376708037e-07, + "loss": 3.0625, + "step": 23098 + }, + { + "epoch": 1.8518570483654366, + "grad_norm": 0.6477730870246887, + "learning_rate": 2.8558356208465697e-07, + "loss": 2.4562, + "step": 23099 + }, + { + "epoch": 1.8519372231464593, + "grad_norm": 1.7876181602478027, + "learning_rate": 2.8527555028014744e-07, + "loss": 3.2172, + "step": 23100 + }, + { + "epoch": 1.8520173979274819, + "grad_norm": 1.7353453636169434, + "learning_rate": 2.8496770226246975e-07, + "loss": 3.1519, + "step": 23101 + }, + { + "epoch": 1.8520975727085045, + "grad_norm": 0.6424052119255066, + "learning_rate": 2.846600180368131e-07, + "loss": 2.4661, + "step": 23102 + }, + { + "epoch": 1.8521777474895271, + "grad_norm": 1.6753231287002563, + "learning_rate": 2.8435249760836223e-07, + "loss": 3.1323, + "step": 23103 + }, + { + "epoch": 1.8522579222705498, + "grad_norm": 1.7887210845947266, + "learning_rate": 2.840451409823053e-07, + "loss": 3.2581, + "step": 23104 + }, + { + "epoch": 1.8523380970515724, + "grad_norm": 1.7184548377990723, + "learning_rate": 2.8373794816382363e-07, + "loss": 3.1245, + "step": 23105 + }, + { + "epoch": 1.8524182718325952, + "grad_norm": 1.678234577178955, + "learning_rate": 2.834309191580953e-07, + "loss": 3.1836, + "step": 23106 + }, + { + "epoch": 1.8524984466136178, + "grad_norm": 1.7649112939834595, + "learning_rate": 2.831240539702973e-07, + "loss": 3.067, + "step": 23107 + }, + { + "epoch": 1.8525786213946405, + "grad_norm": 1.7325023412704468, + "learning_rate": 2.8281735260560327e-07, + "loss": 3.0896, + "step": 23108 + }, + { + "epoch": 1.852658796175663, + "grad_norm": 2.0052273273468018, + "learning_rate": 2.8251081506918353e-07, + "loss": 3.1521, + "step": 23109 + }, + { + "epoch": 1.8527389709566857, + "grad_norm": 1.7603018283843994, + "learning_rate": 2.8220444136620727e-07, + "loss": 3.1266, + "step": 23110 + }, + { + "epoch": 1.8528191457377083, + "grad_norm": 1.7611957788467407, + "learning_rate": 2.8189823150183816e-07, + "loss": 3.0853, + "step": 23111 + }, + { + "epoch": 1.852899320518731, + "grad_norm": 1.7327688932418823, + "learning_rate": 2.81592185481242e-07, + "loss": 3.2184, + "step": 23112 + }, + { + "epoch": 1.8529794952997536, + "grad_norm": 2.025933027267456, + "learning_rate": 2.8128630330957475e-07, + "loss": 3.182, + "step": 23113 + }, + { + "epoch": 1.8530596700807762, + "grad_norm": 1.8033102750778198, + "learning_rate": 2.8098058499199556e-07, + "loss": 3.1434, + "step": 23114 + }, + { + "epoch": 1.8531398448617988, + "grad_norm": 1.7418097257614136, + "learning_rate": 2.806750305336581e-07, + "loss": 3.1202, + "step": 23115 + }, + { + "epoch": 1.8532200196428215, + "grad_norm": 1.7942246198654175, + "learning_rate": 2.803696399397149e-07, + "loss": 3.1624, + "step": 23116 + }, + { + "epoch": 1.853300194423844, + "grad_norm": 1.7560304403305054, + "learning_rate": 2.800644132153141e-07, + "loss": 3.1187, + "step": 23117 + }, + { + "epoch": 1.8533803692048667, + "grad_norm": 1.7980196475982666, + "learning_rate": 2.7975935036560153e-07, + "loss": 3.0997, + "step": 23118 + }, + { + "epoch": 1.8534605439858893, + "grad_norm": 1.7686957120895386, + "learning_rate": 2.794544513957198e-07, + "loss": 3.2256, + "step": 23119 + }, + { + "epoch": 1.853540718766912, + "grad_norm": 1.827730655670166, + "learning_rate": 2.7914971631081036e-07, + "loss": 3.1866, + "step": 23120 + }, + { + "epoch": 1.8536208935479346, + "grad_norm": 1.6939563751220703, + "learning_rate": 2.788451451160112e-07, + "loss": 3.1843, + "step": 23121 + }, + { + "epoch": 1.8537010683289572, + "grad_norm": 1.7566999197006226, + "learning_rate": 2.785407378164573e-07, + "loss": 3.0131, + "step": 23122 + }, + { + "epoch": 1.8537812431099798, + "grad_norm": 1.7006139755249023, + "learning_rate": 2.7823649441727995e-07, + "loss": 3.1554, + "step": 23123 + }, + { + "epoch": 1.8538614178910024, + "grad_norm": 1.7873109579086304, + "learning_rate": 2.7793241492360845e-07, + "loss": 3.0518, + "step": 23124 + }, + { + "epoch": 1.853941592672025, + "grad_norm": 1.8875380754470825, + "learning_rate": 2.776284993405698e-07, + "loss": 3.2522, + "step": 23125 + }, + { + "epoch": 1.8540217674530477, + "grad_norm": 1.7775403261184692, + "learning_rate": 2.7732474767328877e-07, + "loss": 3.2042, + "step": 23126 + }, + { + "epoch": 1.8541019422340703, + "grad_norm": 1.7773655652999878, + "learning_rate": 2.770211599268868e-07, + "loss": 3.0151, + "step": 23127 + }, + { + "epoch": 1.854182117015093, + "grad_norm": 1.8269113302230835, + "learning_rate": 2.767177361064799e-07, + "loss": 3.2189, + "step": 23128 + }, + { + "epoch": 1.8542622917961156, + "grad_norm": 1.9027774333953857, + "learning_rate": 2.764144762171861e-07, + "loss": 3.1873, + "step": 23129 + }, + { + "epoch": 1.8543424665771382, + "grad_norm": 1.7766549587249756, + "learning_rate": 2.7611138026411687e-07, + "loss": 3.2288, + "step": 23130 + }, + { + "epoch": 1.8544226413581608, + "grad_norm": 1.8038159608840942, + "learning_rate": 2.7580844825238264e-07, + "loss": 3.1456, + "step": 23131 + }, + { + "epoch": 1.8545028161391834, + "grad_norm": 1.7567341327667236, + "learning_rate": 2.7550568018709143e-07, + "loss": 3.2766, + "step": 23132 + }, + { + "epoch": 1.854582990920206, + "grad_norm": 1.7482967376708984, + "learning_rate": 2.752030760733471e-07, + "loss": 3.2303, + "step": 23133 + }, + { + "epoch": 1.8546631657012287, + "grad_norm": 1.9156321287155151, + "learning_rate": 2.749006359162509e-07, + "loss": 3.1537, + "step": 23134 + }, + { + "epoch": 1.8547433404822513, + "grad_norm": 1.7425724267959595, + "learning_rate": 2.7459835972090234e-07, + "loss": 3.0911, + "step": 23135 + }, + { + "epoch": 1.854823515263274, + "grad_norm": 1.7738957405090332, + "learning_rate": 2.7429624749239826e-07, + "loss": 3.1083, + "step": 23136 + }, + { + "epoch": 1.8549036900442966, + "grad_norm": 1.754878044128418, + "learning_rate": 2.739942992358324e-07, + "loss": 3.194, + "step": 23137 + }, + { + "epoch": 1.8549838648253192, + "grad_norm": 1.7777968645095825, + "learning_rate": 2.7369251495629525e-07, + "loss": 3.2241, + "step": 23138 + }, + { + "epoch": 1.8550640396063418, + "grad_norm": 1.8095428943634033, + "learning_rate": 2.733908946588726e-07, + "loss": 3.1323, + "step": 23139 + }, + { + "epoch": 1.8551442143873644, + "grad_norm": 1.698551893234253, + "learning_rate": 2.730894383486526e-07, + "loss": 3.188, + "step": 23140 + }, + { + "epoch": 1.855224389168387, + "grad_norm": 1.7893885374069214, + "learning_rate": 2.727881460307158e-07, + "loss": 3.1096, + "step": 23141 + }, + { + "epoch": 1.8553045639494097, + "grad_norm": 1.7400493621826172, + "learning_rate": 2.7248701771014465e-07, + "loss": 3.0645, + "step": 23142 + }, + { + "epoch": 1.8553847387304323, + "grad_norm": 1.7766647338867188, + "learning_rate": 2.7218605339201175e-07, + "loss": 3.2535, + "step": 23143 + }, + { + "epoch": 1.855464913511455, + "grad_norm": 1.8192957639694214, + "learning_rate": 2.7188525308139533e-07, + "loss": 3.1934, + "step": 23144 + }, + { + "epoch": 1.8555450882924776, + "grad_norm": 1.7948838472366333, + "learning_rate": 2.7158461678336356e-07, + "loss": 3.1409, + "step": 23145 + }, + { + "epoch": 1.8556252630735002, + "grad_norm": 1.8487459421157837, + "learning_rate": 2.712841445029868e-07, + "loss": 3.2881, + "step": 23146 + }, + { + "epoch": 1.8557054378545228, + "grad_norm": 1.8667031526565552, + "learning_rate": 2.709838362453321e-07, + "loss": 3.2433, + "step": 23147 + }, + { + "epoch": 1.8557856126355454, + "grad_norm": 1.6078497171401978, + "learning_rate": 2.7068369201545987e-07, + "loss": 3.1588, + "step": 23148 + }, + { + "epoch": 1.855865787416568, + "grad_norm": 1.7504445314407349, + "learning_rate": 2.703837118184316e-07, + "loss": 3.1615, + "step": 23149 + }, + { + "epoch": 1.8559459621975907, + "grad_norm": 1.7162439823150635, + "learning_rate": 2.700838956593044e-07, + "loss": 3.0573, + "step": 23150 + }, + { + "epoch": 1.8560261369786133, + "grad_norm": 1.783660888671875, + "learning_rate": 2.6978424354313416e-07, + "loss": 3.214, + "step": 23151 + }, + { + "epoch": 1.856106311759636, + "grad_norm": 1.7066301107406616, + "learning_rate": 2.6948475547497356e-07, + "loss": 3.2181, + "step": 23152 + }, + { + "epoch": 1.8561864865406585, + "grad_norm": 1.8349014520645142, + "learning_rate": 2.691854314598685e-07, + "loss": 3.2593, + "step": 23153 + }, + { + "epoch": 1.8562666613216812, + "grad_norm": 1.7423800230026245, + "learning_rate": 2.6888627150286837e-07, + "loss": 3.1455, + "step": 23154 + }, + { + "epoch": 1.8563468361027038, + "grad_norm": 1.8420631885528564, + "learning_rate": 2.6858727560901686e-07, + "loss": 3.3243, + "step": 23155 + }, + { + "epoch": 1.8564270108837264, + "grad_norm": 1.7087290287017822, + "learning_rate": 2.6828844378335327e-07, + "loss": 3.1747, + "step": 23156 + }, + { + "epoch": 1.856507185664749, + "grad_norm": 1.7567048072814941, + "learning_rate": 2.67989776030918e-07, + "loss": 3.2247, + "step": 23157 + }, + { + "epoch": 1.8565873604457717, + "grad_norm": 1.706577181816101, + "learning_rate": 2.6769127235674375e-07, + "loss": 3.136, + "step": 23158 + }, + { + "epoch": 1.8566675352267943, + "grad_norm": 1.6759072542190552, + "learning_rate": 2.6739293276586534e-07, + "loss": 3.1171, + "step": 23159 + }, + { + "epoch": 1.856747710007817, + "grad_norm": 1.7826652526855469, + "learning_rate": 2.6709475726331204e-07, + "loss": 3.1808, + "step": 23160 + }, + { + "epoch": 1.8568278847888395, + "grad_norm": 1.7110720872879028, + "learning_rate": 2.667967458541099e-07, + "loss": 3.1567, + "step": 23161 + }, + { + "epoch": 1.8569080595698622, + "grad_norm": 1.8271684646606445, + "learning_rate": 2.664988985432859e-07, + "loss": 3.2723, + "step": 23162 + }, + { + "epoch": 1.8569882343508848, + "grad_norm": 1.7148743867874146, + "learning_rate": 2.662012153358595e-07, + "loss": 3.1573, + "step": 23163 + }, + { + "epoch": 1.8570684091319074, + "grad_norm": 1.8948757648468018, + "learning_rate": 2.6590369623685e-07, + "loss": 3.2004, + "step": 23164 + }, + { + "epoch": 1.85714858391293, + "grad_norm": 1.730147123336792, + "learning_rate": 2.6560634125127213e-07, + "loss": 3.1222, + "step": 23165 + }, + { + "epoch": 1.8572287586939527, + "grad_norm": 1.7637776136398315, + "learning_rate": 2.6530915038414207e-07, + "loss": 3.1595, + "step": 23166 + }, + { + "epoch": 1.8573089334749755, + "grad_norm": 0.6288987398147583, + "learning_rate": 2.6501212364047013e-07, + "loss": 2.4579, + "step": 23167 + }, + { + "epoch": 1.8573891082559981, + "grad_norm": 1.851726770401001, + "learning_rate": 2.6471526102526124e-07, + "loss": 3.1541, + "step": 23168 + }, + { + "epoch": 1.8574692830370207, + "grad_norm": 1.7803504467010498, + "learning_rate": 2.644185625435214e-07, + "loss": 3.2264, + "step": 23169 + }, + { + "epoch": 1.8575494578180434, + "grad_norm": 1.7616124153137207, + "learning_rate": 2.641220282002543e-07, + "loss": 3.1152, + "step": 23170 + }, + { + "epoch": 1.857629632599066, + "grad_norm": 1.7354451417922974, + "learning_rate": 2.638256580004583e-07, + "loss": 3.104, + "step": 23171 + }, + { + "epoch": 1.8577098073800886, + "grad_norm": 1.7515642642974854, + "learning_rate": 2.635294519491316e-07, + "loss": 3.1665, + "step": 23172 + }, + { + "epoch": 1.8577899821611112, + "grad_norm": 1.8546134233474731, + "learning_rate": 2.632334100512657e-07, + "loss": 3.1769, + "step": 23173 + }, + { + "epoch": 1.8578701569421339, + "grad_norm": 0.6628904938697815, + "learning_rate": 2.6293753231185216e-07, + "loss": 2.4595, + "step": 23174 + }, + { + "epoch": 1.8579503317231565, + "grad_norm": 1.7130719423294067, + "learning_rate": 2.626418187358815e-07, + "loss": 3.1448, + "step": 23175 + }, + { + "epoch": 1.8580305065041791, + "grad_norm": 1.7082879543304443, + "learning_rate": 2.623462693283374e-07, + "loss": 3.1001, + "step": 23176 + }, + { + "epoch": 1.8581106812852017, + "grad_norm": 1.8024022579193115, + "learning_rate": 2.6205088409420375e-07, + "loss": 3.1562, + "step": 23177 + }, + { + "epoch": 1.8581908560662244, + "grad_norm": 1.8559746742248535, + "learning_rate": 2.617556630384599e-07, + "loss": 3.2372, + "step": 23178 + }, + { + "epoch": 1.858271030847247, + "grad_norm": 0.640274167060852, + "learning_rate": 2.6146060616608404e-07, + "loss": 2.458, + "step": 23179 + }, + { + "epoch": 1.8583512056282696, + "grad_norm": 1.752250075340271, + "learning_rate": 2.611657134820489e-07, + "loss": 3.2074, + "step": 23180 + }, + { + "epoch": 1.8584313804092922, + "grad_norm": 1.7664872407913208, + "learning_rate": 2.6087098499132824e-07, + "loss": 3.1427, + "step": 23181 + }, + { + "epoch": 1.8585115551903149, + "grad_norm": 1.7179806232452393, + "learning_rate": 2.6057642069889144e-07, + "loss": 3.1548, + "step": 23182 + }, + { + "epoch": 1.8585917299713375, + "grad_norm": 1.7341103553771973, + "learning_rate": 2.6028202060970233e-07, + "loss": 3.0309, + "step": 23183 + }, + { + "epoch": 1.85867190475236, + "grad_norm": 0.6554058194160461, + "learning_rate": 2.599877847287258e-07, + "loss": 2.465, + "step": 23184 + }, + { + "epoch": 1.8587520795333827, + "grad_norm": 1.7187769412994385, + "learning_rate": 2.596937130609223e-07, + "loss": 3.1311, + "step": 23185 + }, + { + "epoch": 1.8588322543144054, + "grad_norm": 1.7582453489303589, + "learning_rate": 2.593998056112501e-07, + "loss": 3.1884, + "step": 23186 + }, + { + "epoch": 1.8589124290954282, + "grad_norm": 1.7522944211959839, + "learning_rate": 2.591060623846653e-07, + "loss": 3.1712, + "step": 23187 + }, + { + "epoch": 1.8589926038764508, + "grad_norm": 1.731398582458496, + "learning_rate": 2.588124833861183e-07, + "loss": 3.0605, + "step": 23188 + }, + { + "epoch": 1.8590727786574734, + "grad_norm": 1.7274346351623535, + "learning_rate": 2.5851906862055965e-07, + "loss": 3.1223, + "step": 23189 + }, + { + "epoch": 1.859152953438496, + "grad_norm": 0.6551118493080139, + "learning_rate": 2.5822581809293643e-07, + "loss": 2.4683, + "step": 23190 + }, + { + "epoch": 1.8592331282195187, + "grad_norm": 1.7705662250518799, + "learning_rate": 2.5793273180819256e-07, + "loss": 3.2112, + "step": 23191 + }, + { + "epoch": 1.8593133030005413, + "grad_norm": 1.693537712097168, + "learning_rate": 2.576398097712707e-07, + "loss": 3.1728, + "step": 23192 + }, + { + "epoch": 1.859393477781564, + "grad_norm": 1.7797913551330566, + "learning_rate": 2.5734705198710685e-07, + "loss": 3.1498, + "step": 23193 + }, + { + "epoch": 1.8594736525625866, + "grad_norm": 1.7885743379592896, + "learning_rate": 2.5705445846063824e-07, + "loss": 3.2211, + "step": 23194 + }, + { + "epoch": 1.8595538273436092, + "grad_norm": 1.8383677005767822, + "learning_rate": 2.567620291967976e-07, + "loss": 3.183, + "step": 23195 + }, + { + "epoch": 1.8596340021246318, + "grad_norm": 1.853107213973999, + "learning_rate": 2.5646976420051537e-07, + "loss": 3.2404, + "step": 23196 + }, + { + "epoch": 1.8597141769056544, + "grad_norm": 0.6709242463111877, + "learning_rate": 2.5617766347671877e-07, + "loss": 2.4719, + "step": 23197 + }, + { + "epoch": 1.859794351686677, + "grad_norm": 1.8366738557815552, + "learning_rate": 2.558857270303328e-07, + "loss": 3.2476, + "step": 23198 + }, + { + "epoch": 1.8598745264676997, + "grad_norm": 1.7249876260757446, + "learning_rate": 2.5559395486628005e-07, + "loss": 3.1508, + "step": 23199 + }, + { + "epoch": 1.8599547012487223, + "grad_norm": 1.8966655731201172, + "learning_rate": 2.5530234698947774e-07, + "loss": 3.1632, + "step": 23200 + }, + { + "epoch": 1.860034876029745, + "grad_norm": 1.7547581195831299, + "learning_rate": 2.5501090340484427e-07, + "loss": 3.2005, + "step": 23201 + }, + { + "epoch": 1.8601150508107676, + "grad_norm": 1.7048420906066895, + "learning_rate": 2.547196241172922e-07, + "loss": 3.145, + "step": 23202 + }, + { + "epoch": 1.8601952255917902, + "grad_norm": 1.7982741594314575, + "learning_rate": 2.5442850913173334e-07, + "loss": 3.1139, + "step": 23203 + }, + { + "epoch": 1.8602754003728128, + "grad_norm": 0.6614364385604858, + "learning_rate": 2.5413755845307364e-07, + "loss": 2.4556, + "step": 23204 + }, + { + "epoch": 1.8603555751538354, + "grad_norm": 1.7059073448181152, + "learning_rate": 2.5384677208622145e-07, + "loss": 3.0848, + "step": 23205 + }, + { + "epoch": 1.860435749934858, + "grad_norm": 0.6371645331382751, + "learning_rate": 2.5355615003607726e-07, + "loss": 2.4679, + "step": 23206 + }, + { + "epoch": 1.8605159247158807, + "grad_norm": 1.6994167566299438, + "learning_rate": 2.532656923075405e-07, + "loss": 3.1033, + "step": 23207 + }, + { + "epoch": 1.8605960994969033, + "grad_norm": 1.678976058959961, + "learning_rate": 2.529753989055106e-07, + "loss": 3.095, + "step": 23208 + }, + { + "epoch": 1.860676274277926, + "grad_norm": 1.8230599164962769, + "learning_rate": 2.526852698348803e-07, + "loss": 3.1898, + "step": 23209 + }, + { + "epoch": 1.8607564490589485, + "grad_norm": 1.8671563863754272, + "learning_rate": 2.5239530510054014e-07, + "loss": 3.1918, + "step": 23210 + }, + { + "epoch": 1.8608366238399712, + "grad_norm": 1.84613835811615, + "learning_rate": 2.521055047073795e-07, + "loss": 3.1821, + "step": 23211 + }, + { + "epoch": 1.8609167986209938, + "grad_norm": 1.8128374814987183, + "learning_rate": 2.518158686602845e-07, + "loss": 3.1421, + "step": 23212 + }, + { + "epoch": 1.8609969734020164, + "grad_norm": 1.736451268196106, + "learning_rate": 2.5152639696414015e-07, + "loss": 3.2009, + "step": 23213 + }, + { + "epoch": 1.861077148183039, + "grad_norm": 1.6817741394042969, + "learning_rate": 2.5123708962382367e-07, + "loss": 3.2227, + "step": 23214 + }, + { + "epoch": 1.8611573229640617, + "grad_norm": 0.6577561497688293, + "learning_rate": 2.509479466442144e-07, + "loss": 2.4729, + "step": 23215 + }, + { + "epoch": 1.8612374977450843, + "grad_norm": 0.6552508473396301, + "learning_rate": 2.506589680301863e-07, + "loss": 2.4493, + "step": 23216 + }, + { + "epoch": 1.861317672526107, + "grad_norm": 1.8307374715805054, + "learning_rate": 2.5037015378661214e-07, + "loss": 3.2292, + "step": 23217 + }, + { + "epoch": 1.8613978473071295, + "grad_norm": 1.9322612285614014, + "learning_rate": 2.500815039183624e-07, + "loss": 3.1929, + "step": 23218 + }, + { + "epoch": 1.8614780220881522, + "grad_norm": 1.767069697380066, + "learning_rate": 2.49793018430301e-07, + "loss": 3.1676, + "step": 23219 + }, + { + "epoch": 1.8615581968691748, + "grad_norm": 1.787987232208252, + "learning_rate": 2.49504697327293e-07, + "loss": 3.1788, + "step": 23220 + }, + { + "epoch": 1.8616383716501974, + "grad_norm": 1.8195449113845825, + "learning_rate": 2.492165406141989e-07, + "loss": 3.2021, + "step": 23221 + }, + { + "epoch": 1.86171854643122, + "grad_norm": 1.8889708518981934, + "learning_rate": 2.489285482958781e-07, + "loss": 3.1992, + "step": 23222 + }, + { + "epoch": 1.8617987212122427, + "grad_norm": 1.8433833122253418, + "learning_rate": 2.486407203771868e-07, + "loss": 3.1754, + "step": 23223 + }, + { + "epoch": 1.8618788959932653, + "grad_norm": 1.6862266063690186, + "learning_rate": 2.483530568629744e-07, + "loss": 3.2463, + "step": 23224 + }, + { + "epoch": 1.861959070774288, + "grad_norm": 1.8287644386291504, + "learning_rate": 2.480655577580926e-07, + "loss": 3.2008, + "step": 23225 + }, + { + "epoch": 1.8620392455553105, + "grad_norm": 1.8304896354675293, + "learning_rate": 2.4777822306738864e-07, + "loss": 3.2134, + "step": 23226 + }, + { + "epoch": 1.8621194203363332, + "grad_norm": 1.7630077600479126, + "learning_rate": 2.474910527957075e-07, + "loss": 3.1272, + "step": 23227 + }, + { + "epoch": 1.8621995951173558, + "grad_norm": 1.7206848859786987, + "learning_rate": 2.472040469478909e-07, + "loss": 3.1006, + "step": 23228 + }, + { + "epoch": 1.8622797698983784, + "grad_norm": 1.7915993928909302, + "learning_rate": 2.4691720552877494e-07, + "loss": 3.1726, + "step": 23229 + }, + { + "epoch": 1.862359944679401, + "grad_norm": 1.6913264989852905, + "learning_rate": 2.4663052854319915e-07, + "loss": 3.1387, + "step": 23230 + }, + { + "epoch": 1.8624401194604236, + "grad_norm": 1.8177998065948486, + "learning_rate": 2.46344015995994e-07, + "loss": 3.1587, + "step": 23231 + }, + { + "epoch": 1.8625202942414463, + "grad_norm": 0.6423341035842896, + "learning_rate": 2.4605766789199126e-07, + "loss": 2.4724, + "step": 23232 + }, + { + "epoch": 1.862600469022469, + "grad_norm": 1.7934831380844116, + "learning_rate": 2.457714842360204e-07, + "loss": 3.0985, + "step": 23233 + }, + { + "epoch": 1.8626806438034915, + "grad_norm": 1.8231728076934814, + "learning_rate": 2.4548546503290307e-07, + "loss": 3.1892, + "step": 23234 + }, + { + "epoch": 1.8627608185845141, + "grad_norm": 1.8337637186050415, + "learning_rate": 2.451996102874632e-07, + "loss": 3.1749, + "step": 23235 + }, + { + "epoch": 1.8628409933655368, + "grad_norm": 1.8300846815109253, + "learning_rate": 2.449139200045192e-07, + "loss": 3.2607, + "step": 23236 + }, + { + "epoch": 1.8629211681465594, + "grad_norm": 1.7959076166152954, + "learning_rate": 2.446283941888894e-07, + "loss": 3.1759, + "step": 23237 + }, + { + "epoch": 1.863001342927582, + "grad_norm": 0.6719410419464111, + "learning_rate": 2.4434303284538776e-07, + "loss": 2.4647, + "step": 23238 + }, + { + "epoch": 1.8630815177086046, + "grad_norm": 1.7830100059509277, + "learning_rate": 2.440578359788226e-07, + "loss": 3.2149, + "step": 23239 + }, + { + "epoch": 1.8631616924896273, + "grad_norm": 1.782101035118103, + "learning_rate": 2.437728035940046e-07, + "loss": 3.201, + "step": 23240 + }, + { + "epoch": 1.8632418672706499, + "grad_norm": 1.7839105129241943, + "learning_rate": 2.434879356957387e-07, + "loss": 3.3067, + "step": 23241 + }, + { + "epoch": 1.8633220420516725, + "grad_norm": 1.7516535520553589, + "learning_rate": 2.4320323228882667e-07, + "loss": 3.1609, + "step": 23242 + }, + { + "epoch": 1.8634022168326951, + "grad_norm": 1.8841508626937866, + "learning_rate": 2.429186933780714e-07, + "loss": 3.2871, + "step": 23243 + }, + { + "epoch": 1.8634823916137178, + "grad_norm": 1.8262442350387573, + "learning_rate": 2.426343189682678e-07, + "loss": 3.082, + "step": 23244 + }, + { + "epoch": 1.8635625663947404, + "grad_norm": 1.712062954902649, + "learning_rate": 2.4235010906421e-07, + "loss": 3.1965, + "step": 23245 + }, + { + "epoch": 1.863642741175763, + "grad_norm": 1.7280453443527222, + "learning_rate": 2.4206606367069065e-07, + "loss": 3.0502, + "step": 23246 + }, + { + "epoch": 1.8637229159567856, + "grad_norm": 1.7199517488479614, + "learning_rate": 2.417821827924982e-07, + "loss": 3.1333, + "step": 23247 + }, + { + "epoch": 1.8638030907378085, + "grad_norm": 1.793812870979309, + "learning_rate": 2.4149846643442e-07, + "loss": 3.2625, + "step": 23248 + }, + { + "epoch": 1.863883265518831, + "grad_norm": 1.8286168575286865, + "learning_rate": 2.412149146012377e-07, + "loss": 3.2174, + "step": 23249 + }, + { + "epoch": 1.8639634402998537, + "grad_norm": 1.731999158859253, + "learning_rate": 2.4093152729773193e-07, + "loss": 3.1642, + "step": 23250 + }, + { + "epoch": 1.8640436150808763, + "grad_norm": 1.8044923543930054, + "learning_rate": 2.4064830452868225e-07, + "loss": 3.2021, + "step": 23251 + }, + { + "epoch": 1.864123789861899, + "grad_norm": 1.7296072244644165, + "learning_rate": 2.403652462988615e-07, + "loss": 3.1431, + "step": 23252 + }, + { + "epoch": 1.8642039646429216, + "grad_norm": 1.9197711944580078, + "learning_rate": 2.4008235261304467e-07, + "loss": 3.1641, + "step": 23253 + }, + { + "epoch": 1.8642841394239442, + "grad_norm": 1.8012192249298096, + "learning_rate": 2.397996234759992e-07, + "loss": 3.1515, + "step": 23254 + }, + { + "epoch": 1.8643643142049668, + "grad_norm": 1.6825710535049438, + "learning_rate": 2.3951705889249224e-07, + "loss": 3.1018, + "step": 23255 + }, + { + "epoch": 1.8644444889859895, + "grad_norm": 1.8592541217803955, + "learning_rate": 2.392346588672867e-07, + "loss": 3.2467, + "step": 23256 + }, + { + "epoch": 1.864524663767012, + "grad_norm": 1.8734098672866821, + "learning_rate": 2.3895242340514545e-07, + "loss": 3.2122, + "step": 23257 + }, + { + "epoch": 1.8646048385480347, + "grad_norm": 1.7670313119888306, + "learning_rate": 2.3867035251082695e-07, + "loss": 3.0962, + "step": 23258 + }, + { + "epoch": 1.8646850133290573, + "grad_norm": 1.7594937086105347, + "learning_rate": 2.3838844618908508e-07, + "loss": 3.0664, + "step": 23259 + }, + { + "epoch": 1.86476518811008, + "grad_norm": 1.81038498878479, + "learning_rate": 2.3810670444467386e-07, + "loss": 3.1854, + "step": 23260 + }, + { + "epoch": 1.8648453628911026, + "grad_norm": 1.7921446561813354, + "learning_rate": 2.3782512728234285e-07, + "loss": 3.1553, + "step": 23261 + }, + { + "epoch": 1.8649255376721252, + "grad_norm": 1.8499433994293213, + "learning_rate": 2.3754371470684047e-07, + "loss": 3.2114, + "step": 23262 + }, + { + "epoch": 1.8650057124531478, + "grad_norm": 1.8161455392837524, + "learning_rate": 2.3726246672291064e-07, + "loss": 3.178, + "step": 23263 + }, + { + "epoch": 1.8650858872341705, + "grad_norm": 1.8160301446914673, + "learning_rate": 2.3698138333529408e-07, + "loss": 3.127, + "step": 23264 + }, + { + "epoch": 1.865166062015193, + "grad_norm": 1.6932055950164795, + "learning_rate": 2.367004645487303e-07, + "loss": 3.2084, + "step": 23265 + }, + { + "epoch": 1.8652462367962157, + "grad_norm": 1.821923017501831, + "learning_rate": 2.3641971036795665e-07, + "loss": 3.1974, + "step": 23266 + }, + { + "epoch": 1.8653264115772383, + "grad_norm": 1.7699342966079712, + "learning_rate": 2.3613912079770485e-07, + "loss": 3.177, + "step": 23267 + }, + { + "epoch": 1.865406586358261, + "grad_norm": 1.80558443069458, + "learning_rate": 2.358586958427078e-07, + "loss": 3.1874, + "step": 23268 + }, + { + "epoch": 1.8654867611392838, + "grad_norm": 1.7443630695343018, + "learning_rate": 2.3557843550769067e-07, + "loss": 3.1766, + "step": 23269 + }, + { + "epoch": 1.8655669359203064, + "grad_norm": 1.7076374292373657, + "learning_rate": 2.352983397973796e-07, + "loss": 3.1582, + "step": 23270 + }, + { + "epoch": 1.865647110701329, + "grad_norm": 0.6531334519386292, + "learning_rate": 2.350184087164975e-07, + "loss": 2.4689, + "step": 23271 + }, + { + "epoch": 1.8657272854823517, + "grad_norm": 1.7306183576583862, + "learning_rate": 2.3473864226976395e-07, + "loss": 3.1184, + "step": 23272 + }, + { + "epoch": 1.8658074602633743, + "grad_norm": 0.6518000960350037, + "learning_rate": 2.3445904046189515e-07, + "loss": 2.4591, + "step": 23273 + }, + { + "epoch": 1.865887635044397, + "grad_norm": 1.7384529113769531, + "learning_rate": 2.3417960329760514e-07, + "loss": 3.0389, + "step": 23274 + }, + { + "epoch": 1.8659678098254195, + "grad_norm": 1.7911624908447266, + "learning_rate": 2.3390033078160458e-07, + "loss": 3.1677, + "step": 23275 + }, + { + "epoch": 1.8660479846064422, + "grad_norm": 1.8186546564102173, + "learning_rate": 2.33621222918603e-07, + "loss": 3.1331, + "step": 23276 + }, + { + "epoch": 1.8661281593874648, + "grad_norm": 1.733022928237915, + "learning_rate": 2.333422797133056e-07, + "loss": 3.2248, + "step": 23277 + }, + { + "epoch": 1.8662083341684874, + "grad_norm": 0.636273980140686, + "learning_rate": 2.3306350117041632e-07, + "loss": 2.458, + "step": 23278 + }, + { + "epoch": 1.86628850894951, + "grad_norm": 1.743173599243164, + "learning_rate": 2.3278488729463257e-07, + "loss": 3.1219, + "step": 23279 + }, + { + "epoch": 1.8663686837305327, + "grad_norm": 1.8414753675460815, + "learning_rate": 2.3250643809065386e-07, + "loss": 3.1356, + "step": 23280 + }, + { + "epoch": 1.8664488585115553, + "grad_norm": 1.6694121360778809, + "learning_rate": 2.322281535631743e-07, + "loss": 3.2225, + "step": 23281 + }, + { + "epoch": 1.866529033292578, + "grad_norm": 1.8209290504455566, + "learning_rate": 2.3195003371688562e-07, + "loss": 3.1848, + "step": 23282 + }, + { + "epoch": 1.8666092080736005, + "grad_norm": 1.7678799629211426, + "learning_rate": 2.3167207855647744e-07, + "loss": 3.1639, + "step": 23283 + }, + { + "epoch": 1.8666893828546232, + "grad_norm": 1.8016635179519653, + "learning_rate": 2.313942880866338e-07, + "loss": 3.1359, + "step": 23284 + }, + { + "epoch": 1.8667695576356458, + "grad_norm": 1.7157572507858276, + "learning_rate": 2.3111666231204089e-07, + "loss": 3.1201, + "step": 23285 + }, + { + "epoch": 1.8668497324166684, + "grad_norm": 1.848547101020813, + "learning_rate": 2.3083920123737723e-07, + "loss": 3.1145, + "step": 23286 + }, + { + "epoch": 1.866929907197691, + "grad_norm": 1.771019697189331, + "learning_rate": 2.3056190486732134e-07, + "loss": 3.1605, + "step": 23287 + }, + { + "epoch": 1.8670100819787137, + "grad_norm": 1.7065485715866089, + "learning_rate": 2.3028477320655052e-07, + "loss": 3.1606, + "step": 23288 + }, + { + "epoch": 1.8670902567597363, + "grad_norm": 1.7563570737838745, + "learning_rate": 2.300078062597344e-07, + "loss": 3.196, + "step": 23289 + }, + { + "epoch": 1.867170431540759, + "grad_norm": 1.8768619298934937, + "learning_rate": 2.2973100403154258e-07, + "loss": 3.3065, + "step": 23290 + }, + { + "epoch": 1.8672506063217815, + "grad_norm": 1.7181531190872192, + "learning_rate": 2.2945436652664354e-07, + "loss": 3.1825, + "step": 23291 + }, + { + "epoch": 1.8673307811028041, + "grad_norm": 1.8347728252410889, + "learning_rate": 2.2917789374969912e-07, + "loss": 3.1986, + "step": 23292 + }, + { + "epoch": 1.8674109558838268, + "grad_norm": 1.8004790544509888, + "learning_rate": 2.2890158570537447e-07, + "loss": 3.1121, + "step": 23293 + }, + { + "epoch": 1.8674911306648494, + "grad_norm": 1.762042760848999, + "learning_rate": 2.2862544239832364e-07, + "loss": 3.1526, + "step": 23294 + }, + { + "epoch": 1.867571305445872, + "grad_norm": 1.8220157623291016, + "learning_rate": 2.28349463833204e-07, + "loss": 3.1726, + "step": 23295 + }, + { + "epoch": 1.8676514802268946, + "grad_norm": 1.7895588874816895, + "learning_rate": 2.2807365001466853e-07, + "loss": 3.2244, + "step": 23296 + }, + { + "epoch": 1.8677316550079173, + "grad_norm": 1.772499442100525, + "learning_rate": 2.2779800094736682e-07, + "loss": 3.1022, + "step": 23297 + }, + { + "epoch": 1.86781182978894, + "grad_norm": 1.696852207183838, + "learning_rate": 2.2752251663594848e-07, + "loss": 3.0927, + "step": 23298 + }, + { + "epoch": 1.8678920045699625, + "grad_norm": 1.7856947183609009, + "learning_rate": 2.272471970850554e-07, + "loss": 3.1398, + "step": 23299 + }, + { + "epoch": 1.8679721793509851, + "grad_norm": 1.7933083772659302, + "learning_rate": 2.2697204229933045e-07, + "loss": 3.1865, + "step": 23300 + }, + { + "epoch": 1.8680523541320078, + "grad_norm": 1.7861371040344238, + "learning_rate": 2.266970522834122e-07, + "loss": 3.2206, + "step": 23301 + }, + { + "epoch": 1.8681325289130304, + "grad_norm": 1.7941946983337402, + "learning_rate": 2.2642222704193694e-07, + "loss": 3.1625, + "step": 23302 + }, + { + "epoch": 1.868212703694053, + "grad_norm": 1.6946674585342407, + "learning_rate": 2.2614756657953985e-07, + "loss": 3.0956, + "step": 23303 + }, + { + "epoch": 1.8682928784750756, + "grad_norm": 1.7522337436676025, + "learning_rate": 2.258730709008483e-07, + "loss": 3.2162, + "step": 23304 + }, + { + "epoch": 1.8683730532560983, + "grad_norm": 1.8107041120529175, + "learning_rate": 2.255987400104931e-07, + "loss": 3.1985, + "step": 23305 + }, + { + "epoch": 1.8684532280371209, + "grad_norm": 1.68313467502594, + "learning_rate": 2.2532457391309715e-07, + "loss": 3.0719, + "step": 23306 + }, + { + "epoch": 1.8685334028181435, + "grad_norm": 1.8386569023132324, + "learning_rate": 2.2505057261328566e-07, + "loss": 3.2166, + "step": 23307 + }, + { + "epoch": 1.8686135775991661, + "grad_norm": 1.7096272706985474, + "learning_rate": 2.2477673611567608e-07, + "loss": 3.0735, + "step": 23308 + }, + { + "epoch": 1.8686937523801888, + "grad_norm": 1.8855862617492676, + "learning_rate": 2.2450306442488467e-07, + "loss": 3.2332, + "step": 23309 + }, + { + "epoch": 1.8687739271612114, + "grad_norm": 1.822426438331604, + "learning_rate": 2.2422955754552667e-07, + "loss": 3.2512, + "step": 23310 + }, + { + "epoch": 1.868854101942234, + "grad_norm": 1.7307566404342651, + "learning_rate": 2.239562154822139e-07, + "loss": 3.2009, + "step": 23311 + }, + { + "epoch": 1.8689342767232566, + "grad_norm": 1.7513028383255005, + "learning_rate": 2.2368303823955273e-07, + "loss": 3.1459, + "step": 23312 + }, + { + "epoch": 1.8690144515042793, + "grad_norm": 1.7292810678482056, + "learning_rate": 2.2341002582215164e-07, + "loss": 3.1554, + "step": 23313 + }, + { + "epoch": 1.8690946262853019, + "grad_norm": 1.8664625883102417, + "learning_rate": 2.2313717823461034e-07, + "loss": 3.1877, + "step": 23314 + }, + { + "epoch": 1.8691748010663245, + "grad_norm": 1.8236467838287354, + "learning_rate": 2.2286449548153066e-07, + "loss": 3.2252, + "step": 23315 + }, + { + "epoch": 1.8692549758473471, + "grad_norm": 1.7224462032318115, + "learning_rate": 2.2259197756751004e-07, + "loss": 3.1942, + "step": 23316 + }, + { + "epoch": 1.8693351506283697, + "grad_norm": 1.6703195571899414, + "learning_rate": 2.2231962449714262e-07, + "loss": 3.1248, + "step": 23317 + }, + { + "epoch": 1.8694153254093924, + "grad_norm": 1.7940384149551392, + "learning_rate": 2.2204743627502135e-07, + "loss": 3.0951, + "step": 23318 + }, + { + "epoch": 1.869495500190415, + "grad_norm": 1.8417435884475708, + "learning_rate": 2.2177541290573368e-07, + "loss": 3.1462, + "step": 23319 + }, + { + "epoch": 1.8695756749714376, + "grad_norm": 1.7269737720489502, + "learning_rate": 2.2150355439386595e-07, + "loss": 3.2068, + "step": 23320 + }, + { + "epoch": 1.8696558497524602, + "grad_norm": 1.7778342962265015, + "learning_rate": 2.2123186074400227e-07, + "loss": 3.1042, + "step": 23321 + }, + { + "epoch": 1.8697360245334829, + "grad_norm": 1.720093846321106, + "learning_rate": 2.2096033196072343e-07, + "loss": 3.1858, + "step": 23322 + }, + { + "epoch": 1.8698161993145055, + "grad_norm": 1.8584617376327515, + "learning_rate": 2.2068896804860684e-07, + "loss": 3.2361, + "step": 23323 + }, + { + "epoch": 1.8698963740955281, + "grad_norm": 1.7576671838760376, + "learning_rate": 2.2041776901222667e-07, + "loss": 3.1324, + "step": 23324 + }, + { + "epoch": 1.8699765488765507, + "grad_norm": 1.7878053188323975, + "learning_rate": 2.20146734856157e-07, + "loss": 3.083, + "step": 23325 + }, + { + "epoch": 1.8700567236575734, + "grad_norm": 1.8548188209533691, + "learning_rate": 2.1987586558496642e-07, + "loss": 3.1642, + "step": 23326 + }, + { + "epoch": 1.870136898438596, + "grad_norm": 1.743775486946106, + "learning_rate": 2.1960516120322238e-07, + "loss": 3.2416, + "step": 23327 + }, + { + "epoch": 1.8702170732196186, + "grad_norm": 1.7971553802490234, + "learning_rate": 2.1933462171548902e-07, + "loss": 3.1739, + "step": 23328 + }, + { + "epoch": 1.8702972480006412, + "grad_norm": 1.7622427940368652, + "learning_rate": 2.190642471263249e-07, + "loss": 3.0909, + "step": 23329 + }, + { + "epoch": 1.870377422781664, + "grad_norm": 0.639017641544342, + "learning_rate": 2.1879403744029193e-07, + "loss": 2.4485, + "step": 23330 + }, + { + "epoch": 1.8704575975626867, + "grad_norm": 1.7383487224578857, + "learning_rate": 2.1852399266194312e-07, + "loss": 3.1806, + "step": 23331 + }, + { + "epoch": 1.8705377723437093, + "grad_norm": 1.7858566045761108, + "learning_rate": 2.1825411279583375e-07, + "loss": 3.0977, + "step": 23332 + }, + { + "epoch": 1.870617947124732, + "grad_norm": 1.7130107879638672, + "learning_rate": 2.1798439784651238e-07, + "loss": 3.1122, + "step": 23333 + }, + { + "epoch": 1.8706981219057546, + "grad_norm": 1.7452430725097656, + "learning_rate": 2.1771484781852647e-07, + "loss": 3.1057, + "step": 23334 + }, + { + "epoch": 1.8707782966867772, + "grad_norm": 1.7193294763565063, + "learning_rate": 2.174454627164202e-07, + "loss": 3.1011, + "step": 23335 + }, + { + "epoch": 1.8708584714677998, + "grad_norm": 1.8748054504394531, + "learning_rate": 2.1717624254473545e-07, + "loss": 3.2067, + "step": 23336 + }, + { + "epoch": 1.8709386462488224, + "grad_norm": 1.8167479038238525, + "learning_rate": 2.1690718730801198e-07, + "loss": 3.1099, + "step": 23337 + }, + { + "epoch": 1.871018821029845, + "grad_norm": 1.8065497875213623, + "learning_rate": 2.166382970107861e-07, + "loss": 3.2786, + "step": 23338 + }, + { + "epoch": 1.8710989958108677, + "grad_norm": 1.8614380359649658, + "learning_rate": 2.1636957165758866e-07, + "loss": 3.2379, + "step": 23339 + }, + { + "epoch": 1.8711791705918903, + "grad_norm": 1.7305052280426025, + "learning_rate": 2.161010112529538e-07, + "loss": 3.1484, + "step": 23340 + }, + { + "epoch": 1.871259345372913, + "grad_norm": 1.7343757152557373, + "learning_rate": 2.158326158014068e-07, + "loss": 3.1384, + "step": 23341 + }, + { + "epoch": 1.8713395201539356, + "grad_norm": 1.795903205871582, + "learning_rate": 2.1556438530747402e-07, + "loss": 3.2062, + "step": 23342 + }, + { + "epoch": 1.8714196949349582, + "grad_norm": 1.7005741596221924, + "learning_rate": 2.1529631977567745e-07, + "loss": 3.1481, + "step": 23343 + }, + { + "epoch": 1.8714998697159808, + "grad_norm": 1.7247999906539917, + "learning_rate": 2.1502841921053673e-07, + "loss": 3.3105, + "step": 23344 + }, + { + "epoch": 1.8715800444970034, + "grad_norm": 1.8125072717666626, + "learning_rate": 2.147606836165672e-07, + "loss": 3.1096, + "step": 23345 + }, + { + "epoch": 1.871660219278026, + "grad_norm": 1.7906098365783691, + "learning_rate": 2.1449311299828412e-07, + "loss": 3.1056, + "step": 23346 + }, + { + "epoch": 1.8717403940590487, + "grad_norm": 1.7556086778640747, + "learning_rate": 2.1422570736019944e-07, + "loss": 3.14, + "step": 23347 + }, + { + "epoch": 1.8718205688400713, + "grad_norm": 1.8068585395812988, + "learning_rate": 2.1395846670682063e-07, + "loss": 3.2113, + "step": 23348 + }, + { + "epoch": 1.871900743621094, + "grad_norm": 1.721580982208252, + "learning_rate": 2.1369139104265191e-07, + "loss": 3.1619, + "step": 23349 + }, + { + "epoch": 1.8719809184021168, + "grad_norm": 1.8099348545074463, + "learning_rate": 2.1342448037219743e-07, + "loss": 3.2083, + "step": 23350 + }, + { + "epoch": 1.8720610931831394, + "grad_norm": 1.7635529041290283, + "learning_rate": 2.1315773469995803e-07, + "loss": 3.1286, + "step": 23351 + }, + { + "epoch": 1.872141267964162, + "grad_norm": 1.8369007110595703, + "learning_rate": 2.1289115403042905e-07, + "loss": 3.2032, + "step": 23352 + }, + { + "epoch": 1.8722214427451846, + "grad_norm": 1.6733059883117676, + "learning_rate": 2.1262473836810682e-07, + "loss": 3.1331, + "step": 23353 + }, + { + "epoch": 1.8723016175262073, + "grad_norm": 1.7806609869003296, + "learning_rate": 2.1235848771748113e-07, + "loss": 3.2137, + "step": 23354 + }, + { + "epoch": 1.87238179230723, + "grad_norm": 1.7604762315750122, + "learning_rate": 2.120924020830428e-07, + "loss": 3.1406, + "step": 23355 + }, + { + "epoch": 1.8724619670882525, + "grad_norm": 1.7473549842834473, + "learning_rate": 2.1182648146927608e-07, + "loss": 3.2141, + "step": 23356 + }, + { + "epoch": 1.8725421418692751, + "grad_norm": 1.8507298231124878, + "learning_rate": 2.115607258806651e-07, + "loss": 3.2551, + "step": 23357 + }, + { + "epoch": 1.8726223166502978, + "grad_norm": 1.8484779596328735, + "learning_rate": 2.112951353216919e-07, + "loss": 3.2102, + "step": 23358 + }, + { + "epoch": 1.8727024914313204, + "grad_norm": 1.81635582447052, + "learning_rate": 2.110297097968317e-07, + "loss": 3.2815, + "step": 23359 + }, + { + "epoch": 1.872782666212343, + "grad_norm": 1.8873353004455566, + "learning_rate": 2.1076444931056096e-07, + "loss": 3.1578, + "step": 23360 + }, + { + "epoch": 1.8728628409933656, + "grad_norm": 1.8270384073257446, + "learning_rate": 2.1049935386735165e-07, + "loss": 3.2006, + "step": 23361 + }, + { + "epoch": 1.8729430157743883, + "grad_norm": 1.6819732189178467, + "learning_rate": 2.1023442347167244e-07, + "loss": 3.2993, + "step": 23362 + }, + { + "epoch": 1.8730231905554109, + "grad_norm": 0.6652172207832336, + "learning_rate": 2.0996965812799197e-07, + "loss": 2.4642, + "step": 23363 + }, + { + "epoch": 1.8731033653364335, + "grad_norm": 1.8348406553268433, + "learning_rate": 2.097050578407711e-07, + "loss": 3.1719, + "step": 23364 + }, + { + "epoch": 1.8731835401174561, + "grad_norm": 1.7764294147491455, + "learning_rate": 2.0944062261447297e-07, + "loss": 3.178, + "step": 23365 + }, + { + "epoch": 1.8732637148984788, + "grad_norm": 1.879507303237915, + "learning_rate": 2.0917635245355504e-07, + "loss": 3.1421, + "step": 23366 + }, + { + "epoch": 1.8733438896795014, + "grad_norm": 1.7281659841537476, + "learning_rate": 2.0891224736247383e-07, + "loss": 3.1445, + "step": 23367 + }, + { + "epoch": 1.873424064460524, + "grad_norm": 1.7759504318237305, + "learning_rate": 2.086483073456824e-07, + "loss": 3.1678, + "step": 23368 + }, + { + "epoch": 1.8735042392415466, + "grad_norm": 1.8087959289550781, + "learning_rate": 2.0838453240762835e-07, + "loss": 3.1965, + "step": 23369 + }, + { + "epoch": 1.8735844140225693, + "grad_norm": 1.7440431118011475, + "learning_rate": 2.0812092255276028e-07, + "loss": 3.11, + "step": 23370 + }, + { + "epoch": 1.8736645888035919, + "grad_norm": 1.8405859470367432, + "learning_rate": 2.0785747778552246e-07, + "loss": 3.2407, + "step": 23371 + }, + { + "epoch": 1.8737447635846145, + "grad_norm": 1.935414433479309, + "learning_rate": 2.0759419811035686e-07, + "loss": 3.262, + "step": 23372 + }, + { + "epoch": 1.8738249383656371, + "grad_norm": 1.7181838750839233, + "learning_rate": 2.0733108353170217e-07, + "loss": 3.1332, + "step": 23373 + }, + { + "epoch": 1.8739051131466598, + "grad_norm": 1.8049830198287964, + "learning_rate": 2.0706813405399374e-07, + "loss": 3.2194, + "step": 23374 + }, + { + "epoch": 1.8739852879276824, + "grad_norm": 1.717527985572815, + "learning_rate": 2.0680534968166467e-07, + "loss": 3.2164, + "step": 23375 + }, + { + "epoch": 1.874065462708705, + "grad_norm": 1.7893131971359253, + "learning_rate": 2.0654273041914697e-07, + "loss": 3.0449, + "step": 23376 + }, + { + "epoch": 1.8741456374897276, + "grad_norm": 1.8287606239318848, + "learning_rate": 2.06280276270866e-07, + "loss": 3.1979, + "step": 23377 + }, + { + "epoch": 1.8742258122707502, + "grad_norm": 1.8237468004226685, + "learning_rate": 2.0601798724124932e-07, + "loss": 3.1479, + "step": 23378 + }, + { + "epoch": 1.8743059870517729, + "grad_norm": 1.7148869037628174, + "learning_rate": 2.0575586333471787e-07, + "loss": 3.1835, + "step": 23379 + }, + { + "epoch": 1.8743861618327955, + "grad_norm": 1.89387845993042, + "learning_rate": 2.0549390455568918e-07, + "loss": 3.0208, + "step": 23380 + }, + { + "epoch": 1.8744663366138181, + "grad_norm": 1.6565535068511963, + "learning_rate": 2.0523211090858198e-07, + "loss": 3.0695, + "step": 23381 + }, + { + "epoch": 1.8745465113948407, + "grad_norm": 1.7279640436172485, + "learning_rate": 2.0497048239780936e-07, + "loss": 3.2112, + "step": 23382 + }, + { + "epoch": 1.8746266861758634, + "grad_norm": 1.9102895259857178, + "learning_rate": 2.047090190277834e-07, + "loss": 3.2326, + "step": 23383 + }, + { + "epoch": 1.874706860956886, + "grad_norm": 1.7110296487808228, + "learning_rate": 2.0444772080290943e-07, + "loss": 3.2526, + "step": 23384 + }, + { + "epoch": 1.8747870357379086, + "grad_norm": 1.81770920753479, + "learning_rate": 2.0418658772759504e-07, + "loss": 3.1316, + "step": 23385 + }, + { + "epoch": 1.8748672105189312, + "grad_norm": 1.742356538772583, + "learning_rate": 2.0392561980624338e-07, + "loss": 3.1311, + "step": 23386 + }, + { + "epoch": 1.8749473852999539, + "grad_norm": 0.6544739603996277, + "learning_rate": 2.0366481704325203e-07, + "loss": 2.4617, + "step": 23387 + }, + { + "epoch": 1.8750275600809765, + "grad_norm": 1.7691762447357178, + "learning_rate": 2.034041794430208e-07, + "loss": 3.2216, + "step": 23388 + }, + { + "epoch": 1.875107734861999, + "grad_norm": 1.8706170320510864, + "learning_rate": 2.031437070099418e-07, + "loss": 3.2297, + "step": 23389 + }, + { + "epoch": 1.8751879096430217, + "grad_norm": 1.7675799131393433, + "learning_rate": 2.0288339974840588e-07, + "loss": 3.2065, + "step": 23390 + }, + { + "epoch": 1.8752680844240444, + "grad_norm": 1.8765827417373657, + "learning_rate": 2.02623257662804e-07, + "loss": 3.1442, + "step": 23391 + }, + { + "epoch": 1.875348259205067, + "grad_norm": 1.8283897638320923, + "learning_rate": 2.0236328075752153e-07, + "loss": 3.1367, + "step": 23392 + }, + { + "epoch": 1.8754284339860896, + "grad_norm": 1.8457496166229248, + "learning_rate": 2.0210346903694056e-07, + "loss": 3.1345, + "step": 23393 + }, + { + "epoch": 1.8755086087671122, + "grad_norm": 1.7376781702041626, + "learning_rate": 2.0184382250544198e-07, + "loss": 3.1028, + "step": 23394 + }, + { + "epoch": 1.8755887835481349, + "grad_norm": 1.8359495401382446, + "learning_rate": 2.0158434116740344e-07, + "loss": 3.0787, + "step": 23395 + }, + { + "epoch": 1.8756689583291575, + "grad_norm": 1.6719846725463867, + "learning_rate": 2.0132502502720031e-07, + "loss": 3.1808, + "step": 23396 + }, + { + "epoch": 1.87574913311018, + "grad_norm": 1.8436081409454346, + "learning_rate": 2.0106587408920242e-07, + "loss": 3.1588, + "step": 23397 + }, + { + "epoch": 1.8758293078912027, + "grad_norm": 1.6662040948867798, + "learning_rate": 2.0080688835778184e-07, + "loss": 3.1635, + "step": 23398 + }, + { + "epoch": 1.8759094826722253, + "grad_norm": 0.6482636332511902, + "learning_rate": 2.0054806783730285e-07, + "loss": 2.4624, + "step": 23399 + }, + { + "epoch": 1.875989657453248, + "grad_norm": 1.7417141199111938, + "learning_rate": 2.0028941253212975e-07, + "loss": 3.0989, + "step": 23400 + }, + { + "epoch": 1.8760698322342706, + "grad_norm": 1.8334026336669922, + "learning_rate": 2.000309224466235e-07, + "loss": 3.2113, + "step": 23401 + }, + { + "epoch": 1.8761500070152932, + "grad_norm": 1.8171443939208984, + "learning_rate": 1.9977259758514168e-07, + "loss": 3.1234, + "step": 23402 + }, + { + "epoch": 1.8762301817963158, + "grad_norm": 1.7860372066497803, + "learning_rate": 1.9951443795204083e-07, + "loss": 3.1385, + "step": 23403 + }, + { + "epoch": 1.8763103565773385, + "grad_norm": 1.7094519138336182, + "learning_rate": 1.9925644355167196e-07, + "loss": 3.1591, + "step": 23404 + }, + { + "epoch": 1.876390531358361, + "grad_norm": 1.840701699256897, + "learning_rate": 1.9899861438838485e-07, + "loss": 3.1998, + "step": 23405 + }, + { + "epoch": 1.8764707061393837, + "grad_norm": 1.7842824459075928, + "learning_rate": 1.987409504665283e-07, + "loss": 3.1183, + "step": 23406 + }, + { + "epoch": 1.8765508809204063, + "grad_norm": 1.7672227621078491, + "learning_rate": 1.9848345179044327e-07, + "loss": 3.1771, + "step": 23407 + }, + { + "epoch": 1.876631055701429, + "grad_norm": 1.7611708641052246, + "learning_rate": 1.9822611836447515e-07, + "loss": 3.2655, + "step": 23408 + }, + { + "epoch": 1.8767112304824516, + "grad_norm": 1.7339574098587036, + "learning_rate": 1.9796895019295826e-07, + "loss": 3.2268, + "step": 23409 + }, + { + "epoch": 1.8767914052634742, + "grad_norm": 1.8677899837493896, + "learning_rate": 1.9771194728023025e-07, + "loss": 3.192, + "step": 23410 + }, + { + "epoch": 1.876871580044497, + "grad_norm": 1.6106964349746704, + "learning_rate": 1.974551096306243e-07, + "loss": 3.1156, + "step": 23411 + }, + { + "epoch": 1.8769517548255197, + "grad_norm": 0.6611196398735046, + "learning_rate": 1.9719843724847144e-07, + "loss": 2.4745, + "step": 23412 + }, + { + "epoch": 1.8770319296065423, + "grad_norm": 2.2978122234344482, + "learning_rate": 1.9694193013809814e-07, + "loss": 3.2025, + "step": 23413 + }, + { + "epoch": 1.877112104387565, + "grad_norm": 1.7424776554107666, + "learning_rate": 1.9668558830382768e-07, + "loss": 3.1353, + "step": 23414 + }, + { + "epoch": 1.8771922791685876, + "grad_norm": 1.810157299041748, + "learning_rate": 1.964294117499843e-07, + "loss": 3.1934, + "step": 23415 + }, + { + "epoch": 1.8772724539496102, + "grad_norm": 1.7589085102081299, + "learning_rate": 1.9617340048088463e-07, + "loss": 3.1694, + "step": 23416 + }, + { + "epoch": 1.8773526287306328, + "grad_norm": 1.7925320863723755, + "learning_rate": 1.9591755450084737e-07, + "loss": 3.2304, + "step": 23417 + }, + { + "epoch": 1.8774328035116554, + "grad_norm": 1.7108385562896729, + "learning_rate": 1.9566187381418468e-07, + "loss": 3.1518, + "step": 23418 + }, + { + "epoch": 1.877512978292678, + "grad_norm": 1.8425909280776978, + "learning_rate": 1.954063584252075e-07, + "loss": 3.2441, + "step": 23419 + }, + { + "epoch": 1.8775931530737007, + "grad_norm": 1.7378613948822021, + "learning_rate": 1.9515100833822354e-07, + "loss": 3.202, + "step": 23420 + }, + { + "epoch": 1.8776733278547233, + "grad_norm": 1.7025002241134644, + "learning_rate": 1.9489582355753712e-07, + "loss": 3.0962, + "step": 23421 + }, + { + "epoch": 1.877753502635746, + "grad_norm": 1.6896073818206787, + "learning_rate": 1.9464080408745256e-07, + "loss": 3.0862, + "step": 23422 + }, + { + "epoch": 1.8778336774167685, + "grad_norm": 1.7506431341171265, + "learning_rate": 1.9438594993226866e-07, + "loss": 3.0648, + "step": 23423 + }, + { + "epoch": 1.8779138521977912, + "grad_norm": 1.8609721660614014, + "learning_rate": 1.9413126109628089e-07, + "loss": 3.1848, + "step": 23424 + }, + { + "epoch": 1.8779940269788138, + "grad_norm": 1.807458758354187, + "learning_rate": 1.9387673758378356e-07, + "loss": 3.1275, + "step": 23425 + }, + { + "epoch": 1.8780742017598364, + "grad_norm": 1.7803844213485718, + "learning_rate": 1.9362237939906992e-07, + "loss": 3.2123, + "step": 23426 + }, + { + "epoch": 1.878154376540859, + "grad_norm": 1.7465643882751465, + "learning_rate": 1.9336818654642543e-07, + "loss": 3.3228, + "step": 23427 + }, + { + "epoch": 1.8782345513218817, + "grad_norm": 1.8413516283035278, + "learning_rate": 1.9311415903013885e-07, + "loss": 3.2062, + "step": 23428 + }, + { + "epoch": 1.8783147261029043, + "grad_norm": 1.7428361177444458, + "learning_rate": 1.9286029685449016e-07, + "loss": 3.0968, + "step": 23429 + }, + { + "epoch": 1.878394900883927, + "grad_norm": 1.818729043006897, + "learning_rate": 1.926066000237603e-07, + "loss": 3.0813, + "step": 23430 + }, + { + "epoch": 1.8784750756649498, + "grad_norm": 1.737235188484192, + "learning_rate": 1.9235306854222701e-07, + "loss": 3.2272, + "step": 23431 + }, + { + "epoch": 1.8785552504459724, + "grad_norm": 1.755913257598877, + "learning_rate": 1.9209970241416466e-07, + "loss": 3.1778, + "step": 23432 + }, + { + "epoch": 1.878635425226995, + "grad_norm": 1.7782025337219238, + "learning_rate": 1.9184650164384422e-07, + "loss": 3.2025, + "step": 23433 + }, + { + "epoch": 1.8787156000080176, + "grad_norm": 1.8133987188339233, + "learning_rate": 1.9159346623553565e-07, + "loss": 3.2057, + "step": 23434 + }, + { + "epoch": 1.8787957747890403, + "grad_norm": 1.7544300556182861, + "learning_rate": 1.913405961935033e-07, + "loss": 3.1044, + "step": 23435 + }, + { + "epoch": 1.8788759495700629, + "grad_norm": 1.7966259717941284, + "learning_rate": 1.910878915220127e-07, + "loss": 3.2511, + "step": 23436 + }, + { + "epoch": 1.8789561243510855, + "grad_norm": 1.8042656183242798, + "learning_rate": 1.9083535222532258e-07, + "loss": 3.2416, + "step": 23437 + }, + { + "epoch": 1.8790362991321081, + "grad_norm": 1.7085013389587402, + "learning_rate": 1.905829783076929e-07, + "loss": 3.1306, + "step": 23438 + }, + { + "epoch": 1.8791164739131307, + "grad_norm": 1.7444543838500977, + "learning_rate": 1.9033076977337472e-07, + "loss": 3.1534, + "step": 23439 + }, + { + "epoch": 1.8791966486941534, + "grad_norm": 1.8054324388504028, + "learning_rate": 1.9007872662662352e-07, + "loss": 3.1797, + "step": 23440 + }, + { + "epoch": 1.879276823475176, + "grad_norm": 1.760912537574768, + "learning_rate": 1.8982684887168812e-07, + "loss": 3.0947, + "step": 23441 + }, + { + "epoch": 1.8793569982561986, + "grad_norm": 1.8526235818862915, + "learning_rate": 1.8957513651281401e-07, + "loss": 3.2184, + "step": 23442 + }, + { + "epoch": 1.8794371730372212, + "grad_norm": 1.735630750656128, + "learning_rate": 1.8932358955424558e-07, + "loss": 3.1748, + "step": 23443 + }, + { + "epoch": 1.8795173478182439, + "grad_norm": 1.7419648170471191, + "learning_rate": 1.8907220800022386e-07, + "loss": 3.137, + "step": 23444 + }, + { + "epoch": 1.8795975225992665, + "grad_norm": 1.7144607305526733, + "learning_rate": 1.888209918549877e-07, + "loss": 3.1591, + "step": 23445 + }, + { + "epoch": 1.8796776973802891, + "grad_norm": 0.666246771812439, + "learning_rate": 1.8856994112277038e-07, + "loss": 2.4832, + "step": 23446 + }, + { + "epoch": 1.8797578721613117, + "grad_norm": 1.7368155717849731, + "learning_rate": 1.883190558078074e-07, + "loss": 3.1625, + "step": 23447 + }, + { + "epoch": 1.8798380469423344, + "grad_norm": 1.7955520153045654, + "learning_rate": 1.880683359143265e-07, + "loss": 3.2467, + "step": 23448 + }, + { + "epoch": 1.879918221723357, + "grad_norm": 1.7583134174346924, + "learning_rate": 1.878177814465554e-07, + "loss": 3.1041, + "step": 23449 + }, + { + "epoch": 1.8799983965043796, + "grad_norm": 1.7997546195983887, + "learning_rate": 1.8756739240871846e-07, + "loss": 3.1132, + "step": 23450 + }, + { + "epoch": 1.8800785712854022, + "grad_norm": 1.833376407623291, + "learning_rate": 1.873171688050368e-07, + "loss": 3.1691, + "step": 23451 + }, + { + "epoch": 1.8801587460664249, + "grad_norm": 1.8462332487106323, + "learning_rate": 1.8706711063972928e-07, + "loss": 3.1535, + "step": 23452 + }, + { + "epoch": 1.8802389208474475, + "grad_norm": 1.7371902465820312, + "learning_rate": 1.8681721791701358e-07, + "loss": 3.0563, + "step": 23453 + }, + { + "epoch": 1.88031909562847, + "grad_norm": 1.6878859996795654, + "learning_rate": 1.8656749064109857e-07, + "loss": 3.119, + "step": 23454 + }, + { + "epoch": 1.8803992704094927, + "grad_norm": 1.709712266921997, + "learning_rate": 1.863179288161987e-07, + "loss": 3.2473, + "step": 23455 + }, + { + "epoch": 1.8804794451905154, + "grad_norm": 1.6855480670928955, + "learning_rate": 1.8606853244651833e-07, + "loss": 3.1759, + "step": 23456 + }, + { + "epoch": 1.880559619971538, + "grad_norm": 0.6450902223587036, + "learning_rate": 1.8581930153626525e-07, + "loss": 2.4528, + "step": 23457 + }, + { + "epoch": 1.8806397947525606, + "grad_norm": 1.711928367614746, + "learning_rate": 1.8557023608963943e-07, + "loss": 3.1286, + "step": 23458 + }, + { + "epoch": 1.8807199695335832, + "grad_norm": 1.7169462442398071, + "learning_rate": 1.853213361108408e-07, + "loss": 3.1805, + "step": 23459 + }, + { + "epoch": 1.8808001443146058, + "grad_norm": 1.830797553062439, + "learning_rate": 1.8507260160406492e-07, + "loss": 3.1616, + "step": 23460 + }, + { + "epoch": 1.8808803190956285, + "grad_norm": 1.6819103956222534, + "learning_rate": 1.8482403257350623e-07, + "loss": 3.147, + "step": 23461 + }, + { + "epoch": 1.880960493876651, + "grad_norm": 1.7070327997207642, + "learning_rate": 1.8457562902335469e-07, + "loss": 3.1383, + "step": 23462 + }, + { + "epoch": 1.8810406686576737, + "grad_norm": 1.6824603080749512, + "learning_rate": 1.8432739095779917e-07, + "loss": 3.1827, + "step": 23463 + }, + { + "epoch": 1.8811208434386963, + "grad_norm": 1.8248915672302246, + "learning_rate": 1.8407931838102412e-07, + "loss": 3.1675, + "step": 23464 + }, + { + "epoch": 1.881201018219719, + "grad_norm": 1.9547748565673828, + "learning_rate": 1.8383141129721283e-07, + "loss": 3.2338, + "step": 23465 + }, + { + "epoch": 1.8812811930007416, + "grad_norm": 1.771413803100586, + "learning_rate": 1.835836697105431e-07, + "loss": 3.2485, + "step": 23466 + }, + { + "epoch": 1.8813613677817642, + "grad_norm": 1.852332592010498, + "learning_rate": 1.833360936251938e-07, + "loss": 3.2153, + "step": 23467 + }, + { + "epoch": 1.8814415425627868, + "grad_norm": 1.772234559059143, + "learning_rate": 1.8308868304533934e-07, + "loss": 3.218, + "step": 23468 + }, + { + "epoch": 1.8815217173438095, + "grad_norm": 1.7392076253890991, + "learning_rate": 1.8284143797514865e-07, + "loss": 3.1711, + "step": 23469 + }, + { + "epoch": 1.881601892124832, + "grad_norm": 1.7664674520492554, + "learning_rate": 1.825943584187917e-07, + "loss": 3.203, + "step": 23470 + }, + { + "epoch": 1.8816820669058547, + "grad_norm": 1.7365745306015015, + "learning_rate": 1.8234744438043406e-07, + "loss": 3.1522, + "step": 23471 + }, + { + "epoch": 1.8817622416868773, + "grad_norm": 1.8788111209869385, + "learning_rate": 1.8210069586423796e-07, + "loss": 3.2026, + "step": 23472 + }, + { + "epoch": 1.8818424164679, + "grad_norm": 1.7583681344985962, + "learning_rate": 1.818541128743645e-07, + "loss": 3.178, + "step": 23473 + }, + { + "epoch": 1.8819225912489226, + "grad_norm": 1.709189772605896, + "learning_rate": 1.8160769541497036e-07, + "loss": 3.0867, + "step": 23474 + }, + { + "epoch": 1.8820027660299452, + "grad_norm": 1.961980938911438, + "learning_rate": 1.8136144349021002e-07, + "loss": 3.2589, + "step": 23475 + }, + { + "epoch": 1.8820829408109678, + "grad_norm": 1.815747857093811, + "learning_rate": 1.8111535710423457e-07, + "loss": 3.1592, + "step": 23476 + }, + { + "epoch": 1.8821631155919905, + "grad_norm": 1.833963394165039, + "learning_rate": 1.8086943626119401e-07, + "loss": 3.1509, + "step": 23477 + }, + { + "epoch": 1.882243290373013, + "grad_norm": 1.8190269470214844, + "learning_rate": 1.806236809652362e-07, + "loss": 3.0947, + "step": 23478 + }, + { + "epoch": 1.8823234651540357, + "grad_norm": 1.6794732809066772, + "learning_rate": 1.8037809122049998e-07, + "loss": 3.1262, + "step": 23479 + }, + { + "epoch": 1.8824036399350583, + "grad_norm": 1.7406822443008423, + "learning_rate": 1.8013266703112875e-07, + "loss": 3.113, + "step": 23480 + }, + { + "epoch": 1.882483814716081, + "grad_norm": 1.7722605466842651, + "learning_rate": 1.7988740840126028e-07, + "loss": 3.2197, + "step": 23481 + }, + { + "epoch": 1.8825639894971036, + "grad_norm": 1.8422666788101196, + "learning_rate": 1.7964231533502908e-07, + "loss": 3.233, + "step": 23482 + }, + { + "epoch": 1.8826441642781262, + "grad_norm": 1.8840285539627075, + "learning_rate": 1.7939738783656847e-07, + "loss": 3.1665, + "step": 23483 + }, + { + "epoch": 1.8827243390591488, + "grad_norm": 1.6998544931411743, + "learning_rate": 1.7915262591000514e-07, + "loss": 3.2862, + "step": 23484 + }, + { + "epoch": 1.8828045138401714, + "grad_norm": 1.8306457996368408, + "learning_rate": 1.7890802955946807e-07, + "loss": 3.1849, + "step": 23485 + }, + { + "epoch": 1.882884688621194, + "grad_norm": 2.0862209796905518, + "learning_rate": 1.7866359878908058e-07, + "loss": 3.2448, + "step": 23486 + }, + { + "epoch": 1.8829648634022167, + "grad_norm": 1.7782013416290283, + "learning_rate": 1.784193336029627e-07, + "loss": 3.1095, + "step": 23487 + }, + { + "epoch": 1.8830450381832393, + "grad_norm": 1.7049874067306519, + "learning_rate": 1.781752340052345e-07, + "loss": 3.239, + "step": 23488 + }, + { + "epoch": 1.883125212964262, + "grad_norm": 1.7598191499710083, + "learning_rate": 1.7793130000000935e-07, + "loss": 3.1522, + "step": 23489 + }, + { + "epoch": 1.8832053877452846, + "grad_norm": 1.8421858549118042, + "learning_rate": 1.7768753159140173e-07, + "loss": 3.1805, + "step": 23490 + }, + { + "epoch": 1.8832855625263072, + "grad_norm": 1.8487282991409302, + "learning_rate": 1.7744392878351947e-07, + "loss": 3.2178, + "step": 23491 + }, + { + "epoch": 1.88336573730733, + "grad_norm": 1.7391940355300903, + "learning_rate": 1.7720049158047147e-07, + "loss": 3.1765, + "step": 23492 + }, + { + "epoch": 1.8834459120883527, + "grad_norm": 1.8539302349090576, + "learning_rate": 1.7695721998636117e-07, + "loss": 3.1262, + "step": 23493 + }, + { + "epoch": 1.8835260868693753, + "grad_norm": 1.8289482593536377, + "learning_rate": 1.7671411400529083e-07, + "loss": 3.2339, + "step": 23494 + }, + { + "epoch": 1.883606261650398, + "grad_norm": 1.9120687246322632, + "learning_rate": 1.7647117364135713e-07, + "loss": 3.2025, + "step": 23495 + }, + { + "epoch": 1.8836864364314205, + "grad_norm": 1.6817090511322021, + "learning_rate": 1.7622839889865796e-07, + "loss": 3.1006, + "step": 23496 + }, + { + "epoch": 1.8837666112124432, + "grad_norm": 1.7615997791290283, + "learning_rate": 1.7598578978128556e-07, + "loss": 3.1091, + "step": 23497 + }, + { + "epoch": 1.8838467859934658, + "grad_norm": 1.869706630706787, + "learning_rate": 1.757433462933311e-07, + "loss": 3.1793, + "step": 23498 + }, + { + "epoch": 1.8839269607744884, + "grad_norm": 1.793554663658142, + "learning_rate": 1.7550106843888137e-07, + "loss": 3.2413, + "step": 23499 + }, + { + "epoch": 1.884007135555511, + "grad_norm": 1.9564932584762573, + "learning_rate": 1.7525895622201972e-07, + "loss": 3.17, + "step": 23500 + }, + { + "epoch": 1.8840873103365336, + "grad_norm": 1.8670399188995361, + "learning_rate": 1.7501700964683067e-07, + "loss": 3.0963, + "step": 23501 + }, + { + "epoch": 1.8841674851175563, + "grad_norm": 1.7185206413269043, + "learning_rate": 1.7477522871739205e-07, + "loss": 3.1897, + "step": 23502 + }, + { + "epoch": 1.884247659898579, + "grad_norm": 1.7660622596740723, + "learning_rate": 1.7453361343778063e-07, + "loss": 3.1625, + "step": 23503 + }, + { + "epoch": 1.8843278346796015, + "grad_norm": 1.7680895328521729, + "learning_rate": 1.742921638120698e-07, + "loss": 3.1162, + "step": 23504 + }, + { + "epoch": 1.8844080094606241, + "grad_norm": 1.6645286083221436, + "learning_rate": 1.7405087984432857e-07, + "loss": 3.1927, + "step": 23505 + }, + { + "epoch": 1.8844881842416468, + "grad_norm": 1.8337253332138062, + "learning_rate": 1.7380976153862807e-07, + "loss": 3.152, + "step": 23506 + }, + { + "epoch": 1.8845683590226694, + "grad_norm": 0.6572601199150085, + "learning_rate": 1.735688088990306e-07, + "loss": 2.4603, + "step": 23507 + }, + { + "epoch": 1.884648533803692, + "grad_norm": 1.6801674365997314, + "learning_rate": 1.7332802192960185e-07, + "loss": 3.166, + "step": 23508 + }, + { + "epoch": 1.8847287085847146, + "grad_norm": 1.6982667446136475, + "learning_rate": 1.7308740063439855e-07, + "loss": 3.1205, + "step": 23509 + }, + { + "epoch": 1.8848088833657373, + "grad_norm": 1.8243706226348877, + "learning_rate": 1.7284694501747746e-07, + "loss": 3.1845, + "step": 23510 + }, + { + "epoch": 1.8848890581467599, + "grad_norm": 1.823079228401184, + "learning_rate": 1.7260665508289422e-07, + "loss": 3.0909, + "step": 23511 + }, + { + "epoch": 1.8849692329277825, + "grad_norm": 1.764473557472229, + "learning_rate": 1.723665308346989e-07, + "loss": 3.1906, + "step": 23512 + }, + { + "epoch": 1.8850494077088054, + "grad_norm": 1.6955137252807617, + "learning_rate": 1.7212657227694052e-07, + "loss": 3.0754, + "step": 23513 + }, + { + "epoch": 1.885129582489828, + "grad_norm": 1.638227939605713, + "learning_rate": 1.7188677941366471e-07, + "loss": 3.0612, + "step": 23514 + }, + { + "epoch": 1.8852097572708506, + "grad_norm": 1.8705686330795288, + "learning_rate": 1.716471522489127e-07, + "loss": 3.1901, + "step": 23515 + }, + { + "epoch": 1.8852899320518732, + "grad_norm": 1.8877569437026978, + "learning_rate": 1.714076907867268e-07, + "loss": 3.2472, + "step": 23516 + }, + { + "epoch": 1.8853701068328959, + "grad_norm": 1.7512931823730469, + "learning_rate": 1.711683950311438e-07, + "loss": 3.156, + "step": 23517 + }, + { + "epoch": 1.8854502816139185, + "grad_norm": 1.763675570487976, + "learning_rate": 1.709292649861971e-07, + "loss": 3.2256, + "step": 23518 + }, + { + "epoch": 1.885530456394941, + "grad_norm": 1.8693405389785767, + "learning_rate": 1.7069030065591908e-07, + "loss": 3.136, + "step": 23519 + }, + { + "epoch": 1.8856106311759637, + "grad_norm": 1.6831622123718262, + "learning_rate": 1.704515020443387e-07, + "loss": 3.1324, + "step": 23520 + }, + { + "epoch": 1.8856908059569863, + "grad_norm": 1.6769213676452637, + "learning_rate": 1.702128691554805e-07, + "loss": 3.1538, + "step": 23521 + }, + { + "epoch": 1.885770980738009, + "grad_norm": 1.6970226764678955, + "learning_rate": 1.6997440199337023e-07, + "loss": 3.0753, + "step": 23522 + }, + { + "epoch": 1.8858511555190316, + "grad_norm": 1.763040542602539, + "learning_rate": 1.6973610056202682e-07, + "loss": 3.2111, + "step": 23523 + }, + { + "epoch": 1.8859313303000542, + "grad_norm": 1.6966615915298462, + "learning_rate": 1.6949796486546822e-07, + "loss": 3.1058, + "step": 23524 + }, + { + "epoch": 1.8860115050810768, + "grad_norm": 1.9032926559448242, + "learning_rate": 1.6925999490770893e-07, + "loss": 3.26, + "step": 23525 + }, + { + "epoch": 1.8860916798620995, + "grad_norm": 1.7514941692352295, + "learning_rate": 1.6902219069276137e-07, + "loss": 3.1676, + "step": 23526 + }, + { + "epoch": 1.886171854643122, + "grad_norm": 1.7729823589324951, + "learning_rate": 1.6878455222463563e-07, + "loss": 3.1993, + "step": 23527 + }, + { + "epoch": 1.8862520294241447, + "grad_norm": 1.8270667791366577, + "learning_rate": 1.685470795073374e-07, + "loss": 3.1414, + "step": 23528 + }, + { + "epoch": 1.8863322042051673, + "grad_norm": 1.8041306734085083, + "learning_rate": 1.6830977254487014e-07, + "loss": 3.2133, + "step": 23529 + }, + { + "epoch": 1.88641237898619, + "grad_norm": 1.848507285118103, + "learning_rate": 1.680726313412351e-07, + "loss": 3.2778, + "step": 23530 + }, + { + "epoch": 1.8864925537672126, + "grad_norm": 1.7726515531539917, + "learning_rate": 1.6783565590043016e-07, + "loss": 3.1912, + "step": 23531 + }, + { + "epoch": 1.8865727285482352, + "grad_norm": 1.825432538986206, + "learning_rate": 1.6759884622645107e-07, + "loss": 3.1487, + "step": 23532 + }, + { + "epoch": 1.8866529033292578, + "grad_norm": 1.7357707023620605, + "learning_rate": 1.6736220232329014e-07, + "loss": 3.1901, + "step": 23533 + }, + { + "epoch": 1.8867330781102805, + "grad_norm": 1.747698187828064, + "learning_rate": 1.6712572419493755e-07, + "loss": 3.1714, + "step": 23534 + }, + { + "epoch": 1.886813252891303, + "grad_norm": 0.6870718598365784, + "learning_rate": 1.66889411845379e-07, + "loss": 2.4803, + "step": 23535 + }, + { + "epoch": 1.8868934276723257, + "grad_norm": 1.7598506212234497, + "learning_rate": 1.6665326527860016e-07, + "loss": 3.1542, + "step": 23536 + }, + { + "epoch": 1.8869736024533483, + "grad_norm": 1.7591978311538696, + "learning_rate": 1.664172844985812e-07, + "loss": 3.2119, + "step": 23537 + }, + { + "epoch": 1.887053777234371, + "grad_norm": 1.7772694826126099, + "learning_rate": 1.6618146950930114e-07, + "loss": 3.1787, + "step": 23538 + }, + { + "epoch": 1.8871339520153936, + "grad_norm": 1.683787226676941, + "learning_rate": 1.6594582031473572e-07, + "loss": 3.0925, + "step": 23539 + }, + { + "epoch": 1.8872141267964162, + "grad_norm": 1.7415927648544312, + "learning_rate": 1.657103369188584e-07, + "loss": 3.1943, + "step": 23540 + }, + { + "epoch": 1.8872943015774388, + "grad_norm": 1.824120044708252, + "learning_rate": 1.6547501932563825e-07, + "loss": 3.2249, + "step": 23541 + }, + { + "epoch": 1.8873744763584615, + "grad_norm": 1.7850812673568726, + "learning_rate": 1.652398675390432e-07, + "loss": 3.164, + "step": 23542 + }, + { + "epoch": 1.887454651139484, + "grad_norm": 1.7457022666931152, + "learning_rate": 1.6500488156303895e-07, + "loss": 3.2615, + "step": 23543 + }, + { + "epoch": 1.8875348259205067, + "grad_norm": 1.756227731704712, + "learning_rate": 1.647700614015857e-07, + "loss": 3.1204, + "step": 23544 + }, + { + "epoch": 1.8876150007015293, + "grad_norm": 1.8168877363204956, + "learning_rate": 1.6453540705864245e-07, + "loss": 3.1473, + "step": 23545 + }, + { + "epoch": 1.887695175482552, + "grad_norm": 1.77947998046875, + "learning_rate": 1.643009185381661e-07, + "loss": 3.2339, + "step": 23546 + }, + { + "epoch": 1.8877753502635746, + "grad_norm": 0.6703505516052246, + "learning_rate": 1.640665958441101e-07, + "loss": 2.4558, + "step": 23547 + }, + { + "epoch": 1.8878555250445972, + "grad_norm": 1.7811030149459839, + "learning_rate": 1.6383243898042466e-07, + "loss": 3.1483, + "step": 23548 + }, + { + "epoch": 1.8879356998256198, + "grad_norm": 1.7251158952713013, + "learning_rate": 1.6359844795105774e-07, + "loss": 3.1071, + "step": 23549 + }, + { + "epoch": 1.8880158746066424, + "grad_norm": 1.7677901983261108, + "learning_rate": 1.633646227599539e-07, + "loss": 3.0945, + "step": 23550 + }, + { + "epoch": 1.888096049387665, + "grad_norm": 1.7103960514068604, + "learning_rate": 1.6313096341105562e-07, + "loss": 3.1853, + "step": 23551 + }, + { + "epoch": 1.8881762241686877, + "grad_norm": 1.800742506980896, + "learning_rate": 1.62897469908303e-07, + "loss": 3.1226, + "step": 23552 + }, + { + "epoch": 1.8882563989497103, + "grad_norm": 1.7823262214660645, + "learning_rate": 1.6266414225563297e-07, + "loss": 3.177, + "step": 23553 + }, + { + "epoch": 1.888336573730733, + "grad_norm": 1.698343276977539, + "learning_rate": 1.624309804569768e-07, + "loss": 3.175, + "step": 23554 + }, + { + "epoch": 1.8884167485117556, + "grad_norm": 1.6996824741363525, + "learning_rate": 1.62197984516268e-07, + "loss": 3.0992, + "step": 23555 + }, + { + "epoch": 1.8884969232927782, + "grad_norm": 0.6594502925872803, + "learning_rate": 1.6196515443743345e-07, + "loss": 2.4601, + "step": 23556 + }, + { + "epoch": 1.8885770980738008, + "grad_norm": 1.7590930461883545, + "learning_rate": 1.6173249022439995e-07, + "loss": 3.1257, + "step": 23557 + }, + { + "epoch": 1.8886572728548234, + "grad_norm": 1.7366546392440796, + "learning_rate": 1.614999918810889e-07, + "loss": 3.2465, + "step": 23558 + }, + { + "epoch": 1.888737447635846, + "grad_norm": 1.7609196901321411, + "learning_rate": 1.612676594114204e-07, + "loss": 3.0643, + "step": 23559 + }, + { + "epoch": 1.8888176224168687, + "grad_norm": 1.8184372186660767, + "learning_rate": 1.610354928193114e-07, + "loss": 3.2073, + "step": 23560 + }, + { + "epoch": 1.8888977971978913, + "grad_norm": 1.8838096857070923, + "learning_rate": 1.6080349210867763e-07, + "loss": 3.2574, + "step": 23561 + }, + { + "epoch": 1.888977971978914, + "grad_norm": 1.8114155530929565, + "learning_rate": 1.6057165728342817e-07, + "loss": 3.2874, + "step": 23562 + }, + { + "epoch": 1.8890581467599366, + "grad_norm": 1.906161904335022, + "learning_rate": 1.6033998834747322e-07, + "loss": 3.2791, + "step": 23563 + }, + { + "epoch": 1.8891383215409592, + "grad_norm": 1.7970483303070068, + "learning_rate": 1.6010848530471857e-07, + "loss": 3.1779, + "step": 23564 + }, + { + "epoch": 1.8892184963219818, + "grad_norm": 1.7747942209243774, + "learning_rate": 1.5987714815906662e-07, + "loss": 3.0491, + "step": 23565 + }, + { + "epoch": 1.8892986711030044, + "grad_norm": 0.6716093420982361, + "learning_rate": 1.5964597691441764e-07, + "loss": 2.4723, + "step": 23566 + }, + { + "epoch": 1.889378845884027, + "grad_norm": 1.714730143547058, + "learning_rate": 1.5941497157466956e-07, + "loss": 3.0925, + "step": 23567 + }, + { + "epoch": 1.8894590206650497, + "grad_norm": 1.7533719539642334, + "learning_rate": 1.591841321437171e-07, + "loss": 3.1067, + "step": 23568 + }, + { + "epoch": 1.8895391954460723, + "grad_norm": 1.7544021606445312, + "learning_rate": 1.5895345862545265e-07, + "loss": 3.1691, + "step": 23569 + }, + { + "epoch": 1.889619370227095, + "grad_norm": 1.8124436140060425, + "learning_rate": 1.587229510237631e-07, + "loss": 3.1064, + "step": 23570 + }, + { + "epoch": 1.8896995450081175, + "grad_norm": 1.6702375411987305, + "learning_rate": 1.5849260934253652e-07, + "loss": 3.2081, + "step": 23571 + }, + { + "epoch": 1.8897797197891402, + "grad_norm": 1.7927216291427612, + "learning_rate": 1.582624335856564e-07, + "loss": 3.1523, + "step": 23572 + }, + { + "epoch": 1.8898598945701628, + "grad_norm": 1.7007339000701904, + "learning_rate": 1.580324237570041e-07, + "loss": 3.1659, + "step": 23573 + }, + { + "epoch": 1.8899400693511856, + "grad_norm": 1.9605519771575928, + "learning_rate": 1.5780257986045544e-07, + "loss": 3.2001, + "step": 23574 + }, + { + "epoch": 1.8900202441322083, + "grad_norm": 0.6406398415565491, + "learning_rate": 1.5757290189988618e-07, + "loss": 2.4603, + "step": 23575 + }, + { + "epoch": 1.8901004189132309, + "grad_norm": 1.7787290811538696, + "learning_rate": 1.5734338987916876e-07, + "loss": 3.1693, + "step": 23576 + }, + { + "epoch": 1.8901805936942535, + "grad_norm": 0.66303950548172, + "learning_rate": 1.5711404380217343e-07, + "loss": 2.4728, + "step": 23577 + }, + { + "epoch": 1.8902607684752761, + "grad_norm": 1.7451235055923462, + "learning_rate": 1.568848636727671e-07, + "loss": 3.1527, + "step": 23578 + }, + { + "epoch": 1.8903409432562988, + "grad_norm": 1.722516655921936, + "learning_rate": 1.5665584949481227e-07, + "loss": 3.151, + "step": 23579 + }, + { + "epoch": 1.8904211180373214, + "grad_norm": 1.9024521112442017, + "learning_rate": 1.5642700127217025e-07, + "loss": 3.1638, + "step": 23580 + }, + { + "epoch": 1.890501292818344, + "grad_norm": 0.6531878709793091, + "learning_rate": 1.561983190087002e-07, + "loss": 2.474, + "step": 23581 + }, + { + "epoch": 1.8905814675993666, + "grad_norm": 1.7117137908935547, + "learning_rate": 1.559698027082579e-07, + "loss": 3.1408, + "step": 23582 + }, + { + "epoch": 1.8906616423803893, + "grad_norm": 1.7806912660598755, + "learning_rate": 1.5574145237469362e-07, + "loss": 3.1395, + "step": 23583 + }, + { + "epoch": 1.8907418171614119, + "grad_norm": 1.7493398189544678, + "learning_rate": 1.5551326801186094e-07, + "loss": 3.1694, + "step": 23584 + }, + { + "epoch": 1.8908219919424345, + "grad_norm": 1.8599958419799805, + "learning_rate": 1.552852496236046e-07, + "loss": 3.2, + "step": 23585 + }, + { + "epoch": 1.8909021667234571, + "grad_norm": 1.8077269792556763, + "learning_rate": 1.550573972137681e-07, + "loss": 3.1292, + "step": 23586 + }, + { + "epoch": 1.8909823415044797, + "grad_norm": 1.7979133129119873, + "learning_rate": 1.5482971078619512e-07, + "loss": 3.2201, + "step": 23587 + }, + { + "epoch": 1.8910625162855024, + "grad_norm": 1.7687383890151978, + "learning_rate": 1.5460219034472367e-07, + "loss": 3.143, + "step": 23588 + }, + { + "epoch": 1.891142691066525, + "grad_norm": 1.7935367822647095, + "learning_rate": 1.543748358931896e-07, + "loss": 3.2285, + "step": 23589 + }, + { + "epoch": 1.8912228658475476, + "grad_norm": 1.8199143409729004, + "learning_rate": 1.541476474354253e-07, + "loss": 3.1685, + "step": 23590 + }, + { + "epoch": 1.8913030406285702, + "grad_norm": 1.7994534969329834, + "learning_rate": 1.5392062497526117e-07, + "loss": 3.1742, + "step": 23591 + }, + { + "epoch": 1.8913832154095929, + "grad_norm": 1.8374513387680054, + "learning_rate": 1.5369376851652517e-07, + "loss": 3.1447, + "step": 23592 + }, + { + "epoch": 1.8914633901906155, + "grad_norm": 1.8325308561325073, + "learning_rate": 1.5346707806304318e-07, + "loss": 3.1907, + "step": 23593 + }, + { + "epoch": 1.8915435649716383, + "grad_norm": 1.8645929098129272, + "learning_rate": 1.5324055361863544e-07, + "loss": 3.1379, + "step": 23594 + }, + { + "epoch": 1.891623739752661, + "grad_norm": 1.818597674369812, + "learning_rate": 1.5301419518712113e-07, + "loss": 3.183, + "step": 23595 + }, + { + "epoch": 1.8917039145336836, + "grad_norm": 1.7139716148376465, + "learning_rate": 1.527880027723183e-07, + "loss": 3.1388, + "step": 23596 + }, + { + "epoch": 1.8917840893147062, + "grad_norm": 1.783684492111206, + "learning_rate": 1.5256197637803838e-07, + "loss": 3.1322, + "step": 23597 + }, + { + "epoch": 1.8918642640957288, + "grad_norm": 1.8053070306777954, + "learning_rate": 1.5233611600809273e-07, + "loss": 3.2646, + "step": 23598 + }, + { + "epoch": 1.8919444388767515, + "grad_norm": 1.802918791770935, + "learning_rate": 1.5211042166629054e-07, + "loss": 3.0769, + "step": 23599 + }, + { + "epoch": 1.892024613657774, + "grad_norm": 1.7720509767532349, + "learning_rate": 1.518848933564343e-07, + "loss": 3.0423, + "step": 23600 + }, + { + "epoch": 1.8921047884387967, + "grad_norm": 1.7407896518707275, + "learning_rate": 1.5165953108232879e-07, + "loss": 3.1821, + "step": 23601 + }, + { + "epoch": 1.8921849632198193, + "grad_norm": 1.6521068811416626, + "learning_rate": 1.5143433484777315e-07, + "loss": 3.2555, + "step": 23602 + }, + { + "epoch": 1.892265138000842, + "grad_norm": 1.861751914024353, + "learning_rate": 1.5120930465656325e-07, + "loss": 3.1519, + "step": 23603 + }, + { + "epoch": 1.8923453127818646, + "grad_norm": 1.7308549880981445, + "learning_rate": 1.5098444051249382e-07, + "loss": 3.1401, + "step": 23604 + }, + { + "epoch": 1.8924254875628872, + "grad_norm": 1.6920409202575684, + "learning_rate": 1.507597424193552e-07, + "loss": 3.2192, + "step": 23605 + }, + { + "epoch": 1.8925056623439098, + "grad_norm": 1.7681218385696411, + "learning_rate": 1.5053521038093654e-07, + "loss": 3.104, + "step": 23606 + }, + { + "epoch": 1.8925858371249324, + "grad_norm": 1.810939073562622, + "learning_rate": 1.503108444010215e-07, + "loss": 3.181, + "step": 23607 + }, + { + "epoch": 1.892666011905955, + "grad_norm": 0.6598049998283386, + "learning_rate": 1.5008664448339595e-07, + "loss": 2.4592, + "step": 23608 + }, + { + "epoch": 1.8927461866869777, + "grad_norm": 1.7752610445022583, + "learning_rate": 1.4986261063183794e-07, + "loss": 3.132, + "step": 23609 + }, + { + "epoch": 1.8928263614680003, + "grad_norm": 1.7093833684921265, + "learning_rate": 1.4963874285012448e-07, + "loss": 3.0793, + "step": 23610 + }, + { + "epoch": 1.892906536249023, + "grad_norm": 1.7023499011993408, + "learning_rate": 1.494150411420292e-07, + "loss": 3.0796, + "step": 23611 + }, + { + "epoch": 1.8929867110300456, + "grad_norm": 1.7894775867462158, + "learning_rate": 1.4919150551132577e-07, + "loss": 3.208, + "step": 23612 + }, + { + "epoch": 1.8930668858110682, + "grad_norm": 1.7498033046722412, + "learning_rate": 1.4896813596178116e-07, + "loss": 3.1398, + "step": 23613 + }, + { + "epoch": 1.8931470605920908, + "grad_norm": 1.7082138061523438, + "learning_rate": 1.487449324971635e-07, + "loss": 3.2363, + "step": 23614 + }, + { + "epoch": 1.8932272353731134, + "grad_norm": 1.713425636291504, + "learning_rate": 1.4852189512123305e-07, + "loss": 3.1652, + "step": 23615 + }, + { + "epoch": 1.893307410154136, + "grad_norm": 1.7227234840393066, + "learning_rate": 1.482990238377513e-07, + "loss": 3.0519, + "step": 23616 + }, + { + "epoch": 1.8933875849351587, + "grad_norm": 1.7891793251037598, + "learning_rate": 1.4807631865047522e-07, + "loss": 3.1705, + "step": 23617 + }, + { + "epoch": 1.8934677597161813, + "grad_norm": 1.7886160612106323, + "learning_rate": 1.4785377956316183e-07, + "loss": 3.2065, + "step": 23618 + }, + { + "epoch": 1.893547934497204, + "grad_norm": 1.7900627851486206, + "learning_rate": 1.4763140657956143e-07, + "loss": 3.2722, + "step": 23619 + }, + { + "epoch": 1.8936281092782266, + "grad_norm": 1.7176191806793213, + "learning_rate": 1.4740919970342216e-07, + "loss": 3.2428, + "step": 23620 + }, + { + "epoch": 1.8937082840592492, + "grad_norm": 1.7432186603546143, + "learning_rate": 1.4718715893849212e-07, + "loss": 3.1282, + "step": 23621 + }, + { + "epoch": 1.8937884588402718, + "grad_norm": 1.6816771030426025, + "learning_rate": 1.4696528428851387e-07, + "loss": 3.0827, + "step": 23622 + }, + { + "epoch": 1.8938686336212944, + "grad_norm": 1.7465169429779053, + "learning_rate": 1.4674357575722776e-07, + "loss": 3.1852, + "step": 23623 + }, + { + "epoch": 1.893948808402317, + "grad_norm": 1.795566201210022, + "learning_rate": 1.4652203334837412e-07, + "loss": 3.0834, + "step": 23624 + }, + { + "epoch": 1.8940289831833397, + "grad_norm": 1.6681315898895264, + "learning_rate": 1.4630065706568442e-07, + "loss": 3.1099, + "step": 23625 + }, + { + "epoch": 1.8941091579643623, + "grad_norm": 1.7601978778839111, + "learning_rate": 1.4607944691289344e-07, + "loss": 3.1266, + "step": 23626 + }, + { + "epoch": 1.894189332745385, + "grad_norm": 1.7520180940628052, + "learning_rate": 1.4585840289373044e-07, + "loss": 3.1095, + "step": 23627 + }, + { + "epoch": 1.8942695075264075, + "grad_norm": 1.712067723274231, + "learning_rate": 1.456375250119213e-07, + "loss": 3.0759, + "step": 23628 + }, + { + "epoch": 1.8943496823074302, + "grad_norm": 1.7490761280059814, + "learning_rate": 1.4541681327119194e-07, + "loss": 3.1909, + "step": 23629 + }, + { + "epoch": 1.8944298570884528, + "grad_norm": 1.815755009651184, + "learning_rate": 1.451962676752605e-07, + "loss": 3.0969, + "step": 23630 + }, + { + "epoch": 1.8945100318694754, + "grad_norm": 1.8218493461608887, + "learning_rate": 1.4497588822784737e-07, + "loss": 3.1681, + "step": 23631 + }, + { + "epoch": 1.894590206650498, + "grad_norm": 1.7860299348831177, + "learning_rate": 1.447556749326673e-07, + "loss": 3.1599, + "step": 23632 + }, + { + "epoch": 1.8946703814315207, + "grad_norm": 1.7976548671722412, + "learning_rate": 1.4453562779343288e-07, + "loss": 3.0873, + "step": 23633 + }, + { + "epoch": 1.8947505562125433, + "grad_norm": 1.8241603374481201, + "learning_rate": 1.4431574681385563e-07, + "loss": 3.1757, + "step": 23634 + }, + { + "epoch": 1.894830730993566, + "grad_norm": 1.6874817609786987, + "learning_rate": 1.4409603199764033e-07, + "loss": 3.1515, + "step": 23635 + }, + { + "epoch": 1.8949109057745885, + "grad_norm": 1.7789926528930664, + "learning_rate": 1.438764833484929e-07, + "loss": 3.2323, + "step": 23636 + }, + { + "epoch": 1.8949910805556112, + "grad_norm": 1.797858715057373, + "learning_rate": 1.4365710087011374e-07, + "loss": 3.1719, + "step": 23637 + }, + { + "epoch": 1.8950712553366338, + "grad_norm": 1.8104369640350342, + "learning_rate": 1.434378845662021e-07, + "loss": 3.2073, + "step": 23638 + }, + { + "epoch": 1.8951514301176564, + "grad_norm": 1.7643085718154907, + "learning_rate": 1.4321883444045503e-07, + "loss": 3.1791, + "step": 23639 + }, + { + "epoch": 1.895231604898679, + "grad_norm": 1.8581722974777222, + "learning_rate": 1.4299995049656401e-07, + "loss": 3.171, + "step": 23640 + }, + { + "epoch": 1.8953117796797017, + "grad_norm": 1.8296706676483154, + "learning_rate": 1.4278123273821942e-07, + "loss": 3.2203, + "step": 23641 + }, + { + "epoch": 1.8953919544607243, + "grad_norm": 1.8956751823425293, + "learning_rate": 1.4256268116910942e-07, + "loss": 3.1096, + "step": 23642 + }, + { + "epoch": 1.895472129241747, + "grad_norm": 1.728779673576355, + "learning_rate": 1.4234429579291776e-07, + "loss": 3.1917, + "step": 23643 + }, + { + "epoch": 1.8955523040227695, + "grad_norm": 1.8446872234344482, + "learning_rate": 1.4212607661332922e-07, + "loss": 3.1383, + "step": 23644 + }, + { + "epoch": 1.8956324788037922, + "grad_norm": 1.7276161909103394, + "learning_rate": 1.4190802363401867e-07, + "loss": 3.1655, + "step": 23645 + }, + { + "epoch": 1.8957126535848148, + "grad_norm": 1.7272002696990967, + "learning_rate": 1.416901368586654e-07, + "loss": 3.1138, + "step": 23646 + }, + { + "epoch": 1.8957928283658374, + "grad_norm": 1.7466497421264648, + "learning_rate": 1.4147241629094199e-07, + "loss": 3.1199, + "step": 23647 + }, + { + "epoch": 1.89587300314686, + "grad_norm": 1.7946199178695679, + "learning_rate": 1.4125486193451888e-07, + "loss": 3.1334, + "step": 23648 + }, + { + "epoch": 1.8959531779278826, + "grad_norm": 1.8203836679458618, + "learning_rate": 1.4103747379306532e-07, + "loss": 3.1721, + "step": 23649 + }, + { + "epoch": 1.8960333527089053, + "grad_norm": 1.8269463777542114, + "learning_rate": 1.4082025187024395e-07, + "loss": 3.167, + "step": 23650 + }, + { + "epoch": 1.896113527489928, + "grad_norm": 1.887634038925171, + "learning_rate": 1.406031961697185e-07, + "loss": 3.1587, + "step": 23651 + }, + { + "epoch": 1.8961937022709505, + "grad_norm": 1.7450364828109741, + "learning_rate": 1.403863066951494e-07, + "loss": 3.1386, + "step": 23652 + }, + { + "epoch": 1.8962738770519731, + "grad_norm": 1.8972235918045044, + "learning_rate": 1.4016958345019149e-07, + "loss": 3.2466, + "step": 23653 + }, + { + "epoch": 1.8963540518329958, + "grad_norm": 1.7832868099212646, + "learning_rate": 1.399530264384996e-07, + "loss": 3.2472, + "step": 23654 + }, + { + "epoch": 1.8964342266140186, + "grad_norm": 1.8180032968521118, + "learning_rate": 1.397366356637253e-07, + "loss": 3.1825, + "step": 23655 + }, + { + "epoch": 1.8965144013950412, + "grad_norm": 1.796406865119934, + "learning_rate": 1.3952041112951566e-07, + "loss": 3.2211, + "step": 23656 + }, + { + "epoch": 1.8965945761760639, + "grad_norm": 1.8223304748535156, + "learning_rate": 1.393043528395166e-07, + "loss": 3.1964, + "step": 23657 + }, + { + "epoch": 1.8966747509570865, + "grad_norm": 1.7852108478546143, + "learning_rate": 1.390884607973708e-07, + "loss": 3.1992, + "step": 23658 + }, + { + "epoch": 1.896754925738109, + "grad_norm": 1.7746059894561768, + "learning_rate": 1.388727350067187e-07, + "loss": 3.0873, + "step": 23659 + }, + { + "epoch": 1.8968351005191317, + "grad_norm": 1.743185043334961, + "learning_rate": 1.3865717547119627e-07, + "loss": 3.2013, + "step": 23660 + }, + { + "epoch": 1.8969152753001544, + "grad_norm": 1.688992977142334, + "learning_rate": 1.3844178219443948e-07, + "loss": 3.2218, + "step": 23661 + }, + { + "epoch": 1.896995450081177, + "grad_norm": 1.8049753904342651, + "learning_rate": 1.3822655518007765e-07, + "loss": 3.2913, + "step": 23662 + }, + { + "epoch": 1.8970756248621996, + "grad_norm": 1.7252297401428223, + "learning_rate": 1.380114944317401e-07, + "loss": 3.14, + "step": 23663 + }, + { + "epoch": 1.8971557996432222, + "grad_norm": 1.8361228704452515, + "learning_rate": 1.3779659995305506e-07, + "loss": 3.1605, + "step": 23664 + }, + { + "epoch": 1.8972359744242449, + "grad_norm": 1.8124754428863525, + "learning_rate": 1.375818717476418e-07, + "loss": 3.1904, + "step": 23665 + }, + { + "epoch": 1.8973161492052675, + "grad_norm": 1.7328349351882935, + "learning_rate": 1.3736730981912305e-07, + "loss": 3.1272, + "step": 23666 + }, + { + "epoch": 1.89739632398629, + "grad_norm": 1.7709264755249023, + "learning_rate": 1.3715291417111475e-07, + "loss": 3.1176, + "step": 23667 + }, + { + "epoch": 1.8974764987673127, + "grad_norm": 1.6632559299468994, + "learning_rate": 1.3693868480723293e-07, + "loss": 3.1562, + "step": 23668 + }, + { + "epoch": 1.8975566735483353, + "grad_norm": 1.8944652080535889, + "learning_rate": 1.3672462173108915e-07, + "loss": 3.2552, + "step": 23669 + }, + { + "epoch": 1.897636848329358, + "grad_norm": 1.7746093273162842, + "learning_rate": 1.3651072494629158e-07, + "loss": 3.1481, + "step": 23670 + }, + { + "epoch": 1.8977170231103806, + "grad_norm": 1.6685373783111572, + "learning_rate": 1.3629699445644628e-07, + "loss": 3.2307, + "step": 23671 + }, + { + "epoch": 1.8977971978914032, + "grad_norm": 1.7978127002716064, + "learning_rate": 1.3608343026515814e-07, + "loss": 3.2274, + "step": 23672 + }, + { + "epoch": 1.8978773726724258, + "grad_norm": 1.6909857988357544, + "learning_rate": 1.358700323760276e-07, + "loss": 3.0967, + "step": 23673 + }, + { + "epoch": 1.8979575474534485, + "grad_norm": 1.857908010482788, + "learning_rate": 1.3565680079265176e-07, + "loss": 3.2183, + "step": 23674 + }, + { + "epoch": 1.898037722234471, + "grad_norm": 1.777782917022705, + "learning_rate": 1.3544373551862445e-07, + "loss": 3.2775, + "step": 23675 + }, + { + "epoch": 1.898117897015494, + "grad_norm": 1.785399317741394, + "learning_rate": 1.3523083655753944e-07, + "loss": 3.1307, + "step": 23676 + }, + { + "epoch": 1.8981980717965166, + "grad_norm": 1.7842055559158325, + "learning_rate": 1.350181039129872e-07, + "loss": 3.2007, + "step": 23677 + }, + { + "epoch": 1.8982782465775392, + "grad_norm": 1.7216464281082153, + "learning_rate": 1.3480553758855154e-07, + "loss": 3.1298, + "step": 23678 + }, + { + "epoch": 1.8983584213585618, + "grad_norm": 1.815267562866211, + "learning_rate": 1.3459313758781843e-07, + "loss": 3.1823, + "step": 23679 + }, + { + "epoch": 1.8984385961395844, + "grad_norm": 1.914174199104309, + "learning_rate": 1.3438090391436842e-07, + "loss": 3.2368, + "step": 23680 + }, + { + "epoch": 1.898518770920607, + "grad_norm": 1.654496431350708, + "learning_rate": 1.3416883657177858e-07, + "loss": 3.1191, + "step": 23681 + }, + { + "epoch": 1.8985989457016297, + "grad_norm": 1.7230944633483887, + "learning_rate": 1.33956935563625e-07, + "loss": 3.1636, + "step": 23682 + }, + { + "epoch": 1.8986791204826523, + "grad_norm": 1.7901217937469482, + "learning_rate": 1.337452008934803e-07, + "loss": 3.1605, + "step": 23683 + }, + { + "epoch": 1.898759295263675, + "grad_norm": 1.7454864978790283, + "learning_rate": 1.3353363256491502e-07, + "loss": 3.1708, + "step": 23684 + }, + { + "epoch": 1.8988394700446976, + "grad_norm": 1.7444692850112915, + "learning_rate": 1.3332223058149518e-07, + "loss": 3.1752, + "step": 23685 + }, + { + "epoch": 1.8989196448257202, + "grad_norm": 1.6984902620315552, + "learning_rate": 1.331109949467857e-07, + "loss": 3.031, + "step": 23686 + }, + { + "epoch": 1.8989998196067428, + "grad_norm": 1.8410779237747192, + "learning_rate": 1.3289992566434596e-07, + "loss": 3.2042, + "step": 23687 + }, + { + "epoch": 1.8990799943877654, + "grad_norm": 1.8208916187286377, + "learning_rate": 1.3268902273773754e-07, + "loss": 3.1517, + "step": 23688 + }, + { + "epoch": 1.899160169168788, + "grad_norm": 1.796663522720337, + "learning_rate": 1.3247828617051428e-07, + "loss": 3.2046, + "step": 23689 + }, + { + "epoch": 1.8992403439498107, + "grad_norm": 1.7937477827072144, + "learning_rate": 1.3226771596622888e-07, + "loss": 3.0923, + "step": 23690 + }, + { + "epoch": 1.8993205187308333, + "grad_norm": 1.7905895709991455, + "learning_rate": 1.3205731212843187e-07, + "loss": 3.1472, + "step": 23691 + }, + { + "epoch": 1.899400693511856, + "grad_norm": 1.7609063386917114, + "learning_rate": 1.318470746606715e-07, + "loss": 3.2189, + "step": 23692 + }, + { + "epoch": 1.8994808682928785, + "grad_norm": 1.7439106702804565, + "learning_rate": 1.3163700356649046e-07, + "loss": 3.0839, + "step": 23693 + }, + { + "epoch": 1.8995610430739012, + "grad_norm": 1.7003810405731201, + "learning_rate": 1.3142709884943372e-07, + "loss": 3.1326, + "step": 23694 + }, + { + "epoch": 1.8996412178549238, + "grad_norm": 1.8125298023223877, + "learning_rate": 1.3121736051303625e-07, + "loss": 3.134, + "step": 23695 + }, + { + "epoch": 1.8997213926359464, + "grad_norm": 1.7925853729248047, + "learning_rate": 1.310077885608363e-07, + "loss": 3.152, + "step": 23696 + }, + { + "epoch": 1.899801567416969, + "grad_norm": 1.8124934434890747, + "learning_rate": 1.307983829963666e-07, + "loss": 3.1974, + "step": 23697 + }, + { + "epoch": 1.8998817421979917, + "grad_norm": 1.76249098777771, + "learning_rate": 1.305891438231588e-07, + "loss": 3.1478, + "step": 23698 + }, + { + "epoch": 1.8999619169790143, + "grad_norm": 1.7815005779266357, + "learning_rate": 1.3038007104474004e-07, + "loss": 3.1818, + "step": 23699 + }, + { + "epoch": 1.900042091760037, + "grad_norm": 1.7433336973190308, + "learning_rate": 1.301711646646342e-07, + "loss": 3.1171, + "step": 23700 + }, + { + "epoch": 1.9001222665410595, + "grad_norm": 1.7510998249053955, + "learning_rate": 1.2996242468636288e-07, + "loss": 3.1805, + "step": 23701 + }, + { + "epoch": 1.9002024413220822, + "grad_norm": 1.779782772064209, + "learning_rate": 1.2975385111344773e-07, + "loss": 3.2328, + "step": 23702 + }, + { + "epoch": 1.9002826161031048, + "grad_norm": 1.6848503351211548, + "learning_rate": 1.2954544394940371e-07, + "loss": 3.2579, + "step": 23703 + }, + { + "epoch": 1.9003627908841274, + "grad_norm": 1.7326620817184448, + "learning_rate": 1.2933720319774578e-07, + "loss": 3.1945, + "step": 23704 + }, + { + "epoch": 1.90044296566515, + "grad_norm": 1.789592981338501, + "learning_rate": 1.2912912886198338e-07, + "loss": 3.1179, + "step": 23705 + }, + { + "epoch": 1.9005231404461727, + "grad_norm": 1.7545056343078613, + "learning_rate": 1.2892122094562475e-07, + "loss": 3.1188, + "step": 23706 + }, + { + "epoch": 1.9006033152271953, + "grad_norm": 1.8087085485458374, + "learning_rate": 1.2871347945217495e-07, + "loss": 3.2552, + "step": 23707 + }, + { + "epoch": 1.900683490008218, + "grad_norm": 1.8693796396255493, + "learning_rate": 1.2850590438513778e-07, + "loss": 3.1722, + "step": 23708 + }, + { + "epoch": 1.9007636647892405, + "grad_norm": 1.7884674072265625, + "learning_rate": 1.2829849574801156e-07, + "loss": 3.2153, + "step": 23709 + }, + { + "epoch": 1.9008438395702631, + "grad_norm": 1.7786203622817993, + "learning_rate": 1.2809125354429463e-07, + "loss": 3.1608, + "step": 23710 + }, + { + "epoch": 1.9009240143512858, + "grad_norm": 1.9931297302246094, + "learning_rate": 1.2788417777747864e-07, + "loss": 3.1834, + "step": 23711 + }, + { + "epoch": 1.9010041891323084, + "grad_norm": 1.814565658569336, + "learning_rate": 1.2767726845105633e-07, + "loss": 3.2645, + "step": 23712 + }, + { + "epoch": 1.901084363913331, + "grad_norm": 1.8593086004257202, + "learning_rate": 1.2747052556851603e-07, + "loss": 3.2576, + "step": 23713 + }, + { + "epoch": 1.9011645386943536, + "grad_norm": 1.8143432140350342, + "learning_rate": 1.2726394913334384e-07, + "loss": 3.1693, + "step": 23714 + }, + { + "epoch": 1.9012447134753763, + "grad_norm": 1.8098721504211426, + "learning_rate": 1.270575391490214e-07, + "loss": 3.2383, + "step": 23715 + }, + { + "epoch": 1.901324888256399, + "grad_norm": 1.6992847919464111, + "learning_rate": 1.2685129561902932e-07, + "loss": 3.179, + "step": 23716 + }, + { + "epoch": 1.9014050630374215, + "grad_norm": 1.6804417371749878, + "learning_rate": 1.2664521854684474e-07, + "loss": 3.1344, + "step": 23717 + }, + { + "epoch": 1.9014852378184441, + "grad_norm": 1.7153791189193726, + "learning_rate": 1.2643930793594162e-07, + "loss": 3.2467, + "step": 23718 + }, + { + "epoch": 1.9015654125994668, + "grad_norm": 0.6745122671127319, + "learning_rate": 1.2623356378979268e-07, + "loss": 2.4594, + "step": 23719 + }, + { + "epoch": 1.9016455873804894, + "grad_norm": 0.6460968255996704, + "learning_rate": 1.260279861118663e-07, + "loss": 2.4672, + "step": 23720 + }, + { + "epoch": 1.901725762161512, + "grad_norm": 1.7698885202407837, + "learning_rate": 1.2582257490562745e-07, + "loss": 3.074, + "step": 23721 + }, + { + "epoch": 1.9018059369425346, + "grad_norm": 1.8042179346084595, + "learning_rate": 1.2561733017454002e-07, + "loss": 3.1257, + "step": 23722 + }, + { + "epoch": 1.9018861117235573, + "grad_norm": 1.7581453323364258, + "learning_rate": 1.2541225192206463e-07, + "loss": 3.134, + "step": 23723 + }, + { + "epoch": 1.9019662865045799, + "grad_norm": 1.824811339378357, + "learning_rate": 1.2520734015165848e-07, + "loss": 3.2584, + "step": 23724 + }, + { + "epoch": 1.9020464612856025, + "grad_norm": 1.6582694053649902, + "learning_rate": 1.2500259486677658e-07, + "loss": 3.1885, + "step": 23725 + }, + { + "epoch": 1.9021266360666251, + "grad_norm": 0.6559280753135681, + "learning_rate": 1.2479801607087062e-07, + "loss": 2.4747, + "step": 23726 + }, + { + "epoch": 1.9022068108476478, + "grad_norm": 1.6838334798812866, + "learning_rate": 1.2459360376738893e-07, + "loss": 3.1649, + "step": 23727 + }, + { + "epoch": 1.9022869856286704, + "grad_norm": 1.7767109870910645, + "learning_rate": 1.243893579597788e-07, + "loss": 3.1538, + "step": 23728 + }, + { + "epoch": 1.902367160409693, + "grad_norm": 1.8335309028625488, + "learning_rate": 1.2418527865148523e-07, + "loss": 3.2655, + "step": 23729 + }, + { + "epoch": 1.9024473351907156, + "grad_norm": 1.891194224357605, + "learning_rate": 1.2398136584594544e-07, + "loss": 3.2047, + "step": 23730 + }, + { + "epoch": 1.9025275099717383, + "grad_norm": 1.8621232509613037, + "learning_rate": 1.2377761954660005e-07, + "loss": 3.2734, + "step": 23731 + }, + { + "epoch": 1.9026076847527609, + "grad_norm": 1.7256896495819092, + "learning_rate": 1.2357403975688297e-07, + "loss": 3.1375, + "step": 23732 + }, + { + "epoch": 1.9026878595337835, + "grad_norm": 1.7451387643814087, + "learning_rate": 1.2337062648022702e-07, + "loss": 3.1654, + "step": 23733 + }, + { + "epoch": 1.9027680343148061, + "grad_norm": 1.696755051612854, + "learning_rate": 1.2316737972006166e-07, + "loss": 3.2021, + "step": 23734 + }, + { + "epoch": 1.9028482090958287, + "grad_norm": 1.8071224689483643, + "learning_rate": 1.2296429947981302e-07, + "loss": 3.1605, + "step": 23735 + }, + { + "epoch": 1.9029283838768514, + "grad_norm": 1.8038660287857056, + "learning_rate": 1.227613857629051e-07, + "loss": 3.1906, + "step": 23736 + }, + { + "epoch": 1.9030085586578742, + "grad_norm": 1.8369499444961548, + "learning_rate": 1.225586385727584e-07, + "loss": 3.2474, + "step": 23737 + }, + { + "epoch": 1.9030887334388968, + "grad_norm": 1.7521908283233643, + "learning_rate": 1.223560579127925e-07, + "loss": 3.1002, + "step": 23738 + }, + { + "epoch": 1.9031689082199195, + "grad_norm": 1.7889825105667114, + "learning_rate": 1.2215364378642348e-07, + "loss": 3.2305, + "step": 23739 + }, + { + "epoch": 1.903249083000942, + "grad_norm": 1.7995458841323853, + "learning_rate": 1.21951396197062e-07, + "loss": 3.1722, + "step": 23740 + }, + { + "epoch": 1.9033292577819647, + "grad_norm": 0.6575941443443298, + "learning_rate": 1.217493151481175e-07, + "loss": 2.4738, + "step": 23741 + }, + { + "epoch": 1.9034094325629873, + "grad_norm": 1.873461365699768, + "learning_rate": 1.2154740064299954e-07, + "loss": 3.202, + "step": 23742 + }, + { + "epoch": 1.90348960734401, + "grad_norm": 1.8291730880737305, + "learning_rate": 1.2134565268510978e-07, + "loss": 3.1677, + "step": 23743 + }, + { + "epoch": 1.9035697821250326, + "grad_norm": 1.8121460676193237, + "learning_rate": 1.2114407127785223e-07, + "loss": 3.1404, + "step": 23744 + }, + { + "epoch": 1.9036499569060552, + "grad_norm": 1.6717594861984253, + "learning_rate": 1.2094265642462299e-07, + "loss": 3.1841, + "step": 23745 + }, + { + "epoch": 1.9037301316870778, + "grad_norm": 1.8756355047225952, + "learning_rate": 1.2074140812881829e-07, + "loss": 3.2366, + "step": 23746 + }, + { + "epoch": 1.9038103064681005, + "grad_norm": 1.8295307159423828, + "learning_rate": 1.2054032639383206e-07, + "loss": 3.1343, + "step": 23747 + }, + { + "epoch": 1.903890481249123, + "grad_norm": 0.6478479504585266, + "learning_rate": 1.2033941122305492e-07, + "loss": 2.4623, + "step": 23748 + }, + { + "epoch": 1.9039706560301457, + "grad_norm": 1.7988747358322144, + "learning_rate": 1.2013866261987306e-07, + "loss": 3.1075, + "step": 23749 + }, + { + "epoch": 1.9040508308111683, + "grad_norm": 1.758408784866333, + "learning_rate": 1.1993808058767154e-07, + "loss": 3.2091, + "step": 23750 + }, + { + "epoch": 1.904131005592191, + "grad_norm": 1.8484015464782715, + "learning_rate": 1.1973766512983097e-07, + "loss": 3.1992, + "step": 23751 + }, + { + "epoch": 1.9042111803732136, + "grad_norm": 1.9083937406539917, + "learning_rate": 1.1953741624973093e-07, + "loss": 3.1306, + "step": 23752 + }, + { + "epoch": 1.9042913551542362, + "grad_norm": 1.7937043905258179, + "learning_rate": 1.1933733395074863e-07, + "loss": 3.0878, + "step": 23753 + }, + { + "epoch": 1.9043715299352588, + "grad_norm": 1.706532597541809, + "learning_rate": 1.1913741823625702e-07, + "loss": 3.1437, + "step": 23754 + }, + { + "epoch": 1.9044517047162814, + "grad_norm": 1.8586077690124512, + "learning_rate": 1.1893766910962556e-07, + "loss": 3.1949, + "step": 23755 + }, + { + "epoch": 1.904531879497304, + "grad_norm": 1.885616660118103, + "learning_rate": 1.187380865742227e-07, + "loss": 3.1702, + "step": 23756 + }, + { + "epoch": 1.904612054278327, + "grad_norm": 1.768962025642395, + "learning_rate": 1.1853867063341351e-07, + "loss": 3.1112, + "step": 23757 + }, + { + "epoch": 1.9046922290593495, + "grad_norm": 1.8277446031570435, + "learning_rate": 1.1833942129055975e-07, + "loss": 3.1932, + "step": 23758 + }, + { + "epoch": 1.9047724038403722, + "grad_norm": 1.8234295845031738, + "learning_rate": 1.1814033854902095e-07, + "loss": 3.1384, + "step": 23759 + }, + { + "epoch": 1.9048525786213948, + "grad_norm": 1.76853346824646, + "learning_rate": 1.1794142241215222e-07, + "loss": 3.2485, + "step": 23760 + }, + { + "epoch": 1.9049327534024174, + "grad_norm": 1.684178113937378, + "learning_rate": 1.1774267288330865e-07, + "loss": 3.1013, + "step": 23761 + }, + { + "epoch": 1.90501292818344, + "grad_norm": 1.8216551542282104, + "learning_rate": 1.1754408996584087e-07, + "loss": 3.1158, + "step": 23762 + }, + { + "epoch": 1.9050931029644627, + "grad_norm": 1.715286374092102, + "learning_rate": 1.1734567366309735e-07, + "loss": 3.2188, + "step": 23763 + }, + { + "epoch": 1.9051732777454853, + "grad_norm": 1.8088442087173462, + "learning_rate": 1.1714742397842316e-07, + "loss": 3.1472, + "step": 23764 + }, + { + "epoch": 1.905253452526508, + "grad_norm": 1.737809181213379, + "learning_rate": 1.1694934091515897e-07, + "loss": 3.1784, + "step": 23765 + }, + { + "epoch": 1.9053336273075305, + "grad_norm": 0.6426467299461365, + "learning_rate": 1.1675142447664656e-07, + "loss": 2.4587, + "step": 23766 + }, + { + "epoch": 1.9054138020885532, + "grad_norm": 1.718289852142334, + "learning_rate": 1.1655367466622214e-07, + "loss": 3.0867, + "step": 23767 + }, + { + "epoch": 1.9054939768695758, + "grad_norm": 1.7386056184768677, + "learning_rate": 1.163560914872186e-07, + "loss": 3.1469, + "step": 23768 + }, + { + "epoch": 1.9055741516505984, + "grad_norm": 1.7600270509719849, + "learning_rate": 1.1615867494296995e-07, + "loss": 3.2032, + "step": 23769 + }, + { + "epoch": 1.905654326431621, + "grad_norm": 1.8896232843399048, + "learning_rate": 1.159614250368013e-07, + "loss": 3.1642, + "step": 23770 + }, + { + "epoch": 1.9057345012126437, + "grad_norm": 1.7909671068191528, + "learning_rate": 1.1576434177203888e-07, + "loss": 3.1868, + "step": 23771 + }, + { + "epoch": 1.9058146759936663, + "grad_norm": 1.608078122138977, + "learning_rate": 1.1556742515200669e-07, + "loss": 3.0517, + "step": 23772 + }, + { + "epoch": 1.905894850774689, + "grad_norm": 1.7029954195022583, + "learning_rate": 1.1537067518002431e-07, + "loss": 3.2089, + "step": 23773 + }, + { + "epoch": 1.9059750255557115, + "grad_norm": 1.7892792224884033, + "learning_rate": 1.1517409185940909e-07, + "loss": 3.1263, + "step": 23774 + }, + { + "epoch": 1.9060552003367341, + "grad_norm": 1.7942805290222168, + "learning_rate": 1.1497767519347392e-07, + "loss": 3.1573, + "step": 23775 + }, + { + "epoch": 1.9061353751177568, + "grad_norm": 1.719781517982483, + "learning_rate": 1.1478142518553171e-07, + "loss": 3.0977, + "step": 23776 + }, + { + "epoch": 1.9062155498987794, + "grad_norm": 1.910577416419983, + "learning_rate": 1.1458534183889092e-07, + "loss": 3.2136, + "step": 23777 + }, + { + "epoch": 1.906295724679802, + "grad_norm": 1.8382024765014648, + "learning_rate": 1.143894251568567e-07, + "loss": 3.2375, + "step": 23778 + }, + { + "epoch": 1.9063758994608246, + "grad_norm": 1.7892063856124878, + "learning_rate": 1.1419367514273417e-07, + "loss": 3.0955, + "step": 23779 + }, + { + "epoch": 1.9064560742418473, + "grad_norm": 1.7522283792495728, + "learning_rate": 1.1399809179982179e-07, + "loss": 3.1409, + "step": 23780 + }, + { + "epoch": 1.9065362490228699, + "grad_norm": 1.749878168106079, + "learning_rate": 1.1380267513141585e-07, + "loss": 3.1004, + "step": 23781 + }, + { + "epoch": 1.9066164238038925, + "grad_norm": 1.8212637901306152, + "learning_rate": 1.1360742514081368e-07, + "loss": 3.1415, + "step": 23782 + }, + { + "epoch": 1.9066965985849151, + "grad_norm": 1.7387226819992065, + "learning_rate": 1.1341234183130601e-07, + "loss": 3.1967, + "step": 23783 + }, + { + "epoch": 1.9067767733659378, + "grad_norm": 1.71195650100708, + "learning_rate": 1.132174252061824e-07, + "loss": 3.204, + "step": 23784 + }, + { + "epoch": 1.9068569481469604, + "grad_norm": 1.7243324518203735, + "learning_rate": 1.1302267526872801e-07, + "loss": 3.1039, + "step": 23785 + }, + { + "epoch": 1.906937122927983, + "grad_norm": 1.8223326206207275, + "learning_rate": 1.1282809202222688e-07, + "loss": 3.2131, + "step": 23786 + }, + { + "epoch": 1.9070172977090056, + "grad_norm": 1.7225008010864258, + "learning_rate": 1.1263367546995974e-07, + "loss": 3.1367, + "step": 23787 + }, + { + "epoch": 1.9070974724900283, + "grad_norm": 1.7576665878295898, + "learning_rate": 1.1243942561520393e-07, + "loss": 3.1226, + "step": 23788 + }, + { + "epoch": 1.9071776472710509, + "grad_norm": 0.6446647644042969, + "learning_rate": 1.1224534246123465e-07, + "loss": 2.4586, + "step": 23789 + }, + { + "epoch": 1.9072578220520735, + "grad_norm": 1.8118188381195068, + "learning_rate": 1.1205142601132479e-07, + "loss": 3.1467, + "step": 23790 + }, + { + "epoch": 1.9073379968330961, + "grad_norm": 1.7668566703796387, + "learning_rate": 1.1185767626874178e-07, + "loss": 3.1371, + "step": 23791 + }, + { + "epoch": 1.9074181716141188, + "grad_norm": 1.8386369943618774, + "learning_rate": 1.1166409323675409e-07, + "loss": 3.2312, + "step": 23792 + }, + { + "epoch": 1.9074983463951414, + "grad_norm": 1.6929506063461304, + "learning_rate": 1.1147067691862467e-07, + "loss": 3.1122, + "step": 23793 + }, + { + "epoch": 1.907578521176164, + "grad_norm": 1.7104692459106445, + "learning_rate": 1.1127742731761537e-07, + "loss": 3.2449, + "step": 23794 + }, + { + "epoch": 1.9076586959571866, + "grad_norm": 1.827953815460205, + "learning_rate": 1.1108434443698247e-07, + "loss": 3.2487, + "step": 23795 + }, + { + "epoch": 1.9077388707382092, + "grad_norm": 1.7332963943481445, + "learning_rate": 1.1089142827998223e-07, + "loss": 3.0948, + "step": 23796 + }, + { + "epoch": 1.9078190455192319, + "grad_norm": 1.8637936115264893, + "learning_rate": 1.1069867884986762e-07, + "loss": 3.0824, + "step": 23797 + }, + { + "epoch": 1.9078992203002545, + "grad_norm": 1.7545452117919922, + "learning_rate": 1.1050609614988828e-07, + "loss": 3.2153, + "step": 23798 + }, + { + "epoch": 1.9079793950812771, + "grad_norm": 1.7160217761993408, + "learning_rate": 1.1031368018329159e-07, + "loss": 3.1534, + "step": 23799 + }, + { + "epoch": 1.9080595698622997, + "grad_norm": 0.6544125080108643, + "learning_rate": 1.1012143095331939e-07, + "loss": 2.4607, + "step": 23800 + }, + { + "epoch": 1.9081397446433224, + "grad_norm": 1.7691351175308228, + "learning_rate": 1.0992934846321468e-07, + "loss": 3.1772, + "step": 23801 + }, + { + "epoch": 1.908219919424345, + "grad_norm": 1.6776812076568604, + "learning_rate": 1.0973743271621484e-07, + "loss": 3.2027, + "step": 23802 + }, + { + "epoch": 1.9083000942053676, + "grad_norm": 1.7571594715118408, + "learning_rate": 1.095456837155573e-07, + "loss": 3.1636, + "step": 23803 + }, + { + "epoch": 1.9083802689863902, + "grad_norm": 1.9656789302825928, + "learning_rate": 1.0935410146447389e-07, + "loss": 3.3398, + "step": 23804 + }, + { + "epoch": 1.9084604437674129, + "grad_norm": 0.6456148028373718, + "learning_rate": 1.0916268596619428e-07, + "loss": 2.455, + "step": 23805 + }, + { + "epoch": 1.9085406185484355, + "grad_norm": 1.7813427448272705, + "learning_rate": 1.0897143722394477e-07, + "loss": 3.2157, + "step": 23806 + }, + { + "epoch": 1.908620793329458, + "grad_norm": 0.6741450428962708, + "learning_rate": 1.0878035524095165e-07, + "loss": 2.4841, + "step": 23807 + }, + { + "epoch": 1.9087009681104807, + "grad_norm": 2.1455156803131104, + "learning_rate": 1.085894400204357e-07, + "loss": 3.175, + "step": 23808 + }, + { + "epoch": 1.9087811428915034, + "grad_norm": 1.807861328125, + "learning_rate": 1.0839869156561655e-07, + "loss": 3.1571, + "step": 23809 + }, + { + "epoch": 1.908861317672526, + "grad_norm": 1.8468345403671265, + "learning_rate": 1.0820810987970942e-07, + "loss": 3.2649, + "step": 23810 + }, + { + "epoch": 1.9089414924535486, + "grad_norm": 1.837408423423767, + "learning_rate": 1.0801769496592618e-07, + "loss": 3.1304, + "step": 23811 + }, + { + "epoch": 1.9090216672345712, + "grad_norm": 1.6642022132873535, + "learning_rate": 1.078274468274787e-07, + "loss": 3.1737, + "step": 23812 + }, + { + "epoch": 1.9091018420155939, + "grad_norm": 1.8825557231903076, + "learning_rate": 1.0763736546757442e-07, + "loss": 3.209, + "step": 23813 + }, + { + "epoch": 1.9091820167966165, + "grad_norm": 0.6553968787193298, + "learning_rate": 1.0744745088941743e-07, + "loss": 2.4723, + "step": 23814 + }, + { + "epoch": 1.909262191577639, + "grad_norm": 1.7592047452926636, + "learning_rate": 1.0725770309621075e-07, + "loss": 3.1654, + "step": 23815 + }, + { + "epoch": 1.9093423663586617, + "grad_norm": 1.7161180973052979, + "learning_rate": 1.070681220911518e-07, + "loss": 3.1162, + "step": 23816 + }, + { + "epoch": 1.9094225411396843, + "grad_norm": 1.7462064027786255, + "learning_rate": 1.0687870787743693e-07, + "loss": 3.1989, + "step": 23817 + }, + { + "epoch": 1.9095027159207072, + "grad_norm": 1.8392261266708374, + "learning_rate": 1.0668946045826134e-07, + "loss": 3.0768, + "step": 23818 + }, + { + "epoch": 1.9095828907017298, + "grad_norm": 1.8265469074249268, + "learning_rate": 1.0650037983681583e-07, + "loss": 3.1285, + "step": 23819 + }, + { + "epoch": 1.9096630654827524, + "grad_norm": 1.7359673976898193, + "learning_rate": 1.0631146601628561e-07, + "loss": 3.1649, + "step": 23820 + }, + { + "epoch": 1.909743240263775, + "grad_norm": 1.8088805675506592, + "learning_rate": 1.0612271899985704e-07, + "loss": 3.1161, + "step": 23821 + }, + { + "epoch": 1.9098234150447977, + "grad_norm": 1.8870352506637573, + "learning_rate": 1.0593413879071313e-07, + "loss": 3.1717, + "step": 23822 + }, + { + "epoch": 1.9099035898258203, + "grad_norm": 1.898169994354248, + "learning_rate": 1.0574572539203243e-07, + "loss": 3.2468, + "step": 23823 + }, + { + "epoch": 1.909983764606843, + "grad_norm": 1.7779957056045532, + "learning_rate": 1.0555747880699241e-07, + "loss": 3.1681, + "step": 23824 + }, + { + "epoch": 1.9100639393878656, + "grad_norm": 1.779228687286377, + "learning_rate": 1.0536939903876498e-07, + "loss": 3.1306, + "step": 23825 + }, + { + "epoch": 1.9101441141688882, + "grad_norm": 1.777232050895691, + "learning_rate": 1.0518148609052315e-07, + "loss": 3.2093, + "step": 23826 + }, + { + "epoch": 1.9102242889499108, + "grad_norm": 1.7792730331420898, + "learning_rate": 1.0499373996543328e-07, + "loss": 3.027, + "step": 23827 + }, + { + "epoch": 1.9103044637309334, + "grad_norm": 1.6934629678726196, + "learning_rate": 1.0480616066666283e-07, + "loss": 3.1253, + "step": 23828 + }, + { + "epoch": 1.910384638511956, + "grad_norm": 0.6615308523178101, + "learning_rate": 1.046187481973726e-07, + "loss": 2.4624, + "step": 23829 + }, + { + "epoch": 1.9104648132929787, + "grad_norm": 1.7507219314575195, + "learning_rate": 1.0443150256072232e-07, + "loss": 3.2074, + "step": 23830 + }, + { + "epoch": 1.9105449880740013, + "grad_norm": 1.7472038269042969, + "learning_rate": 1.0424442375986943e-07, + "loss": 3.1731, + "step": 23831 + }, + { + "epoch": 1.910625162855024, + "grad_norm": 1.7246198654174805, + "learning_rate": 1.0405751179796919e-07, + "loss": 3.1976, + "step": 23832 + }, + { + "epoch": 1.9107053376360466, + "grad_norm": 1.7912184000015259, + "learning_rate": 1.038707666781702e-07, + "loss": 3.1052, + "step": 23833 + }, + { + "epoch": 1.9107855124170692, + "grad_norm": 1.74513578414917, + "learning_rate": 1.0368418840362327e-07, + "loss": 3.1238, + "step": 23834 + }, + { + "epoch": 1.9108656871980918, + "grad_norm": 1.701567530632019, + "learning_rate": 1.0349777697747254e-07, + "loss": 3.1428, + "step": 23835 + }, + { + "epoch": 1.9109458619791144, + "grad_norm": 1.8351507186889648, + "learning_rate": 1.0331153240286107e-07, + "loss": 3.1681, + "step": 23836 + }, + { + "epoch": 1.911026036760137, + "grad_norm": 1.6336860656738281, + "learning_rate": 1.0312545468292967e-07, + "loss": 3.1378, + "step": 23837 + }, + { + "epoch": 1.91110621154116, + "grad_norm": 1.7878621816635132, + "learning_rate": 1.0293954382081584e-07, + "loss": 3.1876, + "step": 23838 + }, + { + "epoch": 1.9111863863221825, + "grad_norm": 1.80190110206604, + "learning_rate": 1.0275379981965261e-07, + "loss": 3.2659, + "step": 23839 + }, + { + "epoch": 1.9112665611032051, + "grad_norm": 1.873775601387024, + "learning_rate": 1.0256822268257194e-07, + "loss": 3.1322, + "step": 23840 + }, + { + "epoch": 1.9113467358842278, + "grad_norm": 1.8267842531204224, + "learning_rate": 1.0238281241270353e-07, + "loss": 3.0502, + "step": 23841 + }, + { + "epoch": 1.9114269106652504, + "grad_norm": 1.8473011255264282, + "learning_rate": 1.0219756901317157e-07, + "loss": 3.2322, + "step": 23842 + }, + { + "epoch": 1.911507085446273, + "grad_norm": 0.6461278200149536, + "learning_rate": 1.0201249248710132e-07, + "loss": 2.467, + "step": 23843 + }, + { + "epoch": 1.9115872602272956, + "grad_norm": 1.8021684885025024, + "learning_rate": 1.0182758283761141e-07, + "loss": 3.1237, + "step": 23844 + }, + { + "epoch": 1.9116674350083183, + "grad_norm": 1.8114264011383057, + "learning_rate": 1.0164284006782044e-07, + "loss": 3.2535, + "step": 23845 + }, + { + "epoch": 1.9117476097893409, + "grad_norm": 1.9515188932418823, + "learning_rate": 1.0145826418084259e-07, + "loss": 3.1901, + "step": 23846 + }, + { + "epoch": 1.9118277845703635, + "grad_norm": 1.7880390882492065, + "learning_rate": 1.0127385517978982e-07, + "loss": 3.1906, + "step": 23847 + }, + { + "epoch": 1.9119079593513861, + "grad_norm": 1.849129557609558, + "learning_rate": 1.0108961306777076e-07, + "loss": 3.1519, + "step": 23848 + }, + { + "epoch": 1.9119881341324088, + "grad_norm": 1.7621455192565918, + "learning_rate": 1.0090553784789292e-07, + "loss": 3.136, + "step": 23849 + }, + { + "epoch": 1.9120683089134314, + "grad_norm": 1.8311551809310913, + "learning_rate": 1.0072162952325825e-07, + "loss": 3.2209, + "step": 23850 + }, + { + "epoch": 1.912148483694454, + "grad_norm": 1.7371752262115479, + "learning_rate": 1.0053788809696763e-07, + "loss": 3.1604, + "step": 23851 + }, + { + "epoch": 1.9122286584754766, + "grad_norm": 1.7673510313034058, + "learning_rate": 1.0035431357211967e-07, + "loss": 3.1525, + "step": 23852 + }, + { + "epoch": 1.9123088332564993, + "grad_norm": 1.6771334409713745, + "learning_rate": 1.0017090595180967e-07, + "loss": 3.1172, + "step": 23853 + }, + { + "epoch": 1.9123890080375219, + "grad_norm": 1.7163971662521362, + "learning_rate": 9.998766523912961e-08, + "loss": 3.158, + "step": 23854 + }, + { + "epoch": 1.9124691828185445, + "grad_norm": 1.8062394857406616, + "learning_rate": 9.980459143716703e-08, + "loss": 3.1907, + "step": 23855 + }, + { + "epoch": 1.9125493575995671, + "grad_norm": 1.733439564704895, + "learning_rate": 9.962168454901055e-08, + "loss": 3.1931, + "step": 23856 + }, + { + "epoch": 1.9126295323805897, + "grad_norm": 1.7530338764190674, + "learning_rate": 9.943894457774328e-08, + "loss": 3.0967, + "step": 23857 + }, + { + "epoch": 1.9127097071616124, + "grad_norm": 1.6968392133712769, + "learning_rate": 9.925637152644496e-08, + "loss": 3.1535, + "step": 23858 + }, + { + "epoch": 1.912789881942635, + "grad_norm": 1.7286698818206787, + "learning_rate": 9.907396539819647e-08, + "loss": 3.2058, + "step": 23859 + }, + { + "epoch": 1.9128700567236576, + "grad_norm": 0.6436238288879395, + "learning_rate": 9.88917261960709e-08, + "loss": 2.4424, + "step": 23860 + }, + { + "epoch": 1.9129502315046802, + "grad_norm": 1.7985870838165283, + "learning_rate": 9.870965392314136e-08, + "loss": 3.0828, + "step": 23861 + }, + { + "epoch": 1.9130304062857029, + "grad_norm": 1.5997591018676758, + "learning_rate": 9.852774858247649e-08, + "loss": 3.0799, + "step": 23862 + }, + { + "epoch": 1.9131105810667255, + "grad_norm": 1.7723716497421265, + "learning_rate": 9.834601017714496e-08, + "loss": 3.2003, + "step": 23863 + }, + { + "epoch": 1.9131907558477481, + "grad_norm": 1.7347722053527832, + "learning_rate": 9.816443871021098e-08, + "loss": 3.1619, + "step": 23864 + }, + { + "epoch": 1.9132709306287707, + "grad_norm": 1.8415223360061646, + "learning_rate": 9.798303418473431e-08, + "loss": 3.1983, + "step": 23865 + }, + { + "epoch": 1.9133511054097934, + "grad_norm": 1.7968308925628662, + "learning_rate": 9.780179660377364e-08, + "loss": 3.1998, + "step": 23866 + }, + { + "epoch": 1.913431280190816, + "grad_norm": 1.8223298788070679, + "learning_rate": 9.76207259703843e-08, + "loss": 3.1096, + "step": 23867 + }, + { + "epoch": 1.9135114549718386, + "grad_norm": 1.8424912691116333, + "learning_rate": 9.743982228761939e-08, + "loss": 3.1914, + "step": 23868 + }, + { + "epoch": 1.9135916297528612, + "grad_norm": 1.7963327169418335, + "learning_rate": 9.725908555853092e-08, + "loss": 3.1837, + "step": 23869 + }, + { + "epoch": 1.9136718045338839, + "grad_norm": 1.8223652839660645, + "learning_rate": 9.7078515786162e-08, + "loss": 3.0907, + "step": 23870 + }, + { + "epoch": 1.9137519793149065, + "grad_norm": 1.9409123659133911, + "learning_rate": 9.689811297356022e-08, + "loss": 3.26, + "step": 23871 + }, + { + "epoch": 1.913832154095929, + "grad_norm": 1.708152174949646, + "learning_rate": 9.671787712376645e-08, + "loss": 3.3387, + "step": 23872 + }, + { + "epoch": 1.9139123288769517, + "grad_norm": 1.748950481414795, + "learning_rate": 9.653780823981829e-08, + "loss": 3.1169, + "step": 23873 + }, + { + "epoch": 1.9139925036579744, + "grad_norm": 1.923262596130371, + "learning_rate": 9.635790632475328e-08, + "loss": 3.3049, + "step": 23874 + }, + { + "epoch": 1.914072678438997, + "grad_norm": 1.7334790229797363, + "learning_rate": 9.617817138160235e-08, + "loss": 3.1593, + "step": 23875 + }, + { + "epoch": 1.9141528532200196, + "grad_norm": 1.7791292667388916, + "learning_rate": 9.59986034133975e-08, + "loss": 3.2153, + "step": 23876 + }, + { + "epoch": 1.9142330280010422, + "grad_norm": 1.709155559539795, + "learning_rate": 9.581920242316523e-08, + "loss": 3.0717, + "step": 23877 + }, + { + "epoch": 1.9143132027820648, + "grad_norm": 1.7035735845565796, + "learning_rate": 9.563996841393197e-08, + "loss": 3.1345, + "step": 23878 + }, + { + "epoch": 1.9143933775630875, + "grad_norm": 1.732901930809021, + "learning_rate": 9.546090138871755e-08, + "loss": 3.1721, + "step": 23879 + }, + { + "epoch": 1.91447355234411, + "grad_norm": 1.8309690952301025, + "learning_rate": 9.528200135054178e-08, + "loss": 3.1442, + "step": 23880 + }, + { + "epoch": 1.9145537271251327, + "grad_norm": 1.829184889793396, + "learning_rate": 9.510326830242e-08, + "loss": 3.2643, + "step": 23881 + }, + { + "epoch": 1.9146339019061553, + "grad_norm": 1.7553845643997192, + "learning_rate": 9.49247022473665e-08, + "loss": 3.0986, + "step": 23882 + }, + { + "epoch": 1.914714076687178, + "grad_norm": 1.7456302642822266, + "learning_rate": 9.474630318839107e-08, + "loss": 3.0376, + "step": 23883 + }, + { + "epoch": 1.9147942514682006, + "grad_norm": 1.7600560188293457, + "learning_rate": 9.456807112850352e-08, + "loss": 3.1833, + "step": 23884 + }, + { + "epoch": 1.9148744262492232, + "grad_norm": 1.7041748762130737, + "learning_rate": 9.439000607070702e-08, + "loss": 3.1552, + "step": 23885 + }, + { + "epoch": 1.9149546010302458, + "grad_norm": 1.81456458568573, + "learning_rate": 9.421210801800251e-08, + "loss": 3.2367, + "step": 23886 + }, + { + "epoch": 1.9150347758112685, + "grad_norm": 0.6621845364570618, + "learning_rate": 9.403437697339202e-08, + "loss": 2.4812, + "step": 23887 + }, + { + "epoch": 1.915114950592291, + "grad_norm": 1.691307783126831, + "learning_rate": 9.385681293987092e-08, + "loss": 3.0749, + "step": 23888 + }, + { + "epoch": 1.9151951253733137, + "grad_norm": 1.7328466176986694, + "learning_rate": 9.367941592043239e-08, + "loss": 3.202, + "step": 23889 + }, + { + "epoch": 1.9152753001543363, + "grad_norm": 1.7531355619430542, + "learning_rate": 9.350218591806736e-08, + "loss": 3.0963, + "step": 23890 + }, + { + "epoch": 1.915355474935359, + "grad_norm": 1.7015193700790405, + "learning_rate": 9.332512293576346e-08, + "loss": 3.0308, + "step": 23891 + }, + { + "epoch": 1.9154356497163816, + "grad_norm": 1.777153491973877, + "learning_rate": 9.314822697650716e-08, + "loss": 3.1509, + "step": 23892 + }, + { + "epoch": 1.9155158244974042, + "grad_norm": 0.6355154514312744, + "learning_rate": 9.297149804328054e-08, + "loss": 2.4637, + "step": 23893 + }, + { + "epoch": 1.9155959992784268, + "grad_norm": 1.776761531829834, + "learning_rate": 9.279493613906232e-08, + "loss": 3.2496, + "step": 23894 + }, + { + "epoch": 1.9156761740594495, + "grad_norm": 1.7560895681381226, + "learning_rate": 9.261854126683012e-08, + "loss": 3.0784, + "step": 23895 + }, + { + "epoch": 1.915756348840472, + "grad_norm": 1.7655285596847534, + "learning_rate": 9.244231342955823e-08, + "loss": 3.246, + "step": 23896 + }, + { + "epoch": 1.9158365236214947, + "grad_norm": 1.688904881477356, + "learning_rate": 9.226625263021538e-08, + "loss": 3.1993, + "step": 23897 + }, + { + "epoch": 1.9159166984025173, + "grad_norm": 1.739711880683899, + "learning_rate": 9.209035887177364e-08, + "loss": 3.2543, + "step": 23898 + }, + { + "epoch": 1.9159968731835402, + "grad_norm": 1.757696509361267, + "learning_rate": 9.19146321571962e-08, + "loss": 3.0287, + "step": 23899 + }, + { + "epoch": 1.9160770479645628, + "grad_norm": 2.002326250076294, + "learning_rate": 9.173907248944513e-08, + "loss": 3.118, + "step": 23900 + }, + { + "epoch": 1.9161572227455854, + "grad_norm": 1.663255214691162, + "learning_rate": 9.15636798714814e-08, + "loss": 3.1548, + "step": 23901 + }, + { + "epoch": 1.916237397526608, + "grad_norm": 1.8827096223831177, + "learning_rate": 9.138845430626265e-08, + "loss": 3.248, + "step": 23902 + }, + { + "epoch": 1.9163175723076307, + "grad_norm": 1.7247527837753296, + "learning_rate": 9.121339579674205e-08, + "loss": 3.1145, + "step": 23903 + }, + { + "epoch": 1.9163977470886533, + "grad_norm": 1.7574412822723389, + "learning_rate": 9.103850434587169e-08, + "loss": 3.2365, + "step": 23904 + }, + { + "epoch": 1.916477921869676, + "grad_norm": 1.822121262550354, + "learning_rate": 9.086377995660034e-08, + "loss": 3.2292, + "step": 23905 + }, + { + "epoch": 1.9165580966506985, + "grad_norm": 0.6646823287010193, + "learning_rate": 9.06892226318723e-08, + "loss": 2.4636, + "step": 23906 + }, + { + "epoch": 1.9166382714317212, + "grad_norm": 1.8183512687683105, + "learning_rate": 9.051483237463187e-08, + "loss": 3.2073, + "step": 23907 + }, + { + "epoch": 1.9167184462127438, + "grad_norm": 0.6406741738319397, + "learning_rate": 9.034060918781895e-08, + "loss": 2.4719, + "step": 23908 + }, + { + "epoch": 1.9167986209937664, + "grad_norm": 1.828202486038208, + "learning_rate": 9.016655307437117e-08, + "loss": 3.1838, + "step": 23909 + }, + { + "epoch": 1.916878795774789, + "grad_norm": 1.700113296508789, + "learning_rate": 8.999266403722284e-08, + "loss": 3.0968, + "step": 23910 + }, + { + "epoch": 1.9169589705558117, + "grad_norm": 1.8980257511138916, + "learning_rate": 8.981894207930497e-08, + "loss": 3.1602, + "step": 23911 + }, + { + "epoch": 1.9170391453368343, + "grad_norm": 1.8216761350631714, + "learning_rate": 8.964538720354742e-08, + "loss": 3.225, + "step": 23912 + }, + { + "epoch": 1.917119320117857, + "grad_norm": 1.7845886945724487, + "learning_rate": 8.947199941287565e-08, + "loss": 3.3209, + "step": 23913 + }, + { + "epoch": 1.9171994948988795, + "grad_norm": 1.8822433948516846, + "learning_rate": 8.929877871021398e-08, + "loss": 3.1401, + "step": 23914 + }, + { + "epoch": 1.9172796696799022, + "grad_norm": 1.7595607042312622, + "learning_rate": 8.912572509848117e-08, + "loss": 3.1614, + "step": 23915 + }, + { + "epoch": 1.9173598444609248, + "grad_norm": 1.7815083265304565, + "learning_rate": 8.8952838580596e-08, + "loss": 3.1654, + "step": 23916 + }, + { + "epoch": 1.9174400192419474, + "grad_norm": 1.7326900959014893, + "learning_rate": 8.878011915947283e-08, + "loss": 3.1504, + "step": 23917 + }, + { + "epoch": 1.91752019402297, + "grad_norm": 1.7140944004058838, + "learning_rate": 8.860756683802374e-08, + "loss": 3.1202, + "step": 23918 + }, + { + "epoch": 1.9176003688039926, + "grad_norm": 1.828428030014038, + "learning_rate": 8.843518161915864e-08, + "loss": 3.1759, + "step": 23919 + }, + { + "epoch": 1.9176805435850155, + "grad_norm": 1.8266663551330566, + "learning_rate": 8.82629635057819e-08, + "loss": 3.2459, + "step": 23920 + }, + { + "epoch": 1.9177607183660381, + "grad_norm": 1.7484112977981567, + "learning_rate": 8.809091250080004e-08, + "loss": 3.196, + "step": 23921 + }, + { + "epoch": 1.9178408931470607, + "grad_norm": 1.7721211910247803, + "learning_rate": 8.791902860711076e-08, + "loss": 3.2044, + "step": 23922 + }, + { + "epoch": 1.9179210679280834, + "grad_norm": 1.6888267993927002, + "learning_rate": 8.774731182761287e-08, + "loss": 3.1885, + "step": 23923 + }, + { + "epoch": 1.918001242709106, + "grad_norm": 1.7772033214569092, + "learning_rate": 8.757576216520402e-08, + "loss": 3.2151, + "step": 23924 + }, + { + "epoch": 1.9180814174901286, + "grad_norm": 1.8081368207931519, + "learning_rate": 8.740437962277304e-08, + "loss": 3.2111, + "step": 23925 + }, + { + "epoch": 1.9181615922711512, + "grad_norm": 1.7752505540847778, + "learning_rate": 8.723316420320982e-08, + "loss": 3.1355, + "step": 23926 + }, + { + "epoch": 1.9182417670521739, + "grad_norm": 1.7256584167480469, + "learning_rate": 8.706211590940317e-08, + "loss": 3.2321, + "step": 23927 + }, + { + "epoch": 1.9183219418331965, + "grad_norm": 1.7878552675247192, + "learning_rate": 8.689123474423522e-08, + "loss": 3.2151, + "step": 23928 + }, + { + "epoch": 1.918402116614219, + "grad_norm": 1.7494066953659058, + "learning_rate": 8.672052071058701e-08, + "loss": 3.1732, + "step": 23929 + }, + { + "epoch": 1.9184822913952417, + "grad_norm": 1.6968883275985718, + "learning_rate": 8.654997381133734e-08, + "loss": 3.1738, + "step": 23930 + }, + { + "epoch": 1.9185624661762644, + "grad_norm": 1.7551292181015015, + "learning_rate": 8.63795940493617e-08, + "loss": 3.0815, + "step": 23931 + }, + { + "epoch": 1.918642640957287, + "grad_norm": 1.6736130714416504, + "learning_rate": 8.62093814275311e-08, + "loss": 3.194, + "step": 23932 + }, + { + "epoch": 1.9187228157383096, + "grad_norm": 1.7915358543395996, + "learning_rate": 8.603933594871661e-08, + "loss": 3.087, + "step": 23933 + }, + { + "epoch": 1.9188029905193322, + "grad_norm": 1.7750550508499146, + "learning_rate": 8.586945761578591e-08, + "loss": 3.176, + "step": 23934 + }, + { + "epoch": 1.9188831653003549, + "grad_norm": 1.8155454397201538, + "learning_rate": 8.569974643160229e-08, + "loss": 3.1481, + "step": 23935 + }, + { + "epoch": 1.9189633400813775, + "grad_norm": 1.72445547580719, + "learning_rate": 8.553020239902566e-08, + "loss": 3.2122, + "step": 23936 + }, + { + "epoch": 1.9190435148624, + "grad_norm": 1.8409361839294434, + "learning_rate": 8.536082552091595e-08, + "loss": 3.135, + "step": 23937 + }, + { + "epoch": 1.9191236896434227, + "grad_norm": 1.9472194910049438, + "learning_rate": 8.519161580012758e-08, + "loss": 3.1661, + "step": 23938 + }, + { + "epoch": 1.9192038644244453, + "grad_norm": 1.7709287405014038, + "learning_rate": 8.502257323951602e-08, + "loss": 3.1138, + "step": 23939 + }, + { + "epoch": 1.919284039205468, + "grad_norm": 1.8376137018203735, + "learning_rate": 8.485369784192788e-08, + "loss": 3.0674, + "step": 23940 + }, + { + "epoch": 1.9193642139864906, + "grad_norm": 1.7669700384140015, + "learning_rate": 8.46849896102131e-08, + "loss": 3.22, + "step": 23941 + }, + { + "epoch": 1.9194443887675132, + "grad_norm": 1.718344807624817, + "learning_rate": 8.451644854721385e-08, + "loss": 3.1967, + "step": 23942 + }, + { + "epoch": 1.9195245635485358, + "grad_norm": 1.868970274925232, + "learning_rate": 8.434807465577344e-08, + "loss": 3.2357, + "step": 23943 + }, + { + "epoch": 1.9196047383295585, + "grad_norm": 0.665860116481781, + "learning_rate": 8.417986793873068e-08, + "loss": 2.4727, + "step": 23944 + }, + { + "epoch": 1.919684913110581, + "grad_norm": 1.8233708143234253, + "learning_rate": 8.401182839891997e-08, + "loss": 3.2844, + "step": 23945 + }, + { + "epoch": 1.9197650878916037, + "grad_norm": 1.6775932312011719, + "learning_rate": 8.384395603917462e-08, + "loss": 3.1827, + "step": 23946 + }, + { + "epoch": 1.9198452626726263, + "grad_norm": 0.6379995346069336, + "learning_rate": 8.367625086232567e-08, + "loss": 2.4611, + "step": 23947 + }, + { + "epoch": 1.919925437453649, + "grad_norm": 1.832881212234497, + "learning_rate": 8.350871287119977e-08, + "loss": 3.2222, + "step": 23948 + }, + { + "epoch": 1.9200056122346716, + "grad_norm": 1.8816858530044556, + "learning_rate": 8.334134206862354e-08, + "loss": 3.2281, + "step": 23949 + }, + { + "epoch": 1.9200857870156942, + "grad_norm": 1.8758476972579956, + "learning_rate": 8.317413845741584e-08, + "loss": 3.1873, + "step": 23950 + }, + { + "epoch": 1.9201659617967168, + "grad_norm": 1.7559754848480225, + "learning_rate": 8.300710204039775e-08, + "loss": 3.1342, + "step": 23951 + }, + { + "epoch": 1.9202461365777395, + "grad_norm": 1.7653549909591675, + "learning_rate": 8.284023282038478e-08, + "loss": 3.1752, + "step": 23952 + }, + { + "epoch": 1.920326311358762, + "grad_norm": 1.7684299945831299, + "learning_rate": 8.267353080019024e-08, + "loss": 3.0922, + "step": 23953 + }, + { + "epoch": 1.9204064861397847, + "grad_norm": 1.75216543674469, + "learning_rate": 8.250699598262523e-08, + "loss": 3.2089, + "step": 23954 + }, + { + "epoch": 1.9204866609208073, + "grad_norm": 1.8001128435134888, + "learning_rate": 8.234062837049639e-08, + "loss": 3.2184, + "step": 23955 + }, + { + "epoch": 1.92056683570183, + "grad_norm": 1.7509024143218994, + "learning_rate": 8.217442796660924e-08, + "loss": 3.2123, + "step": 23956 + }, + { + "epoch": 1.9206470104828526, + "grad_norm": 1.8267321586608887, + "learning_rate": 8.200839477376599e-08, + "loss": 3.2152, + "step": 23957 + }, + { + "epoch": 1.9207271852638752, + "grad_norm": 1.6928125619888306, + "learning_rate": 8.18425287947655e-08, + "loss": 3.0925, + "step": 23958 + }, + { + "epoch": 1.9208073600448978, + "grad_norm": 1.7381802797317505, + "learning_rate": 8.167683003240556e-08, + "loss": 3.1544, + "step": 23959 + }, + { + "epoch": 1.9208875348259205, + "grad_norm": 1.7556304931640625, + "learning_rate": 8.151129848947725e-08, + "loss": 3.1682, + "step": 23960 + }, + { + "epoch": 1.920967709606943, + "grad_norm": 1.7842049598693848, + "learning_rate": 8.13459341687739e-08, + "loss": 3.1381, + "step": 23961 + }, + { + "epoch": 1.9210478843879657, + "grad_norm": 1.7675548791885376, + "learning_rate": 8.118073707308105e-08, + "loss": 3.1843, + "step": 23962 + }, + { + "epoch": 1.9211280591689883, + "grad_norm": 1.723912239074707, + "learning_rate": 8.101570720518648e-08, + "loss": 3.1465, + "step": 23963 + }, + { + "epoch": 1.921208233950011, + "grad_norm": 1.7678402662277222, + "learning_rate": 8.085084456787129e-08, + "loss": 3.2, + "step": 23964 + }, + { + "epoch": 1.9212884087310336, + "grad_norm": 1.7085232734680176, + "learning_rate": 8.068614916391548e-08, + "loss": 3.2187, + "step": 23965 + }, + { + "epoch": 1.9213685835120562, + "grad_norm": 1.7972743511199951, + "learning_rate": 8.052162099609351e-08, + "loss": 3.1846, + "step": 23966 + }, + { + "epoch": 1.9214487582930788, + "grad_norm": 1.7784019708633423, + "learning_rate": 8.035726006718203e-08, + "loss": 3.1274, + "step": 23967 + }, + { + "epoch": 1.9215289330741014, + "grad_norm": 1.808656930923462, + "learning_rate": 8.019306637995216e-08, + "loss": 3.1567, + "step": 23968 + }, + { + "epoch": 1.921609107855124, + "grad_norm": 1.8143346309661865, + "learning_rate": 8.00290399371706e-08, + "loss": 3.1188, + "step": 23969 + }, + { + "epoch": 1.9216892826361467, + "grad_norm": 2.210911273956299, + "learning_rate": 7.986518074160176e-08, + "loss": 3.1594, + "step": 23970 + }, + { + "epoch": 1.9217694574171693, + "grad_norm": 1.7618567943572998, + "learning_rate": 7.970148879601125e-08, + "loss": 3.1909, + "step": 23971 + }, + { + "epoch": 1.921849632198192, + "grad_norm": 1.778444528579712, + "learning_rate": 7.953796410315684e-08, + "loss": 3.2366, + "step": 23972 + }, + { + "epoch": 1.9219298069792146, + "grad_norm": 1.8087238073349, + "learning_rate": 7.937460666579633e-08, + "loss": 3.1555, + "step": 23973 + }, + { + "epoch": 1.9220099817602372, + "grad_norm": 1.8211404085159302, + "learning_rate": 7.921141648668306e-08, + "loss": 3.2448, + "step": 23974 + }, + { + "epoch": 1.9220901565412598, + "grad_norm": 1.767198085784912, + "learning_rate": 7.904839356857042e-08, + "loss": 3.189, + "step": 23975 + }, + { + "epoch": 1.9221703313222824, + "grad_norm": 1.7508127689361572, + "learning_rate": 7.888553791420394e-08, + "loss": 3.1817, + "step": 23976 + }, + { + "epoch": 1.922250506103305, + "grad_norm": 1.7754851579666138, + "learning_rate": 7.872284952633036e-08, + "loss": 3.1978, + "step": 23977 + }, + { + "epoch": 1.9223306808843277, + "grad_norm": 1.8026081323623657, + "learning_rate": 7.856032840769412e-08, + "loss": 3.1963, + "step": 23978 + }, + { + "epoch": 1.9224108556653503, + "grad_norm": 1.8668125867843628, + "learning_rate": 7.839797456103304e-08, + "loss": 3.1379, + "step": 23979 + }, + { + "epoch": 1.922491030446373, + "grad_norm": 1.9210472106933594, + "learning_rate": 7.823578798908716e-08, + "loss": 3.1923, + "step": 23980 + }, + { + "epoch": 1.9225712052273958, + "grad_norm": 1.7136569023132324, + "learning_rate": 7.80737686945876e-08, + "loss": 3.1805, + "step": 23981 + }, + { + "epoch": 1.9226513800084184, + "grad_norm": 1.8059823513031006, + "learning_rate": 7.791191668026777e-08, + "loss": 3.2944, + "step": 23982 + }, + { + "epoch": 1.922731554789441, + "grad_norm": 1.7240443229675293, + "learning_rate": 7.775023194885545e-08, + "loss": 3.0602, + "step": 23983 + }, + { + "epoch": 1.9228117295704636, + "grad_norm": 1.897146463394165, + "learning_rate": 7.758871450307848e-08, + "loss": 3.2843, + "step": 23984 + }, + { + "epoch": 1.9228919043514863, + "grad_norm": 1.7255940437316895, + "learning_rate": 7.742736434565912e-08, + "loss": 3.1332, + "step": 23985 + }, + { + "epoch": 1.922972079132509, + "grad_norm": 1.8350024223327637, + "learning_rate": 7.726618147931631e-08, + "loss": 3.0549, + "step": 23986 + }, + { + "epoch": 1.9230522539135315, + "grad_norm": 1.9075796604156494, + "learning_rate": 7.710516590676786e-08, + "loss": 3.2446, + "step": 23987 + }, + { + "epoch": 1.9231324286945541, + "grad_norm": 1.7040454149246216, + "learning_rate": 7.694431763073051e-08, + "loss": 3.1476, + "step": 23988 + }, + { + "epoch": 1.9232126034755768, + "grad_norm": 1.7427856922149658, + "learning_rate": 7.67836366539143e-08, + "loss": 3.0877, + "step": 23989 + }, + { + "epoch": 1.9232927782565994, + "grad_norm": 1.8503313064575195, + "learning_rate": 7.66231229790293e-08, + "loss": 3.2126, + "step": 23990 + }, + { + "epoch": 1.923372953037622, + "grad_norm": 1.7752361297607422, + "learning_rate": 7.646277660877999e-08, + "loss": 3.1194, + "step": 23991 + }, + { + "epoch": 1.9234531278186446, + "grad_norm": 1.7343882322311401, + "learning_rate": 7.6302597545872e-08, + "loss": 3.1788, + "step": 23992 + }, + { + "epoch": 1.9235333025996673, + "grad_norm": 1.789312481880188, + "learning_rate": 7.614258579300315e-08, + "loss": 3.1853, + "step": 23993 + }, + { + "epoch": 1.9236134773806899, + "grad_norm": 1.894024133682251, + "learning_rate": 7.598274135287464e-08, + "loss": 3.2389, + "step": 23994 + }, + { + "epoch": 1.9236936521617125, + "grad_norm": 1.7319209575653076, + "learning_rate": 7.582306422817876e-08, + "loss": 3.1179, + "step": 23995 + }, + { + "epoch": 1.9237738269427351, + "grad_norm": 1.7627403736114502, + "learning_rate": 7.566355442160778e-08, + "loss": 3.1553, + "step": 23996 + }, + { + "epoch": 1.9238540017237578, + "grad_norm": 1.723097801208496, + "learning_rate": 7.550421193585289e-08, + "loss": 3.0603, + "step": 23997 + }, + { + "epoch": 1.9239341765047804, + "grad_norm": 1.9212819337844849, + "learning_rate": 7.53450367735975e-08, + "loss": 3.2813, + "step": 23998 + }, + { + "epoch": 1.924014351285803, + "grad_norm": 1.783027172088623, + "learning_rate": 7.518602893752835e-08, + "loss": 3.1383, + "step": 23999 + }, + { + "epoch": 1.9240945260668256, + "grad_norm": 0.6396978497505188, + "learning_rate": 7.502718843032553e-08, + "loss": 2.4606, + "step": 24000 + }, + { + "epoch": 1.9241747008478485, + "grad_norm": 1.7496923208236694, + "learning_rate": 7.486851525466466e-08, + "loss": 3.1147, + "step": 24001 + }, + { + "epoch": 1.924254875628871, + "grad_norm": 1.7626439332962036, + "learning_rate": 7.47100094132236e-08, + "loss": 3.3353, + "step": 24002 + }, + { + "epoch": 1.9243350504098937, + "grad_norm": 1.7561326026916504, + "learning_rate": 7.455167090867355e-08, + "loss": 3.1221, + "step": 24003 + }, + { + "epoch": 1.9244152251909163, + "grad_norm": 1.7213823795318604, + "learning_rate": 7.43934997436846e-08, + "loss": 3.1542, + "step": 24004 + }, + { + "epoch": 1.924495399971939, + "grad_norm": 1.83524751663208, + "learning_rate": 7.42354959209235e-08, + "loss": 3.0608, + "step": 24005 + }, + { + "epoch": 1.9245755747529616, + "grad_norm": 1.970983862876892, + "learning_rate": 7.407765944305256e-08, + "loss": 3.1786, + "step": 24006 + }, + { + "epoch": 1.9246557495339842, + "grad_norm": 0.6572141051292419, + "learning_rate": 7.391999031273522e-08, + "loss": 2.4642, + "step": 24007 + }, + { + "epoch": 1.9247359243150068, + "grad_norm": 1.7210744619369507, + "learning_rate": 7.376248853262713e-08, + "loss": 3.1864, + "step": 24008 + }, + { + "epoch": 1.9248160990960295, + "grad_norm": 1.7706623077392578, + "learning_rate": 7.360515410538615e-08, + "loss": 3.2539, + "step": 24009 + }, + { + "epoch": 1.924896273877052, + "grad_norm": 1.8286470174789429, + "learning_rate": 7.344798703366463e-08, + "loss": 3.2352, + "step": 24010 + }, + { + "epoch": 1.9249764486580747, + "grad_norm": 1.816817283630371, + "learning_rate": 7.329098732011042e-08, + "loss": 3.283, + "step": 24011 + }, + { + "epoch": 1.9250566234390973, + "grad_norm": 1.733184576034546, + "learning_rate": 7.313415496737142e-08, + "loss": 3.1376, + "step": 24012 + }, + { + "epoch": 1.92513679822012, + "grad_norm": 1.6808539628982544, + "learning_rate": 7.297748997809218e-08, + "loss": 3.0807, + "step": 24013 + }, + { + "epoch": 1.9252169730011426, + "grad_norm": 1.7163830995559692, + "learning_rate": 7.282099235491391e-08, + "loss": 3.2196, + "step": 24014 + }, + { + "epoch": 1.9252971477821652, + "grad_norm": 1.688106656074524, + "learning_rate": 7.266466210047563e-08, + "loss": 3.2031, + "step": 24015 + }, + { + "epoch": 1.9253773225631878, + "grad_norm": 1.7355098724365234, + "learning_rate": 7.250849921741299e-08, + "loss": 3.2083, + "step": 24016 + }, + { + "epoch": 1.9254574973442105, + "grad_norm": 1.800833821296692, + "learning_rate": 7.235250370835723e-08, + "loss": 3.1604, + "step": 24017 + }, + { + "epoch": 1.925537672125233, + "grad_norm": 1.7716622352600098, + "learning_rate": 7.219667557593846e-08, + "loss": 3.0396, + "step": 24018 + }, + { + "epoch": 1.9256178469062557, + "grad_norm": 0.678102433681488, + "learning_rate": 7.204101482278569e-08, + "loss": 2.4547, + "step": 24019 + }, + { + "epoch": 1.9256980216872783, + "grad_norm": 1.7712452411651611, + "learning_rate": 7.18855214515235e-08, + "loss": 3.1625, + "step": 24020 + }, + { + "epoch": 1.925778196468301, + "grad_norm": 1.744850754737854, + "learning_rate": 7.1730195464772e-08, + "loss": 3.1035, + "step": 24021 + }, + { + "epoch": 1.9258583712493236, + "grad_norm": 1.8606289625167847, + "learning_rate": 7.15750368651491e-08, + "loss": 3.2214, + "step": 24022 + }, + { + "epoch": 1.9259385460303462, + "grad_norm": 1.7178707122802734, + "learning_rate": 7.142004565527271e-08, + "loss": 3.1881, + "step": 24023 + }, + { + "epoch": 1.9260187208113688, + "grad_norm": 1.7797625064849854, + "learning_rate": 7.126522183775519e-08, + "loss": 3.1971, + "step": 24024 + }, + { + "epoch": 1.9260988955923914, + "grad_norm": 1.7584054470062256, + "learning_rate": 7.111056541520777e-08, + "loss": 3.2252, + "step": 24025 + }, + { + "epoch": 1.926179070373414, + "grad_norm": 1.8581756353378296, + "learning_rate": 7.095607639023617e-08, + "loss": 3.1728, + "step": 24026 + }, + { + "epoch": 1.9262592451544367, + "grad_norm": 0.6623292565345764, + "learning_rate": 7.080175476544493e-08, + "loss": 2.4614, + "step": 24027 + }, + { + "epoch": 1.9263394199354593, + "grad_norm": 1.7732712030410767, + "learning_rate": 7.064760054343756e-08, + "loss": 3.1975, + "step": 24028 + }, + { + "epoch": 1.926419594716482, + "grad_norm": 1.6827709674835205, + "learning_rate": 7.049361372681195e-08, + "loss": 3.1126, + "step": 24029 + }, + { + "epoch": 1.9264997694975046, + "grad_norm": 1.750754952430725, + "learning_rate": 7.033979431816495e-08, + "loss": 3.1084, + "step": 24030 + }, + { + "epoch": 1.9265799442785272, + "grad_norm": 1.8042194843292236, + "learning_rate": 7.018614232008891e-08, + "loss": 3.2308, + "step": 24031 + }, + { + "epoch": 1.9266601190595498, + "grad_norm": 1.7401394844055176, + "learning_rate": 7.003265773517509e-08, + "loss": 3.1073, + "step": 24032 + }, + { + "epoch": 1.9267402938405724, + "grad_norm": 1.6926525831222534, + "learning_rate": 6.987934056601143e-08, + "loss": 3.1141, + "step": 24033 + }, + { + "epoch": 1.926820468621595, + "grad_norm": 1.873870849609375, + "learning_rate": 6.97261908151814e-08, + "loss": 3.1526, + "step": 24034 + }, + { + "epoch": 1.9269006434026177, + "grad_norm": 1.7494750022888184, + "learning_rate": 6.95732084852696e-08, + "loss": 3.3274, + "step": 24035 + }, + { + "epoch": 1.9269808181836403, + "grad_norm": 1.7498530149459839, + "learning_rate": 6.942039357885177e-08, + "loss": 3.1837, + "step": 24036 + }, + { + "epoch": 1.927060992964663, + "grad_norm": 1.8614927530288696, + "learning_rate": 6.926774609850807e-08, + "loss": 3.177, + "step": 24037 + }, + { + "epoch": 1.9271411677456856, + "grad_norm": 1.788759708404541, + "learning_rate": 6.911526604680862e-08, + "loss": 3.1599, + "step": 24038 + }, + { + "epoch": 1.9272213425267082, + "grad_norm": 1.7378246784210205, + "learning_rate": 6.896295342632698e-08, + "loss": 3.1602, + "step": 24039 + }, + { + "epoch": 1.9273015173077308, + "grad_norm": 1.7276595830917358, + "learning_rate": 6.881080823962993e-08, + "loss": 3.1687, + "step": 24040 + }, + { + "epoch": 1.9273816920887534, + "grad_norm": 1.7151882648468018, + "learning_rate": 6.865883048928101e-08, + "loss": 3.3496, + "step": 24041 + }, + { + "epoch": 1.927461866869776, + "grad_norm": 1.7400147914886475, + "learning_rate": 6.850702017784594e-08, + "loss": 3.2018, + "step": 24042 + }, + { + "epoch": 1.9275420416507987, + "grad_norm": 1.9224029779434204, + "learning_rate": 6.835537730788156e-08, + "loss": 3.2543, + "step": 24043 + }, + { + "epoch": 1.9276222164318213, + "grad_norm": 1.6883350610733032, + "learning_rate": 6.82039018819447e-08, + "loss": 3.2042, + "step": 24044 + }, + { + "epoch": 1.927702391212844, + "grad_norm": 1.6673470735549927, + "learning_rate": 6.805259390259111e-08, + "loss": 3.1659, + "step": 24045 + }, + { + "epoch": 1.9277825659938665, + "grad_norm": 1.7632375955581665, + "learning_rate": 6.790145337236876e-08, + "loss": 3.1683, + "step": 24046 + }, + { + "epoch": 1.9278627407748892, + "grad_norm": 1.86172616481781, + "learning_rate": 6.775048029382892e-08, + "loss": 3.1787, + "step": 24047 + }, + { + "epoch": 1.9279429155559118, + "grad_norm": 1.7138370275497437, + "learning_rate": 6.7599674669514e-08, + "loss": 3.1198, + "step": 24048 + }, + { + "epoch": 1.9280230903369344, + "grad_norm": 1.8481464385986328, + "learning_rate": 6.744903650196977e-08, + "loss": 3.2433, + "step": 24049 + }, + { + "epoch": 1.928103265117957, + "grad_norm": 1.8192341327667236, + "learning_rate": 6.729856579373306e-08, + "loss": 3.2111, + "step": 24050 + }, + { + "epoch": 1.9281834398989797, + "grad_norm": 1.7969337701797485, + "learning_rate": 6.714826254734186e-08, + "loss": 3.3229, + "step": 24051 + }, + { + "epoch": 1.9282636146800023, + "grad_norm": 1.7218295335769653, + "learning_rate": 6.699812676533079e-08, + "loss": 3.1812, + "step": 24052 + }, + { + "epoch": 1.928343789461025, + "grad_norm": 1.7329137325286865, + "learning_rate": 6.684815845023119e-08, + "loss": 3.1154, + "step": 24053 + }, + { + "epoch": 1.9284239642420475, + "grad_norm": 1.7265658378601074, + "learning_rate": 6.669835760456989e-08, + "loss": 3.0869, + "step": 24054 + }, + { + "epoch": 1.9285041390230702, + "grad_norm": 1.848955512046814, + "learning_rate": 6.654872423087377e-08, + "loss": 3.1882, + "step": 24055 + }, + { + "epoch": 1.9285843138040928, + "grad_norm": 1.8321396112442017, + "learning_rate": 6.639925833166527e-08, + "loss": 3.131, + "step": 24056 + }, + { + "epoch": 1.9286644885851154, + "grad_norm": 1.8863948583602905, + "learning_rate": 6.624995990946348e-08, + "loss": 3.2194, + "step": 24057 + }, + { + "epoch": 1.928744663366138, + "grad_norm": 1.8332816362380981, + "learning_rate": 6.610082896678749e-08, + "loss": 3.2051, + "step": 24058 + }, + { + "epoch": 1.9288248381471607, + "grad_norm": 1.723201036453247, + "learning_rate": 6.595186550614973e-08, + "loss": 3.2446, + "step": 24059 + }, + { + "epoch": 1.9289050129281833, + "grad_norm": 1.7174746990203857, + "learning_rate": 6.580306953006376e-08, + "loss": 3.1969, + "step": 24060 + }, + { + "epoch": 1.928985187709206, + "grad_norm": 1.902708649635315, + "learning_rate": 6.565444104103536e-08, + "loss": 3.1489, + "step": 24061 + }, + { + "epoch": 1.9290653624902288, + "grad_norm": 1.8071949481964111, + "learning_rate": 6.55059800415725e-08, + "loss": 3.1351, + "step": 24062 + }, + { + "epoch": 1.9291455372712514, + "grad_norm": 1.7422631978988647, + "learning_rate": 6.535768653417762e-08, + "loss": 3.2717, + "step": 24063 + }, + { + "epoch": 1.929225712052274, + "grad_norm": 1.768392562866211, + "learning_rate": 6.520956052135097e-08, + "loss": 3.1724, + "step": 24064 + }, + { + "epoch": 1.9293058868332966, + "grad_norm": 0.6586591005325317, + "learning_rate": 6.506160200558941e-08, + "loss": 2.4766, + "step": 24065 + }, + { + "epoch": 1.9293860616143192, + "grad_norm": 1.7898812294006348, + "learning_rate": 6.491381098938765e-08, + "loss": 3.1476, + "step": 24066 + }, + { + "epoch": 1.9294662363953419, + "grad_norm": 1.7654893398284912, + "learning_rate": 6.476618747523811e-08, + "loss": 3.1615, + "step": 24067 + }, + { + "epoch": 1.9295464111763645, + "grad_norm": 1.8642691373825073, + "learning_rate": 6.46187314656288e-08, + "loss": 3.1206, + "step": 24068 + }, + { + "epoch": 1.9296265859573871, + "grad_norm": 1.7340518236160278, + "learning_rate": 6.447144296304664e-08, + "loss": 3.1246, + "step": 24069 + }, + { + "epoch": 1.9297067607384097, + "grad_norm": 1.8035036325454712, + "learning_rate": 6.432432196997406e-08, + "loss": 3.1531, + "step": 24070 + }, + { + "epoch": 1.9297869355194324, + "grad_norm": 1.644032597541809, + "learning_rate": 6.417736848889133e-08, + "loss": 3.1401, + "step": 24071 + }, + { + "epoch": 1.929867110300455, + "grad_norm": 1.7464488744735718, + "learning_rate": 6.403058252227646e-08, + "loss": 3.1388, + "step": 24072 + }, + { + "epoch": 1.9299472850814776, + "grad_norm": 1.8205323219299316, + "learning_rate": 6.388396407260411e-08, + "loss": 3.2234, + "step": 24073 + }, + { + "epoch": 1.9300274598625002, + "grad_norm": 1.7717726230621338, + "learning_rate": 6.373751314234678e-08, + "loss": 3.1993, + "step": 24074 + }, + { + "epoch": 1.9301076346435229, + "grad_norm": 1.7294816970825195, + "learning_rate": 6.359122973397358e-08, + "loss": 3.1975, + "step": 24075 + }, + { + "epoch": 1.9301878094245455, + "grad_norm": 1.793537974357605, + "learning_rate": 6.344511384994923e-08, + "loss": 3.1036, + "step": 24076 + }, + { + "epoch": 1.930267984205568, + "grad_norm": 1.7355287075042725, + "learning_rate": 6.329916549273841e-08, + "loss": 3.2014, + "step": 24077 + }, + { + "epoch": 1.9303481589865907, + "grad_norm": 1.7247412204742432, + "learning_rate": 6.315338466480248e-08, + "loss": 3.0277, + "step": 24078 + }, + { + "epoch": 1.9304283337676134, + "grad_norm": 1.8027374744415283, + "learning_rate": 6.300777136859726e-08, + "loss": 3.1294, + "step": 24079 + }, + { + "epoch": 1.930508508548636, + "grad_norm": 1.937781810760498, + "learning_rate": 6.286232560658079e-08, + "loss": 3.1052, + "step": 24080 + }, + { + "epoch": 1.9305886833296586, + "grad_norm": 1.9071670770645142, + "learning_rate": 6.27170473812011e-08, + "loss": 3.1697, + "step": 24081 + }, + { + "epoch": 1.9306688581106815, + "grad_norm": 1.7671217918395996, + "learning_rate": 6.257193669491068e-08, + "loss": 3.2442, + "step": 24082 + }, + { + "epoch": 1.930749032891704, + "grad_norm": 1.796568751335144, + "learning_rate": 6.242699355015536e-08, + "loss": 3.1849, + "step": 24083 + }, + { + "epoch": 1.9308292076727267, + "grad_norm": 1.7611602544784546, + "learning_rate": 6.228221794937762e-08, + "loss": 3.1886, + "step": 24084 + }, + { + "epoch": 1.9309093824537493, + "grad_norm": 1.8441466093063354, + "learning_rate": 6.213760989502105e-08, + "loss": 3.2158, + "step": 24085 + }, + { + "epoch": 1.930989557234772, + "grad_norm": 1.8654863834381104, + "learning_rate": 6.199316938952038e-08, + "loss": 3.2322, + "step": 24086 + }, + { + "epoch": 1.9310697320157946, + "grad_norm": 1.746087670326233, + "learning_rate": 6.184889643531255e-08, + "loss": 3.2312, + "step": 24087 + }, + { + "epoch": 1.9311499067968172, + "grad_norm": 1.721298098564148, + "learning_rate": 6.170479103483007e-08, + "loss": 3.1506, + "step": 24088 + }, + { + "epoch": 1.9312300815778398, + "grad_norm": 1.7883727550506592, + "learning_rate": 6.15608531905032e-08, + "loss": 3.0696, + "step": 24089 + }, + { + "epoch": 1.9313102563588624, + "grad_norm": 1.6785575151443481, + "learning_rate": 6.141708290475667e-08, + "loss": 3.1283, + "step": 24090 + }, + { + "epoch": 1.931390431139885, + "grad_norm": 1.8392970561981201, + "learning_rate": 6.127348018001522e-08, + "loss": 3.2382, + "step": 24091 + }, + { + "epoch": 1.9314706059209077, + "grad_norm": 1.7775070667266846, + "learning_rate": 6.113004501870023e-08, + "loss": 3.1709, + "step": 24092 + }, + { + "epoch": 1.9315507807019303, + "grad_norm": 1.8073475360870361, + "learning_rate": 6.098677742323089e-08, + "loss": 3.2365, + "step": 24093 + }, + { + "epoch": 1.931630955482953, + "grad_norm": 1.8093822002410889, + "learning_rate": 6.084367739601971e-08, + "loss": 3.2202, + "step": 24094 + }, + { + "epoch": 1.9317111302639756, + "grad_norm": 0.655513346195221, + "learning_rate": 6.070074493948364e-08, + "loss": 2.4413, + "step": 24095 + }, + { + "epoch": 1.9317913050449982, + "grad_norm": 1.7455841302871704, + "learning_rate": 6.055798005602853e-08, + "loss": 3.1134, + "step": 24096 + }, + { + "epoch": 1.9318714798260208, + "grad_norm": 0.6859582662582397, + "learning_rate": 6.041538274806247e-08, + "loss": 2.4722, + "step": 24097 + }, + { + "epoch": 1.9319516546070434, + "grad_norm": 1.7936885356903076, + "learning_rate": 6.02729530179913e-08, + "loss": 3.1527, + "step": 24098 + }, + { + "epoch": 1.932031829388066, + "grad_norm": 1.8184093236923218, + "learning_rate": 6.013069086821421e-08, + "loss": 3.2251, + "step": 24099 + }, + { + "epoch": 1.9321120041690887, + "grad_norm": 1.7581028938293457, + "learning_rate": 5.998859630113151e-08, + "loss": 3.1956, + "step": 24100 + }, + { + "epoch": 1.9321921789501113, + "grad_norm": 1.7233176231384277, + "learning_rate": 5.984666931913685e-08, + "loss": 3.1913, + "step": 24101 + }, + { + "epoch": 1.932272353731134, + "grad_norm": 1.868468999862671, + "learning_rate": 5.970490992462386e-08, + "loss": 3.2168, + "step": 24102 + }, + { + "epoch": 1.9323525285121566, + "grad_norm": 1.8446329832077026, + "learning_rate": 5.9563318119982837e-08, + "loss": 3.1613, + "step": 24103 + }, + { + "epoch": 1.9324327032931792, + "grad_norm": 1.7694387435913086, + "learning_rate": 5.942189390760189e-08, + "loss": 3.2625, + "step": 24104 + }, + { + "epoch": 1.9325128780742018, + "grad_norm": 1.7174453735351562, + "learning_rate": 5.928063728986355e-08, + "loss": 3.1818, + "step": 24105 + }, + { + "epoch": 1.9325930528552244, + "grad_norm": 1.9309899806976318, + "learning_rate": 5.913954826915147e-08, + "loss": 3.1344, + "step": 24106 + }, + { + "epoch": 1.932673227636247, + "grad_norm": 1.7939506769180298, + "learning_rate": 5.8998626847841524e-08, + "loss": 3.1742, + "step": 24107 + }, + { + "epoch": 1.9327534024172697, + "grad_norm": 1.6761631965637207, + "learning_rate": 5.8857873028311804e-08, + "loss": 3.2136, + "step": 24108 + }, + { + "epoch": 1.9328335771982923, + "grad_norm": 1.6548995971679688, + "learning_rate": 5.871728681293487e-08, + "loss": 3.1084, + "step": 24109 + }, + { + "epoch": 1.932913751979315, + "grad_norm": 1.6844711303710938, + "learning_rate": 5.857686820408104e-08, + "loss": 3.0678, + "step": 24110 + }, + { + "epoch": 1.9329939267603375, + "grad_norm": 1.7866672277450562, + "learning_rate": 5.843661720411731e-08, + "loss": 3.1617, + "step": 24111 + }, + { + "epoch": 1.9330741015413602, + "grad_norm": 1.8337856531143188, + "learning_rate": 5.829653381540845e-08, + "loss": 3.2074, + "step": 24112 + }, + { + "epoch": 1.9331542763223828, + "grad_norm": 1.7952916622161865, + "learning_rate": 5.8156618040315916e-08, + "loss": 3.1747, + "step": 24113 + }, + { + "epoch": 1.9332344511034054, + "grad_norm": 1.759858250617981, + "learning_rate": 5.8016869881198925e-08, + "loss": 3.1441, + "step": 24114 + }, + { + "epoch": 1.933314625884428, + "grad_norm": 1.907547950744629, + "learning_rate": 5.787728934041448e-08, + "loss": 3.1771, + "step": 24115 + }, + { + "epoch": 1.9333948006654507, + "grad_norm": 1.6826081275939941, + "learning_rate": 5.773787642031403e-08, + "loss": 3.1674, + "step": 24116 + }, + { + "epoch": 1.9334749754464733, + "grad_norm": 1.7904399633407593, + "learning_rate": 5.759863112324904e-08, + "loss": 3.2109, + "step": 24117 + }, + { + "epoch": 1.933555150227496, + "grad_norm": 1.9432849884033203, + "learning_rate": 5.745955345156762e-08, + "loss": 3.1915, + "step": 24118 + }, + { + "epoch": 1.9336353250085185, + "grad_norm": 1.7711251974105835, + "learning_rate": 5.732064340761345e-08, + "loss": 3.0896, + "step": 24119 + }, + { + "epoch": 1.9337154997895412, + "grad_norm": 0.6542540788650513, + "learning_rate": 5.718190099373022e-08, + "loss": 2.4537, + "step": 24120 + }, + { + "epoch": 1.9337956745705638, + "grad_norm": 1.832716464996338, + "learning_rate": 5.7043326212254944e-08, + "loss": 3.1034, + "step": 24121 + }, + { + "epoch": 1.9338758493515864, + "grad_norm": 0.6622504591941833, + "learning_rate": 5.690491906552576e-08, + "loss": 2.4609, + "step": 24122 + }, + { + "epoch": 1.933956024132609, + "grad_norm": 1.651477575302124, + "learning_rate": 5.6766679555874115e-08, + "loss": 3.204, + "step": 24123 + }, + { + "epoch": 1.9340361989136317, + "grad_norm": 1.8916984796524048, + "learning_rate": 5.662860768563261e-08, + "loss": 3.1809, + "step": 24124 + }, + { + "epoch": 1.9341163736946543, + "grad_norm": 1.806008219718933, + "learning_rate": 5.649070345712826e-08, + "loss": 3.1862, + "step": 24125 + }, + { + "epoch": 1.934196548475677, + "grad_norm": 1.809083342552185, + "learning_rate": 5.6352966872686985e-08, + "loss": 3.0689, + "step": 24126 + }, + { + "epoch": 1.9342767232566995, + "grad_norm": 1.725482702255249, + "learning_rate": 5.621539793462916e-08, + "loss": 3.2455, + "step": 24127 + }, + { + "epoch": 1.9343568980377222, + "grad_norm": 0.6577654480934143, + "learning_rate": 5.607799664527513e-08, + "loss": 2.4849, + "step": 24128 + }, + { + "epoch": 1.9344370728187448, + "grad_norm": 1.7312159538269043, + "learning_rate": 5.5940763006941956e-08, + "loss": 3.2767, + "step": 24129 + }, + { + "epoch": 1.9345172475997674, + "grad_norm": 1.7677316665649414, + "learning_rate": 5.580369702194333e-08, + "loss": 3.0879, + "step": 24130 + }, + { + "epoch": 1.93459742238079, + "grad_norm": 1.68932044506073, + "learning_rate": 5.566679869258962e-08, + "loss": 3.0942, + "step": 24131 + }, + { + "epoch": 1.9346775971618126, + "grad_norm": 1.886303186416626, + "learning_rate": 5.553006802118899e-08, + "loss": 3.24, + "step": 24132 + }, + { + "epoch": 1.9347577719428353, + "grad_norm": 1.7446225881576538, + "learning_rate": 5.5393505010045146e-08, + "loss": 3.1451, + "step": 24133 + }, + { + "epoch": 1.934837946723858, + "grad_norm": 1.7563693523406982, + "learning_rate": 5.5257109661464024e-08, + "loss": 3.2352, + "step": 24134 + }, + { + "epoch": 1.9349181215048805, + "grad_norm": 1.8471463918685913, + "learning_rate": 5.512088197774268e-08, + "loss": 3.1309, + "step": 24135 + }, + { + "epoch": 1.9349982962859031, + "grad_norm": 1.7797764539718628, + "learning_rate": 5.498482196117705e-08, + "loss": 3.1484, + "step": 24136 + }, + { + "epoch": 1.9350784710669258, + "grad_norm": 1.8203061819076538, + "learning_rate": 5.484892961406307e-08, + "loss": 3.1537, + "step": 24137 + }, + { + "epoch": 1.9351586458479484, + "grad_norm": 1.8256670236587524, + "learning_rate": 5.4713204938691144e-08, + "loss": 3.1023, + "step": 24138 + }, + { + "epoch": 1.935238820628971, + "grad_norm": 1.8428243398666382, + "learning_rate": 5.457764793735054e-08, + "loss": 3.2192, + "step": 24139 + }, + { + "epoch": 1.9353189954099936, + "grad_norm": 1.7811561822891235, + "learning_rate": 5.4442258612324996e-08, + "loss": 3.2318, + "step": 24140 + }, + { + "epoch": 1.9353991701910163, + "grad_norm": 1.7051572799682617, + "learning_rate": 5.4307036965897117e-08, + "loss": 3.1868, + "step": 24141 + }, + { + "epoch": 1.9354793449720389, + "grad_norm": 1.8158774375915527, + "learning_rate": 5.417198300034843e-08, + "loss": 3.2791, + "step": 24142 + }, + { + "epoch": 1.9355595197530615, + "grad_norm": 1.7981479167938232, + "learning_rate": 5.403709671795376e-08, + "loss": 3.1554, + "step": 24143 + }, + { + "epoch": 1.9356396945340844, + "grad_norm": 1.8520982265472412, + "learning_rate": 5.3902378120989084e-08, + "loss": 3.169, + "step": 24144 + }, + { + "epoch": 1.935719869315107, + "grad_norm": 1.8786016702651978, + "learning_rate": 5.37678272117248e-08, + "loss": 3.2306, + "step": 24145 + }, + { + "epoch": 1.9358000440961296, + "grad_norm": 1.7451388835906982, + "learning_rate": 5.363344399242909e-08, + "loss": 3.0944, + "step": 24146 + }, + { + "epoch": 1.9358802188771522, + "grad_norm": 1.7458614110946655, + "learning_rate": 5.3499228465369035e-08, + "loss": 3.1669, + "step": 24147 + }, + { + "epoch": 1.9359603936581748, + "grad_norm": 1.7343910932540894, + "learning_rate": 5.336518063280616e-08, + "loss": 3.2134, + "step": 24148 + }, + { + "epoch": 1.9360405684391975, + "grad_norm": 1.8196055889129639, + "learning_rate": 5.323130049699976e-08, + "loss": 3.1651, + "step": 24149 + }, + { + "epoch": 1.93612074322022, + "grad_norm": 1.7017093896865845, + "learning_rate": 5.3097588060210256e-08, + "loss": 3.1194, + "step": 24150 + }, + { + "epoch": 1.9362009180012427, + "grad_norm": 1.827125906944275, + "learning_rate": 5.296404332468807e-08, + "loss": 3.215, + "step": 24151 + }, + { + "epoch": 1.9362810927822653, + "grad_norm": 1.73848295211792, + "learning_rate": 5.283066629268696e-08, + "loss": 3.1346, + "step": 24152 + }, + { + "epoch": 1.936361267563288, + "grad_norm": 1.8174058198928833, + "learning_rate": 5.2697456966455115e-08, + "loss": 3.2001, + "step": 24153 + }, + { + "epoch": 1.9364414423443106, + "grad_norm": 1.793196439743042, + "learning_rate": 5.2564415348237416e-08, + "loss": 3.0656, + "step": 24154 + }, + { + "epoch": 1.9365216171253332, + "grad_norm": 1.8381136655807495, + "learning_rate": 5.243154144027984e-08, + "loss": 3.2197, + "step": 24155 + }, + { + "epoch": 1.9366017919063558, + "grad_norm": 0.6754391193389893, + "learning_rate": 5.229883524481838e-08, + "loss": 2.4579, + "step": 24156 + }, + { + "epoch": 1.9366819666873785, + "grad_norm": 1.65906822681427, + "learning_rate": 5.216629676409457e-08, + "loss": 3.0911, + "step": 24157 + }, + { + "epoch": 1.936762141468401, + "grad_norm": 1.8660322427749634, + "learning_rate": 5.203392600033996e-08, + "loss": 3.1837, + "step": 24158 + }, + { + "epoch": 1.9368423162494237, + "grad_norm": 1.7752684354782104, + "learning_rate": 5.1901722955787214e-08, + "loss": 3.166, + "step": 24159 + }, + { + "epoch": 1.9369224910304463, + "grad_norm": 1.832574725151062, + "learning_rate": 5.176968763266565e-08, + "loss": 3.2175, + "step": 24160 + }, + { + "epoch": 1.937002665811469, + "grad_norm": 1.7559571266174316, + "learning_rate": 5.163782003320128e-08, + "loss": 3.0552, + "step": 24161 + }, + { + "epoch": 1.9370828405924916, + "grad_norm": 1.8448271751403809, + "learning_rate": 5.150612015961676e-08, + "loss": 3.0869, + "step": 24162 + }, + { + "epoch": 1.9371630153735142, + "grad_norm": 0.6445661187171936, + "learning_rate": 5.137458801413364e-08, + "loss": 2.468, + "step": 24163 + }, + { + "epoch": 1.937243190154537, + "grad_norm": 1.7605180740356445, + "learning_rate": 5.124322359896794e-08, + "loss": 3.1604, + "step": 24164 + }, + { + "epoch": 1.9373233649355597, + "grad_norm": 1.7842001914978027, + "learning_rate": 5.111202691633566e-08, + "loss": 3.1458, + "step": 24165 + }, + { + "epoch": 1.9374035397165823, + "grad_norm": 1.9089337587356567, + "learning_rate": 5.098099796844835e-08, + "loss": 3.104, + "step": 24166 + }, + { + "epoch": 1.937483714497605, + "grad_norm": 1.791911244392395, + "learning_rate": 5.085013675751427e-08, + "loss": 3.1575, + "step": 24167 + }, + { + "epoch": 1.9375638892786275, + "grad_norm": 1.7572919130325317, + "learning_rate": 5.071944328574163e-08, + "loss": 3.2042, + "step": 24168 + }, + { + "epoch": 1.9376440640596502, + "grad_norm": 1.8199437856674194, + "learning_rate": 5.058891755533091e-08, + "loss": 3.1132, + "step": 24169 + }, + { + "epoch": 1.9377242388406728, + "grad_norm": 1.771911859512329, + "learning_rate": 5.0458559568487e-08, + "loss": 3.2151, + "step": 24170 + }, + { + "epoch": 1.9378044136216954, + "grad_norm": 1.7421411275863647, + "learning_rate": 5.0328369327403706e-08, + "loss": 3.1686, + "step": 24171 + }, + { + "epoch": 1.937884588402718, + "grad_norm": 1.7650721073150635, + "learning_rate": 5.019834683427704e-08, + "loss": 3.158, + "step": 24172 + }, + { + "epoch": 1.9379647631837407, + "grad_norm": 1.6840662956237793, + "learning_rate": 5.006849209130082e-08, + "loss": 3.2537, + "step": 24173 + }, + { + "epoch": 1.9380449379647633, + "grad_norm": 1.8786202669143677, + "learning_rate": 4.993880510066218e-08, + "loss": 3.2604, + "step": 24174 + }, + { + "epoch": 1.938125112745786, + "grad_norm": 1.7520277500152588, + "learning_rate": 4.9809285864549365e-08, + "loss": 3.0983, + "step": 24175 + }, + { + "epoch": 1.9382052875268085, + "grad_norm": 1.8541064262390137, + "learning_rate": 4.967993438514396e-08, + "loss": 3.2041, + "step": 24176 + }, + { + "epoch": 1.9382854623078312, + "grad_norm": 1.8023886680603027, + "learning_rate": 4.955075066462867e-08, + "loss": 3.1833, + "step": 24177 + }, + { + "epoch": 1.9383656370888538, + "grad_norm": 0.6567152738571167, + "learning_rate": 4.942173470518175e-08, + "loss": 2.4462, + "step": 24178 + }, + { + "epoch": 1.9384458118698764, + "grad_norm": 1.858669638633728, + "learning_rate": 4.9292886508975904e-08, + "loss": 3.1464, + "step": 24179 + }, + { + "epoch": 1.938525986650899, + "grad_norm": 1.6426256895065308, + "learning_rate": 4.916420607818606e-08, + "loss": 3.1975, + "step": 24180 + }, + { + "epoch": 1.9386061614319217, + "grad_norm": 1.849617838859558, + "learning_rate": 4.903569341498049e-08, + "loss": 3.2147, + "step": 24181 + }, + { + "epoch": 1.9386863362129443, + "grad_norm": 1.7944839000701904, + "learning_rate": 4.8907348521526346e-08, + "loss": 3.1694, + "step": 24182 + }, + { + "epoch": 1.938766510993967, + "grad_norm": 1.8252509832382202, + "learning_rate": 4.8779171399986336e-08, + "loss": 3.16, + "step": 24183 + }, + { + "epoch": 1.9388466857749895, + "grad_norm": 1.7658674716949463, + "learning_rate": 4.8651162052523185e-08, + "loss": 3.1731, + "step": 24184 + }, + { + "epoch": 1.9389268605560122, + "grad_norm": 1.7482792139053345, + "learning_rate": 4.852332048129405e-08, + "loss": 3.2032, + "step": 24185 + }, + { + "epoch": 1.9390070353370348, + "grad_norm": 1.7656632661819458, + "learning_rate": 4.839564668845498e-08, + "loss": 3.0853, + "step": 24186 + }, + { + "epoch": 1.9390872101180574, + "grad_norm": 1.7723720073699951, + "learning_rate": 4.826814067615648e-08, + "loss": 3.1455, + "step": 24187 + }, + { + "epoch": 1.93916738489908, + "grad_norm": 1.7657454013824463, + "learning_rate": 4.814080244655128e-08, + "loss": 3.1024, + "step": 24188 + }, + { + "epoch": 1.9392475596801027, + "grad_norm": 1.839219331741333, + "learning_rate": 4.8013632001784325e-08, + "loss": 3.1948, + "step": 24189 + }, + { + "epoch": 1.9393277344611253, + "grad_norm": 1.8041425943374634, + "learning_rate": 4.788662934399946e-08, + "loss": 3.2955, + "step": 24190 + }, + { + "epoch": 1.939407909242148, + "grad_norm": 1.8678534030914307, + "learning_rate": 4.7759794475339405e-08, + "loss": 3.2424, + "step": 24191 + }, + { + "epoch": 1.9394880840231705, + "grad_norm": 1.8251094818115234, + "learning_rate": 4.763312739794135e-08, + "loss": 3.1272, + "step": 24192 + }, + { + "epoch": 1.9395682588041931, + "grad_norm": 1.7824355363845825, + "learning_rate": 4.7506628113941354e-08, + "loss": 3.2242, + "step": 24193 + }, + { + "epoch": 1.9396484335852158, + "grad_norm": 1.6894925832748413, + "learning_rate": 4.7380296625472163e-08, + "loss": 3.256, + "step": 24194 + }, + { + "epoch": 1.9397286083662384, + "grad_norm": 0.6489450931549072, + "learning_rate": 4.725413293466319e-08, + "loss": 2.4484, + "step": 24195 + }, + { + "epoch": 1.939808783147261, + "grad_norm": 1.6541221141815186, + "learning_rate": 4.712813704364161e-08, + "loss": 3.0641, + "step": 24196 + }, + { + "epoch": 1.9398889579282836, + "grad_norm": 1.9393558502197266, + "learning_rate": 4.700230895453239e-08, + "loss": 3.1845, + "step": 24197 + }, + { + "epoch": 1.9399691327093063, + "grad_norm": 1.7430095672607422, + "learning_rate": 4.6876648669456074e-08, + "loss": 3.2729, + "step": 24198 + }, + { + "epoch": 1.940049307490329, + "grad_norm": 1.7693703174591064, + "learning_rate": 4.675115619053206e-08, + "loss": 3.1474, + "step": 24199 + }, + { + "epoch": 1.9401294822713515, + "grad_norm": 1.7881147861480713, + "learning_rate": 4.662583151987643e-08, + "loss": 3.1248, + "step": 24200 + }, + { + "epoch": 1.9402096570523741, + "grad_norm": 1.6918375492095947, + "learning_rate": 4.650067465959973e-08, + "loss": 3.2344, + "step": 24201 + }, + { + "epoch": 1.9402898318333968, + "grad_norm": 1.8215123414993286, + "learning_rate": 4.637568561181472e-08, + "loss": 3.1171, + "step": 24202 + }, + { + "epoch": 1.9403700066144194, + "grad_norm": 1.8741066455841064, + "learning_rate": 4.625086437862747e-08, + "loss": 3.1558, + "step": 24203 + }, + { + "epoch": 1.940450181395442, + "grad_norm": 1.8020806312561035, + "learning_rate": 4.612621096214187e-08, + "loss": 3.1177, + "step": 24204 + }, + { + "epoch": 1.9405303561764646, + "grad_norm": 1.766041874885559, + "learning_rate": 4.600172536446179e-08, + "loss": 3.2107, + "step": 24205 + }, + { + "epoch": 1.9406105309574873, + "grad_norm": 1.7650761604309082, + "learning_rate": 4.587740758768333e-08, + "loss": 3.0892, + "step": 24206 + }, + { + "epoch": 1.9406907057385099, + "grad_norm": 1.799233078956604, + "learning_rate": 4.57532576339037e-08, + "loss": 3.197, + "step": 24207 + }, + { + "epoch": 1.9407708805195325, + "grad_norm": 1.76326322555542, + "learning_rate": 4.5629275505215676e-08, + "loss": 3.166, + "step": 24208 + }, + { + "epoch": 1.9408510553005551, + "grad_norm": 1.7586138248443604, + "learning_rate": 4.550546120371091e-08, + "loss": 3.189, + "step": 24209 + }, + { + "epoch": 1.9409312300815778, + "grad_norm": 1.8511974811553955, + "learning_rate": 4.5381814731475514e-08, + "loss": 3.2024, + "step": 24210 + }, + { + "epoch": 1.9410114048626004, + "grad_norm": 1.8000097274780273, + "learning_rate": 4.525833609059338e-08, + "loss": 3.1239, + "step": 24211 + }, + { + "epoch": 1.941091579643623, + "grad_norm": 1.8687200546264648, + "learning_rate": 4.5135025283148395e-08, + "loss": 3.3282, + "step": 24212 + }, + { + "epoch": 1.9411717544246456, + "grad_norm": 1.650010585784912, + "learning_rate": 4.501188231121778e-08, + "loss": 3.0985, + "step": 24213 + }, + { + "epoch": 1.9412519292056682, + "grad_norm": 1.8388479948043823, + "learning_rate": 4.488890717687766e-08, + "loss": 3.1694, + "step": 24214 + }, + { + "epoch": 1.9413321039866909, + "grad_norm": 1.7573820352554321, + "learning_rate": 4.476609988220304e-08, + "loss": 3.0489, + "step": 24215 + }, + { + "epoch": 1.9414122787677135, + "grad_norm": 1.821196436882019, + "learning_rate": 4.464346042926337e-08, + "loss": 3.2524, + "step": 24216 + }, + { + "epoch": 1.9414924535487361, + "grad_norm": 1.9002501964569092, + "learning_rate": 4.452098882012701e-08, + "loss": 3.2094, + "step": 24217 + }, + { + "epoch": 1.9415726283297587, + "grad_norm": 1.8274359703063965, + "learning_rate": 4.439868505685674e-08, + "loss": 3.1775, + "step": 24218 + }, + { + "epoch": 1.9416528031107814, + "grad_norm": 1.7696666717529297, + "learning_rate": 4.4276549141516466e-08, + "loss": 3.0885, + "step": 24219 + }, + { + "epoch": 1.941732977891804, + "grad_norm": 1.7713623046875, + "learning_rate": 4.415458107616566e-08, + "loss": 3.2096, + "step": 24220 + }, + { + "epoch": 1.9418131526728266, + "grad_norm": 1.6914361715316772, + "learning_rate": 4.4032780862858227e-08, + "loss": 3.1309, + "step": 24221 + }, + { + "epoch": 1.9418933274538492, + "grad_norm": 1.8817874193191528, + "learning_rate": 4.39111485036503e-08, + "loss": 3.2269, + "step": 24222 + }, + { + "epoch": 1.9419735022348719, + "grad_norm": 1.7763973474502563, + "learning_rate": 4.378968400059136e-08, + "loss": 3.1807, + "step": 24223 + }, + { + "epoch": 1.9420536770158945, + "grad_norm": 1.8725098371505737, + "learning_rate": 4.3668387355729756e-08, + "loss": 3.1559, + "step": 24224 + }, + { + "epoch": 1.9421338517969173, + "grad_norm": 1.7530615329742432, + "learning_rate": 4.354725857111053e-08, + "loss": 3.1151, + "step": 24225 + }, + { + "epoch": 1.94221402657794, + "grad_norm": 1.7620034217834473, + "learning_rate": 4.3426297648774265e-08, + "loss": 3.1639, + "step": 24226 + }, + { + "epoch": 1.9422942013589626, + "grad_norm": 1.7879819869995117, + "learning_rate": 4.330550459076266e-08, + "loss": 3.2412, + "step": 24227 + }, + { + "epoch": 1.9423743761399852, + "grad_norm": 1.5977463722229004, + "learning_rate": 4.318487939911076e-08, + "loss": 3.1177, + "step": 24228 + }, + { + "epoch": 1.9424545509210078, + "grad_norm": 1.8381781578063965, + "learning_rate": 4.3064422075852485e-08, + "loss": 3.136, + "step": 24229 + }, + { + "epoch": 1.9425347257020305, + "grad_norm": 1.6281408071517944, + "learning_rate": 4.294413262301955e-08, + "loss": 3.1492, + "step": 24230 + }, + { + "epoch": 1.942614900483053, + "grad_norm": 1.6790366172790527, + "learning_rate": 4.282401104263812e-08, + "loss": 3.1486, + "step": 24231 + }, + { + "epoch": 1.9426950752640757, + "grad_norm": 1.7081409692764282, + "learning_rate": 4.270405733673544e-08, + "loss": 3.1309, + "step": 24232 + }, + { + "epoch": 1.9427752500450983, + "grad_norm": 1.8606526851654053, + "learning_rate": 4.258427150733324e-08, + "loss": 3.2609, + "step": 24233 + }, + { + "epoch": 1.942855424826121, + "grad_norm": 1.927072286605835, + "learning_rate": 4.2464653556449906e-08, + "loss": 3.1265, + "step": 24234 + }, + { + "epoch": 1.9429355996071436, + "grad_norm": 1.8247060775756836, + "learning_rate": 4.234520348610494e-08, + "loss": 3.24, + "step": 24235 + }, + { + "epoch": 1.9430157743881662, + "grad_norm": 1.7565357685089111, + "learning_rate": 4.222592129830894e-08, + "loss": 3.1733, + "step": 24236 + }, + { + "epoch": 1.9430959491691888, + "grad_norm": 1.8039408922195435, + "learning_rate": 4.2106806995074744e-08, + "loss": 3.1773, + "step": 24237 + }, + { + "epoch": 1.9431761239502114, + "grad_norm": 1.7054067850112915, + "learning_rate": 4.1987860578410755e-08, + "loss": 3.1761, + "step": 24238 + }, + { + "epoch": 1.943256298731234, + "grad_norm": 1.8500354290008545, + "learning_rate": 4.186908205032092e-08, + "loss": 3.1331, + "step": 24239 + }, + { + "epoch": 1.9433364735122567, + "grad_norm": 1.7148693799972534, + "learning_rate": 4.17504714128103e-08, + "loss": 3.1155, + "step": 24240 + }, + { + "epoch": 1.9434166482932793, + "grad_norm": 1.7805664539337158, + "learning_rate": 4.163202866787508e-08, + "loss": 3.2145, + "step": 24241 + }, + { + "epoch": 1.943496823074302, + "grad_norm": 1.7669986486434937, + "learning_rate": 4.1513753817515875e-08, + "loss": 3.1595, + "step": 24242 + }, + { + "epoch": 1.9435769978553246, + "grad_norm": 1.7652815580368042, + "learning_rate": 4.1395646863724435e-08, + "loss": 3.1661, + "step": 24243 + }, + { + "epoch": 1.9436571726363472, + "grad_norm": 1.6983225345611572, + "learning_rate": 4.1277707808492495e-08, + "loss": 3.0373, + "step": 24244 + }, + { + "epoch": 1.94373734741737, + "grad_norm": 1.7545897960662842, + "learning_rate": 4.115993665380957e-08, + "loss": 3.2371, + "step": 24245 + }, + { + "epoch": 1.9438175221983927, + "grad_norm": 1.7398544549942017, + "learning_rate": 4.104233340165853e-08, + "loss": 3.0817, + "step": 24246 + }, + { + "epoch": 1.9438976969794153, + "grad_norm": 1.7817291021347046, + "learning_rate": 4.092489805402555e-08, + "loss": 3.1764, + "step": 24247 + }, + { + "epoch": 1.943977871760438, + "grad_norm": 1.7865843772888184, + "learning_rate": 4.080763061288795e-08, + "loss": 3.2012, + "step": 24248 + }, + { + "epoch": 1.9440580465414605, + "grad_norm": 1.8419617414474487, + "learning_rate": 4.069053108022414e-08, + "loss": 3.2253, + "step": 24249 + }, + { + "epoch": 1.9441382213224832, + "grad_norm": 1.756630301475525, + "learning_rate": 4.0573599458008096e-08, + "loss": 3.2344, + "step": 24250 + }, + { + "epoch": 1.9442183961035058, + "grad_norm": 1.7192403078079224, + "learning_rate": 4.045683574821047e-08, + "loss": 3.1305, + "step": 24251 + }, + { + "epoch": 1.9442985708845284, + "grad_norm": 1.7595853805541992, + "learning_rate": 4.034023995280078e-08, + "loss": 3.1748, + "step": 24252 + }, + { + "epoch": 1.944378745665551, + "grad_norm": 1.7387068271636963, + "learning_rate": 4.0223812073744154e-08, + "loss": 3.1495, + "step": 24253 + }, + { + "epoch": 1.9444589204465736, + "grad_norm": 1.7258168458938599, + "learning_rate": 4.0107552113004545e-08, + "loss": 3.1725, + "step": 24254 + }, + { + "epoch": 1.9445390952275963, + "grad_norm": 1.773398995399475, + "learning_rate": 3.999146007254151e-08, + "loss": 3.0781, + "step": 24255 + }, + { + "epoch": 1.944619270008619, + "grad_norm": 0.662235677242279, + "learning_rate": 3.9875535954311265e-08, + "loss": 2.4424, + "step": 24256 + }, + { + "epoch": 1.9446994447896415, + "grad_norm": 1.8331632614135742, + "learning_rate": 3.975977976027001e-08, + "loss": 3.2057, + "step": 24257 + }, + { + "epoch": 1.9447796195706641, + "grad_norm": 1.7322207689285278, + "learning_rate": 3.964419149236842e-08, + "loss": 3.1599, + "step": 24258 + }, + { + "epoch": 1.9448597943516868, + "grad_norm": 1.736255168914795, + "learning_rate": 3.9528771152554934e-08, + "loss": 3.1361, + "step": 24259 + }, + { + "epoch": 1.9449399691327094, + "grad_norm": 1.7852346897125244, + "learning_rate": 3.941351874277688e-08, + "loss": 3.171, + "step": 24260 + }, + { + "epoch": 1.945020143913732, + "grad_norm": 1.7861863374710083, + "learning_rate": 3.929843426497604e-08, + "loss": 3.246, + "step": 24261 + }, + { + "epoch": 1.9451003186947546, + "grad_norm": 0.654621958732605, + "learning_rate": 3.918351772109308e-08, + "loss": 2.4545, + "step": 24262 + }, + { + "epoch": 1.9451804934757773, + "grad_norm": 1.8105219602584839, + "learning_rate": 3.906876911306534e-08, + "loss": 3.1282, + "step": 24263 + }, + { + "epoch": 1.9452606682567999, + "grad_norm": 1.6696792840957642, + "learning_rate": 3.895418844282683e-08, + "loss": 3.0974, + "step": 24264 + }, + { + "epoch": 1.9453408430378225, + "grad_norm": 1.8299401998519897, + "learning_rate": 3.883977571231157e-08, + "loss": 3.2631, + "step": 24265 + }, + { + "epoch": 1.9454210178188451, + "grad_norm": 1.7716844081878662, + "learning_rate": 3.8725530923445776e-08, + "loss": 3.1973, + "step": 24266 + }, + { + "epoch": 1.9455011925998678, + "grad_norm": 1.9686129093170166, + "learning_rate": 3.861145407815792e-08, + "loss": 3.1658, + "step": 24267 + }, + { + "epoch": 1.9455813673808904, + "grad_norm": 1.91056227684021, + "learning_rate": 3.849754517836979e-08, + "loss": 3.181, + "step": 24268 + }, + { + "epoch": 1.945661542161913, + "grad_norm": 1.671480417251587, + "learning_rate": 3.8383804226002074e-08, + "loss": 3.2696, + "step": 24269 + }, + { + "epoch": 1.9457417169429356, + "grad_norm": 1.7753499746322632, + "learning_rate": 3.827023122297324e-08, + "loss": 3.1816, + "step": 24270 + }, + { + "epoch": 1.9458218917239583, + "grad_norm": 1.9204614162445068, + "learning_rate": 3.8156826171196204e-08, + "loss": 3.1876, + "step": 24271 + }, + { + "epoch": 1.9459020665049809, + "grad_norm": 1.7430962324142456, + "learning_rate": 3.8043589072584984e-08, + "loss": 3.0474, + "step": 24272 + }, + { + "epoch": 1.9459822412860035, + "grad_norm": 1.678347110748291, + "learning_rate": 3.793051992904806e-08, + "loss": 3.1294, + "step": 24273 + }, + { + "epoch": 1.9460624160670261, + "grad_norm": 1.6807267665863037, + "learning_rate": 3.781761874249057e-08, + "loss": 3.0977, + "step": 24274 + }, + { + "epoch": 1.9461425908480487, + "grad_norm": 1.7853515148162842, + "learning_rate": 3.770488551481877e-08, + "loss": 3.2617, + "step": 24275 + }, + { + "epoch": 1.9462227656290714, + "grad_norm": 1.706695556640625, + "learning_rate": 3.7592320247930024e-08, + "loss": 3.1819, + "step": 24276 + }, + { + "epoch": 1.946302940410094, + "grad_norm": 1.7210276126861572, + "learning_rate": 3.747992294372393e-08, + "loss": 3.0876, + "step": 24277 + }, + { + "epoch": 1.9463831151911166, + "grad_norm": 1.7850472927093506, + "learning_rate": 3.7367693604095645e-08, + "loss": 3.2131, + "step": 24278 + }, + { + "epoch": 1.9464632899721392, + "grad_norm": 1.7178528308868408, + "learning_rate": 3.7255632230936977e-08, + "loss": 3.2045, + "step": 24279 + }, + { + "epoch": 1.9465434647531619, + "grad_norm": 1.6448575258255005, + "learning_rate": 3.714373882613753e-08, + "loss": 3.1393, + "step": 24280 + }, + { + "epoch": 1.9466236395341845, + "grad_norm": 1.706754446029663, + "learning_rate": 3.703201339158247e-08, + "loss": 3.1555, + "step": 24281 + }, + { + "epoch": 1.9467038143152071, + "grad_norm": 1.8936011791229248, + "learning_rate": 3.692045592915694e-08, + "loss": 3.1566, + "step": 24282 + }, + { + "epoch": 1.9467839890962297, + "grad_norm": 1.7842057943344116, + "learning_rate": 3.680906644074056e-08, + "loss": 3.1766, + "step": 24283 + }, + { + "epoch": 1.9468641638772524, + "grad_norm": 1.7607123851776123, + "learning_rate": 3.669784492821182e-08, + "loss": 3.1962, + "step": 24284 + }, + { + "epoch": 1.946944338658275, + "grad_norm": 1.6700465679168701, + "learning_rate": 3.658679139344701e-08, + "loss": 3.1782, + "step": 24285 + }, + { + "epoch": 1.9470245134392976, + "grad_norm": 1.7036298513412476, + "learning_rate": 3.647590583831684e-08, + "loss": 3.0823, + "step": 24286 + }, + { + "epoch": 1.9471046882203202, + "grad_norm": 0.6465633511543274, + "learning_rate": 3.636518826469204e-08, + "loss": 2.4602, + "step": 24287 + }, + { + "epoch": 1.9471848630013429, + "grad_norm": 1.744966983795166, + "learning_rate": 3.6254638674437795e-08, + "loss": 3.2013, + "step": 24288 + }, + { + "epoch": 1.9472650377823655, + "grad_norm": 1.6812407970428467, + "learning_rate": 3.614425706941926e-08, + "loss": 3.2668, + "step": 24289 + }, + { + "epoch": 1.947345212563388, + "grad_norm": 1.7218222618103027, + "learning_rate": 3.6034043451496084e-08, + "loss": 3.2405, + "step": 24290 + }, + { + "epoch": 1.9474253873444107, + "grad_norm": 0.6495417356491089, + "learning_rate": 3.592399782252787e-08, + "loss": 2.4473, + "step": 24291 + }, + { + "epoch": 1.9475055621254334, + "grad_norm": 1.96772301197052, + "learning_rate": 3.581412018436981e-08, + "loss": 3.1632, + "step": 24292 + }, + { + "epoch": 1.947585736906456, + "grad_norm": 1.7685331106185913, + "learning_rate": 3.5704410538873743e-08, + "loss": 3.1252, + "step": 24293 + }, + { + "epoch": 1.9476659116874786, + "grad_norm": 1.7674002647399902, + "learning_rate": 3.559486888788932e-08, + "loss": 3.2161, + "step": 24294 + }, + { + "epoch": 1.9477460864685012, + "grad_norm": 1.7923251390457153, + "learning_rate": 3.5485495233263945e-08, + "loss": 3.2067, + "step": 24295 + }, + { + "epoch": 1.9478262612495238, + "grad_norm": 1.737197756767273, + "learning_rate": 3.5376289576841694e-08, + "loss": 3.1412, + "step": 24296 + }, + { + "epoch": 1.9479064360305465, + "grad_norm": 1.7511874437332153, + "learning_rate": 3.526725192046221e-08, + "loss": 3.0706, + "step": 24297 + }, + { + "epoch": 1.947986610811569, + "grad_norm": 1.8319878578186035, + "learning_rate": 3.515838226596513e-08, + "loss": 3.1804, + "step": 24298 + }, + { + "epoch": 1.9480667855925917, + "grad_norm": 1.8680825233459473, + "learning_rate": 3.504968061518565e-08, + "loss": 3.1608, + "step": 24299 + }, + { + "epoch": 1.9481469603736143, + "grad_norm": 1.786470651626587, + "learning_rate": 3.494114696995787e-08, + "loss": 3.2156, + "step": 24300 + }, + { + "epoch": 1.948227135154637, + "grad_norm": 1.7509604692459106, + "learning_rate": 3.483278133210921e-08, + "loss": 3.0601, + "step": 24301 + }, + { + "epoch": 1.9483073099356596, + "grad_norm": 1.7021816968917847, + "learning_rate": 3.4724583703468206e-08, + "loss": 3.2153, + "step": 24302 + }, + { + "epoch": 1.9483874847166822, + "grad_norm": 1.7224421501159668, + "learning_rate": 3.461655408585784e-08, + "loss": 3.1353, + "step": 24303 + }, + { + "epoch": 1.9484676594977048, + "grad_norm": 1.7431647777557373, + "learning_rate": 3.450869248110111e-08, + "loss": 3.1632, + "step": 24304 + }, + { + "epoch": 1.9485478342787275, + "grad_norm": 1.7325712442398071, + "learning_rate": 3.440099889101545e-08, + "loss": 3.1344, + "step": 24305 + }, + { + "epoch": 1.9486280090597503, + "grad_norm": 1.7748165130615234, + "learning_rate": 3.429347331741717e-08, + "loss": 3.1072, + "step": 24306 + }, + { + "epoch": 1.948708183840773, + "grad_norm": 1.8305871486663818, + "learning_rate": 3.4186115762118165e-08, + "loss": 3.1862, + "step": 24307 + }, + { + "epoch": 1.9487883586217956, + "grad_norm": 1.825939655303955, + "learning_rate": 3.407892622692921e-08, + "loss": 3.2002, + "step": 24308 + }, + { + "epoch": 1.9488685334028182, + "grad_norm": 1.6932276487350464, + "learning_rate": 3.3971904713656635e-08, + "loss": 3.1249, + "step": 24309 + }, + { + "epoch": 1.9489487081838408, + "grad_norm": 0.6741396188735962, + "learning_rate": 3.386505122410566e-08, + "loss": 2.4873, + "step": 24310 + }, + { + "epoch": 1.9490288829648634, + "grad_norm": 1.7220395803451538, + "learning_rate": 3.3758365760078184e-08, + "loss": 3.1069, + "step": 24311 + }, + { + "epoch": 1.949109057745886, + "grad_norm": 1.7662553787231445, + "learning_rate": 3.365184832337054e-08, + "loss": 3.1756, + "step": 24312 + }, + { + "epoch": 1.9491892325269087, + "grad_norm": 1.7630432844161987, + "learning_rate": 3.354549891578129e-08, + "loss": 3.0879, + "step": 24313 + }, + { + "epoch": 1.9492694073079313, + "grad_norm": 1.7318342924118042, + "learning_rate": 3.343931753910234e-08, + "loss": 3.1148, + "step": 24314 + }, + { + "epoch": 1.949349582088954, + "grad_norm": 1.7927861213684082, + "learning_rate": 3.3333304195124484e-08, + "loss": 3.1779, + "step": 24315 + }, + { + "epoch": 1.9494297568699765, + "grad_norm": 1.9243029356002808, + "learning_rate": 3.3227458885634056e-08, + "loss": 3.1724, + "step": 24316 + }, + { + "epoch": 1.9495099316509992, + "grad_norm": 1.848963737487793, + "learning_rate": 3.3121781612416306e-08, + "loss": 3.1443, + "step": 24317 + }, + { + "epoch": 1.9495901064320218, + "grad_norm": 1.7534390687942505, + "learning_rate": 3.301627237725202e-08, + "loss": 3.0659, + "step": 24318 + }, + { + "epoch": 1.9496702812130444, + "grad_norm": 1.8106061220169067, + "learning_rate": 3.29109311819209e-08, + "loss": 3.0987, + "step": 24319 + }, + { + "epoch": 1.949750455994067, + "grad_norm": 0.6498285531997681, + "learning_rate": 3.28057580281993e-08, + "loss": 2.4611, + "step": 24320 + }, + { + "epoch": 1.9498306307750897, + "grad_norm": 0.667954683303833, + "learning_rate": 3.2700752917859126e-08, + "loss": 2.4673, + "step": 24321 + }, + { + "epoch": 1.9499108055561123, + "grad_norm": 1.7695804834365845, + "learning_rate": 3.25959158526723e-08, + "loss": 3.0567, + "step": 24322 + }, + { + "epoch": 1.949990980337135, + "grad_norm": 1.783374309539795, + "learning_rate": 3.249124683440519e-08, + "loss": 3.1298, + "step": 24323 + }, + { + "epoch": 1.9500711551181575, + "grad_norm": 1.6764410734176636, + "learning_rate": 3.2386745864821935e-08, + "loss": 3.2737, + "step": 24324 + }, + { + "epoch": 1.9501513298991802, + "grad_norm": 1.7089546918869019, + "learning_rate": 3.228241294568668e-08, + "loss": 3.1606, + "step": 24325 + }, + { + "epoch": 1.9502315046802028, + "grad_norm": 1.8594996929168701, + "learning_rate": 3.2178248078755805e-08, + "loss": 3.2221, + "step": 24326 + }, + { + "epoch": 1.9503116794612256, + "grad_norm": 1.711220383644104, + "learning_rate": 3.207425126578678e-08, + "loss": 3.1866, + "step": 24327 + }, + { + "epoch": 1.9503918542422483, + "grad_norm": 0.6556041240692139, + "learning_rate": 3.1970422508532664e-08, + "loss": 2.4625, + "step": 24328 + }, + { + "epoch": 1.9504720290232709, + "grad_norm": 1.735788345336914, + "learning_rate": 3.186676180874426e-08, + "loss": 3.1755, + "step": 24329 + }, + { + "epoch": 1.9505522038042935, + "grad_norm": 1.794686198234558, + "learning_rate": 3.176326916817019e-08, + "loss": 3.1806, + "step": 24330 + }, + { + "epoch": 1.9506323785853161, + "grad_norm": 1.6793333292007446, + "learning_rate": 3.165994458855348e-08, + "loss": 3.2428, + "step": 24331 + }, + { + "epoch": 1.9507125533663388, + "grad_norm": 1.7631924152374268, + "learning_rate": 3.155678807163609e-08, + "loss": 3.2051, + "step": 24332 + }, + { + "epoch": 1.9507927281473614, + "grad_norm": 1.7697081565856934, + "learning_rate": 3.145379961915884e-08, + "loss": 3.1473, + "step": 24333 + }, + { + "epoch": 1.950872902928384, + "grad_norm": 0.6655763983726501, + "learning_rate": 3.135097923285702e-08, + "loss": 2.4636, + "step": 24334 + }, + { + "epoch": 1.9509530777094066, + "grad_norm": 1.7440807819366455, + "learning_rate": 3.124832691446589e-08, + "loss": 3.1339, + "step": 24335 + }, + { + "epoch": 1.9510332524904292, + "grad_norm": 1.8770679235458374, + "learning_rate": 3.114584266571297e-08, + "loss": 3.1788, + "step": 24336 + }, + { + "epoch": 1.9511134272714519, + "grad_norm": 1.6517802476882935, + "learning_rate": 3.104352648832798e-08, + "loss": 3.181, + "step": 24337 + }, + { + "epoch": 1.9511936020524745, + "grad_norm": 1.803544044494629, + "learning_rate": 3.0941378384036215e-08, + "loss": 3.1486, + "step": 24338 + }, + { + "epoch": 1.9512737768334971, + "grad_norm": 1.8912259340286255, + "learning_rate": 3.083939835455851e-08, + "loss": 3.1313, + "step": 24339 + }, + { + "epoch": 1.9513539516145197, + "grad_norm": 1.7751245498657227, + "learning_rate": 3.073758640161573e-08, + "loss": 3.1218, + "step": 24340 + }, + { + "epoch": 1.9514341263955424, + "grad_norm": 1.768571376800537, + "learning_rate": 3.063594252692425e-08, + "loss": 3.1157, + "step": 24341 + }, + { + "epoch": 1.951514301176565, + "grad_norm": 1.6531416177749634, + "learning_rate": 3.053446673219607e-08, + "loss": 3.0773, + "step": 24342 + }, + { + "epoch": 1.9515944759575876, + "grad_norm": 1.728123426437378, + "learning_rate": 3.0433159019144235e-08, + "loss": 3.1428, + "step": 24343 + }, + { + "epoch": 1.9516746507386102, + "grad_norm": 1.7681487798690796, + "learning_rate": 3.0332019389475165e-08, + "loss": 3.1277, + "step": 24344 + }, + { + "epoch": 1.9517548255196329, + "grad_norm": 0.6558793187141418, + "learning_rate": 3.023104784489528e-08, + "loss": 2.4614, + "step": 24345 + }, + { + "epoch": 1.9518350003006555, + "grad_norm": 1.7282109260559082, + "learning_rate": 3.0130244387105436e-08, + "loss": 3.2248, + "step": 24346 + }, + { + "epoch": 1.951915175081678, + "grad_norm": 1.8964793682098389, + "learning_rate": 3.002960901780538e-08, + "loss": 3.2036, + "step": 24347 + }, + { + "epoch": 1.9519953498627007, + "grad_norm": 1.7211869955062866, + "learning_rate": 2.9929141738693765e-08, + "loss": 3.1329, + "step": 24348 + }, + { + "epoch": 1.9520755246437234, + "grad_norm": 0.6593156456947327, + "learning_rate": 2.9828842551462565e-08, + "loss": 2.468, + "step": 24349 + }, + { + "epoch": 1.952155699424746, + "grad_norm": 1.6744297742843628, + "learning_rate": 2.9728711457802652e-08, + "loss": 3.0869, + "step": 24350 + }, + { + "epoch": 1.9522358742057686, + "grad_norm": 1.9150751829147339, + "learning_rate": 2.9628748459403777e-08, + "loss": 3.2637, + "step": 24351 + }, + { + "epoch": 1.9523160489867912, + "grad_norm": 1.6934198141098022, + "learning_rate": 2.9528953557949047e-08, + "loss": 3.0761, + "step": 24352 + }, + { + "epoch": 1.9523962237678139, + "grad_norm": 1.840463638305664, + "learning_rate": 2.9429326755123778e-08, + "loss": 3.1534, + "step": 24353 + }, + { + "epoch": 1.9524763985488365, + "grad_norm": 1.7448829412460327, + "learning_rate": 2.9329868052604405e-08, + "loss": 3.0743, + "step": 24354 + }, + { + "epoch": 1.952556573329859, + "grad_norm": 1.8398066759109497, + "learning_rate": 2.9230577452070695e-08, + "loss": 3.1845, + "step": 24355 + }, + { + "epoch": 1.9526367481108817, + "grad_norm": 1.8338009119033813, + "learning_rate": 2.913145495519576e-08, + "loss": 3.1971, + "step": 24356 + }, + { + "epoch": 1.9527169228919043, + "grad_norm": 1.7391923666000366, + "learning_rate": 2.9032500563649368e-08, + "loss": 3.1087, + "step": 24357 + }, + { + "epoch": 1.952797097672927, + "grad_norm": 1.811363697052002, + "learning_rate": 2.8933714279101298e-08, + "loss": 3.2008, + "step": 24358 + }, + { + "epoch": 1.9528772724539496, + "grad_norm": 0.6453962922096252, + "learning_rate": 2.8835096103215776e-08, + "loss": 2.4584, + "step": 24359 + }, + { + "epoch": 1.9529574472349722, + "grad_norm": 1.959132194519043, + "learning_rate": 2.8736646037658133e-08, + "loss": 3.1874, + "step": 24360 + }, + { + "epoch": 1.9530376220159948, + "grad_norm": 1.8339301347732544, + "learning_rate": 2.8638364084084823e-08, + "loss": 3.0762, + "step": 24361 + }, + { + "epoch": 1.9531177967970175, + "grad_norm": 1.7251296043395996, + "learning_rate": 2.8540250244155632e-08, + "loss": 3.1223, + "step": 24362 + }, + { + "epoch": 1.95319797157804, + "grad_norm": 1.7449747323989868, + "learning_rate": 2.8442304519522568e-08, + "loss": 3.1522, + "step": 24363 + }, + { + "epoch": 1.9532781463590627, + "grad_norm": 1.7729706764221191, + "learning_rate": 2.8344526911838754e-08, + "loss": 3.1819, + "step": 24364 + }, + { + "epoch": 1.9533583211400853, + "grad_norm": 1.7354217767715454, + "learning_rate": 2.824691742275065e-08, + "loss": 3.1521, + "step": 24365 + }, + { + "epoch": 1.953438495921108, + "grad_norm": 0.6724029183387756, + "learning_rate": 2.8149476053906943e-08, + "loss": 2.4637, + "step": 24366 + }, + { + "epoch": 1.9535186707021306, + "grad_norm": 1.8654208183288574, + "learning_rate": 2.8052202806947428e-08, + "loss": 3.1237, + "step": 24367 + }, + { + "epoch": 1.9535988454831532, + "grad_norm": 1.8178315162658691, + "learning_rate": 2.795509768351301e-08, + "loss": 3.1882, + "step": 24368 + }, + { + "epoch": 1.9536790202641758, + "grad_norm": 1.87714421749115, + "learning_rate": 2.7858160685242387e-08, + "loss": 3.1918, + "step": 24369 + }, + { + "epoch": 1.9537591950451985, + "grad_norm": 1.6946901082992554, + "learning_rate": 2.776139181376758e-08, + "loss": 3.2035, + "step": 24370 + }, + { + "epoch": 1.953839369826221, + "grad_norm": 1.7904300689697266, + "learning_rate": 2.766479107072173e-08, + "loss": 3.0815, + "step": 24371 + }, + { + "epoch": 1.9539195446072437, + "grad_norm": 1.72896146774292, + "learning_rate": 2.7568358457732423e-08, + "loss": 3.1828, + "step": 24372 + }, + { + "epoch": 1.9539997193882663, + "grad_norm": 1.888554573059082, + "learning_rate": 2.747209397642614e-08, + "loss": 3.1863, + "step": 24373 + }, + { + "epoch": 1.954079894169289, + "grad_norm": 1.7537908554077148, + "learning_rate": 2.737599762842602e-08, + "loss": 3.1837, + "step": 24374 + }, + { + "epoch": 1.9541600689503116, + "grad_norm": 1.833923101425171, + "learning_rate": 2.7280069415351886e-08, + "loss": 3.2017, + "step": 24375 + }, + { + "epoch": 1.9542402437313342, + "grad_norm": 1.8965206146240234, + "learning_rate": 2.7184309338820213e-08, + "loss": 3.1094, + "step": 24376 + }, + { + "epoch": 1.9543204185123568, + "grad_norm": 0.6592885851860046, + "learning_rate": 2.7088717400447496e-08, + "loss": 2.4678, + "step": 24377 + }, + { + "epoch": 1.9544005932933795, + "grad_norm": 1.8327559232711792, + "learning_rate": 2.699329360184355e-08, + "loss": 3.1401, + "step": 24378 + }, + { + "epoch": 1.954480768074402, + "grad_norm": 1.7584378719329834, + "learning_rate": 2.6898037944618206e-08, + "loss": 3.0962, + "step": 24379 + }, + { + "epoch": 1.9545609428554247, + "grad_norm": 1.668578863143921, + "learning_rate": 2.680295043037684e-08, + "loss": 3.0833, + "step": 24380 + }, + { + "epoch": 1.9546411176364473, + "grad_norm": 1.8306338787078857, + "learning_rate": 2.6708031060722617e-08, + "loss": 3.1915, + "step": 24381 + }, + { + "epoch": 1.95472129241747, + "grad_norm": 1.7466650009155273, + "learning_rate": 2.6613279837256478e-08, + "loss": 3.1588, + "step": 24382 + }, + { + "epoch": 1.9548014671984926, + "grad_norm": 0.6404529213905334, + "learning_rate": 2.6518696761574926e-08, + "loss": 2.4572, + "step": 24383 + }, + { + "epoch": 1.9548816419795152, + "grad_norm": 1.7830488681793213, + "learning_rate": 2.6424281835273346e-08, + "loss": 3.1769, + "step": 24384 + }, + { + "epoch": 1.9549618167605378, + "grad_norm": 1.7998465299606323, + "learning_rate": 2.6330035059943803e-08, + "loss": 3.1857, + "step": 24385 + }, + { + "epoch": 1.9550419915415604, + "grad_norm": 0.6511568427085876, + "learning_rate": 2.6235956437175025e-08, + "loss": 2.4678, + "step": 24386 + }, + { + "epoch": 1.955122166322583, + "grad_norm": 1.7319608926773071, + "learning_rate": 2.61420459685513e-08, + "loss": 3.063, + "step": 24387 + }, + { + "epoch": 1.955202341103606, + "grad_norm": 1.763156533241272, + "learning_rate": 2.6048303655659134e-08, + "loss": 3.1293, + "step": 24388 + }, + { + "epoch": 1.9552825158846285, + "grad_norm": 1.8310972452163696, + "learning_rate": 2.595472950007727e-08, + "loss": 3.1565, + "step": 24389 + }, + { + "epoch": 1.9553626906656512, + "grad_norm": 1.6494802236557007, + "learning_rate": 2.5861323503382217e-08, + "loss": 3.1829, + "step": 24390 + }, + { + "epoch": 1.9554428654466738, + "grad_norm": 1.8001320362091064, + "learning_rate": 2.576808566715161e-08, + "loss": 3.1942, + "step": 24391 + }, + { + "epoch": 1.9555230402276964, + "grad_norm": 1.7839961051940918, + "learning_rate": 2.5675015992955298e-08, + "loss": 3.1912, + "step": 24392 + }, + { + "epoch": 1.955603215008719, + "grad_norm": 1.77200186252594, + "learning_rate": 2.558211448236203e-08, + "loss": 3.1746, + "step": 24393 + }, + { + "epoch": 1.9556833897897417, + "grad_norm": 1.8462436199188232, + "learning_rate": 2.548938113694055e-08, + "loss": 3.2774, + "step": 24394 + }, + { + "epoch": 1.9557635645707643, + "grad_norm": 1.7142313718795776, + "learning_rate": 2.539681595825183e-08, + "loss": 3.1814, + "step": 24395 + }, + { + "epoch": 1.955843739351787, + "grad_norm": 1.8566943407058716, + "learning_rate": 2.5304418947856844e-08, + "loss": 3.1129, + "step": 24396 + }, + { + "epoch": 1.9559239141328095, + "grad_norm": 1.6916677951812744, + "learning_rate": 2.521219010731546e-08, + "loss": 3.0833, + "step": 24397 + }, + { + "epoch": 1.9560040889138322, + "grad_norm": 1.7810192108154297, + "learning_rate": 2.5120129438179763e-08, + "loss": 3.2193, + "step": 24398 + }, + { + "epoch": 1.9560842636948548, + "grad_norm": 1.7342554330825806, + "learning_rate": 2.502823694200296e-08, + "loss": 3.1516, + "step": 24399 + }, + { + "epoch": 1.9561644384758774, + "grad_norm": 1.819468379020691, + "learning_rate": 2.4936512620336027e-08, + "loss": 3.0426, + "step": 24400 + }, + { + "epoch": 1.9562446132569, + "grad_norm": 1.7972135543823242, + "learning_rate": 2.4844956474723292e-08, + "loss": 3.1737, + "step": 24401 + }, + { + "epoch": 1.9563247880379226, + "grad_norm": 1.8487396240234375, + "learning_rate": 2.4753568506709068e-08, + "loss": 3.2317, + "step": 24402 + }, + { + "epoch": 1.9564049628189453, + "grad_norm": 1.8465728759765625, + "learning_rate": 2.466234871783324e-08, + "loss": 3.1055, + "step": 24403 + }, + { + "epoch": 1.956485137599968, + "grad_norm": 1.7878944873809814, + "learning_rate": 2.4571297109635684e-08, + "loss": 3.2153, + "step": 24404 + }, + { + "epoch": 1.9565653123809905, + "grad_norm": 1.8217120170593262, + "learning_rate": 2.4480413683649617e-08, + "loss": 3.2428, + "step": 24405 + }, + { + "epoch": 1.9566454871620131, + "grad_norm": 1.7350536584854126, + "learning_rate": 2.438969844140826e-08, + "loss": 3.1689, + "step": 24406 + }, + { + "epoch": 1.9567256619430358, + "grad_norm": 1.7308703660964966, + "learning_rate": 2.429915138444039e-08, + "loss": 3.1454, + "step": 24407 + }, + { + "epoch": 1.9568058367240586, + "grad_norm": 1.938705563545227, + "learning_rate": 2.420877251427367e-08, + "loss": 3.1766, + "step": 24408 + }, + { + "epoch": 1.9568860115050812, + "grad_norm": 1.8380818367004395, + "learning_rate": 2.411856183243133e-08, + "loss": 3.2477, + "step": 24409 + }, + { + "epoch": 1.9569661862861039, + "grad_norm": 1.7323594093322754, + "learning_rate": 2.4028519340434375e-08, + "loss": 3.1484, + "step": 24410 + }, + { + "epoch": 1.9570463610671265, + "grad_norm": 1.8042508363723755, + "learning_rate": 2.393864503980048e-08, + "loss": 3.2988, + "step": 24411 + }, + { + "epoch": 1.957126535848149, + "grad_norm": 1.7589467763900757, + "learning_rate": 2.3848938932045095e-08, + "loss": 3.1199, + "step": 24412 + }, + { + "epoch": 1.9572067106291717, + "grad_norm": 1.7623772621154785, + "learning_rate": 2.375940101868035e-08, + "loss": 3.0895, + "step": 24413 + }, + { + "epoch": 1.9572868854101944, + "grad_norm": 1.8016427755355835, + "learning_rate": 2.3670031301216145e-08, + "loss": 3.2131, + "step": 24414 + }, + { + "epoch": 1.957367060191217, + "grad_norm": 1.7710130214691162, + "learning_rate": 2.3580829781159052e-08, + "loss": 3.0747, + "step": 24415 + }, + { + "epoch": 1.9574472349722396, + "grad_norm": 1.7811411619186401, + "learning_rate": 2.3491796460013427e-08, + "loss": 3.1376, + "step": 24416 + }, + { + "epoch": 1.9575274097532622, + "grad_norm": 1.7642717361450195, + "learning_rate": 2.3402931339280287e-08, + "loss": 3.1693, + "step": 24417 + }, + { + "epoch": 1.9576075845342849, + "grad_norm": 1.8514635562896729, + "learning_rate": 2.3314234420457326e-08, + "loss": 3.0811, + "step": 24418 + }, + { + "epoch": 1.9576877593153075, + "grad_norm": 1.88590669631958, + "learning_rate": 2.3225705705040014e-08, + "loss": 3.2273, + "step": 24419 + }, + { + "epoch": 1.95776793409633, + "grad_norm": 1.7725677490234375, + "learning_rate": 2.3137345194521598e-08, + "loss": 3.1862, + "step": 24420 + }, + { + "epoch": 1.9578481088773527, + "grad_norm": 1.7708325386047363, + "learning_rate": 2.3049152890392e-08, + "loss": 3.2201, + "step": 24421 + }, + { + "epoch": 1.9579282836583753, + "grad_norm": 1.8681997060775757, + "learning_rate": 2.29611287941367e-08, + "loss": 3.2198, + "step": 24422 + }, + { + "epoch": 1.958008458439398, + "grad_norm": 1.735673189163208, + "learning_rate": 2.2873272907240062e-08, + "loss": 3.2557, + "step": 24423 + }, + { + "epoch": 1.9580886332204206, + "grad_norm": 1.7678755521774292, + "learning_rate": 2.2785585231184236e-08, + "loss": 3.0732, + "step": 24424 + }, + { + "epoch": 1.9581688080014432, + "grad_norm": 1.7700631618499756, + "learning_rate": 2.269806576744693e-08, + "loss": 3.2665, + "step": 24425 + }, + { + "epoch": 1.9582489827824658, + "grad_norm": 0.6733623743057251, + "learning_rate": 2.261071451750474e-08, + "loss": 2.4719, + "step": 24426 + }, + { + "epoch": 1.9583291575634885, + "grad_norm": 1.7773125171661377, + "learning_rate": 2.2523531482828708e-08, + "loss": 3.1469, + "step": 24427 + }, + { + "epoch": 1.958409332344511, + "grad_norm": 1.7361427545547485, + "learning_rate": 2.2436516664889886e-08, + "loss": 3.0825, + "step": 24428 + }, + { + "epoch": 1.9584895071255337, + "grad_norm": 1.8284282684326172, + "learning_rate": 2.2349670065154872e-08, + "loss": 3.0353, + "step": 24429 + }, + { + "epoch": 1.9585696819065563, + "grad_norm": 1.7577852010726929, + "learning_rate": 2.2262991685086944e-08, + "loss": 3.1851, + "step": 24430 + }, + { + "epoch": 1.958649856687579, + "grad_norm": 1.768735647201538, + "learning_rate": 2.2176481526150484e-08, + "loss": 3.2189, + "step": 24431 + }, + { + "epoch": 1.9587300314686016, + "grad_norm": 1.7910301685333252, + "learning_rate": 2.2090139589800995e-08, + "loss": 3.1754, + "step": 24432 + }, + { + "epoch": 1.9588102062496242, + "grad_norm": 1.7015559673309326, + "learning_rate": 2.2003965877495093e-08, + "loss": 3.0654, + "step": 24433 + }, + { + "epoch": 1.9588903810306468, + "grad_norm": 1.7492948770523071, + "learning_rate": 2.1917960390684946e-08, + "loss": 3.1925, + "step": 24434 + }, + { + "epoch": 1.9589705558116695, + "grad_norm": 1.655497431755066, + "learning_rate": 2.1832123130822726e-08, + "loss": 3.1973, + "step": 24435 + }, + { + "epoch": 1.959050730592692, + "grad_norm": 1.8326410055160522, + "learning_rate": 2.174645409935394e-08, + "loss": 3.118, + "step": 24436 + }, + { + "epoch": 1.9591309053737147, + "grad_norm": 1.84154212474823, + "learning_rate": 2.1660953297722998e-08, + "loss": 3.2784, + "step": 24437 + }, + { + "epoch": 1.9592110801547373, + "grad_norm": 1.8126038312911987, + "learning_rate": 2.157562072737096e-08, + "loss": 3.1207, + "step": 24438 + }, + { + "epoch": 1.95929125493576, + "grad_norm": 1.8590062856674194, + "learning_rate": 2.14904563897389e-08, + "loss": 3.1581, + "step": 24439 + }, + { + "epoch": 1.9593714297167826, + "grad_norm": 1.724810242652893, + "learning_rate": 2.1405460286259006e-08, + "loss": 3.2023, + "step": 24440 + }, + { + "epoch": 1.9594516044978052, + "grad_norm": 1.8135900497436523, + "learning_rate": 2.1320632418367904e-08, + "loss": 3.1753, + "step": 24441 + }, + { + "epoch": 1.9595317792788278, + "grad_norm": 1.7263503074645996, + "learning_rate": 2.123597278749334e-08, + "loss": 3.1007, + "step": 24442 + }, + { + "epoch": 1.9596119540598504, + "grad_norm": 1.8401713371276855, + "learning_rate": 2.1151481395064177e-08, + "loss": 3.1844, + "step": 24443 + }, + { + "epoch": 1.959692128840873, + "grad_norm": 1.7690834999084473, + "learning_rate": 2.1067158242503716e-08, + "loss": 3.2001, + "step": 24444 + }, + { + "epoch": 1.9597723036218957, + "grad_norm": 1.74875807762146, + "learning_rate": 2.0983003331233042e-08, + "loss": 3.3075, + "step": 24445 + }, + { + "epoch": 1.9598524784029183, + "grad_norm": 1.7505298852920532, + "learning_rate": 2.0899016662673244e-08, + "loss": 3.2247, + "step": 24446 + }, + { + "epoch": 1.959932653183941, + "grad_norm": 1.7639241218566895, + "learning_rate": 2.0815198238238742e-08, + "loss": 3.1255, + "step": 24447 + }, + { + "epoch": 1.9600128279649636, + "grad_norm": 1.761742353439331, + "learning_rate": 2.0731548059342855e-08, + "loss": 3.0479, + "step": 24448 + }, + { + "epoch": 1.9600930027459862, + "grad_norm": 1.7307695150375366, + "learning_rate": 2.0648066127396672e-08, + "loss": 3.1419, + "step": 24449 + }, + { + "epoch": 1.9601731775270088, + "grad_norm": 1.7954517602920532, + "learning_rate": 2.0564752443805735e-08, + "loss": 3.124, + "step": 24450 + }, + { + "epoch": 1.9602533523080314, + "grad_norm": 1.6912503242492676, + "learning_rate": 2.04816070099767e-08, + "loss": 3.0853, + "step": 24451 + }, + { + "epoch": 1.960333527089054, + "grad_norm": 1.820337176322937, + "learning_rate": 2.039862982731067e-08, + "loss": 3.1304, + "step": 24452 + }, + { + "epoch": 1.9604137018700767, + "grad_norm": 1.6998683214187622, + "learning_rate": 2.0315820897206517e-08, + "loss": 3.1176, + "step": 24453 + }, + { + "epoch": 1.9604938766510993, + "grad_norm": 1.77277410030365, + "learning_rate": 2.023318022106091e-08, + "loss": 3.167, + "step": 24454 + }, + { + "epoch": 1.960574051432122, + "grad_norm": 1.787003993988037, + "learning_rate": 2.0150707800266067e-08, + "loss": 3.2406, + "step": 24455 + }, + { + "epoch": 1.9606542262131446, + "grad_norm": 1.812237024307251, + "learning_rate": 2.0068403636213095e-08, + "loss": 3.15, + "step": 24456 + }, + { + "epoch": 1.9607344009941672, + "grad_norm": 1.7899247407913208, + "learning_rate": 1.9986267730289778e-08, + "loss": 3.1234, + "step": 24457 + }, + { + "epoch": 1.9608145757751898, + "grad_norm": 1.7050398588180542, + "learning_rate": 1.9904300083880558e-08, + "loss": 3.1364, + "step": 24458 + }, + { + "epoch": 1.9608947505562124, + "grad_norm": 1.7574959993362427, + "learning_rate": 1.9822500698367664e-08, + "loss": 3.1663, + "step": 24459 + }, + { + "epoch": 1.960974925337235, + "grad_norm": 0.6460685133934021, + "learning_rate": 1.9740869575129994e-08, + "loss": 2.4657, + "step": 24460 + }, + { + "epoch": 1.9610551001182577, + "grad_norm": 1.7019025087356567, + "learning_rate": 1.9659406715544226e-08, + "loss": 3.1733, + "step": 24461 + }, + { + "epoch": 1.9611352748992803, + "grad_norm": 0.6817312836647034, + "learning_rate": 1.9578112120982594e-08, + "loss": 2.4818, + "step": 24462 + }, + { + "epoch": 1.961215449680303, + "grad_norm": 1.810624122619629, + "learning_rate": 1.9496985792817337e-08, + "loss": 3.2274, + "step": 24463 + }, + { + "epoch": 1.9612956244613255, + "grad_norm": 1.8050673007965088, + "learning_rate": 1.9416027732415134e-08, + "loss": 3.2029, + "step": 24464 + }, + { + "epoch": 1.9613757992423482, + "grad_norm": 1.7584383487701416, + "learning_rate": 1.933523794114045e-08, + "loss": 3.2196, + "step": 24465 + }, + { + "epoch": 1.9614559740233708, + "grad_norm": 1.8149503469467163, + "learning_rate": 1.925461642035775e-08, + "loss": 3.2675, + "step": 24466 + }, + { + "epoch": 1.9615361488043934, + "grad_norm": 1.8007334470748901, + "learning_rate": 1.917416317142373e-08, + "loss": 3.2566, + "step": 24467 + }, + { + "epoch": 1.961616323585416, + "grad_norm": 1.6873300075531006, + "learning_rate": 1.909387819569508e-08, + "loss": 3.1528, + "step": 24468 + }, + { + "epoch": 1.961696498366439, + "grad_norm": 1.8645025491714478, + "learning_rate": 1.9013761494527382e-08, + "loss": 3.1868, + "step": 24469 + }, + { + "epoch": 1.9617766731474615, + "grad_norm": 1.7901277542114258, + "learning_rate": 1.8933813069269556e-08, + "loss": 3.2014, + "step": 24470 + }, + { + "epoch": 1.9618568479284841, + "grad_norm": 1.7403796911239624, + "learning_rate": 1.8854032921269418e-08, + "loss": 3.2284, + "step": 24471 + }, + { + "epoch": 1.9619370227095068, + "grad_norm": 1.853735089302063, + "learning_rate": 1.8774421051872548e-08, + "loss": 3.1935, + "step": 24472 + }, + { + "epoch": 1.9620171974905294, + "grad_norm": 1.7351281642913818, + "learning_rate": 1.8694977462422327e-08, + "loss": 3.1104, + "step": 24473 + }, + { + "epoch": 1.962097372271552, + "grad_norm": 1.857985019683838, + "learning_rate": 1.8615702154255453e-08, + "loss": 3.1564, + "step": 24474 + }, + { + "epoch": 1.9621775470525746, + "grad_norm": 1.6640410423278809, + "learning_rate": 1.8536595128710867e-08, + "loss": 3.1249, + "step": 24475 + }, + { + "epoch": 1.9622577218335973, + "grad_norm": 1.9167141914367676, + "learning_rate": 1.8457656387121935e-08, + "loss": 3.2619, + "step": 24476 + }, + { + "epoch": 1.9623378966146199, + "grad_norm": 1.7811702489852905, + "learning_rate": 1.8378885930818712e-08, + "loss": 3.1704, + "step": 24477 + }, + { + "epoch": 1.9624180713956425, + "grad_norm": 1.7551703453063965, + "learning_rate": 1.8300283761130134e-08, + "loss": 3.1403, + "step": 24478 + }, + { + "epoch": 1.9624982461766651, + "grad_norm": 1.9196927547454834, + "learning_rate": 1.822184987937958e-08, + "loss": 3.2612, + "step": 24479 + }, + { + "epoch": 1.9625784209576878, + "grad_norm": 1.685422420501709, + "learning_rate": 1.8143584286891558e-08, + "loss": 3.0796, + "step": 24480 + }, + { + "epoch": 1.9626585957387104, + "grad_norm": 1.7599784135818481, + "learning_rate": 1.8065486984985004e-08, + "loss": 3.0295, + "step": 24481 + }, + { + "epoch": 1.962738770519733, + "grad_norm": 1.7195677757263184, + "learning_rate": 1.7987557974976645e-08, + "loss": 3.1942, + "step": 24482 + }, + { + "epoch": 1.9628189453007556, + "grad_norm": 1.686323642730713, + "learning_rate": 1.7909797258178763e-08, + "loss": 3.0923, + "step": 24483 + }, + { + "epoch": 1.9628991200817782, + "grad_norm": 1.7288328409194946, + "learning_rate": 1.7832204835904753e-08, + "loss": 3.1428, + "step": 24484 + }, + { + "epoch": 1.9629792948628009, + "grad_norm": 1.869177222251892, + "learning_rate": 1.775478070946135e-08, + "loss": 3.2012, + "step": 24485 + }, + { + "epoch": 1.9630594696438235, + "grad_norm": 1.7744104862213135, + "learning_rate": 1.7677524880154174e-08, + "loss": 3.2073, + "step": 24486 + }, + { + "epoch": 1.9631396444248461, + "grad_norm": 1.7365529537200928, + "learning_rate": 1.7600437349285516e-08, + "loss": 3.1749, + "step": 24487 + }, + { + "epoch": 1.9632198192058687, + "grad_norm": 1.7734155654907227, + "learning_rate": 1.752351811815545e-08, + "loss": 3.2219, + "step": 24488 + }, + { + "epoch": 1.9632999939868916, + "grad_norm": 1.8349974155426025, + "learning_rate": 1.744676718806071e-08, + "loss": 3.156, + "step": 24489 + }, + { + "epoch": 1.9633801687679142, + "grad_norm": 1.7521982192993164, + "learning_rate": 1.7370184560295822e-08, + "loss": 3.1745, + "step": 24490 + }, + { + "epoch": 1.9634603435489368, + "grad_norm": 1.758329153060913, + "learning_rate": 1.7293770236150865e-08, + "loss": 3.215, + "step": 24491 + }, + { + "epoch": 1.9635405183299595, + "grad_norm": 1.7677937746047974, + "learning_rate": 1.7217524216915916e-08, + "loss": 3.1564, + "step": 24492 + }, + { + "epoch": 1.963620693110982, + "grad_norm": 1.7785688638687134, + "learning_rate": 1.7141446503873283e-08, + "loss": 3.1271, + "step": 24493 + }, + { + "epoch": 1.9637008678920047, + "grad_norm": 1.8160775899887085, + "learning_rate": 1.7065537098308606e-08, + "loss": 3.2432, + "step": 24494 + }, + { + "epoch": 1.9637810426730273, + "grad_norm": 1.721838355064392, + "learning_rate": 1.698979600150086e-08, + "loss": 3.0495, + "step": 24495 + }, + { + "epoch": 1.96386121745405, + "grad_norm": 1.7215032577514648, + "learning_rate": 1.69142232147268e-08, + "loss": 3.1746, + "step": 24496 + }, + { + "epoch": 1.9639413922350726, + "grad_norm": 1.753422737121582, + "learning_rate": 1.6838818739259854e-08, + "loss": 3.2104, + "step": 24497 + }, + { + "epoch": 1.9640215670160952, + "grad_norm": 1.766076922416687, + "learning_rate": 1.676358257637234e-08, + "loss": 3.2546, + "step": 24498 + }, + { + "epoch": 1.9641017417971178, + "grad_norm": 1.8781938552856445, + "learning_rate": 1.668851472733213e-08, + "loss": 3.134, + "step": 24499 + }, + { + "epoch": 1.9641819165781405, + "grad_norm": 1.7736376523971558, + "learning_rate": 1.6613615193405986e-08, + "loss": 3.1517, + "step": 24500 + }, + { + "epoch": 1.964262091359163, + "grad_norm": 1.7459067106246948, + "learning_rate": 1.6538883975855123e-08, + "loss": 3.1655, + "step": 24501 + }, + { + "epoch": 1.9643422661401857, + "grad_norm": 0.6573824882507324, + "learning_rate": 1.646432107594076e-08, + "loss": 2.4655, + "step": 24502 + }, + { + "epoch": 1.9644224409212083, + "grad_norm": 1.722790002822876, + "learning_rate": 1.638992649491855e-08, + "loss": 3.2791, + "step": 24503 + }, + { + "epoch": 1.964502615702231, + "grad_norm": 1.8141438961029053, + "learning_rate": 1.6315700234044162e-08, + "loss": 3.1715, + "step": 24504 + }, + { + "epoch": 1.9645827904832536, + "grad_norm": 1.7304370403289795, + "learning_rate": 1.6241642294567704e-08, + "loss": 3.1861, + "step": 24505 + }, + { + "epoch": 1.9646629652642762, + "grad_norm": 1.8347899913787842, + "learning_rate": 1.616775267773929e-08, + "loss": 3.2278, + "step": 24506 + }, + { + "epoch": 1.9647431400452988, + "grad_norm": 1.8111339807510376, + "learning_rate": 1.6094031384804588e-08, + "loss": 3.1141, + "step": 24507 + }, + { + "epoch": 1.9648233148263214, + "grad_norm": 1.6864217519760132, + "learning_rate": 1.6020478417004826e-08, + "loss": 3.1647, + "step": 24508 + }, + { + "epoch": 1.964903489607344, + "grad_norm": 1.7886065244674683, + "learning_rate": 1.5947093775581235e-08, + "loss": 3.1165, + "step": 24509 + }, + { + "epoch": 1.9649836643883667, + "grad_norm": 1.7627549171447754, + "learning_rate": 1.5873877461770603e-08, + "loss": 3.1279, + "step": 24510 + }, + { + "epoch": 1.9650638391693893, + "grad_norm": 1.7181730270385742, + "learning_rate": 1.580082947680861e-08, + "loss": 3.1906, + "step": 24511 + }, + { + "epoch": 1.965144013950412, + "grad_norm": 1.7808889150619507, + "learning_rate": 1.5727949821925382e-08, + "loss": 3.1632, + "step": 24512 + }, + { + "epoch": 1.9652241887314346, + "grad_norm": 1.7348865270614624, + "learning_rate": 1.5655238498349935e-08, + "loss": 3.124, + "step": 24513 + }, + { + "epoch": 1.9653043635124572, + "grad_norm": 1.7504827976226807, + "learning_rate": 1.5582695507307954e-08, + "loss": 3.1148, + "step": 24514 + }, + { + "epoch": 1.9653845382934798, + "grad_norm": 0.6419430375099182, + "learning_rate": 1.551032085002402e-08, + "loss": 2.4625, + "step": 24515 + }, + { + "epoch": 1.9654647130745024, + "grad_norm": 0.635845422744751, + "learning_rate": 1.543811452771604e-08, + "loss": 2.4606, + "step": 24516 + }, + { + "epoch": 1.965544887855525, + "grad_norm": 1.8640625476837158, + "learning_rate": 1.536607654160305e-08, + "loss": 3.2294, + "step": 24517 + }, + { + "epoch": 1.9656250626365477, + "grad_norm": 1.777052402496338, + "learning_rate": 1.529420689289962e-08, + "loss": 3.0736, + "step": 24518 + }, + { + "epoch": 1.9657052374175703, + "grad_norm": 1.916164517402649, + "learning_rate": 1.5222505582815906e-08, + "loss": 3.2266, + "step": 24519 + }, + { + "epoch": 1.965785412198593, + "grad_norm": 1.8428637981414795, + "learning_rate": 1.5150972612562044e-08, + "loss": 3.1266, + "step": 24520 + }, + { + "epoch": 1.9658655869796156, + "grad_norm": 1.92425537109375, + "learning_rate": 1.5079607983344845e-08, + "loss": 3.2732, + "step": 24521 + }, + { + "epoch": 1.9659457617606382, + "grad_norm": 1.8379172086715698, + "learning_rate": 1.5008411696365575e-08, + "loss": 3.1609, + "step": 24522 + }, + { + "epoch": 1.9660259365416608, + "grad_norm": 1.6946091651916504, + "learning_rate": 1.4937383752825496e-08, + "loss": 3.0582, + "step": 24523 + }, + { + "epoch": 1.9661061113226834, + "grad_norm": 1.7695223093032837, + "learning_rate": 1.4866524153921425e-08, + "loss": 3.1326, + "step": 24524 + }, + { + "epoch": 1.966186286103706, + "grad_norm": 1.760617971420288, + "learning_rate": 1.4795832900849071e-08, + "loss": 3.2452, + "step": 24525 + }, + { + "epoch": 1.9662664608847287, + "grad_norm": 1.7792680263519287, + "learning_rate": 1.4725309994799708e-08, + "loss": 3.2585, + "step": 24526 + }, + { + "epoch": 1.9663466356657513, + "grad_norm": 1.763852596282959, + "learning_rate": 1.4654955436963492e-08, + "loss": 3.2236, + "step": 24527 + }, + { + "epoch": 1.966426810446774, + "grad_norm": 1.8268377780914307, + "learning_rate": 1.4584769228525031e-08, + "loss": 3.1306, + "step": 24528 + }, + { + "epoch": 1.9665069852277965, + "grad_norm": 1.7166658639907837, + "learning_rate": 1.4514751370667823e-08, + "loss": 3.1443, + "step": 24529 + }, + { + "epoch": 1.9665871600088192, + "grad_norm": 1.8466870784759521, + "learning_rate": 1.4444901864572036e-08, + "loss": 3.1805, + "step": 24530 + }, + { + "epoch": 1.9666673347898418, + "grad_norm": 1.7941782474517822, + "learning_rate": 1.4375220711416726e-08, + "loss": 3.2535, + "step": 24531 + }, + { + "epoch": 1.9667475095708644, + "grad_norm": 1.7668474912643433, + "learning_rate": 1.430570791237651e-08, + "loss": 3.2653, + "step": 24532 + }, + { + "epoch": 1.966827684351887, + "grad_norm": 0.6888736486434937, + "learning_rate": 1.4236363468621561e-08, + "loss": 2.4758, + "step": 24533 + }, + { + "epoch": 1.9669078591329097, + "grad_norm": 1.7003437280654907, + "learning_rate": 1.4167187381323166e-08, + "loss": 3.0797, + "step": 24534 + }, + { + "epoch": 1.9669880339139323, + "grad_norm": 1.7798813581466675, + "learning_rate": 1.4098179651645949e-08, + "loss": 3.2134, + "step": 24535 + }, + { + "epoch": 1.967068208694955, + "grad_norm": 1.6741409301757812, + "learning_rate": 1.4029340280754533e-08, + "loss": 3.1581, + "step": 24536 + }, + { + "epoch": 1.9671483834759775, + "grad_norm": 1.7957996129989624, + "learning_rate": 1.3960669269809101e-08, + "loss": 3.1947, + "step": 24537 + }, + { + "epoch": 1.9672285582570002, + "grad_norm": 1.7736554145812988, + "learning_rate": 1.3892166619966507e-08, + "loss": 3.1714, + "step": 24538 + }, + { + "epoch": 1.9673087330380228, + "grad_norm": 1.7824383974075317, + "learning_rate": 1.3823832332383603e-08, + "loss": 3.1399, + "step": 24539 + }, + { + "epoch": 1.9673889078190454, + "grad_norm": 0.651398241519928, + "learning_rate": 1.3755666408210577e-08, + "loss": 2.4569, + "step": 24540 + }, + { + "epoch": 1.967469082600068, + "grad_norm": 1.8160325288772583, + "learning_rate": 1.3687668848598734e-08, + "loss": 3.2279, + "step": 24541 + }, + { + "epoch": 1.9675492573810907, + "grad_norm": 1.9539916515350342, + "learning_rate": 1.3619839654692713e-08, + "loss": 3.1628, + "step": 24542 + }, + { + "epoch": 1.9676294321621133, + "grad_norm": 1.690963625907898, + "learning_rate": 1.3552178827636042e-08, + "loss": 2.9983, + "step": 24543 + }, + { + "epoch": 1.967709606943136, + "grad_norm": 1.897063970565796, + "learning_rate": 1.3484686368570032e-08, + "loss": 3.2112, + "step": 24544 + }, + { + "epoch": 1.9677897817241585, + "grad_norm": 1.7572182416915894, + "learning_rate": 1.3417362278632662e-08, + "loss": 3.1911, + "step": 24545 + }, + { + "epoch": 1.9678699565051812, + "grad_norm": 1.8905680179595947, + "learning_rate": 1.3350206558959689e-08, + "loss": 3.2842, + "step": 24546 + }, + { + "epoch": 1.9679501312862038, + "grad_norm": 1.7327978610992432, + "learning_rate": 1.328321921068243e-08, + "loss": 3.2766, + "step": 24547 + }, + { + "epoch": 1.9680303060672264, + "grad_norm": 1.7278975248336792, + "learning_rate": 1.3216400234929982e-08, + "loss": 3.1078, + "step": 24548 + }, + { + "epoch": 1.968110480848249, + "grad_norm": 1.7400870323181152, + "learning_rate": 1.3149749632829223e-08, + "loss": 3.1235, + "step": 24549 + }, + { + "epoch": 1.9681906556292719, + "grad_norm": 1.8250703811645508, + "learning_rate": 1.3083267405504807e-08, + "loss": 3.1831, + "step": 24550 + }, + { + "epoch": 1.9682708304102945, + "grad_norm": 1.8269968032836914, + "learning_rate": 1.301695355407695e-08, + "loss": 3.1199, + "step": 24551 + }, + { + "epoch": 1.9683510051913171, + "grad_norm": 1.7644473314285278, + "learning_rate": 1.2950808079662535e-08, + "loss": 3.1227, + "step": 24552 + }, + { + "epoch": 1.9684311799723397, + "grad_norm": 1.8050764799118042, + "learning_rate": 1.2884830983377339e-08, + "loss": 3.196, + "step": 24553 + }, + { + "epoch": 1.9685113547533624, + "grad_norm": 1.861245036125183, + "learning_rate": 1.2819022266334913e-08, + "loss": 3.1917, + "step": 24554 + }, + { + "epoch": 1.968591529534385, + "grad_norm": 1.6890720129013062, + "learning_rate": 1.2753381929644371e-08, + "loss": 3.137, + "step": 24555 + }, + { + "epoch": 1.9686717043154076, + "grad_norm": 1.8229079246520996, + "learning_rate": 1.2687909974411494e-08, + "loss": 3.295, + "step": 24556 + }, + { + "epoch": 1.9687518790964302, + "grad_norm": 1.775771141052246, + "learning_rate": 1.2622606401739846e-08, + "loss": 3.1318, + "step": 24557 + }, + { + "epoch": 1.9688320538774529, + "grad_norm": 0.6362921595573425, + "learning_rate": 1.2557471212731876e-08, + "loss": 2.4432, + "step": 24558 + }, + { + "epoch": 1.9689122286584755, + "grad_norm": 1.7265722751617432, + "learning_rate": 1.2492504408485595e-08, + "loss": 3.1926, + "step": 24559 + }, + { + "epoch": 1.968992403439498, + "grad_norm": 1.680120587348938, + "learning_rate": 1.2427705990094574e-08, + "loss": 3.118, + "step": 24560 + }, + { + "epoch": 1.9690725782205207, + "grad_norm": 1.8094022274017334, + "learning_rate": 1.2363075958653492e-08, + "loss": 3.2037, + "step": 24561 + }, + { + "epoch": 1.9691527530015434, + "grad_norm": 1.7837700843811035, + "learning_rate": 1.2298614315250368e-08, + "loss": 3.1493, + "step": 24562 + }, + { + "epoch": 1.969232927782566, + "grad_norm": 1.7372349500656128, + "learning_rate": 1.223432106097322e-08, + "loss": 3.1716, + "step": 24563 + }, + { + "epoch": 1.9693131025635886, + "grad_norm": 1.8141298294067383, + "learning_rate": 1.2170196196904515e-08, + "loss": 3.271, + "step": 24564 + }, + { + "epoch": 1.9693932773446112, + "grad_norm": 1.6988625526428223, + "learning_rate": 1.2106239724126722e-08, + "loss": 3.0824, + "step": 24565 + }, + { + "epoch": 1.9694734521256338, + "grad_norm": 1.7053767442703247, + "learning_rate": 1.2042451643717868e-08, + "loss": 3.1437, + "step": 24566 + }, + { + "epoch": 1.9695536269066565, + "grad_norm": 1.6724355220794678, + "learning_rate": 1.1978831956752645e-08, + "loss": 3.0743, + "step": 24567 + }, + { + "epoch": 1.969633801687679, + "grad_norm": 1.756859302520752, + "learning_rate": 1.1915380664304644e-08, + "loss": 3.1079, + "step": 24568 + }, + { + "epoch": 1.9697139764687017, + "grad_norm": 1.7385841608047485, + "learning_rate": 1.1852097767441895e-08, + "loss": 3.2363, + "step": 24569 + }, + { + "epoch": 1.9697941512497243, + "grad_norm": 1.8355540037155151, + "learning_rate": 1.1788983267234655e-08, + "loss": 3.174, + "step": 24570 + }, + { + "epoch": 1.9698743260307472, + "grad_norm": 1.7137826681137085, + "learning_rate": 1.1726037164743187e-08, + "loss": 3.1193, + "step": 24571 + }, + { + "epoch": 1.9699545008117698, + "grad_norm": 1.6969603300094604, + "learning_rate": 1.1663259461032195e-08, + "loss": 3.2438, + "step": 24572 + }, + { + "epoch": 1.9700346755927924, + "grad_norm": 1.7248047590255737, + "learning_rate": 1.16006501571575e-08, + "loss": 3.1547, + "step": 24573 + }, + { + "epoch": 1.970114850373815, + "grad_norm": 1.7173149585723877, + "learning_rate": 1.1538209254174926e-08, + "loss": 3.2625, + "step": 24574 + }, + { + "epoch": 1.9701950251548377, + "grad_norm": 1.7058571577072144, + "learning_rate": 1.1475936753139184e-08, + "loss": 3.2737, + "step": 24575 + }, + { + "epoch": 1.9702751999358603, + "grad_norm": 1.806408405303955, + "learning_rate": 1.1413832655098322e-08, + "loss": 3.3098, + "step": 24576 + }, + { + "epoch": 1.970355374716883, + "grad_norm": 1.8715459108352661, + "learning_rate": 1.1351896961099284e-08, + "loss": 3.1516, + "step": 24577 + }, + { + "epoch": 1.9704355494979056, + "grad_norm": 1.7713500261306763, + "learning_rate": 1.1290129672187899e-08, + "loss": 3.1831, + "step": 24578 + }, + { + "epoch": 1.9705157242789282, + "grad_norm": 1.725005865097046, + "learning_rate": 1.1228530789404445e-08, + "loss": 3.1316, + "step": 24579 + }, + { + "epoch": 1.9705958990599508, + "grad_norm": 1.748050570487976, + "learning_rate": 1.1167100313786982e-08, + "loss": 3.1771, + "step": 24580 + }, + { + "epoch": 1.9706760738409734, + "grad_norm": 1.7951266765594482, + "learning_rate": 1.1105838246372458e-08, + "loss": 3.1686, + "step": 24581 + }, + { + "epoch": 1.970756248621996, + "grad_norm": 1.7000075578689575, + "learning_rate": 1.1044744588192268e-08, + "loss": 3.0767, + "step": 24582 + }, + { + "epoch": 1.9708364234030187, + "grad_norm": 0.6755760908126831, + "learning_rate": 1.0983819340277813e-08, + "loss": 2.4491, + "step": 24583 + }, + { + "epoch": 1.9709165981840413, + "grad_norm": 1.7817438840866089, + "learning_rate": 1.0923062503654935e-08, + "loss": 3.1965, + "step": 24584 + }, + { + "epoch": 1.970996772965064, + "grad_norm": 1.6703217029571533, + "learning_rate": 1.0862474079349483e-08, + "loss": 3.1591, + "step": 24585 + }, + { + "epoch": 1.9710769477460865, + "grad_norm": 1.8067219257354736, + "learning_rate": 1.0802054068382861e-08, + "loss": 3.2251, + "step": 24586 + }, + { + "epoch": 1.9711571225271092, + "grad_norm": 1.7508667707443237, + "learning_rate": 1.0741802471772034e-08, + "loss": 3.1882, + "step": 24587 + }, + { + "epoch": 1.9712372973081318, + "grad_norm": 1.7052675485610962, + "learning_rate": 1.0681719290533964e-08, + "loss": 3.2153, + "step": 24588 + }, + { + "epoch": 1.9713174720891544, + "grad_norm": 1.724077582359314, + "learning_rate": 1.0621804525681178e-08, + "loss": 3.0771, + "step": 24589 + }, + { + "epoch": 1.971397646870177, + "grad_norm": 1.7204617261886597, + "learning_rate": 1.0562058178225087e-08, + "loss": 3.2074, + "step": 24590 + }, + { + "epoch": 1.9714778216511997, + "grad_norm": 1.7119390964508057, + "learning_rate": 1.0502480249171554e-08, + "loss": 3.2463, + "step": 24591 + }, + { + "epoch": 1.9715579964322223, + "grad_norm": 1.7856653928756714, + "learning_rate": 1.0443070739526439e-08, + "loss": 3.0103, + "step": 24592 + }, + { + "epoch": 1.971638171213245, + "grad_norm": 1.8042786121368408, + "learning_rate": 1.0383829650288946e-08, + "loss": 3.1548, + "step": 24593 + }, + { + "epoch": 1.9717183459942675, + "grad_norm": 1.636966347694397, + "learning_rate": 1.0324756982460494e-08, + "loss": 3.0991, + "step": 24594 + }, + { + "epoch": 1.9717985207752902, + "grad_norm": 1.8179316520690918, + "learning_rate": 1.0265852737035842e-08, + "loss": 3.2319, + "step": 24595 + }, + { + "epoch": 1.9718786955563128, + "grad_norm": 1.7328320741653442, + "learning_rate": 1.020711691500753e-08, + "loss": 3.1675, + "step": 24596 + }, + { + "epoch": 1.9719588703373354, + "grad_norm": 1.823589563369751, + "learning_rate": 1.014854951736699e-08, + "loss": 3.2322, + "step": 24597 + }, + { + "epoch": 1.972039045118358, + "grad_norm": 1.8518847227096558, + "learning_rate": 1.0090150545101207e-08, + "loss": 3.0802, + "step": 24598 + }, + { + "epoch": 1.9721192198993807, + "grad_norm": 1.8408578634262085, + "learning_rate": 1.0031919999194951e-08, + "loss": 3.1445, + "step": 24599 + }, + { + "epoch": 1.9721993946804033, + "grad_norm": 1.7415220737457275, + "learning_rate": 9.973857880628545e-09, + "loss": 3.1229, + "step": 24600 + }, + { + "epoch": 1.972279569461426, + "grad_norm": 1.7472476959228516, + "learning_rate": 9.915964190383431e-09, + "loss": 3.1486, + "step": 24601 + }, + { + "epoch": 1.9723597442424485, + "grad_norm": 1.8531025648117065, + "learning_rate": 9.858238929433273e-09, + "loss": 3.2019, + "step": 24602 + }, + { + "epoch": 1.9724399190234712, + "grad_norm": 1.7680158615112305, + "learning_rate": 9.800682098752846e-09, + "loss": 3.1212, + "step": 24603 + }, + { + "epoch": 1.9725200938044938, + "grad_norm": 1.8239611387252808, + "learning_rate": 9.743293699312483e-09, + "loss": 3.1196, + "step": 24604 + }, + { + "epoch": 1.9726002685855164, + "grad_norm": 1.8126649856567383, + "learning_rate": 9.686073732079193e-09, + "loss": 3.1825, + "step": 24605 + }, + { + "epoch": 1.972680443366539, + "grad_norm": 1.749393343925476, + "learning_rate": 9.629022198016646e-09, + "loss": 3.1961, + "step": 24606 + }, + { + "epoch": 1.9727606181475617, + "grad_norm": 1.728750467300415, + "learning_rate": 9.572139098088517e-09, + "loss": 3.1516, + "step": 24607 + }, + { + "epoch": 1.9728407929285843, + "grad_norm": 1.7314199209213257, + "learning_rate": 9.51542443325404e-09, + "loss": 3.1772, + "step": 24608 + }, + { + "epoch": 1.972920967709607, + "grad_norm": 1.7568154335021973, + "learning_rate": 9.458878204466893e-09, + "loss": 3.0773, + "step": 24609 + }, + { + "epoch": 1.9730011424906295, + "grad_norm": 1.798126459121704, + "learning_rate": 9.40250041268298e-09, + "loss": 3.1219, + "step": 24610 + }, + { + "epoch": 1.9730813172716521, + "grad_norm": 1.884649395942688, + "learning_rate": 9.346291058851541e-09, + "loss": 3.2699, + "step": 24611 + }, + { + "epoch": 1.9731614920526748, + "grad_norm": 1.816660761833191, + "learning_rate": 9.29025014392071e-09, + "loss": 3.2203, + "step": 24612 + }, + { + "epoch": 1.9732416668336974, + "grad_norm": 1.7788376808166504, + "learning_rate": 9.234377668835282e-09, + "loss": 3.0951, + "step": 24613 + }, + { + "epoch": 1.97332184161472, + "grad_norm": 1.8863940238952637, + "learning_rate": 9.178673634537839e-09, + "loss": 3.2126, + "step": 24614 + }, + { + "epoch": 1.9734020163957426, + "grad_norm": 1.7689266204833984, + "learning_rate": 9.123138041965406e-09, + "loss": 3.2566, + "step": 24615 + }, + { + "epoch": 1.9734821911767653, + "grad_norm": 1.7784148454666138, + "learning_rate": 9.067770892057237e-09, + "loss": 3.2192, + "step": 24616 + }, + { + "epoch": 1.973562365957788, + "grad_norm": 1.745189905166626, + "learning_rate": 9.012572185744806e-09, + "loss": 3.1524, + "step": 24617 + }, + { + "epoch": 1.9736425407388105, + "grad_norm": 1.798223614692688, + "learning_rate": 8.957541923958479e-09, + "loss": 3.2455, + "step": 24618 + }, + { + "epoch": 1.9737227155198331, + "grad_norm": 0.6478350758552551, + "learning_rate": 8.902680107627514e-09, + "loss": 2.4557, + "step": 24619 + }, + { + "epoch": 1.9738028903008558, + "grad_norm": 1.8160699605941772, + "learning_rate": 8.847986737675618e-09, + "loss": 3.1737, + "step": 24620 + }, + { + "epoch": 1.9738830650818784, + "grad_norm": 1.7711005210876465, + "learning_rate": 8.793461815026493e-09, + "loss": 3.19, + "step": 24621 + }, + { + "epoch": 1.973963239862901, + "grad_norm": 1.769540786743164, + "learning_rate": 8.739105340597188e-09, + "loss": 3.2791, + "step": 24622 + }, + { + "epoch": 1.9740434146439236, + "grad_norm": 1.770675778388977, + "learning_rate": 8.684917315305852e-09, + "loss": 3.1997, + "step": 24623 + }, + { + "epoch": 1.9741235894249463, + "grad_norm": 1.7016252279281616, + "learning_rate": 8.630897740065092e-09, + "loss": 3.0897, + "step": 24624 + }, + { + "epoch": 1.9742037642059689, + "grad_norm": 1.8126022815704346, + "learning_rate": 8.577046615786399e-09, + "loss": 3.1511, + "step": 24625 + }, + { + "epoch": 1.9742839389869915, + "grad_norm": 1.8056455850601196, + "learning_rate": 8.523363943377938e-09, + "loss": 3.1247, + "step": 24626 + }, + { + "epoch": 1.9743641137680141, + "grad_norm": 1.769202709197998, + "learning_rate": 8.469849723743428e-09, + "loss": 3.2146, + "step": 24627 + }, + { + "epoch": 1.9744442885490368, + "grad_norm": 1.8551362752914429, + "learning_rate": 8.416503957786592e-09, + "loss": 3.1193, + "step": 24628 + }, + { + "epoch": 1.9745244633300594, + "grad_norm": 0.6542365550994873, + "learning_rate": 8.3633266464056e-09, + "loss": 2.4531, + "step": 24629 + }, + { + "epoch": 1.974604638111082, + "grad_norm": 1.7980066537857056, + "learning_rate": 8.310317790497513e-09, + "loss": 3.0278, + "step": 24630 + }, + { + "epoch": 1.9746848128921046, + "grad_norm": 1.7309069633483887, + "learning_rate": 8.25747739095606e-09, + "loss": 3.2458, + "step": 24631 + }, + { + "epoch": 1.9747649876731275, + "grad_norm": 1.8134686946868896, + "learning_rate": 8.204805448672749e-09, + "loss": 3.1821, + "step": 24632 + }, + { + "epoch": 1.97484516245415, + "grad_norm": 1.7510288953781128, + "learning_rate": 8.152301964534647e-09, + "loss": 3.2709, + "step": 24633 + }, + { + "epoch": 1.9749253372351727, + "grad_norm": 1.7423312664031982, + "learning_rate": 8.099966939426607e-09, + "loss": 3.1385, + "step": 24634 + }, + { + "epoch": 1.9750055120161953, + "grad_norm": 1.7153170108795166, + "learning_rate": 8.047800374232362e-09, + "loss": 3.1876, + "step": 24635 + }, + { + "epoch": 1.975085686797218, + "grad_norm": 1.7590914964675903, + "learning_rate": 7.9958022698301e-09, + "loss": 3.2557, + "step": 24636 + }, + { + "epoch": 1.9751658615782406, + "grad_norm": 1.8671449422836304, + "learning_rate": 7.943972627098006e-09, + "loss": 3.1335, + "step": 24637 + }, + { + "epoch": 1.9752460363592632, + "grad_norm": 1.7482680082321167, + "learning_rate": 7.892311446907608e-09, + "loss": 3.1848, + "step": 24638 + }, + { + "epoch": 1.9753262111402858, + "grad_norm": 1.9078401327133179, + "learning_rate": 7.840818730132648e-09, + "loss": 3.1433, + "step": 24639 + }, + { + "epoch": 1.9754063859213085, + "grad_norm": 1.7398512363433838, + "learning_rate": 7.789494477639103e-09, + "loss": 3.1848, + "step": 24640 + }, + { + "epoch": 1.975486560702331, + "grad_norm": 1.7719987630844116, + "learning_rate": 7.738338690292946e-09, + "loss": 3.2578, + "step": 24641 + }, + { + "epoch": 1.9755667354833537, + "grad_norm": 1.7724608182907104, + "learning_rate": 7.687351368957929e-09, + "loss": 3.2159, + "step": 24642 + }, + { + "epoch": 1.9756469102643763, + "grad_norm": 1.8212136030197144, + "learning_rate": 7.636532514491146e-09, + "loss": 3.1925, + "step": 24643 + }, + { + "epoch": 1.975727085045399, + "grad_norm": 1.843644142150879, + "learning_rate": 7.58588212775191e-09, + "loss": 3.345, + "step": 24644 + }, + { + "epoch": 1.9758072598264216, + "grad_norm": 1.8183201551437378, + "learning_rate": 7.53540020959176e-09, + "loss": 3.2067, + "step": 24645 + }, + { + "epoch": 1.9758874346074442, + "grad_norm": 1.867790937423706, + "learning_rate": 7.485086760864457e-09, + "loss": 3.1499, + "step": 24646 + }, + { + "epoch": 1.9759676093884668, + "grad_norm": 1.7180567979812622, + "learning_rate": 7.4349417824159944e-09, + "loss": 3.1509, + "step": 24647 + }, + { + "epoch": 1.9760477841694895, + "grad_norm": 1.8711495399475098, + "learning_rate": 7.3849652750934695e-09, + "loss": 3.3042, + "step": 24648 + }, + { + "epoch": 1.976127958950512, + "grad_norm": 1.6965668201446533, + "learning_rate": 7.335157239737323e-09, + "loss": 3.1639, + "step": 24649 + }, + { + "epoch": 1.9762081337315347, + "grad_norm": 1.760794997215271, + "learning_rate": 7.2855176771902125e-09, + "loss": 3.1533, + "step": 24650 + }, + { + "epoch": 1.9762883085125573, + "grad_norm": 1.7697008848190308, + "learning_rate": 7.236046588287027e-09, + "loss": 3.1558, + "step": 24651 + }, + { + "epoch": 1.9763684832935802, + "grad_norm": 1.8081676959991455, + "learning_rate": 7.186743973861543e-09, + "loss": 3.252, + "step": 24652 + }, + { + "epoch": 1.9764486580746028, + "grad_norm": 1.8066316843032837, + "learning_rate": 7.1376098347464285e-09, + "loss": 3.1227, + "step": 24653 + }, + { + "epoch": 1.9765288328556254, + "grad_norm": 1.7803577184677124, + "learning_rate": 7.088644171768799e-09, + "loss": 3.1521, + "step": 24654 + }, + { + "epoch": 1.976609007636648, + "grad_norm": 1.695117712020874, + "learning_rate": 7.039846985755772e-09, + "loss": 3.1559, + "step": 24655 + }, + { + "epoch": 1.9766891824176707, + "grad_norm": 1.8513402938842773, + "learning_rate": 6.99121827752891e-09, + "loss": 3.1284, + "step": 24656 + }, + { + "epoch": 1.9767693571986933, + "grad_norm": 1.8305774927139282, + "learning_rate": 6.942758047907561e-09, + "loss": 3.1481, + "step": 24657 + }, + { + "epoch": 1.976849531979716, + "grad_norm": 1.720800518989563, + "learning_rate": 6.894466297709956e-09, + "loss": 3.1556, + "step": 24658 + }, + { + "epoch": 1.9769297067607385, + "grad_norm": 1.8445181846618652, + "learning_rate": 6.846343027748781e-09, + "loss": 3.2178, + "step": 24659 + }, + { + "epoch": 1.9770098815417612, + "grad_norm": 1.7579381465911865, + "learning_rate": 6.7983882388378275e-09, + "loss": 3.1346, + "step": 24660 + }, + { + "epoch": 1.9770900563227838, + "grad_norm": 1.7544397115707397, + "learning_rate": 6.750601931783118e-09, + "loss": 3.2062, + "step": 24661 + }, + { + "epoch": 1.9771702311038064, + "grad_norm": 1.705424427986145, + "learning_rate": 6.702984107391786e-09, + "loss": 3.0892, + "step": 24662 + }, + { + "epoch": 1.977250405884829, + "grad_norm": 1.828795313835144, + "learning_rate": 6.6555347664654105e-09, + "loss": 3.2616, + "step": 24663 + }, + { + "epoch": 1.9773305806658517, + "grad_norm": 1.7043770551681519, + "learning_rate": 6.608253909805573e-09, + "loss": 3.0854, + "step": 24664 + }, + { + "epoch": 1.9774107554468743, + "grad_norm": 1.710458517074585, + "learning_rate": 6.561141538208304e-09, + "loss": 3.1371, + "step": 24665 + }, + { + "epoch": 1.977490930227897, + "grad_norm": 1.9210669994354248, + "learning_rate": 6.514197652467413e-09, + "loss": 3.1571, + "step": 24666 + }, + { + "epoch": 1.9775711050089195, + "grad_norm": 1.860421895980835, + "learning_rate": 6.467422253375599e-09, + "loss": 3.1387, + "step": 24667 + }, + { + "epoch": 1.9776512797899422, + "grad_norm": 0.6445900797843933, + "learning_rate": 6.4208153417211206e-09, + "loss": 2.4668, + "step": 24668 + }, + { + "epoch": 1.9777314545709648, + "grad_norm": 1.818099021911621, + "learning_rate": 6.3743769182900154e-09, + "loss": 3.2395, + "step": 24669 + }, + { + "epoch": 1.9778116293519874, + "grad_norm": 1.7285410165786743, + "learning_rate": 6.32810698386388e-09, + "loss": 3.2351, + "step": 24670 + }, + { + "epoch": 1.97789180413301, + "grad_norm": 1.7303178310394287, + "learning_rate": 6.282005539224312e-09, + "loss": 3.1394, + "step": 24671 + }, + { + "epoch": 1.9779719789140326, + "grad_norm": 1.718916416168213, + "learning_rate": 6.236072585148467e-09, + "loss": 3.1732, + "step": 24672 + }, + { + "epoch": 1.9780521536950553, + "grad_norm": 1.796716332435608, + "learning_rate": 6.190308122410172e-09, + "loss": 3.272, + "step": 24673 + }, + { + "epoch": 1.978132328476078, + "grad_norm": 1.8151764869689941, + "learning_rate": 6.144712151779919e-09, + "loss": 3.251, + "step": 24674 + }, + { + "epoch": 1.9782125032571005, + "grad_norm": 1.7804033756256104, + "learning_rate": 6.099284674028205e-09, + "loss": 3.157, + "step": 24675 + }, + { + "epoch": 1.9782926780381231, + "grad_norm": 1.8356963396072388, + "learning_rate": 6.054025689921084e-09, + "loss": 3.1678, + "step": 24676 + }, + { + "epoch": 1.9783728528191458, + "grad_norm": 1.8577932119369507, + "learning_rate": 6.008935200220168e-09, + "loss": 3.3431, + "step": 24677 + }, + { + "epoch": 1.9784530276001684, + "grad_norm": 1.8143913745880127, + "learning_rate": 5.964013205687069e-09, + "loss": 3.2042, + "step": 24678 + }, + { + "epoch": 1.978533202381191, + "grad_norm": 1.8006181716918945, + "learning_rate": 5.919259707077851e-09, + "loss": 3.2134, + "step": 24679 + }, + { + "epoch": 1.9786133771622136, + "grad_norm": 1.731241226196289, + "learning_rate": 5.874674705147465e-09, + "loss": 3.2764, + "step": 24680 + }, + { + "epoch": 1.9786935519432363, + "grad_norm": 1.824381947517395, + "learning_rate": 5.830258200648642e-09, + "loss": 3.2231, + "step": 24681 + }, + { + "epoch": 1.9787737267242589, + "grad_norm": 1.6802500486373901, + "learning_rate": 5.7860101943285615e-09, + "loss": 3.0565, + "step": 24682 + }, + { + "epoch": 1.9788539015052815, + "grad_norm": 1.7688729763031006, + "learning_rate": 5.741930686934405e-09, + "loss": 3.0465, + "step": 24683 + }, + { + "epoch": 1.9789340762863041, + "grad_norm": 1.850131869316101, + "learning_rate": 5.69801967920891e-09, + "loss": 3.2454, + "step": 24684 + }, + { + "epoch": 1.9790142510673268, + "grad_norm": 1.8306931257247925, + "learning_rate": 5.654277171892597e-09, + "loss": 3.1763, + "step": 24685 + }, + { + "epoch": 1.9790944258483494, + "grad_norm": 1.6961766481399536, + "learning_rate": 5.610703165722653e-09, + "loss": 3.1746, + "step": 24686 + }, + { + "epoch": 1.979174600629372, + "grad_norm": 1.8102748394012451, + "learning_rate": 5.567297661434046e-09, + "loss": 3.1659, + "step": 24687 + }, + { + "epoch": 1.9792547754103946, + "grad_norm": 0.6488206386566162, + "learning_rate": 5.5240606597584125e-09, + "loss": 2.4663, + "step": 24688 + }, + { + "epoch": 1.9793349501914173, + "grad_norm": 1.7211530208587646, + "learning_rate": 5.4809921614251696e-09, + "loss": 3.2029, + "step": 24689 + }, + { + "epoch": 1.9794151249724399, + "grad_norm": 1.8038712739944458, + "learning_rate": 5.438092167159292e-09, + "loss": 3.1556, + "step": 24690 + }, + { + "epoch": 1.9794952997534625, + "grad_norm": 1.7567390203475952, + "learning_rate": 5.395360677684647e-09, + "loss": 3.1454, + "step": 24691 + }, + { + "epoch": 1.9795754745344851, + "grad_norm": 1.9073151350021362, + "learning_rate": 5.352797693722878e-09, + "loss": 3.2706, + "step": 24692 + }, + { + "epoch": 1.9796556493155077, + "grad_norm": 1.785007357597351, + "learning_rate": 5.310403215988968e-09, + "loss": 3.144, + "step": 24693 + }, + { + "epoch": 1.9797358240965304, + "grad_norm": 1.8453810214996338, + "learning_rate": 5.268177245200123e-09, + "loss": 3.0964, + "step": 24694 + }, + { + "epoch": 1.979815998877553, + "grad_norm": 1.7662806510925293, + "learning_rate": 5.226119782066885e-09, + "loss": 3.1977, + "step": 24695 + }, + { + "epoch": 1.9798961736585756, + "grad_norm": 1.7117750644683838, + "learning_rate": 5.184230827299797e-09, + "loss": 3.145, + "step": 24696 + }, + { + "epoch": 1.9799763484395982, + "grad_norm": 1.8314459323883057, + "learning_rate": 5.142510381602739e-09, + "loss": 3.1911, + "step": 24697 + }, + { + "epoch": 1.9800565232206209, + "grad_norm": 1.7035496234893799, + "learning_rate": 5.100958445680704e-09, + "loss": 3.0987, + "step": 24698 + }, + { + "epoch": 1.9801366980016435, + "grad_norm": 1.8997818231582642, + "learning_rate": 5.059575020234242e-09, + "loss": 3.218, + "step": 24699 + }, + { + "epoch": 1.9802168727826661, + "grad_norm": 1.7482376098632812, + "learning_rate": 5.018360105960574e-09, + "loss": 3.236, + "step": 24700 + }, + { + "epoch": 1.9802970475636887, + "grad_norm": 1.8779760599136353, + "learning_rate": 4.977313703555808e-09, + "loss": 3.1058, + "step": 24701 + }, + { + "epoch": 1.9803772223447114, + "grad_norm": 1.7336846590042114, + "learning_rate": 4.936435813709395e-09, + "loss": 3.0789, + "step": 24702 + }, + { + "epoch": 1.980457397125734, + "grad_norm": 1.7736397981643677, + "learning_rate": 4.895726437113002e-09, + "loss": 3.1302, + "step": 24703 + }, + { + "epoch": 1.9805375719067566, + "grad_norm": 1.7979294061660767, + "learning_rate": 4.8551855744516375e-09, + "loss": 3.1537, + "step": 24704 + }, + { + "epoch": 1.9806177466877792, + "grad_norm": 1.784913182258606, + "learning_rate": 4.814813226409198e-09, + "loss": 3.2343, + "step": 24705 + }, + { + "epoch": 1.9806979214688019, + "grad_norm": 1.7929214239120483, + "learning_rate": 4.774609393667362e-09, + "loss": 3.2172, + "step": 24706 + }, + { + "epoch": 1.9807780962498245, + "grad_norm": 1.7363612651824951, + "learning_rate": 4.734574076902254e-09, + "loss": 3.1294, + "step": 24707 + }, + { + "epoch": 1.980858271030847, + "grad_norm": 0.6572084426879883, + "learning_rate": 4.69470727678889e-09, + "loss": 2.4677, + "step": 24708 + }, + { + "epoch": 1.9809384458118697, + "grad_norm": 0.622192919254303, + "learning_rate": 4.655008994001176e-09, + "loss": 2.4662, + "step": 24709 + }, + { + "epoch": 1.9810186205928924, + "grad_norm": 1.8489524126052856, + "learning_rate": 4.6154792292063545e-09, + "loss": 3.1744, + "step": 24710 + }, + { + "epoch": 1.981098795373915, + "grad_norm": 1.6919127702713013, + "learning_rate": 4.5761179830727806e-09, + "loss": 3.1042, + "step": 24711 + }, + { + "epoch": 1.9811789701549376, + "grad_norm": 1.6784043312072754, + "learning_rate": 4.536925256262148e-09, + "loss": 3.0685, + "step": 24712 + }, + { + "epoch": 1.9812591449359604, + "grad_norm": 1.6652143001556396, + "learning_rate": 4.497901049436149e-09, + "loss": 3.0491, + "step": 24713 + }, + { + "epoch": 1.981339319716983, + "grad_norm": 1.7519854307174683, + "learning_rate": 4.459045363253145e-09, + "loss": 3.2144, + "step": 24714 + }, + { + "epoch": 1.9814194944980057, + "grad_norm": 1.745728850364685, + "learning_rate": 4.420358198368169e-09, + "loss": 3.1546, + "step": 24715 + }, + { + "epoch": 1.9814996692790283, + "grad_norm": 1.700985074043274, + "learning_rate": 4.381839555434031e-09, + "loss": 3.1824, + "step": 24716 + }, + { + "epoch": 1.981579844060051, + "grad_norm": 1.8376463651657104, + "learning_rate": 4.3434894350979915e-09, + "loss": 3.2145, + "step": 24717 + }, + { + "epoch": 1.9816600188410736, + "grad_norm": 1.8669776916503906, + "learning_rate": 4.3053078380084214e-09, + "loss": 3.1331, + "step": 24718 + }, + { + "epoch": 1.9817401936220962, + "grad_norm": 1.7701044082641602, + "learning_rate": 4.2672947648081384e-09, + "loss": 3.1129, + "step": 24719 + }, + { + "epoch": 1.9818203684031188, + "grad_norm": 1.7033206224441528, + "learning_rate": 4.229450216137743e-09, + "loss": 3.0769, + "step": 24720 + }, + { + "epoch": 1.9819005431841414, + "grad_norm": 1.8552008867263794, + "learning_rate": 4.1917741926378316e-09, + "loss": 3.1921, + "step": 24721 + }, + { + "epoch": 1.981980717965164, + "grad_norm": 0.6527384519577026, + "learning_rate": 4.154266694940123e-09, + "loss": 2.4482, + "step": 24722 + }, + { + "epoch": 1.9820608927461867, + "grad_norm": 1.6301534175872803, + "learning_rate": 4.1169277236796645e-09, + "loss": 3.2349, + "step": 24723 + }, + { + "epoch": 1.9821410675272093, + "grad_norm": 1.818979263305664, + "learning_rate": 4.079757279483732e-09, + "loss": 3.2474, + "step": 24724 + }, + { + "epoch": 1.982221242308232, + "grad_norm": 1.8576680421829224, + "learning_rate": 4.04275536298071e-09, + "loss": 3.2178, + "step": 24725 + }, + { + "epoch": 1.9823014170892546, + "grad_norm": 1.6904319524765015, + "learning_rate": 4.0059219747945465e-09, + "loss": 3.1162, + "step": 24726 + }, + { + "epoch": 1.9823815918702772, + "grad_norm": 1.957795262336731, + "learning_rate": 3.9692571155447444e-09, + "loss": 3.2031, + "step": 24727 + }, + { + "epoch": 1.9824617666512998, + "grad_norm": 1.8453842401504517, + "learning_rate": 3.932760785850809e-09, + "loss": 3.2364, + "step": 24728 + }, + { + "epoch": 1.9825419414323224, + "grad_norm": 1.7253497838974, + "learning_rate": 3.896432986327803e-09, + "loss": 3.1328, + "step": 24729 + }, + { + "epoch": 1.982622116213345, + "grad_norm": 1.7084461450576782, + "learning_rate": 3.8602737175874595e-09, + "loss": 3.103, + "step": 24730 + }, + { + "epoch": 1.9827022909943677, + "grad_norm": 1.6790302991867065, + "learning_rate": 3.8242829802404014e-09, + "loss": 3.0449, + "step": 24731 + }, + { + "epoch": 1.9827824657753903, + "grad_norm": 1.900712490081787, + "learning_rate": 3.7884607748928104e-09, + "loss": 3.2928, + "step": 24732 + }, + { + "epoch": 1.982862640556413, + "grad_norm": 1.7427096366882324, + "learning_rate": 3.7528071021486474e-09, + "loss": 3.1498, + "step": 24733 + }, + { + "epoch": 1.9829428153374358, + "grad_norm": 1.794679045677185, + "learning_rate": 3.7173219626085444e-09, + "loss": 3.2201, + "step": 24734 + }, + { + "epoch": 1.9830229901184584, + "grad_norm": 1.7904521226882935, + "learning_rate": 3.6820053568720204e-09, + "loss": 3.1473, + "step": 24735 + }, + { + "epoch": 1.983103164899481, + "grad_norm": 1.667737364768982, + "learning_rate": 3.6468572855341554e-09, + "loss": 3.1354, + "step": 24736 + }, + { + "epoch": 1.9831833396805036, + "grad_norm": 1.6613402366638184, + "learning_rate": 3.6118777491866985e-09, + "loss": 3.106, + "step": 24737 + }, + { + "epoch": 1.9832635144615263, + "grad_norm": 1.859390139579773, + "learning_rate": 3.5770667484202882e-09, + "loss": 3.1849, + "step": 24738 + }, + { + "epoch": 1.983343689242549, + "grad_norm": 1.7377389669418335, + "learning_rate": 3.542424283821122e-09, + "loss": 3.114, + "step": 24739 + }, + { + "epoch": 1.9834238640235715, + "grad_norm": 1.8414949178695679, + "learning_rate": 3.5079503559731777e-09, + "loss": 3.2553, + "step": 24740 + }, + { + "epoch": 1.9835040388045941, + "grad_norm": 1.7532025575637817, + "learning_rate": 3.473644965459322e-09, + "loss": 3.1824, + "step": 24741 + }, + { + "epoch": 1.9835842135856168, + "grad_norm": 1.71692955493927, + "learning_rate": 3.4395081128557607e-09, + "loss": 3.0834, + "step": 24742 + }, + { + "epoch": 1.9836643883666394, + "grad_norm": 1.8697725534439087, + "learning_rate": 3.4055397987386994e-09, + "loss": 3.2061, + "step": 24743 + }, + { + "epoch": 1.983744563147662, + "grad_norm": 1.8652006387710571, + "learning_rate": 3.371740023682124e-09, + "loss": 3.0477, + "step": 24744 + }, + { + "epoch": 1.9838247379286846, + "grad_norm": 0.6596006155014038, + "learning_rate": 3.338108788254468e-09, + "loss": 2.457, + "step": 24745 + }, + { + "epoch": 1.9839049127097073, + "grad_norm": 1.8067469596862793, + "learning_rate": 3.3046460930230564e-09, + "loss": 3.1597, + "step": 24746 + }, + { + "epoch": 1.9839850874907299, + "grad_norm": 1.8388935327529907, + "learning_rate": 3.271351938551881e-09, + "loss": 3.165, + "step": 24747 + }, + { + "epoch": 1.9840652622717525, + "grad_norm": 1.7626447677612305, + "learning_rate": 3.2382263254027157e-09, + "loss": 3.2056, + "step": 24748 + }, + { + "epoch": 1.9841454370527751, + "grad_norm": 0.6662325263023376, + "learning_rate": 3.2052692541328923e-09, + "loss": 2.4575, + "step": 24749 + }, + { + "epoch": 1.9842256118337978, + "grad_norm": 1.751651406288147, + "learning_rate": 3.1724807253008526e-09, + "loss": 3.1412, + "step": 24750 + }, + { + "epoch": 1.9843057866148204, + "grad_norm": 0.6541575193405151, + "learning_rate": 3.1398607394561576e-09, + "loss": 2.4531, + "step": 24751 + }, + { + "epoch": 1.984385961395843, + "grad_norm": 1.7949950695037842, + "learning_rate": 3.1074092971505877e-09, + "loss": 3.111, + "step": 24752 + }, + { + "epoch": 1.9844661361768656, + "grad_norm": 1.7424273490905762, + "learning_rate": 3.075126398931483e-09, + "loss": 3.1239, + "step": 24753 + }, + { + "epoch": 1.9845463109578882, + "grad_norm": 1.712903380393982, + "learning_rate": 3.0430120453417424e-09, + "loss": 3.1443, + "step": 24754 + }, + { + "epoch": 1.9846264857389109, + "grad_norm": 1.8636603355407715, + "learning_rate": 3.0110662369231547e-09, + "loss": 3.1152, + "step": 24755 + }, + { + "epoch": 1.9847066605199335, + "grad_norm": 1.8551089763641357, + "learning_rate": 2.9792889742152885e-09, + "loss": 3.1322, + "step": 24756 + }, + { + "epoch": 1.9847868353009561, + "grad_norm": 1.818830966949463, + "learning_rate": 2.9476802577532714e-09, + "loss": 3.1896, + "step": 24757 + }, + { + "epoch": 1.9848670100819787, + "grad_norm": 1.773939847946167, + "learning_rate": 2.9162400880700102e-09, + "loss": 3.2469, + "step": 24758 + }, + { + "epoch": 1.9849471848630014, + "grad_norm": 1.6932862997055054, + "learning_rate": 2.884968465696192e-09, + "loss": 3.1278, + "step": 24759 + }, + { + "epoch": 1.985027359644024, + "grad_norm": 0.6433948874473572, + "learning_rate": 2.8538653911580617e-09, + "loss": 2.4595, + "step": 24760 + }, + { + "epoch": 1.9851075344250466, + "grad_norm": 1.8127187490463257, + "learning_rate": 2.822930864980755e-09, + "loss": 3.1857, + "step": 24761 + }, + { + "epoch": 1.9851877092060692, + "grad_norm": 1.9216411113739014, + "learning_rate": 2.7921648876849672e-09, + "loss": 3.1982, + "step": 24762 + }, + { + "epoch": 1.9852678839870919, + "grad_norm": 1.8316309452056885, + "learning_rate": 2.761567459790282e-09, + "loss": 3.1159, + "step": 24763 + }, + { + "epoch": 1.9853480587681145, + "grad_norm": 1.7583543062210083, + "learning_rate": 2.7311385818118428e-09, + "loss": 3.168, + "step": 24764 + }, + { + "epoch": 1.9854282335491371, + "grad_norm": 1.743650197982788, + "learning_rate": 2.7008782542625735e-09, + "loss": 3.1326, + "step": 24765 + }, + { + "epoch": 1.9855084083301597, + "grad_norm": 1.9078660011291504, + "learning_rate": 2.670786477654286e-09, + "loss": 3.2721, + "step": 24766 + }, + { + "epoch": 1.9855885831111824, + "grad_norm": 1.899038553237915, + "learning_rate": 2.640863252492132e-09, + "loss": 3.2584, + "step": 24767 + }, + { + "epoch": 1.985668757892205, + "grad_norm": 1.8141769170761108, + "learning_rate": 2.6111085792823732e-09, + "loss": 3.1892, + "step": 24768 + }, + { + "epoch": 1.9857489326732276, + "grad_norm": 1.7147458791732788, + "learning_rate": 2.5815224585257203e-09, + "loss": 3.2637, + "step": 24769 + }, + { + "epoch": 1.9858291074542502, + "grad_norm": 1.7104157209396362, + "learning_rate": 2.5521048907206637e-09, + "loss": 3.1677, + "step": 24770 + }, + { + "epoch": 1.9859092822352729, + "grad_norm": 1.8319036960601807, + "learning_rate": 2.5228558763634725e-09, + "loss": 3.2913, + "step": 24771 + }, + { + "epoch": 1.9859894570162955, + "grad_norm": 1.7869497537612915, + "learning_rate": 2.4937754159481963e-09, + "loss": 3.113, + "step": 24772 + }, + { + "epoch": 1.986069631797318, + "grad_norm": 1.797971487045288, + "learning_rate": 2.4648635099644435e-09, + "loss": 3.2057, + "step": 24773 + }, + { + "epoch": 1.9861498065783407, + "grad_norm": 1.8137273788452148, + "learning_rate": 2.436120158899602e-09, + "loss": 3.1522, + "step": 24774 + }, + { + "epoch": 1.9862299813593634, + "grad_norm": 1.7586594820022583, + "learning_rate": 2.407545363237729e-09, + "loss": 3.0895, + "step": 24775 + }, + { + "epoch": 1.986310156140386, + "grad_norm": 1.8371117115020752, + "learning_rate": 2.379139123461771e-09, + "loss": 3.2175, + "step": 24776 + }, + { + "epoch": 1.9863903309214086, + "grad_norm": 1.631379246711731, + "learning_rate": 2.350901440050235e-09, + "loss": 3.1424, + "step": 24777 + }, + { + "epoch": 1.9864705057024312, + "grad_norm": 1.7399442195892334, + "learning_rate": 2.322832313478296e-09, + "loss": 3.0287, + "step": 24778 + }, + { + "epoch": 1.9865506804834538, + "grad_norm": 1.74868905544281, + "learning_rate": 2.294931744220019e-09, + "loss": 3.1533, + "step": 24779 + }, + { + "epoch": 1.9866308552644765, + "grad_norm": 1.8529448509216309, + "learning_rate": 2.267199732746139e-09, + "loss": 3.2017, + "step": 24780 + }, + { + "epoch": 1.986711030045499, + "grad_norm": 1.74270761013031, + "learning_rate": 2.23963627952295e-09, + "loss": 3.1639, + "step": 24781 + }, + { + "epoch": 1.9867912048265217, + "grad_norm": 1.7156598567962646, + "learning_rate": 2.2122413850167447e-09, + "loss": 3.1161, + "step": 24782 + }, + { + "epoch": 1.9868713796075443, + "grad_norm": 1.8755325078964233, + "learning_rate": 2.1850150496871558e-09, + "loss": 3.2006, + "step": 24783 + }, + { + "epoch": 1.986951554388567, + "grad_norm": 1.7867971658706665, + "learning_rate": 2.1579572739960362e-09, + "loss": 3.1658, + "step": 24784 + }, + { + "epoch": 1.9870317291695896, + "grad_norm": 1.7625253200531006, + "learning_rate": 2.1310680583974676e-09, + "loss": 3.2089, + "step": 24785 + }, + { + "epoch": 1.9871119039506122, + "grad_norm": 1.8210418224334717, + "learning_rate": 2.1043474033466405e-09, + "loss": 3.1675, + "step": 24786 + }, + { + "epoch": 1.9871920787316348, + "grad_norm": 1.7234381437301636, + "learning_rate": 2.0777953092920856e-09, + "loss": 3.1077, + "step": 24787 + }, + { + "epoch": 1.9872722535126575, + "grad_norm": 1.7822455167770386, + "learning_rate": 2.0514117766823325e-09, + "loss": 3.2692, + "step": 24788 + }, + { + "epoch": 1.98735242829368, + "grad_norm": 1.8862595558166504, + "learning_rate": 2.0251968059614715e-09, + "loss": 3.2139, + "step": 24789 + }, + { + "epoch": 1.9874326030747027, + "grad_norm": 1.6611906290054321, + "learning_rate": 1.9991503975735904e-09, + "loss": 3.115, + "step": 24790 + }, + { + "epoch": 1.9875127778557253, + "grad_norm": 1.889707326889038, + "learning_rate": 1.9732725519550076e-09, + "loss": 3.2595, + "step": 24791 + }, + { + "epoch": 1.987592952636748, + "grad_norm": 1.7309551239013672, + "learning_rate": 1.9475632695442613e-09, + "loss": 3.1906, + "step": 24792 + }, + { + "epoch": 1.9876731274177706, + "grad_norm": 1.7482523918151855, + "learning_rate": 1.922022550773228e-09, + "loss": 3.1448, + "step": 24793 + }, + { + "epoch": 1.9877533021987932, + "grad_norm": 1.9137202501296997, + "learning_rate": 1.896650396073785e-09, + "loss": 3.2166, + "step": 24794 + }, + { + "epoch": 1.987833476979816, + "grad_norm": 1.7795296907424927, + "learning_rate": 1.871446805873367e-09, + "loss": 3.0944, + "step": 24795 + }, + { + "epoch": 1.9879136517608387, + "grad_norm": 1.9515585899353027, + "learning_rate": 1.8464117805960801e-09, + "loss": 3.1051, + "step": 24796 + }, + { + "epoch": 1.9879938265418613, + "grad_norm": 1.8056708574295044, + "learning_rate": 1.8215453206649192e-09, + "loss": 3.1101, + "step": 24797 + }, + { + "epoch": 1.988074001322884, + "grad_norm": 1.7936739921569824, + "learning_rate": 1.7968474264995484e-09, + "loss": 3.2095, + "step": 24798 + }, + { + "epoch": 1.9881541761039065, + "grad_norm": 1.7489900588989258, + "learning_rate": 1.7723180985140808e-09, + "loss": 3.1189, + "step": 24799 + }, + { + "epoch": 1.9882343508849292, + "grad_norm": 1.8616975545883179, + "learning_rate": 1.7479573371248504e-09, + "loss": 3.3213, + "step": 24800 + }, + { + "epoch": 1.9883145256659518, + "grad_norm": 1.7894539833068848, + "learning_rate": 1.723765142740419e-09, + "loss": 3.2516, + "step": 24801 + }, + { + "epoch": 1.9883947004469744, + "grad_norm": 1.8518165349960327, + "learning_rate": 1.699741515770459e-09, + "loss": 3.214, + "step": 24802 + }, + { + "epoch": 1.988474875227997, + "grad_norm": 1.7861518859863281, + "learning_rate": 1.6758864566190914e-09, + "loss": 3.2419, + "step": 24803 + }, + { + "epoch": 1.9885550500090197, + "grad_norm": 1.7922688722610474, + "learning_rate": 1.6521999656871068e-09, + "loss": 3.2142, + "step": 24804 + }, + { + "epoch": 1.9886352247900423, + "grad_norm": 1.7380486726760864, + "learning_rate": 1.628682043376406e-09, + "loss": 3.1311, + "step": 24805 + }, + { + "epoch": 1.988715399571065, + "grad_norm": 1.7718490362167358, + "learning_rate": 1.6053326900822286e-09, + "loss": 3.1736, + "step": 24806 + }, + { + "epoch": 1.9887955743520875, + "grad_norm": 1.8686027526855469, + "learning_rate": 1.5821519061987035e-09, + "loss": 3.2029, + "step": 24807 + }, + { + "epoch": 1.9888757491331102, + "grad_norm": 1.8001545667648315, + "learning_rate": 1.5591396921166292e-09, + "loss": 3.1282, + "step": 24808 + }, + { + "epoch": 1.9889559239141328, + "grad_norm": 1.9018948078155518, + "learning_rate": 1.5362960482223632e-09, + "loss": 3.1448, + "step": 24809 + }, + { + "epoch": 1.9890360986951554, + "grad_norm": 1.6962271928787231, + "learning_rate": 1.5136209749033736e-09, + "loss": 3.1843, + "step": 24810 + }, + { + "epoch": 1.989116273476178, + "grad_norm": 1.7894213199615479, + "learning_rate": 1.491114472540467e-09, + "loss": 3.2167, + "step": 24811 + }, + { + "epoch": 1.9891964482572007, + "grad_norm": 1.700587511062622, + "learning_rate": 1.4687765415133392e-09, + "loss": 3.1917, + "step": 24812 + }, + { + "epoch": 1.9892766230382233, + "grad_norm": 1.7772653102874756, + "learning_rate": 1.4466071821994665e-09, + "loss": 3.1736, + "step": 24813 + }, + { + "epoch": 1.989356797819246, + "grad_norm": 1.7836076021194458, + "learning_rate": 1.4246063949718836e-09, + "loss": 3.2362, + "step": 24814 + }, + { + "epoch": 1.9894369726002687, + "grad_norm": 1.7701447010040283, + "learning_rate": 1.4027741802002947e-09, + "loss": 3.1609, + "step": 24815 + }, + { + "epoch": 1.9895171473812914, + "grad_norm": 1.7208198308944702, + "learning_rate": 1.3811105382555145e-09, + "loss": 3.1384, + "step": 24816 + }, + { + "epoch": 1.989597322162314, + "grad_norm": 1.844435214996338, + "learning_rate": 1.359615469500586e-09, + "loss": 3.2678, + "step": 24817 + }, + { + "epoch": 1.9896774969433366, + "grad_norm": 1.7595278024673462, + "learning_rate": 1.3382889742996618e-09, + "loss": 3.2299, + "step": 24818 + }, + { + "epoch": 1.9897576717243592, + "grad_norm": 1.7354549169540405, + "learning_rate": 1.3171310530102344e-09, + "loss": 3.1645, + "step": 24819 + }, + { + "epoch": 1.9898378465053819, + "grad_norm": 1.8940708637237549, + "learning_rate": 1.2961417059909054e-09, + "loss": 3.1908, + "step": 24820 + }, + { + "epoch": 1.9899180212864045, + "grad_norm": 1.8151209354400635, + "learning_rate": 1.2753209335947258e-09, + "loss": 3.1691, + "step": 24821 + }, + { + "epoch": 1.9899981960674271, + "grad_norm": 1.6827855110168457, + "learning_rate": 1.2546687361725263e-09, + "loss": 3.1666, + "step": 24822 + }, + { + "epoch": 1.9900783708484497, + "grad_norm": 1.7501404285430908, + "learning_rate": 1.2341851140718065e-09, + "loss": 3.1451, + "step": 24823 + }, + { + "epoch": 1.9901585456294724, + "grad_norm": 1.8886170387268066, + "learning_rate": 1.213870067640066e-09, + "loss": 3.1864, + "step": 24824 + }, + { + "epoch": 1.990238720410495, + "grad_norm": 1.6884181499481201, + "learning_rate": 1.1937235972181438e-09, + "loss": 3.0942, + "step": 24825 + }, + { + "epoch": 1.9903188951915176, + "grad_norm": 1.8190721273422241, + "learning_rate": 1.1737457031457677e-09, + "loss": 3.118, + "step": 24826 + }, + { + "epoch": 1.9903990699725402, + "grad_norm": 1.7589669227600098, + "learning_rate": 1.1539363857604457e-09, + "loss": 3.1718, + "step": 24827 + }, + { + "epoch": 1.9904792447535629, + "grad_norm": 1.7767784595489502, + "learning_rate": 1.134295645396355e-09, + "loss": 3.1121, + "step": 24828 + }, + { + "epoch": 1.9905594195345855, + "grad_norm": 1.729306697845459, + "learning_rate": 1.1148234823832315e-09, + "loss": 3.1205, + "step": 24829 + }, + { + "epoch": 1.990639594315608, + "grad_norm": 1.7597240209579468, + "learning_rate": 1.095519897050812e-09, + "loss": 3.2172, + "step": 24830 + }, + { + "epoch": 1.9907197690966307, + "grad_norm": 1.7652257680892944, + "learning_rate": 1.076384889723281e-09, + "loss": 3.1309, + "step": 24831 + }, + { + "epoch": 1.9907999438776534, + "grad_norm": 1.8120090961456299, + "learning_rate": 1.0574184607248239e-09, + "loss": 3.2363, + "step": 24832 + }, + { + "epoch": 1.990880118658676, + "grad_norm": 1.7240180969238281, + "learning_rate": 1.0386206103740748e-09, + "loss": 3.199, + "step": 24833 + }, + { + "epoch": 1.9909602934396986, + "grad_norm": 1.6915631294250488, + "learning_rate": 1.0199913389874473e-09, + "loss": 3.2114, + "step": 24834 + }, + { + "epoch": 1.9910404682207212, + "grad_norm": 0.6498587131500244, + "learning_rate": 1.0015306468802444e-09, + "loss": 2.4711, + "step": 24835 + }, + { + "epoch": 1.9911206430017439, + "grad_norm": 1.7358354330062866, + "learning_rate": 9.832385343633288e-10, + "loss": 3.2712, + "step": 24836 + }, + { + "epoch": 1.9912008177827665, + "grad_norm": 1.6873834133148193, + "learning_rate": 9.651150017453425e-10, + "loss": 3.1995, + "step": 24837 + }, + { + "epoch": 1.991280992563789, + "grad_norm": 1.6801282167434692, + "learning_rate": 9.471600493315968e-10, + "loss": 3.1971, + "step": 24838 + }, + { + "epoch": 1.9913611673448117, + "grad_norm": 1.8844302892684937, + "learning_rate": 9.293736774240724e-10, + "loss": 3.1963, + "step": 24839 + }, + { + "epoch": 1.9914413421258343, + "grad_norm": 1.6768680810928345, + "learning_rate": 9.117558863236398e-10, + "loss": 3.1399, + "step": 24840 + }, + { + "epoch": 1.991521516906857, + "grad_norm": 0.6751804947853088, + "learning_rate": 8.943066763267283e-10, + "loss": 2.4824, + "step": 24841 + }, + { + "epoch": 1.9916016916878796, + "grad_norm": 1.8120667934417725, + "learning_rate": 8.770260477275472e-10, + "loss": 3.1707, + "step": 24842 + }, + { + "epoch": 1.9916818664689022, + "grad_norm": 1.8771910667419434, + "learning_rate": 8.599140008180851e-10, + "loss": 3.1341, + "step": 24843 + }, + { + "epoch": 1.9917620412499248, + "grad_norm": 1.7648159265518188, + "learning_rate": 8.42970535887e-10, + "loss": 3.2194, + "step": 24844 + }, + { + "epoch": 1.9918422160309475, + "grad_norm": 1.733551263809204, + "learning_rate": 8.26195653218509e-10, + "loss": 3.2535, + "step": 24845 + }, + { + "epoch": 1.99192239081197, + "grad_norm": 1.7378677129745483, + "learning_rate": 8.095893530968291e-10, + "loss": 3.0991, + "step": 24846 + }, + { + "epoch": 1.9920025655929927, + "grad_norm": 1.742758870124817, + "learning_rate": 7.931516358006264e-10, + "loss": 3.147, + "step": 24847 + }, + { + "epoch": 1.9920827403740153, + "grad_norm": 0.6502209305763245, + "learning_rate": 7.768825016085668e-10, + "loss": 2.4652, + "step": 24848 + }, + { + "epoch": 1.992162915155038, + "grad_norm": 1.690626859664917, + "learning_rate": 7.607819507937652e-10, + "loss": 3.1263, + "step": 24849 + }, + { + "epoch": 1.9922430899360606, + "grad_norm": 1.8022383451461792, + "learning_rate": 7.448499836293366e-10, + "loss": 3.1881, + "step": 24850 + }, + { + "epoch": 1.9923232647170832, + "grad_norm": 1.763091802597046, + "learning_rate": 7.29086600380624e-10, + "loss": 3.1038, + "step": 24851 + }, + { + "epoch": 1.9924034394981058, + "grad_norm": 1.83164381980896, + "learning_rate": 7.134918013174119e-10, + "loss": 3.1935, + "step": 24852 + }, + { + "epoch": 1.9924836142791285, + "grad_norm": 1.7426172494888306, + "learning_rate": 6.980655866994923e-10, + "loss": 3.1087, + "step": 24853 + }, + { + "epoch": 1.992563789060151, + "grad_norm": 1.8265361785888672, + "learning_rate": 6.828079567877677e-10, + "loss": 3.2471, + "step": 24854 + }, + { + "epoch": 1.9926439638411737, + "grad_norm": 1.8085260391235352, + "learning_rate": 6.6771891184092e-10, + "loss": 3.1706, + "step": 24855 + }, + { + "epoch": 1.9927241386221963, + "grad_norm": 0.6379968523979187, + "learning_rate": 6.527984521109698e-10, + "loss": 2.4596, + "step": 24856 + }, + { + "epoch": 1.992804313403219, + "grad_norm": 1.8321897983551025, + "learning_rate": 6.380465778521583e-10, + "loss": 3.1727, + "step": 24857 + }, + { + "epoch": 1.9928844881842416, + "grad_norm": 1.7940553426742554, + "learning_rate": 6.234632893109549e-10, + "loss": 3.2444, + "step": 24858 + }, + { + "epoch": 1.9929646629652642, + "grad_norm": 1.7087793350219727, + "learning_rate": 6.090485867338292e-10, + "loss": 3.2279, + "step": 24859 + }, + { + "epoch": 1.9930448377462868, + "grad_norm": 1.7679072618484497, + "learning_rate": 5.9480247036392e-10, + "loss": 3.1918, + "step": 24860 + }, + { + "epoch": 1.9931250125273094, + "grad_norm": 1.723821759223938, + "learning_rate": 5.807249404421455e-10, + "loss": 3.2209, + "step": 24861 + }, + { + "epoch": 1.993205187308332, + "grad_norm": 1.7913635969161987, + "learning_rate": 5.668159972049836e-10, + "loss": 3.2268, + "step": 24862 + }, + { + "epoch": 1.9932853620893547, + "grad_norm": 1.8301200866699219, + "learning_rate": 5.530756408878013e-10, + "loss": 3.3096, + "step": 24863 + }, + { + "epoch": 1.9933655368703773, + "grad_norm": 1.842492938041687, + "learning_rate": 5.395038717204148e-10, + "loss": 3.2123, + "step": 24864 + }, + { + "epoch": 1.9934457116514, + "grad_norm": 1.8915539979934692, + "learning_rate": 5.261006899337506e-10, + "loss": 3.1704, + "step": 24865 + }, + { + "epoch": 1.9935258864324226, + "grad_norm": 1.7298859357833862, + "learning_rate": 5.128660957520737e-10, + "loss": 3.1749, + "step": 24866 + }, + { + "epoch": 1.9936060612134452, + "grad_norm": 1.8050777912139893, + "learning_rate": 4.998000893996491e-10, + "loss": 3.1144, + "step": 24867 + }, + { + "epoch": 1.9936862359944678, + "grad_norm": 1.6890984773635864, + "learning_rate": 4.86902671096301e-10, + "loss": 3.108, + "step": 24868 + }, + { + "epoch": 1.9937664107754904, + "grad_norm": 1.7279125452041626, + "learning_rate": 4.74173841059633e-10, + "loss": 3.1367, + "step": 24869 + }, + { + "epoch": 1.993846585556513, + "grad_norm": 1.821815848350525, + "learning_rate": 4.616135995050286e-10, + "loss": 3.1836, + "step": 24870 + }, + { + "epoch": 1.9939267603375357, + "grad_norm": 1.7734586000442505, + "learning_rate": 4.4922194664231976e-10, + "loss": 3.2814, + "step": 24871 + }, + { + "epoch": 1.9940069351185583, + "grad_norm": 1.729134440422058, + "learning_rate": 4.3699888268244893e-10, + "loss": 3.1017, + "step": 24872 + }, + { + "epoch": 1.994087109899581, + "grad_norm": 1.8591558933258057, + "learning_rate": 4.2494440782969715e-10, + "loss": 3.2531, + "step": 24873 + }, + { + "epoch": 1.9941672846806036, + "grad_norm": 1.8826485872268677, + "learning_rate": 4.1305852228945565e-10, + "loss": 3.2616, + "step": 24874 + }, + { + "epoch": 1.9942474594616262, + "grad_norm": 1.779883861541748, + "learning_rate": 4.0134122625934413e-10, + "loss": 3.1902, + "step": 24875 + }, + { + "epoch": 1.994327634242649, + "grad_norm": 0.6695825457572937, + "learning_rate": 3.8979251993920275e-10, + "loss": 2.4741, + "step": 24876 + }, + { + "epoch": 1.9944078090236717, + "grad_norm": 1.7433513402938843, + "learning_rate": 3.7841240352332053e-10, + "loss": 3.1945, + "step": 24877 + }, + { + "epoch": 1.9944879838046943, + "grad_norm": 1.8496874570846558, + "learning_rate": 3.6720087720265583e-10, + "loss": 3.243, + "step": 24878 + }, + { + "epoch": 1.994568158585717, + "grad_norm": 1.7074483633041382, + "learning_rate": 3.5615794116705684e-10, + "loss": 3.0369, + "step": 24879 + }, + { + "epoch": 1.9946483333667395, + "grad_norm": 1.7718403339385986, + "learning_rate": 3.452835956019307e-10, + "loss": 3.1859, + "step": 24880 + }, + { + "epoch": 1.9947285081477621, + "grad_norm": 1.8159338235855103, + "learning_rate": 3.3457784069157453e-10, + "loss": 3.1944, + "step": 24881 + }, + { + "epoch": 1.9948086829287848, + "grad_norm": 1.7250896692276, + "learning_rate": 3.240406766158444e-10, + "loss": 3.0884, + "step": 24882 + }, + { + "epoch": 1.9948888577098074, + "grad_norm": 1.8085150718688965, + "learning_rate": 3.13672103552376e-10, + "loss": 3.1565, + "step": 24883 + }, + { + "epoch": 1.99496903249083, + "grad_norm": 0.6446186304092407, + "learning_rate": 3.034721216765846e-10, + "loss": 2.4685, + "step": 24884 + }, + { + "epoch": 1.9950492072718526, + "grad_norm": 1.7765675783157349, + "learning_rate": 2.934407311594445e-10, + "loss": 3.1328, + "step": 24885 + }, + { + "epoch": 1.9951293820528753, + "grad_norm": 1.885644555091858, + "learning_rate": 2.8357793217081987e-10, + "loss": 3.1498, + "step": 24886 + }, + { + "epoch": 1.995209556833898, + "grad_norm": 1.8412638902664185, + "learning_rate": 2.738837248772441e-10, + "loss": 3.2326, + "step": 24887 + }, + { + "epoch": 1.9952897316149205, + "grad_norm": 1.7694153785705566, + "learning_rate": 2.6435810944192007e-10, + "loss": 3.1067, + "step": 24888 + }, + { + "epoch": 1.9953699063959431, + "grad_norm": 1.6498897075653076, + "learning_rate": 2.5500108602471985e-10, + "loss": 3.1937, + "step": 24889 + }, + { + "epoch": 1.9954500811769658, + "grad_norm": 1.6841484308242798, + "learning_rate": 2.458126547844053e-10, + "loss": 3.1776, + "step": 24890 + }, + { + "epoch": 1.9955302559579884, + "grad_norm": 1.8583883047103882, + "learning_rate": 2.367928158741872e-10, + "loss": 3.171, + "step": 24891 + }, + { + "epoch": 1.995610430739011, + "grad_norm": 0.6421600580215454, + "learning_rate": 2.2794156944838663e-10, + "loss": 2.4563, + "step": 24892 + }, + { + "epoch": 1.9956906055200336, + "grad_norm": 1.752533197402954, + "learning_rate": 2.1925891565577338e-10, + "loss": 3.1921, + "step": 24893 + }, + { + "epoch": 1.9957707803010563, + "grad_norm": 1.8667995929718018, + "learning_rate": 2.1074485464178673e-10, + "loss": 3.2562, + "step": 24894 + }, + { + "epoch": 1.9958509550820789, + "grad_norm": 1.8528014421463013, + "learning_rate": 2.0239938654964543e-10, + "loss": 3.2018, + "step": 24895 + }, + { + "epoch": 1.9959311298631017, + "grad_norm": 1.6890158653259277, + "learning_rate": 1.9422251152145798e-10, + "loss": 3.1221, + "step": 24896 + }, + { + "epoch": 1.9960113046441244, + "grad_norm": 1.8466147184371948, + "learning_rate": 1.862142296937819e-10, + "loss": 3.0551, + "step": 24897 + }, + { + "epoch": 1.996091479425147, + "grad_norm": 1.7845368385314941, + "learning_rate": 1.7837454120317454e-10, + "loss": 3.2128, + "step": 24898 + }, + { + "epoch": 1.9961716542061696, + "grad_norm": 1.8319621086120605, + "learning_rate": 1.7070344618064226e-10, + "loss": 3.2888, + "step": 24899 + }, + { + "epoch": 1.9962518289871922, + "grad_norm": 1.8548657894134521, + "learning_rate": 1.6320094475497094e-10, + "loss": 3.139, + "step": 24900 + }, + { + "epoch": 1.9963320037682148, + "grad_norm": 1.7344075441360474, + "learning_rate": 1.558670370538362e-10, + "loss": 3.1775, + "step": 24901 + }, + { + "epoch": 1.9964121785492375, + "grad_norm": 1.9183560609817505, + "learning_rate": 1.4870172320158304e-10, + "loss": 3.1699, + "step": 24902 + }, + { + "epoch": 1.99649235333026, + "grad_norm": 1.8523657321929932, + "learning_rate": 1.4170500331700532e-10, + "loss": 3.2217, + "step": 24903 + }, + { + "epoch": 1.9965725281112827, + "grad_norm": 1.8206969499588013, + "learning_rate": 1.348768775188969e-10, + "loss": 3.1922, + "step": 24904 + }, + { + "epoch": 1.9966527028923053, + "grad_norm": 1.8042018413543701, + "learning_rate": 1.2821734592272094e-10, + "loss": 3.0759, + "step": 24905 + }, + { + "epoch": 1.996732877673328, + "grad_norm": 1.8501999378204346, + "learning_rate": 1.2172640864061002e-10, + "loss": 3.213, + "step": 24906 + }, + { + "epoch": 1.9968130524543506, + "grad_norm": 1.928696632385254, + "learning_rate": 1.1540406578136598e-10, + "loss": 3.2074, + "step": 24907 + }, + { + "epoch": 1.9968932272353732, + "grad_norm": 1.7480429410934448, + "learning_rate": 1.0925031745268044e-10, + "loss": 3.2377, + "step": 24908 + }, + { + "epoch": 1.9969734020163958, + "grad_norm": 1.7484110593795776, + "learning_rate": 1.0326516375669393e-10, + "loss": 3.2495, + "step": 24909 + }, + { + "epoch": 1.9970535767974185, + "grad_norm": 1.6273787021636963, + "learning_rate": 9.744860479665719e-11, + "loss": 3.1304, + "step": 24910 + }, + { + "epoch": 1.997133751578441, + "grad_norm": 0.6456690430641174, + "learning_rate": 9.180064066804939e-11, + "loss": 2.4695, + "step": 24911 + }, + { + "epoch": 1.9972139263594637, + "grad_norm": 1.8286094665527344, + "learning_rate": 8.632127146857017e-11, + "loss": 3.1892, + "step": 24912 + }, + { + "epoch": 1.9972941011404863, + "grad_norm": 1.6978219747543335, + "learning_rate": 8.10104972892578e-11, + "loss": 3.1527, + "step": 24913 + }, + { + "epoch": 1.997374275921509, + "grad_norm": 1.8521363735198975, + "learning_rate": 7.586831821893014e-11, + "loss": 3.1976, + "step": 24914 + }, + { + "epoch": 1.9974544507025316, + "grad_norm": 1.7367768287658691, + "learning_rate": 7.089473434640504e-11, + "loss": 3.1427, + "step": 24915 + }, + { + "epoch": 1.9975346254835542, + "grad_norm": 1.9504386186599731, + "learning_rate": 6.608974575272875e-11, + "loss": 3.2921, + "step": 24916 + }, + { + "epoch": 1.9976148002645768, + "grad_norm": 1.805375576019287, + "learning_rate": 6.145335252227825e-11, + "loss": 3.1935, + "step": 24917 + }, + { + "epoch": 1.9976949750455995, + "grad_norm": 1.9018357992172241, + "learning_rate": 5.698555472943845e-11, + "loss": 3.1115, + "step": 24918 + }, + { + "epoch": 1.997775149826622, + "grad_norm": 1.7999831438064575, + "learning_rate": 5.268635245303522e-11, + "loss": 3.1748, + "step": 24919 + }, + { + "epoch": 1.9978553246076447, + "grad_norm": 1.8577394485473633, + "learning_rate": 4.8555745763012585e-11, + "loss": 3.1733, + "step": 24920 + }, + { + "epoch": 1.9979354993886673, + "grad_norm": 1.7686206102371216, + "learning_rate": 4.459373473042483e-11, + "loss": 3.1471, + "step": 24921 + }, + { + "epoch": 1.99801567416969, + "grad_norm": 1.7081915140151978, + "learning_rate": 4.080031942188534e-11, + "loss": 3.0477, + "step": 24922 + }, + { + "epoch": 1.9980958489507126, + "grad_norm": 1.775012731552124, + "learning_rate": 3.717549990067681e-11, + "loss": 3.2764, + "step": 24923 + }, + { + "epoch": 1.9981760237317352, + "grad_norm": 1.7206308841705322, + "learning_rate": 3.371927622897175e-11, + "loss": 3.0602, + "step": 24924 + }, + { + "epoch": 1.9982561985127578, + "grad_norm": 1.7391258478164673, + "learning_rate": 3.043164846450175e-11, + "loss": 3.127, + "step": 24925 + }, + { + "epoch": 1.9983363732937804, + "grad_norm": 1.8052302598953247, + "learning_rate": 2.7312616661667735e-11, + "loss": 3.1496, + "step": 24926 + }, + { + "epoch": 1.998416548074803, + "grad_norm": 1.7362072467803955, + "learning_rate": 2.4362180874870633e-11, + "loss": 3.1578, + "step": 24927 + }, + { + "epoch": 1.9984967228558257, + "grad_norm": 1.7218022346496582, + "learning_rate": 2.158034115296026e-11, + "loss": 3.0719, + "step": 24928 + }, + { + "epoch": 1.9985768976368483, + "grad_norm": 1.7143175601959229, + "learning_rate": 1.8967097542565982e-11, + "loss": 3.084, + "step": 24929 + }, + { + "epoch": 1.998657072417871, + "grad_norm": 1.7346006631851196, + "learning_rate": 1.652245008809672e-11, + "loss": 3.1211, + "step": 24930 + }, + { + "epoch": 1.9987372471988936, + "grad_norm": 1.7399567365646362, + "learning_rate": 1.4246398830630726e-11, + "loss": 3.1588, + "step": 24931 + }, + { + "epoch": 1.9988174219799162, + "grad_norm": 1.7875056266784668, + "learning_rate": 1.2138943809025804e-11, + "loss": 3.2161, + "step": 24932 + }, + { + "epoch": 1.9988975967609388, + "grad_norm": 1.8564636707305908, + "learning_rate": 1.0200085057698872e-11, + "loss": 3.2246, + "step": 24933 + }, + { + "epoch": 1.9989777715419614, + "grad_norm": 1.848548173904419, + "learning_rate": 8.429822609956618e-12, + "loss": 3.1787, + "step": 24934 + }, + { + "epoch": 1.999057946322984, + "grad_norm": 1.7729980945587158, + "learning_rate": 6.8281564968852856e-12, + "loss": 3.1758, + "step": 24935 + }, + { + "epoch": 1.9991381211040067, + "grad_norm": 1.7523778676986694, + "learning_rate": 5.395086744020006e-12, + "loss": 3.1297, + "step": 24936 + }, + { + "epoch": 1.9992182958850293, + "grad_norm": 1.8871723413467407, + "learning_rate": 4.130613374675463e-12, + "loss": 3.2038, + "step": 24937 + }, + { + "epoch": 1.999298470666052, + "grad_norm": 1.7288553714752197, + "learning_rate": 3.0347364121663393e-12, + "loss": 3.152, + "step": 24938 + }, + { + "epoch": 1.9993786454470746, + "grad_norm": 1.764301061630249, + "learning_rate": 2.1074558742562035e-12, + "loss": 3.1128, + "step": 24939 + }, + { + "epoch": 1.9994588202280972, + "grad_norm": 1.9079736471176147, + "learning_rate": 1.348771776488178e-12, + "loss": 3.2594, + "step": 24940 + }, + { + "epoch": 1.9995389950091198, + "grad_norm": 1.7804877758026123, + "learning_rate": 7.586841321849392e-13, + "loss": 3.2012, + "step": 24941 + }, + { + "epoch": 1.9996191697901424, + "grad_norm": 1.7511546611785889, + "learning_rate": 3.3719295022827114e-13, + "loss": 3.1757, + "step": 24942 + }, + { + "epoch": 1.999699344571165, + "grad_norm": 1.7623884677886963, + "learning_rate": 8.429823727951204e-14, + "loss": 3.0733, + "step": 24943 + }, + { + "epoch": 1.9997795193521877, + "grad_norm": 1.8036004304885864, + "learning_rate": 0.0, + "loss": 3.256, + "step": 24944 + } + ], + "logging_steps": 1.0, + "max_steps": 24944, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.008356382172984e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}