diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6565 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 933,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010718113612004287,
+      "grad_norm": 18.787860870361328,
+      "learning_rate": 0.0002,
+      "loss": 11.4429,
+      "step": 1
+    },
+    {
+      "epoch": 0.0021436227224008574,
+      "grad_norm": 18.825637817382812,
+      "learning_rate": 0.00019978563772775994,
+      "loss": 10.4711,
+      "step": 2
+    },
+    {
+      "epoch": 0.003215434083601286,
+      "grad_norm": 10.722309112548828,
+      "learning_rate": 0.00019957127545551983,
+      "loss": 8.4237,
+      "step": 3
+    },
+    {
+      "epoch": 0.004287245444801715,
+      "grad_norm": 8.985899925231934,
+      "learning_rate": 0.00019935691318327976,
+      "loss": 7.7351,
+      "step": 4
+    },
+    {
+      "epoch": 0.0053590568060021436,
+      "grad_norm": 7.36727237701416,
+      "learning_rate": 0.00019914255091103966,
+      "loss": 7.1636,
+      "step": 5
+    },
+    {
+      "epoch": 0.006430868167202572,
+      "grad_norm": 16.767192840576172,
+      "learning_rate": 0.00019892818863879958,
+      "loss": 7.2813,
+      "step": 6
+    },
+    {
+      "epoch": 0.007502679528403001,
+      "grad_norm": 5.598256587982178,
+      "learning_rate": 0.00019871382636655948,
+      "loss": 6.6949,
+      "step": 7
+    },
+    {
+      "epoch": 0.00857449088960343,
+      "grad_norm": 9.017542839050293,
+      "learning_rate": 0.0001984994640943194,
+      "loss": 6.8995,
+      "step": 8
+    },
+    {
+      "epoch": 0.00964630225080386,
+      "grad_norm": 4.33115816116333,
+      "learning_rate": 0.00019828510182207934,
+      "loss": 6.3657,
+      "step": 9
+    },
+    {
+      "epoch": 0.010718113612004287,
+      "grad_norm": 3.3178164958953857,
+      "learning_rate": 0.00019807073954983923,
+      "loss": 6.3438,
+      "step": 10
+    },
+    {
+      "epoch": 0.011789924973204717,
+      "grad_norm": 2.961005687713623,
+      "learning_rate": 0.00019785637727759916,
+      "loss": 5.9906,
+      "step": 11
+    },
+    {
+      "epoch": 0.012861736334405145,
+      "grad_norm": 4.369665145874023,
+      "learning_rate": 0.00019764201500535906,
+      "loss": 6.3405,
+      "step": 12
+    },
+    {
+      "epoch": 0.013933547695605574,
+      "grad_norm": 7.567108154296875,
+      "learning_rate": 0.00019742765273311899,
+      "loss": 6.1571,
+      "step": 13
+    },
+    {
+      "epoch": 0.015005359056806002,
+      "grad_norm": 6.276333808898926,
+      "learning_rate": 0.00019721329046087888,
+      "loss": 5.9091,
+      "step": 14
+    },
+    {
+      "epoch": 0.01607717041800643,
+      "grad_norm": 3.475311040878296,
+      "learning_rate": 0.0001969989281886388,
+      "loss": 5.7566,
+      "step": 15
+    },
+    {
+      "epoch": 0.01714898177920686,
+      "grad_norm": 3.2999136447906494,
+      "learning_rate": 0.00019678456591639874,
+      "loss": 5.5771,
+      "step": 16
+    },
+    {
+      "epoch": 0.01822079314040729,
+      "grad_norm": 2.9271633625030518,
+      "learning_rate": 0.00019657020364415863,
+      "loss": 5.7828,
+      "step": 17
+    },
+    {
+      "epoch": 0.01929260450160772,
+      "grad_norm": 4.073993682861328,
+      "learning_rate": 0.00019635584137191856,
+      "loss": 5.8397,
+      "step": 18
+    },
+    {
+      "epoch": 0.020364415862808145,
+      "grad_norm": 4.44898796081543,
+      "learning_rate": 0.00019614147909967846,
+      "loss": 5.5915,
+      "step": 19
+    },
+    {
+      "epoch": 0.021436227224008574,
+      "grad_norm": 3.62903094291687,
+      "learning_rate": 0.00019592711682743839,
+      "loss": 5.6345,
+      "step": 20
+    },
+    {
+      "epoch": 0.022508038585209004,
+      "grad_norm": 3.9096498489379883,
+      "learning_rate": 0.00019571275455519828,
+      "loss": 5.2657,
+      "step": 21
+    },
+    {
+      "epoch": 0.023579849946409433,
+      "grad_norm": 5.483320236206055,
+      "learning_rate": 0.0001954983922829582,
+      "loss": 5.7525,
+      "step": 22
+    },
+    {
+      "epoch": 0.02465166130760986,
+      "grad_norm": 4.441135883331299,
+      "learning_rate": 0.00019528403001071814,
+      "loss": 5.1491,
+      "step": 23
+    },
+    {
+      "epoch": 0.02572347266881029,
+      "grad_norm": 4.325745582580566,
+      "learning_rate": 0.00019506966773847803,
+      "loss": 5.3696,
+      "step": 24
+    },
+    {
+      "epoch": 0.02679528403001072,
+      "grad_norm": 5.317447185516357,
+      "learning_rate": 0.00019485530546623796,
+      "loss": 5.1451,
+      "step": 25
+    },
+    {
+      "epoch": 0.027867095391211148,
+      "grad_norm": 2.9565281867980957,
+      "learning_rate": 0.00019464094319399786,
+      "loss": 5.1643,
+      "step": 26
+    },
+    {
+      "epoch": 0.028938906752411574,
+      "grad_norm": 3.596660614013672,
+      "learning_rate": 0.00019442658092175779,
+      "loss": 5.3528,
+      "step": 27
+    },
+    {
+      "epoch": 0.030010718113612004,
+      "grad_norm": 3.135538339614868,
+      "learning_rate": 0.00019421221864951768,
+      "loss": 5.2368,
+      "step": 28
+    },
+    {
+      "epoch": 0.031082529474812434,
+      "grad_norm": 4.410246849060059,
+      "learning_rate": 0.0001939978563772776,
+      "loss": 5.3679,
+      "step": 29
+    },
+    {
+      "epoch": 0.03215434083601286,
+      "grad_norm": 5.01721715927124,
+      "learning_rate": 0.00019378349410503754,
+      "loss": 5.3211,
+      "step": 30
+    },
+    {
+      "epoch": 0.03322615219721329,
+      "grad_norm": 3.6833577156066895,
+      "learning_rate": 0.00019356913183279743,
+      "loss": 5.5033,
+      "step": 31
+    },
+    {
+      "epoch": 0.03429796355841372,
+      "grad_norm": 3.7026724815368652,
+      "learning_rate": 0.00019335476956055736,
+      "loss": 5.2373,
+      "step": 32
+    },
+    {
+      "epoch": 0.03536977491961415,
+      "grad_norm": 2.8100647926330566,
+      "learning_rate": 0.00019314040728831726,
+      "loss": 5.3375,
+      "step": 33
+    },
+    {
+      "epoch": 0.03644158628081458,
+      "grad_norm": 3.55210280418396,
+      "learning_rate": 0.00019292604501607719,
+      "loss": 5.104,
+      "step": 34
+    },
+    {
+      "epoch": 0.03751339764201501,
+      "grad_norm": 3.5566956996917725,
+      "learning_rate": 0.00019271168274383708,
+      "loss": 5.3136,
+      "step": 35
+    },
+    {
+      "epoch": 0.03858520900321544,
+      "grad_norm": 3.579719066619873,
+      "learning_rate": 0.000192497320471597,
+      "loss": 5.4142,
+      "step": 36
+    },
+    {
+      "epoch": 0.03965702036441586,
+      "grad_norm": 3.1719672679901123,
+      "learning_rate": 0.00019228295819935694,
+      "loss": 5.2212,
+      "step": 37
+    },
+    {
+      "epoch": 0.04072883172561629,
+      "grad_norm": 3.3072733879089355,
+      "learning_rate": 0.00019206859592711684,
+      "loss": 5.1418,
+      "step": 38
+    },
+    {
+      "epoch": 0.04180064308681672,
+      "grad_norm": 3.1818485260009766,
+      "learning_rate": 0.00019185423365487676,
+      "loss": 5.0819,
+      "step": 39
+    },
+    {
+      "epoch": 0.04287245444801715,
+      "grad_norm": 2.7018849849700928,
+      "learning_rate": 0.00019163987138263666,
+      "loss": 4.9725,
+      "step": 40
+    },
+    {
+      "epoch": 0.04394426580921758,
+      "grad_norm": 3.215134382247925,
+      "learning_rate": 0.00019142550911039659,
+      "loss": 5.1623,
+      "step": 41
+    },
+    {
+      "epoch": 0.04501607717041801,
+      "grad_norm": 3.6956472396850586,
+      "learning_rate": 0.00019121114683815648,
+      "loss": 5.2277,
+      "step": 42
+    },
+    {
+      "epoch": 0.04608788853161844,
+      "grad_norm": 3.72455096244812,
+      "learning_rate": 0.0001909967845659164,
+      "loss": 5.1207,
+      "step": 43
+    },
+    {
+      "epoch": 0.04715969989281887,
+      "grad_norm": 3.6416709423065186,
+      "learning_rate": 0.00019078242229367634,
+      "loss": 5.4012,
+      "step": 44
+    },
+    {
+      "epoch": 0.04823151125401929,
+      "grad_norm": 3.856046676635742,
+      "learning_rate": 0.00019056806002143624,
+      "loss": 4.9041,
+      "step": 45
+    },
+    {
+      "epoch": 0.04930332261521972,
+      "grad_norm": 3.6323482990264893,
+      "learning_rate": 0.00019035369774919616,
+      "loss": 5.0846,
+      "step": 46
+    },
+    {
+      "epoch": 0.05037513397642015,
+      "grad_norm": 4.227871417999268,
+      "learning_rate": 0.00019013933547695606,
+      "loss": 5.2468,
+      "step": 47
+    },
+    {
+      "epoch": 0.05144694533762058,
+      "grad_norm": 4.334918022155762,
+      "learning_rate": 0.00018992497320471599,
+      "loss": 5.3007,
+      "step": 48
+    },
+    {
+      "epoch": 0.05251875669882101,
+      "grad_norm": 4.2465105056762695,
+      "learning_rate": 0.00018971061093247588,
+      "loss": 4.8531,
+      "step": 49
+    },
+    {
+      "epoch": 0.05359056806002144,
+      "grad_norm": 4.0939106941223145,
+      "learning_rate": 0.0001894962486602358,
+      "loss": 5.2677,
+      "step": 50
+    },
+    {
+      "epoch": 0.05466237942122187,
+      "grad_norm": 4.415087699890137,
+      "learning_rate": 0.00018928188638799574,
+      "loss": 5.1889,
+      "step": 51
+    },
+    {
+      "epoch": 0.055734190782422297,
+      "grad_norm": 2.4377565383911133,
+      "learning_rate": 0.00018906752411575564,
+      "loss": 5.0309,
+      "step": 52
+    },
+    {
+      "epoch": 0.05680600214362272,
+      "grad_norm": 3.9000558853149414,
+      "learning_rate": 0.00018885316184351556,
+      "loss": 5.0311,
+      "step": 53
+    },
+    {
+      "epoch": 0.05787781350482315,
+      "grad_norm": 3.740562915802002,
+      "learning_rate": 0.00018863879957127546,
+      "loss": 5.2185,
+      "step": 54
+    },
+    {
+      "epoch": 0.05894962486602358,
+      "grad_norm": 2.913926124572754,
+      "learning_rate": 0.00018842443729903539,
+      "loss": 4.7942,
+      "step": 55
+    },
+    {
+      "epoch": 0.06002143622722401,
+      "grad_norm": 3.0584166049957275,
+      "learning_rate": 0.00018821007502679528,
+      "loss": 5.1003,
+      "step": 56
+    },
+    {
+      "epoch": 0.06109324758842444,
+      "grad_norm": 3.646268367767334,
+      "learning_rate": 0.0001879957127545552,
+      "loss": 5.3191,
+      "step": 57
+    },
+    {
+      "epoch": 0.06216505894962487,
+      "grad_norm": 4.751884460449219,
+      "learning_rate": 0.00018778135048231514,
+      "loss": 4.9535,
+      "step": 58
+    },
+    {
+      "epoch": 0.0632368703108253,
+      "grad_norm": 3.8159265518188477,
+      "learning_rate": 0.00018756698821007504,
+      "loss": 5.0055,
+      "step": 59
+    },
+    {
+      "epoch": 0.06430868167202572,
+      "grad_norm": 2.9061062335968018,
+      "learning_rate": 0.00018735262593783496,
+      "loss": 5.0264,
+      "step": 60
+    },
+    {
+      "epoch": 0.06538049303322616,
+      "grad_norm": 3.322441577911377,
+      "learning_rate": 0.00018713826366559486,
+      "loss": 5.1263,
+      "step": 61
+    },
+    {
+      "epoch": 0.06645230439442658,
+      "grad_norm": 3.4240949153900146,
+      "learning_rate": 0.00018692390139335479,
+      "loss": 5.0013,
+      "step": 62
+    },
+    {
+      "epoch": 0.06752411575562701,
+      "grad_norm": 3.418482780456543,
+      "learning_rate": 0.00018670953912111469,
+      "loss": 5.4505,
+      "step": 63
+    },
+    {
+      "epoch": 0.06859592711682744,
+      "grad_norm": 3.3597252368927,
+      "learning_rate": 0.0001864951768488746,
+      "loss": 4.9659,
+      "step": 64
+    },
+    {
+      "epoch": 0.06966773847802786,
+      "grad_norm": 3.013861894607544,
+      "learning_rate": 0.0001862808145766345,
+      "loss": 5.1352,
+      "step": 65
+    },
+    {
+      "epoch": 0.0707395498392283,
+      "grad_norm": 5.465812683105469,
+      "learning_rate": 0.00018606645230439444,
+      "loss": 4.8951,
+      "step": 66
+    },
+    {
+      "epoch": 0.07181136120042872,
+      "grad_norm": 4.695466995239258,
+      "learning_rate": 0.00018585209003215436,
+      "loss": 4.7569,
+      "step": 67
+    },
+    {
+      "epoch": 0.07288317256162916,
+      "grad_norm": 4.39758825302124,
+      "learning_rate": 0.00018563772775991426,
+      "loss": 5.0259,
+      "step": 68
+    },
+    {
+      "epoch": 0.07395498392282958,
+      "grad_norm": 3.696704626083374,
+      "learning_rate": 0.00018542336548767419,
+      "loss": 4.7529,
+      "step": 69
+    },
+    {
+      "epoch": 0.07502679528403002,
+      "grad_norm": 5.415925979614258,
+      "learning_rate": 0.00018520900321543409,
+      "loss": 5.3223,
+      "step": 70
+    },
+    {
+      "epoch": 0.07609860664523044,
+      "grad_norm": 3.055694341659546,
+      "learning_rate": 0.000184994640943194,
+      "loss": 5.2108,
+      "step": 71
+    },
+    {
+      "epoch": 0.07717041800643087,
+      "grad_norm": 4.533783912658691,
+      "learning_rate": 0.0001847802786709539,
+      "loss": 4.9434,
+      "step": 72
+    },
+    {
+      "epoch": 0.0782422293676313,
+      "grad_norm": 2.5685031414031982,
+      "learning_rate": 0.00018456591639871384,
+      "loss": 5.1461,
+      "step": 73
+    },
+    {
+      "epoch": 0.07931404072883172,
+      "grad_norm": 3.007826805114746,
+      "learning_rate": 0.00018435155412647376,
+      "loss": 5.058,
+      "step": 74
+    },
+    {
+      "epoch": 0.08038585209003216,
+      "grad_norm": 3.6478159427642822,
+      "learning_rate": 0.00018413719185423366,
+      "loss": 4.7813,
+      "step": 75
+    },
+    {
+      "epoch": 0.08145766345123258,
+      "grad_norm": 2.475228786468506,
+      "learning_rate": 0.0001839228295819936,
+      "loss": 5.0926,
+      "step": 76
+    },
+    {
+      "epoch": 0.08252947481243302,
+      "grad_norm": 2.8211352825164795,
+      "learning_rate": 0.00018370846730975349,
+      "loss": 5.0912,
+      "step": 77
+    },
+    {
+      "epoch": 0.08360128617363344,
+      "grad_norm": 2.761347532272339,
+      "learning_rate": 0.0001834941050375134,
+      "loss": 4.9557,
+      "step": 78
+    },
+    {
+      "epoch": 0.08467309753483387,
+      "grad_norm": 3.7475688457489014,
+      "learning_rate": 0.0001832797427652733,
+      "loss": 5.2675,
+      "step": 79
+    },
+    {
+      "epoch": 0.0857449088960343,
+      "grad_norm": 3.1617653369903564,
+      "learning_rate": 0.00018306538049303324,
+      "loss": 5.1046,
+      "step": 80
+    },
+    {
+      "epoch": 0.08681672025723473,
+      "grad_norm": 2.875410556793213,
+      "learning_rate": 0.00018285101822079316,
+      "loss": 5.0993,
+      "step": 81
+    },
+    {
+      "epoch": 0.08788853161843516,
+      "grad_norm": 3.297821521759033,
+      "learning_rate": 0.00018263665594855306,
+      "loss": 4.8784,
+      "step": 82
+    },
+    {
+      "epoch": 0.08896034297963558,
+      "grad_norm": 2.7489824295043945,
+      "learning_rate": 0.000182422293676313,
+      "loss": 5.0613,
+      "step": 83
+    },
+    {
+      "epoch": 0.09003215434083602,
+      "grad_norm": 4.606520652770996,
+      "learning_rate": 0.00018220793140407289,
+      "loss": 4.795,
+      "step": 84
+    },
+    {
+      "epoch": 0.09110396570203644,
+      "grad_norm": 3.6237289905548096,
+      "learning_rate": 0.0001819935691318328,
+      "loss": 4.9828,
+      "step": 85
+    },
+    {
+      "epoch": 0.09217577706323687,
+      "grad_norm": 3.9898364543914795,
+      "learning_rate": 0.0001817792068595927,
+      "loss": 4.9822,
+      "step": 86
+    },
+    {
+      "epoch": 0.0932475884244373,
+      "grad_norm": 3.16852068901062,
+      "learning_rate": 0.00018156484458735264,
+      "loss": 4.8484,
+      "step": 87
+    },
+    {
+      "epoch": 0.09431939978563773,
+      "grad_norm": 4.019200325012207,
+      "learning_rate": 0.00018135048231511256,
+      "loss": 5.2485,
+      "step": 88
+    },
+    {
+      "epoch": 0.09539121114683816,
+      "grad_norm": 4.304863929748535,
+      "learning_rate": 0.00018113612004287246,
+      "loss": 4.8591,
+      "step": 89
+    },
+    {
+      "epoch": 0.09646302250803858,
+      "grad_norm": 3.7539124488830566,
+      "learning_rate": 0.0001809217577706324,
+      "loss": 5.1811,
+      "step": 90
+    },
+    {
+      "epoch": 0.09753483386923902,
+      "grad_norm": 4.071114540100098,
+      "learning_rate": 0.00018070739549839229,
+      "loss": 4.9069,
+      "step": 91
+    },
+    {
+      "epoch": 0.09860664523043944,
+      "grad_norm": 2.6979987621307373,
+      "learning_rate": 0.0001804930332261522,
+      "loss": 4.95,
+      "step": 92
+    },
+    {
+      "epoch": 0.09967845659163987,
+      "grad_norm": 3.5572869777679443,
+      "learning_rate": 0.0001802786709539121,
+      "loss": 5.0632,
+      "step": 93
+    },
+    {
+      "epoch": 0.1007502679528403,
+      "grad_norm": 3.7195041179656982,
+      "learning_rate": 0.00018006430868167204,
+      "loss": 4.5738,
+      "step": 94
+    },
+    {
+      "epoch": 0.10182207931404073,
+      "grad_norm": 3.0068423748016357,
+      "learning_rate": 0.00017984994640943196,
+      "loss": 5.0092,
+      "step": 95
+    },
+    {
+      "epoch": 0.10289389067524116,
+      "grad_norm": 3.6475560665130615,
+      "learning_rate": 0.00017963558413719186,
+      "loss": 4.932,
+      "step": 96
+    },
+    {
+      "epoch": 0.10396570203644159,
+      "grad_norm": 4.444890022277832,
+      "learning_rate": 0.0001794212218649518,
+      "loss": 5.3889,
+      "step": 97
+    },
+    {
+      "epoch": 0.10503751339764202,
+      "grad_norm": 2.3350093364715576,
+      "learning_rate": 0.00017920685959271169,
+      "loss": 4.8875,
+      "step": 98
+    },
+    {
+      "epoch": 0.10610932475884244,
+      "grad_norm": 5.134820938110352,
+      "learning_rate": 0.0001789924973204716,
+      "loss": 5.1827,
+      "step": 99
+    },
+    {
+      "epoch": 0.10718113612004287,
+      "grad_norm": 2.695220947265625,
+      "learning_rate": 0.0001787781350482315,
+      "loss": 5.082,
+      "step": 100
+    },
+    {
+      "epoch": 0.1082529474812433,
+      "grad_norm": 2.2416651248931885,
+      "learning_rate": 0.00017856377277599144,
+      "loss": 4.8779,
+      "step": 101
+    },
+    {
+      "epoch": 0.10932475884244373,
+      "grad_norm": 4.537072658538818,
+      "learning_rate": 0.00017834941050375136,
+      "loss": 5.1109,
+      "step": 102
+    },
+    {
+      "epoch": 0.11039657020364416,
+      "grad_norm": 3.407715320587158,
+      "learning_rate": 0.00017813504823151126,
+      "loss": 4.7432,
+      "step": 103
+    },
+    {
+      "epoch": 0.11146838156484459,
+      "grad_norm": 3.8071422576904297,
+      "learning_rate": 0.0001779206859592712,
+      "loss": 4.9924,
+      "step": 104
+    },
+    {
+      "epoch": 0.11254019292604502,
+      "grad_norm": 2.7043581008911133,
+      "learning_rate": 0.00017770632368703109,
+      "loss": 5.08,
+      "step": 105
+    },
+    {
+      "epoch": 0.11361200428724544,
+      "grad_norm": 2.9360063076019287,
+      "learning_rate": 0.000177491961414791,
+      "loss": 4.8822,
+      "step": 106
+    },
+    {
+      "epoch": 0.11468381564844587,
+      "grad_norm": 2.972593069076538,
+      "learning_rate": 0.0001772775991425509,
+      "loss": 4.9554,
+      "step": 107
+    },
+    {
+      "epoch": 0.1157556270096463,
+      "grad_norm": 3.9115068912506104,
+      "learning_rate": 0.00017706323687031084,
+      "loss": 4.9983,
+      "step": 108
+    },
+    {
+      "epoch": 0.11682743837084673,
+      "grad_norm": 2.4414217472076416,
+      "learning_rate": 0.00017684887459807076,
+      "loss": 4.8562,
+      "step": 109
+    },
+    {
+      "epoch": 0.11789924973204716,
+      "grad_norm": 4.106845378875732,
+      "learning_rate": 0.00017663451232583066,
+      "loss": 4.9848,
+      "step": 110
+    },
+    {
+      "epoch": 0.1189710610932476,
+      "grad_norm": 2.924886465072632,
+      "learning_rate": 0.0001764201500535906,
+      "loss": 4.7966,
+      "step": 111
+    },
+    {
+      "epoch": 0.12004287245444802,
+      "grad_norm": 3.0221619606018066,
+      "learning_rate": 0.00017620578778135049,
+      "loss": 4.6449,
+      "step": 112
+    },
+    {
+      "epoch": 0.12111468381564845,
+      "grad_norm": 4.032159805297852,
+      "learning_rate": 0.0001759914255091104,
+      "loss": 4.9659,
+      "step": 113
+    },
+    {
+      "epoch": 0.12218649517684887,
+      "grad_norm": 2.689523696899414,
+      "learning_rate": 0.0001757770632368703,
+      "loss": 5.0913,
+      "step": 114
+    },
+    {
+      "epoch": 0.1232583065380493,
+      "grad_norm": 4.315779685974121,
+      "learning_rate": 0.00017556270096463024,
+      "loss": 5.2312,
+      "step": 115
+    },
+    {
+      "epoch": 0.12433011789924973,
+      "grad_norm": 3.1354005336761475,
+      "learning_rate": 0.00017534833869239016,
+      "loss": 5.0489,
+      "step": 116
+    },
+    {
+      "epoch": 0.12540192926045016,
+      "grad_norm": 4.844458103179932,
+      "learning_rate": 0.00017513397642015006,
+      "loss": 5.0938,
+      "step": 117
+    },
+    {
+      "epoch": 0.1264737406216506,
+      "grad_norm": 3.60406756401062,
+      "learning_rate": 0.00017491961414791,
+      "loss": 4.9502,
+      "step": 118
+    },
+    {
+      "epoch": 0.12754555198285103,
+      "grad_norm": 2.6359314918518066,
+      "learning_rate": 0.00017470525187566989,
+      "loss": 4.7598,
+      "step": 119
+    },
+    {
+      "epoch": 0.12861736334405144,
+      "grad_norm": 2.6354403495788574,
+      "learning_rate": 0.0001744908896034298,
+      "loss": 4.7832,
+      "step": 120
+    },
+    {
+      "epoch": 0.12968917470525188,
+      "grad_norm": 3.07201886177063,
+      "learning_rate": 0.0001742765273311897,
+      "loss": 5.0391,
+      "step": 121
+    },
+    {
+      "epoch": 0.1307609860664523,
+      "grad_norm": 4.98795747756958,
+      "learning_rate": 0.00017406216505894964,
+      "loss": 5.3099,
+      "step": 122
+    },
+    {
+      "epoch": 0.13183279742765272,
+      "grad_norm": 2.553927421569824,
+      "learning_rate": 0.00017384780278670956,
+      "loss": 4.8655,
+      "step": 123
+    },
+    {
+      "epoch": 0.13290460878885316,
+      "grad_norm": 5.08751106262207,
+      "learning_rate": 0.00017363344051446946,
+      "loss": 5.2898,
+      "step": 124
+    },
+    {
+      "epoch": 0.1339764201500536,
+      "grad_norm": 3.600430965423584,
+      "learning_rate": 0.0001734190782422294,
+      "loss": 4.8845,
+      "step": 125
+    },
+    {
+      "epoch": 0.13504823151125403,
+      "grad_norm": 3.795992851257324,
+      "learning_rate": 0.00017320471596998929,
+      "loss": 5.0673,
+      "step": 126
+    },
+    {
+      "epoch": 0.13612004287245444,
+      "grad_norm": 4.044764995574951,
+      "learning_rate": 0.0001729903536977492,
+      "loss": 4.9675,
+      "step": 127
+    },
+    {
+      "epoch": 0.13719185423365488,
+      "grad_norm": 3.4079222679138184,
+      "learning_rate": 0.0001727759914255091,
+      "loss": 5.0485,
+      "step": 128
+    },
+    {
+      "epoch": 0.1382636655948553,
+      "grad_norm": 2.5793895721435547,
+      "learning_rate": 0.00017256162915326904,
+      "loss": 5.0764,
+      "step": 129
+    },
+    {
+      "epoch": 0.13933547695605572,
+      "grad_norm": 3.7445921897888184,
+      "learning_rate": 0.00017234726688102896,
+      "loss": 5.44,
+      "step": 130
+    },
+    {
+      "epoch": 0.14040728831725616,
+      "grad_norm": 2.7735092639923096,
+      "learning_rate": 0.00017213290460878886,
+      "loss": 4.5954,
+      "step": 131
+    },
+    {
+      "epoch": 0.1414790996784566,
+      "grad_norm": 4.459575653076172,
+      "learning_rate": 0.0001719185423365488,
+      "loss": 4.6362,
+      "step": 132
+    },
+    {
+      "epoch": 0.14255091103965703,
+      "grad_norm": 4.086177349090576,
+      "learning_rate": 0.0001717041800643087,
+      "loss": 4.8736,
+      "step": 133
+    },
+    {
+      "epoch": 0.14362272240085744,
+      "grad_norm": 2.562479257583618,
+      "learning_rate": 0.0001714898177920686,
+      "loss": 4.9779,
+      "step": 134
+    },
+    {
+      "epoch": 0.14469453376205788,
+      "grad_norm": 2.590363025665283,
+      "learning_rate": 0.0001712754555198285,
+      "loss": 4.8501,
+      "step": 135
+    },
+    {
+      "epoch": 0.1457663451232583,
+      "grad_norm": 4.304504871368408,
+      "learning_rate": 0.00017106109324758844,
+      "loss": 4.7176,
+      "step": 136
+    },
+    {
+      "epoch": 0.14683815648445875,
+      "grad_norm": 2.6717023849487305,
+      "learning_rate": 0.00017084673097534836,
+      "loss": 4.9784,
+      "step": 137
+    },
+    {
+      "epoch": 0.14790996784565916,
+      "grad_norm": 3.3490304946899414,
+      "learning_rate": 0.00017063236870310826,
+      "loss": 5.0833,
+      "step": 138
+    },
+    {
+      "epoch": 0.1489817792068596,
+      "grad_norm": 3.373300790786743,
+      "learning_rate": 0.0001704180064308682,
+      "loss": 5.0491,
+      "step": 139
+    },
+    {
+      "epoch": 0.15005359056806003,
+      "grad_norm": 3.4800350666046143,
+      "learning_rate": 0.0001702036441586281,
+      "loss": 4.8633,
+      "step": 140
+    },
+    {
+      "epoch": 0.15112540192926044,
+      "grad_norm": 4.172618865966797,
+      "learning_rate": 0.000169989281886388,
+      "loss": 4.7306,
+      "step": 141
+    },
+    {
+      "epoch": 0.15219721329046088,
+      "grad_norm": 2.529625654220581,
+      "learning_rate": 0.0001697749196141479,
+      "loss": 5.0736,
+      "step": 142
+    },
+    {
+      "epoch": 0.1532690246516613,
+      "grad_norm": 4.884538650512695,
+      "learning_rate": 0.00016956055734190784,
+      "loss": 4.7077,
+      "step": 143
+    },
+    {
+      "epoch": 0.15434083601286175,
+      "grad_norm": 2.5472567081451416,
+      "learning_rate": 0.00016934619506966776,
+      "loss": 5.2659,
+      "step": 144
+    },
+    {
+      "epoch": 0.15541264737406216,
+      "grad_norm": 2.724351167678833,
+      "learning_rate": 0.00016913183279742766,
+      "loss": 5.1795,
+      "step": 145
+    },
+    {
+      "epoch": 0.1564844587352626,
+      "grad_norm": 2.8223018646240234,
+      "learning_rate": 0.0001689174705251876,
+      "loss": 4.9197,
+      "step": 146
+    },
+    {
+      "epoch": 0.15755627009646303,
+      "grad_norm": 2.990795612335205,
+      "learning_rate": 0.0001687031082529475,
+      "loss": 4.7396,
+      "step": 147
+    },
+    {
+      "epoch": 0.15862808145766344,
+      "grad_norm": 4.808514595031738,
+      "learning_rate": 0.0001684887459807074,
+      "loss": 4.781,
+      "step": 148
+    },
+    {
+      "epoch": 0.15969989281886388,
+      "grad_norm": 2.9120051860809326,
+      "learning_rate": 0.0001682743837084673,
+      "loss": 4.8474,
+      "step": 149
+    },
+    {
+      "epoch": 0.1607717041800643,
+      "grad_norm": 3.2255208492279053,
+      "learning_rate": 0.00016806002143622724,
+      "loss": 4.883,
+      "step": 150
+    },
+    {
+      "epoch": 0.16184351554126475,
+      "grad_norm": 2.5771243572235107,
+      "learning_rate": 0.00016784565916398716,
+      "loss": 4.8337,
+      "step": 151
+    },
+    {
+      "epoch": 0.16291532690246516,
+      "grad_norm": 4.24485969543457,
+      "learning_rate": 0.00016763129689174706,
+      "loss": 5.3985,
+      "step": 152
+    },
+    {
+      "epoch": 0.1639871382636656,
+      "grad_norm": 2.5621776580810547,
+      "learning_rate": 0.000167416934619507,
+      "loss": 5.0145,
+      "step": 153
+    },
+    {
+      "epoch": 0.16505894962486603,
+      "grad_norm": 3.4797775745391846,
+      "learning_rate": 0.0001672025723472669,
+      "loss": 5.2251,
+      "step": 154
+    },
+    {
+      "epoch": 0.16613076098606644,
+      "grad_norm": 3.194912910461426,
+      "learning_rate": 0.0001669882100750268,
+      "loss": 4.6371,
+      "step": 155
+    },
+    {
+      "epoch": 0.16720257234726688,
+      "grad_norm": 3.799504280090332,
+      "learning_rate": 0.0001667738478027867,
+      "loss": 5.0769,
+      "step": 156
+    },
+    {
+      "epoch": 0.1682743837084673,
+      "grad_norm": 3.546942949295044,
+      "learning_rate": 0.00016655948553054664,
+      "loss": 4.6824,
+      "step": 157
+    },
+    {
+      "epoch": 0.16934619506966775,
+      "grad_norm": 3.28645658493042,
+      "learning_rate": 0.00016634512325830656,
+      "loss": 5.1034,
+      "step": 158
+    },
+    {
+      "epoch": 0.17041800643086816,
+      "grad_norm": 3.6776368618011475,
+      "learning_rate": 0.00016613076098606646,
+      "loss": 4.7307,
+      "step": 159
+    },
+    {
+      "epoch": 0.1714898177920686,
+      "grad_norm": 3.408773422241211,
+      "learning_rate": 0.0001659163987138264,
+      "loss": 5.1056,
+      "step": 160
+    },
+    {
+      "epoch": 0.17256162915326903,
+      "grad_norm": 2.734879732131958,
+      "learning_rate": 0.0001657020364415863,
+      "loss": 4.7715,
+      "step": 161
+    },
+    {
+      "epoch": 0.17363344051446947,
+      "grad_norm": 2.385500431060791,
+      "learning_rate": 0.0001654876741693462,
+      "loss": 4.869,
+      "step": 162
+    },
+    {
+      "epoch": 0.17470525187566988,
+      "grad_norm": 2.97015118598938,
+      "learning_rate": 0.0001652733118971061,
+      "loss": 4.7828,
+      "step": 163
+    },
+    {
+      "epoch": 0.1757770632368703,
+      "grad_norm": 5.304216384887695,
+      "learning_rate": 0.00016505894962486604,
+      "loss": 4.7406,
+      "step": 164
+    },
+    {
+      "epoch": 0.17684887459807075,
+      "grad_norm": 4.278355598449707,
+      "learning_rate": 0.00016484458735262596,
+      "loss": 5.1003,
+      "step": 165
+    },
+    {
+      "epoch": 0.17792068595927116,
+      "grad_norm": 3.021716356277466,
+      "learning_rate": 0.00016463022508038586,
+      "loss": 4.7373,
+      "step": 166
+    },
+    {
+      "epoch": 0.1789924973204716,
+      "grad_norm": 3.567626953125,
+      "learning_rate": 0.0001644158628081458,
+      "loss": 5.0243,
+      "step": 167
+    },
+    {
+      "epoch": 0.18006430868167203,
+      "grad_norm": 3.5427238941192627,
+      "learning_rate": 0.0001642015005359057,
+      "loss": 4.6865,
+      "step": 168
+    },
+    {
+      "epoch": 0.18113612004287247,
+      "grad_norm": 3.7450437545776367,
+      "learning_rate": 0.0001639871382636656,
+      "loss": 5.2324,
+      "step": 169
+    },
+    {
+      "epoch": 0.18220793140407288,
+      "grad_norm": 5.276814937591553,
+      "learning_rate": 0.0001637727759914255,
+      "loss": 5.0563,
+      "step": 170
+    },
+    {
+      "epoch": 0.1832797427652733,
+      "grad_norm": 5.184852123260498,
+      "learning_rate": 0.00016355841371918544,
+      "loss": 4.5753,
+      "step": 171
+    },
+    {
+      "epoch": 0.18435155412647375,
+      "grad_norm": 3.9674630165100098,
+      "learning_rate": 0.00016334405144694536,
+      "loss": 5.0476,
+      "step": 172
+    },
+    {
+      "epoch": 0.18542336548767416,
+      "grad_norm": 4.83624792098999,
+      "learning_rate": 0.00016312968917470526,
+      "loss": 4.8303,
+      "step": 173
+    },
+    {
+      "epoch": 0.1864951768488746,
+      "grad_norm": 2.8017940521240234,
+      "learning_rate": 0.0001629153269024652,
+      "loss": 4.7553,
+      "step": 174
+    },
+    {
+      "epoch": 0.18756698821007503,
+      "grad_norm": 3.8941309452056885,
+      "learning_rate": 0.0001627009646302251,
+      "loss": 4.5793,
+      "step": 175
+    },
+    {
+      "epoch": 0.18863879957127547,
+      "grad_norm": 2.9973654747009277,
+      "learning_rate": 0.000162486602357985,
+      "loss": 4.7673,
+      "step": 176
+    },
+    {
+      "epoch": 0.18971061093247588,
+      "grad_norm": 2.4675092697143555,
+      "learning_rate": 0.0001622722400857449,
+      "loss": 4.876,
+      "step": 177
+    },
+    {
+      "epoch": 0.1907824222936763,
+      "grad_norm": 4.017350196838379,
+      "learning_rate": 0.00016205787781350484,
+      "loss": 4.8945,
+      "step": 178
+    },
+    {
+      "epoch": 0.19185423365487675,
+      "grad_norm": 3.6734588146209717,
+      "learning_rate": 0.00016184351554126474,
+      "loss": 5.2249,
+      "step": 179
+    },
+    {
+      "epoch": 0.19292604501607716,
+      "grad_norm": 5.289995193481445,
+      "learning_rate": 0.00016162915326902466,
+      "loss": 4.7494,
+      "step": 180
+    },
+    {
+      "epoch": 0.1939978563772776,
+      "grad_norm": 3.5352132320404053,
+      "learning_rate": 0.0001614147909967846,
+      "loss": 4.9401,
+      "step": 181
+    },
+    {
+      "epoch": 0.19506966773847803,
+      "grad_norm": 3.0050971508026123,
+      "learning_rate": 0.0001612004287245445,
+      "loss": 4.9689,
+      "step": 182
+    },
+    {
+      "epoch": 0.19614147909967847,
+      "grad_norm": 2.5181117057800293,
+      "learning_rate": 0.0001609860664523044,
+      "loss": 4.8451,
+      "step": 183
+    },
+    {
+      "epoch": 0.19721329046087888,
+      "grad_norm": 4.061633110046387,
+      "learning_rate": 0.0001607717041800643,
+      "loss": 4.5857,
+      "step": 184
+    },
+    {
+      "epoch": 0.1982851018220793,
+      "grad_norm": 5.4610981941223145,
+      "learning_rate": 0.00016055734190782424,
+      "loss": 5.0163,
+      "step": 185
+    },
+    {
+      "epoch": 0.19935691318327975,
+      "grad_norm": 4.758736610412598,
+      "learning_rate": 0.00016034297963558414,
+      "loss": 4.5184,
+      "step": 186
+    },
+    {
+      "epoch": 0.20042872454448016,
+      "grad_norm": 3.6250007152557373,
+      "learning_rate": 0.00016012861736334406,
+      "loss": 4.9198,
+      "step": 187
+    },
+    {
+      "epoch": 0.2015005359056806,
+      "grad_norm": 2.938920736312866,
+      "learning_rate": 0.000159914255091104,
+      "loss": 4.6896,
+      "step": 188
+    },
+    {
+      "epoch": 0.20257234726688103,
+      "grad_norm": 3.388197660446167,
+      "learning_rate": 0.0001596998928188639,
+      "loss": 4.8161,
+      "step": 189
+    },
+    {
+      "epoch": 0.20364415862808147,
+      "grad_norm": 2.7313029766082764,
+      "learning_rate": 0.00015948553054662381,
+      "loss": 5.0972,
+      "step": 190
+    },
+    {
+      "epoch": 0.20471596998928188,
+      "grad_norm": 5.155452251434326,
+      "learning_rate": 0.0001592711682743837,
+      "loss": 4.9834,
+      "step": 191
+    },
+    {
+      "epoch": 0.2057877813504823,
+      "grad_norm": 3.554161310195923,
+      "learning_rate": 0.00015905680600214364,
+      "loss": 4.6315,
+      "step": 192
+    },
+    {
+      "epoch": 0.20685959271168275,
+      "grad_norm": 2.219275712966919,
+      "learning_rate": 0.00015884244372990354,
+      "loss": 4.8297,
+      "step": 193
+    },
+    {
+      "epoch": 0.20793140407288319,
+      "grad_norm": 3.4136171340942383,
+      "learning_rate": 0.00015862808145766346,
+      "loss": 4.7078,
+      "step": 194
+    },
+    {
+      "epoch": 0.2090032154340836,
+      "grad_norm": 3.1170814037323,
+      "learning_rate": 0.0001584137191854234,
+      "loss": 4.569,
+      "step": 195
+    },
+    {
+      "epoch": 0.21007502679528403,
+      "grad_norm": 4.157629489898682,
+      "learning_rate": 0.0001581993569131833,
+      "loss": 5.0884,
+      "step": 196
+    },
+    {
+      "epoch": 0.21114683815648447,
+      "grad_norm": 3.6698620319366455,
+      "learning_rate": 0.00015798499464094321,
+      "loss": 4.9795,
+      "step": 197
+    },
+    {
+      "epoch": 0.21221864951768488,
+      "grad_norm": 3.2543110847473145,
+      "learning_rate": 0.0001577706323687031,
+      "loss": 5.4088,
+      "step": 198
+    },
+    {
+      "epoch": 0.2132904608788853,
+      "grad_norm": 5.117916584014893,
+      "learning_rate": 0.00015755627009646304,
+      "loss": 5.0872,
+      "step": 199
+    },
+    {
+      "epoch": 0.21436227224008575,
+      "grad_norm": 2.6757586002349854,
+      "learning_rate": 0.00015734190782422294,
+      "loss": 5.0076,
+      "step": 200
+    },
+    {
+      "epoch": 0.21543408360128619,
+      "grad_norm": 4.662868499755859,
+      "learning_rate": 0.00015712754555198286,
+      "loss": 5.151,
+      "step": 201
+    },
+    {
+      "epoch": 0.2165058949624866,
+      "grad_norm": 2.8466174602508545,
+      "learning_rate": 0.0001569131832797428,
+      "loss": 4.7452,
+      "step": 202
+    },
+    {
+      "epoch": 0.21757770632368703,
+      "grad_norm": 3.703111410140991,
+      "learning_rate": 0.0001566988210075027,
+      "loss": 4.6917,
+      "step": 203
+    },
+    {
+      "epoch": 0.21864951768488747,
+      "grad_norm": 4.147573947906494,
+      "learning_rate": 0.00015648445873526261,
+      "loss": 4.5361,
+      "step": 204
+    },
+    {
+      "epoch": 0.21972132904608788,
+      "grad_norm": 3.213587760925293,
+      "learning_rate": 0.0001562700964630225,
+      "loss": 4.8234,
+      "step": 205
+    },
+    {
+      "epoch": 0.2207931404072883,
+      "grad_norm": 3.092233419418335,
+      "learning_rate": 0.00015605573419078244,
+      "loss": 4.5189,
+      "step": 206
+    },
+    {
+      "epoch": 0.22186495176848875,
+      "grad_norm": 3.0028648376464844,
+      "learning_rate": 0.00015584137191854234,
+      "loss": 5.1736,
+      "step": 207
+    },
+    {
+      "epoch": 0.22293676312968919,
+      "grad_norm": 4.2017998695373535,
+      "learning_rate": 0.00015562700964630226,
+      "loss": 4.9839,
+      "step": 208
+    },
+    {
+      "epoch": 0.2240085744908896,
+      "grad_norm": 2.643603801727295,
+      "learning_rate": 0.0001554126473740622,
+      "loss": 4.6254,
+      "step": 209
+    },
+    {
+      "epoch": 0.22508038585209003,
+      "grad_norm": 4.710816860198975,
+      "learning_rate": 0.0001551982851018221,
+      "loss": 5.0277,
+      "step": 210
+    },
+    {
+      "epoch": 0.22615219721329047,
+      "grad_norm": 2.4961905479431152,
+      "learning_rate": 0.00015498392282958201,
+      "loss": 4.691,
+      "step": 211
+    },
+    {
+      "epoch": 0.22722400857449088,
+      "grad_norm": 5.022608757019043,
+      "learning_rate": 0.0001547695605573419,
+      "loss": 4.7685,
+      "step": 212
+    },
+    {
+      "epoch": 0.2282958199356913,
+      "grad_norm": 3.2233705520629883,
+      "learning_rate": 0.00015455519828510184,
+      "loss": 4.7998,
+      "step": 213
+    },
+    {
+      "epoch": 0.22936763129689175,
+      "grad_norm": 2.526165246963501,
+      "learning_rate": 0.00015434083601286174,
+      "loss": 4.8508,
+      "step": 214
+    },
+    {
+      "epoch": 0.2304394426580922,
+      "grad_norm": 3.2895984649658203,
+      "learning_rate": 0.00015412647374062166,
+      "loss": 4.8211,
+      "step": 215
+    },
+    {
+      "epoch": 0.2315112540192926,
+      "grad_norm": 2.739210367202759,
+      "learning_rate": 0.0001539121114683816,
+      "loss": 5.1027,
+      "step": 216
+    },
+    {
+      "epoch": 0.23258306538049303,
+      "grad_norm": 3.5469467639923096,
+      "learning_rate": 0.0001536977491961415,
+      "loss": 5.2544,
+      "step": 217
+    },
+    {
+      "epoch": 0.23365487674169347,
+      "grad_norm": 2.3437864780426025,
+      "learning_rate": 0.00015348338692390141,
+      "loss": 4.8464,
+      "step": 218
+    },
+    {
+      "epoch": 0.2347266881028939,
+      "grad_norm": 2.8780081272125244,
+      "learning_rate": 0.0001532690246516613,
+      "loss": 4.9309,
+      "step": 219
+    },
+    {
+      "epoch": 0.2357984994640943,
+      "grad_norm": 2.508305788040161,
+      "learning_rate": 0.00015305466237942124,
+      "loss": 4.9276,
+      "step": 220
+    },
+    {
+      "epoch": 0.23687031082529475,
+      "grad_norm": 4.404898166656494,
+      "learning_rate": 0.00015284030010718114,
+      "loss": 4.6623,
+      "step": 221
+    },
+    {
+      "epoch": 0.2379421221864952,
+      "grad_norm": 4.61767578125,
+      "learning_rate": 0.00015262593783494106,
+      "loss": 5.06,
+      "step": 222
+    },
+    {
+      "epoch": 0.2390139335476956,
+      "grad_norm": 4.487622261047363,
+      "learning_rate": 0.000152411575562701,
+      "loss": 5.5357,
+      "step": 223
+    },
+    {
+      "epoch": 0.24008574490889603,
+      "grad_norm": 2.2761528491973877,
+      "learning_rate": 0.0001521972132904609,
+      "loss": 4.7513,
+      "step": 224
+    },
+    {
+      "epoch": 0.24115755627009647,
+      "grad_norm": 3.565575122833252,
+      "learning_rate": 0.00015198285101822081,
+      "loss": 4.4589,
+      "step": 225
+    },
+    {
+      "epoch": 0.2422293676312969,
+      "grad_norm": 3.7335009574890137,
+      "learning_rate": 0.0001517684887459807,
+      "loss": 4.5894,
+      "step": 226
+    },
+    {
+      "epoch": 0.2433011789924973,
+      "grad_norm": 3.367142677307129,
+      "learning_rate": 0.00015155412647374064,
+      "loss": 4.58,
+      "step": 227
+    },
+    {
+      "epoch": 0.24437299035369775,
+      "grad_norm": 2.965182304382324,
+      "learning_rate": 0.00015133976420150054,
+      "loss": 4.9648,
+      "step": 228
+    },
+    {
+      "epoch": 0.2454448017148982,
+      "grad_norm": 2.646235704421997,
+      "learning_rate": 0.00015112540192926046,
+      "loss": 4.8253,
+      "step": 229
+    },
+    {
+      "epoch": 0.2465166130760986,
+      "grad_norm": 3.499410629272461,
+      "learning_rate": 0.0001509110396570204,
+      "loss": 5.3223,
+      "step": 230
+    },
+    {
+      "epoch": 0.24758842443729903,
+      "grad_norm": 4.3262104988098145,
+      "learning_rate": 0.0001506966773847803,
+      "loss": 4.8132,
+      "step": 231
+    },
+    {
+      "epoch": 0.24866023579849947,
+      "grad_norm": 3.103876829147339,
+      "learning_rate": 0.00015048231511254021,
+      "loss": 4.9858,
+      "step": 232
+    },
+    {
+      "epoch": 0.2497320471596999,
+      "grad_norm": 3.5739259719848633,
+      "learning_rate": 0.0001502679528403001,
+      "loss": 4.8506,
+      "step": 233
+    },
+    {
+      "epoch": 0.2508038585209003,
+      "grad_norm": 2.523289918899536,
+      "learning_rate": 0.00015005359056806004,
+      "loss": 4.8562,
+      "step": 234
+    },
+    {
+      "epoch": 0.25187566988210075,
+      "grad_norm": 3.0936124324798584,
+      "learning_rate": 0.00014983922829581994,
+      "loss": 4.8166,
+      "step": 235
+    },
+    {
+      "epoch": 0.2529474812433012,
+      "grad_norm": 2.942096710205078,
+      "learning_rate": 0.00014962486602357986,
+      "loss": 4.7598,
+      "step": 236
+    },
+    {
+      "epoch": 0.2540192926045016,
+      "grad_norm": 3.589401960372925,
+      "learning_rate": 0.0001494105037513398,
+      "loss": 4.6942,
+      "step": 237
+    },
+    {
+      "epoch": 0.25509110396570206,
+      "grad_norm": 3.2308168411254883,
+      "learning_rate": 0.0001491961414790997,
+      "loss": 4.8895,
+      "step": 238
+    },
+    {
+      "epoch": 0.25616291532690244,
+      "grad_norm": 3.224642515182495,
+      "learning_rate": 0.00014898177920685961,
+      "loss": 4.7678,
+      "step": 239
+    },
+    {
+      "epoch": 0.2572347266881029,
+      "grad_norm": 3.0363709926605225,
+      "learning_rate": 0.0001487674169346195,
+      "loss": 4.7585,
+      "step": 240
+    },
+    {
+      "epoch": 0.2583065380493033,
+      "grad_norm": 3.49025821685791,
+      "learning_rate": 0.00014855305466237944,
+      "loss": 4.7028,
+      "step": 241
+    },
+    {
+      "epoch": 0.25937834941050375,
+      "grad_norm": 4.927128314971924,
+      "learning_rate": 0.00014833869239013934,
+      "loss": 4.8932,
+      "step": 242
+    },
+    {
+      "epoch": 0.2604501607717042,
+      "grad_norm": 3.036620855331421,
+      "learning_rate": 0.00014812433011789926,
+      "loss": 4.7612,
+      "step": 243
+    },
+    {
+      "epoch": 0.2615219721329046,
+      "grad_norm": 3.0013017654418945,
+      "learning_rate": 0.0001479099678456592,
+      "loss": 4.701,
+      "step": 244
+    },
+    {
+      "epoch": 0.26259378349410506,
+      "grad_norm": 2.943427324295044,
+      "learning_rate": 0.0001476956055734191,
+      "loss": 4.7791,
+      "step": 245
+    },
+    {
+      "epoch": 0.26366559485530544,
+      "grad_norm": 3.3308539390563965,
+      "learning_rate": 0.00014748124330117901,
+      "loss": 4.9904,
+      "step": 246
+    },
+    {
+      "epoch": 0.2647374062165059,
+      "grad_norm": 3.7442448139190674,
+      "learning_rate": 0.00014726688102893891,
+      "loss": 5.9101,
+      "step": 247
+    },
+    {
+      "epoch": 0.2658092175777063,
+      "grad_norm": 3.206761598587036,
+      "learning_rate": 0.00014705251875669884,
+      "loss": 5.0534,
+      "step": 248
+    },
+    {
+      "epoch": 0.26688102893890675,
+      "grad_norm": 2.8394346237182617,
+      "learning_rate": 0.00014683815648445874,
+      "loss": 4.8381,
+      "step": 249
+    },
+    {
+      "epoch": 0.2679528403001072,
+      "grad_norm": 3.0866284370422363,
+      "learning_rate": 0.00014662379421221866,
+      "loss": 4.9118,
+      "step": 250
+    },
+    {
+      "epoch": 0.2690246516613076,
+      "grad_norm": 3.6339967250823975,
+      "learning_rate": 0.0001464094319399786,
+      "loss": 5.2333,
+      "step": 251
+    },
+    {
+      "epoch": 0.27009646302250806,
+      "grad_norm": 3.3214287757873535,
+      "learning_rate": 0.0001461950696677385,
+      "loss": 4.8577,
+      "step": 252
+    },
+    {
+      "epoch": 0.27116827438370844,
+      "grad_norm": 3.002842903137207,
+      "learning_rate": 0.00014598070739549841,
+      "loss": 4.811,
+      "step": 253
+    },
+    {
+      "epoch": 0.2722400857449089,
+      "grad_norm": 3.2833826541900635,
+      "learning_rate": 0.00014576634512325831,
+      "loss": 4.7479,
+      "step": 254
+    },
+    {
+      "epoch": 0.2733118971061093,
+      "grad_norm": 4.537484645843506,
+      "learning_rate": 0.00014555198285101824,
+      "loss": 4.7998,
+      "step": 255
+    },
+    {
+      "epoch": 0.27438370846730975,
+      "grad_norm": 2.9750728607177734,
+      "learning_rate": 0.00014533762057877814,
+      "loss": 4.8613,
+      "step": 256
+    },
+    {
+      "epoch": 0.2754555198285102,
+      "grad_norm": 2.4807565212249756,
+      "learning_rate": 0.00014512325830653806,
+      "loss": 4.6727,
+      "step": 257
+    },
+    {
+      "epoch": 0.2765273311897106,
+      "grad_norm": 2.890812397003174,
+      "learning_rate": 0.000144908896034298,
+      "loss": 4.8248,
+      "step": 258
+    },
+    {
+      "epoch": 0.27759914255091106,
+      "grad_norm": 2.885880947113037,
+      "learning_rate": 0.0001446945337620579,
+      "loss": 4.7678,
+      "step": 259
+    },
+    {
+      "epoch": 0.27867095391211144,
+      "grad_norm": 3.302708387374878,
+      "learning_rate": 0.00014448017148981782,
+      "loss": 4.709,
+      "step": 260
+    },
+    {
+      "epoch": 0.2797427652733119,
+      "grad_norm": 3.09729266166687,
+      "learning_rate": 0.00014426580921757771,
+      "loss": 4.6741,
+      "step": 261
+    },
+    {
+      "epoch": 0.2808145766345123,
+      "grad_norm": 3.538935899734497,
+      "learning_rate": 0.00014405144694533764,
+      "loss": 4.698,
+      "step": 262
+    },
+    {
+      "epoch": 0.28188638799571275,
+      "grad_norm": 2.7330965995788574,
+      "learning_rate": 0.00014383708467309754,
+      "loss": 4.934,
+      "step": 263
+    },
+    {
+      "epoch": 0.2829581993569132,
+      "grad_norm": 3.0136098861694336,
+      "learning_rate": 0.00014362272240085746,
+      "loss": 4.6817,
+      "step": 264
+    },
+    {
+      "epoch": 0.2840300107181136,
+      "grad_norm": 2.376910448074341,
+      "learning_rate": 0.0001434083601286174,
+      "loss": 4.4482,
+      "step": 265
+    },
+    {
+      "epoch": 0.28510182207931406,
+      "grad_norm": 3.1064581871032715,
+      "learning_rate": 0.0001431939978563773,
+      "loss": 4.6759,
+      "step": 266
+    },
+    {
+      "epoch": 0.2861736334405145,
+      "grad_norm": 3.8472907543182373,
+      "learning_rate": 0.00014297963558413722,
+      "loss": 5.2014,
+      "step": 267
+    },
+    {
+      "epoch": 0.2872454448017149,
+      "grad_norm": 3.489697217941284,
+      "learning_rate": 0.00014276527331189711,
+      "loss": 4.9149,
+      "step": 268
+    },
+    {
+      "epoch": 0.2883172561629153,
+      "grad_norm": 4.324831485748291,
+      "learning_rate": 0.00014255091103965704,
+      "loss": 4.7237,
+      "step": 269
+    },
+    {
+      "epoch": 0.28938906752411575,
+      "grad_norm": 2.6454925537109375,
+      "learning_rate": 0.00014233654876741694,
+      "loss": 4.5799,
+      "step": 270
+    },
+    {
+      "epoch": 0.2904608788853162,
+      "grad_norm": 2.733677387237549,
+      "learning_rate": 0.00014212218649517686,
+      "loss": 5.3109,
+      "step": 271
+    },
+    {
+      "epoch": 0.2915326902465166,
+      "grad_norm": 3.2953786849975586,
+      "learning_rate": 0.0001419078242229368,
+      "loss": 4.7647,
+      "step": 272
+    },
+    {
+      "epoch": 0.29260450160771706,
+      "grad_norm": 3.716773748397827,
+      "learning_rate": 0.0001416934619506967,
+      "loss": 4.7028,
+      "step": 273
+    },
+    {
+      "epoch": 0.2936763129689175,
+      "grad_norm": 4.2017502784729,
+      "learning_rate": 0.00014147909967845662,
+      "loss": 5.0879,
+      "step": 274
+    },
+    {
+      "epoch": 0.2947481243301179,
+      "grad_norm": 3.396883487701416,
+      "learning_rate": 0.00014126473740621651,
+      "loss": 4.5388,
+      "step": 275
+    },
+    {
+      "epoch": 0.2958199356913183,
+      "grad_norm": 4.1809306144714355,
+      "learning_rate": 0.00014105037513397644,
+      "loss": 4.5289,
+      "step": 276
+    },
+    {
+      "epoch": 0.29689174705251875,
+      "grad_norm": 2.7928273677825928,
+      "learning_rate": 0.00014083601286173634,
+      "loss": 4.4894,
+      "step": 277
+    },
+    {
+      "epoch": 0.2979635584137192,
+      "grad_norm": 3.4506866931915283,
+      "learning_rate": 0.00014062165058949626,
+      "loss": 4.7572,
+      "step": 278
+    },
+    {
+      "epoch": 0.2990353697749196,
+      "grad_norm": 2.7902727127075195,
+      "learning_rate": 0.0001404072883172562,
+      "loss": 5.0598,
+      "step": 279
+    },
+    {
+      "epoch": 0.30010718113612006,
+      "grad_norm": 4.7700676918029785,
+      "learning_rate": 0.0001401929260450161,
+      "loss": 4.8324,
+      "step": 280
+    },
+    {
+      "epoch": 0.3011789924973205,
+      "grad_norm": 2.7876083850860596,
+      "learning_rate": 0.00013997856377277602,
+      "loss": 4.7838,
+      "step": 281
+    },
+    {
+      "epoch": 0.3022508038585209,
+      "grad_norm": 3.0817983150482178,
+      "learning_rate": 0.00013976420150053591,
+      "loss": 4.7157,
+      "step": 282
+    },
+    {
+      "epoch": 0.3033226152197213,
+      "grad_norm": 3.8207461833953857,
+      "learning_rate": 0.00013954983922829584,
+      "loss": 4.7388,
+      "step": 283
+    },
+    {
+      "epoch": 0.30439442658092175,
+      "grad_norm": 2.6404857635498047,
+      "learning_rate": 0.00013933547695605574,
+      "loss": 4.65,
+      "step": 284
+    },
+    {
+      "epoch": 0.3054662379421222,
+      "grad_norm": 4.702475070953369,
+      "learning_rate": 0.00013912111468381567,
+      "loss": 5.0798,
+      "step": 285
+    },
+    {
+      "epoch": 0.3065380493033226,
+      "grad_norm": 3.7465784549713135,
+      "learning_rate": 0.0001389067524115756,
+      "loss": 4.5744,
+      "step": 286
+    },
+    {
+      "epoch": 0.30760986066452306,
+      "grad_norm": 3.9282851219177246,
+      "learning_rate": 0.0001386923901393355,
+      "loss": 4.8443,
+      "step": 287
+    },
+    {
+      "epoch": 0.3086816720257235,
+      "grad_norm": 4.132015705108643,
+      "learning_rate": 0.00013847802786709542,
+      "loss": 4.682,
+      "step": 288
+    },
+    {
+      "epoch": 0.3097534833869239,
+      "grad_norm": 2.495121479034424,
+      "learning_rate": 0.00013826366559485531,
+      "loss": 4.6311,
+      "step": 289
+    },
+    {
+      "epoch": 0.3108252947481243,
+      "grad_norm": 3.6948375701904297,
+      "learning_rate": 0.00013804930332261524,
+      "loss": 4.5608,
+      "step": 290
+    },
+    {
+      "epoch": 0.31189710610932475,
+      "grad_norm": 3.2895307540893555,
+      "learning_rate": 0.00013783494105037514,
+      "loss": 4.6748,
+      "step": 291
+    },
+    {
+      "epoch": 0.3129689174705252,
+      "grad_norm": 5.540443420410156,
+      "learning_rate": 0.00013762057877813507,
+      "loss": 4.8687,
+      "step": 292
+    },
+    {
+      "epoch": 0.3140407288317256,
+      "grad_norm": 3.6029367446899414,
+      "learning_rate": 0.00013740621650589496,
+      "loss": 5.135,
+      "step": 293
+    },
+    {
+      "epoch": 0.31511254019292606,
+      "grad_norm": 3.0030689239501953,
+      "learning_rate": 0.0001371918542336549,
+      "loss": 4.5854,
+      "step": 294
+    },
+    {
+      "epoch": 0.3161843515541265,
+      "grad_norm": 2.6009085178375244,
+      "learning_rate": 0.00013697749196141482,
+      "loss": 4.7002,
+      "step": 295
+    },
+    {
+      "epoch": 0.3172561629153269,
+      "grad_norm": 4.169888973236084,
+      "learning_rate": 0.00013676312968917471,
+      "loss": 4.8343,
+      "step": 296
+    },
+    {
+      "epoch": 0.3183279742765273,
+      "grad_norm": 2.8053345680236816,
+      "learning_rate": 0.00013654876741693464,
+      "loss": 4.7164,
+      "step": 297
+    },
+    {
+      "epoch": 0.31939978563772775,
+      "grad_norm": 3.506838321685791,
+      "learning_rate": 0.00013633440514469454,
+      "loss": 4.8534,
+      "step": 298
+    },
+    {
+      "epoch": 0.3204715969989282,
+      "grad_norm": 4.393333911895752,
+      "learning_rate": 0.00013612004287245447,
+      "loss": 4.4928,
+      "step": 299
+    },
+    {
+      "epoch": 0.3215434083601286,
+      "grad_norm": 2.863898277282715,
+      "learning_rate": 0.00013590568060021436,
+      "loss": 4.5477,
+      "step": 300
+    },
+    {
+      "epoch": 0.32261521972132906,
+      "grad_norm": 3.9683938026428223,
+      "learning_rate": 0.0001356913183279743,
+      "loss": 4.7035,
+      "step": 301
+    },
+    {
+      "epoch": 0.3236870310825295,
+      "grad_norm": 3.109039783477783,
+      "learning_rate": 0.00013547695605573422,
+      "loss": 4.6411,
+      "step": 302
+    },
+    {
+      "epoch": 0.3247588424437299,
+      "grad_norm": 3.720078706741333,
+      "learning_rate": 0.00013526259378349411,
+      "loss": 4.9106,
+      "step": 303
+    },
+    {
+      "epoch": 0.3258306538049303,
+      "grad_norm": 3.9917848110198975,
+      "learning_rate": 0.00013504823151125404,
+      "loss": 5.4525,
+      "step": 304
+    },
+    {
+      "epoch": 0.32690246516613075,
+      "grad_norm": 2.6513423919677734,
+      "learning_rate": 0.00013483386923901394,
+      "loss": 4.686,
+      "step": 305
+    },
+    {
+      "epoch": 0.3279742765273312,
+      "grad_norm": 2.915001392364502,
+      "learning_rate": 0.00013461950696677387,
+      "loss": 4.7752,
+      "step": 306
+    },
+    {
+      "epoch": 0.3290460878885316,
+      "grad_norm": 2.989955186843872,
+      "learning_rate": 0.00013440514469453376,
+      "loss": 4.76,
+      "step": 307
+    },
+    {
+      "epoch": 0.33011789924973206,
+      "grad_norm": 5.258363246917725,
+      "learning_rate": 0.0001341907824222937,
+      "loss": 4.7727,
+      "step": 308
+    },
+    {
+      "epoch": 0.3311897106109325,
+      "grad_norm": 2.8469583988189697,
+      "learning_rate": 0.00013397642015005362,
+      "loss": 4.6681,
+      "step": 309
+    },
+    {
+      "epoch": 0.3322615219721329,
+      "grad_norm": 2.9092135429382324,
+      "learning_rate": 0.00013376205787781351,
+      "loss": 4.9489,
+      "step": 310
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.8508243560791016,
+      "learning_rate": 0.00013354769560557344,
+      "loss": 4.8356,
+      "step": 311
+    },
+    {
+      "epoch": 0.33440514469453375,
+      "grad_norm": 2.808962345123291,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 4.6664,
+      "step": 312
+    },
+    {
+      "epoch": 0.3354769560557342,
+      "grad_norm": 3.5812997817993164,
+      "learning_rate": 0.00013311897106109327,
+      "loss": 4.4658,
+      "step": 313
+    },
+    {
+      "epoch": 0.3365487674169346,
+      "grad_norm": 2.3912558555603027,
+      "learning_rate": 0.00013290460878885316,
+      "loss": 4.5262,
+      "step": 314
+    },
+    {
+      "epoch": 0.33762057877813506,
+      "grad_norm": 3.29030704498291,
+      "learning_rate": 0.0001326902465166131,
+      "loss": 4.8285,
+      "step": 315
+    },
+    {
+      "epoch": 0.3386923901393355,
+      "grad_norm": 2.764319658279419,
+      "learning_rate": 0.00013247588424437302,
+      "loss": 4.6389,
+      "step": 316
+    },
+    {
+      "epoch": 0.3397642015005359,
+      "grad_norm": 3.782653331756592,
+      "learning_rate": 0.00013226152197213292,
+      "loss": 4.9516,
+      "step": 317
+    },
+    {
+      "epoch": 0.3408360128617363,
+      "grad_norm": 2.377980947494507,
+      "learning_rate": 0.00013204715969989284,
+      "loss": 4.5456,
+      "step": 318
+    },
+    {
+      "epoch": 0.34190782422293675,
+      "grad_norm": 2.9316773414611816,
+      "learning_rate": 0.00013183279742765274,
+      "loss": 4.4321,
+      "step": 319
+    },
+    {
+      "epoch": 0.3429796355841372,
+      "grad_norm": 3.867056369781494,
+      "learning_rate": 0.00013161843515541267,
+      "loss": 4.7064,
+      "step": 320
+    },
+    {
+      "epoch": 0.3440514469453376,
+      "grad_norm": 2.9708735942840576,
+      "learning_rate": 0.00013140407288317256,
+      "loss": 4.8706,
+      "step": 321
+    },
+    {
+      "epoch": 0.34512325830653806,
+      "grad_norm": 2.9509382247924805,
+      "learning_rate": 0.0001311897106109325,
+      "loss": 5.0689,
+      "step": 322
+    },
+    {
+      "epoch": 0.3461950696677385,
+      "grad_norm": 3.441537857055664,
+      "learning_rate": 0.00013097534833869242,
+      "loss": 5.1403,
+      "step": 323
+    },
+    {
+      "epoch": 0.34726688102893893,
+      "grad_norm": 2.2192604541778564,
+      "learning_rate": 0.00013076098606645232,
+      "loss": 4.8289,
+      "step": 324
+    },
+    {
+      "epoch": 0.3483386923901393,
+      "grad_norm": 6.160877704620361,
+      "learning_rate": 0.00013054662379421224,
+      "loss": 4.6294,
+      "step": 325
+    },
+    {
+      "epoch": 0.34941050375133975,
+      "grad_norm": 2.7882578372955322,
+      "learning_rate": 0.00013033226152197214,
+      "loss": 5.1393,
+      "step": 326
+    },
+    {
+      "epoch": 0.3504823151125402,
+      "grad_norm": 2.55655574798584,
+      "learning_rate": 0.00013011789924973207,
+      "loss": 5.307,
+      "step": 327
+    },
+    {
+      "epoch": 0.3515541264737406,
+      "grad_norm": 4.0760817527771,
+      "learning_rate": 0.00012990353697749196,
+      "loss": 4.4461,
+      "step": 328
+    },
+    {
+      "epoch": 0.35262593783494106,
+      "grad_norm": 2.735928773880005,
+      "learning_rate": 0.0001296891747052519,
+      "loss": 4.4893,
+      "step": 329
+    },
+    {
+      "epoch": 0.3536977491961415,
+      "grad_norm": 3.1315958499908447,
+      "learning_rate": 0.00012947481243301182,
+      "loss": 4.6153,
+      "step": 330
+    },
+    {
+      "epoch": 0.35476956055734193,
+      "grad_norm": 3.7019879817962646,
+      "learning_rate": 0.00012926045016077172,
+      "loss": 4.545,
+      "step": 331
+    },
+    {
+      "epoch": 0.3558413719185423,
+      "grad_norm": 3.004509449005127,
+      "learning_rate": 0.00012904608788853164,
+      "loss": 4.9725,
+      "step": 332
+    },
+    {
+      "epoch": 0.35691318327974275,
+      "grad_norm": 2.3004274368286133,
+      "learning_rate": 0.00012883172561629154,
+      "loss": 4.7199,
+      "step": 333
+    },
+    {
+      "epoch": 0.3579849946409432,
+      "grad_norm": 2.899890184402466,
+      "learning_rate": 0.00012861736334405147,
+      "loss": 4.7919,
+      "step": 334
+    },
+    {
+      "epoch": 0.3590568060021436,
+      "grad_norm": 2.68511962890625,
+      "learning_rate": 0.00012840300107181136,
+      "loss": 4.5754,
+      "step": 335
+    },
+    {
+      "epoch": 0.36012861736334406,
+      "grad_norm": 2.743640899658203,
+      "learning_rate": 0.0001281886387995713,
+      "loss": 4.4467,
+      "step": 336
+    },
+    {
+      "epoch": 0.3612004287245445,
+      "grad_norm": 3.8968470096588135,
+      "learning_rate": 0.00012797427652733122,
+      "loss": 4.8004,
+      "step": 337
+    },
+    {
+      "epoch": 0.36227224008574493,
+      "grad_norm": 2.5243613719940186,
+      "learning_rate": 0.00012775991425509112,
+      "loss": 4.7446,
+      "step": 338
+    },
+    {
+      "epoch": 0.3633440514469453,
+      "grad_norm": 3.5175392627716064,
+      "learning_rate": 0.00012754555198285104,
+      "loss": 4.7064,
+      "step": 339
+    },
+    {
+      "epoch": 0.36441586280814575,
+      "grad_norm": 4.9440155029296875,
+      "learning_rate": 0.00012733118971061094,
+      "loss": 4.729,
+      "step": 340
+    },
+    {
+      "epoch": 0.3654876741693462,
+      "grad_norm": 2.3877315521240234,
+      "learning_rate": 0.00012711682743837087,
+      "loss": 4.9724,
+      "step": 341
+    },
+    {
+      "epoch": 0.3665594855305466,
+      "grad_norm": 2.969658851623535,
+      "learning_rate": 0.00012690246516613077,
+      "loss": 4.7385,
+      "step": 342
+    },
+    {
+      "epoch": 0.36763129689174706,
+      "grad_norm": 3.0826351642608643,
+      "learning_rate": 0.0001266881028938907,
+      "loss": 4.5219,
+      "step": 343
+    },
+    {
+      "epoch": 0.3687031082529475,
+      "grad_norm": 3.592517614364624,
+      "learning_rate": 0.00012647374062165062,
+      "loss": 4.917,
+      "step": 344
+    },
+    {
+      "epoch": 0.36977491961414793,
+      "grad_norm": 2.00095272064209,
+      "learning_rate": 0.00012625937834941052,
+      "loss": 4.7813,
+      "step": 345
+    },
+    {
+      "epoch": 0.3708467309753483,
+      "grad_norm": 3.643803119659424,
+      "learning_rate": 0.00012604501607717044,
+      "loss": 4.7526,
+      "step": 346
+    },
+    {
+      "epoch": 0.37191854233654875,
+      "grad_norm": 2.889460563659668,
+      "learning_rate": 0.00012583065380493034,
+      "loss": 4.804,
+      "step": 347
+    },
+    {
+      "epoch": 0.3729903536977492,
+      "grad_norm": 3.5942623615264893,
+      "learning_rate": 0.00012561629153269027,
+      "loss": 4.418,
+      "step": 348
+    },
+    {
+      "epoch": 0.3740621650589496,
+      "grad_norm": 4.21300745010376,
+      "learning_rate": 0.00012540192926045017,
+      "loss": 4.4909,
+      "step": 349
+    },
+    {
+      "epoch": 0.37513397642015006,
+      "grad_norm": 3.0777478218078613,
+      "learning_rate": 0.0001251875669882101,
+      "loss": 4.4939,
+      "step": 350
+    },
+    {
+      "epoch": 0.3762057877813505,
+      "grad_norm": 2.5130343437194824,
+      "learning_rate": 0.00012497320471597002,
+      "loss": 4.6874,
+      "step": 351
+    },
+    {
+      "epoch": 0.37727759914255093,
+      "grad_norm": 4.369728088378906,
+      "learning_rate": 0.00012475884244372992,
+      "loss": 4.4141,
+      "step": 352
+    },
+    {
+      "epoch": 0.3783494105037513,
+      "grad_norm": 2.8714654445648193,
+      "learning_rate": 0.00012454448017148984,
+      "loss": 4.7199,
+      "step": 353
+    },
+    {
+      "epoch": 0.37942122186495175,
+      "grad_norm": 5.214938163757324,
+      "learning_rate": 0.00012433011789924974,
+      "loss": 4.6146,
+      "step": 354
+    },
+    {
+      "epoch": 0.3804930332261522,
+      "grad_norm": 4.095846176147461,
+      "learning_rate": 0.00012411575562700967,
+      "loss": 4.8992,
+      "step": 355
+    },
+    {
+      "epoch": 0.3815648445873526,
+      "grad_norm": 3.715656042098999,
+      "learning_rate": 0.00012390139335476957,
+      "loss": 4.8978,
+      "step": 356
+    },
+    {
+      "epoch": 0.38263665594855306,
+      "grad_norm": 2.896740198135376,
+      "learning_rate": 0.0001236870310825295,
+      "loss": 4.5217,
+      "step": 357
+    },
+    {
+      "epoch": 0.3837084673097535,
+      "grad_norm": 3.072086811065674,
+      "learning_rate": 0.00012347266881028942,
+      "loss": 4.8439,
+      "step": 358
+    },
+    {
+      "epoch": 0.38478027867095393,
+      "grad_norm": 2.8563904762268066,
+      "learning_rate": 0.00012325830653804932,
+      "loss": 4.8879,
+      "step": 359
+    },
+    {
+      "epoch": 0.3858520900321543,
+      "grad_norm": 2.688250780105591,
+      "learning_rate": 0.00012304394426580924,
+      "loss": 4.472,
+      "step": 360
+    },
+    {
+      "epoch": 0.38692390139335475,
+      "grad_norm": 4.014735698699951,
+      "learning_rate": 0.00012282958199356914,
+      "loss": 5.0148,
+      "step": 361
+    },
+    {
+      "epoch": 0.3879957127545552,
+      "grad_norm": 3.6055967807769775,
+      "learning_rate": 0.00012261521972132907,
+      "loss": 4.4383,
+      "step": 362
+    },
+    {
+      "epoch": 0.3890675241157556,
+      "grad_norm": 2.7861740589141846,
+      "learning_rate": 0.00012240085744908897,
+      "loss": 4.5756,
+      "step": 363
+    },
+    {
+      "epoch": 0.39013933547695606,
+      "grad_norm": 2.6457951068878174,
+      "learning_rate": 0.0001221864951768489,
+      "loss": 4.9015,
+      "step": 364
+    },
+    {
+      "epoch": 0.3912111468381565,
+      "grad_norm": 2.503458023071289,
+      "learning_rate": 0.0001219721329046088,
+      "loss": 4.6681,
+      "step": 365
+    },
+    {
+      "epoch": 0.39228295819935693,
+      "grad_norm": 3.545325517654419,
+      "learning_rate": 0.0001217577706323687,
+      "loss": 4.7556,
+      "step": 366
+    },
+    {
+      "epoch": 0.3933547695605573,
+      "grad_norm": 2.9182682037353516,
+      "learning_rate": 0.00012154340836012863,
+      "loss": 4.7351,
+      "step": 367
+    },
+    {
+      "epoch": 0.39442658092175775,
+      "grad_norm": 3.4011483192443848,
+      "learning_rate": 0.00012132904608788853,
+      "loss": 4.3977,
+      "step": 368
+    },
+    {
+      "epoch": 0.3954983922829582,
+      "grad_norm": 3.131936550140381,
+      "learning_rate": 0.00012111468381564845,
+      "loss": 4.53,
+      "step": 369
+    },
+    {
+      "epoch": 0.3965702036441586,
+      "grad_norm": 2.649606704711914,
+      "learning_rate": 0.00012090032154340835,
+      "loss": 4.5794,
+      "step": 370
+    },
+    {
+      "epoch": 0.39764201500535906,
+      "grad_norm": 3.4589664936065674,
+      "learning_rate": 0.00012068595927116828,
+      "loss": 5.2707,
+      "step": 371
+    },
+    {
+      "epoch": 0.3987138263665595,
+      "grad_norm": 2.252598524093628,
+      "learning_rate": 0.0001204715969989282,
+      "loss": 4.6805,
+      "step": 372
+    },
+    {
+      "epoch": 0.39978563772775993,
+      "grad_norm": 3.345087766647339,
+      "learning_rate": 0.0001202572347266881,
+      "loss": 4.725,
+      "step": 373
+    },
+    {
+      "epoch": 0.4008574490889603,
+      "grad_norm": 2.961495876312256,
+      "learning_rate": 0.00012004287245444803,
+      "loss": 4.5135,
+      "step": 374
+    },
+    {
+      "epoch": 0.40192926045016075,
+      "grad_norm": 2.2345147132873535,
+      "learning_rate": 0.00011982851018220793,
+      "loss": 4.4681,
+      "step": 375
+    },
+    {
+      "epoch": 0.4030010718113612,
+      "grad_norm": 2.904963731765747,
+      "learning_rate": 0.00011961414790996785,
+      "loss": 5.0236,
+      "step": 376
+    },
+    {
+      "epoch": 0.4040728831725616,
+      "grad_norm": 4.671694755554199,
+      "learning_rate": 0.00011939978563772775,
+      "loss": 5.2864,
+      "step": 377
+    },
+    {
+      "epoch": 0.40514469453376206,
+      "grad_norm": 4.501243591308594,
+      "learning_rate": 0.00011918542336548768,
+      "loss": 4.7508,
+      "step": 378
+    },
+    {
+      "epoch": 0.4062165058949625,
+      "grad_norm": 2.389822244644165,
+      "learning_rate": 0.0001189710610932476,
+      "loss": 4.8598,
+      "step": 379
+    },
+    {
+      "epoch": 0.40728831725616294,
+      "grad_norm": 2.789280891418457,
+      "learning_rate": 0.0001187566988210075,
+      "loss": 4.5269,
+      "step": 380
+    },
+    {
+      "epoch": 0.40836012861736337,
+      "grad_norm": 2.0258848667144775,
+      "learning_rate": 0.00011854233654876743,
+      "loss": 4.6753,
+      "step": 381
+    },
+    {
+      "epoch": 0.40943193997856375,
+      "grad_norm": 2.4621434211730957,
+      "learning_rate": 0.00011832797427652733,
+      "loss": 4.5195,
+      "step": 382
+    },
+    {
+      "epoch": 0.4105037513397642,
+      "grad_norm": 2.21687650680542,
+      "learning_rate": 0.00011811361200428725,
+      "loss": 4.5579,
+      "step": 383
+    },
+    {
+      "epoch": 0.4115755627009646,
+      "grad_norm": 2.6150362491607666,
+      "learning_rate": 0.00011789924973204715,
+      "loss": 5.0116,
+      "step": 384
+    },
+    {
+      "epoch": 0.41264737406216506,
+      "grad_norm": 2.5843632221221924,
+      "learning_rate": 0.00011768488745980708,
+      "loss": 4.6433,
+      "step": 385
+    },
+    {
+      "epoch": 0.4137191854233655,
+      "grad_norm": 3.124844551086426,
+      "learning_rate": 0.000117470525187567,
+      "loss": 4.6199,
+      "step": 386
+    },
+    {
+      "epoch": 0.41479099678456594,
+      "grad_norm": 2.8832359313964844,
+      "learning_rate": 0.0001172561629153269,
+      "loss": 4.646,
+      "step": 387
+    },
+    {
+      "epoch": 0.41586280814576637,
+      "grad_norm": 2.497830629348755,
+      "learning_rate": 0.00011704180064308683,
+      "loss": 4.431,
+      "step": 388
+    },
+    {
+      "epoch": 0.41693461950696675,
+      "grad_norm": 2.51007342338562,
+      "learning_rate": 0.00011682743837084673,
+      "loss": 4.8004,
+      "step": 389
+    },
+    {
+      "epoch": 0.4180064308681672,
+      "grad_norm": 3.0995190143585205,
+      "learning_rate": 0.00011661307609860665,
+      "loss": 5.0293,
+      "step": 390
+    },
+    {
+      "epoch": 0.4190782422293676,
+      "grad_norm": 3.0592472553253174,
+      "learning_rate": 0.00011639871382636655,
+      "loss": 4.6023,
+      "step": 391
+    },
+    {
+      "epoch": 0.42015005359056806,
+      "grad_norm": 2.6731066703796387,
+      "learning_rate": 0.00011618435155412648,
+      "loss": 4.5261,
+      "step": 392
+    },
+    {
+      "epoch": 0.4212218649517685,
+      "grad_norm": 2.5948925018310547,
+      "learning_rate": 0.0001159699892818864,
+      "loss": 4.737,
+      "step": 393
+    },
+    {
+      "epoch": 0.42229367631296894,
+      "grad_norm": 4.319209575653076,
+      "learning_rate": 0.0001157556270096463,
+      "loss": 4.7966,
+      "step": 394
+    },
+    {
+      "epoch": 0.42336548767416937,
+      "grad_norm": 2.590317964553833,
+      "learning_rate": 0.00011554126473740623,
+      "loss": 4.6858,
+      "step": 395
+    },
+    {
+      "epoch": 0.42443729903536975,
+      "grad_norm": 2.41375732421875,
+      "learning_rate": 0.00011532690246516613,
+      "loss": 4.3515,
+      "step": 396
+    },
+    {
+      "epoch": 0.4255091103965702,
+      "grad_norm": 2.681856632232666,
+      "learning_rate": 0.00011511254019292605,
+      "loss": 5.0034,
+      "step": 397
+    },
+    {
+      "epoch": 0.4265809217577706,
+      "grad_norm": 2.9265639781951904,
+      "learning_rate": 0.00011489817792068595,
+      "loss": 4.6002,
+      "step": 398
+    },
+    {
+      "epoch": 0.42765273311897106,
+      "grad_norm": 3.013899326324463,
+      "learning_rate": 0.00011468381564844588,
+      "loss": 4.3798,
+      "step": 399
+    },
+    {
+      "epoch": 0.4287245444801715,
+      "grad_norm": 3.1742773056030273,
+      "learning_rate": 0.0001144694533762058,
+      "loss": 4.7765,
+      "step": 400
+    },
+    {
+      "epoch": 0.42979635584137194,
+      "grad_norm": 2.3816874027252197,
+      "learning_rate": 0.0001142550911039657,
+      "loss": 4.5805,
+      "step": 401
+    },
+    {
+      "epoch": 0.43086816720257237,
+      "grad_norm": 5.512743949890137,
+      "learning_rate": 0.00011404072883172563,
+      "loss": 4.8546,
+      "step": 402
+    },
+    {
+      "epoch": 0.43193997856377275,
+      "grad_norm": 3.751079559326172,
+      "learning_rate": 0.00011382636655948553,
+      "loss": 4.5984,
+      "step": 403
+    },
+    {
+      "epoch": 0.4330117899249732,
+      "grad_norm": 3.0676698684692383,
+      "learning_rate": 0.00011361200428724545,
+      "loss": 5.0614,
+      "step": 404
+    },
+    {
+      "epoch": 0.4340836012861736,
+      "grad_norm": 3.396777391433716,
+      "learning_rate": 0.00011339764201500535,
+      "loss": 4.7951,
+      "step": 405
+    },
+    {
+      "epoch": 0.43515541264737406,
+      "grad_norm": 2.4765775203704834,
+      "learning_rate": 0.00011318327974276528,
+      "loss": 5.1756,
+      "step": 406
+    },
+    {
+      "epoch": 0.4362272240085745,
+      "grad_norm": 2.3405649662017822,
+      "learning_rate": 0.0001129689174705252,
+      "loss": 4.5914,
+      "step": 407
+    },
+    {
+      "epoch": 0.43729903536977494,
+      "grad_norm": 3.434269666671753,
+      "learning_rate": 0.0001127545551982851,
+      "loss": 4.4127,
+      "step": 408
+    },
+    {
+      "epoch": 0.43837084673097537,
+      "grad_norm": 2.127887487411499,
+      "learning_rate": 0.00011254019292604503,
+      "loss": 4.6347,
+      "step": 409
+    },
+    {
+      "epoch": 0.43944265809217575,
+      "grad_norm": 4.66085147857666,
+      "learning_rate": 0.00011232583065380493,
+      "loss": 4.9721,
+      "step": 410
+    },
+    {
+      "epoch": 0.4405144694533762,
+      "grad_norm": 4.892344951629639,
+      "learning_rate": 0.00011211146838156485,
+      "loss": 4.5199,
+      "step": 411
+    },
+    {
+      "epoch": 0.4415862808145766,
+      "grad_norm": 3.442005157470703,
+      "learning_rate": 0.00011189710610932475,
+      "loss": 4.502,
+      "step": 412
+    },
+    {
+      "epoch": 0.44265809217577706,
+      "grad_norm": 3.6790285110473633,
+      "learning_rate": 0.00011168274383708468,
+      "loss": 4.2922,
+      "step": 413
+    },
+    {
+      "epoch": 0.4437299035369775,
+      "grad_norm": 4.072107315063477,
+      "learning_rate": 0.00011146838156484458,
+      "loss": 4.5492,
+      "step": 414
+    },
+    {
+      "epoch": 0.44480171489817794,
+      "grad_norm": 2.8195745944976807,
+      "learning_rate": 0.0001112540192926045,
+      "loss": 4.4757,
+      "step": 415
+    },
+    {
+      "epoch": 0.44587352625937837,
+      "grad_norm": 3.7252724170684814,
+      "learning_rate": 0.00011103965702036443,
+      "loss": 4.7089,
+      "step": 416
+    },
+    {
+      "epoch": 0.44694533762057875,
+      "grad_norm": 2.841998338699341,
+      "learning_rate": 0.00011082529474812433,
+      "loss": 4.4443,
+      "step": 417
+    },
+    {
+      "epoch": 0.4480171489817792,
+      "grad_norm": 2.653102159500122,
+      "learning_rate": 0.00011061093247588425,
+      "loss": 4.2913,
+      "step": 418
+    },
+    {
+      "epoch": 0.4490889603429796,
+      "grad_norm": 2.2500383853912354,
+      "learning_rate": 0.00011039657020364415,
+      "loss": 4.8685,
+      "step": 419
+    },
+    {
+      "epoch": 0.45016077170418006,
+      "grad_norm": 3.34576416015625,
+      "learning_rate": 0.00011018220793140408,
+      "loss": 4.9446,
+      "step": 420
+    },
+    {
+      "epoch": 0.4512325830653805,
+      "grad_norm": 5.558518409729004,
+      "learning_rate": 0.00010996784565916398,
+      "loss": 4.6652,
+      "step": 421
+    },
+    {
+      "epoch": 0.45230439442658094,
+      "grad_norm": 2.774711847305298,
+      "learning_rate": 0.0001097534833869239,
+      "loss": 4.5453,
+      "step": 422
+    },
+    {
+      "epoch": 0.4533762057877814,
+      "grad_norm": 2.1220977306365967,
+      "learning_rate": 0.00010953912111468383,
+      "loss": 4.4978,
+      "step": 423
+    },
+    {
+      "epoch": 0.45444801714898175,
+      "grad_norm": 2.5744822025299072,
+      "learning_rate": 0.00010932475884244373,
+      "loss": 4.6972,
+      "step": 424
+    },
+    {
+      "epoch": 0.4555198285101822,
+      "grad_norm": 3.827157974243164,
+      "learning_rate": 0.00010911039657020365,
+      "loss": 4.6225,
+      "step": 425
+    },
+    {
+      "epoch": 0.4565916398713826,
+      "grad_norm": 4.045970439910889,
+      "learning_rate": 0.00010889603429796355,
+      "loss": 4.7059,
+      "step": 426
+    },
+    {
+      "epoch": 0.45766345123258306,
+      "grad_norm": 3.225592851638794,
+      "learning_rate": 0.00010868167202572348,
+      "loss": 4.8028,
+      "step": 427
+    },
+    {
+      "epoch": 0.4587352625937835,
+      "grad_norm": 2.285249948501587,
+      "learning_rate": 0.00010846730975348338,
+      "loss": 4.621,
+      "step": 428
+    },
+    {
+      "epoch": 0.45980707395498394,
+      "grad_norm": 2.890104293823242,
+      "learning_rate": 0.0001082529474812433,
+      "loss": 4.7726,
+      "step": 429
+    },
+    {
+      "epoch": 0.4608788853161844,
+      "grad_norm": 3.505073070526123,
+      "learning_rate": 0.00010803858520900323,
+      "loss": 4.6411,
+      "step": 430
+    },
+    {
+      "epoch": 0.46195069667738475,
+      "grad_norm": 3.7843589782714844,
+      "learning_rate": 0.00010782422293676313,
+      "loss": 4.6693,
+      "step": 431
+    },
+    {
+      "epoch": 0.4630225080385852,
+      "grad_norm": 4.503906726837158,
+      "learning_rate": 0.00010760986066452305,
+      "loss": 4.891,
+      "step": 432
+    },
+    {
+      "epoch": 0.4640943193997856,
+      "grad_norm": 2.121623992919922,
+      "learning_rate": 0.00010739549839228295,
+      "loss": 4.401,
+      "step": 433
+    },
+    {
+      "epoch": 0.46516613076098606,
+      "grad_norm": 3.2020905017852783,
+      "learning_rate": 0.00010718113612004288,
+      "loss": 4.976,
+      "step": 434
+    },
+    {
+      "epoch": 0.4662379421221865,
+      "grad_norm": 2.401608467102051,
+      "learning_rate": 0.00010696677384780278,
+      "loss": 4.5996,
+      "step": 435
+    },
+    {
+      "epoch": 0.46730975348338694,
+      "grad_norm": 4.06643009185791,
+      "learning_rate": 0.0001067524115755627,
+      "loss": 4.9013,
+      "step": 436
+    },
+    {
+      "epoch": 0.4683815648445874,
+      "grad_norm": 2.9890921115875244,
+      "learning_rate": 0.00010653804930332263,
+      "loss": 4.5683,
+      "step": 437
+    },
+    {
+      "epoch": 0.4694533762057878,
+      "grad_norm": 3.724813222885132,
+      "learning_rate": 0.00010632368703108253,
+      "loss": 4.635,
+      "step": 438
+    },
+    {
+      "epoch": 0.4705251875669882,
+      "grad_norm": 4.079103946685791,
+      "learning_rate": 0.00010610932475884245,
+      "loss": 3.6894,
+      "step": 439
+    },
+    {
+      "epoch": 0.4715969989281886,
+      "grad_norm": 3.2234132289886475,
+      "learning_rate": 0.00010589496248660235,
+      "loss": 4.6768,
+      "step": 440
+    },
+    {
+      "epoch": 0.47266881028938906,
+      "grad_norm": 4.110087871551514,
+      "learning_rate": 0.00010568060021436228,
+      "loss": 4.7946,
+      "step": 441
+    },
+    {
+      "epoch": 0.4737406216505895,
+      "grad_norm": 3.6127841472625732,
+      "learning_rate": 0.00010546623794212218,
+      "loss": 4.3852,
+      "step": 442
+    },
+    {
+      "epoch": 0.47481243301178994,
+      "grad_norm": 2.7365081310272217,
+      "learning_rate": 0.0001052518756698821,
+      "loss": 4.7021,
+      "step": 443
+    },
+    {
+      "epoch": 0.4758842443729904,
+      "grad_norm": 2.423485517501831,
+      "learning_rate": 0.00010503751339764203,
+      "loss": 4.5577,
+      "step": 444
+    },
+    {
+      "epoch": 0.4769560557341908,
+      "grad_norm": 4.876895427703857,
+      "learning_rate": 0.00010482315112540193,
+      "loss": 4.4222,
+      "step": 445
+    },
+    {
+      "epoch": 0.4780278670953912,
+      "grad_norm": 4.026849269866943,
+      "learning_rate": 0.00010460878885316185,
+      "loss": 4.7229,
+      "step": 446
+    },
+    {
+      "epoch": 0.4790996784565916,
+      "grad_norm": 3.8580574989318848,
+      "learning_rate": 0.00010439442658092175,
+      "loss": 4.3579,
+      "step": 447
+    },
+    {
+      "epoch": 0.48017148981779206,
+      "grad_norm": 2.972254991531372,
+      "learning_rate": 0.00010418006430868168,
+      "loss": 4.6865,
+      "step": 448
+    },
+    {
+      "epoch": 0.4812433011789925,
+      "grad_norm": 3.2446720600128174,
+      "learning_rate": 0.00010396570203644158,
+      "loss": 5.3246,
+      "step": 449
+    },
+    {
+      "epoch": 0.48231511254019294,
+      "grad_norm": 2.7309093475341797,
+      "learning_rate": 0.0001037513397642015,
+      "loss": 4.5838,
+      "step": 450
+    },
+    {
+      "epoch": 0.4833869239013934,
+      "grad_norm": 2.745063543319702,
+      "learning_rate": 0.00010353697749196143,
+      "loss": 4.802,
+      "step": 451
+    },
+    {
+      "epoch": 0.4844587352625938,
+      "grad_norm": 2.725165367126465,
+      "learning_rate": 0.00010332261521972133,
+      "loss": 4.439,
+      "step": 452
+    },
+    {
+      "epoch": 0.4855305466237942,
+      "grad_norm": 2.230207681655884,
+      "learning_rate": 0.00010310825294748126,
+      "loss": 4.5734,
+      "step": 453
+    },
+    {
+      "epoch": 0.4866023579849946,
+      "grad_norm": 5.293651580810547,
+      "learning_rate": 0.00010289389067524115,
+      "loss": 4.4032,
+      "step": 454
+    },
+    {
+      "epoch": 0.48767416934619506,
+      "grad_norm": 3.427778720855713,
+      "learning_rate": 0.00010267952840300108,
+      "loss": 5.0263,
+      "step": 455
+    },
+    {
+      "epoch": 0.4887459807073955,
+      "grad_norm": 3.034630060195923,
+      "learning_rate": 0.00010246516613076098,
+      "loss": 4.8377,
+      "step": 456
+    },
+    {
+      "epoch": 0.48981779206859594,
+      "grad_norm": 4.529390335083008,
+      "learning_rate": 0.0001022508038585209,
+      "loss": 4.6534,
+      "step": 457
+    },
+    {
+      "epoch": 0.4908896034297964,
+      "grad_norm": 3.547393798828125,
+      "learning_rate": 0.00010203644158628083,
+      "loss": 4.5927,
+      "step": 458
+    },
+    {
+      "epoch": 0.4919614147909968,
+      "grad_norm": 2.7677557468414307,
+      "learning_rate": 0.00010182207931404073,
+      "loss": 4.6297,
+      "step": 459
+    },
+    {
+      "epoch": 0.4930332261521972,
+      "grad_norm": 3.251615285873413,
+      "learning_rate": 0.00010160771704180066,
+      "loss": 4.636,
+      "step": 460
+    },
+    {
+      "epoch": 0.4941050375133976,
+      "grad_norm": 3.7128498554229736,
+      "learning_rate": 0.00010139335476956055,
+      "loss": 4.793,
+      "step": 461
+    },
+    {
+      "epoch": 0.49517684887459806,
+      "grad_norm": 3.7845911979675293,
+      "learning_rate": 0.00010117899249732048,
+      "loss": 4.9675,
+      "step": 462
+    },
+    {
+      "epoch": 0.4962486602357985,
+      "grad_norm": 2.348583698272705,
+      "learning_rate": 0.00010096463022508038,
+      "loss": 4.6437,
+      "step": 463
+    },
+    {
+      "epoch": 0.49732047159699894,
+      "grad_norm": 3.05892014503479,
+      "learning_rate": 0.0001007502679528403,
+      "loss": 4.5614,
+      "step": 464
+    },
+    {
+      "epoch": 0.4983922829581994,
+      "grad_norm": 2.5227441787719727,
+      "learning_rate": 0.00010053590568060023,
+      "loss": 4.7875,
+      "step": 465
+    },
+    {
+      "epoch": 0.4994640943193998,
+      "grad_norm": 2.151101589202881,
+      "learning_rate": 0.00010032154340836013,
+      "loss": 4.431,
+      "step": 466
+    },
+    {
+      "epoch": 0.5005359056806002,
+      "grad_norm": 2.6528310775756836,
+      "learning_rate": 0.00010010718113612006,
+      "loss": 4.6138,
+      "step": 467
+    },
+    {
+      "epoch": 0.5016077170418006,
+      "grad_norm": 3.231062889099121,
+      "learning_rate": 9.989281886387997e-05,
+      "loss": 4.9219,
+      "step": 468
+    },
+    {
+      "epoch": 0.5026795284030011,
+      "grad_norm": 2.9534308910369873,
+      "learning_rate": 9.967845659163988e-05,
+      "loss": 4.6625,
+      "step": 469
+    },
+    {
+      "epoch": 0.5037513397642015,
+      "grad_norm": 5.39685583114624,
+      "learning_rate": 9.946409431939979e-05,
+      "loss": 4.6287,
+      "step": 470
+    },
+    {
+      "epoch": 0.5048231511254019,
+      "grad_norm": 2.1971487998962402,
+      "learning_rate": 9.92497320471597e-05,
+      "loss": 4.7021,
+      "step": 471
+    },
+    {
+      "epoch": 0.5058949624866024,
+      "grad_norm": 4.07705020904541,
+      "learning_rate": 9.903536977491962e-05,
+      "loss": 4.3861,
+      "step": 472
+    },
+    {
+      "epoch": 0.5069667738478028,
+      "grad_norm": 2.779970407485962,
+      "learning_rate": 9.882100750267953e-05,
+      "loss": 4.8171,
+      "step": 473
+    },
+    {
+      "epoch": 0.5080385852090032,
+      "grad_norm": 3.3830039501190186,
+      "learning_rate": 9.860664523043944e-05,
+      "loss": 4.6153,
+      "step": 474
+    },
+    {
+      "epoch": 0.5091103965702036,
+      "grad_norm": 2.822718858718872,
+      "learning_rate": 9.839228295819937e-05,
+      "loss": 4.7932,
+      "step": 475
+    },
+    {
+      "epoch": 0.5101822079314041,
+      "grad_norm": 2.344116687774658,
+      "learning_rate": 9.817792068595928e-05,
+      "loss": 4.6028,
+      "step": 476
+    },
+    {
+      "epoch": 0.5112540192926045,
+      "grad_norm": 4.1745381355285645,
+      "learning_rate": 9.796355841371919e-05,
+      "loss": 4.2435,
+      "step": 477
+    },
+    {
+      "epoch": 0.5123258306538049,
+      "grad_norm": 3.8554046154022217,
+      "learning_rate": 9.77491961414791e-05,
+      "loss": 4.6214,
+      "step": 478
+    },
+    {
+      "epoch": 0.5133976420150054,
+      "grad_norm": 3.1669418811798096,
+      "learning_rate": 9.753483386923902e-05,
+      "loss": 4.4191,
+      "step": 479
+    },
+    {
+      "epoch": 0.5144694533762058,
+      "grad_norm": 2.7627713680267334,
+      "learning_rate": 9.732047159699893e-05,
+      "loss": 4.577,
+      "step": 480
+    },
+    {
+      "epoch": 0.5155412647374062,
+      "grad_norm": 3.044712543487549,
+      "learning_rate": 9.710610932475884e-05,
+      "loss": 4.5155,
+      "step": 481
+    },
+    {
+      "epoch": 0.5166130760986066,
+      "grad_norm": 4.0238356590271,
+      "learning_rate": 9.689174705251877e-05,
+      "loss": 5.1078,
+      "step": 482
+    },
+    {
+      "epoch": 0.5176848874598071,
+      "grad_norm": 2.3172390460968018,
+      "learning_rate": 9.667738478027868e-05,
+      "loss": 4.4444,
+      "step": 483
+    },
+    {
+      "epoch": 0.5187566988210075,
+      "grad_norm": 2.782829999923706,
+      "learning_rate": 9.646302250803859e-05,
+      "loss": 4.3296,
+      "step": 484
+    },
+    {
+      "epoch": 0.5198285101822079,
+      "grad_norm": 2.173595666885376,
+      "learning_rate": 9.62486602357985e-05,
+      "loss": 4.2437,
+      "step": 485
+    },
+    {
+      "epoch": 0.5209003215434084,
+      "grad_norm": 2.6314404010772705,
+      "learning_rate": 9.603429796355842e-05,
+      "loss": 4.4386,
+      "step": 486
+    },
+    {
+      "epoch": 0.5219721329046088,
+      "grad_norm": 2.2896389961242676,
+      "learning_rate": 9.581993569131833e-05,
+      "loss": 4.4891,
+      "step": 487
+    },
+    {
+      "epoch": 0.5230439442658092,
+      "grad_norm": 2.4675259590148926,
+      "learning_rate": 9.560557341907824e-05,
+      "loss": 4.765,
+      "step": 488
+    },
+    {
+      "epoch": 0.5241157556270096,
+      "grad_norm": 3.0715389251708984,
+      "learning_rate": 9.539121114683817e-05,
+      "loss": 4.7717,
+      "step": 489
+    },
+    {
+      "epoch": 0.5251875669882101,
+      "grad_norm": 4.1732635498046875,
+      "learning_rate": 9.517684887459808e-05,
+      "loss": 4.4018,
+      "step": 490
+    },
+    {
+      "epoch": 0.5262593783494105,
+      "grad_norm": 3.266986608505249,
+      "learning_rate": 9.496248660235799e-05,
+      "loss": 4.2295,
+      "step": 491
+    },
+    {
+      "epoch": 0.5273311897106109,
+      "grad_norm": 2.5479917526245117,
+      "learning_rate": 9.47481243301179e-05,
+      "loss": 4.4826,
+      "step": 492
+    },
+    {
+      "epoch": 0.5284030010718114,
+      "grad_norm": 2.021381139755249,
+      "learning_rate": 9.453376205787782e-05,
+      "loss": 4.4129,
+      "step": 493
+    },
+    {
+      "epoch": 0.5294748124330118,
+      "grad_norm": 2.8881123065948486,
+      "learning_rate": 9.431939978563773e-05,
+      "loss": 4.6945,
+      "step": 494
+    },
+    {
+      "epoch": 0.5305466237942122,
+      "grad_norm": 2.44140362739563,
+      "learning_rate": 9.410503751339764e-05,
+      "loss": 4.4454,
+      "step": 495
+    },
+    {
+      "epoch": 0.5316184351554126,
+      "grad_norm": 2.6777150630950928,
+      "learning_rate": 9.389067524115757e-05,
+      "loss": 5.0997,
+      "step": 496
+    },
+    {
+      "epoch": 0.5326902465166131,
+      "grad_norm": 3.453155040740967,
+      "learning_rate": 9.367631296891748e-05,
+      "loss": 4.4479,
+      "step": 497
+    },
+    {
+      "epoch": 0.5337620578778135,
+      "grad_norm": 3.0275356769561768,
+      "learning_rate": 9.346195069667739e-05,
+      "loss": 4.9407,
+      "step": 498
+    },
+    {
+      "epoch": 0.5348338692390139,
+      "grad_norm": 3.3887171745300293,
+      "learning_rate": 9.32475884244373e-05,
+      "loss": 4.8086,
+      "step": 499
+    },
+    {
+      "epoch": 0.5359056806002144,
+      "grad_norm": 3.7157983779907227,
+      "learning_rate": 9.303322615219722e-05,
+      "loss": 4.7977,
+      "step": 500
+    },
+    {
+      "epoch": 0.5369774919614148,
+      "grad_norm": 4.366313457489014,
+      "learning_rate": 9.281886387995713e-05,
+      "loss": 4.9259,
+      "step": 501
+    },
+    {
+      "epoch": 0.5380493033226152,
+      "grad_norm": 2.8501381874084473,
+      "learning_rate": 9.260450160771704e-05,
+      "loss": 4.6289,
+      "step": 502
+    },
+    {
+      "epoch": 0.5391211146838156,
+      "grad_norm": 3.3382182121276855,
+      "learning_rate": 9.239013933547695e-05,
+      "loss": 4.8087,
+      "step": 503
+    },
+    {
+      "epoch": 0.5401929260450161,
+      "grad_norm": 3.3544347286224365,
+      "learning_rate": 9.217577706323688e-05,
+      "loss": 4.515,
+      "step": 504
+    },
+    {
+      "epoch": 0.5412647374062165,
+      "grad_norm": 37.70632553100586,
+      "learning_rate": 9.19614147909968e-05,
+      "loss": 5.0745,
+      "step": 505
+    },
+    {
+      "epoch": 0.5423365487674169,
+      "grad_norm": 2.2493515014648438,
+      "learning_rate": 9.17470525187567e-05,
+      "loss": 4.4932,
+      "step": 506
+    },
+    {
+      "epoch": 0.5434083601286174,
+      "grad_norm": 2.044198751449585,
+      "learning_rate": 9.153269024651662e-05,
+      "loss": 4.4252,
+      "step": 507
+    },
+    {
+      "epoch": 0.5444801714898178,
+      "grad_norm": 3.8044934272766113,
+      "learning_rate": 9.131832797427653e-05,
+      "loss": 4.5577,
+      "step": 508
+    },
+    {
+      "epoch": 0.5455519828510182,
+      "grad_norm": 2.5661163330078125,
+      "learning_rate": 9.110396570203644e-05,
+      "loss": 4.5289,
+      "step": 509
+    },
+    {
+      "epoch": 0.5466237942122186,
+      "grad_norm": 3.2366676330566406,
+      "learning_rate": 9.088960342979636e-05,
+      "loss": 4.5276,
+      "step": 510
+    },
+    {
+      "epoch": 0.5476956055734191,
+      "grad_norm": 2.113185405731201,
+      "learning_rate": 9.067524115755628e-05,
+      "loss": 4.4132,
+      "step": 511
+    },
+    {
+      "epoch": 0.5487674169346195,
+      "grad_norm": 2.1703245639801025,
+      "learning_rate": 9.04608788853162e-05,
+      "loss": 4.4325,
+      "step": 512
+    },
+    {
+      "epoch": 0.5498392282958199,
+      "grad_norm": 2.1970648765563965,
+      "learning_rate": 9.02465166130761e-05,
+      "loss": 4.5207,
+      "step": 513
+    },
+    {
+      "epoch": 0.5509110396570204,
+      "grad_norm": 1.997298240661621,
+      "learning_rate": 9.003215434083602e-05,
+      "loss": 4.4903,
+      "step": 514
+    },
+    {
+      "epoch": 0.5519828510182208,
+      "grad_norm": 2.3199117183685303,
+      "learning_rate": 8.981779206859593e-05,
+      "loss": 4.5107,
+      "step": 515
+    },
+    {
+      "epoch": 0.5530546623794212,
+      "grad_norm": 3.4195363521575928,
+      "learning_rate": 8.960342979635584e-05,
+      "loss": 4.62,
+      "step": 516
+    },
+    {
+      "epoch": 0.5541264737406216,
+      "grad_norm": 4.756778717041016,
+      "learning_rate": 8.938906752411576e-05,
+      "loss": 4.443,
+      "step": 517
+    },
+    {
+      "epoch": 0.5551982851018221,
+      "grad_norm": 2.4822423458099365,
+      "learning_rate": 8.917470525187568e-05,
+      "loss": 4.5553,
+      "step": 518
+    },
+    {
+      "epoch": 0.5562700964630225,
+      "grad_norm": 3.304189682006836,
+      "learning_rate": 8.89603429796356e-05,
+      "loss": 4.6001,
+      "step": 519
+    },
+    {
+      "epoch": 0.5573419078242229,
+      "grad_norm": 2.1849889755249023,
+      "learning_rate": 8.87459807073955e-05,
+      "loss": 4.3251,
+      "step": 520
+    },
+    {
+      "epoch": 0.5584137191854234,
+      "grad_norm": 2.0505974292755127,
+      "learning_rate": 8.853161843515542e-05,
+      "loss": 4.8688,
+      "step": 521
+    },
+    {
+      "epoch": 0.5594855305466238,
+      "grad_norm": 3.737577438354492,
+      "learning_rate": 8.831725616291533e-05,
+      "loss": 4.417,
+      "step": 522
+    },
+    {
+      "epoch": 0.5605573419078242,
+      "grad_norm": 3.136176586151123,
+      "learning_rate": 8.810289389067524e-05,
+      "loss": 4.6331,
+      "step": 523
+    },
+    {
+      "epoch": 0.5616291532690246,
+      "grad_norm": 4.006575107574463,
+      "learning_rate": 8.788853161843516e-05,
+      "loss": 4.084,
+      "step": 524
+    },
+    {
+      "epoch": 0.5627009646302251,
+      "grad_norm": 4.377976417541504,
+      "learning_rate": 8.767416934619508e-05,
+      "loss": 4.6561,
+      "step": 525
+    },
+    {
+      "epoch": 0.5637727759914255,
+      "grad_norm": 5.064045429229736,
+      "learning_rate": 8.7459807073955e-05,
+      "loss": 4.7988,
+      "step": 526
+    },
+    {
+      "epoch": 0.564844587352626,
+      "grad_norm": 2.700467348098755,
+      "learning_rate": 8.72454448017149e-05,
+      "loss": 4.9471,
+      "step": 527
+    },
+    {
+      "epoch": 0.5659163987138264,
+      "grad_norm": 2.9747073650360107,
+      "learning_rate": 8.703108252947482e-05,
+      "loss": 4.7476,
+      "step": 528
+    },
+    {
+      "epoch": 0.5669882100750268,
+      "grad_norm": 2.418670415878296,
+      "learning_rate": 8.681672025723473e-05,
+      "loss": 4.5664,
+      "step": 529
+    },
+    {
+      "epoch": 0.5680600214362272,
+      "grad_norm": 3.5818655490875244,
+      "learning_rate": 8.660235798499464e-05,
+      "loss": 4.6175,
+      "step": 530
+    },
+    {
+      "epoch": 0.5691318327974276,
+      "grad_norm": 2.4983315467834473,
+      "learning_rate": 8.638799571275456e-05,
+      "loss": 4.5348,
+      "step": 531
+    },
+    {
+      "epoch": 0.5702036441586281,
+      "grad_norm": 2.1915688514709473,
+      "learning_rate": 8.617363344051448e-05,
+      "loss": 4.297,
+      "step": 532
+    },
+    {
+      "epoch": 0.5712754555198285,
+      "grad_norm": 2.926217794418335,
+      "learning_rate": 8.59592711682744e-05,
+      "loss": 4.77,
+      "step": 533
+    },
+    {
+      "epoch": 0.572347266881029,
+      "grad_norm": 2.369635820388794,
+      "learning_rate": 8.57449088960343e-05,
+      "loss": 4.712,
+      "step": 534
+    },
+    {
+      "epoch": 0.5734190782422294,
+      "grad_norm": 2.3241968154907227,
+      "learning_rate": 8.553054662379422e-05,
+      "loss": 4.3899,
+      "step": 535
+    },
+    {
+      "epoch": 0.5744908896034298,
+      "grad_norm": 4.746558666229248,
+      "learning_rate": 8.531618435155413e-05,
+      "loss": 4.476,
+      "step": 536
+    },
+    {
+      "epoch": 0.5755627009646302,
+      "grad_norm": 2.7210347652435303,
+      "learning_rate": 8.510182207931404e-05,
+      "loss": 4.3049,
+      "step": 537
+    },
+    {
+      "epoch": 0.5766345123258306,
+      "grad_norm": 3.503418207168579,
+      "learning_rate": 8.488745980707396e-05,
+      "loss": 4.5433,
+      "step": 538
+    },
+    {
+      "epoch": 0.5777063236870311,
+      "grad_norm": 5.652376651763916,
+      "learning_rate": 8.467309753483388e-05,
+      "loss": 4.9893,
+      "step": 539
+    },
+    {
+      "epoch": 0.5787781350482315,
+      "grad_norm": 2.5520389080047607,
+      "learning_rate": 8.44587352625938e-05,
+      "loss": 4.4338,
+      "step": 540
+    },
+    {
+      "epoch": 0.579849946409432,
+      "grad_norm": 3.322006940841675,
+      "learning_rate": 8.42443729903537e-05,
+      "loss": 4.7979,
+      "step": 541
+    },
+    {
+      "epoch": 0.5809217577706324,
+      "grad_norm": 3.558786153793335,
+      "learning_rate": 8.403001071811362e-05,
+      "loss": 4.5882,
+      "step": 542
+    },
+    {
+      "epoch": 0.5819935691318328,
+      "grad_norm": 2.970303773880005,
+      "learning_rate": 8.381564844587353e-05,
+      "loss": 4.4237,
+      "step": 543
+    },
+    {
+      "epoch": 0.5830653804930332,
+      "grad_norm": 2.5710015296936035,
+      "learning_rate": 8.360128617363344e-05,
+      "loss": 4.5529,
+      "step": 544
+    },
+    {
+      "epoch": 0.5841371918542336,
+      "grad_norm": 2.937779664993286,
+      "learning_rate": 8.338692390139336e-05,
+      "loss": 4.7481,
+      "step": 545
+    },
+    {
+      "epoch": 0.5852090032154341,
+      "grad_norm": 2.8150634765625,
+      "learning_rate": 8.317256162915328e-05,
+      "loss": 4.4981,
+      "step": 546
+    },
+    {
+      "epoch": 0.5862808145766345,
+      "grad_norm": 3.412611961364746,
+      "learning_rate": 8.29581993569132e-05,
+      "loss": 4.7558,
+      "step": 547
+    },
+    {
+      "epoch": 0.587352625937835,
+      "grad_norm": 4.309683322906494,
+      "learning_rate": 8.27438370846731e-05,
+      "loss": 4.4244,
+      "step": 548
+    },
+    {
+      "epoch": 0.5884244372990354,
+      "grad_norm": 3.520347833633423,
+      "learning_rate": 8.252947481243302e-05,
+      "loss": 4.4844,
+      "step": 549
+    },
+    {
+      "epoch": 0.5894962486602358,
+      "grad_norm": 3.5350539684295654,
+      "learning_rate": 8.231511254019293e-05,
+      "loss": 5.268,
+      "step": 550
+    },
+    {
+      "epoch": 0.5905680600214362,
+      "grad_norm": 3.6474995613098145,
+      "learning_rate": 8.210075026795284e-05,
+      "loss": 4.3715,
+      "step": 551
+    },
+    {
+      "epoch": 0.5916398713826366,
+      "grad_norm": 2.8854405879974365,
+      "learning_rate": 8.188638799571276e-05,
+      "loss": 4.398,
+      "step": 552
+    },
+    {
+      "epoch": 0.5927116827438371,
+      "grad_norm": 3.763253688812256,
+      "learning_rate": 8.167202572347268e-05,
+      "loss": 4.4518,
+      "step": 553
+    },
+    {
+      "epoch": 0.5937834941050375,
+      "grad_norm": 4.865472316741943,
+      "learning_rate": 8.14576634512326e-05,
+      "loss": 4.1518,
+      "step": 554
+    },
+    {
+      "epoch": 0.594855305466238,
+      "grad_norm": 4.091231822967529,
+      "learning_rate": 8.12433011789925e-05,
+      "loss": 4.786,
+      "step": 555
+    },
+    {
+      "epoch": 0.5959271168274384,
+      "grad_norm": 2.232417106628418,
+      "learning_rate": 8.102893890675242e-05,
+      "loss": 4.2969,
+      "step": 556
+    },
+    {
+      "epoch": 0.5969989281886388,
+      "grad_norm": 2.0849270820617676,
+      "learning_rate": 8.081457663451233e-05,
+      "loss": 4.7915,
+      "step": 557
+    },
+    {
+      "epoch": 0.5980707395498392,
+      "grad_norm": 3.684378147125244,
+      "learning_rate": 8.060021436227224e-05,
+      "loss": 5.0442,
+      "step": 558
+    },
+    {
+      "epoch": 0.5991425509110396,
+      "grad_norm": 2.289313793182373,
+      "learning_rate": 8.038585209003216e-05,
+      "loss": 4.3944,
+      "step": 559
+    },
+    {
+      "epoch": 0.6002143622722401,
+      "grad_norm": 3.844221830368042,
+      "learning_rate": 8.017148981779207e-05,
+      "loss": 4.2145,
+      "step": 560
+    },
+    {
+      "epoch": 0.6012861736334405,
+      "grad_norm": 2.877458333969116,
+      "learning_rate": 7.9957127545552e-05,
+      "loss": 4.8711,
+      "step": 561
+    },
+    {
+      "epoch": 0.602357984994641,
+      "grad_norm": 2.864017963409424,
+      "learning_rate": 7.974276527331191e-05,
+      "loss": 4.457,
+      "step": 562
+    },
+    {
+      "epoch": 0.6034297963558414,
+      "grad_norm": 2.4389705657958984,
+      "learning_rate": 7.952840300107182e-05,
+      "loss": 4.3321,
+      "step": 563
+    },
+    {
+      "epoch": 0.6045016077170418,
+      "grad_norm": 2.7437260150909424,
+      "learning_rate": 7.931404072883173e-05,
+      "loss": 4.2526,
+      "step": 564
+    },
+    {
+      "epoch": 0.6055734190782422,
+      "grad_norm": 2.4014856815338135,
+      "learning_rate": 7.909967845659164e-05,
+      "loss": 4.6251,
+      "step": 565
+    },
+    {
+      "epoch": 0.6066452304394426,
+      "grad_norm": 2.1202969551086426,
+      "learning_rate": 7.888531618435156e-05,
+      "loss": 4.4078,
+      "step": 566
+    },
+    {
+      "epoch": 0.6077170418006431,
+      "grad_norm": 3.2737956047058105,
+      "learning_rate": 7.867095391211147e-05,
+      "loss": 5.3022,
+      "step": 567
+    },
+    {
+      "epoch": 0.6087888531618435,
+      "grad_norm": 4.715353012084961,
+      "learning_rate": 7.84565916398714e-05,
+      "loss": 4.4089,
+      "step": 568
+    },
+    {
+      "epoch": 0.609860664523044,
+      "grad_norm": 3.042727470397949,
+      "learning_rate": 7.824222936763131e-05,
+      "loss": 4.7046,
+      "step": 569
+    },
+    {
+      "epoch": 0.6109324758842444,
+      "grad_norm": 4.115657806396484,
+      "learning_rate": 7.802786709539122e-05,
+      "loss": 4.6974,
+      "step": 570
+    },
+    {
+      "epoch": 0.6120042872454448,
+      "grad_norm": 2.7663843631744385,
+      "learning_rate": 7.781350482315113e-05,
+      "loss": 4.4788,
+      "step": 571
+    },
+    {
+      "epoch": 0.6130760986066452,
+      "grad_norm": 2.7116591930389404,
+      "learning_rate": 7.759914255091104e-05,
+      "loss": 4.4394,
+      "step": 572
+    },
+    {
+      "epoch": 0.6141479099678456,
+      "grad_norm": 2.617058515548706,
+      "learning_rate": 7.738478027867096e-05,
+      "loss": 4.5058,
+      "step": 573
+    },
+    {
+      "epoch": 0.6152197213290461,
+      "grad_norm": 2.9258296489715576,
+      "learning_rate": 7.717041800643087e-05,
+      "loss": 4.638,
+      "step": 574
+    },
+    {
+      "epoch": 0.6162915326902465,
+      "grad_norm": 2.290130376815796,
+      "learning_rate": 7.69560557341908e-05,
+      "loss": 4.494,
+      "step": 575
+    },
+    {
+      "epoch": 0.617363344051447,
+      "grad_norm": 2.9403839111328125,
+      "learning_rate": 7.674169346195071e-05,
+      "loss": 4.4339,
+      "step": 576
+    },
+    {
+      "epoch": 0.6184351554126474,
+      "grad_norm": 5.357972621917725,
+      "learning_rate": 7.652733118971062e-05,
+      "loss": 4.3492,
+      "step": 577
+    },
+    {
+      "epoch": 0.6195069667738478,
+      "grad_norm": 3.641831636428833,
+      "learning_rate": 7.631296891747053e-05,
+      "loss": 4.4423,
+      "step": 578
+    },
+    {
+      "epoch": 0.6205787781350482,
+      "grad_norm": 2.3929615020751953,
+      "learning_rate": 7.609860664523044e-05,
+      "loss": 4.4423,
+      "step": 579
+    },
+    {
+      "epoch": 0.6216505894962486,
+      "grad_norm": 3.540724515914917,
+      "learning_rate": 7.588424437299036e-05,
+      "loss": 4.215,
+      "step": 580
+    },
+    {
+      "epoch": 0.6227224008574491,
+      "grad_norm": 2.61544132232666,
+      "learning_rate": 7.566988210075027e-05,
+      "loss": 4.4321,
+      "step": 581
+    },
+    {
+      "epoch": 0.6237942122186495,
+      "grad_norm": 3.734426498413086,
+      "learning_rate": 7.54555198285102e-05,
+      "loss": 5.0043,
+      "step": 582
+    },
+    {
+      "epoch": 0.62486602357985,
+      "grad_norm": 2.775791645050049,
+      "learning_rate": 7.524115755627011e-05,
+      "loss": 4.2766,
+      "step": 583
+    },
+    {
+      "epoch": 0.6259378349410504,
+      "grad_norm": 2.061469793319702,
+      "learning_rate": 7.502679528403002e-05,
+      "loss": 4.6404,
+      "step": 584
+    },
+    {
+      "epoch": 0.6270096463022508,
+      "grad_norm": 2.8334105014801025,
+      "learning_rate": 7.481243301178993e-05,
+      "loss": 4.8849,
+      "step": 585
+    },
+    {
+      "epoch": 0.6280814576634512,
+      "grad_norm": 2.0020415782928467,
+      "learning_rate": 7.459807073954984e-05,
+      "loss": 4.4276,
+      "step": 586
+    },
+    {
+      "epoch": 0.6291532690246516,
+      "grad_norm": 3.436272621154785,
+      "learning_rate": 7.438370846730976e-05,
+      "loss": 4.2399,
+      "step": 587
+    },
+    {
+      "epoch": 0.6302250803858521,
+      "grad_norm": 3.1034908294677734,
+      "learning_rate": 7.416934619506967e-05,
+      "loss": 4.8148,
+      "step": 588
+    },
+    {
+      "epoch": 0.6312968917470525,
+      "grad_norm": 2.962636947631836,
+      "learning_rate": 7.39549839228296e-05,
+      "loss": 4.2156,
+      "step": 589
+    },
+    {
+      "epoch": 0.632368703108253,
+      "grad_norm": 2.5060412883758545,
+      "learning_rate": 7.374062165058951e-05,
+      "loss": 4.269,
+      "step": 590
+    },
+    {
+      "epoch": 0.6334405144694534,
+      "grad_norm": 3.32008695602417,
+      "learning_rate": 7.352625937834942e-05,
+      "loss": 4.6386,
+      "step": 591
+    },
+    {
+      "epoch": 0.6345123258306538,
+      "grad_norm": 2.796219825744629,
+      "learning_rate": 7.331189710610933e-05,
+      "loss": 4.235,
+      "step": 592
+    },
+    {
+      "epoch": 0.6355841371918542,
+      "grad_norm": 3.8189280033111572,
+      "learning_rate": 7.309753483386924e-05,
+      "loss": 4.4182,
+      "step": 593
+    },
+    {
+      "epoch": 0.6366559485530546,
+      "grad_norm": 2.9340178966522217,
+      "learning_rate": 7.288317256162916e-05,
+      "loss": 4.2946,
+      "step": 594
+    },
+    {
+      "epoch": 0.6377277599142551,
+      "grad_norm": 4.388503551483154,
+      "learning_rate": 7.266881028938907e-05,
+      "loss": 4.5359,
+      "step": 595
+    },
+    {
+      "epoch": 0.6387995712754555,
+      "grad_norm": 3.3181519508361816,
+      "learning_rate": 7.2454448017149e-05,
+      "loss": 4.4088,
+      "step": 596
+    },
+    {
+      "epoch": 0.639871382636656,
+      "grad_norm": 2.9396121501922607,
+      "learning_rate": 7.224008574490891e-05,
+      "loss": 4.8903,
+      "step": 597
+    },
+    {
+      "epoch": 0.6409431939978564,
+      "grad_norm": 4.775658130645752,
+      "learning_rate": 7.202572347266882e-05,
+      "loss": 4.9145,
+      "step": 598
+    },
+    {
+      "epoch": 0.6420150053590568,
+      "grad_norm": 3.0390172004699707,
+      "learning_rate": 7.181136120042873e-05,
+      "loss": 4.3406,
+      "step": 599
+    },
+    {
+      "epoch": 0.6430868167202572,
+      "grad_norm": 3.109550952911377,
+      "learning_rate": 7.159699892818864e-05,
+      "loss": 4.6635,
+      "step": 600
+    },
+    {
+      "epoch": 0.6441586280814576,
+      "grad_norm": 2.880469799041748,
+      "learning_rate": 7.138263665594856e-05,
+      "loss": 4.875,
+      "step": 601
+    },
+    {
+      "epoch": 0.6452304394426581,
+      "grad_norm": 3.215329170227051,
+      "learning_rate": 7.116827438370847e-05,
+      "loss": 4.347,
+      "step": 602
+    },
+    {
+      "epoch": 0.6463022508038585,
+      "grad_norm": 3.218173027038574,
+      "learning_rate": 7.09539121114684e-05,
+      "loss": 4.72,
+      "step": 603
+    },
+    {
+      "epoch": 0.647374062165059,
+      "grad_norm": 2.2390801906585693,
+      "learning_rate": 7.073954983922831e-05,
+      "loss": 4.732,
+      "step": 604
+    },
+    {
+      "epoch": 0.6484458735262594,
+      "grad_norm": 4.779411315917969,
+      "learning_rate": 7.052518756698822e-05,
+      "loss": 4.4878,
+      "step": 605
+    },
+    {
+      "epoch": 0.6495176848874598,
+      "grad_norm": 2.320235252380371,
+      "learning_rate": 7.031082529474813e-05,
+      "loss": 4.4545,
+      "step": 606
+    },
+    {
+      "epoch": 0.6505894962486602,
+      "grad_norm": 2.8826911449432373,
+      "learning_rate": 7.009646302250804e-05,
+      "loss": 4.8964,
+      "step": 607
+    },
+    {
+      "epoch": 0.6516613076098606,
+      "grad_norm": 2.1690621376037598,
+      "learning_rate": 6.988210075026796e-05,
+      "loss": 4.6063,
+      "step": 608
+    },
+    {
+      "epoch": 0.6527331189710611,
+      "grad_norm": 3.2647693157196045,
+      "learning_rate": 6.966773847802787e-05,
+      "loss": 4.8094,
+      "step": 609
+    },
+    {
+      "epoch": 0.6538049303322615,
+      "grad_norm": 3.142394542694092,
+      "learning_rate": 6.94533762057878e-05,
+      "loss": 4.3307,
+      "step": 610
+    },
+    {
+      "epoch": 0.654876741693462,
+      "grad_norm": 3.0853631496429443,
+      "learning_rate": 6.923901393354771e-05,
+      "loss": 4.3164,
+      "step": 611
+    },
+    {
+      "epoch": 0.6559485530546624,
+      "grad_norm": 3.033388614654541,
+      "learning_rate": 6.902465166130762e-05,
+      "loss": 4.4713,
+      "step": 612
+    },
+    {
+      "epoch": 0.6570203644158628,
+      "grad_norm": 2.7106258869171143,
+      "learning_rate": 6.881028938906753e-05,
+      "loss": 4.5005,
+      "step": 613
+    },
+    {
+      "epoch": 0.6580921757770632,
+      "grad_norm": 3.773181200027466,
+      "learning_rate": 6.859592711682744e-05,
+      "loss": 4.5155,
+      "step": 614
+    },
+    {
+      "epoch": 0.6591639871382636,
+      "grad_norm": 5.108589172363281,
+      "learning_rate": 6.838156484458736e-05,
+      "loss": 4.5609,
+      "step": 615
+    },
+    {
+      "epoch": 0.6602357984994641,
+      "grad_norm": 3.9254658222198486,
+      "learning_rate": 6.816720257234727e-05,
+      "loss": 4.5529,
+      "step": 616
+    },
+    {
+      "epoch": 0.6613076098606645,
+      "grad_norm": 3.022003650665283,
+      "learning_rate": 6.795284030010718e-05,
+      "loss": 4.2055,
+      "step": 617
+    },
+    {
+      "epoch": 0.662379421221865,
+      "grad_norm": 2.3257830142974854,
+      "learning_rate": 6.773847802786711e-05,
+      "loss": 4.4332,
+      "step": 618
+    },
+    {
+      "epoch": 0.6634512325830654,
+      "grad_norm": 1.9579797983169556,
+      "learning_rate": 6.752411575562702e-05,
+      "loss": 4.3841,
+      "step": 619
+    },
+    {
+      "epoch": 0.6645230439442658,
+      "grad_norm": 4.0417351722717285,
+      "learning_rate": 6.730975348338693e-05,
+      "loss": 4.7266,
+      "step": 620
+    },
+    {
+      "epoch": 0.6655948553054662,
+      "grad_norm": 2.9920849800109863,
+      "learning_rate": 6.709539121114685e-05,
+      "loss": 4.4271,
+      "step": 621
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 3.686924934387207,
+      "learning_rate": 6.688102893890676e-05,
+      "loss": 4.6426,
+      "step": 622
+    },
+    {
+      "epoch": 0.6677384780278671,
+      "grad_norm": 2.8612165451049805,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 4.4797,
+      "step": 623
+    },
+    {
+      "epoch": 0.6688102893890675,
+      "grad_norm": 2.1640119552612305,
+      "learning_rate": 6.645230439442658e-05,
+      "loss": 4.3597,
+      "step": 624
+    },
+    {
+      "epoch": 0.669882100750268,
+      "grad_norm": 2.755424737930298,
+      "learning_rate": 6.623794212218651e-05,
+      "loss": 4.5758,
+      "step": 625
+    },
+    {
+      "epoch": 0.6709539121114684,
+      "grad_norm": 3.111708402633667,
+      "learning_rate": 6.602357984994642e-05,
+      "loss": 4.3939,
+      "step": 626
+    },
+    {
+      "epoch": 0.6720257234726688,
+      "grad_norm": 2.549870014190674,
+      "learning_rate": 6.580921757770633e-05,
+      "loss": 4.3836,
+      "step": 627
+    },
+    {
+      "epoch": 0.6730975348338692,
+      "grad_norm": 2.126152515411377,
+      "learning_rate": 6.559485530546625e-05,
+      "loss": 4.5625,
+      "step": 628
+    },
+    {
+      "epoch": 0.6741693461950696,
+      "grad_norm": 4.276716709136963,
+      "learning_rate": 6.538049303322616e-05,
+      "loss": 4.5386,
+      "step": 629
+    },
+    {
+      "epoch": 0.6752411575562701,
+      "grad_norm": 3.121826648712158,
+      "learning_rate": 6.516613076098607e-05,
+      "loss": 4.564,
+      "step": 630
+    },
+    {
+      "epoch": 0.6763129689174705,
+      "grad_norm": 2.442767381668091,
+      "learning_rate": 6.495176848874598e-05,
+      "loss": 4.655,
+      "step": 631
+    },
+    {
+      "epoch": 0.677384780278671,
+      "grad_norm": 2.6107630729675293,
+      "learning_rate": 6.473740621650591e-05,
+      "loss": 4.5491,
+      "step": 632
+    },
+    {
+      "epoch": 0.6784565916398714,
+      "grad_norm": 3.7255454063415527,
+      "learning_rate": 6.452304394426582e-05,
+      "loss": 4.3361,
+      "step": 633
+    },
+    {
+      "epoch": 0.6795284030010718,
+      "grad_norm": 2.9900388717651367,
+      "learning_rate": 6.430868167202573e-05,
+      "loss": 4.5003,
+      "step": 634
+    },
+    {
+      "epoch": 0.6806002143622722,
+      "grad_norm": 2.156949758529663,
+      "learning_rate": 6.409431939978565e-05,
+      "loss": 4.348,
+      "step": 635
+    },
+    {
+      "epoch": 0.6816720257234726,
+      "grad_norm": 2.6016056537628174,
+      "learning_rate": 6.387995712754556e-05,
+      "loss": 4.4779,
+      "step": 636
+    },
+    {
+      "epoch": 0.6827438370846731,
+      "grad_norm": 3.6784298419952393,
+      "learning_rate": 6.366559485530547e-05,
+      "loss": 4.4926,
+      "step": 637
+    },
+    {
+      "epoch": 0.6838156484458735,
+      "grad_norm": 3.2673237323760986,
+      "learning_rate": 6.345123258306538e-05,
+      "loss": 4.4138,
+      "step": 638
+    },
+    {
+      "epoch": 0.684887459807074,
+      "grad_norm": 2.430467367172241,
+      "learning_rate": 6.323687031082531e-05,
+      "loss": 4.4007,
+      "step": 639
+    },
+    {
+      "epoch": 0.6859592711682744,
+      "grad_norm": 3.361685276031494,
+      "learning_rate": 6.302250803858522e-05,
+      "loss": 4.1909,
+      "step": 640
+    },
+    {
+      "epoch": 0.6870310825294748,
+      "grad_norm": 2.180561065673828,
+      "learning_rate": 6.280814576634513e-05,
+      "loss": 4.4958,
+      "step": 641
+    },
+    {
+      "epoch": 0.6881028938906752,
+      "grad_norm": 3.7177274227142334,
+      "learning_rate": 6.259378349410505e-05,
+      "loss": 4.5105,
+      "step": 642
+    },
+    {
+      "epoch": 0.6891747052518756,
+      "grad_norm": 3.392599105834961,
+      "learning_rate": 6.237942122186496e-05,
+      "loss": 4.3326,
+      "step": 643
+    },
+    {
+      "epoch": 0.6902465166130761,
+      "grad_norm": 2.894768238067627,
+      "learning_rate": 6.216505894962487e-05,
+      "loss": 4.5509,
+      "step": 644
+    },
+    {
+      "epoch": 0.6913183279742765,
+      "grad_norm": 6.643900394439697,
+      "learning_rate": 6.195069667738478e-05,
+      "loss": 4.8811,
+      "step": 645
+    },
+    {
+      "epoch": 0.692390139335477,
+      "grad_norm": 3.7013251781463623,
+      "learning_rate": 6.173633440514471e-05,
+      "loss": 4.2468,
+      "step": 646
+    },
+    {
+      "epoch": 0.6934619506966774,
+      "grad_norm": 2.7494664192199707,
+      "learning_rate": 6.152197213290462e-05,
+      "loss": 5.0329,
+      "step": 647
+    },
+    {
+      "epoch": 0.6945337620578779,
+      "grad_norm": 2.2148094177246094,
+      "learning_rate": 6.130760986066453e-05,
+      "loss": 4.3135,
+      "step": 648
+    },
+    {
+      "epoch": 0.6956055734190782,
+      "grad_norm": 2.2598962783813477,
+      "learning_rate": 6.109324758842445e-05,
+      "loss": 4.3436,
+      "step": 649
+    },
+    {
+      "epoch": 0.6966773847802786,
+      "grad_norm": 4.410878658294678,
+      "learning_rate": 6.087888531618435e-05,
+      "loss": 4.8447,
+      "step": 650
+    },
+    {
+      "epoch": 0.6977491961414791,
+      "grad_norm": 2.493114709854126,
+      "learning_rate": 6.0664523043944264e-05,
+      "loss": 4.2732,
+      "step": 651
+    },
+    {
+      "epoch": 0.6988210075026795,
+      "grad_norm": 2.972546100616455,
+      "learning_rate": 6.0450160771704176e-05,
+      "loss": 4.5284,
+      "step": 652
+    },
+    {
+      "epoch": 0.69989281886388,
+      "grad_norm": 3.26788330078125,
+      "learning_rate": 6.02357984994641e-05,
+      "loss": 4.7868,
+      "step": 653
+    },
+    {
+      "epoch": 0.7009646302250804,
+      "grad_norm": 4.179505348205566,
+      "learning_rate": 6.0021436227224014e-05,
+      "loss": 4.4146,
+      "step": 654
+    },
+    {
+      "epoch": 0.7020364415862809,
+      "grad_norm": 2.4835996627807617,
+      "learning_rate": 5.980707395498393e-05,
+      "loss": 4.9622,
+      "step": 655
+    },
+    {
+      "epoch": 0.7031082529474812,
+      "grad_norm": 3.4152846336364746,
+      "learning_rate": 5.959271168274384e-05,
+      "loss": 4.8898,
+      "step": 656
+    },
+    {
+      "epoch": 0.7041800643086816,
+      "grad_norm": 3.430803060531616,
+      "learning_rate": 5.937834941050375e-05,
+      "loss": 4.6275,
+      "step": 657
+    },
+    {
+      "epoch": 0.7052518756698821,
+      "grad_norm": 2.6322433948516846,
+      "learning_rate": 5.9163987138263664e-05,
+      "loss": 4.5736,
+      "step": 658
+    },
+    {
+      "epoch": 0.7063236870310825,
+      "grad_norm": 2.5020229816436768,
+      "learning_rate": 5.8949624866023576e-05,
+      "loss": 5.0218,
+      "step": 659
+    },
+    {
+      "epoch": 0.707395498392283,
+      "grad_norm": 2.7531423568725586,
+      "learning_rate": 5.87352625937835e-05,
+      "loss": 4.7368,
+      "step": 660
+    },
+    {
+      "epoch": 0.7084673097534834,
+      "grad_norm": 2.332533597946167,
+      "learning_rate": 5.8520900321543414e-05,
+      "loss": 4.4331,
+      "step": 661
+    },
+    {
+      "epoch": 0.7095391211146839,
+      "grad_norm": 2.4030661582946777,
+      "learning_rate": 5.830653804930333e-05,
+      "loss": 4.3736,
+      "step": 662
+    },
+    {
+      "epoch": 0.7106109324758842,
+      "grad_norm": 2.747347354888916,
+      "learning_rate": 5.809217577706324e-05,
+      "loss": 4.4708,
+      "step": 663
+    },
+    {
+      "epoch": 0.7116827438370846,
+      "grad_norm": 3.930635690689087,
+      "learning_rate": 5.787781350482315e-05,
+      "loss": 4.7282,
+      "step": 664
+    },
+    {
+      "epoch": 0.7127545551982851,
+      "grad_norm": 2.4571895599365234,
+      "learning_rate": 5.7663451232583064e-05,
+      "loss": 4.3438,
+      "step": 665
+    },
+    {
+      "epoch": 0.7138263665594855,
+      "grad_norm": 2.307389497756958,
+      "learning_rate": 5.7449088960342976e-05,
+      "loss": 4.4917,
+      "step": 666
+    },
+    {
+      "epoch": 0.714898177920686,
+      "grad_norm": 3.0858333110809326,
+      "learning_rate": 5.72347266881029e-05,
+      "loss": 4.7639,
+      "step": 667
+    },
+    {
+      "epoch": 0.7159699892818864,
+      "grad_norm": 2.8468878269195557,
+      "learning_rate": 5.7020364415862815e-05,
+      "loss": 4.4349,
+      "step": 668
+    },
+    {
+      "epoch": 0.7170418006430869,
+      "grad_norm": 2.674168825149536,
+      "learning_rate": 5.680600214362273e-05,
+      "loss": 4.6781,
+      "step": 669
+    },
+    {
+      "epoch": 0.7181136120042872,
+      "grad_norm": 3.077220916748047,
+      "learning_rate": 5.659163987138264e-05,
+      "loss": 4.5982,
+      "step": 670
+    },
+    {
+      "epoch": 0.7191854233654876,
+      "grad_norm": 2.4594485759735107,
+      "learning_rate": 5.637727759914255e-05,
+      "loss": 4.4059,
+      "step": 671
+    },
+    {
+      "epoch": 0.7202572347266881,
+      "grad_norm": 3.3636786937713623,
+      "learning_rate": 5.6162915326902464e-05,
+      "loss": 4.5043,
+      "step": 672
+    },
+    {
+      "epoch": 0.7213290460878885,
+      "grad_norm": 5.092872619628906,
+      "learning_rate": 5.5948553054662377e-05,
+      "loss": 4.6357,
+      "step": 673
+    },
+    {
+      "epoch": 0.722400857449089,
+      "grad_norm": 3.4879798889160156,
+      "learning_rate": 5.573419078242229e-05,
+      "loss": 4.4038,
+      "step": 674
+    },
+    {
+      "epoch": 0.7234726688102894,
+      "grad_norm": 3.5704169273376465,
+      "learning_rate": 5.5519828510182215e-05,
+      "loss": 4.2148,
+      "step": 675
+    },
+    {
+      "epoch": 0.7245444801714899,
+      "grad_norm": 3.622905731201172,
+      "learning_rate": 5.530546623794213e-05,
+      "loss": 4.2071,
+      "step": 676
+    },
+    {
+      "epoch": 0.7256162915326902,
+      "grad_norm": 2.2104790210723877,
+      "learning_rate": 5.509110396570204e-05,
+      "loss": 4.3016,
+      "step": 677
+    },
+    {
+      "epoch": 0.7266881028938906,
+      "grad_norm": 2.5049002170562744,
+      "learning_rate": 5.487674169346195e-05,
+      "loss": 4.2184,
+      "step": 678
+    },
+    {
+      "epoch": 0.7277599142550911,
+      "grad_norm": 2.545485019683838,
+      "learning_rate": 5.4662379421221864e-05,
+      "loss": 4.3773,
+      "step": 679
+    },
+    {
+      "epoch": 0.7288317256162915,
+      "grad_norm": 3.1245920658111572,
+      "learning_rate": 5.444801714898178e-05,
+      "loss": 4.6495,
+      "step": 680
+    },
+    {
+      "epoch": 0.729903536977492,
+      "grad_norm": 3.3570711612701416,
+      "learning_rate": 5.423365487674169e-05,
+      "loss": 4.3041,
+      "step": 681
+    },
+    {
+      "epoch": 0.7309753483386924,
+      "grad_norm": 2.5760622024536133,
+      "learning_rate": 5.4019292604501615e-05,
+      "loss": 4.6344,
+      "step": 682
+    },
+    {
+      "epoch": 0.7320471596998929,
+      "grad_norm": 3.143083095550537,
+      "learning_rate": 5.380493033226153e-05,
+      "loss": 4.6799,
+      "step": 683
+    },
+    {
+      "epoch": 0.7331189710610932,
+      "grad_norm": 4.669005870819092,
+      "learning_rate": 5.359056806002144e-05,
+      "loss": 4.3954,
+      "step": 684
+    },
+    {
+      "epoch": 0.7341907824222936,
+      "grad_norm": 3.921560049057007,
+      "learning_rate": 5.337620578778135e-05,
+      "loss": 4.6724,
+      "step": 685
+    },
+    {
+      "epoch": 0.7352625937834941,
+      "grad_norm": 4.419112682342529,
+      "learning_rate": 5.3161843515541264e-05,
+      "loss": 4.6749,
+      "step": 686
+    },
+    {
+      "epoch": 0.7363344051446945,
+      "grad_norm": 3.118290424346924,
+      "learning_rate": 5.294748124330118e-05,
+      "loss": 4.3938,
+      "step": 687
+    },
+    {
+      "epoch": 0.737406216505895,
+      "grad_norm": 2.2908451557159424,
+      "learning_rate": 5.273311897106109e-05,
+      "loss": 4.4141,
+      "step": 688
+    },
+    {
+      "epoch": 0.7384780278670954,
+      "grad_norm": 2.6434786319732666,
+      "learning_rate": 5.2518756698821015e-05,
+      "loss": 4.4886,
+      "step": 689
+    },
+    {
+      "epoch": 0.7395498392282959,
+      "grad_norm": 2.6037230491638184,
+      "learning_rate": 5.230439442658093e-05,
+      "loss": 4.5665,
+      "step": 690
+    },
+    {
+      "epoch": 0.7406216505894962,
+      "grad_norm": 3.525683641433716,
+      "learning_rate": 5.209003215434084e-05,
+      "loss": 4.1823,
+      "step": 691
+    },
+    {
+      "epoch": 0.7416934619506966,
+      "grad_norm": 3.5986504554748535,
+      "learning_rate": 5.187566988210075e-05,
+      "loss": 4.4201,
+      "step": 692
+    },
+    {
+      "epoch": 0.7427652733118971,
+      "grad_norm": 4.4375176429748535,
+      "learning_rate": 5.1661307609860665e-05,
+      "loss": 4.3157,
+      "step": 693
+    },
+    {
+      "epoch": 0.7438370846730975,
+      "grad_norm": 2.6561779975891113,
+      "learning_rate": 5.144694533762058e-05,
+      "loss": 4.3444,
+      "step": 694
+    },
+    {
+      "epoch": 0.744908896034298,
+      "grad_norm": 2.6672708988189697,
+      "learning_rate": 5.123258306538049e-05,
+      "loss": 4.5737,
+      "step": 695
+    },
+    {
+      "epoch": 0.7459807073954984,
+      "grad_norm": 3.335510015487671,
+      "learning_rate": 5.1018220793140415e-05,
+      "loss": 4.4153,
+      "step": 696
+    },
+    {
+      "epoch": 0.7470525187566989,
+      "grad_norm": 2.9571712017059326,
+      "learning_rate": 5.080385852090033e-05,
+      "loss": 4.4999,
+      "step": 697
+    },
+    {
+      "epoch": 0.7481243301178992,
+      "grad_norm": 2.0713906288146973,
+      "learning_rate": 5.058949624866024e-05,
+      "loss": 4.1479,
+      "step": 698
+    },
+    {
+      "epoch": 0.7491961414790996,
+      "grad_norm": 2.731901168823242,
+      "learning_rate": 5.037513397642015e-05,
+      "loss": 4.3306,
+      "step": 699
+    },
+    {
+      "epoch": 0.7502679528403001,
+      "grad_norm": 3.4839823246002197,
+      "learning_rate": 5.0160771704180065e-05,
+      "loss": 4.3706,
+      "step": 700
+    },
+    {
+      "epoch": 0.7513397642015005,
+      "grad_norm": 3.9278030395507812,
+      "learning_rate": 4.9946409431939984e-05,
+      "loss": 4.4768,
+      "step": 701
+    },
+    {
+      "epoch": 0.752411575562701,
+      "grad_norm": 2.87748122215271,
+      "learning_rate": 4.9732047159699896e-05,
+      "loss": 4.463,
+      "step": 702
+    },
+    {
+      "epoch": 0.7534833869239014,
+      "grad_norm": 3.587252616882324,
+      "learning_rate": 4.951768488745981e-05,
+      "loss": 4.7204,
+      "step": 703
+    },
+    {
+      "epoch": 0.7545551982851019,
+      "grad_norm": 2.9082605838775635,
+      "learning_rate": 4.930332261521972e-05,
+      "loss": 4.7662,
+      "step": 704
+    },
+    {
+      "epoch": 0.7556270096463023,
+      "grad_norm": 3.1029810905456543,
+      "learning_rate": 4.908896034297964e-05,
+      "loss": 4.5769,
+      "step": 705
+    },
+    {
+      "epoch": 0.7566988210075026,
+      "grad_norm": 2.8559889793395996,
+      "learning_rate": 4.887459807073955e-05,
+      "loss": 4.2848,
+      "step": 706
+    },
+    {
+      "epoch": 0.7577706323687031,
+      "grad_norm": 2.697112560272217,
+      "learning_rate": 4.8660235798499465e-05,
+      "loss": 4.4089,
+      "step": 707
+    },
+    {
+      "epoch": 0.7588424437299035,
+      "grad_norm": 3.552123546600342,
+      "learning_rate": 4.8445873526259384e-05,
+      "loss": 4.3264,
+      "step": 708
+    },
+    {
+      "epoch": 0.759914255091104,
+      "grad_norm": 4.04207181930542,
+      "learning_rate": 4.8231511254019296e-05,
+      "loss": 4.9773,
+      "step": 709
+    },
+    {
+      "epoch": 0.7609860664523044,
+      "grad_norm": 4.384789943695068,
+      "learning_rate": 4.801714898177921e-05,
+      "loss": 4.2986,
+      "step": 710
+    },
+    {
+      "epoch": 0.7620578778135049,
+      "grad_norm": 3.348785400390625,
+      "learning_rate": 4.780278670953912e-05,
+      "loss": 4.6206,
+      "step": 711
+    },
+    {
+      "epoch": 0.7631296891747053,
+      "grad_norm": 2.0677852630615234,
+      "learning_rate": 4.758842443729904e-05,
+      "loss": 4.4998,
+      "step": 712
+    },
+    {
+      "epoch": 0.7642015005359056,
+      "grad_norm": 3.080993890762329,
+      "learning_rate": 4.737406216505895e-05,
+      "loss": 4.2714,
+      "step": 713
+    },
+    {
+      "epoch": 0.7652733118971061,
+      "grad_norm": 2.621718406677246,
+      "learning_rate": 4.7159699892818865e-05,
+      "loss": 4.508,
+      "step": 714
+    },
+    {
+      "epoch": 0.7663451232583065,
+      "grad_norm": 2.7665212154388428,
+      "learning_rate": 4.6945337620578784e-05,
+      "loss": 4.4284,
+      "step": 715
+    },
+    {
+      "epoch": 0.767416934619507,
+      "grad_norm": 4.301600456237793,
+      "learning_rate": 4.6730975348338697e-05,
+      "loss": 4.8282,
+      "step": 716
+    },
+    {
+      "epoch": 0.7684887459807074,
+      "grad_norm": 4.236885070800781,
+      "learning_rate": 4.651661307609861e-05,
+      "loss": 4.8417,
+      "step": 717
+    },
+    {
+      "epoch": 0.7695605573419079,
+      "grad_norm": 5.495549201965332,
+      "learning_rate": 4.630225080385852e-05,
+      "loss": 4.124,
+      "step": 718
+    },
+    {
+      "epoch": 0.7706323687031083,
+      "grad_norm": 3.4314773082733154,
+      "learning_rate": 4.608788853161844e-05,
+      "loss": 4.5811,
+      "step": 719
+    },
+    {
+      "epoch": 0.7717041800643086,
+      "grad_norm": 2.890056848526001,
+      "learning_rate": 4.587352625937835e-05,
+      "loss": 4.2577,
+      "step": 720
+    },
+    {
+      "epoch": 0.7727759914255091,
+      "grad_norm": 2.566664218902588,
+      "learning_rate": 4.5659163987138265e-05,
+      "loss": 4.3006,
+      "step": 721
+    },
+    {
+      "epoch": 0.7738478027867095,
+      "grad_norm": 3.2805826663970947,
+      "learning_rate": 4.544480171489818e-05,
+      "loss": 4.1904,
+      "step": 722
+    },
+    {
+      "epoch": 0.77491961414791,
+      "grad_norm": 3.1700801849365234,
+      "learning_rate": 4.52304394426581e-05,
+      "loss": 4.2985,
+      "step": 723
+    },
+    {
+      "epoch": 0.7759914255091104,
+      "grad_norm": 2.913239002227783,
+      "learning_rate": 4.501607717041801e-05,
+      "loss": 4.5163,
+      "step": 724
+    },
+    {
+      "epoch": 0.7770632368703109,
+      "grad_norm": 3.6739931106567383,
+      "learning_rate": 4.480171489817792e-05,
+      "loss": 4.0745,
+      "step": 725
+    },
+    {
+      "epoch": 0.7781350482315113,
+      "grad_norm": 4.171450138092041,
+      "learning_rate": 4.458735262593784e-05,
+      "loss": 4.1734,
+      "step": 726
+    },
+    {
+      "epoch": 0.7792068595927116,
+      "grad_norm": 4.786383628845215,
+      "learning_rate": 4.437299035369775e-05,
+      "loss": 4.3863,
+      "step": 727
+    },
+    {
+      "epoch": 0.7802786709539121,
+      "grad_norm": 3.3467090129852295,
+      "learning_rate": 4.4158628081457665e-05,
+      "loss": 4.2379,
+      "step": 728
+    },
+    {
+      "epoch": 0.7813504823151125,
+      "grad_norm": 3.2874300479888916,
+      "learning_rate": 4.394426580921758e-05,
+      "loss": 4.5021,
+      "step": 729
+    },
+    {
+      "epoch": 0.782422293676313,
+      "grad_norm": 2.1295790672302246,
+      "learning_rate": 4.37299035369775e-05,
+      "loss": 4.2291,
+      "step": 730
+    },
+    {
+      "epoch": 0.7834941050375134,
+      "grad_norm": 3.8685142993927,
+      "learning_rate": 4.351554126473741e-05,
+      "loss": 4.2344,
+      "step": 731
+    },
+    {
+      "epoch": 0.7845659163987139,
+      "grad_norm": 2.811497688293457,
+      "learning_rate": 4.330117899249732e-05,
+      "loss": 4.7444,
+      "step": 732
+    },
+    {
+      "epoch": 0.7856377277599143,
+      "grad_norm": 2.16815185546875,
+      "learning_rate": 4.308681672025724e-05,
+      "loss": 4.5432,
+      "step": 733
+    },
+    {
+      "epoch": 0.7867095391211146,
+      "grad_norm": 3.302750587463379,
+      "learning_rate": 4.287245444801715e-05,
+      "loss": 4.1623,
+      "step": 734
+    },
+    {
+      "epoch": 0.7877813504823151,
+      "grad_norm": 3.3594536781311035,
+      "learning_rate": 4.2658092175777065e-05,
+      "loss": 4.9906,
+      "step": 735
+    },
+    {
+      "epoch": 0.7888531618435155,
+      "grad_norm": 2.7700035572052,
+      "learning_rate": 4.244372990353698e-05,
+      "loss": 4.457,
+      "step": 736
+    },
+    {
+      "epoch": 0.789924973204716,
+      "grad_norm": 2.545769691467285,
+      "learning_rate": 4.22293676312969e-05,
+      "loss": 4.3638,
+      "step": 737
+    },
+    {
+      "epoch": 0.7909967845659164,
+      "grad_norm": 3.0002028942108154,
+      "learning_rate": 4.201500535905681e-05,
+      "loss": 4.3884,
+      "step": 738
+    },
+    {
+      "epoch": 0.7920685959271169,
+      "grad_norm": 2.997331142425537,
+      "learning_rate": 4.180064308681672e-05,
+      "loss": 4.3515,
+      "step": 739
+    },
+    {
+      "epoch": 0.7931404072883173,
+      "grad_norm": 2.539663553237915,
+      "learning_rate": 4.158628081457664e-05,
+      "loss": 4.4387,
+      "step": 740
+    },
+    {
+      "epoch": 0.7942122186495176,
+      "grad_norm": 2.4811909198760986,
+      "learning_rate": 4.137191854233655e-05,
+      "loss": 4.2736,
+      "step": 741
+    },
+    {
+      "epoch": 0.7952840300107181,
+      "grad_norm": 3.7457571029663086,
+      "learning_rate": 4.1157556270096466e-05,
+      "loss": 4.4863,
+      "step": 742
+    },
+    {
+      "epoch": 0.7963558413719185,
+      "grad_norm": 2.783266067504883,
+      "learning_rate": 4.094319399785638e-05,
+      "loss": 4.5847,
+      "step": 743
+    },
+    {
+      "epoch": 0.797427652733119,
+      "grad_norm": 2.202946186065674,
+      "learning_rate": 4.07288317256163e-05,
+      "loss": 4.4435,
+      "step": 744
+    },
+    {
+      "epoch": 0.7984994640943194,
+      "grad_norm": 3.0813934803009033,
+      "learning_rate": 4.051446945337621e-05,
+      "loss": 4.1583,
+      "step": 745
+    },
+    {
+      "epoch": 0.7995712754555199,
+      "grad_norm": 3.1837377548217773,
+      "learning_rate": 4.030010718113612e-05,
+      "loss": 4.5823,
+      "step": 746
+    },
+    {
+      "epoch": 0.8006430868167203,
+      "grad_norm": 2.310713529586792,
+      "learning_rate": 4.0085744908896034e-05,
+      "loss": 4.4296,
+      "step": 747
+    },
+    {
+      "epoch": 0.8017148981779206,
+      "grad_norm": 2.4914042949676514,
+      "learning_rate": 3.9871382636655953e-05,
+      "loss": 4.2493,
+      "step": 748
+    },
+    {
+      "epoch": 0.8027867095391211,
+      "grad_norm": 4.564463138580322,
+      "learning_rate": 3.9657020364415866e-05,
+      "loss": 4.5261,
+      "step": 749
+    },
+    {
+      "epoch": 0.8038585209003215,
+      "grad_norm": 3.1176915168762207,
+      "learning_rate": 3.944265809217578e-05,
+      "loss": 4.6123,
+      "step": 750
+    },
+    {
+      "epoch": 0.804930332261522,
+      "grad_norm": 2.2957403659820557,
+      "learning_rate": 3.92282958199357e-05,
+      "loss": 4.6512,
+      "step": 751
+    },
+    {
+      "epoch": 0.8060021436227224,
+      "grad_norm": 3.3911380767822266,
+      "learning_rate": 3.901393354769561e-05,
+      "loss": 4.3699,
+      "step": 752
+    },
+    {
+      "epoch": 0.8070739549839229,
+      "grad_norm": 2.3711419105529785,
+      "learning_rate": 3.879957127545552e-05,
+      "loss": 4.3832,
+      "step": 753
+    },
+    {
+      "epoch": 0.8081457663451233,
+      "grad_norm": 4.154041767120361,
+      "learning_rate": 3.8585209003215434e-05,
+      "loss": 4.4298,
+      "step": 754
+    },
+    {
+      "epoch": 0.8092175777063236,
+      "grad_norm": 4.426170825958252,
+      "learning_rate": 3.8370846730975354e-05,
+      "loss": 4.2205,
+      "step": 755
+    },
+    {
+      "epoch": 0.8102893890675241,
+      "grad_norm": 2.643601179122925,
+      "learning_rate": 3.8156484458735266e-05,
+      "loss": 4.1436,
+      "step": 756
+    },
+    {
+      "epoch": 0.8113612004287245,
+      "grad_norm": 2.698427200317383,
+      "learning_rate": 3.794212218649518e-05,
+      "loss": 4.3915,
+      "step": 757
+    },
+    {
+      "epoch": 0.812433011789925,
+      "grad_norm": 2.6776609420776367,
+      "learning_rate": 3.77277599142551e-05,
+      "loss": 4.7078,
+      "step": 758
+    },
+    {
+      "epoch": 0.8135048231511254,
+      "grad_norm": 3.502988338470459,
+      "learning_rate": 3.751339764201501e-05,
+      "loss": 4.5208,
+      "step": 759
+    },
+    {
+      "epoch": 0.8145766345123259,
+      "grad_norm": 2.7470109462738037,
+      "learning_rate": 3.729903536977492e-05,
+      "loss": 4.3289,
+      "step": 760
+    },
+    {
+      "epoch": 0.8156484458735263,
+      "grad_norm": 3.7509968280792236,
+      "learning_rate": 3.7084673097534835e-05,
+      "loss": 4.3425,
+      "step": 761
+    },
+    {
+      "epoch": 0.8167202572347267,
+      "grad_norm": 2.3047983646392822,
+      "learning_rate": 3.6870310825294754e-05,
+      "loss": 4.2544,
+      "step": 762
+    },
+    {
+      "epoch": 0.8177920685959271,
+      "grad_norm": 3.482459545135498,
+      "learning_rate": 3.6655948553054666e-05,
+      "loss": 4.6132,
+      "step": 763
+    },
+    {
+      "epoch": 0.8188638799571275,
+      "grad_norm": 3.8098599910736084,
+      "learning_rate": 3.644158628081458e-05,
+      "loss": 4.2388,
+      "step": 764
+    },
+    {
+      "epoch": 0.819935691318328,
+      "grad_norm": 3.2833423614501953,
+      "learning_rate": 3.62272240085745e-05,
+      "loss": 4.1095,
+      "step": 765
+    },
+    {
+      "epoch": 0.8210075026795284,
+      "grad_norm": 3.2167413234710693,
+      "learning_rate": 3.601286173633441e-05,
+      "loss": 4.2657,
+      "step": 766
+    },
+    {
+      "epoch": 0.8220793140407289,
+      "grad_norm": 2.6981112957000732,
+      "learning_rate": 3.579849946409432e-05,
+      "loss": 4.2648,
+      "step": 767
+    },
+    {
+      "epoch": 0.8231511254019293,
+      "grad_norm": 2.9271039962768555,
+      "learning_rate": 3.5584137191854235e-05,
+      "loss": 4.2033,
+      "step": 768
+    },
+    {
+      "epoch": 0.8242229367631297,
+      "grad_norm": 3.722834587097168,
+      "learning_rate": 3.5369774919614154e-05,
+      "loss": 4.6958,
+      "step": 769
+    },
+    {
+      "epoch": 0.8252947481243301,
+      "grad_norm": 2.129990339279175,
+      "learning_rate": 3.5155412647374066e-05,
+      "loss": 4.1381,
+      "step": 770
+    },
+    {
+      "epoch": 0.8263665594855305,
+      "grad_norm": 5.136198043823242,
+      "learning_rate": 3.494105037513398e-05,
+      "loss": 4.6291,
+      "step": 771
+    },
+    {
+      "epoch": 0.827438370846731,
+      "grad_norm": 2.7707130908966064,
+      "learning_rate": 3.47266881028939e-05,
+      "loss": 4.0731,
+      "step": 772
+    },
+    {
+      "epoch": 0.8285101822079314,
+      "grad_norm": 3.942397117614746,
+      "learning_rate": 3.451232583065381e-05,
+      "loss": 5.0691,
+      "step": 773
+    },
+    {
+      "epoch": 0.8295819935691319,
+      "grad_norm": 2.749673366546631,
+      "learning_rate": 3.429796355841372e-05,
+      "loss": 4.6547,
+      "step": 774
+    },
+    {
+      "epoch": 0.8306538049303323,
+      "grad_norm": 3.3368706703186035,
+      "learning_rate": 3.4083601286173635e-05,
+      "loss": 4.2318,
+      "step": 775
+    },
+    {
+      "epoch": 0.8317256162915327,
+      "grad_norm": 3.024880886077881,
+      "learning_rate": 3.3869239013933554e-05,
+      "loss": 4.4178,
+      "step": 776
+    },
+    {
+      "epoch": 0.8327974276527331,
+      "grad_norm": 2.661226272583008,
+      "learning_rate": 3.3654876741693466e-05,
+      "loss": 4.3274,
+      "step": 777
+    },
+    {
+      "epoch": 0.8338692390139335,
+      "grad_norm": 2.7595772743225098,
+      "learning_rate": 3.344051446945338e-05,
+      "loss": 4.5459,
+      "step": 778
+    },
+    {
+      "epoch": 0.834941050375134,
+      "grad_norm": 2.113373041152954,
+      "learning_rate": 3.322615219721329e-05,
+      "loss": 4.3255,
+      "step": 779
+    },
+    {
+      "epoch": 0.8360128617363344,
+      "grad_norm": 2.7816619873046875,
+      "learning_rate": 3.301178992497321e-05,
+      "loss": 4.3316,
+      "step": 780
+    },
+    {
+      "epoch": 0.8370846730975349,
+      "grad_norm": 3.1241683959960938,
+      "learning_rate": 3.279742765273312e-05,
+      "loss": 4.4017,
+      "step": 781
+    },
+    {
+      "epoch": 0.8381564844587353,
+      "grad_norm": 3.018803119659424,
+      "learning_rate": 3.2583065380493035e-05,
+      "loss": 4.1559,
+      "step": 782
+    },
+    {
+      "epoch": 0.8392282958199357,
+      "grad_norm": 2.408975601196289,
+      "learning_rate": 3.2368703108252954e-05,
+      "loss": 4.3322,
+      "step": 783
+    },
+    {
+      "epoch": 0.8403001071811361,
+      "grad_norm": 2.97987961769104,
+      "learning_rate": 3.2154340836012867e-05,
+      "loss": 4.2923,
+      "step": 784
+    },
+    {
+      "epoch": 0.8413719185423365,
+      "grad_norm": 2.9940474033355713,
+      "learning_rate": 3.193997856377278e-05,
+      "loss": 4.4587,
+      "step": 785
+    },
+    {
+      "epoch": 0.842443729903537,
+      "grad_norm": 2.686958074569702,
+      "learning_rate": 3.172561629153269e-05,
+      "loss": 4.4801,
+      "step": 786
+    },
+    {
+      "epoch": 0.8435155412647374,
+      "grad_norm": 1.6722944974899292,
+      "learning_rate": 3.151125401929261e-05,
+      "loss": 4.3032,
+      "step": 787
+    },
+    {
+      "epoch": 0.8445873526259379,
+      "grad_norm": 4.333474636077881,
+      "learning_rate": 3.129689174705252e-05,
+      "loss": 4.409,
+      "step": 788
+    },
+    {
+      "epoch": 0.8456591639871383,
+      "grad_norm": 3.1574814319610596,
+      "learning_rate": 3.1082529474812435e-05,
+      "loss": 4.6209,
+      "step": 789
+    },
+    {
+      "epoch": 0.8467309753483387,
+      "grad_norm": 2.261394739151001,
+      "learning_rate": 3.0868167202572354e-05,
+      "loss": 4.3995,
+      "step": 790
+    },
+    {
+      "epoch": 0.8478027867095391,
+      "grad_norm": 2.2780826091766357,
+      "learning_rate": 3.065380493033227e-05,
+      "loss": 4.2624,
+      "step": 791
+    },
+    {
+      "epoch": 0.8488745980707395,
+      "grad_norm": 2.5379581451416016,
+      "learning_rate": 3.0439442658092176e-05,
+      "loss": 4.5917,
+      "step": 792
+    },
+    {
+      "epoch": 0.84994640943194,
+      "grad_norm": 3.24279522895813,
+      "learning_rate": 3.0225080385852088e-05,
+      "loss": 4.4947,
+      "step": 793
+    },
+    {
+      "epoch": 0.8510182207931404,
+      "grad_norm": 3.2784323692321777,
+      "learning_rate": 3.0010718113612007e-05,
+      "loss": 4.3983,
+      "step": 794
+    },
+    {
+      "epoch": 0.8520900321543409,
+      "grad_norm": 2.6092724800109863,
+      "learning_rate": 2.979635584137192e-05,
+      "loss": 4.1586,
+      "step": 795
+    },
+    {
+      "epoch": 0.8531618435155413,
+      "grad_norm": 3.3995344638824463,
+      "learning_rate": 2.9581993569131832e-05,
+      "loss": 4.5867,
+      "step": 796
+    },
+    {
+      "epoch": 0.8542336548767417,
+      "grad_norm": 2.998579978942871,
+      "learning_rate": 2.936763129689175e-05,
+      "loss": 4.6124,
+      "step": 797
+    },
+    {
+      "epoch": 0.8553054662379421,
+      "grad_norm": 3.120178699493408,
+      "learning_rate": 2.9153269024651663e-05,
+      "loss": 4.5091,
+      "step": 798
+    },
+    {
+      "epoch": 0.8563772775991425,
+      "grad_norm": 3.543617010116577,
+      "learning_rate": 2.8938906752411576e-05,
+      "loss": 4.6284,
+      "step": 799
+    },
+    {
+      "epoch": 0.857449088960343,
+      "grad_norm": 3.2729685306549072,
+      "learning_rate": 2.8724544480171488e-05,
+      "loss": 4.2905,
+      "step": 800
+    },
+    {
+      "epoch": 0.8585209003215434,
+      "grad_norm": 3.9517574310302734,
+      "learning_rate": 2.8510182207931407e-05,
+      "loss": 4.7788,
+      "step": 801
+    },
+    {
+      "epoch": 0.8595927116827439,
+      "grad_norm": 3.896101713180542,
+      "learning_rate": 2.829581993569132e-05,
+      "loss": 4.4362,
+      "step": 802
+    },
+    {
+      "epoch": 0.8606645230439443,
+      "grad_norm": 3.16723370552063,
+      "learning_rate": 2.8081457663451232e-05,
+      "loss": 4.5719,
+      "step": 803
+    },
+    {
+      "epoch": 0.8617363344051447,
+      "grad_norm": 3.9187586307525635,
+      "learning_rate": 2.7867095391211144e-05,
+      "loss": 4.4248,
+      "step": 804
+    },
+    {
+      "epoch": 0.8628081457663451,
+      "grad_norm": 2.844324827194214,
+      "learning_rate": 2.7652733118971064e-05,
+      "loss": 4.3487,
+      "step": 805
+    },
+    {
+      "epoch": 0.8638799571275455,
+      "grad_norm": 2.617323398590088,
+      "learning_rate": 2.7438370846730976e-05,
+      "loss": 4.5764,
+      "step": 806
+    },
+    {
+      "epoch": 0.864951768488746,
+      "grad_norm": 3.5526764392852783,
+      "learning_rate": 2.722400857449089e-05,
+      "loss": 4.5931,
+      "step": 807
+    },
+    {
+      "epoch": 0.8660235798499464,
+      "grad_norm": 2.6362595558166504,
+      "learning_rate": 2.7009646302250807e-05,
+      "loss": 4.3441,
+      "step": 808
+    },
+    {
+      "epoch": 0.8670953912111469,
+      "grad_norm": 3.744828701019287,
+      "learning_rate": 2.679528403001072e-05,
+      "loss": 4.4955,
+      "step": 809
+    },
+    {
+      "epoch": 0.8681672025723473,
+      "grad_norm": 2.8340702056884766,
+      "learning_rate": 2.6580921757770632e-05,
+      "loss": 4.2257,
+      "step": 810
+    },
+    {
+      "epoch": 0.8692390139335477,
+      "grad_norm": 2.4332375526428223,
+      "learning_rate": 2.6366559485530545e-05,
+      "loss": 4.1914,
+      "step": 811
+    },
+    {
+      "epoch": 0.8703108252947481,
+      "grad_norm": 2.8110697269439697,
+      "learning_rate": 2.6152197213290464e-05,
+      "loss": 4.4457,
+      "step": 812
+    },
+    {
+      "epoch": 0.8713826366559485,
+      "grad_norm": 3.533874273300171,
+      "learning_rate": 2.5937834941050376e-05,
+      "loss": 4.7375,
+      "step": 813
+    },
+    {
+      "epoch": 0.872454448017149,
+      "grad_norm": 3.4596922397613525,
+      "learning_rate": 2.572347266881029e-05,
+      "loss": 4.3842,
+      "step": 814
+    },
+    {
+      "epoch": 0.8735262593783494,
+      "grad_norm": 4.416886329650879,
+      "learning_rate": 2.5509110396570208e-05,
+      "loss": 5.1532,
+      "step": 815
+    },
+    {
+      "epoch": 0.8745980707395499,
+      "grad_norm": 2.818956136703491,
+      "learning_rate": 2.529474812433012e-05,
+      "loss": 4.2577,
+      "step": 816
+    },
+    {
+      "epoch": 0.8756698821007503,
+      "grad_norm": 4.539247989654541,
+      "learning_rate": 2.5080385852090032e-05,
+      "loss": 4.3223,
+      "step": 817
+    },
+    {
+      "epoch": 0.8767416934619507,
+      "grad_norm": 4.091986656188965,
+      "learning_rate": 2.4866023579849948e-05,
+      "loss": 4.4761,
+      "step": 818
+    },
+    {
+      "epoch": 0.8778135048231511,
+      "grad_norm": 2.9086873531341553,
+      "learning_rate": 2.465166130760986e-05,
+      "loss": 4.3534,
+      "step": 819
+    },
+    {
+      "epoch": 0.8788853161843515,
+      "grad_norm": 3.3018639087677,
+      "learning_rate": 2.4437299035369776e-05,
+      "loss": 4.487,
+      "step": 820
+    },
+    {
+      "epoch": 0.879957127545552,
+      "grad_norm": 2.7818877696990967,
+      "learning_rate": 2.4222936763129692e-05,
+      "loss": 4.5688,
+      "step": 821
+    },
+    {
+      "epoch": 0.8810289389067524,
+      "grad_norm": 2.4334094524383545,
+      "learning_rate": 2.4008574490889604e-05,
+      "loss": 4.8955,
+      "step": 822
+    },
+    {
+      "epoch": 0.8821007502679529,
+      "grad_norm": 2.775305986404419,
+      "learning_rate": 2.379421221864952e-05,
+      "loss": 4.2866,
+      "step": 823
+    },
+    {
+      "epoch": 0.8831725616291533,
+      "grad_norm": 3.648191213607788,
+      "learning_rate": 2.3579849946409433e-05,
+      "loss": 4.2068,
+      "step": 824
+    },
+    {
+      "epoch": 0.8842443729903537,
+      "grad_norm": 4.619109630584717,
+      "learning_rate": 2.3365487674169348e-05,
+      "loss": 4.4062,
+      "step": 825
+    },
+    {
+      "epoch": 0.8853161843515541,
+      "grad_norm": 3.174395799636841,
+      "learning_rate": 2.315112540192926e-05,
+      "loss": 4.199,
+      "step": 826
+    },
+    {
+      "epoch": 0.8863879957127545,
+      "grad_norm": 3.6223504543304443,
+      "learning_rate": 2.2936763129689176e-05,
+      "loss": 4.3869,
+      "step": 827
+    },
+    {
+      "epoch": 0.887459807073955,
+      "grad_norm": 3.9439122676849365,
+      "learning_rate": 2.272240085744909e-05,
+      "loss": 5.0497,
+      "step": 828
+    },
+    {
+      "epoch": 0.8885316184351554,
+      "grad_norm": 3.0752944946289062,
+      "learning_rate": 2.2508038585209005e-05,
+      "loss": 4.4295,
+      "step": 829
+    },
+    {
+      "epoch": 0.8896034297963559,
+      "grad_norm": 2.4576616287231445,
+      "learning_rate": 2.229367631296892e-05,
+      "loss": 4.7185,
+      "step": 830
+    },
+    {
+      "epoch": 0.8906752411575563,
+      "grad_norm": 2.447382688522339,
+      "learning_rate": 2.2079314040728833e-05,
+      "loss": 4.1357,
+      "step": 831
+    },
+    {
+      "epoch": 0.8917470525187567,
+      "grad_norm": 2.2739741802215576,
+      "learning_rate": 2.186495176848875e-05,
+      "loss": 3.9939,
+      "step": 832
+    },
+    {
+      "epoch": 0.8928188638799571,
+      "grad_norm": 3.1691207885742188,
+      "learning_rate": 2.165058949624866e-05,
+      "loss": 4.354,
+      "step": 833
+    },
+    {
+      "epoch": 0.8938906752411575,
+      "grad_norm": 2.7398335933685303,
+      "learning_rate": 2.1436227224008577e-05,
+      "loss": 4.2388,
+      "step": 834
+    },
+    {
+      "epoch": 0.894962486602358,
+      "grad_norm": 2.907942056655884,
+      "learning_rate": 2.122186495176849e-05,
+      "loss": 4.4821,
+      "step": 835
+    },
+    {
+      "epoch": 0.8960342979635584,
+      "grad_norm": 3.2182884216308594,
+      "learning_rate": 2.1007502679528405e-05,
+      "loss": 4.8471,
+      "step": 836
+    },
+    {
+      "epoch": 0.8971061093247589,
+      "grad_norm": 3.2632901668548584,
+      "learning_rate": 2.079314040728832e-05,
+      "loss": 4.7403,
+      "step": 837
+    },
+    {
+      "epoch": 0.8981779206859593,
+      "grad_norm": 2.8337507247924805,
+      "learning_rate": 2.0578778135048233e-05,
+      "loss": 4.8651,
+      "step": 838
+    },
+    {
+      "epoch": 0.8992497320471597,
+      "grad_norm": 2.620939254760742,
+      "learning_rate": 2.036441586280815e-05,
+      "loss": 4.1015,
+      "step": 839
+    },
+    {
+      "epoch": 0.9003215434083601,
+      "grad_norm": 3.049180746078491,
+      "learning_rate": 2.015005359056806e-05,
+      "loss": 4.6333,
+      "step": 840
+    },
+    {
+      "epoch": 0.9013933547695605,
+      "grad_norm": 2.508197546005249,
+      "learning_rate": 1.9935691318327977e-05,
+      "loss": 4.1966,
+      "step": 841
+    },
+    {
+      "epoch": 0.902465166130761,
+      "grad_norm": 4.801202297210693,
+      "learning_rate": 1.972132904608789e-05,
+      "loss": 4.8339,
+      "step": 842
+    },
+    {
+      "epoch": 0.9035369774919614,
+      "grad_norm": 2.1153371334075928,
+      "learning_rate": 1.9506966773847805e-05,
+      "loss": 4.0906,
+      "step": 843
+    },
+    {
+      "epoch": 0.9046087888531619,
+      "grad_norm": 4.357201099395752,
+      "learning_rate": 1.9292604501607717e-05,
+      "loss": 4.6109,
+      "step": 844
+    },
+    {
+      "epoch": 0.9056806002143623,
+      "grad_norm": 4.721092224121094,
+      "learning_rate": 1.9078242229367633e-05,
+      "loss": 4.1943,
+      "step": 845
+    },
+    {
+      "epoch": 0.9067524115755627,
+      "grad_norm": 2.453629970550537,
+      "learning_rate": 1.886387995712755e-05,
+      "loss": 4.5132,
+      "step": 846
+    },
+    {
+      "epoch": 0.9078242229367631,
+      "grad_norm": 2.598621129989624,
+      "learning_rate": 1.864951768488746e-05,
+      "loss": 4.2245,
+      "step": 847
+    },
+    {
+      "epoch": 0.9088960342979635,
+      "grad_norm": 2.0761516094207764,
+      "learning_rate": 1.8435155412647377e-05,
+      "loss": 4.2331,
+      "step": 848
+    },
+    {
+      "epoch": 0.909967845659164,
+      "grad_norm": 2.663645029067993,
+      "learning_rate": 1.822079314040729e-05,
+      "loss": 4.2554,
+      "step": 849
+    },
+    {
+      "epoch": 0.9110396570203644,
+      "grad_norm": 2.794361114501953,
+      "learning_rate": 1.8006430868167205e-05,
+      "loss": 4.0979,
+      "step": 850
+    },
+    {
+      "epoch": 0.9121114683815649,
+      "grad_norm": 4.002750396728516,
+      "learning_rate": 1.7792068595927117e-05,
+      "loss": 4.156,
+      "step": 851
+    },
+    {
+      "epoch": 0.9131832797427653,
+      "grad_norm": 2.1327261924743652,
+      "learning_rate": 1.7577706323687033e-05,
+      "loss": 4.2994,
+      "step": 852
+    },
+    {
+      "epoch": 0.9142550911039657,
+      "grad_norm": 2.218019723892212,
+      "learning_rate": 1.736334405144695e-05,
+      "loss": 4.2681,
+      "step": 853
+    },
+    {
+      "epoch": 0.9153269024651661,
+      "grad_norm": 2.634157419204712,
+      "learning_rate": 1.714898177920686e-05,
+      "loss": 4.457,
+      "step": 854
+    },
+    {
+      "epoch": 0.9163987138263665,
+      "grad_norm": 3.584799289703369,
+      "learning_rate": 1.6934619506966777e-05,
+      "loss": 4.4086,
+      "step": 855
+    },
+    {
+      "epoch": 0.917470525187567,
+      "grad_norm": 3.640207052230835,
+      "learning_rate": 1.672025723472669e-05,
+      "loss": 4.2253,
+      "step": 856
+    },
+    {
+      "epoch": 0.9185423365487674,
+      "grad_norm": 2.656949758529663,
+      "learning_rate": 1.6505894962486605e-05,
+      "loss": 4.3226,
+      "step": 857
+    },
+    {
+      "epoch": 0.9196141479099679,
+      "grad_norm": 2.573221445083618,
+      "learning_rate": 1.6291532690246518e-05,
+      "loss": 4.6633,
+      "step": 858
+    },
+    {
+      "epoch": 0.9206859592711683,
+      "grad_norm": 2.5666069984436035,
+      "learning_rate": 1.6077170418006433e-05,
+      "loss": 4.4142,
+      "step": 859
+    },
+    {
+      "epoch": 0.9217577706323687,
+      "grad_norm": 2.6048295497894287,
+      "learning_rate": 1.5862808145766346e-05,
+      "loss": 4.5767,
+      "step": 860
+    },
+    {
+      "epoch": 0.9228295819935691,
+      "grad_norm": 3.020139694213867,
+      "learning_rate": 1.564844587352626e-05,
+      "loss": 4.2079,
+      "step": 861
+    },
+    {
+      "epoch": 0.9239013933547695,
+      "grad_norm": 3.268181085586548,
+      "learning_rate": 1.5434083601286177e-05,
+      "loss": 4.2529,
+      "step": 862
+    },
+    {
+      "epoch": 0.92497320471597,
+      "grad_norm": 1.9898625612258911,
+      "learning_rate": 1.5219721329046088e-05,
+      "loss": 4.2561,
+      "step": 863
+    },
+    {
+      "epoch": 0.9260450160771704,
+      "grad_norm": 2.4188132286071777,
+      "learning_rate": 1.5005359056806004e-05,
+      "loss": 4.4644,
+      "step": 864
+    },
+    {
+      "epoch": 0.9271168274383709,
+      "grad_norm": 3.1744115352630615,
+      "learning_rate": 1.4790996784565916e-05,
+      "loss": 4.2628,
+      "step": 865
+    },
+    {
+      "epoch": 0.9281886387995713,
+      "grad_norm": 3.263568639755249,
+      "learning_rate": 1.4576634512325832e-05,
+      "loss": 4.2172,
+      "step": 866
+    },
+    {
+      "epoch": 0.9292604501607717,
+      "grad_norm": 2.486051082611084,
+      "learning_rate": 1.4362272240085744e-05,
+      "loss": 4.1906,
+      "step": 867
+    },
+    {
+      "epoch": 0.9303322615219721,
+      "grad_norm": 3.1796326637268066,
+      "learning_rate": 1.414790996784566e-05,
+      "loss": 4.4266,
+      "step": 868
+    },
+    {
+      "epoch": 0.9314040728831725,
+      "grad_norm": 4.484216690063477,
+      "learning_rate": 1.3933547695605572e-05,
+      "loss": 4.5339,
+      "step": 869
+    },
+    {
+      "epoch": 0.932475884244373,
+      "grad_norm": 2.6323423385620117,
+      "learning_rate": 1.3719185423365488e-05,
+      "loss": 4.3935,
+      "step": 870
+    },
+    {
+      "epoch": 0.9335476956055734,
+      "grad_norm": 3.191985845565796,
+      "learning_rate": 1.3504823151125404e-05,
+      "loss": 4.3227,
+      "step": 871
+    },
+    {
+      "epoch": 0.9346195069667739,
+      "grad_norm": 4.297986030578613,
+      "learning_rate": 1.3290460878885316e-05,
+      "loss": 4.2514,
+      "step": 872
+    },
+    {
+      "epoch": 0.9356913183279743,
+      "grad_norm": 3.8759825229644775,
+      "learning_rate": 1.3076098606645232e-05,
+      "loss": 4.1061,
+      "step": 873
+    },
+    {
+      "epoch": 0.9367631296891747,
+      "grad_norm": 2.874577283859253,
+      "learning_rate": 1.2861736334405144e-05,
+      "loss": 4.3546,
+      "step": 874
+    },
+    {
+      "epoch": 0.9378349410503751,
+      "grad_norm": 2.8280208110809326,
+      "learning_rate": 1.264737406216506e-05,
+      "loss": 4.2303,
+      "step": 875
+    },
+    {
+      "epoch": 0.9389067524115756,
+      "grad_norm": 5.017232418060303,
+      "learning_rate": 1.2433011789924974e-05,
+      "loss": 4.4325,
+      "step": 876
+    },
+    {
+      "epoch": 0.939978563772776,
+      "grad_norm": 3.0879180431365967,
+      "learning_rate": 1.2218649517684888e-05,
+      "loss": 4.3345,
+      "step": 877
+    },
+    {
+      "epoch": 0.9410503751339764,
+      "grad_norm": 5.276310920715332,
+      "learning_rate": 1.2004287245444802e-05,
+      "loss": 4.195,
+      "step": 878
+    },
+    {
+      "epoch": 0.9421221864951769,
+      "grad_norm": 3.4308736324310303,
+      "learning_rate": 1.1789924973204716e-05,
+      "loss": 4.1783,
+      "step": 879
+    },
+    {
+      "epoch": 0.9431939978563773,
+      "grad_norm": 4.4123430252075195,
+      "learning_rate": 1.157556270096463e-05,
+      "loss": 4.5029,
+      "step": 880
+    },
+    {
+      "epoch": 0.9442658092175777,
+      "grad_norm": 2.261904239654541,
+      "learning_rate": 1.1361200428724544e-05,
+      "loss": 4.2614,
+      "step": 881
+    },
+    {
+      "epoch": 0.9453376205787781,
+      "grad_norm": 2.2538812160491943,
+      "learning_rate": 1.114683815648446e-05,
+      "loss": 4.2999,
+      "step": 882
+    },
+    {
+      "epoch": 0.9464094319399786,
+      "grad_norm": 3.461613893508911,
+      "learning_rate": 1.0932475884244374e-05,
+      "loss": 4.6115,
+      "step": 883
+    },
+    {
+      "epoch": 0.947481243301179,
+      "grad_norm": 3.164299488067627,
+      "learning_rate": 1.0718113612004288e-05,
+      "loss": 4.0643,
+      "step": 884
+    },
+    {
+      "epoch": 0.9485530546623794,
+      "grad_norm": 2.1744544506073,
+      "learning_rate": 1.0503751339764202e-05,
+      "loss": 4.3665,
+      "step": 885
+    },
+    {
+      "epoch": 0.9496248660235799,
+      "grad_norm": 2.9400393962860107,
+      "learning_rate": 1.0289389067524116e-05,
+      "loss": 3.9939,
+      "step": 886
+    },
+    {
+      "epoch": 0.9506966773847803,
+      "grad_norm": 2.9955875873565674,
+      "learning_rate": 1.007502679528403e-05,
+      "loss": 4.2878,
+      "step": 887
+    },
+    {
+      "epoch": 0.9517684887459807,
+      "grad_norm": 2.6207101345062256,
+      "learning_rate": 9.860664523043945e-06,
+      "loss": 4.0709,
+      "step": 888
+    },
+    {
+      "epoch": 0.9528403001071811,
+      "grad_norm": 2.4389564990997314,
+      "learning_rate": 9.646302250803859e-06,
+      "loss": 4.3193,
+      "step": 889
+    },
+    {
+      "epoch": 0.9539121114683816,
+      "grad_norm": 3.457326889038086,
+      "learning_rate": 9.431939978563774e-06,
+      "loss": 4.7996,
+      "step": 890
+    },
+    {
+      "epoch": 0.954983922829582,
+      "grad_norm": 2.875967502593994,
+      "learning_rate": 9.217577706323688e-06,
+      "loss": 4.5278,
+      "step": 891
+    },
+    {
+      "epoch": 0.9560557341907824,
+      "grad_norm": 4.708974838256836,
+      "learning_rate": 9.003215434083602e-06,
+      "loss": 4.2928,
+      "step": 892
+    },
+    {
+      "epoch": 0.9571275455519829,
+      "grad_norm": 2.4193968772888184,
+      "learning_rate": 8.788853161843517e-06,
+      "loss": 4.0673,
+      "step": 893
+    },
+    {
+      "epoch": 0.9581993569131833,
+      "grad_norm": 2.4473676681518555,
+      "learning_rate": 8.57449088960343e-06,
+      "loss": 4.3235,
+      "step": 894
+    },
+    {
+      "epoch": 0.9592711682743837,
+      "grad_norm": 2.7152135372161865,
+      "learning_rate": 8.360128617363345e-06,
+      "loss": 4.3448,
+      "step": 895
+    },
+    {
+      "epoch": 0.9603429796355841,
+      "grad_norm": 2.9722485542297363,
+      "learning_rate": 8.145766345123259e-06,
+      "loss": 4.6859,
+      "step": 896
+    },
+    {
+      "epoch": 0.9614147909967846,
+      "grad_norm": 4.064876079559326,
+      "learning_rate": 7.931404072883173e-06,
+      "loss": 4.3012,
+      "step": 897
+    },
+    {
+      "epoch": 0.962486602357985,
+      "grad_norm": 2.750589370727539,
+      "learning_rate": 7.717041800643089e-06,
+      "loss": 4.506,
+      "step": 898
+    },
+    {
+      "epoch": 0.9635584137191854,
+      "grad_norm": 3.3813419342041016,
+      "learning_rate": 7.502679528403002e-06,
+      "loss": 4.2005,
+      "step": 899
+    },
+    {
+      "epoch": 0.9646302250803859,
+      "grad_norm": 3.6827099323272705,
+      "learning_rate": 7.288317256162916e-06,
+      "loss": 4.8387,
+      "step": 900
+    },
+    {
+      "epoch": 0.9657020364415863,
+      "grad_norm": 2.949695110321045,
+      "learning_rate": 7.07395498392283e-06,
+      "loss": 4.3218,
+      "step": 901
+    },
+    {
+      "epoch": 0.9667738478027867,
+      "grad_norm": 3.40405535697937,
+      "learning_rate": 6.859592711682744e-06,
+      "loss": 4.2812,
+      "step": 902
+    },
+    {
+      "epoch": 0.9678456591639871,
+      "grad_norm": 3.3691530227661133,
+      "learning_rate": 6.645230439442658e-06,
+      "loss": 4.2832,
+      "step": 903
+    },
+    {
+      "epoch": 0.9689174705251876,
+      "grad_norm": 3.336270332336426,
+      "learning_rate": 6.430868167202572e-06,
+      "loss": 4.2197,
+      "step": 904
+    },
+    {
+      "epoch": 0.969989281886388,
+      "grad_norm": 2.827868700027466,
+      "learning_rate": 6.216505894962487e-06,
+      "loss": 4.3922,
+      "step": 905
+    },
+    {
+      "epoch": 0.9710610932475884,
+      "grad_norm": 4.273472309112549,
+      "learning_rate": 6.002143622722401e-06,
+      "loss": 4.2416,
+      "step": 906
+    },
+    {
+      "epoch": 0.9721329046087889,
+      "grad_norm": 2.3178086280822754,
+      "learning_rate": 5.787781350482315e-06,
+      "loss": 4.6557,
+      "step": 907
+    },
+    {
+      "epoch": 0.9732047159699893,
+      "grad_norm": 2.3149664402008057,
+      "learning_rate": 5.57341907824223e-06,
+      "loss": 4.2855,
+      "step": 908
+    },
+    {
+      "epoch": 0.9742765273311897,
+      "grad_norm": 3.162186622619629,
+      "learning_rate": 5.359056806002144e-06,
+      "loss": 4.4175,
+      "step": 909
+    },
+    {
+      "epoch": 0.9753483386923901,
+      "grad_norm": 3.124436378479004,
+      "learning_rate": 5.144694533762058e-06,
+      "loss": 4.3642,
+      "step": 910
+    },
+    {
+      "epoch": 0.9764201500535906,
+      "grad_norm": 3.431694984436035,
+      "learning_rate": 4.930332261521972e-06,
+      "loss": 4.5494,
+      "step": 911
+    },
+    {
+      "epoch": 0.977491961414791,
+      "grad_norm": 4.276769638061523,
+      "learning_rate": 4.715969989281887e-06,
+      "loss": 4.1405,
+      "step": 912
+    },
+    {
+      "epoch": 0.9785637727759914,
+      "grad_norm": 2.1469695568084717,
+      "learning_rate": 4.501607717041801e-06,
+      "loss": 4.5683,
+      "step": 913
+    },
+    {
+      "epoch": 0.9796355841371919,
+      "grad_norm": 4.689246654510498,
+      "learning_rate": 4.287245444801715e-06,
+      "loss": 4.4474,
+      "step": 914
+    },
+    {
+      "epoch": 0.9807073954983923,
+      "grad_norm": 2.666457176208496,
+      "learning_rate": 4.072883172561629e-06,
+      "loss": 4.4224,
+      "step": 915
+    },
+    {
+      "epoch": 0.9817792068595927,
+      "grad_norm": 4.759450435638428,
+      "learning_rate": 3.858520900321544e-06,
+      "loss": 4.5016,
+      "step": 916
+    },
+    {
+      "epoch": 0.9828510182207931,
+      "grad_norm": 4.514771938323975,
+      "learning_rate": 3.644158628081458e-06,
+      "loss": 4.4132,
+      "step": 917
+    },
+    {
+      "epoch": 0.9839228295819936,
+      "grad_norm": 2.440972089767456,
+      "learning_rate": 3.429796355841372e-06,
+      "loss": 4.2929,
+      "step": 918
+    },
+    {
+      "epoch": 0.984994640943194,
+      "grad_norm": 2.3493099212646484,
+      "learning_rate": 3.215434083601286e-06,
+      "loss": 4.191,
+      "step": 919
+    },
+    {
+      "epoch": 0.9860664523043944,
+      "grad_norm": 3.0151264667510986,
+      "learning_rate": 3.0010718113612005e-06,
+      "loss": 4.3449,
+      "step": 920
+    },
+    {
+      "epoch": 0.9871382636655949,
+      "grad_norm": 2.319638252258301,
+      "learning_rate": 2.786709539121115e-06,
+      "loss": 4.0492,
+      "step": 921
+    },
+    {
+      "epoch": 0.9882100750267953,
+      "grad_norm": 1.8680282831192017,
+      "learning_rate": 2.572347266881029e-06,
+      "loss": 4.1374,
+      "step": 922
+    },
+    {
+      "epoch": 0.9892818863879957,
+      "grad_norm": 5.088656425476074,
+      "learning_rate": 2.3579849946409436e-06,
+      "loss": 4.3019,
+      "step": 923
+    },
+    {
+      "epoch": 0.9903536977491961,
+      "grad_norm": 3.4761383533477783,
+      "learning_rate": 2.1436227224008577e-06,
+      "loss": 4.3048,
+      "step": 924
+    },
+    {
+      "epoch": 0.9914255091103966,
+      "grad_norm": 2.1858043670654297,
+      "learning_rate": 1.929260450160772e-06,
+      "loss": 4.239,
+      "step": 925
+    },
+    {
+      "epoch": 0.992497320471597,
+      "grad_norm": 3.2120463848114014,
+      "learning_rate": 1.714898177920686e-06,
+      "loss": 4.2429,
+      "step": 926
+    },
+    {
+      "epoch": 0.9935691318327974,
+      "grad_norm": 2.1242594718933105,
+      "learning_rate": 1.5005359056806003e-06,
+      "loss": 4.0712,
+      "step": 927
+    },
+    {
+      "epoch": 0.9946409431939979,
+      "grad_norm": 2.9433510303497314,
+      "learning_rate": 1.2861736334405146e-06,
+      "loss": 4.0293,
+      "step": 928
+    },
+    {
+      "epoch": 0.9957127545551983,
+      "grad_norm": 4.833005905151367,
+      "learning_rate": 1.0718113612004288e-06,
+      "loss": 4.081,
+      "step": 929
+    },
+    {
+      "epoch": 0.9967845659163987,
+      "grad_norm": 3.332620620727539,
+      "learning_rate": 8.57449088960343e-07,
+      "loss": 4.4352,
+      "step": 930
+    },
+    {
+      "epoch": 0.9978563772775991,
+      "grad_norm": 2.7460505962371826,
+      "learning_rate": 6.430868167202573e-07,
+      "loss": 4.176,
+      "step": 931
+    },
+    {
+      "epoch": 0.9989281886387996,
+      "grad_norm": 3.637892961502075,
+      "learning_rate": 4.287245444801715e-07,
+      "loss": 4.3091,
+      "step": 932
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8850045204162598,
+      "learning_rate": 2.1436227224008575e-07,
+      "loss": 4.2321,
+      "step": 933
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 933,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1048339206640128.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}