diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18044 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 200,
+  "global_step": 2558,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007818608287724785,
+      "grad_norm": 0.11400622129440308,
+      "learning_rate": 3.90625e-07,
+      "loss": 0.5189,
+      "step": 1
+    },
+    {
+      "epoch": 0.001563721657544957,
+      "grad_norm": 0.28361040353775024,
+      "learning_rate": 7.8125e-07,
+      "loss": 0.7211,
+      "step": 2
+    },
+    {
+      "epoch": 0.0023455824863174357,
+      "grad_norm": 0.12066957354545593,
+      "learning_rate": 1.1718750000000001e-06,
+      "loss": 0.4869,
+      "step": 3
+    },
+    {
+      "epoch": 0.003127443315089914,
+      "grad_norm": 0.14368808269500732,
+      "learning_rate": 1.5625e-06,
+      "loss": 0.6068,
+      "step": 4
+    },
+    {
+      "epoch": 0.003909304143862392,
+      "grad_norm": 0.17054139077663422,
+      "learning_rate": 1.953125e-06,
+      "loss": 0.69,
+      "step": 5
+    },
+    {
+      "epoch": 0.004691164972634871,
+      "grad_norm": 0.13562728464603424,
+      "learning_rate": 2.3437500000000002e-06,
+      "loss": 0.5394,
+      "step": 6
+    },
+    {
+      "epoch": 0.00547302580140735,
+      "grad_norm": 0.1532646119594574,
+      "learning_rate": 2.734375e-06,
+      "loss": 0.5898,
+      "step": 7
+    },
+    {
+      "epoch": 0.006254886630179828,
+      "grad_norm": 0.1191975548863411,
+      "learning_rate": 3.125e-06,
+      "loss": 0.5587,
+      "step": 8
+    },
+    {
+      "epoch": 0.007036747458952306,
+      "grad_norm": 0.13896405696868896,
+      "learning_rate": 3.5156250000000003e-06,
+      "loss": 0.6037,
+      "step": 9
+    },
+    {
+      "epoch": 0.007818608287724784,
+      "grad_norm": 0.12083485722541809,
+      "learning_rate": 3.90625e-06,
+      "loss": 0.516,
+      "step": 10
+    },
+    {
+      "epoch": 0.008600469116497263,
+      "grad_norm": 0.2188330888748169,
+      "learning_rate": 4.296875e-06,
+      "loss": 0.746,
+      "step": 11
+    },
+    {
+      "epoch": 0.009382329945269743,
+      "grad_norm": 0.1677808314561844,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 0.5886,
+      "step": 12
+    },
+    {
+      "epoch": 0.010164190774042221,
+      "grad_norm": 0.1499367654323578,
+      "learning_rate": 5.078125000000001e-06,
+      "loss": 0.6232,
+      "step": 13
+    },
+    {
+      "epoch": 0.0109460516028147,
+      "grad_norm": 0.14739449322223663,
+      "learning_rate": 5.46875e-06,
+      "loss": 0.5995,
+      "step": 14
+    },
+    {
+      "epoch": 0.011727912431587178,
+      "grad_norm": 0.15768533945083618,
+      "learning_rate": 5.859375e-06,
+      "loss": 0.6123,
+      "step": 15
+    },
+    {
+      "epoch": 0.012509773260359656,
+      "grad_norm": 0.1611998975276947,
+      "learning_rate": 6.25e-06,
+      "loss": 0.5923,
+      "step": 16
+    },
+    {
+      "epoch": 0.013291634089132134,
+      "grad_norm": 0.19843457639217377,
+      "learning_rate": 6.6406250000000005e-06,
+      "loss": 0.6369,
+      "step": 17
+    },
+    {
+      "epoch": 0.014073494917904612,
+      "grad_norm": 0.1513635814189911,
+      "learning_rate": 7.031250000000001e-06,
+      "loss": 0.6138,
+      "step": 18
+    },
+    {
+      "epoch": 0.014855355746677092,
+      "grad_norm": 0.15541641414165497,
+      "learning_rate": 7.421875e-06,
+      "loss": 0.5324,
+      "step": 19
+    },
+    {
+      "epoch": 0.01563721657544957,
+      "grad_norm": 0.2085491418838501,
+      "learning_rate": 7.8125e-06,
+      "loss": 0.602,
+      "step": 20
+    },
+    {
+      "epoch": 0.01641907740422205,
+      "grad_norm": 0.21879960596561432,
+      "learning_rate": 8.203125000000001e-06,
+      "loss": 0.5874,
+      "step": 21
+    },
+    {
+      "epoch": 0.017200938232994525,
+      "grad_norm": 0.21974720060825348,
+      "learning_rate": 8.59375e-06,
+      "loss": 0.7044,
+      "step": 22
+    },
+    {
+      "epoch": 0.017982799061767005,
+      "grad_norm": 0.32434722781181335,
+      "learning_rate": 8.984375e-06,
+      "loss": 0.6598,
+      "step": 23
+    },
+    {
+      "epoch": 0.018764659890539485,
+      "grad_norm": 0.20202045142650604,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.5392,
+      "step": 24
+    },
+    {
+      "epoch": 0.019546520719311962,
+      "grad_norm": 0.14759372174739838,
+      "learning_rate": 9.765625e-06,
+      "loss": 0.453,
+      "step": 25
+    },
+    {
+      "epoch": 0.020328381548084442,
+      "grad_norm": 0.28198951482772827,
+      "learning_rate": 1.0156250000000001e-05,
+      "loss": 0.6349,
+      "step": 26
+    },
+    {
+      "epoch": 0.02111024237685692,
+      "grad_norm": 0.2123844176530838,
+      "learning_rate": 1.0546875e-05,
+      "loss": 0.6462,
+      "step": 27
+    },
+    {
+      "epoch": 0.0218921032056294,
+      "grad_norm": 0.19585418701171875,
+      "learning_rate": 1.09375e-05,
+      "loss": 0.4884,
+      "step": 28
+    },
+    {
+      "epoch": 0.022673964034401875,
+      "grad_norm": 0.2947320342063904,
+      "learning_rate": 1.1328125000000001e-05,
+      "loss": 0.6294,
+      "step": 29
+    },
+    {
+      "epoch": 0.023455824863174355,
+      "grad_norm": 0.24433963000774384,
+      "learning_rate": 1.171875e-05,
+      "loss": 0.6965,
+      "step": 30
+    },
+    {
+      "epoch": 0.024237685691946835,
+      "grad_norm": 0.2690712511539459,
+      "learning_rate": 1.2109375000000001e-05,
+      "loss": 0.5956,
+      "step": 31
+    },
+    {
+      "epoch": 0.02501954652071931,
+      "grad_norm": 0.28039294481277466,
+      "learning_rate": 1.25e-05,
+      "loss": 0.4375,
+      "step": 32
+    },
+    {
+      "epoch": 0.02580140734949179,
+      "grad_norm": 0.30447208881378174,
+      "learning_rate": 1.2890625e-05,
+      "loss": 0.5704,
+      "step": 33
+    },
+    {
+      "epoch": 0.026583268178264268,
+      "grad_norm": 0.33207401633262634,
+      "learning_rate": 1.3281250000000001e-05,
+      "loss": 0.6211,
+      "step": 34
+    },
+    {
+      "epoch": 0.027365129007036748,
+      "grad_norm": 0.25266459584236145,
+      "learning_rate": 1.3671875e-05,
+      "loss": 0.5197,
+      "step": 35
+    },
+    {
+      "epoch": 0.028146989835809225,
+      "grad_norm": 0.1736496388912201,
+      "learning_rate": 1.4062500000000001e-05,
+      "loss": 0.4186,
+      "step": 36
+    },
+    {
+      "epoch": 0.028928850664581705,
+      "grad_norm": 0.36728987097740173,
+      "learning_rate": 1.4453125e-05,
+      "loss": 0.5726,
+      "step": 37
+    },
+    {
+      "epoch": 0.029710711493354185,
+      "grad_norm": 0.22769613564014435,
+      "learning_rate": 1.484375e-05,
+      "loss": 0.4935,
+      "step": 38
+    },
+    {
+      "epoch": 0.03049257232212666,
+      "grad_norm": 0.41446453332901,
+      "learning_rate": 1.5234375000000001e-05,
+      "loss": 0.5832,
+      "step": 39
+    },
+    {
+      "epoch": 0.03127443315089914,
+      "grad_norm": 0.24155667424201965,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.5682,
+      "step": 40
+    },
+    {
+      "epoch": 0.03205629397967162,
+      "grad_norm": 0.30262166261672974,
+      "learning_rate": 1.6015625e-05,
+      "loss": 0.5512,
+      "step": 41
+    },
+    {
+      "epoch": 0.0328381548084441,
+      "grad_norm": 0.21015244722366333,
+      "learning_rate": 1.6406250000000002e-05,
+      "loss": 0.4404,
+      "step": 42
+    },
+    {
+      "epoch": 0.033620015637216574,
+      "grad_norm": 0.3361029326915741,
+      "learning_rate": 1.6796875e-05,
+      "loss": 0.5782,
+      "step": 43
+    },
+    {
+      "epoch": 0.03440187646598905,
+      "grad_norm": 0.31037694215774536,
+      "learning_rate": 1.71875e-05,
+      "loss": 0.5791,
+      "step": 44
+    },
+    {
+      "epoch": 0.035183737294761534,
+      "grad_norm": 0.2708664834499359,
+      "learning_rate": 1.7578125000000002e-05,
+      "loss": 0.4308,
+      "step": 45
+    },
+    {
+      "epoch": 0.03596559812353401,
+      "grad_norm": 0.10701970010995865,
+      "learning_rate": 1.796875e-05,
+      "loss": 0.2914,
+      "step": 46
+    },
+    {
+      "epoch": 0.03674745895230649,
+      "grad_norm": 0.26695603132247925,
+      "learning_rate": 1.8359375e-05,
+      "loss": 0.5574,
+      "step": 47
+    },
+    {
+      "epoch": 0.03752931978107897,
+      "grad_norm": 0.1568523496389389,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.4056,
+      "step": 48
+    },
+    {
+      "epoch": 0.03831118060985145,
+      "grad_norm": 0.15718792378902435,
+      "learning_rate": 1.9140625e-05,
+      "loss": 0.3688,
+      "step": 49
+    },
+    {
+      "epoch": 0.039093041438623924,
+      "grad_norm": 0.2712760269641876,
+      "learning_rate": 1.953125e-05,
+      "loss": 0.5625,
+      "step": 50
+    },
+    {
+      "epoch": 0.0398749022673964,
+      "grad_norm": 0.23504135012626648,
+      "learning_rate": 1.9921875e-05,
+      "loss": 0.5107,
+      "step": 51
+    },
+    {
+      "epoch": 0.040656763096168884,
+      "grad_norm": 0.22949786484241486,
+      "learning_rate": 2.0312500000000002e-05,
+      "loss": 0.5666,
+      "step": 52
+    },
+    {
+      "epoch": 0.04143862392494136,
+      "grad_norm": 0.1631242036819458,
+      "learning_rate": 2.0703125e-05,
+      "loss": 0.4784,
+      "step": 53
+    },
+    {
+      "epoch": 0.04222048475371384,
+      "grad_norm": 0.17866574227809906,
+      "learning_rate": 2.109375e-05,
+      "loss": 0.407,
+      "step": 54
+    },
+    {
+      "epoch": 0.04300234558248632,
+      "grad_norm": 0.24938495457172394,
+      "learning_rate": 2.1484375000000002e-05,
+      "loss": 0.5188,
+      "step": 55
+    },
+    {
+      "epoch": 0.0437842064112588,
+      "grad_norm": 0.19512587785720825,
+      "learning_rate": 2.1875e-05,
+      "loss": 0.5769,
+      "step": 56
+    },
+    {
+      "epoch": 0.044566067240031274,
+      "grad_norm": 0.18683001399040222,
+      "learning_rate": 2.2265625e-05,
+      "loss": 0.3868,
+      "step": 57
+    },
+    {
+      "epoch": 0.04534792806880375,
+      "grad_norm": 0.17009177803993225,
+      "learning_rate": 2.2656250000000002e-05,
+      "loss": 0.4581,
+      "step": 58
+    },
+    {
+      "epoch": 0.046129788897576234,
+      "grad_norm": 0.1321457028388977,
+      "learning_rate": 2.3046875e-05,
+      "loss": 0.3902,
+      "step": 59
+    },
+    {
+      "epoch": 0.04691164972634871,
+      "grad_norm": 0.15425974130630493,
+      "learning_rate": 2.34375e-05,
+      "loss": 0.3024,
+      "step": 60
+    },
+    {
+      "epoch": 0.04769351055512119,
+      "grad_norm": 0.19503839313983917,
+      "learning_rate": 2.3828125e-05,
+      "loss": 0.4603,
+      "step": 61
+    },
+    {
+      "epoch": 0.04847537138389367,
+      "grad_norm": 0.1452198475599289,
+      "learning_rate": 2.4218750000000003e-05,
+      "loss": 0.4471,
+      "step": 62
+    },
+    {
+      "epoch": 0.04925723221266615,
+      "grad_norm": 0.18100936710834503,
+      "learning_rate": 2.4609375e-05,
+      "loss": 0.451,
+      "step": 63
+    },
+    {
+      "epoch": 0.05003909304143862,
+      "grad_norm": 0.15564265847206116,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3201,
+      "step": 64
+    },
+    {
+      "epoch": 0.0508209538702111,
+      "grad_norm": 0.1226576417684555,
+      "learning_rate": 2.5390625000000002e-05,
+      "loss": 0.3894,
+      "step": 65
+    },
+    {
+      "epoch": 0.05160281469898358,
+      "grad_norm": 0.14897918701171875,
+      "learning_rate": 2.578125e-05,
+      "loss": 0.3756,
+      "step": 66
+    },
+    {
+      "epoch": 0.05238467552775606,
+      "grad_norm": 0.13855944573879242,
+      "learning_rate": 2.6171875e-05,
+      "loss": 0.3293,
+      "step": 67
+    },
+    {
+      "epoch": 0.053166536356528536,
+      "grad_norm": 0.21206879615783691,
+      "learning_rate": 2.6562500000000002e-05,
+      "loss": 0.4521,
+      "step": 68
+    },
+    {
+      "epoch": 0.05394839718530102,
+      "grad_norm": 0.15101028978824615,
+      "learning_rate": 2.6953125000000003e-05,
+      "loss": 0.3386,
+      "step": 69
+    },
+    {
+      "epoch": 0.054730258014073496,
+      "grad_norm": 0.22186748683452606,
+      "learning_rate": 2.734375e-05,
+      "loss": 0.5313,
+      "step": 70
+    },
+    {
+      "epoch": 0.05551211884284597,
+      "grad_norm": 0.22481444478034973,
+      "learning_rate": 2.7734375e-05,
+      "loss": 0.4278,
+      "step": 71
+    },
+    {
+      "epoch": 0.05629397967161845,
+      "grad_norm": 0.2308468520641327,
+      "learning_rate": 2.8125000000000003e-05,
+      "loss": 0.3959,
+      "step": 72
+    },
+    {
+      "epoch": 0.05707584050039093,
+      "grad_norm": 0.2079354226589203,
+      "learning_rate": 2.8515625e-05,
+      "loss": 0.3525,
+      "step": 73
+    },
+    {
+      "epoch": 0.05785770132916341,
+      "grad_norm": 0.18281999230384827,
+      "learning_rate": 2.890625e-05,
+      "loss": 0.3882,
+      "step": 74
+    },
+    {
+      "epoch": 0.058639562157935886,
+      "grad_norm": 0.17227689921855927,
+      "learning_rate": 2.9296875000000002e-05,
+      "loss": 0.286,
+      "step": 75
+    },
+    {
+      "epoch": 0.05942142298670837,
+      "grad_norm": 0.20047691464424133,
+      "learning_rate": 2.96875e-05,
+      "loss": 0.3934,
+      "step": 76
+    },
+    {
+      "epoch": 0.060203283815480846,
+      "grad_norm": 0.1591797173023224,
+      "learning_rate": 3.0078125e-05,
+      "loss": 0.3894,
+      "step": 77
+    },
+    {
+      "epoch": 0.06098514464425332,
+      "grad_norm": 0.16098585724830627,
+      "learning_rate": 3.0468750000000002e-05,
+      "loss": 0.3237,
+      "step": 78
+    },
+    {
+      "epoch": 0.0617670054730258,
+      "grad_norm": 0.2547154426574707,
+      "learning_rate": 3.0859375e-05,
+      "loss": 0.4338,
+      "step": 79
+    },
+    {
+      "epoch": 0.06254886630179828,
+      "grad_norm": 0.23184354603290558,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3741,
+      "step": 80
+    },
+    {
+      "epoch": 0.06333072713057075,
+      "grad_norm": 0.17047226428985596,
+      "learning_rate": 3.1640625e-05,
+      "loss": 0.3753,
+      "step": 81
+    },
+    {
+      "epoch": 0.06411258795934324,
+      "grad_norm": 0.18740379810333252,
+      "learning_rate": 3.203125e-05,
+      "loss": 0.2879,
+      "step": 82
+    },
+    {
+      "epoch": 0.06489444878811572,
+      "grad_norm": 0.17314107716083527,
+      "learning_rate": 3.2421875e-05,
+      "loss": 0.3541,
+      "step": 83
+    },
+    {
+      "epoch": 0.0656763096168882,
+      "grad_norm": 0.1975407898426056,
+      "learning_rate": 3.2812500000000005e-05,
+      "loss": 0.3109,
+      "step": 84
+    },
+    {
+      "epoch": 0.06645817044566067,
+      "grad_norm": 0.16543574631214142,
+      "learning_rate": 3.3203125e-05,
+      "loss": 0.3602,
+      "step": 85
+    },
+    {
+      "epoch": 0.06724003127443315,
+      "grad_norm": 0.18290646374225616,
+      "learning_rate": 3.359375e-05,
+      "loss": 0.2981,
+      "step": 86
+    },
+    {
+      "epoch": 0.06802189210320563,
+      "grad_norm": 0.24966593086719513,
+      "learning_rate": 3.3984375000000004e-05,
+      "loss": 0.3733,
+      "step": 87
+    },
+    {
+      "epoch": 0.0688037529319781,
+      "grad_norm": 0.20844031870365143,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.3166,
+      "step": 88
+    },
+    {
+      "epoch": 0.06958561376075059,
+      "grad_norm": 0.19846050441265106,
+      "learning_rate": 3.4765625e-05,
+      "loss": 0.3384,
+      "step": 89
+    },
+    {
+      "epoch": 0.07036747458952307,
+      "grad_norm": 0.1682157814502716,
+      "learning_rate": 3.5156250000000004e-05,
+      "loss": 0.3681,
+      "step": 90
+    },
+    {
+      "epoch": 0.07114933541829555,
+      "grad_norm": 0.14380459487438202,
+      "learning_rate": 3.5546875e-05,
+      "loss": 0.269,
+      "step": 91
+    },
+    {
+      "epoch": 0.07193119624706802,
+      "grad_norm": 0.1845722645521164,
+      "learning_rate": 3.59375e-05,
+      "loss": 0.3338,
+      "step": 92
+    },
+    {
+      "epoch": 0.0727130570758405,
+      "grad_norm": 0.20639048516750336,
+      "learning_rate": 3.6328125000000004e-05,
+      "loss": 0.3314,
+      "step": 93
+    },
+    {
+      "epoch": 0.07349491790461297,
+      "grad_norm": 0.17344796657562256,
+      "learning_rate": 3.671875e-05,
+      "loss": 0.2556,
+      "step": 94
+    },
+    {
+      "epoch": 0.07427677873338545,
+      "grad_norm": 0.2165909856557846,
+      "learning_rate": 3.7109375e-05,
+      "loss": 0.2964,
+      "step": 95
+    },
+    {
+      "epoch": 0.07505863956215794,
+      "grad_norm": 0.3408878445625305,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.3842,
+      "step": 96
+    },
+    {
+      "epoch": 0.07584050039093042,
+      "grad_norm": 0.17121568322181702,
+      "learning_rate": 3.7890625e-05,
+      "loss": 0.3365,
+      "step": 97
+    },
+    {
+      "epoch": 0.0766223612197029,
+      "grad_norm": 0.21942703425884247,
+      "learning_rate": 3.828125e-05,
+      "loss": 0.3703,
+      "step": 98
+    },
+    {
+      "epoch": 0.07740422204847537,
+      "grad_norm": 0.22385479509830475,
+      "learning_rate": 3.8671875e-05,
+      "loss": 0.2629,
+      "step": 99
+    },
+    {
+      "epoch": 0.07818608287724785,
+      "grad_norm": 0.2156868726015091,
+      "learning_rate": 3.90625e-05,
+      "loss": 0.2524,
+      "step": 100
+    },
+    {
+      "epoch": 0.07896794370602032,
+      "grad_norm": 0.21917618811130524,
+      "learning_rate": 3.9453125000000005e-05,
+      "loss": 0.3201,
+      "step": 101
+    },
+    {
+      "epoch": 0.0797498045347928,
+      "grad_norm": 0.26034292578697205,
+      "learning_rate": 3.984375e-05,
+      "loss": 0.3169,
+      "step": 102
+    },
+    {
+      "epoch": 0.08053166536356529,
+      "grad_norm": 0.19670964777469635,
+      "learning_rate": 4.0234375e-05,
+      "loss": 0.2685,
+      "step": 103
+    },
+    {
+      "epoch": 0.08131352619233777,
+      "grad_norm": 0.18302002549171448,
+      "learning_rate": 4.0625000000000005e-05,
+      "loss": 0.2827,
+      "step": 104
+    },
+    {
+      "epoch": 0.08209538702111024,
+      "grad_norm": 0.2715568542480469,
+      "learning_rate": 4.1015625e-05,
+      "loss": 0.3265,
+      "step": 105
+    },
+    {
+      "epoch": 0.08287724784988272,
+      "grad_norm": 0.35596245527267456,
+      "learning_rate": 4.140625e-05,
+      "loss": 0.25,
+      "step": 106
+    },
+    {
+      "epoch": 0.0836591086786552,
+      "grad_norm": 0.24965617060661316,
+      "learning_rate": 4.1796875000000005e-05,
+      "loss": 0.2673,
+      "step": 107
+    },
+    {
+      "epoch": 0.08444096950742767,
+      "grad_norm": 0.2757260501384735,
+      "learning_rate": 4.21875e-05,
+      "loss": 0.279,
+      "step": 108
+    },
+    {
+      "epoch": 0.08522283033620015,
+      "grad_norm": 0.19423194229602814,
+      "learning_rate": 4.2578125e-05,
+      "loss": 0.2734,
+      "step": 109
+    },
+    {
+      "epoch": 0.08600469116497264,
+      "grad_norm": 0.25273609161376953,
+      "learning_rate": 4.2968750000000004e-05,
+      "loss": 0.2508,
+      "step": 110
+    },
+    {
+      "epoch": 0.08678655199374512,
+      "grad_norm": 0.26874634623527527,
+      "learning_rate": 4.3359375e-05,
+      "loss": 0.2911,
+      "step": 111
+    },
+    {
+      "epoch": 0.0875684128225176,
+      "grad_norm": 0.2321517914533615,
+      "learning_rate": 4.375e-05,
+      "loss": 0.2583,
+      "step": 112
+    },
+    {
+      "epoch": 0.08835027365129007,
+      "grad_norm": 0.23453573882579803,
+      "learning_rate": 4.4140625000000004e-05,
+      "loss": 0.2177,
+      "step": 113
+    },
+    {
+      "epoch": 0.08913213448006255,
+      "grad_norm": 0.23830734193325043,
+      "learning_rate": 4.453125e-05,
+      "loss": 0.3184,
+      "step": 114
+    },
+    {
+      "epoch": 0.08991399530883502,
+      "grad_norm": 0.28641262650489807,
+      "learning_rate": 4.4921875e-05,
+      "loss": 0.226,
+      "step": 115
+    },
+    {
+      "epoch": 0.0906958561376075,
+      "grad_norm": 0.30627748370170593,
+      "learning_rate": 4.5312500000000004e-05,
+      "loss": 0.2691,
+      "step": 116
+    },
+    {
+      "epoch": 0.09147771696637999,
+      "grad_norm": 0.25366127490997314,
+      "learning_rate": 4.5703125e-05,
+      "loss": 0.2986,
+      "step": 117
+    },
+    {
+      "epoch": 0.09225957779515247,
+      "grad_norm": 0.32924023270606995,
+      "learning_rate": 4.609375e-05,
+      "loss": 0.3162,
+      "step": 118
+    },
+    {
+      "epoch": 0.09304143862392494,
+      "grad_norm": 0.19363336265087128,
+      "learning_rate": 4.6484375e-05,
+      "loss": 0.2461,
+      "step": 119
+    },
+    {
+      "epoch": 0.09382329945269742,
+      "grad_norm": 0.23301272094249725,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.2032,
+      "step": 120
+    },
+    {
+      "epoch": 0.0946051602814699,
+      "grad_norm": 0.2663356363773346,
+      "learning_rate": 4.7265625000000005e-05,
+      "loss": 0.2863,
+      "step": 121
+    },
+    {
+      "epoch": 0.09538702111024237,
+      "grad_norm": 0.28720682859420776,
+      "learning_rate": 4.765625e-05,
+      "loss": 0.2427,
+      "step": 122
+    },
+    {
+      "epoch": 0.09616888193901485,
+      "grad_norm": 0.2564330995082855,
+      "learning_rate": 4.8046875e-05,
+      "loss": 0.1983,
+      "step": 123
+    },
+    {
+      "epoch": 0.09695074276778734,
+      "grad_norm": 0.3134055733680725,
+      "learning_rate": 4.8437500000000005e-05,
+      "loss": 0.3391,
+      "step": 124
+    },
+    {
+      "epoch": 0.09773260359655982,
+      "grad_norm": 0.25326237082481384,
+      "learning_rate": 4.8828125e-05,
+      "loss": 0.2411,
+      "step": 125
+    },
+    {
+      "epoch": 0.0985144644253323,
+      "grad_norm": 0.2458665817975998,
+      "learning_rate": 4.921875e-05,
+      "loss": 0.2508,
+      "step": 126
+    },
+    {
+      "epoch": 0.09929632525410477,
+      "grad_norm": 0.27264484763145447,
+      "learning_rate": 4.9609375000000005e-05,
+      "loss": 0.2081,
+      "step": 127
+    },
+    {
+      "epoch": 0.10007818608287725,
+      "grad_norm": 0.27833232283592224,
+      "learning_rate": 5e-05,
+      "loss": 0.2626,
+      "step": 128
+    },
+    {
+      "epoch": 0.10086004691164972,
+      "grad_norm": 0.26720625162124634,
+      "learning_rate": 5.0390625e-05,
+      "loss": 0.2484,
+      "step": 129
+    },
+    {
+      "epoch": 0.1016419077404222,
+      "grad_norm": 0.21806442737579346,
+      "learning_rate": 5.0781250000000004e-05,
+      "loss": 0.2157,
+      "step": 130
+    },
+    {
+      "epoch": 0.10242376856919469,
+      "grad_norm": 0.3490794599056244,
+      "learning_rate": 5.1171875e-05,
+      "loss": 0.2575,
+      "step": 131
+    },
+    {
+      "epoch": 0.10320562939796717,
+      "grad_norm": 0.2614622712135315,
+      "learning_rate": 5.15625e-05,
+      "loss": 0.2364,
+      "step": 132
+    },
+    {
+      "epoch": 0.10398749022673964,
+      "grad_norm": 0.3352520167827606,
+      "learning_rate": 5.1953125000000004e-05,
+      "loss": 0.2602,
+      "step": 133
+    },
+    {
+      "epoch": 0.10476935105551212,
+      "grad_norm": 0.28158560395240784,
+      "learning_rate": 5.234375e-05,
+      "loss": 0.2229,
+      "step": 134
+    },
+    {
+      "epoch": 0.1055512118842846,
+      "grad_norm": 0.19348321855068207,
+      "learning_rate": 5.2734375e-05,
+      "loss": 0.2265,
+      "step": 135
+    },
+    {
+      "epoch": 0.10633307271305707,
+      "grad_norm": 0.32511651515960693,
+      "learning_rate": 5.3125000000000004e-05,
+      "loss": 0.2831,
+      "step": 136
+    },
+    {
+      "epoch": 0.10711493354182955,
+      "grad_norm": 0.2327871322631836,
+      "learning_rate": 5.3515625e-05,
+      "loss": 0.2626,
+      "step": 137
+    },
+    {
+      "epoch": 0.10789679437060204,
+      "grad_norm": 0.3208716809749603,
+      "learning_rate": 5.3906250000000006e-05,
+      "loss": 0.246,
+      "step": 138
+    },
+    {
+      "epoch": 0.10867865519937452,
+      "grad_norm": 0.18219627439975739,
+      "learning_rate": 5.4296875000000004e-05,
+      "loss": 0.1884,
+      "step": 139
+    },
+    {
+      "epoch": 0.10946051602814699,
+      "grad_norm": 0.3777305483818054,
+      "learning_rate": 5.46875e-05,
+      "loss": 0.254,
+      "step": 140
+    },
+    {
+      "epoch": 0.11024237685691947,
+      "grad_norm": 0.23451992869377136,
+      "learning_rate": 5.5078125000000006e-05,
+      "loss": 0.1793,
+      "step": 141
+    },
+    {
+      "epoch": 0.11102423768569195,
+      "grad_norm": 0.3080081343650818,
+      "learning_rate": 5.546875e-05,
+      "loss": 0.2904,
+      "step": 142
+    },
+    {
+      "epoch": 0.11180609851446442,
+      "grad_norm": 0.20907637476921082,
+      "learning_rate": 5.5859375e-05,
+      "loss": 0.2452,
+      "step": 143
+    },
+    {
+      "epoch": 0.1125879593432369,
+      "grad_norm": 0.2892386317253113,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.2651,
+      "step": 144
+    },
+    {
+      "epoch": 0.11336982017200939,
+      "grad_norm": 0.25630876421928406,
+      "learning_rate": 5.6640625e-05,
+      "loss": 0.264,
+      "step": 145
+    },
+    {
+      "epoch": 0.11415168100078187,
+      "grad_norm": 0.2772256135940552,
+      "learning_rate": 5.703125e-05,
+      "loss": 0.2461,
+      "step": 146
+    },
+    {
+      "epoch": 0.11493354182955434,
+      "grad_norm": 0.28796228766441345,
+      "learning_rate": 5.7421875000000005e-05,
+      "loss": 0.2164,
+      "step": 147
+    },
+    {
+      "epoch": 0.11571540265832682,
+      "grad_norm": 0.272840678691864,
+      "learning_rate": 5.78125e-05,
+      "loss": 0.2057,
+      "step": 148
+    },
+    {
+      "epoch": 0.1164972634870993,
+      "grad_norm": 0.27350762486457825,
+      "learning_rate": 5.8203125e-05,
+      "loss": 0.2567,
+      "step": 149
+    },
+    {
+      "epoch": 0.11727912431587177,
+      "grad_norm": 0.2435077726840973,
+      "learning_rate": 5.8593750000000005e-05,
+      "loss": 0.181,
+      "step": 150
+    },
+    {
+      "epoch": 0.11806098514464425,
+      "grad_norm": 0.3440535068511963,
+      "learning_rate": 5.8984375e-05,
+      "loss": 0.2955,
+      "step": 151
+    },
+    {
+      "epoch": 0.11884284597341674,
+      "grad_norm": 0.30235132575035095,
+      "learning_rate": 5.9375e-05,
+      "loss": 0.249,
+      "step": 152
+    },
+    {
+      "epoch": 0.11962470680218922,
+      "grad_norm": 0.1905151903629303,
+      "learning_rate": 5.9765625000000004e-05,
+      "loss": 0.2057,
+      "step": 153
+    },
+    {
+      "epoch": 0.12040656763096169,
+      "grad_norm": 0.2980440855026245,
+      "learning_rate": 6.015625e-05,
+      "loss": 0.2824,
+      "step": 154
+    },
+    {
+      "epoch": 0.12118842845973417,
+      "grad_norm": 0.21735329926013947,
+      "learning_rate": 6.0546875e-05,
+      "loss": 0.187,
+      "step": 155
+    },
+    {
+      "epoch": 0.12197028928850664,
+      "grad_norm": 0.2793501317501068,
+      "learning_rate": 6.0937500000000004e-05,
+      "loss": 0.2543,
+      "step": 156
+    },
+    {
+      "epoch": 0.12275215011727912,
+      "grad_norm": 0.26220038533210754,
+      "learning_rate": 6.132812500000001e-05,
+      "loss": 0.2218,
+      "step": 157
+    },
+    {
+      "epoch": 0.1235340109460516,
+      "grad_norm": 0.31638094782829285,
+      "learning_rate": 6.171875e-05,
+      "loss": 0.2593,
+      "step": 158
+    },
+    {
+      "epoch": 0.12431587177482409,
+      "grad_norm": 0.2769337296485901,
+      "learning_rate": 6.2109375e-05,
+      "loss": 0.2016,
+      "step": 159
+    },
+    {
+      "epoch": 0.12509773260359655,
+      "grad_norm": 0.42148536443710327,
+      "learning_rate": 6.25e-05,
+      "loss": 0.2441,
+      "step": 160
+    },
+    {
+      "epoch": 0.12587959343236904,
+      "grad_norm": 0.30849018692970276,
+      "learning_rate": 6.2890625e-05,
+      "loss": 0.2421,
+      "step": 161
+    },
+    {
+      "epoch": 0.1266614542611415,
+      "grad_norm": 0.27945342659950256,
+      "learning_rate": 6.328125e-05,
+      "loss": 0.1959,
+      "step": 162
+    },
+    {
+      "epoch": 0.127443315089914,
+      "grad_norm": 0.23380780220031738,
+      "learning_rate": 6.367187500000001e-05,
+      "loss": 0.1951,
+      "step": 163
+    },
+    {
+      "epoch": 0.12822517591868648,
+      "grad_norm": 0.3297896087169647,
+      "learning_rate": 6.40625e-05,
+      "loss": 0.2182,
+      "step": 164
+    },
+    {
+      "epoch": 0.12900703674745895,
+      "grad_norm": 0.4288419783115387,
+      "learning_rate": 6.4453125e-05,
+      "loss": 0.2286,
+      "step": 165
+    },
+    {
+      "epoch": 0.12978889757623144,
+      "grad_norm": 0.4745989143848419,
+      "learning_rate": 6.484375e-05,
+      "loss": 0.1925,
+      "step": 166
+    },
+    {
+      "epoch": 0.1305707584050039,
+      "grad_norm": 0.4383123815059662,
+      "learning_rate": 6.5234375e-05,
+      "loss": 0.1877,
+      "step": 167
+    },
+    {
+      "epoch": 0.1313526192337764,
+      "grad_norm": 0.33036768436431885,
+      "learning_rate": 6.562500000000001e-05,
+      "loss": 0.2172,
+      "step": 168
+    },
+    {
+      "epoch": 0.13213448006254885,
+      "grad_norm": 0.2799517810344696,
+      "learning_rate": 6.601562500000001e-05,
+      "loss": 0.1766,
+      "step": 169
+    },
+    {
+      "epoch": 0.13291634089132134,
+      "grad_norm": 0.3703453242778778,
+      "learning_rate": 6.640625e-05,
+      "loss": 0.207,
+      "step": 170
+    },
+    {
+      "epoch": 0.13369820172009383,
+      "grad_norm": 0.29083386063575745,
+      "learning_rate": 6.6796875e-05,
+      "loss": 0.2675,
+      "step": 171
+    },
+    {
+      "epoch": 0.1344800625488663,
+      "grad_norm": 0.3545566499233246,
+      "learning_rate": 6.71875e-05,
+      "loss": 0.2692,
+      "step": 172
+    },
+    {
+      "epoch": 0.1352619233776388,
+      "grad_norm": 0.2916538119316101,
+      "learning_rate": 6.7578125e-05,
+      "loss": 0.1521,
+      "step": 173
+    },
+    {
+      "epoch": 0.13604378420641125,
+      "grad_norm": 0.3869919776916504,
+      "learning_rate": 6.796875000000001e-05,
+      "loss": 0.2785,
+      "step": 174
+    },
+    {
+      "epoch": 0.13682564503518374,
+      "grad_norm": 0.32206371426582336,
+      "learning_rate": 6.8359375e-05,
+      "loss": 0.2055,
+      "step": 175
+    },
+    {
+      "epoch": 0.1376075058639562,
+      "grad_norm": 0.47468236088752747,
+      "learning_rate": 6.875e-05,
+      "loss": 0.2493,
+      "step": 176
+    },
+    {
+      "epoch": 0.1383893666927287,
+      "grad_norm": 0.23824159801006317,
+      "learning_rate": 6.9140625e-05,
+      "loss": 0.1699,
+      "step": 177
+    },
+    {
+      "epoch": 0.13917122752150118,
+      "grad_norm": 0.5079713463783264,
+      "learning_rate": 6.953125e-05,
+      "loss": 0.2437,
+      "step": 178
+    },
+    {
+      "epoch": 0.13995308835027365,
+      "grad_norm": 0.3343154489994049,
+      "learning_rate": 6.9921875e-05,
+      "loss": 0.1674,
+      "step": 179
+    },
+    {
+      "epoch": 0.14073494917904614,
+      "grad_norm": 0.3303624093532562,
+      "learning_rate": 7.031250000000001e-05,
+      "loss": 0.1942,
+      "step": 180
+    },
+    {
+      "epoch": 0.1415168100078186,
+      "grad_norm": 0.48036229610443115,
+      "learning_rate": 7.0703125e-05,
+      "loss": 0.2186,
+      "step": 181
+    },
+    {
+      "epoch": 0.1422986708365911,
+      "grad_norm": 0.40403860807418823,
+      "learning_rate": 7.109375e-05,
+      "loss": 0.2192,
+      "step": 182
+    },
+    {
+      "epoch": 0.14308053166536355,
+      "grad_norm": 0.4421844184398651,
+      "learning_rate": 7.1484375e-05,
+      "loss": 0.1923,
+      "step": 183
+    },
+    {
+      "epoch": 0.14386239249413604,
+      "grad_norm": 0.39282283186912537,
+      "learning_rate": 7.1875e-05,
+      "loss": 0.2658,
+      "step": 184
+    },
+    {
+      "epoch": 0.14464425332290853,
+      "grad_norm": 0.29992353916168213,
+      "learning_rate": 7.226562500000001e-05,
+      "loss": 0.232,
+      "step": 185
+    },
+    {
+      "epoch": 0.145426114151681,
+      "grad_norm": 0.4302484095096588,
+      "learning_rate": 7.265625000000001e-05,
+      "loss": 0.2118,
+      "step": 186
+    },
+    {
+      "epoch": 0.1462079749804535,
+      "grad_norm": 0.3942473232746124,
+      "learning_rate": 7.3046875e-05,
+      "loss": 0.2664,
+      "step": 187
+    },
+    {
+      "epoch": 0.14698983580922595,
+      "grad_norm": 0.38469308614730835,
+      "learning_rate": 7.34375e-05,
+      "loss": 0.1307,
+      "step": 188
+    },
+    {
+      "epoch": 0.14777169663799844,
+      "grad_norm": 0.2901390790939331,
+      "learning_rate": 7.3828125e-05,
+      "loss": 0.2786,
+      "step": 189
+    },
+    {
+      "epoch": 0.1485535574667709,
+      "grad_norm": 0.28935694694519043,
+      "learning_rate": 7.421875e-05,
+      "loss": 0.1513,
+      "step": 190
+    },
+    {
+      "epoch": 0.1493354182955434,
+      "grad_norm": 0.28370222449302673,
+      "learning_rate": 7.460937500000001e-05,
+      "loss": 0.2446,
+      "step": 191
+    },
+    {
+      "epoch": 0.15011727912431588,
+      "grad_norm": 0.36691758036613464,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.2114,
+      "step": 192
+    },
+    {
+      "epoch": 0.15089913995308835,
+      "grad_norm": 0.41656333208084106,
+      "learning_rate": 7.5390625e-05,
+      "loss": 0.2144,
+      "step": 193
+    },
+    {
+      "epoch": 0.15168100078186084,
+      "grad_norm": 0.34387269616127014,
+      "learning_rate": 7.578125e-05,
+      "loss": 0.2113,
+      "step": 194
+    },
+    {
+      "epoch": 0.1524628616106333,
+      "grad_norm": 0.33604809641838074,
+      "learning_rate": 7.6171875e-05,
+      "loss": 0.1784,
+      "step": 195
+    },
+    {
+      "epoch": 0.1532447224394058,
+      "grad_norm": 0.26703089475631714,
+      "learning_rate": 7.65625e-05,
+      "loss": 0.1617,
+      "step": 196
+    },
+    {
+      "epoch": 0.15402658326817825,
+      "grad_norm": 0.35461607575416565,
+      "learning_rate": 7.695312500000001e-05,
+      "loss": 0.2431,
+      "step": 197
+    },
+    {
+      "epoch": 0.15480844409695074,
+      "grad_norm": 0.35807040333747864,
+      "learning_rate": 7.734375e-05,
+      "loss": 0.1861,
+      "step": 198
+    },
+    {
+      "epoch": 0.15559030492572323,
+      "grad_norm": 0.3112034499645233,
+      "learning_rate": 7.7734375e-05,
+      "loss": 0.2142,
+      "step": 199
+    },
+    {
+      "epoch": 0.1563721657544957,
+      "grad_norm": 0.4752049446105957,
+      "learning_rate": 7.8125e-05,
+      "loss": 0.2322,
+      "step": 200
+    },
+    {
+      "epoch": 0.1563721657544957,
+      "eval_loss": 0.21434074640274048,
+      "eval_runtime": 13.3101,
+      "eval_samples_per_second": 3.907,
+      "eval_steps_per_second": 0.977,
+      "step": 200
+    },
+    {
+      "epoch": 0.15715402658326819,
+      "grad_norm": 0.35235124826431274,
+      "learning_rate": 7.8515625e-05,
+      "loss": 0.1912,
+      "step": 201
+    },
+    {
+      "epoch": 0.15793588741204065,
+      "grad_norm": 0.26190364360809326,
+      "learning_rate": 7.890625000000001e-05,
+      "loss": 0.1864,
+      "step": 202
+    },
+    {
+      "epoch": 0.15871774824081314,
+      "grad_norm": 0.4348473846912384,
+      "learning_rate": 7.929687500000001e-05,
+      "loss": 0.2316,
+      "step": 203
+    },
+    {
+      "epoch": 0.1594996090695856,
+      "grad_norm": 0.3678869605064392,
+      "learning_rate": 7.96875e-05,
+      "loss": 0.2044,
+      "step": 204
+    },
+    {
+      "epoch": 0.1602814698983581,
+      "grad_norm": 0.3102744221687317,
+      "learning_rate": 8.0078125e-05,
+      "loss": 0.2061,
+      "step": 205
+    },
+    {
+      "epoch": 0.16106333072713058,
+      "grad_norm": 0.370443731546402,
+      "learning_rate": 8.046875e-05,
+      "loss": 0.2046,
+      "step": 206
+    },
+    {
+      "epoch": 0.16184519155590305,
+      "grad_norm": 0.4026085138320923,
+      "learning_rate": 8.0859375e-05,
+      "loss": 0.2202,
+      "step": 207
+    },
+    {
+      "epoch": 0.16262705238467554,
+      "grad_norm": 0.30296698212623596,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.1761,
+      "step": 208
+    },
+    {
+      "epoch": 0.163408913213448,
+      "grad_norm": 0.3474259674549103,
+      "learning_rate": 8.164062500000001e-05,
+      "loss": 0.1942,
+      "step": 209
+    },
+    {
+      "epoch": 0.1641907740422205,
+      "grad_norm": 0.32637667655944824,
+      "learning_rate": 8.203125e-05,
+      "loss": 0.1872,
+      "step": 210
+    },
+    {
+      "epoch": 0.16497263487099295,
+      "grad_norm": 0.3576767146587372,
+      "learning_rate": 8.2421875e-05,
+      "loss": 0.1869,
+      "step": 211
+    },
+    {
+      "epoch": 0.16575449569976544,
+      "grad_norm": 0.3988786041736603,
+      "learning_rate": 8.28125e-05,
+      "loss": 0.2351,
+      "step": 212
+    },
+    {
+      "epoch": 0.16653635652853793,
+      "grad_norm": 0.28785213828086853,
+      "learning_rate": 8.3203125e-05,
+      "loss": 0.1558,
+      "step": 213
+    },
+    {
+      "epoch": 0.1673182173573104,
+      "grad_norm": 0.527788519859314,
+      "learning_rate": 8.359375000000001e-05,
+      "loss": 0.2125,
+      "step": 214
+    },
+    {
+      "epoch": 0.16810007818608289,
+      "grad_norm": 0.3742121458053589,
+      "learning_rate": 8.398437500000001e-05,
+      "loss": 0.1884,
+      "step": 215
+    },
+    {
+      "epoch": 0.16888193901485535,
+      "grad_norm": 0.4711468517780304,
+      "learning_rate": 8.4375e-05,
+      "loss": 0.1743,
+      "step": 216
+    },
+    {
+      "epoch": 0.16966379984362784,
+      "grad_norm": 0.6251031756401062,
+      "learning_rate": 8.4765625e-05,
+      "loss": 0.2971,
+      "step": 217
+    },
+    {
+      "epoch": 0.1704456606724003,
+      "grad_norm": 0.4275197684764862,
+      "learning_rate": 8.515625e-05,
+      "loss": 0.1872,
+      "step": 218
+    },
+    {
+      "epoch": 0.1712275215011728,
+      "grad_norm": 0.313466340303421,
+      "learning_rate": 8.5546875e-05,
+      "loss": 0.1916,
+      "step": 219
+    },
+    {
+      "epoch": 0.17200938232994528,
+      "grad_norm": 0.3586588501930237,
+      "learning_rate": 8.593750000000001e-05,
+      "loss": 0.1848,
+      "step": 220
+    },
+    {
+      "epoch": 0.17279124315871774,
+      "grad_norm": 0.4977078139781952,
+      "learning_rate": 8.6328125e-05,
+      "loss": 0.2476,
+      "step": 221
+    },
+    {
+      "epoch": 0.17357310398749023,
+      "grad_norm": 0.35769525170326233,
+      "learning_rate": 8.671875e-05,
+      "loss": 0.199,
+      "step": 222
+    },
+    {
+      "epoch": 0.1743549648162627,
+      "grad_norm": 0.4167306423187256,
+      "learning_rate": 8.7109375e-05,
+      "loss": 0.1861,
+      "step": 223
+    },
+    {
+      "epoch": 0.1751368256450352,
+      "grad_norm": 0.3895793557167053,
+      "learning_rate": 8.75e-05,
+      "loss": 0.1772,
+      "step": 224
+    },
+    {
+      "epoch": 0.17591868647380765,
+      "grad_norm": 0.3502638339996338,
+      "learning_rate": 8.789062500000001e-05,
+      "loss": 0.2047,
+      "step": 225
+    },
+    {
+      "epoch": 0.17670054730258014,
+      "grad_norm": 0.3090845048427582,
+      "learning_rate": 8.828125000000001e-05,
+      "loss": 0.1812,
+      "step": 226
+    },
+    {
+      "epoch": 0.17748240813135263,
+      "grad_norm": 0.40732601284980774,
+      "learning_rate": 8.8671875e-05,
+      "loss": 0.2121,
+      "step": 227
+    },
+    {
+      "epoch": 0.1782642689601251,
+      "grad_norm": 0.42500749230384827,
+      "learning_rate": 8.90625e-05,
+      "loss": 0.19,
+      "step": 228
+    },
+    {
+      "epoch": 0.17904612978889758,
+      "grad_norm": 0.2970224916934967,
+      "learning_rate": 8.9453125e-05,
+      "loss": 0.1517,
+      "step": 229
+    },
+    {
+      "epoch": 0.17982799061767005,
+      "grad_norm": 0.3494272530078888,
+      "learning_rate": 8.984375e-05,
+      "loss": 0.2368,
+      "step": 230
+    },
+    {
+      "epoch": 0.18060985144644254,
+      "grad_norm": 0.4563659727573395,
+      "learning_rate": 9.023437500000001e-05,
+      "loss": 0.1987,
+      "step": 231
+    },
+    {
+      "epoch": 0.181391712275215,
+      "grad_norm": 0.4436132311820984,
+      "learning_rate": 9.062500000000001e-05,
+      "loss": 0.2369,
+      "step": 232
+    },
+    {
+      "epoch": 0.1821735731039875,
+      "grad_norm": 0.4118313789367676,
+      "learning_rate": 9.1015625e-05,
+      "loss": 0.1623,
+      "step": 233
+    },
+    {
+      "epoch": 0.18295543393275998,
+      "grad_norm": 0.535983681678772,
+      "learning_rate": 9.140625e-05,
+      "loss": 0.1718,
+      "step": 234
+    },
+    {
+      "epoch": 0.18373729476153244,
+      "grad_norm": 0.4556559920310974,
+      "learning_rate": 9.1796875e-05,
+      "loss": 0.209,
+      "step": 235
+    },
+    {
+      "epoch": 0.18451915559030493,
+      "grad_norm": 0.43632426857948303,
+      "learning_rate": 9.21875e-05,
+      "loss": 0.1622,
+      "step": 236
+    },
+    {
+      "epoch": 0.1853010164190774,
+      "grad_norm": 0.34906676411628723,
+      "learning_rate": 9.257812500000001e-05,
+      "loss": 0.2026,
+      "step": 237
+    },
+    {
+      "epoch": 0.1860828772478499,
+      "grad_norm": 0.3979455828666687,
+      "learning_rate": 9.296875e-05,
+      "loss": 0.1893,
+      "step": 238
+    },
+    {
+      "epoch": 0.18686473807662235,
+      "grad_norm": 0.31991952657699585,
+      "learning_rate": 9.3359375e-05,
+      "loss": 0.1957,
+      "step": 239
+    },
+    {
+      "epoch": 0.18764659890539484,
+      "grad_norm": 0.3094921112060547,
+      "learning_rate": 9.375e-05,
+      "loss": 0.1545,
+      "step": 240
+    },
+    {
+      "epoch": 0.18842845973416733,
+      "grad_norm": 0.3545514941215515,
+      "learning_rate": 9.4140625e-05,
+      "loss": 0.2365,
+      "step": 241
+    },
+    {
+      "epoch": 0.1892103205629398,
+      "grad_norm": 0.3044324815273285,
+      "learning_rate": 9.453125000000001e-05,
+      "loss": 0.1804,
+      "step": 242
+    },
+    {
+      "epoch": 0.18999218139171228,
+      "grad_norm": 0.3272997736930847,
+      "learning_rate": 9.492187500000001e-05,
+      "loss": 0.1591,
+      "step": 243
+    },
+    {
+      "epoch": 0.19077404222048475,
+      "grad_norm": 0.4019821286201477,
+      "learning_rate": 9.53125e-05,
+      "loss": 0.1935,
+      "step": 244
+    },
+    {
+      "epoch": 0.19155590304925724,
+      "grad_norm": 0.3890402615070343,
+      "learning_rate": 9.5703125e-05,
+      "loss": 0.1816,
+      "step": 245
+    },
+    {
+      "epoch": 0.1923377638780297,
+      "grad_norm": 0.39056506752967834,
+      "learning_rate": 9.609375e-05,
+      "loss": 0.1788,
+      "step": 246
+    },
+    {
+      "epoch": 0.1931196247068022,
+      "grad_norm": 0.47024020552635193,
+      "learning_rate": 9.6484375e-05,
+      "loss": 0.1949,
+      "step": 247
+    },
+    {
+      "epoch": 0.19390148553557468,
+      "grad_norm": 0.31551235914230347,
+      "learning_rate": 9.687500000000001e-05,
+      "loss": 0.1469,
+      "step": 248
+    },
+    {
+      "epoch": 0.19468334636434714,
+      "grad_norm": 0.4270898103713989,
+      "learning_rate": 9.726562500000001e-05,
+      "loss": 0.1873,
+      "step": 249
+    },
+    {
+      "epoch": 0.19546520719311963,
+      "grad_norm": 0.34933722019195557,
+      "learning_rate": 9.765625e-05,
+      "loss": 0.2114,
+      "step": 250
+    },
+    {
+      "epoch": 0.1962470680218921,
+      "grad_norm": 0.32482168078422546,
+      "learning_rate": 9.8046875e-05,
+      "loss": 0.1765,
+      "step": 251
+    },
+    {
+      "epoch": 0.1970289288506646,
+      "grad_norm": 0.36024120450019836,
+      "learning_rate": 9.84375e-05,
+      "loss": 0.185,
+      "step": 252
+    },
+    {
+      "epoch": 0.19781078967943705,
+      "grad_norm": 0.36847880482673645,
+      "learning_rate": 9.8828125e-05,
+      "loss": 0.1791,
+      "step": 253
+    },
+    {
+      "epoch": 0.19859265050820954,
+      "grad_norm": 0.487062007188797,
+      "learning_rate": 9.921875000000001e-05,
+      "loss": 0.2098,
+      "step": 254
+    },
+    {
+      "epoch": 0.19937451133698203,
+      "grad_norm": 0.4240575134754181,
+      "learning_rate": 9.960937500000001e-05,
+      "loss": 0.1558,
+      "step": 255
+    },
+    {
+      "epoch": 0.2001563721657545,
+      "grad_norm": 0.4850272238254547,
+      "learning_rate": 0.0001,
+      "loss": 0.2066,
+      "step": 256
+    },
+    {
+      "epoch": 0.20093823299452698,
+      "grad_norm": 0.2975417375564575,
+      "learning_rate": 9.999995343827644e-05,
+      "loss": 0.1519,
+      "step": 257
+    },
+    {
+      "epoch": 0.20172009382329945,
+      "grad_norm": 0.4383419454097748,
+      "learning_rate": 9.999981375319249e-05,
+      "loss": 0.2041,
+      "step": 258
+    },
+    {
+      "epoch": 0.20250195465207194,
+      "grad_norm": 0.4741787314414978,
+      "learning_rate": 9.99995809450083e-05,
+      "loss": 0.186,
+      "step": 259
+    },
+    {
+      "epoch": 0.2032838154808444,
+      "grad_norm": 0.41490674018859863,
+      "learning_rate": 9.999925501415746e-05,
+      "loss": 0.1978,
+      "step": 260
+    },
+    {
+      "epoch": 0.2040656763096169,
+      "grad_norm": 0.5051758885383606,
+      "learning_rate": 9.9998835961247e-05,
+      "loss": 0.2547,
+      "step": 261
+    },
+    {
+      "epoch": 0.20484753713838938,
+      "grad_norm": 0.2550017535686493,
+      "learning_rate": 9.999832378705743e-05,
+      "loss": 0.1447,
+      "step": 262
+    },
+    {
+      "epoch": 0.20562939796716184,
+      "grad_norm": 0.4408182203769684,
+      "learning_rate": 9.999771849254263e-05,
+      "loss": 0.1724,
+      "step": 263
+    },
+    {
+      "epoch": 0.20641125879593433,
+      "grad_norm": 0.45843225717544556,
+      "learning_rate": 9.999702007882995e-05,
+      "loss": 0.198,
+      "step": 264
+    },
+    {
+      "epoch": 0.2071931196247068,
+      "grad_norm": 0.3172323703765869,
+      "learning_rate": 9.999622854722017e-05,
+      "loss": 0.2108,
+      "step": 265
+    },
+    {
+      "epoch": 0.20797498045347929,
+      "grad_norm": 0.403096467256546,
+      "learning_rate": 9.999534389918747e-05,
+      "loss": 0.1631,
+      "step": 266
+    },
+    {
+      "epoch": 0.20875684128225175,
+      "grad_norm": 0.28646552562713623,
+      "learning_rate": 9.99943661363795e-05,
+      "loss": 0.1384,
+      "step": 267
+    },
+    {
+      "epoch": 0.20953870211102424,
+      "grad_norm": 0.5032647252082825,
+      "learning_rate": 9.999329526061731e-05,
+      "loss": 0.2253,
+      "step": 268
+    },
+    {
+      "epoch": 0.21032056293979673,
+      "grad_norm": 0.4821566045284271,
+      "learning_rate": 9.999213127389536e-05,
+      "loss": 0.2096,
+      "step": 269
+    },
+    {
+      "epoch": 0.2111024237685692,
+      "grad_norm": 0.36564865708351135,
+      "learning_rate": 9.999087417838156e-05,
+      "loss": 0.1496,
+      "step": 270
+    },
+    {
+      "epoch": 0.21188428459734168,
+      "grad_norm": 0.3803997039794922,
+      "learning_rate": 9.99895239764172e-05,
+      "loss": 0.1445,
+      "step": 271
+    },
+    {
+      "epoch": 0.21266614542611414,
+      "grad_norm": 0.3326912522315979,
+      "learning_rate": 9.9988080670517e-05,
+      "loss": 0.1422,
+      "step": 272
+    },
+    {
+      "epoch": 0.21344800625488664,
+      "grad_norm": 0.44342145323753357,
+      "learning_rate": 9.998654426336905e-05,
+      "loss": 0.1902,
+      "step": 273
+    },
+    {
+      "epoch": 0.2142298670836591,
+      "grad_norm": 0.3668600618839264,
+      "learning_rate": 9.998491475783487e-05,
+      "loss": 0.2203,
+      "step": 274
+    },
+    {
+      "epoch": 0.2150117279124316,
+      "grad_norm": 0.49768349528312683,
+      "learning_rate": 9.998319215694936e-05,
+      "loss": 0.1857,
+      "step": 275
+    },
+    {
+      "epoch": 0.21579358874120408,
+      "grad_norm": 0.4042830765247345,
+      "learning_rate": 9.998137646392083e-05,
+      "loss": 0.1858,
+      "step": 276
+    },
+    {
+      "epoch": 0.21657544956997654,
+      "grad_norm": 0.36947476863861084,
+      "learning_rate": 9.997946768213095e-05,
+      "loss": 0.2113,
+      "step": 277
+    },
+    {
+      "epoch": 0.21735731039874903,
+      "grad_norm": 0.32007378339767456,
+      "learning_rate": 9.997746581513474e-05,
+      "loss": 0.2095,
+      "step": 278
+    },
+    {
+      "epoch": 0.2181391712275215,
+      "grad_norm": 0.3947068452835083,
+      "learning_rate": 9.997537086666063e-05,
+      "loss": 0.1986,
+      "step": 279
+    },
+    {
+      "epoch": 0.21892103205629398,
+      "grad_norm": 0.31826138496398926,
+      "learning_rate": 9.997318284061041e-05,
+      "loss": 0.1518,
+      "step": 280
+    },
+    {
+      "epoch": 0.21970289288506645,
+      "grad_norm": 0.5408663749694824,
+      "learning_rate": 9.997090174105919e-05,
+      "loss": 0.1945,
+      "step": 281
+    },
+    {
+      "epoch": 0.22048475371383894,
+      "grad_norm": 0.36804401874542236,
+      "learning_rate": 9.996852757225546e-05,
+      "loss": 0.16,
+      "step": 282
+    },
+    {
+      "epoch": 0.22126661454261143,
+      "grad_norm": 0.3286680579185486,
+      "learning_rate": 9.996606033862102e-05,
+      "loss": 0.1548,
+      "step": 283
+    },
+    {
+      "epoch": 0.2220484753713839,
+      "grad_norm": 0.36132723093032837,
+      "learning_rate": 9.996350004475104e-05,
+      "loss": 0.2245,
+      "step": 284
+    },
+    {
+      "epoch": 0.22283033620015638,
+      "grad_norm": 0.38861459493637085,
+      "learning_rate": 9.996084669541397e-05,
+      "loss": 0.2044,
+      "step": 285
+    },
+    {
+      "epoch": 0.22361219702892884,
+      "grad_norm": 0.5624184608459473,
+      "learning_rate": 9.995810029555159e-05,
+      "loss": 0.1935,
+      "step": 286
+    },
+    {
+      "epoch": 0.22439405785770133,
+      "grad_norm": 0.4298049807548523,
+      "learning_rate": 9.9955260850279e-05,
+      "loss": 0.1332,
+      "step": 287
+    },
+    {
+      "epoch": 0.2251759186864738,
+      "grad_norm": 0.36704859137535095,
+      "learning_rate": 9.995232836488455e-05,
+      "loss": 0.1673,
+      "step": 288
+    },
+    {
+      "epoch": 0.2259577795152463,
+      "grad_norm": 0.2979743182659149,
+      "learning_rate": 9.994930284482993e-05,
+      "loss": 0.1472,
+      "step": 289
+    },
+    {
+      "epoch": 0.22673964034401878,
+      "grad_norm": 0.36498042941093445,
+      "learning_rate": 9.994618429575008e-05,
+      "loss": 0.1409,
+      "step": 290
+    },
+    {
+      "epoch": 0.22752150117279124,
+      "grad_norm": 0.2840629816055298,
+      "learning_rate": 9.994297272345319e-05,
+      "loss": 0.126,
+      "step": 291
+    },
+    {
+      "epoch": 0.22830336200156373,
+      "grad_norm": 0.4682474732398987,
+      "learning_rate": 9.99396681339207e-05,
+      "loss": 0.2315,
+      "step": 292
+    },
+    {
+      "epoch": 0.2290852228303362,
+      "grad_norm": 0.4364871084690094,
+      "learning_rate": 9.993627053330732e-05,
+      "loss": 0.1534,
+      "step": 293
+    },
+    {
+      "epoch": 0.22986708365910868,
+      "grad_norm": 0.4619438350200653,
+      "learning_rate": 9.993277992794096e-05,
+      "loss": 0.186,
+      "step": 294
+    },
+    {
+      "epoch": 0.23064894448788115,
+      "grad_norm": 0.47760748863220215,
+      "learning_rate": 9.99291963243228e-05,
+      "loss": 0.182,
+      "step": 295
+    },
+    {
+      "epoch": 0.23143080531665364,
+      "grad_norm": 0.3896167576313019,
+      "learning_rate": 9.992551972912718e-05,
+      "loss": 0.1883,
+      "step": 296
+    },
+    {
+      "epoch": 0.23221266614542613,
+      "grad_norm": 0.4022518992424011,
+      "learning_rate": 9.992175014920161e-05,
+      "loss": 0.1948,
+      "step": 297
+    },
+    {
+      "epoch": 0.2329945269741986,
+      "grad_norm": 0.382075697183609,
+      "learning_rate": 9.991788759156684e-05,
+      "loss": 0.1836,
+      "step": 298
+    },
+    {
+      "epoch": 0.23377638780297108,
+      "grad_norm": 0.35182997584342957,
+      "learning_rate": 9.991393206341677e-05,
+      "loss": 0.1215,
+      "step": 299
+    },
+    {
+      "epoch": 0.23455824863174354,
+      "grad_norm": 0.2927306890487671,
+      "learning_rate": 9.990988357211843e-05,
+      "loss": 0.127,
+      "step": 300
+    },
+    {
+      "epoch": 0.23534010946051603,
+      "grad_norm": 0.31669214367866516,
+      "learning_rate": 9.990574212521205e-05,
+      "loss": 0.1553,
+      "step": 301
+    },
+    {
+      "epoch": 0.2361219702892885,
+      "grad_norm": 0.39152827858924866,
+      "learning_rate": 9.990150773041089e-05,
+      "loss": 0.1672,
+      "step": 302
+    },
+    {
+      "epoch": 0.236903831118061,
+      "grad_norm": 0.4005658030509949,
+      "learning_rate": 9.98971803956014e-05,
+      "loss": 0.2034,
+      "step": 303
+    },
+    {
+      "epoch": 0.23768569194683348,
+      "grad_norm": 0.3976832330226898,
+      "learning_rate": 9.989276012884313e-05,
+      "loss": 0.1498,
+      "step": 304
+    },
+    {
+      "epoch": 0.23846755277560594,
+      "grad_norm": 0.36635416746139526,
+      "learning_rate": 9.988824693836864e-05,
+      "loss": 0.1765,
+      "step": 305
+    },
+    {
+      "epoch": 0.23924941360437843,
+      "grad_norm": 0.44994375109672546,
+      "learning_rate": 9.988364083258367e-05,
+      "loss": 0.201,
+      "step": 306
+    },
+    {
+      "epoch": 0.2400312744331509,
+      "grad_norm": 0.35641130805015564,
+      "learning_rate": 9.98789418200669e-05,
+      "loss": 0.1594,
+      "step": 307
+    },
+    {
+      "epoch": 0.24081313526192338,
+      "grad_norm": 0.5840378999710083,
+      "learning_rate": 9.987414990957012e-05,
+      "loss": 0.2061,
+      "step": 308
+    },
+    {
+      "epoch": 0.24159499609069585,
+      "grad_norm": 0.3115840256214142,
+      "learning_rate": 9.98692651100181e-05,
+      "loss": 0.1272,
+      "step": 309
+    },
+    {
+      "epoch": 0.24237685691946834,
+      "grad_norm": 0.30972468852996826,
+      "learning_rate": 9.986428743050864e-05,
+      "loss": 0.1711,
+      "step": 310
+    },
+    {
+      "epoch": 0.24315871774824083,
+      "grad_norm": 0.34734436869621277,
+      "learning_rate": 9.985921688031252e-05,
+      "loss": 0.1629,
+      "step": 311
+    },
+    {
+      "epoch": 0.2439405785770133,
+      "grad_norm": 0.3847326338291168,
+      "learning_rate": 9.985405346887345e-05,
+      "loss": 0.2059,
+      "step": 312
+    },
+    {
+      "epoch": 0.24472243940578578,
+      "grad_norm": 0.43368545174598694,
+      "learning_rate": 9.984879720580816e-05,
+      "loss": 0.221,
+      "step": 313
+    },
+    {
+      "epoch": 0.24550430023455824,
+      "grad_norm": 0.3668220043182373,
+      "learning_rate": 9.984344810090623e-05,
+      "loss": 0.1626,
+      "step": 314
+    },
+    {
+      "epoch": 0.24628616106333073,
+      "grad_norm": 0.3731725215911865,
+      "learning_rate": 9.983800616413026e-05,
+      "loss": 0.142,
+      "step": 315
+    },
+    {
+      "epoch": 0.2470680218921032,
+      "grad_norm": 0.4757682681083679,
+      "learning_rate": 9.983247140561565e-05,
+      "loss": 0.1747,
+      "step": 316
+    },
+    {
+      "epoch": 0.24784988272087569,
+      "grad_norm": 0.3016088306903839,
+      "learning_rate": 9.982684383567071e-05,
+      "loss": 0.1177,
+      "step": 317
+    },
+    {
+      "epoch": 0.24863174354964818,
+      "grad_norm": 0.32493770122528076,
+      "learning_rate": 9.982112346477666e-05,
+      "loss": 0.1315,
+      "step": 318
+    },
+    {
+      "epoch": 0.24941360437842064,
+      "grad_norm": 0.2876770794391632,
+      "learning_rate": 9.981531030358746e-05,
+      "loss": 0.1167,
+      "step": 319
+    },
+    {
+      "epoch": 0.2501954652071931,
+      "grad_norm": 0.45965278148651123,
+      "learning_rate": 9.980940436292999e-05,
+      "loss": 0.1888,
+      "step": 320
+    },
+    {
+      "epoch": 0.2509773260359656,
+      "grad_norm": 0.4690196216106415,
+      "learning_rate": 9.980340565380382e-05,
+      "loss": 0.1482,
+      "step": 321
+    },
+    {
+      "epoch": 0.2517591868647381,
+      "grad_norm": 0.4229646623134613,
+      "learning_rate": 9.979731418738143e-05,
+      "loss": 0.1665,
+      "step": 322
+    },
+    {
+      "epoch": 0.25254104769351055,
+      "grad_norm": 0.33682507276535034,
+      "learning_rate": 9.979112997500792e-05,
+      "loss": 0.1573,
+      "step": 323
+    },
+    {
+      "epoch": 0.253322908522283,
+      "grad_norm": 0.44713473320007324,
+      "learning_rate": 9.978485302820125e-05,
+      "loss": 0.1575,
+      "step": 324
+    },
+    {
+      "epoch": 0.2541047693510555,
+      "grad_norm": 0.38661685585975647,
+      "learning_rate": 9.9778483358652e-05,
+      "loss": 0.123,
+      "step": 325
+    },
+    {
+      "epoch": 0.254886630179828,
+      "grad_norm": 0.41950613260269165,
+      "learning_rate": 9.977202097822352e-05,
+      "loss": 0.1819,
+      "step": 326
+    },
+    {
+      "epoch": 0.25566849100860045,
+      "grad_norm": 0.3908550441265106,
+      "learning_rate": 9.976546589895175e-05,
+      "loss": 0.164,
+      "step": 327
+    },
+    {
+      "epoch": 0.25645035183737297,
+      "grad_norm": 0.481545627117157,
+      "learning_rate": 9.975881813304535e-05,
+      "loss": 0.1764,
+      "step": 328
+    },
+    {
+      "epoch": 0.25723221266614543,
+      "grad_norm": 0.4087935984134674,
+      "learning_rate": 9.975207769288556e-05,
+      "loss": 0.1559,
+      "step": 329
+    },
+    {
+      "epoch": 0.2580140734949179,
+      "grad_norm": 0.285412073135376,
+      "learning_rate": 9.974524459102625e-05,
+      "loss": 0.1612,
+      "step": 330
+    },
+    {
+      "epoch": 0.25879593432369036,
+      "grad_norm": 0.413352906703949,
+      "learning_rate": 9.973831884019387e-05,
+      "loss": 0.1504,
+      "step": 331
+    },
+    {
+      "epoch": 0.2595777951524629,
+      "grad_norm": 0.48298683762550354,
+      "learning_rate": 9.97313004532874e-05,
+      "loss": 0.2061,
+      "step": 332
+    },
+    {
+      "epoch": 0.26035965598123534,
+      "grad_norm": 0.4352918863296509,
+      "learning_rate": 9.972418944337835e-05,
+      "loss": 0.2174,
+      "step": 333
+    },
+    {
+      "epoch": 0.2611415168100078,
+      "grad_norm": 0.3502805233001709,
+      "learning_rate": 9.971698582371081e-05,
+      "loss": 0.1882,
+      "step": 334
+    },
+    {
+      "epoch": 0.2619233776387803,
+      "grad_norm": 0.3264360725879669,
+      "learning_rate": 9.970968960770124e-05,
+      "loss": 0.1216,
+      "step": 335
+    },
+    {
+      "epoch": 0.2627052384675528,
+      "grad_norm": 0.42405977845191956,
+      "learning_rate": 9.970230080893866e-05,
+      "loss": 0.1781,
+      "step": 336
+    },
+    {
+      "epoch": 0.26348709929632524,
+      "grad_norm": 0.30436626076698303,
+      "learning_rate": 9.969481944118443e-05,
+      "loss": 0.1665,
+      "step": 337
+    },
+    {
+      "epoch": 0.2642689601250977,
+      "grad_norm": 0.3381073474884033,
+      "learning_rate": 9.968724551837243e-05,
+      "loss": 0.1661,
+      "step": 338
+    },
+    {
+      "epoch": 0.2650508209538702,
+      "grad_norm": 0.3266483545303345,
+      "learning_rate": 9.96795790546088e-05,
+      "loss": 0.1219,
+      "step": 339
+    },
+    {
+      "epoch": 0.2658326817826427,
+      "grad_norm": 0.31733569502830505,
+      "learning_rate": 9.967182006417212e-05,
+      "loss": 0.1555,
+      "step": 340
+    },
+    {
+      "epoch": 0.26661454261141515,
+      "grad_norm": 0.41691651940345764,
+      "learning_rate": 9.966396856151326e-05,
+      "loss": 0.192,
+      "step": 341
+    },
+    {
+      "epoch": 0.26739640344018767,
+      "grad_norm": 0.4389062225818634,
+      "learning_rate": 9.965602456125538e-05,
+      "loss": 0.2015,
+      "step": 342
+    },
+    {
+      "epoch": 0.26817826426896013,
+      "grad_norm": 0.3703081011772156,
+      "learning_rate": 9.964798807819397e-05,
+      "loss": 0.1676,
+      "step": 343
+    },
+    {
+      "epoch": 0.2689601250977326,
+      "grad_norm": 0.4071272909641266,
+      "learning_rate": 9.963985912729671e-05,
+      "loss": 0.1548,
+      "step": 344
+    },
+    {
+      "epoch": 0.26974198592650506,
+      "grad_norm": 0.3323352634906769,
+      "learning_rate": 9.963163772370352e-05,
+      "loss": 0.1208,
+      "step": 345
+    },
+    {
+      "epoch": 0.2705238467552776,
+      "grad_norm": 0.40458884835243225,
+      "learning_rate": 9.962332388272652e-05,
+      "loss": 0.1631,
+      "step": 346
+    },
+    {
+      "epoch": 0.27130570758405004,
+      "grad_norm": 0.3851662576198578,
+      "learning_rate": 9.961491761984996e-05,
+      "loss": 0.1305,
+      "step": 347
+    },
+    {
+      "epoch": 0.2720875684128225,
+      "grad_norm": 0.3678821325302124,
+      "learning_rate": 9.960641895073026e-05,
+      "loss": 0.1621,
+      "step": 348
+    },
+    {
+      "epoch": 0.272869429241595,
+      "grad_norm": 0.37880757451057434,
+      "learning_rate": 9.959782789119592e-05,
+      "loss": 0.1966,
+      "step": 349
+    },
+    {
+      "epoch": 0.2736512900703675,
+      "grad_norm": 0.43384692072868347,
+      "learning_rate": 9.958914445724754e-05,
+      "loss": 0.1812,
+      "step": 350
+    },
+    {
+      "epoch": 0.27443315089913994,
+      "grad_norm": 0.3404906094074249,
+      "learning_rate": 9.958036866505772e-05,
+      "loss": 0.1585,
+      "step": 351
+    },
+    {
+      "epoch": 0.2752150117279124,
+      "grad_norm": 0.2821943759918213,
+      "learning_rate": 9.95715005309711e-05,
+      "loss": 0.1368,
+      "step": 352
+    },
+    {
+      "epoch": 0.2759968725566849,
+      "grad_norm": 0.37595340609550476,
+      "learning_rate": 9.956254007150432e-05,
+      "loss": 0.1723,
+      "step": 353
+    },
+    {
+      "epoch": 0.2767787333854574,
+      "grad_norm": 0.41832900047302246,
+      "learning_rate": 9.955348730334595e-05,
+      "loss": 0.1956,
+      "step": 354
+    },
+    {
+      "epoch": 0.27756059421422985,
+      "grad_norm": 0.40022769570350647,
+      "learning_rate": 9.954434224335649e-05,
+      "loss": 0.1455,
+      "step": 355
+    },
+    {
+      "epoch": 0.27834245504300237,
+      "grad_norm": 0.2701699137687683,
+      "learning_rate": 9.953510490856834e-05,
+      "loss": 0.1547,
+      "step": 356
+    },
+    {
+      "epoch": 0.27912431587177483,
+      "grad_norm": 0.39053916931152344,
+      "learning_rate": 9.952577531618574e-05,
+      "loss": 0.1557,
+      "step": 357
+    },
+    {
+      "epoch": 0.2799061767005473,
+      "grad_norm": 0.3268284797668457,
+      "learning_rate": 9.951635348358476e-05,
+      "loss": 0.1425,
+      "step": 358
+    },
+    {
+      "epoch": 0.28068803752931976,
+      "grad_norm": 0.3364890217781067,
+      "learning_rate": 9.950683942831328e-05,
+      "loss": 0.1423,
+      "step": 359
+    },
+    {
+      "epoch": 0.2814698983580923,
+      "grad_norm": 0.4927947521209717,
+      "learning_rate": 9.949723316809093e-05,
+      "loss": 0.2059,
+      "step": 360
+    },
+    {
+      "epoch": 0.28225175918686474,
+      "grad_norm": 0.40654006600379944,
+      "learning_rate": 9.948753472080907e-05,
+      "loss": 0.1614,
+      "step": 361
+    },
+    {
+      "epoch": 0.2830336200156372,
+      "grad_norm": 0.37723296880722046,
+      "learning_rate": 9.947774410453077e-05,
+      "loss": 0.1289,
+      "step": 362
+    },
+    {
+      "epoch": 0.2838154808444097,
+      "grad_norm": 0.3067987263202667,
+      "learning_rate": 9.946786133749071e-05,
+      "loss": 0.1634,
+      "step": 363
+    },
+    {
+      "epoch": 0.2845973416731822,
+      "grad_norm": 0.49132147431373596,
+      "learning_rate": 9.94578864380953e-05,
+      "loss": 0.15,
+      "step": 364
+    },
+    {
+      "epoch": 0.28537920250195464,
+      "grad_norm": 0.35463565587997437,
+      "learning_rate": 9.944781942492242e-05,
+      "loss": 0.162,
+      "step": 365
+    },
+    {
+      "epoch": 0.2861610633307271,
+      "grad_norm": 0.4066864848136902,
+      "learning_rate": 9.943766031672159e-05,
+      "loss": 0.1448,
+      "step": 366
+    },
+    {
+      "epoch": 0.2869429241594996,
+      "grad_norm": 0.36959394812583923,
+      "learning_rate": 9.942740913241386e-05,
+      "loss": 0.1403,
+      "step": 367
+    },
+    {
+      "epoch": 0.2877247849882721,
+      "grad_norm": 0.3130083382129669,
+      "learning_rate": 9.94170658910917e-05,
+      "loss": 0.1549,
+      "step": 368
+    },
+    {
+      "epoch": 0.28850664581704455,
+      "grad_norm": 0.31939226388931274,
+      "learning_rate": 9.94066306120191e-05,
+      "loss": 0.1266,
+      "step": 369
+    },
+    {
+      "epoch": 0.28928850664581707,
+      "grad_norm": 0.34432169795036316,
+      "learning_rate": 9.939610331463142e-05,
+      "loss": 0.1326,
+      "step": 370
+    },
+    {
+      "epoch": 0.29007036747458953,
+      "grad_norm": 0.6027820706367493,
+      "learning_rate": 9.938548401853547e-05,
+      "loss": 0.1911,
+      "step": 371
+    },
+    {
+      "epoch": 0.290852228303362,
+      "grad_norm": 0.40641260147094727,
+      "learning_rate": 9.937477274350933e-05,
+      "loss": 0.1696,
+      "step": 372
+    },
+    {
+      "epoch": 0.29163408913213446,
+      "grad_norm": 0.4032672643661499,
+      "learning_rate": 9.93639695095024e-05,
+      "loss": 0.1581,
+      "step": 373
+    },
+    {
+      "epoch": 0.292415949960907,
+      "grad_norm": 0.40551677346229553,
+      "learning_rate": 9.93530743366354e-05,
+      "loss": 0.1522,
+      "step": 374
+    },
+    {
+      "epoch": 0.29319781078967944,
+      "grad_norm": 0.41328316926956177,
+      "learning_rate": 9.934208724520024e-05,
+      "loss": 0.1404,
+      "step": 375
+    },
+    {
+      "epoch": 0.2939796716184519,
+      "grad_norm": 0.3744495213031769,
+      "learning_rate": 9.933100825566002e-05,
+      "loss": 0.161,
+      "step": 376
+    },
+    {
+      "epoch": 0.2947615324472244,
+      "grad_norm": 0.3357921242713928,
+      "learning_rate": 9.931983738864904e-05,
+      "loss": 0.1528,
+      "step": 377
+    },
+    {
+      "epoch": 0.2955433932759969,
+      "grad_norm": 0.5647091269493103,
+      "learning_rate": 9.930857466497268e-05,
+      "loss": 0.1388,
+      "step": 378
+    },
+    {
+      "epoch": 0.29632525410476934,
+      "grad_norm": 0.4130406975746155,
+      "learning_rate": 9.92972201056074e-05,
+      "loss": 0.1372,
+      "step": 379
+    },
+    {
+      "epoch": 0.2971071149335418,
+      "grad_norm": 0.4709077775478363,
+      "learning_rate": 9.928577373170075e-05,
+      "loss": 0.197,
+      "step": 380
+    },
+    {
+      "epoch": 0.2978889757623143,
+      "grad_norm": 0.48450398445129395,
+      "learning_rate": 9.927423556457121e-05,
+      "loss": 0.1608,
+      "step": 381
+    },
+    {
+      "epoch": 0.2986708365910868,
+      "grad_norm": 0.35905659198760986,
+      "learning_rate": 9.926260562570829e-05,
+      "loss": 0.2213,
+      "step": 382
+    },
+    {
+      "epoch": 0.29945269741985925,
+      "grad_norm": 0.41346147656440735,
+      "learning_rate": 9.925088393677236e-05,
+      "loss": 0.1616,
+      "step": 383
+    },
+    {
+      "epoch": 0.30023455824863177,
+      "grad_norm": 0.41525787115097046,
+      "learning_rate": 9.92390705195947e-05,
+      "loss": 0.1985,
+      "step": 384
+    },
+    {
+      "epoch": 0.30101641907740423,
+      "grad_norm": 0.34878870844841003,
+      "learning_rate": 9.922716539617746e-05,
+      "loss": 0.1704,
+      "step": 385
+    },
+    {
+      "epoch": 0.3017982799061767,
+      "grad_norm": 0.4187116026878357,
+      "learning_rate": 9.921516858869355e-05,
+      "loss": 0.176,
+      "step": 386
+    },
+    {
+      "epoch": 0.30258014073494915,
+      "grad_norm": 0.35063982009887695,
+      "learning_rate": 9.920308011948665e-05,
+      "loss": 0.1408,
+      "step": 387
+    },
+    {
+      "epoch": 0.3033620015637217,
+      "grad_norm": 0.4096844494342804,
+      "learning_rate": 9.919090001107114e-05,
+      "loss": 0.1191,
+      "step": 388
+    },
+    {
+      "epoch": 0.30414386239249414,
+      "grad_norm": 0.3294246792793274,
+      "learning_rate": 9.917862828613214e-05,
+      "loss": 0.1618,
+      "step": 389
+    },
+    {
+      "epoch": 0.3049257232212666,
+      "grad_norm": 0.32012560963630676,
+      "learning_rate": 9.916626496752532e-05,
+      "loss": 0.1109,
+      "step": 390
+    },
+    {
+      "epoch": 0.3057075840500391,
+      "grad_norm": 0.46081307530403137,
+      "learning_rate": 9.915381007827698e-05,
+      "loss": 0.1631,
+      "step": 391
+    },
+    {
+      "epoch": 0.3064894448788116,
+      "grad_norm": 0.4402189552783966,
+      "learning_rate": 9.914126364158397e-05,
+      "loss": 0.13,
+      "step": 392
+    },
+    {
+      "epoch": 0.30727130570758404,
+      "grad_norm": 0.37083446979522705,
+      "learning_rate": 9.912862568081364e-05,
+      "loss": 0.1409,
+      "step": 393
+    },
+    {
+      "epoch": 0.3080531665363565,
+      "grad_norm": 0.42801108956336975,
+      "learning_rate": 9.91158962195038e-05,
+      "loss": 0.1682,
+      "step": 394
+    },
+    {
+      "epoch": 0.308835027365129,
+      "grad_norm": 0.3493399918079376,
+      "learning_rate": 9.910307528136266e-05,
+      "loss": 0.192,
+      "step": 395
+    },
+    {
+      "epoch": 0.3096168881939015,
+      "grad_norm": 0.42288097739219666,
+      "learning_rate": 9.909016289026886e-05,
+      "loss": 0.1677,
+      "step": 396
+    },
+    {
+      "epoch": 0.31039874902267395,
+      "grad_norm": 0.3308151960372925,
+      "learning_rate": 9.907715907027129e-05,
+      "loss": 0.1634,
+      "step": 397
+    },
+    {
+      "epoch": 0.31118060985144647,
+      "grad_norm": 0.3515646755695343,
+      "learning_rate": 9.906406384558917e-05,
+      "loss": 0.1285,
+      "step": 398
+    },
+    {
+      "epoch": 0.31196247068021893,
+      "grad_norm": 0.41209307312965393,
+      "learning_rate": 9.905087724061195e-05,
+      "loss": 0.1758,
+      "step": 399
+    },
+    {
+      "epoch": 0.3127443315089914,
+      "grad_norm": 0.3443009853363037,
+      "learning_rate": 9.903759927989929e-05,
+      "loss": 0.1419,
+      "step": 400
+    },
+    {
+      "epoch": 0.3127443315089914,
+      "eval_loss": 0.16743157804012299,
+      "eval_runtime": 13.3792,
+      "eval_samples_per_second": 3.887,
+      "eval_steps_per_second": 0.972,
+      "step": 400
+    },
+    {
+      "epoch": 0.31352619233776385,
+      "grad_norm": 0.2995968461036682,
+      "learning_rate": 9.902422998818094e-05,
+      "loss": 0.1303,
+      "step": 401
+    },
+    {
+      "epoch": 0.31430805316653637,
+      "grad_norm": 0.3437633812427521,
+      "learning_rate": 9.901076939035683e-05,
+      "loss": 0.1631,
+      "step": 402
+    },
+    {
+      "epoch": 0.31508991399530883,
+      "grad_norm": 0.3030770421028137,
+      "learning_rate": 9.899721751149688e-05,
+      "loss": 0.1429,
+      "step": 403
+    },
+    {
+      "epoch": 0.3158717748240813,
+      "grad_norm": 0.3427619934082031,
+      "learning_rate": 9.898357437684105e-05,
+      "loss": 0.1401,
+      "step": 404
+    },
+    {
+      "epoch": 0.3166536356528538,
+      "grad_norm": 0.3895941376686096,
+      "learning_rate": 9.896984001179925e-05,
+      "loss": 0.1392,
+      "step": 405
+    },
+    {
+      "epoch": 0.3174354964816263,
+      "grad_norm": 0.40631383657455444,
+      "learning_rate": 9.895601444195133e-05,
+      "loss": 0.1745,
+      "step": 406
+    },
+    {
+      "epoch": 0.31821735731039874,
+      "grad_norm": 0.3306792080402374,
+      "learning_rate": 9.894209769304696e-05,
+      "loss": 0.176,
+      "step": 407
+    },
+    {
+      "epoch": 0.3189992181391712,
+      "grad_norm": 0.27432066202163696,
+      "learning_rate": 9.892808979100567e-05,
+      "loss": 0.1236,
+      "step": 408
+    },
+    {
+      "epoch": 0.3197810789679437,
+      "grad_norm": 0.32954710721969604,
+      "learning_rate": 9.891399076191674e-05,
+      "loss": 0.1191,
+      "step": 409
+    },
+    {
+      "epoch": 0.3205629397967162,
+      "grad_norm": 0.3939608037471771,
+      "learning_rate": 9.889980063203916e-05,
+      "loss": 0.2076,
+      "step": 410
+    },
+    {
+      "epoch": 0.32134480062548865,
+      "grad_norm": 0.4199574291706085,
+      "learning_rate": 9.888551942780162e-05,
+      "loss": 0.1603,
+      "step": 411
+    },
+    {
+      "epoch": 0.32212666145426117,
+      "grad_norm": 0.35778728127479553,
+      "learning_rate": 9.887114717580241e-05,
+      "loss": 0.1764,
+      "step": 412
+    },
+    {
+      "epoch": 0.3229085222830336,
+      "grad_norm": 0.4640763998031616,
+      "learning_rate": 9.885668390280941e-05,
+      "loss": 0.184,
+      "step": 413
+    },
+    {
+      "epoch": 0.3236903831118061,
+      "grad_norm": 0.4196508824825287,
+      "learning_rate": 9.884212963576001e-05,
+      "loss": 0.1607,
+      "step": 414
+    },
+    {
+      "epoch": 0.32447224394057855,
+      "grad_norm": 0.3575003147125244,
+      "learning_rate": 9.882748440176109e-05,
+      "loss": 0.1732,
+      "step": 415
+    },
+    {
+      "epoch": 0.32525410476935107,
+      "grad_norm": 0.34807106852531433,
+      "learning_rate": 9.881274822808893e-05,
+      "loss": 0.1678,
+      "step": 416
+    },
+    {
+      "epoch": 0.32603596559812353,
+      "grad_norm": 0.25956061482429504,
+      "learning_rate": 9.879792114218921e-05,
+      "loss": 0.1139,
+      "step": 417
+    },
+    {
+      "epoch": 0.326817826426896,
+      "grad_norm": 0.2868465781211853,
+      "learning_rate": 9.87830031716769e-05,
+      "loss": 0.1103,
+      "step": 418
+    },
+    {
+      "epoch": 0.3275996872556685,
+      "grad_norm": 0.2630326747894287,
+      "learning_rate": 9.876799434433628e-05,
+      "loss": 0.1424,
+      "step": 419
+    },
+    {
+      "epoch": 0.328381548084441,
+      "grad_norm": 0.30072352290153503,
+      "learning_rate": 9.875289468812081e-05,
+      "loss": 0.1615,
+      "step": 420
+    },
+    {
+      "epoch": 0.32916340891321344,
+      "grad_norm": 0.33478954434394836,
+      "learning_rate": 9.873770423115314e-05,
+      "loss": 0.1799,
+      "step": 421
+    },
+    {
+      "epoch": 0.3299452697419859,
+      "grad_norm": 0.41718190908432007,
+      "learning_rate": 9.8722423001725e-05,
+      "loss": 0.1556,
+      "step": 422
+    },
+    {
+      "epoch": 0.3307271305707584,
+      "grad_norm": 0.3917146623134613,
+      "learning_rate": 9.870705102829723e-05,
+      "loss": 0.1786,
+      "step": 423
+    },
+    {
+      "epoch": 0.3315089913995309,
+      "grad_norm": 0.47362616658210754,
+      "learning_rate": 9.869158833949966e-05,
+      "loss": 0.1621,
+      "step": 424
+    },
+    {
+      "epoch": 0.33229085222830335,
+      "grad_norm": 0.4118567109107971,
+      "learning_rate": 9.867603496413103e-05,
+      "loss": 0.1599,
+      "step": 425
+    },
+    {
+      "epoch": 0.33307271305707586,
+      "grad_norm": 0.421069860458374,
+      "learning_rate": 9.866039093115905e-05,
+      "loss": 0.21,
+      "step": 426
+    },
+    {
+      "epoch": 0.3338545738858483,
+      "grad_norm": 0.33168917894363403,
+      "learning_rate": 9.864465626972023e-05,
+      "loss": 0.1465,
+      "step": 427
+    },
+    {
+      "epoch": 0.3346364347146208,
+      "grad_norm": 0.510787844657898,
+      "learning_rate": 9.862883100911991e-05,
+      "loss": 0.183,
+      "step": 428
+    },
+    {
+      "epoch": 0.33541829554339325,
+      "grad_norm": 0.27205660939216614,
+      "learning_rate": 9.861291517883213e-05,
+      "loss": 0.114,
+      "step": 429
+    },
+    {
+      "epoch": 0.33620015637216577,
+      "grad_norm": 0.38495227694511414,
+      "learning_rate": 9.859690880849962e-05,
+      "loss": 0.1453,
+      "step": 430
+    },
+    {
+      "epoch": 0.33698201720093823,
+      "grad_norm": 0.33769121766090393,
+      "learning_rate": 9.858081192793378e-05,
+      "loss": 0.1366,
+      "step": 431
+    },
+    {
+      "epoch": 0.3377638780297107,
+      "grad_norm": 0.3808351457118988,
+      "learning_rate": 9.856462456711451e-05,
+      "loss": 0.2269,
+      "step": 432
+    },
+    {
+      "epoch": 0.3385457388584832,
+      "grad_norm": 0.33304157853126526,
+      "learning_rate": 9.85483467561903e-05,
+      "loss": 0.1566,
+      "step": 433
+    },
+    {
+      "epoch": 0.3393275996872557,
+      "grad_norm": 0.34193313121795654,
+      "learning_rate": 9.853197852547802e-05,
+      "loss": 0.1855,
+      "step": 434
+    },
+    {
+      "epoch": 0.34010946051602814,
+      "grad_norm": 0.3733983337879181,
+      "learning_rate": 9.851551990546306e-05,
+      "loss": 0.1662,
+      "step": 435
+    },
+    {
+      "epoch": 0.3408913213448006,
+      "grad_norm": 0.3454480469226837,
+      "learning_rate": 9.849897092679903e-05,
+      "loss": 0.1305,
+      "step": 436
+    },
+    {
+      "epoch": 0.3416731821735731,
+      "grad_norm": 0.3219950199127197,
+      "learning_rate": 9.848233162030794e-05,
+      "loss": 0.1215,
+      "step": 437
+    },
+    {
+      "epoch": 0.3424550430023456,
+      "grad_norm": 0.3298224210739136,
+      "learning_rate": 9.846560201697993e-05,
+      "loss": 0.1481,
+      "step": 438
+    },
+    {
+      "epoch": 0.34323690383111805,
+      "grad_norm": 0.4454590976238251,
+      "learning_rate": 9.84487821479734e-05,
+      "loss": 0.116,
+      "step": 439
+    },
+    {
+      "epoch": 0.34401876465989056,
+      "grad_norm": 0.43617334961891174,
+      "learning_rate": 9.843187204461482e-05,
+      "loss": 0.1323,
+      "step": 440
+    },
+    {
+      "epoch": 0.344800625488663,
+      "grad_norm": 0.36367228627204895,
+      "learning_rate": 9.841487173839873e-05,
+      "loss": 0.1167,
+      "step": 441
+    },
+    {
+      "epoch": 0.3455824863174355,
+      "grad_norm": 0.4517843425273895,
+      "learning_rate": 9.839778126098769e-05,
+      "loss": 0.1517,
+      "step": 442
+    },
+    {
+      "epoch": 0.34636434714620795,
+      "grad_norm": 0.39846858382225037,
+      "learning_rate": 9.838060064421217e-05,
+      "loss": 0.147,
+      "step": 443
+    },
+    {
+      "epoch": 0.34714620797498047,
+      "grad_norm": 0.3574373126029968,
+      "learning_rate": 9.836332992007053e-05,
+      "loss": 0.1441,
+      "step": 444
+    },
+    {
+      "epoch": 0.34792806880375293,
+      "grad_norm": 0.3796740770339966,
+      "learning_rate": 9.834596912072897e-05,
+      "loss": 0.1642,
+      "step": 445
+    },
+    {
+      "epoch": 0.3487099296325254,
+      "grad_norm": 0.5189517736434937,
+      "learning_rate": 9.832851827852146e-05,
+      "loss": 0.1503,
+      "step": 446
+    },
+    {
+      "epoch": 0.3494917904612979,
+      "grad_norm": 0.4127172827720642,
+      "learning_rate": 9.831097742594958e-05,
+      "loss": 0.1651,
+      "step": 447
+    },
+    {
+      "epoch": 0.3502736512900704,
+      "grad_norm": 0.4301576316356659,
+      "learning_rate": 9.82933465956827e-05,
+      "loss": 0.1491,
+      "step": 448
+    },
+    {
+      "epoch": 0.35105551211884284,
+      "grad_norm": 0.3040635287761688,
+      "learning_rate": 9.827562582055765e-05,
+      "loss": 0.1265,
+      "step": 449
+    },
+    {
+      "epoch": 0.3518373729476153,
+      "grad_norm": 0.2885948121547699,
+      "learning_rate": 9.825781513357883e-05,
+      "loss": 0.1044,
+      "step": 450
+    },
+    {
+      "epoch": 0.3526192337763878,
+      "grad_norm": 0.43714582920074463,
+      "learning_rate": 9.823991456791811e-05,
+      "loss": 0.1536,
+      "step": 451
+    },
+    {
+      "epoch": 0.3534010946051603,
+      "grad_norm": 0.3937210142612457,
+      "learning_rate": 9.822192415691471e-05,
+      "loss": 0.1558,
+      "step": 452
+    },
+    {
+      "epoch": 0.35418295543393274,
+      "grad_norm": 0.4825761914253235,
+      "learning_rate": 9.820384393407525e-05,
+      "loss": 0.188,
+      "step": 453
+    },
+    {
+      "epoch": 0.35496481626270526,
+      "grad_norm": 0.47098153829574585,
+      "learning_rate": 9.818567393307354e-05,
+      "loss": 0.1939,
+      "step": 454
+    },
+    {
+      "epoch": 0.3557466770914777,
+      "grad_norm": 0.364969402551651,
+      "learning_rate": 9.816741418775066e-05,
+      "loss": 0.1361,
+      "step": 455
+    },
+    {
+      "epoch": 0.3565285379202502,
+      "grad_norm": 0.3509956896305084,
+      "learning_rate": 9.814906473211482e-05,
+      "loss": 0.1622,
+      "step": 456
+    },
+    {
+      "epoch": 0.35731039874902265,
+      "grad_norm": 0.4199284315109253,
+      "learning_rate": 9.813062560034134e-05,
+      "loss": 0.1774,
+      "step": 457
+    },
+    {
+      "epoch": 0.35809225957779517,
+      "grad_norm": 0.44852301478385925,
+      "learning_rate": 9.811209682677247e-05,
+      "loss": 0.1528,
+      "step": 458
+    },
+    {
+      "epoch": 0.35887412040656763,
+      "grad_norm": 0.2535669207572937,
+      "learning_rate": 9.809347844591753e-05,
+      "loss": 0.1465,
+      "step": 459
+    },
+    {
+      "epoch": 0.3596559812353401,
+      "grad_norm": 0.45635074377059937,
+      "learning_rate": 9.807477049245263e-05,
+      "loss": 0.1892,
+      "step": 460
+    },
+    {
+      "epoch": 0.3604378420641126,
+      "grad_norm": 0.45861974358558655,
+      "learning_rate": 9.805597300122081e-05,
+      "loss": 0.1636,
+      "step": 461
+    },
+    {
+      "epoch": 0.3612197028928851,
+      "grad_norm": 0.38760921359062195,
+      "learning_rate": 9.803708600723176e-05,
+      "loss": 0.1571,
+      "step": 462
+    },
+    {
+      "epoch": 0.36200156372165754,
+      "grad_norm": 0.34623679518699646,
+      "learning_rate": 9.801810954566195e-05,
+      "loss": 0.1432,
+      "step": 463
+    },
+    {
+      "epoch": 0.36278342455043,
+      "grad_norm": 0.4060858190059662,
+      "learning_rate": 9.799904365185442e-05,
+      "loss": 0.148,
+      "step": 464
+    },
+    {
+      "epoch": 0.3635652853792025,
+      "grad_norm": 0.3664781451225281,
+      "learning_rate": 9.797988836131884e-05,
+      "loss": 0.1386,
+      "step": 465
+    },
+    {
+      "epoch": 0.364347146207975,
+      "grad_norm": 0.3110964894294739,
+      "learning_rate": 9.796064370973133e-05,
+      "loss": 0.1201,
+      "step": 466
+    },
+    {
+      "epoch": 0.36512900703674744,
+      "grad_norm": 0.31303268671035767,
+      "learning_rate": 9.794130973293445e-05,
+      "loss": 0.1855,
+      "step": 467
+    },
+    {
+      "epoch": 0.36591086786551996,
+      "grad_norm": 0.27726083993911743,
+      "learning_rate": 9.792188646693714e-05,
+      "loss": 0.1037,
+      "step": 468
+    },
+    {
+      "epoch": 0.3666927286942924,
+      "grad_norm": 0.4889364540576935,
+      "learning_rate": 9.790237394791461e-05,
+      "loss": 0.1706,
+      "step": 469
+    },
+    {
+      "epoch": 0.3674745895230649,
+      "grad_norm": 0.41288602352142334,
+      "learning_rate": 9.788277221220836e-05,
+      "loss": 0.114,
+      "step": 470
+    },
+    {
+      "epoch": 0.36825645035183735,
+      "grad_norm": 0.42090147733688354,
+      "learning_rate": 9.786308129632598e-05,
+      "loss": 0.1901,
+      "step": 471
+    },
+    {
+      "epoch": 0.36903831118060987,
+      "grad_norm": 0.511559247970581,
+      "learning_rate": 9.78433012369412e-05,
+      "loss": 0.1699,
+      "step": 472
+    },
+    {
+      "epoch": 0.36982017200938233,
+      "grad_norm": 0.37128859758377075,
+      "learning_rate": 9.782343207089377e-05,
+      "loss": 0.1374,
+      "step": 473
+    },
+    {
+      "epoch": 0.3706020328381548,
+      "grad_norm": 0.34879276156425476,
+      "learning_rate": 9.780347383518937e-05,
+      "loss": 0.1421,
+      "step": 474
+    },
+    {
+      "epoch": 0.3713838936669273,
+      "grad_norm": 0.3927097022533417,
+      "learning_rate": 9.778342656699964e-05,
+      "loss": 0.1767,
+      "step": 475
+    },
+    {
+      "epoch": 0.3721657544956998,
+      "grad_norm": 0.35084983706474304,
+      "learning_rate": 9.776329030366196e-05,
+      "loss": 0.1329,
+      "step": 476
+    },
+    {
+      "epoch": 0.37294761532447224,
+      "grad_norm": 0.3951479494571686,
+      "learning_rate": 9.77430650826795e-05,
+      "loss": 0.1595,
+      "step": 477
+    },
+    {
+      "epoch": 0.3737294761532447,
+      "grad_norm": 0.3984255790710449,
+      "learning_rate": 9.77227509417211e-05,
+      "loss": 0.1838,
+      "step": 478
+    },
+    {
+      "epoch": 0.3745113369820172,
+      "grad_norm": 0.42549511790275574,
+      "learning_rate": 9.770234791862125e-05,
+      "loss": 0.1813,
+      "step": 479
+    },
+    {
+      "epoch": 0.3752931978107897,
+      "grad_norm": 0.35434046387672424,
+      "learning_rate": 9.768185605137992e-05,
+      "loss": 0.1598,
+      "step": 480
+    },
+    {
+      "epoch": 0.37607505863956214,
+      "grad_norm": 0.4132823348045349,
+      "learning_rate": 9.766127537816256e-05,
+      "loss": 0.1416,
+      "step": 481
+    },
+    {
+      "epoch": 0.37685691946833466,
+      "grad_norm": 0.37142178416252136,
+      "learning_rate": 9.764060593730007e-05,
+      "loss": 0.1558,
+      "step": 482
+    },
+    {
+      "epoch": 0.3776387802971071,
+      "grad_norm": 0.23793669044971466,
+      "learning_rate": 9.761984776728864e-05,
+      "loss": 0.0921,
+      "step": 483
+    },
+    {
+      "epoch": 0.3784206411258796,
+      "grad_norm": 0.3711697459220886,
+      "learning_rate": 9.759900090678967e-05,
+      "loss": 0.1345,
+      "step": 484
+    },
+    {
+      "epoch": 0.37920250195465205,
+      "grad_norm": 0.31402236223220825,
+      "learning_rate": 9.757806539462985e-05,
+      "loss": 0.118,
+      "step": 485
+    },
+    {
+      "epoch": 0.37998436278342457,
+      "grad_norm": 0.38717904686927795,
+      "learning_rate": 9.755704126980088e-05,
+      "loss": 0.1424,
+      "step": 486
+    },
+    {
+      "epoch": 0.38076622361219703,
+      "grad_norm": 0.4142398238182068,
+      "learning_rate": 9.753592857145957e-05,
+      "loss": 0.1651,
+      "step": 487
+    },
+    {
+      "epoch": 0.3815480844409695,
+      "grad_norm": 0.433275043964386,
+      "learning_rate": 9.751472733892763e-05,
+      "loss": 0.138,
+      "step": 488
+    },
+    {
+      "epoch": 0.382329945269742,
+      "grad_norm": 0.4147091507911682,
+      "learning_rate": 9.749343761169171e-05,
+      "loss": 0.1581,
+      "step": 489
+    },
+    {
+      "epoch": 0.3831118060985145,
+      "grad_norm": 0.37260568141937256,
+      "learning_rate": 9.74720594294033e-05,
+      "loss": 0.1656,
+      "step": 490
+    },
+    {
+      "epoch": 0.38389366692728694,
+      "grad_norm": 0.42818203568458557,
+      "learning_rate": 9.745059283187857e-05,
+      "loss": 0.1356,
+      "step": 491
+    },
+    {
+      "epoch": 0.3846755277560594,
+      "grad_norm": 0.35563212633132935,
+      "learning_rate": 9.742903785909838e-05,
+      "loss": 0.1773,
+      "step": 492
+    },
+    {
+      "epoch": 0.3854573885848319,
+      "grad_norm": 0.4813738167285919,
+      "learning_rate": 9.74073945512082e-05,
+      "loss": 0.1547,
+      "step": 493
+    },
+    {
+      "epoch": 0.3862392494136044,
+      "grad_norm": 0.29549703001976013,
+      "learning_rate": 9.738566294851805e-05,
+      "loss": 0.1249,
+      "step": 494
+    },
+    {
+      "epoch": 0.38702111024237684,
+      "grad_norm": 0.34138891100883484,
+      "learning_rate": 9.736384309150233e-05,
+      "loss": 0.1532,
+      "step": 495
+    },
+    {
+      "epoch": 0.38780297107114936,
+      "grad_norm": 0.37980690598487854,
+      "learning_rate": 9.734193502079987e-05,
+      "loss": 0.1146,
+      "step": 496
+    },
+    {
+      "epoch": 0.3885848318999218,
+      "grad_norm": 0.2760559916496277,
+      "learning_rate": 9.731993877721377e-05,
+      "loss": 0.1394,
+      "step": 497
+    },
+    {
+      "epoch": 0.3893666927286943,
+      "grad_norm": 0.3500741422176361,
+      "learning_rate": 9.729785440171133e-05,
+      "loss": 0.1481,
+      "step": 498
+    },
+    {
+      "epoch": 0.39014855355746675,
+      "grad_norm": 0.3756244480609894,
+      "learning_rate": 9.727568193542403e-05,
+      "loss": 0.1704,
+      "step": 499
+    },
+    {
+      "epoch": 0.39093041438623927,
+      "grad_norm": 0.36916160583496094,
+      "learning_rate": 9.72534214196474e-05,
+      "loss": 0.15,
+      "step": 500
+    },
+    {
+      "epoch": 0.39171227521501173,
+      "grad_norm": 0.33896347880363464,
+      "learning_rate": 9.723107289584095e-05,
+      "loss": 0.1379,
+      "step": 501
+    },
+    {
+      "epoch": 0.3924941360437842,
+      "grad_norm": 0.3901592791080475,
+      "learning_rate": 9.720863640562812e-05,
+      "loss": 0.1439,
+      "step": 502
+    },
+    {
+      "epoch": 0.3932759968725567,
+      "grad_norm": 0.43926864862442017,
+      "learning_rate": 9.718611199079617e-05,
+      "loss": 0.2371,
+      "step": 503
+    },
+    {
+      "epoch": 0.3940578577013292,
+      "grad_norm": 0.38037753105163574,
+      "learning_rate": 9.716349969329612e-05,
+      "loss": 0.1354,
+      "step": 504
+    },
+    {
+      "epoch": 0.39483971853010164,
+      "grad_norm": 0.37031248211860657,
+      "learning_rate": 9.714079955524269e-05,
+      "loss": 0.1233,
+      "step": 505
+    },
+    {
+      "epoch": 0.3956215793588741,
+      "grad_norm": 0.340681791305542,
+      "learning_rate": 9.711801161891417e-05,
+      "loss": 0.1247,
+      "step": 506
+    },
+    {
+      "epoch": 0.3964034401876466,
+      "grad_norm": 0.47795501351356506,
+      "learning_rate": 9.709513592675236e-05,
+      "loss": 0.1425,
+      "step": 507
+    },
+    {
+      "epoch": 0.3971853010164191,
+      "grad_norm": 0.4588595926761627,
+      "learning_rate": 9.707217252136257e-05,
+      "loss": 0.16,
+      "step": 508
+    },
+    {
+      "epoch": 0.39796716184519154,
+      "grad_norm": 0.41790610551834106,
+      "learning_rate": 9.704912144551341e-05,
+      "loss": 0.1514,
+      "step": 509
+    },
+    {
+      "epoch": 0.39874902267396406,
+      "grad_norm": 0.34573647379875183,
+      "learning_rate": 9.70259827421368e-05,
+      "loss": 0.1614,
+      "step": 510
+    },
+    {
+      "epoch": 0.3995308835027365,
+      "grad_norm": 0.5089054107666016,
+      "learning_rate": 9.700275645432784e-05,
+      "loss": 0.2109,
+      "step": 511
+    },
+    {
+      "epoch": 0.400312744331509,
+      "grad_norm": 0.38729050755500793,
+      "learning_rate": 9.697944262534478e-05,
+      "loss": 0.1341,
+      "step": 512
+    },
+    {
+      "epoch": 0.40109460516028145,
+      "grad_norm": 0.42094919085502625,
+      "learning_rate": 9.695604129860889e-05,
+      "loss": 0.12,
+      "step": 513
+    },
+    {
+      "epoch": 0.40187646598905397,
+      "grad_norm": 0.3481001555919647,
+      "learning_rate": 9.693255251770444e-05,
+      "loss": 0.1285,
+      "step": 514
+    },
+    {
+      "epoch": 0.40265832681782643,
+      "grad_norm": 0.37856054306030273,
+      "learning_rate": 9.690897632637852e-05,
+      "loss": 0.1127,
+      "step": 515
+    },
+    {
+      "epoch": 0.4034401876465989,
+      "grad_norm": 0.4085037112236023,
+      "learning_rate": 9.688531276854109e-05,
+      "loss": 0.1315,
+      "step": 516
+    },
+    {
+      "epoch": 0.4042220484753714,
+      "grad_norm": 0.34440815448760986,
+      "learning_rate": 9.686156188826478e-05,
+      "loss": 0.158,
+      "step": 517
+    },
+    {
+      "epoch": 0.40500390930414387,
+      "grad_norm": 0.4735249876976013,
+      "learning_rate": 9.683772372978485e-05,
+      "loss": 0.1583,
+      "step": 518
+    },
+    {
+      "epoch": 0.40578577013291633,
+      "grad_norm": 0.33127447962760925,
+      "learning_rate": 9.681379833749915e-05,
+      "loss": 0.1002,
+      "step": 519
+    },
+    {
+      "epoch": 0.4065676309616888,
+      "grad_norm": 0.3883433938026428,
+      "learning_rate": 9.678978575596795e-05,
+      "loss": 0.1379,
+      "step": 520
+    },
+    {
+      "epoch": 0.4073494917904613,
+      "grad_norm": 0.3916020691394806,
+      "learning_rate": 9.676568602991399e-05,
+      "loss": 0.1487,
+      "step": 521
+    },
+    {
+      "epoch": 0.4081313526192338,
+      "grad_norm": 0.34118443727493286,
+      "learning_rate": 9.674149920422222e-05,
+      "loss": 0.1419,
+      "step": 522
+    },
+    {
+      "epoch": 0.40891321344800624,
+      "grad_norm": 0.43779224157333374,
+      "learning_rate": 9.671722532393985e-05,
+      "loss": 0.1595,
+      "step": 523
+    },
+    {
+      "epoch": 0.40969507427677876,
+      "grad_norm": 0.43785735964775085,
+      "learning_rate": 9.669286443427625e-05,
+      "loss": 0.1651,
+      "step": 524
+    },
+    {
+      "epoch": 0.4104769351055512,
+      "grad_norm": 0.5933069586753845,
+      "learning_rate": 9.666841658060282e-05,
+      "loss": 0.1758,
+      "step": 525
+    },
+    {
+      "epoch": 0.4112587959343237,
+      "grad_norm": 0.4479697644710541,
+      "learning_rate": 9.66438818084529e-05,
+      "loss": 0.1534,
+      "step": 526
+    },
+    {
+      "epoch": 0.41204065676309615,
+      "grad_norm": 0.3687957525253296,
+      "learning_rate": 9.661926016352178e-05,
+      "loss": 0.1255,
+      "step": 527
+    },
+    {
+      "epoch": 0.41282251759186867,
+      "grad_norm": 0.3272929787635803,
+      "learning_rate": 9.659455169166648e-05,
+      "loss": 0.1403,
+      "step": 528
+    },
+    {
+      "epoch": 0.41360437842064113,
+      "grad_norm": 0.3177148103713989,
+      "learning_rate": 9.656975643890578e-05,
+      "loss": 0.1255,
+      "step": 529
+    },
+    {
+      "epoch": 0.4143862392494136,
+      "grad_norm": 0.44476059079170227,
+      "learning_rate": 9.654487445142004e-05,
+      "loss": 0.1435,
+      "step": 530
+    },
+    {
+      "epoch": 0.4151681000781861,
+      "grad_norm": 0.38959184288978577,
+      "learning_rate": 9.651990577555122e-05,
+      "loss": 0.1252,
+      "step": 531
+    },
+    {
+      "epoch": 0.41594996090695857,
+      "grad_norm": 0.5031436085700989,
+      "learning_rate": 9.64948504578027e-05,
+      "loss": 0.1476,
+      "step": 532
+    },
+    {
+      "epoch": 0.41673182173573103,
+      "grad_norm": 0.41138991713523865,
+      "learning_rate": 9.64697085448392e-05,
+      "loss": 0.1536,
+      "step": 533
+    },
+    {
+      "epoch": 0.4175136825645035,
+      "grad_norm": 0.3746177852153778,
+      "learning_rate": 9.644448008348679e-05,
+      "loss": 0.1369,
+      "step": 534
+    },
+    {
+      "epoch": 0.418295543393276,
+      "grad_norm": 0.29382434487342834,
+      "learning_rate": 9.641916512073268e-05,
+      "loss": 0.1398,
+      "step": 535
+    },
+    {
+      "epoch": 0.4190774042220485,
+      "grad_norm": 0.3424062132835388,
+      "learning_rate": 9.639376370372519e-05,
+      "loss": 0.1229,
+      "step": 536
+    },
+    {
+      "epoch": 0.41985926505082094,
+      "grad_norm": 0.4400944709777832,
+      "learning_rate": 9.636827587977368e-05,
+      "loss": 0.1298,
+      "step": 537
+    },
+    {
+      "epoch": 0.42064112587959346,
+      "grad_norm": 0.36449581384658813,
+      "learning_rate": 9.634270169634845e-05,
+      "loss": 0.1332,
+      "step": 538
+    },
+    {
+      "epoch": 0.4214229867083659,
+      "grad_norm": 0.405576229095459,
+      "learning_rate": 9.63170412010806e-05,
+      "loss": 0.1447,
+      "step": 539
+    },
+    {
+      "epoch": 0.4222048475371384,
+      "grad_norm": 0.35496416687965393,
+      "learning_rate": 9.629129444176202e-05,
+      "loss": 0.1537,
+      "step": 540
+    },
+    {
+      "epoch": 0.42298670836591085,
+      "grad_norm": 0.26653626561164856,
+      "learning_rate": 9.626546146634523e-05,
+      "loss": 0.1199,
+      "step": 541
+    },
+    {
+      "epoch": 0.42376856919468336,
+      "grad_norm": 0.3852273225784302,
+      "learning_rate": 9.623954232294335e-05,
+      "loss": 0.1065,
+      "step": 542
+    },
+    {
+      "epoch": 0.4245504300234558,
+      "grad_norm": 0.30202388763427734,
+      "learning_rate": 9.621353705982998e-05,
+      "loss": 0.1459,
+      "step": 543
+    },
+    {
+      "epoch": 0.4253322908522283,
+      "grad_norm": 0.555022120475769,
+      "learning_rate": 9.618744572543912e-05,
+      "loss": 0.1564,
+      "step": 544
+    },
+    {
+      "epoch": 0.4261141516810008,
+      "grad_norm": 0.3162443935871124,
+      "learning_rate": 9.616126836836508e-05,
+      "loss": 0.1132,
+      "step": 545
+    },
+    {
+      "epoch": 0.42689601250977327,
+      "grad_norm": 0.348391056060791,
+      "learning_rate": 9.613500503736237e-05,
+      "loss": 0.1762,
+      "step": 546
+    },
+    {
+      "epoch": 0.42767787333854573,
+      "grad_norm": 0.34790390729904175,
+      "learning_rate": 9.61086557813456e-05,
+      "loss": 0.1524,
+      "step": 547
+    },
+    {
+      "epoch": 0.4284597341673182,
+      "grad_norm": 0.3102438449859619,
+      "learning_rate": 9.608222064938947e-05,
+      "loss": 0.1189,
+      "step": 548
+    },
+    {
+      "epoch": 0.4292415949960907,
+      "grad_norm": 0.3740433156490326,
+      "learning_rate": 9.60556996907286e-05,
+      "loss": 0.1382,
+      "step": 549
+    },
+    {
+      "epoch": 0.4300234558248632,
+      "grad_norm": 0.3209870755672455,
+      "learning_rate": 9.602909295475743e-05,
+      "loss": 0.1858,
+      "step": 550
+    },
+    {
+      "epoch": 0.43080531665363564,
+      "grad_norm": 0.3138205409049988,
+      "learning_rate": 9.600240049103017e-05,
+      "loss": 0.135,
+      "step": 551
+    },
+    {
+      "epoch": 0.43158717748240816,
+      "grad_norm": 0.2969190180301666,
+      "learning_rate": 9.597562234926074e-05,
+      "loss": 0.1124,
+      "step": 552
+    },
+    {
+      "epoch": 0.4323690383111806,
+      "grad_norm": 0.4288076162338257,
+      "learning_rate": 9.594875857932258e-05,
+      "loss": 0.17,
+      "step": 553
+    },
+    {
+      "epoch": 0.4331508991399531,
+      "grad_norm": 0.36456018686294556,
+      "learning_rate": 9.592180923124861e-05,
+      "loss": 0.1732,
+      "step": 554
+    },
+    {
+      "epoch": 0.43393275996872555,
+      "grad_norm": 0.41333866119384766,
+      "learning_rate": 9.589477435523118e-05,
+      "loss": 0.1762,
+      "step": 555
+    },
+    {
+      "epoch": 0.43471462079749806,
+      "grad_norm": 0.3368997871875763,
+      "learning_rate": 9.58676540016219e-05,
+      "loss": 0.112,
+      "step": 556
+    },
+    {
+      "epoch": 0.4354964816262705,
+      "grad_norm": 0.29240164160728455,
+      "learning_rate": 9.584044822093157e-05,
+      "loss": 0.1659,
+      "step": 557
+    },
+    {
+      "epoch": 0.436278342455043,
+      "grad_norm": 0.29775160551071167,
+      "learning_rate": 9.581315706383013e-05,
+      "loss": 0.1055,
+      "step": 558
+    },
+    {
+      "epoch": 0.4370602032838155,
+      "grad_norm": 0.42036205530166626,
+      "learning_rate": 9.57857805811465e-05,
+      "loss": 0.1232,
+      "step": 559
+    },
+    {
+      "epoch": 0.43784206411258797,
+      "grad_norm": 0.36852604150772095,
+      "learning_rate": 9.575831882386855e-05,
+      "loss": 0.1359,
+      "step": 560
+    },
+    {
+      "epoch": 0.43862392494136043,
+      "grad_norm": 0.3674640357494354,
+      "learning_rate": 9.573077184314294e-05,
+      "loss": 0.1351,
+      "step": 561
+    },
+    {
+      "epoch": 0.4394057857701329,
+      "grad_norm": 0.37468084692955017,
+      "learning_rate": 9.570313969027504e-05,
+      "loss": 0.1353,
+      "step": 562
+    },
+    {
+      "epoch": 0.4401876465989054,
+      "grad_norm": 0.3501920998096466,
+      "learning_rate": 9.567542241672891e-05,
+      "loss": 0.1373,
+      "step": 563
+    },
+    {
+      "epoch": 0.4409695074276779,
+      "grad_norm": 0.5198925137519836,
+      "learning_rate": 9.564762007412711e-05,
+      "loss": 0.1635,
+      "step": 564
+    },
+    {
+      "epoch": 0.44175136825645034,
+      "grad_norm": 0.33644333481788635,
+      "learning_rate": 9.561973271425061e-05,
+      "loss": 0.1384,
+      "step": 565
+    },
+    {
+      "epoch": 0.44253322908522286,
+      "grad_norm": 0.2597680389881134,
+      "learning_rate": 9.559176038903879e-05,
+      "loss": 0.1368,
+      "step": 566
+    },
+    {
+      "epoch": 0.4433150899139953,
+      "grad_norm": 0.38457760214805603,
+      "learning_rate": 9.55637031505892e-05,
+      "loss": 0.1596,
+      "step": 567
+    },
+    {
+      "epoch": 0.4440969507427678,
+      "grad_norm": 0.33996647596359253,
+      "learning_rate": 9.55355610511576e-05,
+      "loss": 0.1253,
+      "step": 568
+    },
+    {
+      "epoch": 0.44487881157154024,
+      "grad_norm": 0.4009721279144287,
+      "learning_rate": 9.550733414315776e-05,
+      "loss": 0.1715,
+      "step": 569
+    },
+    {
+      "epoch": 0.44566067240031276,
+      "grad_norm": 0.4111771285533905,
+      "learning_rate": 9.547902247916143e-05,
+      "loss": 0.1479,
+      "step": 570
+    },
+    {
+      "epoch": 0.4464425332290852,
+      "grad_norm": 0.35021543502807617,
+      "learning_rate": 9.545062611189821e-05,
+      "loss": 0.118,
+      "step": 571
+    },
+    {
+      "epoch": 0.4472243940578577,
+      "grad_norm": 0.41540324687957764,
+      "learning_rate": 9.542214509425544e-05,
+      "loss": 0.1326,
+      "step": 572
+    },
+    {
+      "epoch": 0.4480062548866302,
+      "grad_norm": 0.4157445430755615,
+      "learning_rate": 9.539357947927815e-05,
+      "loss": 0.1728,
+      "step": 573
+    },
+    {
+      "epoch": 0.44878811571540267,
+      "grad_norm": 0.49535229802131653,
+      "learning_rate": 9.536492932016889e-05,
+      "loss": 0.134,
+      "step": 574
+    },
+    {
+      "epoch": 0.44956997654417513,
+      "grad_norm": 0.3090532720088959,
+      "learning_rate": 9.53361946702877e-05,
+      "loss": 0.1258,
+      "step": 575
+    },
+    {
+      "epoch": 0.4503518373729476,
+      "grad_norm": 0.32857659459114075,
+      "learning_rate": 9.530737558315196e-05,
+      "loss": 0.155,
+      "step": 576
+    },
+    {
+      "epoch": 0.4511336982017201,
+      "grad_norm": 0.43488988280296326,
+      "learning_rate": 9.527847211243635e-05,
+      "loss": 0.1295,
+      "step": 577
+    },
+    {
+      "epoch": 0.4519155590304926,
+      "grad_norm": 0.44877803325653076,
+      "learning_rate": 9.524948431197267e-05,
+      "loss": 0.1721,
+      "step": 578
+    },
+    {
+      "epoch": 0.45269741985926504,
+      "grad_norm": 0.32422056794166565,
+      "learning_rate": 9.52204122357498e-05,
+      "loss": 0.1148,
+      "step": 579
+    },
+    {
+      "epoch": 0.45347928068803756,
+      "grad_norm": 0.2760488986968994,
+      "learning_rate": 9.519125593791355e-05,
+      "loss": 0.1013,
+      "step": 580
+    },
+    {
+      "epoch": 0.45426114151681,
+      "grad_norm": 0.41548505425453186,
+      "learning_rate": 9.516201547276668e-05,
+      "loss": 0.1418,
+      "step": 581
+    },
+    {
+      "epoch": 0.4550430023455825,
+      "grad_norm": 0.3416723310947418,
+      "learning_rate": 9.513269089476862e-05,
+      "loss": 0.1955,
+      "step": 582
+    },
+    {
+      "epoch": 0.45582486317435494,
+      "grad_norm": 0.38667765259742737,
+      "learning_rate": 9.510328225853549e-05,
+      "loss": 0.1675,
+      "step": 583
+    },
+    {
+      "epoch": 0.45660672400312746,
+      "grad_norm": 0.3269575834274292,
+      "learning_rate": 9.507378961883992e-05,
+      "loss": 0.1565,
+      "step": 584
+    },
+    {
+      "epoch": 0.4573885848318999,
+      "grad_norm": 0.4353930652141571,
+      "learning_rate": 9.50442130306111e-05,
+      "loss": 0.1567,
+      "step": 585
+    },
+    {
+      "epoch": 0.4581704456606724,
+      "grad_norm": 0.3163772225379944,
+      "learning_rate": 9.501455254893447e-05,
+      "loss": 0.1271,
+      "step": 586
+    },
+    {
+      "epoch": 0.4589523064894449,
+      "grad_norm": 0.36551064252853394,
+      "learning_rate": 9.498480822905176e-05,
+      "loss": 0.1502,
+      "step": 587
+    },
+    {
+      "epoch": 0.45973416731821737,
+      "grad_norm": 0.3503546416759491,
+      "learning_rate": 9.495498012636085e-05,
+      "loss": 0.1323,
+      "step": 588
+    },
+    {
+      "epoch": 0.46051602814698983,
+      "grad_norm": 0.34271863102912903,
+      "learning_rate": 9.492506829641566e-05,
+      "loss": 0.1161,
+      "step": 589
+    },
+    {
+      "epoch": 0.4612978889757623,
+      "grad_norm": 0.3517410159111023,
+      "learning_rate": 9.4895072794926e-05,
+      "loss": 0.138,
+      "step": 590
+    },
+    {
+      "epoch": 0.4620797498045348,
+      "grad_norm": 0.4492360055446625,
+      "learning_rate": 9.486499367775764e-05,
+      "loss": 0.1446,
+      "step": 591
+    },
+    {
+      "epoch": 0.4628616106333073,
+      "grad_norm": 0.4240667521953583,
+      "learning_rate": 9.48348310009319e-05,
+      "loss": 0.1273,
+      "step": 592
+    },
+    {
+      "epoch": 0.46364347146207974,
+      "grad_norm": 0.3540504276752472,
+      "learning_rate": 9.480458482062594e-05,
+      "loss": 0.1645,
+      "step": 593
+    },
+    {
+      "epoch": 0.46442533229085226,
+      "grad_norm": 0.38456183671951294,
+      "learning_rate": 9.477425519317224e-05,
+      "loss": 0.1369,
+      "step": 594
+    },
+    {
+      "epoch": 0.4652071931196247,
+      "grad_norm": 0.36017024517059326,
+      "learning_rate": 9.474384217505883e-05,
+      "loss": 0.1233,
+      "step": 595
+    },
+    {
+      "epoch": 0.4659890539483972,
+      "grad_norm": 0.32265275716781616,
+      "learning_rate": 9.471334582292901e-05,
+      "loss": 0.1219,
+      "step": 596
+    },
+    {
+      "epoch": 0.46677091477716964,
+      "grad_norm": 0.36679038405418396,
+      "learning_rate": 9.468276619358129e-05,
+      "loss": 0.159,
+      "step": 597
+    },
+    {
+      "epoch": 0.46755277560594216,
+      "grad_norm": 0.36844977736473083,
+      "learning_rate": 9.465210334396927e-05,
+      "loss": 0.1405,
+      "step": 598
+    },
+    {
+      "epoch": 0.4683346364347146,
+      "grad_norm": 0.5122119188308716,
+      "learning_rate": 9.462135733120156e-05,
+      "loss": 0.1428,
+      "step": 599
+    },
+    {
+      "epoch": 0.4691164972634871,
+      "grad_norm": 0.35265064239501953,
+      "learning_rate": 9.459052821254166e-05,
+      "loss": 0.1137,
+      "step": 600
+    },
+    {
+      "epoch": 0.4691164972634871,
+      "eval_loss": 0.15098431706428528,
+      "eval_runtime": 13.3435,
+      "eval_samples_per_second": 3.897,
+      "eval_steps_per_second": 0.974,
+      "step": 600
+    },
+    {
+      "epoch": 0.4698983580922596,
+      "grad_norm": 0.38564780354499817,
+      "learning_rate": 9.455961604540784e-05,
+      "loss": 0.1325,
+      "step": 601
+    },
+    {
+      "epoch": 0.47068021892103207,
+      "grad_norm": 0.42470040917396545,
+      "learning_rate": 9.452862088737306e-05,
+      "loss": 0.1747,
+      "step": 602
+    },
+    {
+      "epoch": 0.47146207974980453,
+      "grad_norm": 0.5298458337783813,
+      "learning_rate": 9.449754279616481e-05,
+      "loss": 0.1574,
+      "step": 603
+    },
+    {
+      "epoch": 0.472243940578577,
+      "grad_norm": 0.4499131441116333,
+      "learning_rate": 9.446638182966511e-05,
+      "loss": 0.147,
+      "step": 604
+    },
+    {
+      "epoch": 0.4730258014073495,
+      "grad_norm": 0.3690560460090637,
+      "learning_rate": 9.443513804591026e-05,
+      "loss": 0.1243,
+      "step": 605
+    },
+    {
+      "epoch": 0.473807662236122,
+      "grad_norm": 0.29506915807724,
+      "learning_rate": 9.440381150309085e-05,
+      "loss": 0.1397,
+      "step": 606
+    },
+    {
+      "epoch": 0.47458952306489444,
+      "grad_norm": 0.347539484500885,
+      "learning_rate": 9.43724022595516e-05,
+      "loss": 0.1444,
+      "step": 607
+    },
+    {
+      "epoch": 0.47537138389366695,
+      "grad_norm": 0.3493543863296509,
+      "learning_rate": 9.434091037379125e-05,
+      "loss": 0.1533,
+      "step": 608
+    },
+    {
+      "epoch": 0.4761532447224394,
+      "grad_norm": 0.43107107281684875,
+      "learning_rate": 9.430933590446244e-05,
+      "loss": 0.1394,
+      "step": 609
+    },
+    {
+      "epoch": 0.4769351055512119,
+      "grad_norm": 0.38342204689979553,
+      "learning_rate": 9.427767891037165e-05,
+      "loss": 0.1304,
+      "step": 610
+    },
+    {
+      "epoch": 0.47771696637998434,
+      "grad_norm": 0.44626930356025696,
+      "learning_rate": 9.424593945047906e-05,
+      "loss": 0.1435,
+      "step": 611
+    },
+    {
+      "epoch": 0.47849882720875686,
+      "grad_norm": 0.3243478834629059,
+      "learning_rate": 9.42141175838984e-05,
+      "loss": 0.1265,
+      "step": 612
+    },
+    {
+      "epoch": 0.4792806880375293,
+      "grad_norm": 0.3624386191368103,
+      "learning_rate": 9.418221336989695e-05,
+      "loss": 0.1165,
+      "step": 613
+    },
+    {
+      "epoch": 0.4800625488663018,
+      "grad_norm": 0.5078319907188416,
+      "learning_rate": 9.415022686789528e-05,
+      "loss": 0.1526,
+      "step": 614
+    },
+    {
+      "epoch": 0.4808444096950743,
+      "grad_norm": 0.3112984299659729,
+      "learning_rate": 9.411815813746726e-05,
+      "loss": 0.1167,
+      "step": 615
+    },
+    {
+      "epoch": 0.48162627052384677,
+      "grad_norm": 0.41424402594566345,
+      "learning_rate": 9.408600723833993e-05,
+      "loss": 0.1527,
+      "step": 616
+    },
+    {
+      "epoch": 0.48240813135261923,
+      "grad_norm": 0.5095820426940918,
+      "learning_rate": 9.405377423039331e-05,
+      "loss": 0.1354,
+      "step": 617
+    },
+    {
+      "epoch": 0.4831899921813917,
+      "grad_norm": 0.3398195505142212,
+      "learning_rate": 9.402145917366041e-05,
+      "loss": 0.1319,
+      "step": 618
+    },
+    {
+      "epoch": 0.4839718530101642,
+      "grad_norm": 0.3965277671813965,
+      "learning_rate": 9.398906212832699e-05,
+      "loss": 0.1279,
+      "step": 619
+    },
+    {
+      "epoch": 0.4847537138389367,
+      "grad_norm": 0.38243114948272705,
+      "learning_rate": 9.395658315473154e-05,
+      "loss": 0.1689,
+      "step": 620
+    },
+    {
+      "epoch": 0.48553557466770914,
+      "grad_norm": 0.3739413917064667,
+      "learning_rate": 9.392402231336518e-05,
+      "loss": 0.1593,
+      "step": 621
+    },
+    {
+      "epoch": 0.48631743549648165,
+      "grad_norm": 0.40374016761779785,
+      "learning_rate": 9.389137966487143e-05,
+      "loss": 0.1493,
+      "step": 622
+    },
+    {
+      "epoch": 0.4870992963252541,
+      "grad_norm": 0.3396798074245453,
+      "learning_rate": 9.38586552700462e-05,
+      "loss": 0.1698,
+      "step": 623
+    },
+    {
+      "epoch": 0.4878811571540266,
+      "grad_norm": 0.34499019384384155,
+      "learning_rate": 9.38258491898377e-05,
+      "loss": 0.1315,
+      "step": 624
+    },
+    {
+      "epoch": 0.48866301798279904,
+      "grad_norm": 0.4252553880214691,
+      "learning_rate": 9.379296148534619e-05,
+      "loss": 0.1666,
+      "step": 625
+    },
+    {
+      "epoch": 0.48944487881157156,
+      "grad_norm": 0.3969322144985199,
+      "learning_rate": 9.375999221782402e-05,
+      "loss": 0.1067,
+      "step": 626
+    },
+    {
+      "epoch": 0.490226739640344,
+      "grad_norm": 0.3985759913921356,
+      "learning_rate": 9.372694144867544e-05,
+      "loss": 0.1974,
+      "step": 627
+    },
+    {
+      "epoch": 0.4910086004691165,
+      "grad_norm": 0.3676362931728363,
+      "learning_rate": 9.369380923945645e-05,
+      "loss": 0.1118,
+      "step": 628
+    },
+    {
+      "epoch": 0.491790461297889,
+      "grad_norm": 0.35326021909713745,
+      "learning_rate": 9.36605956518748e-05,
+      "loss": 0.1567,
+      "step": 629
+    },
+    {
+      "epoch": 0.49257232212666147,
+      "grad_norm": 0.3549090623855591,
+      "learning_rate": 9.362730074778973e-05,
+      "loss": 0.1293,
+      "step": 630
+    },
+    {
+      "epoch": 0.49335418295543393,
+      "grad_norm": 0.378653347492218,
+      "learning_rate": 9.359392458921198e-05,
+      "loss": 0.1272,
+      "step": 631
+    },
+    {
+      "epoch": 0.4941360437842064,
+      "grad_norm": 0.32727909088134766,
+      "learning_rate": 9.356046723830361e-05,
+      "loss": 0.1499,
+      "step": 632
+    },
+    {
+      "epoch": 0.4949179046129789,
+      "grad_norm": 0.30692726373672485,
+      "learning_rate": 9.352692875737787e-05,
+      "loss": 0.1832,
+      "step": 633
+    },
+    {
+      "epoch": 0.49569976544175137,
+      "grad_norm": 0.43790024518966675,
+      "learning_rate": 9.349330920889919e-05,
+      "loss": 0.1534,
+      "step": 634
+    },
+    {
+      "epoch": 0.49648162627052383,
+      "grad_norm": 0.2864377200603485,
+      "learning_rate": 9.34596086554829e-05,
+      "loss": 0.131,
+      "step": 635
+    },
+    {
+      "epoch": 0.49726348709929635,
+      "grad_norm": 0.2927132844924927,
+      "learning_rate": 9.342582715989525e-05,
+      "loss": 0.1064,
+      "step": 636
+    },
+    {
+      "epoch": 0.4980453479280688,
+      "grad_norm": 0.2664753794670105,
+      "learning_rate": 9.339196478505321e-05,
+      "loss": 0.0869,
+      "step": 637
+    },
+    {
+      "epoch": 0.4988272087568413,
+      "grad_norm": 0.3081669509410858,
+      "learning_rate": 9.33580215940244e-05,
+      "loss": 0.1209,
+      "step": 638
+    },
+    {
+      "epoch": 0.49960906958561374,
+      "grad_norm": 0.33297160267829895,
+      "learning_rate": 9.332399765002698e-05,
+      "loss": 0.1227,
+      "step": 639
+    },
+    {
+      "epoch": 0.5003909304143862,
+      "grad_norm": 0.4005429148674011,
+      "learning_rate": 9.328989301642947e-05,
+      "loss": 0.1268,
+      "step": 640
+    },
+    {
+      "epoch": 0.5011727912431587,
+      "grad_norm": 0.33951929211616516,
+      "learning_rate": 9.32557077567507e-05,
+      "loss": 0.1343,
+      "step": 641
+    },
+    {
+      "epoch": 0.5019546520719312,
+      "grad_norm": 0.3339715898036957,
+      "learning_rate": 9.322144193465966e-05,
+      "loss": 0.1252,
+      "step": 642
+    },
+    {
+      "epoch": 0.5027365129007036,
+      "grad_norm": 0.300782710313797,
+      "learning_rate": 9.318709561397537e-05,
+      "loss": 0.1014,
+      "step": 643
+    },
+    {
+      "epoch": 0.5035183737294762,
+      "grad_norm": 0.32149744033813477,
+      "learning_rate": 9.315266885866678e-05,
+      "loss": 0.1622,
+      "step": 644
+    },
+    {
+      "epoch": 0.5043002345582487,
+      "grad_norm": 0.3509116470813751,
+      "learning_rate": 9.311816173285268e-05,
+      "loss": 0.1517,
+      "step": 645
+    },
+    {
+      "epoch": 0.5050820953870211,
+      "grad_norm": 0.36596521735191345,
+      "learning_rate": 9.308357430080148e-05,
+      "loss": 0.1331,
+      "step": 646
+    },
+    {
+      "epoch": 0.5058639562157936,
+      "grad_norm": 0.43650323152542114,
+      "learning_rate": 9.304890662693123e-05,
+      "loss": 0.1469,
+      "step": 647
+    },
+    {
+      "epoch": 0.506645817044566,
+      "grad_norm": 0.48330414295196533,
+      "learning_rate": 9.301415877580938e-05,
+      "loss": 0.1492,
+      "step": 648
+    },
+    {
+      "epoch": 0.5074276778733385,
+      "grad_norm": 0.3281579315662384,
+      "learning_rate": 9.297933081215273e-05,
+      "loss": 0.1025,
+      "step": 649
+    },
+    {
+      "epoch": 0.508209538702111,
+      "grad_norm": 0.41272157430648804,
+      "learning_rate": 9.294442280082726e-05,
+      "loss": 0.1362,
+      "step": 650
+    },
+    {
+      "epoch": 0.5089913995308835,
+      "grad_norm": 0.3451278805732727,
+      "learning_rate": 9.29094348068481e-05,
+      "loss": 0.1411,
+      "step": 651
+    },
+    {
+      "epoch": 0.509773260359656,
+      "grad_norm": 0.45731282234191895,
+      "learning_rate": 9.287436689537928e-05,
+      "loss": 0.147,
+      "step": 652
+    },
+    {
+      "epoch": 0.5105551211884285,
+      "grad_norm": 0.47909867763519287,
+      "learning_rate": 9.283921913173368e-05,
+      "loss": 0.157,
+      "step": 653
+    },
+    {
+      "epoch": 0.5113369820172009,
+      "grad_norm": 0.3986799716949463,
+      "learning_rate": 9.280399158137295e-05,
+      "loss": 0.1244,
+      "step": 654
+    },
+    {
+      "epoch": 0.5121188428459734,
+      "grad_norm": 0.41379353404045105,
+      "learning_rate": 9.276868430990726e-05,
+      "loss": 0.1102,
+      "step": 655
+    },
+    {
+      "epoch": 0.5129007036747459,
+      "grad_norm": 0.414865642786026,
+      "learning_rate": 9.273329738309536e-05,
+      "loss": 0.1186,
+      "step": 656
+    },
+    {
+      "epoch": 0.5136825645035183,
+      "grad_norm": 0.2847290337085724,
+      "learning_rate": 9.269783086684428e-05,
+      "loss": 0.1188,
+      "step": 657
+    },
+    {
+      "epoch": 0.5144644253322909,
+      "grad_norm": 0.357957124710083,
+      "learning_rate": 9.266228482720929e-05,
+      "loss": 0.1381,
+      "step": 658
+    },
+    {
+      "epoch": 0.5152462861610634,
+      "grad_norm": 0.46672505140304565,
+      "learning_rate": 9.262665933039381e-05,
+      "loss": 0.1521,
+      "step": 659
+    },
+    {
+      "epoch": 0.5160281469898358,
+      "grad_norm": 0.3336876630783081,
+      "learning_rate": 9.25909544427492e-05,
+      "loss": 0.1222,
+      "step": 660
+    },
+    {
+      "epoch": 0.5168100078186083,
+      "grad_norm": 0.4090306758880615,
+      "learning_rate": 9.255517023077472e-05,
+      "loss": 0.1269,
+      "step": 661
+    },
+    {
+      "epoch": 0.5175918686473807,
+      "grad_norm": 0.42170941829681396,
+      "learning_rate": 9.251930676111735e-05,
+      "loss": 0.1365,
+      "step": 662
+    },
+    {
+      "epoch": 0.5183737294761532,
+      "grad_norm": 0.28856468200683594,
+      "learning_rate": 9.248336410057168e-05,
+      "loss": 0.1134,
+      "step": 663
+    },
+    {
+      "epoch": 0.5191555903049258,
+      "grad_norm": 0.3271030783653259,
+      "learning_rate": 9.244734231607981e-05,
+      "loss": 0.1565,
+      "step": 664
+    },
+    {
+      "epoch": 0.5199374511336982,
+      "grad_norm": 0.43632957339286804,
+      "learning_rate": 9.24112414747312e-05,
+      "loss": 0.1448,
+      "step": 665
+    },
+    {
+      "epoch": 0.5207193119624707,
+      "grad_norm": 0.32105788588523865,
+      "learning_rate": 9.237506164376252e-05,
+      "loss": 0.1092,
+      "step": 666
+    },
+    {
+      "epoch": 0.5215011727912432,
+      "grad_norm": 0.3537161350250244,
+      "learning_rate": 9.233880289055761e-05,
+      "loss": 0.1269,
+      "step": 667
+    },
+    {
+      "epoch": 0.5222830336200156,
+      "grad_norm": 0.37390899658203125,
+      "learning_rate": 9.230246528264726e-05,
+      "loss": 0.1456,
+      "step": 668
+    },
+    {
+      "epoch": 0.5230648944487881,
+      "grad_norm": 0.43096497654914856,
+      "learning_rate": 9.226604888770911e-05,
+      "loss": 0.1593,
+      "step": 669
+    },
+    {
+      "epoch": 0.5238467552775606,
+      "grad_norm": 0.34627631306648254,
+      "learning_rate": 9.222955377356761e-05,
+      "loss": 0.1423,
+      "step": 670
+    },
+    {
+      "epoch": 0.524628616106333,
+      "grad_norm": 0.3825629949569702,
+      "learning_rate": 9.219298000819376e-05,
+      "loss": 0.142,
+      "step": 671
+    },
+    {
+      "epoch": 0.5254104769351056,
+      "grad_norm": 0.338356614112854,
+      "learning_rate": 9.215632765970505e-05,
+      "loss": 0.1156,
+      "step": 672
+    },
+    {
+      "epoch": 0.5261923377638781,
+      "grad_norm": 0.3668399155139923,
+      "learning_rate": 9.211959679636535e-05,
+      "loss": 0.1801,
+      "step": 673
+    },
+    {
+      "epoch": 0.5269741985926505,
+      "grad_norm": 0.365161657333374,
+      "learning_rate": 9.208278748658476e-05,
+      "loss": 0.1168,
+      "step": 674
+    },
+    {
+      "epoch": 0.527756059421423,
+      "grad_norm": 0.4602578580379486,
+      "learning_rate": 9.204589979891946e-05,
+      "loss": 0.1381,
+      "step": 675
+    },
+    {
+      "epoch": 0.5285379202501954,
+      "grad_norm": 0.30687180161476135,
+      "learning_rate": 9.200893380207164e-05,
+      "loss": 0.1067,
+      "step": 676
+    },
+    {
+      "epoch": 0.5293197810789679,
+      "grad_norm": 0.2944718301296234,
+      "learning_rate": 9.197188956488931e-05,
+      "loss": 0.1255,
+      "step": 677
+    },
+    {
+      "epoch": 0.5301016419077405,
+      "grad_norm": 0.42425209283828735,
+      "learning_rate": 9.19347671563662e-05,
+      "loss": 0.1514,
+      "step": 678
+    },
+    {
+      "epoch": 0.5308835027365129,
+      "grad_norm": 0.33718201518058777,
+      "learning_rate": 9.189756664564167e-05,
+      "loss": 0.1645,
+      "step": 679
+    },
+    {
+      "epoch": 0.5316653635652854,
+      "grad_norm": 0.31876128911972046,
+      "learning_rate": 9.186028810200049e-05,
+      "loss": 0.1216,
+      "step": 680
+    },
+    {
+      "epoch": 0.5324472243940579,
+      "grad_norm": 0.3542174994945526,
+      "learning_rate": 9.182293159487281e-05,
+      "loss": 0.1186,
+      "step": 681
+    },
+    {
+      "epoch": 0.5332290852228303,
+      "grad_norm": 0.3566511571407318,
+      "learning_rate": 9.178549719383396e-05,
+      "loss": 0.1061,
+      "step": 682
+    },
+    {
+      "epoch": 0.5340109460516028,
+      "grad_norm": 0.3575066030025482,
+      "learning_rate": 9.174798496860433e-05,
+      "loss": 0.0926,
+      "step": 683
+    },
+    {
+      "epoch": 0.5347928068803753,
+      "grad_norm": 0.3495314121246338,
+      "learning_rate": 9.17103949890493e-05,
+      "loss": 0.1454,
+      "step": 684
+    },
+    {
+      "epoch": 0.5355746677091477,
+      "grad_norm": 0.37512341141700745,
+      "learning_rate": 9.167272732517903e-05,
+      "loss": 0.1329,
+      "step": 685
+    },
+    {
+      "epoch": 0.5363565285379203,
+      "grad_norm": 0.3389855921268463,
+      "learning_rate": 9.163498204714838e-05,
+      "loss": 0.1467,
+      "step": 686
+    },
+    {
+      "epoch": 0.5371383893666928,
+      "grad_norm": 0.40289631485939026,
+      "learning_rate": 9.159715922525673e-05,
+      "loss": 0.126,
+      "step": 687
+    },
+    {
+      "epoch": 0.5379202501954652,
+      "grad_norm": 0.31818968057632446,
+      "learning_rate": 9.155925892994794e-05,
+      "loss": 0.1571,
+      "step": 688
+    },
+    {
+      "epoch": 0.5387021110242377,
+      "grad_norm": 0.34707212448120117,
+      "learning_rate": 9.152128123181013e-05,
+      "loss": 0.1359,
+      "step": 689
+    },
+    {
+      "epoch": 0.5394839718530101,
+      "grad_norm": 0.46505051851272583,
+      "learning_rate": 9.148322620157558e-05,
+      "loss": 0.1432,
+      "step": 690
+    },
+    {
+      "epoch": 0.5402658326817826,
+      "grad_norm": 0.4175328016281128,
+      "learning_rate": 9.14450939101206e-05,
+      "loss": 0.1469,
+      "step": 691
+    },
+    {
+      "epoch": 0.5410476935105551,
+      "grad_norm": 0.3285068869590759,
+      "learning_rate": 9.140688442846539e-05,
+      "loss": 0.1264,
+      "step": 692
+    },
+    {
+      "epoch": 0.5418295543393276,
+      "grad_norm": 0.2715577483177185,
+      "learning_rate": 9.136859782777394e-05,
+      "loss": 0.1005,
+      "step": 693
+    },
+    {
+      "epoch": 0.5426114151681001,
+      "grad_norm": 0.33066704869270325,
+      "learning_rate": 9.133023417935384e-05,
+      "loss": 0.1466,
+      "step": 694
+    },
+    {
+      "epoch": 0.5433932759968726,
+      "grad_norm": 0.3862379491329193,
+      "learning_rate": 9.129179355465621e-05,
+      "loss": 0.156,
+      "step": 695
+    },
+    {
+      "epoch": 0.544175136825645,
+      "grad_norm": 0.5101954340934753,
+      "learning_rate": 9.125327602527551e-05,
+      "loss": 0.1619,
+      "step": 696
+    },
+    {
+      "epoch": 0.5449569976544175,
+      "grad_norm": 0.41626206040382385,
+      "learning_rate": 9.121468166294945e-05,
+      "loss": 0.1543,
+      "step": 697
+    },
+    {
+      "epoch": 0.54573885848319,
+      "grad_norm": 0.48254093527793884,
+      "learning_rate": 9.11760105395588e-05,
+      "loss": 0.1771,
+      "step": 698
+    },
+    {
+      "epoch": 0.5465207193119624,
+      "grad_norm": 0.44880276918411255,
+      "learning_rate": 9.113726272712734e-05,
+      "loss": 0.1401,
+      "step": 699
+    },
+    {
+      "epoch": 0.547302580140735,
+      "grad_norm": 0.3268358111381531,
+      "learning_rate": 9.10984382978217e-05,
+      "loss": 0.1429,
+      "step": 700
+    },
+    {
+      "epoch": 0.5480844409695075,
+      "grad_norm": 0.41424036026000977,
+      "learning_rate": 9.105953732395116e-05,
+      "loss": 0.1199,
+      "step": 701
+    },
+    {
+      "epoch": 0.5488663017982799,
+      "grad_norm": 0.5000854134559631,
+      "learning_rate": 9.102055987796755e-05,
+      "loss": 0.1374,
+      "step": 702
+    },
+    {
+      "epoch": 0.5496481626270524,
+      "grad_norm": 0.3157268464565277,
+      "learning_rate": 9.098150603246517e-05,
+      "loss": 0.1651,
+      "step": 703
+    },
+    {
+      "epoch": 0.5504300234558248,
+      "grad_norm": 0.45736557245254517,
+      "learning_rate": 9.09423758601806e-05,
+      "loss": 0.1301,
+      "step": 704
+    },
+    {
+      "epoch": 0.5512118842845973,
+      "grad_norm": 0.43551793694496155,
+      "learning_rate": 9.090316943399255e-05,
+      "loss": 0.1556,
+      "step": 705
+    },
+    {
+      "epoch": 0.5519937451133698,
+      "grad_norm": 0.3782845139503479,
+      "learning_rate": 9.08638868269218e-05,
+      "loss": 0.1131,
+      "step": 706
+    },
+    {
+      "epoch": 0.5527756059421423,
+      "grad_norm": 0.5090707540512085,
+      "learning_rate": 9.082452811213095e-05,
+      "loss": 0.1539,
+      "step": 707
+    },
+    {
+      "epoch": 0.5535574667709148,
+      "grad_norm": 0.33386969566345215,
+      "learning_rate": 9.078509336292443e-05,
+      "loss": 0.1171,
+      "step": 708
+    },
+    {
+      "epoch": 0.5543393275996873,
+      "grad_norm": 0.3908303678035736,
+      "learning_rate": 9.074558265274819e-05,
+      "loss": 0.1643,
+      "step": 709
+    },
+    {
+      "epoch": 0.5551211884284597,
+      "grad_norm": 0.28425249457359314,
+      "learning_rate": 9.070599605518974e-05,
+      "loss": 0.0953,
+      "step": 710
+    },
+    {
+      "epoch": 0.5559030492572322,
+      "grad_norm": 0.327131062746048,
+      "learning_rate": 9.066633364397786e-05,
+      "loss": 0.0948,
+      "step": 711
+    },
+    {
+      "epoch": 0.5566849100860047,
+      "grad_norm": 0.37248674035072327,
+      "learning_rate": 9.062659549298256e-05,
+      "loss": 0.1285,
+      "step": 712
+    },
+    {
+      "epoch": 0.5574667709147771,
+      "grad_norm": 0.3073717951774597,
+      "learning_rate": 9.058678167621493e-05,
+      "loss": 0.1424,
+      "step": 713
+    },
+    {
+      "epoch": 0.5582486317435497,
+      "grad_norm": 0.44110536575317383,
+      "learning_rate": 9.054689226782695e-05,
+      "loss": 0.1259,
+      "step": 714
+    },
+    {
+      "epoch": 0.5590304925723222,
+      "grad_norm": 0.3249659240245819,
+      "learning_rate": 9.050692734211142e-05,
+      "loss": 0.1455,
+      "step": 715
+    },
+    {
+      "epoch": 0.5598123534010946,
+      "grad_norm": 0.5413174629211426,
+      "learning_rate": 9.046688697350174e-05,
+      "loss": 0.1458,
+      "step": 716
+    },
+    {
+      "epoch": 0.5605942142298671,
+      "grad_norm": 0.3194544017314911,
+      "learning_rate": 9.042677123657191e-05,
+      "loss": 0.0932,
+      "step": 717
+    },
+    {
+      "epoch": 0.5613760750586395,
+      "grad_norm": 0.3347417712211609,
+      "learning_rate": 9.03865802060362e-05,
+      "loss": 0.1552,
+      "step": 718
+    },
+    {
+      "epoch": 0.562157935887412,
+      "grad_norm": 0.32833972573280334,
+      "learning_rate": 9.034631395674917e-05,
+      "loss": 0.1381,
+      "step": 719
+    },
+    {
+      "epoch": 0.5629397967161845,
+      "grad_norm": 0.46047019958496094,
+      "learning_rate": 9.030597256370544e-05,
+      "loss": 0.1887,
+      "step": 720
+    },
+    {
+      "epoch": 0.563721657544957,
+      "grad_norm": 0.48033323884010315,
+      "learning_rate": 9.026555610203964e-05,
+      "loss": 0.1406,
+      "step": 721
+    },
+    {
+      "epoch": 0.5645035183737295,
+      "grad_norm": 0.3175593316555023,
+      "learning_rate": 9.022506464702613e-05,
+      "loss": 0.1145,
+      "step": 722
+    },
+    {
+      "epoch": 0.565285379202502,
+      "grad_norm": 0.37334251403808594,
+      "learning_rate": 9.018449827407905e-05,
+      "loss": 0.1546,
+      "step": 723
+    },
+    {
+      "epoch": 0.5660672400312744,
+      "grad_norm": 0.4068865180015564,
+      "learning_rate": 9.014385705875192e-05,
+      "loss": 0.1292,
+      "step": 724
+    },
+    {
+      "epoch": 0.5668491008600469,
+      "grad_norm": 0.5138534903526306,
+      "learning_rate": 9.010314107673783e-05,
+      "loss": 0.1815,
+      "step": 725
+    },
+    {
+      "epoch": 0.5676309616888194,
+      "grad_norm": 0.4154388904571533,
+      "learning_rate": 9.006235040386897e-05,
+      "loss": 0.1455,
+      "step": 726
+    },
+    {
+      "epoch": 0.5684128225175918,
+      "grad_norm": 0.40405139327049255,
+      "learning_rate": 9.002148511611675e-05,
+      "loss": 0.1154,
+      "step": 727
+    },
+    {
+      "epoch": 0.5691946833463644,
+      "grad_norm": 0.30438432097435,
+      "learning_rate": 8.998054528959145e-05,
+      "loss": 0.1383,
+      "step": 728
+    },
+    {
+      "epoch": 0.5699765441751369,
+      "grad_norm": 0.31808093190193176,
+      "learning_rate": 8.993953100054224e-05,
+      "loss": 0.148,
+      "step": 729
+    },
+    {
+      "epoch": 0.5707584050039093,
+      "grad_norm": 0.4289185702800751,
+      "learning_rate": 8.989844232535699e-05,
+      "loss": 0.1508,
+      "step": 730
+    },
+    {
+      "epoch": 0.5715402658326818,
+      "grad_norm": 0.4019654393196106,
+      "learning_rate": 8.985727934056207e-05,
+      "loss": 0.114,
+      "step": 731
+    },
+    {
+      "epoch": 0.5723221266614542,
+      "grad_norm": 0.4127427339553833,
+      "learning_rate": 8.981604212282222e-05,
+      "loss": 0.1474,
+      "step": 732
+    },
+    {
+      "epoch": 0.5731039874902267,
+      "grad_norm": 0.45876824855804443,
+      "learning_rate": 8.977473074894052e-05,
+      "loss": 0.1567,
+      "step": 733
+    },
+    {
+      "epoch": 0.5738858483189992,
+      "grad_norm": 0.32999297976493835,
+      "learning_rate": 8.973334529585813e-05,
+      "loss": 0.1318,
+      "step": 734
+    },
+    {
+      "epoch": 0.5746677091477717,
+      "grad_norm": 0.378089964389801,
+      "learning_rate": 8.969188584065412e-05,
+      "loss": 0.1084,
+      "step": 735
+    },
+    {
+      "epoch": 0.5754495699765442,
+      "grad_norm": 0.32786768674850464,
+      "learning_rate": 8.965035246054549e-05,
+      "loss": 0.0991,
+      "step": 736
+    },
+    {
+      "epoch": 0.5762314308053167,
+      "grad_norm": 0.31518277525901794,
+      "learning_rate": 8.960874523288683e-05,
+      "loss": 0.1247,
+      "step": 737
+    },
+    {
+      "epoch": 0.5770132916340891,
+      "grad_norm": 0.2880820035934448,
+      "learning_rate": 8.956706423517034e-05,
+      "loss": 0.0881,
+      "step": 738
+    },
+    {
+      "epoch": 0.5777951524628616,
+      "grad_norm": 0.4550432860851288,
+      "learning_rate": 8.952530954502557e-05,
+      "loss": 0.1425,
+      "step": 739
+    },
+    {
+      "epoch": 0.5785770132916341,
+      "grad_norm": 0.29489225149154663,
+      "learning_rate": 8.948348124021933e-05,
+      "loss": 0.0933,
+      "step": 740
+    },
+    {
+      "epoch": 0.5793588741204065,
+      "grad_norm": 0.4262470304965973,
+      "learning_rate": 8.944157939865556e-05,
+      "loss": 0.127,
+      "step": 741
+    },
+    {
+      "epoch": 0.5801407349491791,
+      "grad_norm": 0.32060596346855164,
+      "learning_rate": 8.93996040983751e-05,
+      "loss": 0.1457,
+      "step": 742
+    },
+    {
+      "epoch": 0.5809225957779516,
+      "grad_norm": 0.325571209192276,
+      "learning_rate": 8.935755541755569e-05,
+      "loss": 0.1188,
+      "step": 743
+    },
+    {
+      "epoch": 0.581704456606724,
+      "grad_norm": 0.3205192983150482,
+      "learning_rate": 8.931543343451162e-05,
+      "loss": 0.114,
+      "step": 744
+    },
+    {
+      "epoch": 0.5824863174354965,
+      "grad_norm": 0.36551687121391296,
+      "learning_rate": 8.927323822769386e-05,
+      "loss": 0.1435,
+      "step": 745
+    },
+    {
+      "epoch": 0.5832681782642689,
+      "grad_norm": 0.4025515615940094,
+      "learning_rate": 8.923096987568965e-05,
+      "loss": 0.1171,
+      "step": 746
+    },
+    {
+      "epoch": 0.5840500390930414,
+      "grad_norm": 0.46686384081840515,
+      "learning_rate": 8.918862845722243e-05,
+      "loss": 0.1771,
+      "step": 747
+    },
+    {
+      "epoch": 0.584831899921814,
+      "grad_norm": 0.3949642777442932,
+      "learning_rate": 8.914621405115185e-05,
+      "loss": 0.1385,
+      "step": 748
+    },
+    {
+      "epoch": 0.5856137607505864,
+      "grad_norm": 0.37240198254585266,
+      "learning_rate": 8.910372673647336e-05,
+      "loss": 0.1211,
+      "step": 749
+    },
+    {
+      "epoch": 0.5863956215793589,
+      "grad_norm": 0.4359165132045746,
+      "learning_rate": 8.906116659231829e-05,
+      "loss": 0.1488,
+      "step": 750
+    },
+    {
+      "epoch": 0.5871774824081314,
+      "grad_norm": 0.4202546775341034,
+      "learning_rate": 8.901853369795361e-05,
+      "loss": 0.1468,
+      "step": 751
+    },
+    {
+      "epoch": 0.5879593432369038,
+      "grad_norm": 0.4482746124267578,
+      "learning_rate": 8.897582813278173e-05,
+      "loss": 0.1645,
+      "step": 752
+    },
+    {
+      "epoch": 0.5887412040656763,
+      "grad_norm": 0.4343931972980499,
+      "learning_rate": 8.893304997634045e-05,
+      "loss": 0.1803,
+      "step": 753
+    },
+    {
+      "epoch": 0.5895230648944488,
+      "grad_norm": 0.3664369285106659,
+      "learning_rate": 8.889019930830276e-05,
+      "loss": 0.122,
+      "step": 754
+    },
+    {
+      "epoch": 0.5903049257232212,
+      "grad_norm": 0.3301883637905121,
+      "learning_rate": 8.884727620847669e-05,
+      "loss": 0.1462,
+      "step": 755
+    },
+    {
+      "epoch": 0.5910867865519938,
+      "grad_norm": 0.4846368432044983,
+      "learning_rate": 8.88042807568052e-05,
+      "loss": 0.1361,
+      "step": 756
+    },
+    {
+      "epoch": 0.5918686473807663,
+      "grad_norm": 0.42507967352867126,
+      "learning_rate": 8.876121303336596e-05,
+      "loss": 0.1647,
+      "step": 757
+    },
+    {
+      "epoch": 0.5926505082095387,
+      "grad_norm": 0.2991779148578644,
+      "learning_rate": 8.871807311837128e-05,
+      "loss": 0.1248,
+      "step": 758
+    },
+    {
+      "epoch": 0.5934323690383112,
+      "grad_norm": 0.28716185688972473,
+      "learning_rate": 8.867486109216789e-05,
+      "loss": 0.097,
+      "step": 759
+    },
+    {
+      "epoch": 0.5942142298670836,
+      "grad_norm": 0.3456084430217743,
+      "learning_rate": 8.863157703523689e-05,
+      "loss": 0.124,
+      "step": 760
+    },
+    {
+      "epoch": 0.5949960906958561,
+      "grad_norm": 0.3598391115665436,
+      "learning_rate": 8.858822102819347e-05,
+      "loss": 0.1193,
+      "step": 761
+    },
+    {
+      "epoch": 0.5957779515246286,
+      "grad_norm": 0.42036497592926025,
+      "learning_rate": 8.854479315178681e-05,
+      "loss": 0.1636,
+      "step": 762
+    },
+    {
+      "epoch": 0.596559812353401,
+      "grad_norm": 0.3865693211555481,
+      "learning_rate": 8.850129348690004e-05,
+      "loss": 0.1576,
+      "step": 763
+    },
+    {
+      "epoch": 0.5973416731821736,
+      "grad_norm": 0.425310343503952,
+      "learning_rate": 8.845772211454992e-05,
+      "loss": 0.1141,
+      "step": 764
+    },
+    {
+      "epoch": 0.5981235340109461,
+      "grad_norm": 0.37581977248191833,
+      "learning_rate": 8.841407911588675e-05,
+      "loss": 0.1325,
+      "step": 765
+    },
+    {
+      "epoch": 0.5989053948397185,
+      "grad_norm": 0.2963871955871582,
+      "learning_rate": 8.837036457219427e-05,
+      "loss": 0.1279,
+      "step": 766
+    },
+    {
+      "epoch": 0.599687255668491,
+      "grad_norm": 0.4006364643573761,
+      "learning_rate": 8.832657856488949e-05,
+      "loss": 0.1468,
+      "step": 767
+    },
+    {
+      "epoch": 0.6004691164972635,
+      "grad_norm": 0.381542831659317,
+      "learning_rate": 8.828272117552245e-05,
+      "loss": 0.1094,
+      "step": 768
+    },
+    {
+      "epoch": 0.6012509773260359,
+      "grad_norm": 0.3772754371166229,
+      "learning_rate": 8.82387924857762e-05,
+      "loss": 0.1445,
+      "step": 769
+    },
+    {
+      "epoch": 0.6020328381548085,
+      "grad_norm": 0.34870976209640503,
+      "learning_rate": 8.819479257746655e-05,
+      "loss": 0.1258,
+      "step": 770
+    },
+    {
+      "epoch": 0.602814698983581,
+      "grad_norm": 0.28418296575546265,
+      "learning_rate": 8.815072153254195e-05,
+      "loss": 0.1116,
+      "step": 771
+    },
+    {
+      "epoch": 0.6035965598123534,
+      "grad_norm": 0.36682528257369995,
+      "learning_rate": 8.810657943308338e-05,
+      "loss": 0.1149,
+      "step": 772
+    },
+    {
+      "epoch": 0.6043784206411259,
+      "grad_norm": 0.3400062918663025,
+      "learning_rate": 8.806236636130411e-05,
+      "loss": 0.1356,
+      "step": 773
+    },
+    {
+      "epoch": 0.6051602814698983,
+      "grad_norm": 0.35574647784233093,
+      "learning_rate": 8.801808239954962e-05,
+      "loss": 0.1257,
+      "step": 774
+    },
+    {
+      "epoch": 0.6059421422986708,
+      "grad_norm": 0.3617613911628723,
+      "learning_rate": 8.797372763029742e-05,
+      "loss": 0.1146,
+      "step": 775
+    },
+    {
+      "epoch": 0.6067240031274433,
+      "grad_norm": 0.3895418345928192,
+      "learning_rate": 8.792930213615688e-05,
+      "loss": 0.0915,
+      "step": 776
+    },
+    {
+      "epoch": 0.6075058639562158,
+      "grad_norm": 0.42234644293785095,
+      "learning_rate": 8.78848059998691e-05,
+      "loss": 0.1145,
+      "step": 777
+    },
+    {
+      "epoch": 0.6082877247849883,
+      "grad_norm": 0.37372446060180664,
+      "learning_rate": 8.784023930430676e-05,
+      "loss": 0.1108,
+      "step": 778
+    },
+    {
+      "epoch": 0.6090695856137608,
+      "grad_norm": 0.4088364243507385,
+      "learning_rate": 8.779560213247395e-05,
+      "loss": 0.1355,
+      "step": 779
+    },
+    {
+      "epoch": 0.6098514464425332,
+      "grad_norm": 0.4049277603626251,
+      "learning_rate": 8.7750894567506e-05,
+      "loss": 0.142,
+      "step": 780
+    },
+    {
+      "epoch": 0.6106333072713057,
+      "grad_norm": 0.3075539469718933,
+      "learning_rate": 8.770611669266938e-05,
+      "loss": 0.1148,
+      "step": 781
+    },
+    {
+      "epoch": 0.6114151681000782,
+      "grad_norm": 0.399598091840744,
+      "learning_rate": 8.766126859136147e-05,
+      "loss": 0.129,
+      "step": 782
+    },
+    {
+      "epoch": 0.6121970289288506,
+      "grad_norm": 0.3626682758331299,
+      "learning_rate": 8.76163503471105e-05,
+      "loss": 0.1815,
+      "step": 783
+    },
+    {
+      "epoch": 0.6129788897576232,
+      "grad_norm": 0.3200612962245941,
+      "learning_rate": 8.757136204357527e-05,
+      "loss": 0.1188,
+      "step": 784
+    },
+    {
+      "epoch": 0.6137607505863957,
+      "grad_norm": 0.37934356927871704,
+      "learning_rate": 8.752630376454511e-05,
+      "loss": 0.0982,
+      "step": 785
+    },
+    {
+      "epoch": 0.6145426114151681,
+      "grad_norm": 0.3180946707725525,
+      "learning_rate": 8.748117559393967e-05,
+      "loss": 0.1988,
+      "step": 786
+    },
+    {
+      "epoch": 0.6153244722439406,
+      "grad_norm": 0.33251526951789856,
+      "learning_rate": 8.743597761580877e-05,
+      "loss": 0.1322,
+      "step": 787
+    },
+    {
+      "epoch": 0.616106333072713,
+      "grad_norm": 0.38982832431793213,
+      "learning_rate": 8.739070991433222e-05,
+      "loss": 0.0903,
+      "step": 788
+    },
+    {
+      "epoch": 0.6168881939014855,
+      "grad_norm": 0.3586364984512329,
+      "learning_rate": 8.734537257381973e-05,
+      "loss": 0.1362,
+      "step": 789
+    },
+    {
+      "epoch": 0.617670054730258,
+      "grad_norm": 0.42748382687568665,
+      "learning_rate": 8.729996567871068e-05,
+      "loss": 0.1566,
+      "step": 790
+    },
+    {
+      "epoch": 0.6184519155590305,
+      "grad_norm": 0.29496240615844727,
+      "learning_rate": 8.7254489313574e-05,
+      "loss": 0.1388,
+      "step": 791
+    },
+    {
+      "epoch": 0.619233776387803,
+      "grad_norm": 0.35175201296806335,
+      "learning_rate": 8.7208943563108e-05,
+      "loss": 0.1613,
+      "step": 792
+    },
+    {
+      "epoch": 0.6200156372165755,
+      "grad_norm": 0.3887656629085541,
+      "learning_rate": 8.716332851214024e-05,
+      "loss": 0.1116,
+      "step": 793
+    },
+    {
+      "epoch": 0.6207974980453479,
+      "grad_norm": 0.32240673899650574,
+      "learning_rate": 8.711764424562735e-05,
+      "loss": 0.1502,
+      "step": 794
+    },
+    {
+      "epoch": 0.6215793588741204,
+      "grad_norm": 0.3740167021751404,
+      "learning_rate": 8.707189084865481e-05,
+      "loss": 0.1332,
+      "step": 795
+    },
+    {
+      "epoch": 0.6223612197028929,
+      "grad_norm": 0.35816818475723267,
+      "learning_rate": 8.702606840643694e-05,
+      "loss": 0.1112,
+      "step": 796
+    },
+    {
+      "epoch": 0.6231430805316653,
+      "grad_norm": 0.33102670311927795,
+      "learning_rate": 8.698017700431662e-05,
+      "loss": 0.1437,
+      "step": 797
+    },
+    {
+      "epoch": 0.6239249413604379,
+      "grad_norm": 0.42163562774658203,
+      "learning_rate": 8.693421672776513e-05,
+      "loss": 0.1285,
+      "step": 798
+    },
+    {
+      "epoch": 0.6247068021892104,
+      "grad_norm": 0.2662132680416107,
+      "learning_rate": 8.688818766238208e-05,
+      "loss": 0.1039,
+      "step": 799
+    },
+    {
+      "epoch": 0.6254886630179828,
+      "grad_norm": 0.3629731237888336,
+      "learning_rate": 8.684208989389518e-05,
+      "loss": 0.1457,
+      "step": 800
+    },
+    {
+      "epoch": 0.6254886630179828,
+      "eval_loss": 0.14057856798171997,
+      "eval_runtime": 13.4351,
+      "eval_samples_per_second": 3.87,
+      "eval_steps_per_second": 0.968,
+      "step": 800
+    },
+    {
+      "epoch": 0.6262705238467553,
+      "grad_norm": 0.3074587285518646,
+      "learning_rate": 8.679592350816007e-05,
+      "loss": 0.1022,
+      "step": 801
+    },
+    {
+      "epoch": 0.6270523846755277,
+      "grad_norm": 0.3629996180534363,
+      "learning_rate": 8.67496885911602e-05,
+      "loss": 0.1427,
+      "step": 802
+    },
+    {
+      "epoch": 0.6278342455043002,
+      "grad_norm": 0.3123893439769745,
+      "learning_rate": 8.670338522900672e-05,
+      "loss": 0.1133,
+      "step": 803
+    },
+    {
+      "epoch": 0.6286161063330727,
+      "grad_norm": 0.3146839141845703,
+      "learning_rate": 8.665701350793816e-05,
+      "loss": 0.123,
+      "step": 804
+    },
+    {
+      "epoch": 0.6293979671618452,
+      "grad_norm": 0.40664559602737427,
+      "learning_rate": 8.66105735143204e-05,
+      "loss": 0.1416,
+      "step": 805
+    },
+    {
+      "epoch": 0.6301798279906177,
+      "grad_norm": 0.528860867023468,
+      "learning_rate": 8.656406533464654e-05,
+      "loss": 0.149,
+      "step": 806
+    },
+    {
+      "epoch": 0.6309616888193902,
+      "grad_norm": 0.317931592464447,
+      "learning_rate": 8.651748905553656e-05,
+      "loss": 0.1103,
+      "step": 807
+    },
+    {
+      "epoch": 0.6317435496481626,
+      "grad_norm": 0.40084466338157654,
+      "learning_rate": 8.647084476373737e-05,
+      "loss": 0.1439,
+      "step": 808
+    },
+    {
+      "epoch": 0.6325254104769351,
+      "grad_norm": 0.4050627052783966,
+      "learning_rate": 8.642413254612251e-05,
+      "loss": 0.1535,
+      "step": 809
+    },
+    {
+      "epoch": 0.6333072713057076,
+      "grad_norm": 0.39978843927383423,
+      "learning_rate": 8.637735248969203e-05,
+      "loss": 0.1434,
+      "step": 810
+    },
+    {
+      "epoch": 0.63408913213448,
+      "grad_norm": 0.41895702481269836,
+      "learning_rate": 8.633050468157234e-05,
+      "loss": 0.1349,
+      "step": 811
+    },
+    {
+      "epoch": 0.6348709929632526,
+      "grad_norm": 0.3688557744026184,
+      "learning_rate": 8.628358920901601e-05,
+      "loss": 0.1796,
+      "step": 812
+    },
+    {
+      "epoch": 0.6356528537920251,
+      "grad_norm": 0.3450753390789032,
+      "learning_rate": 8.623660615940165e-05,
+      "loss": 0.104,
+      "step": 813
+    },
+    {
+      "epoch": 0.6364347146207975,
+      "grad_norm": 0.39041540026664734,
+      "learning_rate": 8.618955562023378e-05,
+      "loss": 0.1186,
+      "step": 814
+    },
+    {
+      "epoch": 0.63721657544957,
+      "grad_norm": 0.3215588629245758,
+      "learning_rate": 8.61424376791425e-05,
+      "loss": 0.1145,
+      "step": 815
+    },
+    {
+      "epoch": 0.6379984362783424,
+      "grad_norm": 0.3149644732475281,
+      "learning_rate": 8.609525242388355e-05,
+      "loss": 0.1103,
+      "step": 816
+    },
+    {
+      "epoch": 0.6387802971071149,
+      "grad_norm": 0.3476389944553375,
+      "learning_rate": 8.604799994233798e-05,
+      "loss": 0.1299,
+      "step": 817
+    },
+    {
+      "epoch": 0.6395621579358874,
+      "grad_norm": 0.36129891872406006,
+      "learning_rate": 8.60006803225121e-05,
+      "loss": 0.1372,
+      "step": 818
+    },
+    {
+      "epoch": 0.6403440187646599,
+      "grad_norm": 0.3266359865665436,
+      "learning_rate": 8.595329365253719e-05,
+      "loss": 0.0957,
+      "step": 819
+    },
+    {
+      "epoch": 0.6411258795934324,
+      "grad_norm": 0.4800041913986206,
+      "learning_rate": 8.590584002066949e-05,
+      "loss": 0.1409,
+      "step": 820
+    },
+    {
+      "epoch": 0.6419077404222049,
+      "grad_norm": 0.34764230251312256,
+      "learning_rate": 8.585831951528991e-05,
+      "loss": 0.1204,
+      "step": 821
+    },
+    {
+      "epoch": 0.6426896012509773,
+      "grad_norm": 0.45669716596603394,
+      "learning_rate": 8.58107322249039e-05,
+      "loss": 0.1792,
+      "step": 822
+    },
+    {
+      "epoch": 0.6434714620797498,
+      "grad_norm": 0.3573746681213379,
+      "learning_rate": 8.576307823814132e-05,
+      "loss": 0.1175,
+      "step": 823
+    },
+    {
+      "epoch": 0.6442533229085223,
+      "grad_norm": 0.4184928238391876,
+      "learning_rate": 8.571535764375625e-05,
+      "loss": 0.1392,
+      "step": 824
+    },
+    {
+      "epoch": 0.6450351837372947,
+      "grad_norm": 0.4077909588813782,
+      "learning_rate": 8.566757053062678e-05,
+      "loss": 0.1322,
+      "step": 825
+    },
+    {
+      "epoch": 0.6458170445660673,
+      "grad_norm": 0.2542969584465027,
+      "learning_rate": 8.561971698775496e-05,
+      "loss": 0.0923,
+      "step": 826
+    },
+    {
+      "epoch": 0.6465989053948398,
+      "grad_norm": 0.3500221073627472,
+      "learning_rate": 8.55717971042665e-05,
+      "loss": 0.123,
+      "step": 827
+    },
+    {
+      "epoch": 0.6473807662236122,
+      "grad_norm": 0.3936273753643036,
+      "learning_rate": 8.552381096941074e-05,
+      "loss": 0.1216,
+      "step": 828
+    },
+    {
+      "epoch": 0.6481626270523847,
+      "grad_norm": 0.37407663464546204,
+      "learning_rate": 8.54757586725603e-05,
+      "loss": 0.1401,
+      "step": 829
+    },
+    {
+      "epoch": 0.6489444878811571,
+      "grad_norm": 0.3651149868965149,
+      "learning_rate": 8.542764030321115e-05,
+      "loss": 0.1006,
+      "step": 830
+    },
+    {
+      "epoch": 0.6497263487099296,
+      "grad_norm": 0.3681272268295288,
+      "learning_rate": 8.537945595098222e-05,
+      "loss": 0.1314,
+      "step": 831
+    },
+    {
+      "epoch": 0.6505082095387021,
+      "grad_norm": 0.37228938937187195,
+      "learning_rate": 8.53312057056154e-05,
+      "loss": 0.1523,
+      "step": 832
+    },
+    {
+      "epoch": 0.6512900703674745,
+      "grad_norm": 0.2094591110944748,
+      "learning_rate": 8.528288965697523e-05,
+      "loss": 0.0931,
+      "step": 833
+    },
+    {
+      "epoch": 0.6520719311962471,
+      "grad_norm": 0.39157184958457947,
+      "learning_rate": 8.52345078950489e-05,
+      "loss": 0.1123,
+      "step": 834
+    },
+    {
+      "epoch": 0.6528537920250196,
+      "grad_norm": 0.34393438696861267,
+      "learning_rate": 8.518606050994591e-05,
+      "loss": 0.1333,
+      "step": 835
+    },
+    {
+      "epoch": 0.653635652853792,
+      "grad_norm": 0.2535848021507263,
+      "learning_rate": 8.513754759189801e-05,
+      "loss": 0.0982,
+      "step": 836
+    },
+    {
+      "epoch": 0.6544175136825645,
+      "grad_norm": 0.3018800914287567,
+      "learning_rate": 8.508896923125901e-05,
+      "loss": 0.1341,
+      "step": 837
+    },
+    {
+      "epoch": 0.655199374511337,
+      "grad_norm": 0.3368026316165924,
+      "learning_rate": 8.50403255185046e-05,
+      "loss": 0.1357,
+      "step": 838
+    },
+    {
+      "epoch": 0.6559812353401094,
+      "grad_norm": 0.5133249163627625,
+      "learning_rate": 8.499161654423219e-05,
+      "loss": 0.139,
+      "step": 839
+    },
+    {
+      "epoch": 0.656763096168882,
+      "grad_norm": 0.36635470390319824,
+      "learning_rate": 8.494284239916071e-05,
+      "loss": 0.1016,
+      "step": 840
+    },
+    {
+      "epoch": 0.6575449569976545,
+      "grad_norm": 0.2705834209918976,
+      "learning_rate": 8.48940031741305e-05,
+      "loss": 0.0975,
+      "step": 841
+    },
+    {
+      "epoch": 0.6583268178264269,
+      "grad_norm": 0.4135679602622986,
+      "learning_rate": 8.484509896010311e-05,
+      "loss": 0.1148,
+      "step": 842
+    },
+    {
+      "epoch": 0.6591086786551994,
+      "grad_norm": 0.34528687596321106,
+      "learning_rate": 8.479612984816112e-05,
+      "loss": 0.1309,
+      "step": 843
+    },
+    {
+      "epoch": 0.6598905394839718,
+      "grad_norm": 0.33708104491233826,
+      "learning_rate": 8.474709592950798e-05,
+      "loss": 0.1598,
+      "step": 844
+    },
+    {
+      "epoch": 0.6606724003127443,
+      "grad_norm": 0.2598781883716583,
+      "learning_rate": 8.469799729546781e-05,
+      "loss": 0.0945,
+      "step": 845
+    },
+    {
+      "epoch": 0.6614542611415168,
+      "grad_norm": 0.32785603404045105,
+      "learning_rate": 8.464883403748534e-05,
+      "loss": 0.2023,
+      "step": 846
+    },
+    {
+      "epoch": 0.6622361219702892,
+      "grad_norm": 0.3188624978065491,
+      "learning_rate": 8.459960624712556e-05,
+      "loss": 0.1078,
+      "step": 847
+    },
+    {
+      "epoch": 0.6630179827990618,
+      "grad_norm": 0.36854130029678345,
+      "learning_rate": 8.455031401607375e-05,
+      "loss": 0.1239,
+      "step": 848
+    },
+    {
+      "epoch": 0.6637998436278343,
+      "grad_norm": 0.32640159130096436,
+      "learning_rate": 8.450095743613512e-05,
+      "loss": 0.1128,
+      "step": 849
+    },
+    {
+      "epoch": 0.6645817044566067,
+      "grad_norm": 0.4553893208503723,
+      "learning_rate": 8.44515365992348e-05,
+      "loss": 0.1313,
+      "step": 850
+    },
+    {
+      "epoch": 0.6653635652853792,
+      "grad_norm": 0.3296887278556824,
+      "learning_rate": 8.440205159741752e-05,
+      "loss": 0.1455,
+      "step": 851
+    },
+    {
+      "epoch": 0.6661454261141517,
+      "grad_norm": 0.30615031719207764,
+      "learning_rate": 8.435250252284762e-05,
+      "loss": 0.1123,
+      "step": 852
+    },
+    {
+      "epoch": 0.6669272869429241,
+      "grad_norm": 0.4171629548072815,
+      "learning_rate": 8.430288946780865e-05,
+      "loss": 0.1156,
+      "step": 853
+    },
+    {
+      "epoch": 0.6677091477716967,
+      "grad_norm": 0.42535823583602905,
+      "learning_rate": 8.425321252470342e-05,
+      "loss": 0.1247,
+      "step": 854
+    },
+    {
+      "epoch": 0.6684910086004691,
+      "grad_norm": 0.412070095539093,
+      "learning_rate": 8.420347178605367e-05,
+      "loss": 0.1028,
+      "step": 855
+    },
+    {
+      "epoch": 0.6692728694292416,
+      "grad_norm": 0.3225451707839966,
+      "learning_rate": 8.415366734450003e-05,
+      "loss": 0.1206,
+      "step": 856
+    },
+    {
+      "epoch": 0.6700547302580141,
+      "grad_norm": 0.4608080983161926,
+      "learning_rate": 8.410379929280168e-05,
+      "loss": 0.1485,
+      "step": 857
+    },
+    {
+      "epoch": 0.6708365910867865,
+      "grad_norm": 0.3869389295578003,
+      "learning_rate": 8.405386772383634e-05,
+      "loss": 0.1591,
+      "step": 858
+    },
+    {
+      "epoch": 0.671618451915559,
+      "grad_norm": 0.35010966658592224,
+      "learning_rate": 8.400387273059998e-05,
+      "loss": 0.1151,
+      "step": 859
+    },
+    {
+      "epoch": 0.6724003127443315,
+      "grad_norm": 0.4299210011959076,
+      "learning_rate": 8.395381440620674e-05,
+      "loss": 0.1223,
+      "step": 860
+    },
+    {
+      "epoch": 0.673182173573104,
+      "grad_norm": 0.39979061484336853,
+      "learning_rate": 8.39036928438887e-05,
+      "loss": 0.1436,
+      "step": 861
+    },
+    {
+      "epoch": 0.6739640344018765,
+      "grad_norm": 0.3123779892921448,
+      "learning_rate": 8.385350813699571e-05,
+      "loss": 0.1301,
+      "step": 862
+    },
+    {
+      "epoch": 0.674745895230649,
+      "grad_norm": 0.4949374496936798,
+      "learning_rate": 8.380326037899522e-05,
+      "loss": 0.172,
+      "step": 863
+    },
+    {
+      "epoch": 0.6755277560594214,
+      "grad_norm": 0.3271320164203644,
+      "learning_rate": 8.375294966347212e-05,
+      "loss": 0.1269,
+      "step": 864
+    },
+    {
+      "epoch": 0.6763096168881939,
+      "grad_norm": 0.33784911036491394,
+      "learning_rate": 8.370257608412857e-05,
+      "loss": 0.1374,
+      "step": 865
+    },
+    {
+      "epoch": 0.6770914777169664,
+      "grad_norm": 0.4051656424999237,
+      "learning_rate": 8.365213973478378e-05,
+      "loss": 0.1078,
+      "step": 866
+    },
+    {
+      "epoch": 0.6778733385457388,
+      "grad_norm": 0.40610331296920776,
+      "learning_rate": 8.360164070937389e-05,
+      "loss": 0.116,
+      "step": 867
+    },
+    {
+      "epoch": 0.6786551993745114,
+      "grad_norm": 0.35644418001174927,
+      "learning_rate": 8.355107910195175e-05,
+      "loss": 0.0986,
+      "step": 868
+    },
+    {
+      "epoch": 0.6794370602032838,
+      "grad_norm": 0.4006430208683014,
+      "learning_rate": 8.350045500668681e-05,
+      "loss": 0.1336,
+      "step": 869
+    },
+    {
+      "epoch": 0.6802189210320563,
+      "grad_norm": 0.36977195739746094,
+      "learning_rate": 8.344976851786487e-05,
+      "loss": 0.1294,
+      "step": 870
+    },
+    {
+      "epoch": 0.6810007818608288,
+      "grad_norm": 0.39970162510871887,
+      "learning_rate": 8.339901972988795e-05,
+      "loss": 0.1347,
+      "step": 871
+    },
+    {
+      "epoch": 0.6817826426896012,
+      "grad_norm": 0.36577853560447693,
+      "learning_rate": 8.334820873727406e-05,
+      "loss": 0.1194,
+      "step": 872
+    },
+    {
+      "epoch": 0.6825645035183737,
+      "grad_norm": 0.40536704659461975,
+      "learning_rate": 8.329733563465711e-05,
+      "loss": 0.1342,
+      "step": 873
+    },
+    {
+      "epoch": 0.6833463643471462,
+      "grad_norm": 0.37863144278526306,
+      "learning_rate": 8.324640051678668e-05,
+      "loss": 0.1283,
+      "step": 874
+    },
+    {
+      "epoch": 0.6841282251759186,
+      "grad_norm": 0.36349162459373474,
+      "learning_rate": 8.319540347852786e-05,
+      "loss": 0.138,
+      "step": 875
+    },
+    {
+      "epoch": 0.6849100860046912,
+      "grad_norm": 0.34184351563453674,
+      "learning_rate": 8.314434461486101e-05,
+      "loss": 0.0996,
+      "step": 876
+    },
+    {
+      "epoch": 0.6856919468334637,
+      "grad_norm": 0.4352666735649109,
+      "learning_rate": 8.30932240208817e-05,
+      "loss": 0.1489,
+      "step": 877
+    },
+    {
+      "epoch": 0.6864738076622361,
+      "grad_norm": 0.3389497399330139,
+      "learning_rate": 8.304204179180046e-05,
+      "loss": 0.1134,
+      "step": 878
+    },
+    {
+      "epoch": 0.6872556684910086,
+      "grad_norm": 0.48941102623939514,
+      "learning_rate": 8.299079802294258e-05,
+      "loss": 0.1533,
+      "step": 879
+    },
+    {
+      "epoch": 0.6880375293197811,
+      "grad_norm": 0.3835209608078003,
+      "learning_rate": 8.2939492809748e-05,
+      "loss": 0.1189,
+      "step": 880
+    },
+    {
+      "epoch": 0.6888193901485535,
+      "grad_norm": 0.3911098837852478,
+      "learning_rate": 8.288812624777109e-05,
+      "loss": 0.1267,
+      "step": 881
+    },
+    {
+      "epoch": 0.689601250977326,
+      "grad_norm": 0.3850928843021393,
+      "learning_rate": 8.283669843268047e-05,
+      "loss": 0.1221,
+      "step": 882
+    },
+    {
+      "epoch": 0.6903831118060985,
+      "grad_norm": 0.3491794764995575,
+      "learning_rate": 8.278520946025884e-05,
+      "loss": 0.1099,
+      "step": 883
+    },
+    {
+      "epoch": 0.691164972634871,
+      "grad_norm": 0.26835763454437256,
+      "learning_rate": 8.273365942640283e-05,
+      "loss": 0.1204,
+      "step": 884
+    },
+    {
+      "epoch": 0.6919468334636435,
+      "grad_norm": 0.40667450428009033,
+      "learning_rate": 8.268204842712278e-05,
+      "loss": 0.1487,
+      "step": 885
+    },
+    {
+      "epoch": 0.6927286942924159,
+      "grad_norm": 0.342351496219635,
+      "learning_rate": 8.263037655854254e-05,
+      "loss": 0.1268,
+      "step": 886
+    },
+    {
+      "epoch": 0.6935105551211884,
+      "grad_norm": 0.37458154559135437,
+      "learning_rate": 8.25786439168994e-05,
+      "loss": 0.108,
+      "step": 887
+    },
+    {
+      "epoch": 0.6942924159499609,
+      "grad_norm": 0.4606289863586426,
+      "learning_rate": 8.252685059854376e-05,
+      "loss": 0.1494,
+      "step": 888
+    },
+    {
+      "epoch": 0.6950742767787333,
+      "grad_norm": 0.35914990305900574,
+      "learning_rate": 8.24749966999391e-05,
+      "loss": 0.1142,
+      "step": 889
+    },
+    {
+      "epoch": 0.6958561376075059,
+      "grad_norm": 0.3088033199310303,
+      "learning_rate": 8.242308231766169e-05,
+      "loss": 0.107,
+      "step": 890
+    },
+    {
+      "epoch": 0.6966379984362784,
+      "grad_norm": 0.3469317853450775,
+      "learning_rate": 8.237110754840043e-05,
+      "loss": 0.105,
+      "step": 891
+    },
+    {
+      "epoch": 0.6974198592650508,
+      "grad_norm": 0.38797950744628906,
+      "learning_rate": 8.231907248895674e-05,
+      "loss": 0.1486,
+      "step": 892
+    },
+    {
+      "epoch": 0.6982017200938233,
+      "grad_norm": 0.23797141015529633,
+      "learning_rate": 8.22669772362443e-05,
+      "loss": 0.0972,
+      "step": 893
+    },
+    {
+      "epoch": 0.6989835809225958,
+      "grad_norm": 0.3245048522949219,
+      "learning_rate": 8.221482188728888e-05,
+      "loss": 0.1178,
+      "step": 894
+    },
+    {
+      "epoch": 0.6997654417513682,
+      "grad_norm": 0.30810755491256714,
+      "learning_rate": 8.216260653922823e-05,
+      "loss": 0.139,
+      "step": 895
+    },
+    {
+      "epoch": 0.7005473025801408,
+      "grad_norm": 0.3993600606918335,
+      "learning_rate": 8.211033128931176e-05,
+      "loss": 0.125,
+      "step": 896
+    },
+    {
+      "epoch": 0.7013291634089132,
+      "grad_norm": 0.33253660798072815,
+      "learning_rate": 8.205799623490055e-05,
+      "loss": 0.1001,
+      "step": 897
+    },
+    {
+      "epoch": 0.7021110242376857,
+      "grad_norm": 0.4122001528739929,
+      "learning_rate": 8.200560147346703e-05,
+      "loss": 0.1307,
+      "step": 898
+    },
+    {
+      "epoch": 0.7028928850664582,
+      "grad_norm": 0.3810223340988159,
+      "learning_rate": 8.195314710259475e-05,
+      "loss": 0.1634,
+      "step": 899
+    },
+    {
+      "epoch": 0.7036747458952306,
+      "grad_norm": 0.3160824775695801,
+      "learning_rate": 8.190063321997839e-05,
+      "loss": 0.1637,
+      "step": 900
+    },
+    {
+      "epoch": 0.7044566067240031,
+      "grad_norm": 0.3434925079345703,
+      "learning_rate": 8.184805992342342e-05,
+      "loss": 0.1414,
+      "step": 901
+    },
+    {
+      "epoch": 0.7052384675527756,
+      "grad_norm": 0.3496757447719574,
+      "learning_rate": 8.179542731084595e-05,
+      "loss": 0.1076,
+      "step": 902
+    },
+    {
+      "epoch": 0.706020328381548,
+      "grad_norm": 0.3597007989883423,
+      "learning_rate": 8.174273548027262e-05,
+      "loss": 0.1115,
+      "step": 903
+    },
+    {
+      "epoch": 0.7068021892103206,
+      "grad_norm": 0.3989991843700409,
+      "learning_rate": 8.168998452984031e-05,
+      "loss": 0.1019,
+      "step": 904
+    },
+    {
+      "epoch": 0.7075840500390931,
+      "grad_norm": 0.35645201802253723,
+      "learning_rate": 8.163717455779602e-05,
+      "loss": 0.1351,
+      "step": 905
+    },
+    {
+      "epoch": 0.7083659108678655,
+      "grad_norm": 0.31483447551727295,
+      "learning_rate": 8.158430566249671e-05,
+      "loss": 0.1029,
+      "step": 906
+    },
+    {
+      "epoch": 0.709147771696638,
+      "grad_norm": 0.3804417848587036,
+      "learning_rate": 8.153137794240903e-05,
+      "loss": 0.115,
+      "step": 907
+    },
+    {
+      "epoch": 0.7099296325254105,
+      "grad_norm": 0.33531495928764343,
+      "learning_rate": 8.147839149610922e-05,
+      "loss": 0.1044,
+      "step": 908
+    },
+    {
+      "epoch": 0.7107114933541829,
+      "grad_norm": 0.4194093942642212,
+      "learning_rate": 8.142534642228288e-05,
+      "loss": 0.1219,
+      "step": 909
+    },
+    {
+      "epoch": 0.7114933541829555,
+      "grad_norm": 0.33107230067253113,
+      "learning_rate": 8.137224281972485e-05,
+      "loss": 0.0973,
+      "step": 910
+    },
+    {
+      "epoch": 0.7122752150117279,
+      "grad_norm": 0.34085744619369507,
+      "learning_rate": 8.13190807873389e-05,
+      "loss": 0.1085,
+      "step": 911
+    },
+    {
+      "epoch": 0.7130570758405004,
+      "grad_norm": 0.43536314368247986,
+      "learning_rate": 8.126586042413769e-05,
+      "loss": 0.1466,
+      "step": 912
+    },
+    {
+      "epoch": 0.7138389366692729,
+      "grad_norm": 0.32383957505226135,
+      "learning_rate": 8.121258182924247e-05,
+      "loss": 0.1311,
+      "step": 913
+    },
+    {
+      "epoch": 0.7146207974980453,
+      "grad_norm": 0.3228246569633484,
+      "learning_rate": 8.1159245101883e-05,
+      "loss": 0.1244,
+      "step": 914
+    },
+    {
+      "epoch": 0.7154026583268178,
+      "grad_norm": 0.35498085618019104,
+      "learning_rate": 8.110585034139723e-05,
+      "loss": 0.1174,
+      "step": 915
+    },
+    {
+      "epoch": 0.7161845191555903,
+      "grad_norm": 0.3768846392631531,
+      "learning_rate": 8.105239764723131e-05,
+      "loss": 0.0963,
+      "step": 916
+    },
+    {
+      "epoch": 0.7169663799843627,
+      "grad_norm": 0.26356974244117737,
+      "learning_rate": 8.099888711893917e-05,
+      "loss": 0.1152,
+      "step": 917
+    },
+    {
+      "epoch": 0.7177482408131353,
+      "grad_norm": 0.3726540803909302,
+      "learning_rate": 8.094531885618252e-05,
+      "loss": 0.1068,
+      "step": 918
+    },
+    {
+      "epoch": 0.7185301016419078,
+      "grad_norm": 0.4119989573955536,
+      "learning_rate": 8.089169295873058e-05,
+      "loss": 0.1207,
+      "step": 919
+    },
+    {
+      "epoch": 0.7193119624706802,
+      "grad_norm": 0.3351033329963684,
+      "learning_rate": 8.083800952645994e-05,
+      "loss": 0.0899,
+      "step": 920
+    },
+    {
+      "epoch": 0.7200938232994527,
+      "grad_norm": 0.3526962399482727,
+      "learning_rate": 8.078426865935432e-05,
+      "loss": 0.1478,
+      "step": 921
+    },
+    {
+      "epoch": 0.7208756841282252,
+      "grad_norm": 0.3071853518486023,
+      "learning_rate": 8.07304704575044e-05,
+      "loss": 0.1107,
+      "step": 922
+    },
+    {
+      "epoch": 0.7216575449569976,
+      "grad_norm": 0.3210884630680084,
+      "learning_rate": 8.067661502110768e-05,
+      "loss": 0.0975,
+      "step": 923
+    },
+    {
+      "epoch": 0.7224394057857702,
+      "grad_norm": 0.34034693241119385,
+      "learning_rate": 8.06227024504682e-05,
+      "loss": 0.1235,
+      "step": 924
+    },
+    {
+      "epoch": 0.7232212666145426,
+      "grad_norm": 0.24603362381458282,
+      "learning_rate": 8.056873284599648e-05,
+      "loss": 0.0893,
+      "step": 925
+    },
+    {
+      "epoch": 0.7240031274433151,
+      "grad_norm": 0.5210705399513245,
+      "learning_rate": 8.051470630820925e-05,
+      "loss": 0.1745,
+      "step": 926
+    },
+    {
+      "epoch": 0.7247849882720876,
+      "grad_norm": 0.36895841360092163,
+      "learning_rate": 8.046062293772922e-05,
+      "loss": 0.1238,
+      "step": 927
+    },
+    {
+      "epoch": 0.72556684910086,
+      "grad_norm": 0.3690948784351349,
+      "learning_rate": 8.040648283528501e-05,
+      "loss": 0.1082,
+      "step": 928
+    },
+    {
+      "epoch": 0.7263487099296325,
+      "grad_norm": 0.3088683784008026,
+      "learning_rate": 8.035228610171085e-05,
+      "loss": 0.1122,
+      "step": 929
+    },
+    {
+      "epoch": 0.727130570758405,
+      "grad_norm": 0.316658616065979,
+      "learning_rate": 8.029803283794652e-05,
+      "loss": 0.1129,
+      "step": 930
+    },
+    {
+      "epoch": 0.7279124315871774,
+      "grad_norm": 0.40790560841560364,
+      "learning_rate": 8.024372314503701e-05,
+      "loss": 0.1349,
+      "step": 931
+    },
+    {
+      "epoch": 0.72869429241595,
+      "grad_norm": 0.46086615324020386,
+      "learning_rate": 8.018935712413242e-05,
+      "loss": 0.1845,
+      "step": 932
+    },
+    {
+      "epoch": 0.7294761532447225,
+      "grad_norm": 0.30622750520706177,
+      "learning_rate": 8.013493487648782e-05,
+      "loss": 0.1264,
+      "step": 933
+    },
+    {
+      "epoch": 0.7302580140734949,
+      "grad_norm": 0.42094358801841736,
+      "learning_rate": 8.008045650346294e-05,
+      "loss": 0.1089,
+      "step": 934
+    },
+    {
+      "epoch": 0.7310398749022674,
+      "grad_norm": 0.33936363458633423,
+      "learning_rate": 8.002592210652202e-05,
+      "loss": 0.1073,
+      "step": 935
+    },
+    {
+      "epoch": 0.7318217357310399,
+      "grad_norm": 0.3029439449310303,
+      "learning_rate": 7.997133178723374e-05,
+      "loss": 0.1162,
+      "step": 936
+    },
+    {
+      "epoch": 0.7326035965598123,
+      "grad_norm": 0.2870144844055176,
+      "learning_rate": 7.991668564727082e-05,
+      "loss": 0.1141,
+      "step": 937
+    },
+    {
+      "epoch": 0.7333854573885848,
+      "grad_norm": 0.3112180233001709,
+      "learning_rate": 7.986198378841004e-05,
+      "loss": 0.1262,
+      "step": 938
+    },
+    {
+      "epoch": 0.7341673182173573,
+      "grad_norm": 0.3226885497570038,
+      "learning_rate": 7.98072263125319e-05,
+      "loss": 0.1395,
+      "step": 939
+    },
+    {
+      "epoch": 0.7349491790461298,
+      "grad_norm": 0.47658324241638184,
+      "learning_rate": 7.975241332162048e-05,
+      "loss": 0.1653,
+      "step": 940
+    },
+    {
+      "epoch": 0.7357310398749023,
+      "grad_norm": 0.37937480211257935,
+      "learning_rate": 7.969754491776329e-05,
+      "loss": 0.1274,
+      "step": 941
+    },
+    {
+      "epoch": 0.7365129007036747,
+      "grad_norm": 0.35611534118652344,
+      "learning_rate": 7.964262120315103e-05,
+      "loss": 0.1426,
+      "step": 942
+    },
+    {
+      "epoch": 0.7372947615324472,
+      "grad_norm": 0.38592514395713806,
+      "learning_rate": 7.958764228007741e-05,
+      "loss": 0.1292,
+      "step": 943
+    },
+    {
+      "epoch": 0.7380766223612197,
+      "grad_norm": 0.4212297797203064,
+      "learning_rate": 7.953260825093898e-05,
+      "loss": 0.1283,
+      "step": 944
+    },
+    {
+      "epoch": 0.7388584831899921,
+      "grad_norm": 0.4761105477809906,
+      "learning_rate": 7.947751921823488e-05,
+      "loss": 0.1197,
+      "step": 945
+    },
+    {
+      "epoch": 0.7396403440187647,
+      "grad_norm": 0.351823627948761,
+      "learning_rate": 7.942237528456673e-05,
+      "loss": 0.1195,
+      "step": 946
+    },
+    {
+      "epoch": 0.7404222048475372,
+      "grad_norm": 0.4467895030975342,
+      "learning_rate": 7.936717655263841e-05,
+      "loss": 0.1341,
+      "step": 947
+    },
+    {
+      "epoch": 0.7412040656763096,
+      "grad_norm": 0.35307154059410095,
+      "learning_rate": 7.931192312525584e-05,
+      "loss": 0.1101,
+      "step": 948
+    },
+    {
+      "epoch": 0.7419859265050821,
+      "grad_norm": 0.32904255390167236,
+      "learning_rate": 7.925661510532681e-05,
+      "loss": 0.1324,
+      "step": 949
+    },
+    {
+      "epoch": 0.7427677873338546,
+      "grad_norm": 0.38342636823654175,
+      "learning_rate": 7.920125259586078e-05,
+      "loss": 0.1138,
+      "step": 950
+    },
+    {
+      "epoch": 0.743549648162627,
+      "grad_norm": 0.46515193581581116,
+      "learning_rate": 7.91458356999687e-05,
+      "loss": 0.1151,
+      "step": 951
+    },
+    {
+      "epoch": 0.7443315089913995,
+      "grad_norm": 0.3487924337387085,
+      "learning_rate": 7.909036452086285e-05,
+      "loss": 0.1415,
+      "step": 952
+    },
+    {
+      "epoch": 0.745113369820172,
+      "grad_norm": 0.3488902449607849,
+      "learning_rate": 7.903483916185654e-05,
+      "loss": 0.1466,
+      "step": 953
+    },
+    {
+      "epoch": 0.7458952306489445,
+      "grad_norm": 0.4646369516849518,
+      "learning_rate": 7.897925972636406e-05,
+      "loss": 0.1191,
+      "step": 954
+    },
+    {
+      "epoch": 0.746677091477717,
+      "grad_norm": 0.33017170429229736,
+      "learning_rate": 7.892362631790035e-05,
+      "loss": 0.1032,
+      "step": 955
+    },
+    {
+      "epoch": 0.7474589523064894,
+      "grad_norm": 0.39223426580429077,
+      "learning_rate": 7.886793904008094e-05,
+      "loss": 0.1137,
+      "step": 956
+    },
+    {
+      "epoch": 0.7482408131352619,
+      "grad_norm": 0.30325499176979065,
+      "learning_rate": 7.881219799662164e-05,
+      "loss": 0.1137,
+      "step": 957
+    },
+    {
+      "epoch": 0.7490226739640344,
+      "grad_norm": 0.4288765788078308,
+      "learning_rate": 7.87564032913384e-05,
+      "loss": 0.1458,
+      "step": 958
+    },
+    {
+      "epoch": 0.7498045347928068,
+      "grad_norm": 0.2882708013057709,
+      "learning_rate": 7.870055502814714e-05,
+      "loss": 0.1178,
+      "step": 959
+    },
+    {
+      "epoch": 0.7505863956215794,
+      "grad_norm": 0.36212223768234253,
+      "learning_rate": 7.864465331106349e-05,
+      "loss": 0.1219,
+      "step": 960
+    },
+    {
+      "epoch": 0.7513682564503519,
+      "grad_norm": 0.41355687379837036,
+      "learning_rate": 7.858869824420272e-05,
+      "loss": 0.1486,
+      "step": 961
+    },
+    {
+      "epoch": 0.7521501172791243,
+      "grad_norm": 0.36152246594429016,
+      "learning_rate": 7.853268993177936e-05,
+      "loss": 0.1297,
+      "step": 962
+    },
+    {
+      "epoch": 0.7529319781078968,
+      "grad_norm": 0.4029439389705658,
+      "learning_rate": 7.847662847810713e-05,
+      "loss": 0.1045,
+      "step": 963
+    },
+    {
+      "epoch": 0.7537138389366693,
+      "grad_norm": 0.29506799578666687,
+      "learning_rate": 7.84205139875988e-05,
+      "loss": 0.1051,
+      "step": 964
+    },
+    {
+      "epoch": 0.7544956997654417,
+      "grad_norm": 0.3722326457500458,
+      "learning_rate": 7.836434656476583e-05,
+      "loss": 0.1122,
+      "step": 965
+    },
+    {
+      "epoch": 0.7552775605942142,
+      "grad_norm": 0.326231449842453,
+      "learning_rate": 7.830812631421833e-05,
+      "loss": 0.1399,
+      "step": 966
+    },
+    {
+      "epoch": 0.7560594214229867,
+      "grad_norm": 0.3228454291820526,
+      "learning_rate": 7.825185334066475e-05,
+      "loss": 0.0995,
+      "step": 967
+    },
+    {
+      "epoch": 0.7568412822517592,
+      "grad_norm": 0.4981108605861664,
+      "learning_rate": 7.819552774891174e-05,
+      "loss": 0.1363,
+      "step": 968
+    },
+    {
+      "epoch": 0.7576231430805317,
+      "grad_norm": 0.27120697498321533,
+      "learning_rate": 7.813914964386401e-05,
+      "loss": 0.0893,
+      "step": 969
+    },
+    {
+      "epoch": 0.7584050039093041,
+      "grad_norm": 0.31169626116752625,
+      "learning_rate": 7.808271913052398e-05,
+      "loss": 0.1364,
+      "step": 970
+    },
+    {
+      "epoch": 0.7591868647380766,
+      "grad_norm": 0.3545890152454376,
+      "learning_rate": 7.802623631399176e-05,
+      "loss": 0.1163,
+      "step": 971
+    },
+    {
+      "epoch": 0.7599687255668491,
+      "grad_norm": 0.3600348234176636,
+      "learning_rate": 7.796970129946484e-05,
+      "loss": 0.1292,
+      "step": 972
+    },
+    {
+      "epoch": 0.7607505863956215,
+      "grad_norm": 0.3808040916919708,
+      "learning_rate": 7.791311419223791e-05,
+      "loss": 0.1168,
+      "step": 973
+    },
+    {
+      "epoch": 0.7615324472243941,
+      "grad_norm": 0.44746145606040955,
+      "learning_rate": 7.785647509770272e-05,
+      "loss": 0.1267,
+      "step": 974
+    },
+    {
+      "epoch": 0.7623143080531666,
+      "grad_norm": 0.35534608364105225,
+      "learning_rate": 7.779978412134783e-05,
+      "loss": 0.1183,
+      "step": 975
+    },
+    {
+      "epoch": 0.763096168881939,
+      "grad_norm": 0.34301894903182983,
+      "learning_rate": 7.77430413687584e-05,
+      "loss": 0.1046,
+      "step": 976
+    },
+    {
+      "epoch": 0.7638780297107115,
+      "grad_norm": 0.4878581166267395,
+      "learning_rate": 7.768624694561604e-05,
+      "loss": 0.1592,
+      "step": 977
+    },
+    {
+      "epoch": 0.764659890539484,
+      "grad_norm": 0.43463602662086487,
+      "learning_rate": 7.762940095769861e-05,
+      "loss": 0.1213,
+      "step": 978
+    },
+    {
+      "epoch": 0.7654417513682564,
+      "grad_norm": 0.36322781443595886,
+      "learning_rate": 7.757250351088004e-05,
+      "loss": 0.1183,
+      "step": 979
+    },
+    {
+      "epoch": 0.766223612197029,
+      "grad_norm": 0.28254207968711853,
+      "learning_rate": 7.751555471113e-05,
+      "loss": 0.1063,
+      "step": 980
+    },
+    {
+      "epoch": 0.7670054730258014,
+      "grad_norm": 0.3453584313392639,
+      "learning_rate": 7.745855466451385e-05,
+      "loss": 0.1257,
+      "step": 981
+    },
+    {
+      "epoch": 0.7677873338545739,
+      "grad_norm": 0.35417473316192627,
+      "learning_rate": 7.740150347719246e-05,
+      "loss": 0.1084,
+      "step": 982
+    },
+    {
+      "epoch": 0.7685691946833464,
+      "grad_norm": 0.2651992440223694,
+      "learning_rate": 7.734440125542186e-05,
+      "loss": 0.0937,
+      "step": 983
+    },
+    {
+      "epoch": 0.7693510555121188,
+      "grad_norm": 0.41208094358444214,
+      "learning_rate": 7.728724810555317e-05,
+      "loss": 0.1358,
+      "step": 984
+    },
+    {
+      "epoch": 0.7701329163408913,
+      "grad_norm": 0.3765069246292114,
+      "learning_rate": 7.723004413403238e-05,
+      "loss": 0.1067,
+      "step": 985
+    },
+    {
+      "epoch": 0.7709147771696638,
+      "grad_norm": 0.3152915835380554,
+      "learning_rate": 7.717278944740007e-05,
+      "loss": 0.1083,
+      "step": 986
+    },
+    {
+      "epoch": 0.7716966379984362,
+      "grad_norm": 0.38685888051986694,
+      "learning_rate": 7.711548415229135e-05,
+      "loss": 0.1395,
+      "step": 987
+    },
+    {
+      "epoch": 0.7724784988272088,
+      "grad_norm": 0.3208830654621124,
+      "learning_rate": 7.705812835543553e-05,
+      "loss": 0.1318,
+      "step": 988
+    },
+    {
+      "epoch": 0.7732603596559813,
+      "grad_norm": 0.4066202938556671,
+      "learning_rate": 7.700072216365602e-05,
+      "loss": 0.1192,
+      "step": 989
+    },
+    {
+      "epoch": 0.7740422204847537,
+      "grad_norm": 0.3345940113067627,
+      "learning_rate": 7.694326568387007e-05,
+      "loss": 0.1169,
+      "step": 990
+    },
+    {
+      "epoch": 0.7748240813135262,
+      "grad_norm": 0.3916303813457489,
+      "learning_rate": 7.688575902308854e-05,
+      "loss": 0.1185,
+      "step": 991
+    },
+    {
+      "epoch": 0.7756059421422987,
+      "grad_norm": 0.39679235219955444,
+      "learning_rate": 7.682820228841588e-05,
+      "loss": 0.1277,
+      "step": 992
+    },
+    {
+      "epoch": 0.7763878029710711,
+      "grad_norm": 0.3425503075122833,
+      "learning_rate": 7.677059558704965e-05,
+      "loss": 0.1219,
+      "step": 993
+    },
+    {
+      "epoch": 0.7771696637998436,
+      "grad_norm": 0.2682630121707916,
+      "learning_rate": 7.671293902628058e-05,
+      "loss": 0.12,
+      "step": 994
+    },
+    {
+      "epoch": 0.777951524628616,
+      "grad_norm": 0.4910065233707428,
+      "learning_rate": 7.665523271349221e-05,
+      "loss": 0.1214,
+      "step": 995
+    },
+    {
+      "epoch": 0.7787333854573886,
+      "grad_norm": 0.3416655361652374,
+      "learning_rate": 7.659747675616075e-05,
+      "loss": 0.1107,
+      "step": 996
+    },
+    {
+      "epoch": 0.7795152462861611,
+      "grad_norm": 0.47088661789894104,
+      "learning_rate": 7.65396712618549e-05,
+      "loss": 0.1483,
+      "step": 997
+    },
+    {
+      "epoch": 0.7802971071149335,
+      "grad_norm": 0.36356088519096375,
+      "learning_rate": 7.648181633823559e-05,
+      "loss": 0.1473,
+      "step": 998
+    },
+    {
+      "epoch": 0.781078967943706,
+      "grad_norm": 0.4040223956108093,
+      "learning_rate": 7.642391209305581e-05,
+      "loss": 0.1234,
+      "step": 999
+    },
+    {
+      "epoch": 0.7818608287724785,
+      "grad_norm": 0.27312490344047546,
+      "learning_rate": 7.636595863416041e-05,
+      "loss": 0.0953,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7818608287724785,
+      "eval_loss": 0.13331574201583862,
+      "eval_runtime": 13.3827,
+      "eval_samples_per_second": 3.886,
+      "eval_steps_per_second": 0.971,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7826426896012509,
+      "grad_norm": 0.4011226296424866,
+      "learning_rate": 7.630795606948592e-05,
+      "loss": 0.1279,
+      "step": 1001
+    },
+    {
+      "epoch": 0.7834245504300235,
+      "grad_norm": 0.2801114320755005,
+      "learning_rate": 7.624990450706034e-05,
+      "loss": 0.0881,
+      "step": 1002
+    },
+    {
+      "epoch": 0.784206411258796,
+      "grad_norm": 0.3331213891506195,
+      "learning_rate": 7.619180405500284e-05,
+      "loss": 0.1109,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7849882720875684,
+      "grad_norm": 0.27233004570007324,
+      "learning_rate": 7.613365482152374e-05,
+      "loss": 0.1254,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7857701329163409,
+      "grad_norm": 0.3889213502407074,
+      "learning_rate": 7.607545691492421e-05,
+      "loss": 0.1247,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7865519937451134,
+      "grad_norm": 0.31572502851486206,
+      "learning_rate": 7.601721044359601e-05,
+      "loss": 0.0833,
+      "step": 1006
+    },
+    {
+      "epoch": 0.7873338545738858,
+      "grad_norm": 0.3308155834674835,
+      "learning_rate": 7.595891551602139e-05,
+      "loss": 0.0946,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7881157154026583,
+      "grad_norm": 0.3063320219516754,
+      "learning_rate": 7.590057224077285e-05,
+      "loss": 0.1078,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7888975762314308,
+      "grad_norm": 0.38767197728157043,
+      "learning_rate": 7.584218072651291e-05,
+      "loss": 0.1363,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7896794370602033,
+      "grad_norm": 0.40377962589263916,
+      "learning_rate": 7.578374108199396e-05,
+      "loss": 0.1376,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7904612978889758,
+      "grad_norm": 0.3951176106929779,
+      "learning_rate": 7.572525341605805e-05,
+      "loss": 0.1108,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7912431587177482,
+      "grad_norm": 0.28525015711784363,
+      "learning_rate": 7.56667178376366e-05,
+      "loss": 0.1247,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7920250195465207,
+      "grad_norm": 0.36067599058151245,
+      "learning_rate": 7.560813445575032e-05,
+      "loss": 0.1015,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7928068803752932,
+      "grad_norm": 0.31828776001930237,
+      "learning_rate": 7.554950337950895e-05,
+      "loss": 0.1138,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7935887412040656,
+      "grad_norm": 0.36967208981513977,
+      "learning_rate": 7.549082471811105e-05,
+      "loss": 0.1488,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7943706020328382,
+      "grad_norm": 0.3618142604827881,
+      "learning_rate": 7.54320985808438e-05,
+      "loss": 0.1043,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7951524628616107,
+      "grad_norm": 0.3296690583229065,
+      "learning_rate": 7.53733250770828e-05,
+      "loss": 0.1256,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7959343236903831,
+      "grad_norm": 0.4782399833202362,
+      "learning_rate": 7.531450431629188e-05,
+      "loss": 0.1331,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7967161845191556,
+      "grad_norm": 0.3915999233722687,
+      "learning_rate": 7.525563640802286e-05,
+      "loss": 0.1544,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7974980453479281,
+      "grad_norm": 0.4172796905040741,
+      "learning_rate": 7.519672146191543e-05,
+      "loss": 0.1196,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7982799061767005,
+      "grad_norm": 0.4924179017543793,
+      "learning_rate": 7.513775958769683e-05,
+      "loss": 0.1326,
+      "step": 1021
+    },
+    {
+      "epoch": 0.799061767005473,
+      "grad_norm": 0.32305261492729187,
+      "learning_rate": 7.50787508951817e-05,
+      "loss": 0.1138,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7998436278342455,
+      "grad_norm": 0.2907930314540863,
+      "learning_rate": 7.501969549427195e-05,
+      "loss": 0.1189,
+      "step": 1023
+    },
+    {
+      "epoch": 0.800625488663018,
+      "grad_norm": 0.4041495621204376,
+      "learning_rate": 7.496059349495636e-05,
+      "loss": 0.1088,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8014073494917905,
+      "grad_norm": 0.342549592256546,
+      "learning_rate": 7.49014450073106e-05,
+      "loss": 0.1229,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8021892103205629,
+      "grad_norm": 0.3912782073020935,
+      "learning_rate": 7.484225014149691e-05,
+      "loss": 0.122,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8029710711493354,
+      "grad_norm": 0.38920527696609497,
+      "learning_rate": 7.478300900776387e-05,
+      "loss": 0.1185,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8037529319781079,
+      "grad_norm": 0.38123801350593567,
+      "learning_rate": 7.472372171644626e-05,
+      "loss": 0.151,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8045347928068803,
+      "grad_norm": 0.2902073860168457,
+      "learning_rate": 7.46643883779648e-05,
+      "loss": 0.1127,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8053166536356529,
+      "grad_norm": 0.3496885299682617,
+      "learning_rate": 7.460500910282602e-05,
+      "loss": 0.1319,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8060985144644254,
+      "grad_norm": 0.32722875475883484,
+      "learning_rate": 7.454558400162195e-05,
+      "loss": 0.0927,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8068803752931978,
+      "grad_norm": 0.3707050383090973,
+      "learning_rate": 7.448611318503001e-05,
+      "loss": 0.1008,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8076622361219703,
+      "grad_norm": 0.3069707453250885,
+      "learning_rate": 7.442659676381275e-05,
+      "loss": 0.1218,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8084440969507428,
+      "grad_norm": 0.4122169613838196,
+      "learning_rate": 7.436703484881761e-05,
+      "loss": 0.13,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8092259577795152,
+      "grad_norm": 0.39322778582572937,
+      "learning_rate": 7.430742755097689e-05,
+      "loss": 0.1239,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8100078186082877,
+      "grad_norm": 0.35709503293037415,
+      "learning_rate": 7.424777498130727e-05,
+      "loss": 0.116,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8107896794370602,
+      "grad_norm": 0.3031849265098572,
+      "learning_rate": 7.418807725090983e-05,
+      "loss": 0.098,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8115715402658327,
+      "grad_norm": 0.30318596959114075,
+      "learning_rate": 7.412833447096973e-05,
+      "loss": 0.1031,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8123534010946052,
+      "grad_norm": 0.24984782934188843,
+      "learning_rate": 7.406854675275605e-05,
+      "loss": 0.1119,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8131352619233776,
+      "grad_norm": 0.3230637013912201,
+      "learning_rate": 7.400871420762154e-05,
+      "loss": 0.1415,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8139171227521501,
+      "grad_norm": 0.40312686562538147,
+      "learning_rate": 7.39488369470025e-05,
+      "loss": 0.1162,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8146989835809226,
+      "grad_norm": 0.3000973165035248,
+      "learning_rate": 7.388891508241842e-05,
+      "loss": 0.1341,
+      "step": 1042
+    },
+    {
+      "epoch": 0.815480844409695,
+      "grad_norm": 0.3631311357021332,
+      "learning_rate": 7.382894872547195e-05,
+      "loss": 0.1275,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8162627052384676,
+      "grad_norm": 0.2743859887123108,
+      "learning_rate": 7.376893798784853e-05,
+      "loss": 0.0804,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8170445660672401,
+      "grad_norm": 0.3590327203273773,
+      "learning_rate": 7.370888298131633e-05,
+      "loss": 0.1,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8178264268960125,
+      "grad_norm": 0.42224350571632385,
+      "learning_rate": 7.36487838177259e-05,
+      "loss": 0.134,
+      "step": 1046
+    },
+    {
+      "epoch": 0.818608287724785,
+      "grad_norm": 0.3597356975078583,
+      "learning_rate": 7.35886406090101e-05,
+      "loss": 0.117,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8193901485535575,
+      "grad_norm": 0.30071374773979187,
+      "learning_rate": 7.352845346718378e-05,
+      "loss": 0.103,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8201720093823299,
+      "grad_norm": 0.33568722009658813,
+      "learning_rate": 7.34682225043436e-05,
+      "loss": 0.1102,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8209538702111024,
+      "grad_norm": 0.3519073724746704,
+      "learning_rate": 7.340794783266787e-05,
+      "loss": 0.1545,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8217357310398749,
+      "grad_norm": 0.3975318968296051,
+      "learning_rate": 7.334762956441632e-05,
+      "loss": 0.1136,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8225175918686474,
+      "grad_norm": 0.3154171109199524,
+      "learning_rate": 7.328726781192981e-05,
+      "loss": 0.1,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8232994526974199,
+      "grad_norm": 0.3593648076057434,
+      "learning_rate": 7.322686268763026e-05,
+      "loss": 0.1115,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8240813135261923,
+      "grad_norm": 0.40494292974472046,
+      "learning_rate": 7.31664143040203e-05,
+      "loss": 0.1296,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8248631743549648,
+      "grad_norm": 0.3605891764163971,
+      "learning_rate": 7.310592277368322e-05,
+      "loss": 0.1316,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8256450351837373,
+      "grad_norm": 0.389142245054245,
+      "learning_rate": 7.304538820928257e-05,
+      "loss": 0.0992,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8264268960125097,
+      "grad_norm": 0.3127676844596863,
+      "learning_rate": 7.298481072356214e-05,
+      "loss": 0.0949,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8272087568412823,
+      "grad_norm": 0.2602942883968353,
+      "learning_rate": 7.292419042934556e-05,
+      "loss": 0.1101,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8279906176700548,
+      "grad_norm": 0.40910831093788147,
+      "learning_rate": 7.28635274395363e-05,
+      "loss": 0.1544,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8287724784988272,
+      "grad_norm": 0.38308101892471313,
+      "learning_rate": 7.280282186711726e-05,
+      "loss": 0.1241,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8295543393275997,
+      "grad_norm": 0.2836664021015167,
+      "learning_rate": 7.274207382515071e-05,
+      "loss": 0.1309,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8303362001563722,
+      "grad_norm": 0.3716040551662445,
+      "learning_rate": 7.268128342677796e-05,
+      "loss": 0.1232,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8311180609851446,
+      "grad_norm": 0.4505487382411957,
+      "learning_rate": 7.262045078521924e-05,
+      "loss": 0.1465,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8318999218139171,
+      "grad_norm": 0.45818236470222473,
+      "learning_rate": 7.25595760137735e-05,
+      "loss": 0.1318,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8326817826426895,
+      "grad_norm": 0.33937686681747437,
+      "learning_rate": 7.249865922581807e-05,
+      "loss": 0.1104,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8334636434714621,
+      "grad_norm": 0.5324069857597351,
+      "learning_rate": 7.243770053480859e-05,
+      "loss": 0.1084,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8342455043002346,
+      "grad_norm": 0.3447381854057312,
+      "learning_rate": 7.237670005427872e-05,
+      "loss": 0.0932,
+      "step": 1067
+    },
+    {
+      "epoch": 0.835027365129007,
+      "grad_norm": 0.445389062166214,
+      "learning_rate": 7.231565789783997e-05,
+      "loss": 0.1026,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8358092259577795,
+      "grad_norm": 0.33196380734443665,
+      "learning_rate": 7.225457417918144e-05,
+      "loss": 0.1102,
+      "step": 1069
+    },
+    {
+      "epoch": 0.836591086786552,
+      "grad_norm": 0.37734511494636536,
+      "learning_rate": 7.21934490120697e-05,
+      "loss": 0.0999,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8373729476153244,
+      "grad_norm": 0.4131399095058441,
+      "learning_rate": 7.213228251034844e-05,
+      "loss": 0.1125,
+      "step": 1071
+    },
+    {
+      "epoch": 0.838154808444097,
+      "grad_norm": 0.4269322454929352,
+      "learning_rate": 7.207107478793838e-05,
+      "loss": 0.1709,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8389366692728695,
+      "grad_norm": 0.4380016624927521,
+      "learning_rate": 7.2009825958837e-05,
+      "loss": 0.1252,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8397185301016419,
+      "grad_norm": 0.3470591902732849,
+      "learning_rate": 7.194853613711835e-05,
+      "loss": 0.1178,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8405003909304144,
+      "grad_norm": 0.4182756841182709,
+      "learning_rate": 7.188720543693283e-05,
+      "loss": 0.1041,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8412822517591869,
+      "grad_norm": 0.41240689158439636,
+      "learning_rate": 7.182583397250692e-05,
+      "loss": 0.1316,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8420641125879593,
+      "grad_norm": 0.42284348607063293,
+      "learning_rate": 7.176442185814312e-05,
+      "loss": 0.114,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8428459734167318,
+      "grad_norm": 0.2716412842273712,
+      "learning_rate": 7.170296920821953e-05,
+      "loss": 0.1044,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8436278342455042,
+      "grad_norm": 0.42033228278160095,
+      "learning_rate": 7.164147613718986e-05,
+      "loss": 0.1388,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8444096950742768,
+      "grad_norm": 0.4796225130558014,
+      "learning_rate": 7.157994275958302e-05,
+      "loss": 0.1491,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8451915559030493,
+      "grad_norm": 0.26454976201057434,
+      "learning_rate": 7.151836919000299e-05,
+      "loss": 0.0708,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8459734167318217,
+      "grad_norm": 0.3553334176540375,
+      "learning_rate": 7.145675554312866e-05,
+      "loss": 0.1236,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8467552775605942,
+      "grad_norm": 0.41896483302116394,
+      "learning_rate": 7.139510193371352e-05,
+      "loss": 0.1523,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8475371383893667,
+      "grad_norm": 0.3911309838294983,
+      "learning_rate": 7.133340847658553e-05,
+      "loss": 0.1123,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8483189992181391,
+      "grad_norm": 0.3871746361255646,
+      "learning_rate": 7.127167528664682e-05,
+      "loss": 0.1448,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8491008600469117,
+      "grad_norm": 0.38101163506507874,
+      "learning_rate": 7.120990247887351e-05,
+      "loss": 0.1435,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8498827208756842,
+      "grad_norm": 0.3440212309360504,
+      "learning_rate": 7.114809016831558e-05,
+      "loss": 0.111,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8506645817044566,
+      "grad_norm": 0.34465304017066956,
+      "learning_rate": 7.108623847009651e-05,
+      "loss": 0.1375,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8514464425332291,
+      "grad_norm": 0.3100002706050873,
+      "learning_rate": 7.10243474994132e-05,
+      "loss": 0.0996,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8522283033620016,
+      "grad_norm": 0.2955577075481415,
+      "learning_rate": 7.096241737153562e-05,
+      "loss": 0.0878,
+      "step": 1090
+    },
+    {
+      "epoch": 0.853010164190774,
+      "grad_norm": 0.42580050230026245,
+      "learning_rate": 7.090044820180673e-05,
+      "loss": 0.0961,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8537920250195465,
+      "grad_norm": 0.34347400069236755,
+      "learning_rate": 7.083844010564219e-05,
+      "loss": 0.1231,
+      "step": 1092
+    },
+    {
+      "epoch": 0.854573885848319,
+      "grad_norm": 0.3110819160938263,
+      "learning_rate": 7.077639319853013e-05,
+      "loss": 0.0847,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8553557466770915,
+      "grad_norm": 0.4496898353099823,
+      "learning_rate": 7.071430759603103e-05,
+      "loss": 0.1247,
+      "step": 1094
+    },
+    {
+      "epoch": 0.856137607505864,
+      "grad_norm": 0.3124215602874756,
+      "learning_rate": 7.065218341377734e-05,
+      "loss": 0.0917,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8569194683346364,
+      "grad_norm": 0.3862299621105194,
+      "learning_rate": 7.059002076747348e-05,
+      "loss": 0.103,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8577013291634089,
+      "grad_norm": 0.36825454235076904,
+      "learning_rate": 7.05278197728954e-05,
+      "loss": 0.0993,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8584831899921814,
+      "grad_norm": 0.4220404028892517,
+      "learning_rate": 7.046558054589055e-05,
+      "loss": 0.1358,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8592650508209538,
+      "grad_norm": 0.39067772030830383,
+      "learning_rate": 7.040330320237752e-05,
+      "loss": 0.1061,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8600469116497264,
+      "grad_norm": 0.27759233117103577,
+      "learning_rate": 7.034098785834598e-05,
+      "loss": 0.1062,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8608287724784989,
+      "grad_norm": 0.3399001657962799,
+      "learning_rate": 7.027863462985628e-05,
+      "loss": 0.1167,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8616106333072713,
+      "grad_norm": 0.43628373742103577,
+      "learning_rate": 7.02162436330394e-05,
+      "loss": 0.1563,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8623924941360438,
+      "grad_norm": 0.5306029319763184,
+      "learning_rate": 7.015381498409661e-05,
+      "loss": 0.1653,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8631743549648163,
+      "grad_norm": 0.3642573356628418,
+      "learning_rate": 7.009134879929936e-05,
+      "loss": 0.1072,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8639562157935887,
+      "grad_norm": 0.3719395399093628,
+      "learning_rate": 7.002884519498895e-05,
+      "loss": 0.1017,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8647380766223612,
+      "grad_norm": 0.4951043426990509,
+      "learning_rate": 6.99663042875764e-05,
+      "loss": 0.1516,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8655199374511336,
+      "grad_norm": 0.4145095646381378,
+      "learning_rate": 6.990372619354224e-05,
+      "loss": 0.1188,
+      "step": 1107
+    },
+    {
+      "epoch": 0.8663017982799062,
+      "grad_norm": 0.31760677695274353,
+      "learning_rate": 6.98411110294362e-05,
+      "loss": 0.0957,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8670836591086787,
+      "grad_norm": 0.48377615213394165,
+      "learning_rate": 6.977845891187708e-05,
+      "loss": 0.142,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8678655199374511,
+      "grad_norm": 0.3468191921710968,
+      "learning_rate": 6.971576995755253e-05,
+      "loss": 0.0911,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8686473807662236,
+      "grad_norm": 0.38117656111717224,
+      "learning_rate": 6.965304428321874e-05,
+      "loss": 0.1185,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8694292415949961,
+      "grad_norm": 0.4549887478351593,
+      "learning_rate": 6.959028200570034e-05,
+      "loss": 0.1416,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8702111024237685,
+      "grad_norm": 0.3793928027153015,
+      "learning_rate": 6.952748324189016e-05,
+      "loss": 0.1584,
+      "step": 1113
+    },
+    {
+      "epoch": 0.870992963252541,
+      "grad_norm": 0.430199533700943,
+      "learning_rate": 6.94646481087489e-05,
+      "loss": 0.1863,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8717748240813136,
+      "grad_norm": 0.37007904052734375,
+      "learning_rate": 6.940177672330508e-05,
+      "loss": 0.1413,
+      "step": 1115
+    },
+    {
+      "epoch": 0.872556684910086,
+      "grad_norm": 0.29991209506988525,
+      "learning_rate": 6.933886920265468e-05,
+      "loss": 0.1143,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8733385457388585,
+      "grad_norm": 0.36572265625,
+      "learning_rate": 6.9275925663961e-05,
+      "loss": 0.1286,
+      "step": 1117
+    },
+    {
+      "epoch": 0.874120406567631,
+      "grad_norm": 0.43926042318344116,
+      "learning_rate": 6.921294622445444e-05,
+      "loss": 0.1329,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8749022673964034,
+      "grad_norm": 0.35618895292282104,
+      "learning_rate": 6.914993100143224e-05,
+      "loss": 0.1193,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8756841282251759,
+      "grad_norm": 0.3571595847606659,
+      "learning_rate": 6.908688011225831e-05,
+      "loss": 0.105,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8764659890539483,
+      "grad_norm": 0.40551918745040894,
+      "learning_rate": 6.902379367436296e-05,
+      "loss": 0.1064,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8772478498827209,
+      "grad_norm": 0.33147695660591125,
+      "learning_rate": 6.896067180524273e-05,
+      "loss": 0.091,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8780297107114934,
+      "grad_norm": 0.3313888609409332,
+      "learning_rate": 6.889751462246013e-05,
+      "loss": 0.1345,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8788115715402658,
+      "grad_norm": 0.4676271378993988,
+      "learning_rate": 6.883432224364346e-05,
+      "loss": 0.1433,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8795934323690383,
+      "grad_norm": 0.3496025502681732,
+      "learning_rate": 6.877109478648656e-05,
+      "loss": 0.1565,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8803752931978108,
+      "grad_norm": 0.3045591115951538,
+      "learning_rate": 6.870783236874861e-05,
+      "loss": 0.1122,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8811571540265832,
+      "grad_norm": 0.39115655422210693,
+      "learning_rate": 6.864453510825388e-05,
+      "loss": 0.1323,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8819390148553558,
+      "grad_norm": 0.3827655613422394,
+      "learning_rate": 6.85812031228916e-05,
+      "loss": 0.1326,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8827208756841283,
+      "grad_norm": 0.4689268469810486,
+      "learning_rate": 6.851783653061555e-05,
+      "loss": 0.1199,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8835027365129007,
+      "grad_norm": 0.36318734288215637,
+      "learning_rate": 6.845443544944411e-05,
+      "loss": 0.0897,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8842845973416732,
+      "grad_norm": 0.3593188524246216,
+      "learning_rate": 6.83909999974598e-05,
+      "loss": 0.118,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8850664581704457,
+      "grad_norm": 0.2524106502532959,
+      "learning_rate": 6.832753029280913e-05,
+      "loss": 0.0675,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8858483189992181,
+      "grad_norm": 0.3099701404571533,
+      "learning_rate": 6.826402645370256e-05,
+      "loss": 0.1163,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8866301798279906,
+      "grad_norm": 0.4418666958808899,
+      "learning_rate": 6.820048859841393e-05,
+      "loss": 0.1369,
+      "step": 1134
+    },
+    {
+      "epoch": 0.887412040656763,
+      "grad_norm": 0.35631266236305237,
+      "learning_rate": 6.813691684528054e-05,
+      "loss": 0.0867,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8881939014855356,
+      "grad_norm": 0.40591201186180115,
+      "learning_rate": 6.807331131270283e-05,
+      "loss": 0.124,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8889757623143081,
+      "grad_norm": 0.4229952394962311,
+      "learning_rate": 6.800967211914409e-05,
+      "loss": 0.1235,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8897576231430805,
+      "grad_norm": 0.4261278510093689,
+      "learning_rate": 6.794599938313041e-05,
+      "loss": 0.0969,
+      "step": 1138
+    },
+    {
+      "epoch": 0.890539483971853,
+      "grad_norm": 0.35794153809547424,
+      "learning_rate": 6.788229322325022e-05,
+      "loss": 0.1119,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8913213448006255,
+      "grad_norm": 0.32628294825553894,
+      "learning_rate": 6.781855375815427e-05,
+      "loss": 0.094,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8921032056293979,
+      "grad_norm": 0.4180498719215393,
+      "learning_rate": 6.775478110655535e-05,
+      "loss": 0.1502,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8928850664581705,
+      "grad_norm": 0.28836068511009216,
+      "learning_rate": 6.769097538722806e-05,
+      "loss": 0.1117,
+      "step": 1142
+    },
+    {
+      "epoch": 0.893666927286943,
+      "grad_norm": 0.3085721731185913,
+      "learning_rate": 6.762713671900853e-05,
+      "loss": 0.0979,
+      "step": 1143
+    },
+    {
+      "epoch": 0.8944487881157154,
+      "grad_norm": 0.3328137993812561,
+      "learning_rate": 6.756326522079433e-05,
+      "loss": 0.1306,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8952306489444879,
+      "grad_norm": 0.4145255386829376,
+      "learning_rate": 6.74993610115441e-05,
+      "loss": 0.1007,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8960125097732604,
+      "grad_norm": 0.31838345527648926,
+      "learning_rate": 6.743542421027751e-05,
+      "loss": 0.1043,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8967943706020328,
+      "grad_norm": 0.3162778317928314,
+      "learning_rate": 6.737145493607482e-05,
+      "loss": 0.0856,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8975762314308053,
+      "grad_norm": 0.44549447298049927,
+      "learning_rate": 6.730745330807684e-05,
+      "loss": 0.1045,
+      "step": 1148
+    },
+    {
+      "epoch": 0.8983580922595777,
+      "grad_norm": 0.3551749885082245,
+      "learning_rate": 6.724341944548459e-05,
+      "loss": 0.1258,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8991399530883503,
+      "grad_norm": 0.3578115999698639,
+      "learning_rate": 6.717935346755915e-05,
+      "loss": 0.0974,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8999218139171228,
+      "grad_norm": 0.3265388607978821,
+      "learning_rate": 6.711525549362144e-05,
+      "loss": 0.0956,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9007036747458952,
+      "grad_norm": 0.30331680178642273,
+      "learning_rate": 6.705112564305194e-05,
+      "loss": 0.1004,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9014855355746677,
+      "grad_norm": 0.3490830063819885,
+      "learning_rate": 6.698696403529049e-05,
+      "loss": 0.1187,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9022673964034402,
+      "grad_norm": 0.2992035448551178,
+      "learning_rate": 6.69227707898361e-05,
+      "loss": 0.0808,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9030492572322126,
+      "grad_norm": 0.4635206460952759,
+      "learning_rate": 6.685854602624668e-05,
+      "loss": 0.1041,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9038311180609852,
+      "grad_norm": 0.3826647102832794,
+      "learning_rate": 6.679428986413889e-05,
+      "loss": 0.0935,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9046129788897577,
+      "grad_norm": 0.4464735984802246,
+      "learning_rate": 6.673000242318782e-05,
+      "loss": 0.1552,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9053948397185301,
+      "grad_norm": 0.4731329083442688,
+      "learning_rate": 6.666568382312684e-05,
+      "loss": 0.119,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9061767005473026,
+      "grad_norm": 0.4805479645729065,
+      "learning_rate": 6.660133418374732e-05,
+      "loss": 0.146,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9069585613760751,
+      "grad_norm": 0.34081944823265076,
+      "learning_rate": 6.653695362489846e-05,
+      "loss": 0.1157,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9077404222048475,
+      "grad_norm": 0.38758766651153564,
+      "learning_rate": 6.647254226648711e-05,
+      "loss": 0.1181,
+      "step": 1161
+    },
+    {
+      "epoch": 0.90852228303362,
+      "grad_norm": 0.29969069361686707,
+      "learning_rate": 6.640810022847738e-05,
+      "loss": 0.1082,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9093041438623924,
+      "grad_norm": 0.3649740219116211,
+      "learning_rate": 6.634362763089056e-05,
+      "loss": 0.0969,
+      "step": 1163
+    },
+    {
+      "epoch": 0.910086004691165,
+      "grad_norm": 0.24078141152858734,
+      "learning_rate": 6.627912459380487e-05,
+      "loss": 0.0791,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9108678655199375,
+      "grad_norm": 0.41361549496650696,
+      "learning_rate": 6.621459123735522e-05,
+      "loss": 0.1202,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9116497263487099,
+      "grad_norm": 0.31104376912117004,
+      "learning_rate": 6.615002768173299e-05,
+      "loss": 0.0979,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9124315871774824,
+      "grad_norm": 0.2964945435523987,
+      "learning_rate": 6.608543404718578e-05,
+      "loss": 0.1185,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9132134480062549,
+      "grad_norm": 0.37027430534362793,
+      "learning_rate": 6.60208104540172e-05,
+      "loss": 0.1269,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9139953088350273,
+      "grad_norm": 0.40137842297554016,
+      "learning_rate": 6.595615702258676e-05,
+      "loss": 0.127,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9147771696637998,
+      "grad_norm": 0.41014501452445984,
+      "learning_rate": 6.589147387330939e-05,
+      "loss": 0.0935,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9155590304925724,
+      "grad_norm": 0.3405829071998596,
+      "learning_rate": 6.58267611266555e-05,
+      "loss": 0.1261,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9163408913213448,
+      "grad_norm": 0.32811757922172546,
+      "learning_rate": 6.576201890315058e-05,
+      "loss": 0.1491,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9171227521501173,
+      "grad_norm": 0.3566417098045349,
+      "learning_rate": 6.569724732337495e-05,
+      "loss": 0.1167,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9179046129788898,
+      "grad_norm": 0.31817594170570374,
+      "learning_rate": 6.563244650796371e-05,
+      "loss": 0.1016,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9186864738076622,
+      "grad_norm": 0.291289359331131,
+      "learning_rate": 6.556761657760635e-05,
+      "loss": 0.0951,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9194683346364347,
+      "grad_norm": 0.2821560204029083,
+      "learning_rate": 6.550275765304661e-05,
+      "loss": 0.0743,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9202501954652071,
+      "grad_norm": 0.39526405930519104,
+      "learning_rate": 6.543786985508223e-05,
+      "loss": 0.1109,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9210320562939797,
+      "grad_norm": 0.44412076473236084,
+      "learning_rate": 6.537295330456469e-05,
+      "loss": 0.1296,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9218139171227522,
+      "grad_norm": 0.3970738649368286,
+      "learning_rate": 6.53080081223991e-05,
+      "loss": 0.1113,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9225957779515246,
+      "grad_norm": 0.4147748351097107,
+      "learning_rate": 6.524303442954378e-05,
+      "loss": 0.1096,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9233776387802971,
+      "grad_norm": 0.37349894642829895,
+      "learning_rate": 6.517803234701025e-05,
+      "loss": 0.1036,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9241594996090696,
+      "grad_norm": 0.35803303122520447,
+      "learning_rate": 6.511300199586289e-05,
+      "loss": 0.1096,
+      "step": 1182
+    },
+    {
+      "epoch": 0.924941360437842,
+      "grad_norm": 0.5214382410049438,
+      "learning_rate": 6.504794349721866e-05,
+      "loss": 0.1325,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9257232212666145,
+      "grad_norm": 0.40093794465065,
+      "learning_rate": 6.498285697224703e-05,
+      "loss": 0.1131,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9265050820953871,
+      "grad_norm": 0.39705073833465576,
+      "learning_rate": 6.491774254216963e-05,
+      "loss": 0.109,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9272869429241595,
+      "grad_norm": 0.3166132867336273,
+      "learning_rate": 6.485260032826005e-05,
+      "loss": 0.1104,
+      "step": 1186
+    },
+    {
+      "epoch": 0.928068803752932,
+      "grad_norm": 0.278113454580307,
+      "learning_rate": 6.478743045184365e-05,
+      "loss": 0.093,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9288506645817045,
+      "grad_norm": 0.3798607587814331,
+      "learning_rate": 6.472223303429727e-05,
+      "loss": 0.1601,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9296325254104769,
+      "grad_norm": 0.43413954973220825,
+      "learning_rate": 6.465700819704913e-05,
+      "loss": 0.1416,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9304143862392494,
+      "grad_norm": 0.3416893780231476,
+      "learning_rate": 6.459175606157842e-05,
+      "loss": 0.1145,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9311962470680218,
+      "grad_norm": 0.33215364813804626,
+      "learning_rate": 6.452647674941524e-05,
+      "loss": 0.1113,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9319781078967944,
+      "grad_norm": 0.34538665413856506,
+      "learning_rate": 6.446117038214027e-05,
+      "loss": 0.09,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9327599687255669,
+      "grad_norm": 0.4208803176879883,
+      "learning_rate": 6.439583708138459e-05,
+      "loss": 0.1198,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9335418295543393,
+      "grad_norm": 0.35723745822906494,
+      "learning_rate": 6.433047696882945e-05,
+      "loss": 0.1337,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9343236903831118,
+      "grad_norm": 0.1974477916955948,
+      "learning_rate": 6.426509016620603e-05,
+      "loss": 0.069,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9351055512118843,
+      "grad_norm": 0.3032665252685547,
+      "learning_rate": 6.419967679529522e-05,
+      "loss": 0.1063,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9358874120406567,
+      "grad_norm": 0.4309205114841461,
+      "learning_rate": 6.413423697792737e-05,
+      "loss": 0.1138,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9366692728694292,
+      "grad_norm": 0.31886643171310425,
+      "learning_rate": 6.406877083598213e-05,
+      "loss": 0.1061,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9374511336982018,
+      "grad_norm": 0.45376572012901306,
+      "learning_rate": 6.400327849138814e-05,
+      "loss": 0.1211,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9382329945269742,
+      "grad_norm": 0.37083959579467773,
+      "learning_rate": 6.393776006612288e-05,
+      "loss": 0.1201,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9382329945269742,
+      "eval_loss": 0.12676353752613068,
+      "eval_runtime": 13.3499,
+      "eval_samples_per_second": 3.895,
+      "eval_steps_per_second": 0.974,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9390148553557467,
+      "grad_norm": 0.47506844997406006,
+      "learning_rate": 6.387221568221238e-05,
+      "loss": 0.1607,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9397967161845192,
+      "grad_norm": 0.4289664030075073,
+      "learning_rate": 6.380664546173101e-05,
+      "loss": 0.1491,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9405785770132916,
+      "grad_norm": 0.3842359185218811,
+      "learning_rate": 6.374104952680125e-05,
+      "loss": 0.1403,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9413604378420641,
+      "grad_norm": 0.36879363656044006,
+      "learning_rate": 6.367542799959352e-05,
+      "loss": 0.1256,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9421422986708365,
+      "grad_norm": 0.5137174725532532,
+      "learning_rate": 6.360978100232587e-05,
+      "loss": 0.1216,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9429241594996091,
+      "grad_norm": 0.41332346200942993,
+      "learning_rate": 6.35441086572638e-05,
+      "loss": 0.1551,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9437060203283816,
+      "grad_norm": 0.3252861201763153,
+      "learning_rate": 6.347841108672e-05,
+      "loss": 0.1116,
+      "step": 1207
+    },
+    {
+      "epoch": 0.944487881157154,
+      "grad_norm": 0.30747750401496887,
+      "learning_rate": 6.341268841305415e-05,
+      "loss": 0.1039,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9452697419859265,
+      "grad_norm": 0.3144964575767517,
+      "learning_rate": 6.334694075867269e-05,
+      "loss": 0.1119,
+      "step": 1209
+    },
+    {
+      "epoch": 0.946051602814699,
+      "grad_norm": 0.31715381145477295,
+      "learning_rate": 6.328116824602859e-05,
+      "loss": 0.105,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9468334636434714,
+      "grad_norm": 0.3612911105155945,
+      "learning_rate": 6.321537099762114e-05,
+      "loss": 0.1276,
+      "step": 1211
+    },
+    {
+      "epoch": 0.947615324472244,
+      "grad_norm": 0.40377581119537354,
+      "learning_rate": 6.314954913599563e-05,
+      "loss": 0.1227,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9483971853010165,
+      "grad_norm": 0.35488733649253845,
+      "learning_rate": 6.308370278374325e-05,
+      "loss": 0.1183,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9491790461297889,
+      "grad_norm": 0.4154433012008667,
+      "learning_rate": 6.301783206350076e-05,
+      "loss": 0.137,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9499609069585614,
+      "grad_norm": 0.3620409667491913,
+      "learning_rate": 6.295193709795037e-05,
+      "loss": 0.1155,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9507427677873339,
+      "grad_norm": 0.4153326451778412,
+      "learning_rate": 6.288601800981942e-05,
+      "loss": 0.1002,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9515246286161063,
+      "grad_norm": 0.38747620582580566,
+      "learning_rate": 6.282007492188011e-05,
+      "loss": 0.1343,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9523064894448788,
+      "grad_norm": 0.3720448911190033,
+      "learning_rate": 6.275410795694943e-05,
+      "loss": 0.0982,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9530883502736512,
+      "grad_norm": 0.27825435996055603,
+      "learning_rate": 6.268811723788877e-05,
+      "loss": 0.0688,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9538702111024238,
+      "grad_norm": 0.4155515730381012,
+      "learning_rate": 6.262210288760383e-05,
+      "loss": 0.1256,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9546520719311963,
+      "grad_norm": 0.26842448115348816,
+      "learning_rate": 6.255606502904429e-05,
+      "loss": 0.1326,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9554339327599687,
+      "grad_norm": 0.377493679523468,
+      "learning_rate": 6.249000378520358e-05,
+      "loss": 0.1258,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9562157935887412,
+      "grad_norm": 0.3844210207462311,
+      "learning_rate": 6.242391927911872e-05,
+      "loss": 0.1402,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9569976544175137,
+      "grad_norm": 0.4269979000091553,
+      "learning_rate": 6.235781163387007e-05,
+      "loss": 0.1253,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9577795152462861,
+      "grad_norm": 0.4109889566898346,
+      "learning_rate": 6.229168097258106e-05,
+      "loss": 0.1343,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9585613760750586,
+      "grad_norm": 0.3345010578632355,
+      "learning_rate": 6.2225527418418e-05,
+      "loss": 0.0976,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9593432369038312,
+      "grad_norm": 0.45198965072631836,
+      "learning_rate": 6.21593510945898e-05,
+      "loss": 0.1088,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9601250977326036,
+      "grad_norm": 0.3664204180240631,
+      "learning_rate": 6.209315212434783e-05,
+      "loss": 0.0907,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9609069585613761,
+      "grad_norm": 0.3774789571762085,
+      "learning_rate": 6.202693063098561e-05,
+      "loss": 0.1208,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9616888193901486,
+      "grad_norm": 0.3721179664134979,
+      "learning_rate": 6.196068673783863e-05,
+      "loss": 0.1295,
+      "step": 1230
+    },
+    {
+      "epoch": 0.962470680218921,
+      "grad_norm": 0.3367028534412384,
+      "learning_rate": 6.189442056828407e-05,
+      "loss": 0.1699,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9632525410476935,
+      "grad_norm": 0.3180219531059265,
+      "learning_rate": 6.182813224574061e-05,
+      "loss": 0.1318,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9640344018764659,
+      "grad_norm": 0.357023686170578,
+      "learning_rate": 6.176182189366819e-05,
+      "loss": 0.1029,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9648162627052385,
+      "grad_norm": 0.34663912653923035,
+      "learning_rate": 6.169548963556779e-05,
+      "loss": 0.1013,
+      "step": 1234
+    },
+    {
+      "epoch": 0.965598123534011,
+      "grad_norm": 0.3682876527309418,
+      "learning_rate": 6.16291355949812e-05,
+      "loss": 0.0902,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9663799843627834,
+      "grad_norm": 0.3583768904209137,
+      "learning_rate": 6.156275989549072e-05,
+      "loss": 0.092,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9671618451915559,
+      "grad_norm": 0.3423864543437958,
+      "learning_rate": 6.149636266071904e-05,
+      "loss": 0.1074,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9679437060203284,
+      "grad_norm": 0.42394763231277466,
+      "learning_rate": 6.142994401432896e-05,
+      "loss": 0.1143,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9687255668491008,
+      "grad_norm": 0.3888026177883148,
+      "learning_rate": 6.136350408002314e-05,
+      "loss": 0.1147,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9695074276778733,
+      "grad_norm": 0.3200819492340088,
+      "learning_rate": 6.12970429815439e-05,
+      "loss": 0.1197,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9702892885066459,
+      "grad_norm": 0.37223726511001587,
+      "learning_rate": 6.123056084267296e-05,
+      "loss": 0.0876,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9710711493354183,
+      "grad_norm": 0.39913463592529297,
+      "learning_rate": 6.116405778723123e-05,
+      "loss": 0.1182,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9718530101641908,
+      "grad_norm": 0.47016170620918274,
+      "learning_rate": 6.109753393907862e-05,
+      "loss": 0.1375,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9726348709929633,
+      "grad_norm": 0.38198941946029663,
+      "learning_rate": 6.1030989422113715e-05,
+      "loss": 0.1543,
+      "step": 1244
+    },
+    {
+      "epoch": 0.9734167318217357,
+      "grad_norm": 0.3025572597980499,
+      "learning_rate": 6.09644243602736e-05,
+      "loss": 0.0988,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9741985926505082,
+      "grad_norm": 0.4157261848449707,
+      "learning_rate": 6.0897838877533644e-05,
+      "loss": 0.1167,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9749804534792806,
+      "grad_norm": 0.5874128341674805,
+      "learning_rate": 6.0831233097907236e-05,
+      "loss": 0.1105,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9757623143080532,
+      "grad_norm": 0.4182235300540924,
+      "learning_rate": 6.0764607145445576e-05,
+      "loss": 0.1205,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9765441751368257,
+      "grad_norm": 0.41899439692497253,
+      "learning_rate": 6.069796114423743e-05,
+      "loss": 0.1544,
+      "step": 1249
+    },
+    {
+      "epoch": 0.9773260359655981,
+      "grad_norm": 0.3662065863609314,
+      "learning_rate": 6.0631295218408926e-05,
+      "loss": 0.1081,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9781078967943706,
+      "grad_norm": 0.3846765458583832,
+      "learning_rate": 6.056460949212324e-05,
+      "loss": 0.1128,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9788897576231431,
+      "grad_norm": 0.31117817759513855,
+      "learning_rate": 6.049790408958048e-05,
+      "loss": 0.1114,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9796716184519155,
+      "grad_norm": 0.359997421503067,
+      "learning_rate": 6.043117913501741e-05,
+      "loss": 0.1078,
+      "step": 1253
+    },
+    {
+      "epoch": 0.980453479280688,
+      "grad_norm": 0.30968189239501953,
+      "learning_rate": 6.036443475270717e-05,
+      "loss": 0.1071,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9812353401094606,
+      "grad_norm": 0.378490686416626,
+      "learning_rate": 6.029767106695909e-05,
+      "loss": 0.1085,
+      "step": 1255
+    },
+    {
+      "epoch": 0.982017200938233,
+      "grad_norm": 0.32061898708343506,
+      "learning_rate": 6.0230888202118473e-05,
+      "loss": 0.108,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9827990617670055,
+      "grad_norm": 0.42287373542785645,
+      "learning_rate": 6.0164086282566326e-05,
+      "loss": 0.129,
+      "step": 1257
+    },
+    {
+      "epoch": 0.983580922595778,
+      "grad_norm": 0.3686187267303467,
+      "learning_rate": 6.009726543271914e-05,
+      "loss": 0.0982,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9843627834245504,
+      "grad_norm": 0.41652533411979675,
+      "learning_rate": 6.0030425777028685e-05,
+      "loss": 0.1222,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9851446442533229,
+      "grad_norm": 0.33978602290153503,
+      "learning_rate": 5.996356743998175e-05,
+      "loss": 0.1171,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9859265050820953,
+      "grad_norm": 0.2727195918560028,
+      "learning_rate": 5.9896690546099906e-05,
+      "loss": 0.0973,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9867083659108679,
+      "grad_norm": 0.3532811105251312,
+      "learning_rate": 5.982979521993929e-05,
+      "loss": 0.1165,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9874902267396404,
+      "grad_norm": 0.43228256702423096,
+      "learning_rate": 5.9762881586090344e-05,
+      "loss": 0.1413,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9882720875684128,
+      "grad_norm": 0.34505805373191833,
+      "learning_rate": 5.969594976917768e-05,
+      "loss": 0.1569,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9890539483971853,
+      "grad_norm": 0.40569958090782166,
+      "learning_rate": 5.962899989385969e-05,
+      "loss": 0.1049,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9898358092259578,
+      "grad_norm": 0.39335259795188904,
+      "learning_rate": 5.9562032084828457e-05,
+      "loss": 0.1012,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9906176700547302,
+      "grad_norm": 0.39652782678604126,
+      "learning_rate": 5.9495046466809444e-05,
+      "loss": 0.1355,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9913995308835027,
+      "grad_norm": 0.32174959778785706,
+      "learning_rate": 5.942804316456126e-05,
+      "loss": 0.1018,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9921813917122753,
+      "grad_norm": 0.34195002913475037,
+      "learning_rate": 5.936102230287553e-05,
+      "loss": 0.1363,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9929632525410477,
+      "grad_norm": 0.306333065032959,
+      "learning_rate": 5.929398400657647e-05,
+      "loss": 0.0909,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9937451133698202,
+      "grad_norm": 0.4631345272064209,
+      "learning_rate": 5.9226928400520854e-05,
+      "loss": 0.1437,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9945269741985927,
+      "grad_norm": 0.42312151193618774,
+      "learning_rate": 5.9159855609597645e-05,
+      "loss": 0.1745,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9953088350273651,
+      "grad_norm": 0.34158191084861755,
+      "learning_rate": 5.9092765758727854e-05,
+      "loss": 0.1221,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9960906958561376,
+      "grad_norm": 0.38159799575805664,
+      "learning_rate": 5.902565897286425e-05,
+      "loss": 0.103,
+      "step": 1274
+    },
+    {
+      "epoch": 0.99687255668491,
+      "grad_norm": 0.3136163651943207,
+      "learning_rate": 5.8958535376991106e-05,
+      "loss": 0.1218,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9976544175136826,
+      "grad_norm": 0.3725820481777191,
+      "learning_rate": 5.8891395096124067e-05,
+      "loss": 0.1232,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9984362783424551,
+      "grad_norm": 0.3924759030342102,
+      "learning_rate": 5.882423825530981e-05,
+      "loss": 0.1176,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9992181391712275,
+      "grad_norm": 0.35742858052253723,
+      "learning_rate": 5.875706497962584e-05,
+      "loss": 0.1124,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42548516392707825,
+      "learning_rate": 5.8689875394180335e-05,
+      "loss": 0.14,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0007818608287724,
+      "grad_norm": 0.3003665506839752,
+      "learning_rate": 5.8622669624111804e-05,
+      "loss": 0.0545,
+      "step": 1280
+    },
+    {
+      "epoch": 1.001563721657545,
+      "grad_norm": 0.24529042840003967,
+      "learning_rate": 5.855544779458887e-05,
+      "loss": 0.0673,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0023455824863174,
+      "grad_norm": 0.3029482066631317,
+      "learning_rate": 5.8488210030810153e-05,
+      "loss": 0.0856,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0031274433150899,
+      "grad_norm": 0.3220517933368683,
+      "learning_rate": 5.8420956458003876e-05,
+      "loss": 0.0782,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0039093041438625,
+      "grad_norm": 0.37451228499412537,
+      "learning_rate": 5.835368720142773e-05,
+      "loss": 0.1183,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0046911649726349,
+      "grad_norm": 0.33108869194984436,
+      "learning_rate": 5.828640238636861e-05,
+      "loss": 0.1057,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0054730258014073,
+      "grad_norm": 0.2879885137081146,
+      "learning_rate": 5.82191021381424e-05,
+      "loss": 0.0831,
+      "step": 1286
+    },
+    {
+      "epoch": 1.00625488663018,
+      "grad_norm": 0.40735185146331787,
+      "learning_rate": 5.815178658209372e-05,
+      "loss": 0.1013,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0070367474589523,
+      "grad_norm": 0.31678634881973267,
+      "learning_rate": 5.808445584359572e-05,
+      "loss": 0.0781,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0078186082877247,
+      "grad_norm": 0.3188799023628235,
+      "learning_rate": 5.801711004804979e-05,
+      "loss": 0.0491,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0086004691164974,
+      "grad_norm": 0.3288949728012085,
+      "learning_rate": 5.7949749320885394e-05,
+      "loss": 0.0748,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0093823299452698,
+      "grad_norm": 0.3360097110271454,
+      "learning_rate": 5.7882373787559775e-05,
+      "loss": 0.0603,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0101641907740422,
+      "grad_norm": 0.34202682971954346,
+      "learning_rate": 5.781498357355779e-05,
+      "loss": 0.0639,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0109460516028146,
+      "grad_norm": 0.3513847589492798,
+      "learning_rate": 5.7747578804391624e-05,
+      "loss": 0.0636,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0117279124315872,
+      "grad_norm": 0.3393630087375641,
+      "learning_rate": 5.768015960560055e-05,
+      "loss": 0.0515,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0125097732603596,
+      "grad_norm": 0.4196714162826538,
+      "learning_rate": 5.761272610275074e-05,
+      "loss": 0.0524,
+      "step": 1295
+    },
+    {
+      "epoch": 1.013291634089132,
+      "grad_norm": 0.33013924956321716,
+      "learning_rate": 5.754527842143498e-05,
+      "loss": 0.0503,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0140734949179047,
+      "grad_norm": 0.3442160189151764,
+      "learning_rate": 5.747781668727251e-05,
+      "loss": 0.0461,
+      "step": 1297
+    },
+    {
+      "epoch": 1.014855355746677,
+      "grad_norm": 0.3586715757846832,
+      "learning_rate": 5.741034102590871e-05,
+      "loss": 0.0656,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0156372165754495,
+      "grad_norm": 0.3679417073726654,
+      "learning_rate": 5.73428515630149e-05,
+      "loss": 0.0644,
+      "step": 1299
+    },
+    {
+      "epoch": 1.016419077404222,
+      "grad_norm": 0.3188945949077606,
+      "learning_rate": 5.72753484242881e-05,
+      "loss": 0.0712,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0172009382329945,
+      "grad_norm": 0.35757797956466675,
+      "learning_rate": 5.72078317354508e-05,
+      "loss": 0.0607,
+      "step": 1301
+    },
+    {
+      "epoch": 1.017982799061767,
+      "grad_norm": 0.4698977768421173,
+      "learning_rate": 5.714030162225077e-05,
+      "loss": 0.0533,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0187646598905395,
+      "grad_norm": 0.3897135853767395,
+      "learning_rate": 5.7072758210460716e-05,
+      "loss": 0.0516,
+      "step": 1303
+    },
+    {
+      "epoch": 1.019546520719312,
+      "grad_norm": 0.6064049005508423,
+      "learning_rate": 5.700520162587817e-05,
+      "loss": 0.1096,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0203283815480844,
+      "grad_norm": 0.3872586488723755,
+      "learning_rate": 5.693763199432516e-05,
+      "loss": 0.0424,
+      "step": 1305
+    },
+    {
+      "epoch": 1.021110242376857,
+      "grad_norm": 0.3485041558742523,
+      "learning_rate": 5.687004944164801e-05,
+      "loss": 0.0542,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0218921032056294,
+      "grad_norm": 0.38680729269981384,
+      "learning_rate": 5.680245409371716e-05,
+      "loss": 0.0648,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0226739640344018,
+      "grad_norm": 0.3012762665748596,
+      "learning_rate": 5.673484607642684e-05,
+      "loss": 0.0547,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0234558248631744,
+      "grad_norm": 0.41317009925842285,
+      "learning_rate": 5.666722551569484e-05,
+      "loss": 0.0519,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0242376856919468,
+      "grad_norm": 0.3628830313682556,
+      "learning_rate": 5.6599592537462406e-05,
+      "loss": 0.0963,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0250195465207192,
+      "grad_norm": 0.39497920870780945,
+      "learning_rate": 5.653194726769382e-05,
+      "loss": 0.072,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0258014073494919,
+      "grad_norm": 0.4044725298881531,
+      "learning_rate": 5.646428983237633e-05,
+      "loss": 0.1094,
+      "step": 1312
+    },
+    {
+      "epoch": 1.0265832681782643,
+      "grad_norm": 0.3730235695838928,
+      "learning_rate": 5.63966203575198e-05,
+      "loss": 0.0848,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0273651290070367,
+      "grad_norm": 0.2993086874485016,
+      "learning_rate": 5.63289389691565e-05,
+      "loss": 0.0731,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0281469898358093,
+      "grad_norm": 0.35841402411460876,
+      "learning_rate": 5.6261245793340944e-05,
+      "loss": 0.0795,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0289288506645817,
+      "grad_norm": 0.39184918999671936,
+      "learning_rate": 5.619354095614955e-05,
+      "loss": 0.0848,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0297107114933541,
+      "grad_norm": 0.30408361554145813,
+      "learning_rate": 5.612582458368047e-05,
+      "loss": 0.0419,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0304925723221268,
+      "grad_norm": 0.33947086334228516,
+      "learning_rate": 5.60580968020534e-05,
+      "loss": 0.0504,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0312744331508992,
+      "grad_norm": 0.31831854581832886,
+      "learning_rate": 5.599035773740915e-05,
+      "loss": 0.0588,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0320562939796716,
+      "grad_norm": 0.38083136081695557,
+      "learning_rate": 5.592260751590967e-05,
+      "loss": 0.0768,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0328381548084442,
+      "grad_norm": 0.31160804629325867,
+      "learning_rate": 5.5854846263737625e-05,
+      "loss": 0.1107,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0336200156372166,
+      "grad_norm": 0.3518250584602356,
+      "learning_rate": 5.578707410709626e-05,
+      "loss": 0.0498,
+      "step": 1322
+    },
+    {
+      "epoch": 1.034401876465989,
+      "grad_norm": 0.43158453702926636,
+      "learning_rate": 5.571929117220911e-05,
+      "loss": 0.0467,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0351837372947614,
+      "grad_norm": 0.47497236728668213,
+      "learning_rate": 5.5651497585319756e-05,
+      "loss": 0.0749,
+      "step": 1324
+    },
+    {
+      "epoch": 1.035965598123534,
+      "grad_norm": 0.32614821195602417,
+      "learning_rate": 5.558369347269169e-05,
+      "loss": 0.0593,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0367474589523065,
+      "grad_norm": 0.30701157450675964,
+      "learning_rate": 5.551587896060795e-05,
+      "loss": 0.0771,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0375293197810789,
+      "grad_norm": 0.394361674785614,
+      "learning_rate": 5.544805417537096e-05,
+      "loss": 0.0752,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0383111806098515,
+      "grad_norm": 0.32267239689826965,
+      "learning_rate": 5.538021924330225e-05,
+      "loss": 0.0449,
+      "step": 1328
+    },
+    {
+      "epoch": 1.039093041438624,
+      "grad_norm": 0.3949149250984192,
+      "learning_rate": 5.531237429074231e-05,
+      "loss": 0.0756,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0398749022673963,
+      "grad_norm": 0.3293861448764801,
+      "learning_rate": 5.5244519444050234e-05,
+      "loss": 0.1048,
+      "step": 1330
+    },
+    {
+      "epoch": 1.040656763096169,
+      "grad_norm": 0.3384082019329071,
+      "learning_rate": 5.517665482960359e-05,
+      "loss": 0.0579,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0414386239249414,
+      "grad_norm": 0.29830190539360046,
+      "learning_rate": 5.510878057379809e-05,
+      "loss": 0.0593,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0422204847537138,
+      "grad_norm": 0.5683103203773499,
+      "learning_rate": 5.504089680304745e-05,
+      "loss": 0.0614,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0430023455824864,
+      "grad_norm": 0.3690869212150574,
+      "learning_rate": 5.497300364378306e-05,
+      "loss": 0.0643,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0437842064112588,
+      "grad_norm": 0.32530248165130615,
+      "learning_rate": 5.490510122245384e-05,
+      "loss": 0.0569,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0445660672400312,
+      "grad_norm": 0.4188139736652374,
+      "learning_rate": 5.483718966552594e-05,
+      "loss": 0.1026,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0453479280688038,
+      "grad_norm": 0.4417498707771301,
+      "learning_rate": 5.47692690994825e-05,
+      "loss": 0.0764,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0461297888975762,
+      "grad_norm": 0.312726229429245,
+      "learning_rate": 5.47013396508235e-05,
+      "loss": 0.0339,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0469116497263486,
+      "grad_norm": 0.46260857582092285,
+      "learning_rate": 5.463340144606541e-05,
+      "loss": 0.103,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0476935105551213,
+      "grad_norm": 0.4564020335674286,
+      "learning_rate": 5.456545461174102e-05,
+      "loss": 0.0427,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0484753713838937,
+      "grad_norm": 0.2873257100582123,
+      "learning_rate": 5.449749927439922e-05,
+      "loss": 0.0531,
+      "step": 1341
+    },
+    {
+      "epoch": 1.049257232212666,
+      "grad_norm": 0.4216899871826172,
+      "learning_rate": 5.44295355606047e-05,
+      "loss": 0.0673,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0500390930414387,
+      "grad_norm": 0.4410907030105591,
+      "learning_rate": 5.436156359693777e-05,
+      "loss": 0.0768,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0508209538702111,
+      "grad_norm": 0.4740668833255768,
+      "learning_rate": 5.42935835099941e-05,
+      "loss": 0.0923,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0516028146989835,
+      "grad_norm": 0.34221503138542175,
+      "learning_rate": 5.422559542638448e-05,
+      "loss": 0.0407,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0523846755277562,
+      "grad_norm": 0.35219067335128784,
+      "learning_rate": 5.415759947273462e-05,
+      "loss": 0.0716,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0531665363565286,
+      "grad_norm": 0.3162952959537506,
+      "learning_rate": 5.4089595775684886e-05,
+      "loss": 0.0487,
+      "step": 1347
+    },
+    {
+      "epoch": 1.053948397185301,
+      "grad_norm": 0.3372567594051361,
+      "learning_rate": 5.4021584461890015e-05,
+      "loss": 0.0387,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0547302580140734,
+      "grad_norm": 0.40237340331077576,
+      "learning_rate": 5.395356565801899e-05,
+      "loss": 0.0505,
+      "step": 1349
+    },
+    {
+      "epoch": 1.055512118842846,
+      "grad_norm": 0.3668085038661957,
+      "learning_rate": 5.388553949075471e-05,
+      "loss": 0.0367,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0562939796716184,
+      "grad_norm": 0.4131850302219391,
+      "learning_rate": 5.3817506086793813e-05,
+      "loss": 0.058,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0570758405003908,
+      "grad_norm": 0.43202531337738037,
+      "learning_rate": 5.374946557284638e-05,
+      "loss": 0.066,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0578577013291635,
+      "grad_norm": 0.41850730776786804,
+      "learning_rate": 5.368141807563578e-05,
+      "loss": 0.0603,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0586395621579359,
+      "grad_norm": 0.5610138773918152,
+      "learning_rate": 5.361336372189835e-05,
+      "loss": 0.0686,
+      "step": 1354
+    },
+    {
+      "epoch": 1.0594214229867083,
+      "grad_norm": 0.4340571165084839,
+      "learning_rate": 5.35453026383832e-05,
+      "loss": 0.0597,
+      "step": 1355
+    },
+    {
+      "epoch": 1.060203283815481,
+      "grad_norm": 0.526792049407959,
+      "learning_rate": 5.3477234951852e-05,
+      "loss": 0.0695,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0609851446442533,
+      "grad_norm": 0.3665534555912018,
+      "learning_rate": 5.34091607890787e-05,
+      "loss": 0.0816,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0617670054730257,
+      "grad_norm": 0.29408949613571167,
+      "learning_rate": 5.334108027684931e-05,
+      "loss": 0.0321,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0625488663017983,
+      "grad_norm": 0.5186415910720825,
+      "learning_rate": 5.327299354196167e-05,
+      "loss": 0.1035,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0633307271305708,
+      "grad_norm": 0.3105923533439636,
+      "learning_rate": 5.320490071122519e-05,
+      "loss": 0.0377,
+      "step": 1360
+    },
+    {
+      "epoch": 1.0641125879593432,
+      "grad_norm": 0.3671916723251343,
+      "learning_rate": 5.313680191146071e-05,
+      "loss": 0.0622,
+      "step": 1361
+    },
+    {
+      "epoch": 1.0648944487881158,
+      "grad_norm": 0.38357749581336975,
+      "learning_rate": 5.306869726950007e-05,
+      "loss": 0.0636,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0656763096168882,
+      "grad_norm": 0.41577282547950745,
+      "learning_rate": 5.300058691218607e-05,
+      "loss": 0.0387,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0664581704456606,
+      "grad_norm": 0.3620113134384155,
+      "learning_rate": 5.293247096637214e-05,
+      "loss": 0.0534,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0672400312744332,
+      "grad_norm": 0.3379097878932953,
+      "learning_rate": 5.28643495589221e-05,
+      "loss": 0.0686,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0680218921032056,
+      "grad_norm": 0.26274728775024414,
+      "learning_rate": 5.279622281670998e-05,
+      "loss": 0.0368,
+      "step": 1366
+    },
+    {
+      "epoch": 1.068803752931978,
+      "grad_norm": 0.35920795798301697,
+      "learning_rate": 5.272809086661972e-05,
+      "loss": 0.0625,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0695856137607507,
+      "grad_norm": 0.32537350058555603,
+      "learning_rate": 5.2659953835544916e-05,
+      "loss": 0.0647,
+      "step": 1368
+    },
+    {
+      "epoch": 1.070367474589523,
+      "grad_norm": 0.38874363899230957,
+      "learning_rate": 5.2591811850388726e-05,
+      "loss": 0.0664,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0711493354182955,
+      "grad_norm": 0.3997640013694763,
+      "learning_rate": 5.252366503806345e-05,
+      "loss": 0.0453,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0719311962470681,
+      "grad_norm": 0.4044627845287323,
+      "learning_rate": 5.245551352549044e-05,
+      "loss": 0.0725,
+      "step": 1371
+    },
+    {
+      "epoch": 1.0727130570758405,
+      "grad_norm": 0.40725669264793396,
+      "learning_rate": 5.238735743959974e-05,
+      "loss": 0.0528,
+      "step": 1372
+    },
+    {
+      "epoch": 1.073494917904613,
+      "grad_norm": 0.3452327847480774,
+      "learning_rate": 5.231919690732995e-05,
+      "loss": 0.0454,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0742767787333856,
+      "grad_norm": 0.33204716444015503,
+      "learning_rate": 5.2251032055627946e-05,
+      "loss": 0.0335,
+      "step": 1374
+    },
+    {
+      "epoch": 1.075058639562158,
+      "grad_norm": 0.39199399948120117,
+      "learning_rate": 5.218286301144867e-05,
+      "loss": 0.038,
+      "step": 1375
+    },
+    {
+      "epoch": 1.0758405003909304,
+      "grad_norm": 0.44110965728759766,
+      "learning_rate": 5.211468990175481e-05,
+      "loss": 0.0997,
+      "step": 1376
+    },
+    {
+      "epoch": 1.076622361219703,
+      "grad_norm": 0.5448676347732544,
+      "learning_rate": 5.2046512853516696e-05,
+      "loss": 0.0656,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0774042220484754,
+      "grad_norm": 0.450676828622818,
+      "learning_rate": 5.197833199371194e-05,
+      "loss": 0.0515,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0781860828772478,
+      "grad_norm": 0.373246967792511,
+      "learning_rate": 5.1910147449325295e-05,
+      "loss": 0.0403,
+      "step": 1379
+    },
+    {
+      "epoch": 1.0789679437060202,
+      "grad_norm": 0.4239073693752289,
+      "learning_rate": 5.184195934734835e-05,
+      "loss": 0.0699,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0797498045347929,
+      "grad_norm": 0.48097920417785645,
+      "learning_rate": 5.177376781477933e-05,
+      "loss": 0.0779,
+      "step": 1381
+    },
+    {
+      "epoch": 1.0805316653635653,
+      "grad_norm": 0.3312666118144989,
+      "learning_rate": 5.170557297862284e-05,
+      "loss": 0.0444,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0813135261923377,
+      "grad_norm": 0.328531414270401,
+      "learning_rate": 5.163737496588964e-05,
+      "loss": 0.0436,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0820953870211103,
+      "grad_norm": 0.303060382604599,
+      "learning_rate": 5.156917390359643e-05,
+      "loss": 0.0375,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0828772478498827,
+      "grad_norm": 0.3380538821220398,
+      "learning_rate": 5.150096991876556e-05,
+      "loss": 0.0416,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0836591086786551,
+      "grad_norm": 0.3253069519996643,
+      "learning_rate": 5.143276313842485e-05,
+      "loss": 0.058,
+      "step": 1386
+    },
+    {
+      "epoch": 1.0844409695074277,
+      "grad_norm": 0.38924887776374817,
+      "learning_rate": 5.1364553689607264e-05,
+      "loss": 0.074,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0852228303362002,
+      "grad_norm": 0.39946839213371277,
+      "learning_rate": 5.129634169935081e-05,
+      "loss": 0.0764,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0860046911649726,
+      "grad_norm": 0.3759361803531647,
+      "learning_rate": 5.1228127294698225e-05,
+      "loss": 0.0538,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0867865519937452,
+      "grad_norm": 0.3306618928909302,
+      "learning_rate": 5.1159910602696683e-05,
+      "loss": 0.0857,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0875684128225176,
+      "grad_norm": 0.5133655071258545,
+      "learning_rate": 5.1091691750397675e-05,
+      "loss": 0.0727,
+      "step": 1391
+    },
+    {
+      "epoch": 1.08835027365129,
+      "grad_norm": 0.43473249673843384,
+      "learning_rate": 5.1023470864856684e-05,
+      "loss": 0.0567,
+      "step": 1392
+    },
+    {
+      "epoch": 1.0891321344800626,
+      "grad_norm": 0.4632827043533325,
+      "learning_rate": 5.0955248073132975e-05,
+      "loss": 0.0531,
+      "step": 1393
+    },
+    {
+      "epoch": 1.089913995308835,
+      "grad_norm": 0.4766741991043091,
+      "learning_rate": 5.088702350228942e-05,
+      "loss": 0.0995,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0906958561376074,
+      "grad_norm": 0.37831687927246094,
+      "learning_rate": 5.081879727939214e-05,
+      "loss": 0.0374,
+      "step": 1395
+    },
+    {
+      "epoch": 1.09147771696638,
+      "grad_norm": 0.4387483298778534,
+      "learning_rate": 5.075056953151035e-05,
+      "loss": 0.0639,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0922595777951525,
+      "grad_norm": 0.3122713565826416,
+      "learning_rate": 5.068234038571612e-05,
+      "loss": 0.0737,
+      "step": 1397
+    },
+    {
+      "epoch": 1.0930414386239249,
+      "grad_norm": 0.39683401584625244,
+      "learning_rate": 5.0614109969084125e-05,
+      "loss": 0.1093,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0938232994526975,
+      "grad_norm": 0.34212830662727356,
+      "learning_rate": 5.054587840869136e-05,
+      "loss": 0.047,
+      "step": 1399
+    },
+    {
+      "epoch": 1.09460516028147,
+      "grad_norm": 0.4159735143184662,
+      "learning_rate": 5.047764583161704e-05,
+      "loss": 0.0899,
+      "step": 1400
+    },
+    {
+      "epoch": 1.09460516028147,
+      "eval_loss": 0.12889541685581207,
+      "eval_runtime": 13.3979,
+      "eval_samples_per_second": 3.881,
+      "eval_steps_per_second": 0.97,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0953870211102423,
+      "grad_norm": 0.3841370940208435,
+      "learning_rate": 5.0409412364942165e-05,
+      "loss": 0.0606,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0961688819390147,
+      "grad_norm": 0.41264137625694275,
+      "learning_rate": 5.03411781357495e-05,
+      "loss": 0.0637,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0969507427677874,
+      "grad_norm": 0.3105151057243347,
+      "learning_rate": 5.027294327112314e-05,
+      "loss": 0.0486,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0977326035965598,
+      "grad_norm": 0.31089651584625244,
+      "learning_rate": 5.020470789814842e-05,
+      "loss": 0.055,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0985144644253322,
+      "grad_norm": 0.3997081518173218,
+      "learning_rate": 5.01364721439116e-05,
+      "loss": 0.0696,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0992963252541048,
+      "grad_norm": 0.4661003649234772,
+      "learning_rate": 5.006823613549964e-05,
+      "loss": 0.0712,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1000781860828772,
+      "grad_norm": 0.4002636969089508,
+      "learning_rate": 5e-05,
+      "loss": 0.0556,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1008600469116496,
+      "grad_norm": 0.402055025100708,
+      "learning_rate": 4.993176386450036e-05,
+      "loss": 0.0556,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1016419077404223,
+      "grad_norm": 0.4962114095687866,
+      "learning_rate": 4.986352785608842e-05,
+      "loss": 0.0714,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1024237685691947,
+      "grad_norm": 0.40611883997917175,
+      "learning_rate": 4.97952921018516e-05,
+      "loss": 0.0681,
+      "step": 1410
+    },
+    {
+      "epoch": 1.103205629397967,
+      "grad_norm": 0.5475478172302246,
+      "learning_rate": 4.9727056728876865e-05,
+      "loss": 0.1258,
+      "step": 1411
+    },
+    {
+      "epoch": 1.1039874902267397,
+      "grad_norm": 0.33019861578941345,
+      "learning_rate": 4.965882186425053e-05,
+      "loss": 0.0628,
+      "step": 1412
+    },
+    {
+      "epoch": 1.104769351055512,
+      "grad_norm": 0.2784503698348999,
+      "learning_rate": 4.959058763505784e-05,
+      "loss": 0.0396,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1055512118842845,
+      "grad_norm": 0.4385380446910858,
+      "learning_rate": 4.9522354168382973e-05,
+      "loss": 0.0941,
+      "step": 1414
+    },
+    {
+      "epoch": 1.1063330727130571,
+      "grad_norm": 0.33498090505599976,
+      "learning_rate": 4.945412159130864e-05,
+      "loss": 0.0423,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1071149335418295,
+      "grad_norm": 0.3903992474079132,
+      "learning_rate": 4.9385890030915907e-05,
+      "loss": 0.0581,
+      "step": 1416
+    },
+    {
+      "epoch": 1.107896794370602,
+      "grad_norm": 0.31511014699935913,
+      "learning_rate": 4.931765961428389e-05,
+      "loss": 0.063,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1086786551993746,
+      "grad_norm": 0.4896702170372009,
+      "learning_rate": 4.9249430468489646e-05,
+      "loss": 0.0814,
+      "step": 1418
+    },
+    {
+      "epoch": 1.109460516028147,
+      "grad_norm": 0.3954057991504669,
+      "learning_rate": 4.9181202720607874e-05,
+      "loss": 0.0563,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1102423768569194,
+      "grad_norm": 0.36815062165260315,
+      "learning_rate": 4.9112976497710586e-05,
+      "loss": 0.0694,
+      "step": 1420
+    },
+    {
+      "epoch": 1.111024237685692,
+      "grad_norm": 0.4068279564380646,
+      "learning_rate": 4.904475192686702e-05,
+      "loss": 0.0636,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1118060985144644,
+      "grad_norm": 0.42344993352890015,
+      "learning_rate": 4.897652913514334e-05,
+      "loss": 0.0479,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1125879593432368,
+      "grad_norm": 0.43622395396232605,
+      "learning_rate": 4.890830824960234e-05,
+      "loss": 0.0585,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1133698201720095,
+      "grad_norm": 0.2758655250072479,
+      "learning_rate": 4.884008939730333e-05,
+      "loss": 0.047,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1141516810007819,
+      "grad_norm": 0.39345523715019226,
+      "learning_rate": 4.877187270530178e-05,
+      "loss": 0.0497,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1149335418295543,
+      "grad_norm": 0.48883023858070374,
+      "learning_rate": 4.87036583006492e-05,
+      "loss": 0.0757,
+      "step": 1426
+    },
+    {
+      "epoch": 1.115715402658327,
+      "grad_norm": 0.34832096099853516,
+      "learning_rate": 4.863544631039275e-05,
+      "loss": 0.0448,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1164972634870993,
+      "grad_norm": 0.4675982594490051,
+      "learning_rate": 4.856723686157516e-05,
+      "loss": 0.0671,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1172791243158717,
+      "grad_norm": 0.37189602851867676,
+      "learning_rate": 4.8499030081234444e-05,
+      "loss": 0.0909,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1180609851446444,
+      "grad_norm": 0.3896830976009369,
+      "learning_rate": 4.843082609640357e-05,
+      "loss": 0.0654,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1188428459734168,
+      "grad_norm": 0.31397873163223267,
+      "learning_rate": 4.8362625034110354e-05,
+      "loss": 0.0473,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1196247068021892,
+      "grad_norm": 0.45029687881469727,
+      "learning_rate": 4.829442702137718e-05,
+      "loss": 0.0544,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1204065676309618,
+      "grad_norm": 0.3392208516597748,
+      "learning_rate": 4.8226232185220684e-05,
+      "loss": 0.043,
+      "step": 1433
+    },
+    {
+      "epoch": 1.1211884284597342,
+      "grad_norm": 0.4472123384475708,
+      "learning_rate": 4.815804065265166e-05,
+      "loss": 0.0761,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1219702892885066,
+      "grad_norm": 0.35860681533813477,
+      "learning_rate": 4.808985255067473e-05,
+      "loss": 0.0746,
+      "step": 1435
+    },
+    {
+      "epoch": 1.122752150117279,
+      "grad_norm": 0.4250498414039612,
+      "learning_rate": 4.8021668006288075e-05,
+      "loss": 0.0512,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1235340109460517,
+      "grad_norm": 0.3914888799190521,
+      "learning_rate": 4.7953487146483315e-05,
+      "loss": 0.0869,
+      "step": 1437
+    },
+    {
+      "epoch": 1.124315871774824,
+      "grad_norm": 0.544806957244873,
+      "learning_rate": 4.78853100982452e-05,
+      "loss": 0.0859,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1250977326035965,
+      "grad_norm": 0.4872533679008484,
+      "learning_rate": 4.781713698855135e-05,
+      "loss": 0.069,
+      "step": 1439
+    },
+    {
+      "epoch": 1.125879593432369,
+      "grad_norm": 0.41077733039855957,
+      "learning_rate": 4.774896794437206e-05,
+      "loss": 0.0658,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1266614542611415,
+      "grad_norm": 0.5026431083679199,
+      "learning_rate": 4.7680803092670054e-05,
+      "loss": 0.0768,
+      "step": 1441
+    },
+    {
+      "epoch": 1.127443315089914,
+      "grad_norm": 0.45740193128585815,
+      "learning_rate": 4.761264256040028e-05,
+      "loss": 0.0676,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1282251759186865,
+      "grad_norm": 0.3656143844127655,
+      "learning_rate": 4.754448647450957e-05,
+      "loss": 0.0688,
+      "step": 1443
+    },
+    {
+      "epoch": 1.129007036747459,
+      "grad_norm": 0.4475756883621216,
+      "learning_rate": 4.747633496193655e-05,
+      "loss": 0.0507,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1297888975762314,
+      "grad_norm": 0.33670148253440857,
+      "learning_rate": 4.740818814961129e-05,
+      "loss": 0.032,
+      "step": 1445
+    },
+    {
+      "epoch": 1.130570758405004,
+      "grad_norm": 0.5176746249198914,
+      "learning_rate": 4.7340046164455096e-05,
+      "loss": 0.0778,
+      "step": 1446
+    },
+    {
+      "epoch": 1.1313526192337764,
+      "grad_norm": 0.2944004237651825,
+      "learning_rate": 4.72719091333803e-05,
+      "loss": 0.081,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1321344800625488,
+      "grad_norm": 0.36607813835144043,
+      "learning_rate": 4.720377718329003e-05,
+      "loss": 0.062,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1329163408913214,
+      "grad_norm": 0.38473501801490784,
+      "learning_rate": 4.713565044107792e-05,
+      "loss": 0.0809,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1336982017200938,
+      "grad_norm": 0.37212830781936646,
+      "learning_rate": 4.7067529033627874e-05,
+      "loss": 0.0573,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1344800625488662,
+      "grad_norm": 0.35405033826828003,
+      "learning_rate": 4.699941308781394e-05,
+      "loss": 0.053,
+      "step": 1451
+    },
+    {
+      "epoch": 1.1352619233776389,
+      "grad_norm": 0.469467431306839,
+      "learning_rate": 4.693130273049995e-05,
+      "loss": 0.065,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1360437842064113,
+      "grad_norm": 0.37710899114608765,
+      "learning_rate": 4.686319808853931e-05,
+      "loss": 0.0766,
+      "step": 1453
+    },
+    {
+      "epoch": 1.1368256450351837,
+      "grad_norm": 0.49938127398490906,
+      "learning_rate": 4.679509928877481e-05,
+      "loss": 0.0463,
+      "step": 1454
+    },
+    {
+      "epoch": 1.137607505863956,
+      "grad_norm": 0.3544851839542389,
+      "learning_rate": 4.6727006458038357e-05,
+      "loss": 0.0523,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1383893666927287,
+      "grad_norm": 0.26120486855506897,
+      "learning_rate": 4.6658919723150706e-05,
+      "loss": 0.0476,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1391712275215011,
+      "grad_norm": 0.4471639096736908,
+      "learning_rate": 4.659083921092131e-05,
+      "loss": 0.076,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1399530883502735,
+      "grad_norm": 0.3284335434436798,
+      "learning_rate": 4.6522765048148015e-05,
+      "loss": 0.0456,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1407349491790462,
+      "grad_norm": 0.35543978214263916,
+      "learning_rate": 4.6454697361616816e-05,
+      "loss": 0.0578,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1415168100078186,
+      "grad_norm": 0.3903660476207733,
+      "learning_rate": 4.6386636278101665e-05,
+      "loss": 0.0421,
+      "step": 1460
+    },
+    {
+      "epoch": 1.142298670836591,
+      "grad_norm": 0.39665699005126953,
+      "learning_rate": 4.631858192436422e-05,
+      "loss": 0.0684,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1430805316653636,
+      "grad_norm": 0.44944366812705994,
+      "learning_rate": 4.6250534427153626e-05,
+      "loss": 0.0639,
+      "step": 1462
+    },
+    {
+      "epoch": 1.143862392494136,
+      "grad_norm": 0.3654981553554535,
+      "learning_rate": 4.6182493913206205e-05,
+      "loss": 0.0533,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1446442533229084,
+      "grad_norm": 0.38796553015708923,
+      "learning_rate": 4.611446050924529e-05,
+      "loss": 0.0969,
+      "step": 1464
+    },
+    {
+      "epoch": 1.145426114151681,
+      "grad_norm": 0.31265556812286377,
+      "learning_rate": 4.6046434341981034e-05,
+      "loss": 0.0347,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1462079749804535,
+      "grad_norm": 0.4331945776939392,
+      "learning_rate": 4.597841553810999e-05,
+      "loss": 0.0629,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1469898358092259,
+      "grad_norm": 0.2969951927661896,
+      "learning_rate": 4.591040422431512e-05,
+      "loss": 0.0649,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1477716966379985,
+      "grad_norm": 0.3629123270511627,
+      "learning_rate": 4.5842400527265386e-05,
+      "loss": 0.0794,
+      "step": 1468
+    },
+    {
+      "epoch": 1.148553557466771,
+      "grad_norm": 0.33106693625450134,
+      "learning_rate": 4.5774404573615534e-05,
+      "loss": 0.0347,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1493354182955433,
+      "grad_norm": 0.3500162363052368,
+      "learning_rate": 4.570641649000591e-05,
+      "loss": 0.031,
+      "step": 1470
+    },
+    {
+      "epoch": 1.150117279124316,
+      "grad_norm": 0.39276114106178284,
+      "learning_rate": 4.563843640306225e-05,
+      "loss": 0.0432,
+      "step": 1471
+    },
+    {
+      "epoch": 1.1508991399530883,
+      "grad_norm": 0.5838002562522888,
+      "learning_rate": 4.557046443939531e-05,
+      "loss": 0.1249,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1516810007818608,
+      "grad_norm": 0.5215751528739929,
+      "learning_rate": 4.550250072560079e-05,
+      "loss": 0.0673,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1524628616106334,
+      "grad_norm": 0.30646979808807373,
+      "learning_rate": 4.543454538825898e-05,
+      "loss": 0.0486,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1532447224394058,
+      "grad_norm": 0.4018358588218689,
+      "learning_rate": 4.5366598553934606e-05,
+      "loss": 0.0595,
+      "step": 1475
+    },
+    {
+      "epoch": 1.1540265832681782,
+      "grad_norm": 0.34716930985450745,
+      "learning_rate": 4.529866034917651e-05,
+      "loss": 0.0528,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1548084440969508,
+      "grad_norm": 0.46272578835487366,
+      "learning_rate": 4.523073090051751e-05,
+      "loss": 0.0574,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1555903049257232,
+      "grad_norm": 0.4649121165275574,
+      "learning_rate": 4.516281033447408e-05,
+      "loss": 0.0682,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1563721657544956,
+      "grad_norm": 0.43052923679351807,
+      "learning_rate": 4.5094898777546175e-05,
+      "loss": 0.0839,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1571540265832683,
+      "grad_norm": 0.3807370960712433,
+      "learning_rate": 4.5026996356216945e-05,
+      "loss": 0.0934,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1579358874120407,
+      "grad_norm": 0.3976602554321289,
+      "learning_rate": 4.495910319695257e-05,
+      "loss": 0.0474,
+      "step": 1481
+    },
+    {
+      "epoch": 1.158717748240813,
+      "grad_norm": 0.3352206349372864,
+      "learning_rate": 4.4891219426201914e-05,
+      "loss": 0.0811,
+      "step": 1482
+    },
+    {
+      "epoch": 1.1594996090695857,
+      "grad_norm": 0.2787337005138397,
+      "learning_rate": 4.4823345170396417e-05,
+      "loss": 0.0863,
+      "step": 1483
+    },
+    {
+      "epoch": 1.1602814698983581,
+      "grad_norm": 0.4110407531261444,
+      "learning_rate": 4.475548055594976e-05,
+      "loss": 0.061,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1610633307271305,
+      "grad_norm": 0.3591212034225464,
+      "learning_rate": 4.468762570925771e-05,
+      "loss": 0.0481,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1618451915559032,
+      "grad_norm": 0.48095250129699707,
+      "learning_rate": 4.461978075669775e-05,
+      "loss": 0.0568,
+      "step": 1486
+    },
+    {
+      "epoch": 1.1626270523846756,
+      "grad_norm": 0.4325644075870514,
+      "learning_rate": 4.455194582462906e-05,
+      "loss": 0.0506,
+      "step": 1487
+    },
+    {
+      "epoch": 1.163408913213448,
+      "grad_norm": 0.3393900692462921,
+      "learning_rate": 4.4484121039392074e-05,
+      "loss": 0.0424,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1641907740422206,
+      "grad_norm": 0.38791218400001526,
+      "learning_rate": 4.441630652730831e-05,
+      "loss": 0.0393,
+      "step": 1489
+    },
+    {
+      "epoch": 1.164972634870993,
+      "grad_norm": 0.4067123532295227,
+      "learning_rate": 4.4348502414680236e-05,
+      "loss": 0.0613,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1657544956997654,
+      "grad_norm": 0.3250060975551605,
+      "learning_rate": 4.428070882779091e-05,
+      "loss": 0.0518,
+      "step": 1491
+    },
+    {
+      "epoch": 1.166536356528538,
+      "grad_norm": 0.3601875901222229,
+      "learning_rate": 4.4212925892903745e-05,
+      "loss": 0.0491,
+      "step": 1492
+    },
+    {
+      "epoch": 1.1673182173573105,
+      "grad_norm": 0.42251959443092346,
+      "learning_rate": 4.4145153736262387e-05,
+      "loss": 0.0437,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1681000781860829,
+      "grad_norm": 0.3853507936000824,
+      "learning_rate": 4.407739248409036e-05,
+      "loss": 0.0344,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1688819390148553,
+      "grad_norm": 0.40151551365852356,
+      "learning_rate": 4.400964226259087e-05,
+      "loss": 0.0763,
+      "step": 1495
+    },
+    {
+      "epoch": 1.169663799843628,
+      "grad_norm": 0.2914046049118042,
+      "learning_rate": 4.3941903197946614e-05,
+      "loss": 0.0283,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1704456606724003,
+      "grad_norm": 0.3761674463748932,
+      "learning_rate": 4.387417541631952e-05,
+      "loss": 0.0524,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1712275215011727,
+      "grad_norm": 0.5237846374511719,
+      "learning_rate": 4.3806459043850465e-05,
+      "loss": 0.0456,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1720093823299453,
+      "grad_norm": 0.40119150280952454,
+      "learning_rate": 4.373875420665907e-05,
+      "loss": 0.0599,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1727912431587177,
+      "grad_norm": 0.4992891252040863,
+      "learning_rate": 4.3671061030843495e-05,
+      "loss": 0.0484,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1735731039874902,
+      "grad_norm": 0.45423582196235657,
+      "learning_rate": 4.3603379642480216e-05,
+      "loss": 0.0905,
+      "step": 1501
+    },
+    {
+      "epoch": 1.1743549648162628,
+      "grad_norm": 0.47261035442352295,
+      "learning_rate": 4.353571016762368e-05,
+      "loss": 0.0506,
+      "step": 1502
+    },
+    {
+      "epoch": 1.1751368256450352,
+      "grad_norm": 0.40774956345558167,
+      "learning_rate": 4.3468052732306184e-05,
+      "loss": 0.0532,
+      "step": 1503
+    },
+    {
+      "epoch": 1.1759186864738076,
+      "grad_norm": 0.4656347632408142,
+      "learning_rate": 4.340040746253762e-05,
+      "loss": 0.0451,
+      "step": 1504
+    },
+    {
+      "epoch": 1.1767005473025802,
+      "grad_norm": 0.2806099057197571,
+      "learning_rate": 4.333277448430517e-05,
+      "loss": 0.0388,
+      "step": 1505
+    },
+    {
+      "epoch": 1.1774824081313526,
+      "grad_norm": 0.37031981348991394,
+      "learning_rate": 4.3265153923573174e-05,
+      "loss": 0.0373,
+      "step": 1506
+    },
+    {
+      "epoch": 1.178264268960125,
+      "grad_norm": 0.4706210792064667,
+      "learning_rate": 4.3197545906282845e-05,
+      "loss": 0.0478,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1790461297888977,
+      "grad_norm": 0.4227662980556488,
+      "learning_rate": 4.3129950558352e-05,
+      "loss": 0.0451,
+      "step": 1508
+    },
+    {
+      "epoch": 1.17982799061767,
+      "grad_norm": 0.5277842283248901,
+      "learning_rate": 4.306236800567485e-05,
+      "loss": 0.0586,
+      "step": 1509
+    },
+    {
+      "epoch": 1.1806098514464425,
+      "grad_norm": 0.5246910452842712,
+      "learning_rate": 4.2994798374121834e-05,
+      "loss": 0.053,
+      "step": 1510
+    },
+    {
+      "epoch": 1.181391712275215,
+      "grad_norm": 0.4553016126155853,
+      "learning_rate": 4.292724178953929e-05,
+      "loss": 0.083,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1821735731039875,
+      "grad_norm": 0.36565476655960083,
+      "learning_rate": 4.2859698377749244e-05,
+      "loss": 0.0493,
+      "step": 1512
+    },
+    {
+      "epoch": 1.18295543393276,
+      "grad_norm": 0.38505685329437256,
+      "learning_rate": 4.27921682645492e-05,
+      "loss": 0.085,
+      "step": 1513
+    },
+    {
+      "epoch": 1.1837372947615323,
+      "grad_norm": 0.46367672085762024,
+      "learning_rate": 4.272465157571193e-05,
+      "loss": 0.0626,
+      "step": 1514
+    },
+    {
+      "epoch": 1.184519155590305,
+      "grad_norm": 0.34231388568878174,
+      "learning_rate": 4.265714843698511e-05,
+      "loss": 0.0646,
+      "step": 1515
+    },
+    {
+      "epoch": 1.1853010164190774,
+      "grad_norm": 0.3470911979675293,
+      "learning_rate": 4.258965897409129e-05,
+      "loss": 0.0529,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1860828772478498,
+      "grad_norm": 0.5022766590118408,
+      "learning_rate": 4.2522183312727496e-05,
+      "loss": 0.0774,
+      "step": 1517
+    },
+    {
+      "epoch": 1.1868647380766224,
+      "grad_norm": 0.40239912271499634,
+      "learning_rate": 4.2454721578565024e-05,
+      "loss": 0.0573,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1876465989053948,
+      "grad_norm": 0.4047490358352661,
+      "learning_rate": 4.238727389724927e-05,
+      "loss": 0.069,
+      "step": 1519
+    },
+    {
+      "epoch": 1.1884284597341672,
+      "grad_norm": 0.3190065622329712,
+      "learning_rate": 4.231984039439945e-05,
+      "loss": 0.0597,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1892103205629398,
+      "grad_norm": 0.38884586095809937,
+      "learning_rate": 4.225242119560839e-05,
+      "loss": 0.0433,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1899921813917123,
+      "grad_norm": 0.4333357512950897,
+      "learning_rate": 4.2185016426442214e-05,
+      "loss": 0.0714,
+      "step": 1522
+    },
+    {
+      "epoch": 1.1907740422204847,
+      "grad_norm": 0.4471852779388428,
+      "learning_rate": 4.211762621244022e-05,
+      "loss": 0.0771,
+      "step": 1523
+    },
+    {
+      "epoch": 1.1915559030492573,
+      "grad_norm": 0.3137734532356262,
+      "learning_rate": 4.2050250679114625e-05,
+      "loss": 0.0404,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1923377638780297,
+      "grad_norm": 0.3927496671676636,
+      "learning_rate": 4.198288995195021e-05,
+      "loss": 0.0459,
+      "step": 1525
+    },
+    {
+      "epoch": 1.193119624706802,
+      "grad_norm": 0.3484167754650116,
+      "learning_rate": 4.1915544156404284e-05,
+      "loss": 0.0555,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1939014855355747,
+      "grad_norm": 0.3492136299610138,
+      "learning_rate": 4.184821341790629e-05,
+      "loss": 0.0363,
+      "step": 1527
+    },
+    {
+      "epoch": 1.1946833463643471,
+      "grad_norm": 0.34315451979637146,
+      "learning_rate": 4.178089786185761e-05,
+      "loss": 0.0568,
+      "step": 1528
+    },
+    {
+      "epoch": 1.1954652071931196,
+      "grad_norm": 0.36132681369781494,
+      "learning_rate": 4.17135976136314e-05,
+      "loss": 0.0434,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1962470680218922,
+      "grad_norm": 0.6220799088478088,
+      "learning_rate": 4.1646312798572296e-05,
+      "loss": 0.0789,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1970289288506646,
+      "grad_norm": 0.3326771855354309,
+      "learning_rate": 4.1579043541996136e-05,
+      "loss": 0.0583,
+      "step": 1531
+    },
+    {
+      "epoch": 1.197810789679437,
+      "grad_norm": 0.5193848609924316,
+      "learning_rate": 4.151178996918986e-05,
+      "loss": 0.0977,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1985926505082096,
+      "grad_norm": 0.3856359124183655,
+      "learning_rate": 4.144455220541112e-05,
+      "loss": 0.0419,
+      "step": 1533
+    },
+    {
+      "epoch": 1.199374511336982,
+      "grad_norm": 0.3599039614200592,
+      "learning_rate": 4.137733037588822e-05,
+      "loss": 0.0917,
+      "step": 1534
+    },
+    {
+      "epoch": 1.2001563721657544,
+      "grad_norm": 0.3522965908050537,
+      "learning_rate": 4.131012460581967e-05,
+      "loss": 0.0446,
+      "step": 1535
+    },
+    {
+      "epoch": 1.200938232994527,
+      "grad_norm": 0.38580793142318726,
+      "learning_rate": 4.1242935020374165e-05,
+      "loss": 0.062,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2017200938232995,
+      "grad_norm": 0.35813602805137634,
+      "learning_rate": 4.1175761744690225e-05,
+      "loss": 0.0825,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2025019546520719,
+      "grad_norm": 0.38828739523887634,
+      "learning_rate": 4.1108604903875945e-05,
+      "loss": 0.0868,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2032838154808445,
+      "grad_norm": 0.40260881185531616,
+      "learning_rate": 4.10414646230089e-05,
+      "loss": 0.0488,
+      "step": 1539
+    },
+    {
+      "epoch": 1.204065676309617,
+      "grad_norm": 0.33392858505249023,
+      "learning_rate": 4.097434102713577e-05,
+      "loss": 0.0438,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2048475371383893,
+      "grad_norm": 0.4407254159450531,
+      "learning_rate": 4.090723424127216e-05,
+      "loss": 0.0632,
+      "step": 1541
+    },
+    {
+      "epoch": 1.205629397967162,
+      "grad_norm": 0.5629727244377136,
+      "learning_rate": 4.084014439040236e-05,
+      "loss": 0.0962,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2064112587959344,
+      "grad_norm": 0.3747621476650238,
+      "learning_rate": 4.077307159947915e-05,
+      "loss": 0.0482,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2071931196247068,
+      "grad_norm": 0.424283891916275,
+      "learning_rate": 4.070601599342354e-05,
+      "loss": 0.054,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2079749804534794,
+      "grad_norm": 0.39465880393981934,
+      "learning_rate": 4.063897769712448e-05,
+      "loss": 0.0419,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2087568412822518,
+      "grad_norm": 0.30999326705932617,
+      "learning_rate": 4.057195683543873e-05,
+      "loss": 0.0436,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2095387021110242,
+      "grad_norm": 0.4226630628108978,
+      "learning_rate": 4.0504953533190575e-05,
+      "loss": 0.0725,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2103205629397968,
+      "grad_norm": 0.31118202209472656,
+      "learning_rate": 4.043796791517155e-05,
+      "loss": 0.0439,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2111024237685692,
+      "grad_norm": 0.36616459488868713,
+      "learning_rate": 4.037100010614031e-05,
+      "loss": 0.0534,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2118842845973417,
+      "grad_norm": 0.47533512115478516,
+      "learning_rate": 4.0304050230822335e-05,
+      "loss": 0.0636,
+      "step": 1550
+    },
+    {
+      "epoch": 1.212666145426114,
+      "grad_norm": 0.396884560585022,
+      "learning_rate": 4.023711841390966e-05,
+      "loss": 0.0996,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2134480062548867,
+      "grad_norm": 0.3450510501861572,
+      "learning_rate": 4.017020478006073e-05,
+      "loss": 0.035,
+      "step": 1552
+    },
+    {
+      "epoch": 1.214229867083659,
+      "grad_norm": 0.48043736815452576,
+      "learning_rate": 4.010330945390012e-05,
+      "loss": 0.0527,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2150117279124315,
+      "grad_norm": 0.4231407940387726,
+      "learning_rate": 4.0036432560018255e-05,
+      "loss": 0.0549,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2157935887412041,
+      "grad_norm": 0.3957667648792267,
+      "learning_rate": 3.996957422297132e-05,
+      "loss": 0.0545,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2165754495699765,
+      "grad_norm": 0.4996613562107086,
+      "learning_rate": 3.990273456728086e-05,
+      "loss": 0.0701,
+      "step": 1556
+    },
+    {
+      "epoch": 1.217357310398749,
+      "grad_norm": 0.42404982447624207,
+      "learning_rate": 3.98359137174337e-05,
+      "loss": 0.0573,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2181391712275216,
+      "grad_norm": 0.30868658423423767,
+      "learning_rate": 3.9769111797881545e-05,
+      "loss": 0.0423,
+      "step": 1558
+    },
+    {
+      "epoch": 1.218921032056294,
+      "grad_norm": 0.40342530608177185,
+      "learning_rate": 3.9702328933040914e-05,
+      "loss": 0.0433,
+      "step": 1559
+    },
+    {
+      "epoch": 1.2197028928850664,
+      "grad_norm": 0.3019714653491974,
+      "learning_rate": 3.963556524729284e-05,
+      "loss": 0.0413,
+      "step": 1560
+    },
+    {
+      "epoch": 1.220484753713839,
+      "grad_norm": 0.4296371042728424,
+      "learning_rate": 3.9568820864982606e-05,
+      "loss": 0.0688,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2212666145426114,
+      "grad_norm": 0.2896515130996704,
+      "learning_rate": 3.9502095910419515e-05,
+      "loss": 0.0657,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2220484753713838,
+      "grad_norm": 0.38489270210266113,
+      "learning_rate": 3.943539050787678e-05,
+      "loss": 0.0783,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2228303362001565,
+      "grad_norm": 0.3944641351699829,
+      "learning_rate": 3.936870478159109e-05,
+      "loss": 0.0469,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2236121970289289,
+      "grad_norm": 0.3929797410964966,
+      "learning_rate": 3.930203885576257e-05,
+      "loss": 0.0563,
+      "step": 1565
+    },
+    {
+      "epoch": 1.2243940578577013,
+      "grad_norm": 0.6391537189483643,
+      "learning_rate": 3.923539285455442e-05,
+      "loss": 0.0864,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2251759186864737,
+      "grad_norm": 0.43830859661102295,
+      "learning_rate": 3.9168766902092776e-05,
+      "loss": 0.07,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2259577795152463,
+      "grad_norm": 0.35931962728500366,
+      "learning_rate": 3.910216112246636e-05,
+      "loss": 0.0723,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2267396403440187,
+      "grad_norm": 0.3321268856525421,
+      "learning_rate": 3.903557563972641e-05,
+      "loss": 0.0599,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2275215011727911,
+      "grad_norm": 0.33983114361763,
+      "learning_rate": 3.89690105778863e-05,
+      "loss": 0.0489,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2283033620015638,
+      "grad_norm": 0.36031773686408997,
+      "learning_rate": 3.890246606092139e-05,
+      "loss": 0.0968,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2290852228303362,
+      "grad_norm": 0.3760223388671875,
+      "learning_rate": 3.883594221276877e-05,
+      "loss": 0.0501,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2298670836591086,
+      "grad_norm": 0.3305293023586273,
+      "learning_rate": 3.876943915732706e-05,
+      "loss": 0.0593,
+      "step": 1573
+    },
+    {
+      "epoch": 1.2306489444878812,
+      "grad_norm": 0.32184192538261414,
+      "learning_rate": 3.8702957018456124e-05,
+      "loss": 0.0507,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2314308053166536,
+      "grad_norm": 0.6274586915969849,
+      "learning_rate": 3.863649591997688e-05,
+      "loss": 0.107,
+      "step": 1575
+    },
+    {
+      "epoch": 1.232212666145426,
+      "grad_norm": 0.4004760980606079,
+      "learning_rate": 3.857005598567107e-05,
+      "loss": 0.064,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2329945269741986,
+      "grad_norm": 0.47167232632637024,
+      "learning_rate": 3.850363733928098e-05,
+      "loss": 0.0611,
+      "step": 1577
+    },
+    {
+      "epoch": 1.233776387802971,
+      "grad_norm": 0.41693300008773804,
+      "learning_rate": 3.84372401045093e-05,
+      "loss": 0.0771,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2345582486317435,
+      "grad_norm": 0.32655128836631775,
+      "learning_rate": 3.8370864405018816e-05,
+      "loss": 0.0345,
+      "step": 1579
+    },
+    {
+      "epoch": 1.235340109460516,
+      "grad_norm": 0.34897708892822266,
+      "learning_rate": 3.8304510364432225e-05,
+      "loss": 0.0551,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2361219702892885,
+      "grad_norm": 0.3144432306289673,
+      "learning_rate": 3.823817810633181e-05,
+      "loss": 0.0554,
+      "step": 1581
+    },
+    {
+      "epoch": 1.236903831118061,
+      "grad_norm": 0.3374468684196472,
+      "learning_rate": 3.8171867754259396e-05,
+      "loss": 0.0503,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2376856919468335,
+      "grad_norm": 0.6711187958717346,
+      "learning_rate": 3.810557943171594e-05,
+      "loss": 0.0912,
+      "step": 1583
+    },
+    {
+      "epoch": 1.238467552775606,
+      "grad_norm": 0.3551695644855499,
+      "learning_rate": 3.803931326216137e-05,
+      "loss": 0.0482,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2392494136043783,
+      "grad_norm": 0.29928022623062134,
+      "learning_rate": 3.797306936901439e-05,
+      "loss": 0.0591,
+      "step": 1585
+    },
+    {
+      "epoch": 1.240031274433151,
+      "grad_norm": 0.29246705770492554,
+      "learning_rate": 3.790684787565219e-05,
+      "loss": 0.0671,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2408131352619234,
+      "grad_norm": 0.442971795797348,
+      "learning_rate": 3.784064890541022e-05,
+      "loss": 0.0714,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2415949960906958,
+      "grad_norm": 0.4673009216785431,
+      "learning_rate": 3.777447258158201e-05,
+      "loss": 0.0862,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2423768569194684,
+      "grad_norm": 0.3445381224155426,
+      "learning_rate": 3.770831902741895e-05,
+      "loss": 0.0425,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2431587177482408,
+      "grad_norm": 0.43028736114501953,
+      "learning_rate": 3.7642188366129935e-05,
+      "loss": 0.085,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2439405785770132,
+      "grad_norm": 0.45845282077789307,
+      "learning_rate": 3.757608072088129e-05,
+      "loss": 0.0623,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2447224394057859,
+      "grad_norm": 0.3896902799606323,
+      "learning_rate": 3.750999621479643e-05,
+      "loss": 0.0776,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2455043002345583,
+      "grad_norm": 0.38196709752082825,
+      "learning_rate": 3.744393497095573e-05,
+      "loss": 0.0466,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2462861610633307,
+      "grad_norm": 0.34162676334381104,
+      "learning_rate": 3.7377897112396175e-05,
+      "loss": 0.0342,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2470680218921033,
+      "grad_norm": 0.4113738238811493,
+      "learning_rate": 3.7311882762111235e-05,
+      "loss": 0.0761,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2478498827208757,
+      "grad_norm": 0.5574212670326233,
+      "learning_rate": 3.72458920430506e-05,
+      "loss": 0.0513,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2486317435496481,
+      "grad_norm": 0.3796621561050415,
+      "learning_rate": 3.7179925078119905e-05,
+      "loss": 0.0641,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2494136043784208,
+      "grad_norm": 0.6831467151641846,
+      "learning_rate": 3.711398199018059e-05,
+      "loss": 0.0753,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2501954652071932,
+      "grad_norm": 0.3368784487247467,
+      "learning_rate": 3.704806290204963e-05,
+      "loss": 0.0386,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2509773260359656,
+      "grad_norm": 0.3436744809150696,
+      "learning_rate": 3.698216793649925e-05,
+      "loss": 0.0548,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2509773260359656,
+      "eval_loss": 0.12688221037387848,
+      "eval_runtime": 13.396,
+      "eval_samples_per_second": 3.882,
+      "eval_steps_per_second": 0.97,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2517591868647382,
+      "grad_norm": 0.3372763693332672,
+      "learning_rate": 3.691629721625677e-05,
+      "loss": 0.0867,
+      "step": 1601
+    },
+    {
+      "epoch": 1.2525410476935106,
+      "grad_norm": 0.39428460597991943,
+      "learning_rate": 3.6850450864004376e-05,
+      "loss": 0.0566,
+      "step": 1602
+    },
+    {
+      "epoch": 1.253322908522283,
+      "grad_norm": 0.39720073342323303,
+      "learning_rate": 3.678462900237888e-05,
+      "loss": 0.0585,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2541047693510556,
+      "grad_norm": 0.5266603827476501,
+      "learning_rate": 3.671883175397142e-05,
+      "loss": 0.0686,
+      "step": 1604
+    },
+    {
+      "epoch": 1.254886630179828,
+      "grad_norm": 0.3272291123867035,
+      "learning_rate": 3.6653059241327314e-05,
+      "loss": 0.0601,
+      "step": 1605
+    },
+    {
+      "epoch": 1.2556684910086005,
+      "grad_norm": 0.5012699961662292,
+      "learning_rate": 3.6587311586945875e-05,
+      "loss": 0.0562,
+      "step": 1606
+    },
+    {
+      "epoch": 1.256450351837373,
+      "grad_norm": 0.3750763237476349,
+      "learning_rate": 3.6521588913280024e-05,
+      "loss": 0.0462,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2572322126661455,
+      "grad_norm": 0.5167805552482605,
+      "learning_rate": 3.6455891342736215e-05,
+      "loss": 0.066,
+      "step": 1608
+    },
+    {
+      "epoch": 1.258014073494918,
+      "grad_norm": 0.31887608766555786,
+      "learning_rate": 3.6390218997674144e-05,
+      "loss": 0.0815,
+      "step": 1609
+    },
+    {
+      "epoch": 1.2587959343236903,
+      "grad_norm": 0.3656921684741974,
+      "learning_rate": 3.6324572000406486e-05,
+      "loss": 0.0558,
+      "step": 1610
+    },
+    {
+      "epoch": 1.259577795152463,
+      "grad_norm": 0.4891717731952667,
+      "learning_rate": 3.625895047319875e-05,
+      "loss": 0.0633,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2603596559812353,
+      "grad_norm": 0.3321298658847809,
+      "learning_rate": 3.619335453826901e-05,
+      "loss": 0.0392,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2611415168100077,
+      "grad_norm": 0.3826375901699066,
+      "learning_rate": 3.6127784317787625e-05,
+      "loss": 0.0726,
+      "step": 1613
+    },
+    {
+      "epoch": 1.2619233776387804,
+      "grad_norm": 0.5942705869674683,
+      "learning_rate": 3.6062239933877116e-05,
+      "loss": 0.0724,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2627052384675528,
+      "grad_norm": 0.44333893060684204,
+      "learning_rate": 3.5996721508611854e-05,
+      "loss": 0.0612,
+      "step": 1615
+    },
+    {
+      "epoch": 1.2634870992963252,
+      "grad_norm": 0.4055474102497101,
+      "learning_rate": 3.5931229164017886e-05,
+      "loss": 0.0503,
+      "step": 1616
+    },
+    {
+      "epoch": 1.2642689601250976,
+      "grad_norm": 0.34622856974601746,
+      "learning_rate": 3.586576302207264e-05,
+      "loss": 0.0618,
+      "step": 1617
+    },
+    {
+      "epoch": 1.2650508209538702,
+      "grad_norm": 0.33253592252731323,
+      "learning_rate": 3.58003232047048e-05,
+      "loss": 0.032,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2658326817826426,
+      "grad_norm": 0.30954334139823914,
+      "learning_rate": 3.573490983379399e-05,
+      "loss": 0.0387,
+      "step": 1619
+    },
+    {
+      "epoch": 1.266614542611415,
+      "grad_norm": 0.38260695338249207,
+      "learning_rate": 3.566952303117056e-05,
+      "loss": 0.0829,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2673964034401877,
+      "grad_norm": 0.3483465313911438,
+      "learning_rate": 3.5604162918615413e-05,
+      "loss": 0.0441,
+      "step": 1621
+    },
+    {
+      "epoch": 1.26817826426896,
+      "grad_norm": 0.49197712540626526,
+      "learning_rate": 3.5538829617859746e-05,
+      "loss": 0.0676,
+      "step": 1622
+    },
+    {
+      "epoch": 1.2689601250977325,
+      "grad_norm": 0.29698431491851807,
+      "learning_rate": 3.5473523250584765e-05,
+      "loss": 0.0313,
+      "step": 1623
+    },
+    {
+      "epoch": 1.2697419859265051,
+      "grad_norm": 0.4328691363334656,
+      "learning_rate": 3.5408243938421584e-05,
+      "loss": 0.0637,
+      "step": 1624
+    },
+    {
+      "epoch": 1.2705238467552775,
+      "grad_norm": 0.41575777530670166,
+      "learning_rate": 3.534299180295087e-05,
+      "loss": 0.0948,
+      "step": 1625
+    },
+    {
+      "epoch": 1.27130570758405,
+      "grad_norm": 0.42420047521591187,
+      "learning_rate": 3.5277766965702734e-05,
+      "loss": 0.0786,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2720875684128226,
+      "grad_norm": 0.3792000114917755,
+      "learning_rate": 3.5212569548156366e-05,
+      "loss": 0.0797,
+      "step": 1627
+    },
+    {
+      "epoch": 1.272869429241595,
+      "grad_norm": 0.2699577808380127,
+      "learning_rate": 3.514739967173996e-05,
+      "loss": 0.0615,
+      "step": 1628
+    },
+    {
+      "epoch": 1.2736512900703674,
+      "grad_norm": 0.37993887066841125,
+      "learning_rate": 3.508225745783039e-05,
+      "loss": 0.0734,
+      "step": 1629
+    },
+    {
+      "epoch": 1.27443315089914,
+      "grad_norm": 0.3054560720920563,
+      "learning_rate": 3.501714302775297e-05,
+      "loss": 0.0452,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2752150117279124,
+      "grad_norm": 0.33489546179771423,
+      "learning_rate": 3.4952056502781336e-05,
+      "loss": 0.0378,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2759968725566848,
+      "grad_norm": 0.365928053855896,
+      "learning_rate": 3.488699800413714e-05,
+      "loss": 0.0532,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2767787333854574,
+      "grad_norm": 0.3355133831501007,
+      "learning_rate": 3.4821967652989764e-05,
+      "loss": 0.0428,
+      "step": 1633
+    },
+    {
+      "epoch": 1.2775605942142298,
+      "grad_norm": 0.3714124262332916,
+      "learning_rate": 3.4756965570456236e-05,
+      "loss": 0.0655,
+      "step": 1634
+    },
+    {
+      "epoch": 1.2783424550430023,
+      "grad_norm": 0.3428504765033722,
+      "learning_rate": 3.469199187760094e-05,
+      "loss": 0.0665,
+      "step": 1635
+    },
+    {
+      "epoch": 1.2791243158717749,
+      "grad_norm": 0.36663559079170227,
+      "learning_rate": 3.462704669543532e-05,
+      "loss": 0.0497,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2799061767005473,
+      "grad_norm": 0.38190585374832153,
+      "learning_rate": 3.456213014491778e-05,
+      "loss": 0.0827,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2806880375293197,
+      "grad_norm": 0.3864405155181885,
+      "learning_rate": 3.449724234695339e-05,
+      "loss": 0.0711,
+      "step": 1638
+    },
+    {
+      "epoch": 1.2814698983580923,
+      "grad_norm": 0.40252625942230225,
+      "learning_rate": 3.4432383422393666e-05,
+      "loss": 0.0581,
+      "step": 1639
+    },
+    {
+      "epoch": 1.2822517591868647,
+      "grad_norm": 0.4013853967189789,
+      "learning_rate": 3.43675534920363e-05,
+      "loss": 0.0414,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2830336200156371,
+      "grad_norm": 0.37079882621765137,
+      "learning_rate": 3.430275267662505e-05,
+      "loss": 0.0458,
+      "step": 1641
+    },
+    {
+      "epoch": 1.2838154808444098,
+      "grad_norm": 0.4159170091152191,
+      "learning_rate": 3.4237981096849443e-05,
+      "loss": 0.0513,
+      "step": 1642
+    },
+    {
+      "epoch": 1.2845973416731822,
+      "grad_norm": 0.34236016869544983,
+      "learning_rate": 3.41732388733445e-05,
+      "loss": 0.0569,
+      "step": 1643
+    },
+    {
+      "epoch": 1.2853792025019546,
+      "grad_norm": 0.4664420485496521,
+      "learning_rate": 3.41085261266906e-05,
+      "loss": 0.0694,
+      "step": 1644
+    },
+    {
+      "epoch": 1.2861610633307272,
+      "grad_norm": 0.4391569495201111,
+      "learning_rate": 3.404384297741326e-05,
+      "loss": 0.0553,
+      "step": 1645
+    },
+    {
+      "epoch": 1.2869429241594996,
+      "grad_norm": 0.4465954899787903,
+      "learning_rate": 3.39791895459828e-05,
+      "loss": 0.0657,
+      "step": 1646
+    },
+    {
+      "epoch": 1.287724784988272,
+      "grad_norm": 0.32883110642433167,
+      "learning_rate": 3.3914565952814237e-05,
+      "loss": 0.0649,
+      "step": 1647
+    },
+    {
+      "epoch": 1.2885066458170447,
+      "grad_norm": 0.4377223551273346,
+      "learning_rate": 3.384997231826702e-05,
+      "loss": 0.0604,
+      "step": 1648
+    },
+    {
+      "epoch": 1.289288506645817,
+      "grad_norm": 0.4015994369983673,
+      "learning_rate": 3.378540876264479e-05,
+      "loss": 0.0699,
+      "step": 1649
+    },
+    {
+      "epoch": 1.2900703674745895,
+      "grad_norm": 0.46162915229797363,
+      "learning_rate": 3.372087540619513e-05,
+      "loss": 0.0503,
+      "step": 1650
+    },
+    {
+      "epoch": 1.290852228303362,
+      "grad_norm": 0.5002314448356628,
+      "learning_rate": 3.365637236910944e-05,
+      "loss": 0.07,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2916340891321345,
+      "grad_norm": 0.4243912994861603,
+      "learning_rate": 3.359189977152264e-05,
+      "loss": 0.0558,
+      "step": 1652
+    },
+    {
+      "epoch": 1.292415949960907,
+      "grad_norm": 0.32526394724845886,
+      "learning_rate": 3.35274577335129e-05,
+      "loss": 0.049,
+      "step": 1653
+    },
+    {
+      "epoch": 1.2931978107896795,
+      "grad_norm": 0.4108484089374542,
+      "learning_rate": 3.346304637510153e-05,
+      "loss": 0.0648,
+      "step": 1654
+    },
+    {
+      "epoch": 1.293979671618452,
+      "grad_norm": 0.4533168077468872,
+      "learning_rate": 3.3398665816252705e-05,
+      "loss": 0.1103,
+      "step": 1655
+    },
+    {
+      "epoch": 1.2947615324472244,
+      "grad_norm": 0.3625294864177704,
+      "learning_rate": 3.333431617687318e-05,
+      "loss": 0.0387,
+      "step": 1656
+    },
+    {
+      "epoch": 1.295543393275997,
+      "grad_norm": 0.32721203565597534,
+      "learning_rate": 3.326999757681218e-05,
+      "loss": 0.0432,
+      "step": 1657
+    },
+    {
+      "epoch": 1.2963252541047694,
+      "grad_norm": 0.31126266717910767,
+      "learning_rate": 3.320571013586111e-05,
+      "loss": 0.0412,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2971071149335418,
+      "grad_norm": 0.4804350435733795,
+      "learning_rate": 3.3141453973753324e-05,
+      "loss": 0.0841,
+      "step": 1659
+    },
+    {
+      "epoch": 1.2978889757623144,
+      "grad_norm": 0.38014477491378784,
+      "learning_rate": 3.3077229210163906e-05,
+      "loss": 0.0645,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2986708365910868,
+      "grad_norm": 0.41697919368743896,
+      "learning_rate": 3.301303596470951e-05,
+      "loss": 0.0868,
+      "step": 1661
+    },
+    {
+      "epoch": 1.2994526974198592,
+      "grad_norm": 0.3734145760536194,
+      "learning_rate": 3.2948874356948074e-05,
+      "loss": 0.0526,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3002345582486319,
+      "grad_norm": 0.3617169260978699,
+      "learning_rate": 3.288474450637857e-05,
+      "loss": 0.0427,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3010164190774043,
+      "grad_norm": 0.43751880526542664,
+      "learning_rate": 3.2820646532440846e-05,
+      "loss": 0.0608,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3017982799061767,
+      "grad_norm": 0.3905401825904846,
+      "learning_rate": 3.2756580554515434e-05,
+      "loss": 0.1049,
+      "step": 1665
+    },
+    {
+      "epoch": 1.302580140734949,
+      "grad_norm": 0.5550699234008789,
+      "learning_rate": 3.2692546691923174e-05,
+      "loss": 0.0957,
+      "step": 1666
+    },
+    {
+      "epoch": 1.3033620015637217,
+      "grad_norm": 0.47704434394836426,
+      "learning_rate": 3.262854506392519e-05,
+      "loss": 0.0875,
+      "step": 1667
+    },
+    {
+      "epoch": 1.3041438623924941,
+      "grad_norm": 0.44277340173721313,
+      "learning_rate": 3.256457578972251e-05,
+      "loss": 0.0608,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3049257232212665,
+      "grad_norm": 0.30996251106262207,
+      "learning_rate": 3.25006389884559e-05,
+      "loss": 0.029,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3057075840500392,
+      "grad_norm": 0.4079422652721405,
+      "learning_rate": 3.243673477920569e-05,
+      "loss": 0.0435,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3064894448788116,
+      "grad_norm": 0.34746474027633667,
+      "learning_rate": 3.237286328099149e-05,
+      "loss": 0.0406,
+      "step": 1671
+    },
+    {
+      "epoch": 1.307271305707584,
+      "grad_norm": 0.3626313805580139,
+      "learning_rate": 3.230902461277196e-05,
+      "loss": 0.0358,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3080531665363564,
+      "grad_norm": 0.40878042578697205,
+      "learning_rate": 3.2245218893444664e-05,
+      "loss": 0.0539,
+      "step": 1673
+    },
+    {
+      "epoch": 1.308835027365129,
+      "grad_norm": 0.6351624727249146,
+      "learning_rate": 3.218144624184574e-05,
+      "loss": 0.0833,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3096168881939014,
+      "grad_norm": 0.40278536081314087,
+      "learning_rate": 3.2117706776749815e-05,
+      "loss": 0.0458,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3103987490226738,
+      "grad_norm": 0.3833138942718506,
+      "learning_rate": 3.205400061686961e-05,
+      "loss": 0.0441,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3111806098514465,
+      "grad_norm": 0.3146219551563263,
+      "learning_rate": 3.19903278808559e-05,
+      "loss": 0.0815,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3119624706802189,
+      "grad_norm": 0.38004133105278015,
+      "learning_rate": 3.192668868729719e-05,
+      "loss": 0.0356,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3127443315089913,
+      "grad_norm": 0.41840022802352905,
+      "learning_rate": 3.186308315471947e-05,
+      "loss": 0.0433,
+      "step": 1679
+    },
+    {
+      "epoch": 1.313526192337764,
+      "grad_norm": 0.5004205107688904,
+      "learning_rate": 3.179951140158608e-05,
+      "loss": 0.0702,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3143080531665363,
+      "grad_norm": 0.39277809858322144,
+      "learning_rate": 3.173597354629746e-05,
+      "loss": 0.0623,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3150899139953087,
+      "grad_norm": 0.41982904076576233,
+      "learning_rate": 3.1672469707190864e-05,
+      "loss": 0.0658,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3158717748240814,
+      "grad_norm": 0.41825437545776367,
+      "learning_rate": 3.1609000002540226e-05,
+      "loss": 0.0576,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3166536356528538,
+      "grad_norm": 0.391253262758255,
+      "learning_rate": 3.15455645505559e-05,
+      "loss": 0.0411,
+      "step": 1684
+    },
+    {
+      "epoch": 1.3174354964816262,
+      "grad_norm": 0.587203323841095,
+      "learning_rate": 3.148216346938446e-05,
+      "loss": 0.0591,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3182173573103988,
+      "grad_norm": 0.39162901043891907,
+      "learning_rate": 3.141879687710841e-05,
+      "loss": 0.0608,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3189992181391712,
+      "grad_norm": 0.4081566035747528,
+      "learning_rate": 3.135546489174612e-05,
+      "loss": 0.0481,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3197810789679436,
+      "grad_norm": 0.35251811146736145,
+      "learning_rate": 3.1292167631251414e-05,
+      "loss": 0.0419,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3205629397967162,
+      "grad_norm": 0.45182615518569946,
+      "learning_rate": 3.122890521351345e-05,
+      "loss": 0.0776,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3213448006254886,
+      "grad_norm": 0.39525046944618225,
+      "learning_rate": 3.116567775635654e-05,
+      "loss": 0.0627,
+      "step": 1690
+    },
+    {
+      "epoch": 1.322126661454261,
+      "grad_norm": 0.5914624929428101,
+      "learning_rate": 3.1102485377539886e-05,
+      "loss": 0.0611,
+      "step": 1691
+    },
+    {
+      "epoch": 1.3229085222830337,
+      "grad_norm": 0.43132856488227844,
+      "learning_rate": 3.103932819475729e-05,
+      "loss": 0.0544,
+      "step": 1692
+    },
+    {
+      "epoch": 1.323690383111806,
+      "grad_norm": 0.4183245301246643,
+      "learning_rate": 3.0976206325637046e-05,
+      "loss": 0.0684,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3244722439405785,
+      "grad_norm": 0.4382965564727783,
+      "learning_rate": 3.0913119887741716e-05,
+      "loss": 0.0352,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3252541047693511,
+      "grad_norm": 0.5051260590553284,
+      "learning_rate": 3.085006899856777e-05,
+      "loss": 0.0728,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3260359655981235,
+      "grad_norm": 0.39223799109458923,
+      "learning_rate": 3.078705377554558e-05,
+      "loss": 0.0612,
+      "step": 1696
+    },
+    {
+      "epoch": 1.326817826426896,
+      "grad_norm": 0.44544515013694763,
+      "learning_rate": 3.072407433603901e-05,
+      "loss": 0.0232,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3275996872556686,
+      "grad_norm": 0.540951669216156,
+      "learning_rate": 3.066113079734535e-05,
+      "loss": 0.0704,
+      "step": 1698
+    },
+    {
+      "epoch": 1.328381548084441,
+      "grad_norm": 0.3875698149204254,
+      "learning_rate": 3.059822327669494e-05,
+      "loss": 0.0508,
+      "step": 1699
+    },
+    {
+      "epoch": 1.3291634089132134,
+      "grad_norm": 0.4373841881752014,
+      "learning_rate": 3.0535351891251094e-05,
+      "loss": 0.0275,
+      "step": 1700
+    },
+    {
+      "epoch": 1.329945269741986,
+      "grad_norm": 0.5520462393760681,
+      "learning_rate": 3.0472516758109847e-05,
+      "loss": 0.0665,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3307271305707584,
+      "grad_norm": 0.5406703352928162,
+      "learning_rate": 3.040971799429966e-05,
+      "loss": 0.0559,
+      "step": 1702
+    },
+    {
+      "epoch": 1.3315089913995308,
+      "grad_norm": 0.4071379005908966,
+      "learning_rate": 3.034695571678127e-05,
+      "loss": 0.0572,
+      "step": 1703
+    },
+    {
+      "epoch": 1.3322908522283035,
+      "grad_norm": 0.37903717160224915,
+      "learning_rate": 3.0284230042447493e-05,
+      "loss": 0.046,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3330727130570759,
+      "grad_norm": 0.4394493103027344,
+      "learning_rate": 3.0221541088122917e-05,
+      "loss": 0.0527,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3338545738858483,
+      "grad_norm": 0.404144287109375,
+      "learning_rate": 3.015888897056381e-05,
+      "loss": 0.0903,
+      "step": 1706
+    },
+    {
+      "epoch": 1.334636434714621,
+      "grad_norm": 0.453784704208374,
+      "learning_rate": 3.0096273806457763e-05,
+      "loss": 0.0466,
+      "step": 1707
+    },
+    {
+      "epoch": 1.3354182955433933,
+      "grad_norm": 0.4025675654411316,
+      "learning_rate": 3.0033695712423603e-05,
+      "loss": 0.119,
+      "step": 1708
+    },
+    {
+      "epoch": 1.3362001563721657,
+      "grad_norm": 0.4505664110183716,
+      "learning_rate": 2.997115480501106e-05,
+      "loss": 0.092,
+      "step": 1709
+    },
+    {
+      "epoch": 1.3369820172009383,
+      "grad_norm": 0.4255506992340088,
+      "learning_rate": 2.9908651200700655e-05,
+      "loss": 0.0506,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3377638780297108,
+      "grad_norm": 0.30817708373069763,
+      "learning_rate": 2.9846185015903394e-05,
+      "loss": 0.0429,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3385457388584832,
+      "grad_norm": 0.3303023874759674,
+      "learning_rate": 2.9783756366960613e-05,
+      "loss": 0.0814,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3393275996872558,
+      "grad_norm": 0.4083544909954071,
+      "learning_rate": 2.9721365370143722e-05,
+      "loss": 0.1116,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3401094605160282,
+      "grad_norm": 0.321193665266037,
+      "learning_rate": 2.965901214165404e-05,
+      "loss": 0.0355,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3408913213448006,
+      "grad_norm": 0.30505916476249695,
+      "learning_rate": 2.959669679762248e-05,
+      "loss": 0.0242,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3416731821735732,
+      "grad_norm": 0.3857214152812958,
+      "learning_rate": 2.9534419454109476e-05,
+      "loss": 0.0725,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3424550430023456,
+      "grad_norm": 0.3174574673175812,
+      "learning_rate": 2.9472180227104628e-05,
+      "loss": 0.0465,
+      "step": 1717
+    },
+    {
+      "epoch": 1.343236903831118,
+      "grad_norm": 0.4131394028663635,
+      "learning_rate": 2.940997923252654e-05,
+      "loss": 0.079,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3440187646598907,
+      "grad_norm": 0.3730389475822449,
+      "learning_rate": 2.9347816586222654e-05,
+      "loss": 0.0444,
+      "step": 1719
+    },
+    {
+      "epoch": 1.344800625488663,
+      "grad_norm": 0.39177075028419495,
+      "learning_rate": 2.9285692403968977e-05,
+      "loss": 0.0683,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3455824863174355,
+      "grad_norm": 0.31527620553970337,
+      "learning_rate": 2.9223606801469882e-05,
+      "loss": 0.0887,
+      "step": 1721
+    },
+    {
+      "epoch": 1.346364347146208,
+      "grad_norm": 0.3214835524559021,
+      "learning_rate": 2.9161559894357827e-05,
+      "loss": 0.0491,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3471462079749805,
+      "grad_norm": 0.4698956608772278,
+      "learning_rate": 2.909955179819328e-05,
+      "loss": 0.0489,
+      "step": 1723
+    },
+    {
+      "epoch": 1.347928068803753,
+      "grad_norm": 0.42727044224739075,
+      "learning_rate": 2.9037582628464398e-05,
+      "loss": 0.0557,
+      "step": 1724
+    },
+    {
+      "epoch": 1.3487099296325253,
+      "grad_norm": 0.6430052518844604,
+      "learning_rate": 2.8975652500586826e-05,
+      "loss": 0.0917,
+      "step": 1725
+    },
+    {
+      "epoch": 1.349491790461298,
+      "grad_norm": 0.36423027515411377,
+      "learning_rate": 2.8913761529903484e-05,
+      "loss": 0.0952,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3502736512900704,
+      "grad_norm": 0.3790799677371979,
+      "learning_rate": 2.885190983168444e-05,
+      "loss": 0.068,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3510555121188428,
+      "grad_norm": 0.45250916481018066,
+      "learning_rate": 2.8790097521126507e-05,
+      "loss": 0.0534,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3518373729476152,
+      "grad_norm": 0.3944171965122223,
+      "learning_rate": 2.8728324713353194e-05,
+      "loss": 0.0531,
+      "step": 1729
+    },
+    {
+      "epoch": 1.3526192337763878,
+      "grad_norm": 0.3291238844394684,
+      "learning_rate": 2.8666591523414475e-05,
+      "loss": 0.0625,
+      "step": 1730
+    },
+    {
+      "epoch": 1.3534010946051602,
+      "grad_norm": 0.5208644866943359,
+      "learning_rate": 2.860489806628648e-05,
+      "loss": 0.0516,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3541829554339326,
+      "grad_norm": 0.3720276951789856,
+      "learning_rate": 2.854324445687135e-05,
+      "loss": 0.052,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3549648162627053,
+      "grad_norm": 0.3863518536090851,
+      "learning_rate": 2.8481630809997022e-05,
+      "loss": 0.0671,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3557466770914777,
+      "grad_norm": 0.4168684482574463,
+      "learning_rate": 2.842005724041701e-05,
+      "loss": 0.0646,
+      "step": 1734
+    },
+    {
+      "epoch": 1.35652853792025,
+      "grad_norm": 0.33291158080101013,
+      "learning_rate": 2.8358523862810155e-05,
+      "loss": 0.0565,
+      "step": 1735
+    },
+    {
+      "epoch": 1.3573103987490227,
+      "grad_norm": 0.36476758122444153,
+      "learning_rate": 2.8297030791780455e-05,
+      "loss": 0.0538,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3580922595777951,
+      "grad_norm": 0.3390056788921356,
+      "learning_rate": 2.823557814185691e-05,
+      "loss": 0.0636,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3588741204065675,
+      "grad_norm": 0.5852319598197937,
+      "learning_rate": 2.81741660274931e-05,
+      "loss": 0.1102,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3596559812353401,
+      "grad_norm": 0.34502658247947693,
+      "learning_rate": 2.811279456306718e-05,
+      "loss": 0.0382,
+      "step": 1739
+    },
+    {
+      "epoch": 1.3604378420641126,
+      "grad_norm": 0.3496917188167572,
+      "learning_rate": 2.8051463862881665e-05,
+      "loss": 0.0833,
+      "step": 1740
+    },
+    {
+      "epoch": 1.361219702892885,
+      "grad_norm": 0.5584290027618408,
+      "learning_rate": 2.7990174041162997e-05,
+      "loss": 0.0919,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3620015637216576,
+      "grad_norm": 0.3309808671474457,
+      "learning_rate": 2.792892521206163e-05,
+      "loss": 0.0522,
+      "step": 1742
+    },
+    {
+      "epoch": 1.36278342455043,
+      "grad_norm": 0.4036805331707001,
+      "learning_rate": 2.7867717489651567e-05,
+      "loss": 0.0644,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3635652853792024,
+      "grad_norm": 0.2975353002548218,
+      "learning_rate": 2.7806550987930314e-05,
+      "loss": 0.061,
+      "step": 1744
+    },
+    {
+      "epoch": 1.364347146207975,
+      "grad_norm": 0.4004823863506317,
+      "learning_rate": 2.7745425820818567e-05,
+      "loss": 0.0501,
+      "step": 1745
+    },
+    {
+      "epoch": 1.3651290070367474,
+      "grad_norm": 0.4823578894138336,
+      "learning_rate": 2.7684342102160034e-05,
+      "loss": 0.0705,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3659108678655199,
+      "grad_norm": 0.40492135286331177,
+      "learning_rate": 2.7623299945721302e-05,
+      "loss": 0.0546,
+      "step": 1747
+    },
+    {
+      "epoch": 1.3666927286942925,
+      "grad_norm": 0.3871229887008667,
+      "learning_rate": 2.7562299465191432e-05,
+      "loss": 0.0514,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3674745895230649,
+      "grad_norm": 0.43121662735939026,
+      "learning_rate": 2.7501340774181934e-05,
+      "loss": 0.0724,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3682564503518373,
+      "grad_norm": 0.4135430157184601,
+      "learning_rate": 2.7440423986226527e-05,
+      "loss": 0.0584,
+      "step": 1750
+    },
+    {
+      "epoch": 1.36903831118061,
+      "grad_norm": 0.36338895559310913,
+      "learning_rate": 2.7379549214780754e-05,
+      "loss": 0.0534,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3698201720093823,
+      "grad_norm": 0.46101894974708557,
+      "learning_rate": 2.7318716573222053e-05,
+      "loss": 0.0596,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3706020328381547,
+      "grad_norm": 0.3271227180957794,
+      "learning_rate": 2.7257926174849328e-05,
+      "loss": 0.0326,
+      "step": 1753
+    },
+    {
+      "epoch": 1.3713838936669274,
+      "grad_norm": 0.3329024612903595,
+      "learning_rate": 2.719717813288275e-05,
+      "loss": 0.0499,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3721657544956998,
+      "grad_norm": 0.49644121527671814,
+      "learning_rate": 2.713647256046371e-05,
+      "loss": 0.069,
+      "step": 1755
+    },
+    {
+      "epoch": 1.3729476153244722,
+      "grad_norm": 0.32194784283638,
+      "learning_rate": 2.707580957065445e-05,
+      "loss": 0.0671,
+      "step": 1756
+    },
+    {
+      "epoch": 1.3737294761532448,
+      "grad_norm": 0.31142404675483704,
+      "learning_rate": 2.7015189276437884e-05,
+      "loss": 0.0714,
+      "step": 1757
+    },
+    {
+      "epoch": 1.3745113369820172,
+      "grad_norm": 0.4320639967918396,
+      "learning_rate": 2.6954611790717445e-05,
+      "loss": 0.0679,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3752931978107896,
+      "grad_norm": 0.4968913495540619,
+      "learning_rate": 2.6894077226316783e-05,
+      "loss": 0.067,
+      "step": 1759
+    },
+    {
+      "epoch": 1.3760750586395623,
+      "grad_norm": 0.44248491525650024,
+      "learning_rate": 2.6833585695979714e-05,
+      "loss": 0.0791,
+      "step": 1760
+    },
+    {
+      "epoch": 1.3768569194683347,
+      "grad_norm": 0.3351903259754181,
+      "learning_rate": 2.6773137312369756e-05,
+      "loss": 0.0408,
+      "step": 1761
+    },
+    {
+      "epoch": 1.377638780297107,
+      "grad_norm": 0.3830302059650421,
+      "learning_rate": 2.6712732188070198e-05,
+      "loss": 0.0586,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3784206411258797,
+      "grad_norm": 0.30592599511146545,
+      "learning_rate": 2.6652370435583712e-05,
+      "loss": 0.0456,
+      "step": 1763
+    },
+    {
+      "epoch": 1.379202501954652,
+      "grad_norm": 0.33059370517730713,
+      "learning_rate": 2.6592052167332126e-05,
+      "loss": 0.0443,
+      "step": 1764
+    },
+    {
+      "epoch": 1.3799843627834245,
+      "grad_norm": 0.3415541350841522,
+      "learning_rate": 2.6531777495656406e-05,
+      "loss": 0.0372,
+      "step": 1765
+    },
+    {
+      "epoch": 1.3807662236121971,
+      "grad_norm": 0.5239612460136414,
+      "learning_rate": 2.6471546532816238e-05,
+      "loss": 0.0576,
+      "step": 1766
+    },
+    {
+      "epoch": 1.3815480844409695,
+      "grad_norm": 0.4342079162597656,
+      "learning_rate": 2.641135939098991e-05,
+      "loss": 0.0621,
+      "step": 1767
+    },
+    {
+      "epoch": 1.382329945269742,
+      "grad_norm": 0.42560240626335144,
+      "learning_rate": 2.6351216182274107e-05,
+      "loss": 0.0423,
+      "step": 1768
+    },
+    {
+      "epoch": 1.3831118060985146,
+      "grad_norm": 0.36572739481925964,
+      "learning_rate": 2.6291117018683676e-05,
+      "loss": 0.0657,
+      "step": 1769
+    },
+    {
+      "epoch": 1.383893666927287,
+      "grad_norm": 0.46662601828575134,
+      "learning_rate": 2.6231062012151487e-05,
+      "loss": 0.0828,
+      "step": 1770
+    },
+    {
+      "epoch": 1.3846755277560594,
+      "grad_norm": 0.34814974665641785,
+      "learning_rate": 2.6171051274528057e-05,
+      "loss": 0.0488,
+      "step": 1771
+    },
+    {
+      "epoch": 1.385457388584832,
+      "grad_norm": 0.4473120868206024,
+      "learning_rate": 2.611108491758158e-05,
+      "loss": 0.0564,
+      "step": 1772
+    },
+    {
+      "epoch": 1.3862392494136044,
+      "grad_norm": 0.5653979778289795,
+      "learning_rate": 2.6051163052997528e-05,
+      "loss": 0.0885,
+      "step": 1773
+    },
+    {
+      "epoch": 1.3870211102423768,
+      "grad_norm": 0.38065004348754883,
+      "learning_rate": 2.5991285792378457e-05,
+      "loss": 0.0526,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3878029710711495,
+      "grad_norm": 0.4245047867298126,
+      "learning_rate": 2.5931453247243963e-05,
+      "loss": 0.0553,
+      "step": 1775
+    },
+    {
+      "epoch": 1.3885848318999219,
+      "grad_norm": 0.4021928310394287,
+      "learning_rate": 2.5871665529030286e-05,
+      "loss": 0.05,
+      "step": 1776
+    },
+    {
+      "epoch": 1.3893666927286943,
+      "grad_norm": 0.4062913656234741,
+      "learning_rate": 2.5811922749090188e-05,
+      "loss": 0.0593,
+      "step": 1777
+    },
+    {
+      "epoch": 1.3901485535574667,
+      "grad_norm": 0.3696041405200958,
+      "learning_rate": 2.5752225018692745e-05,
+      "loss": 0.0567,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3909304143862393,
+      "grad_norm": 0.42013299465179443,
+      "learning_rate": 2.569257244902311e-05,
+      "loss": 0.0666,
+      "step": 1779
+    },
+    {
+      "epoch": 1.3917122752150117,
+      "grad_norm": 0.530455470085144,
+      "learning_rate": 2.56329651511824e-05,
+      "loss": 0.0424,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3924941360437841,
+      "grad_norm": 0.46744510531425476,
+      "learning_rate": 2.5573403236187287e-05,
+      "loss": 0.0484,
+      "step": 1781
+    },
+    {
+      "epoch": 1.3932759968725568,
+      "grad_norm": 0.39260637760162354,
+      "learning_rate": 2.551388681497e-05,
+      "loss": 0.0457,
+      "step": 1782
+    },
+    {
+      "epoch": 1.3940578577013292,
+      "grad_norm": 0.379202276468277,
+      "learning_rate": 2.5454415998378073e-05,
+      "loss": 0.057,
+      "step": 1783
+    },
+    {
+      "epoch": 1.3948397185301016,
+      "grad_norm": 0.4000501036643982,
+      "learning_rate": 2.5394990897173987e-05,
+      "loss": 0.0752,
+      "step": 1784
+    },
+    {
+      "epoch": 1.395621579358874,
+      "grad_norm": 0.3499116897583008,
+      "learning_rate": 2.5335611622035198e-05,
+      "loss": 0.0469,
+      "step": 1785
+    },
+    {
+      "epoch": 1.3964034401876466,
+      "grad_norm": 0.4184105694293976,
+      "learning_rate": 2.5276278283553746e-05,
+      "loss": 0.0524,
+      "step": 1786
+    },
+    {
+      "epoch": 1.397185301016419,
+      "grad_norm": 0.31106022000312805,
+      "learning_rate": 2.5216990992236135e-05,
+      "loss": 0.0615,
+      "step": 1787
+    },
+    {
+      "epoch": 1.3979671618451914,
+      "grad_norm": 0.44269534945487976,
+      "learning_rate": 2.51577498585031e-05,
+      "loss": 0.0368,
+      "step": 1788
+    },
+    {
+      "epoch": 1.398749022673964,
+      "grad_norm": 0.3069053888320923,
+      "learning_rate": 2.509855499268938e-05,
+      "loss": 0.026,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3995308835027365,
+      "grad_norm": 0.38774728775024414,
+      "learning_rate": 2.5039406505043662e-05,
+      "loss": 0.0392,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4003127443315089,
+      "grad_norm": 0.5503056645393372,
+      "learning_rate": 2.498030450572808e-05,
+      "loss": 0.0639,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4010946051602815,
+      "grad_norm": 0.2796127498149872,
+      "learning_rate": 2.492124910481829e-05,
+      "loss": 0.0671,
+      "step": 1792
+    },
+    {
+      "epoch": 1.401876465989054,
+      "grad_norm": 0.811257541179657,
+      "learning_rate": 2.486224041230319e-05,
+      "loss": 0.0987,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4026583268178263,
+      "grad_norm": 0.27601099014282227,
+      "learning_rate": 2.4803278538084568e-05,
+      "loss": 0.0403,
+      "step": 1794
+    },
+    {
+      "epoch": 1.403440187646599,
+      "grad_norm": 0.3818424642086029,
+      "learning_rate": 2.474436359197714e-05,
+      "loss": 0.0441,
+      "step": 1795
+    },
+    {
+      "epoch": 1.4042220484753714,
+      "grad_norm": 0.37289881706237793,
+      "learning_rate": 2.468549568370813e-05,
+      "loss": 0.0528,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4050039093041438,
+      "grad_norm": 0.3854784369468689,
+      "learning_rate": 2.4626674922917207e-05,
+      "loss": 0.0645,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4057857701329164,
+      "grad_norm": 0.38945889472961426,
+      "learning_rate": 2.4567901419156207e-05,
+      "loss": 0.072,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4065676309616888,
+      "grad_norm": 0.35413599014282227,
+      "learning_rate": 2.4509175281888957e-05,
+      "loss": 0.0319,
+      "step": 1799
+    },
+    {
+      "epoch": 1.4073494917904612,
+      "grad_norm": 0.32834237813949585,
+      "learning_rate": 2.4450496620491053e-05,
+      "loss": 0.0323,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4073494917904612,
+      "eval_loss": 0.12402170151472092,
+      "eval_runtime": 13.3717,
+      "eval_samples_per_second": 3.889,
+      "eval_steps_per_second": 0.972,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4081313526192338,
+      "grad_norm": 0.4939782917499542,
+      "learning_rate": 2.4391865544249687e-05,
+      "loss": 0.0804,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4089132134480062,
+      "grad_norm": 0.4043434262275696,
+      "learning_rate": 2.4333282162363402e-05,
+      "loss": 0.0754,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4096950742767786,
+      "grad_norm": 0.46937301754951477,
+      "learning_rate": 2.4274746583941975e-05,
+      "loss": 0.0391,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4104769351055513,
+      "grad_norm": 0.3977530002593994,
+      "learning_rate": 2.4216258918006033e-05,
+      "loss": 0.0621,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4112587959343237,
+      "grad_norm": 0.4665568768978119,
+      "learning_rate": 2.415781927348709e-05,
+      "loss": 0.06,
+      "step": 1805
+    },
+    {
+      "epoch": 1.412040656763096,
+      "grad_norm": 0.34306639432907104,
+      "learning_rate": 2.4099427759227183e-05,
+      "loss": 0.0427,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4128225175918687,
+      "grad_norm": 0.3104557394981384,
+      "learning_rate": 2.4041084483978616e-05,
+      "loss": 0.038,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4136043784206411,
+      "grad_norm": 0.45473772287368774,
+      "learning_rate": 2.3982789556404e-05,
+      "loss": 0.0842,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4143862392494135,
+      "grad_norm": 0.43577784299850464,
+      "learning_rate": 2.39245430850758e-05,
+      "loss": 0.0577,
+      "step": 1809
+    },
+    {
+      "epoch": 1.4151681000781862,
+      "grad_norm": 0.33145925402641296,
+      "learning_rate": 2.386634517847626e-05,
+      "loss": 0.0459,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4159499609069586,
+      "grad_norm": 0.4155854284763336,
+      "learning_rate": 2.380819594499718e-05,
+      "loss": 0.0449,
+      "step": 1811
+    },
+    {
+      "epoch": 1.416731821735731,
+      "grad_norm": 0.33247607946395874,
+      "learning_rate": 2.3750095492939695e-05,
+      "loss": 0.0786,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4175136825645036,
+      "grad_norm": 0.398334264755249,
+      "learning_rate": 2.369204393051409e-05,
+      "loss": 0.0849,
+      "step": 1813
+    },
+    {
+      "epoch": 1.418295543393276,
+      "grad_norm": 0.4067325294017792,
+      "learning_rate": 2.3634041365839588e-05,
+      "loss": 0.0603,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4190774042220484,
+      "grad_norm": 0.4016716778278351,
+      "learning_rate": 2.3576087906944195e-05,
+      "loss": 0.0745,
+      "step": 1815
+    },
+    {
+      "epoch": 1.419859265050821,
+      "grad_norm": 0.36571478843688965,
+      "learning_rate": 2.351818366176443e-05,
+      "loss": 0.0616,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4206411258795935,
+      "grad_norm": 0.3647540509700775,
+      "learning_rate": 2.34603287381451e-05,
+      "loss": 0.0727,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4214229867083659,
+      "grad_norm": 0.32372230291366577,
+      "learning_rate": 2.340252324383925e-05,
+      "loss": 0.0571,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4222048475371385,
+      "grad_norm": 0.43298980593681335,
+      "learning_rate": 2.33447672865078e-05,
+      "loss": 0.0498,
+      "step": 1819
+    },
+    {
+      "epoch": 1.422986708365911,
+      "grad_norm": 0.40937963128089905,
+      "learning_rate": 2.3287060973719433e-05,
+      "loss": 0.0609,
+      "step": 1820
+    },
+    {
+      "epoch": 1.4237685691946833,
+      "grad_norm": 0.42888909578323364,
+      "learning_rate": 2.322940441295036e-05,
+      "loss": 0.0647,
+      "step": 1821
+    },
+    {
+      "epoch": 1.424550430023456,
+      "grad_norm": 0.5512627363204956,
+      "learning_rate": 2.317179771158414e-05,
+      "loss": 0.0887,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4253322908522283,
+      "grad_norm": 0.4589550793170929,
+      "learning_rate": 2.3114240976911466e-05,
+      "loss": 0.0585,
+      "step": 1823
+    },
+    {
+      "epoch": 1.4261141516810008,
+      "grad_norm": 0.3198803663253784,
+      "learning_rate": 2.305673431612994e-05,
+      "loss": 0.0577,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4268960125097734,
+      "grad_norm": 0.33503422141075134,
+      "learning_rate": 2.2999277836343973e-05,
+      "loss": 0.0504,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4276778733385458,
+      "grad_norm": 0.3085525333881378,
+      "learning_rate": 2.2941871644564482e-05,
+      "loss": 0.065,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4284597341673182,
+      "grad_norm": 0.4282321035861969,
+      "learning_rate": 2.2884515847708648e-05,
+      "loss": 0.0543,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4292415949960908,
+      "grad_norm": 0.3238259255886078,
+      "learning_rate": 2.2827210552599925e-05,
+      "loss": 0.0507,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4300234558248632,
+      "grad_norm": 0.37012743949890137,
+      "learning_rate": 2.276995586596763e-05,
+      "loss": 0.0297,
+      "step": 1829
+    },
+    {
+      "epoch": 1.4308053166536356,
+      "grad_norm": 0.3611026108264923,
+      "learning_rate": 2.2712751894446833e-05,
+      "loss": 0.0389,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4315871774824083,
+      "grad_norm": 0.26761916279792786,
+      "learning_rate": 2.265559874457815e-05,
+      "loss": 0.0505,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4323690383111807,
+      "grad_norm": 0.4100571274757385,
+      "learning_rate": 2.2598496522807556e-05,
+      "loss": 0.0591,
+      "step": 1832
+    },
+    {
+      "epoch": 1.433150899139953,
+      "grad_norm": 0.2951042652130127,
+      "learning_rate": 2.254144533548616e-05,
+      "loss": 0.0343,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4339327599687255,
+      "grad_norm": 0.49010226130485535,
+      "learning_rate": 2.2484445288870037e-05,
+      "loss": 0.079,
+      "step": 1834
+    },
+    {
+      "epoch": 1.4347146207974981,
+      "grad_norm": 0.37976258993148804,
+      "learning_rate": 2.2427496489119986e-05,
+      "loss": 0.037,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4354964816262705,
+      "grad_norm": 0.4197808802127838,
+      "learning_rate": 2.2370599042301394e-05,
+      "loss": 0.0572,
+      "step": 1836
+    },
+    {
+      "epoch": 1.436278342455043,
+      "grad_norm": 0.30215969681739807,
+      "learning_rate": 2.2313753054383958e-05,
+      "loss": 0.0342,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4370602032838156,
+      "grad_norm": 0.3675067126750946,
+      "learning_rate": 2.225695863124161e-05,
+      "loss": 0.0676,
+      "step": 1838
+    },
+    {
+      "epoch": 1.437842064112588,
+      "grad_norm": 0.550208568572998,
+      "learning_rate": 2.220021587865218e-05,
+      "loss": 0.0797,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4386239249413604,
+      "grad_norm": 0.35796505212783813,
+      "learning_rate": 2.214352490229728e-05,
+      "loss": 0.0376,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4394057857701328,
+      "grad_norm": 0.3726497292518616,
+      "learning_rate": 2.2086885807762093e-05,
+      "loss": 0.0602,
+      "step": 1841
+    },
+    {
+      "epoch": 1.4401876465989054,
+      "grad_norm": 0.47628331184387207,
+      "learning_rate": 2.203029870053517e-05,
+      "loss": 0.0671,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4409695074276778,
+      "grad_norm": 0.3403119444847107,
+      "learning_rate": 2.197376368600825e-05,
+      "loss": 0.0853,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4417513682564502,
+      "grad_norm": 0.39009755849838257,
+      "learning_rate": 2.191728086947603e-05,
+      "loss": 0.0405,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4425332290852229,
+      "grad_norm": 0.340263307094574,
+      "learning_rate": 2.1860850356136015e-05,
+      "loss": 0.0566,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4433150899139953,
+      "grad_norm": 0.3942875862121582,
+      "learning_rate": 2.180447225108827e-05,
+      "loss": 0.0804,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4440969507427677,
+      "grad_norm": 0.44318339228630066,
+      "learning_rate": 2.1748146659335256e-05,
+      "loss": 0.0542,
+      "step": 1847
+    },
+    {
+      "epoch": 1.4448788115715403,
+      "grad_norm": 0.36085474491119385,
+      "learning_rate": 2.1691873685781673e-05,
+      "loss": 0.06,
+      "step": 1848
+    },
+    {
+      "epoch": 1.4456606724003127,
+      "grad_norm": 0.3341085612773895,
+      "learning_rate": 2.163565343523416e-05,
+      "loss": 0.0467,
+      "step": 1849
+    },
+    {
+      "epoch": 1.4464425332290851,
+      "grad_norm": 0.48646220564842224,
+      "learning_rate": 2.15794860124012e-05,
+      "loss": 0.0559,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4472243940578577,
+      "grad_norm": 0.3500666618347168,
+      "learning_rate": 2.152337152189287e-05,
+      "loss": 0.0617,
+      "step": 1851
+    },
+    {
+      "epoch": 1.4480062548866302,
+      "grad_norm": 0.34813258051872253,
+      "learning_rate": 2.1467310068220658e-05,
+      "loss": 0.0397,
+      "step": 1852
+    },
+    {
+      "epoch": 1.4487881157154026,
+      "grad_norm": 0.5163322687149048,
+      "learning_rate": 2.1411301755797293e-05,
+      "loss": 0.0591,
+      "step": 1853
+    },
+    {
+      "epoch": 1.4495699765441752,
+      "grad_norm": 0.37251824140548706,
+      "learning_rate": 2.1355346688936505e-05,
+      "loss": 0.0317,
+      "step": 1854
+    },
+    {
+      "epoch": 1.4503518373729476,
+      "grad_norm": 0.42122918367385864,
+      "learning_rate": 2.1299444971852876e-05,
+      "loss": 0.0363,
+      "step": 1855
+    },
+    {
+      "epoch": 1.45113369820172,
+      "grad_norm": 0.3831481337547302,
+      "learning_rate": 2.1243596708661618e-05,
+      "loss": 0.0396,
+      "step": 1856
+    },
+    {
+      "epoch": 1.4519155590304926,
+      "grad_norm": 0.40503552556037903,
+      "learning_rate": 2.118780200337836e-05,
+      "loss": 0.0491,
+      "step": 1857
+    },
+    {
+      "epoch": 1.452697419859265,
+      "grad_norm": 0.6177302598953247,
+      "learning_rate": 2.113206095991907e-05,
+      "loss": 0.0611,
+      "step": 1858
+    },
+    {
+      "epoch": 1.4534792806880374,
+      "grad_norm": 0.4156329035758972,
+      "learning_rate": 2.107637368209966e-05,
+      "loss": 0.0616,
+      "step": 1859
+    },
+    {
+      "epoch": 1.45426114151681,
+      "grad_norm": 0.41171953082084656,
+      "learning_rate": 2.102074027363594e-05,
+      "loss": 0.0472,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4550430023455825,
+      "grad_norm": 0.5042744278907776,
+      "learning_rate": 2.096516083814346e-05,
+      "loss": 0.0731,
+      "step": 1861
+    },
+    {
+      "epoch": 1.4558248631743549,
+      "grad_norm": 0.36022377014160156,
+      "learning_rate": 2.0909635479137163e-05,
+      "loss": 0.0465,
+      "step": 1862
+    },
+    {
+      "epoch": 1.4566067240031275,
+      "grad_norm": 0.45423850417137146,
+      "learning_rate": 2.085416430003131e-05,
+      "loss": 0.0557,
+      "step": 1863
+    },
+    {
+      "epoch": 1.4573885848319,
+      "grad_norm": 0.46697577834129333,
+      "learning_rate": 2.079874740413924e-05,
+      "loss": 0.1132,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4581704456606723,
+      "grad_norm": 0.46780556440353394,
+      "learning_rate": 2.074338489467322e-05,
+      "loss": 0.0849,
+      "step": 1865
+    },
+    {
+      "epoch": 1.458952306489445,
+      "grad_norm": 0.5109332203865051,
+      "learning_rate": 2.0688076874744183e-05,
+      "loss": 0.0803,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4597341673182174,
+      "grad_norm": 0.3422967493534088,
+      "learning_rate": 2.0632823447361593e-05,
+      "loss": 0.0786,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4605160281469898,
+      "grad_norm": 0.26753708720207214,
+      "learning_rate": 2.057762471543329e-05,
+      "loss": 0.0287,
+      "step": 1868
+    },
+    {
+      "epoch": 1.4612978889757624,
+      "grad_norm": 0.4200882911682129,
+      "learning_rate": 2.0522480781765153e-05,
+      "loss": 0.0504,
+      "step": 1869
+    },
+    {
+      "epoch": 1.4620797498045348,
+      "grad_norm": 0.5091989040374756,
+      "learning_rate": 2.0467391749061032e-05,
+      "loss": 0.0666,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4628616106333072,
+      "grad_norm": 0.3979477286338806,
+      "learning_rate": 2.0412357719922593e-05,
+      "loss": 0.0457,
+      "step": 1871
+    },
+    {
+      "epoch": 1.4636434714620798,
+      "grad_norm": 0.32263174653053284,
+      "learning_rate": 2.0357378796848968e-05,
+      "loss": 0.0695,
+      "step": 1872
+    },
+    {
+      "epoch": 1.4644253322908523,
+      "grad_norm": 0.32924485206604004,
+      "learning_rate": 2.0302455082236716e-05,
+      "loss": 0.07,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4652071931196247,
+      "grad_norm": 0.3112393021583557,
+      "learning_rate": 2.0247586678379534e-05,
+      "loss": 0.054,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4659890539483973,
+      "grad_norm": 0.3827929198741913,
+      "learning_rate": 2.019277368746812e-05,
+      "loss": 0.064,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4667709147771697,
+      "grad_norm": 0.4558732509613037,
+      "learning_rate": 2.0138016211589972e-05,
+      "loss": 0.0433,
+      "step": 1876
+    },
+    {
+      "epoch": 1.467552775605942,
+      "grad_norm": 0.5064982175827026,
+      "learning_rate": 2.008331435272917e-05,
+      "loss": 0.0723,
+      "step": 1877
+    },
+    {
+      "epoch": 1.4683346364347147,
+      "grad_norm": 0.3777099549770355,
+      "learning_rate": 2.0028668212766283e-05,
+      "loss": 0.0542,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4691164972634871,
+      "grad_norm": 0.3097599744796753,
+      "learning_rate": 1.997407789347799e-05,
+      "loss": 0.0495,
+      "step": 1879
+    },
+    {
+      "epoch": 1.4698983580922595,
+      "grad_norm": 0.2744178771972656,
+      "learning_rate": 1.991954349653707e-05,
+      "loss": 0.1154,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4706802189210322,
+      "grad_norm": 0.3737988770008087,
+      "learning_rate": 1.9865065123512194e-05,
+      "loss": 0.0682,
+      "step": 1881
+    },
+    {
+      "epoch": 1.4714620797498046,
+      "grad_norm": 0.5198439955711365,
+      "learning_rate": 1.9810642875867573e-05,
+      "loss": 0.0913,
+      "step": 1882
+    },
+    {
+      "epoch": 1.472243940578577,
+      "grad_norm": 0.3445640802383423,
+      "learning_rate": 1.9756276854963002e-05,
+      "loss": 0.0829,
+      "step": 1883
+    },
+    {
+      "epoch": 1.4730258014073496,
+      "grad_norm": 0.3063513934612274,
+      "learning_rate": 1.970196716205349e-05,
+      "loss": 0.0344,
+      "step": 1884
+    },
+    {
+      "epoch": 1.473807662236122,
+      "grad_norm": 0.4947914481163025,
+      "learning_rate": 1.9647713898289154e-05,
+      "loss": 0.0694,
+      "step": 1885
+    },
+    {
+      "epoch": 1.4745895230648944,
+      "grad_norm": 0.35459840297698975,
+      "learning_rate": 1.9593517164715014e-05,
+      "loss": 0.054,
+      "step": 1886
+    },
+    {
+      "epoch": 1.475371383893667,
+      "grad_norm": 0.4725291132926941,
+      "learning_rate": 1.953937706227078e-05,
+      "loss": 0.0764,
+      "step": 1887
+    },
+    {
+      "epoch": 1.4761532447224395,
+      "grad_norm": 0.47036585211753845,
+      "learning_rate": 1.9485293691790774e-05,
+      "loss": 0.0679,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4769351055512119,
+      "grad_norm": 0.4962479770183563,
+      "learning_rate": 1.943126715400353e-05,
+      "loss": 0.0753,
+      "step": 1889
+    },
+    {
+      "epoch": 1.4777169663799843,
+      "grad_norm": 0.33534544706344604,
+      "learning_rate": 1.9377297549531805e-05,
+      "loss": 0.0316,
+      "step": 1890
+    },
+    {
+      "epoch": 1.478498827208757,
+      "grad_norm": 0.5018200874328613,
+      "learning_rate": 1.9323384978892357e-05,
+      "loss": 0.053,
+      "step": 1891
+    },
+    {
+      "epoch": 1.4792806880375293,
+      "grad_norm": 0.3958439230918884,
+      "learning_rate": 1.9269529542495605e-05,
+      "loss": 0.0628,
+      "step": 1892
+    },
+    {
+      "epoch": 1.4800625488663017,
+      "grad_norm": 0.40492334961891174,
+      "learning_rate": 1.921573134064569e-05,
+      "loss": 0.0504,
+      "step": 1893
+    },
+    {
+      "epoch": 1.4808444096950744,
+      "grad_norm": 0.4669480323791504,
+      "learning_rate": 1.9161990473540074e-05,
+      "loss": 0.0658,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4816262705238468,
+      "grad_norm": 0.40686583518981934,
+      "learning_rate": 1.9108307041269418e-05,
+      "loss": 0.0595,
+      "step": 1895
+    },
+    {
+      "epoch": 1.4824081313526192,
+      "grad_norm": 0.37184271216392517,
+      "learning_rate": 1.9054681143817487e-05,
+      "loss": 0.0571,
+      "step": 1896
+    },
+    {
+      "epoch": 1.4831899921813916,
+      "grad_norm": 0.4182341396808624,
+      "learning_rate": 1.9001112881060845e-05,
+      "loss": 0.03,
+      "step": 1897
+    },
+    {
+      "epoch": 1.4839718530101642,
+      "grad_norm": 0.42824220657348633,
+      "learning_rate": 1.89476023527687e-05,
+      "loss": 0.0436,
+      "step": 1898
+    },
+    {
+      "epoch": 1.4847537138389366,
+      "grad_norm": 0.4151003658771515,
+      "learning_rate": 1.8894149658602767e-05,
+      "loss": 0.0515,
+      "step": 1899
+    },
+    {
+      "epoch": 1.485535574667709,
+      "grad_norm": 0.38467374444007874,
+      "learning_rate": 1.8840754898117e-05,
+      "loss": 0.0334,
+      "step": 1900
+    },
+    {
+      "epoch": 1.4863174354964817,
+      "grad_norm": 0.6196097731590271,
+      "learning_rate": 1.878741817075754e-05,
+      "loss": 0.0773,
+      "step": 1901
+    },
+    {
+      "epoch": 1.487099296325254,
+      "grad_norm": 0.3245231509208679,
+      "learning_rate": 1.8734139575862313e-05,
+      "loss": 0.0367,
+      "step": 1902
+    },
+    {
+      "epoch": 1.4878811571540265,
+      "grad_norm": 0.45020565390586853,
+      "learning_rate": 1.8680919212661097e-05,
+      "loss": 0.04,
+      "step": 1903
+    },
+    {
+      "epoch": 1.488663017982799,
+      "grad_norm": 0.5742055773735046,
+      "learning_rate": 1.8627757180275175e-05,
+      "loss": 0.0592,
+      "step": 1904
+    },
+    {
+      "epoch": 1.4894448788115715,
+      "grad_norm": 0.35358932614326477,
+      "learning_rate": 1.8574653577717116e-05,
+      "loss": 0.0777,
+      "step": 1905
+    },
+    {
+      "epoch": 1.490226739640344,
+      "grad_norm": 0.3724954426288605,
+      "learning_rate": 1.8521608503890787e-05,
+      "loss": 0.0436,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4910086004691165,
+      "grad_norm": 0.4322279393672943,
+      "learning_rate": 1.8468622057590978e-05,
+      "loss": 0.0396,
+      "step": 1907
+    },
+    {
+      "epoch": 1.491790461297889,
+      "grad_norm": 0.297240287065506,
+      "learning_rate": 1.8415694337503297e-05,
+      "loss": 0.0313,
+      "step": 1908
+    },
+    {
+      "epoch": 1.4925723221266614,
+      "grad_norm": 0.2714248299598694,
+      "learning_rate": 1.836282544220398e-05,
+      "loss": 0.0375,
+      "step": 1909
+    },
+    {
+      "epoch": 1.493354182955434,
+      "grad_norm": 0.5247171521186829,
+      "learning_rate": 1.8310015470159685e-05,
+      "loss": 0.0567,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4941360437842064,
+      "grad_norm": 0.30854910612106323,
+      "learning_rate": 1.825726451972739e-05,
+      "loss": 0.0451,
+      "step": 1911
+    },
+    {
+      "epoch": 1.4949179046129788,
+      "grad_norm": 0.4068269431591034,
+      "learning_rate": 1.820457268915404e-05,
+      "loss": 0.0481,
+      "step": 1912
+    },
+    {
+      "epoch": 1.4956997654417514,
+      "grad_norm": 0.39392808079719543,
+      "learning_rate": 1.815194007657659e-05,
+      "loss": 0.057,
+      "step": 1913
+    },
+    {
+      "epoch": 1.4964816262705238,
+      "grad_norm": 0.31934913992881775,
+      "learning_rate": 1.809936678002163e-05,
+      "loss": 0.0219,
+      "step": 1914
+    },
+    {
+      "epoch": 1.4972634870992962,
+      "grad_norm": 0.5628071427345276,
+      "learning_rate": 1.804685289740526e-05,
+      "loss": 0.0834,
+      "step": 1915
+    },
+    {
+      "epoch": 1.4980453479280689,
+      "grad_norm": 0.3296336829662323,
+      "learning_rate": 1.7994398526532986e-05,
+      "loss": 0.0268,
+      "step": 1916
+    },
+    {
+      "epoch": 1.4988272087568413,
+      "grad_norm": 0.40787026286125183,
+      "learning_rate": 1.794200376509944e-05,
+      "loss": 0.0746,
+      "step": 1917
+    },
+    {
+      "epoch": 1.4996090695856137,
+      "grad_norm": 0.4122026860713959,
+      "learning_rate": 1.788966871068824e-05,
+      "loss": 0.0591,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5003909304143863,
+      "grad_norm": 0.4837484359741211,
+      "learning_rate": 1.7837393460771795e-05,
+      "loss": 0.0713,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5011727912431587,
+      "grad_norm": 0.36081263422966003,
+      "learning_rate": 1.7785178112711114e-05,
+      "loss": 0.0536,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5019546520719311,
+      "grad_norm": 0.5664728879928589,
+      "learning_rate": 1.7733022763755725e-05,
+      "loss": 0.0981,
+      "step": 1921
+    },
+    {
+      "epoch": 1.5027365129007038,
+      "grad_norm": 0.46622321009635925,
+      "learning_rate": 1.7680927511043277e-05,
+      "loss": 0.0736,
+      "step": 1922
+    },
+    {
+      "epoch": 1.5035183737294762,
+      "grad_norm": 0.4234619438648224,
+      "learning_rate": 1.762889245159957e-05,
+      "loss": 0.0952,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5043002345582486,
+      "grad_norm": 0.3387734591960907,
+      "learning_rate": 1.757691768233834e-05,
+      "loss": 0.0601,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5050820953870212,
+      "grad_norm": 0.35551008582115173,
+      "learning_rate": 1.7525003300060904e-05,
+      "loss": 0.0887,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5058639562157936,
+      "grad_norm": 0.5371754765510559,
+      "learning_rate": 1.747314940145624e-05,
+      "loss": 0.0658,
+      "step": 1926
+    },
+    {
+      "epoch": 1.506645817044566,
+      "grad_norm": 0.3989940881729126,
+      "learning_rate": 1.7421356083100615e-05,
+      "loss": 0.0381,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5074276778733386,
+      "grad_norm": 0.44846925139427185,
+      "learning_rate": 1.7369623441457465e-05,
+      "loss": 0.0607,
+      "step": 1928
+    },
+    {
+      "epoch": 1.508209538702111,
+      "grad_norm": 0.3646467328071594,
+      "learning_rate": 1.7317951572877237e-05,
+      "loss": 0.0551,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5089913995308835,
+      "grad_norm": 0.41458413004875183,
+      "learning_rate": 1.7266340573597163e-05,
+      "loss": 0.0536,
+      "step": 1930
+    },
+    {
+      "epoch": 1.509773260359656,
+      "grad_norm": 0.3311718702316284,
+      "learning_rate": 1.7214790539741167e-05,
+      "loss": 0.0459,
+      "step": 1931
+    },
+    {
+      "epoch": 1.5105551211884285,
+      "grad_norm": 0.3195326626300812,
+      "learning_rate": 1.7163301567319546e-05,
+      "loss": 0.0285,
+      "step": 1932
+    },
+    {
+      "epoch": 1.511336982017201,
+      "grad_norm": 0.29825884103775024,
+      "learning_rate": 1.7111873752228907e-05,
+      "loss": 0.0555,
+      "step": 1933
+    },
+    {
+      "epoch": 1.5121188428459735,
+      "grad_norm": 0.4819906949996948,
+      "learning_rate": 1.7060507190252012e-05,
+      "loss": 0.0894,
+      "step": 1934
+    },
+    {
+      "epoch": 1.512900703674746,
+      "grad_norm": 0.4403591454029083,
+      "learning_rate": 1.700920197705742e-05,
+      "loss": 0.0653,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5136825645035183,
+      "grad_norm": 0.3573175370693207,
+      "learning_rate": 1.6957958208199548e-05,
+      "loss": 0.0453,
+      "step": 1936
+    },
+    {
+      "epoch": 1.514464425332291,
+      "grad_norm": 0.3273763656616211,
+      "learning_rate": 1.6906775979118307e-05,
+      "loss": 0.0469,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5152462861610634,
+      "grad_norm": 0.29220449924468994,
+      "learning_rate": 1.6855655385139002e-05,
+      "loss": 0.0337,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5160281469898358,
+      "grad_norm": 0.4651147425174713,
+      "learning_rate": 1.680459652147216e-05,
+      "loss": 0.1163,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5168100078186084,
+      "grad_norm": 0.43985429406166077,
+      "learning_rate": 1.675359948321333e-05,
+      "loss": 0.0678,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5175918686473806,
+      "grad_norm": 0.2881910800933838,
+      "learning_rate": 1.67026643653429e-05,
+      "loss": 0.0332,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5183737294761532,
+      "grad_norm": 0.7812718152999878,
+      "learning_rate": 1.6651791262725958e-05,
+      "loss": 0.0723,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5191555903049259,
+      "grad_norm": 0.4525265097618103,
+      "learning_rate": 1.6600980270112055e-05,
+      "loss": 0.05,
+      "step": 1943
+    },
+    {
+      "epoch": 1.519937451133698,
+      "grad_norm": 0.4249601662158966,
+      "learning_rate": 1.6550231482135136e-05,
+      "loss": 0.0504,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5207193119624707,
+      "grad_norm": 0.40071019530296326,
+      "learning_rate": 1.6499544993313183e-05,
+      "loss": 0.0466,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5215011727912433,
+      "grad_norm": 0.3151744306087494,
+      "learning_rate": 1.6448920898048247e-05,
+      "loss": 0.046,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5222830336200155,
+      "grad_norm": 0.21613501012325287,
+      "learning_rate": 1.6398359290626135e-05,
+      "loss": 0.0159,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5230648944487881,
+      "grad_norm": 0.38744068145751953,
+      "learning_rate": 1.634786026521623e-05,
+      "loss": 0.0569,
+      "step": 1948
+    },
+    {
+      "epoch": 1.5238467552775607,
+      "grad_norm": 0.33948633074760437,
+      "learning_rate": 1.629742391587144e-05,
+      "loss": 0.0263,
+      "step": 1949
+    },
+    {
+      "epoch": 1.524628616106333,
+      "grad_norm": 0.558428168296814,
+      "learning_rate": 1.6247050336527886e-05,
+      "loss": 0.056,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5254104769351056,
+      "grad_norm": 0.3884804844856262,
+      "learning_rate": 1.619673962100479e-05,
+      "loss": 0.0611,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5261923377638782,
+      "grad_norm": 0.4065593481063843,
+      "learning_rate": 1.6146491863004303e-05,
+      "loss": 0.0371,
+      "step": 1952
+    },
+    {
+      "epoch": 1.5269741985926504,
+      "grad_norm": 0.4196559190750122,
+      "learning_rate": 1.6096307156111312e-05,
+      "loss": 0.0551,
+      "step": 1953
+    },
+    {
+      "epoch": 1.527756059421423,
+      "grad_norm": 0.3707648515701294,
+      "learning_rate": 1.6046185593793272e-05,
+      "loss": 0.0424,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5285379202501954,
+      "grad_norm": 0.3814994990825653,
+      "learning_rate": 1.5996127269400023e-05,
+      "loss": 0.0238,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5293197810789678,
+      "grad_norm": 0.4228754937648773,
+      "learning_rate": 1.594613227616367e-05,
+      "loss": 0.0558,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5301016419077405,
+      "grad_norm": 0.5330584645271301,
+      "learning_rate": 1.589620070719834e-05,
+      "loss": 0.1043,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5308835027365129,
+      "grad_norm": 0.5210996866226196,
+      "learning_rate": 1.5846332655499972e-05,
+      "loss": 0.0895,
+      "step": 1958
+    },
+    {
+      "epoch": 1.5316653635652853,
+      "grad_norm": 0.5705960988998413,
+      "learning_rate": 1.579652821394632e-05,
+      "loss": 0.0708,
+      "step": 1959
+    },
+    {
+      "epoch": 1.532447224394058,
+      "grad_norm": 0.420893132686615,
+      "learning_rate": 1.574678747529659e-05,
+      "loss": 0.0559,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5332290852228303,
+      "grad_norm": 0.36799362301826477,
+      "learning_rate": 1.5697110532191366e-05,
+      "loss": 0.0472,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5340109460516027,
+      "grad_norm": 0.2931898832321167,
+      "learning_rate": 1.5647497477152406e-05,
+      "loss": 0.0376,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5347928068803753,
+      "grad_norm": 0.36866313219070435,
+      "learning_rate": 1.559794840258249e-05,
+      "loss": 0.0419,
+      "step": 1963
+    },
+    {
+      "epoch": 1.5355746677091477,
+      "grad_norm": 0.4287290871143341,
+      "learning_rate": 1.5548463400765227e-05,
+      "loss": 0.0564,
+      "step": 1964
+    },
+    {
+      "epoch": 1.5363565285379202,
+      "grad_norm": 0.38672366738319397,
+      "learning_rate": 1.549904256386488e-05,
+      "loss": 0.0691,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5371383893666928,
+      "grad_norm": 0.4107593894004822,
+      "learning_rate": 1.544968598392626e-05,
+      "loss": 0.0998,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5379202501954652,
+      "grad_norm": 0.3715873658657074,
+      "learning_rate": 1.5400393752874454e-05,
+      "loss": 0.0748,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5387021110242376,
+      "grad_norm": 0.4931233525276184,
+      "learning_rate": 1.5351165962514675e-05,
+      "loss": 0.0277,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5394839718530102,
+      "grad_norm": 0.4674701392650604,
+      "learning_rate": 1.5302002704532192e-05,
+      "loss": 0.0462,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5402658326817826,
+      "grad_norm": 0.4781097173690796,
+      "learning_rate": 1.5252904070492035e-05,
+      "loss": 0.0506,
+      "step": 1970
+    },
+    {
+      "epoch": 1.541047693510555,
+      "grad_norm": 0.2934138774871826,
+      "learning_rate": 1.5203870151838884e-05,
+      "loss": 0.0477,
+      "step": 1971
+    },
+    {
+      "epoch": 1.5418295543393277,
+      "grad_norm": 0.5006688833236694,
+      "learning_rate": 1.515490103989689e-05,
+      "loss": 0.0495,
+      "step": 1972
+    },
+    {
+      "epoch": 1.5426114151681,
+      "grad_norm": 0.3465513586997986,
+      "learning_rate": 1.51059968258695e-05,
+      "loss": 0.0478,
+      "step": 1973
+    },
+    {
+      "epoch": 1.5433932759968725,
+      "grad_norm": 0.3733910620212555,
+      "learning_rate": 1.5057157600839305e-05,
+      "loss": 0.0645,
+      "step": 1974
+    },
+    {
+      "epoch": 1.5441751368256451,
+      "grad_norm": 0.37032419443130493,
+      "learning_rate": 1.5008383455767828e-05,
+      "loss": 0.0594,
+      "step": 1975
+    },
+    {
+      "epoch": 1.5449569976544175,
+      "grad_norm": 0.41008704900741577,
+      "learning_rate": 1.4959674481495412e-05,
+      "loss": 0.0647,
+      "step": 1976
+    },
+    {
+      "epoch": 1.54573885848319,
+      "grad_norm": 0.37868285179138184,
+      "learning_rate": 1.4911030768741003e-05,
+      "loss": 0.0403,
+      "step": 1977
+    },
+    {
+      "epoch": 1.5465207193119626,
+      "grad_norm": 0.3398186266422272,
+      "learning_rate": 1.4862452408101996e-05,
+      "loss": 0.0485,
+      "step": 1978
+    },
+    {
+      "epoch": 1.547302580140735,
+      "grad_norm": 0.37676557898521423,
+      "learning_rate": 1.4813939490054095e-05,
+      "loss": 0.0696,
+      "step": 1979
+    },
+    {
+      "epoch": 1.5480844409695074,
+      "grad_norm": 0.3662286698818207,
+      "learning_rate": 1.4765492104951107e-05,
+      "loss": 0.0509,
+      "step": 1980
+    },
+    {
+      "epoch": 1.54886630179828,
+      "grad_norm": 0.37704092264175415,
+      "learning_rate": 1.471711034302477e-05,
+      "loss": 0.0449,
+      "step": 1981
+    },
+    {
+      "epoch": 1.5496481626270524,
+      "grad_norm": 0.37568730115890503,
+      "learning_rate": 1.466879429438462e-05,
+      "loss": 0.0686,
+      "step": 1982
+    },
+    {
+      "epoch": 1.5504300234558248,
+      "grad_norm": 0.3329296112060547,
+      "learning_rate": 1.4620544049017787e-05,
+      "loss": 0.0606,
+      "step": 1983
+    },
+    {
+      "epoch": 1.5512118842845974,
+      "grad_norm": 0.47085657715797424,
+      "learning_rate": 1.4572359696788868e-05,
+      "loss": 0.0732,
+      "step": 1984
+    },
+    {
+      "epoch": 1.5519937451133698,
+      "grad_norm": 0.6710319519042969,
+      "learning_rate": 1.4524241327439708e-05,
+      "loss": 0.1023,
+      "step": 1985
+    },
+    {
+      "epoch": 1.5527756059421423,
+      "grad_norm": 0.363131046295166,
+      "learning_rate": 1.4476189030589287e-05,
+      "loss": 0.0549,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5535574667709149,
+      "grad_norm": 0.334349662065506,
+      "learning_rate": 1.4428202895733505e-05,
+      "loss": 0.0427,
+      "step": 1987
+    },
+    {
+      "epoch": 1.5543393275996873,
+      "grad_norm": 0.3576214611530304,
+      "learning_rate": 1.4380283012245044e-05,
+      "loss": 0.0437,
+      "step": 1988
+    },
+    {
+      "epoch": 1.5551211884284597,
+      "grad_norm": 0.47771093249320984,
+      "learning_rate": 1.4332429469373226e-05,
+      "loss": 0.0388,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5559030492572323,
+      "grad_norm": 0.3168260157108307,
+      "learning_rate": 1.4284642356243765e-05,
+      "loss": 0.0596,
+      "step": 1990
+    },
+    {
+      "epoch": 1.5566849100860047,
+      "grad_norm": 0.38129621744155884,
+      "learning_rate": 1.4236921761858685e-05,
+      "loss": 0.0956,
+      "step": 1991
+    },
+    {
+      "epoch": 1.5574667709147771,
+      "grad_norm": 0.39686211943626404,
+      "learning_rate": 1.4189267775096104e-05,
+      "loss": 0.0273,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5582486317435498,
+      "grad_norm": 0.39223000407218933,
+      "learning_rate": 1.4141680484710095e-05,
+      "loss": 0.05,
+      "step": 1993
+    },
+    {
+      "epoch": 1.5590304925723222,
+      "grad_norm": 0.3337058424949646,
+      "learning_rate": 1.4094159979330513e-05,
+      "loss": 0.0445,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5598123534010946,
+      "grad_norm": 0.37635454535484314,
+      "learning_rate": 1.4046706347462819e-05,
+      "loss": 0.0341,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5605942142298672,
+      "grad_norm": 0.35981684923171997,
+      "learning_rate": 1.3999319677487926e-05,
+      "loss": 0.0569,
+      "step": 1996
+    },
+    {
+      "epoch": 1.5613760750586394,
+      "grad_norm": 0.5779023170471191,
+      "learning_rate": 1.395200005766204e-05,
+      "loss": 0.079,
+      "step": 1997
+    },
+    {
+      "epoch": 1.562157935887412,
+      "grad_norm": 0.3641951382160187,
+      "learning_rate": 1.3904747576116462e-05,
+      "loss": 0.1088,
+      "step": 1998
+    },
+    {
+      "epoch": 1.5629397967161847,
+      "grad_norm": 0.5276214480400085,
+      "learning_rate": 1.3857562320857526e-05,
+      "loss": 0.072,
+      "step": 1999
+    },
+    {
+      "epoch": 1.5637216575449568,
+      "grad_norm": 0.3834344446659088,
+      "learning_rate": 1.3810444379766246e-05,
+      "loss": 0.0414,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5637216575449568,
+      "eval_loss": 0.12067654728889465,
+      "eval_runtime": 13.4086,
+      "eval_samples_per_second": 3.878,
+      "eval_steps_per_second": 0.97,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5645035183737295,
+      "grad_norm": 0.46198686957359314,
+      "learning_rate": 1.3763393840598338e-05,
+      "loss": 0.039,
+      "step": 2001
+    },
+    {
+      "epoch": 1.565285379202502,
+      "grad_norm": 0.3458808958530426,
+      "learning_rate": 1.3716410790983991e-05,
+      "loss": 0.0597,
+      "step": 2002
+    },
+    {
+      "epoch": 1.5660672400312743,
+      "grad_norm": 0.4747830033302307,
+      "learning_rate": 1.3669495318427666e-05,
+      "loss": 0.0455,
+      "step": 2003
+    },
+    {
+      "epoch": 1.566849100860047,
+      "grad_norm": 0.5279855132102966,
+      "learning_rate": 1.3622647510307968e-05,
+      "loss": 0.044,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5676309616888195,
+      "grad_norm": 0.5206402540206909,
+      "learning_rate": 1.3575867453877488e-05,
+      "loss": 0.0562,
+      "step": 2005
+    },
+    {
+      "epoch": 1.5684128225175917,
+      "grad_norm": 0.44958412647247314,
+      "learning_rate": 1.352915523626263e-05,
+      "loss": 0.0669,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5691946833463644,
+      "grad_norm": 0.30790039896965027,
+      "learning_rate": 1.3482510944463445e-05,
+      "loss": 0.0207,
+      "step": 2007
+    },
+    {
+      "epoch": 1.569976544175137,
+      "grad_norm": 0.4307234585285187,
+      "learning_rate": 1.3435934665353466e-05,
+      "loss": 0.0678,
+      "step": 2008
+    },
+    {
+      "epoch": 1.5707584050039092,
+      "grad_norm": 0.5340322852134705,
+      "learning_rate": 1.3389426485679607e-05,
+      "loss": 0.0401,
+      "step": 2009
+    },
+    {
+      "epoch": 1.5715402658326818,
+      "grad_norm": 0.3630266487598419,
+      "learning_rate": 1.334298649206187e-05,
+      "loss": 0.07,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5723221266614542,
+      "grad_norm": 0.437450110912323,
+      "learning_rate": 1.3296614770993293e-05,
+      "loss": 0.0637,
+      "step": 2011
+    },
+    {
+      "epoch": 1.5731039874902266,
+      "grad_norm": 0.3910987377166748,
+      "learning_rate": 1.32503114088398e-05,
+      "loss": 0.051,
+      "step": 2012
+    },
+    {
+      "epoch": 1.5738858483189992,
+      "grad_norm": 0.38609978556632996,
+      "learning_rate": 1.320407649183995e-05,
+      "loss": 0.0498,
+      "step": 2013
+    },
+    {
+      "epoch": 1.5746677091477717,
+      "grad_norm": 0.44294193387031555,
+      "learning_rate": 1.3157910106104836e-05,
+      "loss": 0.0445,
+      "step": 2014
+    },
+    {
+      "epoch": 1.575449569976544,
+      "grad_norm": 0.6343628764152527,
+      "learning_rate": 1.3111812337617924e-05,
+      "loss": 0.0584,
+      "step": 2015
+    },
+    {
+      "epoch": 1.5762314308053167,
+      "grad_norm": 0.3056366741657257,
+      "learning_rate": 1.3065783272234878e-05,
+      "loss": 0.0639,
+      "step": 2016
+    },
+    {
+      "epoch": 1.577013291634089,
+      "grad_norm": 0.3845876157283783,
+      "learning_rate": 1.3019822995683395e-05,
+      "loss": 0.0352,
+      "step": 2017
+    },
+    {
+      "epoch": 1.5777951524628615,
+      "grad_norm": 0.40483787655830383,
+      "learning_rate": 1.2973931593563048e-05,
+      "loss": 0.0618,
+      "step": 2018
+    },
+    {
+      "epoch": 1.5785770132916341,
+      "grad_norm": 0.4752943813800812,
+      "learning_rate": 1.2928109151345196e-05,
+      "loss": 0.0547,
+      "step": 2019
+    },
+    {
+      "epoch": 1.5793588741204065,
+      "grad_norm": 0.4694390296936035,
+      "learning_rate": 1.2882355754372672e-05,
+      "loss": 0.0574,
+      "step": 2020
+    },
+    {
+      "epoch": 1.580140734949179,
+      "grad_norm": 0.38313254714012146,
+      "learning_rate": 1.2836671487859754e-05,
+      "loss": 0.05,
+      "step": 2021
+    },
+    {
+      "epoch": 1.5809225957779516,
+      "grad_norm": 0.6127452254295349,
+      "learning_rate": 1.2791056436892013e-05,
+      "loss": 0.067,
+      "step": 2022
+    },
+    {
+      "epoch": 1.581704456606724,
+      "grad_norm": 0.4264466464519501,
+      "learning_rate": 1.274551068642601e-05,
+      "loss": 0.0547,
+      "step": 2023
+    },
+    {
+      "epoch": 1.5824863174354964,
+      "grad_norm": 0.37022411823272705,
+      "learning_rate": 1.2700034321289333e-05,
+      "loss": 0.0604,
+      "step": 2024
+    },
+    {
+      "epoch": 1.583268178264269,
+      "grad_norm": 0.4471640884876251,
+      "learning_rate": 1.2654627426180277e-05,
+      "loss": 0.0732,
+      "step": 2025
+    },
+    {
+      "epoch": 1.5840500390930414,
+      "grad_norm": 0.358602374792099,
+      "learning_rate": 1.2609290085667785e-05,
+      "loss": 0.0555,
+      "step": 2026
+    },
+    {
+      "epoch": 1.5848318999218138,
+      "grad_norm": 0.49344927072525024,
+      "learning_rate": 1.2564022384191243e-05,
+      "loss": 0.0525,
+      "step": 2027
+    },
+    {
+      "epoch": 1.5856137607505865,
+      "grad_norm": 0.43709656596183777,
+      "learning_rate": 1.2518824406060336e-05,
+      "loss": 0.0855,
+      "step": 2028
+    },
+    {
+      "epoch": 1.5863956215793589,
+      "grad_norm": 0.38000303506851196,
+      "learning_rate": 1.2473696235454896e-05,
+      "loss": 0.0459,
+      "step": 2029
+    },
+    {
+      "epoch": 1.5871774824081313,
+      "grad_norm": 0.3962831497192383,
+      "learning_rate": 1.2428637956424743e-05,
+      "loss": 0.0604,
+      "step": 2030
+    },
+    {
+      "epoch": 1.587959343236904,
+      "grad_norm": 0.42037636041641235,
+      "learning_rate": 1.2383649652889501e-05,
+      "loss": 0.0399,
+      "step": 2031
+    },
+    {
+      "epoch": 1.5887412040656763,
+      "grad_norm": 0.4292720556259155,
+      "learning_rate": 1.2338731408638532e-05,
+      "loss": 0.0737,
+      "step": 2032
+    },
+    {
+      "epoch": 1.5895230648944487,
+      "grad_norm": 0.496189683675766,
+      "learning_rate": 1.2293883307330622e-05,
+      "loss": 0.0595,
+      "step": 2033
+    },
+    {
+      "epoch": 1.5903049257232214,
+      "grad_norm": 0.49012449383735657,
+      "learning_rate": 1.2249105432493996e-05,
+      "loss": 0.0721,
+      "step": 2034
+    },
+    {
+      "epoch": 1.5910867865519938,
+      "grad_norm": 0.3142451345920563,
+      "learning_rate": 1.2204397867526069e-05,
+      "loss": 0.0889,
+      "step": 2035
+    },
+    {
+      "epoch": 1.5918686473807662,
+      "grad_norm": 0.48869624733924866,
+      "learning_rate": 1.2159760695693239e-05,
+      "loss": 0.0468,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5926505082095388,
+      "grad_norm": 0.3193237781524658,
+      "learning_rate": 1.2115194000130903e-05,
+      "loss": 0.0226,
+      "step": 2037
+    },
+    {
+      "epoch": 1.5934323690383112,
+      "grad_norm": 0.37936335802078247,
+      "learning_rate": 1.2070697863843127e-05,
+      "loss": 0.048,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5942142298670836,
+      "grad_norm": 0.3389032185077667,
+      "learning_rate": 1.202627236970259e-05,
+      "loss": 0.039,
+      "step": 2039
+    },
+    {
+      "epoch": 1.5949960906958562,
+      "grad_norm": 0.2734612226486206,
+      "learning_rate": 1.1981917600450387e-05,
+      "loss": 0.0353,
+      "step": 2040
+    },
+    {
+      "epoch": 1.5957779515246286,
+      "grad_norm": 0.3461781144142151,
+      "learning_rate": 1.1937633638695883e-05,
+      "loss": 0.0575,
+      "step": 2041
+    },
+    {
+      "epoch": 1.596559812353401,
+      "grad_norm": 0.3470762073993683,
+      "learning_rate": 1.1893420566916636e-05,
+      "loss": 0.0592,
+      "step": 2042
+    },
+    {
+      "epoch": 1.5973416731821737,
+      "grad_norm": 0.41323909163475037,
+      "learning_rate": 1.1849278467458048e-05,
+      "loss": 0.033,
+      "step": 2043
+    },
+    {
+      "epoch": 1.598123534010946,
+      "grad_norm": 0.3009279668331146,
+      "learning_rate": 1.1805207422533459e-05,
+      "loss": 0.0318,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5989053948397185,
+      "grad_norm": 0.3985391855239868,
+      "learning_rate": 1.1761207514223822e-05,
+      "loss": 0.0451,
+      "step": 2045
+    },
+    {
+      "epoch": 1.5996872556684911,
+      "grad_norm": 0.43245643377304077,
+      "learning_rate": 1.1717278824477556e-05,
+      "loss": 0.0732,
+      "step": 2046
+    },
+    {
+      "epoch": 1.6004691164972635,
+      "grad_norm": 0.5485223531723022,
+      "learning_rate": 1.1673421435110522e-05,
+      "loss": 0.08,
+      "step": 2047
+    },
+    {
+      "epoch": 1.601250977326036,
+      "grad_norm": 0.30518728494644165,
+      "learning_rate": 1.162963542780573e-05,
+      "loss": 0.0364,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6020328381548086,
+      "grad_norm": 0.32443472743034363,
+      "learning_rate": 1.1585920884113261e-05,
+      "loss": 0.0504,
+      "step": 2049
+    },
+    {
+      "epoch": 1.602814698983581,
+      "grad_norm": 0.5871180295944214,
+      "learning_rate": 1.1542277885450098e-05,
+      "loss": 0.0784,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6035965598123534,
+      "grad_norm": 0.321475088596344,
+      "learning_rate": 1.1498706513099949e-05,
+      "loss": 0.0671,
+      "step": 2051
+    },
+    {
+      "epoch": 1.604378420641126,
+      "grad_norm": 0.35943475365638733,
+      "learning_rate": 1.1455206848213196e-05,
+      "loss": 0.0512,
+      "step": 2052
+    },
+    {
+      "epoch": 1.6051602814698982,
+      "grad_norm": 0.40514686703681946,
+      "learning_rate": 1.1411778971806558e-05,
+      "loss": 0.0645,
+      "step": 2053
+    },
+    {
+      "epoch": 1.6059421422986708,
+      "grad_norm": 0.3564876317977905,
+      "learning_rate": 1.1368422964763115e-05,
+      "loss": 0.0744,
+      "step": 2054
+    },
+    {
+      "epoch": 1.6067240031274435,
+      "grad_norm": 0.533642590045929,
+      "learning_rate": 1.1325138907832122e-05,
+      "loss": 0.0548,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6075058639562156,
+      "grad_norm": 0.4948970377445221,
+      "learning_rate": 1.1281926881628735e-05,
+      "loss": 0.0604,
+      "step": 2056
+    },
+    {
+      "epoch": 1.6082877247849883,
+      "grad_norm": 0.5011610984802246,
+      "learning_rate": 1.1238786966634052e-05,
+      "loss": 0.0722,
+      "step": 2057
+    },
+    {
+      "epoch": 1.609069585613761,
+      "grad_norm": 0.28366154432296753,
+      "learning_rate": 1.1195719243194813e-05,
+      "loss": 0.0627,
+      "step": 2058
+    },
+    {
+      "epoch": 1.609851446442533,
+      "grad_norm": 0.3866322636604309,
+      "learning_rate": 1.1152723791523318e-05,
+      "loss": 0.073,
+      "step": 2059
+    },
+    {
+      "epoch": 1.6106333072713057,
+      "grad_norm": 0.30931589007377625,
+      "learning_rate": 1.1109800691697254e-05,
+      "loss": 0.0321,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6114151681000783,
+      "grad_norm": 0.3482475280761719,
+      "learning_rate": 1.1066950023659545e-05,
+      "loss": 0.0546,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6121970289288505,
+      "grad_norm": 0.4044942259788513,
+      "learning_rate": 1.102417186721828e-05,
+      "loss": 0.0679,
+      "step": 2062
+    },
+    {
+      "epoch": 1.6129788897576232,
+      "grad_norm": 0.34607014060020447,
+      "learning_rate": 1.0981466302046406e-05,
+      "loss": 0.0416,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6137607505863958,
+      "grad_norm": 0.34637823700904846,
+      "learning_rate": 1.0938833407681704e-05,
+      "loss": 0.0751,
+      "step": 2064
+    },
+    {
+      "epoch": 1.614542611415168,
+      "grad_norm": 0.43607819080352783,
+      "learning_rate": 1.0896273263526663e-05,
+      "loss": 0.0462,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6153244722439406,
+      "grad_norm": 0.37277066707611084,
+      "learning_rate": 1.0853785948848166e-05,
+      "loss": 0.0649,
+      "step": 2066
+    },
+    {
+      "epoch": 1.616106333072713,
+      "grad_norm": 0.3339214324951172,
+      "learning_rate": 1.0811371542777571e-05,
+      "loss": 0.0581,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6168881939014854,
+      "grad_norm": 0.398568719625473,
+      "learning_rate": 1.0769030124310364e-05,
+      "loss": 0.0612,
+      "step": 2068
+    },
+    {
+      "epoch": 1.617670054730258,
+      "grad_norm": 0.3037922978401184,
+      "learning_rate": 1.0726761772306137e-05,
+      "loss": 0.0366,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6184519155590305,
+      "grad_norm": 0.41374528408050537,
+      "learning_rate": 1.0684566565488374e-05,
+      "loss": 0.0875,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6192337763878029,
+      "grad_norm": 0.4728708863258362,
+      "learning_rate": 1.0642444582444322e-05,
+      "loss": 0.0492,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6200156372165755,
+      "grad_norm": 0.49290919303894043,
+      "learning_rate": 1.060039590162491e-05,
+      "loss": 0.0442,
+      "step": 2072
+    },
+    {
+      "epoch": 1.620797498045348,
+      "grad_norm": 0.382946252822876,
+      "learning_rate": 1.055842060134446e-05,
+      "loss": 0.0516,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6215793588741203,
+      "grad_norm": 0.4095887243747711,
+      "learning_rate": 1.0516518759780664e-05,
+      "loss": 0.0466,
+      "step": 2074
+    },
+    {
+      "epoch": 1.622361219702893,
+      "grad_norm": 0.48378729820251465,
+      "learning_rate": 1.0474690454974445e-05,
+      "loss": 0.0736,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6231430805316653,
+      "grad_norm": 0.4431527554988861,
+      "learning_rate": 1.043293576482966e-05,
+      "loss": 0.082,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6239249413604377,
+      "grad_norm": 0.40211930871009827,
+      "learning_rate": 1.0391254767113169e-05,
+      "loss": 0.0581,
+      "step": 2077
+    },
+    {
+      "epoch": 1.6247068021892104,
+      "grad_norm": 0.42459335923194885,
+      "learning_rate": 1.0349647539454532e-05,
+      "loss": 0.0887,
+      "step": 2078
+    },
+    {
+      "epoch": 1.6254886630179828,
+      "grad_norm": 0.4116522967815399,
+      "learning_rate": 1.0308114159345883e-05,
+      "loss": 0.0566,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6262705238467552,
+      "grad_norm": 0.5448094010353088,
+      "learning_rate": 1.0266654704141888e-05,
+      "loss": 0.0721,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6270523846755278,
+      "grad_norm": 0.45197612047195435,
+      "learning_rate": 1.0225269251059483e-05,
+      "loss": 0.0558,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6278342455043002,
+      "grad_norm": 0.2840746343135834,
+      "learning_rate": 1.0183957877177786e-05,
+      "loss": 0.0485,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6286161063330726,
+      "grad_norm": 0.4308018982410431,
+      "learning_rate": 1.0142720659437955e-05,
+      "loss": 0.0955,
+      "step": 2083
+    },
+    {
+      "epoch": 1.6293979671618453,
+      "grad_norm": 0.4573703110218048,
+      "learning_rate": 1.0101557674643003e-05,
+      "loss": 0.0686,
+      "step": 2084
+    },
+    {
+      "epoch": 1.6301798279906177,
+      "grad_norm": 0.4302104115486145,
+      "learning_rate": 1.0060468999457767e-05,
+      "loss": 0.0454,
+      "step": 2085
+    },
+    {
+      "epoch": 1.63096168881939,
+      "grad_norm": 0.388776034116745,
+      "learning_rate": 1.001945471040856e-05,
+      "loss": 0.0576,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6317435496481627,
+      "grad_norm": 0.5159154534339905,
+      "learning_rate": 9.978514883883266e-06,
+      "loss": 0.1033,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6325254104769351,
+      "grad_norm": 0.31607142090797424,
+      "learning_rate": 9.937649596131043e-06,
+      "loss": 0.0583,
+      "step": 2088
+    },
+    {
+      "epoch": 1.6333072713057075,
+      "grad_norm": 0.33039212226867676,
+      "learning_rate": 9.89685892326218e-06,
+      "loss": 0.0262,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6340891321344801,
+      "grad_norm": 0.376118540763855,
+      "learning_rate": 9.856142941248075e-06,
+      "loss": 0.0573,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6348709929632526,
+      "grad_norm": 0.46004632115364075,
+      "learning_rate": 9.815501725920972e-06,
+      "loss": 0.0527,
+      "step": 2091
+    },
+    {
+      "epoch": 1.635652853792025,
+      "grad_norm": 0.40226930379867554,
+      "learning_rate": 9.774935352973868e-06,
+      "loss": 0.0464,
+      "step": 2092
+    },
+    {
+      "epoch": 1.6364347146207976,
+      "grad_norm": 0.4005904197692871,
+      "learning_rate": 9.734443897960372e-06,
+      "loss": 0.0839,
+      "step": 2093
+    },
+    {
+      "epoch": 1.63721657544957,
+      "grad_norm": 0.39641985297203064,
+      "learning_rate": 9.694027436294567e-06,
+      "loss": 0.0678,
+      "step": 2094
+    },
+    {
+      "epoch": 1.6379984362783424,
+      "grad_norm": 0.6553030610084534,
+      "learning_rate": 9.653686043250848e-06,
+      "loss": 0.0562,
+      "step": 2095
+    },
+    {
+      "epoch": 1.638780297107115,
+      "grad_norm": 0.4170481860637665,
+      "learning_rate": 9.613419793963808e-06,
+      "loss": 0.0661,
+      "step": 2096
+    },
+    {
+      "epoch": 1.6395621579358874,
+      "grad_norm": 0.4887992739677429,
+      "learning_rate": 9.573228763428093e-06,
+      "loss": 0.0658,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6403440187646599,
+      "grad_norm": 0.40965110063552856,
+      "learning_rate": 9.533113026498264e-06,
+      "loss": 0.0377,
+      "step": 2098
+    },
+    {
+      "epoch": 1.6411258795934325,
+      "grad_norm": 0.3160760700702667,
+      "learning_rate": 9.493072657888597e-06,
+      "loss": 0.0654,
+      "step": 2099
+    },
+    {
+      "epoch": 1.6419077404222049,
+      "grad_norm": 0.3468204140663147,
+      "learning_rate": 9.45310773217306e-06,
+      "loss": 0.0482,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6426896012509773,
+      "grad_norm": 0.5230024456977844,
+      "learning_rate": 9.413218323785084e-06,
+      "loss": 0.081,
+      "step": 2101
+    },
+    {
+      "epoch": 1.64347146207975,
+      "grad_norm": 0.3241496980190277,
+      "learning_rate": 9.373404507017453e-06,
+      "loss": 0.0389,
+      "step": 2102
+    },
+    {
+      "epoch": 1.6442533229085223,
+      "grad_norm": 0.4153323173522949,
+      "learning_rate": 9.333666356022158e-06,
+      "loss": 0.0658,
+      "step": 2103
+    },
+    {
+      "epoch": 1.6450351837372947,
+      "grad_norm": 0.4620719254016876,
+      "learning_rate": 9.294003944810276e-06,
+      "loss": 0.0496,
+      "step": 2104
+    },
+    {
+      "epoch": 1.6458170445660674,
+      "grad_norm": 0.3673686385154724,
+      "learning_rate": 9.254417347251815e-06,
+      "loss": 0.0826,
+      "step": 2105
+    },
+    {
+      "epoch": 1.6465989053948398,
+      "grad_norm": 0.3582766056060791,
+      "learning_rate": 9.21490663707557e-06,
+      "loss": 0.0673,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6473807662236122,
+      "grad_norm": 0.3266817331314087,
+      "learning_rate": 9.175471887869042e-06,
+      "loss": 0.0259,
+      "step": 2107
+    },
+    {
+      "epoch": 1.6481626270523848,
+      "grad_norm": 0.4447278082370758,
+      "learning_rate": 9.136113173078221e-06,
+      "loss": 0.0419,
+      "step": 2108
+    },
+    {
+      "epoch": 1.648944487881157,
+      "grad_norm": 0.3423357605934143,
+      "learning_rate": 9.096830566007452e-06,
+      "loss": 0.0458,
+      "step": 2109
+    },
+    {
+      "epoch": 1.6497263487099296,
+      "grad_norm": 0.41473570466041565,
+      "learning_rate": 9.05762413981941e-06,
+      "loss": 0.0686,
+      "step": 2110
+    },
+    {
+      "epoch": 1.6505082095387023,
+      "grad_norm": 0.3200038969516754,
+      "learning_rate": 9.018493967534835e-06,
+      "loss": 0.0623,
+      "step": 2111
+    },
+    {
+      "epoch": 1.6512900703674744,
+      "grad_norm": 0.29377713799476624,
+      "learning_rate": 8.979440122032457e-06,
+      "loss": 0.0449,
+      "step": 2112
+    },
+    {
+      "epoch": 1.652071931196247,
+      "grad_norm": 0.3613704442977905,
+      "learning_rate": 8.940462676048855e-06,
+      "loss": 0.0421,
+      "step": 2113
+    },
+    {
+      "epoch": 1.6528537920250197,
+      "grad_norm": 0.3492453992366791,
+      "learning_rate": 8.901561702178302e-06,
+      "loss": 0.0306,
+      "step": 2114
+    },
+    {
+      "epoch": 1.6536356528537919,
+      "grad_norm": 0.38384905457496643,
+      "learning_rate": 8.862737272872657e-06,
+      "loss": 0.042,
+      "step": 2115
+    },
+    {
+      "epoch": 1.6544175136825645,
+      "grad_norm": 0.33163630962371826,
+      "learning_rate": 8.823989460441217e-06,
+      "loss": 0.0475,
+      "step": 2116
+    },
+    {
+      "epoch": 1.6551993745113371,
+      "grad_norm": 0.40485212206840515,
+      "learning_rate": 8.78531833705058e-06,
+      "loss": 0.0485,
+      "step": 2117
+    },
+    {
+      "epoch": 1.6559812353401093,
+      "grad_norm": 0.3084823191165924,
+      "learning_rate": 8.746723974724507e-06,
+      "loss": 0.0573,
+      "step": 2118
+    },
+    {
+      "epoch": 1.656763096168882,
+      "grad_norm": 0.34011852741241455,
+      "learning_rate": 8.708206445343791e-06,
+      "loss": 0.0419,
+      "step": 2119
+    },
+    {
+      "epoch": 1.6575449569976546,
+      "grad_norm": 0.4085788428783417,
+      "learning_rate": 8.669765820646159e-06,
+      "loss": 0.0725,
+      "step": 2120
+    },
+    {
+      "epoch": 1.6583268178264268,
+      "grad_norm": 0.35617566108703613,
+      "learning_rate": 8.631402172226061e-06,
+      "loss": 0.0325,
+      "step": 2121
+    },
+    {
+      "epoch": 1.6591086786551994,
+      "grad_norm": 0.40245357155799866,
+      "learning_rate": 8.59311557153461e-06,
+      "loss": 0.0632,
+      "step": 2122
+    },
+    {
+      "epoch": 1.6598905394839718,
+      "grad_norm": 0.3330320715904236,
+      "learning_rate": 8.554906089879411e-06,
+      "loss": 0.037,
+      "step": 2123
+    },
+    {
+      "epoch": 1.6606724003127442,
+      "grad_norm": 0.33040401339530945,
+      "learning_rate": 8.516773798424427e-06,
+      "loss": 0.0518,
+      "step": 2124
+    },
+    {
+      "epoch": 1.6614542611415168,
+      "grad_norm": 0.307596892118454,
+      "learning_rate": 8.478718768189875e-06,
+      "loss": 0.0298,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6622361219702892,
+      "grad_norm": 0.400358110666275,
+      "learning_rate": 8.440741070052067e-06,
+      "loss": 0.0582,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6630179827990617,
+      "grad_norm": 0.5020748376846313,
+      "learning_rate": 8.402840774743281e-06,
+      "loss": 0.0568,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6637998436278343,
+      "grad_norm": 0.41084474325180054,
+      "learning_rate": 8.365017952851645e-06,
+      "loss": 0.0482,
+      "step": 2128
+    },
+    {
+      "epoch": 1.6645817044566067,
+      "grad_norm": 0.46043017506599426,
+      "learning_rate": 8.327272674820974e-06,
+      "loss": 0.0481,
+      "step": 2129
+    },
+    {
+      "epoch": 1.665363565285379,
+      "grad_norm": 0.3039909899234772,
+      "learning_rate": 8.289605010950703e-06,
+      "loss": 0.0422,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6661454261141517,
+      "grad_norm": 0.4253079891204834,
+      "learning_rate": 8.252015031395672e-06,
+      "loss": 0.0292,
+      "step": 2131
+    },
+    {
+      "epoch": 1.6669272869429241,
+      "grad_norm": 0.38544708490371704,
+      "learning_rate": 8.214502806166052e-06,
+      "loss": 0.0468,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6677091477716965,
+      "grad_norm": 0.30641862750053406,
+      "learning_rate": 8.177068405127198e-06,
+      "loss": 0.0376,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6684910086004692,
+      "grad_norm": 0.45573583245277405,
+      "learning_rate": 8.139711897999514e-06,
+      "loss": 0.0327,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6692728694292416,
+      "grad_norm": 0.4065832197666168,
+      "learning_rate": 8.10243335435834e-06,
+      "loss": 0.1065,
+      "step": 2135
+    },
+    {
+      "epoch": 1.670054730258014,
+      "grad_norm": 0.36566224694252014,
+      "learning_rate": 8.065232843633807e-06,
+      "loss": 0.0524,
+      "step": 2136
+    },
+    {
+      "epoch": 1.6708365910867866,
+      "grad_norm": 0.3225889801979065,
+      "learning_rate": 8.028110435110709e-06,
+      "loss": 0.059,
+      "step": 2137
+    },
+    {
+      "epoch": 1.671618451915559,
+      "grad_norm": 0.5427262187004089,
+      "learning_rate": 7.991066197928372e-06,
+      "loss": 0.0879,
+      "step": 2138
+    },
+    {
+      "epoch": 1.6724003127443314,
+      "grad_norm": 0.42064690589904785,
+      "learning_rate": 7.954100201080538e-06,
+      "loss": 0.0341,
+      "step": 2139
+    },
+    {
+      "epoch": 1.673182173573104,
+      "grad_norm": 0.33844882249832153,
+      "learning_rate": 7.91721251341525e-06,
+      "loss": 0.0657,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6739640344018765,
+      "grad_norm": 0.41567763686180115,
+      "learning_rate": 7.880403203634657e-06,
+      "loss": 0.0418,
+      "step": 2141
+    },
+    {
+      "epoch": 1.6747458952306489,
+      "grad_norm": 0.4430031180381775,
+      "learning_rate": 7.84367234029495e-06,
+      "loss": 0.0918,
+      "step": 2142
+    },
+    {
+      "epoch": 1.6755277560594215,
+      "grad_norm": 0.4221014976501465,
+      "learning_rate": 7.807019991806247e-06,
+      "loss": 0.0791,
+      "step": 2143
+    },
+    {
+      "epoch": 1.676309616888194,
+      "grad_norm": 0.4085227847099304,
+      "learning_rate": 7.770446226432393e-06,
+      "loss": 0.0404,
+      "step": 2144
+    },
+    {
+      "epoch": 1.6770914777169663,
+      "grad_norm": 0.41148459911346436,
+      "learning_rate": 7.733951112290894e-06,
+      "loss": 0.0617,
+      "step": 2145
+    },
+    {
+      "epoch": 1.677873338545739,
+      "grad_norm": 0.3634051978588104,
+      "learning_rate": 7.69753471735276e-06,
+      "loss": 0.0581,
+      "step": 2146
+    },
+    {
+      "epoch": 1.6786551993745114,
+      "grad_norm": 0.5589910745620728,
+      "learning_rate": 7.661197109442409e-06,
+      "loss": 0.0699,
+      "step": 2147
+    },
+    {
+      "epoch": 1.6794370602032838,
+      "grad_norm": 0.321353942155838,
+      "learning_rate": 7.624938356237493e-06,
+      "loss": 0.0262,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6802189210320564,
+      "grad_norm": 0.43340280652046204,
+      "learning_rate": 7.588758525268808e-06,
+      "loss": 0.0697,
+      "step": 2149
+    },
+    {
+      "epoch": 1.6810007818608288,
+      "grad_norm": 0.42228564620018005,
+      "learning_rate": 7.5526576839202e-06,
+      "loss": 0.0596,
+      "step": 2150
+    },
+    {
+      "epoch": 1.6817826426896012,
+      "grad_norm": 0.39023756980895996,
+      "learning_rate": 7.516635899428331e-06,
+      "loss": 0.0439,
+      "step": 2151
+    },
+    {
+      "epoch": 1.6825645035183738,
+      "grad_norm": 0.4297489523887634,
+      "learning_rate": 7.480693238882658e-06,
+      "loss": 0.0642,
+      "step": 2152
+    },
+    {
+      "epoch": 1.6833463643471462,
+      "grad_norm": 0.3283405601978302,
+      "learning_rate": 7.444829769225286e-06,
+      "loss": 0.0585,
+      "step": 2153
+    },
+    {
+      "epoch": 1.6841282251759186,
+      "grad_norm": 0.3524952232837677,
+      "learning_rate": 7.4090455572508044e-06,
+      "loss": 0.0281,
+      "step": 2154
+    },
+    {
+      "epoch": 1.6849100860046913,
+      "grad_norm": 0.34404295682907104,
+      "learning_rate": 7.373340669606205e-06,
+      "loss": 0.0488,
+      "step": 2155
+    },
+    {
+      "epoch": 1.6856919468334637,
+      "grad_norm": 0.3423098623752594,
+      "learning_rate": 7.337715172790721e-06,
+      "loss": 0.0451,
+      "step": 2156
+    },
+    {
+      "epoch": 1.686473807662236,
+      "grad_norm": 0.3290472626686096,
+      "learning_rate": 7.30216913315574e-06,
+      "loss": 0.0489,
+      "step": 2157
+    },
+    {
+      "epoch": 1.6872556684910087,
+      "grad_norm": 0.38891029357910156,
+      "learning_rate": 7.266702616904653e-06,
+      "loss": 0.0571,
+      "step": 2158
+    },
+    {
+      "epoch": 1.6880375293197811,
+      "grad_norm": 0.36488309502601624,
+      "learning_rate": 7.231315690092733e-06,
+      "loss": 0.0782,
+      "step": 2159
+    },
+    {
+      "epoch": 1.6888193901485535,
+      "grad_norm": 0.3920009136199951,
+      "learning_rate": 7.19600841862707e-06,
+      "loss": 0.0322,
+      "step": 2160
+    },
+    {
+      "epoch": 1.6896012509773262,
+      "grad_norm": 0.3348820209503174,
+      "learning_rate": 7.1607808682663315e-06,
+      "loss": 0.0661,
+      "step": 2161
+    },
+    {
+      "epoch": 1.6903831118060983,
+      "grad_norm": 0.3700244426727295,
+      "learning_rate": 7.125633104620722e-06,
+      "loss": 0.0241,
+      "step": 2162
+    },
+    {
+      "epoch": 1.691164972634871,
+      "grad_norm": 0.32485297322273254,
+      "learning_rate": 7.090565193151905e-06,
+      "loss": 0.0433,
+      "step": 2163
+    },
+    {
+      "epoch": 1.6919468334636436,
+      "grad_norm": 0.411578506231308,
+      "learning_rate": 7.05557719917273e-06,
+      "loss": 0.0608,
+      "step": 2164
+    },
+    {
+      "epoch": 1.6927286942924158,
+      "grad_norm": 0.2988610863685608,
+      "learning_rate": 7.020669187847278e-06,
+      "loss": 0.0583,
+      "step": 2165
+    },
+    {
+      "epoch": 1.6935105551211884,
+      "grad_norm": 0.31292635202407837,
+      "learning_rate": 6.985841224190625e-06,
+      "loss": 0.0471,
+      "step": 2166
+    },
+    {
+      "epoch": 1.694292415949961,
+      "grad_norm": 0.3985885679721832,
+      "learning_rate": 6.951093373068779e-06,
+      "loss": 0.036,
+      "step": 2167
+    },
+    {
+      "epoch": 1.6950742767787332,
+      "grad_norm": 0.3803912401199341,
+      "learning_rate": 6.916425699198526e-06,
+      "loss": 0.0593,
+      "step": 2168
+    },
+    {
+      "epoch": 1.6958561376075059,
+      "grad_norm": 0.5552504658699036,
+      "learning_rate": 6.881838267147334e-06,
+      "loss": 0.0904,
+      "step": 2169
+    },
+    {
+      "epoch": 1.6966379984362785,
+      "grad_norm": 0.32160112261772156,
+      "learning_rate": 6.847331141333224e-06,
+      "loss": 0.0445,
+      "step": 2170
+    },
+    {
+      "epoch": 1.6974198592650507,
+      "grad_norm": 0.33420655131340027,
+      "learning_rate": 6.812904386024644e-06,
+      "loss": 0.0653,
+      "step": 2171
+    },
+    {
+      "epoch": 1.6982017200938233,
+      "grad_norm": 0.43210169672966003,
+      "learning_rate": 6.77855806534034e-06,
+      "loss": 0.0518,
+      "step": 2172
+    },
+    {
+      "epoch": 1.698983580922596,
+      "grad_norm": 0.3932843804359436,
+      "learning_rate": 6.744292243249306e-06,
+      "loss": 0.0666,
+      "step": 2173
+    },
+    {
+      "epoch": 1.6997654417513681,
+      "grad_norm": 0.37392768263816833,
+      "learning_rate": 6.71010698357053e-06,
+      "loss": 0.0488,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7005473025801408,
+      "grad_norm": 0.4128730595111847,
+      "learning_rate": 6.676002349973027e-06,
+      "loss": 0.0503,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7013291634089132,
+      "grad_norm": 0.511454164981842,
+      "learning_rate": 6.641978405975618e-06,
+      "loss": 0.0639,
+      "step": 2176
+    },
+    {
+      "epoch": 1.7021110242376856,
+      "grad_norm": 0.4749528169631958,
+      "learning_rate": 6.608035214946806e-06,
+      "loss": 0.0356,
+      "step": 2177
+    },
+    {
+      "epoch": 1.7028928850664582,
+      "grad_norm": 0.35844314098358154,
+      "learning_rate": 6.574172840104764e-06,
+      "loss": 0.0618,
+      "step": 2178
+    },
+    {
+      "epoch": 1.7036747458952306,
+      "grad_norm": 0.361444890499115,
+      "learning_rate": 6.540391344517105e-06,
+      "loss": 0.0326,
+      "step": 2179
+    },
+    {
+      "epoch": 1.704456606724003,
+      "grad_norm": 0.4428097605705261,
+      "learning_rate": 6.5066907911008165e-06,
+      "loss": 0.0527,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7052384675527756,
+      "grad_norm": 0.3516511023044586,
+      "learning_rate": 6.473071242622131e-06,
+      "loss": 0.0325,
+      "step": 2181
+    },
+    {
+      "epoch": 1.706020328381548,
+      "grad_norm": 0.3570480942726135,
+      "learning_rate": 6.439532761696399e-06,
+      "loss": 0.0347,
+      "step": 2182
+    },
+    {
+      "epoch": 1.7068021892103205,
+      "grad_norm": 0.49399954080581665,
+      "learning_rate": 6.406075410788037e-06,
+      "loss": 0.0473,
+      "step": 2183
+    },
+    {
+      "epoch": 1.707584050039093,
+      "grad_norm": 0.36502349376678467,
+      "learning_rate": 6.372699252210274e-06,
+      "loss": 0.0332,
+      "step": 2184
+    },
+    {
+      "epoch": 1.7083659108678655,
+      "grad_norm": 0.30636587738990784,
+      "learning_rate": 6.339404348125205e-06,
+      "loss": 0.0461,
+      "step": 2185
+    },
+    {
+      "epoch": 1.709147771696638,
+      "grad_norm": 0.4163917899131775,
+      "learning_rate": 6.3061907605435565e-06,
+      "loss": 0.0454,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7099296325254105,
+      "grad_norm": 0.33639055490493774,
+      "learning_rate": 6.273058551324568e-06,
+      "loss": 0.0515,
+      "step": 2187
+    },
+    {
+      "epoch": 1.710711493354183,
+      "grad_norm": 0.3798970580101013,
+      "learning_rate": 6.240007782175983e-06,
+      "loss": 0.0692,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7114933541829553,
+      "grad_norm": 0.5719696283340454,
+      "learning_rate": 6.207038514653818e-06,
+      "loss": 0.0819,
+      "step": 2189
+    },
+    {
+      "epoch": 1.712275215011728,
+      "grad_norm": 0.36644482612609863,
+      "learning_rate": 6.174150810162315e-06,
+      "loss": 0.0527,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7130570758405004,
+      "grad_norm": 0.34109747409820557,
+      "learning_rate": 6.141344729953802e-06,
+      "loss": 0.0265,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7138389366692728,
+      "grad_norm": 0.46717700362205505,
+      "learning_rate": 6.108620335128573e-06,
+      "loss": 0.0647,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7146207974980454,
+      "grad_norm": 0.28352048993110657,
+      "learning_rate": 6.075977686634831e-06,
+      "loss": 0.0276,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7154026583268178,
+      "grad_norm": 0.4177485406398773,
+      "learning_rate": 6.043416845268457e-06,
+      "loss": 0.0571,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7161845191555902,
+      "grad_norm": 0.4173346161842346,
+      "learning_rate": 6.010937871673017e-06,
+      "loss": 0.0509,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7169663799843629,
+      "grad_norm": 0.3966216444969177,
+      "learning_rate": 5.978540826339607e-06,
+      "loss": 0.0503,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7177482408131353,
+      "grad_norm": 0.36653560400009155,
+      "learning_rate": 5.94622576960669e-06,
+      "loss": 0.0518,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7185301016419077,
+      "grad_norm": 0.31493353843688965,
+      "learning_rate": 5.913992761660081e-06,
+      "loss": 0.0784,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7193119624706803,
+      "grad_norm": 0.5782929062843323,
+      "learning_rate": 5.88184186253275e-06,
+      "loss": 0.0672,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7200938232994527,
+      "grad_norm": 0.3939312696456909,
+      "learning_rate": 5.849773132104741e-06,
+      "loss": 0.0375,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7200938232994527,
+      "eval_loss": 0.12024590373039246,
+      "eval_runtime": 13.3539,
+      "eval_samples_per_second": 3.894,
+      "eval_steps_per_second": 0.973,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7208756841282251,
+      "grad_norm": 0.4152418375015259,
+      "learning_rate": 5.817786630103067e-06,
+      "loss": 0.0504,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7216575449569977,
+      "grad_norm": 0.4392826557159424,
+      "learning_rate": 5.785882416101595e-06,
+      "loss": 0.0873,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7224394057857702,
+      "grad_norm": 0.4337979853153229,
+      "learning_rate": 5.754060549520956e-06,
+      "loss": 0.0365,
+      "step": 2203
+    },
+    {
+      "epoch": 1.7232212666145426,
+      "grad_norm": 0.4514532685279846,
+      "learning_rate": 5.722321089628363e-06,
+      "loss": 0.059,
+      "step": 2204
+    },
+    {
+      "epoch": 1.7240031274433152,
+      "grad_norm": 0.5406986474990845,
+      "learning_rate": 5.690664095537568e-06,
+      "loss": 0.0937,
+      "step": 2205
+    },
+    {
+      "epoch": 1.7247849882720876,
+      "grad_norm": 0.470359742641449,
+      "learning_rate": 5.659089626208769e-06,
+      "loss": 0.0601,
+      "step": 2206
+    },
+    {
+      "epoch": 1.72556684910086,
+      "grad_norm": 0.39090287685394287,
+      "learning_rate": 5.627597740448398e-06,
+      "loss": 0.0379,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7263487099296326,
+      "grad_norm": 0.4238806664943695,
+      "learning_rate": 5.59618849690915e-06,
+      "loss": 0.0821,
+      "step": 2208
+    },
+    {
+      "epoch": 1.727130570758405,
+      "grad_norm": 0.41596564650535583,
+      "learning_rate": 5.5648619540897395e-06,
+      "loss": 0.0316,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7279124315871774,
+      "grad_norm": 0.47083380818367004,
+      "learning_rate": 5.533618170334898e-06,
+      "loss": 0.0437,
+      "step": 2210
+    },
+    {
+      "epoch": 1.72869429241595,
+      "grad_norm": 0.3975255787372589,
+      "learning_rate": 5.502457203835187e-06,
+      "loss": 0.0536,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7294761532447225,
+      "grad_norm": 0.3355897068977356,
+      "learning_rate": 5.47137911262694e-06,
+      "loss": 0.0505,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7302580140734949,
+      "grad_norm": 0.3098110854625702,
+      "learning_rate": 5.4403839545921595e-06,
+      "loss": 0.0384,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7310398749022675,
+      "grad_norm": 0.5140456557273865,
+      "learning_rate": 5.409471787458342e-06,
+      "loss": 0.0551,
+      "step": 2214
+    },
+    {
+      "epoch": 1.73182173573104,
+      "grad_norm": 0.3814539611339569,
+      "learning_rate": 5.378642668798428e-06,
+      "loss": 0.0835,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7326035965598123,
+      "grad_norm": 0.4870288670063019,
+      "learning_rate": 5.347896656030738e-06,
+      "loss": 0.0863,
+      "step": 2216
+    },
+    {
+      "epoch": 1.733385457388585,
+      "grad_norm": 0.406101793050766,
+      "learning_rate": 5.317233806418708e-06,
+      "loss": 0.0615,
+      "step": 2217
+    },
+    {
+      "epoch": 1.7341673182173571,
+      "grad_norm": 0.3656727373600006,
+      "learning_rate": 5.286654177070988e-06,
+      "loss": 0.05,
+      "step": 2218
+    },
+    {
+      "epoch": 1.7349491790461298,
+      "grad_norm": 0.5171136856079102,
+      "learning_rate": 5.2561578249411824e-06,
+      "loss": 0.0475,
+      "step": 2219
+    },
+    {
+      "epoch": 1.7357310398749024,
+      "grad_norm": 0.5292237997055054,
+      "learning_rate": 5.2257448068277705e-06,
+      "loss": 0.047,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7365129007036746,
+      "grad_norm": 0.40704137086868286,
+      "learning_rate": 5.19541517937408e-06,
+      "loss": 0.0386,
+      "step": 2221
+    },
+    {
+      "epoch": 1.7372947615324472,
+      "grad_norm": 0.36376097798347473,
+      "learning_rate": 5.165168999068098e-06,
+      "loss": 0.0565,
+      "step": 2222
+    },
+    {
+      "epoch": 1.7380766223612198,
+      "grad_norm": 0.5038173794746399,
+      "learning_rate": 5.135006322242386e-06,
+      "loss": 0.0557,
+      "step": 2223
+    },
+    {
+      "epoch": 1.738858483189992,
+      "grad_norm": 0.3662266433238983,
+      "learning_rate": 5.1049272050740005e-06,
+      "loss": 0.0541,
+      "step": 2224
+    },
+    {
+      "epoch": 1.7396403440187647,
+      "grad_norm": 0.483214795589447,
+      "learning_rate": 5.074931703584352e-06,
+      "loss": 0.038,
+      "step": 2225
+    },
+    {
+      "epoch": 1.7404222048475373,
+      "grad_norm": 0.31406158208847046,
+      "learning_rate": 5.045019873639162e-06,
+      "loss": 0.0234,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7412040656763095,
+      "grad_norm": 0.4109937250614166,
+      "learning_rate": 5.015191770948241e-06,
+      "loss": 0.0489,
+      "step": 2227
+    },
+    {
+      "epoch": 1.741985926505082,
+      "grad_norm": 0.30827614665031433,
+      "learning_rate": 4.985447451065528e-06,
+      "loss": 0.0388,
+      "step": 2228
+    },
+    {
+      "epoch": 1.7427677873338547,
+      "grad_norm": 0.44419851899147034,
+      "learning_rate": 4.955786969388909e-06,
+      "loss": 0.0542,
+      "step": 2229
+    },
+    {
+      "epoch": 1.743549648162627,
+      "grad_norm": 0.3911994695663452,
+      "learning_rate": 4.926210381160074e-06,
+      "loss": 0.0371,
+      "step": 2230
+    },
+    {
+      "epoch": 1.7443315089913995,
+      "grad_norm": 0.30713513493537903,
+      "learning_rate": 4.896717741464524e-06,
+      "loss": 0.0444,
+      "step": 2231
+    },
+    {
+      "epoch": 1.745113369820172,
+      "grad_norm": 0.3942602276802063,
+      "learning_rate": 4.8673091052313826e-06,
+      "loss": 0.0603,
+      "step": 2232
+    },
+    {
+      "epoch": 1.7458952306489444,
+      "grad_norm": 0.36613914370536804,
+      "learning_rate": 4.837984527233314e-06,
+      "loss": 0.0444,
+      "step": 2233
+    },
+    {
+      "epoch": 1.746677091477717,
+      "grad_norm": 0.32284703850746155,
+      "learning_rate": 4.808744062086446e-06,
+      "loss": 0.0366,
+      "step": 2234
+    },
+    {
+      "epoch": 1.7474589523064894,
+      "grad_norm": 0.39046886563301086,
+      "learning_rate": 4.77958776425022e-06,
+      "loss": 0.0509,
+      "step": 2235
+    },
+    {
+      "epoch": 1.7482408131352618,
+      "grad_norm": 0.5282332301139832,
+      "learning_rate": 4.750515688027351e-06,
+      "loss": 0.0912,
+      "step": 2236
+    },
+    {
+      "epoch": 1.7490226739640344,
+      "grad_norm": 0.490951806306839,
+      "learning_rate": 4.721527887563659e-06,
+      "loss": 0.0695,
+      "step": 2237
+    },
+    {
+      "epoch": 1.7498045347928068,
+      "grad_norm": 0.3287993371486664,
+      "learning_rate": 4.69262441684804e-06,
+      "loss": 0.0414,
+      "step": 2238
+    },
+    {
+      "epoch": 1.7505863956215792,
+      "grad_norm": 0.45448336005210876,
+      "learning_rate": 4.663805329712318e-06,
+      "loss": 0.0385,
+      "step": 2239
+    },
+    {
+      "epoch": 1.7513682564503519,
+      "grad_norm": 0.49318641424179077,
+      "learning_rate": 4.635070679831116e-06,
+      "loss": 0.0635,
+      "step": 2240
+    },
+    {
+      "epoch": 1.7521501172791243,
+      "grad_norm": 0.38292163610458374,
+      "learning_rate": 4.6064205207218546e-06,
+      "loss": 0.0357,
+      "step": 2241
+    },
+    {
+      "epoch": 1.7529319781078967,
+      "grad_norm": 0.3284981846809387,
+      "learning_rate": 4.5778549057445556e-06,
+      "loss": 0.024,
+      "step": 2242
+    },
+    {
+      "epoch": 1.7537138389366693,
+      "grad_norm": 0.35669028759002686,
+      "learning_rate": 4.549373888101793e-06,
+      "loss": 0.0276,
+      "step": 2243
+    },
+    {
+      "epoch": 1.7544956997654417,
+      "grad_norm": 0.4276807904243469,
+      "learning_rate": 4.5209775208385704e-06,
+      "loss": 0.039,
+      "step": 2244
+    },
+    {
+      "epoch": 1.7552775605942141,
+      "grad_norm": 0.36494016647338867,
+      "learning_rate": 4.492665856842249e-06,
+      "loss": 0.0509,
+      "step": 2245
+    },
+    {
+      "epoch": 1.7560594214229868,
+      "grad_norm": 0.30909621715545654,
+      "learning_rate": 4.464438948842414e-06,
+      "loss": 0.0242,
+      "step": 2246
+    },
+    {
+      "epoch": 1.7568412822517592,
+      "grad_norm": 0.3083038032054901,
+      "learning_rate": 4.4362968494108145e-06,
+      "loss": 0.0909,
+      "step": 2247
+    },
+    {
+      "epoch": 1.7576231430805316,
+      "grad_norm": 0.35672706365585327,
+      "learning_rate": 4.408239610961223e-06,
+      "loss": 0.0487,
+      "step": 2248
+    },
+    {
+      "epoch": 1.7584050039093042,
+      "grad_norm": 0.40354442596435547,
+      "learning_rate": 4.3802672857494006e-06,
+      "loss": 0.0621,
+      "step": 2249
+    },
+    {
+      "epoch": 1.7591868647380766,
+      "grad_norm": 0.532529354095459,
+      "learning_rate": 4.352379925872901e-06,
+      "loss": 0.0866,
+      "step": 2250
+    },
+    {
+      "epoch": 1.759968725566849,
+      "grad_norm": 0.39215683937072754,
+      "learning_rate": 4.324577583271089e-06,
+      "loss": 0.0339,
+      "step": 2251
+    },
+    {
+      "epoch": 1.7607505863956217,
+      "grad_norm": 0.383793443441391,
+      "learning_rate": 4.29686030972496e-06,
+      "loss": 0.0491,
+      "step": 2252
+    },
+    {
+      "epoch": 1.761532447224394,
+      "grad_norm": 0.49607357382774353,
+      "learning_rate": 4.269228156857069e-06,
+      "loss": 0.0486,
+      "step": 2253
+    },
+    {
+      "epoch": 1.7623143080531665,
+      "grad_norm": 0.43889257311820984,
+      "learning_rate": 4.241681176131451e-06,
+      "loss": 0.0564,
+      "step": 2254
+    },
+    {
+      "epoch": 1.763096168881939,
+      "grad_norm": 0.3303699195384979,
+      "learning_rate": 4.2142194188534934e-06,
+      "loss": 0.046,
+      "step": 2255
+    },
+    {
+      "epoch": 1.7638780297107115,
+      "grad_norm": 0.3569354712963104,
+      "learning_rate": 4.186842936169877e-06,
+      "loss": 0.0357,
+      "step": 2256
+    },
+    {
+      "epoch": 1.764659890539484,
+      "grad_norm": 0.3961658179759979,
+      "learning_rate": 4.159551779068438e-06,
+      "loss": 0.0538,
+      "step": 2257
+    },
+    {
+      "epoch": 1.7654417513682565,
+      "grad_norm": 0.38139376044273376,
+      "learning_rate": 4.132345998378112e-06,
+      "loss": 0.0565,
+      "step": 2258
+    },
+    {
+      "epoch": 1.766223612197029,
+      "grad_norm": 0.38264214992523193,
+      "learning_rate": 4.1052256447688285e-06,
+      "loss": 0.0461,
+      "step": 2259
+    },
+    {
+      "epoch": 1.7670054730258014,
+      "grad_norm": 0.34434980154037476,
+      "learning_rate": 4.078190768751389e-06,
+      "loss": 0.0373,
+      "step": 2260
+    },
+    {
+      "epoch": 1.767787333854574,
+      "grad_norm": 0.4508064091205597,
+      "learning_rate": 4.051241420677427e-06,
+      "loss": 0.0995,
+      "step": 2261
+    },
+    {
+      "epoch": 1.7685691946833464,
+      "grad_norm": 0.6263595223426819,
+      "learning_rate": 4.024377650739264e-06,
+      "loss": 0.0725,
+      "step": 2262
+    },
+    {
+      "epoch": 1.7693510555121188,
+      "grad_norm": 0.3575042486190796,
+      "learning_rate": 3.997599508969829e-06,
+      "loss": 0.0541,
+      "step": 2263
+    },
+    {
+      "epoch": 1.7701329163408914,
+      "grad_norm": 0.5154090523719788,
+      "learning_rate": 3.970907045242583e-06,
+      "loss": 0.0544,
+      "step": 2264
+    },
+    {
+      "epoch": 1.7709147771696638,
+      "grad_norm": 0.362200528383255,
+      "learning_rate": 3.9443003092714095e-06,
+      "loss": 0.0471,
+      "step": 2265
+    },
+    {
+      "epoch": 1.7716966379984362,
+      "grad_norm": 0.4568684995174408,
+      "learning_rate": 3.9177793506105285e-06,
+      "loss": 0.0969,
+      "step": 2266
+    },
+    {
+      "epoch": 1.7724784988272089,
+      "grad_norm": 0.3077591359615326,
+      "learning_rate": 3.891344218654403e-06,
+      "loss": 0.0279,
+      "step": 2267
+    },
+    {
+      "epoch": 1.7732603596559813,
+      "grad_norm": 0.4141885042190552,
+      "learning_rate": 3.864994962637647e-06,
+      "loss": 0.0481,
+      "step": 2268
+    },
+    {
+      "epoch": 1.7740422204847537,
+      "grad_norm": 0.4456418454647064,
+      "learning_rate": 3.8387316316349285e-06,
+      "loss": 0.0602,
+      "step": 2269
+    },
+    {
+      "epoch": 1.7748240813135263,
+      "grad_norm": 0.32831570506095886,
+      "learning_rate": 3.8125542745608756e-06,
+      "loss": 0.0255,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7756059421422987,
+      "grad_norm": 0.429839164018631,
+      "learning_rate": 3.7864629401700214e-06,
+      "loss": 0.0412,
+      "step": 2271
+    },
+    {
+      "epoch": 1.7763878029710711,
+      "grad_norm": 0.4667169749736786,
+      "learning_rate": 3.7604576770566724e-06,
+      "loss": 0.0715,
+      "step": 2272
+    },
+    {
+      "epoch": 1.7771696637998438,
+      "grad_norm": 0.34892037510871887,
+      "learning_rate": 3.7345385336547856e-06,
+      "loss": 0.0806,
+      "step": 2273
+    },
+    {
+      "epoch": 1.777951524628616,
+      "grad_norm": 0.5732429623603821,
+      "learning_rate": 3.7087055582379957e-06,
+      "loss": 0.062,
+      "step": 2274
+    },
+    {
+      "epoch": 1.7787333854573886,
+      "grad_norm": 0.44712191820144653,
+      "learning_rate": 3.682958798919406e-06,
+      "loss": 0.0449,
+      "step": 2275
+    },
+    {
+      "epoch": 1.7795152462861612,
+      "grad_norm": 0.28738516569137573,
+      "learning_rate": 3.6572983036515572e-06,
+      "loss": 0.058,
+      "step": 2276
+    },
+    {
+      "epoch": 1.7802971071149334,
+      "grad_norm": 0.401889443397522,
+      "learning_rate": 3.631724120226321e-06,
+      "loss": 0.0501,
+      "step": 2277
+    },
+    {
+      "epoch": 1.781078967943706,
+      "grad_norm": 0.37569504976272583,
+      "learning_rate": 3.6062362962748176e-06,
+      "loss": 0.0419,
+      "step": 2278
+    },
+    {
+      "epoch": 1.7818608287724786,
+      "grad_norm": 0.39430519938468933,
+      "learning_rate": 3.5808348792673364e-06,
+      "loss": 0.0551,
+      "step": 2279
+    },
+    {
+      "epoch": 1.7826426896012508,
+      "grad_norm": 0.3892035484313965,
+      "learning_rate": 3.5555199165132168e-06,
+      "loss": 0.0325,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7834245504300235,
+      "grad_norm": 0.4657302796840668,
+      "learning_rate": 3.5302914551608112e-06,
+      "loss": 0.0612,
+      "step": 2281
+    },
+    {
+      "epoch": 1.784206411258796,
+      "grad_norm": 0.4596342444419861,
+      "learning_rate": 3.5051495421973223e-06,
+      "loss": 0.0566,
+      "step": 2282
+    },
+    {
+      "epoch": 1.7849882720875683,
+      "grad_norm": 0.39085909724235535,
+      "learning_rate": 3.480094224448788e-06,
+      "loss": 0.0394,
+      "step": 2283
+    },
+    {
+      "epoch": 1.785770132916341,
+      "grad_norm": 0.4541622996330261,
+      "learning_rate": 3.4551255485799627e-06,
+      "loss": 0.0432,
+      "step": 2284
+    },
+    {
+      "epoch": 1.7865519937451135,
+      "grad_norm": 0.4123796224594116,
+      "learning_rate": 3.4302435610942372e-06,
+      "loss": 0.0525,
+      "step": 2285
+    },
+    {
+      "epoch": 1.7873338545738857,
+      "grad_norm": 0.5204493403434753,
+      "learning_rate": 3.4054483083335255e-06,
+      "loss": 0.0732,
+      "step": 2286
+    },
+    {
+      "epoch": 1.7881157154026583,
+      "grad_norm": 0.41093361377716064,
+      "learning_rate": 3.3807398364782307e-06,
+      "loss": 0.04,
+      "step": 2287
+    },
+    {
+      "epoch": 1.7888975762314308,
+      "grad_norm": 0.4056856334209442,
+      "learning_rate": 3.3561181915471028e-06,
+      "loss": 0.0627,
+      "step": 2288
+    },
+    {
+      "epoch": 1.7896794370602032,
+      "grad_norm": 0.3035411238670349,
+      "learning_rate": 3.331583419397194e-06,
+      "loss": 0.0433,
+      "step": 2289
+    },
+    {
+      "epoch": 1.7904612978889758,
+      "grad_norm": 0.42910125851631165,
+      "learning_rate": 3.307135565723746e-06,
+      "loss": 0.067,
+      "step": 2290
+    },
+    {
+      "epoch": 1.7912431587177482,
+      "grad_norm": 0.38780465722084045,
+      "learning_rate": 3.2827746760601573e-06,
+      "loss": 0.0511,
+      "step": 2291
+    },
+    {
+      "epoch": 1.7920250195465206,
+      "grad_norm": 0.3865264654159546,
+      "learning_rate": 3.2585007957777967e-06,
+      "loss": 0.0544,
+      "step": 2292
+    },
+    {
+      "epoch": 1.7928068803752932,
+      "grad_norm": 0.354108065366745,
+      "learning_rate": 3.2343139700860168e-06,
+      "loss": 0.0591,
+      "step": 2293
+    },
+    {
+      "epoch": 1.7935887412040656,
+      "grad_norm": 0.37514495849609375,
+      "learning_rate": 3.2102142440320447e-06,
+      "loss": 0.0441,
+      "step": 2294
+    },
+    {
+      "epoch": 1.794370602032838,
+      "grad_norm": 0.38472533226013184,
+      "learning_rate": 3.186201662500865e-06,
+      "loss": 0.0433,
+      "step": 2295
+    },
+    {
+      "epoch": 1.7951524628616107,
+      "grad_norm": 0.41590240597724915,
+      "learning_rate": 3.1622762702151585e-06,
+      "loss": 0.0644,
+      "step": 2296
+    },
+    {
+      "epoch": 1.795934323690383,
+      "grad_norm": 0.43187642097473145,
+      "learning_rate": 3.138438111735231e-06,
+      "loss": 0.0474,
+      "step": 2297
+    },
+    {
+      "epoch": 1.7967161845191555,
+      "grad_norm": 0.4443909525871277,
+      "learning_rate": 3.1146872314589127e-06,
+      "loss": 0.0495,
+      "step": 2298
+    },
+    {
+      "epoch": 1.7974980453479281,
+      "grad_norm": 0.38686707615852356,
+      "learning_rate": 3.0910236736214794e-06,
+      "loss": 0.0621,
+      "step": 2299
+    },
+    {
+      "epoch": 1.7982799061767005,
+      "grad_norm": 0.47035327553749084,
+      "learning_rate": 3.067447482295577e-06,
+      "loss": 0.0824,
+      "step": 2300
+    },
+    {
+      "epoch": 1.799061767005473,
+      "grad_norm": 0.4474782645702362,
+      "learning_rate": 3.043958701391114e-06,
+      "loss": 0.0807,
+      "step": 2301
+    },
+    {
+      "epoch": 1.7998436278342456,
+      "grad_norm": 0.3506847023963928,
+      "learning_rate": 3.0205573746552364e-06,
+      "loss": 0.0351,
+      "step": 2302
+    },
+    {
+      "epoch": 1.800625488663018,
+      "grad_norm": 0.39736175537109375,
+      "learning_rate": 2.9972435456721627e-06,
+      "loss": 0.0495,
+      "step": 2303
+    },
+    {
+      "epoch": 1.8014073494917904,
+      "grad_norm": 0.35211628675460815,
+      "learning_rate": 2.9740172578632096e-06,
+      "loss": 0.0339,
+      "step": 2304
+    },
+    {
+      "epoch": 1.802189210320563,
+      "grad_norm": 0.42447611689567566,
+      "learning_rate": 2.9508785544865856e-06,
+      "loss": 0.0452,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8029710711493354,
+      "grad_norm": 0.426352858543396,
+      "learning_rate": 2.9278274786374237e-06,
+      "loss": 0.043,
+      "step": 2306
+    },
+    {
+      "epoch": 1.8037529319781078,
+      "grad_norm": 0.4002751111984253,
+      "learning_rate": 2.9048640732476317e-06,
+      "loss": 0.0583,
+      "step": 2307
+    },
+    {
+      "epoch": 1.8045347928068804,
+      "grad_norm": 0.4455183148384094,
+      "learning_rate": 2.8819883810858428e-06,
+      "loss": 0.065,
+      "step": 2308
+    },
+    {
+      "epoch": 1.8053166536356529,
+      "grad_norm": 0.432157039642334,
+      "learning_rate": 2.8592004447573207e-06,
+      "loss": 0.0337,
+      "step": 2309
+    },
+    {
+      "epoch": 1.8060985144644253,
+      "grad_norm": 0.3376258611679077,
+      "learning_rate": 2.836500306703882e-06,
+      "loss": 0.0523,
+      "step": 2310
+    },
+    {
+      "epoch": 1.806880375293198,
+      "grad_norm": 0.3891136050224304,
+      "learning_rate": 2.813888009203841e-06,
+      "loss": 0.0566,
+      "step": 2311
+    },
+    {
+      "epoch": 1.8076622361219703,
+      "grad_norm": 0.48112839460372925,
+      "learning_rate": 2.791363594371893e-06,
+      "loss": 0.0487,
+      "step": 2312
+    },
+    {
+      "epoch": 1.8084440969507427,
+      "grad_norm": 0.422264039516449,
+      "learning_rate": 2.768927104159058e-06,
+      "loss": 0.0889,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8092259577795153,
+      "grad_norm": 0.37410929799079895,
+      "learning_rate": 2.7465785803526155e-06,
+      "loss": 0.0477,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8100078186082877,
+      "grad_norm": 0.4331490099430084,
+      "learning_rate": 2.724318064575976e-06,
+      "loss": 0.0385,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8107896794370602,
+      "grad_norm": 0.3773462772369385,
+      "learning_rate": 2.7021455982886747e-06,
+      "loss": 0.0225,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8115715402658328,
+      "grad_norm": 0.40394389629364014,
+      "learning_rate": 2.6800612227862453e-06,
+      "loss": 0.0407,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8123534010946052,
+      "grad_norm": 0.6558087468147278,
+      "learning_rate": 2.658064979200131e-06,
+      "loss": 0.0767,
+      "step": 2318
+    },
+    {
+      "epoch": 1.8131352619233776,
+      "grad_norm": 0.359089732170105,
+      "learning_rate": 2.6361569084976723e-06,
+      "loss": 0.0374,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8139171227521502,
+      "grad_norm": 0.47283199429512024,
+      "learning_rate": 2.6143370514819634e-06,
+      "loss": 0.0693,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8146989835809226,
+      "grad_norm": 0.35094255208969116,
+      "learning_rate": 2.592605448791807e-06,
+      "loss": 0.0371,
+      "step": 2321
+    },
+    {
+      "epoch": 1.815480844409695,
+      "grad_norm": 0.2955813705921173,
+      "learning_rate": 2.5709621409016394e-06,
+      "loss": 0.0491,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8162627052384677,
+      "grad_norm": 0.46730920672416687,
+      "learning_rate": 2.5494071681214483e-06,
+      "loss": 0.0713,
+      "step": 2323
+    },
+    {
+      "epoch": 1.81704456606724,
+      "grad_norm": 0.36728736758232117,
+      "learning_rate": 2.5279405705967097e-06,
+      "loss": 0.0405,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8178264268960125,
+      "grad_norm": 0.42313656210899353,
+      "learning_rate": 2.5065623883082867e-06,
+      "loss": 0.0345,
+      "step": 2325
+    },
+    {
+      "epoch": 1.818608287724785,
+      "grad_norm": 0.33062589168548584,
+      "learning_rate": 2.4852726610723743e-06,
+      "loss": 0.0648,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8193901485535575,
+      "grad_norm": 0.37844690680503845,
+      "learning_rate": 2.4640714285404544e-06,
+      "loss": 0.0443,
+      "step": 2327
+    },
+    {
+      "epoch": 1.82017200938233,
+      "grad_norm": 0.4094715118408203,
+      "learning_rate": 2.442958730199124e-06,
+      "loss": 0.0486,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8209538702111026,
+      "grad_norm": 0.4287872016429901,
+      "learning_rate": 2.421934605370163e-06,
+      "loss": 0.0509,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8217357310398747,
+      "grad_norm": 0.38323545455932617,
+      "learning_rate": 2.4009990932103322e-06,
+      "loss": 0.0498,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8225175918686474,
+      "grad_norm": 0.45034465193748474,
+      "learning_rate": 2.3801522327113802e-06,
+      "loss": 0.0921,
+      "step": 2331
+    },
+    {
+      "epoch": 1.82329945269742,
+      "grad_norm": 0.4322207272052765,
+      "learning_rate": 2.359394062699932e-06,
+      "loss": 0.0529,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8240813135261922,
+      "grad_norm": 0.5762638449668884,
+      "learning_rate": 2.338724621837435e-06,
+      "loss": 0.0752,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8248631743549648,
+      "grad_norm": 0.5516983270645142,
+      "learning_rate": 2.318143948620094e-06,
+      "loss": 0.0571,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8256450351837374,
+      "grad_norm": 0.3905304968357086,
+      "learning_rate": 2.2976520813787594e-06,
+      "loss": 0.0457,
+      "step": 2335
+    },
+    {
+      "epoch": 1.8264268960125096,
+      "grad_norm": 0.4931924343109131,
+      "learning_rate": 2.2772490582788965e-06,
+      "loss": 0.0542,
+      "step": 2336
+    },
+    {
+      "epoch": 1.8272087568412823,
+      "grad_norm": 0.42153871059417725,
+      "learning_rate": 2.2569349173205133e-06,
+      "loss": 0.0591,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8279906176700549,
+      "grad_norm": 0.35189369320869446,
+      "learning_rate": 2.236709696338052e-06,
+      "loss": 0.0484,
+      "step": 2338
+    },
+    {
+      "epoch": 1.828772478498827,
+      "grad_norm": 0.4322154223918915,
+      "learning_rate": 2.2165734330003686e-06,
+      "loss": 0.0376,
+      "step": 2339
+    },
+    {
+      "epoch": 1.8295543393275997,
+      "grad_norm": 0.3702199459075928,
+      "learning_rate": 2.1965261648106305e-06,
+      "loss": 0.045,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8303362001563723,
+      "grad_norm": 0.3591911792755127,
+      "learning_rate": 2.176567929106249e-06,
+      "loss": 0.0333,
+      "step": 2341
+    },
+    {
+      "epoch": 1.8311180609851445,
+      "grad_norm": 0.4833166003227234,
+      "learning_rate": 2.156698763058812e-06,
+      "loss": 0.0701,
+      "step": 2342
+    },
+    {
+      "epoch": 1.8318999218139171,
+      "grad_norm": 0.38630276918411255,
+      "learning_rate": 2.1369187036740235e-06,
+      "loss": 0.0843,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8326817826426895,
+      "grad_norm": 0.33287540078163147,
+      "learning_rate": 2.117227787791648e-06,
+      "loss": 0.0382,
+      "step": 2344
+    },
+    {
+      "epoch": 1.833463643471462,
+      "grad_norm": 0.3747866451740265,
+      "learning_rate": 2.0976260520853886e-06,
+      "loss": 0.0516,
+      "step": 2345
+    },
+    {
+      "epoch": 1.8342455043002346,
+      "grad_norm": 0.502683699131012,
+      "learning_rate": 2.078113533062864e-06,
+      "loss": 0.0794,
+      "step": 2346
+    },
+    {
+      "epoch": 1.835027365129007,
+      "grad_norm": 0.40196695923805237,
+      "learning_rate": 2.0586902670655606e-06,
+      "loss": 0.0402,
+      "step": 2347
+    },
+    {
+      "epoch": 1.8358092259577794,
+      "grad_norm": 0.3394660949707031,
+      "learning_rate": 2.0393562902686737e-06,
+      "loss": 0.0576,
+      "step": 2348
+    },
+    {
+      "epoch": 1.836591086786552,
+      "grad_norm": 0.3652074337005615,
+      "learning_rate": 2.02011163868116e-06,
+      "loss": 0.0459,
+      "step": 2349
+    },
+    {
+      "epoch": 1.8373729476153244,
+      "grad_norm": 0.3400147557258606,
+      "learning_rate": 2.0009563481455775e-06,
+      "loss": 0.0484,
+      "step": 2350
+    },
+    {
+      "epoch": 1.8381548084440968,
+      "grad_norm": 0.43868163228034973,
+      "learning_rate": 1.9818904543380643e-06,
+      "loss": 0.0645,
+      "step": 2351
+    },
+    {
+      "epoch": 1.8389366692728695,
+      "grad_norm": 0.5077487826347351,
+      "learning_rate": 1.962913992768245e-06,
+      "loss": 0.0913,
+      "step": 2352
+    },
+    {
+      "epoch": 1.8397185301016419,
+      "grad_norm": 0.34455615282058716,
+      "learning_rate": 1.9440269987791914e-06,
+      "loss": 0.0661,
+      "step": 2353
+    },
+    {
+      "epoch": 1.8405003909304143,
+      "grad_norm": 0.3783716857433319,
+      "learning_rate": 1.92522950754736e-06,
+      "loss": 0.0284,
+      "step": 2354
+    },
+    {
+      "epoch": 1.841282251759187,
+      "grad_norm": 0.5611537098884583,
+      "learning_rate": 1.90652155408248e-06,
+      "loss": 0.0454,
+      "step": 2355
+    },
+    {
+      "epoch": 1.8420641125879593,
+      "grad_norm": 0.39524754881858826,
+      "learning_rate": 1.8879031732275231e-06,
+      "loss": 0.0629,
+      "step": 2356
+    },
+    {
+      "epoch": 1.8428459734167317,
+      "grad_norm": 0.5189679265022278,
+      "learning_rate": 1.8693743996586742e-06,
+      "loss": 0.0333,
+      "step": 2357
+    },
+    {
+      "epoch": 1.8436278342455044,
+      "grad_norm": 0.2966693341732025,
+      "learning_rate": 1.8509352678851732e-06,
+      "loss": 0.0427,
+      "step": 2358
+    },
+    {
+      "epoch": 1.8444096950742768,
+      "grad_norm": 0.38795343041419983,
+      "learning_rate": 1.8325858122493432e-06,
+      "loss": 0.0576,
+      "step": 2359
+    },
+    {
+      "epoch": 1.8451915559030492,
+      "grad_norm": 0.44317272305488586,
+      "learning_rate": 1.814326066926475e-06,
+      "loss": 0.0593,
+      "step": 2360
+    },
+    {
+      "epoch": 1.8459734167318218,
+      "grad_norm": 0.3784867525100708,
+      "learning_rate": 1.7961560659247646e-06,
+      "loss": 0.0499,
+      "step": 2361
+    },
+    {
+      "epoch": 1.8467552775605942,
+      "grad_norm": 0.39091184735298157,
+      "learning_rate": 1.7780758430852917e-06,
+      "loss": 0.0553,
+      "step": 2362
+    },
+    {
+      "epoch": 1.8475371383893666,
+      "grad_norm": 0.3573996424674988,
+      "learning_rate": 1.7600854320819038e-06,
+      "loss": 0.0552,
+      "step": 2363
+    },
+    {
+      "epoch": 1.8483189992181392,
+      "grad_norm": 0.40562674403190613,
+      "learning_rate": 1.7421848664211759e-06,
+      "loss": 0.0969,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8491008600469117,
+      "grad_norm": 0.35688963532447815,
+      "learning_rate": 1.7243741794423619e-06,
+      "loss": 0.0608,
+      "step": 2365
+    },
+    {
+      "epoch": 1.849882720875684,
+      "grad_norm": 0.44463032484054565,
+      "learning_rate": 1.7066534043173098e-06,
+      "loss": 0.0487,
+      "step": 2366
+    },
+    {
+      "epoch": 1.8506645817044567,
+      "grad_norm": 0.44885513186454773,
+      "learning_rate": 1.6890225740504251e-06,
+      "loss": 0.0674,
+      "step": 2367
+    },
+    {
+      "epoch": 1.851446442533229,
+      "grad_norm": 0.4468289911746979,
+      "learning_rate": 1.6714817214785572e-06,
+      "loss": 0.0468,
+      "step": 2368
+    },
+    {
+      "epoch": 1.8522283033620015,
+      "grad_norm": 0.3201262652873993,
+      "learning_rate": 1.6540308792710235e-06,
+      "loss": 0.0447,
+      "step": 2369
+    },
+    {
+      "epoch": 1.8530101641907741,
+      "grad_norm": 0.4483521282672882,
+      "learning_rate": 1.63667007992947e-06,
+      "loss": 0.0535,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8537920250195465,
+      "grad_norm": 0.39991673827171326,
+      "learning_rate": 1.6193993557878317e-06,
+      "loss": 0.0374,
+      "step": 2371
+    },
+    {
+      "epoch": 1.854573885848319,
+      "grad_norm": 0.33672747015953064,
+      "learning_rate": 1.6022187390123067e-06,
+      "loss": 0.0369,
+      "step": 2372
+    },
+    {
+      "epoch": 1.8553557466770916,
+      "grad_norm": 0.4697551131248474,
+      "learning_rate": 1.5851282616012653e-06,
+      "loss": 0.0679,
+      "step": 2373
+    },
+    {
+      "epoch": 1.856137607505864,
+      "grad_norm": 0.3470526933670044,
+      "learning_rate": 1.5681279553851846e-06,
+      "loss": 0.0429,
+      "step": 2374
+    },
+    {
+      "epoch": 1.8569194683346364,
+      "grad_norm": 0.340726375579834,
+      "learning_rate": 1.5512178520266096e-06,
+      "loss": 0.0374,
+      "step": 2375
+    },
+    {
+      "epoch": 1.857701329163409,
+      "grad_norm": 0.4430573582649231,
+      "learning_rate": 1.5343979830200695e-06,
+      "loss": 0.0436,
+      "step": 2376
+    },
+    {
+      "epoch": 1.8584831899921814,
+      "grad_norm": 0.44432032108306885,
+      "learning_rate": 1.5176683796920721e-06,
+      "loss": 0.0379,
+      "step": 2377
+    },
+    {
+      "epoch": 1.8592650508209538,
+      "grad_norm": 0.6266927719116211,
+      "learning_rate": 1.5010290732009658e-06,
+      "loss": 0.0946,
+      "step": 2378
+    },
+    {
+      "epoch": 1.8600469116497265,
+      "grad_norm": 0.40188395977020264,
+      "learning_rate": 1.4844800945369498e-06,
+      "loss": 0.0716,
+      "step": 2379
+    },
+    {
+      "epoch": 1.8608287724784989,
+      "grad_norm": 0.35556766390800476,
+      "learning_rate": 1.46802147452198e-06,
+      "loss": 0.0546,
+      "step": 2380
+    },
+    {
+      "epoch": 1.8616106333072713,
+      "grad_norm": 0.4273186922073364,
+      "learning_rate": 1.4516532438097196e-06,
+      "loss": 0.0794,
+      "step": 2381
+    },
+    {
+      "epoch": 1.862392494136044,
+      "grad_norm": 0.3742946982383728,
+      "learning_rate": 1.435375432885505e-06,
+      "loss": 0.0438,
+      "step": 2382
+    },
+    {
+      "epoch": 1.8631743549648163,
+      "grad_norm": 0.32148027420043945,
+      "learning_rate": 1.419188072066241e-06,
+      "loss": 0.044,
+      "step": 2383
+    },
+    {
+      "epoch": 1.8639562157935887,
+      "grad_norm": 0.4895723760128021,
+      "learning_rate": 1.403091191500383e-06,
+      "loss": 0.0544,
+      "step": 2384
+    },
+    {
+      "epoch": 1.8647380766223614,
+      "grad_norm": 0.390939325094223,
+      "learning_rate": 1.3870848211678833e-06,
+      "loss": 0.0558,
+      "step": 2385
+    },
+    {
+      "epoch": 1.8655199374511335,
+      "grad_norm": 0.35615217685699463,
+      "learning_rate": 1.371168990880095e-06,
+      "loss": 0.0446,
+      "step": 2386
+    },
+    {
+      "epoch": 1.8663017982799062,
+      "grad_norm": 0.3843987286090851,
+      "learning_rate": 1.3553437302797733e-06,
+      "loss": 0.0667,
+      "step": 2387
+    },
+    {
+      "epoch": 1.8670836591086788,
+      "grad_norm": 0.3831058442592621,
+      "learning_rate": 1.339609068840958e-06,
+      "loss": 0.0407,
+      "step": 2388
+    },
+    {
+      "epoch": 1.867865519937451,
+      "grad_norm": 0.4777224361896515,
+      "learning_rate": 1.3239650358689737e-06,
+      "loss": 0.0642,
+      "step": 2389
+    },
+    {
+      "epoch": 1.8686473807662236,
+      "grad_norm": 0.2945762574672699,
+      "learning_rate": 1.3084116605003582e-06,
+      "loss": 0.0595,
+      "step": 2390
+    },
+    {
+      "epoch": 1.8694292415949962,
+      "grad_norm": 0.35123440623283386,
+      "learning_rate": 1.292948971702773e-06,
+      "loss": 0.0378,
+      "step": 2391
+    },
+    {
+      "epoch": 1.8702111024237684,
+      "grad_norm": 0.287575900554657,
+      "learning_rate": 1.2775769982750041e-06,
+      "loss": 0.0515,
+      "step": 2392
+    },
+    {
+      "epoch": 1.870992963252541,
+      "grad_norm": 0.3514576256275177,
+      "learning_rate": 1.2622957688468717e-06,
+      "loss": 0.083,
+      "step": 2393
+    },
+    {
+      "epoch": 1.8717748240813137,
+      "grad_norm": 0.45359691977500916,
+      "learning_rate": 1.247105311879193e-06,
+      "loss": 0.0826,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8725566849100859,
+      "grad_norm": 0.3990536332130432,
+      "learning_rate": 1.2320056556637205e-06,
+      "loss": 0.0373,
+      "step": 2395
+    },
+    {
+      "epoch": 1.8733385457388585,
+      "grad_norm": 0.5825475454330444,
+      "learning_rate": 1.216996828323097e-06,
+      "loss": 0.0875,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8741204065676311,
+      "grad_norm": 0.3500322997570038,
+      "learning_rate": 1.2020788578107956e-06,
+      "loss": 0.0368,
+      "step": 2397
+    },
+    {
+      "epoch": 1.8749022673964033,
+      "grad_norm": 0.4058021605014801,
+      "learning_rate": 1.1872517719110743e-06,
+      "loss": 0.0564,
+      "step": 2398
+    },
+    {
+      "epoch": 1.875684128225176,
+      "grad_norm": 0.551173746585846,
+      "learning_rate": 1.1725155982389158e-06,
+      "loss": 0.0547,
+      "step": 2399
+    },
+    {
+      "epoch": 1.8764659890539483,
+      "grad_norm": 0.3550439476966858,
+      "learning_rate": 1.1578703642399935e-06,
+      "loss": 0.0512,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8764659890539483,
+      "eval_loss": 0.12009644508361816,
+      "eval_runtime": 13.4244,
+      "eval_samples_per_second": 3.874,
+      "eval_steps_per_second": 0.968,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8772478498827208,
+      "grad_norm": 0.4831383526325226,
+      "learning_rate": 1.1433160971905942e-06,
+      "loss": 0.0737,
+      "step": 2401
+    },
+    {
+      "epoch": 1.8780297107114934,
+      "grad_norm": 0.3398386836051941,
+      "learning_rate": 1.1288528241975904e-06,
+      "loss": 0.0464,
+      "step": 2402
+    },
+    {
+      "epoch": 1.8788115715402658,
+      "grad_norm": 0.31560376286506653,
+      "learning_rate": 1.1144805721983841e-06,
+      "loss": 0.0351,
+      "step": 2403
+    },
+    {
+      "epoch": 1.8795934323690382,
+      "grad_norm": 0.47334128618240356,
+      "learning_rate": 1.1001993679608414e-06,
+      "loss": 0.0692,
+      "step": 2404
+    },
+    {
+      "epoch": 1.8803752931978108,
+      "grad_norm": 0.4987127482891083,
+      "learning_rate": 1.0860092380832632e-06,
+      "loss": 0.067,
+      "step": 2405
+    },
+    {
+      "epoch": 1.8811571540265832,
+      "grad_norm": 0.4054931402206421,
+      "learning_rate": 1.0719102089943257e-06,
+      "loss": 0.0585,
+      "step": 2406
+    },
+    {
+      "epoch": 1.8819390148553556,
+      "grad_norm": 0.5126636624336243,
+      "learning_rate": 1.0579023069530346e-06,
+      "loss": 0.0557,
+      "step": 2407
+    },
+    {
+      "epoch": 1.8827208756841283,
+      "grad_norm": 0.3793666958808899,
+      "learning_rate": 1.0439855580486714e-06,
+      "loss": 0.0457,
+      "step": 2408
+    },
+    {
+      "epoch": 1.8835027365129007,
+      "grad_norm": 0.5941369533538818,
+      "learning_rate": 1.0301599882007462e-06,
+      "loss": 0.0597,
+      "step": 2409
+    },
+    {
+      "epoch": 1.884284597341673,
+      "grad_norm": 0.396608829498291,
+      "learning_rate": 1.0164256231589563e-06,
+      "loss": 0.072,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8850664581704457,
+      "grad_norm": 0.3238108456134796,
+      "learning_rate": 1.0027824885031288e-06,
+      "loss": 0.0526,
+      "step": 2411
+    },
+    {
+      "epoch": 1.8858483189992181,
+      "grad_norm": 0.664912760257721,
+      "learning_rate": 9.892306096431826e-07,
+      "loss": 0.0993,
+      "step": 2412
+    },
+    {
+      "epoch": 1.8866301798279905,
+      "grad_norm": 0.38779711723327637,
+      "learning_rate": 9.75770011819066e-07,
+      "loss": 0.0276,
+      "step": 2413
+    },
+    {
+      "epoch": 1.8874120406567632,
+      "grad_norm": 0.42406222224235535,
+      "learning_rate": 9.624007201007257e-07,
+      "loss": 0.0347,
+      "step": 2414
+    },
+    {
+      "epoch": 1.8881939014855356,
+      "grad_norm": 0.4085008502006531,
+      "learning_rate": 9.491227593880492e-07,
+      "loss": 0.0391,
+      "step": 2415
+    },
+    {
+      "epoch": 1.888975762314308,
+      "grad_norm": 0.3823298513889313,
+      "learning_rate": 9.359361544108324e-07,
+      "loss": 0.0489,
+      "step": 2416
+    },
+    {
+      "epoch": 1.8897576231430806,
+      "grad_norm": 0.5838673114776611,
+      "learning_rate": 9.228409297287132e-07,
+      "loss": 0.0562,
+      "step": 2417
+    },
+    {
+      "epoch": 1.890539483971853,
+      "grad_norm": 0.41118496656417847,
+      "learning_rate": 9.098371097311431e-07,
+      "loss": 0.0428,
+      "step": 2418
+    },
+    {
+      "epoch": 1.8913213448006254,
+      "grad_norm": 0.3399001955986023,
+      "learning_rate": 8.96924718637332e-07,
+      "loss": 0.0582,
+      "step": 2419
+    },
+    {
+      "epoch": 1.892103205629398,
+      "grad_norm": 0.4081033170223236,
+      "learning_rate": 8.841037804962094e-07,
+      "loss": 0.0682,
+      "step": 2420
+    },
+    {
+      "epoch": 1.8928850664581705,
+      "grad_norm": 0.4025924801826477,
+      "learning_rate": 8.713743191863633e-07,
+      "loss": 0.0282,
+      "step": 2421
+    },
+    {
+      "epoch": 1.8936669272869429,
+      "grad_norm": 0.36863648891448975,
+      "learning_rate": 8.587363584160401e-07,
+      "loss": 0.0491,
+      "step": 2422
+    },
+    {
+      "epoch": 1.8944487881157155,
+      "grad_norm": 0.42005249857902527,
+      "learning_rate": 8.46189921723034e-07,
+      "loss": 0.0637,
+      "step": 2423
+    },
+    {
+      "epoch": 1.895230648944488,
+      "grad_norm": 0.48608678579330444,
+      "learning_rate": 8.337350324746917e-07,
+      "loss": 0.0535,
+      "step": 2424
+    },
+    {
+      "epoch": 1.8960125097732603,
+      "grad_norm": 0.3172799348831177,
+      "learning_rate": 8.21371713867869e-07,
+      "loss": 0.0356,
+      "step": 2425
+    },
+    {
+      "epoch": 1.896794370602033,
+      "grad_norm": 0.49111539125442505,
+      "learning_rate": 8.090999889288576e-07,
+      "loss": 0.0542,
+      "step": 2426
+    },
+    {
+      "epoch": 1.8975762314308053,
+      "grad_norm": 0.38987061381340027,
+      "learning_rate": 7.969198805133638e-07,
+      "loss": 0.0634,
+      "step": 2427
+    },
+    {
+      "epoch": 1.8983580922595777,
+      "grad_norm": 0.3516485393047333,
+      "learning_rate": 7.848314113064636e-07,
+      "loss": 0.0278,
+      "step": 2428
+    },
+    {
+      "epoch": 1.8991399530883504,
+      "grad_norm": 0.32461583614349365,
+      "learning_rate": 7.728346038225475e-07,
+      "loss": 0.0242,
+      "step": 2429
+    },
+    {
+      "epoch": 1.8999218139171228,
+      "grad_norm": 0.8025784492492676,
+      "learning_rate": 7.609294804053091e-07,
+      "loss": 0.0712,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9007036747458952,
+      "grad_norm": 0.33313748240470886,
+      "learning_rate": 7.491160632276562e-07,
+      "loss": 0.0563,
+      "step": 2431
+    },
+    {
+      "epoch": 1.9014855355746678,
+      "grad_norm": 0.39373156428337097,
+      "learning_rate": 7.373943742917277e-07,
+      "loss": 0.0381,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9022673964034402,
+      "grad_norm": 0.37192046642303467,
+      "learning_rate": 7.25764435428794e-07,
+      "loss": 0.0945,
+      "step": 2433
+    },
+    {
+      "epoch": 1.9030492572322126,
+      "grad_norm": 0.31872573494911194,
+      "learning_rate": 7.142262682992562e-07,
+      "loss": 0.0275,
+      "step": 2434
+    },
+    {
+      "epoch": 1.9038311180609853,
+      "grad_norm": 0.5576297044754028,
+      "learning_rate": 7.027798943925967e-07,
+      "loss": 0.0466,
+      "step": 2435
+    },
+    {
+      "epoch": 1.9046129788897577,
+      "grad_norm": 0.4052869975566864,
+      "learning_rate": 6.914253350273292e-07,
+      "loss": 0.0395,
+      "step": 2436
+    },
+    {
+      "epoch": 1.90539483971853,
+      "grad_norm": 0.30455952882766724,
+      "learning_rate": 6.801626113509651e-07,
+      "loss": 0.0558,
+      "step": 2437
+    },
+    {
+      "epoch": 1.9061767005473027,
+      "grad_norm": 0.6548801064491272,
+      "learning_rate": 6.689917443399862e-07,
+      "loss": 0.0499,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9069585613760751,
+      "grad_norm": 0.3970543444156647,
+      "learning_rate": 6.579127547997721e-07,
+      "loss": 0.0741,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9077404222048475,
+      "grad_norm": 0.4393087327480316,
+      "learning_rate": 6.469256633646059e-07,
+      "loss": 0.0449,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9085222830336201,
+      "grad_norm": 0.319876492023468,
+      "learning_rate": 6.360304904976022e-07,
+      "loss": 0.0605,
+      "step": 2441
+    },
+    {
+      "epoch": 1.9093041438623923,
+      "grad_norm": 0.4261433482170105,
+      "learning_rate": 6.252272564906848e-07,
+      "loss": 0.0418,
+      "step": 2442
+    },
+    {
+      "epoch": 1.910086004691165,
+      "grad_norm": 0.44413477182388306,
+      "learning_rate": 6.145159814645362e-07,
+      "loss": 0.0762,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9108678655199376,
+      "grad_norm": 0.3411540985107422,
+      "learning_rate": 6.038966853685712e-07,
+      "loss": 0.0413,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9116497263487098,
+      "grad_norm": 0.4325558543205261,
+      "learning_rate": 5.933693879809132e-07,
+      "loss": 0.0458,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9124315871774824,
+      "grad_norm": 0.5270968675613403,
+      "learning_rate": 5.829341089083118e-07,
+      "loss": 0.0636,
+      "step": 2446
+    },
+    {
+      "epoch": 1.913213448006255,
+      "grad_norm": 0.43004176020622253,
+      "learning_rate": 5.725908675861535e-07,
+      "loss": 0.0581,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9139953088350272,
+      "grad_norm": 0.5290845036506653,
+      "learning_rate": 5.623396832784123e-07,
+      "loss": 0.0673,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9147771696637998,
+      "grad_norm": 0.3586988151073456,
+      "learning_rate": 5.521805750775877e-07,
+      "loss": 0.0387,
+      "step": 2449
+    },
+    {
+      "epoch": 1.9155590304925725,
+      "grad_norm": 0.4046856760978699,
+      "learning_rate": 5.421135619047113e-07,
+      "loss": 0.0678,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9163408913213447,
+      "grad_norm": 0.3092903196811676,
+      "learning_rate": 5.32138662509285e-07,
+      "loss": 0.0489,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9171227521501173,
+      "grad_norm": 0.42022499442100525,
+      "learning_rate": 5.222558954692424e-07,
+      "loss": 0.0447,
+      "step": 2452
+    },
+    {
+      "epoch": 1.91790461297889,
+      "grad_norm": 0.4458214044570923,
+      "learning_rate": 5.124652791909324e-07,
+      "loss": 0.0447,
+      "step": 2453
+    },
+    {
+      "epoch": 1.918686473807662,
+      "grad_norm": 0.37331241369247437,
+      "learning_rate": 5.027668319090739e-07,
+      "loss": 0.0407,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9194683346364347,
+      "grad_norm": 0.37612366676330566,
+      "learning_rate": 4.931605716867293e-07,
+      "loss": 0.0404,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9202501954652071,
+      "grad_norm": 0.4228704273700714,
+      "learning_rate": 4.836465164152426e-07,
+      "loss": 0.0311,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9210320562939796,
+      "grad_norm": 0.4672371745109558,
+      "learning_rate": 4.742246838142672e-07,
+      "loss": 0.0624,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9218139171227522,
+      "grad_norm": 0.40904104709625244,
+      "learning_rate": 4.6489509143166633e-07,
+      "loss": 0.0491,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9225957779515246,
+      "grad_norm": 0.5268110632896423,
+      "learning_rate": 4.5565775664351275e-07,
+      "loss": 0.0688,
+      "step": 2459
+    },
+    {
+      "epoch": 1.923377638780297,
+      "grad_norm": 0.3935900628566742,
+      "learning_rate": 4.465126966540556e-07,
+      "loss": 0.0488,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9241594996090696,
+      "grad_norm": 0.4083102345466614,
+      "learning_rate": 4.3745992849568707e-07,
+      "loss": 0.0442,
+      "step": 2461
+    },
+    {
+      "epoch": 1.924941360437842,
+      "grad_norm": 0.3193533718585968,
+      "learning_rate": 4.28499469028909e-07,
+      "loss": 0.0385,
+      "step": 2462
+    },
+    {
+      "epoch": 1.9257232212666144,
+      "grad_norm": 0.48666051030158997,
+      "learning_rate": 4.196313349422942e-07,
+      "loss": 0.068,
+      "step": 2463
+    },
+    {
+      "epoch": 1.926505082095387,
+      "grad_norm": 0.42172297835350037,
+      "learning_rate": 4.1085554275246965e-07,
+      "loss": 0.0441,
+      "step": 2464
+    },
+    {
+      "epoch": 1.9272869429241595,
+      "grad_norm": 0.47432318329811096,
+      "learning_rate": 4.021721088040775e-07,
+      "loss": 0.0652,
+      "step": 2465
+    },
+    {
+      "epoch": 1.9280688037529319,
+      "grad_norm": 0.3693954646587372,
+      "learning_rate": 3.9358104926974225e-07,
+      "loss": 0.0727,
+      "step": 2466
+    },
+    {
+      "epoch": 1.9288506645817045,
+      "grad_norm": 0.5284464359283447,
+      "learning_rate": 3.8508238015003697e-07,
+      "loss": 0.0715,
+      "step": 2467
+    },
+    {
+      "epoch": 1.929632525410477,
+      "grad_norm": 0.43904706835746765,
+      "learning_rate": 3.7667611727348894e-07,
+      "loss": 0.0329,
+      "step": 2468
+    },
+    {
+      "epoch": 1.9304143862392493,
+      "grad_norm": 0.33485591411590576,
+      "learning_rate": 3.6836227629648e-07,
+      "loss": 0.0525,
+      "step": 2469
+    },
+    {
+      "epoch": 1.931196247068022,
+      "grad_norm": 0.34851041436195374,
+      "learning_rate": 3.6014087270329064e-07,
+      "loss": 0.0236,
+      "step": 2470
+    },
+    {
+      "epoch": 1.9319781078967944,
+      "grad_norm": 0.4184808135032654,
+      "learning_rate": 3.520119218060336e-07,
+      "loss": 0.0586,
+      "step": 2471
+    },
+    {
+      "epoch": 1.9327599687255668,
+      "grad_norm": 0.3958164155483246,
+      "learning_rate": 3.4397543874462037e-07,
+      "loss": 0.0626,
+      "step": 2472
+    },
+    {
+      "epoch": 1.9335418295543394,
+      "grad_norm": 0.3597150146961212,
+      "learning_rate": 3.360314384867558e-07,
+      "loss": 0.0873,
+      "step": 2473
+    },
+    {
+      "epoch": 1.9343236903831118,
+      "grad_norm": 0.4362090229988098,
+      "learning_rate": 3.281799358278881e-07,
+      "loss": 0.0284,
+      "step": 2474
+    },
+    {
+      "epoch": 1.9351055512118842,
+      "grad_norm": 0.4249899387359619,
+      "learning_rate": 3.2042094539120883e-07,
+      "loss": 0.0444,
+      "step": 2475
+    },
+    {
+      "epoch": 1.9358874120406568,
+      "grad_norm": 0.38876742124557495,
+      "learning_rate": 3.127544816275807e-07,
+      "loss": 0.0519,
+      "step": 2476
+    },
+    {
+      "epoch": 1.9366692728694292,
+      "grad_norm": 0.29063284397125244,
+      "learning_rate": 3.051805588155654e-07,
+      "loss": 0.0363,
+      "step": 2477
+    },
+    {
+      "epoch": 1.9374511336982017,
+      "grad_norm": 0.3670947849750519,
+      "learning_rate": 2.976991910613569e-07,
+      "loss": 0.0454,
+      "step": 2478
+    },
+    {
+      "epoch": 1.9382329945269743,
+      "grad_norm": 0.4057813286781311,
+      "learning_rate": 2.903103922987649e-07,
+      "loss": 0.0766,
+      "step": 2479
+    },
+    {
+      "epoch": 1.9390148553557467,
+      "grad_norm": 0.319303035736084,
+      "learning_rate": 2.8301417628920357e-07,
+      "loss": 0.0424,
+      "step": 2480
+    },
+    {
+      "epoch": 1.939796716184519,
+      "grad_norm": 0.33699148893356323,
+      "learning_rate": 2.7581055662164736e-07,
+      "loss": 0.057,
+      "step": 2481
+    },
+    {
+      "epoch": 1.9405785770132917,
+      "grad_norm": 0.3684520125389099,
+      "learning_rate": 2.6869954671261963e-07,
+      "loss": 0.0534,
+      "step": 2482
+    },
+    {
+      "epoch": 1.9413604378420641,
+      "grad_norm": 0.42329129576683044,
+      "learning_rate": 2.6168115980614303e-07,
+      "loss": 0.0476,
+      "step": 2483
+    },
+    {
+      "epoch": 1.9421422986708365,
+      "grad_norm": 0.34844499826431274,
+      "learning_rate": 2.547554089737503e-07,
+      "loss": 0.031,
+      "step": 2484
+    },
+    {
+      "epoch": 1.9429241594996092,
+      "grad_norm": 0.3760523796081543,
+      "learning_rate": 2.4792230711444544e-07,
+      "loss": 0.0544,
+      "step": 2485
+    },
+    {
+      "epoch": 1.9437060203283816,
+      "grad_norm": 0.45071330666542053,
+      "learning_rate": 2.411818669546595e-07,
+      "loss": 0.0371,
+      "step": 2486
+    },
+    {
+      "epoch": 1.944487881157154,
+      "grad_norm": 0.3400191068649292,
+      "learning_rate": 2.3453410104825046e-07,
+      "loss": 0.0458,
+      "step": 2487
+    },
+    {
+      "epoch": 1.9452697419859266,
+      "grad_norm": 0.45273348689079285,
+      "learning_rate": 2.2797902177648657e-07,
+      "loss": 0.0435,
+      "step": 2488
+    },
+    {
+      "epoch": 1.946051602814699,
+      "grad_norm": 0.49700218439102173,
+      "learning_rate": 2.2151664134799076e-07,
+      "loss": 0.0761,
+      "step": 2489
+    },
+    {
+      "epoch": 1.9468334636434714,
+      "grad_norm": 0.3172673285007477,
+      "learning_rate": 2.1514697179875198e-07,
+      "loss": 0.0512,
+      "step": 2490
+    },
+    {
+      "epoch": 1.947615324472244,
+      "grad_norm": 0.367340087890625,
+      "learning_rate": 2.0887002499207498e-07,
+      "loss": 0.0585,
+      "step": 2491
+    },
+    {
+      "epoch": 1.9483971853010165,
+      "grad_norm": 0.3881520926952362,
+      "learning_rate": 2.02685812618586e-07,
+      "loss": 0.042,
+      "step": 2492
+    },
+    {
+      "epoch": 1.9491790461297889,
+      "grad_norm": 0.5710947513580322,
+      "learning_rate": 1.9659434619617723e-07,
+      "loss": 0.0846,
+      "step": 2493
+    },
+    {
+      "epoch": 1.9499609069585615,
+      "grad_norm": 0.5186334252357483,
+      "learning_rate": 1.9059563707002903e-07,
+      "loss": 0.052,
+      "step": 2494
+    },
+    {
+      "epoch": 1.950742767787334,
+      "grad_norm": 0.3576108515262604,
+      "learning_rate": 1.846896964125433e-07,
+      "loss": 0.0623,
+      "step": 2495
+    },
+    {
+      "epoch": 1.9515246286161063,
+      "grad_norm": 0.3158591389656067,
+      "learning_rate": 1.78876535223349e-07,
+      "loss": 0.054,
+      "step": 2496
+    },
+    {
+      "epoch": 1.952306489444879,
+      "grad_norm": 0.3423866331577301,
+      "learning_rate": 1.731561643292856e-07,
+      "loss": 0.0535,
+      "step": 2497
+    },
+    {
+      "epoch": 1.9530883502736511,
+      "grad_norm": 0.4671078622341156,
+      "learning_rate": 1.6752859438435854e-07,
+      "loss": 0.0585,
+      "step": 2498
+    },
+    {
+      "epoch": 1.9538702111024238,
+      "grad_norm": 0.4123114049434662,
+      "learning_rate": 1.6199383586975037e-07,
+      "loss": 0.0384,
+      "step": 2499
+    },
+    {
+      "epoch": 1.9546520719311964,
+      "grad_norm": 0.4386757016181946,
+      "learning_rate": 1.5655189909377088e-07,
+      "loss": 0.0864,
+      "step": 2500
+    },
+    {
+      "epoch": 1.9554339327599686,
+      "grad_norm": 0.4988419711589813,
+      "learning_rate": 1.5120279419185701e-07,
+      "loss": 0.0741,
+      "step": 2501
+    },
+    {
+      "epoch": 1.9562157935887412,
+      "grad_norm": 0.37020736932754517,
+      "learning_rate": 1.459465311265562e-07,
+      "loss": 0.0434,
+      "step": 2502
+    },
+    {
+      "epoch": 1.9569976544175138,
+      "grad_norm": 0.32374125719070435,
+      "learning_rate": 1.4078311968749313e-07,
+      "loss": 0.0402,
+      "step": 2503
+    },
+    {
+      "epoch": 1.957779515246286,
+      "grad_norm": 0.4434984028339386,
+      "learning_rate": 1.357125694913586e-07,
+      "loss": 0.0483,
+      "step": 2504
+    },
+    {
+      "epoch": 1.9585613760750586,
+      "grad_norm": 0.3723717927932739,
+      "learning_rate": 1.307348899818983e-07,
+      "loss": 0.0529,
+      "step": 2505
+    },
+    {
+      "epoch": 1.9593432369038313,
+      "grad_norm": 0.45597562193870544,
+      "learning_rate": 1.2585009042987982e-07,
+      "loss": 0.0595,
+      "step": 2506
+    },
+    {
+      "epoch": 1.9601250977326035,
+      "grad_norm": 0.4516230821609497,
+      "learning_rate": 1.2105817993309786e-07,
+      "loss": 0.0558,
+      "step": 2507
+    },
+    {
+      "epoch": 1.960906958561376,
+      "grad_norm": 0.39071428775787354,
+      "learning_rate": 1.1635916741633557e-07,
+      "loss": 0.059,
+      "step": 2508
+    },
+    {
+      "epoch": 1.9616888193901487,
+      "grad_norm": 0.38830170035362244,
+      "learning_rate": 1.1175306163135335e-07,
+      "loss": 0.037,
+      "step": 2509
+    },
+    {
+      "epoch": 1.962470680218921,
+      "grad_norm": 0.50648432970047,
+      "learning_rate": 1.0723987115688339e-07,
+      "loss": 0.0698,
+      "step": 2510
+    },
+    {
+      "epoch": 1.9632525410476935,
+      "grad_norm": 0.5163297653198242,
+      "learning_rate": 1.0281960439860739e-07,
+      "loss": 0.0494,
+      "step": 2511
+    },
+    {
+      "epoch": 1.964034401876466,
+      "grad_norm": 0.3927709460258484,
+      "learning_rate": 9.849226958912328e-08,
+      "loss": 0.068,
+      "step": 2512
+    },
+    {
+      "epoch": 1.9648162627052383,
+      "grad_norm": 0.3887207806110382,
+      "learning_rate": 9.425787478796744e-08,
+      "loss": 0.0552,
+      "step": 2513
+    },
+    {
+      "epoch": 1.965598123534011,
+      "grad_norm": 0.35503289103507996,
+      "learning_rate": 9.011642788156472e-08,
+      "loss": 0.0488,
+      "step": 2514
+    },
+    {
+      "epoch": 1.9663799843627834,
+      "grad_norm": 0.5193012952804565,
+      "learning_rate": 8.606793658323398e-08,
+      "loss": 0.093,
+      "step": 2515
+    },
+    {
+      "epoch": 1.9671618451915558,
+      "grad_norm": 0.4710843861103058,
+      "learning_rate": 8.211240843316037e-08,
+      "loss": 0.0571,
+      "step": 2516
+    },
+    {
+      "epoch": 1.9679437060203284,
+      "grad_norm": 0.288714200258255,
+      "learning_rate": 7.824985079839531e-08,
+      "loss": 0.0294,
+      "step": 2517
+    },
+    {
+      "epoch": 1.9687255668491008,
+      "grad_norm": 0.5777271389961243,
+      "learning_rate": 7.448027087283426e-08,
+      "loss": 0.0718,
+      "step": 2518
+    },
+    {
+      "epoch": 1.9695074276778732,
+      "grad_norm": 0.4299383759498596,
+      "learning_rate": 7.08036756771946e-08,
+      "loss": 0.0661,
+      "step": 2519
+    },
+    {
+      "epoch": 1.9702892885066459,
+      "grad_norm": 0.3604923486709595,
+      "learning_rate": 6.722007205903213e-08,
+      "loss": 0.0648,
+      "step": 2520
+    },
+    {
+      "epoch": 1.9710711493354183,
+      "grad_norm": 0.3248717784881592,
+      "learning_rate": 6.372946669269131e-08,
+      "loss": 0.0364,
+      "step": 2521
+    },
+    {
+      "epoch": 1.9718530101641907,
+      "grad_norm": 0.4972895085811615,
+      "learning_rate": 6.033186607931062e-08,
+      "loss": 0.0711,
+      "step": 2522
+    },
+    {
+      "epoch": 1.9726348709929633,
+      "grad_norm": 0.39111319184303284,
+      "learning_rate": 5.702727654682272e-08,
+      "loss": 0.0584,
+      "step": 2523
+    },
+    {
+      "epoch": 1.9734167318217357,
+      "grad_norm": 0.3655970096588135,
+      "learning_rate": 5.381570424992655e-08,
+      "loss": 0.0478,
+      "step": 2524
+    },
+    {
+      "epoch": 1.9741985926505081,
+      "grad_norm": 0.4365689754486084,
+      "learning_rate": 5.069715517007079e-08,
+      "loss": 0.0401,
+      "step": 2525
+    },
+    {
+      "epoch": 1.9749804534792808,
+      "grad_norm": 0.6145164966583252,
+      "learning_rate": 4.7671635115453805e-08,
+      "loss": 0.0792,
+      "step": 2526
+    },
+    {
+      "epoch": 1.9757623143080532,
+      "grad_norm": 0.3456154763698578,
+      "learning_rate": 4.473914972101256e-08,
+      "loss": 0.0381,
+      "step": 2527
+    },
+    {
+      "epoch": 1.9765441751368256,
+      "grad_norm": 0.47349321842193604,
+      "learning_rate": 4.189970444841707e-08,
+      "loss": 0.0817,
+      "step": 2528
+    },
+    {
+      "epoch": 1.9773260359655982,
+      "grad_norm": 0.40150561928749084,
+      "learning_rate": 3.9153304586042605e-08,
+      "loss": 0.0846,
+      "step": 2529
+    },
+    {
+      "epoch": 1.9781078967943706,
+      "grad_norm": 0.49519282579421997,
+      "learning_rate": 3.649995524896976e-08,
+      "loss": 0.0631,
+      "step": 2530
+    },
+    {
+      "epoch": 1.978889757623143,
+      "grad_norm": 0.4056538939476013,
+      "learning_rate": 3.393966137898441e-08,
+      "loss": 0.0373,
+      "step": 2531
+    },
+    {
+      "epoch": 1.9796716184519156,
+      "grad_norm": 0.35286810994148254,
+      "learning_rate": 3.147242774454995e-08,
+      "loss": 0.0364,
+      "step": 2532
+    },
+    {
+      "epoch": 1.980453479280688,
+      "grad_norm": 0.3447190523147583,
+      "learning_rate": 2.9098258940818414e-08,
+      "loss": 0.0326,
+      "step": 2533
+    },
+    {
+      "epoch": 1.9812353401094605,
+      "grad_norm": 0.3452252745628357,
+      "learning_rate": 2.6817159389597167e-08,
+      "loss": 0.0362,
+      "step": 2534
+    },
+    {
+      "epoch": 1.982017200938233,
+      "grad_norm": 0.4439004361629486,
+      "learning_rate": 2.4629133339371113e-08,
+      "loss": 0.0566,
+      "step": 2535
+    },
+    {
+      "epoch": 1.9827990617670055,
+      "grad_norm": 0.37382808327674866,
+      "learning_rate": 2.253418486526382e-08,
+      "loss": 0.0406,
+      "step": 2536
+    },
+    {
+      "epoch": 1.983580922595778,
+      "grad_norm": 0.327451229095459,
+      "learning_rate": 2.0532317869059735e-08,
+      "loss": 0.0353,
+      "step": 2537
+    },
+    {
+      "epoch": 1.9843627834245505,
+      "grad_norm": 0.3801063001155853,
+      "learning_rate": 1.862353607917089e-08,
+      "loss": 0.0353,
+      "step": 2538
+    },
+    {
+      "epoch": 1.985144644253323,
+      "grad_norm": 0.45267173647880554,
+      "learning_rate": 1.6807843050636874e-08,
+      "loss": 0.0283,
+      "step": 2539
+    },
+    {
+      "epoch": 1.9859265050820953,
+      "grad_norm": 0.41038188338279724,
+      "learning_rate": 1.508524216514151e-08,
+      "loss": 0.0328,
+      "step": 2540
+    },
+    {
+      "epoch": 1.986708365910868,
+      "grad_norm": 0.3300861716270447,
+      "learning_rate": 1.345573663096289e-08,
+      "loss": 0.0451,
+      "step": 2541
+    },
+    {
+      "epoch": 1.9874902267396404,
+      "grad_norm": 0.32301145792007446,
+      "learning_rate": 1.1919329483017772e-08,
+      "loss": 0.0481,
+      "step": 2542
+    },
+    {
+      "epoch": 1.9882720875684128,
+      "grad_norm": 0.3235101103782654,
+      "learning_rate": 1.0476023582806083e-08,
+      "loss": 0.0572,
+      "step": 2543
+    },
+    {
+      "epoch": 1.9890539483971854,
+      "grad_norm": 0.3280138373374939,
+      "learning_rate": 9.12582161844422e-09,
+      "loss": 0.0518,
+      "step": 2544
+    },
+    {
+      "epoch": 1.9898358092259578,
+      "grad_norm": 0.4068671762943268,
+      "learning_rate": 7.868726104642844e-09,
+      "loss": 0.0492,
+      "step": 2545
+    },
+    {
+      "epoch": 1.9906176700547302,
+      "grad_norm": 0.4065364897251129,
+      "learning_rate": 6.704739382701331e-09,
+      "loss": 0.048,
+      "step": 2546
+    },
+    {
+      "epoch": 1.9913995308835029,
+      "grad_norm": 0.35321515798568726,
+      "learning_rate": 5.633863620507773e-09,
+      "loss": 0.0613,
+      "step": 2547
+    },
+    {
+      "epoch": 1.9921813917122753,
+      "grad_norm": 0.3954537510871887,
+      "learning_rate": 4.656100812538977e-09,
+      "loss": 0.0329,
+      "step": 2548
+    },
+    {
+      "epoch": 1.9929632525410477,
+      "grad_norm": 0.32833099365234375,
+      "learning_rate": 3.7714527798438095e-09,
+      "loss": 0.0346,
+      "step": 2549
+    },
+    {
+      "epoch": 1.9937451133698203,
+      "grad_norm": 0.4241968095302582,
+      "learning_rate": 2.9799211700543005e-09,
+      "loss": 0.0452,
+      "step": 2550
+    },
+    {
+      "epoch": 1.9945269741985927,
+      "grad_norm": 0.4431023597717285,
+      "learning_rate": 2.2815074573745434e-09,
+      "loss": 0.0458,
+      "step": 2551
+    },
+    {
+      "epoch": 1.9953088350273651,
+      "grad_norm": 0.42436906695365906,
+      "learning_rate": 1.6762129425751394e-09,
+      "loss": 0.0471,
+      "step": 2552
+    },
+    {
+      "epoch": 1.9960906958561377,
+      "grad_norm": 0.4154678285121918,
+      "learning_rate": 1.164038752998753e-09,
+      "loss": 0.0637,
+      "step": 2553
+    },
+    {
+      "epoch": 1.99687255668491,
+      "grad_norm": 0.3614688813686371,
+      "learning_rate": 7.449858425545575e-10,
+      "loss": 0.0862,
+      "step": 2554
+    },
+    {
+      "epoch": 1.9976544175136826,
+      "grad_norm": 0.3825763165950775,
+      "learning_rate": 4.1905499171268626e-10,
+      "loss": 0.0959,
+      "step": 2555
+    },
+    {
+      "epoch": 1.9984362783424552,
+      "grad_norm": 0.3862949311733246,
+      "learning_rate": 1.8624680751533342e-10,
+      "loss": 0.0502,
+      "step": 2556
+    },
+    {
+      "epoch": 1.9992181391712274,
+      "grad_norm": 0.36724698543548584,
+      "learning_rate": 4.6561723560101245e-11,
+      "loss": 0.0472,
+      "step": 2557
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4372091293334961,
+      "learning_rate": 0.0,
+      "loss": 0.0345,
+      "step": 2558
+    },
+    {
+      "epoch": 2.0,
+      "step": 2558,
+      "total_flos": 2.7985345273774735e+18,
+      "train_loss": 0.11478913314727902,
+      "train_runtime": 9036.0615,
+      "train_samples_per_second": 1.132,
+      "train_steps_per_second": 0.283
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2558,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.7985345273774735e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}