diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25984 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3706,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002698327037236913,
+      "grad_norm": 296.13665771484375,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 1.7116,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005396654074473826,
+      "grad_norm": 251.30067443847656,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 1.6891,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008094981111710739,
+      "grad_norm": 258.78179931640625,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 1.7201,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010793308148947653,
+      "grad_norm": 235.64593505859375,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.5884,
+      "step": 4
+    },
+    {
+      "epoch": 0.0013491635186184566,
+      "grad_norm": 244.46585083007812,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 1.4946,
+      "step": 5
+    },
+    {
+      "epoch": 0.0016189962223421479,
+      "grad_norm": 202.9253387451172,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 1.3626,
+      "step": 6
+    },
+    {
+      "epoch": 0.0018888289260658392,
+      "grad_norm": 54.758338928222656,
+      "learning_rate": 1.25e-06,
+      "loss": 0.9884,
+      "step": 7
+    },
+    {
+      "epoch": 0.0021586616297895305,
+      "grad_norm": 44.69416046142578,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.7653,
+      "step": 8
+    },
+    {
+      "epoch": 0.0024284943335132216,
+      "grad_norm": 68.98052978515625,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 1.3046,
+      "step": 9
+    },
+    {
+      "epoch": 0.002698327037236913,
+      "grad_norm": 86.93709564208984,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 1.4205,
+      "step": 10
+    },
+    {
+      "epoch": 0.0029681597409606042,
+      "grad_norm": 76.9502944946289,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 1.5194,
+      "step": 11
+    },
+    {
+      "epoch": 0.0032379924446842958,
+      "grad_norm": 76.57403564453125,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.5864,
+      "step": 12
+    },
+    {
+      "epoch": 0.003507825148407987,
+      "grad_norm": 68.76900482177734,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 1.5735,
+      "step": 13
+    },
+    {
+      "epoch": 0.0037776578521316784,
+      "grad_norm": 61.87789535522461,
+      "learning_rate": 2.5e-06,
+      "loss": 1.4026,
+      "step": 14
+    },
+    {
+      "epoch": 0.0040474905558553695,
+      "grad_norm": 76.0875473022461,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 1.4448,
+      "step": 15
+    },
+    {
+      "epoch": 0.004317323259579061,
+      "grad_norm": 46.45457077026367,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.9371,
+      "step": 16
+    },
+    {
+      "epoch": 0.0045871559633027525,
+      "grad_norm": 9.54983901977539,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 0.6708,
+      "step": 17
+    },
+    {
+      "epoch": 0.004856988667026443,
+      "grad_norm": 74.91175079345703,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 0.8029,
+      "step": 18
+    },
+    {
+      "epoch": 0.005126821370750135,
+      "grad_norm": 88.67768096923828,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 0.8459,
+      "step": 19
+    },
+    {
+      "epoch": 0.005396654074473826,
+      "grad_norm": 53.243507385253906,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.7042,
+      "step": 20
+    },
+    {
+      "epoch": 0.005666486778197518,
+      "grad_norm": 16.148927688598633,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6082,
+      "step": 21
+    },
+    {
+      "epoch": 0.0059363194819212085,
+      "grad_norm": 14.280265808105469,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.5361,
+      "step": 22
+    },
+    {
+      "epoch": 0.0062061521856449,
+      "grad_norm": 27.236541748046875,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 0.5692,
+      "step": 23
+    },
+    {
+      "epoch": 0.0064759848893685915,
+      "grad_norm": 35.92571258544922,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.6986,
+      "step": 24
+    },
+    {
+      "epoch": 0.006745817593092283,
+      "grad_norm": 28.47654914855957,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 0.7209,
+      "step": 25
+    },
+    {
+      "epoch": 0.007015650296815974,
+      "grad_norm": 25.38446044921875,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 0.631,
+      "step": 26
+    },
+    {
+      "epoch": 0.007285483000539665,
+      "grad_norm": 14.109930038452148,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 0.5741,
+      "step": 27
+    },
+    {
+      "epoch": 0.007555315704263357,
+      "grad_norm": 3.8552215099334717,
+      "learning_rate": 5e-06,
+      "loss": 0.5243,
+      "step": 28
+    },
+    {
+      "epoch": 0.007825148407987047,
+      "grad_norm": 18.07439422607422,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 0.5378,
+      "step": 29
+    },
+    {
+      "epoch": 0.008094981111710739,
+      "grad_norm": 30.23243522644043,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.5744,
+      "step": 30
+    },
+    {
+      "epoch": 0.00836481381543443,
+      "grad_norm": 28.361143112182617,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 0.5957,
+      "step": 31
+    },
+    {
+      "epoch": 0.008634646519158122,
+      "grad_norm": 10.887151718139648,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.5019,
+      "step": 32
+    },
+    {
+      "epoch": 0.008904479222881814,
+      "grad_norm": 5.808140754699707,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 0.5631,
+      "step": 33
+    },
+    {
+      "epoch": 0.009174311926605505,
+      "grad_norm": 10.890024185180664,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 0.5257,
+      "step": 34
+    },
+    {
+      "epoch": 0.009444144630329197,
+      "grad_norm": 11.746404647827148,
+      "learning_rate": 6.25e-06,
+      "loss": 0.5256,
+      "step": 35
+    },
+    {
+      "epoch": 0.009713977334052886,
+      "grad_norm": 16.615083694458008,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.5483,
+      "step": 36
+    },
+    {
+      "epoch": 0.009983810037776578,
+      "grad_norm": 14.429203033447266,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 0.492,
+      "step": 37
+    },
+    {
+      "epoch": 0.01025364274150027,
+      "grad_norm": 6.11021089553833,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 0.433,
+      "step": 38
+    },
+    {
+      "epoch": 0.010523475445223961,
+      "grad_norm": 2.218346118927002,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 0.4802,
+      "step": 39
+    },
+    {
+      "epoch": 0.010793308148947653,
+      "grad_norm": 9.668712615966797,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.4565,
+      "step": 40
+    },
+    {
+      "epoch": 0.011063140852671344,
+      "grad_norm": 15.203758239746094,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 0.4876,
+      "step": 41
+    },
+    {
+      "epoch": 0.011332973556395036,
+      "grad_norm": 12.916653633117676,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4588,
+      "step": 42
+    },
+    {
+      "epoch": 0.011602806260118727,
+      "grad_norm": 4.884962558746338,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 0.4386,
+      "step": 43
+    },
+    {
+      "epoch": 0.011872638963842417,
+      "grad_norm": 3.218534231185913,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.4619,
+      "step": 44
+    },
+    {
+      "epoch": 0.012142471667566108,
+      "grad_norm": 9.778227806091309,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 0.5115,
+      "step": 45
+    },
+    {
+      "epoch": 0.0124123043712898,
+      "grad_norm": 10.435979843139648,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.4458,
+      "step": 46
+    },
+    {
+      "epoch": 0.012682137075013492,
+      "grad_norm": 9.991037368774414,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 0.4403,
+      "step": 47
+    },
+    {
+      "epoch": 0.012951969778737183,
+      "grad_norm": 4.812756061553955,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.4482,
+      "step": 48
+    },
+    {
+      "epoch": 0.013221802482460875,
+      "grad_norm": 8.234877586364746,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.4232,
+      "step": 49
+    },
+    {
+      "epoch": 0.013491635186184566,
+      "grad_norm": 9.28339672088623,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.4326,
+      "step": 50
+    },
+    {
+      "epoch": 0.013761467889908258,
+      "grad_norm": 9.527681350708008,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 0.4143,
+      "step": 51
+    },
+    {
+      "epoch": 0.014031300593631947,
+      "grad_norm": 5.326992511749268,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.4324,
+      "step": 52
+    },
+    {
+      "epoch": 0.014301133297355639,
+      "grad_norm": 3.2621424198150635,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 0.4478,
+      "step": 53
+    },
+    {
+      "epoch": 0.01457096600107933,
+      "grad_norm": 7.041463851928711,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.4723,
+      "step": 54
+    },
+    {
+      "epoch": 0.014840798704803022,
+      "grad_norm": 8.163468360900879,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 0.3992,
+      "step": 55
+    },
+    {
+      "epoch": 0.015110631408526714,
+      "grad_norm": 9.231693267822266,
+      "learning_rate": 1e-05,
+      "loss": 0.4419,
+      "step": 56
+    },
+    {
+      "epoch": 0.015380464112250405,
+      "grad_norm": 1.2798889875411987,
+      "learning_rate": 1.0178571428571429e-05,
+      "loss": 0.43,
+      "step": 57
+    },
+    {
+      "epoch": 0.015650296815974095,
+      "grad_norm": 10.91202163696289,
+      "learning_rate": 1.0357142857142859e-05,
+      "loss": 0.4609,
+      "step": 58
+    },
+    {
+      "epoch": 0.015920129519697786,
+      "grad_norm": 9.247339248657227,
+      "learning_rate": 1.0535714285714287e-05,
+      "loss": 0.4312,
+      "step": 59
+    },
+    {
+      "epoch": 0.016189962223421478,
+      "grad_norm": 2.4968299865722656,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.4071,
+      "step": 60
+    },
+    {
+      "epoch": 0.01645979492714517,
+      "grad_norm": 3.81266713142395,
+      "learning_rate": 1.0892857142857142e-05,
+      "loss": 0.433,
+      "step": 61
+    },
+    {
+      "epoch": 0.01672962763086886,
+      "grad_norm": 7.424432754516602,
+      "learning_rate": 1.1071428571428572e-05,
+      "loss": 0.4597,
+      "step": 62
+    },
+    {
+      "epoch": 0.016999460334592553,
+      "grad_norm": 2.6121442317962646,
+      "learning_rate": 1.125e-05,
+      "loss": 0.4081,
+      "step": 63
+    },
+    {
+      "epoch": 0.017269293038316244,
+      "grad_norm": 4.168766021728516,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.4293,
+      "step": 64
+    },
+    {
+      "epoch": 0.017539125742039936,
+      "grad_norm": 2.467888116836548,
+      "learning_rate": 1.1607142857142859e-05,
+      "loss": 0.4052,
+      "step": 65
+    },
+    {
+      "epoch": 0.017808958445763627,
+      "grad_norm": 6.4459919929504395,
+      "learning_rate": 1.1785714285714287e-05,
+      "loss": 0.4208,
+      "step": 66
+    },
+    {
+      "epoch": 0.01807879114948732,
+      "grad_norm": 8.185012817382812,
+      "learning_rate": 1.1964285714285716e-05,
+      "loss": 0.431,
+      "step": 67
+    },
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 1.0100297927856445,
+      "learning_rate": 1.2142857142857142e-05,
+      "loss": 0.4293,
+      "step": 68
+    },
+    {
+      "epoch": 0.0186184565569347,
+      "grad_norm": 8.321513175964355,
+      "learning_rate": 1.2321428571428572e-05,
+      "loss": 0.442,
+      "step": 69
+    },
+    {
+      "epoch": 0.018888289260658393,
+      "grad_norm": 9.19276237487793,
+      "learning_rate": 1.25e-05,
+      "loss": 0.4609,
+      "step": 70
+    },
+    {
+      "epoch": 0.019158121964382085,
+      "grad_norm": 0.6837041974067688,
+      "learning_rate": 1.2678571428571429e-05,
+      "loss": 0.4065,
+      "step": 71
+    },
+    {
+      "epoch": 0.019427954668105773,
+      "grad_norm": 6.580166339874268,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 0.4337,
+      "step": 72
+    },
+    {
+      "epoch": 0.019697787371829464,
+      "grad_norm": 10.928820610046387,
+      "learning_rate": 1.3035714285714287e-05,
+      "loss": 0.4497,
+      "step": 73
+    },
+    {
+      "epoch": 0.019967620075553156,
+      "grad_norm": 2.5008108615875244,
+      "learning_rate": 1.3214285714285716e-05,
+      "loss": 0.4194,
+      "step": 74
+    },
+    {
+      "epoch": 0.020237452779276847,
+      "grad_norm": 6.543095111846924,
+      "learning_rate": 1.3392857142857142e-05,
+      "loss": 0.386,
+      "step": 75
+    },
+    {
+      "epoch": 0.02050728548300054,
+      "grad_norm": 8.731756210327148,
+      "learning_rate": 1.3571428571428574e-05,
+      "loss": 0.4347,
+      "step": 76
+    },
+    {
+      "epoch": 0.02077711818672423,
+      "grad_norm": 6.255819320678711,
+      "learning_rate": 1.375e-05,
+      "loss": 0.4186,
+      "step": 77
+    },
+    {
+      "epoch": 0.021046950890447922,
+      "grad_norm": 1.6862869262695312,
+      "learning_rate": 1.3928571428571429e-05,
+      "loss": 0.4432,
+      "step": 78
+    },
+    {
+      "epoch": 0.021316783594171614,
+      "grad_norm": 8.184305191040039,
+      "learning_rate": 1.4107142857142859e-05,
+      "loss": 0.3882,
+      "step": 79
+    },
+    {
+      "epoch": 0.021586616297895305,
+      "grad_norm": 13.66966438293457,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.4437,
+      "step": 80
+    },
+    {
+      "epoch": 0.021856449001618997,
+      "grad_norm": 2.432039976119995,
+      "learning_rate": 1.4464285714285715e-05,
+      "loss": 0.3967,
+      "step": 81
+    },
+    {
+      "epoch": 0.022126281705342688,
+      "grad_norm": 12.502958297729492,
+      "learning_rate": 1.4642857142857144e-05,
+      "loss": 0.4736,
+      "step": 82
+    },
+    {
+      "epoch": 0.02239611440906638,
+      "grad_norm": 9.629276275634766,
+      "learning_rate": 1.4821428571428574e-05,
+      "loss": 0.3966,
+      "step": 83
+    },
+    {
+      "epoch": 0.02266594711279007,
+      "grad_norm": 3.4270660877227783,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3917,
+      "step": 84
+    },
+    {
+      "epoch": 0.022935779816513763,
+      "grad_norm": 6.486330032348633,
+      "learning_rate": 1.5178571428571429e-05,
+      "loss": 0.3876,
+      "step": 85
+    },
+    {
+      "epoch": 0.023205612520237454,
+      "grad_norm": 11.248103141784668,
+      "learning_rate": 1.535714285714286e-05,
+      "loss": 0.4314,
+      "step": 86
+    },
+    {
+      "epoch": 0.023475445223961146,
+      "grad_norm": 2.5970327854156494,
+      "learning_rate": 1.553571428571429e-05,
+      "loss": 0.411,
+      "step": 87
+    },
+    {
+      "epoch": 0.023745277927684834,
+      "grad_norm": 8.5509614944458,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.4489,
+      "step": 88
+    },
+    {
+      "epoch": 0.024015110631408525,
+      "grad_norm": 6.770622253417969,
+      "learning_rate": 1.5892857142857142e-05,
+      "loss": 0.4117,
+      "step": 89
+    },
+    {
+      "epoch": 0.024284943335132217,
+      "grad_norm": 1.8935930728912354,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.3831,
+      "step": 90
+    },
+    {
+      "epoch": 0.02455477603885591,
+      "grad_norm": 6.624536991119385,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.4002,
+      "step": 91
+    },
+    {
+      "epoch": 0.0248246087425796,
+      "grad_norm": 3.456044912338257,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.4218,
+      "step": 92
+    },
+    {
+      "epoch": 0.02509444144630329,
+      "grad_norm": 2.4928646087646484,
+      "learning_rate": 1.660714285714286e-05,
+      "loss": 0.413,
+      "step": 93
+    },
+    {
+      "epoch": 0.025364274150026983,
+      "grad_norm": 3.4704556465148926,
+      "learning_rate": 1.678571428571429e-05,
+      "loss": 0.4015,
+      "step": 94
+    },
+    {
+      "epoch": 0.025634106853750675,
+      "grad_norm": 5.672900676727295,
+      "learning_rate": 1.6964285714285715e-05,
+      "loss": 0.4116,
+      "step": 95
+    },
+    {
+      "epoch": 0.025903939557474366,
+      "grad_norm": 1.1748989820480347,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.4057,
+      "step": 96
+    },
+    {
+      "epoch": 0.026173772261198058,
+      "grad_norm": 4.132370471954346,
+      "learning_rate": 1.7321428571428572e-05,
+      "loss": 0.3799,
+      "step": 97
+    },
+    {
+      "epoch": 0.02644360496492175,
+      "grad_norm": 3.563218116760254,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.423,
+      "step": 98
+    },
+    {
+      "epoch": 0.02671343766864544,
+      "grad_norm": 0.7231237888336182,
+      "learning_rate": 1.767857142857143e-05,
+      "loss": 0.3889,
+      "step": 99
+    },
+    {
+      "epoch": 0.026983270372369132,
+      "grad_norm": 3.450093984603882,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.4238,
+      "step": 100
+    },
+    {
+      "epoch": 0.027253103076092824,
+      "grad_norm": 1.069626808166504,
+      "learning_rate": 1.803571428571429e-05,
+      "loss": 0.3861,
+      "step": 101
+    },
+    {
+      "epoch": 0.027522935779816515,
+      "grad_norm": 0.6431038975715637,
+      "learning_rate": 1.8214285714285715e-05,
+      "loss": 0.4021,
+      "step": 102
+    },
+    {
+      "epoch": 0.027792768483540203,
+      "grad_norm": 1.3995107412338257,
+      "learning_rate": 1.8392857142857142e-05,
+      "loss": 0.3923,
+      "step": 103
+    },
+    {
+      "epoch": 0.028062601187263895,
+      "grad_norm": 1.0395961999893188,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.393,
+      "step": 104
+    },
+    {
+      "epoch": 0.028332433890987586,
+      "grad_norm": 3.872102737426758,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.4123,
+      "step": 105
+    },
+    {
+      "epoch": 0.028602266594711278,
+      "grad_norm": 2.5545907020568848,
+      "learning_rate": 1.892857142857143e-05,
+      "loss": 0.4468,
+      "step": 106
+    },
+    {
+      "epoch": 0.02887209929843497,
+      "grad_norm": 9.189969062805176,
+      "learning_rate": 1.910714285714286e-05,
+      "loss": 0.4089,
+      "step": 107
+    },
+    {
+      "epoch": 0.02914193200215866,
+      "grad_norm": 8.54588508605957,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.4215,
+      "step": 108
+    },
+    {
+      "epoch": 0.029411764705882353,
+      "grad_norm": 3.4994733333587646,
+      "learning_rate": 1.9464285714285715e-05,
+      "loss": 0.4058,
+      "step": 109
+    },
+    {
+      "epoch": 0.029681597409606044,
+      "grad_norm": 1.1470156908035278,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 0.3984,
+      "step": 110
+    },
+    {
+      "epoch": 0.029951430113329736,
+      "grad_norm": 9.102022171020508,
+      "learning_rate": 1.9821428571428575e-05,
+      "loss": 0.3897,
+      "step": 111
+    },
+    {
+      "epoch": 0.030221262817053427,
+      "grad_norm": 9.201180458068848,
+      "learning_rate": 2e-05,
+      "loss": 0.4272,
+      "step": 112
+    },
+    {
+      "epoch": 0.03049109552077712,
+      "grad_norm": 0.808781623840332,
+      "learning_rate": 1.9999996179558304e-05,
+      "loss": 0.4192,
+      "step": 113
+    },
+    {
+      "epoch": 0.03076092822450081,
+      "grad_norm": 5.8410468101501465,
+      "learning_rate": 1.999998471823613e-05,
+      "loss": 0.3964,
+      "step": 114
+    },
+    {
+      "epoch": 0.0310307609282245,
+      "grad_norm": 2.4389750957489014,
+      "learning_rate": 1.999996561604223e-05,
+      "loss": 0.4095,
+      "step": 115
+    },
+    {
+      "epoch": 0.03130059363194819,
+      "grad_norm": 2.6165452003479004,
+      "learning_rate": 1.9999938872991212e-05,
+      "loss": 0.4175,
+      "step": 116
+    },
+    {
+      "epoch": 0.031570426335671885,
+      "grad_norm": 0.8905782699584961,
+      "learning_rate": 1.9999904489103504e-05,
+      "loss": 0.3781,
+      "step": 117
+    },
+    {
+      "epoch": 0.03184025903939557,
+      "grad_norm": 4.104673385620117,
+      "learning_rate": 1.9999862464405377e-05,
+      "loss": 0.3891,
+      "step": 118
+    },
+    {
+      "epoch": 0.03211009174311927,
+      "grad_norm": 3.547621250152588,
+      "learning_rate": 1.9999812798928944e-05,
+      "loss": 0.3975,
+      "step": 119
+    },
+    {
+      "epoch": 0.032379924446842956,
+      "grad_norm": 10.695901870727539,
+      "learning_rate": 1.9999755492712148e-05,
+      "loss": 0.425,
+      "step": 120
+    },
+    {
+      "epoch": 0.03264975715056665,
+      "grad_norm": 4.4934258460998535,
+      "learning_rate": 1.9999690545798785e-05,
+      "loss": 0.3942,
+      "step": 121
+    },
+    {
+      "epoch": 0.03291958985429034,
+      "grad_norm": 1.6885639429092407,
+      "learning_rate": 1.9999617958238477e-05,
+      "loss": 0.3762,
+      "step": 122
+    },
+    {
+      "epoch": 0.033189422558014034,
+      "grad_norm": 6.010866641998291,
+      "learning_rate": 1.9999537730086684e-05,
+      "loss": 0.36,
+      "step": 123
+    },
+    {
+      "epoch": 0.03345925526173772,
+      "grad_norm": 5.377583026885986,
+      "learning_rate": 1.9999449861404716e-05,
+      "loss": 0.4113,
+      "step": 124
+    },
+    {
+      "epoch": 0.03372908796546142,
+      "grad_norm": 5.2750959396362305,
+      "learning_rate": 1.9999354352259696e-05,
+      "loss": 0.3775,
+      "step": 125
+    },
+    {
+      "epoch": 0.033998920669185105,
+      "grad_norm": 4.2186665534973145,
+      "learning_rate": 1.9999251202724616e-05,
+      "loss": 0.4436,
+      "step": 126
+    },
+    {
+      "epoch": 0.03426875337290879,
+      "grad_norm": 14.201189041137695,
+      "learning_rate": 1.9999140412878288e-05,
+      "loss": 0.3975,
+      "step": 127
+    },
+    {
+      "epoch": 0.03453858607663249,
+      "grad_norm": 2.002364158630371,
+      "learning_rate": 1.9999021982805363e-05,
+      "loss": 0.4263,
+      "step": 128
+    },
+    {
+      "epoch": 0.034808418780356176,
+      "grad_norm": 8.503058433532715,
+      "learning_rate": 1.9998895912596332e-05,
+      "loss": 0.4481,
+      "step": 129
+    },
+    {
+      "epoch": 0.03507825148407987,
+      "grad_norm": 0.7332853674888611,
+      "learning_rate": 1.999876220234753e-05,
+      "loss": 0.3929,
+      "step": 130
+    },
+    {
+      "epoch": 0.03534808418780356,
+      "grad_norm": 4.720513820648193,
+      "learning_rate": 1.999862085216111e-05,
+      "loss": 0.3912,
+      "step": 131
+    },
+    {
+      "epoch": 0.035617916891527254,
+      "grad_norm": 0.6026013493537903,
+      "learning_rate": 1.9998471862145088e-05,
+      "loss": 0.3942,
+      "step": 132
+    },
+    {
+      "epoch": 0.03588774959525094,
+      "grad_norm": 0.9120808839797974,
+      "learning_rate": 1.9998315232413304e-05,
+      "loss": 0.3938,
+      "step": 133
+    },
+    {
+      "epoch": 0.03615758229897464,
+      "grad_norm": 3.827704668045044,
+      "learning_rate": 1.9998150963085434e-05,
+      "loss": 0.3979,
+      "step": 134
+    },
+    {
+      "epoch": 0.036427415002698325,
+      "grad_norm": 2.3384830951690674,
+      "learning_rate": 1.9997979054286992e-05,
+      "loss": 0.3993,
+      "step": 135
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 3.886779546737671,
+      "learning_rate": 1.999779950614934e-05,
+      "loss": 0.3935,
+      "step": 136
+    },
+    {
+      "epoch": 0.03696708041014571,
+      "grad_norm": 2.6251606941223145,
+      "learning_rate": 1.999761231880966e-05,
+      "loss": 0.3622,
+      "step": 137
+    },
+    {
+      "epoch": 0.0372369131138694,
+      "grad_norm": 2.4224445819854736,
+      "learning_rate": 1.9997417492410985e-05,
+      "loss": 0.3694,
+      "step": 138
+    },
+    {
+      "epoch": 0.03750674581759309,
+      "grad_norm": 0.951027512550354,
+      "learning_rate": 1.999721502710217e-05,
+      "loss": 0.3674,
+      "step": 139
+    },
+    {
+      "epoch": 0.037776578521316787,
+      "grad_norm": 2.7989346981048584,
+      "learning_rate": 1.999700492303793e-05,
+      "loss": 0.4139,
+      "step": 140
+    },
+    {
+      "epoch": 0.038046411225040475,
+      "grad_norm": 2.7216851711273193,
+      "learning_rate": 1.99967871803788e-05,
+      "loss": 0.404,
+      "step": 141
+    },
+    {
+      "epoch": 0.03831624392876417,
+      "grad_norm": 3.0476880073547363,
+      "learning_rate": 1.999656179929115e-05,
+      "loss": 0.3847,
+      "step": 142
+    },
+    {
+      "epoch": 0.03858607663248786,
+      "grad_norm": 4.825053691864014,
+      "learning_rate": 1.999632877994719e-05,
+      "loss": 0.3889,
+      "step": 143
+    },
+    {
+      "epoch": 0.038855909336211546,
+      "grad_norm": 2.0866966247558594,
+      "learning_rate": 1.9996088122524973e-05,
+      "loss": 0.3884,
+      "step": 144
+    },
+    {
+      "epoch": 0.03912574203993524,
+      "grad_norm": 4.13264799118042,
+      "learning_rate": 1.9995839827208378e-05,
+      "loss": 0.4117,
+      "step": 145
+    },
+    {
+      "epoch": 0.03939557474365893,
+      "grad_norm": 0.6439782977104187,
+      "learning_rate": 1.9995583894187125e-05,
+      "loss": 0.38,
+      "step": 146
+    },
+    {
+      "epoch": 0.039665407447382624,
+      "grad_norm": 2.5739409923553467,
+      "learning_rate": 1.999532032365677e-05,
+      "loss": 0.4045,
+      "step": 147
+    },
+    {
+      "epoch": 0.03993524015110631,
+      "grad_norm": 4.372972011566162,
+      "learning_rate": 1.9995049115818706e-05,
+      "loss": 0.4033,
+      "step": 148
+    },
+    {
+      "epoch": 0.04020507285483001,
+      "grad_norm": 4.483029842376709,
+      "learning_rate": 1.999477027088016e-05,
+      "loss": 0.4001,
+      "step": 149
+    },
+    {
+      "epoch": 0.040474905558553695,
+      "grad_norm": 3.930680990219116,
+      "learning_rate": 1.9994483789054192e-05,
+      "loss": 0.3931,
+      "step": 150
+    },
+    {
+      "epoch": 0.04074473826227739,
+      "grad_norm": 5.181520462036133,
+      "learning_rate": 1.99941896705597e-05,
+      "loss": 0.4105,
+      "step": 151
+    },
+    {
+      "epoch": 0.04101457096600108,
+      "grad_norm": 1.050384521484375,
+      "learning_rate": 1.9993887915621416e-05,
+      "loss": 0.3856,
+      "step": 152
+    },
+    {
+      "epoch": 0.04128440366972477,
+      "grad_norm": 1.123969554901123,
+      "learning_rate": 1.9993578524469907e-05,
+      "loss": 0.3901,
+      "step": 153
+    },
+    {
+      "epoch": 0.04155423637344846,
+      "grad_norm": 5.739140510559082,
+      "learning_rate": 1.9993261497341575e-05,
+      "loss": 0.3833,
+      "step": 154
+    },
+    {
+      "epoch": 0.041824069077172156,
+      "grad_norm": 4.126053333282471,
+      "learning_rate": 1.999293683447866e-05,
+      "loss": 0.3847,
+      "step": 155
+    },
+    {
+      "epoch": 0.042093901780895844,
+      "grad_norm": 2.2701728343963623,
+      "learning_rate": 1.999260453612923e-05,
+      "loss": 0.3652,
+      "step": 156
+    },
+    {
+      "epoch": 0.04236373448461954,
+      "grad_norm": 3.938037633895874,
+      "learning_rate": 1.9992264602547194e-05,
+      "loss": 0.3915,
+      "step": 157
+    },
+    {
+      "epoch": 0.04263356718834323,
+      "grad_norm": 10.784651756286621,
+      "learning_rate": 1.9991917033992287e-05,
+      "loss": 0.3863,
+      "step": 158
+    },
+    {
+      "epoch": 0.042903399892066915,
+      "grad_norm": 1.4693799018859863,
+      "learning_rate": 1.999156183073008e-05,
+      "loss": 0.3685,
+      "step": 159
+    },
+    {
+      "epoch": 0.04317323259579061,
+      "grad_norm": 6.2212653160095215,
+      "learning_rate": 1.9991198993031992e-05,
+      "loss": 0.3836,
+      "step": 160
+    },
+    {
+      "epoch": 0.0434430652995143,
+      "grad_norm": 6.595179557800293,
+      "learning_rate": 1.9990828521175247e-05,
+      "loss": 0.3939,
+      "step": 161
+    },
+    {
+      "epoch": 0.04371289800323799,
+      "grad_norm": 12.49017333984375,
+      "learning_rate": 1.999045041544293e-05,
+      "loss": 0.4486,
+      "step": 162
+    },
+    {
+      "epoch": 0.04398273070696168,
+      "grad_norm": 6.004252910614014,
+      "learning_rate": 1.999006467612394e-05,
+      "loss": 0.3939,
+      "step": 163
+    },
+    {
+      "epoch": 0.044252563410685376,
+      "grad_norm": 10.004898071289062,
+      "learning_rate": 1.998967130351302e-05,
+      "loss": 0.4165,
+      "step": 164
+    },
+    {
+      "epoch": 0.044522396114409064,
+      "grad_norm": 4.274568557739258,
+      "learning_rate": 1.9989270297910737e-05,
+      "loss": 0.3916,
+      "step": 165
+    },
+    {
+      "epoch": 0.04479222881813276,
+      "grad_norm": 1.8940659761428833,
+      "learning_rate": 1.99888616596235e-05,
+      "loss": 0.4038,
+      "step": 166
+    },
+    {
+      "epoch": 0.04506206152185645,
+      "grad_norm": 6.905783176422119,
+      "learning_rate": 1.9988445388963543e-05,
+      "loss": 0.4036,
+      "step": 167
+    },
+    {
+      "epoch": 0.04533189422558014,
+      "grad_norm": 1.6764140129089355,
+      "learning_rate": 1.9988021486248933e-05,
+      "loss": 0.4084,
+      "step": 168
+    },
+    {
+      "epoch": 0.04560172692930383,
+      "grad_norm": 3.911954402923584,
+      "learning_rate": 1.998758995180357e-05,
+      "loss": 0.4018,
+      "step": 169
+    },
+    {
+      "epoch": 0.045871559633027525,
+      "grad_norm": 1.0182290077209473,
+      "learning_rate": 1.9987150785957183e-05,
+      "loss": 0.3856,
+      "step": 170
+    },
+    {
+      "epoch": 0.046141392336751214,
+      "grad_norm": 1.1737697124481201,
+      "learning_rate": 1.9986703989045334e-05,
+      "loss": 0.3802,
+      "step": 171
+    },
+    {
+      "epoch": 0.04641122504047491,
+      "grad_norm": 0.8012648224830627,
+      "learning_rate": 1.9986249561409415e-05,
+      "loss": 0.4115,
+      "step": 172
+    },
+    {
+      "epoch": 0.0466810577441986,
+      "grad_norm": 2.823913335800171,
+      "learning_rate": 1.998578750339665e-05,
+      "loss": 0.4165,
+      "step": 173
+    },
+    {
+      "epoch": 0.04695089044792229,
+      "grad_norm": 2.148939371109009,
+      "learning_rate": 1.998531781536009e-05,
+      "loss": 0.413,
+      "step": 174
+    },
+    {
+      "epoch": 0.04722072315164598,
+      "grad_norm": 3.65297269821167,
+      "learning_rate": 1.9984840497658624e-05,
+      "loss": 0.3467,
+      "step": 175
+    },
+    {
+      "epoch": 0.04749055585536967,
+      "grad_norm": 1.9458959102630615,
+      "learning_rate": 1.9984355550656957e-05,
+      "loss": 0.3842,
+      "step": 176
+    },
+    {
+      "epoch": 0.04776038855909336,
+      "grad_norm": 5.1969757080078125,
+      "learning_rate": 1.9983862974725637e-05,
+      "loss": 0.3923,
+      "step": 177
+    },
+    {
+      "epoch": 0.04803022126281705,
+      "grad_norm": 1.7149699926376343,
+      "learning_rate": 1.998336277024103e-05,
+      "loss": 0.3754,
+      "step": 178
+    },
+    {
+      "epoch": 0.048300053966540746,
+      "grad_norm": 7.781804084777832,
+      "learning_rate": 1.9982854937585342e-05,
+      "loss": 0.4016,
+      "step": 179
+    },
+    {
+      "epoch": 0.048569886670264434,
+      "grad_norm": 3.0358760356903076,
+      "learning_rate": 1.9982339477146602e-05,
+      "loss": 0.3937,
+      "step": 180
+    },
+    {
+      "epoch": 0.04883971937398813,
+      "grad_norm": 9.285788536071777,
+      "learning_rate": 1.9981816389318663e-05,
+      "loss": 0.4107,
+      "step": 181
+    },
+    {
+      "epoch": 0.04910955207771182,
+      "grad_norm": 1.1116142272949219,
+      "learning_rate": 1.9981285674501213e-05,
+      "loss": 0.3704,
+      "step": 182
+    },
+    {
+      "epoch": 0.04937938478143551,
+      "grad_norm": 8.136933326721191,
+      "learning_rate": 1.9980747333099764e-05,
+      "loss": 0.3806,
+      "step": 183
+    },
+    {
+      "epoch": 0.0496492174851592,
+      "grad_norm": 0.7685086727142334,
+      "learning_rate": 1.998020136552566e-05,
+      "loss": 0.3944,
+      "step": 184
+    },
+    {
+      "epoch": 0.049919050188882895,
+      "grad_norm": 6.753932952880859,
+      "learning_rate": 1.997964777219606e-05,
+      "loss": 0.372,
+      "step": 185
+    },
+    {
+      "epoch": 0.05018888289260658,
+      "grad_norm": 2.9645187854766846,
+      "learning_rate": 1.9979086553533966e-05,
+      "loss": 0.3919,
+      "step": 186
+    },
+    {
+      "epoch": 0.05045871559633028,
+      "grad_norm": 7.841409683227539,
+      "learning_rate": 1.99785177099682e-05,
+      "loss": 0.4121,
+      "step": 187
+    },
+    {
+      "epoch": 0.050728548300053966,
+      "grad_norm": 4.415322303771973,
+      "learning_rate": 1.9977941241933395e-05,
+      "loss": 0.3924,
+      "step": 188
+    },
+    {
+      "epoch": 0.05099838100377766,
+      "grad_norm": 5.308473110198975,
+      "learning_rate": 1.997735714987004e-05,
+      "loss": 0.3938,
+      "step": 189
+    },
+    {
+      "epoch": 0.05126821370750135,
+      "grad_norm": 6.0477213859558105,
+      "learning_rate": 1.9976765434224426e-05,
+      "loss": 0.3943,
+      "step": 190
+    },
+    {
+      "epoch": 0.05153804641122504,
+      "grad_norm": 2.717771530151367,
+      "learning_rate": 1.9976166095448677e-05,
+      "loss": 0.4073,
+      "step": 191
+    },
+    {
+      "epoch": 0.05180787911494873,
+      "grad_norm": 7.7683563232421875,
+      "learning_rate": 1.9975559134000738e-05,
+      "loss": 0.3902,
+      "step": 192
+    },
+    {
+      "epoch": 0.05207771181867242,
+      "grad_norm": 1.4602460861206055,
+      "learning_rate": 1.9974944550344385e-05,
+      "loss": 0.3858,
+      "step": 193
+    },
+    {
+      "epoch": 0.052347544522396115,
+      "grad_norm": 7.551769256591797,
+      "learning_rate": 1.9974322344949214e-05,
+      "loss": 0.4059,
+      "step": 194
+    },
+    {
+      "epoch": 0.0526173772261198,
+      "grad_norm": 1.7735443115234375,
+      "learning_rate": 1.997369251829064e-05,
+      "loss": 0.3895,
+      "step": 195
+    },
+    {
+      "epoch": 0.0528872099298435,
+      "grad_norm": 9.310518264770508,
+      "learning_rate": 1.9973055070849912e-05,
+      "loss": 0.4012,
+      "step": 196
+    },
+    {
+      "epoch": 0.053157042633567186,
+      "grad_norm": 0.7109113335609436,
+      "learning_rate": 1.9972410003114092e-05,
+      "loss": 0.3969,
+      "step": 197
+    },
+    {
+      "epoch": 0.05342687533729088,
+      "grad_norm": 6.293333530426025,
+      "learning_rate": 1.9971757315576073e-05,
+      "loss": 0.4128,
+      "step": 198
+    },
+    {
+      "epoch": 0.05369670804101457,
+      "grad_norm": 0.6900772452354431,
+      "learning_rate": 1.997109700873456e-05,
+      "loss": 0.3725,
+      "step": 199
+    },
+    {
+      "epoch": 0.053966540744738264,
+      "grad_norm": 5.082067012786865,
+      "learning_rate": 1.9970429083094092e-05,
+      "loss": 0.3816,
+      "step": 200
+    },
+    {
+      "epoch": 0.05423637344846195,
+      "grad_norm": 3.0355701446533203,
+      "learning_rate": 1.996975353916502e-05,
+      "loss": 0.3921,
+      "step": 201
+    },
+    {
+      "epoch": 0.05450620615218565,
+      "grad_norm": 1.9523979425430298,
+      "learning_rate": 1.996907037746352e-05,
+      "loss": 0.3787,
+      "step": 202
+    },
+    {
+      "epoch": 0.054776038855909336,
+      "grad_norm": 2.740537643432617,
+      "learning_rate": 1.9968379598511586e-05,
+      "loss": 0.3896,
+      "step": 203
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 3.0266401767730713,
+      "learning_rate": 1.9967681202837035e-05,
+      "loss": 0.3783,
+      "step": 204
+    },
+    {
+      "epoch": 0.05531570426335672,
+      "grad_norm": 4.1108293533325195,
+      "learning_rate": 1.9966975190973507e-05,
+      "loss": 0.3943,
+      "step": 205
+    },
+    {
+      "epoch": 0.05558553696708041,
+      "grad_norm": 1.816755771636963,
+      "learning_rate": 1.996626156346045e-05,
+      "loss": 0.3601,
+      "step": 206
+    },
+    {
+      "epoch": 0.0558553696708041,
+      "grad_norm": 1.842166781425476,
+      "learning_rate": 1.9965540320843148e-05,
+      "loss": 0.3548,
+      "step": 207
+    },
+    {
+      "epoch": 0.05612520237452779,
+      "grad_norm": 0.7184922099113464,
+      "learning_rate": 1.9964811463672685e-05,
+      "loss": 0.3816,
+      "step": 208
+    },
+    {
+      "epoch": 0.056395035078251485,
+      "grad_norm": 0.9205771088600159,
+      "learning_rate": 1.9964074992505977e-05,
+      "loss": 0.3974,
+      "step": 209
+    },
+    {
+      "epoch": 0.05666486778197517,
+      "grad_norm": 1.0519591569900513,
+      "learning_rate": 1.9963330907905747e-05,
+      "loss": 0.394,
+      "step": 210
+    },
+    {
+      "epoch": 0.05693470048569887,
+      "grad_norm": 4.6511101722717285,
+      "learning_rate": 1.9962579210440552e-05,
+      "loss": 0.3925,
+      "step": 211
+    },
+    {
+      "epoch": 0.057204533189422556,
+      "grad_norm": 4.01237154006958,
+      "learning_rate": 1.9961819900684747e-05,
+      "loss": 0.354,
+      "step": 212
+    },
+    {
+      "epoch": 0.05747436589314625,
+      "grad_norm": 1.7215228080749512,
+      "learning_rate": 1.9961052979218517e-05,
+      "loss": 0.3492,
+      "step": 213
+    },
+    {
+      "epoch": 0.05774419859686994,
+      "grad_norm": 0.8879185914993286,
+      "learning_rate": 1.996027844662785e-05,
+      "loss": 0.3502,
+      "step": 214
+    },
+    {
+      "epoch": 0.058014031300593634,
+      "grad_norm": 1.9626739025115967,
+      "learning_rate": 1.995949630350457e-05,
+      "loss": 0.3948,
+      "step": 215
+    },
+    {
+      "epoch": 0.05828386400431732,
+      "grad_norm": 1.8283271789550781,
+      "learning_rate": 1.995870655044629e-05,
+      "loss": 0.3596,
+      "step": 216
+    },
+    {
+      "epoch": 0.05855369670804102,
+      "grad_norm": 3.1209323406219482,
+      "learning_rate": 1.9957909188056458e-05,
+      "loss": 0.3492,
+      "step": 217
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 2.706901788711548,
+      "learning_rate": 1.9957104216944328e-05,
+      "loss": 0.4011,
+      "step": 218
+    },
+    {
+      "epoch": 0.0590933621154884,
+      "grad_norm": 4.873446941375732,
+      "learning_rate": 1.995629163772497e-05,
+      "loss": 0.4057,
+      "step": 219
+    },
+    {
+      "epoch": 0.05936319481921209,
+      "grad_norm": 0.8274926543235779,
+      "learning_rate": 1.9955471451019264e-05,
+      "loss": 0.367,
+      "step": 220
+    },
+    {
+      "epoch": 0.05963302752293578,
+      "grad_norm": 1.5360764265060425,
+      "learning_rate": 1.9954643657453907e-05,
+      "loss": 0.3996,
+      "step": 221
+    },
+    {
+      "epoch": 0.05990286022665947,
+      "grad_norm": 1.9188793897628784,
+      "learning_rate": 1.995380825766141e-05,
+      "loss": 0.3506,
+      "step": 222
+    },
+    {
+      "epoch": 0.06017269293038316,
+      "grad_norm": 0.9776981472969055,
+      "learning_rate": 1.9952965252280084e-05,
+      "loss": 0.386,
+      "step": 223
+    },
+    {
+      "epoch": 0.060442525634106854,
+      "grad_norm": 6.003381252288818,
+      "learning_rate": 1.9952114641954067e-05,
+      "loss": 0.353,
+      "step": 224
+    },
+    {
+      "epoch": 0.06071235833783054,
+      "grad_norm": 5.819228172302246,
+      "learning_rate": 1.9951256427333292e-05,
+      "loss": 0.3785,
+      "step": 225
+    },
+    {
+      "epoch": 0.06098219104155424,
+      "grad_norm": 4.405791282653809,
+      "learning_rate": 1.995039060907352e-05,
+      "loss": 0.3889,
+      "step": 226
+    },
+    {
+      "epoch": 0.061252023745277925,
+      "grad_norm": 9.892199516296387,
+      "learning_rate": 1.9949517187836308e-05,
+      "loss": 0.3574,
+      "step": 227
+    },
+    {
+      "epoch": 0.06152185644900162,
+      "grad_norm": 4.544127464294434,
+      "learning_rate": 1.994863616428903e-05,
+      "loss": 0.341,
+      "step": 228
+    },
+    {
+      "epoch": 0.06179168915272531,
+      "grad_norm": 5.731225490570068,
+      "learning_rate": 1.994774753910486e-05,
+      "loss": 0.3806,
+      "step": 229
+    },
+    {
+      "epoch": 0.062061521856449,
+      "grad_norm": 5.26816463470459,
+      "learning_rate": 1.994685131296279e-05,
+      "loss": 0.3562,
+      "step": 230
+    },
+    {
+      "epoch": 0.06233135456017269,
+      "grad_norm": 4.272116184234619,
+      "learning_rate": 1.9945947486547617e-05,
+      "loss": 0.3725,
+      "step": 231
+    },
+    {
+      "epoch": 0.06260118726389638,
+      "grad_norm": 5.120586395263672,
+      "learning_rate": 1.994503606054994e-05,
+      "loss": 0.3918,
+      "step": 232
+    },
+    {
+      "epoch": 0.06287101996762008,
+      "grad_norm": 1.8483742475509644,
+      "learning_rate": 1.994411703566618e-05,
+      "loss": 0.3564,
+      "step": 233
+    },
+    {
+      "epoch": 0.06314085267134377,
+      "grad_norm": 1.8815995454788208,
+      "learning_rate": 1.994319041259854e-05,
+      "loss": 0.3622,
+      "step": 234
+    },
+    {
+      "epoch": 0.06341068537506746,
+      "grad_norm": 4.939080715179443,
+      "learning_rate": 1.9942256192055044e-05,
+      "loss": 0.3222,
+      "step": 235
+    },
+    {
+      "epoch": 0.06368051807879115,
+      "grad_norm": 3.727351665496826,
+      "learning_rate": 1.9941314374749524e-05,
+      "loss": 0.362,
+      "step": 236
+    },
+    {
+      "epoch": 0.06395035078251485,
+      "grad_norm": 3.959965467453003,
+      "learning_rate": 1.994036496140161e-05,
+      "loss": 0.3712,
+      "step": 237
+    },
+    {
+      "epoch": 0.06422018348623854,
+      "grad_norm": 5.5117621421813965,
+      "learning_rate": 1.9939407952736737e-05,
+      "loss": 0.3835,
+      "step": 238
+    },
+    {
+      "epoch": 0.06449001618996222,
+      "grad_norm": 0.9606245756149292,
+      "learning_rate": 1.9938443349486146e-05,
+      "loss": 0.3885,
+      "step": 239
+    },
+    {
+      "epoch": 0.06475984889368591,
+      "grad_norm": 2.727975368499756,
+      "learning_rate": 1.9937471152386876e-05,
+      "loss": 0.3712,
+      "step": 240
+    },
+    {
+      "epoch": 0.0650296815974096,
+      "grad_norm": 1.702797770500183,
+      "learning_rate": 1.9936491362181774e-05,
+      "loss": 0.3913,
+      "step": 241
+    },
+    {
+      "epoch": 0.0652995143011333,
+      "grad_norm": 1.4128336906433105,
+      "learning_rate": 1.9935503979619483e-05,
+      "loss": 0.3647,
+      "step": 242
+    },
+    {
+      "epoch": 0.06556934700485699,
+      "grad_norm": 3.3170485496520996,
+      "learning_rate": 1.9934509005454453e-05,
+      "loss": 0.3776,
+      "step": 243
+    },
+    {
+      "epoch": 0.06583917970858068,
+      "grad_norm": 1.3686492443084717,
+      "learning_rate": 1.9933506440446932e-05,
+      "loss": 0.3867,
+      "step": 244
+    },
+    {
+      "epoch": 0.06610901241230437,
+      "grad_norm": 7.256997585296631,
+      "learning_rate": 1.993249628536297e-05,
+      "loss": 0.3676,
+      "step": 245
+    },
+    {
+      "epoch": 0.06637884511602807,
+      "grad_norm": 2.9992611408233643,
+      "learning_rate": 1.9931478540974414e-05,
+      "loss": 0.3525,
+      "step": 246
+    },
+    {
+      "epoch": 0.06664867781975176,
+      "grad_norm": 9.596076011657715,
+      "learning_rate": 1.9930453208058905e-05,
+      "loss": 0.4155,
+      "step": 247
+    },
+    {
+      "epoch": 0.06691851052347544,
+      "grad_norm": 8.130996704101562,
+      "learning_rate": 1.9929420287399896e-05,
+      "loss": 0.3652,
+      "step": 248
+    },
+    {
+      "epoch": 0.06718834322719913,
+      "grad_norm": 6.9948015213012695,
+      "learning_rate": 1.9928379779786625e-05,
+      "loss": 0.3586,
+      "step": 249
+    },
+    {
+      "epoch": 0.06745817593092283,
+      "grad_norm": 8.444416046142578,
+      "learning_rate": 1.992733168601413e-05,
+      "loss": 0.3942,
+      "step": 250
+    },
+    {
+      "epoch": 0.06772800863464652,
+      "grad_norm": 8.40573787689209,
+      "learning_rate": 1.992627600688325e-05,
+      "loss": 0.3757,
+      "step": 251
+    },
+    {
+      "epoch": 0.06799784133837021,
+      "grad_norm": 9.380939483642578,
+      "learning_rate": 1.9925212743200623e-05,
+      "loss": 0.3687,
+      "step": 252
+    },
+    {
+      "epoch": 0.0682676740420939,
+      "grad_norm": 8.251864433288574,
+      "learning_rate": 1.9924141895778667e-05,
+      "loss": 0.3213,
+      "step": 253
+    },
+    {
+      "epoch": 0.06853750674581759,
+      "grad_norm": 8.825217247009277,
+      "learning_rate": 1.9923063465435607e-05,
+      "loss": 0.3858,
+      "step": 254
+    },
+    {
+      "epoch": 0.06880733944954129,
+      "grad_norm": 6.803030967712402,
+      "learning_rate": 1.9921977452995457e-05,
+      "loss": 0.3602,
+      "step": 255
+    },
+    {
+      "epoch": 0.06907717215326498,
+      "grad_norm": 5.716035842895508,
+      "learning_rate": 1.9920883859288035e-05,
+      "loss": 0.3636,
+      "step": 256
+    },
+    {
+      "epoch": 0.06934700485698866,
+      "grad_norm": 6.331687927246094,
+      "learning_rate": 1.991978268514893e-05,
+      "loss": 0.3982,
+      "step": 257
+    },
+    {
+      "epoch": 0.06961683756071235,
+      "grad_norm": 1.3983004093170166,
+      "learning_rate": 1.9918673931419552e-05,
+      "loss": 0.3296,
+      "step": 258
+    },
+    {
+      "epoch": 0.06988667026443605,
+      "grad_norm": 5.970175743103027,
+      "learning_rate": 1.9917557598947073e-05,
+      "loss": 0.3692,
+      "step": 259
+    },
+    {
+      "epoch": 0.07015650296815974,
+      "grad_norm": 1.1065860986709595,
+      "learning_rate": 1.9916433688584475e-05,
+      "loss": 0.3413,
+      "step": 260
+    },
+    {
+      "epoch": 0.07042633567188343,
+      "grad_norm": 4.800811290740967,
+      "learning_rate": 1.9915302201190526e-05,
+      "loss": 0.3599,
+      "step": 261
+    },
+    {
+      "epoch": 0.07069616837560712,
+      "grad_norm": 2.2967529296875,
+      "learning_rate": 1.991416313762978e-05,
+      "loss": 0.3397,
+      "step": 262
+    },
+    {
+      "epoch": 0.07096600107933082,
+      "grad_norm": 5.960307598114014,
+      "learning_rate": 1.9913016498772584e-05,
+      "loss": 0.3501,
+      "step": 263
+    },
+    {
+      "epoch": 0.07123583378305451,
+      "grad_norm": 1.2800002098083496,
+      "learning_rate": 1.9911862285495068e-05,
+      "loss": 0.3404,
+      "step": 264
+    },
+    {
+      "epoch": 0.0715056664867782,
+      "grad_norm": 1.382997989654541,
+      "learning_rate": 1.991070049867916e-05,
+      "loss": 0.3268,
+      "step": 265
+    },
+    {
+      "epoch": 0.07177549919050188,
+      "grad_norm": 1.562153935432434,
+      "learning_rate": 1.990953113921256e-05,
+      "loss": 0.328,
+      "step": 266
+    },
+    {
+      "epoch": 0.07204533189422559,
+      "grad_norm": 1.422719120979309,
+      "learning_rate": 1.990835420798877e-05,
+      "loss": 0.3598,
+      "step": 267
+    },
+    {
+      "epoch": 0.07231516459794927,
+      "grad_norm": 3.7587075233459473,
+      "learning_rate": 1.990716970590706e-05,
+      "loss": 0.3391,
+      "step": 268
+    },
+    {
+      "epoch": 0.07258499730167296,
+      "grad_norm": 3.140716791152954,
+      "learning_rate": 1.99059776338725e-05,
+      "loss": 0.3562,
+      "step": 269
+    },
+    {
+      "epoch": 0.07285483000539665,
+      "grad_norm": 1.8365578651428223,
+      "learning_rate": 1.9904777992795937e-05,
+      "loss": 0.3165,
+      "step": 270
+    },
+    {
+      "epoch": 0.07312466270912034,
+      "grad_norm": 1.7543565034866333,
+      "learning_rate": 1.9903570783594003e-05,
+      "loss": 0.3494,
+      "step": 271
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 4.742021560668945,
+      "learning_rate": 1.990235600718911e-05,
+      "loss": 0.3361,
+      "step": 272
+    },
+    {
+      "epoch": 0.07366432811656773,
+      "grad_norm": 2.0521650314331055,
+      "learning_rate": 1.990113366450946e-05,
+      "loss": 0.397,
+      "step": 273
+    },
+    {
+      "epoch": 0.07393416082029142,
+      "grad_norm": 5.402493476867676,
+      "learning_rate": 1.989990375648903e-05,
+      "loss": 0.3509,
+      "step": 274
+    },
+    {
+      "epoch": 0.0742039935240151,
+      "grad_norm": 3.3207602500915527,
+      "learning_rate": 1.9898666284067573e-05,
+      "loss": 0.279,
+      "step": 275
+    },
+    {
+      "epoch": 0.0744738262277388,
+      "grad_norm": 11.143855094909668,
+      "learning_rate": 1.989742124819063e-05,
+      "loss": 0.3707,
+      "step": 276
+    },
+    {
+      "epoch": 0.0747436589314625,
+      "grad_norm": 2.225947380065918,
+      "learning_rate": 1.9896168649809518e-05,
+      "loss": 0.3337,
+      "step": 277
+    },
+    {
+      "epoch": 0.07501349163518618,
+      "grad_norm": 14.69095516204834,
+      "learning_rate": 1.9894908489881336e-05,
+      "loss": 0.3684,
+      "step": 278
+    },
+    {
+      "epoch": 0.07528332433890987,
+      "grad_norm": 1.4998265504837036,
+      "learning_rate": 1.9893640769368955e-05,
+      "loss": 0.3355,
+      "step": 279
+    },
+    {
+      "epoch": 0.07555315704263357,
+      "grad_norm": 10.582429885864258,
+      "learning_rate": 1.9892365489241023e-05,
+      "loss": 0.3797,
+      "step": 280
+    },
+    {
+      "epoch": 0.07582298974635726,
+      "grad_norm": 3.837320566177368,
+      "learning_rate": 1.989108265047197e-05,
+      "loss": 0.3446,
+      "step": 281
+    },
+    {
+      "epoch": 0.07609282245008095,
+      "grad_norm": 10.683239936828613,
+      "learning_rate": 1.9889792254041998e-05,
+      "loss": 0.3287,
+      "step": 282
+    },
+    {
+      "epoch": 0.07636265515380464,
+      "grad_norm": 3.991856098175049,
+      "learning_rate": 1.9888494300937086e-05,
+      "loss": 0.3327,
+      "step": 283
+    },
+    {
+      "epoch": 0.07663248785752834,
+      "grad_norm": 6.164737701416016,
+      "learning_rate": 1.988718879214898e-05,
+      "loss": 0.3394,
+      "step": 284
+    },
+    {
+      "epoch": 0.07690232056125203,
+      "grad_norm": 5.921388626098633,
+      "learning_rate": 1.9885875728675204e-05,
+      "loss": 0.3359,
+      "step": 285
+    },
+    {
+      "epoch": 0.07717215326497572,
+      "grad_norm": 1.8078041076660156,
+      "learning_rate": 1.988455511151906e-05,
+      "loss": 0.3454,
+      "step": 286
+    },
+    {
+      "epoch": 0.0774419859686994,
+      "grad_norm": 11.108833312988281,
+      "learning_rate": 1.988322694168961e-05,
+      "loss": 0.3358,
+      "step": 287
+    },
+    {
+      "epoch": 0.07771181867242309,
+      "grad_norm": 1.682905673980713,
+      "learning_rate": 1.9881891220201697e-05,
+      "loss": 0.3245,
+      "step": 288
+    },
+    {
+      "epoch": 0.0779816513761468,
+      "grad_norm": 8.088534355163574,
+      "learning_rate": 1.9880547948075927e-05,
+      "loss": 0.3535,
+      "step": 289
+    },
+    {
+      "epoch": 0.07825148407987048,
+      "grad_norm": 6.454010963439941,
+      "learning_rate": 1.9879197126338684e-05,
+      "loss": 0.3202,
+      "step": 290
+    },
+    {
+      "epoch": 0.07852131678359417,
+      "grad_norm": 2.5465567111968994,
+      "learning_rate": 1.987783875602211e-05,
+      "loss": 0.3011,
+      "step": 291
+    },
+    {
+      "epoch": 0.07879114948731786,
+      "grad_norm": 13.461112022399902,
+      "learning_rate": 1.987647283816412e-05,
+      "loss": 0.3464,
+      "step": 292
+    },
+    {
+      "epoch": 0.07906098219104156,
+      "grad_norm": 1.1166651248931885,
+      "learning_rate": 1.98750993738084e-05,
+      "loss": 0.321,
+      "step": 293
+    },
+    {
+      "epoch": 0.07933081489476525,
+      "grad_norm": 8.680593490600586,
+      "learning_rate": 1.9873718364004398e-05,
+      "loss": 0.3711,
+      "step": 294
+    },
+    {
+      "epoch": 0.07960064759848894,
+      "grad_norm": 6.934235095977783,
+      "learning_rate": 1.9872329809807322e-05,
+      "loss": 0.3946,
+      "step": 295
+    },
+    {
+      "epoch": 0.07987048030221262,
+      "grad_norm": 7.020078659057617,
+      "learning_rate": 1.987093371227815e-05,
+      "loss": 0.3379,
+      "step": 296
+    },
+    {
+      "epoch": 0.08014031300593633,
+      "grad_norm": 9.334239959716797,
+      "learning_rate": 1.986953007248363e-05,
+      "loss": 0.3415,
+      "step": 297
+    },
+    {
+      "epoch": 0.08041014570966001,
+      "grad_norm": 3.428124189376831,
+      "learning_rate": 1.9868118891496268e-05,
+      "loss": 0.3407,
+      "step": 298
+    },
+    {
+      "epoch": 0.0806799784133837,
+      "grad_norm": 4.523345947265625,
+      "learning_rate": 1.9866700170394322e-05,
+      "loss": 0.301,
+      "step": 299
+    },
+    {
+      "epoch": 0.08094981111710739,
+      "grad_norm": 5.485284328460693,
+      "learning_rate": 1.9865273910261826e-05,
+      "loss": 0.3168,
+      "step": 300
+    },
+    {
+      "epoch": 0.08121964382083108,
+      "grad_norm": 4.186789035797119,
+      "learning_rate": 1.986384011218857e-05,
+      "loss": 0.3622,
+      "step": 301
+    },
+    {
+      "epoch": 0.08148947652455478,
+      "grad_norm": 16.344541549682617,
+      "learning_rate": 1.9862398777270097e-05,
+      "loss": 0.3753,
+      "step": 302
+    },
+    {
+      "epoch": 0.08175930922827847,
+      "grad_norm": 2.1884119510650635,
+      "learning_rate": 1.9860949906607718e-05,
+      "loss": 0.3722,
+      "step": 303
+    },
+    {
+      "epoch": 0.08202914193200216,
+      "grad_norm": 7.600675106048584,
+      "learning_rate": 1.98594935013085e-05,
+      "loss": 0.347,
+      "step": 304
+    },
+    {
+      "epoch": 0.08229897463572584,
+      "grad_norm": 8.81036376953125,
+      "learning_rate": 1.985802956248526e-05,
+      "loss": 0.3361,
+      "step": 305
+    },
+    {
+      "epoch": 0.08256880733944955,
+      "grad_norm": 1.6337834596633911,
+      "learning_rate": 1.9856558091256583e-05,
+      "loss": 0.3011,
+      "step": 306
+    },
+    {
+      "epoch": 0.08283864004317323,
+      "grad_norm": 6.589818954467773,
+      "learning_rate": 1.9855079088746794e-05,
+      "loss": 0.3205,
+      "step": 307
+    },
+    {
+      "epoch": 0.08310847274689692,
+      "grad_norm": 9.288436889648438,
+      "learning_rate": 1.9853592556085994e-05,
+      "loss": 0.337,
+      "step": 308
+    },
+    {
+      "epoch": 0.08337830545062061,
+      "grad_norm": 3.517489433288574,
+      "learning_rate": 1.985209849441001e-05,
+      "loss": 0.3799,
+      "step": 309
+    },
+    {
+      "epoch": 0.08364813815434431,
+      "grad_norm": 4.501574993133545,
+      "learning_rate": 1.985059690486045e-05,
+      "loss": 0.3204,
+      "step": 310
+    },
+    {
+      "epoch": 0.083917970858068,
+      "grad_norm": 5.554238319396973,
+      "learning_rate": 1.984908778858466e-05,
+      "loss": 0.3455,
+      "step": 311
+    },
+    {
+      "epoch": 0.08418780356179169,
+      "grad_norm": 1.5607404708862305,
+      "learning_rate": 1.984757114673573e-05,
+      "loss": 0.3387,
+      "step": 312
+    },
+    {
+      "epoch": 0.08445763626551538,
+      "grad_norm": 8.945670127868652,
+      "learning_rate": 1.9846046980472514e-05,
+      "loss": 0.3345,
+      "step": 313
+    },
+    {
+      "epoch": 0.08472746896923908,
+      "grad_norm": 0.9406250715255737,
+      "learning_rate": 1.9844515290959606e-05,
+      "loss": 0.3028,
+      "step": 314
+    },
+    {
+      "epoch": 0.08499730167296277,
+      "grad_norm": 1.963525414466858,
+      "learning_rate": 1.984297607936736e-05,
+      "loss": 0.3301,
+      "step": 315
+    },
+    {
+      "epoch": 0.08526713437668645,
+      "grad_norm": 5.224861145019531,
+      "learning_rate": 1.9841429346871863e-05,
+      "loss": 0.318,
+      "step": 316
+    },
+    {
+      "epoch": 0.08553696708041014,
+      "grad_norm": 3.2502832412719727,
+      "learning_rate": 1.9839875094654954e-05,
+      "loss": 0.2964,
+      "step": 317
+    },
+    {
+      "epoch": 0.08580679978413383,
+      "grad_norm": 2.45579195022583,
+      "learning_rate": 1.9838313323904225e-05,
+      "loss": 0.3364,
+      "step": 318
+    },
+    {
+      "epoch": 0.08607663248785753,
+      "grad_norm": 9.502269744873047,
+      "learning_rate": 1.9836744035812997e-05,
+      "loss": 0.3434,
+      "step": 319
+    },
+    {
+      "epoch": 0.08634646519158122,
+      "grad_norm": 2.209676742553711,
+      "learning_rate": 1.9835167231580355e-05,
+      "loss": 0.3514,
+      "step": 320
+    },
+    {
+      "epoch": 0.08661629789530491,
+      "grad_norm": 6.199123859405518,
+      "learning_rate": 1.9833582912411118e-05,
+      "loss": 0.3526,
+      "step": 321
+    },
+    {
+      "epoch": 0.0868861305990286,
+      "grad_norm": 6.509483814239502,
+      "learning_rate": 1.9831991079515836e-05,
+      "loss": 0.3528,
+      "step": 322
+    },
+    {
+      "epoch": 0.0871559633027523,
+      "grad_norm": 2.103527069091797,
+      "learning_rate": 1.9830391734110815e-05,
+      "loss": 0.2766,
+      "step": 323
+    },
+    {
+      "epoch": 0.08742579600647599,
+      "grad_norm": 8.02919864654541,
+      "learning_rate": 1.9828784877418098e-05,
+      "loss": 0.3615,
+      "step": 324
+    },
+    {
+      "epoch": 0.08769562871019967,
+      "grad_norm": 4.050201416015625,
+      "learning_rate": 1.9827170510665458e-05,
+      "loss": 0.3129,
+      "step": 325
+    },
+    {
+      "epoch": 0.08796546141392336,
+      "grad_norm": 2.9818665981292725,
+      "learning_rate": 1.9825548635086425e-05,
+      "loss": 0.2978,
+      "step": 326
+    },
+    {
+      "epoch": 0.08823529411764706,
+      "grad_norm": 5.805616855621338,
+      "learning_rate": 1.9823919251920247e-05,
+      "loss": 0.3141,
+      "step": 327
+    },
+    {
+      "epoch": 0.08850512682137075,
+      "grad_norm": 3.442366123199463,
+      "learning_rate": 1.982228236241192e-05,
+      "loss": 0.3356,
+      "step": 328
+    },
+    {
+      "epoch": 0.08877495952509444,
+      "grad_norm": 1.8598415851593018,
+      "learning_rate": 1.9820637967812167e-05,
+      "loss": 0.3226,
+      "step": 329
+    },
+    {
+      "epoch": 0.08904479222881813,
+      "grad_norm": 6.714463710784912,
+      "learning_rate": 1.981898606937746e-05,
+      "loss": 0.3251,
+      "step": 330
+    },
+    {
+      "epoch": 0.08931462493254183,
+      "grad_norm": 3.679516553878784,
+      "learning_rate": 1.9817326668369987e-05,
+      "loss": 0.3534,
+      "step": 331
+    },
+    {
+      "epoch": 0.08958445763626552,
+      "grad_norm": 3.2387406826019287,
+      "learning_rate": 1.9815659766057684e-05,
+      "loss": 0.278,
+      "step": 332
+    },
+    {
+      "epoch": 0.0898542903399892,
+      "grad_norm": 5.020092964172363,
+      "learning_rate": 1.9813985363714202e-05,
+      "loss": 0.2993,
+      "step": 333
+    },
+    {
+      "epoch": 0.0901241230437129,
+      "grad_norm": 2.6905510425567627,
+      "learning_rate": 1.9812303462618945e-05,
+      "loss": 0.3518,
+      "step": 334
+    },
+    {
+      "epoch": 0.09039395574743658,
+      "grad_norm": 2.8398759365081787,
+      "learning_rate": 1.981061406405702e-05,
+      "loss": 0.3153,
+      "step": 335
+    },
+    {
+      "epoch": 0.09066378845116028,
+      "grad_norm": 2.825042724609375,
+      "learning_rate": 1.980891716931929e-05,
+      "loss": 0.3191,
+      "step": 336
+    },
+    {
+      "epoch": 0.09093362115488397,
+      "grad_norm": 2.141221284866333,
+      "learning_rate": 1.9807212779702324e-05,
+      "loss": 0.3014,
+      "step": 337
+    },
+    {
+      "epoch": 0.09120345385860766,
+      "grad_norm": 1.3840841054916382,
+      "learning_rate": 1.980550089650843e-05,
+      "loss": 0.2944,
+      "step": 338
+    },
+    {
+      "epoch": 0.09147328656233135,
+      "grad_norm": 6.046553134918213,
+      "learning_rate": 1.9803781521045636e-05,
+      "loss": 0.3627,
+      "step": 339
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 1.110737681388855,
+      "learning_rate": 1.9802054654627694e-05,
+      "loss": 0.3203,
+      "step": 340
+    },
+    {
+      "epoch": 0.09201295196977874,
+      "grad_norm": 4.467282295227051,
+      "learning_rate": 1.980032029857409e-05,
+      "loss": 0.3307,
+      "step": 341
+    },
+    {
+      "epoch": 0.09228278467350243,
+      "grad_norm": 2.7430059909820557,
+      "learning_rate": 1.9798578454210024e-05,
+      "loss": 0.2906,
+      "step": 342
+    },
+    {
+      "epoch": 0.09255261737722612,
+      "grad_norm": 1.3973308801651,
+      "learning_rate": 1.979682912286641e-05,
+      "loss": 0.2834,
+      "step": 343
+    },
+    {
+      "epoch": 0.09282245008094982,
+      "grad_norm": 3.5317599773406982,
+      "learning_rate": 1.9795072305879897e-05,
+      "loss": 0.291,
+      "step": 344
+    },
+    {
+      "epoch": 0.0930922827846735,
+      "grad_norm": 2.7944188117980957,
+      "learning_rate": 1.9793308004592854e-05,
+      "loss": 0.3368,
+      "step": 345
+    },
+    {
+      "epoch": 0.0933621154883972,
+      "grad_norm": 2.910141944885254,
+      "learning_rate": 1.9791536220353355e-05,
+      "loss": 0.3639,
+      "step": 346
+    },
+    {
+      "epoch": 0.09363194819212088,
+      "grad_norm": 4.570195198059082,
+      "learning_rate": 1.9789756954515205e-05,
+      "loss": 0.3089,
+      "step": 347
+    },
+    {
+      "epoch": 0.09390178089584458,
+      "grad_norm": 1.7247203588485718,
+      "learning_rate": 1.9787970208437917e-05,
+      "loss": 0.3188,
+      "step": 348
+    },
+    {
+      "epoch": 0.09417161359956827,
+      "grad_norm": 2.3574259281158447,
+      "learning_rate": 1.9786175983486725e-05,
+      "loss": 0.3494,
+      "step": 349
+    },
+    {
+      "epoch": 0.09444144630329196,
+      "grad_norm": 6.170711517333984,
+      "learning_rate": 1.978437428103257e-05,
+      "loss": 0.3241,
+      "step": 350
+    },
+    {
+      "epoch": 0.09471127900701565,
+      "grad_norm": 2.2345502376556396,
+      "learning_rate": 1.978256510245212e-05,
+      "loss": 0.3499,
+      "step": 351
+    },
+    {
+      "epoch": 0.09498111171073934,
+      "grad_norm": 5.901663780212402,
+      "learning_rate": 1.9780748449127745e-05,
+      "loss": 0.3219,
+      "step": 352
+    },
+    {
+      "epoch": 0.09525094441446304,
+      "grad_norm": 5.652328014373779,
+      "learning_rate": 1.9778924322447527e-05,
+      "loss": 0.2768,
+      "step": 353
+    },
+    {
+      "epoch": 0.09552077711818673,
+      "grad_norm": 2.8446595668792725,
+      "learning_rate": 1.9777092723805258e-05,
+      "loss": 0.3522,
+      "step": 354
+    },
+    {
+      "epoch": 0.09579060982191041,
+      "grad_norm": 4.041268825531006,
+      "learning_rate": 1.9775253654600443e-05,
+      "loss": 0.3087,
+      "step": 355
+    },
+    {
+      "epoch": 0.0960604425256341,
+      "grad_norm": 5.432677268981934,
+      "learning_rate": 1.9773407116238293e-05,
+      "loss": 0.3203,
+      "step": 356
+    },
+    {
+      "epoch": 0.0963302752293578,
+      "grad_norm": 1.7260606288909912,
+      "learning_rate": 1.9771553110129727e-05,
+      "loss": 0.2989,
+      "step": 357
+    },
+    {
+      "epoch": 0.09660010793308149,
+      "grad_norm": 7.146731376647949,
+      "learning_rate": 1.976969163769137e-05,
+      "loss": 0.3218,
+      "step": 358
+    },
+    {
+      "epoch": 0.09686994063680518,
+      "grad_norm": 3.925907850265503,
+      "learning_rate": 1.976782270034555e-05,
+      "loss": 0.2954,
+      "step": 359
+    },
+    {
+      "epoch": 0.09713977334052887,
+      "grad_norm": 6.000894069671631,
+      "learning_rate": 1.9765946299520302e-05,
+      "loss": 0.3594,
+      "step": 360
+    },
+    {
+      "epoch": 0.09740960604425257,
+      "grad_norm": 3.1758873462677,
+      "learning_rate": 1.9764062436649362e-05,
+      "loss": 0.3077,
+      "step": 361
+    },
+    {
+      "epoch": 0.09767943874797626,
+      "grad_norm": 1.028581142425537,
+      "learning_rate": 1.9762171113172166e-05,
+      "loss": 0.3389,
+      "step": 362
+    },
+    {
+      "epoch": 0.09794927145169995,
+      "grad_norm": 1.4201823472976685,
+      "learning_rate": 1.9760272330533853e-05,
+      "loss": 0.3324,
+      "step": 363
+    },
+    {
+      "epoch": 0.09821910415542363,
+      "grad_norm": 2.846605062484741,
+      "learning_rate": 1.9758366090185255e-05,
+      "loss": 0.3704,
+      "step": 364
+    },
+    {
+      "epoch": 0.09848893685914732,
+      "grad_norm": 1.7962620258331299,
+      "learning_rate": 1.975645239358292e-05,
+      "loss": 0.3064,
+      "step": 365
+    },
+    {
+      "epoch": 0.09875876956287102,
+      "grad_norm": 2.338270664215088,
+      "learning_rate": 1.9754531242189075e-05,
+      "loss": 0.3019,
+      "step": 366
+    },
+    {
+      "epoch": 0.09902860226659471,
+      "grad_norm": 4.45302152633667,
+      "learning_rate": 1.9752602637471646e-05,
+      "loss": 0.3248,
+      "step": 367
+    },
+    {
+      "epoch": 0.0992984349703184,
+      "grad_norm": 1.3446452617645264,
+      "learning_rate": 1.9750666580904263e-05,
+      "loss": 0.3455,
+      "step": 368
+    },
+    {
+      "epoch": 0.09956826767404209,
+      "grad_norm": 5.155446529388428,
+      "learning_rate": 1.974872307396624e-05,
+      "loss": 0.3448,
+      "step": 369
+    },
+    {
+      "epoch": 0.09983810037776579,
+      "grad_norm": 2.0688929557800293,
+      "learning_rate": 1.974677211814259e-05,
+      "loss": 0.2858,
+      "step": 370
+    },
+    {
+      "epoch": 0.10010793308148948,
+      "grad_norm": 1.3265455961227417,
+      "learning_rate": 1.9744813714924017e-05,
+      "loss": 0.3302,
+      "step": 371
+    },
+    {
+      "epoch": 0.10037776578521317,
+      "grad_norm": 4.006172180175781,
+      "learning_rate": 1.9742847865806913e-05,
+      "loss": 0.3407,
+      "step": 372
+    },
+    {
+      "epoch": 0.10064759848893685,
+      "grad_norm": 1.429436445236206,
+      "learning_rate": 1.9740874572293357e-05,
+      "loss": 0.3159,
+      "step": 373
+    },
+    {
+      "epoch": 0.10091743119266056,
+      "grad_norm": 1.9584721326828003,
+      "learning_rate": 1.9738893835891122e-05,
+      "loss": 0.3225,
+      "step": 374
+    },
+    {
+      "epoch": 0.10118726389638424,
+      "grad_norm": 6.1032185554504395,
+      "learning_rate": 1.9736905658113667e-05,
+      "loss": 0.3223,
+      "step": 375
+    },
+    {
+      "epoch": 0.10145709660010793,
+      "grad_norm": 1.8192338943481445,
+      "learning_rate": 1.973491004048014e-05,
+      "loss": 0.3099,
+      "step": 376
+    },
+    {
+      "epoch": 0.10172692930383162,
+      "grad_norm": 4.014854431152344,
+      "learning_rate": 1.9732906984515355e-05,
+      "loss": 0.2953,
+      "step": 377
+    },
+    {
+      "epoch": 0.10199676200755532,
+      "grad_norm": 3.2682042121887207,
+      "learning_rate": 1.9730896491749836e-05,
+      "loss": 0.2948,
+      "step": 378
+    },
+    {
+      "epoch": 0.10226659471127901,
+      "grad_norm": 2.437673807144165,
+      "learning_rate": 1.9728878563719776e-05,
+      "loss": 0.3218,
+      "step": 379
+    },
+    {
+      "epoch": 0.1025364274150027,
+      "grad_norm": 5.997743129730225,
+      "learning_rate": 1.972685320196704e-05,
+      "loss": 0.3109,
+      "step": 380
+    },
+    {
+      "epoch": 0.10280626011872639,
+      "grad_norm": 2.6550042629241943,
+      "learning_rate": 1.97248204080392e-05,
+      "loss": 0.3296,
+      "step": 381
+    },
+    {
+      "epoch": 0.10307609282245007,
+      "grad_norm": 3.3859405517578125,
+      "learning_rate": 1.9722780183489477e-05,
+      "loss": 0.339,
+      "step": 382
+    },
+    {
+      "epoch": 0.10334592552617378,
+      "grad_norm": 1.1447583436965942,
+      "learning_rate": 1.9720732529876785e-05,
+      "loss": 0.2798,
+      "step": 383
+    },
+    {
+      "epoch": 0.10361575822989746,
+      "grad_norm": 2.2558467388153076,
+      "learning_rate": 1.9718677448765718e-05,
+      "loss": 0.3062,
+      "step": 384
+    },
+    {
+      "epoch": 0.10388559093362115,
+      "grad_norm": 1.7118128538131714,
+      "learning_rate": 1.9716614941726533e-05,
+      "loss": 0.2998,
+      "step": 385
+    },
+    {
+      "epoch": 0.10415542363734484,
+      "grad_norm": 2.5867927074432373,
+      "learning_rate": 1.971454501033517e-05,
+      "loss": 0.324,
+      "step": 386
+    },
+    {
+      "epoch": 0.10442525634106854,
+      "grad_norm": 2.621339797973633,
+      "learning_rate": 1.971246765617324e-05,
+      "loss": 0.315,
+      "step": 387
+    },
+    {
+      "epoch": 0.10469508904479223,
+      "grad_norm": 2.1263391971588135,
+      "learning_rate": 1.9710382880828028e-05,
+      "loss": 0.3456,
+      "step": 388
+    },
+    {
+      "epoch": 0.10496492174851592,
+      "grad_norm": 1.3193202018737793,
+      "learning_rate": 1.970829068589248e-05,
+      "loss": 0.3196,
+      "step": 389
+    },
+    {
+      "epoch": 0.1052347544522396,
+      "grad_norm": 2.072350263595581,
+      "learning_rate": 1.970619107296522e-05,
+      "loss": 0.3008,
+      "step": 390
+    },
+    {
+      "epoch": 0.10550458715596331,
+      "grad_norm": 1.3679946660995483,
+      "learning_rate": 1.9704084043650545e-05,
+      "loss": 0.3106,
+      "step": 391
+    },
+    {
+      "epoch": 0.105774419859687,
+      "grad_norm": 1.7540134191513062,
+      "learning_rate": 1.9701969599558398e-05,
+      "loss": 0.2757,
+      "step": 392
+    },
+    {
+      "epoch": 0.10604425256341068,
+      "grad_norm": 3.070902109146118,
+      "learning_rate": 1.9699847742304414e-05,
+      "loss": 0.3152,
+      "step": 393
+    },
+    {
+      "epoch": 0.10631408526713437,
+      "grad_norm": 1.2945635318756104,
+      "learning_rate": 1.969771847350987e-05,
+      "loss": 0.3061,
+      "step": 394
+    },
+    {
+      "epoch": 0.10658391797085807,
+      "grad_norm": 1.9487013816833496,
+      "learning_rate": 1.969558179480172e-05,
+      "loss": 0.3545,
+      "step": 395
+    },
+    {
+      "epoch": 0.10685375067458176,
+      "grad_norm": 1.8175294399261475,
+      "learning_rate": 1.9693437707812577e-05,
+      "loss": 0.3175,
+      "step": 396
+    },
+    {
+      "epoch": 0.10712358337830545,
+      "grad_norm": 1.0505638122558594,
+      "learning_rate": 1.9691286214180708e-05,
+      "loss": 0.2875,
+      "step": 397
+    },
+    {
+      "epoch": 0.10739341608202914,
+      "grad_norm": 2.234910726547241,
+      "learning_rate": 1.9689127315550047e-05,
+      "loss": 0.2899,
+      "step": 398
+    },
+    {
+      "epoch": 0.10766324878575283,
+      "grad_norm": 3.4617631435394287,
+      "learning_rate": 1.9686961013570186e-05,
+      "loss": 0.3334,
+      "step": 399
+    },
+    {
+      "epoch": 0.10793308148947653,
+      "grad_norm": 1.361401915550232,
+      "learning_rate": 1.968478730989636e-05,
+      "loss": 0.309,
+      "step": 400
+    },
+    {
+      "epoch": 0.10820291419320022,
+      "grad_norm": 2.939791202545166,
+      "learning_rate": 1.9682606206189486e-05,
+      "loss": 0.3341,
+      "step": 401
+    },
+    {
+      "epoch": 0.1084727468969239,
+      "grad_norm": 1.6604571342468262,
+      "learning_rate": 1.968041770411611e-05,
+      "loss": 0.2991,
+      "step": 402
+    },
+    {
+      "epoch": 0.1087425796006476,
+      "grad_norm": 0.9691702127456665,
+      "learning_rate": 1.9678221805348445e-05,
+      "loss": 0.2974,
+      "step": 403
+    },
+    {
+      "epoch": 0.1090124123043713,
+      "grad_norm": 3.353623390197754,
+      "learning_rate": 1.967601851156435e-05,
+      "loss": 0.3149,
+      "step": 404
+    },
+    {
+      "epoch": 0.10928224500809498,
+      "grad_norm": 1.2448843717575073,
+      "learning_rate": 1.9673807824447333e-05,
+      "loss": 0.3352,
+      "step": 405
+    },
+    {
+      "epoch": 0.10955207771181867,
+      "grad_norm": 1.8166934251785278,
+      "learning_rate": 1.9671589745686563e-05,
+      "loss": 0.2835,
+      "step": 406
+    },
+    {
+      "epoch": 0.10982191041554236,
+      "grad_norm": 1.444082498550415,
+      "learning_rate": 1.9669364276976836e-05,
+      "loss": 0.3126,
+      "step": 407
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 1.9386342763900757,
+      "learning_rate": 1.9667131420018615e-05,
+      "loss": 0.3003,
+      "step": 408
+    },
+    {
+      "epoch": 0.11036157582298975,
+      "grad_norm": 1.1757546663284302,
+      "learning_rate": 1.9664891176518e-05,
+      "loss": 0.2951,
+      "step": 409
+    },
+    {
+      "epoch": 0.11063140852671344,
+      "grad_norm": 2.7657134532928467,
+      "learning_rate": 1.966264354818673e-05,
+      "loss": 0.3008,
+      "step": 410
+    },
+    {
+      "epoch": 0.11090124123043713,
+      "grad_norm": 3.027733564376831,
+      "learning_rate": 1.96603885367422e-05,
+      "loss": 0.3238,
+      "step": 411
+    },
+    {
+      "epoch": 0.11117107393416081,
+      "grad_norm": 3.135148048400879,
+      "learning_rate": 1.965812614390743e-05,
+      "loss": 0.316,
+      "step": 412
+    },
+    {
+      "epoch": 0.11144090663788452,
+      "grad_norm": 2.194401979446411,
+      "learning_rate": 1.965585637141109e-05,
+      "loss": 0.3019,
+      "step": 413
+    },
+    {
+      "epoch": 0.1117107393416082,
+      "grad_norm": 1.0952454805374146,
+      "learning_rate": 1.9653579220987488e-05,
+      "loss": 0.3228,
+      "step": 414
+    },
+    {
+      "epoch": 0.11198057204533189,
+      "grad_norm": 2.3813652992248535,
+      "learning_rate": 1.9651294694376566e-05,
+      "loss": 0.3289,
+      "step": 415
+    },
+    {
+      "epoch": 0.11225040474905558,
+      "grad_norm": 1.100504755973816,
+      "learning_rate": 1.964900279332391e-05,
+      "loss": 0.3224,
+      "step": 416
+    },
+    {
+      "epoch": 0.11252023745277928,
+      "grad_norm": 2.431295394897461,
+      "learning_rate": 1.9646703519580724e-05,
+      "loss": 0.3124,
+      "step": 417
+    },
+    {
+      "epoch": 0.11279007015650297,
+      "grad_norm": 1.9856544733047485,
+      "learning_rate": 1.9644396874903865e-05,
+      "loss": 0.3194,
+      "step": 418
+    },
+    {
+      "epoch": 0.11305990286022666,
+      "grad_norm": 3.094207763671875,
+      "learning_rate": 1.9642082861055812e-05,
+      "loss": 0.3288,
+      "step": 419
+    },
+    {
+      "epoch": 0.11332973556395035,
+      "grad_norm": 1.5447479486465454,
+      "learning_rate": 1.9639761479804677e-05,
+      "loss": 0.3305,
+      "step": 420
+    },
+    {
+      "epoch": 0.11359956826767405,
+      "grad_norm": 1.751523733139038,
+      "learning_rate": 1.9637432732924196e-05,
+      "loss": 0.3193,
+      "step": 421
+    },
+    {
+      "epoch": 0.11386940097139774,
+      "grad_norm": 2.4769203662872314,
+      "learning_rate": 1.963509662219374e-05,
+      "loss": 0.292,
+      "step": 422
+    },
+    {
+      "epoch": 0.11413923367512142,
+      "grad_norm": 1.4808107614517212,
+      "learning_rate": 1.9632753149398303e-05,
+      "loss": 0.2925,
+      "step": 423
+    },
+    {
+      "epoch": 0.11440906637884511,
+      "grad_norm": 2.1203739643096924,
+      "learning_rate": 1.9630402316328506e-05,
+      "loss": 0.3203,
+      "step": 424
+    },
+    {
+      "epoch": 0.11467889908256881,
+      "grad_norm": 0.968853235244751,
+      "learning_rate": 1.9628044124780597e-05,
+      "loss": 0.2994,
+      "step": 425
+    },
+    {
+      "epoch": 0.1149487317862925,
+      "grad_norm": 1.6296722888946533,
+      "learning_rate": 1.9625678576556436e-05,
+      "loss": 0.3026,
+      "step": 426
+    },
+    {
+      "epoch": 0.11521856449001619,
+      "grad_norm": 3.231654167175293,
+      "learning_rate": 1.9623305673463512e-05,
+      "loss": 0.3271,
+      "step": 427
+    },
+    {
+      "epoch": 0.11548839719373988,
+      "grad_norm": 3.565884590148926,
+      "learning_rate": 1.9620925417314934e-05,
+      "loss": 0.2728,
+      "step": 428
+    },
+    {
+      "epoch": 0.11575822989746357,
+      "grad_norm": 1.1069263219833374,
+      "learning_rate": 1.961853780992943e-05,
+      "loss": 0.3263,
+      "step": 429
+    },
+    {
+      "epoch": 0.11602806260118727,
+      "grad_norm": 0.9990012049674988,
+      "learning_rate": 1.9616142853131342e-05,
+      "loss": 0.27,
+      "step": 430
+    },
+    {
+      "epoch": 0.11629789530491096,
+      "grad_norm": 2.134874105453491,
+      "learning_rate": 1.9613740548750625e-05,
+      "loss": 0.3179,
+      "step": 431
+    },
+    {
+      "epoch": 0.11656772800863464,
+      "grad_norm": 4.530993938446045,
+      "learning_rate": 1.9611330898622856e-05,
+      "loss": 0.3718,
+      "step": 432
+    },
+    {
+      "epoch": 0.11683756071235833,
+      "grad_norm": 3.061640977859497,
+      "learning_rate": 1.9608913904589216e-05,
+      "loss": 0.2602,
+      "step": 433
+    },
+    {
+      "epoch": 0.11710739341608203,
+      "grad_norm": 4.465636253356934,
+      "learning_rate": 1.9606489568496506e-05,
+      "loss": 0.3265,
+      "step": 434
+    },
+    {
+      "epoch": 0.11737722611980572,
+      "grad_norm": 3.3101086616516113,
+      "learning_rate": 1.9604057892197134e-05,
+      "loss": 0.3295,
+      "step": 435
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 0.9853208065032959,
+      "learning_rate": 1.9601618877549113e-05,
+      "loss": 0.2645,
+      "step": 436
+    },
+    {
+      "epoch": 0.1179168915272531,
+      "grad_norm": 1.778331995010376,
+      "learning_rate": 1.9599172526416063e-05,
+      "loss": 0.3187,
+      "step": 437
+    },
+    {
+      "epoch": 0.1181867242309768,
+      "grad_norm": 1.9473369121551514,
+      "learning_rate": 1.9596718840667218e-05,
+      "loss": 0.2958,
+      "step": 438
+    },
+    {
+      "epoch": 0.11845655693470049,
+      "grad_norm": 2.7204747200012207,
+      "learning_rate": 1.9594257822177408e-05,
+      "loss": 0.2588,
+      "step": 439
+    },
+    {
+      "epoch": 0.11872638963842418,
+      "grad_norm": 0.9458572864532471,
+      "learning_rate": 1.959178947282707e-05,
+      "loss": 0.296,
+      "step": 440
+    },
+    {
+      "epoch": 0.11899622234214786,
+      "grad_norm": 4.038306713104248,
+      "learning_rate": 1.9589313794502235e-05,
+      "loss": 0.348,
+      "step": 441
+    },
+    {
+      "epoch": 0.11926605504587157,
+      "grad_norm": 1.005253791809082,
+      "learning_rate": 1.9586830789094548e-05,
+      "loss": 0.2802,
+      "step": 442
+    },
+    {
+      "epoch": 0.11953588774959525,
+      "grad_norm": 3.966348648071289,
+      "learning_rate": 1.958434045850124e-05,
+      "loss": 0.3654,
+      "step": 443
+    },
+    {
+      "epoch": 0.11980572045331894,
+      "grad_norm": 1.3045215606689453,
+      "learning_rate": 1.9581842804625146e-05,
+      "loss": 0.3003,
+      "step": 444
+    },
+    {
+      "epoch": 0.12007555315704263,
+      "grad_norm": 0.9951353669166565,
+      "learning_rate": 1.957933782937469e-05,
+      "loss": 0.3061,
+      "step": 445
+    },
+    {
+      "epoch": 0.12034538586076632,
+      "grad_norm": 4.687140464782715,
+      "learning_rate": 1.95768255346639e-05,
+      "loss": 0.3382,
+      "step": 446
+    },
+    {
+      "epoch": 0.12061521856449002,
+      "grad_norm": 3.2950189113616943,
+      "learning_rate": 1.9574305922412388e-05,
+      "loss": 0.3234,
+      "step": 447
+    },
+    {
+      "epoch": 0.12088505126821371,
+      "grad_norm": 3.57102108001709,
+      "learning_rate": 1.9571778994545356e-05,
+      "loss": 0.3142,
+      "step": 448
+    },
+    {
+      "epoch": 0.1211548839719374,
+      "grad_norm": 6.206331253051758,
+      "learning_rate": 1.9569244752993607e-05,
+      "loss": 0.3411,
+      "step": 449
+    },
+    {
+      "epoch": 0.12142471667566108,
+      "grad_norm": 1.2747764587402344,
+      "learning_rate": 1.9566703199693525e-05,
+      "loss": 0.3061,
+      "step": 450
+    },
+    {
+      "epoch": 0.12169454937938479,
+      "grad_norm": 5.835274696350098,
+      "learning_rate": 1.956415433658708e-05,
+      "loss": 0.3127,
+      "step": 451
+    },
+    {
+      "epoch": 0.12196438208310847,
+      "grad_norm": 3.900926351547241,
+      "learning_rate": 1.9561598165621823e-05,
+      "loss": 0.3012,
+      "step": 452
+    },
+    {
+      "epoch": 0.12223421478683216,
+      "grad_norm": 1.1747643947601318,
+      "learning_rate": 1.9559034688750906e-05,
+      "loss": 0.3148,
+      "step": 453
+    },
+    {
+      "epoch": 0.12250404749055585,
+      "grad_norm": 3.9242682456970215,
+      "learning_rate": 1.9556463907933038e-05,
+      "loss": 0.3116,
+      "step": 454
+    },
+    {
+      "epoch": 0.12277388019427955,
+      "grad_norm": 1.3817052841186523,
+      "learning_rate": 1.9553885825132536e-05,
+      "loss": 0.2986,
+      "step": 455
+    },
+    {
+      "epoch": 0.12304371289800324,
+      "grad_norm": 2.6693453788757324,
+      "learning_rate": 1.955130044231927e-05,
+      "loss": 0.284,
+      "step": 456
+    },
+    {
+      "epoch": 0.12331354560172693,
+      "grad_norm": 2.526277780532837,
+      "learning_rate": 1.9548707761468715e-05,
+      "loss": 0.2927,
+      "step": 457
+    },
+    {
+      "epoch": 0.12358337830545062,
+      "grad_norm": 7.084099292755127,
+      "learning_rate": 1.9546107784561897e-05,
+      "loss": 0.3653,
+      "step": 458
+    },
+    {
+      "epoch": 0.12385321100917432,
+      "grad_norm": 1.7276091575622559,
+      "learning_rate": 1.954350051358543e-05,
+      "loss": 0.2926,
+      "step": 459
+    },
+    {
+      "epoch": 0.124123043712898,
+      "grad_norm": 5.75747537612915,
+      "learning_rate": 1.9540885950531507e-05,
+      "loss": 0.2896,
+      "step": 460
+    },
+    {
+      "epoch": 0.1243928764166217,
+      "grad_norm": 1.3438332080841064,
+      "learning_rate": 1.9538264097397873e-05,
+      "loss": 0.3321,
+      "step": 461
+    },
+    {
+      "epoch": 0.12466270912034538,
+      "grad_norm": 1.146366834640503,
+      "learning_rate": 1.9535634956187867e-05,
+      "loss": 0.2959,
+      "step": 462
+    },
+    {
+      "epoch": 0.12493254182406907,
+      "grad_norm": 4.385280132293701,
+      "learning_rate": 1.953299852891038e-05,
+      "loss": 0.3166,
+      "step": 463
+    },
+    {
+      "epoch": 0.12520237452779276,
+      "grad_norm": 1.4171056747436523,
+      "learning_rate": 1.953035481757987e-05,
+      "loss": 0.3123,
+      "step": 464
+    },
+    {
+      "epoch": 0.12547220723151645,
+      "grad_norm": 3.098209857940674,
+      "learning_rate": 1.9527703824216374e-05,
+      "loss": 0.3114,
+      "step": 465
+    },
+    {
+      "epoch": 0.12574203993524016,
+      "grad_norm": 1.539747714996338,
+      "learning_rate": 1.9525045550845482e-05,
+      "loss": 0.3373,
+      "step": 466
+    },
+    {
+      "epoch": 0.12601187263896385,
+      "grad_norm": 0.8720614910125732,
+      "learning_rate": 1.952237999949835e-05,
+      "loss": 0.295,
+      "step": 467
+    },
+    {
+      "epoch": 0.12628170534268754,
+      "grad_norm": 2.24847149848938,
+      "learning_rate": 1.9519707172211696e-05,
+      "loss": 0.306,
+      "step": 468
+    },
+    {
+      "epoch": 0.12655153804641123,
+      "grad_norm": 0.8678748607635498,
+      "learning_rate": 1.9517027071027796e-05,
+      "loss": 0.3293,
+      "step": 469
+    },
+    {
+      "epoch": 0.12682137075013492,
+      "grad_norm": 2.57658052444458,
+      "learning_rate": 1.951433969799448e-05,
+      "loss": 0.3332,
+      "step": 470
+    },
+    {
+      "epoch": 0.1270912034538586,
+      "grad_norm": 2.2076215744018555,
+      "learning_rate": 1.951164505516514e-05,
+      "loss": 0.329,
+      "step": 471
+    },
+    {
+      "epoch": 0.1273610361575823,
+      "grad_norm": 1.8399533033370972,
+      "learning_rate": 1.9508943144598726e-05,
+      "loss": 0.289,
+      "step": 472
+    },
+    {
+      "epoch": 0.12763086886130598,
+      "grad_norm": 1.1825848817825317,
+      "learning_rate": 1.9506233968359734e-05,
+      "loss": 0.3292,
+      "step": 473
+    },
+    {
+      "epoch": 0.1279007015650297,
+      "grad_norm": 4.158936977386475,
+      "learning_rate": 1.950351752851821e-05,
+      "loss": 0.2838,
+      "step": 474
+    },
+    {
+      "epoch": 0.12817053426875338,
+      "grad_norm": 0.9815038442611694,
+      "learning_rate": 1.9500793827149756e-05,
+      "loss": 0.3322,
+      "step": 475
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.9764858484268188,
+      "learning_rate": 1.9498062866335522e-05,
+      "loss": 0.2591,
+      "step": 476
+    },
+    {
+      "epoch": 0.12871019967620076,
+      "grad_norm": 1.9159437417984009,
+      "learning_rate": 1.9495324648162207e-05,
+      "loss": 0.3117,
+      "step": 477
+    },
+    {
+      "epoch": 0.12898003237992445,
+      "grad_norm": 2.1342008113861084,
+      "learning_rate": 1.9492579174722043e-05,
+      "loss": 0.3226,
+      "step": 478
+    },
+    {
+      "epoch": 0.12924986508364814,
+      "grad_norm": 1.6109669208526611,
+      "learning_rate": 1.948982644811282e-05,
+      "loss": 0.2958,
+      "step": 479
+    },
+    {
+      "epoch": 0.12951969778737182,
+      "grad_norm": 4.04025936126709,
+      "learning_rate": 1.9487066470437863e-05,
+      "loss": 0.3247,
+      "step": 480
+    },
+    {
+      "epoch": 0.1297895304910955,
+      "grad_norm": 1.217788815498352,
+      "learning_rate": 1.9484299243806036e-05,
+      "loss": 0.3112,
+      "step": 481
+    },
+    {
+      "epoch": 0.1300593631948192,
+      "grad_norm": 3.1797890663146973,
+      "learning_rate": 1.9481524770331748e-05,
+      "loss": 0.3012,
+      "step": 482
+    },
+    {
+      "epoch": 0.13032919589854292,
+      "grad_norm": 1.6939276456832886,
+      "learning_rate": 1.9478743052134944e-05,
+      "loss": 0.2848,
+      "step": 483
+    },
+    {
+      "epoch": 0.1305990286022666,
+      "grad_norm": 1.0560511350631714,
+      "learning_rate": 1.9475954091341098e-05,
+      "loss": 0.2867,
+      "step": 484
+    },
+    {
+      "epoch": 0.1308688613059903,
+      "grad_norm": 3.809028148651123,
+      "learning_rate": 1.9473157890081223e-05,
+      "loss": 0.3168,
+      "step": 485
+    },
+    {
+      "epoch": 0.13113869400971398,
+      "grad_norm": 3.545024871826172,
+      "learning_rate": 1.947035445049186e-05,
+      "loss": 0.3189,
+      "step": 486
+    },
+    {
+      "epoch": 0.13140852671343767,
+      "grad_norm": 3.876812219619751,
+      "learning_rate": 1.9467543774715098e-05,
+      "loss": 0.2954,
+      "step": 487
+    },
+    {
+      "epoch": 0.13167835941716136,
+      "grad_norm": 2.816450595855713,
+      "learning_rate": 1.9464725864898527e-05,
+      "loss": 0.316,
+      "step": 488
+    },
+    {
+      "epoch": 0.13194819212088504,
+      "grad_norm": 1.6212997436523438,
+      "learning_rate": 1.9461900723195284e-05,
+      "loss": 0.3355,
+      "step": 489
+    },
+    {
+      "epoch": 0.13221802482460873,
+      "grad_norm": 4.111842632293701,
+      "learning_rate": 1.9459068351764032e-05,
+      "loss": 0.3136,
+      "step": 490
+    },
+    {
+      "epoch": 0.13248785752833242,
+      "grad_norm": 2.541093111038208,
+      "learning_rate": 1.9456228752768946e-05,
+      "loss": 0.2904,
+      "step": 491
+    },
+    {
+      "epoch": 0.13275769023205614,
+      "grad_norm": 3.381216526031494,
+      "learning_rate": 1.9453381928379732e-05,
+      "loss": 0.3622,
+      "step": 492
+    },
+    {
+      "epoch": 0.13302752293577982,
+      "grad_norm": 3.2044320106506348,
+      "learning_rate": 1.9450527880771618e-05,
+      "loss": 0.3135,
+      "step": 493
+    },
+    {
+      "epoch": 0.1332973556395035,
+      "grad_norm": 1.1519147157669067,
+      "learning_rate": 1.9447666612125348e-05,
+      "loss": 0.3414,
+      "step": 494
+    },
+    {
+      "epoch": 0.1335671883432272,
+      "grad_norm": 3.9146101474761963,
+      "learning_rate": 1.944479812462718e-05,
+      "loss": 0.3093,
+      "step": 495
+    },
+    {
+      "epoch": 0.1338370210469509,
+      "grad_norm": 1.8818269968032837,
+      "learning_rate": 1.94419224204689e-05,
+      "loss": 0.3135,
+      "step": 496
+    },
+    {
+      "epoch": 0.13410685375067458,
+      "grad_norm": 4.183235168457031,
+      "learning_rate": 1.9439039501847788e-05,
+      "loss": 0.3153,
+      "step": 497
+    },
+    {
+      "epoch": 0.13437668645439826,
+      "grad_norm": 2.2878146171569824,
+      "learning_rate": 1.943614937096666e-05,
+      "loss": 0.2884,
+      "step": 498
+    },
+    {
+      "epoch": 0.13464651915812195,
+      "grad_norm": 4.997314929962158,
+      "learning_rate": 1.9433252030033824e-05,
+      "loss": 0.3466,
+      "step": 499
+    },
+    {
+      "epoch": 0.13491635186184567,
+      "grad_norm": 0.9487417340278625,
+      "learning_rate": 1.943034748126311e-05,
+      "loss": 0.3255,
+      "step": 500
+    },
+    {
+      "epoch": 0.13518618456556936,
+      "grad_norm": 1.6898459196090698,
+      "learning_rate": 1.942743572687384e-05,
+      "loss": 0.3358,
+      "step": 501
+    },
+    {
+      "epoch": 0.13545601726929304,
+      "grad_norm": 0.8151451349258423,
+      "learning_rate": 1.9424516769090863e-05,
+      "loss": 0.3118,
+      "step": 502
+    },
+    {
+      "epoch": 0.13572584997301673,
+      "grad_norm": 0.9528963565826416,
+      "learning_rate": 1.942159061014452e-05,
+      "loss": 0.3253,
+      "step": 503
+    },
+    {
+      "epoch": 0.13599568267674042,
+      "grad_norm": 0.9173166155815125,
+      "learning_rate": 1.9418657252270644e-05,
+      "loss": 0.2941,
+      "step": 504
+    },
+    {
+      "epoch": 0.1362655153804641,
+      "grad_norm": 2.54341459274292,
+      "learning_rate": 1.9415716697710588e-05,
+      "loss": 0.2871,
+      "step": 505
+    },
+    {
+      "epoch": 0.1365353480841878,
+      "grad_norm": 0.9867715239524841,
+      "learning_rate": 1.94127689487112e-05,
+      "loss": 0.2987,
+      "step": 506
+    },
+    {
+      "epoch": 0.13680518078791148,
+      "grad_norm": 4.665298938751221,
+      "learning_rate": 1.9409814007524807e-05,
+      "loss": 0.3105,
+      "step": 507
+    },
+    {
+      "epoch": 0.13707501349163517,
+      "grad_norm": 1.494543194770813,
+      "learning_rate": 1.9406851876409254e-05,
+      "loss": 0.3092,
+      "step": 508
+    },
+    {
+      "epoch": 0.1373448461953589,
+      "grad_norm": 2.042158603668213,
+      "learning_rate": 1.940388255762787e-05,
+      "loss": 0.3088,
+      "step": 509
+    },
+    {
+      "epoch": 0.13761467889908258,
+      "grad_norm": 3.7526259422302246,
+      "learning_rate": 1.940090605344948e-05,
+      "loss": 0.3175,
+      "step": 510
+    },
+    {
+      "epoch": 0.13788451160280626,
+      "grad_norm": 2.0237467288970947,
+      "learning_rate": 1.939792236614839e-05,
+      "loss": 0.3365,
+      "step": 511
+    },
+    {
+      "epoch": 0.13815434430652995,
+      "grad_norm": 1.7773805856704712,
+      "learning_rate": 1.9394931498004403e-05,
+      "loss": 0.2983,
+      "step": 512
+    },
+    {
+      "epoch": 0.13842417701025364,
+      "grad_norm": 6.7792582511901855,
+      "learning_rate": 1.9391933451302803e-05,
+      "loss": 0.3487,
+      "step": 513
+    },
+    {
+      "epoch": 0.13869400971397733,
+      "grad_norm": 1.2407934665679932,
+      "learning_rate": 1.938892822833437e-05,
+      "loss": 0.3297,
+      "step": 514
+    },
+    {
+      "epoch": 0.13896384241770102,
+      "grad_norm": 6.714599609375,
+      "learning_rate": 1.9385915831395358e-05,
+      "loss": 0.2884,
+      "step": 515
+    },
+    {
+      "epoch": 0.1392336751214247,
+      "grad_norm": 3.378161668777466,
+      "learning_rate": 1.93828962627875e-05,
+      "loss": 0.3254,
+      "step": 516
+    },
+    {
+      "epoch": 0.13950350782514842,
+      "grad_norm": 2.1945202350616455,
+      "learning_rate": 1.9379869524818015e-05,
+      "loss": 0.2773,
+      "step": 517
+    },
+    {
+      "epoch": 0.1397733405288721,
+      "grad_norm": 5.837981700897217,
+      "learning_rate": 1.9376835619799598e-05,
+      "loss": 0.2959,
+      "step": 518
+    },
+    {
+      "epoch": 0.1400431732325958,
+      "grad_norm": 2.479004383087158,
+      "learning_rate": 1.9373794550050424e-05,
+      "loss": 0.2907,
+      "step": 519
+    },
+    {
+      "epoch": 0.14031300593631948,
+      "grad_norm": 1.481318473815918,
+      "learning_rate": 1.9370746317894135e-05,
+      "loss": 0.3131,
+      "step": 520
+    },
+    {
+      "epoch": 0.14058283864004317,
+      "grad_norm": 6.100139141082764,
+      "learning_rate": 1.936769092565985e-05,
+      "loss": 0.3257,
+      "step": 521
+    },
+    {
+      "epoch": 0.14085267134376686,
+      "grad_norm": 1.891465663909912,
+      "learning_rate": 1.936462837568216e-05,
+      "loss": 0.2967,
+      "step": 522
+    },
+    {
+      "epoch": 0.14112250404749055,
+      "grad_norm": 5.506683349609375,
+      "learning_rate": 1.9361558670301125e-05,
+      "loss": 0.3245,
+      "step": 523
+    },
+    {
+      "epoch": 0.14139233675121424,
+      "grad_norm": 2.8681142330169678,
+      "learning_rate": 1.9358481811862265e-05,
+      "loss": 0.2908,
+      "step": 524
+    },
+    {
+      "epoch": 0.14166216945493793,
+      "grad_norm": 1.3085130453109741,
+      "learning_rate": 1.935539780271658e-05,
+      "loss": 0.2881,
+      "step": 525
+    },
+    {
+      "epoch": 0.14193200215866164,
+      "grad_norm": 2.8253307342529297,
+      "learning_rate": 1.9352306645220518e-05,
+      "loss": 0.2929,
+      "step": 526
+    },
+    {
+      "epoch": 0.14220183486238533,
+      "grad_norm": 3.9635913372039795,
+      "learning_rate": 1.9349208341736e-05,
+      "loss": 0.3124,
+      "step": 527
+    },
+    {
+      "epoch": 0.14247166756610902,
+      "grad_norm": 2.6544647216796875,
+      "learning_rate": 1.9346102894630405e-05,
+      "loss": 0.2877,
+      "step": 528
+    },
+    {
+      "epoch": 0.1427415002698327,
+      "grad_norm": 1.152215838432312,
+      "learning_rate": 1.9342990306276567e-05,
+      "loss": 0.3135,
+      "step": 529
+    },
+    {
+      "epoch": 0.1430113329735564,
+      "grad_norm": 0.9675891399383545,
+      "learning_rate": 1.933987057905278e-05,
+      "loss": 0.3233,
+      "step": 530
+    },
+    {
+      "epoch": 0.14328116567728008,
+      "grad_norm": 1.1975467205047607,
+      "learning_rate": 1.9336743715342785e-05,
+      "loss": 0.3375,
+      "step": 531
+    },
+    {
+      "epoch": 0.14355099838100377,
+      "grad_norm": 2.24670147895813,
+      "learning_rate": 1.9333609717535788e-05,
+      "loss": 0.3016,
+      "step": 532
+    },
+    {
+      "epoch": 0.14382083108472746,
+      "grad_norm": 1.0070465803146362,
+      "learning_rate": 1.933046858802644e-05,
+      "loss": 0.3269,
+      "step": 533
+    },
+    {
+      "epoch": 0.14409066378845117,
+      "grad_norm": 1.9072781801223755,
+      "learning_rate": 1.932732032921484e-05,
+      "loss": 0.2988,
+      "step": 534
+    },
+    {
+      "epoch": 0.14436049649217486,
+      "grad_norm": 1.0057188272476196,
+      "learning_rate": 1.932416494350654e-05,
+      "loss": 0.2935,
+      "step": 535
+    },
+    {
+      "epoch": 0.14463032919589855,
+      "grad_norm": 2.1650004386901855,
+      "learning_rate": 1.9321002433312527e-05,
+      "loss": 0.2956,
+      "step": 536
+    },
+    {
+      "epoch": 0.14490016189962224,
+      "grad_norm": 1.5116249322891235,
+      "learning_rate": 1.931783280104924e-05,
+      "loss": 0.2872,
+      "step": 537
+    },
+    {
+      "epoch": 0.14516999460334593,
+      "grad_norm": 1.6434077024459839,
+      "learning_rate": 1.931465604913856e-05,
+      "loss": 0.3133,
+      "step": 538
+    },
+    {
+      "epoch": 0.1454398273070696,
+      "grad_norm": 0.9260255694389343,
+      "learning_rate": 1.9311472180007803e-05,
+      "loss": 0.2752,
+      "step": 539
+    },
+    {
+      "epoch": 0.1457096600107933,
+      "grad_norm": 1.1168843507766724,
+      "learning_rate": 1.930828119608973e-05,
+      "loss": 0.2872,
+      "step": 540
+    },
+    {
+      "epoch": 0.145979492714517,
+      "grad_norm": 1.3670954704284668,
+      "learning_rate": 1.9305083099822534e-05,
+      "loss": 0.331,
+      "step": 541
+    },
+    {
+      "epoch": 0.14624932541824068,
+      "grad_norm": 2.735773801803589,
+      "learning_rate": 1.9301877893649842e-05,
+      "loss": 0.3152,
+      "step": 542
+    },
+    {
+      "epoch": 0.1465191581219644,
+      "grad_norm": 2.7743208408355713,
+      "learning_rate": 1.9298665580020715e-05,
+      "loss": 0.34,
+      "step": 543
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 2.1040079593658447,
+      "learning_rate": 1.9295446161389644e-05,
+      "loss": 0.3161,
+      "step": 544
+    },
+    {
+      "epoch": 0.14705882352941177,
+      "grad_norm": 2.754239559173584,
+      "learning_rate": 1.9292219640216548e-05,
+      "loss": 0.3145,
+      "step": 545
+    },
+    {
+      "epoch": 0.14732865623313546,
+      "grad_norm": 1.29515540599823,
+      "learning_rate": 1.9288986018966778e-05,
+      "loss": 0.3021,
+      "step": 546
+    },
+    {
+      "epoch": 0.14759848893685915,
+      "grad_norm": 1.3120839595794678,
+      "learning_rate": 1.9285745300111102e-05,
+      "loss": 0.3104,
+      "step": 547
+    },
+    {
+      "epoch": 0.14786832164058283,
+      "grad_norm": 1.3073720932006836,
+      "learning_rate": 1.9282497486125716e-05,
+      "loss": 0.3192,
+      "step": 548
+    },
+    {
+      "epoch": 0.14813815434430652,
+      "grad_norm": 1.6458323001861572,
+      "learning_rate": 1.927924257949224e-05,
+      "loss": 0.281,
+      "step": 549
+    },
+    {
+      "epoch": 0.1484079870480302,
+      "grad_norm": 3.4357223510742188,
+      "learning_rate": 1.9275980582697707e-05,
+      "loss": 0.3088,
+      "step": 550
+    },
+    {
+      "epoch": 0.14867781975175393,
+      "grad_norm": 0.7410882711410522,
+      "learning_rate": 1.9272711498234574e-05,
+      "loss": 0.2938,
+      "step": 551
+    },
+    {
+      "epoch": 0.1489476524554776,
+      "grad_norm": 3.7744996547698975,
+      "learning_rate": 1.9269435328600704e-05,
+      "loss": 0.293,
+      "step": 552
+    },
+    {
+      "epoch": 0.1492174851592013,
+      "grad_norm": 2.957030773162842,
+      "learning_rate": 1.926615207629939e-05,
+      "loss": 0.287,
+      "step": 553
+    },
+    {
+      "epoch": 0.149487317862925,
+      "grad_norm": 1.530582070350647,
+      "learning_rate": 1.9262861743839318e-05,
+      "loss": 0.3048,
+      "step": 554
+    },
+    {
+      "epoch": 0.14975715056664868,
+      "grad_norm": 5.090281963348389,
+      "learning_rate": 1.9259564333734596e-05,
+      "loss": 0.3126,
+      "step": 555
+    },
+    {
+      "epoch": 0.15002698327037237,
+      "grad_norm": 1.0566977262496948,
+      "learning_rate": 1.9256259848504737e-05,
+      "loss": 0.2959,
+      "step": 556
+    },
+    {
+      "epoch": 0.15029681597409605,
+      "grad_norm": 4.5524468421936035,
+      "learning_rate": 1.925294829067466e-05,
+      "loss": 0.3405,
+      "step": 557
+    },
+    {
+      "epoch": 0.15056664867781974,
+      "grad_norm": 1.6716094017028809,
+      "learning_rate": 1.9249629662774684e-05,
+      "loss": 0.305,
+      "step": 558
+    },
+    {
+      "epoch": 0.15083648138154343,
+      "grad_norm": 2.78855299949646,
+      "learning_rate": 1.9246303967340537e-05,
+      "loss": 0.3051,
+      "step": 559
+    },
+    {
+      "epoch": 0.15110631408526715,
+      "grad_norm": 4.550906181335449,
+      "learning_rate": 1.9242971206913347e-05,
+      "loss": 0.3133,
+      "step": 560
+    },
+    {
+      "epoch": 0.15137614678899083,
+      "grad_norm": 1.1869055032730103,
+      "learning_rate": 1.9239631384039632e-05,
+      "loss": 0.3022,
+      "step": 561
+    },
+    {
+      "epoch": 0.15164597949271452,
+      "grad_norm": 4.059884071350098,
+      "learning_rate": 1.9236284501271317e-05,
+      "loss": 0.3037,
+      "step": 562
+    },
+    {
+      "epoch": 0.1519158121964382,
+      "grad_norm": 3.412703275680542,
+      "learning_rate": 1.923293056116571e-05,
+      "loss": 0.3303,
+      "step": 563
+    },
+    {
+      "epoch": 0.1521856449001619,
+      "grad_norm": 4.285120010375977,
+      "learning_rate": 1.9229569566285518e-05,
+      "loss": 0.278,
+      "step": 564
+    },
+    {
+      "epoch": 0.1524554776038856,
+      "grad_norm": 2.9242851734161377,
+      "learning_rate": 1.9226201519198845e-05,
+      "loss": 0.3122,
+      "step": 565
+    },
+    {
+      "epoch": 0.15272531030760927,
+      "grad_norm": 1.8546960353851318,
+      "learning_rate": 1.922282642247917e-05,
+      "loss": 0.3016,
+      "step": 566
+    },
+    {
+      "epoch": 0.15299514301133296,
+      "grad_norm": 3.218536853790283,
+      "learning_rate": 1.9219444278705364e-05,
+      "loss": 0.3212,
+      "step": 567
+    },
+    {
+      "epoch": 0.15326497571505668,
+      "grad_norm": 1.0564050674438477,
+      "learning_rate": 1.9216055090461693e-05,
+      "loss": 0.3045,
+      "step": 568
+    },
+    {
+      "epoch": 0.15353480841878037,
+      "grad_norm": 1.2863322496414185,
+      "learning_rate": 1.9212658860337787e-05,
+      "loss": 0.3363,
+      "step": 569
+    },
+    {
+      "epoch": 0.15380464112250405,
+      "grad_norm": 5.057046413421631,
+      "learning_rate": 1.920925559092867e-05,
+      "loss": 0.3087,
+      "step": 570
+    },
+    {
+      "epoch": 0.15407447382622774,
+      "grad_norm": 1.3052369356155396,
+      "learning_rate": 1.9205845284834735e-05,
+      "loss": 0.2966,
+      "step": 571
+    },
+    {
+      "epoch": 0.15434430652995143,
+      "grad_norm": 3.2919483184814453,
+      "learning_rate": 1.9202427944661766e-05,
+      "loss": 0.3243,
+      "step": 572
+    },
+    {
+      "epoch": 0.15461413923367512,
+      "grad_norm": 2.863386631011963,
+      "learning_rate": 1.9199003573020904e-05,
+      "loss": 0.2934,
+      "step": 573
+    },
+    {
+      "epoch": 0.1548839719373988,
+      "grad_norm": 1.330075740814209,
+      "learning_rate": 1.9195572172528678e-05,
+      "loss": 0.3087,
+      "step": 574
+    },
+    {
+      "epoch": 0.1551538046411225,
+      "grad_norm": 4.211341857910156,
+      "learning_rate": 1.9192133745806976e-05,
+      "loss": 0.3034,
+      "step": 575
+    },
+    {
+      "epoch": 0.15542363734484618,
+      "grad_norm": 1.3641526699066162,
+      "learning_rate": 1.9188688295483068e-05,
+      "loss": 0.3478,
+      "step": 576
+    },
+    {
+      "epoch": 0.1556934700485699,
+      "grad_norm": 1.1217687129974365,
+      "learning_rate": 1.9185235824189573e-05,
+      "loss": 0.296,
+      "step": 577
+    },
+    {
+      "epoch": 0.1559633027522936,
+      "grad_norm": 1.030380129814148,
+      "learning_rate": 1.918177633456449e-05,
+      "loss": 0.2627,
+      "step": 578
+    },
+    {
+      "epoch": 0.15623313545601727,
+      "grad_norm": 1.3523330688476562,
+      "learning_rate": 1.9178309829251168e-05,
+      "loss": 0.2788,
+      "step": 579
+    },
+    {
+      "epoch": 0.15650296815974096,
+      "grad_norm": 2.812425136566162,
+      "learning_rate": 1.9174836310898334e-05,
+      "loss": 0.3186,
+      "step": 580
+    },
+    {
+      "epoch": 0.15677280086346465,
+      "grad_norm": 0.9522241950035095,
+      "learning_rate": 1.917135578216005e-05,
+      "loss": 0.2915,
+      "step": 581
+    },
+    {
+      "epoch": 0.15704263356718834,
+      "grad_norm": 1.2210774421691895,
+      "learning_rate": 1.916786824569576e-05,
+      "loss": 0.2959,
+      "step": 582
+    },
+    {
+      "epoch": 0.15731246627091203,
+      "grad_norm": 4.049483299255371,
+      "learning_rate": 1.916437370417024e-05,
+      "loss": 0.3022,
+      "step": 583
+    },
+    {
+      "epoch": 0.15758229897463572,
+      "grad_norm": 1.730484127998352,
+      "learning_rate": 1.9160872160253638e-05,
+      "loss": 0.3355,
+      "step": 584
+    },
+    {
+      "epoch": 0.15785213167835943,
+      "grad_norm": 2.0946056842803955,
+      "learning_rate": 1.915736361662143e-05,
+      "loss": 0.3038,
+      "step": 585
+    },
+    {
+      "epoch": 0.15812196438208312,
+      "grad_norm": 4.3383965492248535,
+      "learning_rate": 1.9153848075954465e-05,
+      "loss": 0.3225,
+      "step": 586
+    },
+    {
+      "epoch": 0.1583917970858068,
+      "grad_norm": 1.5763877630233765,
+      "learning_rate": 1.9150325540938923e-05,
+      "loss": 0.2857,
+      "step": 587
+    },
+    {
+      "epoch": 0.1586616297895305,
+      "grad_norm": 6.045645236968994,
+      "learning_rate": 1.9146796014266332e-05,
+      "loss": 0.2892,
+      "step": 588
+    },
+    {
+      "epoch": 0.15893146249325418,
+      "grad_norm": 1.2092574834823608,
+      "learning_rate": 1.914325949863356e-05,
+      "loss": 0.3183,
+      "step": 589
+    },
+    {
+      "epoch": 0.15920129519697787,
+      "grad_norm": 4.499578475952148,
+      "learning_rate": 1.9139715996742816e-05,
+      "loss": 0.3459,
+      "step": 590
+    },
+    {
+      "epoch": 0.15947112790070156,
+      "grad_norm": 1.3380200862884521,
+      "learning_rate": 1.9136165511301653e-05,
+      "loss": 0.3042,
+      "step": 591
+    },
+    {
+      "epoch": 0.15974096060442525,
+      "grad_norm": 2.238821268081665,
+      "learning_rate": 1.9132608045022954e-05,
+      "loss": 0.2708,
+      "step": 592
+    },
+    {
+      "epoch": 0.16001079330814894,
+      "grad_norm": 1.2946232557296753,
+      "learning_rate": 1.912904360062494e-05,
+      "loss": 0.301,
+      "step": 593
+    },
+    {
+      "epoch": 0.16028062601187265,
+      "grad_norm": 0.9654742479324341,
+      "learning_rate": 1.9125472180831155e-05,
+      "loss": 0.2792,
+      "step": 594
+    },
+    {
+      "epoch": 0.16055045871559634,
+      "grad_norm": 1.0027459859848022,
+      "learning_rate": 1.9121893788370486e-05,
+      "loss": 0.279,
+      "step": 595
+    },
+    {
+      "epoch": 0.16082029141932003,
+      "grad_norm": 1.293976068496704,
+      "learning_rate": 1.9118308425977133e-05,
+      "loss": 0.3311,
+      "step": 596
+    },
+    {
+      "epoch": 0.16109012412304372,
+      "grad_norm": 1.8101601600646973,
+      "learning_rate": 1.911471609639064e-05,
+      "loss": 0.3071,
+      "step": 597
+    },
+    {
+      "epoch": 0.1613599568267674,
+      "grad_norm": 1.5578137636184692,
+      "learning_rate": 1.9111116802355853e-05,
+      "loss": 0.2933,
+      "step": 598
+    },
+    {
+      "epoch": 0.1616297895304911,
+      "grad_norm": 1.7257205247879028,
+      "learning_rate": 1.9107510546622962e-05,
+      "loss": 0.3146,
+      "step": 599
+    },
+    {
+      "epoch": 0.16189962223421478,
+      "grad_norm": 0.9653056859970093,
+      "learning_rate": 1.9103897331947455e-05,
+      "loss": 0.2881,
+      "step": 600
+    },
+    {
+      "epoch": 0.16216945493793847,
+      "grad_norm": 2.3949053287506104,
+      "learning_rate": 1.9100277161090153e-05,
+      "loss": 0.327,
+      "step": 601
+    },
+    {
+      "epoch": 0.16243928764166216,
+      "grad_norm": 2.5193660259246826,
+      "learning_rate": 1.9096650036817184e-05,
+      "loss": 0.2901,
+      "step": 602
+    },
+    {
+      "epoch": 0.16270912034538587,
+      "grad_norm": 1.1664611101150513,
+      "learning_rate": 1.9093015961899996e-05,
+      "loss": 0.3267,
+      "step": 603
+    },
+    {
+      "epoch": 0.16297895304910956,
+      "grad_norm": 3.726238489151001,
+      "learning_rate": 1.9089374939115335e-05,
+      "loss": 0.3284,
+      "step": 604
+    },
+    {
+      "epoch": 0.16324878575283325,
+      "grad_norm": 4.697918891906738,
+      "learning_rate": 1.9085726971245267e-05,
+      "loss": 0.2762,
+      "step": 605
+    },
+    {
+      "epoch": 0.16351861845655694,
+      "grad_norm": 3.893923044204712,
+      "learning_rate": 1.908207206107717e-05,
+      "loss": 0.3233,
+      "step": 606
+    },
+    {
+      "epoch": 0.16378845116028062,
+      "grad_norm": 1.9809579849243164,
+      "learning_rate": 1.9078410211403704e-05,
+      "loss": 0.3086,
+      "step": 607
+    },
+    {
+      "epoch": 0.1640582838640043,
+      "grad_norm": 1.5440902709960938,
+      "learning_rate": 1.9074741425022855e-05,
+      "loss": 0.3144,
+      "step": 608
+    },
+    {
+      "epoch": 0.164328116567728,
+      "grad_norm": 2.904125452041626,
+      "learning_rate": 1.90710657047379e-05,
+      "loss": 0.3192,
+      "step": 609
+    },
+    {
+      "epoch": 0.1645979492714517,
+      "grad_norm": 2.7370400428771973,
+      "learning_rate": 1.906738305335741e-05,
+      "loss": 0.3092,
+      "step": 610
+    },
+    {
+      "epoch": 0.1648677819751754,
+      "grad_norm": 0.725981593132019,
+      "learning_rate": 1.906369347369526e-05,
+      "loss": 0.283,
+      "step": 611
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 2.591895341873169,
+      "learning_rate": 1.905999696857061e-05,
+      "loss": 0.3121,
+      "step": 612
+    },
+    {
+      "epoch": 0.16540744738262278,
+      "grad_norm": 1.7620854377746582,
+      "learning_rate": 1.905629354080792e-05,
+      "loss": 0.29,
+      "step": 613
+    },
+    {
+      "epoch": 0.16567728008634647,
+      "grad_norm": 3.460005283355713,
+      "learning_rate": 1.9052583193236938e-05,
+      "loss": 0.3266,
+      "step": 614
+    },
+    {
+      "epoch": 0.16594711279007016,
+      "grad_norm": 7.2651543617248535,
+      "learning_rate": 1.904886592869269e-05,
+      "loss": 0.3314,
+      "step": 615
+    },
+    {
+      "epoch": 0.16621694549379384,
+      "grad_norm": 2.3988196849823,
+      "learning_rate": 1.90451417500155e-05,
+      "loss": 0.3107,
+      "step": 616
+    },
+    {
+      "epoch": 0.16648677819751753,
+      "grad_norm": 1.0958611965179443,
+      "learning_rate": 1.904141066005097e-05,
+      "loss": 0.2895,
+      "step": 617
+    },
+    {
+      "epoch": 0.16675661090124122,
+      "grad_norm": 2.601902723312378,
+      "learning_rate": 1.903767266164998e-05,
+      "loss": 0.2858,
+      "step": 618
+    },
+    {
+      "epoch": 0.1670264436049649,
+      "grad_norm": 0.97732013463974,
+      "learning_rate": 1.9033927757668692e-05,
+      "loss": 0.3114,
+      "step": 619
+    },
+    {
+      "epoch": 0.16729627630868862,
+      "grad_norm": 4.074997425079346,
+      "learning_rate": 1.9030175950968543e-05,
+      "loss": 0.3046,
+      "step": 620
+    },
+    {
+      "epoch": 0.1675661090124123,
+      "grad_norm": 1.4154969453811646,
+      "learning_rate": 1.9026417244416247e-05,
+      "loss": 0.3536,
+      "step": 621
+    },
+    {
+      "epoch": 0.167835941716136,
+      "grad_norm": 1.7893296480178833,
+      "learning_rate": 1.902265164088378e-05,
+      "loss": 0.294,
+      "step": 622
+    },
+    {
+      "epoch": 0.1681057744198597,
+      "grad_norm": 1.5126819610595703,
+      "learning_rate": 1.9018879143248407e-05,
+      "loss": 0.2845,
+      "step": 623
+    },
+    {
+      "epoch": 0.16837560712358338,
+      "grad_norm": 1.250545620918274,
+      "learning_rate": 1.901509975439264e-05,
+      "loss": 0.3015,
+      "step": 624
+    },
+    {
+      "epoch": 0.16864543982730706,
+      "grad_norm": 1.2928470373153687,
+      "learning_rate": 1.9011313477204276e-05,
+      "loss": 0.2876,
+      "step": 625
+    },
+    {
+      "epoch": 0.16891527253103075,
+      "grad_norm": 1.2828264236450195,
+      "learning_rate": 1.9007520314576356e-05,
+      "loss": 0.2635,
+      "step": 626
+    },
+    {
+      "epoch": 0.16918510523475444,
+      "grad_norm": 1.2343151569366455,
+      "learning_rate": 1.900372026940719e-05,
+      "loss": 0.2961,
+      "step": 627
+    },
+    {
+      "epoch": 0.16945493793847816,
+      "grad_norm": 1.5760513544082642,
+      "learning_rate": 1.899991334460036e-05,
+      "loss": 0.326,
+      "step": 628
+    },
+    {
+      "epoch": 0.16972477064220184,
+      "grad_norm": 3.338301420211792,
+      "learning_rate": 1.899609954306468e-05,
+      "loss": 0.2813,
+      "step": 629
+    },
+    {
+      "epoch": 0.16999460334592553,
+      "grad_norm": 1.0559368133544922,
+      "learning_rate": 1.899227886771424e-05,
+      "loss": 0.3193,
+      "step": 630
+    },
+    {
+      "epoch": 0.17026443604964922,
+      "grad_norm": 0.9462928175926208,
+      "learning_rate": 1.8988451321468372e-05,
+      "loss": 0.261,
+      "step": 631
+    },
+    {
+      "epoch": 0.1705342687533729,
+      "grad_norm": 0.9721528887748718,
+      "learning_rate": 1.8984616907251655e-05,
+      "loss": 0.3037,
+      "step": 632
+    },
+    {
+      "epoch": 0.1708041014570966,
+      "grad_norm": 1.9441041946411133,
+      "learning_rate": 1.8980775627993925e-05,
+      "loss": 0.2784,
+      "step": 633
+    },
+    {
+      "epoch": 0.17107393416082028,
+      "grad_norm": 2.730639934539795,
+      "learning_rate": 1.8976927486630252e-05,
+      "loss": 0.2766,
+      "step": 634
+    },
+    {
+      "epoch": 0.17134376686454397,
+      "grad_norm": 2.2268755435943604,
+      "learning_rate": 1.8973072486100963e-05,
+      "loss": 0.3082,
+      "step": 635
+    },
+    {
+      "epoch": 0.17161359956826766,
+      "grad_norm": 1.545017957687378,
+      "learning_rate": 1.896921062935162e-05,
+      "loss": 0.2911,
+      "step": 636
+    },
+    {
+      "epoch": 0.17188343227199138,
+      "grad_norm": 2.25679874420166,
+      "learning_rate": 1.8965341919333015e-05,
+      "loss": 0.3306,
+      "step": 637
+    },
+    {
+      "epoch": 0.17215326497571506,
+      "grad_norm": 1.2802419662475586,
+      "learning_rate": 1.896146635900119e-05,
+      "loss": 0.3493,
+      "step": 638
+    },
+    {
+      "epoch": 0.17242309767943875,
+      "grad_norm": 1.1736631393432617,
+      "learning_rate": 1.8957583951317414e-05,
+      "loss": 0.3051,
+      "step": 639
+    },
+    {
+      "epoch": 0.17269293038316244,
+      "grad_norm": 0.9208096265792847,
+      "learning_rate": 1.8953694699248193e-05,
+      "loss": 0.2853,
+      "step": 640
+    },
+    {
+      "epoch": 0.17296276308688613,
+      "grad_norm": 1.6063733100891113,
+      "learning_rate": 1.8949798605765252e-05,
+      "loss": 0.3045,
+      "step": 641
+    },
+    {
+      "epoch": 0.17323259579060982,
+      "grad_norm": 1.590314269065857,
+      "learning_rate": 1.8945895673845552e-05,
+      "loss": 0.3256,
+      "step": 642
+    },
+    {
+      "epoch": 0.1735024284943335,
+      "grad_norm": 3.4414167404174805,
+      "learning_rate": 1.8941985906471284e-05,
+      "loss": 0.3006,
+      "step": 643
+    },
+    {
+      "epoch": 0.1737722611980572,
+      "grad_norm": 2.230807304382324,
+      "learning_rate": 1.8938069306629853e-05,
+      "loss": 0.2906,
+      "step": 644
+    },
+    {
+      "epoch": 0.1740420939017809,
+      "grad_norm": 2.3978118896484375,
+      "learning_rate": 1.8934145877313882e-05,
+      "loss": 0.3061,
+      "step": 645
+    },
+    {
+      "epoch": 0.1743119266055046,
+      "grad_norm": 2.528841257095337,
+      "learning_rate": 1.893021562152122e-05,
+      "loss": 0.293,
+      "step": 646
+    },
+    {
+      "epoch": 0.17458175930922828,
+      "grad_norm": 1.1000761985778809,
+      "learning_rate": 1.8926278542254936e-05,
+      "loss": 0.2763,
+      "step": 647
+    },
+    {
+      "epoch": 0.17485159201295197,
+      "grad_norm": 0.9161582589149475,
+      "learning_rate": 1.89223346425233e-05,
+      "loss": 0.2909,
+      "step": 648
+    },
+    {
+      "epoch": 0.17512142471667566,
+      "grad_norm": 2.603771448135376,
+      "learning_rate": 1.89183839253398e-05,
+      "loss": 0.267,
+      "step": 649
+    },
+    {
+      "epoch": 0.17539125742039935,
+      "grad_norm": 2.0195446014404297,
+      "learning_rate": 1.8914426393723134e-05,
+      "loss": 0.3014,
+      "step": 650
+    },
+    {
+      "epoch": 0.17566109012412304,
+      "grad_norm": 1.293984055519104,
+      "learning_rate": 1.8910462050697206e-05,
+      "loss": 0.2612,
+      "step": 651
+    },
+    {
+      "epoch": 0.17593092282784673,
+      "grad_norm": 1.3261752128601074,
+      "learning_rate": 1.8906490899291125e-05,
+      "loss": 0.3075,
+      "step": 652
+    },
+    {
+      "epoch": 0.1762007555315704,
+      "grad_norm": 1.176215410232544,
+      "learning_rate": 1.8902512942539202e-05,
+      "loss": 0.275,
+      "step": 653
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 1.445433259010315,
+      "learning_rate": 1.8898528183480945e-05,
+      "loss": 0.3127,
+      "step": 654
+    },
+    {
+      "epoch": 0.17674042093901782,
+      "grad_norm": 0.990032434463501,
+      "learning_rate": 1.889453662516106e-05,
+      "loss": 0.2827,
+      "step": 655
+    },
+    {
+      "epoch": 0.1770102536427415,
+      "grad_norm": 2.4735045433044434,
+      "learning_rate": 1.8890538270629458e-05,
+      "loss": 0.2961,
+      "step": 656
+    },
+    {
+      "epoch": 0.1772800863464652,
+      "grad_norm": 2.3039443492889404,
+      "learning_rate": 1.8886533122941228e-05,
+      "loss": 0.312,
+      "step": 657
+    },
+    {
+      "epoch": 0.17754991905018888,
+      "grad_norm": 0.9664218425750732,
+      "learning_rate": 1.888252118515666e-05,
+      "loss": 0.296,
+      "step": 658
+    },
+    {
+      "epoch": 0.17781975175391257,
+      "grad_norm": 5.479343414306641,
+      "learning_rate": 1.887850246034123e-05,
+      "loss": 0.2861,
+      "step": 659
+    },
+    {
+      "epoch": 0.17808958445763626,
+      "grad_norm": 0.8309643268585205,
+      "learning_rate": 1.887447695156559e-05,
+      "loss": 0.2883,
+      "step": 660
+    },
+    {
+      "epoch": 0.17835941716135995,
+      "grad_norm": 4.44924259185791,
+      "learning_rate": 1.8870444661905598e-05,
+      "loss": 0.3141,
+      "step": 661
+    },
+    {
+      "epoch": 0.17862924986508366,
+      "grad_norm": 2.9890189170837402,
+      "learning_rate": 1.8866405594442264e-05,
+      "loss": 0.2586,
+      "step": 662
+    },
+    {
+      "epoch": 0.17889908256880735,
+      "grad_norm": 2.3052284717559814,
+      "learning_rate": 1.886235975226181e-05,
+      "loss": 0.32,
+      "step": 663
+    },
+    {
+      "epoch": 0.17916891527253104,
+      "grad_norm": 3.493809700012207,
+      "learning_rate": 1.88583071384556e-05,
+      "loss": 0.2827,
+      "step": 664
+    },
+    {
+      "epoch": 0.17943874797625473,
+      "grad_norm": 0.878583550453186,
+      "learning_rate": 1.88542477561202e-05,
+      "loss": 0.2837,
+      "step": 665
+    },
+    {
+      "epoch": 0.1797085806799784,
+      "grad_norm": 1.6771315336227417,
+      "learning_rate": 1.8850181608357334e-05,
+      "loss": 0.3255,
+      "step": 666
+    },
+    {
+      "epoch": 0.1799784133837021,
+      "grad_norm": 1.877223253250122,
+      "learning_rate": 1.8846108698273898e-05,
+      "loss": 0.2809,
+      "step": 667
+    },
+    {
+      "epoch": 0.1802482460874258,
+      "grad_norm": 1.1803197860717773,
+      "learning_rate": 1.8842029028981953e-05,
+      "loss": 0.3075,
+      "step": 668
+    },
+    {
+      "epoch": 0.18051807879114948,
+      "grad_norm": 4.53121280670166,
+      "learning_rate": 1.8837942603598727e-05,
+      "loss": 0.3035,
+      "step": 669
+    },
+    {
+      "epoch": 0.18078791149487317,
+      "grad_norm": 1.7798503637313843,
+      "learning_rate": 1.883384942524661e-05,
+      "loss": 0.2803,
+      "step": 670
+    },
+    {
+      "epoch": 0.18105774419859688,
+      "grad_norm": 4.019486427307129,
+      "learning_rate": 1.8829749497053158e-05,
+      "loss": 0.3292,
+      "step": 671
+    },
+    {
+      "epoch": 0.18132757690232057,
+      "grad_norm": 4.033787727355957,
+      "learning_rate": 1.882564282215107e-05,
+      "loss": 0.302,
+      "step": 672
+    },
+    {
+      "epoch": 0.18159740960604426,
+      "grad_norm": 1.0440713167190552,
+      "learning_rate": 1.8821529403678212e-05,
+      "loss": 0.3114,
+      "step": 673
+    },
+    {
+      "epoch": 0.18186724230976795,
+      "grad_norm": 4.608468532562256,
+      "learning_rate": 1.88174092447776e-05,
+      "loss": 0.301,
+      "step": 674
+    },
+    {
+      "epoch": 0.18213707501349163,
+      "grad_norm": 2.9315829277038574,
+      "learning_rate": 1.8813282348597398e-05,
+      "loss": 0.2783,
+      "step": 675
+    },
+    {
+      "epoch": 0.18240690771721532,
+      "grad_norm": 4.216452121734619,
+      "learning_rate": 1.880914871829092e-05,
+      "loss": 0.3324,
+      "step": 676
+    },
+    {
+      "epoch": 0.182676740420939,
+      "grad_norm": 3.650352954864502,
+      "learning_rate": 1.8805008357016624e-05,
+      "loss": 0.2948,
+      "step": 677
+    },
+    {
+      "epoch": 0.1829465731246627,
+      "grad_norm": 2.1062357425689697,
+      "learning_rate": 1.880086126793811e-05,
+      "loss": 0.2763,
+      "step": 678
+    },
+    {
+      "epoch": 0.1832164058283864,
+      "grad_norm": 5.303624153137207,
+      "learning_rate": 1.8796707454224123e-05,
+      "loss": 0.2989,
+      "step": 679
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 1.723644733428955,
+      "learning_rate": 1.8792546919048545e-05,
+      "loss": 0.3193,
+      "step": 680
+    },
+    {
+      "epoch": 0.1837560712358338,
+      "grad_norm": 3.3395888805389404,
+      "learning_rate": 1.8788379665590386e-05,
+      "loss": 0.284,
+      "step": 681
+    },
+    {
+      "epoch": 0.18402590393955748,
+      "grad_norm": 1.1692960262298584,
+      "learning_rate": 1.8784205697033803e-05,
+      "loss": 0.3135,
+      "step": 682
+    },
+    {
+      "epoch": 0.18429573664328117,
+      "grad_norm": 1.4019652605056763,
+      "learning_rate": 1.8780025016568074e-05,
+      "loss": 0.2647,
+      "step": 683
+    },
+    {
+      "epoch": 0.18456556934700485,
+      "grad_norm": 3.875499725341797,
+      "learning_rate": 1.877583762738761e-05,
+      "loss": 0.3027,
+      "step": 684
+    },
+    {
+      "epoch": 0.18483540205072854,
+      "grad_norm": 1.4832490682601929,
+      "learning_rate": 1.8771643532691942e-05,
+      "loss": 0.2949,
+      "step": 685
+    },
+    {
+      "epoch": 0.18510523475445223,
+      "grad_norm": 2.513218879699707,
+      "learning_rate": 1.8767442735685734e-05,
+      "loss": 0.3005,
+      "step": 686
+    },
+    {
+      "epoch": 0.18537506745817592,
+      "grad_norm": 2.952930212020874,
+      "learning_rate": 1.8763235239578755e-05,
+      "loss": 0.3238,
+      "step": 687
+    },
+    {
+      "epoch": 0.18564490016189963,
+      "grad_norm": 2.5589730739593506,
+      "learning_rate": 1.875902104758592e-05,
+      "loss": 0.2823,
+      "step": 688
+    },
+    {
+      "epoch": 0.18591473286562332,
+      "grad_norm": 4.821250915527344,
+      "learning_rate": 1.8754800162927233e-05,
+      "loss": 0.325,
+      "step": 689
+    },
+    {
+      "epoch": 0.186184565569347,
+      "grad_norm": 1.751145362854004,
+      "learning_rate": 1.8750572588827828e-05,
+      "loss": 0.3103,
+      "step": 690
+    },
+    {
+      "epoch": 0.1864543982730707,
+      "grad_norm": 1.9546347856521606,
+      "learning_rate": 1.8746338328517942e-05,
+      "loss": 0.2927,
+      "step": 691
+    },
+    {
+      "epoch": 0.1867242309767944,
+      "grad_norm": 1.0354565382003784,
+      "learning_rate": 1.8742097385232922e-05,
+      "loss": 0.2754,
+      "step": 692
+    },
+    {
+      "epoch": 0.18699406368051807,
+      "grad_norm": 1.5183359384536743,
+      "learning_rate": 1.8737849762213226e-05,
+      "loss": 0.3163,
+      "step": 693
+    },
+    {
+      "epoch": 0.18726389638424176,
+      "grad_norm": 2.0842161178588867,
+      "learning_rate": 1.873359546270442e-05,
+      "loss": 0.3259,
+      "step": 694
+    },
+    {
+      "epoch": 0.18753372908796545,
+      "grad_norm": 3.3596270084381104,
+      "learning_rate": 1.872933448995715e-05,
+      "loss": 0.3195,
+      "step": 695
+    },
+    {
+      "epoch": 0.18780356179168917,
+      "grad_norm": 3.520615577697754,
+      "learning_rate": 1.8725066847227184e-05,
+      "loss": 0.3154,
+      "step": 696
+    },
+    {
+      "epoch": 0.18807339449541285,
+      "grad_norm": 1.3651080131530762,
+      "learning_rate": 1.872079253777538e-05,
+      "loss": 0.2873,
+      "step": 697
+    },
+    {
+      "epoch": 0.18834322719913654,
+      "grad_norm": 3.021942615509033,
+      "learning_rate": 1.8716511564867685e-05,
+      "loss": 0.2921,
+      "step": 698
+    },
+    {
+      "epoch": 0.18861305990286023,
+      "grad_norm": 1.0197726488113403,
+      "learning_rate": 1.871222393177514e-05,
+      "loss": 0.3349,
+      "step": 699
+    },
+    {
+      "epoch": 0.18888289260658392,
+      "grad_norm": 2.443998336791992,
+      "learning_rate": 1.8707929641773876e-05,
+      "loss": 0.3069,
+      "step": 700
+    },
+    {
+      "epoch": 0.1891527253103076,
+      "grad_norm": 3.488307237625122,
+      "learning_rate": 1.8703628698145112e-05,
+      "loss": 0.3196,
+      "step": 701
+    },
+    {
+      "epoch": 0.1894225580140313,
+      "grad_norm": 4.2047624588012695,
+      "learning_rate": 1.869932110417515e-05,
+      "loss": 0.2962,
+      "step": 702
+    },
+    {
+      "epoch": 0.18969239071775498,
+      "grad_norm": 3.586852550506592,
+      "learning_rate": 1.8695006863155365e-05,
+      "loss": 0.2711,
+      "step": 703
+    },
+    {
+      "epoch": 0.18996222342147867,
+      "grad_norm": 3.138190984725952,
+      "learning_rate": 1.869068597838222e-05,
+      "loss": 0.2963,
+      "step": 704
+    },
+    {
+      "epoch": 0.1902320561252024,
+      "grad_norm": 1.1812266111373901,
+      "learning_rate": 1.8686358453157263e-05,
+      "loss": 0.301,
+      "step": 705
+    },
+    {
+      "epoch": 0.19050188882892607,
+      "grad_norm": 2.5653138160705566,
+      "learning_rate": 1.8682024290787092e-05,
+      "loss": 0.2803,
+      "step": 706
+    },
+    {
+      "epoch": 0.19077172153264976,
+      "grad_norm": 1.805107831954956,
+      "learning_rate": 1.8677683494583398e-05,
+      "loss": 0.316,
+      "step": 707
+    },
+    {
+      "epoch": 0.19104155423637345,
+      "grad_norm": 0.805758535861969,
+      "learning_rate": 1.8673336067862933e-05,
+      "loss": 0.2696,
+      "step": 708
+    },
+    {
+      "epoch": 0.19131138694009714,
+      "grad_norm": 1.1082613468170166,
+      "learning_rate": 1.866898201394751e-05,
+      "loss": 0.2565,
+      "step": 709
+    },
+    {
+      "epoch": 0.19158121964382083,
+      "grad_norm": 3.7487876415252686,
+      "learning_rate": 1.8664621336164014e-05,
+      "loss": 0.2949,
+      "step": 710
+    },
+    {
+      "epoch": 0.19185105234754452,
+      "grad_norm": 1.5922000408172607,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.3231,
+      "step": 711
+    },
+    {
+      "epoch": 0.1921208850512682,
+      "grad_norm": 4.760123252868652,
+      "learning_rate": 1.8655880122325633e-05,
+      "loss": 0.2937,
+      "step": 712
+    },
+    {
+      "epoch": 0.1923907177549919,
+      "grad_norm": 2.6837947368621826,
+      "learning_rate": 1.8651499592949805e-05,
+      "loss": 0.2555,
+      "step": 713
+    },
+    {
+      "epoch": 0.1926605504587156,
+      "grad_norm": 1.957506775856018,
+      "learning_rate": 1.864711245306402e-05,
+      "loss": 0.2878,
+      "step": 714
+    },
+    {
+      "epoch": 0.1929303831624393,
+      "grad_norm": 3.272042989730835,
+      "learning_rate": 1.8642718706020443e-05,
+      "loss": 0.3215,
+      "step": 715
+    },
+    {
+      "epoch": 0.19320021586616298,
+      "grad_norm": 1.5925581455230713,
+      "learning_rate": 1.8638318355176273e-05,
+      "loss": 0.3374,
+      "step": 716
+    },
+    {
+      "epoch": 0.19347004856988667,
+      "grad_norm": 1.1061913967132568,
+      "learning_rate": 1.8633911403893778e-05,
+      "loss": 0.2924,
+      "step": 717
+    },
+    {
+      "epoch": 0.19373988127361036,
+      "grad_norm": 1.4847605228424072,
+      "learning_rate": 1.862949785554025e-05,
+      "loss": 0.3092,
+      "step": 718
+    },
+    {
+      "epoch": 0.19400971397733405,
+      "grad_norm": 2.826857566833496,
+      "learning_rate": 1.8625077713488035e-05,
+      "loss": 0.3134,
+      "step": 719
+    },
+    {
+      "epoch": 0.19427954668105774,
+      "grad_norm": 2.4468297958374023,
+      "learning_rate": 1.8620650981114508e-05,
+      "loss": 0.3293,
+      "step": 720
+    },
+    {
+      "epoch": 0.19454937938478142,
+      "grad_norm": 1.8880223035812378,
+      "learning_rate": 1.861621766180209e-05,
+      "loss": 0.2786,
+      "step": 721
+    },
+    {
+      "epoch": 0.19481921208850514,
+      "grad_norm": 2.9317545890808105,
+      "learning_rate": 1.8611777758938225e-05,
+      "loss": 0.3118,
+      "step": 722
+    },
+    {
+      "epoch": 0.19508904479222883,
+      "grad_norm": 1.3060197830200195,
+      "learning_rate": 1.860733127591539e-05,
+      "loss": 0.2816,
+      "step": 723
+    },
+    {
+      "epoch": 0.19535887749595252,
+      "grad_norm": 1.1916656494140625,
+      "learning_rate": 1.8602878216131093e-05,
+      "loss": 0.3176,
+      "step": 724
+    },
+    {
+      "epoch": 0.1956287101996762,
+      "grad_norm": 4.151193618774414,
+      "learning_rate": 1.8598418582987862e-05,
+      "loss": 0.3178,
+      "step": 725
+    },
+    {
+      "epoch": 0.1958985429033999,
+      "grad_norm": 1.6275237798690796,
+      "learning_rate": 1.859395237989325e-05,
+      "loss": 0.3278,
+      "step": 726
+    },
+    {
+      "epoch": 0.19616837560712358,
+      "grad_norm": 3.4714293479919434,
+      "learning_rate": 1.8589479610259837e-05,
+      "loss": 0.2962,
+      "step": 727
+    },
+    {
+      "epoch": 0.19643820831084727,
+      "grad_norm": 1.0343096256256104,
+      "learning_rate": 1.858500027750521e-05,
+      "loss": 0.2853,
+      "step": 728
+    },
+    {
+      "epoch": 0.19670804101457096,
+      "grad_norm": 1.1632835865020752,
+      "learning_rate": 1.8580514385051973e-05,
+      "loss": 0.3234,
+      "step": 729
+    },
+    {
+      "epoch": 0.19697787371829464,
+      "grad_norm": 2.4407503604888916,
+      "learning_rate": 1.8576021936327747e-05,
+      "loss": 0.2939,
+      "step": 730
+    },
+    {
+      "epoch": 0.19724770642201836,
+      "grad_norm": 1.5555877685546875,
+      "learning_rate": 1.8571522934765157e-05,
+      "loss": 0.2899,
+      "step": 731
+    },
+    {
+      "epoch": 0.19751753912574205,
+      "grad_norm": 1.3930840492248535,
+      "learning_rate": 1.8567017383801843e-05,
+      "loss": 0.3041,
+      "step": 732
+    },
+    {
+      "epoch": 0.19778737182946574,
+      "grad_norm": 2.6021604537963867,
+      "learning_rate": 1.8562505286880438e-05,
+      "loss": 0.294,
+      "step": 733
+    },
+    {
+      "epoch": 0.19805720453318942,
+      "grad_norm": 1.2145638465881348,
+      "learning_rate": 1.8557986647448586e-05,
+      "loss": 0.2705,
+      "step": 734
+    },
+    {
+      "epoch": 0.1983270372369131,
+      "grad_norm": 1.5432404279708862,
+      "learning_rate": 1.855346146895893e-05,
+      "loss": 0.3218,
+      "step": 735
+    },
+    {
+      "epoch": 0.1985968699406368,
+      "grad_norm": 6.226170539855957,
+      "learning_rate": 1.8548929754869095e-05,
+      "loss": 0.3181,
+      "step": 736
+    },
+    {
+      "epoch": 0.1988667026443605,
+      "grad_norm": 0.8017795085906982,
+      "learning_rate": 1.854439150864172e-05,
+      "loss": 0.2827,
+      "step": 737
+    },
+    {
+      "epoch": 0.19913653534808418,
+      "grad_norm": 3.283099889755249,
+      "learning_rate": 1.8539846733744426e-05,
+      "loss": 0.2984,
+      "step": 738
+    },
+    {
+      "epoch": 0.1994063680518079,
+      "grad_norm": 3.9565212726593018,
+      "learning_rate": 1.8535295433649822e-05,
+      "loss": 0.3312,
+      "step": 739
+    },
+    {
+      "epoch": 0.19967620075553158,
+      "grad_norm": 1.959877610206604,
+      "learning_rate": 1.85307376118355e-05,
+      "loss": 0.2905,
+      "step": 740
+    },
+    {
+      "epoch": 0.19994603345925527,
+      "grad_norm": 4.504112243652344,
+      "learning_rate": 1.8526173271784038e-05,
+      "loss": 0.3044,
+      "step": 741
+    },
+    {
+      "epoch": 0.20021586616297896,
+      "grad_norm": 1.4390770196914673,
+      "learning_rate": 1.8521602416982998e-05,
+      "loss": 0.2737,
+      "step": 742
+    },
+    {
+      "epoch": 0.20048569886670264,
+      "grad_norm": 2.5497422218322754,
+      "learning_rate": 1.8517025050924916e-05,
+      "loss": 0.291,
+      "step": 743
+    },
+    {
+      "epoch": 0.20075553157042633,
+      "grad_norm": 2.9591593742370605,
+      "learning_rate": 1.8512441177107306e-05,
+      "loss": 0.3216,
+      "step": 744
+    },
+    {
+      "epoch": 0.20102536427415002,
+      "grad_norm": 1.8294364213943481,
+      "learning_rate": 1.8507850799032653e-05,
+      "loss": 0.276,
+      "step": 745
+    },
+    {
+      "epoch": 0.2012951969778737,
+      "grad_norm": 2.586883783340454,
+      "learning_rate": 1.8503253920208402e-05,
+      "loss": 0.3374,
+      "step": 746
+    },
+    {
+      "epoch": 0.2015650296815974,
+      "grad_norm": 2.468473434448242,
+      "learning_rate": 1.8498650544146985e-05,
+      "loss": 0.2844,
+      "step": 747
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 2.6255712509155273,
+      "learning_rate": 1.8494040674365785e-05,
+      "loss": 0.2649,
+      "step": 748
+    },
+    {
+      "epoch": 0.2021046950890448,
+      "grad_norm": 3.8263967037200928,
+      "learning_rate": 1.8489424314387153e-05,
+      "loss": 0.3004,
+      "step": 749
+    },
+    {
+      "epoch": 0.2023745277927685,
+      "grad_norm": 3.075228691101074,
+      "learning_rate": 1.8484801467738386e-05,
+      "loss": 0.284,
+      "step": 750
+    },
+    {
+      "epoch": 0.20264436049649218,
+      "grad_norm": 1.2499473094940186,
+      "learning_rate": 1.8480172137951756e-05,
+      "loss": 0.3107,
+      "step": 751
+    },
+    {
+      "epoch": 0.20291419320021586,
+      "grad_norm": 3.487081527709961,
+      "learning_rate": 1.8475536328564478e-05,
+      "loss": 0.2483,
+      "step": 752
+    },
+    {
+      "epoch": 0.20318402590393955,
+      "grad_norm": 2.7405643463134766,
+      "learning_rate": 1.8470894043118717e-05,
+      "loss": 0.3108,
+      "step": 753
+    },
+    {
+      "epoch": 0.20345385860766324,
+      "grad_norm": 1.4685109853744507,
+      "learning_rate": 1.8466245285161593e-05,
+      "loss": 0.2885,
+      "step": 754
+    },
+    {
+      "epoch": 0.20372369131138693,
+      "grad_norm": 2.4970309734344482,
+      "learning_rate": 1.8461590058245166e-05,
+      "loss": 0.3022,
+      "step": 755
+    },
+    {
+      "epoch": 0.20399352401511064,
+      "grad_norm": 2.764352560043335,
+      "learning_rate": 1.845692836592644e-05,
+      "loss": 0.3086,
+      "step": 756
+    },
+    {
+      "epoch": 0.20426335671883433,
+      "grad_norm": 3.5047285556793213,
+      "learning_rate": 1.8452260211767358e-05,
+      "loss": 0.2912,
+      "step": 757
+    },
+    {
+      "epoch": 0.20453318942255802,
+      "grad_norm": 3.6481828689575195,
+      "learning_rate": 1.8447585599334806e-05,
+      "loss": 0.2715,
+      "step": 758
+    },
+    {
+      "epoch": 0.2048030221262817,
+      "grad_norm": 0.7495086789131165,
+      "learning_rate": 1.8442904532200603e-05,
+      "loss": 0.2993,
+      "step": 759
+    },
+    {
+      "epoch": 0.2050728548300054,
+      "grad_norm": 2.1055328845977783,
+      "learning_rate": 1.8438217013941494e-05,
+      "loss": 0.2976,
+      "step": 760
+    },
+    {
+      "epoch": 0.20534268753372908,
+      "grad_norm": 3.7416014671325684,
+      "learning_rate": 1.843352304813915e-05,
+      "loss": 0.2819,
+      "step": 761
+    },
+    {
+      "epoch": 0.20561252023745277,
+      "grad_norm": 1.9901988506317139,
+      "learning_rate": 1.842882263838019e-05,
+      "loss": 0.2705,
+      "step": 762
+    },
+    {
+      "epoch": 0.20588235294117646,
+      "grad_norm": 3.9372899532318115,
+      "learning_rate": 1.8424115788256135e-05,
+      "loss": 0.3265,
+      "step": 763
+    },
+    {
+      "epoch": 0.20615218564490015,
+      "grad_norm": 3.6716463565826416,
+      "learning_rate": 1.841940250136343e-05,
+      "loss": 0.2961,
+      "step": 764
+    },
+    {
+      "epoch": 0.20642201834862386,
+      "grad_norm": 3.034686803817749,
+      "learning_rate": 1.841468278130345e-05,
+      "loss": 0.2882,
+      "step": 765
+    },
+    {
+      "epoch": 0.20669185105234755,
+      "grad_norm": 1.1620547771453857,
+      "learning_rate": 1.8409956631682475e-05,
+      "loss": 0.2694,
+      "step": 766
+    },
+    {
+      "epoch": 0.20696168375607124,
+      "grad_norm": 3.552187442779541,
+      "learning_rate": 1.84052240561117e-05,
+      "loss": 0.3327,
+      "step": 767
+    },
+    {
+      "epoch": 0.20723151645979493,
+      "grad_norm": 1.9691848754882812,
+      "learning_rate": 1.840048505820723e-05,
+      "loss": 0.3193,
+      "step": 768
+    },
+    {
+      "epoch": 0.20750134916351862,
+      "grad_norm": 2.1665396690368652,
+      "learning_rate": 1.8395739641590084e-05,
+      "loss": 0.3027,
+      "step": 769
+    },
+    {
+      "epoch": 0.2077711818672423,
+      "grad_norm": 2.4780161380767822,
+      "learning_rate": 1.8390987809886173e-05,
+      "loss": 0.2969,
+      "step": 770
+    },
+    {
+      "epoch": 0.208041014570966,
+      "grad_norm": 0.6979816555976868,
+      "learning_rate": 1.838622956672632e-05,
+      "loss": 0.2727,
+      "step": 771
+    },
+    {
+      "epoch": 0.20831084727468968,
+      "grad_norm": 2.6476891040802,
+      "learning_rate": 1.838146491574624e-05,
+      "loss": 0.2924,
+      "step": 772
+    },
+    {
+      "epoch": 0.2085806799784134,
+      "grad_norm": 1.4862687587738037,
+      "learning_rate": 1.8376693860586546e-05,
+      "loss": 0.2762,
+      "step": 773
+    },
+    {
+      "epoch": 0.20885051268213708,
+      "grad_norm": 1.8782343864440918,
+      "learning_rate": 1.837191640489275e-05,
+      "loss": 0.3157,
+      "step": 774
+    },
+    {
+      "epoch": 0.20912034538586077,
+      "grad_norm": 3.6275768280029297,
+      "learning_rate": 1.836713255231525e-05,
+      "loss": 0.2993,
+      "step": 775
+    },
+    {
+      "epoch": 0.20939017808958446,
+      "grad_norm": 2.128124475479126,
+      "learning_rate": 1.836234230650933e-05,
+      "loss": 0.2846,
+      "step": 776
+    },
+    {
+      "epoch": 0.20966001079330815,
+      "grad_norm": 1.1387020349502563,
+      "learning_rate": 1.8357545671135163e-05,
+      "loss": 0.2736,
+      "step": 777
+    },
+    {
+      "epoch": 0.20992984349703184,
+      "grad_norm": 2.2182345390319824,
+      "learning_rate": 1.83527426498578e-05,
+      "loss": 0.3181,
+      "step": 778
+    },
+    {
+      "epoch": 0.21019967620075553,
+      "grad_norm": 1.6657764911651611,
+      "learning_rate": 1.8347933246347174e-05,
+      "loss": 0.2988,
+      "step": 779
+    },
+    {
+      "epoch": 0.2104695089044792,
+      "grad_norm": 1.0416992902755737,
+      "learning_rate": 1.834311746427809e-05,
+      "loss": 0.3057,
+      "step": 780
+    },
+    {
+      "epoch": 0.2107393416082029,
+      "grad_norm": 2.0946953296661377,
+      "learning_rate": 1.833829530733024e-05,
+      "loss": 0.2847,
+      "step": 781
+    },
+    {
+      "epoch": 0.21100917431192662,
+      "grad_norm": 1.821840524673462,
+      "learning_rate": 1.8333466779188173e-05,
+      "loss": 0.2971,
+      "step": 782
+    },
+    {
+      "epoch": 0.2112790070156503,
+      "grad_norm": 1.121606469154358,
+      "learning_rate": 1.832863188354131e-05,
+      "loss": 0.2626,
+      "step": 783
+    },
+    {
+      "epoch": 0.211548839719374,
+      "grad_norm": 3.6977665424346924,
+      "learning_rate": 1.832379062408394e-05,
+      "loss": 0.3309,
+      "step": 784
+    },
+    {
+      "epoch": 0.21181867242309768,
+      "grad_norm": 1.2279757261276245,
+      "learning_rate": 1.8318943004515214e-05,
+      "loss": 0.3225,
+      "step": 785
+    },
+    {
+      "epoch": 0.21208850512682137,
+      "grad_norm": 2.5102174282073975,
+      "learning_rate": 1.831408902853914e-05,
+      "loss": 0.3099,
+      "step": 786
+    },
+    {
+      "epoch": 0.21235833783054506,
+      "grad_norm": 0.8546987771987915,
+      "learning_rate": 1.8309228699864582e-05,
+      "loss": 0.2937,
+      "step": 787
+    },
+    {
+      "epoch": 0.21262817053426875,
+      "grad_norm": 0.7433753609657288,
+      "learning_rate": 1.8304362022205263e-05,
+      "loss": 0.3319,
+      "step": 788
+    },
+    {
+      "epoch": 0.21289800323799243,
+      "grad_norm": 2.1169779300689697,
+      "learning_rate": 1.8299488999279758e-05,
+      "loss": 0.3121,
+      "step": 789
+    },
+    {
+      "epoch": 0.21316783594171615,
+      "grad_norm": 0.8133589029312134,
+      "learning_rate": 1.8294609634811482e-05,
+      "loss": 0.2837,
+      "step": 790
+    },
+    {
+      "epoch": 0.21343766864543984,
+      "grad_norm": 1.609074592590332,
+      "learning_rate": 1.8289723932528704e-05,
+      "loss": 0.3139,
+      "step": 791
+    },
+    {
+      "epoch": 0.21370750134916353,
+      "grad_norm": 1.202027678489685,
+      "learning_rate": 1.8284831896164527e-05,
+      "loss": 0.2883,
+      "step": 792
+    },
+    {
+      "epoch": 0.2139773340528872,
+      "grad_norm": 0.9812056422233582,
+      "learning_rate": 1.8279933529456905e-05,
+      "loss": 0.301,
+      "step": 793
+    },
+    {
+      "epoch": 0.2142471667566109,
+      "grad_norm": 0.9920928478240967,
+      "learning_rate": 1.8275028836148615e-05,
+      "loss": 0.2861,
+      "step": 794
+    },
+    {
+      "epoch": 0.2145169994603346,
+      "grad_norm": 1.6611597537994385,
+      "learning_rate": 1.8270117819987287e-05,
+      "loss": 0.2712,
+      "step": 795
+    },
+    {
+      "epoch": 0.21478683216405828,
+      "grad_norm": 1.0533279180526733,
+      "learning_rate": 1.8265200484725364e-05,
+      "loss": 0.3111,
+      "step": 796
+    },
+    {
+      "epoch": 0.21505666486778197,
+      "grad_norm": 3.3801465034484863,
+      "learning_rate": 1.8260276834120123e-05,
+      "loss": 0.3101,
+      "step": 797
+    },
+    {
+      "epoch": 0.21532649757150565,
+      "grad_norm": 1.328139066696167,
+      "learning_rate": 1.8255346871933673e-05,
+      "loss": 0.3394,
+      "step": 798
+    },
+    {
+      "epoch": 0.21559633027522937,
+      "grad_norm": 3.592576265335083,
+      "learning_rate": 1.8250410601932937e-05,
+      "loss": 0.2541,
+      "step": 799
+    },
+    {
+      "epoch": 0.21586616297895306,
+      "grad_norm": 1.3668057918548584,
+      "learning_rate": 1.8245468027889663e-05,
+      "loss": 0.296,
+      "step": 800
+    },
+    {
+      "epoch": 0.21613599568267675,
+      "grad_norm": 2.606140375137329,
+      "learning_rate": 1.8240519153580417e-05,
+      "loss": 0.3135,
+      "step": 801
+    },
+    {
+      "epoch": 0.21640582838640043,
+      "grad_norm": 1.8793245553970337,
+      "learning_rate": 1.823556398278657e-05,
+      "loss": 0.2855,
+      "step": 802
+    },
+    {
+      "epoch": 0.21667566109012412,
+      "grad_norm": 1.1196082830429077,
+      "learning_rate": 1.8230602519294314e-05,
+      "loss": 0.302,
+      "step": 803
+    },
+    {
+      "epoch": 0.2169454937938478,
+      "grad_norm": 0.8586265444755554,
+      "learning_rate": 1.8225634766894645e-05,
+      "loss": 0.3059,
+      "step": 804
+    },
+    {
+      "epoch": 0.2172153264975715,
+      "grad_norm": 1.1635435819625854,
+      "learning_rate": 1.8220660729383365e-05,
+      "loss": 0.3223,
+      "step": 805
+    },
+    {
+      "epoch": 0.2174851592012952,
+      "grad_norm": 2.7796363830566406,
+      "learning_rate": 1.821568041056108e-05,
+      "loss": 0.2895,
+      "step": 806
+    },
+    {
+      "epoch": 0.2177549919050189,
+      "grad_norm": 2.6912918090820312,
+      "learning_rate": 1.8210693814233188e-05,
+      "loss": 0.284,
+      "step": 807
+    },
+    {
+      "epoch": 0.2180248246087426,
+      "grad_norm": 1.7719309329986572,
+      "learning_rate": 1.820570094420989e-05,
+      "loss": 0.2901,
+      "step": 808
+    },
+    {
+      "epoch": 0.21829465731246628,
+      "grad_norm": 1.6466035842895508,
+      "learning_rate": 1.8200701804306186e-05,
+      "loss": 0.3027,
+      "step": 809
+    },
+    {
+      "epoch": 0.21856449001618997,
+      "grad_norm": 1.1211652755737305,
+      "learning_rate": 1.8195696398341857e-05,
+      "loss": 0.3002,
+      "step": 810
+    },
+    {
+      "epoch": 0.21883432271991365,
+      "grad_norm": 3.222228527069092,
+      "learning_rate": 1.819068473014147e-05,
+      "loss": 0.3048,
+      "step": 811
+    },
+    {
+      "epoch": 0.21910415542363734,
+      "grad_norm": 1.4517183303833008,
+      "learning_rate": 1.8185666803534388e-05,
+      "loss": 0.2597,
+      "step": 812
+    },
+    {
+      "epoch": 0.21937398812736103,
+      "grad_norm": 2.813934087753296,
+      "learning_rate": 1.8180642622354753e-05,
+      "loss": 0.3133,
+      "step": 813
+    },
+    {
+      "epoch": 0.21964382083108472,
+      "grad_norm": 3.341914415359497,
+      "learning_rate": 1.817561219044148e-05,
+      "loss": 0.3021,
+      "step": 814
+    },
+    {
+      "epoch": 0.2199136535348084,
+      "grad_norm": 1.2155262231826782,
+      "learning_rate": 1.8170575511638258e-05,
+      "loss": 0.2973,
+      "step": 815
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 7.271151065826416,
+      "learning_rate": 1.8165532589793563e-05,
+      "loss": 0.3131,
+      "step": 816
+    },
+    {
+      "epoch": 0.2204533189422558,
+      "grad_norm": 2.0311813354492188,
+      "learning_rate": 1.8160483428760628e-05,
+      "loss": 0.3034,
+      "step": 817
+    },
+    {
+      "epoch": 0.2207231516459795,
+      "grad_norm": 4.808474540710449,
+      "learning_rate": 1.8155428032397457e-05,
+      "loss": 0.2882,
+      "step": 818
+    },
+    {
+      "epoch": 0.2209929843497032,
+      "grad_norm": 5.231053829193115,
+      "learning_rate": 1.8150366404566827e-05,
+      "loss": 0.2986,
+      "step": 819
+    },
+    {
+      "epoch": 0.22126281705342687,
+      "grad_norm": 2.4351983070373535,
+      "learning_rate": 1.814529854913626e-05,
+      "loss": 0.3208,
+      "step": 820
+    },
+    {
+      "epoch": 0.22153264975715056,
+      "grad_norm": 6.930006980895996,
+      "learning_rate": 1.814022446997805e-05,
+      "loss": 0.2971,
+      "step": 821
+    },
+    {
+      "epoch": 0.22180248246087425,
+      "grad_norm": 4.602273941040039,
+      "learning_rate": 1.8135144170969235e-05,
+      "loss": 0.3322,
+      "step": 822
+    },
+    {
+      "epoch": 0.22207231516459794,
+      "grad_norm": 1.506335973739624,
+      "learning_rate": 1.8130057655991625e-05,
+      "loss": 0.2882,
+      "step": 823
+    },
+    {
+      "epoch": 0.22234214786832163,
+      "grad_norm": 3.4634897708892822,
+      "learning_rate": 1.8124964928931755e-05,
+      "loss": 0.2802,
+      "step": 824
+    },
+    {
+      "epoch": 0.22261198057204534,
+      "grad_norm": 3.555952310562134,
+      "learning_rate": 1.8119865993680924e-05,
+      "loss": 0.2681,
+      "step": 825
+    },
+    {
+      "epoch": 0.22288181327576903,
+      "grad_norm": 0.8892737627029419,
+      "learning_rate": 1.8114760854135168e-05,
+      "loss": 0.3056,
+      "step": 826
+    },
+    {
+      "epoch": 0.22315164597949272,
+      "grad_norm": 5.157120227813721,
+      "learning_rate": 1.8109649514195268e-05,
+      "loss": 0.2658,
+      "step": 827
+    },
+    {
+      "epoch": 0.2234214786832164,
+      "grad_norm": 2.303499698638916,
+      "learning_rate": 1.810453197776673e-05,
+      "loss": 0.3001,
+      "step": 828
+    },
+    {
+      "epoch": 0.2236913113869401,
+      "grad_norm": 1.5508161783218384,
+      "learning_rate": 1.809940824875981e-05,
+      "loss": 0.2627,
+      "step": 829
+    },
+    {
+      "epoch": 0.22396114409066378,
+      "grad_norm": 3.9749433994293213,
+      "learning_rate": 1.8094278331089496e-05,
+      "loss": 0.2964,
+      "step": 830
+    },
+    {
+      "epoch": 0.22423097679438747,
+      "grad_norm": 1.2338846921920776,
+      "learning_rate": 1.808914222867548e-05,
+      "loss": 0.2942,
+      "step": 831
+    },
+    {
+      "epoch": 0.22450080949811116,
+      "grad_norm": 1.1691101789474487,
+      "learning_rate": 1.808399994544222e-05,
+      "loss": 0.3077,
+      "step": 832
+    },
+    {
+      "epoch": 0.22477064220183487,
+      "grad_norm": 2.5266199111938477,
+      "learning_rate": 1.8078851485318857e-05,
+      "loss": 0.3243,
+      "step": 833
+    },
+    {
+      "epoch": 0.22504047490555856,
+      "grad_norm": 2.4897053241729736,
+      "learning_rate": 1.807369685223928e-05,
+      "loss": 0.2978,
+      "step": 834
+    },
+    {
+      "epoch": 0.22531030760928225,
+      "grad_norm": 1.3095663785934448,
+      "learning_rate": 1.806853605014208e-05,
+      "loss": 0.2852,
+      "step": 835
+    },
+    {
+      "epoch": 0.22558014031300594,
+      "grad_norm": 2.604254722595215,
+      "learning_rate": 1.806336908297057e-05,
+      "loss": 0.2729,
+      "step": 836
+    },
+    {
+      "epoch": 0.22584997301672963,
+      "grad_norm": 1.1099776029586792,
+      "learning_rate": 1.8058195954672758e-05,
+      "loss": 0.3021,
+      "step": 837
+    },
+    {
+      "epoch": 0.22611980572045332,
+      "grad_norm": 3.178739070892334,
+      "learning_rate": 1.805301666920138e-05,
+      "loss": 0.2664,
+      "step": 838
+    },
+    {
+      "epoch": 0.226389638424177,
+      "grad_norm": 2.026181221008301,
+      "learning_rate": 1.804783123051387e-05,
+      "loss": 0.2947,
+      "step": 839
+    },
+    {
+      "epoch": 0.2266594711279007,
+      "grad_norm": 1.9811091423034668,
+      "learning_rate": 1.8042639642572352e-05,
+      "loss": 0.2888,
+      "step": 840
+    },
+    {
+      "epoch": 0.22692930383162438,
+      "grad_norm": 3.24641752243042,
+      "learning_rate": 1.8037441909343666e-05,
+      "loss": 0.3019,
+      "step": 841
+    },
+    {
+      "epoch": 0.2271991365353481,
+      "grad_norm": 2.3258750438690186,
+      "learning_rate": 1.803223803479934e-05,
+      "loss": 0.3021,
+      "step": 842
+    },
+    {
+      "epoch": 0.22746896923907178,
+      "grad_norm": 1.1767789125442505,
+      "learning_rate": 1.8027028022915587e-05,
+      "loss": 0.266,
+      "step": 843
+    },
+    {
+      "epoch": 0.22773880194279547,
+      "grad_norm": 3.1815693378448486,
+      "learning_rate": 1.802181187767332e-05,
+      "loss": 0.3122,
+      "step": 844
+    },
+    {
+      "epoch": 0.22800863464651916,
+      "grad_norm": 2.461888551712036,
+      "learning_rate": 1.8016589603058134e-05,
+      "loss": 0.2852,
+      "step": 845
+    },
+    {
+      "epoch": 0.22827846735024285,
+      "grad_norm": 2.00380539894104,
+      "learning_rate": 1.8011361203060308e-05,
+      "loss": 0.2937,
+      "step": 846
+    },
+    {
+      "epoch": 0.22854830005396654,
+      "grad_norm": 4.0061726570129395,
+      "learning_rate": 1.8006126681674804e-05,
+      "loss": 0.259,
+      "step": 847
+    },
+    {
+      "epoch": 0.22881813275769022,
+      "grad_norm": 1.3365811109542847,
+      "learning_rate": 1.8000886042901254e-05,
+      "loss": 0.2932,
+      "step": 848
+    },
+    {
+      "epoch": 0.2290879654614139,
+      "grad_norm": 2.8306777477264404,
+      "learning_rate": 1.7995639290743975e-05,
+      "loss": 0.313,
+      "step": 849
+    },
+    {
+      "epoch": 0.22935779816513763,
+      "grad_norm": 2.398937463760376,
+      "learning_rate": 1.7990386429211945e-05,
+      "loss": 0.307,
+      "step": 850
+    },
+    {
+      "epoch": 0.22962763086886132,
+      "grad_norm": 0.8818338513374329,
+      "learning_rate": 1.7985127462318815e-05,
+      "loss": 0.2936,
+      "step": 851
+    },
+    {
+      "epoch": 0.229897463572585,
+      "grad_norm": 1.6002426147460938,
+      "learning_rate": 1.7979862394082902e-05,
+      "loss": 0.2876,
+      "step": 852
+    },
+    {
+      "epoch": 0.2301672962763087,
+      "grad_norm": 3.8820383548736572,
+      "learning_rate": 1.797459122852718e-05,
+      "loss": 0.2676,
+      "step": 853
+    },
+    {
+      "epoch": 0.23043712898003238,
+      "grad_norm": 0.9613157510757446,
+      "learning_rate": 1.7969313969679286e-05,
+      "loss": 0.3274,
+      "step": 854
+    },
+    {
+      "epoch": 0.23070696168375607,
+      "grad_norm": 4.258617401123047,
+      "learning_rate": 1.7964030621571515e-05,
+      "loss": 0.307,
+      "step": 855
+    },
+    {
+      "epoch": 0.23097679438747976,
+      "grad_norm": 2.0114426612854004,
+      "learning_rate": 1.7958741188240808e-05,
+      "loss": 0.2769,
+      "step": 856
+    },
+    {
+      "epoch": 0.23124662709120344,
+      "grad_norm": 1.8643455505371094,
+      "learning_rate": 1.795344567372876e-05,
+      "loss": 0.2919,
+      "step": 857
+    },
+    {
+      "epoch": 0.23151645979492713,
+      "grad_norm": 3.118746519088745,
+      "learning_rate": 1.7948144082081615e-05,
+      "loss": 0.2976,
+      "step": 858
+    },
+    {
+      "epoch": 0.23178629249865085,
+      "grad_norm": 1.6401116847991943,
+      "learning_rate": 1.7942836417350252e-05,
+      "loss": 0.2871,
+      "step": 859
+    },
+    {
+      "epoch": 0.23205612520237454,
+      "grad_norm": 1.883307933807373,
+      "learning_rate": 1.7937522683590204e-05,
+      "loss": 0.2949,
+      "step": 860
+    },
+    {
+      "epoch": 0.23232595790609822,
+      "grad_norm": 3.200051784515381,
+      "learning_rate": 1.7932202884861625e-05,
+      "loss": 0.3095,
+      "step": 861
+    },
+    {
+      "epoch": 0.2325957906098219,
+      "grad_norm": 2.1455490589141846,
+      "learning_rate": 1.7926877025229313e-05,
+      "loss": 0.325,
+      "step": 862
+    },
+    {
+      "epoch": 0.2328656233135456,
+      "grad_norm": 3.2392144203186035,
+      "learning_rate": 1.7921545108762694e-05,
+      "loss": 0.2896,
+      "step": 863
+    },
+    {
+      "epoch": 0.2331354560172693,
+      "grad_norm": 3.9804303646087646,
+      "learning_rate": 1.791620713953583e-05,
+      "loss": 0.3044,
+      "step": 864
+    },
+    {
+      "epoch": 0.23340528872099298,
+      "grad_norm": 0.9719670414924622,
+      "learning_rate": 1.7910863121627396e-05,
+      "loss": 0.2941,
+      "step": 865
+    },
+    {
+      "epoch": 0.23367512142471666,
+      "grad_norm": 4.572351455688477,
+      "learning_rate": 1.7905513059120693e-05,
+      "loss": 0.2635,
+      "step": 866
+    },
+    {
+      "epoch": 0.23394495412844038,
+      "grad_norm": 0.9175219535827637,
+      "learning_rate": 1.7900156956103643e-05,
+      "loss": 0.2853,
+      "step": 867
+    },
+    {
+      "epoch": 0.23421478683216407,
+      "grad_norm": 2.314467668533325,
+      "learning_rate": 1.789479481666878e-05,
+      "loss": 0.2691,
+      "step": 868
+    },
+    {
+      "epoch": 0.23448461953588776,
+      "grad_norm": 1.9501879215240479,
+      "learning_rate": 1.7889426644913257e-05,
+      "loss": 0.3141,
+      "step": 869
+    },
+    {
+      "epoch": 0.23475445223961144,
+      "grad_norm": 1.4802796840667725,
+      "learning_rate": 1.7884052444938827e-05,
+      "loss": 0.3027,
+      "step": 870
+    },
+    {
+      "epoch": 0.23502428494333513,
+      "grad_norm": 2.9978981018066406,
+      "learning_rate": 1.7878672220851853e-05,
+      "loss": 0.2738,
+      "step": 871
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 2.3386003971099854,
+      "learning_rate": 1.7873285976763303e-05,
+      "loss": 0.2909,
+      "step": 872
+    },
+    {
+      "epoch": 0.2355639503507825,
+      "grad_norm": 0.8048591017723083,
+      "learning_rate": 1.7867893716788746e-05,
+      "loss": 0.2942,
+      "step": 873
+    },
+    {
+      "epoch": 0.2358337830545062,
+      "grad_norm": 3.168987989425659,
+      "learning_rate": 1.786249544504834e-05,
+      "loss": 0.2787,
+      "step": 874
+    },
+    {
+      "epoch": 0.23610361575822988,
+      "grad_norm": 0.8916498422622681,
+      "learning_rate": 1.7857091165666845e-05,
+      "loss": 0.2643,
+      "step": 875
+    },
+    {
+      "epoch": 0.2363734484619536,
+      "grad_norm": 1.286604404449463,
+      "learning_rate": 1.7851680882773607e-05,
+      "loss": 0.3192,
+      "step": 876
+    },
+    {
+      "epoch": 0.2366432811656773,
+      "grad_norm": 3.530160427093506,
+      "learning_rate": 1.784626460050256e-05,
+      "loss": 0.3177,
+      "step": 877
+    },
+    {
+      "epoch": 0.23691311386940098,
+      "grad_norm": 2.8858065605163574,
+      "learning_rate": 1.784084232299222e-05,
+      "loss": 0.2855,
+      "step": 878
+    },
+    {
+      "epoch": 0.23718294657312466,
+      "grad_norm": 1.1932474374771118,
+      "learning_rate": 1.7835414054385692e-05,
+      "loss": 0.3229,
+      "step": 879
+    },
+    {
+      "epoch": 0.23745277927684835,
+      "grad_norm": 0.9355245232582092,
+      "learning_rate": 1.7829979798830646e-05,
+      "loss": 0.2777,
+      "step": 880
+    },
+    {
+      "epoch": 0.23772261198057204,
+      "grad_norm": 1.0392200946807861,
+      "learning_rate": 1.7824539560479338e-05,
+      "loss": 0.3105,
+      "step": 881
+    },
+    {
+      "epoch": 0.23799244468429573,
+      "grad_norm": 0.9828007221221924,
+      "learning_rate": 1.781909334348859e-05,
+      "loss": 0.3163,
+      "step": 882
+    },
+    {
+      "epoch": 0.23826227738801942,
+      "grad_norm": 1.9239410161972046,
+      "learning_rate": 1.781364115201979e-05,
+      "loss": 0.2633,
+      "step": 883
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 1.6579258441925049,
+      "learning_rate": 1.7808182990238894e-05,
+      "loss": 0.2863,
+      "step": 884
+    },
+    {
+      "epoch": 0.23880194279546682,
+      "grad_norm": 2.4309847354888916,
+      "learning_rate": 1.780271886231642e-05,
+      "loss": 0.2903,
+      "step": 885
+    },
+    {
+      "epoch": 0.2390717754991905,
+      "grad_norm": 0.9034820199012756,
+      "learning_rate": 1.779724877242745e-05,
+      "loss": 0.2927,
+      "step": 886
+    },
+    {
+      "epoch": 0.2393416082029142,
+      "grad_norm": 2.8854312896728516,
+      "learning_rate": 1.7791772724751606e-05,
+      "loss": 0.2805,
+      "step": 887
+    },
+    {
+      "epoch": 0.23961144090663788,
+      "grad_norm": 1.9433176517486572,
+      "learning_rate": 1.7786290723473082e-05,
+      "loss": 0.2649,
+      "step": 888
+    },
+    {
+      "epoch": 0.23988127361036157,
+      "grad_norm": 1.016381859779358,
+      "learning_rate": 1.778080277278061e-05,
+      "loss": 0.2899,
+      "step": 889
+    },
+    {
+      "epoch": 0.24015110631408526,
+      "grad_norm": 2.688339948654175,
+      "learning_rate": 1.777530887686746e-05,
+      "loss": 0.2569,
+      "step": 890
+    },
+    {
+      "epoch": 0.24042093901780895,
+      "grad_norm": 2.536651849746704,
+      "learning_rate": 1.7769809039931466e-05,
+      "loss": 0.3222,
+      "step": 891
+    },
+    {
+      "epoch": 0.24069077172153264,
+      "grad_norm": 2.3051199913024902,
+      "learning_rate": 1.776430326617498e-05,
+      "loss": 0.3093,
+      "step": 892
+    },
+    {
+      "epoch": 0.24096060442525635,
+      "grad_norm": 1.4840691089630127,
+      "learning_rate": 1.7758791559804904e-05,
+      "loss": 0.2611,
+      "step": 893
+    },
+    {
+      "epoch": 0.24123043712898004,
+      "grad_norm": 2.5153181552886963,
+      "learning_rate": 1.7753273925032665e-05,
+      "loss": 0.2982,
+      "step": 894
+    },
+    {
+      "epoch": 0.24150026983270373,
+      "grad_norm": 1.8463948965072632,
+      "learning_rate": 1.774775036607423e-05,
+      "loss": 0.312,
+      "step": 895
+    },
+    {
+      "epoch": 0.24177010253642742,
+      "grad_norm": 3.3008482456207275,
+      "learning_rate": 1.774222088715008e-05,
+      "loss": 0.3246,
+      "step": 896
+    },
+    {
+      "epoch": 0.2420399352401511,
+      "grad_norm": 1.3531514406204224,
+      "learning_rate": 1.7736685492485225e-05,
+      "loss": 0.3538,
+      "step": 897
+    },
+    {
+      "epoch": 0.2423097679438748,
+      "grad_norm": 0.9987041354179382,
+      "learning_rate": 1.77311441863092e-05,
+      "loss": 0.2699,
+      "step": 898
+    },
+    {
+      "epoch": 0.24257960064759848,
+      "grad_norm": 3.598644495010376,
+      "learning_rate": 1.7725596972856047e-05,
+      "loss": 0.3312,
+      "step": 899
+    },
+    {
+      "epoch": 0.24284943335132217,
+      "grad_norm": 1.4154469966888428,
+      "learning_rate": 1.7720043856364334e-05,
+      "loss": 0.2858,
+      "step": 900
+    },
+    {
+      "epoch": 0.24311926605504589,
+      "grad_norm": 1.3847718238830566,
+      "learning_rate": 1.7714484841077127e-05,
+      "loss": 0.295,
+      "step": 901
+    },
+    {
+      "epoch": 0.24338909875876957,
+      "grad_norm": 1.0586942434310913,
+      "learning_rate": 1.770891993124201e-05,
+      "loss": 0.3019,
+      "step": 902
+    },
+    {
+      "epoch": 0.24365893146249326,
+      "grad_norm": 3.7392098903656006,
+      "learning_rate": 1.770334913111106e-05,
+      "loss": 0.3464,
+      "step": 903
+    },
+    {
+      "epoch": 0.24392876416621695,
+      "grad_norm": 1.1105849742889404,
+      "learning_rate": 1.769777244494086e-05,
+      "loss": 0.2888,
+      "step": 904
+    },
+    {
+      "epoch": 0.24419859686994064,
+      "grad_norm": 2.235166311264038,
+      "learning_rate": 1.7692189876992496e-05,
+      "loss": 0.3071,
+      "step": 905
+    },
+    {
+      "epoch": 0.24446842957366433,
+      "grad_norm": 1.2054306268692017,
+      "learning_rate": 1.768660143153154e-05,
+      "loss": 0.2873,
+      "step": 906
+    },
+    {
+      "epoch": 0.244738262277388,
+      "grad_norm": 1.2031320333480835,
+      "learning_rate": 1.7681007112828063e-05,
+      "loss": 0.2725,
+      "step": 907
+    },
+    {
+      "epoch": 0.2450080949811117,
+      "grad_norm": 1.5726971626281738,
+      "learning_rate": 1.767540692515661e-05,
+      "loss": 0.3191,
+      "step": 908
+    },
+    {
+      "epoch": 0.2452779276848354,
+      "grad_norm": 1.0206282138824463,
+      "learning_rate": 1.766980087279622e-05,
+      "loss": 0.2929,
+      "step": 909
+    },
+    {
+      "epoch": 0.2455477603885591,
+      "grad_norm": 1.2062270641326904,
+      "learning_rate": 1.7664188960030422e-05,
+      "loss": 0.3299,
+      "step": 910
+    },
+    {
+      "epoch": 0.2458175930922828,
+      "grad_norm": 1.2147338390350342,
+      "learning_rate": 1.7658571191147206e-05,
+      "loss": 0.2983,
+      "step": 911
+    },
+    {
+      "epoch": 0.24608742579600648,
+      "grad_norm": 1.7091821432113647,
+      "learning_rate": 1.765294757043904e-05,
+      "loss": 0.307,
+      "step": 912
+    },
+    {
+      "epoch": 0.24635725849973017,
+      "grad_norm": 1.1049470901489258,
+      "learning_rate": 1.7647318102202876e-05,
+      "loss": 0.2582,
+      "step": 913
+    },
+    {
+      "epoch": 0.24662709120345386,
+      "grad_norm": 1.083500862121582,
+      "learning_rate": 1.7641682790740122e-05,
+      "loss": 0.2773,
+      "step": 914
+    },
+    {
+      "epoch": 0.24689692390717755,
+      "grad_norm": 2.2466421127319336,
+      "learning_rate": 1.7636041640356648e-05,
+      "loss": 0.2847,
+      "step": 915
+    },
+    {
+      "epoch": 0.24716675661090123,
+      "grad_norm": 0.9754478931427002,
+      "learning_rate": 1.7630394655362798e-05,
+      "loss": 0.3462,
+      "step": 916
+    },
+    {
+      "epoch": 0.24743658931462492,
+      "grad_norm": 2.1321017742156982,
+      "learning_rate": 1.7624741840073365e-05,
+      "loss": 0.3304,
+      "step": 917
+    },
+    {
+      "epoch": 0.24770642201834864,
+      "grad_norm": 2.1501708030700684,
+      "learning_rate": 1.76190831988076e-05,
+      "loss": 0.2931,
+      "step": 918
+    },
+    {
+      "epoch": 0.24797625472207233,
+      "grad_norm": 0.72209632396698,
+      "learning_rate": 1.76134187358892e-05,
+      "loss": 0.2786,
+      "step": 919
+    },
+    {
+      "epoch": 0.248246087425796,
+      "grad_norm": 1.4570835828781128,
+      "learning_rate": 1.7607748455646323e-05,
+      "loss": 0.2866,
+      "step": 920
+    },
+    {
+      "epoch": 0.2485159201295197,
+      "grad_norm": 2.0870180130004883,
+      "learning_rate": 1.760207236241156e-05,
+      "loss": 0.3067,
+      "step": 921
+    },
+    {
+      "epoch": 0.2487857528332434,
+      "grad_norm": 0.9873558282852173,
+      "learning_rate": 1.7596390460521946e-05,
+      "loss": 0.2554,
+      "step": 922
+    },
+    {
+      "epoch": 0.24905558553696708,
+      "grad_norm": 3.806623697280884,
+      "learning_rate": 1.7590702754318957e-05,
+      "loss": 0.3053,
+      "step": 923
+    },
+    {
+      "epoch": 0.24932541824069077,
+      "grad_norm": 0.801768958568573,
+      "learning_rate": 1.7585009248148507e-05,
+      "loss": 0.2982,
+      "step": 924
+    },
+    {
+      "epoch": 0.24959525094441445,
+      "grad_norm": 2.8146047592163086,
+      "learning_rate": 1.757930994636093e-05,
+      "loss": 0.3152,
+      "step": 925
+    },
+    {
+      "epoch": 0.24986508364813814,
+      "grad_norm": 0.8977971076965332,
+      "learning_rate": 1.7573604853311005e-05,
+      "loss": 0.3012,
+      "step": 926
+    },
+    {
+      "epoch": 0.25013491635186186,
+      "grad_norm": 1.3070673942565918,
+      "learning_rate": 1.756789397335792e-05,
+      "loss": 0.2924,
+      "step": 927
+    },
+    {
+      "epoch": 0.2504047490555855,
+      "grad_norm": 1.2641279697418213,
+      "learning_rate": 1.7562177310865296e-05,
+      "loss": 0.311,
+      "step": 928
+    },
+    {
+      "epoch": 0.25067458175930923,
+      "grad_norm": 1.546240210533142,
+      "learning_rate": 1.7556454870201166e-05,
+      "loss": 0.306,
+      "step": 929
+    },
+    {
+      "epoch": 0.2509444144630329,
+      "grad_norm": 1.0503368377685547,
+      "learning_rate": 1.755072665573798e-05,
+      "loss": 0.2732,
+      "step": 930
+    },
+    {
+      "epoch": 0.2512142471667566,
+      "grad_norm": 1.7501614093780518,
+      "learning_rate": 1.75449926718526e-05,
+      "loss": 0.2732,
+      "step": 931
+    },
+    {
+      "epoch": 0.2514840798704803,
+      "grad_norm": 1.0615034103393555,
+      "learning_rate": 1.7539252922926295e-05,
+      "loss": 0.2852,
+      "step": 932
+    },
+    {
+      "epoch": 0.251753912574204,
+      "grad_norm": 1.0441597700119019,
+      "learning_rate": 1.7533507413344746e-05,
+      "loss": 0.2734,
+      "step": 933
+    },
+    {
+      "epoch": 0.2520237452779277,
+      "grad_norm": 1.1869550943374634,
+      "learning_rate": 1.7527756147498026e-05,
+      "loss": 0.281,
+      "step": 934
+    },
+    {
+      "epoch": 0.25229357798165136,
+      "grad_norm": 1.675325632095337,
+      "learning_rate": 1.7521999129780606e-05,
+      "loss": 0.3105,
+      "step": 935
+    },
+    {
+      "epoch": 0.2525634106853751,
+      "grad_norm": 1.3030115365982056,
+      "learning_rate": 1.7516236364591365e-05,
+      "loss": 0.3177,
+      "step": 936
+    },
+    {
+      "epoch": 0.25283324338909874,
+      "grad_norm": 1.1539233922958374,
+      "learning_rate": 1.7510467856333556e-05,
+      "loss": 0.2929,
+      "step": 937
+    },
+    {
+      "epoch": 0.25310307609282245,
+      "grad_norm": 1.0007246732711792,
+      "learning_rate": 1.7504693609414835e-05,
+      "loss": 0.2893,
+      "step": 938
+    },
+    {
+      "epoch": 0.2533729087965461,
+      "grad_norm": 1.9604766368865967,
+      "learning_rate": 1.7498913628247232e-05,
+      "loss": 0.2754,
+      "step": 939
+    },
+    {
+      "epoch": 0.25364274150026983,
+      "grad_norm": 1.084153652191162,
+      "learning_rate": 1.7493127917247168e-05,
+      "loss": 0.2739,
+      "step": 940
+    },
+    {
+      "epoch": 0.25391257420399355,
+      "grad_norm": 1.1053500175476074,
+      "learning_rate": 1.748733648083543e-05,
+      "loss": 0.2671,
+      "step": 941
+    },
+    {
+      "epoch": 0.2541824069077172,
+      "grad_norm": 0.9310557842254639,
+      "learning_rate": 1.74815393234372e-05,
+      "loss": 0.3104,
+      "step": 942
+    },
+    {
+      "epoch": 0.2544522396114409,
+      "grad_norm": 0.8322042226791382,
+      "learning_rate": 1.7475736449482006e-05,
+      "loss": 0.275,
+      "step": 943
+    },
+    {
+      "epoch": 0.2547220723151646,
+      "grad_norm": 0.6906194090843201,
+      "learning_rate": 1.746992786340376e-05,
+      "loss": 0.2645,
+      "step": 944
+    },
+    {
+      "epoch": 0.2549919050188883,
+      "grad_norm": 1.0623716115951538,
+      "learning_rate": 1.7464113569640736e-05,
+      "loss": 0.2895,
+      "step": 945
+    },
+    {
+      "epoch": 0.25526173772261196,
+      "grad_norm": 1.5662360191345215,
+      "learning_rate": 1.7458293572635573e-05,
+      "loss": 0.2563,
+      "step": 946
+    },
+    {
+      "epoch": 0.2555315704263357,
+      "grad_norm": 1.2521657943725586,
+      "learning_rate": 1.7452467876835252e-05,
+      "loss": 0.3112,
+      "step": 947
+    },
+    {
+      "epoch": 0.2558014031300594,
+      "grad_norm": 1.2860701084136963,
+      "learning_rate": 1.7446636486691123e-05,
+      "loss": 0.2805,
+      "step": 948
+    },
+    {
+      "epoch": 0.25607123583378305,
+      "grad_norm": 1.7004153728485107,
+      "learning_rate": 1.744079940665889e-05,
+      "loss": 0.2872,
+      "step": 949
+    },
+    {
+      "epoch": 0.25634106853750677,
+      "grad_norm": 0.699410617351532,
+      "learning_rate": 1.743495664119859e-05,
+      "loss": 0.2477,
+      "step": 950
+    },
+    {
+      "epoch": 0.2566109012412304,
+      "grad_norm": 2.7868244647979736,
+      "learning_rate": 1.7429108194774614e-05,
+      "loss": 0.3166,
+      "step": 951
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 2.367600440979004,
+      "learning_rate": 1.7423254071855696e-05,
+      "loss": 0.2951,
+      "step": 952
+    },
+    {
+      "epoch": 0.2571505666486778,
+      "grad_norm": 1.098291039466858,
+      "learning_rate": 1.7417394276914895e-05,
+      "loss": 0.3099,
+      "step": 953
+    },
+    {
+      "epoch": 0.2574203993524015,
+      "grad_norm": 0.9729587435722351,
+      "learning_rate": 1.741152881442962e-05,
+      "loss": 0.2652,
+      "step": 954
+    },
+    {
+      "epoch": 0.2576902320561252,
+      "grad_norm": 0.9850955009460449,
+      "learning_rate": 1.7405657688881597e-05,
+      "loss": 0.2758,
+      "step": 955
+    },
+    {
+      "epoch": 0.2579600647598489,
+      "grad_norm": 1.0178335905075073,
+      "learning_rate": 1.7399780904756885e-05,
+      "loss": 0.2741,
+      "step": 956
+    },
+    {
+      "epoch": 0.2582298974635726,
+      "grad_norm": 1.4973105192184448,
+      "learning_rate": 1.739389846654587e-05,
+      "loss": 0.3124,
+      "step": 957
+    },
+    {
+      "epoch": 0.25849973016729627,
+      "grad_norm": 1.0455963611602783,
+      "learning_rate": 1.7388010378743255e-05,
+      "loss": 0.2656,
+      "step": 958
+    },
+    {
+      "epoch": 0.25876956287102,
+      "grad_norm": 1.4307544231414795,
+      "learning_rate": 1.738211664584805e-05,
+      "loss": 0.3006,
+      "step": 959
+    },
+    {
+      "epoch": 0.25903939557474365,
+      "grad_norm": 1.649924635887146,
+      "learning_rate": 1.73762172723636e-05,
+      "loss": 0.3177,
+      "step": 960
+    },
+    {
+      "epoch": 0.25930922827846736,
+      "grad_norm": 1.6095399856567383,
+      "learning_rate": 1.737031226279754e-05,
+      "loss": 0.3086,
+      "step": 961
+    },
+    {
+      "epoch": 0.259579060982191,
+      "grad_norm": 1.0704795122146606,
+      "learning_rate": 1.736440162166182e-05,
+      "loss": 0.3327,
+      "step": 962
+    },
+    {
+      "epoch": 0.25984889368591474,
+      "grad_norm": 0.9887911677360535,
+      "learning_rate": 1.7358485353472692e-05,
+      "loss": 0.2593,
+      "step": 963
+    },
+    {
+      "epoch": 0.2601187263896384,
+      "grad_norm": 2.1962804794311523,
+      "learning_rate": 1.735256346275071e-05,
+      "loss": 0.3036,
+      "step": 964
+    },
+    {
+      "epoch": 0.2603885590933621,
+      "grad_norm": 0.7615513801574707,
+      "learning_rate": 1.7346635954020717e-05,
+      "loss": 0.2874,
+      "step": 965
+    },
+    {
+      "epoch": 0.26065839179708583,
+      "grad_norm": 0.9293563961982727,
+      "learning_rate": 1.7340702831811857e-05,
+      "loss": 0.2885,
+      "step": 966
+    },
+    {
+      "epoch": 0.2609282245008095,
+      "grad_norm": 1.112978458404541,
+      "learning_rate": 1.7334764100657557e-05,
+      "loss": 0.274,
+      "step": 967
+    },
+    {
+      "epoch": 0.2611980572045332,
+      "grad_norm": 1.8749356269836426,
+      "learning_rate": 1.7328819765095538e-05,
+      "loss": 0.333,
+      "step": 968
+    },
+    {
+      "epoch": 0.26146788990825687,
+      "grad_norm": 3.1324448585510254,
+      "learning_rate": 1.7322869829667788e-05,
+      "loss": 0.3049,
+      "step": 969
+    },
+    {
+      "epoch": 0.2617377226119806,
+      "grad_norm": 1.9288402795791626,
+      "learning_rate": 1.7316914298920592e-05,
+      "loss": 0.2964,
+      "step": 970
+    },
+    {
+      "epoch": 0.26200755531570424,
+      "grad_norm": 0.9363199472427368,
+      "learning_rate": 1.7310953177404494e-05,
+      "loss": 0.2769,
+      "step": 971
+    },
+    {
+      "epoch": 0.26227738801942796,
+      "grad_norm": 3.821183204650879,
+      "learning_rate": 1.7304986469674324e-05,
+      "loss": 0.3189,
+      "step": 972
+    },
+    {
+      "epoch": 0.2625472207231516,
+      "grad_norm": 1.8012478351593018,
+      "learning_rate": 1.7299014180289174e-05,
+      "loss": 0.2882,
+      "step": 973
+    },
+    {
+      "epoch": 0.26281705342687534,
+      "grad_norm": 1.3391838073730469,
+      "learning_rate": 1.7293036313812395e-05,
+      "loss": 0.2603,
+      "step": 974
+    },
+    {
+      "epoch": 0.26308688613059905,
+      "grad_norm": 1.4790602922439575,
+      "learning_rate": 1.7287052874811607e-05,
+      "loss": 0.2864,
+      "step": 975
+    },
+    {
+      "epoch": 0.2633567188343227,
+      "grad_norm": 0.912143886089325,
+      "learning_rate": 1.7281063867858687e-05,
+      "loss": 0.2929,
+      "step": 976
+    },
+    {
+      "epoch": 0.2636265515380464,
+      "grad_norm": 2.150841474533081,
+      "learning_rate": 1.727506929752977e-05,
+      "loss": 0.2602,
+      "step": 977
+    },
+    {
+      "epoch": 0.2638963842417701,
+      "grad_norm": 4.132073402404785,
+      "learning_rate": 1.726906916840523e-05,
+      "loss": 0.3004,
+      "step": 978
+    },
+    {
+      "epoch": 0.2641662169454938,
+      "grad_norm": 0.8109918832778931,
+      "learning_rate": 1.7263063485069697e-05,
+      "loss": 0.2767,
+      "step": 979
+    },
+    {
+      "epoch": 0.26443604964921746,
+      "grad_norm": 1.1391271352767944,
+      "learning_rate": 1.7257052252112045e-05,
+      "loss": 0.2722,
+      "step": 980
+    },
+    {
+      "epoch": 0.2647058823529412,
+      "grad_norm": 2.84055233001709,
+      "learning_rate": 1.7251035474125388e-05,
+      "loss": 0.298,
+      "step": 981
+    },
+    {
+      "epoch": 0.26497571505666484,
+      "grad_norm": 1.3901035785675049,
+      "learning_rate": 1.7245013155707076e-05,
+      "loss": 0.331,
+      "step": 982
+    },
+    {
+      "epoch": 0.26524554776038856,
+      "grad_norm": 1.034536600112915,
+      "learning_rate": 1.7238985301458688e-05,
+      "loss": 0.2815,
+      "step": 983
+    },
+    {
+      "epoch": 0.26551538046411227,
+      "grad_norm": 0.7150821089744568,
+      "learning_rate": 1.7232951915986042e-05,
+      "loss": 0.259,
+      "step": 984
+    },
+    {
+      "epoch": 0.26578521316783593,
+      "grad_norm": 1.944658875465393,
+      "learning_rate": 1.722691300389918e-05,
+      "loss": 0.2778,
+      "step": 985
+    },
+    {
+      "epoch": 0.26605504587155965,
+      "grad_norm": 1.9129546880722046,
+      "learning_rate": 1.722086856981235e-05,
+      "loss": 0.2988,
+      "step": 986
+    },
+    {
+      "epoch": 0.2663248785752833,
+      "grad_norm": 1.0918543338775635,
+      "learning_rate": 1.7214818618344052e-05,
+      "loss": 0.2654,
+      "step": 987
+    },
+    {
+      "epoch": 0.266594711279007,
+      "grad_norm": 0.8844354152679443,
+      "learning_rate": 1.7208763154116973e-05,
+      "loss": 0.3073,
+      "step": 988
+    },
+    {
+      "epoch": 0.2668645439827307,
+      "grad_norm": 1.9242589473724365,
+      "learning_rate": 1.7202702181758027e-05,
+      "loss": 0.2608,
+      "step": 989
+    },
+    {
+      "epoch": 0.2671343766864544,
+      "grad_norm": 0.9070780873298645,
+      "learning_rate": 1.7196635705898325e-05,
+      "loss": 0.2766,
+      "step": 990
+    },
+    {
+      "epoch": 0.2674042093901781,
+      "grad_norm": 1.5097455978393555,
+      "learning_rate": 1.71905637311732e-05,
+      "loss": 0.3381,
+      "step": 991
+    },
+    {
+      "epoch": 0.2676740420939018,
+      "grad_norm": 1.9708846807479858,
+      "learning_rate": 1.718448626222217e-05,
+      "loss": 0.2529,
+      "step": 992
+    },
+    {
+      "epoch": 0.2679438747976255,
+      "grad_norm": 1.1412370204925537,
+      "learning_rate": 1.7178403303688962e-05,
+      "loss": 0.2824,
+      "step": 993
+    },
+    {
+      "epoch": 0.26821370750134915,
+      "grad_norm": 1.6000938415527344,
+      "learning_rate": 1.7172314860221494e-05,
+      "loss": 0.2864,
+      "step": 994
+    },
+    {
+      "epoch": 0.26848354020507287,
+      "grad_norm": 1.3167824745178223,
+      "learning_rate": 1.7166220936471867e-05,
+      "loss": 0.2907,
+      "step": 995
+    },
+    {
+      "epoch": 0.26875337290879653,
+      "grad_norm": 0.9368878602981567,
+      "learning_rate": 1.7160121537096388e-05,
+      "loss": 0.3386,
+      "step": 996
+    },
+    {
+      "epoch": 0.26902320561252024,
+      "grad_norm": 0.8059290647506714,
+      "learning_rate": 1.7154016666755533e-05,
+      "loss": 0.3085,
+      "step": 997
+    },
+    {
+      "epoch": 0.2692930383162439,
+      "grad_norm": 1.0368785858154297,
+      "learning_rate": 1.7147906330113956e-05,
+      "loss": 0.2688,
+      "step": 998
+    },
+    {
+      "epoch": 0.2695628710199676,
+      "grad_norm": 1.774554967880249,
+      "learning_rate": 1.7141790531840507e-05,
+      "loss": 0.2729,
+      "step": 999
+    },
+    {
+      "epoch": 0.26983270372369134,
+      "grad_norm": 0.9203168153762817,
+      "learning_rate": 1.713566927660818e-05,
+      "loss": 0.288,
+      "step": 1000
+    },
+    {
+      "epoch": 0.270102536427415,
+      "grad_norm": 1.8485814332962036,
+      "learning_rate": 1.7129542569094167e-05,
+      "loss": 0.2979,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2703723691311387,
+      "grad_norm": 1.4690961837768555,
+      "learning_rate": 1.7123410413979814e-05,
+      "loss": 0.2866,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2706422018348624,
+      "grad_norm": 1.3313214778900146,
+      "learning_rate": 1.7117272815950622e-05,
+      "loss": 0.28,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2709120345385861,
+      "grad_norm": 2.090742588043213,
+      "learning_rate": 1.711112977969626e-05,
+      "loss": 0.3151,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27118186724230975,
+      "grad_norm": 2.0711236000061035,
+      "learning_rate": 1.7104981309910552e-05,
+      "loss": 0.2868,
+      "step": 1005
+    },
+    {
+      "epoch": 0.27145169994603346,
+      "grad_norm": 0.9303967356681824,
+      "learning_rate": 1.7098827411291474e-05,
+      "loss": 0.285,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2717215326497571,
+      "grad_norm": 1.109462022781372,
+      "learning_rate": 1.7092668088541146e-05,
+      "loss": 0.297,
+      "step": 1007
+    },
+    {
+      "epoch": 0.27199136535348084,
+      "grad_norm": 1.7291449308395386,
+      "learning_rate": 1.708650334636583e-05,
+      "loss": 0.277,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27226119805720456,
+      "grad_norm": 2.20208740234375,
+      "learning_rate": 1.708033318947594e-05,
+      "loss": 0.2732,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2725310307609282,
+      "grad_norm": 1.8430044651031494,
+      "learning_rate": 1.7074157622586024e-05,
+      "loss": 0.2798,
+      "step": 1010
+    },
+    {
+      "epoch": 0.27280086346465193,
+      "grad_norm": 2.777773141860962,
+      "learning_rate": 1.706797665041475e-05,
+      "loss": 0.275,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2730706961683756,
+      "grad_norm": 1.5487868785858154,
+      "learning_rate": 1.7061790277684935e-05,
+      "loss": 0.3094,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2733405288720993,
+      "grad_norm": 2.02055025100708,
+      "learning_rate": 1.7055598509123513e-05,
+      "loss": 0.2501,
+      "step": 1013
+    },
+    {
+      "epoch": 0.27361036157582297,
+      "grad_norm": 2.7567241191864014,
+      "learning_rate": 1.7049401349461538e-05,
+      "loss": 0.2935,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2738801942795467,
+      "grad_norm": 1.0148948431015015,
+      "learning_rate": 1.704319880343419e-05,
+      "loss": 0.2476,
+      "step": 1015
+    },
+    {
+      "epoch": 0.27415002698327035,
+      "grad_norm": 2.8275060653686523,
+      "learning_rate": 1.7036990875780767e-05,
+      "loss": 0.2929,
+      "step": 1016
+    },
+    {
+      "epoch": 0.27441985968699406,
+      "grad_norm": 1.7588363885879517,
+      "learning_rate": 1.7030777571244668e-05,
+      "loss": 0.317,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2746896923907178,
+      "grad_norm": 1.9434889554977417,
+      "learning_rate": 1.7024558894573408e-05,
+      "loss": 0.2904,
+      "step": 1018
+    },
+    {
+      "epoch": 0.27495952509444144,
+      "grad_norm": 3.709277391433716,
+      "learning_rate": 1.7018334850518602e-05,
+      "loss": 0.3072,
+      "step": 1019
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 2.6865744590759277,
+      "learning_rate": 1.7012105443835977e-05,
+      "loss": 0.289,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2754991905018888,
+      "grad_norm": 4.1840691566467285,
+      "learning_rate": 1.7005870679285347e-05,
+      "loss": 0.2913,
+      "step": 1021
+    },
+    {
+      "epoch": 0.27576902320561253,
+      "grad_norm": 4.668893814086914,
+      "learning_rate": 1.6999630561630618e-05,
+      "loss": 0.2943,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2760388559093362,
+      "grad_norm": 0.7585011720657349,
+      "learning_rate": 1.6993385095639796e-05,
+      "loss": 0.2655,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2763086886130599,
+      "grad_norm": 3.005267858505249,
+      "learning_rate": 1.698713428608497e-05,
+      "loss": 0.297,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2765785213167836,
+      "grad_norm": 3.517792224884033,
+      "learning_rate": 1.6980878137742304e-05,
+      "loss": 0.3065,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2768483540205073,
+      "grad_norm": 0.911779522895813,
+      "learning_rate": 1.6974616655392057e-05,
+      "loss": 0.2902,
+      "step": 1026
+    },
+    {
+      "epoch": 0.277118186724231,
+      "grad_norm": 1.4038212299346924,
+      "learning_rate": 1.696834984381855e-05,
+      "loss": 0.2912,
+      "step": 1027
+    },
+    {
+      "epoch": 0.27738801942795466,
+      "grad_norm": 3.1971986293792725,
+      "learning_rate": 1.6962077707810182e-05,
+      "loss": 0.2919,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2776578521316784,
+      "grad_norm": 1.0454946756362915,
+      "learning_rate": 1.6955800252159412e-05,
+      "loss": 0.3301,
+      "step": 1029
+    },
+    {
+      "epoch": 0.27792768483540203,
+      "grad_norm": 1.0592483282089233,
+      "learning_rate": 1.694951748166278e-05,
+      "loss": 0.2627,
+      "step": 1030
+    },
+    {
+      "epoch": 0.27819751753912575,
+      "grad_norm": 3.003589153289795,
+      "learning_rate": 1.6943229401120875e-05,
+      "loss": 0.281,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2784673502428494,
+      "grad_norm": 1.1354464292526245,
+      "learning_rate": 1.6936936015338346e-05,
+      "loss": 0.3135,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2787371829465731,
+      "grad_norm": 2.2453935146331787,
+      "learning_rate": 1.6930637329123892e-05,
+      "loss": 0.3014,
+      "step": 1033
+    },
+    {
+      "epoch": 0.27900701565029684,
+      "grad_norm": 1.2543076276779175,
+      "learning_rate": 1.6924333347290268e-05,
+      "loss": 0.3122,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2792768483540205,
+      "grad_norm": 2.2762584686279297,
+      "learning_rate": 1.6918024074654277e-05,
+      "loss": 0.2785,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2795466810577442,
+      "grad_norm": 1.764620065689087,
+      "learning_rate": 1.6911709516036755e-05,
+      "loss": 0.2573,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2798165137614679,
+      "grad_norm": 0.9675930142402649,
+      "learning_rate": 1.6905389676262583e-05,
+      "loss": 0.2871,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2800863464651916,
+      "grad_norm": 1.0223536491394043,
+      "learning_rate": 1.6899064560160683e-05,
+      "loss": 0.3099,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28035617916891525,
+      "grad_norm": 2.4096648693084717,
+      "learning_rate": 1.6892734172563994e-05,
+      "loss": 0.2659,
+      "step": 1039
+    },
+    {
+      "epoch": 0.28062601187263897,
+      "grad_norm": 1.0753589868545532,
+      "learning_rate": 1.6886398518309495e-05,
+      "loss": 0.2865,
+      "step": 1040
+    },
+    {
+      "epoch": 0.28089584457636263,
+      "grad_norm": 1.2452812194824219,
+      "learning_rate": 1.688005760223819e-05,
+      "loss": 0.2952,
+      "step": 1041
+    },
+    {
+      "epoch": 0.28116567728008635,
+      "grad_norm": 2.601379871368408,
+      "learning_rate": 1.6873711429195095e-05,
+      "loss": 0.2873,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28143550998381006,
+      "grad_norm": 1.3118112087249756,
+      "learning_rate": 1.686736000402925e-05,
+      "loss": 0.2933,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2817053426875337,
+      "grad_norm": 1.6555988788604736,
+      "learning_rate": 1.686100333159369e-05,
+      "loss": 0.2851,
+      "step": 1044
+    },
+    {
+      "epoch": 0.28197517539125744,
+      "grad_norm": 1.369454264640808,
+      "learning_rate": 1.6854641416745497e-05,
+      "loss": 0.2827,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2822450080949811,
+      "grad_norm": 0.9721022844314575,
+      "learning_rate": 1.684827426434572e-05,
+      "loss": 0.28,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2825148407987048,
+      "grad_norm": 0.983004093170166,
+      "learning_rate": 1.6841901879259435e-05,
+      "loss": 0.2895,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2827846735024285,
+      "grad_norm": 3.494870662689209,
+      "learning_rate": 1.6835524266355698e-05,
+      "loss": 0.2944,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2830545062061522,
+      "grad_norm": 1.388425350189209,
+      "learning_rate": 1.682914143050758e-05,
+      "loss": 0.2788,
+      "step": 1049
+    },
+    {
+      "epoch": 0.28332433890987585,
+      "grad_norm": 2.550611734390259,
+      "learning_rate": 1.682275337659212e-05,
+      "loss": 0.2885,
+      "step": 1050
+    },
+    {
+      "epoch": 0.28359417161359957,
+      "grad_norm": 1.8122187852859497,
+      "learning_rate": 1.6816360109490363e-05,
+      "loss": 0.2788,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2838640043173233,
+      "grad_norm": 4.739626407623291,
+      "learning_rate": 1.6809961634087323e-05,
+      "loss": 0.3574,
+      "step": 1052
+    },
+    {
+      "epoch": 0.28413383702104694,
+      "grad_norm": 4.018714427947998,
+      "learning_rate": 1.680355795527201e-05,
+      "loss": 0.2636,
+      "step": 1053
+    },
+    {
+      "epoch": 0.28440366972477066,
+      "grad_norm": 3.9142704010009766,
+      "learning_rate": 1.6797149077937395e-05,
+      "loss": 0.2923,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2846735024284943,
+      "grad_norm": 1.2436619997024536,
+      "learning_rate": 1.6790735006980426e-05,
+      "loss": 0.2937,
+      "step": 1055
+    },
+    {
+      "epoch": 0.28494333513221803,
+      "grad_norm": 1.4356327056884766,
+      "learning_rate": 1.678431574730202e-05,
+      "loss": 0.2666,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2852131678359417,
+      "grad_norm": 4.362776756286621,
+      "learning_rate": 1.6777891303807058e-05,
+      "loss": 0.3007,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2854830005396654,
+      "grad_norm": 2.225059747695923,
+      "learning_rate": 1.6771461681404386e-05,
+      "loss": 0.3034,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2857528332433891,
+      "grad_norm": 3.7078065872192383,
+      "learning_rate": 1.67650268850068e-05,
+      "loss": 0.2624,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2860226659471128,
+      "grad_norm": 2.7021095752716064,
+      "learning_rate": 1.6758586919531054e-05,
+      "loss": 0.3008,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2862924986508365,
+      "grad_norm": 1.0223181247711182,
+      "learning_rate": 1.6752141789897848e-05,
+      "loss": 0.2887,
+      "step": 1061
+    },
+    {
+      "epoch": 0.28656233135456016,
+      "grad_norm": 5.139009475708008,
+      "learning_rate": 1.6745691501031833e-05,
+      "loss": 0.3381,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2868321640582839,
+      "grad_norm": 1.5818520784378052,
+      "learning_rate": 1.6739236057861602e-05,
+      "loss": 0.2821,
+      "step": 1063
+    },
+    {
+      "epoch": 0.28710199676200754,
+      "grad_norm": 1.1995737552642822,
+      "learning_rate": 1.673277546531968e-05,
+      "loss": 0.2616,
+      "step": 1064
+    },
+    {
+      "epoch": 0.28737182946573125,
+      "grad_norm": 2.589512586593628,
+      "learning_rate": 1.672630972834253e-05,
+      "loss": 0.2792,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2876416621694549,
+      "grad_norm": 1.591948390007019,
+      "learning_rate": 1.671983885187055e-05,
+      "loss": 0.2787,
+      "step": 1066
+    },
+    {
+      "epoch": 0.28791149487317863,
+      "grad_norm": 2.2742249965667725,
+      "learning_rate": 1.671336284084806e-05,
+      "loss": 0.2984,
+      "step": 1067
+    },
+    {
+      "epoch": 0.28818132757690235,
+      "grad_norm": 3.4528071880340576,
+      "learning_rate": 1.6706881700223302e-05,
+      "loss": 0.2644,
+      "step": 1068
+    },
+    {
+      "epoch": 0.288451160280626,
+      "grad_norm": 4.148963451385498,
+      "learning_rate": 1.670039543494844e-05,
+      "loss": 0.3257,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2887209929843497,
+      "grad_norm": 1.6394826173782349,
+      "learning_rate": 1.6693904049979556e-05,
+      "loss": 0.2385,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2889908256880734,
+      "grad_norm": 3.297025203704834,
+      "learning_rate": 1.6687407550276643e-05,
+      "loss": 0.2807,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2892606583917971,
+      "grad_norm": 2.022799253463745,
+      "learning_rate": 1.6680905940803596e-05,
+      "loss": 0.2937,
+      "step": 1072
+    },
+    {
+      "epoch": 0.28953049109552076,
+      "grad_norm": 1.9114232063293457,
+      "learning_rate": 1.6674399226528223e-05,
+      "loss": 0.3073,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2898003237992445,
+      "grad_norm": 1.8456062078475952,
+      "learning_rate": 1.6667887412422224e-05,
+      "loss": 0.283,
+      "step": 1074
+    },
+    {
+      "epoch": 0.29007015650296814,
+      "grad_norm": 1.4161187410354614,
+      "learning_rate": 1.6661370503461206e-05,
+      "loss": 0.2806,
+      "step": 1075
+    },
+    {
+      "epoch": 0.29033998920669185,
+      "grad_norm": 3.1105146408081055,
+      "learning_rate": 1.6654848504624655e-05,
+      "loss": 0.2994,
+      "step": 1076
+    },
+    {
+      "epoch": 0.29060982191041557,
+      "grad_norm": 0.8557801246643066,
+      "learning_rate": 1.664832142089597e-05,
+      "loss": 0.2746,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2908796546141392,
+      "grad_norm": 4.628997325897217,
+      "learning_rate": 1.66417892572624e-05,
+      "loss": 0.3421,
+      "step": 1078
+    },
+    {
+      "epoch": 0.29114948731786294,
+      "grad_norm": 0.8681191802024841,
+      "learning_rate": 1.6635252018715107e-05,
+      "loss": 0.2985,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2914193200215866,
+      "grad_norm": 2.7058420181274414,
+      "learning_rate": 1.6628709710249117e-05,
+      "loss": 0.2818,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2916891527253103,
+      "grad_norm": 2.3108468055725098,
+      "learning_rate": 1.662216233686333e-05,
+      "loss": 0.2968,
+      "step": 1081
+    },
+    {
+      "epoch": 0.291958985429034,
+      "grad_norm": 1.1351710557937622,
+      "learning_rate": 1.6615609903560516e-05,
+      "loss": 0.2388,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2922288181327577,
+      "grad_norm": 2.801759958267212,
+      "learning_rate": 1.6609052415347317e-05,
+      "loss": 0.291,
+      "step": 1083
+    },
+    {
+      "epoch": 0.29249865083648136,
+      "grad_norm": 2.699573278427124,
+      "learning_rate": 1.6602489877234235e-05,
+      "loss": 0.3115,
+      "step": 1084
+    },
+    {
+      "epoch": 0.29276848354020507,
+      "grad_norm": 1.2947930097579956,
+      "learning_rate": 1.659592229423562e-05,
+      "loss": 0.3056,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2930383162439288,
+      "grad_norm": 5.561519145965576,
+      "learning_rate": 1.6589349671369693e-05,
+      "loss": 0.2813,
+      "step": 1086
+    },
+    {
+      "epoch": 0.29330814894765245,
+      "grad_norm": 1.7087205648422241,
+      "learning_rate": 1.658277201365851e-05,
+      "loss": 0.2985,
+      "step": 1087
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.9423991441726685,
+      "learning_rate": 1.6576189326128e-05,
+      "loss": 0.3,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2938478143550998,
+      "grad_norm": 3.5517001152038574,
+      "learning_rate": 1.65696016138079e-05,
+      "loss": 0.2986,
+      "step": 1089
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 1.9861531257629395,
+      "learning_rate": 1.656300888173181e-05,
+      "loss": 0.3262,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2943874797625472,
+      "grad_norm": 1.2046643495559692,
+      "learning_rate": 1.655641113493716e-05,
+      "loss": 0.2873,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2946573124662709,
+      "grad_norm": 1.9210561513900757,
+      "learning_rate": 1.6549808378465218e-05,
+      "loss": 0.3077,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2949271451699946,
+      "grad_norm": 3.493391990661621,
+      "learning_rate": 1.6543200617361062e-05,
+      "loss": 0.3044,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2951969778737183,
+      "grad_norm": 1.297688603401184,
+      "learning_rate": 1.653658785667361e-05,
+      "loss": 0.2977,
+      "step": 1094
+    },
+    {
+      "epoch": 0.295466810577442,
+      "grad_norm": 1.1604455709457397,
+      "learning_rate": 1.6529970101455598e-05,
+      "loss": 0.2881,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29573664328116567,
+      "grad_norm": 0.8922545313835144,
+      "learning_rate": 1.6523347356763572e-05,
+      "loss": 0.2839,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2960064759848894,
+      "grad_norm": 2.1038970947265625,
+      "learning_rate": 1.6516719627657898e-05,
+      "loss": 0.2929,
+      "step": 1097
+    },
+    {
+      "epoch": 0.29627630868861304,
+      "grad_norm": 0.7419105172157288,
+      "learning_rate": 1.651008691920274e-05,
+      "loss": 0.2627,
+      "step": 1098
+    },
+    {
+      "epoch": 0.29654614139233676,
+      "grad_norm": 1.9074819087982178,
+      "learning_rate": 1.650344923646608e-05,
+      "loss": 0.2448,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2968159740960604,
+      "grad_norm": 1.221846103668213,
+      "learning_rate": 1.649680658451969e-05,
+      "loss": 0.2901,
+      "step": 1100
+    },
+    {
+      "epoch": 0.29708580679978414,
+      "grad_norm": 1.521204948425293,
+      "learning_rate": 1.6490158968439144e-05,
+      "loss": 0.3153,
+      "step": 1101
+    },
+    {
+      "epoch": 0.29735563950350785,
+      "grad_norm": 1.3695560693740845,
+      "learning_rate": 1.6483506393303807e-05,
+      "loss": 0.3555,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2976254722072315,
+      "grad_norm": 0.9769427180290222,
+      "learning_rate": 1.6476848864196836e-05,
+      "loss": 0.2962,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2978953049109552,
+      "grad_norm": 1.3921418190002441,
+      "learning_rate": 1.6470186386205166e-05,
+      "loss": 0.2811,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2981651376146789,
+      "grad_norm": 0.9750059247016907,
+      "learning_rate": 1.646351896441953e-05,
+      "loss": 0.2942,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2984349703184026,
+      "grad_norm": 1.2545908689498901,
+      "learning_rate": 1.6456846603934414e-05,
+      "loss": 0.2963,
+      "step": 1106
+    },
+    {
+      "epoch": 0.29870480302212626,
+      "grad_norm": 1.1744868755340576,
+      "learning_rate": 1.64501693098481e-05,
+      "loss": 0.2752,
+      "step": 1107
+    },
+    {
+      "epoch": 0.29897463572585,
+      "grad_norm": 0.7558862566947937,
+      "learning_rate": 1.644348708726263e-05,
+      "loss": 0.2586,
+      "step": 1108
+    },
+    {
+      "epoch": 0.29924446842957364,
+      "grad_norm": 0.8458996415138245,
+      "learning_rate": 1.6436799941283806e-05,
+      "loss": 0.2759,
+      "step": 1109
+    },
+    {
+      "epoch": 0.29951430113329736,
+      "grad_norm": 1.9976427555084229,
+      "learning_rate": 1.6430107877021203e-05,
+      "loss": 0.3051,
+      "step": 1110
+    },
+    {
+      "epoch": 0.29978413383702107,
+      "grad_norm": 1.0971769094467163,
+      "learning_rate": 1.6423410899588147e-05,
+      "loss": 0.2937,
+      "step": 1111
+    },
+    {
+      "epoch": 0.30005396654074473,
+      "grad_norm": 1.6530709266662598,
+      "learning_rate": 1.641670901410173e-05,
+      "loss": 0.2897,
+      "step": 1112
+    },
+    {
+      "epoch": 0.30032379924446845,
+      "grad_norm": 1.7303764820098877,
+      "learning_rate": 1.6410002225682767e-05,
+      "loss": 0.2567,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3005936319481921,
+      "grad_norm": 0.7390519976615906,
+      "learning_rate": 1.640329053945585e-05,
+      "loss": 0.2779,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3008634646519158,
+      "grad_norm": 1.4708294868469238,
+      "learning_rate": 1.6396573960549295e-05,
+      "loss": 0.2697,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3011332973556395,
+      "grad_norm": 3.127549886703491,
+      "learning_rate": 1.6389852494095168e-05,
+      "loss": 0.2878,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3014031300593632,
+      "grad_norm": 2.3425137996673584,
+      "learning_rate": 1.638312614522925e-05,
+      "loss": 0.3001,
+      "step": 1117
+    },
+    {
+      "epoch": 0.30167296276308686,
+      "grad_norm": 3.571821451187134,
+      "learning_rate": 1.6376394919091082e-05,
+      "loss": 0.2777,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3019427954668106,
+      "grad_norm": 2.1612343788146973,
+      "learning_rate": 1.6369658820823903e-05,
+      "loss": 0.3084,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3022126281705343,
+      "grad_norm": 1.1699120998382568,
+      "learning_rate": 1.6362917855574694e-05,
+      "loss": 0.2466,
+      "step": 1120
+    },
+    {
+      "epoch": 0.30248246087425795,
+      "grad_norm": 1.727301836013794,
+      "learning_rate": 1.6356172028494142e-05,
+      "loss": 0.2417,
+      "step": 1121
+    },
+    {
+      "epoch": 0.30275229357798167,
+      "grad_norm": 1.0532854795455933,
+      "learning_rate": 1.634942134473666e-05,
+      "loss": 0.357,
+      "step": 1122
+    },
+    {
+      "epoch": 0.30302212628170533,
+      "grad_norm": 1.4044809341430664,
+      "learning_rate": 1.6342665809460364e-05,
+      "loss": 0.2548,
+      "step": 1123
+    },
+    {
+      "epoch": 0.30329195898542904,
+      "grad_norm": 0.9448745250701904,
+      "learning_rate": 1.6335905427827083e-05,
+      "loss": 0.2731,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3035617916891527,
+      "grad_norm": 0.7795705795288086,
+      "learning_rate": 1.632914020500234e-05,
+      "loss": 0.2808,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3038316243928764,
+      "grad_norm": 0.8469337821006775,
+      "learning_rate": 1.6322370146155372e-05,
+      "loss": 0.3193,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3041014570966001,
+      "grad_norm": 1.0282151699066162,
+      "learning_rate": 1.6315595256459093e-05,
+      "loss": 0.2868,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3043712898003238,
+      "grad_norm": 1.089608907699585,
+      "learning_rate": 1.6308815541090122e-05,
+      "loss": 0.3164,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3046411225040475,
+      "grad_norm": 1.0757147073745728,
+      "learning_rate": 1.6302031005228757e-05,
+      "loss": 0.3148,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3049109552077712,
+      "grad_norm": 0.8068697452545166,
+      "learning_rate": 1.6295241654058986e-05,
+      "loss": 0.3192,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3051807879114949,
+      "grad_norm": 1.1698812246322632,
+      "learning_rate": 1.6288447492768475e-05,
+      "loss": 0.2596,
+      "step": 1131
+    },
+    {
+      "epoch": 0.30545062061521855,
+      "grad_norm": 1.3373805284500122,
+      "learning_rate": 1.6281648526548556e-05,
+      "loss": 0.2858,
+      "step": 1132
+    },
+    {
+      "epoch": 0.30572045331894226,
+      "grad_norm": 0.7622947692871094,
+      "learning_rate": 1.6274844760594244e-05,
+      "loss": 0.3025,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3059902860226659,
+      "grad_norm": 3.0909955501556396,
+      "learning_rate": 1.626803620010422e-05,
+      "loss": 0.2936,
+      "step": 1134
+    },
+    {
+      "epoch": 0.30626011872638964,
+      "grad_norm": 1.5170855522155762,
+      "learning_rate": 1.6261222850280824e-05,
+      "loss": 0.3291,
+      "step": 1135
+    },
+    {
+      "epoch": 0.30652995143011336,
+      "grad_norm": 3.5147502422332764,
+      "learning_rate": 1.625440471633005e-05,
+      "loss": 0.2645,
+      "step": 1136
+    },
+    {
+      "epoch": 0.306799784133837,
+      "grad_norm": 3.604261636734009,
+      "learning_rate": 1.6247581803461567e-05,
+      "loss": 0.287,
+      "step": 1137
+    },
+    {
+      "epoch": 0.30706961683756073,
+      "grad_norm": 1.0742744207382202,
+      "learning_rate": 1.6240754116888673e-05,
+      "loss": 0.273,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3073394495412844,
+      "grad_norm": 3.2584402561187744,
+      "learning_rate": 1.623392166182833e-05,
+      "loss": 0.2883,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3076092822450081,
+      "grad_norm": 2.897963047027588,
+      "learning_rate": 1.6227084443501134e-05,
+      "loss": 0.2852,
+      "step": 1140
+    },
+    {
+      "epoch": 0.30787911494873177,
+      "grad_norm": 2.668905735015869,
+      "learning_rate": 1.6220242467131328e-05,
+      "loss": 0.326,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3081489476524555,
+      "grad_norm": 6.1003618240356445,
+      "learning_rate": 1.621339573794678e-05,
+      "loss": 0.2683,
+      "step": 1142
+    },
+    {
+      "epoch": 0.30841878035617915,
+      "grad_norm": 2.4495933055877686,
+      "learning_rate": 1.6206544261179e-05,
+      "loss": 0.2737,
+      "step": 1143
+    },
+    {
+      "epoch": 0.30868861305990286,
+      "grad_norm": 1.2295517921447754,
+      "learning_rate": 1.6199688042063118e-05,
+      "loss": 0.2702,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3089584457636266,
+      "grad_norm": 3.528616428375244,
+      "learning_rate": 1.6192827085837898e-05,
+      "loss": 0.3018,
+      "step": 1145
+    },
+    {
+      "epoch": 0.30922827846735024,
+      "grad_norm": 3.522688150405884,
+      "learning_rate": 1.618596139774571e-05,
+      "loss": 0.2757,
+      "step": 1146
+    },
+    {
+      "epoch": 0.30949811117107395,
+      "grad_norm": 0.7978090047836304,
+      "learning_rate": 1.6179090983032545e-05,
+      "loss": 0.3053,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3097679438747976,
+      "grad_norm": 2.500594139099121,
+      "learning_rate": 1.6172215846948015e-05,
+      "loss": 0.2643,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31003777657852133,
+      "grad_norm": 3.071995258331299,
+      "learning_rate": 1.6165335994745327e-05,
+      "loss": 0.2647,
+      "step": 1149
+    },
+    {
+      "epoch": 0.310307609282245,
+      "grad_norm": 1.6877248287200928,
+      "learning_rate": 1.6158451431681292e-05,
+      "loss": 0.3317,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3105774419859687,
+      "grad_norm": 3.452859401702881,
+      "learning_rate": 1.615156216301633e-05,
+      "loss": 0.318,
+      "step": 1151
+    },
+    {
+      "epoch": 0.31084727468969237,
+      "grad_norm": 3.4556431770324707,
+      "learning_rate": 1.614466819401445e-05,
+      "loss": 0.2941,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3111171073934161,
+      "grad_norm": 2.643317699432373,
+      "learning_rate": 1.613776952994325e-05,
+      "loss": 0.2586,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3113869400971398,
+      "grad_norm": 3.9198124408721924,
+      "learning_rate": 1.613086617607392e-05,
+      "loss": 0.2899,
+      "step": 1154
+    },
+    {
+      "epoch": 0.31165677280086346,
+      "grad_norm": 4.792860507965088,
+      "learning_rate": 1.6123958137681233e-05,
+      "loss": 0.2972,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 1.7262635231018066,
+      "learning_rate": 1.6117045420043545e-05,
+      "loss": 0.2978,
+      "step": 1156
+    },
+    {
+      "epoch": 0.31219643820831083,
+      "grad_norm": 3.2826592922210693,
+      "learning_rate": 1.6110128028442777e-05,
+      "loss": 0.3162,
+      "step": 1157
+    },
+    {
+      "epoch": 0.31246627091203455,
+      "grad_norm": 3.3966078758239746,
+      "learning_rate": 1.6103205968164424e-05,
+      "loss": 0.2888,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3127361036157582,
+      "grad_norm": 4.212197780609131,
+      "learning_rate": 1.609627924449756e-05,
+      "loss": 0.3182,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3130059363194819,
+      "grad_norm": 2.483081340789795,
+      "learning_rate": 1.6089347862734807e-05,
+      "loss": 0.3515,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3132757690232056,
+      "grad_norm": 6.53769063949585,
+      "learning_rate": 1.608241182817236e-05,
+      "loss": 0.3202,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3135456017269293,
+      "grad_norm": 3.147894859313965,
+      "learning_rate": 1.6075471146109957e-05,
+      "loss": 0.2985,
+      "step": 1162
+    },
+    {
+      "epoch": 0.313815434430653,
+      "grad_norm": 1.827409029006958,
+      "learning_rate": 1.6068525821850894e-05,
+      "loss": 0.2854,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3140852671343767,
+      "grad_norm": 5.771702289581299,
+      "learning_rate": 1.606157586070201e-05,
+      "loss": 0.3345,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3143550998381004,
+      "grad_norm": 2.9276320934295654,
+      "learning_rate": 1.605462126797369e-05,
+      "loss": 0.2841,
+      "step": 1165
+    },
+    {
+      "epoch": 0.31462493254182405,
+      "grad_norm": 1.1686172485351562,
+      "learning_rate": 1.604766204897986e-05,
+      "loss": 0.2611,
+      "step": 1166
+    },
+    {
+      "epoch": 0.31489476524554777,
+      "grad_norm": 4.583543300628662,
+      "learning_rate": 1.604069820903798e-05,
+      "loss": 0.2631,
+      "step": 1167
+    },
+    {
+      "epoch": 0.31516459794927143,
+      "grad_norm": 0.8960361480712891,
+      "learning_rate": 1.603372975346903e-05,
+      "loss": 0.2823,
+      "step": 1168
+    },
+    {
+      "epoch": 0.31543443065299515,
+      "grad_norm": 1.6184470653533936,
+      "learning_rate": 1.602675668759753e-05,
+      "loss": 0.2969,
+      "step": 1169
+    },
+    {
+      "epoch": 0.31570426335671886,
+      "grad_norm": 0.8032336831092834,
+      "learning_rate": 1.601977901675152e-05,
+      "loss": 0.2548,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3159740960604425,
+      "grad_norm": 2.202359914779663,
+      "learning_rate": 1.601279674626256e-05,
+      "loss": 0.3243,
+      "step": 1171
+    },
+    {
+      "epoch": 0.31624392876416624,
+      "grad_norm": 1.2234760522842407,
+      "learning_rate": 1.6005809881465714e-05,
+      "loss": 0.2964,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3165137614678899,
+      "grad_norm": 1.2580511569976807,
+      "learning_rate": 1.599881842769957e-05,
+      "loss": 0.3112,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3167835941716136,
+      "grad_norm": 1.2169339656829834,
+      "learning_rate": 1.599182239030621e-05,
+      "loss": 0.2969,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3170534268753373,
+      "grad_norm": 2.3081207275390625,
+      "learning_rate": 1.598482177463123e-05,
+      "loss": 0.2604,
+      "step": 1175
+    },
+    {
+      "epoch": 0.317323259579061,
+      "grad_norm": 1.0179927349090576,
+      "learning_rate": 1.597781658602372e-05,
+      "loss": 0.2436,
+      "step": 1176
+    },
+    {
+      "epoch": 0.31759309228278465,
+      "grad_norm": 1.3726062774658203,
+      "learning_rate": 1.5970806829836257e-05,
+      "loss": 0.2683,
+      "step": 1177
+    },
+    {
+      "epoch": 0.31786292498650837,
+      "grad_norm": 1.0786725282669067,
+      "learning_rate": 1.5963792511424917e-05,
+      "loss": 0.2834,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3181327576902321,
+      "grad_norm": 1.7551662921905518,
+      "learning_rate": 1.595677363614926e-05,
+      "loss": 0.2578,
+      "step": 1179
+    },
+    {
+      "epoch": 0.31840259039395574,
+      "grad_norm": 0.8687933087348938,
+      "learning_rate": 1.594975020937233e-05,
+      "loss": 0.2664,
+      "step": 1180
+    },
+    {
+      "epoch": 0.31867242309767946,
+      "grad_norm": 1.8023662567138672,
+      "learning_rate": 1.5942722236460634e-05,
+      "loss": 0.3163,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3189422558014031,
+      "grad_norm": 1.8106436729431152,
+      "learning_rate": 1.5935689722784175e-05,
+      "loss": 0.2612,
+      "step": 1182
+    },
+    {
+      "epoch": 0.31921208850512683,
+      "grad_norm": 0.8823711276054382,
+      "learning_rate": 1.592865267371641e-05,
+      "loss": 0.2793,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3194819212088505,
+      "grad_norm": 2.831777811050415,
+      "learning_rate": 1.592161109463427e-05,
+      "loss": 0.3319,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3197517539125742,
+      "grad_norm": 0.976580798625946,
+      "learning_rate": 1.591456499091814e-05,
+      "loss": 0.2939,
+      "step": 1185
+    },
+    {
+      "epoch": 0.32002158661629787,
+      "grad_norm": 0.9627212285995483,
+      "learning_rate": 1.590751436795186e-05,
+      "loss": 0.2896,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3202914193200216,
+      "grad_norm": 1.5700981616973877,
+      "learning_rate": 1.5900459231122744e-05,
+      "loss": 0.2703,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3205612520237453,
+      "grad_norm": 1.1212780475616455,
+      "learning_rate": 1.5893399585821522e-05,
+      "loss": 0.2726,
+      "step": 1188
+    },
+    {
+      "epoch": 0.32083108472746896,
+      "grad_norm": 2.0198781490325928,
+      "learning_rate": 1.5886335437442403e-05,
+      "loss": 0.2916,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3211009174311927,
+      "grad_norm": 0.8582085967063904,
+      "learning_rate": 1.587926679138301e-05,
+      "loss": 0.2514,
+      "step": 1190
+    },
+    {
+      "epoch": 0.32137075013491634,
+      "grad_norm": 0.7578006386756897,
+      "learning_rate": 1.5872193653044416e-05,
+      "loss": 0.2907,
+      "step": 1191
+    },
+    {
+      "epoch": 0.32164058283864005,
+      "grad_norm": 0.866920530796051,
+      "learning_rate": 1.5865116027831123e-05,
+      "loss": 0.2701,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3219104155423637,
+      "grad_norm": 0.9974144101142883,
+      "learning_rate": 1.5858033921151064e-05,
+      "loss": 0.2748,
+      "step": 1193
+    },
+    {
+      "epoch": 0.32218024824608743,
+      "grad_norm": 0.998055100440979,
+      "learning_rate": 1.5850947338415595e-05,
+      "loss": 0.3293,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3224500809498111,
+      "grad_norm": 0.9223896861076355,
+      "learning_rate": 1.5843856285039484e-05,
+      "loss": 0.3061,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3227199136535348,
+      "grad_norm": 2.585747480392456,
+      "learning_rate": 1.583676076644093e-05,
+      "loss": 0.3164,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3229897463572585,
+      "grad_norm": 1.225725531578064,
+      "learning_rate": 1.5829660788041533e-05,
+      "loss": 0.2879,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3232595790609822,
+      "grad_norm": 2.0412421226501465,
+      "learning_rate": 1.5822556355266302e-05,
+      "loss": 0.3115,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3235294117647059,
+      "grad_norm": 2.4211738109588623,
+      "learning_rate": 1.5815447473543655e-05,
+      "loss": 0.3003,
+      "step": 1199
+    },
+    {
+      "epoch": 0.32379924446842956,
+      "grad_norm": 1.468252420425415,
+      "learning_rate": 1.5808334148305405e-05,
+      "loss": 0.3238,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3240690771721533,
+      "grad_norm": 2.086937665939331,
+      "learning_rate": 1.5801216384986755e-05,
+      "loss": 0.2892,
+      "step": 1201
+    },
+    {
+      "epoch": 0.32433890987587694,
+      "grad_norm": 1.1095391511917114,
+      "learning_rate": 1.5794094189026314e-05,
+      "loss": 0.2862,
+      "step": 1202
+    },
+    {
+      "epoch": 0.32460874257960065,
+      "grad_norm": 2.0404703617095947,
+      "learning_rate": 1.578696756586606e-05,
+      "loss": 0.3242,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3248785752833243,
+      "grad_norm": 1.0868204832077026,
+      "learning_rate": 1.577983652095137e-05,
+      "loss": 0.2772,
+      "step": 1204
+    },
+    {
+      "epoch": 0.325148407987048,
+      "grad_norm": 1.0613921880722046,
+      "learning_rate": 1.577270105973099e-05,
+      "loss": 0.2842,
+      "step": 1205
+    },
+    {
+      "epoch": 0.32541824069077174,
+      "grad_norm": 0.7662310600280762,
+      "learning_rate": 1.576556118765704e-05,
+      "loss": 0.269,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3256880733944954,
+      "grad_norm": 1.5874828100204468,
+      "learning_rate": 1.575841691018502e-05,
+      "loss": 0.3163,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3259579060982191,
+      "grad_norm": 0.9267633557319641,
+      "learning_rate": 1.5751268232773777e-05,
+      "loss": 0.2723,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3262277388019428,
+      "grad_norm": 2.4684653282165527,
+      "learning_rate": 1.574411516088554e-05,
+      "loss": 0.2539,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3264975715056665,
+      "grad_norm": 1.5678800344467163,
+      "learning_rate": 1.5736957699985887e-05,
+      "loss": 0.2737,
+      "step": 1210
+    },
+    {
+      "epoch": 0.32676740420939016,
+      "grad_norm": 0.733814001083374,
+      "learning_rate": 1.5729795855543754e-05,
+      "loss": 0.2838,
+      "step": 1211
+    },
+    {
+      "epoch": 0.32703723691311387,
+      "grad_norm": 1.2354410886764526,
+      "learning_rate": 1.572262963303142e-05,
+      "loss": 0.2739,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3273070696168376,
+      "grad_norm": 2.8745903968811035,
+      "learning_rate": 1.5715459037924508e-05,
+      "loss": 0.2929,
+      "step": 1213
+    },
+    {
+      "epoch": 0.32757690232056125,
+      "grad_norm": 1.5085886716842651,
+      "learning_rate": 1.570828407570199e-05,
+      "loss": 0.2693,
+      "step": 1214
+    },
+    {
+      "epoch": 0.32784673502428496,
+      "grad_norm": 2.4612183570861816,
+      "learning_rate": 1.5701104751846172e-05,
+      "loss": 0.3067,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3281165677280086,
+      "grad_norm": 0.8595640659332275,
+      "learning_rate": 1.5693921071842688e-05,
+      "loss": 0.3083,
+      "step": 1216
+    },
+    {
+      "epoch": 0.32838640043173234,
+      "grad_norm": 2.0822885036468506,
+      "learning_rate": 1.568673304118051e-05,
+      "loss": 0.2931,
+      "step": 1217
+    },
+    {
+      "epoch": 0.328656233135456,
+      "grad_norm": 0.8887134194374084,
+      "learning_rate": 1.567954066535192e-05,
+      "loss": 0.2731,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3289260658391797,
+      "grad_norm": 1.3438900709152222,
+      "learning_rate": 1.5672343949852535e-05,
+      "loss": 0.2936,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3291958985429034,
+      "grad_norm": 2.140310764312744,
+      "learning_rate": 1.5665142900181278e-05,
+      "loss": 0.2759,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3294657312466271,
+      "grad_norm": 2.1858975887298584,
+      "learning_rate": 1.565793752184039e-05,
+      "loss": 0.2843,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3297355639503508,
+      "grad_norm": 2.2848000526428223,
+      "learning_rate": 1.5650727820335417e-05,
+      "loss": 0.2608,
+      "step": 1222
+    },
+    {
+      "epoch": 0.33000539665407447,
+      "grad_norm": 3.297050952911377,
+      "learning_rate": 1.56435138011752e-05,
+      "loss": 0.2983,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 2.0053343772888184,
+      "learning_rate": 1.56362954698719e-05,
+      "loss": 0.2872,
+      "step": 1224
+    },
+    {
+      "epoch": 0.33054506206152184,
+      "grad_norm": 1.1255773305892944,
+      "learning_rate": 1.562907283194095e-05,
+      "loss": 0.3116,
+      "step": 1225
+    },
+    {
+      "epoch": 0.33081489476524556,
+      "grad_norm": 4.5522942543029785,
+      "learning_rate": 1.5621845892901086e-05,
+      "loss": 0.2894,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3310847274689692,
+      "grad_norm": 2.5927252769470215,
+      "learning_rate": 1.5614614658274328e-05,
+      "loss": 0.2809,
+      "step": 1227
+    },
+    {
+      "epoch": 0.33135456017269294,
+      "grad_norm": 1.453437089920044,
+      "learning_rate": 1.5607379133585978e-05,
+      "loss": 0.2862,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3316243928764166,
+      "grad_norm": 2.8194732666015625,
+      "learning_rate": 1.5600139324364613e-05,
+      "loss": 0.2789,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3318942255801403,
+      "grad_norm": 2.036109685897827,
+      "learning_rate": 1.5592895236142093e-05,
+      "loss": 0.2623,
+      "step": 1230
+    },
+    {
+      "epoch": 0.332164058283864,
+      "grad_norm": 1.9524617195129395,
+      "learning_rate": 1.558564687445354e-05,
+      "loss": 0.3032,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3324338909875877,
+      "grad_norm": 4.607323169708252,
+      "learning_rate": 1.5578394244837337e-05,
+      "loss": 0.2936,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3327037236913114,
+      "grad_norm": 2.728323221206665,
+      "learning_rate": 1.557113735283514e-05,
+      "loss": 0.2847,
+      "step": 1233
+    },
+    {
+      "epoch": 0.33297355639503506,
+      "grad_norm": 0.8903769254684448,
+      "learning_rate": 1.5563876203991856e-05,
+      "loss": 0.2582,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3332433890987588,
+      "grad_norm": 2.8456432819366455,
+      "learning_rate": 1.5556610803855633e-05,
+      "loss": 0.2824,
+      "step": 1235
+    },
+    {
+      "epoch": 0.33351322180248244,
+      "grad_norm": 3.709645986557007,
+      "learning_rate": 1.554934115797789e-05,
+      "loss": 0.297,
+      "step": 1236
+    },
+    {
+      "epoch": 0.33378305450620616,
+      "grad_norm": 0.9881486892700195,
+      "learning_rate": 1.5542067271913278e-05,
+      "loss": 0.3156,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3340528872099298,
+      "grad_norm": 3.4304251670837402,
+      "learning_rate": 1.5534789151219687e-05,
+      "loss": 0.2806,
+      "step": 1238
+    },
+    {
+      "epoch": 0.33432271991365353,
+      "grad_norm": 2.2984488010406494,
+      "learning_rate": 1.5527506801458244e-05,
+      "loss": 0.3142,
+      "step": 1239
+    },
+    {
+      "epoch": 0.33459255261737725,
+      "grad_norm": 1.0510340929031372,
+      "learning_rate": 1.55202202281933e-05,
+      "loss": 0.2852,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3348623853211009,
+      "grad_norm": 2.1932425498962402,
+      "learning_rate": 1.5512929436992454e-05,
+      "loss": 0.2812,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3351322180248246,
+      "grad_norm": 1.7568073272705078,
+      "learning_rate": 1.5505634433426507e-05,
+      "loss": 0.2679,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3354020507285483,
+      "grad_norm": 3.13419508934021,
+      "learning_rate": 1.5498335223069484e-05,
+      "loss": 0.2788,
+      "step": 1243
+    },
+    {
+      "epoch": 0.335671883432272,
+      "grad_norm": 1.052015781402588,
+      "learning_rate": 1.549103181149863e-05,
+      "loss": 0.3062,
+      "step": 1244
+    },
+    {
+      "epoch": 0.33594171613599566,
+      "grad_norm": 4.9879231452941895,
+      "learning_rate": 1.54837242042944e-05,
+      "loss": 0.2992,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3362115488397194,
+      "grad_norm": 3.466933488845825,
+      "learning_rate": 1.5476412407040445e-05,
+      "loss": 0.2951,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3364813815434431,
+      "grad_norm": 1.5323662757873535,
+      "learning_rate": 1.5469096425323625e-05,
+      "loss": 0.3067,
+      "step": 1247
+    },
+    {
+      "epoch": 0.33675121424716675,
+      "grad_norm": 2.792513132095337,
+      "learning_rate": 1.5461776264734e-05,
+      "loss": 0.2787,
+      "step": 1248
+    },
+    {
+      "epoch": 0.33702104695089047,
+      "grad_norm": 3.5921523571014404,
+      "learning_rate": 1.5454451930864816e-05,
+      "loss": 0.283,
+      "step": 1249
+    },
+    {
+      "epoch": 0.33729087965461413,
+      "grad_norm": 2.660120964050293,
+      "learning_rate": 1.5447123429312516e-05,
+      "loss": 0.3069,
+      "step": 1250
+    },
+    {
+      "epoch": 0.33756071235833784,
+      "grad_norm": 3.9040002822875977,
+      "learning_rate": 1.5439790765676717e-05,
+      "loss": 0.2993,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3378305450620615,
+      "grad_norm": 3.4792227745056152,
+      "learning_rate": 1.5432453945560223e-05,
+      "loss": 0.3089,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3381003777657852,
+      "grad_norm": 2.7260146141052246,
+      "learning_rate": 1.5425112974569016e-05,
+      "loss": 0.3252,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3383702104695089,
+      "grad_norm": 1.5019865036010742,
+      "learning_rate": 1.541776785831224e-05,
+      "loss": 0.305,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3386400431732326,
+      "grad_norm": 2.7494187355041504,
+      "learning_rate": 1.541041860240222e-05,
+      "loss": 0.2662,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3389098758769563,
+      "grad_norm": 4.726777076721191,
+      "learning_rate": 1.5403065212454433e-05,
+      "loss": 0.3109,
+      "step": 1256
+    },
+    {
+      "epoch": 0.33917970858068,
+      "grad_norm": 0.9377835392951965,
+      "learning_rate": 1.539570769408752e-05,
+      "loss": 0.2676,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3394495412844037,
+      "grad_norm": 2.1195931434631348,
+      "learning_rate": 1.5388346052923268e-05,
+      "loss": 0.2861,
+      "step": 1258
+    },
+    {
+      "epoch": 0.33971937398812735,
+      "grad_norm": 3.8360671997070312,
+      "learning_rate": 1.538098029458663e-05,
+      "loss": 0.2859,
+      "step": 1259
+    },
+    {
+      "epoch": 0.33998920669185106,
+      "grad_norm": 1.0446707010269165,
+      "learning_rate": 1.5373610424705696e-05,
+      "loss": 0.3246,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3402590393955747,
+      "grad_norm": 1.5027743577957153,
+      "learning_rate": 1.536623644891169e-05,
+      "loss": 0.2721,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34052887209929844,
+      "grad_norm": 1.8928422927856445,
+      "learning_rate": 1.5358858372838992e-05,
+      "loss": 0.2959,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3407987048030221,
+      "grad_norm": 1.9361720085144043,
+      "learning_rate": 1.5351476202125092e-05,
+      "loss": 0.3078,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3410685375067458,
+      "grad_norm": 0.9238438606262207,
+      "learning_rate": 1.534408994241063e-05,
+      "loss": 0.3265,
+      "step": 1264
+    },
+    {
+      "epoch": 0.34133837021046953,
+      "grad_norm": 1.05008065700531,
+      "learning_rate": 1.5336699599339357e-05,
+      "loss": 0.2606,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3416082029141932,
+      "grad_norm": 1.1025454998016357,
+      "learning_rate": 1.5329305178558143e-05,
+      "loss": 0.3167,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3418780356179169,
+      "grad_norm": 1.2295783758163452,
+      "learning_rate": 1.5321906685716986e-05,
+      "loss": 0.2999,
+      "step": 1267
+    },
+    {
+      "epoch": 0.34214786832164057,
+      "grad_norm": 1.9748789072036743,
+      "learning_rate": 1.5314504126468986e-05,
+      "loss": 0.3303,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3424177010253643,
+      "grad_norm": 1.2329014539718628,
+      "learning_rate": 1.5307097506470354e-05,
+      "loss": 0.2886,
+      "step": 1269
+    },
+    {
+      "epoch": 0.34268753372908795,
+      "grad_norm": 2.0341386795043945,
+      "learning_rate": 1.5299686831380395e-05,
+      "loss": 0.2596,
+      "step": 1270
+    },
+    {
+      "epoch": 0.34295736643281166,
+      "grad_norm": 1.0573580265045166,
+      "learning_rate": 1.5292272106861527e-05,
+      "loss": 0.2684,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3432271991365353,
+      "grad_norm": 2.3345043659210205,
+      "learning_rate": 1.5284853338579254e-05,
+      "loss": 0.2791,
+      "step": 1272
+    },
+    {
+      "epoch": 0.34349703184025904,
+      "grad_norm": 2.819807529449463,
+      "learning_rate": 1.5277430532202167e-05,
+      "loss": 0.2834,
+      "step": 1273
+    },
+    {
+      "epoch": 0.34376686454398275,
+      "grad_norm": 0.967500627040863,
+      "learning_rate": 1.5270003693401946e-05,
+      "loss": 0.2831,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3440366972477064,
+      "grad_norm": 2.777731418609619,
+      "learning_rate": 1.5262572827853355e-05,
+      "loss": 0.3091,
+      "step": 1275
+    },
+    {
+      "epoch": 0.34430652995143013,
+      "grad_norm": 0.7706719040870667,
+      "learning_rate": 1.5255137941234228e-05,
+      "loss": 0.2978,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3445763626551538,
+      "grad_norm": 1.6521650552749634,
+      "learning_rate": 1.524769903922548e-05,
+      "loss": 0.2843,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3448461953588775,
+      "grad_norm": 1.362475872039795,
+      "learning_rate": 1.5240256127511081e-05,
+      "loss": 0.2991,
+      "step": 1278
+    },
+    {
+      "epoch": 0.34511602806260117,
+      "grad_norm": 1.4100948572158813,
+      "learning_rate": 1.5232809211778082e-05,
+      "loss": 0.2668,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3453858607663249,
+      "grad_norm": 2.592756748199463,
+      "learning_rate": 1.5225358297716582e-05,
+      "loss": 0.273,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3456556934700486,
+      "grad_norm": 1.6819162368774414,
+      "learning_rate": 1.5217903391019733e-05,
+      "loss": 0.2924,
+      "step": 1281
+    },
+    {
+      "epoch": 0.34592552617377226,
+      "grad_norm": 1.564993143081665,
+      "learning_rate": 1.5210444497383745e-05,
+      "loss": 0.2894,
+      "step": 1282
+    },
+    {
+      "epoch": 0.346195358877496,
+      "grad_norm": 2.3107454776763916,
+      "learning_rate": 1.5202981622507875e-05,
+      "loss": 0.3428,
+      "step": 1283
+    },
+    {
+      "epoch": 0.34646519158121963,
+      "grad_norm": 2.376851797103882,
+      "learning_rate": 1.5195514772094413e-05,
+      "loss": 0.2696,
+      "step": 1284
+    },
+    {
+      "epoch": 0.34673502428494335,
+      "grad_norm": 1.0994371175765991,
+      "learning_rate": 1.5188043951848699e-05,
+      "loss": 0.2864,
+      "step": 1285
+    },
+    {
+      "epoch": 0.347004856988667,
+      "grad_norm": 1.234992504119873,
+      "learning_rate": 1.5180569167479093e-05,
+      "loss": 0.2677,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3472746896923907,
+      "grad_norm": 3.1713850498199463,
+      "learning_rate": 1.5173090424696995e-05,
+      "loss": 0.3186,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3475445223961144,
+      "grad_norm": 1.1779358386993408,
+      "learning_rate": 1.5165607729216822e-05,
+      "loss": 0.3113,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3478143550998381,
+      "grad_norm": 0.7678874135017395,
+      "learning_rate": 1.5158121086756017e-05,
+      "loss": 0.2639,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3480841878035618,
+      "grad_norm": 3.6494147777557373,
+      "learning_rate": 1.5150630503035033e-05,
+      "loss": 0.2935,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3483540205072855,
+      "grad_norm": 1.0797173976898193,
+      "learning_rate": 1.5143135983777341e-05,
+      "loss": 0.3082,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.8066907525062561,
+      "learning_rate": 1.5135637534709417e-05,
+      "loss": 0.2856,
+      "step": 1292
+    },
+    {
+      "epoch": 0.34889368591473285,
+      "grad_norm": 3.473174810409546,
+      "learning_rate": 1.5128135161560736e-05,
+      "loss": 0.3136,
+      "step": 1293
+    },
+    {
+      "epoch": 0.34916351861845657,
+      "grad_norm": 1.4678212404251099,
+      "learning_rate": 1.5120628870063772e-05,
+      "loss": 0.2717,
+      "step": 1294
+    },
+    {
+      "epoch": 0.34943335132218023,
+      "grad_norm": 0.9784295558929443,
+      "learning_rate": 1.5113118665953997e-05,
+      "loss": 0.2711,
+      "step": 1295
+    },
+    {
+      "epoch": 0.34970318402590395,
+      "grad_norm": 1.3884719610214233,
+      "learning_rate": 1.5105604554969869e-05,
+      "loss": 0.2775,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3499730167296276,
+      "grad_norm": 2.100548505783081,
+      "learning_rate": 1.5098086542852835e-05,
+      "loss": 0.2577,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3502428494333513,
+      "grad_norm": 0.7610272765159607,
+      "learning_rate": 1.509056463534732e-05,
+      "loss": 0.2697,
+      "step": 1298
+    },
+    {
+      "epoch": 0.35051268213707504,
+      "grad_norm": 2.457855224609375,
+      "learning_rate": 1.5083038838200725e-05,
+      "loss": 0.3147,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3507825148407987,
+      "grad_norm": 1.0059278011322021,
+      "learning_rate": 1.5075509157163422e-05,
+      "loss": 0.2742,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3510523475445224,
+      "grad_norm": 0.7582268714904785,
+      "learning_rate": 1.5067975597988756e-05,
+      "loss": 0.2562,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3513221802482461,
+      "grad_norm": 1.1288223266601562,
+      "learning_rate": 1.5060438166433024e-05,
+      "loss": 0.2983,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3515920129519698,
+      "grad_norm": 2.3228025436401367,
+      "learning_rate": 1.5052896868255504e-05,
+      "loss": 0.2633,
+      "step": 1303
+    },
+    {
+      "epoch": 0.35186184565569345,
+      "grad_norm": 0.9308859705924988,
+      "learning_rate": 1.50453517092184e-05,
+      "loss": 0.2656,
+      "step": 1304
+    },
+    {
+      "epoch": 0.35213167835941717,
+      "grad_norm": 1.9447647333145142,
+      "learning_rate": 1.5037802695086887e-05,
+      "loss": 0.2979,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3524015110631408,
+      "grad_norm": 1.147905945777893,
+      "learning_rate": 1.503024983162908e-05,
+      "loss": 0.2793,
+      "step": 1306
+    },
+    {
+      "epoch": 0.35267134376686454,
+      "grad_norm": 1.0662331581115723,
+      "learning_rate": 1.502269312461603e-05,
+      "loss": 0.3219,
+      "step": 1307
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.9681627154350281,
+      "learning_rate": 1.5015132579821727e-05,
+      "loss": 0.2677,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3532110091743119,
+      "grad_norm": 0.8060582280158997,
+      "learning_rate": 1.5007568203023102e-05,
+      "loss": 0.2873,
+      "step": 1309
+    },
+    {
+      "epoch": 0.35348084187803563,
+      "grad_norm": 1.0703715085983276,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.2705,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3537506745817593,
+      "grad_norm": 0.8657724261283875,
+      "learning_rate": 1.4992427976535202e-05,
+      "loss": 0.2784,
+      "step": 1311
+    },
+    {
+      "epoch": 0.354020507285483,
+      "grad_norm": 1.1321289539337158,
+      "learning_rate": 1.49848521384144e-05,
+      "loss": 0.2666,
+      "step": 1312
+    },
+    {
+      "epoch": 0.35429033998920667,
+      "grad_norm": 1.280087947845459,
+      "learning_rate": 1.4977272491426205e-05,
+      "loss": 0.2762,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3545601726929304,
+      "grad_norm": 0.7937597036361694,
+      "learning_rate": 1.4969689041362132e-05,
+      "loss": 0.2582,
+      "step": 1314
+    },
+    {
+      "epoch": 0.35483000539665405,
+      "grad_norm": 1.7915090322494507,
+      "learning_rate": 1.4962101794016616e-05,
+      "loss": 0.2719,
+      "step": 1315
+    },
+    {
+      "epoch": 0.35509983810037776,
+      "grad_norm": 1.479893684387207,
+      "learning_rate": 1.4954510755186977e-05,
+      "loss": 0.2835,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3553696708041015,
+      "grad_norm": 2.5772297382354736,
+      "learning_rate": 1.494691593067344e-05,
+      "loss": 0.309,
+      "step": 1317
+    },
+    {
+      "epoch": 0.35563950350782514,
+      "grad_norm": 1.1295808553695679,
+      "learning_rate": 1.4939317326279125e-05,
+      "loss": 0.272,
+      "step": 1318
+    },
+    {
+      "epoch": 0.35590933621154885,
+      "grad_norm": 0.9618966579437256,
+      "learning_rate": 1.4931714947810034e-05,
+      "loss": 0.3102,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3561791689152725,
+      "grad_norm": 1.7224066257476807,
+      "learning_rate": 1.4924108801075057e-05,
+      "loss": 0.2918,
+      "step": 1320
+    },
+    {
+      "epoch": 0.35644900161899623,
+      "grad_norm": 1.9361594915390015,
+      "learning_rate": 1.4916498891885964e-05,
+      "loss": 0.2773,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3567188343227199,
+      "grad_norm": 0.8640527725219727,
+      "learning_rate": 1.4908885226057392e-05,
+      "loss": 0.2847,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3569886670264436,
+      "grad_norm": 1.8357332944869995,
+      "learning_rate": 1.490126780940686e-05,
+      "loss": 0.2809,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3572584997301673,
+      "grad_norm": 1.5443099737167358,
+      "learning_rate": 1.489364664775475e-05,
+      "loss": 0.2844,
+      "step": 1324
+    },
+    {
+      "epoch": 0.357528332433891,
+      "grad_norm": 1.424877643585205,
+      "learning_rate": 1.4886021746924294e-05,
+      "loss": 0.2671,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3577981651376147,
+      "grad_norm": 0.8965582251548767,
+      "learning_rate": 1.4878393112741592e-05,
+      "loss": 0.2879,
+      "step": 1326
+    },
+    {
+      "epoch": 0.35806799784133836,
+      "grad_norm": 2.2153196334838867,
+      "learning_rate": 1.48707607510356e-05,
+      "loss": 0.2766,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3583378305450621,
+      "grad_norm": 1.2405903339385986,
+      "learning_rate": 1.4863124667638109e-05,
+      "loss": 0.2781,
+      "step": 1328
+    },
+    {
+      "epoch": 0.35860766324878574,
+      "grad_norm": 1.3171685934066772,
+      "learning_rate": 1.485548486838377e-05,
+      "loss": 0.2847,
+      "step": 1329
+    },
+    {
+      "epoch": 0.35887749595250945,
+      "grad_norm": 0.9232844710350037,
+      "learning_rate": 1.4847841359110058e-05,
+      "loss": 0.2656,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3591473286562331,
+      "grad_norm": 1.056105613708496,
+      "learning_rate": 1.4840194145657287e-05,
+      "loss": 0.2552,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3594171613599568,
+      "grad_norm": 0.7128060460090637,
+      "learning_rate": 1.4832543233868616e-05,
+      "loss": 0.2875,
+      "step": 1332
+    },
+    {
+      "epoch": 0.35968699406368054,
+      "grad_norm": 1.2447583675384521,
+      "learning_rate": 1.4824888629590001e-05,
+      "loss": 0.2903,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3599568267674042,
+      "grad_norm": 1.2196476459503174,
+      "learning_rate": 1.4817230338670247e-05,
+      "loss": 0.3024,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3602266594711279,
+      "grad_norm": 1.506929636001587,
+      "learning_rate": 1.4809568366960965e-05,
+      "loss": 0.2712,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3604964921748516,
+      "grad_norm": 1.284228801727295,
+      "learning_rate": 1.480190272031657e-05,
+      "loss": 0.3193,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3607663248785753,
+      "grad_norm": 1.0691475868225098,
+      "learning_rate": 1.47942334045943e-05,
+      "loss": 0.2941,
+      "step": 1337
+    },
+    {
+      "epoch": 0.36103615758229896,
+      "grad_norm": 1.821616530418396,
+      "learning_rate": 1.478656042565419e-05,
+      "loss": 0.3,
+      "step": 1338
+    },
+    {
+      "epoch": 0.36130599028602267,
+      "grad_norm": 2.008531093597412,
+      "learning_rate": 1.4778883789359069e-05,
+      "loss": 0.2642,
+      "step": 1339
+    },
+    {
+      "epoch": 0.36157582298974633,
+      "grad_norm": 0.7319455742835999,
+      "learning_rate": 1.477120350157457e-05,
+      "loss": 0.2706,
+      "step": 1340
+    },
+    {
+      "epoch": 0.36184565569347005,
+      "grad_norm": 1.6054025888442993,
+      "learning_rate": 1.4763519568169112e-05,
+      "loss": 0.2584,
+      "step": 1341
+    },
+    {
+      "epoch": 0.36211548839719376,
+      "grad_norm": 0.8019745945930481,
+      "learning_rate": 1.475583199501389e-05,
+      "loss": 0.2412,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3623853211009174,
+      "grad_norm": 3.180929183959961,
+      "learning_rate": 1.4748140787982899e-05,
+      "loss": 0.2766,
+      "step": 1343
+    },
+    {
+      "epoch": 0.36265515380464114,
+      "grad_norm": 1.2832595109939575,
+      "learning_rate": 1.4740445952952898e-05,
+      "loss": 0.3072,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3629249865083648,
+      "grad_norm": 1.8657560348510742,
+      "learning_rate": 1.4732747495803416e-05,
+      "loss": 0.3032,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3631948192120885,
+      "grad_norm": 3.480280876159668,
+      "learning_rate": 1.4725045422416759e-05,
+      "loss": 0.2847,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3634646519158122,
+      "grad_norm": 1.155091404914856,
+      "learning_rate": 1.4717339738677994e-05,
+      "loss": 0.2687,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3637344846195359,
+      "grad_norm": 1.27252197265625,
+      "learning_rate": 1.4709630450474936e-05,
+      "loss": 0.2767,
+      "step": 1348
+    },
+    {
+      "epoch": 0.36400431732325955,
+      "grad_norm": 3.807859420776367,
+      "learning_rate": 1.4701917563698169e-05,
+      "loss": 0.3006,
+      "step": 1349
+    },
+    {
+      "epoch": 0.36427415002698327,
+      "grad_norm": 1.3168023824691772,
+      "learning_rate": 1.469420108424102e-05,
+      "loss": 0.2835,
+      "step": 1350
+    },
+    {
+      "epoch": 0.364543982730707,
+      "grad_norm": 0.7568921446800232,
+      "learning_rate": 1.4686481017999554e-05,
+      "loss": 0.3047,
+      "step": 1351
+    },
+    {
+      "epoch": 0.36481381543443064,
+      "grad_norm": 4.224237442016602,
+      "learning_rate": 1.4678757370872591e-05,
+      "loss": 0.2758,
+      "step": 1352
+    },
+    {
+      "epoch": 0.36508364813815436,
+      "grad_norm": 2.793552875518799,
+      "learning_rate": 1.4671030148761675e-05,
+      "loss": 0.2712,
+      "step": 1353
+    },
+    {
+      "epoch": 0.365353480841878,
+      "grad_norm": 1.3387013673782349,
+      "learning_rate": 1.466329935757109e-05,
+      "loss": 0.2662,
+      "step": 1354
+    },
+    {
+      "epoch": 0.36562331354560174,
+      "grad_norm": 2.5022404193878174,
+      "learning_rate": 1.465556500320784e-05,
+      "loss": 0.2618,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3658931462493254,
+      "grad_norm": 3.273979663848877,
+      "learning_rate": 1.4647827091581659e-05,
+      "loss": 0.2756,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3661629789530491,
+      "grad_norm": 1.1062389612197876,
+      "learning_rate": 1.4640085628604991e-05,
+      "loss": 0.2807,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3664328116567728,
+      "grad_norm": 0.9365298748016357,
+      "learning_rate": 1.4632340620192999e-05,
+      "loss": 0.3142,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3667026443604965,
+      "grad_norm": 4.8092570304870605,
+      "learning_rate": 1.4624592072263554e-05,
+      "loss": 0.2755,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 3.4967336654663086,
+      "learning_rate": 1.4616839990737232e-05,
+      "loss": 0.2824,
+      "step": 1360
+    },
+    {
+      "epoch": 0.36724230976794386,
+      "grad_norm": 1.8004776239395142,
+      "learning_rate": 1.4609084381537306e-05,
+      "loss": 0.2685,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3675121424716676,
+      "grad_norm": 3.2120912075042725,
+      "learning_rate": 1.4601325250589748e-05,
+      "loss": 0.3012,
+      "step": 1362
+    },
+    {
+      "epoch": 0.36778197517539124,
+      "grad_norm": 3.696263074874878,
+      "learning_rate": 1.4593562603823221e-05,
+      "loss": 0.2897,
+      "step": 1363
+    },
+    {
+      "epoch": 0.36805180787911496,
+      "grad_norm": 1.7866644859313965,
+      "learning_rate": 1.458579644716907e-05,
+      "loss": 0.304,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3683216405828386,
+      "grad_norm": 2.076159715652466,
+      "learning_rate": 1.4578026786561323e-05,
+      "loss": 0.3069,
+      "step": 1365
+    },
+    {
+      "epoch": 0.36859147328656233,
+      "grad_norm": 5.294987201690674,
+      "learning_rate": 1.4570253627936693e-05,
+      "loss": 0.3236,
+      "step": 1366
+    },
+    {
+      "epoch": 0.36886130599028605,
+      "grad_norm": 3.591904640197754,
+      "learning_rate": 1.4562476977234557e-05,
+      "loss": 0.2539,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3691311386940097,
+      "grad_norm": 1.596648097038269,
+      "learning_rate": 1.455469684039696e-05,
+      "loss": 0.264,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3694009713977334,
+      "grad_norm": 4.913113594055176,
+      "learning_rate": 1.4546913223368618e-05,
+      "loss": 0.3105,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3696708041014571,
+      "grad_norm": 2.177978992462158,
+      "learning_rate": 1.4539126132096901e-05,
+      "loss": 0.3011,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3699406368051808,
+      "grad_norm": 1.1363049745559692,
+      "learning_rate": 1.4531335572531836e-05,
+      "loss": 0.2359,
+      "step": 1371
+    },
+    {
+      "epoch": 0.37021046950890446,
+      "grad_norm": 3.0824575424194336,
+      "learning_rate": 1.4523541550626093e-05,
+      "loss": 0.2556,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3704803022126282,
+      "grad_norm": 1.0375487804412842,
+      "learning_rate": 1.4515744072334999e-05,
+      "loss": 0.2944,
+      "step": 1373
+    },
+    {
+      "epoch": 0.37075013491635184,
+      "grad_norm": 1.5148178339004517,
+      "learning_rate": 1.4507943143616516e-05,
+      "loss": 0.3046,
+      "step": 1374
+    },
+    {
+      "epoch": 0.37101996762007555,
+      "grad_norm": 0.9822885990142822,
+      "learning_rate": 1.4500138770431237e-05,
+      "loss": 0.2796,
+      "step": 1375
+    },
+    {
+      "epoch": 0.37128980032379927,
+      "grad_norm": 0.7670736908912659,
+      "learning_rate": 1.4492330958742402e-05,
+      "loss": 0.2609,
+      "step": 1376
+    },
+    {
+      "epoch": 0.37155963302752293,
+      "grad_norm": 1.8025751113891602,
+      "learning_rate": 1.4484519714515861e-05,
+      "loss": 0.3001,
+      "step": 1377
+    },
+    {
+      "epoch": 0.37182946573124664,
+      "grad_norm": 1.6635991334915161,
+      "learning_rate": 1.4476705043720099e-05,
+      "loss": 0.3057,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3720992984349703,
+      "grad_norm": 1.1496137380599976,
+      "learning_rate": 1.4468886952326212e-05,
+      "loss": 0.2536,
+      "step": 1379
+    },
+    {
+      "epoch": 0.372369131138694,
+      "grad_norm": 1.903809666633606,
+      "learning_rate": 1.4461065446307911e-05,
+      "loss": 0.3077,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3726389638424177,
+      "grad_norm": 1.1117746829986572,
+      "learning_rate": 1.4453240531641524e-05,
+      "loss": 0.2643,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3729087965461414,
+      "grad_norm": 1.7307173013687134,
+      "learning_rate": 1.4445412214305971e-05,
+      "loss": 0.2479,
+      "step": 1382
+    },
+    {
+      "epoch": 0.37317862924986506,
+      "grad_norm": 2.0965075492858887,
+      "learning_rate": 1.4437580500282779e-05,
+      "loss": 0.3273,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3734484619535888,
+      "grad_norm": 2.49078106880188,
+      "learning_rate": 1.4429745395556073e-05,
+      "loss": 0.2853,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3737182946573125,
+      "grad_norm": 1.8522762060165405,
+      "learning_rate": 1.4421906906112563e-05,
+      "loss": 0.2944,
+      "step": 1385
+    },
+    {
+      "epoch": 0.37398812736103615,
+      "grad_norm": 1.7792532444000244,
+      "learning_rate": 1.4414065037941547e-05,
+      "loss": 0.2984,
+      "step": 1386
+    },
+    {
+      "epoch": 0.37425796006475986,
+      "grad_norm": 3.347324848175049,
+      "learning_rate": 1.4406219797034901e-05,
+      "loss": 0.2719,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3745277927684835,
+      "grad_norm": 1.7051727771759033,
+      "learning_rate": 1.439837118938709e-05,
+      "loss": 0.3142,
+      "step": 1388
+    },
+    {
+      "epoch": 0.37479762547220724,
+      "grad_norm": 1.9594447612762451,
+      "learning_rate": 1.439051922099514e-05,
+      "loss": 0.2457,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3750674581759309,
+      "grad_norm": 4.55794095993042,
+      "learning_rate": 1.4382663897858647e-05,
+      "loss": 0.2913,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3753372908796546,
+      "grad_norm": 0.9579979181289673,
+      "learning_rate": 1.4374805225979773e-05,
+      "loss": 0.2468,
+      "step": 1391
+    },
+    {
+      "epoch": 0.37560712358337833,
+      "grad_norm": 2.1451096534729004,
+      "learning_rate": 1.436694321136324e-05,
+      "loss": 0.2845,
+      "step": 1392
+    },
+    {
+      "epoch": 0.375876956287102,
+      "grad_norm": 2.696444511413574,
+      "learning_rate": 1.4359077860016316e-05,
+      "loss": 0.2974,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3761467889908257,
+      "grad_norm": 2.354314088821411,
+      "learning_rate": 1.4351209177948829e-05,
+      "loss": 0.2652,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37641662169454937,
+      "grad_norm": 2.007080554962158,
+      "learning_rate": 1.4343337171173146e-05,
+      "loss": 0.3155,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3766864543982731,
+      "grad_norm": 4.615767955780029,
+      "learning_rate": 1.4335461845704173e-05,
+      "loss": 0.3317,
+      "step": 1396
+    },
+    {
+      "epoch": 0.37695628710199675,
+      "grad_norm": 1.1148803234100342,
+      "learning_rate": 1.432758320755936e-05,
+      "loss": 0.2416,
+      "step": 1397
+    },
+    {
+      "epoch": 0.37722611980572046,
+      "grad_norm": 2.858940362930298,
+      "learning_rate": 1.4319701262758679e-05,
+      "loss": 0.2746,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3774959525094441,
+      "grad_norm": 2.4148190021514893,
+      "learning_rate": 1.4311816017324632e-05,
+      "loss": 0.271,
+      "step": 1399
+    },
+    {
+      "epoch": 0.37776578521316784,
+      "grad_norm": 1.043639063835144,
+      "learning_rate": 1.4303927477282242e-05,
+      "loss": 0.2607,
+      "step": 1400
+    },
+    {
+      "epoch": 0.37803561791689155,
+      "grad_norm": 1.8447080850601196,
+      "learning_rate": 1.4296035648659055e-05,
+      "loss": 0.3114,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3783054506206152,
+      "grad_norm": 0.7396567463874817,
+      "learning_rate": 1.428814053748512e-05,
+      "loss": 0.2544,
+      "step": 1402
+    },
+    {
+      "epoch": 0.37857528332433893,
+      "grad_norm": 0.8814095854759216,
+      "learning_rate": 1.4280242149793e-05,
+      "loss": 0.281,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3788451160280626,
+      "grad_norm": 1.2064263820648193,
+      "learning_rate": 1.4272340491617762e-05,
+      "loss": 0.2832,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3791149487317863,
+      "grad_norm": 1.2073777914047241,
+      "learning_rate": 1.4264435568996977e-05,
+      "loss": 0.2464,
+      "step": 1405
+    },
+    {
+      "epoch": 0.37938478143550997,
+      "grad_norm": 1.0936381816864014,
+      "learning_rate": 1.4256527387970694e-05,
+      "loss": 0.2879,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3796546141392337,
+      "grad_norm": 2.090101957321167,
+      "learning_rate": 1.4248615954581466e-05,
+      "loss": 0.2871,
+      "step": 1407
+    },
+    {
+      "epoch": 0.37992444684295734,
+      "grad_norm": 1.4196090698242188,
+      "learning_rate": 1.4240701274874331e-05,
+      "loss": 0.2649,
+      "step": 1408
+    },
+    {
+      "epoch": 0.38019427954668106,
+      "grad_norm": 0.7955494523048401,
+      "learning_rate": 1.42327833548968e-05,
+      "loss": 0.2856,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3804641122504048,
+      "grad_norm": 2.783764600753784,
+      "learning_rate": 1.4224862200698857e-05,
+      "loss": 0.2845,
+      "step": 1410
+    },
+    {
+      "epoch": 0.38073394495412843,
+      "grad_norm": 0.7839375734329224,
+      "learning_rate": 1.4216937818332976e-05,
+      "loss": 0.2951,
+      "step": 1411
+    },
+    {
+      "epoch": 0.38100377765785215,
+      "grad_norm": 0.8038012981414795,
+      "learning_rate": 1.4209010213854079e-05,
+      "loss": 0.3105,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3812736103615758,
+      "grad_norm": 0.9589053392410278,
+      "learning_rate": 1.4201079393319552e-05,
+      "loss": 0.2983,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3815434430652995,
+      "grad_norm": 1.5275932550430298,
+      "learning_rate": 1.419314536278925e-05,
+      "loss": 0.2791,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3818132757690232,
+      "grad_norm": 1.8686482906341553,
+      "learning_rate": 1.4185208128325469e-05,
+      "loss": 0.289,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3820831084727469,
+      "grad_norm": 1.051590085029602,
+      "learning_rate": 1.4177267695992958e-05,
+      "loss": 0.31,
+      "step": 1416
+    },
+    {
+      "epoch": 0.38235294117647056,
+      "grad_norm": 1.4000166654586792,
+      "learning_rate": 1.416932407185891e-05,
+      "loss": 0.3316,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3826227738801943,
+      "grad_norm": 1.6888964176177979,
+      "learning_rate": 1.4161377261992954e-05,
+      "loss": 0.2926,
+      "step": 1418
+    },
+    {
+      "epoch": 0.382892606583918,
+      "grad_norm": 0.8326447010040283,
+      "learning_rate": 1.4153427272467153e-05,
+      "loss": 0.2837,
+      "step": 1419
+    },
+    {
+      "epoch": 0.38316243928764165,
+      "grad_norm": 0.7444689273834229,
+      "learning_rate": 1.4145474109356008e-05,
+      "loss": 0.2691,
+      "step": 1420
+    },
+    {
+      "epoch": 0.38343227199136537,
+      "grad_norm": 0.8119150400161743,
+      "learning_rate": 1.413751777873643e-05,
+      "loss": 0.2534,
+      "step": 1421
+    },
+    {
+      "epoch": 0.38370210469508903,
+      "grad_norm": 1.3293696641921997,
+      "learning_rate": 1.4129558286687762e-05,
+      "loss": 0.2726,
+      "step": 1422
+    },
+    {
+      "epoch": 0.38397193739881275,
+      "grad_norm": 1.1912062168121338,
+      "learning_rate": 1.4121595639291763e-05,
+      "loss": 0.2771,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3842417701025364,
+      "grad_norm": 2.248249053955078,
+      "learning_rate": 1.411362984263259e-05,
+      "loss": 0.334,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3845116028062601,
+      "grad_norm": 1.0264694690704346,
+      "learning_rate": 1.4105660902796823e-05,
+      "loss": 0.2398,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3847814355099838,
+      "grad_norm": 1.5661494731903076,
+      "learning_rate": 1.4097688825873437e-05,
+      "loss": 0.3492,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3850512682137075,
+      "grad_norm": 2.2138843536376953,
+      "learning_rate": 1.4089713617953795e-05,
+      "loss": 0.2771,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 1.021709680557251,
+      "learning_rate": 1.4081735285131668e-05,
+      "loss": 0.2775,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3855909336211549,
+      "grad_norm": 1.2108277082443237,
+      "learning_rate": 1.4073753833503206e-05,
+      "loss": 0.2941,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3858607663248786,
+      "grad_norm": 1.4432761669158936,
+      "learning_rate": 1.4065769269166935e-05,
+      "loss": 0.2637,
+      "step": 1430
+    },
+    {
+      "epoch": 0.38613059902860225,
+      "grad_norm": 1.5312235355377197,
+      "learning_rate": 1.4057781598223776e-05,
+      "loss": 0.2819,
+      "step": 1431
+    },
+    {
+      "epoch": 0.38640043173232597,
+      "grad_norm": 1.011212944984436,
+      "learning_rate": 1.4049790826777016e-05,
+      "loss": 0.3201,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3866702644360496,
+      "grad_norm": 2.49552845954895,
+      "learning_rate": 1.4041796960932303e-05,
+      "loss": 0.2988,
+      "step": 1433
+    },
+    {
+      "epoch": 0.38694009713977334,
+      "grad_norm": 2.3510475158691406,
+      "learning_rate": 1.4033800006797664e-05,
+      "loss": 0.3043,
+      "step": 1434
+    },
+    {
+      "epoch": 0.38720992984349706,
+      "grad_norm": 1.4112496376037598,
+      "learning_rate": 1.4025799970483476e-05,
+      "loss": 0.2816,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3874797625472207,
+      "grad_norm": 2.3633015155792236,
+      "learning_rate": 1.401779685810247e-05,
+      "loss": 0.294,
+      "step": 1436
+    },
+    {
+      "epoch": 0.38774959525094443,
+      "grad_norm": 3.1169912815093994,
+      "learning_rate": 1.4009790675769736e-05,
+      "loss": 0.2811,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3880194279546681,
+      "grad_norm": 1.1797826290130615,
+      "learning_rate": 1.4001781429602704e-05,
+      "loss": 0.2922,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3882892606583918,
+      "grad_norm": 1.5838346481323242,
+      "learning_rate": 1.399376912572114e-05,
+      "loss": 0.2628,
+      "step": 1439
+    },
+    {
+      "epoch": 0.38855909336211547,
+      "grad_norm": 1.6916989088058472,
+      "learning_rate": 1.3985753770247156e-05,
+      "loss": 0.2867,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3888289260658392,
+      "grad_norm": 2.031893253326416,
+      "learning_rate": 1.3977735369305192e-05,
+      "loss": 0.2743,
+      "step": 1441
+    },
+    {
+      "epoch": 0.38909875876956285,
+      "grad_norm": 1.4210376739501953,
+      "learning_rate": 1.3969713929022015e-05,
+      "loss": 0.2936,
+      "step": 1442
+    },
+    {
+      "epoch": 0.38936859147328656,
+      "grad_norm": 1.9989560842514038,
+      "learning_rate": 1.3961689455526713e-05,
+      "loss": 0.2712,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3896384241770103,
+      "grad_norm": 1.4333336353302002,
+      "learning_rate": 1.3953661954950693e-05,
+      "loss": 0.2812,
+      "step": 1444
+    },
+    {
+      "epoch": 0.38990825688073394,
+      "grad_norm": 1.217286229133606,
+      "learning_rate": 1.3945631433427673e-05,
+      "loss": 0.2552,
+      "step": 1445
+    },
+    {
+      "epoch": 0.39017808958445765,
+      "grad_norm": 1.9379582405090332,
+      "learning_rate": 1.3937597897093684e-05,
+      "loss": 0.2953,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3904479222881813,
+      "grad_norm": 0.8678966164588928,
+      "learning_rate": 1.3929561352087056e-05,
+      "loss": 0.3161,
+      "step": 1447
+    },
+    {
+      "epoch": 0.39071775499190503,
+      "grad_norm": 1.6537083387374878,
+      "learning_rate": 1.3921521804548417e-05,
+      "loss": 0.254,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3909875876956287,
+      "grad_norm": 3.927506923675537,
+      "learning_rate": 1.3913479260620695e-05,
+      "loss": 0.3278,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3912574203993524,
+      "grad_norm": 1.3323986530303955,
+      "learning_rate": 1.3905433726449102e-05,
+      "loss": 0.2651,
+      "step": 1450
+    },
+    {
+      "epoch": 0.39152725310307607,
+      "grad_norm": 1.1926218271255493,
+      "learning_rate": 1.3897385208181136e-05,
+      "loss": 0.2442,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3917970858067998,
+      "grad_norm": 1.5048848390579224,
+      "learning_rate": 1.3889333711966579e-05,
+      "loss": 0.299,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3920669185105235,
+      "grad_norm": 1.1701581478118896,
+      "learning_rate": 1.3881279243957484e-05,
+      "loss": 0.2738,
+      "step": 1453
+    },
+    {
+      "epoch": 0.39233675121424716,
+      "grad_norm": 0.7826051712036133,
+      "learning_rate": 1.3873221810308176e-05,
+      "loss": 0.2748,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3926065839179709,
+      "grad_norm": 0.9467362761497498,
+      "learning_rate": 1.3865161417175243e-05,
+      "loss": 0.3076,
+      "step": 1455
+    },
+    {
+      "epoch": 0.39287641662169454,
+      "grad_norm": 2.264704942703247,
+      "learning_rate": 1.3857098070717543e-05,
+      "loss": 0.2881,
+      "step": 1456
+    },
+    {
+      "epoch": 0.39314624932541825,
+      "grad_norm": 0.8245582580566406,
+      "learning_rate": 1.3849031777096182e-05,
+      "loss": 0.2264,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3934160820291419,
+      "grad_norm": 0.849230945110321,
+      "learning_rate": 1.384096254247452e-05,
+      "loss": 0.2691,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3936859147328656,
+      "grad_norm": 1.0220624208450317,
+      "learning_rate": 1.3832890373018167e-05,
+      "loss": 0.302,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3939557474365893,
+      "grad_norm": 1.9688712358474731,
+      "learning_rate": 1.3824815274894974e-05,
+      "loss": 0.2706,
+      "step": 1460
+    },
+    {
+      "epoch": 0.394225580140313,
+      "grad_norm": 1.0542528629302979,
+      "learning_rate": 1.3816737254275026e-05,
+      "loss": 0.2973,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3944954128440367,
+      "grad_norm": 0.8931406140327454,
+      "learning_rate": 1.3808656317330646e-05,
+      "loss": 0.2537,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3947652455477604,
+      "grad_norm": 2.861748695373535,
+      "learning_rate": 1.3800572470236388e-05,
+      "loss": 0.2638,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3950350782514841,
+      "grad_norm": 1.677796721458435,
+      "learning_rate": 1.379248571916902e-05,
+      "loss": 0.2904,
+      "step": 1464
+    },
+    {
+      "epoch": 0.39530491095520776,
+      "grad_norm": 1.8913992643356323,
+      "learning_rate": 1.3784396070307535e-05,
+      "loss": 0.2958,
+      "step": 1465
+    },
+    {
+      "epoch": 0.39557474365893147,
+      "grad_norm": 3.177823781967163,
+      "learning_rate": 1.377630352983314e-05,
+      "loss": 0.3012,
+      "step": 1466
+    },
+    {
+      "epoch": 0.39584457636265513,
+      "grad_norm": 2.041222095489502,
+      "learning_rate": 1.376820810392925e-05,
+      "loss": 0.3014,
+      "step": 1467
+    },
+    {
+      "epoch": 0.39611440906637885,
+      "grad_norm": 0.8287718892097473,
+      "learning_rate": 1.3760109798781489e-05,
+      "loss": 0.3091,
+      "step": 1468
+    },
+    {
+      "epoch": 0.39638424177010256,
+      "grad_norm": 2.3357393741607666,
+      "learning_rate": 1.3752008620577672e-05,
+      "loss": 0.2574,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3966540744738262,
+      "grad_norm": 3.5601601600646973,
+      "learning_rate": 1.374390457550782e-05,
+      "loss": 0.2498,
+      "step": 1470
+    },
+    {
+      "epoch": 0.39692390717754994,
+      "grad_norm": 0.8133686184883118,
+      "learning_rate": 1.3735797669764137e-05,
+      "loss": 0.2523,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3971937398812736,
+      "grad_norm": 2.6341896057128906,
+      "learning_rate": 1.3727687909541014e-05,
+      "loss": 0.2626,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3974635725849973,
+      "grad_norm": 1.4977757930755615,
+      "learning_rate": 1.3719575301035026e-05,
+      "loss": 0.3147,
+      "step": 1473
+    },
+    {
+      "epoch": 0.397733405288721,
+      "grad_norm": 1.814617395401001,
+      "learning_rate": 1.3711459850444923e-05,
+      "loss": 0.2908,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3980032379924447,
+      "grad_norm": 1.3439204692840576,
+      "learning_rate": 1.3703341563971625e-05,
+      "loss": 0.2726,
+      "step": 1475
+    },
+    {
+      "epoch": 0.39827307069616835,
+      "grad_norm": 2.5131890773773193,
+      "learning_rate": 1.3695220447818219e-05,
+      "loss": 0.2933,
+      "step": 1476
+    },
+    {
+      "epoch": 0.39854290339989207,
+      "grad_norm": 1.5285488367080688,
+      "learning_rate": 1.3687096508189957e-05,
+      "loss": 0.266,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3988127361036158,
+      "grad_norm": 0.8122151494026184,
+      "learning_rate": 1.3678969751294247e-05,
+      "loss": 0.2595,
+      "step": 1478
+    },
+    {
+      "epoch": 0.39908256880733944,
+      "grad_norm": 0.9343953132629395,
+      "learning_rate": 1.3670840183340648e-05,
+      "loss": 0.2759,
+      "step": 1479
+    },
+    {
+      "epoch": 0.39935240151106316,
+      "grad_norm": 2.8351471424102783,
+      "learning_rate": 1.3662707810540867e-05,
+      "loss": 0.2433,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3996222342147868,
+      "grad_norm": 3.9198031425476074,
+      "learning_rate": 1.3654572639108756e-05,
+      "loss": 0.308,
+      "step": 1481
+    },
+    {
+      "epoch": 0.39989206691851054,
+      "grad_norm": 0.7934620976448059,
+      "learning_rate": 1.3646434675260307e-05,
+      "loss": 0.2608,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4001618996222342,
+      "grad_norm": 1.3549673557281494,
+      "learning_rate": 1.363829392521364e-05,
+      "loss": 0.2855,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4004317323259579,
+      "grad_norm": 1.0404765605926514,
+      "learning_rate": 1.363015039518901e-05,
+      "loss": 0.2821,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4007015650296816,
+      "grad_norm": 0.9348905682563782,
+      "learning_rate": 1.362200409140879e-05,
+      "loss": 0.2778,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4009713977334053,
+      "grad_norm": 1.9686415195465088,
+      "learning_rate": 1.3613855020097477e-05,
+      "loss": 0.27,
+      "step": 1486
+    },
+    {
+      "epoch": 0.401241230437129,
+      "grad_norm": 0.7990533113479614,
+      "learning_rate": 1.3605703187481685e-05,
+      "loss": 0.28,
+      "step": 1487
+    },
+    {
+      "epoch": 0.40151106314085266,
+      "grad_norm": 2.5923070907592773,
+      "learning_rate": 1.3597548599790131e-05,
+      "loss": 0.2774,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4017808958445764,
+      "grad_norm": 2.4670286178588867,
+      "learning_rate": 1.3589391263253638e-05,
+      "loss": 0.2814,
+      "step": 1489
+    },
+    {
+      "epoch": 0.40205072854830004,
+      "grad_norm": 2.573228120803833,
+      "learning_rate": 1.3581231184105136e-05,
+      "loss": 0.2646,
+      "step": 1490
+    },
+    {
+      "epoch": 0.40232056125202376,
+      "grad_norm": 3.002721071243286,
+      "learning_rate": 1.3573068368579644e-05,
+      "loss": 0.2723,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4025903939557474,
+      "grad_norm": 3.166978597640991,
+      "learning_rate": 1.3564902822914274e-05,
+      "loss": 0.2657,
+      "step": 1492
+    },
+    {
+      "epoch": 0.40286022665947113,
+      "grad_norm": 1.5848630666732788,
+      "learning_rate": 1.3556734553348227e-05,
+      "loss": 0.2724,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4031300593631948,
+      "grad_norm": 3.3456015586853027,
+      "learning_rate": 1.3548563566122781e-05,
+      "loss": 0.3015,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4033998920669185,
+      "grad_norm": 1.700993537902832,
+      "learning_rate": 1.354038986748129e-05,
+      "loss": 0.3221,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 1.3336894512176514,
+      "learning_rate": 1.3532213463669184e-05,
+      "loss": 0.2802,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4039395574743659,
+      "grad_norm": 2.1244664192199707,
+      "learning_rate": 1.3524034360933957e-05,
+      "loss": 0.2744,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4042093901780896,
+      "grad_norm": 2.1234378814697266,
+      "learning_rate": 1.3515852565525167e-05,
+      "loss": 0.2424,
+      "step": 1498
+    },
+    {
+      "epoch": 0.40447922288181326,
+      "grad_norm": 2.1078450679779053,
+      "learning_rate": 1.3507668083694428e-05,
+      "loss": 0.2943,
+      "step": 1499
+    },
+    {
+      "epoch": 0.404749055585537,
+      "grad_norm": 2.262376546859741,
+      "learning_rate": 1.3499480921695407e-05,
+      "loss": 0.2972,
+      "step": 1500
+    },
+    {
+      "epoch": 0.40501888828926064,
+      "grad_norm": 1.2333877086639404,
+      "learning_rate": 1.3491291085783815e-05,
+      "loss": 0.2905,
+      "step": 1501
+    },
+    {
+      "epoch": 0.40528872099298435,
+      "grad_norm": 2.2646324634552,
+      "learning_rate": 1.348309858221742e-05,
+      "loss": 0.304,
+      "step": 1502
+    },
+    {
+      "epoch": 0.40555855369670807,
+      "grad_norm": 0.929746150970459,
+      "learning_rate": 1.347490341725601e-05,
+      "loss": 0.2774,
+      "step": 1503
+    },
+    {
+      "epoch": 0.40582838640043173,
+      "grad_norm": 1.8068088293075562,
+      "learning_rate": 1.3466705597161416e-05,
+      "loss": 0.3021,
+      "step": 1504
+    },
+    {
+      "epoch": 0.40609821910415544,
+      "grad_norm": 0.7326279282569885,
+      "learning_rate": 1.3458505128197502e-05,
+      "loss": 0.3073,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4063680518078791,
+      "grad_norm": 1.233330488204956,
+      "learning_rate": 1.3450302016630141e-05,
+      "loss": 0.2534,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4066378845116028,
+      "grad_norm": 1.0014569759368896,
+      "learning_rate": 1.3442096268727246e-05,
+      "loss": 0.2886,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4069077172153265,
+      "grad_norm": 0.8554761409759521,
+      "learning_rate": 1.3433887890758727e-05,
+      "loss": 0.241,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4071775499190502,
+      "grad_norm": 1.557159662246704,
+      "learning_rate": 1.3425676888996508e-05,
+      "loss": 0.2961,
+      "step": 1509
+    },
+    {
+      "epoch": 0.40744738262277386,
+      "grad_norm": 0.8422708511352539,
+      "learning_rate": 1.3417463269714525e-05,
+      "loss": 0.2403,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4077172153264976,
+      "grad_norm": 1.6500035524368286,
+      "learning_rate": 1.3409247039188712e-05,
+      "loss": 0.2702,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4079870480302213,
+      "grad_norm": 0.9386299848556519,
+      "learning_rate": 1.340102820369698e-05,
+      "loss": 0.2831,
+      "step": 1512
+    },
+    {
+      "epoch": 0.40825688073394495,
+      "grad_norm": 0.9903119802474976,
+      "learning_rate": 1.339280676951926e-05,
+      "loss": 0.2935,
+      "step": 1513
+    },
+    {
+      "epoch": 0.40852671343766866,
+      "grad_norm": 1.4045350551605225,
+      "learning_rate": 1.3384582742937448e-05,
+      "loss": 0.282,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4087965461413923,
+      "grad_norm": 1.5634765625,
+      "learning_rate": 1.3376356130235426e-05,
+      "loss": 0.3084,
+      "step": 1515
+    },
+    {
+      "epoch": 0.40906637884511604,
+      "grad_norm": 1.5632762908935547,
+      "learning_rate": 1.3368126937699055e-05,
+      "loss": 0.2812,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4093362115488397,
+      "grad_norm": 0.7643537521362305,
+      "learning_rate": 1.3359895171616168e-05,
+      "loss": 0.294,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4096060442525634,
+      "grad_norm": 0.8108052015304565,
+      "learning_rate": 1.3351660838276553e-05,
+      "loss": 0.2638,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4098758769562871,
+      "grad_norm": 1.245322823524475,
+      "learning_rate": 1.3343423943971976e-05,
+      "loss": 0.2491,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4101457096600108,
+      "grad_norm": 0.894798755645752,
+      "learning_rate": 1.3335184494996148e-05,
+      "loss": 0.2745,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4104155423637345,
+      "grad_norm": 1.0814299583435059,
+      "learning_rate": 1.3326942497644737e-05,
+      "loss": 0.2999,
+      "step": 1521
+    },
+    {
+      "epoch": 0.41068537506745817,
+      "grad_norm": 1.2975765466690063,
+      "learning_rate": 1.3318697958215358e-05,
+      "loss": 0.2561,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4109552077711819,
+      "grad_norm": 1.3861722946166992,
+      "learning_rate": 1.3310450883007562e-05,
+      "loss": 0.2739,
+      "step": 1523
+    },
+    {
+      "epoch": 0.41122504047490555,
+      "grad_norm": 0.9870875477790833,
+      "learning_rate": 1.330220127832285e-05,
+      "loss": 0.2621,
+      "step": 1524
+    },
+    {
+      "epoch": 0.41149487317862926,
+      "grad_norm": 3.050025224685669,
+      "learning_rate": 1.3293949150464651e-05,
+      "loss": 0.2326,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 1.0873831510543823,
+      "learning_rate": 1.3285694505738307e-05,
+      "loss": 0.2715,
+      "step": 1526
+    },
+    {
+      "epoch": 0.41203453858607664,
+      "grad_norm": 1.426272988319397,
+      "learning_rate": 1.3277437350451107e-05,
+      "loss": 0.3007,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4123043712898003,
+      "grad_norm": 2.20251202583313,
+      "learning_rate": 1.3269177690912244e-05,
+      "loss": 0.2516,
+      "step": 1528
+    },
+    {
+      "epoch": 0.412574203993524,
+      "grad_norm": 1.3408479690551758,
+      "learning_rate": 1.3260915533432827e-05,
+      "loss": 0.2529,
+      "step": 1529
+    },
+    {
+      "epoch": 0.41284403669724773,
+      "grad_norm": 1.262690544128418,
+      "learning_rate": 1.3252650884325873e-05,
+      "loss": 0.2818,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4131138694009714,
+      "grad_norm": 1.107474446296692,
+      "learning_rate": 1.3244383749906308e-05,
+      "loss": 0.2849,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4133837021046951,
+      "grad_norm": 1.2090812921524048,
+      "learning_rate": 1.3236114136490947e-05,
+      "loss": 0.2652,
+      "step": 1532
+    },
+    {
+      "epoch": 0.41365353480841877,
+      "grad_norm": 0.8837510943412781,
+      "learning_rate": 1.322784205039851e-05,
+      "loss": 0.2794,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4139233675121425,
+      "grad_norm": 0.8789448738098145,
+      "learning_rate": 1.3219567497949603e-05,
+      "loss": 0.281,
+      "step": 1534
+    },
+    {
+      "epoch": 0.41419320021586614,
+      "grad_norm": 0.9842005372047424,
+      "learning_rate": 1.3211290485466708e-05,
+      "loss": 0.2583,
+      "step": 1535
+    },
+    {
+      "epoch": 0.41446303291958986,
+      "grad_norm": 1.2743475437164307,
+      "learning_rate": 1.32030110192742e-05,
+      "loss": 0.2948,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4147328656233135,
+      "grad_norm": 0.8208792209625244,
+      "learning_rate": 1.3194729105698318e-05,
+      "loss": 0.3003,
+      "step": 1537
+    },
+    {
+      "epoch": 0.41500269832703723,
+      "grad_norm": 1.4075411558151245,
+      "learning_rate": 1.318644475106718e-05,
+      "loss": 0.2673,
+      "step": 1538
+    },
+    {
+      "epoch": 0.41527253103076095,
+      "grad_norm": 1.5867305994033813,
+      "learning_rate": 1.3178157961710762e-05,
+      "loss": 0.2237,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4155423637344846,
+      "grad_norm": 3.088475465774536,
+      "learning_rate": 1.3169868743960904e-05,
+      "loss": 0.3113,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4158121964382083,
+      "grad_norm": 2.2844786643981934,
+      "learning_rate": 1.3161577104151303e-05,
+      "loss": 0.3038,
+      "step": 1541
+    },
+    {
+      "epoch": 0.416082029141932,
+      "grad_norm": 1.3498417139053345,
+      "learning_rate": 1.31532830486175e-05,
+      "loss": 0.2992,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4163518618456557,
+      "grad_norm": 4.195821285247803,
+      "learning_rate": 1.3144986583696888e-05,
+      "loss": 0.2831,
+      "step": 1543
+    },
+    {
+      "epoch": 0.41662169454937936,
+      "grad_norm": 1.7130485773086548,
+      "learning_rate": 1.31366877157287e-05,
+      "loss": 0.2678,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4168915272531031,
+      "grad_norm": 1.0272133350372314,
+      "learning_rate": 1.3128386451054e-05,
+      "loss": 0.2809,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4171613599568268,
+      "grad_norm": 2.556349039077759,
+      "learning_rate": 1.3120082796015694e-05,
+      "loss": 0.3341,
+      "step": 1546
+    },
+    {
+      "epoch": 0.41743119266055045,
+      "grad_norm": 0.749023973941803,
+      "learning_rate": 1.3111776756958503e-05,
+      "loss": 0.2609,
+      "step": 1547
+    },
+    {
+      "epoch": 0.41770102536427417,
+      "grad_norm": 2.112528085708618,
+      "learning_rate": 1.310346834022898e-05,
+      "loss": 0.319,
+      "step": 1548
+    },
+    {
+      "epoch": 0.41797085806799783,
+      "grad_norm": 1.4362165927886963,
+      "learning_rate": 1.3095157552175481e-05,
+      "loss": 0.2892,
+      "step": 1549
+    },
+    {
+      "epoch": 0.41824069077172155,
+      "grad_norm": 1.1248252391815186,
+      "learning_rate": 1.3086844399148191e-05,
+      "loss": 0.2854,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4185105234754452,
+      "grad_norm": 2.0591673851013184,
+      "learning_rate": 1.3078528887499088e-05,
+      "loss": 0.2501,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4187803561791689,
+      "grad_norm": 1.534560203552246,
+      "learning_rate": 1.3070211023581959e-05,
+      "loss": 0.2449,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4190501888828926,
+      "grad_norm": 0.8788701295852661,
+      "learning_rate": 1.3061890813752384e-05,
+      "loss": 0.2803,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4193200215866163,
+      "grad_norm": 2.063279151916504,
+      "learning_rate": 1.3053568264367744e-05,
+      "loss": 0.2846,
+      "step": 1554
+    },
+    {
+      "epoch": 0.41958985429034,
+      "grad_norm": 0.7388185858726501,
+      "learning_rate": 1.3045243381787197e-05,
+      "loss": 0.2774,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4198596869940637,
+      "grad_norm": 3.0363991260528564,
+      "learning_rate": 1.3036916172371692e-05,
+      "loss": 0.3086,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4201295196977874,
+      "grad_norm": 1.0844616889953613,
+      "learning_rate": 1.3028586642483951e-05,
+      "loss": 0.2612,
+      "step": 1557
+    },
+    {
+      "epoch": 0.42039935240151105,
+      "grad_norm": 2.309138059616089,
+      "learning_rate": 1.302025479848847e-05,
+      "loss": 0.2928,
+      "step": 1558
+    },
+    {
+      "epoch": 0.42066918510523477,
+      "grad_norm": 0.7124882936477661,
+      "learning_rate": 1.3011920646751515e-05,
+      "loss": 0.2612,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4209390178089584,
+      "grad_norm": 0.9218741059303284,
+      "learning_rate": 1.3003584193641115e-05,
+      "loss": 0.2681,
+      "step": 1560
+    },
+    {
+      "epoch": 0.42120885051268214,
+      "grad_norm": 0.8616629838943481,
+      "learning_rate": 1.2995245445527054e-05,
+      "loss": 0.2782,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4214786832164058,
+      "grad_norm": 0.9681334495544434,
+      "learning_rate": 1.2986904408780875e-05,
+      "loss": 0.3005,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4217485159201295,
+      "grad_norm": 1.6181387901306152,
+      "learning_rate": 1.2978561089775865e-05,
+      "loss": 0.2877,
+      "step": 1563
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.7624027729034424,
+      "learning_rate": 1.2970215494887057e-05,
+      "loss": 0.2592,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4222881813275769,
+      "grad_norm": 0.8228490352630615,
+      "learning_rate": 1.2961867630491222e-05,
+      "loss": 0.2794,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4225580140313006,
+      "grad_norm": 0.9140209555625916,
+      "learning_rate": 1.295351750296687e-05,
+      "loss": 0.2764,
+      "step": 1566
+    },
+    {
+      "epoch": 0.42282784673502427,
+      "grad_norm": 1.1879830360412598,
+      "learning_rate": 1.2945165118694232e-05,
+      "loss": 0.2738,
+      "step": 1567
+    },
+    {
+      "epoch": 0.423097679438748,
+      "grad_norm": 1.793563723564148,
+      "learning_rate": 1.2936810484055267e-05,
+      "loss": 0.2897,
+      "step": 1568
+    },
+    {
+      "epoch": 0.42336751214247165,
+      "grad_norm": 1.3230788707733154,
+      "learning_rate": 1.2928453605433658e-05,
+      "loss": 0.2812,
+      "step": 1569
+    },
+    {
+      "epoch": 0.42363734484619536,
+      "grad_norm": 0.7468932271003723,
+      "learning_rate": 1.2920094489214794e-05,
+      "loss": 0.2967,
+      "step": 1570
+    },
+    {
+      "epoch": 0.423907177549919,
+      "grad_norm": 1.990308165550232,
+      "learning_rate": 1.2911733141785779e-05,
+      "loss": 0.2767,
+      "step": 1571
+    },
+    {
+      "epoch": 0.42417701025364274,
+      "grad_norm": 1.0316516160964966,
+      "learning_rate": 1.2903369569535427e-05,
+      "loss": 0.2995,
+      "step": 1572
+    },
+    {
+      "epoch": 0.42444684295736645,
+      "grad_norm": 1.7481430768966675,
+      "learning_rate": 1.2895003778854238e-05,
+      "loss": 0.2763,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4247166756610901,
+      "grad_norm": 0.7182138562202454,
+      "learning_rate": 1.2886635776134415e-05,
+      "loss": 0.2723,
+      "step": 1574
+    },
+    {
+      "epoch": 0.42498650836481383,
+      "grad_norm": 1.3726974725723267,
+      "learning_rate": 1.2878265567769862e-05,
+      "loss": 0.2831,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4252563410685375,
+      "grad_norm": 0.9556413292884827,
+      "learning_rate": 1.2869893160156144e-05,
+      "loss": 0.2875,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4255261737722612,
+      "grad_norm": 1.8461720943450928,
+      "learning_rate": 1.286151855969053e-05,
+      "loss": 0.2448,
+      "step": 1577
+    },
+    {
+      "epoch": 0.42579600647598487,
+      "grad_norm": 0.9314191937446594,
+      "learning_rate": 1.2853141772771949e-05,
+      "loss": 0.2712,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4260658391797086,
+      "grad_norm": 1.144457459449768,
+      "learning_rate": 1.2844762805801008e-05,
+      "loss": 0.2819,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4263356718834323,
+      "grad_norm": 0.8922066688537598,
+      "learning_rate": 1.2836381665179977e-05,
+      "loss": 0.277,
+      "step": 1580
+    },
+    {
+      "epoch": 0.42660550458715596,
+      "grad_norm": 1.5883222818374634,
+      "learning_rate": 1.2827998357312791e-05,
+      "loss": 0.2939,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4268753372908797,
+      "grad_norm": 0.9617652297019958,
+      "learning_rate": 1.2819612888605038e-05,
+      "loss": 0.2554,
+      "step": 1582
+    },
+    {
+      "epoch": 0.42714516999460334,
+      "grad_norm": 0.9037154316902161,
+      "learning_rate": 1.281122526546395e-05,
+      "loss": 0.2551,
+      "step": 1583
+    },
+    {
+      "epoch": 0.42741500269832705,
+      "grad_norm": 3.3124070167541504,
+      "learning_rate": 1.280283549429842e-05,
+      "loss": 0.2957,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4276848354020507,
+      "grad_norm": 1.1585896015167236,
+      "learning_rate": 1.2794443581518967e-05,
+      "loss": 0.2958,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4279546681057744,
+      "grad_norm": 2.1725046634674072,
+      "learning_rate": 1.278604953353776e-05,
+      "loss": 0.2908,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4282245008094981,
+      "grad_norm": 2.907588481903076,
+      "learning_rate": 1.2777653356768591e-05,
+      "loss": 0.2695,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4284943335132218,
+      "grad_norm": 0.7130913734436035,
+      "learning_rate": 1.2769255057626879e-05,
+      "loss": 0.2493,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4287641662169455,
+      "grad_norm": 0.9983488321304321,
+      "learning_rate": 1.276085464252967e-05,
+      "loss": 0.2727,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4290339989206692,
+      "grad_norm": 1.5578442811965942,
+      "learning_rate": 1.2752452117895621e-05,
+      "loss": 0.2878,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4293038316243929,
+      "grad_norm": 2.873304843902588,
+      "learning_rate": 1.2744047490144999e-05,
+      "loss": 0.2872,
+      "step": 1591
+    },
+    {
+      "epoch": 0.42957366432811656,
+      "grad_norm": 1.3378244638442993,
+      "learning_rate": 1.2735640765699693e-05,
+      "loss": 0.3076,
+      "step": 1592
+    },
+    {
+      "epoch": 0.42984349703184027,
+      "grad_norm": 2.0782854557037354,
+      "learning_rate": 1.2727231950983174e-05,
+      "loss": 0.3178,
+      "step": 1593
+    },
+    {
+      "epoch": 0.43011332973556393,
+      "grad_norm": 3.6254141330718994,
+      "learning_rate": 1.2718821052420518e-05,
+      "loss": 0.2695,
+      "step": 1594
+    },
+    {
+      "epoch": 0.43038316243928765,
+      "grad_norm": 1.0718744993209839,
+      "learning_rate": 1.2710408076438406e-05,
+      "loss": 0.2906,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4306529951430113,
+      "grad_norm": 1.1864875555038452,
+      "learning_rate": 1.270199302946508e-05,
+      "loss": 0.2895,
+      "step": 1596
+    },
+    {
+      "epoch": 0.430922827846735,
+      "grad_norm": 1.4502627849578857,
+      "learning_rate": 1.2693575917930387e-05,
+      "loss": 0.2744,
+      "step": 1597
+    },
+    {
+      "epoch": 0.43119266055045874,
+      "grad_norm": 3.9329752922058105,
+      "learning_rate": 1.2685156748265748e-05,
+      "loss": 0.2858,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4314624932541824,
+      "grad_norm": 0.7731364369392395,
+      "learning_rate": 1.2676735526904145e-05,
+      "loss": 0.2609,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4317323259579061,
+      "grad_norm": 1.220463752746582,
+      "learning_rate": 1.2668312260280136e-05,
+      "loss": 0.2999,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4320021586616298,
+      "grad_norm": 2.912548065185547,
+      "learning_rate": 1.2659886954829844e-05,
+      "loss": 0.282,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4322719913653535,
+      "grad_norm": 1.076962947845459,
+      "learning_rate": 1.2651459616990942e-05,
+      "loss": 0.2973,
+      "step": 1602
+    },
+    {
+      "epoch": 0.43254182406907715,
+      "grad_norm": 1.6514934301376343,
+      "learning_rate": 1.2643030253202669e-05,
+      "loss": 0.2697,
+      "step": 1603
+    },
+    {
+      "epoch": 0.43281165677280087,
+      "grad_norm": 0.8611807823181152,
+      "learning_rate": 1.2634598869905797e-05,
+      "loss": 0.2795,
+      "step": 1604
+    },
+    {
+      "epoch": 0.43308148947652453,
+      "grad_norm": 1.3867688179016113,
+      "learning_rate": 1.2626165473542645e-05,
+      "loss": 0.2819,
+      "step": 1605
+    },
+    {
+      "epoch": 0.43335132218024824,
+      "grad_norm": 0.9957842826843262,
+      "learning_rate": 1.2617730070557079e-05,
+      "loss": 0.305,
+      "step": 1606
+    },
+    {
+      "epoch": 0.43362115488397196,
+      "grad_norm": 0.8424469232559204,
+      "learning_rate": 1.260929266739449e-05,
+      "loss": 0.2573,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4338909875876956,
+      "grad_norm": 0.7613505125045776,
+      "learning_rate": 1.2600853270501798e-05,
+      "loss": 0.2618,
+      "step": 1608
+    },
+    {
+      "epoch": 0.43416082029141934,
+      "grad_norm": 3.1709396839141846,
+      "learning_rate": 1.259241188632745e-05,
+      "loss": 0.2703,
+      "step": 1609
+    },
+    {
+      "epoch": 0.434430652995143,
+      "grad_norm": 0.8743679523468018,
+      "learning_rate": 1.258396852132141e-05,
+      "loss": 0.2916,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4347004856988667,
+      "grad_norm": 1.100263237953186,
+      "learning_rate": 1.2575523181935148e-05,
+      "loss": 0.2979,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4349703184025904,
+      "grad_norm": 1.1564009189605713,
+      "learning_rate": 1.2567075874621658e-05,
+      "loss": 0.2541,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4352401511063141,
+      "grad_norm": 1.540154218673706,
+      "learning_rate": 1.2558626605835426e-05,
+      "loss": 0.3024,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4355099838100378,
+      "grad_norm": 2.4456286430358887,
+      "learning_rate": 1.2550175382032434e-05,
+      "loss": 0.3061,
+      "step": 1614
+    },
+    {
+      "epoch": 0.43577981651376146,
+      "grad_norm": 1.008794903755188,
+      "learning_rate": 1.2541722209670173e-05,
+      "loss": 0.2568,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4360496492174852,
+      "grad_norm": 2.1274492740631104,
+      "learning_rate": 1.2533267095207608e-05,
+      "loss": 0.2814,
+      "step": 1616
+    },
+    {
+      "epoch": 0.43631948192120884,
+      "grad_norm": 1.5166007280349731,
+      "learning_rate": 1.2524810045105194e-05,
+      "loss": 0.2711,
+      "step": 1617
+    },
+    {
+      "epoch": 0.43658931462493256,
+      "grad_norm": 1.7389001846313477,
+      "learning_rate": 1.2516351065824864e-05,
+      "loss": 0.2637,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4368591473286562,
+      "grad_norm": 1.9732087850570679,
+      "learning_rate": 1.2507890163830028e-05,
+      "loss": 0.3067,
+      "step": 1619
+    },
+    {
+      "epoch": 0.43712898003237993,
+      "grad_norm": 1.8479074239730835,
+      "learning_rate": 1.249942734558556e-05,
+      "loss": 0.2619,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4373988127361036,
+      "grad_norm": 1.9939231872558594,
+      "learning_rate": 1.24909626175578e-05,
+      "loss": 0.3091,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4376686454398273,
+      "grad_norm": 0.9880063533782959,
+      "learning_rate": 1.2482495986214551e-05,
+      "loss": 0.3025,
+      "step": 1622
+    },
+    {
+      "epoch": 0.437938478143551,
+      "grad_norm": 1.4989172220230103,
+      "learning_rate": 1.2474027458025067e-05,
+      "loss": 0.2893,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4382083108472747,
+      "grad_norm": 1.6497864723205566,
+      "learning_rate": 1.2465557039460048e-05,
+      "loss": 0.3056,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4384781435509984,
+      "grad_norm": 2.66267991065979,
+      "learning_rate": 1.2457084736991645e-05,
+      "loss": 0.2463,
+      "step": 1625
+    },
+    {
+      "epoch": 0.43874797625472206,
+      "grad_norm": 0.7502012252807617,
+      "learning_rate": 1.2448610557093446e-05,
+      "loss": 0.2565,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4390178089584458,
+      "grad_norm": 3.9167354106903076,
+      "learning_rate": 1.2440134506240469e-05,
+      "loss": 0.3142,
+      "step": 1627
+    },
+    {
+      "epoch": 0.43928764166216944,
+      "grad_norm": 0.9917934536933899,
+      "learning_rate": 1.243165659090917e-05,
+      "loss": 0.242,
+      "step": 1628
+    },
+    {
+      "epoch": 0.43955747436589315,
+      "grad_norm": 2.2707622051239014,
+      "learning_rate": 1.2423176817577427e-05,
+      "loss": 0.3169,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4398273070696168,
+      "grad_norm": 0.9845569729804993,
+      "learning_rate": 1.241469519272453e-05,
+      "loss": 0.306,
+      "step": 1630
+    },
+    {
+      "epoch": 0.44009713977334053,
+      "grad_norm": 1.951632022857666,
+      "learning_rate": 1.2406211722831192e-05,
+      "loss": 0.2868,
+      "step": 1631
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 2.628763437271118,
+      "learning_rate": 1.2397726414379535e-05,
+      "loss": 0.3086,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4406368051807879,
+      "grad_norm": 1.4184796810150146,
+      "learning_rate": 1.2389239273853084e-05,
+      "loss": 0.2823,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4409066378845116,
+      "grad_norm": 1.7122094631195068,
+      "learning_rate": 1.2380750307736761e-05,
+      "loss": 0.2675,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4411764705882353,
+      "grad_norm": 1.7353466749191284,
+      "learning_rate": 1.237225952251689e-05,
+      "loss": 0.2785,
+      "step": 1635
+    },
+    {
+      "epoch": 0.441446303291959,
+      "grad_norm": 2.1962356567382812,
+      "learning_rate": 1.2363766924681178e-05,
+      "loss": 0.2687,
+      "step": 1636
+    },
+    {
+      "epoch": 0.44171613599568266,
+      "grad_norm": 0.7292801737785339,
+      "learning_rate": 1.2355272520718722e-05,
+      "loss": 0.2567,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4419859686994064,
+      "grad_norm": 1.4787192344665527,
+      "learning_rate": 1.2346776317119997e-05,
+      "loss": 0.3094,
+      "step": 1638
+    },
+    {
+      "epoch": 0.44225580140313003,
+      "grad_norm": 2.063849925994873,
+      "learning_rate": 1.2338278320376851e-05,
+      "loss": 0.2867,
+      "step": 1639
+    },
+    {
+      "epoch": 0.44252563410685375,
+      "grad_norm": 2.1483333110809326,
+      "learning_rate": 1.2329778536982507e-05,
+      "loss": 0.2873,
+      "step": 1640
+    },
+    {
+      "epoch": 0.44279546681057746,
+      "grad_norm": 0.9305816888809204,
+      "learning_rate": 1.2321276973431547e-05,
+      "loss": 0.2672,
+      "step": 1641
+    },
+    {
+      "epoch": 0.4430652995143011,
+      "grad_norm": 2.329916477203369,
+      "learning_rate": 1.2312773636219919e-05,
+      "loss": 0.2717,
+      "step": 1642
+    },
+    {
+      "epoch": 0.44333513221802484,
+      "grad_norm": 2.497776508331299,
+      "learning_rate": 1.2304268531844926e-05,
+      "loss": 0.2951,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4436049649217485,
+      "grad_norm": 1.4716198444366455,
+      "learning_rate": 1.2295761666805212e-05,
+      "loss": 0.2508,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4438747976254722,
+      "grad_norm": 1.6698970794677734,
+      "learning_rate": 1.228725304760078e-05,
+      "loss": 0.2701,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4441446303291959,
+      "grad_norm": 2.357048273086548,
+      "learning_rate": 1.2278742680732964e-05,
+      "loss": 0.2952,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4444144630329196,
+      "grad_norm": 0.8967030644416809,
+      "learning_rate": 1.2270230572704438e-05,
+      "loss": 0.2615,
+      "step": 1647
+    },
+    {
+      "epoch": 0.44468429573664325,
+      "grad_norm": 4.08506965637207,
+      "learning_rate": 1.2261716730019202e-05,
+      "loss": 0.3016,
+      "step": 1648
+    },
+    {
+      "epoch": 0.44495412844036697,
+      "grad_norm": 2.9097177982330322,
+      "learning_rate": 1.2253201159182586e-05,
+      "loss": 0.2899,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4452239611440907,
+      "grad_norm": 0.8899257183074951,
+      "learning_rate": 1.2244683866701237e-05,
+      "loss": 0.274,
+      "step": 1650
+    },
+    {
+      "epoch": 0.44549379384781435,
+      "grad_norm": 2.2423558235168457,
+      "learning_rate": 1.2236164859083117e-05,
+      "loss": 0.3261,
+      "step": 1651
+    },
+    {
+      "epoch": 0.44576362655153806,
+      "grad_norm": 3.8785812854766846,
+      "learning_rate": 1.2227644142837507e-05,
+      "loss": 0.3066,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4460334592552617,
+      "grad_norm": 2.0214996337890625,
+      "learning_rate": 1.2219121724474976e-05,
+      "loss": 0.2716,
+      "step": 1653
+    },
+    {
+      "epoch": 0.44630329195898544,
+      "grad_norm": 1.0493419170379639,
+      "learning_rate": 1.2210597610507418e-05,
+      "loss": 0.2803,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4465731246627091,
+      "grad_norm": 4.234036922454834,
+      "learning_rate": 1.2202071807448e-05,
+      "loss": 0.2756,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4468429573664328,
+      "grad_norm": 3.3894639015197754,
+      "learning_rate": 1.2193544321811191e-05,
+      "loss": 0.3201,
+      "step": 1656
+    },
+    {
+      "epoch": 0.44711279007015653,
+      "grad_norm": 0.678377091884613,
+      "learning_rate": 1.2185015160112741e-05,
+      "loss": 0.2598,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4473826227738802,
+      "grad_norm": 2.4082858562469482,
+      "learning_rate": 1.2176484328869687e-05,
+      "loss": 0.3157,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4476524554776039,
+      "grad_norm": 3.0508840084075928,
+      "learning_rate": 1.2167951834600335e-05,
+      "loss": 0.2702,
+      "step": 1659
+    },
+    {
+      "epoch": 0.44792228818132757,
+      "grad_norm": 1.4309051036834717,
+      "learning_rate": 1.2159417683824266e-05,
+      "loss": 0.2948,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4481921208850513,
+      "grad_norm": 1.4191372394561768,
+      "learning_rate": 1.2150881883062326e-05,
+      "loss": 0.2689,
+      "step": 1661
+    },
+    {
+      "epoch": 0.44846195358877494,
+      "grad_norm": 2.3413186073303223,
+      "learning_rate": 1.2142344438836617e-05,
+      "loss": 0.2732,
+      "step": 1662
+    },
+    {
+      "epoch": 0.44873178629249866,
+      "grad_norm": 2.1374967098236084,
+      "learning_rate": 1.2133805357670503e-05,
+      "loss": 0.2713,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4490016189962223,
+      "grad_norm": 0.8924049735069275,
+      "learning_rate": 1.2125264646088598e-05,
+      "loss": 0.2578,
+      "step": 1664
+    },
+    {
+      "epoch": 0.44927145169994603,
+      "grad_norm": 2.805912494659424,
+      "learning_rate": 1.2116722310616755e-05,
+      "loss": 0.285,
+      "step": 1665
+    },
+    {
+      "epoch": 0.44954128440366975,
+      "grad_norm": 1.8601393699645996,
+      "learning_rate": 1.2108178357782079e-05,
+      "loss": 0.2751,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4498111171073934,
+      "grad_norm": 0.8004688024520874,
+      "learning_rate": 1.2099632794112901e-05,
+      "loss": 0.2666,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4500809498111171,
+      "grad_norm": 1.9667503833770752,
+      "learning_rate": 1.2091085626138788e-05,
+      "loss": 0.2774,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4503507825148408,
+      "grad_norm": 2.0271430015563965,
+      "learning_rate": 1.2082536860390528e-05,
+      "loss": 0.2873,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4506206152185645,
+      "grad_norm": 2.2222788333892822,
+      "learning_rate": 1.2073986503400137e-05,
+      "loss": 0.2919,
+      "step": 1670
+    },
+    {
+      "epoch": 0.45089044792228816,
+      "grad_norm": 1.9875798225402832,
+      "learning_rate": 1.2065434561700847e-05,
+      "loss": 0.2877,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4511602806260119,
+      "grad_norm": 2.405111074447632,
+      "learning_rate": 1.205688104182709e-05,
+      "loss": 0.3129,
+      "step": 1672
+    },
+    {
+      "epoch": 0.45143011332973554,
+      "grad_norm": 1.2459466457366943,
+      "learning_rate": 1.2048325950314508e-05,
+      "loss": 0.2869,
+      "step": 1673
+    },
+    {
+      "epoch": 0.45169994603345925,
+      "grad_norm": 1.734710693359375,
+      "learning_rate": 1.2039769293699956e-05,
+      "loss": 0.3143,
+      "step": 1674
+    },
+    {
+      "epoch": 0.45196977873718297,
+      "grad_norm": 1.8248006105422974,
+      "learning_rate": 1.2031211078521471e-05,
+      "loss": 0.307,
+      "step": 1675
+    },
+    {
+      "epoch": 0.45223961144090663,
+      "grad_norm": 1.095033884048462,
+      "learning_rate": 1.2022651311318283e-05,
+      "loss": 0.3176,
+      "step": 1676
+    },
+    {
+      "epoch": 0.45250944414463035,
+      "grad_norm": 2.2088842391967773,
+      "learning_rate": 1.2014089998630815e-05,
+      "loss": 0.248,
+      "step": 1677
+    },
+    {
+      "epoch": 0.452779276848354,
+      "grad_norm": 0.8687427639961243,
+      "learning_rate": 1.2005527147000663e-05,
+      "loss": 0.3092,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4530491095520777,
+      "grad_norm": 1.0797159671783447,
+      "learning_rate": 1.1996962762970604e-05,
+      "loss": 0.2992,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4533189422558014,
+      "grad_norm": 0.7450168132781982,
+      "learning_rate": 1.1988396853084585e-05,
+      "loss": 0.2502,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4535887749595251,
+      "grad_norm": 0.7660818696022034,
+      "learning_rate": 1.1979829423887714e-05,
+      "loss": 0.2892,
+      "step": 1681
+    },
+    {
+      "epoch": 0.45385860766324876,
+      "grad_norm": 0.8350481986999512,
+      "learning_rate": 1.1971260481926264e-05,
+      "loss": 0.2739,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4541284403669725,
+      "grad_norm": 1.1423898935317993,
+      "learning_rate": 1.1962690033747669e-05,
+      "loss": 0.2812,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4543982730706962,
+      "grad_norm": 2.1746318340301514,
+      "learning_rate": 1.1954118085900503e-05,
+      "loss": 0.3103,
+      "step": 1684
+    },
+    {
+      "epoch": 0.45466810577441985,
+      "grad_norm": 0.8126717805862427,
+      "learning_rate": 1.194554464493449e-05,
+      "loss": 0.2884,
+      "step": 1685
+    },
+    {
+      "epoch": 0.45493793847814357,
+      "grad_norm": 0.9850130081176758,
+      "learning_rate": 1.1936969717400506e-05,
+      "loss": 0.3034,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4552077711818672,
+      "grad_norm": 2.3839621543884277,
+      "learning_rate": 1.192839330985054e-05,
+      "loss": 0.3049,
+      "step": 1687
+    },
+    {
+      "epoch": 0.45547760388559094,
+      "grad_norm": 1.591545820236206,
+      "learning_rate": 1.1919815428837736e-05,
+      "loss": 0.2849,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4557474365893146,
+      "grad_norm": 1.1181315183639526,
+      "learning_rate": 1.1911236080916352e-05,
+      "loss": 0.2858,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4560172692930383,
+      "grad_norm": 1.3874965906143188,
+      "learning_rate": 1.1902655272641756e-05,
+      "loss": 0.2605,
+      "step": 1690
+    },
+    {
+      "epoch": 0.45628710199676203,
+      "grad_norm": 2.4412620067596436,
+      "learning_rate": 1.1894073010570458e-05,
+      "loss": 0.2968,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4565569347004857,
+      "grad_norm": 1.1957857608795166,
+      "learning_rate": 1.1885489301260056e-05,
+      "loss": 0.3267,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4568267674042094,
+      "grad_norm": 1.5795114040374756,
+      "learning_rate": 1.1876904151269262e-05,
+      "loss": 0.249,
+      "step": 1693
+    },
+    {
+      "epoch": 0.45709660010793307,
+      "grad_norm": 2.022439956665039,
+      "learning_rate": 1.1868317567157894e-05,
+      "loss": 0.2688,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4573664328116568,
+      "grad_norm": 1.3771320581436157,
+      "learning_rate": 1.185972955548686e-05,
+      "loss": 0.2646,
+      "step": 1695
+    },
+    {
+      "epoch": 0.45763626551538045,
+      "grad_norm": 0.8737799525260925,
+      "learning_rate": 1.1851140122818155e-05,
+      "loss": 0.2383,
+      "step": 1696
+    },
+    {
+      "epoch": 0.45790609821910416,
+      "grad_norm": 1.0503333806991577,
+      "learning_rate": 1.1842549275714867e-05,
+      "loss": 0.2514,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4581759309228278,
+      "grad_norm": 1.521773338317871,
+      "learning_rate": 1.1833957020741165e-05,
+      "loss": 0.3131,
+      "step": 1698
+    },
+    {
+      "epoch": 0.45844576362655154,
+      "grad_norm": 1.7766767740249634,
+      "learning_rate": 1.1825363364462286e-05,
+      "loss": 0.2915,
+      "step": 1699
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 1.7590800523757935,
+      "learning_rate": 1.1816768313444541e-05,
+      "loss": 0.2484,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4589854290339989,
+      "grad_norm": 0.9944455623626709,
+      "learning_rate": 1.1808171874255321e-05,
+      "loss": 0.2645,
+      "step": 1701
+    },
+    {
+      "epoch": 0.45925526173772263,
+      "grad_norm": 0.9902780055999756,
+      "learning_rate": 1.1799574053463048e-05,
+      "loss": 0.2812,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4595250944414463,
+      "grad_norm": 2.913714647293091,
+      "learning_rate": 1.1790974857637227e-05,
+      "loss": 0.2743,
+      "step": 1703
+    },
+    {
+      "epoch": 0.45979492714517,
+      "grad_norm": 2.5103585720062256,
+      "learning_rate": 1.17823742933484e-05,
+      "loss": 0.2818,
+      "step": 1704
+    },
+    {
+      "epoch": 0.46006475984889367,
+      "grad_norm": 1.1543667316436768,
+      "learning_rate": 1.177377236716816e-05,
+      "loss": 0.2959,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4603345925526174,
+      "grad_norm": 2.870126724243164,
+      "learning_rate": 1.1765169085669135e-05,
+      "loss": 0.2617,
+      "step": 1706
+    },
+    {
+      "epoch": 0.46060442525634104,
+      "grad_norm": 2.43091082572937,
+      "learning_rate": 1.1756564455424994e-05,
+      "loss": 0.2738,
+      "step": 1707
+    },
+    {
+      "epoch": 0.46087425796006476,
+      "grad_norm": 0.6852809190750122,
+      "learning_rate": 1.1747958483010438e-05,
+      "loss": 0.2646,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4611440906637885,
+      "grad_norm": 2.273150682449341,
+      "learning_rate": 1.1739351175001185e-05,
+      "loss": 0.2789,
+      "step": 1709
+    },
+    {
+      "epoch": 0.46141392336751214,
+      "grad_norm": 1.502650499343872,
+      "learning_rate": 1.173074253797398e-05,
+      "loss": 0.3076,
+      "step": 1710
+    },
+    {
+      "epoch": 0.46168375607123585,
+      "grad_norm": 0.8386364579200745,
+      "learning_rate": 1.1722132578506584e-05,
+      "loss": 0.2618,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4619535887749595,
+      "grad_norm": 2.289910078048706,
+      "learning_rate": 1.1713521303177764e-05,
+      "loss": 0.2447,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4622234214786832,
+      "grad_norm": 2.597780704498291,
+      "learning_rate": 1.1704908718567298e-05,
+      "loss": 0.2576,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4624932541824069,
+      "grad_norm": 3.9267778396606445,
+      "learning_rate": 1.1696294831255961e-05,
+      "loss": 0.304,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4627630868861306,
+      "grad_norm": 1.0608516931533813,
+      "learning_rate": 1.1687679647825522e-05,
+      "loss": 0.2434,
+      "step": 1715
+    },
+    {
+      "epoch": 0.46303291958985426,
+      "grad_norm": 1.2459750175476074,
+      "learning_rate": 1.167906317485874e-05,
+      "loss": 0.3113,
+      "step": 1716
+    },
+    {
+      "epoch": 0.463302752293578,
+      "grad_norm": 4.775570392608643,
+      "learning_rate": 1.1670445418939367e-05,
+      "loss": 0.283,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4635725849973017,
+      "grad_norm": 1.6547666788101196,
+      "learning_rate": 1.1661826386652128e-05,
+      "loss": 0.2725,
+      "step": 1718
+    },
+    {
+      "epoch": 0.46384241770102536,
+      "grad_norm": 2.50020694732666,
+      "learning_rate": 1.1653206084582725e-05,
+      "loss": 0.2795,
+      "step": 1719
+    },
+    {
+      "epoch": 0.46411225040474907,
+      "grad_norm": 2.4893627166748047,
+      "learning_rate": 1.1644584519317828e-05,
+      "loss": 0.2758,
+      "step": 1720
+    },
+    {
+      "epoch": 0.46438208310847273,
+      "grad_norm": 1.1422513723373413,
+      "learning_rate": 1.1635961697445077e-05,
+      "loss": 0.2646,
+      "step": 1721
+    },
+    {
+      "epoch": 0.46465191581219645,
+      "grad_norm": 0.9962664842605591,
+      "learning_rate": 1.1627337625553069e-05,
+      "loss": 0.2748,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4649217485159201,
+      "grad_norm": 2.1642072200775146,
+      "learning_rate": 1.1618712310231355e-05,
+      "loss": 0.2763,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4651915812196438,
+      "grad_norm": 1.4143232107162476,
+      "learning_rate": 1.1610085758070442e-05,
+      "loss": 0.26,
+      "step": 1724
+    },
+    {
+      "epoch": 0.46546141392336754,
+      "grad_norm": 1.2307811975479126,
+      "learning_rate": 1.1601457975661774e-05,
+      "loss": 0.2671,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4657312466270912,
+      "grad_norm": 0.7937163710594177,
+      "learning_rate": 1.159282896959774e-05,
+      "loss": 0.2594,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4660010793308149,
+      "grad_norm": 2.4448554515838623,
+      "learning_rate": 1.1584198746471664e-05,
+      "loss": 0.2987,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4662709120345386,
+      "grad_norm": 1.842415452003479,
+      "learning_rate": 1.1575567312877799e-05,
+      "loss": 0.2788,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4665407447382623,
+      "grad_norm": 1.7508078813552856,
+      "learning_rate": 1.1566934675411322e-05,
+      "loss": 0.2715,
+      "step": 1729
+    },
+    {
+      "epoch": 0.46681057744198595,
+      "grad_norm": 1.2374294996261597,
+      "learning_rate": 1.1558300840668329e-05,
+      "loss": 0.2552,
+      "step": 1730
+    },
+    {
+      "epoch": 0.46708041014570967,
+      "grad_norm": 1.4465796947479248,
+      "learning_rate": 1.1549665815245833e-05,
+      "loss": 0.2848,
+      "step": 1731
+    },
+    {
+      "epoch": 0.46735024284943333,
+      "grad_norm": 0.975462019443512,
+      "learning_rate": 1.1541029605741758e-05,
+      "loss": 0.284,
+      "step": 1732
+    },
+    {
+      "epoch": 0.46762007555315704,
+      "grad_norm": 1.498844861984253,
+      "learning_rate": 1.1532392218754934e-05,
+      "loss": 0.2712,
+      "step": 1733
+    },
+    {
+      "epoch": 0.46788990825688076,
+      "grad_norm": 1.3734400272369385,
+      "learning_rate": 1.152375366088508e-05,
+      "loss": 0.2568,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4681597409606044,
+      "grad_norm": 1.2517064809799194,
+      "learning_rate": 1.151511393873282e-05,
+      "loss": 0.2543,
+      "step": 1735
+    },
+    {
+      "epoch": 0.46842957366432814,
+      "grad_norm": 2.2151060104370117,
+      "learning_rate": 1.150647305889967e-05,
+      "loss": 0.2692,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4686994063680518,
+      "grad_norm": 1.5583680868148804,
+      "learning_rate": 1.149783102798802e-05,
+      "loss": 0.2681,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4689692390717755,
+      "grad_norm": 0.7892103791236877,
+      "learning_rate": 1.1489187852601147e-05,
+      "loss": 0.2642,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4692390717754992,
+      "grad_norm": 2.092423677444458,
+      "learning_rate": 1.1480543539343198e-05,
+      "loss": 0.2539,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4695089044792229,
+      "grad_norm": 2.0396716594696045,
+      "learning_rate": 1.1471898094819198e-05,
+      "loss": 0.3242,
+      "step": 1740
+    },
+    {
+      "epoch": 0.46977873718294655,
+      "grad_norm": 1.0554331541061401,
+      "learning_rate": 1.1463251525635023e-05,
+      "loss": 0.328,
+      "step": 1741
+    },
+    {
+      "epoch": 0.47004856988667026,
+      "grad_norm": 1.705673098564148,
+      "learning_rate": 1.1454603838397423e-05,
+      "loss": 0.2838,
+      "step": 1742
+    },
+    {
+      "epoch": 0.470318402590394,
+      "grad_norm": 3.157578706741333,
+      "learning_rate": 1.1445955039713989e-05,
+      "loss": 0.2819,
+      "step": 1743
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 2.497765302658081,
+      "learning_rate": 1.143730513619317e-05,
+      "loss": 0.2787,
+      "step": 1744
+    },
+    {
+      "epoch": 0.47085806799784136,
+      "grad_norm": 1.6466450691223145,
+      "learning_rate": 1.1428654134444256e-05,
+      "loss": 0.2899,
+      "step": 1745
+    },
+    {
+      "epoch": 0.471127900701565,
+      "grad_norm": 3.530923366546631,
+      "learning_rate": 1.1420002041077376e-05,
+      "loss": 0.2835,
+      "step": 1746
+    },
+    {
+      "epoch": 0.47139773340528873,
+      "grad_norm": 1.959504246711731,
+      "learning_rate": 1.1411348862703496e-05,
+      "loss": 0.308,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4716675661090124,
+      "grad_norm": 0.8502026796340942,
+      "learning_rate": 1.1402694605934406e-05,
+      "loss": 0.2692,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4719373988127361,
+      "grad_norm": 2.766505002975464,
+      "learning_rate": 1.1394039277382722e-05,
+      "loss": 0.2593,
+      "step": 1749
+    },
+    {
+      "epoch": 0.47220723151645977,
+      "grad_norm": 1.799454689025879,
+      "learning_rate": 1.1385382883661881e-05,
+      "loss": 0.3045,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4724770642201835,
+      "grad_norm": 1.1983394622802734,
+      "learning_rate": 1.1376725431386136e-05,
+      "loss": 0.2685,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4727468969239072,
+      "grad_norm": 0.6879530549049377,
+      "learning_rate": 1.1368066927170535e-05,
+      "loss": 0.2575,
+      "step": 1752
+    },
+    {
+      "epoch": 0.47301672962763086,
+      "grad_norm": 2.838103771209717,
+      "learning_rate": 1.1359407377630955e-05,
+      "loss": 0.2834,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4732865623313546,
+      "grad_norm": 2.573545455932617,
+      "learning_rate": 1.1350746789384046e-05,
+      "loss": 0.2786,
+      "step": 1754
+    },
+    {
+      "epoch": 0.47355639503507824,
+      "grad_norm": 1.4904471635818481,
+      "learning_rate": 1.1342085169047263e-05,
+      "loss": 0.2517,
+      "step": 1755
+    },
+    {
+      "epoch": 0.47382622773880195,
+      "grad_norm": 2.3375802040100098,
+      "learning_rate": 1.1333422523238858e-05,
+      "loss": 0.2878,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4740960604425256,
+      "grad_norm": 1.0425461530685425,
+      "learning_rate": 1.1324758858577844e-05,
+      "loss": 0.245,
+      "step": 1757
+    },
+    {
+      "epoch": 0.47436589314624933,
+      "grad_norm": 0.9921801686286926,
+      "learning_rate": 1.1316094181684034e-05,
+      "loss": 0.2462,
+      "step": 1758
+    },
+    {
+      "epoch": 0.474635725849973,
+      "grad_norm": 1.532133936882019,
+      "learning_rate": 1.1307428499178013e-05,
+      "loss": 0.2517,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4749055585536967,
+      "grad_norm": 2.8816311359405518,
+      "learning_rate": 1.1298761817681114e-05,
+      "loss": 0.2742,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4751753912574204,
+      "grad_norm": 0.9000155925750732,
+      "learning_rate": 1.1290094143815455e-05,
+      "loss": 0.2644,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4754452239611441,
+      "grad_norm": 1.0089839696884155,
+      "learning_rate": 1.1281425484203908e-05,
+      "loss": 0.2674,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4757150566648678,
+      "grad_norm": 1.1171197891235352,
+      "learning_rate": 1.1272755845470085e-05,
+      "loss": 0.3097,
+      "step": 1763
+    },
+    {
+      "epoch": 0.47598488936859146,
+      "grad_norm": 2.3801891803741455,
+      "learning_rate": 1.1264085234238363e-05,
+      "loss": 0.269,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4762547220723152,
+      "grad_norm": 0.8138959407806396,
+      "learning_rate": 1.1255413657133855e-05,
+      "loss": 0.3071,
+      "step": 1765
+    },
+    {
+      "epoch": 0.47652455477603883,
+      "grad_norm": 0.9955488443374634,
+      "learning_rate": 1.124674112078241e-05,
+      "loss": 0.2662,
+      "step": 1766
+    },
+    {
+      "epoch": 0.47679438747976255,
+      "grad_norm": 0.9701579213142395,
+      "learning_rate": 1.1238067631810608e-05,
+      "loss": 0.2613,
+      "step": 1767
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 1.0805946588516235,
+      "learning_rate": 1.122939319684577e-05,
+      "loss": 0.2904,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4773340528872099,
+      "grad_norm": 0.9450655579566956,
+      "learning_rate": 1.1220717822515921e-05,
+      "loss": 0.2887,
+      "step": 1769
+    },
+    {
+      "epoch": 0.47760388559093364,
+      "grad_norm": 1.1033295392990112,
+      "learning_rate": 1.1212041515449819e-05,
+      "loss": 0.262,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4778737182946573,
+      "grad_norm": 0.7690514922142029,
+      "learning_rate": 1.1203364282276931e-05,
+      "loss": 0.2852,
+      "step": 1771
+    },
+    {
+      "epoch": 0.478143550998381,
+      "grad_norm": 0.9530160427093506,
+      "learning_rate": 1.1194686129627423e-05,
+      "loss": 0.2886,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4784133837021047,
+      "grad_norm": 1.6271306276321411,
+      "learning_rate": 1.1186007064132176e-05,
+      "loss": 0.2322,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4786832164058284,
+      "grad_norm": 0.6973333954811096,
+      "learning_rate": 1.1177327092422761e-05,
+      "loss": 0.2393,
+      "step": 1774
+    },
+    {
+      "epoch": 0.47895304910955205,
+      "grad_norm": 0.9680898189544678,
+      "learning_rate": 1.116864622113144e-05,
+      "loss": 0.2839,
+      "step": 1775
+    },
+    {
+      "epoch": 0.47922288181327577,
+      "grad_norm": 1.800846815109253,
+      "learning_rate": 1.115996445689117e-05,
+      "loss": 0.2878,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4794927145169995,
+      "grad_norm": 0.9115430116653442,
+      "learning_rate": 1.1151281806335585e-05,
+      "loss": 0.2429,
+      "step": 1777
+    },
+    {
+      "epoch": 0.47976254722072315,
+      "grad_norm": 0.8797100782394409,
+      "learning_rate": 1.1142598276098995e-05,
+      "loss": 0.2866,
+      "step": 1778
+    },
+    {
+      "epoch": 0.48003237992444686,
+      "grad_norm": 1.0538309812545776,
+      "learning_rate": 1.1133913872816384e-05,
+      "loss": 0.2663,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4803022126281705,
+      "grad_norm": 1.9310979843139648,
+      "learning_rate": 1.1125228603123408e-05,
+      "loss": 0.2699,
+      "step": 1780
+    },
+    {
+      "epoch": 0.48057204533189424,
+      "grad_norm": 0.8080328702926636,
+      "learning_rate": 1.1116542473656373e-05,
+      "loss": 0.2567,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4808418780356179,
+      "grad_norm": 1.8147269487380981,
+      "learning_rate": 1.1107855491052258e-05,
+      "loss": 0.2511,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4811117107393416,
+      "grad_norm": 1.965612769126892,
+      "learning_rate": 1.1099167661948678e-05,
+      "loss": 0.284,
+      "step": 1783
+    },
+    {
+      "epoch": 0.4813815434430653,
+      "grad_norm": 1.097743272781372,
+      "learning_rate": 1.10904789929839e-05,
+      "loss": 0.2573,
+      "step": 1784
+    },
+    {
+      "epoch": 0.481651376146789,
+      "grad_norm": 1.652754545211792,
+      "learning_rate": 1.1081789490796846e-05,
+      "loss": 0.2718,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4819212088505127,
+      "grad_norm": 1.8054450750350952,
+      "learning_rate": 1.107309916202705e-05,
+      "loss": 0.2242,
+      "step": 1786
+    },
+    {
+      "epoch": 0.48219104155423637,
+      "grad_norm": 0.9417752027511597,
+      "learning_rate": 1.10644080133147e-05,
+      "loss": 0.293,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4824608742579601,
+      "grad_norm": 1.640665054321289,
+      "learning_rate": 1.10557160513006e-05,
+      "loss": 0.2861,
+      "step": 1788
+    },
+    {
+      "epoch": 0.48273070696168374,
+      "grad_norm": 2.110673189163208,
+      "learning_rate": 1.104702328262617e-05,
+      "loss": 0.2877,
+      "step": 1789
+    },
+    {
+      "epoch": 0.48300053966540746,
+      "grad_norm": 1.0449599027633667,
+      "learning_rate": 1.1038329713933463e-05,
+      "loss": 0.3242,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4832703723691311,
+      "grad_norm": 3.216430425643921,
+      "learning_rate": 1.102963535186513e-05,
+      "loss": 0.2853,
+      "step": 1791
+    },
+    {
+      "epoch": 0.48354020507285483,
+      "grad_norm": 4.057866096496582,
+      "learning_rate": 1.1020940203064425e-05,
+      "loss": 0.2657,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4838100377765785,
+      "grad_norm": 1.0532896518707275,
+      "learning_rate": 1.1012244274175219e-05,
+      "loss": 0.2516,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4840798704803022,
+      "grad_norm": 3.8121988773345947,
+      "learning_rate": 1.1003547571841969e-05,
+      "loss": 0.3046,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4843497031840259,
+      "grad_norm": 3.254439115524292,
+      "learning_rate": 1.0994850102709713e-05,
+      "loss": 0.2781,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4846195358877496,
+      "grad_norm": 1.0805805921554565,
+      "learning_rate": 1.09861518734241e-05,
+      "loss": 0.2586,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4848893685914733,
+      "grad_norm": 3.3879220485687256,
+      "learning_rate": 1.0977452890631339e-05,
+      "loss": 0.2831,
+      "step": 1797
+    },
+    {
+      "epoch": 0.48515920129519696,
+      "grad_norm": 2.4869940280914307,
+      "learning_rate": 1.096875316097822e-05,
+      "loss": 0.2852,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4854290339989207,
+      "grad_norm": 1.0137519836425781,
+      "learning_rate": 1.0960052691112106e-05,
+      "loss": 0.2933,
+      "step": 1799
+    },
+    {
+      "epoch": 0.48569886670264434,
+      "grad_norm": 0.8460010886192322,
+      "learning_rate": 1.095135148768093e-05,
+      "loss": 0.2573,
+      "step": 1800
+    },
+    {
+      "epoch": 0.48596869940636805,
+      "grad_norm": 2.2469210624694824,
+      "learning_rate": 1.094264955733317e-05,
+      "loss": 0.2654,
+      "step": 1801
+    },
+    {
+      "epoch": 0.48623853211009177,
+      "grad_norm": 2.3379745483398438,
+      "learning_rate": 1.0933946906717877e-05,
+      "loss": 0.2474,
+      "step": 1802
+    },
+    {
+      "epoch": 0.48650836481381543,
+      "grad_norm": 1.5732976198196411,
+      "learning_rate": 1.0925243542484642e-05,
+      "loss": 0.2923,
+      "step": 1803
+    },
+    {
+      "epoch": 0.48677819751753915,
+      "grad_norm": 2.1926026344299316,
+      "learning_rate": 1.0916539471283607e-05,
+      "loss": 0.2734,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4870480302212628,
+      "grad_norm": 3.0244252681732178,
+      "learning_rate": 1.0907834699765446e-05,
+      "loss": 0.2916,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4873178629249865,
+      "grad_norm": 1.9182188510894775,
+      "learning_rate": 1.089912923458138e-05,
+      "loss": 0.2815,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4875876956287102,
+      "grad_norm": 0.9087111353874207,
+      "learning_rate": 1.0890423082383147e-05,
+      "loss": 0.2606,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4878575283324339,
+      "grad_norm": 2.225037097930908,
+      "learning_rate": 1.0881716249823022e-05,
+      "loss": 0.2915,
+      "step": 1808
+    },
+    {
+      "epoch": 0.48812736103615756,
+      "grad_norm": 1.5827850103378296,
+      "learning_rate": 1.087300874355379e-05,
+      "loss": 0.2451,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4883971937398813,
+      "grad_norm": 1.0031403303146362,
+      "learning_rate": 1.0864300570228757e-05,
+      "loss": 0.2602,
+      "step": 1810
+    },
+    {
+      "epoch": 0.488667026443605,
+      "grad_norm": 0.7273039817810059,
+      "learning_rate": 1.0855591736501736e-05,
+      "loss": 0.2395,
+      "step": 1811
+    },
+    {
+      "epoch": 0.48893685914732865,
+      "grad_norm": 2.031121015548706,
+      "learning_rate": 1.0846882249027046e-05,
+      "loss": 0.2645,
+      "step": 1812
+    },
+    {
+      "epoch": 0.48920669185105237,
+      "grad_norm": 1.0803141593933105,
+      "learning_rate": 1.0838172114459504e-05,
+      "loss": 0.2634,
+      "step": 1813
+    },
+    {
+      "epoch": 0.489476524554776,
+      "grad_norm": 1.4245915412902832,
+      "learning_rate": 1.0829461339454424e-05,
+      "loss": 0.2667,
+      "step": 1814
+    },
+    {
+      "epoch": 0.48974635725849974,
+      "grad_norm": 1.4667080640792847,
+      "learning_rate": 1.0820749930667607e-05,
+      "loss": 0.2541,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4900161899622234,
+      "grad_norm": 1.165995478630066,
+      "learning_rate": 1.0812037894755336e-05,
+      "loss": 0.2839,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4902860226659471,
+      "grad_norm": 0.8899661898612976,
+      "learning_rate": 1.0803325238374377e-05,
+      "loss": 0.2709,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4905558553696708,
+      "grad_norm": 1.330762267112732,
+      "learning_rate": 1.0794611968181974e-05,
+      "loss": 0.2638,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4908256880733945,
+      "grad_norm": 1.1550347805023193,
+      "learning_rate": 1.0785898090835831e-05,
+      "loss": 0.2417,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4910955207771182,
+      "grad_norm": 1.2064788341522217,
+      "learning_rate": 1.077718361299412e-05,
+      "loss": 0.2802,
+      "step": 1820
+    },
+    {
+      "epoch": 0.49136535348084187,
+      "grad_norm": 1.0787068605422974,
+      "learning_rate": 1.0768468541315471e-05,
+      "loss": 0.2955,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4916351861845656,
+      "grad_norm": 1.585891842842102,
+      "learning_rate": 1.0759752882458972e-05,
+      "loss": 0.2861,
+      "step": 1822
+    },
+    {
+      "epoch": 0.49190501888828925,
+      "grad_norm": 1.304799199104309,
+      "learning_rate": 1.0751036643084154e-05,
+      "loss": 0.2874,
+      "step": 1823
+    },
+    {
+      "epoch": 0.49217485159201296,
+      "grad_norm": 1.6443769931793213,
+      "learning_rate": 1.0742319829850993e-05,
+      "loss": 0.2503,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4924446842957366,
+      "grad_norm": 1.199611783027649,
+      "learning_rate": 1.0733602449419909e-05,
+      "loss": 0.2816,
+      "step": 1825
+    },
+    {
+      "epoch": 0.49271451699946034,
+      "grad_norm": 2.170712471008301,
+      "learning_rate": 1.0724884508451744e-05,
+      "loss": 0.2657,
+      "step": 1826
+    },
+    {
+      "epoch": 0.492984349703184,
+      "grad_norm": 3.0990400314331055,
+      "learning_rate": 1.071616601360778e-05,
+      "loss": 0.3129,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4932541824069077,
+      "grad_norm": 0.9433331489562988,
+      "learning_rate": 1.0707446971549717e-05,
+      "loss": 0.3068,
+      "step": 1828
+    },
+    {
+      "epoch": 0.49352401511063143,
+      "grad_norm": 1.203244924545288,
+      "learning_rate": 1.069872738893967e-05,
+      "loss": 0.2777,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4937938478143551,
+      "grad_norm": 1.2429840564727783,
+      "learning_rate": 1.0690007272440175e-05,
+      "loss": 0.3053,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4940636805180788,
+      "grad_norm": 1.1717705726623535,
+      "learning_rate": 1.0681286628714166e-05,
+      "loss": 0.2644,
+      "step": 1831
+    },
+    {
+      "epoch": 0.49433351322180247,
+      "grad_norm": 3.115442991256714,
+      "learning_rate": 1.0672565464424988e-05,
+      "loss": 0.2855,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4946033459255262,
+      "grad_norm": 1.8445583581924438,
+      "learning_rate": 1.0663843786236384e-05,
+      "loss": 0.2771,
+      "step": 1833
+    },
+    {
+      "epoch": 0.49487317862924984,
+      "grad_norm": 1.4395403861999512,
+      "learning_rate": 1.0655121600812482e-05,
+      "loss": 0.3017,
+      "step": 1834
+    },
+    {
+      "epoch": 0.49514301133297356,
+      "grad_norm": 3.0359134674072266,
+      "learning_rate": 1.0646398914817804e-05,
+      "loss": 0.3,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.8712506890296936,
+      "learning_rate": 1.0637675734917252e-05,
+      "loss": 0.2605,
+      "step": 1836
+    },
+    {
+      "epoch": 0.49568267674042094,
+      "grad_norm": 1.0268560647964478,
+      "learning_rate": 1.0628952067776103e-05,
+      "loss": 0.2928,
+      "step": 1837
+    },
+    {
+      "epoch": 0.49595250944414465,
+      "grad_norm": 2.191091299057007,
+      "learning_rate": 1.062022792006002e-05,
+      "loss": 0.301,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4962223421478683,
+      "grad_norm": 2.003588914871216,
+      "learning_rate": 1.0611503298435011e-05,
+      "loss": 0.2399,
+      "step": 1839
+    },
+    {
+      "epoch": 0.496492174851592,
+      "grad_norm": 0.9982341527938843,
+      "learning_rate": 1.0602778209567462e-05,
+      "loss": 0.3058,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4967620075553157,
+      "grad_norm": 2.5027620792388916,
+      "learning_rate": 1.0594052660124114e-05,
+      "loss": 0.2823,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4970318402590394,
+      "grad_norm": 2.5282483100891113,
+      "learning_rate": 1.0585326656772056e-05,
+      "loss": 0.3325,
+      "step": 1842
+    },
+    {
+      "epoch": 0.49730167296276306,
+      "grad_norm": 0.7055755257606506,
+      "learning_rate": 1.0576600206178717e-05,
+      "loss": 0.2627,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4975715056664868,
+      "grad_norm": 2.4516420364379883,
+      "learning_rate": 1.0567873315011892e-05,
+      "loss": 0.3201,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4978413383702105,
+      "grad_norm": 1.522295355796814,
+      "learning_rate": 1.0559145989939686e-05,
+      "loss": 0.2944,
+      "step": 1845
+    },
+    {
+      "epoch": 0.49811117107393416,
+      "grad_norm": 2.1098620891571045,
+      "learning_rate": 1.0550418237630547e-05,
+      "loss": 0.2788,
+      "step": 1846
+    },
+    {
+      "epoch": 0.49838100377765787,
+      "grad_norm": 0.7936456203460693,
+      "learning_rate": 1.0541690064753253e-05,
+      "loss": 0.2828,
+      "step": 1847
+    },
+    {
+      "epoch": 0.49865083648138153,
+      "grad_norm": 2.506153106689453,
+      "learning_rate": 1.0532961477976898e-05,
+      "loss": 0.2757,
+      "step": 1848
+    },
+    {
+      "epoch": 0.49892066918510525,
+      "grad_norm": 2.464245319366455,
+      "learning_rate": 1.0524232483970891e-05,
+      "loss": 0.2753,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4991905018888289,
+      "grad_norm": 0.8341727256774902,
+      "learning_rate": 1.0515503089404955e-05,
+      "loss": 0.2661,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4994603345925526,
+      "grad_norm": 0.9577743411064148,
+      "learning_rate": 1.0506773300949117e-05,
+      "loss": 0.3161,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4997301672962763,
+      "grad_norm": 2.721142292022705,
+      "learning_rate": 1.0498043125273714e-05,
+      "loss": 0.2965,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.645790934562683,
+      "learning_rate": 1.0489312569049364e-05,
+      "loss": 0.2949,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5002698327037237,
+      "grad_norm": 0.9959818124771118,
+      "learning_rate": 1.0480581638946983e-05,
+      "loss": 0.2867,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5005396654074474,
+      "grad_norm": 2.227278232574463,
+      "learning_rate": 1.047185034163778e-05,
+      "loss": 0.2753,
+      "step": 1855
+    },
+    {
+      "epoch": 0.500809498111171,
+      "grad_norm": 2.1624414920806885,
+      "learning_rate": 1.046311868379323e-05,
+      "loss": 0.3021,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5010793308148948,
+      "grad_norm": 1.0196051597595215,
+      "learning_rate": 1.0454386672085093e-05,
+      "loss": 0.323,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5013491635186185,
+      "grad_norm": 0.8915191292762756,
+      "learning_rate": 1.0445654313185402e-05,
+      "loss": 0.2873,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5016189962223422,
+      "grad_norm": 1.5180319547653198,
+      "learning_rate": 1.0436921613766444e-05,
+      "loss": 0.2774,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5018888289260658,
+      "grad_norm": 1.4688845872879028,
+      "learning_rate": 1.0428188580500773e-05,
+      "loss": 0.3003,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5021586616297895,
+      "grad_norm": 1.3301665782928467,
+      "learning_rate": 1.0419455220061207e-05,
+      "loss": 0.2629,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5024284943335132,
+      "grad_norm": 1.2865246534347534,
+      "learning_rate": 1.0410721539120797e-05,
+      "loss": 0.2297,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5026983270372369,
+      "grad_norm": 0.772615373134613,
+      "learning_rate": 1.0401987544352843e-05,
+      "loss": 0.2605,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5029681597409607,
+      "grad_norm": 0.992931604385376,
+      "learning_rate": 1.0393253242430898e-05,
+      "loss": 0.2592,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5032379924446843,
+      "grad_norm": 3.7021427154541016,
+      "learning_rate": 1.0384518640028733e-05,
+      "loss": 0.3109,
+      "step": 1865
+    },
+    {
+      "epoch": 0.503507825148408,
+      "grad_norm": 1.1142998933792114,
+      "learning_rate": 1.037578374382036e-05,
+      "loss": 0.2475,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5037776578521317,
+      "grad_norm": 2.480664014816284,
+      "learning_rate": 1.0367048560480013e-05,
+      "loss": 0.3042,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5040474905558554,
+      "grad_norm": 3.98327898979187,
+      "learning_rate": 1.0358313096682136e-05,
+      "loss": 0.2538,
+      "step": 1868
+    },
+    {
+      "epoch": 0.504317323259579,
+      "grad_norm": 1.7896349430084229,
+      "learning_rate": 1.0349577359101402e-05,
+      "loss": 0.267,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5045871559633027,
+      "grad_norm": 0.973347544670105,
+      "learning_rate": 1.0340841354412688e-05,
+      "loss": 0.2662,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5048569886670264,
+      "grad_norm": 4.028376579284668,
+      "learning_rate": 1.0332105089291062e-05,
+      "loss": 0.2943,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5051268213707502,
+      "grad_norm": 3.4477927684783936,
+      "learning_rate": 1.0323368570411813e-05,
+      "loss": 0.3071,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5053966540744739,
+      "grad_norm": 0.7780229449272156,
+      "learning_rate": 1.0314631804450415e-05,
+      "loss": 0.2486,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5056664867781975,
+      "grad_norm": 1.9514127969741821,
+      "learning_rate": 1.0305894798082516e-05,
+      "loss": 0.2777,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5059363194819212,
+      "grad_norm": 3.587014675140381,
+      "learning_rate": 1.0297157557983971e-05,
+      "loss": 0.2761,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5062061521856449,
+      "grad_norm": 2.5316450595855713,
+      "learning_rate": 1.0288420090830803e-05,
+      "loss": 0.2643,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5064759848893686,
+      "grad_norm": 1.3704602718353271,
+      "learning_rate": 1.0279682403299205e-05,
+      "loss": 0.274,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5067458175930922,
+      "grad_norm": 3.46244740486145,
+      "learning_rate": 1.0270944502065544e-05,
+      "loss": 0.263,
+      "step": 1878
+    },
+    {
+      "epoch": 0.507015650296816,
+      "grad_norm": 3.69081711769104,
+      "learning_rate": 1.026220639380635e-05,
+      "loss": 0.2747,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5072854830005397,
+      "grad_norm": 2.5850625038146973,
+      "learning_rate": 1.0253468085198306e-05,
+      "loss": 0.2352,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5075553157042634,
+      "grad_norm": 0.8160683512687683,
+      "learning_rate": 1.0244729582918255e-05,
+      "loss": 0.2645,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5078251484079871,
+      "grad_norm": 2.962550640106201,
+      "learning_rate": 1.0235990893643184e-05,
+      "loss": 0.2799,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5080949811117107,
+      "grad_norm": 4.338951110839844,
+      "learning_rate": 1.022725202405022e-05,
+      "loss": 0.3244,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5083648138154344,
+      "grad_norm": 1.1171197891235352,
+      "learning_rate": 1.0218512980816634e-05,
+      "loss": 0.2884,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5086346465191581,
+      "grad_norm": 0.8100233674049377,
+      "learning_rate": 1.0209773770619831e-05,
+      "loss": 0.2721,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5089044792228818,
+      "grad_norm": 1.5780545473098755,
+      "learning_rate": 1.0201034400137335e-05,
+      "loss": 0.2541,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5091743119266054,
+      "grad_norm": 1.9539085626602173,
+      "learning_rate": 1.0192294876046796e-05,
+      "loss": 0.241,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5094441446303292,
+      "grad_norm": 2.0911612510681152,
+      "learning_rate": 1.0183555205025986e-05,
+      "loss": 0.2882,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5097139773340529,
+      "grad_norm": 1.7805238962173462,
+      "learning_rate": 1.0174815393752783e-05,
+      "loss": 0.3023,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5099838100377766,
+      "grad_norm": 1.4473052024841309,
+      "learning_rate": 1.0166075448905177e-05,
+      "loss": 0.301,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5102536427415003,
+      "grad_norm": 2.7354073524475098,
+      "learning_rate": 1.0157335377161256e-05,
+      "loss": 0.2578,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5105234754452239,
+      "grad_norm": 2.145749807357788,
+      "learning_rate": 1.0148595185199208e-05,
+      "loss": 0.2661,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5107933081489476,
+      "grad_norm": 1.6383830308914185,
+      "learning_rate": 1.0139854879697313e-05,
+      "loss": 0.2785,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5110631408526713,
+      "grad_norm": 2.321411371231079,
+      "learning_rate": 1.0131114467333935e-05,
+      "loss": 0.3145,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5113329735563951,
+      "grad_norm": 2.6663870811462402,
+      "learning_rate": 1.0122373954787521e-05,
+      "loss": 0.2826,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5116028062601188,
+      "grad_norm": 3.5124351978302,
+      "learning_rate": 1.0113633348736593e-05,
+      "loss": 0.2934,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5118726389638424,
+      "grad_norm": 0.734468936920166,
+      "learning_rate": 1.0104892655859751e-05,
+      "loss": 0.2518,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5121424716675661,
+      "grad_norm": 1.7605562210083008,
+      "learning_rate": 1.0096151882835654e-05,
+      "loss": 0.2541,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5124123043712898,
+      "grad_norm": 4.232354164123535,
+      "learning_rate": 1.0087411036343022e-05,
+      "loss": 0.2802,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5126821370750135,
+      "grad_norm": 1.5926586389541626,
+      "learning_rate": 1.0078670123060638e-05,
+      "loss": 0.2513,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5129519697787371,
+      "grad_norm": 3.878392219543457,
+      "learning_rate": 1.006992914966733e-05,
+      "loss": 0.3021,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5132218024824609,
+      "grad_norm": 3.891751527786255,
+      "learning_rate": 1.0061188122841975e-05,
+      "loss": 0.3042,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5134916351861846,
+      "grad_norm": 4.581977844238281,
+      "learning_rate": 1.0052447049263488e-05,
+      "loss": 0.3182,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 2.9654715061187744,
+      "learning_rate": 1.0043705935610822e-05,
+      "loss": 0.2672,
+      "step": 1904
+    },
+    {
+      "epoch": 0.514031300593632,
+      "grad_norm": 0.8368027210235596,
+      "learning_rate": 1.003496478856296e-05,
+      "loss": 0.2645,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5143011332973556,
+      "grad_norm": 4.046614646911621,
+      "learning_rate": 1.002622361479891e-05,
+      "loss": 0.3004,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5145709660010793,
+      "grad_norm": 3.70094633102417,
+      "learning_rate": 1.0017482420997704e-05,
+      "loss": 0.3023,
+      "step": 1907
+    },
+    {
+      "epoch": 0.514840798704803,
+      "grad_norm": 1.7231991291046143,
+      "learning_rate": 1.0008741213838382e-05,
+      "loss": 0.2688,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5151106314085268,
+      "grad_norm": 1.4420623779296875,
+      "learning_rate": 1e-05,
+      "loss": 0.3086,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5153804641122504,
+      "grad_norm": 4.643080234527588,
+      "learning_rate": 9.991258786161618e-06,
+      "loss": 0.3387,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5156502968159741,
+      "grad_norm": 0.9490359425544739,
+      "learning_rate": 9.982517579002301e-06,
+      "loss": 0.2508,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5159201295196978,
+      "grad_norm": 1.2798312902450562,
+      "learning_rate": 9.973776385201093e-06,
+      "loss": 0.2689,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5161899622234215,
+      "grad_norm": 0.8091087937355042,
+      "learning_rate": 9.965035211437041e-06,
+      "loss": 0.2783,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5164597949271452,
+      "grad_norm": 1.319119930267334,
+      "learning_rate": 9.956294064389183e-06,
+      "loss": 0.2501,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5167296276308688,
+      "grad_norm": 2.8179919719696045,
+      "learning_rate": 9.947552950736516e-06,
+      "loss": 0.286,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5169994603345925,
+      "grad_norm": 1.760055661201477,
+      "learning_rate": 9.938811877158029e-06,
+      "loss": 0.2836,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5172692930383163,
+      "grad_norm": 1.0237386226654053,
+      "learning_rate": 9.930070850332676e-06,
+      "loss": 0.2761,
+      "step": 1917
+    },
+    {
+      "epoch": 0.51753912574204,
+      "grad_norm": 1.580460548400879,
+      "learning_rate": 9.921329876939365e-06,
+      "loss": 0.2685,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5178089584457636,
+      "grad_norm": 1.436585545539856,
+      "learning_rate": 9.912588963656981e-06,
+      "loss": 0.2297,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5180787911494873,
+      "grad_norm": 2.802029609680176,
+      "learning_rate": 9.903848117164351e-06,
+      "loss": 0.2821,
+      "step": 1920
+    },
+    {
+      "epoch": 0.518348623853211,
+      "grad_norm": 1.7763844728469849,
+      "learning_rate": 9.89510734414025e-06,
+      "loss": 0.2861,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5186184565569347,
+      "grad_norm": 1.4216418266296387,
+      "learning_rate": 9.886366651263409e-06,
+      "loss": 0.2694,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5188882892606584,
+      "grad_norm": 2.754307508468628,
+      "learning_rate": 9.877626045212485e-06,
+      "loss": 0.2515,
+      "step": 1923
+    },
+    {
+      "epoch": 0.519158121964382,
+      "grad_norm": 4.254525184631348,
+      "learning_rate": 9.868885532666068e-06,
+      "loss": 0.2617,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5194279546681058,
+      "grad_norm": 1.132094383239746,
+      "learning_rate": 9.86014512030269e-06,
+      "loss": 0.277,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5196977873718295,
+      "grad_norm": 2.200253963470459,
+      "learning_rate": 9.851404814800792e-06,
+      "loss": 0.2881,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5199676200755532,
+      "grad_norm": 2.914752244949341,
+      "learning_rate": 9.842664622838747e-06,
+      "loss": 0.2938,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5202374527792768,
+      "grad_norm": 1.6566078662872314,
+      "learning_rate": 9.833924551094826e-06,
+      "loss": 0.2631,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5205072854830005,
+      "grad_norm": 1.960695743560791,
+      "learning_rate": 9.825184606247218e-06,
+      "loss": 0.2444,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5207771181867242,
+      "grad_norm": 0.9039238691329956,
+      "learning_rate": 9.816444794974018e-06,
+      "loss": 0.2483,
+      "step": 1930
+    },
+    {
+      "epoch": 0.521046950890448,
+      "grad_norm": 3.4154627323150635,
+      "learning_rate": 9.807705123953205e-06,
+      "loss": 0.264,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5213167835941717,
+      "grad_norm": 2.545193910598755,
+      "learning_rate": 9.798965599862667e-06,
+      "loss": 0.2646,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5215866162978953,
+      "grad_norm": 1.7519617080688477,
+      "learning_rate": 9.790226229380172e-06,
+      "loss": 0.3073,
+      "step": 1933
+    },
+    {
+      "epoch": 0.521856449001619,
+      "grad_norm": 1.40316641330719,
+      "learning_rate": 9.781487019183367e-06,
+      "loss": 0.2916,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5221262817053427,
+      "grad_norm": 1.2153196334838867,
+      "learning_rate": 9.77274797594978e-06,
+      "loss": 0.2645,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5223961144090664,
+      "grad_norm": 2.9920527935028076,
+      "learning_rate": 9.76400910635682e-06,
+      "loss": 0.2867,
+      "step": 1936
+    },
+    {
+      "epoch": 0.52266594711279,
+      "grad_norm": 1.0534918308258057,
+      "learning_rate": 9.755270417081748e-06,
+      "loss": 0.2805,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5229357798165137,
+      "grad_norm": 1.1542798280715942,
+      "learning_rate": 9.746531914801694e-06,
+      "loss": 0.2522,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5232056125202375,
+      "grad_norm": 0.8744438886642456,
+      "learning_rate": 9.737793606193653e-06,
+      "loss": 0.2627,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5234754452239612,
+      "grad_norm": 2.749849557876587,
+      "learning_rate": 9.729055497934457e-06,
+      "loss": 0.2508,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5237452779276849,
+      "grad_norm": 0.6660797595977783,
+      "learning_rate": 9.720317596700795e-06,
+      "loss": 0.2791,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5240151106314085,
+      "grad_norm": 0.7394640445709229,
+      "learning_rate": 9.7115799091692e-06,
+      "loss": 0.272,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5242849433351322,
+      "grad_norm": 1.355002760887146,
+      "learning_rate": 9.70284244201603e-06,
+      "loss": 0.2275,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5245547760388559,
+      "grad_norm": 1.5360618829727173,
+      "learning_rate": 9.694105201917486e-06,
+      "loss": 0.278,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5248246087425796,
+      "grad_norm": 1.2514899969100952,
+      "learning_rate": 9.685368195549592e-06,
+      "loss": 0.2694,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5250944414463032,
+      "grad_norm": 1.069603681564331,
+      "learning_rate": 9.676631429588188e-06,
+      "loss": 0.2633,
+      "step": 1946
+    },
+    {
+      "epoch": 0.525364274150027,
+      "grad_norm": 1.2547860145568848,
+      "learning_rate": 9.667894910708938e-06,
+      "loss": 0.271,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5256341068537507,
+      "grad_norm": 0.8257637619972229,
+      "learning_rate": 9.659158645587319e-06,
+      "loss": 0.2794,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5259039395574744,
+      "grad_norm": 1.3755170106887817,
+      "learning_rate": 9.650422640898601e-06,
+      "loss": 0.3051,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5261737722611981,
+      "grad_norm": 1.1156259775161743,
+      "learning_rate": 9.641686903317864e-06,
+      "loss": 0.2497,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5264436049649217,
+      "grad_norm": 1.074972152709961,
+      "learning_rate": 9.632951439519992e-06,
+      "loss": 0.2732,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5267134376686454,
+      "grad_norm": 0.8749028444290161,
+      "learning_rate": 9.624216256179642e-06,
+      "loss": 0.288,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5269832703723691,
+      "grad_norm": 1.2709335088729858,
+      "learning_rate": 9.615481359971268e-06,
+      "loss": 0.2299,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5272531030760929,
+      "grad_norm": 1.66628098487854,
+      "learning_rate": 9.606746757569107e-06,
+      "loss": 0.2703,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5275229357798165,
+      "grad_norm": 1.249106764793396,
+      "learning_rate": 9.59801245564716e-06,
+      "loss": 0.2552,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5277927684835402,
+      "grad_norm": 0.9405361413955688,
+      "learning_rate": 9.589278460879208e-06,
+      "loss": 0.2676,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5280626011872639,
+      "grad_norm": 1.627439022064209,
+      "learning_rate": 9.580544779938798e-06,
+      "loss": 0.2828,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5283324338909876,
+      "grad_norm": 0.7585939764976501,
+      "learning_rate": 9.571811419499229e-06,
+      "loss": 0.2546,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5286022665947113,
+      "grad_norm": 1.0234673023223877,
+      "learning_rate": 9.56307838623356e-06,
+      "loss": 0.2968,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5288720992984349,
+      "grad_norm": 1.0350896120071411,
+      "learning_rate": 9.554345686814601e-06,
+      "loss": 0.267,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5291419320021586,
+      "grad_norm": 0.8712798953056335,
+      "learning_rate": 9.54561332791491e-06,
+      "loss": 0.2999,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 0.9526864886283875,
+      "learning_rate": 9.536881316206772e-06,
+      "loss": 0.2557,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5296815974096061,
+      "grad_norm": 0.9466046094894409,
+      "learning_rate": 9.528149658362223e-06,
+      "loss": 0.2916,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5299514301133297,
+      "grad_norm": 1.0030120611190796,
+      "learning_rate": 9.51941836105302e-06,
+      "loss": 0.3036,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5302212628170534,
+      "grad_norm": 1.6553846597671509,
+      "learning_rate": 9.51068743095064e-06,
+      "loss": 0.2505,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5304910955207771,
+      "grad_norm": 1.203987956047058,
+      "learning_rate": 9.501956874726289e-06,
+      "loss": 0.286,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5307609282245008,
+      "grad_norm": 0.9015994071960449,
+      "learning_rate": 9.493226699050886e-06,
+      "loss": 0.2984,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5310307609282245,
+      "grad_norm": 0.8772453665733337,
+      "learning_rate": 9.484496910595048e-06,
+      "loss": 0.2758,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5313005936319481,
+      "grad_norm": 1.252602458000183,
+      "learning_rate": 9.475767516029114e-06,
+      "loss": 0.2669,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5315704263356719,
+      "grad_norm": 0.9488521814346313,
+      "learning_rate": 9.467038522023107e-06,
+      "loss": 0.297,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5318402590393956,
+      "grad_norm": 1.2883851528167725,
+      "learning_rate": 9.458309935246748e-06,
+      "loss": 0.2731,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 1.1098257303237915,
+      "learning_rate": 9.449581762369454e-06,
+      "loss": 0.2684,
+      "step": 1972
+    },
+    {
+      "epoch": 0.532379924446843,
+      "grad_norm": 0.7633166909217834,
+      "learning_rate": 9.44085401006032e-06,
+      "loss": 0.2549,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5326497571505666,
+      "grad_norm": 1.8115533590316772,
+      "learning_rate": 9.43212668498811e-06,
+      "loss": 0.2598,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5329195898542903,
+      "grad_norm": 1.0547151565551758,
+      "learning_rate": 9.423399793821283e-06,
+      "loss": 0.2947,
+      "step": 1975
+    },
+    {
+      "epoch": 0.533189422558014,
+      "grad_norm": 0.7810186147689819,
+      "learning_rate": 9.414673343227951e-06,
+      "loss": 0.2878,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5334592552617378,
+      "grad_norm": 0.712790310382843,
+      "learning_rate": 9.40594733987589e-06,
+      "loss": 0.242,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5337290879654614,
+      "grad_norm": 0.7988012433052063,
+      "learning_rate": 9.39722179043254e-06,
+      "loss": 0.2297,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5339989206691851,
+      "grad_norm": 0.8248799443244934,
+      "learning_rate": 9.388496701564994e-06,
+      "loss": 0.2493,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5342687533729088,
+      "grad_norm": 1.3304588794708252,
+      "learning_rate": 9.379772079939983e-06,
+      "loss": 0.2606,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5345385860766325,
+      "grad_norm": 1.5471069812774658,
+      "learning_rate": 9.371047932223897e-06,
+      "loss": 0.2534,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5348084187803562,
+      "grad_norm": 2.872340679168701,
+      "learning_rate": 9.362324265082755e-06,
+      "loss": 0.3018,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5350782514840798,
+      "grad_norm": 1.1670695543289185,
+      "learning_rate": 9.3536010851822e-06,
+      "loss": 0.305,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5353480841878036,
+      "grad_norm": 2.4095921516418457,
+      "learning_rate": 9.344878399187521e-06,
+      "loss": 0.3019,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5356179168915273,
+      "grad_norm": 1.8594398498535156,
+      "learning_rate": 9.336156213763621e-06,
+      "loss": 0.2506,
+      "step": 1985
+    },
+    {
+      "epoch": 0.535887749595251,
+      "grad_norm": 1.2704347372055054,
+      "learning_rate": 9.327434535575014e-06,
+      "loss": 0.2489,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5361575822989746,
+      "grad_norm": 0.9457646608352661,
+      "learning_rate": 9.318713371285837e-06,
+      "loss": 0.2768,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5364274150026983,
+      "grad_norm": 1.4462155103683472,
+      "learning_rate": 9.309992727559831e-06,
+      "loss": 0.2483,
+      "step": 1988
+    },
+    {
+      "epoch": 0.536697247706422,
+      "grad_norm": 0.9293270111083984,
+      "learning_rate": 9.301272611060333e-06,
+      "loss": 0.2396,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5369670804101457,
+      "grad_norm": 2.01761531829834,
+      "learning_rate": 9.292553028450286e-06,
+      "loss": 0.2953,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5372369131138695,
+      "grad_norm": 2.156665086746216,
+      "learning_rate": 9.283833986392225e-06,
+      "loss": 0.2943,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5375067458175931,
+      "grad_norm": 0.9269663095474243,
+      "learning_rate": 9.275115491548257e-06,
+      "loss": 0.258,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5377765785213168,
+      "grad_norm": 1.9710363149642944,
+      "learning_rate": 9.266397550580094e-06,
+      "loss": 0.2278,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5380464112250405,
+      "grad_norm": 1.968768835067749,
+      "learning_rate": 9.257680170149007e-06,
+      "loss": 0.2605,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5383162439287642,
+      "grad_norm": 1.6021437644958496,
+      "learning_rate": 9.248963356915848e-06,
+      "loss": 0.3,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5385860766324878,
+      "grad_norm": 0.8516154289245605,
+      "learning_rate": 9.240247117541031e-06,
+      "loss": 0.2571,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5388559093362115,
+      "grad_norm": 2.3496510982513428,
+      "learning_rate": 9.231531458684529e-06,
+      "loss": 0.2742,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5391257420399352,
+      "grad_norm": 3.2700724601745605,
+      "learning_rate": 9.222816387005884e-06,
+      "loss": 0.2907,
+      "step": 1998
+    },
+    {
+      "epoch": 0.539395574743659,
+      "grad_norm": 1.0170637369155884,
+      "learning_rate": 9.214101909164172e-06,
+      "loss": 0.2581,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5396654074473827,
+      "grad_norm": 0.8277434706687927,
+      "learning_rate": 9.205388031818027e-06,
+      "loss": 0.2565,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5399352401511063,
+      "grad_norm": 1.9436687231063843,
+      "learning_rate": 9.196674761625624e-06,
+      "loss": 0.2748,
+      "step": 2001
+    },
+    {
+      "epoch": 0.54020507285483,
+      "grad_norm": 1.9049378633499146,
+      "learning_rate": 9.187962105244667e-06,
+      "loss": 0.2879,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5404749055585537,
+      "grad_norm": 1.5403751134872437,
+      "learning_rate": 9.179250069332395e-06,
+      "loss": 0.2802,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5407447382622774,
+      "grad_norm": 1.4591472148895264,
+      "learning_rate": 9.170538660545578e-06,
+      "loss": 0.2651,
+      "step": 2004
+    },
+    {
+      "epoch": 0.541014570966001,
+      "grad_norm": 2.2130520343780518,
+      "learning_rate": 9.161827885540498e-06,
+      "loss": 0.2979,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5412844036697247,
+      "grad_norm": 1.2650607824325562,
+      "learning_rate": 9.153117750972954e-06,
+      "loss": 0.2588,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5415542363734485,
+      "grad_norm": 1.2762384414672852,
+      "learning_rate": 9.144408263498267e-06,
+      "loss": 0.2744,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5418240690771722,
+      "grad_norm": 0.8685076236724854,
+      "learning_rate": 9.135699429771245e-06,
+      "loss": 0.2751,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5420939017808959,
+      "grad_norm": 2.2113068103790283,
+      "learning_rate": 9.12699125644621e-06,
+      "loss": 0.2734,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5423637344846195,
+      "grad_norm": 1.0810248851776123,
+      "learning_rate": 9.118283750176982e-06,
+      "loss": 0.2786,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5426335671883432,
+      "grad_norm": 0.8902052044868469,
+      "learning_rate": 9.109576917616854e-06,
+      "loss": 0.2912,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5429033998920669,
+      "grad_norm": 1.242580533027649,
+      "learning_rate": 9.100870765418622e-06,
+      "loss": 0.2467,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5431732325957906,
+      "grad_norm": 1.3017221689224243,
+      "learning_rate": 9.092165300234555e-06,
+      "loss": 0.2846,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5434430652995142,
+      "grad_norm": 1.5688929557800293,
+      "learning_rate": 9.083460528716396e-06,
+      "loss": 0.2789,
+      "step": 2014
+    },
+    {
+      "epoch": 0.543712898003238,
+      "grad_norm": 0.8973223567008972,
+      "learning_rate": 9.074756457515358e-06,
+      "loss": 0.27,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5439827307069617,
+      "grad_norm": 0.7863004803657532,
+      "learning_rate": 9.066053093282125e-06,
+      "loss": 0.2866,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5442525634106854,
+      "grad_norm": 1.5343027114868164,
+      "learning_rate": 9.057350442666832e-06,
+      "loss": 0.2674,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5445223961144091,
+      "grad_norm": 0.8919948935508728,
+      "learning_rate": 9.048648512319072e-06,
+      "loss": 0.2733,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5447922288181327,
+      "grad_norm": 1.1473737955093384,
+      "learning_rate": 9.039947308887896e-06,
+      "loss": 0.2809,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5450620615218564,
+      "grad_norm": 0.8636385202407837,
+      "learning_rate": 9.031246839021783e-06,
+      "loss": 0.2714,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5453318942255801,
+      "grad_norm": 0.7791317701339722,
+      "learning_rate": 9.022547109368663e-06,
+      "loss": 0.2532,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5456017269293039,
+      "grad_norm": 0.9588449001312256,
+      "learning_rate": 9.013848126575902e-06,
+      "loss": 0.2905,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5458715596330275,
+      "grad_norm": 0.7931052446365356,
+      "learning_rate": 9.005149897290289e-06,
+      "loss": 0.2466,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5461413923367512,
+      "grad_norm": 0.7705451846122742,
+      "learning_rate": 8.996452428158036e-06,
+      "loss": 0.246,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5464112250404749,
+      "grad_norm": 1.2396680116653442,
+      "learning_rate": 8.987755725824783e-06,
+      "loss": 0.289,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5466810577441986,
+      "grad_norm": 1.4026050567626953,
+      "learning_rate": 8.979059796935578e-06,
+      "loss": 0.2474,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5469508904479223,
+      "grad_norm": 2.200256586074829,
+      "learning_rate": 8.970364648134874e-06,
+      "loss": 0.2578,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5472207231516459,
+      "grad_norm": 1.0405282974243164,
+      "learning_rate": 8.961670286066539e-06,
+      "loss": 0.2263,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5474905558553697,
+      "grad_norm": 1.068292498588562,
+      "learning_rate": 8.952976717373833e-06,
+      "loss": 0.2829,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5477603885590934,
+      "grad_norm": 1.4279407262802124,
+      "learning_rate": 8.944283948699403e-06,
+      "loss": 0.2467,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5480302212628171,
+      "grad_norm": 1.8920745849609375,
+      "learning_rate": 8.935591986685302e-06,
+      "loss": 0.2882,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5483000539665407,
+      "grad_norm": 0.7814832329750061,
+      "learning_rate": 8.926900837972953e-06,
+      "loss": 0.2461,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5485698866702644,
+      "grad_norm": 2.254251480102539,
+      "learning_rate": 8.918210509203158e-06,
+      "loss": 0.297,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5488397193739881,
+      "grad_norm": 1.0486804246902466,
+      "learning_rate": 8.9095210070161e-06,
+      "loss": 0.2991,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5491095520777118,
+      "grad_norm": 1.52339506149292,
+      "learning_rate": 8.900832338051329e-06,
+      "loss": 0.2798,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5493793847814356,
+      "grad_norm": 1.0586086511611938,
+      "learning_rate": 8.892144508947745e-06,
+      "loss": 0.2827,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5496492174851592,
+      "grad_norm": 1.5417141914367676,
+      "learning_rate": 8.883457526343627e-06,
+      "loss": 0.2784,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5499190501888829,
+      "grad_norm": 1.2257753610610962,
+      "learning_rate": 8.874771396876597e-06,
+      "loss": 0.2608,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5501888828926066,
+      "grad_norm": 0.7058821320533752,
+      "learning_rate": 8.866086127183618e-06,
+      "loss": 0.2478,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 1.8226131200790405,
+      "learning_rate": 8.857401723901006e-06,
+      "loss": 0.2787,
+      "step": 2040
+    },
+    {
+      "epoch": 0.550728548300054,
+      "grad_norm": 0.982736349105835,
+      "learning_rate": 8.84871819366442e-06,
+      "loss": 0.249,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5509983810037776,
+      "grad_norm": 0.8325945138931274,
+      "learning_rate": 8.840035543108834e-06,
+      "loss": 0.2768,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5512682137075013,
+      "grad_norm": 1.030625343322754,
+      "learning_rate": 8.831353778868562e-06,
+      "loss": 0.2687,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5515380464112251,
+      "grad_norm": 1.179280400276184,
+      "learning_rate": 8.822672907577244e-06,
+      "loss": 0.25,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5518078791149488,
+      "grad_norm": 0.9092696905136108,
+      "learning_rate": 8.813992935867829e-06,
+      "loss": 0.3053,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5520777118186724,
+      "grad_norm": 1.1160078048706055,
+      "learning_rate": 8.805313870372578e-06,
+      "loss": 0.2712,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5523475445223961,
+      "grad_norm": 2.031663179397583,
+      "learning_rate": 8.796635717723074e-06,
+      "loss": 0.2513,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5526173772261198,
+      "grad_norm": 1.5580111742019653,
+      "learning_rate": 8.787958484550183e-06,
+      "loss": 0.2946,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5528872099298435,
+      "grad_norm": 0.7856748104095459,
+      "learning_rate": 8.77928217748408e-06,
+      "loss": 0.2885,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5531570426335672,
+      "grad_norm": 1.480365514755249,
+      "learning_rate": 8.770606803154235e-06,
+      "loss": 0.2928,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5534268753372908,
+      "grad_norm": 1.0418368577957153,
+      "learning_rate": 8.761932368189395e-06,
+      "loss": 0.2781,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5536967080410146,
+      "grad_norm": 1.320803165435791,
+      "learning_rate": 8.753258879217594e-06,
+      "loss": 0.265,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5539665407447383,
+      "grad_norm": 0.902825653553009,
+      "learning_rate": 8.744586342866151e-06,
+      "loss": 0.245,
+      "step": 2053
+    },
+    {
+      "epoch": 0.554236373448462,
+      "grad_norm": 0.9388259053230286,
+      "learning_rate": 8.73591476576164e-06,
+      "loss": 0.3099,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5545062061521856,
+      "grad_norm": 1.3321231603622437,
+      "learning_rate": 8.727244154529916e-06,
+      "loss": 0.2594,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5547760388559093,
+      "grad_norm": 1.9373490810394287,
+      "learning_rate": 8.718574515796099e-06,
+      "loss": 0.2467,
+      "step": 2056
+    },
+    {
+      "epoch": 0.555045871559633,
+      "grad_norm": 0.9597958326339722,
+      "learning_rate": 8.709905856184548e-06,
+      "loss": 0.2592,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5553157042633567,
+      "grad_norm": 1.2154154777526855,
+      "learning_rate": 8.701238182318888e-06,
+      "loss": 0.2798,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5555855369670805,
+      "grad_norm": 1.2180938720703125,
+      "learning_rate": 8.692571500821992e-06,
+      "loss": 0.2523,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5558553696708041,
+      "grad_norm": 0.9010483622550964,
+      "learning_rate": 8.683905818315968e-06,
+      "loss": 0.2763,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5561252023745278,
+      "grad_norm": 1.2069810628890991,
+      "learning_rate": 8.675241141422157e-06,
+      "loss": 0.2735,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5563950350782515,
+      "grad_norm": 0.9211894869804382,
+      "learning_rate": 8.666577476761147e-06,
+      "loss": 0.2882,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5566648677819752,
+      "grad_norm": 2.3285701274871826,
+      "learning_rate": 8.657914830952739e-06,
+      "loss": 0.2449,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5569347004856988,
+      "grad_norm": 1.1538158655166626,
+      "learning_rate": 8.649253210615957e-06,
+      "loss": 0.284,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5572045331894225,
+      "grad_norm": 1.5228290557861328,
+      "learning_rate": 8.640592622369047e-06,
+      "loss": 0.2846,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5574743658931463,
+      "grad_norm": 1.2883540391921997,
+      "learning_rate": 8.631933072829466e-06,
+      "loss": 0.2607,
+      "step": 2066
+    },
+    {
+      "epoch": 0.55774419859687,
+      "grad_norm": 0.6729283332824707,
+      "learning_rate": 8.623274568613868e-06,
+      "loss": 0.2562,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5580140313005937,
+      "grad_norm": 1.5743610858917236,
+      "learning_rate": 8.61461711633812e-06,
+      "loss": 0.3122,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5582838640043173,
+      "grad_norm": 1.2146888971328735,
+      "learning_rate": 8.605960722617284e-06,
+      "loss": 0.239,
+      "step": 2069
+    },
+    {
+      "epoch": 0.558553696708041,
+      "grad_norm": 0.9326139688491821,
+      "learning_rate": 8.597305394065597e-06,
+      "loss": 0.2579,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5588235294117647,
+      "grad_norm": 1.2776931524276733,
+      "learning_rate": 8.588651137296508e-06,
+      "loss": 0.2593,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5590933621154884,
+      "grad_norm": 2.162252902984619,
+      "learning_rate": 8.579997958922628e-06,
+      "loss": 0.2778,
+      "step": 2072
+    },
+    {
+      "epoch": 0.559363194819212,
+      "grad_norm": 1.8505454063415527,
+      "learning_rate": 8.571345865555747e-06,
+      "loss": 0.2643,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5596330275229358,
+      "grad_norm": 0.9259780645370483,
+      "learning_rate": 8.562694863806833e-06,
+      "loss": 0.28,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5599028602266595,
+      "grad_norm": 0.8102869391441345,
+      "learning_rate": 8.554044960286016e-06,
+      "loss": 0.2696,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5601726929303832,
+      "grad_norm": 1.6406450271606445,
+      "learning_rate": 8.545396161602579e-06,
+      "loss": 0.2579,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5604425256341069,
+      "grad_norm": 2.1807913780212402,
+      "learning_rate": 8.536748474364978e-06,
+      "loss": 0.2701,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5607123583378305,
+      "grad_norm": 3.136784315109253,
+      "learning_rate": 8.528101905180809e-06,
+      "loss": 0.3278,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5609821910415542,
+      "grad_norm": 1.2836761474609375,
+      "learning_rate": 8.519456460656803e-06,
+      "loss": 0.2778,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5612520237452779,
+      "grad_norm": 2.8312458992004395,
+      "learning_rate": 8.510812147398857e-06,
+      "loss": 0.256,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5615218564490017,
+      "grad_norm": 2.603703737258911,
+      "learning_rate": 8.502168972011987e-06,
+      "loss": 0.2383,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5617916891527253,
+      "grad_norm": 1.8114206790924072,
+      "learning_rate": 8.493526941100334e-06,
+      "loss": 0.2538,
+      "step": 2082
+    },
+    {
+      "epoch": 0.562061521856449,
+      "grad_norm": 2.266664505004883,
+      "learning_rate": 8.484886061267181e-06,
+      "loss": 0.2961,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5623313545601727,
+      "grad_norm": 2.5040090084075928,
+      "learning_rate": 8.476246339114927e-06,
+      "loss": 0.2709,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5626011872638964,
+      "grad_norm": 1.0933821201324463,
+      "learning_rate": 8.467607781245071e-06,
+      "loss": 0.2752,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5628710199676201,
+      "grad_norm": 1.5068005323410034,
+      "learning_rate": 8.458970394258244e-06,
+      "loss": 0.2479,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5631408526713437,
+      "grad_norm": 1.1946412324905396,
+      "learning_rate": 8.450334184754172e-06,
+      "loss": 0.2599,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5634106853750674,
+      "grad_norm": 1.7061984539031982,
+      "learning_rate": 8.441699159331676e-06,
+      "loss": 0.3206,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5636805180787912,
+      "grad_norm": 1.6053333282470703,
+      "learning_rate": 8.433065324588683e-06,
+      "loss": 0.2656,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5639503507825149,
+      "grad_norm": 0.9481534361839294,
+      "learning_rate": 8.424432687122206e-06,
+      "loss": 0.3229,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5642201834862385,
+      "grad_norm": 1.1614317893981934,
+      "learning_rate": 8.415801253528338e-06,
+      "loss": 0.2839,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5644900161899622,
+      "grad_norm": 0.8945530652999878,
+      "learning_rate": 8.407171030402263e-06,
+      "loss": 0.273,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5647598488936859,
+      "grad_norm": 1.2546594142913818,
+      "learning_rate": 8.398542024338227e-06,
+      "loss": 0.2744,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5650296815974096,
+      "grad_norm": 0.9939213395118713,
+      "learning_rate": 8.389914241929561e-06,
+      "loss": 0.2749,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5652995143011333,
+      "grad_norm": 0.8794467449188232,
+      "learning_rate": 8.381287689768646e-06,
+      "loss": 0.2666,
+      "step": 2095
+    },
+    {
+      "epoch": 0.565569347004857,
+      "grad_norm": 1.801514744758606,
+      "learning_rate": 8.372662374446933e-06,
+      "loss": 0.2968,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5658391797085807,
+      "grad_norm": 0.8529360294342041,
+      "learning_rate": 8.364038302554926e-06,
+      "loss": 0.2524,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5661090124123044,
+      "grad_norm": 1.2572522163391113,
+      "learning_rate": 8.355415480682176e-06,
+      "loss": 0.2576,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5663788451160281,
+      "grad_norm": 0.9506404995918274,
+      "learning_rate": 8.346793915417277e-06,
+      "loss": 0.3007,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5666486778197517,
+      "grad_norm": 0.8658832311630249,
+      "learning_rate": 8.338173613347876e-06,
+      "loss": 0.2901,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5669185105234754,
+      "grad_norm": 1.0784802436828613,
+      "learning_rate": 8.329554581060634e-06,
+      "loss": 0.2781,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5671883432271991,
+      "grad_norm": 1.1565532684326172,
+      "learning_rate": 8.320936825141259e-06,
+      "loss": 0.2707,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5674581759309228,
+      "grad_norm": 1.3576381206512451,
+      "learning_rate": 8.312320352174483e-06,
+      "loss": 0.2546,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5677280086346466,
+      "grad_norm": 1.4048036336898804,
+      "learning_rate": 8.303705168744042e-06,
+      "loss": 0.2451,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5679978413383702,
+      "grad_norm": 0.9486166834831238,
+      "learning_rate": 8.295091281432702e-06,
+      "loss": 0.2735,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5682676740420939,
+      "grad_norm": 0.9981785416603088,
+      "learning_rate": 8.286478696822238e-06,
+      "loss": 0.2457,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5685375067458176,
+      "grad_norm": 0.9450013637542725,
+      "learning_rate": 8.277867421493419e-06,
+      "loss": 0.2508,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 1.0122876167297363,
+      "learning_rate": 8.26925746202602e-06,
+      "loss": 0.2386,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5690771721532649,
+      "grad_norm": 2.072953701019287,
+      "learning_rate": 8.260648824998818e-06,
+      "loss": 0.2924,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5693470048569886,
+      "grad_norm": 1.4291988611221313,
+      "learning_rate": 8.252041516989565e-06,
+      "loss": 0.2562,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5696168375607124,
+      "grad_norm": 0.9888383150100708,
+      "learning_rate": 8.243435544575004e-06,
+      "loss": 0.3093,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5698866702644361,
+      "grad_norm": 2.5357089042663574,
+      "learning_rate": 8.234830914330868e-06,
+      "loss": 0.2382,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5701565029681598,
+      "grad_norm": 1.7603410482406616,
+      "learning_rate": 8.226227632831844e-06,
+      "loss": 0.2923,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5704263356718834,
+      "grad_norm": 2.312013864517212,
+      "learning_rate": 8.217625706651601e-06,
+      "loss": 0.2318,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5706961683756071,
+      "grad_norm": 2.5861153602600098,
+      "learning_rate": 8.209025142362776e-06,
+      "loss": 0.2972,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5709660010793308,
+      "grad_norm": 2.533484697341919,
+      "learning_rate": 8.200425946536956e-06,
+      "loss": 0.2973,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5712358337830545,
+      "grad_norm": 0.8722811937332153,
+      "learning_rate": 8.191828125744684e-06,
+      "loss": 0.2636,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5715056664867783,
+      "grad_norm": 1.4765410423278809,
+      "learning_rate": 8.18323168655546e-06,
+      "loss": 0.3236,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5717754991905019,
+      "grad_norm": 1.56089186668396,
+      "learning_rate": 8.174636635537719e-06,
+      "loss": 0.2936,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5720453318942256,
+      "grad_norm": 2.328164577484131,
+      "learning_rate": 8.166042979258837e-06,
+      "loss": 0.2853,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5723151645979493,
+      "grad_norm": 0.7223979234695435,
+      "learning_rate": 8.157450724285137e-06,
+      "loss": 0.2747,
+      "step": 2121
+    },
+    {
+      "epoch": 0.572584997301673,
+      "grad_norm": 1.0296223163604736,
+      "learning_rate": 8.148859877181849e-06,
+      "loss": 0.2597,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5728548300053966,
+      "grad_norm": 0.9047951698303223,
+      "learning_rate": 8.140270444513142e-06,
+      "loss": 0.2411,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5731246627091203,
+      "grad_norm": 1.7256417274475098,
+      "learning_rate": 8.131682432842108e-06,
+      "loss": 0.2497,
+      "step": 2124
+    },
+    {
+      "epoch": 0.573394495412844,
+      "grad_norm": 2.768732786178589,
+      "learning_rate": 8.12309584873074e-06,
+      "loss": 0.2631,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5736643281165678,
+      "grad_norm": 1.042403221130371,
+      "learning_rate": 8.114510698739947e-06,
+      "loss": 0.2558,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5739341608202915,
+      "grad_norm": 0.7608386278152466,
+      "learning_rate": 8.105926989429544e-06,
+      "loss": 0.2711,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5742039935240151,
+      "grad_norm": 2.185546875,
+      "learning_rate": 8.097344727358247e-06,
+      "loss": 0.2629,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5744738262277388,
+      "grad_norm": 1.7382022142410278,
+      "learning_rate": 8.088763919083653e-06,
+      "loss": 0.2456,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5747436589314625,
+      "grad_norm": 1.396360158920288,
+      "learning_rate": 8.080184571162264e-06,
+      "loss": 0.2958,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5750134916351862,
+      "grad_norm": 1.2695201635360718,
+      "learning_rate": 8.071606690149461e-06,
+      "loss": 0.263,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5752833243389098,
+      "grad_norm": 0.9906688332557678,
+      "learning_rate": 8.063030282599496e-06,
+      "loss": 0.2912,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5755531570426335,
+      "grad_norm": 1.4720076322555542,
+      "learning_rate": 8.054455355065511e-06,
+      "loss": 0.276,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5758229897463573,
+      "grad_norm": 1.246246337890625,
+      "learning_rate": 8.045881914099503e-06,
+      "loss": 0.2778,
+      "step": 2134
+    },
+    {
+      "epoch": 0.576092822450081,
+      "grad_norm": 1.080122709274292,
+      "learning_rate": 8.037309966252333e-06,
+      "loss": 0.2741,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5763626551538047,
+      "grad_norm": 0.872567892074585,
+      "learning_rate": 8.028739518073738e-06,
+      "loss": 0.2895,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5766324878575283,
+      "grad_norm": 0.770826518535614,
+      "learning_rate": 8.020170576112291e-06,
+      "loss": 0.2353,
+      "step": 2137
+    },
+    {
+      "epoch": 0.576902320561252,
+      "grad_norm": 1.2850362062454224,
+      "learning_rate": 8.011603146915418e-06,
+      "loss": 0.2401,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5771721532649757,
+      "grad_norm": 1.5641871690750122,
+      "learning_rate": 8.003037237029396e-06,
+      "loss": 0.2727,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5774419859686994,
+      "grad_norm": 1.339402437210083,
+      "learning_rate": 7.99447285299934e-06,
+      "loss": 0.2472,
+      "step": 2140
+    },
+    {
+      "epoch": 0.577711818672423,
+      "grad_norm": 1.3952178955078125,
+      "learning_rate": 7.985910001369186e-06,
+      "loss": 0.2885,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5779816513761468,
+      "grad_norm": 1.2408580780029297,
+      "learning_rate": 7.977348688681719e-06,
+      "loss": 0.2734,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5782514840798705,
+      "grad_norm": 1.3924349546432495,
+      "learning_rate": 7.968788921478534e-06,
+      "loss": 0.2835,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5785213167835942,
+      "grad_norm": 1.786377191543579,
+      "learning_rate": 7.960230706300047e-06,
+      "loss": 0.2663,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5787911494873179,
+      "grad_norm": 0.8870909214019775,
+      "learning_rate": 7.951674049685493e-06,
+      "loss": 0.2444,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5790609821910415,
+      "grad_norm": 1.971816062927246,
+      "learning_rate": 7.943118958172917e-06,
+      "loss": 0.2476,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5793308148947652,
+      "grad_norm": 1.753965139389038,
+      "learning_rate": 7.934565438299156e-06,
+      "loss": 0.2683,
+      "step": 2147
+    },
+    {
+      "epoch": 0.579600647598489,
+      "grad_norm": 1.5364638566970825,
+      "learning_rate": 7.926013496599864e-06,
+      "loss": 0.2515,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5798704803022127,
+      "grad_norm": 0.968102216720581,
+      "learning_rate": 7.917463139609475e-06,
+      "loss": 0.2744,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5801403130059363,
+      "grad_norm": 1.5126365423202515,
+      "learning_rate": 7.908914373861217e-06,
+      "loss": 0.243,
+      "step": 2150
+    },
+    {
+      "epoch": 0.58041014570966,
+      "grad_norm": 0.8365370631217957,
+      "learning_rate": 7.900367205887102e-06,
+      "loss": 0.2671,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5806799784133837,
+      "grad_norm": 0.8668344616889954,
+      "learning_rate": 7.891821642217926e-06,
+      "loss": 0.2613,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5809498111171074,
+      "grad_norm": 1.5775479078292847,
+      "learning_rate": 7.883277689383248e-06,
+      "loss": 0.2756,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5812196438208311,
+      "grad_norm": 1.465959072113037,
+      "learning_rate": 7.874735353911406e-06,
+      "loss": 0.2485,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5814894765245547,
+      "grad_norm": 2.1123433113098145,
+      "learning_rate": 7.866194642329502e-06,
+      "loss": 0.2974,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5817593092282785,
+      "grad_norm": 1.380141019821167,
+      "learning_rate": 7.857655561163386e-06,
+      "loss": 0.2447,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5820291419320022,
+      "grad_norm": 0.9314815402030945,
+      "learning_rate": 7.849118116937677e-06,
+      "loss": 0.2619,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5822989746357259,
+      "grad_norm": 1.2609533071517944,
+      "learning_rate": 7.840582316175737e-06,
+      "loss": 0.2433,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5825688073394495,
+      "grad_norm": 1.6926270723342896,
+      "learning_rate": 7.832048165399666e-06,
+      "loss": 0.2851,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5828386400431732,
+      "grad_norm": 0.947239875793457,
+      "learning_rate": 7.823515671130315e-06,
+      "loss": 0.2943,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5831084727468969,
+      "grad_norm": 0.9032384157180786,
+      "learning_rate": 7.814984839887259e-06,
+      "loss": 0.3039,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5833783054506206,
+      "grad_norm": 2.483617067337036,
+      "learning_rate": 7.806455678188812e-06,
+      "loss": 0.2876,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5836481381543444,
+      "grad_norm": 1.2645118236541748,
+      "learning_rate": 7.797928192552002e-06,
+      "loss": 0.2609,
+      "step": 2163
+    },
+    {
+      "epoch": 0.583917970858068,
+      "grad_norm": 0.8077886700630188,
+      "learning_rate": 7.789402389492582e-06,
+      "loss": 0.245,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5841878035617917,
+      "grad_norm": 1.146843433380127,
+      "learning_rate": 7.780878275525025e-06,
+      "loss": 0.2503,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5844576362655154,
+      "grad_norm": 2.3331503868103027,
+      "learning_rate": 7.772355857162496e-06,
+      "loss": 0.2326,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5847274689692391,
+      "grad_norm": 1.2163803577423096,
+      "learning_rate": 7.763835140916883e-06,
+      "loss": 0.2679,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5849973016729627,
+      "grad_norm": 2.6890928745269775,
+      "learning_rate": 7.75531613329877e-06,
+      "loss": 0.3185,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5852671343766864,
+      "grad_norm": 0.9884248375892639,
+      "learning_rate": 7.746798840817417e-06,
+      "loss": 0.285,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5855369670804101,
+      "grad_norm": 1.7594841718673706,
+      "learning_rate": 7.738283269980798e-06,
+      "loss": 0.2887,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5858067997841339,
+      "grad_norm": 4.495017051696777,
+      "learning_rate": 7.729769427295567e-06,
+      "loss": 0.2655,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5860766324878576,
+      "grad_norm": 1.4083465337753296,
+      "learning_rate": 7.721257319267039e-06,
+      "loss": 0.3096,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5863464651915812,
+      "grad_norm": 0.7818588614463806,
+      "learning_rate": 7.712746952399223e-06,
+      "loss": 0.2798,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5866162978953049,
+      "grad_norm": 1.9579530954360962,
+      "learning_rate": 7.704238333194793e-06,
+      "loss": 0.268,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5868861305990286,
+      "grad_norm": 2.4031636714935303,
+      "learning_rate": 7.695731468155079e-06,
+      "loss": 0.2747,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 2.389094352722168,
+      "learning_rate": 7.687226363780084e-06,
+      "loss": 0.2495,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5874257960064759,
+      "grad_norm": 3.0080816745758057,
+      "learning_rate": 7.678723026568458e-06,
+      "loss": 0.2945,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5876956287101996,
+      "grad_norm": 1.250423789024353,
+      "learning_rate": 7.670221463017498e-06,
+      "loss": 0.2504,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5879654614139234,
+      "grad_norm": 3.044999122619629,
+      "learning_rate": 7.661721679623152e-06,
+      "loss": 0.2732,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 3.1750996112823486,
+      "learning_rate": 7.65322368288001e-06,
+      "loss": 0.2836,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5885051268213708,
+      "grad_norm": 2.5236637592315674,
+      "learning_rate": 7.644727479281281e-06,
+      "loss": 0.2358,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5887749595250944,
+      "grad_norm": 0.7176179885864258,
+      "learning_rate": 7.636233075318824e-06,
+      "loss": 0.2633,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5890447922288181,
+      "grad_norm": 2.5804672241210938,
+      "learning_rate": 7.6277404774831154e-06,
+      "loss": 0.2868,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5893146249325418,
+      "grad_norm": 2.872629165649414,
+      "learning_rate": 7.619249692263242e-06,
+      "loss": 0.2478,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5895844576362655,
+      "grad_norm": 2.415127754211426,
+      "learning_rate": 7.61076072614692e-06,
+      "loss": 0.2904,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5898542903399892,
+      "grad_norm": 1.3119151592254639,
+      "learning_rate": 7.60227358562047e-06,
+      "loss": 0.2727,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5901241230437129,
+      "grad_norm": 2.760488271713257,
+      "learning_rate": 7.59378827716881e-06,
+      "loss": 0.309,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5903939557474366,
+      "grad_norm": 5.0148396492004395,
+      "learning_rate": 7.585304807275473e-06,
+      "loss": 0.3137,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5906637884511603,
+      "grad_norm": 2.893136978149414,
+      "learning_rate": 7.576823182422579e-06,
+      "loss": 0.2643,
+      "step": 2189
+    },
+    {
+      "epoch": 0.590933621154884,
+      "grad_norm": 1.1754097938537598,
+      "learning_rate": 7.568343409090832e-06,
+      "loss": 0.2513,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5912034538586076,
+      "grad_norm": 3.4868860244750977,
+      "learning_rate": 7.559865493759533e-06,
+      "loss": 0.2982,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5914732865623313,
+      "grad_norm": 2.6867358684539795,
+      "learning_rate": 7.551389442906556e-06,
+      "loss": 0.2513,
+      "step": 2192
+    },
+    {
+      "epoch": 0.591743119266055,
+      "grad_norm": 2.357210397720337,
+      "learning_rate": 7.542915263008358e-06,
+      "loss": 0.2723,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5920129519697788,
+      "grad_norm": 1.4523780345916748,
+      "learning_rate": 7.534442960539956e-06,
+      "loss": 0.2565,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5922827846735025,
+      "grad_norm": 1.292918086051941,
+      "learning_rate": 7.525972541974936e-06,
+      "loss": 0.2724,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5925526173772261,
+      "grad_norm": 3.4017155170440674,
+      "learning_rate": 7.517504013785452e-06,
+      "loss": 0.2655,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5928224500809498,
+      "grad_norm": 2.935624837875366,
+      "learning_rate": 7.509037382442203e-06,
+      "loss": 0.2693,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5930922827846735,
+      "grad_norm": 1.1360735893249512,
+      "learning_rate": 7.500572654414442e-06,
+      "loss": 0.2558,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5933621154883972,
+      "grad_norm": 0.9711629152297974,
+      "learning_rate": 7.492109836169975e-06,
+      "loss": 0.2396,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5936319481921208,
+      "grad_norm": 2.6750757694244385,
+      "learning_rate": 7.483648934175138e-06,
+      "loss": 0.2642,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5939017808958446,
+      "grad_norm": 2.7978060245513916,
+      "learning_rate": 7.475189954894807e-06,
+      "loss": 0.2943,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5941716135995683,
+      "grad_norm": 2.373189926147461,
+      "learning_rate": 7.466732904792395e-06,
+      "loss": 0.2791,
+      "step": 2202
+    },
+    {
+      "epoch": 0.594441446303292,
+      "grad_norm": 0.7604328393936157,
+      "learning_rate": 7.458277790329831e-06,
+      "loss": 0.233,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5947112790070157,
+      "grad_norm": 2.893235206604004,
+      "learning_rate": 7.449824617967566e-06,
+      "loss": 0.2388,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5949811117107393,
+      "grad_norm": 3.570953607559204,
+      "learning_rate": 7.44137339416458e-06,
+      "loss": 0.2587,
+      "step": 2205
+    },
+    {
+      "epoch": 0.595250944414463,
+      "grad_norm": 2.4148929119110107,
+      "learning_rate": 7.432924125378345e-06,
+      "loss": 0.2752,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5955207771181867,
+      "grad_norm": 1.887665867805481,
+      "learning_rate": 7.424476818064852e-06,
+      "loss": 0.2526,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5957906098219105,
+      "grad_norm": 2.7964742183685303,
+      "learning_rate": 7.416031478678596e-06,
+      "loss": 0.2678,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5960604425256341,
+      "grad_norm": 1.5072870254516602,
+      "learning_rate": 7.407588113672552e-06,
+      "loss": 0.2609,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5963302752293578,
+      "grad_norm": 1.1576237678527832,
+      "learning_rate": 7.399146729498203e-06,
+      "loss": 0.2673,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5966001079330815,
+      "grad_norm": 1.6817265748977661,
+      "learning_rate": 7.3907073326055135e-06,
+      "loss": 0.3011,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5968699406368052,
+      "grad_norm": 2.118372917175293,
+      "learning_rate": 7.382269929442925e-06,
+      "loss": 0.2602,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5971397733405289,
+      "grad_norm": 1.247247576713562,
+      "learning_rate": 7.373834526457358e-06,
+      "loss": 0.277,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5974096060442525,
+      "grad_norm": 1.089403748512268,
+      "learning_rate": 7.36540113009421e-06,
+      "loss": 0.2759,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5976794387479762,
+      "grad_norm": 1.3078670501708984,
+      "learning_rate": 7.356969746797335e-06,
+      "loss": 0.2984,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5979492714517,
+      "grad_norm": 0.8953466415405273,
+      "learning_rate": 7.348540383009057e-06,
+      "loss": 0.2765,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5982191041554237,
+      "grad_norm": 0.8842044472694397,
+      "learning_rate": 7.3401130451701605e-06,
+      "loss": 0.287,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5984889368591473,
+      "grad_norm": 0.8571217060089111,
+      "learning_rate": 7.331687739719868e-06,
+      "loss": 0.2752,
+      "step": 2218
+    },
+    {
+      "epoch": 0.598758769562871,
+      "grad_norm": 1.4221107959747314,
+      "learning_rate": 7.323264473095859e-06,
+      "loss": 0.2435,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5990286022665947,
+      "grad_norm": 1.4553788900375366,
+      "learning_rate": 7.314843251734257e-06,
+      "loss": 0.2686,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5992984349703184,
+      "grad_norm": 0.7243186831474304,
+      "learning_rate": 7.306424082069615e-06,
+      "loss": 0.3021,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5995682676740421,
+      "grad_norm": 0.8463066816329956,
+      "learning_rate": 7.298006970534922e-06,
+      "loss": 0.2493,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5998381003777657,
+      "grad_norm": 0.8127794861793518,
+      "learning_rate": 7.289591923561601e-06,
+      "loss": 0.2748,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6001079330814895,
+      "grad_norm": 1.232803463935852,
+      "learning_rate": 7.281178947579484e-06,
+      "loss": 0.2515,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6003777657852132,
+      "grad_norm": 0.8982816934585571,
+      "learning_rate": 7.272768049016829e-06,
+      "loss": 0.2683,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6006475984889369,
+      "grad_norm": 0.7264969348907471,
+      "learning_rate": 7.264359234300309e-06,
+      "loss": 0.2986,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6009174311926605,
+      "grad_norm": 1.1590924263000488,
+      "learning_rate": 7.255952509855002e-06,
+      "loss": 0.2662,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6011872638963842,
+      "grad_norm": 1.2683281898498535,
+      "learning_rate": 7.247547882104382e-06,
+      "loss": 0.2704,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6014570966001079,
+      "grad_norm": 0.8444262742996216,
+      "learning_rate": 7.239145357470332e-06,
+      "loss": 0.2598,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6017269293038316,
+      "grad_norm": 1.5139031410217285,
+      "learning_rate": 7.230744942373125e-06,
+      "loss": 0.2886,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6019967620075554,
+      "grad_norm": 1.161885380744934,
+      "learning_rate": 7.2223466432314104e-06,
+      "loss": 0.2718,
+      "step": 2231
+    },
+    {
+      "epoch": 0.602266594711279,
+      "grad_norm": 1.5684939622879028,
+      "learning_rate": 7.213950466462241e-06,
+      "loss": 0.2801,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6025364274150027,
+      "grad_norm": 1.6250052452087402,
+      "learning_rate": 7.205556418481037e-06,
+      "loss": 0.2997,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6028062601187264,
+      "grad_norm": 2.5640835762023926,
+      "learning_rate": 7.197164505701584e-06,
+      "loss": 0.3013,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6030760928224501,
+      "grad_norm": 0.8120871782302856,
+      "learning_rate": 7.188774734536051e-06,
+      "loss": 0.2431,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6033459255261737,
+      "grad_norm": 0.9405423998832703,
+      "learning_rate": 7.1803871113949675e-06,
+      "loss": 0.2689,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6036157582298974,
+      "grad_norm": 0.7726659178733826,
+      "learning_rate": 7.17200164268721e-06,
+      "loss": 0.2571,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6038855909336212,
+      "grad_norm": 0.9076669812202454,
+      "learning_rate": 7.163618334820024e-06,
+      "loss": 0.2282,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6041554236373449,
+      "grad_norm": 0.9585376977920532,
+      "learning_rate": 7.1552371941989966e-06,
+      "loss": 0.2822,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6044252563410686,
+      "grad_norm": 1.798622727394104,
+      "learning_rate": 7.146858227228053e-06,
+      "loss": 0.2506,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6046950890447922,
+      "grad_norm": 1.125195026397705,
+      "learning_rate": 7.138481440309473e-06,
+      "loss": 0.2634,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6049649217485159,
+      "grad_norm": 0.9385486245155334,
+      "learning_rate": 7.13010683984386e-06,
+      "loss": 0.29,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6052347544522396,
+      "grad_norm": 1.8910548686981201,
+      "learning_rate": 7.1217344322301426e-06,
+      "loss": 0.2517,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 2.2433581352233887,
+      "learning_rate": 7.113364223865585e-06,
+      "loss": 0.259,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6057744198596869,
+      "grad_norm": 0.9464813470840454,
+      "learning_rate": 7.104996221145768e-06,
+      "loss": 0.2565,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6060442525634107,
+      "grad_norm": 2.5237033367156982,
+      "learning_rate": 7.096630430464577e-06,
+      "loss": 0.2968,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6063140852671344,
+      "grad_norm": 2.3155696392059326,
+      "learning_rate": 7.088266858214221e-06,
+      "loss": 0.2935,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6065839179708581,
+      "grad_norm": 0.7948070764541626,
+      "learning_rate": 7.07990551078521e-06,
+      "loss": 0.2713,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6068537506745818,
+      "grad_norm": 1.4712588787078857,
+      "learning_rate": 7.071546394566344e-06,
+      "loss": 0.2549,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6071235833783054,
+      "grad_norm": 1.3041421175003052,
+      "learning_rate": 7.063189515944735e-06,
+      "loss": 0.2419,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6073934160820291,
+      "grad_norm": 1.5080175399780273,
+      "learning_rate": 7.054834881305772e-06,
+      "loss": 0.2465,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6076632487857528,
+      "grad_norm": 0.7494954466819763,
+      "learning_rate": 7.046482497033131e-06,
+      "loss": 0.275,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6079330814894766,
+      "grad_norm": 0.6209778785705566,
+      "learning_rate": 7.038132369508779e-06,
+      "loss": 0.2388,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6082029141932002,
+      "grad_norm": 1.0005371570587158,
+      "learning_rate": 7.029784505112948e-06,
+      "loss": 0.2645,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6084727468969239,
+      "grad_norm": 1.892678141593933,
+      "learning_rate": 7.021438910224138e-06,
+      "loss": 0.2532,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6087425796006476,
+      "grad_norm": 1.1618189811706543,
+      "learning_rate": 7.013095591219128e-06,
+      "loss": 0.239,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6090124123043713,
+      "grad_norm": 1.6371448040008545,
+      "learning_rate": 7.00475455447295e-06,
+      "loss": 0.2679,
+      "step": 2257
+    },
+    {
+      "epoch": 0.609282245008095,
+      "grad_norm": 1.245186448097229,
+      "learning_rate": 6.996415806358888e-06,
+      "loss": 0.2835,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6095520777118186,
+      "grad_norm": 1.9404029846191406,
+      "learning_rate": 6.988079353248487e-06,
+      "loss": 0.2272,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6098219104155423,
+      "grad_norm": 2.848073959350586,
+      "learning_rate": 6.979745201511531e-06,
+      "loss": 0.2421,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6100917431192661,
+      "grad_norm": 0.9602468609809875,
+      "learning_rate": 6.971413357516052e-06,
+      "loss": 0.292,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6103615758229898,
+      "grad_norm": 1.5286660194396973,
+      "learning_rate": 6.963083827628312e-06,
+      "loss": 0.2871,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6106314085267135,
+      "grad_norm": 1.606199026107788,
+      "learning_rate": 6.954756618212803e-06,
+      "loss": 0.2476,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6109012412304371,
+      "grad_norm": 2.179142713546753,
+      "learning_rate": 6.94643173563226e-06,
+      "loss": 0.2581,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6111710739341608,
+      "grad_norm": 2.123885154724121,
+      "learning_rate": 6.938109186247619e-06,
+      "loss": 0.268,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6114409066378845,
+      "grad_norm": 0.9528677463531494,
+      "learning_rate": 6.929788976418044e-06,
+      "loss": 0.2674,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6117107393416082,
+      "grad_norm": 2.7092506885528564,
+      "learning_rate": 6.921471112500916e-06,
+      "loss": 0.2689,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6119805720453318,
+      "grad_norm": 2.963897228240967,
+      "learning_rate": 6.913155600851813e-06,
+      "loss": 0.2704,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6122504047490556,
+      "grad_norm": 1.3243041038513184,
+      "learning_rate": 6.904842447824518e-06,
+      "loss": 0.2707,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6125202374527793,
+      "grad_norm": 1.3842357397079468,
+      "learning_rate": 6.896531659771024e-06,
+      "loss": 0.2566,
+      "step": 2270
+    },
+    {
+      "epoch": 0.612790070156503,
+      "grad_norm": 1.5946691036224365,
+      "learning_rate": 6.888223243041499e-06,
+      "loss": 0.2594,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6130599028602267,
+      "grad_norm": 1.4231644868850708,
+      "learning_rate": 6.879917203984306e-06,
+      "loss": 0.2617,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6133297355639503,
+      "grad_norm": 1.6424612998962402,
+      "learning_rate": 6.871613548946001e-06,
+      "loss": 0.2568,
+      "step": 2273
+    },
+    {
+      "epoch": 0.613599568267674,
+      "grad_norm": 1.0543789863586426,
+      "learning_rate": 6.863312284271304e-06,
+      "loss": 0.2799,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6138694009713977,
+      "grad_norm": 1.0277252197265625,
+      "learning_rate": 6.855013416303113e-06,
+      "loss": 0.2585,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6141392336751215,
+      "grad_norm": 2.5486814975738525,
+      "learning_rate": 6.8467169513825025e-06,
+      "loss": 0.2731,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6144090663788451,
+      "grad_norm": 1.2918025255203247,
+      "learning_rate": 6.8384228958487e-06,
+      "loss": 0.3211,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6146788990825688,
+      "grad_norm": 0.9801622033119202,
+      "learning_rate": 6.830131256039094e-06,
+      "loss": 0.276,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6149487317862925,
+      "grad_norm": 0.9228232502937317,
+      "learning_rate": 6.82184203828924e-06,
+      "loss": 0.2515,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6152185644900162,
+      "grad_norm": 0.9755334258079529,
+      "learning_rate": 6.813555248932822e-06,
+      "loss": 0.2533,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6154883971937399,
+      "grad_norm": 2.355743646621704,
+      "learning_rate": 6.805270894301682e-06,
+      "loss": 0.2678,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6157582298974635,
+      "grad_norm": 2.686023235321045,
+      "learning_rate": 6.796988980725804e-06,
+      "loss": 0.3052,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6160280626011873,
+      "grad_norm": 0.9963394999504089,
+      "learning_rate": 6.788709514533295e-06,
+      "loss": 0.2721,
+      "step": 2283
+    },
+    {
+      "epoch": 0.616297895304911,
+      "grad_norm": 2.1087453365325928,
+      "learning_rate": 6.7804325020504e-06,
+      "loss": 0.2282,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6165677280086347,
+      "grad_norm": 1.1521786451339722,
+      "learning_rate": 6.7721579496014925e-06,
+      "loss": 0.2856,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6168375607123583,
+      "grad_norm": 1.185594081878662,
+      "learning_rate": 6.763885863509056e-06,
+      "loss": 0.2509,
+      "step": 2286
+    },
+    {
+      "epoch": 0.617107393416082,
+      "grad_norm": 1.367430329322815,
+      "learning_rate": 6.755616250093694e-06,
+      "loss": 0.2441,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6173772261198057,
+      "grad_norm": 1.0292530059814453,
+      "learning_rate": 6.747349115674131e-06,
+      "loss": 0.2838,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6176470588235294,
+      "grad_norm": 0.9090302586555481,
+      "learning_rate": 6.739084466567176e-06,
+      "loss": 0.2238,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6179168915272532,
+      "grad_norm": 0.9569499492645264,
+      "learning_rate": 6.730822309087756e-06,
+      "loss": 0.269,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6181867242309768,
+      "grad_norm": 1.1813760995864868,
+      "learning_rate": 6.722562649548898e-06,
+      "loss": 0.2328,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6184565569347005,
+      "grad_norm": 1.613837480545044,
+      "learning_rate": 6.714305494261695e-06,
+      "loss": 0.3059,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6187263896384242,
+      "grad_norm": 0.9070197343826294,
+      "learning_rate": 6.706050849535353e-06,
+      "loss": 0.2789,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6189962223421479,
+      "grad_norm": 2.252887010574341,
+      "learning_rate": 6.697798721677149e-06,
+      "loss": 0.2349,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6192660550458715,
+      "grad_norm": 0.872302234172821,
+      "learning_rate": 6.68954911699244e-06,
+      "loss": 0.2503,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6195358877495952,
+      "grad_norm": 1.492230772972107,
+      "learning_rate": 6.6813020417846456e-06,
+      "loss": 0.2633,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6198057204533189,
+      "grad_norm": 0.762330949306488,
+      "learning_rate": 6.673057502355265e-06,
+      "loss": 0.2345,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6200755531570427,
+      "grad_norm": 3.457165479660034,
+      "learning_rate": 6.664815505003856e-06,
+      "loss": 0.3276,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6203453858607664,
+      "grad_norm": 1.3414520025253296,
+      "learning_rate": 6.656576056028028e-06,
+      "loss": 0.2605,
+      "step": 2299
+    },
+    {
+      "epoch": 0.62061521856449,
+      "grad_norm": 0.8384405374526978,
+      "learning_rate": 6.648339161723447e-06,
+      "loss": 0.2785,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6208850512682137,
+      "grad_norm": 2.1462364196777344,
+      "learning_rate": 6.6401048283838375e-06,
+      "loss": 0.2646,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6211548839719374,
+      "grad_norm": 1.3320510387420654,
+      "learning_rate": 6.6318730623009465e-06,
+      "loss": 0.2787,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6214247166756611,
+      "grad_norm": 1.2961645126342773,
+      "learning_rate": 6.623643869764575e-06,
+      "loss": 0.2838,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6216945493793847,
+      "grad_norm": 0.7948213815689087,
+      "learning_rate": 6.615417257062556e-06,
+      "loss": 0.2605,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6219643820831084,
+      "grad_norm": 0.8254593014717102,
+      "learning_rate": 6.607193230480744e-06,
+      "loss": 0.2994,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6222342147868322,
+      "grad_norm": 0.8213491439819336,
+      "learning_rate": 6.5989717963030206e-06,
+      "loss": 0.255,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6225040474905559,
+      "grad_norm": 1.1966582536697388,
+      "learning_rate": 6.590752960811295e-06,
+      "loss": 0.267,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6227738801942796,
+      "grad_norm": 0.820884644985199,
+      "learning_rate": 6.582536730285476e-06,
+      "loss": 0.2564,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6230437128980032,
+      "grad_norm": 1.020833134651184,
+      "learning_rate": 6.574323111003491e-06,
+      "loss": 0.2904,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6233135456017269,
+      "grad_norm": 1.7857580184936523,
+      "learning_rate": 6.5661121092412785e-06,
+      "loss": 0.2635,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6235833783054506,
+      "grad_norm": 0.8182295560836792,
+      "learning_rate": 6.557903731272758e-06,
+      "loss": 0.2847,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.8526229858398438,
+      "learning_rate": 6.54969798336986e-06,
+      "loss": 0.2913,
+      "step": 2312
+    },
+    {
+      "epoch": 0.624123043712898,
+      "grad_norm": 1.1285094022750854,
+      "learning_rate": 6.541494871802504e-06,
+      "loss": 0.2968,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6243928764166217,
+      "grad_norm": 1.411910891532898,
+      "learning_rate": 6.5332944028385885e-06,
+      "loss": 0.3047,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6246627091203454,
+      "grad_norm": 0.7462607622146606,
+      "learning_rate": 6.525096582743992e-06,
+      "loss": 0.2654,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6249325418240691,
+      "grad_norm": 0.8665904998779297,
+      "learning_rate": 6.516901417782584e-06,
+      "loss": 0.2333,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6252023745277928,
+      "grad_norm": 1.178702712059021,
+      "learning_rate": 6.5087089142161865e-06,
+      "loss": 0.2796,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6254722072315164,
+      "grad_norm": 1.134119987487793,
+      "learning_rate": 6.500519078304597e-06,
+      "loss": 0.2964,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6257420399352401,
+      "grad_norm": 0.9187453389167786,
+      "learning_rate": 6.492331916305576e-06,
+      "loss": 0.242,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6260118726389639,
+      "grad_norm": 1.391666293144226,
+      "learning_rate": 6.484147434474837e-06,
+      "loss": 0.2373,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6262817053426876,
+      "grad_norm": 2.018650531768799,
+      "learning_rate": 6.475965639066044e-06,
+      "loss": 0.2693,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6265515380464112,
+      "grad_norm": 0.8021873831748962,
+      "learning_rate": 6.46778653633082e-06,
+      "loss": 0.2796,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6268213707501349,
+      "grad_norm": 1.2555994987487793,
+      "learning_rate": 6.459610132518714e-06,
+      "loss": 0.2526,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6270912034538586,
+      "grad_norm": 1.840186595916748,
+      "learning_rate": 6.4514364338772205e-06,
+      "loss": 0.2948,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6273610361575823,
+      "grad_norm": 2.638143539428711,
+      "learning_rate": 6.443265446651774e-06,
+      "loss": 0.2858,
+      "step": 2325
+    },
+    {
+      "epoch": 0.627630868861306,
+      "grad_norm": 1.1777907609939575,
+      "learning_rate": 6.435097177085728e-06,
+      "loss": 0.2791,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6279007015650296,
+      "grad_norm": 0.9402373433113098,
+      "learning_rate": 6.426931631420358e-06,
+      "loss": 0.2444,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6281705342687534,
+      "grad_norm": 1.4587208032608032,
+      "learning_rate": 6.418768815894867e-06,
+      "loss": 0.2531,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6284403669724771,
+      "grad_norm": 1.2955989837646484,
+      "learning_rate": 6.410608736746367e-06,
+      "loss": 0.234,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6287101996762008,
+      "grad_norm": 0.8608735203742981,
+      "learning_rate": 6.402451400209873e-06,
+      "loss": 0.2659,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6289800323799244,
+      "grad_norm": 0.7598680257797241,
+      "learning_rate": 6.394296812518317e-06,
+      "loss": 0.2486,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6292498650836481,
+      "grad_norm": 1.7671480178833008,
+      "learning_rate": 6.386144979902527e-06,
+      "loss": 0.2877,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6295196977873718,
+      "grad_norm": 0.7645602226257324,
+      "learning_rate": 6.377995908591213e-06,
+      "loss": 0.2329,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6297895304910955,
+      "grad_norm": 1.4483979940414429,
+      "learning_rate": 6.369849604810994e-06,
+      "loss": 0.2656,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6300593631948193,
+      "grad_norm": 1.3618022203445435,
+      "learning_rate": 6.3617060747863644e-06,
+      "loss": 0.2508,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6303291958985429,
+      "grad_norm": 0.9737834334373474,
+      "learning_rate": 6.3535653247396965e-06,
+      "loss": 0.2719,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6305990286022666,
+      "grad_norm": 1.1365381479263306,
+      "learning_rate": 6.345427360891245e-06,
+      "loss": 0.2838,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6308688613059903,
+      "grad_norm": 1.5925894975662231,
+      "learning_rate": 6.337292189459139e-06,
+      "loss": 0.2538,
+      "step": 2338
+    },
+    {
+      "epoch": 0.631138694009714,
+      "grad_norm": 1.6314568519592285,
+      "learning_rate": 6.329159816659356e-06,
+      "loss": 0.2652,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6314085267134377,
+      "grad_norm": 0.9832297563552856,
+      "learning_rate": 6.321030248705756e-06,
+      "loss": 0.2588,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6316783594171613,
+      "grad_norm": 1.0588876008987427,
+      "learning_rate": 6.3129034918100475e-06,
+      "loss": 0.297,
+      "step": 2341
+    },
+    {
+      "epoch": 0.631948192120885,
+      "grad_norm": 0.6630373597145081,
+      "learning_rate": 6.304779552181783e-06,
+      "loss": 0.2472,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6322180248246088,
+      "grad_norm": 1.2848323583602905,
+      "learning_rate": 6.296658436028378e-06,
+      "loss": 0.2456,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6324878575283325,
+      "grad_norm": 1.1359772682189941,
+      "learning_rate": 6.2885401495550826e-06,
+      "loss": 0.265,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6327576902320561,
+      "grad_norm": 1.1673206090927124,
+      "learning_rate": 6.280424698964977e-06,
+      "loss": 0.2617,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6330275229357798,
+      "grad_norm": 1.3849053382873535,
+      "learning_rate": 6.272312090458988e-06,
+      "loss": 0.2473,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6332973556395035,
+      "grad_norm": 0.7756108045578003,
+      "learning_rate": 6.264202330235869e-06,
+      "loss": 0.2686,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6335671883432272,
+      "grad_norm": 1.7736762762069702,
+      "learning_rate": 6.256095424492183e-06,
+      "loss": 0.2383,
+      "step": 2348
+    },
+    {
+      "epoch": 0.6338370210469509,
+      "grad_norm": 1.1570478677749634,
+      "learning_rate": 6.24799137942233e-06,
+      "loss": 0.2763,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6341068537506745,
+      "grad_norm": 0.9947857856750488,
+      "learning_rate": 6.239890201218517e-06,
+      "loss": 0.2762,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6343766864543983,
+      "grad_norm": 1.6437578201293945,
+      "learning_rate": 6.231791896070752e-06,
+      "loss": 0.2444,
+      "step": 2351
+    },
+    {
+      "epoch": 0.634646519158122,
+      "grad_norm": 1.3034433126449585,
+      "learning_rate": 6.223696470166863e-06,
+      "loss": 0.233,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6349163518618457,
+      "grad_norm": 1.1192454099655151,
+      "learning_rate": 6.215603929692471e-06,
+      "loss": 0.2935,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6351861845655693,
+      "grad_norm": 1.4182828664779663,
+      "learning_rate": 6.207514280830984e-06,
+      "loss": 0.2816,
+      "step": 2354
+    },
+    {
+      "epoch": 0.635456017269293,
+      "grad_norm": 1.2007042169570923,
+      "learning_rate": 6.199427529763615e-06,
+      "loss": 0.2705,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6357258499730167,
+      "grad_norm": 1.416608452796936,
+      "learning_rate": 6.191343682669357e-06,
+      "loss": 0.2892,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6359956826767404,
+      "grad_norm": 0.9986120462417603,
+      "learning_rate": 6.183262745724976e-06,
+      "loss": 0.2522,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6362655153804642,
+      "grad_norm": 1.1393169164657593,
+      "learning_rate": 6.17518472510503e-06,
+      "loss": 0.2151,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6365353480841878,
+      "grad_norm": 1.0892841815948486,
+      "learning_rate": 6.167109626981833e-06,
+      "loss": 0.3089,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6368051807879115,
+      "grad_norm": 1.1441901922225952,
+      "learning_rate": 6.159037457525483e-06,
+      "loss": 0.2755,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6370750134916352,
+      "grad_norm": 1.6062510013580322,
+      "learning_rate": 6.150968222903822e-06,
+      "loss": 0.2984,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6373448461953589,
+      "grad_norm": 0.8166984915733337,
+      "learning_rate": 6.142901929282459e-06,
+      "loss": 0.2422,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6376146788990825,
+      "grad_norm": 1.1426907777786255,
+      "learning_rate": 6.13483858282476e-06,
+      "loss": 0.2576,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6378845116028062,
+      "grad_norm": 0.9255439639091492,
+      "learning_rate": 6.1267781896918285e-06,
+      "loss": 0.2547,
+      "step": 2364
+    },
+    {
+      "epoch": 0.63815434430653,
+      "grad_norm": 0.7929574847221375,
+      "learning_rate": 6.1187207560425175e-06,
+      "loss": 0.2522,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6384241770102537,
+      "grad_norm": 1.123069167137146,
+      "learning_rate": 6.110666288033424e-06,
+      "loss": 0.2763,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6386940097139774,
+      "grad_norm": 1.05642569065094,
+      "learning_rate": 6.102614791818866e-06,
+      "loss": 0.2345,
+      "step": 2367
+    },
+    {
+      "epoch": 0.638963842417701,
+      "grad_norm": 1.4229053258895874,
+      "learning_rate": 6.094566273550899e-06,
+      "loss": 0.2442,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6392336751214247,
+      "grad_norm": 0.8854284286499023,
+      "learning_rate": 6.086520739379309e-06,
+      "loss": 0.2329,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6395035078251484,
+      "grad_norm": 1.4309442043304443,
+      "learning_rate": 6.0784781954515856e-06,
+      "loss": 0.2489,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6397733405288721,
+      "grad_norm": 1.151275396347046,
+      "learning_rate": 6.070438647912945e-06,
+      "loss": 0.2446,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6400431732325957,
+      "grad_norm": 1.0651471614837646,
+      "learning_rate": 6.0624021029063184e-06,
+      "loss": 0.2778,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6403130059363195,
+      "grad_norm": 0.9093769192695618,
+      "learning_rate": 6.054368566572328e-06,
+      "loss": 0.2304,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6405828386400432,
+      "grad_norm": 1.3571912050247192,
+      "learning_rate": 6.046338045049307e-06,
+      "loss": 0.2835,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6408526713437669,
+      "grad_norm": 0.8759908676147461,
+      "learning_rate": 6.038310544473289e-06,
+      "loss": 0.2683,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6411225040474906,
+      "grad_norm": 3.288300037384033,
+      "learning_rate": 6.030286070977986e-06,
+      "loss": 0.2268,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6413923367512142,
+      "grad_norm": 1.690068244934082,
+      "learning_rate": 6.022264630694807e-06,
+      "loss": 0.2182,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6416621694549379,
+      "grad_norm": 0.8688433766365051,
+      "learning_rate": 6.014246229752847e-06,
+      "loss": 0.2467,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6419320021586616,
+      "grad_norm": 1.0472760200500488,
+      "learning_rate": 6.006230874278863e-06,
+      "loss": 0.2768,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 1.7069085836410522,
+      "learning_rate": 5.998218570397298e-06,
+      "loss": 0.2636,
+      "step": 2380
+    },
+    {
+      "epoch": 0.642471667566109,
+      "grad_norm": 1.9888927936553955,
+      "learning_rate": 5.990209324230266e-06,
+      "loss": 0.2915,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6427415002698327,
+      "grad_norm": 0.8581491112709045,
+      "learning_rate": 5.982203141897531e-06,
+      "loss": 0.2542,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6430113329735564,
+      "grad_norm": 1.1668142080307007,
+      "learning_rate": 5.974200029516526e-06,
+      "loss": 0.2621,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6432811656772801,
+      "grad_norm": 1.3992540836334229,
+      "learning_rate": 5.9661999932023376e-06,
+      "loss": 0.2701,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6435509983810038,
+      "grad_norm": 2.4744982719421387,
+      "learning_rate": 5.958203039067698e-06,
+      "loss": 0.2868,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6438208310847274,
+      "grad_norm": 2.3034074306488037,
+      "learning_rate": 5.950209173222985e-06,
+      "loss": 0.262,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6440906637884511,
+      "grad_norm": 1.8448705673217773,
+      "learning_rate": 5.942218401776225e-06,
+      "loss": 0.2547,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6443604964921749,
+      "grad_norm": 1.4008986949920654,
+      "learning_rate": 5.934230730833068e-06,
+      "loss": 0.2729,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6446303291958986,
+      "grad_norm": 1.5237371921539307,
+      "learning_rate": 5.9262461664967985e-06,
+      "loss": 0.2478,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6449001618996222,
+      "grad_norm": 0.9780192375183105,
+      "learning_rate": 5.918264714868336e-06,
+      "loss": 0.2516,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6451699946033459,
+      "grad_norm": 0.7068191170692444,
+      "learning_rate": 5.910286382046207e-06,
+      "loss": 0.2316,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6454398273070696,
+      "grad_norm": 3.07895827293396,
+      "learning_rate": 5.902311174126565e-06,
+      "loss": 0.2766,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6457096600107933,
+      "grad_norm": 0.8056784272193909,
+      "learning_rate": 5.894339097203176e-06,
+      "loss": 0.233,
+      "step": 2393
+    },
+    {
+      "epoch": 0.645979492714517,
+      "grad_norm": 0.8796300888061523,
+      "learning_rate": 5.8863701573674125e-06,
+      "loss": 0.2713,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6462493254182406,
+      "grad_norm": 0.9664385914802551,
+      "learning_rate": 5.878404360708241e-06,
+      "loss": 0.2423,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6465191581219644,
+      "grad_norm": 2.098768711090088,
+      "learning_rate": 5.870441713312238e-06,
+      "loss": 0.2474,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6467889908256881,
+      "grad_norm": 0.9904431104660034,
+      "learning_rate": 5.862482221263575e-06,
+      "loss": 0.2668,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 1.4273697137832642,
+      "learning_rate": 5.854525890643996e-06,
+      "loss": 0.267,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6473286562331354,
+      "grad_norm": 1.6860491037368774,
+      "learning_rate": 5.846572727532849e-06,
+      "loss": 0.2861,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6475984889368591,
+      "grad_norm": 0.9392491579055786,
+      "learning_rate": 5.838622738007051e-06,
+      "loss": 0.2686,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6478683216405828,
+      "grad_norm": 0.8313483595848083,
+      "learning_rate": 5.830675928141092e-06,
+      "loss": 0.2532,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6481381543443065,
+      "grad_norm": 1.0184792280197144,
+      "learning_rate": 5.8227323040070435e-06,
+      "loss": 0.2584,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6484079870480303,
+      "grad_norm": 1.0025299787521362,
+      "learning_rate": 5.814791871674532e-06,
+      "loss": 0.246,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6486778197517539,
+      "grad_norm": 1.7561419010162354,
+      "learning_rate": 5.806854637210752e-06,
+      "loss": 0.2699,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6489476524554776,
+      "grad_norm": 1.1539487838745117,
+      "learning_rate": 5.79892060668045e-06,
+      "loss": 0.2542,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6492174851592013,
+      "grad_norm": 1.1399242877960205,
+      "learning_rate": 5.790989786145924e-06,
+      "loss": 0.2412,
+      "step": 2406
+    },
+    {
+      "epoch": 0.649487317862925,
+      "grad_norm": 0.8627252578735352,
+      "learning_rate": 5.783062181667025e-06,
+      "loss": 0.2858,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6497571505666486,
+      "grad_norm": 1.3148490190505981,
+      "learning_rate": 5.775137799301143e-06,
+      "loss": 0.2878,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6500269832703723,
+      "grad_norm": 1.4426440000534058,
+      "learning_rate": 5.767216645103204e-06,
+      "loss": 0.2846,
+      "step": 2409
+    },
+    {
+      "epoch": 0.650296815974096,
+      "grad_norm": 1.121668815612793,
+      "learning_rate": 5.759298725125671e-06,
+      "loss": 0.2368,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6505666486778198,
+      "grad_norm": 1.6011898517608643,
+      "learning_rate": 5.751384045418534e-06,
+      "loss": 0.2841,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6508364813815435,
+      "grad_norm": 1.8797924518585205,
+      "learning_rate": 5.7434726120293085e-06,
+      "loss": 0.2518,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6511063140852671,
+      "grad_norm": 0.9497119188308716,
+      "learning_rate": 5.7355644310030265e-06,
+      "loss": 0.2498,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6513761467889908,
+      "grad_norm": 0.8765178322792053,
+      "learning_rate": 5.7276595083822375e-06,
+      "loss": 0.2841,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6516459794927145,
+      "grad_norm": 1.7587532997131348,
+      "learning_rate": 5.719757850207001e-06,
+      "loss": 0.2836,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6519158121964382,
+      "grad_norm": 1.3646577596664429,
+      "learning_rate": 5.711859462514883e-06,
+      "loss": 0.2665,
+      "step": 2416
+    },
+    {
+      "epoch": 0.652185644900162,
+      "grad_norm": 1.0520551204681396,
+      "learning_rate": 5.703964351340949e-06,
+      "loss": 0.2642,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6524554776038856,
+      "grad_norm": 1.693588376045227,
+      "learning_rate": 5.6960725227177635e-06,
+      "loss": 0.2968,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6527253103076093,
+      "grad_norm": 1.4760632514953613,
+      "learning_rate": 5.6881839826753704e-06,
+      "loss": 0.2505,
+      "step": 2419
+    },
+    {
+      "epoch": 0.652995143011333,
+      "grad_norm": 0.7768751978874207,
+      "learning_rate": 5.6802987372413235e-06,
+      "loss": 0.3045,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6532649757150567,
+      "grad_norm": 0.9819402098655701,
+      "learning_rate": 5.672416792440645e-06,
+      "loss": 0.2846,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6535348084187803,
+      "grad_norm": 0.9052737951278687,
+      "learning_rate": 5.664538154295827e-06,
+      "loss": 0.2764,
+      "step": 2422
+    },
+    {
+      "epoch": 0.653804641122504,
+      "grad_norm": 1.5336061716079712,
+      "learning_rate": 5.6566628288268564e-06,
+      "loss": 0.2593,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6540744738262277,
+      "grad_norm": 1.366075873374939,
+      "learning_rate": 5.648790822051178e-06,
+      "loss": 0.2302,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6543443065299515,
+      "grad_norm": 0.9524087309837341,
+      "learning_rate": 5.640922139983687e-06,
+      "loss": 0.271,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6546141392336752,
+      "grad_norm": 0.8853236436843872,
+      "learning_rate": 5.633056788636764e-06,
+      "loss": 0.2809,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6548839719373988,
+      "grad_norm": 0.809134840965271,
+      "learning_rate": 5.625194774020229e-06,
+      "loss": 0.2584,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6551538046411225,
+      "grad_norm": 0.8997631072998047,
+      "learning_rate": 5.617336102141356e-06,
+      "loss": 0.2685,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6554236373448462,
+      "grad_norm": 1.702239990234375,
+      "learning_rate": 5.609480779004862e-06,
+      "loss": 0.2721,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6556934700485699,
+      "grad_norm": 1.2478662729263306,
+      "learning_rate": 5.601628810612911e-06,
+      "loss": 0.267,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6559633027522935,
+      "grad_norm": 0.764225959777832,
+      "learning_rate": 5.5937802029651e-06,
+      "loss": 0.2547,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6562331354560172,
+      "grad_norm": 0.870741069316864,
+      "learning_rate": 5.585934962058457e-06,
+      "loss": 0.2704,
+      "step": 2432
+    },
+    {
+      "epoch": 0.656502968159741,
+      "grad_norm": 1.0013765096664429,
+      "learning_rate": 5.57809309388744e-06,
+      "loss": 0.2678,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6567728008634647,
+      "grad_norm": 0.7918844223022461,
+      "learning_rate": 5.570254604443929e-06,
+      "loss": 0.2599,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6570426335671884,
+      "grad_norm": 0.8641981482505798,
+      "learning_rate": 5.562419499717222e-06,
+      "loss": 0.2532,
+      "step": 2435
+    },
+    {
+      "epoch": 0.657312466270912,
+      "grad_norm": 1.2443643808364868,
+      "learning_rate": 5.554587785694032e-06,
+      "loss": 0.2689,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6575822989746357,
+      "grad_norm": 0.8148622512817383,
+      "learning_rate": 5.5467594683584795e-06,
+      "loss": 0.2291,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6578521316783594,
+      "grad_norm": 1.1282086372375488,
+      "learning_rate": 5.53893455369209e-06,
+      "loss": 0.275,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6581219643820831,
+      "grad_norm": 1.1173254251480103,
+      "learning_rate": 5.531113047673792e-06,
+      "loss": 0.2343,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6583917970858068,
+      "grad_norm": 1.279711365699768,
+      "learning_rate": 5.5232949562799055e-06,
+      "loss": 0.2732,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6586616297895305,
+      "grad_norm": 0.9675107598304749,
+      "learning_rate": 5.515480285484141e-06,
+      "loss": 0.2831,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6589314624932542,
+      "grad_norm": 1.5094584226608276,
+      "learning_rate": 5.507669041257601e-06,
+      "loss": 0.273,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6592012951969779,
+      "grad_norm": 1.029333472251892,
+      "learning_rate": 5.4998612295687635e-06,
+      "loss": 0.2915,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6594711279007016,
+      "grad_norm": 1.0688508749008179,
+      "learning_rate": 5.492056856383488e-06,
+      "loss": 0.2719,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6597409606044252,
+      "grad_norm": 1.4233678579330444,
+      "learning_rate": 5.484255927665003e-06,
+      "loss": 0.2511,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6600107933081489,
+      "grad_norm": 0.8227878212928772,
+      "learning_rate": 5.4764584493739095e-06,
+      "loss": 0.2769,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6602806260118727,
+      "grad_norm": 2.7820680141448975,
+      "learning_rate": 5.468664427468169e-06,
+      "loss": 0.2385,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 1.0660181045532227,
+      "learning_rate": 5.460873867903103e-06,
+      "loss": 0.2534,
+      "step": 2448
+    },
+    {
+      "epoch": 0.66082029141932,
+      "grad_norm": 0.9603192806243896,
+      "learning_rate": 5.453086776631384e-06,
+      "loss": 0.2541,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6610901241230437,
+      "grad_norm": 1.2765765190124512,
+      "learning_rate": 5.445303159603043e-06,
+      "loss": 0.3037,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6613599568267674,
+      "grad_norm": 1.129353404045105,
+      "learning_rate": 5.4375230227654475e-06,
+      "loss": 0.2538,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6616297895304911,
+      "grad_norm": 1.2160744667053223,
+      "learning_rate": 5.429746372063309e-06,
+      "loss": 0.2451,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6618996222342148,
+      "grad_norm": 1.2882336378097534,
+      "learning_rate": 5.421973213438678e-06,
+      "loss": 0.2581,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6621694549379384,
+      "grad_norm": 1.101177453994751,
+      "learning_rate": 5.414203552830934e-06,
+      "loss": 0.2458,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6624392876416622,
+      "grad_norm": 1.3069547414779663,
+      "learning_rate": 5.406437396176783e-06,
+      "loss": 0.2774,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6627091203453859,
+      "grad_norm": 1.2091983556747437,
+      "learning_rate": 5.398674749410253e-06,
+      "loss": 0.279,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6629789530491096,
+      "grad_norm": 2.9212164878845215,
+      "learning_rate": 5.390915618462695e-06,
+      "loss": 0.3295,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6632487857528332,
+      "grad_norm": 0.7125205397605896,
+      "learning_rate": 5.3831600092627704e-06,
+      "loss": 0.2486,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6635186184565569,
+      "grad_norm": 2.45274019241333,
+      "learning_rate": 5.375407927736447e-06,
+      "loss": 0.25,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6637884511602806,
+      "grad_norm": 2.066992998123169,
+      "learning_rate": 5.367659379807002e-06,
+      "loss": 0.277,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6640582838640043,
+      "grad_norm": 0.9218307733535767,
+      "learning_rate": 5.359914371395012e-06,
+      "loss": 0.2486,
+      "step": 2461
+    },
+    {
+      "epoch": 0.664328116567728,
+      "grad_norm": 1.5415490865707397,
+      "learning_rate": 5.352172908418343e-06,
+      "loss": 0.298,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6645979492714517,
+      "grad_norm": 1.0419403314590454,
+      "learning_rate": 5.344434996792162e-06,
+      "loss": 0.2781,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6648677819751754,
+      "grad_norm": 1.2006274461746216,
+      "learning_rate": 5.336700642428913e-06,
+      "loss": 0.2643,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6651376146788991,
+      "grad_norm": 0.7162570357322693,
+      "learning_rate": 5.328969851238327e-06,
+      "loss": 0.2742,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6654074473826228,
+      "grad_norm": 1.5005978345870972,
+      "learning_rate": 5.321242629127412e-06,
+      "loss": 0.2931,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6656772800863464,
+      "grad_norm": 1.0908997058868408,
+      "learning_rate": 5.3135189820004486e-06,
+      "loss": 0.2629,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6659471127900701,
+      "grad_norm": 0.9958769083023071,
+      "learning_rate": 5.305798915758984e-06,
+      "loss": 0.2875,
+      "step": 2468
+    },
+    {
+      "epoch": 0.6662169454937938,
+      "grad_norm": 1.667133092880249,
+      "learning_rate": 5.298082436301832e-06,
+      "loss": 0.2373,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6664867781975176,
+      "grad_norm": 1.0325188636779785,
+      "learning_rate": 5.290369549525066e-06,
+      "loss": 0.2438,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6667566109012413,
+      "grad_norm": 0.8477824330329895,
+      "learning_rate": 5.28266026132201e-06,
+      "loss": 0.2907,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6670264436049649,
+      "grad_norm": 1.0365746021270752,
+      "learning_rate": 5.2749545775832425e-06,
+      "loss": 0.239,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6672962763086886,
+      "grad_norm": 0.9639760255813599,
+      "learning_rate": 5.267252504196587e-06,
+      "loss": 0.2904,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6675661090124123,
+      "grad_norm": 1.041070580482483,
+      "learning_rate": 5.259554047047106e-06,
+      "loss": 0.2673,
+      "step": 2474
+    },
+    {
+      "epoch": 0.667835941716136,
+      "grad_norm": 1.130446434020996,
+      "learning_rate": 5.2518592120171045e-06,
+      "loss": 0.2565,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6681057744198596,
+      "grad_norm": 2.1641616821289062,
+      "learning_rate": 5.2441680049861125e-06,
+      "loss": 0.3295,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6683756071235833,
+      "grad_norm": 1.2218416929244995,
+      "learning_rate": 5.236480431830895e-06,
+      "loss": 0.2647,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6686454398273071,
+      "grad_norm": 1.997772216796875,
+      "learning_rate": 5.228796498425432e-06,
+      "loss": 0.263,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6689152725310308,
+      "grad_norm": 1.2077445983886719,
+      "learning_rate": 5.221116210640933e-06,
+      "loss": 0.2385,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6691851052347545,
+      "grad_norm": 0.9158920645713806,
+      "learning_rate": 5.213439574345814e-06,
+      "loss": 0.279,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6694549379384781,
+      "grad_norm": 0.6433935165405273,
+      "learning_rate": 5.205766595405703e-06,
+      "loss": 0.2375,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6697247706422018,
+      "grad_norm": 0.8731844425201416,
+      "learning_rate": 5.198097279683434e-06,
+      "loss": 0.3143,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6699946033459255,
+      "grad_norm": 1.3500096797943115,
+      "learning_rate": 5.190431633039042e-06,
+      "loss": 0.2956,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6702644360496492,
+      "grad_norm": 0.7880909442901611,
+      "learning_rate": 5.182769661329756e-06,
+      "loss": 0.2623,
+      "step": 2484
+    },
+    {
+      "epoch": 0.670534268753373,
+      "grad_norm": 0.9401893019676208,
+      "learning_rate": 5.175111370410002e-06,
+      "loss": 0.2398,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6708041014570966,
+      "grad_norm": 0.8522136807441711,
+      "learning_rate": 5.167456766131391e-06,
+      "loss": 0.2536,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6710739341608203,
+      "grad_norm": 1.1613191366195679,
+      "learning_rate": 5.159805854342714e-06,
+      "loss": 0.2483,
+      "step": 2487
+    },
+    {
+      "epoch": 0.671343766864544,
+      "grad_norm": 1.2855637073516846,
+      "learning_rate": 5.152158640889947e-06,
+      "loss": 0.2768,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6716135995682677,
+      "grad_norm": 1.2939449548721313,
+      "learning_rate": 5.144515131616235e-06,
+      "loss": 0.2348,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6718834322719913,
+      "grad_norm": 0.8970277905464172,
+      "learning_rate": 5.136875332361893e-06,
+      "loss": 0.2521,
+      "step": 2490
+    },
+    {
+      "epoch": 0.672153264975715,
+      "grad_norm": 1.52006995677948,
+      "learning_rate": 5.129239248964404e-06,
+      "loss": 0.2877,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6724230976794388,
+      "grad_norm": 0.818020761013031,
+      "learning_rate": 5.121606887258407e-06,
+      "loss": 0.2545,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6726929303831625,
+      "grad_norm": 0.9843150973320007,
+      "learning_rate": 5.113978253075711e-06,
+      "loss": 0.2416,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6729627630868862,
+      "grad_norm": 2.3811402320861816,
+      "learning_rate": 5.106353352245254e-06,
+      "loss": 0.304,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6732325957906098,
+      "grad_norm": 2.4794695377349854,
+      "learning_rate": 5.098732190593137e-06,
+      "loss": 0.2584,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6735024284943335,
+      "grad_norm": 1.0273422002792358,
+      "learning_rate": 5.091114773942609e-06,
+      "loss": 0.2779,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6737722611980572,
+      "grad_norm": 2.211319923400879,
+      "learning_rate": 5.083501108114039e-06,
+      "loss": 0.2658,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6740420939017809,
+      "grad_norm": 1.1483089923858643,
+      "learning_rate": 5.075891198924942e-06,
+      "loss": 0.2769,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6743119266055045,
+      "grad_norm": 0.9274725914001465,
+      "learning_rate": 5.068285052189968e-06,
+      "loss": 0.2699,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6745817593092283,
+      "grad_norm": 0.8508108258247375,
+      "learning_rate": 5.060682673720878e-06,
+      "loss": 0.2596,
+      "step": 2500
+    },
+    {
+      "epoch": 0.674851592012952,
+      "grad_norm": 1.4897525310516357,
+      "learning_rate": 5.05308406932656e-06,
+      "loss": 0.2572,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6751214247166757,
+      "grad_norm": 0.9482914209365845,
+      "learning_rate": 5.045489244813027e-06,
+      "loss": 0.2395,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6753912574203994,
+      "grad_norm": 0.7715896964073181,
+      "learning_rate": 5.037898205983388e-06,
+      "loss": 0.2714,
+      "step": 2503
+    },
+    {
+      "epoch": 0.675661090124123,
+      "grad_norm": 0.7780461311340332,
+      "learning_rate": 5.030310958637865e-06,
+      "loss": 0.2603,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6759309228278467,
+      "grad_norm": 0.7754076719284058,
+      "learning_rate": 5.022727508573799e-06,
+      "loss": 0.2581,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6762007555315704,
+      "grad_norm": 0.9156253933906555,
+      "learning_rate": 5.015147861585603e-06,
+      "loss": 0.2467,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6764705882352942,
+      "grad_norm": 1.5604363679885864,
+      "learning_rate": 5.0075720234647974e-06,
+      "loss": 0.2581,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6767404209390178,
+      "grad_norm": 1.7143417596817017,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.3174,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6770102536427415,
+      "grad_norm": 1.7577524185180664,
+      "learning_rate": 4.992431796976902e-06,
+      "loss": 0.28,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6772800863464652,
+      "grad_norm": 1.554797649383545,
+      "learning_rate": 4.984867420178271e-06,
+      "loss": 0.2617,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6775499190501889,
+      "grad_norm": 1.6056751012802124,
+      "learning_rate": 4.977306875383974e-06,
+      "loss": 0.2624,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6778197517539126,
+      "grad_norm": 0.9899097681045532,
+      "learning_rate": 4.969750168370924e-06,
+      "loss": 0.2596,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6780895844576362,
+      "grad_norm": 0.9200010299682617,
+      "learning_rate": 4.96219730491311e-06,
+      "loss": 0.2569,
+      "step": 2513
+    },
+    {
+      "epoch": 0.67835941716136,
+      "grad_norm": 1.8135788440704346,
+      "learning_rate": 4.954648290781602e-06,
+      "loss": 0.2557,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6786292498650837,
+      "grad_norm": 0.8333318829536438,
+      "learning_rate": 4.9471031317445005e-06,
+      "loss": 0.2121,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 1.1304751634597778,
+      "learning_rate": 4.9395618335669726e-06,
+      "loss": 0.2833,
+      "step": 2516
+    },
+    {
+      "epoch": 0.679168915272531,
+      "grad_norm": 1.1979104280471802,
+      "learning_rate": 4.932024402011249e-06,
+      "loss": 0.2411,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6794387479762547,
+      "grad_norm": 1.2152214050292969,
+      "learning_rate": 4.924490842836584e-06,
+      "loss": 0.2653,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6797085806799784,
+      "grad_norm": 0.696350634098053,
+      "learning_rate": 4.916961161799278e-06,
+      "loss": 0.195,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6799784133837021,
+      "grad_norm": 1.9317229986190796,
+      "learning_rate": 4.9094353646526845e-06,
+      "loss": 0.2584,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6802482460874258,
+      "grad_norm": 1.6010396480560303,
+      "learning_rate": 4.901913457147169e-06,
+      "loss": 0.2236,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6805180787911494,
+      "grad_norm": 0.9393137097358704,
+      "learning_rate": 4.894395445030131e-06,
+      "loss": 0.2558,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6807879114948732,
+      "grad_norm": 1.2070720195770264,
+      "learning_rate": 4.886881334046008e-06,
+      "loss": 0.267,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6810577441985969,
+      "grad_norm": 1.781626582145691,
+      "learning_rate": 4.879371129936233e-06,
+      "loss": 0.2347,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6813275769023206,
+      "grad_norm": 1.0960925817489624,
+      "learning_rate": 4.871864838439267e-06,
+      "loss": 0.2698,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6815974096060442,
+      "grad_norm": 0.8313362002372742,
+      "learning_rate": 4.864362465290583e-06,
+      "loss": 0.2365,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6818672423097679,
+      "grad_norm": 1.427254557609558,
+      "learning_rate": 4.856864016222661e-06,
+      "loss": 0.2298,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6821370750134916,
+      "grad_norm": 1.2723969221115112,
+      "learning_rate": 4.849369496964966e-06,
+      "loss": 0.2482,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6824069077172153,
+      "grad_norm": 1.3138600587844849,
+      "learning_rate": 4.841878913243984e-06,
+      "loss": 0.27,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6826767404209391,
+      "grad_norm": 1.2733210325241089,
+      "learning_rate": 4.834392270783183e-06,
+      "loss": 0.2641,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6829465731246627,
+      "grad_norm": 1.075350046157837,
+      "learning_rate": 4.826909575303006e-06,
+      "loss": 0.2764,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6832164058283864,
+      "grad_norm": 1.4895039796829224,
+      "learning_rate": 4.8194308325209075e-06,
+      "loss": 0.2734,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6834862385321101,
+      "grad_norm": 2.034773588180542,
+      "learning_rate": 4.811956048151306e-06,
+      "loss": 0.2551,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6837560712358338,
+      "grad_norm": 2.048279285430908,
+      "learning_rate": 4.804485227905587e-06,
+      "loss": 0.2286,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6840259039395574,
+      "grad_norm": 1.2363805770874023,
+      "learning_rate": 4.797018377492125e-06,
+      "loss": 0.2709,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6842957366432811,
+      "grad_norm": 2.438185453414917,
+      "learning_rate": 4.789555502616258e-06,
+      "loss": 0.2783,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6845655693470049,
+      "grad_norm": 2.483915328979492,
+      "learning_rate": 4.782096608980268e-06,
+      "loss": 0.3057,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6848354020507286,
+      "grad_norm": 1.1009361743927002,
+      "learning_rate": 4.7746417022834205e-06,
+      "loss": 0.2292,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6851052347544523,
+      "grad_norm": 0.8804078698158264,
+      "learning_rate": 4.767190788221922e-06,
+      "loss": 0.2647,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6853750674581759,
+      "grad_norm": 1.4404693841934204,
+      "learning_rate": 4.759743872488918e-06,
+      "loss": 0.2856,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6856449001618996,
+      "grad_norm": 0.7835676074028015,
+      "learning_rate": 4.752300960774522e-06,
+      "loss": 0.2859,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6859147328656233,
+      "grad_norm": 1.0020006895065308,
+      "learning_rate": 4.744862058765776e-06,
+      "loss": 0.2584,
+      "step": 2542
+    },
+    {
+      "epoch": 0.686184565569347,
+      "grad_norm": 0.7443634867668152,
+      "learning_rate": 4.737427172146645e-06,
+      "loss": 0.2772,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6864543982730706,
+      "grad_norm": 0.9347329139709473,
+      "learning_rate": 4.729996306598054e-06,
+      "loss": 0.2686,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6867242309767944,
+      "grad_norm": 1.2908965349197388,
+      "learning_rate": 4.722569467797837e-06,
+      "loss": 0.2509,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6869940636805181,
+      "grad_norm": 0.8737359642982483,
+      "learning_rate": 4.715146661420746e-06,
+      "loss": 0.2172,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6872638963842418,
+      "grad_norm": 1.0590243339538574,
+      "learning_rate": 4.707727893138471e-06,
+      "loss": 0.2723,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6875337290879655,
+      "grad_norm": 0.8968498706817627,
+      "learning_rate": 4.700313168619608e-06,
+      "loss": 0.2828,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6878035617916891,
+      "grad_norm": 1.0944011211395264,
+      "learning_rate": 4.692902493529649e-06,
+      "loss": 0.248,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6880733944954128,
+      "grad_norm": 1.4295835494995117,
+      "learning_rate": 4.685495873531015e-06,
+      "loss": 0.2495,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6883432271991365,
+      "grad_norm": 1.2530772686004639,
+      "learning_rate": 4.6780933142830185e-06,
+      "loss": 0.2731,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6886130599028603,
+      "grad_norm": 1.0650092363357544,
+      "learning_rate": 4.670694821441862e-06,
+      "loss": 0.2746,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6888828926065839,
+      "grad_norm": 1.0256997346878052,
+      "learning_rate": 4.663300400660647e-06,
+      "loss": 0.2744,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6891527253103076,
+      "grad_norm": 1.226362943649292,
+      "learning_rate": 4.655910057589377e-06,
+      "loss": 0.2538,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6894225580140313,
+      "grad_norm": 0.791283369064331,
+      "learning_rate": 4.648523797874912e-06,
+      "loss": 0.2768,
+      "step": 2555
+    },
+    {
+      "epoch": 0.689692390717755,
+      "grad_norm": 0.9433553814888,
+      "learning_rate": 4.64114162716101e-06,
+      "loss": 0.2698,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6899622234214787,
+      "grad_norm": 1.126847743988037,
+      "learning_rate": 4.6337635510883125e-06,
+      "loss": 0.2776,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6902320561252023,
+      "grad_norm": 1.635595440864563,
+      "learning_rate": 4.626389575294309e-06,
+      "loss": 0.2564,
+      "step": 2558
+    },
+    {
+      "epoch": 0.690501888828926,
+      "grad_norm": 0.9314450025558472,
+      "learning_rate": 4.619019705413369e-06,
+      "loss": 0.2433,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6907717215326498,
+      "grad_norm": 1.374998927116394,
+      "learning_rate": 4.611653947076732e-06,
+      "loss": 0.2474,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6910415542363735,
+      "grad_norm": 0.7961729168891907,
+      "learning_rate": 4.604292305912488e-06,
+      "loss": 0.2651,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6913113869400972,
+      "grad_norm": 1.2278457880020142,
+      "learning_rate": 4.596934787545568e-06,
+      "loss": 0.2903,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6915812196438208,
+      "grad_norm": 1.2137348651885986,
+      "learning_rate": 4.589581397597781e-06,
+      "loss": 0.2929,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6918510523475445,
+      "grad_norm": 1.4576600790023804,
+      "learning_rate": 4.582232141687763e-06,
+      "loss": 0.2603,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6921208850512682,
+      "grad_norm": 1.0567086935043335,
+      "learning_rate": 4.574887025430986e-06,
+      "loss": 0.2615,
+      "step": 2565
+    },
+    {
+      "epoch": 0.692390717754992,
+      "grad_norm": 1.06019127368927,
+      "learning_rate": 4.567546054439777e-06,
+      "loss": 0.2592,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6926605504587156,
+      "grad_norm": 1.0018277168273926,
+      "learning_rate": 4.560209234323288e-06,
+      "loss": 0.2787,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6929303831624393,
+      "grad_norm": 1.063462257385254,
+      "learning_rate": 4.552876570687485e-06,
+      "loss": 0.2713,
+      "step": 2568
+    },
+    {
+      "epoch": 0.693200215866163,
+      "grad_norm": 1.3597807884216309,
+      "learning_rate": 4.545548069135183e-06,
+      "loss": 0.2495,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6934700485698867,
+      "grad_norm": 0.8036149740219116,
+      "learning_rate": 4.538223735266004e-06,
+      "loss": 0.2359,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6937398812736104,
+      "grad_norm": 1.3266559839248657,
+      "learning_rate": 4.530903574676375e-06,
+      "loss": 0.2505,
+      "step": 2571
+    },
+    {
+      "epoch": 0.694009713977334,
+      "grad_norm": 1.3789161443710327,
+      "learning_rate": 4.523587592959557e-06,
+      "loss": 0.2988,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6942795466810577,
+      "grad_norm": 1.075484275817871,
+      "learning_rate": 4.516275795705606e-06,
+      "loss": 0.2658,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6945493793847815,
+      "grad_norm": 1.5469354391098022,
+      "learning_rate": 4.50896818850137e-06,
+      "loss": 0.2493,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6948192120885052,
+      "grad_norm": 1.3242665529251099,
+      "learning_rate": 4.501664776930516e-06,
+      "loss": 0.2199,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6950890447922288,
+      "grad_norm": 1.0644514560699463,
+      "learning_rate": 4.4943655665735e-06,
+      "loss": 0.2803,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6953588774959525,
+      "grad_norm": 0.8293008804321289,
+      "learning_rate": 4.487070563007547e-06,
+      "loss": 0.2589,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6956287101996762,
+      "grad_norm": 1.6318359375,
+      "learning_rate": 4.479779771806699e-06,
+      "loss": 0.2628,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6958985429033999,
+      "grad_norm": 0.7406837940216064,
+      "learning_rate": 4.472493198541763e-06,
+      "loss": 0.2422,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6961683756071236,
+      "grad_norm": 1.021897792816162,
+      "learning_rate": 4.4652108487803134e-06,
+      "loss": 0.2598,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6964382083108472,
+      "grad_norm": 1.3043224811553955,
+      "learning_rate": 4.457932728086721e-06,
+      "loss": 0.2859,
+      "step": 2581
+    },
+    {
+      "epoch": 0.696708041014571,
+      "grad_norm": 1.6670291423797607,
+      "learning_rate": 4.450658842022112e-06,
+      "loss": 0.2506,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6969778737182947,
+      "grad_norm": 1.0219188928604126,
+      "learning_rate": 4.443389196144367e-06,
+      "loss": 0.242,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 1.2464905977249146,
+      "learning_rate": 4.436123796008149e-06,
+      "loss": 0.2643,
+      "step": 2584
+    },
+    {
+      "epoch": 0.697517539125742,
+      "grad_norm": 1.1516501903533936,
+      "learning_rate": 4.428862647164865e-06,
+      "loss": 0.2743,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6977873718294657,
+      "grad_norm": 1.091568112373352,
+      "learning_rate": 4.421605755162664e-06,
+      "loss": 0.2493,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6980572045331894,
+      "grad_norm": 1.0625407695770264,
+      "learning_rate": 4.414353125546463e-06,
+      "loss": 0.2397,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6983270372369131,
+      "grad_norm": 1.0322537422180176,
+      "learning_rate": 4.407104763857912e-06,
+      "loss": 0.3204,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6985968699406369,
+      "grad_norm": 1.0884380340576172,
+      "learning_rate": 4.399860675635388e-06,
+      "loss": 0.286,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6988667026443605,
+      "grad_norm": 1.6380945444107056,
+      "learning_rate": 4.392620866414026e-06,
+      "loss": 0.2728,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6991365353480842,
+      "grad_norm": 1.0981178283691406,
+      "learning_rate": 4.385385341725679e-06,
+      "loss": 0.2759,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6994063680518079,
+      "grad_norm": 1.2444257736206055,
+      "learning_rate": 4.378154107098918e-06,
+      "loss": 0.3017,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6996762007555316,
+      "grad_norm": 0.9715368747711182,
+      "learning_rate": 4.370927168059053e-06,
+      "loss": 0.2812,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6999460334592552,
+      "grad_norm": 1.6477843523025513,
+      "learning_rate": 4.363704530128102e-06,
+      "loss": 0.2786,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7002158661629789,
+      "grad_norm": 1.5219430923461914,
+      "learning_rate": 4.3564861988248e-06,
+      "loss": 0.2785,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7004856988667026,
+      "grad_norm": 0.7957215905189514,
+      "learning_rate": 4.349272179664586e-06,
+      "loss": 0.2563,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7007555315704264,
+      "grad_norm": 0.8375627398490906,
+      "learning_rate": 4.342062478159612e-06,
+      "loss": 0.2465,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7010253642741501,
+      "grad_norm": 0.8976210951805115,
+      "learning_rate": 4.3348570998187235e-06,
+      "loss": 0.2428,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7012951969778737,
+      "grad_norm": 0.984353244304657,
+      "learning_rate": 4.327656050147467e-06,
+      "loss": 0.2734,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7015650296815974,
+      "grad_norm": 1.8998918533325195,
+      "learning_rate": 4.320459334648083e-06,
+      "loss": 0.2923,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7018348623853211,
+      "grad_norm": 0.8569320440292358,
+      "learning_rate": 4.313266958819494e-06,
+      "loss": 0.2611,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7021046950890448,
+      "grad_norm": 0.7905661463737488,
+      "learning_rate": 4.3060789281573135e-06,
+      "loss": 0.2544,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7023745277927684,
+      "grad_norm": 1.3752340078353882,
+      "learning_rate": 4.298895248153831e-06,
+      "loss": 0.2446,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7026443604964921,
+      "grad_norm": 1.1675658226013184,
+      "learning_rate": 4.291715924298011e-06,
+      "loss": 0.2428,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7029141932002159,
+      "grad_norm": 0.8075810074806213,
+      "learning_rate": 4.284540962075494e-06,
+      "loss": 0.2361,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7031840259039396,
+      "grad_norm": 0.7594898343086243,
+      "learning_rate": 4.277370366968583e-06,
+      "loss": 0.2513,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7034538586076633,
+      "grad_norm": 1.0171515941619873,
+      "learning_rate": 4.270204144456247e-06,
+      "loss": 0.2473,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7037236913113869,
+      "grad_norm": 1.1782780885696411,
+      "learning_rate": 4.263042300014112e-06,
+      "loss": 0.2611,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7039935240151106,
+      "grad_norm": 1.2921159267425537,
+      "learning_rate": 4.255884839114461e-06,
+      "loss": 0.255,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7042633567188343,
+      "grad_norm": 0.9987583756446838,
+      "learning_rate": 4.248731767226227e-06,
+      "loss": 0.2501,
+      "step": 2610
+    },
+    {
+      "epoch": 0.704533189422558,
+      "grad_norm": 1.8600718975067139,
+      "learning_rate": 4.2415830898149864e-06,
+      "loss": 0.2324,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7048030221262817,
+      "grad_norm": 1.1190707683563232,
+      "learning_rate": 4.234438812342962e-06,
+      "loss": 0.2594,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7050728548300054,
+      "grad_norm": 0.9465727806091309,
+      "learning_rate": 4.2272989402690136e-06,
+      "loss": 0.255,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7053426875337291,
+      "grad_norm": 0.9407065510749817,
+      "learning_rate": 4.220163479048632e-06,
+      "loss": 0.2647,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7056125202374528,
+      "grad_norm": 0.9355469346046448,
+      "learning_rate": 4.213032434133941e-06,
+      "loss": 0.2577,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 1.3126226663589478,
+      "learning_rate": 4.205905810973689e-06,
+      "loss": 0.251,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7061521856449001,
+      "grad_norm": 1.1453746557235718,
+      "learning_rate": 4.198783615013247e-06,
+      "loss": 0.2692,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7064220183486238,
+      "grad_norm": 0.8964697122573853,
+      "learning_rate": 4.191665851694598e-06,
+      "loss": 0.2513,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7066918510523476,
+      "grad_norm": 0.8926539421081543,
+      "learning_rate": 4.184552526456347e-06,
+      "loss": 0.2768,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7069616837560713,
+      "grad_norm": 0.9268461465835571,
+      "learning_rate": 4.177443644733699e-06,
+      "loss": 0.2411,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7072315164597949,
+      "grad_norm": 1.4377031326293945,
+      "learning_rate": 4.170339211958469e-06,
+      "loss": 0.2622,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7075013491635186,
+      "grad_norm": 0.7510713338851929,
+      "learning_rate": 4.163239233559072e-06,
+      "loss": 0.2254,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7077711818672423,
+      "grad_norm": 1.3356330394744873,
+      "learning_rate": 4.156143714960517e-06,
+      "loss": 0.2604,
+      "step": 2623
+    },
+    {
+      "epoch": 0.708041014570966,
+      "grad_norm": 1.069567084312439,
+      "learning_rate": 4.149052661584409e-06,
+      "loss": 0.2518,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7083108472746897,
+      "grad_norm": 1.7971376180648804,
+      "learning_rate": 4.1419660788489365e-06,
+      "loss": 0.2673,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7085806799784133,
+      "grad_norm": 2.441920757293701,
+      "learning_rate": 4.134883972168877e-06,
+      "loss": 0.2511,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7088505126821371,
+      "grad_norm": 1.0382018089294434,
+      "learning_rate": 4.127806346955587e-06,
+      "loss": 0.2462,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7091203453858608,
+      "grad_norm": 0.9174507856369019,
+      "learning_rate": 4.120733208616993e-06,
+      "loss": 0.2327,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7093901780895845,
+      "grad_norm": 1.8250417709350586,
+      "learning_rate": 4.1136645625576e-06,
+      "loss": 0.2611,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7096600107933081,
+      "grad_norm": 1.4865002632141113,
+      "learning_rate": 4.106600414178479e-06,
+      "loss": 0.2584,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7099298434970318,
+      "grad_norm": 1.0687217712402344,
+      "learning_rate": 4.099540768877262e-06,
+      "loss": 0.2461,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7101996762007555,
+      "grad_norm": 1.1378074884414673,
+      "learning_rate": 4.092485632048142e-06,
+      "loss": 0.2527,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7104695089044792,
+      "grad_norm": 1.2257922887802124,
+      "learning_rate": 4.085435009081866e-06,
+      "loss": 0.2854,
+      "step": 2633
+    },
+    {
+      "epoch": 0.710739341608203,
+      "grad_norm": 1.0108143091201782,
+      "learning_rate": 4.078388905365734e-06,
+      "loss": 0.2534,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7110091743119266,
+      "grad_norm": 0.9591968655586243,
+      "learning_rate": 4.071347326283592e-06,
+      "loss": 0.2756,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7112790070156503,
+      "grad_norm": 0.9902868270874023,
+      "learning_rate": 4.064310277215828e-06,
+      "loss": 0.2648,
+      "step": 2636
+    },
+    {
+      "epoch": 0.711548839719374,
+      "grad_norm": 0.9005113840103149,
+      "learning_rate": 4.057277763539368e-06,
+      "loss": 0.2571,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7118186724230977,
+      "grad_norm": 1.7900928258895874,
+      "learning_rate": 4.050249790627675e-06,
+      "loss": 0.2549,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7120885051268214,
+      "grad_norm": 1.2770901918411255,
+      "learning_rate": 4.043226363850741e-06,
+      "loss": 0.2659,
+      "step": 2639
+    },
+    {
+      "epoch": 0.712358337830545,
+      "grad_norm": 1.763883113861084,
+      "learning_rate": 4.036207488575083e-06,
+      "loss": 0.2659,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7126281705342687,
+      "grad_norm": 1.38455069065094,
+      "learning_rate": 4.029193170163746e-06,
+      "loss": 0.2909,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7128980032379925,
+      "grad_norm": 1.1126168966293335,
+      "learning_rate": 4.022183413976283e-06,
+      "loss": 0.2431,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7131678359417162,
+      "grad_norm": 0.9228575229644775,
+      "learning_rate": 4.015178225368771e-06,
+      "loss": 0.2496,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7134376686454398,
+      "grad_norm": 1.1832232475280762,
+      "learning_rate": 4.008177609693791e-06,
+      "loss": 0.2442,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7137075013491635,
+      "grad_norm": 0.9664711952209473,
+      "learning_rate": 4.001181572300434e-06,
+      "loss": 0.295,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7139773340528872,
+      "grad_norm": 1.4989123344421387,
+      "learning_rate": 3.994190118534288e-06,
+      "loss": 0.2675,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7142471667566109,
+      "grad_norm": 1.0096102952957153,
+      "learning_rate": 3.9872032537374425e-06,
+      "loss": 0.2682,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7145169994603346,
+      "grad_norm": 1.4161955118179321,
+      "learning_rate": 3.9802209832484805e-06,
+      "loss": 0.2289,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7147868321640583,
+      "grad_norm": 0.988493025302887,
+      "learning_rate": 3.973243312402472e-06,
+      "loss": 0.2896,
+      "step": 2649
+    },
+    {
+      "epoch": 0.715056664867782,
+      "grad_norm": 1.7877819538116455,
+      "learning_rate": 3.966270246530975e-06,
+      "loss": 0.2487,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7153264975715057,
+      "grad_norm": 0.7501274347305298,
+      "learning_rate": 3.959301790962026e-06,
+      "loss": 0.2358,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 1.570910930633545,
+      "learning_rate": 3.952337951020142e-06,
+      "loss": 0.2793,
+      "step": 2652
+    },
+    {
+      "epoch": 0.715866162978953,
+      "grad_norm": 0.820387601852417,
+      "learning_rate": 3.945378732026311e-06,
+      "loss": 0.2662,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7161359956826767,
+      "grad_norm": 1.7187386751174927,
+      "learning_rate": 3.938424139297994e-06,
+      "loss": 0.2726,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7164058283864004,
+      "grad_norm": 1.0269886255264282,
+      "learning_rate": 3.9314741781491105e-06,
+      "loss": 0.2714,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7166756610901241,
+      "grad_norm": 1.0020525455474854,
+      "learning_rate": 3.924528853890046e-06,
+      "loss": 0.2564,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7169454937938479,
+      "grad_norm": 1.6605523824691772,
+      "learning_rate": 3.9175881718276435e-06,
+      "loss": 0.2544,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7172153264975715,
+      "grad_norm": 0.8622308373451233,
+      "learning_rate": 3.910652137265194e-06,
+      "loss": 0.2753,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7174851592012952,
+      "grad_norm": 0.9320884346961975,
+      "learning_rate": 3.90372075550244e-06,
+      "loss": 0.2728,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7177549919050189,
+      "grad_norm": 0.8287668228149414,
+      "learning_rate": 3.896794031835579e-06,
+      "loss": 0.249,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7180248246087426,
+      "grad_norm": 0.8364244103431702,
+      "learning_rate": 3.889871971557229e-06,
+      "loss": 0.2836,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7182946573124662,
+      "grad_norm": 1.969671607017517,
+      "learning_rate": 3.882954579956455e-06,
+      "loss": 0.2449,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7185644900161899,
+      "grad_norm": 1.7685270309448242,
+      "learning_rate": 3.8760418623187676e-06,
+      "loss": 0.2797,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7188343227199137,
+      "grad_norm": 0.7836909294128418,
+      "learning_rate": 3.869133823926083e-06,
+      "loss": 0.2427,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7191041554236374,
+      "grad_norm": 2.2423346042633057,
+      "learning_rate": 3.862230470056749e-06,
+      "loss": 0.3098,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7193739881273611,
+      "grad_norm": 2.4500069618225098,
+      "learning_rate": 3.8553318059855536e-06,
+      "loss": 0.2946,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7196438208310847,
+      "grad_norm": 1.6622202396392822,
+      "learning_rate": 3.848437836983672e-06,
+      "loss": 0.2653,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7199136535348084,
+      "grad_norm": 1.4512710571289062,
+      "learning_rate": 3.841548568318706e-06,
+      "loss": 0.2508,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7201834862385321,
+      "grad_norm": 1.8699828386306763,
+      "learning_rate": 3.834664005254677e-06,
+      "loss": 0.2592,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7204533189422558,
+      "grad_norm": 1.0899070501327515,
+      "learning_rate": 3.827784153051986e-06,
+      "loss": 0.2778,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7207231516459794,
+      "grad_norm": 1.6079155206680298,
+      "learning_rate": 3.820909016967452e-06,
+      "loss": 0.2722,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7209929843497032,
+      "grad_norm": 1.191772222518921,
+      "learning_rate": 3.814038602254294e-06,
+      "loss": 0.2501,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7212628170534269,
+      "grad_norm": 1.0726553201675415,
+      "learning_rate": 3.807172914162106e-06,
+      "loss": 0.2465,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7215326497571506,
+      "grad_norm": 1.470635175704956,
+      "learning_rate": 3.8003119579368806e-06,
+      "loss": 0.2708,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7218024824608743,
+      "grad_norm": 1.5601539611816406,
+      "learning_rate": 3.793455738821005e-06,
+      "loss": 0.2719,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7220723151645979,
+      "grad_norm": 1.5410610437393188,
+      "learning_rate": 3.786604262053224e-06,
+      "loss": 0.2358,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7223421478683216,
+      "grad_norm": 1.0174764394760132,
+      "learning_rate": 3.7797575328686733e-06,
+      "loss": 0.2814,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7226119805720453,
+      "grad_norm": 0.8663463592529297,
+      "learning_rate": 3.7729155564988685e-06,
+      "loss": 0.2661,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7228818132757691,
+      "grad_norm": 0.8009067177772522,
+      "learning_rate": 3.7660783381716726e-06,
+      "loss": 0.2328,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7231516459794927,
+      "grad_norm": 1.9528971910476685,
+      "learning_rate": 3.7592458831113256e-06,
+      "loss": 0.2352,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7234214786832164,
+      "grad_norm": 0.895017147064209,
+      "learning_rate": 3.7524181965384375e-06,
+      "loss": 0.2624,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7236913113869401,
+      "grad_norm": 1.104075312614441,
+      "learning_rate": 3.7455952836699515e-06,
+      "loss": 0.2529,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7239611440906638,
+      "grad_norm": 1.6323492527008057,
+      "learning_rate": 3.738777149719178e-06,
+      "loss": 0.3105,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7242309767943875,
+      "grad_norm": 0.873004674911499,
+      "learning_rate": 3.7319637998957836e-06,
+      "loss": 0.2506,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7245008094981111,
+      "grad_norm": 1.0012136697769165,
+      "learning_rate": 3.7251552394057587e-06,
+      "loss": 0.2734,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7247706422018348,
+      "grad_norm": 1.7199792861938477,
+      "learning_rate": 3.718351473451448e-06,
+      "loss": 0.2388,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7250404749055586,
+      "grad_norm": 1.70649254322052,
+      "learning_rate": 3.7115525072315305e-06,
+      "loss": 0.2141,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7253103076092823,
+      "grad_norm": 1.118070125579834,
+      "learning_rate": 3.7047583459410164e-06,
+      "loss": 0.2501,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7255801403130059,
+      "grad_norm": 1.7676512002944946,
+      "learning_rate": 3.697968994771246e-06,
+      "loss": 0.2655,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7258499730167296,
+      "grad_norm": 0.9134480953216553,
+      "learning_rate": 3.691184458909882e-06,
+      "loss": 0.2497,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7261198057204533,
+      "grad_norm": 1.642382264137268,
+      "learning_rate": 3.6844047435409104e-06,
+      "loss": 0.2753,
+      "step": 2691
+    },
+    {
+      "epoch": 0.726389638424177,
+      "grad_norm": 1.1990115642547607,
+      "learning_rate": 3.6776298538446307e-06,
+      "loss": 0.2312,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7266594711279007,
+      "grad_norm": 0.8796782493591309,
+      "learning_rate": 3.6708597949976565e-06,
+      "loss": 0.2521,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7269293038316244,
+      "grad_norm": 0.9950150847434998,
+      "learning_rate": 3.6640945721729192e-06,
+      "loss": 0.2849,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7271991365353481,
+      "grad_norm": 0.9386069774627686,
+      "learning_rate": 3.6573341905396378e-06,
+      "loss": 0.2488,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7274689692390718,
+      "grad_norm": 1.0978105068206787,
+      "learning_rate": 3.650578655263339e-06,
+      "loss": 0.2448,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7277388019427955,
+      "grad_norm": 1.0399280786514282,
+      "learning_rate": 3.6438279715058613e-06,
+      "loss": 0.2569,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7280086346465191,
+      "grad_norm": 0.9988282322883606,
+      "learning_rate": 3.6370821444253112e-06,
+      "loss": 0.2854,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7282784673502428,
+      "grad_norm": 1.0492562055587769,
+      "learning_rate": 3.630341179176098e-06,
+      "loss": 0.2192,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7285483000539665,
+      "grad_norm": 0.9541296362876892,
+      "learning_rate": 3.623605080908923e-06,
+      "loss": 0.2425,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7288181327576903,
+      "grad_norm": 0.76021808385849,
+      "learning_rate": 3.6168738547707516e-06,
+      "loss": 0.2534,
+      "step": 2701
+    },
+    {
+      "epoch": 0.729087965461414,
+      "grad_norm": 3.112220287322998,
+      "learning_rate": 3.610147505904835e-06,
+      "loss": 0.3194,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7293577981651376,
+      "grad_norm": 1.8213475942611694,
+      "learning_rate": 3.603426039450708e-06,
+      "loss": 0.2753,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7296276308688613,
+      "grad_norm": 1.0664585828781128,
+      "learning_rate": 3.5967094605441545e-06,
+      "loss": 0.2737,
+      "step": 2704
+    },
+    {
+      "epoch": 0.729897463572585,
+      "grad_norm": 0.92326420545578,
+      "learning_rate": 3.589997774317233e-06,
+      "loss": 0.2581,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7301672962763087,
+      "grad_norm": 2.693870782852173,
+      "learning_rate": 3.583290985898278e-06,
+      "loss": 0.3278,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7304371289800324,
+      "grad_norm": 1.9616189002990723,
+      "learning_rate": 3.5765891004118546e-06,
+      "loss": 0.2652,
+      "step": 2707
+    },
+    {
+      "epoch": 0.730706961683756,
+      "grad_norm": 2.002969264984131,
+      "learning_rate": 3.5698921229787973e-06,
+      "loss": 0.2212,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7309767943874798,
+      "grad_norm": 0.8798991441726685,
+      "learning_rate": 3.5632000587161986e-06,
+      "loss": 0.2988,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7312466270912035,
+      "grad_norm": 0.7765748500823975,
+      "learning_rate": 3.5565129127373765e-06,
+      "loss": 0.2668,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7315164597949272,
+      "grad_norm": 1.9220932722091675,
+      "learning_rate": 3.5498306901518997e-06,
+      "loss": 0.2364,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7317862924986508,
+      "grad_norm": 1.700018286705017,
+      "learning_rate": 3.5431533960655893e-06,
+      "loss": 0.2821,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7320561252023745,
+      "grad_norm": 1.344620704650879,
+      "learning_rate": 3.5364810355804758e-06,
+      "loss": 0.2599,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7323259579060982,
+      "grad_norm": 1.0655697584152222,
+      "learning_rate": 3.5298136137948335e-06,
+      "loss": 0.2581,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7325957906098219,
+      "grad_norm": 0.8452352285385132,
+      "learning_rate": 3.52315113580317e-06,
+      "loss": 0.2572,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7328656233135457,
+      "grad_norm": 1.2864820957183838,
+      "learning_rate": 3.5164936066961984e-06,
+      "loss": 0.2338,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7331354560172693,
+      "grad_norm": 1.024914026260376,
+      "learning_rate": 3.509841031560858e-06,
+      "loss": 0.264,
+      "step": 2717
+    },
+    {
+      "epoch": 0.733405288720993,
+      "grad_norm": 1.0051769018173218,
+      "learning_rate": 3.503193415480315e-06,
+      "loss": 0.2792,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7336751214247167,
+      "grad_norm": 1.7980308532714844,
+      "learning_rate": 3.4965507635339246e-06,
+      "loss": 0.2524,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.9470153450965881,
+      "learning_rate": 3.4899130807972603e-06,
+      "loss": 0.266,
+      "step": 2720
+    },
+    {
+      "epoch": 0.734214786832164,
+      "grad_norm": 1.384989857673645,
+      "learning_rate": 3.483280372342107e-06,
+      "loss": 0.2242,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7344846195358877,
+      "grad_norm": 0.8876177072525024,
+      "learning_rate": 3.476652643236431e-06,
+      "loss": 0.2239,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7347544522396114,
+      "grad_norm": 2.4033467769622803,
+      "learning_rate": 3.4700298985444025e-06,
+      "loss": 0.2783,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7350242849433352,
+      "grad_norm": 1.4728809595108032,
+      "learning_rate": 3.4634121433263934e-06,
+      "loss": 0.2454,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7352941176470589,
+      "grad_norm": 1.3175711631774902,
+      "learning_rate": 3.456799382638942e-06,
+      "loss": 0.2328,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7355639503507825,
+      "grad_norm": 2.014012336730957,
+      "learning_rate": 3.450191621534784e-06,
+      "loss": 0.3091,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7358337830545062,
+      "grad_norm": 1.0086828470230103,
+      "learning_rate": 3.443588865062839e-06,
+      "loss": 0.2672,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7361036157582299,
+      "grad_norm": 1.289802074432373,
+      "learning_rate": 3.436991118268195e-06,
+      "loss": 0.2439,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7363734484619536,
+      "grad_norm": 1.3743854761123657,
+      "learning_rate": 3.430398386192103e-06,
+      "loss": 0.2463,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7366432811656772,
+      "grad_norm": 2.017167568206787,
+      "learning_rate": 3.4238106738720045e-06,
+      "loss": 0.2912,
+      "step": 2730
+    },
+    {
+      "epoch": 0.736913113869401,
+      "grad_norm": 0.8100365400314331,
+      "learning_rate": 3.4172279863414912e-06,
+      "loss": 0.2746,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7371829465731247,
+      "grad_norm": 1.2810996770858765,
+      "learning_rate": 3.4106503286303107e-06,
+      "loss": 0.276,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7374527792768484,
+      "grad_norm": 0.9085292816162109,
+      "learning_rate": 3.4040777057643825e-06,
+      "loss": 0.2564,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7377226119805721,
+      "grad_norm": 0.8259239196777344,
+      "learning_rate": 3.3975101227657726e-06,
+      "loss": 0.2638,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7379924446842957,
+      "grad_norm": 2.2571685314178467,
+      "learning_rate": 3.390947584652684e-06,
+      "loss": 0.2877,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7382622773880194,
+      "grad_norm": 0.8403110504150391,
+      "learning_rate": 3.384390096439485e-06,
+      "loss": 0.282,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7385321100917431,
+      "grad_norm": 0.9671818614006042,
+      "learning_rate": 3.377837663136676e-06,
+      "loss": 0.265,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7388019427954668,
+      "grad_norm": 1.1827454566955566,
+      "learning_rate": 3.3712902897508846e-06,
+      "loss": 0.2586,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7390717754991905,
+      "grad_norm": 0.8864468336105347,
+      "learning_rate": 3.3647479812848936e-06,
+      "loss": 0.2616,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7393416082029142,
+      "grad_norm": 1.479189157485962,
+      "learning_rate": 3.3582107427376044e-06,
+      "loss": 0.2349,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7396114409066379,
+      "grad_norm": 0.9792450070381165,
+      "learning_rate": 3.3516785791040342e-06,
+      "loss": 0.2514,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7398812736103616,
+      "grad_norm": 1.1871862411499023,
+      "learning_rate": 3.345151495375343e-06,
+      "loss": 0.2323,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7401511063140853,
+      "grad_norm": 0.9173151850700378,
+      "learning_rate": 3.338629496538799e-06,
+      "loss": 0.2364,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7404209390178089,
+      "grad_norm": 0.9113356471061707,
+      "learning_rate": 3.332112587577777e-06,
+      "loss": 0.2876,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7406907717215326,
+      "grad_norm": 0.9618027806282043,
+      "learning_rate": 3.3256007734717797e-06,
+      "loss": 0.2477,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7409606044252564,
+      "grad_norm": 1.0229202508926392,
+      "learning_rate": 3.3190940591964094e-06,
+      "loss": 0.2344,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7412304371289801,
+      "grad_norm": 0.9451637268066406,
+      "learning_rate": 3.3125924497233597e-06,
+      "loss": 0.3342,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7415002698327037,
+      "grad_norm": 0.7947325706481934,
+      "learning_rate": 3.3060959500204447e-06,
+      "loss": 0.2634,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7417701025364274,
+      "grad_norm": 1.0273754596710205,
+      "learning_rate": 3.299604565051564e-06,
+      "loss": 0.2775,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7420399352401511,
+      "grad_norm": 1.2836337089538574,
+      "learning_rate": 3.2931182997767007e-06,
+      "loss": 0.2795,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7423097679438748,
+      "grad_norm": 0.8906709551811218,
+      "learning_rate": 3.286637159151942e-06,
+      "loss": 0.2621,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7425796006475985,
+      "grad_norm": 0.8027698397636414,
+      "learning_rate": 3.2801611481294538e-06,
+      "loss": 0.2527,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7428494333513221,
+      "grad_norm": 1.1168071031570435,
+      "learning_rate": 3.273690271657469e-06,
+      "loss": 0.2368,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7431192660550459,
+      "grad_norm": 1.4753191471099854,
+      "learning_rate": 3.2672245346803213e-06,
+      "loss": 0.2262,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7433890987587696,
+      "grad_norm": 0.8558721542358398,
+      "learning_rate": 3.260763942138402e-06,
+      "loss": 0.2754,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7436589314624933,
+      "grad_norm": 0.8160566091537476,
+      "learning_rate": 3.2543084989681663e-06,
+      "loss": 0.2431,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7439287641662169,
+      "grad_norm": 0.749883234500885,
+      "learning_rate": 3.2478582101021527e-06,
+      "loss": 0.2261,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7441985968699406,
+      "grad_norm": 0.8594847321510315,
+      "learning_rate": 3.2414130804689492e-06,
+      "loss": 0.2476,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7444684295736643,
+      "grad_norm": 2.026195764541626,
+      "learning_rate": 3.234973114993203e-06,
+      "loss": 0.2706,
+      "step": 2759
+    },
+    {
+      "epoch": 0.744738262277388,
+      "grad_norm": 0.9667326211929321,
+      "learning_rate": 3.2285383185956166e-06,
+      "loss": 0.2468,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7450080949811118,
+      "grad_norm": 1.1130951642990112,
+      "learning_rate": 3.2221086961929425e-06,
+      "loss": 0.2853,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7452779276848354,
+      "grad_norm": 1.220273494720459,
+      "learning_rate": 3.2156842526979827e-06,
+      "loss": 0.2366,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7455477603885591,
+      "grad_norm": 1.1015572547912598,
+      "learning_rate": 3.209264993019576e-06,
+      "loss": 0.2324,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7458175930922828,
+      "grad_norm": 1.2186126708984375,
+      "learning_rate": 3.202850922062607e-06,
+      "loss": 0.2366,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7460874257960065,
+      "grad_norm": 0.7686904668807983,
+      "learning_rate": 3.196442044727991e-06,
+      "loss": 0.2405,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7463572584997301,
+      "grad_norm": 1.2904678583145142,
+      "learning_rate": 3.1900383659126764e-06,
+      "loss": 0.248,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7466270912034538,
+      "grad_norm": 1.295802354812622,
+      "learning_rate": 3.183639890509641e-06,
+      "loss": 0.2684,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7468969239071775,
+      "grad_norm": 1.321737289428711,
+      "learning_rate": 3.1772466234078824e-06,
+      "loss": 0.2636,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7471667566109013,
+      "grad_norm": 1.7181298732757568,
+      "learning_rate": 3.1708585694924242e-06,
+      "loss": 0.3037,
+      "step": 2769
+    },
+    {
+      "epoch": 0.747436589314625,
+      "grad_norm": 1.1747217178344727,
+      "learning_rate": 3.1644757336443023e-06,
+      "loss": 0.2728,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7477064220183486,
+      "grad_norm": 0.8694581985473633,
+      "learning_rate": 3.158098120740568e-06,
+      "loss": 0.23,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7479762547220723,
+      "grad_norm": 0.909735918045044,
+      "learning_rate": 3.1517257356542805e-06,
+      "loss": 0.2749,
+      "step": 2772
+    },
+    {
+      "epoch": 0.748246087425796,
+      "grad_norm": 1.4220868349075317,
+      "learning_rate": 3.1453585832545052e-06,
+      "loss": 0.2415,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7485159201295197,
+      "grad_norm": 1.4815943241119385,
+      "learning_rate": 3.138996668406308e-06,
+      "loss": 0.2687,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7487857528332433,
+      "grad_norm": 1.6655601263046265,
+      "learning_rate": 3.1326399959707555e-06,
+      "loss": 0.2205,
+      "step": 2775
+    },
+    {
+      "epoch": 0.749055585536967,
+      "grad_norm": 1.4859131574630737,
+      "learning_rate": 3.126288570804906e-06,
+      "loss": 0.2295,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7493254182406908,
+      "grad_norm": 0.8820031881332397,
+      "learning_rate": 3.1199423977618095e-06,
+      "loss": 0.2419,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7495952509444145,
+      "grad_norm": 0.9406993389129639,
+      "learning_rate": 3.1136014816905037e-06,
+      "loss": 0.261,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7498650836481382,
+      "grad_norm": 1.9998524188995361,
+      "learning_rate": 3.107265827436007e-06,
+      "loss": 0.2352,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7501349163518618,
+      "grad_norm": 2.139923095703125,
+      "learning_rate": 3.10093543983932e-06,
+      "loss": 0.2322,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7504047490555855,
+      "grad_norm": 2.2339587211608887,
+      "learning_rate": 3.0946103237374183e-06,
+      "loss": 0.2529,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7506745817593092,
+      "grad_norm": 1.09945809841156,
+      "learning_rate": 3.0882904839632476e-06,
+      "loss": 0.2684,
+      "step": 2782
+    },
+    {
+      "epoch": 0.750944414463033,
+      "grad_norm": 1.1904523372650146,
+      "learning_rate": 3.081975925345726e-06,
+      "loss": 0.2906,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7512142471667567,
+      "grad_norm": 2.182112693786621,
+      "learning_rate": 3.0756666527097333e-06,
+      "loss": 0.283,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7514840798704803,
+      "grad_norm": 2.4022836685180664,
+      "learning_rate": 3.0693626708761104e-06,
+      "loss": 0.2728,
+      "step": 2785
+    },
+    {
+      "epoch": 0.751753912574204,
+      "grad_norm": 1.6211256980895996,
+      "learning_rate": 3.0630639846616583e-06,
+      "loss": 0.2271,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7520237452779277,
+      "grad_norm": 0.8532425165176392,
+      "learning_rate": 3.0567705988791274e-06,
+      "loss": 0.2784,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 1.295923113822937,
+      "learning_rate": 3.050482518337221e-06,
+      "loss": 0.2893,
+      "step": 2788
+    },
+    {
+      "epoch": 0.752563410685375,
+      "grad_norm": 0.9625117182731628,
+      "learning_rate": 3.044199747840589e-06,
+      "loss": 0.2416,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7528332433890987,
+      "grad_norm": 0.9042655825614929,
+      "learning_rate": 3.0379222921898223e-06,
+      "loss": 0.2387,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7531030760928225,
+      "grad_norm": 1.0394724607467651,
+      "learning_rate": 3.031650156181453e-06,
+      "loss": 0.2605,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7533729087965462,
+      "grad_norm": 1.8897844552993774,
+      "learning_rate": 3.025383344607944e-06,
+      "loss": 0.2451,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7536427415002699,
+      "grad_norm": 1.3438433408737183,
+      "learning_rate": 3.0191218622576957e-06,
+      "loss": 0.2542,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7539125742039935,
+      "grad_norm": 0.8242205381393433,
+      "learning_rate": 3.012865713915033e-06,
+      "loss": 0.2403,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7541824069077172,
+      "grad_norm": 0.9281934499740601,
+      "learning_rate": 3.0066149043602054e-06,
+      "loss": 0.2417,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7544522396114409,
+      "grad_norm": 1.8561586141586304,
+      "learning_rate": 3.0003694383693847e-06,
+      "loss": 0.2548,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7547220723151646,
+      "grad_norm": 0.934240996837616,
+      "learning_rate": 2.9941293207146573e-06,
+      "loss": 0.2566,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7549919050188882,
+      "grad_norm": 1.5952359437942505,
+      "learning_rate": 2.987894556164025e-06,
+      "loss": 0.2677,
+      "step": 2798
+    },
+    {
+      "epoch": 0.755261737722612,
+      "grad_norm": 1.8446345329284668,
+      "learning_rate": 2.9816651494813996e-06,
+      "loss": 0.2559,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7555315704263357,
+      "grad_norm": 1.0338975191116333,
+      "learning_rate": 2.9754411054265966e-06,
+      "loss": 0.2497,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7558014031300594,
+      "grad_norm": 1.0647785663604736,
+      "learning_rate": 2.969222428755335e-06,
+      "loss": 0.2576,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7560712358337831,
+      "grad_norm": 2.071849822998047,
+      "learning_rate": 2.963009124219236e-06,
+      "loss": 0.2892,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7563410685375067,
+      "grad_norm": 1.0721668004989624,
+      "learning_rate": 2.9568011965658107e-06,
+      "loss": 0.2601,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7566109012412304,
+      "grad_norm": 1.6454716920852661,
+      "learning_rate": 2.950598650538464e-06,
+      "loss": 0.2829,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7568807339449541,
+      "grad_norm": 1.4703394174575806,
+      "learning_rate": 2.9444014908764905e-06,
+      "loss": 0.247,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7571505666486779,
+      "grad_norm": 1.385879635810852,
+      "learning_rate": 2.9382097223150675e-06,
+      "loss": 0.2656,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7574203993524015,
+      "grad_norm": 1.1229740381240845,
+      "learning_rate": 2.9320233495852523e-06,
+      "loss": 0.2715,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7576902320561252,
+      "grad_norm": 1.036169409751892,
+      "learning_rate": 2.92584237741398e-06,
+      "loss": 0.3038,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7579600647598489,
+      "grad_norm": 1.5405993461608887,
+      "learning_rate": 2.91966681052406e-06,
+      "loss": 0.2475,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7582298974635726,
+      "grad_norm": 1.6029067039489746,
+      "learning_rate": 2.913496653634171e-06,
+      "loss": 0.2368,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7584997301672963,
+      "grad_norm": 1.3309963941574097,
+      "learning_rate": 2.9073319114588584e-06,
+      "loss": 0.2698,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7587695628710199,
+      "grad_norm": 1.1400134563446045,
+      "learning_rate": 2.9011725887085286e-06,
+      "loss": 0.2381,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7590393955747436,
+      "grad_norm": 1.4676406383514404,
+      "learning_rate": 2.8950186900894493e-06,
+      "loss": 0.2272,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7593092282784674,
+      "grad_norm": 0.9424062371253967,
+      "learning_rate": 2.888870220303742e-06,
+      "loss": 0.2509,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7595790609821911,
+      "grad_norm": 1.3440684080123901,
+      "learning_rate": 2.882727184049381e-06,
+      "loss": 0.2507,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7598488936859147,
+      "grad_norm": 1.9160703420639038,
+      "learning_rate": 2.876589586020189e-06,
+      "loss": 0.2713,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7601187263896384,
+      "grad_norm": 3.020200252532959,
+      "learning_rate": 2.870457430905832e-06,
+      "loss": 0.2862,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7603885590933621,
+      "grad_norm": 0.8580182790756226,
+      "learning_rate": 2.8643307233918192e-06,
+      "loss": 0.2494,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7606583917970858,
+      "grad_norm": 1.0419074296951294,
+      "learning_rate": 2.858209468159497e-06,
+      "loss": 0.2783,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7609282245008095,
+      "grad_norm": 1.5060081481933594,
+      "learning_rate": 2.8520936698860447e-06,
+      "loss": 0.2117,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7611980572045332,
+      "grad_norm": 1.1949522495269775,
+      "learning_rate": 2.845983333244471e-06,
+      "loss": 0.2494,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7614678899082569,
+      "grad_norm": 1.63998281955719,
+      "learning_rate": 2.8398784629036136e-06,
+      "loss": 0.2386,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7617377226119806,
+      "grad_norm": 1.3019367456436157,
+      "learning_rate": 2.8337790635281338e-06,
+      "loss": 0.2739,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7620075553157043,
+      "grad_norm": 0.9391539692878723,
+      "learning_rate": 2.827685139778511e-06,
+      "loss": 0.2586,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7622773880194279,
+      "grad_norm": 0.8642283082008362,
+      "learning_rate": 2.8215966963110408e-06,
+      "loss": 0.2373,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7625472207231516,
+      "grad_norm": 0.8586753010749817,
+      "learning_rate": 2.815513737777833e-06,
+      "loss": 0.2802,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7628170534268753,
+      "grad_norm": 0.9102429151535034,
+      "learning_rate": 2.809436268826804e-06,
+      "loss": 0.2594,
+      "step": 2827
+    },
+    {
+      "epoch": 0.763086886130599,
+      "grad_norm": 1.029500126838684,
+      "learning_rate": 2.803364294101677e-06,
+      "loss": 0.2217,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7633567188343228,
+      "grad_norm": 1.3859851360321045,
+      "learning_rate": 2.7972978182419776e-06,
+      "loss": 0.239,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7636265515380464,
+      "grad_norm": 1.8983263969421387,
+      "learning_rate": 2.7912368458830295e-06,
+      "loss": 0.262,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7638963842417701,
+      "grad_norm": 2.0265555381774902,
+      "learning_rate": 2.7851813816559505e-06,
+      "loss": 0.2419,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7641662169454938,
+      "grad_norm": 0.8934813141822815,
+      "learning_rate": 2.7791314301876505e-06,
+      "loss": 0.2255,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7644360496492175,
+      "grad_norm": 0.7845723032951355,
+      "learning_rate": 2.7730869961008266e-06,
+      "loss": 0.2656,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 2.2529332637786865,
+      "learning_rate": 2.7670480840139602e-06,
+      "loss": 0.2593,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7649757150566648,
+      "grad_norm": 1.4895058870315552,
+      "learning_rate": 2.7610146985413143e-06,
+      "loss": 0.2584,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7652455477603886,
+      "grad_norm": 2.1284494400024414,
+      "learning_rate": 2.7549868442929286e-06,
+      "loss": 0.2613,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7655153804641123,
+      "grad_norm": 1.8878580331802368,
+      "learning_rate": 2.7489645258746155e-06,
+      "loss": 0.2686,
+      "step": 2837
+    },
+    {
+      "epoch": 0.765785213167836,
+      "grad_norm": 1.267595887184143,
+      "learning_rate": 2.7429477478879583e-06,
+      "loss": 0.2764,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7660550458715596,
+      "grad_norm": 1.5108109712600708,
+      "learning_rate": 2.7369365149303074e-06,
+      "loss": 0.2725,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7663248785752833,
+      "grad_norm": 1.823224425315857,
+      "learning_rate": 2.730930831594775e-06,
+      "loss": 0.278,
+      "step": 2840
+    },
+    {
+      "epoch": 0.766594711279007,
+      "grad_norm": 1.6569315195083618,
+      "learning_rate": 2.7249307024702343e-06,
+      "loss": 0.2396,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7668645439827307,
+      "grad_norm": 1.056585669517517,
+      "learning_rate": 2.7189361321413144e-06,
+      "loss": 0.2481,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7671343766864543,
+      "grad_norm": 0.8570457100868225,
+      "learning_rate": 2.7129471251883954e-06,
+      "loss": 0.2091,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7674042093901781,
+      "grad_norm": 1.024570345878601,
+      "learning_rate": 2.706963686187609e-06,
+      "loss": 0.2458,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7676740420939018,
+      "grad_norm": 1.2781559228897095,
+      "learning_rate": 2.700985819710831e-06,
+      "loss": 0.2465,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7679438747976255,
+      "grad_norm": 0.8759855031967163,
+      "learning_rate": 2.6950135303256776e-06,
+      "loss": 0.269,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7682137075013492,
+      "grad_norm": 0.7685315608978271,
+      "learning_rate": 2.689046822595508e-06,
+      "loss": 0.2334,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7684835402050728,
+      "grad_norm": 1.350806713104248,
+      "learning_rate": 2.683085701079412e-06,
+      "loss": 0.2666,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7687533729087965,
+      "grad_norm": 1.2115999460220337,
+      "learning_rate": 2.677130170332215e-06,
+      "loss": 0.2481,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7690232056125202,
+      "grad_norm": 0.9161841869354248,
+      "learning_rate": 2.671180234904467e-06,
+      "loss": 0.2835,
+      "step": 2850
+    },
+    {
+      "epoch": 0.769293038316244,
+      "grad_norm": 1.1432185173034668,
+      "learning_rate": 2.6652358993424455e-06,
+      "loss": 0.2701,
+      "step": 2851
+    },
+    {
+      "epoch": 0.7695628710199676,
+      "grad_norm": 0.8995304703712463,
+      "learning_rate": 2.6592971681881463e-06,
+      "loss": 0.2577,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7698327037236913,
+      "grad_norm": 1.249277949333191,
+      "learning_rate": 2.6533640459792863e-06,
+      "loss": 0.2776,
+      "step": 2853
+    },
+    {
+      "epoch": 0.770102536427415,
+      "grad_norm": 1.158490538597107,
+      "learning_rate": 2.647436537249294e-06,
+      "loss": 0.2302,
+      "step": 2854
+    },
+    {
+      "epoch": 0.7703723691311387,
+      "grad_norm": 1.1928026676177979,
+      "learning_rate": 2.641514646527311e-06,
+      "loss": 0.2317,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 1.7682350873947144,
+      "learning_rate": 2.6355983783381835e-06,
+      "loss": 0.2839,
+      "step": 2856
+    },
+    {
+      "epoch": 0.770912034538586,
+      "grad_norm": 0.9130220413208008,
+      "learning_rate": 2.629687737202463e-06,
+      "loss": 0.2518,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7711818672423097,
+      "grad_norm": 0.7952666282653809,
+      "learning_rate": 2.6237827276364024e-06,
+      "loss": 0.2254,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7714516999460335,
+      "grad_norm": 0.9822012782096863,
+      "learning_rate": 2.61788335415195e-06,
+      "loss": 0.2491,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7717215326497572,
+      "grad_norm": 1.0469549894332886,
+      "learning_rate": 2.611989621256745e-06,
+      "loss": 0.2567,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7719913653534809,
+      "grad_norm": 1.067867636680603,
+      "learning_rate": 2.6061015334541307e-06,
+      "loss": 0.2764,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7722611980572045,
+      "grad_norm": 0.9471949934959412,
+      "learning_rate": 2.6002190952431163e-06,
+      "loss": 0.2713,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7725310307609282,
+      "grad_norm": 1.0239307880401611,
+      "learning_rate": 2.594342311118403e-06,
+      "loss": 0.2503,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7728008634646519,
+      "grad_norm": 0.8670828342437744,
+      "learning_rate": 2.5884711855703838e-06,
+      "loss": 0.2314,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7730706961683756,
+      "grad_norm": 1.0919568538665771,
+      "learning_rate": 2.5826057230851075e-06,
+      "loss": 0.2538,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7733405288720993,
+      "grad_norm": 0.8379662036895752,
+      "learning_rate": 2.5767459281443064e-06,
+      "loss": 0.2569,
+      "step": 2866
+    },
+    {
+      "epoch": 0.773610361575823,
+      "grad_norm": 1.2212305068969727,
+      "learning_rate": 2.5708918052253883e-06,
+      "loss": 0.2498,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7738801942795467,
+      "grad_norm": 0.7223842740058899,
+      "learning_rate": 2.5650433588014147e-06,
+      "loss": 0.2267,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7741500269832704,
+      "grad_norm": 0.8385284543037415,
+      "learning_rate": 2.5592005933411124e-06,
+      "loss": 0.2337,
+      "step": 2869
+    },
+    {
+      "epoch": 0.7744198596869941,
+      "grad_norm": 1.359933614730835,
+      "learning_rate": 2.5533635133088795e-06,
+      "loss": 0.2184,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7746896923907177,
+      "grad_norm": 1.1184347867965698,
+      "learning_rate": 2.5475321231647533e-06,
+      "loss": 0.2715,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7749595250944414,
+      "grad_norm": 0.8148548603057861,
+      "learning_rate": 2.541706427364431e-06,
+      "loss": 0.2556,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7752293577981652,
+      "grad_norm": 0.9654533863067627,
+      "learning_rate": 2.5358864303592655e-06,
+      "loss": 0.2618,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7754991905018889,
+      "grad_norm": 1.5279268026351929,
+      "learning_rate": 2.5300721365962433e-06,
+      "loss": 0.2492,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7757690232056125,
+      "grad_norm": 0.8118475079536438,
+      "learning_rate": 2.5242635505179956e-06,
+      "loss": 0.2136,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7760388559093362,
+      "grad_norm": 1.3985683917999268,
+      "learning_rate": 2.518460676562805e-06,
+      "loss": 0.2185,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7763086886130599,
+      "grad_norm": 0.8151246309280396,
+      "learning_rate": 2.512663519164571e-06,
+      "loss": 0.2498,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7765785213167836,
+      "grad_norm": 0.9439542293548584,
+      "learning_rate": 2.506872082752834e-06,
+      "loss": 0.2787,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7768483540205073,
+      "grad_norm": 1.1584993600845337,
+      "learning_rate": 2.5010863717527724e-06,
+      "loss": 0.2508,
+      "step": 2879
+    },
+    {
+      "epoch": 0.7771181867242309,
+      "grad_norm": 0.7469279170036316,
+      "learning_rate": 2.4953063905851703e-06,
+      "loss": 0.2288,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7773880194279547,
+      "grad_norm": 1.0640738010406494,
+      "learning_rate": 2.4895321436664454e-06,
+      "loss": 0.2708,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7776578521316784,
+      "grad_norm": 1.1203536987304688,
+      "learning_rate": 2.4837636354086415e-06,
+      "loss": 0.2722,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7779276848354021,
+      "grad_norm": 0.9807953238487244,
+      "learning_rate": 2.478000870219397e-06,
+      "loss": 0.2597,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7781975175391257,
+      "grad_norm": 0.9756600260734558,
+      "learning_rate": 2.4722438525019764e-06,
+      "loss": 0.266,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7784673502428494,
+      "grad_norm": 1.4076931476593018,
+      "learning_rate": 2.466492586655258e-06,
+      "loss": 0.2521,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7787371829465731,
+      "grad_norm": 0.8982987403869629,
+      "learning_rate": 2.4607470770737075e-06,
+      "loss": 0.2631,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7790070156502968,
+      "grad_norm": 1.2331753969192505,
+      "learning_rate": 2.4550073281474016e-06,
+      "loss": 0.2718,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7792768483540206,
+      "grad_norm": 1.2121976613998413,
+      "learning_rate": 2.4492733442620253e-06,
+      "loss": 0.2503,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7795466810577442,
+      "grad_norm": 1.0134589672088623,
+      "learning_rate": 2.443545129798839e-06,
+      "loss": 0.2577,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7798165137614679,
+      "grad_norm": 0.8620551824569702,
+      "learning_rate": 2.4378226891347056e-06,
+      "loss": 0.2155,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7800863464651916,
+      "grad_norm": 0.9157201647758484,
+      "learning_rate": 2.4321060266420802e-06,
+      "loss": 0.2446,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7803561791689153,
+      "grad_norm": 0.8729531168937683,
+      "learning_rate": 2.426395146688999e-06,
+      "loss": 0.2534,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7806260118726389,
+      "grad_norm": 2.109168291091919,
+      "learning_rate": 2.42069005363907e-06,
+      "loss": 0.2351,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7808958445763626,
+      "grad_norm": 1.0341562032699585,
+      "learning_rate": 2.414990751851495e-06,
+      "loss": 0.2958,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7811656772800863,
+      "grad_norm": 0.8622056245803833,
+      "learning_rate": 2.409297245681046e-06,
+      "loss": 0.2179,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7814355099838101,
+      "grad_norm": 1.3834373950958252,
+      "learning_rate": 2.403609539478056e-06,
+      "loss": 0.2556,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7817053426875338,
+      "grad_norm": 1.176883339881897,
+      "learning_rate": 2.397927637588442e-06,
+      "loss": 0.2752,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7819751753912574,
+      "grad_norm": 0.8752351999282837,
+      "learning_rate": 2.392251544353681e-06,
+      "loss": 0.2521,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7822450080949811,
+      "grad_norm": 0.9197589159011841,
+      "learning_rate": 2.3865812641108e-06,
+      "loss": 0.2485,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7825148407987048,
+      "grad_norm": 1.0460208654403687,
+      "learning_rate": 2.380916801192402e-06,
+      "loss": 0.2369,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7827846735024285,
+      "grad_norm": 1.0845636129379272,
+      "learning_rate": 2.3752581599266387e-06,
+      "loss": 0.2793,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7830545062061521,
+      "grad_norm": 1.3572076559066772,
+      "learning_rate": 2.3696053446372026e-06,
+      "loss": 0.2514,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7833243389098759,
+      "grad_norm": 0.8141172528266907,
+      "learning_rate": 2.3639583596433524e-06,
+      "loss": 0.2572,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7835941716135996,
+      "grad_norm": 1.4322822093963623,
+      "learning_rate": 2.358317209259883e-06,
+      "loss": 0.2759,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7838640043173233,
+      "grad_norm": 0.8642662763595581,
+      "learning_rate": 2.3526818977971235e-06,
+      "loss": 0.2326,
+      "step": 2905
+    },
+    {
+      "epoch": 0.784133837021047,
+      "grad_norm": 1.7199457883834839,
+      "learning_rate": 2.3470524295609586e-06,
+      "loss": 0.2332,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7844036697247706,
+      "grad_norm": 0.8535600304603577,
+      "learning_rate": 2.341428808852799e-06,
+      "loss": 0.2959,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7846735024284943,
+      "grad_norm": 1.7828149795532227,
+      "learning_rate": 2.3358110399695788e-06,
+      "loss": 0.2574,
+      "step": 2908
+    },
+    {
+      "epoch": 0.784943335132218,
+      "grad_norm": 1.7031919956207275,
+      "learning_rate": 2.3301991272037784e-06,
+      "loss": 0.254,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7852131678359417,
+      "grad_norm": 1.5151183605194092,
+      "learning_rate": 2.324593074843395e-06,
+      "loss": 0.2582,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7854830005396654,
+      "grad_norm": 1.2216993570327759,
+      "learning_rate": 2.3189928871719404e-06,
+      "loss": 0.2728,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7857528332433891,
+      "grad_norm": 1.265013575553894,
+      "learning_rate": 2.31339856846846e-06,
+      "loss": 0.243,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7860226659471128,
+      "grad_norm": 1.7832986116409302,
+      "learning_rate": 2.307810123007508e-06,
+      "loss": 0.2388,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7862924986508365,
+      "grad_norm": 1.2688908576965332,
+      "learning_rate": 2.302227555059141e-06,
+      "loss": 0.2533,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7865623313545602,
+      "grad_norm": 1.2569414377212524,
+      "learning_rate": 2.2966508688889433e-06,
+      "loss": 0.2346,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7868321640582838,
+      "grad_norm": 1.0619051456451416,
+      "learning_rate": 2.291080068757996e-06,
+      "loss": 0.2478,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7871019967620075,
+      "grad_norm": 1.1263363361358643,
+      "learning_rate": 2.285515158922873e-06,
+      "loss": 0.2662,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7873718294657313,
+      "grad_norm": 1.0899661779403687,
+      "learning_rate": 2.2799561436356665e-06,
+      "loss": 0.2155,
+      "step": 2918
+    },
+    {
+      "epoch": 0.787641662169455,
+      "grad_norm": 1.8365812301635742,
+      "learning_rate": 2.2744030271439556e-06,
+      "loss": 0.2375,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7879114948731786,
+      "grad_norm": 0.9622383117675781,
+      "learning_rate": 2.2688558136908025e-06,
+      "loss": 0.2048,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7881813275769023,
+      "grad_norm": 2.311168670654297,
+      "learning_rate": 2.263314507514777e-06,
+      "loss": 0.2165,
+      "step": 2921
+    },
+    {
+      "epoch": 0.788451160280626,
+      "grad_norm": 1.422361135482788,
+      "learning_rate": 2.257779112849926e-06,
+      "loss": 0.242,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7887209929843497,
+      "grad_norm": 1.0336179733276367,
+      "learning_rate": 2.252249633925773e-06,
+      "loss": 0.2362,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 1.717832088470459,
+      "learning_rate": 2.2467260749673357e-06,
+      "loss": 0.2388,
+      "step": 2924
+    },
+    {
+      "epoch": 0.789260658391797,
+      "grad_norm": 1.3378719091415405,
+      "learning_rate": 2.2412084401950983e-06,
+      "loss": 0.2439,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7895304910955208,
+      "grad_norm": 2.1422884464263916,
+      "learning_rate": 2.2356967338250223e-06,
+      "loss": 0.2405,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7898003237992445,
+      "grad_norm": 1.4154285192489624,
+      "learning_rate": 2.230190960068537e-06,
+      "loss": 0.2656,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7900701565029682,
+      "grad_norm": 1.058073878288269,
+      "learning_rate": 2.2246911231325396e-06,
+      "loss": 0.2609,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7903399892066919,
+      "grad_norm": 1.317310094833374,
+      "learning_rate": 2.219197227219393e-06,
+      "loss": 0.2389,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7906098219104155,
+      "grad_norm": 0.937031090259552,
+      "learning_rate": 2.2137092765269174e-06,
+      "loss": 0.2349,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7908796546141392,
+      "grad_norm": 0.9941629767417908,
+      "learning_rate": 2.2082272752483935e-06,
+      "loss": 0.2813,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7911494873178629,
+      "grad_norm": 1.5534207820892334,
+      "learning_rate": 2.202751227572556e-06,
+      "loss": 0.2264,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7914193200215867,
+      "grad_norm": 1.5839259624481201,
+      "learning_rate": 2.197281137683581e-06,
+      "loss": 0.2716,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7916891527253103,
+      "grad_norm": 2.135491371154785,
+      "learning_rate": 2.191817009761109e-06,
+      "loss": 0.2599,
+      "step": 2934
+    },
+    {
+      "epoch": 0.791958985429034,
+      "grad_norm": 1.1375353336334229,
+      "learning_rate": 2.186358847980217e-06,
+      "loss": 0.2393,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7922288181327577,
+      "grad_norm": 0.9597465395927429,
+      "learning_rate": 2.1809066565114132e-06,
+      "loss": 0.2726,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7924986508364814,
+      "grad_norm": 1.1295933723449707,
+      "learning_rate": 2.1754604395206635e-06,
+      "loss": 0.2404,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7927684835402051,
+      "grad_norm": 1.0045075416564941,
+      "learning_rate": 2.1700202011693573e-06,
+      "loss": 0.2282,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7930383162439287,
+      "grad_norm": 1.7756739854812622,
+      "learning_rate": 2.164585945614309e-06,
+      "loss": 0.2467,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7933081489476524,
+      "grad_norm": 1.4644638299942017,
+      "learning_rate": 2.1591576770077784e-06,
+      "loss": 0.2788,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7935779816513762,
+      "grad_norm": 1.3074876070022583,
+      "learning_rate": 2.153735399497443e-06,
+      "loss": 0.2489,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7938478143550999,
+      "grad_norm": 1.802755355834961,
+      "learning_rate": 2.148319117226394e-06,
+      "loss": 0.2784,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7941176470588235,
+      "grad_norm": 1.0552555322647095,
+      "learning_rate": 2.142908834333155e-06,
+      "loss": 0.2885,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7943874797625472,
+      "grad_norm": 1.1824276447296143,
+      "learning_rate": 2.1375045549516636e-06,
+      "loss": 0.2631,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7946573124662709,
+      "grad_norm": 1.2791507244110107,
+      "learning_rate": 2.132106283211256e-06,
+      "loss": 0.3076,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7949271451699946,
+      "grad_norm": 2.0106682777404785,
+      "learning_rate": 2.126714023236698e-06,
+      "loss": 0.2728,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7951969778737183,
+      "grad_norm": 1.9162076711654663,
+      "learning_rate": 2.121327779148151e-06,
+      "loss": 0.2794,
+      "step": 2947
+    },
+    {
+      "epoch": 0.795466810577442,
+      "grad_norm": 1.0659760236740112,
+      "learning_rate": 2.1159475550611763e-06,
+      "loss": 0.2719,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7957366432811657,
+      "grad_norm": 0.9203602075576782,
+      "learning_rate": 2.1105733550867457e-06,
+      "loss": 0.2384,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7960064759848894,
+      "grad_norm": 0.7790428400039673,
+      "learning_rate": 2.105205183331224e-06,
+      "loss": 0.2551,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7962763086886131,
+      "grad_norm": 1.5077874660491943,
+      "learning_rate": 2.0998430438963592e-06,
+      "loss": 0.2209,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7965461413923367,
+      "grad_norm": 1.486519455909729,
+      "learning_rate": 2.094486940879309e-06,
+      "loss": 0.2323,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7968159740960604,
+      "grad_norm": 1.261599063873291,
+      "learning_rate": 2.0891368783726083e-06,
+      "loss": 0.2645,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7970858067997841,
+      "grad_norm": 0.8529981970787048,
+      "learning_rate": 2.083792860464171e-06,
+      "loss": 0.2647,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7973556395035079,
+      "grad_norm": 0.9199646711349487,
+      "learning_rate": 2.078454891237306e-06,
+      "loss": 0.2408,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7976254722072316,
+      "grad_norm": 1.2477949857711792,
+      "learning_rate": 2.0731229747706926e-06,
+      "loss": 0.2356,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7978953049109552,
+      "grad_norm": 1.4132107496261597,
+      "learning_rate": 2.067797115138379e-06,
+      "loss": 0.2535,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7981651376146789,
+      "grad_norm": 1.00215482711792,
+      "learning_rate": 2.062477316409799e-06,
+      "loss": 0.2425,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7984349703184026,
+      "grad_norm": 1.4233955144882202,
+      "learning_rate": 2.057163582649748e-06,
+      "loss": 0.2349,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7987048030221263,
+      "grad_norm": 1.5263769626617432,
+      "learning_rate": 2.0518559179183873e-06,
+      "loss": 0.2315,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7989746357258499,
+      "grad_norm": 1.217991828918457,
+      "learning_rate": 2.0465543262712417e-06,
+      "loss": 0.2764,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7992444684295736,
+      "grad_norm": 1.2172868251800537,
+      "learning_rate": 2.041258811759195e-06,
+      "loss": 0.2833,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7995143011332974,
+      "grad_norm": 0.9673815965652466,
+      "learning_rate": 2.0359693784284883e-06,
+      "loss": 0.2713,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7997841338370211,
+      "grad_norm": 0.8467391133308411,
+      "learning_rate": 2.0306860303207166e-06,
+      "loss": 0.2625,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8000539665407448,
+      "grad_norm": 1.120172142982483,
+      "learning_rate": 2.0254087714728234e-06,
+      "loss": 0.2467,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8003237992444684,
+      "grad_norm": 1.4225075244903564,
+      "learning_rate": 2.020137605917102e-06,
+      "loss": 0.289,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8005936319481921,
+      "grad_norm": 1.3648520708084106,
+      "learning_rate": 2.0148725376811875e-06,
+      "loss": 0.2767,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8008634646519158,
+      "grad_norm": 1.3091508150100708,
+      "learning_rate": 2.009613570788057e-06,
+      "loss": 0.306,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8011332973556395,
+      "grad_norm": 0.8872712850570679,
+      "learning_rate": 2.004360709256027e-06,
+      "loss": 0.258,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8014031300593631,
+      "grad_norm": 0.9025729894638062,
+      "learning_rate": 1.999113957098746e-06,
+      "loss": 0.246,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8016729627630869,
+      "grad_norm": 0.9187671542167664,
+      "learning_rate": 1.993873318325199e-06,
+      "loss": 0.2459,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8019427954668106,
+      "grad_norm": 0.8485140800476074,
+      "learning_rate": 1.9886387969396937e-06,
+      "loss": 0.2724,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8022126281705343,
+      "grad_norm": 1.1865864992141724,
+      "learning_rate": 1.9834103969418684e-06,
+      "loss": 0.2225,
+      "step": 2973
+    },
+    {
+      "epoch": 0.802482460874258,
+      "grad_norm": 1.2929097414016724,
+      "learning_rate": 1.978188122326683e-06,
+      "loss": 0.2719,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8027522935779816,
+      "grad_norm": 1.0871570110321045,
+      "learning_rate": 1.9729719770844156e-06,
+      "loss": 0.2366,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8030221262817053,
+      "grad_norm": 0.9727485179901123,
+      "learning_rate": 1.967761965200663e-06,
+      "loss": 0.2627,
+      "step": 2976
+    },
+    {
+      "epoch": 0.803291958985429,
+      "grad_norm": 1.0222405195236206,
+      "learning_rate": 1.9625580906563348e-06,
+      "loss": 0.2668,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8035617916891528,
+      "grad_norm": 0.8128262758255005,
+      "learning_rate": 1.957360357427649e-06,
+      "loss": 0.2736,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8038316243928764,
+      "grad_norm": 0.9023306965827942,
+      "learning_rate": 1.9521687694861336e-06,
+      "loss": 0.2618,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8041014570966001,
+      "grad_norm": 1.067893147468567,
+      "learning_rate": 1.946983330798621e-06,
+      "loss": 0.2825,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8043712898003238,
+      "grad_norm": 1.102923035621643,
+      "learning_rate": 1.9418040453272448e-06,
+      "loss": 0.2563,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8046411225040475,
+      "grad_norm": 1.569082498550415,
+      "learning_rate": 1.936630917029436e-06,
+      "loss": 0.2155,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8049109552077712,
+      "grad_norm": 1.1191504001617432,
+      "learning_rate": 1.931463949857921e-06,
+      "loss": 0.2558,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8051807879114948,
+      "grad_norm": 1.409506916999817,
+      "learning_rate": 1.926303147760721e-06,
+      "loss": 0.2528,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8054506206152185,
+      "grad_norm": 1.3394663333892822,
+      "learning_rate": 1.921148514681144e-06,
+      "loss": 0.239,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8057204533189423,
+      "grad_norm": 1.8142637014389038,
+      "learning_rate": 1.916000054557783e-06,
+      "loss": 0.2836,
+      "step": 2986
+    },
+    {
+      "epoch": 0.805990286022666,
+      "grad_norm": 1.0419009923934937,
+      "learning_rate": 1.910857771324519e-06,
+      "loss": 0.2644,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8062601187263896,
+      "grad_norm": 1.0996564626693726,
+      "learning_rate": 1.9057216689105097e-06,
+      "loss": 0.2467,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8065299514301133,
+      "grad_norm": 0.9447081089019775,
+      "learning_rate": 1.9005917512401905e-06,
+      "loss": 0.274,
+      "step": 2989
+    },
+    {
+      "epoch": 0.806799784133837,
+      "grad_norm": 1.258445143699646,
+      "learning_rate": 1.8954680222332722e-06,
+      "loss": 0.2061,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8070696168375607,
+      "grad_norm": 1.4685399532318115,
+      "learning_rate": 1.8903504858047361e-06,
+      "loss": 0.2749,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 1.5901374816894531,
+      "learning_rate": 1.8852391458648323e-06,
+      "loss": 0.2451,
+      "step": 2992
+    },
+    {
+      "epoch": 0.807609282245008,
+      "grad_norm": 1.2240535020828247,
+      "learning_rate": 1.8801340063190766e-06,
+      "loss": 0.2539,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8078791149487318,
+      "grad_norm": 0.9628891944885254,
+      "learning_rate": 1.8750350710682464e-06,
+      "loss": 0.2209,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8081489476524555,
+      "grad_norm": 0.8044888973236084,
+      "learning_rate": 1.8699423440083786e-06,
+      "loss": 0.2473,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8084187803561792,
+      "grad_norm": 0.8723772168159485,
+      "learning_rate": 1.864855829030765e-06,
+      "loss": 0.2343,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8086886130599028,
+      "grad_norm": 1.0535954236984253,
+      "learning_rate": 1.8597755300219544e-06,
+      "loss": 0.2364,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8089584457636265,
+      "grad_norm": 2.2117059230804443,
+      "learning_rate": 1.854701450863744e-06,
+      "loss": 0.2865,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8092282784673502,
+      "grad_norm": 1.0591880083084106,
+      "learning_rate": 1.8496335954331767e-06,
+      "loss": 0.254,
+      "step": 2999
+    },
+    {
+      "epoch": 0.809498111171074,
+      "grad_norm": 0.9559292197227478,
+      "learning_rate": 1.8445719676025442e-06,
+      "loss": 0.2336,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8097679438747977,
+      "grad_norm": 1.1796631813049316,
+      "learning_rate": 1.839516571239376e-06,
+      "loss": 0.2384,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8100377765785213,
+      "grad_norm": 1.778013825416565,
+      "learning_rate": 1.834467410206441e-06,
+      "loss": 0.2613,
+      "step": 3002
+    },
+    {
+      "epoch": 0.810307609282245,
+      "grad_norm": 0.8403366804122925,
+      "learning_rate": 1.829424488361744e-06,
+      "loss": 0.2701,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8105774419859687,
+      "grad_norm": 1.1803412437438965,
+      "learning_rate": 1.8243878095585244e-06,
+      "loss": 0.2547,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8108472746896924,
+      "grad_norm": 1.2165312767028809,
+      "learning_rate": 1.819357377645249e-06,
+      "loss": 0.2583,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8111171073934161,
+      "grad_norm": 0.8605884909629822,
+      "learning_rate": 1.814333196465612e-06,
+      "loss": 0.2881,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8113869400971397,
+      "grad_norm": 0.9720732569694519,
+      "learning_rate": 1.8093152698585325e-06,
+      "loss": 0.2285,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8116567728008635,
+      "grad_norm": 1.2440024614334106,
+      "learning_rate": 1.8043036016581483e-06,
+      "loss": 0.2503,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8119266055045872,
+      "grad_norm": 1.2146265506744385,
+      "learning_rate": 1.7992981956938172e-06,
+      "loss": 0.2852,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8121964382083109,
+      "grad_norm": 1.1288738250732422,
+      "learning_rate": 1.7942990557901119e-06,
+      "loss": 0.2304,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8124662709120345,
+      "grad_norm": 0.9449646472930908,
+      "learning_rate": 1.789306185766816e-06,
+      "loss": 0.2386,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8127361036157582,
+      "grad_norm": 0.9483136534690857,
+      "learning_rate": 1.7843195894389242e-06,
+      "loss": 0.2889,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8130059363194819,
+      "grad_norm": 0.9870485067367554,
+      "learning_rate": 1.7793392706166368e-06,
+      "loss": 0.2107,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8132757690232056,
+      "grad_norm": 1.2763394117355347,
+      "learning_rate": 1.7743652331053562e-06,
+      "loss": 0.2613,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8135456017269294,
+      "grad_norm": 1.398007869720459,
+      "learning_rate": 1.769397480705688e-06,
+      "loss": 0.2724,
+      "step": 3015
+    },
+    {
+      "epoch": 0.813815434430653,
+      "grad_norm": 0.906068742275238,
+      "learning_rate": 1.7644360172134323e-06,
+      "loss": 0.253,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8140852671343767,
+      "grad_norm": 1.4642654657363892,
+      "learning_rate": 1.7594808464195867e-06,
+      "loss": 0.2266,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8143550998381004,
+      "grad_norm": 1.2873657941818237,
+      "learning_rate": 1.7545319721103381e-06,
+      "loss": 0.2403,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8146249325418241,
+      "grad_norm": 0.832125723361969,
+      "learning_rate": 1.7495893980670653e-06,
+      "loss": 0.2507,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8148947652455477,
+      "grad_norm": 1.2334462404251099,
+      "learning_rate": 1.7446531280663303e-06,
+      "loss": 0.231,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8151645979492714,
+      "grad_norm": 1.611694574356079,
+      "learning_rate": 1.7397231658798808e-06,
+      "loss": 0.2548,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8154344306529951,
+      "grad_norm": 1.1768789291381836,
+      "learning_rate": 1.734799515274641e-06,
+      "loss": 0.2625,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8157042633567189,
+      "grad_norm": 1.0607473850250244,
+      "learning_rate": 1.7298821800127175e-06,
+      "loss": 0.2751,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8159740960604426,
+      "grad_norm": 0.9140863418579102,
+      "learning_rate": 1.7249711638513845e-06,
+      "loss": 0.2498,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8162439287641662,
+      "grad_norm": 1.2152066230773926,
+      "learning_rate": 1.7200664705431003e-06,
+      "loss": 0.2878,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8165137614678899,
+      "grad_norm": 1.3145592212677002,
+      "learning_rate": 1.7151681038354772e-06,
+      "loss": 0.28,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8167835941716136,
+      "grad_norm": 1.0914944410324097,
+      "learning_rate": 1.7102760674712994e-06,
+      "loss": 0.3044,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8170534268753373,
+      "grad_norm": 1.4639004468917847,
+      "learning_rate": 1.7053903651885217e-06,
+      "loss": 0.2629,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8173232595790609,
+      "grad_norm": 1.9587889909744263,
+      "learning_rate": 1.7005110007202452e-06,
+      "loss": 0.2392,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8175930922827847,
+      "grad_norm": 1.5423575639724731,
+      "learning_rate": 1.6956379777947352e-06,
+      "loss": 0.2363,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8178629249865084,
+      "grad_norm": 1.8476983308792114,
+      "learning_rate": 1.690771300135421e-06,
+      "loss": 0.2074,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8181327576902321,
+      "grad_norm": 1.485688328742981,
+      "learning_rate": 1.6859109714608635e-06,
+      "loss": 0.2586,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8184025903939558,
+      "grad_norm": 1.2554693222045898,
+      "learning_rate": 1.6810569954847867e-06,
+      "loss": 0.2418,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8186724230976794,
+      "grad_norm": 1.0497089624404907,
+      "learning_rate": 1.6762093759160614e-06,
+      "loss": 0.2674,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8189422558014031,
+      "grad_norm": 1.8688005208969116,
+      "learning_rate": 1.6713681164586914e-06,
+      "loss": 0.239,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8192120885051268,
+      "grad_norm": 1.2148860692977905,
+      "learning_rate": 1.6665332208118267e-06,
+      "loss": 0.2641,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8194819212088506,
+      "grad_norm": 1.53981351852417,
+      "learning_rate": 1.661704692669761e-06,
+      "loss": 0.252,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8197517539125742,
+      "grad_norm": 1.3042700290679932,
+      "learning_rate": 1.656882535721911e-06,
+      "loss": 0.2483,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8200215866162979,
+      "grad_norm": 1.3406310081481934,
+      "learning_rate": 1.6520667536528278e-06,
+      "loss": 0.2629,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8202914193200216,
+      "grad_norm": 0.9211012721061707,
+      "learning_rate": 1.647257350142204e-06,
+      "loss": 0.2688,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8205612520237453,
+      "grad_norm": 1.368143081665039,
+      "learning_rate": 1.6424543288648398e-06,
+      "loss": 0.2284,
+      "step": 3041
+    },
+    {
+      "epoch": 0.820831084727469,
+      "grad_norm": 1.584941029548645,
+      "learning_rate": 1.6376576934906696e-06,
+      "loss": 0.2555,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8211009174311926,
+      "grad_norm": 1.880881428718567,
+      "learning_rate": 1.6328674476847516e-06,
+      "loss": 0.2702,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8213707501349163,
+      "grad_norm": 1.130049228668213,
+      "learning_rate": 1.6280835951072515e-06,
+      "loss": 0.2406,
+      "step": 3044
+    },
+    {
+      "epoch": 0.82164058283864,
+      "grad_norm": 1.7763452529907227,
+      "learning_rate": 1.623306139413454e-06,
+      "loss": 0.2735,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8219104155423638,
+      "grad_norm": 1.0071371793746948,
+      "learning_rate": 1.618535084253765e-06,
+      "loss": 0.2575,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8221802482460874,
+      "grad_norm": 0.9252629280090332,
+      "learning_rate": 1.6137704332736847e-06,
+      "loss": 0.2473,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8224500809498111,
+      "grad_norm": 2.556549072265625,
+      "learning_rate": 1.6090121901138278e-06,
+      "loss": 0.3167,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8227199136535348,
+      "grad_norm": 1.0139598846435547,
+      "learning_rate": 1.604260358409918e-06,
+      "loss": 0.2233,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8229897463572585,
+      "grad_norm": 2.3814890384674072,
+      "learning_rate": 1.5995149417927703e-06,
+      "loss": 0.2905,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8232595790609822,
+      "grad_norm": 1.496466875076294,
+      "learning_rate": 1.5947759438883004e-06,
+      "loss": 0.2681,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 1.5073753595352173,
+      "learning_rate": 1.5900433683175277e-06,
+      "loss": 0.2526,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8237992444684296,
+      "grad_norm": 1.2024897336959839,
+      "learning_rate": 1.5853172186965515e-06,
+      "loss": 0.2593,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8240690771721533,
+      "grad_norm": 0.7716065049171448,
+      "learning_rate": 1.5805974986365692e-06,
+      "loss": 0.2783,
+      "step": 3054
+    },
+    {
+      "epoch": 0.824338909875877,
+      "grad_norm": 1.3857553005218506,
+      "learning_rate": 1.5758842117438688e-06,
+      "loss": 0.2743,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8246087425796006,
+      "grad_norm": 1.987254023551941,
+      "learning_rate": 1.5711773616198124e-06,
+      "loss": 0.2649,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8248785752833243,
+      "grad_norm": 0.9485495686531067,
+      "learning_rate": 1.566476951860848e-06,
+      "loss": 0.2487,
+      "step": 3057
+    },
+    {
+      "epoch": 0.825148407987048,
+      "grad_norm": 1.4368356466293335,
+      "learning_rate": 1.5617829860585087e-06,
+      "loss": 0.2948,
+      "step": 3058
+    },
+    {
+      "epoch": 0.8254182406907717,
+      "grad_norm": 1.0704184770584106,
+      "learning_rate": 1.5570954677994e-06,
+      "loss": 0.2375,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.8070966601371765,
+      "learning_rate": 1.552414400665192e-06,
+      "loss": 0.2257,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8259579060982191,
+      "grad_norm": 1.0417391061782837,
+      "learning_rate": 1.5477397882326418e-06,
+      "loss": 0.237,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8262277388019428,
+      "grad_norm": 1.0329066514968872,
+      "learning_rate": 1.543071634073564e-06,
+      "loss": 0.2865,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8264975715056665,
+      "grad_norm": 1.6353132724761963,
+      "learning_rate": 1.5384099417548359e-06,
+      "loss": 0.2704,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8267674042093902,
+      "grad_norm": 0.9926678538322449,
+      "learning_rate": 1.533754714838408e-06,
+      "loss": 0.2964,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8270372369131138,
+      "grad_norm": 1.048696756362915,
+      "learning_rate": 1.529105956881286e-06,
+      "loss": 0.2254,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8273070696168375,
+      "grad_norm": 1.0285667181015015,
+      "learning_rate": 1.5244636714355266e-06,
+      "loss": 0.2781,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8275769023205612,
+      "grad_norm": 1.5292519330978394,
+      "learning_rate": 1.5198278620482454e-06,
+      "loss": 0.21,
+      "step": 3067
+    },
+    {
+      "epoch": 0.827846735024285,
+      "grad_norm": 1.3913131952285767,
+      "learning_rate": 1.515198532261617e-06,
+      "loss": 0.2495,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8281165677280087,
+      "grad_norm": 1.2348507642745972,
+      "learning_rate": 1.5105756856128529e-06,
+      "loss": 0.2643,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8283864004317323,
+      "grad_norm": 1.3613711595535278,
+      "learning_rate": 1.5059593256342142e-06,
+      "loss": 0.2138,
+      "step": 3070
+    },
+    {
+      "epoch": 0.828656233135456,
+      "grad_norm": 1.1131243705749512,
+      "learning_rate": 1.5013494558530162e-06,
+      "loss": 0.281,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8289260658391797,
+      "grad_norm": 0.8918585181236267,
+      "learning_rate": 1.4967460797916001e-06,
+      "loss": 0.2473,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8291958985429034,
+      "grad_norm": 0.9085717797279358,
+      "learning_rate": 1.492149200967351e-06,
+      "loss": 0.2152,
+      "step": 3073
+    },
+    {
+      "epoch": 0.829465731246627,
+      "grad_norm": 1.3287911415100098,
+      "learning_rate": 1.4875588228926964e-06,
+      "loss": 0.2438,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8297355639503508,
+      "grad_norm": 1.5275652408599854,
+      "learning_rate": 1.4829749490750856e-06,
+      "loss": 0.2519,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8300053966540745,
+      "grad_norm": 1.0149762630462646,
+      "learning_rate": 1.4783975830170028e-06,
+      "loss": 0.2356,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8302752293577982,
+      "grad_norm": 0.817284882068634,
+      "learning_rate": 1.4738267282159658e-06,
+      "loss": 0.2323,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8305450620615219,
+      "grad_norm": 1.323622703552246,
+      "learning_rate": 1.4692623881645052e-06,
+      "loss": 0.2485,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8308148947652455,
+      "grad_norm": 1.072948932647705,
+      "learning_rate": 1.4647045663501801e-06,
+      "loss": 0.274,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8310847274689692,
+      "grad_norm": 1.2872915267944336,
+      "learning_rate": 1.4601532662555762e-06,
+      "loss": 0.2446,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8313545601726929,
+      "grad_norm": 0.8632007241249084,
+      "learning_rate": 1.4556084913582813e-06,
+      "loss": 0.2601,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8316243928764167,
+      "grad_norm": 0.8919570446014404,
+      "learning_rate": 1.4510702451309055e-06,
+      "loss": 0.2536,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8318942255801404,
+      "grad_norm": 1.6977310180664062,
+      "learning_rate": 1.4465385310410751e-06,
+      "loss": 0.2539,
+      "step": 3083
+    },
+    {
+      "epoch": 0.832164058283864,
+      "grad_norm": 1.6512761116027832,
+      "learning_rate": 1.4420133525514158e-06,
+      "loss": 0.2924,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8324338909875877,
+      "grad_norm": 1.299375295639038,
+      "learning_rate": 1.4374947131195617e-06,
+      "loss": 0.2282,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8327037236913114,
+      "grad_norm": 2.332383871078491,
+      "learning_rate": 1.4329826161981607e-06,
+      "loss": 0.2789,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8329735563950351,
+      "grad_norm": 1.0186880826950073,
+      "learning_rate": 1.4284770652348457e-06,
+      "loss": 0.2542,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8332433890987587,
+      "grad_norm": 1.6572496891021729,
+      "learning_rate": 1.4239780636722555e-06,
+      "loss": 0.2254,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8335132218024824,
+      "grad_norm": 0.805113673210144,
+      "learning_rate": 1.419485614948032e-06,
+      "loss": 0.2371,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8337830545062062,
+      "grad_norm": 1.1359212398529053,
+      "learning_rate": 1.4149997224947954e-06,
+      "loss": 0.2699,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8340528872099299,
+      "grad_norm": 2.010716199874878,
+      "learning_rate": 1.4105203897401654e-06,
+      "loss": 0.2535,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8343227199136536,
+      "grad_norm": 0.928462028503418,
+      "learning_rate": 1.4060476201067508e-06,
+      "loss": 0.2366,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8345925526173772,
+      "grad_norm": 1.3978617191314697,
+      "learning_rate": 1.4015814170121434e-06,
+      "loss": 0.2292,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8348623853211009,
+      "grad_norm": 0.9662782549858093,
+      "learning_rate": 1.39712178386891e-06,
+      "loss": 0.2409,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8351322180248246,
+      "grad_norm": 1.137027621269226,
+      "learning_rate": 1.3926687240846115e-06,
+      "loss": 0.2503,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8354020507285483,
+      "grad_norm": 1.4905757904052734,
+      "learning_rate": 1.3882222410617786e-06,
+      "loss": 0.2429,
+      "step": 3096
+    },
+    {
+      "epoch": 0.835671883432272,
+      "grad_norm": 1.6609336137771606,
+      "learning_rate": 1.3837823381979098e-06,
+      "loss": 0.2815,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8359417161359957,
+      "grad_norm": 1.2367022037506104,
+      "learning_rate": 1.3793490188854908e-06,
+      "loss": 0.2748,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8362115488397194,
+      "grad_norm": 1.6461222171783447,
+      "learning_rate": 1.3749222865119693e-06,
+      "loss": 0.2691,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8364813815434431,
+      "grad_norm": 1.5357296466827393,
+      "learning_rate": 1.3705021444597521e-06,
+      "loss": 0.2321,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8367512142471668,
+      "grad_norm": 1.427616834640503,
+      "learning_rate": 1.3660885961062254e-06,
+      "loss": 0.2374,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8370210469508904,
+      "grad_norm": 1.0070767402648926,
+      "learning_rate": 1.3616816448237302e-06,
+      "loss": 0.239,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8372908796546141,
+      "grad_norm": 0.9238376617431641,
+      "learning_rate": 1.3572812939795598e-06,
+      "loss": 0.2833,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8375607123583378,
+      "grad_norm": 1.2639409303665161,
+      "learning_rate": 1.3528875469359781e-06,
+      "loss": 0.2673,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8378305450620616,
+      "grad_norm": 2.0981032848358154,
+      "learning_rate": 1.3485004070501962e-06,
+      "loss": 0.2972,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8381003777657852,
+      "grad_norm": 1.0110149383544922,
+      "learning_rate": 1.344119877674368e-06,
+      "loss": 0.2446,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8383702104695089,
+      "grad_norm": 0.8722718358039856,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.2664,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8386400431732326,
+      "grad_norm": 0.9161373972892761,
+      "learning_rate": 1.33537866383599e-06,
+      "loss": 0.2374,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8389098758769563,
+      "grad_norm": 1.1237530708312988,
+      "learning_rate": 1.331017986052492e-06,
+      "loss": 0.2532,
+      "step": 3109
+    },
+    {
+      "epoch": 0.83917970858068,
+      "grad_norm": 0.9259228706359863,
+      "learning_rate": 1.326663932137069e-06,
+      "loss": 0.279,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8394495412844036,
+      "grad_norm": 0.8383319973945618,
+      "learning_rate": 1.3223165054166043e-06,
+      "loss": 0.2212,
+      "step": 3111
+    },
+    {
+      "epoch": 0.8397193739881273,
+      "grad_norm": 1.6744608879089355,
+      "learning_rate": 1.3179757092129087e-06,
+      "loss": 0.2116,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8399892066918511,
+      "grad_norm": 0.9502532482147217,
+      "learning_rate": 1.31364154684274e-06,
+      "loss": 0.2456,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8402590393955748,
+      "grad_norm": 1.7954977750778198,
+      "learning_rate": 1.3093140216177814e-06,
+      "loss": 0.2659,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8405288720992984,
+      "grad_norm": 1.0558723211288452,
+      "learning_rate": 1.3049931368446378e-06,
+      "loss": 0.2651,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8407987048030221,
+      "grad_norm": 0.8701751828193665,
+      "learning_rate": 1.300678895824854e-06,
+      "loss": 0.2393,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8410685375067458,
+      "grad_norm": 1.1922876834869385,
+      "learning_rate": 1.2963713018548906e-06,
+      "loss": 0.2385,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8413383702104695,
+      "grad_norm": 1.3241044282913208,
+      "learning_rate": 1.292070358226124e-06,
+      "loss": 0.2508,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8416082029141932,
+      "grad_norm": 0.8277075290679932,
+      "learning_rate": 1.2877760682248608e-06,
+      "loss": 0.2272,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8418780356179169,
+      "grad_norm": 1.3088337182998657,
+      "learning_rate": 1.2834884351323196e-06,
+      "loss": 0.2338,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8421478683216406,
+      "grad_norm": 1.0497779846191406,
+      "learning_rate": 1.279207462224622e-06,
+      "loss": 0.275,
+      "step": 3121
+    },
+    {
+      "epoch": 0.8424177010253643,
+      "grad_norm": 1.2748932838439941,
+      "learning_rate": 1.274933152772817e-06,
+      "loss": 0.2442,
+      "step": 3122
+    },
+    {
+      "epoch": 0.842687533729088,
+      "grad_norm": 0.813736617565155,
+      "learning_rate": 1.2706655100428556e-06,
+      "loss": 0.2507,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8429573664328116,
+      "grad_norm": 1.0318999290466309,
+      "learning_rate": 1.2664045372955858e-06,
+      "loss": 0.2612,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8432271991365353,
+      "grad_norm": 0.9960605502128601,
+      "learning_rate": 1.2621502377867746e-06,
+      "loss": 0.2584,
+      "step": 3125
+    },
+    {
+      "epoch": 0.843497031840259,
+      "grad_norm": 0.8731059432029724,
+      "learning_rate": 1.2579026147670792e-06,
+      "loss": 0.2436,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8437668645439828,
+      "grad_norm": 1.8202052116394043,
+      "learning_rate": 1.2536616714820614e-06,
+      "loss": 0.2624,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 1.3580933809280396,
+      "learning_rate": 1.2494274111721737e-06,
+      "loss": 0.2439,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8443065299514301,
+      "grad_norm": 1.9957126379013062,
+      "learning_rate": 1.245199837072767e-06,
+      "loss": 0.2718,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8445763626551538,
+      "grad_norm": 1.1930474042892456,
+      "learning_rate": 1.2409789524140813e-06,
+      "loss": 0.264,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8448461953588775,
+      "grad_norm": 0.9323124289512634,
+      "learning_rate": 1.2367647604212441e-06,
+      "loss": 0.2291,
+      "step": 3131
+    },
+    {
+      "epoch": 0.8451160280626012,
+      "grad_norm": 1.4825975894927979,
+      "learning_rate": 1.2325572643142702e-06,
+      "loss": 0.2527,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8453858607663248,
+      "grad_norm": 1.1822201013565063,
+      "learning_rate": 1.2283564673080605e-06,
+      "loss": 0.2297,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8456556934700485,
+      "grad_norm": 1.5013542175292969,
+      "learning_rate": 1.224162372612393e-06,
+      "loss": 0.2874,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8459255261737723,
+      "grad_norm": 1.0640422105789185,
+      "learning_rate": 1.2199749834319264e-06,
+      "loss": 0.2451,
+      "step": 3135
+    },
+    {
+      "epoch": 0.846195358877496,
+      "grad_norm": 1.020533561706543,
+      "learning_rate": 1.2157943029661977e-06,
+      "loss": 0.2652,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8464651915812197,
+      "grad_norm": 1.1189326047897339,
+      "learning_rate": 1.2116203344096144e-06,
+      "loss": 0.2313,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8467350242849433,
+      "grad_norm": 1.023314356803894,
+      "learning_rate": 1.2074530809514585e-06,
+      "loss": 0.2643,
+      "step": 3138
+    },
+    {
+      "epoch": 0.847004856988667,
+      "grad_norm": 1.196414828300476,
+      "learning_rate": 1.2032925457758794e-06,
+      "loss": 0.2617,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8472746896923907,
+      "grad_norm": 1.7107715606689453,
+      "learning_rate": 1.199138732061893e-06,
+      "loss": 0.2777,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8475445223961144,
+      "grad_norm": 1.0318551063537598,
+      "learning_rate": 1.1949916429833796e-06,
+      "loss": 0.2701,
+      "step": 3141
+    },
+    {
+      "epoch": 0.847814355099838,
+      "grad_norm": 1.2388701438903809,
+      "learning_rate": 1.1908512817090833e-06,
+      "loss": 0.2529,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8480841878035618,
+      "grad_norm": 0.9064311981201172,
+      "learning_rate": 1.1867176514026047e-06,
+      "loss": 0.2385,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8483540205072855,
+      "grad_norm": 0.7733277678489685,
+      "learning_rate": 1.1825907552224014e-06,
+      "loss": 0.213,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8486238532110092,
+      "grad_norm": 2.1889336109161377,
+      "learning_rate": 1.178470596321789e-06,
+      "loss": 0.2706,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8488936859147329,
+      "grad_norm": 1.0966073274612427,
+      "learning_rate": 1.1743571778489315e-06,
+      "loss": 0.2447,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8491635186184565,
+      "grad_norm": 1.2415790557861328,
+      "learning_rate": 1.1702505029468448e-06,
+      "loss": 0.2293,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8494333513221802,
+      "grad_norm": 1.1009591817855835,
+      "learning_rate": 1.1661505747533897e-06,
+      "loss": 0.2317,
+      "step": 3148
+    },
+    {
+      "epoch": 0.849703184025904,
+      "grad_norm": 1.0442945957183838,
+      "learning_rate": 1.162057396401275e-06,
+      "loss": 0.2359,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8499730167296277,
+      "grad_norm": 0.8923771977424622,
+      "learning_rate": 1.1579709710180499e-06,
+      "loss": 0.2481,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8502428494333514,
+      "grad_norm": 1.1245121955871582,
+      "learning_rate": 1.1538913017261056e-06,
+      "loss": 0.2663,
+      "step": 3151
+    },
+    {
+      "epoch": 0.850512682137075,
+      "grad_norm": 0.9606658816337585,
+      "learning_rate": 1.1498183916426675e-06,
+      "loss": 0.2517,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8507825148407987,
+      "grad_norm": 0.866111695766449,
+      "learning_rate": 1.1457522438798008e-06,
+      "loss": 0.2488,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8510523475445224,
+      "grad_norm": 0.9385152459144592,
+      "learning_rate": 1.1416928615444013e-06,
+      "loss": 0.2252,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8513221802482461,
+      "grad_norm": 1.1607122421264648,
+      "learning_rate": 1.1376402477381942e-06,
+      "loss": 0.262,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8515920129519697,
+      "grad_norm": 1.5690966844558716,
+      "learning_rate": 1.1335944055577364e-06,
+      "loss": 0.2446,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8518618456556935,
+      "grad_norm": 1.0012922286987305,
+      "learning_rate": 1.1295553380944068e-06,
+      "loss": 0.2537,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8521316783594172,
+      "grad_norm": 0.8946221470832825,
+      "learning_rate": 1.1255230484344115e-06,
+      "loss": 0.2773,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8524015110631409,
+      "grad_norm": 0.8122677803039551,
+      "learning_rate": 1.1214975396587746e-06,
+      "loss": 0.2148,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8526713437668646,
+      "grad_norm": 0.8336051106452942,
+      "learning_rate": 1.1174788148433423e-06,
+      "loss": 0.2621,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8529411764705882,
+      "grad_norm": 1.189158320426941,
+      "learning_rate": 1.1134668770587731e-06,
+      "loss": 0.2697,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8532110091743119,
+      "grad_norm": 1.1680986881256104,
+      "learning_rate": 1.1094617293705434e-06,
+      "loss": 0.2692,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8534808418780356,
+      "grad_norm": 0.9973505139350891,
+      "learning_rate": 1.10546337483894e-06,
+      "loss": 0.2461,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8537506745817594,
+      "grad_norm": 1.5087894201278687,
+      "learning_rate": 1.1014718165190573e-06,
+      "loss": 0.2383,
+      "step": 3164
+    },
+    {
+      "epoch": 0.854020507285483,
+      "grad_norm": 0.8700072765350342,
+      "learning_rate": 1.097487057460801e-06,
+      "loss": 0.2364,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8542903399892067,
+      "grad_norm": 0.9738239049911499,
+      "learning_rate": 1.0935091007088761e-06,
+      "loss": 0.2564,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8545601726929304,
+      "grad_norm": 0.8632180094718933,
+      "learning_rate": 1.0895379493027957e-06,
+      "loss": 0.2495,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8548300053966541,
+      "grad_norm": 0.9393500685691833,
+      "learning_rate": 1.0855736062768684e-06,
+      "loss": 0.243,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8550998381003778,
+      "grad_norm": 1.1241251230239868,
+      "learning_rate": 1.0816160746602033e-06,
+      "loss": 0.2674,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8553696708041014,
+      "grad_norm": 1.0607774257659912,
+      "learning_rate": 1.0776653574767037e-06,
+      "loss": 0.2472,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8556395035078251,
+      "grad_norm": 1.4021737575531006,
+      "learning_rate": 1.0737214577450671e-06,
+      "loss": 0.2812,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8559093362115489,
+      "grad_norm": 0.9835765361785889,
+      "learning_rate": 1.069784378478781e-06,
+      "loss": 0.2233,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8561791689152726,
+      "grad_norm": 0.7896925806999207,
+      "learning_rate": 1.0658541226861207e-06,
+      "loss": 0.2793,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8564490016189962,
+      "grad_norm": 1.5671240091323853,
+      "learning_rate": 1.0619306933701513e-06,
+      "loss": 0.2439,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8567188343227199,
+      "grad_norm": 0.8586401343345642,
+      "learning_rate": 1.0580140935287176e-06,
+      "loss": 0.2399,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8569886670264436,
+      "grad_norm": 2.012503147125244,
+      "learning_rate": 1.0541043261544482e-06,
+      "loss": 0.2611,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8572584997301673,
+      "grad_norm": 1.155657172203064,
+      "learning_rate": 1.050201394234751e-06,
+      "loss": 0.2599,
+      "step": 3177
+    },
+    {
+      "epoch": 0.857528332433891,
+      "grad_norm": 1.100838303565979,
+      "learning_rate": 1.046305300751811e-06,
+      "loss": 0.2105,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8577981651376146,
+      "grad_norm": 0.941137433052063,
+      "learning_rate": 1.0424160486825863e-06,
+      "loss": 0.2267,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8580679978413384,
+      "grad_norm": 1.4465572834014893,
+      "learning_rate": 1.0385336409988112e-06,
+      "loss": 0.2768,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8583378305450621,
+      "grad_norm": 1.5805350542068481,
+      "learning_rate": 1.034658080666986e-06,
+      "loss": 0.2352,
+      "step": 3181
+    },
+    {
+      "epoch": 0.8586076632487858,
+      "grad_norm": 0.87136310338974,
+      "learning_rate": 1.0307893706483829e-06,
+      "loss": 0.2165,
+      "step": 3182
+    },
+    {
+      "epoch": 0.8588774959525094,
+      "grad_norm": 0.9064762592315674,
+      "learning_rate": 1.0269275138990376e-06,
+      "loss": 0.2726,
+      "step": 3183
+    },
+    {
+      "epoch": 0.8591473286562331,
+      "grad_norm": 0.9657045602798462,
+      "learning_rate": 1.0230725133697495e-06,
+      "loss": 0.2674,
+      "step": 3184
+    },
+    {
+      "epoch": 0.8594171613599568,
+      "grad_norm": 1.4019486904144287,
+      "learning_rate": 1.0192243720060802e-06,
+      "loss": 0.2666,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8596869940636805,
+      "grad_norm": 1.0275709629058838,
+      "learning_rate": 1.0153830927483489e-06,
+      "loss": 0.2405,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8599568267674043,
+      "grad_norm": 0.9673107862472534,
+      "learning_rate": 1.011548678531632e-06,
+      "loss": 0.2497,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8602266594711279,
+      "grad_norm": 0.9517835378646851,
+      "learning_rate": 1.0077211322857617e-06,
+      "loss": 0.2171,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8604964921748516,
+      "grad_norm": 1.0294504165649414,
+      "learning_rate": 1.0039004569353217e-06,
+      "loss": 0.2339,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8607663248785753,
+      "grad_norm": 1.7520965337753296,
+      "learning_rate": 1.0000866553996436e-06,
+      "loss": 0.2579,
+      "step": 3190
+    },
+    {
+      "epoch": 0.861036157582299,
+      "grad_norm": 1.1611276865005493,
+      "learning_rate": 9.962797305928086e-07,
+      "loss": 0.2549,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8613059902860226,
+      "grad_norm": 1.1429972648620605,
+      "learning_rate": 9.924796854236485e-07,
+      "loss": 0.2878,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8615758229897463,
+      "grad_norm": 0.9611542224884033,
+      "learning_rate": 9.886865227957277e-07,
+      "loss": 0.2586,
+      "step": 3193
+    },
+    {
+      "epoch": 0.86184565569347,
+      "grad_norm": 2.0560853481292725,
+      "learning_rate": 9.849002456073586e-07,
+      "loss": 0.2845,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8621154883971938,
+      "grad_norm": 1.0641931295394897,
+      "learning_rate": 9.81120856751595e-07,
+      "loss": 0.24,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.8908579349517822,
+      "learning_rate": 9.773483591162203e-07,
+      "loss": 0.2688,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8626551538046411,
+      "grad_norm": 1.854861855506897,
+      "learning_rate": 9.73582755583755e-07,
+      "loss": 0.2683,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8629249865083648,
+      "grad_norm": 1.7097156047821045,
+      "learning_rate": 9.698240490314591e-07,
+      "loss": 0.2327,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8631948192120885,
+      "grad_norm": 1.680273413658142,
+      "learning_rate": 9.660722423313106e-07,
+      "loss": 0.2517,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8634646519158122,
+      "grad_norm": 0.9566216468811035,
+      "learning_rate": 9.62327338350022e-07,
+      "loss": 0.2849,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8637344846195358,
+      "grad_norm": 0.958699643611908,
+      "learning_rate": 9.58589339949032e-07,
+      "loss": 0.2537,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8640043173232596,
+      "grad_norm": 0.8262258172035217,
+      "learning_rate": 9.548582499845015e-07,
+      "loss": 0.2535,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8642741500269833,
+      "grad_norm": 1.48940110206604,
+      "learning_rate": 9.511340713073125e-07,
+      "loss": 0.2407,
+      "step": 3203
+    },
+    {
+      "epoch": 0.864543982730707,
+      "grad_norm": 1.2870721817016602,
+      "learning_rate": 9.474168067630663e-07,
+      "loss": 0.2434,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8648138154344307,
+      "grad_norm": 1.191909670829773,
+      "learning_rate": 9.43706459192082e-07,
+      "loss": 0.2177,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8650836481381543,
+      "grad_norm": 1.16819167137146,
+      "learning_rate": 9.400030314293918e-07,
+      "loss": 0.23,
+      "step": 3206
+    },
+    {
+      "epoch": 0.865353480841878,
+      "grad_norm": 1.7302374839782715,
+      "learning_rate": 9.363065263047433e-07,
+      "loss": 0.2633,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8656233135456017,
+      "grad_norm": 2.289484739303589,
+      "learning_rate": 9.326169466425916e-07,
+      "loss": 0.2581,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8658931462493255,
+      "grad_norm": 1.1625593900680542,
+      "learning_rate": 9.289342952621028e-07,
+      "loss": 0.251,
+      "step": 3209
+    },
+    {
+      "epoch": 0.8661629789530491,
+      "grad_norm": 1.246585488319397,
+      "learning_rate": 9.252585749771459e-07,
+      "loss": 0.22,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8664328116567728,
+      "grad_norm": 0.9701352119445801,
+      "learning_rate": 9.215897885962988e-07,
+      "loss": 0.2562,
+      "step": 3211
+    },
+    {
+      "epoch": 0.8667026443604965,
+      "grad_norm": 0.9250827431678772,
+      "learning_rate": 9.179279389228357e-07,
+      "loss": 0.2334,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8669724770642202,
+      "grad_norm": 1.380204200744629,
+      "learning_rate": 9.142730287547341e-07,
+      "loss": 0.2807,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8672423097679439,
+      "grad_norm": 1.4393200874328613,
+      "learning_rate": 9.106250608846679e-07,
+      "loss": 0.277,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8675121424716675,
+      "grad_norm": 0.9013667702674866,
+      "learning_rate": 9.069840381000083e-07,
+      "loss": 0.2589,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8677819751753912,
+      "grad_norm": 1.0673787593841553,
+      "learning_rate": 9.033499631828169e-07,
+      "loss": 0.2656,
+      "step": 3216
+    },
+    {
+      "epoch": 0.868051807879115,
+      "grad_norm": 0.8352909088134766,
+      "learning_rate": 8.997228389098489e-07,
+      "loss": 0.2193,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8683216405828387,
+      "grad_norm": 1.1712286472320557,
+      "learning_rate": 8.961026680525465e-07,
+      "loss": 0.2491,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8685914732865623,
+      "grad_norm": 1.0909615755081177,
+      "learning_rate": 8.924894533770411e-07,
+      "loss": 0.2492,
+      "step": 3219
+    },
+    {
+      "epoch": 0.868861305990286,
+      "grad_norm": 1.0856380462646484,
+      "learning_rate": 8.888831976441481e-07,
+      "loss": 0.2485,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8691311386940097,
+      "grad_norm": 1.5610781908035278,
+      "learning_rate": 8.852839036093642e-07,
+      "loss": 0.226,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8694009713977334,
+      "grad_norm": 1.0179191827774048,
+      "learning_rate": 8.816915740228682e-07,
+      "loss": 0.2379,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8696708041014571,
+      "grad_norm": 1.1423559188842773,
+      "learning_rate": 8.78106211629518e-07,
+      "loss": 0.2493,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8699406368051807,
+      "grad_norm": 0.8643580675125122,
+      "learning_rate": 8.745278191688477e-07,
+      "loss": 0.214,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8702104695089045,
+      "grad_norm": 1.0923187732696533,
+      "learning_rate": 8.709563993750625e-07,
+      "loss": 0.2454,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8704803022126282,
+      "grad_norm": 0.9153228998184204,
+      "learning_rate": 8.673919549770483e-07,
+      "loss": 0.2298,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8707501349163519,
+      "grad_norm": 0.9973602890968323,
+      "learning_rate": 8.638344886983496e-07,
+      "loss": 0.2288,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8710199676200756,
+      "grad_norm": 0.9440615177154541,
+      "learning_rate": 8.602840032571847e-07,
+      "loss": 0.2256,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8712898003237992,
+      "grad_norm": 2.0998215675354004,
+      "learning_rate": 8.567405013664454e-07,
+      "loss": 0.2716,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8715596330275229,
+      "grad_norm": 1.1322108507156372,
+      "learning_rate": 8.532039857336727e-07,
+      "loss": 0.2543,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8718294657312466,
+      "grad_norm": 0.8539922833442688,
+      "learning_rate": 8.496744590610784e-07,
+      "loss": 0.2314,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8720992984349704,
+      "grad_norm": 1.4085053205490112,
+      "learning_rate": 8.461519240455362e-07,
+      "loss": 0.2488,
+      "step": 3232
+    },
+    {
+      "epoch": 0.872369131138694,
+      "grad_norm": 1.0167498588562012,
+      "learning_rate": 8.426363833785711e-07,
+      "loss": 0.24,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8726389638424177,
+      "grad_norm": 0.9038625955581665,
+      "learning_rate": 8.391278397463653e-07,
+      "loss": 0.2393,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8729087965461414,
+      "grad_norm": 1.0785290002822876,
+      "learning_rate": 8.356262958297612e-07,
+      "loss": 0.2808,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8731786292498651,
+      "grad_norm": 0.8531122207641602,
+      "learning_rate": 8.321317543042428e-07,
+      "loss": 0.2388,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8734484619535888,
+      "grad_norm": 1.180879831314087,
+      "learning_rate": 8.28644217839949e-07,
+      "loss": 0.2784,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8737182946573124,
+      "grad_norm": 1.4731197357177734,
+      "learning_rate": 8.251636891016702e-07,
+      "loss": 0.2349,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8739881273610361,
+      "grad_norm": 0.9897095561027527,
+      "learning_rate": 8.216901707488334e-07,
+      "loss": 0.2433,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8742579600647599,
+      "grad_norm": 1.5730280876159668,
+      "learning_rate": 8.182236654355125e-07,
+      "loss": 0.288,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8745277927684836,
+      "grad_norm": 1.0827752351760864,
+      "learning_rate": 8.147641758104297e-07,
+      "loss": 0.2386,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8747976254722072,
+      "grad_norm": 0.8646848201751709,
+      "learning_rate": 8.113117045169361e-07,
+      "loss": 0.2543,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8750674581759309,
+      "grad_norm": 1.0002316236495972,
+      "learning_rate": 8.078662541930226e-07,
+      "loss": 0.2313,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8753372908796546,
+      "grad_norm": 0.9016987085342407,
+      "learning_rate": 8.044278274713246e-07,
+      "loss": 0.2453,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8756071235833783,
+      "grad_norm": 1.6987379789352417,
+      "learning_rate": 8.009964269790982e-07,
+      "loss": 0.2456,
+      "step": 3245
+    },
+    {
+      "epoch": 0.875876956287102,
+      "grad_norm": 0.9269284009933472,
+      "learning_rate": 7.975720553382371e-07,
+      "loss": 0.2395,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8761467889908257,
+      "grad_norm": 1.3558449745178223,
+      "learning_rate": 7.941547151652674e-07,
+      "loss": 0.2878,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8764166216945494,
+      "grad_norm": 0.8435084223747253,
+      "learning_rate": 7.907444090713357e-07,
+      "loss": 0.2278,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8766864543982731,
+      "grad_norm": 1.364548921585083,
+      "learning_rate": 7.873411396622155e-07,
+      "loss": 0.2371,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8769562871019968,
+      "grad_norm": 0.8614928126335144,
+      "learning_rate": 7.839449095383111e-07,
+      "loss": 0.2347,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8772261198057204,
+      "grad_norm": 1.3455554246902466,
+      "learning_rate": 7.805557212946369e-07,
+      "loss": 0.2536,
+      "step": 3251
+    },
+    {
+      "epoch": 0.8774959525094441,
+      "grad_norm": 1.7070667743682861,
+      "learning_rate": 7.771735775208322e-07,
+      "loss": 0.2454,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8777657852131678,
+      "grad_norm": 0.779703676700592,
+      "learning_rate": 7.737984808011589e-07,
+      "loss": 0.2206,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8780356179168916,
+      "grad_norm": 1.5412483215332031,
+      "learning_rate": 7.704304337144841e-07,
+      "loss": 0.2477,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8783054506206153,
+      "grad_norm": 1.5307430028915405,
+      "learning_rate": 7.670694388342936e-07,
+      "loss": 0.2498,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8785752833243389,
+      "grad_norm": 1.3116364479064941,
+      "learning_rate": 7.637154987286888e-07,
+      "loss": 0.2465,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8788451160280626,
+      "grad_norm": 1.0793054103851318,
+      "learning_rate": 7.603686159603708e-07,
+      "loss": 0.2363,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8791149487317863,
+      "grad_norm": 0.9958755373954773,
+      "learning_rate": 7.570287930866538e-07,
+      "loss": 0.2471,
+      "step": 3258
+    },
+    {
+      "epoch": 0.87938478143551,
+      "grad_norm": 1.7137540578842163,
+      "learning_rate": 7.536960326594623e-07,
+      "loss": 0.211,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8796546141392336,
+      "grad_norm": 1.74440336227417,
+      "learning_rate": 7.50370337225318e-07,
+      "loss": 0.2294,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8799244468429573,
+      "grad_norm": 0.9219018816947937,
+      "learning_rate": 7.470517093253427e-07,
+      "loss": 0.2379,
+      "step": 3261
+    },
+    {
+      "epoch": 0.8801942795466811,
+      "grad_norm": 0.9635429382324219,
+      "learning_rate": 7.437401514952646e-07,
+      "loss": 0.2473,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8804641122504048,
+      "grad_norm": 0.842802882194519,
+      "learning_rate": 7.404356662654067e-07,
+      "loss": 0.2627,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 1.537770390510559,
+      "learning_rate": 7.371382561606833e-07,
+      "loss": 0.2824,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8810037776578521,
+      "grad_norm": 1.4165436029434204,
+      "learning_rate": 7.338479237006125e-07,
+      "loss": 0.2494,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8812736103615758,
+      "grad_norm": 1.2259719371795654,
+      "learning_rate": 7.305646713992976e-07,
+      "loss": 0.2556,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8815434430652995,
+      "grad_norm": 1.226577639579773,
+      "learning_rate": 7.272885017654285e-07,
+      "loss": 0.2334,
+      "step": 3267
+    },
+    {
+      "epoch": 0.8818132757690232,
+      "grad_norm": 0.9988632798194885,
+      "learning_rate": 7.240194173022941e-07,
+      "loss": 0.2535,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8820831084727468,
+      "grad_norm": 1.3131110668182373,
+      "learning_rate": 7.207574205077639e-07,
+      "loss": 0.2697,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.8588821887969971,
+      "learning_rate": 7.175025138742853e-07,
+      "loss": 0.2128,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8826227738801943,
+      "grad_norm": 1.1451349258422852,
+      "learning_rate": 7.14254699888901e-07,
+      "loss": 0.2399,
+      "step": 3271
+    },
+    {
+      "epoch": 0.882892606583918,
+      "grad_norm": 0.9475056529045105,
+      "learning_rate": 7.110139810332262e-07,
+      "loss": 0.2296,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8831624392876417,
+      "grad_norm": 0.9315005540847778,
+      "learning_rate": 7.077803597834532e-07,
+      "loss": 0.2398,
+      "step": 3273
+    },
+    {
+      "epoch": 0.8834322719913653,
+      "grad_norm": 1.1523480415344238,
+      "learning_rate": 7.045538386103579e-07,
+      "loss": 0.2649,
+      "step": 3274
+    },
+    {
+      "epoch": 0.883702104695089,
+      "grad_norm": 1.1371192932128906,
+      "learning_rate": 7.013344199792871e-07,
+      "loss": 0.2342,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8839719373988127,
+      "grad_norm": 1.022342562675476,
+      "learning_rate": 6.981221063501576e-07,
+      "loss": 0.2441,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8842417701025365,
+      "grad_norm": 0.9332855939865112,
+      "learning_rate": 6.949169001774658e-07,
+      "loss": 0.2662,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8845116028062601,
+      "grad_norm": 0.9957926273345947,
+      "learning_rate": 6.917188039102707e-07,
+      "loss": 0.2558,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8847814355099838,
+      "grad_norm": 1.4324427843093872,
+      "learning_rate": 6.885278199921964e-07,
+      "loss": 0.2664,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8850512682137075,
+      "grad_norm": 0.9028330445289612,
+      "learning_rate": 6.853439508614412e-07,
+      "loss": 0.2374,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8853211009174312,
+      "grad_norm": 1.0852539539337158,
+      "learning_rate": 6.821671989507627e-07,
+      "loss": 0.2526,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8855909336211549,
+      "grad_norm": 1.0124531984329224,
+      "learning_rate": 6.789975666874759e-07,
+      "loss": 0.2559,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8858607663248785,
+      "grad_norm": 1.2266390323638916,
+      "learning_rate": 6.758350564934624e-07,
+      "loss": 0.2456,
+      "step": 3283
+    },
+    {
+      "epoch": 0.8861305990286023,
+      "grad_norm": 1.2077546119689941,
+      "learning_rate": 6.726796707851613e-07,
+      "loss": 0.2483,
+      "step": 3284
+    },
+    {
+      "epoch": 0.886400431732326,
+      "grad_norm": 0.9771092534065247,
+      "learning_rate": 6.695314119735607e-07,
+      "loss": 0.2516,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8866702644360497,
+      "grad_norm": 1.158041000366211,
+      "learning_rate": 6.663902824642132e-07,
+      "loss": 0.2889,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8869400971397733,
+      "grad_norm": 1.3121812343597412,
+      "learning_rate": 6.6325628465722e-07,
+      "loss": 0.2503,
+      "step": 3287
+    },
+    {
+      "epoch": 0.887209929843497,
+      "grad_norm": 1.4837623834609985,
+      "learning_rate": 6.601294209472253e-07,
+      "loss": 0.2642,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8874797625472207,
+      "grad_norm": 1.3000531196594238,
+      "learning_rate": 6.57009693723436e-07,
+      "loss": 0.2577,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8877495952509444,
+      "grad_norm": 0.938453197479248,
+      "learning_rate": 6.538971053695986e-07,
+      "loss": 0.2245,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8880194279546682,
+      "grad_norm": 1.0897566080093384,
+      "learning_rate": 6.507916582640006e-07,
+      "loss": 0.2719,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8882892606583918,
+      "grad_norm": 1.263576865196228,
+      "learning_rate": 6.47693354779484e-07,
+      "loss": 0.2895,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8885590933621155,
+      "grad_norm": 0.8857362270355225,
+      "learning_rate": 6.446021972834227e-07,
+      "loss": 0.2468,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8888289260658392,
+      "grad_norm": 1.4022942781448364,
+      "learning_rate": 6.415181881377364e-07,
+      "loss": 0.2978,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8890987587695629,
+      "grad_norm": 0.9230780005455017,
+      "learning_rate": 6.384413296988779e-07,
+      "loss": 0.2651,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8893685914732865,
+      "grad_norm": 1.071678638458252,
+      "learning_rate": 6.353716243178409e-07,
+      "loss": 0.2978,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8896384241770102,
+      "grad_norm": 0.8500325679779053,
+      "learning_rate": 6.323090743401505e-07,
+      "loss": 0.236,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8899082568807339,
+      "grad_norm": 0.9790142178535461,
+      "learning_rate": 6.29253682105866e-07,
+      "loss": 0.2686,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8901780895844577,
+      "grad_norm": 0.8554282784461975,
+      "learning_rate": 6.262054499495773e-07,
+      "loss": 0.2484,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8904479222881814,
+      "grad_norm": 1.6469272375106812,
+      "learning_rate": 6.231643802004017e-07,
+      "loss": 0.2713,
+      "step": 3300
+    },
+    {
+      "epoch": 0.890717754991905,
+      "grad_norm": 0.8707671165466309,
+      "learning_rate": 6.201304751819859e-07,
+      "loss": 0.2285,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8909875876956287,
+      "grad_norm": 1.1698243618011475,
+      "learning_rate": 6.171037372125011e-07,
+      "loss": 0.2311,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8912574203993524,
+      "grad_norm": 0.9607957601547241,
+      "learning_rate": 6.140841686046439e-07,
+      "loss": 0.2607,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8915272531030761,
+      "grad_norm": 1.509322166442871,
+      "learning_rate": 6.110717716656289e-07,
+      "loss": 0.2274,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8917970858067998,
+      "grad_norm": 1.7903807163238525,
+      "learning_rate": 6.080665486971959e-07,
+      "loss": 0.2044,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8920669185105234,
+      "grad_norm": 0.8146570324897766,
+      "learning_rate": 6.050685019955993e-07,
+      "loss": 0.2243,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8923367512142472,
+      "grad_norm": 1.2276428937911987,
+      "learning_rate": 6.020776338516133e-07,
+      "loss": 0.234,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8926065839179709,
+      "grad_norm": 0.9314084053039551,
+      "learning_rate": 5.99093946550523e-07,
+      "loss": 0.2354,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8928764166216946,
+      "grad_norm": 1.4974128007888794,
+      "learning_rate": 5.961174423721295e-07,
+      "loss": 0.2782,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8931462493254182,
+      "grad_norm": 1.94919753074646,
+      "learning_rate": 5.931481235907466e-07,
+      "loss": 0.2667,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8934160820291419,
+      "grad_norm": 0.955417275428772,
+      "learning_rate": 5.901859924751951e-07,
+      "loss": 0.2636,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8936859147328656,
+      "grad_norm": 1.0281600952148438,
+      "learning_rate": 5.872310512888046e-07,
+      "loss": 0.2628,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8939557474365893,
+      "grad_norm": 1.4251620769500732,
+      "learning_rate": 5.842833022894124e-07,
+      "loss": 0.2629,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8942255801403131,
+      "grad_norm": 1.0249937772750854,
+      "learning_rate": 5.813427477293565e-07,
+      "loss": 0.2589,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8944954128440367,
+      "grad_norm": 1.1251027584075928,
+      "learning_rate": 5.784093898554855e-07,
+      "loss": 0.2269,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8947652455477604,
+      "grad_norm": 1.3190134763717651,
+      "learning_rate": 5.754832309091362e-07,
+      "loss": 0.2294,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8950350782514841,
+      "grad_norm": 0.8861879110336304,
+      "learning_rate": 5.725642731261593e-07,
+      "loss": 0.2696,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8953049109552078,
+      "grad_norm": 1.2382797002792358,
+      "learning_rate": 5.696525187368951e-07,
+      "loss": 0.2326,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8955747436589314,
+      "grad_norm": 1.0307719707489014,
+      "learning_rate": 5.667479699661771e-07,
+      "loss": 0.2283,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8958445763626551,
+      "grad_norm": 0.8295007348060608,
+      "learning_rate": 5.638506290333424e-07,
+      "loss": 0.2231,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8961144090663788,
+      "grad_norm": 0.8128833770751953,
+      "learning_rate": 5.609604981522143e-07,
+      "loss": 0.2448,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8963842417701026,
+      "grad_norm": 0.8469763398170471,
+      "learning_rate": 5.580775795311033e-07,
+      "loss": 0.238,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8966540744738263,
+      "grad_norm": 0.8922017812728882,
+      "learning_rate": 5.552018753728195e-07,
+      "loss": 0.2125,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8969239071775499,
+      "grad_norm": 1.8300600051879883,
+      "learning_rate": 5.523333878746529e-07,
+      "loss": 0.2757,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8971937398812736,
+      "grad_norm": 1.1821825504302979,
+      "learning_rate": 5.494721192283814e-07,
+      "loss": 0.239,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8974635725849973,
+      "grad_norm": 0.8093709945678711,
+      "learning_rate": 5.466180716202674e-07,
+      "loss": 0.2352,
+      "step": 3326
+    },
+    {
+      "epoch": 0.897733405288721,
+      "grad_norm": 1.1772170066833496,
+      "learning_rate": 5.437712472310552e-07,
+      "loss": 0.2282,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8980032379924446,
+      "grad_norm": 1.0345453023910522,
+      "learning_rate": 5.409316482359694e-07,
+      "loss": 0.2279,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8982730706961684,
+      "grad_norm": 1.1100983619689941,
+      "learning_rate": 5.380992768047155e-07,
+      "loss": 0.2231,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8985429033998921,
+      "grad_norm": 1.709969162940979,
+      "learning_rate": 5.352741351014745e-07,
+      "loss": 0.2987,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8988127361036158,
+      "grad_norm": 1.2787803411483765,
+      "learning_rate": 5.324562252849053e-07,
+      "loss": 0.2537,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.9516897797584534,
+      "learning_rate": 5.296455495081387e-07,
+      "loss": 0.2496,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8993524015110631,
+      "grad_norm": 1.0533305406570435,
+      "learning_rate": 5.268421099187804e-07,
+      "loss": 0.2637,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8996222342147868,
+      "grad_norm": 0.9942564368247986,
+      "learning_rate": 5.240459086589056e-07,
+      "loss": 0.2315,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8998920669185105,
+      "grad_norm": 1.1651999950408936,
+      "learning_rate": 5.212569478650586e-07,
+      "loss": 0.2546,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9001618996222343,
+      "grad_norm": 0.8621547222137451,
+      "learning_rate": 5.184752296682527e-07,
+      "loss": 0.2787,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9004317323259579,
+      "grad_norm": 0.9615342020988464,
+      "learning_rate": 5.157007561939653e-07,
+      "loss": 0.231,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9007015650296816,
+      "grad_norm": 1.1469202041625977,
+      "learning_rate": 5.129335295621407e-07,
+      "loss": 0.2786,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9009713977334053,
+      "grad_norm": 1.6831563711166382,
+      "learning_rate": 5.101735518871831e-07,
+      "loss": 0.2589,
+      "step": 3339
+    },
+    {
+      "epoch": 0.901241230437129,
+      "grad_norm": 1.670232892036438,
+      "learning_rate": 5.074208252779589e-07,
+      "loss": 0.2419,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9015110631408527,
+      "grad_norm": 1.0447397232055664,
+      "learning_rate": 5.046753518377966e-07,
+      "loss": 0.261,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9017808958445763,
+      "grad_norm": 0.8965407609939575,
+      "learning_rate": 5.019371336644774e-07,
+      "loss": 0.2717,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9020507285483,
+      "grad_norm": 1.0223729610443115,
+      "learning_rate": 4.992061728502451e-07,
+      "loss": 0.27,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9023205612520238,
+      "grad_norm": 0.9343674182891846,
+      "learning_rate": 4.964824714817928e-07,
+      "loss": 0.2681,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9025903939557475,
+      "grad_norm": 0.7298147082328796,
+      "learning_rate": 4.937660316402693e-07,
+      "loss": 0.2069,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9028602266594711,
+      "grad_norm": 0.9183060526847839,
+      "learning_rate": 4.910568554012751e-07,
+      "loss": 0.2856,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9031300593631948,
+      "grad_norm": 0.891181468963623,
+      "learning_rate": 4.883549448348601e-07,
+      "loss": 0.2476,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9033998920669185,
+      "grad_norm": 0.7883742451667786,
+      "learning_rate": 4.856603020055228e-07,
+      "loss": 0.2478,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9036697247706422,
+      "grad_norm": 1.1982215642929077,
+      "learning_rate": 4.829729289722085e-07,
+      "loss": 0.2766,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9039395574743659,
+      "grad_norm": 1.9756836891174316,
+      "learning_rate": 4.802928277883056e-07,
+      "loss": 0.2468,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9042093901780895,
+      "grad_norm": 1.207680344581604,
+      "learning_rate": 4.776200005016518e-07,
+      "loss": 0.2623,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9044792228818133,
+      "grad_norm": 1.0128222703933716,
+      "learning_rate": 4.749544491545199e-07,
+      "loss": 0.261,
+      "step": 3352
+    },
+    {
+      "epoch": 0.904749055585537,
+      "grad_norm": 1.037597894668579,
+      "learning_rate": 4.722961757836286e-07,
+      "loss": 0.2421,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9050188882892607,
+      "grad_norm": 1.1478517055511475,
+      "learning_rate": 4.69645182420132e-07,
+      "loss": 0.2488,
+      "step": 3354
+    },
+    {
+      "epoch": 0.9052887209929843,
+      "grad_norm": 0.8490355014801025,
+      "learning_rate": 4.670014710896253e-07,
+      "loss": 0.2728,
+      "step": 3355
+    },
+    {
+      "epoch": 0.905558553696708,
+      "grad_norm": 1.1801671981811523,
+      "learning_rate": 4.643650438121361e-07,
+      "loss": 0.2522,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9058283864004317,
+      "grad_norm": 0.8276525735855103,
+      "learning_rate": 4.617359026021273e-07,
+      "loss": 0.2357,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9060982191041554,
+      "grad_norm": 0.9166259765625,
+      "learning_rate": 4.591140494684965e-07,
+      "loss": 0.2346,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9063680518078792,
+      "grad_norm": 1.5225701332092285,
+      "learning_rate": 4.5649948641457e-07,
+      "loss": 0.2475,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9066378845116028,
+      "grad_norm": 0.8455879092216492,
+      "learning_rate": 4.5389221543810533e-07,
+      "loss": 0.2511,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9069077172153265,
+      "grad_norm": 0.986233115196228,
+      "learning_rate": 4.512922385312868e-07,
+      "loss": 0.2343,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9071775499190502,
+      "grad_norm": 1.0894365310668945,
+      "learning_rate": 4.486995576807285e-07,
+      "loss": 0.2423,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9074473826227739,
+      "grad_norm": 0.8907444477081299,
+      "learning_rate": 4.461141748674669e-07,
+      "loss": 0.223,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9077172153264975,
+      "grad_norm": 0.8875721096992493,
+      "learning_rate": 4.435360920669618e-07,
+      "loss": 0.2648,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9079870480302212,
+      "grad_norm": 0.9952379465103149,
+      "learning_rate": 4.4096531124909746e-07,
+      "loss": 0.2237,
+      "step": 3365
+    },
+    {
+      "epoch": 0.908256880733945,
+      "grad_norm": 0.9354422688484192,
+      "learning_rate": 4.3840183437817707e-07,
+      "loss": 0.2543,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9085267134376687,
+      "grad_norm": 1.17440664768219,
+      "learning_rate": 4.3584566341292265e-07,
+      "loss": 0.2718,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9087965461413924,
+      "grad_norm": 0.8992283344268799,
+      "learning_rate": 4.3329680030647635e-07,
+      "loss": 0.2496,
+      "step": 3368
+    },
+    {
+      "epoch": 0.909066378845116,
+      "grad_norm": 1.2564210891723633,
+      "learning_rate": 4.307552470063936e-07,
+      "loss": 0.2513,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9093362115488397,
+      "grad_norm": 0.8898721933364868,
+      "learning_rate": 4.282210054546454e-07,
+      "loss": 0.2544,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9096060442525634,
+      "grad_norm": 0.9556162357330322,
+      "learning_rate": 4.2569407758761615e-07,
+      "loss": 0.2569,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9098758769562871,
+      "grad_norm": 0.9676016569137573,
+      "learning_rate": 4.231744653361025e-07,
+      "loss": 0.257,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9101457096600108,
+      "grad_norm": 0.9668446779251099,
+      "learning_rate": 4.206621706253111e-07,
+      "loss": 0.2591,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9104155423637345,
+      "grad_norm": 2.9116499423980713,
+      "learning_rate": 4.1815719537485647e-07,
+      "loss": 0.3121,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9106853750674582,
+      "grad_norm": 1.9127647876739502,
+      "learning_rate": 4.1565954149876077e-07,
+      "loss": 0.2622,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9109552077711819,
+      "grad_norm": 0.9674962162971497,
+      "learning_rate": 4.1316921090545305e-07,
+      "loss": 0.2618,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9112250404749056,
+      "grad_norm": 1.0007799863815308,
+      "learning_rate": 4.106862054977656e-07,
+      "loss": 0.248,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9114948731786292,
+      "grad_norm": 1.2965437173843384,
+      "learning_rate": 4.082105271729331e-07,
+      "loss": 0.2586,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9117647058823529,
+      "grad_norm": 0.9328044056892395,
+      "learning_rate": 4.057421778225934e-07,
+      "loss": 0.2808,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9120345385860766,
+      "grad_norm": 1.0520893335342407,
+      "learning_rate": 4.032811593327823e-07,
+      "loss": 0.2505,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9123043712898004,
+      "grad_norm": 1.693113923072815,
+      "learning_rate": 4.008274735839368e-07,
+      "loss": 0.239,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9125742039935241,
+      "grad_norm": 1.1616286039352417,
+      "learning_rate": 3.9838112245088934e-07,
+      "loss": 0.259,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9128440366972477,
+      "grad_norm": 0.9374892711639404,
+      "learning_rate": 3.95942107802868e-07,
+      "loss": 0.2265,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9131138694009714,
+      "grad_norm": 0.9705605506896973,
+      "learning_rate": 3.9351043150349434e-07,
+      "loss": 0.2602,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9133837021046951,
+      "grad_norm": 1.5652246475219727,
+      "learning_rate": 3.9108609541078537e-07,
+      "loss": 0.2778,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9136535348084188,
+      "grad_norm": 1.515113115310669,
+      "learning_rate": 3.886691013771482e-07,
+      "loss": 0.2345,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9139233675121424,
+      "grad_norm": 1.144887089729309,
+      "learning_rate": 3.862594512493778e-07,
+      "loss": 0.2825,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9141932002158661,
+      "grad_norm": 1.366786241531372,
+      "learning_rate": 3.8385714686866137e-07,
+      "loss": 0.2826,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9144630329195899,
+      "grad_norm": 1.0038634538650513,
+      "learning_rate": 3.8146219007057175e-07,
+      "loss": 0.2209,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9147328656233136,
+      "grad_norm": 1.011917233467102,
+      "learning_rate": 3.7907458268506724e-07,
+      "loss": 0.2338,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9150026983270373,
+      "grad_norm": 1.3334553241729736,
+      "learning_rate": 3.7669432653649086e-07,
+      "loss": 0.2632,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9152725310307609,
+      "grad_norm": 1.1774754524230957,
+      "learning_rate": 3.743214234435666e-07,
+      "loss": 0.2644,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9155423637344846,
+      "grad_norm": 1.1296664476394653,
+      "learning_rate": 3.7195587521940633e-07,
+      "loss": 0.2665,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9158121964382083,
+      "grad_norm": 1.0365549325942993,
+      "learning_rate": 3.695976836714932e-07,
+      "loss": 0.2689,
+      "step": 3394
+    },
+    {
+      "epoch": 0.916082029141932,
+      "grad_norm": 0.7891403436660767,
+      "learning_rate": 3.6724685060169575e-07,
+      "loss": 0.2361,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9163518618456556,
+      "grad_norm": 1.1821138858795166,
+      "learning_rate": 3.649033778062616e-07,
+      "loss": 0.2337,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9166216945493794,
+      "grad_norm": 1.0615520477294922,
+      "learning_rate": 3.6256726707580514e-07,
+      "loss": 0.2533,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9168915272531031,
+      "grad_norm": 1.9463436603546143,
+      "learning_rate": 3.6023852019532403e-07,
+      "loss": 0.2617,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9171613599568268,
+      "grad_norm": 1.2681214809417725,
+      "learning_rate": 3.579171389441882e-07,
+      "loss": 0.2666,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 1.5652275085449219,
+      "learning_rate": 3.556031250961356e-07,
+      "loss": 0.2847,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9177010253642741,
+      "grad_norm": 1.0154640674591064,
+      "learning_rate": 3.5329648041927624e-07,
+      "loss": 0.2451,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9179708580679978,
+      "grad_norm": 1.518041729927063,
+      "learning_rate": 3.509972066760947e-07,
+      "loss": 0.2935,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9182406907717215,
+      "grad_norm": 0.9477723836898804,
+      "learning_rate": 3.4870530562343575e-07,
+      "loss": 0.238,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9185105234754453,
+      "grad_norm": 1.7321922779083252,
+      "learning_rate": 3.4642077901251403e-07,
+      "loss": 0.2561,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9187803561791689,
+      "grad_norm": 1.0394009351730347,
+      "learning_rate": 3.4414362858891214e-07,
+      "loss": 0.2375,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9190501888828926,
+      "grad_norm": 1.2746214866638184,
+      "learning_rate": 3.4187385609257275e-07,
+      "loss": 0.2691,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9193200215866163,
+      "grad_norm": 0.8944178819656372,
+      "learning_rate": 3.396114632578018e-07,
+      "loss": 0.2794,
+      "step": 3407
+    },
+    {
+      "epoch": 0.91958985429034,
+      "grad_norm": 1.1912651062011719,
+      "learning_rate": 3.3735645181326994e-07,
+      "loss": 0.2367,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9198596869940637,
+      "grad_norm": 1.0775517225265503,
+      "learning_rate": 3.3510882348200215e-07,
+      "loss": 0.2762,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9201295196977873,
+      "grad_norm": 0.9770665168762207,
+      "learning_rate": 3.328685799813858e-07,
+      "loss": 0.2592,
+      "step": 3410
+    },
+    {
+      "epoch": 0.920399352401511,
+      "grad_norm": 1.0982683897018433,
+      "learning_rate": 3.3063572302316714e-07,
+      "loss": 0.2178,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9206691851052348,
+      "grad_norm": 1.0858373641967773,
+      "learning_rate": 3.284102543134426e-07,
+      "loss": 0.2409,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9209390178089585,
+      "grad_norm": 1.112441062927246,
+      "learning_rate": 3.261921755526676e-07,
+      "loss": 0.2552,
+      "step": 3413
+    },
+    {
+      "epoch": 0.9212088505126821,
+      "grad_norm": 1.0210357904434204,
+      "learning_rate": 3.2398148843565405e-07,
+      "loss": 0.2807,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9214786832164058,
+      "grad_norm": 1.7997992038726807,
+      "learning_rate": 3.2177819465155746e-07,
+      "loss": 0.2386,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9217485159201295,
+      "grad_norm": 1.0974987745285034,
+      "learning_rate": 3.195822958838901e-07,
+      "loss": 0.2825,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9220183486238532,
+      "grad_norm": 0.8938255310058594,
+      "learning_rate": 3.173937938105154e-07,
+      "loss": 0.2428,
+      "step": 3417
+    },
+    {
+      "epoch": 0.922288181327577,
+      "grad_norm": 1.0202192068099976,
+      "learning_rate": 3.152126901036401e-07,
+      "loss": 0.2494,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9225580140313006,
+      "grad_norm": 0.8347780704498291,
+      "learning_rate": 3.130389864298178e-07,
+      "loss": 0.2862,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9228278467350243,
+      "grad_norm": 1.306592345237732,
+      "learning_rate": 3.1087268444995436e-07,
+      "loss": 0.2626,
+      "step": 3420
+    },
+    {
+      "epoch": 0.923097679438748,
+      "grad_norm": 0.926452100276947,
+      "learning_rate": 3.0871378581929456e-07,
+      "loss": 0.2351,
+      "step": 3421
+    },
+    {
+      "epoch": 0.9233675121424717,
+      "grad_norm": 1.2555633783340454,
+      "learning_rate": 3.065622921874245e-07,
+      "loss": 0.2353,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9236373448461953,
+      "grad_norm": 0.9639153480529785,
+      "learning_rate": 3.044182051982814e-07,
+      "loss": 0.2572,
+      "step": 3423
+    },
+    {
+      "epoch": 0.923907177549919,
+      "grad_norm": 1.00819730758667,
+      "learning_rate": 3.0228152649013133e-07,
+      "loss": 0.2514,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9241770102536427,
+      "grad_norm": 0.8635022640228271,
+      "learning_rate": 3.001522576955884e-07,
+      "loss": 0.2409,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9244468429573665,
+      "grad_norm": 1.0054231882095337,
+      "learning_rate": 2.9803040044160126e-07,
+      "loss": 0.2548,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9247166756610902,
+      "grad_norm": 1.3867589235305786,
+      "learning_rate": 2.959159563494585e-07,
+      "loss": 0.2473,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9249865083648138,
+      "grad_norm": 1.0444684028625488,
+      "learning_rate": 2.9380892703477884e-07,
+      "loss": 0.2496,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9252563410685375,
+      "grad_norm": 1.2539169788360596,
+      "learning_rate": 2.9170931410752e-07,
+      "loss": 0.2717,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9255261737722612,
+      "grad_norm": 1.0261061191558838,
+      "learning_rate": 2.896171191719743e-07,
+      "loss": 0.2593,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9257960064759849,
+      "grad_norm": 1.6193578243255615,
+      "learning_rate": 2.8753234382675854e-07,
+      "loss": 0.2732,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9260658391797085,
+      "grad_norm": 0.8564831018447876,
+      "learning_rate": 2.8545498966482954e-07,
+      "loss": 0.2161,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9263356718834322,
+      "grad_norm": 1.6475627422332764,
+      "learning_rate": 2.8338505827346873e-07,
+      "loss": 0.282,
+      "step": 3433
+    },
+    {
+      "epoch": 0.926605504587156,
+      "grad_norm": 0.8187236785888672,
+      "learning_rate": 2.813225512342843e-07,
+      "loss": 0.2229,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9268753372908797,
+      "grad_norm": 0.9141151309013367,
+      "learning_rate": 2.792674701232156e-07,
+      "loss": 0.2289,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9271451699946034,
+      "grad_norm": 0.9707819223403931,
+      "learning_rate": 2.772198165105267e-07,
+      "loss": 0.2704,
+      "step": 3436
+    },
+    {
+      "epoch": 0.927415002698327,
+      "grad_norm": 0.8442431688308716,
+      "learning_rate": 2.7517959196080245e-07,
+      "loss": 0.2193,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9276848354020507,
+      "grad_norm": 1.0632582902908325,
+      "learning_rate": 2.731467980329583e-07,
+      "loss": 0.2752,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9279546681057744,
+      "grad_norm": 1.0130881071090698,
+      "learning_rate": 2.7112143628022833e-07,
+      "loss": 0.2438,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9282245008094981,
+      "grad_norm": 0.9056159853935242,
+      "learning_rate": 2.6910350825016474e-07,
+      "loss": 0.2108,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9284943335132217,
+      "grad_norm": 0.9626255035400391,
+      "learning_rate": 2.670930154846463e-07,
+      "loss": 0.27,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9287641662169455,
+      "grad_norm": 0.9918268918991089,
+      "learning_rate": 2.6508995951986526e-07,
+      "loss": 0.2593,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9290339989206692,
+      "grad_norm": 1.1631016731262207,
+      "learning_rate": 2.63094341886333e-07,
+      "loss": 0.2505,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9293038316243929,
+      "grad_norm": 1.2613874673843384,
+      "learning_rate": 2.6110616410887856e-07,
+      "loss": 0.2695,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9295736643281166,
+      "grad_norm": 1.2744042873382568,
+      "learning_rate": 2.591254277066457e-07,
+      "loss": 0.2326,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9298434970318402,
+      "grad_norm": 1.053614854812622,
+      "learning_rate": 2.571521341930894e-07,
+      "loss": 0.2342,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9301133297355639,
+      "grad_norm": 1.3415077924728394,
+      "learning_rate": 2.5518628507598454e-07,
+      "loss": 0.2305,
+      "step": 3447
+    },
+    {
+      "epoch": 0.9303831624392876,
+      "grad_norm": 1.05692720413208,
+      "learning_rate": 2.532278818574108e-07,
+      "loss": 0.2711,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9306529951430114,
+      "grad_norm": 0.8377360701560974,
+      "learning_rate": 2.5127692603376107e-07,
+      "loss": 0.2513,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9309228278467351,
+      "grad_norm": 1.4690423011779785,
+      "learning_rate": 2.493334190957386e-07,
+      "loss": 0.2289,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9311926605504587,
+      "grad_norm": 0.8457438945770264,
+      "learning_rate": 2.4739736252835545e-07,
+      "loss": 0.273,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9314624932541824,
+      "grad_norm": 1.561444878578186,
+      "learning_rate": 2.4546875781092714e-07,
+      "loss": 0.2149,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9317323259579061,
+      "grad_norm": 1.0397751331329346,
+      "learning_rate": 2.435476064170794e-07,
+      "loss": 0.2352,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9320021586616298,
+      "grad_norm": 1.3445332050323486,
+      "learning_rate": 2.4163390981474354e-07,
+      "loss": 0.2592,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9322719913653534,
+      "grad_norm": 0.9900785088539124,
+      "learning_rate": 2.39727669466151e-07,
+      "loss": 0.254,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9325418240690772,
+      "grad_norm": 1.3373383283615112,
+      "learning_rate": 2.3782888682783557e-07,
+      "loss": 0.2516,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9328116567728009,
+      "grad_norm": 1.1434911489486694,
+      "learning_rate": 2.3593756335064e-07,
+      "loss": 0.2621,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9330814894765246,
+      "grad_norm": 1.058342456817627,
+      "learning_rate": 2.3405370047969944e-07,
+      "loss": 0.2674,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9333513221802483,
+      "grad_norm": 1.0170992612838745,
+      "learning_rate": 2.3217729965445024e-07,
+      "loss": 0.2878,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9336211548839719,
+      "grad_norm": 0.7673280239105225,
+      "learning_rate": 2.3030836230863108e-07,
+      "loss": 0.23,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9338909875876956,
+      "grad_norm": 1.2992603778839111,
+      "learning_rate": 2.284468898702752e-07,
+      "loss": 0.2648,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9341608202914193,
+      "grad_norm": 0.8483427166938782,
+      "learning_rate": 2.265928837617082e-07,
+      "loss": 0.2397,
+      "step": 3462
+    },
+    {
+      "epoch": 0.934430652995143,
+      "grad_norm": 1.1370737552642822,
+      "learning_rate": 2.2474634539955907e-07,
+      "loss": 0.2416,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9347004856988667,
+      "grad_norm": 1.1507833003997803,
+      "learning_rate": 2.2290727619474594e-07,
+      "loss": 0.2471,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9349703184025904,
+      "grad_norm": 1.4040553569793701,
+      "learning_rate": 2.2107567755247584e-07,
+      "loss": 0.2445,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9352401511063141,
+      "grad_norm": 1.9227079153060913,
+      "learning_rate": 2.192515508722559e-07,
+      "loss": 0.2713,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9355099838100378,
+      "grad_norm": 1.0273640155792236,
+      "learning_rate": 2.1743489754788017e-07,
+      "loss": 0.2311,
+      "step": 3467
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 1.285233736038208,
+      "learning_rate": 2.1562571896742935e-07,
+      "loss": 0.2242,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9360496492174851,
+      "grad_norm": 1.434043526649475,
+      "learning_rate": 2.1382401651327767e-07,
+      "loss": 0.2436,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9363194819212088,
+      "grad_norm": 0.9344869256019592,
+      "learning_rate": 2.1202979156208504e-07,
+      "loss": 0.2351,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9365893146249326,
+      "grad_norm": 0.966553807258606,
+      "learning_rate": 2.1024304548479702e-07,
+      "loss": 0.235,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9368591473286563,
+      "grad_norm": 1.5438873767852783,
+      "learning_rate": 2.08463779646646e-07,
+      "loss": 0.2274,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9371289800323799,
+      "grad_norm": 1.0499929189682007,
+      "learning_rate": 2.066919954071489e-07,
+      "loss": 0.2138,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9373988127361036,
+      "grad_norm": 0.8389066457748413,
+      "learning_rate": 2.0492769412010284e-07,
+      "loss": 0.2155,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9376686454398273,
+      "grad_norm": 1.0444698333740234,
+      "learning_rate": 2.0317087713359273e-07,
+      "loss": 0.2273,
+      "step": 3475
+    },
+    {
+      "epoch": 0.937938478143551,
+      "grad_norm": 1.6911917924880981,
+      "learning_rate": 2.0142154578998153e-07,
+      "loss": 0.2466,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9382083108472747,
+      "grad_norm": 0.9807485342025757,
+      "learning_rate": 1.996797014259111e-07,
+      "loss": 0.2304,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9384781435509983,
+      "grad_norm": 0.9428440928459167,
+      "learning_rate": 1.979453453723057e-07,
+      "loss": 0.2633,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9387479762547221,
+      "grad_norm": 0.9982620477676392,
+      "learning_rate": 1.962184789543675e-07,
+      "loss": 0.2253,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9390178089584458,
+      "grad_norm": 1.6853564977645874,
+      "learning_rate": 1.9449910349157219e-07,
+      "loss": 0.2529,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9392876416621695,
+      "grad_norm": 1.0075082778930664,
+      "learning_rate": 1.927872202976766e-07,
+      "loss": 0.2104,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9395574743658931,
+      "grad_norm": 1.1361993551254272,
+      "learning_rate": 1.9108283068071221e-07,
+      "loss": 0.2197,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9398273070696168,
+      "grad_norm": 1.1917293071746826,
+      "learning_rate": 1.893859359429795e-07,
+      "loss": 0.2428,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9400971397733405,
+      "grad_norm": 0.9808080792427063,
+      "learning_rate": 1.8769653738105797e-07,
+      "loss": 0.2373,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9403669724770642,
+      "grad_norm": 0.8490698933601379,
+      "learning_rate": 1.860146362857984e-07,
+      "loss": 0.2244,
+      "step": 3485
+    },
+    {
+      "epoch": 0.940636805180788,
+      "grad_norm": 0.8340795040130615,
+      "learning_rate": 1.843402339423195e-07,
+      "loss": 0.2256,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9409066378845116,
+      "grad_norm": 0.9318832755088806,
+      "learning_rate": 1.8267333163001444e-07,
+      "loss": 0.2555,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 1.0371506214141846,
+      "learning_rate": 1.8101393062254335e-07,
+      "loss": 0.2407,
+      "step": 3488
+    },
+    {
+      "epoch": 0.941446303291959,
+      "grad_norm": 1.1280317306518555,
+      "learning_rate": 1.7936203218783421e-07,
+      "loss": 0.2308,
+      "step": 3489
+    },
+    {
+      "epoch": 0.9417161359956827,
+      "grad_norm": 0.9925305843353271,
+      "learning_rate": 1.7771763758808403e-07,
+      "loss": 0.258,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9419859686994063,
+      "grad_norm": 0.9233106374740601,
+      "learning_rate": 1.760807480797555e-07,
+      "loss": 0.2492,
+      "step": 3491
+    },
+    {
+      "epoch": 0.94225580140313,
+      "grad_norm": 1.1781879663467407,
+      "learning_rate": 1.7445136491357816e-07,
+      "loss": 0.2544,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9425256341068537,
+      "grad_norm": 0.8746750354766846,
+      "learning_rate": 1.728294893345428e-07,
+      "loss": 0.2411,
+      "step": 3493
+    },
+    {
+      "epoch": 0.9427954668105775,
+      "grad_norm": 0.9626951217651367,
+      "learning_rate": 1.7121512258190587e-07,
+      "loss": 0.2337,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9430652995143012,
+      "grad_norm": 1.416587471961975,
+      "learning_rate": 1.6960826588918732e-07,
+      "loss": 0.2448,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9433351322180248,
+      "grad_norm": 1.2790323495864868,
+      "learning_rate": 1.6800892048416618e-07,
+      "loss": 0.2468,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9436049649217485,
+      "grad_norm": 1.0324482917785645,
+      "learning_rate": 1.6641708758888486e-07,
+      "loss": 0.2423,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9438747976254722,
+      "grad_norm": 1.0843931436538696,
+      "learning_rate": 1.648327684196438e-07,
+      "loss": 0.2489,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9441446303291959,
+      "grad_norm": 1.472177267074585,
+      "learning_rate": 1.632559641870024e-07,
+      "loss": 0.2598,
+      "step": 3499
+    },
+    {
+      "epoch": 0.9444144630329195,
+      "grad_norm": 0.9559295773506165,
+      "learning_rate": 1.6168667609577915e-07,
+      "loss": 0.255,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9446842957366433,
+      "grad_norm": 1.214861512184143,
+      "learning_rate": 1.601249053450482e-07,
+      "loss": 0.2492,
+      "step": 3501
+    },
+    {
+      "epoch": 0.944954128440367,
+      "grad_norm": 1.5962530374526978,
+      "learning_rate": 1.5857065312814058e-07,
+      "loss": 0.2824,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9452239611440907,
+      "grad_norm": 0.9132072329521179,
+      "learning_rate": 1.5702392063264294e-07,
+      "loss": 0.241,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9454937938478144,
+      "grad_norm": 1.0570907592773438,
+      "learning_rate": 1.5548470904039436e-07,
+      "loss": 0.2469,
+      "step": 3504
+    },
+    {
+      "epoch": 0.945763626551538,
+      "grad_norm": 1.5786651372909546,
+      "learning_rate": 1.5395301952748854e-07,
+      "loss": 0.2409,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9460334592552617,
+      "grad_norm": 1.4660522937774658,
+      "learning_rate": 1.524288532642726e-07,
+      "loss": 0.2864,
+      "step": 3506
+    },
+    {
+      "epoch": 0.9463032919589854,
+      "grad_norm": 0.9599338173866272,
+      "learning_rate": 1.5091221141534384e-07,
+      "loss": 0.2643,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9465731246627092,
+      "grad_norm": 1.0511517524719238,
+      "learning_rate": 1.4940309513955088e-07,
+      "loss": 0.2715,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9468429573664328,
+      "grad_norm": 1.6058752536773682,
+      "learning_rate": 1.4790150558999016e-07,
+      "loss": 0.2304,
+      "step": 3509
+    },
+    {
+      "epoch": 0.9471127900701565,
+      "grad_norm": 0.9196997284889221,
+      "learning_rate": 1.4640744391401064e-07,
+      "loss": 0.2515,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9473826227738802,
+      "grad_norm": 0.9891975522041321,
+      "learning_rate": 1.4492091125320684e-07,
+      "loss": 0.2712,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9476524554776039,
+      "grad_norm": 1.164265513420105,
+      "learning_rate": 1.4344190874342134e-07,
+      "loss": 0.2452,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9479222881813276,
+      "grad_norm": 1.042662262916565,
+      "learning_rate": 1.419704375147424e-07,
+      "loss": 0.2921,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9481921208850512,
+      "grad_norm": 1.165993332862854,
+      "learning_rate": 1.405064986915028e-07,
+      "loss": 0.2421,
+      "step": 3514
+    },
+    {
+      "epoch": 0.9484619535887749,
+      "grad_norm": 0.9909040331840515,
+      "learning_rate": 1.3905009339228336e-07,
+      "loss": 0.2781,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9487317862924987,
+      "grad_norm": 0.9671472311019897,
+      "learning_rate": 1.3760122272990505e-07,
+      "loss": 0.2479,
+      "step": 3516
+    },
+    {
+      "epoch": 0.9490016189962224,
+      "grad_norm": 0.900456964969635,
+      "learning_rate": 1.361598878114334e-07,
+      "loss": 0.2451,
+      "step": 3517
+    },
+    {
+      "epoch": 0.949271451699946,
+      "grad_norm": 1.2012739181518555,
+      "learning_rate": 1.3472608973817525e-07,
+      "loss": 0.2289,
+      "step": 3518
+    },
+    {
+      "epoch": 0.9495412844036697,
+      "grad_norm": 1.2713685035705566,
+      "learning_rate": 1.3329982960567977e-07,
+      "loss": 0.2334,
+      "step": 3519
+    },
+    {
+      "epoch": 0.9498111171073934,
+      "grad_norm": 1.0354576110839844,
+      "learning_rate": 1.3188110850373527e-07,
+      "loss": 0.2226,
+      "step": 3520
+    },
+    {
+      "epoch": 0.9500809498111171,
+      "grad_norm": 0.9257195591926575,
+      "learning_rate": 1.304699275163701e-07,
+      "loss": 0.2422,
+      "step": 3521
+    },
+    {
+      "epoch": 0.9503507825148408,
+      "grad_norm": 1.0016839504241943,
+      "learning_rate": 1.290662877218507e-07,
+      "loss": 0.2585,
+      "step": 3522
+    },
+    {
+      "epoch": 0.9506206152185644,
+      "grad_norm": 1.2661397457122803,
+      "learning_rate": 1.2767019019268134e-07,
+      "loss": 0.2338,
+      "step": 3523
+    },
+    {
+      "epoch": 0.9508904479222882,
+      "grad_norm": 0.919041633605957,
+      "learning_rate": 1.2628163599560538e-07,
+      "loss": 0.2797,
+      "step": 3524
+    },
+    {
+      "epoch": 0.9511602806260119,
+      "grad_norm": 1.0429943799972534,
+      "learning_rate": 1.2490062619160082e-07,
+      "loss": 0.2868,
+      "step": 3525
+    },
+    {
+      "epoch": 0.9514301133297356,
+      "grad_norm": 0.886855959892273,
+      "learning_rate": 1.2352716183588022e-07,
+      "loss": 0.2174,
+      "step": 3526
+    },
+    {
+      "epoch": 0.9516999460334593,
+      "grad_norm": 1.5107280015945435,
+      "learning_rate": 1.2216124397789187e-07,
+      "loss": 0.2654,
+      "step": 3527
+    },
+    {
+      "epoch": 0.9519697787371829,
+      "grad_norm": 0.8899127840995789,
+      "learning_rate": 1.208028736613187e-07,
+      "loss": 0.2598,
+      "step": 3528
+    },
+    {
+      "epoch": 0.9522396114409066,
+      "grad_norm": 1.813210368156433,
+      "learning_rate": 1.1945205192407383e-07,
+      "loss": 0.23,
+      "step": 3529
+    },
+    {
+      "epoch": 0.9525094441446303,
+      "grad_norm": 1.5451568365097046,
+      "learning_rate": 1.1810877979830492e-07,
+      "loss": 0.2635,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9527792768483541,
+      "grad_norm": 1.5778871774673462,
+      "learning_rate": 1.167730583103921e-07,
+      "loss": 0.2365,
+      "step": 3531
+    },
+    {
+      "epoch": 0.9530491095520777,
+      "grad_norm": 1.1266709566116333,
+      "learning_rate": 1.1544488848094338e-07,
+      "loss": 0.2555,
+      "step": 3532
+    },
+    {
+      "epoch": 0.9533189422558014,
+      "grad_norm": 1.0937931537628174,
+      "learning_rate": 1.1412427132479698e-07,
+      "loss": 0.2249,
+      "step": 3533
+    },
+    {
+      "epoch": 0.9535887749595251,
+      "grad_norm": 0.9257707595825195,
+      "learning_rate": 1.1281120785102239e-07,
+      "loss": 0.2396,
+      "step": 3534
+    },
+    {
+      "epoch": 0.9538586076632488,
+      "grad_norm": 1.661321997642517,
+      "learning_rate": 1.1150569906291598e-07,
+      "loss": 0.2878,
+      "step": 3535
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 1.0626643896102905,
+      "learning_rate": 1.1020774595800198e-07,
+      "loss": 0.2232,
+      "step": 3536
+    },
+    {
+      "epoch": 0.9543982730706961,
+      "grad_norm": 0.9862203598022461,
+      "learning_rate": 1.0891734952802934e-07,
+      "loss": 0.232,
+      "step": 3537
+    },
+    {
+      "epoch": 0.9546681057744199,
+      "grad_norm": 0.8893401026725769,
+      "learning_rate": 1.0763451075897713e-07,
+      "loss": 0.2463,
+      "step": 3538
+    },
+    {
+      "epoch": 0.9549379384781436,
+      "grad_norm": 1.04206120967865,
+      "learning_rate": 1.063592306310468e-07,
+      "loss": 0.2463,
+      "step": 3539
+    },
+    {
+      "epoch": 0.9552077711818673,
+      "grad_norm": 0.9988157153129578,
+      "learning_rate": 1.0509151011866558e-07,
+      "loss": 0.2246,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9554776038855909,
+      "grad_norm": 0.9415999054908752,
+      "learning_rate": 1.0383135019048196e-07,
+      "loss": 0.2577,
+      "step": 3541
+    },
+    {
+      "epoch": 0.9557474365893146,
+      "grad_norm": 1.2408337593078613,
+      "learning_rate": 1.0257875180937127e-07,
+      "loss": 0.2561,
+      "step": 3542
+    },
+    {
+      "epoch": 0.9560172692930383,
+      "grad_norm": 1.147184133529663,
+      "learning_rate": 1.0133371593242902e-07,
+      "loss": 0.2328,
+      "step": 3543
+    },
+    {
+      "epoch": 0.956287101996762,
+      "grad_norm": 1.0100079774856567,
+      "learning_rate": 1.0009624351097313e-07,
+      "loss": 0.2833,
+      "step": 3544
+    },
+    {
+      "epoch": 0.9565569347004858,
+      "grad_norm": 1.1013966798782349,
+      "learning_rate": 9.88663354905406e-08,
+      "loss": 0.2035,
+      "step": 3545
+    },
+    {
+      "epoch": 0.9568267674042094,
+      "grad_norm": 1.0230672359466553,
+      "learning_rate": 9.764399281088965e-08,
+      "loss": 0.2247,
+      "step": 3546
+    },
+    {
+      "epoch": 0.9570966001079331,
+      "grad_norm": 0.9613625407218933,
+      "learning_rate": 9.642921640599879e-08,
+      "loss": 0.2596,
+      "step": 3547
+    },
+    {
+      "epoch": 0.9573664328116568,
+      "grad_norm": 1.2595155239105225,
+      "learning_rate": 9.522200720406549e-08,
+      "loss": 0.2626,
+      "step": 3548
+    },
+    {
+      "epoch": 0.9576362655153805,
+      "grad_norm": 0.9021123051643372,
+      "learning_rate": 9.402236612750305e-08,
+      "loss": 0.2493,
+      "step": 3549
+    },
+    {
+      "epoch": 0.9579060982191041,
+      "grad_norm": 0.981185793876648,
+      "learning_rate": 9.283029409294263e-08,
+      "loss": 0.2227,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9581759309228278,
+      "grad_norm": 0.8714508414268494,
+      "learning_rate": 9.164579201123347e-08,
+      "loss": 0.2497,
+      "step": 3551
+    },
+    {
+      "epoch": 0.9584457636265515,
+      "grad_norm": 0.9475069642066956,
+      "learning_rate": 9.046886078744044e-08,
+      "loss": 0.2447,
+      "step": 3552
+    },
+    {
+      "epoch": 0.9587155963302753,
+      "grad_norm": 0.9200780987739563,
+      "learning_rate": 8.929950132084086e-08,
+      "loss": 0.2958,
+      "step": 3553
+    },
+    {
+      "epoch": 0.958985429033999,
+      "grad_norm": 1.234106421470642,
+      "learning_rate": 8.813771450493114e-08,
+      "loss": 0.2383,
+      "step": 3554
+    },
+    {
+      "epoch": 0.9592552617377226,
+      "grad_norm": 1.5173259973526,
+      "learning_rate": 8.698350122741783e-08,
+      "loss": 0.2564,
+      "step": 3555
+    },
+    {
+      "epoch": 0.9595250944414463,
+      "grad_norm": 0.8406164646148682,
+      "learning_rate": 8.583686237022105e-08,
+      "loss": 0.2359,
+      "step": 3556
+    },
+    {
+      "epoch": 0.95979492714517,
+      "grad_norm": 1.924187421798706,
+      "learning_rate": 8.46977988094766e-08,
+      "loss": 0.2875,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9600647598488937,
+      "grad_norm": 1.035154938697815,
+      "learning_rate": 8.356631141552607e-08,
+      "loss": 0.268,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9603345925526173,
+      "grad_norm": 1.7517836093902588,
+      "learning_rate": 8.2442401052929e-08,
+      "loss": 0.2665,
+      "step": 3559
+    },
+    {
+      "epoch": 0.960604425256341,
+      "grad_norm": 1.2423560619354248,
+      "learning_rate": 8.132606858045067e-08,
+      "loss": 0.2606,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9608742579600648,
+      "grad_norm": 1.04481041431427,
+      "learning_rate": 8.021731485106877e-08,
+      "loss": 0.2234,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9611440906637885,
+      "grad_norm": 1.1929190158843994,
+      "learning_rate": 7.911614071196671e-08,
+      "loss": 0.2579,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9614139233675122,
+      "grad_norm": 0.9982279539108276,
+      "learning_rate": 7.802254700454259e-08,
+      "loss": 0.2649,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9616837560712358,
+      "grad_norm": 1.1076874732971191,
+      "learning_rate": 7.693653456439576e-08,
+      "loss": 0.2536,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9619535887749595,
+      "grad_norm": 1.132493257522583,
+      "learning_rate": 7.585810422133466e-08,
+      "loss": 0.2815,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9622234214786832,
+      "grad_norm": 1.0886120796203613,
+      "learning_rate": 7.478725679937904e-08,
+      "loss": 0.2516,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9624932541824069,
+      "grad_norm": 0.9007965922355652,
+      "learning_rate": 7.372399311674882e-08,
+      "loss": 0.2477,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9627630868861305,
+      "grad_norm": 1.0126781463623047,
+      "learning_rate": 7.266831398587082e-08,
+      "loss": 0.2733,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9630329195898543,
+      "grad_norm": 1.0329698324203491,
+      "learning_rate": 7.162022021337867e-08,
+      "loss": 0.2542,
+      "step": 3569
+    },
+    {
+      "epoch": 0.963302752293578,
+      "grad_norm": 0.8857415914535522,
+      "learning_rate": 7.057971260010732e-08,
+      "loss": 0.2219,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9635725849973017,
+      "grad_norm": 0.8997820019721985,
+      "learning_rate": 6.954679194109637e-08,
+      "loss": 0.2387,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9638424177010254,
+      "grad_norm": 1.258575677871704,
+      "learning_rate": 6.852145902559004e-08,
+      "loss": 0.2642,
+      "step": 3572
+    },
+    {
+      "epoch": 0.964112250404749,
+      "grad_norm": 0.9814557433128357,
+      "learning_rate": 6.750371463703164e-08,
+      "loss": 0.2727,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9643820831084727,
+      "grad_norm": 1.0974315404891968,
+      "learning_rate": 6.649355955306802e-08,
+      "loss": 0.2293,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9646519158121964,
+      "grad_norm": 1.1178288459777832,
+      "learning_rate": 6.549099454554841e-08,
+      "loss": 0.2001,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9649217485159202,
+      "grad_norm": 1.0497643947601318,
+      "learning_rate": 6.449602038051894e-08,
+      "loss": 0.2527,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9651915812196438,
+      "grad_norm": 1.3208602666854858,
+      "learning_rate": 6.350863781822813e-08,
+      "loss": 0.2687,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9654614139233675,
+      "grad_norm": 0.9483137726783752,
+      "learning_rate": 6.252884761312584e-08,
+      "loss": 0.2059,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9657312466270912,
+      "grad_norm": 1.6079738140106201,
+      "learning_rate": 6.155665051385651e-08,
+      "loss": 0.2602,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9660010793308149,
+      "grad_norm": 0.9147442579269409,
+      "learning_rate": 6.059204726326373e-08,
+      "loss": 0.2436,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9662709120345386,
+      "grad_norm": 0.9747893810272217,
+      "learning_rate": 5.963503859839125e-08,
+      "loss": 0.2569,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9665407447382622,
+      "grad_norm": 1.208474040031433,
+      "learning_rate": 5.868562525047639e-08,
+      "loss": 0.2592,
+      "step": 3582
+    },
+    {
+      "epoch": 0.966810577441986,
+      "grad_norm": 0.8668001890182495,
+      "learning_rate": 5.774380794495771e-08,
+      "loss": 0.2299,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9670804101457097,
+      "grad_norm": 0.909814715385437,
+      "learning_rate": 5.6809587401464025e-08,
+      "loss": 0.2603,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9673502428494334,
+      "grad_norm": 1.062694787979126,
+      "learning_rate": 5.588296433382323e-08,
+      "loss": 0.2217,
+      "step": 3585
+    },
+    {
+      "epoch": 0.967620075553157,
+      "grad_norm": 0.95551997423172,
+      "learning_rate": 5.4963939450057846e-08,
+      "loss": 0.2233,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9678899082568807,
+      "grad_norm": 0.9697267413139343,
+      "learning_rate": 5.4052513452383933e-08,
+      "loss": 0.2528,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9681597409606044,
+      "grad_norm": 1.0810906887054443,
+      "learning_rate": 5.314868703720999e-08,
+      "loss": 0.2611,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9684295736643281,
+      "grad_norm": 1.4063448905944824,
+      "learning_rate": 5.225246089514135e-08,
+      "loss": 0.2799,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9686994063680519,
+      "grad_norm": 1.139684796333313,
+      "learning_rate": 5.1363835710972476e-08,
+      "loss": 0.268,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9689692390717755,
+      "grad_norm": 1.0735059976577759,
+      "learning_rate": 5.048281216369355e-08,
+      "loss": 0.2532,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9692390717754992,
+      "grad_norm": 1.1969654560089111,
+      "learning_rate": 4.960939092648165e-08,
+      "loss": 0.2632,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9695089044792229,
+      "grad_norm": 1.0117276906967163,
+      "learning_rate": 4.8743572666708486e-08,
+      "loss": 0.2455,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9697787371829466,
+      "grad_norm": 1.0563112497329712,
+      "learning_rate": 4.788535804593708e-08,
+      "loss": 0.2243,
+      "step": 3594
+    },
+    {
+      "epoch": 0.9700485698866703,
+      "grad_norm": 0.8439617156982422,
+      "learning_rate": 4.703474771991845e-08,
+      "loss": 0.2497,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9703184025903939,
+      "grad_norm": 0.8904944658279419,
+      "learning_rate": 4.61917423385938e-08,
+      "loss": 0.2238,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9705882352941176,
+      "grad_norm": 0.957848072052002,
+      "learning_rate": 4.5356342546093446e-08,
+      "loss": 0.2749,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9708580679978414,
+      "grad_norm": 1.1660819053649902,
+      "learning_rate": 4.452854898073788e-08,
+      "loss": 0.2265,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9711279007015651,
+      "grad_norm": 1.002392292022705,
+      "learning_rate": 4.370836227503228e-08,
+      "loss": 0.2611,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9713977334052887,
+      "grad_norm": 1.8921759128570557,
+      "learning_rate": 4.2895783055674214e-08,
+      "loss": 0.2224,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9716675661090124,
+      "grad_norm": 1.0716564655303955,
+      "learning_rate": 4.2090811943544807e-08,
+      "loss": 0.2314,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9719373988127361,
+      "grad_norm": 0.939160943031311,
+      "learning_rate": 4.129344955371317e-08,
+      "loss": 0.2408,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9722072315164598,
+      "grad_norm": 1.5893704891204834,
+      "learning_rate": 4.050369649543529e-08,
+      "loss": 0.2347,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 1.098116397857666,
+      "learning_rate": 3.9721553372150665e-08,
+      "loss": 0.2349,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9727468969239071,
+      "grad_norm": 0.8978264331817627,
+      "learning_rate": 3.894702078148571e-08,
+      "loss": 0.231,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9730167296276309,
+      "grad_norm": 1.2091326713562012,
+      "learning_rate": 3.8180099315254794e-08,
+      "loss": 0.2592,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9732865623313546,
+      "grad_norm": 0.9311931133270264,
+      "learning_rate": 3.742078955945028e-08,
+      "loss": 0.2309,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9735563950350783,
+      "grad_norm": 0.870396077632904,
+      "learning_rate": 3.666909209425362e-08,
+      "loss": 0.2612,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9738262277388019,
+      "grad_norm": 1.1579692363739014,
+      "learning_rate": 3.592500749402761e-08,
+      "loss": 0.2814,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9740960604425256,
+      "grad_norm": 1.1642515659332275,
+      "learning_rate": 3.5188536327318554e-08,
+      "loss": 0.2627,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9743658931462493,
+      "grad_norm": 0.9623153209686279,
+      "learning_rate": 3.4459679156856326e-08,
+      "loss": 0.2593,
+      "step": 3611
+    },
+    {
+      "epoch": 0.974635725849973,
+      "grad_norm": 0.9747045636177063,
+      "learning_rate": 3.3738436539551e-08,
+      "loss": 0.2807,
+      "step": 3612
+    },
+    {
+      "epoch": 0.9749055585536968,
+      "grad_norm": 1.0174964666366577,
+      "learning_rate": 3.3024809026496185e-08,
+      "loss": 0.2566,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9751753912574204,
+      "grad_norm": 0.9386301636695862,
+      "learning_rate": 3.231879716296571e-08,
+      "loss": 0.2545,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9754452239611441,
+      "grad_norm": 1.1665081977844238,
+      "learning_rate": 3.1620401488415834e-08,
+      "loss": 0.2524,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9757150566648678,
+      "grad_norm": 0.862600564956665,
+      "learning_rate": 3.092962253648302e-08,
+      "loss": 0.2361,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9759848893685915,
+      "grad_norm": 1.375880241394043,
+      "learning_rate": 3.0246460834981726e-08,
+      "loss": 0.2607,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9762547220723151,
+      "grad_norm": 1.0406776666641235,
+      "learning_rate": 2.9570916905908854e-08,
+      "loss": 0.2353,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9765245547760388,
+      "grad_norm": 1.1515228748321533,
+      "learning_rate": 2.8902991265440384e-08,
+      "loss": 0.2926,
+      "step": 3619
+    },
+    {
+      "epoch": 0.9767943874797625,
+      "grad_norm": 2.6442675590515137,
+      "learning_rate": 2.824268442392919e-08,
+      "loss": 0.2579,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9770642201834863,
+      "grad_norm": 1.1480995416641235,
+      "learning_rate": 2.7589996885908354e-08,
+      "loss": 0.2682,
+      "step": 3621
+    },
+    {
+      "epoch": 0.97733405288721,
+      "grad_norm": 0.9370139241218567,
+      "learning_rate": 2.694492915009006e-08,
+      "loss": 0.3089,
+      "step": 3622
+    },
+    {
+      "epoch": 0.9776038855909336,
+      "grad_norm": 1.5124880075454712,
+      "learning_rate": 2.6307481709361148e-08,
+      "loss": 0.2539,
+      "step": 3623
+    },
+    {
+      "epoch": 0.9778737182946573,
+      "grad_norm": 0.9607952833175659,
+      "learning_rate": 2.5677655050788674e-08,
+      "loss": 0.2384,
+      "step": 3624
+    },
+    {
+      "epoch": 0.978143550998381,
+      "grad_norm": 1.638617992401123,
+      "learning_rate": 2.505544965561546e-08,
+      "loss": 0.2562,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9784133837021047,
+      "grad_norm": 1.4915692806243896,
+      "learning_rate": 2.444086599926232e-08,
+      "loss": 0.2234,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9786832164058283,
+      "grad_norm": 1.0803436040878296,
+      "learning_rate": 2.383390455132473e-08,
+      "loss": 0.2986,
+      "step": 3627
+    },
+    {
+      "epoch": 0.978953049109552,
+      "grad_norm": 1.4764043092727661,
+      "learning_rate": 2.3234565775575034e-08,
+      "loss": 0.2225,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9792228818132758,
+      "grad_norm": 0.9830582141876221,
+      "learning_rate": 2.264285012996026e-08,
+      "loss": 0.2234,
+      "step": 3629
+    },
+    {
+      "epoch": 0.9794927145169995,
+      "grad_norm": 1.1383365392684937,
+      "learning_rate": 2.205875806660429e-08,
+      "loss": 0.2616,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9797625472207232,
+      "grad_norm": 1.0615745782852173,
+      "learning_rate": 2.1482290031805687e-08,
+      "loss": 0.2618,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9800323799244468,
+      "grad_norm": 0.9715497493743896,
+      "learning_rate": 2.0913446466035436e-08,
+      "loss": 0.2516,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9803022126281705,
+      "grad_norm": 1.0376906394958496,
+      "learning_rate": 2.0352227803941415e-08,
+      "loss": 0.2451,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9805720453318942,
+      "grad_norm": 0.9235926866531372,
+      "learning_rate": 1.9798634474345048e-08,
+      "loss": 0.2648,
+      "step": 3634
+    },
+    {
+      "epoch": 0.980841878035618,
+      "grad_norm": 1.0748919248580933,
+      "learning_rate": 1.925266690023686e-08,
+      "loss": 0.201,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9811117107393416,
+      "grad_norm": 0.7897247076034546,
+      "learning_rate": 1.871432549878871e-08,
+      "loss": 0.2228,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9813815434430653,
+      "grad_norm": 1.0417091846466064,
+      "learning_rate": 1.8183610681338226e-08,
+      "loss": 0.2807,
+      "step": 3637
+    },
+    {
+      "epoch": 0.981651376146789,
+      "grad_norm": 0.9404270052909851,
+      "learning_rate": 1.766052285339992e-08,
+      "loss": 0.222,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9819212088505127,
+      "grad_norm": 1.038517713546753,
+      "learning_rate": 1.714506241465741e-08,
+      "loss": 0.2351,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9821910415542364,
+      "grad_norm": 1.4498341083526611,
+      "learning_rate": 1.6637229758970087e-08,
+      "loss": 0.224,
+      "step": 3640
+    },
+    {
+      "epoch": 0.98246087425796,
+      "grad_norm": 0.9809703826904297,
+      "learning_rate": 1.6137025274365336e-08,
+      "loss": 0.2488,
+      "step": 3641
+    },
+    {
+      "epoch": 0.9827307069616837,
+      "grad_norm": 1.4311909675598145,
+      "learning_rate": 1.56444493430441e-08,
+      "loss": 0.2187,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9830005396654075,
+      "grad_norm": 1.012400507926941,
+      "learning_rate": 1.515950234137753e-08,
+      "loss": 0.2876,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9832703723691312,
+      "grad_norm": 0.8363939523696899,
+      "learning_rate": 1.4682184639909224e-08,
+      "loss": 0.2083,
+      "step": 3644
+    },
+    {
+      "epoch": 0.9835402050728548,
+      "grad_norm": 1.0199151039123535,
+      "learning_rate": 1.4212496603350778e-08,
+      "loss": 0.2686,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9838100377765785,
+      "grad_norm": 1.261934757232666,
+      "learning_rate": 1.3750438590586223e-08,
+      "loss": 0.2569,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9840798704803022,
+      "grad_norm": 0.8835524916648865,
+      "learning_rate": 1.3296010954667593e-08,
+      "loss": 0.2008,
+      "step": 3647
+    },
+    {
+      "epoch": 0.9843497031840259,
+      "grad_norm": 0.9338524341583252,
+      "learning_rate": 1.2849214042819359e-08,
+      "loss": 0.268,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9846195358877496,
+      "grad_norm": 1.8936468362808228,
+      "learning_rate": 1.2410048196432878e-08,
+      "loss": 0.2485,
+      "step": 3649
+    },
+    {
+      "epoch": 0.9848893685914732,
+      "grad_norm": 0.9044299721717834,
+      "learning_rate": 1.197851375106862e-08,
+      "loss": 0.2377,
+      "step": 3650
+    },
+    {
+      "epoch": 0.985159201295197,
+      "grad_norm": 1.0701262950897217,
+      "learning_rate": 1.1554611036458386e-08,
+      "loss": 0.2349,
+      "step": 3651
+    },
+    {
+      "epoch": 0.9854290339989207,
+      "grad_norm": 1.0875827074050903,
+      "learning_rate": 1.1138340376501966e-08,
+      "loss": 0.2456,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9856988667026444,
+      "grad_norm": 1.3053686618804932,
+      "learning_rate": 1.0729702089264938e-08,
+      "loss": 0.2375,
+      "step": 3653
+    },
+    {
+      "epoch": 0.985968699406368,
+      "grad_norm": 1.035393238067627,
+      "learning_rate": 1.0328696486984202e-08,
+      "loss": 0.2566,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9862385321100917,
+      "grad_norm": 0.9066647291183472,
+      "learning_rate": 9.935323876063552e-09,
+      "loss": 0.2295,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9865083648138154,
+      "grad_norm": 0.9023276567459106,
+      "learning_rate": 9.549584557074775e-09,
+      "loss": 0.2505,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9867781975175391,
+      "grad_norm": 1.095625400543213,
+      "learning_rate": 9.171478824755441e-09,
+      "loss": 0.2667,
+      "step": 3657
+    },
+    {
+      "epoch": 0.9870480302212629,
+      "grad_norm": 1.1310292482376099,
+      "learning_rate": 8.801006968012227e-09,
+      "loss": 0.2387,
+      "step": 3658
+    },
+    {
+      "epoch": 0.9873178629249865,
+      "grad_norm": 0.8289709091186523,
+      "learning_rate": 8.438169269918694e-09,
+      "loss": 0.2247,
+      "step": 3659
+    },
+    {
+      "epoch": 0.9875876956287102,
+      "grad_norm": 1.5531152486801147,
+      "learning_rate": 8.082966007715298e-09,
+      "loss": 0.2459,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9878575283324339,
+      "grad_norm": 1.0354291200637817,
+      "learning_rate": 7.73539745280827e-09,
+      "loss": 0.2653,
+      "step": 3661
+    },
+    {
+      "epoch": 0.9881273610361576,
+      "grad_norm": 0.9499836564064026,
+      "learning_rate": 7.395463870770725e-09,
+      "loss": 0.2536,
+      "step": 3662
+    },
+    {
+      "epoch": 0.9883971937398812,
+      "grad_norm": 1.436198353767395,
+      "learning_rate": 7.063165521341564e-09,
+      "loss": 0.2408,
+      "step": 3663
+    },
+    {
+      "epoch": 0.9886670264436049,
+      "grad_norm": 1.3272536993026733,
+      "learning_rate": 6.738502658426571e-09,
+      "loss": 0.2695,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9889368591473287,
+      "grad_norm": 1.108034610748291,
+      "learning_rate": 6.421475530096199e-09,
+      "loss": 0.2071,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9892066918510524,
+      "grad_norm": 1.5205875635147095,
+      "learning_rate": 6.1120843785877904e-09,
+      "loss": 0.2321,
+      "step": 3666
+    },
+    {
+      "epoch": 0.9894765245547761,
+      "grad_norm": 0.8838106989860535,
+      "learning_rate": 5.8103294403033525e-09,
+      "loss": 0.2219,
+      "step": 3667
+    },
+    {
+      "epoch": 0.9897463572584997,
+      "grad_norm": 1.0022790431976318,
+      "learning_rate": 5.516210945810674e-09,
+      "loss": 0.2306,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9900161899622234,
+      "grad_norm": 1.1648350954055786,
+      "learning_rate": 5.229729119842208e-09,
+      "loss": 0.2387,
+      "step": 3669
+    },
+    {
+      "epoch": 0.9902860226659471,
+      "grad_norm": 1.02855384349823,
+      "learning_rate": 4.950884181295079e-09,
+      "loss": 0.2721,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9905558553696708,
+      "grad_norm": 1.1199991703033447,
+      "learning_rate": 4.679676343231077e-09,
+      "loss": 0.2293,
+      "step": 3671
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 1.168959617614746,
+      "learning_rate": 4.416105812877769e-09,
+      "loss": 0.2532,
+      "step": 3672
+    },
+    {
+      "epoch": 0.9910955207771182,
+      "grad_norm": 1.047879934310913,
+      "learning_rate": 4.160172791625172e-09,
+      "loss": 0.236,
+      "step": 3673
+    },
+    {
+      "epoch": 0.9913653534808419,
+      "grad_norm": 1.1724107265472412,
+      "learning_rate": 3.91187747503019e-09,
+      "loss": 0.2282,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9916351861845656,
+      "grad_norm": 1.0215120315551758,
+      "learning_rate": 3.6712200528110642e-09,
+      "loss": 0.2545,
+      "step": 3675
+    },
+    {
+      "epoch": 0.9919050188882893,
+      "grad_norm": 1.6420207023620605,
+      "learning_rate": 3.4382007088518134e-09,
+      "loss": 0.2295,
+      "step": 3676
+    },
+    {
+      "epoch": 0.9921748515920129,
+      "grad_norm": 0.9902812838554382,
+      "learning_rate": 3.212819621201124e-09,
+      "loss": 0.2722,
+      "step": 3677
+    },
+    {
+      "epoch": 0.9924446842957366,
+      "grad_norm": 1.1094350814819336,
+      "learning_rate": 2.995076962067911e-09,
+      "loss": 0.2342,
+      "step": 3678
+    },
+    {
+      "epoch": 0.9927145169994603,
+      "grad_norm": 0.7946183085441589,
+      "learning_rate": 2.7849728978279753e-09,
+      "loss": 0.2381,
+      "step": 3679
+    },
+    {
+      "epoch": 0.992984349703184,
+      "grad_norm": 1.1771692037582397,
+      "learning_rate": 2.582507589018457e-09,
+      "loss": 0.2581,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9932541824069078,
+      "grad_norm": 1.1895432472229004,
+      "learning_rate": 2.3876811903422723e-09,
+      "loss": 0.2746,
+      "step": 3681
+    },
+    {
+      "epoch": 0.9935240151106314,
+      "grad_norm": 0.9848347902297974,
+      "learning_rate": 2.200493850662566e-09,
+      "loss": 0.2418,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9937938478143551,
+      "grad_norm": 1.1719189882278442,
+      "learning_rate": 2.0209457130071496e-09,
+      "loss": 0.198,
+      "step": 3683
+    },
+    {
+      "epoch": 0.9940636805180788,
+      "grad_norm": 1.4688509702682495,
+      "learning_rate": 1.8490369145685027e-09,
+      "loss": 0.238,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9943335132218025,
+      "grad_norm": 1.0212900638580322,
+      "learning_rate": 1.6847675866971114e-09,
+      "loss": 0.2037,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9946033459255261,
+      "grad_norm": 0.9387383460998535,
+      "learning_rate": 1.5281378549125702e-09,
+      "loss": 0.2527,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9948731786292498,
+      "grad_norm": 1.6200354099273682,
+      "learning_rate": 1.3791478388913704e-09,
+      "loss": 0.2798,
+      "step": 3687
+    },
+    {
+      "epoch": 0.9951430113329736,
+      "grad_norm": 0.8763278126716614,
+      "learning_rate": 1.2377976524746705e-09,
+      "loss": 0.2134,
+      "step": 3688
+    },
+    {
+      "epoch": 0.9954128440366973,
+      "grad_norm": 1.3627339601516724,
+      "learning_rate": 1.1040874036694072e-09,
+      "loss": 0.2638,
+      "step": 3689
+    },
+    {
+      "epoch": 0.995682676740421,
+      "grad_norm": 1.106690764427185,
+      "learning_rate": 9.780171946394135e-10,
+      "loss": 0.2476,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9959525094441446,
+      "grad_norm": 0.9294391870498657,
+      "learning_rate": 8.595871217143004e-10,
+      "loss": 0.2209,
+      "step": 3691
+    },
+    {
+      "epoch": 0.9962223421478683,
+      "grad_norm": 1.2955615520477295,
+      "learning_rate": 7.487972753850159e-10,
+      "loss": 0.266,
+      "step": 3692
+    },
+    {
+      "epoch": 0.996492174851592,
+      "grad_norm": 1.0936665534973145,
+      "learning_rate": 6.456477403049554e-10,
+      "loss": 0.2639,
+      "step": 3693
+    },
+    {
+      "epoch": 0.9967620075553157,
+      "grad_norm": 1.0047434568405151,
+      "learning_rate": 5.501385952888516e-10,
+      "loss": 0.2392,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9970318402590393,
+      "grad_norm": 0.8788974285125732,
+      "learning_rate": 4.622699133161046e-10,
+      "loss": 0.2338,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9973016729627631,
+      "grad_norm": 0.9008697867393494,
+      "learning_rate": 3.82041761523011e-10,
+      "loss": 0.2493,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9975715056664868,
+      "grad_norm": 1.1299192905426025,
+      "learning_rate": 3.0945420121386573e-10,
+      "loss": 0.2499,
+      "step": 3697
+    },
+    {
+      "epoch": 0.9978413383702105,
+      "grad_norm": 1.1186836957931519,
+      "learning_rate": 2.4450728785097023e-10,
+      "loss": 0.2338,
+      "step": 3698
+    },
+    {
+      "epoch": 0.9981111710739342,
+      "grad_norm": 1.0582714080810547,
+      "learning_rate": 1.8720107105796304e-10,
+      "loss": 0.2807,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9983810037776578,
+      "grad_norm": 0.9211650490760803,
+      "learning_rate": 1.375355946242607e-10,
+      "loss": 0.2531,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9986508364813815,
+      "grad_norm": 1.1477832794189453,
+      "learning_rate": 9.551089649728617e-11,
+      "loss": 0.2478,
+      "step": 3701
+    },
+    {
+      "epoch": 0.9989206691851052,
+      "grad_norm": 1.017176628112793,
+      "learning_rate": 6.112700878802003e-11,
+      "loss": 0.2599,
+      "step": 3702
+    },
+    {
+      "epoch": 0.999190501888829,
+      "grad_norm": 1.152467966079712,
+      "learning_rate": 3.4383957768779895e-11,
+      "loss": 0.2486,
+      "step": 3703
+    },
+    {
+      "epoch": 0.9994603345925526,
+      "grad_norm": 0.9113025069236755,
+      "learning_rate": 1.5281763873220556e-11,
+      "loss": 0.2861,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9997301672962763,
+      "grad_norm": 1.6283750534057617,
+      "learning_rate": 3.820441698554334e-12,
+      "loss": 0.2864,
+      "step": 3705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0231740474700928,
+      "learning_rate": 0.0,
+      "loss": 0.2314,
+      "step": 3706
+    },
+    {
+      "epoch": 1.0,
+      "step": 3706,
+      "total_flos": 4.432216601164513e+17,
+      "train_loss": 0.0,
+      "train_runtime": 0.0081,
+      "train_samples_per_second": 58634471.284,
+      "train_steps_per_second": 458095.327
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3706,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.432216601164513e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}