diff --git "a/checkpoints/checkpoint-2500/trainer_state.json" "b/checkpoints/checkpoint-2500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoints/checkpoint-2500/trainer_state.json"
@@ -0,0 +1,17533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.881422924901186,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003952569169960474,
+      "grad_norm": 10.89438533782959,
+      "learning_rate": 5e-06,
+      "loss": 4.1054,
+      "step": 1
+    },
+    {
+      "epoch": 0.007905138339920948,
+      "grad_norm": 10.555335998535156,
+      "learning_rate": 1e-05,
+      "loss": 4.1308,
+      "step": 2
+    },
+    {
+      "epoch": 0.011857707509881422,
+      "grad_norm": 10.703575134277344,
+      "learning_rate": 1.5e-05,
+      "loss": 4.3599,
+      "step": 3
+    },
+    {
+      "epoch": 0.015810276679841896,
+      "grad_norm": 5.956073760986328,
+      "learning_rate": 2e-05,
+      "loss": 3.8205,
+      "step": 4
+    },
+    {
+      "epoch": 0.019762845849802372,
+      "grad_norm": 6.856590270996094,
+      "learning_rate": 2.5e-05,
+      "loss": 3.5323,
+      "step": 5
+    },
+    {
+      "epoch": 0.023715415019762844,
+      "grad_norm": 7.316745281219482,
+      "learning_rate": 3e-05,
+      "loss": 3.7808,
+      "step": 6
+    },
+    {
+      "epoch": 0.02766798418972332,
+      "grad_norm": 4.9175705909729,
+      "learning_rate": 3.5e-05,
+      "loss": 3.4615,
+      "step": 7
+    },
+    {
+      "epoch": 0.03162055335968379,
+      "grad_norm": 4.404887676239014,
+      "learning_rate": 4e-05,
+      "loss": 3.3232,
+      "step": 8
+    },
+    {
+      "epoch": 0.03557312252964427,
+      "grad_norm": 4.211629390716553,
+      "learning_rate": 4.5e-05,
+      "loss": 3.2175,
+      "step": 9
+    },
+    {
+      "epoch": 0.039525691699604744,
+      "grad_norm": 4.854459762573242,
+      "learning_rate": 5e-05,
+      "loss": 3.3368,
+      "step": 10
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 3.71142840385437,
+      "learning_rate": 4.998015873015873e-05,
+      "loss": 3.2278,
+      "step": 11
+    },
+    {
+      "epoch": 0.04743083003952569,
+      "grad_norm": 2.8285276889801025,
+      "learning_rate": 4.996031746031746e-05,
+      "loss": 2.9991,
+      "step": 12
+    },
+    {
+      "epoch": 0.05138339920948617,
+      "grad_norm": 2.7281620502471924,
+      "learning_rate": 4.994047619047619e-05,
+      "loss": 3.2112,
+      "step": 13
+    },
+    {
+      "epoch": 0.05533596837944664,
+      "grad_norm": 2.6310040950775146,
+      "learning_rate": 4.9920634920634924e-05,
+      "loss": 3.3289,
+      "step": 14
+    },
+    {
+      "epoch": 0.05928853754940711,
+      "grad_norm": 2.928968667984009,
+      "learning_rate": 4.990079365079365e-05,
+      "loss": 3.3052,
+      "step": 15
+    },
+    {
+      "epoch": 0.06324110671936758,
+      "grad_norm": 2.7853801250457764,
+      "learning_rate": 4.9880952380952385e-05,
+      "loss": 3.0065,
+      "step": 16
+    },
+    {
+      "epoch": 0.06719367588932806,
+      "grad_norm": 2.0545976161956787,
+      "learning_rate": 4.986111111111111e-05,
+      "loss": 3.0645,
+      "step": 17
+    },
+    {
+      "epoch": 0.07114624505928854,
+      "grad_norm": 2.3543589115142822,
+      "learning_rate": 4.9841269841269845e-05,
+      "loss": 3.0591,
+      "step": 18
+    },
+    {
+      "epoch": 0.07509881422924901,
+      "grad_norm": 2.1679906845092773,
+      "learning_rate": 4.982142857142857e-05,
+      "loss": 2.9374,
+      "step": 19
+    },
+    {
+      "epoch": 0.07905138339920949,
+      "grad_norm": 2.5650746822357178,
+      "learning_rate": 4.9801587301587306e-05,
+      "loss": 2.9292,
+      "step": 20
+    },
+    {
+      "epoch": 0.08300395256916997,
+      "grad_norm": 2.4944090843200684,
+      "learning_rate": 4.978174603174603e-05,
+      "loss": 3.0967,
+      "step": 21
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 2.8087055683135986,
+      "learning_rate": 4.976190476190477e-05,
+      "loss": 3.076,
+      "step": 22
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 2.570450782775879,
+      "learning_rate": 4.9742063492063494e-05,
+      "loss": 2.8585,
+      "step": 23
+    },
+    {
+      "epoch": 0.09486166007905138,
+      "grad_norm": 2.3296656608581543,
+      "learning_rate": 4.972222222222223e-05,
+      "loss": 2.7182,
+      "step": 24
+    },
+    {
+      "epoch": 0.09881422924901186,
+      "grad_norm": 2.1972436904907227,
+      "learning_rate": 4.9702380952380955e-05,
+      "loss": 2.8991,
+      "step": 25
+    },
+    {
+      "epoch": 0.10276679841897234,
+      "grad_norm": 2.7606372833251953,
+      "learning_rate": 4.968253968253969e-05,
+      "loss": 2.912,
+      "step": 26
+    },
+    {
+      "epoch": 0.1067193675889328,
+      "grad_norm": 2.7917637825012207,
+      "learning_rate": 4.9662698412698415e-05,
+      "loss": 2.8944,
+      "step": 27
+    },
+    {
+      "epoch": 0.11067193675889328,
+      "grad_norm": 2.271846055984497,
+      "learning_rate": 4.964285714285715e-05,
+      "loss": 2.7234,
+      "step": 28
+    },
+    {
+      "epoch": 0.11462450592885376,
+      "grad_norm": 2.8581902980804443,
+      "learning_rate": 4.9623015873015876e-05,
+      "loss": 2.662,
+      "step": 29
+    },
+    {
+      "epoch": 0.11857707509881422,
+      "grad_norm": 2.4018585681915283,
+      "learning_rate": 4.960317460317461e-05,
+      "loss": 2.554,
+      "step": 30
+    },
+    {
+      "epoch": 0.1225296442687747,
+      "grad_norm": 3.0106191635131836,
+      "learning_rate": 4.958333333333334e-05,
+      "loss": 2.6187,
+      "step": 31
+    },
+    {
+      "epoch": 0.12648221343873517,
+      "grad_norm": 2.4871628284454346,
+      "learning_rate": 4.956349206349207e-05,
+      "loss": 2.5486,
+      "step": 32
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 3.1579999923706055,
+      "learning_rate": 4.95436507936508e-05,
+      "loss": 2.4142,
+      "step": 33
+    },
+    {
+      "epoch": 0.13438735177865613,
+      "grad_norm": 2.4816365242004395,
+      "learning_rate": 4.9523809523809525e-05,
+      "loss": 2.8092,
+      "step": 34
+    },
+    {
+      "epoch": 0.1383399209486166,
+      "grad_norm": 2.4598770141601562,
+      "learning_rate": 4.950396825396826e-05,
+      "loss": 2.5606,
+      "step": 35
+    },
+    {
+      "epoch": 0.1422924901185771,
+      "grad_norm": 2.9210352897644043,
+      "learning_rate": 4.9484126984126985e-05,
+      "loss": 2.6477,
+      "step": 36
+    },
+    {
+      "epoch": 0.14624505928853754,
+      "grad_norm": 2.082595109939575,
+      "learning_rate": 4.946428571428572e-05,
+      "loss": 2.6539,
+      "step": 37
+    },
+    {
+      "epoch": 0.15019762845849802,
+      "grad_norm": 2.171468734741211,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 2.8466,
+      "step": 38
+    },
+    {
+      "epoch": 0.1541501976284585,
+      "grad_norm": 2.395892858505249,
+      "learning_rate": 4.942460317460318e-05,
+      "loss": 2.4224,
+      "step": 39
+    },
+    {
+      "epoch": 0.15810276679841898,
+      "grad_norm": 3.123197317123413,
+      "learning_rate": 4.940476190476191e-05,
+      "loss": 2.9204,
+      "step": 40
+    },
+    {
+      "epoch": 0.16205533596837945,
+      "grad_norm": 2.659156084060669,
+      "learning_rate": 4.938492063492064e-05,
+      "loss": 2.43,
+      "step": 41
+    },
+    {
+      "epoch": 0.16600790513833993,
+      "grad_norm": 2.520598888397217,
+      "learning_rate": 4.936507936507937e-05,
+      "loss": 2.4932,
+      "step": 42
+    },
+    {
+      "epoch": 0.16996047430830039,
+      "grad_norm": 2.3670473098754883,
+      "learning_rate": 4.93452380952381e-05,
+      "loss": 2.5526,
+      "step": 43
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 2.4440720081329346,
+      "learning_rate": 4.932539682539683e-05,
+      "loss": 2.7987,
+      "step": 44
+    },
+    {
+      "epoch": 0.17786561264822134,
+      "grad_norm": 2.4289233684539795,
+      "learning_rate": 4.930555555555556e-05,
+      "loss": 2.5825,
+      "step": 45
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 2.3470001220703125,
+      "learning_rate": 4.928571428571429e-05,
+      "loss": 2.6192,
+      "step": 46
+    },
+    {
+      "epoch": 0.1857707509881423,
+      "grad_norm": 2.0881083011627197,
+      "learning_rate": 4.926587301587302e-05,
+      "loss": 2.5765,
+      "step": 47
+    },
+    {
+      "epoch": 0.18972332015810275,
+      "grad_norm": 2.184652328491211,
+      "learning_rate": 4.924603174603175e-05,
+      "loss": 2.5345,
+      "step": 48
+    },
+    {
+      "epoch": 0.19367588932806323,
+      "grad_norm": 2.1539316177368164,
+      "learning_rate": 4.9226190476190484e-05,
+      "loss": 2.7078,
+      "step": 49
+    },
+    {
+      "epoch": 0.1976284584980237,
+      "grad_norm": 2.476987361907959,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 2.5304,
+      "step": 50
+    },
+    {
+      "epoch": 0.2015810276679842,
+      "grad_norm": 2.591413974761963,
+      "learning_rate": 4.918650793650794e-05,
+      "loss": 2.2591,
+      "step": 51
+    },
+    {
+      "epoch": 0.20553359683794467,
+      "grad_norm": 6.341241836547852,
+      "learning_rate": 4.9166666666666665e-05,
+      "loss": 2.4837,
+      "step": 52
+    },
+    {
+      "epoch": 0.20948616600790515,
+      "grad_norm": 2.9513611793518066,
+      "learning_rate": 4.91468253968254e-05,
+      "loss": 2.6514,
+      "step": 53
+    },
+    {
+      "epoch": 0.2134387351778656,
+      "grad_norm": 2.2543134689331055,
+      "learning_rate": 4.9126984126984125e-05,
+      "loss": 2.657,
+      "step": 54
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 2.0405731201171875,
+      "learning_rate": 4.910714285714286e-05,
+      "loss": 2.5088,
+      "step": 55
+    },
+    {
+      "epoch": 0.22134387351778656,
+      "grad_norm": 2.6079225540161133,
+      "learning_rate": 4.9087301587301586e-05,
+      "loss": 2.2703,
+      "step": 56
+    },
+    {
+      "epoch": 0.22529644268774704,
+      "grad_norm": 2.587153673171997,
+      "learning_rate": 4.906746031746032e-05,
+      "loss": 2.5027,
+      "step": 57
+    },
+    {
+      "epoch": 0.22924901185770752,
+      "grad_norm": 2.083115577697754,
+      "learning_rate": 4.904761904761905e-05,
+      "loss": 2.7224,
+      "step": 58
+    },
+    {
+      "epoch": 0.233201581027668,
+      "grad_norm": 2.291238784790039,
+      "learning_rate": 4.902777777777778e-05,
+      "loss": 2.7383,
+      "step": 59
+    },
+    {
+      "epoch": 0.23715415019762845,
+      "grad_norm": 2.4876511096954346,
+      "learning_rate": 4.900793650793651e-05,
+      "loss": 2.3814,
+      "step": 60
+    },
+    {
+      "epoch": 0.24110671936758893,
+      "grad_norm": 2.6496829986572266,
+      "learning_rate": 4.898809523809524e-05,
+      "loss": 2.2271,
+      "step": 61
+    },
+    {
+      "epoch": 0.2450592885375494,
+      "grad_norm": 2.8697926998138428,
+      "learning_rate": 4.896825396825397e-05,
+      "loss": 2.5395,
+      "step": 62
+    },
+    {
+      "epoch": 0.2490118577075099,
+      "grad_norm": 2.5211477279663086,
+      "learning_rate": 4.89484126984127e-05,
+      "loss": 2.0096,
+      "step": 63
+    },
+    {
+      "epoch": 0.25296442687747034,
+      "grad_norm": 2.0613839626312256,
+      "learning_rate": 4.892857142857143e-05,
+      "loss": 2.1318,
+      "step": 64
+    },
+    {
+      "epoch": 0.25691699604743085,
+      "grad_norm": 2.4938395023345947,
+      "learning_rate": 4.8908730158730156e-05,
+      "loss": 2.4198,
+      "step": 65
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 2.5162112712860107,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 2.2697,
+      "step": 66
+    },
+    {
+      "epoch": 0.2648221343873518,
+      "grad_norm": 2.361053943634033,
+      "learning_rate": 4.886904761904762e-05,
+      "loss": 2.6438,
+      "step": 67
+    },
+    {
+      "epoch": 0.26877470355731226,
+      "grad_norm": 2.257228374481201,
+      "learning_rate": 4.884920634920635e-05,
+      "loss": 2.4411,
+      "step": 68
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 3.0840468406677246,
+      "learning_rate": 4.882936507936508e-05,
+      "loss": 2.1741,
+      "step": 69
+    },
+    {
+      "epoch": 0.2766798418972332,
+      "grad_norm": 2.4883201122283936,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 2.2851,
+      "step": 70
+    },
+    {
+      "epoch": 0.28063241106719367,
+      "grad_norm": 2.0108885765075684,
+      "learning_rate": 4.878968253968254e-05,
+      "loss": 2.3053,
+      "step": 71
+    },
+    {
+      "epoch": 0.2845849802371542,
+      "grad_norm": 2.4046590328216553,
+      "learning_rate": 4.876984126984127e-05,
+      "loss": 2.2728,
+      "step": 72
+    },
+    {
+      "epoch": 0.2885375494071146,
+      "grad_norm": 2.492821216583252,
+      "learning_rate": 4.875e-05,
+      "loss": 2.463,
+      "step": 73
+    },
+    {
+      "epoch": 0.2924901185770751,
+      "grad_norm": 1.9024349451065063,
+      "learning_rate": 4.873015873015873e-05,
+      "loss": 2.5196,
+      "step": 74
+    },
+    {
+      "epoch": 0.2964426877470356,
+      "grad_norm": 2.240678310394287,
+      "learning_rate": 4.871031746031746e-05,
+      "loss": 2.2548,
+      "step": 75
+    },
+    {
+      "epoch": 0.30039525691699603,
+      "grad_norm": 2.447615146636963,
+      "learning_rate": 4.8690476190476194e-05,
+      "loss": 2.9293,
+      "step": 76
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 1.805576205253601,
+      "learning_rate": 4.867063492063492e-05,
+      "loss": 2.3533,
+      "step": 77
+    },
+    {
+      "epoch": 0.308300395256917,
+      "grad_norm": 2.2282891273498535,
+      "learning_rate": 4.8650793650793654e-05,
+      "loss": 2.5451,
+      "step": 78
+    },
+    {
+      "epoch": 0.31225296442687744,
+      "grad_norm": 2.568352699279785,
+      "learning_rate": 4.863095238095238e-05,
+      "loss": 1.9444,
+      "step": 79
+    },
+    {
+      "epoch": 0.31620553359683795,
+      "grad_norm": 1.9792273044586182,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 2.4367,
+      "step": 80
+    },
+    {
+      "epoch": 0.3201581027667984,
+      "grad_norm": 2.3930373191833496,
+      "learning_rate": 4.859126984126984e-05,
+      "loss": 2.3669,
+      "step": 81
+    },
+    {
+      "epoch": 0.3241106719367589,
+      "grad_norm": 2.2622337341308594,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 2.4811,
+      "step": 82
+    },
+    {
+      "epoch": 0.32806324110671936,
+      "grad_norm": 2.4643726348876953,
+      "learning_rate": 4.85515873015873e-05,
+      "loss": 2.3285,
+      "step": 83
+    },
+    {
+      "epoch": 0.33201581027667987,
+      "grad_norm": 2.4243311882019043,
+      "learning_rate": 4.853174603174604e-05,
+      "loss": 2.579,
+      "step": 84
+    },
+    {
+      "epoch": 0.3359683794466403,
+      "grad_norm": 3.0440781116485596,
+      "learning_rate": 4.8511904761904764e-05,
+      "loss": 2.0484,
+      "step": 85
+    },
+    {
+      "epoch": 0.33992094861660077,
+      "grad_norm": 2.282578706741333,
+      "learning_rate": 4.84920634920635e-05,
+      "loss": 2.2485,
+      "step": 86
+    },
+    {
+      "epoch": 0.3438735177865613,
+      "grad_norm": 2.2174232006073,
+      "learning_rate": 4.8472222222222224e-05,
+      "loss": 2.6938,
+      "step": 87
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 2.4350028038024902,
+      "learning_rate": 4.845238095238095e-05,
+      "loss": 2.1194,
+      "step": 88
+    },
+    {
+      "epoch": 0.35177865612648224,
+      "grad_norm": 2.424428939819336,
+      "learning_rate": 4.8432539682539685e-05,
+      "loss": 2.1156,
+      "step": 89
+    },
+    {
+      "epoch": 0.3557312252964427,
+      "grad_norm": 2.444096326828003,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 2.2003,
+      "step": 90
+    },
+    {
+      "epoch": 0.35968379446640314,
+      "grad_norm": 2.2972753047943115,
+      "learning_rate": 4.8392857142857146e-05,
+      "loss": 2.273,
+      "step": 91
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 2.599595546722412,
+      "learning_rate": 4.837301587301587e-05,
+      "loss": 2.1506,
+      "step": 92
+    },
+    {
+      "epoch": 0.3675889328063241,
+      "grad_norm": 2.1127512454986572,
+      "learning_rate": 4.835317460317461e-05,
+      "loss": 2.2395,
+      "step": 93
+    },
+    {
+      "epoch": 0.3715415019762846,
+      "grad_norm": 1.8674228191375732,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 2.4755,
+      "step": 94
+    },
+    {
+      "epoch": 0.37549407114624506,
+      "grad_norm": 2.062201499938965,
+      "learning_rate": 4.831349206349207e-05,
+      "loss": 2.8275,
+      "step": 95
+    },
+    {
+      "epoch": 0.3794466403162055,
+      "grad_norm": 2.196667432785034,
+      "learning_rate": 4.8293650793650794e-05,
+      "loss": 2.3944,
+      "step": 96
+    },
+    {
+      "epoch": 0.383399209486166,
+      "grad_norm": 2.0713300704956055,
+      "learning_rate": 4.827380952380953e-05,
+      "loss": 2.4696,
+      "step": 97
+    },
+    {
+      "epoch": 0.38735177865612647,
+      "grad_norm": 2.0104503631591797,
+      "learning_rate": 4.8253968253968255e-05,
+      "loss": 2.334,
+      "step": 98
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 2.1917121410369873,
+      "learning_rate": 4.823412698412699e-05,
+      "loss": 2.2834,
+      "step": 99
+    },
+    {
+      "epoch": 0.3952569169960474,
+      "grad_norm": 2.4141860008239746,
+      "learning_rate": 4.8214285714285716e-05,
+      "loss": 2.3086,
+      "step": 100
+    },
+    {
+      "epoch": 0.39920948616600793,
+      "grad_norm": 2.384432792663574,
+      "learning_rate": 4.819444444444445e-05,
+      "loss": 2.1402,
+      "step": 101
+    },
+    {
+      "epoch": 0.4031620553359684,
+      "grad_norm": 2.4520158767700195,
+      "learning_rate": 4.817460317460318e-05,
+      "loss": 2.033,
+      "step": 102
+    },
+    {
+      "epoch": 0.40711462450592883,
+      "grad_norm": 2.809053659439087,
+      "learning_rate": 4.815476190476191e-05,
+      "loss": 2.0992,
+      "step": 103
+    },
+    {
+      "epoch": 0.41106719367588934,
+      "grad_norm": 2.3782355785369873,
+      "learning_rate": 4.813492063492064e-05,
+      "loss": 2.3137,
+      "step": 104
+    },
+    {
+      "epoch": 0.4150197628458498,
+      "grad_norm": 1.9347542524337769,
+      "learning_rate": 4.811507936507937e-05,
+      "loss": 2.3339,
+      "step": 105
+    },
+    {
+      "epoch": 0.4189723320158103,
+      "grad_norm": 2.0949320793151855,
+      "learning_rate": 4.80952380952381e-05,
+      "loss": 2.4536,
+      "step": 106
+    },
+    {
+      "epoch": 0.42292490118577075,
+      "grad_norm": 1.8690857887268066,
+      "learning_rate": 4.807539682539683e-05,
+      "loss": 2.0155,
+      "step": 107
+    },
+    {
+      "epoch": 0.4268774703557312,
+      "grad_norm": 2.1170482635498047,
+      "learning_rate": 4.805555555555556e-05,
+      "loss": 2.2891,
+      "step": 108
+    },
+    {
+      "epoch": 0.4308300395256917,
+      "grad_norm": 2.1500346660614014,
+      "learning_rate": 4.803571428571429e-05,
+      "loss": 2.4996,
+      "step": 109
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 1.9133753776550293,
+      "learning_rate": 4.801587301587302e-05,
+      "loss": 2.57,
+      "step": 110
+    },
+    {
+      "epoch": 0.43873517786561267,
+      "grad_norm": 2.0104360580444336,
+      "learning_rate": 4.799603174603175e-05,
+      "loss": 2.1764,
+      "step": 111
+    },
+    {
+      "epoch": 0.4426877470355731,
+      "grad_norm": 2.2259364128112793,
+      "learning_rate": 4.797619047619048e-05,
+      "loss": 2.4447,
+      "step": 112
+    },
+    {
+      "epoch": 0.44664031620553357,
+      "grad_norm": 1.6618857383728027,
+      "learning_rate": 4.795634920634921e-05,
+      "loss": 2.3628,
+      "step": 113
+    },
+    {
+      "epoch": 0.4505928853754941,
+      "grad_norm": 2.0611374378204346,
+      "learning_rate": 4.793650793650794e-05,
+      "loss": 2.7009,
+      "step": 114
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 2.3287601470947266,
+      "learning_rate": 4.791666666666667e-05,
+      "loss": 2.5049,
+      "step": 115
+    },
+    {
+      "epoch": 0.45849802371541504,
+      "grad_norm": 2.238161563873291,
+      "learning_rate": 4.78968253968254e-05,
+      "loss": 2.3077,
+      "step": 116
+    },
+    {
+      "epoch": 0.4624505928853755,
+      "grad_norm": 2.3234920501708984,
+      "learning_rate": 4.787698412698413e-05,
+      "loss": 2.3173,
+      "step": 117
+    },
+    {
+      "epoch": 0.466403162055336,
+      "grad_norm": 2.1455657482147217,
+      "learning_rate": 4.785714285714286e-05,
+      "loss": 2.329,
+      "step": 118
+    },
+    {
+      "epoch": 0.47035573122529645,
+      "grad_norm": 2.4944536685943604,
+      "learning_rate": 4.783730158730159e-05,
+      "loss": 2.1652,
+      "step": 119
+    },
+    {
+      "epoch": 0.4743083003952569,
+      "grad_norm": 1.9373809099197388,
+      "learning_rate": 4.781746031746032e-05,
+      "loss": 2.3349,
+      "step": 120
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 2.0994389057159424,
+      "learning_rate": 4.779761904761905e-05,
+      "loss": 2.27,
+      "step": 121
+    },
+    {
+      "epoch": 0.48221343873517786,
+      "grad_norm": 1.697857141494751,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 2.3405,
+      "step": 122
+    },
+    {
+      "epoch": 0.48616600790513836,
+      "grad_norm": 2.085141658782959,
+      "learning_rate": 4.775793650793651e-05,
+      "loss": 1.9531,
+      "step": 123
+    },
+    {
+      "epoch": 0.4901185770750988,
+      "grad_norm": 2.2885854244232178,
+      "learning_rate": 4.7738095238095245e-05,
+      "loss": 2.2837,
+      "step": 124
+    },
+    {
+      "epoch": 0.49407114624505927,
+      "grad_norm": 2.0135693550109863,
+      "learning_rate": 4.771825396825397e-05,
+      "loss": 2.3009,
+      "step": 125
+    },
+    {
+      "epoch": 0.4980237154150198,
+      "grad_norm": 2.485074520111084,
+      "learning_rate": 4.7698412698412706e-05,
+      "loss": 1.9311,
+      "step": 126
+    },
+    {
+      "epoch": 0.5019762845849802,
+      "grad_norm": 2.2360167503356934,
+      "learning_rate": 4.767857142857143e-05,
+      "loss": 2.426,
+      "step": 127
+    },
+    {
+      "epoch": 0.5059288537549407,
+      "grad_norm": 2.3205668926239014,
+      "learning_rate": 4.7658730158730166e-05,
+      "loss": 1.9649,
+      "step": 128
+    },
+    {
+      "epoch": 0.5098814229249012,
+      "grad_norm": 2.3518431186676025,
+      "learning_rate": 4.7638888888888887e-05,
+      "loss": 2.5223,
+      "step": 129
+    },
+    {
+      "epoch": 0.5138339920948617,
+      "grad_norm": 2.197111129760742,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 2.303,
+      "step": 130
+    },
+    {
+      "epoch": 0.5177865612648221,
+      "grad_norm": 2.4568827152252197,
+      "learning_rate": 4.759920634920635e-05,
+      "loss": 2.0753,
+      "step": 131
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 2.004725217819214,
+      "learning_rate": 4.757936507936508e-05,
+      "loss": 2.3092,
+      "step": 132
+    },
+    {
+      "epoch": 0.525691699604743,
+      "grad_norm": 2.0977554321289062,
+      "learning_rate": 4.755952380952381e-05,
+      "loss": 2.4393,
+      "step": 133
+    },
+    {
+      "epoch": 0.5296442687747036,
+      "grad_norm": 1.903745412826538,
+      "learning_rate": 4.753968253968254e-05,
+      "loss": 2.0935,
+      "step": 134
+    },
+    {
+      "epoch": 0.5335968379446641,
+      "grad_norm": 2.0144717693328857,
+      "learning_rate": 4.751984126984127e-05,
+      "loss": 2.2653,
+      "step": 135
+    },
+    {
+      "epoch": 0.5375494071146245,
+      "grad_norm": 2.0766761302948,
+      "learning_rate": 4.75e-05,
+      "loss": 2.2034,
+      "step": 136
+    },
+    {
+      "epoch": 0.541501976284585,
+      "grad_norm": 2.109747886657715,
+      "learning_rate": 4.748015873015873e-05,
+      "loss": 2.2778,
+      "step": 137
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 8.714595794677734,
+      "learning_rate": 4.746031746031746e-05,
+      "loss": 1.945,
+      "step": 138
+    },
+    {
+      "epoch": 0.549407114624506,
+      "grad_norm": 2.571970224380493,
+      "learning_rate": 4.744047619047619e-05,
+      "loss": 2.0047,
+      "step": 139
+    },
+    {
+      "epoch": 0.5533596837944664,
+      "grad_norm": 2.0782594680786133,
+      "learning_rate": 4.7420634920634924e-05,
+      "loss": 2.2821,
+      "step": 140
+    },
+    {
+      "epoch": 0.5573122529644269,
+      "grad_norm": 1.9552282094955444,
+      "learning_rate": 4.740079365079365e-05,
+      "loss": 1.9381,
+      "step": 141
+    },
+    {
+      "epoch": 0.5612648221343873,
+      "grad_norm": 2.5030784606933594,
+      "learning_rate": 4.738095238095238e-05,
+      "loss": 1.9552,
+      "step": 142
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 1.996016263961792,
+      "learning_rate": 4.736111111111111e-05,
+      "loss": 2.3815,
+      "step": 143
+    },
+    {
+      "epoch": 0.5691699604743083,
+      "grad_norm": 2.2575206756591797,
+      "learning_rate": 4.734126984126984e-05,
+      "loss": 2.3244,
+      "step": 144
+    },
+    {
+      "epoch": 0.5731225296442688,
+      "grad_norm": 2.067824363708496,
+      "learning_rate": 4.732142857142857e-05,
+      "loss": 2.338,
+      "step": 145
+    },
+    {
+      "epoch": 0.5770750988142292,
+      "grad_norm": 1.8195747137069702,
+      "learning_rate": 4.73015873015873e-05,
+      "loss": 2.4185,
+      "step": 146
+    },
+    {
+      "epoch": 0.5810276679841897,
+      "grad_norm": 2.230973958969116,
+      "learning_rate": 4.728174603174603e-05,
+      "loss": 2.2955,
+      "step": 147
+    },
+    {
+      "epoch": 0.5849802371541502,
+      "grad_norm": 2.303642749786377,
+      "learning_rate": 4.726190476190476e-05,
+      "loss": 2.6163,
+      "step": 148
+    },
+    {
+      "epoch": 0.5889328063241107,
+      "grad_norm": 2.1462321281433105,
+      "learning_rate": 4.7242063492063494e-05,
+      "loss": 2.1224,
+      "step": 149
+    },
+    {
+      "epoch": 0.5928853754940712,
+      "grad_norm": 1.8279284238815308,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 2.1556,
+      "step": 150
+    },
+    {
+      "epoch": 0.5968379446640316,
+      "grad_norm": 2.0945279598236084,
+      "learning_rate": 4.7202380952380955e-05,
+      "loss": 2.2004,
+      "step": 151
+    },
+    {
+      "epoch": 0.6007905138339921,
+      "grad_norm": 2.0842084884643555,
+      "learning_rate": 4.718253968253968e-05,
+      "loss": 1.9211,
+      "step": 152
+    },
+    {
+      "epoch": 0.6047430830039525,
+      "grad_norm": 1.9995882511138916,
+      "learning_rate": 4.7162698412698416e-05,
+      "loss": 2.116,
+      "step": 153
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 2.471076250076294,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 2.035,
+      "step": 154
+    },
+    {
+      "epoch": 0.6126482213438735,
+      "grad_norm": 2.1906332969665527,
+      "learning_rate": 4.7123015873015876e-05,
+      "loss": 1.8779,
+      "step": 155
+    },
+    {
+      "epoch": 0.616600790513834,
+      "grad_norm": 2.511838674545288,
+      "learning_rate": 4.71031746031746e-05,
+      "loss": 1.9335,
+      "step": 156
+    },
+    {
+      "epoch": 0.6205533596837944,
+      "grad_norm": 1.754230260848999,
+      "learning_rate": 4.708333333333334e-05,
+      "loss": 2.1955,
+      "step": 157
+    },
+    {
+      "epoch": 0.6245059288537549,
+      "grad_norm": 1.9571726322174072,
+      "learning_rate": 4.7063492063492064e-05,
+      "loss": 2.0802,
+      "step": 158
+    },
+    {
+      "epoch": 0.6284584980237155,
+      "grad_norm": 2.271517038345337,
+      "learning_rate": 4.70436507936508e-05,
+      "loss": 2.2352,
+      "step": 159
+    },
+    {
+      "epoch": 0.6324110671936759,
+      "grad_norm": 1.9034878015518188,
+      "learning_rate": 4.7023809523809525e-05,
+      "loss": 2.2499,
+      "step": 160
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 1.9912493228912354,
+      "learning_rate": 4.700396825396826e-05,
+      "loss": 2.6197,
+      "step": 161
+    },
+    {
+      "epoch": 0.6403162055335968,
+      "grad_norm": 2.255777597427368,
+      "learning_rate": 4.6984126984126986e-05,
+      "loss": 2.36,
+      "step": 162
+    },
+    {
+      "epoch": 0.6442687747035574,
+      "grad_norm": 2.2219536304473877,
+      "learning_rate": 4.696428571428572e-05,
+      "loss": 1.9835,
+      "step": 163
+    },
+    {
+      "epoch": 0.6482213438735178,
+      "grad_norm": 2.551605224609375,
+      "learning_rate": 4.6944444444444446e-05,
+      "loss": 2.2187,
+      "step": 164
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 2.3967275619506836,
+      "learning_rate": 4.692460317460317e-05,
+      "loss": 1.8407,
+      "step": 165
+    },
+    {
+      "epoch": 0.6561264822134387,
+      "grad_norm": 2.194493293762207,
+      "learning_rate": 4.690476190476191e-05,
+      "loss": 2.3458,
+      "step": 166
+    },
+    {
+      "epoch": 0.6600790513833992,
+      "grad_norm": 1.9432865381240845,
+      "learning_rate": 4.6884920634920634e-05,
+      "loss": 2.2172,
+      "step": 167
+    },
+    {
+      "epoch": 0.6640316205533597,
+      "grad_norm": 2.18040132522583,
+      "learning_rate": 4.686507936507937e-05,
+      "loss": 2.1371,
+      "step": 168
+    },
+    {
+      "epoch": 0.6679841897233202,
+      "grad_norm": 2.4075417518615723,
+      "learning_rate": 4.6845238095238095e-05,
+      "loss": 2.2766,
+      "step": 169
+    },
+    {
+      "epoch": 0.6719367588932806,
+      "grad_norm": 1.8353605270385742,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 1.809,
+      "step": 170
+    },
+    {
+      "epoch": 0.6758893280632411,
+      "grad_norm": 2.305044651031494,
+      "learning_rate": 4.6805555555555556e-05,
+      "loss": 1.9724,
+      "step": 171
+    },
+    {
+      "epoch": 0.6798418972332015,
+      "grad_norm": 2.7235798835754395,
+      "learning_rate": 4.678571428571429e-05,
+      "loss": 1.6704,
+      "step": 172
+    },
+    {
+      "epoch": 0.6837944664031621,
+      "grad_norm": 2.5702693462371826,
+      "learning_rate": 4.6765873015873016e-05,
+      "loss": 2.0057,
+      "step": 173
+    },
+    {
+      "epoch": 0.6877470355731226,
+      "grad_norm": 1.952614188194275,
+      "learning_rate": 4.674603174603175e-05,
+      "loss": 2.3162,
+      "step": 174
+    },
+    {
+      "epoch": 0.691699604743083,
+      "grad_norm": 2.334252119064331,
+      "learning_rate": 4.672619047619048e-05,
+      "loss": 2.2679,
+      "step": 175
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 2.064568519592285,
+      "learning_rate": 4.670634920634921e-05,
+      "loss": 2.2213,
+      "step": 176
+    },
+    {
+      "epoch": 0.6996047430830039,
+      "grad_norm": 1.8959503173828125,
+      "learning_rate": 4.668650793650794e-05,
+      "loss": 2.4685,
+      "step": 177
+    },
+    {
+      "epoch": 0.7035573122529645,
+      "grad_norm": 2.5481746196746826,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 2.2007,
+      "step": 178
+    },
+    {
+      "epoch": 0.7075098814229249,
+      "grad_norm": 1.9843651056289673,
+      "learning_rate": 4.66468253968254e-05,
+      "loss": 2.2955,
+      "step": 179
+    },
+    {
+      "epoch": 0.7114624505928854,
+      "grad_norm": 1.9685429334640503,
+      "learning_rate": 4.662698412698413e-05,
+      "loss": 2.0954,
+      "step": 180
+    },
+    {
+      "epoch": 0.7154150197628458,
+      "grad_norm": 2.216379165649414,
+      "learning_rate": 4.660714285714286e-05,
+      "loss": 2.3786,
+      "step": 181
+    },
+    {
+      "epoch": 0.7193675889328063,
+      "grad_norm": 2.144599437713623,
+      "learning_rate": 4.658730158730159e-05,
+      "loss": 2.1757,
+      "step": 182
+    },
+    {
+      "epoch": 0.7233201581027668,
+      "grad_norm": 1.829734206199646,
+      "learning_rate": 4.656746031746032e-05,
+      "loss": 2.2259,
+      "step": 183
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 1.8877259492874146,
+      "learning_rate": 4.6547619047619054e-05,
+      "loss": 2.1431,
+      "step": 184
+    },
+    {
+      "epoch": 0.7312252964426877,
+      "grad_norm": 1.9971317052841187,
+      "learning_rate": 4.652777777777778e-05,
+      "loss": 2.263,
+      "step": 185
+    },
+    {
+      "epoch": 0.7351778656126482,
+      "grad_norm": 2.129324436187744,
+      "learning_rate": 4.6507936507936515e-05,
+      "loss": 2.1523,
+      "step": 186
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 2.2705445289611816,
+      "learning_rate": 4.648809523809524e-05,
+      "loss": 2.0131,
+      "step": 187
+    },
+    {
+      "epoch": 0.7430830039525692,
+      "grad_norm": 1.9856712818145752,
+      "learning_rate": 4.646825396825397e-05,
+      "loss": 1.8903,
+      "step": 188
+    },
+    {
+      "epoch": 0.7470355731225297,
+      "grad_norm": 2.1891982555389404,
+      "learning_rate": 4.64484126984127e-05,
+      "loss": 2.2185,
+      "step": 189
+    },
+    {
+      "epoch": 0.7509881422924901,
+      "grad_norm": 2.514817476272583,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 2.3529,
+      "step": 190
+    },
+    {
+      "epoch": 0.7549407114624506,
+      "grad_norm": 1.9026365280151367,
+      "learning_rate": 4.640873015873016e-05,
+      "loss": 2.4147,
+      "step": 191
+    },
+    {
+      "epoch": 0.758893280632411,
+      "grad_norm": 2.04667329788208,
+      "learning_rate": 4.638888888888889e-05,
+      "loss": 1.9511,
+      "step": 192
+    },
+    {
+      "epoch": 0.7628458498023716,
+      "grad_norm": 1.8381803035736084,
+      "learning_rate": 4.6369047619047624e-05,
+      "loss": 2.1575,
+      "step": 193
+    },
+    {
+      "epoch": 0.766798418972332,
+      "grad_norm": 2.05672550201416,
+      "learning_rate": 4.634920634920635e-05,
+      "loss": 2.0799,
+      "step": 194
+    },
+    {
+      "epoch": 0.7707509881422925,
+      "grad_norm": 1.973151445388794,
+      "learning_rate": 4.6329365079365085e-05,
+      "loss": 2.2552,
+      "step": 195
+    },
+    {
+      "epoch": 0.7747035573122529,
+      "grad_norm": 1.8159013986587524,
+      "learning_rate": 4.630952380952381e-05,
+      "loss": 1.8409,
+      "step": 196
+    },
+    {
+      "epoch": 0.7786561264822134,
+      "grad_norm": 2.1178860664367676,
+      "learning_rate": 4.6289682539682545e-05,
+      "loss": 2.1587,
+      "step": 197
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 2.200260877609253,
+      "learning_rate": 4.626984126984127e-05,
+      "loss": 1.5963,
+      "step": 198
+    },
+    {
+      "epoch": 0.7865612648221344,
+      "grad_norm": 2.7356910705566406,
+      "learning_rate": 4.6250000000000006e-05,
+      "loss": 1.7795,
+      "step": 199
+    },
+    {
+      "epoch": 0.7905138339920948,
+      "grad_norm": 2.6005921363830566,
+      "learning_rate": 4.623015873015873e-05,
+      "loss": 2.07,
+      "step": 200
+    },
+    {
+      "epoch": 0.7944664031620553,
+      "grad_norm": 2.155571937561035,
+      "learning_rate": 4.621031746031747e-05,
+      "loss": 1.9797,
+      "step": 201
+    },
+    {
+      "epoch": 0.7984189723320159,
+      "grad_norm": 1.8139017820358276,
+      "learning_rate": 4.6190476190476194e-05,
+      "loss": 2.0915,
+      "step": 202
+    },
+    {
+      "epoch": 0.8023715415019763,
+      "grad_norm": 2.1978001594543457,
+      "learning_rate": 4.617063492063493e-05,
+      "loss": 2.0828,
+      "step": 203
+    },
+    {
+      "epoch": 0.8063241106719368,
+      "grad_norm": 2.4255006313323975,
+      "learning_rate": 4.6150793650793655e-05,
+      "loss": 1.8046,
+      "step": 204
+    },
+    {
+      "epoch": 0.8102766798418972,
+      "grad_norm": 2.3868203163146973,
+      "learning_rate": 4.613095238095239e-05,
+      "loss": 2.0577,
+      "step": 205
+    },
+    {
+      "epoch": 0.8142292490118577,
+      "grad_norm": 1.7391860485076904,
+      "learning_rate": 4.6111111111111115e-05,
+      "loss": 2.292,
+      "step": 206
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 1.9303078651428223,
+      "learning_rate": 4.609126984126984e-05,
+      "loss": 1.8133,
+      "step": 207
+    },
+    {
+      "epoch": 0.8221343873517787,
+      "grad_norm": 1.9964845180511475,
+      "learning_rate": 4.607142857142857e-05,
+      "loss": 2.1007,
+      "step": 208
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 2.102794885635376,
+      "learning_rate": 4.60515873015873e-05,
+      "loss": 2.2607,
+      "step": 209
+    },
+    {
+      "epoch": 0.8300395256916996,
+      "grad_norm": 2.2853715419769287,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 2.2519,
+      "step": 210
+    },
+    {
+      "epoch": 0.83399209486166,
+      "grad_norm": 1.8562860488891602,
+      "learning_rate": 4.6011904761904764e-05,
+      "loss": 2.1676,
+      "step": 211
+    },
+    {
+      "epoch": 0.8379446640316206,
+      "grad_norm": 2.448793649673462,
+      "learning_rate": 4.599206349206349e-05,
+      "loss": 2.1657,
+      "step": 212
+    },
+    {
+      "epoch": 0.841897233201581,
+      "grad_norm": 1.8922572135925293,
+      "learning_rate": 4.5972222222222225e-05,
+      "loss": 1.7849,
+      "step": 213
+    },
+    {
+      "epoch": 0.8458498023715415,
+      "grad_norm": 2.190263509750366,
+      "learning_rate": 4.595238095238095e-05,
+      "loss": 2.1807,
+      "step": 214
+    },
+    {
+      "epoch": 0.849802371541502,
+      "grad_norm": 2.0355405807495117,
+      "learning_rate": 4.5932539682539685e-05,
+      "loss": 2.163,
+      "step": 215
+    },
+    {
+      "epoch": 0.8537549407114624,
+      "grad_norm": 2.3606069087982178,
+      "learning_rate": 4.591269841269841e-05,
+      "loss": 2.3872,
+      "step": 216
+    },
+    {
+      "epoch": 0.857707509881423,
+      "grad_norm": 2.091801166534424,
+      "learning_rate": 4.5892857142857146e-05,
+      "loss": 2.0303,
+      "step": 217
+    },
+    {
+      "epoch": 0.8616600790513834,
+      "grad_norm": 2.5148980617523193,
+      "learning_rate": 4.587301587301587e-05,
+      "loss": 1.7546,
+      "step": 218
+    },
+    {
+      "epoch": 0.8656126482213439,
+      "grad_norm": 2.172477960586548,
+      "learning_rate": 4.58531746031746e-05,
+      "loss": 2.0569,
+      "step": 219
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 1.820509433746338,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 2.2322,
+      "step": 220
+    },
+    {
+      "epoch": 0.8735177865612648,
+      "grad_norm": 1.5621439218521118,
+      "learning_rate": 4.581349206349206e-05,
+      "loss": 2.0088,
+      "step": 221
+    },
+    {
+      "epoch": 0.8774703557312253,
+      "grad_norm": 2.2147130966186523,
+      "learning_rate": 4.5793650793650795e-05,
+      "loss": 1.9092,
+      "step": 222
+    },
+    {
+      "epoch": 0.8814229249011858,
+      "grad_norm": 2.103334426879883,
+      "learning_rate": 4.577380952380952e-05,
+      "loss": 1.9992,
+      "step": 223
+    },
+    {
+      "epoch": 0.8853754940711462,
+      "grad_norm": 1.926761507987976,
+      "learning_rate": 4.5753968253968255e-05,
+      "loss": 1.8496,
+      "step": 224
+    },
+    {
+      "epoch": 0.8893280632411067,
+      "grad_norm": 2.040013313293457,
+      "learning_rate": 4.573412698412698e-05,
+      "loss": 1.9573,
+      "step": 225
+    },
+    {
+      "epoch": 0.8932806324110671,
+      "grad_norm": 1.969488263130188,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 2.1783,
+      "step": 226
+    },
+    {
+      "epoch": 0.8972332015810277,
+      "grad_norm": 2.0519626140594482,
+      "learning_rate": 4.569444444444444e-05,
+      "loss": 1.9827,
+      "step": 227
+    },
+    {
+      "epoch": 0.9011857707509882,
+      "grad_norm": 2.2971911430358887,
+      "learning_rate": 4.567460317460318e-05,
+      "loss": 1.9498,
+      "step": 228
+    },
+    {
+      "epoch": 0.9051383399209486,
+      "grad_norm": 2.0618982315063477,
+      "learning_rate": 4.5654761904761904e-05,
+      "loss": 1.8099,
+      "step": 229
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.670811414718628,
+      "learning_rate": 4.563492063492064e-05,
+      "loss": 2.1808,
+      "step": 230
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 3.3800909519195557,
+      "learning_rate": 4.5615079365079365e-05,
+      "loss": 1.458,
+      "step": 231
+    },
+    {
+      "epoch": 0.9169960474308301,
+      "grad_norm": 2.106755495071411,
+      "learning_rate": 4.55952380952381e-05,
+      "loss": 2.1909,
+      "step": 232
+    },
+    {
+      "epoch": 0.9209486166007905,
+      "grad_norm": 2.3558948040008545,
+      "learning_rate": 4.5575396825396825e-05,
+      "loss": 2.1529,
+      "step": 233
+    },
+    {
+      "epoch": 0.924901185770751,
+      "grad_norm": 2.1475250720977783,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 2.0272,
+      "step": 234
+    },
+    {
+      "epoch": 0.9288537549407114,
+      "grad_norm": 2.1931400299072266,
+      "learning_rate": 4.5535714285714286e-05,
+      "loss": 2.0567,
+      "step": 235
+    },
+    {
+      "epoch": 0.932806324110672,
+      "grad_norm": 2.34619140625,
+      "learning_rate": 4.551587301587302e-05,
+      "loss": 2.36,
+      "step": 236
+    },
+    {
+      "epoch": 0.9367588932806324,
+      "grad_norm": 1.8363399505615234,
+      "learning_rate": 4.549603174603175e-05,
+      "loss": 2.2903,
+      "step": 237
+    },
+    {
+      "epoch": 0.9407114624505929,
+      "grad_norm": 2.0586445331573486,
+      "learning_rate": 4.547619047619048e-05,
+      "loss": 1.8766,
+      "step": 238
+    },
+    {
+      "epoch": 0.9446640316205533,
+      "grad_norm": 1.7671842575073242,
+      "learning_rate": 4.545634920634921e-05,
+      "loss": 1.7877,
+      "step": 239
+    },
+    {
+      "epoch": 0.9486166007905138,
+      "grad_norm": 2.8485286235809326,
+      "learning_rate": 4.543650793650794e-05,
+      "loss": 1.9066,
+      "step": 240
+    },
+    {
+      "epoch": 0.9525691699604744,
+      "grad_norm": 2.1801576614379883,
+      "learning_rate": 4.541666666666667e-05,
+      "loss": 2.2792,
+      "step": 241
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 1.9892218112945557,
+      "learning_rate": 4.5396825396825395e-05,
+      "loss": 2.1138,
+      "step": 242
+    },
+    {
+      "epoch": 0.9604743083003953,
+      "grad_norm": 2.0810766220092773,
+      "learning_rate": 4.537698412698413e-05,
+      "loss": 1.8441,
+      "step": 243
+    },
+    {
+      "epoch": 0.9644268774703557,
+      "grad_norm": 2.243373155593872,
+      "learning_rate": 4.5357142857142856e-05,
+      "loss": 2.0694,
+      "step": 244
+    },
+    {
+      "epoch": 0.9683794466403162,
+      "grad_norm": 2.5647904872894287,
+      "learning_rate": 4.533730158730159e-05,
+      "loss": 1.861,
+      "step": 245
+    },
+    {
+      "epoch": 0.9723320158102767,
+      "grad_norm": 1.625081181526184,
+      "learning_rate": 4.531746031746032e-05,
+      "loss": 2.1496,
+      "step": 246
+    },
+    {
+      "epoch": 0.9762845849802372,
+      "grad_norm": 1.7020546197891235,
+      "learning_rate": 4.529761904761905e-05,
+      "loss": 1.9332,
+      "step": 247
+    },
+    {
+      "epoch": 0.9802371541501976,
+      "grad_norm": 1.802681565284729,
+      "learning_rate": 4.527777777777778e-05,
+      "loss": 1.8204,
+      "step": 248
+    },
+    {
+      "epoch": 0.9841897233201581,
+      "grad_norm": 1.7810888290405273,
+      "learning_rate": 4.525793650793651e-05,
+      "loss": 1.8915,
+      "step": 249
+    },
+    {
+      "epoch": 0.9881422924901185,
+      "grad_norm": 1.6715744733810425,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 2.0593,
+      "step": 250
+    },
+    {
+      "epoch": 0.9920948616600791,
+      "grad_norm": 2.5212719440460205,
+      "learning_rate": 4.521825396825397e-05,
+      "loss": 2.0725,
+      "step": 251
+    },
+    {
+      "epoch": 0.9960474308300395,
+      "grad_norm": 2.189178943634033,
+      "learning_rate": 4.51984126984127e-05,
+      "loss": 2.0336,
+      "step": 252
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2733006477355957,
+      "learning_rate": 4.517857142857143e-05,
+      "loss": 1.9606,
+      "step": 253
+    },
+    {
+      "epoch": 1.0039525691699605,
+      "grad_norm": 1.8183317184448242,
+      "learning_rate": 4.515873015873016e-05,
+      "loss": 1.9106,
+      "step": 254
+    },
+    {
+      "epoch": 1.007905138339921,
+      "grad_norm": 2.471332550048828,
+      "learning_rate": 4.5138888888888894e-05,
+      "loss": 1.9963,
+      "step": 255
+    },
+    {
+      "epoch": 1.0118577075098814,
+      "grad_norm": 1.8422023057937622,
+      "learning_rate": 4.511904761904762e-05,
+      "loss": 2.0978,
+      "step": 256
+    },
+    {
+      "epoch": 1.0158102766798418,
+      "grad_norm": 1.79561185836792,
+      "learning_rate": 4.5099206349206354e-05,
+      "loss": 1.4994,
+      "step": 257
+    },
+    {
+      "epoch": 1.0197628458498025,
+      "grad_norm": 2.271358013153076,
+      "learning_rate": 4.507936507936508e-05,
+      "loss": 1.7649,
+      "step": 258
+    },
+    {
+      "epoch": 1.023715415019763,
+      "grad_norm": 2.6813673973083496,
+      "learning_rate": 4.5059523809523815e-05,
+      "loss": 1.8252,
+      "step": 259
+    },
+    {
+      "epoch": 1.0276679841897234,
+      "grad_norm": 1.8112016916275024,
+      "learning_rate": 4.503968253968254e-05,
+      "loss": 1.7883,
+      "step": 260
+    },
+    {
+      "epoch": 1.0316205533596838,
+      "grad_norm": 2.4050188064575195,
+      "learning_rate": 4.5019841269841276e-05,
+      "loss": 1.9855,
+      "step": 261
+    },
+    {
+      "epoch": 1.0355731225296443,
+      "grad_norm": 2.38291072845459,
+      "learning_rate": 4.5e-05,
+      "loss": 1.8171,
+      "step": 262
+    },
+    {
+      "epoch": 1.0395256916996047,
+      "grad_norm": 2.221317768096924,
+      "learning_rate": 4.4980158730158737e-05,
+      "loss": 1.5565,
+      "step": 263
+    },
+    {
+      "epoch": 1.0434782608695652,
+      "grad_norm": 2.3104920387268066,
+      "learning_rate": 4.4960317460317464e-05,
+      "loss": 1.926,
+      "step": 264
+    },
+    {
+      "epoch": 1.0474308300395256,
+      "grad_norm": 2.0159404277801514,
+      "learning_rate": 4.494047619047619e-05,
+      "loss": 1.6123,
+      "step": 265
+    },
+    {
+      "epoch": 1.051383399209486,
+      "grad_norm": 2.0315968990325928,
+      "learning_rate": 4.4920634920634924e-05,
+      "loss": 1.9577,
+      "step": 266
+    },
+    {
+      "epoch": 1.0553359683794465,
+      "grad_norm": 2.021657943725586,
+      "learning_rate": 4.490079365079365e-05,
+      "loss": 1.7845,
+      "step": 267
+    },
+    {
+      "epoch": 1.0592885375494072,
+      "grad_norm": 2.1020936965942383,
+      "learning_rate": 4.4880952380952385e-05,
+      "loss": 1.8233,
+      "step": 268
+    },
+    {
+      "epoch": 1.0632411067193677,
+      "grad_norm": 2.0355613231658936,
+      "learning_rate": 4.486111111111111e-05,
+      "loss": 1.9241,
+      "step": 269
+    },
+    {
+      "epoch": 1.0671936758893281,
+      "grad_norm": 1.797027349472046,
+      "learning_rate": 4.4841269841269846e-05,
+      "loss": 1.7056,
+      "step": 270
+    },
+    {
+      "epoch": 1.0711462450592886,
+      "grad_norm": 2.725644826889038,
+      "learning_rate": 4.482142857142857e-05,
+      "loss": 1.4221,
+      "step": 271
+    },
+    {
+      "epoch": 1.075098814229249,
+      "grad_norm": 2.5590972900390625,
+      "learning_rate": 4.4801587301587307e-05,
+      "loss": 1.7194,
+      "step": 272
+    },
+    {
+      "epoch": 1.0790513833992095,
+      "grad_norm": 2.4488916397094727,
+      "learning_rate": 4.4781746031746034e-05,
+      "loss": 1.3487,
+      "step": 273
+    },
+    {
+      "epoch": 1.08300395256917,
+      "grad_norm": 2.245779037475586,
+      "learning_rate": 4.476190476190477e-05,
+      "loss": 1.6895,
+      "step": 274
+    },
+    {
+      "epoch": 1.0869565217391304,
+      "grad_norm": 2.3138208389282227,
+      "learning_rate": 4.4742063492063494e-05,
+      "loss": 2.0797,
+      "step": 275
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 2.106358528137207,
+      "learning_rate": 4.472222222222223e-05,
+      "loss": 1.7057,
+      "step": 276
+    },
+    {
+      "epoch": 1.0948616600790513,
+      "grad_norm": 2.0810418128967285,
+      "learning_rate": 4.4702380952380955e-05,
+      "loss": 1.6139,
+      "step": 277
+    },
+    {
+      "epoch": 1.098814229249012,
+      "grad_norm": 2.3179874420166016,
+      "learning_rate": 4.468253968253969e-05,
+      "loss": 1.7748,
+      "step": 278
+    },
+    {
+      "epoch": 1.1027667984189724,
+      "grad_norm": 2.4948816299438477,
+      "learning_rate": 4.4662698412698416e-05,
+      "loss": 2.0072,
+      "step": 279
+    },
+    {
+      "epoch": 1.1067193675889329,
+      "grad_norm": 2.1853582859039307,
+      "learning_rate": 4.464285714285715e-05,
+      "loss": 2.017,
+      "step": 280
+    },
+    {
+      "epoch": 1.1106719367588933,
+      "grad_norm": 2.066575288772583,
+      "learning_rate": 4.4623015873015877e-05,
+      "loss": 1.7979,
+      "step": 281
+    },
+    {
+      "epoch": 1.1146245059288538,
+      "grad_norm": 2.454230785369873,
+      "learning_rate": 4.460317460317461e-05,
+      "loss": 1.6153,
+      "step": 282
+    },
+    {
+      "epoch": 1.1185770750988142,
+      "grad_norm": 2.088303804397583,
+      "learning_rate": 4.458333333333334e-05,
+      "loss": 1.7366,
+      "step": 283
+    },
+    {
+      "epoch": 1.1225296442687747,
+      "grad_norm": 2.0203330516815186,
+      "learning_rate": 4.456349206349207e-05,
+      "loss": 1.8878,
+      "step": 284
+    },
+    {
+      "epoch": 1.1264822134387351,
+      "grad_norm": 1.8328367471694946,
+      "learning_rate": 4.45436507936508e-05,
+      "loss": 1.4853,
+      "step": 285
+    },
+    {
+      "epoch": 1.1304347826086956,
+      "grad_norm": 2.1855709552764893,
+      "learning_rate": 4.4523809523809525e-05,
+      "loss": 2.1256,
+      "step": 286
+    },
+    {
+      "epoch": 1.1343873517786562,
+      "grad_norm": 2.0757203102111816,
+      "learning_rate": 4.450396825396825e-05,
+      "loss": 2.1395,
+      "step": 287
+    },
+    {
+      "epoch": 1.1383399209486167,
+      "grad_norm": 2.217658519744873,
+      "learning_rate": 4.4484126984126986e-05,
+      "loss": 1.6217,
+      "step": 288
+    },
+    {
+      "epoch": 1.1422924901185771,
+      "grad_norm": 2.346212387084961,
+      "learning_rate": 4.446428571428571e-05,
+      "loss": 1.6561,
+      "step": 289
+    },
+    {
+      "epoch": 1.1462450592885376,
+      "grad_norm": 2.340932607650757,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.5699,
+      "step": 290
+    },
+    {
+      "epoch": 1.150197628458498,
+      "grad_norm": 2.2137064933776855,
+      "learning_rate": 4.4424603174603174e-05,
+      "loss": 1.9642,
+      "step": 291
+    },
+    {
+      "epoch": 1.1541501976284585,
+      "grad_norm": 1.6653351783752441,
+      "learning_rate": 4.440476190476191e-05,
+      "loss": 1.8973,
+      "step": 292
+    },
+    {
+      "epoch": 1.158102766798419,
+      "grad_norm": 2.238419771194458,
+      "learning_rate": 4.4384920634920634e-05,
+      "loss": 1.7091,
+      "step": 293
+    },
+    {
+      "epoch": 1.1620553359683794,
+      "grad_norm": 1.8844563961029053,
+      "learning_rate": 4.436507936507937e-05,
+      "loss": 1.8355,
+      "step": 294
+    },
+    {
+      "epoch": 1.1660079051383399,
+      "grad_norm": 1.5482860803604126,
+      "learning_rate": 4.4345238095238095e-05,
+      "loss": 1.8152,
+      "step": 295
+    },
+    {
+      "epoch": 1.1699604743083003,
+      "grad_norm": 2.046618700027466,
+      "learning_rate": 4.432539682539683e-05,
+      "loss": 1.5727,
+      "step": 296
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 1.9557174444198608,
+      "learning_rate": 4.4305555555555556e-05,
+      "loss": 1.2086,
+      "step": 297
+    },
+    {
+      "epoch": 1.1778656126482214,
+      "grad_norm": 2.5918216705322266,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 1.8132,
+      "step": 298
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 2.1008517742156982,
+      "learning_rate": 4.4265873015873017e-05,
+      "loss": 1.7421,
+      "step": 299
+    },
+    {
+      "epoch": 1.1857707509881423,
+      "grad_norm": 2.3811569213867188,
+      "learning_rate": 4.4246031746031744e-05,
+      "loss": 1.9358,
+      "step": 300
+    },
+    {
+      "epoch": 1.1897233201581028,
+      "grad_norm": 2.317112922668457,
+      "learning_rate": 4.422619047619048e-05,
+      "loss": 1.8341,
+      "step": 301
+    },
+    {
+      "epoch": 1.1936758893280632,
+      "grad_norm": 1.8912357091903687,
+      "learning_rate": 4.4206349206349204e-05,
+      "loss": 1.7062,
+      "step": 302
+    },
+    {
+      "epoch": 1.1976284584980237,
+      "grad_norm": 1.7493529319763184,
+      "learning_rate": 4.418650793650794e-05,
+      "loss": 1.9519,
+      "step": 303
+    },
+    {
+      "epoch": 1.2015810276679841,
+      "grad_norm": 2.0161383152008057,
+      "learning_rate": 4.4166666666666665e-05,
+      "loss": 1.6137,
+      "step": 304
+    },
+    {
+      "epoch": 1.2055335968379446,
+      "grad_norm": 1.7949028015136719,
+      "learning_rate": 4.41468253968254e-05,
+      "loss": 1.7514,
+      "step": 305
+    },
+    {
+      "epoch": 1.2094861660079053,
+      "grad_norm": 2.1296987533569336,
+      "learning_rate": 4.4126984126984126e-05,
+      "loss": 1.6876,
+      "step": 306
+    },
+    {
+      "epoch": 1.2134387351778657,
+      "grad_norm": 2.3091883659362793,
+      "learning_rate": 4.410714285714286e-05,
+      "loss": 1.9752,
+      "step": 307
+    },
+    {
+      "epoch": 1.2173913043478262,
+      "grad_norm": 2.804625988006592,
+      "learning_rate": 4.4087301587301587e-05,
+      "loss": 1.7697,
+      "step": 308
+    },
+    {
+      "epoch": 1.2213438735177866,
+      "grad_norm": 1.7799369096755981,
+      "learning_rate": 4.406746031746032e-05,
+      "loss": 1.8301,
+      "step": 309
+    },
+    {
+      "epoch": 1.225296442687747,
+      "grad_norm": 2.0240299701690674,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 1.6604,
+      "step": 310
+    },
+    {
+      "epoch": 1.2292490118577075,
+      "grad_norm": 2.3005123138427734,
+      "learning_rate": 4.402777777777778e-05,
+      "loss": 1.8875,
+      "step": 311
+    },
+    {
+      "epoch": 1.233201581027668,
+      "grad_norm": 2.1108436584472656,
+      "learning_rate": 4.400793650793651e-05,
+      "loss": 1.8508,
+      "step": 312
+    },
+    {
+      "epoch": 1.2371541501976284,
+      "grad_norm": 2.7670435905456543,
+      "learning_rate": 4.398809523809524e-05,
+      "loss": 1.8427,
+      "step": 313
+    },
+    {
+      "epoch": 1.2411067193675889,
+      "grad_norm": 2.1103644371032715,
+      "learning_rate": 4.396825396825397e-05,
+      "loss": 1.5623,
+      "step": 314
+    },
+    {
+      "epoch": 1.2450592885375493,
+      "grad_norm": 2.517120838165283,
+      "learning_rate": 4.39484126984127e-05,
+      "loss": 1.688,
+      "step": 315
+    },
+    {
+      "epoch": 1.2490118577075098,
+      "grad_norm": 1.799034595489502,
+      "learning_rate": 4.392857142857143e-05,
+      "loss": 1.7609,
+      "step": 316
+    },
+    {
+      "epoch": 1.2529644268774702,
+      "grad_norm": 2.450300455093384,
+      "learning_rate": 4.390873015873016e-05,
+      "loss": 1.6296,
+      "step": 317
+    },
+    {
+      "epoch": 1.256916996047431,
+      "grad_norm": 2.148935317993164,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 1.849,
+      "step": 318
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 2.3157055377960205,
+      "learning_rate": 4.3869047619047624e-05,
+      "loss": 1.62,
+      "step": 319
+    },
+    {
+      "epoch": 1.2648221343873518,
+      "grad_norm": 1.6752326488494873,
+      "learning_rate": 4.384920634920635e-05,
+      "loss": 1.594,
+      "step": 320
+    },
+    {
+      "epoch": 1.2687747035573123,
+      "grad_norm": 2.3844189643859863,
+      "learning_rate": 4.382936507936508e-05,
+      "loss": 2.033,
+      "step": 321
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 2.3986222743988037,
+      "learning_rate": 4.380952380952381e-05,
+      "loss": 1.4159,
+      "step": 322
+    },
+    {
+      "epoch": 1.2766798418972332,
+      "grad_norm": 1.919364094734192,
+      "learning_rate": 4.378968253968254e-05,
+      "loss": 1.9236,
+      "step": 323
+    },
+    {
+      "epoch": 1.2806324110671936,
+      "grad_norm": 2.3783788681030273,
+      "learning_rate": 4.376984126984127e-05,
+      "loss": 1.8965,
+      "step": 324
+    },
+    {
+      "epoch": 1.2845849802371543,
+      "grad_norm": 2.122201919555664,
+      "learning_rate": 4.375e-05,
+      "loss": 1.7124,
+      "step": 325
+    },
+    {
+      "epoch": 1.2885375494071147,
+      "grad_norm": 1.829590082168579,
+      "learning_rate": 4.373015873015873e-05,
+      "loss": 1.7605,
+      "step": 326
+    },
+    {
+      "epoch": 1.2924901185770752,
+      "grad_norm": 1.9862433671951294,
+      "learning_rate": 4.371031746031746e-05,
+      "loss": 1.6903,
+      "step": 327
+    },
+    {
+      "epoch": 1.2964426877470356,
+      "grad_norm": 2.1671435832977295,
+      "learning_rate": 4.3690476190476194e-05,
+      "loss": 1.9038,
+      "step": 328
+    },
+    {
+      "epoch": 1.300395256916996,
+      "grad_norm": 2.0836524963378906,
+      "learning_rate": 4.367063492063492e-05,
+      "loss": 1.8675,
+      "step": 329
+    },
+    {
+      "epoch": 1.3043478260869565,
+      "grad_norm": 2.2062249183654785,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 1.7196,
+      "step": 330
+    },
+    {
+      "epoch": 1.308300395256917,
+      "grad_norm": 2.018446683883667,
+      "learning_rate": 4.363095238095238e-05,
+      "loss": 1.6995,
+      "step": 331
+    },
+    {
+      "epoch": 1.3122529644268774,
+      "grad_norm": 2.521754026412964,
+      "learning_rate": 4.3611111111111116e-05,
+      "loss": 1.649,
+      "step": 332
+    },
+    {
+      "epoch": 1.316205533596838,
+      "grad_norm": 2.223747968673706,
+      "learning_rate": 4.359126984126984e-05,
+      "loss": 1.5682,
+      "step": 333
+    },
+    {
+      "epoch": 1.3201581027667983,
+      "grad_norm": 2.356834650039673,
+      "learning_rate": 4.3571428571428576e-05,
+      "loss": 1.7477,
+      "step": 334
+    },
+    {
+      "epoch": 1.3241106719367588,
+      "grad_norm": 1.9272010326385498,
+      "learning_rate": 4.35515873015873e-05,
+      "loss": 1.8585,
+      "step": 335
+    },
+    {
+      "epoch": 1.3280632411067192,
+      "grad_norm": 2.0459604263305664,
+      "learning_rate": 4.353174603174604e-05,
+      "loss": 1.5119,
+      "step": 336
+    },
+    {
+      "epoch": 1.33201581027668,
+      "grad_norm": 1.9021974802017212,
+      "learning_rate": 4.3511904761904764e-05,
+      "loss": 1.9379,
+      "step": 337
+    },
+    {
+      "epoch": 1.3359683794466404,
+      "grad_norm": 2.5137360095977783,
+      "learning_rate": 4.34920634920635e-05,
+      "loss": 1.8224,
+      "step": 338
+    },
+    {
+      "epoch": 1.3399209486166008,
+      "grad_norm": 1.5489245653152466,
+      "learning_rate": 4.3472222222222225e-05,
+      "loss": 1.8488,
+      "step": 339
+    },
+    {
+      "epoch": 1.3438735177865613,
+      "grad_norm": 1.811347484588623,
+      "learning_rate": 4.345238095238096e-05,
+      "loss": 1.5281,
+      "step": 340
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 2.5799477100372314,
+      "learning_rate": 4.3432539682539686e-05,
+      "loss": 1.7175,
+      "step": 341
+    },
+    {
+      "epoch": 1.3517786561264822,
+      "grad_norm": 2.4223809242248535,
+      "learning_rate": 4.341269841269842e-05,
+      "loss": 1.6518,
+      "step": 342
+    },
+    {
+      "epoch": 1.3557312252964426,
+      "grad_norm": 2.713998556137085,
+      "learning_rate": 4.3392857142857146e-05,
+      "loss": 1.6456,
+      "step": 343
+    },
+    {
+      "epoch": 1.359683794466403,
+      "grad_norm": 2.6156938076019287,
+      "learning_rate": 4.337301587301587e-05,
+      "loss": 1.6932,
+      "step": 344
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 2.451413154602051,
+      "learning_rate": 4.335317460317461e-05,
+      "loss": 1.7987,
+      "step": 345
+    },
+    {
+      "epoch": 1.3675889328063242,
+      "grad_norm": 1.895689845085144,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 1.7549,
+      "step": 346
+    },
+    {
+      "epoch": 1.3715415019762847,
+      "grad_norm": 2.4354727268218994,
+      "learning_rate": 4.331349206349207e-05,
+      "loss": 1.5661,
+      "step": 347
+    },
+    {
+      "epoch": 1.3754940711462451,
+      "grad_norm": 2.5790798664093018,
+      "learning_rate": 4.3293650793650795e-05,
+      "loss": 1.89,
+      "step": 348
+    },
+    {
+      "epoch": 1.3794466403162056,
+      "grad_norm": 1.9985342025756836,
+      "learning_rate": 4.327380952380953e-05,
+      "loss": 1.5705,
+      "step": 349
+    },
+    {
+      "epoch": 1.383399209486166,
+      "grad_norm": 1.87380850315094,
+      "learning_rate": 4.3253968253968256e-05,
+      "loss": 1.8884,
+      "step": 350
+    },
+    {
+      "epoch": 1.3873517786561265,
+      "grad_norm": 2.4583888053894043,
+      "learning_rate": 4.323412698412699e-05,
+      "loss": 1.51,
+      "step": 351
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 2.0746865272521973,
+      "learning_rate": 4.3214285714285716e-05,
+      "loss": 1.688,
+      "step": 352
+    },
+    {
+      "epoch": 1.3952569169960474,
+      "grad_norm": 2.3721249103546143,
+      "learning_rate": 4.319444444444445e-05,
+      "loss": 1.2734,
+      "step": 353
+    },
+    {
+      "epoch": 1.3992094861660078,
+      "grad_norm": 2.44508695602417,
+      "learning_rate": 4.317460317460318e-05,
+      "loss": 1.8527,
+      "step": 354
+    },
+    {
+      "epoch": 1.4031620553359683,
+      "grad_norm": 2.2311763763427734,
+      "learning_rate": 4.315476190476191e-05,
+      "loss": 1.6745,
+      "step": 355
+    },
+    {
+      "epoch": 1.4071146245059287,
+      "grad_norm": 1.867505431175232,
+      "learning_rate": 4.313492063492064e-05,
+      "loss": 1.9571,
+      "step": 356
+    },
+    {
+      "epoch": 1.4110671936758894,
+      "grad_norm": 2.2237112522125244,
+      "learning_rate": 4.311507936507937e-05,
+      "loss": 2.133,
+      "step": 357
+    },
+    {
+      "epoch": 1.4150197628458498,
+      "grad_norm": 1.834166169166565,
+      "learning_rate": 4.30952380952381e-05,
+      "loss": 2.0071,
+      "step": 358
+    },
+    {
+      "epoch": 1.4189723320158103,
+      "grad_norm": 1.9390510320663452,
+      "learning_rate": 4.307539682539683e-05,
+      "loss": 1.5187,
+      "step": 359
+    },
+    {
+      "epoch": 1.4229249011857708,
+      "grad_norm": 2.7471730709075928,
+      "learning_rate": 4.305555555555556e-05,
+      "loss": 1.7342,
+      "step": 360
+    },
+    {
+      "epoch": 1.4268774703557312,
+      "grad_norm": 1.8465523719787598,
+      "learning_rate": 4.303571428571429e-05,
+      "loss": 1.9512,
+      "step": 361
+    },
+    {
+      "epoch": 1.4308300395256917,
+      "grad_norm": 2.100477933883667,
+      "learning_rate": 4.301587301587302e-05,
+      "loss": 1.6373,
+      "step": 362
+    },
+    {
+      "epoch": 1.434782608695652,
+      "grad_norm": 2.298740863800049,
+      "learning_rate": 4.2996031746031754e-05,
+      "loss": 1.7411,
+      "step": 363
+    },
+    {
+      "epoch": 1.4387351778656128,
+      "grad_norm": 1.8366012573242188,
+      "learning_rate": 4.297619047619048e-05,
+      "loss": 1.89,
+      "step": 364
+    },
+    {
+      "epoch": 1.4426877470355732,
+      "grad_norm": 2.246143341064453,
+      "learning_rate": 4.295634920634921e-05,
+      "loss": 1.6918,
+      "step": 365
+    },
+    {
+      "epoch": 1.4466403162055337,
+      "grad_norm": 1.8526705503463745,
+      "learning_rate": 4.2936507936507935e-05,
+      "loss": 1.6653,
+      "step": 366
+    },
+    {
+      "epoch": 1.4505928853754941,
+      "grad_norm": 2.273653984069824,
+      "learning_rate": 4.291666666666667e-05,
+      "loss": 2.0948,
+      "step": 367
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 2.1943068504333496,
+      "learning_rate": 4.2896825396825396e-05,
+      "loss": 1.6798,
+      "step": 368
+    },
+    {
+      "epoch": 1.458498023715415,
+      "grad_norm": 1.9659279584884644,
+      "learning_rate": 4.287698412698413e-05,
+      "loss": 1.9008,
+      "step": 369
+    },
+    {
+      "epoch": 1.4624505928853755,
+      "grad_norm": 1.792734980583191,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 1.6039,
+      "step": 370
+    },
+    {
+      "epoch": 1.466403162055336,
+      "grad_norm": 2.0952234268188477,
+      "learning_rate": 4.283730158730159e-05,
+      "loss": 1.5815,
+      "step": 371
+    },
+    {
+      "epoch": 1.4703557312252964,
+      "grad_norm": 1.9501980543136597,
+      "learning_rate": 4.281746031746032e-05,
+      "loss": 1.5774,
+      "step": 372
+    },
+    {
+      "epoch": 1.4743083003952568,
+      "grad_norm": 1.960719347000122,
+      "learning_rate": 4.279761904761905e-05,
+      "loss": 1.5171,
+      "step": 373
+    },
+    {
+      "epoch": 1.4782608695652173,
+      "grad_norm": 2.022639751434326,
+      "learning_rate": 4.277777777777778e-05,
+      "loss": 1.8212,
+      "step": 374
+    },
+    {
+      "epoch": 1.4822134387351777,
+      "grad_norm": 1.9371771812438965,
+      "learning_rate": 4.2757936507936505e-05,
+      "loss": 1.6054,
+      "step": 375
+    },
+    {
+      "epoch": 1.4861660079051384,
+      "grad_norm": 2.072070598602295,
+      "learning_rate": 4.273809523809524e-05,
+      "loss": 1.6589,
+      "step": 376
+    },
+    {
+      "epoch": 1.4901185770750989,
+      "grad_norm": 1.8878108263015747,
+      "learning_rate": 4.2718253968253966e-05,
+      "loss": 1.9921,
+      "step": 377
+    },
+    {
+      "epoch": 1.4940711462450593,
+      "grad_norm": 1.9379626512527466,
+      "learning_rate": 4.26984126984127e-05,
+      "loss": 1.8345,
+      "step": 378
+    },
+    {
+      "epoch": 1.4980237154150198,
+      "grad_norm": 2.775383234024048,
+      "learning_rate": 4.2678571428571426e-05,
+      "loss": 1.8299,
+      "step": 379
+    },
+    {
+      "epoch": 1.5019762845849802,
+      "grad_norm": 2.041731119155884,
+      "learning_rate": 4.265873015873016e-05,
+      "loss": 1.4022,
+      "step": 380
+    },
+    {
+      "epoch": 1.5059288537549407,
+      "grad_norm": 2.0958447456359863,
+      "learning_rate": 4.263888888888889e-05,
+      "loss": 1.8758,
+      "step": 381
+    },
+    {
+      "epoch": 1.5098814229249014,
+      "grad_norm": 2.6933484077453613,
+      "learning_rate": 4.261904761904762e-05,
+      "loss": 2.0389,
+      "step": 382
+    },
+    {
+      "epoch": 1.5138339920948618,
+      "grad_norm": 2.594853401184082,
+      "learning_rate": 4.259920634920635e-05,
+      "loss": 1.5752,
+      "step": 383
+    },
+    {
+      "epoch": 1.5177865612648223,
+      "grad_norm": 2.6426239013671875,
+      "learning_rate": 4.257936507936508e-05,
+      "loss": 1.9273,
+      "step": 384
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 1.837033748626709,
+      "learning_rate": 4.255952380952381e-05,
+      "loss": 1.774,
+      "step": 385
+    },
+    {
+      "epoch": 1.5256916996047432,
+      "grad_norm": 2.4687540531158447,
+      "learning_rate": 4.253968253968254e-05,
+      "loss": 2.3856,
+      "step": 386
+    },
+    {
+      "epoch": 1.5296442687747036,
+      "grad_norm": 1.7977819442749023,
+      "learning_rate": 4.251984126984127e-05,
+      "loss": 1.6884,
+      "step": 387
+    },
+    {
+      "epoch": 1.533596837944664,
+      "grad_norm": 2.1651384830474854,
+      "learning_rate": 4.25e-05,
+      "loss": 2.0945,
+      "step": 388
+    },
+    {
+      "epoch": 1.5375494071146245,
+      "grad_norm": 1.796292781829834,
+      "learning_rate": 4.248015873015873e-05,
+      "loss": 1.7548,
+      "step": 389
+    },
+    {
+      "epoch": 1.541501976284585,
+      "grad_norm": 3.0309319496154785,
+      "learning_rate": 4.2460317460317464e-05,
+      "loss": 1.9628,
+      "step": 390
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 1.8279919624328613,
+      "learning_rate": 4.244047619047619e-05,
+      "loss": 1.9005,
+      "step": 391
+    },
+    {
+      "epoch": 1.5494071146245059,
+      "grad_norm": 1.9936280250549316,
+      "learning_rate": 4.2420634920634925e-05,
+      "loss": 1.7502,
+      "step": 392
+    },
+    {
+      "epoch": 1.5533596837944663,
+      "grad_norm": 1.728090763092041,
+      "learning_rate": 4.240079365079365e-05,
+      "loss": 1.8107,
+      "step": 393
+    },
+    {
+      "epoch": 1.5573122529644268,
+      "grad_norm": 1.614793300628662,
+      "learning_rate": 4.2380952380952385e-05,
+      "loss": 1.6821,
+      "step": 394
+    },
+    {
+      "epoch": 1.5612648221343872,
+      "grad_norm": 2.016058921813965,
+      "learning_rate": 4.236111111111111e-05,
+      "loss": 1.6821,
+      "step": 395
+    },
+    {
+      "epoch": 1.5652173913043477,
+      "grad_norm": 2.421929359436035,
+      "learning_rate": 4.2341269841269846e-05,
+      "loss": 2.0802,
+      "step": 396
+    },
+    {
+      "epoch": 1.5691699604743083,
+      "grad_norm": 1.9785467386245728,
+      "learning_rate": 4.232142857142857e-05,
+      "loss": 1.7954,
+      "step": 397
+    },
+    {
+      "epoch": 1.5731225296442688,
+      "grad_norm": 2.4265410900115967,
+      "learning_rate": 4.23015873015873e-05,
+      "loss": 1.6911,
+      "step": 398
+    },
+    {
+      "epoch": 1.5770750988142292,
+      "grad_norm": 2.7018771171569824,
+      "learning_rate": 4.2281746031746034e-05,
+      "loss": 1.7161,
+      "step": 399
+    },
+    {
+      "epoch": 1.5810276679841897,
+      "grad_norm": 1.9163415431976318,
+      "learning_rate": 4.226190476190476e-05,
+      "loss": 1.6967,
+      "step": 400
+    },
+    {
+      "epoch": 1.5849802371541502,
+      "grad_norm": 1.7845193147659302,
+      "learning_rate": 4.2242063492063495e-05,
+      "loss": 2.0456,
+      "step": 401
+    },
+    {
+      "epoch": 1.5889328063241108,
+      "grad_norm": 2.364121675491333,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 1.8798,
+      "step": 402
+    },
+    {
+      "epoch": 1.5928853754940713,
+      "grad_norm": 1.998205542564392,
+      "learning_rate": 4.2202380952380955e-05,
+      "loss": 1.6054,
+      "step": 403
+    },
+    {
+      "epoch": 1.5968379446640317,
+      "grad_norm": 2.847519874572754,
+      "learning_rate": 4.218253968253968e-05,
+      "loss": 1.723,
+      "step": 404
+    },
+    {
+      "epoch": 1.6007905138339922,
+      "grad_norm": 2.2237510681152344,
+      "learning_rate": 4.2162698412698416e-05,
+      "loss": 1.5771,
+      "step": 405
+    },
+    {
+      "epoch": 1.6047430830039526,
+      "grad_norm": 2.4652063846588135,
+      "learning_rate": 4.214285714285714e-05,
+      "loss": 1.9165,
+      "step": 406
+    },
+    {
+      "epoch": 1.608695652173913,
+      "grad_norm": 1.9992256164550781,
+      "learning_rate": 4.212301587301588e-05,
+      "loss": 1.9398,
+      "step": 407
+    },
+    {
+      "epoch": 1.6126482213438735,
+      "grad_norm": 1.7840253114700317,
+      "learning_rate": 4.2103174603174604e-05,
+      "loss": 1.623,
+      "step": 408
+    },
+    {
+      "epoch": 1.616600790513834,
+      "grad_norm": 1.5577056407928467,
+      "learning_rate": 4.208333333333334e-05,
+      "loss": 1.6815,
+      "step": 409
+    },
+    {
+      "epoch": 1.6205533596837944,
+      "grad_norm": 2.6008100509643555,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 1.6165,
+      "step": 410
+    },
+    {
+      "epoch": 1.6245059288537549,
+      "grad_norm": 1.9897792339324951,
+      "learning_rate": 4.20436507936508e-05,
+      "loss": 1.6959,
+      "step": 411
+    },
+    {
+      "epoch": 1.6284584980237153,
+      "grad_norm": 1.9050272703170776,
+      "learning_rate": 4.2023809523809525e-05,
+      "loss": 1.9483,
+      "step": 412
+    },
+    {
+      "epoch": 1.6324110671936758,
+      "grad_norm": 2.1082065105438232,
+      "learning_rate": 4.200396825396826e-05,
+      "loss": 1.6326,
+      "step": 413
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 1.8587292432785034,
+      "learning_rate": 4.1984126984126986e-05,
+      "loss": 1.5015,
+      "step": 414
+    },
+    {
+      "epoch": 1.6403162055335967,
+      "grad_norm": 2.440154552459717,
+      "learning_rate": 4.196428571428572e-05,
+      "loss": 1.8917,
+      "step": 415
+    },
+    {
+      "epoch": 1.6442687747035574,
+      "grad_norm": 2.1441152095794678,
+      "learning_rate": 4.194444444444445e-05,
+      "loss": 1.8155,
+      "step": 416
+    },
+    {
+      "epoch": 1.6482213438735178,
+      "grad_norm": 1.9019662141799927,
+      "learning_rate": 4.192460317460318e-05,
+      "loss": 1.9238,
+      "step": 417
+    },
+    {
+      "epoch": 1.6521739130434783,
+      "grad_norm": 1.8844928741455078,
+      "learning_rate": 4.190476190476191e-05,
+      "loss": 1.6799,
+      "step": 418
+    },
+    {
+      "epoch": 1.6561264822134387,
+      "grad_norm": 2.373283624649048,
+      "learning_rate": 4.188492063492064e-05,
+      "loss": 1.6916,
+      "step": 419
+    },
+    {
+      "epoch": 1.6600790513833992,
+      "grad_norm": 2.1077182292938232,
+      "learning_rate": 4.186507936507937e-05,
+      "loss": 1.924,
+      "step": 420
+    },
+    {
+      "epoch": 1.6640316205533598,
+      "grad_norm": 2.5388450622558594,
+      "learning_rate": 4.1845238095238095e-05,
+      "loss": 1.7846,
+      "step": 421
+    },
+    {
+      "epoch": 1.6679841897233203,
+      "grad_norm": 2.8166000843048096,
+      "learning_rate": 4.182539682539683e-05,
+      "loss": 1.5143,
+      "step": 422
+    },
+    {
+      "epoch": 1.6719367588932808,
+      "grad_norm": 1.919756531715393,
+      "learning_rate": 4.1805555555555556e-05,
+      "loss": 1.7846,
+      "step": 423
+    },
+    {
+      "epoch": 1.6758893280632412,
+      "grad_norm": 2.265761375427246,
+      "learning_rate": 4.178571428571429e-05,
+      "loss": 1.2774,
+      "step": 424
+    },
+    {
+      "epoch": 1.6798418972332017,
+      "grad_norm": 2.4432921409606934,
+      "learning_rate": 4.176587301587302e-05,
+      "loss": 1.637,
+      "step": 425
+    },
+    {
+      "epoch": 1.683794466403162,
+      "grad_norm": 2.095233678817749,
+      "learning_rate": 4.174603174603175e-05,
+      "loss": 1.4832,
+      "step": 426
+    },
+    {
+      "epoch": 1.6877470355731226,
+      "grad_norm": 2.602231740951538,
+      "learning_rate": 4.172619047619048e-05,
+      "loss": 1.5574,
+      "step": 427
+    },
+    {
+      "epoch": 1.691699604743083,
+      "grad_norm": 1.9906206130981445,
+      "learning_rate": 4.170634920634921e-05,
+      "loss": 1.5909,
+      "step": 428
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 2.3297181129455566,
+      "learning_rate": 4.168650793650794e-05,
+      "loss": 1.7061,
+      "step": 429
+    },
+    {
+      "epoch": 1.699604743083004,
+      "grad_norm": 2.3630523681640625,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.6981,
+      "step": 430
+    },
+    {
+      "epoch": 1.7035573122529644,
+      "grad_norm": 1.9802945852279663,
+      "learning_rate": 4.16468253968254e-05,
+      "loss": 1.7784,
+      "step": 431
+    },
+    {
+      "epoch": 1.7075098814229248,
+      "grad_norm": 2.8566882610321045,
+      "learning_rate": 4.162698412698413e-05,
+      "loss": 1.5392,
+      "step": 432
+    },
+    {
+      "epoch": 1.7114624505928853,
+      "grad_norm": 2.2854273319244385,
+      "learning_rate": 4.160714285714286e-05,
+      "loss": 1.7431,
+      "step": 433
+    },
+    {
+      "epoch": 1.7154150197628457,
+      "grad_norm": 1.9989619255065918,
+      "learning_rate": 4.1587301587301594e-05,
+      "loss": 1.8095,
+      "step": 434
+    },
+    {
+      "epoch": 1.7193675889328062,
+      "grad_norm": 2.6097142696380615,
+      "learning_rate": 4.156746031746032e-05,
+      "loss": 1.464,
+      "step": 435
+    },
+    {
+      "epoch": 1.7233201581027668,
+      "grad_norm": 2.048234462738037,
+      "learning_rate": 4.1547619047619054e-05,
+      "loss": 2.0557,
+      "step": 436
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 2.046173334121704,
+      "learning_rate": 4.152777777777778e-05,
+      "loss": 1.4132,
+      "step": 437
+    },
+    {
+      "epoch": 1.7312252964426877,
+      "grad_norm": 2.1724562644958496,
+      "learning_rate": 4.1507936507936515e-05,
+      "loss": 1.6773,
+      "step": 438
+    },
+    {
+      "epoch": 1.7351778656126482,
+      "grad_norm": 2.2624452114105225,
+      "learning_rate": 4.148809523809524e-05,
+      "loss": 1.6013,
+      "step": 439
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "grad_norm": 2.2074332237243652,
+      "learning_rate": 4.1468253968253976e-05,
+      "loss": 1.7686,
+      "step": 440
+    },
+    {
+      "epoch": 1.7430830039525693,
+      "grad_norm": 2.1703858375549316,
+      "learning_rate": 4.14484126984127e-05,
+      "loss": 1.5108,
+      "step": 441
+    },
+    {
+      "epoch": 1.7470355731225298,
+      "grad_norm": 1.9463564157485962,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 1.9879,
+      "step": 442
+    },
+    {
+      "epoch": 1.7509881422924902,
+      "grad_norm": 2.4043517112731934,
+      "learning_rate": 4.1408730158730164e-05,
+      "loss": 1.6002,
+      "step": 443
+    },
+    {
+      "epoch": 1.7549407114624507,
+      "grad_norm": 1.906467080116272,
+      "learning_rate": 4.138888888888889e-05,
+      "loss": 1.8149,
+      "step": 444
+    },
+    {
+      "epoch": 1.7588932806324111,
+      "grad_norm": 2.612128257751465,
+      "learning_rate": 4.136904761904762e-05,
+      "loss": 1.8805,
+      "step": 445
+    },
+    {
+      "epoch": 1.7628458498023716,
+      "grad_norm": 2.056048631668091,
+      "learning_rate": 4.134920634920635e-05,
+      "loss": 1.6064,
+      "step": 446
+    },
+    {
+      "epoch": 1.766798418972332,
+      "grad_norm": 1.9283335208892822,
+      "learning_rate": 4.132936507936508e-05,
+      "loss": 2.2151,
+      "step": 447
+    },
+    {
+      "epoch": 1.7707509881422925,
+      "grad_norm": 1.8714007139205933,
+      "learning_rate": 4.130952380952381e-05,
+      "loss": 1.6642,
+      "step": 448
+    },
+    {
+      "epoch": 1.774703557312253,
+      "grad_norm": 2.0939621925354004,
+      "learning_rate": 4.128968253968254e-05,
+      "loss": 1.558,
+      "step": 449
+    },
+    {
+      "epoch": 1.7786561264822134,
+      "grad_norm": 2.0754241943359375,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 1.6465,
+      "step": 450
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 2.7905728816986084,
+      "learning_rate": 4.125e-05,
+      "loss": 1.7298,
+      "step": 451
+    },
+    {
+      "epoch": 1.7865612648221343,
+      "grad_norm": 2.2117226123809814,
+      "learning_rate": 4.123015873015873e-05,
+      "loss": 1.5478,
+      "step": 452
+    },
+    {
+      "epoch": 1.7905138339920947,
+      "grad_norm": 2.22251033782959,
+      "learning_rate": 4.121031746031746e-05,
+      "loss": 1.4483,
+      "step": 453
+    },
+    {
+      "epoch": 1.7944664031620552,
+      "grad_norm": 2.636388063430786,
+      "learning_rate": 4.119047619047619e-05,
+      "loss": 1.3413,
+      "step": 454
+    },
+    {
+      "epoch": 1.7984189723320159,
+      "grad_norm": 2.417893409729004,
+      "learning_rate": 4.117063492063492e-05,
+      "loss": 1.5273,
+      "step": 455
+    },
+    {
+      "epoch": 1.8023715415019763,
+      "grad_norm": 1.960670828819275,
+      "learning_rate": 4.115079365079365e-05,
+      "loss": 1.8887,
+      "step": 456
+    },
+    {
+      "epoch": 1.8063241106719368,
+      "grad_norm": 2.0714080333709717,
+      "learning_rate": 4.113095238095238e-05,
+      "loss": 1.9395,
+      "step": 457
+    },
+    {
+      "epoch": 1.8102766798418972,
+      "grad_norm": 2.6367220878601074,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 1.475,
+      "step": 458
+    },
+    {
+      "epoch": 1.8142292490118577,
+      "grad_norm": 2.8209869861602783,
+      "learning_rate": 4.109126984126984e-05,
+      "loss": 1.4326,
+      "step": 459
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 1.884318470954895,
+      "learning_rate": 4.107142857142857e-05,
+      "loss": 1.5297,
+      "step": 460
+    },
+    {
+      "epoch": 1.8221343873517788,
+      "grad_norm": 1.7990609407424927,
+      "learning_rate": 4.1051587301587304e-05,
+      "loss": 1.9113,
+      "step": 461
+    },
+    {
+      "epoch": 1.8260869565217392,
+      "grad_norm": 1.8307974338531494,
+      "learning_rate": 4.103174603174603e-05,
+      "loss": 1.5365,
+      "step": 462
+    },
+    {
+      "epoch": 1.8300395256916997,
+      "grad_norm": 2.2995738983154297,
+      "learning_rate": 4.1011904761904764e-05,
+      "loss": 1.4421,
+      "step": 463
+    },
+    {
+      "epoch": 1.8339920948616601,
+      "grad_norm": 2.2058722972869873,
+      "learning_rate": 4.099206349206349e-05,
+      "loss": 1.5384,
+      "step": 464
+    },
+    {
+      "epoch": 1.8379446640316206,
+      "grad_norm": 2.011327028274536,
+      "learning_rate": 4.0972222222222225e-05,
+      "loss": 1.6523,
+      "step": 465
+    },
+    {
+      "epoch": 1.841897233201581,
+      "grad_norm": 2.0535199642181396,
+      "learning_rate": 4.095238095238095e-05,
+      "loss": 1.6645,
+      "step": 466
+    },
+    {
+      "epoch": 1.8458498023715415,
+      "grad_norm": 2.485356092453003,
+      "learning_rate": 4.0932539682539686e-05,
+      "loss": 1.5562,
+      "step": 467
+    },
+    {
+      "epoch": 1.849802371541502,
+      "grad_norm": 2.368373155593872,
+      "learning_rate": 4.091269841269841e-05,
+      "loss": 1.806,
+      "step": 468
+    },
+    {
+      "epoch": 1.8537549407114624,
+      "grad_norm": 2.3104066848754883,
+      "learning_rate": 4.0892857142857147e-05,
+      "loss": 1.9633,
+      "step": 469
+    },
+    {
+      "epoch": 1.8577075098814229,
+      "grad_norm": 2.3505473136901855,
+      "learning_rate": 4.0873015873015874e-05,
+      "loss": 1.6175,
+      "step": 470
+    },
+    {
+      "epoch": 1.8616600790513833,
+      "grad_norm": 1.9146708250045776,
+      "learning_rate": 4.085317460317461e-05,
+      "loss": 1.5535,
+      "step": 471
+    },
+    {
+      "epoch": 1.8656126482213438,
+      "grad_norm": 1.624837040901184,
+      "learning_rate": 4.0833333333333334e-05,
+      "loss": 1.6677,
+      "step": 472
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 1.7742847204208374,
+      "learning_rate": 4.081349206349207e-05,
+      "loss": 1.5509,
+      "step": 473
+    },
+    {
+      "epoch": 1.8735177865612647,
+      "grad_norm": 1.9143329858779907,
+      "learning_rate": 4.0793650793650795e-05,
+      "loss": 1.8877,
+      "step": 474
+    },
+    {
+      "epoch": 1.8774703557312253,
+      "grad_norm": 2.1041526794433594,
+      "learning_rate": 4.077380952380952e-05,
+      "loss": 1.6796,
+      "step": 475
+    },
+    {
+      "epoch": 1.8814229249011858,
+      "grad_norm": 2.0280792713165283,
+      "learning_rate": 4.0753968253968256e-05,
+      "loss": 1.7892,
+      "step": 476
+    },
+    {
+      "epoch": 1.8853754940711462,
+      "grad_norm": 2.8694331645965576,
+      "learning_rate": 4.073412698412698e-05,
+      "loss": 1.9448,
+      "step": 477
+    },
+    {
+      "epoch": 1.8893280632411067,
+      "grad_norm": 2.471470832824707,
+      "learning_rate": 4.0714285714285717e-05,
+      "loss": 1.7856,
+      "step": 478
+    },
+    {
+      "epoch": 1.8932806324110671,
+      "grad_norm": 2.2044591903686523,
+      "learning_rate": 4.0694444444444444e-05,
+      "loss": 1.7688,
+      "step": 479
+    },
+    {
+      "epoch": 1.8972332015810278,
+      "grad_norm": 2.3773610591888428,
+      "learning_rate": 4.067460317460318e-05,
+      "loss": 1.908,
+      "step": 480
+    },
+    {
+      "epoch": 1.9011857707509883,
+      "grad_norm": 2.1004207134246826,
+      "learning_rate": 4.0654761904761904e-05,
+      "loss": 1.4418,
+      "step": 481
+    },
+    {
+      "epoch": 1.9051383399209487,
+      "grad_norm": 2.415409803390503,
+      "learning_rate": 4.063492063492064e-05,
+      "loss": 1.7695,
+      "step": 482
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 1.89458167552948,
+      "learning_rate": 4.0615079365079365e-05,
+      "loss": 1.8081,
+      "step": 483
+    },
+    {
+      "epoch": 1.9130434782608696,
+      "grad_norm": 2.560452699661255,
+      "learning_rate": 4.05952380952381e-05,
+      "loss": 2.0119,
+      "step": 484
+    },
+    {
+      "epoch": 1.91699604743083,
+      "grad_norm": 2.115086793899536,
+      "learning_rate": 4.0575396825396826e-05,
+      "loss": 1.7012,
+      "step": 485
+    },
+    {
+      "epoch": 1.9209486166007905,
+      "grad_norm": 2.1731040477752686,
+      "learning_rate": 4.055555555555556e-05,
+      "loss": 1.7267,
+      "step": 486
+    },
+    {
+      "epoch": 1.924901185770751,
+      "grad_norm": 1.9672155380249023,
+      "learning_rate": 4.0535714285714287e-05,
+      "loss": 1.7396,
+      "step": 487
+    },
+    {
+      "epoch": 1.9288537549407114,
+      "grad_norm": 2.380613327026367,
+      "learning_rate": 4.051587301587302e-05,
+      "loss": 1.4355,
+      "step": 488
+    },
+    {
+      "epoch": 1.9328063241106719,
+      "grad_norm": 2.087024450302124,
+      "learning_rate": 4.049603174603175e-05,
+      "loss": 1.812,
+      "step": 489
+    },
+    {
+      "epoch": 1.9367588932806323,
+      "grad_norm": 2.704306125640869,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 1.7029,
+      "step": 490
+    },
+    {
+      "epoch": 1.9407114624505928,
+      "grad_norm": 1.8128267526626587,
+      "learning_rate": 4.045634920634921e-05,
+      "loss": 1.616,
+      "step": 491
+    },
+    {
+      "epoch": 1.9446640316205532,
+      "grad_norm": 1.7133640050888062,
+      "learning_rate": 4.043650793650794e-05,
+      "loss": 1.4945,
+      "step": 492
+    },
+    {
+      "epoch": 1.9486166007905137,
+      "grad_norm": 1.8677951097488403,
+      "learning_rate": 4.041666666666667e-05,
+      "loss": 1.6594,
+      "step": 493
+    },
+    {
+      "epoch": 1.9525691699604744,
+      "grad_norm": 1.908327579498291,
+      "learning_rate": 4.03968253968254e-05,
+      "loss": 1.7492,
+      "step": 494
+    },
+    {
+      "epoch": 1.9565217391304348,
+      "grad_norm": 2.1175668239593506,
+      "learning_rate": 4.037698412698413e-05,
+      "loss": 1.6509,
+      "step": 495
+    },
+    {
+      "epoch": 1.9604743083003953,
+      "grad_norm": 2.284240961074829,
+      "learning_rate": 4.035714285714286e-05,
+      "loss": 1.4284,
+      "step": 496
+    },
+    {
+      "epoch": 1.9644268774703557,
+      "grad_norm": 1.581749677658081,
+      "learning_rate": 4.033730158730159e-05,
+      "loss": 1.8258,
+      "step": 497
+    },
+    {
+      "epoch": 1.9683794466403162,
+      "grad_norm": 2.200153112411499,
+      "learning_rate": 4.031746031746032e-05,
+      "loss": 1.573,
+      "step": 498
+    },
+    {
+      "epoch": 1.9723320158102768,
+      "grad_norm": 2.2217061519622803,
+      "learning_rate": 4.029761904761905e-05,
+      "loss": 1.6952,
+      "step": 499
+    },
+    {
+      "epoch": 1.9762845849802373,
+      "grad_norm": 2.2044413089752197,
+      "learning_rate": 4.027777777777778e-05,
+      "loss": 1.5415,
+      "step": 500
+    },
+    {
+      "epoch": 1.9802371541501977,
+      "grad_norm": 2.001845359802246,
+      "learning_rate": 4.025793650793651e-05,
+      "loss": 1.8791,
+      "step": 501
+    },
+    {
+      "epoch": 1.9841897233201582,
+      "grad_norm": 1.7531094551086426,
+      "learning_rate": 4.023809523809524e-05,
+      "loss": 1.7529,
+      "step": 502
+    },
+    {
+      "epoch": 1.9881422924901186,
+      "grad_norm": 3.204590320587158,
+      "learning_rate": 4.021825396825397e-05,
+      "loss": 1.6866,
+      "step": 503
+    },
+    {
+      "epoch": 1.992094861660079,
+      "grad_norm": 2.3752427101135254,
+      "learning_rate": 4.01984126984127e-05,
+      "loss": 1.4053,
+      "step": 504
+    },
+    {
+      "epoch": 1.9960474308300395,
+      "grad_norm": 1.789313793182373,
+      "learning_rate": 4.017857142857143e-05,
+      "loss": 1.6106,
+      "step": 505
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.152024269104004,
+      "learning_rate": 4.015873015873016e-05,
+      "loss": 1.7857,
+      "step": 506
+    },
+    {
+      "epoch": 2.0039525691699605,
+      "grad_norm": 1.7887808084487915,
+      "learning_rate": 4.0138888888888894e-05,
+      "loss": 1.2694,
+      "step": 507
+    },
+    {
+      "epoch": 2.007905138339921,
+      "grad_norm": 2.0157063007354736,
+      "learning_rate": 4.011904761904762e-05,
+      "loss": 1.3274,
+      "step": 508
+    },
+    {
+      "epoch": 2.0118577075098814,
+      "grad_norm": 2.462636709213257,
+      "learning_rate": 4.0099206349206355e-05,
+      "loss": 1.5154,
+      "step": 509
+    },
+    {
+      "epoch": 2.015810276679842,
+      "grad_norm": 2.299966812133789,
+      "learning_rate": 4.007936507936508e-05,
+      "loss": 1.378,
+      "step": 510
+    },
+    {
+      "epoch": 2.0197628458498023,
+      "grad_norm": 2.748854398727417,
+      "learning_rate": 4.0059523809523816e-05,
+      "loss": 1.5578,
+      "step": 511
+    },
+    {
+      "epoch": 2.0237154150197627,
+      "grad_norm": 2.8967697620391846,
+      "learning_rate": 4.003968253968254e-05,
+      "loss": 1.2523,
+      "step": 512
+    },
+    {
+      "epoch": 2.027667984189723,
+      "grad_norm": 3.2559096813201904,
+      "learning_rate": 4.0019841269841276e-05,
+      "loss": 1.3687,
+      "step": 513
+    },
+    {
+      "epoch": 2.0316205533596836,
+      "grad_norm": 3.0115792751312256,
+      "learning_rate": 4e-05,
+      "loss": 1.6751,
+      "step": 514
+    },
+    {
+      "epoch": 2.035573122529644,
+      "grad_norm": 2.7409708499908447,
+      "learning_rate": 3.998015873015874e-05,
+      "loss": 1.5692,
+      "step": 515
+    },
+    {
+      "epoch": 2.039525691699605,
+      "grad_norm": 2.488542318344116,
+      "learning_rate": 3.9960317460317464e-05,
+      "loss": 1.6312,
+      "step": 516
+    },
+    {
+      "epoch": 2.0434782608695654,
+      "grad_norm": 2.267803192138672,
+      "learning_rate": 3.99404761904762e-05,
+      "loss": 1.3411,
+      "step": 517
+    },
+    {
+      "epoch": 2.047430830039526,
+      "grad_norm": 2.1117396354675293,
+      "learning_rate": 3.9920634920634925e-05,
+      "loss": 1.3751,
+      "step": 518
+    },
+    {
+      "epoch": 2.0513833992094863,
+      "grad_norm": 2.5073764324188232,
+      "learning_rate": 3.990079365079366e-05,
+      "loss": 1.3868,
+      "step": 519
+    },
+    {
+      "epoch": 2.0553359683794468,
+      "grad_norm": 2.462871789932251,
+      "learning_rate": 3.9880952380952386e-05,
+      "loss": 1.2461,
+      "step": 520
+    },
+    {
+      "epoch": 2.059288537549407,
+      "grad_norm": 2.81300950050354,
+      "learning_rate": 3.986111111111111e-05,
+      "loss": 1.6863,
+      "step": 521
+    },
+    {
+      "epoch": 2.0632411067193677,
+      "grad_norm": 2.345595359802246,
+      "learning_rate": 3.984126984126984e-05,
+      "loss": 1.6056,
+      "step": 522
+    },
+    {
+      "epoch": 2.067193675889328,
+      "grad_norm": 3.0314559936523438,
+      "learning_rate": 3.982142857142857e-05,
+      "loss": 1.4429,
+      "step": 523
+    },
+    {
+      "epoch": 2.0711462450592886,
+      "grad_norm": 2.563462257385254,
+      "learning_rate": 3.98015873015873e-05,
+      "loss": 1.4249,
+      "step": 524
+    },
+    {
+      "epoch": 2.075098814229249,
+      "grad_norm": 3.088543653488159,
+      "learning_rate": 3.9781746031746034e-05,
+      "loss": 1.6226,
+      "step": 525
+    },
+    {
+      "epoch": 2.0790513833992095,
+      "grad_norm": 2.6812891960144043,
+      "learning_rate": 3.976190476190476e-05,
+      "loss": 1.4069,
+      "step": 526
+    },
+    {
+      "epoch": 2.08300395256917,
+      "grad_norm": 2.729356288909912,
+      "learning_rate": 3.9742063492063495e-05,
+      "loss": 1.4145,
+      "step": 527
+    },
+    {
+      "epoch": 2.0869565217391304,
+      "grad_norm": 1.9687360525131226,
+      "learning_rate": 3.972222222222222e-05,
+      "loss": 1.2861,
+      "step": 528
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 2.8874728679656982,
+      "learning_rate": 3.970238095238095e-05,
+      "loss": 1.0838,
+      "step": 529
+    },
+    {
+      "epoch": 2.0948616600790513,
+      "grad_norm": 2.1731650829315186,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 1.5371,
+      "step": 530
+    },
+    {
+      "epoch": 2.0988142292490117,
+      "grad_norm": 2.3168811798095703,
+      "learning_rate": 3.966269841269841e-05,
+      "loss": 1.6941,
+      "step": 531
+    },
+    {
+      "epoch": 2.102766798418972,
+      "grad_norm": 2.0623104572296143,
+      "learning_rate": 3.964285714285714e-05,
+      "loss": 1.2772,
+      "step": 532
+    },
+    {
+      "epoch": 2.1067193675889326,
+      "grad_norm": 3.4326181411743164,
+      "learning_rate": 3.962301587301587e-05,
+      "loss": 1.3626,
+      "step": 533
+    },
+    {
+      "epoch": 2.110671936758893,
+      "grad_norm": 2.1339476108551025,
+      "learning_rate": 3.9603174603174604e-05,
+      "loss": 1.4972,
+      "step": 534
+    },
+    {
+      "epoch": 2.1146245059288535,
+      "grad_norm": 2.1961209774017334,
+      "learning_rate": 3.958333333333333e-05,
+      "loss": 1.2863,
+      "step": 535
+    },
+    {
+      "epoch": 2.1185770750988144,
+      "grad_norm": 2.13753604888916,
+      "learning_rate": 3.9563492063492065e-05,
+      "loss": 1.3903,
+      "step": 536
+    },
+    {
+      "epoch": 2.122529644268775,
+      "grad_norm": 2.521846055984497,
+      "learning_rate": 3.954365079365079e-05,
+      "loss": 1.2818,
+      "step": 537
+    },
+    {
+      "epoch": 2.1264822134387353,
+      "grad_norm": 2.1714985370635986,
+      "learning_rate": 3.9523809523809526e-05,
+      "loss": 1.4502,
+      "step": 538
+    },
+    {
+      "epoch": 2.130434782608696,
+      "grad_norm": 3.092069149017334,
+      "learning_rate": 3.950396825396825e-05,
+      "loss": 1.5637,
+      "step": 539
+    },
+    {
+      "epoch": 2.1343873517786562,
+      "grad_norm": 2.6351075172424316,
+      "learning_rate": 3.9484126984126986e-05,
+      "loss": 1.4462,
+      "step": 540
+    },
+    {
+      "epoch": 2.1383399209486167,
+      "grad_norm": 1.9999125003814697,
+      "learning_rate": 3.946428571428571e-05,
+      "loss": 1.4445,
+      "step": 541
+    },
+    {
+      "epoch": 2.142292490118577,
+      "grad_norm": 3.0483391284942627,
+      "learning_rate": 3.944444444444445e-05,
+      "loss": 1.206,
+      "step": 542
+    },
+    {
+      "epoch": 2.1462450592885376,
+      "grad_norm": 2.455862522125244,
+      "learning_rate": 3.9424603174603174e-05,
+      "loss": 1.3128,
+      "step": 543
+    },
+    {
+      "epoch": 2.150197628458498,
+      "grad_norm": 2.6878578662872314,
+      "learning_rate": 3.940476190476191e-05,
+      "loss": 1.3592,
+      "step": 544
+    },
+    {
+      "epoch": 2.1541501976284585,
+      "grad_norm": 2.25331449508667,
+      "learning_rate": 3.9384920634920635e-05,
+      "loss": 1.2749,
+      "step": 545
+    },
+    {
+      "epoch": 2.158102766798419,
+      "grad_norm": 2.7239346504211426,
+      "learning_rate": 3.936507936507937e-05,
+      "loss": 1.0487,
+      "step": 546
+    },
+    {
+      "epoch": 2.1620553359683794,
+      "grad_norm": 2.8506832122802734,
+      "learning_rate": 3.9345238095238096e-05,
+      "loss": 1.2713,
+      "step": 547
+    },
+    {
+      "epoch": 2.16600790513834,
+      "grad_norm": 2.3467752933502197,
+      "learning_rate": 3.932539682539683e-05,
+      "loss": 1.1959,
+      "step": 548
+    },
+    {
+      "epoch": 2.1699604743083003,
+      "grad_norm": 1.99912428855896,
+      "learning_rate": 3.9305555555555556e-05,
+      "loss": 1.2617,
+      "step": 549
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 2.6946754455566406,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 1.5949,
+      "step": 550
+    },
+    {
+      "epoch": 2.177865612648221,
+      "grad_norm": 2.771515369415283,
+      "learning_rate": 3.926587301587302e-05,
+      "loss": 1.4533,
+      "step": 551
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 2.188755989074707,
+      "learning_rate": 3.9246031746031744e-05,
+      "loss": 1.3215,
+      "step": 552
+    },
+    {
+      "epoch": 2.185770750988142,
+      "grad_norm": 2.7638049125671387,
+      "learning_rate": 3.922619047619048e-05,
+      "loss": 1.1885,
+      "step": 553
+    },
+    {
+      "epoch": 2.1897233201581026,
+      "grad_norm": 2.2438464164733887,
+      "learning_rate": 3.9206349206349205e-05,
+      "loss": 1.2809,
+      "step": 554
+    },
+    {
+      "epoch": 2.1936758893280635,
+      "grad_norm": 2.8813867568969727,
+      "learning_rate": 3.918650793650794e-05,
+      "loss": 1.3764,
+      "step": 555
+    },
+    {
+      "epoch": 2.197628458498024,
+      "grad_norm": 2.219583034515381,
+      "learning_rate": 3.9166666666666665e-05,
+      "loss": 1.5119,
+      "step": 556
+    },
+    {
+      "epoch": 2.2015810276679844,
+      "grad_norm": 2.888500452041626,
+      "learning_rate": 3.91468253968254e-05,
+      "loss": 1.5976,
+      "step": 557
+    },
+    {
+      "epoch": 2.205533596837945,
+      "grad_norm": 2.0659193992614746,
+      "learning_rate": 3.9126984126984126e-05,
+      "loss": 1.3095,
+      "step": 558
+    },
+    {
+      "epoch": 2.2094861660079053,
+      "grad_norm": 2.3203117847442627,
+      "learning_rate": 3.910714285714286e-05,
+      "loss": 1.2698,
+      "step": 559
+    },
+    {
+      "epoch": 2.2134387351778657,
+      "grad_norm": 2.8916614055633545,
+      "learning_rate": 3.908730158730159e-05,
+      "loss": 1.4077,
+      "step": 560
+    },
+    {
+      "epoch": 2.217391304347826,
+      "grad_norm": 2.6932151317596436,
+      "learning_rate": 3.906746031746032e-05,
+      "loss": 1.1033,
+      "step": 561
+    },
+    {
+      "epoch": 2.2213438735177866,
+      "grad_norm": 2.4444422721862793,
+      "learning_rate": 3.904761904761905e-05,
+      "loss": 1.2782,
+      "step": 562
+    },
+    {
+      "epoch": 2.225296442687747,
+      "grad_norm": 2.334657669067383,
+      "learning_rate": 3.902777777777778e-05,
+      "loss": 1.5071,
+      "step": 563
+    },
+    {
+      "epoch": 2.2292490118577075,
+      "grad_norm": 2.4118669033050537,
+      "learning_rate": 3.900793650793651e-05,
+      "loss": 1.2665,
+      "step": 564
+    },
+    {
+      "epoch": 2.233201581027668,
+      "grad_norm": 2.160811185836792,
+      "learning_rate": 3.898809523809524e-05,
+      "loss": 1.4489,
+      "step": 565
+    },
+    {
+      "epoch": 2.2371541501976284,
+      "grad_norm": 2.49159574508667,
+      "learning_rate": 3.896825396825397e-05,
+      "loss": 1.3127,
+      "step": 566
+    },
+    {
+      "epoch": 2.241106719367589,
+      "grad_norm": 2.5408575534820557,
+      "learning_rate": 3.89484126984127e-05,
+      "loss": 1.0498,
+      "step": 567
+    },
+    {
+      "epoch": 2.2450592885375493,
+      "grad_norm": 2.121267557144165,
+      "learning_rate": 3.892857142857143e-05,
+      "loss": 1.3147,
+      "step": 568
+    },
+    {
+      "epoch": 2.2490118577075098,
+      "grad_norm": 2.7093420028686523,
+      "learning_rate": 3.8908730158730164e-05,
+      "loss": 1.3408,
+      "step": 569
+    },
+    {
+      "epoch": 2.2529644268774702,
+      "grad_norm": 3.397962808609009,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 1.4971,
+      "step": 570
+    },
+    {
+      "epoch": 2.2569169960474307,
+      "grad_norm": 2.7930893898010254,
+      "learning_rate": 3.8869047619047625e-05,
+      "loss": 1.4294,
+      "step": 571
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 2.7343547344207764,
+      "learning_rate": 3.884920634920635e-05,
+      "loss": 0.9542,
+      "step": 572
+    },
+    {
+      "epoch": 2.2648221343873516,
+      "grad_norm": 2.365590810775757,
+      "learning_rate": 3.8829365079365085e-05,
+      "loss": 1.2086,
+      "step": 573
+    },
+    {
+      "epoch": 2.2687747035573125,
+      "grad_norm": 2.1030609607696533,
+      "learning_rate": 3.880952380952381e-05,
+      "loss": 1.3605,
+      "step": 574
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 2.1257293224334717,
+      "learning_rate": 3.878968253968254e-05,
+      "loss": 1.4352,
+      "step": 575
+    },
+    {
+      "epoch": 2.2766798418972334,
+      "grad_norm": 2.316331386566162,
+      "learning_rate": 3.876984126984127e-05,
+      "loss": 1.4796,
+      "step": 576
+    },
+    {
+      "epoch": 2.280632411067194,
+      "grad_norm": 2.8514187335968018,
+      "learning_rate": 3.875e-05,
+      "loss": 1.3073,
+      "step": 577
+    },
+    {
+      "epoch": 2.2845849802371543,
+      "grad_norm": 2.8949341773986816,
+      "learning_rate": 3.8730158730158734e-05,
+      "loss": 1.3915,
+      "step": 578
+    },
+    {
+      "epoch": 2.2885375494071147,
+      "grad_norm": 2.150175094604492,
+      "learning_rate": 3.871031746031746e-05,
+      "loss": 1.2567,
+      "step": 579
+    },
+    {
+      "epoch": 2.292490118577075,
+      "grad_norm": 2.0244216918945312,
+      "learning_rate": 3.8690476190476195e-05,
+      "loss": 1.3674,
+      "step": 580
+    },
+    {
+      "epoch": 2.2964426877470356,
+      "grad_norm": 2.2279505729675293,
+      "learning_rate": 3.867063492063492e-05,
+      "loss": 1.4776,
+      "step": 581
+    },
+    {
+      "epoch": 2.300395256916996,
+      "grad_norm": 2.4603147506713867,
+      "learning_rate": 3.8650793650793655e-05,
+      "loss": 1.6039,
+      "step": 582
+    },
+    {
+      "epoch": 2.3043478260869565,
+      "grad_norm": 2.8002820014953613,
+      "learning_rate": 3.863095238095238e-05,
+      "loss": 1.2765,
+      "step": 583
+    },
+    {
+      "epoch": 2.308300395256917,
+      "grad_norm": 2.505256414413452,
+      "learning_rate": 3.8611111111111116e-05,
+      "loss": 1.1921,
+      "step": 584
+    },
+    {
+      "epoch": 2.3122529644268774,
+      "grad_norm": 2.458230972290039,
+      "learning_rate": 3.859126984126984e-05,
+      "loss": 1.7022,
+      "step": 585
+    },
+    {
+      "epoch": 2.316205533596838,
+      "grad_norm": 2.3297207355499268,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 1.5739,
+      "step": 586
+    },
+    {
+      "epoch": 2.3201581027667983,
+      "grad_norm": 2.038970708847046,
+      "learning_rate": 3.8551587301587304e-05,
+      "loss": 1.2446,
+      "step": 587
+    },
+    {
+      "epoch": 2.324110671936759,
+      "grad_norm": 2.4906201362609863,
+      "learning_rate": 3.853174603174604e-05,
+      "loss": 1.3053,
+      "step": 588
+    },
+    {
+      "epoch": 2.3280632411067192,
+      "grad_norm": 2.42885160446167,
+      "learning_rate": 3.8511904761904765e-05,
+      "loss": 1.1066,
+      "step": 589
+    },
+    {
+      "epoch": 2.3320158102766797,
+      "grad_norm": 2.2669129371643066,
+      "learning_rate": 3.84920634920635e-05,
+      "loss": 1.2192,
+      "step": 590
+    },
+    {
+      "epoch": 2.33596837944664,
+      "grad_norm": 4.123607635498047,
+      "learning_rate": 3.8472222222222225e-05,
+      "loss": 1.3616,
+      "step": 591
+    },
+    {
+      "epoch": 2.3399209486166006,
+      "grad_norm": 2.522677183151245,
+      "learning_rate": 3.845238095238096e-05,
+      "loss": 1.3161,
+      "step": 592
+    },
+    {
+      "epoch": 2.3438735177865615,
+      "grad_norm": 2.6520330905914307,
+      "learning_rate": 3.8432539682539686e-05,
+      "loss": 1.1572,
+      "step": 593
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 2.5755152702331543,
+      "learning_rate": 3.841269841269842e-05,
+      "loss": 1.5446,
+      "step": 594
+    },
+    {
+      "epoch": 2.3517786561264824,
+      "grad_norm": 2.2885093688964844,
+      "learning_rate": 3.839285714285715e-05,
+      "loss": 1.4885,
+      "step": 595
+    },
+    {
+      "epoch": 2.355731225296443,
+      "grad_norm": 2.176738977432251,
+      "learning_rate": 3.837301587301588e-05,
+      "loss": 1.4043,
+      "step": 596
+    },
+    {
+      "epoch": 2.3596837944664033,
+      "grad_norm": 2.7747554779052734,
+      "learning_rate": 3.835317460317461e-05,
+      "loss": 1.6517,
+      "step": 597
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 1.9070560932159424,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 1.0572,
+      "step": 598
+    },
+    {
+      "epoch": 2.367588932806324,
+      "grad_norm": 2.161180019378662,
+      "learning_rate": 3.831349206349207e-05,
+      "loss": 1.3089,
+      "step": 599
+    },
+    {
+      "epoch": 2.3715415019762847,
+      "grad_norm": 2.5658626556396484,
+      "learning_rate": 3.8293650793650795e-05,
+      "loss": 1.1435,
+      "step": 600
+    },
+    {
+      "epoch": 2.375494071146245,
+      "grad_norm": 2.733940839767456,
+      "learning_rate": 3.827380952380952e-05,
+      "loss": 1.4425,
+      "step": 601
+    },
+    {
+      "epoch": 2.3794466403162056,
+      "grad_norm": 2.494096279144287,
+      "learning_rate": 3.8253968253968256e-05,
+      "loss": 1.5981,
+      "step": 602
+    },
+    {
+      "epoch": 2.383399209486166,
+      "grad_norm": 2.405909299850464,
+      "learning_rate": 3.823412698412698e-05,
+      "loss": 1.4946,
+      "step": 603
+    },
+    {
+      "epoch": 2.3873517786561265,
+      "grad_norm": 2.067415952682495,
+      "learning_rate": 3.821428571428572e-05,
+      "loss": 1.3633,
+      "step": 604
+    },
+    {
+      "epoch": 2.391304347826087,
+      "grad_norm": 2.1548092365264893,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 1.2963,
+      "step": 605
+    },
+    {
+      "epoch": 2.3952569169960474,
+      "grad_norm": 2.004725694656372,
+      "learning_rate": 3.817460317460317e-05,
+      "loss": 1.4675,
+      "step": 606
+    },
+    {
+      "epoch": 2.399209486166008,
+      "grad_norm": 2.1438629627227783,
+      "learning_rate": 3.8154761904761904e-05,
+      "loss": 1.5303,
+      "step": 607
+    },
+    {
+      "epoch": 2.4031620553359683,
+      "grad_norm": 2.6474621295928955,
+      "learning_rate": 3.813492063492063e-05,
+      "loss": 1.5247,
+      "step": 608
+    },
+    {
+      "epoch": 2.4071146245059287,
+      "grad_norm": 2.7593584060668945,
+      "learning_rate": 3.8115079365079365e-05,
+      "loss": 1.2225,
+      "step": 609
+    },
+    {
+      "epoch": 2.411067193675889,
+      "grad_norm": 2.53412127494812,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 1.3592,
+      "step": 610
+    },
+    {
+      "epoch": 2.4150197628458496,
+      "grad_norm": 2.477109909057617,
+      "learning_rate": 3.8075396825396826e-05,
+      "loss": 1.3446,
+      "step": 611
+    },
+    {
+      "epoch": 2.4189723320158105,
+      "grad_norm": 2.8138701915740967,
+      "learning_rate": 3.805555555555555e-05,
+      "loss": 1.3014,
+      "step": 612
+    },
+    {
+      "epoch": 2.4229249011857705,
+      "grad_norm": 2.805239200592041,
+      "learning_rate": 3.803571428571429e-05,
+      "loss": 1.3786,
+      "step": 613
+    },
+    {
+      "epoch": 2.4268774703557314,
+      "grad_norm": 2.7038285732269287,
+      "learning_rate": 3.8015873015873014e-05,
+      "loss": 1.1213,
+      "step": 614
+    },
+    {
+      "epoch": 2.430830039525692,
+      "grad_norm": 2.978872299194336,
+      "learning_rate": 3.799603174603175e-05,
+      "loss": 1.2804,
+      "step": 615
+    },
+    {
+      "epoch": 2.4347826086956523,
+      "grad_norm": 2.0930662155151367,
+      "learning_rate": 3.7976190476190474e-05,
+      "loss": 1.6642,
+      "step": 616
+    },
+    {
+      "epoch": 2.438735177865613,
+      "grad_norm": 2.3781399726867676,
+      "learning_rate": 3.795634920634921e-05,
+      "loss": 1.2398,
+      "step": 617
+    },
+    {
+      "epoch": 2.4426877470355732,
+      "grad_norm": 2.1842458248138428,
+      "learning_rate": 3.7936507936507935e-05,
+      "loss": 1.5295,
+      "step": 618
+    },
+    {
+      "epoch": 2.4466403162055337,
+      "grad_norm": 3.6338818073272705,
+      "learning_rate": 3.791666666666667e-05,
+      "loss": 1.0698,
+      "step": 619
+    },
+    {
+      "epoch": 2.450592885375494,
+      "grad_norm": 2.5560874938964844,
+      "learning_rate": 3.7896825396825396e-05,
+      "loss": 1.3759,
+      "step": 620
+    },
+    {
+      "epoch": 2.4545454545454546,
+      "grad_norm": 2.4692065715789795,
+      "learning_rate": 3.787698412698413e-05,
+      "loss": 1.2892,
+      "step": 621
+    },
+    {
+      "epoch": 2.458498023715415,
+      "grad_norm": 2.5481793880462646,
+      "learning_rate": 3.785714285714286e-05,
+      "loss": 1.5718,
+      "step": 622
+    },
+    {
+      "epoch": 2.4624505928853755,
+      "grad_norm": 2.608428716659546,
+      "learning_rate": 3.783730158730159e-05,
+      "loss": 1.3108,
+      "step": 623
+    },
+    {
+      "epoch": 2.466403162055336,
+      "grad_norm": 2.8005776405334473,
+      "learning_rate": 3.781746031746032e-05,
+      "loss": 1.5068,
+      "step": 624
+    },
+    {
+      "epoch": 2.4703557312252964,
+      "grad_norm": 2.2495791912078857,
+      "learning_rate": 3.779761904761905e-05,
+      "loss": 1.3082,
+      "step": 625
+    },
+    {
+      "epoch": 2.474308300395257,
+      "grad_norm": 2.567267656326294,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 1.4442,
+      "step": 626
+    },
+    {
+      "epoch": 2.4782608695652173,
+      "grad_norm": 2.4826018810272217,
+      "learning_rate": 3.775793650793651e-05,
+      "loss": 1.2581,
+      "step": 627
+    },
+    {
+      "epoch": 2.4822134387351777,
+      "grad_norm": 2.3756072521209717,
+      "learning_rate": 3.773809523809524e-05,
+      "loss": 1.2695,
+      "step": 628
+    },
+    {
+      "epoch": 2.486166007905138,
+      "grad_norm": 2.6527416706085205,
+      "learning_rate": 3.7718253968253966e-05,
+      "loss": 1.2956,
+      "step": 629
+    },
+    {
+      "epoch": 2.4901185770750986,
+      "grad_norm": 2.467822790145874,
+      "learning_rate": 3.76984126984127e-05,
+      "loss": 1.4364,
+      "step": 630
+    },
+    {
+      "epoch": 2.494071146245059,
+      "grad_norm": 2.6203055381774902,
+      "learning_rate": 3.767857142857143e-05,
+      "loss": 1.4897,
+      "step": 631
+    },
+    {
+      "epoch": 2.4980237154150196,
+      "grad_norm": 2.233952045440674,
+      "learning_rate": 3.765873015873016e-05,
+      "loss": 1.3318,
+      "step": 632
+    },
+    {
+      "epoch": 2.5019762845849804,
+      "grad_norm": 1.96635901927948,
+      "learning_rate": 3.763888888888889e-05,
+      "loss": 1.2446,
+      "step": 633
+    },
+    {
+      "epoch": 2.5059288537549405,
+      "grad_norm": 2.349234104156494,
+      "learning_rate": 3.761904761904762e-05,
+      "loss": 1.3324,
+      "step": 634
+    },
+    {
+      "epoch": 2.5098814229249014,
+      "grad_norm": 2.961324691772461,
+      "learning_rate": 3.759920634920635e-05,
+      "loss": 1.1765,
+      "step": 635
+    },
+    {
+      "epoch": 2.513833992094862,
+      "grad_norm": 2.8292768001556396,
+      "learning_rate": 3.757936507936508e-05,
+      "loss": 1.0077,
+      "step": 636
+    },
+    {
+      "epoch": 2.5177865612648223,
+      "grad_norm": 2.9311156272888184,
+      "learning_rate": 3.755952380952381e-05,
+      "loss": 1.3392,
+      "step": 637
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 2.6130383014678955,
+      "learning_rate": 3.753968253968254e-05,
+      "loss": 1.2328,
+      "step": 638
+    },
+    {
+      "epoch": 2.525691699604743,
+      "grad_norm": 2.0875937938690186,
+      "learning_rate": 3.751984126984127e-05,
+      "loss": 1.3491,
+      "step": 639
+    },
+    {
+      "epoch": 2.5296442687747036,
+      "grad_norm": 2.593751907348633,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.1862,
+      "step": 640
+    },
+    {
+      "epoch": 2.533596837944664,
+      "grad_norm": 2.0226974487304688,
+      "learning_rate": 3.748015873015873e-05,
+      "loss": 1.3762,
+      "step": 641
+    },
+    {
+      "epoch": 2.5375494071146245,
+      "grad_norm": 2.4336142539978027,
+      "learning_rate": 3.7460317460317464e-05,
+      "loss": 1.3271,
+      "step": 642
+    },
+    {
+      "epoch": 2.541501976284585,
+      "grad_norm": 1.8397362232208252,
+      "learning_rate": 3.744047619047619e-05,
+      "loss": 1.2235,
+      "step": 643
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 2.0357463359832764,
+      "learning_rate": 3.7420634920634925e-05,
+      "loss": 1.3871,
+      "step": 644
+    },
+    {
+      "epoch": 2.549407114624506,
+      "grad_norm": 3.3774683475494385,
+      "learning_rate": 3.740079365079365e-05,
+      "loss": 1.3579,
+      "step": 645
+    },
+    {
+      "epoch": 2.5533596837944663,
+      "grad_norm": 3.041534662246704,
+      "learning_rate": 3.7380952380952386e-05,
+      "loss": 1.1318,
+      "step": 646
+    },
+    {
+      "epoch": 2.5573122529644268,
+      "grad_norm": 2.484518527984619,
+      "learning_rate": 3.736111111111111e-05,
+      "loss": 1.3531,
+      "step": 647
+    },
+    {
+      "epoch": 2.561264822134387,
+      "grad_norm": 2.612365245819092,
+      "learning_rate": 3.7341269841269846e-05,
+      "loss": 1.3075,
+      "step": 648
+    },
+    {
+      "epoch": 2.5652173913043477,
+      "grad_norm": 2.3737142086029053,
+      "learning_rate": 3.7321428571428573e-05,
+      "loss": 1.272,
+      "step": 649
+    },
+    {
+      "epoch": 2.5691699604743086,
+      "grad_norm": 2.600964069366455,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 1.0839,
+      "step": 650
+    },
+    {
+      "epoch": 2.5731225296442686,
+      "grad_norm": 2.697070598602295,
+      "learning_rate": 3.7281746031746034e-05,
+      "loss": 1.1383,
+      "step": 651
+    },
+    {
+      "epoch": 2.5770750988142295,
+      "grad_norm": 2.752293109893799,
+      "learning_rate": 3.726190476190476e-05,
+      "loss": 1.2823,
+      "step": 652
+    },
+    {
+      "epoch": 2.5810276679841895,
+      "grad_norm": 2.47784161567688,
+      "learning_rate": 3.7242063492063495e-05,
+      "loss": 1.5547,
+      "step": 653
+    },
+    {
+      "epoch": 2.5849802371541504,
+      "grad_norm": 2.5010085105895996,
+      "learning_rate": 3.722222222222222e-05,
+      "loss": 1.2051,
+      "step": 654
+    },
+    {
+      "epoch": 2.588932806324111,
+      "grad_norm": 2.3102123737335205,
+      "learning_rate": 3.7202380952380956e-05,
+      "loss": 1.1448,
+      "step": 655
+    },
+    {
+      "epoch": 2.5928853754940713,
+      "grad_norm": 2.760524034500122,
+      "learning_rate": 3.718253968253968e-05,
+      "loss": 1.3238,
+      "step": 656
+    },
+    {
+      "epoch": 2.5968379446640317,
+      "grad_norm": 3.336056709289551,
+      "learning_rate": 3.7162698412698416e-05,
+      "loss": 1.2137,
+      "step": 657
+    },
+    {
+      "epoch": 2.600790513833992,
+      "grad_norm": 3.0085883140563965,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 1.0954,
+      "step": 658
+    },
+    {
+      "epoch": 2.6047430830039526,
+      "grad_norm": 2.819323778152466,
+      "learning_rate": 3.712301587301588e-05,
+      "loss": 1.4706,
+      "step": 659
+    },
+    {
+      "epoch": 2.608695652173913,
+      "grad_norm": 2.527916193008423,
+      "learning_rate": 3.7103174603174604e-05,
+      "loss": 1.2759,
+      "step": 660
+    },
+    {
+      "epoch": 2.6126482213438735,
+      "grad_norm": 3.203950881958008,
+      "learning_rate": 3.708333333333334e-05,
+      "loss": 1.907,
+      "step": 661
+    },
+    {
+      "epoch": 2.616600790513834,
+      "grad_norm": 2.3258347511291504,
+      "learning_rate": 3.7063492063492065e-05,
+      "loss": 1.0956,
+      "step": 662
+    },
+    {
+      "epoch": 2.6205533596837944,
+      "grad_norm": 2.3885722160339355,
+      "learning_rate": 3.70436507936508e-05,
+      "loss": 1.3588,
+      "step": 663
+    },
+    {
+      "epoch": 2.624505928853755,
+      "grad_norm": 2.46718168258667,
+      "learning_rate": 3.7023809523809526e-05,
+      "loss": 1.5113,
+      "step": 664
+    },
+    {
+      "epoch": 2.6284584980237153,
+      "grad_norm": 2.780848503112793,
+      "learning_rate": 3.700396825396826e-05,
+      "loss": 1.3804,
+      "step": 665
+    },
+    {
+      "epoch": 2.632411067193676,
+      "grad_norm": 2.919987440109253,
+      "learning_rate": 3.6984126984126986e-05,
+      "loss": 1.4946,
+      "step": 666
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 2.093374252319336,
+      "learning_rate": 3.696428571428572e-05,
+      "loss": 1.3501,
+      "step": 667
+    },
+    {
+      "epoch": 2.6403162055335967,
+      "grad_norm": 3.430272102355957,
+      "learning_rate": 3.694444444444445e-05,
+      "loss": 0.956,
+      "step": 668
+    },
+    {
+      "epoch": 2.6442687747035576,
+      "grad_norm": 2.6946187019348145,
+      "learning_rate": 3.692460317460318e-05,
+      "loss": 1.3491,
+      "step": 669
+    },
+    {
+      "epoch": 2.6482213438735176,
+      "grad_norm": 2.1861889362335205,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 1.3425,
+      "step": 670
+    },
+    {
+      "epoch": 2.6521739130434785,
+      "grad_norm": 2.343151569366455,
+      "learning_rate": 3.688492063492064e-05,
+      "loss": 1.3148,
+      "step": 671
+    },
+    {
+      "epoch": 2.6561264822134385,
+      "grad_norm": 2.8744852542877197,
+      "learning_rate": 3.686507936507937e-05,
+      "loss": 1.1524,
+      "step": 672
+    },
+    {
+      "epoch": 2.6600790513833994,
+      "grad_norm": 2.830291271209717,
+      "learning_rate": 3.68452380952381e-05,
+      "loss": 1.5512,
+      "step": 673
+    },
+    {
+      "epoch": 2.66403162055336,
+      "grad_norm": 2.655128002166748,
+      "learning_rate": 3.682539682539683e-05,
+      "loss": 1.2045,
+      "step": 674
+    },
+    {
+      "epoch": 2.6679841897233203,
+      "grad_norm": 2.4060311317443848,
+      "learning_rate": 3.6805555555555556e-05,
+      "loss": 1.0609,
+      "step": 675
+    },
+    {
+      "epoch": 2.6719367588932808,
+      "grad_norm": 2.0129261016845703,
+      "learning_rate": 3.678571428571429e-05,
+      "loss": 1.2873,
+      "step": 676
+    },
+    {
+      "epoch": 2.675889328063241,
+      "grad_norm": 2.645951271057129,
+      "learning_rate": 3.676587301587302e-05,
+      "loss": 1.1951,
+      "step": 677
+    },
+    {
+      "epoch": 2.6798418972332017,
+      "grad_norm": 2.4416396617889404,
+      "learning_rate": 3.674603174603175e-05,
+      "loss": 1.3103,
+      "step": 678
+    },
+    {
+      "epoch": 2.683794466403162,
+      "grad_norm": 3.599039077758789,
+      "learning_rate": 3.672619047619048e-05,
+      "loss": 1.4431,
+      "step": 679
+    },
+    {
+      "epoch": 2.6877470355731226,
+      "grad_norm": 2.941194534301758,
+      "learning_rate": 3.6706349206349205e-05,
+      "loss": 1.0918,
+      "step": 680
+    },
+    {
+      "epoch": 2.691699604743083,
+      "grad_norm": 2.7332816123962402,
+      "learning_rate": 3.668650793650794e-05,
+      "loss": 1.3429,
+      "step": 681
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 3.3440287113189697,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.4627,
+      "step": 682
+    },
+    {
+      "epoch": 2.699604743083004,
+      "grad_norm": 2.5131101608276367,
+      "learning_rate": 3.664682539682539e-05,
+      "loss": 1.5036,
+      "step": 683
+    },
+    {
+      "epoch": 2.7035573122529644,
+      "grad_norm": 2.991786241531372,
+      "learning_rate": 3.6626984126984126e-05,
+      "loss": 1.8342,
+      "step": 684
+    },
+    {
+      "epoch": 2.707509881422925,
+      "grad_norm": 2.5615289211273193,
+      "learning_rate": 3.6607142857142853e-05,
+      "loss": 1.3832,
+      "step": 685
+    },
+    {
+      "epoch": 2.7114624505928853,
+      "grad_norm": 2.467275857925415,
+      "learning_rate": 3.658730158730159e-05,
+      "loss": 1.2683,
+      "step": 686
+    },
+    {
+      "epoch": 2.7154150197628457,
+      "grad_norm": 2.156137228012085,
+      "learning_rate": 3.6567460317460314e-05,
+      "loss": 1.2575,
+      "step": 687
+    },
+    {
+      "epoch": 2.719367588932806,
+      "grad_norm": 1.868895411491394,
+      "learning_rate": 3.654761904761905e-05,
+      "loss": 1.4412,
+      "step": 688
+    },
+    {
+      "epoch": 2.7233201581027666,
+      "grad_norm": 2.4768173694610596,
+      "learning_rate": 3.6527777777777775e-05,
+      "loss": 1.2711,
+      "step": 689
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 3.286557912826538,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 1.4438,
+      "step": 690
+    },
+    {
+      "epoch": 2.7312252964426875,
+      "grad_norm": 2.219813823699951,
+      "learning_rate": 3.6488095238095236e-05,
+      "loss": 1.309,
+      "step": 691
+    },
+    {
+      "epoch": 2.7351778656126484,
+      "grad_norm": 2.425921678543091,
+      "learning_rate": 3.646825396825397e-05,
+      "loss": 1.4898,
+      "step": 692
+    },
+    {
+      "epoch": 2.7391304347826084,
+      "grad_norm": 2.3334920406341553,
+      "learning_rate": 3.6448412698412696e-05,
+      "loss": 1.3917,
+      "step": 693
+    },
+    {
+      "epoch": 2.7430830039525693,
+      "grad_norm": 2.4505045413970947,
+      "learning_rate": 3.642857142857143e-05,
+      "loss": 1.2184,
+      "step": 694
+    },
+    {
+      "epoch": 2.7470355731225298,
+      "grad_norm": 2.223360538482666,
+      "learning_rate": 3.640873015873016e-05,
+      "loss": 1.4157,
+      "step": 695
+    },
+    {
+      "epoch": 2.7509881422924902,
+      "grad_norm": 2.872119665145874,
+      "learning_rate": 3.638888888888889e-05,
+      "loss": 1.1279,
+      "step": 696
+    },
+    {
+      "epoch": 2.7549407114624507,
+      "grad_norm": 2.502791404724121,
+      "learning_rate": 3.636904761904762e-05,
+      "loss": 1.1267,
+      "step": 697
+    },
+    {
+      "epoch": 2.758893280632411,
+      "grad_norm": 2.880258083343506,
+      "learning_rate": 3.634920634920635e-05,
+      "loss": 1.2936,
+      "step": 698
+    },
+    {
+      "epoch": 2.7628458498023716,
+      "grad_norm": 2.5304715633392334,
+      "learning_rate": 3.632936507936508e-05,
+      "loss": 1.0587,
+      "step": 699
+    },
+    {
+      "epoch": 2.766798418972332,
+      "grad_norm": 2.664189338684082,
+      "learning_rate": 3.630952380952381e-05,
+      "loss": 1.4785,
+      "step": 700
+    },
+    {
+      "epoch": 2.7707509881422925,
+      "grad_norm": 2.274841070175171,
+      "learning_rate": 3.628968253968254e-05,
+      "loss": 1.3508,
+      "step": 701
+    },
+    {
+      "epoch": 2.774703557312253,
+      "grad_norm": 2.5693907737731934,
+      "learning_rate": 3.626984126984127e-05,
+      "loss": 1.3676,
+      "step": 702
+    },
+    {
+      "epoch": 2.7786561264822134,
+      "grad_norm": 2.505685806274414,
+      "learning_rate": 3.625e-05,
+      "loss": 1.3922,
+      "step": 703
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 2.220625400543213,
+      "learning_rate": 3.6230158730158734e-05,
+      "loss": 1.2811,
+      "step": 704
+    },
+    {
+      "epoch": 2.7865612648221343,
+      "grad_norm": 2.694793939590454,
+      "learning_rate": 3.621031746031746e-05,
+      "loss": 1.4419,
+      "step": 705
+    },
+    {
+      "epoch": 2.7905138339920947,
+      "grad_norm": 2.996187925338745,
+      "learning_rate": 3.619047619047619e-05,
+      "loss": 1.4447,
+      "step": 706
+    },
+    {
+      "epoch": 2.794466403162055,
+      "grad_norm": 2.177954912185669,
+      "learning_rate": 3.617063492063492e-05,
+      "loss": 1.3629,
+      "step": 707
+    },
+    {
+      "epoch": 2.7984189723320156,
+      "grad_norm": 2.743326187133789,
+      "learning_rate": 3.615079365079365e-05,
+      "loss": 1.4397,
+      "step": 708
+    },
+    {
+      "epoch": 2.8023715415019765,
+      "grad_norm": 2.6118264198303223,
+      "learning_rate": 3.613095238095238e-05,
+      "loss": 1.1212,
+      "step": 709
+    },
+    {
+      "epoch": 2.8063241106719365,
+      "grad_norm": 2.1388731002807617,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 1.5192,
+      "step": 710
+    },
+    {
+      "epoch": 2.8102766798418974,
+      "grad_norm": 2.721480131149292,
+      "learning_rate": 3.609126984126984e-05,
+      "loss": 1.579,
+      "step": 711
+    },
+    {
+      "epoch": 2.8142292490118574,
+      "grad_norm": 2.726956844329834,
+      "learning_rate": 3.607142857142857e-05,
+      "loss": 1.6523,
+      "step": 712
+    },
+    {
+      "epoch": 2.8181818181818183,
+      "grad_norm": 2.369964599609375,
+      "learning_rate": 3.6051587301587304e-05,
+      "loss": 1.3389,
+      "step": 713
+    },
+    {
+      "epoch": 2.822134387351779,
+      "grad_norm": 2.3911476135253906,
+      "learning_rate": 3.603174603174603e-05,
+      "loss": 1.4589,
+      "step": 714
+    },
+    {
+      "epoch": 2.8260869565217392,
+      "grad_norm": 2.717634916305542,
+      "learning_rate": 3.6011904761904765e-05,
+      "loss": 1.4241,
+      "step": 715
+    },
+    {
+      "epoch": 2.8300395256916997,
+      "grad_norm": 2.3168764114379883,
+      "learning_rate": 3.599206349206349e-05,
+      "loss": 1.3172,
+      "step": 716
+    },
+    {
+      "epoch": 2.83399209486166,
+      "grad_norm": 2.34859037399292,
+      "learning_rate": 3.5972222222222225e-05,
+      "loss": 1.2871,
+      "step": 717
+    },
+    {
+      "epoch": 2.8379446640316206,
+      "grad_norm": 2.4705185890197754,
+      "learning_rate": 3.595238095238095e-05,
+      "loss": 1.524,
+      "step": 718
+    },
+    {
+      "epoch": 2.841897233201581,
+      "grad_norm": 2.9651718139648438,
+      "learning_rate": 3.5932539682539686e-05,
+      "loss": 1.4915,
+      "step": 719
+    },
+    {
+      "epoch": 2.8458498023715415,
+      "grad_norm": 2.442030668258667,
+      "learning_rate": 3.591269841269841e-05,
+      "loss": 1.5001,
+      "step": 720
+    },
+    {
+      "epoch": 2.849802371541502,
+      "grad_norm": 4.075084686279297,
+      "learning_rate": 3.589285714285715e-05,
+      "loss": 0.9039,
+      "step": 721
+    },
+    {
+      "epoch": 2.8537549407114624,
+      "grad_norm": 3.0871315002441406,
+      "learning_rate": 3.5873015873015874e-05,
+      "loss": 1.6743,
+      "step": 722
+    },
+    {
+      "epoch": 2.857707509881423,
+      "grad_norm": 2.6841776371002197,
+      "learning_rate": 3.585317460317461e-05,
+      "loss": 1.1567,
+      "step": 723
+    },
+    {
+      "epoch": 2.8616600790513833,
+      "grad_norm": 2.2139689922332764,
+      "learning_rate": 3.5833333333333335e-05,
+      "loss": 1.4276,
+      "step": 724
+    },
+    {
+      "epoch": 2.8656126482213438,
+      "grad_norm": 2.8722684383392334,
+      "learning_rate": 3.581349206349207e-05,
+      "loss": 1.2222,
+      "step": 725
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 3.5767440795898438,
+      "learning_rate": 3.5793650793650795e-05,
+      "loss": 1.421,
+      "step": 726
+    },
+    {
+      "epoch": 2.8735177865612647,
+      "grad_norm": 3.3746676445007324,
+      "learning_rate": 3.577380952380953e-05,
+      "loss": 1.9383,
+      "step": 727
+    },
+    {
+      "epoch": 2.8774703557312256,
+      "grad_norm": 3.810504913330078,
+      "learning_rate": 3.5753968253968256e-05,
+      "loss": 1.0087,
+      "step": 728
+    },
+    {
+      "epoch": 2.8814229249011856,
+      "grad_norm": 2.2538998126983643,
+      "learning_rate": 3.573412698412698e-05,
+      "loss": 1.4981,
+      "step": 729
+    },
+    {
+      "epoch": 2.8853754940711465,
+      "grad_norm": 2.282931327819824,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 1.2655,
+      "step": 730
+    },
+    {
+      "epoch": 2.8893280632411065,
+      "grad_norm": 2.3477354049682617,
+      "learning_rate": 3.5694444444444444e-05,
+      "loss": 1.4641,
+      "step": 731
+    },
+    {
+      "epoch": 2.8932806324110674,
+      "grad_norm": 2.141794204711914,
+      "learning_rate": 3.567460317460318e-05,
+      "loss": 1.5262,
+      "step": 732
+    },
+    {
+      "epoch": 2.897233201581028,
+      "grad_norm": 2.0820016860961914,
+      "learning_rate": 3.5654761904761905e-05,
+      "loss": 1.3743,
+      "step": 733
+    },
+    {
+      "epoch": 2.9011857707509883,
+      "grad_norm": 2.0662853717803955,
+      "learning_rate": 3.563492063492064e-05,
+      "loss": 1.3662,
+      "step": 734
+    },
+    {
+      "epoch": 2.9051383399209487,
+      "grad_norm": 2.205324411392212,
+      "learning_rate": 3.5615079365079365e-05,
+      "loss": 1.4079,
+      "step": 735
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 2.201070785522461,
+      "learning_rate": 3.55952380952381e-05,
+      "loss": 1.1008,
+      "step": 736
+    },
+    {
+      "epoch": 2.9130434782608696,
+      "grad_norm": 2.159719944000244,
+      "learning_rate": 3.5575396825396826e-05,
+      "loss": 1.2646,
+      "step": 737
+    },
+    {
+      "epoch": 2.91699604743083,
+      "grad_norm": 2.3559587001800537,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 1.4704,
+      "step": 738
+    },
+    {
+      "epoch": 2.9209486166007905,
+      "grad_norm": 3.3258392810821533,
+      "learning_rate": 3.553571428571429e-05,
+      "loss": 1.1226,
+      "step": 739
+    },
+    {
+      "epoch": 2.924901185770751,
+      "grad_norm": 2.6636176109313965,
+      "learning_rate": 3.551587301587302e-05,
+      "loss": 1.3304,
+      "step": 740
+    },
+    {
+      "epoch": 2.9288537549407114,
+      "grad_norm": 2.8491475582122803,
+      "learning_rate": 3.549603174603175e-05,
+      "loss": 1.4011,
+      "step": 741
+    },
+    {
+      "epoch": 2.932806324110672,
+      "grad_norm": 3.1546342372894287,
+      "learning_rate": 3.547619047619048e-05,
+      "loss": 1.0591,
+      "step": 742
+    },
+    {
+      "epoch": 2.9367588932806323,
+      "grad_norm": 3.2184884548187256,
+      "learning_rate": 3.545634920634921e-05,
+      "loss": 1.5376,
+      "step": 743
+    },
+    {
+      "epoch": 2.940711462450593,
+      "grad_norm": 2.49813175201416,
+      "learning_rate": 3.543650793650794e-05,
+      "loss": 1.2826,
+      "step": 744
+    },
+    {
+      "epoch": 2.9446640316205532,
+      "grad_norm": 2.511040210723877,
+      "learning_rate": 3.541666666666667e-05,
+      "loss": 1.4989,
+      "step": 745
+    },
+    {
+      "epoch": 2.9486166007905137,
+      "grad_norm": 2.233522415161133,
+      "learning_rate": 3.53968253968254e-05,
+      "loss": 1.1618,
+      "step": 746
+    },
+    {
+      "epoch": 2.9525691699604746,
+      "grad_norm": 2.288743495941162,
+      "learning_rate": 3.537698412698413e-05,
+      "loss": 0.9788,
+      "step": 747
+    },
+    {
+      "epoch": 2.9565217391304346,
+      "grad_norm": 3.0124926567077637,
+      "learning_rate": 3.5357142857142864e-05,
+      "loss": 0.9291,
+      "step": 748
+    },
+    {
+      "epoch": 2.9604743083003955,
+      "grad_norm": 3.1643810272216797,
+      "learning_rate": 3.533730158730159e-05,
+      "loss": 0.789,
+      "step": 749
+    },
+    {
+      "epoch": 2.9644268774703555,
+      "grad_norm": 2.8564984798431396,
+      "learning_rate": 3.5317460317460324e-05,
+      "loss": 1.4296,
+      "step": 750
+    },
+    {
+      "epoch": 2.9683794466403164,
+      "grad_norm": 2.2462079524993896,
+      "learning_rate": 3.529761904761905e-05,
+      "loss": 1.275,
+      "step": 751
+    },
+    {
+      "epoch": 2.972332015810277,
+      "grad_norm": 2.5663106441497803,
+      "learning_rate": 3.527777777777778e-05,
+      "loss": 1.7537,
+      "step": 752
+    },
+    {
+      "epoch": 2.9762845849802373,
+      "grad_norm": 2.8393218517303467,
+      "learning_rate": 3.525793650793651e-05,
+      "loss": 1.6171,
+      "step": 753
+    },
+    {
+      "epoch": 2.9802371541501977,
+      "grad_norm": 2.4854910373687744,
+      "learning_rate": 3.523809523809524e-05,
+      "loss": 1.4017,
+      "step": 754
+    },
+    {
+      "epoch": 2.984189723320158,
+      "grad_norm": 2.6834123134613037,
+      "learning_rate": 3.521825396825397e-05,
+      "loss": 1.4678,
+      "step": 755
+    },
+    {
+      "epoch": 2.9881422924901186,
+      "grad_norm": 2.4412403106689453,
+      "learning_rate": 3.51984126984127e-05,
+      "loss": 1.2793,
+      "step": 756
+    },
+    {
+      "epoch": 2.992094861660079,
+      "grad_norm": 2.7471697330474854,
+      "learning_rate": 3.5178571428571434e-05,
+      "loss": 1.4015,
+      "step": 757
+    },
+    {
+      "epoch": 2.9960474308300395,
+      "grad_norm": 2.6294939517974854,
+      "learning_rate": 3.515873015873016e-05,
+      "loss": 1.4328,
+      "step": 758
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.844717502593994,
+      "learning_rate": 3.513888888888889e-05,
+      "loss": 1.2477,
+      "step": 759
+    },
+    {
+      "epoch": 3.0039525691699605,
+      "grad_norm": 2.496687173843384,
+      "learning_rate": 3.511904761904762e-05,
+      "loss": 0.9741,
+      "step": 760
+    },
+    {
+      "epoch": 3.007905138339921,
+      "grad_norm": 2.5280258655548096,
+      "learning_rate": 3.509920634920635e-05,
+      "loss": 0.8558,
+      "step": 761
+    },
+    {
+      "epoch": 3.0118577075098814,
+      "grad_norm": 2.257899522781372,
+      "learning_rate": 3.5079365079365075e-05,
+      "loss": 1.1979,
+      "step": 762
+    },
+    {
+      "epoch": 3.015810276679842,
+      "grad_norm": 2.1623339653015137,
+      "learning_rate": 3.505952380952381e-05,
+      "loss": 0.9799,
+      "step": 763
+    },
+    {
+      "epoch": 3.0197628458498023,
+      "grad_norm": 2.9412381649017334,
+      "learning_rate": 3.5039682539682536e-05,
+      "loss": 0.8915,
+      "step": 764
+    },
+    {
+      "epoch": 3.0237154150197627,
+      "grad_norm": 3.138514995574951,
+      "learning_rate": 3.501984126984127e-05,
+      "loss": 0.8636,
+      "step": 765
+    },
+    {
+      "epoch": 3.027667984189723,
+      "grad_norm": 4.3793416023254395,
+      "learning_rate": 3.5e-05,
+      "loss": 0.7668,
+      "step": 766
+    },
+    {
+      "epoch": 3.0316205533596836,
+      "grad_norm": 4.136168003082275,
+      "learning_rate": 3.498015873015873e-05,
+      "loss": 0.9277,
+      "step": 767
+    },
+    {
+      "epoch": 3.035573122529644,
+      "grad_norm": 4.092403888702393,
+      "learning_rate": 3.496031746031746e-05,
+      "loss": 1.1269,
+      "step": 768
+    },
+    {
+      "epoch": 3.039525691699605,
+      "grad_norm": 4.120564937591553,
+      "learning_rate": 3.494047619047619e-05,
+      "loss": 0.9644,
+      "step": 769
+    },
+    {
+      "epoch": 3.0434782608695654,
+      "grad_norm": 2.608088254928589,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.8488,
+      "step": 770
+    },
+    {
+      "epoch": 3.047430830039526,
+      "grad_norm": 3.257272481918335,
+      "learning_rate": 3.490079365079365e-05,
+      "loss": 0.8844,
+      "step": 771
+    },
+    {
+      "epoch": 3.0513833992094863,
+      "grad_norm": 3.9354476928710938,
+      "learning_rate": 3.488095238095238e-05,
+      "loss": 0.9306,
+      "step": 772
+    },
+    {
+      "epoch": 3.0553359683794468,
+      "grad_norm": 3.038783550262451,
+      "learning_rate": 3.486111111111111e-05,
+      "loss": 0.9639,
+      "step": 773
+    },
+    {
+      "epoch": 3.059288537549407,
+      "grad_norm": 2.9691848754882812,
+      "learning_rate": 3.484126984126984e-05,
+      "loss": 1.0708,
+      "step": 774
+    },
+    {
+      "epoch": 3.0632411067193677,
+      "grad_norm": 2.9549548625946045,
+      "learning_rate": 3.4821428571428574e-05,
+      "loss": 0.6093,
+      "step": 775
+    },
+    {
+      "epoch": 3.067193675889328,
+      "grad_norm": 2.429551839828491,
+      "learning_rate": 3.48015873015873e-05,
+      "loss": 0.9417,
+      "step": 776
+    },
+    {
+      "epoch": 3.0711462450592886,
+      "grad_norm": 3.1636385917663574,
+      "learning_rate": 3.4781746031746034e-05,
+      "loss": 0.7358,
+      "step": 777
+    },
+    {
+      "epoch": 3.075098814229249,
+      "grad_norm": 2.7946548461914062,
+      "learning_rate": 3.476190476190476e-05,
+      "loss": 1.0406,
+      "step": 778
+    },
+    {
+      "epoch": 3.0790513833992095,
+      "grad_norm": 2.363107442855835,
+      "learning_rate": 3.4742063492063495e-05,
+      "loss": 0.7761,
+      "step": 779
+    },
+    {
+      "epoch": 3.08300395256917,
+      "grad_norm": 2.957887649536133,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.947,
+      "step": 780
+    },
+    {
+      "epoch": 3.0869565217391304,
+      "grad_norm": 2.936105489730835,
+      "learning_rate": 3.4702380952380956e-05,
+      "loss": 0.9311,
+      "step": 781
+    },
+    {
+      "epoch": 3.090909090909091,
+      "grad_norm": 3.723759174346924,
+      "learning_rate": 3.468253968253968e-05,
+      "loss": 0.7467,
+      "step": 782
+    },
+    {
+      "epoch": 3.0948616600790513,
+      "grad_norm": 2.8762381076812744,
+      "learning_rate": 3.466269841269842e-05,
+      "loss": 1.078,
+      "step": 783
+    },
+    {
+      "epoch": 3.0988142292490117,
+      "grad_norm": 3.5455434322357178,
+      "learning_rate": 3.4642857142857144e-05,
+      "loss": 0.9024,
+      "step": 784
+    },
+    {
+      "epoch": 3.102766798418972,
+      "grad_norm": 3.4952311515808105,
+      "learning_rate": 3.462301587301587e-05,
+      "loss": 1.1114,
+      "step": 785
+    },
+    {
+      "epoch": 3.1067193675889326,
+      "grad_norm": 3.0181546211242676,
+      "learning_rate": 3.4603174603174604e-05,
+      "loss": 1.0401,
+      "step": 786
+    },
+    {
+      "epoch": 3.110671936758893,
+      "grad_norm": 2.7165963649749756,
+      "learning_rate": 3.458333333333333e-05,
+      "loss": 1.0207,
+      "step": 787
+    },
+    {
+      "epoch": 3.1146245059288535,
+      "grad_norm": 3.077030897140503,
+      "learning_rate": 3.4563492063492065e-05,
+      "loss": 0.8701,
+      "step": 788
+    },
+    {
+      "epoch": 3.1185770750988144,
+      "grad_norm": 2.9918665885925293,
+      "learning_rate": 3.454365079365079e-05,
+      "loss": 1.2665,
+      "step": 789
+    },
+    {
+      "epoch": 3.122529644268775,
+      "grad_norm": 3.622313976287842,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 1.1394,
+      "step": 790
+    },
+    {
+      "epoch": 3.1264822134387353,
+      "grad_norm": 2.705162286758423,
+      "learning_rate": 3.450396825396825e-05,
+      "loss": 0.8924,
+      "step": 791
+    },
+    {
+      "epoch": 3.130434782608696,
+      "grad_norm": 3.595820426940918,
+      "learning_rate": 3.448412698412699e-05,
+      "loss": 1.057,
+      "step": 792
+    },
+    {
+      "epoch": 3.1343873517786562,
+      "grad_norm": 3.4480679035186768,
+      "learning_rate": 3.4464285714285714e-05,
+      "loss": 0.8167,
+      "step": 793
+    },
+    {
+      "epoch": 3.1383399209486167,
+      "grad_norm": 2.8485183715820312,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.8327,
+      "step": 794
+    },
+    {
+      "epoch": 3.142292490118577,
+      "grad_norm": 3.1256802082061768,
+      "learning_rate": 3.4424603174603174e-05,
+      "loss": 1.0551,
+      "step": 795
+    },
+    {
+      "epoch": 3.1462450592885376,
+      "grad_norm": 3.3785347938537598,
+      "learning_rate": 3.440476190476191e-05,
+      "loss": 1.2833,
+      "step": 796
+    },
+    {
+      "epoch": 3.150197628458498,
+      "grad_norm": 2.9685938358306885,
+      "learning_rate": 3.4384920634920635e-05,
+      "loss": 1.0435,
+      "step": 797
+    },
+    {
+      "epoch": 3.1541501976284585,
+      "grad_norm": 2.9943830966949463,
+      "learning_rate": 3.436507936507937e-05,
+      "loss": 0.7879,
+      "step": 798
+    },
+    {
+      "epoch": 3.158102766798419,
+      "grad_norm": 2.655322790145874,
+      "learning_rate": 3.4345238095238096e-05,
+      "loss": 0.8366,
+      "step": 799
+    },
+    {
+      "epoch": 3.1620553359683794,
+      "grad_norm": 3.5305442810058594,
+      "learning_rate": 3.432539682539683e-05,
+      "loss": 0.8311,
+      "step": 800
+    },
+    {
+      "epoch": 3.16600790513834,
+      "grad_norm": 2.946333169937134,
+      "learning_rate": 3.430555555555556e-05,
+      "loss": 1.037,
+      "step": 801
+    },
+    {
+      "epoch": 3.1699604743083003,
+      "grad_norm": 3.5695722103118896,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.7785,
+      "step": 802
+    },
+    {
+      "epoch": 3.1739130434782608,
+      "grad_norm": 3.318615436553955,
+      "learning_rate": 3.426587301587302e-05,
+      "loss": 0.8239,
+      "step": 803
+    },
+    {
+      "epoch": 3.177865612648221,
+      "grad_norm": 3.4505763053894043,
+      "learning_rate": 3.424603174603175e-05,
+      "loss": 0.8528,
+      "step": 804
+    },
+    {
+      "epoch": 3.1818181818181817,
+      "grad_norm": 3.004587411880493,
+      "learning_rate": 3.422619047619048e-05,
+      "loss": 1.123,
+      "step": 805
+    },
+    {
+      "epoch": 3.185770750988142,
+      "grad_norm": 3.96537709236145,
+      "learning_rate": 3.420634920634921e-05,
+      "loss": 0.9566,
+      "step": 806
+    },
+    {
+      "epoch": 3.1897233201581026,
+      "grad_norm": 3.715635061264038,
+      "learning_rate": 3.418650793650794e-05,
+      "loss": 0.8639,
+      "step": 807
+    },
+    {
+      "epoch": 3.1936758893280635,
+      "grad_norm": 2.9800610542297363,
+      "learning_rate": 3.4166666666666666e-05,
+      "loss": 0.9108,
+      "step": 808
+    },
+    {
+      "epoch": 3.197628458498024,
+      "grad_norm": 3.7264318466186523,
+      "learning_rate": 3.41468253968254e-05,
+      "loss": 1.0782,
+      "step": 809
+    },
+    {
+      "epoch": 3.2015810276679844,
+      "grad_norm": 2.8650896549224854,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.7647,
+      "step": 810
+    },
+    {
+      "epoch": 3.205533596837945,
+      "grad_norm": 4.19207763671875,
+      "learning_rate": 3.410714285714286e-05,
+      "loss": 1.0105,
+      "step": 811
+    },
+    {
+      "epoch": 3.2094861660079053,
+      "grad_norm": 2.8715367317199707,
+      "learning_rate": 3.408730158730159e-05,
+      "loss": 1.0023,
+      "step": 812
+    },
+    {
+      "epoch": 3.2134387351778657,
+      "grad_norm": 5.560529708862305,
+      "learning_rate": 3.406746031746032e-05,
+      "loss": 0.8081,
+      "step": 813
+    },
+    {
+      "epoch": 3.217391304347826,
+      "grad_norm": 3.2364330291748047,
+      "learning_rate": 3.404761904761905e-05,
+      "loss": 0.8591,
+      "step": 814
+    },
+    {
+      "epoch": 3.2213438735177866,
+      "grad_norm": 2.6285791397094727,
+      "learning_rate": 3.402777777777778e-05,
+      "loss": 1.0187,
+      "step": 815
+    },
+    {
+      "epoch": 3.225296442687747,
+      "grad_norm": 3.7776012420654297,
+      "learning_rate": 3.400793650793651e-05,
+      "loss": 0.819,
+      "step": 816
+    },
+    {
+      "epoch": 3.2292490118577075,
+      "grad_norm": 2.7896721363067627,
+      "learning_rate": 3.398809523809524e-05,
+      "loss": 0.975,
+      "step": 817
+    },
+    {
+      "epoch": 3.233201581027668,
+      "grad_norm": 2.8350744247436523,
+      "learning_rate": 3.396825396825397e-05,
+      "loss": 1.3311,
+      "step": 818
+    },
+    {
+      "epoch": 3.2371541501976284,
+      "grad_norm": 3.2776432037353516,
+      "learning_rate": 3.3948412698412703e-05,
+      "loss": 0.8914,
+      "step": 819
+    },
+    {
+      "epoch": 3.241106719367589,
+      "grad_norm": 2.9362752437591553,
+      "learning_rate": 3.392857142857143e-05,
+      "loss": 1.0549,
+      "step": 820
+    },
+    {
+      "epoch": 3.2450592885375493,
+      "grad_norm": 2.536288022994995,
+      "learning_rate": 3.3908730158730164e-05,
+      "loss": 0.8083,
+      "step": 821
+    },
+    {
+      "epoch": 3.2490118577075098,
+      "grad_norm": 3.4325833320617676,
+      "learning_rate": 3.388888888888889e-05,
+      "loss": 0.9863,
+      "step": 822
+    },
+    {
+      "epoch": 3.2529644268774702,
+      "grad_norm": 2.966909646987915,
+      "learning_rate": 3.3869047619047625e-05,
+      "loss": 0.9928,
+      "step": 823
+    },
+    {
+      "epoch": 3.2569169960474307,
+      "grad_norm": 3.627739191055298,
+      "learning_rate": 3.384920634920635e-05,
+      "loss": 1.0572,
+      "step": 824
+    },
+    {
+      "epoch": 3.260869565217391,
+      "grad_norm": 3.5908288955688477,
+      "learning_rate": 3.3829365079365086e-05,
+      "loss": 1.1389,
+      "step": 825
+    },
+    {
+      "epoch": 3.2648221343873516,
+      "grad_norm": 2.7947778701782227,
+      "learning_rate": 3.380952380952381e-05,
+      "loss": 0.9508,
+      "step": 826
+    },
+    {
+      "epoch": 3.2687747035573125,
+      "grad_norm": 2.9585814476013184,
+      "learning_rate": 3.3789682539682546e-05,
+      "loss": 0.8809,
+      "step": 827
+    },
+    {
+      "epoch": 3.2727272727272725,
+      "grad_norm": 2.968035936355591,
+      "learning_rate": 3.3769841269841273e-05,
+      "loss": 1.0937,
+      "step": 828
+    },
+    {
+      "epoch": 3.2766798418972334,
+      "grad_norm": 4.545178413391113,
+      "learning_rate": 3.375000000000001e-05,
+      "loss": 1.3998,
+      "step": 829
+    },
+    {
+      "epoch": 3.280632411067194,
+      "grad_norm": 3.4058380126953125,
+      "learning_rate": 3.3730158730158734e-05,
+      "loss": 1.269,
+      "step": 830
+    },
+    {
+      "epoch": 3.2845849802371543,
+      "grad_norm": 3.8424315452575684,
+      "learning_rate": 3.371031746031746e-05,
+      "loss": 0.7128,
+      "step": 831
+    },
+    {
+      "epoch": 3.2885375494071147,
+      "grad_norm": 2.7813680171966553,
+      "learning_rate": 3.3690476190476195e-05,
+      "loss": 0.8108,
+      "step": 832
+    },
+    {
+      "epoch": 3.292490118577075,
+      "grad_norm": 2.6684579849243164,
+      "learning_rate": 3.367063492063492e-05,
+      "loss": 0.8819,
+      "step": 833
+    },
+    {
+      "epoch": 3.2964426877470356,
+      "grad_norm": 3.1254680156707764,
+      "learning_rate": 3.3650793650793656e-05,
+      "loss": 0.9083,
+      "step": 834
+    },
+    {
+      "epoch": 3.300395256916996,
+      "grad_norm": 2.6201884746551514,
+      "learning_rate": 3.363095238095238e-05,
+      "loss": 0.751,
+      "step": 835
+    },
+    {
+      "epoch": 3.3043478260869565,
+      "grad_norm": 2.976027250289917,
+      "learning_rate": 3.3611111111111116e-05,
+      "loss": 0.7687,
+      "step": 836
+    },
+    {
+      "epoch": 3.308300395256917,
+      "grad_norm": 4.267848968505859,
+      "learning_rate": 3.3591269841269843e-05,
+      "loss": 0.8482,
+      "step": 837
+    },
+    {
+      "epoch": 3.3122529644268774,
+      "grad_norm": 4.229562759399414,
+      "learning_rate": 3.357142857142857e-05,
+      "loss": 0.9975,
+      "step": 838
+    },
+    {
+      "epoch": 3.316205533596838,
+      "grad_norm": 3.5600497722625732,
+      "learning_rate": 3.35515873015873e-05,
+      "loss": 1.197,
+      "step": 839
+    },
+    {
+      "epoch": 3.3201581027667983,
+      "grad_norm": 3.5733120441436768,
+      "learning_rate": 3.353174603174603e-05,
+      "loss": 0.9723,
+      "step": 840
+    },
+    {
+      "epoch": 3.324110671936759,
+      "grad_norm": 4.316638469696045,
+      "learning_rate": 3.351190476190476e-05,
+      "loss": 1.0709,
+      "step": 841
+    },
+    {
+      "epoch": 3.3280632411067192,
+      "grad_norm": 2.8753249645233154,
+      "learning_rate": 3.349206349206349e-05,
+      "loss": 1.044,
+      "step": 842
+    },
+    {
+      "epoch": 3.3320158102766797,
+      "grad_norm": 3.8951146602630615,
+      "learning_rate": 3.347222222222222e-05,
+      "loss": 1.2154,
+      "step": 843
+    },
+    {
+      "epoch": 3.33596837944664,
+      "grad_norm": 2.5250723361968994,
+      "learning_rate": 3.345238095238095e-05,
+      "loss": 0.7862,
+      "step": 844
+    },
+    {
+      "epoch": 3.3399209486166006,
+      "grad_norm": 2.821502447128296,
+      "learning_rate": 3.343253968253968e-05,
+      "loss": 1.0455,
+      "step": 845
+    },
+    {
+      "epoch": 3.3438735177865615,
+      "grad_norm": 2.9029202461242676,
+      "learning_rate": 3.3412698412698413e-05,
+      "loss": 1.1058,
+      "step": 846
+    },
+    {
+      "epoch": 3.3478260869565215,
+      "grad_norm": 2.247542381286621,
+      "learning_rate": 3.339285714285714e-05,
+      "loss": 0.9376,
+      "step": 847
+    },
+    {
+      "epoch": 3.3517786561264824,
+      "grad_norm": 4.182517051696777,
+      "learning_rate": 3.3373015873015874e-05,
+      "loss": 0.7921,
+      "step": 848
+    },
+    {
+      "epoch": 3.355731225296443,
+      "grad_norm": 2.482083797454834,
+      "learning_rate": 3.33531746031746e-05,
+      "loss": 0.9948,
+      "step": 849
+    },
+    {
+      "epoch": 3.3596837944664033,
+      "grad_norm": 2.999898672103882,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.8631,
+      "step": 850
+    },
+    {
+      "epoch": 3.3636363636363638,
+      "grad_norm": 4.035496234893799,
+      "learning_rate": 3.331349206349206e-05,
+      "loss": 0.8496,
+      "step": 851
+    },
+    {
+      "epoch": 3.367588932806324,
+      "grad_norm": 3.46382737159729,
+      "learning_rate": 3.3293650793650796e-05,
+      "loss": 0.7815,
+      "step": 852
+    },
+    {
+      "epoch": 3.3715415019762847,
+      "grad_norm": 2.803903818130493,
+      "learning_rate": 3.327380952380952e-05,
+      "loss": 0.8237,
+      "step": 853
+    },
+    {
+      "epoch": 3.375494071146245,
+      "grad_norm": 3.75998592376709,
+      "learning_rate": 3.3253968253968256e-05,
+      "loss": 1.183,
+      "step": 854
+    },
+    {
+      "epoch": 3.3794466403162056,
+      "grad_norm": 3.8531124591827393,
+      "learning_rate": 3.3234126984126983e-05,
+      "loss": 1.105,
+      "step": 855
+    },
+    {
+      "epoch": 3.383399209486166,
+      "grad_norm": 4.231212615966797,
+      "learning_rate": 3.321428571428572e-05,
+      "loss": 0.8909,
+      "step": 856
+    },
+    {
+      "epoch": 3.3873517786561265,
+      "grad_norm": 3.3275704383850098,
+      "learning_rate": 3.3194444444444444e-05,
+      "loss": 0.8102,
+      "step": 857
+    },
+    {
+      "epoch": 3.391304347826087,
+      "grad_norm": 3.103105306625366,
+      "learning_rate": 3.317460317460318e-05,
+      "loss": 0.6757,
+      "step": 858
+    },
+    {
+      "epoch": 3.3952569169960474,
+      "grad_norm": 4.611384391784668,
+      "learning_rate": 3.3154761904761905e-05,
+      "loss": 1.113,
+      "step": 859
+    },
+    {
+      "epoch": 3.399209486166008,
+      "grad_norm": 2.466308355331421,
+      "learning_rate": 3.313492063492064e-05,
+      "loss": 0.758,
+      "step": 860
+    },
+    {
+      "epoch": 3.4031620553359683,
+      "grad_norm": 2.4484543800354004,
+      "learning_rate": 3.3115079365079366e-05,
+      "loss": 0.9824,
+      "step": 861
+    },
+    {
+      "epoch": 3.4071146245059287,
+      "grad_norm": 2.8706154823303223,
+      "learning_rate": 3.309523809523809e-05,
+      "loss": 0.9201,
+      "step": 862
+    },
+    {
+      "epoch": 3.411067193675889,
+      "grad_norm": 3.8132359981536865,
+      "learning_rate": 3.3075396825396826e-05,
+      "loss": 0.6403,
+      "step": 863
+    },
+    {
+      "epoch": 3.4150197628458496,
+      "grad_norm": 3.109771251678467,
+      "learning_rate": 3.3055555555555553e-05,
+      "loss": 0.756,
+      "step": 864
+    },
+    {
+      "epoch": 3.4189723320158105,
+      "grad_norm": 3.4301555156707764,
+      "learning_rate": 3.303571428571429e-05,
+      "loss": 1.0933,
+      "step": 865
+    },
+    {
+      "epoch": 3.4229249011857705,
+      "grad_norm": 3.461763620376587,
+      "learning_rate": 3.3015873015873014e-05,
+      "loss": 1.2169,
+      "step": 866
+    },
+    {
+      "epoch": 3.4268774703557314,
+      "grad_norm": 4.107053756713867,
+      "learning_rate": 3.299603174603175e-05,
+      "loss": 1.0824,
+      "step": 867
+    },
+    {
+      "epoch": 3.430830039525692,
+      "grad_norm": 3.434462070465088,
+      "learning_rate": 3.2976190476190475e-05,
+      "loss": 1.0065,
+      "step": 868
+    },
+    {
+      "epoch": 3.4347826086956523,
+      "grad_norm": 3.3090224266052246,
+      "learning_rate": 3.295634920634921e-05,
+      "loss": 0.9509,
+      "step": 869
+    },
+    {
+      "epoch": 3.438735177865613,
+      "grad_norm": 3.5097386837005615,
+      "learning_rate": 3.2936507936507936e-05,
+      "loss": 0.8331,
+      "step": 870
+    },
+    {
+      "epoch": 3.4426877470355732,
+      "grad_norm": 3.1316442489624023,
+      "learning_rate": 3.291666666666667e-05,
+      "loss": 0.8634,
+      "step": 871
+    },
+    {
+      "epoch": 3.4466403162055337,
+      "grad_norm": 2.5135436058044434,
+      "learning_rate": 3.2896825396825396e-05,
+      "loss": 0.6734,
+      "step": 872
+    },
+    {
+      "epoch": 3.450592885375494,
+      "grad_norm": 3.5739667415618896,
+      "learning_rate": 3.287698412698413e-05,
+      "loss": 0.8149,
+      "step": 873
+    },
+    {
+      "epoch": 3.4545454545454546,
+      "grad_norm": 2.84611439704895,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.8658,
+      "step": 874
+    },
+    {
+      "epoch": 3.458498023715415,
+      "grad_norm": 2.680215835571289,
+      "learning_rate": 3.283730158730159e-05,
+      "loss": 1.1295,
+      "step": 875
+    },
+    {
+      "epoch": 3.4624505928853755,
+      "grad_norm": 3.028846263885498,
+      "learning_rate": 3.281746031746032e-05,
+      "loss": 1.0244,
+      "step": 876
+    },
+    {
+      "epoch": 3.466403162055336,
+      "grad_norm": 2.5386555194854736,
+      "learning_rate": 3.279761904761905e-05,
+      "loss": 1.0598,
+      "step": 877
+    },
+    {
+      "epoch": 3.4703557312252964,
+      "grad_norm": 2.7928504943847656,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 1.181,
+      "step": 878
+    },
+    {
+      "epoch": 3.474308300395257,
+      "grad_norm": 3.864605665206909,
+      "learning_rate": 3.275793650793651e-05,
+      "loss": 0.9425,
+      "step": 879
+    },
+    {
+      "epoch": 3.4782608695652173,
+      "grad_norm": 2.8199121952056885,
+      "learning_rate": 3.273809523809524e-05,
+      "loss": 1.0021,
+      "step": 880
+    },
+    {
+      "epoch": 3.4822134387351777,
+      "grad_norm": 3.323481321334839,
+      "learning_rate": 3.271825396825397e-05,
+      "loss": 1.141,
+      "step": 881
+    },
+    {
+      "epoch": 3.486166007905138,
+      "grad_norm": 3.6544790267944336,
+      "learning_rate": 3.26984126984127e-05,
+      "loss": 0.9603,
+      "step": 882
+    },
+    {
+      "epoch": 3.4901185770750986,
+      "grad_norm": 3.930521011352539,
+      "learning_rate": 3.2678571428571434e-05,
+      "loss": 0.9759,
+      "step": 883
+    },
+    {
+      "epoch": 3.494071146245059,
+      "grad_norm": 2.638461112976074,
+      "learning_rate": 3.265873015873016e-05,
+      "loss": 1.0179,
+      "step": 884
+    },
+    {
+      "epoch": 3.4980237154150196,
+      "grad_norm": 3.1710309982299805,
+      "learning_rate": 3.263888888888889e-05,
+      "loss": 0.9966,
+      "step": 885
+    },
+    {
+      "epoch": 3.5019762845849804,
+      "grad_norm": 4.251243591308594,
+      "learning_rate": 3.261904761904762e-05,
+      "loss": 0.9437,
+      "step": 886
+    },
+    {
+      "epoch": 3.5059288537549405,
+      "grad_norm": 3.0481278896331787,
+      "learning_rate": 3.259920634920635e-05,
+      "loss": 0.99,
+      "step": 887
+    },
+    {
+      "epoch": 3.5098814229249014,
+      "grad_norm": 2.843862295150757,
+      "learning_rate": 3.257936507936508e-05,
+      "loss": 1.057,
+      "step": 888
+    },
+    {
+      "epoch": 3.513833992094862,
+      "grad_norm": 2.6303539276123047,
+      "learning_rate": 3.255952380952381e-05,
+      "loss": 1.0664,
+      "step": 889
+    },
+    {
+      "epoch": 3.5177865612648223,
+      "grad_norm": 2.792405128479004,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 1.0506,
+      "step": 890
+    },
+    {
+      "epoch": 3.5217391304347827,
+      "grad_norm": 5.069064140319824,
+      "learning_rate": 3.251984126984127e-05,
+      "loss": 0.815,
+      "step": 891
+    },
+    {
+      "epoch": 3.525691699604743,
+      "grad_norm": 2.9551217555999756,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.1707,
+      "step": 892
+    },
+    {
+      "epoch": 3.5296442687747036,
+      "grad_norm": 3.7232065200805664,
+      "learning_rate": 3.248015873015873e-05,
+      "loss": 1.0063,
+      "step": 893
+    },
+    {
+      "epoch": 3.533596837944664,
+      "grad_norm": 3.990532875061035,
+      "learning_rate": 3.2460317460317465e-05,
+      "loss": 0.8298,
+      "step": 894
+    },
+    {
+      "epoch": 3.5375494071146245,
+      "grad_norm": 4.859503269195557,
+      "learning_rate": 3.244047619047619e-05,
+      "loss": 0.8827,
+      "step": 895
+    },
+    {
+      "epoch": 3.541501976284585,
+      "grad_norm": 3.32344913482666,
+      "learning_rate": 3.2420634920634925e-05,
+      "loss": 0.8536,
+      "step": 896
+    },
+    {
+      "epoch": 3.5454545454545454,
+      "grad_norm": 3.2624435424804688,
+      "learning_rate": 3.240079365079365e-05,
+      "loss": 1.2324,
+      "step": 897
+    },
+    {
+      "epoch": 3.549407114624506,
+      "grad_norm": 4.429110050201416,
+      "learning_rate": 3.2380952380952386e-05,
+      "loss": 1.1154,
+      "step": 898
+    },
+    {
+      "epoch": 3.5533596837944663,
+      "grad_norm": 3.8388755321502686,
+      "learning_rate": 3.236111111111111e-05,
+      "loss": 1.2359,
+      "step": 899
+    },
+    {
+      "epoch": 3.5573122529644268,
+      "grad_norm": 3.2584800720214844,
+      "learning_rate": 3.234126984126985e-05,
+      "loss": 0.9469,
+      "step": 900
+    },
+    {
+      "epoch": 3.561264822134387,
+      "grad_norm": 3.2762997150421143,
+      "learning_rate": 3.2321428571428574e-05,
+      "loss": 0.7991,
+      "step": 901
+    },
+    {
+      "epoch": 3.5652173913043477,
+      "grad_norm": 3.214747905731201,
+      "learning_rate": 3.230158730158731e-05,
+      "loss": 1.2799,
+      "step": 902
+    },
+    {
+      "epoch": 3.5691699604743086,
+      "grad_norm": 3.528118133544922,
+      "learning_rate": 3.2281746031746035e-05,
+      "loss": 0.9679,
+      "step": 903
+    },
+    {
+      "epoch": 3.5731225296442686,
+      "grad_norm": 2.9673147201538086,
+      "learning_rate": 3.226190476190477e-05,
+      "loss": 1.0996,
+      "step": 904
+    },
+    {
+      "epoch": 3.5770750988142295,
+      "grad_norm": 2.345867395401001,
+      "learning_rate": 3.2242063492063495e-05,
+      "loss": 0.8981,
+      "step": 905
+    },
+    {
+      "epoch": 3.5810276679841895,
+      "grad_norm": 2.914339065551758,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 1.2681,
+      "step": 906
+    },
+    {
+      "epoch": 3.5849802371541504,
+      "grad_norm": 2.6776278018951416,
+      "learning_rate": 3.2202380952380956e-05,
+      "loss": 0.9516,
+      "step": 907
+    },
+    {
+      "epoch": 3.588932806324111,
+      "grad_norm": 3.282893180847168,
+      "learning_rate": 3.218253968253968e-05,
+      "loss": 1.1996,
+      "step": 908
+    },
+    {
+      "epoch": 3.5928853754940713,
+      "grad_norm": 2.855419635772705,
+      "learning_rate": 3.216269841269842e-05,
+      "loss": 0.9228,
+      "step": 909
+    },
+    {
+      "epoch": 3.5968379446640317,
+      "grad_norm": 3.372144937515259,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 1.1176,
+      "step": 910
+    },
+    {
+      "epoch": 3.600790513833992,
+      "grad_norm": 2.9246695041656494,
+      "learning_rate": 3.212301587301588e-05,
+      "loss": 0.8848,
+      "step": 911
+    },
+    {
+      "epoch": 3.6047430830039526,
+      "grad_norm": 3.258700370788574,
+      "learning_rate": 3.2103174603174605e-05,
+      "loss": 1.1232,
+      "step": 912
+    },
+    {
+      "epoch": 3.608695652173913,
+      "grad_norm": 3.726252794265747,
+      "learning_rate": 3.208333333333334e-05,
+      "loss": 0.9653,
+      "step": 913
+    },
+    {
+      "epoch": 3.6126482213438735,
+      "grad_norm": 3.1601107120513916,
+      "learning_rate": 3.2063492063492065e-05,
+      "loss": 1.039,
+      "step": 914
+    },
+    {
+      "epoch": 3.616600790513834,
+      "grad_norm": 3.5057191848754883,
+      "learning_rate": 3.20436507936508e-05,
+      "loss": 0.9399,
+      "step": 915
+    },
+    {
+      "epoch": 3.6205533596837944,
+      "grad_norm": 3.562908411026001,
+      "learning_rate": 3.202380952380952e-05,
+      "loss": 1.0264,
+      "step": 916
+    },
+    {
+      "epoch": 3.624505928853755,
+      "grad_norm": 3.443648099899292,
+      "learning_rate": 3.200396825396825e-05,
+      "loss": 0.8603,
+      "step": 917
+    },
+    {
+      "epoch": 3.6284584980237153,
+      "grad_norm": 3.095889091491699,
+      "learning_rate": 3.198412698412698e-05,
+      "loss": 1.1861,
+      "step": 918
+    },
+    {
+      "epoch": 3.632411067193676,
+      "grad_norm": 3.4708924293518066,
+      "learning_rate": 3.1964285714285714e-05,
+      "loss": 1.1105,
+      "step": 919
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 2.894131660461426,
+      "learning_rate": 3.194444444444444e-05,
+      "loss": 1.07,
+      "step": 920
+    },
+    {
+      "epoch": 3.6403162055335967,
+      "grad_norm": 4.143770694732666,
+      "learning_rate": 3.1924603174603175e-05,
+      "loss": 0.9666,
+      "step": 921
+    },
+    {
+      "epoch": 3.6442687747035576,
+      "grad_norm": 2.9045121669769287,
+      "learning_rate": 3.19047619047619e-05,
+      "loss": 1.0069,
+      "step": 922
+    },
+    {
+      "epoch": 3.6482213438735176,
+      "grad_norm": 2.5982000827789307,
+      "learning_rate": 3.1884920634920635e-05,
+      "loss": 1.2217,
+      "step": 923
+    },
+    {
+      "epoch": 3.6521739130434785,
+      "grad_norm": 2.874178647994995,
+      "learning_rate": 3.186507936507936e-05,
+      "loss": 1.1494,
+      "step": 924
+    },
+    {
+      "epoch": 3.6561264822134385,
+      "grad_norm": 3.3155198097229004,
+      "learning_rate": 3.1845238095238096e-05,
+      "loss": 1.1825,
+      "step": 925
+    },
+    {
+      "epoch": 3.6600790513833994,
+      "grad_norm": 3.4435360431671143,
+      "learning_rate": 3.182539682539682e-05,
+      "loss": 0.7864,
+      "step": 926
+    },
+    {
+      "epoch": 3.66403162055336,
+      "grad_norm": 3.355100393295288,
+      "learning_rate": 3.180555555555556e-05,
+      "loss": 0.8861,
+      "step": 927
+    },
+    {
+      "epoch": 3.6679841897233203,
+      "grad_norm": 3.6266121864318848,
+      "learning_rate": 3.1785714285714284e-05,
+      "loss": 1.0996,
+      "step": 928
+    },
+    {
+      "epoch": 3.6719367588932808,
+      "grad_norm": 3.77498459815979,
+      "learning_rate": 3.176587301587302e-05,
+      "loss": 0.7629,
+      "step": 929
+    },
+    {
+      "epoch": 3.675889328063241,
+      "grad_norm": 3.8526268005371094,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 1.1476,
+      "step": 930
+    },
+    {
+      "epoch": 3.6798418972332017,
+      "grad_norm": 2.970158338546753,
+      "learning_rate": 3.172619047619048e-05,
+      "loss": 1.1171,
+      "step": 931
+    },
+    {
+      "epoch": 3.683794466403162,
+      "grad_norm": 2.934819459915161,
+      "learning_rate": 3.1706349206349205e-05,
+      "loss": 0.964,
+      "step": 932
+    },
+    {
+      "epoch": 3.6877470355731226,
+      "grad_norm": 3.0377979278564453,
+      "learning_rate": 3.168650793650794e-05,
+      "loss": 1.1309,
+      "step": 933
+    },
+    {
+      "epoch": 3.691699604743083,
+      "grad_norm": 3.5154223442077637,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.9907,
+      "step": 934
+    },
+    {
+      "epoch": 3.6956521739130435,
+      "grad_norm": 3.415923833847046,
+      "learning_rate": 3.16468253968254e-05,
+      "loss": 0.9697,
+      "step": 935
+    },
+    {
+      "epoch": 3.699604743083004,
+      "grad_norm": 3.0694198608398438,
+      "learning_rate": 3.162698412698413e-05,
+      "loss": 0.9632,
+      "step": 936
+    },
+    {
+      "epoch": 3.7035573122529644,
+      "grad_norm": 2.25923490524292,
+      "learning_rate": 3.160714285714286e-05,
+      "loss": 0.9753,
+      "step": 937
+    },
+    {
+      "epoch": 3.707509881422925,
+      "grad_norm": 4.5253190994262695,
+      "learning_rate": 3.158730158730159e-05,
+      "loss": 0.9479,
+      "step": 938
+    },
+    {
+      "epoch": 3.7114624505928853,
+      "grad_norm": 2.794191598892212,
+      "learning_rate": 3.1567460317460315e-05,
+      "loss": 0.9286,
+      "step": 939
+    },
+    {
+      "epoch": 3.7154150197628457,
+      "grad_norm": 3.3554208278656006,
+      "learning_rate": 3.154761904761905e-05,
+      "loss": 1.0205,
+      "step": 940
+    },
+    {
+      "epoch": 3.719367588932806,
+      "grad_norm": 3.562750816345215,
+      "learning_rate": 3.1527777777777775e-05,
+      "loss": 0.7624,
+      "step": 941
+    },
+    {
+      "epoch": 3.7233201581027666,
+      "grad_norm": 2.868648052215576,
+      "learning_rate": 3.150793650793651e-05,
+      "loss": 0.8968,
+      "step": 942
+    },
+    {
+      "epoch": 3.7272727272727275,
+      "grad_norm": 3.128340721130371,
+      "learning_rate": 3.1488095238095236e-05,
+      "loss": 0.9745,
+      "step": 943
+    },
+    {
+      "epoch": 3.7312252964426875,
+      "grad_norm": 3.5969157218933105,
+      "learning_rate": 3.146825396825397e-05,
+      "loss": 1.0403,
+      "step": 944
+    },
+    {
+      "epoch": 3.7351778656126484,
+      "grad_norm": 3.2540881633758545,
+      "learning_rate": 3.14484126984127e-05,
+      "loss": 1.1106,
+      "step": 945
+    },
+    {
+      "epoch": 3.7391304347826084,
+      "grad_norm": 2.8699069023132324,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.9645,
+      "step": 946
+    },
+    {
+      "epoch": 3.7430830039525693,
+      "grad_norm": 3.2061824798583984,
+      "learning_rate": 3.140873015873016e-05,
+      "loss": 0.864,
+      "step": 947
+    },
+    {
+      "epoch": 3.7470355731225298,
+      "grad_norm": 2.7899482250213623,
+      "learning_rate": 3.138888888888889e-05,
+      "loss": 1.1192,
+      "step": 948
+    },
+    {
+      "epoch": 3.7509881422924902,
+      "grad_norm": 4.122161388397217,
+      "learning_rate": 3.136904761904762e-05,
+      "loss": 0.8513,
+      "step": 949
+    },
+    {
+      "epoch": 3.7549407114624507,
+      "grad_norm": 2.6482725143432617,
+      "learning_rate": 3.134920634920635e-05,
+      "loss": 0.811,
+      "step": 950
+    },
+    {
+      "epoch": 3.758893280632411,
+      "grad_norm": 2.7796828746795654,
+      "learning_rate": 3.132936507936508e-05,
+      "loss": 0.9356,
+      "step": 951
+    },
+    {
+      "epoch": 3.7628458498023716,
+      "grad_norm": 3.4659996032714844,
+      "learning_rate": 3.130952380952381e-05,
+      "loss": 1.0248,
+      "step": 952
+    },
+    {
+      "epoch": 3.766798418972332,
+      "grad_norm": 3.140477418899536,
+      "learning_rate": 3.128968253968254e-05,
+      "loss": 1.2593,
+      "step": 953
+    },
+    {
+      "epoch": 3.7707509881422925,
+      "grad_norm": 3.8717284202575684,
+      "learning_rate": 3.1269841269841274e-05,
+      "loss": 1.0904,
+      "step": 954
+    },
+    {
+      "epoch": 3.774703557312253,
+      "grad_norm": 2.737684488296509,
+      "learning_rate": 3.125e-05,
+      "loss": 0.8346,
+      "step": 955
+    },
+    {
+      "epoch": 3.7786561264822134,
+      "grad_norm": 3.515249013900757,
+      "learning_rate": 3.1230158730158734e-05,
+      "loss": 1.1003,
+      "step": 956
+    },
+    {
+      "epoch": 3.782608695652174,
+      "grad_norm": 3.8520002365112305,
+      "learning_rate": 3.121031746031746e-05,
+      "loss": 0.8998,
+      "step": 957
+    },
+    {
+      "epoch": 3.7865612648221343,
+      "grad_norm": 3.644209384918213,
+      "learning_rate": 3.1190476190476195e-05,
+      "loss": 1.0063,
+      "step": 958
+    },
+    {
+      "epoch": 3.7905138339920947,
+      "grad_norm": 3.2415544986724854,
+      "learning_rate": 3.117063492063492e-05,
+      "loss": 0.8328,
+      "step": 959
+    },
+    {
+      "epoch": 3.794466403162055,
+      "grad_norm": 3.8536369800567627,
+      "learning_rate": 3.1150793650793656e-05,
+      "loss": 0.9605,
+      "step": 960
+    },
+    {
+      "epoch": 3.7984189723320156,
+      "grad_norm": 2.8492162227630615,
+      "learning_rate": 3.113095238095238e-05,
+      "loss": 0.8059,
+      "step": 961
+    },
+    {
+      "epoch": 3.8023715415019765,
+      "grad_norm": 3.2966291904449463,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.9345,
+      "step": 962
+    },
+    {
+      "epoch": 3.8063241106719365,
+      "grad_norm": 3.0209009647369385,
+      "learning_rate": 3.1091269841269844e-05,
+      "loss": 0.9933,
+      "step": 963
+    },
+    {
+      "epoch": 3.8102766798418974,
+      "grad_norm": 2.6817831993103027,
+      "learning_rate": 3.107142857142857e-05,
+      "loss": 0.9633,
+      "step": 964
+    },
+    {
+      "epoch": 3.8142292490118574,
+      "grad_norm": 3.1547319889068604,
+      "learning_rate": 3.1051587301587304e-05,
+      "loss": 1.1855,
+      "step": 965
+    },
+    {
+      "epoch": 3.8181818181818183,
+      "grad_norm": 3.485544443130493,
+      "learning_rate": 3.103174603174603e-05,
+      "loss": 1.1107,
+      "step": 966
+    },
+    {
+      "epoch": 3.822134387351779,
+      "grad_norm": 3.0962092876434326,
+      "learning_rate": 3.1011904761904765e-05,
+      "loss": 1.2013,
+      "step": 967
+    },
+    {
+      "epoch": 3.8260869565217392,
+      "grad_norm": 3.373776912689209,
+      "learning_rate": 3.099206349206349e-05,
+      "loss": 0.7553,
+      "step": 968
+    },
+    {
+      "epoch": 3.8300395256916997,
+      "grad_norm": 3.174527883529663,
+      "learning_rate": 3.0972222222222226e-05,
+      "loss": 1.1787,
+      "step": 969
+    },
+    {
+      "epoch": 3.83399209486166,
+      "grad_norm": 3.290992498397827,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.9103,
+      "step": 970
+    },
+    {
+      "epoch": 3.8379446640316206,
+      "grad_norm": 3.8229081630706787,
+      "learning_rate": 3.093253968253969e-05,
+      "loss": 0.9309,
+      "step": 971
+    },
+    {
+      "epoch": 3.841897233201581,
+      "grad_norm": 3.5504062175750732,
+      "learning_rate": 3.0912698412698414e-05,
+      "loss": 1.0507,
+      "step": 972
+    },
+    {
+      "epoch": 3.8458498023715415,
+      "grad_norm": 3.4466750621795654,
+      "learning_rate": 3.089285714285715e-05,
+      "loss": 1.056,
+      "step": 973
+    },
+    {
+      "epoch": 3.849802371541502,
+      "grad_norm": 4.083953857421875,
+      "learning_rate": 3.0873015873015874e-05,
+      "loss": 0.9524,
+      "step": 974
+    },
+    {
+      "epoch": 3.8537549407114624,
+      "grad_norm": 4.008061408996582,
+      "learning_rate": 3.085317460317461e-05,
+      "loss": 1.0339,
+      "step": 975
+    },
+    {
+      "epoch": 3.857707509881423,
+      "grad_norm": 2.528346300125122,
+      "learning_rate": 3.0833333333333335e-05,
+      "loss": 1.0405,
+      "step": 976
+    },
+    {
+      "epoch": 3.8616600790513833,
+      "grad_norm": 2.7856969833374023,
+      "learning_rate": 3.081349206349207e-05,
+      "loss": 0.878,
+      "step": 977
+    },
+    {
+      "epoch": 3.8656126482213438,
+      "grad_norm": 3.5250959396362305,
+      "learning_rate": 3.0793650793650796e-05,
+      "loss": 1.0533,
+      "step": 978
+    },
+    {
+      "epoch": 3.869565217391304,
+      "grad_norm": 3.1857500076293945,
+      "learning_rate": 3.077380952380953e-05,
+      "loss": 1.1474,
+      "step": 979
+    },
+    {
+      "epoch": 3.8735177865612647,
+      "grad_norm": 3.111074209213257,
+      "learning_rate": 3.075396825396826e-05,
+      "loss": 1.0187,
+      "step": 980
+    },
+    {
+      "epoch": 3.8774703557312256,
+      "grad_norm": 3.149482250213623,
+      "learning_rate": 3.073412698412699e-05,
+      "loss": 0.8251,
+      "step": 981
+    },
+    {
+      "epoch": 3.8814229249011856,
+      "grad_norm": 3.7290894985198975,
+      "learning_rate": 3.071428571428572e-05,
+      "loss": 0.8504,
+      "step": 982
+    },
+    {
+      "epoch": 3.8853754940711465,
+      "grad_norm": 2.8733174800872803,
+      "learning_rate": 3.069444444444445e-05,
+      "loss": 0.7323,
+      "step": 983
+    },
+    {
+      "epoch": 3.8893280632411065,
+      "grad_norm": 3.1158406734466553,
+      "learning_rate": 3.067460317460318e-05,
+      "loss": 0.8583,
+      "step": 984
+    },
+    {
+      "epoch": 3.8932806324110674,
+      "grad_norm": 3.126250743865967,
+      "learning_rate": 3.0654761904761905e-05,
+      "loss": 1.0742,
+      "step": 985
+    },
+    {
+      "epoch": 3.897233201581028,
+      "grad_norm": 4.984484672546387,
+      "learning_rate": 3.063492063492064e-05,
+      "loss": 1.0604,
+      "step": 986
+    },
+    {
+      "epoch": 3.9011857707509883,
+      "grad_norm": 3.3782174587249756,
+      "learning_rate": 3.0615079365079366e-05,
+      "loss": 0.8303,
+      "step": 987
+    },
+    {
+      "epoch": 3.9051383399209487,
+      "grad_norm": 3.0348994731903076,
+      "learning_rate": 3.05952380952381e-05,
+      "loss": 1.0685,
+      "step": 988
+    },
+    {
+      "epoch": 3.909090909090909,
+      "grad_norm": 3.0283362865448,
+      "learning_rate": 3.057539682539683e-05,
+      "loss": 1.1548,
+      "step": 989
+    },
+    {
+      "epoch": 3.9130434782608696,
+      "grad_norm": 3.938385009765625,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.7831,
+      "step": 990
+    },
+    {
+      "epoch": 3.91699604743083,
+      "grad_norm": 3.3874075412750244,
+      "learning_rate": 3.053571428571429e-05,
+      "loss": 0.7838,
+      "step": 991
+    },
+    {
+      "epoch": 3.9209486166007905,
+      "grad_norm": 3.09395432472229,
+      "learning_rate": 3.051587301587302e-05,
+      "loss": 0.7563,
+      "step": 992
+    },
+    {
+      "epoch": 3.924901185770751,
+      "grad_norm": 3.3949618339538574,
+      "learning_rate": 3.049603174603175e-05,
+      "loss": 1.2432,
+      "step": 993
+    },
+    {
+      "epoch": 3.9288537549407114,
+      "grad_norm": 3.91082501411438,
+      "learning_rate": 3.0476190476190482e-05,
+      "loss": 0.8746,
+      "step": 994
+    },
+    {
+      "epoch": 3.932806324110672,
+      "grad_norm": 3.182863235473633,
+      "learning_rate": 3.0456349206349206e-05,
+      "loss": 0.9189,
+      "step": 995
+    },
+    {
+      "epoch": 3.9367588932806323,
+      "grad_norm": 3.7233152389526367,
+      "learning_rate": 3.0436507936507936e-05,
+      "loss": 0.9538,
+      "step": 996
+    },
+    {
+      "epoch": 3.940711462450593,
+      "grad_norm": 4.027876853942871,
+      "learning_rate": 3.0416666666666666e-05,
+      "loss": 0.8883,
+      "step": 997
+    },
+    {
+      "epoch": 3.9446640316205532,
+      "grad_norm": 3.228454113006592,
+      "learning_rate": 3.0396825396825397e-05,
+      "loss": 0.9645,
+      "step": 998
+    },
+    {
+      "epoch": 3.9486166007905137,
+      "grad_norm": 3.264758348464966,
+      "learning_rate": 3.0376984126984127e-05,
+      "loss": 0.9099,
+      "step": 999
+    },
+    {
+      "epoch": 3.9525691699604746,
+      "grad_norm": 3.349977731704712,
+      "learning_rate": 3.0357142857142857e-05,
+      "loss": 0.731,
+      "step": 1000
+    },
+    {
+      "epoch": 3.9565217391304346,
+      "grad_norm": 3.3020308017730713,
+      "learning_rate": 3.0337301587301588e-05,
+      "loss": 0.9018,
+      "step": 1001
+    },
+    {
+      "epoch": 3.9604743083003955,
+      "grad_norm": 3.4217529296875,
+      "learning_rate": 3.0317460317460318e-05,
+      "loss": 1.1229,
+      "step": 1002
+    },
+    {
+      "epoch": 3.9644268774703555,
+      "grad_norm": 3.2917253971099854,
+      "learning_rate": 3.029761904761905e-05,
+      "loss": 1.0188,
+      "step": 1003
+    },
+    {
+      "epoch": 3.9683794466403164,
+      "grad_norm": 3.4840149879455566,
+      "learning_rate": 3.0277777777777776e-05,
+      "loss": 0.987,
+      "step": 1004
+    },
+    {
+      "epoch": 3.972332015810277,
+      "grad_norm": 3.3869576454162598,
+      "learning_rate": 3.0257936507936506e-05,
+      "loss": 0.8482,
+      "step": 1005
+    },
+    {
+      "epoch": 3.9762845849802373,
+      "grad_norm": 4.020000457763672,
+      "learning_rate": 3.0238095238095236e-05,
+      "loss": 0.9846,
+      "step": 1006
+    },
+    {
+      "epoch": 3.9802371541501977,
+      "grad_norm": 3.028305768966675,
+      "learning_rate": 3.0218253968253967e-05,
+      "loss": 0.8622,
+      "step": 1007
+    },
+    {
+      "epoch": 3.984189723320158,
+      "grad_norm": 2.8138604164123535,
+      "learning_rate": 3.0198412698412697e-05,
+      "loss": 1.2293,
+      "step": 1008
+    },
+    {
+      "epoch": 3.9881422924901186,
+      "grad_norm": 3.196972370147705,
+      "learning_rate": 3.0178571428571427e-05,
+      "loss": 1.0762,
+      "step": 1009
+    },
+    {
+      "epoch": 3.992094861660079,
+      "grad_norm": 4.044905185699463,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 1.01,
+      "step": 1010
+    },
+    {
+      "epoch": 3.9960474308300395,
+      "grad_norm": 2.731539011001587,
+      "learning_rate": 3.0138888888888888e-05,
+      "loss": 0.9882,
+      "step": 1011
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.787231683731079,
+      "learning_rate": 3.011904761904762e-05,
+      "loss": 0.7786,
+      "step": 1012
+    },
+    {
+      "epoch": 4.003952569169961,
+      "grad_norm": 2.597419023513794,
+      "learning_rate": 3.009920634920635e-05,
+      "loss": 0.5683,
+      "step": 1013
+    },
+    {
+      "epoch": 4.007905138339921,
+      "grad_norm": 3.2427003383636475,
+      "learning_rate": 3.007936507936508e-05,
+      "loss": 0.9022,
+      "step": 1014
+    },
+    {
+      "epoch": 4.011857707509882,
+      "grad_norm": 2.5612714290618896,
+      "learning_rate": 3.005952380952381e-05,
+      "loss": 0.4494,
+      "step": 1015
+    },
+    {
+      "epoch": 4.015810276679842,
+      "grad_norm": 3.5918335914611816,
+      "learning_rate": 3.003968253968254e-05,
+      "loss": 0.5691,
+      "step": 1016
+    },
+    {
+      "epoch": 4.019762845849803,
+      "grad_norm": 4.1184892654418945,
+      "learning_rate": 3.001984126984127e-05,
+      "loss": 0.6702,
+      "step": 1017
+    },
+    {
+      "epoch": 4.023715415019763,
+      "grad_norm": 4.948250770568848,
+      "learning_rate": 3e-05,
+      "loss": 0.7939,
+      "step": 1018
+    },
+    {
+      "epoch": 4.027667984189724,
+      "grad_norm": 4.541839122772217,
+      "learning_rate": 2.998015873015873e-05,
+      "loss": 0.5846,
+      "step": 1019
+    },
+    {
+      "epoch": 4.031620553359684,
+      "grad_norm": 6.365063190460205,
+      "learning_rate": 2.996031746031746e-05,
+      "loss": 1.0476,
+      "step": 1020
+    },
+    {
+      "epoch": 4.0355731225296445,
+      "grad_norm": 5.4133992195129395,
+      "learning_rate": 2.9940476190476192e-05,
+      "loss": 0.5169,
+      "step": 1021
+    },
+    {
+      "epoch": 4.0395256916996045,
+      "grad_norm": 5.79783821105957,
+      "learning_rate": 2.9920634920634922e-05,
+      "loss": 0.5218,
+      "step": 1022
+    },
+    {
+      "epoch": 4.043478260869565,
+      "grad_norm": 3.8281731605529785,
+      "learning_rate": 2.9900793650793653e-05,
+      "loss": 0.5262,
+      "step": 1023
+    },
+    {
+      "epoch": 4.047430830039525,
+      "grad_norm": 3.4406702518463135,
+      "learning_rate": 2.9880952380952383e-05,
+      "loss": 0.6195,
+      "step": 1024
+    },
+    {
+      "epoch": 4.051383399209486,
+      "grad_norm": 2.6785409450531006,
+      "learning_rate": 2.9861111111111113e-05,
+      "loss": 0.6092,
+      "step": 1025
+    },
+    {
+      "epoch": 4.055335968379446,
+      "grad_norm": 3.221266508102417,
+      "learning_rate": 2.9841269841269844e-05,
+      "loss": 0.5122,
+      "step": 1026
+    },
+    {
+      "epoch": 4.059288537549407,
+      "grad_norm": 3.454709053039551,
+      "learning_rate": 2.982142857142857e-05,
+      "loss": 0.364,
+      "step": 1027
+    },
+    {
+      "epoch": 4.063241106719367,
+      "grad_norm": 3.4353463649749756,
+      "learning_rate": 2.98015873015873e-05,
+      "loss": 0.4952,
+      "step": 1028
+    },
+    {
+      "epoch": 4.067193675889328,
+      "grad_norm": 3.565638303756714,
+      "learning_rate": 2.978174603174603e-05,
+      "loss": 0.7452,
+      "step": 1029
+    },
+    {
+      "epoch": 4.071146245059288,
+      "grad_norm": 4.197516918182373,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 0.5709,
+      "step": 1030
+    },
+    {
+      "epoch": 4.075098814229249,
+      "grad_norm": 3.723968744277954,
+      "learning_rate": 2.9742063492063492e-05,
+      "loss": 0.4972,
+      "step": 1031
+    },
+    {
+      "epoch": 4.07905138339921,
+      "grad_norm": 3.2937045097351074,
+      "learning_rate": 2.9722222222222223e-05,
+      "loss": 0.5963,
+      "step": 1032
+    },
+    {
+      "epoch": 4.08300395256917,
+      "grad_norm": 3.6638174057006836,
+      "learning_rate": 2.9702380952380953e-05,
+      "loss": 0.6339,
+      "step": 1033
+    },
+    {
+      "epoch": 4.086956521739131,
+      "grad_norm": 3.8052680492401123,
+      "learning_rate": 2.9682539682539683e-05,
+      "loss": 0.4921,
+      "step": 1034
+    },
+    {
+      "epoch": 4.090909090909091,
+      "grad_norm": 3.9429097175598145,
+      "learning_rate": 2.9662698412698414e-05,
+      "loss": 0.5685,
+      "step": 1035
+    },
+    {
+      "epoch": 4.094861660079052,
+      "grad_norm": 4.403009414672852,
+      "learning_rate": 2.9642857142857144e-05,
+      "loss": 0.8721,
+      "step": 1036
+    },
+    {
+      "epoch": 4.098814229249012,
+      "grad_norm": 3.999182939529419,
+      "learning_rate": 2.9623015873015875e-05,
+      "loss": 0.768,
+      "step": 1037
+    },
+    {
+      "epoch": 4.102766798418973,
+      "grad_norm": 3.5793051719665527,
+      "learning_rate": 2.9603174603174605e-05,
+      "loss": 0.465,
+      "step": 1038
+    },
+    {
+      "epoch": 4.106719367588933,
+      "grad_norm": 3.8434765338897705,
+      "learning_rate": 2.9583333333333335e-05,
+      "loss": 0.5206,
+      "step": 1039
+    },
+    {
+      "epoch": 4.1106719367588935,
+      "grad_norm": 4.839815139770508,
+      "learning_rate": 2.9563492063492066e-05,
+      "loss": 0.4424,
+      "step": 1040
+    },
+    {
+      "epoch": 4.1146245059288535,
+      "grad_norm": 4.106775283813477,
+      "learning_rate": 2.9543650793650796e-05,
+      "loss": 0.5181,
+      "step": 1041
+    },
+    {
+      "epoch": 4.118577075098814,
+      "grad_norm": 3.0763916969299316,
+      "learning_rate": 2.9523809523809526e-05,
+      "loss": 0.5855,
+      "step": 1042
+    },
+    {
+      "epoch": 4.122529644268774,
+      "grad_norm": 3.8926162719726562,
+      "learning_rate": 2.9503968253968257e-05,
+      "loss": 0.6347,
+      "step": 1043
+    },
+    {
+      "epoch": 4.126482213438735,
+      "grad_norm": 4.27288293838501,
+      "learning_rate": 2.9484126984126987e-05,
+      "loss": 0.4295,
+      "step": 1044
+    },
+    {
+      "epoch": 4.130434782608695,
+      "grad_norm": 2.7551262378692627,
+      "learning_rate": 2.9464285714285718e-05,
+      "loss": 0.733,
+      "step": 1045
+    },
+    {
+      "epoch": 4.134387351778656,
+      "grad_norm": 3.992636203765869,
+      "learning_rate": 2.9444444444444448e-05,
+      "loss": 0.5781,
+      "step": 1046
+    },
+    {
+      "epoch": 4.138339920948616,
+      "grad_norm": 5.2175469398498535,
+      "learning_rate": 2.9424603174603178e-05,
+      "loss": 0.5791,
+      "step": 1047
+    },
+    {
+      "epoch": 4.142292490118577,
+      "grad_norm": 3.3808295726776123,
+      "learning_rate": 2.940476190476191e-05,
+      "loss": 0.5001,
+      "step": 1048
+    },
+    {
+      "epoch": 4.146245059288537,
+      "grad_norm": 3.7828593254089355,
+      "learning_rate": 2.938492063492064e-05,
+      "loss": 0.5229,
+      "step": 1049
+    },
+    {
+      "epoch": 4.150197628458498,
+      "grad_norm": 3.9474239349365234,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.6843,
+      "step": 1050
+    },
+    {
+      "epoch": 4.154150197628459,
+      "grad_norm": 3.5939300060272217,
+      "learning_rate": 2.9345238095238096e-05,
+      "loss": 0.5496,
+      "step": 1051
+    },
+    {
+      "epoch": 4.158102766798419,
+      "grad_norm": 4.338398456573486,
+      "learning_rate": 2.9325396825396827e-05,
+      "loss": 0.5746,
+      "step": 1052
+    },
+    {
+      "epoch": 4.16205533596838,
+      "grad_norm": 3.5342512130737305,
+      "learning_rate": 2.9305555555555557e-05,
+      "loss": 0.5925,
+      "step": 1053
+    },
+    {
+      "epoch": 4.16600790513834,
+      "grad_norm": 3.970322608947754,
+      "learning_rate": 2.9285714285714288e-05,
+      "loss": 0.5174,
+      "step": 1054
+    },
+    {
+      "epoch": 4.169960474308301,
+      "grad_norm": 3.5085713863372803,
+      "learning_rate": 2.9265873015873018e-05,
+      "loss": 0.5177,
+      "step": 1055
+    },
+    {
+      "epoch": 4.173913043478261,
+      "grad_norm": 4.02853536605835,
+      "learning_rate": 2.9246031746031748e-05,
+      "loss": 0.58,
+      "step": 1056
+    },
+    {
+      "epoch": 4.177865612648222,
+      "grad_norm": 3.9853994846343994,
+      "learning_rate": 2.922619047619048e-05,
+      "loss": 0.8849,
+      "step": 1057
+    },
+    {
+      "epoch": 4.181818181818182,
+      "grad_norm": 3.8760087490081787,
+      "learning_rate": 2.920634920634921e-05,
+      "loss": 0.3916,
+      "step": 1058
+    },
+    {
+      "epoch": 4.1857707509881426,
+      "grad_norm": 4.411335468292236,
+      "learning_rate": 2.918650793650794e-05,
+      "loss": 0.5398,
+      "step": 1059
+    },
+    {
+      "epoch": 4.189723320158103,
+      "grad_norm": 2.8598296642303467,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.5319,
+      "step": 1060
+    },
+    {
+      "epoch": 4.1936758893280635,
+      "grad_norm": 3.225334405899048,
+      "learning_rate": 2.91468253968254e-05,
+      "loss": 0.6801,
+      "step": 1061
+    },
+    {
+      "epoch": 4.1976284584980235,
+      "grad_norm": 3.511744976043701,
+      "learning_rate": 2.912698412698413e-05,
+      "loss": 0.6467,
+      "step": 1062
+    },
+    {
+      "epoch": 4.201581027667984,
+      "grad_norm": 5.1000165939331055,
+      "learning_rate": 2.910714285714286e-05,
+      "loss": 0.6979,
+      "step": 1063
+    },
+    {
+      "epoch": 4.205533596837944,
+      "grad_norm": 4.431115627288818,
+      "learning_rate": 2.908730158730159e-05,
+      "loss": 0.7414,
+      "step": 1064
+    },
+    {
+      "epoch": 4.209486166007905,
+      "grad_norm": 3.7748608589172363,
+      "learning_rate": 2.906746031746032e-05,
+      "loss": 0.5364,
+      "step": 1065
+    },
+    {
+      "epoch": 4.213438735177865,
+      "grad_norm": 3.4133846759796143,
+      "learning_rate": 2.9047619047619052e-05,
+      "loss": 0.766,
+      "step": 1066
+    },
+    {
+      "epoch": 4.217391304347826,
+      "grad_norm": 4.544011116027832,
+      "learning_rate": 2.9027777777777782e-05,
+      "loss": 0.6298,
+      "step": 1067
+    },
+    {
+      "epoch": 4.221343873517786,
+      "grad_norm": 3.5001301765441895,
+      "learning_rate": 2.9007936507936513e-05,
+      "loss": 0.6486,
+      "step": 1068
+    },
+    {
+      "epoch": 4.225296442687747,
+      "grad_norm": 4.15241003036499,
+      "learning_rate": 2.8988095238095243e-05,
+      "loss": 0.6689,
+      "step": 1069
+    },
+    {
+      "epoch": 4.229249011857707,
+      "grad_norm": 4.282387733459473,
+      "learning_rate": 2.8968253968253974e-05,
+      "loss": 0.6515,
+      "step": 1070
+    },
+    {
+      "epoch": 4.233201581027668,
+      "grad_norm": 3.4778153896331787,
+      "learning_rate": 2.8948412698412704e-05,
+      "loss": 0.6742,
+      "step": 1071
+    },
+    {
+      "epoch": 4.237154150197629,
+      "grad_norm": 3.6013362407684326,
+      "learning_rate": 2.8928571428571434e-05,
+      "loss": 0.5932,
+      "step": 1072
+    },
+    {
+      "epoch": 4.241106719367589,
+      "grad_norm": 3.0561881065368652,
+      "learning_rate": 2.890873015873016e-05,
+      "loss": 0.5181,
+      "step": 1073
+    },
+    {
+      "epoch": 4.24505928853755,
+      "grad_norm": 3.436709403991699,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.6392,
+      "step": 1074
+    },
+    {
+      "epoch": 4.24901185770751,
+      "grad_norm": 3.798583984375,
+      "learning_rate": 2.886904761904762e-05,
+      "loss": 0.3719,
+      "step": 1075
+    },
+    {
+      "epoch": 4.252964426877471,
+      "grad_norm": 4.340543746948242,
+      "learning_rate": 2.884920634920635e-05,
+      "loss": 0.7213,
+      "step": 1076
+    },
+    {
+      "epoch": 4.256916996047431,
+      "grad_norm": 3.0846059322357178,
+      "learning_rate": 2.882936507936508e-05,
+      "loss": 0.6184,
+      "step": 1077
+    },
+    {
+      "epoch": 4.260869565217392,
+      "grad_norm": 3.127023696899414,
+      "learning_rate": 2.880952380952381e-05,
+      "loss": 0.8165,
+      "step": 1078
+    },
+    {
+      "epoch": 4.264822134387352,
+      "grad_norm": 4.435762405395508,
+      "learning_rate": 2.878968253968254e-05,
+      "loss": 0.8462,
+      "step": 1079
+    },
+    {
+      "epoch": 4.2687747035573125,
+      "grad_norm": 3.7268333435058594,
+      "learning_rate": 2.876984126984127e-05,
+      "loss": 0.5501,
+      "step": 1080
+    },
+    {
+      "epoch": 4.2727272727272725,
+      "grad_norm": 4.605154514312744,
+      "learning_rate": 2.8749999999999997e-05,
+      "loss": 0.4942,
+      "step": 1081
+    },
+    {
+      "epoch": 4.276679841897233,
+      "grad_norm": 4.297336101531982,
+      "learning_rate": 2.8730158730158728e-05,
+      "loss": 0.7144,
+      "step": 1082
+    },
+    {
+      "epoch": 4.280632411067193,
+      "grad_norm": 3.4597883224487305,
+      "learning_rate": 2.8710317460317458e-05,
+      "loss": 0.6571,
+      "step": 1083
+    },
+    {
+      "epoch": 4.284584980237154,
+      "grad_norm": 3.579210042953491,
+      "learning_rate": 2.869047619047619e-05,
+      "loss": 0.8167,
+      "step": 1084
+    },
+    {
+      "epoch": 4.288537549407114,
+      "grad_norm": 4.63372802734375,
+      "learning_rate": 2.867063492063492e-05,
+      "loss": 0.5739,
+      "step": 1085
+    },
+    {
+      "epoch": 4.292490118577075,
+      "grad_norm": 4.892091751098633,
+      "learning_rate": 2.865079365079365e-05,
+      "loss": 0.7786,
+      "step": 1086
+    },
+    {
+      "epoch": 4.296442687747035,
+      "grad_norm": 3.682457208633423,
+      "learning_rate": 2.863095238095238e-05,
+      "loss": 0.6753,
+      "step": 1087
+    },
+    {
+      "epoch": 4.300395256916996,
+      "grad_norm": 4.4599432945251465,
+      "learning_rate": 2.861111111111111e-05,
+      "loss": 0.5335,
+      "step": 1088
+    },
+    {
+      "epoch": 4.304347826086957,
+      "grad_norm": 4.139975070953369,
+      "learning_rate": 2.859126984126984e-05,
+      "loss": 0.631,
+      "step": 1089
+    },
+    {
+      "epoch": 4.308300395256917,
+      "grad_norm": 2.9012610912323,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.6686,
+      "step": 1090
+    },
+    {
+      "epoch": 4.312252964426877,
+      "grad_norm": 4.059675216674805,
+      "learning_rate": 2.85515873015873e-05,
+      "loss": 0.5449,
+      "step": 1091
+    },
+    {
+      "epoch": 4.316205533596838,
+      "grad_norm": 3.3060367107391357,
+      "learning_rate": 2.853174603174603e-05,
+      "loss": 0.5399,
+      "step": 1092
+    },
+    {
+      "epoch": 4.320158102766799,
+      "grad_norm": 4.22020149230957,
+      "learning_rate": 2.8511904761904762e-05,
+      "loss": 0.5664,
+      "step": 1093
+    },
+    {
+      "epoch": 4.324110671936759,
+      "grad_norm": 3.2028045654296875,
+      "learning_rate": 2.8492063492063492e-05,
+      "loss": 0.5289,
+      "step": 1094
+    },
+    {
+      "epoch": 4.32806324110672,
+      "grad_norm": 4.202157497406006,
+      "learning_rate": 2.8472222222222223e-05,
+      "loss": 0.6845,
+      "step": 1095
+    },
+    {
+      "epoch": 4.33201581027668,
+      "grad_norm": 3.3916430473327637,
+      "learning_rate": 2.8452380952380953e-05,
+      "loss": 0.7733,
+      "step": 1096
+    },
+    {
+      "epoch": 4.335968379446641,
+      "grad_norm": 3.6081862449645996,
+      "learning_rate": 2.8432539682539683e-05,
+      "loss": 0.7535,
+      "step": 1097
+    },
+    {
+      "epoch": 4.339920948616601,
+      "grad_norm": 3.1682872772216797,
+      "learning_rate": 2.8412698412698414e-05,
+      "loss": 0.5639,
+      "step": 1098
+    },
+    {
+      "epoch": 4.3438735177865615,
+      "grad_norm": 3.9137520790100098,
+      "learning_rate": 2.8392857142857144e-05,
+      "loss": 0.5085,
+      "step": 1099
+    },
+    {
+      "epoch": 4.3478260869565215,
+      "grad_norm": 3.902578353881836,
+      "learning_rate": 2.8373015873015875e-05,
+      "loss": 0.7883,
+      "step": 1100
+    },
+    {
+      "epoch": 4.351778656126482,
+      "grad_norm": 3.498415946960449,
+      "learning_rate": 2.8353174603174605e-05,
+      "loss": 0.7754,
+      "step": 1101
+    },
+    {
+      "epoch": 4.355731225296442,
+      "grad_norm": 3.8628177642822266,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.7349,
+      "step": 1102
+    },
+    {
+      "epoch": 4.359683794466403,
+      "grad_norm": 3.9302592277526855,
+      "learning_rate": 2.8313492063492066e-05,
+      "loss": 0.5988,
+      "step": 1103
+    },
+    {
+      "epoch": 4.363636363636363,
+      "grad_norm": 3.7080790996551514,
+      "learning_rate": 2.8293650793650793e-05,
+      "loss": 0.7116,
+      "step": 1104
+    },
+    {
+      "epoch": 4.367588932806324,
+      "grad_norm": 3.222238779067993,
+      "learning_rate": 2.8273809523809523e-05,
+      "loss": 0.4634,
+      "step": 1105
+    },
+    {
+      "epoch": 4.371541501976284,
+      "grad_norm": 3.111239194869995,
+      "learning_rate": 2.8253968253968253e-05,
+      "loss": 0.4384,
+      "step": 1106
+    },
+    {
+      "epoch": 4.375494071146245,
+      "grad_norm": 3.0986177921295166,
+      "learning_rate": 2.8234126984126984e-05,
+      "loss": 0.7854,
+      "step": 1107
+    },
+    {
+      "epoch": 4.379446640316205,
+      "grad_norm": 4.473976135253906,
+      "learning_rate": 2.8214285714285714e-05,
+      "loss": 0.6981,
+      "step": 1108
+    },
+    {
+      "epoch": 4.383399209486166,
+      "grad_norm": 3.3967599868774414,
+      "learning_rate": 2.8194444444444445e-05,
+      "loss": 0.7922,
+      "step": 1109
+    },
+    {
+      "epoch": 4.387351778656127,
+      "grad_norm": 5.092434406280518,
+      "learning_rate": 2.8174603174603175e-05,
+      "loss": 0.6217,
+      "step": 1110
+    },
+    {
+      "epoch": 4.391304347826087,
+      "grad_norm": 6.223644256591797,
+      "learning_rate": 2.8154761904761905e-05,
+      "loss": 0.6414,
+      "step": 1111
+    },
+    {
+      "epoch": 4.395256916996048,
+      "grad_norm": 4.24446439743042,
+      "learning_rate": 2.8134920634920636e-05,
+      "loss": 0.9485,
+      "step": 1112
+    },
+    {
+      "epoch": 4.399209486166008,
+      "grad_norm": 4.5990986824035645,
+      "learning_rate": 2.8115079365079366e-05,
+      "loss": 0.7239,
+      "step": 1113
+    },
+    {
+      "epoch": 4.403162055335969,
+      "grad_norm": 4.005640506744385,
+      "learning_rate": 2.8095238095238096e-05,
+      "loss": 0.9314,
+      "step": 1114
+    },
+    {
+      "epoch": 4.407114624505929,
+      "grad_norm": 4.990167140960693,
+      "learning_rate": 2.8075396825396827e-05,
+      "loss": 0.6125,
+      "step": 1115
+    },
+    {
+      "epoch": 4.41106719367589,
+      "grad_norm": 4.539425849914551,
+      "learning_rate": 2.8055555555555557e-05,
+      "loss": 0.4673,
+      "step": 1116
+    },
+    {
+      "epoch": 4.41501976284585,
+      "grad_norm": 5.004159927368164,
+      "learning_rate": 2.8035714285714288e-05,
+      "loss": 0.6748,
+      "step": 1117
+    },
+    {
+      "epoch": 4.4189723320158105,
+      "grad_norm": 5.46227502822876,
+      "learning_rate": 2.8015873015873018e-05,
+      "loss": 0.6671,
+      "step": 1118
+    },
+    {
+      "epoch": 4.4229249011857705,
+      "grad_norm": 3.3494784832000732,
+      "learning_rate": 2.799603174603175e-05,
+      "loss": 0.5736,
+      "step": 1119
+    },
+    {
+      "epoch": 4.426877470355731,
+      "grad_norm": 4.099065780639648,
+      "learning_rate": 2.797619047619048e-05,
+      "loss": 0.6317,
+      "step": 1120
+    },
+    {
+      "epoch": 4.430830039525691,
+      "grad_norm": 3.662707805633545,
+      "learning_rate": 2.795634920634921e-05,
+      "loss": 0.6749,
+      "step": 1121
+    },
+    {
+      "epoch": 4.434782608695652,
+      "grad_norm": 3.4078757762908936,
+      "learning_rate": 2.793650793650794e-05,
+      "loss": 0.723,
+      "step": 1122
+    },
+    {
+      "epoch": 4.438735177865612,
+      "grad_norm": 3.4741852283477783,
+      "learning_rate": 2.791666666666667e-05,
+      "loss": 0.7619,
+      "step": 1123
+    },
+    {
+      "epoch": 4.442687747035573,
+      "grad_norm": 4.361887454986572,
+      "learning_rate": 2.78968253968254e-05,
+      "loss": 0.7173,
+      "step": 1124
+    },
+    {
+      "epoch": 4.446640316205533,
+      "grad_norm": 3.31022047996521,
+      "learning_rate": 2.787698412698413e-05,
+      "loss": 0.4076,
+      "step": 1125
+    },
+    {
+      "epoch": 4.450592885375494,
+      "grad_norm": 3.635115623474121,
+      "learning_rate": 2.785714285714286e-05,
+      "loss": 0.5728,
+      "step": 1126
+    },
+    {
+      "epoch": 4.454545454545454,
+      "grad_norm": 5.300922870635986,
+      "learning_rate": 2.7837301587301588e-05,
+      "loss": 0.8788,
+      "step": 1127
+    },
+    {
+      "epoch": 4.458498023715415,
+      "grad_norm": 4.8898491859436035,
+      "learning_rate": 2.781746031746032e-05,
+      "loss": 0.6955,
+      "step": 1128
+    },
+    {
+      "epoch": 4.462450592885375,
+      "grad_norm": 3.6756420135498047,
+      "learning_rate": 2.779761904761905e-05,
+      "loss": 0.6531,
+      "step": 1129
+    },
+    {
+      "epoch": 4.466403162055336,
+      "grad_norm": 4.139333724975586,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.4227,
+      "step": 1130
+    },
+    {
+      "epoch": 4.470355731225297,
+      "grad_norm": 4.150503158569336,
+      "learning_rate": 2.775793650793651e-05,
+      "loss": 0.5596,
+      "step": 1131
+    },
+    {
+      "epoch": 4.474308300395257,
+      "grad_norm": 3.9440436363220215,
+      "learning_rate": 2.773809523809524e-05,
+      "loss": 0.5473,
+      "step": 1132
+    },
+    {
+      "epoch": 4.478260869565218,
+      "grad_norm": 4.698122978210449,
+      "learning_rate": 2.771825396825397e-05,
+      "loss": 0.5331,
+      "step": 1133
+    },
+    {
+      "epoch": 4.482213438735178,
+      "grad_norm": 4.4642133712768555,
+      "learning_rate": 2.76984126984127e-05,
+      "loss": 0.5267,
+      "step": 1134
+    },
+    {
+      "epoch": 4.486166007905139,
+      "grad_norm": 3.5104897022247314,
+      "learning_rate": 2.767857142857143e-05,
+      "loss": 0.7138,
+      "step": 1135
+    },
+    {
+      "epoch": 4.490118577075099,
+      "grad_norm": 4.170843601226807,
+      "learning_rate": 2.765873015873016e-05,
+      "loss": 0.8175,
+      "step": 1136
+    },
+    {
+      "epoch": 4.4940711462450595,
+      "grad_norm": 3.3033299446105957,
+      "learning_rate": 2.7638888888888892e-05,
+      "loss": 0.5711,
+      "step": 1137
+    },
+    {
+      "epoch": 4.4980237154150196,
+      "grad_norm": 4.042965412139893,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.7944,
+      "step": 1138
+    },
+    {
+      "epoch": 4.5019762845849804,
+      "grad_norm": 3.890293836593628,
+      "learning_rate": 2.7599206349206352e-05,
+      "loss": 0.7309,
+      "step": 1139
+    },
+    {
+      "epoch": 4.5059288537549405,
+      "grad_norm": 3.5573506355285645,
+      "learning_rate": 2.7579365079365083e-05,
+      "loss": 0.2866,
+      "step": 1140
+    },
+    {
+      "epoch": 4.509881422924901,
+      "grad_norm": 3.3874566555023193,
+      "learning_rate": 2.7559523809523813e-05,
+      "loss": 0.5363,
+      "step": 1141
+    },
+    {
+      "epoch": 4.513833992094861,
+      "grad_norm": 3.777653217315674,
+      "learning_rate": 2.7539682539682544e-05,
+      "loss": 0.8195,
+      "step": 1142
+    },
+    {
+      "epoch": 4.517786561264822,
+      "grad_norm": 3.0119121074676514,
+      "learning_rate": 2.7519841269841274e-05,
+      "loss": 0.4722,
+      "step": 1143
+    },
+    {
+      "epoch": 4.521739130434782,
+      "grad_norm": 3.57424259185791,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.6739,
+      "step": 1144
+    },
+    {
+      "epoch": 4.525691699604743,
+      "grad_norm": 4.8777174949646,
+      "learning_rate": 2.7480158730158735e-05,
+      "loss": 0.5557,
+      "step": 1145
+    },
+    {
+      "epoch": 4.529644268774703,
+      "grad_norm": 5.821610927581787,
+      "learning_rate": 2.7460317460317465e-05,
+      "loss": 0.7279,
+      "step": 1146
+    },
+    {
+      "epoch": 4.533596837944664,
+      "grad_norm": 3.617403268814087,
+      "learning_rate": 2.7440476190476195e-05,
+      "loss": 0.7618,
+      "step": 1147
+    },
+    {
+      "epoch": 4.537549407114625,
+      "grad_norm": 4.069850921630859,
+      "learning_rate": 2.7420634920634926e-05,
+      "loss": 0.6691,
+      "step": 1148
+    },
+    {
+      "epoch": 4.541501976284585,
+      "grad_norm": 3.8116888999938965,
+      "learning_rate": 2.7400793650793656e-05,
+      "loss": 0.6847,
+      "step": 1149
+    },
+    {
+      "epoch": 4.545454545454545,
+      "grad_norm": 4.080233573913574,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 0.7181,
+      "step": 1150
+    },
+    {
+      "epoch": 4.549407114624506,
+      "grad_norm": 3.930443048477173,
+      "learning_rate": 2.7361111111111114e-05,
+      "loss": 0.4972,
+      "step": 1151
+    },
+    {
+      "epoch": 4.553359683794467,
+      "grad_norm": 3.8190906047821045,
+      "learning_rate": 2.734126984126984e-05,
+      "loss": 0.7407,
+      "step": 1152
+    },
+    {
+      "epoch": 4.557312252964427,
+      "grad_norm": 4.792582988739014,
+      "learning_rate": 2.732142857142857e-05,
+      "loss": 0.5122,
+      "step": 1153
+    },
+    {
+      "epoch": 4.561264822134388,
+      "grad_norm": 3.7988221645355225,
+      "learning_rate": 2.73015873015873e-05,
+      "loss": 0.7225,
+      "step": 1154
+    },
+    {
+      "epoch": 4.565217391304348,
+      "grad_norm": 4.3564605712890625,
+      "learning_rate": 2.7281746031746032e-05,
+      "loss": 0.5053,
+      "step": 1155
+    },
+    {
+      "epoch": 4.569169960474309,
+      "grad_norm": 3.3426055908203125,
+      "learning_rate": 2.7261904761904762e-05,
+      "loss": 0.5994,
+      "step": 1156
+    },
+    {
+      "epoch": 4.573122529644269,
+      "grad_norm": 3.6366496086120605,
+      "learning_rate": 2.7242063492063492e-05,
+      "loss": 0.7183,
+      "step": 1157
+    },
+    {
+      "epoch": 4.5770750988142295,
+      "grad_norm": 4.141812801361084,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 0.8983,
+      "step": 1158
+    },
+    {
+      "epoch": 4.5810276679841895,
+      "grad_norm": 3.6468398571014404,
+      "learning_rate": 2.720238095238095e-05,
+      "loss": 0.7188,
+      "step": 1159
+    },
+    {
+      "epoch": 4.58498023715415,
+      "grad_norm": 3.9180054664611816,
+      "learning_rate": 2.718253968253968e-05,
+      "loss": 0.5211,
+      "step": 1160
+    },
+    {
+      "epoch": 4.58893280632411,
+      "grad_norm": 3.7029590606689453,
+      "learning_rate": 2.716269841269841e-05,
+      "loss": 0.6044,
+      "step": 1161
+    },
+    {
+      "epoch": 4.592885375494071,
+      "grad_norm": 4.28466272354126,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.6413,
+      "step": 1162
+    },
+    {
+      "epoch": 4.596837944664031,
+      "grad_norm": 5.500331401824951,
+      "learning_rate": 2.712301587301587e-05,
+      "loss": 0.7009,
+      "step": 1163
+    },
+    {
+      "epoch": 4.600790513833992,
+      "grad_norm": 4.083467960357666,
+      "learning_rate": 2.7103174603174602e-05,
+      "loss": 0.822,
+      "step": 1164
+    },
+    {
+      "epoch": 4.604743083003952,
+      "grad_norm": 2.7674543857574463,
+      "learning_rate": 2.7083333333333332e-05,
+      "loss": 0.6164,
+      "step": 1165
+    },
+    {
+      "epoch": 4.608695652173913,
+      "grad_norm": 3.0514588356018066,
+      "learning_rate": 2.7063492063492062e-05,
+      "loss": 0.4806,
+      "step": 1166
+    },
+    {
+      "epoch": 4.612648221343873,
+      "grad_norm": 4.635437965393066,
+      "learning_rate": 2.7043650793650793e-05,
+      "loss": 0.5268,
+      "step": 1167
+    },
+    {
+      "epoch": 4.616600790513834,
+      "grad_norm": 3.791935443878174,
+      "learning_rate": 2.7023809523809523e-05,
+      "loss": 0.6394,
+      "step": 1168
+    },
+    {
+      "epoch": 4.620553359683795,
+      "grad_norm": 3.0412096977233887,
+      "learning_rate": 2.7003968253968254e-05,
+      "loss": 0.5388,
+      "step": 1169
+    },
+    {
+      "epoch": 4.624505928853755,
+      "grad_norm": 3.3811123371124268,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.5911,
+      "step": 1170
+    },
+    {
+      "epoch": 4.628458498023716,
+      "grad_norm": 3.5328590869903564,
+      "learning_rate": 2.6964285714285714e-05,
+      "loss": 0.5826,
+      "step": 1171
+    },
+    {
+      "epoch": 4.632411067193676,
+      "grad_norm": 3.483593702316284,
+      "learning_rate": 2.6944444444444445e-05,
+      "loss": 0.5478,
+      "step": 1172
+    },
+    {
+      "epoch": 4.636363636363637,
+      "grad_norm": 4.339010238647461,
+      "learning_rate": 2.6924603174603175e-05,
+      "loss": 0.5807,
+      "step": 1173
+    },
+    {
+      "epoch": 4.640316205533597,
+      "grad_norm": 3.775643825531006,
+      "learning_rate": 2.6904761904761905e-05,
+      "loss": 0.7021,
+      "step": 1174
+    },
+    {
+      "epoch": 4.644268774703558,
+      "grad_norm": 4.458033561706543,
+      "learning_rate": 2.6884920634920636e-05,
+      "loss": 0.6987,
+      "step": 1175
+    },
+    {
+      "epoch": 4.648221343873518,
+      "grad_norm": 4.082224369049072,
+      "learning_rate": 2.6865079365079366e-05,
+      "loss": 0.5305,
+      "step": 1176
+    },
+    {
+      "epoch": 4.6521739130434785,
+      "grad_norm": 4.834577560424805,
+      "learning_rate": 2.6845238095238097e-05,
+      "loss": 0.5664,
+      "step": 1177
+    },
+    {
+      "epoch": 4.6561264822134385,
+      "grad_norm": 4.269159317016602,
+      "learning_rate": 2.6825396825396827e-05,
+      "loss": 0.5704,
+      "step": 1178
+    },
+    {
+      "epoch": 4.660079051383399,
+      "grad_norm": 3.2594568729400635,
+      "learning_rate": 2.6805555555555557e-05,
+      "loss": 0.5927,
+      "step": 1179
+    },
+    {
+      "epoch": 4.664031620553359,
+      "grad_norm": 3.458214521408081,
+      "learning_rate": 2.6785714285714288e-05,
+      "loss": 0.5754,
+      "step": 1180
+    },
+    {
+      "epoch": 4.66798418972332,
+      "grad_norm": 4.189966201782227,
+      "learning_rate": 2.6765873015873018e-05,
+      "loss": 0.8723,
+      "step": 1181
+    },
+    {
+      "epoch": 4.67193675889328,
+      "grad_norm": 3.8360981941223145,
+      "learning_rate": 2.6746031746031745e-05,
+      "loss": 0.4775,
+      "step": 1182
+    },
+    {
+      "epoch": 4.675889328063241,
+      "grad_norm": 4.381925582885742,
+      "learning_rate": 2.6726190476190475e-05,
+      "loss": 0.607,
+      "step": 1183
+    },
+    {
+      "epoch": 4.679841897233201,
+      "grad_norm": 4.080252647399902,
+      "learning_rate": 2.6706349206349206e-05,
+      "loss": 0.8249,
+      "step": 1184
+    },
+    {
+      "epoch": 4.683794466403162,
+      "grad_norm": 3.6359405517578125,
+      "learning_rate": 2.6686507936507936e-05,
+      "loss": 0.3814,
+      "step": 1185
+    },
+    {
+      "epoch": 4.687747035573123,
+      "grad_norm": 3.182612895965576,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5989,
+      "step": 1186
+    },
+    {
+      "epoch": 4.691699604743083,
+      "grad_norm": 4.81362247467041,
+      "learning_rate": 2.6646825396825397e-05,
+      "loss": 0.7298,
+      "step": 1187
+    },
+    {
+      "epoch": 4.695652173913043,
+      "grad_norm": 3.2109711170196533,
+      "learning_rate": 2.6626984126984127e-05,
+      "loss": 0.3586,
+      "step": 1188
+    },
+    {
+      "epoch": 4.699604743083004,
+      "grad_norm": 3.968430280685425,
+      "learning_rate": 2.6607142857142858e-05,
+      "loss": 0.5076,
+      "step": 1189
+    },
+    {
+      "epoch": 4.703557312252965,
+      "grad_norm": 3.057274341583252,
+      "learning_rate": 2.6587301587301588e-05,
+      "loss": 0.4669,
+      "step": 1190
+    },
+    {
+      "epoch": 4.707509881422925,
+      "grad_norm": 4.017573356628418,
+      "learning_rate": 2.656746031746032e-05,
+      "loss": 0.6747,
+      "step": 1191
+    },
+    {
+      "epoch": 4.711462450592886,
+      "grad_norm": 3.6146085262298584,
+      "learning_rate": 2.654761904761905e-05,
+      "loss": 0.4618,
+      "step": 1192
+    },
+    {
+      "epoch": 4.715415019762846,
+      "grad_norm": 3.433858871459961,
+      "learning_rate": 2.652777777777778e-05,
+      "loss": 0.4446,
+      "step": 1193
+    },
+    {
+      "epoch": 4.719367588932807,
+      "grad_norm": 3.7666232585906982,
+      "learning_rate": 2.650793650793651e-05,
+      "loss": 0.5724,
+      "step": 1194
+    },
+    {
+      "epoch": 4.723320158102767,
+      "grad_norm": 3.7725718021392822,
+      "learning_rate": 2.648809523809524e-05,
+      "loss": 0.7872,
+      "step": 1195
+    },
+    {
+      "epoch": 4.7272727272727275,
+      "grad_norm": 4.2439446449279785,
+      "learning_rate": 2.646825396825397e-05,
+      "loss": 0.7158,
+      "step": 1196
+    },
+    {
+      "epoch": 4.7312252964426875,
+      "grad_norm": 5.448159694671631,
+      "learning_rate": 2.64484126984127e-05,
+      "loss": 0.4721,
+      "step": 1197
+    },
+    {
+      "epoch": 4.735177865612648,
+      "grad_norm": 4.358859062194824,
+      "learning_rate": 2.642857142857143e-05,
+      "loss": 0.6254,
+      "step": 1198
+    },
+    {
+      "epoch": 4.739130434782608,
+      "grad_norm": 3.7496564388275146,
+      "learning_rate": 2.640873015873016e-05,
+      "loss": 0.7183,
+      "step": 1199
+    },
+    {
+      "epoch": 4.743083003952569,
+      "grad_norm": 4.28209114074707,
+      "learning_rate": 2.6388888888888892e-05,
+      "loss": 0.5722,
+      "step": 1200
+    },
+    {
+      "epoch": 4.747035573122529,
+      "grad_norm": 3.856718063354492,
+      "learning_rate": 2.6369047619047622e-05,
+      "loss": 0.5289,
+      "step": 1201
+    },
+    {
+      "epoch": 4.75098814229249,
+      "grad_norm": 3.4435012340545654,
+      "learning_rate": 2.6349206349206353e-05,
+      "loss": 0.7542,
+      "step": 1202
+    },
+    {
+      "epoch": 4.75494071146245,
+      "grad_norm": 4.4995436668396,
+      "learning_rate": 2.6329365079365083e-05,
+      "loss": 0.7093,
+      "step": 1203
+    },
+    {
+      "epoch": 4.758893280632411,
+      "grad_norm": 3.929421901702881,
+      "learning_rate": 2.6309523809523813e-05,
+      "loss": 0.8889,
+      "step": 1204
+    },
+    {
+      "epoch": 4.762845849802371,
+      "grad_norm": 3.778069496154785,
+      "learning_rate": 2.628968253968254e-05,
+      "loss": 0.766,
+      "step": 1205
+    },
+    {
+      "epoch": 4.766798418972332,
+      "grad_norm": 3.344264030456543,
+      "learning_rate": 2.626984126984127e-05,
+      "loss": 0.5133,
+      "step": 1206
+    },
+    {
+      "epoch": 4.770750988142293,
+      "grad_norm": 3.597881317138672,
+      "learning_rate": 2.625e-05,
+      "loss": 0.7008,
+      "step": 1207
+    },
+    {
+      "epoch": 4.774703557312253,
+      "grad_norm": 3.753389358520508,
+      "learning_rate": 2.623015873015873e-05,
+      "loss": 0.5307,
+      "step": 1208
+    },
+    {
+      "epoch": 4.778656126482213,
+      "grad_norm": 3.206299066543579,
+      "learning_rate": 2.6210317460317462e-05,
+      "loss": 0.5639,
+      "step": 1209
+    },
+    {
+      "epoch": 4.782608695652174,
+      "grad_norm": 3.630187749862671,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.4498,
+      "step": 1210
+    },
+    {
+      "epoch": 4.786561264822135,
+      "grad_norm": 3.9658334255218506,
+      "learning_rate": 2.6170634920634923e-05,
+      "loss": 0.6379,
+      "step": 1211
+    },
+    {
+      "epoch": 4.790513833992095,
+      "grad_norm": 4.101767539978027,
+      "learning_rate": 2.6150793650793653e-05,
+      "loss": 0.6574,
+      "step": 1212
+    },
+    {
+      "epoch": 4.794466403162056,
+      "grad_norm": 3.9470412731170654,
+      "learning_rate": 2.6130952380952383e-05,
+      "loss": 0.5475,
+      "step": 1213
+    },
+    {
+      "epoch": 4.798418972332016,
+      "grad_norm": 4.430981159210205,
+      "learning_rate": 2.6111111111111114e-05,
+      "loss": 0.4462,
+      "step": 1214
+    },
+    {
+      "epoch": 4.8023715415019765,
+      "grad_norm": 5.369266033172607,
+      "learning_rate": 2.6091269841269844e-05,
+      "loss": 0.7793,
+      "step": 1215
+    },
+    {
+      "epoch": 4.8063241106719365,
+      "grad_norm": 3.2877326011657715,
+      "learning_rate": 2.6071428571428574e-05,
+      "loss": 0.5941,
+      "step": 1216
+    },
+    {
+      "epoch": 4.810276679841897,
+      "grad_norm": 4.576911449432373,
+      "learning_rate": 2.6051587301587305e-05,
+      "loss": 0.582,
+      "step": 1217
+    },
+    {
+      "epoch": 4.8142292490118574,
+      "grad_norm": 3.5742104053497314,
+      "learning_rate": 2.6031746031746035e-05,
+      "loss": 0.6526,
+      "step": 1218
+    },
+    {
+      "epoch": 4.818181818181818,
+      "grad_norm": 2.9173500537872314,
+      "learning_rate": 2.6011904761904766e-05,
+      "loss": 0.4235,
+      "step": 1219
+    },
+    {
+      "epoch": 4.822134387351778,
+      "grad_norm": 3.32147479057312,
+      "learning_rate": 2.5992063492063496e-05,
+      "loss": 0.5129,
+      "step": 1220
+    },
+    {
+      "epoch": 4.826086956521739,
+      "grad_norm": 3.451444387435913,
+      "learning_rate": 2.5972222222222226e-05,
+      "loss": 0.5899,
+      "step": 1221
+    },
+    {
+      "epoch": 4.830039525691699,
+      "grad_norm": 4.481076717376709,
+      "learning_rate": 2.5952380952380957e-05,
+      "loss": 0.7092,
+      "step": 1222
+    },
+    {
+      "epoch": 4.83399209486166,
+      "grad_norm": 5.769997596740723,
+      "learning_rate": 2.5932539682539687e-05,
+      "loss": 0.8478,
+      "step": 1223
+    },
+    {
+      "epoch": 4.837944664031621,
+      "grad_norm": 4.319329738616943,
+      "learning_rate": 2.5912698412698417e-05,
+      "loss": 0.7591,
+      "step": 1224
+    },
+    {
+      "epoch": 4.841897233201581,
+      "grad_norm": 3.661302328109741,
+      "learning_rate": 2.5892857142857148e-05,
+      "loss": 0.6261,
+      "step": 1225
+    },
+    {
+      "epoch": 4.845849802371541,
+      "grad_norm": 3.7250123023986816,
+      "learning_rate": 2.5873015873015878e-05,
+      "loss": 0.8497,
+      "step": 1226
+    },
+    {
+      "epoch": 4.849802371541502,
+      "grad_norm": 4.192583084106445,
+      "learning_rate": 2.585317460317461e-05,
+      "loss": 0.6283,
+      "step": 1227
+    },
+    {
+      "epoch": 4.853754940711463,
+      "grad_norm": 3.978309154510498,
+      "learning_rate": 2.5833333333333336e-05,
+      "loss": 0.7002,
+      "step": 1228
+    },
+    {
+      "epoch": 4.857707509881423,
+      "grad_norm": 3.473998546600342,
+      "learning_rate": 2.5813492063492066e-05,
+      "loss": 0.6287,
+      "step": 1229
+    },
+    {
+      "epoch": 4.861660079051384,
+      "grad_norm": 3.9068286418914795,
+      "learning_rate": 2.5793650793650796e-05,
+      "loss": 0.4956,
+      "step": 1230
+    },
+    {
+      "epoch": 4.865612648221344,
+      "grad_norm": 2.8710238933563232,
+      "learning_rate": 2.5773809523809523e-05,
+      "loss": 0.6042,
+      "step": 1231
+    },
+    {
+      "epoch": 4.869565217391305,
+      "grad_norm": 3.4524970054626465,
+      "learning_rate": 2.5753968253968254e-05,
+      "loss": 0.4019,
+      "step": 1232
+    },
+    {
+      "epoch": 4.873517786561265,
+      "grad_norm": 3.4277803897857666,
+      "learning_rate": 2.5734126984126984e-05,
+      "loss": 0.4765,
+      "step": 1233
+    },
+    {
+      "epoch": 4.877470355731226,
+      "grad_norm": 3.466019868850708,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.6861,
+      "step": 1234
+    },
+    {
+      "epoch": 4.881422924901186,
+      "grad_norm": 4.309812068939209,
+      "learning_rate": 2.5694444444444445e-05,
+      "loss": 0.5055,
+      "step": 1235
+    },
+    {
+      "epoch": 4.8853754940711465,
+      "grad_norm": 3.608254909515381,
+      "learning_rate": 2.5674603174603172e-05,
+      "loss": 0.5123,
+      "step": 1236
+    },
+    {
+      "epoch": 4.8893280632411065,
+      "grad_norm": 4.4674787521362305,
+      "learning_rate": 2.5654761904761902e-05,
+      "loss": 0.8411,
+      "step": 1237
+    },
+    {
+      "epoch": 4.893280632411067,
+      "grad_norm": 4.53634786605835,
+      "learning_rate": 2.5634920634920633e-05,
+      "loss": 0.5074,
+      "step": 1238
+    },
+    {
+      "epoch": 4.897233201581027,
+      "grad_norm": 3.6150360107421875,
+      "learning_rate": 2.5615079365079363e-05,
+      "loss": 0.7875,
+      "step": 1239
+    },
+    {
+      "epoch": 4.901185770750988,
+      "grad_norm": 4.096851348876953,
+      "learning_rate": 2.5595238095238093e-05,
+      "loss": 0.6804,
+      "step": 1240
+    },
+    {
+      "epoch": 4.905138339920948,
+      "grad_norm": 3.8696417808532715,
+      "learning_rate": 2.5575396825396824e-05,
+      "loss": 0.6159,
+      "step": 1241
+    },
+    {
+      "epoch": 4.909090909090909,
+      "grad_norm": 3.823349714279175,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.6646,
+      "step": 1242
+    },
+    {
+      "epoch": 4.913043478260869,
+      "grad_norm": 4.550106048583984,
+      "learning_rate": 2.5535714285714284e-05,
+      "loss": 0.6771,
+      "step": 1243
+    },
+    {
+      "epoch": 4.91699604743083,
+      "grad_norm": 4.155416011810303,
+      "learning_rate": 2.5515873015873015e-05,
+      "loss": 0.6493,
+      "step": 1244
+    },
+    {
+      "epoch": 4.920948616600791,
+      "grad_norm": 3.774624824523926,
+      "learning_rate": 2.5496031746031745e-05,
+      "loss": 0.6771,
+      "step": 1245
+    },
+    {
+      "epoch": 4.924901185770751,
+      "grad_norm": 3.5552561283111572,
+      "learning_rate": 2.5476190476190476e-05,
+      "loss": 0.6626,
+      "step": 1246
+    },
+    {
+      "epoch": 4.928853754940711,
+      "grad_norm": 3.7037765979766846,
+      "learning_rate": 2.5456349206349206e-05,
+      "loss": 0.5427,
+      "step": 1247
+    },
+    {
+      "epoch": 4.932806324110672,
+      "grad_norm": 4.090572357177734,
+      "learning_rate": 2.5436507936507936e-05,
+      "loss": 0.7206,
+      "step": 1248
+    },
+    {
+      "epoch": 4.936758893280633,
+      "grad_norm": 3.817936420440674,
+      "learning_rate": 2.5416666666666667e-05,
+      "loss": 0.4127,
+      "step": 1249
+    },
+    {
+      "epoch": 4.940711462450593,
+      "grad_norm": 3.522655487060547,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.6837,
+      "step": 1250
+    },
+    {
+      "epoch": 4.944664031620554,
+      "grad_norm": 3.069960832595825,
+      "learning_rate": 2.5376984126984127e-05,
+      "loss": 0.5466,
+      "step": 1251
+    },
+    {
+      "epoch": 4.948616600790514,
+      "grad_norm": 2.7633731365203857,
+      "learning_rate": 2.5357142857142858e-05,
+      "loss": 0.3194,
+      "step": 1252
+    },
+    {
+      "epoch": 4.952569169960475,
+      "grad_norm": 4.602902889251709,
+      "learning_rate": 2.5337301587301588e-05,
+      "loss": 0.5539,
+      "step": 1253
+    },
+    {
+      "epoch": 4.956521739130435,
+      "grad_norm": 4.454743385314941,
+      "learning_rate": 2.531746031746032e-05,
+      "loss": 0.5853,
+      "step": 1254
+    },
+    {
+      "epoch": 4.9604743083003955,
+      "grad_norm": 3.8360097408294678,
+      "learning_rate": 2.529761904761905e-05,
+      "loss": 0.6157,
+      "step": 1255
+    },
+    {
+      "epoch": 4.9644268774703555,
+      "grad_norm": 3.1883623600006104,
+      "learning_rate": 2.527777777777778e-05,
+      "loss": 0.6483,
+      "step": 1256
+    },
+    {
+      "epoch": 4.968379446640316,
+      "grad_norm": 3.8525397777557373,
+      "learning_rate": 2.525793650793651e-05,
+      "loss": 0.4996,
+      "step": 1257
+    },
+    {
+      "epoch": 4.972332015810276,
+      "grad_norm": 3.8935108184814453,
+      "learning_rate": 2.523809523809524e-05,
+      "loss": 0.5236,
+      "step": 1258
+    },
+    {
+      "epoch": 4.976284584980237,
+      "grad_norm": 3.436164140701294,
+      "learning_rate": 2.5218253968253967e-05,
+      "loss": 0.6688,
+      "step": 1259
+    },
+    {
+      "epoch": 4.980237154150197,
+      "grad_norm": 3.803886651992798,
+      "learning_rate": 2.5198412698412697e-05,
+      "loss": 0.4893,
+      "step": 1260
+    },
+    {
+      "epoch": 4.984189723320158,
+      "grad_norm": 4.310186862945557,
+      "learning_rate": 2.5178571428571428e-05,
+      "loss": 0.3476,
+      "step": 1261
+    },
+    {
+      "epoch": 4.988142292490118,
+      "grad_norm": 4.136441707611084,
+      "learning_rate": 2.5158730158730158e-05,
+      "loss": 0.6233,
+      "step": 1262
+    },
+    {
+      "epoch": 4.992094861660079,
+      "grad_norm": 3.9039924144744873,
+      "learning_rate": 2.513888888888889e-05,
+      "loss": 0.6625,
+      "step": 1263
+    },
+    {
+      "epoch": 4.996047430830039,
+      "grad_norm": 5.175858497619629,
+      "learning_rate": 2.511904761904762e-05,
+      "loss": 0.5538,
+      "step": 1264
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 4.218173503875732,
+      "learning_rate": 2.509920634920635e-05,
+      "loss": 0.7123,
+      "step": 1265
+    },
+    {
+      "epoch": 5.003952569169961,
+      "grad_norm": 3.1178061962127686,
+      "learning_rate": 2.507936507936508e-05,
+      "loss": 0.4178,
+      "step": 1266
+    },
+    {
+      "epoch": 5.007905138339921,
+      "grad_norm": 2.8245718479156494,
+      "learning_rate": 2.505952380952381e-05,
+      "loss": 0.2365,
+      "step": 1267
+    },
+    {
+      "epoch": 5.011857707509882,
+      "grad_norm": 2.6951630115509033,
+      "learning_rate": 2.503968253968254e-05,
+      "loss": 0.4007,
+      "step": 1268
+    },
+    {
+      "epoch": 5.015810276679842,
+      "grad_norm": 2.6765530109405518,
+      "learning_rate": 2.501984126984127e-05,
+      "loss": 0.2377,
+      "step": 1269
+    },
+    {
+      "epoch": 5.019762845849803,
+      "grad_norm": 4.122332572937012,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3555,
+      "step": 1270
+    },
+    {
+      "epoch": 5.023715415019763,
+      "grad_norm": 4.3077712059021,
+      "learning_rate": 2.498015873015873e-05,
+      "loss": 0.2416,
+      "step": 1271
+    },
+    {
+      "epoch": 5.027667984189724,
+      "grad_norm": 4.278382778167725,
+      "learning_rate": 2.4960317460317462e-05,
+      "loss": 0.2413,
+      "step": 1272
+    },
+    {
+      "epoch": 5.031620553359684,
+      "grad_norm": 5.61036491394043,
+      "learning_rate": 2.4940476190476192e-05,
+      "loss": 0.4181,
+      "step": 1273
+    },
+    {
+      "epoch": 5.0355731225296445,
+      "grad_norm": 5.716897487640381,
+      "learning_rate": 2.4920634920634923e-05,
+      "loss": 0.4539,
+      "step": 1274
+    },
+    {
+      "epoch": 5.0395256916996045,
+      "grad_norm": 4.314607620239258,
+      "learning_rate": 2.4900793650793653e-05,
+      "loss": 0.3872,
+      "step": 1275
+    },
+    {
+      "epoch": 5.043478260869565,
+      "grad_norm": 4.847557067871094,
+      "learning_rate": 2.4880952380952383e-05,
+      "loss": 0.4242,
+      "step": 1276
+    },
+    {
+      "epoch": 5.047430830039525,
+      "grad_norm": 4.569025039672852,
+      "learning_rate": 2.4861111111111114e-05,
+      "loss": 0.3746,
+      "step": 1277
+    },
+    {
+      "epoch": 5.051383399209486,
+      "grad_norm": 3.309478282928467,
+      "learning_rate": 2.4841269841269844e-05,
+      "loss": 0.263,
+      "step": 1278
+    },
+    {
+      "epoch": 5.055335968379446,
+      "grad_norm": 3.7712574005126953,
+      "learning_rate": 2.4821428571428575e-05,
+      "loss": 0.3596,
+      "step": 1279
+    },
+    {
+      "epoch": 5.059288537549407,
+      "grad_norm": 4.986356258392334,
+      "learning_rate": 2.4801587301587305e-05,
+      "loss": 0.324,
+      "step": 1280
+    },
+    {
+      "epoch": 5.063241106719367,
+      "grad_norm": 3.729706048965454,
+      "learning_rate": 2.4781746031746035e-05,
+      "loss": 0.476,
+      "step": 1281
+    },
+    {
+      "epoch": 5.067193675889328,
+      "grad_norm": 3.7136807441711426,
+      "learning_rate": 2.4761904761904762e-05,
+      "loss": 0.2756,
+      "step": 1282
+    },
+    {
+      "epoch": 5.071146245059288,
+      "grad_norm": 2.8954904079437256,
+      "learning_rate": 2.4742063492063493e-05,
+      "loss": 0.3505,
+      "step": 1283
+    },
+    {
+      "epoch": 5.075098814229249,
+      "grad_norm": 4.071838855743408,
+      "learning_rate": 2.4722222222222223e-05,
+      "loss": 0.447,
+      "step": 1284
+    },
+    {
+      "epoch": 5.07905138339921,
+      "grad_norm": 3.134131908416748,
+      "learning_rate": 2.4702380952380953e-05,
+      "loss": 0.3407,
+      "step": 1285
+    },
+    {
+      "epoch": 5.08300395256917,
+      "grad_norm": 3.690873861312866,
+      "learning_rate": 2.4682539682539684e-05,
+      "loss": 0.3665,
+      "step": 1286
+    },
+    {
+      "epoch": 5.086956521739131,
+      "grad_norm": 4.1070051193237305,
+      "learning_rate": 2.4662698412698414e-05,
+      "loss": 0.3099,
+      "step": 1287
+    },
+    {
+      "epoch": 5.090909090909091,
+      "grad_norm": 3.0999696254730225,
+      "learning_rate": 2.4642857142857145e-05,
+      "loss": 0.278,
+      "step": 1288
+    },
+    {
+      "epoch": 5.094861660079052,
+      "grad_norm": 4.00131893157959,
+      "learning_rate": 2.4623015873015875e-05,
+      "loss": 0.4586,
+      "step": 1289
+    },
+    {
+      "epoch": 5.098814229249012,
+      "grad_norm": 4.4364776611328125,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.5131,
+      "step": 1290
+    },
+    {
+      "epoch": 5.102766798418973,
+      "grad_norm": 4.154432773590088,
+      "learning_rate": 2.4583333333333332e-05,
+      "loss": 0.2289,
+      "step": 1291
+    },
+    {
+      "epoch": 5.106719367588933,
+      "grad_norm": 4.669454574584961,
+      "learning_rate": 2.4563492063492063e-05,
+      "loss": 0.3081,
+      "step": 1292
+    },
+    {
+      "epoch": 5.1106719367588935,
+      "grad_norm": 3.263782024383545,
+      "learning_rate": 2.4543650793650793e-05,
+      "loss": 0.2814,
+      "step": 1293
+    },
+    {
+      "epoch": 5.1146245059288535,
+      "grad_norm": 4.402275085449219,
+      "learning_rate": 2.4523809523809523e-05,
+      "loss": 0.2797,
+      "step": 1294
+    },
+    {
+      "epoch": 5.118577075098814,
+      "grad_norm": 3.0900776386260986,
+      "learning_rate": 2.4503968253968254e-05,
+      "loss": 0.2825,
+      "step": 1295
+    },
+    {
+      "epoch": 5.122529644268774,
+      "grad_norm": 3.7776150703430176,
+      "learning_rate": 2.4484126984126984e-05,
+      "loss": 0.3906,
+      "step": 1296
+    },
+    {
+      "epoch": 5.126482213438735,
+      "grad_norm": 3.909672975540161,
+      "learning_rate": 2.4464285714285715e-05,
+      "loss": 0.3245,
+      "step": 1297
+    },
+    {
+      "epoch": 5.130434782608695,
+      "grad_norm": 3.4614851474761963,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.364,
+      "step": 1298
+    },
+    {
+      "epoch": 5.134387351778656,
+      "grad_norm": 3.2292017936706543,
+      "learning_rate": 2.4424603174603175e-05,
+      "loss": 0.3371,
+      "step": 1299
+    },
+    {
+      "epoch": 5.138339920948616,
+      "grad_norm": 4.739649772644043,
+      "learning_rate": 2.4404761904761906e-05,
+      "loss": 0.3461,
+      "step": 1300
+    },
+    {
+      "epoch": 5.142292490118577,
+      "grad_norm": 3.756049156188965,
+      "learning_rate": 2.4384920634920636e-05,
+      "loss": 0.3143,
+      "step": 1301
+    },
+    {
+      "epoch": 5.146245059288537,
+      "grad_norm": 4.8067145347595215,
+      "learning_rate": 2.4365079365079366e-05,
+      "loss": 0.4448,
+      "step": 1302
+    },
+    {
+      "epoch": 5.150197628458498,
+      "grad_norm": 3.8932623863220215,
+      "learning_rate": 2.4345238095238097e-05,
+      "loss": 0.2526,
+      "step": 1303
+    },
+    {
+      "epoch": 5.154150197628459,
+      "grad_norm": 3.6755340099334717,
+      "learning_rate": 2.4325396825396827e-05,
+      "loss": 0.397,
+      "step": 1304
+    },
+    {
+      "epoch": 5.158102766798419,
+      "grad_norm": 3.7219412326812744,
+      "learning_rate": 2.4305555555555558e-05,
+      "loss": 0.1992,
+      "step": 1305
+    },
+    {
+      "epoch": 5.16205533596838,
+      "grad_norm": 3.744370698928833,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.2522,
+      "step": 1306
+    },
+    {
+      "epoch": 5.16600790513834,
+      "grad_norm": 3.2599740028381348,
+      "learning_rate": 2.426587301587302e-05,
+      "loss": 0.3581,
+      "step": 1307
+    },
+    {
+      "epoch": 5.169960474308301,
+      "grad_norm": 3.48649525642395,
+      "learning_rate": 2.424603174603175e-05,
+      "loss": 0.3864,
+      "step": 1308
+    },
+    {
+      "epoch": 5.173913043478261,
+      "grad_norm": 3.1932923793792725,
+      "learning_rate": 2.4226190476190476e-05,
+      "loss": 0.3054,
+      "step": 1309
+    },
+    {
+      "epoch": 5.177865612648222,
+      "grad_norm": 4.015415191650391,
+      "learning_rate": 2.4206349206349206e-05,
+      "loss": 0.2925,
+      "step": 1310
+    },
+    {
+      "epoch": 5.181818181818182,
+      "grad_norm": 3.9713363647460938,
+      "learning_rate": 2.4186507936507936e-05,
+      "loss": 0.2362,
+      "step": 1311
+    },
+    {
+      "epoch": 5.1857707509881426,
+      "grad_norm": 4.697850704193115,
+      "learning_rate": 2.4166666666666667e-05,
+      "loss": 0.2583,
+      "step": 1312
+    },
+    {
+      "epoch": 5.189723320158103,
+      "grad_norm": 3.067460298538208,
+      "learning_rate": 2.4146825396825397e-05,
+      "loss": 0.2453,
+      "step": 1313
+    },
+    {
+      "epoch": 5.1936758893280635,
+      "grad_norm": 4.411377906799316,
+      "learning_rate": 2.4126984126984128e-05,
+      "loss": 0.1776,
+      "step": 1314
+    },
+    {
+      "epoch": 5.1976284584980235,
+      "grad_norm": 3.420597791671753,
+      "learning_rate": 2.4107142857142858e-05,
+      "loss": 0.3165,
+      "step": 1315
+    },
+    {
+      "epoch": 5.201581027667984,
+      "grad_norm": 4.028104305267334,
+      "learning_rate": 2.408730158730159e-05,
+      "loss": 0.4102,
+      "step": 1316
+    },
+    {
+      "epoch": 5.205533596837944,
+      "grad_norm": 3.9082906246185303,
+      "learning_rate": 2.406746031746032e-05,
+      "loss": 0.3382,
+      "step": 1317
+    },
+    {
+      "epoch": 5.209486166007905,
+      "grad_norm": 4.633881092071533,
+      "learning_rate": 2.404761904761905e-05,
+      "loss": 0.2822,
+      "step": 1318
+    },
+    {
+      "epoch": 5.213438735177865,
+      "grad_norm": 3.6297354698181152,
+      "learning_rate": 2.402777777777778e-05,
+      "loss": 0.2072,
+      "step": 1319
+    },
+    {
+      "epoch": 5.217391304347826,
+      "grad_norm": 3.1377158164978027,
+      "learning_rate": 2.400793650793651e-05,
+      "loss": 0.271,
+      "step": 1320
+    },
+    {
+      "epoch": 5.221343873517786,
+      "grad_norm": 3.472032308578491,
+      "learning_rate": 2.398809523809524e-05,
+      "loss": 0.2801,
+      "step": 1321
+    },
+    {
+      "epoch": 5.225296442687747,
+      "grad_norm": 3.7419073581695557,
+      "learning_rate": 2.396825396825397e-05,
+      "loss": 0.3191,
+      "step": 1322
+    },
+    {
+      "epoch": 5.229249011857707,
+      "grad_norm": 3.5220224857330322,
+      "learning_rate": 2.39484126984127e-05,
+      "loss": 0.4082,
+      "step": 1323
+    },
+    {
+      "epoch": 5.233201581027668,
+      "grad_norm": 3.874562978744507,
+      "learning_rate": 2.392857142857143e-05,
+      "loss": 0.3281,
+      "step": 1324
+    },
+    {
+      "epoch": 5.237154150197629,
+      "grad_norm": 4.159897804260254,
+      "learning_rate": 2.390873015873016e-05,
+      "loss": 0.437,
+      "step": 1325
+    },
+    {
+      "epoch": 5.241106719367589,
+      "grad_norm": 4.653357982635498,
+      "learning_rate": 2.3888888888888892e-05,
+      "loss": 0.4198,
+      "step": 1326
+    },
+    {
+      "epoch": 5.24505928853755,
+      "grad_norm": 5.037525177001953,
+      "learning_rate": 2.3869047619047622e-05,
+      "loss": 0.4139,
+      "step": 1327
+    },
+    {
+      "epoch": 5.24901185770751,
+      "grad_norm": 4.0110368728637695,
+      "learning_rate": 2.3849206349206353e-05,
+      "loss": 0.2786,
+      "step": 1328
+    },
+    {
+      "epoch": 5.252964426877471,
+      "grad_norm": 3.9994056224823,
+      "learning_rate": 2.3829365079365083e-05,
+      "loss": 0.3709,
+      "step": 1329
+    },
+    {
+      "epoch": 5.256916996047431,
+      "grad_norm": 3.936352014541626,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.3744,
+      "step": 1330
+    },
+    {
+      "epoch": 5.260869565217392,
+      "grad_norm": 3.9300754070281982,
+      "learning_rate": 2.378968253968254e-05,
+      "loss": 0.5034,
+      "step": 1331
+    },
+    {
+      "epoch": 5.264822134387352,
+      "grad_norm": 4.13817024230957,
+      "learning_rate": 2.376984126984127e-05,
+      "loss": 0.4842,
+      "step": 1332
+    },
+    {
+      "epoch": 5.2687747035573125,
+      "grad_norm": 3.0476675033569336,
+      "learning_rate": 2.375e-05,
+      "loss": 0.2846,
+      "step": 1333
+    },
+    {
+      "epoch": 5.2727272727272725,
+      "grad_norm": 4.827993869781494,
+      "learning_rate": 2.373015873015873e-05,
+      "loss": 0.4083,
+      "step": 1334
+    },
+    {
+      "epoch": 5.276679841897233,
+      "grad_norm": 3.111452341079712,
+      "learning_rate": 2.3710317460317462e-05,
+      "loss": 0.3334,
+      "step": 1335
+    },
+    {
+      "epoch": 5.280632411067193,
+      "grad_norm": 4.097339630126953,
+      "learning_rate": 2.369047619047619e-05,
+      "loss": 0.3375,
+      "step": 1336
+    },
+    {
+      "epoch": 5.284584980237154,
+      "grad_norm": 4.1747822761535645,
+      "learning_rate": 2.367063492063492e-05,
+      "loss": 0.3817,
+      "step": 1337
+    },
+    {
+      "epoch": 5.288537549407114,
+      "grad_norm": 3.4249610900878906,
+      "learning_rate": 2.365079365079365e-05,
+      "loss": 0.3121,
+      "step": 1338
+    },
+    {
+      "epoch": 5.292490118577075,
+      "grad_norm": 3.0609562397003174,
+      "learning_rate": 2.363095238095238e-05,
+      "loss": 0.3,
+      "step": 1339
+    },
+    {
+      "epoch": 5.296442687747035,
+      "grad_norm": 4.675295829772949,
+      "learning_rate": 2.361111111111111e-05,
+      "loss": 0.4734,
+      "step": 1340
+    },
+    {
+      "epoch": 5.300395256916996,
+      "grad_norm": 4.56593656539917,
+      "learning_rate": 2.359126984126984e-05,
+      "loss": 0.386,
+      "step": 1341
+    },
+    {
+      "epoch": 5.304347826086957,
+      "grad_norm": 5.29105806350708,
+      "learning_rate": 2.357142857142857e-05,
+      "loss": 0.3133,
+      "step": 1342
+    },
+    {
+      "epoch": 5.308300395256917,
+      "grad_norm": 4.2684712409973145,
+      "learning_rate": 2.35515873015873e-05,
+      "loss": 0.4087,
+      "step": 1343
+    },
+    {
+      "epoch": 5.312252964426877,
+      "grad_norm": 5.04403018951416,
+      "learning_rate": 2.3531746031746032e-05,
+      "loss": 0.4299,
+      "step": 1344
+    },
+    {
+      "epoch": 5.316205533596838,
+      "grad_norm": 4.923758506774902,
+      "learning_rate": 2.3511904761904762e-05,
+      "loss": 0.3231,
+      "step": 1345
+    },
+    {
+      "epoch": 5.320158102766799,
+      "grad_norm": 2.760050058364868,
+      "learning_rate": 2.3492063492063493e-05,
+      "loss": 0.2268,
+      "step": 1346
+    },
+    {
+      "epoch": 5.324110671936759,
+      "grad_norm": 3.437049150466919,
+      "learning_rate": 2.3472222222222223e-05,
+      "loss": 0.391,
+      "step": 1347
+    },
+    {
+      "epoch": 5.32806324110672,
+      "grad_norm": 4.473261833190918,
+      "learning_rate": 2.3452380952380954e-05,
+      "loss": 0.4342,
+      "step": 1348
+    },
+    {
+      "epoch": 5.33201581027668,
+      "grad_norm": 4.289645195007324,
+      "learning_rate": 2.3432539682539684e-05,
+      "loss": 0.4789,
+      "step": 1349
+    },
+    {
+      "epoch": 5.335968379446641,
+      "grad_norm": 3.814570426940918,
+      "learning_rate": 2.3412698412698414e-05,
+      "loss": 0.3432,
+      "step": 1350
+    },
+    {
+      "epoch": 5.339920948616601,
+      "grad_norm": 3.5725395679473877,
+      "learning_rate": 2.3392857142857145e-05,
+      "loss": 0.2394,
+      "step": 1351
+    },
+    {
+      "epoch": 5.3438735177865615,
+      "grad_norm": 3.9966726303100586,
+      "learning_rate": 2.3373015873015875e-05,
+      "loss": 0.3657,
+      "step": 1352
+    },
+    {
+      "epoch": 5.3478260869565215,
+      "grad_norm": 3.7222578525543213,
+      "learning_rate": 2.3353174603174605e-05,
+      "loss": 0.2929,
+      "step": 1353
+    },
+    {
+      "epoch": 5.351778656126482,
+      "grad_norm": 3.659536361694336,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.5094,
+      "step": 1354
+    },
+    {
+      "epoch": 5.355731225296442,
+      "grad_norm": 4.850214004516602,
+      "learning_rate": 2.3313492063492066e-05,
+      "loss": 0.4599,
+      "step": 1355
+    },
+    {
+      "epoch": 5.359683794466403,
+      "grad_norm": 3.3052451610565186,
+      "learning_rate": 2.3293650793650797e-05,
+      "loss": 0.1597,
+      "step": 1356
+    },
+    {
+      "epoch": 5.363636363636363,
+      "grad_norm": 3.6496050357818604,
+      "learning_rate": 2.3273809523809527e-05,
+      "loss": 0.3862,
+      "step": 1357
+    },
+    {
+      "epoch": 5.367588932806324,
+      "grad_norm": 4.302274703979492,
+      "learning_rate": 2.3253968253968257e-05,
+      "loss": 0.2688,
+      "step": 1358
+    },
+    {
+      "epoch": 5.371541501976284,
+      "grad_norm": 4.052597522735596,
+      "learning_rate": 2.3234126984126984e-05,
+      "loss": 0.328,
+      "step": 1359
+    },
+    {
+      "epoch": 5.375494071146245,
+      "grad_norm": 4.061760425567627,
+      "learning_rate": 2.3214285714285715e-05,
+      "loss": 0.394,
+      "step": 1360
+    },
+    {
+      "epoch": 5.379446640316205,
+      "grad_norm": 4.183040618896484,
+      "learning_rate": 2.3194444444444445e-05,
+      "loss": 0.4477,
+      "step": 1361
+    },
+    {
+      "epoch": 5.383399209486166,
+      "grad_norm": 2.766695737838745,
+      "learning_rate": 2.3174603174603175e-05,
+      "loss": 0.1786,
+      "step": 1362
+    },
+    {
+      "epoch": 5.387351778656127,
+      "grad_norm": 5.040890693664551,
+      "learning_rate": 2.3154761904761906e-05,
+      "loss": 0.4192,
+      "step": 1363
+    },
+    {
+      "epoch": 5.391304347826087,
+      "grad_norm": 4.445863246917725,
+      "learning_rate": 2.3134920634920636e-05,
+      "loss": 0.5025,
+      "step": 1364
+    },
+    {
+      "epoch": 5.395256916996048,
+      "grad_norm": 3.4918015003204346,
+      "learning_rate": 2.3115079365079367e-05,
+      "loss": 0.2392,
+      "step": 1365
+    },
+    {
+      "epoch": 5.399209486166008,
+      "grad_norm": 3.366082191467285,
+      "learning_rate": 2.3095238095238097e-05,
+      "loss": 0.4595,
+      "step": 1366
+    },
+    {
+      "epoch": 5.403162055335969,
+      "grad_norm": 3.521512508392334,
+      "learning_rate": 2.3075396825396827e-05,
+      "loss": 0.422,
+      "step": 1367
+    },
+    {
+      "epoch": 5.407114624505929,
+      "grad_norm": 3.9632761478424072,
+      "learning_rate": 2.3055555555555558e-05,
+      "loss": 0.4579,
+      "step": 1368
+    },
+    {
+      "epoch": 5.41106719367589,
+      "grad_norm": 3.9189956188201904,
+      "learning_rate": 2.3035714285714285e-05,
+      "loss": 0.3785,
+      "step": 1369
+    },
+    {
+      "epoch": 5.41501976284585,
+      "grad_norm": 3.7866268157958984,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.223,
+      "step": 1370
+    },
+    {
+      "epoch": 5.4189723320158105,
+      "grad_norm": 3.4151864051818848,
+      "learning_rate": 2.2996031746031745e-05,
+      "loss": 0.1987,
+      "step": 1371
+    },
+    {
+      "epoch": 5.4229249011857705,
+      "grad_norm": 3.8373048305511475,
+      "learning_rate": 2.2976190476190476e-05,
+      "loss": 0.3249,
+      "step": 1372
+    },
+    {
+      "epoch": 5.426877470355731,
+      "grad_norm": 4.3949103355407715,
+      "learning_rate": 2.2956349206349206e-05,
+      "loss": 0.3704,
+      "step": 1373
+    },
+    {
+      "epoch": 5.430830039525691,
+      "grad_norm": 3.6524100303649902,
+      "learning_rate": 2.2936507936507937e-05,
+      "loss": 0.4462,
+      "step": 1374
+    },
+    {
+      "epoch": 5.434782608695652,
+      "grad_norm": 3.927030086517334,
+      "learning_rate": 2.2916666666666667e-05,
+      "loss": 0.3642,
+      "step": 1375
+    },
+    {
+      "epoch": 5.438735177865612,
+      "grad_norm": 4.005973815917969,
+      "learning_rate": 2.2896825396825397e-05,
+      "loss": 0.4237,
+      "step": 1376
+    },
+    {
+      "epoch": 5.442687747035573,
+      "grad_norm": 3.733210325241089,
+      "learning_rate": 2.2876984126984128e-05,
+      "loss": 0.3089,
+      "step": 1377
+    },
+    {
+      "epoch": 5.446640316205533,
+      "grad_norm": 5.140311241149902,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.4288,
+      "step": 1378
+    },
+    {
+      "epoch": 5.450592885375494,
+      "grad_norm": 3.3404946327209473,
+      "learning_rate": 2.283730158730159e-05,
+      "loss": 0.2962,
+      "step": 1379
+    },
+    {
+      "epoch": 5.454545454545454,
+      "grad_norm": 4.286531448364258,
+      "learning_rate": 2.281746031746032e-05,
+      "loss": 0.3931,
+      "step": 1380
+    },
+    {
+      "epoch": 5.458498023715415,
+      "grad_norm": 3.68621826171875,
+      "learning_rate": 2.279761904761905e-05,
+      "loss": 0.2881,
+      "step": 1381
+    },
+    {
+      "epoch": 5.462450592885375,
+      "grad_norm": 3.6594183444976807,
+      "learning_rate": 2.277777777777778e-05,
+      "loss": 0.293,
+      "step": 1382
+    },
+    {
+      "epoch": 5.466403162055336,
+      "grad_norm": 3.7897143363952637,
+      "learning_rate": 2.275793650793651e-05,
+      "loss": 0.207,
+      "step": 1383
+    },
+    {
+      "epoch": 5.470355731225297,
+      "grad_norm": 4.473965644836426,
+      "learning_rate": 2.273809523809524e-05,
+      "loss": 0.2935,
+      "step": 1384
+    },
+    {
+      "epoch": 5.474308300395257,
+      "grad_norm": 3.8163721561431885,
+      "learning_rate": 2.271825396825397e-05,
+      "loss": 0.4574,
+      "step": 1385
+    },
+    {
+      "epoch": 5.478260869565218,
+      "grad_norm": 3.263646125793457,
+      "learning_rate": 2.2698412698412698e-05,
+      "loss": 0.3372,
+      "step": 1386
+    },
+    {
+      "epoch": 5.482213438735178,
+      "grad_norm": 4.062425136566162,
+      "learning_rate": 2.2678571428571428e-05,
+      "loss": 0.2997,
+      "step": 1387
+    },
+    {
+      "epoch": 5.486166007905139,
+      "grad_norm": 3.691974401473999,
+      "learning_rate": 2.265873015873016e-05,
+      "loss": 0.3711,
+      "step": 1388
+    },
+    {
+      "epoch": 5.490118577075099,
+      "grad_norm": 3.3711211681365967,
+      "learning_rate": 2.263888888888889e-05,
+      "loss": 0.293,
+      "step": 1389
+    },
+    {
+      "epoch": 5.4940711462450595,
+      "grad_norm": 3.659691572189331,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 0.2519,
+      "step": 1390
+    },
+    {
+      "epoch": 5.4980237154150196,
+      "grad_norm": 4.6761088371276855,
+      "learning_rate": 2.259920634920635e-05,
+      "loss": 0.3788,
+      "step": 1391
+    },
+    {
+      "epoch": 5.5019762845849804,
+      "grad_norm": 4.013514995574951,
+      "learning_rate": 2.257936507936508e-05,
+      "loss": 0.3685,
+      "step": 1392
+    },
+    {
+      "epoch": 5.5059288537549405,
+      "grad_norm": 3.272243022918701,
+      "learning_rate": 2.255952380952381e-05,
+      "loss": 0.2541,
+      "step": 1393
+    },
+    {
+      "epoch": 5.509881422924901,
+      "grad_norm": 4.194965362548828,
+      "learning_rate": 2.253968253968254e-05,
+      "loss": 0.3032,
+      "step": 1394
+    },
+    {
+      "epoch": 5.513833992094861,
+      "grad_norm": 5.20955228805542,
+      "learning_rate": 2.251984126984127e-05,
+      "loss": 0.4339,
+      "step": 1395
+    },
+    {
+      "epoch": 5.517786561264822,
+      "grad_norm": 3.9413256645202637,
+      "learning_rate": 2.25e-05,
+      "loss": 0.2921,
+      "step": 1396
+    },
+    {
+      "epoch": 5.521739130434782,
+      "grad_norm": 3.783141613006592,
+      "learning_rate": 2.2480158730158732e-05,
+      "loss": 0.2763,
+      "step": 1397
+    },
+    {
+      "epoch": 5.525691699604743,
+      "grad_norm": 3.764054298400879,
+      "learning_rate": 2.2460317460317462e-05,
+      "loss": 0.4252,
+      "step": 1398
+    },
+    {
+      "epoch": 5.529644268774703,
+      "grad_norm": 3.668379545211792,
+      "learning_rate": 2.2440476190476193e-05,
+      "loss": 0.3577,
+      "step": 1399
+    },
+    {
+      "epoch": 5.533596837944664,
+      "grad_norm": 4.717470645904541,
+      "learning_rate": 2.2420634920634923e-05,
+      "loss": 0.2837,
+      "step": 1400
+    },
+    {
+      "epoch": 5.537549407114625,
+      "grad_norm": 4.919825553894043,
+      "learning_rate": 2.2400793650793653e-05,
+      "loss": 0.4905,
+      "step": 1401
+    },
+    {
+      "epoch": 5.541501976284585,
+      "grad_norm": 3.827908754348755,
+      "learning_rate": 2.2380952380952384e-05,
+      "loss": 0.219,
+      "step": 1402
+    },
+    {
+      "epoch": 5.545454545454545,
+      "grad_norm": 3.300463914871216,
+      "learning_rate": 2.2361111111111114e-05,
+      "loss": 0.2114,
+      "step": 1403
+    },
+    {
+      "epoch": 5.549407114624506,
+      "grad_norm": 4.180085182189941,
+      "learning_rate": 2.2341269841269844e-05,
+      "loss": 0.4172,
+      "step": 1404
+    },
+    {
+      "epoch": 5.553359683794467,
+      "grad_norm": 3.7211532592773438,
+      "learning_rate": 2.2321428571428575e-05,
+      "loss": 0.3994,
+      "step": 1405
+    },
+    {
+      "epoch": 5.557312252964427,
+      "grad_norm": 4.4185991287231445,
+      "learning_rate": 2.2301587301587305e-05,
+      "loss": 0.4922,
+      "step": 1406
+    },
+    {
+      "epoch": 5.561264822134388,
+      "grad_norm": 5.1140456199646,
+      "learning_rate": 2.2281746031746036e-05,
+      "loss": 0.4038,
+      "step": 1407
+    },
+    {
+      "epoch": 5.565217391304348,
+      "grad_norm": 3.7456719875335693,
+      "learning_rate": 2.2261904761904763e-05,
+      "loss": 0.2587,
+      "step": 1408
+    },
+    {
+      "epoch": 5.569169960474309,
+      "grad_norm": 4.17236852645874,
+      "learning_rate": 2.2242063492063493e-05,
+      "loss": 0.3107,
+      "step": 1409
+    },
+    {
+      "epoch": 5.573122529644269,
+      "grad_norm": 3.580836057662964,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3648,
+      "step": 1410
+    },
+    {
+      "epoch": 5.5770750988142295,
+      "grad_norm": 4.266388893127441,
+      "learning_rate": 2.2202380952380954e-05,
+      "loss": 0.3403,
+      "step": 1411
+    },
+    {
+      "epoch": 5.5810276679841895,
+      "grad_norm": 2.99090576171875,
+      "learning_rate": 2.2182539682539684e-05,
+      "loss": 0.3277,
+      "step": 1412
+    },
+    {
+      "epoch": 5.58498023715415,
+      "grad_norm": 4.490349292755127,
+      "learning_rate": 2.2162698412698414e-05,
+      "loss": 0.4371,
+      "step": 1413
+    },
+    {
+      "epoch": 5.58893280632411,
+      "grad_norm": 4.330888271331787,
+      "learning_rate": 2.214285714285714e-05,
+      "loss": 0.2821,
+      "step": 1414
+    },
+    {
+      "epoch": 5.592885375494071,
+      "grad_norm": 3.7015597820281982,
+      "learning_rate": 2.2123015873015872e-05,
+      "loss": 0.2986,
+      "step": 1415
+    },
+    {
+      "epoch": 5.596837944664031,
+      "grad_norm": 4.108189582824707,
+      "learning_rate": 2.2103174603174602e-05,
+      "loss": 0.4689,
+      "step": 1416
+    },
+    {
+      "epoch": 5.600790513833992,
+      "grad_norm": 3.0757765769958496,
+      "learning_rate": 2.2083333333333333e-05,
+      "loss": 0.2335,
+      "step": 1417
+    },
+    {
+      "epoch": 5.604743083003952,
+      "grad_norm": 3.065613031387329,
+      "learning_rate": 2.2063492063492063e-05,
+      "loss": 0.3694,
+      "step": 1418
+    },
+    {
+      "epoch": 5.608695652173913,
+      "grad_norm": 4.539591312408447,
+      "learning_rate": 2.2043650793650793e-05,
+      "loss": 0.3448,
+      "step": 1419
+    },
+    {
+      "epoch": 5.612648221343873,
+      "grad_norm": 3.5507845878601074,
+      "learning_rate": 2.2023809523809524e-05,
+      "loss": 0.4036,
+      "step": 1420
+    },
+    {
+      "epoch": 5.616600790513834,
+      "grad_norm": 3.766512632369995,
+      "learning_rate": 2.2003968253968254e-05,
+      "loss": 0.3682,
+      "step": 1421
+    },
+    {
+      "epoch": 5.620553359683795,
+      "grad_norm": 3.268449306488037,
+      "learning_rate": 2.1984126984126984e-05,
+      "loss": 0.2589,
+      "step": 1422
+    },
+    {
+      "epoch": 5.624505928853755,
+      "grad_norm": 3.850033760070801,
+      "learning_rate": 2.1964285714285715e-05,
+      "loss": 0.2949,
+      "step": 1423
+    },
+    {
+      "epoch": 5.628458498023716,
+      "grad_norm": 4.563751220703125,
+      "learning_rate": 2.1944444444444445e-05,
+      "loss": 0.387,
+      "step": 1424
+    },
+    {
+      "epoch": 5.632411067193676,
+      "grad_norm": 3.109978199005127,
+      "learning_rate": 2.1924603174603176e-05,
+      "loss": 0.2175,
+      "step": 1425
+    },
+    {
+      "epoch": 5.636363636363637,
+      "grad_norm": 4.669355869293213,
+      "learning_rate": 2.1904761904761906e-05,
+      "loss": 0.2422,
+      "step": 1426
+    },
+    {
+      "epoch": 5.640316205533597,
+      "grad_norm": 3.95523738861084,
+      "learning_rate": 2.1884920634920636e-05,
+      "loss": 0.3867,
+      "step": 1427
+    },
+    {
+      "epoch": 5.644268774703558,
+      "grad_norm": 3.9912211894989014,
+      "learning_rate": 2.1865079365079367e-05,
+      "loss": 0.2815,
+      "step": 1428
+    },
+    {
+      "epoch": 5.648221343873518,
+      "grad_norm": 3.4305765628814697,
+      "learning_rate": 2.1845238095238097e-05,
+      "loss": 0.263,
+      "step": 1429
+    },
+    {
+      "epoch": 5.6521739130434785,
+      "grad_norm": 4.1827216148376465,
+      "learning_rate": 2.1825396825396827e-05,
+      "loss": 0.292,
+      "step": 1430
+    },
+    {
+      "epoch": 5.6561264822134385,
+      "grad_norm": 3.6767704486846924,
+      "learning_rate": 2.1805555555555558e-05,
+      "loss": 0.2732,
+      "step": 1431
+    },
+    {
+      "epoch": 5.660079051383399,
+      "grad_norm": 4.459062099456787,
+      "learning_rate": 2.1785714285714288e-05,
+      "loss": 0.3492,
+      "step": 1432
+    },
+    {
+      "epoch": 5.664031620553359,
+      "grad_norm": 3.5659234523773193,
+      "learning_rate": 2.176587301587302e-05,
+      "loss": 0.2712,
+      "step": 1433
+    },
+    {
+      "epoch": 5.66798418972332,
+      "grad_norm": 4.487940788269043,
+      "learning_rate": 2.174603174603175e-05,
+      "loss": 0.2831,
+      "step": 1434
+    },
+    {
+      "epoch": 5.67193675889328,
+      "grad_norm": 5.612968921661377,
+      "learning_rate": 2.172619047619048e-05,
+      "loss": 0.5013,
+      "step": 1435
+    },
+    {
+      "epoch": 5.675889328063241,
+      "grad_norm": 5.958313465118408,
+      "learning_rate": 2.170634920634921e-05,
+      "loss": 0.4195,
+      "step": 1436
+    },
+    {
+      "epoch": 5.679841897233201,
+      "grad_norm": 4.63142728805542,
+      "learning_rate": 2.1686507936507937e-05,
+      "loss": 0.421,
+      "step": 1437
+    },
+    {
+      "epoch": 5.683794466403162,
+      "grad_norm": 3.549577236175537,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.3966,
+      "step": 1438
+    },
+    {
+      "epoch": 5.687747035573123,
+      "grad_norm": 3.38968825340271,
+      "learning_rate": 2.1646825396825397e-05,
+      "loss": 0.3352,
+      "step": 1439
+    },
+    {
+      "epoch": 5.691699604743083,
+      "grad_norm": 3.199084520339966,
+      "learning_rate": 2.1626984126984128e-05,
+      "loss": 0.2996,
+      "step": 1440
+    },
+    {
+      "epoch": 5.695652173913043,
+      "grad_norm": 3.6920907497406006,
+      "learning_rate": 2.1607142857142858e-05,
+      "loss": 0.329,
+      "step": 1441
+    },
+    {
+      "epoch": 5.699604743083004,
+      "grad_norm": 3.676727533340454,
+      "learning_rate": 2.158730158730159e-05,
+      "loss": 0.5122,
+      "step": 1442
+    },
+    {
+      "epoch": 5.703557312252965,
+      "grad_norm": 3.520972728729248,
+      "learning_rate": 2.156746031746032e-05,
+      "loss": 0.2829,
+      "step": 1443
+    },
+    {
+      "epoch": 5.707509881422925,
+      "grad_norm": 4.055224895477295,
+      "learning_rate": 2.154761904761905e-05,
+      "loss": 0.3474,
+      "step": 1444
+    },
+    {
+      "epoch": 5.711462450592886,
+      "grad_norm": 4.417123317718506,
+      "learning_rate": 2.152777777777778e-05,
+      "loss": 0.3876,
+      "step": 1445
+    },
+    {
+      "epoch": 5.715415019762846,
+      "grad_norm": 3.9218266010284424,
+      "learning_rate": 2.150793650793651e-05,
+      "loss": 0.4128,
+      "step": 1446
+    },
+    {
+      "epoch": 5.719367588932807,
+      "grad_norm": 4.506389141082764,
+      "learning_rate": 2.148809523809524e-05,
+      "loss": 0.4672,
+      "step": 1447
+    },
+    {
+      "epoch": 5.723320158102767,
+      "grad_norm": 3.930652141571045,
+      "learning_rate": 2.1468253968253967e-05,
+      "loss": 0.2926,
+      "step": 1448
+    },
+    {
+      "epoch": 5.7272727272727275,
+      "grad_norm": 4.012630939483643,
+      "learning_rate": 2.1448412698412698e-05,
+      "loss": 0.4217,
+      "step": 1449
+    },
+    {
+      "epoch": 5.7312252964426875,
+      "grad_norm": 4.338024139404297,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.2772,
+      "step": 1450
+    },
+    {
+      "epoch": 5.735177865612648,
+      "grad_norm": 3.886202096939087,
+      "learning_rate": 2.140873015873016e-05,
+      "loss": 0.3482,
+      "step": 1451
+    },
+    {
+      "epoch": 5.739130434782608,
+      "grad_norm": 4.083372116088867,
+      "learning_rate": 2.138888888888889e-05,
+      "loss": 0.2735,
+      "step": 1452
+    },
+    {
+      "epoch": 5.743083003952569,
+      "grad_norm": 3.7557191848754883,
+      "learning_rate": 2.136904761904762e-05,
+      "loss": 0.4228,
+      "step": 1453
+    },
+    {
+      "epoch": 5.747035573122529,
+      "grad_norm": 3.172152519226074,
+      "learning_rate": 2.134920634920635e-05,
+      "loss": 0.3562,
+      "step": 1454
+    },
+    {
+      "epoch": 5.75098814229249,
+      "grad_norm": 3.7699217796325684,
+      "learning_rate": 2.132936507936508e-05,
+      "loss": 0.2838,
+      "step": 1455
+    },
+    {
+      "epoch": 5.75494071146245,
+      "grad_norm": 3.8559303283691406,
+      "learning_rate": 2.130952380952381e-05,
+      "loss": 0.5692,
+      "step": 1456
+    },
+    {
+      "epoch": 5.758893280632411,
+      "grad_norm": 5.278477191925049,
+      "learning_rate": 2.128968253968254e-05,
+      "loss": 0.5161,
+      "step": 1457
+    },
+    {
+      "epoch": 5.762845849802371,
+      "grad_norm": 4.712038993835449,
+      "learning_rate": 2.126984126984127e-05,
+      "loss": 0.473,
+      "step": 1458
+    },
+    {
+      "epoch": 5.766798418972332,
+      "grad_norm": 3.714198112487793,
+      "learning_rate": 2.125e-05,
+      "loss": 0.2426,
+      "step": 1459
+    },
+    {
+      "epoch": 5.770750988142293,
+      "grad_norm": 3.246757745742798,
+      "learning_rate": 2.1230158730158732e-05,
+      "loss": 0.4447,
+      "step": 1460
+    },
+    {
+      "epoch": 5.774703557312253,
+      "grad_norm": 4.785327434539795,
+      "learning_rate": 2.1210317460317462e-05,
+      "loss": 0.3517,
+      "step": 1461
+    },
+    {
+      "epoch": 5.778656126482213,
+      "grad_norm": 4.272199630737305,
+      "learning_rate": 2.1190476190476193e-05,
+      "loss": 0.329,
+      "step": 1462
+    },
+    {
+      "epoch": 5.782608695652174,
+      "grad_norm": 4.346473217010498,
+      "learning_rate": 2.1170634920634923e-05,
+      "loss": 0.4102,
+      "step": 1463
+    },
+    {
+      "epoch": 5.786561264822135,
+      "grad_norm": 3.4643688201904297,
+      "learning_rate": 2.115079365079365e-05,
+      "loss": 0.2851,
+      "step": 1464
+    },
+    {
+      "epoch": 5.790513833992095,
+      "grad_norm": 4.473227024078369,
+      "learning_rate": 2.113095238095238e-05,
+      "loss": 0.3341,
+      "step": 1465
+    },
+    {
+      "epoch": 5.794466403162056,
+      "grad_norm": 3.813743829727173,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.4491,
+      "step": 1466
+    },
+    {
+      "epoch": 5.798418972332016,
+      "grad_norm": 4.01721715927124,
+      "learning_rate": 2.109126984126984e-05,
+      "loss": 0.1942,
+      "step": 1467
+    },
+    {
+      "epoch": 5.8023715415019765,
+      "grad_norm": 3.757263422012329,
+      "learning_rate": 2.107142857142857e-05,
+      "loss": 0.3666,
+      "step": 1468
+    },
+    {
+      "epoch": 5.8063241106719365,
+      "grad_norm": 3.4394917488098145,
+      "learning_rate": 2.1051587301587302e-05,
+      "loss": 0.3211,
+      "step": 1469
+    },
+    {
+      "epoch": 5.810276679841897,
+      "grad_norm": 5.106522083282471,
+      "learning_rate": 2.1031746031746032e-05,
+      "loss": 0.3757,
+      "step": 1470
+    },
+    {
+      "epoch": 5.8142292490118574,
+      "grad_norm": 4.171218395233154,
+      "learning_rate": 2.1011904761904763e-05,
+      "loss": 0.4219,
+      "step": 1471
+    },
+    {
+      "epoch": 5.818181818181818,
+      "grad_norm": 4.083078384399414,
+      "learning_rate": 2.0992063492063493e-05,
+      "loss": 0.3451,
+      "step": 1472
+    },
+    {
+      "epoch": 5.822134387351778,
+      "grad_norm": 4.33717679977417,
+      "learning_rate": 2.0972222222222223e-05,
+      "loss": 0.5292,
+      "step": 1473
+    },
+    {
+      "epoch": 5.826086956521739,
+      "grad_norm": 4.866550445556641,
+      "learning_rate": 2.0952380952380954e-05,
+      "loss": 0.3334,
+      "step": 1474
+    },
+    {
+      "epoch": 5.830039525691699,
+      "grad_norm": 3.197094202041626,
+      "learning_rate": 2.0932539682539684e-05,
+      "loss": 0.2206,
+      "step": 1475
+    },
+    {
+      "epoch": 5.83399209486166,
+      "grad_norm": 3.3910369873046875,
+      "learning_rate": 2.0912698412698415e-05,
+      "loss": 0.1893,
+      "step": 1476
+    },
+    {
+      "epoch": 5.837944664031621,
+      "grad_norm": 4.107388496398926,
+      "learning_rate": 2.0892857142857145e-05,
+      "loss": 0.4357,
+      "step": 1477
+    },
+    {
+      "epoch": 5.841897233201581,
+      "grad_norm": 2.6555662155151367,
+      "learning_rate": 2.0873015873015875e-05,
+      "loss": 0.2589,
+      "step": 1478
+    },
+    {
+      "epoch": 5.845849802371541,
+      "grad_norm": 3.3115289211273193,
+      "learning_rate": 2.0853174603174606e-05,
+      "loss": 0.3587,
+      "step": 1479
+    },
+    {
+      "epoch": 5.849802371541502,
+      "grad_norm": 3.604375123977661,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.273,
+      "step": 1480
+    },
+    {
+      "epoch": 5.853754940711463,
+      "grad_norm": 4.312830448150635,
+      "learning_rate": 2.0813492063492066e-05,
+      "loss": 0.2748,
+      "step": 1481
+    },
+    {
+      "epoch": 5.857707509881423,
+      "grad_norm": 4.150437831878662,
+      "learning_rate": 2.0793650793650797e-05,
+      "loss": 0.3099,
+      "step": 1482
+    },
+    {
+      "epoch": 5.861660079051384,
+      "grad_norm": 4.916041851043701,
+      "learning_rate": 2.0773809523809527e-05,
+      "loss": 0.3229,
+      "step": 1483
+    },
+    {
+      "epoch": 5.865612648221344,
+      "grad_norm": 4.0450663566589355,
+      "learning_rate": 2.0753968253968258e-05,
+      "loss": 0.3658,
+      "step": 1484
+    },
+    {
+      "epoch": 5.869565217391305,
+      "grad_norm": 3.5350306034088135,
+      "learning_rate": 2.0734126984126988e-05,
+      "loss": 0.3391,
+      "step": 1485
+    },
+    {
+      "epoch": 5.873517786561265,
+      "grad_norm": 3.364154100418091,
+      "learning_rate": 2.0714285714285718e-05,
+      "loss": 0.1964,
+      "step": 1486
+    },
+    {
+      "epoch": 5.877470355731226,
+      "grad_norm": 5.209838390350342,
+      "learning_rate": 2.0694444444444445e-05,
+      "loss": 0.3585,
+      "step": 1487
+    },
+    {
+      "epoch": 5.881422924901186,
+      "grad_norm": 3.9065868854522705,
+      "learning_rate": 2.0674603174603176e-05,
+      "loss": 0.2676,
+      "step": 1488
+    },
+    {
+      "epoch": 5.8853754940711465,
+      "grad_norm": 4.029334545135498,
+      "learning_rate": 2.0654761904761906e-05,
+      "loss": 0.2837,
+      "step": 1489
+    },
+    {
+      "epoch": 5.8893280632411065,
+      "grad_norm": 4.0023112297058105,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.4846,
+      "step": 1490
+    },
+    {
+      "epoch": 5.893280632411067,
+      "grad_norm": 4.869349479675293,
+      "learning_rate": 2.0615079365079363e-05,
+      "loss": 0.3852,
+      "step": 1491
+    },
+    {
+      "epoch": 5.897233201581027,
+      "grad_norm": 4.35573148727417,
+      "learning_rate": 2.0595238095238094e-05,
+      "loss": 0.2365,
+      "step": 1492
+    },
+    {
+      "epoch": 5.901185770750988,
+      "grad_norm": 4.35988187789917,
+      "learning_rate": 2.0575396825396824e-05,
+      "loss": 0.6209,
+      "step": 1493
+    },
+    {
+      "epoch": 5.905138339920948,
+      "grad_norm": 3.8130955696105957,
+      "learning_rate": 2.0555555555555555e-05,
+      "loss": 0.3111,
+      "step": 1494
+    },
+    {
+      "epoch": 5.909090909090909,
+      "grad_norm": 3.3116374015808105,
+      "learning_rate": 2.0535714285714285e-05,
+      "loss": 0.261,
+      "step": 1495
+    },
+    {
+      "epoch": 5.913043478260869,
+      "grad_norm": 4.634829521179199,
+      "learning_rate": 2.0515873015873015e-05,
+      "loss": 0.3943,
+      "step": 1496
+    },
+    {
+      "epoch": 5.91699604743083,
+      "grad_norm": 3.972263813018799,
+      "learning_rate": 2.0496031746031746e-05,
+      "loss": 0.3012,
+      "step": 1497
+    },
+    {
+      "epoch": 5.920948616600791,
+      "grad_norm": 4.382541656494141,
+      "learning_rate": 2.0476190476190476e-05,
+      "loss": 0.2957,
+      "step": 1498
+    },
+    {
+      "epoch": 5.924901185770751,
+      "grad_norm": 4.666507720947266,
+      "learning_rate": 2.0456349206349206e-05,
+      "loss": 0.4242,
+      "step": 1499
+    },
+    {
+      "epoch": 5.928853754940711,
+      "grad_norm": 3.7338948249816895,
+      "learning_rate": 2.0436507936507937e-05,
+      "loss": 0.2565,
+      "step": 1500
+    },
+    {
+      "epoch": 5.932806324110672,
+      "grad_norm": 3.5870771408081055,
+      "learning_rate": 2.0416666666666667e-05,
+      "loss": 0.4244,
+      "step": 1501
+    },
+    {
+      "epoch": 5.936758893280633,
+      "grad_norm": 4.410994052886963,
+      "learning_rate": 2.0396825396825398e-05,
+      "loss": 0.3985,
+      "step": 1502
+    },
+    {
+      "epoch": 5.940711462450593,
+      "grad_norm": 5.01856803894043,
+      "learning_rate": 2.0376984126984128e-05,
+      "loss": 0.354,
+      "step": 1503
+    },
+    {
+      "epoch": 5.944664031620554,
+      "grad_norm": 3.5459818840026855,
+      "learning_rate": 2.0357142857142858e-05,
+      "loss": 0.3448,
+      "step": 1504
+    },
+    {
+      "epoch": 5.948616600790514,
+      "grad_norm": 4.243325233459473,
+      "learning_rate": 2.033730158730159e-05,
+      "loss": 0.4499,
+      "step": 1505
+    },
+    {
+      "epoch": 5.952569169960475,
+      "grad_norm": 4.445835113525391,
+      "learning_rate": 2.031746031746032e-05,
+      "loss": 0.4184,
+      "step": 1506
+    },
+    {
+      "epoch": 5.956521739130435,
+      "grad_norm": 4.013076305389404,
+      "learning_rate": 2.029761904761905e-05,
+      "loss": 0.3721,
+      "step": 1507
+    },
+    {
+      "epoch": 5.9604743083003955,
+      "grad_norm": 4.0771284103393555,
+      "learning_rate": 2.027777777777778e-05,
+      "loss": 0.414,
+      "step": 1508
+    },
+    {
+      "epoch": 5.9644268774703555,
+      "grad_norm": 4.353712558746338,
+      "learning_rate": 2.025793650793651e-05,
+      "loss": 0.4212,
+      "step": 1509
+    },
+    {
+      "epoch": 5.968379446640316,
+      "grad_norm": 3.816298007965088,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 0.2885,
+      "step": 1510
+    },
+    {
+      "epoch": 5.972332015810276,
+      "grad_norm": 5.043026447296143,
+      "learning_rate": 2.021825396825397e-05,
+      "loss": 0.4901,
+      "step": 1511
+    },
+    {
+      "epoch": 5.976284584980237,
+      "grad_norm": 3.432596206665039,
+      "learning_rate": 2.01984126984127e-05,
+      "loss": 0.4023,
+      "step": 1512
+    },
+    {
+      "epoch": 5.980237154150197,
+      "grad_norm": 5.160770893096924,
+      "learning_rate": 2.017857142857143e-05,
+      "loss": 0.435,
+      "step": 1513
+    },
+    {
+      "epoch": 5.984189723320158,
+      "grad_norm": 3.863649845123291,
+      "learning_rate": 2.015873015873016e-05,
+      "loss": 0.2929,
+      "step": 1514
+    },
+    {
+      "epoch": 5.988142292490118,
+      "grad_norm": 3.960110902786255,
+      "learning_rate": 2.013888888888889e-05,
+      "loss": 0.3358,
+      "step": 1515
+    },
+    {
+      "epoch": 5.992094861660079,
+      "grad_norm": 3.597496747970581,
+      "learning_rate": 2.011904761904762e-05,
+      "loss": 0.3232,
+      "step": 1516
+    },
+    {
+      "epoch": 5.996047430830039,
+      "grad_norm": 3.676575183868408,
+      "learning_rate": 2.009920634920635e-05,
+      "loss": 0.3356,
+      "step": 1517
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.146391868591309,
+      "learning_rate": 2.007936507936508e-05,
+      "loss": 0.3818,
+      "step": 1518
+    },
+    {
+      "epoch": 6.003952569169961,
+      "grad_norm": 1.9677680730819702,
+      "learning_rate": 2.005952380952381e-05,
+      "loss": 0.1328,
+      "step": 1519
+    },
+    {
+      "epoch": 6.007905138339921,
+      "grad_norm": 3.2990286350250244,
+      "learning_rate": 2.003968253968254e-05,
+      "loss": 0.1407,
+      "step": 1520
+    },
+    {
+      "epoch": 6.011857707509882,
+      "grad_norm": 3.0897316932678223,
+      "learning_rate": 2.001984126984127e-05,
+      "loss": 0.23,
+      "step": 1521
+    },
+    {
+      "epoch": 6.015810276679842,
+      "grad_norm": 3.0987026691436768,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 1522
+    },
+    {
+      "epoch": 6.019762845849803,
+      "grad_norm": 2.3966803550720215,
+      "learning_rate": 1.9980158730158732e-05,
+      "loss": 0.1346,
+      "step": 1523
+    },
+    {
+      "epoch": 6.023715415019763,
+      "grad_norm": 2.7668561935424805,
+      "learning_rate": 1.9960317460317462e-05,
+      "loss": 0.1102,
+      "step": 1524
+    },
+    {
+      "epoch": 6.027667984189724,
+      "grad_norm": 3.623638868331909,
+      "learning_rate": 1.9940476190476193e-05,
+      "loss": 0.1318,
+      "step": 1525
+    },
+    {
+      "epoch": 6.031620553359684,
+      "grad_norm": 3.124091148376465,
+      "learning_rate": 1.992063492063492e-05,
+      "loss": 0.2056,
+      "step": 1526
+    },
+    {
+      "epoch": 6.0355731225296445,
+      "grad_norm": 3.1599087715148926,
+      "learning_rate": 1.990079365079365e-05,
+      "loss": 0.1375,
+      "step": 1527
+    },
+    {
+      "epoch": 6.0395256916996045,
+      "grad_norm": 3.4084415435791016,
+      "learning_rate": 1.988095238095238e-05,
+      "loss": 0.1443,
+      "step": 1528
+    },
+    {
+      "epoch": 6.043478260869565,
+      "grad_norm": 3.607503652572632,
+      "learning_rate": 1.986111111111111e-05,
+      "loss": 0.1846,
+      "step": 1529
+    },
+    {
+      "epoch": 6.047430830039525,
+      "grad_norm": 3.6665408611297607,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.1801,
+      "step": 1530
+    },
+    {
+      "epoch": 6.051383399209486,
+      "grad_norm": 3.5593857765197754,
+      "learning_rate": 1.982142857142857e-05,
+      "loss": 0.2119,
+      "step": 1531
+    },
+    {
+      "epoch": 6.055335968379446,
+      "grad_norm": 2.55330228805542,
+      "learning_rate": 1.9801587301587302e-05,
+      "loss": 0.119,
+      "step": 1532
+    },
+    {
+      "epoch": 6.059288537549407,
+      "grad_norm": 4.402235984802246,
+      "learning_rate": 1.9781746031746032e-05,
+      "loss": 0.2068,
+      "step": 1533
+    },
+    {
+      "epoch": 6.063241106719367,
+      "grad_norm": 2.9708304405212402,
+      "learning_rate": 1.9761904761904763e-05,
+      "loss": 0.1736,
+      "step": 1534
+    },
+    {
+      "epoch": 6.067193675889328,
+      "grad_norm": 4.265127182006836,
+      "learning_rate": 1.9742063492063493e-05,
+      "loss": 0.2059,
+      "step": 1535
+    },
+    {
+      "epoch": 6.071146245059288,
+      "grad_norm": 3.346165895462036,
+      "learning_rate": 1.9722222222222224e-05,
+      "loss": 0.2927,
+      "step": 1536
+    },
+    {
+      "epoch": 6.075098814229249,
+      "grad_norm": 3.0124001502990723,
+      "learning_rate": 1.9702380952380954e-05,
+      "loss": 0.1643,
+      "step": 1537
+    },
+    {
+      "epoch": 6.07905138339921,
+      "grad_norm": 3.65307879447937,
+      "learning_rate": 1.9682539682539684e-05,
+      "loss": 0.1432,
+      "step": 1538
+    },
+    {
+      "epoch": 6.08300395256917,
+      "grad_norm": 3.4866161346435547,
+      "learning_rate": 1.9662698412698415e-05,
+      "loss": 0.1515,
+      "step": 1539
+    },
+    {
+      "epoch": 6.086956521739131,
+      "grad_norm": 3.2024996280670166,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 0.1209,
+      "step": 1540
+    },
+    {
+      "epoch": 6.090909090909091,
+      "grad_norm": 3.161350965499878,
+      "learning_rate": 1.9623015873015872e-05,
+      "loss": 0.1511,
+      "step": 1541
+    },
+    {
+      "epoch": 6.094861660079052,
+      "grad_norm": 2.771225929260254,
+      "learning_rate": 1.9603174603174602e-05,
+      "loss": 0.1782,
+      "step": 1542
+    },
+    {
+      "epoch": 6.098814229249012,
+      "grad_norm": 2.3539700508117676,
+      "learning_rate": 1.9583333333333333e-05,
+      "loss": 0.1533,
+      "step": 1543
+    },
+    {
+      "epoch": 6.102766798418973,
+      "grad_norm": 3.1462249755859375,
+      "learning_rate": 1.9563492063492063e-05,
+      "loss": 0.1733,
+      "step": 1544
+    },
+    {
+      "epoch": 6.106719367588933,
+      "grad_norm": 3.8993260860443115,
+      "learning_rate": 1.9543650793650793e-05,
+      "loss": 0.2576,
+      "step": 1545
+    },
+    {
+      "epoch": 6.1106719367588935,
+      "grad_norm": 4.060632228851318,
+      "learning_rate": 1.9523809523809524e-05,
+      "loss": 0.2333,
+      "step": 1546
+    },
+    {
+      "epoch": 6.1146245059288535,
+      "grad_norm": 3.7323641777038574,
+      "learning_rate": 1.9503968253968254e-05,
+      "loss": 0.256,
+      "step": 1547
+    },
+    {
+      "epoch": 6.118577075098814,
+      "grad_norm": 2.8360302448272705,
+      "learning_rate": 1.9484126984126985e-05,
+      "loss": 0.1162,
+      "step": 1548
+    },
+    {
+      "epoch": 6.122529644268774,
+      "grad_norm": 4.867652416229248,
+      "learning_rate": 1.9464285714285715e-05,
+      "loss": 0.2682,
+      "step": 1549
+    },
+    {
+      "epoch": 6.126482213438735,
+      "grad_norm": 3.90091872215271,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.2084,
+      "step": 1550
+    },
+    {
+      "epoch": 6.130434782608695,
+      "grad_norm": 3.1003096103668213,
+      "learning_rate": 1.9424603174603176e-05,
+      "loss": 0.1783,
+      "step": 1551
+    },
+    {
+      "epoch": 6.134387351778656,
+      "grad_norm": 4.112887859344482,
+      "learning_rate": 1.9404761904761906e-05,
+      "loss": 0.2177,
+      "step": 1552
+    },
+    {
+      "epoch": 6.138339920948616,
+      "grad_norm": 2.7363293170928955,
+      "learning_rate": 1.9384920634920637e-05,
+      "loss": 0.1403,
+      "step": 1553
+    },
+    {
+      "epoch": 6.142292490118577,
+      "grad_norm": 3.061474561691284,
+      "learning_rate": 1.9365079365079367e-05,
+      "loss": 0.1204,
+      "step": 1554
+    },
+    {
+      "epoch": 6.146245059288537,
+      "grad_norm": 3.5960800647735596,
+      "learning_rate": 1.9345238095238097e-05,
+      "loss": 0.1865,
+      "step": 1555
+    },
+    {
+      "epoch": 6.150197628458498,
+      "grad_norm": 3.1782121658325195,
+      "learning_rate": 1.9325396825396828e-05,
+      "loss": 0.1481,
+      "step": 1556
+    },
+    {
+      "epoch": 6.154150197628459,
+      "grad_norm": 3.249864339828491,
+      "learning_rate": 1.9305555555555558e-05,
+      "loss": 0.2116,
+      "step": 1557
+    },
+    {
+      "epoch": 6.158102766798419,
+      "grad_norm": 4.330471515655518,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.158,
+      "step": 1558
+    },
+    {
+      "epoch": 6.16205533596838,
+      "grad_norm": 4.093515872955322,
+      "learning_rate": 1.926587301587302e-05,
+      "loss": 0.1871,
+      "step": 1559
+    },
+    {
+      "epoch": 6.16600790513834,
+      "grad_norm": 3.096252918243408,
+      "learning_rate": 1.924603174603175e-05,
+      "loss": 0.1601,
+      "step": 1560
+    },
+    {
+      "epoch": 6.169960474308301,
+      "grad_norm": 3.766211748123169,
+      "learning_rate": 1.922619047619048e-05,
+      "loss": 0.1666,
+      "step": 1561
+    },
+    {
+      "epoch": 6.173913043478261,
+      "grad_norm": 3.036038637161255,
+      "learning_rate": 1.920634920634921e-05,
+      "loss": 0.1687,
+      "step": 1562
+    },
+    {
+      "epoch": 6.177865612648222,
+      "grad_norm": 3.8249239921569824,
+      "learning_rate": 1.918650793650794e-05,
+      "loss": 0.1887,
+      "step": 1563
+    },
+    {
+      "epoch": 6.181818181818182,
+      "grad_norm": 3.461127281188965,
+      "learning_rate": 1.9166666666666667e-05,
+      "loss": 0.1666,
+      "step": 1564
+    },
+    {
+      "epoch": 6.1857707509881426,
+      "grad_norm": 3.262033462524414,
+      "learning_rate": 1.9146825396825398e-05,
+      "loss": 0.1546,
+      "step": 1565
+    },
+    {
+      "epoch": 6.189723320158103,
+      "grad_norm": 3.1641769409179688,
+      "learning_rate": 1.9126984126984128e-05,
+      "loss": 0.1628,
+      "step": 1566
+    },
+    {
+      "epoch": 6.1936758893280635,
+      "grad_norm": 2.8139045238494873,
+      "learning_rate": 1.910714285714286e-05,
+      "loss": 0.1658,
+      "step": 1567
+    },
+    {
+      "epoch": 6.1976284584980235,
+      "grad_norm": 4.408447265625,
+      "learning_rate": 1.9087301587301585e-05,
+      "loss": 0.2102,
+      "step": 1568
+    },
+    {
+      "epoch": 6.201581027667984,
+      "grad_norm": 2.9684667587280273,
+      "learning_rate": 1.9067460317460316e-05,
+      "loss": 0.2325,
+      "step": 1569
+    },
+    {
+      "epoch": 6.205533596837944,
+      "grad_norm": 2.9627156257629395,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.1779,
+      "step": 1570
+    },
+    {
+      "epoch": 6.209486166007905,
+      "grad_norm": 3.4896349906921387,
+      "learning_rate": 1.9027777777777776e-05,
+      "loss": 0.1404,
+      "step": 1571
+    },
+    {
+      "epoch": 6.213438735177865,
+      "grad_norm": 2.7135539054870605,
+      "learning_rate": 1.9007936507936507e-05,
+      "loss": 0.1057,
+      "step": 1572
+    },
+    {
+      "epoch": 6.217391304347826,
+      "grad_norm": 3.230480909347534,
+      "learning_rate": 1.8988095238095237e-05,
+      "loss": 0.1485,
+      "step": 1573
+    },
+    {
+      "epoch": 6.221343873517786,
+      "grad_norm": 2.9966936111450195,
+      "learning_rate": 1.8968253968253968e-05,
+      "loss": 0.1521,
+      "step": 1574
+    },
+    {
+      "epoch": 6.225296442687747,
+      "grad_norm": 2.451078414916992,
+      "learning_rate": 1.8948412698412698e-05,
+      "loss": 0.1289,
+      "step": 1575
+    },
+    {
+      "epoch": 6.229249011857707,
+      "grad_norm": 3.682922124862671,
+      "learning_rate": 1.892857142857143e-05,
+      "loss": 0.1696,
+      "step": 1576
+    },
+    {
+      "epoch": 6.233201581027668,
+      "grad_norm": 4.325542449951172,
+      "learning_rate": 1.890873015873016e-05,
+      "loss": 0.2367,
+      "step": 1577
+    },
+    {
+      "epoch": 6.237154150197629,
+      "grad_norm": 3.799766778945923,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.1636,
+      "step": 1578
+    },
+    {
+      "epoch": 6.241106719367589,
+      "grad_norm": 2.6366379261016846,
+      "learning_rate": 1.886904761904762e-05,
+      "loss": 0.126,
+      "step": 1579
+    },
+    {
+      "epoch": 6.24505928853755,
+      "grad_norm": 3.575343608856201,
+      "learning_rate": 1.884920634920635e-05,
+      "loss": 0.1621,
+      "step": 1580
+    },
+    {
+      "epoch": 6.24901185770751,
+      "grad_norm": 4.663724422454834,
+      "learning_rate": 1.882936507936508e-05,
+      "loss": 0.2898,
+      "step": 1581
+    },
+    {
+      "epoch": 6.252964426877471,
+      "grad_norm": 3.3895034790039062,
+      "learning_rate": 1.880952380952381e-05,
+      "loss": 0.2493,
+      "step": 1582
+    },
+    {
+      "epoch": 6.256916996047431,
+      "grad_norm": 4.064142227172852,
+      "learning_rate": 1.878968253968254e-05,
+      "loss": 0.2136,
+      "step": 1583
+    },
+    {
+      "epoch": 6.260869565217392,
+      "grad_norm": 4.431739330291748,
+      "learning_rate": 1.876984126984127e-05,
+      "loss": 0.168,
+      "step": 1584
+    },
+    {
+      "epoch": 6.264822134387352,
+      "grad_norm": 3.109384536743164,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.1606,
+      "step": 1585
+    },
+    {
+      "epoch": 6.2687747035573125,
+      "grad_norm": 3.7452895641326904,
+      "learning_rate": 1.8730158730158732e-05,
+      "loss": 0.2287,
+      "step": 1586
+    },
+    {
+      "epoch": 6.2727272727272725,
+      "grad_norm": 3.320481777191162,
+      "learning_rate": 1.8710317460317462e-05,
+      "loss": 0.1221,
+      "step": 1587
+    },
+    {
+      "epoch": 6.276679841897233,
+      "grad_norm": 3.2335190773010254,
+      "learning_rate": 1.8690476190476193e-05,
+      "loss": 0.1522,
+      "step": 1588
+    },
+    {
+      "epoch": 6.280632411067193,
+      "grad_norm": 2.431941509246826,
+      "learning_rate": 1.8670634920634923e-05,
+      "loss": 0.1787,
+      "step": 1589
+    },
+    {
+      "epoch": 6.284584980237154,
+      "grad_norm": 2.7418622970581055,
+      "learning_rate": 1.8650793650793654e-05,
+      "loss": 0.1862,
+      "step": 1590
+    },
+    {
+      "epoch": 6.288537549407114,
+      "grad_norm": 3.3152060508728027,
+      "learning_rate": 1.863095238095238e-05,
+      "loss": 0.1879,
+      "step": 1591
+    },
+    {
+      "epoch": 6.292490118577075,
+      "grad_norm": 2.946201801300049,
+      "learning_rate": 1.861111111111111e-05,
+      "loss": 0.1429,
+      "step": 1592
+    },
+    {
+      "epoch": 6.296442687747035,
+      "grad_norm": 4.4527587890625,
+      "learning_rate": 1.859126984126984e-05,
+      "loss": 0.1634,
+      "step": 1593
+    },
+    {
+      "epoch": 6.300395256916996,
+      "grad_norm": 2.763129711151123,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.1338,
+      "step": 1594
+    },
+    {
+      "epoch": 6.304347826086957,
+      "grad_norm": 3.8917641639709473,
+      "learning_rate": 1.8551587301587302e-05,
+      "loss": 0.1526,
+      "step": 1595
+    },
+    {
+      "epoch": 6.308300395256917,
+      "grad_norm": 3.10577130317688,
+      "learning_rate": 1.8531746031746032e-05,
+      "loss": 0.1993,
+      "step": 1596
+    },
+    {
+      "epoch": 6.312252964426877,
+      "grad_norm": 4.823581695556641,
+      "learning_rate": 1.8511904761904763e-05,
+      "loss": 0.1888,
+      "step": 1597
+    },
+    {
+      "epoch": 6.316205533596838,
+      "grad_norm": 3.8852086067199707,
+      "learning_rate": 1.8492063492063493e-05,
+      "loss": 0.1369,
+      "step": 1598
+    },
+    {
+      "epoch": 6.320158102766799,
+      "grad_norm": 3.8083715438842773,
+      "learning_rate": 1.8472222222222224e-05,
+      "loss": 0.1808,
+      "step": 1599
+    },
+    {
+      "epoch": 6.324110671936759,
+      "grad_norm": 3.3586747646331787,
+      "learning_rate": 1.8452380952380954e-05,
+      "loss": 0.1479,
+      "step": 1600
+    },
+    {
+      "epoch": 6.32806324110672,
+      "grad_norm": 2.599363088607788,
+      "learning_rate": 1.8432539682539684e-05,
+      "loss": 0.1342,
+      "step": 1601
+    },
+    {
+      "epoch": 6.33201581027668,
+      "grad_norm": 3.6683595180511475,
+      "learning_rate": 1.8412698412698415e-05,
+      "loss": 0.1694,
+      "step": 1602
+    },
+    {
+      "epoch": 6.335968379446641,
+      "grad_norm": 2.9072253704071045,
+      "learning_rate": 1.8392857142857145e-05,
+      "loss": 0.1598,
+      "step": 1603
+    },
+    {
+      "epoch": 6.339920948616601,
+      "grad_norm": 4.043822288513184,
+      "learning_rate": 1.8373015873015875e-05,
+      "loss": 0.2375,
+      "step": 1604
+    },
+    {
+      "epoch": 6.3438735177865615,
+      "grad_norm": 4.241188049316406,
+      "learning_rate": 1.8353174603174602e-05,
+      "loss": 0.2391,
+      "step": 1605
+    },
+    {
+      "epoch": 6.3478260869565215,
+      "grad_norm": 3.922586441040039,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.2149,
+      "step": 1606
+    },
+    {
+      "epoch": 6.351778656126482,
+      "grad_norm": 4.914572715759277,
+      "learning_rate": 1.8313492063492063e-05,
+      "loss": 0.2859,
+      "step": 1607
+    },
+    {
+      "epoch": 6.355731225296442,
+      "grad_norm": 3.6631100177764893,
+      "learning_rate": 1.8293650793650794e-05,
+      "loss": 0.2134,
+      "step": 1608
+    },
+    {
+      "epoch": 6.359683794466403,
+      "grad_norm": 2.8764913082122803,
+      "learning_rate": 1.8273809523809524e-05,
+      "loss": 0.1453,
+      "step": 1609
+    },
+    {
+      "epoch": 6.363636363636363,
+      "grad_norm": 2.821441173553467,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.1457,
+      "step": 1610
+    },
+    {
+      "epoch": 6.367588932806324,
+      "grad_norm": 2.9946677684783936,
+      "learning_rate": 1.8234126984126985e-05,
+      "loss": 0.144,
+      "step": 1611
+    },
+    {
+      "epoch": 6.371541501976284,
+      "grad_norm": 2.6403133869171143,
+      "learning_rate": 1.8214285714285715e-05,
+      "loss": 0.1484,
+      "step": 1612
+    },
+    {
+      "epoch": 6.375494071146245,
+      "grad_norm": 2.8457889556884766,
+      "learning_rate": 1.8194444444444445e-05,
+      "loss": 0.1816,
+      "step": 1613
+    },
+    {
+      "epoch": 6.379446640316205,
+      "grad_norm": 2.93542742729187,
+      "learning_rate": 1.8174603174603176e-05,
+      "loss": 0.1347,
+      "step": 1614
+    },
+    {
+      "epoch": 6.383399209486166,
+      "grad_norm": 2.890878915786743,
+      "learning_rate": 1.8154761904761906e-05,
+      "loss": 0.1245,
+      "step": 1615
+    },
+    {
+      "epoch": 6.387351778656127,
+      "grad_norm": 2.7771224975585938,
+      "learning_rate": 1.8134920634920637e-05,
+      "loss": 0.1573,
+      "step": 1616
+    },
+    {
+      "epoch": 6.391304347826087,
+      "grad_norm": 3.317066192626953,
+      "learning_rate": 1.8115079365079367e-05,
+      "loss": 0.1826,
+      "step": 1617
+    },
+    {
+      "epoch": 6.395256916996048,
+      "grad_norm": 3.2389166355133057,
+      "learning_rate": 1.8095238095238094e-05,
+      "loss": 0.1539,
+      "step": 1618
+    },
+    {
+      "epoch": 6.399209486166008,
+      "grad_norm": 3.585541009902954,
+      "learning_rate": 1.8075396825396824e-05,
+      "loss": 0.1731,
+      "step": 1619
+    },
+    {
+      "epoch": 6.403162055335969,
+      "grad_norm": 3.805875778198242,
+      "learning_rate": 1.8055555555555555e-05,
+      "loss": 0.2165,
+      "step": 1620
+    },
+    {
+      "epoch": 6.407114624505929,
+      "grad_norm": 4.0997090339660645,
+      "learning_rate": 1.8035714285714285e-05,
+      "loss": 0.178,
+      "step": 1621
+    },
+    {
+      "epoch": 6.41106719367589,
+      "grad_norm": 3.204512596130371,
+      "learning_rate": 1.8015873015873015e-05,
+      "loss": 0.1188,
+      "step": 1622
+    },
+    {
+      "epoch": 6.41501976284585,
+      "grad_norm": 3.9086551666259766,
+      "learning_rate": 1.7996031746031746e-05,
+      "loss": 0.1424,
+      "step": 1623
+    },
+    {
+      "epoch": 6.4189723320158105,
+      "grad_norm": 3.208350658416748,
+      "learning_rate": 1.7976190476190476e-05,
+      "loss": 0.1119,
+      "step": 1624
+    },
+    {
+      "epoch": 6.4229249011857705,
+      "grad_norm": 4.048145771026611,
+      "learning_rate": 1.7956349206349207e-05,
+      "loss": 0.2034,
+      "step": 1625
+    },
+    {
+      "epoch": 6.426877470355731,
+      "grad_norm": 3.807363986968994,
+      "learning_rate": 1.7936507936507937e-05,
+      "loss": 0.1894,
+      "step": 1626
+    },
+    {
+      "epoch": 6.430830039525691,
+      "grad_norm": 3.468174457550049,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 0.1755,
+      "step": 1627
+    },
+    {
+      "epoch": 6.434782608695652,
+      "grad_norm": 3.572981595993042,
+      "learning_rate": 1.7896825396825398e-05,
+      "loss": 0.2043,
+      "step": 1628
+    },
+    {
+      "epoch": 6.438735177865612,
+      "grad_norm": 4.217936992645264,
+      "learning_rate": 1.7876984126984128e-05,
+      "loss": 0.3312,
+      "step": 1629
+    },
+    {
+      "epoch": 6.442687747035573,
+      "grad_norm": 3.2552435398101807,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.1874,
+      "step": 1630
+    },
+    {
+      "epoch": 6.446640316205533,
+      "grad_norm": 3.2607510089874268,
+      "learning_rate": 1.783730158730159e-05,
+      "loss": 0.1648,
+      "step": 1631
+    },
+    {
+      "epoch": 6.450592885375494,
+      "grad_norm": 4.080394268035889,
+      "learning_rate": 1.781746031746032e-05,
+      "loss": 0.1779,
+      "step": 1632
+    },
+    {
+      "epoch": 6.454545454545454,
+      "grad_norm": 3.7420568466186523,
+      "learning_rate": 1.779761904761905e-05,
+      "loss": 0.161,
+      "step": 1633
+    },
+    {
+      "epoch": 6.458498023715415,
+      "grad_norm": 3.199740171432495,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.1783,
+      "step": 1634
+    },
+    {
+      "epoch": 6.462450592885375,
+      "grad_norm": 3.5963351726531982,
+      "learning_rate": 1.775793650793651e-05,
+      "loss": 0.1577,
+      "step": 1635
+    },
+    {
+      "epoch": 6.466403162055336,
+      "grad_norm": 2.5002620220184326,
+      "learning_rate": 1.773809523809524e-05,
+      "loss": 0.1206,
+      "step": 1636
+    },
+    {
+      "epoch": 6.470355731225297,
+      "grad_norm": 3.222224235534668,
+      "learning_rate": 1.771825396825397e-05,
+      "loss": 0.1532,
+      "step": 1637
+    },
+    {
+      "epoch": 6.474308300395257,
+      "grad_norm": 2.948617935180664,
+      "learning_rate": 1.76984126984127e-05,
+      "loss": 0.1807,
+      "step": 1638
+    },
+    {
+      "epoch": 6.478260869565218,
+      "grad_norm": 3.9954729080200195,
+      "learning_rate": 1.7678571428571432e-05,
+      "loss": 0.1606,
+      "step": 1639
+    },
+    {
+      "epoch": 6.482213438735178,
+      "grad_norm": 3.8689181804656982,
+      "learning_rate": 1.7658730158730162e-05,
+      "loss": 0.188,
+      "step": 1640
+    },
+    {
+      "epoch": 6.486166007905139,
+      "grad_norm": 2.888376235961914,
+      "learning_rate": 1.763888888888889e-05,
+      "loss": 0.1553,
+      "step": 1641
+    },
+    {
+      "epoch": 6.490118577075099,
+      "grad_norm": 3.6249587535858154,
+      "learning_rate": 1.761904761904762e-05,
+      "loss": 0.178,
+      "step": 1642
+    },
+    {
+      "epoch": 6.4940711462450595,
+      "grad_norm": 2.787454843521118,
+      "learning_rate": 1.759920634920635e-05,
+      "loss": 0.129,
+      "step": 1643
+    },
+    {
+      "epoch": 6.4980237154150196,
+      "grad_norm": 4.260427474975586,
+      "learning_rate": 1.757936507936508e-05,
+      "loss": 0.2073,
+      "step": 1644
+    },
+    {
+      "epoch": 6.5019762845849804,
+      "grad_norm": 4.095132827758789,
+      "learning_rate": 1.755952380952381e-05,
+      "loss": 0.1609,
+      "step": 1645
+    },
+    {
+      "epoch": 6.5059288537549405,
+      "grad_norm": 3.89748477935791,
+      "learning_rate": 1.7539682539682538e-05,
+      "loss": 0.2223,
+      "step": 1646
+    },
+    {
+      "epoch": 6.509881422924901,
+      "grad_norm": 3.197842597961426,
+      "learning_rate": 1.7519841269841268e-05,
+      "loss": 0.1804,
+      "step": 1647
+    },
+    {
+      "epoch": 6.513833992094861,
+      "grad_norm": 3.7351438999176025,
+      "learning_rate": 1.75e-05,
+      "loss": 0.1659,
+      "step": 1648
+    },
+    {
+      "epoch": 6.517786561264822,
+      "grad_norm": 3.417168378829956,
+      "learning_rate": 1.748015873015873e-05,
+      "loss": 0.1493,
+      "step": 1649
+    },
+    {
+      "epoch": 6.521739130434782,
+      "grad_norm": 2.904148578643799,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.1689,
+      "step": 1650
+    },
+    {
+      "epoch": 6.525691699604743,
+      "grad_norm": 3.141493558883667,
+      "learning_rate": 1.744047619047619e-05,
+      "loss": 0.1433,
+      "step": 1651
+    },
+    {
+      "epoch": 6.529644268774703,
+      "grad_norm": 3.594679832458496,
+      "learning_rate": 1.742063492063492e-05,
+      "loss": 0.2365,
+      "step": 1652
+    },
+    {
+      "epoch": 6.533596837944664,
+      "grad_norm": 3.7037243843078613,
+      "learning_rate": 1.740079365079365e-05,
+      "loss": 0.2317,
+      "step": 1653
+    },
+    {
+      "epoch": 6.537549407114625,
+      "grad_norm": 3.6561882495880127,
+      "learning_rate": 1.738095238095238e-05,
+      "loss": 0.1495,
+      "step": 1654
+    },
+    {
+      "epoch": 6.541501976284585,
+      "grad_norm": 3.278259754180908,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 0.1314,
+      "step": 1655
+    },
+    {
+      "epoch": 6.545454545454545,
+      "grad_norm": 3.1691975593566895,
+      "learning_rate": 1.734126984126984e-05,
+      "loss": 0.1566,
+      "step": 1656
+    },
+    {
+      "epoch": 6.549407114624506,
+      "grad_norm": 3.14813494682312,
+      "learning_rate": 1.7321428571428572e-05,
+      "loss": 0.1585,
+      "step": 1657
+    },
+    {
+      "epoch": 6.553359683794467,
+      "grad_norm": 3.2934324741363525,
+      "learning_rate": 1.7301587301587302e-05,
+      "loss": 0.1568,
+      "step": 1658
+    },
+    {
+      "epoch": 6.557312252964427,
+      "grad_norm": 4.18383264541626,
+      "learning_rate": 1.7281746031746033e-05,
+      "loss": 0.3083,
+      "step": 1659
+    },
+    {
+      "epoch": 6.561264822134388,
+      "grad_norm": 4.086765766143799,
+      "learning_rate": 1.7261904761904763e-05,
+      "loss": 0.1687,
+      "step": 1660
+    },
+    {
+      "epoch": 6.565217391304348,
+      "grad_norm": 3.70186185836792,
+      "learning_rate": 1.7242063492063493e-05,
+      "loss": 0.1741,
+      "step": 1661
+    },
+    {
+      "epoch": 6.569169960474309,
+      "grad_norm": 3.012298583984375,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 0.1249,
+      "step": 1662
+    },
+    {
+      "epoch": 6.573122529644269,
+      "grad_norm": 5.418375492095947,
+      "learning_rate": 1.7202380952380954e-05,
+      "loss": 0.328,
+      "step": 1663
+    },
+    {
+      "epoch": 6.5770750988142295,
+      "grad_norm": 2.400665521621704,
+      "learning_rate": 1.7182539682539684e-05,
+      "loss": 0.1125,
+      "step": 1664
+    },
+    {
+      "epoch": 6.5810276679841895,
+      "grad_norm": 3.7247157096862793,
+      "learning_rate": 1.7162698412698415e-05,
+      "loss": 0.1577,
+      "step": 1665
+    },
+    {
+      "epoch": 6.58498023715415,
+      "grad_norm": 2.6763904094696045,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.1269,
+      "step": 1666
+    },
+    {
+      "epoch": 6.58893280632411,
+      "grad_norm": 2.851741313934326,
+      "learning_rate": 1.7123015873015876e-05,
+      "loss": 0.141,
+      "step": 1667
+    },
+    {
+      "epoch": 6.592885375494071,
+      "grad_norm": 3.190314769744873,
+      "learning_rate": 1.7103174603174606e-05,
+      "loss": 0.2027,
+      "step": 1668
+    },
+    {
+      "epoch": 6.596837944664031,
+      "grad_norm": 3.6213324069976807,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 0.2126,
+      "step": 1669
+    },
+    {
+      "epoch": 6.600790513833992,
+      "grad_norm": 3.6748242378234863,
+      "learning_rate": 1.7063492063492063e-05,
+      "loss": 0.1946,
+      "step": 1670
+    },
+    {
+      "epoch": 6.604743083003952,
+      "grad_norm": 2.7625160217285156,
+      "learning_rate": 1.7043650793650794e-05,
+      "loss": 0.1886,
+      "step": 1671
+    },
+    {
+      "epoch": 6.608695652173913,
+      "grad_norm": 3.2063212394714355,
+      "learning_rate": 1.7023809523809524e-05,
+      "loss": 0.1705,
+      "step": 1672
+    },
+    {
+      "epoch": 6.612648221343873,
+      "grad_norm": 3.1696741580963135,
+      "learning_rate": 1.7003968253968254e-05,
+      "loss": 0.1476,
+      "step": 1673
+    },
+    {
+      "epoch": 6.616600790513834,
+      "grad_norm": 2.514961004257202,
+      "learning_rate": 1.6984126984126985e-05,
+      "loss": 0.1324,
+      "step": 1674
+    },
+    {
+      "epoch": 6.620553359683795,
+      "grad_norm": 3.8006770610809326,
+      "learning_rate": 1.6964285714285715e-05,
+      "loss": 0.2031,
+      "step": 1675
+    },
+    {
+      "epoch": 6.624505928853755,
+      "grad_norm": 2.094867706298828,
+      "learning_rate": 1.6944444444444446e-05,
+      "loss": 0.0909,
+      "step": 1676
+    },
+    {
+      "epoch": 6.628458498023716,
+      "grad_norm": 5.729726314544678,
+      "learning_rate": 1.6924603174603176e-05,
+      "loss": 0.2006,
+      "step": 1677
+    },
+    {
+      "epoch": 6.632411067193676,
+      "grad_norm": 4.788626194000244,
+      "learning_rate": 1.6904761904761906e-05,
+      "loss": 0.188,
+      "step": 1678
+    },
+    {
+      "epoch": 6.636363636363637,
+      "grad_norm": 3.7394933700561523,
+      "learning_rate": 1.6884920634920637e-05,
+      "loss": 0.1474,
+      "step": 1679
+    },
+    {
+      "epoch": 6.640316205533597,
+      "grad_norm": 3.15619158744812,
+      "learning_rate": 1.6865079365079367e-05,
+      "loss": 0.1226,
+      "step": 1680
+    },
+    {
+      "epoch": 6.644268774703558,
+      "grad_norm": 3.1773312091827393,
+      "learning_rate": 1.6845238095238097e-05,
+      "loss": 0.1683,
+      "step": 1681
+    },
+    {
+      "epoch": 6.648221343873518,
+      "grad_norm": 3.101545572280884,
+      "learning_rate": 1.6825396825396828e-05,
+      "loss": 0.1786,
+      "step": 1682
+    },
+    {
+      "epoch": 6.6521739130434785,
+      "grad_norm": 2.9866223335266113,
+      "learning_rate": 1.6805555555555558e-05,
+      "loss": 0.1385,
+      "step": 1683
+    },
+    {
+      "epoch": 6.6561264822134385,
+      "grad_norm": 2.350311517715454,
+      "learning_rate": 1.6785714285714285e-05,
+      "loss": 0.1332,
+      "step": 1684
+    },
+    {
+      "epoch": 6.660079051383399,
+      "grad_norm": 3.1310527324676514,
+      "learning_rate": 1.6765873015873016e-05,
+      "loss": 0.1336,
+      "step": 1685
+    },
+    {
+      "epoch": 6.664031620553359,
+      "grad_norm": 3.0900039672851562,
+      "learning_rate": 1.6746031746031746e-05,
+      "loss": 0.1304,
+      "step": 1686
+    },
+    {
+      "epoch": 6.66798418972332,
+      "grad_norm": 3.2894325256347656,
+      "learning_rate": 1.6726190476190476e-05,
+      "loss": 0.1869,
+      "step": 1687
+    },
+    {
+      "epoch": 6.67193675889328,
+      "grad_norm": 3.1792759895324707,
+      "learning_rate": 1.6706349206349207e-05,
+      "loss": 0.145,
+      "step": 1688
+    },
+    {
+      "epoch": 6.675889328063241,
+      "grad_norm": 2.9144246578216553,
+      "learning_rate": 1.6686507936507937e-05,
+      "loss": 0.1617,
+      "step": 1689
+    },
+    {
+      "epoch": 6.679841897233201,
+      "grad_norm": 3.1634509563446045,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.159,
+      "step": 1690
+    },
+    {
+      "epoch": 6.683794466403162,
+      "grad_norm": 3.328819990158081,
+      "learning_rate": 1.6646825396825398e-05,
+      "loss": 0.2096,
+      "step": 1691
+    },
+    {
+      "epoch": 6.687747035573123,
+      "grad_norm": 3.085995674133301,
+      "learning_rate": 1.6626984126984128e-05,
+      "loss": 0.1649,
+      "step": 1692
+    },
+    {
+      "epoch": 6.691699604743083,
+      "grad_norm": 2.8405370712280273,
+      "learning_rate": 1.660714285714286e-05,
+      "loss": 0.1666,
+      "step": 1693
+    },
+    {
+      "epoch": 6.695652173913043,
+      "grad_norm": 3.161557912826538,
+      "learning_rate": 1.658730158730159e-05,
+      "loss": 0.1811,
+      "step": 1694
+    },
+    {
+      "epoch": 6.699604743083004,
+      "grad_norm": 4.0587158203125,
+      "learning_rate": 1.656746031746032e-05,
+      "loss": 0.2273,
+      "step": 1695
+    },
+    {
+      "epoch": 6.703557312252965,
+      "grad_norm": 3.1666932106018066,
+      "learning_rate": 1.6547619047619046e-05,
+      "loss": 0.2019,
+      "step": 1696
+    },
+    {
+      "epoch": 6.707509881422925,
+      "grad_norm": 4.08314847946167,
+      "learning_rate": 1.6527777777777777e-05,
+      "loss": 0.2561,
+      "step": 1697
+    },
+    {
+      "epoch": 6.711462450592886,
+      "grad_norm": 3.302401065826416,
+      "learning_rate": 1.6507936507936507e-05,
+      "loss": 0.1543,
+      "step": 1698
+    },
+    {
+      "epoch": 6.715415019762846,
+      "grad_norm": 3.8679957389831543,
+      "learning_rate": 1.6488095238095237e-05,
+      "loss": 0.2805,
+      "step": 1699
+    },
+    {
+      "epoch": 6.719367588932807,
+      "grad_norm": 2.699409008026123,
+      "learning_rate": 1.6468253968253968e-05,
+      "loss": 0.1732,
+      "step": 1700
+    },
+    {
+      "epoch": 6.723320158102767,
+      "grad_norm": 4.578848838806152,
+      "learning_rate": 1.6448412698412698e-05,
+      "loss": 0.1573,
+      "step": 1701
+    },
+    {
+      "epoch": 6.7272727272727275,
+      "grad_norm": 4.305120468139648,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.1459,
+      "step": 1702
+    },
+    {
+      "epoch": 6.7312252964426875,
+      "grad_norm": 2.981783628463745,
+      "learning_rate": 1.640873015873016e-05,
+      "loss": 0.1605,
+      "step": 1703
+    },
+    {
+      "epoch": 6.735177865612648,
+      "grad_norm": 3.4200551509857178,
+      "learning_rate": 1.638888888888889e-05,
+      "loss": 0.1136,
+      "step": 1704
+    },
+    {
+      "epoch": 6.739130434782608,
+      "grad_norm": 3.0042500495910645,
+      "learning_rate": 1.636904761904762e-05,
+      "loss": 0.1645,
+      "step": 1705
+    },
+    {
+      "epoch": 6.743083003952569,
+      "grad_norm": 5.0387043952941895,
+      "learning_rate": 1.634920634920635e-05,
+      "loss": 0.3057,
+      "step": 1706
+    },
+    {
+      "epoch": 6.747035573122529,
+      "grad_norm": 3.7466354370117188,
+      "learning_rate": 1.632936507936508e-05,
+      "loss": 0.1289,
+      "step": 1707
+    },
+    {
+      "epoch": 6.75098814229249,
+      "grad_norm": 2.57358980178833,
+      "learning_rate": 1.630952380952381e-05,
+      "loss": 0.1363,
+      "step": 1708
+    },
+    {
+      "epoch": 6.75494071146245,
+      "grad_norm": 3.13360333442688,
+      "learning_rate": 1.628968253968254e-05,
+      "loss": 0.1979,
+      "step": 1709
+    },
+    {
+      "epoch": 6.758893280632411,
+      "grad_norm": 3.5185632705688477,
+      "learning_rate": 1.626984126984127e-05,
+      "loss": 0.1732,
+      "step": 1710
+    },
+    {
+      "epoch": 6.762845849802371,
+      "grad_norm": 4.430140018463135,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.1609,
+      "step": 1711
+    },
+    {
+      "epoch": 6.766798418972332,
+      "grad_norm": 3.1439976692199707,
+      "learning_rate": 1.6230158730158732e-05,
+      "loss": 0.1765,
+      "step": 1712
+    },
+    {
+      "epoch": 6.770750988142293,
+      "grad_norm": 2.766545534133911,
+      "learning_rate": 1.6210317460317463e-05,
+      "loss": 0.173,
+      "step": 1713
+    },
+    {
+      "epoch": 6.774703557312253,
+      "grad_norm": 3.4417989253997803,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.1748,
+      "step": 1714
+    },
+    {
+      "epoch": 6.778656126482213,
+      "grad_norm": 3.091768503189087,
+      "learning_rate": 1.6170634920634923e-05,
+      "loss": 0.1947,
+      "step": 1715
+    },
+    {
+      "epoch": 6.782608695652174,
+      "grad_norm": 2.7375543117523193,
+      "learning_rate": 1.6150793650793654e-05,
+      "loss": 0.1786,
+      "step": 1716
+    },
+    {
+      "epoch": 6.786561264822135,
+      "grad_norm": 3.463697671890259,
+      "learning_rate": 1.6130952380952384e-05,
+      "loss": 0.2595,
+      "step": 1717
+    },
+    {
+      "epoch": 6.790513833992095,
+      "grad_norm": 3.0348737239837646,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 0.123,
+      "step": 1718
+    },
+    {
+      "epoch": 6.794466403162056,
+      "grad_norm": 3.646333694458008,
+      "learning_rate": 1.609126984126984e-05,
+      "loss": 0.1529,
+      "step": 1719
+    },
+    {
+      "epoch": 6.798418972332016,
+      "grad_norm": 4.054477691650391,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.1475,
+      "step": 1720
+    },
+    {
+      "epoch": 6.8023715415019765,
+      "grad_norm": 3.619706869125366,
+      "learning_rate": 1.6051587301587302e-05,
+      "loss": 0.1682,
+      "step": 1721
+    },
+    {
+      "epoch": 6.8063241106719365,
+      "grad_norm": 3.0783016681671143,
+      "learning_rate": 1.6031746031746033e-05,
+      "loss": 0.1772,
+      "step": 1722
+    },
+    {
+      "epoch": 6.810276679841897,
+      "grad_norm": 2.9383182525634766,
+      "learning_rate": 1.601190476190476e-05,
+      "loss": 0.149,
+      "step": 1723
+    },
+    {
+      "epoch": 6.8142292490118574,
+      "grad_norm": 3.6608943939208984,
+      "learning_rate": 1.599206349206349e-05,
+      "loss": 0.152,
+      "step": 1724
+    },
+    {
+      "epoch": 6.818181818181818,
+      "grad_norm": 3.396385908126831,
+      "learning_rate": 1.597222222222222e-05,
+      "loss": 0.1656,
+      "step": 1725
+    },
+    {
+      "epoch": 6.822134387351778,
+      "grad_norm": 2.923689126968384,
+      "learning_rate": 1.595238095238095e-05,
+      "loss": 0.1529,
+      "step": 1726
+    },
+    {
+      "epoch": 6.826086956521739,
+      "grad_norm": 4.30761194229126,
+      "learning_rate": 1.593253968253968e-05,
+      "loss": 0.3032,
+      "step": 1727
+    },
+    {
+      "epoch": 6.830039525691699,
+      "grad_norm": 3.8314170837402344,
+      "learning_rate": 1.591269841269841e-05,
+      "loss": 0.1874,
+      "step": 1728
+    },
+    {
+      "epoch": 6.83399209486166,
+      "grad_norm": 2.5451226234436035,
+      "learning_rate": 1.5892857142857142e-05,
+      "loss": 0.1155,
+      "step": 1729
+    },
+    {
+      "epoch": 6.837944664031621,
+      "grad_norm": 3.4286904335021973,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.2253,
+      "step": 1730
+    },
+    {
+      "epoch": 6.841897233201581,
+      "grad_norm": 3.8259549140930176,
+      "learning_rate": 1.5853174603174603e-05,
+      "loss": 0.2408,
+      "step": 1731
+    },
+    {
+      "epoch": 6.845849802371541,
+      "grad_norm": 3.6046714782714844,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 0.1758,
+      "step": 1732
+    },
+    {
+      "epoch": 6.849802371541502,
+      "grad_norm": 3.0990259647369385,
+      "learning_rate": 1.5813492063492063e-05,
+      "loss": 0.1622,
+      "step": 1733
+    },
+    {
+      "epoch": 6.853754940711463,
+      "grad_norm": 2.931340217590332,
+      "learning_rate": 1.5793650793650794e-05,
+      "loss": 0.1598,
+      "step": 1734
+    },
+    {
+      "epoch": 6.857707509881423,
+      "grad_norm": 2.4774787425994873,
+      "learning_rate": 1.5773809523809524e-05,
+      "loss": 0.141,
+      "step": 1735
+    },
+    {
+      "epoch": 6.861660079051384,
+      "grad_norm": 2.9939541816711426,
+      "learning_rate": 1.5753968253968255e-05,
+      "loss": 0.135,
+      "step": 1736
+    },
+    {
+      "epoch": 6.865612648221344,
+      "grad_norm": 2.929865598678589,
+      "learning_rate": 1.5734126984126985e-05,
+      "loss": 0.1947,
+      "step": 1737
+    },
+    {
+      "epoch": 6.869565217391305,
+      "grad_norm": 3.4848814010620117,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.1808,
+      "step": 1738
+    },
+    {
+      "epoch": 6.873517786561265,
+      "grad_norm": 2.8920042514801025,
+      "learning_rate": 1.5694444444444446e-05,
+      "loss": 0.102,
+      "step": 1739
+    },
+    {
+      "epoch": 6.877470355731226,
+      "grad_norm": 4.020946979522705,
+      "learning_rate": 1.5674603174603176e-05,
+      "loss": 0.1483,
+      "step": 1740
+    },
+    {
+      "epoch": 6.881422924901186,
+      "grad_norm": 3.165574073791504,
+      "learning_rate": 1.5654761904761906e-05,
+      "loss": 0.1437,
+      "step": 1741
+    },
+    {
+      "epoch": 6.8853754940711465,
+      "grad_norm": 3.595200538635254,
+      "learning_rate": 1.5634920634920637e-05,
+      "loss": 0.1561,
+      "step": 1742
+    },
+    {
+      "epoch": 6.8893280632411065,
+      "grad_norm": 2.952357530593872,
+      "learning_rate": 1.5615079365079367e-05,
+      "loss": 0.1786,
+      "step": 1743
+    },
+    {
+      "epoch": 6.893280632411067,
+      "grad_norm": 3.0202043056488037,
+      "learning_rate": 1.5595238095238098e-05,
+      "loss": 0.1493,
+      "step": 1744
+    },
+    {
+      "epoch": 6.897233201581027,
+      "grad_norm": 3.46177339553833,
+      "learning_rate": 1.5575396825396828e-05,
+      "loss": 0.1578,
+      "step": 1745
+    },
+    {
+      "epoch": 6.901185770750988,
+      "grad_norm": 4.415173053741455,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.205,
+      "step": 1746
+    },
+    {
+      "epoch": 6.905138339920948,
+      "grad_norm": 3.0862183570861816,
+      "learning_rate": 1.5535714285714285e-05,
+      "loss": 0.1837,
+      "step": 1747
+    },
+    {
+      "epoch": 6.909090909090909,
+      "grad_norm": 3.788036584854126,
+      "learning_rate": 1.5515873015873016e-05,
+      "loss": 0.1423,
+      "step": 1748
+    },
+    {
+      "epoch": 6.913043478260869,
+      "grad_norm": 3.3133058547973633,
+      "learning_rate": 1.5496031746031746e-05,
+      "loss": 0.2186,
+      "step": 1749
+    },
+    {
+      "epoch": 6.91699604743083,
+      "grad_norm": 4.402420520782471,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 0.1938,
+      "step": 1750
+    },
+    {
+      "epoch": 6.920948616600791,
+      "grad_norm": 3.363860607147217,
+      "learning_rate": 1.5456349206349207e-05,
+      "loss": 0.1936,
+      "step": 1751
+    },
+    {
+      "epoch": 6.924901185770751,
+      "grad_norm": 3.4065957069396973,
+      "learning_rate": 1.5436507936507937e-05,
+      "loss": 0.174,
+      "step": 1752
+    },
+    {
+      "epoch": 6.928853754940711,
+      "grad_norm": 2.393531322479248,
+      "learning_rate": 1.5416666666666668e-05,
+      "loss": 0.1511,
+      "step": 1753
+    },
+    {
+      "epoch": 6.932806324110672,
+      "grad_norm": 3.180001974105835,
+      "learning_rate": 1.5396825396825398e-05,
+      "loss": 0.1551,
+      "step": 1754
+    },
+    {
+      "epoch": 6.936758893280633,
+      "grad_norm": 2.9649102687835693,
+      "learning_rate": 1.537698412698413e-05,
+      "loss": 0.1588,
+      "step": 1755
+    },
+    {
+      "epoch": 6.940711462450593,
+      "grad_norm": 3.6014418601989746,
+      "learning_rate": 1.535714285714286e-05,
+      "loss": 0.3112,
+      "step": 1756
+    },
+    {
+      "epoch": 6.944664031620554,
+      "grad_norm": 3.6775028705596924,
+      "learning_rate": 1.533730158730159e-05,
+      "loss": 0.2638,
+      "step": 1757
+    },
+    {
+      "epoch": 6.948616600790514,
+      "grad_norm": 2.682131052017212,
+      "learning_rate": 1.531746031746032e-05,
+      "loss": 0.1377,
+      "step": 1758
+    },
+    {
+      "epoch": 6.952569169960475,
+      "grad_norm": 3.296804904937744,
+      "learning_rate": 1.529761904761905e-05,
+      "loss": 0.1325,
+      "step": 1759
+    },
+    {
+      "epoch": 6.956521739130435,
+      "grad_norm": 3.019824743270874,
+      "learning_rate": 1.527777777777778e-05,
+      "loss": 0.1358,
+      "step": 1760
+    },
+    {
+      "epoch": 6.9604743083003955,
+      "grad_norm": 3.4432945251464844,
+      "learning_rate": 1.525793650793651e-05,
+      "loss": 0.1674,
+      "step": 1761
+    },
+    {
+      "epoch": 6.9644268774703555,
+      "grad_norm": 4.434787273406982,
+      "learning_rate": 1.5238095238095241e-05,
+      "loss": 0.2198,
+      "step": 1762
+    },
+    {
+      "epoch": 6.968379446640316,
+      "grad_norm": 2.954920530319214,
+      "learning_rate": 1.5218253968253968e-05,
+      "loss": 0.1315,
+      "step": 1763
+    },
+    {
+      "epoch": 6.972332015810276,
+      "grad_norm": 3.6730740070343018,
+      "learning_rate": 1.5198412698412698e-05,
+      "loss": 0.1586,
+      "step": 1764
+    },
+    {
+      "epoch": 6.976284584980237,
+      "grad_norm": 4.036146640777588,
+      "learning_rate": 1.5178571428571429e-05,
+      "loss": 0.2249,
+      "step": 1765
+    },
+    {
+      "epoch": 6.980237154150197,
+      "grad_norm": 3.2338225841522217,
+      "learning_rate": 1.5158730158730159e-05,
+      "loss": 0.138,
+      "step": 1766
+    },
+    {
+      "epoch": 6.984189723320158,
+      "grad_norm": 3.692736864089966,
+      "learning_rate": 1.5138888888888888e-05,
+      "loss": 0.1937,
+      "step": 1767
+    },
+    {
+      "epoch": 6.988142292490118,
+      "grad_norm": 2.946046829223633,
+      "learning_rate": 1.5119047619047618e-05,
+      "loss": 0.1506,
+      "step": 1768
+    },
+    {
+      "epoch": 6.992094861660079,
+      "grad_norm": 4.145671367645264,
+      "learning_rate": 1.5099206349206349e-05,
+      "loss": 0.1591,
+      "step": 1769
+    },
+    {
+      "epoch": 6.996047430830039,
+      "grad_norm": 3.8622443675994873,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.2398,
+      "step": 1770
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 3.0714492797851562,
+      "learning_rate": 1.505952380952381e-05,
+      "loss": 0.1849,
+      "step": 1771
+    },
+    {
+      "epoch": 7.003952569169961,
+      "grad_norm": 1.683645248413086,
+      "learning_rate": 1.503968253968254e-05,
+      "loss": 0.0714,
+      "step": 1772
+    },
+    {
+      "epoch": 7.007905138339921,
+      "grad_norm": 2.0551791191101074,
+      "learning_rate": 1.501984126984127e-05,
+      "loss": 0.0825,
+      "step": 1773
+    },
+    {
+      "epoch": 7.011857707509882,
+      "grad_norm": 2.248523712158203,
+      "learning_rate": 1.5e-05,
+      "loss": 0.0954,
+      "step": 1774
+    },
+    {
+      "epoch": 7.015810276679842,
+      "grad_norm": 2.069105863571167,
+      "learning_rate": 1.498015873015873e-05,
+      "loss": 0.1108,
+      "step": 1775
+    },
+    {
+      "epoch": 7.019762845849803,
+      "grad_norm": 2.3756532669067383,
+      "learning_rate": 1.4960317460317461e-05,
+      "loss": 0.1123,
+      "step": 1776
+    },
+    {
+      "epoch": 7.023715415019763,
+      "grad_norm": 2.0308547019958496,
+      "learning_rate": 1.4940476190476192e-05,
+      "loss": 0.0715,
+      "step": 1777
+    },
+    {
+      "epoch": 7.027667984189724,
+      "grad_norm": 2.602980613708496,
+      "learning_rate": 1.4920634920634922e-05,
+      "loss": 0.1107,
+      "step": 1778
+    },
+    {
+      "epoch": 7.031620553359684,
+      "grad_norm": 1.8657187223434448,
+      "learning_rate": 1.490079365079365e-05,
+      "loss": 0.0681,
+      "step": 1779
+    },
+    {
+      "epoch": 7.0355731225296445,
+      "grad_norm": 1.9880305528640747,
+      "learning_rate": 1.4880952380952381e-05,
+      "loss": 0.0644,
+      "step": 1780
+    },
+    {
+      "epoch": 7.0395256916996045,
+      "grad_norm": 1.8279154300689697,
+      "learning_rate": 1.4861111111111111e-05,
+      "loss": 0.0662,
+      "step": 1781
+    },
+    {
+      "epoch": 7.043478260869565,
+      "grad_norm": 2.1131718158721924,
+      "learning_rate": 1.4841269841269842e-05,
+      "loss": 0.0787,
+      "step": 1782
+    },
+    {
+      "epoch": 7.047430830039525,
+      "grad_norm": 2.5316834449768066,
+      "learning_rate": 1.4821428571428572e-05,
+      "loss": 0.1016,
+      "step": 1783
+    },
+    {
+      "epoch": 7.051383399209486,
+      "grad_norm": 2.494107961654663,
+      "learning_rate": 1.4801587301587302e-05,
+      "loss": 0.0835,
+      "step": 1784
+    },
+    {
+      "epoch": 7.055335968379446,
+      "grad_norm": 2.633171796798706,
+      "learning_rate": 1.4781746031746033e-05,
+      "loss": 0.0724,
+      "step": 1785
+    },
+    {
+      "epoch": 7.059288537549407,
+      "grad_norm": 1.709030270576477,
+      "learning_rate": 1.4761904761904763e-05,
+      "loss": 0.0618,
+      "step": 1786
+    },
+    {
+      "epoch": 7.063241106719367,
+      "grad_norm": 2.1349966526031494,
+      "learning_rate": 1.4742063492063494e-05,
+      "loss": 0.0763,
+      "step": 1787
+    },
+    {
+      "epoch": 7.067193675889328,
+      "grad_norm": 3.602543354034424,
+      "learning_rate": 1.4722222222222224e-05,
+      "loss": 0.1131,
+      "step": 1788
+    },
+    {
+      "epoch": 7.071146245059288,
+      "grad_norm": 3.5014562606811523,
+      "learning_rate": 1.4702380952380954e-05,
+      "loss": 0.1475,
+      "step": 1789
+    },
+    {
+      "epoch": 7.075098814229249,
+      "grad_norm": 1.7112377882003784,
+      "learning_rate": 1.4682539682539683e-05,
+      "loss": 0.0655,
+      "step": 1790
+    },
+    {
+      "epoch": 7.07905138339921,
+      "grad_norm": 2.5485458374023438,
+      "learning_rate": 1.4662698412698413e-05,
+      "loss": 0.0878,
+      "step": 1791
+    },
+    {
+      "epoch": 7.08300395256917,
+      "grad_norm": 2.0620596408843994,
+      "learning_rate": 1.4642857142857144e-05,
+      "loss": 0.0801,
+      "step": 1792
+    },
+    {
+      "epoch": 7.086956521739131,
+      "grad_norm": 2.537457227706909,
+      "learning_rate": 1.4623015873015874e-05,
+      "loss": 0.0698,
+      "step": 1793
+    },
+    {
+      "epoch": 7.090909090909091,
+      "grad_norm": 1.8265228271484375,
+      "learning_rate": 1.4603174603174605e-05,
+      "loss": 0.0559,
+      "step": 1794
+    },
+    {
+      "epoch": 7.094861660079052,
+      "grad_norm": 1.819351315498352,
+      "learning_rate": 1.4583333333333335e-05,
+      "loss": 0.0665,
+      "step": 1795
+    },
+    {
+      "epoch": 7.098814229249012,
+      "grad_norm": 2.4396214485168457,
+      "learning_rate": 1.4563492063492065e-05,
+      "loss": 0.0786,
+      "step": 1796
+    },
+    {
+      "epoch": 7.102766798418973,
+      "grad_norm": 2.4278268814086914,
+      "learning_rate": 1.4543650793650796e-05,
+      "loss": 0.0957,
+      "step": 1797
+    },
+    {
+      "epoch": 7.106719367588933,
+      "grad_norm": 2.5539965629577637,
+      "learning_rate": 1.4523809523809526e-05,
+      "loss": 0.0876,
+      "step": 1798
+    },
+    {
+      "epoch": 7.1106719367588935,
+      "grad_norm": 2.3676304817199707,
+      "learning_rate": 1.4503968253968256e-05,
+      "loss": 0.0801,
+      "step": 1799
+    },
+    {
+      "epoch": 7.1146245059288535,
+      "grad_norm": 2.803798198699951,
+      "learning_rate": 1.4484126984126987e-05,
+      "loss": 0.1065,
+      "step": 1800
+    },
+    {
+      "epoch": 7.118577075098814,
+      "grad_norm": 2.0833263397216797,
+      "learning_rate": 1.4464285714285717e-05,
+      "loss": 0.0701,
+      "step": 1801
+    },
+    {
+      "epoch": 7.122529644268774,
+      "grad_norm": 1.9213995933532715,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.0689,
+      "step": 1802
+    },
+    {
+      "epoch": 7.126482213438735,
+      "grad_norm": 2.999032974243164,
+      "learning_rate": 1.4424603174603174e-05,
+      "loss": 0.1407,
+      "step": 1803
+    },
+    {
+      "epoch": 7.130434782608695,
+      "grad_norm": 2.492380142211914,
+      "learning_rate": 1.4404761904761905e-05,
+      "loss": 0.0919,
+      "step": 1804
+    },
+    {
+      "epoch": 7.134387351778656,
+      "grad_norm": 2.170208692550659,
+      "learning_rate": 1.4384920634920635e-05,
+      "loss": 0.0862,
+      "step": 1805
+    },
+    {
+      "epoch": 7.138339920948616,
+      "grad_norm": 2.5370352268218994,
+      "learning_rate": 1.4365079365079364e-05,
+      "loss": 0.0899,
+      "step": 1806
+    },
+    {
+      "epoch": 7.142292490118577,
+      "grad_norm": 2.430377960205078,
+      "learning_rate": 1.4345238095238094e-05,
+      "loss": 0.1082,
+      "step": 1807
+    },
+    {
+      "epoch": 7.146245059288537,
+      "grad_norm": 3.0296268463134766,
+      "learning_rate": 1.4325396825396825e-05,
+      "loss": 0.1017,
+      "step": 1808
+    },
+    {
+      "epoch": 7.150197628458498,
+      "grad_norm": 2.192507743835449,
+      "learning_rate": 1.4305555555555555e-05,
+      "loss": 0.0969,
+      "step": 1809
+    },
+    {
+      "epoch": 7.154150197628459,
+      "grad_norm": 2.191904306411743,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.0905,
+      "step": 1810
+    },
+    {
+      "epoch": 7.158102766798419,
+      "grad_norm": 1.9413102865219116,
+      "learning_rate": 1.4265873015873016e-05,
+      "loss": 0.0746,
+      "step": 1811
+    },
+    {
+      "epoch": 7.16205533596838,
+      "grad_norm": 2.1803011894226074,
+      "learning_rate": 1.4246031746031746e-05,
+      "loss": 0.0679,
+      "step": 1812
+    },
+    {
+      "epoch": 7.16600790513834,
+      "grad_norm": 1.9626377820968628,
+      "learning_rate": 1.4226190476190477e-05,
+      "loss": 0.0784,
+      "step": 1813
+    },
+    {
+      "epoch": 7.169960474308301,
+      "grad_norm": 3.6826534271240234,
+      "learning_rate": 1.4206349206349207e-05,
+      "loss": 0.1193,
+      "step": 1814
+    },
+    {
+      "epoch": 7.173913043478261,
+      "grad_norm": 2.018355131149292,
+      "learning_rate": 1.4186507936507937e-05,
+      "loss": 0.0614,
+      "step": 1815
+    },
+    {
+      "epoch": 7.177865612648222,
+      "grad_norm": 2.3391740322113037,
+      "learning_rate": 1.4166666666666668e-05,
+      "loss": 0.1052,
+      "step": 1816
+    },
+    {
+      "epoch": 7.181818181818182,
+      "grad_norm": 3.039984941482544,
+      "learning_rate": 1.4146825396825396e-05,
+      "loss": 0.1077,
+      "step": 1817
+    },
+    {
+      "epoch": 7.1857707509881426,
+      "grad_norm": 2.2779464721679688,
+      "learning_rate": 1.4126984126984127e-05,
+      "loss": 0.0654,
+      "step": 1818
+    },
+    {
+      "epoch": 7.189723320158103,
+      "grad_norm": 1.959446907043457,
+      "learning_rate": 1.4107142857142857e-05,
+      "loss": 0.0676,
+      "step": 1819
+    },
+    {
+      "epoch": 7.1936758893280635,
+      "grad_norm": 3.0260629653930664,
+      "learning_rate": 1.4087301587301587e-05,
+      "loss": 0.0924,
+      "step": 1820
+    },
+    {
+      "epoch": 7.1976284584980235,
+      "grad_norm": 1.552370548248291,
+      "learning_rate": 1.4067460317460318e-05,
+      "loss": 0.0557,
+      "step": 1821
+    },
+    {
+      "epoch": 7.201581027667984,
+      "grad_norm": 1.851386308670044,
+      "learning_rate": 1.4047619047619048e-05,
+      "loss": 0.0842,
+      "step": 1822
+    },
+    {
+      "epoch": 7.205533596837944,
+      "grad_norm": 1.901151180267334,
+      "learning_rate": 1.4027777777777779e-05,
+      "loss": 0.0602,
+      "step": 1823
+    },
+    {
+      "epoch": 7.209486166007905,
+      "grad_norm": 2.256410598754883,
+      "learning_rate": 1.4007936507936509e-05,
+      "loss": 0.0779,
+      "step": 1824
+    },
+    {
+      "epoch": 7.213438735177865,
+      "grad_norm": 2.2768027782440186,
+      "learning_rate": 1.398809523809524e-05,
+      "loss": 0.0701,
+      "step": 1825
+    },
+    {
+      "epoch": 7.217391304347826,
+      "grad_norm": 1.8319358825683594,
+      "learning_rate": 1.396825396825397e-05,
+      "loss": 0.0711,
+      "step": 1826
+    },
+    {
+      "epoch": 7.221343873517786,
+      "grad_norm": 1.6305453777313232,
+      "learning_rate": 1.39484126984127e-05,
+      "loss": 0.0601,
+      "step": 1827
+    },
+    {
+      "epoch": 7.225296442687747,
+      "grad_norm": 2.377932071685791,
+      "learning_rate": 1.392857142857143e-05,
+      "loss": 0.084,
+      "step": 1828
+    },
+    {
+      "epoch": 7.229249011857707,
+      "grad_norm": 2.3718206882476807,
+      "learning_rate": 1.390873015873016e-05,
+      "loss": 0.1073,
+      "step": 1829
+    },
+    {
+      "epoch": 7.233201581027668,
+      "grad_norm": 1.8948851823806763,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.066,
+      "step": 1830
+    },
+    {
+      "epoch": 7.237154150197629,
+      "grad_norm": 2.6554150581359863,
+      "learning_rate": 1.386904761904762e-05,
+      "loss": 0.1004,
+      "step": 1831
+    },
+    {
+      "epoch": 7.241106719367589,
+      "grad_norm": 1.7534379959106445,
+      "learning_rate": 1.384920634920635e-05,
+      "loss": 0.0686,
+      "step": 1832
+    },
+    {
+      "epoch": 7.24505928853755,
+      "grad_norm": 2.8044631481170654,
+      "learning_rate": 1.382936507936508e-05,
+      "loss": 0.0981,
+      "step": 1833
+    },
+    {
+      "epoch": 7.24901185770751,
+      "grad_norm": 1.962010145187378,
+      "learning_rate": 1.3809523809523811e-05,
+      "loss": 0.0606,
+      "step": 1834
+    },
+    {
+      "epoch": 7.252964426877471,
+      "grad_norm": 2.598278522491455,
+      "learning_rate": 1.3789682539682541e-05,
+      "loss": 0.0823,
+      "step": 1835
+    },
+    {
+      "epoch": 7.256916996047431,
+      "grad_norm": 2.0581893920898438,
+      "learning_rate": 1.3769841269841272e-05,
+      "loss": 0.0724,
+      "step": 1836
+    },
+    {
+      "epoch": 7.260869565217392,
+      "grad_norm": 2.1700425148010254,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 0.0747,
+      "step": 1837
+    },
+    {
+      "epoch": 7.264822134387352,
+      "grad_norm": 2.3965978622436523,
+      "learning_rate": 1.3730158730158733e-05,
+      "loss": 0.0756,
+      "step": 1838
+    },
+    {
+      "epoch": 7.2687747035573125,
+      "grad_norm": 2.8531904220581055,
+      "learning_rate": 1.3710317460317463e-05,
+      "loss": 0.0915,
+      "step": 1839
+    },
+    {
+      "epoch": 7.2727272727272725,
+      "grad_norm": 1.8420287370681763,
+      "learning_rate": 1.3690476190476192e-05,
+      "loss": 0.0674,
+      "step": 1840
+    },
+    {
+      "epoch": 7.276679841897233,
+      "grad_norm": 2.040949583053589,
+      "learning_rate": 1.367063492063492e-05,
+      "loss": 0.0662,
+      "step": 1841
+    },
+    {
+      "epoch": 7.280632411067193,
+      "grad_norm": 4.134637832641602,
+      "learning_rate": 1.365079365079365e-05,
+      "loss": 0.1148,
+      "step": 1842
+    },
+    {
+      "epoch": 7.284584980237154,
+      "grad_norm": 2.1330273151397705,
+      "learning_rate": 1.3630952380952381e-05,
+      "loss": 0.0594,
+      "step": 1843
+    },
+    {
+      "epoch": 7.288537549407114,
+      "grad_norm": 2.3376622200012207,
+      "learning_rate": 1.3611111111111111e-05,
+      "loss": 0.1026,
+      "step": 1844
+    },
+    {
+      "epoch": 7.292490118577075,
+      "grad_norm": 2.1795151233673096,
+      "learning_rate": 1.359126984126984e-05,
+      "loss": 0.0685,
+      "step": 1845
+    },
+    {
+      "epoch": 7.296442687747035,
+      "grad_norm": 2.4371635913848877,
+      "learning_rate": 1.357142857142857e-05,
+      "loss": 0.0856,
+      "step": 1846
+    },
+    {
+      "epoch": 7.300395256916996,
+      "grad_norm": 1.8547295331954956,
+      "learning_rate": 1.3551587301587301e-05,
+      "loss": 0.0593,
+      "step": 1847
+    },
+    {
+      "epoch": 7.304347826086957,
+      "grad_norm": 2.3141019344329834,
+      "learning_rate": 1.3531746031746031e-05,
+      "loss": 0.0935,
+      "step": 1848
+    },
+    {
+      "epoch": 7.308300395256917,
+      "grad_norm": 1.9820470809936523,
+      "learning_rate": 1.3511904761904762e-05,
+      "loss": 0.069,
+      "step": 1849
+    },
+    {
+      "epoch": 7.312252964426877,
+      "grad_norm": 2.5655643939971924,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.0899,
+      "step": 1850
+    },
+    {
+      "epoch": 7.316205533596838,
+      "grad_norm": 2.5492866039276123,
+      "learning_rate": 1.3472222222222222e-05,
+      "loss": 0.0874,
+      "step": 1851
+    },
+    {
+      "epoch": 7.320158102766799,
+      "grad_norm": 2.2358458042144775,
+      "learning_rate": 1.3452380952380953e-05,
+      "loss": 0.0769,
+      "step": 1852
+    },
+    {
+      "epoch": 7.324110671936759,
+      "grad_norm": 1.9372227191925049,
+      "learning_rate": 1.3432539682539683e-05,
+      "loss": 0.0729,
+      "step": 1853
+    },
+    {
+      "epoch": 7.32806324110672,
+      "grad_norm": 2.8513784408569336,
+      "learning_rate": 1.3412698412698413e-05,
+      "loss": 0.0988,
+      "step": 1854
+    },
+    {
+      "epoch": 7.33201581027668,
+      "grad_norm": 2.437175989151001,
+      "learning_rate": 1.3392857142857144e-05,
+      "loss": 0.0891,
+      "step": 1855
+    },
+    {
+      "epoch": 7.335968379446641,
+      "grad_norm": 1.8044086694717407,
+      "learning_rate": 1.3373015873015873e-05,
+      "loss": 0.0611,
+      "step": 1856
+    },
+    {
+      "epoch": 7.339920948616601,
+      "grad_norm": 1.8136906623840332,
+      "learning_rate": 1.3353174603174603e-05,
+      "loss": 0.0604,
+      "step": 1857
+    },
+    {
+      "epoch": 7.3438735177865615,
+      "grad_norm": 2.4834377765655518,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.126,
+      "step": 1858
+    },
+    {
+      "epoch": 7.3478260869565215,
+      "grad_norm": 2.810823678970337,
+      "learning_rate": 1.3313492063492064e-05,
+      "loss": 0.1165,
+      "step": 1859
+    },
+    {
+      "epoch": 7.351778656126482,
+      "grad_norm": 3.216146945953369,
+      "learning_rate": 1.3293650793650794e-05,
+      "loss": 0.0965,
+      "step": 1860
+    },
+    {
+      "epoch": 7.355731225296442,
+      "grad_norm": 2.910884141921997,
+      "learning_rate": 1.3273809523809524e-05,
+      "loss": 0.103,
+      "step": 1861
+    },
+    {
+      "epoch": 7.359683794466403,
+      "grad_norm": 1.8496427536010742,
+      "learning_rate": 1.3253968253968255e-05,
+      "loss": 0.0602,
+      "step": 1862
+    },
+    {
+      "epoch": 7.363636363636363,
+      "grad_norm": 2.382444381713867,
+      "learning_rate": 1.3234126984126985e-05,
+      "loss": 0.1,
+      "step": 1863
+    },
+    {
+      "epoch": 7.367588932806324,
+      "grad_norm": 2.162414789199829,
+      "learning_rate": 1.3214285714285716e-05,
+      "loss": 0.0704,
+      "step": 1864
+    },
+    {
+      "epoch": 7.371541501976284,
+      "grad_norm": 2.582324981689453,
+      "learning_rate": 1.3194444444444446e-05,
+      "loss": 0.0735,
+      "step": 1865
+    },
+    {
+      "epoch": 7.375494071146245,
+      "grad_norm": 2.0686068534851074,
+      "learning_rate": 1.3174603174603176e-05,
+      "loss": 0.0691,
+      "step": 1866
+    },
+    {
+      "epoch": 7.379446640316205,
+      "grad_norm": 2.73799467086792,
+      "learning_rate": 1.3154761904761907e-05,
+      "loss": 0.0977,
+      "step": 1867
+    },
+    {
+      "epoch": 7.383399209486166,
+      "grad_norm": 2.5663845539093018,
+      "learning_rate": 1.3134920634920635e-05,
+      "loss": 0.0964,
+      "step": 1868
+    },
+    {
+      "epoch": 7.387351778656127,
+      "grad_norm": 2.602886199951172,
+      "learning_rate": 1.3115079365079366e-05,
+      "loss": 0.0834,
+      "step": 1869
+    },
+    {
+      "epoch": 7.391304347826087,
+      "grad_norm": 2.9359493255615234,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 0.1205,
+      "step": 1870
+    },
+    {
+      "epoch": 7.395256916996048,
+      "grad_norm": 2.4510388374328613,
+      "learning_rate": 1.3075396825396826e-05,
+      "loss": 0.0765,
+      "step": 1871
+    },
+    {
+      "epoch": 7.399209486166008,
+      "grad_norm": 1.9874929189682007,
+      "learning_rate": 1.3055555555555557e-05,
+      "loss": 0.0798,
+      "step": 1872
+    },
+    {
+      "epoch": 7.403162055335969,
+      "grad_norm": 1.9240570068359375,
+      "learning_rate": 1.3035714285714287e-05,
+      "loss": 0.0649,
+      "step": 1873
+    },
+    {
+      "epoch": 7.407114624505929,
+      "grad_norm": 1.8569307327270508,
+      "learning_rate": 1.3015873015873018e-05,
+      "loss": 0.061,
+      "step": 1874
+    },
+    {
+      "epoch": 7.41106719367589,
+      "grad_norm": 2.232877016067505,
+      "learning_rate": 1.2996031746031748e-05,
+      "loss": 0.0778,
+      "step": 1875
+    },
+    {
+      "epoch": 7.41501976284585,
+      "grad_norm": 2.330413818359375,
+      "learning_rate": 1.2976190476190478e-05,
+      "loss": 0.0897,
+      "step": 1876
+    },
+    {
+      "epoch": 7.4189723320158105,
+      "grad_norm": 2.543062686920166,
+      "learning_rate": 1.2956349206349209e-05,
+      "loss": 0.1039,
+      "step": 1877
+    },
+    {
+      "epoch": 7.4229249011857705,
+      "grad_norm": 2.075242757797241,
+      "learning_rate": 1.2936507936507939e-05,
+      "loss": 0.0918,
+      "step": 1878
+    },
+    {
+      "epoch": 7.426877470355731,
+      "grad_norm": 2.339674234390259,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 0.0905,
+      "step": 1879
+    },
+    {
+      "epoch": 7.430830039525691,
+      "grad_norm": 2.0178771018981934,
+      "learning_rate": 1.2896825396825398e-05,
+      "loss": 0.074,
+      "step": 1880
+    },
+    {
+      "epoch": 7.434782608695652,
+      "grad_norm": 2.169339179992676,
+      "learning_rate": 1.2876984126984127e-05,
+      "loss": 0.0627,
+      "step": 1881
+    },
+    {
+      "epoch": 7.438735177865612,
+      "grad_norm": 3.798720121383667,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.1162,
+      "step": 1882
+    },
+    {
+      "epoch": 7.442687747035573,
+      "grad_norm": 3.5218403339385986,
+      "learning_rate": 1.2837301587301586e-05,
+      "loss": 0.1009,
+      "step": 1883
+    },
+    {
+      "epoch": 7.446640316205533,
+      "grad_norm": 2.855560302734375,
+      "learning_rate": 1.2817460317460316e-05,
+      "loss": 0.1037,
+      "step": 1884
+    },
+    {
+      "epoch": 7.450592885375494,
+      "grad_norm": 1.8869415521621704,
+      "learning_rate": 1.2797619047619047e-05,
+      "loss": 0.0598,
+      "step": 1885
+    },
+    {
+      "epoch": 7.454545454545454,
+      "grad_norm": 1.6059690713882446,
+      "learning_rate": 1.2777777777777777e-05,
+      "loss": 0.0565,
+      "step": 1886
+    },
+    {
+      "epoch": 7.458498023715415,
+      "grad_norm": 3.0044198036193848,
+      "learning_rate": 1.2757936507936507e-05,
+      "loss": 0.131,
+      "step": 1887
+    },
+    {
+      "epoch": 7.462450592885375,
+      "grad_norm": 2.2843515872955322,
+      "learning_rate": 1.2738095238095238e-05,
+      "loss": 0.0856,
+      "step": 1888
+    },
+    {
+      "epoch": 7.466403162055336,
+      "grad_norm": 1.9446831941604614,
+      "learning_rate": 1.2718253968253968e-05,
+      "loss": 0.0657,
+      "step": 1889
+    },
+    {
+      "epoch": 7.470355731225297,
+      "grad_norm": 2.3453097343444824,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.0929,
+      "step": 1890
+    },
+    {
+      "epoch": 7.474308300395257,
+      "grad_norm": 2.4141080379486084,
+      "learning_rate": 1.2678571428571429e-05,
+      "loss": 0.0882,
+      "step": 1891
+    },
+    {
+      "epoch": 7.478260869565218,
+      "grad_norm": 3.0322489738464355,
+      "learning_rate": 1.265873015873016e-05,
+      "loss": 0.0845,
+      "step": 1892
+    },
+    {
+      "epoch": 7.482213438735178,
+      "grad_norm": 2.348433017730713,
+      "learning_rate": 1.263888888888889e-05,
+      "loss": 0.0828,
+      "step": 1893
+    },
+    {
+      "epoch": 7.486166007905139,
+      "grad_norm": 2.1347391605377197,
+      "learning_rate": 1.261904761904762e-05,
+      "loss": 0.0667,
+      "step": 1894
+    },
+    {
+      "epoch": 7.490118577075099,
+      "grad_norm": 2.272301435470581,
+      "learning_rate": 1.2599206349206349e-05,
+      "loss": 0.0898,
+      "step": 1895
+    },
+    {
+      "epoch": 7.4940711462450595,
+      "grad_norm": 2.3849878311157227,
+      "learning_rate": 1.2579365079365079e-05,
+      "loss": 0.0747,
+      "step": 1896
+    },
+    {
+      "epoch": 7.4980237154150196,
+      "grad_norm": 2.220501661300659,
+      "learning_rate": 1.255952380952381e-05,
+      "loss": 0.0794,
+      "step": 1897
+    },
+    {
+      "epoch": 7.5019762845849804,
+      "grad_norm": 2.281405448913574,
+      "learning_rate": 1.253968253968254e-05,
+      "loss": 0.0915,
+      "step": 1898
+    },
+    {
+      "epoch": 7.5059288537549405,
+      "grad_norm": 2.4831249713897705,
+      "learning_rate": 1.251984126984127e-05,
+      "loss": 0.0725,
+      "step": 1899
+    },
+    {
+      "epoch": 7.509881422924901,
+      "grad_norm": 2.52744197845459,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0719,
+      "step": 1900
+    },
+    {
+      "epoch": 7.513833992094861,
+      "grad_norm": 2.3339502811431885,
+      "learning_rate": 1.2480158730158731e-05,
+      "loss": 0.0821,
+      "step": 1901
+    },
+    {
+      "epoch": 7.517786561264822,
+      "grad_norm": 2.408015012741089,
+      "learning_rate": 1.2460317460317461e-05,
+      "loss": 0.0929,
+      "step": 1902
+    },
+    {
+      "epoch": 7.521739130434782,
+      "grad_norm": 2.307608127593994,
+      "learning_rate": 1.2440476190476192e-05,
+      "loss": 0.0884,
+      "step": 1903
+    },
+    {
+      "epoch": 7.525691699604743,
+      "grad_norm": 2.454751491546631,
+      "learning_rate": 1.2420634920634922e-05,
+      "loss": 0.0851,
+      "step": 1904
+    },
+    {
+      "epoch": 7.529644268774703,
+      "grad_norm": 1.6989669799804688,
+      "learning_rate": 1.2400793650793652e-05,
+      "loss": 0.0765,
+      "step": 1905
+    },
+    {
+      "epoch": 7.533596837944664,
+      "grad_norm": 3.3208930492401123,
+      "learning_rate": 1.2380952380952381e-05,
+      "loss": 0.1033,
+      "step": 1906
+    },
+    {
+      "epoch": 7.537549407114625,
+      "grad_norm": 2.317495822906494,
+      "learning_rate": 1.2361111111111112e-05,
+      "loss": 0.0911,
+      "step": 1907
+    },
+    {
+      "epoch": 7.541501976284585,
+      "grad_norm": 2.3214991092681885,
+      "learning_rate": 1.2341269841269842e-05,
+      "loss": 0.0985,
+      "step": 1908
+    },
+    {
+      "epoch": 7.545454545454545,
+      "grad_norm": 2.3527603149414062,
+      "learning_rate": 1.2321428571428572e-05,
+      "loss": 0.077,
+      "step": 1909
+    },
+    {
+      "epoch": 7.549407114624506,
+      "grad_norm": 3.5933728218078613,
+      "learning_rate": 1.2301587301587301e-05,
+      "loss": 0.1284,
+      "step": 1910
+    },
+    {
+      "epoch": 7.553359683794467,
+      "grad_norm": 2.274735450744629,
+      "learning_rate": 1.2281746031746031e-05,
+      "loss": 0.1062,
+      "step": 1911
+    },
+    {
+      "epoch": 7.557312252964427,
+      "grad_norm": 3.133971691131592,
+      "learning_rate": 1.2261904761904762e-05,
+      "loss": 0.1185,
+      "step": 1912
+    },
+    {
+      "epoch": 7.561264822134388,
+      "grad_norm": 3.7312350273132324,
+      "learning_rate": 1.2242063492063492e-05,
+      "loss": 0.1343,
+      "step": 1913
+    },
+    {
+      "epoch": 7.565217391304348,
+      "grad_norm": 1.8403165340423584,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.0699,
+      "step": 1914
+    },
+    {
+      "epoch": 7.569169960474309,
+      "grad_norm": 3.350177049636841,
+      "learning_rate": 1.2202380952380953e-05,
+      "loss": 0.1266,
+      "step": 1915
+    },
+    {
+      "epoch": 7.573122529644269,
+      "grad_norm": 1.6349577903747559,
+      "learning_rate": 1.2182539682539683e-05,
+      "loss": 0.0628,
+      "step": 1916
+    },
+    {
+      "epoch": 7.5770750988142295,
+      "grad_norm": 1.906072974205017,
+      "learning_rate": 1.2162698412698414e-05,
+      "loss": 0.0835,
+      "step": 1917
+    },
+    {
+      "epoch": 7.5810276679841895,
+      "grad_norm": 2.1547391414642334,
+      "learning_rate": 1.2142857142857144e-05,
+      "loss": 0.0803,
+      "step": 1918
+    },
+    {
+      "epoch": 7.58498023715415,
+      "grad_norm": 2.4813320636749268,
+      "learning_rate": 1.2123015873015874e-05,
+      "loss": 0.1215,
+      "step": 1919
+    },
+    {
+      "epoch": 7.58893280632411,
+      "grad_norm": 2.235426664352417,
+      "learning_rate": 1.2103174603174603e-05,
+      "loss": 0.0785,
+      "step": 1920
+    },
+    {
+      "epoch": 7.592885375494071,
+      "grad_norm": 2.80841326713562,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 0.1276,
+      "step": 1921
+    },
+    {
+      "epoch": 7.596837944664031,
+      "grad_norm": 2.305530548095703,
+      "learning_rate": 1.2063492063492064e-05,
+      "loss": 0.065,
+      "step": 1922
+    },
+    {
+      "epoch": 7.600790513833992,
+      "grad_norm": 1.6414098739624023,
+      "learning_rate": 1.2043650793650794e-05,
+      "loss": 0.0607,
+      "step": 1923
+    },
+    {
+      "epoch": 7.604743083003952,
+      "grad_norm": 2.6492977142333984,
+      "learning_rate": 1.2023809523809525e-05,
+      "loss": 0.0877,
+      "step": 1924
+    },
+    {
+      "epoch": 7.608695652173913,
+      "grad_norm": 2.1286118030548096,
+      "learning_rate": 1.2003968253968255e-05,
+      "loss": 0.0843,
+      "step": 1925
+    },
+    {
+      "epoch": 7.612648221343873,
+      "grad_norm": 2.137827157974243,
+      "learning_rate": 1.1984126984126985e-05,
+      "loss": 0.0691,
+      "step": 1926
+    },
+    {
+      "epoch": 7.616600790513834,
+      "grad_norm": 2.486067056655884,
+      "learning_rate": 1.1964285714285716e-05,
+      "loss": 0.0795,
+      "step": 1927
+    },
+    {
+      "epoch": 7.620553359683795,
+      "grad_norm": 3.5269148349761963,
+      "learning_rate": 1.1944444444444446e-05,
+      "loss": 0.1021,
+      "step": 1928
+    },
+    {
+      "epoch": 7.624505928853755,
+      "grad_norm": 3.005458116531372,
+      "learning_rate": 1.1924603174603176e-05,
+      "loss": 0.0863,
+      "step": 1929
+    },
+    {
+      "epoch": 7.628458498023716,
+      "grad_norm": 1.5922969579696655,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.0562,
+      "step": 1930
+    },
+    {
+      "epoch": 7.632411067193676,
+      "grad_norm": 2.0607211589813232,
+      "learning_rate": 1.1884920634920635e-05,
+      "loss": 0.0796,
+      "step": 1931
+    },
+    {
+      "epoch": 7.636363636363637,
+      "grad_norm": 2.8598814010620117,
+      "learning_rate": 1.1865079365079366e-05,
+      "loss": 0.1132,
+      "step": 1932
+    },
+    {
+      "epoch": 7.640316205533597,
+      "grad_norm": 2.145017147064209,
+      "learning_rate": 1.1845238095238095e-05,
+      "loss": 0.081,
+      "step": 1933
+    },
+    {
+      "epoch": 7.644268774703558,
+      "grad_norm": 1.9191386699676514,
+      "learning_rate": 1.1825396825396825e-05,
+      "loss": 0.0867,
+      "step": 1934
+    },
+    {
+      "epoch": 7.648221343873518,
+      "grad_norm": 2.149658441543579,
+      "learning_rate": 1.1805555555555555e-05,
+      "loss": 0.0706,
+      "step": 1935
+    },
+    {
+      "epoch": 7.6521739130434785,
+      "grad_norm": 2.7651522159576416,
+      "learning_rate": 1.1785714285714286e-05,
+      "loss": 0.0835,
+      "step": 1936
+    },
+    {
+      "epoch": 7.6561264822134385,
+      "grad_norm": 3.162454605102539,
+      "learning_rate": 1.1765873015873016e-05,
+      "loss": 0.1092,
+      "step": 1937
+    },
+    {
+      "epoch": 7.660079051383399,
+      "grad_norm": 2.8850865364074707,
+      "learning_rate": 1.1746031746031746e-05,
+      "loss": 0.0828,
+      "step": 1938
+    },
+    {
+      "epoch": 7.664031620553359,
+      "grad_norm": 1.7384766340255737,
+      "learning_rate": 1.1726190476190477e-05,
+      "loss": 0.057,
+      "step": 1939
+    },
+    {
+      "epoch": 7.66798418972332,
+      "grad_norm": 1.8011753559112549,
+      "learning_rate": 1.1706349206349207e-05,
+      "loss": 0.0737,
+      "step": 1940
+    },
+    {
+      "epoch": 7.67193675889328,
+      "grad_norm": 3.141075372695923,
+      "learning_rate": 1.1686507936507938e-05,
+      "loss": 0.0868,
+      "step": 1941
+    },
+    {
+      "epoch": 7.675889328063241,
+      "grad_norm": 3.0187206268310547,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.1188,
+      "step": 1942
+    },
+    {
+      "epoch": 7.679841897233201,
+      "grad_norm": 2.71610689163208,
+      "learning_rate": 1.1646825396825398e-05,
+      "loss": 0.0873,
+      "step": 1943
+    },
+    {
+      "epoch": 7.683794466403162,
+      "grad_norm": 3.3756630420684814,
+      "learning_rate": 1.1626984126984129e-05,
+      "loss": 0.1259,
+      "step": 1944
+    },
+    {
+      "epoch": 7.687747035573123,
+      "grad_norm": 2.53981351852417,
+      "learning_rate": 1.1607142857142857e-05,
+      "loss": 0.0935,
+      "step": 1945
+    },
+    {
+      "epoch": 7.691699604743083,
+      "grad_norm": 1.9116166830062866,
+      "learning_rate": 1.1587301587301588e-05,
+      "loss": 0.0657,
+      "step": 1946
+    },
+    {
+      "epoch": 7.695652173913043,
+      "grad_norm": 2.145357370376587,
+      "learning_rate": 1.1567460317460318e-05,
+      "loss": 0.0928,
+      "step": 1947
+    },
+    {
+      "epoch": 7.699604743083004,
+      "grad_norm": 2.8089849948883057,
+      "learning_rate": 1.1547619047619048e-05,
+      "loss": 0.0938,
+      "step": 1948
+    },
+    {
+      "epoch": 7.703557312252965,
+      "grad_norm": 2.510547399520874,
+      "learning_rate": 1.1527777777777779e-05,
+      "loss": 0.0942,
+      "step": 1949
+    },
+    {
+      "epoch": 7.707509881422925,
+      "grad_norm": 2.142611503601074,
+      "learning_rate": 1.1507936507936508e-05,
+      "loss": 0.0776,
+      "step": 1950
+    },
+    {
+      "epoch": 7.711462450592886,
+      "grad_norm": 2.8611981868743896,
+      "learning_rate": 1.1488095238095238e-05,
+      "loss": 0.0874,
+      "step": 1951
+    },
+    {
+      "epoch": 7.715415019762846,
+      "grad_norm": 2.1545233726501465,
+      "learning_rate": 1.1468253968253968e-05,
+      "loss": 0.0846,
+      "step": 1952
+    },
+    {
+      "epoch": 7.719367588932807,
+      "grad_norm": 2.4546070098876953,
+      "learning_rate": 1.1448412698412699e-05,
+      "loss": 0.0745,
+      "step": 1953
+    },
+    {
+      "epoch": 7.723320158102767,
+      "grad_norm": 3.605870008468628,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.1321,
+      "step": 1954
+    },
+    {
+      "epoch": 7.7272727272727275,
+      "grad_norm": 3.173536539077759,
+      "learning_rate": 1.140873015873016e-05,
+      "loss": 0.1024,
+      "step": 1955
+    },
+    {
+      "epoch": 7.7312252964426875,
+      "grad_norm": 2.163079023361206,
+      "learning_rate": 1.138888888888889e-05,
+      "loss": 0.0626,
+      "step": 1956
+    },
+    {
+      "epoch": 7.735177865612648,
+      "grad_norm": 2.247568130493164,
+      "learning_rate": 1.136904761904762e-05,
+      "loss": 0.0801,
+      "step": 1957
+    },
+    {
+      "epoch": 7.739130434782608,
+      "grad_norm": 2.0294995307922363,
+      "learning_rate": 1.1349206349206349e-05,
+      "loss": 0.0987,
+      "step": 1958
+    },
+    {
+      "epoch": 7.743083003952569,
+      "grad_norm": 2.5226917266845703,
+      "learning_rate": 1.132936507936508e-05,
+      "loss": 0.0815,
+      "step": 1959
+    },
+    {
+      "epoch": 7.747035573122529,
+      "grad_norm": 2.68029522895813,
+      "learning_rate": 1.130952380952381e-05,
+      "loss": 0.1003,
+      "step": 1960
+    },
+    {
+      "epoch": 7.75098814229249,
+      "grad_norm": 2.617349624633789,
+      "learning_rate": 1.128968253968254e-05,
+      "loss": 0.0843,
+      "step": 1961
+    },
+    {
+      "epoch": 7.75494071146245,
+      "grad_norm": 2.5157854557037354,
+      "learning_rate": 1.126984126984127e-05,
+      "loss": 0.1024,
+      "step": 1962
+    },
+    {
+      "epoch": 7.758893280632411,
+      "grad_norm": 2.2654972076416016,
+      "learning_rate": 1.125e-05,
+      "loss": 0.0801,
+      "step": 1963
+    },
+    {
+      "epoch": 7.762845849802371,
+      "grad_norm": 2.273188352584839,
+      "learning_rate": 1.1230158730158731e-05,
+      "loss": 0.0774,
+      "step": 1964
+    },
+    {
+      "epoch": 7.766798418972332,
+      "grad_norm": 2.1782922744750977,
+      "learning_rate": 1.1210317460317461e-05,
+      "loss": 0.069,
+      "step": 1965
+    },
+    {
+      "epoch": 7.770750988142293,
+      "grad_norm": 2.23232102394104,
+      "learning_rate": 1.1190476190476192e-05,
+      "loss": 0.0868,
+      "step": 1966
+    },
+    {
+      "epoch": 7.774703557312253,
+      "grad_norm": 2.0577633380889893,
+      "learning_rate": 1.1170634920634922e-05,
+      "loss": 0.0718,
+      "step": 1967
+    },
+    {
+      "epoch": 7.778656126482213,
+      "grad_norm": 2.2454826831817627,
+      "learning_rate": 1.1150793650793653e-05,
+      "loss": 0.0789,
+      "step": 1968
+    },
+    {
+      "epoch": 7.782608695652174,
+      "grad_norm": 3.088087558746338,
+      "learning_rate": 1.1130952380952381e-05,
+      "loss": 0.0944,
+      "step": 1969
+    },
+    {
+      "epoch": 7.786561264822135,
+      "grad_norm": 2.3378899097442627,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.0908,
+      "step": 1970
+    },
+    {
+      "epoch": 7.790513833992095,
+      "grad_norm": 2.6260411739349365,
+      "learning_rate": 1.1091269841269842e-05,
+      "loss": 0.087,
+      "step": 1971
+    },
+    {
+      "epoch": 7.794466403162056,
+      "grad_norm": 1.8749479055404663,
+      "learning_rate": 1.107142857142857e-05,
+      "loss": 0.0641,
+      "step": 1972
+    },
+    {
+      "epoch": 7.798418972332016,
+      "grad_norm": 2.9281997680664062,
+      "learning_rate": 1.1051587301587301e-05,
+      "loss": 0.1033,
+      "step": 1973
+    },
+    {
+      "epoch": 7.8023715415019765,
+      "grad_norm": 1.7113523483276367,
+      "learning_rate": 1.1031746031746031e-05,
+      "loss": 0.0573,
+      "step": 1974
+    },
+    {
+      "epoch": 7.8063241106719365,
+      "grad_norm": 4.461965560913086,
+      "learning_rate": 1.1011904761904762e-05,
+      "loss": 0.1805,
+      "step": 1975
+    },
+    {
+      "epoch": 7.810276679841897,
+      "grad_norm": 2.3460776805877686,
+      "learning_rate": 1.0992063492063492e-05,
+      "loss": 0.0981,
+      "step": 1976
+    },
+    {
+      "epoch": 7.8142292490118574,
+      "grad_norm": 2.7546355724334717,
+      "learning_rate": 1.0972222222222223e-05,
+      "loss": 0.0963,
+      "step": 1977
+    },
+    {
+      "epoch": 7.818181818181818,
+      "grad_norm": 3.3164448738098145,
+      "learning_rate": 1.0952380952380953e-05,
+      "loss": 0.096,
+      "step": 1978
+    },
+    {
+      "epoch": 7.822134387351778,
+      "grad_norm": 1.8524004220962524,
+      "learning_rate": 1.0932539682539683e-05,
+      "loss": 0.0725,
+      "step": 1979
+    },
+    {
+      "epoch": 7.826086956521739,
+      "grad_norm": 2.2797439098358154,
+      "learning_rate": 1.0912698412698414e-05,
+      "loss": 0.0792,
+      "step": 1980
+    },
+    {
+      "epoch": 7.830039525691699,
+      "grad_norm": 2.863020420074463,
+      "learning_rate": 1.0892857142857144e-05,
+      "loss": 0.0749,
+      "step": 1981
+    },
+    {
+      "epoch": 7.83399209486166,
+      "grad_norm": 2.2198400497436523,
+      "learning_rate": 1.0873015873015874e-05,
+      "loss": 0.0986,
+      "step": 1982
+    },
+    {
+      "epoch": 7.837944664031621,
+      "grad_norm": 1.540998935699463,
+      "learning_rate": 1.0853174603174605e-05,
+      "loss": 0.0654,
+      "step": 1983
+    },
+    {
+      "epoch": 7.841897233201581,
+      "grad_norm": 1.856723666191101,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 0.0751,
+      "step": 1984
+    },
+    {
+      "epoch": 7.845849802371541,
+      "grad_norm": 2.4738926887512207,
+      "learning_rate": 1.0813492063492064e-05,
+      "loss": 0.0848,
+      "step": 1985
+    },
+    {
+      "epoch": 7.849802371541502,
+      "grad_norm": 1.9589910507202148,
+      "learning_rate": 1.0793650793650794e-05,
+      "loss": 0.0674,
+      "step": 1986
+    },
+    {
+      "epoch": 7.853754940711463,
+      "grad_norm": 2.1947100162506104,
+      "learning_rate": 1.0773809523809525e-05,
+      "loss": 0.077,
+      "step": 1987
+    },
+    {
+      "epoch": 7.857707509881423,
+      "grad_norm": 3.0187575817108154,
+      "learning_rate": 1.0753968253968255e-05,
+      "loss": 0.1128,
+      "step": 1988
+    },
+    {
+      "epoch": 7.861660079051384,
+      "grad_norm": 1.8199398517608643,
+      "learning_rate": 1.0734126984126984e-05,
+      "loss": 0.077,
+      "step": 1989
+    },
+    {
+      "epoch": 7.865612648221344,
+      "grad_norm": 2.459689140319824,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.0834,
+      "step": 1990
+    },
+    {
+      "epoch": 7.869565217391305,
+      "grad_norm": 2.451312303543091,
+      "learning_rate": 1.0694444444444444e-05,
+      "loss": 0.1036,
+      "step": 1991
+    },
+    {
+      "epoch": 7.873517786561265,
+      "grad_norm": 2.400502920150757,
+      "learning_rate": 1.0674603174603175e-05,
+      "loss": 0.0821,
+      "step": 1992
+    },
+    {
+      "epoch": 7.877470355731226,
+      "grad_norm": 1.9994014501571655,
+      "learning_rate": 1.0654761904761905e-05,
+      "loss": 0.072,
+      "step": 1993
+    },
+    {
+      "epoch": 7.881422924901186,
+      "grad_norm": 1.9804929494857788,
+      "learning_rate": 1.0634920634920636e-05,
+      "loss": 0.0723,
+      "step": 1994
+    },
+    {
+      "epoch": 7.8853754940711465,
+      "grad_norm": 1.7949855327606201,
+      "learning_rate": 1.0615079365079366e-05,
+      "loss": 0.0725,
+      "step": 1995
+    },
+    {
+      "epoch": 7.8893280632411065,
+      "grad_norm": 3.4392995834350586,
+      "learning_rate": 1.0595238095238096e-05,
+      "loss": 0.1088,
+      "step": 1996
+    },
+    {
+      "epoch": 7.893280632411067,
+      "grad_norm": 2.1142258644104004,
+      "learning_rate": 1.0575396825396825e-05,
+      "loss": 0.0653,
+      "step": 1997
+    },
+    {
+      "epoch": 7.897233201581027,
+      "grad_norm": 1.7416729927062988,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 0.0623,
+      "step": 1998
+    },
+    {
+      "epoch": 7.901185770750988,
+      "grad_norm": 2.438103437423706,
+      "learning_rate": 1.0535714285714286e-05,
+      "loss": 0.0712,
+      "step": 1999
+    },
+    {
+      "epoch": 7.905138339920948,
+      "grad_norm": 2.4873170852661133,
+      "learning_rate": 1.0515873015873016e-05,
+      "loss": 0.0704,
+      "step": 2000
+    },
+    {
+      "epoch": 7.909090909090909,
+      "grad_norm": 2.9187819957733154,
+      "learning_rate": 1.0496031746031747e-05,
+      "loss": 0.1055,
+      "step": 2001
+    },
+    {
+      "epoch": 7.913043478260869,
+      "grad_norm": 2.2539193630218506,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.0777,
+      "step": 2002
+    },
+    {
+      "epoch": 7.91699604743083,
+      "grad_norm": 1.9834545850753784,
+      "learning_rate": 1.0456349206349207e-05,
+      "loss": 0.0651,
+      "step": 2003
+    },
+    {
+      "epoch": 7.920948616600791,
+      "grad_norm": 2.7357428073883057,
+      "learning_rate": 1.0436507936507938e-05,
+      "loss": 0.0943,
+      "step": 2004
+    },
+    {
+      "epoch": 7.924901185770751,
+      "grad_norm": 2.9565341472625732,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.1032,
+      "step": 2005
+    },
+    {
+      "epoch": 7.928853754940711,
+      "grad_norm": 2.176581382751465,
+      "learning_rate": 1.0396825396825398e-05,
+      "loss": 0.0774,
+      "step": 2006
+    },
+    {
+      "epoch": 7.932806324110672,
+      "grad_norm": 2.717416524887085,
+      "learning_rate": 1.0376984126984129e-05,
+      "loss": 0.1105,
+      "step": 2007
+    },
+    {
+      "epoch": 7.936758893280633,
+      "grad_norm": 2.976388931274414,
+      "learning_rate": 1.0357142857142859e-05,
+      "loss": 0.1036,
+      "step": 2008
+    },
+    {
+      "epoch": 7.940711462450593,
+      "grad_norm": 4.3491411209106445,
+      "learning_rate": 1.0337301587301588e-05,
+      "loss": 0.166,
+      "step": 2009
+    },
+    {
+      "epoch": 7.944664031620554,
+      "grad_norm": 2.4038586616516113,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.0833,
+      "step": 2010
+    },
+    {
+      "epoch": 7.948616600790514,
+      "grad_norm": 2.2004966735839844,
+      "learning_rate": 1.0297619047619047e-05,
+      "loss": 0.079,
+      "step": 2011
+    },
+    {
+      "epoch": 7.952569169960475,
+      "grad_norm": 2.817812919616699,
+      "learning_rate": 1.0277777777777777e-05,
+      "loss": 0.1185,
+      "step": 2012
+    },
+    {
+      "epoch": 7.956521739130435,
+      "grad_norm": 2.230272054672241,
+      "learning_rate": 1.0257936507936508e-05,
+      "loss": 0.0932,
+      "step": 2013
+    },
+    {
+      "epoch": 7.9604743083003955,
+      "grad_norm": 2.4217472076416016,
+      "learning_rate": 1.0238095238095238e-05,
+      "loss": 0.0912,
+      "step": 2014
+    },
+    {
+      "epoch": 7.9644268774703555,
+      "grad_norm": 2.3574728965759277,
+      "learning_rate": 1.0218253968253968e-05,
+      "loss": 0.0844,
+      "step": 2015
+    },
+    {
+      "epoch": 7.968379446640316,
+      "grad_norm": 1.9760262966156006,
+      "learning_rate": 1.0198412698412699e-05,
+      "loss": 0.0852,
+      "step": 2016
+    },
+    {
+      "epoch": 7.972332015810276,
+      "grad_norm": 3.1597955226898193,
+      "learning_rate": 1.0178571428571429e-05,
+      "loss": 0.1101,
+      "step": 2017
+    },
+    {
+      "epoch": 7.976284584980237,
+      "grad_norm": 2.150995969772339,
+      "learning_rate": 1.015873015873016e-05,
+      "loss": 0.0867,
+      "step": 2018
+    },
+    {
+      "epoch": 7.980237154150197,
+      "grad_norm": 2.0937557220458984,
+      "learning_rate": 1.013888888888889e-05,
+      "loss": 0.0701,
+      "step": 2019
+    },
+    {
+      "epoch": 7.984189723320158,
+      "grad_norm": 3.2109408378601074,
+      "learning_rate": 1.011904761904762e-05,
+      "loss": 0.0974,
+      "step": 2020
+    },
+    {
+      "epoch": 7.988142292490118,
+      "grad_norm": 3.7526028156280518,
+      "learning_rate": 1.009920634920635e-05,
+      "loss": 0.1281,
+      "step": 2021
+    },
+    {
+      "epoch": 7.992094861660079,
+      "grad_norm": 2.220501184463501,
+      "learning_rate": 1.007936507936508e-05,
+      "loss": 0.1216,
+      "step": 2022
+    },
+    {
+      "epoch": 7.996047430830039,
+      "grad_norm": 2.583831548690796,
+      "learning_rate": 1.005952380952381e-05,
+      "loss": 0.0817,
+      "step": 2023
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.142831802368164,
+      "learning_rate": 1.003968253968254e-05,
+      "loss": 0.0659,
+      "step": 2024
+    },
+    {
+      "epoch": 8.003952569169961,
+      "grad_norm": 0.9592176079750061,
+      "learning_rate": 1.001984126984127e-05,
+      "loss": 0.0366,
+      "step": 2025
+    },
+    {
+      "epoch": 8.007905138339922,
+      "grad_norm": 1.110743522644043,
+      "learning_rate": 1e-05,
+      "loss": 0.0397,
+      "step": 2026
+    },
+    {
+      "epoch": 8.011857707509881,
+      "grad_norm": 1.0800786018371582,
+      "learning_rate": 9.980158730158731e-06,
+      "loss": 0.04,
+      "step": 2027
+    },
+    {
+      "epoch": 8.015810276679842,
+      "grad_norm": 1.1953082084655762,
+      "learning_rate": 9.96031746031746e-06,
+      "loss": 0.04,
+      "step": 2028
+    },
+    {
+      "epoch": 8.019762845849803,
+      "grad_norm": 1.4953047037124634,
+      "learning_rate": 9.94047619047619e-06,
+      "loss": 0.0726,
+      "step": 2029
+    },
+    {
+      "epoch": 8.023715415019764,
+      "grad_norm": 1.4194068908691406,
+      "learning_rate": 9.92063492063492e-06,
+      "loss": 0.0474,
+      "step": 2030
+    },
+    {
+      "epoch": 8.027667984189723,
+      "grad_norm": 1.2994885444641113,
+      "learning_rate": 9.900793650793651e-06,
+      "loss": 0.0462,
+      "step": 2031
+    },
+    {
+      "epoch": 8.031620553359684,
+      "grad_norm": 1.5428236722946167,
+      "learning_rate": 9.880952380952381e-06,
+      "loss": 0.0681,
+      "step": 2032
+    },
+    {
+      "epoch": 8.035573122529645,
+      "grad_norm": 1.375458836555481,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 0.0437,
+      "step": 2033
+    },
+    {
+      "epoch": 8.039525691699605,
+      "grad_norm": 1.303333044052124,
+      "learning_rate": 9.841269841269842e-06,
+      "loss": 0.0418,
+      "step": 2034
+    },
+    {
+      "epoch": 8.043478260869565,
+      "grad_norm": 1.6367980241775513,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 0.0603,
+      "step": 2035
+    },
+    {
+      "epoch": 8.047430830039525,
+      "grad_norm": 1.2272205352783203,
+      "learning_rate": 9.801587301587301e-06,
+      "loss": 0.0373,
+      "step": 2036
+    },
+    {
+      "epoch": 8.051383399209486,
+      "grad_norm": 1.42149019241333,
+      "learning_rate": 9.781746031746032e-06,
+      "loss": 0.0376,
+      "step": 2037
+    },
+    {
+      "epoch": 8.055335968379447,
+      "grad_norm": 1.757165789604187,
+      "learning_rate": 9.761904761904762e-06,
+      "loss": 0.051,
+      "step": 2038
+    },
+    {
+      "epoch": 8.059288537549406,
+      "grad_norm": 2.7419040203094482,
+      "learning_rate": 9.742063492063492e-06,
+      "loss": 0.0709,
+      "step": 2039
+    },
+    {
+      "epoch": 8.063241106719367,
+      "grad_norm": 1.15654718875885,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.0328,
+      "step": 2040
+    },
+    {
+      "epoch": 8.067193675889328,
+      "grad_norm": 1.6456338167190552,
+      "learning_rate": 9.702380952380953e-06,
+      "loss": 0.0584,
+      "step": 2041
+    },
+    {
+      "epoch": 8.071146245059289,
+      "grad_norm": 1.8168511390686035,
+      "learning_rate": 9.682539682539683e-06,
+      "loss": 0.0547,
+      "step": 2042
+    },
+    {
+      "epoch": 8.075098814229248,
+      "grad_norm": 1.2571635246276855,
+      "learning_rate": 9.662698412698414e-06,
+      "loss": 0.0409,
+      "step": 2043
+    },
+    {
+      "epoch": 8.079051383399209,
+      "grad_norm": 1.5656439065933228,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.0416,
+      "step": 2044
+    },
+    {
+      "epoch": 8.08300395256917,
+      "grad_norm": 1.297942042350769,
+      "learning_rate": 9.623015873015875e-06,
+      "loss": 0.0449,
+      "step": 2045
+    },
+    {
+      "epoch": 8.08695652173913,
+      "grad_norm": 1.577046275138855,
+      "learning_rate": 9.603174603174605e-06,
+      "loss": 0.0613,
+      "step": 2046
+    },
+    {
+      "epoch": 8.090909090909092,
+      "grad_norm": 1.2949413061141968,
+      "learning_rate": 9.583333333333334e-06,
+      "loss": 0.0421,
+      "step": 2047
+    },
+    {
+      "epoch": 8.09486166007905,
+      "grad_norm": 1.3732746839523315,
+      "learning_rate": 9.563492063492064e-06,
+      "loss": 0.0487,
+      "step": 2048
+    },
+    {
+      "epoch": 8.098814229249012,
+      "grad_norm": 1.6629955768585205,
+      "learning_rate": 9.543650793650793e-06,
+      "loss": 0.0535,
+      "step": 2049
+    },
+    {
+      "epoch": 8.102766798418973,
+      "grad_norm": 1.3761000633239746,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.0475,
+      "step": 2050
+    },
+    {
+      "epoch": 8.106719367588934,
+      "grad_norm": 1.721295952796936,
+      "learning_rate": 9.503968253968253e-06,
+      "loss": 0.0532,
+      "step": 2051
+    },
+    {
+      "epoch": 8.110671936758893,
+      "grad_norm": 2.0054855346679688,
+      "learning_rate": 9.484126984126984e-06,
+      "loss": 0.0652,
+      "step": 2052
+    },
+    {
+      "epoch": 8.114624505928854,
+      "grad_norm": 1.1261333227157593,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 0.0385,
+      "step": 2053
+    },
+    {
+      "epoch": 8.118577075098814,
+      "grad_norm": 1.6062073707580566,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.0692,
+      "step": 2054
+    },
+    {
+      "epoch": 8.122529644268775,
+      "grad_norm": 1.7869125604629517,
+      "learning_rate": 9.424603174603175e-06,
+      "loss": 0.0545,
+      "step": 2055
+    },
+    {
+      "epoch": 8.126482213438734,
+      "grad_norm": 1.4991074800491333,
+      "learning_rate": 9.404761904761905e-06,
+      "loss": 0.0564,
+      "step": 2056
+    },
+    {
+      "epoch": 8.130434782608695,
+      "grad_norm": 1.527508020401001,
+      "learning_rate": 9.384920634920636e-06,
+      "loss": 0.0515,
+      "step": 2057
+    },
+    {
+      "epoch": 8.134387351778656,
+      "grad_norm": 1.497931957244873,
+      "learning_rate": 9.365079365079366e-06,
+      "loss": 0.0542,
+      "step": 2058
+    },
+    {
+      "epoch": 8.138339920948617,
+      "grad_norm": 1.5828138589859009,
+      "learning_rate": 9.345238095238096e-06,
+      "loss": 0.0502,
+      "step": 2059
+    },
+    {
+      "epoch": 8.142292490118576,
+      "grad_norm": 2.0395078659057617,
+      "learning_rate": 9.325396825396827e-06,
+      "loss": 0.0642,
+      "step": 2060
+    },
+    {
+      "epoch": 8.146245059288537,
+      "grad_norm": 1.7669570446014404,
+      "learning_rate": 9.305555555555555e-06,
+      "loss": 0.0685,
+      "step": 2061
+    },
+    {
+      "epoch": 8.150197628458498,
+      "grad_norm": 1.2871284484863281,
+      "learning_rate": 9.285714285714286e-06,
+      "loss": 0.0494,
+      "step": 2062
+    },
+    {
+      "epoch": 8.154150197628459,
+      "grad_norm": 1.5461068153381348,
+      "learning_rate": 9.265873015873016e-06,
+      "loss": 0.0486,
+      "step": 2063
+    },
+    {
+      "epoch": 8.15810276679842,
+      "grad_norm": 3.5550456047058105,
+      "learning_rate": 9.246031746031747e-06,
+      "loss": 0.0718,
+      "step": 2064
+    },
+    {
+      "epoch": 8.162055335968379,
+      "grad_norm": 1.4119137525558472,
+      "learning_rate": 9.226190476190477e-06,
+      "loss": 0.039,
+      "step": 2065
+    },
+    {
+      "epoch": 8.16600790513834,
+      "grad_norm": 1.7155872583389282,
+      "learning_rate": 9.206349206349207e-06,
+      "loss": 0.0476,
+      "step": 2066
+    },
+    {
+      "epoch": 8.1699604743083,
+      "grad_norm": 1.5434550046920776,
+      "learning_rate": 9.186507936507938e-06,
+      "loss": 0.0549,
+      "step": 2067
+    },
+    {
+      "epoch": 8.173913043478262,
+      "grad_norm": 1.677038550376892,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.0592,
+      "step": 2068
+    },
+    {
+      "epoch": 8.17786561264822,
+      "grad_norm": 1.7242430448532104,
+      "learning_rate": 9.146825396825397e-06,
+      "loss": 0.0461,
+      "step": 2069
+    },
+    {
+      "epoch": 8.181818181818182,
+      "grad_norm": 1.5962724685668945,
+      "learning_rate": 9.126984126984127e-06,
+      "loss": 0.056,
+      "step": 2070
+    },
+    {
+      "epoch": 8.185770750988143,
+      "grad_norm": 1.2169913053512573,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 0.0414,
+      "step": 2071
+    },
+    {
+      "epoch": 8.189723320158103,
+      "grad_norm": 1.2942997217178345,
+      "learning_rate": 9.087301587301588e-06,
+      "loss": 0.0397,
+      "step": 2072
+    },
+    {
+      "epoch": 8.193675889328063,
+      "grad_norm": 1.064394474029541,
+      "learning_rate": 9.067460317460318e-06,
+      "loss": 0.0412,
+      "step": 2073
+    },
+    {
+      "epoch": 8.197628458498023,
+      "grad_norm": 1.550001859664917,
+      "learning_rate": 9.047619047619047e-06,
+      "loss": 0.0496,
+      "step": 2074
+    },
+    {
+      "epoch": 8.201581027667984,
+      "grad_norm": 0.9107499718666077,
+      "learning_rate": 9.027777777777777e-06,
+      "loss": 0.036,
+      "step": 2075
+    },
+    {
+      "epoch": 8.205533596837945,
+      "grad_norm": 1.4544087648391724,
+      "learning_rate": 9.007936507936508e-06,
+      "loss": 0.0486,
+      "step": 2076
+    },
+    {
+      "epoch": 8.209486166007904,
+      "grad_norm": 1.8703144788742065,
+      "learning_rate": 8.988095238095238e-06,
+      "loss": 0.0581,
+      "step": 2077
+    },
+    {
+      "epoch": 8.213438735177865,
+      "grad_norm": 1.3876852989196777,
+      "learning_rate": 8.968253968253968e-06,
+      "loss": 0.0463,
+      "step": 2078
+    },
+    {
+      "epoch": 8.217391304347826,
+      "grad_norm": 1.303426742553711,
+      "learning_rate": 8.948412698412699e-06,
+      "loss": 0.0374,
+      "step": 2079
+    },
+    {
+      "epoch": 8.221343873517787,
+      "grad_norm": 1.1986994743347168,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.0455,
+      "step": 2080
+    },
+    {
+      "epoch": 8.225296442687746,
+      "grad_norm": 1.7828227281570435,
+      "learning_rate": 8.90873015873016e-06,
+      "loss": 0.0522,
+      "step": 2081
+    },
+    {
+      "epoch": 8.229249011857707,
+      "grad_norm": 1.5747590065002441,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.0524,
+      "step": 2082
+    },
+    {
+      "epoch": 8.233201581027668,
+      "grad_norm": 1.693088412284851,
+      "learning_rate": 8.86904761904762e-06,
+      "loss": 0.0467,
+      "step": 2083
+    },
+    {
+      "epoch": 8.237154150197629,
+      "grad_norm": 1.7213762998580933,
+      "learning_rate": 8.84920634920635e-06,
+      "loss": 0.0479,
+      "step": 2084
+    },
+    {
+      "epoch": 8.24110671936759,
+      "grad_norm": 1.4260462522506714,
+      "learning_rate": 8.829365079365081e-06,
+      "loss": 0.0515,
+      "step": 2085
+    },
+    {
+      "epoch": 8.245059288537549,
+      "grad_norm": 1.4755898714065552,
+      "learning_rate": 8.80952380952381e-06,
+      "loss": 0.0434,
+      "step": 2086
+    },
+    {
+      "epoch": 8.24901185770751,
+      "grad_norm": 1.2368816137313843,
+      "learning_rate": 8.78968253968254e-06,
+      "loss": 0.0404,
+      "step": 2087
+    },
+    {
+      "epoch": 8.25296442687747,
+      "grad_norm": 1.3295128345489502,
+      "learning_rate": 8.769841269841269e-06,
+      "loss": 0.0443,
+      "step": 2088
+    },
+    {
+      "epoch": 8.256916996047432,
+      "grad_norm": 1.3453065156936646,
+      "learning_rate": 8.75e-06,
+      "loss": 0.044,
+      "step": 2089
+    },
+    {
+      "epoch": 8.26086956521739,
+      "grad_norm": 1.3558346033096313,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.0477,
+      "step": 2090
+    },
+    {
+      "epoch": 8.264822134387352,
+      "grad_norm": 1.5338172912597656,
+      "learning_rate": 8.71031746031746e-06,
+      "loss": 0.0489,
+      "step": 2091
+    },
+    {
+      "epoch": 8.268774703557312,
+      "grad_norm": 1.350160002708435,
+      "learning_rate": 8.69047619047619e-06,
+      "loss": 0.0508,
+      "step": 2092
+    },
+    {
+      "epoch": 8.272727272727273,
+      "grad_norm": 1.5303537845611572,
+      "learning_rate": 8.67063492063492e-06,
+      "loss": 0.0465,
+      "step": 2093
+    },
+    {
+      "epoch": 8.276679841897232,
+      "grad_norm": 1.5438976287841797,
+      "learning_rate": 8.650793650793651e-06,
+      "loss": 0.0692,
+      "step": 2094
+    },
+    {
+      "epoch": 8.280632411067193,
+      "grad_norm": 1.1079212427139282,
+      "learning_rate": 8.630952380952381e-06,
+      "loss": 0.044,
+      "step": 2095
+    },
+    {
+      "epoch": 8.284584980237154,
+      "grad_norm": 1.2162022590637207,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 0.0369,
+      "step": 2096
+    },
+    {
+      "epoch": 8.288537549407115,
+      "grad_norm": 1.4534077644348145,
+      "learning_rate": 8.591269841269842e-06,
+      "loss": 0.0549,
+      "step": 2097
+    },
+    {
+      "epoch": 8.292490118577074,
+      "grad_norm": 1.3834608793258667,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.0423,
+      "step": 2098
+    },
+    {
+      "epoch": 8.296442687747035,
+      "grad_norm": 1.7133499383926392,
+      "learning_rate": 8.551587301587303e-06,
+      "loss": 0.0612,
+      "step": 2099
+    },
+    {
+      "epoch": 8.300395256916996,
+      "grad_norm": 1.966752052307129,
+      "learning_rate": 8.531746031746032e-06,
+      "loss": 0.0524,
+      "step": 2100
+    },
+    {
+      "epoch": 8.304347826086957,
+      "grad_norm": 2.143880844116211,
+      "learning_rate": 8.511904761904762e-06,
+      "loss": 0.0515,
+      "step": 2101
+    },
+    {
+      "epoch": 8.308300395256918,
+      "grad_norm": 2.052530527114868,
+      "learning_rate": 8.492063492063492e-06,
+      "loss": 0.0721,
+      "step": 2102
+    },
+    {
+      "epoch": 8.312252964426877,
+      "grad_norm": 2.0002057552337646,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 0.0559,
+      "step": 2103
+    },
+    {
+      "epoch": 8.316205533596838,
+      "grad_norm": 1.0597388744354248,
+      "learning_rate": 8.452380952380953e-06,
+      "loss": 0.0444,
+      "step": 2104
+    },
+    {
+      "epoch": 8.320158102766799,
+      "grad_norm": 1.3378074169158936,
+      "learning_rate": 8.432539682539684e-06,
+      "loss": 0.0408,
+      "step": 2105
+    },
+    {
+      "epoch": 8.32411067193676,
+      "grad_norm": 1.926456093788147,
+      "learning_rate": 8.412698412698414e-06,
+      "loss": 0.0592,
+      "step": 2106
+    },
+    {
+      "epoch": 8.328063241106719,
+      "grad_norm": 1.6579585075378418,
+      "learning_rate": 8.392857142857143e-06,
+      "loss": 0.0448,
+      "step": 2107
+    },
+    {
+      "epoch": 8.33201581027668,
+      "grad_norm": 1.6691573858261108,
+      "learning_rate": 8.373015873015873e-06,
+      "loss": 0.0476,
+      "step": 2108
+    },
+    {
+      "epoch": 8.33596837944664,
+      "grad_norm": 1.7489898204803467,
+      "learning_rate": 8.353174603174603e-06,
+      "loss": 0.067,
+      "step": 2109
+    },
+    {
+      "epoch": 8.339920948616601,
+      "grad_norm": 1.1511108875274658,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.0477,
+      "step": 2110
+    },
+    {
+      "epoch": 8.34387351778656,
+      "grad_norm": 1.527985692024231,
+      "learning_rate": 8.313492063492064e-06,
+      "loss": 0.041,
+      "step": 2111
+    },
+    {
+      "epoch": 8.347826086956522,
+      "grad_norm": 1.5950675010681152,
+      "learning_rate": 8.293650793650794e-06,
+      "loss": 0.0588,
+      "step": 2112
+    },
+    {
+      "epoch": 8.351778656126482,
+      "grad_norm": 1.2112717628479004,
+      "learning_rate": 8.273809523809523e-06,
+      "loss": 0.0397,
+      "step": 2113
+    },
+    {
+      "epoch": 8.355731225296443,
+      "grad_norm": 1.6547131538391113,
+      "learning_rate": 8.253968253968254e-06,
+      "loss": 0.0609,
+      "step": 2114
+    },
+    {
+      "epoch": 8.359683794466402,
+      "grad_norm": 1.6982663869857788,
+      "learning_rate": 8.234126984126984e-06,
+      "loss": 0.0554,
+      "step": 2115
+    },
+    {
+      "epoch": 8.363636363636363,
+      "grad_norm": 1.5355052947998047,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.0663,
+      "step": 2116
+    },
+    {
+      "epoch": 8.367588932806324,
+      "grad_norm": 0.8851954340934753,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 0.0346,
+      "step": 2117
+    },
+    {
+      "epoch": 8.371541501976285,
+      "grad_norm": 1.6623241901397705,
+      "learning_rate": 8.174603174603175e-06,
+      "loss": 0.0488,
+      "step": 2118
+    },
+    {
+      "epoch": 8.375494071146244,
+      "grad_norm": 0.9663978219032288,
+      "learning_rate": 8.154761904761905e-06,
+      "loss": 0.0372,
+      "step": 2119
+    },
+    {
+      "epoch": 8.379446640316205,
+      "grad_norm": 1.2707703113555908,
+      "learning_rate": 8.134920634920636e-06,
+      "loss": 0.0439,
+      "step": 2120
+    },
+    {
+      "epoch": 8.383399209486166,
+      "grad_norm": 2.081395149230957,
+      "learning_rate": 8.115079365079366e-06,
+      "loss": 0.0474,
+      "step": 2121
+    },
+    {
+      "epoch": 8.387351778656127,
+      "grad_norm": 1.8119603395462036,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.0496,
+      "step": 2122
+    },
+    {
+      "epoch": 8.391304347826088,
+      "grad_norm": 1.7686362266540527,
+      "learning_rate": 8.075396825396827e-06,
+      "loss": 0.0549,
+      "step": 2123
+    },
+    {
+      "epoch": 8.395256916996047,
+      "grad_norm": 1.752198338508606,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 0.044,
+      "step": 2124
+    },
+    {
+      "epoch": 8.399209486166008,
+      "grad_norm": 1.523292064666748,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 0.0491,
+      "step": 2125
+    },
+    {
+      "epoch": 8.403162055335969,
+      "grad_norm": 1.8821699619293213,
+      "learning_rate": 8.015873015873016e-06,
+      "loss": 0.0473,
+      "step": 2126
+    },
+    {
+      "epoch": 8.40711462450593,
+      "grad_norm": 1.762847661972046,
+      "learning_rate": 7.996031746031745e-06,
+      "loss": 0.0498,
+      "step": 2127
+    },
+    {
+      "epoch": 8.411067193675889,
+      "grad_norm": 1.716994285583496,
+      "learning_rate": 7.976190476190475e-06,
+      "loss": 0.0536,
+      "step": 2128
+    },
+    {
+      "epoch": 8.41501976284585,
+      "grad_norm": 1.1499348878860474,
+      "learning_rate": 7.956349206349206e-06,
+      "loss": 0.0351,
+      "step": 2129
+    },
+    {
+      "epoch": 8.41897233201581,
+      "grad_norm": 0.8600573539733887,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.036,
+      "step": 2130
+    },
+    {
+      "epoch": 8.422924901185771,
+      "grad_norm": 1.5126005411148071,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.0523,
+      "step": 2131
+    },
+    {
+      "epoch": 8.42687747035573,
+      "grad_norm": 0.9862467050552368,
+      "learning_rate": 7.896825396825397e-06,
+      "loss": 0.0395,
+      "step": 2132
+    },
+    {
+      "epoch": 8.430830039525691,
+      "grad_norm": 1.1598824262619019,
+      "learning_rate": 7.876984126984127e-06,
+      "loss": 0.0387,
+      "step": 2133
+    },
+    {
+      "epoch": 8.434782608695652,
+      "grad_norm": 2.440631151199341,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.0462,
+      "step": 2134
+    },
+    {
+      "epoch": 8.438735177865613,
+      "grad_norm": 1.5269445180892944,
+      "learning_rate": 7.837301587301588e-06,
+      "loss": 0.0576,
+      "step": 2135
+    },
+    {
+      "epoch": 8.442687747035572,
+      "grad_norm": 1.6497447490692139,
+      "learning_rate": 7.817460317460318e-06,
+      "loss": 0.0695,
+      "step": 2136
+    },
+    {
+      "epoch": 8.446640316205533,
+      "grad_norm": 2.240730047225952,
+      "learning_rate": 7.797619047619049e-06,
+      "loss": 0.0558,
+      "step": 2137
+    },
+    {
+      "epoch": 8.450592885375494,
+      "grad_norm": 1.2452744245529175,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.0861,
+      "step": 2138
+    },
+    {
+      "epoch": 8.454545454545455,
+      "grad_norm": 1.3848949670791626,
+      "learning_rate": 7.757936507936508e-06,
+      "loss": 0.0522,
+      "step": 2139
+    },
+    {
+      "epoch": 8.458498023715414,
+      "grad_norm": 1.5077615976333618,
+      "learning_rate": 7.738095238095238e-06,
+      "loss": 0.051,
+      "step": 2140
+    },
+    {
+      "epoch": 8.462450592885375,
+      "grad_norm": 1.707553505897522,
+      "learning_rate": 7.718253968253969e-06,
+      "loss": 0.0592,
+      "step": 2141
+    },
+    {
+      "epoch": 8.466403162055336,
+      "grad_norm": 1.6871522665023804,
+      "learning_rate": 7.698412698412699e-06,
+      "loss": 0.0518,
+      "step": 2142
+    },
+    {
+      "epoch": 8.470355731225297,
+      "grad_norm": 2.1698107719421387,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 0.0599,
+      "step": 2143
+    },
+    {
+      "epoch": 8.474308300395258,
+      "grad_norm": 1.105559229850769,
+      "learning_rate": 7.65873015873016e-06,
+      "loss": 0.0368,
+      "step": 2144
+    },
+    {
+      "epoch": 8.478260869565217,
+      "grad_norm": 1.6224021911621094,
+      "learning_rate": 7.63888888888889e-06,
+      "loss": 0.0513,
+      "step": 2145
+    },
+    {
+      "epoch": 8.482213438735178,
+      "grad_norm": 1.7027612924575806,
+      "learning_rate": 7.6190476190476205e-06,
+      "loss": 0.0502,
+      "step": 2146
+    },
+    {
+      "epoch": 8.486166007905139,
+      "grad_norm": 1.5517146587371826,
+      "learning_rate": 7.599206349206349e-06,
+      "loss": 0.0515,
+      "step": 2147
+    },
+    {
+      "epoch": 8.4901185770751,
+      "grad_norm": 1.5171629190444946,
+      "learning_rate": 7.5793650793650795e-06,
+      "loss": 0.0509,
+      "step": 2148
+    },
+    {
+      "epoch": 8.494071146245059,
+      "grad_norm": 1.476023554801941,
+      "learning_rate": 7.559523809523809e-06,
+      "loss": 0.0498,
+      "step": 2149
+    },
+    {
+      "epoch": 8.49802371541502,
+      "grad_norm": 1.404647707939148,
+      "learning_rate": 7.5396825396825394e-06,
+      "loss": 0.0449,
+      "step": 2150
+    },
+    {
+      "epoch": 8.50197628458498,
+      "grad_norm": 1.2494837045669556,
+      "learning_rate": 7.51984126984127e-06,
+      "loss": 0.0503,
+      "step": 2151
+    },
+    {
+      "epoch": 8.505928853754941,
+      "grad_norm": 1.9429749250411987,
+      "learning_rate": 7.5e-06,
+      "loss": 0.0454,
+      "step": 2152
+    },
+    {
+      "epoch": 8.5098814229249,
+      "grad_norm": 1.3501893281936646,
+      "learning_rate": 7.4801587301587306e-06,
+      "loss": 0.0435,
+      "step": 2153
+    },
+    {
+      "epoch": 8.513833992094861,
+      "grad_norm": 1.3887948989868164,
+      "learning_rate": 7.460317460317461e-06,
+      "loss": 0.0479,
+      "step": 2154
+    },
+    {
+      "epoch": 8.517786561264822,
+      "grad_norm": 1.7705752849578857,
+      "learning_rate": 7.4404761904761905e-06,
+      "loss": 0.0579,
+      "step": 2155
+    },
+    {
+      "epoch": 8.521739130434783,
+      "grad_norm": 1.6018643379211426,
+      "learning_rate": 7.420634920634921e-06,
+      "loss": 0.0488,
+      "step": 2156
+    },
+    {
+      "epoch": 8.525691699604742,
+      "grad_norm": 1.2142655849456787,
+      "learning_rate": 7.400793650793651e-06,
+      "loss": 0.0474,
+      "step": 2157
+    },
+    {
+      "epoch": 8.529644268774703,
+      "grad_norm": 1.2533504962921143,
+      "learning_rate": 7.380952380952382e-06,
+      "loss": 0.0428,
+      "step": 2158
+    },
+    {
+      "epoch": 8.533596837944664,
+      "grad_norm": 1.5056265592575073,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 0.0412,
+      "step": 2159
+    },
+    {
+      "epoch": 8.537549407114625,
+      "grad_norm": 1.0978323221206665,
+      "learning_rate": 7.3412698412698415e-06,
+      "loss": 0.0429,
+      "step": 2160
+    },
+    {
+      "epoch": 8.541501976284586,
+      "grad_norm": 1.600511908531189,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 0.0506,
+      "step": 2161
+    },
+    {
+      "epoch": 8.545454545454545,
+      "grad_norm": 1.2966439723968506,
+      "learning_rate": 7.301587301587302e-06,
+      "loss": 0.0445,
+      "step": 2162
+    },
+    {
+      "epoch": 8.549407114624506,
+      "grad_norm": 1.1138356924057007,
+      "learning_rate": 7.281746031746033e-06,
+      "loss": 0.0362,
+      "step": 2163
+    },
+    {
+      "epoch": 8.553359683794467,
+      "grad_norm": 1.2294458150863647,
+      "learning_rate": 7.261904761904763e-06,
+      "loss": 0.0418,
+      "step": 2164
+    },
+    {
+      "epoch": 8.557312252964428,
+      "grad_norm": 2.5378119945526123,
+      "learning_rate": 7.242063492063493e-06,
+      "loss": 0.0703,
+      "step": 2165
+    },
+    {
+      "epoch": 8.561264822134387,
+      "grad_norm": 1.068594217300415,
+      "learning_rate": 7.222222222222222e-06,
+      "loss": 0.0407,
+      "step": 2166
+    },
+    {
+      "epoch": 8.565217391304348,
+      "grad_norm": 1.1735312938690186,
+      "learning_rate": 7.2023809523809524e-06,
+      "loss": 0.0435,
+      "step": 2167
+    },
+    {
+      "epoch": 8.569169960474309,
+      "grad_norm": 2.6890134811401367,
+      "learning_rate": 7.182539682539682e-06,
+      "loss": 0.0897,
+      "step": 2168
+    },
+    {
+      "epoch": 8.57312252964427,
+      "grad_norm": 1.2781907320022583,
+      "learning_rate": 7.162698412698412e-06,
+      "loss": 0.0517,
+      "step": 2169
+    },
+    {
+      "epoch": 8.577075098814229,
+      "grad_norm": 1.8715581893920898,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.0598,
+      "step": 2170
+    },
+    {
+      "epoch": 8.58102766798419,
+      "grad_norm": 2.0492424964904785,
+      "learning_rate": 7.123015873015873e-06,
+      "loss": 0.0498,
+      "step": 2171
+    },
+    {
+      "epoch": 8.58498023715415,
+      "grad_norm": 2.0802016258239746,
+      "learning_rate": 7.1031746031746035e-06,
+      "loss": 0.0543,
+      "step": 2172
+    },
+    {
+      "epoch": 8.588932806324111,
+      "grad_norm": 1.2858587503433228,
+      "learning_rate": 7.083333333333334e-06,
+      "loss": 0.0445,
+      "step": 2173
+    },
+    {
+      "epoch": 8.59288537549407,
+      "grad_norm": 1.2407563924789429,
+      "learning_rate": 7.063492063492063e-06,
+      "loss": 0.0393,
+      "step": 2174
+    },
+    {
+      "epoch": 8.596837944664031,
+      "grad_norm": 1.4286069869995117,
+      "learning_rate": 7.043650793650794e-06,
+      "loss": 0.0472,
+      "step": 2175
+    },
+    {
+      "epoch": 8.600790513833992,
+      "grad_norm": 2.2047770023345947,
+      "learning_rate": 7.023809523809524e-06,
+      "loss": 0.0527,
+      "step": 2176
+    },
+    {
+      "epoch": 8.604743083003953,
+      "grad_norm": 1.2905137538909912,
+      "learning_rate": 7.0039682539682545e-06,
+      "loss": 0.0412,
+      "step": 2177
+    },
+    {
+      "epoch": 8.608695652173914,
+      "grad_norm": 2.2183566093444824,
+      "learning_rate": 6.984126984126985e-06,
+      "loss": 0.0679,
+      "step": 2178
+    },
+    {
+      "epoch": 8.612648221343873,
+      "grad_norm": 1.4202773571014404,
+      "learning_rate": 6.964285714285715e-06,
+      "loss": 0.0462,
+      "step": 2179
+    },
+    {
+      "epoch": 8.616600790513834,
+      "grad_norm": 1.6715538501739502,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.0637,
+      "step": 2180
+    },
+    {
+      "epoch": 8.620553359683795,
+      "grad_norm": 1.480151891708374,
+      "learning_rate": 6.924603174603175e-06,
+      "loss": 0.0496,
+      "step": 2181
+    },
+    {
+      "epoch": 8.624505928853754,
+      "grad_norm": 1.6258138418197632,
+      "learning_rate": 6.9047619047619055e-06,
+      "loss": 0.0608,
+      "step": 2182
+    },
+    {
+      "epoch": 8.628458498023715,
+      "grad_norm": 1.6677876710891724,
+      "learning_rate": 6.884920634920636e-06,
+      "loss": 0.0547,
+      "step": 2183
+    },
+    {
+      "epoch": 8.632411067193676,
+      "grad_norm": 1.3751989603042603,
+      "learning_rate": 6.865079365079366e-06,
+      "loss": 0.0441,
+      "step": 2184
+    },
+    {
+      "epoch": 8.636363636363637,
+      "grad_norm": 1.4697294235229492,
+      "learning_rate": 6.845238095238096e-06,
+      "loss": 0.0485,
+      "step": 2185
+    },
+    {
+      "epoch": 8.640316205533598,
+      "grad_norm": 1.547059178352356,
+      "learning_rate": 6.825396825396825e-06,
+      "loss": 0.0515,
+      "step": 2186
+    },
+    {
+      "epoch": 8.644268774703557,
+      "grad_norm": 1.3061259984970093,
+      "learning_rate": 6.805555555555556e-06,
+      "loss": 0.0428,
+      "step": 2187
+    },
+    {
+      "epoch": 8.648221343873518,
+      "grad_norm": 1.2875500917434692,
+      "learning_rate": 6.785714285714285e-06,
+      "loss": 0.0422,
+      "step": 2188
+    },
+    {
+      "epoch": 8.652173913043478,
+      "grad_norm": 1.4154988527297974,
+      "learning_rate": 6.765873015873016e-06,
+      "loss": 0.051,
+      "step": 2189
+    },
+    {
+      "epoch": 8.65612648221344,
+      "grad_norm": 1.7555774450302124,
+      "learning_rate": 6.746031746031746e-06,
+      "loss": 0.0484,
+      "step": 2190
+    },
+    {
+      "epoch": 8.660079051383399,
+      "grad_norm": 1.8239620923995972,
+      "learning_rate": 6.726190476190476e-06,
+      "loss": 0.0498,
+      "step": 2191
+    },
+    {
+      "epoch": 8.66403162055336,
+      "grad_norm": 1.1520577669143677,
+      "learning_rate": 6.706349206349207e-06,
+      "loss": 0.042,
+      "step": 2192
+    },
+    {
+      "epoch": 8.66798418972332,
+      "grad_norm": 1.5023807287216187,
+      "learning_rate": 6.686507936507936e-06,
+      "loss": 0.0578,
+      "step": 2193
+    },
+    {
+      "epoch": 8.671936758893281,
+      "grad_norm": 2.02333927154541,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.05,
+      "step": 2194
+    },
+    {
+      "epoch": 8.67588932806324,
+      "grad_norm": 1.4692609310150146,
+      "learning_rate": 6.646825396825397e-06,
+      "loss": 0.0454,
+      "step": 2195
+    },
+    {
+      "epoch": 8.679841897233201,
+      "grad_norm": 1.0825679302215576,
+      "learning_rate": 6.626984126984127e-06,
+      "loss": 0.0362,
+      "step": 2196
+    },
+    {
+      "epoch": 8.683794466403162,
+      "grad_norm": 1.203744649887085,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 0.0464,
+      "step": 2197
+    },
+    {
+      "epoch": 8.687747035573123,
+      "grad_norm": 2.339820623397827,
+      "learning_rate": 6.587301587301588e-06,
+      "loss": 0.0745,
+      "step": 2198
+    },
+    {
+      "epoch": 8.691699604743082,
+      "grad_norm": 1.2836312055587769,
+      "learning_rate": 6.567460317460318e-06,
+      "loss": 0.0443,
+      "step": 2199
+    },
+    {
+      "epoch": 8.695652173913043,
+      "grad_norm": 2.841017723083496,
+      "learning_rate": 6.547619047619048e-06,
+      "loss": 0.0571,
+      "step": 2200
+    },
+    {
+      "epoch": 8.699604743083004,
+      "grad_norm": 1.3378969430923462,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 0.0398,
+      "step": 2201
+    },
+    {
+      "epoch": 8.703557312252965,
+      "grad_norm": 3.261378526687622,
+      "learning_rate": 6.507936507936509e-06,
+      "loss": 0.044,
+      "step": 2202
+    },
+    {
+      "epoch": 8.707509881422926,
+      "grad_norm": 1.5358128547668457,
+      "learning_rate": 6.488095238095239e-06,
+      "loss": 0.0509,
+      "step": 2203
+    },
+    {
+      "epoch": 8.711462450592885,
+      "grad_norm": 1.134779930114746,
+      "learning_rate": 6.4682539682539696e-06,
+      "loss": 0.0338,
+      "step": 2204
+    },
+    {
+      "epoch": 8.715415019762846,
+      "grad_norm": 1.4963998794555664,
+      "learning_rate": 6.448412698412699e-06,
+      "loss": 0.0444,
+      "step": 2205
+    },
+    {
+      "epoch": 8.719367588932807,
+      "grad_norm": 1.054849624633789,
+      "learning_rate": 6.428571428571429e-06,
+      "loss": 0.0365,
+      "step": 2206
+    },
+    {
+      "epoch": 8.723320158102768,
+      "grad_norm": 1.3599568605422974,
+      "learning_rate": 6.408730158730158e-06,
+      "loss": 0.0443,
+      "step": 2207
+    },
+    {
+      "epoch": 8.727272727272727,
+      "grad_norm": 2.381058931350708,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.0691,
+      "step": 2208
+    },
+    {
+      "epoch": 8.731225296442688,
+      "grad_norm": 1.278258204460144,
+      "learning_rate": 6.369047619047619e-06,
+      "loss": 0.051,
+      "step": 2209
+    },
+    {
+      "epoch": 8.735177865612648,
+      "grad_norm": 1.1571305990219116,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.0556,
+      "step": 2210
+    },
+    {
+      "epoch": 8.73913043478261,
+      "grad_norm": 1.4655886888504028,
+      "learning_rate": 6.32936507936508e-06,
+      "loss": 0.0484,
+      "step": 2211
+    },
+    {
+      "epoch": 8.743083003952568,
+      "grad_norm": 1.1832677125930786,
+      "learning_rate": 6.30952380952381e-06,
+      "loss": 0.0396,
+      "step": 2212
+    },
+    {
+      "epoch": 8.74703557312253,
+      "grad_norm": 1.5218229293823242,
+      "learning_rate": 6.2896825396825395e-06,
+      "loss": 0.0535,
+      "step": 2213
+    },
+    {
+      "epoch": 8.75098814229249,
+      "grad_norm": 1.2771645784378052,
+      "learning_rate": 6.26984126984127e-06,
+      "loss": 0.0454,
+      "step": 2214
+    },
+    {
+      "epoch": 8.754940711462451,
+      "grad_norm": 1.9904732704162598,
+      "learning_rate": 6.25e-06,
+      "loss": 0.0582,
+      "step": 2215
+    },
+    {
+      "epoch": 8.75889328063241,
+      "grad_norm": 1.5673877000808716,
+      "learning_rate": 6.230158730158731e-06,
+      "loss": 0.0481,
+      "step": 2216
+    },
+    {
+      "epoch": 8.762845849802371,
+      "grad_norm": 1.2909679412841797,
+      "learning_rate": 6.210317460317461e-06,
+      "loss": 0.0427,
+      "step": 2217
+    },
+    {
+      "epoch": 8.766798418972332,
+      "grad_norm": 1.0783694982528687,
+      "learning_rate": 6.190476190476191e-06,
+      "loss": 0.0379,
+      "step": 2218
+    },
+    {
+      "epoch": 8.770750988142293,
+      "grad_norm": 1.3569505214691162,
+      "learning_rate": 6.170634920634921e-06,
+      "loss": 0.0438,
+      "step": 2219
+    },
+    {
+      "epoch": 8.774703557312254,
+      "grad_norm": 1.4521833658218384,
+      "learning_rate": 6.1507936507936505e-06,
+      "loss": 0.0533,
+      "step": 2220
+    },
+    {
+      "epoch": 8.778656126482213,
+      "grad_norm": 1.4218260049819946,
+      "learning_rate": 6.130952380952381e-06,
+      "loss": 0.0499,
+      "step": 2221
+    },
+    {
+      "epoch": 8.782608695652174,
+      "grad_norm": 1.4923803806304932,
+      "learning_rate": 6.111111111111111e-06,
+      "loss": 0.066,
+      "step": 2222
+    },
+    {
+      "epoch": 8.786561264822135,
+      "grad_norm": 1.5490056276321411,
+      "learning_rate": 6.091269841269842e-06,
+      "loss": 0.0537,
+      "step": 2223
+    },
+    {
+      "epoch": 8.790513833992096,
+      "grad_norm": 2.071575164794922,
+      "learning_rate": 6.071428571428572e-06,
+      "loss": 0.0601,
+      "step": 2224
+    },
+    {
+      "epoch": 8.794466403162055,
+      "grad_norm": 1.5150130987167358,
+      "learning_rate": 6.0515873015873015e-06,
+      "loss": 0.0486,
+      "step": 2225
+    },
+    {
+      "epoch": 8.798418972332016,
+      "grad_norm": 1.5356022119522095,
+      "learning_rate": 6.031746031746032e-06,
+      "loss": 0.0446,
+      "step": 2226
+    },
+    {
+      "epoch": 8.802371541501977,
+      "grad_norm": 1.9417153596878052,
+      "learning_rate": 6.011904761904762e-06,
+      "loss": 0.0542,
+      "step": 2227
+    },
+    {
+      "epoch": 8.806324110671937,
+      "grad_norm": 1.9926491975784302,
+      "learning_rate": 5.992063492063493e-06,
+      "loss": 0.0664,
+      "step": 2228
+    },
+    {
+      "epoch": 8.810276679841897,
+      "grad_norm": 1.1277952194213867,
+      "learning_rate": 5.972222222222223e-06,
+      "loss": 0.0478,
+      "step": 2229
+    },
+    {
+      "epoch": 8.814229249011857,
+      "grad_norm": 1.2221788167953491,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 0.0366,
+      "step": 2230
+    },
+    {
+      "epoch": 8.818181818181818,
+      "grad_norm": 1.5764946937561035,
+      "learning_rate": 5.932539682539683e-06,
+      "loss": 0.0477,
+      "step": 2231
+    },
+    {
+      "epoch": 8.82213438735178,
+      "grad_norm": 1.4645206928253174,
+      "learning_rate": 5.9126984126984124e-06,
+      "loss": 0.0418,
+      "step": 2232
+    },
+    {
+      "epoch": 8.826086956521738,
+      "grad_norm": 1.9131215810775757,
+      "learning_rate": 5.892857142857143e-06,
+      "loss": 0.0689,
+      "step": 2233
+    },
+    {
+      "epoch": 8.8300395256917,
+      "grad_norm": 1.1039915084838867,
+      "learning_rate": 5.873015873015873e-06,
+      "loss": 0.035,
+      "step": 2234
+    },
+    {
+      "epoch": 8.83399209486166,
+      "grad_norm": 1.4914485216140747,
+      "learning_rate": 5.8531746031746036e-06,
+      "loss": 0.0451,
+      "step": 2235
+    },
+    {
+      "epoch": 8.837944664031621,
+      "grad_norm": 2.195981502532959,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.0672,
+      "step": 2236
+    },
+    {
+      "epoch": 8.841897233201582,
+      "grad_norm": 1.5078537464141846,
+      "learning_rate": 5.813492063492064e-06,
+      "loss": 0.0416,
+      "step": 2237
+    },
+    {
+      "epoch": 8.845849802371541,
+      "grad_norm": 1.2279962301254272,
+      "learning_rate": 5.793650793650794e-06,
+      "loss": 0.0414,
+      "step": 2238
+    },
+    {
+      "epoch": 8.849802371541502,
+      "grad_norm": 1.8902143239974976,
+      "learning_rate": 5.773809523809524e-06,
+      "loss": 0.0677,
+      "step": 2239
+    },
+    {
+      "epoch": 8.853754940711463,
+      "grad_norm": 0.9590296745300293,
+      "learning_rate": 5.753968253968254e-06,
+      "loss": 0.0366,
+      "step": 2240
+    },
+    {
+      "epoch": 8.857707509881424,
+      "grad_norm": 1.2233220338821411,
+      "learning_rate": 5.734126984126984e-06,
+      "loss": 0.0498,
+      "step": 2241
+    },
+    {
+      "epoch": 8.861660079051383,
+      "grad_norm": 1.215208888053894,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.049,
+      "step": 2242
+    },
+    {
+      "epoch": 8.865612648221344,
+      "grad_norm": 1.1311190128326416,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 0.0442,
+      "step": 2243
+    },
+    {
+      "epoch": 8.869565217391305,
+      "grad_norm": 1.8418809175491333,
+      "learning_rate": 5.674603174603174e-06,
+      "loss": 0.0535,
+      "step": 2244
+    },
+    {
+      "epoch": 8.873517786561266,
+      "grad_norm": 1.3625643253326416,
+      "learning_rate": 5.654761904761905e-06,
+      "loss": 0.0461,
+      "step": 2245
+    },
+    {
+      "epoch": 8.877470355731225,
+      "grad_norm": 1.0108672380447388,
+      "learning_rate": 5.634920634920635e-06,
+      "loss": 0.0362,
+      "step": 2246
+    },
+    {
+      "epoch": 8.881422924901186,
+      "grad_norm": 1.096070647239685,
+      "learning_rate": 5.6150793650793655e-06,
+      "loss": 0.0467,
+      "step": 2247
+    },
+    {
+      "epoch": 8.885375494071146,
+      "grad_norm": 1.4030896425247192,
+      "learning_rate": 5.595238095238096e-06,
+      "loss": 0.0459,
+      "step": 2248
+    },
+    {
+      "epoch": 8.889328063241107,
+      "grad_norm": 1.9264237880706787,
+      "learning_rate": 5.575396825396826e-06,
+      "loss": 0.0593,
+      "step": 2249
+    },
+    {
+      "epoch": 8.893280632411066,
+      "grad_norm": 1.8848391771316528,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.0654,
+      "step": 2250
+    },
+    {
+      "epoch": 8.897233201581027,
+      "grad_norm": 1.1368132829666138,
+      "learning_rate": 5.535714285714285e-06,
+      "loss": 0.0356,
+      "step": 2251
+    },
+    {
+      "epoch": 8.901185770750988,
+      "grad_norm": 1.4050102233886719,
+      "learning_rate": 5.515873015873016e-06,
+      "loss": 0.0568,
+      "step": 2252
+    },
+    {
+      "epoch": 8.90513833992095,
+      "grad_norm": 1.6391522884368896,
+      "learning_rate": 5.496031746031746e-06,
+      "loss": 0.0512,
+      "step": 2253
+    },
+    {
+      "epoch": 8.909090909090908,
+      "grad_norm": 1.2339924573898315,
+      "learning_rate": 5.4761904761904765e-06,
+      "loss": 0.0423,
+      "step": 2254
+    },
+    {
+      "epoch": 8.91304347826087,
+      "grad_norm": 1.334693431854248,
+      "learning_rate": 5.456349206349207e-06,
+      "loss": 0.043,
+      "step": 2255
+    },
+    {
+      "epoch": 8.91699604743083,
+      "grad_norm": 1.6208196878433228,
+      "learning_rate": 5.436507936507937e-06,
+      "loss": 0.0515,
+      "step": 2256
+    },
+    {
+      "epoch": 8.920948616600791,
+      "grad_norm": 1.959010124206543,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.0591,
+      "step": 2257
+    },
+    {
+      "epoch": 8.92490118577075,
+      "grad_norm": 1.2464101314544678,
+      "learning_rate": 5.396825396825397e-06,
+      "loss": 0.0444,
+      "step": 2258
+    },
+    {
+      "epoch": 8.928853754940711,
+      "grad_norm": 1.113886833190918,
+      "learning_rate": 5.3769841269841275e-06,
+      "loss": 0.045,
+      "step": 2259
+    },
+    {
+      "epoch": 8.932806324110672,
+      "grad_norm": 1.9017833471298218,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.0586,
+      "step": 2260
+    },
+    {
+      "epoch": 8.936758893280633,
+      "grad_norm": 1.6921342611312866,
+      "learning_rate": 5.337301587301587e-06,
+      "loss": 0.0678,
+      "step": 2261
+    },
+    {
+      "epoch": 8.940711462450594,
+      "grad_norm": 1.1940516233444214,
+      "learning_rate": 5.317460317460318e-06,
+      "loss": 0.0476,
+      "step": 2262
+    },
+    {
+      "epoch": 8.944664031620553,
+      "grad_norm": 1.899581789970398,
+      "learning_rate": 5.297619047619048e-06,
+      "loss": 0.0688,
+      "step": 2263
+    },
+    {
+      "epoch": 8.948616600790514,
+      "grad_norm": 1.1794816255569458,
+      "learning_rate": 5.277777777777778e-06,
+      "loss": 0.0459,
+      "step": 2264
+    },
+    {
+      "epoch": 8.952569169960475,
+      "grad_norm": 1.233652949333191,
+      "learning_rate": 5.257936507936508e-06,
+      "loss": 0.0436,
+      "step": 2265
+    },
+    {
+      "epoch": 8.956521739130435,
+      "grad_norm": 1.3263018131256104,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.046,
+      "step": 2266
+    },
+    {
+      "epoch": 8.960474308300395,
+      "grad_norm": 1.1982510089874268,
+      "learning_rate": 5.218253968253969e-06,
+      "loss": 0.0414,
+      "step": 2267
+    },
+    {
+      "epoch": 8.964426877470355,
+      "grad_norm": 2.1782376766204834,
+      "learning_rate": 5.198412698412699e-06,
+      "loss": 0.075,
+      "step": 2268
+    },
+    {
+      "epoch": 8.968379446640316,
+      "grad_norm": 1.3456032276153564,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 0.0451,
+      "step": 2269
+    },
+    {
+      "epoch": 8.972332015810277,
+      "grad_norm": 1.7961373329162598,
+      "learning_rate": 5.158730158730159e-06,
+      "loss": 0.0464,
+      "step": 2270
+    },
+    {
+      "epoch": 8.976284584980236,
+      "grad_norm": 1.0362279415130615,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 0.0374,
+      "step": 2271
+    },
+    {
+      "epoch": 8.980237154150197,
+      "grad_norm": 1.464831829071045,
+      "learning_rate": 5.119047619047619e-06,
+      "loss": 0.0419,
+      "step": 2272
+    },
+    {
+      "epoch": 8.984189723320158,
+      "grad_norm": 1.7607347965240479,
+      "learning_rate": 5.099206349206349e-06,
+      "loss": 0.053,
+      "step": 2273
+    },
+    {
+      "epoch": 8.988142292490119,
+      "grad_norm": 1.2373989820480347,
+      "learning_rate": 5.07936507936508e-06,
+      "loss": 0.05,
+      "step": 2274
+    },
+    {
+      "epoch": 8.992094861660078,
+      "grad_norm": 1.8209704160690308,
+      "learning_rate": 5.05952380952381e-06,
+      "loss": 0.0494,
+      "step": 2275
+    },
+    {
+      "epoch": 8.996047430830039,
+      "grad_norm": 1.3406612873077393,
+      "learning_rate": 5.03968253968254e-06,
+      "loss": 0.0461,
+      "step": 2276
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.2392350435256958,
+      "learning_rate": 5.01984126984127e-06,
+      "loss": 0.0467,
+      "step": 2277
+    },
+    {
+      "epoch": 9.003952569169961,
+      "grad_norm": 0.9584967494010925,
+      "learning_rate": 5e-06,
+      "loss": 0.0403,
+      "step": 2278
+    },
+    {
+      "epoch": 9.007905138339922,
+      "grad_norm": 0.5832052826881409,
+      "learning_rate": 4.98015873015873e-06,
+      "loss": 0.0321,
+      "step": 2279
+    },
+    {
+      "epoch": 9.011857707509881,
+      "grad_norm": 1.2194710969924927,
+      "learning_rate": 4.96031746031746e-06,
+      "loss": 0.0532,
+      "step": 2280
+    },
+    {
+      "epoch": 9.015810276679842,
+      "grad_norm": 0.5771675705909729,
+      "learning_rate": 4.940476190476191e-06,
+      "loss": 0.028,
+      "step": 2281
+    },
+    {
+      "epoch": 9.019762845849803,
+      "grad_norm": 1.116558313369751,
+      "learning_rate": 4.920634920634921e-06,
+      "loss": 0.0355,
+      "step": 2282
+    },
+    {
+      "epoch": 9.023715415019764,
+      "grad_norm": 0.4172620475292206,
+      "learning_rate": 4.900793650793651e-06,
+      "loss": 0.0239,
+      "step": 2283
+    },
+    {
+      "epoch": 9.027667984189723,
+      "grad_norm": 0.6424721479415894,
+      "learning_rate": 4.880952380952381e-06,
+      "loss": 0.027,
+      "step": 2284
+    },
+    {
+      "epoch": 9.031620553359684,
+      "grad_norm": 1.7028534412384033,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 0.0455,
+      "step": 2285
+    },
+    {
+      "epoch": 9.035573122529645,
+      "grad_norm": 0.9933211803436279,
+      "learning_rate": 4.841269841269842e-06,
+      "loss": 0.033,
+      "step": 2286
+    },
+    {
+      "epoch": 9.039525691699605,
+      "grad_norm": 0.9646832942962646,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 0.0386,
+      "step": 2287
+    },
+    {
+      "epoch": 9.043478260869565,
+      "grad_norm": 0.6727697253227234,
+      "learning_rate": 4.8015873015873025e-06,
+      "loss": 0.0352,
+      "step": 2288
+    },
+    {
+      "epoch": 9.047430830039525,
+      "grad_norm": 0.7543931603431702,
+      "learning_rate": 4.781746031746032e-06,
+      "loss": 0.0295,
+      "step": 2289
+    },
+    {
+      "epoch": 9.051383399209486,
+      "grad_norm": 0.46685945987701416,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.0243,
+      "step": 2290
+    },
+    {
+      "epoch": 9.055335968379447,
+      "grad_norm": 0.4817000925540924,
+      "learning_rate": 4.742063492063492e-06,
+      "loss": 0.0292,
+      "step": 2291
+    },
+    {
+      "epoch": 9.059288537549406,
+      "grad_norm": 0.8889188170433044,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.0369,
+      "step": 2292
+    },
+    {
+      "epoch": 9.063241106719367,
+      "grad_norm": 0.5639324188232422,
+      "learning_rate": 4.702380952380953e-06,
+      "loss": 0.0307,
+      "step": 2293
+    },
+    {
+      "epoch": 9.067193675889328,
+      "grad_norm": 0.6505725383758545,
+      "learning_rate": 4.682539682539683e-06,
+      "loss": 0.0313,
+      "step": 2294
+    },
+    {
+      "epoch": 9.071146245059289,
+      "grad_norm": 0.9556856751441956,
+      "learning_rate": 4.662698412698413e-06,
+      "loss": 0.0463,
+      "step": 2295
+    },
+    {
+      "epoch": 9.075098814229248,
+      "grad_norm": 0.49648115038871765,
+      "learning_rate": 4.642857142857143e-06,
+      "loss": 0.0246,
+      "step": 2296
+    },
+    {
+      "epoch": 9.079051383399209,
+      "grad_norm": 0.877655029296875,
+      "learning_rate": 4.623015873015873e-06,
+      "loss": 0.0309,
+      "step": 2297
+    },
+    {
+      "epoch": 9.08300395256917,
+      "grad_norm": 1.1071048974990845,
+      "learning_rate": 4.603174603174604e-06,
+      "loss": 0.0437,
+      "step": 2298
+    },
+    {
+      "epoch": 9.08695652173913,
+      "grad_norm": 1.238307237625122,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.0346,
+      "step": 2299
+    },
+    {
+      "epoch": 9.090909090909092,
+      "grad_norm": 0.9343061447143555,
+      "learning_rate": 4.563492063492064e-06,
+      "loss": 0.0306,
+      "step": 2300
+    },
+    {
+      "epoch": 9.09486166007905,
+      "grad_norm": 1.3237608671188354,
+      "learning_rate": 4.543650793650794e-06,
+      "loss": 0.0405,
+      "step": 2301
+    },
+    {
+      "epoch": 9.098814229249012,
+      "grad_norm": 1.2943129539489746,
+      "learning_rate": 4.5238095238095235e-06,
+      "loss": 0.0414,
+      "step": 2302
+    },
+    {
+      "epoch": 9.102766798418973,
+      "grad_norm": 0.4701171815395355,
+      "learning_rate": 4.503968253968254e-06,
+      "loss": 0.0275,
+      "step": 2303
+    },
+    {
+      "epoch": 9.106719367588934,
+      "grad_norm": 1.5329245328903198,
+      "learning_rate": 4.484126984126984e-06,
+      "loss": 0.0412,
+      "step": 2304
+    },
+    {
+      "epoch": 9.110671936758893,
+      "grad_norm": 1.2468199729919434,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 0.0361,
+      "step": 2305
+    },
+    {
+      "epoch": 9.114624505928854,
+      "grad_norm": 0.4569074511528015,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.0251,
+      "step": 2306
+    },
+    {
+      "epoch": 9.118577075098814,
+      "grad_norm": 0.7745764255523682,
+      "learning_rate": 4.424603174603175e-06,
+      "loss": 0.033,
+      "step": 2307
+    },
+    {
+      "epoch": 9.122529644268775,
+      "grad_norm": 0.8108665943145752,
+      "learning_rate": 4.404761904761905e-06,
+      "loss": 0.0239,
+      "step": 2308
+    },
+    {
+      "epoch": 9.126482213438734,
+      "grad_norm": 0.8088359832763672,
+      "learning_rate": 4.3849206349206344e-06,
+      "loss": 0.0303,
+      "step": 2309
+    },
+    {
+      "epoch": 9.130434782608695,
+      "grad_norm": 0.9375463724136353,
+      "learning_rate": 4.365079365079365e-06,
+      "loss": 0.0593,
+      "step": 2310
+    },
+    {
+      "epoch": 9.134387351778656,
+      "grad_norm": 1.1133482456207275,
+      "learning_rate": 4.345238095238095e-06,
+      "loss": 0.0357,
+      "step": 2311
+    },
+    {
+      "epoch": 9.138339920948617,
+      "grad_norm": 0.5319989919662476,
+      "learning_rate": 4.3253968253968256e-06,
+      "loss": 0.029,
+      "step": 2312
+    },
+    {
+      "epoch": 9.142292490118576,
+      "grad_norm": 0.9559447169303894,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 0.0377,
+      "step": 2313
+    },
+    {
+      "epoch": 9.146245059288537,
+      "grad_norm": 1.1195396184921265,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.0254,
+      "step": 2314
+    },
+    {
+      "epoch": 9.150197628458498,
+      "grad_norm": 0.6906282901763916,
+      "learning_rate": 4.265873015873016e-06,
+      "loss": 0.0417,
+      "step": 2315
+    },
+    {
+      "epoch": 9.154150197628459,
+      "grad_norm": 0.730162501335144,
+      "learning_rate": 4.246031746031746e-06,
+      "loss": 0.0304,
+      "step": 2316
+    },
+    {
+      "epoch": 9.15810276679842,
+      "grad_norm": 0.8155921697616577,
+      "learning_rate": 4.226190476190477e-06,
+      "loss": 0.0334,
+      "step": 2317
+    },
+    {
+      "epoch": 9.162055335968379,
+      "grad_norm": 1.145893931388855,
+      "learning_rate": 4.206349206349207e-06,
+      "loss": 0.0472,
+      "step": 2318
+    },
+    {
+      "epoch": 9.16600790513834,
+      "grad_norm": 0.6284162998199463,
+      "learning_rate": 4.1865079365079365e-06,
+      "loss": 0.0302,
+      "step": 2319
+    },
+    {
+      "epoch": 9.1699604743083,
+      "grad_norm": 3.132331371307373,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.0671,
+      "step": 2320
+    },
+    {
+      "epoch": 9.173913043478262,
+      "grad_norm": 0.725439190864563,
+      "learning_rate": 4.146825396825397e-06,
+      "loss": 0.029,
+      "step": 2321
+    },
+    {
+      "epoch": 9.17786561264822,
+      "grad_norm": 0.6807708144187927,
+      "learning_rate": 4.126984126984127e-06,
+      "loss": 0.0265,
+      "step": 2322
+    },
+    {
+      "epoch": 9.181818181818182,
+      "grad_norm": 1.1239207983016968,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 0.0428,
+      "step": 2323
+    },
+    {
+      "epoch": 9.185770750988143,
+      "grad_norm": 0.6186617612838745,
+      "learning_rate": 4.0873015873015875e-06,
+      "loss": 0.026,
+      "step": 2324
+    },
+    {
+      "epoch": 9.189723320158103,
+      "grad_norm": 0.5834595561027527,
+      "learning_rate": 4.067460317460318e-06,
+      "loss": 0.0281,
+      "step": 2325
+    },
+    {
+      "epoch": 9.193675889328063,
+      "grad_norm": 0.7775494456291199,
+      "learning_rate": 4.047619047619048e-06,
+      "loss": 0.027,
+      "step": 2326
+    },
+    {
+      "epoch": 9.197628458498023,
+      "grad_norm": 0.5415392518043518,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 0.0291,
+      "step": 2327
+    },
+    {
+      "epoch": 9.201581027667984,
+      "grad_norm": 1.0766407251358032,
+      "learning_rate": 4.007936507936508e-06,
+      "loss": 0.0347,
+      "step": 2328
+    },
+    {
+      "epoch": 9.205533596837945,
+      "grad_norm": 0.705011785030365,
+      "learning_rate": 3.988095238095238e-06,
+      "loss": 0.0274,
+      "step": 2329
+    },
+    {
+      "epoch": 9.209486166007904,
+      "grad_norm": 1.0642330646514893,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.0432,
+      "step": 2330
+    },
+    {
+      "epoch": 9.213438735177865,
+      "grad_norm": 0.8075931072235107,
+      "learning_rate": 3.9484126984126985e-06,
+      "loss": 0.031,
+      "step": 2331
+    },
+    {
+      "epoch": 9.217391304347826,
+      "grad_norm": 0.5113716721534729,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.0224,
+      "step": 2332
+    },
+    {
+      "epoch": 9.221343873517787,
+      "grad_norm": 0.9608832597732544,
+      "learning_rate": 3.908730158730159e-06,
+      "loss": 0.0319,
+      "step": 2333
+    },
+    {
+      "epoch": 9.225296442687746,
+      "grad_norm": 1.2569172382354736,
+      "learning_rate": 3.888888888888889e-06,
+      "loss": 0.0328,
+      "step": 2334
+    },
+    {
+      "epoch": 9.229249011857707,
+      "grad_norm": 0.7299090027809143,
+      "learning_rate": 3.869047619047619e-06,
+      "loss": 0.0258,
+      "step": 2335
+    },
+    {
+      "epoch": 9.233201581027668,
+      "grad_norm": 0.5903069972991943,
+      "learning_rate": 3.8492063492063495e-06,
+      "loss": 0.0313,
+      "step": 2336
+    },
+    {
+      "epoch": 9.237154150197629,
+      "grad_norm": 0.715298056602478,
+      "learning_rate": 3.82936507936508e-06,
+      "loss": 0.042,
+      "step": 2337
+    },
+    {
+      "epoch": 9.24110671936759,
+      "grad_norm": 0.6976611018180847,
+      "learning_rate": 3.8095238095238102e-06,
+      "loss": 0.0399,
+      "step": 2338
+    },
+    {
+      "epoch": 9.245059288537549,
+      "grad_norm": 0.9303411245346069,
+      "learning_rate": 3.7896825396825398e-06,
+      "loss": 0.0319,
+      "step": 2339
+    },
+    {
+      "epoch": 9.24901185770751,
+      "grad_norm": 0.706676721572876,
+      "learning_rate": 3.7698412698412697e-06,
+      "loss": 0.0332,
+      "step": 2340
+    },
+    {
+      "epoch": 9.25296442687747,
+      "grad_norm": 0.553431510925293,
+      "learning_rate": 3.75e-06,
+      "loss": 0.0339,
+      "step": 2341
+    },
+    {
+      "epoch": 9.256916996047432,
+      "grad_norm": 1.1647964715957642,
+      "learning_rate": 3.7301587301587305e-06,
+      "loss": 0.0277,
+      "step": 2342
+    },
+    {
+      "epoch": 9.26086956521739,
+      "grad_norm": 0.5181522369384766,
+      "learning_rate": 3.7103174603174604e-06,
+      "loss": 0.0231,
+      "step": 2343
+    },
+    {
+      "epoch": 9.264822134387352,
+      "grad_norm": 1.1019365787506104,
+      "learning_rate": 3.690476190476191e-06,
+      "loss": 0.0654,
+      "step": 2344
+    },
+    {
+      "epoch": 9.268774703557312,
+      "grad_norm": 0.6852990984916687,
+      "learning_rate": 3.6706349206349208e-06,
+      "loss": 0.0254,
+      "step": 2345
+    },
+    {
+      "epoch": 9.272727272727273,
+      "grad_norm": 0.6852630972862244,
+      "learning_rate": 3.650793650793651e-06,
+      "loss": 0.028,
+      "step": 2346
+    },
+    {
+      "epoch": 9.276679841897232,
+      "grad_norm": 0.79278165102005,
+      "learning_rate": 3.6309523809523815e-06,
+      "loss": 0.0321,
+      "step": 2347
+    },
+    {
+      "epoch": 9.280632411067193,
+      "grad_norm": 1.3321659564971924,
+      "learning_rate": 3.611111111111111e-06,
+      "loss": 0.0305,
+      "step": 2348
+    },
+    {
+      "epoch": 9.284584980237154,
+      "grad_norm": 1.0156902074813843,
+      "learning_rate": 3.591269841269841e-06,
+      "loss": 0.0331,
+      "step": 2349
+    },
+    {
+      "epoch": 9.288537549407115,
+      "grad_norm": 0.9449909329414368,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.0291,
+      "step": 2350
+    },
+    {
+      "epoch": 9.292490118577074,
+      "grad_norm": 0.6974058747291565,
+      "learning_rate": 3.5515873015873017e-06,
+      "loss": 0.0369,
+      "step": 2351
+    },
+    {
+      "epoch": 9.296442687747035,
+      "grad_norm": 0.6614970564842224,
+      "learning_rate": 3.5317460317460317e-06,
+      "loss": 0.0245,
+      "step": 2352
+    },
+    {
+      "epoch": 9.300395256916996,
+      "grad_norm": 1.0674023628234863,
+      "learning_rate": 3.511904761904762e-06,
+      "loss": 0.0361,
+      "step": 2353
+    },
+    {
+      "epoch": 9.304347826086957,
+      "grad_norm": 0.8966097831726074,
+      "learning_rate": 3.4920634920634924e-06,
+      "loss": 0.0399,
+      "step": 2354
+    },
+    {
+      "epoch": 9.308300395256918,
+      "grad_norm": 0.8403730988502502,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.0403,
+      "step": 2355
+    },
+    {
+      "epoch": 9.312252964426877,
+      "grad_norm": 0.9505050182342529,
+      "learning_rate": 3.4523809523809528e-06,
+      "loss": 0.0277,
+      "step": 2356
+    },
+    {
+      "epoch": 9.316205533596838,
+      "grad_norm": 0.5641964673995972,
+      "learning_rate": 3.432539682539683e-06,
+      "loss": 0.0316,
+      "step": 2357
+    },
+    {
+      "epoch": 9.320158102766799,
+      "grad_norm": 0.9463332891464233,
+      "learning_rate": 3.4126984126984127e-06,
+      "loss": 0.0395,
+      "step": 2358
+    },
+    {
+      "epoch": 9.32411067193676,
+      "grad_norm": 1.2878776788711548,
+      "learning_rate": 3.3928571428571426e-06,
+      "loss": 0.0397,
+      "step": 2359
+    },
+    {
+      "epoch": 9.328063241106719,
+      "grad_norm": 1.2284893989562988,
+      "learning_rate": 3.373015873015873e-06,
+      "loss": 0.0421,
+      "step": 2360
+    },
+    {
+      "epoch": 9.33201581027668,
+      "grad_norm": 0.9104984402656555,
+      "learning_rate": 3.3531746031746034e-06,
+      "loss": 0.0384,
+      "step": 2361
+    },
+    {
+      "epoch": 9.33596837944664,
+      "grad_norm": 0.7004300951957703,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.0506,
+      "step": 2362
+    },
+    {
+      "epoch": 9.339920948616601,
+      "grad_norm": 0.9767426252365112,
+      "learning_rate": 3.3134920634920637e-06,
+      "loss": 0.0415,
+      "step": 2363
+    },
+    {
+      "epoch": 9.34387351778656,
+      "grad_norm": 0.7821481227874756,
+      "learning_rate": 3.293650793650794e-06,
+      "loss": 0.031,
+      "step": 2364
+    },
+    {
+      "epoch": 9.347826086956522,
+      "grad_norm": 0.9590467810630798,
+      "learning_rate": 3.273809523809524e-06,
+      "loss": 0.0393,
+      "step": 2365
+    },
+    {
+      "epoch": 9.351778656126482,
+      "grad_norm": 0.6908400654792786,
+      "learning_rate": 3.2539682539682544e-06,
+      "loss": 0.0405,
+      "step": 2366
+    },
+    {
+      "epoch": 9.355731225296443,
+      "grad_norm": 0.8684214949607849,
+      "learning_rate": 3.2341269841269848e-06,
+      "loss": 0.0313,
+      "step": 2367
+    },
+    {
+      "epoch": 9.359683794466402,
+      "grad_norm": 0.5750323534011841,
+      "learning_rate": 3.2142857142857143e-06,
+      "loss": 0.0323,
+      "step": 2368
+    },
+    {
+      "epoch": 9.363636363636363,
+      "grad_norm": 0.7478132247924805,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 0.0299,
+      "step": 2369
+    },
+    {
+      "epoch": 9.367588932806324,
+      "grad_norm": 0.4561312794685364,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.0263,
+      "step": 2370
+    },
+    {
+      "epoch": 9.371541501976285,
+      "grad_norm": 0.856050968170166,
+      "learning_rate": 3.154761904761905e-06,
+      "loss": 0.032,
+      "step": 2371
+    },
+    {
+      "epoch": 9.375494071146244,
+      "grad_norm": 0.4932451844215393,
+      "learning_rate": 3.134920634920635e-06,
+      "loss": 0.0264,
+      "step": 2372
+    },
+    {
+      "epoch": 9.379446640316205,
+      "grad_norm": 0.6153606176376343,
+      "learning_rate": 3.1150793650793653e-06,
+      "loss": 0.0281,
+      "step": 2373
+    },
+    {
+      "epoch": 9.383399209486166,
+      "grad_norm": 0.8731127381324768,
+      "learning_rate": 3.0952380952380953e-06,
+      "loss": 0.0324,
+      "step": 2374
+    },
+    {
+      "epoch": 9.387351778656127,
+      "grad_norm": 0.7539214491844177,
+      "learning_rate": 3.0753968253968252e-06,
+      "loss": 0.0375,
+      "step": 2375
+    },
+    {
+      "epoch": 9.391304347826088,
+      "grad_norm": 1.1730583906173706,
+      "learning_rate": 3.0555555555555556e-06,
+      "loss": 0.0359,
+      "step": 2376
+    },
+    {
+      "epoch": 9.395256916996047,
+      "grad_norm": 0.7655700445175171,
+      "learning_rate": 3.035714285714286e-06,
+      "loss": 0.0297,
+      "step": 2377
+    },
+    {
+      "epoch": 9.399209486166008,
+      "grad_norm": 0.5464181900024414,
+      "learning_rate": 3.015873015873016e-06,
+      "loss": 0.0272,
+      "step": 2378
+    },
+    {
+      "epoch": 9.403162055335969,
+      "grad_norm": 1.2985944747924805,
+      "learning_rate": 2.9960317460317463e-06,
+      "loss": 0.0368,
+      "step": 2379
+    },
+    {
+      "epoch": 9.40711462450593,
+      "grad_norm": 0.7202474474906921,
+      "learning_rate": 2.9761904761904763e-06,
+      "loss": 0.0431,
+      "step": 2380
+    },
+    {
+      "epoch": 9.411067193675889,
+      "grad_norm": 0.859272301197052,
+      "learning_rate": 2.9563492063492062e-06,
+      "loss": 0.0339,
+      "step": 2381
+    },
+    {
+      "epoch": 9.41501976284585,
+      "grad_norm": 0.9698725342750549,
+      "learning_rate": 2.9365079365079366e-06,
+      "loss": 0.0264,
+      "step": 2382
+    },
+    {
+      "epoch": 9.41897233201581,
+      "grad_norm": 0.8615301251411438,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.0291,
+      "step": 2383
+    },
+    {
+      "epoch": 9.422924901185771,
+      "grad_norm": 0.5525624752044678,
+      "learning_rate": 2.896825396825397e-06,
+      "loss": 0.0368,
+      "step": 2384
+    },
+    {
+      "epoch": 9.42687747035573,
+      "grad_norm": 1.0065622329711914,
+      "learning_rate": 2.876984126984127e-06,
+      "loss": 0.0386,
+      "step": 2385
+    },
+    {
+      "epoch": 9.430830039525691,
+      "grad_norm": 1.2710402011871338,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.0432,
+      "step": 2386
+    },
+    {
+      "epoch": 9.434782608695652,
+      "grad_norm": 0.5823608636856079,
+      "learning_rate": 2.837301587301587e-06,
+      "loss": 0.0302,
+      "step": 2387
+    },
+    {
+      "epoch": 9.438735177865613,
+      "grad_norm": 0.8949801325798035,
+      "learning_rate": 2.8174603174603176e-06,
+      "loss": 0.0464,
+      "step": 2388
+    },
+    {
+      "epoch": 9.442687747035572,
+      "grad_norm": 0.6641007661819458,
+      "learning_rate": 2.797619047619048e-06,
+      "loss": 0.0232,
+      "step": 2389
+    },
+    {
+      "epoch": 9.446640316205533,
+      "grad_norm": 0.9579842686653137,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.0421,
+      "step": 2390
+    },
+    {
+      "epoch": 9.450592885375494,
+      "grad_norm": 0.843721866607666,
+      "learning_rate": 2.757936507936508e-06,
+      "loss": 0.0329,
+      "step": 2391
+    },
+    {
+      "epoch": 9.454545454545455,
+      "grad_norm": 0.8626483082771301,
+      "learning_rate": 2.7380952380952382e-06,
+      "loss": 0.0407,
+      "step": 2392
+    },
+    {
+      "epoch": 9.458498023715414,
+      "grad_norm": 0.610456109046936,
+      "learning_rate": 2.7182539682539686e-06,
+      "loss": 0.0307,
+      "step": 2393
+    },
+    {
+      "epoch": 9.462450592885375,
+      "grad_norm": 0.6262418031692505,
+      "learning_rate": 2.6984126984126986e-06,
+      "loss": 0.0261,
+      "step": 2394
+    },
+    {
+      "epoch": 9.466403162055336,
+      "grad_norm": 0.5422096848487854,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 0.0271,
+      "step": 2395
+    },
+    {
+      "epoch": 9.470355731225297,
+      "grad_norm": 0.46671536564826965,
+      "learning_rate": 2.658730158730159e-06,
+      "loss": 0.0212,
+      "step": 2396
+    },
+    {
+      "epoch": 9.474308300395258,
+      "grad_norm": 0.6522403359413147,
+      "learning_rate": 2.638888888888889e-06,
+      "loss": 0.0346,
+      "step": 2397
+    },
+    {
+      "epoch": 9.478260869565217,
+      "grad_norm": 1.04923677444458,
+      "learning_rate": 2.6190476190476192e-06,
+      "loss": 0.0386,
+      "step": 2398
+    },
+    {
+      "epoch": 9.482213438735178,
+      "grad_norm": 0.817672610282898,
+      "learning_rate": 2.5992063492063496e-06,
+      "loss": 0.0375,
+      "step": 2399
+    },
+    {
+      "epoch": 9.486166007905139,
+      "grad_norm": 0.4436691105365753,
+      "learning_rate": 2.5793650793650795e-06,
+      "loss": 0.0337,
+      "step": 2400
+    },
+    {
+      "epoch": 9.4901185770751,
+      "grad_norm": 0.5380107164382935,
+      "learning_rate": 2.5595238095238095e-06,
+      "loss": 0.0252,
+      "step": 2401
+    },
+    {
+      "epoch": 9.494071146245059,
+      "grad_norm": 0.5638983845710754,
+      "learning_rate": 2.53968253968254e-06,
+      "loss": 0.0258,
+      "step": 2402
+    },
+    {
+      "epoch": 9.49802371541502,
+      "grad_norm": 0.5976006388664246,
+      "learning_rate": 2.51984126984127e-06,
+      "loss": 0.026,
+      "step": 2403
+    },
+    {
+      "epoch": 9.50197628458498,
+      "grad_norm": 0.8063969612121582,
+      "learning_rate": 2.5e-06,
+      "loss": 0.035,
+      "step": 2404
+    },
+    {
+      "epoch": 9.505928853754941,
+      "grad_norm": 0.7016451954841614,
+      "learning_rate": 2.48015873015873e-06,
+      "loss": 0.031,
+      "step": 2405
+    },
+    {
+      "epoch": 9.5098814229249,
+      "grad_norm": 0.6681998372077942,
+      "learning_rate": 2.4603174603174605e-06,
+      "loss": 0.0293,
+      "step": 2406
+    },
+    {
+      "epoch": 9.513833992094861,
+      "grad_norm": 0.8126972913742065,
+      "learning_rate": 2.4404761904761905e-06,
+      "loss": 0.0296,
+      "step": 2407
+    },
+    {
+      "epoch": 9.517786561264822,
+      "grad_norm": 0.6644697189331055,
+      "learning_rate": 2.420634920634921e-06,
+      "loss": 0.028,
+      "step": 2408
+    },
+    {
+      "epoch": 9.521739130434783,
+      "grad_norm": 0.7953290939331055,
+      "learning_rate": 2.4007936507936512e-06,
+      "loss": 0.0244,
+      "step": 2409
+    },
+    {
+      "epoch": 9.525691699604742,
+      "grad_norm": 0.8905209302902222,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.0629,
+      "step": 2410
+    },
+    {
+      "epoch": 9.529644268774703,
+      "grad_norm": 0.6507531404495239,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 0.026,
+      "step": 2411
+    },
+    {
+      "epoch": 9.533596837944664,
+      "grad_norm": 0.6166356205940247,
+      "learning_rate": 2.3412698412698415e-06,
+      "loss": 0.0342,
+      "step": 2412
+    },
+    {
+      "epoch": 9.537549407114625,
+      "grad_norm": 0.6299740076065063,
+      "learning_rate": 2.3214285714285715e-06,
+      "loss": 0.0307,
+      "step": 2413
+    },
+    {
+      "epoch": 9.541501976284586,
+      "grad_norm": 0.6078210473060608,
+      "learning_rate": 2.301587301587302e-06,
+      "loss": 0.0341,
+      "step": 2414
+    },
+    {
+      "epoch": 9.545454545454545,
+      "grad_norm": 1.263504981994629,
+      "learning_rate": 2.281746031746032e-06,
+      "loss": 0.0389,
+      "step": 2415
+    },
+    {
+      "epoch": 9.549407114624506,
+      "grad_norm": 0.6518561244010925,
+      "learning_rate": 2.2619047619047617e-06,
+      "loss": 0.0344,
+      "step": 2416
+    },
+    {
+      "epoch": 9.553359683794467,
+      "grad_norm": 0.9337310194969177,
+      "learning_rate": 2.242063492063492e-06,
+      "loss": 0.0299,
+      "step": 2417
+    },
+    {
+      "epoch": 9.557312252964428,
+      "grad_norm": 0.6526855826377869,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 0.0307,
+      "step": 2418
+    },
+    {
+      "epoch": 9.561264822134387,
+      "grad_norm": 0.8223960399627686,
+      "learning_rate": 2.2023809523809525e-06,
+      "loss": 0.0279,
+      "step": 2419
+    },
+    {
+      "epoch": 9.565217391304348,
+      "grad_norm": 0.6811292171478271,
+      "learning_rate": 2.1825396825396824e-06,
+      "loss": 0.0301,
+      "step": 2420
+    },
+    {
+      "epoch": 9.569169960474309,
+      "grad_norm": 0.9670735597610474,
+      "learning_rate": 2.1626984126984128e-06,
+      "loss": 0.031,
+      "step": 2421
+    },
+    {
+      "epoch": 9.57312252964427,
+      "grad_norm": 0.5966360569000244,
+      "learning_rate": 2.142857142857143e-06,
+      "loss": 0.0266,
+      "step": 2422
+    },
+    {
+      "epoch": 9.577075098814229,
+      "grad_norm": 0.6706477999687195,
+      "learning_rate": 2.123015873015873e-06,
+      "loss": 0.0356,
+      "step": 2423
+    },
+    {
+      "epoch": 9.58102766798419,
+      "grad_norm": 0.7263142466545105,
+      "learning_rate": 2.1031746031746035e-06,
+      "loss": 0.0336,
+      "step": 2424
+    },
+    {
+      "epoch": 9.58498023715415,
+      "grad_norm": 0.509760856628418,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.0276,
+      "step": 2425
+    },
+    {
+      "epoch": 9.588932806324111,
+      "grad_norm": 0.6894228458404541,
+      "learning_rate": 2.0634920634920634e-06,
+      "loss": 0.0322,
+      "step": 2426
+    },
+    {
+      "epoch": 9.59288537549407,
+      "grad_norm": 0.751865029335022,
+      "learning_rate": 2.0436507936507938e-06,
+      "loss": 0.0364,
+      "step": 2427
+    },
+    {
+      "epoch": 9.596837944664031,
+      "grad_norm": 0.9960070848464966,
+      "learning_rate": 2.023809523809524e-06,
+      "loss": 0.0268,
+      "step": 2428
+    },
+    {
+      "epoch": 9.600790513833992,
+      "grad_norm": 1.3808013200759888,
+      "learning_rate": 2.003968253968254e-06,
+      "loss": 0.0395,
+      "step": 2429
+    },
+    {
+      "epoch": 9.604743083003953,
+      "grad_norm": 0.6000566482543945,
+      "learning_rate": 1.984126984126984e-06,
+      "loss": 0.0259,
+      "step": 2430
+    },
+    {
+      "epoch": 9.608695652173914,
+      "grad_norm": 0.715231716632843,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 0.0297,
+      "step": 2431
+    },
+    {
+      "epoch": 9.612648221343873,
+      "grad_norm": 0.8936877250671387,
+      "learning_rate": 1.9444444444444444e-06,
+      "loss": 0.0346,
+      "step": 2432
+    },
+    {
+      "epoch": 9.616600790513834,
+      "grad_norm": 0.7350375056266785,
+      "learning_rate": 1.9246031746031747e-06,
+      "loss": 0.0375,
+      "step": 2433
+    },
+    {
+      "epoch": 9.620553359683795,
+      "grad_norm": 0.7897645235061646,
+      "learning_rate": 1.9047619047619051e-06,
+      "loss": 0.0312,
+      "step": 2434
+    },
+    {
+      "epoch": 9.624505928853754,
+      "grad_norm": 0.4184766113758087,
+      "learning_rate": 1.8849206349206349e-06,
+      "loss": 0.0244,
+      "step": 2435
+    },
+    {
+      "epoch": 9.628458498023715,
+      "grad_norm": 0.7881268262863159,
+      "learning_rate": 1.8650793650793652e-06,
+      "loss": 0.0385,
+      "step": 2436
+    },
+    {
+      "epoch": 9.632411067193676,
+      "grad_norm": 0.83879554271698,
+      "learning_rate": 1.8452380952380954e-06,
+      "loss": 0.0278,
+      "step": 2437
+    },
+    {
+      "epoch": 9.636363636363637,
+      "grad_norm": 0.9279481768608093,
+      "learning_rate": 1.8253968253968256e-06,
+      "loss": 0.0288,
+      "step": 2438
+    },
+    {
+      "epoch": 9.640316205533598,
+      "grad_norm": 1.2790271043777466,
+      "learning_rate": 1.8055555555555555e-06,
+      "loss": 0.0289,
+      "step": 2439
+    },
+    {
+      "epoch": 9.644268774703557,
+      "grad_norm": 0.5256231427192688,
+      "learning_rate": 1.7857142857142857e-06,
+      "loss": 0.0262,
+      "step": 2440
+    },
+    {
+      "epoch": 9.648221343873518,
+      "grad_norm": 0.5460281372070312,
+      "learning_rate": 1.7658730158730158e-06,
+      "loss": 0.0312,
+      "step": 2441
+    },
+    {
+      "epoch": 9.652173913043478,
+      "grad_norm": 0.8667499423027039,
+      "learning_rate": 1.7460317460317462e-06,
+      "loss": 0.0355,
+      "step": 2442
+    },
+    {
+      "epoch": 9.65612648221344,
+      "grad_norm": 1.1661618947982788,
+      "learning_rate": 1.7261904761904764e-06,
+      "loss": 0.0397,
+      "step": 2443
+    },
+    {
+      "epoch": 9.660079051383399,
+      "grad_norm": 0.5786067843437195,
+      "learning_rate": 1.7063492063492063e-06,
+      "loss": 0.0255,
+      "step": 2444
+    },
+    {
+      "epoch": 9.66403162055336,
+      "grad_norm": 1.2098567485809326,
+      "learning_rate": 1.6865079365079365e-06,
+      "loss": 0.0829,
+      "step": 2445
+    },
+    {
+      "epoch": 9.66798418972332,
+      "grad_norm": 1.0478078126907349,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.0351,
+      "step": 2446
+    },
+    {
+      "epoch": 9.671936758893281,
+      "grad_norm": 0.5507435202598572,
+      "learning_rate": 1.646825396825397e-06,
+      "loss": 0.0268,
+      "step": 2447
+    },
+    {
+      "epoch": 9.67588932806324,
+      "grad_norm": 0.4946918785572052,
+      "learning_rate": 1.6269841269841272e-06,
+      "loss": 0.0291,
+      "step": 2448
+    },
+    {
+      "epoch": 9.679841897233201,
+      "grad_norm": 1.1714826822280884,
+      "learning_rate": 1.6071428571428572e-06,
+      "loss": 0.0418,
+      "step": 2449
+    },
+    {
+      "epoch": 9.683794466403162,
+      "grad_norm": 0.6348810791969299,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.0329,
+      "step": 2450
+    },
+    {
+      "epoch": 9.687747035573123,
+      "grad_norm": 1.8041942119598389,
+      "learning_rate": 1.5674603174603175e-06,
+      "loss": 0.0434,
+      "step": 2451
+    },
+    {
+      "epoch": 9.691699604743082,
+      "grad_norm": 0.9931659698486328,
+      "learning_rate": 1.5476190476190476e-06,
+      "loss": 0.0405,
+      "step": 2452
+    },
+    {
+      "epoch": 9.695652173913043,
+      "grad_norm": 0.7843307256698608,
+      "learning_rate": 1.5277777777777778e-06,
+      "loss": 0.0316,
+      "step": 2453
+    },
+    {
+      "epoch": 9.699604743083004,
+      "grad_norm": 0.6582829356193542,
+      "learning_rate": 1.507936507936508e-06,
+      "loss": 0.0373,
+      "step": 2454
+    },
+    {
+      "epoch": 9.703557312252965,
+      "grad_norm": 1.3511526584625244,
+      "learning_rate": 1.4880952380952381e-06,
+      "loss": 0.0397,
+      "step": 2455
+    },
+    {
+      "epoch": 9.707509881422926,
+      "grad_norm": 0.6811285614967346,
+      "learning_rate": 1.4682539682539683e-06,
+      "loss": 0.0257,
+      "step": 2456
+    },
+    {
+      "epoch": 9.711462450592885,
+      "grad_norm": 1.0743825435638428,
+      "learning_rate": 1.4484126984126985e-06,
+      "loss": 0.0378,
+      "step": 2457
+    },
+    {
+      "epoch": 9.715415019762846,
+      "grad_norm": 0.5769475102424622,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.0271,
+      "step": 2458
+    },
+    {
+      "epoch": 9.719367588932807,
+      "grad_norm": 0.5640432834625244,
+      "learning_rate": 1.4087301587301588e-06,
+      "loss": 0.023,
+      "step": 2459
+    },
+    {
+      "epoch": 9.723320158102768,
+      "grad_norm": 0.6095978021621704,
+      "learning_rate": 1.388888888888889e-06,
+      "loss": 0.0368,
+      "step": 2460
+    },
+    {
+      "epoch": 9.727272727272727,
+      "grad_norm": 0.6505516767501831,
+      "learning_rate": 1.3690476190476191e-06,
+      "loss": 0.025,
+      "step": 2461
+    },
+    {
+      "epoch": 9.731225296442688,
+      "grad_norm": 0.7247635722160339,
+      "learning_rate": 1.3492063492063493e-06,
+      "loss": 0.0271,
+      "step": 2462
+    },
+    {
+      "epoch": 9.735177865612648,
+      "grad_norm": 0.7161766290664673,
+      "learning_rate": 1.3293650793650794e-06,
+      "loss": 0.0316,
+      "step": 2463
+    },
+    {
+      "epoch": 9.73913043478261,
+      "grad_norm": 0.8741920590400696,
+      "learning_rate": 1.3095238095238096e-06,
+      "loss": 0.0356,
+      "step": 2464
+    },
+    {
+      "epoch": 9.743083003952568,
+      "grad_norm": 0.7110796570777893,
+      "learning_rate": 1.2896825396825398e-06,
+      "loss": 0.0312,
+      "step": 2465
+    },
+    {
+      "epoch": 9.74703557312253,
+      "grad_norm": 0.5976476669311523,
+      "learning_rate": 1.26984126984127e-06,
+      "loss": 0.0341,
+      "step": 2466
+    },
+    {
+      "epoch": 9.75098814229249,
+      "grad_norm": 1.2879040241241455,
+      "learning_rate": 1.25e-06,
+      "loss": 0.0361,
+      "step": 2467
+    },
+    {
+      "epoch": 9.754940711462451,
+      "grad_norm": 0.5123355388641357,
+      "learning_rate": 1.2301587301587303e-06,
+      "loss": 0.0235,
+      "step": 2468
+    },
+    {
+      "epoch": 9.75889328063241,
+      "grad_norm": 1.0054389238357544,
+      "learning_rate": 1.2103174603174604e-06,
+      "loss": 0.0441,
+      "step": 2469
+    },
+    {
+      "epoch": 9.762845849802371,
+      "grad_norm": 0.8575003147125244,
+      "learning_rate": 1.1904761904761904e-06,
+      "loss": 0.0337,
+      "step": 2470
+    },
+    {
+      "epoch": 9.766798418972332,
+      "grad_norm": 0.6102113127708435,
+      "learning_rate": 1.1706349206349208e-06,
+      "loss": 0.0359,
+      "step": 2471
+    },
+    {
+      "epoch": 9.770750988142293,
+      "grad_norm": 1.0554832220077515,
+      "learning_rate": 1.150793650793651e-06,
+      "loss": 0.0388,
+      "step": 2472
+    },
+    {
+      "epoch": 9.774703557312254,
+      "grad_norm": 0.7735835909843445,
+      "learning_rate": 1.1309523809523809e-06,
+      "loss": 0.0271,
+      "step": 2473
+    },
+    {
+      "epoch": 9.778656126482213,
+      "grad_norm": 1.1842035055160522,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 0.0413,
+      "step": 2474
+    },
+    {
+      "epoch": 9.782608695652174,
+      "grad_norm": 0.8943471312522888,
+      "learning_rate": 1.0912698412698412e-06,
+      "loss": 0.0396,
+      "step": 2475
+    },
+    {
+      "epoch": 9.786561264822135,
+      "grad_norm": 0.7007603049278259,
+      "learning_rate": 1.0714285714285716e-06,
+      "loss": 0.0315,
+      "step": 2476
+    },
+    {
+      "epoch": 9.790513833992096,
+      "grad_norm": 0.545301616191864,
+      "learning_rate": 1.0515873015873017e-06,
+      "loss": 0.0243,
+      "step": 2477
+    },
+    {
+      "epoch": 9.794466403162055,
+      "grad_norm": 0.3348066210746765,
+      "learning_rate": 1.0317460317460317e-06,
+      "loss": 0.0259,
+      "step": 2478
+    },
+    {
+      "epoch": 9.798418972332016,
+      "grad_norm": 0.856115996837616,
+      "learning_rate": 1.011904761904762e-06,
+      "loss": 0.0496,
+      "step": 2479
+    },
+    {
+      "epoch": 9.802371541501977,
+      "grad_norm": 1.026168704032898,
+      "learning_rate": 9.92063492063492e-07,
+      "loss": 0.0343,
+      "step": 2480
+    },
+    {
+      "epoch": 9.806324110671937,
+      "grad_norm": 0.5965343117713928,
+      "learning_rate": 9.722222222222222e-07,
+      "loss": 0.0248,
+      "step": 2481
+    },
+    {
+      "epoch": 9.810276679841897,
+      "grad_norm": 0.6260294318199158,
+      "learning_rate": 9.523809523809526e-07,
+      "loss": 0.0229,
+      "step": 2482
+    },
+    {
+      "epoch": 9.814229249011857,
+      "grad_norm": 0.7057980298995972,
+      "learning_rate": 9.325396825396826e-07,
+      "loss": 0.0344,
+      "step": 2483
+    },
+    {
+      "epoch": 9.818181818181818,
+      "grad_norm": 1.1280661821365356,
+      "learning_rate": 9.126984126984128e-07,
+      "loss": 0.0297,
+      "step": 2484
+    },
+    {
+      "epoch": 9.82213438735178,
+      "grad_norm": 0.6508181095123291,
+      "learning_rate": 8.928571428571428e-07,
+      "loss": 0.0312,
+      "step": 2485
+    },
+    {
+      "epoch": 9.826086956521738,
+      "grad_norm": 2.5763347148895264,
+      "learning_rate": 8.730158730158731e-07,
+      "loss": 0.0786,
+      "step": 2486
+    },
+    {
+      "epoch": 9.8300395256917,
+      "grad_norm": 0.6393975615501404,
+      "learning_rate": 8.531746031746032e-07,
+      "loss": 0.0292,
+      "step": 2487
+    },
+    {
+      "epoch": 9.83399209486166,
+      "grad_norm": 0.559859037399292,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.0322,
+      "step": 2488
+    },
+    {
+      "epoch": 9.837944664031621,
+      "grad_norm": 0.8873099088668823,
+      "learning_rate": 8.134920634920636e-07,
+      "loss": 0.0402,
+      "step": 2489
+    },
+    {
+      "epoch": 9.841897233201582,
+      "grad_norm": 1.0293989181518555,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.0487,
+      "step": 2490
+    },
+    {
+      "epoch": 9.845849802371541,
+      "grad_norm": 0.8084055185317993,
+      "learning_rate": 7.738095238095238e-07,
+      "loss": 0.0415,
+      "step": 2491
+    },
+    {
+      "epoch": 9.849802371541502,
+      "grad_norm": 0.8011215329170227,
+      "learning_rate": 7.53968253968254e-07,
+      "loss": 0.0332,
+      "step": 2492
+    },
+    {
+      "epoch": 9.853754940711463,
+      "grad_norm": 0.6973430514335632,
+      "learning_rate": 7.341269841269842e-07,
+      "loss": 0.0338,
+      "step": 2493
+    },
+    {
+      "epoch": 9.857707509881424,
+      "grad_norm": 1.018647313117981,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.0305,
+      "step": 2494
+    },
+    {
+      "epoch": 9.861660079051383,
+      "grad_norm": 0.7031568884849548,
+      "learning_rate": 6.944444444444445e-07,
+      "loss": 0.0297,
+      "step": 2495
+    },
+    {
+      "epoch": 9.865612648221344,
+      "grad_norm": 0.6655853986740112,
+      "learning_rate": 6.746031746031746e-07,
+      "loss": 0.0298,
+      "step": 2496
+    },
+    {
+      "epoch": 9.869565217391305,
+      "grad_norm": 1.6888632774353027,
+      "learning_rate": 6.547619047619048e-07,
+      "loss": 0.0578,
+      "step": 2497
+    },
+    {
+      "epoch": 9.873517786561266,
+      "grad_norm": 1.1527448892593384,
+      "learning_rate": 6.34920634920635e-07,
+      "loss": 0.0444,
+      "step": 2498
+    },
+    {
+      "epoch": 9.877470355731225,
+      "grad_norm": 0.6494206190109253,
+      "learning_rate": 6.150793650793651e-07,
+      "loss": 0.034,
+      "step": 2499
+    },
+    {
+      "epoch": 9.881422924901186,
+      "grad_norm": 0.9560872912406921,
+      "learning_rate": 5.952380952380952e-07,
+      "loss": 0.0283,
+      "step": 2500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2530,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.237898477266954e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}