diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16036 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2286,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013123359580052493,
+      "grad_norm": 2.9126638029304415,
+      "learning_rate": 0.0,
+      "loss": 0.7832,
+      "step": 1
+    },
+    {
+      "epoch": 0.0026246719160104987,
+      "grad_norm": 2.9535107814310413,
+      "learning_rate": 2.183406113537118e-07,
+      "loss": 0.782,
+      "step": 2
+    },
+    {
+      "epoch": 0.003937007874015748,
+      "grad_norm": 2.891225516031155,
+      "learning_rate": 4.366812227074236e-07,
+      "loss": 0.7873,
+      "step": 3
+    },
+    {
+      "epoch": 0.005249343832020997,
+      "grad_norm": 2.8828155882085245,
+      "learning_rate": 6.550218340611353e-07,
+      "loss": 0.7822,
+      "step": 4
+    },
+    {
+      "epoch": 0.006561679790026247,
+      "grad_norm": 2.9010802980384534,
+      "learning_rate": 8.733624454148472e-07,
+      "loss": 0.7915,
+      "step": 5
+    },
+    {
+      "epoch": 0.007874015748031496,
+      "grad_norm": 2.7602762133965606,
+      "learning_rate": 1.0917030567685589e-06,
+      "loss": 0.7785,
+      "step": 6
+    },
+    {
+      "epoch": 0.009186351706036745,
+      "grad_norm": 2.715618019973422,
+      "learning_rate": 1.3100436681222706e-06,
+      "loss": 0.7733,
+      "step": 7
+    },
+    {
+      "epoch": 0.010498687664041995,
+      "grad_norm": 2.191730331558692,
+      "learning_rate": 1.5283842794759824e-06,
+      "loss": 0.7496,
+      "step": 8
+    },
+    {
+      "epoch": 0.011811023622047244,
+      "grad_norm": 2.1595811141480215,
+      "learning_rate": 1.7467248908296944e-06,
+      "loss": 0.7521,
+      "step": 9
+    },
+    {
+      "epoch": 0.013123359580052493,
+      "grad_norm": 1.390575887265111,
+      "learning_rate": 1.9650655021834064e-06,
+      "loss": 0.7016,
+      "step": 10
+    },
+    {
+      "epoch": 0.014435695538057743,
+      "grad_norm": 1.33171849772304,
+      "learning_rate": 2.1834061135371177e-06,
+      "loss": 0.7155,
+      "step": 11
+    },
+    {
+      "epoch": 0.015748031496062992,
+      "grad_norm": 1.2409870217345158,
+      "learning_rate": 2.40174672489083e-06,
+      "loss": 0.6929,
+      "step": 12
+    },
+    {
+      "epoch": 0.01706036745406824,
+      "grad_norm": 1.2084793545043495,
+      "learning_rate": 2.6200873362445413e-06,
+      "loss": 0.7009,
+      "step": 13
+    },
+    {
+      "epoch": 0.01837270341207349,
+      "grad_norm": 1.1680208954849955,
+      "learning_rate": 2.8384279475982535e-06,
+      "loss": 0.6812,
+      "step": 14
+    },
+    {
+      "epoch": 0.01968503937007874,
+      "grad_norm": 1.1355743447203603,
+      "learning_rate": 3.056768558951965e-06,
+      "loss": 0.688,
+      "step": 15
+    },
+    {
+      "epoch": 0.02099737532808399,
+      "grad_norm": 1.0317165420549248,
+      "learning_rate": 3.275109170305677e-06,
+      "loss": 0.6593,
+      "step": 16
+    },
+    {
+      "epoch": 0.02230971128608924,
+      "grad_norm": 0.9387184587707322,
+      "learning_rate": 3.493449781659389e-06,
+      "loss": 0.6541,
+      "step": 17
+    },
+    {
+      "epoch": 0.023622047244094488,
+      "grad_norm": 0.7541005074924774,
+      "learning_rate": 3.711790393013101e-06,
+      "loss": 0.6408,
+      "step": 18
+    },
+    {
+      "epoch": 0.024934383202099737,
+      "grad_norm": 0.7728105133543528,
+      "learning_rate": 3.930131004366813e-06,
+      "loss": 0.6203,
+      "step": 19
+    },
+    {
+      "epoch": 0.026246719160104987,
+      "grad_norm": 0.7950890795970847,
+      "learning_rate": 4.1484716157205246e-06,
+      "loss": 0.6202,
+      "step": 20
+    },
+    {
+      "epoch": 0.027559055118110236,
+      "grad_norm": 0.6863748218310068,
+      "learning_rate": 4.3668122270742355e-06,
+      "loss": 0.6002,
+      "step": 21
+    },
+    {
+      "epoch": 0.028871391076115485,
+      "grad_norm": 0.5739835561108916,
+      "learning_rate": 4.585152838427948e-06,
+      "loss": 0.6062,
+      "step": 22
+    },
+    {
+      "epoch": 0.030183727034120734,
+      "grad_norm": 0.5473276582136811,
+      "learning_rate": 4.80349344978166e-06,
+      "loss": 0.5867,
+      "step": 23
+    },
+    {
+      "epoch": 0.031496062992125984,
+      "grad_norm": 0.5300550512987406,
+      "learning_rate": 5.021834061135372e-06,
+      "loss": 0.5829,
+      "step": 24
+    },
+    {
+      "epoch": 0.03280839895013123,
+      "grad_norm": 0.49955414085231453,
+      "learning_rate": 5.240174672489083e-06,
+      "loss": 0.5794,
+      "step": 25
+    },
+    {
+      "epoch": 0.03412073490813648,
+      "grad_norm": 0.5671716120602697,
+      "learning_rate": 5.458515283842795e-06,
+      "loss": 0.5761,
+      "step": 26
+    },
+    {
+      "epoch": 0.03543307086614173,
+      "grad_norm": 0.5615134663772937,
+      "learning_rate": 5.676855895196507e-06,
+      "loss": 0.5739,
+      "step": 27
+    },
+    {
+      "epoch": 0.03674540682414698,
+      "grad_norm": 0.5134460794996384,
+      "learning_rate": 5.895196506550219e-06,
+      "loss": 0.566,
+      "step": 28
+    },
+    {
+      "epoch": 0.03805774278215223,
+      "grad_norm": 0.47674093745871055,
+      "learning_rate": 6.11353711790393e-06,
+      "loss": 0.562,
+      "step": 29
+    },
+    {
+      "epoch": 0.03937007874015748,
+      "grad_norm": 0.47630236844797613,
+      "learning_rate": 6.3318777292576415e-06,
+      "loss": 0.5586,
+      "step": 30
+    },
+    {
+      "epoch": 0.04068241469816273,
+      "grad_norm": 0.4306875048655715,
+      "learning_rate": 6.550218340611354e-06,
+      "loss": 0.5488,
+      "step": 31
+    },
+    {
+      "epoch": 0.04199475065616798,
+      "grad_norm": 0.3634824566362368,
+      "learning_rate": 6.768558951965067e-06,
+      "loss": 0.5386,
+      "step": 32
+    },
+    {
+      "epoch": 0.04330708661417323,
+      "grad_norm": 0.34218667667665775,
+      "learning_rate": 6.986899563318778e-06,
+      "loss": 0.54,
+      "step": 33
+    },
+    {
+      "epoch": 0.04461942257217848,
+      "grad_norm": 0.29365855132502194,
+      "learning_rate": 7.205240174672489e-06,
+      "loss": 0.5409,
+      "step": 34
+    },
+    {
+      "epoch": 0.045931758530183726,
+      "grad_norm": 0.27485677104927425,
+      "learning_rate": 7.423580786026202e-06,
+      "loss": 0.5315,
+      "step": 35
+    },
+    {
+      "epoch": 0.047244094488188976,
+      "grad_norm": 0.2613916460435711,
+      "learning_rate": 7.641921397379913e-06,
+      "loss": 0.5284,
+      "step": 36
+    },
+    {
+      "epoch": 0.048556430446194225,
+      "grad_norm": 0.2656332908906874,
+      "learning_rate": 7.860262008733626e-06,
+      "loss": 0.5269,
+      "step": 37
+    },
+    {
+      "epoch": 0.049868766404199474,
+      "grad_norm": 0.2697843516829197,
+      "learning_rate": 8.078602620087337e-06,
+      "loss": 0.5167,
+      "step": 38
+    },
+    {
+      "epoch": 0.051181102362204724,
+      "grad_norm": 0.26375539361466077,
+      "learning_rate": 8.296943231441049e-06,
+      "loss": 0.5304,
+      "step": 39
+    },
+    {
+      "epoch": 0.05249343832020997,
+      "grad_norm": 0.23774888735675084,
+      "learning_rate": 8.51528384279476e-06,
+      "loss": 0.5325,
+      "step": 40
+    },
+    {
+      "epoch": 0.05380577427821522,
+      "grad_norm": 0.2145391226341411,
+      "learning_rate": 8.733624454148471e-06,
+      "loss": 0.5221,
+      "step": 41
+    },
+    {
+      "epoch": 0.05511811023622047,
+      "grad_norm": 0.21938685904995558,
+      "learning_rate": 8.951965065502184e-06,
+      "loss": 0.5156,
+      "step": 42
+    },
+    {
+      "epoch": 0.05643044619422572,
+      "grad_norm": 0.2096698333984364,
+      "learning_rate": 9.170305676855896e-06,
+      "loss": 0.5106,
+      "step": 43
+    },
+    {
+      "epoch": 0.05774278215223097,
+      "grad_norm": 0.16824175251999954,
+      "learning_rate": 9.388646288209607e-06,
+      "loss": 0.5119,
+      "step": 44
+    },
+    {
+      "epoch": 0.05905511811023622,
+      "grad_norm": 0.17263429526248827,
+      "learning_rate": 9.60698689956332e-06,
+      "loss": 0.5102,
+      "step": 45
+    },
+    {
+      "epoch": 0.06036745406824147,
+      "grad_norm": 0.17059708077767888,
+      "learning_rate": 9.82532751091703e-06,
+      "loss": 0.5127,
+      "step": 46
+    },
+    {
+      "epoch": 0.06167979002624672,
+      "grad_norm": 0.16307457196564992,
+      "learning_rate": 1.0043668122270743e-05,
+      "loss": 0.505,
+      "step": 47
+    },
+    {
+      "epoch": 0.06299212598425197,
+      "grad_norm": 0.1641709790762573,
+      "learning_rate": 1.0262008733624454e-05,
+      "loss": 0.5001,
+      "step": 48
+    },
+    {
+      "epoch": 0.06430446194225722,
+      "grad_norm": 0.13863292834299282,
+      "learning_rate": 1.0480349344978165e-05,
+      "loss": 0.4933,
+      "step": 49
+    },
+    {
+      "epoch": 0.06561679790026247,
+      "grad_norm": 0.13774484876880888,
+      "learning_rate": 1.0698689956331878e-05,
+      "loss": 0.5053,
+      "step": 50
+    },
+    {
+      "epoch": 0.06692913385826772,
+      "grad_norm": 0.13646253100134526,
+      "learning_rate": 1.091703056768559e-05,
+      "loss": 0.4943,
+      "step": 51
+    },
+    {
+      "epoch": 0.06824146981627296,
+      "grad_norm": 0.13146937474580844,
+      "learning_rate": 1.1135371179039301e-05,
+      "loss": 0.5017,
+      "step": 52
+    },
+    {
+      "epoch": 0.06955380577427822,
+      "grad_norm": 0.13454561638467452,
+      "learning_rate": 1.1353711790393014e-05,
+      "loss": 0.4862,
+      "step": 53
+    },
+    {
+      "epoch": 0.07086614173228346,
+      "grad_norm": 0.1262045634655757,
+      "learning_rate": 1.1572052401746727e-05,
+      "loss": 0.4809,
+      "step": 54
+    },
+    {
+      "epoch": 0.07217847769028872,
+      "grad_norm": 0.13082459415917275,
+      "learning_rate": 1.1790393013100438e-05,
+      "loss": 0.4838,
+      "step": 55
+    },
+    {
+      "epoch": 0.07349081364829396,
+      "grad_norm": 0.1366110192360569,
+      "learning_rate": 1.2008733624454148e-05,
+      "loss": 0.4862,
+      "step": 56
+    },
+    {
+      "epoch": 0.07480314960629922,
+      "grad_norm": 0.11537830792870955,
+      "learning_rate": 1.222707423580786e-05,
+      "loss": 0.4808,
+      "step": 57
+    },
+    {
+      "epoch": 0.07611548556430446,
+      "grad_norm": 0.13240358914758546,
+      "learning_rate": 1.2445414847161572e-05,
+      "loss": 0.4855,
+      "step": 58
+    },
+    {
+      "epoch": 0.07742782152230972,
+      "grad_norm": 0.118908529253653,
+      "learning_rate": 1.2663755458515283e-05,
+      "loss": 0.4818,
+      "step": 59
+    },
+    {
+      "epoch": 0.07874015748031496,
+      "grad_norm": 0.11935352020616406,
+      "learning_rate": 1.2882096069868996e-05,
+      "loss": 0.4925,
+      "step": 60
+    },
+    {
+      "epoch": 0.08005249343832022,
+      "grad_norm": 0.13312193504193806,
+      "learning_rate": 1.3100436681222708e-05,
+      "loss": 0.4824,
+      "step": 61
+    },
+    {
+      "epoch": 0.08136482939632546,
+      "grad_norm": 0.11484004783498163,
+      "learning_rate": 1.331877729257642e-05,
+      "loss": 0.4794,
+      "step": 62
+    },
+    {
+      "epoch": 0.08267716535433071,
+      "grad_norm": 0.11235536943619709,
+      "learning_rate": 1.3537117903930133e-05,
+      "loss": 0.4752,
+      "step": 63
+    },
+    {
+      "epoch": 0.08398950131233596,
+      "grad_norm": 0.11950214866650528,
+      "learning_rate": 1.3755458515283843e-05,
+      "loss": 0.4783,
+      "step": 64
+    },
+    {
+      "epoch": 0.08530183727034121,
+      "grad_norm": 0.10153907388237406,
+      "learning_rate": 1.3973799126637555e-05,
+      "loss": 0.4677,
+      "step": 65
+    },
+    {
+      "epoch": 0.08661417322834646,
+      "grad_norm": 0.1113306589375119,
+      "learning_rate": 1.4192139737991266e-05,
+      "loss": 0.4765,
+      "step": 66
+    },
+    {
+      "epoch": 0.08792650918635171,
+      "grad_norm": 0.10465439810210157,
+      "learning_rate": 1.4410480349344979e-05,
+      "loss": 0.4737,
+      "step": 67
+    },
+    {
+      "epoch": 0.08923884514435695,
+      "grad_norm": 0.11642351561425732,
+      "learning_rate": 1.4628820960698691e-05,
+      "loss": 0.4757,
+      "step": 68
+    },
+    {
+      "epoch": 0.09055118110236221,
+      "grad_norm": 0.10724702478562735,
+      "learning_rate": 1.4847161572052404e-05,
+      "loss": 0.4748,
+      "step": 69
+    },
+    {
+      "epoch": 0.09186351706036745,
+      "grad_norm": 0.10856693170594509,
+      "learning_rate": 1.5065502183406113e-05,
+      "loss": 0.4627,
+      "step": 70
+    },
+    {
+      "epoch": 0.09317585301837271,
+      "grad_norm": 0.11923368607754395,
+      "learning_rate": 1.5283842794759826e-05,
+      "loss": 0.4684,
+      "step": 71
+    },
+    {
+      "epoch": 0.09448818897637795,
+      "grad_norm": 0.14581179221689672,
+      "learning_rate": 1.5502183406113537e-05,
+      "loss": 0.4671,
+      "step": 72
+    },
+    {
+      "epoch": 0.09580052493438321,
+      "grad_norm": 0.16515971207995384,
+      "learning_rate": 1.572052401746725e-05,
+      "loss": 0.4721,
+      "step": 73
+    },
+    {
+      "epoch": 0.09711286089238845,
+      "grad_norm": 0.22604749161373452,
+      "learning_rate": 1.5938864628820962e-05,
+      "loss": 0.4732,
+      "step": 74
+    },
+    {
+      "epoch": 0.0984251968503937,
+      "grad_norm": 0.22992702267989845,
+      "learning_rate": 1.6157205240174673e-05,
+      "loss": 0.4566,
+      "step": 75
+    },
+    {
+      "epoch": 0.09973753280839895,
+      "grad_norm": 0.17380008449474543,
+      "learning_rate": 1.6375545851528384e-05,
+      "loss": 0.4591,
+      "step": 76
+    },
+    {
+      "epoch": 0.1010498687664042,
+      "grad_norm": 0.13475250853185414,
+      "learning_rate": 1.6593886462882098e-05,
+      "loss": 0.4627,
+      "step": 77
+    },
+    {
+      "epoch": 0.10236220472440945,
+      "grad_norm": 0.14708913695551645,
+      "learning_rate": 1.681222707423581e-05,
+      "loss": 0.4647,
+      "step": 78
+    },
+    {
+      "epoch": 0.1036745406824147,
+      "grad_norm": 0.15159946837457072,
+      "learning_rate": 1.703056768558952e-05,
+      "loss": 0.4674,
+      "step": 79
+    },
+    {
+      "epoch": 0.10498687664041995,
+      "grad_norm": 0.14381291486066278,
+      "learning_rate": 1.7248908296943234e-05,
+      "loss": 0.4495,
+      "step": 80
+    },
+    {
+      "epoch": 0.1062992125984252,
+      "grad_norm": 0.1968221290209024,
+      "learning_rate": 1.7467248908296942e-05,
+      "loss": 0.4655,
+      "step": 81
+    },
+    {
+      "epoch": 0.10761154855643044,
+      "grad_norm": 0.17899788688278773,
+      "learning_rate": 1.7685589519650656e-05,
+      "loss": 0.4636,
+      "step": 82
+    },
+    {
+      "epoch": 0.1089238845144357,
+      "grad_norm": 0.15918650276014326,
+      "learning_rate": 1.7903930131004367e-05,
+      "loss": 0.4692,
+      "step": 83
+    },
+    {
+      "epoch": 0.11023622047244094,
+      "grad_norm": 0.15065790353307687,
+      "learning_rate": 1.812227074235808e-05,
+      "loss": 0.4747,
+      "step": 84
+    },
+    {
+      "epoch": 0.1115485564304462,
+      "grad_norm": 0.12641296870850482,
+      "learning_rate": 1.8340611353711792e-05,
+      "loss": 0.4586,
+      "step": 85
+    },
+    {
+      "epoch": 0.11286089238845144,
+      "grad_norm": 0.1395171761113485,
+      "learning_rate": 1.8558951965065503e-05,
+      "loss": 0.4567,
+      "step": 86
+    },
+    {
+      "epoch": 0.1141732283464567,
+      "grad_norm": 0.18422006487300874,
+      "learning_rate": 1.8777292576419214e-05,
+      "loss": 0.4523,
+      "step": 87
+    },
+    {
+      "epoch": 0.11548556430446194,
+      "grad_norm": 0.20458604260133284,
+      "learning_rate": 1.8995633187772925e-05,
+      "loss": 0.466,
+      "step": 88
+    },
+    {
+      "epoch": 0.1167979002624672,
+      "grad_norm": 0.21416566482810911,
+      "learning_rate": 1.921397379912664e-05,
+      "loss": 0.4597,
+      "step": 89
+    },
+    {
+      "epoch": 0.11811023622047244,
+      "grad_norm": 0.3404980245161944,
+      "learning_rate": 1.943231441048035e-05,
+      "loss": 0.4516,
+      "step": 90
+    },
+    {
+      "epoch": 0.1194225721784777,
+      "grad_norm": 0.47883401640310325,
+      "learning_rate": 1.965065502183406e-05,
+      "loss": 0.4585,
+      "step": 91
+    },
+    {
+      "epoch": 0.12073490813648294,
+      "grad_norm": 0.419629206304247,
+      "learning_rate": 1.9868995633187772e-05,
+      "loss": 0.4634,
+      "step": 92
+    },
+    {
+      "epoch": 0.1220472440944882,
+      "grad_norm": 0.267151389749082,
+      "learning_rate": 2.0087336244541487e-05,
+      "loss": 0.4589,
+      "step": 93
+    },
+    {
+      "epoch": 0.12335958005249344,
+      "grad_norm": 0.2792738180746428,
+      "learning_rate": 2.0305676855895198e-05,
+      "loss": 0.4432,
+      "step": 94
+    },
+    {
+      "epoch": 0.12467191601049869,
+      "grad_norm": 0.32860034753423023,
+      "learning_rate": 2.052401746724891e-05,
+      "loss": 0.4533,
+      "step": 95
+    },
+    {
+      "epoch": 0.12598425196850394,
+      "grad_norm": 0.28138240243534185,
+      "learning_rate": 2.0742358078602623e-05,
+      "loss": 0.4415,
+      "step": 96
+    },
+    {
+      "epoch": 0.1272965879265092,
+      "grad_norm": 0.23267116334172264,
+      "learning_rate": 2.096069868995633e-05,
+      "loss": 0.4495,
+      "step": 97
+    },
+    {
+      "epoch": 0.12860892388451445,
+      "grad_norm": 0.26399179767212094,
+      "learning_rate": 2.1179039301310045e-05,
+      "loss": 0.4605,
+      "step": 98
+    },
+    {
+      "epoch": 0.12992125984251968,
+      "grad_norm": 0.26014780504864593,
+      "learning_rate": 2.1397379912663756e-05,
+      "loss": 0.4546,
+      "step": 99
+    },
+    {
+      "epoch": 0.13123359580052493,
+      "grad_norm": 0.17534351139068388,
+      "learning_rate": 2.161572052401747e-05,
+      "loss": 0.461,
+      "step": 100
+    },
+    {
+      "epoch": 0.1325459317585302,
+      "grad_norm": 0.2476528990243175,
+      "learning_rate": 2.183406113537118e-05,
+      "loss": 0.4411,
+      "step": 101
+    },
+    {
+      "epoch": 0.13385826771653545,
+      "grad_norm": 0.22724440840701302,
+      "learning_rate": 2.2052401746724892e-05,
+      "loss": 0.4379,
+      "step": 102
+    },
+    {
+      "epoch": 0.13517060367454067,
+      "grad_norm": 0.1806446716929541,
+      "learning_rate": 2.2270742358078603e-05,
+      "loss": 0.4516,
+      "step": 103
+    },
+    {
+      "epoch": 0.13648293963254593,
+      "grad_norm": 0.19257462997189592,
+      "learning_rate": 2.2489082969432314e-05,
+      "loss": 0.4403,
+      "step": 104
+    },
+    {
+      "epoch": 0.1377952755905512,
+      "grad_norm": 0.2174925123446844,
+      "learning_rate": 2.2707423580786028e-05,
+      "loss": 0.4504,
+      "step": 105
+    },
+    {
+      "epoch": 0.13910761154855644,
+      "grad_norm": 0.22509543794425005,
+      "learning_rate": 2.292576419213974e-05,
+      "loss": 0.4458,
+      "step": 106
+    },
+    {
+      "epoch": 0.14041994750656167,
+      "grad_norm": 0.1992229229109345,
+      "learning_rate": 2.3144104803493453e-05,
+      "loss": 0.4553,
+      "step": 107
+    },
+    {
+      "epoch": 0.14173228346456693,
+      "grad_norm": 0.20615256431073567,
+      "learning_rate": 2.336244541484716e-05,
+      "loss": 0.4448,
+      "step": 108
+    },
+    {
+      "epoch": 0.14304461942257218,
+      "grad_norm": 0.24694817090896407,
+      "learning_rate": 2.3580786026200875e-05,
+      "loss": 0.4414,
+      "step": 109
+    },
+    {
+      "epoch": 0.14435695538057744,
+      "grad_norm": 0.23900524227389022,
+      "learning_rate": 2.3799126637554586e-05,
+      "loss": 0.4542,
+      "step": 110
+    },
+    {
+      "epoch": 0.14566929133858267,
+      "grad_norm": 0.2608990561024207,
+      "learning_rate": 2.4017467248908297e-05,
+      "loss": 0.4447,
+      "step": 111
+    },
+    {
+      "epoch": 0.14698162729658792,
+      "grad_norm": 0.3519009362219106,
+      "learning_rate": 2.423580786026201e-05,
+      "loss": 0.4424,
+      "step": 112
+    },
+    {
+      "epoch": 0.14829396325459318,
+      "grad_norm": 0.5354693143812376,
+      "learning_rate": 2.445414847161572e-05,
+      "loss": 0.4461,
+      "step": 113
+    },
+    {
+      "epoch": 0.14960629921259844,
+      "grad_norm": 0.7528660374301389,
+      "learning_rate": 2.4672489082969433e-05,
+      "loss": 0.4533,
+      "step": 114
+    },
+    {
+      "epoch": 0.15091863517060367,
+      "grad_norm": 0.7218382935213049,
+      "learning_rate": 2.4890829694323144e-05,
+      "loss": 0.4482,
+      "step": 115
+    },
+    {
+      "epoch": 0.15223097112860892,
+      "grad_norm": 0.40421740086763935,
+      "learning_rate": 2.5109170305676855e-05,
+      "loss": 0.4477,
+      "step": 116
+    },
+    {
+      "epoch": 0.15354330708661418,
+      "grad_norm": 0.4384378397023311,
+      "learning_rate": 2.5327510917030566e-05,
+      "loss": 0.4476,
+      "step": 117
+    },
+    {
+      "epoch": 0.15485564304461943,
+      "grad_norm": 0.4584510425545346,
+      "learning_rate": 2.554585152838428e-05,
+      "loss": 0.454,
+      "step": 118
+    },
+    {
+      "epoch": 0.15616797900262466,
+      "grad_norm": 0.3535613181149153,
+      "learning_rate": 2.576419213973799e-05,
+      "loss": 0.4379,
+      "step": 119
+    },
+    {
+      "epoch": 0.15748031496062992,
+      "grad_norm": 0.40893327153676035,
+      "learning_rate": 2.5982532751091705e-05,
+      "loss": 0.4436,
+      "step": 120
+    },
+    {
+      "epoch": 0.15879265091863518,
+      "grad_norm": 0.26712551625851066,
+      "learning_rate": 2.6200873362445416e-05,
+      "loss": 0.439,
+      "step": 121
+    },
+    {
+      "epoch": 0.16010498687664043,
+      "grad_norm": 0.34778283107317803,
+      "learning_rate": 2.6419213973799127e-05,
+      "loss": 0.4416,
+      "step": 122
+    },
+    {
+      "epoch": 0.16141732283464566,
+      "grad_norm": 0.39361390949810415,
+      "learning_rate": 2.663755458515284e-05,
+      "loss": 0.4406,
+      "step": 123
+    },
+    {
+      "epoch": 0.16272965879265092,
+      "grad_norm": 0.20152417022980623,
+      "learning_rate": 2.6855895196506553e-05,
+      "loss": 0.4301,
+      "step": 124
+    },
+    {
+      "epoch": 0.16404199475065617,
+      "grad_norm": 0.3205126563367627,
+      "learning_rate": 2.7074235807860267e-05,
+      "loss": 0.4343,
+      "step": 125
+    },
+    {
+      "epoch": 0.16535433070866143,
+      "grad_norm": 0.2909053587232463,
+      "learning_rate": 2.7292576419213978e-05,
+      "loss": 0.4418,
+      "step": 126
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.3863749502887551,
+      "learning_rate": 2.7510917030567685e-05,
+      "loss": 0.4447,
+      "step": 127
+    },
+    {
+      "epoch": 0.1679790026246719,
+      "grad_norm": 0.3923157330285858,
+      "learning_rate": 2.7729257641921396e-05,
+      "loss": 0.4337,
+      "step": 128
+    },
+    {
+      "epoch": 0.16929133858267717,
+      "grad_norm": 0.3385816658118167,
+      "learning_rate": 2.794759825327511e-05,
+      "loss": 0.4285,
+      "step": 129
+    },
+    {
+      "epoch": 0.17060367454068243,
+      "grad_norm": 0.35913717075847734,
+      "learning_rate": 2.816593886462882e-05,
+      "loss": 0.4473,
+      "step": 130
+    },
+    {
+      "epoch": 0.17191601049868765,
+      "grad_norm": 0.29966256400659497,
+      "learning_rate": 2.8384279475982532e-05,
+      "loss": 0.4437,
+      "step": 131
+    },
+    {
+      "epoch": 0.1732283464566929,
+      "grad_norm": 0.28606781037381884,
+      "learning_rate": 2.8602620087336247e-05,
+      "loss": 0.441,
+      "step": 132
+    },
+    {
+      "epoch": 0.17454068241469817,
+      "grad_norm": 0.32184323094558953,
+      "learning_rate": 2.8820960698689958e-05,
+      "loss": 0.4421,
+      "step": 133
+    },
+    {
+      "epoch": 0.17585301837270342,
+      "grad_norm": 0.2533912390964874,
+      "learning_rate": 2.9039301310043672e-05,
+      "loss": 0.4343,
+      "step": 134
+    },
+    {
+      "epoch": 0.17716535433070865,
+      "grad_norm": 0.2685186276210848,
+      "learning_rate": 2.9257641921397383e-05,
+      "loss": 0.4339,
+      "step": 135
+    },
+    {
+      "epoch": 0.1784776902887139,
+      "grad_norm": 0.22471833697417554,
+      "learning_rate": 2.9475982532751094e-05,
+      "loss": 0.4335,
+      "step": 136
+    },
+    {
+      "epoch": 0.17979002624671916,
+      "grad_norm": 0.28974785662997526,
+      "learning_rate": 2.9694323144104808e-05,
+      "loss": 0.4326,
+      "step": 137
+    },
+    {
+      "epoch": 0.18110236220472442,
+      "grad_norm": 0.2877052678755618,
+      "learning_rate": 2.9912663755458516e-05,
+      "loss": 0.4451,
+      "step": 138
+    },
+    {
+      "epoch": 0.18241469816272965,
+      "grad_norm": 0.34979218257630096,
+      "learning_rate": 3.0131004366812227e-05,
+      "loss": 0.4393,
+      "step": 139
+    },
+    {
+      "epoch": 0.1837270341207349,
+      "grad_norm": 0.4170783149855214,
+      "learning_rate": 3.0349344978165938e-05,
+      "loss": 0.4384,
+      "step": 140
+    },
+    {
+      "epoch": 0.18503937007874016,
+      "grad_norm": 0.41466178167696083,
+      "learning_rate": 3.056768558951965e-05,
+      "loss": 0.4441,
+      "step": 141
+    },
+    {
+      "epoch": 0.18635170603674542,
+      "grad_norm": 0.39635901760898173,
+      "learning_rate": 3.0786026200873366e-05,
+      "loss": 0.4417,
+      "step": 142
+    },
+    {
+      "epoch": 0.18766404199475065,
+      "grad_norm": 0.36401877790616766,
+      "learning_rate": 3.1004366812227074e-05,
+      "loss": 0.4329,
+      "step": 143
+    },
+    {
+      "epoch": 0.1889763779527559,
+      "grad_norm": 0.3457905246423825,
+      "learning_rate": 3.122270742358079e-05,
+      "loss": 0.4491,
+      "step": 144
+    },
+    {
+      "epoch": 0.19028871391076116,
+      "grad_norm": 0.3385565182452068,
+      "learning_rate": 3.14410480349345e-05,
+      "loss": 0.4362,
+      "step": 145
+    },
+    {
+      "epoch": 0.19160104986876642,
+      "grad_norm": 0.3175877390027479,
+      "learning_rate": 3.165938864628821e-05,
+      "loss": 0.4167,
+      "step": 146
+    },
+    {
+      "epoch": 0.19291338582677164,
+      "grad_norm": 0.41240863653213106,
+      "learning_rate": 3.1877729257641924e-05,
+      "loss": 0.4362,
+      "step": 147
+    },
+    {
+      "epoch": 0.1942257217847769,
+      "grad_norm": 0.5112019612538,
+      "learning_rate": 3.209606986899564e-05,
+      "loss": 0.4342,
+      "step": 148
+    },
+    {
+      "epoch": 0.19553805774278216,
+      "grad_norm": 0.45477790069950064,
+      "learning_rate": 3.2314410480349346e-05,
+      "loss": 0.4346,
+      "step": 149
+    },
+    {
+      "epoch": 0.1968503937007874,
+      "grad_norm": 0.4077263883425018,
+      "learning_rate": 3.2532751091703054e-05,
+      "loss": 0.4417,
+      "step": 150
+    },
+    {
+      "epoch": 0.19816272965879264,
+      "grad_norm": 0.33909528485936735,
+      "learning_rate": 3.275109170305677e-05,
+      "loss": 0.4319,
+      "step": 151
+    },
+    {
+      "epoch": 0.1994750656167979,
+      "grad_norm": 0.4258896632427699,
+      "learning_rate": 3.296943231441048e-05,
+      "loss": 0.4465,
+      "step": 152
+    },
+    {
+      "epoch": 0.20078740157480315,
+      "grad_norm": 0.4213150816343299,
+      "learning_rate": 3.3187772925764197e-05,
+      "loss": 0.4286,
+      "step": 153
+    },
+    {
+      "epoch": 0.2020997375328084,
+      "grad_norm": 0.36852978308823964,
+      "learning_rate": 3.3406113537117904e-05,
+      "loss": 0.4328,
+      "step": 154
+    },
+    {
+      "epoch": 0.20341207349081364,
+      "grad_norm": 1.2832130801330608,
+      "learning_rate": 3.362445414847162e-05,
+      "loss": 0.435,
+      "step": 155
+    },
+    {
+      "epoch": 0.2047244094488189,
+      "grad_norm": 0.24075002771480297,
+      "learning_rate": 3.384279475982533e-05,
+      "loss": 0.436,
+      "step": 156
+    },
+    {
+      "epoch": 0.20603674540682415,
+      "grad_norm": 0.33291751771111,
+      "learning_rate": 3.406113537117904e-05,
+      "loss": 0.4326,
+      "step": 157
+    },
+    {
+      "epoch": 0.2073490813648294,
+      "grad_norm": 0.41709080856171965,
+      "learning_rate": 3.4279475982532755e-05,
+      "loss": 0.441,
+      "step": 158
+    },
+    {
+      "epoch": 0.20866141732283464,
+      "grad_norm": 0.41849809829936896,
+      "learning_rate": 3.449781659388647e-05,
+      "loss": 0.4288,
+      "step": 159
+    },
+    {
+      "epoch": 0.2099737532808399,
+      "grad_norm": 0.44592039334062705,
+      "learning_rate": 3.4716157205240176e-05,
+      "loss": 0.4348,
+      "step": 160
+    },
+    {
+      "epoch": 0.21128608923884515,
+      "grad_norm": 0.5414015502039145,
+      "learning_rate": 3.4934497816593884e-05,
+      "loss": 0.4367,
+      "step": 161
+    },
+    {
+      "epoch": 0.2125984251968504,
+      "grad_norm": 0.515695106560996,
+      "learning_rate": 3.51528384279476e-05,
+      "loss": 0.431,
+      "step": 162
+    },
+    {
+      "epoch": 0.21391076115485563,
+      "grad_norm": 0.43623300240664803,
+      "learning_rate": 3.537117903930131e-05,
+      "loss": 0.4394,
+      "step": 163
+    },
+    {
+      "epoch": 0.2152230971128609,
+      "grad_norm": 0.5588533427701409,
+      "learning_rate": 3.558951965065502e-05,
+      "loss": 0.4369,
+      "step": 164
+    },
+    {
+      "epoch": 0.21653543307086615,
+      "grad_norm": 0.5628529179948005,
+      "learning_rate": 3.5807860262008734e-05,
+      "loss": 0.4374,
+      "step": 165
+    },
+    {
+      "epoch": 0.2178477690288714,
+      "grad_norm": 0.3714552107731112,
+      "learning_rate": 3.602620087336245e-05,
+      "loss": 0.4318,
+      "step": 166
+    },
+    {
+      "epoch": 0.21916010498687663,
+      "grad_norm": 0.4446519737939915,
+      "learning_rate": 3.624454148471616e-05,
+      "loss": 0.4441,
+      "step": 167
+    },
+    {
+      "epoch": 0.2204724409448819,
+      "grad_norm": 0.40029572581957534,
+      "learning_rate": 3.646288209606987e-05,
+      "loss": 0.421,
+      "step": 168
+    },
+    {
+      "epoch": 0.22178477690288714,
+      "grad_norm": 0.40583712831734026,
+      "learning_rate": 3.6681222707423585e-05,
+      "loss": 0.4453,
+      "step": 169
+    },
+    {
+      "epoch": 0.2230971128608924,
+      "grad_norm": 0.39131529667763315,
+      "learning_rate": 3.68995633187773e-05,
+      "loss": 0.4362,
+      "step": 170
+    },
+    {
+      "epoch": 0.22440944881889763,
+      "grad_norm": 0.28434889284502946,
+      "learning_rate": 3.711790393013101e-05,
+      "loss": 0.4366,
+      "step": 171
+    },
+    {
+      "epoch": 0.22572178477690288,
+      "grad_norm": 0.3972538994490993,
+      "learning_rate": 3.7336244541484714e-05,
+      "loss": 0.4278,
+      "step": 172
+    },
+    {
+      "epoch": 0.22703412073490814,
+      "grad_norm": 0.34948914712195395,
+      "learning_rate": 3.755458515283843e-05,
+      "loss": 0.4315,
+      "step": 173
+    },
+    {
+      "epoch": 0.2283464566929134,
+      "grad_norm": 0.29667572940298825,
+      "learning_rate": 3.777292576419214e-05,
+      "loss": 0.4268,
+      "step": 174
+    },
+    {
+      "epoch": 0.22965879265091863,
+      "grad_norm": 0.3018305594473432,
+      "learning_rate": 3.799126637554585e-05,
+      "loss": 0.4271,
+      "step": 175
+    },
+    {
+      "epoch": 0.23097112860892388,
+      "grad_norm": 0.31762171333445116,
+      "learning_rate": 3.8209606986899565e-05,
+      "loss": 0.4201,
+      "step": 176
+    },
+    {
+      "epoch": 0.23228346456692914,
+      "grad_norm": 0.3709639700296948,
+      "learning_rate": 3.842794759825328e-05,
+      "loss": 0.4282,
+      "step": 177
+    },
+    {
+      "epoch": 0.2335958005249344,
+      "grad_norm": 0.2927680356653586,
+      "learning_rate": 3.864628820960699e-05,
+      "loss": 0.4351,
+      "step": 178
+    },
+    {
+      "epoch": 0.23490813648293962,
+      "grad_norm": 0.27169843649693876,
+      "learning_rate": 3.88646288209607e-05,
+      "loss": 0.4323,
+      "step": 179
+    },
+    {
+      "epoch": 0.23622047244094488,
+      "grad_norm": 0.3859443921331653,
+      "learning_rate": 3.9082969432314415e-05,
+      "loss": 0.4323,
+      "step": 180
+    },
+    {
+      "epoch": 0.23753280839895013,
+      "grad_norm": 0.299486361291648,
+      "learning_rate": 3.930131004366812e-05,
+      "loss": 0.4295,
+      "step": 181
+    },
+    {
+      "epoch": 0.2388451443569554,
+      "grad_norm": 0.29971315956283956,
+      "learning_rate": 3.951965065502183e-05,
+      "loss": 0.4224,
+      "step": 182
+    },
+    {
+      "epoch": 0.24015748031496062,
+      "grad_norm": 0.3644764317014543,
+      "learning_rate": 3.9737991266375545e-05,
+      "loss": 0.4321,
+      "step": 183
+    },
+    {
+      "epoch": 0.24146981627296588,
+      "grad_norm": 0.3973383506336532,
+      "learning_rate": 3.995633187772926e-05,
+      "loss": 0.431,
+      "step": 184
+    },
+    {
+      "epoch": 0.24278215223097113,
+      "grad_norm": 0.5223246852029854,
+      "learning_rate": 4.017467248908297e-05,
+      "loss": 0.4357,
+      "step": 185
+    },
+    {
+      "epoch": 0.2440944881889764,
+      "grad_norm": 0.5891610574112952,
+      "learning_rate": 4.039301310043668e-05,
+      "loss": 0.4411,
+      "step": 186
+    },
+    {
+      "epoch": 0.24540682414698162,
+      "grad_norm": 0.645227747524698,
+      "learning_rate": 4.0611353711790395e-05,
+      "loss": 0.4261,
+      "step": 187
+    },
+    {
+      "epoch": 0.24671916010498687,
+      "grad_norm": 0.6203939897248906,
+      "learning_rate": 4.082969432314411e-05,
+      "loss": 0.4314,
+      "step": 188
+    },
+    {
+      "epoch": 0.24803149606299213,
+      "grad_norm": 0.5276788672173057,
+      "learning_rate": 4.104803493449782e-05,
+      "loss": 0.4241,
+      "step": 189
+    },
+    {
+      "epoch": 0.24934383202099739,
+      "grad_norm": 0.4304640998760914,
+      "learning_rate": 4.126637554585153e-05,
+      "loss": 0.4235,
+      "step": 190
+    },
+    {
+      "epoch": 0.25065616797900264,
+      "grad_norm": 0.2905343038703453,
+      "learning_rate": 4.1484716157205246e-05,
+      "loss": 0.421,
+      "step": 191
+    },
+    {
+      "epoch": 0.25196850393700787,
+      "grad_norm": 0.30208551349085533,
+      "learning_rate": 4.170305676855895e-05,
+      "loss": 0.4273,
+      "step": 192
+    },
+    {
+      "epoch": 0.2532808398950131,
+      "grad_norm": 0.43232518517168844,
+      "learning_rate": 4.192139737991266e-05,
+      "loss": 0.4174,
+      "step": 193
+    },
+    {
+      "epoch": 0.2545931758530184,
+      "grad_norm": 0.32769352629792814,
+      "learning_rate": 4.2139737991266375e-05,
+      "loss": 0.4297,
+      "step": 194
+    },
+    {
+      "epoch": 0.2559055118110236,
+      "grad_norm": 0.23088163215948715,
+      "learning_rate": 4.235807860262009e-05,
+      "loss": 0.4233,
+      "step": 195
+    },
+    {
+      "epoch": 0.2572178477690289,
+      "grad_norm": 0.41789475859042535,
+      "learning_rate": 4.2576419213973804e-05,
+      "loss": 0.4328,
+      "step": 196
+    },
+    {
+      "epoch": 0.2585301837270341,
+      "grad_norm": 0.4316452382437703,
+      "learning_rate": 4.279475982532751e-05,
+      "loss": 0.4332,
+      "step": 197
+    },
+    {
+      "epoch": 0.25984251968503935,
+      "grad_norm": 0.32960223678621026,
+      "learning_rate": 4.3013100436681226e-05,
+      "loss": 0.4301,
+      "step": 198
+    },
+    {
+      "epoch": 0.26115485564304464,
+      "grad_norm": 0.38316761120390197,
+      "learning_rate": 4.323144104803494e-05,
+      "loss": 0.4192,
+      "step": 199
+    },
+    {
+      "epoch": 0.26246719160104987,
+      "grad_norm": 0.5287472732087778,
+      "learning_rate": 4.344978165938865e-05,
+      "loss": 0.4267,
+      "step": 200
+    },
+    {
+      "epoch": 0.2637795275590551,
+      "grad_norm": 0.6241761636162948,
+      "learning_rate": 4.366812227074236e-05,
+      "loss": 0.433,
+      "step": 201
+    },
+    {
+      "epoch": 0.2650918635170604,
+      "grad_norm": 0.7259293698745072,
+      "learning_rate": 4.3886462882096076e-05,
+      "loss": 0.4307,
+      "step": 202
+    },
+    {
+      "epoch": 0.2664041994750656,
+      "grad_norm": 0.6132962224392898,
+      "learning_rate": 4.4104803493449784e-05,
+      "loss": 0.4247,
+      "step": 203
+    },
+    {
+      "epoch": 0.2677165354330709,
+      "grad_norm": 0.3164007843680363,
+      "learning_rate": 4.432314410480349e-05,
+      "loss": 0.4281,
+      "step": 204
+    },
+    {
+      "epoch": 0.2690288713910761,
+      "grad_norm": 0.6912600514873588,
+      "learning_rate": 4.4541484716157205e-05,
+      "loss": 0.4267,
+      "step": 205
+    },
+    {
+      "epoch": 0.27034120734908135,
+      "grad_norm": 0.5867601329886866,
+      "learning_rate": 4.475982532751092e-05,
+      "loss": 0.4222,
+      "step": 206
+    },
+    {
+      "epoch": 0.27165354330708663,
+      "grad_norm": 0.46285862341048506,
+      "learning_rate": 4.497816593886463e-05,
+      "loss": 0.4284,
+      "step": 207
+    },
+    {
+      "epoch": 0.27296587926509186,
+      "grad_norm": 0.5087313985093849,
+      "learning_rate": 4.519650655021834e-05,
+      "loss": 0.4352,
+      "step": 208
+    },
+    {
+      "epoch": 0.2742782152230971,
+      "grad_norm": 0.5107365252700129,
+      "learning_rate": 4.5414847161572056e-05,
+      "loss": 0.4288,
+      "step": 209
+    },
+    {
+      "epoch": 0.2755905511811024,
+      "grad_norm": 0.4346262598712386,
+      "learning_rate": 4.563318777292577e-05,
+      "loss": 0.4354,
+      "step": 210
+    },
+    {
+      "epoch": 0.2769028871391076,
+      "grad_norm": 0.4823026625359119,
+      "learning_rate": 4.585152838427948e-05,
+      "loss": 0.4296,
+      "step": 211
+    },
+    {
+      "epoch": 0.2782152230971129,
+      "grad_norm": 0.5515178986258251,
+      "learning_rate": 4.606986899563319e-05,
+      "loss": 0.4296,
+      "step": 212
+    },
+    {
+      "epoch": 0.2795275590551181,
+      "grad_norm": 0.48670569413107573,
+      "learning_rate": 4.6288209606986906e-05,
+      "loss": 0.4138,
+      "step": 213
+    },
+    {
+      "epoch": 0.28083989501312334,
+      "grad_norm": 0.37074402981462806,
+      "learning_rate": 4.6506550218340614e-05,
+      "loss": 0.4237,
+      "step": 214
+    },
+    {
+      "epoch": 0.2821522309711286,
+      "grad_norm": 0.46622486705150035,
+      "learning_rate": 4.672489082969432e-05,
+      "loss": 0.4277,
+      "step": 215
+    },
+    {
+      "epoch": 0.28346456692913385,
+      "grad_norm": 0.35623970631959745,
+      "learning_rate": 4.6943231441048036e-05,
+      "loss": 0.4321,
+      "step": 216
+    },
+    {
+      "epoch": 0.2847769028871391,
+      "grad_norm": 0.321982251432209,
+      "learning_rate": 4.716157205240175e-05,
+      "loss": 0.4275,
+      "step": 217
+    },
+    {
+      "epoch": 0.28608923884514437,
+      "grad_norm": 0.3196543934296738,
+      "learning_rate": 4.737991266375546e-05,
+      "loss": 0.4225,
+      "step": 218
+    },
+    {
+      "epoch": 0.2874015748031496,
+      "grad_norm": 0.3167551868254657,
+      "learning_rate": 4.759825327510917e-05,
+      "loss": 0.4312,
+      "step": 219
+    },
+    {
+      "epoch": 0.2887139107611549,
+      "grad_norm": 0.3353815275487663,
+      "learning_rate": 4.7816593886462886e-05,
+      "loss": 0.4172,
+      "step": 220
+    },
+    {
+      "epoch": 0.2900262467191601,
+      "grad_norm": 0.312101245109464,
+      "learning_rate": 4.8034934497816594e-05,
+      "loss": 0.4252,
+      "step": 221
+    },
+    {
+      "epoch": 0.29133858267716534,
+      "grad_norm": 0.3284542724911117,
+      "learning_rate": 4.825327510917031e-05,
+      "loss": 0.4245,
+      "step": 222
+    },
+    {
+      "epoch": 0.2926509186351706,
+      "grad_norm": 0.2735627766759545,
+      "learning_rate": 4.847161572052402e-05,
+      "loss": 0.4148,
+      "step": 223
+    },
+    {
+      "epoch": 0.29396325459317585,
+      "grad_norm": 0.2712414752969883,
+      "learning_rate": 4.868995633187774e-05,
+      "loss": 0.4114,
+      "step": 224
+    },
+    {
+      "epoch": 0.2952755905511811,
+      "grad_norm": 0.2522783224586585,
+      "learning_rate": 4.890829694323144e-05,
+      "loss": 0.418,
+      "step": 225
+    },
+    {
+      "epoch": 0.29658792650918636,
+      "grad_norm": 0.27520990311574206,
+      "learning_rate": 4.912663755458515e-05,
+      "loss": 0.4304,
+      "step": 226
+    },
+    {
+      "epoch": 0.2979002624671916,
+      "grad_norm": 0.27783818639172664,
+      "learning_rate": 4.9344978165938866e-05,
+      "loss": 0.4209,
+      "step": 227
+    },
+    {
+      "epoch": 0.2992125984251969,
+      "grad_norm": 0.287654561514498,
+      "learning_rate": 4.956331877729258e-05,
+      "loss": 0.4301,
+      "step": 228
+    },
+    {
+      "epoch": 0.3005249343832021,
+      "grad_norm": 0.3330723422509161,
+      "learning_rate": 4.978165938864629e-05,
+      "loss": 0.4295,
+      "step": 229
+    },
+    {
+      "epoch": 0.30183727034120733,
+      "grad_norm": 0.2920109456724688,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 230
+    },
+    {
+      "epoch": 0.3031496062992126,
+      "grad_norm": 0.3074418477211762,
+      "learning_rate": 4.9999970843117236e-05,
+      "loss": 0.4158,
+      "step": 231
+    },
+    {
+      "epoch": 0.30446194225721784,
+      "grad_norm": 0.39457598687482043,
+      "learning_rate": 4.999988337253696e-05,
+      "loss": 0.4208,
+      "step": 232
+    },
+    {
+      "epoch": 0.30577427821522307,
+      "grad_norm": 0.4161478912136869,
+      "learning_rate": 4.9999737588463194e-05,
+      "loss": 0.4203,
+      "step": 233
+    },
+    {
+      "epoch": 0.30708661417322836,
+      "grad_norm": 0.5091257894738094,
+      "learning_rate": 4.9999533491236e-05,
+      "loss": 0.4336,
+      "step": 234
+    },
+    {
+      "epoch": 0.3083989501312336,
+      "grad_norm": 0.5749088647648364,
+      "learning_rate": 4.999927108133142e-05,
+      "loss": 0.4255,
+      "step": 235
+    },
+    {
+      "epoch": 0.30971128608923887,
+      "grad_norm": 0.4723166447854288,
+      "learning_rate": 4.9998950359361565e-05,
+      "loss": 0.4198,
+      "step": 236
+    },
+    {
+      "epoch": 0.3110236220472441,
+      "grad_norm": 0.350189515325989,
+      "learning_rate": 4.999857132607452e-05,
+      "loss": 0.4351,
+      "step": 237
+    },
+    {
+      "epoch": 0.3123359580052493,
+      "grad_norm": 0.3757192695968391,
+      "learning_rate": 4.999813398235441e-05,
+      "loss": 0.4313,
+      "step": 238
+    },
+    {
+      "epoch": 0.3136482939632546,
+      "grad_norm": 0.4869678669911018,
+      "learning_rate": 4.9997638329221345e-05,
+      "loss": 0.4338,
+      "step": 239
+    },
+    {
+      "epoch": 0.31496062992125984,
+      "grad_norm": 0.5154360464178366,
+      "learning_rate": 4.999708436783148e-05,
+      "loss": 0.423,
+      "step": 240
+    },
+    {
+      "epoch": 0.31627296587926507,
+      "grad_norm": 0.4959587234525514,
+      "learning_rate": 4.9996472099476943e-05,
+      "loss": 0.4144,
+      "step": 241
+    },
+    {
+      "epoch": 0.31758530183727035,
+      "grad_norm": 0.5168768672437032,
+      "learning_rate": 4.99958015255859e-05,
+      "loss": 0.4271,
+      "step": 242
+    },
+    {
+      "epoch": 0.3188976377952756,
+      "grad_norm": 0.4506819016310133,
+      "learning_rate": 4.999507264772247e-05,
+      "loss": 0.4168,
+      "step": 243
+    },
+    {
+      "epoch": 0.32020997375328086,
+      "grad_norm": 0.2932029319832407,
+      "learning_rate": 4.9994285467586814e-05,
+      "loss": 0.4147,
+      "step": 244
+    },
+    {
+      "epoch": 0.3215223097112861,
+      "grad_norm": 0.4150281423309139,
+      "learning_rate": 4.999343998701508e-05,
+      "loss": 0.417,
+      "step": 245
+    },
+    {
+      "epoch": 0.3228346456692913,
+      "grad_norm": 0.4896272965688593,
+      "learning_rate": 4.999253620797937e-05,
+      "loss": 0.4156,
+      "step": 246
+    },
+    {
+      "epoch": 0.3241469816272966,
+      "grad_norm": 0.2967630386826462,
+      "learning_rate": 4.9991574132587815e-05,
+      "loss": 0.4141,
+      "step": 247
+    },
+    {
+      "epoch": 0.32545931758530183,
+      "grad_norm": 0.2233132963441932,
+      "learning_rate": 4.999055376308449e-05,
+      "loss": 0.4245,
+      "step": 248
+    },
+    {
+      "epoch": 0.32677165354330706,
+      "grad_norm": 0.3575722249747852,
+      "learning_rate": 4.998947510184947e-05,
+      "loss": 0.4306,
+      "step": 249
+    },
+    {
+      "epoch": 0.32808398950131235,
+      "grad_norm": 0.27438977367735756,
+      "learning_rate": 4.9988338151398785e-05,
+      "loss": 0.4129,
+      "step": 250
+    },
+    {
+      "epoch": 0.3293963254593176,
+      "grad_norm": 0.21247964296725702,
+      "learning_rate": 4.998714291438442e-05,
+      "loss": 0.4342,
+      "step": 251
+    },
+    {
+      "epoch": 0.33070866141732286,
+      "grad_norm": 0.26348819065101026,
+      "learning_rate": 4.9985889393594345e-05,
+      "loss": 0.4237,
+      "step": 252
+    },
+    {
+      "epoch": 0.3320209973753281,
+      "grad_norm": 0.23629497019681334,
+      "learning_rate": 4.998457759195244e-05,
+      "loss": 0.4147,
+      "step": 253
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.2615681739753958,
+      "learning_rate": 4.998320751251855e-05,
+      "loss": 0.4186,
+      "step": 254
+    },
+    {
+      "epoch": 0.3346456692913386,
+      "grad_norm": 0.270423229034784,
+      "learning_rate": 4.9981779158488476e-05,
+      "loss": 0.3971,
+      "step": 255
+    },
+    {
+      "epoch": 0.3359580052493438,
+      "grad_norm": 0.21882737636975502,
+      "learning_rate": 4.9980292533193896e-05,
+      "loss": 0.4212,
+      "step": 256
+    },
+    {
+      "epoch": 0.33727034120734906,
+      "grad_norm": 0.246752071477264,
+      "learning_rate": 4.9978747640102464e-05,
+      "loss": 0.414,
+      "step": 257
+    },
+    {
+      "epoch": 0.33858267716535434,
+      "grad_norm": 0.2091286490424575,
+      "learning_rate": 4.997714448281771e-05,
+      "loss": 0.4116,
+      "step": 258
+    },
+    {
+      "epoch": 0.33989501312335957,
+      "grad_norm": 0.21741622660278365,
+      "learning_rate": 4.997548306507908e-05,
+      "loss": 0.4099,
+      "step": 259
+    },
+    {
+      "epoch": 0.34120734908136485,
+      "grad_norm": 0.27587733609309706,
+      "learning_rate": 4.997376339076192e-05,
+      "loss": 0.4184,
+      "step": 260
+    },
+    {
+      "epoch": 0.3425196850393701,
+      "grad_norm": 0.20241814119386456,
+      "learning_rate": 4.997198546387745e-05,
+      "loss": 0.4182,
+      "step": 261
+    },
+    {
+      "epoch": 0.3438320209973753,
+      "grad_norm": 0.19867489518455475,
+      "learning_rate": 4.997014928857278e-05,
+      "loss": 0.4273,
+      "step": 262
+    },
+    {
+      "epoch": 0.3451443569553806,
+      "grad_norm": 0.2755494920364151,
+      "learning_rate": 4.996825486913088e-05,
+      "loss": 0.4159,
+      "step": 263
+    },
+    {
+      "epoch": 0.3464566929133858,
+      "grad_norm": 0.221250901318171,
+      "learning_rate": 4.996630220997058e-05,
+      "loss": 0.4171,
+      "step": 264
+    },
+    {
+      "epoch": 0.34776902887139105,
+      "grad_norm": 0.19127451196396072,
+      "learning_rate": 4.996429131564655e-05,
+      "loss": 0.4154,
+      "step": 265
+    },
+    {
+      "epoch": 0.34908136482939633,
+      "grad_norm": 0.22385088753723567,
+      "learning_rate": 4.996222219084932e-05,
+      "loss": 0.4165,
+      "step": 266
+    },
+    {
+      "epoch": 0.35039370078740156,
+      "grad_norm": 0.19392113259193244,
+      "learning_rate": 4.9960094840405206e-05,
+      "loss": 0.4163,
+      "step": 267
+    },
+    {
+      "epoch": 0.35170603674540685,
+      "grad_norm": 0.21141253611317148,
+      "learning_rate": 4.995790926927637e-05,
+      "loss": 0.4123,
+      "step": 268
+    },
+    {
+      "epoch": 0.3530183727034121,
+      "grad_norm": 0.21547637831100683,
+      "learning_rate": 4.995566548256077e-05,
+      "loss": 0.4157,
+      "step": 269
+    },
+    {
+      "epoch": 0.3543307086614173,
+      "grad_norm": 0.2564989074889852,
+      "learning_rate": 4.9953363485492154e-05,
+      "loss": 0.4238,
+      "step": 270
+    },
+    {
+      "epoch": 0.3556430446194226,
+      "grad_norm": 0.3934015228208072,
+      "learning_rate": 4.9951003283440045e-05,
+      "loss": 0.4128,
+      "step": 271
+    },
+    {
+      "epoch": 0.3569553805774278,
+      "grad_norm": 0.5348641187932306,
+      "learning_rate": 4.994858488190973e-05,
+      "loss": 0.4141,
+      "step": 272
+    },
+    {
+      "epoch": 0.35826771653543305,
+      "grad_norm": 0.6513379830377857,
+      "learning_rate": 4.994610828654226e-05,
+      "loss": 0.4114,
+      "step": 273
+    },
+    {
+      "epoch": 0.35958005249343833,
+      "grad_norm": 0.6204154061983626,
+      "learning_rate": 4.99435735031144e-05,
+      "loss": 0.4201,
+      "step": 274
+    },
+    {
+      "epoch": 0.36089238845144356,
+      "grad_norm": 0.3369340398123888,
+      "learning_rate": 4.9940980537538696e-05,
+      "loss": 0.4217,
+      "step": 275
+    },
+    {
+      "epoch": 0.36220472440944884,
+      "grad_norm": 0.41135128043280306,
+      "learning_rate": 4.993832939586334e-05,
+      "loss": 0.4255,
+      "step": 276
+    },
+    {
+      "epoch": 0.36351706036745407,
+      "grad_norm": 0.5201936959973461,
+      "learning_rate": 4.993562008427226e-05,
+      "loss": 0.419,
+      "step": 277
+    },
+    {
+      "epoch": 0.3648293963254593,
+      "grad_norm": 0.3238326272899072,
+      "learning_rate": 4.993285260908507e-05,
+      "loss": 0.4189,
+      "step": 278
+    },
+    {
+      "epoch": 0.3661417322834646,
+      "grad_norm": 0.35700081511537773,
+      "learning_rate": 4.9930026976757035e-05,
+      "loss": 0.4138,
+      "step": 279
+    },
+    {
+      "epoch": 0.3674540682414698,
+      "grad_norm": 0.4412381093027184,
+      "learning_rate": 4.992714319387911e-05,
+      "loss": 0.4184,
+      "step": 280
+    },
+    {
+      "epoch": 0.36876640419947504,
+      "grad_norm": 0.25495019600332747,
+      "learning_rate": 4.992420126717784e-05,
+      "loss": 0.421,
+      "step": 281
+    },
+    {
+      "epoch": 0.3700787401574803,
+      "grad_norm": 0.39513738681838073,
+      "learning_rate": 4.992120120351542e-05,
+      "loss": 0.4113,
+      "step": 282
+    },
+    {
+      "epoch": 0.37139107611548555,
+      "grad_norm": 0.36234309933066755,
+      "learning_rate": 4.991814300988966e-05,
+      "loss": 0.4095,
+      "step": 283
+    },
+    {
+      "epoch": 0.37270341207349084,
+      "grad_norm": 0.28002733303566674,
+      "learning_rate": 4.9915026693433954e-05,
+      "loss": 0.4173,
+      "step": 284
+    },
+    {
+      "epoch": 0.37401574803149606,
+      "grad_norm": 0.4242150031821422,
+      "learning_rate": 4.991185226141725e-05,
+      "loss": 0.4205,
+      "step": 285
+    },
+    {
+      "epoch": 0.3753280839895013,
+      "grad_norm": 0.2533704263096715,
+      "learning_rate": 4.9908619721244086e-05,
+      "loss": 0.4215,
+      "step": 286
+    },
+    {
+      "epoch": 0.3766404199475066,
+      "grad_norm": 0.3135931983505598,
+      "learning_rate": 4.9905329080454534e-05,
+      "loss": 0.4133,
+      "step": 287
+    },
+    {
+      "epoch": 0.3779527559055118,
+      "grad_norm": 0.36499564815306074,
+      "learning_rate": 4.990198034672415e-05,
+      "loss": 0.4195,
+      "step": 288
+    },
+    {
+      "epoch": 0.37926509186351703,
+      "grad_norm": 0.28902587776213085,
+      "learning_rate": 4.989857352786406e-05,
+      "loss": 0.42,
+      "step": 289
+    },
+    {
+      "epoch": 0.3805774278215223,
+      "grad_norm": 0.306260717180519,
+      "learning_rate": 4.989510863182082e-05,
+      "loss": 0.4181,
+      "step": 290
+    },
+    {
+      "epoch": 0.38188976377952755,
+      "grad_norm": 0.2424310181517759,
+      "learning_rate": 4.989158566667648e-05,
+      "loss": 0.4101,
+      "step": 291
+    },
+    {
+      "epoch": 0.38320209973753283,
+      "grad_norm": 0.30227946612012235,
+      "learning_rate": 4.9888004640648537e-05,
+      "loss": 0.4161,
+      "step": 292
+    },
+    {
+      "epoch": 0.38451443569553806,
+      "grad_norm": 0.2396130739314267,
+      "learning_rate": 4.988436556208991e-05,
+      "loss": 0.417,
+      "step": 293
+    },
+    {
+      "epoch": 0.3858267716535433,
+      "grad_norm": 0.2466959829737264,
+      "learning_rate": 4.988066843948893e-05,
+      "loss": 0.4163,
+      "step": 294
+    },
+    {
+      "epoch": 0.38713910761154857,
+      "grad_norm": 0.2723817067062934,
+      "learning_rate": 4.987691328146934e-05,
+      "loss": 0.4225,
+      "step": 295
+    },
+    {
+      "epoch": 0.3884514435695538,
+      "grad_norm": 0.19355499232002882,
+      "learning_rate": 4.9873100096790226e-05,
+      "loss": 0.4155,
+      "step": 296
+    },
+    {
+      "epoch": 0.38976377952755903,
+      "grad_norm": 0.2357989228432189,
+      "learning_rate": 4.986922889434603e-05,
+      "loss": 0.4122,
+      "step": 297
+    },
+    {
+      "epoch": 0.3910761154855643,
+      "grad_norm": 0.2510606238968301,
+      "learning_rate": 4.986529968316653e-05,
+      "loss": 0.4107,
+      "step": 298
+    },
+    {
+      "epoch": 0.39238845144356954,
+      "grad_norm": 0.1762036795759902,
+      "learning_rate": 4.986131247241682e-05,
+      "loss": 0.4177,
+      "step": 299
+    },
+    {
+      "epoch": 0.3937007874015748,
+      "grad_norm": 0.21976478268997046,
+      "learning_rate": 4.985726727139727e-05,
+      "loss": 0.4163,
+      "step": 300
+    },
+    {
+      "epoch": 0.39501312335958005,
+      "grad_norm": 0.1544550767068837,
+      "learning_rate": 4.985316408954349e-05,
+      "loss": 0.4163,
+      "step": 301
+    },
+    {
+      "epoch": 0.3963254593175853,
+      "grad_norm": 0.18664384792849828,
+      "learning_rate": 4.9849002936426394e-05,
+      "loss": 0.4111,
+      "step": 302
+    },
+    {
+      "epoch": 0.39763779527559057,
+      "grad_norm": 0.21755201822555292,
+      "learning_rate": 4.9844783821752064e-05,
+      "loss": 0.418,
+      "step": 303
+    },
+    {
+      "epoch": 0.3989501312335958,
+      "grad_norm": 0.22814146729580476,
+      "learning_rate": 4.984050675536179e-05,
+      "loss": 0.4072,
+      "step": 304
+    },
+    {
+      "epoch": 0.400262467191601,
+      "grad_norm": 0.27432501736083154,
+      "learning_rate": 4.9836171747232066e-05,
+      "loss": 0.4118,
+      "step": 305
+    },
+    {
+      "epoch": 0.4015748031496063,
+      "grad_norm": 0.3728490128102477,
+      "learning_rate": 4.983177880747451e-05,
+      "loss": 0.4199,
+      "step": 306
+    },
+    {
+      "epoch": 0.40288713910761154,
+      "grad_norm": 0.5165294602670208,
+      "learning_rate": 4.9827327946335875e-05,
+      "loss": 0.4153,
+      "step": 307
+    },
+    {
+      "epoch": 0.4041994750656168,
+      "grad_norm": 0.6176576418316814,
+      "learning_rate": 4.982281917419802e-05,
+      "loss": 0.4176,
+      "step": 308
+    },
+    {
+      "epoch": 0.40551181102362205,
+      "grad_norm": 0.5602097511019141,
+      "learning_rate": 4.981825250157789e-05,
+      "loss": 0.4185,
+      "step": 309
+    },
+    {
+      "epoch": 0.4068241469816273,
+      "grad_norm": 0.3443485241516124,
+      "learning_rate": 4.981362793912746e-05,
+      "loss": 0.405,
+      "step": 310
+    },
+    {
+      "epoch": 0.40813648293963256,
+      "grad_norm": 0.19438480691804602,
+      "learning_rate": 4.980894549763379e-05,
+      "loss": 0.4092,
+      "step": 311
+    },
+    {
+      "epoch": 0.4094488188976378,
+      "grad_norm": 0.38649677964751666,
+      "learning_rate": 4.9804205188018885e-05,
+      "loss": 0.4243,
+      "step": 312
+    },
+    {
+      "epoch": 0.410761154855643,
+      "grad_norm": 0.3761067912185831,
+      "learning_rate": 4.979940702133976e-05,
+      "loss": 0.4212,
+      "step": 313
+    },
+    {
+      "epoch": 0.4120734908136483,
+      "grad_norm": 0.20629513384668874,
+      "learning_rate": 4.979455100878838e-05,
+      "loss": 0.4145,
+      "step": 314
+    },
+    {
+      "epoch": 0.41338582677165353,
+      "grad_norm": 0.307627471049785,
+      "learning_rate": 4.978963716169166e-05,
+      "loss": 0.4079,
+      "step": 315
+    },
+    {
+      "epoch": 0.4146981627296588,
+      "grad_norm": 0.33049263655217054,
+      "learning_rate": 4.978466549151138e-05,
+      "loss": 0.4164,
+      "step": 316
+    },
+    {
+      "epoch": 0.41601049868766404,
+      "grad_norm": 0.2527706844879446,
+      "learning_rate": 4.9779636009844216e-05,
+      "loss": 0.4226,
+      "step": 317
+    },
+    {
+      "epoch": 0.41732283464566927,
+      "grad_norm": 0.2824133662052566,
+      "learning_rate": 4.977454872842169e-05,
+      "loss": 0.4133,
+      "step": 318
+    },
+    {
+      "epoch": 0.41863517060367456,
+      "grad_norm": 0.3548926281269489,
+      "learning_rate": 4.9769403659110136e-05,
+      "loss": 0.4195,
+      "step": 319
+    },
+    {
+      "epoch": 0.4199475065616798,
+      "grad_norm": 0.24948153471376328,
+      "learning_rate": 4.97642008139107e-05,
+      "loss": 0.4079,
+      "step": 320
+    },
+    {
+      "epoch": 0.421259842519685,
+      "grad_norm": 0.5099945257216063,
+      "learning_rate": 4.975894020495928e-05,
+      "loss": 0.4308,
+      "step": 321
+    },
+    {
+      "epoch": 0.4225721784776903,
+      "grad_norm": 0.3143511840707888,
+      "learning_rate": 4.97536218445265e-05,
+      "loss": 0.4142,
+      "step": 322
+    },
+    {
+      "epoch": 0.4238845144356955,
+      "grad_norm": 1.5385788158459532,
+      "learning_rate": 4.974824574501772e-05,
+      "loss": 0.4509,
+      "step": 323
+    },
+    {
+      "epoch": 0.4251968503937008,
+      "grad_norm": 0.3350842132152767,
+      "learning_rate": 4.974281191897295e-05,
+      "loss": 0.4257,
+      "step": 324
+    },
+    {
+      "epoch": 0.42650918635170604,
+      "grad_norm": 0.5050210455315485,
+      "learning_rate": 4.973732037906688e-05,
+      "loss": 0.4342,
+      "step": 325
+    },
+    {
+      "epoch": 0.42782152230971127,
+      "grad_norm": 0.5162211802296551,
+      "learning_rate": 4.9731771138108794e-05,
+      "loss": 0.4278,
+      "step": 326
+    },
+    {
+      "epoch": 0.42913385826771655,
+      "grad_norm": 0.48324327822037666,
+      "learning_rate": 4.972616420904259e-05,
+      "loss": 0.424,
+      "step": 327
+    },
+    {
+      "epoch": 0.4304461942257218,
+      "grad_norm": 0.5483145790568777,
+      "learning_rate": 4.97204996049467e-05,
+      "loss": 0.4262,
+      "step": 328
+    },
+    {
+      "epoch": 0.431758530183727,
+      "grad_norm": 0.4511827009187962,
+      "learning_rate": 4.97147773390341e-05,
+      "loss": 0.4304,
+      "step": 329
+    },
+    {
+      "epoch": 0.4330708661417323,
+      "grad_norm": 0.35459365295412737,
+      "learning_rate": 4.970899742465227e-05,
+      "loss": 0.4304,
+      "step": 330
+    },
+    {
+      "epoch": 0.4343832020997375,
+      "grad_norm": 0.3324889035279798,
+      "learning_rate": 4.970315987528316e-05,
+      "loss": 0.4207,
+      "step": 331
+    },
+    {
+      "epoch": 0.4356955380577428,
+      "grad_norm": 0.32812036264350497,
+      "learning_rate": 4.969726470454313e-05,
+      "loss": 0.4281,
+      "step": 332
+    },
+    {
+      "epoch": 0.43700787401574803,
+      "grad_norm": 0.30379229289811516,
+      "learning_rate": 4.969131192618298e-05,
+      "loss": 0.4229,
+      "step": 333
+    },
+    {
+      "epoch": 0.43832020997375326,
+      "grad_norm": 0.24754107507181816,
+      "learning_rate": 4.968530155408787e-05,
+      "loss": 0.4189,
+      "step": 334
+    },
+    {
+      "epoch": 0.43963254593175854,
+      "grad_norm": 0.2874375639613842,
+      "learning_rate": 4.967923360227728e-05,
+      "loss": 0.4268,
+      "step": 335
+    },
+    {
+      "epoch": 0.4409448818897638,
+      "grad_norm": 0.24843816314965084,
+      "learning_rate": 4.967310808490503e-05,
+      "loss": 0.4047,
+      "step": 336
+    },
+    {
+      "epoch": 0.442257217847769,
+      "grad_norm": 0.28427141699819647,
+      "learning_rate": 4.966692501625919e-05,
+      "loss": 0.4234,
+      "step": 337
+    },
+    {
+      "epoch": 0.4435695538057743,
+      "grad_norm": 0.27496483605003824,
+      "learning_rate": 4.9660684410762095e-05,
+      "loss": 0.4162,
+      "step": 338
+    },
+    {
+      "epoch": 0.4448818897637795,
+      "grad_norm": 0.21448667310488012,
+      "learning_rate": 4.9654386282970254e-05,
+      "loss": 0.415,
+      "step": 339
+    },
+    {
+      "epoch": 0.4461942257217848,
+      "grad_norm": 0.2814578094974153,
+      "learning_rate": 4.964803064757438e-05,
+      "loss": 0.4143,
+      "step": 340
+    },
+    {
+      "epoch": 0.44750656167979,
+      "grad_norm": 0.19025326537868012,
+      "learning_rate": 4.964161751939931e-05,
+      "loss": 0.414,
+      "step": 341
+    },
+    {
+      "epoch": 0.44881889763779526,
+      "grad_norm": 0.2251974198723888,
+      "learning_rate": 4.9635146913404e-05,
+      "loss": 0.4092,
+      "step": 342
+    },
+    {
+      "epoch": 0.45013123359580054,
+      "grad_norm": 0.18260058306627053,
+      "learning_rate": 4.9628618844681445e-05,
+      "loss": 0.4245,
+      "step": 343
+    },
+    {
+      "epoch": 0.45144356955380577,
+      "grad_norm": 0.19871825730727236,
+      "learning_rate": 4.962203332845872e-05,
+      "loss": 0.4265,
+      "step": 344
+    },
+    {
+      "epoch": 0.452755905511811,
+      "grad_norm": 0.1591697551993324,
+      "learning_rate": 4.961539038009686e-05,
+      "loss": 0.4107,
+      "step": 345
+    },
+    {
+      "epoch": 0.4540682414698163,
+      "grad_norm": 0.1914094079341871,
+      "learning_rate": 4.960869001509088e-05,
+      "loss": 0.4209,
+      "step": 346
+    },
+    {
+      "epoch": 0.4553805774278215,
+      "grad_norm": 0.1502743833808356,
+      "learning_rate": 4.9601932249069724e-05,
+      "loss": 0.4131,
+      "step": 347
+    },
+    {
+      "epoch": 0.4566929133858268,
+      "grad_norm": 0.18876626357230278,
+      "learning_rate": 4.9595117097796224e-05,
+      "loss": 0.4141,
+      "step": 348
+    },
+    {
+      "epoch": 0.458005249343832,
+      "grad_norm": 0.1733121557107134,
+      "learning_rate": 4.9588244577167065e-05,
+      "loss": 0.4099,
+      "step": 349
+    },
+    {
+      "epoch": 0.45931758530183725,
+      "grad_norm": 0.18136895624370508,
+      "learning_rate": 4.958131470321275e-05,
+      "loss": 0.4153,
+      "step": 350
+    },
+    {
+      "epoch": 0.46062992125984253,
+      "grad_norm": 0.14913805090623186,
+      "learning_rate": 4.957432749209755e-05,
+      "loss": 0.4123,
+      "step": 351
+    },
+    {
+      "epoch": 0.46194225721784776,
+      "grad_norm": 0.20394966187263736,
+      "learning_rate": 4.9567282960119496e-05,
+      "loss": 0.4155,
+      "step": 352
+    },
+    {
+      "epoch": 0.463254593175853,
+      "grad_norm": 0.18848898269004555,
+      "learning_rate": 4.956018112371032e-05,
+      "loss": 0.4209,
+      "step": 353
+    },
+    {
+      "epoch": 0.4645669291338583,
+      "grad_norm": 0.2564245040689335,
+      "learning_rate": 4.955302199943542e-05,
+      "loss": 0.4169,
+      "step": 354
+    },
+    {
+      "epoch": 0.4658792650918635,
+      "grad_norm": 0.16080469043821982,
+      "learning_rate": 4.95458056039938e-05,
+      "loss": 0.4119,
+      "step": 355
+    },
+    {
+      "epoch": 0.4671916010498688,
+      "grad_norm": 0.17887056386904304,
+      "learning_rate": 4.9538531954218073e-05,
+      "loss": 0.419,
+      "step": 356
+    },
+    {
+      "epoch": 0.468503937007874,
+      "grad_norm": 0.49382525603577804,
+      "learning_rate": 4.95312010670744e-05,
+      "loss": 0.4124,
+      "step": 357
+    },
+    {
+      "epoch": 0.46981627296587924,
+      "grad_norm": 0.43817635845220515,
+      "learning_rate": 4.952381295966244e-05,
+      "loss": 0.4288,
+      "step": 358
+    },
+    {
+      "epoch": 0.47112860892388453,
+      "grad_norm": 3.199729677597055,
+      "learning_rate": 4.951636764921534e-05,
+      "loss": 0.4175,
+      "step": 359
+    },
+    {
+      "epoch": 0.47244094488188976,
+      "grad_norm": 0.6918281776515283,
+      "learning_rate": 4.950886515309965e-05,
+      "loss": 0.4346,
+      "step": 360
+    },
+    {
+      "epoch": 0.473753280839895,
+      "grad_norm": 1.2350806453602328,
+      "learning_rate": 4.950130548881532e-05,
+      "loss": 0.4405,
+      "step": 361
+    },
+    {
+      "epoch": 0.47506561679790027,
+      "grad_norm": 1.1911499335779887,
+      "learning_rate": 4.949368867399567e-05,
+      "loss": 0.431,
+      "step": 362
+    },
+    {
+      "epoch": 0.4763779527559055,
+      "grad_norm": 0.8565353831080061,
+      "learning_rate": 4.9486014726407275e-05,
+      "loss": 0.4245,
+      "step": 363
+    },
+    {
+      "epoch": 0.4776902887139108,
+      "grad_norm": 0.6115172138880747,
+      "learning_rate": 4.947828366395003e-05,
+      "loss": 0.4278,
+      "step": 364
+    },
+    {
+      "epoch": 0.479002624671916,
+      "grad_norm": 0.5916262772813263,
+      "learning_rate": 4.947049550465703e-05,
+      "loss": 0.4194,
+      "step": 365
+    },
+    {
+      "epoch": 0.48031496062992124,
+      "grad_norm": 0.5405828896918433,
+      "learning_rate": 4.9462650266694544e-05,
+      "loss": 0.4192,
+      "step": 366
+    },
+    {
+      "epoch": 0.4816272965879265,
+      "grad_norm": 0.4115716300186974,
+      "learning_rate": 4.945474796836198e-05,
+      "loss": 0.4244,
+      "step": 367
+    },
+    {
+      "epoch": 0.48293963254593175,
+      "grad_norm": 0.4278858465994355,
+      "learning_rate": 4.9446788628091854e-05,
+      "loss": 0.4209,
+      "step": 368
+    },
+    {
+      "epoch": 0.484251968503937,
+      "grad_norm": 0.34630508339985644,
+      "learning_rate": 4.943877226444974e-05,
+      "loss": 0.4208,
+      "step": 369
+    },
+    {
+      "epoch": 0.48556430446194226,
+      "grad_norm": 0.36489134975801907,
+      "learning_rate": 4.9430698896134196e-05,
+      "loss": 0.4138,
+      "step": 370
+    },
+    {
+      "epoch": 0.4868766404199475,
+      "grad_norm": 0.3385270012221531,
+      "learning_rate": 4.942256854197678e-05,
+      "loss": 0.4163,
+      "step": 371
+    },
+    {
+      "epoch": 0.4881889763779528,
+      "grad_norm": 0.3725071896327031,
+      "learning_rate": 4.941438122094193e-05,
+      "loss": 0.4216,
+      "step": 372
+    },
+    {
+      "epoch": 0.489501312335958,
+      "grad_norm": 0.3499790963615952,
+      "learning_rate": 4.9406136952127015e-05,
+      "loss": 0.4157,
+      "step": 373
+    },
+    {
+      "epoch": 0.49081364829396323,
+      "grad_norm": 0.3109234946072283,
+      "learning_rate": 4.9397835754762185e-05,
+      "loss": 0.4181,
+      "step": 374
+    },
+    {
+      "epoch": 0.4921259842519685,
+      "grad_norm": 0.3433140900694529,
+      "learning_rate": 4.938947764821042e-05,
+      "loss": 0.4109,
+      "step": 375
+    },
+    {
+      "epoch": 0.49343832020997375,
+      "grad_norm": 0.31220564143920554,
+      "learning_rate": 4.9381062651967416e-05,
+      "loss": 0.4118,
+      "step": 376
+    },
+    {
+      "epoch": 0.494750656167979,
+      "grad_norm": 0.3124549422975957,
+      "learning_rate": 4.937259078566159e-05,
+      "loss": 0.4196,
+      "step": 377
+    },
+    {
+      "epoch": 0.49606299212598426,
+      "grad_norm": 0.27385421078561056,
+      "learning_rate": 4.936406206905398e-05,
+      "loss": 0.4161,
+      "step": 378
+    },
+    {
+      "epoch": 0.4973753280839895,
+      "grad_norm": 0.2308265393435439,
+      "learning_rate": 4.935547652203827e-05,
+      "loss": 0.4217,
+      "step": 379
+    },
+    {
+      "epoch": 0.49868766404199477,
+      "grad_norm": 0.2751933750526806,
+      "learning_rate": 4.934683416464067e-05,
+      "loss": 0.4153,
+      "step": 380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21345194482569194,
+      "learning_rate": 4.933813501701992e-05,
+      "loss": 0.4065,
+      "step": 381
+    },
+    {
+      "epoch": 0.5013123359580053,
+      "grad_norm": 0.2544912768709972,
+      "learning_rate": 4.932937909946721e-05,
+      "loss": 0.4168,
+      "step": 382
+    },
+    {
+      "epoch": 0.5026246719160105,
+      "grad_norm": 0.21318431094778914,
+      "learning_rate": 4.932056643240618e-05,
+      "loss": 0.3968,
+      "step": 383
+    },
+    {
+      "epoch": 0.5039370078740157,
+      "grad_norm": 0.2529942562737851,
+      "learning_rate": 4.931169703639282e-05,
+      "loss": 0.4075,
+      "step": 384
+    },
+    {
+      "epoch": 0.505249343832021,
+      "grad_norm": 0.18085071588976748,
+      "learning_rate": 4.930277093211544e-05,
+      "loss": 0.4027,
+      "step": 385
+    },
+    {
+      "epoch": 0.5065616797900262,
+      "grad_norm": 0.1996294584655515,
+      "learning_rate": 4.9293788140394626e-05,
+      "loss": 0.406,
+      "step": 386
+    },
+    {
+      "epoch": 0.5078740157480315,
+      "grad_norm": 0.21546767093577585,
+      "learning_rate": 4.9284748682183196e-05,
+      "loss": 0.4098,
+      "step": 387
+    },
+    {
+      "epoch": 0.5091863517060368,
+      "grad_norm": 0.15617501153284202,
+      "learning_rate": 4.927565257856615e-05,
+      "loss": 0.4088,
+      "step": 388
+    },
+    {
+      "epoch": 0.510498687664042,
+      "grad_norm": 0.18951679905366606,
+      "learning_rate": 4.926649985076061e-05,
+      "loss": 0.4078,
+      "step": 389
+    },
+    {
+      "epoch": 0.5118110236220472,
+      "grad_norm": 0.15189234462587017,
+      "learning_rate": 4.925729052011576e-05,
+      "loss": 0.4106,
+      "step": 390
+    },
+    {
+      "epoch": 0.5131233595800525,
+      "grad_norm": 0.18858504393820436,
+      "learning_rate": 4.924802460811285e-05,
+      "loss": 0.4096,
+      "step": 391
+    },
+    {
+      "epoch": 0.5144356955380578,
+      "grad_norm": 0.1910420289495145,
+      "learning_rate": 4.923870213636509e-05,
+      "loss": 0.4099,
+      "step": 392
+    },
+    {
+      "epoch": 0.515748031496063,
+      "grad_norm": 0.18002133641180482,
+      "learning_rate": 4.92293231266176e-05,
+      "loss": 0.4095,
+      "step": 393
+    },
+    {
+      "epoch": 0.5170603674540682,
+      "grad_norm": 0.24801878351108522,
+      "learning_rate": 4.921988760074741e-05,
+      "loss": 0.4137,
+      "step": 394
+    },
+    {
+      "epoch": 0.5183727034120735,
+      "grad_norm": 0.15224254805942775,
+      "learning_rate": 4.921039558076335e-05,
+      "loss": 0.4164,
+      "step": 395
+    },
+    {
+      "epoch": 0.5196850393700787,
+      "grad_norm": 0.1267352802156717,
+      "learning_rate": 4.920084708880604e-05,
+      "loss": 0.4098,
+      "step": 396
+    },
+    {
+      "epoch": 0.520997375328084,
+      "grad_norm": 0.1671887905763498,
+      "learning_rate": 4.919124214714783e-05,
+      "loss": 0.4143,
+      "step": 397
+    },
+    {
+      "epoch": 0.5223097112860893,
+      "grad_norm": 0.1343305582471269,
+      "learning_rate": 4.9181580778192725e-05,
+      "loss": 0.4186,
+      "step": 398
+    },
+    {
+      "epoch": 0.5236220472440944,
+      "grad_norm": 0.16900113916501724,
+      "learning_rate": 4.917186300447636e-05,
+      "loss": 0.4201,
+      "step": 399
+    },
+    {
+      "epoch": 0.5249343832020997,
+      "grad_norm": 0.15783114058974643,
+      "learning_rate": 4.916208884866593e-05,
+      "loss": 0.4,
+      "step": 400
+    },
+    {
+      "epoch": 0.526246719160105,
+      "grad_norm": 0.12891495044483428,
+      "learning_rate": 4.915225833356015e-05,
+      "loss": 0.4067,
+      "step": 401
+    },
+    {
+      "epoch": 0.5275590551181102,
+      "grad_norm": 0.18589587146653347,
+      "learning_rate": 4.9142371482089195e-05,
+      "loss": 0.4093,
+      "step": 402
+    },
+    {
+      "epoch": 0.5288713910761155,
+      "grad_norm": 0.15762459779454557,
+      "learning_rate": 4.913242831731465e-05,
+      "loss": 0.4162,
+      "step": 403
+    },
+    {
+      "epoch": 0.5301837270341208,
+      "grad_norm": 0.11544676956124467,
+      "learning_rate": 4.912242886242945e-05,
+      "loss": 0.4139,
+      "step": 404
+    },
+    {
+      "epoch": 0.531496062992126,
+      "grad_norm": 0.17560184523345174,
+      "learning_rate": 4.9112373140757815e-05,
+      "loss": 0.4038,
+      "step": 405
+    },
+    {
+      "epoch": 0.5328083989501312,
+      "grad_norm": 0.16057388563787123,
+      "learning_rate": 4.910226117575524e-05,
+      "loss": 0.4189,
+      "step": 406
+    },
+    {
+      "epoch": 0.5341207349081365,
+      "grad_norm": 0.12296678488188359,
+      "learning_rate": 4.90920929910084e-05,
+      "loss": 0.4043,
+      "step": 407
+    },
+    {
+      "epoch": 0.5354330708661418,
+      "grad_norm": 0.14664690116397058,
+      "learning_rate": 4.908186861023508e-05,
+      "loss": 0.4147,
+      "step": 408
+    },
+    {
+      "epoch": 0.536745406824147,
+      "grad_norm": 0.13585052888290033,
+      "learning_rate": 4.907158805728419e-05,
+      "loss": 0.4044,
+      "step": 409
+    },
+    {
+      "epoch": 0.5380577427821522,
+      "grad_norm": 0.1096981216308042,
+      "learning_rate": 4.9061251356135616e-05,
+      "loss": 0.4105,
+      "step": 410
+    },
+    {
+      "epoch": 0.5393700787401575,
+      "grad_norm": 0.1212992838964438,
+      "learning_rate": 4.9050858530900255e-05,
+      "loss": 0.4016,
+      "step": 411
+    },
+    {
+      "epoch": 0.5406824146981627,
+      "grad_norm": 0.11443627290280393,
+      "learning_rate": 4.9040409605819895e-05,
+      "loss": 0.4099,
+      "step": 412
+    },
+    {
+      "epoch": 0.541994750656168,
+      "grad_norm": 0.11938820291554479,
+      "learning_rate": 4.902990460526717e-05,
+      "loss": 0.4062,
+      "step": 413
+    },
+    {
+      "epoch": 0.5433070866141733,
+      "grad_norm": 0.14375927542495223,
+      "learning_rate": 4.9019343553745544e-05,
+      "loss": 0.4022,
+      "step": 414
+    },
+    {
+      "epoch": 0.5446194225721784,
+      "grad_norm": 0.1507466203655345,
+      "learning_rate": 4.90087264758892e-05,
+      "loss": 0.4044,
+      "step": 415
+    },
+    {
+      "epoch": 0.5459317585301837,
+      "grad_norm": 0.1741057126056145,
+      "learning_rate": 4.899805339646299e-05,
+      "loss": 0.4148,
+      "step": 416
+    },
+    {
+      "epoch": 0.547244094488189,
+      "grad_norm": 0.15087890526753722,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 0.4131,
+      "step": 417
+    },
+    {
+      "epoch": 0.5485564304461942,
+      "grad_norm": 0.11535651112492006,
+      "learning_rate": 4.897653933261359e-05,
+      "loss": 0.4062,
+      "step": 418
+    },
+    {
+      "epoch": 0.5498687664041995,
+      "grad_norm": 0.11442948252905392,
+      "learning_rate": 4.896569839837304e-05,
+      "loss": 0.401,
+      "step": 419
+    },
+    {
+      "epoch": 0.5511811023622047,
+      "grad_norm": 0.12129241041654187,
+      "learning_rate": 4.895480156292781e-05,
+      "loss": 0.4013,
+      "step": 420
+    },
+    {
+      "epoch": 0.55249343832021,
+      "grad_norm": 0.13267647136455357,
+      "learning_rate": 4.89438488516953e-05,
+      "loss": 0.4082,
+      "step": 421
+    },
+    {
+      "epoch": 0.5538057742782152,
+      "grad_norm": 0.13276140140902706,
+      "learning_rate": 4.8932840290223306e-05,
+      "loss": 0.4012,
+      "step": 422
+    },
+    {
+      "epoch": 0.5551181102362205,
+      "grad_norm": 0.1273151697693719,
+      "learning_rate": 4.892177590418982e-05,
+      "loss": 0.4061,
+      "step": 423
+    },
+    {
+      "epoch": 0.5564304461942258,
+      "grad_norm": 0.1278270792491001,
+      "learning_rate": 4.891065571940309e-05,
+      "loss": 0.4015,
+      "step": 424
+    },
+    {
+      "epoch": 0.5577427821522309,
+      "grad_norm": 0.12238183153423313,
+      "learning_rate": 4.889947976180151e-05,
+      "loss": 0.418,
+      "step": 425
+    },
+    {
+      "epoch": 0.5590551181102362,
+      "grad_norm": 0.15166084925258094,
+      "learning_rate": 4.888824805745357e-05,
+      "loss": 0.4192,
+      "step": 426
+    },
+    {
+      "epoch": 0.5603674540682415,
+      "grad_norm": 0.14220616180288803,
+      "learning_rate": 4.887696063255779e-05,
+      "loss": 0.4028,
+      "step": 427
+    },
+    {
+      "epoch": 0.5616797900262467,
+      "grad_norm": 0.15494642386445662,
+      "learning_rate": 4.886561751344266e-05,
+      "loss": 0.4117,
+      "step": 428
+    },
+    {
+      "epoch": 0.562992125984252,
+      "grad_norm": 0.13331060573936604,
+      "learning_rate": 4.885421872656658e-05,
+      "loss": 0.4069,
+      "step": 429
+    },
+    {
+      "epoch": 0.5643044619422573,
+      "grad_norm": 0.13121870597103924,
+      "learning_rate": 4.8842764298517796e-05,
+      "loss": 0.4145,
+      "step": 430
+    },
+    {
+      "epoch": 0.5656167979002624,
+      "grad_norm": 0.13879675433276448,
+      "learning_rate": 4.8831254256014335e-05,
+      "loss": 0.4156,
+      "step": 431
+    },
+    {
+      "epoch": 0.5669291338582677,
+      "grad_norm": 0.16619848040804633,
+      "learning_rate": 4.881968862590396e-05,
+      "loss": 0.404,
+      "step": 432
+    },
+    {
+      "epoch": 0.568241469816273,
+      "grad_norm": 0.18080366916429533,
+      "learning_rate": 4.880806743516409e-05,
+      "loss": 0.4184,
+      "step": 433
+    },
+    {
+      "epoch": 0.5695538057742782,
+      "grad_norm": 0.1610141740715539,
+      "learning_rate": 4.879639071090174e-05,
+      "loss": 0.4037,
+      "step": 434
+    },
+    {
+      "epoch": 0.5708661417322834,
+      "grad_norm": 0.1249113661644865,
+      "learning_rate": 4.878465848035346e-05,
+      "loss": 0.396,
+      "step": 435
+    },
+    {
+      "epoch": 0.5721784776902887,
+      "grad_norm": 0.1620802916325431,
+      "learning_rate": 4.877287077088527e-05,
+      "loss": 0.4099,
+      "step": 436
+    },
+    {
+      "epoch": 0.573490813648294,
+      "grad_norm": 0.7634409669193102,
+      "learning_rate": 4.8761027609992596e-05,
+      "loss": 0.4072,
+      "step": 437
+    },
+    {
+      "epoch": 0.5748031496062992,
+      "grad_norm": 0.14374272104593716,
+      "learning_rate": 4.874912902530022e-05,
+      "loss": 0.4115,
+      "step": 438
+    },
+    {
+      "epoch": 0.5761154855643045,
+      "grad_norm": 0.16607378060419561,
+      "learning_rate": 4.873717504456219e-05,
+      "loss": 0.4118,
+      "step": 439
+    },
+    {
+      "epoch": 0.5774278215223098,
+      "grad_norm": 0.17785007398830333,
+      "learning_rate": 4.872516569566176e-05,
+      "loss": 0.4148,
+      "step": 440
+    },
+    {
+      "epoch": 0.5787401574803149,
+      "grad_norm": 0.1979436248025201,
+      "learning_rate": 4.871310100661136e-05,
+      "loss": 0.4089,
+      "step": 441
+    },
+    {
+      "epoch": 0.5800524934383202,
+      "grad_norm": 0.20906513565273158,
+      "learning_rate": 4.870098100555247e-05,
+      "loss": 0.4115,
+      "step": 442
+    },
+    {
+      "epoch": 0.5813648293963255,
+      "grad_norm": 0.22791305351196015,
+      "learning_rate": 4.868880572075562e-05,
+      "loss": 0.3999,
+      "step": 443
+    },
+    {
+      "epoch": 0.5826771653543307,
+      "grad_norm": 0.21704435726901897,
+      "learning_rate": 4.867657518062029e-05,
+      "loss": 0.4038,
+      "step": 444
+    },
+    {
+      "epoch": 0.583989501312336,
+      "grad_norm": 0.19410624605254911,
+      "learning_rate": 4.86642894136748e-05,
+      "loss": 0.4053,
+      "step": 445
+    },
+    {
+      "epoch": 0.5853018372703412,
+      "grad_norm": 0.15968443933580007,
+      "learning_rate": 4.865194844857635e-05,
+      "loss": 0.4015,
+      "step": 446
+    },
+    {
+      "epoch": 0.5866141732283464,
+      "grad_norm": 0.17970635280894542,
+      "learning_rate": 4.863955231411086e-05,
+      "loss": 0.4011,
+      "step": 447
+    },
+    {
+      "epoch": 0.5879265091863517,
+      "grad_norm": 0.15502104653870377,
+      "learning_rate": 4.8627101039192936e-05,
+      "loss": 0.4077,
+      "step": 448
+    },
+    {
+      "epoch": 0.589238845144357,
+      "grad_norm": 0.1065813605525407,
+      "learning_rate": 4.861459465286581e-05,
+      "loss": 0.4108,
+      "step": 449
+    },
+    {
+      "epoch": 0.5905511811023622,
+      "grad_norm": 0.17267684648963488,
+      "learning_rate": 4.860203318430126e-05,
+      "loss": 0.4055,
+      "step": 450
+    },
+    {
+      "epoch": 0.5918635170603674,
+      "grad_norm": 0.22872108955084253,
+      "learning_rate": 4.858941666279955e-05,
+      "loss": 0.4139,
+      "step": 451
+    },
+    {
+      "epoch": 0.5931758530183727,
+      "grad_norm": 0.22353322341081291,
+      "learning_rate": 4.857674511778936e-05,
+      "loss": 0.4032,
+      "step": 452
+    },
+    {
+      "epoch": 0.594488188976378,
+      "grad_norm": 0.17726102193595658,
+      "learning_rate": 4.85640185788277e-05,
+      "loss": 0.4097,
+      "step": 453
+    },
+    {
+      "epoch": 0.5958005249343832,
+      "grad_norm": 0.15878060948243783,
+      "learning_rate": 4.8551237075599864e-05,
+      "loss": 0.4019,
+      "step": 454
+    },
+    {
+      "epoch": 0.5971128608923885,
+      "grad_norm": 0.1547374714490182,
+      "learning_rate": 4.853840063791936e-05,
+      "loss": 0.4033,
+      "step": 455
+    },
+    {
+      "epoch": 0.5984251968503937,
+      "grad_norm": 0.14114080055333156,
+      "learning_rate": 4.852550929572784e-05,
+      "loss": 0.4036,
+      "step": 456
+    },
+    {
+      "epoch": 0.5997375328083989,
+      "grad_norm": 0.1497988009515872,
+      "learning_rate": 4.8512563079094994e-05,
+      "loss": 0.4048,
+      "step": 457
+    },
+    {
+      "epoch": 0.6010498687664042,
+      "grad_norm": 0.1793595280264377,
+      "learning_rate": 4.849956201821855e-05,
+      "loss": 0.4034,
+      "step": 458
+    },
+    {
+      "epoch": 0.6023622047244095,
+      "grad_norm": 0.16882615570538992,
+      "learning_rate": 4.848650614342411e-05,
+      "loss": 0.4031,
+      "step": 459
+    },
+    {
+      "epoch": 0.6036745406824147,
+      "grad_norm": 0.14096299468442725,
+      "learning_rate": 4.847339548516519e-05,
+      "loss": 0.3906,
+      "step": 460
+    },
+    {
+      "epoch": 0.60498687664042,
+      "grad_norm": 0.16629400809145173,
+      "learning_rate": 4.846023007402305e-05,
+      "loss": 0.4044,
+      "step": 461
+    },
+    {
+      "epoch": 0.6062992125984252,
+      "grad_norm": 0.1885150781995525,
+      "learning_rate": 4.844700994070668e-05,
+      "loss": 0.4089,
+      "step": 462
+    },
+    {
+      "epoch": 0.6076115485564304,
+      "grad_norm": 0.15253727552993576,
+      "learning_rate": 4.8433735116052715e-05,
+      "loss": 0.4078,
+      "step": 463
+    },
+    {
+      "epoch": 0.6089238845144357,
+      "grad_norm": 0.13614938910819904,
+      "learning_rate": 4.8420405631025344e-05,
+      "loss": 0.4033,
+      "step": 464
+    },
+    {
+      "epoch": 0.610236220472441,
+      "grad_norm": 0.14778278320165314,
+      "learning_rate": 4.8407021516716285e-05,
+      "loss": 0.4068,
+      "step": 465
+    },
+    {
+      "epoch": 0.6115485564304461,
+      "grad_norm": 0.15120024747617894,
+      "learning_rate": 4.839358280434464e-05,
+      "loss": 0.4114,
+      "step": 466
+    },
+    {
+      "epoch": 0.6128608923884514,
+      "grad_norm": 0.15682303992219795,
+      "learning_rate": 4.838008952525691e-05,
+      "loss": 0.399,
+      "step": 467
+    },
+    {
+      "epoch": 0.6141732283464567,
+      "grad_norm": 0.134750172636718,
+      "learning_rate": 4.8366541710926825e-05,
+      "loss": 0.3942,
+      "step": 468
+    },
+    {
+      "epoch": 0.615485564304462,
+      "grad_norm": 0.17194824938366796,
+      "learning_rate": 4.835293939295538e-05,
+      "loss": 0.3985,
+      "step": 469
+    },
+    {
+      "epoch": 0.6167979002624672,
+      "grad_norm": 0.1725968305155504,
+      "learning_rate": 4.833928260307064e-05,
+      "loss": 0.399,
+      "step": 470
+    },
+    {
+      "epoch": 0.6181102362204725,
+      "grad_norm": 0.16021164543722893,
+      "learning_rate": 4.832557137312776e-05,
+      "loss": 0.4084,
+      "step": 471
+    },
+    {
+      "epoch": 0.6194225721784777,
+      "grad_norm": 0.14819698960977304,
+      "learning_rate": 4.8311805735108894e-05,
+      "loss": 0.4008,
+      "step": 472
+    },
+    {
+      "epoch": 0.6207349081364829,
+      "grad_norm": 0.17664353577922517,
+      "learning_rate": 4.829798572112309e-05,
+      "loss": 0.4038,
+      "step": 473
+    },
+    {
+      "epoch": 0.6220472440944882,
+      "grad_norm": 0.14789993883822575,
+      "learning_rate": 4.828411136340622e-05,
+      "loss": 0.4089,
+      "step": 474
+    },
+    {
+      "epoch": 0.6233595800524935,
+      "grad_norm": 0.16663531762051542,
+      "learning_rate": 4.827018269432093e-05,
+      "loss": 0.4132,
+      "step": 475
+    },
+    {
+      "epoch": 0.6246719160104987,
+      "grad_norm": 0.18025273459427973,
+      "learning_rate": 4.825619974635655e-05,
+      "loss": 0.3989,
+      "step": 476
+    },
+    {
+      "epoch": 0.6259842519685039,
+      "grad_norm": 0.1719437001360498,
+      "learning_rate": 4.8242162552129e-05,
+      "loss": 0.4049,
+      "step": 477
+    },
+    {
+      "epoch": 0.6272965879265092,
+      "grad_norm": 0.18893098546126064,
+      "learning_rate": 4.822807114438076e-05,
+      "loss": 0.404,
+      "step": 478
+    },
+    {
+      "epoch": 0.6286089238845144,
+      "grad_norm": 0.18328819210780264,
+      "learning_rate": 4.821392555598074e-05,
+      "loss": 0.4051,
+      "step": 479
+    },
+    {
+      "epoch": 0.6299212598425197,
+      "grad_norm": 0.16632898383891148,
+      "learning_rate": 4.8199725819924254e-05,
+      "loss": 0.4144,
+      "step": 480
+    },
+    {
+      "epoch": 0.631233595800525,
+      "grad_norm": 0.15628465949618872,
+      "learning_rate": 4.81854719693329e-05,
+      "loss": 0.4124,
+      "step": 481
+    },
+    {
+      "epoch": 0.6325459317585301,
+      "grad_norm": 0.16218796520436057,
+      "learning_rate": 4.817116403745451e-05,
+      "loss": 0.4017,
+      "step": 482
+    },
+    {
+      "epoch": 0.6338582677165354,
+      "grad_norm": 0.16943827330192357,
+      "learning_rate": 4.815680205766304e-05,
+      "loss": 0.4015,
+      "step": 483
+    },
+    {
+      "epoch": 0.6351706036745407,
+      "grad_norm": 0.12191190452290061,
+      "learning_rate": 4.814238606345857e-05,
+      "loss": 0.4026,
+      "step": 484
+    },
+    {
+      "epoch": 0.636482939632546,
+      "grad_norm": 0.11558599778192816,
+      "learning_rate": 4.81279160884671e-05,
+      "loss": 0.3961,
+      "step": 485
+    },
+    {
+      "epoch": 0.6377952755905512,
+      "grad_norm": 0.13787835127779655,
+      "learning_rate": 4.8113392166440596e-05,
+      "loss": 0.4091,
+      "step": 486
+    },
+    {
+      "epoch": 0.6391076115485564,
+      "grad_norm": 0.12368655018030755,
+      "learning_rate": 4.8098814331256845e-05,
+      "loss": 0.4059,
+      "step": 487
+    },
+    {
+      "epoch": 0.6404199475065617,
+      "grad_norm": 0.1331078903490741,
+      "learning_rate": 4.808418261691938e-05,
+      "loss": 0.4036,
+      "step": 488
+    },
+    {
+      "epoch": 0.6417322834645669,
+      "grad_norm": 0.11889665106436335,
+      "learning_rate": 4.806949705755741e-05,
+      "loss": 0.4009,
+      "step": 489
+    },
+    {
+      "epoch": 0.6430446194225722,
+      "grad_norm": 0.1447248479134384,
+      "learning_rate": 4.805475768742576e-05,
+      "loss": 0.4003,
+      "step": 490
+    },
+    {
+      "epoch": 0.6443569553805775,
+      "grad_norm": 0.1794178917692529,
+      "learning_rate": 4.8039964540904746e-05,
+      "loss": 0.4102,
+      "step": 491
+    },
+    {
+      "epoch": 0.6456692913385826,
+      "grad_norm": 0.16849972652063158,
+      "learning_rate": 4.802511765250014e-05,
+      "loss": 0.406,
+      "step": 492
+    },
+    {
+      "epoch": 0.6469816272965879,
+      "grad_norm": 0.1482820474812675,
+      "learning_rate": 4.801021705684304e-05,
+      "loss": 0.4095,
+      "step": 493
+    },
+    {
+      "epoch": 0.6482939632545932,
+      "grad_norm": 0.16492521865517187,
+      "learning_rate": 4.799526278868987e-05,
+      "loss": 0.4002,
+      "step": 494
+    },
+    {
+      "epoch": 0.6496062992125984,
+      "grad_norm": 0.19680477988404224,
+      "learning_rate": 4.7980254882922194e-05,
+      "loss": 0.4025,
+      "step": 495
+    },
+    {
+      "epoch": 0.6509186351706037,
+      "grad_norm": 0.19339006624224625,
+      "learning_rate": 4.7965193374546714e-05,
+      "loss": 0.4004,
+      "step": 496
+    },
+    {
+      "epoch": 0.652230971128609,
+      "grad_norm": 0.1427463161644861,
+      "learning_rate": 4.795007829869518e-05,
+      "loss": 0.4031,
+      "step": 497
+    },
+    {
+      "epoch": 0.6535433070866141,
+      "grad_norm": 0.11417785579506787,
+      "learning_rate": 4.793490969062424e-05,
+      "loss": 0.4058,
+      "step": 498
+    },
+    {
+      "epoch": 0.6548556430446194,
+      "grad_norm": 0.1396933280731565,
+      "learning_rate": 4.7919687585715476e-05,
+      "loss": 0.4024,
+      "step": 499
+    },
+    {
+      "epoch": 0.6561679790026247,
+      "grad_norm": 0.1518144061964623,
+      "learning_rate": 4.7904412019475186e-05,
+      "loss": 0.414,
+      "step": 500
+    },
+    {
+      "epoch": 0.65748031496063,
+      "grad_norm": 0.1534085899564797,
+      "learning_rate": 4.788908302753442e-05,
+      "loss": 0.4064,
+      "step": 501
+    },
+    {
+      "epoch": 0.6587926509186351,
+      "grad_norm": 0.1356754796990492,
+      "learning_rate": 4.787370064564883e-05,
+      "loss": 0.4135,
+      "step": 502
+    },
+    {
+      "epoch": 0.6601049868766404,
+      "grad_norm": 0.13466043885547815,
+      "learning_rate": 4.785826490969859e-05,
+      "loss": 0.4081,
+      "step": 503
+    },
+    {
+      "epoch": 0.6614173228346457,
+      "grad_norm": 0.1334919264432201,
+      "learning_rate": 4.784277585568834e-05,
+      "loss": 0.4063,
+      "step": 504
+    },
+    {
+      "epoch": 0.6627296587926509,
+      "grad_norm": 0.14512693876258922,
+      "learning_rate": 4.7827233519747086e-05,
+      "loss": 0.4083,
+      "step": 505
+    },
+    {
+      "epoch": 0.6640419947506562,
+      "grad_norm": 0.1547509151010453,
+      "learning_rate": 4.7811637938128104e-05,
+      "loss": 0.4092,
+      "step": 506
+    },
+    {
+      "epoch": 0.6653543307086615,
+      "grad_norm": 0.15632952592640692,
+      "learning_rate": 4.779598914720888e-05,
+      "loss": 0.4114,
+      "step": 507
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.2528733175540486,
+      "learning_rate": 4.778028718349102e-05,
+      "loss": 0.3984,
+      "step": 508
+    },
+    {
+      "epoch": 0.6679790026246719,
+      "grad_norm": 0.1319242920503639,
+      "learning_rate": 4.776453208360014e-05,
+      "loss": 0.4045,
+      "step": 509
+    },
+    {
+      "epoch": 0.6692913385826772,
+      "grad_norm": 0.19018565841116652,
+      "learning_rate": 4.7748723884285804e-05,
+      "loss": 0.4031,
+      "step": 510
+    },
+    {
+      "epoch": 0.6706036745406824,
+      "grad_norm": 0.251341580117547,
+      "learning_rate": 4.773286262242146e-05,
+      "loss": 0.4084,
+      "step": 511
+    },
+    {
+      "epoch": 0.6719160104986877,
+      "grad_norm": 0.3047883870253671,
+      "learning_rate": 4.771694833500428e-05,
+      "loss": 0.4034,
+      "step": 512
+    },
+    {
+      "epoch": 0.6732283464566929,
+      "grad_norm": 0.3212654324054873,
+      "learning_rate": 4.770098105915515e-05,
+      "loss": 0.4035,
+      "step": 513
+    },
+    {
+      "epoch": 0.6745406824146981,
+      "grad_norm": 0.31025875803465713,
+      "learning_rate": 4.768496083211856e-05,
+      "loss": 0.412,
+      "step": 514
+    },
+    {
+      "epoch": 0.6758530183727034,
+      "grad_norm": 0.2713916776745543,
+      "learning_rate": 4.766888769126249e-05,
+      "loss": 0.4057,
+      "step": 515
+    },
+    {
+      "epoch": 0.6771653543307087,
+      "grad_norm": 0.2671285186921146,
+      "learning_rate": 4.765276167407835e-05,
+      "loss": 0.3938,
+      "step": 516
+    },
+    {
+      "epoch": 0.678477690288714,
+      "grad_norm": 0.2850292073741704,
+      "learning_rate": 4.763658281818091e-05,
+      "loss": 0.4037,
+      "step": 517
+    },
+    {
+      "epoch": 0.6797900262467191,
+      "grad_norm": 0.31712502831093003,
+      "learning_rate": 4.762035116130816e-05,
+      "loss": 0.4063,
+      "step": 518
+    },
+    {
+      "epoch": 0.6811023622047244,
+      "grad_norm": 0.32769732743333563,
+      "learning_rate": 4.760406674132126e-05,
+      "loss": 0.3991,
+      "step": 519
+    },
+    {
+      "epoch": 0.6824146981627297,
+      "grad_norm": 0.24553405178701068,
+      "learning_rate": 4.7587729596204446e-05,
+      "loss": 0.4008,
+      "step": 520
+    },
+    {
+      "epoch": 0.6837270341207349,
+      "grad_norm": 0.21483438223076357,
+      "learning_rate": 4.757133976406493e-05,
+      "loss": 0.3948,
+      "step": 521
+    },
+    {
+      "epoch": 0.6850393700787402,
+      "grad_norm": 0.1633085374973873,
+      "learning_rate": 4.7554897283132824e-05,
+      "loss": 0.399,
+      "step": 522
+    },
+    {
+      "epoch": 0.6863517060367454,
+      "grad_norm": 0.16080266623057232,
+      "learning_rate": 4.753840219176107e-05,
+      "loss": 0.3951,
+      "step": 523
+    },
+    {
+      "epoch": 0.6876640419947506,
+      "grad_norm": 0.19845393228509373,
+      "learning_rate": 4.752185452842527e-05,
+      "loss": 0.4066,
+      "step": 524
+    },
+    {
+      "epoch": 0.6889763779527559,
+      "grad_norm": 0.2287931212030834,
+      "learning_rate": 4.750525433172371e-05,
+      "loss": 0.4121,
+      "step": 525
+    },
+    {
+      "epoch": 0.6902887139107612,
+      "grad_norm": 0.21088832361067936,
+      "learning_rate": 4.7488601640377183e-05,
+      "loss": 0.4052,
+      "step": 526
+    },
+    {
+      "epoch": 0.6916010498687664,
+      "grad_norm": 0.1479675439023272,
+      "learning_rate": 4.747189649322894e-05,
+      "loss": 0.3998,
+      "step": 527
+    },
+    {
+      "epoch": 0.6929133858267716,
+      "grad_norm": 0.15689381666324903,
+      "learning_rate": 4.745513892924457e-05,
+      "loss": 0.4052,
+      "step": 528
+    },
+    {
+      "epoch": 0.6942257217847769,
+      "grad_norm": 0.19264370129710942,
+      "learning_rate": 4.7438328987511945e-05,
+      "loss": 0.4046,
+      "step": 529
+    },
+    {
+      "epoch": 0.6955380577427821,
+      "grad_norm": 0.18420360797453247,
+      "learning_rate": 4.74214667072411e-05,
+      "loss": 0.4097,
+      "step": 530
+    },
+    {
+      "epoch": 0.6968503937007874,
+      "grad_norm": 0.20505773567867194,
+      "learning_rate": 4.740455212776417e-05,
+      "loss": 0.3978,
+      "step": 531
+    },
+    {
+      "epoch": 0.6981627296587927,
+      "grad_norm": 0.19049880004924197,
+      "learning_rate": 4.738758528853526e-05,
+      "loss": 0.404,
+      "step": 532
+    },
+    {
+      "epoch": 0.699475065616798,
+      "grad_norm": 0.14404379508271933,
+      "learning_rate": 4.7370566229130385e-05,
+      "loss": 0.4108,
+      "step": 533
+    },
+    {
+      "epoch": 0.7007874015748031,
+      "grad_norm": 0.20710015023862424,
+      "learning_rate": 4.735349498924736e-05,
+      "loss": 0.3968,
+      "step": 534
+    },
+    {
+      "epoch": 0.7020997375328084,
+      "grad_norm": 0.15471958499621985,
+      "learning_rate": 4.733637160870571e-05,
+      "loss": 0.4017,
+      "step": 535
+    },
+    {
+      "epoch": 0.7034120734908137,
+      "grad_norm": 0.1320725477996808,
+      "learning_rate": 4.73191961274466e-05,
+      "loss": 0.4076,
+      "step": 536
+    },
+    {
+      "epoch": 0.7047244094488189,
+      "grad_norm": 0.16691161873522453,
+      "learning_rate": 4.730196858553269e-05,
+      "loss": 0.4108,
+      "step": 537
+    },
+    {
+      "epoch": 0.7060367454068242,
+      "grad_norm": 0.23764639701032267,
+      "learning_rate": 4.728468902314812e-05,
+      "loss": 0.4109,
+      "step": 538
+    },
+    {
+      "epoch": 0.7073490813648294,
+      "grad_norm": 0.19337465013129762,
+      "learning_rate": 4.726735748059832e-05,
+      "loss": 0.405,
+      "step": 539
+    },
+    {
+      "epoch": 0.7086614173228346,
+      "grad_norm": 0.1733836701851396,
+      "learning_rate": 4.7249973998310005e-05,
+      "loss": 0.4071,
+      "step": 540
+    },
+    {
+      "epoch": 0.7099737532808399,
+      "grad_norm": 0.16604467674129533,
+      "learning_rate": 4.723253861683101e-05,
+      "loss": 0.4019,
+      "step": 541
+    },
+    {
+      "epoch": 0.7112860892388452,
+      "grad_norm": 0.15023476649364237,
+      "learning_rate": 4.7215051376830275e-05,
+      "loss": 0.4003,
+      "step": 542
+    },
+    {
+      "epoch": 0.7125984251968503,
+      "grad_norm": 0.4566236388676907,
+      "learning_rate": 4.719751231909765e-05,
+      "loss": 0.4121,
+      "step": 543
+    },
+    {
+      "epoch": 0.7139107611548556,
+      "grad_norm": 0.12534760245375368,
+      "learning_rate": 4.717992148454388e-05,
+      "loss": 0.3945,
+      "step": 544
+    },
+    {
+      "epoch": 0.7152230971128609,
+      "grad_norm": 0.13681433907479482,
+      "learning_rate": 4.716227891420047e-05,
+      "loss": 0.4044,
+      "step": 545
+    },
+    {
+      "epoch": 0.7165354330708661,
+      "grad_norm": 0.16129092654710506,
+      "learning_rate": 4.714458464921962e-05,
+      "loss": 0.4063,
+      "step": 546
+    },
+    {
+      "epoch": 0.7178477690288714,
+      "grad_norm": 0.16879936312135807,
+      "learning_rate": 4.7126838730874096e-05,
+      "loss": 0.3992,
+      "step": 547
+    },
+    {
+      "epoch": 0.7191601049868767,
+      "grad_norm": 0.16696781529725252,
+      "learning_rate": 4.7109041200557146e-05,
+      "loss": 0.3873,
+      "step": 548
+    },
+    {
+      "epoch": 0.7204724409448819,
+      "grad_norm": 0.3216133178608258,
+      "learning_rate": 4.709119209978242e-05,
+      "loss": 0.402,
+      "step": 549
+    },
+    {
+      "epoch": 0.7217847769028871,
+      "grad_norm": 0.16705909045785755,
+      "learning_rate": 4.707329147018383e-05,
+      "loss": 0.4079,
+      "step": 550
+    },
+    {
+      "epoch": 0.7230971128608924,
+      "grad_norm": 0.1535385992840315,
+      "learning_rate": 4.7055339353515526e-05,
+      "loss": 0.3948,
+      "step": 551
+    },
+    {
+      "epoch": 0.7244094488188977,
+      "grad_norm": 0.1608491448302763,
+      "learning_rate": 4.703733579165171e-05,
+      "loss": 0.4017,
+      "step": 552
+    },
+    {
+      "epoch": 0.7257217847769029,
+      "grad_norm": 0.17579798760103818,
+      "learning_rate": 4.7019280826586606e-05,
+      "loss": 0.416,
+      "step": 553
+    },
+    {
+      "epoch": 0.7270341207349081,
+      "grad_norm": 0.17008683684566311,
+      "learning_rate": 4.7001174500434345e-05,
+      "loss": 0.4081,
+      "step": 554
+    },
+    {
+      "epoch": 0.7283464566929134,
+      "grad_norm": 0.14677761197073813,
+      "learning_rate": 4.698301685542883e-05,
+      "loss": 0.4075,
+      "step": 555
+    },
+    {
+      "epoch": 0.7296587926509186,
+      "grad_norm": 0.15324775903832252,
+      "learning_rate": 4.69648079339237e-05,
+      "loss": 0.3976,
+      "step": 556
+    },
+    {
+      "epoch": 0.7309711286089239,
+      "grad_norm": 0.15004022785462737,
+      "learning_rate": 4.694654777839218e-05,
+      "loss": 0.4064,
+      "step": 557
+    },
+    {
+      "epoch": 0.7322834645669292,
+      "grad_norm": 0.17477899222792434,
+      "learning_rate": 4.692823643142702e-05,
+      "loss": 0.4042,
+      "step": 558
+    },
+    {
+      "epoch": 0.7335958005249343,
+      "grad_norm": 0.20064098249424359,
+      "learning_rate": 4.6909873935740345e-05,
+      "loss": 0.407,
+      "step": 559
+    },
+    {
+      "epoch": 0.7349081364829396,
+      "grad_norm": 0.18522086342154886,
+      "learning_rate": 4.689146033416362e-05,
+      "loss": 0.4097,
+      "step": 560
+    },
+    {
+      "epoch": 0.7362204724409449,
+      "grad_norm": 0.13380290909288708,
+      "learning_rate": 4.6872995669647493e-05,
+      "loss": 0.3957,
+      "step": 561
+    },
+    {
+      "epoch": 0.7375328083989501,
+      "grad_norm": 0.1676001947376276,
+      "learning_rate": 4.685447998526174e-05,
+      "loss": 0.406,
+      "step": 562
+    },
+    {
+      "epoch": 0.7388451443569554,
+      "grad_norm": 0.1647264099821427,
+      "learning_rate": 4.683591332419511e-05,
+      "loss": 0.4055,
+      "step": 563
+    },
+    {
+      "epoch": 0.7401574803149606,
+      "grad_norm": 0.17171830933452345,
+      "learning_rate": 4.681729572975531e-05,
+      "loss": 0.397,
+      "step": 564
+    },
+    {
+      "epoch": 0.7414698162729659,
+      "grad_norm": 0.18768862804100056,
+      "learning_rate": 4.67986272453688e-05,
+      "loss": 0.4084,
+      "step": 565
+    },
+    {
+      "epoch": 0.7427821522309711,
+      "grad_norm": 0.2067316342492148,
+      "learning_rate": 4.677990791458077e-05,
+      "loss": 0.4101,
+      "step": 566
+    },
+    {
+      "epoch": 0.7440944881889764,
+      "grad_norm": 0.180162534790244,
+      "learning_rate": 4.676113778105501e-05,
+      "loss": 0.3946,
+      "step": 567
+    },
+    {
+      "epoch": 0.7454068241469817,
+      "grad_norm": 0.17290889670386309,
+      "learning_rate": 4.674231688857381e-05,
+      "loss": 0.403,
+      "step": 568
+    },
+    {
+      "epoch": 0.7467191601049868,
+      "grad_norm": 0.18303252073626708,
+      "learning_rate": 4.6723445281037845e-05,
+      "loss": 0.4049,
+      "step": 569
+    },
+    {
+      "epoch": 0.7480314960629921,
+      "grad_norm": 0.1994522468254952,
+      "learning_rate": 4.67045230024661e-05,
+      "loss": 0.4009,
+      "step": 570
+    },
+    {
+      "epoch": 0.7493438320209974,
+      "grad_norm": 0.20400140432804176,
+      "learning_rate": 4.6685550096995744e-05,
+      "loss": 0.3998,
+      "step": 571
+    },
+    {
+      "epoch": 0.7506561679790026,
+      "grad_norm": 0.18058308139906912,
+      "learning_rate": 4.666652660888205e-05,
+      "loss": 0.3982,
+      "step": 572
+    },
+    {
+      "epoch": 0.7519685039370079,
+      "grad_norm": 0.14252446986078984,
+      "learning_rate": 4.664745258249825e-05,
+      "loss": 0.4034,
+      "step": 573
+    },
+    {
+      "epoch": 0.7532808398950132,
+      "grad_norm": 0.1301415141119463,
+      "learning_rate": 4.6628328062335484e-05,
+      "loss": 0.4113,
+      "step": 574
+    },
+    {
+      "epoch": 0.7545931758530183,
+      "grad_norm": 0.1679422734287755,
+      "learning_rate": 4.660915309300267e-05,
+      "loss": 0.3996,
+      "step": 575
+    },
+    {
+      "epoch": 0.7559055118110236,
+      "grad_norm": 0.19774963224413805,
+      "learning_rate": 4.658992771922639e-05,
+      "loss": 0.4135,
+      "step": 576
+    },
+    {
+      "epoch": 0.7572178477690289,
+      "grad_norm": 0.15986427487720803,
+      "learning_rate": 4.65706519858508e-05,
+      "loss": 0.4,
+      "step": 577
+    },
+    {
+      "epoch": 0.7585301837270341,
+      "grad_norm": 0.20150812520875988,
+      "learning_rate": 4.655132593783752e-05,
+      "loss": 0.4049,
+      "step": 578
+    },
+    {
+      "epoch": 0.7598425196850394,
+      "grad_norm": 0.23150066574247125,
+      "learning_rate": 4.653194962026554e-05,
+      "loss": 0.3962,
+      "step": 579
+    },
+    {
+      "epoch": 0.7611548556430446,
+      "grad_norm": 0.20221605014221597,
+      "learning_rate": 4.65125230783311e-05,
+      "loss": 0.4035,
+      "step": 580
+    },
+    {
+      "epoch": 0.7624671916010499,
+      "grad_norm": 0.1558911307340604,
+      "learning_rate": 4.6493046357347604e-05,
+      "loss": 0.3964,
+      "step": 581
+    },
+    {
+      "epoch": 0.7637795275590551,
+      "grad_norm": 0.15684646094723179,
+      "learning_rate": 4.6473519502745476e-05,
+      "loss": 0.3932,
+      "step": 582
+    },
+    {
+      "epoch": 0.7650918635170604,
+      "grad_norm": 0.14940863357263953,
+      "learning_rate": 4.64539425600721e-05,
+      "loss": 0.4064,
+      "step": 583
+    },
+    {
+      "epoch": 0.7664041994750657,
+      "grad_norm": 0.12502456865628658,
+      "learning_rate": 4.643431557499167e-05,
+      "loss": 0.4007,
+      "step": 584
+    },
+    {
+      "epoch": 0.7677165354330708,
+      "grad_norm": 0.14333228236421242,
+      "learning_rate": 4.6414638593285144e-05,
+      "loss": 0.4019,
+      "step": 585
+    },
+    {
+      "epoch": 0.7690288713910761,
+      "grad_norm": 0.20068410220224983,
+      "learning_rate": 4.639491166085006e-05,
+      "loss": 0.4044,
+      "step": 586
+    },
+    {
+      "epoch": 0.7703412073490814,
+      "grad_norm": 0.16462143911858243,
+      "learning_rate": 4.6375134823700505e-05,
+      "loss": 0.4087,
+      "step": 587
+    },
+    {
+      "epoch": 0.7716535433070866,
+      "grad_norm": 0.15022837164072808,
+      "learning_rate": 4.6355308127966934e-05,
+      "loss": 0.3988,
+      "step": 588
+    },
+    {
+      "epoch": 0.7729658792650919,
+      "grad_norm": 0.15905274837245795,
+      "learning_rate": 4.633543161989613e-05,
+      "loss": 0.4061,
+      "step": 589
+    },
+    {
+      "epoch": 0.7742782152230971,
+      "grad_norm": 0.151151698795356,
+      "learning_rate": 4.631550534585105e-05,
+      "loss": 0.3962,
+      "step": 590
+    },
+    {
+      "epoch": 0.7755905511811023,
+      "grad_norm": 0.13011423710965503,
+      "learning_rate": 4.629552935231074e-05,
+      "loss": 0.398,
+      "step": 591
+    },
+    {
+      "epoch": 0.7769028871391076,
+      "grad_norm": 0.1384868471032219,
+      "learning_rate": 4.6275503685870214e-05,
+      "loss": 0.3913,
+      "step": 592
+    },
+    {
+      "epoch": 0.7782152230971129,
+      "grad_norm": 0.15356838866322492,
+      "learning_rate": 4.625542839324036e-05,
+      "loss": 0.3993,
+      "step": 593
+    },
+    {
+      "epoch": 0.7795275590551181,
+      "grad_norm": 0.13953515263529315,
+      "learning_rate": 4.6235303521247795e-05,
+      "loss": 0.4092,
+      "step": 594
+    },
+    {
+      "epoch": 0.7808398950131233,
+      "grad_norm": 0.1313591049938149,
+      "learning_rate": 4.621512911683482e-05,
+      "loss": 0.4062,
+      "step": 595
+    },
+    {
+      "epoch": 0.7821522309711286,
+      "grad_norm": 0.13242167787265785,
+      "learning_rate": 4.619490522705925e-05,
+      "loss": 0.3948,
+      "step": 596
+    },
+    {
+      "epoch": 0.7834645669291339,
+      "grad_norm": 0.12851210216145587,
+      "learning_rate": 4.6174631899094325e-05,
+      "loss": 0.4091,
+      "step": 597
+    },
+    {
+      "epoch": 0.7847769028871391,
+      "grad_norm": 0.12219656020148367,
+      "learning_rate": 4.615430918022862e-05,
+      "loss": 0.3966,
+      "step": 598
+    },
+    {
+      "epoch": 0.7860892388451444,
+      "grad_norm": 0.1469543885780674,
+      "learning_rate": 4.61339371178659e-05,
+      "loss": 0.41,
+      "step": 599
+    },
+    {
+      "epoch": 0.7874015748031497,
+      "grad_norm": 0.14806058340641817,
+      "learning_rate": 4.6113515759525024e-05,
+      "loss": 0.4057,
+      "step": 600
+    },
+    {
+      "epoch": 0.7887139107611548,
+      "grad_norm": 0.15146206599499495,
+      "learning_rate": 4.6093045152839855e-05,
+      "loss": 0.4038,
+      "step": 601
+    },
+    {
+      "epoch": 0.7900262467191601,
+      "grad_norm": 0.13859644725287792,
+      "learning_rate": 4.607252534555912e-05,
+      "loss": 0.4002,
+      "step": 602
+    },
+    {
+      "epoch": 0.7913385826771654,
+      "grad_norm": 0.14208116708274968,
+      "learning_rate": 4.605195638554629e-05,
+      "loss": 0.3986,
+      "step": 603
+    },
+    {
+      "epoch": 0.7926509186351706,
+      "grad_norm": 0.15673191965606809,
+      "learning_rate": 4.6031338320779534e-05,
+      "loss": 0.4043,
+      "step": 604
+    },
+    {
+      "epoch": 0.7939632545931758,
+      "grad_norm": 0.15491848144053638,
+      "learning_rate": 4.6010671199351505e-05,
+      "loss": 0.4017,
+      "step": 605
+    },
+    {
+      "epoch": 0.7952755905511811,
+      "grad_norm": 0.17569637253138956,
+      "learning_rate": 4.5989955069469335e-05,
+      "loss": 0.4007,
+      "step": 606
+    },
+    {
+      "epoch": 0.7965879265091863,
+      "grad_norm": 0.21734211146545088,
+      "learning_rate": 4.5969189979454416e-05,
+      "loss": 0.4041,
+      "step": 607
+    },
+    {
+      "epoch": 0.7979002624671916,
+      "grad_norm": 0.22986521497306112,
+      "learning_rate": 4.594837597774241e-05,
+      "loss": 0.4001,
+      "step": 608
+    },
+    {
+      "epoch": 0.7992125984251969,
+      "grad_norm": 0.23772424600140243,
+      "learning_rate": 4.592751311288299e-05,
+      "loss": 0.405,
+      "step": 609
+    },
+    {
+      "epoch": 0.800524934383202,
+      "grad_norm": 0.2206511565821667,
+      "learning_rate": 4.590660143353986e-05,
+      "loss": 0.3985,
+      "step": 610
+    },
+    {
+      "epoch": 0.8018372703412073,
+      "grad_norm": 0.21345096825914708,
+      "learning_rate": 4.588564098849058e-05,
+      "loss": 0.3978,
+      "step": 611
+    },
+    {
+      "epoch": 0.8031496062992126,
+      "grad_norm": 0.23381516348557552,
+      "learning_rate": 4.5864631826626446e-05,
+      "loss": 0.405,
+      "step": 612
+    },
+    {
+      "epoch": 0.8044619422572179,
+      "grad_norm": 0.22826592457436318,
+      "learning_rate": 4.584357399695239e-05,
+      "loss": 0.3958,
+      "step": 613
+    },
+    {
+      "epoch": 0.8057742782152231,
+      "grad_norm": 0.1939760354685595,
+      "learning_rate": 4.582246754858686e-05,
+      "loss": 0.3946,
+      "step": 614
+    },
+    {
+      "epoch": 0.8070866141732284,
+      "grad_norm": 0.16572417809049966,
+      "learning_rate": 4.580131253076171e-05,
+      "loss": 0.3933,
+      "step": 615
+    },
+    {
+      "epoch": 0.8083989501312336,
+      "grad_norm": 0.1569591940989404,
+      "learning_rate": 4.5780108992822104e-05,
+      "loss": 0.3944,
+      "step": 616
+    },
+    {
+      "epoch": 0.8097112860892388,
+      "grad_norm": 0.15795427666562675,
+      "learning_rate": 4.575885698422636e-05,
+      "loss": 0.3963,
+      "step": 617
+    },
+    {
+      "epoch": 0.8110236220472441,
+      "grad_norm": 0.1407443245510736,
+      "learning_rate": 4.573755655454587e-05,
+      "loss": 0.4014,
+      "step": 618
+    },
+    {
+      "epoch": 0.8123359580052494,
+      "grad_norm": 0.14083019329246735,
+      "learning_rate": 4.571620775346496e-05,
+      "loss": 0.3933,
+      "step": 619
+    },
+    {
+      "epoch": 0.8136482939632546,
+      "grad_norm": 0.16616975730610875,
+      "learning_rate": 4.569481063078078e-05,
+      "loss": 0.3929,
+      "step": 620
+    },
+    {
+      "epoch": 0.8149606299212598,
+      "grad_norm": 0.190451031480055,
+      "learning_rate": 4.567336523640322e-05,
+      "loss": 0.378,
+      "step": 621
+    },
+    {
+      "epoch": 0.8162729658792651,
+      "grad_norm": 0.21508731365965086,
+      "learning_rate": 4.5651871620354736e-05,
+      "loss": 0.3977,
+      "step": 622
+    },
+    {
+      "epoch": 0.8175853018372703,
+      "grad_norm": 0.21356406674894615,
+      "learning_rate": 4.563032983277028e-05,
+      "loss": 0.4036,
+      "step": 623
+    },
+    {
+      "epoch": 0.8188976377952756,
+      "grad_norm": 0.1727394689008456,
+      "learning_rate": 4.560873992389716e-05,
+      "loss": 0.402,
+      "step": 624
+    },
+    {
+      "epoch": 0.8202099737532809,
+      "grad_norm": 0.1426083825308613,
+      "learning_rate": 4.558710194409493e-05,
+      "loss": 0.4051,
+      "step": 625
+    },
+    {
+      "epoch": 0.821522309711286,
+      "grad_norm": 0.11046138455050303,
+      "learning_rate": 4.5565415943835285e-05,
+      "loss": 0.3956,
+      "step": 626
+    },
+    {
+      "epoch": 0.8228346456692913,
+      "grad_norm": 0.11313828876324225,
+      "learning_rate": 4.5543681973701904e-05,
+      "loss": 0.3946,
+      "step": 627
+    },
+    {
+      "epoch": 0.8241469816272966,
+      "grad_norm": 0.1326562568179498,
+      "learning_rate": 4.552190008439037e-05,
+      "loss": 0.3977,
+      "step": 628
+    },
+    {
+      "epoch": 0.8254593175853019,
+      "grad_norm": 0.16743769066497496,
+      "learning_rate": 4.5500070326708054e-05,
+      "loss": 0.3989,
+      "step": 629
+    },
+    {
+      "epoch": 0.8267716535433071,
+      "grad_norm": 0.1779071078128041,
+      "learning_rate": 4.5478192751573966e-05,
+      "loss": 0.4063,
+      "step": 630
+    },
+    {
+      "epoch": 0.8280839895013123,
+      "grad_norm": 0.16738572077414987,
+      "learning_rate": 4.545626741001866e-05,
+      "loss": 0.3965,
+      "step": 631
+    },
+    {
+      "epoch": 0.8293963254593176,
+      "grad_norm": 0.14477174242994487,
+      "learning_rate": 4.5434294353184106e-05,
+      "loss": 0.3999,
+      "step": 632
+    },
+    {
+      "epoch": 0.8307086614173228,
+      "grad_norm": 0.15331328174901218,
+      "learning_rate": 4.541227363232356e-05,
+      "loss": 0.4054,
+      "step": 633
+    },
+    {
+      "epoch": 0.8320209973753281,
+      "grad_norm": 0.14249398577581956,
+      "learning_rate": 4.539020529880148e-05,
+      "loss": 0.401,
+      "step": 634
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.13849833455521612,
+      "learning_rate": 4.536808940409337e-05,
+      "loss": 0.4091,
+      "step": 635
+    },
+    {
+      "epoch": 0.8346456692913385,
+      "grad_norm": 0.14171689530489762,
+      "learning_rate": 4.5345925999785674e-05,
+      "loss": 0.3995,
+      "step": 636
+    },
+    {
+      "epoch": 0.8359580052493438,
+      "grad_norm": 0.14027774821237535,
+      "learning_rate": 4.532371513757564e-05,
+      "loss": 0.3945,
+      "step": 637
+    },
+    {
+      "epoch": 0.8372703412073491,
+      "grad_norm": 0.1283081431880198,
+      "learning_rate": 4.5301456869271255e-05,
+      "loss": 0.3977,
+      "step": 638
+    },
+    {
+      "epoch": 0.8385826771653543,
+      "grad_norm": 0.12061453115060244,
+      "learning_rate": 4.527915124679103e-05,
+      "loss": 0.3971,
+      "step": 639
+    },
+    {
+      "epoch": 0.8398950131233596,
+      "grad_norm": 0.12639162992339384,
+      "learning_rate": 4.525679832216398e-05,
+      "loss": 0.3987,
+      "step": 640
+    },
+    {
+      "epoch": 0.8412073490813649,
+      "grad_norm": 0.14667463926733962,
+      "learning_rate": 4.523439814752942e-05,
+      "loss": 0.3971,
+      "step": 641
+    },
+    {
+      "epoch": 0.84251968503937,
+      "grad_norm": 0.12141066219754858,
+      "learning_rate": 4.521195077513689e-05,
+      "loss": 0.3954,
+      "step": 642
+    },
+    {
+      "epoch": 0.8438320209973753,
+      "grad_norm": 0.11553596235107075,
+      "learning_rate": 4.518945625734603e-05,
+      "loss": 0.3955,
+      "step": 643
+    },
+    {
+      "epoch": 0.8451443569553806,
+      "grad_norm": 0.12261053774466445,
+      "learning_rate": 4.5166914646626444e-05,
+      "loss": 0.4017,
+      "step": 644
+    },
+    {
+      "epoch": 0.8464566929133859,
+      "grad_norm": 0.15473699644617164,
+      "learning_rate": 4.514432599555757e-05,
+      "loss": 0.3948,
+      "step": 645
+    },
+    {
+      "epoch": 0.847769028871391,
+      "grad_norm": 0.1371734374161911,
+      "learning_rate": 4.512169035682859e-05,
+      "loss": 0.4035,
+      "step": 646
+    },
+    {
+      "epoch": 0.8490813648293963,
+      "grad_norm": 0.15341669182879902,
+      "learning_rate": 4.509900778323827e-05,
+      "loss": 0.3901,
+      "step": 647
+    },
+    {
+      "epoch": 0.8503937007874016,
+      "grad_norm": 0.13938291433815733,
+      "learning_rate": 4.507627832769486e-05,
+      "loss": 0.3965,
+      "step": 648
+    },
+    {
+      "epoch": 0.8517060367454068,
+      "grad_norm": 0.1343599979683778,
+      "learning_rate": 4.505350204321597e-05,
+      "loss": 0.3886,
+      "step": 649
+    },
+    {
+      "epoch": 0.8530183727034121,
+      "grad_norm": 0.18097792958323536,
+      "learning_rate": 4.5030678982928435e-05,
+      "loss": 0.3984,
+      "step": 650
+    },
+    {
+      "epoch": 0.8543307086614174,
+      "grad_norm": 0.17203497489477457,
+      "learning_rate": 4.5007809200068204e-05,
+      "loss": 0.4027,
+      "step": 651
+    },
+    {
+      "epoch": 0.8556430446194225,
+      "grad_norm": 0.18102031195113322,
+      "learning_rate": 4.498489274798019e-05,
+      "loss": 0.3973,
+      "step": 652
+    },
+    {
+      "epoch": 0.8569553805774278,
+      "grad_norm": 0.17342660387026565,
+      "learning_rate": 4.496192968011819e-05,
+      "loss": 0.3978,
+      "step": 653
+    },
+    {
+      "epoch": 0.8582677165354331,
+      "grad_norm": 0.13678661269152725,
+      "learning_rate": 4.493892005004472e-05,
+      "loss": 0.3871,
+      "step": 654
+    },
+    {
+      "epoch": 0.8595800524934383,
+      "grad_norm": 0.13189064167453163,
+      "learning_rate": 4.49158639114309e-05,
+      "loss": 0.4034,
+      "step": 655
+    },
+    {
+      "epoch": 0.8608923884514436,
+      "grad_norm": 0.15252827788466908,
+      "learning_rate": 4.489276131805635e-05,
+      "loss": 0.3963,
+      "step": 656
+    },
+    {
+      "epoch": 0.8622047244094488,
+      "grad_norm": 0.1551700275884995,
+      "learning_rate": 4.4869612323809035e-05,
+      "loss": 0.3944,
+      "step": 657
+    },
+    {
+      "epoch": 0.863517060367454,
+      "grad_norm": 0.17235777196486823,
+      "learning_rate": 4.4846416982685154e-05,
+      "loss": 0.4039,
+      "step": 658
+    },
+    {
+      "epoch": 0.8648293963254593,
+      "grad_norm": 0.16013549000555466,
+      "learning_rate": 4.482317534878901e-05,
+      "loss": 0.4004,
+      "step": 659
+    },
+    {
+      "epoch": 0.8661417322834646,
+      "grad_norm": 0.15245307609329442,
+      "learning_rate": 4.4799887476332904e-05,
+      "loss": 0.3988,
+      "step": 660
+    },
+    {
+      "epoch": 0.8674540682414699,
+      "grad_norm": 0.1455863170887436,
+      "learning_rate": 4.477655341963696e-05,
+      "loss": 0.3961,
+      "step": 661
+    },
+    {
+      "epoch": 0.868766404199475,
+      "grad_norm": 0.15047617895116375,
+      "learning_rate": 4.475317323312905e-05,
+      "loss": 0.3989,
+      "step": 662
+    },
+    {
+      "epoch": 0.8700787401574803,
+      "grad_norm": 0.12705556280228789,
+      "learning_rate": 4.472974697134466e-05,
+      "loss": 0.3973,
+      "step": 663
+    },
+    {
+      "epoch": 0.8713910761154856,
+      "grad_norm": 0.12944107678919511,
+      "learning_rate": 4.470627468892672e-05,
+      "loss": 0.395,
+      "step": 664
+    },
+    {
+      "epoch": 0.8727034120734908,
+      "grad_norm": 0.11369909341124088,
+      "learning_rate": 4.468275644062551e-05,
+      "loss": 0.3867,
+      "step": 665
+    },
+    {
+      "epoch": 0.8740157480314961,
+      "grad_norm": 0.11821869044646137,
+      "learning_rate": 4.465919228129854e-05,
+      "loss": 0.3997,
+      "step": 666
+    },
+    {
+      "epoch": 0.8753280839895013,
+      "grad_norm": 0.14082748894704317,
+      "learning_rate": 4.4635582265910416e-05,
+      "loss": 0.4079,
+      "step": 667
+    },
+    {
+      "epoch": 0.8766404199475065,
+      "grad_norm": 0.149142520878773,
+      "learning_rate": 4.4611926449532674e-05,
+      "loss": 0.4029,
+      "step": 668
+    },
+    {
+      "epoch": 0.8779527559055118,
+      "grad_norm": 0.11968224332739391,
+      "learning_rate": 4.458822488734371e-05,
+      "loss": 0.3966,
+      "step": 669
+    },
+    {
+      "epoch": 0.8792650918635171,
+      "grad_norm": 0.13664248379725827,
+      "learning_rate": 4.456447763462863e-05,
+      "loss": 0.4001,
+      "step": 670
+    },
+    {
+      "epoch": 0.8805774278215223,
+      "grad_norm": 0.13948210655042376,
+      "learning_rate": 4.4540684746779085e-05,
+      "loss": 0.3964,
+      "step": 671
+    },
+    {
+      "epoch": 0.8818897637795275,
+      "grad_norm": 0.12378798527199207,
+      "learning_rate": 4.45168462792932e-05,
+      "loss": 0.3956,
+      "step": 672
+    },
+    {
+      "epoch": 0.8832020997375328,
+      "grad_norm": 0.13803995147108036,
+      "learning_rate": 4.449296228777541e-05,
+      "loss": 0.3965,
+      "step": 673
+    },
+    {
+      "epoch": 0.884514435695538,
+      "grad_norm": 0.23753396912040864,
+      "learning_rate": 4.446903282793633e-05,
+      "loss": 0.3968,
+      "step": 674
+    },
+    {
+      "epoch": 0.8858267716535433,
+      "grad_norm": 0.2858951021935361,
+      "learning_rate": 4.4445057955592635e-05,
+      "loss": 0.395,
+      "step": 675
+    },
+    {
+      "epoch": 0.8871391076115486,
+      "grad_norm": 0.12052990345429232,
+      "learning_rate": 4.442103772666693e-05,
+      "loss": 0.3954,
+      "step": 676
+    },
+    {
+      "epoch": 0.8884514435695539,
+      "grad_norm": 0.1567804564366801,
+      "learning_rate": 4.4396972197187615e-05,
+      "loss": 0.4003,
+      "step": 677
+    },
+    {
+      "epoch": 0.889763779527559,
+      "grad_norm": 0.19000097946733707,
+      "learning_rate": 4.437286142328876e-05,
+      "loss": 0.3948,
+      "step": 678
+    },
+    {
+      "epoch": 0.8910761154855643,
+      "grad_norm": 0.21814417770694522,
+      "learning_rate": 4.434870546120995e-05,
+      "loss": 0.4028,
+      "step": 679
+    },
+    {
+      "epoch": 0.8923884514435696,
+      "grad_norm": 0.2040204325043419,
+      "learning_rate": 4.43245043672962e-05,
+      "loss": 0.3942,
+      "step": 680
+    },
+    {
+      "epoch": 0.8937007874015748,
+      "grad_norm": 0.22218776645630703,
+      "learning_rate": 4.43002581979978e-05,
+      "loss": 0.3885,
+      "step": 681
+    },
+    {
+      "epoch": 0.89501312335958,
+      "grad_norm": 0.21734349786074847,
+      "learning_rate": 4.4275967009870154e-05,
+      "loss": 0.3977,
+      "step": 682
+    },
+    {
+      "epoch": 0.8963254593175853,
+      "grad_norm": 0.22664690617398855,
+      "learning_rate": 4.425163085957368e-05,
+      "loss": 0.4056,
+      "step": 683
+    },
+    {
+      "epoch": 0.8976377952755905,
+      "grad_norm": 0.24637850143397205,
+      "learning_rate": 4.4227249803873694e-05,
+      "loss": 0.3986,
+      "step": 684
+    },
+    {
+      "epoch": 0.8989501312335958,
+      "grad_norm": 0.2342944234841161,
+      "learning_rate": 4.420282389964023e-05,
+      "loss": 0.402,
+      "step": 685
+    },
+    {
+      "epoch": 0.9002624671916011,
+      "grad_norm": 0.2766835351760319,
+      "learning_rate": 4.4178353203847954e-05,
+      "loss": 0.3944,
+      "step": 686
+    },
+    {
+      "epoch": 0.9015748031496063,
+      "grad_norm": 0.29876962382910066,
+      "learning_rate": 4.415383777357601e-05,
+      "loss": 0.4023,
+      "step": 687
+    },
+    {
+      "epoch": 0.9028871391076115,
+      "grad_norm": 0.21263898513469146,
+      "learning_rate": 4.412927766600787e-05,
+      "loss": 0.3915,
+      "step": 688
+    },
+    {
+      "epoch": 0.9041994750656168,
+      "grad_norm": 0.17783249895569256,
+      "learning_rate": 4.410467293843123e-05,
+      "loss": 0.401,
+      "step": 689
+    },
+    {
+      "epoch": 0.905511811023622,
+      "grad_norm": 0.22762938071870445,
+      "learning_rate": 4.4080023648237866e-05,
+      "loss": 0.3972,
+      "step": 690
+    },
+    {
+      "epoch": 0.9068241469816273,
+      "grad_norm": 0.2567089628440319,
+      "learning_rate": 4.405532985292349e-05,
+      "loss": 0.4081,
+      "step": 691
+    },
+    {
+      "epoch": 0.9081364829396326,
+      "grad_norm": 0.2630614601060006,
+      "learning_rate": 4.4030591610087624e-05,
+      "loss": 0.3932,
+      "step": 692
+    },
+    {
+      "epoch": 0.9094488188976378,
+      "grad_norm": 0.2309316854301495,
+      "learning_rate": 4.4005808977433486e-05,
+      "loss": 0.4033,
+      "step": 693
+    },
+    {
+      "epoch": 0.910761154855643,
+      "grad_norm": 0.17096153757825372,
+      "learning_rate": 4.398098201276782e-05,
+      "loss": 0.3861,
+      "step": 694
+    },
+    {
+      "epoch": 0.9120734908136483,
+      "grad_norm": 0.14332251374033492,
+      "learning_rate": 4.395611077400076e-05,
+      "loss": 0.4039,
+      "step": 695
+    },
+    {
+      "epoch": 0.9133858267716536,
+      "grad_norm": 0.20305688495851315,
+      "learning_rate": 4.3931195319145755e-05,
+      "loss": 0.4041,
+      "step": 696
+    },
+    {
+      "epoch": 0.9146981627296588,
+      "grad_norm": 0.22157216688725265,
+      "learning_rate": 4.390623570631934e-05,
+      "loss": 0.4041,
+      "step": 697
+    },
+    {
+      "epoch": 0.916010498687664,
+      "grad_norm": 0.1798587890958251,
+      "learning_rate": 4.38812319937411e-05,
+      "loss": 0.3928,
+      "step": 698
+    },
+    {
+      "epoch": 0.9173228346456693,
+      "grad_norm": 0.17199245757243142,
+      "learning_rate": 4.3856184239733436e-05,
+      "loss": 0.3963,
+      "step": 699
+    },
+    {
+      "epoch": 0.9186351706036745,
+      "grad_norm": 0.19478756715991516,
+      "learning_rate": 4.383109250272152e-05,
+      "loss": 0.393,
+      "step": 700
+    },
+    {
+      "epoch": 0.9199475065616798,
+      "grad_norm": 0.19749584116371233,
+      "learning_rate": 4.380595684123309e-05,
+      "loss": 0.4082,
+      "step": 701
+    },
+    {
+      "epoch": 0.9212598425196851,
+      "grad_norm": 0.19167965412700064,
+      "learning_rate": 4.378077731389835e-05,
+      "loss": 0.3946,
+      "step": 702
+    },
+    {
+      "epoch": 0.9225721784776902,
+      "grad_norm": 0.17143472755069564,
+      "learning_rate": 4.375555397944983e-05,
+      "loss": 0.4063,
+      "step": 703
+    },
+    {
+      "epoch": 0.9238845144356955,
+      "grad_norm": 0.18325072477635432,
+      "learning_rate": 4.3730286896722216e-05,
+      "loss": 0.4035,
+      "step": 704
+    },
+    {
+      "epoch": 0.9251968503937008,
+      "grad_norm": 0.17939411310251857,
+      "learning_rate": 4.370497612465228e-05,
+      "loss": 0.3903,
+      "step": 705
+    },
+    {
+      "epoch": 0.926509186351706,
+      "grad_norm": 0.19534539266762807,
+      "learning_rate": 4.367962172227866e-05,
+      "loss": 0.3987,
+      "step": 706
+    },
+    {
+      "epoch": 0.9278215223097113,
+      "grad_norm": 0.18222197882316082,
+      "learning_rate": 4.3654223748741796e-05,
+      "loss": 0.3977,
+      "step": 707
+    },
+    {
+      "epoch": 0.9291338582677166,
+      "grad_norm": 0.14846820127760232,
+      "learning_rate": 4.362878226328374e-05,
+      "loss": 0.3954,
+      "step": 708
+    },
+    {
+      "epoch": 0.9304461942257218,
+      "grad_norm": 0.13635358156527239,
+      "learning_rate": 4.360329732524805e-05,
+      "loss": 0.3924,
+      "step": 709
+    },
+    {
+      "epoch": 0.931758530183727,
+      "grad_norm": 0.12290100765390052,
+      "learning_rate": 4.357776899407963e-05,
+      "loss": 0.3966,
+      "step": 710
+    },
+    {
+      "epoch": 0.9330708661417323,
+      "grad_norm": 0.491682468878477,
+      "learning_rate": 4.3552197329324605e-05,
+      "loss": 0.4265,
+      "step": 711
+    },
+    {
+      "epoch": 0.9343832020997376,
+      "grad_norm": 0.1540325970058068,
+      "learning_rate": 4.352658239063018e-05,
+      "loss": 0.4033,
+      "step": 712
+    },
+    {
+      "epoch": 0.9356955380577427,
+      "grad_norm": 0.19087797740452758,
+      "learning_rate": 4.350092423774449e-05,
+      "loss": 0.3867,
+      "step": 713
+    },
+    {
+      "epoch": 0.937007874015748,
+      "grad_norm": 0.6544704112017427,
+      "learning_rate": 4.347522293051648e-05,
+      "loss": 0.399,
+      "step": 714
+    },
+    {
+      "epoch": 0.9383202099737533,
+      "grad_norm": 0.1999146760083894,
+      "learning_rate": 4.344947852889575e-05,
+      "loss": 0.3925,
+      "step": 715
+    },
+    {
+      "epoch": 0.9396325459317585,
+      "grad_norm": 0.3411522346062135,
+      "learning_rate": 4.342369109293242e-05,
+      "loss": 0.4,
+      "step": 716
+    },
+    {
+      "epoch": 0.9409448818897638,
+      "grad_norm": 0.1818555712289427,
+      "learning_rate": 4.3397860682777e-05,
+      "loss": 0.4024,
+      "step": 717
+    },
+    {
+      "epoch": 0.9422572178477691,
+      "grad_norm": 0.27683629901559076,
+      "learning_rate": 4.33719873586802e-05,
+      "loss": 0.4102,
+      "step": 718
+    },
+    {
+      "epoch": 0.9435695538057742,
+      "grad_norm": 0.20684767996982859,
+      "learning_rate": 4.334607118099289e-05,
+      "loss": 0.3888,
+      "step": 719
+    },
+    {
+      "epoch": 0.9448818897637795,
+      "grad_norm": 0.34520172215819267,
+      "learning_rate": 4.3320112210165854e-05,
+      "loss": 0.4059,
+      "step": 720
+    },
+    {
+      "epoch": 0.9461942257217848,
+      "grad_norm": 0.3559035152167433,
+      "learning_rate": 4.32941105067497e-05,
+      "loss": 0.3967,
+      "step": 721
+    },
+    {
+      "epoch": 0.94750656167979,
+      "grad_norm": 0.2706837532111342,
+      "learning_rate": 4.3268066131394726e-05,
+      "loss": 0.4082,
+      "step": 722
+    },
+    {
+      "epoch": 0.9488188976377953,
+      "grad_norm": 0.2311813321277985,
+      "learning_rate": 4.324197914485075e-05,
+      "loss": 0.4012,
+      "step": 723
+    },
+    {
+      "epoch": 0.9501312335958005,
+      "grad_norm": 0.3523196336872336,
+      "learning_rate": 4.321584960796699e-05,
+      "loss": 0.406,
+      "step": 724
+    },
+    {
+      "epoch": 0.9514435695538058,
+      "grad_norm": 0.25994471496050625,
+      "learning_rate": 4.318967758169192e-05,
+      "loss": 0.4049,
+      "step": 725
+    },
+    {
+      "epoch": 0.952755905511811,
+      "grad_norm": 0.20934865585421772,
+      "learning_rate": 4.316346312707311e-05,
+      "loss": 0.3997,
+      "step": 726
+    },
+    {
+      "epoch": 0.9540682414698163,
+      "grad_norm": 0.2211056134676872,
+      "learning_rate": 4.3137206305257095e-05,
+      "loss": 0.408,
+      "step": 727
+    },
+    {
+      "epoch": 0.9553805774278216,
+      "grad_norm": 0.23206342154685894,
+      "learning_rate": 4.3110907177489256e-05,
+      "loss": 0.4103,
+      "step": 728
+    },
+    {
+      "epoch": 0.9566929133858267,
+      "grad_norm": 0.17134377109858123,
+      "learning_rate": 4.308456580511363e-05,
+      "loss": 0.4013,
+      "step": 729
+    },
+    {
+      "epoch": 0.958005249343832,
+      "grad_norm": 0.17064125514566136,
+      "learning_rate": 4.305818224957281e-05,
+      "loss": 0.4043,
+      "step": 730
+    },
+    {
+      "epoch": 0.9593175853018373,
+      "grad_norm": 0.17041542032652893,
+      "learning_rate": 4.3031756572407766e-05,
+      "loss": 0.4011,
+      "step": 731
+    },
+    {
+      "epoch": 0.9606299212598425,
+      "grad_norm": 0.15891273126532682,
+      "learning_rate": 4.3005288835257726e-05,
+      "loss": 0.4062,
+      "step": 732
+    },
+    {
+      "epoch": 0.9619422572178478,
+      "grad_norm": 0.14630871467202639,
+      "learning_rate": 4.297877909986003e-05,
+      "loss": 0.3975,
+      "step": 733
+    },
+    {
+      "epoch": 0.963254593175853,
+      "grad_norm": 0.14575123796960698,
+      "learning_rate": 4.295222742804999e-05,
+      "loss": 0.3999,
+      "step": 734
+    },
+    {
+      "epoch": 0.9645669291338582,
+      "grad_norm": 0.13506344958260905,
+      "learning_rate": 4.2925633881760707e-05,
+      "loss": 0.4039,
+      "step": 735
+    },
+    {
+      "epoch": 0.9658792650918635,
+      "grad_norm": 0.13620545018167585,
+      "learning_rate": 4.289899852302298e-05,
+      "loss": 0.4034,
+      "step": 736
+    },
+    {
+      "epoch": 0.9671916010498688,
+      "grad_norm": 0.1333004567433019,
+      "learning_rate": 4.2872321413965144e-05,
+      "loss": 0.3986,
+      "step": 737
+    },
+    {
+      "epoch": 0.968503937007874,
+      "grad_norm": 0.14234862671520887,
+      "learning_rate": 4.284560261681288e-05,
+      "loss": 0.3931,
+      "step": 738
+    },
+    {
+      "epoch": 0.9698162729658792,
+      "grad_norm": 0.2679490963264685,
+      "learning_rate": 4.281884219388916e-05,
+      "loss": 0.421,
+      "step": 739
+    },
+    {
+      "epoch": 0.9711286089238845,
+      "grad_norm": 0.1263847747323097,
+      "learning_rate": 4.2792040207614005e-05,
+      "loss": 0.3929,
+      "step": 740
+    },
+    {
+      "epoch": 0.9724409448818898,
+      "grad_norm": 0.12854753583860543,
+      "learning_rate": 4.2765196720504423e-05,
+      "loss": 0.4019,
+      "step": 741
+    },
+    {
+      "epoch": 0.973753280839895,
+      "grad_norm": 0.11765040308487919,
+      "learning_rate": 4.273831179517419e-05,
+      "loss": 0.3867,
+      "step": 742
+    },
+    {
+      "epoch": 0.9750656167979003,
+      "grad_norm": 0.13376437704201996,
+      "learning_rate": 4.271138549433377e-05,
+      "loss": 0.4009,
+      "step": 743
+    },
+    {
+      "epoch": 0.9763779527559056,
+      "grad_norm": 0.13251829295123227,
+      "learning_rate": 4.268441788079012e-05,
+      "loss": 0.3957,
+      "step": 744
+    },
+    {
+      "epoch": 0.9776902887139107,
+      "grad_norm": 0.14109386792291295,
+      "learning_rate": 4.265740901744655e-05,
+      "loss": 0.3957,
+      "step": 745
+    },
+    {
+      "epoch": 0.979002624671916,
+      "grad_norm": 0.1376537228861701,
+      "learning_rate": 4.2630358967302616e-05,
+      "loss": 0.4063,
+      "step": 746
+    },
+    {
+      "epoch": 0.9803149606299213,
+      "grad_norm": 0.11186957113337022,
+      "learning_rate": 4.260326779345393e-05,
+      "loss": 0.3977,
+      "step": 747
+    },
+    {
+      "epoch": 0.9816272965879265,
+      "grad_norm": 0.10396532020021984,
+      "learning_rate": 4.257613555909201e-05,
+      "loss": 0.3939,
+      "step": 748
+    },
+    {
+      "epoch": 0.9829396325459318,
+      "grad_norm": 0.10427601519856881,
+      "learning_rate": 4.2548962327504176e-05,
+      "loss": 0.4016,
+      "step": 749
+    },
+    {
+      "epoch": 0.984251968503937,
+      "grad_norm": 0.12873312947128498,
+      "learning_rate": 4.252174816207337e-05,
+      "loss": 0.3919,
+      "step": 750
+    },
+    {
+      "epoch": 0.9855643044619422,
+      "grad_norm": 0.09126584792605437,
+      "learning_rate": 4.2494493126278015e-05,
+      "loss": 0.3954,
+      "step": 751
+    },
+    {
+      "epoch": 0.9868766404199475,
+      "grad_norm": 0.11715822786722277,
+      "learning_rate": 4.2467197283691836e-05,
+      "loss": 0.3952,
+      "step": 752
+    },
+    {
+      "epoch": 0.9881889763779528,
+      "grad_norm": 0.10276179962434762,
+      "learning_rate": 4.24398606979838e-05,
+      "loss": 0.4038,
+      "step": 753
+    },
+    {
+      "epoch": 0.989501312335958,
+      "grad_norm": 0.13198186007686902,
+      "learning_rate": 4.241248343291786e-05,
+      "loss": 0.3931,
+      "step": 754
+    },
+    {
+      "epoch": 0.9908136482939632,
+      "grad_norm": 0.11078495967334587,
+      "learning_rate": 4.238506555235287e-05,
+      "loss": 0.3945,
+      "step": 755
+    },
+    {
+      "epoch": 0.9921259842519685,
+      "grad_norm": 0.1356660258018273,
+      "learning_rate": 4.235760712024244e-05,
+      "loss": 0.4042,
+      "step": 756
+    },
+    {
+      "epoch": 0.9934383202099738,
+      "grad_norm": 0.13460163025730645,
+      "learning_rate": 4.233010820063473e-05,
+      "loss": 0.4001,
+      "step": 757
+    },
+    {
+      "epoch": 0.994750656167979,
+      "grad_norm": 0.10905562872613399,
+      "learning_rate": 4.2302568857672375e-05,
+      "loss": 0.3993,
+      "step": 758
+    },
+    {
+      "epoch": 0.9960629921259843,
+      "grad_norm": 0.10396146974399824,
+      "learning_rate": 4.2274989155592296e-05,
+      "loss": 0.395,
+      "step": 759
+    },
+    {
+      "epoch": 0.9973753280839895,
+      "grad_norm": 0.13252306455512947,
+      "learning_rate": 4.2247369158725524e-05,
+      "loss": 0.3974,
+      "step": 760
+    },
+    {
+      "epoch": 0.9986876640419947,
+      "grad_norm": 0.11940131888782154,
+      "learning_rate": 4.2219708931497115e-05,
+      "loss": 0.3988,
+      "step": 761
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.12731567415239461,
+      "learning_rate": 4.219200853842594e-05,
+      "loss": 0.3987,
+      "step": 762
+    },
+    {
+      "epoch": 1.0013123359580052,
+      "grad_norm": 0.1269144626047017,
+      "learning_rate": 4.216426804412458e-05,
+      "loss": 0.3863,
+      "step": 763
+    },
+    {
+      "epoch": 1.0026246719160106,
+      "grad_norm": 0.11366178614792022,
+      "learning_rate": 4.213648751329913e-05,
+      "loss": 0.3761,
+      "step": 764
+    },
+    {
+      "epoch": 1.0039370078740157,
+      "grad_norm": 0.1282739151124855,
+      "learning_rate": 4.210866701074908e-05,
+      "loss": 0.379,
+      "step": 765
+    },
+    {
+      "epoch": 1.005249343832021,
+      "grad_norm": 0.14117814503015133,
+      "learning_rate": 4.2080806601367184e-05,
+      "loss": 0.3845,
+      "step": 766
+    },
+    {
+      "epoch": 1.0065616797900263,
+      "grad_norm": 0.1093995131999839,
+      "learning_rate": 4.205290635013923e-05,
+      "loss": 0.3961,
+      "step": 767
+    },
+    {
+      "epoch": 1.0078740157480315,
+      "grad_norm": 0.11524996847303756,
+      "learning_rate": 4.202496632214399e-05,
+      "loss": 0.3818,
+      "step": 768
+    },
+    {
+      "epoch": 1.0091863517060367,
+      "grad_norm": 0.11144797111346402,
+      "learning_rate": 4.199698658255298e-05,
+      "loss": 0.3852,
+      "step": 769
+    },
+    {
+      "epoch": 1.010498687664042,
+      "grad_norm": 0.15390306308344284,
+      "learning_rate": 4.196896719663036e-05,
+      "loss": 0.3798,
+      "step": 770
+    },
+    {
+      "epoch": 1.0118110236220472,
+      "grad_norm": 0.11575178084356152,
+      "learning_rate": 4.1940908229732756e-05,
+      "loss": 0.379,
+      "step": 771
+    },
+    {
+      "epoch": 1.0131233595800524,
+      "grad_norm": 0.12168049538796198,
+      "learning_rate": 4.191280974730915e-05,
+      "loss": 0.3879,
+      "step": 772
+    },
+    {
+      "epoch": 1.0144356955380578,
+      "grad_norm": 0.12403463833223415,
+      "learning_rate": 4.188467181490067e-05,
+      "loss": 0.3865,
+      "step": 773
+    },
+    {
+      "epoch": 1.015748031496063,
+      "grad_norm": 0.10431006223859703,
+      "learning_rate": 4.1856494498140454e-05,
+      "loss": 0.3824,
+      "step": 774
+    },
+    {
+      "epoch": 1.0170603674540681,
+      "grad_norm": 0.1394112360275521,
+      "learning_rate": 4.182827786275353e-05,
+      "loss": 0.3826,
+      "step": 775
+    },
+    {
+      "epoch": 1.0183727034120735,
+      "grad_norm": 0.10719191735399457,
+      "learning_rate": 4.180002197455663e-05,
+      "loss": 0.3863,
+      "step": 776
+    },
+    {
+      "epoch": 1.0196850393700787,
+      "grad_norm": 0.1041813768870706,
+      "learning_rate": 4.1771726899458035e-05,
+      "loss": 0.3776,
+      "step": 777
+    },
+    {
+      "epoch": 1.020997375328084,
+      "grad_norm": 0.1469384724182806,
+      "learning_rate": 4.174339270345745e-05,
+      "loss": 0.3825,
+      "step": 778
+    },
+    {
+      "epoch": 1.0223097112860893,
+      "grad_norm": 0.14590329061691035,
+      "learning_rate": 4.171501945264581e-05,
+      "loss": 0.3847,
+      "step": 779
+    },
+    {
+      "epoch": 1.0236220472440944,
+      "grad_norm": 0.14737741283679037,
+      "learning_rate": 4.168660721320517e-05,
+      "loss": 0.3857,
+      "step": 780
+    },
+    {
+      "epoch": 1.0249343832020998,
+      "grad_norm": 0.14566408719410875,
+      "learning_rate": 4.165815605140852e-05,
+      "loss": 0.3867,
+      "step": 781
+    },
+    {
+      "epoch": 1.026246719160105,
+      "grad_norm": 0.11461086944314398,
+      "learning_rate": 4.162966603361961e-05,
+      "loss": 0.375,
+      "step": 782
+    },
+    {
+      "epoch": 1.0275590551181102,
+      "grad_norm": 0.1102882831128391,
+      "learning_rate": 4.160113722629287e-05,
+      "loss": 0.3767,
+      "step": 783
+    },
+    {
+      "epoch": 1.0288713910761156,
+      "grad_norm": 0.10091193682570963,
+      "learning_rate": 4.15725696959732e-05,
+      "loss": 0.3793,
+      "step": 784
+    },
+    {
+      "epoch": 1.0301837270341208,
+      "grad_norm": 0.11497577051641998,
+      "learning_rate": 4.154396350929578e-05,
+      "loss": 0.3764,
+      "step": 785
+    },
+    {
+      "epoch": 1.031496062992126,
+      "grad_norm": 0.9014442936934395,
+      "learning_rate": 4.1515318732986e-05,
+      "loss": 0.3847,
+      "step": 786
+    },
+    {
+      "epoch": 1.0328083989501313,
+      "grad_norm": 0.1322974831656431,
+      "learning_rate": 4.148663543385925e-05,
+      "loss": 0.3837,
+      "step": 787
+    },
+    {
+      "epoch": 1.0341207349081365,
+      "grad_norm": 0.12854417654252512,
+      "learning_rate": 4.14579136788208e-05,
+      "loss": 0.3786,
+      "step": 788
+    },
+    {
+      "epoch": 1.0354330708661417,
+      "grad_norm": 0.11350917374656776,
+      "learning_rate": 4.142915353486556e-05,
+      "loss": 0.3785,
+      "step": 789
+    },
+    {
+      "epoch": 1.036745406824147,
+      "grad_norm": 0.12613946100448467,
+      "learning_rate": 4.1400355069078045e-05,
+      "loss": 0.3799,
+      "step": 790
+    },
+    {
+      "epoch": 1.0380577427821522,
+      "grad_norm": 0.11738528626012777,
+      "learning_rate": 4.137151834863213e-05,
+      "loss": 0.3784,
+      "step": 791
+    },
+    {
+      "epoch": 1.0393700787401574,
+      "grad_norm": 0.13218860952549757,
+      "learning_rate": 4.1342643440790915e-05,
+      "loss": 0.3806,
+      "step": 792
+    },
+    {
+      "epoch": 1.0406824146981628,
+      "grad_norm": 0.13736085502986534,
+      "learning_rate": 4.1313730412906606e-05,
+      "loss": 0.3866,
+      "step": 793
+    },
+    {
+      "epoch": 1.041994750656168,
+      "grad_norm": 0.13941297945145958,
+      "learning_rate": 4.1284779332420287e-05,
+      "loss": 0.3756,
+      "step": 794
+    },
+    {
+      "epoch": 1.0433070866141732,
+      "grad_norm": 0.14740260121504903,
+      "learning_rate": 4.1255790266861814e-05,
+      "loss": 0.3753,
+      "step": 795
+    },
+    {
+      "epoch": 1.0446194225721785,
+      "grad_norm": 0.27500763727564465,
+      "learning_rate": 4.1226763283849674e-05,
+      "loss": 0.381,
+      "step": 796
+    },
+    {
+      "epoch": 1.0459317585301837,
+      "grad_norm": 0.2013889770653256,
+      "learning_rate": 4.119769845109075e-05,
+      "loss": 0.3919,
+      "step": 797
+    },
+    {
+      "epoch": 1.047244094488189,
+      "grad_norm": 0.17020327103736033,
+      "learning_rate": 4.1168595836380245e-05,
+      "loss": 0.3878,
+      "step": 798
+    },
+    {
+      "epoch": 1.0485564304461943,
+      "grad_norm": 0.14046773667732435,
+      "learning_rate": 4.113945550760148e-05,
+      "loss": 0.384,
+      "step": 799
+    },
+    {
+      "epoch": 1.0498687664041995,
+      "grad_norm": 0.13090528117445205,
+      "learning_rate": 4.111027753272575e-05,
+      "loss": 0.3792,
+      "step": 800
+    },
+    {
+      "epoch": 1.0511811023622046,
+      "grad_norm": 0.15849053852972686,
+      "learning_rate": 4.1081061979812154e-05,
+      "loss": 0.3864,
+      "step": 801
+    },
+    {
+      "epoch": 1.05249343832021,
+      "grad_norm": 0.16120167427765578,
+      "learning_rate": 4.105180891700746e-05,
+      "loss": 0.3915,
+      "step": 802
+    },
+    {
+      "epoch": 1.0538057742782152,
+      "grad_norm": 0.1809898033360453,
+      "learning_rate": 4.10225184125459e-05,
+      "loss": 0.3766,
+      "step": 803
+    },
+    {
+      "epoch": 1.0551181102362204,
+      "grad_norm": 0.21924689745935558,
+      "learning_rate": 4.099319053474906e-05,
+      "loss": 0.3771,
+      "step": 804
+    },
+    {
+      "epoch": 1.0564304461942258,
+      "grad_norm": 0.21273671553980056,
+      "learning_rate": 4.096382535202572e-05,
+      "loss": 0.3859,
+      "step": 805
+    },
+    {
+      "epoch": 1.057742782152231,
+      "grad_norm": 0.14103203403619483,
+      "learning_rate": 4.093442293287163e-05,
+      "loss": 0.3857,
+      "step": 806
+    },
+    {
+      "epoch": 1.0590551181102361,
+      "grad_norm": 0.1050572699951881,
+      "learning_rate": 4.0904983345869427e-05,
+      "loss": 0.3804,
+      "step": 807
+    },
+    {
+      "epoch": 1.0603674540682415,
+      "grad_norm": 0.14584668568587006,
+      "learning_rate": 4.0875506659688465e-05,
+      "loss": 0.3824,
+      "step": 808
+    },
+    {
+      "epoch": 1.0616797900262467,
+      "grad_norm": 0.16253726001997285,
+      "learning_rate": 4.0845992943084565e-05,
+      "loss": 0.3817,
+      "step": 809
+    },
+    {
+      "epoch": 1.0629921259842519,
+      "grad_norm": 0.1682088446262236,
+      "learning_rate": 4.08164422649e-05,
+      "loss": 0.3847,
+      "step": 810
+    },
+    {
+      "epoch": 1.0643044619422573,
+      "grad_norm": 0.15926949429818862,
+      "learning_rate": 4.0786854694063196e-05,
+      "loss": 0.3806,
+      "step": 811
+    },
+    {
+      "epoch": 1.0656167979002624,
+      "grad_norm": 0.14416695644098762,
+      "learning_rate": 4.075723029958868e-05,
+      "loss": 0.376,
+      "step": 812
+    },
+    {
+      "epoch": 1.0669291338582678,
+      "grad_norm": 0.11826970913237331,
+      "learning_rate": 4.072756915057683e-05,
+      "loss": 0.3787,
+      "step": 813
+    },
+    {
+      "epoch": 1.068241469816273,
+      "grad_norm": 0.11267730984032903,
+      "learning_rate": 4.06978713162138e-05,
+      "loss": 0.3755,
+      "step": 814
+    },
+    {
+      "epoch": 1.0695538057742782,
+      "grad_norm": 0.15945638510731194,
+      "learning_rate": 4.066813686577129e-05,
+      "loss": 0.3814,
+      "step": 815
+    },
+    {
+      "epoch": 1.0708661417322836,
+      "grad_norm": 0.13083821906466664,
+      "learning_rate": 4.063836586860639e-05,
+      "loss": 0.3806,
+      "step": 816
+    },
+    {
+      "epoch": 1.0721784776902887,
+      "grad_norm": 0.13052147975939385,
+      "learning_rate": 4.060855839416148e-05,
+      "loss": 0.3766,
+      "step": 817
+    },
+    {
+      "epoch": 1.073490813648294,
+      "grad_norm": 0.14276236748321663,
+      "learning_rate": 4.057871451196399e-05,
+      "loss": 0.3768,
+      "step": 818
+    },
+    {
+      "epoch": 1.0748031496062993,
+      "grad_norm": 0.103405145651711,
+      "learning_rate": 4.0548834291626295e-05,
+      "loss": 0.3782,
+      "step": 819
+    },
+    {
+      "epoch": 1.0761154855643045,
+      "grad_norm": 0.19275942381342648,
+      "learning_rate": 4.051891780284551e-05,
+      "loss": 0.3847,
+      "step": 820
+    },
+    {
+      "epoch": 1.0774278215223096,
+      "grad_norm": 0.16244340591478362,
+      "learning_rate": 4.048896511540337e-05,
+      "loss": 0.3899,
+      "step": 821
+    },
+    {
+      "epoch": 1.078740157480315,
+      "grad_norm": 0.11846909120402306,
+      "learning_rate": 4.045897629916603e-05,
+      "loss": 0.3882,
+      "step": 822
+    },
+    {
+      "epoch": 1.0800524934383202,
+      "grad_norm": 0.13498943037677982,
+      "learning_rate": 4.0428951424083926e-05,
+      "loss": 0.3769,
+      "step": 823
+    },
+    {
+      "epoch": 1.0813648293963254,
+      "grad_norm": 0.1769651879880087,
+      "learning_rate": 4.039889056019159e-05,
+      "loss": 0.383,
+      "step": 824
+    },
+    {
+      "epoch": 1.0826771653543308,
+      "grad_norm": 0.13848281455098185,
+      "learning_rate": 4.0368793777607524e-05,
+      "loss": 0.3831,
+      "step": 825
+    },
+    {
+      "epoch": 1.083989501312336,
+      "grad_norm": 0.10422000495616023,
+      "learning_rate": 4.033866114653398e-05,
+      "loss": 0.3788,
+      "step": 826
+    },
+    {
+      "epoch": 1.0853018372703411,
+      "grad_norm": 0.16791212029256555,
+      "learning_rate": 4.030849273725686e-05,
+      "loss": 0.3849,
+      "step": 827
+    },
+    {
+      "epoch": 1.0866141732283465,
+      "grad_norm": 0.16563472126230502,
+      "learning_rate": 4.027828862014549e-05,
+      "loss": 0.3808,
+      "step": 828
+    },
+    {
+      "epoch": 1.0879265091863517,
+      "grad_norm": 0.09343742515826432,
+      "learning_rate": 4.024804886565252e-05,
+      "loss": 0.3788,
+      "step": 829
+    },
+    {
+      "epoch": 1.0892388451443569,
+      "grad_norm": 0.14043311972134345,
+      "learning_rate": 4.02177735443137e-05,
+      "loss": 0.3865,
+      "step": 830
+    },
+    {
+      "epoch": 1.0905511811023623,
+      "grad_norm": 0.11071633680191442,
+      "learning_rate": 4.018746272674774e-05,
+      "loss": 0.3822,
+      "step": 831
+    },
+    {
+      "epoch": 1.0918635170603674,
+      "grad_norm": 0.1147731750874795,
+      "learning_rate": 4.0157116483656164e-05,
+      "loss": 0.3809,
+      "step": 832
+    },
+    {
+      "epoch": 1.0931758530183726,
+      "grad_norm": 0.11854883669693143,
+      "learning_rate": 4.012673488582312e-05,
+      "loss": 0.3812,
+      "step": 833
+    },
+    {
+      "epoch": 1.094488188976378,
+      "grad_norm": 0.115328772798458,
+      "learning_rate": 4.009631800411524e-05,
+      "loss": 0.3822,
+      "step": 834
+    },
+    {
+      "epoch": 1.0958005249343832,
+      "grad_norm": 0.12567993718937578,
+      "learning_rate": 4.0065865909481417e-05,
+      "loss": 0.3827,
+      "step": 835
+    },
+    {
+      "epoch": 1.0971128608923884,
+      "grad_norm": 1.5179641050787926,
+      "learning_rate": 4.003537867295271e-05,
+      "loss": 0.3831,
+      "step": 836
+    },
+    {
+      "epoch": 1.0984251968503937,
+      "grad_norm": 0.13289868788609308,
+      "learning_rate": 4.000485636564213e-05,
+      "loss": 0.3788,
+      "step": 837
+    },
+    {
+      "epoch": 1.099737532808399,
+      "grad_norm": 0.23323804722519567,
+      "learning_rate": 3.9974299058744525e-05,
+      "loss": 0.3792,
+      "step": 838
+    },
+    {
+      "epoch": 1.1010498687664043,
+      "grad_norm": 0.14485429197880426,
+      "learning_rate": 3.994370682353635e-05,
+      "loss": 0.3794,
+      "step": 839
+    },
+    {
+      "epoch": 1.1023622047244095,
+      "grad_norm": 0.14420783227135758,
+      "learning_rate": 3.991307973137555e-05,
+      "loss": 0.3892,
+      "step": 840
+    },
+    {
+      "epoch": 1.1036745406824147,
+      "grad_norm": 0.14100605009319553,
+      "learning_rate": 3.988241785370136e-05,
+      "loss": 0.3754,
+      "step": 841
+    },
+    {
+      "epoch": 1.10498687664042,
+      "grad_norm": 0.14173714056410816,
+      "learning_rate": 3.9851721262034156e-05,
+      "loss": 0.3834,
+      "step": 842
+    },
+    {
+      "epoch": 1.1062992125984252,
+      "grad_norm": 0.1217088309565284,
+      "learning_rate": 3.98209900279753e-05,
+      "loss": 0.3708,
+      "step": 843
+    },
+    {
+      "epoch": 1.1076115485564304,
+      "grad_norm": 0.16543203336839635,
+      "learning_rate": 3.979022422320696e-05,
+      "loss": 0.3805,
+      "step": 844
+    },
+    {
+      "epoch": 1.1089238845144358,
+      "grad_norm": 0.1825944276190469,
+      "learning_rate": 3.9759423919491914e-05,
+      "loss": 0.376,
+      "step": 845
+    },
+    {
+      "epoch": 1.110236220472441,
+      "grad_norm": 0.1445039608870268,
+      "learning_rate": 3.972858918867344e-05,
+      "loss": 0.3801,
+      "step": 846
+    },
+    {
+      "epoch": 1.1115485564304461,
+      "grad_norm": 0.15584250356454818,
+      "learning_rate": 3.9697720102675104e-05,
+      "loss": 0.3882,
+      "step": 847
+    },
+    {
+      "epoch": 1.1128608923884515,
+      "grad_norm": 0.11744610036539122,
+      "learning_rate": 3.9666816733500623e-05,
+      "loss": 0.386,
+      "step": 848
+    },
+    {
+      "epoch": 1.1141732283464567,
+      "grad_norm": 0.11311211892191116,
+      "learning_rate": 3.9635879153233654e-05,
+      "loss": 0.3867,
+      "step": 849
+    },
+    {
+      "epoch": 1.1154855643044619,
+      "grad_norm": 0.1263548217325208,
+      "learning_rate": 3.960490743403768e-05,
+      "loss": 0.3828,
+      "step": 850
+    },
+    {
+      "epoch": 1.1167979002624673,
+      "grad_norm": 0.13621685132379793,
+      "learning_rate": 3.957390164815581e-05,
+      "loss": 0.3822,
+      "step": 851
+    },
+    {
+      "epoch": 1.1181102362204725,
+      "grad_norm": 0.14062203293694947,
+      "learning_rate": 3.9542861867910585e-05,
+      "loss": 0.3771,
+      "step": 852
+    },
+    {
+      "epoch": 1.1194225721784776,
+      "grad_norm": 0.13502152612219423,
+      "learning_rate": 3.951178816570389e-05,
+      "loss": 0.3784,
+      "step": 853
+    },
+    {
+      "epoch": 1.120734908136483,
+      "grad_norm": 0.13009706688345393,
+      "learning_rate": 3.9480680614016686e-05,
+      "loss": 0.3808,
+      "step": 854
+    },
+    {
+      "epoch": 1.1220472440944882,
+      "grad_norm": 0.11598474365252126,
+      "learning_rate": 3.944953928540893e-05,
+      "loss": 0.3726,
+      "step": 855
+    },
+    {
+      "epoch": 1.1233595800524934,
+      "grad_norm": 0.16487460883781624,
+      "learning_rate": 3.9418364252519336e-05,
+      "loss": 0.38,
+      "step": 856
+    },
+    {
+      "epoch": 1.1246719160104988,
+      "grad_norm": 0.14899500441374636,
+      "learning_rate": 3.938715558806525e-05,
+      "loss": 0.3773,
+      "step": 857
+    },
+    {
+      "epoch": 1.125984251968504,
+      "grad_norm": 0.14024384789875913,
+      "learning_rate": 3.935591336484247e-05,
+      "loss": 0.371,
+      "step": 858
+    },
+    {
+      "epoch": 1.127296587926509,
+      "grad_norm": 0.11327010831140966,
+      "learning_rate": 3.9324637655725055e-05,
+      "loss": 0.3775,
+      "step": 859
+    },
+    {
+      "epoch": 1.1286089238845145,
+      "grad_norm": 0.11844849663559481,
+      "learning_rate": 3.9293328533665175e-05,
+      "loss": 0.373,
+      "step": 860
+    },
+    {
+      "epoch": 1.1299212598425197,
+      "grad_norm": 0.14838636835330302,
+      "learning_rate": 3.926198607169295e-05,
+      "loss": 0.379,
+      "step": 861
+    },
+    {
+      "epoch": 1.1312335958005248,
+      "grad_norm": 0.14875044958174075,
+      "learning_rate": 3.923061034291626e-05,
+      "loss": 0.3788,
+      "step": 862
+    },
+    {
+      "epoch": 1.1325459317585302,
+      "grad_norm": 0.125451942449121,
+      "learning_rate": 3.919920142052057e-05,
+      "loss": 0.3796,
+      "step": 863
+    },
+    {
+      "epoch": 1.1338582677165354,
+      "grad_norm": 0.14235808695553756,
+      "learning_rate": 3.916775937776879e-05,
+      "loss": 0.373,
+      "step": 864
+    },
+    {
+      "epoch": 1.1351706036745406,
+      "grad_norm": 0.12064503110691056,
+      "learning_rate": 3.913628428800108e-05,
+      "loss": 0.3734,
+      "step": 865
+    },
+    {
+      "epoch": 1.136482939632546,
+      "grad_norm": 0.11190121806421426,
+      "learning_rate": 3.910477622463466e-05,
+      "loss": 0.3787,
+      "step": 866
+    },
+    {
+      "epoch": 1.1377952755905512,
+      "grad_norm": 0.11161302045982086,
+      "learning_rate": 3.907323526116372e-05,
+      "loss": 0.3858,
+      "step": 867
+    },
+    {
+      "epoch": 1.1391076115485563,
+      "grad_norm": 0.1364551881992253,
+      "learning_rate": 3.904166147115912e-05,
+      "loss": 0.3794,
+      "step": 868
+    },
+    {
+      "epoch": 1.1404199475065617,
+      "grad_norm": 0.12330113367303193,
+      "learning_rate": 3.9010054928268335e-05,
+      "loss": 0.3782,
+      "step": 869
+    },
+    {
+      "epoch": 1.141732283464567,
+      "grad_norm": 0.11328436921539649,
+      "learning_rate": 3.8978415706215225e-05,
+      "loss": 0.3752,
+      "step": 870
+    },
+    {
+      "epoch": 1.143044619422572,
+      "grad_norm": 0.12886406898433855,
+      "learning_rate": 3.894674387879988e-05,
+      "loss": 0.376,
+      "step": 871
+    },
+    {
+      "epoch": 1.1443569553805775,
+      "grad_norm": 0.1135574986692543,
+      "learning_rate": 3.8915039519898444e-05,
+      "loss": 0.3751,
+      "step": 872
+    },
+    {
+      "epoch": 1.1456692913385826,
+      "grad_norm": 0.12044750910275365,
+      "learning_rate": 3.888330270346293e-05,
+      "loss": 0.3813,
+      "step": 873
+    },
+    {
+      "epoch": 1.1469816272965878,
+      "grad_norm": 0.1197242996358705,
+      "learning_rate": 3.8851533503521076e-05,
+      "loss": 0.3731,
+      "step": 874
+    },
+    {
+      "epoch": 1.1482939632545932,
+      "grad_norm": 0.11293749242851599,
+      "learning_rate": 3.8819731994176145e-05,
+      "loss": 0.3819,
+      "step": 875
+    },
+    {
+      "epoch": 1.1496062992125984,
+      "grad_norm": 0.1001039781535333,
+      "learning_rate": 3.878789824960677e-05,
+      "loss": 0.3801,
+      "step": 876
+    },
+    {
+      "epoch": 1.1509186351706036,
+      "grad_norm": 0.11530706158726432,
+      "learning_rate": 3.875603234406677e-05,
+      "loss": 0.3841,
+      "step": 877
+    },
+    {
+      "epoch": 1.152230971128609,
+      "grad_norm": 0.1176163646037514,
+      "learning_rate": 3.8724134351884975e-05,
+      "loss": 0.3783,
+      "step": 878
+    },
+    {
+      "epoch": 1.1535433070866141,
+      "grad_norm": 0.11418626402091171,
+      "learning_rate": 3.869220434746509e-05,
+      "loss": 0.3756,
+      "step": 879
+    },
+    {
+      "epoch": 1.1548556430446195,
+      "grad_norm": 0.12577225794704405,
+      "learning_rate": 3.866024240528544e-05,
+      "loss": 0.3807,
+      "step": 880
+    },
+    {
+      "epoch": 1.1561679790026247,
+      "grad_norm": 0.14480374129239335,
+      "learning_rate": 3.862824859989888e-05,
+      "loss": 0.3817,
+      "step": 881
+    },
+    {
+      "epoch": 1.1574803149606299,
+      "grad_norm": 0.10325857962461234,
+      "learning_rate": 3.8596223005932596e-05,
+      "loss": 0.3704,
+      "step": 882
+    },
+    {
+      "epoch": 1.1587926509186353,
+      "grad_norm": 0.10894921111421742,
+      "learning_rate": 3.856416569808789e-05,
+      "loss": 0.3716,
+      "step": 883
+    },
+    {
+      "epoch": 1.1601049868766404,
+      "grad_norm": 0.12031451920344342,
+      "learning_rate": 3.853207675114006e-05,
+      "loss": 0.3716,
+      "step": 884
+    },
+    {
+      "epoch": 1.1614173228346456,
+      "grad_norm": 0.11257312187440273,
+      "learning_rate": 3.84999562399382e-05,
+      "loss": 0.3732,
+      "step": 885
+    },
+    {
+      "epoch": 1.162729658792651,
+      "grad_norm": 0.13030518395087778,
+      "learning_rate": 3.8467804239405035e-05,
+      "loss": 0.3756,
+      "step": 886
+    },
+    {
+      "epoch": 1.1640419947506562,
+      "grad_norm": 0.12374277010064194,
+      "learning_rate": 3.843562082453673e-05,
+      "loss": 0.3737,
+      "step": 887
+    },
+    {
+      "epoch": 1.1653543307086613,
+      "grad_norm": 0.09289290210090062,
+      "learning_rate": 3.840340607040272e-05,
+      "loss": 0.3728,
+      "step": 888
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.09659410132009372,
+      "learning_rate": 3.837116005214556e-05,
+      "loss": 0.3691,
+      "step": 889
+    },
+    {
+      "epoch": 1.167979002624672,
+      "grad_norm": 0.10168642344086551,
+      "learning_rate": 3.833888284498071e-05,
+      "loss": 0.3694,
+      "step": 890
+    },
+    {
+      "epoch": 1.169291338582677,
+      "grad_norm": 0.11037038994641724,
+      "learning_rate": 3.83065745241964e-05,
+      "loss": 0.3811,
+      "step": 891
+    },
+    {
+      "epoch": 1.1706036745406825,
+      "grad_norm": 0.11931553172358055,
+      "learning_rate": 3.827423516515343e-05,
+      "loss": 0.3726,
+      "step": 892
+    },
+    {
+      "epoch": 1.1719160104986877,
+      "grad_norm": 0.10809054828273294,
+      "learning_rate": 3.824186484328497e-05,
+      "loss": 0.3771,
+      "step": 893
+    },
+    {
+      "epoch": 1.1732283464566928,
+      "grad_norm": 0.1123185864332617,
+      "learning_rate": 3.8209463634096446e-05,
+      "loss": 0.3729,
+      "step": 894
+    },
+    {
+      "epoch": 1.1745406824146982,
+      "grad_norm": 0.1116557157769846,
+      "learning_rate": 3.817703161316532e-05,
+      "loss": 0.3855,
+      "step": 895
+    },
+    {
+      "epoch": 1.1758530183727034,
+      "grad_norm": 0.10960674130027341,
+      "learning_rate": 3.814456885614093e-05,
+      "loss": 0.38,
+      "step": 896
+    },
+    {
+      "epoch": 1.1771653543307086,
+      "grad_norm": 0.09292029845598376,
+      "learning_rate": 3.811207543874428e-05,
+      "loss": 0.3707,
+      "step": 897
+    },
+    {
+      "epoch": 1.178477690288714,
+      "grad_norm": 0.1129803712258862,
+      "learning_rate": 3.807955143676793e-05,
+      "loss": 0.3839,
+      "step": 898
+    },
+    {
+      "epoch": 1.1797900262467191,
+      "grad_norm": 0.13605484077385574,
+      "learning_rate": 3.804699692607575e-05,
+      "loss": 0.3773,
+      "step": 899
+    },
+    {
+      "epoch": 1.1811023622047245,
+      "grad_norm": 0.128555898138646,
+      "learning_rate": 3.8014411982602796e-05,
+      "loss": 0.3762,
+      "step": 900
+    },
+    {
+      "epoch": 1.1824146981627297,
+      "grad_norm": 0.11794283843382125,
+      "learning_rate": 3.798179668235508e-05,
+      "loss": 0.3766,
+      "step": 901
+    },
+    {
+      "epoch": 1.1837270341207349,
+      "grad_norm": 0.13054229691785532,
+      "learning_rate": 3.7949151101409446e-05,
+      "loss": 0.3773,
+      "step": 902
+    },
+    {
+      "epoch": 1.1850393700787403,
+      "grad_norm": 0.14574371982657017,
+      "learning_rate": 3.791647531591338e-05,
+      "loss": 0.3741,
+      "step": 903
+    },
+    {
+      "epoch": 1.1863517060367454,
+      "grad_norm": 0.15761596223167829,
+      "learning_rate": 3.788376940208479e-05,
+      "loss": 0.3782,
+      "step": 904
+    },
+    {
+      "epoch": 1.1876640419947506,
+      "grad_norm": 0.13447152488293831,
+      "learning_rate": 3.785103343621188e-05,
+      "loss": 0.3788,
+      "step": 905
+    },
+    {
+      "epoch": 1.188976377952756,
+      "grad_norm": 0.10102402545192958,
+      "learning_rate": 3.781826749465295e-05,
+      "loss": 0.3776,
+      "step": 906
+    },
+    {
+      "epoch": 1.1902887139107612,
+      "grad_norm": 0.1158939751431388,
+      "learning_rate": 3.778547165383621e-05,
+      "loss": 0.3802,
+      "step": 907
+    },
+    {
+      "epoch": 1.1916010498687664,
+      "grad_norm": 0.12553267829727421,
+      "learning_rate": 3.775264599025963e-05,
+      "loss": 0.3706,
+      "step": 908
+    },
+    {
+      "epoch": 1.1929133858267718,
+      "grad_norm": 0.1735473080167772,
+      "learning_rate": 3.771979058049072e-05,
+      "loss": 0.3834,
+      "step": 909
+    },
+    {
+      "epoch": 1.194225721784777,
+      "grad_norm": 0.14641134482154497,
+      "learning_rate": 3.768690550116639e-05,
+      "loss": 0.3773,
+      "step": 910
+    },
+    {
+      "epoch": 1.195538057742782,
+      "grad_norm": 0.10301602512221351,
+      "learning_rate": 3.765399082899277e-05,
+      "loss": 0.3739,
+      "step": 911
+    },
+    {
+      "epoch": 1.1968503937007875,
+      "grad_norm": 0.11279527169867605,
+      "learning_rate": 3.7621046640744975e-05,
+      "loss": 0.3753,
+      "step": 912
+    },
+    {
+      "epoch": 1.1981627296587927,
+      "grad_norm": 0.10257162453635466,
+      "learning_rate": 3.758807301326699e-05,
+      "loss": 0.3805,
+      "step": 913
+    },
+    {
+      "epoch": 1.1994750656167978,
+      "grad_norm": 0.12745893524312207,
+      "learning_rate": 3.7555070023471486e-05,
+      "loss": 0.3807,
+      "step": 914
+    },
+    {
+      "epoch": 1.2007874015748032,
+      "grad_norm": 0.09659952284042561,
+      "learning_rate": 3.75220377483396e-05,
+      "loss": 0.3786,
+      "step": 915
+    },
+    {
+      "epoch": 1.2020997375328084,
+      "grad_norm": 0.10917644155598942,
+      "learning_rate": 3.74889762649208e-05,
+      "loss": 0.3807,
+      "step": 916
+    },
+    {
+      "epoch": 1.2034120734908136,
+      "grad_norm": 0.12091253194684407,
+      "learning_rate": 3.7455885650332646e-05,
+      "loss": 0.3732,
+      "step": 917
+    },
+    {
+      "epoch": 1.204724409448819,
+      "grad_norm": 0.11313893589233741,
+      "learning_rate": 3.742276598176069e-05,
+      "loss": 0.3878,
+      "step": 918
+    },
+    {
+      "epoch": 1.2060367454068242,
+      "grad_norm": 0.10975686683834837,
+      "learning_rate": 3.738961733645822e-05,
+      "loss": 0.3838,
+      "step": 919
+    },
+    {
+      "epoch": 1.2073490813648293,
+      "grad_norm": 0.12523817564351208,
+      "learning_rate": 3.7356439791746145e-05,
+      "loss": 0.3796,
+      "step": 920
+    },
+    {
+      "epoch": 1.2086614173228347,
+      "grad_norm": 0.12086100642115866,
+      "learning_rate": 3.732323342501275e-05,
+      "loss": 0.3749,
+      "step": 921
+    },
+    {
+      "epoch": 1.20997375328084,
+      "grad_norm": 0.10054773789300171,
+      "learning_rate": 3.7289998313713585e-05,
+      "loss": 0.3709,
+      "step": 922
+    },
+    {
+      "epoch": 1.211286089238845,
+      "grad_norm": 0.10859323890462756,
+      "learning_rate": 3.7256734535371225e-05,
+      "loss": 0.383,
+      "step": 923
+    },
+    {
+      "epoch": 1.2125984251968505,
+      "grad_norm": 0.12989219987447922,
+      "learning_rate": 3.722344216757511e-05,
+      "loss": 0.373,
+      "step": 924
+    },
+    {
+      "epoch": 1.2139107611548556,
+      "grad_norm": 0.12360028618695715,
+      "learning_rate": 3.719012128798137e-05,
+      "loss": 0.3795,
+      "step": 925
+    },
+    {
+      "epoch": 1.2152230971128608,
+      "grad_norm": 0.10791022231197318,
+      "learning_rate": 3.7156771974312655e-05,
+      "loss": 0.3753,
+      "step": 926
+    },
+    {
+      "epoch": 1.2165354330708662,
+      "grad_norm": 0.1366461139857178,
+      "learning_rate": 3.712339430435792e-05,
+      "loss": 0.3798,
+      "step": 927
+    },
+    {
+      "epoch": 1.2178477690288714,
+      "grad_norm": 0.12539555293526783,
+      "learning_rate": 3.708998835597227e-05,
+      "loss": 0.3847,
+      "step": 928
+    },
+    {
+      "epoch": 1.2191601049868765,
+      "grad_norm": 0.1124172583147763,
+      "learning_rate": 3.705655420707677e-05,
+      "loss": 0.38,
+      "step": 929
+    },
+    {
+      "epoch": 1.220472440944882,
+      "grad_norm": 0.09662508666320244,
+      "learning_rate": 3.702309193565827e-05,
+      "loss": 0.3831,
+      "step": 930
+    },
+    {
+      "epoch": 1.2217847769028871,
+      "grad_norm": 0.11617719186919961,
+      "learning_rate": 3.698960161976921e-05,
+      "loss": 0.374,
+      "step": 931
+    },
+    {
+      "epoch": 1.2230971128608923,
+      "grad_norm": 0.11882439256433475,
+      "learning_rate": 3.695608333752745e-05,
+      "loss": 0.3704,
+      "step": 932
+    },
+    {
+      "epoch": 1.2244094488188977,
+      "grad_norm": 0.09548581456724556,
+      "learning_rate": 3.692253716711607e-05,
+      "loss": 0.3779,
+      "step": 933
+    },
+    {
+      "epoch": 1.2257217847769029,
+      "grad_norm": 0.11906747342618301,
+      "learning_rate": 3.6888963186783224e-05,
+      "loss": 0.3869,
+      "step": 934
+    },
+    {
+      "epoch": 1.227034120734908,
+      "grad_norm": 0.13003752794566134,
+      "learning_rate": 3.68553614748419e-05,
+      "loss": 0.3787,
+      "step": 935
+    },
+    {
+      "epoch": 1.2283464566929134,
+      "grad_norm": 0.10731668162845412,
+      "learning_rate": 3.6821732109669816e-05,
+      "loss": 0.3837,
+      "step": 936
+    },
+    {
+      "epoch": 1.2296587926509186,
+      "grad_norm": 0.13050484684092575,
+      "learning_rate": 3.678807516970915e-05,
+      "loss": 0.381,
+      "step": 937
+    },
+    {
+      "epoch": 1.2309711286089238,
+      "grad_norm": 0.16917683009815182,
+      "learning_rate": 3.675439073346643e-05,
+      "loss": 0.3767,
+      "step": 938
+    },
+    {
+      "epoch": 1.2322834645669292,
+      "grad_norm": 0.14885830135625253,
+      "learning_rate": 3.672067887951231e-05,
+      "loss": 0.373,
+      "step": 939
+    },
+    {
+      "epoch": 1.2335958005249343,
+      "grad_norm": 0.09657945742539464,
+      "learning_rate": 3.6686939686481385e-05,
+      "loss": 0.3798,
+      "step": 940
+    },
+    {
+      "epoch": 1.2349081364829395,
+      "grad_norm": 0.12773651540876566,
+      "learning_rate": 3.665317323307204e-05,
+      "loss": 0.379,
+      "step": 941
+    },
+    {
+      "epoch": 1.236220472440945,
+      "grad_norm": 0.12962903420737892,
+      "learning_rate": 3.661937959804623e-05,
+      "loss": 0.3828,
+      "step": 942
+    },
+    {
+      "epoch": 1.23753280839895,
+      "grad_norm": 0.10854446229312453,
+      "learning_rate": 3.6585558860229323e-05,
+      "loss": 0.3838,
+      "step": 943
+    },
+    {
+      "epoch": 1.2388451443569555,
+      "grad_norm": 0.13113703101266241,
+      "learning_rate": 3.6551711098509906e-05,
+      "loss": 0.3698,
+      "step": 944
+    },
+    {
+      "epoch": 1.2401574803149606,
+      "grad_norm": 0.12724068713666717,
+      "learning_rate": 3.6517836391839586e-05,
+      "loss": 0.376,
+      "step": 945
+    },
+    {
+      "epoch": 1.2414698162729658,
+      "grad_norm": 0.13482644660393756,
+      "learning_rate": 3.648393481923284e-05,
+      "loss": 0.3777,
+      "step": 946
+    },
+    {
+      "epoch": 1.2427821522309712,
+      "grad_norm": 0.13652191200455815,
+      "learning_rate": 3.645000645976681e-05,
+      "loss": 0.3836,
+      "step": 947
+    },
+    {
+      "epoch": 1.2440944881889764,
+      "grad_norm": 0.11463910317124391,
+      "learning_rate": 3.641605139258109e-05,
+      "loss": 0.3766,
+      "step": 948
+    },
+    {
+      "epoch": 1.2454068241469816,
+      "grad_norm": 0.12532110590761025,
+      "learning_rate": 3.6382069696877594e-05,
+      "loss": 0.3766,
+      "step": 949
+    },
+    {
+      "epoch": 1.246719160104987,
+      "grad_norm": 0.13202135901604484,
+      "learning_rate": 3.634806145192038e-05,
+      "loss": 0.3762,
+      "step": 950
+    },
+    {
+      "epoch": 1.2480314960629921,
+      "grad_norm": 0.12828040412830571,
+      "learning_rate": 3.631402673703537e-05,
+      "loss": 0.3744,
+      "step": 951
+    },
+    {
+      "epoch": 1.2493438320209973,
+      "grad_norm": 0.11870672686361022,
+      "learning_rate": 3.6279965631610266e-05,
+      "loss": 0.3707,
+      "step": 952
+    },
+    {
+      "epoch": 1.2506561679790027,
+      "grad_norm": 0.12622702669042504,
+      "learning_rate": 3.6245878215094325e-05,
+      "loss": 0.3806,
+      "step": 953
+    },
+    {
+      "epoch": 1.2519685039370079,
+      "grad_norm": 0.10018405272267897,
+      "learning_rate": 3.6211764566998176e-05,
+      "loss": 0.3788,
+      "step": 954
+    },
+    {
+      "epoch": 1.253280839895013,
+      "grad_norm": 0.0990762544683016,
+      "learning_rate": 3.617762476689361e-05,
+      "loss": 0.3869,
+      "step": 955
+    },
+    {
+      "epoch": 1.2545931758530184,
+      "grad_norm": 0.10951365019839676,
+      "learning_rate": 3.6143458894413465e-05,
+      "loss": 0.3804,
+      "step": 956
+    },
+    {
+      "epoch": 1.2559055118110236,
+      "grad_norm": 0.09213215068017436,
+      "learning_rate": 3.610926702925134e-05,
+      "loss": 0.3885,
+      "step": 957
+    },
+    {
+      "epoch": 1.257217847769029,
+      "grad_norm": 0.11069256244073072,
+      "learning_rate": 3.607504925116152e-05,
+      "loss": 0.3735,
+      "step": 958
+    },
+    {
+      "epoch": 1.2585301837270342,
+      "grad_norm": 0.11237642352265312,
+      "learning_rate": 3.604080563995869e-05,
+      "loss": 0.3758,
+      "step": 959
+    },
+    {
+      "epoch": 1.2598425196850394,
+      "grad_norm": 0.09205804780868952,
+      "learning_rate": 3.60065362755178e-05,
+      "loss": 0.3737,
+      "step": 960
+    },
+    {
+      "epoch": 1.2611548556430447,
+      "grad_norm": 0.09965022333756572,
+      "learning_rate": 3.59722412377739e-05,
+      "loss": 0.3759,
+      "step": 961
+    },
+    {
+      "epoch": 1.26246719160105,
+      "grad_norm": 0.09294206500270538,
+      "learning_rate": 3.5937920606721875e-05,
+      "loss": 0.3929,
+      "step": 962
+    },
+    {
+      "epoch": 1.263779527559055,
+      "grad_norm": 0.12611593567988108,
+      "learning_rate": 3.590357446241635e-05,
+      "loss": 0.3725,
+      "step": 963
+    },
+    {
+      "epoch": 1.2650918635170605,
+      "grad_norm": 0.1458511307894957,
+      "learning_rate": 3.586920288497143e-05,
+      "loss": 0.3774,
+      "step": 964
+    },
+    {
+      "epoch": 1.2664041994750657,
+      "grad_norm": 0.11841865443076866,
+      "learning_rate": 3.583480595456059e-05,
+      "loss": 0.38,
+      "step": 965
+    },
+    {
+      "epoch": 1.2677165354330708,
+      "grad_norm": 0.14680457995172153,
+      "learning_rate": 3.580038375141638e-05,
+      "loss": 0.375,
+      "step": 966
+    },
+    {
+      "epoch": 1.2690288713910762,
+      "grad_norm": 0.09797752183126215,
+      "learning_rate": 3.5765936355830356e-05,
+      "loss": 0.3796,
+      "step": 967
+    },
+    {
+      "epoch": 1.2703412073490814,
+      "grad_norm": 0.0970780533962701,
+      "learning_rate": 3.573146384815278e-05,
+      "loss": 0.3781,
+      "step": 968
+    },
+    {
+      "epoch": 1.2716535433070866,
+      "grad_norm": 0.09414535294224183,
+      "learning_rate": 3.569696630879256e-05,
+      "loss": 0.3758,
+      "step": 969
+    },
+    {
+      "epoch": 1.272965879265092,
+      "grad_norm": 0.10985708737264833,
+      "learning_rate": 3.566244381821692e-05,
+      "loss": 0.3717,
+      "step": 970
+    },
+    {
+      "epoch": 1.2742782152230971,
+      "grad_norm": 0.11249179697430559,
+      "learning_rate": 3.562789645695133e-05,
+      "loss": 0.3808,
+      "step": 971
+    },
+    {
+      "epoch": 1.2755905511811023,
+      "grad_norm": 0.09829372380406715,
+      "learning_rate": 3.559332430557926e-05,
+      "loss": 0.3886,
+      "step": 972
+    },
+    {
+      "epoch": 1.2769028871391077,
+      "grad_norm": 0.09796206483662481,
+      "learning_rate": 3.5558727444742004e-05,
+      "loss": 0.3762,
+      "step": 973
+    },
+    {
+      "epoch": 1.2782152230971129,
+      "grad_norm": 0.09172536562009581,
+      "learning_rate": 3.5524105955138496e-05,
+      "loss": 0.3772,
+      "step": 974
+    },
+    {
+      "epoch": 1.279527559055118,
+      "grad_norm": 0.09965218099063002,
+      "learning_rate": 3.5489459917525103e-05,
+      "loss": 0.3851,
+      "step": 975
+    },
+    {
+      "epoch": 1.2808398950131235,
+      "grad_norm": 0.1019649685708242,
+      "learning_rate": 3.545478941271546e-05,
+      "loss": 0.3862,
+      "step": 976
+    },
+    {
+      "epoch": 1.2821522309711286,
+      "grad_norm": 0.09544746821874413,
+      "learning_rate": 3.5420094521580283e-05,
+      "loss": 0.3764,
+      "step": 977
+    },
+    {
+      "epoch": 1.2834645669291338,
+      "grad_norm": 0.09510678703083722,
+      "learning_rate": 3.5385375325047166e-05,
+      "loss": 0.3755,
+      "step": 978
+    },
+    {
+      "epoch": 1.2847769028871392,
+      "grad_norm": 0.10163809712132466,
+      "learning_rate": 3.535063190410038e-05,
+      "loss": 0.3791,
+      "step": 979
+    },
+    {
+      "epoch": 1.2860892388451444,
+      "grad_norm": 0.10553872030351516,
+      "learning_rate": 3.5315864339780714e-05,
+      "loss": 0.3683,
+      "step": 980
+    },
+    {
+      "epoch": 1.2874015748031495,
+      "grad_norm": 0.08937057574986522,
+      "learning_rate": 3.5281072713185275e-05,
+      "loss": 0.3766,
+      "step": 981
+    },
+    {
+      "epoch": 1.288713910761155,
+      "grad_norm": 0.09669394133995492,
+      "learning_rate": 3.52462571054673e-05,
+      "loss": 0.381,
+      "step": 982
+    },
+    {
+      "epoch": 1.29002624671916,
+      "grad_norm": 0.1039453142536142,
+      "learning_rate": 3.521141759783595e-05,
+      "loss": 0.3686,
+      "step": 983
+    },
+    {
+      "epoch": 1.2913385826771653,
+      "grad_norm": 0.11558078495163236,
+      "learning_rate": 3.517655427155613e-05,
+      "loss": 0.3786,
+      "step": 984
+    },
+    {
+      "epoch": 1.2926509186351707,
+      "grad_norm": 0.0880731749810649,
+      "learning_rate": 3.5141667207948336e-05,
+      "loss": 0.3697,
+      "step": 985
+    },
+    {
+      "epoch": 1.2939632545931758,
+      "grad_norm": 0.11466425071165062,
+      "learning_rate": 3.5106756488388396e-05,
+      "loss": 0.3703,
+      "step": 986
+    },
+    {
+      "epoch": 1.295275590551181,
+      "grad_norm": 0.09778248221216157,
+      "learning_rate": 3.507182219430734e-05,
+      "loss": 0.3733,
+      "step": 987
+    },
+    {
+      "epoch": 1.2965879265091864,
+      "grad_norm": 0.1021742776626833,
+      "learning_rate": 3.503686440719115e-05,
+      "loss": 0.3737,
+      "step": 988
+    },
+    {
+      "epoch": 1.2979002624671916,
+      "grad_norm": 0.15439055996675835,
+      "learning_rate": 3.5001883208580665e-05,
+      "loss": 0.3762,
+      "step": 989
+    },
+    {
+      "epoch": 1.2992125984251968,
+      "grad_norm": 0.096555155926866,
+      "learning_rate": 3.4966878680071293e-05,
+      "loss": 0.3774,
+      "step": 990
+    },
+    {
+      "epoch": 1.3005249343832022,
+      "grad_norm": 0.10415107914007558,
+      "learning_rate": 3.4931850903312866e-05,
+      "loss": 0.3768,
+      "step": 991
+    },
+    {
+      "epoch": 1.3018372703412073,
+      "grad_norm": 0.13560605148967278,
+      "learning_rate": 3.489679996000944e-05,
+      "loss": 0.3733,
+      "step": 992
+    },
+    {
+      "epoch": 1.3031496062992125,
+      "grad_norm": 0.11174531408139572,
+      "learning_rate": 3.486172593191913e-05,
+      "loss": 0.3771,
+      "step": 993
+    },
+    {
+      "epoch": 1.304461942257218,
+      "grad_norm": 0.10239048594698844,
+      "learning_rate": 3.482662890085386e-05,
+      "loss": 0.3746,
+      "step": 994
+    },
+    {
+      "epoch": 1.305774278215223,
+      "grad_norm": 0.11864125880435909,
+      "learning_rate": 3.479150894867926e-05,
+      "loss": 0.391,
+      "step": 995
+    },
+    {
+      "epoch": 1.3070866141732282,
+      "grad_norm": 0.09223984983912663,
+      "learning_rate": 3.475636615731437e-05,
+      "loss": 0.3817,
+      "step": 996
+    },
+    {
+      "epoch": 1.3083989501312336,
+      "grad_norm": 0.10240073944851918,
+      "learning_rate": 3.472120060873154e-05,
+      "loss": 0.3864,
+      "step": 997
+    },
+    {
+      "epoch": 1.3097112860892388,
+      "grad_norm": 0.1054981588761247,
+      "learning_rate": 3.46860123849562e-05,
+      "loss": 0.3725,
+      "step": 998
+    },
+    {
+      "epoch": 1.311023622047244,
+      "grad_norm": 0.09104185291302025,
+      "learning_rate": 3.465080156806666e-05,
+      "loss": 0.3784,
+      "step": 999
+    },
+    {
+      "epoch": 1.3123359580052494,
+      "grad_norm": 0.09234550039014487,
+      "learning_rate": 3.461556824019392e-05,
+      "loss": 0.3787,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3136482939632546,
+      "grad_norm": 0.09560890720629979,
+      "learning_rate": 3.458031248352152e-05,
+      "loss": 0.3732,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3149606299212597,
+      "grad_norm": 0.7324439984455627,
+      "learning_rate": 3.454503438028528e-05,
+      "loss": 0.3925,
+      "step": 1002
+    },
+    {
+      "epoch": 1.3162729658792651,
+      "grad_norm": 0.09474926527346442,
+      "learning_rate": 3.450973401277317e-05,
+      "loss": 0.3767,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3175853018372703,
+      "grad_norm": 0.10355272439698986,
+      "learning_rate": 3.447441146332507e-05,
+      "loss": 0.3803,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3188976377952755,
+      "grad_norm": 0.10149233205002853,
+      "learning_rate": 3.4439066814332645e-05,
+      "loss": 0.3736,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3202099737532809,
+      "grad_norm": 0.10339343323585012,
+      "learning_rate": 3.440370014823906e-05,
+      "loss": 0.3959,
+      "step": 1006
+    },
+    {
+      "epoch": 1.321522309711286,
+      "grad_norm": 0.1781728328709455,
+      "learning_rate": 3.4368311547538846e-05,
+      "loss": 0.3873,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3228346456692912,
+      "grad_norm": 0.10323008381306104,
+      "learning_rate": 3.4332901094777715e-05,
+      "loss": 0.377,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3241469816272966,
+      "grad_norm": 0.10867731974068413,
+      "learning_rate": 3.429746887255234e-05,
+      "loss": 0.3726,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3254593175853018,
+      "grad_norm": 0.10492751917669109,
+      "learning_rate": 3.426201496351017e-05,
+      "loss": 0.3795,
+      "step": 1010
+    },
+    {
+      "epoch": 1.326771653543307,
+      "grad_norm": 0.10955687832896575,
+      "learning_rate": 3.422653945034925e-05,
+      "loss": 0.3802,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3280839895013123,
+      "grad_norm": 0.10933201986658457,
+      "learning_rate": 3.4191042415818e-05,
+      "loss": 0.3745,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3293963254593175,
+      "grad_norm": 0.10673462552752362,
+      "learning_rate": 3.415552394271506e-05,
+      "loss": 0.3752,
+      "step": 1013
+    },
+    {
+      "epoch": 1.330708661417323,
+      "grad_norm": 0.1080014944935063,
+      "learning_rate": 3.411998411388907e-05,
+      "loss": 0.378,
+      "step": 1014
+    },
+    {
+      "epoch": 1.332020997375328,
+      "grad_norm": 0.12470699097192006,
+      "learning_rate": 3.408442301223846e-05,
+      "loss": 0.3803,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.12167934980432343,
+      "learning_rate": 3.404884072071131e-05,
+      "loss": 0.3832,
+      "step": 1016
+    },
+    {
+      "epoch": 1.3346456692913387,
+      "grad_norm": 0.10371428782001373,
+      "learning_rate": 3.401323732230512e-05,
+      "loss": 0.3768,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3359580052493438,
+      "grad_norm": 0.139628783900874,
+      "learning_rate": 3.397761290006661e-05,
+      "loss": 0.3736,
+      "step": 1018
+    },
+    {
+      "epoch": 1.337270341207349,
+      "grad_norm": 0.11751600594457484,
+      "learning_rate": 3.394196753709157e-05,
+      "loss": 0.3809,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3385826771653544,
+      "grad_norm": 0.10722878013602477,
+      "learning_rate": 3.3906301316524585e-05,
+      "loss": 0.3738,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3398950131233596,
+      "grad_norm": 0.12527947923350663,
+      "learning_rate": 3.387061432155894e-05,
+      "loss": 0.38,
+      "step": 1021
+    },
+    {
+      "epoch": 1.341207349081365,
+      "grad_norm": 0.12847494670390583,
+      "learning_rate": 3.383490663543635e-05,
+      "loss": 0.377,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3425196850393701,
+      "grad_norm": 0.09498242225071009,
+      "learning_rate": 3.3799178341446794e-05,
+      "loss": 0.377,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3438320209973753,
+      "grad_norm": 0.10485074222980127,
+      "learning_rate": 3.376342952292834e-05,
+      "loss": 0.3773,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3451443569553807,
+      "grad_norm": 0.10411996710036212,
+      "learning_rate": 3.372766026326691e-05,
+      "loss": 0.3808,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3464566929133859,
+      "grad_norm": 0.11643184004176785,
+      "learning_rate": 3.369187064589611e-05,
+      "loss": 0.3859,
+      "step": 1026
+    },
+    {
+      "epoch": 1.347769028871391,
+      "grad_norm": 0.10368172118224732,
+      "learning_rate": 3.3656060754297044e-05,
+      "loss": 0.3759,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3490813648293964,
+      "grad_norm": 0.09977533908524072,
+      "learning_rate": 3.3620230671998084e-05,
+      "loss": 0.3805,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3503937007874016,
+      "grad_norm": 0.10194262300452411,
+      "learning_rate": 3.3584380482574716e-05,
+      "loss": 0.3764,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3517060367454068,
+      "grad_norm": 0.10916456606990443,
+      "learning_rate": 3.354851026964933e-05,
+      "loss": 0.3727,
+      "step": 1030
+    },
+    {
+      "epoch": 1.3530183727034122,
+      "grad_norm": 0.11725282260330819,
+      "learning_rate": 3.351262011689099e-05,
+      "loss": 0.3894,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3543307086614174,
+      "grad_norm": 0.09272589611640758,
+      "learning_rate": 3.347671010801532e-05,
+      "loss": 0.3712,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3556430446194225,
+      "grad_norm": 0.12660056759757232,
+      "learning_rate": 3.344078032678422e-05,
+      "loss": 0.3755,
+      "step": 1033
+    },
+    {
+      "epoch": 1.356955380577428,
+      "grad_norm": 0.12369341663721592,
+      "learning_rate": 3.340483085700573e-05,
+      "loss": 0.3788,
+      "step": 1034
+    },
+    {
+      "epoch": 1.358267716535433,
+      "grad_norm": 0.10141924002266478,
+      "learning_rate": 3.33688617825338e-05,
+      "loss": 0.3761,
+      "step": 1035
+    },
+    {
+      "epoch": 1.3595800524934383,
+      "grad_norm": 0.1191807605319902,
+      "learning_rate": 3.333287318726813e-05,
+      "loss": 0.3772,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3608923884514437,
+      "grad_norm": 0.1005043783267694,
+      "learning_rate": 3.329686515515393e-05,
+      "loss": 0.3712,
+      "step": 1037
+    },
+    {
+      "epoch": 1.3622047244094488,
+      "grad_norm": 0.11019900550164984,
+      "learning_rate": 3.3260837770181756e-05,
+      "loss": 0.3809,
+      "step": 1038
+    },
+    {
+      "epoch": 1.363517060367454,
+      "grad_norm": 0.10423404471263267,
+      "learning_rate": 3.322479111638732e-05,
+      "loss": 0.381,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3648293963254594,
+      "grad_norm": 0.21365999296787339,
+      "learning_rate": 3.318872527785124e-05,
+      "loss": 0.3894,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3661417322834646,
+      "grad_norm": 0.11217834341230006,
+      "learning_rate": 3.3152640338698946e-05,
+      "loss": 0.3849,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3674540682414698,
+      "grad_norm": 0.0985325343622421,
+      "learning_rate": 3.311653638310036e-05,
+      "loss": 0.3669,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3687664041994752,
+      "grad_norm": 0.1324225163005603,
+      "learning_rate": 3.308041349526979e-05,
+      "loss": 0.3744,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3700787401574803,
+      "grad_norm": 0.10412044130846862,
+      "learning_rate": 3.30442717594657e-05,
+      "loss": 0.377,
+      "step": 1044
+    },
+    {
+      "epoch": 1.3713910761154855,
+      "grad_norm": 0.10124014943311734,
+      "learning_rate": 3.300811125999053e-05,
+      "loss": 0.3795,
+      "step": 1045
+    },
+    {
+      "epoch": 1.372703412073491,
+      "grad_norm": 0.1041646240636703,
+      "learning_rate": 3.297193208119047e-05,
+      "loss": 0.3675,
+      "step": 1046
+    },
+    {
+      "epoch": 1.374015748031496,
+      "grad_norm": 0.11433695551815642,
+      "learning_rate": 3.293573430745527e-05,
+      "loss": 0.3801,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3753280839895012,
+      "grad_norm": 0.9691314280001578,
+      "learning_rate": 3.2899518023218096e-05,
+      "loss": 0.5048,
+      "step": 1048
+    },
+    {
+      "epoch": 1.3766404199475066,
+      "grad_norm": 0.09699987643766399,
+      "learning_rate": 3.286328331295525e-05,
+      "loss": 0.3746,
+      "step": 1049
+    },
+    {
+      "epoch": 1.3779527559055118,
+      "grad_norm": 0.09331852885503358,
+      "learning_rate": 3.282703026118602e-05,
+      "loss": 0.3769,
+      "step": 1050
+    },
+    {
+      "epoch": 1.379265091863517,
+      "grad_norm": 0.10297162001582057,
+      "learning_rate": 3.27907589524725e-05,
+      "loss": 0.3734,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3805774278215224,
+      "grad_norm": 0.10488589376633052,
+      "learning_rate": 3.2754469471419346e-05,
+      "loss": 0.3692,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3818897637795275,
+      "grad_norm": 0.10967328040967199,
+      "learning_rate": 3.271816190267361e-05,
+      "loss": 0.3748,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3832020997375327,
+      "grad_norm": 0.2404147262354525,
+      "learning_rate": 3.2681836330924545e-05,
+      "loss": 0.3775,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3845144356955381,
+      "grad_norm": 0.11350567880485413,
+      "learning_rate": 3.264549284090336e-05,
+      "loss": 0.367,
+      "step": 1055
+    },
+    {
+      "epoch": 1.3858267716535433,
+      "grad_norm": 0.13611887718441043,
+      "learning_rate": 3.2609131517383104e-05,
+      "loss": 0.3722,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3871391076115485,
+      "grad_norm": 0.17088459475391177,
+      "learning_rate": 3.257275244517841e-05,
+      "loss": 0.3742,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3884514435695539,
+      "grad_norm": 0.10426072228157263,
+      "learning_rate": 3.2536355709145287e-05,
+      "loss": 0.3743,
+      "step": 1058
+    },
+    {
+      "epoch": 1.389763779527559,
+      "grad_norm": 0.13692462744192516,
+      "learning_rate": 3.2499941394180974e-05,
+      "loss": 0.3776,
+      "step": 1059
+    },
+    {
+      "epoch": 1.3910761154855642,
+      "grad_norm": 0.1161052047724405,
+      "learning_rate": 3.246350958522372e-05,
+      "loss": 0.3711,
+      "step": 1060
+    },
+    {
+      "epoch": 1.3923884514435696,
+      "grad_norm": 0.8624604277698171,
+      "learning_rate": 3.242706036725254e-05,
+      "loss": 0.4133,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3937007874015748,
+      "grad_norm": 0.20330009891456213,
+      "learning_rate": 3.2390593825287086e-05,
+      "loss": 0.3766,
+      "step": 1062
+    },
+    {
+      "epoch": 1.39501312335958,
+      "grad_norm": 0.1593378220919011,
+      "learning_rate": 3.235411004438741e-05,
+      "loss": 0.3823,
+      "step": 1063
+    },
+    {
+      "epoch": 1.3963254593175853,
+      "grad_norm": 0.1460689057018055,
+      "learning_rate": 3.231760910965379e-05,
+      "loss": 0.3796,
+      "step": 1064
+    },
+    {
+      "epoch": 1.3976377952755905,
+      "grad_norm": 0.11117174371535123,
+      "learning_rate": 3.228109110622649e-05,
+      "loss": 0.372,
+      "step": 1065
+    },
+    {
+      "epoch": 1.3989501312335957,
+      "grad_norm": 0.13037065914356483,
+      "learning_rate": 3.2244556119285616e-05,
+      "loss": 0.3791,
+      "step": 1066
+    },
+    {
+      "epoch": 1.400262467191601,
+      "grad_norm": 0.13008612751972362,
+      "learning_rate": 3.220800423405086e-05,
+      "loss": 0.3798,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4015748031496063,
+      "grad_norm": 0.11406725164495116,
+      "learning_rate": 3.2171435535781364e-05,
+      "loss": 0.3779,
+      "step": 1068
+    },
+    {
+      "epoch": 1.4028871391076114,
+      "grad_norm": 0.15623125227882076,
+      "learning_rate": 3.213485010977545e-05,
+      "loss": 0.3773,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4041994750656168,
+      "grad_norm": 0.10264102217648577,
+      "learning_rate": 3.209824804137048e-05,
+      "loss": 0.3768,
+      "step": 1070
+    },
+    {
+      "epoch": 1.405511811023622,
+      "grad_norm": 0.11941947655792956,
+      "learning_rate": 3.206162941594263e-05,
+      "loss": 0.3831,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4068241469816272,
+      "grad_norm": 0.09718110930060342,
+      "learning_rate": 3.20249943189067e-05,
+      "loss": 0.3718,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4081364829396326,
+      "grad_norm": 0.10403879854600272,
+      "learning_rate": 3.198834283571591e-05,
+      "loss": 0.383,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4094488188976377,
+      "grad_norm": 0.09961947392256104,
+      "learning_rate": 3.19516750518617e-05,
+      "loss": 0.3838,
+      "step": 1074
+    },
+    {
+      "epoch": 1.410761154855643,
+      "grad_norm": 0.09375858705456702,
+      "learning_rate": 3.191499105287353e-05,
+      "loss": 0.3853,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4120734908136483,
+      "grad_norm": 0.10898539467131445,
+      "learning_rate": 3.187829092431869e-05,
+      "loss": 0.3754,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4133858267716535,
+      "grad_norm": 0.08436820235580599,
+      "learning_rate": 3.1841574751802076e-05,
+      "loss": 0.3802,
+      "step": 1077
+    },
+    {
+      "epoch": 1.4146981627296589,
+      "grad_norm": 0.09119724739685257,
+      "learning_rate": 3.180484262096603e-05,
+      "loss": 0.3701,
+      "step": 1078
+    },
+    {
+      "epoch": 1.416010498687664,
+      "grad_norm": 0.0961315477802874,
+      "learning_rate": 3.17680946174901e-05,
+      "loss": 0.3819,
+      "step": 1079
+    },
+    {
+      "epoch": 1.4173228346456692,
+      "grad_norm": 0.09601471543158668,
+      "learning_rate": 3.1731330827090865e-05,
+      "loss": 0.3713,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4186351706036746,
+      "grad_norm": 0.10119147961156184,
+      "learning_rate": 3.169455133552173e-05,
+      "loss": 0.372,
+      "step": 1081
+    },
+    {
+      "epoch": 1.4199475065616798,
+      "grad_norm": 0.09053850565993789,
+      "learning_rate": 3.165775622857273e-05,
+      "loss": 0.3769,
+      "step": 1082
+    },
+    {
+      "epoch": 1.421259842519685,
+      "grad_norm": 0.1252310225475832,
+      "learning_rate": 3.162094559207029e-05,
+      "loss": 0.3792,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4225721784776904,
+      "grad_norm": 0.09637914168993317,
+      "learning_rate": 3.158411951187711e-05,
+      "loss": 0.3791,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4238845144356955,
+      "grad_norm": 0.10109819632885503,
+      "learning_rate": 3.154727807389186e-05,
+      "loss": 0.3739,
+      "step": 1085
+    },
+    {
+      "epoch": 1.425196850393701,
+      "grad_norm": 0.10432212714059884,
+      "learning_rate": 3.151042136404908e-05,
+      "loss": 0.3709,
+      "step": 1086
+    },
+    {
+      "epoch": 1.426509186351706,
+      "grad_norm": 0.09328785141889541,
+      "learning_rate": 3.14735494683189e-05,
+      "loss": 0.3778,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4278215223097113,
+      "grad_norm": 0.12145741398664167,
+      "learning_rate": 3.1436662472706895e-05,
+      "loss": 0.3721,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4291338582677167,
+      "grad_norm": 0.08871496973104057,
+      "learning_rate": 3.139976046325384e-05,
+      "loss": 0.3775,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4304461942257218,
+      "grad_norm": 0.11404444024534809,
+      "learning_rate": 3.136284352603554e-05,
+      "loss": 0.3805,
+      "step": 1090
+    },
+    {
+      "epoch": 1.431758530183727,
+      "grad_norm": 0.09081272556664162,
+      "learning_rate": 3.132591174716262e-05,
+      "loss": 0.3824,
+      "step": 1091
+    },
+    {
+      "epoch": 1.4330708661417324,
+      "grad_norm": 0.09361365477549906,
+      "learning_rate": 3.128896521278033e-05,
+      "loss": 0.3818,
+      "step": 1092
+    },
+    {
+      "epoch": 1.4343832020997376,
+      "grad_norm": 0.09652248643551763,
+      "learning_rate": 3.1252004009068324e-05,
+      "loss": 0.3794,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4356955380577427,
+      "grad_norm": 0.09729184539371502,
+      "learning_rate": 3.121502822224049e-05,
+      "loss": 0.3726,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4370078740157481,
+      "grad_norm": 0.09766883029463666,
+      "learning_rate": 3.1178037938544706e-05,
+      "loss": 0.3816,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4383202099737533,
+      "grad_norm": 0.08998963901736301,
+      "learning_rate": 3.11410332442627e-05,
+      "loss": 0.373,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4396325459317585,
+      "grad_norm": 0.10143101926483168,
+      "learning_rate": 3.110401422570979e-05,
+      "loss": 0.3736,
+      "step": 1097
+    },
+    {
+      "epoch": 1.4409448818897639,
+      "grad_norm": 0.10060749479807819,
+      "learning_rate": 3.10669809692347e-05,
+      "loss": 0.3776,
+      "step": 1098
+    },
+    {
+      "epoch": 1.442257217847769,
+      "grad_norm": 0.09786360335601098,
+      "learning_rate": 3.1029933561219375e-05,
+      "loss": 0.3747,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4435695538057742,
+      "grad_norm": 0.1343606678998333,
+      "learning_rate": 3.099287208807879e-05,
+      "loss": 0.3792,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4448818897637796,
+      "grad_norm": 0.10190065251230013,
+      "learning_rate": 3.095579663626069e-05,
+      "loss": 0.3725,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4461942257217848,
+      "grad_norm": 0.10383346915620609,
+      "learning_rate": 3.0918707292245454e-05,
+      "loss": 0.372,
+      "step": 1102
+    },
+    {
+      "epoch": 1.44750656167979,
+      "grad_norm": 0.11567885764241202,
+      "learning_rate": 3.0881604142545837e-05,
+      "loss": 0.3778,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4488188976377954,
+      "grad_norm": 0.1000983222987359,
+      "learning_rate": 3.084448727370683e-05,
+      "loss": 0.3774,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4501312335958005,
+      "grad_norm": 0.11702315213546698,
+      "learning_rate": 3.080735677230541e-05,
+      "loss": 0.3733,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4514435695538057,
+      "grad_norm": 0.09278223833647567,
+      "learning_rate": 3.077021272495033e-05,
+      "loss": 0.3778,
+      "step": 1106
+    },
+    {
+      "epoch": 1.452755905511811,
+      "grad_norm": 0.09481615397468997,
+      "learning_rate": 3.073305521828198e-05,
+      "loss": 0.3766,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4540682414698163,
+      "grad_norm": 0.08513825472030663,
+      "learning_rate": 3.069588433897212e-05,
+      "loss": 0.3846,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4553805774278215,
+      "grad_norm": 0.08239419705036141,
+      "learning_rate": 3.065870017372371e-05,
+      "loss": 0.382,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4566929133858268,
+      "grad_norm": 0.07834025880278521,
+      "learning_rate": 3.06215028092707e-05,
+      "loss": 0.3754,
+      "step": 1110
+    },
+    {
+      "epoch": 1.458005249343832,
+      "grad_norm": 0.08658541285591478,
+      "learning_rate": 3.058429233237781e-05,
+      "loss": 0.375,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4593175853018372,
+      "grad_norm": 0.07997808416530298,
+      "learning_rate": 3.0547068829840384e-05,
+      "loss": 0.3842,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4606299212598426,
+      "grad_norm": 0.0794746757240125,
+      "learning_rate": 3.0509832388484104e-05,
+      "loss": 0.3867,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4619422572178478,
+      "grad_norm": 0.08416925930506677,
+      "learning_rate": 3.0472583095164874e-05,
+      "loss": 0.3752,
+      "step": 1114
+    },
+    {
+      "epoch": 1.463254593175853,
+      "grad_norm": 0.08806782737850741,
+      "learning_rate": 3.043532103676854e-05,
+      "loss": 0.3754,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4645669291338583,
+      "grad_norm": 0.08038913759077383,
+      "learning_rate": 3.039804630021075e-05,
+      "loss": 0.3817,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4658792650918635,
+      "grad_norm": 0.08403690802991205,
+      "learning_rate": 3.0360758972436708e-05,
+      "loss": 0.3757,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4671916010498687,
+      "grad_norm": 0.07793661915965237,
+      "learning_rate": 3.0323459140420996e-05,
+      "loss": 0.3757,
+      "step": 1118
+    },
+    {
+      "epoch": 1.468503937007874,
+      "grad_norm": 0.7305605133197076,
+      "learning_rate": 3.0286146891167355e-05,
+      "loss": 0.3908,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4698162729658792,
+      "grad_norm": 0.09388975435573238,
+      "learning_rate": 3.024882231170851e-05,
+      "loss": 0.3827,
+      "step": 1120
+    },
+    {
+      "epoch": 1.4711286089238844,
+      "grad_norm": 0.08747306962302862,
+      "learning_rate": 3.0211485489105917e-05,
+      "loss": 0.3774,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4724409448818898,
+      "grad_norm": 0.08494796923533174,
+      "learning_rate": 3.0174136510449615e-05,
+      "loss": 0.3749,
+      "step": 1122
+    },
+    {
+      "epoch": 1.473753280839895,
+      "grad_norm": 0.08890594744949992,
+      "learning_rate": 3.0136775462857973e-05,
+      "loss": 0.3804,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4750656167979002,
+      "grad_norm": 0.0926281994412685,
+      "learning_rate": 3.009940243347754e-05,
+      "loss": 0.378,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4763779527559056,
+      "grad_norm": 0.09803863768505795,
+      "learning_rate": 3.0062017509482797e-05,
+      "loss": 0.379,
+      "step": 1125
+    },
+    {
+      "epoch": 1.4776902887139107,
+      "grad_norm": 0.09032068872345834,
+      "learning_rate": 3.0024620778075963e-05,
+      "loss": 0.389,
+      "step": 1126
+    },
+    {
+      "epoch": 1.479002624671916,
+      "grad_norm": 0.11636300077155064,
+      "learning_rate": 2.9987212326486813e-05,
+      "loss": 0.3781,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4803149606299213,
+      "grad_norm": 0.09180964116329138,
+      "learning_rate": 2.9949792241972458e-05,
+      "loss": 0.376,
+      "step": 1128
+    },
+    {
+      "epoch": 1.4816272965879265,
+      "grad_norm": 0.1709845725368337,
+      "learning_rate": 2.9912360611817135e-05,
+      "loss": 0.3692,
+      "step": 1129
+    },
+    {
+      "epoch": 1.4829396325459316,
+      "grad_norm": 0.1048519794912273,
+      "learning_rate": 2.987491752333201e-05,
+      "loss": 0.3762,
+      "step": 1130
+    },
+    {
+      "epoch": 1.484251968503937,
+      "grad_norm": 0.09181514883252197,
+      "learning_rate": 2.9837463063854994e-05,
+      "loss": 0.3854,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4855643044619422,
+      "grad_norm": 0.09107778544862581,
+      "learning_rate": 2.9799997320750507e-05,
+      "loss": 0.3688,
+      "step": 1132
+    },
+    {
+      "epoch": 1.4868766404199474,
+      "grad_norm": 0.10719426826015938,
+      "learning_rate": 2.976252038140928e-05,
+      "loss": 0.3803,
+      "step": 1133
+    },
+    {
+      "epoch": 1.4881889763779528,
+      "grad_norm": 0.09775401779693955,
+      "learning_rate": 2.9725032333248188e-05,
+      "loss": 0.3828,
+      "step": 1134
+    },
+    {
+      "epoch": 1.489501312335958,
+      "grad_norm": 0.15980025281173965,
+      "learning_rate": 2.9687533263709983e-05,
+      "loss": 0.3821,
+      "step": 1135
+    },
+    {
+      "epoch": 1.4908136482939631,
+      "grad_norm": 0.10359093562217811,
+      "learning_rate": 2.965002326026316e-05,
+      "loss": 0.3798,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4921259842519685,
+      "grad_norm": 0.0960779687352872,
+      "learning_rate": 2.9612502410401693e-05,
+      "loss": 0.3718,
+      "step": 1137
+    },
+    {
+      "epoch": 1.4934383202099737,
+      "grad_norm": 0.09730681871388,
+      "learning_rate": 2.9574970801644865e-05,
+      "loss": 0.3723,
+      "step": 1138
+    },
+    {
+      "epoch": 1.4947506561679789,
+      "grad_norm": 0.08363280660402565,
+      "learning_rate": 2.9537428521537054e-05,
+      "loss": 0.3748,
+      "step": 1139
+    },
+    {
+      "epoch": 1.4960629921259843,
+      "grad_norm": 0.09069477483222212,
+      "learning_rate": 2.9499875657647517e-05,
+      "loss": 0.3828,
+      "step": 1140
+    },
+    {
+      "epoch": 1.4973753280839894,
+      "grad_norm": 0.09130081678499127,
+      "learning_rate": 2.9462312297570227e-05,
+      "loss": 0.3795,
+      "step": 1141
+    },
+    {
+      "epoch": 1.4986876640419948,
+      "grad_norm": 0.08805263622529404,
+      "learning_rate": 2.942473852892362e-05,
+      "loss": 0.3768,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.09803441343893382,
+      "learning_rate": 2.9387154439350406e-05,
+      "loss": 0.3795,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5013123359580054,
+      "grad_norm": 0.09102511722644539,
+      "learning_rate": 2.9349560116517383e-05,
+      "loss": 0.3705,
+      "step": 1144
+    },
+    {
+      "epoch": 1.5026246719160103,
+      "grad_norm": 0.08512203410749762,
+      "learning_rate": 2.9311955648115208e-05,
+      "loss": 0.3756,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5039370078740157,
+      "grad_norm": 0.09854358486838942,
+      "learning_rate": 2.9274341121858203e-05,
+      "loss": 0.3772,
+      "step": 1146
+    },
+    {
+      "epoch": 1.5052493438320211,
+      "grad_norm": 0.08768365177189426,
+      "learning_rate": 2.9236716625484156e-05,
+      "loss": 0.3796,
+      "step": 1147
+    },
+    {
+      "epoch": 1.506561679790026,
+      "grad_norm": 0.09496518465950153,
+      "learning_rate": 2.9199082246754122e-05,
+      "loss": 0.3763,
+      "step": 1148
+    },
+    {
+      "epoch": 1.5078740157480315,
+      "grad_norm": 0.09506781258153221,
+      "learning_rate": 2.9161438073452186e-05,
+      "loss": 0.3726,
+      "step": 1149
+    },
+    {
+      "epoch": 1.5091863517060369,
+      "grad_norm": 0.09830588329702851,
+      "learning_rate": 2.9123784193385283e-05,
+      "loss": 0.3682,
+      "step": 1150
+    },
+    {
+      "epoch": 1.510498687664042,
+      "grad_norm": 0.09558495881404815,
+      "learning_rate": 2.9086120694383e-05,
+      "loss": 0.3745,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5118110236220472,
+      "grad_norm": 0.09134572706220935,
+      "learning_rate": 2.9048447664297356e-05,
+      "loss": 0.3796,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5131233595800526,
+      "grad_norm": 0.10739447129456987,
+      "learning_rate": 2.9010765191002597e-05,
+      "loss": 0.38,
+      "step": 1153
+    },
+    {
+      "epoch": 1.5144356955380578,
+      "grad_norm": 0.08722725789869759,
+      "learning_rate": 2.8973073362394998e-05,
+      "loss": 0.3812,
+      "step": 1154
+    },
+    {
+      "epoch": 1.515748031496063,
+      "grad_norm": 0.10255732030543267,
+      "learning_rate": 2.8935372266392662e-05,
+      "loss": 0.378,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5170603674540684,
+      "grad_norm": 0.21617459470670783,
+      "learning_rate": 2.889766199093531e-05,
+      "loss": 0.3857,
+      "step": 1156
+    },
+    {
+      "epoch": 1.5183727034120735,
+      "grad_norm": 0.09730867461010395,
+      "learning_rate": 2.8859942623984056e-05,
+      "loss": 0.3863,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5196850393700787,
+      "grad_norm": 0.10231375482316067,
+      "learning_rate": 2.882221425352123e-05,
+      "loss": 0.3717,
+      "step": 1158
+    },
+    {
+      "epoch": 1.520997375328084,
+      "grad_norm": 0.08909669119976227,
+      "learning_rate": 2.8784476967550177e-05,
+      "loss": 0.3732,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5223097112860893,
+      "grad_norm": 0.09383199057381207,
+      "learning_rate": 2.8746730854095026e-05,
+      "loss": 0.3707,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5236220472440944,
+      "grad_norm": 0.08506376079467293,
+      "learning_rate": 2.8708976001200493e-05,
+      "loss": 0.3727,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5249343832020998,
+      "grad_norm": 0.08760345377189876,
+      "learning_rate": 2.8671212496931682e-05,
+      "loss": 0.3743,
+      "step": 1162
+    },
+    {
+      "epoch": 1.526246719160105,
+      "grad_norm": 0.0834232232754643,
+      "learning_rate": 2.8633440429373888e-05,
+      "loss": 0.3684,
+      "step": 1163
+    },
+    {
+      "epoch": 1.5275590551181102,
+      "grad_norm": 0.08922114678439605,
+      "learning_rate": 2.8595659886632363e-05,
+      "loss": 0.3791,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5288713910761156,
+      "grad_norm": 0.09098253398567528,
+      "learning_rate": 2.8557870956832132e-05,
+      "loss": 0.3798,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5301837270341208,
+      "grad_norm": 0.08077113721364407,
+      "learning_rate": 2.852007372811779e-05,
+      "loss": 0.385,
+      "step": 1166
+    },
+    {
+      "epoch": 1.531496062992126,
+      "grad_norm": 0.08244668471857224,
+      "learning_rate": 2.848226828865329e-05,
+      "loss": 0.3816,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5328083989501313,
+      "grad_norm": 0.07899460318801674,
+      "learning_rate": 2.8444454726621734e-05,
+      "loss": 0.3702,
+      "step": 1168
+    },
+    {
+      "epoch": 1.5341207349081365,
+      "grad_norm": 0.9057865480560622,
+      "learning_rate": 2.8406633130225163e-05,
+      "loss": 0.5108,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5354330708661417,
+      "grad_norm": 0.08617977284028872,
+      "learning_rate": 2.8368803587684368e-05,
+      "loss": 0.3794,
+      "step": 1170
+    },
+    {
+      "epoch": 1.536745406824147,
+      "grad_norm": 0.08239502011885261,
+      "learning_rate": 2.8330966187238673e-05,
+      "loss": 0.3833,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5380577427821522,
+      "grad_norm": 0.09039190839168151,
+      "learning_rate": 2.8293121017145735e-05,
+      "loss": 0.3793,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5393700787401574,
+      "grad_norm": 0.0894352807706549,
+      "learning_rate": 2.825526816568132e-05,
+      "loss": 0.3707,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5406824146981628,
+      "grad_norm": 0.215839231796849,
+      "learning_rate": 2.8217407721139112e-05,
+      "loss": 0.3768,
+      "step": 1174
+    },
+    {
+      "epoch": 1.541994750656168,
+      "grad_norm": 0.08008701601159633,
+      "learning_rate": 2.8179539771830537e-05,
+      "loss": 0.3777,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5433070866141732,
+      "grad_norm": 0.09016156074333795,
+      "learning_rate": 2.8141664406084495e-05,
+      "loss": 0.3705,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5446194225721785,
+      "grad_norm": 0.08879402025083412,
+      "learning_rate": 2.810378171224719e-05,
+      "loss": 0.3799,
+      "step": 1177
+    },
+    {
+      "epoch": 1.5459317585301837,
+      "grad_norm": 0.0863506212152667,
+      "learning_rate": 2.806589177868192e-05,
+      "loss": 0.3722,
+      "step": 1178
+    },
+    {
+      "epoch": 1.547244094488189,
+      "grad_norm": 0.09372942558988334,
+      "learning_rate": 2.802799469376888e-05,
+      "loss": 0.378,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5485564304461943,
+      "grad_norm": 0.08267592912632861,
+      "learning_rate": 2.799009054590494e-05,
+      "loss": 0.3788,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5498687664041995,
+      "grad_norm": 0.09118272978416449,
+      "learning_rate": 2.7952179423503432e-05,
+      "loss": 0.3858,
+      "step": 1181
+    },
+    {
+      "epoch": 1.5511811023622046,
+      "grad_norm": 0.07988562172565089,
+      "learning_rate": 2.7914261414993982e-05,
+      "loss": 0.3707,
+      "step": 1182
+    },
+    {
+      "epoch": 1.55249343832021,
+      "grad_norm": 0.09468697091911725,
+      "learning_rate": 2.7876336608822256e-05,
+      "loss": 0.3708,
+      "step": 1183
+    },
+    {
+      "epoch": 1.5538057742782152,
+      "grad_norm": 0.08263365085750876,
+      "learning_rate": 2.7838405093449792e-05,
+      "loss": 0.3748,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5551181102362204,
+      "grad_norm": 0.8130094534934841,
+      "learning_rate": 2.7800466957353767e-05,
+      "loss": 0.3941,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5564304461942258,
+      "grad_norm": 0.09359157528669675,
+      "learning_rate": 2.776252228902679e-05,
+      "loss": 0.3767,
+      "step": 1186
+    },
+    {
+      "epoch": 1.557742782152231,
+      "grad_norm": 0.08227960770079872,
+      "learning_rate": 2.7724571176976732e-05,
+      "loss": 0.3816,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5590551181102361,
+      "grad_norm": 0.12069178774097711,
+      "learning_rate": 2.7686613709726493e-05,
+      "loss": 0.3887,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5603674540682415,
+      "grad_norm": 0.2841702404072889,
+      "learning_rate": 2.764864997581378e-05,
+      "loss": 0.3761,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5616797900262467,
+      "grad_norm": 0.1872024664829523,
+      "learning_rate": 2.7610680063790907e-05,
+      "loss": 0.3814,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5629921259842519,
+      "grad_norm": 0.09591871767125988,
+      "learning_rate": 2.757270406222464e-05,
+      "loss": 0.3798,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5643044619422573,
+      "grad_norm": 0.09073369354787368,
+      "learning_rate": 2.7534722059695917e-05,
+      "loss": 0.3752,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5656167979002624,
+      "grad_norm": 0.14507392298144337,
+      "learning_rate": 2.749673414479967e-05,
+      "loss": 0.3771,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5669291338582676,
+      "grad_norm": 0.07940002502751882,
+      "learning_rate": 2.745874040614464e-05,
+      "loss": 0.3728,
+      "step": 1194
+    },
+    {
+      "epoch": 1.568241469816273,
+      "grad_norm": 0.08046361832348527,
+      "learning_rate": 2.7420740932353162e-05,
+      "loss": 0.3736,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5695538057742782,
+      "grad_norm": 0.08424069862029145,
+      "learning_rate": 2.7382735812060912e-05,
+      "loss": 0.3747,
+      "step": 1196
+    },
+    {
+      "epoch": 1.5708661417322833,
+      "grad_norm": 0.08896438501181192,
+      "learning_rate": 2.734472513391677e-05,
+      "loss": 0.3835,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5721784776902887,
+      "grad_norm": 0.08406943091636061,
+      "learning_rate": 2.7306708986582553e-05,
+      "loss": 0.3709,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5734908136482941,
+      "grad_norm": 0.08075053407925999,
+      "learning_rate": 2.726868745873286e-05,
+      "loss": 0.3851,
+      "step": 1199
+    },
+    {
+      "epoch": 1.574803149606299,
+      "grad_norm": 0.08508438568748895,
+      "learning_rate": 2.723066063905483e-05,
+      "loss": 0.3846,
+      "step": 1200
+    },
+    {
+      "epoch": 1.5761154855643045,
+      "grad_norm": 0.08241881731782234,
+      "learning_rate": 2.7192628616247932e-05,
+      "loss": 0.3697,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5774278215223099,
+      "grad_norm": 0.08803968552470688,
+      "learning_rate": 2.7154591479023794e-05,
+      "loss": 0.3794,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5787401574803148,
+      "grad_norm": 0.08205523339950967,
+      "learning_rate": 2.711654931610597e-05,
+      "loss": 0.3764,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5800524934383202,
+      "grad_norm": 0.08658270862711326,
+      "learning_rate": 2.7078502216229728e-05,
+      "loss": 0.3719,
+      "step": 1204
+    },
+    {
+      "epoch": 1.5813648293963256,
+      "grad_norm": 0.08712326412413146,
+      "learning_rate": 2.704045026814184e-05,
+      "loss": 0.378,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5826771653543306,
+      "grad_norm": 0.09768017274743301,
+      "learning_rate": 2.700239356060042e-05,
+      "loss": 0.3805,
+      "step": 1206
+    },
+    {
+      "epoch": 1.583989501312336,
+      "grad_norm": 0.08711658600889674,
+      "learning_rate": 2.6964332182374653e-05,
+      "loss": 0.3691,
+      "step": 1207
+    },
+    {
+      "epoch": 1.5853018372703414,
+      "grad_norm": 0.10293102312919887,
+      "learning_rate": 2.6926266222244634e-05,
+      "loss": 0.3788,
+      "step": 1208
+    },
+    {
+      "epoch": 1.5866141732283463,
+      "grad_norm": 0.09516481708905597,
+      "learning_rate": 2.6888195769001146e-05,
+      "loss": 0.3844,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5879265091863517,
+      "grad_norm": 0.0954646645591328,
+      "learning_rate": 2.6850120911445436e-05,
+      "loss": 0.376,
+      "step": 1210
+    },
+    {
+      "epoch": 1.589238845144357,
+      "grad_norm": 0.13402116917794715,
+      "learning_rate": 2.6812041738389055e-05,
+      "loss": 0.3834,
+      "step": 1211
+    },
+    {
+      "epoch": 1.590551181102362,
+      "grad_norm": 0.0787225226871965,
+      "learning_rate": 2.6773958338653588e-05,
+      "loss": 0.373,
+      "step": 1212
+    },
+    {
+      "epoch": 1.5918635170603674,
+      "grad_norm": 0.12422610422025135,
+      "learning_rate": 2.673587080107049e-05,
+      "loss": 0.373,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5931758530183728,
+      "grad_norm": 0.08872415218809131,
+      "learning_rate": 2.6697779214480885e-05,
+      "loss": 0.3782,
+      "step": 1214
+    },
+    {
+      "epoch": 1.594488188976378,
+      "grad_norm": 0.10362118378424036,
+      "learning_rate": 2.6659683667735312e-05,
+      "loss": 0.3784,
+      "step": 1215
+    },
+    {
+      "epoch": 1.5958005249343832,
+      "grad_norm": 0.09658647128866811,
+      "learning_rate": 2.6621584249693575e-05,
+      "loss": 0.3734,
+      "step": 1216
+    },
+    {
+      "epoch": 1.5971128608923886,
+      "grad_norm": 0.09569211991776817,
+      "learning_rate": 2.658348104922449e-05,
+      "loss": 0.3721,
+      "step": 1217
+    },
+    {
+      "epoch": 1.5984251968503937,
+      "grad_norm": 0.07991546450322744,
+      "learning_rate": 2.65453741552057e-05,
+      "loss": 0.3771,
+      "step": 1218
+    },
+    {
+      "epoch": 1.599737532808399,
+      "grad_norm": 0.08763818840681191,
+      "learning_rate": 2.6507263656523462e-05,
+      "loss": 0.3739,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6010498687664043,
+      "grad_norm": 0.0961226059747525,
+      "learning_rate": 2.646914964207245e-05,
+      "loss": 0.3739,
+      "step": 1220
+    },
+    {
+      "epoch": 1.6023622047244095,
+      "grad_norm": 0.0965053070232335,
+      "learning_rate": 2.643103220075552e-05,
+      "loss": 0.3785,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6036745406824147,
+      "grad_norm": 0.0903860670640885,
+      "learning_rate": 2.639291142148354e-05,
+      "loss": 0.3821,
+      "step": 1222
+    },
+    {
+      "epoch": 1.60498687664042,
+      "grad_norm": 0.11718883460228136,
+      "learning_rate": 2.635478739317517e-05,
+      "loss": 0.3801,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6062992125984252,
+      "grad_norm": 0.0911812375921744,
+      "learning_rate": 2.631666020475662e-05,
+      "loss": 0.3743,
+      "step": 1224
+    },
+    {
+      "epoch": 1.6076115485564304,
+      "grad_norm": 0.09940809001574305,
+      "learning_rate": 2.627852994516149e-05,
+      "loss": 0.3714,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6089238845144358,
+      "grad_norm": 0.10009347924352205,
+      "learning_rate": 2.6240396703330544e-05,
+      "loss": 0.3689,
+      "step": 1226
+    },
+    {
+      "epoch": 1.610236220472441,
+      "grad_norm": 0.08001953168794321,
+      "learning_rate": 2.6202260568211502e-05,
+      "loss": 0.3874,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6115485564304461,
+      "grad_norm": 0.08401243224458588,
+      "learning_rate": 2.6164121628758814e-05,
+      "loss": 0.376,
+      "step": 1228
+    },
+    {
+      "epoch": 1.6128608923884515,
+      "grad_norm": 0.08223260988020036,
+      "learning_rate": 2.6125979973933505e-05,
+      "loss": 0.3858,
+      "step": 1229
+    },
+    {
+      "epoch": 1.6141732283464567,
+      "grad_norm": 0.0900424528540787,
+      "learning_rate": 2.6087835692702904e-05,
+      "loss": 0.3684,
+      "step": 1230
+    },
+    {
+      "epoch": 1.6154855643044619,
+      "grad_norm": 0.07419047236621268,
+      "learning_rate": 2.604968887404049e-05,
+      "loss": 0.3769,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6167979002624673,
+      "grad_norm": 0.08437772509945372,
+      "learning_rate": 2.601153960692564e-05,
+      "loss": 0.3816,
+      "step": 1232
+    },
+    {
+      "epoch": 1.6181102362204725,
+      "grad_norm": 0.08602115594234694,
+      "learning_rate": 2.5973387980343443e-05,
+      "loss": 0.3809,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6194225721784776,
+      "grad_norm": 0.07804170814167065,
+      "learning_rate": 2.5935234083284516e-05,
+      "loss": 0.3802,
+      "step": 1234
+    },
+    {
+      "epoch": 1.620734908136483,
+      "grad_norm": 0.08622840574355366,
+      "learning_rate": 2.589707800474474e-05,
+      "loss": 0.3705,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6220472440944882,
+      "grad_norm": 0.07808393242398147,
+      "learning_rate": 2.5858919833725114e-05,
+      "loss": 0.3771,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6233595800524934,
+      "grad_norm": 0.09186694524115407,
+      "learning_rate": 2.5820759659231492e-05,
+      "loss": 0.3804,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6246719160104988,
+      "grad_norm": 0.09354010538434414,
+      "learning_rate": 2.5782597570274427e-05,
+      "loss": 0.3678,
+      "step": 1238
+    },
+    {
+      "epoch": 1.625984251968504,
+      "grad_norm": 0.0769703926946228,
+      "learning_rate": 2.5744433655868904e-05,
+      "loss": 0.3718,
+      "step": 1239
+    },
+    {
+      "epoch": 1.627296587926509,
+      "grad_norm": 0.08763436739104581,
+      "learning_rate": 2.5706268005034202e-05,
+      "loss": 0.3753,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6286089238845145,
+      "grad_norm": 0.0899239927217872,
+      "learning_rate": 2.5668100706793615e-05,
+      "loss": 0.3747,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6299212598425197,
+      "grad_norm": 0.0821758854756211,
+      "learning_rate": 2.562993185017431e-05,
+      "loss": 0.3808,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6312335958005248,
+      "grad_norm": 0.07987854743831367,
+      "learning_rate": 2.5591761524207086e-05,
+      "loss": 0.3824,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6325459317585302,
+      "grad_norm": 0.08263263147990764,
+      "learning_rate": 2.555358981792615e-05,
+      "loss": 0.3794,
+      "step": 1244
+    },
+    {
+      "epoch": 1.6338582677165354,
+      "grad_norm": 0.08216677383066037,
+      "learning_rate": 2.5515416820368933e-05,
+      "loss": 0.3789,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6351706036745406,
+      "grad_norm": 0.07956052914830104,
+      "learning_rate": 2.5477242620575896e-05,
+      "loss": 0.3797,
+      "step": 1246
+    },
+    {
+      "epoch": 1.636482939632546,
+      "grad_norm": 0.07577110495931663,
+      "learning_rate": 2.5439067307590292e-05,
+      "loss": 0.3731,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6377952755905512,
+      "grad_norm": 0.08179033818903186,
+      "learning_rate": 2.5400890970457962e-05,
+      "loss": 0.3755,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6391076115485563,
+      "grad_norm": 0.3274917122510175,
+      "learning_rate": 2.536271369822714e-05,
+      "loss": 0.3775,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6404199475065617,
+      "grad_norm": 0.09118629824996985,
+      "learning_rate": 2.5324535579948274e-05,
+      "loss": 0.3791,
+      "step": 1250
+    },
+    {
+      "epoch": 1.641732283464567,
+      "grad_norm": 0.07666869495192381,
+      "learning_rate": 2.5286356704673735e-05,
+      "loss": 0.3819,
+      "step": 1251
+    },
+    {
+      "epoch": 1.643044619422572,
+      "grad_norm": 0.08964730818644759,
+      "learning_rate": 2.5248177161457687e-05,
+      "loss": 0.3824,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6443569553805775,
+      "grad_norm": 0.0881691923842221,
+      "learning_rate": 2.5209997039355837e-05,
+      "loss": 0.3825,
+      "step": 1253
+    },
+    {
+      "epoch": 1.6456692913385826,
+      "grad_norm": 0.08313603178747564,
+      "learning_rate": 2.5171816427425272e-05,
+      "loss": 0.3756,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6469816272965878,
+      "grad_norm": 0.0788118591547596,
+      "learning_rate": 2.5133635414724192e-05,
+      "loss": 0.3715,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6482939632545932,
+      "grad_norm": 0.08341469698040833,
+      "learning_rate": 2.509545409031174e-05,
+      "loss": 0.3867,
+      "step": 1256
+    },
+    {
+      "epoch": 1.6496062992125984,
+      "grad_norm": 0.07657248386491622,
+      "learning_rate": 2.5057272543247784e-05,
+      "loss": 0.3736,
+      "step": 1257
+    },
+    {
+      "epoch": 1.6509186351706036,
+      "grad_norm": 0.25782485374690445,
+      "learning_rate": 2.501909086259273e-05,
+      "loss": 0.379,
+      "step": 1258
+    },
+    {
+      "epoch": 1.652230971128609,
+      "grad_norm": 0.09139410174894938,
+      "learning_rate": 2.4980909137407268e-05,
+      "loss": 0.368,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6535433070866141,
+      "grad_norm": 0.08939825446930573,
+      "learning_rate": 2.4942727456752215e-05,
+      "loss": 0.3817,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6548556430446193,
+      "grad_norm": 0.08211824508809352,
+      "learning_rate": 2.4904545909688267e-05,
+      "loss": 0.3662,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6561679790026247,
+      "grad_norm": 0.08050927305987655,
+      "learning_rate": 2.486636458527581e-05,
+      "loss": 0.3762,
+      "step": 1262
+    },
+    {
+      "epoch": 1.65748031496063,
+      "grad_norm": 0.0967071636357244,
+      "learning_rate": 2.482818357257473e-05,
+      "loss": 0.3783,
+      "step": 1263
+    },
+    {
+      "epoch": 1.658792650918635,
+      "grad_norm": 0.08648088073663057,
+      "learning_rate": 2.479000296064417e-05,
+      "loss": 0.3791,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6601049868766404,
+      "grad_norm": 0.09009731788673898,
+      "learning_rate": 2.4751822838542322e-05,
+      "loss": 0.369,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6614173228346458,
+      "grad_norm": 0.07933174886314172,
+      "learning_rate": 2.471364329532627e-05,
+      "loss": 0.3746,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6627296587926508,
+      "grad_norm": 0.08222410217069696,
+      "learning_rate": 2.4675464420051732e-05,
+      "loss": 0.3712,
+      "step": 1267
+    },
+    {
+      "epoch": 1.6640419947506562,
+      "grad_norm": 0.08184402539537622,
+      "learning_rate": 2.4637286301772862e-05,
+      "loss": 0.3692,
+      "step": 1268
+    },
+    {
+      "epoch": 1.6653543307086616,
+      "grad_norm": 0.08515457499365685,
+      "learning_rate": 2.4599109029542047e-05,
+      "loss": 0.3697,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.09379055134956528,
+      "learning_rate": 2.456093269240972e-05,
+      "loss": 0.3835,
+      "step": 1270
+    },
+    {
+      "epoch": 1.667979002624672,
+      "grad_norm": 0.07345299099181218,
+      "learning_rate": 2.4522757379424113e-05,
+      "loss": 0.373,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6692913385826773,
+      "grad_norm": 0.08960288728054544,
+      "learning_rate": 2.4484583179631073e-05,
+      "loss": 0.3656,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6706036745406823,
+      "grad_norm": 0.08717569982468386,
+      "learning_rate": 2.444641018207386e-05,
+      "loss": 0.3732,
+      "step": 1273
+    },
+    {
+      "epoch": 1.6719160104986877,
+      "grad_norm": 0.08724628549635252,
+      "learning_rate": 2.4408238475792916e-05,
+      "loss": 0.3825,
+      "step": 1274
+    },
+    {
+      "epoch": 1.673228346456693,
+      "grad_norm": 0.10996041908151194,
+      "learning_rate": 2.437006814982569e-05,
+      "loss": 0.376,
+      "step": 1275
+    },
+    {
+      "epoch": 1.674540682414698,
+      "grad_norm": 0.07748534738863194,
+      "learning_rate": 2.433189929320639e-05,
+      "loss": 0.372,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6758530183727034,
+      "grad_norm": 0.0840877427883847,
+      "learning_rate": 2.4293731994965807e-05,
+      "loss": 0.3729,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6771653543307088,
+      "grad_norm": 0.08056129478902133,
+      "learning_rate": 2.42555663441311e-05,
+      "loss": 0.373,
+      "step": 1278
+    },
+    {
+      "epoch": 1.678477690288714,
+      "grad_norm": 0.07847288854367504,
+      "learning_rate": 2.4217402429725583e-05,
+      "loss": 0.3793,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6797900262467191,
+      "grad_norm": 0.07742753174551084,
+      "learning_rate": 2.417924034076851e-05,
+      "loss": 0.3793,
+      "step": 1280
+    },
+    {
+      "epoch": 1.6811023622047245,
+      "grad_norm": 0.07747721491969639,
+      "learning_rate": 2.414108016627489e-05,
+      "loss": 0.3754,
+      "step": 1281
+    },
+    {
+      "epoch": 1.6824146981627297,
+      "grad_norm": 0.08006111241476797,
+      "learning_rate": 2.4102921995255263e-05,
+      "loss": 0.375,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6837270341207349,
+      "grad_norm": 0.07830181373731351,
+      "learning_rate": 2.4064765916715493e-05,
+      "loss": 0.385,
+      "step": 1283
+    },
+    {
+      "epoch": 1.6850393700787403,
+      "grad_norm": 0.07761661249553933,
+      "learning_rate": 2.4026612019656562e-05,
+      "loss": 0.3795,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6863517060367454,
+      "grad_norm": 0.07579883949991571,
+      "learning_rate": 2.398846039307437e-05,
+      "loss": 0.3681,
+      "step": 1285
+    },
+    {
+      "epoch": 1.6876640419947506,
+      "grad_norm": 0.0754139181977833,
+      "learning_rate": 2.395031112595951e-05,
+      "loss": 0.3709,
+      "step": 1286
+    },
+    {
+      "epoch": 1.688976377952756,
+      "grad_norm": 0.07801362723429854,
+      "learning_rate": 2.3912164307297095e-05,
+      "loss": 0.376,
+      "step": 1287
+    },
+    {
+      "epoch": 1.6902887139107612,
+      "grad_norm": 0.08263202214760902,
+      "learning_rate": 2.3874020026066497e-05,
+      "loss": 0.3787,
+      "step": 1288
+    },
+    {
+      "epoch": 1.6916010498687664,
+      "grad_norm": 0.07154159949307194,
+      "learning_rate": 2.3835878371241185e-05,
+      "loss": 0.3668,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6929133858267718,
+      "grad_norm": 0.08548998059083211,
+      "learning_rate": 2.3797739431788503e-05,
+      "loss": 0.3855,
+      "step": 1290
+    },
+    {
+      "epoch": 1.694225721784777,
+      "grad_norm": 0.08028266621289991,
+      "learning_rate": 2.375960329666946e-05,
+      "loss": 0.3733,
+      "step": 1291
+    },
+    {
+      "epoch": 1.695538057742782,
+      "grad_norm": 0.08783859983900946,
+      "learning_rate": 2.3721470054838517e-05,
+      "loss": 0.381,
+      "step": 1292
+    },
+    {
+      "epoch": 1.6968503937007875,
+      "grad_norm": 0.08117781733646746,
+      "learning_rate": 2.3683339795243385e-05,
+      "loss": 0.3674,
+      "step": 1293
+    },
+    {
+      "epoch": 1.6981627296587927,
+      "grad_norm": 0.07391562184023422,
+      "learning_rate": 2.3645212606824833e-05,
+      "loss": 0.3781,
+      "step": 1294
+    },
+    {
+      "epoch": 1.6994750656167978,
+      "grad_norm": 0.08039774104873211,
+      "learning_rate": 2.3607088578516463e-05,
+      "loss": 0.3786,
+      "step": 1295
+    },
+    {
+      "epoch": 1.7007874015748032,
+      "grad_norm": 0.07332826118214995,
+      "learning_rate": 2.3568967799244484e-05,
+      "loss": 0.3782,
+      "step": 1296
+    },
+    {
+      "epoch": 1.7020997375328084,
+      "grad_norm": 0.07531323166341349,
+      "learning_rate": 2.353085035792756e-05,
+      "loss": 0.3734,
+      "step": 1297
+    },
+    {
+      "epoch": 1.7034120734908136,
+      "grad_norm": 0.07576417853980534,
+      "learning_rate": 2.3492736343476547e-05,
+      "loss": 0.3653,
+      "step": 1298
+    },
+    {
+      "epoch": 1.704724409448819,
+      "grad_norm": 0.07913255586615861,
+      "learning_rate": 2.3454625844794316e-05,
+      "loss": 0.3773,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7060367454068242,
+      "grad_norm": 0.07031398577022911,
+      "learning_rate": 2.341651895077551e-05,
+      "loss": 0.3838,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7073490813648293,
+      "grad_norm": 0.07486239487503384,
+      "learning_rate": 2.3378415750306424e-05,
+      "loss": 0.3699,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7086614173228347,
+      "grad_norm": 0.07188844408731149,
+      "learning_rate": 2.3340316332264687e-05,
+      "loss": 0.3699,
+      "step": 1302
+    },
+    {
+      "epoch": 1.70997375328084,
+      "grad_norm": 0.07487726447740582,
+      "learning_rate": 2.3302220785519117e-05,
+      "loss": 0.382,
+      "step": 1303
+    },
+    {
+      "epoch": 1.711286089238845,
+      "grad_norm": 0.06902566948420971,
+      "learning_rate": 2.3264129198929515e-05,
+      "loss": 0.3703,
+      "step": 1304
+    },
+    {
+      "epoch": 1.7125984251968505,
+      "grad_norm": 0.06816822860994712,
+      "learning_rate": 2.322604166134642e-05,
+      "loss": 0.3756,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7139107611548556,
+      "grad_norm": 0.07741247423155766,
+      "learning_rate": 2.3187958261610955e-05,
+      "loss": 0.3736,
+      "step": 1306
+    },
+    {
+      "epoch": 1.7152230971128608,
+      "grad_norm": 0.06563040078549623,
+      "learning_rate": 2.314987908855457e-05,
+      "loss": 0.3667,
+      "step": 1307
+    },
+    {
+      "epoch": 1.7165354330708662,
+      "grad_norm": 0.07805978594462037,
+      "learning_rate": 2.3111804230998863e-05,
+      "loss": 0.372,
+      "step": 1308
+    },
+    {
+      "epoch": 1.7178477690288714,
+      "grad_norm": 0.07595346050020149,
+      "learning_rate": 2.3073733777755372e-05,
+      "loss": 0.3707,
+      "step": 1309
+    },
+    {
+      "epoch": 1.7191601049868765,
+      "grad_norm": 0.07623758891556112,
+      "learning_rate": 2.3035667817625356e-05,
+      "loss": 0.374,
+      "step": 1310
+    },
+    {
+      "epoch": 1.720472440944882,
+      "grad_norm": 0.09237031393203983,
+      "learning_rate": 2.2997606439399593e-05,
+      "loss": 0.3784,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7217847769028871,
+      "grad_norm": 0.08081281278482788,
+      "learning_rate": 2.2959549731858164e-05,
+      "loss": 0.3711,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7230971128608923,
+      "grad_norm": 0.11105939780817552,
+      "learning_rate": 2.2921497783770275e-05,
+      "loss": 0.3817,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7244094488188977,
+      "grad_norm": 0.08010715856121602,
+      "learning_rate": 2.2883450683894027e-05,
+      "loss": 0.3747,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7257217847769029,
+      "grad_norm": 0.10684589118915717,
+      "learning_rate": 2.2845408520976205e-05,
+      "loss": 0.3717,
+      "step": 1315
+    },
+    {
+      "epoch": 1.727034120734908,
+      "grad_norm": 0.07445182594676578,
+      "learning_rate": 2.2807371383752073e-05,
+      "loss": 0.37,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7283464566929134,
+      "grad_norm": 0.1018649105385139,
+      "learning_rate": 2.2769339360945176e-05,
+      "loss": 0.3772,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7296587926509186,
+      "grad_norm": 0.0816702865365565,
+      "learning_rate": 2.2731312541267145e-05,
+      "loss": 0.3737,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7309711286089238,
+      "grad_norm": 0.0823310275045568,
+      "learning_rate": 2.2693291013417453e-05,
+      "loss": 0.3713,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7322834645669292,
+      "grad_norm": 0.07743475240081751,
+      "learning_rate": 2.265527486608324e-05,
+      "loss": 0.3826,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7335958005249343,
+      "grad_norm": 0.09931263909236925,
+      "learning_rate": 2.261726418793909e-05,
+      "loss": 0.3791,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7349081364829395,
+      "grad_norm": 0.08585392905067733,
+      "learning_rate": 2.2579259067646847e-05,
+      "loss": 0.377,
+      "step": 1322
+    },
+    {
+      "epoch": 1.736220472440945,
+      "grad_norm": 0.09429519708910096,
+      "learning_rate": 2.2541259593855363e-05,
+      "loss": 0.375,
+      "step": 1323
+    },
+    {
+      "epoch": 1.73753280839895,
+      "grad_norm": 0.08511868473928784,
+      "learning_rate": 2.2503265855200337e-05,
+      "loss": 0.3698,
+      "step": 1324
+    },
+    {
+      "epoch": 1.7388451443569553,
+      "grad_norm": 0.08566532817459108,
+      "learning_rate": 2.2465277940304096e-05,
+      "loss": 0.3782,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7401574803149606,
+      "grad_norm": 0.08236261544752937,
+      "learning_rate": 2.242729593777537e-05,
+      "loss": 0.3749,
+      "step": 1326
+    },
+    {
+      "epoch": 1.741469816272966,
+      "grad_norm": 0.0810484140129979,
+      "learning_rate": 2.2389319936209092e-05,
+      "loss": 0.3763,
+      "step": 1327
+    },
+    {
+      "epoch": 1.742782152230971,
+      "grad_norm": 0.07890289109134699,
+      "learning_rate": 2.235135002418623e-05,
+      "loss": 0.3717,
+      "step": 1328
+    },
+    {
+      "epoch": 1.7440944881889764,
+      "grad_norm": 0.0957206464421913,
+      "learning_rate": 2.2313386290273506e-05,
+      "loss": 0.3783,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7454068241469818,
+      "grad_norm": 0.08902122052209789,
+      "learning_rate": 2.2275428823023263e-05,
+      "loss": 0.3823,
+      "step": 1330
+    },
+    {
+      "epoch": 1.7467191601049867,
+      "grad_norm": 0.09063279671668352,
+      "learning_rate": 2.2237477710973214e-05,
+      "loss": 0.3663,
+      "step": 1331
+    },
+    {
+      "epoch": 1.7480314960629921,
+      "grad_norm": 0.0737993014310086,
+      "learning_rate": 2.2199533042646242e-05,
+      "loss": 0.3712,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7493438320209975,
+      "grad_norm": 0.1065882303074127,
+      "learning_rate": 2.2161594906550214e-05,
+      "loss": 0.3676,
+      "step": 1333
+    },
+    {
+      "epoch": 1.7506561679790025,
+      "grad_norm": 0.07674876046656787,
+      "learning_rate": 2.2123663391177747e-05,
+      "loss": 0.3751,
+      "step": 1334
+    },
+    {
+      "epoch": 1.7519685039370079,
+      "grad_norm": 0.0997320585914179,
+      "learning_rate": 2.2085738585006024e-05,
+      "loss": 0.374,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7532808398950133,
+      "grad_norm": 0.07769112215776183,
+      "learning_rate": 2.204782057649657e-05,
+      "loss": 0.3703,
+      "step": 1336
+    },
+    {
+      "epoch": 1.7545931758530182,
+      "grad_norm": 0.08438297730479576,
+      "learning_rate": 2.200990945409507e-05,
+      "loss": 0.3728,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7559055118110236,
+      "grad_norm": 0.08309535347496295,
+      "learning_rate": 2.197200530623113e-05,
+      "loss": 0.3831,
+      "step": 1338
+    },
+    {
+      "epoch": 1.757217847769029,
+      "grad_norm": 0.07854679448058476,
+      "learning_rate": 2.1934108221318087e-05,
+      "loss": 0.3757,
+      "step": 1339
+    },
+    {
+      "epoch": 1.758530183727034,
+      "grad_norm": 0.09746662575933818,
+      "learning_rate": 2.189621828775282e-05,
+      "loss": 0.3806,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7598425196850394,
+      "grad_norm": 0.08265976769483324,
+      "learning_rate": 2.1858335593915504e-05,
+      "loss": 0.3818,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7611548556430447,
+      "grad_norm": 0.08785870371389862,
+      "learning_rate": 2.1820460228169462e-05,
+      "loss": 0.3825,
+      "step": 1342
+    },
+    {
+      "epoch": 1.76246719160105,
+      "grad_norm": 0.08520589643104393,
+      "learning_rate": 2.1782592278860887e-05,
+      "loss": 0.3853,
+      "step": 1343
+    },
+    {
+      "epoch": 1.763779527559055,
+      "grad_norm": 0.07776346033263665,
+      "learning_rate": 2.1744731834318685e-05,
+      "loss": 0.3625,
+      "step": 1344
+    },
+    {
+      "epoch": 1.7650918635170605,
+      "grad_norm": 0.08201078162938198,
+      "learning_rate": 2.1706878982854274e-05,
+      "loss": 0.3722,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7664041994750657,
+      "grad_norm": 0.12504325973170183,
+      "learning_rate": 2.166903381276133e-05,
+      "loss": 0.3737,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7677165354330708,
+      "grad_norm": 0.07790222624079127,
+      "learning_rate": 2.1631196412315638e-05,
+      "loss": 0.3687,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7690288713910762,
+      "grad_norm": 0.071899313008755,
+      "learning_rate": 2.159336686977484e-05,
+      "loss": 0.367,
+      "step": 1348
+    },
+    {
+      "epoch": 1.7703412073490814,
+      "grad_norm": 0.07306540792403066,
+      "learning_rate": 2.155554527337827e-05,
+      "loss": 0.3708,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7716535433070866,
+      "grad_norm": 0.07140365673272449,
+      "learning_rate": 2.1517731711346716e-05,
+      "loss": 0.366,
+      "step": 1350
+    },
+    {
+      "epoch": 1.772965879265092,
+      "grad_norm": 0.0732764885461719,
+      "learning_rate": 2.1479926271882213e-05,
+      "loss": 0.3759,
+      "step": 1351
+    },
+    {
+      "epoch": 1.7742782152230971,
+      "grad_norm": 0.07447716241725034,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 0.3713,
+      "step": 1352
+    },
+    {
+      "epoch": 1.7755905511811023,
+      "grad_norm": 0.0813104212076147,
+      "learning_rate": 2.140434011336765e-05,
+      "loss": 0.3714,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7769028871391077,
+      "grad_norm": 0.06557721595453007,
+      "learning_rate": 2.1366559570626114e-05,
+      "loss": 0.3714,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7782152230971129,
+      "grad_norm": 0.07548028247508365,
+      "learning_rate": 2.1328787503068317e-05,
+      "loss": 0.3716,
+      "step": 1355
+    },
+    {
+      "epoch": 1.779527559055118,
+      "grad_norm": 0.08149309681509104,
+      "learning_rate": 2.1291023998799506e-05,
+      "loss": 0.3744,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7808398950131235,
+      "grad_norm": 0.06988163973591179,
+      "learning_rate": 2.1253269145904976e-05,
+      "loss": 0.3766,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7821522309711286,
+      "grad_norm": 0.07974520007359683,
+      "learning_rate": 2.1215523032449825e-05,
+      "loss": 0.3827,
+      "step": 1358
+    },
+    {
+      "epoch": 1.7834645669291338,
+      "grad_norm": 0.0841985469606396,
+      "learning_rate": 2.1177785746478777e-05,
+      "loss": 0.3756,
+      "step": 1359
+    },
+    {
+      "epoch": 1.7847769028871392,
+      "grad_norm": 0.07747037350813625,
+      "learning_rate": 2.114005737601595e-05,
+      "loss": 0.3711,
+      "step": 1360
+    },
+    {
+      "epoch": 1.7860892388451444,
+      "grad_norm": 0.6265280618277046,
+      "learning_rate": 2.11023380090647e-05,
+      "loss": 0.3812,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7874015748031495,
+      "grad_norm": 0.0934255280423324,
+      "learning_rate": 2.106462773360734e-05,
+      "loss": 0.3717,
+      "step": 1362
+    },
+    {
+      "epoch": 1.788713910761155,
+      "grad_norm": 0.08606255787915999,
+      "learning_rate": 2.1026926637605008e-05,
+      "loss": 0.3757,
+      "step": 1363
+    },
+    {
+      "epoch": 1.79002624671916,
+      "grad_norm": 0.08608098994177468,
+      "learning_rate": 2.098923480899741e-05,
+      "loss": 0.385,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7913385826771653,
+      "grad_norm": 0.08114117346975096,
+      "learning_rate": 2.095155233570265e-05,
+      "loss": 0.3744,
+      "step": 1365
+    },
+    {
+      "epoch": 1.7926509186351707,
+      "grad_norm": 0.10552022404611737,
+      "learning_rate": 2.0913879305617008e-05,
+      "loss": 0.3878,
+      "step": 1366
+    },
+    {
+      "epoch": 1.7939632545931758,
+      "grad_norm": 0.10083953542027302,
+      "learning_rate": 2.0876215806614723e-05,
+      "loss": 0.3657,
+      "step": 1367
+    },
+    {
+      "epoch": 1.795275590551181,
+      "grad_norm": 0.08295259800189925,
+      "learning_rate": 2.0838561926547817e-05,
+      "loss": 0.3839,
+      "step": 1368
+    },
+    {
+      "epoch": 1.7965879265091864,
+      "grad_norm": 0.09273390042642027,
+      "learning_rate": 2.0800917753245877e-05,
+      "loss": 0.3816,
+      "step": 1369
+    },
+    {
+      "epoch": 1.7979002624671916,
+      "grad_norm": 0.07877308296524815,
+      "learning_rate": 2.0763283374515843e-05,
+      "loss": 0.3761,
+      "step": 1370
+    },
+    {
+      "epoch": 1.7992125984251968,
+      "grad_norm": 0.08295613919764896,
+      "learning_rate": 2.0725658878141803e-05,
+      "loss": 0.3832,
+      "step": 1371
+    },
+    {
+      "epoch": 1.8005249343832022,
+      "grad_norm": 0.07347289002768054,
+      "learning_rate": 2.06880443518848e-05,
+      "loss": 0.3781,
+      "step": 1372
+    },
+    {
+      "epoch": 1.8018372703412073,
+      "grad_norm": 0.08752219193208338,
+      "learning_rate": 2.0650439883482626e-05,
+      "loss": 0.3701,
+      "step": 1373
+    },
+    {
+      "epoch": 1.8031496062992125,
+      "grad_norm": 0.07639282473217078,
+      "learning_rate": 2.0612845560649603e-05,
+      "loss": 0.3701,
+      "step": 1374
+    },
+    {
+      "epoch": 1.804461942257218,
+      "grad_norm": 0.08629127746976152,
+      "learning_rate": 2.057526147107639e-05,
+      "loss": 0.3744,
+      "step": 1375
+    },
+    {
+      "epoch": 1.805774278215223,
+      "grad_norm": 0.06706434220288687,
+      "learning_rate": 2.0537687702429776e-05,
+      "loss": 0.3661,
+      "step": 1376
+    },
+    {
+      "epoch": 1.8070866141732282,
+      "grad_norm": 0.0774369348678832,
+      "learning_rate": 2.050012434235249e-05,
+      "loss": 0.3752,
+      "step": 1377
+    },
+    {
+      "epoch": 1.8083989501312336,
+      "grad_norm": 0.0807618836584794,
+      "learning_rate": 2.0462571478462962e-05,
+      "loss": 0.3749,
+      "step": 1378
+    },
+    {
+      "epoch": 1.8097112860892388,
+      "grad_norm": 0.07415598832647952,
+      "learning_rate": 2.042502919835514e-05,
+      "loss": 0.3831,
+      "step": 1379
+    },
+    {
+      "epoch": 1.811023622047244,
+      "grad_norm": 0.078138739387594,
+      "learning_rate": 2.0387497589598316e-05,
+      "loss": 0.3691,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8123359580052494,
+      "grad_norm": 0.07203136024551042,
+      "learning_rate": 2.034997673973685e-05,
+      "loss": 0.3765,
+      "step": 1381
+    },
+    {
+      "epoch": 1.8136482939632546,
+      "grad_norm": 0.08699091662849108,
+      "learning_rate": 2.0312466736290016e-05,
+      "loss": 0.3764,
+      "step": 1382
+    },
+    {
+      "epoch": 1.8149606299212597,
+      "grad_norm": 0.07703577033158868,
+      "learning_rate": 2.027496766675182e-05,
+      "loss": 0.3764,
+      "step": 1383
+    },
+    {
+      "epoch": 1.8162729658792651,
+      "grad_norm": 0.06967008017874017,
+      "learning_rate": 2.0237479618590723e-05,
+      "loss": 0.3778,
+      "step": 1384
+    },
+    {
+      "epoch": 1.8175853018372703,
+      "grad_norm": 0.0824496978699296,
+      "learning_rate": 2.02000026792495e-05,
+      "loss": 0.3701,
+      "step": 1385
+    },
+    {
+      "epoch": 1.8188976377952755,
+      "grad_norm": 0.07019715207611438,
+      "learning_rate": 2.016253693614501e-05,
+      "loss": 0.3737,
+      "step": 1386
+    },
+    {
+      "epoch": 1.8202099737532809,
+      "grad_norm": 0.08227058040906289,
+      "learning_rate": 2.0125082476667993e-05,
+      "loss": 0.377,
+      "step": 1387
+    },
+    {
+      "epoch": 1.821522309711286,
+      "grad_norm": 0.07402095552810128,
+      "learning_rate": 2.008763938818287e-05,
+      "loss": 0.3733,
+      "step": 1388
+    },
+    {
+      "epoch": 1.8228346456692912,
+      "grad_norm": 0.07566943983102979,
+      "learning_rate": 2.005020775802755e-05,
+      "loss": 0.3659,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8241469816272966,
+      "grad_norm": 0.06954620281810747,
+      "learning_rate": 2.0012787673513193e-05,
+      "loss": 0.373,
+      "step": 1390
+    },
+    {
+      "epoch": 1.825459317585302,
+      "grad_norm": 0.07737326240185356,
+      "learning_rate": 1.9975379221924043e-05,
+      "loss": 0.3736,
+      "step": 1391
+    },
+    {
+      "epoch": 1.826771653543307,
+      "grad_norm": 0.08198754549432477,
+      "learning_rate": 1.9937982490517212e-05,
+      "loss": 0.373,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8280839895013123,
+      "grad_norm": 0.10790414407489009,
+      "learning_rate": 1.9900597566522468e-05,
+      "loss": 0.3763,
+      "step": 1393
+    },
+    {
+      "epoch": 1.8293963254593177,
+      "grad_norm": 0.08015833774589441,
+      "learning_rate": 1.986322453714204e-05,
+      "loss": 0.3765,
+      "step": 1394
+    },
+    {
+      "epoch": 1.8307086614173227,
+      "grad_norm": 0.0781623289177399,
+      "learning_rate": 1.982586348955039e-05,
+      "loss": 0.3732,
+      "step": 1395
+    },
+    {
+      "epoch": 1.832020997375328,
+      "grad_norm": 0.06767072103856987,
+      "learning_rate": 1.9788514510894082e-05,
+      "loss": 0.3679,
+      "step": 1396
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.07293559442022682,
+      "learning_rate": 1.975117768829149e-05,
+      "loss": 0.3689,
+      "step": 1397
+    },
+    {
+      "epoch": 1.8346456692913384,
+      "grad_norm": 0.06788200629060834,
+      "learning_rate": 1.9713853108832644e-05,
+      "loss": 0.3747,
+      "step": 1398
+    },
+    {
+      "epoch": 1.8359580052493438,
+      "grad_norm": 0.07329711579916914,
+      "learning_rate": 1.9676540859579013e-05,
+      "loss": 0.3699,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8372703412073492,
+      "grad_norm": 0.06870097642001094,
+      "learning_rate": 1.9639241027563294e-05,
+      "loss": 0.3754,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8385826771653542,
+      "grad_norm": 0.07120934777419474,
+      "learning_rate": 1.9601953699789254e-05,
+      "loss": 0.3859,
+      "step": 1401
+    },
+    {
+      "epoch": 1.8398950131233596,
+      "grad_norm": 0.07325908312437746,
+      "learning_rate": 1.9564678963231467e-05,
+      "loss": 0.3736,
+      "step": 1402
+    },
+    {
+      "epoch": 1.841207349081365,
+      "grad_norm": 0.06933120362539917,
+      "learning_rate": 1.9527416904835132e-05,
+      "loss": 0.3698,
+      "step": 1403
+    },
+    {
+      "epoch": 1.84251968503937,
+      "grad_norm": 0.06756204419926463,
+      "learning_rate": 1.94901676115159e-05,
+      "loss": 0.3701,
+      "step": 1404
+    },
+    {
+      "epoch": 1.8438320209973753,
+      "grad_norm": 0.06740768925975776,
+      "learning_rate": 1.945293117015963e-05,
+      "loss": 0.3848,
+      "step": 1405
+    },
+    {
+      "epoch": 1.8451443569553807,
+      "grad_norm": 0.07194684301093532,
+      "learning_rate": 1.9415707667622198e-05,
+      "loss": 0.3788,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8464566929133859,
+      "grad_norm": 0.06661773543975859,
+      "learning_rate": 1.937849719072931e-05,
+      "loss": 0.3671,
+      "step": 1407
+    },
+    {
+      "epoch": 1.847769028871391,
+      "grad_norm": 0.08076469732706866,
+      "learning_rate": 1.9341299826276298e-05,
+      "loss": 0.3733,
+      "step": 1408
+    },
+    {
+      "epoch": 1.8490813648293964,
+      "grad_norm": 0.06798563236949896,
+      "learning_rate": 1.9304115661027878e-05,
+      "loss": 0.3695,
+      "step": 1409
+    },
+    {
+      "epoch": 1.8503937007874016,
+      "grad_norm": 0.08174705618398645,
+      "learning_rate": 1.926694478171802e-05,
+      "loss": 0.3715,
+      "step": 1410
+    },
+    {
+      "epoch": 1.8517060367454068,
+      "grad_norm": 0.07096602175178422,
+      "learning_rate": 1.9229787275049678e-05,
+      "loss": 0.375,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8530183727034122,
+      "grad_norm": 0.06678517731035415,
+      "learning_rate": 1.91926432276946e-05,
+      "loss": 0.3676,
+      "step": 1412
+    },
+    {
+      "epoch": 1.8543307086614174,
+      "grad_norm": 0.06445922538819156,
+      "learning_rate": 1.9155512726293174e-05,
+      "loss": 0.3722,
+      "step": 1413
+    },
+    {
+      "epoch": 1.8556430446194225,
+      "grad_norm": 0.06827685878280779,
+      "learning_rate": 1.911839585745417e-05,
+      "loss": 0.3684,
+      "step": 1414
+    },
+    {
+      "epoch": 1.856955380577428,
+      "grad_norm": 0.06811776523935577,
+      "learning_rate": 1.9081292707754552e-05,
+      "loss": 0.3701,
+      "step": 1415
+    },
+    {
+      "epoch": 1.858267716535433,
+      "grad_norm": 0.07263691954993988,
+      "learning_rate": 1.9044203363739313e-05,
+      "loss": 0.3796,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8595800524934383,
+      "grad_norm": 0.07433967654849472,
+      "learning_rate": 1.9007127911921214e-05,
+      "loss": 0.3656,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8608923884514437,
+      "grad_norm": 0.06984698129054637,
+      "learning_rate": 1.8970066438780628e-05,
+      "loss": 0.3665,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8622047244094488,
+      "grad_norm": 0.07100221997103474,
+      "learning_rate": 1.893301903076531e-05,
+      "loss": 0.3786,
+      "step": 1419
+    },
+    {
+      "epoch": 1.863517060367454,
+      "grad_norm": 0.07062998608860808,
+      "learning_rate": 1.889598577429022e-05,
+      "loss": 0.3772,
+      "step": 1420
+    },
+    {
+      "epoch": 1.8648293963254594,
+      "grad_norm": 0.07691360277245567,
+      "learning_rate": 1.8858966755737307e-05,
+      "loss": 0.3685,
+      "step": 1421
+    },
+    {
+      "epoch": 1.8661417322834646,
+      "grad_norm": 0.07432759754680626,
+      "learning_rate": 1.8821962061455293e-05,
+      "loss": 0.3749,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8674540682414698,
+      "grad_norm": 0.22087195930967818,
+      "learning_rate": 1.8784971777759518e-05,
+      "loss": 0.3787,
+      "step": 1423
+    },
+    {
+      "epoch": 1.8687664041994752,
+      "grad_norm": 0.0868039053616512,
+      "learning_rate": 1.8747995990931675e-05,
+      "loss": 0.3729,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8700787401574803,
+      "grad_norm": 0.07002606010649444,
+      "learning_rate": 1.8711034787219672e-05,
+      "loss": 0.3715,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8713910761154855,
+      "grad_norm": 0.09452352161096572,
+      "learning_rate": 1.867408825283738e-05,
+      "loss": 0.38,
+      "step": 1426
+    },
+    {
+      "epoch": 1.872703412073491,
+      "grad_norm": 0.07232270976951251,
+      "learning_rate": 1.8637156473964463e-05,
+      "loss": 0.381,
+      "step": 1427
+    },
+    {
+      "epoch": 1.874015748031496,
+      "grad_norm": 0.09422669839816034,
+      "learning_rate": 1.8600239536746163e-05,
+      "loss": 0.3789,
+      "step": 1428
+    },
+    {
+      "epoch": 1.8753280839895012,
+      "grad_norm": 0.07908161811883255,
+      "learning_rate": 1.856333752729311e-05,
+      "loss": 0.376,
+      "step": 1429
+    },
+    {
+      "epoch": 1.8766404199475066,
+      "grad_norm": 0.08789722733280858,
+      "learning_rate": 1.8526450531681107e-05,
+      "loss": 0.3774,
+      "step": 1430
+    },
+    {
+      "epoch": 1.8779527559055118,
+      "grad_norm": 0.08030136787827258,
+      "learning_rate": 1.8489578635950924e-05,
+      "loss": 0.3756,
+      "step": 1431
+    },
+    {
+      "epoch": 1.879265091863517,
+      "grad_norm": 0.08011762706699267,
+      "learning_rate": 1.8452721926108146e-05,
+      "loss": 0.3738,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8805774278215224,
+      "grad_norm": 0.0747249919831693,
+      "learning_rate": 1.8415880488122905e-05,
+      "loss": 0.3779,
+      "step": 1433
+    },
+    {
+      "epoch": 1.8818897637795275,
+      "grad_norm": 0.06769357394179615,
+      "learning_rate": 1.8379054407929715e-05,
+      "loss": 0.3707,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8832020997375327,
+      "grad_norm": 0.07088040945199324,
+      "learning_rate": 1.8342243771427284e-05,
+      "loss": 0.3743,
+      "step": 1435
+    },
+    {
+      "epoch": 1.8845144356955381,
+      "grad_norm": 0.06604620865684069,
+      "learning_rate": 1.8305448664478266e-05,
+      "loss": 0.3778,
+      "step": 1436
+    },
+    {
+      "epoch": 1.8858267716535433,
+      "grad_norm": 0.07070692448991929,
+      "learning_rate": 1.8268669172909137e-05,
+      "loss": 0.3801,
+      "step": 1437
+    },
+    {
+      "epoch": 1.8871391076115485,
+      "grad_norm": 0.07240940649965957,
+      "learning_rate": 1.8231905382509905e-05,
+      "loss": 0.3675,
+      "step": 1438
+    },
+    {
+      "epoch": 1.8884514435695539,
+      "grad_norm": 0.06868792471378343,
+      "learning_rate": 1.8195157379033973e-05,
+      "loss": 0.38,
+      "step": 1439
+    },
+    {
+      "epoch": 1.889763779527559,
+      "grad_norm": 0.5202792420422951,
+      "learning_rate": 1.815842524819793e-05,
+      "loss": 0.3742,
+      "step": 1440
+    },
+    {
+      "epoch": 1.8910761154855642,
+      "grad_norm": 0.07409138855659138,
+      "learning_rate": 1.812170907568132e-05,
+      "loss": 0.3757,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8923884514435696,
+      "grad_norm": 0.07543994480100888,
+      "learning_rate": 1.8085008947126474e-05,
+      "loss": 0.373,
+      "step": 1442
+    },
+    {
+      "epoch": 1.8937007874015748,
+      "grad_norm": 0.0779179892296895,
+      "learning_rate": 1.8048324948138303e-05,
+      "loss": 0.3711,
+      "step": 1443
+    },
+    {
+      "epoch": 1.89501312335958,
+      "grad_norm": 0.06948389975050286,
+      "learning_rate": 1.8011657164284095e-05,
+      "loss": 0.3693,
+      "step": 1444
+    },
+    {
+      "epoch": 1.8963254593175853,
+      "grad_norm": 0.07088013450067471,
+      "learning_rate": 1.797500568109331e-05,
+      "loss": 0.3739,
+      "step": 1445
+    },
+    {
+      "epoch": 1.8976377952755905,
+      "grad_norm": 0.06871257678767896,
+      "learning_rate": 1.7938370584057378e-05,
+      "loss": 0.3744,
+      "step": 1446
+    },
+    {
+      "epoch": 1.8989501312335957,
+      "grad_norm": 0.07698541248435493,
+      "learning_rate": 1.7901751958629533e-05,
+      "loss": 0.3739,
+      "step": 1447
+    },
+    {
+      "epoch": 1.900262467191601,
+      "grad_norm": 0.07549578339644179,
+      "learning_rate": 1.7865149890224563e-05,
+      "loss": 0.3837,
+      "step": 1448
+    },
+    {
+      "epoch": 1.9015748031496063,
+      "grad_norm": 0.07254914354263563,
+      "learning_rate": 1.782856446421865e-05,
+      "loss": 0.3837,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9028871391076114,
+      "grad_norm": 0.07192473624445811,
+      "learning_rate": 1.7791995765949138e-05,
+      "loss": 0.3717,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9041994750656168,
+      "grad_norm": 0.16674459755063153,
+      "learning_rate": 1.7755443880714383e-05,
+      "loss": 0.3798,
+      "step": 1451
+    },
+    {
+      "epoch": 1.905511811023622,
+      "grad_norm": 0.07855937113943308,
+      "learning_rate": 1.7718908893773505e-05,
+      "loss": 0.3801,
+      "step": 1452
+    },
+    {
+      "epoch": 1.9068241469816272,
+      "grad_norm": 0.07749522428709266,
+      "learning_rate": 1.7682390890346212e-05,
+      "loss": 0.3748,
+      "step": 1453
+    },
+    {
+      "epoch": 1.9081364829396326,
+      "grad_norm": 0.06918129629851116,
+      "learning_rate": 1.7645889955612593e-05,
+      "loss": 0.3736,
+      "step": 1454
+    },
+    {
+      "epoch": 1.909448818897638,
+      "grad_norm": 0.06658216407644153,
+      "learning_rate": 1.760940617471292e-05,
+      "loss": 0.3756,
+      "step": 1455
+    },
+    {
+      "epoch": 1.910761154855643,
+      "grad_norm": 0.0670434120188208,
+      "learning_rate": 1.7572939632747468e-05,
+      "loss": 0.3778,
+      "step": 1456
+    },
+    {
+      "epoch": 1.9120734908136483,
+      "grad_norm": 0.06559813925120576,
+      "learning_rate": 1.753649041477629e-05,
+      "loss": 0.3699,
+      "step": 1457
+    },
+    {
+      "epoch": 1.9133858267716537,
+      "grad_norm": 0.06775221245749306,
+      "learning_rate": 1.7500058605819025e-05,
+      "loss": 0.3838,
+      "step": 1458
+    },
+    {
+      "epoch": 1.9146981627296586,
+      "grad_norm": 0.06839370659978926,
+      "learning_rate": 1.7463644290854712e-05,
+      "loss": 0.3835,
+      "step": 1459
+    },
+    {
+      "epoch": 1.916010498687664,
+      "grad_norm": 0.0714727179486522,
+      "learning_rate": 1.7427247554821603e-05,
+      "loss": 0.368,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9173228346456694,
+      "grad_norm": 0.06468935032374366,
+      "learning_rate": 1.7390868482616902e-05,
+      "loss": 0.3735,
+      "step": 1461
+    },
+    {
+      "epoch": 1.9186351706036744,
+      "grad_norm": 0.07188768609811998,
+      "learning_rate": 1.7354507159096647e-05,
+      "loss": 0.3759,
+      "step": 1462
+    },
+    {
+      "epoch": 1.9199475065616798,
+      "grad_norm": 0.06797843498799402,
+      "learning_rate": 1.7318163669075467e-05,
+      "loss": 0.378,
+      "step": 1463
+    },
+    {
+      "epoch": 1.9212598425196852,
+      "grad_norm": 0.06799943401026322,
+      "learning_rate": 1.7281838097326388e-05,
+      "loss": 0.3694,
+      "step": 1464
+    },
+    {
+      "epoch": 1.9225721784776901,
+      "grad_norm": 0.07120767161336886,
+      "learning_rate": 1.7245530528580656e-05,
+      "loss": 0.3672,
+      "step": 1465
+    },
+    {
+      "epoch": 1.9238845144356955,
+      "grad_norm": 0.07179901349884693,
+      "learning_rate": 1.7209241047527503e-05,
+      "loss": 0.3622,
+      "step": 1466
+    },
+    {
+      "epoch": 1.925196850393701,
+      "grad_norm": 0.06923870385410505,
+      "learning_rate": 1.717296973881398e-05,
+      "loss": 0.3863,
+      "step": 1467
+    },
+    {
+      "epoch": 1.9265091863517059,
+      "grad_norm": 0.06924875465666393,
+      "learning_rate": 1.7136716687044756e-05,
+      "loss": 0.3737,
+      "step": 1468
+    },
+    {
+      "epoch": 1.9278215223097113,
+      "grad_norm": 0.11660004845275591,
+      "learning_rate": 1.710048197678191e-05,
+      "loss": 0.3971,
+      "step": 1469
+    },
+    {
+      "epoch": 1.9291338582677167,
+      "grad_norm": 0.07079485263247276,
+      "learning_rate": 1.7064265692544735e-05,
+      "loss": 0.3742,
+      "step": 1470
+    },
+    {
+      "epoch": 1.9304461942257218,
+      "grad_norm": 0.07282433227374298,
+      "learning_rate": 1.7028067918809537e-05,
+      "loss": 0.3719,
+      "step": 1471
+    },
+    {
+      "epoch": 1.931758530183727,
+      "grad_norm": 0.07343490898777169,
+      "learning_rate": 1.6991888740009477e-05,
+      "loss": 0.3747,
+      "step": 1472
+    },
+    {
+      "epoch": 1.9330708661417324,
+      "grad_norm": 0.0837445276170927,
+      "learning_rate": 1.6955728240534305e-05,
+      "loss": 0.3703,
+      "step": 1473
+    },
+    {
+      "epoch": 1.9343832020997376,
+      "grad_norm": 0.07377731463596557,
+      "learning_rate": 1.6919586504730218e-05,
+      "loss": 0.3687,
+      "step": 1474
+    },
+    {
+      "epoch": 1.9356955380577427,
+      "grad_norm": 0.10252401567856152,
+      "learning_rate": 1.6883463616899653e-05,
+      "loss": 0.3785,
+      "step": 1475
+    },
+    {
+      "epoch": 1.9370078740157481,
+      "grad_norm": 0.07794839536781521,
+      "learning_rate": 1.6847359661301067e-05,
+      "loss": 0.3788,
+      "step": 1476
+    },
+    {
+      "epoch": 1.9383202099737533,
+      "grad_norm": 0.09243723553533967,
+      "learning_rate": 1.6811274722148756e-05,
+      "loss": 0.3734,
+      "step": 1477
+    },
+    {
+      "epoch": 1.9396325459317585,
+      "grad_norm": 0.07379465195371976,
+      "learning_rate": 1.6775208883612692e-05,
+      "loss": 0.3763,
+      "step": 1478
+    },
+    {
+      "epoch": 1.9409448818897639,
+      "grad_norm": 0.08158232911155025,
+      "learning_rate": 1.6739162229818246e-05,
+      "loss": 0.3741,
+      "step": 1479
+    },
+    {
+      "epoch": 1.942257217847769,
+      "grad_norm": 0.0707975908754488,
+      "learning_rate": 1.6703134844846073e-05,
+      "loss": 0.3695,
+      "step": 1480
+    },
+    {
+      "epoch": 1.9435695538057742,
+      "grad_norm": 0.07696163896803351,
+      "learning_rate": 1.6667126812731877e-05,
+      "loss": 0.3753,
+      "step": 1481
+    },
+    {
+      "epoch": 1.9448818897637796,
+      "grad_norm": 0.07619971099960819,
+      "learning_rate": 1.6631138217466203e-05,
+      "loss": 0.3759,
+      "step": 1482
+    },
+    {
+      "epoch": 1.9461942257217848,
+      "grad_norm": 0.07140622798039638,
+      "learning_rate": 1.659516914299427e-05,
+      "loss": 0.3779,
+      "step": 1483
+    },
+    {
+      "epoch": 1.94750656167979,
+      "grad_norm": 0.09460980957917928,
+      "learning_rate": 1.6559219673215784e-05,
+      "loss": 0.381,
+      "step": 1484
+    },
+    {
+      "epoch": 1.9488188976377954,
+      "grad_norm": 0.0693242646302972,
+      "learning_rate": 1.652328989198469e-05,
+      "loss": 0.3793,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9501312335958005,
+      "grad_norm": 0.07679266906219018,
+      "learning_rate": 1.6487379883109012e-05,
+      "loss": 0.3643,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9514435695538057,
+      "grad_norm": 0.07815201814941125,
+      "learning_rate": 1.6451489730350682e-05,
+      "loss": 0.3673,
+      "step": 1487
+    },
+    {
+      "epoch": 1.952755905511811,
+      "grad_norm": 0.0698634633136468,
+      "learning_rate": 1.6415619517425296e-05,
+      "loss": 0.3781,
+      "step": 1488
+    },
+    {
+      "epoch": 1.9540682414698163,
+      "grad_norm": 0.07494775384883079,
+      "learning_rate": 1.637976932800193e-05,
+      "loss": 0.379,
+      "step": 1489
+    },
+    {
+      "epoch": 1.9553805774278215,
+      "grad_norm": 0.06975829351029696,
+      "learning_rate": 1.6343939245702965e-05,
+      "loss": 0.3736,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9566929133858268,
+      "grad_norm": 0.06582980790912202,
+      "learning_rate": 1.6308129354103885e-05,
+      "loss": 0.3717,
+      "step": 1491
+    },
+    {
+      "epoch": 1.958005249343832,
+      "grad_norm": 0.07540709062945551,
+      "learning_rate": 1.627233973673309e-05,
+      "loss": 0.3618,
+      "step": 1492
+    },
+    {
+      "epoch": 1.9593175853018372,
+      "grad_norm": 0.06555160898183686,
+      "learning_rate": 1.623657047707166e-05,
+      "loss": 0.3664,
+      "step": 1493
+    },
+    {
+      "epoch": 1.9606299212598426,
+      "grad_norm": 0.07035215036150988,
+      "learning_rate": 1.6200821658553205e-05,
+      "loss": 0.3782,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9619422572178478,
+      "grad_norm": 0.06526383465022488,
+      "learning_rate": 1.6165093364563654e-05,
+      "loss": 0.3683,
+      "step": 1495
+    },
+    {
+      "epoch": 1.963254593175853,
+      "grad_norm": 0.0790789721608903,
+      "learning_rate": 1.6129385678441062e-05,
+      "loss": 0.3747,
+      "step": 1496
+    },
+    {
+      "epoch": 1.9645669291338583,
+      "grad_norm": 0.06920402884881423,
+      "learning_rate": 1.6093698683475418e-05,
+      "loss": 0.3722,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9658792650918635,
+      "grad_norm": 0.07100418773464605,
+      "learning_rate": 1.6058032462908434e-05,
+      "loss": 0.3848,
+      "step": 1498
+    },
+    {
+      "epoch": 1.9671916010498687,
+      "grad_norm": 0.07069789787571622,
+      "learning_rate": 1.602238709993339e-05,
+      "loss": 0.3743,
+      "step": 1499
+    },
+    {
+      "epoch": 1.968503937007874,
+      "grad_norm": 0.06496172272798555,
+      "learning_rate": 1.598676267769489e-05,
+      "loss": 0.3759,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9698162729658792,
+      "grad_norm": 0.07864730127628033,
+      "learning_rate": 1.5951159279288703e-05,
+      "loss": 0.3725,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9711286089238844,
+      "grad_norm": 0.07119986876469131,
+      "learning_rate": 1.591557698776155e-05,
+      "loss": 0.3752,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9724409448818898,
+      "grad_norm": 0.07559015372198125,
+      "learning_rate": 1.5880015886110942e-05,
+      "loss": 0.381,
+      "step": 1503
+    },
+    {
+      "epoch": 1.973753280839895,
+      "grad_norm": 0.06905642392784156,
+      "learning_rate": 1.5844476057284935e-05,
+      "loss": 0.3723,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9750656167979002,
+      "grad_norm": 0.06838448984316826,
+      "learning_rate": 1.5808957584181998e-05,
+      "loss": 0.3675,
+      "step": 1505
+    },
+    {
+      "epoch": 1.9763779527559056,
+      "grad_norm": 0.0738342037375548,
+      "learning_rate": 1.5773460549650753e-05,
+      "loss": 0.3719,
+      "step": 1506
+    },
+    {
+      "epoch": 1.9776902887139107,
+      "grad_norm": 0.0664241968847619,
+      "learning_rate": 1.5737985036489827e-05,
+      "loss": 0.3752,
+      "step": 1507
+    },
+    {
+      "epoch": 1.979002624671916,
+      "grad_norm": 0.0659536107451796,
+      "learning_rate": 1.5702531127447663e-05,
+      "loss": 0.372,
+      "step": 1508
+    },
+    {
+      "epoch": 1.9803149606299213,
+      "grad_norm": 0.06986175562397576,
+      "learning_rate": 1.566709890522229e-05,
+      "loss": 0.3749,
+      "step": 1509
+    },
+    {
+      "epoch": 1.9816272965879265,
+      "grad_norm": 0.0718556286592205,
+      "learning_rate": 1.5631688452461157e-05,
+      "loss": 0.3658,
+      "step": 1510
+    },
+    {
+      "epoch": 1.9829396325459316,
+      "grad_norm": 0.0692515157011358,
+      "learning_rate": 1.559629985176095e-05,
+      "loss": 0.3795,
+      "step": 1511
+    },
+    {
+      "epoch": 1.984251968503937,
+      "grad_norm": 0.07766796262738447,
+      "learning_rate": 1.556093318566736e-05,
+      "loss": 0.3763,
+      "step": 1512
+    },
+    {
+      "epoch": 1.9855643044619422,
+      "grad_norm": 0.06394995161446897,
+      "learning_rate": 1.5525588536674935e-05,
+      "loss": 0.3778,
+      "step": 1513
+    },
+    {
+      "epoch": 1.9868766404199474,
+      "grad_norm": 0.07895268316336373,
+      "learning_rate": 1.549026598722684e-05,
+      "loss": 0.3777,
+      "step": 1514
+    },
+    {
+      "epoch": 1.9881889763779528,
+      "grad_norm": 0.06636902964147873,
+      "learning_rate": 1.545496561971473e-05,
+      "loss": 0.3703,
+      "step": 1515
+    },
+    {
+      "epoch": 1.989501312335958,
+      "grad_norm": 0.07403454627064664,
+      "learning_rate": 1.5419687516478494e-05,
+      "loss": 0.3677,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9908136482939631,
+      "grad_norm": 0.0657179049263881,
+      "learning_rate": 1.5384431759806083e-05,
+      "loss": 0.373,
+      "step": 1517
+    },
+    {
+      "epoch": 1.9921259842519685,
+      "grad_norm": 0.07859523314189777,
+      "learning_rate": 1.534919843193334e-05,
+      "loss": 0.3839,
+      "step": 1518
+    },
+    {
+      "epoch": 1.993438320209974,
+      "grad_norm": 0.0652447425786888,
+      "learning_rate": 1.5313987615043797e-05,
+      "loss": 0.3732,
+      "step": 1519
+    },
+    {
+      "epoch": 1.9947506561679789,
+      "grad_norm": 0.07116332706356729,
+      "learning_rate": 1.5278799391268456e-05,
+      "loss": 0.3742,
+      "step": 1520
+    },
+    {
+      "epoch": 1.9960629921259843,
+      "grad_norm": 0.06793069415863877,
+      "learning_rate": 1.5243633842685635e-05,
+      "loss": 0.3713,
+      "step": 1521
+    },
+    {
+      "epoch": 1.9973753280839897,
+      "grad_norm": 0.0753073459914751,
+      "learning_rate": 1.5208491051320744e-05,
+      "loss": 0.3767,
+      "step": 1522
+    },
+    {
+      "epoch": 1.9986876640419946,
+      "grad_norm": 0.06675173348238607,
+      "learning_rate": 1.5173371099146138e-05,
+      "loss": 0.3729,
+      "step": 1523
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.09579498674661296,
+      "learning_rate": 1.5138274068080879e-05,
+      "loss": 0.3772,
+      "step": 1524
+    },
+    {
+      "epoch": 2.0013123359580054,
+      "grad_norm": 0.0827592271985169,
+      "learning_rate": 1.5103200039990566e-05,
+      "loss": 0.3551,
+      "step": 1525
+    },
+    {
+      "epoch": 2.0026246719160103,
+      "grad_norm": 0.08423905201765781,
+      "learning_rate": 1.506814909668714e-05,
+      "loss": 0.3557,
+      "step": 1526
+    },
+    {
+      "epoch": 2.0039370078740157,
+      "grad_norm": 0.1386154415632436,
+      "learning_rate": 1.5033121319928714e-05,
+      "loss": 0.3552,
+      "step": 1527
+    },
+    {
+      "epoch": 2.005249343832021,
+      "grad_norm": 0.0787190234275724,
+      "learning_rate": 1.4998116791419342e-05,
+      "loss": 0.3565,
+      "step": 1528
+    },
+    {
+      "epoch": 2.006561679790026,
+      "grad_norm": 0.0783698126684778,
+      "learning_rate": 1.4963135592808852e-05,
+      "loss": 0.3489,
+      "step": 1529
+    },
+    {
+      "epoch": 2.0078740157480315,
+      "grad_norm": 0.07816210063089371,
+      "learning_rate": 1.4928177805692678e-05,
+      "loss": 0.3501,
+      "step": 1530
+    },
+    {
+      "epoch": 2.009186351706037,
+      "grad_norm": 0.24117808164266982,
+      "learning_rate": 1.4893243511611604e-05,
+      "loss": 0.3655,
+      "step": 1531
+    },
+    {
+      "epoch": 2.010498687664042,
+      "grad_norm": 0.07891421191533145,
+      "learning_rate": 1.4858332792051663e-05,
+      "loss": 0.3513,
+      "step": 1532
+    },
+    {
+      "epoch": 2.0118110236220472,
+      "grad_norm": 0.08622476687137787,
+      "learning_rate": 1.4823445728443868e-05,
+      "loss": 0.3561,
+      "step": 1533
+    },
+    {
+      "epoch": 2.0131233595800526,
+      "grad_norm": 0.07886293630554841,
+      "learning_rate": 1.4788582402164052e-05,
+      "loss": 0.3596,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0144356955380576,
+      "grad_norm": 0.08472329069616033,
+      "learning_rate": 1.4753742894532702e-05,
+      "loss": 0.357,
+      "step": 1535
+    },
+    {
+      "epoch": 2.015748031496063,
+      "grad_norm": 0.0665257963828778,
+      "learning_rate": 1.4718927286814727e-05,
+      "loss": 0.3623,
+      "step": 1536
+    },
+    {
+      "epoch": 2.0170603674540684,
+      "grad_norm": 0.08225386302913881,
+      "learning_rate": 1.4684135660219295e-05,
+      "loss": 0.3536,
+      "step": 1537
+    },
+    {
+      "epoch": 2.0183727034120733,
+      "grad_norm": 0.07146260739547013,
+      "learning_rate": 1.464936809589963e-05,
+      "loss": 0.3595,
+      "step": 1538
+    },
+    {
+      "epoch": 2.0196850393700787,
+      "grad_norm": 0.08457755257020916,
+      "learning_rate": 1.4614624674952842e-05,
+      "loss": 0.3636,
+      "step": 1539
+    },
+    {
+      "epoch": 2.020997375328084,
+      "grad_norm": 0.0753958573401714,
+      "learning_rate": 1.4579905478419714e-05,
+      "loss": 0.3752,
+      "step": 1540
+    },
+    {
+      "epoch": 2.022309711286089,
+      "grad_norm": 0.07257210083053495,
+      "learning_rate": 1.4545210587284541e-05,
+      "loss": 0.3621,
+      "step": 1541
+    },
+    {
+      "epoch": 2.0236220472440944,
+      "grad_norm": 0.0772202097015638,
+      "learning_rate": 1.4510540082474906e-05,
+      "loss": 0.3612,
+      "step": 1542
+    },
+    {
+      "epoch": 2.0249343832021,
+      "grad_norm": 0.06876241630813322,
+      "learning_rate": 1.4475894044861513e-05,
+      "loss": 0.3617,
+      "step": 1543
+    },
+    {
+      "epoch": 2.026246719160105,
+      "grad_norm": 0.0742810565391507,
+      "learning_rate": 1.4441272555258e-05,
+      "loss": 0.3593,
+      "step": 1544
+    },
+    {
+      "epoch": 2.02755905511811,
+      "grad_norm": 0.06823779578046398,
+      "learning_rate": 1.440667569442074e-05,
+      "loss": 0.3469,
+      "step": 1545
+    },
+    {
+      "epoch": 2.0288713910761156,
+      "grad_norm": 0.07403804789247229,
+      "learning_rate": 1.4372103543048673e-05,
+      "loss": 0.3544,
+      "step": 1546
+    },
+    {
+      "epoch": 2.0301837270341205,
+      "grad_norm": 0.07111451786859072,
+      "learning_rate": 1.433755618178309e-05,
+      "loss": 0.3626,
+      "step": 1547
+    },
+    {
+      "epoch": 2.031496062992126,
+      "grad_norm": 0.07544957326917047,
+      "learning_rate": 1.4303033691207446e-05,
+      "loss": 0.3561,
+      "step": 1548
+    },
+    {
+      "epoch": 2.0328083989501313,
+      "grad_norm": 0.07269199256522595,
+      "learning_rate": 1.4268536151847218e-05,
+      "loss": 0.3635,
+      "step": 1549
+    },
+    {
+      "epoch": 2.0341207349081363,
+      "grad_norm": 0.06887855344373077,
+      "learning_rate": 1.4234063644169654e-05,
+      "loss": 0.3478,
+      "step": 1550
+    },
+    {
+      "epoch": 2.0354330708661417,
+      "grad_norm": 0.07046389975959205,
+      "learning_rate": 1.4199616248583622e-05,
+      "loss": 0.3545,
+      "step": 1551
+    },
+    {
+      "epoch": 2.036745406824147,
+      "grad_norm": 0.0662336056157472,
+      "learning_rate": 1.4165194045439418e-05,
+      "loss": 0.3655,
+      "step": 1552
+    },
+    {
+      "epoch": 2.038057742782152,
+      "grad_norm": 0.05990479963884375,
+      "learning_rate": 1.413079711502857e-05,
+      "loss": 0.356,
+      "step": 1553
+    },
+    {
+      "epoch": 2.0393700787401574,
+      "grad_norm": 0.0748297170499018,
+      "learning_rate": 1.4096425537583662e-05,
+      "loss": 0.3561,
+      "step": 1554
+    },
+    {
+      "epoch": 2.040682414698163,
+      "grad_norm": 0.06236151160161714,
+      "learning_rate": 1.4062079393278127e-05,
+      "loss": 0.3597,
+      "step": 1555
+    },
+    {
+      "epoch": 2.041994750656168,
+      "grad_norm": 0.07080408614783967,
+      "learning_rate": 1.402775876222611e-05,
+      "loss": 0.362,
+      "step": 1556
+    },
+    {
+      "epoch": 2.043307086614173,
+      "grad_norm": 0.057701242736967785,
+      "learning_rate": 1.3993463724482202e-05,
+      "loss": 0.3563,
+      "step": 1557
+    },
+    {
+      "epoch": 2.0446194225721785,
+      "grad_norm": 0.06678885263331123,
+      "learning_rate": 1.3959194360041319e-05,
+      "loss": 0.3547,
+      "step": 1558
+    },
+    {
+      "epoch": 2.045931758530184,
+      "grad_norm": 0.06077058137732822,
+      "learning_rate": 1.392495074883848e-05,
+      "loss": 0.3542,
+      "step": 1559
+    },
+    {
+      "epoch": 2.047244094488189,
+      "grad_norm": 0.05860709080723558,
+      "learning_rate": 1.389073297074866e-05,
+      "loss": 0.3557,
+      "step": 1560
+    },
+    {
+      "epoch": 2.0485564304461943,
+      "grad_norm": 0.0673590978083707,
+      "learning_rate": 1.3856541105586545e-05,
+      "loss": 0.3532,
+      "step": 1561
+    },
+    {
+      "epoch": 2.0498687664041997,
+      "grad_norm": 0.06285387805348613,
+      "learning_rate": 1.3822375233106396e-05,
+      "loss": 0.3462,
+      "step": 1562
+    },
+    {
+      "epoch": 2.0511811023622046,
+      "grad_norm": 0.062660968430688,
+      "learning_rate": 1.3788235433001839e-05,
+      "loss": 0.3564,
+      "step": 1563
+    },
+    {
+      "epoch": 2.05249343832021,
+      "grad_norm": 0.06571645868553211,
+      "learning_rate": 1.3754121784905678e-05,
+      "loss": 0.3595,
+      "step": 1564
+    },
+    {
+      "epoch": 2.0538057742782154,
+      "grad_norm": 0.06428663892945402,
+      "learning_rate": 1.3720034368389737e-05,
+      "loss": 0.3623,
+      "step": 1565
+    },
+    {
+      "epoch": 2.0551181102362204,
+      "grad_norm": 0.06998915703160963,
+      "learning_rate": 1.3685973262964635e-05,
+      "loss": 0.3577,
+      "step": 1566
+    },
+    {
+      "epoch": 2.0564304461942258,
+      "grad_norm": 0.0687506043548039,
+      "learning_rate": 1.3651938548079629e-05,
+      "loss": 0.3639,
+      "step": 1567
+    },
+    {
+      "epoch": 2.057742782152231,
+      "grad_norm": 0.06637641566245839,
+      "learning_rate": 1.3617930303122406e-05,
+      "loss": 0.3617,
+      "step": 1568
+    },
+    {
+      "epoch": 2.059055118110236,
+      "grad_norm": 0.06651285424178163,
+      "learning_rate": 1.3583948607418926e-05,
+      "loss": 0.3549,
+      "step": 1569
+    },
+    {
+      "epoch": 2.0603674540682415,
+      "grad_norm": 0.06360845250238177,
+      "learning_rate": 1.3549993540233207e-05,
+      "loss": 0.3545,
+      "step": 1570
+    },
+    {
+      "epoch": 2.061679790026247,
+      "grad_norm": 0.06695885569006574,
+      "learning_rate": 1.3516065180767162e-05,
+      "loss": 0.3487,
+      "step": 1571
+    },
+    {
+      "epoch": 2.062992125984252,
+      "grad_norm": 0.0625538164546679,
+      "learning_rate": 1.3482163608160408e-05,
+      "loss": 0.3519,
+      "step": 1572
+    },
+    {
+      "epoch": 2.0643044619422573,
+      "grad_norm": 0.06902758034084451,
+      "learning_rate": 1.3448288901490095e-05,
+      "loss": 0.3546,
+      "step": 1573
+    },
+    {
+      "epoch": 2.0656167979002626,
+      "grad_norm": 0.05946897390780854,
+      "learning_rate": 1.3414441139770676e-05,
+      "loss": 0.3663,
+      "step": 1574
+    },
+    {
+      "epoch": 2.0669291338582676,
+      "grad_norm": 0.07058605036972201,
+      "learning_rate": 1.3380620401953773e-05,
+      "loss": 0.3587,
+      "step": 1575
+    },
+    {
+      "epoch": 2.068241469816273,
+      "grad_norm": 0.06381501421194796,
+      "learning_rate": 1.3346826766927968e-05,
+      "loss": 0.3576,
+      "step": 1576
+    },
+    {
+      "epoch": 2.0695538057742784,
+      "grad_norm": 0.06283831102008329,
+      "learning_rate": 1.3313060313518622e-05,
+      "loss": 0.3607,
+      "step": 1577
+    },
+    {
+      "epoch": 2.0708661417322833,
+      "grad_norm": 0.059482274466446924,
+      "learning_rate": 1.3279321120487698e-05,
+      "loss": 0.3513,
+      "step": 1578
+    },
+    {
+      "epoch": 2.0721784776902887,
+      "grad_norm": 0.06065959480315156,
+      "learning_rate": 1.3245609266533567e-05,
+      "loss": 0.3602,
+      "step": 1579
+    },
+    {
+      "epoch": 2.073490813648294,
+      "grad_norm": 0.060350112568170036,
+      "learning_rate": 1.321192483029085e-05,
+      "loss": 0.3597,
+      "step": 1580
+    },
+    {
+      "epoch": 2.074803149606299,
+      "grad_norm": 0.05864468588177573,
+      "learning_rate": 1.3178267890330192e-05,
+      "loss": 0.3555,
+      "step": 1581
+    },
+    {
+      "epoch": 2.0761154855643045,
+      "grad_norm": 0.05752631428038145,
+      "learning_rate": 1.3144638525158107e-05,
+      "loss": 0.3574,
+      "step": 1582
+    },
+    {
+      "epoch": 2.07742782152231,
+      "grad_norm": 0.12856496850798016,
+      "learning_rate": 1.3111036813216792e-05,
+      "loss": 0.3624,
+      "step": 1583
+    },
+    {
+      "epoch": 2.078740157480315,
+      "grad_norm": 0.057271145407274314,
+      "learning_rate": 1.307746283288394e-05,
+      "loss": 0.358,
+      "step": 1584
+    },
+    {
+      "epoch": 2.08005249343832,
+      "grad_norm": 0.07101777181476644,
+      "learning_rate": 1.3043916662472567e-05,
+      "loss": 0.3661,
+      "step": 1585
+    },
+    {
+      "epoch": 2.0813648293963256,
+      "grad_norm": 0.05922230169966975,
+      "learning_rate": 1.3010398380230792e-05,
+      "loss": 0.3559,
+      "step": 1586
+    },
+    {
+      "epoch": 2.0826771653543306,
+      "grad_norm": 0.06845845819789113,
+      "learning_rate": 1.2976908064341734e-05,
+      "loss": 0.3534,
+      "step": 1587
+    },
+    {
+      "epoch": 2.083989501312336,
+      "grad_norm": 0.062353436370822246,
+      "learning_rate": 1.2943445792923226e-05,
+      "loss": 0.355,
+      "step": 1588
+    },
+    {
+      "epoch": 2.0853018372703414,
+      "grad_norm": 0.07323653283550258,
+      "learning_rate": 1.291001164402773e-05,
+      "loss": 0.3541,
+      "step": 1589
+    },
+    {
+      "epoch": 2.0866141732283463,
+      "grad_norm": 0.057896442281030636,
+      "learning_rate": 1.2876605695642086e-05,
+      "loss": 0.3593,
+      "step": 1590
+    },
+    {
+      "epoch": 2.0879265091863517,
+      "grad_norm": 0.068891463156724,
+      "learning_rate": 1.284322802568735e-05,
+      "loss": 0.3686,
+      "step": 1591
+    },
+    {
+      "epoch": 2.089238845144357,
+      "grad_norm": 0.06245693400141888,
+      "learning_rate": 1.2809878712018635e-05,
+      "loss": 0.36,
+      "step": 1592
+    },
+    {
+      "epoch": 2.090551181102362,
+      "grad_norm": 0.06391499581866825,
+      "learning_rate": 1.2776557832424902e-05,
+      "loss": 0.3483,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0918635170603674,
+      "grad_norm": 1.1363851782164776,
+      "learning_rate": 1.2743265464628786e-05,
+      "loss": 0.385,
+      "step": 1594
+    },
+    {
+      "epoch": 2.093175853018373,
+      "grad_norm": 0.07001097223355543,
+      "learning_rate": 1.2710001686286412e-05,
+      "loss": 0.3582,
+      "step": 1595
+    },
+    {
+      "epoch": 2.094488188976378,
+      "grad_norm": 0.061774834907104906,
+      "learning_rate": 1.2676766574987248e-05,
+      "loss": 0.3597,
+      "step": 1596
+    },
+    {
+      "epoch": 2.095800524934383,
+      "grad_norm": 0.06232340825317771,
+      "learning_rate": 1.2643560208253863e-05,
+      "loss": 0.3605,
+      "step": 1597
+    },
+    {
+      "epoch": 2.0971128608923886,
+      "grad_norm": 0.06677228570149688,
+      "learning_rate": 1.2610382663541787e-05,
+      "loss": 0.3578,
+      "step": 1598
+    },
+    {
+      "epoch": 2.0984251968503935,
+      "grad_norm": 0.0603029302943236,
+      "learning_rate": 1.2577234018239325e-05,
+      "loss": 0.3588,
+      "step": 1599
+    },
+    {
+      "epoch": 2.099737532808399,
+      "grad_norm": 0.06021574117839872,
+      "learning_rate": 1.2544114349667358e-05,
+      "loss": 0.3576,
+      "step": 1600
+    },
+    {
+      "epoch": 2.1010498687664043,
+      "grad_norm": 0.06296379282093427,
+      "learning_rate": 1.2511023735079207e-05,
+      "loss": 0.3617,
+      "step": 1601
+    },
+    {
+      "epoch": 2.1023622047244093,
+      "grad_norm": 0.05727256100712065,
+      "learning_rate": 1.24779622516604e-05,
+      "loss": 0.3456,
+      "step": 1602
+    },
+    {
+      "epoch": 2.1036745406824147,
+      "grad_norm": 0.06019486117102856,
+      "learning_rate": 1.2444929976528521e-05,
+      "loss": 0.3557,
+      "step": 1603
+    },
+    {
+      "epoch": 2.10498687664042,
+      "grad_norm": 0.0683965201289296,
+      "learning_rate": 1.241192698673301e-05,
+      "loss": 0.363,
+      "step": 1604
+    },
+    {
+      "epoch": 2.106299212598425,
+      "grad_norm": 0.059485617129202355,
+      "learning_rate": 1.2378953359255036e-05,
+      "loss": 0.3572,
+      "step": 1605
+    },
+    {
+      "epoch": 2.1076115485564304,
+      "grad_norm": 0.12546545660589056,
+      "learning_rate": 1.2346009171007243e-05,
+      "loss": 0.3647,
+      "step": 1606
+    },
+    {
+      "epoch": 2.108923884514436,
+      "grad_norm": 0.06318988490693984,
+      "learning_rate": 1.2313094498833611e-05,
+      "loss": 0.3638,
+      "step": 1607
+    },
+    {
+      "epoch": 2.1102362204724407,
+      "grad_norm": 0.06123396477677397,
+      "learning_rate": 1.2280209419509289e-05,
+      "loss": 0.3474,
+      "step": 1608
+    },
+    {
+      "epoch": 2.111548556430446,
+      "grad_norm": 0.06400854662980139,
+      "learning_rate": 1.2247354009740384e-05,
+      "loss": 0.3558,
+      "step": 1609
+    },
+    {
+      "epoch": 2.1128608923884515,
+      "grad_norm": 0.059322117580979467,
+      "learning_rate": 1.22145283461638e-05,
+      "loss": 0.3666,
+      "step": 1610
+    },
+    {
+      "epoch": 2.1141732283464565,
+      "grad_norm": 0.06443611855677478,
+      "learning_rate": 1.2181732505347058e-05,
+      "loss": 0.366,
+      "step": 1611
+    },
+    {
+      "epoch": 2.115485564304462,
+      "grad_norm": 0.06449521555518296,
+      "learning_rate": 1.2148966563788129e-05,
+      "loss": 0.3668,
+      "step": 1612
+    },
+    {
+      "epoch": 2.1167979002624673,
+      "grad_norm": 0.06166357061197335,
+      "learning_rate": 1.2116230597915209e-05,
+      "loss": 0.3734,
+      "step": 1613
+    },
+    {
+      "epoch": 2.1181102362204722,
+      "grad_norm": 0.06363606732189205,
+      "learning_rate": 1.2083524684086622e-05,
+      "loss": 0.3531,
+      "step": 1614
+    },
+    {
+      "epoch": 2.1194225721784776,
+      "grad_norm": 0.07811065651597078,
+      "learning_rate": 1.2050848898590553e-05,
+      "loss": 0.3543,
+      "step": 1615
+    },
+    {
+      "epoch": 2.120734908136483,
+      "grad_norm": 0.06779271267356968,
+      "learning_rate": 1.201820331764493e-05,
+      "loss": 0.3582,
+      "step": 1616
+    },
+    {
+      "epoch": 2.122047244094488,
+      "grad_norm": 0.062161777661376844,
+      "learning_rate": 1.1985588017397218e-05,
+      "loss": 0.3573,
+      "step": 1617
+    },
+    {
+      "epoch": 2.1233595800524934,
+      "grad_norm": 0.0683819975051944,
+      "learning_rate": 1.1953003073924257e-05,
+      "loss": 0.3552,
+      "step": 1618
+    },
+    {
+      "epoch": 2.1246719160104988,
+      "grad_norm": 0.05857740185720876,
+      "learning_rate": 1.1920448563232073e-05,
+      "loss": 0.3591,
+      "step": 1619
+    },
+    {
+      "epoch": 2.1259842519685037,
+      "grad_norm": 0.06646161293771743,
+      "learning_rate": 1.188792456125572e-05,
+      "loss": 0.3545,
+      "step": 1620
+    },
+    {
+      "epoch": 2.127296587926509,
+      "grad_norm": 0.06435941141779361,
+      "learning_rate": 1.1855431143859078e-05,
+      "loss": 0.3555,
+      "step": 1621
+    },
+    {
+      "epoch": 2.1286089238845145,
+      "grad_norm": 0.05800308793696974,
+      "learning_rate": 1.1822968386834684e-05,
+      "loss": 0.3526,
+      "step": 1622
+    },
+    {
+      "epoch": 2.1299212598425195,
+      "grad_norm": 0.062493014033742134,
+      "learning_rate": 1.1790536365903563e-05,
+      "loss": 0.3517,
+      "step": 1623
+    },
+    {
+      "epoch": 2.131233595800525,
+      "grad_norm": 0.060421839180095416,
+      "learning_rate": 1.1758135156715042e-05,
+      "loss": 0.3569,
+      "step": 1624
+    },
+    {
+      "epoch": 2.1325459317585302,
+      "grad_norm": 0.06552449433771859,
+      "learning_rate": 1.1725764834846586e-05,
+      "loss": 0.3579,
+      "step": 1625
+    },
+    {
+      "epoch": 2.1338582677165356,
+      "grad_norm": 0.0568105809811458,
+      "learning_rate": 1.1693425475803601e-05,
+      "loss": 0.359,
+      "step": 1626
+    },
+    {
+      "epoch": 2.1351706036745406,
+      "grad_norm": 0.05863595153986424,
+      "learning_rate": 1.1661117155019293e-05,
+      "loss": 0.3568,
+      "step": 1627
+    },
+    {
+      "epoch": 2.136482939632546,
+      "grad_norm": 0.058378602009385584,
+      "learning_rate": 1.1628839947854442e-05,
+      "loss": 0.3529,
+      "step": 1628
+    },
+    {
+      "epoch": 2.1377952755905514,
+      "grad_norm": 0.05979206038117951,
+      "learning_rate": 1.159659392959728e-05,
+      "loss": 0.3613,
+      "step": 1629
+    },
+    {
+      "epoch": 2.1391076115485563,
+      "grad_norm": 0.055041083951442006,
+      "learning_rate": 1.1564379175463275e-05,
+      "loss": 0.3519,
+      "step": 1630
+    },
+    {
+      "epoch": 2.1404199475065617,
+      "grad_norm": 0.053457442322164615,
+      "learning_rate": 1.1532195760594965e-05,
+      "loss": 0.3538,
+      "step": 1631
+    },
+    {
+      "epoch": 2.141732283464567,
+      "grad_norm": 0.062189817203737695,
+      "learning_rate": 1.1500043760061799e-05,
+      "loss": 0.3611,
+      "step": 1632
+    },
+    {
+      "epoch": 2.143044619422572,
+      "grad_norm": 0.055347538317683145,
+      "learning_rate": 1.1467923248859944e-05,
+      "loss": 0.3536,
+      "step": 1633
+    },
+    {
+      "epoch": 2.1443569553805775,
+      "grad_norm": 0.06209216174674383,
+      "learning_rate": 1.1435834301912119e-05,
+      "loss": 0.3591,
+      "step": 1634
+    },
+    {
+      "epoch": 2.145669291338583,
+      "grad_norm": 0.06281816598747232,
+      "learning_rate": 1.1403776994067408e-05,
+      "loss": 0.3556,
+      "step": 1635
+    },
+    {
+      "epoch": 2.146981627296588,
+      "grad_norm": 0.05839462874231476,
+      "learning_rate": 1.1371751400101121e-05,
+      "loss": 0.3565,
+      "step": 1636
+    },
+    {
+      "epoch": 2.148293963254593,
+      "grad_norm": 0.058306362868531254,
+      "learning_rate": 1.133975759471457e-05,
+      "loss": 0.3583,
+      "step": 1637
+    },
+    {
+      "epoch": 2.1496062992125986,
+      "grad_norm": 0.06125681747369518,
+      "learning_rate": 1.1307795652534923e-05,
+      "loss": 0.3586,
+      "step": 1638
+    },
+    {
+      "epoch": 2.1509186351706036,
+      "grad_norm": 0.0624920449125678,
+      "learning_rate": 1.1275865648115031e-05,
+      "loss": 0.3559,
+      "step": 1639
+    },
+    {
+      "epoch": 2.152230971128609,
+      "grad_norm": 0.06032693755839303,
+      "learning_rate": 1.1243967655933235e-05,
+      "loss": 0.3622,
+      "step": 1640
+    },
+    {
+      "epoch": 2.1535433070866143,
+      "grad_norm": 0.06144216211127043,
+      "learning_rate": 1.1212101750393237e-05,
+      "loss": 0.3549,
+      "step": 1641
+    },
+    {
+      "epoch": 2.1548556430446193,
+      "grad_norm": 0.05991097704659593,
+      "learning_rate": 1.1180268005823861e-05,
+      "loss": 0.3567,
+      "step": 1642
+    },
+    {
+      "epoch": 2.1561679790026247,
+      "grad_norm": 0.05698750182609437,
+      "learning_rate": 1.1148466496478921e-05,
+      "loss": 0.3649,
+      "step": 1643
+    },
+    {
+      "epoch": 2.15748031496063,
+      "grad_norm": 0.05999491704983101,
+      "learning_rate": 1.1116697296537068e-05,
+      "loss": 0.355,
+      "step": 1644
+    },
+    {
+      "epoch": 2.158792650918635,
+      "grad_norm": 0.06342837978028584,
+      "learning_rate": 1.1084960480101558e-05,
+      "loss": 0.3562,
+      "step": 1645
+    },
+    {
+      "epoch": 2.1601049868766404,
+      "grad_norm": 0.06110762864710089,
+      "learning_rate": 1.1053256121200121e-05,
+      "loss": 0.3534,
+      "step": 1646
+    },
+    {
+      "epoch": 2.161417322834646,
+      "grad_norm": 0.06170488319010425,
+      "learning_rate": 1.1021584293784778e-05,
+      "loss": 0.3624,
+      "step": 1647
+    },
+    {
+      "epoch": 2.1627296587926508,
+      "grad_norm": 0.058528317282955974,
+      "learning_rate": 1.0989945071731672e-05,
+      "loss": 0.3571,
+      "step": 1648
+    },
+    {
+      "epoch": 2.164041994750656,
+      "grad_norm": 0.05946488152593141,
+      "learning_rate": 1.0958338528840893e-05,
+      "loss": 0.3643,
+      "step": 1649
+    },
+    {
+      "epoch": 2.1653543307086616,
+      "grad_norm": 0.3038442398988041,
+      "learning_rate": 1.0926764738836284e-05,
+      "loss": 0.3639,
+      "step": 1650
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.0653030862894503,
+      "learning_rate": 1.0895223775365337e-05,
+      "loss": 0.3596,
+      "step": 1651
+    },
+    {
+      "epoch": 2.167979002624672,
+      "grad_norm": 0.0600984028953686,
+      "learning_rate": 1.0863715711998928e-05,
+      "loss": 0.3567,
+      "step": 1652
+    },
+    {
+      "epoch": 2.1692913385826773,
+      "grad_norm": 0.07288108048864743,
+      "learning_rate": 1.0832240622231218e-05,
+      "loss": 0.3597,
+      "step": 1653
+    },
+    {
+      "epoch": 2.1706036745406823,
+      "grad_norm": 0.06407079684880841,
+      "learning_rate": 1.0800798579479432e-05,
+      "loss": 0.3567,
+      "step": 1654
+    },
+    {
+      "epoch": 2.1719160104986877,
+      "grad_norm": 0.07184505511231944,
+      "learning_rate": 1.0769389657083748e-05,
+      "loss": 0.3604,
+      "step": 1655
+    },
+    {
+      "epoch": 2.173228346456693,
+      "grad_norm": 0.06543753816342587,
+      "learning_rate": 1.0738013928307053e-05,
+      "loss": 0.3583,
+      "step": 1656
+    },
+    {
+      "epoch": 2.174540682414698,
+      "grad_norm": 0.07465088358614212,
+      "learning_rate": 1.070667146633483e-05,
+      "loss": 0.3619,
+      "step": 1657
+    },
+    {
+      "epoch": 2.1758530183727034,
+      "grad_norm": 0.061722975252185286,
+      "learning_rate": 1.0675362344274953e-05,
+      "loss": 0.3566,
+      "step": 1658
+    },
+    {
+      "epoch": 2.177165354330709,
+      "grad_norm": 0.07658385893949687,
+      "learning_rate": 1.0644086635157527e-05,
+      "loss": 0.3578,
+      "step": 1659
+    },
+    {
+      "epoch": 2.1784776902887137,
+      "grad_norm": 0.06112746376772535,
+      "learning_rate": 1.0612844411934746e-05,
+      "loss": 0.3504,
+      "step": 1660
+    },
+    {
+      "epoch": 2.179790026246719,
+      "grad_norm": 0.06676268482702369,
+      "learning_rate": 1.0581635747480665e-05,
+      "loss": 0.3605,
+      "step": 1661
+    },
+    {
+      "epoch": 2.1811023622047245,
+      "grad_norm": 0.0595160611109363,
+      "learning_rate": 1.0550460714591076e-05,
+      "loss": 0.3588,
+      "step": 1662
+    },
+    {
+      "epoch": 2.1824146981627295,
+      "grad_norm": 0.06454611858947089,
+      "learning_rate": 1.051931938598332e-05,
+      "loss": 0.365,
+      "step": 1663
+    },
+    {
+      "epoch": 2.183727034120735,
+      "grad_norm": 0.05875480159094306,
+      "learning_rate": 1.0488211834296124e-05,
+      "loss": 0.3541,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1850393700787403,
+      "grad_norm": 0.06104849796614371,
+      "learning_rate": 1.0457138132089426e-05,
+      "loss": 0.3515,
+      "step": 1665
+    },
+    {
+      "epoch": 2.1863517060367452,
+      "grad_norm": 0.05961841268279423,
+      "learning_rate": 1.04260983518442e-05,
+      "loss": 0.3542,
+      "step": 1666
+    },
+    {
+      "epoch": 2.1876640419947506,
+      "grad_norm": 0.05908199112738232,
+      "learning_rate": 1.0395092565962322e-05,
+      "loss": 0.3651,
+      "step": 1667
+    },
+    {
+      "epoch": 2.188976377952756,
+      "grad_norm": 0.06726326237231821,
+      "learning_rate": 1.0364120846766343e-05,
+      "loss": 0.3559,
+      "step": 1668
+    },
+    {
+      "epoch": 2.190288713910761,
+      "grad_norm": 0.05563151935714622,
+      "learning_rate": 1.033318326649938e-05,
+      "loss": 0.356,
+      "step": 1669
+    },
+    {
+      "epoch": 2.1916010498687664,
+      "grad_norm": 0.06543915508425273,
+      "learning_rate": 1.0302279897324895e-05,
+      "loss": 0.3596,
+      "step": 1670
+    },
+    {
+      "epoch": 2.1929133858267718,
+      "grad_norm": 0.05957030319607274,
+      "learning_rate": 1.0271410811326565e-05,
+      "loss": 0.358,
+      "step": 1671
+    },
+    {
+      "epoch": 2.1942257217847767,
+      "grad_norm": 0.06254187568854164,
+      "learning_rate": 1.0240576080508094e-05,
+      "loss": 0.3561,
+      "step": 1672
+    },
+    {
+      "epoch": 2.195538057742782,
+      "grad_norm": 0.06834850897952761,
+      "learning_rate": 1.0209775776793052e-05,
+      "loss": 0.3529,
+      "step": 1673
+    },
+    {
+      "epoch": 2.1968503937007875,
+      "grad_norm": 0.0555461238699706,
+      "learning_rate": 1.0179009972024697e-05,
+      "loss": 0.3578,
+      "step": 1674
+    },
+    {
+      "epoch": 2.1981627296587924,
+      "grad_norm": 0.061176122628641035,
+      "learning_rate": 1.0148278737965845e-05,
+      "loss": 0.3653,
+      "step": 1675
+    },
+    {
+      "epoch": 2.199475065616798,
+      "grad_norm": 0.059827569386772514,
+      "learning_rate": 1.0117582146298647e-05,
+      "loss": 0.3506,
+      "step": 1676
+    },
+    {
+      "epoch": 2.2007874015748032,
+      "grad_norm": 0.058861398767757686,
+      "learning_rate": 1.0086920268624456e-05,
+      "loss": 0.3589,
+      "step": 1677
+    },
+    {
+      "epoch": 2.2020997375328086,
+      "grad_norm": 0.055413777157859376,
+      "learning_rate": 1.0056293176463653e-05,
+      "loss": 0.3542,
+      "step": 1678
+    },
+    {
+      "epoch": 2.2034120734908136,
+      "grad_norm": 0.05229055372588018,
+      "learning_rate": 1.002570094125548e-05,
+      "loss": 0.3586,
+      "step": 1679
+    },
+    {
+      "epoch": 2.204724409448819,
+      "grad_norm": 0.0585366135245372,
+      "learning_rate": 9.995143634357878e-06,
+      "loss": 0.357,
+      "step": 1680
+    },
+    {
+      "epoch": 2.2060367454068244,
+      "grad_norm": 0.05686522695066687,
+      "learning_rate": 9.964621327047302e-06,
+      "loss": 0.3557,
+      "step": 1681
+    },
+    {
+      "epoch": 2.2073490813648293,
+      "grad_norm": 0.05875881658351451,
+      "learning_rate": 9.934134090518593e-06,
+      "loss": 0.3484,
+      "step": 1682
+    },
+    {
+      "epoch": 2.2086614173228347,
+      "grad_norm": 0.05819120866049562,
+      "learning_rate": 9.903681995884762e-06,
+      "loss": 0.3537,
+      "step": 1683
+    },
+    {
+      "epoch": 2.20997375328084,
+      "grad_norm": 0.056032848834714155,
+      "learning_rate": 9.873265114176872e-06,
+      "loss": 0.3648,
+      "step": 1684
+    },
+    {
+      "epoch": 2.211286089238845,
+      "grad_norm": 0.05670085831441742,
+      "learning_rate": 9.842883516343835e-06,
+      "loss": 0.3518,
+      "step": 1685
+    },
+    {
+      "epoch": 2.2125984251968505,
+      "grad_norm": 0.05463644001950396,
+      "learning_rate": 9.812537273252264e-06,
+      "loss": 0.3521,
+      "step": 1686
+    },
+    {
+      "epoch": 2.213910761154856,
+      "grad_norm": 0.05598860989268237,
+      "learning_rate": 9.782226455686311e-06,
+      "loss": 0.3639,
+      "step": 1687
+    },
+    {
+      "epoch": 2.215223097112861,
+      "grad_norm": 0.05429860125289025,
+      "learning_rate": 9.751951134347486e-06,
+      "loss": 0.3581,
+      "step": 1688
+    },
+    {
+      "epoch": 2.216535433070866,
+      "grad_norm": 0.06581908023570862,
+      "learning_rate": 9.721711379854514e-06,
+      "loss": 0.3578,
+      "step": 1689
+    },
+    {
+      "epoch": 2.2178477690288716,
+      "grad_norm": 0.053091964985044586,
+      "learning_rate": 9.691507262743143e-06,
+      "loss": 0.3585,
+      "step": 1690
+    },
+    {
+      "epoch": 2.2191601049868765,
+      "grad_norm": 0.06196089000500889,
+      "learning_rate": 9.661338853466021e-06,
+      "loss": 0.3573,
+      "step": 1691
+    },
+    {
+      "epoch": 2.220472440944882,
+      "grad_norm": 0.05488306377911171,
+      "learning_rate": 9.63120622239248e-06,
+      "loss": 0.3659,
+      "step": 1692
+    },
+    {
+      "epoch": 2.2217847769028873,
+      "grad_norm": 0.05764587976711867,
+      "learning_rate": 9.601109439808412e-06,
+      "loss": 0.3554,
+      "step": 1693
+    },
+    {
+      "epoch": 2.2230971128608923,
+      "grad_norm": 0.05803746829221256,
+      "learning_rate": 9.571048575916081e-06,
+      "loss": 0.357,
+      "step": 1694
+    },
+    {
+      "epoch": 2.2244094488188977,
+      "grad_norm": 0.060122810843399425,
+      "learning_rate": 9.54102370083397e-06,
+      "loss": 0.3619,
+      "step": 1695
+    },
+    {
+      "epoch": 2.225721784776903,
+      "grad_norm": 0.05611627877592681,
+      "learning_rate": 9.511034884596631e-06,
+      "loss": 0.3609,
+      "step": 1696
+    },
+    {
+      "epoch": 2.227034120734908,
+      "grad_norm": 0.0586929723275656,
+      "learning_rate": 9.481082197154495e-06,
+      "loss": 0.3604,
+      "step": 1697
+    },
+    {
+      "epoch": 2.2283464566929134,
+      "grad_norm": 0.05753807160978978,
+      "learning_rate": 9.451165708373704e-06,
+      "loss": 0.3546,
+      "step": 1698
+    },
+    {
+      "epoch": 2.229658792650919,
+      "grad_norm": 0.0640000326821599,
+      "learning_rate": 9.421285488036011e-06,
+      "loss": 0.3621,
+      "step": 1699
+    },
+    {
+      "epoch": 2.2309711286089238,
+      "grad_norm": 0.06779206727563135,
+      "learning_rate": 9.391441605838522e-06,
+      "loss": 0.3573,
+      "step": 1700
+    },
+    {
+      "epoch": 2.232283464566929,
+      "grad_norm": 0.05595306861709388,
+      "learning_rate": 9.361634131393613e-06,
+      "loss": 0.3477,
+      "step": 1701
+    },
+    {
+      "epoch": 2.2335958005249346,
+      "grad_norm": 0.06386013600880185,
+      "learning_rate": 9.331863134228718e-06,
+      "loss": 0.3602,
+      "step": 1702
+    },
+    {
+      "epoch": 2.2349081364829395,
+      "grad_norm": 0.06086102014915878,
+      "learning_rate": 9.302128683786202e-06,
+      "loss": 0.3616,
+      "step": 1703
+    },
+    {
+      "epoch": 2.236220472440945,
+      "grad_norm": 0.058312960646176325,
+      "learning_rate": 9.272430849423174e-06,
+      "loss": 0.3472,
+      "step": 1704
+    },
+    {
+      "epoch": 2.2375328083989503,
+      "grad_norm": 0.06394302901507053,
+      "learning_rate": 9.242769700411335e-06,
+      "loss": 0.3625,
+      "step": 1705
+    },
+    {
+      "epoch": 2.2388451443569553,
+      "grad_norm": 0.053938361294237853,
+      "learning_rate": 9.213145305936808e-06,
+      "loss": 0.3536,
+      "step": 1706
+    },
+    {
+      "epoch": 2.2401574803149606,
+      "grad_norm": 0.07014166858304133,
+      "learning_rate": 9.18355773510001e-06,
+      "loss": 0.3575,
+      "step": 1707
+    },
+    {
+      "epoch": 2.241469816272966,
+      "grad_norm": 0.05474173316995641,
+      "learning_rate": 9.15400705691544e-06,
+      "loss": 0.3551,
+      "step": 1708
+    },
+    {
+      "epoch": 2.242782152230971,
+      "grad_norm": 0.06163593476381391,
+      "learning_rate": 9.124493340311537e-06,
+      "loss": 0.3523,
+      "step": 1709
+    },
+    {
+      "epoch": 2.2440944881889764,
+      "grad_norm": 0.06477224774416773,
+      "learning_rate": 9.095016654130567e-06,
+      "loss": 0.363,
+      "step": 1710
+    },
+    {
+      "epoch": 2.245406824146982,
+      "grad_norm": 0.056558526410529254,
+      "learning_rate": 9.065577067128375e-06,
+      "loss": 0.3591,
+      "step": 1711
+    },
+    {
+      "epoch": 2.2467191601049867,
+      "grad_norm": 0.06400533245576914,
+      "learning_rate": 9.036174647974291e-06,
+      "loss": 0.3616,
+      "step": 1712
+    },
+    {
+      "epoch": 2.248031496062992,
+      "grad_norm": 0.059663535030641264,
+      "learning_rate": 9.006809465250945e-06,
+      "loss": 0.3551,
+      "step": 1713
+    },
+    {
+      "epoch": 2.2493438320209975,
+      "grad_norm": 0.05842294829056049,
+      "learning_rate": 8.977481587454106e-06,
+      "loss": 0.3546,
+      "step": 1714
+    },
+    {
+      "epoch": 2.2506561679790025,
+      "grad_norm": 0.06437723556762957,
+      "learning_rate": 8.948191082992546e-06,
+      "loss": 0.359,
+      "step": 1715
+    },
+    {
+      "epoch": 2.251968503937008,
+      "grad_norm": 0.06011363331671689,
+      "learning_rate": 8.918938020187847e-06,
+      "loss": 0.3586,
+      "step": 1716
+    },
+    {
+      "epoch": 2.2532808398950133,
+      "grad_norm": 0.05615395935226597,
+      "learning_rate": 8.889722467274256e-06,
+      "loss": 0.3629,
+      "step": 1717
+    },
+    {
+      "epoch": 2.254593175853018,
+      "grad_norm": 0.05885779256208603,
+      "learning_rate": 8.860544492398526e-06,
+      "loss": 0.3526,
+      "step": 1718
+    },
+    {
+      "epoch": 2.2559055118110236,
+      "grad_norm": 0.05730045837439229,
+      "learning_rate": 8.831404163619763e-06,
+      "loss": 0.3528,
+      "step": 1719
+    },
+    {
+      "epoch": 2.257217847769029,
+      "grad_norm": 0.056536309476499805,
+      "learning_rate": 8.80230154890926e-06,
+      "loss": 0.3544,
+      "step": 1720
+    },
+    {
+      "epoch": 2.258530183727034,
+      "grad_norm": 0.05513151429034815,
+      "learning_rate": 8.773236716150332e-06,
+      "loss": 0.3543,
+      "step": 1721
+    },
+    {
+      "epoch": 2.2598425196850394,
+      "grad_norm": 0.058865859671354985,
+      "learning_rate": 8.744209733138186e-06,
+      "loss": 0.3588,
+      "step": 1722
+    },
+    {
+      "epoch": 2.2611548556430447,
+      "grad_norm": 0.057375154566669515,
+      "learning_rate": 8.715220667579712e-06,
+      "loss": 0.3503,
+      "step": 1723
+    },
+    {
+      "epoch": 2.2624671916010497,
+      "grad_norm": 0.05559730696595684,
+      "learning_rate": 8.686269587093393e-06,
+      "loss": 0.3604,
+      "step": 1724
+    },
+    {
+      "epoch": 2.263779527559055,
+      "grad_norm": 0.06246625883394754,
+      "learning_rate": 8.657356559209082e-06,
+      "loss": 0.355,
+      "step": 1725
+    },
+    {
+      "epoch": 2.2650918635170605,
+      "grad_norm": 0.05965109073820155,
+      "learning_rate": 8.628481651367876e-06,
+      "loss": 0.3586,
+      "step": 1726
+    },
+    {
+      "epoch": 2.2664041994750654,
+      "grad_norm": 0.056916394201248666,
+      "learning_rate": 8.59964493092196e-06,
+      "loss": 0.3527,
+      "step": 1727
+    },
+    {
+      "epoch": 2.267716535433071,
+      "grad_norm": 0.060369564111622254,
+      "learning_rate": 8.570846465134447e-06,
+      "loss": 0.3549,
+      "step": 1728
+    },
+    {
+      "epoch": 2.2690288713910762,
+      "grad_norm": 0.05788151173742681,
+      "learning_rate": 8.542086321179212e-06,
+      "loss": 0.3584,
+      "step": 1729
+    },
+    {
+      "epoch": 2.270341207349081,
+      "grad_norm": 0.06090691995371832,
+      "learning_rate": 8.513364566140744e-06,
+      "loss": 0.3579,
+      "step": 1730
+    },
+    {
+      "epoch": 2.2716535433070866,
+      "grad_norm": 0.06341081495959844,
+      "learning_rate": 8.484681267014002e-06,
+      "loss": 0.3626,
+      "step": 1731
+    },
+    {
+      "epoch": 2.272965879265092,
+      "grad_norm": 0.05421988950531622,
+      "learning_rate": 8.456036490704228e-06,
+      "loss": 0.3685,
+      "step": 1732
+    },
+    {
+      "epoch": 2.274278215223097,
+      "grad_norm": 0.0689666421196173,
+      "learning_rate": 8.427430304026812e-06,
+      "loss": 0.3547,
+      "step": 1733
+    },
+    {
+      "epoch": 2.2755905511811023,
+      "grad_norm": 0.05910014623909422,
+      "learning_rate": 8.39886277370713e-06,
+      "loss": 0.3533,
+      "step": 1734
+    },
+    {
+      "epoch": 2.2769028871391077,
+      "grad_norm": 0.057795256598455705,
+      "learning_rate": 8.370333966380398e-06,
+      "loss": 0.3624,
+      "step": 1735
+    },
+    {
+      "epoch": 2.2782152230971127,
+      "grad_norm": 0.06328932215394228,
+      "learning_rate": 8.34184394859149e-06,
+      "loss": 0.366,
+      "step": 1736
+    },
+    {
+      "epoch": 2.279527559055118,
+      "grad_norm": 0.05467349856759057,
+      "learning_rate": 8.313392786794833e-06,
+      "loss": 0.3596,
+      "step": 1737
+    },
+    {
+      "epoch": 2.2808398950131235,
+      "grad_norm": 0.05406373735377541,
+      "learning_rate": 8.284980547354184e-06,
+      "loss": 0.3564,
+      "step": 1738
+    },
+    {
+      "epoch": 2.2821522309711284,
+      "grad_norm": 0.05609156211087191,
+      "learning_rate": 8.256607296542551e-06,
+      "loss": 0.3545,
+      "step": 1739
+    },
+    {
+      "epoch": 2.283464566929134,
+      "grad_norm": 0.05217085946753699,
+      "learning_rate": 8.228273100541967e-06,
+      "loss": 0.3473,
+      "step": 1740
+    },
+    {
+      "epoch": 2.284776902887139,
+      "grad_norm": 0.05934948019570391,
+      "learning_rate": 8.199978025443375e-06,
+      "loss": 0.359,
+      "step": 1741
+    },
+    {
+      "epoch": 2.286089238845144,
+      "grad_norm": 0.05401868075800795,
+      "learning_rate": 8.171722137246473e-06,
+      "loss": 0.3633,
+      "step": 1742
+    },
+    {
+      "epoch": 2.2874015748031495,
+      "grad_norm": 0.055798859903180224,
+      "learning_rate": 8.143505501859552e-06,
+      "loss": 0.3609,
+      "step": 1743
+    },
+    {
+      "epoch": 2.288713910761155,
+      "grad_norm": 0.05944985514906621,
+      "learning_rate": 8.115328185099342e-06,
+      "loss": 0.3622,
+      "step": 1744
+    },
+    {
+      "epoch": 2.29002624671916,
+      "grad_norm": 0.062006145367526064,
+      "learning_rate": 8.087190252690849e-06,
+      "loss": 0.3557,
+      "step": 1745
+    },
+    {
+      "epoch": 2.2913385826771653,
+      "grad_norm": 0.05859603509220162,
+      "learning_rate": 8.059091770267247e-06,
+      "loss": 0.3588,
+      "step": 1746
+    },
+    {
+      "epoch": 2.2926509186351707,
+      "grad_norm": 0.06074406098201543,
+      "learning_rate": 8.031032803369653e-06,
+      "loss": 0.3532,
+      "step": 1747
+    },
+    {
+      "epoch": 2.2939632545931756,
+      "grad_norm": 0.05640541894792236,
+      "learning_rate": 8.003013417447034e-06,
+      "loss": 0.3501,
+      "step": 1748
+    },
+    {
+      "epoch": 2.295275590551181,
+      "grad_norm": 0.05481747447179637,
+      "learning_rate": 7.975033677856015e-06,
+      "loss": 0.3582,
+      "step": 1749
+    },
+    {
+      "epoch": 2.2965879265091864,
+      "grad_norm": 0.06185481770957466,
+      "learning_rate": 7.947093649860768e-06,
+      "loss": 0.3635,
+      "step": 1750
+    },
+    {
+      "epoch": 2.2979002624671914,
+      "grad_norm": 0.05460868924292492,
+      "learning_rate": 7.919193398632824e-06,
+      "loss": 0.3583,
+      "step": 1751
+    },
+    {
+      "epoch": 2.2992125984251968,
+      "grad_norm": 0.056337584156473074,
+      "learning_rate": 7.891332989250922e-06,
+      "loss": 0.3476,
+      "step": 1752
+    },
+    {
+      "epoch": 2.300524934383202,
+      "grad_norm": 0.0602222365977504,
+      "learning_rate": 7.86351248670088e-06,
+      "loss": 0.3539,
+      "step": 1753
+    },
+    {
+      "epoch": 2.301837270341207,
+      "grad_norm": 0.05515196350980559,
+      "learning_rate": 7.835731955875426e-06,
+      "loss": 0.3603,
+      "step": 1754
+    },
+    {
+      "epoch": 2.3031496062992125,
+      "grad_norm": 0.05424841396281651,
+      "learning_rate": 7.80799146157406e-06,
+      "loss": 0.3605,
+      "step": 1755
+    },
+    {
+      "epoch": 2.304461942257218,
+      "grad_norm": 0.14167444792968842,
+      "learning_rate": 7.78029106850289e-06,
+      "loss": 0.3764,
+      "step": 1756
+    },
+    {
+      "epoch": 2.305774278215223,
+      "grad_norm": 0.05519441182211243,
+      "learning_rate": 7.75263084127448e-06,
+      "loss": 0.3532,
+      "step": 1757
+    },
+    {
+      "epoch": 2.3070866141732282,
+      "grad_norm": 0.05301109802030151,
+      "learning_rate": 7.725010844407715e-06,
+      "loss": 0.3608,
+      "step": 1758
+    },
+    {
+      "epoch": 2.3083989501312336,
+      "grad_norm": 0.05749184683832112,
+      "learning_rate": 7.697431142327632e-06,
+      "loss": 0.3639,
+      "step": 1759
+    },
+    {
+      "epoch": 2.309711286089239,
+      "grad_norm": 0.057009546148056825,
+      "learning_rate": 7.669891799365283e-06,
+      "loss": 0.3591,
+      "step": 1760
+    },
+    {
+      "epoch": 2.311023622047244,
+      "grad_norm": 0.06090992899034897,
+      "learning_rate": 7.64239287975757e-06,
+      "loss": 0.3465,
+      "step": 1761
+    },
+    {
+      "epoch": 2.3123359580052494,
+      "grad_norm": 0.062124767700075786,
+      "learning_rate": 7.614934447647135e-06,
+      "loss": 0.3538,
+      "step": 1762
+    },
+    {
+      "epoch": 2.3136482939632548,
+      "grad_norm": 0.05406760696576841,
+      "learning_rate": 7.587516567082142e-06,
+      "loss": 0.3666,
+      "step": 1763
+    },
+    {
+      "epoch": 2.3149606299212597,
+      "grad_norm": 0.055621810600947084,
+      "learning_rate": 7.560139302016203e-06,
+      "loss": 0.3503,
+      "step": 1764
+    },
+    {
+      "epoch": 2.316272965879265,
+      "grad_norm": 0.05920877946892147,
+      "learning_rate": 7.532802716308163e-06,
+      "loss": 0.3552,
+      "step": 1765
+    },
+    {
+      "epoch": 2.3175853018372705,
+      "grad_norm": 0.05485938578126322,
+      "learning_rate": 7.505506873721996e-06,
+      "loss": 0.3618,
+      "step": 1766
+    },
+    {
+      "epoch": 2.3188976377952755,
+      "grad_norm": 0.05470468549984193,
+      "learning_rate": 7.478251837926637e-06,
+      "loss": 0.3674,
+      "step": 1767
+    },
+    {
+      "epoch": 2.320209973753281,
+      "grad_norm": 0.05261539846255244,
+      "learning_rate": 7.451037672495831e-06,
+      "loss": 0.352,
+      "step": 1768
+    },
+    {
+      "epoch": 2.3215223097112863,
+      "grad_norm": 0.056392989070201884,
+      "learning_rate": 7.423864440907996e-06,
+      "loss": 0.3539,
+      "step": 1769
+    },
+    {
+      "epoch": 2.322834645669291,
+      "grad_norm": 0.05520519984925455,
+      "learning_rate": 7.396732206546081e-06,
+      "loss": 0.3643,
+      "step": 1770
+    },
+    {
+      "epoch": 2.3241469816272966,
+      "grad_norm": 0.05137153146465011,
+      "learning_rate": 7.369641032697389e-06,
+      "loss": 0.3539,
+      "step": 1771
+    },
+    {
+      "epoch": 2.325459317585302,
+      "grad_norm": 0.05616729548859629,
+      "learning_rate": 7.342590982553454e-06,
+      "loss": 0.3538,
+      "step": 1772
+    },
+    {
+      "epoch": 2.326771653543307,
+      "grad_norm": 0.053078990209235,
+      "learning_rate": 7.3155821192098905e-06,
+      "loss": 0.3586,
+      "step": 1773
+    },
+    {
+      "epoch": 2.3280839895013123,
+      "grad_norm": 0.05309993288401357,
+      "learning_rate": 7.288614505666236e-06,
+      "loss": 0.359,
+      "step": 1774
+    },
+    {
+      "epoch": 2.3293963254593177,
+      "grad_norm": 0.058432705857401616,
+      "learning_rate": 7.261688204825814e-06,
+      "loss": 0.3617,
+      "step": 1775
+    },
+    {
+      "epoch": 2.3307086614173227,
+      "grad_norm": 0.052128556073050954,
+      "learning_rate": 7.234803279495581e-06,
+      "loss": 0.3564,
+      "step": 1776
+    },
+    {
+      "epoch": 2.332020997375328,
+      "grad_norm": 0.19121722571020022,
+      "learning_rate": 7.207959792385999e-06,
+      "loss": 0.3489,
+      "step": 1777
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.058720497329515745,
+      "learning_rate": 7.1811578061108415e-06,
+      "loss": 0.3518,
+      "step": 1778
+    },
+    {
+      "epoch": 2.3346456692913384,
+      "grad_norm": 0.05262950751383933,
+      "learning_rate": 7.15439738318712e-06,
+      "loss": 0.3553,
+      "step": 1779
+    },
+    {
+      "epoch": 2.335958005249344,
+      "grad_norm": 0.05598612414340612,
+      "learning_rate": 7.1276785860348634e-06,
+      "loss": 0.3619,
+      "step": 1780
+    },
+    {
+      "epoch": 2.3372703412073492,
+      "grad_norm": 0.05842002921533752,
+      "learning_rate": 7.1010014769770204e-06,
+      "loss": 0.3544,
+      "step": 1781
+    },
+    {
+      "epoch": 2.338582677165354,
+      "grad_norm": 0.053798931269893324,
+      "learning_rate": 7.0743661182392975e-06,
+      "loss": 0.3571,
+      "step": 1782
+    },
+    {
+      "epoch": 2.3398950131233596,
+      "grad_norm": 0.06638210503594781,
+      "learning_rate": 7.047772571950018e-06,
+      "loss": 0.3614,
+      "step": 1783
+    },
+    {
+      "epoch": 2.341207349081365,
+      "grad_norm": 0.052898037057204625,
+      "learning_rate": 7.0212209001399745e-06,
+      "loss": 0.3595,
+      "step": 1784
+    },
+    {
+      "epoch": 2.34251968503937,
+      "grad_norm": 0.058759701679475425,
+      "learning_rate": 6.9947111647422765e-06,
+      "loss": 0.3533,
+      "step": 1785
+    },
+    {
+      "epoch": 2.3438320209973753,
+      "grad_norm": 0.05651603191455769,
+      "learning_rate": 6.968243427592242e-06,
+      "loss": 0.3539,
+      "step": 1786
+    },
+    {
+      "epoch": 2.3451443569553807,
+      "grad_norm": 0.05539653708935374,
+      "learning_rate": 6.941817750427199e-06,
+      "loss": 0.358,
+      "step": 1787
+    },
+    {
+      "epoch": 2.3464566929133857,
+      "grad_norm": 0.06136121664086897,
+      "learning_rate": 6.915434194886375e-06,
+      "loss": 0.3616,
+      "step": 1788
+    },
+    {
+      "epoch": 2.347769028871391,
+      "grad_norm": 0.05298550425728424,
+      "learning_rate": 6.889092822510751e-06,
+      "loss": 0.3606,
+      "step": 1789
+    },
+    {
+      "epoch": 2.3490813648293964,
+      "grad_norm": 0.052645404611534274,
+      "learning_rate": 6.8627936947429055e-06,
+      "loss": 0.3651,
+      "step": 1790
+    },
+    {
+      "epoch": 2.3503937007874014,
+      "grad_norm": 0.05117746342501379,
+      "learning_rate": 6.836536872926897e-06,
+      "loss": 0.3569,
+      "step": 1791
+    },
+    {
+      "epoch": 2.351706036745407,
+      "grad_norm": 0.052470752452150424,
+      "learning_rate": 6.810322418308085e-06,
+      "loss": 0.3545,
+      "step": 1792
+    },
+    {
+      "epoch": 2.353018372703412,
+      "grad_norm": 0.056954705023547526,
+      "learning_rate": 6.784150392033006e-06,
+      "loss": 0.3581,
+      "step": 1793
+    },
+    {
+      "epoch": 2.354330708661417,
+      "grad_norm": 0.052835464527275236,
+      "learning_rate": 6.75802085514925e-06,
+      "loss": 0.3592,
+      "step": 1794
+    },
+    {
+      "epoch": 2.3556430446194225,
+      "grad_norm": 0.055295806985964334,
+      "learning_rate": 6.731933868605275e-06,
+      "loss": 0.3599,
+      "step": 1795
+    },
+    {
+      "epoch": 2.356955380577428,
+      "grad_norm": 0.050875244324891405,
+      "learning_rate": 6.7058894932503e-06,
+      "loss": 0.3555,
+      "step": 1796
+    },
+    {
+      "epoch": 2.358267716535433,
+      "grad_norm": 0.054399255063301274,
+      "learning_rate": 6.679887789834149e-06,
+      "loss": 0.3579,
+      "step": 1797
+    },
+    {
+      "epoch": 2.3595800524934383,
+      "grad_norm": 0.05473399916890118,
+      "learning_rate": 6.653928819007113e-06,
+      "loss": 0.3537,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3608923884514437,
+      "grad_norm": 0.049954268132163186,
+      "learning_rate": 6.628012641319803e-06,
+      "loss": 0.3652,
+      "step": 1799
+    },
+    {
+      "epoch": 2.362204724409449,
+      "grad_norm": 0.05331839147479487,
+      "learning_rate": 6.602139317223014e-06,
+      "loss": 0.3619,
+      "step": 1800
+    },
+    {
+      "epoch": 2.363517060367454,
+      "grad_norm": 0.05388046012709417,
+      "learning_rate": 6.57630890706758e-06,
+      "loss": 0.3554,
+      "step": 1801
+    },
+    {
+      "epoch": 2.3648293963254594,
+      "grad_norm": 0.05274228879812333,
+      "learning_rate": 6.550521471104254e-06,
+      "loss": 0.36,
+      "step": 1802
+    },
+    {
+      "epoch": 2.366141732283465,
+      "grad_norm": 0.051444509185234265,
+      "learning_rate": 6.524777069483526e-06,
+      "loss": 0.3589,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3674540682414698,
+      "grad_norm": 0.05245261393554264,
+      "learning_rate": 6.4990757622555135e-06,
+      "loss": 0.3638,
+      "step": 1804
+    },
+    {
+      "epoch": 2.368766404199475,
+      "grad_norm": 0.05155623902237255,
+      "learning_rate": 6.473417609369825e-06,
+      "loss": 0.3505,
+      "step": 1805
+    },
+    {
+      "epoch": 2.3700787401574805,
+      "grad_norm": 0.05584024939408547,
+      "learning_rate": 6.447802670675401e-06,
+      "loss": 0.3517,
+      "step": 1806
+    },
+    {
+      "epoch": 2.3713910761154855,
+      "grad_norm": 0.05529418607661528,
+      "learning_rate": 6.422231005920376e-06,
+      "loss": 0.3578,
+      "step": 1807
+    },
+    {
+      "epoch": 2.372703412073491,
+      "grad_norm": 0.05495093133335771,
+      "learning_rate": 6.396702674751956e-06,
+      "loss": 0.3499,
+      "step": 1808
+    },
+    {
+      "epoch": 2.3740157480314963,
+      "grad_norm": 0.05576155812900595,
+      "learning_rate": 6.371217736716259e-06,
+      "loss": 0.3587,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3753280839895012,
+      "grad_norm": 0.05191764604830467,
+      "learning_rate": 6.345776251258207e-06,
+      "loss": 0.3637,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3766404199475066,
+      "grad_norm": 0.05427039732812151,
+      "learning_rate": 6.320378277721342e-06,
+      "loss": 0.3578,
+      "step": 1811
+    },
+    {
+      "epoch": 2.377952755905512,
+      "grad_norm": 0.056408666125740484,
+      "learning_rate": 6.2950238753477275e-06,
+      "loss": 0.3622,
+      "step": 1812
+    },
+    {
+      "epoch": 2.379265091863517,
+      "grad_norm": 0.054573231774436216,
+      "learning_rate": 6.269713103277789e-06,
+      "loss": 0.3571,
+      "step": 1813
+    },
+    {
+      "epoch": 2.3805774278215224,
+      "grad_norm": 0.06024554578776597,
+      "learning_rate": 6.244446020550182e-06,
+      "loss": 0.345,
+      "step": 1814
+    },
+    {
+      "epoch": 2.3818897637795278,
+      "grad_norm": 0.05166865185644,
+      "learning_rate": 6.219222686101656e-06,
+      "loss": 0.349,
+      "step": 1815
+    },
+    {
+      "epoch": 2.3832020997375327,
+      "grad_norm": 0.05050262169286038,
+      "learning_rate": 6.1940431587669125e-06,
+      "loss": 0.3591,
+      "step": 1816
+    },
+    {
+      "epoch": 2.384514435695538,
+      "grad_norm": 0.05559392139464911,
+      "learning_rate": 6.168907497278484e-06,
+      "loss": 0.3615,
+      "step": 1817
+    },
+    {
+      "epoch": 2.3858267716535435,
+      "grad_norm": 0.05122170980137,
+      "learning_rate": 6.143815760266561e-06,
+      "loss": 0.361,
+      "step": 1818
+    },
+    {
+      "epoch": 2.3871391076115485,
+      "grad_norm": 0.050450794901557676,
+      "learning_rate": 6.1187680062589015e-06,
+      "loss": 0.3585,
+      "step": 1819
+    },
+    {
+      "epoch": 2.388451443569554,
+      "grad_norm": 0.05470623918358004,
+      "learning_rate": 6.093764293680657e-06,
+      "loss": 0.3539,
+      "step": 1820
+    },
+    {
+      "epoch": 2.3897637795275593,
+      "grad_norm": 0.05216466100768102,
+      "learning_rate": 6.068804680854248e-06,
+      "loss": 0.3531,
+      "step": 1821
+    },
+    {
+      "epoch": 2.391076115485564,
+      "grad_norm": 0.04996226460657203,
+      "learning_rate": 6.043889225999239e-06,
+      "loss": 0.3526,
+      "step": 1822
+    },
+    {
+      "epoch": 2.3923884514435696,
+      "grad_norm": 0.053888458150525204,
+      "learning_rate": 6.019017987232189e-06,
+      "loss": 0.3617,
+      "step": 1823
+    },
+    {
+      "epoch": 2.393700787401575,
+      "grad_norm": 0.06618454400244207,
+      "learning_rate": 5.99419102256652e-06,
+      "loss": 0.3649,
+      "step": 1824
+    },
+    {
+      "epoch": 2.39501312335958,
+      "grad_norm": 0.04976182039309913,
+      "learning_rate": 5.969408389912376e-06,
+      "loss": 0.3616,
+      "step": 1825
+    },
+    {
+      "epoch": 2.3963254593175853,
+      "grad_norm": 0.05494262034311611,
+      "learning_rate": 5.944670147076517e-06,
+      "loss": 0.3628,
+      "step": 1826
+    },
+    {
+      "epoch": 2.3976377952755907,
+      "grad_norm": 0.0533475607031257,
+      "learning_rate": 5.919976351762141e-06,
+      "loss": 0.3569,
+      "step": 1827
+    },
+    {
+      "epoch": 2.3989501312335957,
+      "grad_norm": 0.05365872356879417,
+      "learning_rate": 5.895327061568776e-06,
+      "loss": 0.3593,
+      "step": 1828
+    },
+    {
+      "epoch": 2.400262467191601,
+      "grad_norm": 0.05282593866879189,
+      "learning_rate": 5.870722333992137e-06,
+      "loss": 0.3496,
+      "step": 1829
+    },
+    {
+      "epoch": 2.4015748031496065,
+      "grad_norm": 0.06369121342650719,
+      "learning_rate": 5.846162226423996e-06,
+      "loss": 0.3578,
+      "step": 1830
+    },
+    {
+      "epoch": 2.4028871391076114,
+      "grad_norm": 0.05415674197946112,
+      "learning_rate": 5.821646796152044e-06,
+      "loss": 0.3478,
+      "step": 1831
+    },
+    {
+      "epoch": 2.404199475065617,
+      "grad_norm": 0.052279050934511635,
+      "learning_rate": 5.797176100359775e-06,
+      "loss": 0.3546,
+      "step": 1832
+    },
+    {
+      "epoch": 2.405511811023622,
+      "grad_norm": 0.05798020753822551,
+      "learning_rate": 5.772750196126309e-06,
+      "loss": 0.3689,
+      "step": 1833
+    },
+    {
+      "epoch": 2.406824146981627,
+      "grad_norm": 0.05653231012930724,
+      "learning_rate": 5.748369140426321e-06,
+      "loss": 0.3538,
+      "step": 1834
+    },
+    {
+      "epoch": 2.4081364829396326,
+      "grad_norm": 0.05097819223025257,
+      "learning_rate": 5.724032990129852e-06,
+      "loss": 0.3633,
+      "step": 1835
+    },
+    {
+      "epoch": 2.409448818897638,
+      "grad_norm": 0.07063461913854532,
+      "learning_rate": 5.699741802002201e-06,
+      "loss": 0.358,
+      "step": 1836
+    },
+    {
+      "epoch": 2.410761154855643,
+      "grad_norm": 0.05856703454577752,
+      "learning_rate": 5.6754956327037975e-06,
+      "loss": 0.362,
+      "step": 1837
+    },
+    {
+      "epoch": 2.4120734908136483,
+      "grad_norm": 0.054011605617699424,
+      "learning_rate": 5.6512945387900555e-06,
+      "loss": 0.353,
+      "step": 1838
+    },
+    {
+      "epoch": 2.4133858267716537,
+      "grad_norm": 0.04976559636744368,
+      "learning_rate": 5.627138576711252e-06,
+      "loss": 0.3605,
+      "step": 1839
+    },
+    {
+      "epoch": 2.4146981627296586,
+      "grad_norm": 0.05508285380110518,
+      "learning_rate": 5.603027802812389e-06,
+      "loss": 0.3642,
+      "step": 1840
+    },
+    {
+      "epoch": 2.416010498687664,
+      "grad_norm": 0.05976025974612839,
+      "learning_rate": 5.578962273333075e-06,
+      "loss": 0.3548,
+      "step": 1841
+    },
+    {
+      "epoch": 2.4173228346456694,
+      "grad_norm": 0.05206776981764562,
+      "learning_rate": 5.554942044407368e-06,
+      "loss": 0.3437,
+      "step": 1842
+    },
+    {
+      "epoch": 2.4186351706036744,
+      "grad_norm": 0.052158136637304786,
+      "learning_rate": 5.530967172063675e-06,
+      "loss": 0.3567,
+      "step": 1843
+    },
+    {
+      "epoch": 2.41994750656168,
+      "grad_norm": 0.05181038875692037,
+      "learning_rate": 5.507037712224597e-06,
+      "loss": 0.3655,
+      "step": 1844
+    },
+    {
+      "epoch": 2.421259842519685,
+      "grad_norm": 0.054960093087651395,
+      "learning_rate": 5.483153720706799e-06,
+      "loss": 0.3638,
+      "step": 1845
+    },
+    {
+      "epoch": 2.42257217847769,
+      "grad_norm": 0.04839545200588311,
+      "learning_rate": 5.459315253220918e-06,
+      "loss": 0.3487,
+      "step": 1846
+    },
+    {
+      "epoch": 2.4238845144356955,
+      "grad_norm": 0.048699192904103655,
+      "learning_rate": 5.435522365371376e-06,
+      "loss": 0.3529,
+      "step": 1847
+    },
+    {
+      "epoch": 2.425196850393701,
+      "grad_norm": 0.05034076399957775,
+      "learning_rate": 5.411775112656295e-06,
+      "loss": 0.3565,
+      "step": 1848
+    },
+    {
+      "epoch": 2.426509186351706,
+      "grad_norm": 0.049240227219214464,
+      "learning_rate": 5.388073550467329e-06,
+      "loss": 0.3572,
+      "step": 1849
+    },
+    {
+      "epoch": 2.4278215223097113,
+      "grad_norm": 0.05271699461483889,
+      "learning_rate": 5.3644177340895915e-06,
+      "loss": 0.3641,
+      "step": 1850
+    },
+    {
+      "epoch": 2.4291338582677167,
+      "grad_norm": 0.05512709495892897,
+      "learning_rate": 5.34080771870146e-06,
+      "loss": 0.364,
+      "step": 1851
+    },
+    {
+      "epoch": 2.4304461942257216,
+      "grad_norm": 0.0506990555949897,
+      "learning_rate": 5.317243559374496e-06,
+      "loss": 0.3514,
+      "step": 1852
+    },
+    {
+      "epoch": 2.431758530183727,
+      "grad_norm": 0.0497205002268577,
+      "learning_rate": 5.2937253110732885e-06,
+      "loss": 0.3642,
+      "step": 1853
+    },
+    {
+      "epoch": 2.4330708661417324,
+      "grad_norm": 0.04979816200557152,
+      "learning_rate": 5.270253028655345e-06,
+      "loss": 0.3661,
+      "step": 1854
+    },
+    {
+      "epoch": 2.4343832020997374,
+      "grad_norm": 0.049707121147874206,
+      "learning_rate": 5.246826766870955e-06,
+      "loss": 0.3633,
+      "step": 1855
+    },
+    {
+      "epoch": 2.4356955380577427,
+      "grad_norm": 0.05243156282572297,
+      "learning_rate": 5.2234465803630475e-06,
+      "loss": 0.3624,
+      "step": 1856
+    },
+    {
+      "epoch": 2.437007874015748,
+      "grad_norm": 0.050492860009923465,
+      "learning_rate": 5.200112523667108e-06,
+      "loss": 0.3559,
+      "step": 1857
+    },
+    {
+      "epoch": 2.438320209973753,
+      "grad_norm": 0.0541871907183149,
+      "learning_rate": 5.17682465121099e-06,
+      "loss": 0.3631,
+      "step": 1858
+    },
+    {
+      "epoch": 2.4396325459317585,
+      "grad_norm": 0.05283739697640613,
+      "learning_rate": 5.15358301731485e-06,
+      "loss": 0.3496,
+      "step": 1859
+    },
+    {
+      "epoch": 2.440944881889764,
+      "grad_norm": 0.04742701090106659,
+      "learning_rate": 5.130387676190968e-06,
+      "loss": 0.3578,
+      "step": 1860
+    },
+    {
+      "epoch": 2.442257217847769,
+      "grad_norm": 0.057724577512166,
+      "learning_rate": 5.1072386819436515e-06,
+      "loss": 0.3593,
+      "step": 1861
+    },
+    {
+      "epoch": 2.4435695538057742,
+      "grad_norm": 0.051370996403714064,
+      "learning_rate": 5.0841360885691e-06,
+      "loss": 0.3546,
+      "step": 1862
+    },
+    {
+      "epoch": 2.4448818897637796,
+      "grad_norm": 0.05174415120534029,
+      "learning_rate": 5.061079949955286e-06,
+      "loss": 0.3626,
+      "step": 1863
+    },
+    {
+      "epoch": 2.4461942257217846,
+      "grad_norm": 0.049074469936230845,
+      "learning_rate": 5.0380703198818076e-06,
+      "loss": 0.354,
+      "step": 1864
+    },
+    {
+      "epoch": 2.44750656167979,
+      "grad_norm": 0.049736375626042884,
+      "learning_rate": 5.01510725201981e-06,
+      "loss": 0.3592,
+      "step": 1865
+    },
+    {
+      "epoch": 2.4488188976377954,
+      "grad_norm": 0.05758159238378674,
+      "learning_rate": 4.992190799931801e-06,
+      "loss": 0.3558,
+      "step": 1866
+    },
+    {
+      "epoch": 2.4501312335958003,
+      "grad_norm": 0.05311923244538085,
+      "learning_rate": 4.969321017071568e-06,
+      "loss": 0.3609,
+      "step": 1867
+    },
+    {
+      "epoch": 2.4514435695538057,
+      "grad_norm": 0.050429425958331546,
+      "learning_rate": 4.9464979567840355e-06,
+      "loss": 0.3601,
+      "step": 1868
+    },
+    {
+      "epoch": 2.452755905511811,
+      "grad_norm": 0.05294453424110676,
+      "learning_rate": 4.9237216723051485e-06,
+      "loss": 0.3629,
+      "step": 1869
+    },
+    {
+      "epoch": 2.454068241469816,
+      "grad_norm": 0.05071440732052766,
+      "learning_rate": 4.900992216761741e-06,
+      "loss": 0.3632,
+      "step": 1870
+    },
+    {
+      "epoch": 2.4553805774278215,
+      "grad_norm": 0.0509252969416955,
+      "learning_rate": 4.878309643171414e-06,
+      "loss": 0.3567,
+      "step": 1871
+    },
+    {
+      "epoch": 2.456692913385827,
+      "grad_norm": 0.048456184235244004,
+      "learning_rate": 4.855674004442434e-06,
+      "loss": 0.3564,
+      "step": 1872
+    },
+    {
+      "epoch": 2.458005249343832,
+      "grad_norm": 0.05319173664629956,
+      "learning_rate": 4.833085353373557e-06,
+      "loss": 0.3584,
+      "step": 1873
+    },
+    {
+      "epoch": 2.459317585301837,
+      "grad_norm": 0.051452499918587656,
+      "learning_rate": 4.810543742653967e-06,
+      "loss": 0.3572,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4606299212598426,
+      "grad_norm": 0.05732914352461673,
+      "learning_rate": 4.788049224863112e-06,
+      "loss": 0.3558,
+      "step": 1875
+    },
+    {
+      "epoch": 2.4619422572178475,
+      "grad_norm": 0.3118116720462278,
+      "learning_rate": 4.765601852470586e-06,
+      "loss": 0.3521,
+      "step": 1876
+    },
+    {
+      "epoch": 2.463254593175853,
+      "grad_norm": 0.05747966874788123,
+      "learning_rate": 4.743201677836026e-06,
+      "loss": 0.3602,
+      "step": 1877
+    },
+    {
+      "epoch": 2.4645669291338583,
+      "grad_norm": 0.05601631627298742,
+      "learning_rate": 4.720848753208973e-06,
+      "loss": 0.35,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4658792650918633,
+      "grad_norm": 0.04913812516645295,
+      "learning_rate": 4.698543130728755e-06,
+      "loss": 0.3581,
+      "step": 1879
+    },
+    {
+      "epoch": 2.4671916010498687,
+      "grad_norm": 0.049399705388106945,
+      "learning_rate": 4.6762848624243575e-06,
+      "loss": 0.3597,
+      "step": 1880
+    },
+    {
+      "epoch": 2.468503937007874,
+      "grad_norm": 0.059459385230214545,
+      "learning_rate": 4.654074000214334e-06,
+      "loss": 0.3597,
+      "step": 1881
+    },
+    {
+      "epoch": 2.469816272965879,
+      "grad_norm": 0.055037808854798274,
+      "learning_rate": 4.631910595906635e-06,
+      "loss": 0.3542,
+      "step": 1882
+    },
+    {
+      "epoch": 2.4711286089238844,
+      "grad_norm": 0.054325562139931734,
+      "learning_rate": 4.609794701198525e-06,
+      "loss": 0.3523,
+      "step": 1883
+    },
+    {
+      "epoch": 2.47244094488189,
+      "grad_norm": 0.04994679902573007,
+      "learning_rate": 4.587726367676445e-06,
+      "loss": 0.3608,
+      "step": 1884
+    },
+    {
+      "epoch": 2.4737532808398948,
+      "grad_norm": 0.05104811636668923,
+      "learning_rate": 4.565705646815907e-06,
+      "loss": 0.367,
+      "step": 1885
+    },
+    {
+      "epoch": 2.4750656167979,
+      "grad_norm": 0.0526994407213885,
+      "learning_rate": 4.5437325899813425e-06,
+      "loss": 0.3624,
+      "step": 1886
+    },
+    {
+      "epoch": 2.4763779527559056,
+      "grad_norm": 0.05504719950026765,
+      "learning_rate": 4.521807248426039e-06,
+      "loss": 0.3561,
+      "step": 1887
+    },
+    {
+      "epoch": 2.477690288713911,
+      "grad_norm": 0.05167455878919545,
+      "learning_rate": 4.499929673291945e-06,
+      "loss": 0.3536,
+      "step": 1888
+    },
+    {
+      "epoch": 2.479002624671916,
+      "grad_norm": 0.05081673944288187,
+      "learning_rate": 4.478099915609632e-06,
+      "loss": 0.3597,
+      "step": 1889
+    },
+    {
+      "epoch": 2.4803149606299213,
+      "grad_norm": 0.0520362063107889,
+      "learning_rate": 4.456318026298104e-06,
+      "loss": 0.3594,
+      "step": 1890
+    },
+    {
+      "epoch": 2.4816272965879267,
+      "grad_norm": 0.056897690531467864,
+      "learning_rate": 4.4345840561647204e-06,
+      "loss": 0.3582,
+      "step": 1891
+    },
+    {
+      "epoch": 2.4829396325459316,
+      "grad_norm": 0.0505141938144976,
+      "learning_rate": 4.412898055905071e-06,
+      "loss": 0.3571,
+      "step": 1892
+    },
+    {
+      "epoch": 2.484251968503937,
+      "grad_norm": 0.04782556770920617,
+      "learning_rate": 4.391260076102844e-06,
+      "loss": 0.3546,
+      "step": 1893
+    },
+    {
+      "epoch": 2.4855643044619424,
+      "grad_norm": 0.04916196333810083,
+      "learning_rate": 4.369670167229725e-06,
+      "loss": 0.3585,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4868766404199474,
+      "grad_norm": 0.05392247995211765,
+      "learning_rate": 4.348128379645267e-06,
+      "loss": 0.3565,
+      "step": 1895
+    },
+    {
+      "epoch": 2.4881889763779528,
+      "grad_norm": 0.05001551013389653,
+      "learning_rate": 4.326634763596785e-06,
+      "loss": 0.3515,
+      "step": 1896
+    },
+    {
+      "epoch": 2.489501312335958,
+      "grad_norm": 0.05122300342624965,
+      "learning_rate": 4.305189369219223e-06,
+      "loss": 0.3554,
+      "step": 1897
+    },
+    {
+      "epoch": 2.490813648293963,
+      "grad_norm": 0.051348352378895436,
+      "learning_rate": 4.283792246535051e-06,
+      "loss": 0.3539,
+      "step": 1898
+    },
+    {
+      "epoch": 2.4921259842519685,
+      "grad_norm": 0.05117448705322551,
+      "learning_rate": 4.262443445454134e-06,
+      "loss": 0.3576,
+      "step": 1899
+    },
+    {
+      "epoch": 2.493438320209974,
+      "grad_norm": 0.0482770786225259,
+      "learning_rate": 4.241143015773641e-06,
+      "loss": 0.3483,
+      "step": 1900
+    },
+    {
+      "epoch": 2.494750656167979,
+      "grad_norm": 0.049446231651436606,
+      "learning_rate": 4.219891007177901e-06,
+      "loss": 0.3577,
+      "step": 1901
+    },
+    {
+      "epoch": 2.4960629921259843,
+      "grad_norm": 0.04976924665024374,
+      "learning_rate": 4.198687469238297e-06,
+      "loss": 0.3622,
+      "step": 1902
+    },
+    {
+      "epoch": 2.4973753280839897,
+      "grad_norm": 0.05111674215608954,
+      "learning_rate": 4.177532451413152e-06,
+      "loss": 0.3641,
+      "step": 1903
+    },
+    {
+      "epoch": 2.4986876640419946,
+      "grad_norm": 0.04859573797520916,
+      "learning_rate": 4.156426003047614e-06,
+      "loss": 0.3589,
+      "step": 1904
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.04900528448503481,
+      "learning_rate": 4.1353681733735524e-06,
+      "loss": 0.3571,
+      "step": 1905
+    },
+    {
+      "epoch": 2.5013123359580054,
+      "grad_norm": 0.04963475985020977,
+      "learning_rate": 4.114359011509416e-06,
+      "loss": 0.3551,
+      "step": 1906
+    },
+    {
+      "epoch": 2.5026246719160103,
+      "grad_norm": 0.05182468297025475,
+      "learning_rate": 4.0933985664601385e-06,
+      "loss": 0.3608,
+      "step": 1907
+    },
+    {
+      "epoch": 2.5039370078740157,
+      "grad_norm": 0.04809399407051528,
+      "learning_rate": 4.072486887117017e-06,
+      "loss": 0.3538,
+      "step": 1908
+    },
+    {
+      "epoch": 2.505249343832021,
+      "grad_norm": 0.04798158048868202,
+      "learning_rate": 4.051624022257603e-06,
+      "loss": 0.3537,
+      "step": 1909
+    },
+    {
+      "epoch": 2.506561679790026,
+      "grad_norm": 0.049011185214178424,
+      "learning_rate": 4.0308100205455866e-06,
+      "loss": 0.356,
+      "step": 1910
+    },
+    {
+      "epoch": 2.5078740157480315,
+      "grad_norm": 0.049204871577410575,
+      "learning_rate": 4.010044930530671e-06,
+      "loss": 0.3559,
+      "step": 1911
+    },
+    {
+      "epoch": 2.509186351706037,
+      "grad_norm": 0.046785659172418864,
+      "learning_rate": 3.9893288006484956e-06,
+      "loss": 0.3658,
+      "step": 1912
+    },
+    {
+      "epoch": 2.5104986876640423,
+      "grad_norm": 0.04910057943764418,
+      "learning_rate": 3.968661679220468e-06,
+      "loss": 0.3547,
+      "step": 1913
+    },
+    {
+      "epoch": 2.5118110236220472,
+      "grad_norm": 0.05352416775583909,
+      "learning_rate": 3.948043614453708e-06,
+      "loss": 0.36,
+      "step": 1914
+    },
+    {
+      "epoch": 2.5131233595800526,
+      "grad_norm": 0.05501280419975566,
+      "learning_rate": 3.927474654440888e-06,
+      "loss": 0.3577,
+      "step": 1915
+    },
+    {
+      "epoch": 2.514435695538058,
+      "grad_norm": 0.04542052347059095,
+      "learning_rate": 3.906954847160149e-06,
+      "loss": 0.3529,
+      "step": 1916
+    },
+    {
+      "epoch": 2.515748031496063,
+      "grad_norm": 0.05514277097647821,
+      "learning_rate": 3.88648424047498e-06,
+      "loss": 0.353,
+      "step": 1917
+    },
+    {
+      "epoch": 2.5170603674540684,
+      "grad_norm": 0.05024748907946007,
+      "learning_rate": 3.866062882134108e-06,
+      "loss": 0.3577,
+      "step": 1918
+    },
+    {
+      "epoch": 2.5183727034120738,
+      "grad_norm": 0.048089418673997256,
+      "learning_rate": 3.845690819771386e-06,
+      "loss": 0.3546,
+      "step": 1919
+    },
+    {
+      "epoch": 2.5196850393700787,
+      "grad_norm": 0.04788903845959175,
+      "learning_rate": 3.8253681009056745e-06,
+      "loss": 0.3511,
+      "step": 1920
+    },
+    {
+      "epoch": 2.520997375328084,
+      "grad_norm": 0.05158815078036811,
+      "learning_rate": 3.8050947729407536e-06,
+      "loss": 0.3615,
+      "step": 1921
+    },
+    {
+      "epoch": 2.5223097112860895,
+      "grad_norm": 0.08362604211704089,
+      "learning_rate": 3.784870883165184e-06,
+      "loss": 0.3556,
+      "step": 1922
+    },
+    {
+      "epoch": 2.5236220472440944,
+      "grad_norm": 0.04675332926300411,
+      "learning_rate": 3.7646964787522083e-06,
+      "loss": 0.3608,
+      "step": 1923
+    },
+    {
+      "epoch": 2.5249343832021,
+      "grad_norm": 0.051561931564519284,
+      "learning_rate": 3.7445716067596503e-06,
+      "loss": 0.3567,
+      "step": 1924
+    },
+    {
+      "epoch": 2.5262467191601052,
+      "grad_norm": 0.04726270166338027,
+      "learning_rate": 3.7244963141297913e-06,
+      "loss": 0.3534,
+      "step": 1925
+    },
+    {
+      "epoch": 2.52755905511811,
+      "grad_norm": 0.04933178008805301,
+      "learning_rate": 3.7044706476892678e-06,
+      "loss": 0.348,
+      "step": 1926
+    },
+    {
+      "epoch": 2.5288713910761156,
+      "grad_norm": 0.04917271348309979,
+      "learning_rate": 3.684494654148951e-06,
+      "loss": 0.36,
+      "step": 1927
+    },
+    {
+      "epoch": 2.530183727034121,
+      "grad_norm": 0.05069251162394361,
+      "learning_rate": 3.6645683801038692e-06,
+      "loss": 0.357,
+      "step": 1928
+    },
+    {
+      "epoch": 2.531496062992126,
+      "grad_norm": 0.049138990035736636,
+      "learning_rate": 3.6446918720330654e-06,
+      "loss": 0.348,
+      "step": 1929
+    },
+    {
+      "epoch": 2.5328083989501313,
+      "grad_norm": 0.047252565247883474,
+      "learning_rate": 3.6248651762994995e-06,
+      "loss": 0.3617,
+      "step": 1930
+    },
+    {
+      "epoch": 2.5341207349081367,
+      "grad_norm": 0.04994249215605769,
+      "learning_rate": 3.6050883391499424e-06,
+      "loss": 0.3564,
+      "step": 1931
+    },
+    {
+      "epoch": 2.5354330708661417,
+      "grad_norm": 0.049228514158026544,
+      "learning_rate": 3.5853614067148628e-06,
+      "loss": 0.3639,
+      "step": 1932
+    },
+    {
+      "epoch": 2.536745406824147,
+      "grad_norm": 0.04870508536241451,
+      "learning_rate": 3.5656844250083367e-06,
+      "loss": 0.3578,
+      "step": 1933
+    },
+    {
+      "epoch": 2.5380577427821525,
+      "grad_norm": 0.050696683325101195,
+      "learning_rate": 3.5460574399279133e-06,
+      "loss": 0.3583,
+      "step": 1934
+    },
+    {
+      "epoch": 2.5393700787401574,
+      "grad_norm": 0.04895184594474329,
+      "learning_rate": 3.526480497254528e-06,
+      "loss": 0.359,
+      "step": 1935
+    },
+    {
+      "epoch": 2.540682414698163,
+      "grad_norm": 0.0520290116731824,
+      "learning_rate": 3.506953642652397e-06,
+      "loss": 0.3587,
+      "step": 1936
+    },
+    {
+      "epoch": 2.541994750656168,
+      "grad_norm": 0.0494684962615316,
+      "learning_rate": 3.4874769216688974e-06,
+      "loss": 0.3592,
+      "step": 1937
+    },
+    {
+      "epoch": 2.543307086614173,
+      "grad_norm": 0.05249558154868743,
+      "learning_rate": 3.468050379734464e-06,
+      "loss": 0.3596,
+      "step": 1938
+    },
+    {
+      "epoch": 2.5446194225721785,
+      "grad_norm": 0.04584913348101994,
+      "learning_rate": 3.448674062162488e-06,
+      "loss": 0.3566,
+      "step": 1939
+    },
+    {
+      "epoch": 2.545931758530184,
+      "grad_norm": 0.05072028394785099,
+      "learning_rate": 3.4293480141492055e-06,
+      "loss": 0.3575,
+      "step": 1940
+    },
+    {
+      "epoch": 2.547244094488189,
+      "grad_norm": 0.04724639514331244,
+      "learning_rate": 3.410072280773616e-06,
+      "loss": 0.3584,
+      "step": 1941
+    },
+    {
+      "epoch": 2.5485564304461943,
+      "grad_norm": 0.049436505212745,
+      "learning_rate": 3.390846906997333e-06,
+      "loss": 0.3544,
+      "step": 1942
+    },
+    {
+      "epoch": 2.5498687664041997,
+      "grad_norm": 0.0457946110306144,
+      "learning_rate": 3.3716719376645194e-06,
+      "loss": 0.354,
+      "step": 1943
+    },
+    {
+      "epoch": 2.5511811023622046,
+      "grad_norm": 0.04755840344348723,
+      "learning_rate": 3.352547417501753e-06,
+      "loss": 0.3589,
+      "step": 1944
+    },
+    {
+      "epoch": 2.55249343832021,
+      "grad_norm": 0.0461854336538385,
+      "learning_rate": 3.3334733911179583e-06,
+      "loss": 0.363,
+      "step": 1945
+    },
+    {
+      "epoch": 2.5538057742782154,
+      "grad_norm": 0.0448888146060998,
+      "learning_rate": 3.3144499030042575e-06,
+      "loss": 0.3519,
+      "step": 1946
+    },
+    {
+      "epoch": 2.5551181102362204,
+      "grad_norm": 0.047008485346021055,
+      "learning_rate": 3.2954769975339055e-06,
+      "loss": 0.3589,
+      "step": 1947
+    },
+    {
+      "epoch": 2.5564304461942258,
+      "grad_norm": 0.04706897262345459,
+      "learning_rate": 3.2765547189621604e-06,
+      "loss": 0.3517,
+      "step": 1948
+    },
+    {
+      "epoch": 2.557742782152231,
+      "grad_norm": 0.04796711607739274,
+      "learning_rate": 3.2576831114261965e-06,
+      "loss": 0.3591,
+      "step": 1949
+    },
+    {
+      "epoch": 2.559055118110236,
+      "grad_norm": 0.046363738585271895,
+      "learning_rate": 3.2388622189449936e-06,
+      "loss": 0.3541,
+      "step": 1950
+    },
+    {
+      "epoch": 2.5603674540682415,
+      "grad_norm": 0.04865841251996419,
+      "learning_rate": 3.2200920854192295e-06,
+      "loss": 0.3575,
+      "step": 1951
+    },
+    {
+      "epoch": 2.561679790026247,
+      "grad_norm": 0.04734039851728647,
+      "learning_rate": 3.2013727546312043e-06,
+      "loss": 0.3577,
+      "step": 1952
+    },
+    {
+      "epoch": 2.562992125984252,
+      "grad_norm": 0.04795576726425437,
+      "learning_rate": 3.1827042702446965e-06,
+      "loss": 0.3568,
+      "step": 1953
+    },
+    {
+      "epoch": 2.5643044619422573,
+      "grad_norm": 0.047668658784770175,
+      "learning_rate": 3.164086675804889e-06,
+      "loss": 0.3484,
+      "step": 1954
+    },
+    {
+      "epoch": 2.5656167979002626,
+      "grad_norm": 0.05141474365970435,
+      "learning_rate": 3.1455200147382707e-06,
+      "loss": 0.3577,
+      "step": 1955
+    },
+    {
+      "epoch": 2.5669291338582676,
+      "grad_norm": 0.04814259908869827,
+      "learning_rate": 3.127004330352512e-06,
+      "loss": 0.3647,
+      "step": 1956
+    },
+    {
+      "epoch": 2.568241469816273,
+      "grad_norm": 0.04934459870601553,
+      "learning_rate": 3.108539665836388e-06,
+      "loss": 0.3612,
+      "step": 1957
+    },
+    {
+      "epoch": 2.5695538057742784,
+      "grad_norm": 0.04753958689181884,
+      "learning_rate": 3.09012606425966e-06,
+      "loss": 0.3645,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5708661417322833,
+      "grad_norm": 0.04863946722173951,
+      "learning_rate": 3.0717635685729855e-06,
+      "loss": 0.3535,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5721784776902887,
+      "grad_norm": 0.05019344130458152,
+      "learning_rate": 3.0534522216078193e-06,
+      "loss": 0.3548,
+      "step": 1960
+    },
+    {
+      "epoch": 2.573490813648294,
+      "grad_norm": 0.04864407304682342,
+      "learning_rate": 3.0351920660763033e-06,
+      "loss": 0.3598,
+      "step": 1961
+    },
+    {
+      "epoch": 2.574803149606299,
+      "grad_norm": 0.048380086593796646,
+      "learning_rate": 3.0169831445711754e-06,
+      "loss": 0.3518,
+      "step": 1962
+    },
+    {
+      "epoch": 2.5761154855643045,
+      "grad_norm": 0.048596741882073316,
+      "learning_rate": 2.9988254995656633e-06,
+      "loss": 0.3599,
+      "step": 1963
+    },
+    {
+      "epoch": 2.57742782152231,
+      "grad_norm": 0.04751968533957247,
+      "learning_rate": 2.9807191734133962e-06,
+      "loss": 0.3589,
+      "step": 1964
+    },
+    {
+      "epoch": 2.578740157480315,
+      "grad_norm": 0.04927200034250536,
+      "learning_rate": 2.962664208348295e-06,
+      "loss": 0.367,
+      "step": 1965
+    },
+    {
+      "epoch": 2.58005249343832,
+      "grad_norm": 0.0486707203304219,
+      "learning_rate": 2.944660646484479e-06,
+      "loss": 0.3637,
+      "step": 1966
+    },
+    {
+      "epoch": 2.5813648293963256,
+      "grad_norm": 0.047185555546423924,
+      "learning_rate": 2.926708529816169e-06,
+      "loss": 0.3666,
+      "step": 1967
+    },
+    {
+      "epoch": 2.5826771653543306,
+      "grad_norm": 0.048955909948813096,
+      "learning_rate": 2.908807900217583e-06,
+      "loss": 0.354,
+      "step": 1968
+    },
+    {
+      "epoch": 2.583989501312336,
+      "grad_norm": 0.04765877171244196,
+      "learning_rate": 2.8909587994428526e-06,
+      "loss": 0.3575,
+      "step": 1969
+    },
+    {
+      "epoch": 2.5853018372703414,
+      "grad_norm": 0.04805823974508138,
+      "learning_rate": 2.873161269125907e-06,
+      "loss": 0.3501,
+      "step": 1970
+    },
+    {
+      "epoch": 2.5866141732283463,
+      "grad_norm": 0.046842129345452226,
+      "learning_rate": 2.8554153507803827e-06,
+      "loss": 0.3543,
+      "step": 1971
+    },
+    {
+      "epoch": 2.5879265091863517,
+      "grad_norm": 0.04758851175083853,
+      "learning_rate": 2.8377210857995327e-06,
+      "loss": 0.3569,
+      "step": 1972
+    },
+    {
+      "epoch": 2.589238845144357,
+      "grad_norm": 0.048923708415539315,
+      "learning_rate": 2.820078515456129e-06,
+      "loss": 0.3536,
+      "step": 1973
+    },
+    {
+      "epoch": 2.590551181102362,
+      "grad_norm": 0.050466281720126675,
+      "learning_rate": 2.8024876809023574e-06,
+      "loss": 0.357,
+      "step": 1974
+    },
+    {
+      "epoch": 2.5918635170603674,
+      "grad_norm": 0.047226678555284146,
+      "learning_rate": 2.784948623169728e-06,
+      "loss": 0.3607,
+      "step": 1975
+    },
+    {
+      "epoch": 2.593175853018373,
+      "grad_norm": 0.05143493421057677,
+      "learning_rate": 2.7674613831689873e-06,
+      "loss": 0.3579,
+      "step": 1976
+    },
+    {
+      "epoch": 2.594488188976378,
+      "grad_norm": 0.05054128862843281,
+      "learning_rate": 2.7500260016900037e-06,
+      "loss": 0.3564,
+      "step": 1977
+    },
+    {
+      "epoch": 2.595800524934383,
+      "grad_norm": 0.04786540264569638,
+      "learning_rate": 2.7326425194016877e-06,
+      "loss": 0.3586,
+      "step": 1978
+    },
+    {
+      "epoch": 2.5971128608923886,
+      "grad_norm": 0.07854314736807709,
+      "learning_rate": 2.7153109768518925e-06,
+      "loss": 0.3521,
+      "step": 1979
+    },
+    {
+      "epoch": 2.5984251968503935,
+      "grad_norm": 0.04992040818926543,
+      "learning_rate": 2.698031414467317e-06,
+      "loss": 0.3532,
+      "step": 1980
+    },
+    {
+      "epoch": 2.599737532808399,
+      "grad_norm": 0.05132938735416898,
+      "learning_rate": 2.6808038725534084e-06,
+      "loss": 0.3523,
+      "step": 1981
+    },
+    {
+      "epoch": 2.6010498687664043,
+      "grad_norm": 0.04916384856187757,
+      "learning_rate": 2.6636283912942956e-06,
+      "loss": 0.3521,
+      "step": 1982
+    },
+    {
+      "epoch": 2.6023622047244093,
+      "grad_norm": 0.050767021581270055,
+      "learning_rate": 2.6465050107526423e-06,
+      "loss": 0.3614,
+      "step": 1983
+    },
+    {
+      "epoch": 2.6036745406824147,
+      "grad_norm": 0.051365538516161095,
+      "learning_rate": 2.6294337708696158e-06,
+      "loss": 0.3688,
+      "step": 1984
+    },
+    {
+      "epoch": 2.60498687664042,
+      "grad_norm": 0.11312910727004709,
+      "learning_rate": 2.612414711464739e-06,
+      "loss": 0.3602,
+      "step": 1985
+    },
+    {
+      "epoch": 2.606299212598425,
+      "grad_norm": 0.04919821468816635,
+      "learning_rate": 2.595447872235829e-06,
+      "loss": 0.3617,
+      "step": 1986
+    },
+    {
+      "epoch": 2.6076115485564304,
+      "grad_norm": 0.06438142238315288,
+      "learning_rate": 2.5785332927589e-06,
+      "loss": 0.3598,
+      "step": 1987
+    },
+    {
+      "epoch": 2.608923884514436,
+      "grad_norm": 0.048418290710139596,
+      "learning_rate": 2.5616710124880623e-06,
+      "loss": 0.3572,
+      "step": 1988
+    },
+    {
+      "epoch": 2.6102362204724407,
+      "grad_norm": 0.045425697395792754,
+      "learning_rate": 2.5448610707554386e-06,
+      "loss": 0.3532,
+      "step": 1989
+    },
+    {
+      "epoch": 2.611548556430446,
+      "grad_norm": 0.04706963283553229,
+      "learning_rate": 2.5281035067710653e-06,
+      "loss": 0.3553,
+      "step": 1990
+    },
+    {
+      "epoch": 2.6128608923884515,
+      "grad_norm": 0.05205644279284315,
+      "learning_rate": 2.5113983596228156e-06,
+      "loss": 0.3535,
+      "step": 1991
+    },
+    {
+      "epoch": 2.6141732283464565,
+      "grad_norm": 0.04735245731694125,
+      "learning_rate": 2.494745668276291e-06,
+      "loss": 0.3586,
+      "step": 1992
+    },
+    {
+      "epoch": 2.615485564304462,
+      "grad_norm": 0.047537948526743715,
+      "learning_rate": 2.4781454715747334e-06,
+      "loss": 0.3532,
+      "step": 1993
+    },
+    {
+      "epoch": 2.6167979002624673,
+      "grad_norm": 0.04918149697038648,
+      "learning_rate": 2.461597808238941e-06,
+      "loss": 0.3563,
+      "step": 1994
+    },
+    {
+      "epoch": 2.6181102362204722,
+      "grad_norm": 0.05636953542257394,
+      "learning_rate": 2.445102716867173e-06,
+      "loss": 0.3475,
+      "step": 1995
+    },
+    {
+      "epoch": 2.6194225721784776,
+      "grad_norm": 0.04985760177381763,
+      "learning_rate": 2.428660235935076e-06,
+      "loss": 0.3666,
+      "step": 1996
+    },
+    {
+      "epoch": 2.620734908136483,
+      "grad_norm": 0.04628967703283908,
+      "learning_rate": 2.4122704037955635e-06,
+      "loss": 0.357,
+      "step": 1997
+    },
+    {
+      "epoch": 2.622047244094488,
+      "grad_norm": 0.04715787248070312,
+      "learning_rate": 2.3959332586787452e-06,
+      "loss": 0.3624,
+      "step": 1998
+    },
+    {
+      "epoch": 2.6233595800524934,
+      "grad_norm": 0.05032588779927888,
+      "learning_rate": 2.379648838691842e-06,
+      "loss": 0.3514,
+      "step": 1999
+    },
+    {
+      "epoch": 2.6246719160104988,
+      "grad_norm": 0.04958485882850621,
+      "learning_rate": 2.3634171818190886e-06,
+      "loss": 0.3602,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6259842519685037,
+      "grad_norm": 0.04838879843827294,
+      "learning_rate": 2.347238325921647e-06,
+      "loss": 0.353,
+      "step": 2001
+    },
+    {
+      "epoch": 2.627296587926509,
+      "grad_norm": 0.04500685514631901,
+      "learning_rate": 2.331112308737515e-06,
+      "loss": 0.3539,
+      "step": 2002
+    },
+    {
+      "epoch": 2.6286089238845145,
+      "grad_norm": 0.04543837212583877,
+      "learning_rate": 2.3150391678814472e-06,
+      "loss": 0.3505,
+      "step": 2003
+    },
+    {
+      "epoch": 2.6299212598425195,
+      "grad_norm": 0.04973866730745355,
+      "learning_rate": 2.299018940844855e-06,
+      "loss": 0.3562,
+      "step": 2004
+    },
+    {
+      "epoch": 2.631233595800525,
+      "grad_norm": 0.04754827793784138,
+      "learning_rate": 2.2830516649957277e-06,
+      "loss": 0.3532,
+      "step": 2005
+    },
+    {
+      "epoch": 2.6325459317585302,
+      "grad_norm": 0.04828521280255231,
+      "learning_rate": 2.2671373775785458e-06,
+      "loss": 0.3559,
+      "step": 2006
+    },
+    {
+      "epoch": 2.633858267716535,
+      "grad_norm": 0.04858518747578076,
+      "learning_rate": 2.251276115714193e-06,
+      "loss": 0.3563,
+      "step": 2007
+    },
+    {
+      "epoch": 2.6351706036745406,
+      "grad_norm": 0.04521120679397871,
+      "learning_rate": 2.235467916399861e-06,
+      "loss": 0.3515,
+      "step": 2008
+    },
+    {
+      "epoch": 2.636482939632546,
+      "grad_norm": 0.046714585278710134,
+      "learning_rate": 2.2197128165089842e-06,
+      "loss": 0.3609,
+      "step": 2009
+    },
+    {
+      "epoch": 2.637795275590551,
+      "grad_norm": 0.04634600520311642,
+      "learning_rate": 2.204010852791122e-06,
+      "loss": 0.3621,
+      "step": 2010
+    },
+    {
+      "epoch": 2.6391076115485563,
+      "grad_norm": 0.046593029016280485,
+      "learning_rate": 2.1883620618719014e-06,
+      "loss": 0.3485,
+      "step": 2011
+    },
+    {
+      "epoch": 2.6404199475065617,
+      "grad_norm": 0.046963717649899143,
+      "learning_rate": 2.1727664802529216e-06,
+      "loss": 0.3554,
+      "step": 2012
+    },
+    {
+      "epoch": 2.6417322834645667,
+      "grad_norm": 0.04544709114437087,
+      "learning_rate": 2.1572241443116647e-06,
+      "loss": 0.349,
+      "step": 2013
+    },
+    {
+      "epoch": 2.643044619422572,
+      "grad_norm": 0.04905034110501083,
+      "learning_rate": 2.141735090301414e-06,
+      "loss": 0.3582,
+      "step": 2014
+    },
+    {
+      "epoch": 2.6443569553805775,
+      "grad_norm": 0.045595434838754785,
+      "learning_rate": 2.1262993543511717e-06,
+      "loss": 0.3652,
+      "step": 2015
+    },
+    {
+      "epoch": 2.6456692913385824,
+      "grad_norm": 0.04675152823705787,
+      "learning_rate": 2.110916972465579e-06,
+      "loss": 0.3597,
+      "step": 2016
+    },
+    {
+      "epoch": 2.646981627296588,
+      "grad_norm": 0.050425312809195294,
+      "learning_rate": 2.0955879805248156e-06,
+      "loss": 0.3569,
+      "step": 2017
+    },
+    {
+      "epoch": 2.648293963254593,
+      "grad_norm": 0.04582224460196934,
+      "learning_rate": 2.0803124142845325e-06,
+      "loss": 0.3597,
+      "step": 2018
+    },
+    {
+      "epoch": 2.649606299212598,
+      "grad_norm": 0.04865292185673152,
+      "learning_rate": 2.065090309375761e-06,
+      "loss": 0.3625,
+      "step": 2019
+    },
+    {
+      "epoch": 2.6509186351706036,
+      "grad_norm": 0.047033941312324204,
+      "learning_rate": 2.0499217013048327e-06,
+      "loss": 0.3542,
+      "step": 2020
+    },
+    {
+      "epoch": 2.652230971128609,
+      "grad_norm": 0.04378348477543762,
+      "learning_rate": 2.034806625453292e-06,
+      "loss": 0.352,
+      "step": 2021
+    },
+    {
+      "epoch": 2.653543307086614,
+      "grad_norm": 0.046365190173255556,
+      "learning_rate": 2.0197451170778148e-06,
+      "loss": 0.365,
+      "step": 2022
+    },
+    {
+      "epoch": 2.6548556430446193,
+      "grad_norm": 0.047511516958288195,
+      "learning_rate": 2.0047372113101344e-06,
+      "loss": 0.3533,
+      "step": 2023
+    },
+    {
+      "epoch": 2.6561679790026247,
+      "grad_norm": 0.052168255610815464,
+      "learning_rate": 1.989782943156959e-06,
+      "loss": 0.3568,
+      "step": 2024
+    },
+    {
+      "epoch": 2.65748031496063,
+      "grad_norm": 0.05114024892490732,
+      "learning_rate": 1.974882347499868e-06,
+      "loss": 0.3479,
+      "step": 2025
+    },
+    {
+      "epoch": 2.658792650918635,
+      "grad_norm": 0.048962803119342015,
+      "learning_rate": 1.960035459095255e-06,
+      "loss": 0.3562,
+      "step": 2026
+    },
+    {
+      "epoch": 2.6601049868766404,
+      "grad_norm": 0.04424194839825734,
+      "learning_rate": 1.945242312574244e-06,
+      "loss": 0.3683,
+      "step": 2027
+    },
+    {
+      "epoch": 2.661417322834646,
+      "grad_norm": 0.045404051609959015,
+      "learning_rate": 1.9305029424425915e-06,
+      "loss": 0.3557,
+      "step": 2028
+    },
+    {
+      "epoch": 2.6627296587926508,
+      "grad_norm": 0.3631843962684382,
+      "learning_rate": 1.9158173830806286e-06,
+      "loss": 0.3681,
+      "step": 2029
+    },
+    {
+      "epoch": 2.664041994750656,
+      "grad_norm": 0.04958697850857475,
+      "learning_rate": 1.901185668743158e-06,
+      "loss": 0.3576,
+      "step": 2030
+    },
+    {
+      "epoch": 2.6653543307086616,
+      "grad_norm": 0.047750668393314585,
+      "learning_rate": 1.8866078335594062e-06,
+      "loss": 0.3682,
+      "step": 2031
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.0463713181099126,
+      "learning_rate": 1.872083911532907e-06,
+      "loss": 0.3598,
+      "step": 2032
+    },
+    {
+      "epoch": 2.667979002624672,
+      "grad_norm": 0.04380732187327547,
+      "learning_rate": 1.8576139365414414e-06,
+      "loss": 0.3561,
+      "step": 2033
+    },
+    {
+      "epoch": 2.6692913385826773,
+      "grad_norm": 0.04614519191686969,
+      "learning_rate": 1.8431979423369604e-06,
+      "loss": 0.3553,
+      "step": 2034
+    },
+    {
+      "epoch": 2.6706036745406823,
+      "grad_norm": 0.049632118327091676,
+      "learning_rate": 1.828835962545497e-06,
+      "loss": 0.3511,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6719160104986877,
+      "grad_norm": 0.046863997617341994,
+      "learning_rate": 1.8145280306671037e-06,
+      "loss": 0.3542,
+      "step": 2036
+    },
+    {
+      "epoch": 2.673228346456693,
+      "grad_norm": 0.0453293673772525,
+      "learning_rate": 1.8002741800757477e-06,
+      "loss": 0.3547,
+      "step": 2037
+    },
+    {
+      "epoch": 2.674540682414698,
+      "grad_norm": 0.0453615581450037,
+      "learning_rate": 1.7860744440192617e-06,
+      "loss": 0.3535,
+      "step": 2038
+    },
+    {
+      "epoch": 2.6758530183727034,
+      "grad_norm": 0.04664840489887822,
+      "learning_rate": 1.7719288556192453e-06,
+      "loss": 0.3657,
+      "step": 2039
+    },
+    {
+      "epoch": 2.677165354330709,
+      "grad_norm": 0.04785977642356595,
+      "learning_rate": 1.7578374478710052e-06,
+      "loss": 0.3588,
+      "step": 2040
+    },
+    {
+      "epoch": 2.678477690288714,
+      "grad_norm": 0.04652581019174764,
+      "learning_rate": 1.74380025364346e-06,
+      "loss": 0.347,
+      "step": 2041
+    },
+    {
+      "epoch": 2.679790026246719,
+      "grad_norm": 0.046391679723569654,
+      "learning_rate": 1.7298173056790739e-06,
+      "loss": 0.3528,
+      "step": 2042
+    },
+    {
+      "epoch": 2.6811023622047245,
+      "grad_norm": 0.048794831224333986,
+      "learning_rate": 1.715888636593785e-06,
+      "loss": 0.3571,
+      "step": 2043
+    },
+    {
+      "epoch": 2.68241469816273,
+      "grad_norm": 0.04722559230452588,
+      "learning_rate": 1.702014278876915e-06,
+      "loss": 0.3528,
+      "step": 2044
+    },
+    {
+      "epoch": 2.683727034120735,
+      "grad_norm": 0.04398100863576829,
+      "learning_rate": 1.6881942648911076e-06,
+      "loss": 0.3614,
+      "step": 2045
+    },
+    {
+      "epoch": 2.6850393700787403,
+      "grad_norm": 0.048810010312926616,
+      "learning_rate": 1.674428626872243e-06,
+      "loss": 0.36,
+      "step": 2046
+    },
+    {
+      "epoch": 2.6863517060367457,
+      "grad_norm": 0.04737639569451151,
+      "learning_rate": 1.6607173969293733e-06,
+      "loss": 0.3551,
+      "step": 2047
+    },
+    {
+      "epoch": 2.6876640419947506,
+      "grad_norm": 0.045337722264438114,
+      "learning_rate": 1.6470606070446314e-06,
+      "loss": 0.3588,
+      "step": 2048
+    },
+    {
+      "epoch": 2.688976377952756,
+      "grad_norm": 0.04358946189810038,
+      "learning_rate": 1.6334582890731698e-06,
+      "loss": 0.3577,
+      "step": 2049
+    },
+    {
+      "epoch": 2.6902887139107614,
+      "grad_norm": 0.048963195668697966,
+      "learning_rate": 1.6199104747430944e-06,
+      "loss": 0.3541,
+      "step": 2050
+    },
+    {
+      "epoch": 2.6916010498687664,
+      "grad_norm": 0.043459158912144705,
+      "learning_rate": 1.6064171956553574e-06,
+      "loss": 0.3616,
+      "step": 2051
+    },
+    {
+      "epoch": 2.6929133858267718,
+      "grad_norm": 0.04508839464329014,
+      "learning_rate": 1.5929784832837192e-06,
+      "loss": 0.3587,
+      "step": 2052
+    },
+    {
+      "epoch": 2.694225721784777,
+      "grad_norm": 0.04618483452552496,
+      "learning_rate": 1.5795943689746573e-06,
+      "loss": 0.3497,
+      "step": 2053
+    },
+    {
+      "epoch": 2.695538057742782,
+      "grad_norm": 0.04595464976246346,
+      "learning_rate": 1.5662648839472881e-06,
+      "loss": 0.3594,
+      "step": 2054
+    },
+    {
+      "epoch": 2.6968503937007875,
+      "grad_norm": 0.04594699607593378,
+      "learning_rate": 1.5529900592933244e-06,
+      "loss": 0.3633,
+      "step": 2055
+    },
+    {
+      "epoch": 2.698162729658793,
+      "grad_norm": 0.04562234876946421,
+      "learning_rate": 1.5397699259769554e-06,
+      "loss": 0.3573,
+      "step": 2056
+    },
+    {
+      "epoch": 2.699475065616798,
+      "grad_norm": 0.046828276517912736,
+      "learning_rate": 1.5266045148348174e-06,
+      "loss": 0.358,
+      "step": 2057
+    },
+    {
+      "epoch": 2.7007874015748032,
+      "grad_norm": 0.04470527840863635,
+      "learning_rate": 1.5134938565758928e-06,
+      "loss": 0.3513,
+      "step": 2058
+    },
+    {
+      "epoch": 2.7020997375328086,
+      "grad_norm": 0.047134727250948,
+      "learning_rate": 1.5004379817814606e-06,
+      "loss": 0.3609,
+      "step": 2059
+    },
+    {
+      "epoch": 2.7034120734908136,
+      "grad_norm": 0.04535365375015032,
+      "learning_rate": 1.4874369209050082e-06,
+      "loss": 0.3599,
+      "step": 2060
+    },
+    {
+      "epoch": 2.704724409448819,
+      "grad_norm": 0.04691569890129092,
+      "learning_rate": 1.4744907042721633e-06,
+      "loss": 0.3552,
+      "step": 2061
+    },
+    {
+      "epoch": 2.7060367454068244,
+      "grad_norm": 0.04654002642448677,
+      "learning_rate": 1.46159936208064e-06,
+      "loss": 0.3549,
+      "step": 2062
+    },
+    {
+      "epoch": 2.7073490813648293,
+      "grad_norm": 0.04387593832891074,
+      "learning_rate": 1.4487629244001377e-06,
+      "loss": 0.3568,
+      "step": 2063
+    },
+    {
+      "epoch": 2.7086614173228347,
+      "grad_norm": 0.04489470098541252,
+      "learning_rate": 1.4359814211723083e-06,
+      "loss": 0.3495,
+      "step": 2064
+    },
+    {
+      "epoch": 2.70997375328084,
+      "grad_norm": 0.05025774465487149,
+      "learning_rate": 1.4232548822106478e-06,
+      "loss": 0.3583,
+      "step": 2065
+    },
+    {
+      "epoch": 2.711286089238845,
+      "grad_norm": 0.04310524958772868,
+      "learning_rate": 1.4105833372004524e-06,
+      "loss": 0.3527,
+      "step": 2066
+    },
+    {
+      "epoch": 2.7125984251968505,
+      "grad_norm": 0.04350203615228184,
+      "learning_rate": 1.3979668156987425e-06,
+      "loss": 0.3508,
+      "step": 2067
+    },
+    {
+      "epoch": 2.713910761154856,
+      "grad_norm": 0.045195290876123576,
+      "learning_rate": 1.3854053471341943e-06,
+      "loss": 0.3657,
+      "step": 2068
+    },
+    {
+      "epoch": 2.715223097112861,
+      "grad_norm": 0.045372172070249295,
+      "learning_rate": 1.3728989608070702e-06,
+      "loss": 0.3601,
+      "step": 2069
+    },
+    {
+      "epoch": 2.716535433070866,
+      "grad_norm": 0.04358985995994657,
+      "learning_rate": 1.3604476858891434e-06,
+      "loss": 0.3563,
+      "step": 2070
+    },
+    {
+      "epoch": 2.7178477690288716,
+      "grad_norm": 0.047331650247305256,
+      "learning_rate": 1.3480515514236513e-06,
+      "loss": 0.3593,
+      "step": 2071
+    },
+    {
+      "epoch": 2.7191601049868765,
+      "grad_norm": 0.0451672316198721,
+      "learning_rate": 1.3357105863252006e-06,
+      "loss": 0.3541,
+      "step": 2072
+    },
+    {
+      "epoch": 2.720472440944882,
+      "grad_norm": 0.04233765090131497,
+      "learning_rate": 1.3234248193797205e-06,
+      "loss": 0.3493,
+      "step": 2073
+    },
+    {
+      "epoch": 2.7217847769028873,
+      "grad_norm": 0.04640197329331548,
+      "learning_rate": 1.3111942792443793e-06,
+      "loss": 0.3643,
+      "step": 2074
+    },
+    {
+      "epoch": 2.7230971128608923,
+      "grad_norm": 0.0448242420747861,
+      "learning_rate": 1.2990189944475318e-06,
+      "loss": 0.3587,
+      "step": 2075
+    },
+    {
+      "epoch": 2.7244094488188977,
+      "grad_norm": 0.04374969664622727,
+      "learning_rate": 1.2868989933886466e-06,
+      "loss": 0.3614,
+      "step": 2076
+    },
+    {
+      "epoch": 2.725721784776903,
+      "grad_norm": 0.05928590015392598,
+      "learning_rate": 1.2748343043382433e-06,
+      "loss": 0.3554,
+      "step": 2077
+    },
+    {
+      "epoch": 2.727034120734908,
+      "grad_norm": 0.045434077701312185,
+      "learning_rate": 1.2628249554378135e-06,
+      "loss": 0.3552,
+      "step": 2078
+    },
+    {
+      "epoch": 2.7283464566929134,
+      "grad_norm": 0.047252310591636905,
+      "learning_rate": 1.2508709746997804e-06,
+      "loss": 0.3567,
+      "step": 2079
+    },
+    {
+      "epoch": 2.729658792650919,
+      "grad_norm": 0.04454540345371954,
+      "learning_rate": 1.2389723900074034e-06,
+      "loss": 0.3614,
+      "step": 2080
+    },
+    {
+      "epoch": 2.7309711286089238,
+      "grad_norm": 0.04341783547475662,
+      "learning_rate": 1.2271292291147318e-06,
+      "loss": 0.3547,
+      "step": 2081
+    },
+    {
+      "epoch": 2.732283464566929,
+      "grad_norm": 0.045274852768379996,
+      "learning_rate": 1.2153415196465428e-06,
+      "loss": 0.3554,
+      "step": 2082
+    },
+    {
+      "epoch": 2.7335958005249346,
+      "grad_norm": 0.04470738318087039,
+      "learning_rate": 1.2036092890982619e-06,
+      "loss": 0.3532,
+      "step": 2083
+    },
+    {
+      "epoch": 2.7349081364829395,
+      "grad_norm": 0.04448402062504038,
+      "learning_rate": 1.1919325648359126e-06,
+      "loss": 0.3537,
+      "step": 2084
+    },
+    {
+      "epoch": 2.736220472440945,
+      "grad_norm": 0.04305456409424968,
+      "learning_rate": 1.180311374096038e-06,
+      "loss": 0.3568,
+      "step": 2085
+    },
+    {
+      "epoch": 2.7375328083989503,
+      "grad_norm": 0.04909102927393153,
+      "learning_rate": 1.1687457439856663e-06,
+      "loss": 0.3528,
+      "step": 2086
+    },
+    {
+      "epoch": 2.7388451443569553,
+      "grad_norm": 0.04433031145453869,
+      "learning_rate": 1.1572357014822094e-06,
+      "loss": 0.3543,
+      "step": 2087
+    },
+    {
+      "epoch": 2.7401574803149606,
+      "grad_norm": 0.04454138569827031,
+      "learning_rate": 1.1457812734334245e-06,
+      "loss": 0.3599,
+      "step": 2088
+    },
+    {
+      "epoch": 2.741469816272966,
+      "grad_norm": 0.04484005603296987,
+      "learning_rate": 1.134382486557342e-06,
+      "loss": 0.3631,
+      "step": 2089
+    },
+    {
+      "epoch": 2.742782152230971,
+      "grad_norm": 0.04685537873906327,
+      "learning_rate": 1.1230393674422108e-06,
+      "loss": 0.3655,
+      "step": 2090
+    },
+    {
+      "epoch": 2.7440944881889764,
+      "grad_norm": 0.04440526019842769,
+      "learning_rate": 1.1117519425464324e-06,
+      "loss": 0.3566,
+      "step": 2091
+    },
+    {
+      "epoch": 2.745406824146982,
+      "grad_norm": 0.04180455247877345,
+      "learning_rate": 1.1005202381984963e-06,
+      "loss": 0.3562,
+      "step": 2092
+    },
+    {
+      "epoch": 2.7467191601049867,
+      "grad_norm": 0.04517016772257302,
+      "learning_rate": 1.0893442805969213e-06,
+      "loss": 0.3587,
+      "step": 2093
+    },
+    {
+      "epoch": 2.748031496062992,
+      "grad_norm": 0.04790519663497479,
+      "learning_rate": 1.0782240958101874e-06,
+      "loss": 0.3503,
+      "step": 2094
+    },
+    {
+      "epoch": 2.7493438320209975,
+      "grad_norm": 0.04263330894096276,
+      "learning_rate": 1.0671597097766993e-06,
+      "loss": 0.3496,
+      "step": 2095
+    },
+    {
+      "epoch": 2.7506561679790025,
+      "grad_norm": 0.04414263366547313,
+      "learning_rate": 1.056151148304696e-06,
+      "loss": 0.358,
+      "step": 2096
+    },
+    {
+      "epoch": 2.751968503937008,
+      "grad_norm": 0.04665131013354414,
+      "learning_rate": 1.0451984370721978e-06,
+      "loss": 0.3519,
+      "step": 2097
+    },
+    {
+      "epoch": 2.7532808398950133,
+      "grad_norm": 0.058897130750229586,
+      "learning_rate": 1.0343016016269636e-06,
+      "loss": 0.354,
+      "step": 2098
+    },
+    {
+      "epoch": 2.754593175853018,
+      "grad_norm": 0.08468043494059783,
+      "learning_rate": 1.023460667386411e-06,
+      "loss": 0.3534,
+      "step": 2099
+    },
+    {
+      "epoch": 2.7559055118110236,
+      "grad_norm": 0.04633245701717214,
+      "learning_rate": 1.0126756596375686e-06,
+      "loss": 0.3578,
+      "step": 2100
+    },
+    {
+      "epoch": 2.757217847769029,
+      "grad_norm": 0.04542911552637304,
+      "learning_rate": 1.00194660353701e-06,
+      "loss": 0.3562,
+      "step": 2101
+    },
+    {
+      "epoch": 2.758530183727034,
+      "grad_norm": 0.0429360319339708,
+      "learning_rate": 9.912735241108057e-07,
+      "loss": 0.3478,
+      "step": 2102
+    },
+    {
+      "epoch": 2.7598425196850394,
+      "grad_norm": 0.045200231882010986,
+      "learning_rate": 9.806564462544548e-07,
+      "loss": 0.3494,
+      "step": 2103
+    },
+    {
+      "epoch": 2.7611548556430447,
+      "grad_norm": 0.04268740186984134,
+      "learning_rate": 9.70095394732823e-07,
+      "loss": 0.3587,
+      "step": 2104
+    },
+    {
+      "epoch": 2.7624671916010497,
+      "grad_norm": 0.05030933030146666,
+      "learning_rate": 9.595903941801071e-07,
+      "loss": 0.3616,
+      "step": 2105
+    },
+    {
+      "epoch": 2.763779527559055,
+      "grad_norm": 0.043532530503483303,
+      "learning_rate": 9.49141469099743e-07,
+      "loss": 0.3491,
+      "step": 2106
+    },
+    {
+      "epoch": 2.7650918635170605,
+      "grad_norm": 0.04686576204719352,
+      "learning_rate": 9.387486438643816e-07,
+      "loss": 0.3518,
+      "step": 2107
+    },
+    {
+      "epoch": 2.7664041994750654,
+      "grad_norm": 0.044457969085546806,
+      "learning_rate": 9.28411942715815e-07,
+      "loss": 0.3571,
+      "step": 2108
+    },
+    {
+      "epoch": 2.767716535433071,
+      "grad_norm": 0.04307094724383611,
+      "learning_rate": 9.181313897649202e-07,
+      "loss": 0.3559,
+      "step": 2109
+    },
+    {
+      "epoch": 2.7690288713910762,
+      "grad_norm": 0.04280477223437526,
+      "learning_rate": 9.079070089916041e-07,
+      "loss": 0.3594,
+      "step": 2110
+    },
+    {
+      "epoch": 2.770341207349081,
+      "grad_norm": 0.042711494404853494,
+      "learning_rate": 8.977388242447593e-07,
+      "loss": 0.36,
+      "step": 2111
+    },
+    {
+      "epoch": 2.7716535433070866,
+      "grad_norm": 0.043718017545715665,
+      "learning_rate": 8.876268592421882e-07,
+      "loss": 0.3515,
+      "step": 2112
+    },
+    {
+      "epoch": 2.772965879265092,
+      "grad_norm": 0.04494860654662509,
+      "learning_rate": 8.775711375705598e-07,
+      "loss": 0.3613,
+      "step": 2113
+    },
+    {
+      "epoch": 2.774278215223097,
+      "grad_norm": 0.044505580912473204,
+      "learning_rate": 8.675716826853558e-07,
+      "loss": 0.3572,
+      "step": 2114
+    },
+    {
+      "epoch": 2.7755905511811023,
+      "grad_norm": 0.06060424287142137,
+      "learning_rate": 8.576285179108107e-07,
+      "loss": 0.3537,
+      "step": 2115
+    },
+    {
+      "epoch": 2.7769028871391077,
+      "grad_norm": 0.04362964810850725,
+      "learning_rate": 8.47741666439858e-07,
+      "loss": 0.3547,
+      "step": 2116
+    },
+    {
+      "epoch": 2.7782152230971127,
+      "grad_norm": 0.04401476627703584,
+      "learning_rate": 8.379111513340754e-07,
+      "loss": 0.349,
+      "step": 2117
+    },
+    {
+      "epoch": 2.779527559055118,
+      "grad_norm": 0.04459235338360295,
+      "learning_rate": 8.281369955236456e-07,
+      "loss": 0.3483,
+      "step": 2118
+    },
+    {
+      "epoch": 2.7808398950131235,
+      "grad_norm": 0.041937767824518524,
+      "learning_rate": 8.18419221807279e-07,
+      "loss": 0.3568,
+      "step": 2119
+    },
+    {
+      "epoch": 2.7821522309711284,
+      "grad_norm": 0.04321704530551636,
+      "learning_rate": 8.087578528521739e-07,
+      "loss": 0.3446,
+      "step": 2120
+    },
+    {
+      "epoch": 2.783464566929134,
+      "grad_norm": 0.044667423671111685,
+      "learning_rate": 7.991529111939622e-07,
+      "loss": 0.3631,
+      "step": 2121
+    },
+    {
+      "epoch": 2.784776902887139,
+      "grad_norm": 0.04629646037275347,
+      "learning_rate": 7.896044192366586e-07,
+      "loss": 0.3636,
+      "step": 2122
+    },
+    {
+      "epoch": 2.786089238845144,
+      "grad_norm": 0.04380770621415317,
+      "learning_rate": 7.801123992526e-07,
+      "loss": 0.3599,
+      "step": 2123
+    },
+    {
+      "epoch": 2.7874015748031495,
+      "grad_norm": 0.043688285621081506,
+      "learning_rate": 7.70676873382406e-07,
+      "loss": 0.352,
+      "step": 2124
+    },
+    {
+      "epoch": 2.788713910761155,
+      "grad_norm": 0.0710731504795418,
+      "learning_rate": 7.612978636349161e-07,
+      "loss": 0.3524,
+      "step": 2125
+    },
+    {
+      "epoch": 2.79002624671916,
+      "grad_norm": 0.044472474490278795,
+      "learning_rate": 7.519753918871497e-07,
+      "loss": 0.3652,
+      "step": 2126
+    },
+    {
+      "epoch": 2.7913385826771653,
+      "grad_norm": 0.042924250780580944,
+      "learning_rate": 7.427094798842404e-07,
+      "loss": 0.365,
+      "step": 2127
+    },
+    {
+      "epoch": 2.7926509186351707,
+      "grad_norm": 0.041994981420152074,
+      "learning_rate": 7.335001492393994e-07,
+      "loss": 0.3582,
+      "step": 2128
+    },
+    {
+      "epoch": 2.7939632545931756,
+      "grad_norm": 0.04431362503693609,
+      "learning_rate": 7.243474214338547e-07,
+      "loss": 0.3573,
+      "step": 2129
+    },
+    {
+      "epoch": 2.795275590551181,
+      "grad_norm": 0.04556937959454002,
+      "learning_rate": 7.152513178168063e-07,
+      "loss": 0.3651,
+      "step": 2130
+    },
+    {
+      "epoch": 2.7965879265091864,
+      "grad_norm": 0.04329582474110909,
+      "learning_rate": 7.062118596053768e-07,
+      "loss": 0.3597,
+      "step": 2131
+    },
+    {
+      "epoch": 2.7979002624671914,
+      "grad_norm": 0.04277219841921721,
+      "learning_rate": 6.972290678845639e-07,
+      "loss": 0.3558,
+      "step": 2132
+    },
+    {
+      "epoch": 2.7992125984251968,
+      "grad_norm": 0.0424055339879228,
+      "learning_rate": 6.883029636071819e-07,
+      "loss": 0.3528,
+      "step": 2133
+    },
+    {
+      "epoch": 2.800524934383202,
+      "grad_norm": 0.05799868469786029,
+      "learning_rate": 6.79433567593818e-07,
+      "loss": 0.3642,
+      "step": 2134
+    },
+    {
+      "epoch": 2.801837270341207,
+      "grad_norm": 0.04637003563712031,
+      "learning_rate": 6.706209005327895e-07,
+      "loss": 0.3574,
+      "step": 2135
+    },
+    {
+      "epoch": 2.8031496062992125,
+      "grad_norm": 0.04151010007202843,
+      "learning_rate": 6.6186498298009e-07,
+      "loss": 0.3526,
+      "step": 2136
+    },
+    {
+      "epoch": 2.804461942257218,
+      "grad_norm": 0.04223301307224779,
+      "learning_rate": 6.531658353593373e-07,
+      "loss": 0.3561,
+      "step": 2137
+    },
+    {
+      "epoch": 2.805774278215223,
+      "grad_norm": 0.04359846609429753,
+      "learning_rate": 6.445234779617365e-07,
+      "loss": 0.3615,
+      "step": 2138
+    },
+    {
+      "epoch": 2.8070866141732282,
+      "grad_norm": 0.042997062302553936,
+      "learning_rate": 6.359379309460234e-07,
+      "loss": 0.3612,
+      "step": 2139
+    },
+    {
+      "epoch": 2.8083989501312336,
+      "grad_norm": 0.04241276214780555,
+      "learning_rate": 6.274092143384202e-07,
+      "loss": 0.3559,
+      "step": 2140
+    },
+    {
+      "epoch": 2.8097112860892386,
+      "grad_norm": 0.04531278618957648,
+      "learning_rate": 6.189373480325861e-07,
+      "loss": 0.3622,
+      "step": 2141
+    },
+    {
+      "epoch": 2.811023622047244,
+      "grad_norm": 0.04271985135023751,
+      "learning_rate": 6.105223517895858e-07,
+      "loss": 0.3558,
+      "step": 2142
+    },
+    {
+      "epoch": 2.8123359580052494,
+      "grad_norm": 0.041779553146376165,
+      "learning_rate": 6.021642452378184e-07,
+      "loss": 0.3552,
+      "step": 2143
+    },
+    {
+      "epoch": 2.8136482939632543,
+      "grad_norm": 0.04498367352787201,
+      "learning_rate": 5.938630478729917e-07,
+      "loss": 0.358,
+      "step": 2144
+    },
+    {
+      "epoch": 2.8149606299212597,
+      "grad_norm": 0.04374953729260444,
+      "learning_rate": 5.856187790580697e-07,
+      "loss": 0.3589,
+      "step": 2145
+    },
+    {
+      "epoch": 2.816272965879265,
+      "grad_norm": 0.04421792152446158,
+      "learning_rate": 5.774314580232281e-07,
+      "loss": 0.3479,
+      "step": 2146
+    },
+    {
+      "epoch": 2.81758530183727,
+      "grad_norm": 0.04441579741380491,
+      "learning_rate": 5.693011038658048e-07,
+      "loss": 0.359,
+      "step": 2147
+    },
+    {
+      "epoch": 2.8188976377952755,
+      "grad_norm": 0.04182478305110252,
+      "learning_rate": 5.612277355502655e-07,
+      "loss": 0.3467,
+      "step": 2148
+    },
+    {
+      "epoch": 2.820209973753281,
+      "grad_norm": 0.043968323085362436,
+      "learning_rate": 5.532113719081438e-07,
+      "loss": 0.3473,
+      "step": 2149
+    },
+    {
+      "epoch": 2.821522309711286,
+      "grad_norm": 0.0427332842353786,
+      "learning_rate": 5.452520316380244e-07,
+      "loss": 0.3524,
+      "step": 2150
+    },
+    {
+      "epoch": 2.822834645669291,
+      "grad_norm": 0.04460458422879059,
+      "learning_rate": 5.373497333054617e-07,
+      "loss": 0.3576,
+      "step": 2151
+    },
+    {
+      "epoch": 2.8241469816272966,
+      "grad_norm": 0.043320050873474125,
+      "learning_rate": 5.295044953429723e-07,
+      "loss": 0.3528,
+      "step": 2152
+    },
+    {
+      "epoch": 2.825459317585302,
+      "grad_norm": 0.041766850704729803,
+      "learning_rate": 5.217163360499683e-07,
+      "loss": 0.3518,
+      "step": 2153
+    },
+    {
+      "epoch": 2.826771653543307,
+      "grad_norm": 0.04168417585590033,
+      "learning_rate": 5.139852735927265e-07,
+      "loss": 0.3564,
+      "step": 2154
+    },
+    {
+      "epoch": 2.8280839895013123,
+      "grad_norm": 0.04394473631252587,
+      "learning_rate": 5.063113260043412e-07,
+      "loss": 0.3559,
+      "step": 2155
+    },
+    {
+      "epoch": 2.8293963254593177,
+      "grad_norm": 0.04298501686177732,
+      "learning_rate": 4.98694511184683e-07,
+      "loss": 0.3588,
+      "step": 2156
+    },
+    {
+      "epoch": 2.8307086614173227,
+      "grad_norm": 0.042201409110942056,
+      "learning_rate": 4.911348469003568e-07,
+      "loss": 0.3587,
+      "step": 2157
+    },
+    {
+      "epoch": 2.832020997375328,
+      "grad_norm": 0.05323662193242019,
+      "learning_rate": 4.836323507846658e-07,
+      "loss": 0.356,
+      "step": 2158
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.04177862839665858,
+      "learning_rate": 4.7618704033756125e-07,
+      "loss": 0.3512,
+      "step": 2159
+    },
+    {
+      "epoch": 2.8346456692913384,
+      "grad_norm": 0.0550421785206551,
+      "learning_rate": 4.6879893292560426e-07,
+      "loss": 0.358,
+      "step": 2160
+    },
+    {
+      "epoch": 2.835958005249344,
+      "grad_norm": 0.041254348082554355,
+      "learning_rate": 4.6146804578192927e-07,
+      "loss": 0.3639,
+      "step": 2161
+    },
+    {
+      "epoch": 2.8372703412073492,
+      "grad_norm": 0.04353478725625004,
+      "learning_rate": 4.541943960062051e-07,
+      "loss": 0.3621,
+      "step": 2162
+    },
+    {
+      "epoch": 2.838582677165354,
+      "grad_norm": 0.043257381731781695,
+      "learning_rate": 4.4697800056458795e-07,
+      "loss": 0.3558,
+      "step": 2163
+    },
+    {
+      "epoch": 2.8398950131233596,
+      "grad_norm": 0.04309193147341983,
+      "learning_rate": 4.398188762896799e-07,
+      "loss": 0.3515,
+      "step": 2164
+    },
+    {
+      "epoch": 2.841207349081365,
+      "grad_norm": 0.04536669899754826,
+      "learning_rate": 4.327170398805036e-07,
+      "loss": 0.3527,
+      "step": 2165
+    },
+    {
+      "epoch": 2.84251968503937,
+      "grad_norm": 0.04164663062961032,
+      "learning_rate": 4.256725079024554e-07,
+      "loss": 0.351,
+      "step": 2166
+    },
+    {
+      "epoch": 2.8438320209973753,
+      "grad_norm": 0.04287308517212222,
+      "learning_rate": 4.1868529678726067e-07,
+      "loss": 0.3462,
+      "step": 2167
+    },
+    {
+      "epoch": 2.8451443569553807,
+      "grad_norm": 0.053675684043672664,
+      "learning_rate": 4.1175542283294064e-07,
+      "loss": 0.3546,
+      "step": 2168
+    },
+    {
+      "epoch": 2.846456692913386,
+      "grad_norm": 0.0435449971409896,
+      "learning_rate": 4.0488290220377924e-07,
+      "loss": 0.3597,
+      "step": 2169
+    },
+    {
+      "epoch": 2.847769028871391,
+      "grad_norm": 0.04513777853441207,
+      "learning_rate": 3.9806775093027824e-07,
+      "loss": 0.3672,
+      "step": 2170
+    },
+    {
+      "epoch": 2.8490813648293964,
+      "grad_norm": 0.04340924966566775,
+      "learning_rate": 3.913099849091245e-07,
+      "loss": 0.365,
+      "step": 2171
+    },
+    {
+      "epoch": 2.850393700787402,
+      "grad_norm": 0.04565998600373819,
+      "learning_rate": 3.8460961990314517e-07,
+      "loss": 0.3561,
+      "step": 2172
+    },
+    {
+      "epoch": 2.851706036745407,
+      "grad_norm": 0.04236491473970563,
+      "learning_rate": 3.7796667154128273e-07,
+      "loss": 0.3517,
+      "step": 2173
+    },
+    {
+      "epoch": 2.853018372703412,
+      "grad_norm": 0.04343261640859382,
+      "learning_rate": 3.7138115531855643e-07,
+      "loss": 0.357,
+      "step": 2174
+    },
+    {
+      "epoch": 2.8543307086614176,
+      "grad_norm": 0.04063199574309751,
+      "learning_rate": 3.648530865960065e-07,
+      "loss": 0.3526,
+      "step": 2175
+    },
+    {
+      "epoch": 2.8556430446194225,
+      "grad_norm": 0.04246342735453163,
+      "learning_rate": 3.5838248060069137e-07,
+      "loss": 0.3557,
+      "step": 2176
+    },
+    {
+      "epoch": 2.856955380577428,
+      "grad_norm": 0.04242901897903004,
+      "learning_rate": 3.51969352425624e-07,
+      "loss": 0.3635,
+      "step": 2177
+    },
+    {
+      "epoch": 2.8582677165354333,
+      "grad_norm": 0.04360223162578459,
+      "learning_rate": 3.4561371702974967e-07,
+      "loss": 0.3518,
+      "step": 2178
+    },
+    {
+      "epoch": 2.8595800524934383,
+      "grad_norm": 0.04389920118523791,
+      "learning_rate": 3.3931558923791243e-07,
+      "loss": 0.3606,
+      "step": 2179
+    },
+    {
+      "epoch": 2.8608923884514437,
+      "grad_norm": 0.04409769300103001,
+      "learning_rate": 3.3307498374080816e-07,
+      "loss": 0.3645,
+      "step": 2180
+    },
+    {
+      "epoch": 2.862204724409449,
+      "grad_norm": 0.04249583288945692,
+      "learning_rate": 3.2689191509497065e-07,
+      "loss": 0.3519,
+      "step": 2181
+    },
+    {
+      "epoch": 2.863517060367454,
+      "grad_norm": 0.04430374443593538,
+      "learning_rate": 3.2076639772272153e-07,
+      "loss": 0.3582,
+      "step": 2182
+    },
+    {
+      "epoch": 2.8648293963254594,
+      "grad_norm": 0.17754538512899629,
+      "learning_rate": 3.1469844591213717e-07,
+      "loss": 0.3596,
+      "step": 2183
+    },
+    {
+      "epoch": 2.866141732283465,
+      "grad_norm": 0.04508723803249319,
+      "learning_rate": 3.086880738170206e-07,
+      "loss": 0.3617,
+      "step": 2184
+    },
+    {
+      "epoch": 2.8674540682414698,
+      "grad_norm": 0.0418342670772758,
+      "learning_rate": 3.027352954568713e-07,
+      "loss": 0.3573,
+      "step": 2185
+    },
+    {
+      "epoch": 2.868766404199475,
+      "grad_norm": 0.2693460620509927,
+      "learning_rate": 2.96840124716849e-07,
+      "loss": 0.3766,
+      "step": 2186
+    },
+    {
+      "epoch": 2.8700787401574805,
+      "grad_norm": 0.04377974429215082,
+      "learning_rate": 2.910025753477319e-07,
+      "loss": 0.3528,
+      "step": 2187
+    },
+    {
+      "epoch": 2.8713910761154855,
+      "grad_norm": 0.042308563600515096,
+      "learning_rate": 2.852226609659059e-07,
+      "loss": 0.3577,
+      "step": 2188
+    },
+    {
+      "epoch": 2.872703412073491,
+      "grad_norm": 0.0427041341429207,
+      "learning_rate": 2.7950039505330594e-07,
+      "loss": 0.3499,
+      "step": 2189
+    },
+    {
+      "epoch": 2.8740157480314963,
+      "grad_norm": 0.042049470924498206,
+      "learning_rate": 2.738357909574135e-07,
+      "loss": 0.3551,
+      "step": 2190
+    },
+    {
+      "epoch": 2.8753280839895012,
+      "grad_norm": 0.043406443213581315,
+      "learning_rate": 2.6822886189120377e-07,
+      "loss": 0.3567,
+      "step": 2191
+    },
+    {
+      "epoch": 2.8766404199475066,
+      "grad_norm": 0.04098317814805982,
+      "learning_rate": 2.626796209331206e-07,
+      "loss": 0.3594,
+      "step": 2192
+    },
+    {
+      "epoch": 2.877952755905512,
+      "grad_norm": 0.04330622328810624,
+      "learning_rate": 2.571880810270516e-07,
+      "loss": 0.358,
+      "step": 2193
+    },
+    {
+      "epoch": 2.879265091863517,
+      "grad_norm": 0.04216148707637403,
+      "learning_rate": 2.5175425498228657e-07,
+      "loss": 0.3586,
+      "step": 2194
+    },
+    {
+      "epoch": 2.8805774278215224,
+      "grad_norm": 0.0446785001087786,
+      "learning_rate": 2.463781554735062e-07,
+      "loss": 0.36,
+      "step": 2195
+    },
+    {
+      "epoch": 2.8818897637795278,
+      "grad_norm": 0.04489462069670692,
+      "learning_rate": 2.410597950407295e-07,
+      "loss": 0.3571,
+      "step": 2196
+    },
+    {
+      "epoch": 2.8832020997375327,
+      "grad_norm": 0.041779218754766145,
+      "learning_rate": 2.3579918608930273e-07,
+      "loss": 0.3519,
+      "step": 2197
+    },
+    {
+      "epoch": 2.884514435695538,
+      "grad_norm": 0.044579835510484875,
+      "learning_rate": 2.3059634088986592e-07,
+      "loss": 0.3577,
+      "step": 2198
+    },
+    {
+      "epoch": 2.8858267716535435,
+      "grad_norm": 0.04346799266434987,
+      "learning_rate": 2.2545127157831413e-07,
+      "loss": 0.3532,
+      "step": 2199
+    },
+    {
+      "epoch": 2.8871391076115485,
+      "grad_norm": 0.0426533497643065,
+      "learning_rate": 2.2036399015578634e-07,
+      "loss": 0.3525,
+      "step": 2200
+    },
+    {
+      "epoch": 2.888451443569554,
+      "grad_norm": 0.043903665317712726,
+      "learning_rate": 2.153345084886238e-07,
+      "loss": 0.3512,
+      "step": 2201
+    },
+    {
+      "epoch": 2.8897637795275593,
+      "grad_norm": 0.04312412565060085,
+      "learning_rate": 2.1036283830834224e-07,
+      "loss": 0.3541,
+      "step": 2202
+    },
+    {
+      "epoch": 2.891076115485564,
+      "grad_norm": 0.042669538654930894,
+      "learning_rate": 2.054489912116181e-07,
+      "loss": 0.3495,
+      "step": 2203
+    },
+    {
+      "epoch": 2.8923884514435696,
+      "grad_norm": 0.04341938668078426,
+      "learning_rate": 2.0059297866024673e-07,
+      "loss": 0.3593,
+      "step": 2204
+    },
+    {
+      "epoch": 2.893700787401575,
+      "grad_norm": 0.04475104516431718,
+      "learning_rate": 1.9579481198112315e-07,
+      "loss": 0.3603,
+      "step": 2205
+    },
+    {
+      "epoch": 2.89501312335958,
+      "grad_norm": 0.041304500552404116,
+      "learning_rate": 1.9105450236621692e-07,
+      "loss": 0.3514,
+      "step": 2206
+    },
+    {
+      "epoch": 2.8963254593175853,
+      "grad_norm": 0.04231105172295359,
+      "learning_rate": 1.8637206087253888e-07,
+      "loss": 0.3546,
+      "step": 2207
+    },
+    {
+      "epoch": 2.8976377952755907,
+      "grad_norm": 0.042579406946531236,
+      "learning_rate": 1.8174749842211903e-07,
+      "loss": 0.359,
+      "step": 2208
+    },
+    {
+      "epoch": 2.8989501312335957,
+      "grad_norm": 0.04191012437008877,
+      "learning_rate": 1.7718082580198414e-07,
+      "loss": 0.3571,
+      "step": 2209
+    },
+    {
+      "epoch": 2.900262467191601,
+      "grad_norm": 0.04370444201581222,
+      "learning_rate": 1.726720536641274e-07,
+      "loss": 0.359,
+      "step": 2210
+    },
+    {
+      "epoch": 2.9015748031496065,
+      "grad_norm": 0.0420271782605382,
+      "learning_rate": 1.6822119252549174e-07,
+      "loss": 0.35,
+      "step": 2211
+    },
+    {
+      "epoch": 2.9028871391076114,
+      "grad_norm": 0.044005308848892435,
+      "learning_rate": 1.6382825276793356e-07,
+      "loss": 0.3583,
+      "step": 2212
+    },
+    {
+      "epoch": 2.904199475065617,
+      "grad_norm": 0.042487690387185945,
+      "learning_rate": 1.5949324463820915e-07,
+      "loss": 0.3538,
+      "step": 2213
+    },
+    {
+      "epoch": 2.905511811023622,
+      "grad_norm": 0.04214561215882684,
+      "learning_rate": 1.5521617824794387e-07,
+      "loss": 0.3536,
+      "step": 2214
+    },
+    {
+      "epoch": 2.906824146981627,
+      "grad_norm": 0.04078601040224686,
+      "learning_rate": 1.509970635736102e-07,
+      "loss": 0.3579,
+      "step": 2215
+    },
+    {
+      "epoch": 2.9081364829396326,
+      "grad_norm": 0.04195741740943959,
+      "learning_rate": 1.4683591045650814e-07,
+      "loss": 0.3589,
+      "step": 2216
+    },
+    {
+      "epoch": 2.909448818897638,
+      "grad_norm": 0.04057723401143457,
+      "learning_rate": 1.4273272860274023e-07,
+      "loss": 0.3535,
+      "step": 2217
+    },
+    {
+      "epoch": 2.910761154855643,
+      "grad_norm": 0.040496079076247256,
+      "learning_rate": 1.3868752758317838e-07,
+      "loss": 0.3593,
+      "step": 2218
+    },
+    {
+      "epoch": 2.9120734908136483,
+      "grad_norm": 0.04124944207741271,
+      "learning_rate": 1.3470031683346651e-07,
+      "loss": 0.3583,
+      "step": 2219
+    },
+    {
+      "epoch": 2.9133858267716537,
+      "grad_norm": 0.04443671982215531,
+      "learning_rate": 1.3077110565397067e-07,
+      "loss": 0.361,
+      "step": 2220
+    },
+    {
+      "epoch": 2.9146981627296586,
+      "grad_norm": 0.04079350127107068,
+      "learning_rate": 1.2689990320977896e-07,
+      "loss": 0.3525,
+      "step": 2221
+    },
+    {
+      "epoch": 2.916010498687664,
+      "grad_norm": 0.04259480953398898,
+      "learning_rate": 1.2308671853066278e-07,
+      "loss": 0.3632,
+      "step": 2222
+    },
+    {
+      "epoch": 2.9173228346456694,
+      "grad_norm": 0.04168547691941991,
+      "learning_rate": 1.193315605110712e-07,
+      "loss": 0.3584,
+      "step": 2223
+    },
+    {
+      "epoch": 2.9186351706036744,
+      "grad_norm": 0.04103260246066425,
+      "learning_rate": 1.1563443791009765e-07,
+      "loss": 0.3513,
+      "step": 2224
+    },
+    {
+      "epoch": 2.91994750656168,
+      "grad_norm": 0.04095257620744735,
+      "learning_rate": 1.1199535935147165e-07,
+      "loss": 0.3489,
+      "step": 2225
+    },
+    {
+      "epoch": 2.921259842519685,
+      "grad_norm": 0.04607567464852418,
+      "learning_rate": 1.0841433332352546e-07,
+      "loss": 0.3559,
+      "step": 2226
+    },
+    {
+      "epoch": 2.92257217847769,
+      "grad_norm": 0.04255326304026008,
+      "learning_rate": 1.0489136817918576e-07,
+      "loss": 0.3605,
+      "step": 2227
+    },
+    {
+      "epoch": 2.9238845144356955,
+      "grad_norm": 0.04105675727880427,
+      "learning_rate": 1.0142647213594314e-07,
+      "loss": 0.354,
+      "step": 2228
+    },
+    {
+      "epoch": 2.925196850393701,
+      "grad_norm": 0.04126728467089348,
+      "learning_rate": 9.80196532758465e-08,
+      "loss": 0.3597,
+      "step": 2229
+    },
+    {
+      "epoch": 2.926509186351706,
+      "grad_norm": 0.042190610026549155,
+      "learning_rate": 9.467091954547258e-08,
+      "loss": 0.3502,
+      "step": 2230
+    },
+    {
+      "epoch": 2.9278215223097113,
+      "grad_norm": 0.04140919008909004,
+      "learning_rate": 9.138027875591205e-08,
+      "loss": 0.3499,
+      "step": 2231
+    },
+    {
+      "epoch": 2.9291338582677167,
+      "grad_norm": 0.0430428838509556,
+      "learning_rate": 8.814773858275004e-08,
+      "loss": 0.3627,
+      "step": 2232
+    },
+    {
+      "epoch": 2.9304461942257216,
+      "grad_norm": 0.042809550864385496,
+      "learning_rate": 8.49733065660524e-08,
+      "loss": 0.3653,
+      "step": 2233
+    },
+    {
+      "epoch": 2.931758530183727,
+      "grad_norm": 0.04179475737611144,
+      "learning_rate": 8.18569901103433e-08,
+      "loss": 0.3587,
+      "step": 2234
+    },
+    {
+      "epoch": 2.9330708661417324,
+      "grad_norm": 0.043200203787640994,
+      "learning_rate": 7.879879648458321e-08,
+      "loss": 0.3528,
+      "step": 2235
+    },
+    {
+      "epoch": 2.9343832020997374,
+      "grad_norm": 0.04175559424938238,
+      "learning_rate": 7.579873282216599e-08,
+      "loss": 0.3607,
+      "step": 2236
+    },
+    {
+      "epoch": 2.9356955380577427,
+      "grad_norm": 0.042890549957018506,
+      "learning_rate": 7.285680612089673e-08,
+      "loss": 0.3488,
+      "step": 2237
+    },
+    {
+      "epoch": 2.937007874015748,
+      "grad_norm": 0.04166542139417922,
+      "learning_rate": 6.99730232429613e-08,
+      "loss": 0.3538,
+      "step": 2238
+    },
+    {
+      "epoch": 2.938320209973753,
+      "grad_norm": 0.041445913577836044,
+      "learning_rate": 6.71473909149345e-08,
+      "loss": 0.347,
+      "step": 2239
+    },
+    {
+      "epoch": 2.9396325459317585,
+      "grad_norm": 0.041489026212925995,
+      "learning_rate": 6.437991572774415e-08,
+      "loss": 0.3559,
+      "step": 2240
+    },
+    {
+      "epoch": 2.940944881889764,
+      "grad_norm": 0.04206243118958933,
+      "learning_rate": 6.167060413666825e-08,
+      "loss": 0.3596,
+      "step": 2241
+    },
+    {
+      "epoch": 2.942257217847769,
+      "grad_norm": 0.04057559463105476,
+      "learning_rate": 5.901946246130996e-08,
+      "loss": 0.3531,
+      "step": 2242
+    },
+    {
+      "epoch": 2.9435695538057742,
+      "grad_norm": 0.04283603126591131,
+      "learning_rate": 5.642649688559487e-08,
+      "loss": 0.3582,
+      "step": 2243
+    },
+    {
+      "epoch": 2.9448818897637796,
+      "grad_norm": 0.04402566243856241,
+      "learning_rate": 5.3891713457743244e-08,
+      "loss": 0.3517,
+      "step": 2244
+    },
+    {
+      "epoch": 2.9461942257217846,
+      "grad_norm": 0.04248937543554805,
+      "learning_rate": 5.141511809027e-08,
+      "loss": 0.3512,
+      "step": 2245
+    },
+    {
+      "epoch": 2.94750656167979,
+      "grad_norm": 0.0412665717059966,
+      "learning_rate": 4.899671655995419e-08,
+      "loss": 0.3598,
+      "step": 2246
+    },
+    {
+      "epoch": 2.9488188976377954,
+      "grad_norm": 0.04336398210648325,
+      "learning_rate": 4.6636514507844566e-08,
+      "loss": 0.357,
+      "step": 2247
+    },
+    {
+      "epoch": 2.9501312335958003,
+      "grad_norm": 0.04317626972320007,
+      "learning_rate": 4.433451743922623e-08,
+      "loss": 0.3526,
+      "step": 2248
+    },
+    {
+      "epoch": 2.9514435695538057,
+      "grad_norm": 0.04400118394363825,
+      "learning_rate": 4.209073072362901e-08,
+      "loss": 0.3557,
+      "step": 2249
+    },
+    {
+      "epoch": 2.952755905511811,
+      "grad_norm": 0.04284019826086388,
+      "learning_rate": 3.990515959479968e-08,
+      "loss": 0.348,
+      "step": 2250
+    },
+    {
+      "epoch": 2.954068241469816,
+      "grad_norm": 0.041114948726323416,
+      "learning_rate": 3.777780915068807e-08,
+      "loss": 0.364,
+      "step": 2251
+    },
+    {
+      "epoch": 2.9553805774278215,
+      "grad_norm": 0.043065163512995254,
+      "learning_rate": 3.570868435344987e-08,
+      "loss": 0.353,
+      "step": 2252
+    },
+    {
+      "epoch": 2.956692913385827,
+      "grad_norm": 0.04467226755384039,
+      "learning_rate": 3.369779002942442e-08,
+      "loss": 0.3599,
+      "step": 2253
+    },
+    {
+      "epoch": 2.958005249343832,
+      "grad_norm": 0.04269137212842603,
+      "learning_rate": 3.1745130869123566e-08,
+      "loss": 0.3524,
+      "step": 2254
+    },
+    {
+      "epoch": 2.959317585301837,
+      "grad_norm": 0.041564758520685016,
+      "learning_rate": 2.9850711427220626e-08,
+      "loss": 0.362,
+      "step": 2255
+    },
+    {
+      "epoch": 2.9606299212598426,
+      "grad_norm": 0.04292994261667837,
+      "learning_rate": 2.801453612255034e-08,
+      "loss": 0.3581,
+      "step": 2256
+    },
+    {
+      "epoch": 2.9619422572178475,
+      "grad_norm": 0.04080861037580894,
+      "learning_rate": 2.6236609238083908e-08,
+      "loss": 0.3606,
+      "step": 2257
+    },
+    {
+      "epoch": 2.963254593175853,
+      "grad_norm": 0.04688724360316701,
+      "learning_rate": 2.4516934920920663e-08,
+      "loss": 0.3598,
+      "step": 2258
+    },
+    {
+      "epoch": 2.9645669291338583,
+      "grad_norm": 0.04077535943968583,
+      "learning_rate": 2.2855517182290842e-08,
+      "loss": 0.357,
+      "step": 2259
+    },
+    {
+      "epoch": 2.9658792650918633,
+      "grad_norm": 0.0430244602585007,
+      "learning_rate": 2.1252359897536154e-08,
+      "loss": 0.3551,
+      "step": 2260
+    },
+    {
+      "epoch": 2.9671916010498687,
+      "grad_norm": 0.042344742178167784,
+      "learning_rate": 1.9707466806104248e-08,
+      "loss": 0.3601,
+      "step": 2261
+    },
+    {
+      "epoch": 2.968503937007874,
+      "grad_norm": 0.041111142768341144,
+      "learning_rate": 1.8220841511532028e-08,
+      "loss": 0.3542,
+      "step": 2262
+    },
+    {
+      "epoch": 2.969816272965879,
+      "grad_norm": 0.04262908203114843,
+      "learning_rate": 1.6792487481451236e-08,
+      "loss": 0.3616,
+      "step": 2263
+    },
+    {
+      "epoch": 2.9711286089238844,
+      "grad_norm": 0.04232819349283561,
+      "learning_rate": 1.542240804756623e-08,
+      "loss": 0.3556,
+      "step": 2264
+    },
+    {
+      "epoch": 2.97244094488189,
+      "grad_norm": 0.041666090579798754,
+      "learning_rate": 1.4110606405662308e-08,
+      "loss": 0.3588,
+      "step": 2265
+    },
+    {
+      "epoch": 2.9737532808398948,
+      "grad_norm": 0.0416867998324897,
+      "learning_rate": 1.2857085615577968e-08,
+      "loss": 0.3595,
+      "step": 2266
+    },
+    {
+      "epoch": 2.9750656167979,
+      "grad_norm": 0.045017697220355986,
+      "learning_rate": 1.1661848601215997e-08,
+      "loss": 0.3571,
+      "step": 2267
+    },
+    {
+      "epoch": 2.9763779527559056,
+      "grad_norm": 0.04237838537790356,
+      "learning_rate": 1.0524898150529594e-08,
+      "loss": 0.3563,
+      "step": 2268
+    },
+    {
+      "epoch": 2.9776902887139105,
+      "grad_norm": 0.042434011968408764,
+      "learning_rate": 9.4462369155085e-09,
+      "loss": 0.3518,
+      "step": 2269
+    },
+    {
+      "epoch": 2.979002624671916,
+      "grad_norm": 0.04181642537450737,
+      "learning_rate": 8.42586741219009e-09,
+      "loss": 0.3515,
+      "step": 2270
+    },
+    {
+      "epoch": 2.9803149606299213,
+      "grad_norm": 0.041462716259445094,
+      "learning_rate": 7.463792020631633e-09,
+      "loss": 0.363,
+      "step": 2271
+    },
+    {
+      "epoch": 2.9816272965879262,
+      "grad_norm": 0.042048048251802736,
+      "learning_rate": 6.560012984926922e-09,
+      "loss": 0.3655,
+      "step": 2272
+    },
+    {
+      "epoch": 2.9829396325459316,
+      "grad_norm": 0.04271360068598683,
+      "learning_rate": 5.714532413186868e-09,
+      "loss": 0.3566,
+      "step": 2273
+    },
+    {
+      "epoch": 2.984251968503937,
+      "grad_norm": 0.0424775385592596,
+      "learning_rate": 4.927352277536712e-09,
+      "loss": 0.3568,
+      "step": 2274
+    },
+    {
+      "epoch": 2.985564304461942,
+      "grad_norm": 0.043560111550625676,
+      "learning_rate": 4.198474414110476e-09,
+      "loss": 0.3554,
+      "step": 2275
+    },
+    {
+      "epoch": 2.9868766404199474,
+      "grad_norm": 0.04066774761223068,
+      "learning_rate": 3.5279005230565155e-09,
+      "loss": 0.3567,
+      "step": 2276
+    },
+    {
+      "epoch": 2.9881889763779528,
+      "grad_norm": 0.04634562286346347,
+      "learning_rate": 2.915632168523641e-09,
+      "loss": 0.3654,
+      "step": 2277
+    },
+    {
+      "epoch": 2.9895013123359577,
+      "grad_norm": 0.04266116628911044,
+      "learning_rate": 2.361670778655567e-09,
+      "loss": 0.357,
+      "step": 2278
+    },
+    {
+      "epoch": 2.990813648293963,
+      "grad_norm": 0.04219000781785991,
+      "learning_rate": 1.866017645596463e-09,
+      "loss": 0.3602,
+      "step": 2279
+    },
+    {
+      "epoch": 2.9921259842519685,
+      "grad_norm": 0.046397454936986854,
+      "learning_rate": 1.428673925479851e-09,
+      "loss": 0.3547,
+      "step": 2280
+    },
+    {
+      "epoch": 2.993438320209974,
+      "grad_norm": 0.04217233734545585,
+      "learning_rate": 1.049640638436933e-09,
+      "loss": 0.3578,
+      "step": 2281
+    },
+    {
+      "epoch": 2.994750656167979,
+      "grad_norm": 0.04167866515890675,
+      "learning_rate": 7.289186685799365e-10,
+      "loss": 0.3587,
+      "step": 2282
+    },
+    {
+      "epoch": 2.9960629921259843,
+      "grad_norm": 0.040555992008321884,
+      "learning_rate": 4.66508764007667e-10,
+      "loss": 0.3563,
+      "step": 2283
+    },
+    {
+      "epoch": 2.9973753280839897,
+      "grad_norm": 0.04250366163982152,
+      "learning_rate": 2.624115368055069e-10,
+      "loss": 0.3556,
+      "step": 2284
+    },
+    {
+      "epoch": 2.9986876640419946,
+      "grad_norm": 0.05112946273633736,
+      "learning_rate": 1.166274630398645e-10,
+      "loss": 0.3603,
+      "step": 2285
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.047543132569776575,
+      "learning_rate": 2.915688276372563e-11,
+      "loss": 0.3418,
+      "step": 2286
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2286,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.287767603502121e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}