diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4859676831490706,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002429838415745353,
+      "grad_norm": 0.9283741116523743,
+      "learning_rate": 8e-05,
+      "loss": 1.9395,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004859676831490706,
+      "grad_norm": 1.3144824504852295,
+      "learning_rate": 8e-05,
+      "loss": 1.9039,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007289515247236059,
+      "grad_norm": 0.5672217011451721,
+      "learning_rate": 8e-05,
+      "loss": 1.8385,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009719353662981412,
+      "grad_norm": 0.6657972931861877,
+      "learning_rate": 8e-05,
+      "loss": 1.9482,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012149192078726764,
+      "grad_norm": 0.512556791305542,
+      "learning_rate": 8e-05,
+      "loss": 1.9696,
+      "step": 5
+    },
+    {
+      "epoch": 0.0014579030494472117,
+      "grad_norm": 0.5714458227157593,
+      "learning_rate": 8e-05,
+      "loss": 1.8831,
+      "step": 6
+    },
+    {
+      "epoch": 0.001700886891021747,
+      "grad_norm": 0.6022382378578186,
+      "learning_rate": 8e-05,
+      "loss": 1.9216,
+      "step": 7
+    },
+    {
+      "epoch": 0.0019438707325962824,
+      "grad_norm": 0.5293871164321899,
+      "learning_rate": 8e-05,
+      "loss": 1.8539,
+      "step": 8
+    },
+    {
+      "epoch": 0.0021868545741708175,
+      "grad_norm": 0.5279130339622498,
+      "learning_rate": 8e-05,
+      "loss": 1.7731,
+      "step": 9
+    },
+    {
+      "epoch": 0.0024298384157453528,
+      "grad_norm": 0.4825509190559387,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 10
+    },
+    {
+      "epoch": 0.002672822257319888,
+      "grad_norm": 0.464072048664093,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 11
+    },
+    {
+      "epoch": 0.0029158060988944234,
+      "grad_norm": 0.514319658279419,
+      "learning_rate": 8e-05,
+      "loss": 1.9682,
+      "step": 12
+    },
+    {
+      "epoch": 0.0031587899404689587,
+      "grad_norm": 0.4752178490161896,
+      "learning_rate": 8e-05,
+      "loss": 1.7164,
+      "step": 13
+    },
+    {
+      "epoch": 0.003401773782043494,
+      "grad_norm": 0.5416842103004456,
+      "learning_rate": 8e-05,
+      "loss": 1.8342,
+      "step": 14
+    },
+    {
+      "epoch": 0.0036447576236180294,
+      "grad_norm": 0.513598620891571,
+      "learning_rate": 8e-05,
+      "loss": 1.9164,
+      "step": 15
+    },
+    {
+      "epoch": 0.0038877414651925647,
+      "grad_norm": 0.4487161934375763,
+      "learning_rate": 8e-05,
+      "loss": 1.7595,
+      "step": 16
+    },
+    {
+      "epoch": 0.0041307253067671,
+      "grad_norm": 0.5007575750350952,
+      "learning_rate": 8e-05,
+      "loss": 1.9731,
+      "step": 17
+    },
+    {
+      "epoch": 0.004373709148341635,
+      "grad_norm": 0.4565035402774811,
+      "learning_rate": 8e-05,
+      "loss": 1.7485,
+      "step": 18
+    },
+    {
+      "epoch": 0.00461669298991617,
+      "grad_norm": 0.4516490399837494,
+      "learning_rate": 8e-05,
+      "loss": 1.7877,
+      "step": 19
+    },
+    {
+      "epoch": 0.0048596768314907056,
+      "grad_norm": 0.4720084071159363,
+      "learning_rate": 8e-05,
+      "loss": 1.8198,
+      "step": 20
+    },
+    {
+      "epoch": 0.005102660673065241,
+      "grad_norm": 0.46458321809768677,
+      "learning_rate": 8e-05,
+      "loss": 1.7779,
+      "step": 21
+    },
+    {
+      "epoch": 0.005345644514639776,
+      "grad_norm": 0.46023452281951904,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 22
+    },
+    {
+      "epoch": 0.0055886283562143115,
+      "grad_norm": 0.45890012383461,
+      "learning_rate": 8e-05,
+      "loss": 1.8359,
+      "step": 23
+    },
+    {
+      "epoch": 0.005831612197788847,
+      "grad_norm": 0.4736693799495697,
+      "learning_rate": 8e-05,
+      "loss": 1.8834,
+      "step": 24
+    },
+    {
+      "epoch": 0.006074596039363382,
+      "grad_norm": 0.47544634342193604,
+      "learning_rate": 8e-05,
+      "loss": 1.8574,
+      "step": 25
+    },
+    {
+      "epoch": 0.0063175798809379175,
+      "grad_norm": 0.48038774728775024,
+      "learning_rate": 8e-05,
+      "loss": 1.9607,
+      "step": 26
+    },
+    {
+      "epoch": 0.006560563722512453,
+      "grad_norm": 0.4817793369293213,
+      "learning_rate": 8e-05,
+      "loss": 1.8488,
+      "step": 27
+    },
+    {
+      "epoch": 0.006803547564086988,
+      "grad_norm": 0.42538994550704956,
+      "learning_rate": 8e-05,
+      "loss": 1.7048,
+      "step": 28
+    },
+    {
+      "epoch": 0.0070465314056615234,
+      "grad_norm": 0.4736614525318146,
+      "learning_rate": 8e-05,
+      "loss": 1.882,
+      "step": 29
+    },
+    {
+      "epoch": 0.007289515247236059,
+      "grad_norm": 0.42797160148620605,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 30
+    },
+    {
+      "epoch": 0.007532499088810594,
+      "grad_norm": 0.4591984450817108,
+      "learning_rate": 8e-05,
+      "loss": 1.8623,
+      "step": 31
+    },
+    {
+      "epoch": 0.007775482930385129,
+      "grad_norm": 0.4495055377483368,
+      "learning_rate": 8e-05,
+      "loss": 1.8061,
+      "step": 32
+    },
+    {
+      "epoch": 0.008018466771959665,
+      "grad_norm": 0.44170689582824707,
+      "learning_rate": 8e-05,
+      "loss": 1.8883,
+      "step": 33
+    },
+    {
+      "epoch": 0.0082614506135342,
+      "grad_norm": 0.4610385298728943,
+      "learning_rate": 8e-05,
+      "loss": 1.9759,
+      "step": 34
+    },
+    {
+      "epoch": 0.008504434455108735,
+      "grad_norm": 0.4396963119506836,
+      "learning_rate": 8e-05,
+      "loss": 1.7499,
+      "step": 35
+    },
+    {
+      "epoch": 0.00874741829668327,
+      "grad_norm": 0.4422125220298767,
+      "learning_rate": 8e-05,
+      "loss": 1.8818,
+      "step": 36
+    },
+    {
+      "epoch": 0.008990402138257806,
+      "grad_norm": 0.42213165760040283,
+      "learning_rate": 8e-05,
+      "loss": 1.7612,
+      "step": 37
+    },
+    {
+      "epoch": 0.00923338597983234,
+      "grad_norm": 0.4536469280719757,
+      "learning_rate": 8e-05,
+      "loss": 1.8168,
+      "step": 38
+    },
+    {
+      "epoch": 0.009476369821406877,
+      "grad_norm": 0.4328950345516205,
+      "learning_rate": 8e-05,
+      "loss": 1.8431,
+      "step": 39
+    },
+    {
+      "epoch": 0.009719353662981411,
+      "grad_norm": 0.4586104452610016,
+      "learning_rate": 8e-05,
+      "loss": 1.9572,
+      "step": 40
+    },
+    {
+      "epoch": 0.009962337504555947,
+      "grad_norm": 0.4931277632713318,
+      "learning_rate": 8e-05,
+      "loss": 2.0045,
+      "step": 41
+    },
+    {
+      "epoch": 0.010205321346130482,
+      "grad_norm": 0.4379693269729614,
+      "learning_rate": 8e-05,
+      "loss": 1.7844,
+      "step": 42
+    },
+    {
+      "epoch": 0.010448305187705018,
+      "grad_norm": 0.4282945394515991,
+      "learning_rate": 8e-05,
+      "loss": 1.7884,
+      "step": 43
+    },
+    {
+      "epoch": 0.010691289029279552,
+      "grad_norm": 0.4561641812324524,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 44
+    },
+    {
+      "epoch": 0.010934272870854089,
+      "grad_norm": 0.4802621901035309,
+      "learning_rate": 8e-05,
+      "loss": 1.8825,
+      "step": 45
+    },
+    {
+      "epoch": 0.011177256712428623,
+      "grad_norm": 0.45682236552238464,
+      "learning_rate": 8e-05,
+      "loss": 1.9142,
+      "step": 46
+    },
+    {
+      "epoch": 0.01142024055400316,
+      "grad_norm": 0.4659646451473236,
+      "learning_rate": 8e-05,
+      "loss": 1.9928,
+      "step": 47
+    },
+    {
+      "epoch": 0.011663224395577694,
+      "grad_norm": 0.46792858839035034,
+      "learning_rate": 8e-05,
+      "loss": 1.7975,
+      "step": 48
+    },
+    {
+      "epoch": 0.01190620823715223,
+      "grad_norm": 0.42964687943458557,
+      "learning_rate": 8e-05,
+      "loss": 1.7913,
+      "step": 49
+    },
+    {
+      "epoch": 0.012149192078726764,
+      "grad_norm": 0.4474690854549408,
+      "learning_rate": 8e-05,
+      "loss": 1.8821,
+      "step": 50
+    },
+    {
+      "epoch": 0.0123921759203013,
+      "grad_norm": 0.43518373370170593,
+      "learning_rate": 8e-05,
+      "loss": 1.815,
+      "step": 51
+    },
+    {
+      "epoch": 0.012635159761875835,
+      "grad_norm": 0.41885894536972046,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 52
+    },
+    {
+      "epoch": 0.012878143603450371,
+      "grad_norm": 0.42882251739501953,
+      "learning_rate": 8e-05,
+      "loss": 1.8297,
+      "step": 53
+    },
+    {
+      "epoch": 0.013121127445024906,
+      "grad_norm": 0.4267278015613556,
+      "learning_rate": 8e-05,
+      "loss": 1.8158,
+      "step": 54
+    },
+    {
+      "epoch": 0.013364111286599442,
+      "grad_norm": 0.4405764937400818,
+      "learning_rate": 8e-05,
+      "loss": 1.9295,
+      "step": 55
+    },
+    {
+      "epoch": 0.013607095128173976,
+      "grad_norm": 0.45413410663604736,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 56
+    },
+    {
+      "epoch": 0.013850078969748512,
+      "grad_norm": 0.4386683404445648,
+      "learning_rate": 8e-05,
+      "loss": 1.8841,
+      "step": 57
+    },
+    {
+      "epoch": 0.014093062811323047,
+      "grad_norm": 0.4813304841518402,
+      "learning_rate": 8e-05,
+      "loss": 1.9317,
+      "step": 58
+    },
+    {
+      "epoch": 0.014336046652897583,
+      "grad_norm": 0.4510192573070526,
+      "learning_rate": 8e-05,
+      "loss": 1.7689,
+      "step": 59
+    },
+    {
+      "epoch": 0.014579030494472118,
+      "grad_norm": 0.42841628193855286,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 60
+    },
+    {
+      "epoch": 0.014822014336046654,
+      "grad_norm": 0.4288236200809479,
+      "learning_rate": 8e-05,
+      "loss": 1.8038,
+      "step": 61
+    },
+    {
+      "epoch": 0.015064998177621188,
+      "grad_norm": 0.41781896352767944,
+      "learning_rate": 8e-05,
+      "loss": 1.8636,
+      "step": 62
+    },
+    {
+      "epoch": 0.015307982019195723,
+      "grad_norm": 0.4112628996372223,
+      "learning_rate": 8e-05,
+      "loss": 1.7913,
+      "step": 63
+    },
+    {
+      "epoch": 0.015550965860770259,
+      "grad_norm": 0.4400574564933777,
+      "learning_rate": 8e-05,
+      "loss": 1.78,
+      "step": 64
+    },
+    {
+      "epoch": 0.015793949702344795,
+      "grad_norm": 0.4312531352043152,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 65
+    },
+    {
+      "epoch": 0.01603693354391933,
+      "grad_norm": 0.4226854145526886,
+      "learning_rate": 8e-05,
+      "loss": 1.8446,
+      "step": 66
+    },
+    {
+      "epoch": 0.016279917385493864,
+      "grad_norm": 0.4536939561367035,
+      "learning_rate": 8e-05,
+      "loss": 1.8761,
+      "step": 67
+    },
+    {
+      "epoch": 0.0165229012270684,
+      "grad_norm": 0.44555526971817017,
+      "learning_rate": 8e-05,
+      "loss": 1.7852,
+      "step": 68
+    },
+    {
+      "epoch": 0.016765885068642936,
+      "grad_norm": 0.44413813948631287,
+      "learning_rate": 8e-05,
+      "loss": 1.7404,
+      "step": 69
+    },
+    {
+      "epoch": 0.01700886891021747,
+      "grad_norm": 0.4578434228897095,
+      "learning_rate": 8e-05,
+      "loss": 1.8257,
+      "step": 70
+    },
+    {
+      "epoch": 0.017251852751792005,
+      "grad_norm": 0.4213903546333313,
+      "learning_rate": 8e-05,
+      "loss": 1.7382,
+      "step": 71
+    },
+    {
+      "epoch": 0.01749483659336654,
+      "grad_norm": 0.4393966495990753,
+      "learning_rate": 8e-05,
+      "loss": 1.8235,
+      "step": 72
+    },
+    {
+      "epoch": 0.017737820434941078,
+      "grad_norm": 0.43298789858818054,
+      "learning_rate": 8e-05,
+      "loss": 1.9501,
+      "step": 73
+    },
+    {
+      "epoch": 0.017980804276515612,
+      "grad_norm": 0.43117642402648926,
+      "learning_rate": 8e-05,
+      "loss": 1.8529,
+      "step": 74
+    },
+    {
+      "epoch": 0.018223788118090146,
+      "grad_norm": 0.4453331530094147,
+      "learning_rate": 8e-05,
+      "loss": 1.8208,
+      "step": 75
+    },
+    {
+      "epoch": 0.01846677195966468,
+      "grad_norm": 0.4348840117454529,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 76
+    },
+    {
+      "epoch": 0.01870975580123922,
+      "grad_norm": 0.4274575412273407,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 77
+    },
+    {
+      "epoch": 0.018952739642813753,
+      "grad_norm": 0.45439624786376953,
+      "learning_rate": 8e-05,
+      "loss": 1.8004,
+      "step": 78
+    },
+    {
+      "epoch": 0.019195723484388288,
+      "grad_norm": 0.431682288646698,
+      "learning_rate": 8e-05,
+      "loss": 1.7191,
+      "step": 79
+    },
+    {
+      "epoch": 0.019438707325962822,
+      "grad_norm": 0.43740177154541016,
+      "learning_rate": 8e-05,
+      "loss": 1.7335,
+      "step": 80
+    },
+    {
+      "epoch": 0.01968169116753736,
+      "grad_norm": 0.44585227966308594,
+      "learning_rate": 8e-05,
+      "loss": 1.8385,
+      "step": 81
+    },
+    {
+      "epoch": 0.019924675009111895,
+      "grad_norm": 0.41222020983695984,
+      "learning_rate": 8e-05,
+      "loss": 1.8581,
+      "step": 82
+    },
+    {
+      "epoch": 0.02016765885068643,
+      "grad_norm": 0.4409661889076233,
+      "learning_rate": 8e-05,
+      "loss": 1.7644,
+      "step": 83
+    },
+    {
+      "epoch": 0.020410642692260964,
+      "grad_norm": 0.4678371846675873,
+      "learning_rate": 8e-05,
+      "loss": 1.9272,
+      "step": 84
+    },
+    {
+      "epoch": 0.0206536265338355,
+      "grad_norm": 0.43374109268188477,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 85
+    },
+    {
+      "epoch": 0.020896610375410036,
+      "grad_norm": 0.4561867117881775,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 86
+    },
+    {
+      "epoch": 0.02113959421698457,
+      "grad_norm": 0.4501771628856659,
+      "learning_rate": 8e-05,
+      "loss": 1.791,
+      "step": 87
+    },
+    {
+      "epoch": 0.021382578058559105,
+      "grad_norm": 0.4467450678348541,
+      "learning_rate": 8e-05,
+      "loss": 1.6327,
+      "step": 88
+    },
+    {
+      "epoch": 0.021625561900133643,
+      "grad_norm": 0.4093148112297058,
+      "learning_rate": 8e-05,
+      "loss": 1.682,
+      "step": 89
+    },
+    {
+      "epoch": 0.021868545741708177,
+      "grad_norm": 0.4533080458641052,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 90
+    },
+    {
+      "epoch": 0.02211152958328271,
+      "grad_norm": 0.48888301849365234,
+      "learning_rate": 8e-05,
+      "loss": 1.7748,
+      "step": 91
+    },
+    {
+      "epoch": 0.022354513424857246,
+      "grad_norm": 0.43746045231819153,
+      "learning_rate": 8e-05,
+      "loss": 1.868,
+      "step": 92
+    },
+    {
+      "epoch": 0.022597497266431784,
+      "grad_norm": 0.4325760304927826,
+      "learning_rate": 8e-05,
+      "loss": 1.9195,
+      "step": 93
+    },
+    {
+      "epoch": 0.02284048110800632,
+      "grad_norm": 0.4253198802471161,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 94
+    },
+    {
+      "epoch": 0.023083464949580853,
+      "grad_norm": 0.44341036677360535,
+      "learning_rate": 8e-05,
+      "loss": 1.7678,
+      "step": 95
+    },
+    {
+      "epoch": 0.023326448791155387,
+      "grad_norm": 0.40490806102752686,
+      "learning_rate": 8e-05,
+      "loss": 1.7993,
+      "step": 96
+    },
+    {
+      "epoch": 0.023569432632729922,
+      "grad_norm": 0.4402480125427246,
+      "learning_rate": 8e-05,
+      "loss": 1.8517,
+      "step": 97
+    },
+    {
+      "epoch": 0.02381241647430446,
+      "grad_norm": 0.4148694574832916,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 98
+    },
+    {
+      "epoch": 0.024055400315878994,
+      "grad_norm": 0.4118325412273407,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 99
+    },
+    {
+      "epoch": 0.02429838415745353,
+      "grad_norm": 0.4318928122520447,
+      "learning_rate": 8e-05,
+      "loss": 1.9232,
+      "step": 100
+    },
+    {
+      "epoch": 0.024541367999028063,
+      "grad_norm": 0.43065184354782104,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 101
+    },
+    {
+      "epoch": 0.0247843518406026,
+      "grad_norm": 0.4622381627559662,
+      "learning_rate": 8e-05,
+      "loss": 1.7769,
+      "step": 102
+    },
+    {
+      "epoch": 0.025027335682177135,
+      "grad_norm": 0.4534037411212921,
+      "learning_rate": 8e-05,
+      "loss": 1.9383,
+      "step": 103
+    },
+    {
+      "epoch": 0.02527031952375167,
+      "grad_norm": 0.4303297996520996,
+      "learning_rate": 8e-05,
+      "loss": 1.8807,
+      "step": 104
+    },
+    {
+      "epoch": 0.025513303365326204,
+      "grad_norm": 0.44953134655952454,
+      "learning_rate": 8e-05,
+      "loss": 1.9548,
+      "step": 105
+    },
+    {
+      "epoch": 0.025756287206900742,
+      "grad_norm": 0.4983407258987427,
+      "learning_rate": 8e-05,
+      "loss": 1.8985,
+      "step": 106
+    },
+    {
+      "epoch": 0.025999271048475277,
+      "grad_norm": 0.4147990942001343,
+      "learning_rate": 8e-05,
+      "loss": 1.74,
+      "step": 107
+    },
+    {
+      "epoch": 0.02624225489004981,
+      "grad_norm": 0.41655194759368896,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 108
+    },
+    {
+      "epoch": 0.026485238731624346,
+      "grad_norm": 0.43446749448776245,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 109
+    },
+    {
+      "epoch": 0.026728222573198884,
+      "grad_norm": 0.43117472529411316,
+      "learning_rate": 8e-05,
+      "loss": 1.8309,
+      "step": 110
+    },
+    {
+      "epoch": 0.026971206414773418,
+      "grad_norm": 0.4523254334926605,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 111
+    },
+    {
+      "epoch": 0.027214190256347952,
+      "grad_norm": 0.495615690946579,
+      "learning_rate": 8e-05,
+      "loss": 2.0216,
+      "step": 112
+    },
+    {
+      "epoch": 0.027457174097922487,
+      "grad_norm": 0.437045693397522,
+      "learning_rate": 8e-05,
+      "loss": 1.81,
+      "step": 113
+    },
+    {
+      "epoch": 0.027700157939497025,
+      "grad_norm": 0.44665664434432983,
+      "learning_rate": 8e-05,
+      "loss": 1.8155,
+      "step": 114
+    },
+    {
+      "epoch": 0.02794314178107156,
+      "grad_norm": 0.3989410102367401,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 115
+    },
+    {
+      "epoch": 0.028186125622646094,
+      "grad_norm": 0.4348151981830597,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 116
+    },
+    {
+      "epoch": 0.028429109464220628,
+      "grad_norm": 0.4325806498527527,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 117
+    },
+    {
+      "epoch": 0.028672093305795166,
+      "grad_norm": 0.44066253304481506,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 118
+    },
+    {
+      "epoch": 0.0289150771473697,
+      "grad_norm": 0.4325154423713684,
+      "learning_rate": 8e-05,
+      "loss": 1.8724,
+      "step": 119
+    },
+    {
+      "epoch": 0.029158060988944235,
+      "grad_norm": 0.4403970539569855,
+      "learning_rate": 8e-05,
+      "loss": 1.7267,
+      "step": 120
+    },
+    {
+      "epoch": 0.02940104483051877,
+      "grad_norm": 0.44233599305152893,
+      "learning_rate": 8e-05,
+      "loss": 1.8199,
+      "step": 121
+    },
+    {
+      "epoch": 0.029644028672093307,
+      "grad_norm": 0.41565191745758057,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 122
+    },
+    {
+      "epoch": 0.029887012513667842,
+      "grad_norm": 0.4531630277633667,
+      "learning_rate": 8e-05,
+      "loss": 1.9043,
+      "step": 123
+    },
+    {
+      "epoch": 0.030129996355242376,
+      "grad_norm": 0.3977494239807129,
+      "learning_rate": 8e-05,
+      "loss": 1.6552,
+      "step": 124
+    },
+    {
+      "epoch": 0.03037298019681691,
+      "grad_norm": 0.40935423970222473,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 125
+    },
+    {
+      "epoch": 0.030615964038391445,
+      "grad_norm": 0.4547923505306244,
+      "learning_rate": 8e-05,
+      "loss": 1.8269,
+      "step": 126
+    },
+    {
+      "epoch": 0.030858947879965983,
+      "grad_norm": 0.4384150505065918,
+      "learning_rate": 8e-05,
+      "loss": 1.9686,
+      "step": 127
+    },
+    {
+      "epoch": 0.031101931721540518,
+      "grad_norm": 0.4189033508300781,
+      "learning_rate": 8e-05,
+      "loss": 1.5945,
+      "step": 128
+    },
+    {
+      "epoch": 0.031344915563115056,
+      "grad_norm": 0.4254000186920166,
+      "learning_rate": 8e-05,
+      "loss": 1.8124,
+      "step": 129
+    },
+    {
+      "epoch": 0.03158789940468959,
+      "grad_norm": 0.44288694858551025,
+      "learning_rate": 8e-05,
+      "loss": 1.8364,
+      "step": 130
+    },
+    {
+      "epoch": 0.031830883246264124,
+      "grad_norm": 0.4456717073917389,
+      "learning_rate": 8e-05,
+      "loss": 1.8809,
+      "step": 131
+    },
+    {
+      "epoch": 0.03207386708783866,
+      "grad_norm": 0.41588082909584045,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 132
+    },
+    {
+      "epoch": 0.03231685092941319,
+      "grad_norm": 0.43358904123306274,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 133
+    },
+    {
+      "epoch": 0.03255983477098773,
+      "grad_norm": 0.4369245171546936,
+      "learning_rate": 8e-05,
+      "loss": 1.8364,
+      "step": 134
+    },
+    {
+      "epoch": 0.03280281861256226,
+      "grad_norm": 0.4606565237045288,
+      "learning_rate": 8e-05,
+      "loss": 1.9564,
+      "step": 135
+    },
+    {
+      "epoch": 0.0330458024541368,
+      "grad_norm": 0.44513610005378723,
+      "learning_rate": 8e-05,
+      "loss": 1.8491,
+      "step": 136
+    },
+    {
+      "epoch": 0.03328878629571134,
+      "grad_norm": 0.41755104064941406,
+      "learning_rate": 8e-05,
+      "loss": 1.7142,
+      "step": 137
+    },
+    {
+      "epoch": 0.03353177013728587,
+      "grad_norm": 0.42471590638160706,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 138
+    },
+    {
+      "epoch": 0.03377475397886041,
+      "grad_norm": 0.43392664194107056,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 139
+    },
+    {
+      "epoch": 0.03401773782043494,
+      "grad_norm": 0.44673675298690796,
+      "learning_rate": 8e-05,
+      "loss": 1.8783,
+      "step": 140
+    },
+    {
+      "epoch": 0.034260721662009476,
+      "grad_norm": 0.43131935596466064,
+      "learning_rate": 8e-05,
+      "loss": 1.7653,
+      "step": 141
+    },
+    {
+      "epoch": 0.03450370550358401,
+      "grad_norm": 0.39308279752731323,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 142
+    },
+    {
+      "epoch": 0.034746689345158545,
+      "grad_norm": 0.42498987913131714,
+      "learning_rate": 8e-05,
+      "loss": 1.7597,
+      "step": 143
+    },
+    {
+      "epoch": 0.03498967318673308,
+      "grad_norm": 0.43941235542297363,
+      "learning_rate": 8e-05,
+      "loss": 1.7055,
+      "step": 144
+    },
+    {
+      "epoch": 0.03523265702830762,
+      "grad_norm": 0.4112749695777893,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 145
+    },
+    {
+      "epoch": 0.035475640869882155,
+      "grad_norm": 0.39101311564445496,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 146
+    },
+    {
+      "epoch": 0.03571862471145669,
+      "grad_norm": 0.429481565952301,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 147
+    },
+    {
+      "epoch": 0.035961608553031224,
+      "grad_norm": 0.40940847992897034,
+      "learning_rate": 8e-05,
+      "loss": 1.7589,
+      "step": 148
+    },
+    {
+      "epoch": 0.03620459239460576,
+      "grad_norm": 0.42588290572166443,
+      "learning_rate": 8e-05,
+      "loss": 1.8461,
+      "step": 149
+    },
+    {
+      "epoch": 0.03644757623618029,
+      "grad_norm": 0.4170544743537903,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 150
+    },
+    {
+      "epoch": 0.03669056007775483,
+      "grad_norm": 0.41155344247817993,
+      "learning_rate": 8e-05,
+      "loss": 1.7645,
+      "step": 151
+    },
+    {
+      "epoch": 0.03693354391932936,
+      "grad_norm": 0.4302254617214203,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 152
+    },
+    {
+      "epoch": 0.0371765277609039,
+      "grad_norm": 0.43057015538215637,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 153
+    },
+    {
+      "epoch": 0.03741951160247844,
+      "grad_norm": 0.41998055577278137,
+      "learning_rate": 8e-05,
+      "loss": 1.8225,
+      "step": 154
+    },
+    {
+      "epoch": 0.03766249544405297,
+      "grad_norm": 0.41916316747665405,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 155
+    },
+    {
+      "epoch": 0.03790547928562751,
+      "grad_norm": 0.42347627878189087,
+      "learning_rate": 8e-05,
+      "loss": 1.8304,
+      "step": 156
+    },
+    {
+      "epoch": 0.03814846312720204,
+      "grad_norm": 0.4198334515094757,
+      "learning_rate": 8e-05,
+      "loss": 1.7716,
+      "step": 157
+    },
+    {
+      "epoch": 0.038391446968776576,
+      "grad_norm": 0.4451964497566223,
+      "learning_rate": 8e-05,
+      "loss": 1.8368,
+      "step": 158
+    },
+    {
+      "epoch": 0.03863443081035111,
+      "grad_norm": 0.4325767755508423,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 159
+    },
+    {
+      "epoch": 0.038877414651925644,
+      "grad_norm": 0.43794333934783936,
+      "learning_rate": 8e-05,
+      "loss": 1.858,
+      "step": 160
+    },
+    {
+      "epoch": 0.03912039849350018,
+      "grad_norm": 0.4283236861228943,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 161
+    },
+    {
+      "epoch": 0.03936338233507472,
+      "grad_norm": 0.4178604185581207,
+      "learning_rate": 8e-05,
+      "loss": 1.6726,
+      "step": 162
+    },
+    {
+      "epoch": 0.039606366176649255,
+      "grad_norm": 0.43650540709495544,
+      "learning_rate": 8e-05,
+      "loss": 1.8255,
+      "step": 163
+    },
+    {
+      "epoch": 0.03984935001822379,
+      "grad_norm": 0.40848660469055176,
+      "learning_rate": 8e-05,
+      "loss": 1.6742,
+      "step": 164
+    },
+    {
+      "epoch": 0.040092333859798324,
+      "grad_norm": 0.4333600699901581,
+      "learning_rate": 8e-05,
+      "loss": 1.7795,
+      "step": 165
+    },
+    {
+      "epoch": 0.04033531770137286,
+      "grad_norm": 0.43268805742263794,
+      "learning_rate": 8e-05,
+      "loss": 1.7503,
+      "step": 166
+    },
+    {
+      "epoch": 0.04057830154294739,
+      "grad_norm": 0.39446377754211426,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 167
+    },
+    {
+      "epoch": 0.04082128538452193,
+      "grad_norm": 0.40882980823516846,
+      "learning_rate": 8e-05,
+      "loss": 1.6174,
+      "step": 168
+    },
+    {
+      "epoch": 0.04106426922609646,
+      "grad_norm": 0.43151581287384033,
+      "learning_rate": 8e-05,
+      "loss": 1.8183,
+      "step": 169
+    },
+    {
+      "epoch": 0.041307253067671,
+      "grad_norm": 0.43037307262420654,
+      "learning_rate": 8e-05,
+      "loss": 1.7799,
+      "step": 170
+    },
+    {
+      "epoch": 0.04155023690924554,
+      "grad_norm": 0.42919039726257324,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 171
+    },
+    {
+      "epoch": 0.04179322075082007,
+      "grad_norm": 0.41983699798583984,
+      "learning_rate": 8e-05,
+      "loss": 1.7359,
+      "step": 172
+    },
+    {
+      "epoch": 0.042036204592394606,
+      "grad_norm": 0.432204931974411,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 173
+    },
+    {
+      "epoch": 0.04227918843396914,
+      "grad_norm": 0.4395346939563751,
+      "learning_rate": 8e-05,
+      "loss": 1.7703,
+      "step": 174
+    },
+    {
+      "epoch": 0.042522172275543675,
+      "grad_norm": 0.4153292775154114,
+      "learning_rate": 8e-05,
+      "loss": 1.7165,
+      "step": 175
+    },
+    {
+      "epoch": 0.04276515611711821,
+      "grad_norm": 0.39776793122291565,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 176
+    },
+    {
+      "epoch": 0.043008139958692744,
+      "grad_norm": 0.4581263065338135,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 177
+    },
+    {
+      "epoch": 0.043251123800267285,
+      "grad_norm": 0.4424501657485962,
+      "learning_rate": 8e-05,
+      "loss": 1.8035,
+      "step": 178
+    },
+    {
+      "epoch": 0.04349410764184182,
+      "grad_norm": 0.480151891708374,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 179
+    },
+    {
+      "epoch": 0.043737091483416354,
+      "grad_norm": 0.4894304871559143,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 180
+    },
+    {
+      "epoch": 0.04398007532499089,
+      "grad_norm": 0.41660112142562866,
+      "learning_rate": 8e-05,
+      "loss": 1.7438,
+      "step": 181
+    },
+    {
+      "epoch": 0.04422305916656542,
+      "grad_norm": 0.41020017862319946,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 182
+    },
+    {
+      "epoch": 0.04446604300813996,
+      "grad_norm": 0.4876617193222046,
+      "learning_rate": 8e-05,
+      "loss": 1.9507,
+      "step": 183
+    },
+    {
+      "epoch": 0.04470902684971449,
+      "grad_norm": 0.4428721070289612,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 184
+    },
+    {
+      "epoch": 0.04495201069128903,
+      "grad_norm": 0.40488573908805847,
+      "learning_rate": 8e-05,
+      "loss": 1.6729,
+      "step": 185
+    },
+    {
+      "epoch": 0.04519499453286357,
+      "grad_norm": 0.4551525413990021,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 186
+    },
+    {
+      "epoch": 0.0454379783744381,
+      "grad_norm": 0.4685042202472687,
+      "learning_rate": 8e-05,
+      "loss": 1.8417,
+      "step": 187
+    },
+    {
+      "epoch": 0.04568096221601264,
+      "grad_norm": 0.43833059072494507,
+      "learning_rate": 8e-05,
+      "loss": 1.8575,
+      "step": 188
+    },
+    {
+      "epoch": 0.04592394605758717,
+      "grad_norm": 0.44860225915908813,
+      "learning_rate": 8e-05,
+      "loss": 1.8212,
+      "step": 189
+    },
+    {
+      "epoch": 0.046166929899161706,
+      "grad_norm": 0.4465605914592743,
+      "learning_rate": 8e-05,
+      "loss": 1.7541,
+      "step": 190
+    },
+    {
+      "epoch": 0.04640991374073624,
+      "grad_norm": 0.4442957639694214,
+      "learning_rate": 8e-05,
+      "loss": 1.9162,
+      "step": 191
+    },
+    {
+      "epoch": 0.046652897582310775,
+      "grad_norm": 0.4390282928943634,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 192
+    },
+    {
+      "epoch": 0.04689588142388531,
+      "grad_norm": 0.454426646232605,
+      "learning_rate": 8e-05,
+      "loss": 1.8302,
+      "step": 193
+    },
+    {
+      "epoch": 0.047138865265459844,
+      "grad_norm": 0.43619468808174133,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 194
+    },
+    {
+      "epoch": 0.047381849107034385,
+      "grad_norm": 0.40817612409591675,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 195
+    },
+    {
+      "epoch": 0.04762483294860892,
+      "grad_norm": 0.42603886127471924,
+      "learning_rate": 8e-05,
+      "loss": 1.8151,
+      "step": 196
+    },
+    {
+      "epoch": 0.047867816790183454,
+      "grad_norm": 0.4507157504558563,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 197
+    },
+    {
+      "epoch": 0.04811080063175799,
+      "grad_norm": 0.4105568826198578,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 198
+    },
+    {
+      "epoch": 0.04835378447333252,
+      "grad_norm": 0.40932849049568176,
+      "learning_rate": 8e-05,
+      "loss": 1.6927,
+      "step": 199
+    },
+    {
+      "epoch": 0.04859676831490706,
+      "grad_norm": 0.4452005624771118,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 200
+    },
+    {
+      "epoch": 0.04883975215648159,
+      "grad_norm": 0.44663047790527344,
+      "learning_rate": 8e-05,
+      "loss": 1.7947,
+      "step": 201
+    },
+    {
+      "epoch": 0.049082735998056126,
+      "grad_norm": 0.4227569103240967,
+      "learning_rate": 8e-05,
+      "loss": 1.8643,
+      "step": 202
+    },
+    {
+      "epoch": 0.04932571983963067,
+      "grad_norm": 0.42726418375968933,
+      "learning_rate": 8e-05,
+      "loss": 1.7908,
+      "step": 203
+    },
+    {
+      "epoch": 0.0495687036812052,
+      "grad_norm": 0.44869542121887207,
+      "learning_rate": 8e-05,
+      "loss": 1.8512,
+      "step": 204
+    },
+    {
+      "epoch": 0.049811687522779736,
+      "grad_norm": 0.43665412068367004,
+      "learning_rate": 8e-05,
+      "loss": 1.7925,
+      "step": 205
+    },
+    {
+      "epoch": 0.05005467136435427,
+      "grad_norm": 0.4294425845146179,
+      "learning_rate": 8e-05,
+      "loss": 1.7406,
+      "step": 206
+    },
+    {
+      "epoch": 0.050297655205928805,
+      "grad_norm": 0.4532032310962677,
+      "learning_rate": 8e-05,
+      "loss": 1.8743,
+      "step": 207
+    },
+    {
+      "epoch": 0.05054063904750334,
+      "grad_norm": 0.4643804132938385,
+      "learning_rate": 8e-05,
+      "loss": 1.7961,
+      "step": 208
+    },
+    {
+      "epoch": 0.050783622889077874,
+      "grad_norm": 0.412218302488327,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 209
+    },
+    {
+      "epoch": 0.05102660673065241,
+      "grad_norm": 0.45216429233551025,
+      "learning_rate": 8e-05,
+      "loss": 1.8609,
+      "step": 210
+    },
+    {
+      "epoch": 0.05126959057222695,
+      "grad_norm": 0.42403680086135864,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 211
+    },
+    {
+      "epoch": 0.051512574413801485,
+      "grad_norm": 0.4293997883796692,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 212
+    },
+    {
+      "epoch": 0.05175555825537602,
+      "grad_norm": 0.4341531991958618,
+      "learning_rate": 8e-05,
+      "loss": 1.8109,
+      "step": 213
+    },
+    {
+      "epoch": 0.051998542096950554,
+      "grad_norm": 0.40185922384262085,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 214
+    },
+    {
+      "epoch": 0.05224152593852509,
+      "grad_norm": 0.4393584728240967,
+      "learning_rate": 8e-05,
+      "loss": 1.8533,
+      "step": 215
+    },
+    {
+      "epoch": 0.05248450978009962,
+      "grad_norm": 0.42899441719055176,
+      "learning_rate": 8e-05,
+      "loss": 1.7643,
+      "step": 216
+    },
+    {
+      "epoch": 0.05272749362167416,
+      "grad_norm": 0.4237852692604065,
+      "learning_rate": 8e-05,
+      "loss": 1.7457,
+      "step": 217
+    },
+    {
+      "epoch": 0.05297047746324869,
+      "grad_norm": 0.42249593138694763,
+      "learning_rate": 8e-05,
+      "loss": 1.7127,
+      "step": 218
+    },
+    {
+      "epoch": 0.05321346130482323,
+      "grad_norm": 0.421320378780365,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 219
+    },
+    {
+      "epoch": 0.05345644514639777,
+      "grad_norm": 0.4289996325969696,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 220
+    },
+    {
+      "epoch": 0.0536994289879723,
+      "grad_norm": 0.4348432421684265,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 221
+    },
+    {
+      "epoch": 0.053942412829546836,
+      "grad_norm": 0.41037532687187195,
+      "learning_rate": 8e-05,
+      "loss": 1.8001,
+      "step": 222
+    },
+    {
+      "epoch": 0.05418539667112137,
+      "grad_norm": 0.4165085256099701,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 223
+    },
+    {
+      "epoch": 0.054428380512695905,
+      "grad_norm": 0.44498410820961,
+      "learning_rate": 8e-05,
+      "loss": 1.8597,
+      "step": 224
+    },
+    {
+      "epoch": 0.05467136435427044,
+      "grad_norm": 0.40958845615386963,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 225
+    },
+    {
+      "epoch": 0.054914348195844974,
+      "grad_norm": 0.4312472343444824,
+      "learning_rate": 8e-05,
+      "loss": 1.7577,
+      "step": 226
+    },
+    {
+      "epoch": 0.05515733203741951,
+      "grad_norm": 0.4492066204547882,
+      "learning_rate": 8e-05,
+      "loss": 1.84,
+      "step": 227
+    },
+    {
+      "epoch": 0.05540031587899405,
+      "grad_norm": 0.40647533535957336,
+      "learning_rate": 8e-05,
+      "loss": 1.6446,
+      "step": 228
+    },
+    {
+      "epoch": 0.055643299720568584,
+      "grad_norm": 0.429390549659729,
+      "learning_rate": 8e-05,
+      "loss": 1.7794,
+      "step": 229
+    },
+    {
+      "epoch": 0.05588628356214312,
+      "grad_norm": 0.41633883118629456,
+      "learning_rate": 8e-05,
+      "loss": 1.8394,
+      "step": 230
+    },
+    {
+      "epoch": 0.05612926740371765,
+      "grad_norm": 0.40792158246040344,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 231
+    },
+    {
+      "epoch": 0.05637225124529219,
+      "grad_norm": 0.41727787256240845,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 232
+    },
+    {
+      "epoch": 0.05661523508686672,
+      "grad_norm": 0.4122094213962555,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 233
+    },
+    {
+      "epoch": 0.056858218928441256,
+      "grad_norm": 0.43074655532836914,
+      "learning_rate": 8e-05,
+      "loss": 1.8505,
+      "step": 234
+    },
+    {
+      "epoch": 0.05710120277001579,
+      "grad_norm": 0.4109654426574707,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 235
+    },
+    {
+      "epoch": 0.05734418661159033,
+      "grad_norm": 0.40236973762512207,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 236
+    },
+    {
+      "epoch": 0.05758717045316487,
+      "grad_norm": 0.43539148569107056,
+      "learning_rate": 8e-05,
+      "loss": 1.8052,
+      "step": 237
+    },
+    {
+      "epoch": 0.0578301542947394,
+      "grad_norm": 0.41733142733573914,
+      "learning_rate": 8e-05,
+      "loss": 1.7533,
+      "step": 238
+    },
+    {
+      "epoch": 0.058073138136313936,
+      "grad_norm": 0.4331442415714264,
+      "learning_rate": 8e-05,
+      "loss": 1.7759,
+      "step": 239
+    },
+    {
+      "epoch": 0.05831612197788847,
+      "grad_norm": 0.4151133894920349,
+      "learning_rate": 8e-05,
+      "loss": 1.8155,
+      "step": 240
+    },
+    {
+      "epoch": 0.058559105819463005,
+      "grad_norm": 0.42025041580200195,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 241
+    },
+    {
+      "epoch": 0.05880208966103754,
+      "grad_norm": 0.4147380590438843,
+      "learning_rate": 8e-05,
+      "loss": 1.7317,
+      "step": 242
+    },
+    {
+      "epoch": 0.059045073502612073,
+      "grad_norm": 0.3916904330253601,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 243
+    },
+    {
+      "epoch": 0.059288057344186615,
+      "grad_norm": 0.4432324469089508,
+      "learning_rate": 8e-05,
+      "loss": 1.8159,
+      "step": 244
+    },
+    {
+      "epoch": 0.05953104118576115,
+      "grad_norm": 0.4161258935928345,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 245
+    },
+    {
+      "epoch": 0.059774025027335684,
+      "grad_norm": 0.4148889482021332,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 246
+    },
+    {
+      "epoch": 0.06001700886891022,
+      "grad_norm": 0.4157002866268158,
+      "learning_rate": 8e-05,
+      "loss": 1.7166,
+      "step": 247
+    },
+    {
+      "epoch": 0.06025999271048475,
+      "grad_norm": 0.39465364813804626,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 248
+    },
+    {
+      "epoch": 0.06050297655205929,
+      "grad_norm": 0.41584497690200806,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 249
+    },
+    {
+      "epoch": 0.06074596039363382,
+      "grad_norm": 0.4148780107498169,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 250
+    },
+    {
+      "epoch": 0.060988944235208356,
+      "grad_norm": 0.43678727746009827,
+      "learning_rate": 8e-05,
+      "loss": 1.8074,
+      "step": 251
+    },
+    {
+      "epoch": 0.06123192807678289,
+      "grad_norm": 0.4008216857910156,
+      "learning_rate": 8e-05,
+      "loss": 1.6168,
+      "step": 252
+    },
+    {
+      "epoch": 0.06147491191835743,
+      "grad_norm": 0.4190528094768524,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 253
+    },
+    {
+      "epoch": 0.061717895759931966,
+      "grad_norm": 0.4158279001712799,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 254
+    },
+    {
+      "epoch": 0.0619608796015065,
+      "grad_norm": 0.4088147282600403,
+      "learning_rate": 8e-05,
+      "loss": 1.7789,
+      "step": 255
+    },
+    {
+      "epoch": 0.062203863443081035,
+      "grad_norm": 0.41537022590637207,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 256
+    },
+    {
+      "epoch": 0.06244684728465557,
+      "grad_norm": 0.3931873142719269,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 257
+    },
+    {
+      "epoch": 0.06268983112623011,
+      "grad_norm": 0.4231928884983063,
+      "learning_rate": 8e-05,
+      "loss": 1.7634,
+      "step": 258
+    },
+    {
+      "epoch": 0.06293281496780465,
+      "grad_norm": 0.396416574716568,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 259
+    },
+    {
+      "epoch": 0.06317579880937918,
+      "grad_norm": 0.39217904210090637,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 260
+    },
+    {
+      "epoch": 0.06341878265095371,
+      "grad_norm": 0.4122065305709839,
+      "learning_rate": 8e-05,
+      "loss": 1.874,
+      "step": 261
+    },
+    {
+      "epoch": 0.06366176649252825,
+      "grad_norm": 0.4038335680961609,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 262
+    },
+    {
+      "epoch": 0.06390475033410278,
+      "grad_norm": 0.4242023229598999,
+      "learning_rate": 8e-05,
+      "loss": 1.8315,
+      "step": 263
+    },
+    {
+      "epoch": 0.06414773417567732,
+      "grad_norm": 0.41710829734802246,
+      "learning_rate": 8e-05,
+      "loss": 1.6869,
+      "step": 264
+    },
+    {
+      "epoch": 0.06439071801725185,
+      "grad_norm": 0.4024234116077423,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 265
+    },
+    {
+      "epoch": 0.06463370185882639,
+      "grad_norm": 0.43158861994743347,
+      "learning_rate": 8e-05,
+      "loss": 1.8,
+      "step": 266
+    },
+    {
+      "epoch": 0.06487668570040092,
+      "grad_norm": 0.4473307132720947,
+      "learning_rate": 8e-05,
+      "loss": 1.8952,
+      "step": 267
+    },
+    {
+      "epoch": 0.06511966954197546,
+      "grad_norm": 0.4519525170326233,
+      "learning_rate": 8e-05,
+      "loss": 1.9715,
+      "step": 268
+    },
+    {
+      "epoch": 0.06536265338354999,
+      "grad_norm": 0.42093324661254883,
+      "learning_rate": 8e-05,
+      "loss": 1.848,
+      "step": 269
+    },
+    {
+      "epoch": 0.06560563722512452,
+      "grad_norm": 0.44148561358451843,
+      "learning_rate": 8e-05,
+      "loss": 1.7559,
+      "step": 270
+    },
+    {
+      "epoch": 0.06584862106669906,
+      "grad_norm": 0.4208815097808838,
+      "learning_rate": 8e-05,
+      "loss": 1.8728,
+      "step": 271
+    },
+    {
+      "epoch": 0.0660916049082736,
+      "grad_norm": 0.4542776942253113,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 272
+    },
+    {
+      "epoch": 0.06633458874984814,
+      "grad_norm": 0.4561680257320404,
+      "learning_rate": 8e-05,
+      "loss": 1.8184,
+      "step": 273
+    },
+    {
+      "epoch": 0.06657757259142268,
+      "grad_norm": 0.4309070408344269,
+      "learning_rate": 8e-05,
+      "loss": 1.7915,
+      "step": 274
+    },
+    {
+      "epoch": 0.06682055643299721,
+      "grad_norm": 0.4592060446739197,
+      "learning_rate": 8e-05,
+      "loss": 1.8771,
+      "step": 275
+    },
+    {
+      "epoch": 0.06706354027457175,
+      "grad_norm": 0.4334331452846527,
+      "learning_rate": 8e-05,
+      "loss": 1.7187,
+      "step": 276
+    },
+    {
+      "epoch": 0.06730652411614628,
+      "grad_norm": 0.41513165831565857,
+      "learning_rate": 8e-05,
+      "loss": 1.7187,
+      "step": 277
+    },
+    {
+      "epoch": 0.06754950795772081,
+      "grad_norm": 0.42725256085395813,
+      "learning_rate": 8e-05,
+      "loss": 1.6586,
+      "step": 278
+    },
+    {
+      "epoch": 0.06779249179929535,
+      "grad_norm": 0.42010682821273804,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 279
+    },
+    {
+      "epoch": 0.06803547564086988,
+      "grad_norm": 0.41037890315055847,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 280
+    },
+    {
+      "epoch": 0.06827845948244442,
+      "grad_norm": 0.4202909469604492,
+      "learning_rate": 8e-05,
+      "loss": 1.8023,
+      "step": 281
+    },
+    {
+      "epoch": 0.06852144332401895,
+      "grad_norm": 0.38481223583221436,
+      "learning_rate": 8e-05,
+      "loss": 1.6706,
+      "step": 282
+    },
+    {
+      "epoch": 0.06876442716559349,
+      "grad_norm": 0.4297863245010376,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 283
+    },
+    {
+      "epoch": 0.06900741100716802,
+      "grad_norm": 0.448222279548645,
+      "learning_rate": 8e-05,
+      "loss": 1.9037,
+      "step": 284
+    },
+    {
+      "epoch": 0.06925039484874256,
+      "grad_norm": 0.42773011326789856,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 285
+    },
+    {
+      "epoch": 0.06949337869031709,
+      "grad_norm": 0.425235778093338,
+      "learning_rate": 8e-05,
+      "loss": 1.8267,
+      "step": 286
+    },
+    {
+      "epoch": 0.06973636253189162,
+      "grad_norm": 0.42985713481903076,
+      "learning_rate": 8e-05,
+      "loss": 1.8545,
+      "step": 287
+    },
+    {
+      "epoch": 0.06997934637346616,
+      "grad_norm": 0.4162665009498596,
+      "learning_rate": 8e-05,
+      "loss": 1.8069,
+      "step": 288
+    },
+    {
+      "epoch": 0.07022233021504069,
+      "grad_norm": 0.4128991961479187,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 289
+    },
+    {
+      "epoch": 0.07046531405661524,
+      "grad_norm": 0.41411450505256653,
+      "learning_rate": 8e-05,
+      "loss": 1.8225,
+      "step": 290
+    },
+    {
+      "epoch": 0.07070829789818978,
+      "grad_norm": 0.4319620132446289,
+      "learning_rate": 8e-05,
+      "loss": 1.7133,
+      "step": 291
+    },
+    {
+      "epoch": 0.07095128173976431,
+      "grad_norm": 0.42714792490005493,
+      "learning_rate": 8e-05,
+      "loss": 1.7042,
+      "step": 292
+    },
+    {
+      "epoch": 0.07119426558133884,
+      "grad_norm": 0.41195017099380493,
+      "learning_rate": 8e-05,
+      "loss": 1.747,
+      "step": 293
+    },
+    {
+      "epoch": 0.07143724942291338,
+      "grad_norm": 0.411868155002594,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 294
+    },
+    {
+      "epoch": 0.07168023326448791,
+      "grad_norm": 2.195003032684326,
+      "learning_rate": 8e-05,
+      "loss": 1.7605,
+      "step": 295
+    },
+    {
+      "epoch": 0.07192321710606245,
+      "grad_norm": 0.4210272431373596,
+      "learning_rate": 8e-05,
+      "loss": 1.7884,
+      "step": 296
+    },
+    {
+      "epoch": 0.07216620094763698,
+      "grad_norm": 0.41573718190193176,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 297
+    },
+    {
+      "epoch": 0.07240918478921152,
+      "grad_norm": 0.42812490463256836,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 298
+    },
+    {
+      "epoch": 0.07265216863078605,
+      "grad_norm": 0.4160097539424896,
+      "learning_rate": 8e-05,
+      "loss": 1.6746,
+      "step": 299
+    },
+    {
+      "epoch": 0.07289515247236059,
+      "grad_norm": 0.4339464008808136,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 300
+    },
+    {
+      "epoch": 0.07313813631393512,
+      "grad_norm": 0.40688636898994446,
+      "learning_rate": 8e-05,
+      "loss": 1.7263,
+      "step": 301
+    },
+    {
+      "epoch": 0.07338112015550965,
+      "grad_norm": 0.4286927282810211,
+      "learning_rate": 8e-05,
+      "loss": 1.8026,
+      "step": 302
+    },
+    {
+      "epoch": 0.07362410399708419,
+      "grad_norm": 0.4497995972633362,
+      "learning_rate": 8e-05,
+      "loss": 1.8929,
+      "step": 303
+    },
+    {
+      "epoch": 0.07386708783865872,
+      "grad_norm": 0.39630329608917236,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 304
+    },
+    {
+      "epoch": 0.07411007168023326,
+      "grad_norm": 0.43036988377571106,
+      "learning_rate": 8e-05,
+      "loss": 1.7689,
+      "step": 305
+    },
+    {
+      "epoch": 0.0743530555218078,
+      "grad_norm": 0.41594815254211426,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 306
+    },
+    {
+      "epoch": 0.07459603936338234,
+      "grad_norm": 0.42958587408065796,
+      "learning_rate": 8e-05,
+      "loss": 1.8022,
+      "step": 307
+    },
+    {
+      "epoch": 0.07483902320495688,
+      "grad_norm": 0.41847291588783264,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 308
+    },
+    {
+      "epoch": 0.07508200704653141,
+      "grad_norm": 0.4142318665981293,
+      "learning_rate": 8e-05,
+      "loss": 1.8017,
+      "step": 309
+    },
+    {
+      "epoch": 0.07532499088810594,
+      "grad_norm": 0.41507381200790405,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 310
+    },
+    {
+      "epoch": 0.07556797472968048,
+      "grad_norm": 0.4372384548187256,
+      "learning_rate": 8e-05,
+      "loss": 1.9386,
+      "step": 311
+    },
+    {
+      "epoch": 0.07581095857125501,
+      "grad_norm": 0.4396492838859558,
+      "learning_rate": 8e-05,
+      "loss": 1.8556,
+      "step": 312
+    },
+    {
+      "epoch": 0.07605394241282955,
+      "grad_norm": 0.40758559107780457,
+      "learning_rate": 8e-05,
+      "loss": 1.809,
+      "step": 313
+    },
+    {
+      "epoch": 0.07629692625440408,
+      "grad_norm": 0.4153490662574768,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 314
+    },
+    {
+      "epoch": 0.07653991009597862,
+      "grad_norm": 0.43232935667037964,
+      "learning_rate": 8e-05,
+      "loss": 1.6996,
+      "step": 315
+    },
+    {
+      "epoch": 0.07678289393755315,
+      "grad_norm": 0.4399256706237793,
+      "learning_rate": 8e-05,
+      "loss": 1.7999,
+      "step": 316
+    },
+    {
+      "epoch": 0.07702587777912769,
+      "grad_norm": 0.4185955822467804,
+      "learning_rate": 8e-05,
+      "loss": 1.8538,
+      "step": 317
+    },
+    {
+      "epoch": 0.07726886162070222,
+      "grad_norm": 0.42307454347610474,
+      "learning_rate": 8e-05,
+      "loss": 1.7355,
+      "step": 318
+    },
+    {
+      "epoch": 0.07751184546227675,
+      "grad_norm": 0.42181700468063354,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 319
+    },
+    {
+      "epoch": 0.07775482930385129,
+      "grad_norm": 0.42582136392593384,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 320
+    },
+    {
+      "epoch": 0.07799781314542582,
+      "grad_norm": 0.40768539905548096,
+      "learning_rate": 8e-05,
+      "loss": 1.779,
+      "step": 321
+    },
+    {
+      "epoch": 0.07824079698700036,
+      "grad_norm": 0.42627450823783875,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 322
+    },
+    {
+      "epoch": 0.0784837808285749,
+      "grad_norm": 0.4287555515766144,
+      "learning_rate": 8e-05,
+      "loss": 1.6869,
+      "step": 323
+    },
+    {
+      "epoch": 0.07872676467014944,
+      "grad_norm": 0.4250035285949707,
+      "learning_rate": 8e-05,
+      "loss": 1.7786,
+      "step": 324
+    },
+    {
+      "epoch": 0.07896974851172398,
+      "grad_norm": 0.42147719860076904,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 325
+    },
+    {
+      "epoch": 0.07921273235329851,
+      "grad_norm": 0.4258410334587097,
+      "learning_rate": 8e-05,
+      "loss": 1.7664,
+      "step": 326
+    },
+    {
+      "epoch": 0.07945571619487304,
+      "grad_norm": 0.42165327072143555,
+      "learning_rate": 8e-05,
+      "loss": 1.857,
+      "step": 327
+    },
+    {
+      "epoch": 0.07969870003644758,
+      "grad_norm": 0.3965833783149719,
+      "learning_rate": 8e-05,
+      "loss": 1.712,
+      "step": 328
+    },
+    {
+      "epoch": 0.07994168387802211,
+      "grad_norm": 0.4351164698600769,
+      "learning_rate": 8e-05,
+      "loss": 1.8151,
+      "step": 329
+    },
+    {
+      "epoch": 0.08018466771959665,
+      "grad_norm": 0.41587620973587036,
+      "learning_rate": 8e-05,
+      "loss": 1.7757,
+      "step": 330
+    },
+    {
+      "epoch": 0.08042765156117118,
+      "grad_norm": 0.422070175409317,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 331
+    },
+    {
+      "epoch": 0.08067063540274572,
+      "grad_norm": 0.4227633476257324,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 332
+    },
+    {
+      "epoch": 0.08091361924432025,
+      "grad_norm": 0.42356976866722107,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 333
+    },
+    {
+      "epoch": 0.08115660308589479,
+      "grad_norm": 0.4465808868408203,
+      "learning_rate": 8e-05,
+      "loss": 1.8332,
+      "step": 334
+    },
+    {
+      "epoch": 0.08139958692746932,
+      "grad_norm": 0.4090607762336731,
+      "learning_rate": 8e-05,
+      "loss": 1.7735,
+      "step": 335
+    },
+    {
+      "epoch": 0.08164257076904385,
+      "grad_norm": 0.4155711233615875,
+      "learning_rate": 8e-05,
+      "loss": 1.8427,
+      "step": 336
+    },
+    {
+      "epoch": 0.08188555461061839,
+      "grad_norm": 0.42290425300598145,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 337
+    },
+    {
+      "epoch": 0.08212853845219292,
+      "grad_norm": 0.427473783493042,
+      "learning_rate": 8e-05,
+      "loss": 1.8271,
+      "step": 338
+    },
+    {
+      "epoch": 0.08237152229376747,
+      "grad_norm": 0.41899219155311584,
+      "learning_rate": 8e-05,
+      "loss": 1.866,
+      "step": 339
+    },
+    {
+      "epoch": 0.082614506135342,
+      "grad_norm": 0.4216976463794708,
+      "learning_rate": 8e-05,
+      "loss": 1.8554,
+      "step": 340
+    },
+    {
+      "epoch": 0.08285748997691654,
+      "grad_norm": 0.40997281670570374,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 341
+    },
+    {
+      "epoch": 0.08310047381849107,
+      "grad_norm": 0.4109158217906952,
+      "learning_rate": 8e-05,
+      "loss": 1.7338,
+      "step": 342
+    },
+    {
+      "epoch": 0.08334345766006561,
+      "grad_norm": 0.4242565333843231,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 343
+    },
+    {
+      "epoch": 0.08358644150164014,
+      "grad_norm": 0.4627560079097748,
+      "learning_rate": 8e-05,
+      "loss": 1.8587,
+      "step": 344
+    },
+    {
+      "epoch": 0.08382942534321468,
+      "grad_norm": 0.39504504203796387,
+      "learning_rate": 8e-05,
+      "loss": 1.5618,
+      "step": 345
+    },
+    {
+      "epoch": 0.08407240918478921,
+      "grad_norm": 0.5971431136131287,
+      "learning_rate": 8e-05,
+      "loss": 1.984,
+      "step": 346
+    },
+    {
+      "epoch": 0.08431539302636375,
+      "grad_norm": 0.43111053109169006,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 347
+    },
+    {
+      "epoch": 0.08455837686793828,
+      "grad_norm": 0.44461020827293396,
+      "learning_rate": 8e-05,
+      "loss": 1.9002,
+      "step": 348
+    },
+    {
+      "epoch": 0.08480136070951282,
+      "grad_norm": 0.40719398856163025,
+      "learning_rate": 8e-05,
+      "loss": 1.6941,
+      "step": 349
+    },
+    {
+      "epoch": 0.08504434455108735,
+      "grad_norm": 0.41791296005249023,
+      "learning_rate": 8e-05,
+      "loss": 1.7223,
+      "step": 350
+    },
+    {
+      "epoch": 0.08528732839266188,
+      "grad_norm": 0.43790993094444275,
+      "learning_rate": 8e-05,
+      "loss": 1.7557,
+      "step": 351
+    },
+    {
+      "epoch": 0.08553031223423642,
+      "grad_norm": 0.3951578438282013,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 352
+    },
+    {
+      "epoch": 0.08577329607581095,
+      "grad_norm": 0.4710818827152252,
+      "learning_rate": 8e-05,
+      "loss": 1.8857,
+      "step": 353
+    },
+    {
+      "epoch": 0.08601627991738549,
+      "grad_norm": 0.4196699857711792,
+      "learning_rate": 8e-05,
+      "loss": 1.7204,
+      "step": 354
+    },
+    {
+      "epoch": 0.08625926375896002,
+      "grad_norm": 0.4194652736186981,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 355
+    },
+    {
+      "epoch": 0.08650224760053457,
+      "grad_norm": 0.438880980014801,
+      "learning_rate": 8e-05,
+      "loss": 1.6351,
+      "step": 356
+    },
+    {
+      "epoch": 0.0867452314421091,
+      "grad_norm": 0.41452550888061523,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 357
+    },
+    {
+      "epoch": 0.08698821528368364,
+      "grad_norm": 0.4164050221443176,
+      "learning_rate": 8e-05,
+      "loss": 1.7089,
+      "step": 358
+    },
+    {
+      "epoch": 0.08723119912525817,
+      "grad_norm": 0.4484345018863678,
+      "learning_rate": 8e-05,
+      "loss": 1.8237,
+      "step": 359
+    },
+    {
+      "epoch": 0.08747418296683271,
+      "grad_norm": 0.43570855259895325,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 360
+    },
+    {
+      "epoch": 0.08771716680840724,
+      "grad_norm": 0.4059920907020569,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 361
+    },
+    {
+      "epoch": 0.08796015064998178,
+      "grad_norm": 0.4080146253108978,
+      "learning_rate": 8e-05,
+      "loss": 1.641,
+      "step": 362
+    },
+    {
+      "epoch": 0.08820313449155631,
+      "grad_norm": 0.39081236720085144,
+      "learning_rate": 8e-05,
+      "loss": 1.6033,
+      "step": 363
+    },
+    {
+      "epoch": 0.08844611833313085,
+      "grad_norm": 0.4051390290260315,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 364
+    },
+    {
+      "epoch": 0.08868910217470538,
+      "grad_norm": 0.4073815643787384,
+      "learning_rate": 8e-05,
+      "loss": 1.8466,
+      "step": 365
+    },
+    {
+      "epoch": 0.08893208601627992,
+      "grad_norm": 0.4258662760257721,
+      "learning_rate": 8e-05,
+      "loss": 1.9736,
+      "step": 366
+    },
+    {
+      "epoch": 0.08917506985785445,
+      "grad_norm": 0.4329228401184082,
+      "learning_rate": 8e-05,
+      "loss": 1.7617,
+      "step": 367
+    },
+    {
+      "epoch": 0.08941805369942898,
+      "grad_norm": 0.4149377644062042,
+      "learning_rate": 8e-05,
+      "loss": 1.7838,
+      "step": 368
+    },
+    {
+      "epoch": 0.08966103754100352,
+      "grad_norm": 0.4258325397968292,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 369
+    },
+    {
+      "epoch": 0.08990402138257805,
+      "grad_norm": 0.4353642761707306,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 370
+    },
+    {
+      "epoch": 0.09014700522415259,
+      "grad_norm": 0.4435424506664276,
+      "learning_rate": 8e-05,
+      "loss": 1.8068,
+      "step": 371
+    },
+    {
+      "epoch": 0.09038998906572714,
+      "grad_norm": 0.43207329511642456,
+      "learning_rate": 8e-05,
+      "loss": 1.8337,
+      "step": 372
+    },
+    {
+      "epoch": 0.09063297290730167,
+      "grad_norm": 0.4225546717643738,
+      "learning_rate": 8e-05,
+      "loss": 1.8111,
+      "step": 373
+    },
+    {
+      "epoch": 0.0908759567488762,
+      "grad_norm": 0.41650712490081787,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 374
+    },
+    {
+      "epoch": 0.09111894059045074,
+      "grad_norm": 0.44362857937812805,
+      "learning_rate": 8e-05,
+      "loss": 1.8781,
+      "step": 375
+    },
+    {
+      "epoch": 0.09136192443202527,
+      "grad_norm": 0.4244385063648224,
+      "learning_rate": 8e-05,
+      "loss": 1.7318,
+      "step": 376
+    },
+    {
+      "epoch": 0.09160490827359981,
+      "grad_norm": 0.4230808615684509,
+      "learning_rate": 8e-05,
+      "loss": 1.868,
+      "step": 377
+    },
+    {
+      "epoch": 0.09184789211517434,
+      "grad_norm": 0.4087420403957367,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 378
+    },
+    {
+      "epoch": 0.09209087595674888,
+      "grad_norm": 0.4241844117641449,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 379
+    },
+    {
+      "epoch": 0.09233385979832341,
+      "grad_norm": 0.4437425136566162,
+      "learning_rate": 8e-05,
+      "loss": 1.8141,
+      "step": 380
+    },
+    {
+      "epoch": 0.09257684363989795,
+      "grad_norm": 0.43334922194480896,
+      "learning_rate": 8e-05,
+      "loss": 1.7747,
+      "step": 381
+    },
+    {
+      "epoch": 0.09281982748147248,
+      "grad_norm": 0.4382667541503906,
+      "learning_rate": 8e-05,
+      "loss": 1.7946,
+      "step": 382
+    },
+    {
+      "epoch": 0.09306281132304701,
+      "grad_norm": 0.4285142719745636,
+      "learning_rate": 8e-05,
+      "loss": 1.7562,
+      "step": 383
+    },
+    {
+      "epoch": 0.09330579516462155,
+      "grad_norm": 0.4092921018600464,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 384
+    },
+    {
+      "epoch": 0.09354877900619608,
+      "grad_norm": 0.4280427396297455,
+      "learning_rate": 8e-05,
+      "loss": 1.8365,
+      "step": 385
+    },
+    {
+      "epoch": 0.09379176284777062,
+      "grad_norm": 0.41424787044525146,
+      "learning_rate": 8e-05,
+      "loss": 1.6779,
+      "step": 386
+    },
+    {
+      "epoch": 0.09403474668934515,
+      "grad_norm": 0.4082753658294678,
+      "learning_rate": 8e-05,
+      "loss": 1.6822,
+      "step": 387
+    },
+    {
+      "epoch": 0.09427773053091969,
+      "grad_norm": 0.4185827374458313,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 388
+    },
+    {
+      "epoch": 0.09452071437249424,
+      "grad_norm": 0.4068378508090973,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 389
+    },
+    {
+      "epoch": 0.09476369821406877,
+      "grad_norm": 0.4201984405517578,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 390
+    },
+    {
+      "epoch": 0.0950066820556433,
+      "grad_norm": 0.4467129707336426,
+      "learning_rate": 8e-05,
+      "loss": 1.9124,
+      "step": 391
+    },
+    {
+      "epoch": 0.09524966589721784,
+      "grad_norm": 0.4684138596057892,
+      "learning_rate": 8e-05,
+      "loss": 1.8845,
+      "step": 392
+    },
+    {
+      "epoch": 0.09549264973879237,
+      "grad_norm": 0.391348659992218,
+      "learning_rate": 8e-05,
+      "loss": 1.6064,
+      "step": 393
+    },
+    {
+      "epoch": 0.09573563358036691,
+      "grad_norm": 0.43737274408340454,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 394
+    },
+    {
+      "epoch": 0.09597861742194144,
+      "grad_norm": 0.4272935688495636,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 395
+    },
+    {
+      "epoch": 0.09622160126351598,
+      "grad_norm": 0.41889452934265137,
+      "learning_rate": 8e-05,
+      "loss": 1.674,
+      "step": 396
+    },
+    {
+      "epoch": 0.09646458510509051,
+      "grad_norm": 0.4118969440460205,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 397
+    },
+    {
+      "epoch": 0.09670756894666505,
+      "grad_norm": 0.440357506275177,
+      "learning_rate": 8e-05,
+      "loss": 1.7683,
+      "step": 398
+    },
+    {
+      "epoch": 0.09695055278823958,
+      "grad_norm": 0.40678104758262634,
+      "learning_rate": 8e-05,
+      "loss": 1.6479,
+      "step": 399
+    },
+    {
+      "epoch": 0.09719353662981411,
+      "grad_norm": 0.42315101623535156,
+      "learning_rate": 8e-05,
+      "loss": 1.7731,
+      "step": 400
+    },
+    {
+      "epoch": 0.09743652047138865,
+      "grad_norm": 0.44656410813331604,
+      "learning_rate": 8e-05,
+      "loss": 1.7927,
+      "step": 401
+    },
+    {
+      "epoch": 0.09767950431296318,
+      "grad_norm": 0.4181327819824219,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 402
+    },
+    {
+      "epoch": 0.09792248815453772,
+      "grad_norm": 0.4143099784851074,
+      "learning_rate": 8e-05,
+      "loss": 1.7421,
+      "step": 403
+    },
+    {
+      "epoch": 0.09816547199611225,
+      "grad_norm": 0.399997353553772,
+      "learning_rate": 8e-05,
+      "loss": 1.732,
+      "step": 404
+    },
+    {
+      "epoch": 0.0984084558376868,
+      "grad_norm": 0.41090577840805054,
+      "learning_rate": 8e-05,
+      "loss": 1.8132,
+      "step": 405
+    },
+    {
+      "epoch": 0.09865143967926134,
+      "grad_norm": 0.4159490764141083,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 406
+    },
+    {
+      "epoch": 0.09889442352083587,
+      "grad_norm": 0.4340929687023163,
+      "learning_rate": 8e-05,
+      "loss": 1.7595,
+      "step": 407
+    },
+    {
+      "epoch": 0.0991374073624104,
+      "grad_norm": 0.41250815987586975,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 408
+    },
+    {
+      "epoch": 0.09938039120398494,
+      "grad_norm": 0.3895319402217865,
+      "learning_rate": 8e-05,
+      "loss": 1.6261,
+      "step": 409
+    },
+    {
+      "epoch": 0.09962337504555947,
+      "grad_norm": 0.42065054178237915,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 410
+    },
+    {
+      "epoch": 0.09986635888713401,
+      "grad_norm": 0.4263005554676056,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 411
+    },
+    {
+      "epoch": 0.10010934272870854,
+      "grad_norm": 0.4143153131008148,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 412
+    },
+    {
+      "epoch": 0.10035232657028308,
+      "grad_norm": 0.4837665259838104,
+      "learning_rate": 8e-05,
+      "loss": 1.5294,
+      "step": 413
+    },
+    {
+      "epoch": 0.10059531041185761,
+      "grad_norm": 0.4138377606868744,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 414
+    },
+    {
+      "epoch": 0.10083829425343215,
+      "grad_norm": 0.5110868215560913,
+      "learning_rate": 8e-05,
+      "loss": 1.7925,
+      "step": 415
+    },
+    {
+      "epoch": 0.10108127809500668,
+      "grad_norm": 0.40476730465888977,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 416
+    },
+    {
+      "epoch": 0.10132426193658121,
+      "grad_norm": 0.42230334877967834,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 417
+    },
+    {
+      "epoch": 0.10156724577815575,
+      "grad_norm": 0.4212943911552429,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 418
+    },
+    {
+      "epoch": 0.10181022961973028,
+      "grad_norm": 0.4509720206260681,
+      "learning_rate": 8e-05,
+      "loss": 1.8692,
+      "step": 419
+    },
+    {
+      "epoch": 0.10205321346130482,
+      "grad_norm": 0.39649084210395813,
+      "learning_rate": 8e-05,
+      "loss": 1.7349,
+      "step": 420
+    },
+    {
+      "epoch": 0.10229619730287935,
+      "grad_norm": 0.4376465678215027,
+      "learning_rate": 8e-05,
+      "loss": 1.6797,
+      "step": 421
+    },
+    {
+      "epoch": 0.1025391811444539,
+      "grad_norm": 0.4412459135055542,
+      "learning_rate": 8e-05,
+      "loss": 1.769,
+      "step": 422
+    },
+    {
+      "epoch": 0.10278216498602843,
+      "grad_norm": 0.4239305853843689,
+      "learning_rate": 8e-05,
+      "loss": 1.7809,
+      "step": 423
+    },
+    {
+      "epoch": 0.10302514882760297,
+      "grad_norm": 0.4495208263397217,
+      "learning_rate": 8e-05,
+      "loss": 1.8249,
+      "step": 424
+    },
+    {
+      "epoch": 0.1032681326691775,
+      "grad_norm": 0.4457133710384369,
+      "learning_rate": 8e-05,
+      "loss": 1.8118,
+      "step": 425
+    },
+    {
+      "epoch": 0.10351111651075204,
+      "grad_norm": 0.4061901569366455,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 426
+    },
+    {
+      "epoch": 0.10375410035232657,
+      "grad_norm": 0.41500920057296753,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 427
+    },
+    {
+      "epoch": 0.10399708419390111,
+      "grad_norm": 0.46426844596862793,
+      "learning_rate": 8e-05,
+      "loss": 1.7675,
+      "step": 428
+    },
+    {
+      "epoch": 0.10424006803547564,
+      "grad_norm": 0.4445520043373108,
+      "learning_rate": 8e-05,
+      "loss": 1.7938,
+      "step": 429
+    },
+    {
+      "epoch": 0.10448305187705018,
+      "grad_norm": 0.42322006821632385,
+      "learning_rate": 8e-05,
+      "loss": 1.8434,
+      "step": 430
+    },
+    {
+      "epoch": 0.10472603571862471,
+      "grad_norm": 0.4529503881931305,
+      "learning_rate": 8e-05,
+      "loss": 1.7494,
+      "step": 431
+    },
+    {
+      "epoch": 0.10496901956019924,
+      "grad_norm": 0.4172574579715729,
+      "learning_rate": 8e-05,
+      "loss": 1.7325,
+      "step": 432
+    },
+    {
+      "epoch": 0.10521200340177378,
+      "grad_norm": 0.4366316795349121,
+      "learning_rate": 8e-05,
+      "loss": 1.7598,
+      "step": 433
+    },
+    {
+      "epoch": 0.10545498724334831,
+      "grad_norm": 0.4232217073440552,
+      "learning_rate": 8e-05,
+      "loss": 1.7331,
+      "step": 434
+    },
+    {
+      "epoch": 0.10569797108492285,
+      "grad_norm": 0.4281679093837738,
+      "learning_rate": 8e-05,
+      "loss": 1.8115,
+      "step": 435
+    },
+    {
+      "epoch": 0.10594095492649738,
+      "grad_norm": 0.4335026144981384,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 436
+    },
+    {
+      "epoch": 0.10618393876807192,
+      "grad_norm": 0.4126567244529724,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 437
+    },
+    {
+      "epoch": 0.10642692260964647,
+      "grad_norm": 0.4086419343948364,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 438
+    },
+    {
+      "epoch": 0.106669906451221,
+      "grad_norm": 0.4197182357311249,
+      "learning_rate": 8e-05,
+      "loss": 1.8021,
+      "step": 439
+    },
+    {
+      "epoch": 0.10691289029279553,
+      "grad_norm": 0.3942872881889343,
+      "learning_rate": 8e-05,
+      "loss": 1.4548,
+      "step": 440
+    },
+    {
+      "epoch": 0.10715587413437007,
+      "grad_norm": 0.41451960802078247,
+      "learning_rate": 8e-05,
+      "loss": 1.5907,
+      "step": 441
+    },
+    {
+      "epoch": 0.1073988579759446,
+      "grad_norm": 0.4235744774341583,
+      "learning_rate": 8e-05,
+      "loss": 1.9298,
+      "step": 442
+    },
+    {
+      "epoch": 0.10764184181751914,
+      "grad_norm": 0.44768843054771423,
+      "learning_rate": 8e-05,
+      "loss": 1.8039,
+      "step": 443
+    },
+    {
+      "epoch": 0.10788482565909367,
+      "grad_norm": 0.40303924679756165,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 444
+    },
+    {
+      "epoch": 0.1081278095006682,
+      "grad_norm": 0.42123013734817505,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 445
+    },
+    {
+      "epoch": 0.10837079334224274,
+      "grad_norm": 0.4180941581726074,
+      "learning_rate": 8e-05,
+      "loss": 1.7982,
+      "step": 446
+    },
+    {
+      "epoch": 0.10861377718381728,
+      "grad_norm": 0.41173458099365234,
+      "learning_rate": 8e-05,
+      "loss": 1.6573,
+      "step": 447
+    },
+    {
+      "epoch": 0.10885676102539181,
+      "grad_norm": 0.41863924264907837,
+      "learning_rate": 8e-05,
+      "loss": 1.766,
+      "step": 448
+    },
+    {
+      "epoch": 0.10909974486696634,
+      "grad_norm": 0.4276777505874634,
+      "learning_rate": 8e-05,
+      "loss": 1.7029,
+      "step": 449
+    },
+    {
+      "epoch": 0.10934272870854088,
+      "grad_norm": 0.42458993196487427,
+      "learning_rate": 8e-05,
+      "loss": 1.6814,
+      "step": 450
+    },
+    {
+      "epoch": 0.10958571255011541,
+      "grad_norm": 0.41878148913383484,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 451
+    },
+    {
+      "epoch": 0.10982869639168995,
+      "grad_norm": 0.4032799303531647,
+      "learning_rate": 8e-05,
+      "loss": 1.6357,
+      "step": 452
+    },
+    {
+      "epoch": 0.11007168023326448,
+      "grad_norm": 0.41101089119911194,
+      "learning_rate": 8e-05,
+      "loss": 1.6992,
+      "step": 453
+    },
+    {
+      "epoch": 0.11031466407483902,
+      "grad_norm": 0.41978752613067627,
+      "learning_rate": 8e-05,
+      "loss": 1.8756,
+      "step": 454
+    },
+    {
+      "epoch": 0.11055764791641357,
+      "grad_norm": 0.40360498428344727,
+      "learning_rate": 8e-05,
+      "loss": 1.6878,
+      "step": 455
+    },
+    {
+      "epoch": 0.1108006317579881,
+      "grad_norm": 0.4039525091648102,
+      "learning_rate": 8e-05,
+      "loss": 1.7367,
+      "step": 456
+    },
+    {
+      "epoch": 0.11104361559956263,
+      "grad_norm": 0.43515491485595703,
+      "learning_rate": 8e-05,
+      "loss": 1.8582,
+      "step": 457
+    },
+    {
+      "epoch": 0.11128659944113717,
+      "grad_norm": 0.4115177392959595,
+      "learning_rate": 8e-05,
+      "loss": 1.7892,
+      "step": 458
+    },
+    {
+      "epoch": 0.1115295832827117,
+      "grad_norm": 0.4229578673839569,
+      "learning_rate": 8e-05,
+      "loss": 1.8702,
+      "step": 459
+    },
+    {
+      "epoch": 0.11177256712428624,
+      "grad_norm": 0.4015786945819855,
+      "learning_rate": 8e-05,
+      "loss": 1.7798,
+      "step": 460
+    },
+    {
+      "epoch": 0.11201555096586077,
+      "grad_norm": 0.40968620777130127,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 461
+    },
+    {
+      "epoch": 0.1122585348074353,
+      "grad_norm": 0.4068968594074249,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 462
+    },
+    {
+      "epoch": 0.11250151864900984,
+      "grad_norm": 0.40810808539390564,
+      "learning_rate": 8e-05,
+      "loss": 1.8082,
+      "step": 463
+    },
+    {
+      "epoch": 0.11274450249058438,
+      "grad_norm": 0.4733920097351074,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 464
+    },
+    {
+      "epoch": 0.11298748633215891,
+      "grad_norm": 0.45223650336265564,
+      "learning_rate": 8e-05,
+      "loss": 1.7298,
+      "step": 465
+    },
+    {
+      "epoch": 0.11323047017373344,
+      "grad_norm": 0.4065145254135132,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 466
+    },
+    {
+      "epoch": 0.11347345401530798,
+      "grad_norm": 0.43968337774276733,
+      "learning_rate": 8e-05,
+      "loss": 1.8832,
+      "step": 467
+    },
+    {
+      "epoch": 0.11371643785688251,
+      "grad_norm": 0.4068659245967865,
+      "learning_rate": 8e-05,
+      "loss": 1.6833,
+      "step": 468
+    },
+    {
+      "epoch": 0.11395942169845705,
+      "grad_norm": 0.44198158383369446,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 469
+    },
+    {
+      "epoch": 0.11420240554003158,
+      "grad_norm": 0.4050433039665222,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 470
+    },
+    {
+      "epoch": 0.11444538938160612,
+      "grad_norm": 0.423566609621048,
+      "learning_rate": 8e-05,
+      "loss": 1.7293,
+      "step": 471
+    },
+    {
+      "epoch": 0.11468837322318066,
+      "grad_norm": 0.4198954105377197,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 472
+    },
+    {
+      "epoch": 0.1149313570647552,
+      "grad_norm": 0.43905210494995117,
+      "learning_rate": 8e-05,
+      "loss": 1.8544,
+      "step": 473
+    },
+    {
+      "epoch": 0.11517434090632973,
+      "grad_norm": 0.403835654258728,
+      "learning_rate": 8e-05,
+      "loss": 1.7008,
+      "step": 474
+    },
+    {
+      "epoch": 0.11541732474790427,
+      "grad_norm": 0.4185623228549957,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 475
+    },
+    {
+      "epoch": 0.1156603085894788,
+      "grad_norm": 0.3955186605453491,
+      "learning_rate": 8e-05,
+      "loss": 1.6089,
+      "step": 476
+    },
+    {
+      "epoch": 0.11590329243105334,
+      "grad_norm": 0.40133073925971985,
+      "learning_rate": 8e-05,
+      "loss": 1.5766,
+      "step": 477
+    },
+    {
+      "epoch": 0.11614627627262787,
+      "grad_norm": 0.4133303165435791,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 478
+    },
+    {
+      "epoch": 0.1163892601142024,
+      "grad_norm": 0.4125986397266388,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 479
+    },
+    {
+      "epoch": 0.11663224395577694,
+      "grad_norm": 0.467380553483963,
+      "learning_rate": 8e-05,
+      "loss": 1.753,
+      "step": 480
+    },
+    {
+      "epoch": 0.11687522779735147,
+      "grad_norm": 0.4279719293117523,
+      "learning_rate": 8e-05,
+      "loss": 1.7712,
+      "step": 481
+    },
+    {
+      "epoch": 0.11711821163892601,
+      "grad_norm": 0.4266670048236847,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 482
+    },
+    {
+      "epoch": 0.11736119548050054,
+      "grad_norm": 0.42196038365364075,
+      "learning_rate": 8e-05,
+      "loss": 1.7975,
+      "step": 483
+    },
+    {
+      "epoch": 0.11760417932207508,
+      "grad_norm": 0.4362437129020691,
+      "learning_rate": 8e-05,
+      "loss": 1.8205,
+      "step": 484
+    },
+    {
+      "epoch": 0.11784716316364961,
+      "grad_norm": 0.4456392824649811,
+      "learning_rate": 8e-05,
+      "loss": 1.9124,
+      "step": 485
+    },
+    {
+      "epoch": 0.11809014700522415,
+      "grad_norm": 0.40454182028770447,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 486
+    },
+    {
+      "epoch": 0.11833313084679868,
+      "grad_norm": 0.4022403061389923,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 487
+    },
+    {
+      "epoch": 0.11857611468837323,
+      "grad_norm": 0.40457379817962646,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 488
+    },
+    {
+      "epoch": 0.11881909852994776,
+      "grad_norm": 0.4198242127895355,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 489
+    },
+    {
+      "epoch": 0.1190620823715223,
+      "grad_norm": 0.40101826190948486,
+      "learning_rate": 8e-05,
+      "loss": 1.621,
+      "step": 490
+    },
+    {
+      "epoch": 0.11930506621309683,
+      "grad_norm": 0.438236266374588,
+      "learning_rate": 8e-05,
+      "loss": 1.777,
+      "step": 491
+    },
+    {
+      "epoch": 0.11954805005467137,
+      "grad_norm": 0.4342082738876343,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 492
+    },
+    {
+      "epoch": 0.1197910338962459,
+      "grad_norm": 0.4013380706310272,
+      "learning_rate": 8e-05,
+      "loss": 1.5459,
+      "step": 493
+    },
+    {
+      "epoch": 0.12003401773782044,
+      "grad_norm": 0.4234047830104828,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 494
+    },
+    {
+      "epoch": 0.12027700157939497,
+      "grad_norm": 0.4206213653087616,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 495
+    },
+    {
+      "epoch": 0.1205199854209695,
+      "grad_norm": 0.4329131543636322,
+      "learning_rate": 8e-05,
+      "loss": 1.7233,
+      "step": 496
+    },
+    {
+      "epoch": 0.12076296926254404,
+      "grad_norm": 0.4249413013458252,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 497
+    },
+    {
+      "epoch": 0.12100595310411857,
+      "grad_norm": 0.4211755096912384,
+      "learning_rate": 8e-05,
+      "loss": 1.683,
+      "step": 498
+    },
+    {
+      "epoch": 0.12124893694569311,
+      "grad_norm": 0.4254203140735626,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 499
+    },
+    {
+      "epoch": 0.12149192078726764,
+      "grad_norm": 0.41103148460388184,
+      "learning_rate": 8e-05,
+      "loss": 1.7606,
+      "step": 500
+    },
+    {
+      "epoch": 0.12173490462884218,
+      "grad_norm": 0.4685738980770111,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 501
+    },
+    {
+      "epoch": 0.12197788847041671,
+      "grad_norm": 0.4338212013244629,
+      "learning_rate": 8e-05,
+      "loss": 1.7965,
+      "step": 502
+    },
+    {
+      "epoch": 0.12222087231199125,
+      "grad_norm": 0.44578397274017334,
+      "learning_rate": 8e-05,
+      "loss": 1.907,
+      "step": 503
+    },
+    {
+      "epoch": 0.12246385615356578,
+      "grad_norm": 0.4282798171043396,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 504
+    },
+    {
+      "epoch": 0.12270683999514033,
+      "grad_norm": 0.42436984181404114,
+      "learning_rate": 8e-05,
+      "loss": 1.8776,
+      "step": 505
+    },
+    {
+      "epoch": 0.12294982383671486,
+      "grad_norm": 0.4542418420314789,
+      "learning_rate": 8e-05,
+      "loss": 1.7844,
+      "step": 506
+    },
+    {
+      "epoch": 0.1231928076782894,
+      "grad_norm": 0.43623366951942444,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 507
+    },
+    {
+      "epoch": 0.12343579151986393,
+      "grad_norm": 0.41631045937538147,
+      "learning_rate": 8e-05,
+      "loss": 1.6828,
+      "step": 508
+    },
+    {
+      "epoch": 0.12367877536143847,
+      "grad_norm": 0.4436653256416321,
+      "learning_rate": 8e-05,
+      "loss": 1.81,
+      "step": 509
+    },
+    {
+      "epoch": 0.123921759203013,
+      "grad_norm": 0.4137571454048157,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 510
+    },
+    {
+      "epoch": 0.12416474304458754,
+      "grad_norm": 0.4194542467594147,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 511
+    },
+    {
+      "epoch": 0.12440772688616207,
+      "grad_norm": 0.43399301171302795,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 512
+    },
+    {
+      "epoch": 0.1246507107277366,
+      "grad_norm": 0.4126221537590027,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 513
+    },
+    {
+      "epoch": 0.12489369456931114,
+      "grad_norm": 0.44673866033554077,
+      "learning_rate": 8e-05,
+      "loss": 1.8053,
+      "step": 514
+    },
+    {
+      "epoch": 0.1251366784108857,
+      "grad_norm": 0.417203813791275,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 515
+    },
+    {
+      "epoch": 0.12537966225246022,
+      "grad_norm": 0.43842193484306335,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 516
+    },
+    {
+      "epoch": 0.12562264609403476,
+      "grad_norm": 0.40844860672950745,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 517
+    },
+    {
+      "epoch": 0.1258656299356093,
+      "grad_norm": 0.46497642993927,
+      "learning_rate": 8e-05,
+      "loss": 1.8363,
+      "step": 518
+    },
+    {
+      "epoch": 0.12610861377718383,
+      "grad_norm": 0.4338991343975067,
+      "learning_rate": 8e-05,
+      "loss": 1.8075,
+      "step": 519
+    },
+    {
+      "epoch": 0.12635159761875836,
+      "grad_norm": 0.42398321628570557,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 520
+    },
+    {
+      "epoch": 0.1265945814603329,
+      "grad_norm": 0.4285203218460083,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 521
+    },
+    {
+      "epoch": 0.12683756530190743,
+      "grad_norm": 0.43014469742774963,
+      "learning_rate": 8e-05,
+      "loss": 1.8067,
+      "step": 522
+    },
+    {
+      "epoch": 0.12708054914348196,
+      "grad_norm": 0.43716058135032654,
+      "learning_rate": 8e-05,
+      "loss": 1.8106,
+      "step": 523
+    },
+    {
+      "epoch": 0.1273235329850565,
+      "grad_norm": 0.43253129720687866,
+      "learning_rate": 8e-05,
+      "loss": 1.8653,
+      "step": 524
+    },
+    {
+      "epoch": 0.12756651682663103,
+      "grad_norm": 0.4163695275783539,
+      "learning_rate": 8e-05,
+      "loss": 1.6138,
+      "step": 525
+    },
+    {
+      "epoch": 0.12780950066820557,
+      "grad_norm": 0.39770838618278503,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 526
+    },
+    {
+      "epoch": 0.1280524845097801,
+      "grad_norm": 0.416168212890625,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 527
+    },
+    {
+      "epoch": 0.12829546835135464,
+      "grad_norm": 0.393230140209198,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 528
+    },
+    {
+      "epoch": 0.12853845219292917,
+      "grad_norm": 0.4247177839279175,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 529
+    },
+    {
+      "epoch": 0.1287814360345037,
+      "grad_norm": 0.41639310121536255,
+      "learning_rate": 8e-05,
+      "loss": 1.7591,
+      "step": 530
+    },
+    {
+      "epoch": 0.12902441987607824,
+      "grad_norm": 0.4266749322414398,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 531
+    },
+    {
+      "epoch": 0.12926740371765277,
+      "grad_norm": 0.40807127952575684,
+      "learning_rate": 8e-05,
+      "loss": 1.6978,
+      "step": 532
+    },
+    {
+      "epoch": 0.1295103875592273,
+      "grad_norm": 0.426586389541626,
+      "learning_rate": 8e-05,
+      "loss": 1.756,
+      "step": 533
+    },
+    {
+      "epoch": 0.12975337140080184,
+      "grad_norm": 0.41720592975616455,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 534
+    },
+    {
+      "epoch": 0.12999635524237638,
+      "grad_norm": 0.4057789146900177,
+      "learning_rate": 8e-05,
+      "loss": 1.7339,
+      "step": 535
+    },
+    {
+      "epoch": 0.1302393390839509,
+      "grad_norm": 0.43273788690567017,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 536
+    },
+    {
+      "epoch": 0.13048232292552545,
+      "grad_norm": 0.42253121733665466,
+      "learning_rate": 8e-05,
+      "loss": 1.8726,
+      "step": 537
+    },
+    {
+      "epoch": 0.13072530676709998,
+      "grad_norm": 0.4509675204753876,
+      "learning_rate": 8e-05,
+      "loss": 1.8426,
+      "step": 538
+    },
+    {
+      "epoch": 0.13096829060867451,
+      "grad_norm": 0.4126783311367035,
+      "learning_rate": 8e-05,
+      "loss": 1.8826,
+      "step": 539
+    },
+    {
+      "epoch": 0.13121127445024905,
+      "grad_norm": 0.4048231840133667,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 540
+    },
+    {
+      "epoch": 0.13145425829182358,
+      "grad_norm": 0.4085652828216553,
+      "learning_rate": 8e-05,
+      "loss": 1.7383,
+      "step": 541
+    },
+    {
+      "epoch": 0.13169724213339812,
+      "grad_norm": 0.41886651515960693,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 542
+    },
+    {
+      "epoch": 0.13194022597497265,
+      "grad_norm": 0.4263283610343933,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 543
+    },
+    {
+      "epoch": 0.1321832098165472,
+      "grad_norm": 0.409944623708725,
+      "learning_rate": 8e-05,
+      "loss": 1.7645,
+      "step": 544
+    },
+    {
+      "epoch": 0.13242619365812172,
+      "grad_norm": 0.45405811071395874,
+      "learning_rate": 8e-05,
+      "loss": 1.8067,
+      "step": 545
+    },
+    {
+      "epoch": 0.13266917749969628,
+      "grad_norm": 0.4124011695384979,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 546
+    },
+    {
+      "epoch": 0.13291216134127082,
+      "grad_norm": 0.4117543399333954,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 547
+    },
+    {
+      "epoch": 0.13315514518284535,
+      "grad_norm": 0.4148860275745392,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 548
+    },
+    {
+      "epoch": 0.1333981290244199,
+      "grad_norm": 0.39591357111930847,
+      "learning_rate": 8e-05,
+      "loss": 1.7578,
+      "step": 549
+    },
+    {
+      "epoch": 0.13364111286599442,
+      "grad_norm": 0.41340890526771545,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 550
+    },
+    {
+      "epoch": 0.13388409670756896,
+      "grad_norm": 0.41868317127227783,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 551
+    },
+    {
+      "epoch": 0.1341270805491435,
+      "grad_norm": 0.392217218875885,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 552
+    },
+    {
+      "epoch": 0.13437006439071802,
+      "grad_norm": 0.41015952825546265,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 553
+    },
+    {
+      "epoch": 0.13461304823229256,
+      "grad_norm": 0.4273151457309723,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 554
+    },
+    {
+      "epoch": 0.1348560320738671,
+      "grad_norm": 0.4185205101966858,
+      "learning_rate": 8e-05,
+      "loss": 1.7621,
+      "step": 555
+    },
+    {
+      "epoch": 0.13509901591544163,
+      "grad_norm": 0.3908994197845459,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 556
+    },
+    {
+      "epoch": 0.13534199975701616,
+      "grad_norm": 0.3928578794002533,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 557
+    },
+    {
+      "epoch": 0.1355849835985907,
+      "grad_norm": 0.4071027636528015,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 558
+    },
+    {
+      "epoch": 0.13582796744016523,
+      "grad_norm": 0.4001491963863373,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 559
+    },
+    {
+      "epoch": 0.13607095128173977,
+      "grad_norm": 0.42042168974876404,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 560
+    },
+    {
+      "epoch": 0.1363139351233143,
+      "grad_norm": 0.4091740846633911,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 561
+    },
+    {
+      "epoch": 0.13655691896488883,
+      "grad_norm": 0.4089222848415375,
+      "learning_rate": 8e-05,
+      "loss": 1.8531,
+      "step": 562
+    },
+    {
+      "epoch": 0.13679990280646337,
+      "grad_norm": 0.39700132608413696,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 563
+    },
+    {
+      "epoch": 0.1370428866480379,
+      "grad_norm": 0.3812037408351898,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 564
+    },
+    {
+      "epoch": 0.13728587048961244,
+      "grad_norm": 0.4124666750431061,
+      "learning_rate": 8e-05,
+      "loss": 1.7889,
+      "step": 565
+    },
+    {
+      "epoch": 0.13752885433118697,
+      "grad_norm": 0.4233843982219696,
+      "learning_rate": 8e-05,
+      "loss": 1.8359,
+      "step": 566
+    },
+    {
+      "epoch": 0.1377718381727615,
+      "grad_norm": 0.40482261776924133,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 567
+    },
+    {
+      "epoch": 0.13801482201433604,
+      "grad_norm": 0.40571871399879456,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 568
+    },
+    {
+      "epoch": 0.13825780585591058,
+      "grad_norm": 0.4105265140533447,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 569
+    },
+    {
+      "epoch": 0.1385007896974851,
+      "grad_norm": 0.3944711685180664,
+      "learning_rate": 8e-05,
+      "loss": 1.7262,
+      "step": 570
+    },
+    {
+      "epoch": 0.13874377353905964,
+      "grad_norm": 0.415972501039505,
+      "learning_rate": 8e-05,
+      "loss": 1.797,
+      "step": 571
+    },
+    {
+      "epoch": 0.13898675738063418,
+      "grad_norm": 0.41803696751594543,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 572
+    },
+    {
+      "epoch": 0.1392297412222087,
+      "grad_norm": 0.42807522416114807,
+      "learning_rate": 8e-05,
+      "loss": 1.8867,
+      "step": 573
+    },
+    {
+      "epoch": 0.13947272506378325,
+      "grad_norm": 0.4009261727333069,
+      "learning_rate": 8e-05,
+      "loss": 1.7295,
+      "step": 574
+    },
+    {
+      "epoch": 0.13971570890535778,
+      "grad_norm": 0.4152818024158478,
+      "learning_rate": 8e-05,
+      "loss": 1.7609,
+      "step": 575
+    },
+    {
+      "epoch": 0.13995869274693232,
+      "grad_norm": 0.47868460416793823,
+      "learning_rate": 8e-05,
+      "loss": 1.8627,
+      "step": 576
+    },
+    {
+      "epoch": 0.14020167658850685,
+      "grad_norm": 0.409762442111969,
+      "learning_rate": 8e-05,
+      "loss": 1.8045,
+      "step": 577
+    },
+    {
+      "epoch": 0.14044466043008139,
+      "grad_norm": 0.6949756145477295,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 578
+    },
+    {
+      "epoch": 0.14068764427165595,
+      "grad_norm": 0.45087161660194397,
+      "learning_rate": 8e-05,
+      "loss": 1.7881,
+      "step": 579
+    },
+    {
+      "epoch": 0.14093062811323048,
+      "grad_norm": 0.46107062697410583,
+      "learning_rate": 8e-05,
+      "loss": 1.8546,
+      "step": 580
+    },
+    {
+      "epoch": 0.14117361195480502,
+      "grad_norm": 0.39177265763282776,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 581
+    },
+    {
+      "epoch": 0.14141659579637955,
+      "grad_norm": 0.4238806962966919,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 582
+    },
+    {
+      "epoch": 0.1416595796379541,
+      "grad_norm": 0.4344601333141327,
+      "learning_rate": 8e-05,
+      "loss": 1.7946,
+      "step": 583
+    },
+    {
+      "epoch": 0.14190256347952862,
+      "grad_norm": 0.39927345514297485,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 584
+    },
+    {
+      "epoch": 0.14214554732110316,
+      "grad_norm": 0.44042500853538513,
+      "learning_rate": 8e-05,
+      "loss": 1.6398,
+      "step": 585
+    },
+    {
+      "epoch": 0.1423885311626777,
+      "grad_norm": 0.41562238335609436,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 586
+    },
+    {
+      "epoch": 0.14263151500425222,
+      "grad_norm": 0.41256847977638245,
+      "learning_rate": 8e-05,
+      "loss": 1.7209,
+      "step": 587
+    },
+    {
+      "epoch": 0.14287449884582676,
+      "grad_norm": 0.4304940104484558,
+      "learning_rate": 8e-05,
+      "loss": 1.8157,
+      "step": 588
+    },
+    {
+      "epoch": 0.1431174826874013,
+      "grad_norm": 0.4045521914958954,
+      "learning_rate": 8e-05,
+      "loss": 1.5891,
+      "step": 589
+    },
+    {
+      "epoch": 0.14336046652897583,
+      "grad_norm": 0.41248899698257446,
+      "learning_rate": 8e-05,
+      "loss": 1.8079,
+      "step": 590
+    },
+    {
+      "epoch": 0.14360345037055036,
+      "grad_norm": 0.40122056007385254,
+      "learning_rate": 8e-05,
+      "loss": 1.6265,
+      "step": 591
+    },
+    {
+      "epoch": 0.1438464342121249,
+      "grad_norm": 0.4546814560890198,
+      "learning_rate": 8e-05,
+      "loss": 1.922,
+      "step": 592
+    },
+    {
+      "epoch": 0.14408941805369943,
+      "grad_norm": 0.4200947880744934,
+      "learning_rate": 8e-05,
+      "loss": 1.7937,
+      "step": 593
+    },
+    {
+      "epoch": 0.14433240189527397,
+      "grad_norm": 0.42049646377563477,
+      "learning_rate": 8e-05,
+      "loss": 1.7916,
+      "step": 594
+    },
+    {
+      "epoch": 0.1445753857368485,
+      "grad_norm": 0.4661625921726227,
+      "learning_rate": 8e-05,
+      "loss": 1.8344,
+      "step": 595
+    },
+    {
+      "epoch": 0.14481836957842303,
+      "grad_norm": 0.4149019718170166,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 596
+    },
+    {
+      "epoch": 0.14506135341999757,
+      "grad_norm": 0.41684791445732117,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 597
+    },
+    {
+      "epoch": 0.1453043372615721,
+      "grad_norm": 0.4358704686164856,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 598
+    },
+    {
+      "epoch": 0.14554732110314664,
+      "grad_norm": 0.4250600337982178,
+      "learning_rate": 8e-05,
+      "loss": 1.6269,
+      "step": 599
+    },
+    {
+      "epoch": 0.14579030494472117,
+      "grad_norm": 0.42459315061569214,
+      "learning_rate": 8e-05,
+      "loss": 1.6598,
+      "step": 600
+    },
+    {
+      "epoch": 0.1460332887862957,
+      "grad_norm": 0.4363904893398285,
+      "learning_rate": 8e-05,
+      "loss": 1.8471,
+      "step": 601
+    },
+    {
+      "epoch": 0.14627627262787024,
+      "grad_norm": 0.41599100828170776,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 602
+    },
+    {
+      "epoch": 0.14651925646944478,
+      "grad_norm": 0.3999607264995575,
+      "learning_rate": 8e-05,
+      "loss": 1.6048,
+      "step": 603
+    },
+    {
+      "epoch": 0.1467622403110193,
+      "grad_norm": 0.4132905900478363,
+      "learning_rate": 8e-05,
+      "loss": 1.784,
+      "step": 604
+    },
+    {
+      "epoch": 0.14700522415259384,
+      "grad_norm": 0.40192580223083496,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 605
+    },
+    {
+      "epoch": 0.14724820799416838,
+      "grad_norm": 0.4473113715648651,
+      "learning_rate": 8e-05,
+      "loss": 1.7678,
+      "step": 606
+    },
+    {
+      "epoch": 0.1474911918357429,
+      "grad_norm": 0.4752279222011566,
+      "learning_rate": 8e-05,
+      "loss": 1.7444,
+      "step": 607
+    },
+    {
+      "epoch": 0.14773417567731745,
+      "grad_norm": 0.4193718135356903,
+      "learning_rate": 8e-05,
+      "loss": 1.6125,
+      "step": 608
+    },
+    {
+      "epoch": 0.14797715951889198,
+      "grad_norm": 0.4258435368537903,
+      "learning_rate": 8e-05,
+      "loss": 1.8557,
+      "step": 609
+    },
+    {
+      "epoch": 0.14822014336046652,
+      "grad_norm": 0.4286644756793976,
+      "learning_rate": 8e-05,
+      "loss": 1.6663,
+      "step": 610
+    },
+    {
+      "epoch": 0.14846312720204105,
+      "grad_norm": 0.41051027178764343,
+      "learning_rate": 8e-05,
+      "loss": 1.6868,
+      "step": 611
+    },
+    {
+      "epoch": 0.1487061110436156,
+      "grad_norm": 0.4174746870994568,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 612
+    },
+    {
+      "epoch": 0.14894909488519015,
+      "grad_norm": 0.41176313161849976,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 613
+    },
+    {
+      "epoch": 0.14919207872676468,
+      "grad_norm": 0.4219014644622803,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 614
+    },
+    {
+      "epoch": 0.14943506256833922,
+      "grad_norm": 0.41391685605049133,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 615
+    },
+    {
+      "epoch": 0.14967804640991375,
+      "grad_norm": 0.4116523563861847,
+      "learning_rate": 8e-05,
+      "loss": 1.8255,
+      "step": 616
+    },
+    {
+      "epoch": 0.14992103025148829,
+      "grad_norm": 0.4185222089290619,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 617
+    },
+    {
+      "epoch": 0.15016401409306282,
+      "grad_norm": 0.4165772497653961,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 618
+    },
+    {
+      "epoch": 0.15040699793463735,
+      "grad_norm": 0.42195072770118713,
+      "learning_rate": 8e-05,
+      "loss": 1.8882,
+      "step": 619
+    },
+    {
+      "epoch": 0.1506499817762119,
+      "grad_norm": 0.43403923511505127,
+      "learning_rate": 8e-05,
+      "loss": 1.7437,
+      "step": 620
+    },
+    {
+      "epoch": 0.15089296561778642,
+      "grad_norm": 0.422005832195282,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 621
+    },
+    {
+      "epoch": 0.15113594945936096,
+      "grad_norm": 0.4126865565776825,
+      "learning_rate": 8e-05,
+      "loss": 1.6739,
+      "step": 622
+    },
+    {
+      "epoch": 0.1513789333009355,
+      "grad_norm": 0.42506495118141174,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 623
+    },
+    {
+      "epoch": 0.15162191714251003,
+      "grad_norm": 0.4353766441345215,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 624
+    },
+    {
+      "epoch": 0.15186490098408456,
+      "grad_norm": 0.43494346737861633,
+      "learning_rate": 8e-05,
+      "loss": 1.8059,
+      "step": 625
+    },
+    {
+      "epoch": 0.1521078848256591,
+      "grad_norm": 0.4109300374984741,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 626
+    },
+    {
+      "epoch": 0.15235086866723363,
+      "grad_norm": 0.4274158477783203,
+      "learning_rate": 8e-05,
+      "loss": 1.877,
+      "step": 627
+    },
+    {
+      "epoch": 0.15259385250880816,
+      "grad_norm": 0.4044157862663269,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 628
+    },
+    {
+      "epoch": 0.1528368363503827,
+      "grad_norm": 0.41327598690986633,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 629
+    },
+    {
+      "epoch": 0.15307982019195723,
+      "grad_norm": 0.4174918830394745,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 630
+    },
+    {
+      "epoch": 0.15332280403353177,
+      "grad_norm": 0.4256085157394409,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 631
+    },
+    {
+      "epoch": 0.1535657878751063,
+      "grad_norm": 0.40840238332748413,
+      "learning_rate": 8e-05,
+      "loss": 1.7003,
+      "step": 632
+    },
+    {
+      "epoch": 0.15380877171668084,
+      "grad_norm": 0.43425703048706055,
+      "learning_rate": 8e-05,
+      "loss": 1.7694,
+      "step": 633
+    },
+    {
+      "epoch": 0.15405175555825537,
+      "grad_norm": 0.4183935821056366,
+      "learning_rate": 8e-05,
+      "loss": 1.7218,
+      "step": 634
+    },
+    {
+      "epoch": 0.1542947393998299,
+      "grad_norm": 0.40338075160980225,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 635
+    },
+    {
+      "epoch": 0.15453772324140444,
+      "grad_norm": 0.4134522080421448,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 636
+    },
+    {
+      "epoch": 0.15478070708297897,
+      "grad_norm": 0.4205785393714905,
+      "learning_rate": 8e-05,
+      "loss": 1.7742,
+      "step": 637
+    },
+    {
+      "epoch": 0.1550236909245535,
+      "grad_norm": 0.39972954988479614,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 638
+    },
+    {
+      "epoch": 0.15526667476612804,
+      "grad_norm": 0.4296969473361969,
+      "learning_rate": 8e-05,
+      "loss": 1.9122,
+      "step": 639
+    },
+    {
+      "epoch": 0.15550965860770258,
+      "grad_norm": 0.45631009340286255,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 640
+    },
+    {
+      "epoch": 0.1557526424492771,
+      "grad_norm": 0.4262036979198456,
+      "learning_rate": 8e-05,
+      "loss": 1.7179,
+      "step": 641
+    },
+    {
+      "epoch": 0.15599562629085165,
+      "grad_norm": 0.4598737955093384,
+      "learning_rate": 8e-05,
+      "loss": 1.697,
+      "step": 642
+    },
+    {
+      "epoch": 0.15623861013242618,
+      "grad_norm": 0.4666830897331238,
+      "learning_rate": 8e-05,
+      "loss": 1.7301,
+      "step": 643
+    },
+    {
+      "epoch": 0.15648159397400072,
+      "grad_norm": 0.41235294938087463,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 644
+    },
+    {
+      "epoch": 0.15672457781557528,
+      "grad_norm": 0.45842328667640686,
+      "learning_rate": 8e-05,
+      "loss": 1.72,
+      "step": 645
+    },
+    {
+      "epoch": 0.1569675616571498,
+      "grad_norm": 0.448718786239624,
+      "learning_rate": 8e-05,
+      "loss": 1.9079,
+      "step": 646
+    },
+    {
+      "epoch": 0.15721054549872435,
+      "grad_norm": 0.42486071586608887,
+      "learning_rate": 8e-05,
+      "loss": 1.7585,
+      "step": 647
+    },
+    {
+      "epoch": 0.15745352934029888,
+      "grad_norm": 0.45865485072135925,
+      "learning_rate": 8e-05,
+      "loss": 1.786,
+      "step": 648
+    },
+    {
+      "epoch": 0.15769651318187342,
+      "grad_norm": 0.4131222367286682,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 649
+    },
+    {
+      "epoch": 0.15793949702344795,
+      "grad_norm": 0.42305582761764526,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 650
+    },
+    {
+      "epoch": 0.15818248086502248,
+      "grad_norm": 0.42484596371650696,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 651
+    },
+    {
+      "epoch": 0.15842546470659702,
+      "grad_norm": 0.41841161251068115,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 652
+    },
+    {
+      "epoch": 0.15866844854817155,
+      "grad_norm": 0.42979368567466736,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 653
+    },
+    {
+      "epoch": 0.1589114323897461,
+      "grad_norm": 0.4110826551914215,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 654
+    },
+    {
+      "epoch": 0.15915441623132062,
+      "grad_norm": 0.4090655446052551,
+      "learning_rate": 8e-05,
+      "loss": 1.7052,
+      "step": 655
+    },
+    {
+      "epoch": 0.15939740007289516,
+      "grad_norm": 0.4239189624786377,
+      "learning_rate": 8e-05,
+      "loss": 1.9034,
+      "step": 656
+    },
+    {
+      "epoch": 0.1596403839144697,
+      "grad_norm": 0.4068964421749115,
+      "learning_rate": 8e-05,
+      "loss": 1.7131,
+      "step": 657
+    },
+    {
+      "epoch": 0.15988336775604423,
+      "grad_norm": 0.4201562702655792,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 658
+    },
+    {
+      "epoch": 0.16012635159761876,
+      "grad_norm": 0.4384630620479584,
+      "learning_rate": 8e-05,
+      "loss": 1.8174,
+      "step": 659
+    },
+    {
+      "epoch": 0.1603693354391933,
+      "grad_norm": 0.4189380407333374,
+      "learning_rate": 8e-05,
+      "loss": 1.7293,
+      "step": 660
+    },
+    {
+      "epoch": 0.16061231928076783,
+      "grad_norm": 0.43468591570854187,
+      "learning_rate": 8e-05,
+      "loss": 1.8541,
+      "step": 661
+    },
+    {
+      "epoch": 0.16085530312234236,
+      "grad_norm": 0.44042378664016724,
+      "learning_rate": 8e-05,
+      "loss": 1.8672,
+      "step": 662
+    },
+    {
+      "epoch": 0.1610982869639169,
+      "grad_norm": 0.39210212230682373,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 663
+    },
+    {
+      "epoch": 0.16134127080549143,
+      "grad_norm": 0.4077080488204956,
+      "learning_rate": 8e-05,
+      "loss": 1.7045,
+      "step": 664
+    },
+    {
+      "epoch": 0.16158425464706597,
+      "grad_norm": 0.413170725107193,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 665
+    },
+    {
+      "epoch": 0.1618272384886405,
+      "grad_norm": 0.4146006107330322,
+      "learning_rate": 8e-05,
+      "loss": 1.736,
+      "step": 666
+    },
+    {
+      "epoch": 0.16207022233021504,
+      "grad_norm": 0.41123950481414795,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 667
+    },
+    {
+      "epoch": 0.16231320617178957,
+      "grad_norm": 0.4749496579170227,
+      "learning_rate": 8e-05,
+      "loss": 1.7646,
+      "step": 668
+    },
+    {
+      "epoch": 0.1625561900133641,
+      "grad_norm": 0.4240420162677765,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 669
+    },
+    {
+      "epoch": 0.16279917385493864,
+      "grad_norm": 0.3952424228191376,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 670
+    },
+    {
+      "epoch": 0.16304215769651317,
+      "grad_norm": 0.39755764603614807,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 671
+    },
+    {
+      "epoch": 0.1632851415380877,
+      "grad_norm": 0.41174235939979553,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 672
+    },
+    {
+      "epoch": 0.16352812537966224,
+      "grad_norm": 0.4327336847782135,
+      "learning_rate": 8e-05,
+      "loss": 1.8273,
+      "step": 673
+    },
+    {
+      "epoch": 0.16377110922123678,
+      "grad_norm": 0.3779991865158081,
+      "learning_rate": 8e-05,
+      "loss": 1.3962,
+      "step": 674
+    },
+    {
+      "epoch": 0.1640140930628113,
+      "grad_norm": 0.40893781185150146,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 675
+    },
+    {
+      "epoch": 0.16425707690438585,
+      "grad_norm": 0.41329607367515564,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 676
+    },
+    {
+      "epoch": 0.16450006074596038,
+      "grad_norm": 0.401868999004364,
+      "learning_rate": 8e-05,
+      "loss": 1.8449,
+      "step": 677
+    },
+    {
+      "epoch": 0.16474304458753494,
+      "grad_norm": 0.4251278340816498,
+      "learning_rate": 8e-05,
+      "loss": 1.8646,
+      "step": 678
+    },
+    {
+      "epoch": 0.16498602842910948,
+      "grad_norm": 0.4233865737915039,
+      "learning_rate": 8e-05,
+      "loss": 1.8867,
+      "step": 679
+    },
+    {
+      "epoch": 0.165229012270684,
+      "grad_norm": 0.4191146194934845,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 680
+    },
+    {
+      "epoch": 0.16547199611225855,
+      "grad_norm": 0.41691043972969055,
+      "learning_rate": 8e-05,
+      "loss": 1.8327,
+      "step": 681
+    },
+    {
+      "epoch": 0.16571497995383308,
+      "grad_norm": 0.41075795888900757,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 682
+    },
+    {
+      "epoch": 0.16595796379540761,
+      "grad_norm": 0.41517287492752075,
+      "learning_rate": 8e-05,
+      "loss": 1.7031,
+      "step": 683
+    },
+    {
+      "epoch": 0.16620094763698215,
+      "grad_norm": 0.4045945703983307,
+      "learning_rate": 8e-05,
+      "loss": 1.7285,
+      "step": 684
+    },
+    {
+      "epoch": 0.16644393147855668,
+      "grad_norm": 0.40752533078193665,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 685
+    },
+    {
+      "epoch": 0.16668691532013122,
+      "grad_norm": 0.5462080240249634,
+      "learning_rate": 8e-05,
+      "loss": 1.741,
+      "step": 686
+    },
+    {
+      "epoch": 0.16692989916170575,
+      "grad_norm": 0.43143707513809204,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 687
+    },
+    {
+      "epoch": 0.1671728830032803,
+      "grad_norm": 0.4070568382740021,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 688
+    },
+    {
+      "epoch": 0.16741586684485482,
+      "grad_norm": 0.4276026785373688,
+      "learning_rate": 8e-05,
+      "loss": 1.8271,
+      "step": 689
+    },
+    {
+      "epoch": 0.16765885068642936,
+      "grad_norm": 0.4280378520488739,
+      "learning_rate": 8e-05,
+      "loss": 1.8633,
+      "step": 690
+    },
+    {
+      "epoch": 0.1679018345280039,
+      "grad_norm": 0.43015697598457336,
+      "learning_rate": 8e-05,
+      "loss": 1.7541,
+      "step": 691
+    },
+    {
+      "epoch": 0.16814481836957842,
+      "grad_norm": 0.41025182604789734,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 692
+    },
+    {
+      "epoch": 0.16838780221115296,
+      "grad_norm": 0.41091296076774597,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 693
+    },
+    {
+      "epoch": 0.1686307860527275,
+      "grad_norm": 0.3976031243801117,
+      "learning_rate": 8e-05,
+      "loss": 1.6933,
+      "step": 694
+    },
+    {
+      "epoch": 0.16887376989430203,
+      "grad_norm": 0.3983782231807709,
+      "learning_rate": 8e-05,
+      "loss": 1.7087,
+      "step": 695
+    },
+    {
+      "epoch": 0.16911675373587656,
+      "grad_norm": 0.3997262716293335,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 696
+    },
+    {
+      "epoch": 0.1693597375774511,
+      "grad_norm": 0.4466289281845093,
+      "learning_rate": 8e-05,
+      "loss": 1.9315,
+      "step": 697
+    },
+    {
+      "epoch": 0.16960272141902563,
+      "grad_norm": 0.40420153737068176,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 698
+    },
+    {
+      "epoch": 0.16984570526060017,
+      "grad_norm": 0.39683201909065247,
+      "learning_rate": 8e-05,
+      "loss": 1.7654,
+      "step": 699
+    },
+    {
+      "epoch": 0.1700886891021747,
+      "grad_norm": 0.39208847284317017,
+      "learning_rate": 8e-05,
+      "loss": 1.6486,
+      "step": 700
+    },
+    {
+      "epoch": 0.17033167294374923,
+      "grad_norm": 0.40302395820617676,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 701
+    },
+    {
+      "epoch": 0.17057465678532377,
+      "grad_norm": 0.40285834670066833,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 702
+    },
+    {
+      "epoch": 0.1708176406268983,
+      "grad_norm": 0.43496444821357727,
+      "learning_rate": 8e-05,
+      "loss": 1.7652,
+      "step": 703
+    },
+    {
+      "epoch": 0.17106062446847284,
+      "grad_norm": 0.408661812543869,
+      "learning_rate": 8e-05,
+      "loss": 1.8078,
+      "step": 704
+    },
+    {
+      "epoch": 0.17130360831004737,
+      "grad_norm": 0.4155483841896057,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 705
+    },
+    {
+      "epoch": 0.1715465921516219,
+      "grad_norm": 0.4108770787715912,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 706
+    },
+    {
+      "epoch": 0.17178957599319644,
+      "grad_norm": 0.43399226665496826,
+      "learning_rate": 8e-05,
+      "loss": 1.8858,
+      "step": 707
+    },
+    {
+      "epoch": 0.17203255983477098,
+      "grad_norm": 0.43061381578445435,
+      "learning_rate": 8e-05,
+      "loss": 1.6588,
+      "step": 708
+    },
+    {
+      "epoch": 0.1722755436763455,
+      "grad_norm": 0.4035395681858063,
+      "learning_rate": 8e-05,
+      "loss": 1.7147,
+      "step": 709
+    },
+    {
+      "epoch": 0.17251852751792005,
+      "grad_norm": 0.4377632439136505,
+      "learning_rate": 8e-05,
+      "loss": 1.7836,
+      "step": 710
+    },
+    {
+      "epoch": 0.1727615113594946,
+      "grad_norm": 0.4374580383300781,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 711
+    },
+    {
+      "epoch": 0.17300449520106914,
+      "grad_norm": 0.43939241766929626,
+      "learning_rate": 8e-05,
+      "loss": 1.8797,
+      "step": 712
+    },
+    {
+      "epoch": 0.17324747904264368,
+      "grad_norm": 0.4374196231365204,
+      "learning_rate": 8e-05,
+      "loss": 1.7044,
+      "step": 713
+    },
+    {
+      "epoch": 0.1734904628842182,
+      "grad_norm": 0.4204579293727875,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 714
+    },
+    {
+      "epoch": 0.17373344672579275,
+      "grad_norm": 0.4335678815841675,
+      "learning_rate": 8e-05,
+      "loss": 1.8633,
+      "step": 715
+    },
+    {
+      "epoch": 0.17397643056736728,
+      "grad_norm": 0.39257657527923584,
+      "learning_rate": 8e-05,
+      "loss": 1.5843,
+      "step": 716
+    },
+    {
+      "epoch": 0.17421941440894181,
+      "grad_norm": 0.4656222462654114,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 717
+    },
+    {
+      "epoch": 0.17446239825051635,
+      "grad_norm": 0.42730045318603516,
+      "learning_rate": 8e-05,
+      "loss": 1.8616,
+      "step": 718
+    },
+    {
+      "epoch": 0.17470538209209088,
+      "grad_norm": 0.38788118958473206,
+      "learning_rate": 8e-05,
+      "loss": 1.6818,
+      "step": 719
+    },
+    {
+      "epoch": 0.17494836593366542,
+      "grad_norm": 0.4097652733325958,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 720
+    },
+    {
+      "epoch": 0.17519134977523995,
+      "grad_norm": 0.4379699230194092,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 721
+    },
+    {
+      "epoch": 0.1754343336168145,
+      "grad_norm": 0.44912099838256836,
+      "learning_rate": 8e-05,
+      "loss": 1.7851,
+      "step": 722
+    },
+    {
+      "epoch": 0.17567731745838902,
+      "grad_norm": 0.4183383882045746,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 723
+    },
+    {
+      "epoch": 0.17592030129996356,
+      "grad_norm": 0.4096759259700775,
+      "learning_rate": 8e-05,
+      "loss": 1.6585,
+      "step": 724
+    },
+    {
+      "epoch": 0.1761632851415381,
+      "grad_norm": 0.4122353196144104,
+      "learning_rate": 8e-05,
+      "loss": 1.7693,
+      "step": 725
+    },
+    {
+      "epoch": 0.17640626898311262,
+      "grad_norm": 0.39519456028938293,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 726
+    },
+    {
+      "epoch": 0.17664925282468716,
+      "grad_norm": 0.419357031583786,
+      "learning_rate": 8e-05,
+      "loss": 1.7755,
+      "step": 727
+    },
+    {
+      "epoch": 0.1768922366662617,
+      "grad_norm": 0.39010873436927795,
+      "learning_rate": 8e-05,
+      "loss": 1.6475,
+      "step": 728
+    },
+    {
+      "epoch": 0.17713522050783623,
+      "grad_norm": 0.41785943508148193,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 729
+    },
+    {
+      "epoch": 0.17737820434941076,
+      "grad_norm": 0.4477573037147522,
+      "learning_rate": 8e-05,
+      "loss": 1.822,
+      "step": 730
+    },
+    {
+      "epoch": 0.1776211881909853,
+      "grad_norm": 0.40994468331336975,
+      "learning_rate": 8e-05,
+      "loss": 1.7401,
+      "step": 731
+    },
+    {
+      "epoch": 0.17786417203255983,
+      "grad_norm": 0.4095614552497864,
+      "learning_rate": 8e-05,
+      "loss": 1.7347,
+      "step": 732
+    },
+    {
+      "epoch": 0.17810715587413437,
+      "grad_norm": 0.42191195487976074,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 733
+    },
+    {
+      "epoch": 0.1783501397157089,
+      "grad_norm": 0.4296643137931824,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 734
+    },
+    {
+      "epoch": 0.17859312355728343,
+      "grad_norm": 0.41722437739372253,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 735
+    },
+    {
+      "epoch": 0.17883610739885797,
+      "grad_norm": 0.41322436928749084,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 736
+    },
+    {
+      "epoch": 0.1790790912404325,
+      "grad_norm": 0.4126609265804291,
+      "learning_rate": 8e-05,
+      "loss": 1.7489,
+      "step": 737
+    },
+    {
+      "epoch": 0.17932207508200704,
+      "grad_norm": 0.41765400767326355,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 738
+    },
+    {
+      "epoch": 0.17956505892358157,
+      "grad_norm": 0.41525664925575256,
+      "learning_rate": 8e-05,
+      "loss": 1.777,
+      "step": 739
+    },
+    {
+      "epoch": 0.1798080427651561,
+      "grad_norm": 0.42361241579055786,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 740
+    },
+    {
+      "epoch": 0.18005102660673064,
+      "grad_norm": 0.40086159110069275,
+      "learning_rate": 8e-05,
+      "loss": 1.6845,
+      "step": 741
+    },
+    {
+      "epoch": 0.18029401044830518,
+      "grad_norm": 0.433889240026474,
+      "learning_rate": 8e-05,
+      "loss": 1.7732,
+      "step": 742
+    },
+    {
+      "epoch": 0.1805369942898797,
+      "grad_norm": 0.4261779487133026,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 743
+    },
+    {
+      "epoch": 0.18077997813145427,
+      "grad_norm": 0.41455700993537903,
+      "learning_rate": 8e-05,
+      "loss": 1.8022,
+      "step": 744
+    },
+    {
+      "epoch": 0.1810229619730288,
+      "grad_norm": 0.41002756357192993,
+      "learning_rate": 8e-05,
+      "loss": 1.8093,
+      "step": 745
+    },
+    {
+      "epoch": 0.18126594581460334,
+      "grad_norm": 0.4073259234428406,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 746
+    },
+    {
+      "epoch": 0.18150892965617788,
+      "grad_norm": 0.4292190372943878,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 747
+    },
+    {
+      "epoch": 0.1817519134977524,
+      "grad_norm": 0.3857293725013733,
+      "learning_rate": 8e-05,
+      "loss": 1.7038,
+      "step": 748
+    },
+    {
+      "epoch": 0.18199489733932694,
+      "grad_norm": 0.4110948145389557,
+      "learning_rate": 8e-05,
+      "loss": 1.7761,
+      "step": 749
+    },
+    {
+      "epoch": 0.18223788118090148,
+      "grad_norm": 0.4361400306224823,
+      "learning_rate": 8e-05,
+      "loss": 1.5824,
+      "step": 750
+    },
+    {
+      "epoch": 0.182480865022476,
+      "grad_norm": 0.4137819707393646,
+      "learning_rate": 8e-05,
+      "loss": 1.6897,
+      "step": 751
+    },
+    {
+      "epoch": 0.18272384886405055,
+      "grad_norm": 0.4077703058719635,
+      "learning_rate": 8e-05,
+      "loss": 1.7019,
+      "step": 752
+    },
+    {
+      "epoch": 0.18296683270562508,
+      "grad_norm": 0.44059401750564575,
+      "learning_rate": 8e-05,
+      "loss": 1.6948,
+      "step": 753
+    },
+    {
+      "epoch": 0.18320981654719962,
+      "grad_norm": 0.4075104296207428,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 754
+    },
+    {
+      "epoch": 0.18345280038877415,
+      "grad_norm": 0.4188830554485321,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 755
+    },
+    {
+      "epoch": 0.18369578423034869,
+      "grad_norm": 0.41404202580451965,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 756
+    },
+    {
+      "epoch": 0.18393876807192322,
+      "grad_norm": 0.41878440976142883,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 757
+    },
+    {
+      "epoch": 0.18418175191349775,
+      "grad_norm": 0.43462273478507996,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 758
+    },
+    {
+      "epoch": 0.1844247357550723,
+      "grad_norm": 0.39942023158073425,
+      "learning_rate": 8e-05,
+      "loss": 1.6298,
+      "step": 759
+    },
+    {
+      "epoch": 0.18466771959664682,
+      "grad_norm": 0.4408327341079712,
+      "learning_rate": 8e-05,
+      "loss": 1.7053,
+      "step": 760
+    },
+    {
+      "epoch": 0.18491070343822136,
+      "grad_norm": 0.45880624651908875,
+      "learning_rate": 8e-05,
+      "loss": 1.8,
+      "step": 761
+    },
+    {
+      "epoch": 0.1851536872797959,
+      "grad_norm": 0.4250109791755676,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 762
+    },
+    {
+      "epoch": 0.18539667112137043,
+      "grad_norm": 0.43293994665145874,
+      "learning_rate": 8e-05,
+      "loss": 1.8144,
+      "step": 763
+    },
+    {
+      "epoch": 0.18563965496294496,
+      "grad_norm": 0.3994579613208771,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 764
+    },
+    {
+      "epoch": 0.1858826388045195,
+      "grad_norm": 0.427898645401001,
+      "learning_rate": 8e-05,
+      "loss": 1.6265,
+      "step": 765
+    },
+    {
+      "epoch": 0.18612562264609403,
+      "grad_norm": 0.4274621307849884,
+      "learning_rate": 8e-05,
+      "loss": 1.8205,
+      "step": 766
+    },
+    {
+      "epoch": 0.18636860648766856,
+      "grad_norm": 0.42581209540367126,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 767
+    },
+    {
+      "epoch": 0.1866115903292431,
+      "grad_norm": 0.4216351807117462,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 768
+    },
+    {
+      "epoch": 0.18685457417081763,
+      "grad_norm": 0.41380995512008667,
+      "learning_rate": 8e-05,
+      "loss": 1.7582,
+      "step": 769
+    },
+    {
+      "epoch": 0.18709755801239217,
+      "grad_norm": 0.3948034346103668,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 770
+    },
+    {
+      "epoch": 0.1873405418539667,
+      "grad_norm": 0.40538617968559265,
+      "learning_rate": 8e-05,
+      "loss": 1.7171,
+      "step": 771
+    },
+    {
+      "epoch": 0.18758352569554124,
+      "grad_norm": 0.40522703528404236,
+      "learning_rate": 8e-05,
+      "loss": 1.742,
+      "step": 772
+    },
+    {
+      "epoch": 0.18782650953711577,
+      "grad_norm": 0.40100088715553284,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 773
+    },
+    {
+      "epoch": 0.1880694933786903,
+      "grad_norm": 0.4220276176929474,
+      "learning_rate": 8e-05,
+      "loss": 1.7515,
+      "step": 774
+    },
+    {
+      "epoch": 0.18831247722026484,
+      "grad_norm": 0.4018454849720001,
+      "learning_rate": 8e-05,
+      "loss": 1.4636,
+      "step": 775
+    },
+    {
+      "epoch": 0.18855546106183937,
+      "grad_norm": 0.4422740638256073,
+      "learning_rate": 8e-05,
+      "loss": 1.7838,
+      "step": 776
+    },
+    {
+      "epoch": 0.18879844490341394,
+      "grad_norm": 0.40587881207466125,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 777
+    },
+    {
+      "epoch": 0.18904142874498847,
+      "grad_norm": 0.4551006853580475,
+      "learning_rate": 8e-05,
+      "loss": 1.8645,
+      "step": 778
+    },
+    {
+      "epoch": 0.189284412586563,
+      "grad_norm": 0.41088271141052246,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 779
+    },
+    {
+      "epoch": 0.18952739642813754,
+      "grad_norm": 0.4191667139530182,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 780
+    },
+    {
+      "epoch": 0.18977038026971207,
+      "grad_norm": 0.4077453911304474,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 781
+    },
+    {
+      "epoch": 0.1900133641112866,
+      "grad_norm": 0.3883300721645355,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 782
+    },
+    {
+      "epoch": 0.19025634795286114,
+      "grad_norm": 0.4219571053981781,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 783
+    },
+    {
+      "epoch": 0.19049933179443568,
+      "grad_norm": 0.4366585910320282,
+      "learning_rate": 8e-05,
+      "loss": 1.7601,
+      "step": 784
+    },
+    {
+      "epoch": 0.1907423156360102,
+      "grad_norm": 0.4421660900115967,
+      "learning_rate": 8e-05,
+      "loss": 1.8607,
+      "step": 785
+    },
+    {
+      "epoch": 0.19098529947758475,
+      "grad_norm": 0.44615718722343445,
+      "learning_rate": 8e-05,
+      "loss": 1.7898,
+      "step": 786
+    },
+    {
+      "epoch": 0.19122828331915928,
+      "grad_norm": 0.39300891757011414,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 787
+    },
+    {
+      "epoch": 0.19147126716073382,
+      "grad_norm": 0.42134273052215576,
+      "learning_rate": 8e-05,
+      "loss": 1.7447,
+      "step": 788
+    },
+    {
+      "epoch": 0.19171425100230835,
+      "grad_norm": 0.4182814955711365,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 789
+    },
+    {
+      "epoch": 0.19195723484388288,
+      "grad_norm": 0.41383031010627747,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 790
+    },
+    {
+      "epoch": 0.19220021868545742,
+      "grad_norm": 0.41922250390052795,
+      "learning_rate": 8e-05,
+      "loss": 1.7863,
+      "step": 791
+    },
+    {
+      "epoch": 0.19244320252703195,
+      "grad_norm": 0.4196271300315857,
+      "learning_rate": 8e-05,
+      "loss": 1.8236,
+      "step": 792
+    },
+    {
+      "epoch": 0.1926861863686065,
+      "grad_norm": 0.41270729899406433,
+      "learning_rate": 8e-05,
+      "loss": 1.823,
+      "step": 793
+    },
+    {
+      "epoch": 0.19292917021018102,
+      "grad_norm": 0.40714430809020996,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 794
+    },
+    {
+      "epoch": 0.19317215405175556,
+      "grad_norm": 0.4369906783103943,
+      "learning_rate": 8e-05,
+      "loss": 1.8872,
+      "step": 795
+    },
+    {
+      "epoch": 0.1934151378933301,
+      "grad_norm": 0.4148472249507904,
+      "learning_rate": 8e-05,
+      "loss": 1.519,
+      "step": 796
+    },
+    {
+      "epoch": 0.19365812173490463,
+      "grad_norm": 0.4084674119949341,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 797
+    },
+    {
+      "epoch": 0.19390110557647916,
+      "grad_norm": 0.40457743406295776,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 798
+    },
+    {
+      "epoch": 0.1941440894180537,
+      "grad_norm": 0.4159713387489319,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 799
+    },
+    {
+      "epoch": 0.19438707325962823,
+      "grad_norm": 0.42744749784469604,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 800
+    },
+    {
+      "epoch": 0.19463005710120276,
+      "grad_norm": 0.40813565254211426,
+      "learning_rate": 8e-05,
+      "loss": 1.5935,
+      "step": 801
+    },
+    {
+      "epoch": 0.1948730409427773,
+      "grad_norm": 0.42466405034065247,
+      "learning_rate": 8e-05,
+      "loss": 1.7304,
+      "step": 802
+    },
+    {
+      "epoch": 0.19511602478435183,
+      "grad_norm": 0.41067975759506226,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 803
+    },
+    {
+      "epoch": 0.19535900862592637,
+      "grad_norm": 0.415198415517807,
+      "learning_rate": 8e-05,
+      "loss": 1.678,
+      "step": 804
+    },
+    {
+      "epoch": 0.1956019924675009,
+      "grad_norm": 0.43364110589027405,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 805
+    },
+    {
+      "epoch": 0.19584497630907544,
+      "grad_norm": 0.42752882838249207,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 806
+    },
+    {
+      "epoch": 0.19608796015064997,
+      "grad_norm": 0.40812936425209045,
+      "learning_rate": 8e-05,
+      "loss": 1.5766,
+      "step": 807
+    },
+    {
+      "epoch": 0.1963309439922245,
+      "grad_norm": 0.4319069981575012,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 808
+    },
+    {
+      "epoch": 0.19657392783379904,
+      "grad_norm": 0.40930795669555664,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 809
+    },
+    {
+      "epoch": 0.1968169116753736,
+      "grad_norm": 0.4289321303367615,
+      "learning_rate": 8e-05,
+      "loss": 1.7135,
+      "step": 810
+    },
+    {
+      "epoch": 0.19705989551694814,
+      "grad_norm": 0.4316776692867279,
+      "learning_rate": 8e-05,
+      "loss": 1.8198,
+      "step": 811
+    },
+    {
+      "epoch": 0.19730287935852267,
+      "grad_norm": 0.4388635456562042,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 812
+    },
+    {
+      "epoch": 0.1975458632000972,
+      "grad_norm": 0.425304114818573,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 813
+    },
+    {
+      "epoch": 0.19778884704167174,
+      "grad_norm": 0.4264742136001587,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 814
+    },
+    {
+      "epoch": 0.19803183088324627,
+      "grad_norm": 0.3999445140361786,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 815
+    },
+    {
+      "epoch": 0.1982748147248208,
+      "grad_norm": 0.39516952633857727,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 816
+    },
+    {
+      "epoch": 0.19851779856639534,
+      "grad_norm": 0.41314324736595154,
+      "learning_rate": 8e-05,
+      "loss": 1.6587,
+      "step": 817
+    },
+    {
+      "epoch": 0.19876078240796988,
+      "grad_norm": 0.4036077558994293,
+      "learning_rate": 8e-05,
+      "loss": 1.6872,
+      "step": 818
+    },
+    {
+      "epoch": 0.1990037662495444,
+      "grad_norm": 0.41270309686660767,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 819
+    },
+    {
+      "epoch": 0.19924675009111895,
+      "grad_norm": 0.42153701186180115,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 820
+    },
+    {
+      "epoch": 0.19948973393269348,
+      "grad_norm": 0.40164998173713684,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 821
+    },
+    {
+      "epoch": 0.19973271777426801,
+      "grad_norm": 0.42216864228248596,
+      "learning_rate": 8e-05,
+      "loss": 1.7704,
+      "step": 822
+    },
+    {
+      "epoch": 0.19997570161584255,
+      "grad_norm": 0.4152696132659912,
+      "learning_rate": 8e-05,
+      "loss": 1.6546,
+      "step": 823
+    },
+    {
+      "epoch": 0.20021868545741708,
+      "grad_norm": 0.44916486740112305,
+      "learning_rate": 8e-05,
+      "loss": 1.7995,
+      "step": 824
+    },
+    {
+      "epoch": 0.20046166929899162,
+      "grad_norm": 0.40811580419540405,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 825
+    },
+    {
+      "epoch": 0.20070465314056615,
+      "grad_norm": 0.4276411831378937,
+      "learning_rate": 8e-05,
+      "loss": 1.7476,
+      "step": 826
+    },
+    {
+      "epoch": 0.2009476369821407,
+      "grad_norm": 0.43811461329460144,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 827
+    },
+    {
+      "epoch": 0.20119062082371522,
+      "grad_norm": 0.43729057908058167,
+      "learning_rate": 8e-05,
+      "loss": 1.7084,
+      "step": 828
+    },
+    {
+      "epoch": 0.20143360466528976,
+      "grad_norm": 0.4092738628387451,
+      "learning_rate": 8e-05,
+      "loss": 1.7276,
+      "step": 829
+    },
+    {
+      "epoch": 0.2016765885068643,
+      "grad_norm": 0.42942163348197937,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 830
+    },
+    {
+      "epoch": 0.20191957234843882,
+      "grad_norm": 0.4280286729335785,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 831
+    },
+    {
+      "epoch": 0.20216255619001336,
+      "grad_norm": 0.4240659475326538,
+      "learning_rate": 8e-05,
+      "loss": 1.811,
+      "step": 832
+    },
+    {
+      "epoch": 0.2024055400315879,
+      "grad_norm": 0.40779125690460205,
+      "learning_rate": 8e-05,
+      "loss": 1.6911,
+      "step": 833
+    },
+    {
+      "epoch": 0.20264852387316243,
+      "grad_norm": 0.4293864965438843,
+      "learning_rate": 8e-05,
+      "loss": 1.7607,
+      "step": 834
+    },
+    {
+      "epoch": 0.20289150771473696,
+      "grad_norm": 0.4241393506526947,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 835
+    },
+    {
+      "epoch": 0.2031344915563115,
+      "grad_norm": 0.3994300067424774,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 836
+    },
+    {
+      "epoch": 0.20337747539788603,
+      "grad_norm": 0.4241347908973694,
+      "learning_rate": 8e-05,
+      "loss": 1.723,
+      "step": 837
+    },
+    {
+      "epoch": 0.20362045923946057,
+      "grad_norm": 0.4182228744029999,
+      "learning_rate": 8e-05,
+      "loss": 1.8148,
+      "step": 838
+    },
+    {
+      "epoch": 0.2038634430810351,
+      "grad_norm": 0.44692403078079224,
+      "learning_rate": 8e-05,
+      "loss": 1.7862,
+      "step": 839
+    },
+    {
+      "epoch": 0.20410642692260964,
+      "grad_norm": 0.4088408052921295,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 840
+    },
+    {
+      "epoch": 0.20434941076418417,
+      "grad_norm": 0.4297780692577362,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 841
+    },
+    {
+      "epoch": 0.2045923946057587,
+      "grad_norm": 0.43446582555770874,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 842
+    },
+    {
+      "epoch": 0.20483537844733327,
+      "grad_norm": 0.4295143187046051,
+      "learning_rate": 8e-05,
+      "loss": 1.7383,
+      "step": 843
+    },
+    {
+      "epoch": 0.2050783622889078,
+      "grad_norm": 0.3991616666316986,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 844
+    },
+    {
+      "epoch": 0.20532134613048234,
+      "grad_norm": 0.4088374078273773,
+      "learning_rate": 8e-05,
+      "loss": 1.828,
+      "step": 845
+    },
+    {
+      "epoch": 0.20556432997205687,
+      "grad_norm": 0.4084298014640808,
+      "learning_rate": 8e-05,
+      "loss": 1.714,
+      "step": 846
+    },
+    {
+      "epoch": 0.2058073138136314,
+      "grad_norm": 0.4130483567714691,
+      "learning_rate": 8e-05,
+      "loss": 1.5867,
+      "step": 847
+    },
+    {
+      "epoch": 0.20605029765520594,
+      "grad_norm": 0.415761798620224,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 848
+    },
+    {
+      "epoch": 0.20629328149678047,
+      "grad_norm": 0.4126366674900055,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 849
+    },
+    {
+      "epoch": 0.206536265338355,
+      "grad_norm": 0.41798070073127747,
+      "learning_rate": 8e-05,
+      "loss": 1.8068,
+      "step": 850
+    },
+    {
+      "epoch": 0.20677924917992954,
+      "grad_norm": 0.41303834319114685,
+      "learning_rate": 8e-05,
+      "loss": 1.7239,
+      "step": 851
+    },
+    {
+      "epoch": 0.20702223302150408,
+      "grad_norm": 0.40471598505973816,
+      "learning_rate": 8e-05,
+      "loss": 1.737,
+      "step": 852
+    },
+    {
+      "epoch": 0.2072652168630786,
+      "grad_norm": 0.4138995409011841,
+      "learning_rate": 8e-05,
+      "loss": 1.6055,
+      "step": 853
+    },
+    {
+      "epoch": 0.20750820070465315,
+      "grad_norm": 0.42502766847610474,
+      "learning_rate": 8e-05,
+      "loss": 1.7967,
+      "step": 854
+    },
+    {
+      "epoch": 0.20775118454622768,
+      "grad_norm": 0.4025301933288574,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 855
+    },
+    {
+      "epoch": 0.20799416838780221,
+      "grad_norm": 0.40375009179115295,
+      "learning_rate": 8e-05,
+      "loss": 1.7121,
+      "step": 856
+    },
+    {
+      "epoch": 0.20823715222937675,
+      "grad_norm": 0.4137362241744995,
+      "learning_rate": 8e-05,
+      "loss": 1.5489,
+      "step": 857
+    },
+    {
+      "epoch": 0.20848013607095128,
+      "grad_norm": 0.4193675220012665,
+      "learning_rate": 8e-05,
+      "loss": 1.746,
+      "step": 858
+    },
+    {
+      "epoch": 0.20872311991252582,
+      "grad_norm": 0.4016789197921753,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 859
+    },
+    {
+      "epoch": 0.20896610375410035,
+      "grad_norm": 0.45497262477874756,
+      "learning_rate": 8e-05,
+      "loss": 1.6858,
+      "step": 860
+    },
+    {
+      "epoch": 0.2092090875956749,
+      "grad_norm": 0.43340960144996643,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 861
+    },
+    {
+      "epoch": 0.20945207143724942,
+      "grad_norm": 0.39834776520729065,
+      "learning_rate": 8e-05,
+      "loss": 1.6862,
+      "step": 862
+    },
+    {
+      "epoch": 0.20969505527882396,
+      "grad_norm": 0.4076816439628601,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 863
+    },
+    {
+      "epoch": 0.2099380391203985,
+      "grad_norm": 0.4182765483856201,
+      "learning_rate": 8e-05,
+      "loss": 1.7633,
+      "step": 864
+    },
+    {
+      "epoch": 0.21018102296197302,
+      "grad_norm": 0.38607677817344666,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 865
+    },
+    {
+      "epoch": 0.21042400680354756,
+      "grad_norm": 0.4406508803367615,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 866
+    },
+    {
+      "epoch": 0.2106669906451221,
+      "grad_norm": 0.4337162971496582,
+      "learning_rate": 8e-05,
+      "loss": 1.6993,
+      "step": 867
+    },
+    {
+      "epoch": 0.21090997448669663,
+      "grad_norm": 0.40655967593193054,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 868
+    },
+    {
+      "epoch": 0.21115295832827116,
+      "grad_norm": 0.4192434549331665,
+      "learning_rate": 8e-05,
+      "loss": 1.8316,
+      "step": 869
+    },
+    {
+      "epoch": 0.2113959421698457,
+      "grad_norm": 0.41924941539764404,
+      "learning_rate": 8e-05,
+      "loss": 1.9092,
+      "step": 870
+    },
+    {
+      "epoch": 0.21163892601142023,
+      "grad_norm": 0.4139121472835541,
+      "learning_rate": 8e-05,
+      "loss": 1.7908,
+      "step": 871
+    },
+    {
+      "epoch": 0.21188190985299477,
+      "grad_norm": 0.42128536105155945,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 872
+    },
+    {
+      "epoch": 0.2121248936945693,
+      "grad_norm": 0.419964075088501,
+      "learning_rate": 8e-05,
+      "loss": 1.7105,
+      "step": 873
+    },
+    {
+      "epoch": 0.21236787753614383,
+      "grad_norm": 0.44340789318084717,
+      "learning_rate": 8e-05,
+      "loss": 1.8523,
+      "step": 874
+    },
+    {
+      "epoch": 0.21261086137771837,
+      "grad_norm": 0.3913477957248688,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 875
+    },
+    {
+      "epoch": 0.21285384521929293,
+      "grad_norm": 0.41680392622947693,
+      "learning_rate": 8e-05,
+      "loss": 1.7872,
+      "step": 876
+    },
+    {
+      "epoch": 0.21309682906086747,
+      "grad_norm": 0.40231987833976746,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 877
+    },
+    {
+      "epoch": 0.213339812902442,
+      "grad_norm": 0.42542412877082825,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 878
+    },
+    {
+      "epoch": 0.21358279674401653,
+      "grad_norm": 0.40235117077827454,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 879
+    },
+    {
+      "epoch": 0.21382578058559107,
+      "grad_norm": 0.41061437129974365,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 880
+    },
+    {
+      "epoch": 0.2140687644271656,
+      "grad_norm": 0.42108458280563354,
+      "learning_rate": 8e-05,
+      "loss": 1.7367,
+      "step": 881
+    },
+    {
+      "epoch": 0.21431174826874014,
+      "grad_norm": 0.4302060008049011,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 882
+    },
+    {
+      "epoch": 0.21455473211031467,
+      "grad_norm": 0.4124979078769684,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 883
+    },
+    {
+      "epoch": 0.2147977159518892,
+      "grad_norm": 0.40241193771362305,
+      "learning_rate": 8e-05,
+      "loss": 1.6988,
+      "step": 884
+    },
+    {
+      "epoch": 0.21504069979346374,
+      "grad_norm": 0.3977680802345276,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 885
+    },
+    {
+      "epoch": 0.21528368363503828,
+      "grad_norm": 0.4362911283969879,
+      "learning_rate": 8e-05,
+      "loss": 1.8247,
+      "step": 886
+    },
+    {
+      "epoch": 0.2155266674766128,
+      "grad_norm": 0.4307892322540283,
+      "learning_rate": 8e-05,
+      "loss": 1.7296,
+      "step": 887
+    },
+    {
+      "epoch": 0.21576965131818734,
+      "grad_norm": 0.4166683852672577,
+      "learning_rate": 8e-05,
+      "loss": 1.7336,
+      "step": 888
+    },
+    {
+      "epoch": 0.21601263515976188,
+      "grad_norm": 0.4215519428253174,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 889
+    },
+    {
+      "epoch": 0.2162556190013364,
+      "grad_norm": 0.40709665417671204,
+      "learning_rate": 8e-05,
+      "loss": 1.5952,
+      "step": 890
+    },
+    {
+      "epoch": 0.21649860284291095,
+      "grad_norm": 0.42042356729507446,
+      "learning_rate": 8e-05,
+      "loss": 1.7194,
+      "step": 891
+    },
+    {
+      "epoch": 0.21674158668448548,
+      "grad_norm": 0.43915465474128723,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 892
+    },
+    {
+      "epoch": 0.21698457052606002,
+      "grad_norm": 0.40834030508995056,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 893
+    },
+    {
+      "epoch": 0.21722755436763455,
+      "grad_norm": 0.4310332238674164,
+      "learning_rate": 8e-05,
+      "loss": 1.8158,
+      "step": 894
+    },
+    {
+      "epoch": 0.21747053820920909,
+      "grad_norm": 0.39638951420783997,
+      "learning_rate": 8e-05,
+      "loss": 1.6619,
+      "step": 895
+    },
+    {
+      "epoch": 0.21771352205078362,
+      "grad_norm": 0.421747088432312,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 896
+    },
+    {
+      "epoch": 0.21795650589235815,
+      "grad_norm": 0.4247625470161438,
+      "learning_rate": 8e-05,
+      "loss": 1.8315,
+      "step": 897
+    },
+    {
+      "epoch": 0.2181994897339327,
+      "grad_norm": 0.3869495689868927,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 898
+    },
+    {
+      "epoch": 0.21844247357550722,
+      "grad_norm": 0.4097820520401001,
+      "learning_rate": 8e-05,
+      "loss": 1.6738,
+      "step": 899
+    },
+    {
+      "epoch": 0.21868545741708176,
+      "grad_norm": 0.40507251024246216,
+      "learning_rate": 8e-05,
+      "loss": 1.6671,
+      "step": 900
+    },
+    {
+      "epoch": 0.2189284412586563,
+      "grad_norm": 0.4169073700904846,
+      "learning_rate": 8e-05,
+      "loss": 1.7553,
+      "step": 901
+    },
+    {
+      "epoch": 0.21917142510023083,
+      "grad_norm": 0.42916202545166016,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 902
+    },
+    {
+      "epoch": 0.21941440894180536,
+      "grad_norm": 0.4131448268890381,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 903
+    },
+    {
+      "epoch": 0.2196573927833799,
+      "grad_norm": 0.4290483891963959,
+      "learning_rate": 8e-05,
+      "loss": 1.7547,
+      "step": 904
+    },
+    {
+      "epoch": 0.21990037662495443,
+      "grad_norm": 0.44039106369018555,
+      "learning_rate": 8e-05,
+      "loss": 1.708,
+      "step": 905
+    },
+    {
+      "epoch": 0.22014336046652896,
+      "grad_norm": 0.42198899388313293,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 906
+    },
+    {
+      "epoch": 0.2203863443081035,
+      "grad_norm": 0.4307883679866791,
+      "learning_rate": 8e-05,
+      "loss": 1.748,
+      "step": 907
+    },
+    {
+      "epoch": 0.22062932814967803,
+      "grad_norm": 0.43735718727111816,
+      "learning_rate": 8e-05,
+      "loss": 1.7796,
+      "step": 908
+    },
+    {
+      "epoch": 0.22087231199125257,
+      "grad_norm": 0.4238229990005493,
+      "learning_rate": 8e-05,
+      "loss": 1.8024,
+      "step": 909
+    },
+    {
+      "epoch": 0.22111529583282713,
+      "grad_norm": 0.44213277101516724,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 910
+    },
+    {
+      "epoch": 0.22135827967440166,
+      "grad_norm": 0.450988233089447,
+      "learning_rate": 8e-05,
+      "loss": 1.9036,
+      "step": 911
+    },
+    {
+      "epoch": 0.2216012635159762,
+      "grad_norm": 0.40055665373802185,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 912
+    },
+    {
+      "epoch": 0.22184424735755073,
+      "grad_norm": 0.4240516126155853,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 913
+    },
+    {
+      "epoch": 0.22208723119912527,
+      "grad_norm": 0.4209631383419037,
+      "learning_rate": 8e-05,
+      "loss": 1.7108,
+      "step": 914
+    },
+    {
+      "epoch": 0.2223302150406998,
+      "grad_norm": 0.40620243549346924,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 915
+    },
+    {
+      "epoch": 0.22257319888227434,
+      "grad_norm": 0.4493052661418915,
+      "learning_rate": 8e-05,
+      "loss": 1.904,
+      "step": 916
+    },
+    {
+      "epoch": 0.22281618272384887,
+      "grad_norm": 0.4233781099319458,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 917
+    },
+    {
+      "epoch": 0.2230591665654234,
+      "grad_norm": 0.4173276722431183,
+      "learning_rate": 8e-05,
+      "loss": 1.785,
+      "step": 918
+    },
+    {
+      "epoch": 0.22330215040699794,
+      "grad_norm": 0.4131208062171936,
+      "learning_rate": 8e-05,
+      "loss": 1.8299,
+      "step": 919
+    },
+    {
+      "epoch": 0.22354513424857247,
+      "grad_norm": 0.4253666400909424,
+      "learning_rate": 8e-05,
+      "loss": 1.7282,
+      "step": 920
+    },
+    {
+      "epoch": 0.223788118090147,
+      "grad_norm": 0.40283799171447754,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 921
+    },
+    {
+      "epoch": 0.22403110193172154,
+      "grad_norm": 0.40612563490867615,
+      "learning_rate": 8e-05,
+      "loss": 1.7855,
+      "step": 922
+    },
+    {
+      "epoch": 0.22427408577329608,
+      "grad_norm": 0.4286128282546997,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 923
+    },
+    {
+      "epoch": 0.2245170696148706,
+      "grad_norm": 0.44093555212020874,
+      "learning_rate": 8e-05,
+      "loss": 1.7744,
+      "step": 924
+    },
+    {
+      "epoch": 0.22476005345644515,
+      "grad_norm": 0.431942343711853,
+      "learning_rate": 8e-05,
+      "loss": 1.7882,
+      "step": 925
+    },
+    {
+      "epoch": 0.22500303729801968,
+      "grad_norm": 0.44907891750335693,
+      "learning_rate": 8e-05,
+      "loss": 1.7911,
+      "step": 926
+    },
+    {
+      "epoch": 0.22524602113959422,
+      "grad_norm": 0.4033655524253845,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 927
+    },
+    {
+      "epoch": 0.22548900498116875,
+      "grad_norm": 0.42380547523498535,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 928
+    },
+    {
+      "epoch": 0.22573198882274328,
+      "grad_norm": 0.4354460537433624,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 929
+    },
+    {
+      "epoch": 0.22597497266431782,
+      "grad_norm": 0.446147084236145,
+      "learning_rate": 8e-05,
+      "loss": 1.778,
+      "step": 930
+    },
+    {
+      "epoch": 0.22621795650589235,
+      "grad_norm": 0.41288965940475464,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 931
+    },
+    {
+      "epoch": 0.2264609403474669,
+      "grad_norm": 0.4251484274864197,
+      "learning_rate": 8e-05,
+      "loss": 1.7162,
+      "step": 932
+    },
+    {
+      "epoch": 0.22670392418904142,
+      "grad_norm": 0.39323094487190247,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 933
+    },
+    {
+      "epoch": 0.22694690803061596,
+      "grad_norm": 0.4114528298377991,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 934
+    },
+    {
+      "epoch": 0.2271898918721905,
+      "grad_norm": 0.40497374534606934,
+      "learning_rate": 8e-05,
+      "loss": 1.6789,
+      "step": 935
+    },
+    {
+      "epoch": 0.22743287571376503,
+      "grad_norm": 0.4167117178440094,
+      "learning_rate": 8e-05,
+      "loss": 1.7146,
+      "step": 936
+    },
+    {
+      "epoch": 0.22767585955533956,
+      "grad_norm": 0.45462724566459656,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 937
+    },
+    {
+      "epoch": 0.2279188433969141,
+      "grad_norm": 0.39964571595191956,
+      "learning_rate": 8e-05,
+      "loss": 1.6931,
+      "step": 938
+    },
+    {
+      "epoch": 0.22816182723848863,
+      "grad_norm": 0.4638615548610687,
+      "learning_rate": 8e-05,
+      "loss": 1.9907,
+      "step": 939
+    },
+    {
+      "epoch": 0.22840481108006316,
+      "grad_norm": 0.4084910750389099,
+      "learning_rate": 8e-05,
+      "loss": 1.7023,
+      "step": 940
+    },
+    {
+      "epoch": 0.2286477949216377,
+      "grad_norm": 0.4054889380931854,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 941
+    },
+    {
+      "epoch": 0.22889077876321223,
+      "grad_norm": 0.40866798162460327,
+      "learning_rate": 8e-05,
+      "loss": 1.6645,
+      "step": 942
+    },
+    {
+      "epoch": 0.2291337626047868,
+      "grad_norm": 0.41853010654449463,
+      "learning_rate": 8e-05,
+      "loss": 1.7759,
+      "step": 943
+    },
+    {
+      "epoch": 0.22937674644636133,
+      "grad_norm": 0.4199506938457489,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 944
+    },
+    {
+      "epoch": 0.22961973028793586,
+      "grad_norm": 0.4339092969894409,
+      "learning_rate": 8e-05,
+      "loss": 1.7896,
+      "step": 945
+    },
+    {
+      "epoch": 0.2298627141295104,
+      "grad_norm": 0.41555607318878174,
+      "learning_rate": 8e-05,
+      "loss": 1.5698,
+      "step": 946
+    },
+    {
+      "epoch": 0.23010569797108493,
+      "grad_norm": 0.4364539086818695,
+      "learning_rate": 8e-05,
+      "loss": 1.8479,
+      "step": 947
+    },
+    {
+      "epoch": 0.23034868181265947,
+      "grad_norm": 0.40731021761894226,
+      "learning_rate": 8e-05,
+      "loss": 1.6878,
+      "step": 948
+    },
+    {
+      "epoch": 0.230591665654234,
+      "grad_norm": 0.4170721471309662,
+      "learning_rate": 8e-05,
+      "loss": 1.7859,
+      "step": 949
+    },
+    {
+      "epoch": 0.23083464949580854,
+      "grad_norm": 0.39958837628364563,
+      "learning_rate": 8e-05,
+      "loss": 1.7188,
+      "step": 950
+    },
+    {
+      "epoch": 0.23107763333738307,
+      "grad_norm": 0.4102901518344879,
+      "learning_rate": 8e-05,
+      "loss": 1.7157,
+      "step": 951
+    },
+    {
+      "epoch": 0.2313206171789576,
+      "grad_norm": 0.42558231949806213,
+      "learning_rate": 8e-05,
+      "loss": 1.8472,
+      "step": 952
+    },
+    {
+      "epoch": 0.23156360102053214,
+      "grad_norm": 0.4051419496536255,
+      "learning_rate": 8e-05,
+      "loss": 1.7987,
+      "step": 953
+    },
+    {
+      "epoch": 0.23180658486210667,
+      "grad_norm": 0.39528441429138184,
+      "learning_rate": 8e-05,
+      "loss": 1.6232,
+      "step": 954
+    },
+    {
+      "epoch": 0.2320495687036812,
+      "grad_norm": 0.41888582706451416,
+      "learning_rate": 8e-05,
+      "loss": 1.7172,
+      "step": 955
+    },
+    {
+      "epoch": 0.23229255254525574,
+      "grad_norm": 0.567353367805481,
+      "learning_rate": 8e-05,
+      "loss": 1.8401,
+      "step": 956
+    },
+    {
+      "epoch": 0.23253553638683028,
+      "grad_norm": 0.43099990487098694,
+      "learning_rate": 8e-05,
+      "loss": 1.9614,
+      "step": 957
+    },
+    {
+      "epoch": 0.2327785202284048,
+      "grad_norm": 0.40936943888664246,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 958
+    },
+    {
+      "epoch": 0.23302150406997935,
+      "grad_norm": 0.42363786697387695,
+      "learning_rate": 8e-05,
+      "loss": 1.7326,
+      "step": 959
+    },
+    {
+      "epoch": 0.23326448791155388,
+      "grad_norm": 0.39604732394218445,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 960
+    },
+    {
+      "epoch": 0.23350747175312841,
+      "grad_norm": 0.3995433449745178,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 961
+    },
+    {
+      "epoch": 0.23375045559470295,
+      "grad_norm": 0.4398478865623474,
+      "learning_rate": 8e-05,
+      "loss": 1.7686,
+      "step": 962
+    },
+    {
+      "epoch": 0.23399343943627748,
+      "grad_norm": 0.4354543387889862,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 963
+    },
+    {
+      "epoch": 0.23423642327785202,
+      "grad_norm": 0.4096083641052246,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 964
+    },
+    {
+      "epoch": 0.23447940711942655,
+      "grad_norm": 0.3929365873336792,
+      "learning_rate": 8e-05,
+      "loss": 1.4504,
+      "step": 965
+    },
+    {
+      "epoch": 0.2347223909610011,
+      "grad_norm": 0.41313865780830383,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 966
+    },
+    {
+      "epoch": 0.23496537480257562,
+      "grad_norm": 0.41436493396759033,
+      "learning_rate": 8e-05,
+      "loss": 1.7809,
+      "step": 967
+    },
+    {
+      "epoch": 0.23520835864415016,
+      "grad_norm": 0.4562958776950836,
+      "learning_rate": 8e-05,
+      "loss": 2.001,
+      "step": 968
+    },
+    {
+      "epoch": 0.2354513424857247,
+      "grad_norm": 0.4054759442806244,
+      "learning_rate": 8e-05,
+      "loss": 1.6621,
+      "step": 969
+    },
+    {
+      "epoch": 0.23569432632729923,
+      "grad_norm": 0.40638554096221924,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 970
+    },
+    {
+      "epoch": 0.23593731016887376,
+      "grad_norm": 0.41580459475517273,
+      "learning_rate": 8e-05,
+      "loss": 1.7458,
+      "step": 971
+    },
+    {
+      "epoch": 0.2361802940104483,
+      "grad_norm": 0.41007718443870544,
+      "learning_rate": 8e-05,
+      "loss": 1.6175,
+      "step": 972
+    },
+    {
+      "epoch": 0.23642327785202283,
+      "grad_norm": 0.38843604922294617,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 973
+    },
+    {
+      "epoch": 0.23666626169359736,
+      "grad_norm": 0.4144339859485626,
+      "learning_rate": 8e-05,
+      "loss": 1.8184,
+      "step": 974
+    },
+    {
+      "epoch": 0.2369092455351719,
+      "grad_norm": 0.4222639799118042,
+      "learning_rate": 8e-05,
+      "loss": 1.6873,
+      "step": 975
+    },
+    {
+      "epoch": 0.23715222937674646,
+      "grad_norm": 0.428195983171463,
+      "learning_rate": 8e-05,
+      "loss": 1.6401,
+      "step": 976
+    },
+    {
+      "epoch": 0.237395213218321,
+      "grad_norm": 0.425357460975647,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 977
+    },
+    {
+      "epoch": 0.23763819705989553,
+      "grad_norm": 0.42763423919677734,
+      "learning_rate": 8e-05,
+      "loss": 1.8252,
+      "step": 978
+    },
+    {
+      "epoch": 0.23788118090147006,
+      "grad_norm": 0.42000171542167664,
+      "learning_rate": 8e-05,
+      "loss": 1.7491,
+      "step": 979
+    },
+    {
+      "epoch": 0.2381241647430446,
+      "grad_norm": 0.4197124242782593,
+      "learning_rate": 8e-05,
+      "loss": 1.7387,
+      "step": 980
+    },
+    {
+      "epoch": 0.23836714858461913,
+      "grad_norm": 0.42555204033851624,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 981
+    },
+    {
+      "epoch": 0.23861013242619367,
+      "grad_norm": 0.42190057039260864,
+      "learning_rate": 8e-05,
+      "loss": 1.6623,
+      "step": 982
+    },
+    {
+      "epoch": 0.2388531162677682,
+      "grad_norm": 0.41451454162597656,
+      "learning_rate": 8e-05,
+      "loss": 1.6618,
+      "step": 983
+    },
+    {
+      "epoch": 0.23909610010934274,
+      "grad_norm": 0.4178244471549988,
+      "learning_rate": 8e-05,
+      "loss": 1.7211,
+      "step": 984
+    },
+    {
+      "epoch": 0.23933908395091727,
+      "grad_norm": 0.42830926179885864,
+      "learning_rate": 8e-05,
+      "loss": 1.8328,
+      "step": 985
+    },
+    {
+      "epoch": 0.2395820677924918,
+      "grad_norm": 0.42744654417037964,
+      "learning_rate": 8e-05,
+      "loss": 1.7594,
+      "step": 986
+    },
+    {
+      "epoch": 0.23982505163406634,
+      "grad_norm": 0.4242519736289978,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 987
+    },
+    {
+      "epoch": 0.24006803547564087,
+      "grad_norm": 0.41713517904281616,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 988
+    },
+    {
+      "epoch": 0.2403110193172154,
+      "grad_norm": 0.40585222840309143,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 989
+    },
+    {
+      "epoch": 0.24055400315878994,
+      "grad_norm": 0.4133054316043854,
+      "learning_rate": 8e-05,
+      "loss": 1.772,
+      "step": 990
+    },
+    {
+      "epoch": 0.24079698700036448,
+      "grad_norm": 0.39868175983428955,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 991
+    },
+    {
+      "epoch": 0.241039970841939,
+      "grad_norm": 0.40420007705688477,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 992
+    },
+    {
+      "epoch": 0.24128295468351355,
+      "grad_norm": 0.41098877787590027,
+      "learning_rate": 8e-05,
+      "loss": 1.6195,
+      "step": 993
+    },
+    {
+      "epoch": 0.24152593852508808,
+      "grad_norm": 0.40686550736427307,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 994
+    },
+    {
+      "epoch": 0.24176892236666261,
+      "grad_norm": 0.43064266443252563,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 995
+    },
+    {
+      "epoch": 0.24201190620823715,
+      "grad_norm": 0.40894436836242676,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 996
+    },
+    {
+      "epoch": 0.24225489004981168,
+      "grad_norm": 0.405504047870636,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 997
+    },
+    {
+      "epoch": 0.24249787389138622,
+      "grad_norm": 0.39840230345726013,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 998
+    },
+    {
+      "epoch": 0.24274085773296075,
+      "grad_norm": 0.4055247902870178,
+      "learning_rate": 8e-05,
+      "loss": 1.657,
+      "step": 999
+    },
+    {
+      "epoch": 0.2429838415745353,
+      "grad_norm": 0.4019063711166382,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 1000
+    },
+    {
+      "epoch": 0.24322682541610982,
+      "grad_norm": 0.4296441376209259,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 1001
+    },
+    {
+      "epoch": 0.24346980925768436,
+      "grad_norm": 0.43681925535202026,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2437127930992589,
+      "grad_norm": 0.42198410630226135,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 1003
+    },
+    {
+      "epoch": 0.24395577694083342,
+      "grad_norm": 0.4400230050086975,
+      "learning_rate": 8e-05,
+      "loss": 1.8853,
+      "step": 1004
+    },
+    {
+      "epoch": 0.24419876078240796,
+      "grad_norm": 0.4218568503856659,
+      "learning_rate": 8e-05,
+      "loss": 1.7897,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2444417446239825,
+      "grad_norm": 0.4169543981552124,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1006
+    },
+    {
+      "epoch": 0.24468472846555703,
+      "grad_norm": 0.3962773084640503,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 1007
+    },
+    {
+      "epoch": 0.24492771230713156,
+      "grad_norm": 0.4171653687953949,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 1008
+    },
+    {
+      "epoch": 0.24517069614870612,
+      "grad_norm": 0.3938234746456146,
+      "learning_rate": 8e-05,
+      "loss": 1.5476,
+      "step": 1009
+    },
+    {
+      "epoch": 0.24541367999028066,
+      "grad_norm": 0.426641583442688,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2456566638318552,
+      "grad_norm": 0.44361692667007446,
+      "learning_rate": 8e-05,
+      "loss": 1.7383,
+      "step": 1011
+    },
+    {
+      "epoch": 0.24589964767342973,
+      "grad_norm": 0.43024498224258423,
+      "learning_rate": 8e-05,
+      "loss": 1.8564,
+      "step": 1012
+    },
+    {
+      "epoch": 0.24614263151500426,
+      "grad_norm": 0.39949142932891846,
+      "learning_rate": 8e-05,
+      "loss": 1.7309,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2463856153565788,
+      "grad_norm": 0.39729544520378113,
+      "learning_rate": 8e-05,
+      "loss": 1.6329,
+      "step": 1014
+    },
+    {
+      "epoch": 0.24662859919815333,
+      "grad_norm": 0.406109094619751,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 1015
+    },
+    {
+      "epoch": 0.24687158303972787,
+      "grad_norm": 0.40192610025405884,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2471145668813024,
+      "grad_norm": 0.40629956126213074,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 1017
+    },
+    {
+      "epoch": 0.24735755072287693,
+      "grad_norm": 0.4411627948284149,
+      "learning_rate": 8e-05,
+      "loss": 1.8105,
+      "step": 1018
+    },
+    {
+      "epoch": 0.24760053456445147,
+      "grad_norm": 0.4106011986732483,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 1019
+    },
+    {
+      "epoch": 0.247843518406026,
+      "grad_norm": 0.4224635660648346,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 1020
+    },
+    {
+      "epoch": 0.24808650224760054,
+      "grad_norm": 0.4095596969127655,
+      "learning_rate": 8e-05,
+      "loss": 1.7653,
+      "step": 1021
+    },
+    {
+      "epoch": 0.24832948608917507,
+      "grad_norm": 0.41897109150886536,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2485724699307496,
+      "grad_norm": 0.42478346824645996,
+      "learning_rate": 8e-05,
+      "loss": 1.7572,
+      "step": 1023
+    },
+    {
+      "epoch": 0.24881545377232414,
+      "grad_norm": 0.42339861392974854,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1024
+    },
+    {
+      "epoch": 0.24905843761389868,
+      "grad_norm": 0.3910640478134155,
+      "learning_rate": 8e-05,
+      "loss": 1.4181,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2493014214554732,
+      "grad_norm": 0.41935157775878906,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 1026
+    },
+    {
+      "epoch": 0.24954440529704774,
+      "grad_norm": 0.4243759512901306,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24978738913862228,
+      "grad_norm": 0.40742623805999756,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2500303729801968,
+      "grad_norm": 0.3955259323120117,
+      "learning_rate": 8e-05,
+      "loss": 1.6848,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2502733568217714,
+      "grad_norm": 0.38137251138687134,
+      "learning_rate": 8e-05,
+      "loss": 1.4792,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2505163406633459,
+      "grad_norm": 0.41162389516830444,
+      "learning_rate": 8e-05,
+      "loss": 1.8394,
+      "step": 1031
+    },
+    {
+      "epoch": 0.25075932450492044,
+      "grad_norm": 0.4106503129005432,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 1032
+    },
+    {
+      "epoch": 0.25100230834649495,
+      "grad_norm": 0.4271000623703003,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2512452921880695,
+      "grad_norm": 0.4363470673561096,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 1034
+    },
+    {
+      "epoch": 0.251488276029644,
+      "grad_norm": 0.40387797355651855,
+      "learning_rate": 8e-05,
+      "loss": 1.7888,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2517312598712186,
+      "grad_norm": 0.3888157904148102,
+      "learning_rate": 8e-05,
+      "loss": 1.5436,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2519742437127931,
+      "grad_norm": 0.4226617217063904,
+      "learning_rate": 8e-05,
+      "loss": 1.7919,
+      "step": 1037
+    },
+    {
+      "epoch": 0.25221722755436765,
+      "grad_norm": 0.44716712832450867,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 1038
+    },
+    {
+      "epoch": 0.25246021139594216,
+      "grad_norm": 0.42776966094970703,
+      "learning_rate": 8e-05,
+      "loss": 1.7561,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2527031952375167,
+      "grad_norm": 0.4059743583202362,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2529461790790912,
+      "grad_norm": 0.4273454248905182,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2531891629206658,
+      "grad_norm": 0.41879257559776306,
+      "learning_rate": 8e-05,
+      "loss": 1.786,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2534321467622403,
+      "grad_norm": 0.42721250653266907,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 1043
+    },
+    {
+      "epoch": 0.25367513060381486,
+      "grad_norm": 0.42117446660995483,
+      "learning_rate": 8e-05,
+      "loss": 1.6064,
+      "step": 1044
+    },
+    {
+      "epoch": 0.25391811444538936,
+      "grad_norm": 0.4064764678478241,
+      "learning_rate": 8e-05,
+      "loss": 1.8178,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2541610982869639,
+      "grad_norm": 0.3869502544403076,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 1046
+    },
+    {
+      "epoch": 0.25440408212853843,
+      "grad_norm": 0.41754618287086487,
+      "learning_rate": 8e-05,
+      "loss": 1.8209,
+      "step": 1047
+    },
+    {
+      "epoch": 0.254647065970113,
+      "grad_norm": 0.41012975573539734,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2548900498116875,
+      "grad_norm": 0.39602041244506836,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 1049
+    },
+    {
+      "epoch": 0.25513303365326206,
+      "grad_norm": 0.4394192099571228,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1050
+    },
+    {
+      "epoch": 0.25537601749483657,
+      "grad_norm": 0.4520413875579834,
+      "learning_rate": 8e-05,
+      "loss": 1.8656,
+      "step": 1051
+    },
+    {
+      "epoch": 0.25561900133641113,
+      "grad_norm": 0.4225905239582062,
+      "learning_rate": 8e-05,
+      "loss": 1.8081,
+      "step": 1052
+    },
+    {
+      "epoch": 0.25586198517798564,
+      "grad_norm": 0.4238746464252472,
+      "learning_rate": 8e-05,
+      "loss": 1.7315,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2561049690195602,
+      "grad_norm": 0.40032291412353516,
+      "learning_rate": 8e-05,
+      "loss": 1.6789,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2563479528611347,
+      "grad_norm": 0.3927189111709595,
+      "learning_rate": 8e-05,
+      "loss": 1.6654,
+      "step": 1055
+    },
+    {
+      "epoch": 0.25659093670270927,
+      "grad_norm": 0.4187318682670593,
+      "learning_rate": 8e-05,
+      "loss": 1.8698,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2568339205442838,
+      "grad_norm": 0.43324345350265503,
+      "learning_rate": 8e-05,
+      "loss": 1.608,
+      "step": 1057
+    },
+    {
+      "epoch": 0.25707690438585834,
+      "grad_norm": 0.40138164162635803,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2573198882274329,
+      "grad_norm": 0.43308112025260925,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2575628720690074,
+      "grad_norm": 0.43418794870376587,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 1060
+    },
+    {
+      "epoch": 0.25780585591058197,
+      "grad_norm": 0.4303111732006073,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2580488397521565,
+      "grad_norm": 0.4321209192276001,
+      "learning_rate": 8e-05,
+      "loss": 1.7081,
+      "step": 1062
+    },
+    {
+      "epoch": 0.25829182359373104,
+      "grad_norm": 0.46187859773635864,
+      "learning_rate": 8e-05,
+      "loss": 1.978,
+      "step": 1063
+    },
+    {
+      "epoch": 0.25853480743530555,
+      "grad_norm": 0.41593241691589355,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2587777912768801,
+      "grad_norm": 0.4512927830219269,
+      "learning_rate": 8e-05,
+      "loss": 1.8112,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2590207751184546,
+      "grad_norm": 0.41697239875793457,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2592637589600292,
+      "grad_norm": 0.4299595355987549,
+      "learning_rate": 8e-05,
+      "loss": 1.7994,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2595067428016037,
+      "grad_norm": 0.43223658204078674,
+      "learning_rate": 8e-05,
+      "loss": 1.8077,
+      "step": 1068
+    },
+    {
+      "epoch": 0.25974972664317825,
+      "grad_norm": 0.4131694436073303,
+      "learning_rate": 8e-05,
+      "loss": 1.8751,
+      "step": 1069
+    },
+    {
+      "epoch": 0.25999271048475275,
+      "grad_norm": 0.4038296341896057,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2602356943263273,
+      "grad_norm": 0.41273900866508484,
+      "learning_rate": 8e-05,
+      "loss": 1.6093,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2604786781679018,
+      "grad_norm": 0.4254191517829895,
+      "learning_rate": 8e-05,
+      "loss": 1.7514,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2607216620094764,
+      "grad_norm": 0.4102127254009247,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2609646458510509,
+      "grad_norm": 0.42967456579208374,
+      "learning_rate": 8e-05,
+      "loss": 1.8004,
+      "step": 1074
+    },
+    {
+      "epoch": 0.26120762969262545,
+      "grad_norm": 0.42619723081588745,
+      "learning_rate": 8e-05,
+      "loss": 1.7346,
+      "step": 1075
+    },
+    {
+      "epoch": 0.26145061353419996,
+      "grad_norm": 0.42867016792297363,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2616935973757745,
+      "grad_norm": 0.435442715883255,
+      "learning_rate": 8e-05,
+      "loss": 1.8384,
+      "step": 1077
+    },
+    {
+      "epoch": 0.26193658121734903,
+      "grad_norm": 0.41969239711761475,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2621795650589236,
+      "grad_norm": 0.4229139983654022,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2624225489004981,
+      "grad_norm": 0.42010724544525146,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 1080
+    },
+    {
+      "epoch": 0.26266553274207266,
+      "grad_norm": 0.4319921135902405,
+      "learning_rate": 8e-05,
+      "loss": 1.7998,
+      "step": 1081
+    },
+    {
+      "epoch": 0.26290851658364717,
+      "grad_norm": 0.4249647259712219,
+      "learning_rate": 8e-05,
+      "loss": 1.771,
+      "step": 1082
+    },
+    {
+      "epoch": 0.26315150042522173,
+      "grad_norm": 0.4056098759174347,
+      "learning_rate": 8e-05,
+      "loss": 1.4846,
+      "step": 1083
+    },
+    {
+      "epoch": 0.26339448426679624,
+      "grad_norm": 0.4061814546585083,
+      "learning_rate": 8e-05,
+      "loss": 1.401,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2636374681083708,
+      "grad_norm": 0.4380955696105957,
+      "learning_rate": 8e-05,
+      "loss": 1.8282,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2638804519499453,
+      "grad_norm": 0.4158300757408142,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 1086
+    },
+    {
+      "epoch": 0.26412343579151987,
+      "grad_norm": 0.39589807391166687,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2643664196330944,
+      "grad_norm": 0.39479830861091614,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 1088
+    },
+    {
+      "epoch": 0.26460940347466894,
+      "grad_norm": 0.4172505736351013,
+      "learning_rate": 8e-05,
+      "loss": 1.783,
+      "step": 1089
+    },
+    {
+      "epoch": 0.26485238731624344,
+      "grad_norm": 0.4129602313041687,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 1090
+    },
+    {
+      "epoch": 0.265095371157818,
+      "grad_norm": 0.40379106998443604,
+      "learning_rate": 8e-05,
+      "loss": 1.6992,
+      "step": 1091
+    },
+    {
+      "epoch": 0.26533835499939257,
+      "grad_norm": 0.4196067154407501,
+      "learning_rate": 8e-05,
+      "loss": 1.7052,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2655813388409671,
+      "grad_norm": 0.4030078947544098,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 1093
+    },
+    {
+      "epoch": 0.26582432268254164,
+      "grad_norm": 0.4095304608345032,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 1094
+    },
+    {
+      "epoch": 0.26606730652411614,
+      "grad_norm": 0.4298233091831207,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2663102903656907,
+      "grad_norm": 0.41325706243515015,
+      "learning_rate": 8e-05,
+      "loss": 1.7149,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2665532742072652,
+      "grad_norm": 0.41717031598091125,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2667962580488398,
+      "grad_norm": 0.4257058799266815,
+      "learning_rate": 8e-05,
+      "loss": 1.7694,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2670392418904143,
+      "grad_norm": 0.4161388874053955,
+      "learning_rate": 8e-05,
+      "loss": 1.6243,
+      "step": 1099
+    },
+    {
+      "epoch": 0.26728222573198884,
+      "grad_norm": 0.43736934661865234,
+      "learning_rate": 8e-05,
+      "loss": 1.8092,
+      "step": 1100
+    },
+    {
+      "epoch": 0.26752520957356335,
+      "grad_norm": 0.4007851183414459,
+      "learning_rate": 8e-05,
+      "loss": 1.5047,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2677681934151379,
+      "grad_norm": 0.4082740843296051,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2680111772567124,
+      "grad_norm": 0.42005470395088196,
+      "learning_rate": 8e-05,
+      "loss": 1.7195,
+      "step": 1103
+    },
+    {
+      "epoch": 0.268254161098287,
+      "grad_norm": 0.4103584885597229,
+      "learning_rate": 8e-05,
+      "loss": 1.4854,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2684971449398615,
+      "grad_norm": 0.4048874080181122,
+      "learning_rate": 8e-05,
+      "loss": 1.8195,
+      "step": 1105
+    },
+    {
+      "epoch": 0.26874012878143605,
+      "grad_norm": 0.4084217846393585,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1106
+    },
+    {
+      "epoch": 0.26898311262301056,
+      "grad_norm": 0.4224699139595032,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2692260964645851,
+      "grad_norm": 0.43159735202789307,
+      "learning_rate": 8e-05,
+      "loss": 1.7818,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2694690803061596,
+      "grad_norm": 0.40543919801712036,
+      "learning_rate": 8e-05,
+      "loss": 1.5697,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2697120641477342,
+      "grad_norm": 0.4308815002441406,
+      "learning_rate": 8e-05,
+      "loss": 1.7352,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2699550479893087,
+      "grad_norm": 0.44923198223114014,
+      "learning_rate": 8e-05,
+      "loss": 1.7749,
+      "step": 1111
+    },
+    {
+      "epoch": 0.27019803183088326,
+      "grad_norm": 0.42572107911109924,
+      "learning_rate": 8e-05,
+      "loss": 1.8659,
+      "step": 1112
+    },
+    {
+      "epoch": 0.27044101567245776,
+      "grad_norm": 0.41339996457099915,
+      "learning_rate": 8e-05,
+      "loss": 1.7747,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2706839995140323,
+      "grad_norm": 0.4206458032131195,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 1114
+    },
+    {
+      "epoch": 0.27092698335560683,
+      "grad_norm": 0.429023414850235,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2711699671971814,
+      "grad_norm": 0.40702301263809204,
+      "learning_rate": 8e-05,
+      "loss": 1.79,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2714129510387559,
+      "grad_norm": 0.4375455379486084,
+      "learning_rate": 8e-05,
+      "loss": 1.8345,
+      "step": 1117
+    },
+    {
+      "epoch": 0.27165593488033046,
+      "grad_norm": 0.3962376117706299,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 1118
+    },
+    {
+      "epoch": 0.27189891872190497,
+      "grad_norm": 0.40322914719581604,
+      "learning_rate": 8e-05,
+      "loss": 1.6283,
+      "step": 1119
+    },
+    {
+      "epoch": 0.27214190256347953,
+      "grad_norm": 0.43046557903289795,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 1120
+    },
+    {
+      "epoch": 0.27238488640505404,
+      "grad_norm": 0.4123181700706482,
+      "learning_rate": 8e-05,
+      "loss": 1.7929,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2726278702466286,
+      "grad_norm": 0.41423794627189636,
+      "learning_rate": 8e-05,
+      "loss": 1.7222,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2728708540882031,
+      "grad_norm": 0.40966901183128357,
+      "learning_rate": 8e-05,
+      "loss": 1.6769,
+      "step": 1123
+    },
+    {
+      "epoch": 0.27311383792977767,
+      "grad_norm": 0.43045124411582947,
+      "learning_rate": 8e-05,
+      "loss": 1.6193,
+      "step": 1124
+    },
+    {
+      "epoch": 0.27335682177135223,
+      "grad_norm": 0.41753265261650085,
+      "learning_rate": 8e-05,
+      "loss": 1.6965,
+      "step": 1125
+    },
+    {
+      "epoch": 0.27359980561292674,
+      "grad_norm": 0.4054257273674011,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2738427894545013,
+      "grad_norm": 0.4242555499076843,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2740857732960758,
+      "grad_norm": 0.4189755618572235,
+      "learning_rate": 8e-05,
+      "loss": 1.6478,
+      "step": 1128
+    },
+    {
+      "epoch": 0.27432875713765037,
+      "grad_norm": 0.4091269373893738,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2745717409792249,
+      "grad_norm": 0.4247717559337616,
+      "learning_rate": 8e-05,
+      "loss": 1.7505,
+      "step": 1130
+    },
+    {
+      "epoch": 0.27481472482079944,
+      "grad_norm": 0.4193531572818756,
+      "learning_rate": 8e-05,
+      "loss": 1.8572,
+      "step": 1131
+    },
+    {
+      "epoch": 0.27505770866237395,
+      "grad_norm": 0.42771103978157043,
+      "learning_rate": 8e-05,
+      "loss": 1.8101,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2753006925039485,
+      "grad_norm": 0.4443416893482208,
+      "learning_rate": 8e-05,
+      "loss": 1.9342,
+      "step": 1133
+    },
+    {
+      "epoch": 0.275543676345523,
+      "grad_norm": 0.452569842338562,
+      "learning_rate": 8e-05,
+      "loss": 1.9045,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2757866601870976,
+      "grad_norm": 0.41667863726615906,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2760296440286721,
+      "grad_norm": 0.4566730856895447,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 1136
+    },
+    {
+      "epoch": 0.27627262787024665,
+      "grad_norm": 0.42623674869537354,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 1137
+    },
+    {
+      "epoch": 0.27651561171182115,
+      "grad_norm": 0.4352518618106842,
+      "learning_rate": 8e-05,
+      "loss": 1.811,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2767585955533957,
+      "grad_norm": 0.4077364206314087,
+      "learning_rate": 8e-05,
+      "loss": 1.7585,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2770015793949702,
+      "grad_norm": 0.427761435508728,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2772445632365448,
+      "grad_norm": 0.41399678587913513,
+      "learning_rate": 8e-05,
+      "loss": 1.7356,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2774875470781193,
+      "grad_norm": 0.3881222605705261,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 1142
+    },
+    {
+      "epoch": 0.27773053091969385,
+      "grad_norm": 0.41863733530044556,
+      "learning_rate": 8e-05,
+      "loss": 1.7983,
+      "step": 1143
+    },
+    {
+      "epoch": 0.27797351476126836,
+      "grad_norm": 0.41757330298423767,
+      "learning_rate": 8e-05,
+      "loss": 1.7363,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2782164986028429,
+      "grad_norm": 0.4343527555465698,
+      "learning_rate": 8e-05,
+      "loss": 1.7453,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2784594824444174,
+      "grad_norm": 0.411081463098526,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 1146
+    },
+    {
+      "epoch": 0.278702466285992,
+      "grad_norm": 0.42007976770401,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2789454501275665,
+      "grad_norm": 0.41134604811668396,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 1148
+    },
+    {
+      "epoch": 0.27918843396914106,
+      "grad_norm": 0.41881364583969116,
+      "learning_rate": 8e-05,
+      "loss": 1.6334,
+      "step": 1149
+    },
+    {
+      "epoch": 0.27943141781071557,
+      "grad_norm": 0.4174276888370514,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2796744016522901,
+      "grad_norm": 0.4109209477901459,
+      "learning_rate": 8e-05,
+      "loss": 1.6622,
+      "step": 1151
+    },
+    {
+      "epoch": 0.27991738549386463,
+      "grad_norm": 0.4209733307361603,
+      "learning_rate": 8e-05,
+      "loss": 1.6243,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2801603693354392,
+      "grad_norm": 0.43282854557037354,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2804033531770137,
+      "grad_norm": 0.4117317199707031,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 1154
+    },
+    {
+      "epoch": 0.28064633701858827,
+      "grad_norm": 0.44751831889152527,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 1155
+    },
+    {
+      "epoch": 0.28088932086016277,
+      "grad_norm": 0.401326447725296,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 1156
+    },
+    {
+      "epoch": 0.28113230470173733,
+      "grad_norm": 0.4252309203147888,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2813752885433119,
+      "grad_norm": 0.42324867844581604,
+      "learning_rate": 8e-05,
+      "loss": 1.7284,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2816182723848864,
+      "grad_norm": 0.42442786693573,
+      "learning_rate": 8e-05,
+      "loss": 1.6773,
+      "step": 1159
+    },
+    {
+      "epoch": 0.28186125622646097,
+      "grad_norm": 0.39588966965675354,
+      "learning_rate": 8e-05,
+      "loss": 1.7574,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2821042400680355,
+      "grad_norm": 0.42665570974349976,
+      "learning_rate": 8e-05,
+      "loss": 1.779,
+      "step": 1161
+    },
+    {
+      "epoch": 0.28234722390961003,
+      "grad_norm": 0.4189774692058563,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 1162
+    },
+    {
+      "epoch": 0.28259020775118454,
+      "grad_norm": 0.41338545083999634,
+      "learning_rate": 8e-05,
+      "loss": 1.701,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2828331915927591,
+      "grad_norm": 0.4207904040813446,
+      "learning_rate": 8e-05,
+      "loss": 1.5652,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2830761754343336,
+      "grad_norm": 0.4322403371334076,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2833191592759082,
+      "grad_norm": 0.43739262223243713,
+      "learning_rate": 8e-05,
+      "loss": 1.8341,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2835621431174827,
+      "grad_norm": 0.4155855178833008,
+      "learning_rate": 8e-05,
+      "loss": 1.7949,
+      "step": 1167
+    },
+    {
+      "epoch": 0.28380512695905724,
+      "grad_norm": 0.42609381675720215,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 1168
+    },
+    {
+      "epoch": 0.28404811080063175,
+      "grad_norm": 0.4329025149345398,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2842910946422063,
+      "grad_norm": 0.4329528510570526,
+      "learning_rate": 8e-05,
+      "loss": 1.9396,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2845340784837808,
+      "grad_norm": 0.44219064712524414,
+      "learning_rate": 8e-05,
+      "loss": 1.6349,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2847770623253554,
+      "grad_norm": 0.44023966789245605,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2850200461669299,
+      "grad_norm": 0.4343285858631134,
+      "learning_rate": 8e-05,
+      "loss": 1.9336,
+      "step": 1173
+    },
+    {
+      "epoch": 0.28526303000850445,
+      "grad_norm": 0.41535404324531555,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 1174
+    },
+    {
+      "epoch": 0.28550601385007895,
+      "grad_norm": 0.4356030523777008,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2857489976916535,
+      "grad_norm": 0.4417020082473755,
+      "learning_rate": 8e-05,
+      "loss": 1.7016,
+      "step": 1176
+    },
+    {
+      "epoch": 0.285991981533228,
+      "grad_norm": 0.39354732632637024,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2862349653748026,
+      "grad_norm": 0.4548605680465698,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2864779492163771,
+      "grad_norm": 0.4130997061729431,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 1179
+    },
+    {
+      "epoch": 0.28672093305795165,
+      "grad_norm": 0.4564334750175476,
+      "learning_rate": 8e-05,
+      "loss": 1.8025,
+      "step": 1180
+    },
+    {
+      "epoch": 0.28696391689952616,
+      "grad_norm": 0.4132091701030731,
+      "learning_rate": 8e-05,
+      "loss": 1.6924,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2872069007411007,
+      "grad_norm": 0.43107539415359497,
+      "learning_rate": 8e-05,
+      "loss": 1.8584,
+      "step": 1182
+    },
+    {
+      "epoch": 0.28744988458267523,
+      "grad_norm": 0.4062747061252594,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2876928684242498,
+      "grad_norm": 0.4347899556159973,
+      "learning_rate": 8e-05,
+      "loss": 1.8802,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2879358522658243,
+      "grad_norm": 0.45099252462387085,
+      "learning_rate": 8e-05,
+      "loss": 1.6991,
+      "step": 1185
+    },
+    {
+      "epoch": 0.28817883610739886,
+      "grad_norm": 0.41353484988212585,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 1186
+    },
+    {
+      "epoch": 0.28842181994897337,
+      "grad_norm": 0.4440639317035675,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 1187
+    },
+    {
+      "epoch": 0.28866480379054793,
+      "grad_norm": 0.4112803041934967,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 1188
+    },
+    {
+      "epoch": 0.28890778763212244,
+      "grad_norm": 0.38698825240135193,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1189
+    },
+    {
+      "epoch": 0.289150771473697,
+      "grad_norm": 0.40388596057891846,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 1190
+    },
+    {
+      "epoch": 0.28939375531527156,
+      "grad_norm": 0.39465680718421936,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 1191
+    },
+    {
+      "epoch": 0.28963673915684607,
+      "grad_norm": 0.4267415702342987,
+      "learning_rate": 8e-05,
+      "loss": 1.7361,
+      "step": 1192
+    },
+    {
+      "epoch": 0.28987972299842063,
+      "grad_norm": 0.42168205976486206,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 1193
+    },
+    {
+      "epoch": 0.29012270683999514,
+      "grad_norm": 0.41416439414024353,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2903656906815697,
+      "grad_norm": 0.4276784658432007,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2906086745231442,
+      "grad_norm": 0.4125884175300598,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1196
+    },
+    {
+      "epoch": 0.29085165836471877,
+      "grad_norm": 0.4149490296840668,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2910946422062933,
+      "grad_norm": 0.42758074402809143,
+      "learning_rate": 8e-05,
+      "loss": 1.7272,
+      "step": 1198
+    },
+    {
+      "epoch": 0.29133762604786784,
+      "grad_norm": 0.4167897701263428,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 1199
+    },
+    {
+      "epoch": 0.29158060988944234,
+      "grad_norm": 0.4239620864391327,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2918235937310169,
+      "grad_norm": 0.4121783673763275,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2920665775725914,
+      "grad_norm": 0.43166282773017883,
+      "learning_rate": 8e-05,
+      "loss": 1.6699,
+      "step": 1202
+    },
+    {
+      "epoch": 0.292309561414166,
+      "grad_norm": 0.442334920167923,
+      "learning_rate": 8e-05,
+      "loss": 1.8271,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2925525452557405,
+      "grad_norm": 0.4301314055919647,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 1204
+    },
+    {
+      "epoch": 0.29279552909731504,
+      "grad_norm": 0.45724913477897644,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 1205
+    },
+    {
+      "epoch": 0.29303851293888955,
+      "grad_norm": 0.4075496792793274,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2932814967804641,
+      "grad_norm": 0.42279496788978577,
+      "learning_rate": 8e-05,
+      "loss": 1.8865,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2935244806220386,
+      "grad_norm": 0.4176923930644989,
+      "learning_rate": 8e-05,
+      "loss": 1.7014,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2937674644636132,
+      "grad_norm": 0.4202452301979065,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2940104483051877,
+      "grad_norm": 0.41166600584983826,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 1210
+    },
+    {
+      "epoch": 0.29425343214676225,
+      "grad_norm": 0.4162682592868805,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 1211
+    },
+    {
+      "epoch": 0.29449641598833676,
+      "grad_norm": 0.42768990993499756,
+      "learning_rate": 8e-05,
+      "loss": 1.7041,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2947393998299113,
+      "grad_norm": 0.4126628637313843,
+      "learning_rate": 8e-05,
+      "loss": 1.8088,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2949823836714858,
+      "grad_norm": 0.41666892170906067,
+      "learning_rate": 8e-05,
+      "loss": 1.7258,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2952253675130604,
+      "grad_norm": 0.39374640583992004,
+      "learning_rate": 8e-05,
+      "loss": 1.5282,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2954683513546349,
+      "grad_norm": 0.40578532218933105,
+      "learning_rate": 8e-05,
+      "loss": 1.7309,
+      "step": 1216
+    },
+    {
+      "epoch": 0.29571133519620946,
+      "grad_norm": 0.4230765104293823,
+      "learning_rate": 8e-05,
+      "loss": 1.7824,
+      "step": 1217
+    },
+    {
+      "epoch": 0.29595431903778396,
+      "grad_norm": 0.4090322256088257,
+      "learning_rate": 8e-05,
+      "loss": 1.4695,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2961973028793585,
+      "grad_norm": 0.39290255308151245,
+      "learning_rate": 8e-05,
+      "loss": 1.6753,
+      "step": 1219
+    },
+    {
+      "epoch": 0.29644028672093303,
+      "grad_norm": 0.4163476228713989,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2966832705625076,
+      "grad_norm": 0.4184444844722748,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2969262544040821,
+      "grad_norm": 0.46152612566947937,
+      "learning_rate": 8e-05,
+      "loss": 1.8103,
+      "step": 1222
+    },
+    {
+      "epoch": 0.29716923824565666,
+      "grad_norm": 0.4153369069099426,
+      "learning_rate": 8e-05,
+      "loss": 1.6935,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2974122220872312,
+      "grad_norm": 0.5140358209609985,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 1224
+    },
+    {
+      "epoch": 0.29765520592880573,
+      "grad_norm": 0.473117470741272,
+      "learning_rate": 8e-05,
+      "loss": 1.7152,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2978981897703803,
+      "grad_norm": 0.4486581087112427,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2981411736119548,
+      "grad_norm": 0.4636811912059784,
+      "learning_rate": 8e-05,
+      "loss": 1.9135,
+      "step": 1227
+    },
+    {
+      "epoch": 0.29838415745352936,
+      "grad_norm": 0.4324636459350586,
+      "learning_rate": 8e-05,
+      "loss": 1.5375,
+      "step": 1228
+    },
+    {
+      "epoch": 0.29862714129510387,
+      "grad_norm": 0.41699549555778503,
+      "learning_rate": 8e-05,
+      "loss": 1.6287,
+      "step": 1229
+    },
+    {
+      "epoch": 0.29887012513667843,
+      "grad_norm": 0.42031776905059814,
+      "learning_rate": 8e-05,
+      "loss": 1.8032,
+      "step": 1230
+    },
+    {
+      "epoch": 0.29911310897825294,
+      "grad_norm": 0.43739062547683716,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2993560928198275,
+      "grad_norm": 0.4243648052215576,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 1232
+    },
+    {
+      "epoch": 0.299599076661402,
+      "grad_norm": 0.40939441323280334,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 1233
+    },
+    {
+      "epoch": 0.29984206050297657,
+      "grad_norm": 0.4723411500453949,
+      "learning_rate": 8e-05,
+      "loss": 1.8328,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3000850443445511,
+      "grad_norm": 0.4084712564945221,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 1235
+    },
+    {
+      "epoch": 0.30032802818612564,
+      "grad_norm": 0.44527482986450195,
+      "learning_rate": 8e-05,
+      "loss": 1.7545,
+      "step": 1236
+    },
+    {
+      "epoch": 0.30057101202770015,
+      "grad_norm": 0.40758949518203735,
+      "learning_rate": 8e-05,
+      "loss": 1.5875,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3008139958692747,
+      "grad_norm": 0.4355928599834442,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3010569797108492,
+      "grad_norm": 0.4212299585342407,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3012999635524238,
+      "grad_norm": 0.40415745973587036,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3015429473939983,
+      "grad_norm": 0.43318644165992737,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1241
+    },
+    {
+      "epoch": 0.30178593123557285,
+      "grad_norm": 0.3988751173019409,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 1242
+    },
+    {
+      "epoch": 0.30202891507714735,
+      "grad_norm": 0.46518203616142273,
+      "learning_rate": 8e-05,
+      "loss": 1.7732,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3022718989187219,
+      "grad_norm": 0.4167146384716034,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3025148827602964,
+      "grad_norm": 0.4274185299873352,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 1245
+    },
+    {
+      "epoch": 0.302757866601871,
+      "grad_norm": 0.41032370924949646,
+      "learning_rate": 8e-05,
+      "loss": 1.7365,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3030008504434455,
+      "grad_norm": 0.4476938843727112,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 1247
+    },
+    {
+      "epoch": 0.30324383428502005,
+      "grad_norm": 0.3983435034751892,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 1248
+    },
+    {
+      "epoch": 0.30348681812659456,
+      "grad_norm": 0.4122038185596466,
+      "learning_rate": 8e-05,
+      "loss": 1.7713,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3037298019681691,
+      "grad_norm": 0.4572095572948456,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 1250
+    },
+    {
+      "epoch": 0.30397278580974363,
+      "grad_norm": 0.4236910343170166,
+      "learning_rate": 8e-05,
+      "loss": 1.7273,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3042157696513182,
+      "grad_norm": 0.42044517397880554,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3044587534928927,
+      "grad_norm": 0.4279285967350006,
+      "learning_rate": 8e-05,
+      "loss": 1.7749,
+      "step": 1253
+    },
+    {
+      "epoch": 0.30470173733446726,
+      "grad_norm": 0.43583187460899353,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 1254
+    },
+    {
+      "epoch": 0.30494472117604177,
+      "grad_norm": 0.4059058427810669,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 1255
+    },
+    {
+      "epoch": 0.30518770501761633,
+      "grad_norm": 0.43845224380493164,
+      "learning_rate": 8e-05,
+      "loss": 1.7272,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3054306888591909,
+      "grad_norm": 0.4243011772632599,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3056736727007654,
+      "grad_norm": 0.43698588013648987,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 1258
+    },
+    {
+      "epoch": 0.30591665654233996,
+      "grad_norm": 0.4046631157398224,
+      "learning_rate": 8e-05,
+      "loss": 1.7473,
+      "step": 1259
+    },
+    {
+      "epoch": 0.30615964038391447,
+      "grad_norm": 0.4501211941242218,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 1260
+    },
+    {
+      "epoch": 0.30640262422548903,
+      "grad_norm": 0.40824440121650696,
+      "learning_rate": 8e-05,
+      "loss": 1.7088,
+      "step": 1261
+    },
+    {
+      "epoch": 0.30664560806706354,
+      "grad_norm": 0.3860519230365753,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3068885919086381,
+      "grad_norm": 0.43458759784698486,
+      "learning_rate": 8e-05,
+      "loss": 1.7604,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3071315757502126,
+      "grad_norm": 0.4188707768917084,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 1264
+    },
+    {
+      "epoch": 0.30737455959178717,
+      "grad_norm": 0.42965972423553467,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3076175434333617,
+      "grad_norm": 0.45659002661705017,
+      "learning_rate": 8e-05,
+      "loss": 1.8528,
+      "step": 1266
+    },
+    {
+      "epoch": 0.30786052727493624,
+      "grad_norm": 0.4213427007198334,
+      "learning_rate": 8e-05,
+      "loss": 1.7526,
+      "step": 1267
+    },
+    {
+      "epoch": 0.30810351111651074,
+      "grad_norm": 0.44389426708221436,
+      "learning_rate": 8e-05,
+      "loss": 1.7928,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3083464949580853,
+      "grad_norm": 0.3961242437362671,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3085894787996598,
+      "grad_norm": 0.389335960149765,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3088324626412344,
+      "grad_norm": 0.42223578691482544,
+      "learning_rate": 8e-05,
+      "loss": 1.7528,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3090754464828089,
+      "grad_norm": 0.40308067202568054,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1272
+    },
+    {
+      "epoch": 0.30931843032438344,
+      "grad_norm": 0.41871175169944763,
+      "learning_rate": 8e-05,
+      "loss": 1.8313,
+      "step": 1273
+    },
+    {
+      "epoch": 0.30956141416595795,
+      "grad_norm": 0.404159277677536,
+      "learning_rate": 8e-05,
+      "loss": 1.5875,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3098043980075325,
+      "grad_norm": 0.4272163212299347,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 1275
+    },
+    {
+      "epoch": 0.310047381849107,
+      "grad_norm": 0.4222562313079834,
+      "learning_rate": 8e-05,
+      "loss": 1.6618,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3102903656906816,
+      "grad_norm": 0.42326620221138,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3105333495322561,
+      "grad_norm": 0.4882233142852783,
+      "learning_rate": 8e-05,
+      "loss": 1.7169,
+      "step": 1278
+    },
+    {
+      "epoch": 0.31077633337383065,
+      "grad_norm": 0.4253446161746979,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 1279
+    },
+    {
+      "epoch": 0.31101931721540516,
+      "grad_norm": 0.4078899323940277,
+      "learning_rate": 8e-05,
+      "loss": 1.7789,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3112623010569797,
+      "grad_norm": 0.4250544011592865,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3115052848985542,
+      "grad_norm": 0.42535340785980225,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3117482687401288,
+      "grad_norm": 0.42241019010543823,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3119912525817033,
+      "grad_norm": 0.4194417893886566,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 1284
+    },
+    {
+      "epoch": 0.31223423642327786,
+      "grad_norm": 0.42229947447776794,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 1285
+    },
+    {
+      "epoch": 0.31247722026485236,
+      "grad_norm": 0.4440034031867981,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3127202041064269,
+      "grad_norm": 0.4309479296207428,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 1287
+    },
+    {
+      "epoch": 0.31296318794800143,
+      "grad_norm": 0.4373340606689453,
+      "learning_rate": 8e-05,
+      "loss": 1.7831,
+      "step": 1288
+    },
+    {
+      "epoch": 0.313206171789576,
+      "grad_norm": 0.40313777327537537,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 1289
+    },
+    {
+      "epoch": 0.31344915563115056,
+      "grad_norm": 0.4201229214668274,
+      "learning_rate": 8e-05,
+      "loss": 1.7659,
+      "step": 1290
+    },
+    {
+      "epoch": 0.31369213947272506,
+      "grad_norm": 0.43222776055336,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3139351233142996,
+      "grad_norm": 0.40938252210617065,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 1292
+    },
+    {
+      "epoch": 0.31417810715587413,
+      "grad_norm": 0.43402594327926636,
+      "learning_rate": 8e-05,
+      "loss": 1.7937,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3144210909974487,
+      "grad_norm": 0.4150988459587097,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3146640748390232,
+      "grad_norm": 0.4323060214519501,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 1295
+    },
+    {
+      "epoch": 0.31490705868059776,
+      "grad_norm": 0.4141565263271332,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 1296
+    },
+    {
+      "epoch": 0.31515004252217227,
+      "grad_norm": 0.41137444972991943,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 1297
+    },
+    {
+      "epoch": 0.31539302636374683,
+      "grad_norm": 0.40606027841567993,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 1298
+    },
+    {
+      "epoch": 0.31563601020532134,
+      "grad_norm": 0.41533470153808594,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3158789940468959,
+      "grad_norm": 0.39931249618530273,
+      "learning_rate": 8e-05,
+      "loss": 1.4294,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3161219778884704,
+      "grad_norm": 0.3959195613861084,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 1301
+    },
+    {
+      "epoch": 0.31636496173004497,
+      "grad_norm": 0.42165106534957886,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3166079455716195,
+      "grad_norm": 0.40195026993751526,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 1303
+    },
+    {
+      "epoch": 0.31685092941319404,
+      "grad_norm": 0.41727834939956665,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 1304
+    },
+    {
+      "epoch": 0.31709391325476854,
+      "grad_norm": 0.42682093381881714,
+      "learning_rate": 8e-05,
+      "loss": 1.8175,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3173368970963431,
+      "grad_norm": 0.4234749376773834,
+      "learning_rate": 8e-05,
+      "loss": 1.6421,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3175798809379176,
+      "grad_norm": 0.43160662055015564,
+      "learning_rate": 8e-05,
+      "loss": 1.7788,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3178228647794922,
+      "grad_norm": 0.39189645648002625,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3180658486210667,
+      "grad_norm": 0.41887709498405457,
+      "learning_rate": 8e-05,
+      "loss": 1.7226,
+      "step": 1309
+    },
+    {
+      "epoch": 0.31830883246264124,
+      "grad_norm": 0.42342591285705566,
+      "learning_rate": 8e-05,
+      "loss": 1.7036,
+      "step": 1310
+    },
+    {
+      "epoch": 0.31855181630421575,
+      "grad_norm": 0.4288804233074188,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3187948001457903,
+      "grad_norm": 0.4197126030921936,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3190377839873648,
+      "grad_norm": 0.43244338035583496,
+      "learning_rate": 8e-05,
+      "loss": 1.7697,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3192807678289394,
+      "grad_norm": 0.42230987548828125,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3195237516705139,
+      "grad_norm": 0.4212227463722229,
+      "learning_rate": 8e-05,
+      "loss": 1.7405,
+      "step": 1315
+    },
+    {
+      "epoch": 0.31976673551208845,
+      "grad_norm": 0.4391201436519623,
+      "learning_rate": 8e-05,
+      "loss": 1.7155,
+      "step": 1316
+    },
+    {
+      "epoch": 0.32000971935366296,
+      "grad_norm": 0.42658093571662903,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3202527031952375,
+      "grad_norm": 0.44827648997306824,
+      "learning_rate": 8e-05,
+      "loss": 1.8623,
+      "step": 1318
+    },
+    {
+      "epoch": 0.320495687036812,
+      "grad_norm": 0.5477005243301392,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3207386708783866,
+      "grad_norm": 0.4057164490222931,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3209816547199611,
+      "grad_norm": 0.438320517539978,
+      "learning_rate": 8e-05,
+      "loss": 1.7614,
+      "step": 1321
+    },
+    {
+      "epoch": 0.32122463856153566,
+      "grad_norm": 0.40908926725387573,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3214676224031102,
+      "grad_norm": 0.43416181206703186,
+      "learning_rate": 8e-05,
+      "loss": 1.7809,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3217106062446847,
+      "grad_norm": 0.40206679701805115,
+      "learning_rate": 8e-05,
+      "loss": 1.5453,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3219535900862593,
+      "grad_norm": 0.39225509762763977,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3221965739278338,
+      "grad_norm": 0.41371697187423706,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 1326
+    },
+    {
+      "epoch": 0.32243955776940836,
+      "grad_norm": 0.42055219411849976,
+      "learning_rate": 8e-05,
+      "loss": 1.8141,
+      "step": 1327
+    },
+    {
+      "epoch": 0.32268254161098286,
+      "grad_norm": 0.41531890630722046,
+      "learning_rate": 8e-05,
+      "loss": 1.7458,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3229255254525574,
+      "grad_norm": 0.39801040291786194,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 1329
+    },
+    {
+      "epoch": 0.32316850929413193,
+      "grad_norm": 0.4303327798843384,
+      "learning_rate": 8e-05,
+      "loss": 1.7254,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3234114931357065,
+      "grad_norm": 0.40039440989494324,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 1331
+    },
+    {
+      "epoch": 0.323654476977281,
+      "grad_norm": 0.42339256405830383,
+      "learning_rate": 8e-05,
+      "loss": 1.7048,
+      "step": 1332
+    },
+    {
+      "epoch": 0.32389746081885556,
+      "grad_norm": 0.4127884805202484,
+      "learning_rate": 8e-05,
+      "loss": 1.7663,
+      "step": 1333
+    },
+    {
+      "epoch": 0.32414044466043007,
+      "grad_norm": 0.3969059884548187,
+      "learning_rate": 8e-05,
+      "loss": 1.6196,
+      "step": 1334
+    },
+    {
+      "epoch": 0.32438342850200463,
+      "grad_norm": 0.4179375469684601,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 1335
+    },
+    {
+      "epoch": 0.32462641234357914,
+      "grad_norm": 0.42289412021636963,
+      "learning_rate": 8e-05,
+      "loss": 1.7619,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3248693961851537,
+      "grad_norm": 0.4024028182029724,
+      "learning_rate": 8e-05,
+      "loss": 1.6723,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3251123800267282,
+      "grad_norm": 0.42601507902145386,
+      "learning_rate": 8e-05,
+      "loss": 1.7478,
+      "step": 1338
+    },
+    {
+      "epoch": 0.32535536386830277,
+      "grad_norm": 0.4272320866584778,
+      "learning_rate": 8e-05,
+      "loss": 1.7722,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3255983477098773,
+      "grad_norm": 0.41830718517303467,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 1340
+    },
+    {
+      "epoch": 0.32584133155145184,
+      "grad_norm": 0.41066452860832214,
+      "learning_rate": 8e-05,
+      "loss": 1.7447,
+      "step": 1341
+    },
+    {
+      "epoch": 0.32608431539302635,
+      "grad_norm": 0.4213522970676422,
+      "learning_rate": 8e-05,
+      "loss": 1.7642,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3263272992346009,
+      "grad_norm": 0.4362061321735382,
+      "learning_rate": 8e-05,
+      "loss": 1.9346,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3265702830761754,
+      "grad_norm": 0.40942779183387756,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 1344
+    },
+    {
+      "epoch": 0.32681326691775,
+      "grad_norm": 0.41534700989723206,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3270562507593245,
+      "grad_norm": 0.43242648243904114,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 1346
+    },
+    {
+      "epoch": 0.32729923460089905,
+      "grad_norm": 0.45282021164894104,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 1347
+    },
+    {
+      "epoch": 0.32754221844247355,
+      "grad_norm": 0.41814547777175903,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3277852022840481,
+      "grad_norm": 0.433636873960495,
+      "learning_rate": 8e-05,
+      "loss": 1.8099,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3280281861256226,
+      "grad_norm": 0.42978718876838684,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3282711699671972,
+      "grad_norm": 0.4175388813018799,
+      "learning_rate": 8e-05,
+      "loss": 1.6847,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3285141538087717,
+      "grad_norm": 0.40037235617637634,
+      "learning_rate": 8e-05,
+      "loss": 1.7295,
+      "step": 1352
+    },
+    {
+      "epoch": 0.32875713765034625,
+      "grad_norm": 0.4433320462703705,
+      "learning_rate": 8e-05,
+      "loss": 1.8396,
+      "step": 1353
+    },
+    {
+      "epoch": 0.32900012149192076,
+      "grad_norm": 0.4098261594772339,
+      "learning_rate": 8e-05,
+      "loss": 1.6605,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3292431053334953,
+      "grad_norm": 0.4211520552635193,
+      "learning_rate": 8e-05,
+      "loss": 1.6127,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3294860891750699,
+      "grad_norm": 0.43561968207359314,
+      "learning_rate": 8e-05,
+      "loss": 1.784,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3297290730166444,
+      "grad_norm": 0.40554532408714294,
+      "learning_rate": 8e-05,
+      "loss": 1.4564,
+      "step": 1357
+    },
+    {
+      "epoch": 0.32997205685821895,
+      "grad_norm": 0.434696763753891,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 1358
+    },
+    {
+      "epoch": 0.33021504069979346,
+      "grad_norm": 0.4481261670589447,
+      "learning_rate": 8e-05,
+      "loss": 1.7727,
+      "step": 1359
+    },
+    {
+      "epoch": 0.330458024541368,
+      "grad_norm": 0.4229499399662018,
+      "learning_rate": 8e-05,
+      "loss": 1.7349,
+      "step": 1360
+    },
+    {
+      "epoch": 0.33070100838294253,
+      "grad_norm": 0.4054132103919983,
+      "learning_rate": 8e-05,
+      "loss": 1.4939,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3309439922245171,
+      "grad_norm": 0.4409117102622986,
+      "learning_rate": 8e-05,
+      "loss": 1.9241,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3311869760660916,
+      "grad_norm": 0.4396663010120392,
+      "learning_rate": 8e-05,
+      "loss": 1.732,
+      "step": 1363
+    },
+    {
+      "epoch": 0.33142995990766616,
+      "grad_norm": 0.42291024327278137,
+      "learning_rate": 8e-05,
+      "loss": 1.655,
+      "step": 1364
+    },
+    {
+      "epoch": 0.33167294374924067,
+      "grad_norm": 0.416900098323822,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 1365
+    },
+    {
+      "epoch": 0.33191592759081523,
+      "grad_norm": 0.41162773966789246,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 1366
+    },
+    {
+      "epoch": 0.33215891143238974,
+      "grad_norm": 0.4159378409385681,
+      "learning_rate": 8e-05,
+      "loss": 1.5593,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3324018952739643,
+      "grad_norm": 0.4554840326309204,
+      "learning_rate": 8e-05,
+      "loss": 1.886,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3326448791155388,
+      "grad_norm": 0.4456634521484375,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 1369
+    },
+    {
+      "epoch": 0.33288786295711337,
+      "grad_norm": 0.4127257466316223,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3331308467986879,
+      "grad_norm": 0.4211070239543915,
+      "learning_rate": 8e-05,
+      "loss": 1.5673,
+      "step": 1371
+    },
+    {
+      "epoch": 0.33337383064026244,
+      "grad_norm": 0.4010700285434723,
+      "learning_rate": 8e-05,
+      "loss": 1.6978,
+      "step": 1372
+    },
+    {
+      "epoch": 0.33361681448183694,
+      "grad_norm": 0.3907526135444641,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3338597983234115,
+      "grad_norm": 0.40844234824180603,
+      "learning_rate": 8e-05,
+      "loss": 1.7085,
+      "step": 1374
+    },
+    {
+      "epoch": 0.334102782164986,
+      "grad_norm": 0.4103162884712219,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3343457660065606,
+      "grad_norm": 0.4243805408477783,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3345887498481351,
+      "grad_norm": 0.4297308921813965,
+      "learning_rate": 8e-05,
+      "loss": 1.8386,
+      "step": 1377
+    },
+    {
+      "epoch": 0.33483173368970964,
+      "grad_norm": 0.4148638844490051,
+      "learning_rate": 8e-05,
+      "loss": 1.7938,
+      "step": 1378
+    },
+    {
+      "epoch": 0.33507471753128415,
+      "grad_norm": 0.4163691997528076,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3353177013728587,
+      "grad_norm": 0.41445234417915344,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3355606852144332,
+      "grad_norm": 0.4312629997730255,
+      "learning_rate": 8e-05,
+      "loss": 1.7591,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3358036690560078,
+      "grad_norm": 0.41140955686569214,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3360466528975823,
+      "grad_norm": 0.4554731249809265,
+      "learning_rate": 8e-05,
+      "loss": 1.7677,
+      "step": 1383
+    },
+    {
+      "epoch": 0.33628963673915685,
+      "grad_norm": 0.44381895661354065,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 1384
+    },
+    {
+      "epoch": 0.33653262058073136,
+      "grad_norm": 0.4227243661880493,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3367756044223059,
+      "grad_norm": 0.4265260100364685,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3370185882638804,
+      "grad_norm": 0.4079815149307251,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1387
+    },
+    {
+      "epoch": 0.337261572105455,
+      "grad_norm": 0.4142341911792755,
+      "learning_rate": 8e-05,
+      "loss": 1.766,
+      "step": 1388
+    },
+    {
+      "epoch": 0.33750455594702955,
+      "grad_norm": 0.424040824174881,
+      "learning_rate": 8e-05,
+      "loss": 1.8513,
+      "step": 1389
+    },
+    {
+      "epoch": 0.33774753978860406,
+      "grad_norm": 0.41974717378616333,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3379905236301786,
+      "grad_norm": 0.42985066771507263,
+      "learning_rate": 8e-05,
+      "loss": 1.8263,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3382335074717531,
+      "grad_norm": 0.4142277240753174,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3384764913133277,
+      "grad_norm": 0.398746132850647,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3387194751549022,
+      "grad_norm": 0.40721410512924194,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 1394
+    },
+    {
+      "epoch": 0.33896245899647676,
+      "grad_norm": 0.421593576669693,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 1395
+    },
+    {
+      "epoch": 0.33920544283805126,
+      "grad_norm": 0.4076847732067108,
+      "learning_rate": 8e-05,
+      "loss": 1.7031,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3394484266796258,
+      "grad_norm": 0.44437170028686523,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 1397
+    },
+    {
+      "epoch": 0.33969141052120033,
+      "grad_norm": 0.4115573465824127,
+      "learning_rate": 8e-05,
+      "loss": 1.65,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3399343943627749,
+      "grad_norm": 0.42203497886657715,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3401773782043494,
+      "grad_norm": 0.42462655901908875,
+      "learning_rate": 8e-05,
+      "loss": 1.6971,
+      "step": 1400
+    },
+    {
+      "epoch": 0.34042036204592396,
+      "grad_norm": 0.42710965871810913,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 1401
+    },
+    {
+      "epoch": 0.34066334588749847,
+      "grad_norm": 0.41822290420532227,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 1402
+    },
+    {
+      "epoch": 0.34090632972907303,
+      "grad_norm": 0.40744471549987793,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1403
+    },
+    {
+      "epoch": 0.34114931357064754,
+      "grad_norm": 0.4046993553638458,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3413922974122221,
+      "grad_norm": 0.4165453612804413,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3416352812537966,
+      "grad_norm": 0.4177655279636383,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 1406
+    },
+    {
+      "epoch": 0.34187826509537117,
+      "grad_norm": 0.43630239367485046,
+      "learning_rate": 8e-05,
+      "loss": 1.7462,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3421212489369457,
+      "grad_norm": 0.43209436535835266,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1408
+    },
+    {
+      "epoch": 0.34236423277852024,
+      "grad_norm": 0.4308101534843445,
+      "learning_rate": 8e-05,
+      "loss": 1.7389,
+      "step": 1409
+    },
+    {
+      "epoch": 0.34260721662009475,
+      "grad_norm": 0.40700584650039673,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3428502004616693,
+      "grad_norm": 0.4164145290851593,
+      "learning_rate": 8e-05,
+      "loss": 1.7506,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3430931843032438,
+      "grad_norm": 0.40757638216018677,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3433361681448184,
+      "grad_norm": 0.4093579053878784,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3435791519863929,
+      "grad_norm": 0.41316044330596924,
+      "learning_rate": 8e-05,
+      "loss": 1.7199,
+      "step": 1414
+    },
+    {
+      "epoch": 0.34382213582796745,
+      "grad_norm": 0.41529348492622375,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 1415
+    },
+    {
+      "epoch": 0.34406511966954195,
+      "grad_norm": 0.410444051027298,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3443081035111165,
+      "grad_norm": 0.4090292453765869,
+      "learning_rate": 8e-05,
+      "loss": 1.8186,
+      "step": 1417
+    },
+    {
+      "epoch": 0.344551087352691,
+      "grad_norm": 0.44185879826545715,
+      "learning_rate": 8e-05,
+      "loss": 1.8422,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3447940711942656,
+      "grad_norm": 0.4232048988342285,
+      "learning_rate": 8e-05,
+      "loss": 1.7903,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3450370550358401,
+      "grad_norm": 0.4253289997577667,
+      "learning_rate": 8e-05,
+      "loss": 1.789,
+      "step": 1420
+    },
+    {
+      "epoch": 0.34528003887741465,
+      "grad_norm": 0.3991374373435974,
+      "learning_rate": 8e-05,
+      "loss": 1.5593,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3455230227189892,
+      "grad_norm": 0.41997548937797546,
+      "learning_rate": 8e-05,
+      "loss": 1.7432,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3457660065605637,
+      "grad_norm": 0.4185141921043396,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3460089904021383,
+      "grad_norm": 0.41045713424682617,
+      "learning_rate": 8e-05,
+      "loss": 1.7732,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3462519742437128,
+      "grad_norm": 0.4068754315376282,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 1425
+    },
+    {
+      "epoch": 0.34649495808528735,
+      "grad_norm": 0.40424397587776184,
+      "learning_rate": 8e-05,
+      "loss": 1.6652,
+      "step": 1426
+    },
+    {
+      "epoch": 0.34673794192686186,
+      "grad_norm": 0.39907094836235046,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3469809257684364,
+      "grad_norm": 0.42224764823913574,
+      "learning_rate": 8e-05,
+      "loss": 1.8078,
+      "step": 1428
+    },
+    {
+      "epoch": 0.34722390961001093,
+      "grad_norm": 0.42266082763671875,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3474668934515855,
+      "grad_norm": 0.4065650701522827,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1430
+    },
+    {
+      "epoch": 0.34770987729316,
+      "grad_norm": 0.42584428191185,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 1431
+    },
+    {
+      "epoch": 0.34795286113473456,
+      "grad_norm": 0.43668845295906067,
+      "learning_rate": 8e-05,
+      "loss": 1.8497,
+      "step": 1432
+    },
+    {
+      "epoch": 0.34819584497630907,
+      "grad_norm": 0.422085702419281,
+      "learning_rate": 8e-05,
+      "loss": 1.5448,
+      "step": 1433
+    },
+    {
+      "epoch": 0.34843882881788363,
+      "grad_norm": 0.438243567943573,
+      "learning_rate": 8e-05,
+      "loss": 1.9311,
+      "step": 1434
+    },
+    {
+      "epoch": 0.34868181265945813,
+      "grad_norm": 0.412178099155426,
+      "learning_rate": 8e-05,
+      "loss": 1.7131,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3489247965010327,
+      "grad_norm": 0.4246040880680084,
+      "learning_rate": 8e-05,
+      "loss": 1.7761,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3491677803426072,
+      "grad_norm": 0.44907382130622864,
+      "learning_rate": 8e-05,
+      "loss": 1.8752,
+      "step": 1437
+    },
+    {
+      "epoch": 0.34941076418418177,
+      "grad_norm": 0.4161170423030853,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3496537480257563,
+      "grad_norm": 0.4451916813850403,
+      "learning_rate": 8e-05,
+      "loss": 1.7319,
+      "step": 1439
+    },
+    {
+      "epoch": 0.34989673186733083,
+      "grad_norm": 0.4271717369556427,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 1440
+    },
+    {
+      "epoch": 0.35013971570890534,
+      "grad_norm": 0.43731844425201416,
+      "learning_rate": 8e-05,
+      "loss": 1.8988,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3503826995504799,
+      "grad_norm": 0.41044342517852783,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3506256833920544,
+      "grad_norm": 0.4042555093765259,
+      "learning_rate": 8e-05,
+      "loss": 1.757,
+      "step": 1443
+    },
+    {
+      "epoch": 0.350868667233629,
+      "grad_norm": 0.44009044766426086,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3511116510752035,
+      "grad_norm": 0.3960078954696655,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 1445
+    },
+    {
+      "epoch": 0.35135463491677804,
+      "grad_norm": 0.4207194745540619,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 1446
+    },
+    {
+      "epoch": 0.35159761875835255,
+      "grad_norm": 0.41871359944343567,
+      "learning_rate": 8e-05,
+      "loss": 1.6439,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3518406025999271,
+      "grad_norm": 0.42891284823417664,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3520835864415016,
+      "grad_norm": 0.4187850058078766,
+      "learning_rate": 8e-05,
+      "loss": 1.6498,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3523265702830762,
+      "grad_norm": 0.4176000952720642,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3525695541246507,
+      "grad_norm": 0.4076887369155884,
+      "learning_rate": 8e-05,
+      "loss": 1.6618,
+      "step": 1451
+    },
+    {
+      "epoch": 0.35281253796622525,
+      "grad_norm": 0.4008098244667053,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 1452
+    },
+    {
+      "epoch": 0.35305552180779975,
+      "grad_norm": 0.42970818281173706,
+      "learning_rate": 8e-05,
+      "loss": 1.7839,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3532985056493743,
+      "grad_norm": 0.43289485573768616,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3535414894909489,
+      "grad_norm": 0.43016672134399414,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3537844733325234,
+      "grad_norm": 0.41671884059906006,
+      "learning_rate": 8e-05,
+      "loss": 1.7256,
+      "step": 1456
+    },
+    {
+      "epoch": 0.35402745717409795,
+      "grad_norm": 0.4214855432510376,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 1457
+    },
+    {
+      "epoch": 0.35427044101567245,
+      "grad_norm": 0.4305601716041565,
+      "learning_rate": 8e-05,
+      "loss": 1.804,
+      "step": 1458
+    },
+    {
+      "epoch": 0.354513424857247,
+      "grad_norm": 0.42480790615081787,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3547564086988215,
+      "grad_norm": 0.41388338804244995,
+      "learning_rate": 8e-05,
+      "loss": 1.714,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3549993925403961,
+      "grad_norm": 0.418966680765152,
+      "learning_rate": 8e-05,
+      "loss": 1.7532,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3552423763819706,
+      "grad_norm": 0.4355233907699585,
+      "learning_rate": 8e-05,
+      "loss": 1.6545,
+      "step": 1462
+    },
+    {
+      "epoch": 0.35548536022354515,
+      "grad_norm": 0.45201870799064636,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 1463
+    },
+    {
+      "epoch": 0.35572834406511966,
+      "grad_norm": 0.43834418058395386,
+      "learning_rate": 8e-05,
+      "loss": 1.8601,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3559713279066942,
+      "grad_norm": 0.4327118694782257,
+      "learning_rate": 8e-05,
+      "loss": 1.6826,
+      "step": 1465
+    },
+    {
+      "epoch": 0.35621431174826873,
+      "grad_norm": 0.4124552011489868,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3564572955898433,
+      "grad_norm": 0.4168725907802582,
+      "learning_rate": 8e-05,
+      "loss": 1.6836,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3567002794314178,
+      "grad_norm": 0.4121659994125366,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 1468
+    },
+    {
+      "epoch": 0.35694326327299236,
+      "grad_norm": 0.40972062945365906,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 1469
+    },
+    {
+      "epoch": 0.35718624711456687,
+      "grad_norm": 0.43188977241516113,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 1470
+    },
+    {
+      "epoch": 0.35742923095614143,
+      "grad_norm": 0.4164832532405853,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 1471
+    },
+    {
+      "epoch": 0.35767221479771594,
+      "grad_norm": 0.41471636295318604,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3579151986392905,
+      "grad_norm": 0.42676058411598206,
+      "learning_rate": 8e-05,
+      "loss": 1.7083,
+      "step": 1473
+    },
+    {
+      "epoch": 0.358158182480865,
+      "grad_norm": 0.4020005464553833,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 1474
+    },
+    {
+      "epoch": 0.35840116632243957,
+      "grad_norm": 0.43297043442726135,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3586441501640141,
+      "grad_norm": 0.4124680459499359,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 1476
+    },
+    {
+      "epoch": 0.35888713400558864,
+      "grad_norm": 0.42065784335136414,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 1477
+    },
+    {
+      "epoch": 0.35913011784716314,
+      "grad_norm": 0.439014196395874,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3593731016887377,
+      "grad_norm": 0.41923463344573975,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3596160855303122,
+      "grad_norm": 0.40060892701148987,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3598590693718868,
+      "grad_norm": 0.41042670607566833,
+      "learning_rate": 8e-05,
+      "loss": 1.7557,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3601020532134613,
+      "grad_norm": 0.45834240317344666,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 1482
+    },
+    {
+      "epoch": 0.36034503705503584,
+      "grad_norm": 0.44385936856269836,
+      "learning_rate": 8e-05,
+      "loss": 1.7123,
+      "step": 1483
+    },
+    {
+      "epoch": 0.36058802089661035,
+      "grad_norm": 0.4490148425102234,
+      "learning_rate": 8e-05,
+      "loss": 1.7722,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3608310047381849,
+      "grad_norm": 0.44380930066108704,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3610739885797594,
+      "grad_norm": 0.45426803827285767,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 1486
+    },
+    {
+      "epoch": 0.361316972421334,
+      "grad_norm": 0.42598724365234375,
+      "learning_rate": 8e-05,
+      "loss": 1.8526,
+      "step": 1487
+    },
+    {
+      "epoch": 0.36155995626290854,
+      "grad_norm": 0.41523346304893494,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 1488
+    },
+    {
+      "epoch": 0.36180294010448305,
+      "grad_norm": 0.40627723932266235,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3620459239460576,
+      "grad_norm": 0.44283729791641235,
+      "learning_rate": 8e-05,
+      "loss": 1.7863,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3622889077876321,
+      "grad_norm": 0.42527687549591064,
+      "learning_rate": 8e-05,
+      "loss": 1.7138,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3625318916292067,
+      "grad_norm": 0.4094756543636322,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3627748754707812,
+      "grad_norm": 0.44145467877388,
+      "learning_rate": 8e-05,
+      "loss": 1.7635,
+      "step": 1493
+    },
+    {
+      "epoch": 0.36301785931235575,
+      "grad_norm": 0.41469717025756836,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 1494
+    },
+    {
+      "epoch": 0.36326084315393026,
+      "grad_norm": 0.4286789298057556,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3635038269955048,
+      "grad_norm": 0.39633095264434814,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3637468108370793,
+      "grad_norm": 0.44227567315101624,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3639897946786539,
+      "grad_norm": 0.42462319135665894,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3642327785202284,
+      "grad_norm": 0.5036365985870361,
+      "learning_rate": 8e-05,
+      "loss": 1.7893,
+      "step": 1499
+    },
+    {
+      "epoch": 0.36447576236180296,
+      "grad_norm": 0.40516528487205505,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36471874620337746,
+      "grad_norm": 0.4498935341835022,
+      "learning_rate": 8e-05,
+      "loss": 1.8523,
+      "step": 1501
+    },
+    {
+      "epoch": 0.364961730044952,
+      "grad_norm": 0.44853898882865906,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 1502
+    },
+    {
+      "epoch": 0.36520471388652653,
+      "grad_norm": 0.43535977602005005,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3654476977281011,
+      "grad_norm": 0.4103722870349884,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3656906815696756,
+      "grad_norm": 0.4043563902378082,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 1505
+    },
+    {
+      "epoch": 0.36593366541125016,
+      "grad_norm": 0.4265104830265045,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 1506
+    },
+    {
+      "epoch": 0.36617664925282467,
+      "grad_norm": 0.45437732338905334,
+      "learning_rate": 8e-05,
+      "loss": 1.7729,
+      "step": 1507
+    },
+    {
+      "epoch": 0.36641963309439923,
+      "grad_norm": 0.4220513105392456,
+      "learning_rate": 8e-05,
+      "loss": 1.6852,
+      "step": 1508
+    },
+    {
+      "epoch": 0.36666261693597374,
+      "grad_norm": 0.4355618953704834,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3669056007775483,
+      "grad_norm": 0.3921411335468292,
+      "learning_rate": 8e-05,
+      "loss": 1.6566,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3671485846191228,
+      "grad_norm": 0.4296509921550751,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 1511
+    },
+    {
+      "epoch": 0.36739156846069737,
+      "grad_norm": 0.42648905515670776,
+      "learning_rate": 8e-05,
+      "loss": 1.7684,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3676345523022719,
+      "grad_norm": 0.41103217005729675,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 1513
+    },
+    {
+      "epoch": 0.36787753614384644,
+      "grad_norm": 0.4502837657928467,
+      "learning_rate": 8e-05,
+      "loss": 1.8414,
+      "step": 1514
+    },
+    {
+      "epoch": 0.36812051998542095,
+      "grad_norm": 0.4428488612174988,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3683635038269955,
+      "grad_norm": 0.434501588344574,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 1516
+    },
+    {
+      "epoch": 0.36860648766857,
+      "grad_norm": 0.429174542427063,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3688494715101446,
+      "grad_norm": 0.4316585063934326,
+      "learning_rate": 8e-05,
+      "loss": 1.6992,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3690924553517191,
+      "grad_norm": 0.43607017397880554,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 1519
+    },
+    {
+      "epoch": 0.36933543919329365,
+      "grad_norm": 0.3982023596763611,
+      "learning_rate": 8e-05,
+      "loss": 1.558,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3695784230348682,
+      "grad_norm": 0.4116176962852478,
+      "learning_rate": 8e-05,
+      "loss": 1.6175,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3698214068764427,
+      "grad_norm": 0.45402878522872925,
+      "learning_rate": 8e-05,
+      "loss": 1.7713,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3700643907180173,
+      "grad_norm": 0.42919039726257324,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3703073745595918,
+      "grad_norm": 0.42670485377311707,
+      "learning_rate": 8e-05,
+      "loss": 1.7326,
+      "step": 1524
+    },
+    {
+      "epoch": 0.37055035840116635,
+      "grad_norm": 0.4170015752315521,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 1525
+    },
+    {
+      "epoch": 0.37079334224274085,
+      "grad_norm": 0.42701733112335205,
+      "learning_rate": 8e-05,
+      "loss": 1.7559,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3710363260843154,
+      "grad_norm": 0.4310395419597626,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3712793099258899,
+      "grad_norm": 0.40354016423225403,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3715222937674645,
+      "grad_norm": 0.4501798748970032,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 1529
+    },
+    {
+      "epoch": 0.371765277609039,
+      "grad_norm": 0.4462423026561737,
+      "learning_rate": 8e-05,
+      "loss": 1.8189,
+      "step": 1530
+    },
+    {
+      "epoch": 0.37200826145061355,
+      "grad_norm": 0.4353228211402893,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 1531
+    },
+    {
+      "epoch": 0.37225124529218806,
+      "grad_norm": 0.4250049591064453,
+      "learning_rate": 8e-05,
+      "loss": 1.6773,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3724942291337626,
+      "grad_norm": 0.4075191915035248,
+      "learning_rate": 8e-05,
+      "loss": 1.7537,
+      "step": 1533
+    },
+    {
+      "epoch": 0.37273721297533713,
+      "grad_norm": 0.4113391935825348,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3729801968169117,
+      "grad_norm": 0.41470906138420105,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3732231806584862,
+      "grad_norm": 0.4519250690937042,
+      "learning_rate": 8e-05,
+      "loss": 1.755,
+      "step": 1536
+    },
+    {
+      "epoch": 0.37346616450006076,
+      "grad_norm": 0.41602808237075806,
+      "learning_rate": 8e-05,
+      "loss": 1.6691,
+      "step": 1537
+    },
+    {
+      "epoch": 0.37370914834163527,
+      "grad_norm": 0.4330156147480011,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 1538
+    },
+    {
+      "epoch": 0.37395213218320983,
+      "grad_norm": 0.428617000579834,
+      "learning_rate": 8e-05,
+      "loss": 1.8051,
+      "step": 1539
+    },
+    {
+      "epoch": 0.37419511602478434,
+      "grad_norm": 0.42976975440979004,
+      "learning_rate": 8e-05,
+      "loss": 1.5662,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3744380998663589,
+      "grad_norm": 0.40064191818237305,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3746810837079334,
+      "grad_norm": 0.4101344645023346,
+      "learning_rate": 8e-05,
+      "loss": 1.6954,
+      "step": 1542
+    },
+    {
+      "epoch": 0.37492406754950797,
+      "grad_norm": 0.4178948998451233,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3751670513910825,
+      "grad_norm": 0.4114849865436554,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 1544
+    },
+    {
+      "epoch": 0.37541003523265704,
+      "grad_norm": 0.4005047082901001,
+      "learning_rate": 8e-05,
+      "loss": 1.5517,
+      "step": 1545
+    },
+    {
+      "epoch": 0.37565301907423154,
+      "grad_norm": 0.4733780026435852,
+      "learning_rate": 8e-05,
+      "loss": 1.8076,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3758960029158061,
+      "grad_norm": 0.4235011041164398,
+      "learning_rate": 8e-05,
+      "loss": 1.625,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3761389867573806,
+      "grad_norm": 0.4334684908390045,
+      "learning_rate": 8e-05,
+      "loss": 1.747,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3763819705989552,
+      "grad_norm": 0.42389872670173645,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3766249544405297,
+      "grad_norm": 0.4101676046848297,
+      "learning_rate": 8e-05,
+      "loss": 1.7444,
+      "step": 1550
+    },
+    {
+      "epoch": 0.37686793828210424,
+      "grad_norm": 0.41118088364601135,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 1551
+    },
+    {
+      "epoch": 0.37711092212367875,
+      "grad_norm": 0.431918203830719,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3773539059652533,
+      "grad_norm": 0.41058656573295593,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3775968898068279,
+      "grad_norm": 0.4221169650554657,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3778398736484024,
+      "grad_norm": 0.43766945600509644,
+      "learning_rate": 8e-05,
+      "loss": 1.8801,
+      "step": 1555
+    },
+    {
+      "epoch": 0.37808285748997694,
+      "grad_norm": 0.4249259829521179,
+      "learning_rate": 8e-05,
+      "loss": 1.5557,
+      "step": 1556
+    },
+    {
+      "epoch": 0.37832584133155145,
+      "grad_norm": 0.4021548330783844,
+      "learning_rate": 8e-05,
+      "loss": 1.5998,
+      "step": 1557
+    },
+    {
+      "epoch": 0.378568825173126,
+      "grad_norm": 0.402353972196579,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3788118090147005,
+      "grad_norm": 0.42179930210113525,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3790547928562751,
+      "grad_norm": 0.4151405394077301,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3792977766978496,
+      "grad_norm": 0.43462592363357544,
+      "learning_rate": 8e-05,
+      "loss": 1.5964,
+      "step": 1561
+    },
+    {
+      "epoch": 0.37954076053942415,
+      "grad_norm": 0.424850732088089,
+      "learning_rate": 8e-05,
+      "loss": 1.6482,
+      "step": 1562
+    },
+    {
+      "epoch": 0.37978374438099866,
+      "grad_norm": 0.41227251291275024,
+      "learning_rate": 8e-05,
+      "loss": 1.5886,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3800267282225732,
+      "grad_norm": 0.4173636734485626,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3802697120641477,
+      "grad_norm": 0.40978628396987915,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3805126959057223,
+      "grad_norm": 0.40448302030563354,
+      "learning_rate": 8e-05,
+      "loss": 1.4954,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3807556797472968,
+      "grad_norm": 0.42568641901016235,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 1567
+    },
+    {
+      "epoch": 0.38099866358887136,
+      "grad_norm": 0.4181409180164337,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 1568
+    },
+    {
+      "epoch": 0.38124164743044586,
+      "grad_norm": 0.4193849265575409,
+      "learning_rate": 8e-05,
+      "loss": 1.7605,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3814846312720204,
+      "grad_norm": 0.4197647273540497,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 1570
+    },
+    {
+      "epoch": 0.38172761511359493,
+      "grad_norm": 0.4191533923149109,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3819705989551695,
+      "grad_norm": 0.4143694341182709,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 1572
+    },
+    {
+      "epoch": 0.382213582796744,
+      "grad_norm": 0.41417399048805237,
+      "learning_rate": 8e-05,
+      "loss": 1.6817,
+      "step": 1573
+    },
+    {
+      "epoch": 0.38245656663831856,
+      "grad_norm": 0.4393376111984253,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 1574
+    },
+    {
+      "epoch": 0.38269955047989307,
+      "grad_norm": 0.4121435284614563,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 1575
+    },
+    {
+      "epoch": 0.38294253432146763,
+      "grad_norm": 0.41746968030929565,
+      "learning_rate": 8e-05,
+      "loss": 1.7227,
+      "step": 1576
+    },
+    {
+      "epoch": 0.38318551816304214,
+      "grad_norm": 0.43105581402778625,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3834285020046167,
+      "grad_norm": 0.43507346510887146,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3836714858461912,
+      "grad_norm": 0.3958384692668915,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 1579
+    },
+    {
+      "epoch": 0.38391446968776577,
+      "grad_norm": 0.44736096262931824,
+      "learning_rate": 8e-05,
+      "loss": 1.8793,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3841574535293403,
+      "grad_norm": 0.4385405480861664,
+      "learning_rate": 8e-05,
+      "loss": 1.8495,
+      "step": 1581
+    },
+    {
+      "epoch": 0.38440043737091484,
+      "grad_norm": 0.438483864068985,
+      "learning_rate": 8e-05,
+      "loss": 1.8367,
+      "step": 1582
+    },
+    {
+      "epoch": 0.38464342121248934,
+      "grad_norm": 0.4386710822582245,
+      "learning_rate": 8e-05,
+      "loss": 1.7457,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3848864050540639,
+      "grad_norm": 0.44526001811027527,
+      "learning_rate": 8e-05,
+      "loss": 1.7394,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3851293888956384,
+      "grad_norm": 0.43033620715141296,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 1585
+    },
+    {
+      "epoch": 0.385372372737213,
+      "grad_norm": 0.43658843636512756,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 1586
+    },
+    {
+      "epoch": 0.38561535657878754,
+      "grad_norm": 0.4354395866394043,
+      "learning_rate": 8e-05,
+      "loss": 1.7699,
+      "step": 1587
+    },
+    {
+      "epoch": 0.38585834042036204,
+      "grad_norm": 0.40769627690315247,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3861013242619366,
+      "grad_norm": 0.4172569215297699,
+      "learning_rate": 8e-05,
+      "loss": 1.6343,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3863443081035111,
+      "grad_norm": 0.4266487956047058,
+      "learning_rate": 8e-05,
+      "loss": 1.7551,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3865872919450857,
+      "grad_norm": 0.4560794234275818,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3868302757866602,
+      "grad_norm": 0.4079170525074005,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 1592
+    },
+    {
+      "epoch": 0.38707325962823474,
+      "grad_norm": 0.43422430753707886,
+      "learning_rate": 8e-05,
+      "loss": 1.7242,
+      "step": 1593
+    },
+    {
+      "epoch": 0.38731624346980925,
+      "grad_norm": 0.4020366370677948,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3875592273113838,
+      "grad_norm": 0.42723163962364197,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3878022111529583,
+      "grad_norm": 0.44158118963241577,
+      "learning_rate": 8e-05,
+      "loss": 1.7811,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3880451949945329,
+      "grad_norm": 0.3944547176361084,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3882881788361074,
+      "grad_norm": 0.43721675872802734,
+      "learning_rate": 8e-05,
+      "loss": 1.6909,
+      "step": 1598
+    },
+    {
+      "epoch": 0.38853116267768195,
+      "grad_norm": 0.44996586441993713,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 1599
+    },
+    {
+      "epoch": 0.38877414651925646,
+      "grad_norm": 0.42338284850120544,
+      "learning_rate": 8e-05,
+      "loss": 1.7383,
+      "step": 1600
+    },
+    {
+      "epoch": 0.389017130360831,
+      "grad_norm": 0.44629910588264465,
+      "learning_rate": 8e-05,
+      "loss": 1.7169,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3892601142024055,
+      "grad_norm": 0.4520544111728668,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3895030980439801,
+      "grad_norm": 0.396907776594162,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3897460818855546,
+      "grad_norm": 0.4147047996520996,
+      "learning_rate": 8e-05,
+      "loss": 1.7224,
+      "step": 1604
+    },
+    {
+      "epoch": 0.38998906572712916,
+      "grad_norm": 0.42678162455558777,
+      "learning_rate": 8e-05,
+      "loss": 1.7471,
+      "step": 1605
+    },
+    {
+      "epoch": 0.39023204956870366,
+      "grad_norm": 0.42349156737327576,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3904750334102782,
+      "grad_norm": 0.40292665362358093,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 1607
+    },
+    {
+      "epoch": 0.39071801725185273,
+      "grad_norm": 0.44564610719680786,
+      "learning_rate": 8e-05,
+      "loss": 1.8699,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3909610010934273,
+      "grad_norm": 0.41767212748527527,
+      "learning_rate": 8e-05,
+      "loss": 1.7606,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3912039849350018,
+      "grad_norm": 0.4166221022605896,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 1610
+    },
+    {
+      "epoch": 0.39144696877657637,
+      "grad_norm": 0.4434347152709961,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 1611
+    },
+    {
+      "epoch": 0.39168995261815087,
+      "grad_norm": 0.4146440923213959,
+      "learning_rate": 8e-05,
+      "loss": 1.7159,
+      "step": 1612
+    },
+    {
+      "epoch": 0.39193293645972543,
+      "grad_norm": 0.3920593559741974,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 1613
+    },
+    {
+      "epoch": 0.39217592030129994,
+      "grad_norm": 0.4225471317768097,
+      "learning_rate": 8e-05,
+      "loss": 1.6693,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3924189041428745,
+      "grad_norm": 0.44872719049453735,
+      "learning_rate": 8e-05,
+      "loss": 1.6434,
+      "step": 1615
+    },
+    {
+      "epoch": 0.392661887984449,
+      "grad_norm": 0.3961496353149414,
+      "learning_rate": 8e-05,
+      "loss": 1.6164,
+      "step": 1616
+    },
+    {
+      "epoch": 0.39290487182602357,
+      "grad_norm": 0.38982248306274414,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3931478556675981,
+      "grad_norm": 0.4293288588523865,
+      "learning_rate": 8e-05,
+      "loss": 1.8845,
+      "step": 1618
+    },
+    {
+      "epoch": 0.39339083950917264,
+      "grad_norm": 0.4216848611831665,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3936338233507472,
+      "grad_norm": 0.3969649374485016,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3938768071923217,
+      "grad_norm": 0.4064768850803375,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 1621
+    },
+    {
+      "epoch": 0.39411979103389627,
+      "grad_norm": 0.41068586707115173,
+      "learning_rate": 8e-05,
+      "loss": 1.7272,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3943627748754708,
+      "grad_norm": 0.4052871763706207,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 1623
+    },
+    {
+      "epoch": 0.39460575871704534,
+      "grad_norm": 0.435359925031662,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 1624
+    },
+    {
+      "epoch": 0.39484874255861985,
+      "grad_norm": 0.4205849766731262,
+      "learning_rate": 8e-05,
+      "loss": 1.7329,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3950917264001944,
+      "grad_norm": 0.43587934970855713,
+      "learning_rate": 8e-05,
+      "loss": 1.7942,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3953347102417689,
+      "grad_norm": 0.41418036818504333,
+      "learning_rate": 8e-05,
+      "loss": 1.6616,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3955776940833435,
+      "grad_norm": 0.3877982795238495,
+      "learning_rate": 8e-05,
+      "loss": 1.6862,
+      "step": 1628
+    },
+    {
+      "epoch": 0.395820677924918,
+      "grad_norm": 0.4145828187465668,
+      "learning_rate": 8e-05,
+      "loss": 1.5951,
+      "step": 1629
+    },
+    {
+      "epoch": 0.39606366176649255,
+      "grad_norm": 0.4185433089733124,
+      "learning_rate": 8e-05,
+      "loss": 1.7882,
+      "step": 1630
+    },
+    {
+      "epoch": 0.39630664560806705,
+      "grad_norm": 0.41606849431991577,
+      "learning_rate": 8e-05,
+      "loss": 1.6905,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3965496294496416,
+      "grad_norm": 0.4242626130580902,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3967926132912161,
+      "grad_norm": 0.39382022619247437,
+      "learning_rate": 8e-05,
+      "loss": 1.5491,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3970355971327907,
+      "grad_norm": 0.4146677553653717,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3972785809743652,
+      "grad_norm": 0.4308668076992035,
+      "learning_rate": 8e-05,
+      "loss": 1.8234,
+      "step": 1635
+    },
+    {
+      "epoch": 0.39752156481593975,
+      "grad_norm": 0.39495551586151123,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 1636
+    },
+    {
+      "epoch": 0.39776454865751426,
+      "grad_norm": 0.4327457845211029,
+      "learning_rate": 8e-05,
+      "loss": 1.8254,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3980075324990888,
+      "grad_norm": 0.4194384217262268,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 1638
+    },
+    {
+      "epoch": 0.39825051634066333,
+      "grad_norm": 0.4201095998287201,
+      "learning_rate": 8e-05,
+      "loss": 1.6589,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3984935001822379,
+      "grad_norm": 0.42406922578811646,
+      "learning_rate": 8e-05,
+      "loss": 1.8066,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3987364840238124,
+      "grad_norm": 0.4152491092681885,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 1641
+    },
+    {
+      "epoch": 0.39897946786538696,
+      "grad_norm": 0.4010281562805176,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 1642
+    },
+    {
+      "epoch": 0.39922245170696147,
+      "grad_norm": 0.4447481036186218,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 1643
+    },
+    {
+      "epoch": 0.39946543554853603,
+      "grad_norm": 0.4013897478580475,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 1644
+    },
+    {
+      "epoch": 0.39970841939011054,
+      "grad_norm": 0.400485098361969,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3999514032316851,
+      "grad_norm": 0.42620760202407837,
+      "learning_rate": 8e-05,
+      "loss": 1.7478,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4001943870732596,
+      "grad_norm": 0.45568305253982544,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 1647
+    },
+    {
+      "epoch": 0.40043737091483417,
+      "grad_norm": 0.42114564776420593,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 1648
+    },
+    {
+      "epoch": 0.4006803547564087,
+      "grad_norm": 0.4360169470310211,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 1649
+    },
+    {
+      "epoch": 0.40092333859798324,
+      "grad_norm": 0.4264334738254547,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 1650
+    },
+    {
+      "epoch": 0.40116632243955774,
+      "grad_norm": 0.4771878123283386,
+      "learning_rate": 8e-05,
+      "loss": 1.9623,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4014093062811323,
+      "grad_norm": 0.40234559774398804,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 1652
+    },
+    {
+      "epoch": 0.40165229012270687,
+      "grad_norm": 0.4067186415195465,
+      "learning_rate": 8e-05,
+      "loss": 1.6945,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4018952739642814,
+      "grad_norm": 0.4087909758090973,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 1654
+    },
+    {
+      "epoch": 0.40213825780585594,
+      "grad_norm": 0.42809879779815674,
+      "learning_rate": 8e-05,
+      "loss": 1.6883,
+      "step": 1655
+    },
+    {
+      "epoch": 0.40238124164743044,
+      "grad_norm": 0.4224081337451935,
+      "learning_rate": 8e-05,
+      "loss": 1.7737,
+      "step": 1656
+    },
+    {
+      "epoch": 0.402624225489005,
+      "grad_norm": 0.41031697392463684,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4028672093305795,
+      "grad_norm": 0.43192222714424133,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4031101931721541,
+      "grad_norm": 0.4231877624988556,
+      "learning_rate": 8e-05,
+      "loss": 1.7694,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4033531770137286,
+      "grad_norm": 0.4189487397670746,
+      "learning_rate": 8e-05,
+      "loss": 1.723,
+      "step": 1660
+    },
+    {
+      "epoch": 0.40359616085530314,
+      "grad_norm": 0.43119171261787415,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 1661
+    },
+    {
+      "epoch": 0.40383914469687765,
+      "grad_norm": 0.39486396312713623,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4040821285384522,
+      "grad_norm": 0.42586183547973633,
+      "learning_rate": 8e-05,
+      "loss": 1.87,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4043251123800267,
+      "grad_norm": 0.42093515396118164,
+      "learning_rate": 8e-05,
+      "loss": 1.6897,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4045680962216013,
+      "grad_norm": 0.418782114982605,
+      "learning_rate": 8e-05,
+      "loss": 1.7326,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4048110800631758,
+      "grad_norm": 0.43015027046203613,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 1666
+    },
+    {
+      "epoch": 0.40505406390475035,
+      "grad_norm": 0.44864341616630554,
+      "learning_rate": 8e-05,
+      "loss": 1.7804,
+      "step": 1667
+    },
+    {
+      "epoch": 0.40529704774632486,
+      "grad_norm": 0.41778072714805603,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4055400315878994,
+      "grad_norm": 0.3932303786277771,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4057830154294739,
+      "grad_norm": 0.41271570324897766,
+      "learning_rate": 8e-05,
+      "loss": 1.6061,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4060259992710485,
+      "grad_norm": 0.44117751717567444,
+      "learning_rate": 8e-05,
+      "loss": 1.5682,
+      "step": 1671
+    },
+    {
+      "epoch": 0.406268983112623,
+      "grad_norm": 0.4247724711894989,
+      "learning_rate": 8e-05,
+      "loss": 1.7314,
+      "step": 1672
+    },
+    {
+      "epoch": 0.40651196695419756,
+      "grad_norm": 0.41147905588150024,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 1673
+    },
+    {
+      "epoch": 0.40675495079577206,
+      "grad_norm": 0.4455786943435669,
+      "learning_rate": 8e-05,
+      "loss": 1.7796,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4069979346373466,
+      "grad_norm": 0.4278455674648285,
+      "learning_rate": 8e-05,
+      "loss": 1.6492,
+      "step": 1675
+    },
+    {
+      "epoch": 0.40724091847892113,
+      "grad_norm": 0.42609670758247375,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4074839023204957,
+      "grad_norm": 0.41262224316596985,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4077268861620702,
+      "grad_norm": 0.4027808904647827,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1678
+    },
+    {
+      "epoch": 0.40796987000364476,
+      "grad_norm": 0.42386099696159363,
+      "learning_rate": 8e-05,
+      "loss": 1.8049,
+      "step": 1679
+    },
+    {
+      "epoch": 0.40821285384521927,
+      "grad_norm": 0.43816202878952026,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 1680
+    },
+    {
+      "epoch": 0.40845583768679383,
+      "grad_norm": 0.4120200276374817,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 1681
+    },
+    {
+      "epoch": 0.40869882152836834,
+      "grad_norm": 0.42139312624931335,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4089418053699429,
+      "grad_norm": 0.44489359855651855,
+      "learning_rate": 8e-05,
+      "loss": 1.8039,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4091847892115174,
+      "grad_norm": 0.400413453578949,
+      "learning_rate": 8e-05,
+      "loss": 1.4418,
+      "step": 1684
+    },
+    {
+      "epoch": 0.40942777305309197,
+      "grad_norm": 0.45639127492904663,
+      "learning_rate": 8e-05,
+      "loss": 1.7726,
+      "step": 1685
+    },
+    {
+      "epoch": 0.40967075689466653,
+      "grad_norm": 0.415913462638855,
+      "learning_rate": 8e-05,
+      "loss": 1.7282,
+      "step": 1686
+    },
+    {
+      "epoch": 0.40991374073624104,
+      "grad_norm": 0.4201204180717468,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4101567245778156,
+      "grad_norm": 0.44380196928977966,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4103997084193901,
+      "grad_norm": 0.4186372458934784,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 1689
+    },
+    {
+      "epoch": 0.41064269226096467,
+      "grad_norm": 0.4141177535057068,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4108856761025392,
+      "grad_norm": 0.4486232101917267,
+      "learning_rate": 8e-05,
+      "loss": 1.8352,
+      "step": 1691
+    },
+    {
+      "epoch": 0.41112865994411374,
+      "grad_norm": 0.41571044921875,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 1692
+    },
+    {
+      "epoch": 0.41137164378568825,
+      "grad_norm": 0.4206002950668335,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4116146276272628,
+      "grad_norm": 0.41686561703681946,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4118576114688373,
+      "grad_norm": 0.41292232275009155,
+      "learning_rate": 8e-05,
+      "loss": 1.6059,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4121005953104119,
+      "grad_norm": 0.4010278880596161,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4123435791519864,
+      "grad_norm": 0.40597307682037354,
+      "learning_rate": 8e-05,
+      "loss": 1.6319,
+      "step": 1697
+    },
+    {
+      "epoch": 0.41258656299356095,
+      "grad_norm": 0.41981416940689087,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 1698
+    },
+    {
+      "epoch": 0.41282954683513545,
+      "grad_norm": 0.4284166991710663,
+      "learning_rate": 8e-05,
+      "loss": 1.7544,
+      "step": 1699
+    },
+    {
+      "epoch": 0.41307253067671,
+      "grad_norm": 0.4121367037296295,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4133155145182845,
+      "grad_norm": 0.40503549575805664,
+      "learning_rate": 8e-05,
+      "loss": 1.7035,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4135584983598591,
+      "grad_norm": 0.4065505564212799,
+      "learning_rate": 8e-05,
+      "loss": 1.7314,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4138014822014336,
+      "grad_norm": 0.397203654050827,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 1703
+    },
+    {
+      "epoch": 0.41404446604300815,
+      "grad_norm": 0.38829803466796875,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 1704
+    },
+    {
+      "epoch": 0.41428744988458266,
+      "grad_norm": 0.42386242747306824,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4145304337261572,
+      "grad_norm": 0.44385597109794617,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 1706
+    },
+    {
+      "epoch": 0.41477341756773173,
+      "grad_norm": 0.39557069540023804,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4150164014093063,
+      "grad_norm": 0.4212467074394226,
+      "learning_rate": 8e-05,
+      "loss": 1.7743,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4152593852508808,
+      "grad_norm": 0.4524325430393219,
+      "learning_rate": 8e-05,
+      "loss": 1.7753,
+      "step": 1709
+    },
+    {
+      "epoch": 0.41550236909245536,
+      "grad_norm": 0.39658069610595703,
+      "learning_rate": 8e-05,
+      "loss": 1.6624,
+      "step": 1710
+    },
+    {
+      "epoch": 0.41574535293402987,
+      "grad_norm": 0.43072929978370667,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 1711
+    },
+    {
+      "epoch": 0.41598833677560443,
+      "grad_norm": 0.4198828935623169,
+      "learning_rate": 8e-05,
+      "loss": 1.7076,
+      "step": 1712
+    },
+    {
+      "epoch": 0.41623132061717893,
+      "grad_norm": 0.4408133029937744,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4164743044587535,
+      "grad_norm": 0.4172877073287964,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 1714
+    },
+    {
+      "epoch": 0.416717288300328,
+      "grad_norm": 0.4034704566001892,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 1715
+    },
+    {
+      "epoch": 0.41696027214190257,
+      "grad_norm": 0.4334101676940918,
+      "learning_rate": 8e-05,
+      "loss": 1.6583,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4172032559834771,
+      "grad_norm": 0.4380352795124054,
+      "learning_rate": 8e-05,
+      "loss": 1.6863,
+      "step": 1717
+    },
+    {
+      "epoch": 0.41744623982505163,
+      "grad_norm": 0.4561653137207031,
+      "learning_rate": 8e-05,
+      "loss": 1.6515,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4176892236666262,
+      "grad_norm": 0.41770467162132263,
+      "learning_rate": 8e-05,
+      "loss": 1.829,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4179322075082007,
+      "grad_norm": 0.4551490843296051,
+      "learning_rate": 8e-05,
+      "loss": 1.9915,
+      "step": 1720
+    },
+    {
+      "epoch": 0.41817519134977527,
+      "grad_norm": 0.4066857397556305,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4184181751913498,
+      "grad_norm": 0.4038218557834625,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 1722
+    },
+    {
+      "epoch": 0.41866115903292433,
+      "grad_norm": 0.4305240213871002,
+      "learning_rate": 8e-05,
+      "loss": 1.7119,
+      "step": 1723
+    },
+    {
+      "epoch": 0.41890414287449884,
+      "grad_norm": 0.4003041386604309,
+      "learning_rate": 8e-05,
+      "loss": 1.5079,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4191471267160734,
+      "grad_norm": 0.39947882294654846,
+      "learning_rate": 8e-05,
+      "loss": 1.6268,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4193901105576479,
+      "grad_norm": 0.43016764521598816,
+      "learning_rate": 8e-05,
+      "loss": 1.5147,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4196330943992225,
+      "grad_norm": 0.4064975082874298,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 1727
+    },
+    {
+      "epoch": 0.419876078240797,
+      "grad_norm": 0.40748926997184753,
+      "learning_rate": 8e-05,
+      "loss": 1.6469,
+      "step": 1728
+    },
+    {
+      "epoch": 0.42011906208237154,
+      "grad_norm": 0.40836912393569946,
+      "learning_rate": 8e-05,
+      "loss": 1.5056,
+      "step": 1729
+    },
+    {
+      "epoch": 0.42036204592394605,
+      "grad_norm": 0.4225461483001709,
+      "learning_rate": 8e-05,
+      "loss": 1.774,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4206050297655206,
+      "grad_norm": 0.4032735824584961,
+      "learning_rate": 8e-05,
+      "loss": 1.6572,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4208480136070951,
+      "grad_norm": 0.44586724042892456,
+      "learning_rate": 8e-05,
+      "loss": 1.7298,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4210909974486697,
+      "grad_norm": 0.4447050094604492,
+      "learning_rate": 8e-05,
+      "loss": 1.7359,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4213339812902442,
+      "grad_norm": 0.4337802231311798,
+      "learning_rate": 8e-05,
+      "loss": 1.7008,
+      "step": 1734
+    },
+    {
+      "epoch": 0.42157696513181875,
+      "grad_norm": 0.4286479353904724,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 1735
+    },
+    {
+      "epoch": 0.42181994897339326,
+      "grad_norm": 0.41877228021621704,
+      "learning_rate": 8e-05,
+      "loss": 1.54,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4220629328149678,
+      "grad_norm": 0.4385734498500824,
+      "learning_rate": 8e-05,
+      "loss": 1.8678,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4223059166565423,
+      "grad_norm": 0.4281429052352905,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4225489004981169,
+      "grad_norm": 0.4258042573928833,
+      "learning_rate": 8e-05,
+      "loss": 1.8062,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4227918843396914,
+      "grad_norm": 0.4337453246116638,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 1740
+    },
+    {
+      "epoch": 0.42303486818126596,
+      "grad_norm": 0.3949694037437439,
+      "learning_rate": 8e-05,
+      "loss": 1.6143,
+      "step": 1741
+    },
+    {
+      "epoch": 0.42327785202284046,
+      "grad_norm": 0.45012426376342773,
+      "learning_rate": 8e-05,
+      "loss": 1.9184,
+      "step": 1742
+    },
+    {
+      "epoch": 0.423520835864415,
+      "grad_norm": 0.44694384932518005,
+      "learning_rate": 8e-05,
+      "loss": 1.7543,
+      "step": 1743
+    },
+    {
+      "epoch": 0.42376381970598953,
+      "grad_norm": 0.4394371211528778,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4240068035475641,
+      "grad_norm": 0.41531896591186523,
+      "learning_rate": 8e-05,
+      "loss": 1.7072,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4242497873891386,
+      "grad_norm": 0.3919975757598877,
+      "learning_rate": 8e-05,
+      "loss": 1.5487,
+      "step": 1746
+    },
+    {
+      "epoch": 0.42449277123071316,
+      "grad_norm": 0.4876977503299713,
+      "learning_rate": 8e-05,
+      "loss": 1.819,
+      "step": 1747
+    },
+    {
+      "epoch": 0.42473575507228767,
+      "grad_norm": 0.4127189517021179,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1748
+    },
+    {
+      "epoch": 0.42497873891386223,
+      "grad_norm": 0.4112125635147095,
+      "learning_rate": 8e-05,
+      "loss": 1.5441,
+      "step": 1749
+    },
+    {
+      "epoch": 0.42522172275543674,
+      "grad_norm": 0.39865005016326904,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4254647065970113,
+      "grad_norm": 0.41315123438835144,
+      "learning_rate": 8e-05,
+      "loss": 1.5477,
+      "step": 1751
+    },
+    {
+      "epoch": 0.42570769043858586,
+      "grad_norm": 0.4037797451019287,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 1752
+    },
+    {
+      "epoch": 0.42595067428016037,
+      "grad_norm": 0.4321102201938629,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 1753
+    },
+    {
+      "epoch": 0.42619365812173493,
+      "grad_norm": 0.434779554605484,
+      "learning_rate": 8e-05,
+      "loss": 1.8621,
+      "step": 1754
+    },
+    {
+      "epoch": 0.42643664196330944,
+      "grad_norm": 0.398060142993927,
+      "learning_rate": 8e-05,
+      "loss": 1.7119,
+      "step": 1755
+    },
+    {
+      "epoch": 0.426679625804884,
+      "grad_norm": 0.4613003432750702,
+      "learning_rate": 8e-05,
+      "loss": 1.7898,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4269226096464585,
+      "grad_norm": 0.4197688698768616,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 1757
+    },
+    {
+      "epoch": 0.42716559348803307,
+      "grad_norm": 0.38842418789863586,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4274085773296076,
+      "grad_norm": 0.42749500274658203,
+      "learning_rate": 8e-05,
+      "loss": 1.751,
+      "step": 1759
+    },
+    {
+      "epoch": 0.42765156117118214,
+      "grad_norm": 0.4103221595287323,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 1760
+    },
+    {
+      "epoch": 0.42789454501275664,
+      "grad_norm": 0.4110293388366699,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4281375288543312,
+      "grad_norm": 0.4347650408744812,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4283805126959057,
+      "grad_norm": 0.42236021161079407,
+      "learning_rate": 8e-05,
+      "loss": 1.6019,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4286234965374803,
+      "grad_norm": 0.4228074252605438,
+      "learning_rate": 8e-05,
+      "loss": 1.7376,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4288664803790548,
+      "grad_norm": 0.43777650594711304,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 1765
+    },
+    {
+      "epoch": 0.42910946422062934,
+      "grad_norm": 0.43366578221321106,
+      "learning_rate": 8e-05,
+      "loss": 1.7939,
+      "step": 1766
+    },
+    {
+      "epoch": 0.42935244806220385,
+      "grad_norm": 0.4141087830066681,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4295954319037784,
+      "grad_norm": 0.40956902503967285,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4298384157453529,
+      "grad_norm": 0.4060921370983124,
+      "learning_rate": 8e-05,
+      "loss": 1.7002,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4300813995869275,
+      "grad_norm": 0.4277002215385437,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 1770
+    },
+    {
+      "epoch": 0.430324383428502,
+      "grad_norm": 0.4249592423439026,
+      "learning_rate": 8e-05,
+      "loss": 1.6289,
+      "step": 1771
+    },
+    {
+      "epoch": 0.43056736727007655,
+      "grad_norm": 0.41829076409339905,
+      "learning_rate": 8e-05,
+      "loss": 1.6168,
+      "step": 1772
+    },
+    {
+      "epoch": 0.43081035111165106,
+      "grad_norm": 0.4147033393383026,
+      "learning_rate": 8e-05,
+      "loss": 1.5973,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4310533349532256,
+      "grad_norm": 0.43540245294570923,
+      "learning_rate": 8e-05,
+      "loss": 1.8268,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4312963187948001,
+      "grad_norm": 0.40885448455810547,
+      "learning_rate": 8e-05,
+      "loss": 1.6481,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4315393026363747,
+      "grad_norm": 0.41620758175849915,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4317822864779492,
+      "grad_norm": 0.4079805314540863,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 1777
+    },
+    {
+      "epoch": 0.43202527031952376,
+      "grad_norm": 0.446587473154068,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 1778
+    },
+    {
+      "epoch": 0.43226825416109826,
+      "grad_norm": 0.43002545833587646,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4325112380026728,
+      "grad_norm": 0.41290929913520813,
+      "learning_rate": 8e-05,
+      "loss": 1.755,
+      "step": 1780
+    },
+    {
+      "epoch": 0.43275422184424733,
+      "grad_norm": 0.4275924265384674,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4329972056858219,
+      "grad_norm": 0.4202406108379364,
+      "learning_rate": 8e-05,
+      "loss": 1.7452,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4332401895273964,
+      "grad_norm": 0.4105161726474762,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 1783
+    },
+    {
+      "epoch": 0.43348317336897096,
+      "grad_norm": 0.4252176284790039,
+      "learning_rate": 8e-05,
+      "loss": 1.678,
+      "step": 1784
+    },
+    {
+      "epoch": 0.43372615721054547,
+      "grad_norm": 0.41096925735473633,
+      "learning_rate": 8e-05,
+      "loss": 1.7689,
+      "step": 1785
+    },
+    {
+      "epoch": 0.43396914105212003,
+      "grad_norm": 0.4047700762748718,
+      "learning_rate": 8e-05,
+      "loss": 1.5749,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4342121248936946,
+      "grad_norm": 0.44662734866142273,
+      "learning_rate": 8e-05,
+      "loss": 1.8453,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4344551087352691,
+      "grad_norm": 0.41828441619873047,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 1788
+    },
+    {
+      "epoch": 0.43469809257684366,
+      "grad_norm": 0.4287610650062561,
+      "learning_rate": 8e-05,
+      "loss": 1.7657,
+      "step": 1789
+    },
+    {
+      "epoch": 0.43494107641841817,
+      "grad_norm": 0.4535541534423828,
+      "learning_rate": 8e-05,
+      "loss": 1.8044,
+      "step": 1790
+    },
+    {
+      "epoch": 0.43518406025999273,
+      "grad_norm": 0.4219343662261963,
+      "learning_rate": 8e-05,
+      "loss": 1.7335,
+      "step": 1791
+    },
+    {
+      "epoch": 0.43542704410156724,
+      "grad_norm": 0.42163732647895813,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4356700279431418,
+      "grad_norm": 0.43030476570129395,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4359130117847163,
+      "grad_norm": 0.4406364858150482,
+      "learning_rate": 8e-05,
+      "loss": 1.8768,
+      "step": 1794
+    },
+    {
+      "epoch": 0.43615599562629087,
+      "grad_norm": 0.4096386730670929,
+      "learning_rate": 8e-05,
+      "loss": 1.6918,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4363989794678654,
+      "grad_norm": 0.418416291475296,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 1796
+    },
+    {
+      "epoch": 0.43664196330943994,
+      "grad_norm": 0.44120216369628906,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 1797
+    },
+    {
+      "epoch": 0.43688494715101445,
+      "grad_norm": 0.41503068804740906,
+      "learning_rate": 8e-05,
+      "loss": 1.5663,
+      "step": 1798
+    },
+    {
+      "epoch": 0.437127930992589,
+      "grad_norm": 0.3758297860622406,
+      "learning_rate": 8e-05,
+      "loss": 1.5509,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4373709148341635,
+      "grad_norm": 0.41770219802856445,
+      "learning_rate": 8e-05,
+      "loss": 1.772,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4376138986757381,
+      "grad_norm": 0.44644829630851746,
+      "learning_rate": 8e-05,
+      "loss": 1.8186,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4378568825173126,
+      "grad_norm": 0.4562210738658905,
+      "learning_rate": 8e-05,
+      "loss": 1.8294,
+      "step": 1802
+    },
+    {
+      "epoch": 0.43809986635888715,
+      "grad_norm": 0.43727368116378784,
+      "learning_rate": 8e-05,
+      "loss": 1.7658,
+      "step": 1803
+    },
+    {
+      "epoch": 0.43834285020046165,
+      "grad_norm": 0.4097563326358795,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4385858340420362,
+      "grad_norm": 0.446963906288147,
+      "learning_rate": 8e-05,
+      "loss": 1.7977,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4388288178836107,
+      "grad_norm": 0.4568128287792206,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4390718017251853,
+      "grad_norm": 0.4310816824436188,
+      "learning_rate": 8e-05,
+      "loss": 1.7771,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4393147855667598,
+      "grad_norm": 0.42967939376831055,
+      "learning_rate": 8e-05,
+      "loss": 1.7093,
+      "step": 1808
+    },
+    {
+      "epoch": 0.43955776940833435,
+      "grad_norm": 0.46501290798187256,
+      "learning_rate": 8e-05,
+      "loss": 1.6964,
+      "step": 1809
+    },
+    {
+      "epoch": 0.43980075324990886,
+      "grad_norm": 0.41305580735206604,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4400437370914834,
+      "grad_norm": 0.43250951170921326,
+      "learning_rate": 8e-05,
+      "loss": 1.7467,
+      "step": 1811
+    },
+    {
+      "epoch": 0.44028672093305793,
+      "grad_norm": 0.41424018144607544,
+      "learning_rate": 8e-05,
+      "loss": 1.652,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4405297047746325,
+      "grad_norm": 0.40682876110076904,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 1813
+    },
+    {
+      "epoch": 0.440772688616207,
+      "grad_norm": 0.4290339946746826,
+      "learning_rate": 8e-05,
+      "loss": 1.7227,
+      "step": 1814
+    },
+    {
+      "epoch": 0.44101567245778156,
+      "grad_norm": 0.4426180124282837,
+      "learning_rate": 8e-05,
+      "loss": 1.6162,
+      "step": 1815
+    },
+    {
+      "epoch": 0.44125865629935607,
+      "grad_norm": 0.4229072332382202,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 1816
+    },
+    {
+      "epoch": 0.44150164014093063,
+      "grad_norm": 0.41639086604118347,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 1817
+    },
+    {
+      "epoch": 0.44174462398250514,
+      "grad_norm": 0.41044601798057556,
+      "learning_rate": 8e-05,
+      "loss": 1.6975,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4419876078240797,
+      "grad_norm": 0.41082099080085754,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 1819
+    },
+    {
+      "epoch": 0.44223059166565426,
+      "grad_norm": 0.4211667478084564,
+      "learning_rate": 8e-05,
+      "loss": 1.7065,
+      "step": 1820
+    },
+    {
+      "epoch": 0.44247357550722877,
+      "grad_norm": 0.41173863410949707,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 1821
+    },
+    {
+      "epoch": 0.44271655934880333,
+      "grad_norm": 0.40657758712768555,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 1822
+    },
+    {
+      "epoch": 0.44295954319037784,
+      "grad_norm": 0.4356505870819092,
+      "learning_rate": 8e-05,
+      "loss": 1.7657,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4432025270319524,
+      "grad_norm": 0.42441052198410034,
+      "learning_rate": 8e-05,
+      "loss": 1.7506,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4434455108735269,
+      "grad_norm": 0.42454221844673157,
+      "learning_rate": 8e-05,
+      "loss": 1.7831,
+      "step": 1825
+    },
+    {
+      "epoch": 0.44368849471510147,
+      "grad_norm": 0.41915127635002136,
+      "learning_rate": 8e-05,
+      "loss": 1.7567,
+      "step": 1826
+    },
+    {
+      "epoch": 0.443931478556676,
+      "grad_norm": 0.39879941940307617,
+      "learning_rate": 8e-05,
+      "loss": 1.6248,
+      "step": 1827
+    },
+    {
+      "epoch": 0.44417446239825054,
+      "grad_norm": 0.43807122111320496,
+      "learning_rate": 8e-05,
+      "loss": 1.7819,
+      "step": 1828
+    },
+    {
+      "epoch": 0.44441744623982504,
+      "grad_norm": 0.4479237496852875,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4446604300813996,
+      "grad_norm": 0.4309643805027008,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4449034139229741,
+      "grad_norm": 0.426151841878891,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4451463977645487,
+      "grad_norm": 0.45159122347831726,
+      "learning_rate": 8e-05,
+      "loss": 1.8247,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4453893816061232,
+      "grad_norm": 0.4200173020362854,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 1833
+    },
+    {
+      "epoch": 0.44563236544769774,
+      "grad_norm": 0.45498377084732056,
+      "learning_rate": 8e-05,
+      "loss": 1.9266,
+      "step": 1834
+    },
+    {
+      "epoch": 0.44587534928927225,
+      "grad_norm": 0.4178752899169922,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4461183331308468,
+      "grad_norm": 0.4127987325191498,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4463613169724213,
+      "grad_norm": 0.3950411379337311,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4466043008139959,
+      "grad_norm": 0.4156026840209961,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4468472846555704,
+      "grad_norm": 0.4114047586917877,
+      "learning_rate": 8e-05,
+      "loss": 1.7528,
+      "step": 1839
+    },
+    {
+      "epoch": 0.44709026849714495,
+      "grad_norm": 0.40903884172439575,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 1840
+    },
+    {
+      "epoch": 0.44733325233871946,
+      "grad_norm": 0.4279289245605469,
+      "learning_rate": 8e-05,
+      "loss": 1.7351,
+      "step": 1841
+    },
+    {
+      "epoch": 0.447576236180294,
+      "grad_norm": 0.4179602861404419,
+      "learning_rate": 8e-05,
+      "loss": 1.647,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4478192200218685,
+      "grad_norm": 0.4204224646091461,
+      "learning_rate": 8e-05,
+      "loss": 1.7088,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4480622038634431,
+      "grad_norm": 0.4532315135002136,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4483051877050176,
+      "grad_norm": 0.42938992381095886,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 1845
+    },
+    {
+      "epoch": 0.44854817154659216,
+      "grad_norm": 0.5924028158187866,
+      "learning_rate": 8e-05,
+      "loss": 1.7198,
+      "step": 1846
+    },
+    {
+      "epoch": 0.44879115538816666,
+      "grad_norm": 0.4050762355327606,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4490341392297412,
+      "grad_norm": 0.4267370402812958,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 1848
+    },
+    {
+      "epoch": 0.44927712307131573,
+      "grad_norm": 0.4041881263256073,
+      "learning_rate": 8e-05,
+      "loss": 1.5839,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4495201069128903,
+      "grad_norm": 0.4267805814743042,
+      "learning_rate": 8e-05,
+      "loss": 1.7747,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4497630907544648,
+      "grad_norm": 0.414093941450119,
+      "learning_rate": 8e-05,
+      "loss": 1.5164,
+      "step": 1851
+    },
+    {
+      "epoch": 0.45000607459603936,
+      "grad_norm": 0.43252500891685486,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4502490584376139,
+      "grad_norm": 0.42252010107040405,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 1853
+    },
+    {
+      "epoch": 0.45049204227918843,
+      "grad_norm": 0.4307325482368469,
+      "learning_rate": 8e-05,
+      "loss": 1.7401,
+      "step": 1854
+    },
+    {
+      "epoch": 0.450735026120763,
+      "grad_norm": 0.41866233944892883,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4509780099623375,
+      "grad_norm": 0.4151945412158966,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 1856
+    },
+    {
+      "epoch": 0.45122099380391206,
+      "grad_norm": 0.4459803104400635,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 1857
+    },
+    {
+      "epoch": 0.45146397764548657,
+      "grad_norm": 0.4404464662075043,
+      "learning_rate": 8e-05,
+      "loss": 1.8168,
+      "step": 1858
+    },
+    {
+      "epoch": 0.45170696148706113,
+      "grad_norm": 0.43660715222358704,
+      "learning_rate": 8e-05,
+      "loss": 1.691,
+      "step": 1859
+    },
+    {
+      "epoch": 0.45194994532863564,
+      "grad_norm": 0.4186517000198364,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4521929291702102,
+      "grad_norm": 0.46955999732017517,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4524359130117847,
+      "grad_norm": 0.3999137878417969,
+      "learning_rate": 8e-05,
+      "loss": 1.5466,
+      "step": 1862
+    },
+    {
+      "epoch": 0.45267889685335927,
+      "grad_norm": 0.39286574721336365,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4529218806949338,
+      "grad_norm": 0.4328494071960449,
+      "learning_rate": 8e-05,
+      "loss": 1.7226,
+      "step": 1864
+    },
+    {
+      "epoch": 0.45316486453650834,
+      "grad_norm": 0.4277777075767517,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 1865
+    },
+    {
+      "epoch": 0.45340784837808285,
+      "grad_norm": 0.4166783392429352,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4536508322196574,
+      "grad_norm": 0.439242959022522,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 1867
+    },
+    {
+      "epoch": 0.4538938160612319,
+      "grad_norm": 0.40764227509498596,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4541367999028065,
+      "grad_norm": 0.4133087992668152,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 1869
+    },
+    {
+      "epoch": 0.454379783744381,
+      "grad_norm": 0.4220365583896637,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 1870
+    },
+    {
+      "epoch": 0.45462276758595555,
+      "grad_norm": 0.4321078360080719,
+      "learning_rate": 8e-05,
+      "loss": 1.8009,
+      "step": 1871
+    },
+    {
+      "epoch": 0.45486575142753005,
+      "grad_norm": 0.43721655011177063,
+      "learning_rate": 8e-05,
+      "loss": 1.7187,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4551087352691046,
+      "grad_norm": 0.43518969416618347,
+      "learning_rate": 8e-05,
+      "loss": 1.8886,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4553517191106791,
+      "grad_norm": 0.4107803702354431,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4555947029522537,
+      "grad_norm": 0.4013316333293915,
+      "learning_rate": 8e-05,
+      "loss": 1.6816,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4558376867938282,
+      "grad_norm": 0.41884687542915344,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 1876
+    },
+    {
+      "epoch": 0.45608067063540275,
+      "grad_norm": 0.4207857549190521,
+      "learning_rate": 8e-05,
+      "loss": 1.8294,
+      "step": 1877
+    },
+    {
+      "epoch": 0.45632365447697726,
+      "grad_norm": 0.4228103458881378,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 1878
+    },
+    {
+      "epoch": 0.4565666383185518,
+      "grad_norm": 0.4231003522872925,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4568096221601263,
+      "grad_norm": 0.45048895478248596,
+      "learning_rate": 8e-05,
+      "loss": 1.6999,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4570526060017009,
+      "grad_norm": 0.4198697507381439,
+      "learning_rate": 8e-05,
+      "loss": 1.7461,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4572955898432754,
+      "grad_norm": 0.4145090878009796,
+      "learning_rate": 8e-05,
+      "loss": 1.5858,
+      "step": 1882
+    },
+    {
+      "epoch": 0.45753857368484996,
+      "grad_norm": 0.442281574010849,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 1883
+    },
+    {
+      "epoch": 0.45778155752642447,
+      "grad_norm": 0.41116663813591003,
+      "learning_rate": 8e-05,
+      "loss": 1.6314,
+      "step": 1884
+    },
+    {
+      "epoch": 0.458024541367999,
+      "grad_norm": 0.42355984449386597,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4582675252095736,
+      "grad_norm": 0.44513821601867676,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4585105090511481,
+      "grad_norm": 0.47635918855667114,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 1887
+    },
+    {
+      "epoch": 0.45875349289272266,
+      "grad_norm": 0.4152633845806122,
+      "learning_rate": 8e-05,
+      "loss": 1.6731,
+      "step": 1888
+    },
+    {
+      "epoch": 0.45899647673429717,
+      "grad_norm": 0.5451321601867676,
+      "learning_rate": 8e-05,
+      "loss": 1.7647,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4592394605758717,
+      "grad_norm": 0.4499322772026062,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 1890
+    },
+    {
+      "epoch": 0.45948244441744623,
+      "grad_norm": 0.42485854029655457,
+      "learning_rate": 8e-05,
+      "loss": 1.7115,
+      "step": 1891
+    },
+    {
+      "epoch": 0.4597254282590208,
+      "grad_norm": 0.4471605122089386,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4599684121005953,
+      "grad_norm": 0.43966439366340637,
+      "learning_rate": 8e-05,
+      "loss": 1.6775,
+      "step": 1893
+    },
+    {
+      "epoch": 0.46021139594216987,
+      "grad_norm": 0.4498927593231201,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 1894
+    },
+    {
+      "epoch": 0.46045437978374437,
+      "grad_norm": 0.4585667550563812,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 1895
+    },
+    {
+      "epoch": 0.46069736362531893,
+      "grad_norm": 0.4273296296596527,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 1896
+    },
+    {
+      "epoch": 0.46094034746689344,
+      "grad_norm": 0.42229199409484863,
+      "learning_rate": 8e-05,
+      "loss": 1.6637,
+      "step": 1897
+    },
+    {
+      "epoch": 0.461183331308468,
+      "grad_norm": 0.4054836928844452,
+      "learning_rate": 8e-05,
+      "loss": 1.641,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4614263151500425,
+      "grad_norm": 0.39605993032455444,
+      "learning_rate": 8e-05,
+      "loss": 1.5741,
+      "step": 1899
+    },
+    {
+      "epoch": 0.46166929899161707,
+      "grad_norm": 0.41373077034950256,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4619122828331916,
+      "grad_norm": 0.4387223720550537,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 1901
+    },
+    {
+      "epoch": 0.46215526667476614,
+      "grad_norm": 0.40724092721939087,
+      "learning_rate": 8e-05,
+      "loss": 1.5232,
+      "step": 1902
+    },
+    {
+      "epoch": 0.46239825051634065,
+      "grad_norm": 0.4248490631580353,
+      "learning_rate": 8e-05,
+      "loss": 1.7604,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4626412343579152,
+      "grad_norm": 0.4261873960494995,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4628842181994897,
+      "grad_norm": 0.405947208404541,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4631272020410643,
+      "grad_norm": 0.43242183327674866,
+      "learning_rate": 8e-05,
+      "loss": 1.7289,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4633701858826388,
+      "grad_norm": 0.42454537749290466,
+      "learning_rate": 8e-05,
+      "loss": 1.7662,
+      "step": 1907
+    },
+    {
+      "epoch": 0.46361316972421335,
+      "grad_norm": 0.43956276774406433,
+      "learning_rate": 8e-05,
+      "loss": 1.7926,
+      "step": 1908
+    },
+    {
+      "epoch": 0.46385615356578785,
+      "grad_norm": 0.4338242709636688,
+      "learning_rate": 8e-05,
+      "loss": 1.8531,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4640991374073624,
+      "grad_norm": 0.40287795662879944,
+      "learning_rate": 8e-05,
+      "loss": 1.7194,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4643421212489369,
+      "grad_norm": 0.40767234563827515,
+      "learning_rate": 8e-05,
+      "loss": 1.6324,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4645851050905115,
+      "grad_norm": 0.4035324156284332,
+      "learning_rate": 8e-05,
+      "loss": 1.6616,
+      "step": 1912
+    },
+    {
+      "epoch": 0.464828088932086,
+      "grad_norm": 0.43699395656585693,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 1913
+    },
+    {
+      "epoch": 0.46507107277366055,
+      "grad_norm": 0.41658124327659607,
+      "learning_rate": 8e-05,
+      "loss": 1.6999,
+      "step": 1914
+    },
+    {
+      "epoch": 0.46531405661523506,
+      "grad_norm": 0.41529208421707153,
+      "learning_rate": 8e-05,
+      "loss": 1.7192,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4655570404568096,
+      "grad_norm": 0.4511304795742035,
+      "learning_rate": 8e-05,
+      "loss": 1.8472,
+      "step": 1916
+    },
+    {
+      "epoch": 0.46580002429838413,
+      "grad_norm": 0.470687597990036,
+      "learning_rate": 8e-05,
+      "loss": 1.9462,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4660430081399587,
+      "grad_norm": 0.41925495862960815,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1918
+    },
+    {
+      "epoch": 0.46628599198153325,
+      "grad_norm": 0.41691452264785767,
+      "learning_rate": 8e-05,
+      "loss": 1.5893,
+      "step": 1919
+    },
+    {
+      "epoch": 0.46652897582310776,
+      "grad_norm": 0.44769784808158875,
+      "learning_rate": 8e-05,
+      "loss": 1.6974,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4667719596646823,
+      "grad_norm": 0.45075252652168274,
+      "learning_rate": 8e-05,
+      "loss": 1.9332,
+      "step": 1921
+    },
+    {
+      "epoch": 0.46701494350625683,
+      "grad_norm": 0.42499783635139465,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4672579273478314,
+      "grad_norm": 0.4107694923877716,
+      "learning_rate": 8e-05,
+      "loss": 1.7142,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4675009111894059,
+      "grad_norm": 0.386578232049942,
+      "learning_rate": 8e-05,
+      "loss": 1.5295,
+      "step": 1924
+    },
+    {
+      "epoch": 0.46774389503098046,
+      "grad_norm": 0.39867669343948364,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 1925
+    },
+    {
+      "epoch": 0.46798687887255497,
+      "grad_norm": 0.43957293033599854,
+      "learning_rate": 8e-05,
+      "loss": 1.7702,
+      "step": 1926
+    },
+    {
+      "epoch": 0.46822986271412953,
+      "grad_norm": 0.4263380169868469,
+      "learning_rate": 8e-05,
+      "loss": 1.5876,
+      "step": 1927
+    },
+    {
+      "epoch": 0.46847284655570404,
+      "grad_norm": 0.4032500982284546,
+      "learning_rate": 8e-05,
+      "loss": 1.5244,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4687158303972786,
+      "grad_norm": 0.42151403427124023,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4689588142388531,
+      "grad_norm": 0.4025190770626068,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 1930
+    },
+    {
+      "epoch": 0.46920179808042767,
+      "grad_norm": 0.4104991853237152,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4694447819220022,
+      "grad_norm": 0.41498640179634094,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 1932
+    },
+    {
+      "epoch": 0.46968776576357674,
+      "grad_norm": 0.426425576210022,
+      "learning_rate": 8e-05,
+      "loss": 1.7485,
+      "step": 1933
+    },
+    {
+      "epoch": 0.46993074960515124,
+      "grad_norm": 0.4150788187980652,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4701737334467258,
+      "grad_norm": 0.3989652693271637,
+      "learning_rate": 8e-05,
+      "loss": 1.5937,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4704167172883003,
+      "grad_norm": 0.42093151807785034,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4706597011298749,
+      "grad_norm": 0.4280698299407959,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4709026849714494,
+      "grad_norm": 0.4426701068878174,
+      "learning_rate": 8e-05,
+      "loss": 1.8398,
+      "step": 1938
+    },
+    {
+      "epoch": 0.47114566881302394,
+      "grad_norm": 0.4073024392127991,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 1939
+    },
+    {
+      "epoch": 0.47138865265459845,
+      "grad_norm": 0.4276946783065796,
+      "learning_rate": 8e-05,
+      "loss": 1.7999,
+      "step": 1940
+    },
+    {
+      "epoch": 0.471631636496173,
+      "grad_norm": 0.3996224105358124,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4718746203377475,
+      "grad_norm": 0.4300505220890045,
+      "learning_rate": 8e-05,
+      "loss": 1.8833,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4721176041793221,
+      "grad_norm": 0.3979191482067108,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4723605880208966,
+      "grad_norm": 0.44216442108154297,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 1944
+    },
+    {
+      "epoch": 0.47260357186247115,
+      "grad_norm": 0.4506628215312958,
+      "learning_rate": 8e-05,
+      "loss": 1.8527,
+      "step": 1945
+    },
+    {
+      "epoch": 0.47284655570404566,
+      "grad_norm": 0.42636045813560486,
+      "learning_rate": 8e-05,
+      "loss": 1.6937,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4730895395456202,
+      "grad_norm": 0.44184669852256775,
+      "learning_rate": 8e-05,
+      "loss": 1.8348,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4733325233871947,
+      "grad_norm": 0.4105511009693146,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4735755072287693,
+      "grad_norm": 0.449083149433136,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4738184910703438,
+      "grad_norm": 0.43708378076553345,
+      "learning_rate": 8e-05,
+      "loss": 1.7683,
+      "step": 1950
+    },
+    {
+      "epoch": 0.47406147491191836,
+      "grad_norm": 0.5191806554794312,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4743044587534929,
+      "grad_norm": 0.42893365025520325,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4745474425950674,
+      "grad_norm": 0.5133644342422485,
+      "learning_rate": 8e-05,
+      "loss": 1.6852,
+      "step": 1953
+    },
+    {
+      "epoch": 0.474790426436642,
+      "grad_norm": 0.42633509635925293,
+      "learning_rate": 8e-05,
+      "loss": 1.6697,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4750334102782165,
+      "grad_norm": 0.4043295383453369,
+      "learning_rate": 8e-05,
+      "loss": 1.5707,
+      "step": 1955
+    },
+    {
+      "epoch": 0.47527639411979106,
+      "grad_norm": 0.4424762725830078,
+      "learning_rate": 8e-05,
+      "loss": 1.7666,
+      "step": 1956
+    },
+    {
+      "epoch": 0.47551937796136556,
+      "grad_norm": 0.4177377223968506,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4757623618029401,
+      "grad_norm": 0.4313500225543976,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 1958
+    },
+    {
+      "epoch": 0.47600534564451463,
+      "grad_norm": 0.4116400182247162,
+      "learning_rate": 8e-05,
+      "loss": 1.6701,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4762483294860892,
+      "grad_norm": 0.43881407380104065,
+      "learning_rate": 8e-05,
+      "loss": 1.8674,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4764913133276637,
+      "grad_norm": 0.4389816224575043,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 1961
+    },
+    {
+      "epoch": 0.47673429716923826,
+      "grad_norm": 0.41141510009765625,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 1962
+    },
+    {
+      "epoch": 0.47697728101081277,
+      "grad_norm": 0.45347926020622253,
+      "learning_rate": 8e-05,
+      "loss": 1.7938,
+      "step": 1963
+    },
+    {
+      "epoch": 0.47722026485238733,
+      "grad_norm": 0.44667941331863403,
+      "learning_rate": 8e-05,
+      "loss": 1.6149,
+      "step": 1964
+    },
+    {
+      "epoch": 0.47746324869396184,
+      "grad_norm": 0.427518367767334,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4777062325355364,
+      "grad_norm": 0.4317607283592224,
+      "learning_rate": 8e-05,
+      "loss": 1.7672,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4779492163771109,
+      "grad_norm": 0.46663787961006165,
+      "learning_rate": 8e-05,
+      "loss": 1.7708,
+      "step": 1967
+    },
+    {
+      "epoch": 0.47819220021868547,
+      "grad_norm": 0.44258251786231995,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 1968
+    },
+    {
+      "epoch": 0.47843518406026,
+      "grad_norm": 0.4117269814014435,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 1969
+    },
+    {
+      "epoch": 0.47867816790183454,
+      "grad_norm": 0.4820593595504761,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 1970
+    },
+    {
+      "epoch": 0.47892115174340905,
+      "grad_norm": 0.4440200626850128,
+      "learning_rate": 8e-05,
+      "loss": 1.7023,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4791641355849836,
+      "grad_norm": 0.4371243715286255,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4794071194265581,
+      "grad_norm": 0.4171777069568634,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4796501032681327,
+      "grad_norm": 0.4147702753543854,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4798930871097072,
+      "grad_norm": 0.4282766878604889,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 1975
+    },
+    {
+      "epoch": 0.48013607095128175,
+      "grad_norm": 0.4245712459087372,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 1976
+    },
+    {
+      "epoch": 0.48037905479285625,
+      "grad_norm": 0.427929550409317,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4806220386344308,
+      "grad_norm": 0.4307423532009125,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4808650224760053,
+      "grad_norm": 0.41224634647369385,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4811080063175799,
+      "grad_norm": 0.4373134970664978,
+      "learning_rate": 8e-05,
+      "loss": 1.8249,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4813509901591544,
+      "grad_norm": 0.41638660430908203,
+      "learning_rate": 8e-05,
+      "loss": 1.697,
+      "step": 1981
+    },
+    {
+      "epoch": 0.48159397400072895,
+      "grad_norm": 0.40082883834838867,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 1982
+    },
+    {
+      "epoch": 0.48183695784230346,
+      "grad_norm": 0.4369618594646454,
+      "learning_rate": 8e-05,
+      "loss": 1.763,
+      "step": 1983
+    },
+    {
+      "epoch": 0.482079941683878,
+      "grad_norm": 0.4087012708187103,
+      "learning_rate": 8e-05,
+      "loss": 1.5982,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4823229255254526,
+      "grad_norm": 0.408522367477417,
+      "learning_rate": 8e-05,
+      "loss": 1.6309,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4825659093670271,
+      "grad_norm": 0.44090622663497925,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 1986
+    },
+    {
+      "epoch": 0.48280889320860165,
+      "grad_norm": 0.4151834547519684,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 1987
+    },
+    {
+      "epoch": 0.48305187705017616,
+      "grad_norm": 0.44949987530708313,
+      "learning_rate": 8e-05,
+      "loss": 1.8235,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4832948608917507,
+      "grad_norm": 0.42424458265304565,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 1989
+    },
+    {
+      "epoch": 0.48353784473332523,
+      "grad_norm": 0.45849746465682983,
+      "learning_rate": 8e-05,
+      "loss": 1.8579,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4837808285748998,
+      "grad_norm": 0.4297718405723572,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4840238124164743,
+      "grad_norm": 0.4386619031429291,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 1992
+    },
+    {
+      "epoch": 0.48426679625804886,
+      "grad_norm": 0.4270291030406952,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 1993
+    },
+    {
+      "epoch": 0.48450978009962337,
+      "grad_norm": 0.4325200319290161,
+      "learning_rate": 8e-05,
+      "loss": 1.5759,
+      "step": 1994
+    },
+    {
+      "epoch": 0.48475276394119793,
+      "grad_norm": 0.43285664916038513,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 1995
+    },
+    {
+      "epoch": 0.48499574778277244,
+      "grad_norm": 0.4182582199573517,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 1996
+    },
+    {
+      "epoch": 0.485238731624347,
+      "grad_norm": 0.4355359673500061,
+      "learning_rate": 8e-05,
+      "loss": 1.7369,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4854817154659215,
+      "grad_norm": 0.4323694705963135,
+      "learning_rate": 8e-05,
+      "loss": 1.7227,
+      "step": 1998
+    },
+    {
+      "epoch": 0.48572469930749607,
+      "grad_norm": 0.40371614694595337,
+      "learning_rate": 8e-05,
+      "loss": 1.7561,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4859676831490706,
+      "grad_norm": 0.4079720675945282,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4115,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.81251563290624e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}