diff --git "a/checkpoint-5000/trainer_state.json" "b/checkpoint-5000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-5000/trainer_state.json"
@@ -0,0 +1,35033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2150406997934637,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "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
+    },
+    {
+      "epoch": 0.48621066699064514,
+      "grad_norm": 0.4067677855491638,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 2001
+    },
+    {
+      "epoch": 0.48645365083221964,
+      "grad_norm": 0.42427200078964233,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4866966346737942,
+      "grad_norm": 0.4287988543510437,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4869396185153687,
+      "grad_norm": 0.4463859498500824,
+      "learning_rate": 8e-05,
+      "loss": 1.7649,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4871826023569433,
+      "grad_norm": 0.42050695419311523,
+      "learning_rate": 8e-05,
+      "loss": 1.7177,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4874255861985178,
+      "grad_norm": 0.4367334544658661,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 2006
+    },
+    {
+      "epoch": 0.48766857004009234,
+      "grad_norm": 0.4281326234340668,
+      "learning_rate": 8e-05,
+      "loss": 1.8229,
+      "step": 2007
+    },
+    {
+      "epoch": 0.48791155388166685,
+      "grad_norm": 0.4476991295814514,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4881545377232414,
+      "grad_norm": 0.4020079970359802,
+      "learning_rate": 8e-05,
+      "loss": 1.5858,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4883975215648159,
+      "grad_norm": 0.5277190208435059,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4886405054063905,
+      "grad_norm": 0.4083438217639923,
+      "learning_rate": 8e-05,
+      "loss": 1.6041,
+      "step": 2011
+    },
+    {
+      "epoch": 0.488883489247965,
+      "grad_norm": 0.3972097933292389,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 2012
+    },
+    {
+      "epoch": 0.48912647308953955,
+      "grad_norm": 0.40560653805732727,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 2013
+    },
+    {
+      "epoch": 0.48936945693111406,
+      "grad_norm": 0.41275861859321594,
+      "learning_rate": 8e-05,
+      "loss": 1.7268,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4896124407726886,
+      "grad_norm": 0.4319611191749573,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4898554246142631,
+      "grad_norm": 0.4500797986984253,
+      "learning_rate": 8e-05,
+      "loss": 1.8008,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4900984084558377,
+      "grad_norm": 0.43417975306510925,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 2017
+    },
+    {
+      "epoch": 0.49034139229741225,
+      "grad_norm": 0.44237813353538513,
+      "learning_rate": 8e-05,
+      "loss": 1.8711,
+      "step": 2018
+    },
+    {
+      "epoch": 0.49058437613898676,
+      "grad_norm": 0.4223446249961853,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4908273599805613,
+      "grad_norm": 0.4409749209880829,
+      "learning_rate": 8e-05,
+      "loss": 1.831,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4910703438221358,
+      "grad_norm": 0.39405280351638794,
+      "learning_rate": 8e-05,
+      "loss": 1.5467,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4913133276637104,
+      "grad_norm": 0.4100813865661621,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4915563115052849,
+      "grad_norm": 0.43508127331733704,
+      "learning_rate": 8e-05,
+      "loss": 1.8118,
+      "step": 2023
+    },
+    {
+      "epoch": 0.49179929534685946,
+      "grad_norm": 0.4335186183452606,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 2024
+    },
+    {
+      "epoch": 0.49204227918843396,
+      "grad_norm": 0.4278884828090668,
+      "learning_rate": 8e-05,
+      "loss": 1.7866,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4922852630300085,
+      "grad_norm": 0.4198608994483948,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 2026
+    },
+    {
+      "epoch": 0.49252824687158303,
+      "grad_norm": 0.44235336780548096,
+      "learning_rate": 8e-05,
+      "loss": 1.6914,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4927712307131576,
+      "grad_norm": 0.4087814688682556,
+      "learning_rate": 8e-05,
+      "loss": 1.6539,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4930142145547321,
+      "grad_norm": 0.44606995582580566,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 2029
+    },
+    {
+      "epoch": 0.49325719839630666,
+      "grad_norm": 0.41591018438339233,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 2030
+    },
+    {
+      "epoch": 0.49350018223788117,
+      "grad_norm": 0.4187583029270172,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 2031
+    },
+    {
+      "epoch": 0.49374316607945573,
+      "grad_norm": 0.4315899610519409,
+      "learning_rate": 8e-05,
+      "loss": 1.7002,
+      "step": 2032
+    },
+    {
+      "epoch": 0.49398614992103024,
+      "grad_norm": 0.4039921164512634,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4942291337626048,
+      "grad_norm": 0.4449801743030548,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4944721176041793,
+      "grad_norm": 0.4388439655303955,
+      "learning_rate": 8e-05,
+      "loss": 1.5253,
+      "step": 2035
+    },
+    {
+      "epoch": 0.49471510144575387,
+      "grad_norm": 0.42010697722435,
+      "learning_rate": 8e-05,
+      "loss": 1.7391,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4949580852873284,
+      "grad_norm": 0.40051358938217163,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 2037
+    },
+    {
+      "epoch": 0.49520106912890294,
+      "grad_norm": 0.4096914231777191,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 2038
+    },
+    {
+      "epoch": 0.49544405297047744,
+      "grad_norm": 0.4256300926208496,
+      "learning_rate": 8e-05,
+      "loss": 1.5667,
+      "step": 2039
+    },
+    {
+      "epoch": 0.495687036812052,
+      "grad_norm": 0.41908881068229675,
+      "learning_rate": 8e-05,
+      "loss": 1.6786,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4959300206536265,
+      "grad_norm": 0.4249550998210907,
+      "learning_rate": 8e-05,
+      "loss": 1.7477,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4961730044952011,
+      "grad_norm": 0.4326035678386688,
+      "learning_rate": 8e-05,
+      "loss": 1.7515,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4964159883367756,
+      "grad_norm": 0.39607805013656616,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 2043
+    },
+    {
+      "epoch": 0.49665897217835014,
+      "grad_norm": 0.4403810203075409,
+      "learning_rate": 8e-05,
+      "loss": 1.8264,
+      "step": 2044
+    },
+    {
+      "epoch": 0.49690195601992465,
+      "grad_norm": 0.4221819043159485,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4971449398614992,
+      "grad_norm": 0.42985081672668457,
+      "learning_rate": 8e-05,
+      "loss": 1.7506,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4973879237030737,
+      "grad_norm": 0.428672194480896,
+      "learning_rate": 8e-05,
+      "loss": 1.7116,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4976309075446483,
+      "grad_norm": 0.42397749423980713,
+      "learning_rate": 8e-05,
+      "loss": 1.8315,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4978738913862228,
+      "grad_norm": 0.40653225779533386,
+      "learning_rate": 8e-05,
+      "loss": 1.5583,
+      "step": 2049
+    },
+    {
+      "epoch": 0.49811687522779735,
+      "grad_norm": 0.4188259541988373,
+      "learning_rate": 8e-05,
+      "loss": 1.775,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4983598590693719,
+      "grad_norm": 0.42173391580581665,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4986028429109464,
+      "grad_norm": 0.424783855676651,
+      "learning_rate": 8e-05,
+      "loss": 1.8405,
+      "step": 2052
+    },
+    {
+      "epoch": 0.498845826752521,
+      "grad_norm": 0.4394426941871643,
+      "learning_rate": 8e-05,
+      "loss": 1.6298,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4990888105940955,
+      "grad_norm": 0.4615374505519867,
+      "learning_rate": 8e-05,
+      "loss": 1.8431,
+      "step": 2054
+    },
+    {
+      "epoch": 0.49933179443567005,
+      "grad_norm": 0.42615365982055664,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 2055
+    },
+    {
+      "epoch": 0.49957477827724456,
+      "grad_norm": 0.42110222578048706,
+      "learning_rate": 8e-05,
+      "loss": 1.7516,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4998177621188191,
+      "grad_norm": 0.4481637179851532,
+      "learning_rate": 8e-05,
+      "loss": 1.8262,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5000607459603936,
+      "grad_norm": 0.4159661829471588,
+      "learning_rate": 8e-05,
+      "loss": 1.7217,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5003037298019681,
+      "grad_norm": 0.4300229549407959,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5005467136435428,
+      "grad_norm": 0.41252321004867554,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5007896974851173,
+      "grad_norm": 0.4080057740211487,
+      "learning_rate": 8e-05,
+      "loss": 1.6099,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5010326813266918,
+      "grad_norm": 0.43958574533462524,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5012756651682663,
+      "grad_norm": 0.41116398572921753,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5015186490098409,
+      "grad_norm": 0.42529818415641785,
+      "learning_rate": 8e-05,
+      "loss": 1.6823,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5017616328514154,
+      "grad_norm": 0.42211756110191345,
+      "learning_rate": 8e-05,
+      "loss": 1.6732,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5020046166929899,
+      "grad_norm": 0.4227226674556732,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5022476005345644,
+      "grad_norm": 0.4149601459503174,
+      "learning_rate": 8e-05,
+      "loss": 1.7683,
+      "step": 2067
+    },
+    {
+      "epoch": 0.502490584376139,
+      "grad_norm": 0.41580531001091003,
+      "learning_rate": 8e-05,
+      "loss": 1.7407,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5027335682177135,
+      "grad_norm": 0.4171891510486603,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 2069
+    },
+    {
+      "epoch": 0.502976552059288,
+      "grad_norm": 0.46244126558303833,
+      "learning_rate": 8e-05,
+      "loss": 1.7397,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5032195359008625,
+      "grad_norm": 0.4327504634857178,
+      "learning_rate": 8e-05,
+      "loss": 1.7112,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5034625197424372,
+      "grad_norm": 0.4318760335445404,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5037055035840117,
+      "grad_norm": 0.4132522642612457,
+      "learning_rate": 8e-05,
+      "loss": 1.5867,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5039484874255862,
+      "grad_norm": 0.44223716855049133,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5041914712671607,
+      "grad_norm": 0.4249269366264343,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5044344551087353,
+      "grad_norm": 0.42366063594818115,
+      "learning_rate": 8e-05,
+      "loss": 1.7786,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5046774389503098,
+      "grad_norm": 0.40331336855888367,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5049204227918843,
+      "grad_norm": 0.4033096134662628,
+      "learning_rate": 8e-05,
+      "loss": 1.5853,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5051634066334588,
+      "grad_norm": 0.4134315550327301,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5054063904750334,
+      "grad_norm": 0.43576759099960327,
+      "learning_rate": 8e-05,
+      "loss": 1.8256,
+      "step": 2080
+    },
+    {
+      "epoch": 0.505649374316608,
+      "grad_norm": 0.41187170147895813,
+      "learning_rate": 8e-05,
+      "loss": 1.6962,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5058923581581825,
+      "grad_norm": 0.44993796944618225,
+      "learning_rate": 8e-05,
+      "loss": 1.8513,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5061353419997571,
+      "grad_norm": 0.42347627878189087,
+      "learning_rate": 8e-05,
+      "loss": 1.7441,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5063783258413316,
+      "grad_norm": 0.42224588990211487,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5066213096829061,
+      "grad_norm": 0.42265740036964417,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5068642935244806,
+      "grad_norm": 0.4238162636756897,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5071072773660552,
+      "grad_norm": 0.4217587411403656,
+      "learning_rate": 8e-05,
+      "loss": 1.7298,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5073502612076297,
+      "grad_norm": 0.40629303455352783,
+      "learning_rate": 8e-05,
+      "loss": 1.5576,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5075932450492042,
+      "grad_norm": 0.4425939619541168,
+      "learning_rate": 8e-05,
+      "loss": 1.8597,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5078362288907787,
+      "grad_norm": 0.46037763357162476,
+      "learning_rate": 8e-05,
+      "loss": 1.9015,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5080792127323533,
+      "grad_norm": 0.435923308134079,
+      "learning_rate": 8e-05,
+      "loss": 1.6172,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5083221965739279,
+      "grad_norm": 0.4095153510570526,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5085651804155024,
+      "grad_norm": 0.4168741703033447,
+      "learning_rate": 8e-05,
+      "loss": 1.6067,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5088081642570769,
+      "grad_norm": 0.4301334023475647,
+      "learning_rate": 8e-05,
+      "loss": 1.5698,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5090511480986515,
+      "grad_norm": 0.41287705302238464,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 2095
+    },
+    {
+      "epoch": 0.509294131940226,
+      "grad_norm": 0.44751450419425964,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5095371157818005,
+      "grad_norm": 0.4126018285751343,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 2097
+    },
+    {
+      "epoch": 0.509780099623375,
+      "grad_norm": 0.4485846757888794,
+      "learning_rate": 8e-05,
+      "loss": 1.7399,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5100230834649496,
+      "grad_norm": 0.45301029086112976,
+      "learning_rate": 8e-05,
+      "loss": 1.7872,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5102660673065241,
+      "grad_norm": 0.42355573177337646,
+      "learning_rate": 8e-05,
+      "loss": 1.6792,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5105090511480986,
+      "grad_norm": 0.42952632904052734,
+      "learning_rate": 8e-05,
+      "loss": 1.6962,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5107520349896731,
+      "grad_norm": 0.4603313207626343,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5109950188312478,
+      "grad_norm": 0.4149731695652008,
+      "learning_rate": 8e-05,
+      "loss": 1.5566,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5112380026728223,
+      "grad_norm": 0.42397817969322205,
+      "learning_rate": 8e-05,
+      "loss": 1.8187,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5114809865143968,
+      "grad_norm": 0.4256420135498047,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5117239703559713,
+      "grad_norm": 0.4320630729198456,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5119669541975459,
+      "grad_norm": 0.45823419094085693,
+      "learning_rate": 8e-05,
+      "loss": 1.7925,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5122099380391204,
+      "grad_norm": 0.43793827295303345,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5124529218806949,
+      "grad_norm": 0.415707528591156,
+      "learning_rate": 8e-05,
+      "loss": 1.7277,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5126959057222694,
+      "grad_norm": 0.4325178861618042,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 2110
+    },
+    {
+      "epoch": 0.512938889563844,
+      "grad_norm": 0.4661019742488861,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5131818734054185,
+      "grad_norm": 0.43143412470817566,
+      "learning_rate": 8e-05,
+      "loss": 1.8569,
+      "step": 2112
+    },
+    {
+      "epoch": 0.513424857246993,
+      "grad_norm": 0.4136514663696289,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5136678410885676,
+      "grad_norm": 0.4750009775161743,
+      "learning_rate": 8e-05,
+      "loss": 1.6042,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5139108249301422,
+      "grad_norm": 0.41810303926467896,
+      "learning_rate": 8e-05,
+      "loss": 1.7655,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5141538087717167,
+      "grad_norm": 0.4353654682636261,
+      "learning_rate": 8e-05,
+      "loss": 1.8453,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5143967926132912,
+      "grad_norm": 0.45224910974502563,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5146397764548658,
+      "grad_norm": 0.4224438965320587,
+      "learning_rate": 8e-05,
+      "loss": 1.6636,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5148827602964403,
+      "grad_norm": 0.42598116397857666,
+      "learning_rate": 8e-05,
+      "loss": 1.6627,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5151257441380148,
+      "grad_norm": 0.4862707555294037,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5153687279795893,
+      "grad_norm": 0.44141459465026855,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5156117118211639,
+      "grad_norm": 0.4234525263309479,
+      "learning_rate": 8e-05,
+      "loss": 1.6515,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5158546956627384,
+      "grad_norm": 0.42503517866134644,
+      "learning_rate": 8e-05,
+      "loss": 1.6863,
+      "step": 2123
+    },
+    {
+      "epoch": 0.516097679504313,
+      "grad_norm": 0.43979698419570923,
+      "learning_rate": 8e-05,
+      "loss": 1.7135,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5163406633458875,
+      "grad_norm": 0.42391321063041687,
+      "learning_rate": 8e-05,
+      "loss": 1.7751,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5165836471874621,
+      "grad_norm": 0.4411396086215973,
+      "learning_rate": 8e-05,
+      "loss": 1.8186,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5168266310290366,
+      "grad_norm": 0.4332398474216461,
+      "learning_rate": 8e-05,
+      "loss": 1.7627,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5170696148706111,
+      "grad_norm": 0.45441141724586487,
+      "learning_rate": 8e-05,
+      "loss": 1.7027,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5173125987121856,
+      "grad_norm": 0.42690712213516235,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5175555825537602,
+      "grad_norm": 0.44065365195274353,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5177985663953347,
+      "grad_norm": 0.4382827877998352,
+      "learning_rate": 8e-05,
+      "loss": 1.788,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5180415502369092,
+      "grad_norm": 0.4443352222442627,
+      "learning_rate": 8e-05,
+      "loss": 1.5818,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5182845340784837,
+      "grad_norm": 0.41484344005584717,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5185275179200584,
+      "grad_norm": 0.4568868577480316,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5187705017616329,
+      "grad_norm": 0.4405900835990906,
+      "learning_rate": 8e-05,
+      "loss": 1.8059,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5190134856032074,
+      "grad_norm": 0.4328583776950836,
+      "learning_rate": 8e-05,
+      "loss": 1.7569,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5192564694447819,
+      "grad_norm": 0.4324393570423126,
+      "learning_rate": 8e-05,
+      "loss": 1.7312,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5194994532863565,
+      "grad_norm": 0.4428865909576416,
+      "learning_rate": 8e-05,
+      "loss": 1.6585,
+      "step": 2138
+    },
+    {
+      "epoch": 0.519742437127931,
+      "grad_norm": 0.43051040172576904,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5199854209695055,
+      "grad_norm": 0.4162987470626831,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 2140
+    },
+    {
+      "epoch": 0.52022840481108,
+      "grad_norm": 0.44039788842201233,
+      "learning_rate": 8e-05,
+      "loss": 1.7926,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5204713886526546,
+      "grad_norm": 0.41748008131980896,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5207143724942291,
+      "grad_norm": 0.40175241231918335,
+      "learning_rate": 8e-05,
+      "loss": 1.4881,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5209573563358036,
+      "grad_norm": 0.47015997767448425,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5212003401773782,
+      "grad_norm": 0.4361039102077484,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5214433240189528,
+      "grad_norm": 0.41764092445373535,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5216863078605273,
+      "grad_norm": 0.4543766379356384,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5219292917021018,
+      "grad_norm": 0.451860636472702,
+      "learning_rate": 8e-05,
+      "loss": 1.5975,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5221722755436764,
+      "grad_norm": 0.42808955907821655,
+      "learning_rate": 8e-05,
+      "loss": 1.8195,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5224152593852509,
+      "grad_norm": 0.4308494031429291,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5226582432268254,
+      "grad_norm": 0.42901811003685,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5229012270683999,
+      "grad_norm": 0.45172786712646484,
+      "learning_rate": 8e-05,
+      "loss": 1.8201,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5231442109099745,
+      "grad_norm": 0.42531678080558777,
+      "learning_rate": 8e-05,
+      "loss": 1.8085,
+      "step": 2153
+    },
+    {
+      "epoch": 0.523387194751549,
+      "grad_norm": 0.4457677900791168,
+      "learning_rate": 8e-05,
+      "loss": 1.745,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5236301785931236,
+      "grad_norm": 0.4440343379974365,
+      "learning_rate": 8e-05,
+      "loss": 1.7508,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5238731624346981,
+      "grad_norm": 0.4284655451774597,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5241161462762727,
+      "grad_norm": 0.39911216497421265,
+      "learning_rate": 8e-05,
+      "loss": 1.5078,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5243591301178472,
+      "grad_norm": 0.472358763217926,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5246021139594217,
+      "grad_norm": 0.45089733600616455,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5248450978009962,
+      "grad_norm": 0.5574978590011597,
+      "learning_rate": 8e-05,
+      "loss": 1.7546,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5250880816425708,
+      "grad_norm": 0.4636007249355316,
+      "learning_rate": 8e-05,
+      "loss": 1.7547,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5253310654841453,
+      "grad_norm": 0.4396299123764038,
+      "learning_rate": 8e-05,
+      "loss": 1.6925,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5255740493257198,
+      "grad_norm": 0.4171754717826843,
+      "learning_rate": 8e-05,
+      "loss": 1.7085,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5258170331672943,
+      "grad_norm": 0.48295915126800537,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 2164
+    },
+    {
+      "epoch": 0.526060017008869,
+      "grad_norm": 0.45748254656791687,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5263030008504435,
+      "grad_norm": 0.464437335729599,
+      "learning_rate": 8e-05,
+      "loss": 1.8113,
+      "step": 2166
+    },
+    {
+      "epoch": 0.526545984692018,
+      "grad_norm": 0.5038710832595825,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5267889685335925,
+      "grad_norm": 0.43993282318115234,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5270319523751671,
+      "grad_norm": 0.45278286933898926,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5272749362167416,
+      "grad_norm": 0.4331422746181488,
+      "learning_rate": 8e-05,
+      "loss": 1.85,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5275179200583161,
+      "grad_norm": 0.4202103316783905,
+      "learning_rate": 8e-05,
+      "loss": 1.6456,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5277609038998906,
+      "grad_norm": 0.450734943151474,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5280038877414652,
+      "grad_norm": 0.44593900442123413,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5282468715830397,
+      "grad_norm": 0.4255177676677704,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5284898554246142,
+      "grad_norm": 0.4444335699081421,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5287328392661887,
+      "grad_norm": 0.44298461079597473,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5289758231077634,
+      "grad_norm": 0.43717876076698303,
+      "learning_rate": 8e-05,
+      "loss": 1.7975,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5292188069493379,
+      "grad_norm": 0.4377862811088562,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5294617907909124,
+      "grad_norm": 0.4273555278778076,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5297047746324869,
+      "grad_norm": 0.44019001722335815,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5299477584740615,
+      "grad_norm": 0.41915464401245117,
+      "learning_rate": 8e-05,
+      "loss": 1.6117,
+      "step": 2181
+    },
+    {
+      "epoch": 0.530190742315636,
+      "grad_norm": 0.4198468029499054,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5304337261572105,
+      "grad_norm": 0.44378358125686646,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5306767099987851,
+      "grad_norm": 0.4340839982032776,
+      "learning_rate": 8e-05,
+      "loss": 1.703,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5309196938403596,
+      "grad_norm": 0.45091456174850464,
+      "learning_rate": 8e-05,
+      "loss": 1.7635,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5311626776819341,
+      "grad_norm": 0.420701801776886,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5314056615235087,
+      "grad_norm": 0.4314432442188263,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5316486453650833,
+      "grad_norm": 0.4679161012172699,
+      "learning_rate": 8e-05,
+      "loss": 1.7367,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5318916292066578,
+      "grad_norm": 0.4167570173740387,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5321346130482323,
+      "grad_norm": 0.49057233333587646,
+      "learning_rate": 8e-05,
+      "loss": 1.7376,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5323775968898068,
+      "grad_norm": 0.45474299788475037,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5326205807313814,
+      "grad_norm": 0.4404684007167816,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5328635645729559,
+      "grad_norm": 0.40527915954589844,
+      "learning_rate": 8e-05,
+      "loss": 1.5696,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5331065484145304,
+      "grad_norm": 0.43138939142227173,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5333495322561049,
+      "grad_norm": 0.436314195394516,
+      "learning_rate": 8e-05,
+      "loss": 1.7466,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5335925160976795,
+      "grad_norm": 0.4222269058227539,
+      "learning_rate": 8e-05,
+      "loss": 1.6579,
+      "step": 2196
+    },
+    {
+      "epoch": 0.533835499939254,
+      "grad_norm": 0.4563753604888916,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5340784837808286,
+      "grad_norm": 0.4227203130722046,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5343214676224031,
+      "grad_norm": 0.4319281280040741,
+      "learning_rate": 8e-05,
+      "loss": 1.7583,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5345644514639777,
+      "grad_norm": 0.4201965928077698,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5348074353055522,
+      "grad_norm": 0.45120948553085327,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5350504191471267,
+      "grad_norm": 0.42846375703811646,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5352934029887012,
+      "grad_norm": 0.434852659702301,
+      "learning_rate": 8e-05,
+      "loss": 1.7648,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5355363868302758,
+      "grad_norm": 0.43229424953460693,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5357793706718503,
+      "grad_norm": 0.4069499671459198,
+      "learning_rate": 8e-05,
+      "loss": 1.674,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5360223545134248,
+      "grad_norm": 0.4106616973876953,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5362653383549993,
+      "grad_norm": 0.41744673252105713,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 2207
+    },
+    {
+      "epoch": 0.536508322196574,
+      "grad_norm": 0.4047800898551941,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5367513060381485,
+      "grad_norm": 0.44462546706199646,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 2209
+    },
+    {
+      "epoch": 0.536994289879723,
+      "grad_norm": 0.45628994703292847,
+      "learning_rate": 8e-05,
+      "loss": 1.8221,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5372372737212975,
+      "grad_norm": 0.4441223740577698,
+      "learning_rate": 8e-05,
+      "loss": 1.7519,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5374802575628721,
+      "grad_norm": 0.44681769609451294,
+      "learning_rate": 8e-05,
+      "loss": 1.778,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5377232414044466,
+      "grad_norm": 0.45145052671432495,
+      "learning_rate": 8e-05,
+      "loss": 1.6996,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5379662252460211,
+      "grad_norm": 0.4367946982383728,
+      "learning_rate": 8e-05,
+      "loss": 1.7484,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5382092090875957,
+      "grad_norm": 0.42780929803848267,
+      "learning_rate": 8e-05,
+      "loss": 1.8319,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5384521929291702,
+      "grad_norm": 0.42155957221984863,
+      "learning_rate": 8e-05,
+      "loss": 1.6683,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5386951767707447,
+      "grad_norm": 0.4169459044933319,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5389381606123193,
+      "grad_norm": 0.4357196092605591,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5391811444538939,
+      "grad_norm": 0.41908493638038635,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5394241282954684,
+      "grad_norm": 0.442971795797348,
+      "learning_rate": 8e-05,
+      "loss": 1.7417,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5396671121370429,
+      "grad_norm": 0.40678754448890686,
+      "learning_rate": 8e-05,
+      "loss": 1.584,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5399100959786174,
+      "grad_norm": 0.46350914239883423,
+      "learning_rate": 8e-05,
+      "loss": 1.7753,
+      "step": 2222
+    },
+    {
+      "epoch": 0.540153079820192,
+      "grad_norm": 0.4378567934036255,
+      "learning_rate": 8e-05,
+      "loss": 1.7361,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5403960636617665,
+      "grad_norm": 0.4252557158470154,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 2224
+    },
+    {
+      "epoch": 0.540639047503341,
+      "grad_norm": 0.4332224428653717,
+      "learning_rate": 8e-05,
+      "loss": 1.675,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5408820313449155,
+      "grad_norm": 0.42885562777519226,
+      "learning_rate": 8e-05,
+      "loss": 1.7691,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5411250151864901,
+      "grad_norm": 0.4333006739616394,
+      "learning_rate": 8e-05,
+      "loss": 1.7016,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5413679990280647,
+      "grad_norm": 0.4079436957836151,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5416109828696392,
+      "grad_norm": 0.4096442759037018,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5418539667112137,
+      "grad_norm": 0.4148697555065155,
+      "learning_rate": 8e-05,
+      "loss": 1.6398,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5420969505527883,
+      "grad_norm": 0.44697436690330505,
+      "learning_rate": 8e-05,
+      "loss": 1.7731,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5423399343943628,
+      "grad_norm": 0.4449930191040039,
+      "learning_rate": 8e-05,
+      "loss": 1.8066,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5425829182359373,
+      "grad_norm": 0.4291306734085083,
+      "learning_rate": 8e-05,
+      "loss": 1.7369,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5428259020775118,
+      "grad_norm": 0.43089258670806885,
+      "learning_rate": 8e-05,
+      "loss": 1.8702,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5430688859190864,
+      "grad_norm": 0.4485584795475006,
+      "learning_rate": 8e-05,
+      "loss": 1.7447,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5433118697606609,
+      "grad_norm": 0.4498457908630371,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5435548536022354,
+      "grad_norm": 0.4413444697856903,
+      "learning_rate": 8e-05,
+      "loss": 1.6148,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5437978374438099,
+      "grad_norm": 0.42494112253189087,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5440408212853846,
+      "grad_norm": 0.4272076189517975,
+      "learning_rate": 8e-05,
+      "loss": 1.5895,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5442838051269591,
+      "grad_norm": 0.42755207419395447,
+      "learning_rate": 8e-05,
+      "loss": 1.7665,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5445267889685336,
+      "grad_norm": 0.4191838800907135,
+      "learning_rate": 8e-05,
+      "loss": 1.5821,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5447697728101081,
+      "grad_norm": 0.40706369280815125,
+      "learning_rate": 8e-05,
+      "loss": 1.6577,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5450127566516827,
+      "grad_norm": 0.4298412501811981,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5452557404932572,
+      "grad_norm": 0.43807289004325867,
+      "learning_rate": 8e-05,
+      "loss": 1.8956,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5454987243348317,
+      "grad_norm": 0.430306077003479,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5457417081764062,
+      "grad_norm": 0.44380131363868713,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5459846920179808,
+      "grad_norm": 0.41122177243232727,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5462276758595553,
+      "grad_norm": 0.4258728325366974,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5464706597011298,
+      "grad_norm": 0.4211385250091553,
+      "learning_rate": 8e-05,
+      "loss": 1.7448,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5467136435427045,
+      "grad_norm": 0.41565054655075073,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 2250
+    },
+    {
+      "epoch": 0.546956627384279,
+      "grad_norm": 0.46474114060401917,
+      "learning_rate": 8e-05,
+      "loss": 1.7761,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5471996112258535,
+      "grad_norm": 0.41311749815940857,
+      "learning_rate": 8e-05,
+      "loss": 1.5347,
+      "step": 2252
+    },
+    {
+      "epoch": 0.547442595067428,
+      "grad_norm": 0.41779565811157227,
+      "learning_rate": 8e-05,
+      "loss": 1.5277,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5476855789090026,
+      "grad_norm": 0.4209821820259094,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5479285627505771,
+      "grad_norm": 0.4146195352077484,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5481715465921516,
+      "grad_norm": 0.4297250807285309,
+      "learning_rate": 8e-05,
+      "loss": 1.6377,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5484145304337261,
+      "grad_norm": 0.41996052861213684,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5486575142753007,
+      "grad_norm": 0.4262883961200714,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5489004981168752,
+      "grad_norm": 0.46587684750556946,
+      "learning_rate": 8e-05,
+      "loss": 1.756,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5491434819584498,
+      "grad_norm": 0.4192856252193451,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5493864658000243,
+      "grad_norm": 0.4181446433067322,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5496294496415989,
+      "grad_norm": 0.4725036025047302,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5498724334831734,
+      "grad_norm": 0.40109264850616455,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5501154173247479,
+      "grad_norm": 0.426868200302124,
+      "learning_rate": 8e-05,
+      "loss": 1.6949,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5503584011663224,
+      "grad_norm": 0.41913172602653503,
+      "learning_rate": 8e-05,
+      "loss": 1.6287,
+      "step": 2265
+    },
+    {
+      "epoch": 0.550601385007897,
+      "grad_norm": 0.4026455283164978,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5508443688494715,
+      "grad_norm": 0.4266178011894226,
+      "learning_rate": 8e-05,
+      "loss": 1.7352,
+      "step": 2267
+    },
+    {
+      "epoch": 0.551087352691046,
+      "grad_norm": 0.4161139130592346,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5513303365326205,
+      "grad_norm": 0.4157556891441345,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5515733203741952,
+      "grad_norm": 0.434096097946167,
+      "learning_rate": 8e-05,
+      "loss": 1.7711,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5518163042157697,
+      "grad_norm": 0.4041610062122345,
+      "learning_rate": 8e-05,
+      "loss": 1.3591,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5520592880573442,
+      "grad_norm": 0.43938177824020386,
+      "learning_rate": 8e-05,
+      "loss": 1.8242,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5523022718989187,
+      "grad_norm": 0.42942848801612854,
+      "learning_rate": 8e-05,
+      "loss": 1.7418,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5525452557404933,
+      "grad_norm": 0.4261589050292969,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5527882395820678,
+      "grad_norm": 0.42738568782806396,
+      "learning_rate": 8e-05,
+      "loss": 1.776,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5530312234236423,
+      "grad_norm": 0.41965100169181824,
+      "learning_rate": 8e-05,
+      "loss": 1.7852,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5532742072652168,
+      "grad_norm": 0.41479334235191345,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5535171911067914,
+      "grad_norm": 0.4221387207508087,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5537601749483659,
+      "grad_norm": 0.43190494179725647,
+      "learning_rate": 8e-05,
+      "loss": 1.7375,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5540031587899404,
+      "grad_norm": 0.41732457280158997,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5542461426315151,
+      "grad_norm": 0.41478410363197327,
+      "learning_rate": 8e-05,
+      "loss": 1.726,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5544891264730896,
+      "grad_norm": 0.4239521026611328,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5547321103146641,
+      "grad_norm": 0.435709148645401,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5549750941562386,
+      "grad_norm": 0.41457170248031616,
+      "learning_rate": 8e-05,
+      "loss": 1.6399,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5552180779978132,
+      "grad_norm": 0.44268688559532166,
+      "learning_rate": 8e-05,
+      "loss": 1.8558,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5554610618393877,
+      "grad_norm": 0.44362714886665344,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5557040456809622,
+      "grad_norm": 0.4304722845554352,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5559470295225367,
+      "grad_norm": 0.40472716093063354,
+      "learning_rate": 8e-05,
+      "loss": 1.6467,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5561900133641113,
+      "grad_norm": 0.42298266291618347,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5564329972056858,
+      "grad_norm": 0.4108637869358063,
+      "learning_rate": 8e-05,
+      "loss": 1.6849,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5566759810472603,
+      "grad_norm": 0.42652082443237305,
+      "learning_rate": 8e-05,
+      "loss": 1.7884,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5569189648888349,
+      "grad_norm": 0.4225563406944275,
+      "learning_rate": 8e-05,
+      "loss": 1.6253,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5571619487304095,
+      "grad_norm": 0.4448477625846863,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 2293
+    },
+    {
+      "epoch": 0.557404932571984,
+      "grad_norm": 0.4309808611869812,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5576479164135585,
+      "grad_norm": 0.41815412044525146,
+      "learning_rate": 8e-05,
+      "loss": 1.5367,
+      "step": 2295
+    },
+    {
+      "epoch": 0.557890900255133,
+      "grad_norm": 0.4293038845062256,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5581338840967076,
+      "grad_norm": 0.4345656633377075,
+      "learning_rate": 8e-05,
+      "loss": 1.7053,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5583768679382821,
+      "grad_norm": 0.408460408449173,
+      "learning_rate": 8e-05,
+      "loss": 1.606,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5586198517798566,
+      "grad_norm": 0.4121434688568115,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5588628356214311,
+      "grad_norm": 0.42664140462875366,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5591058194630057,
+      "grad_norm": 0.42563116550445557,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5593488033045803,
+      "grad_norm": 0.4168783128261566,
+      "learning_rate": 8e-05,
+      "loss": 1.5279,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5595917871461548,
+      "grad_norm": 0.4008963108062744,
+      "learning_rate": 8e-05,
+      "loss": 1.533,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5598347709877293,
+      "grad_norm": 0.3969338834285736,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5600777548293039,
+      "grad_norm": 0.40849944949150085,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5603207386708784,
+      "grad_norm": 0.4392300546169281,
+      "learning_rate": 8e-05,
+      "loss": 1.8035,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5605637225124529,
+      "grad_norm": 0.42521360516548157,
+      "learning_rate": 8e-05,
+      "loss": 1.7565,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5608067063540274,
+      "grad_norm": 0.4263232350349426,
+      "learning_rate": 8e-05,
+      "loss": 1.6666,
+      "step": 2308
+    },
+    {
+      "epoch": 0.561049690195602,
+      "grad_norm": 0.41891422867774963,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5612926740371765,
+      "grad_norm": 0.4274829626083374,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 2310
+    },
+    {
+      "epoch": 0.561535657878751,
+      "grad_norm": 0.4220094382762909,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5617786417203255,
+      "grad_norm": 0.4187576174736023,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5620216255619002,
+      "grad_norm": 0.4324800670146942,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5622646094034747,
+      "grad_norm": 0.4349084198474884,
+      "learning_rate": 8e-05,
+      "loss": 1.7288,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5625075932450492,
+      "grad_norm": 0.45561450719833374,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5627505770866238,
+      "grad_norm": 0.4237786829471588,
+      "learning_rate": 8e-05,
+      "loss": 1.7117,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5629935609281983,
+      "grad_norm": 0.41226062178611755,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5632365447697728,
+      "grad_norm": 0.41445454955101013,
+      "learning_rate": 8e-05,
+      "loss": 1.6895,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5634795286113473,
+      "grad_norm": 0.4144866466522217,
+      "learning_rate": 8e-05,
+      "loss": 1.6924,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5637225124529219,
+      "grad_norm": 0.4397876262664795,
+      "learning_rate": 8e-05,
+      "loss": 1.7511,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5639654962944964,
+      "grad_norm": 0.4511023759841919,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 2321
+    },
+    {
+      "epoch": 0.564208480136071,
+      "grad_norm": 0.41083458065986633,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5644514639776455,
+      "grad_norm": 0.41399556398391724,
+      "learning_rate": 8e-05,
+      "loss": 1.5223,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5646944478192201,
+      "grad_norm": 0.4346249997615814,
+      "learning_rate": 8e-05,
+      "loss": 1.6987,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5649374316607946,
+      "grad_norm": 0.43831732869148254,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5651804155023691,
+      "grad_norm": 0.4231765866279602,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5654233993439436,
+      "grad_norm": 0.4603574872016907,
+      "learning_rate": 8e-05,
+      "loss": 1.9103,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5656663831855182,
+      "grad_norm": 0.4295811057090759,
+      "learning_rate": 8e-05,
+      "loss": 1.7513,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5659093670270927,
+      "grad_norm": 0.412395179271698,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5661523508686672,
+      "grad_norm": 0.418568879365921,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5663953347102417,
+      "grad_norm": 0.4386887550354004,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5666383185518163,
+      "grad_norm": 0.43578046560287476,
+      "learning_rate": 8e-05,
+      "loss": 1.652,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5668813023933909,
+      "grad_norm": 0.43578222393989563,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5671242862349654,
+      "grad_norm": 0.42693188786506653,
+      "learning_rate": 8e-05,
+      "loss": 1.6273,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5673672700765399,
+      "grad_norm": 0.4326976537704468,
+      "learning_rate": 8e-05,
+      "loss": 1.7058,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5676102539181145,
+      "grad_norm": 0.4334565997123718,
+      "learning_rate": 8e-05,
+      "loss": 1.7088,
+      "step": 2336
+    },
+    {
+      "epoch": 0.567853237759689,
+      "grad_norm": 0.4255800247192383,
+      "learning_rate": 8e-05,
+      "loss": 1.777,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5680962216012635,
+      "grad_norm": 0.4347940981388092,
+      "learning_rate": 8e-05,
+      "loss": 1.7433,
+      "step": 2338
+    },
+    {
+      "epoch": 0.568339205442838,
+      "grad_norm": 0.4247983694076538,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5685821892844126,
+      "grad_norm": 0.44486117362976074,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5688251731259871,
+      "grad_norm": 0.41745704412460327,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5690681569675616,
+      "grad_norm": 0.43471214175224304,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5693111408091361,
+      "grad_norm": 0.41162189841270447,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5695541246507108,
+      "grad_norm": 0.427015095949173,
+      "learning_rate": 8e-05,
+      "loss": 1.7251,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5697971084922853,
+      "grad_norm": 0.4154822528362274,
+      "learning_rate": 8e-05,
+      "loss": 1.7248,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5700400923338598,
+      "grad_norm": 0.4321574866771698,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5702830761754344,
+      "grad_norm": 0.43180641531944275,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5705260600170089,
+      "grad_norm": 0.4145722985267639,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5707690438585834,
+      "grad_norm": 0.4477742314338684,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5710120277001579,
+      "grad_norm": 0.44234955310821533,
+      "learning_rate": 8e-05,
+      "loss": 1.6225,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5712550115417325,
+      "grad_norm": 0.43275707960128784,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 2351
+    },
+    {
+      "epoch": 0.571497995383307,
+      "grad_norm": 0.4226607382297516,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5717409792248815,
+      "grad_norm": 0.4446256458759308,
+      "learning_rate": 8e-05,
+      "loss": 1.8143,
+      "step": 2353
+    },
+    {
+      "epoch": 0.571983963066456,
+      "grad_norm": 0.42211976647377014,
+      "learning_rate": 8e-05,
+      "loss": 1.7168,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5722269469080307,
+      "grad_norm": 0.4262460768222809,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5724699307496052,
+      "grad_norm": 0.44543465971946716,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5727129145911797,
+      "grad_norm": 0.46873605251312256,
+      "learning_rate": 8e-05,
+      "loss": 1.8398,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5729558984327542,
+      "grad_norm": 0.41115307807922363,
+      "learning_rate": 8e-05,
+      "loss": 1.5915,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5731988822743288,
+      "grad_norm": 0.4282917082309723,
+      "learning_rate": 8e-05,
+      "loss": 1.7388,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5734418661159033,
+      "grad_norm": 0.41556891798973083,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5736848499574778,
+      "grad_norm": 0.42523348331451416,
+      "learning_rate": 8e-05,
+      "loss": 1.6314,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5739278337990523,
+      "grad_norm": 0.44694674015045166,
+      "learning_rate": 8e-05,
+      "loss": 1.7248,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5741708176406269,
+      "grad_norm": 0.41888970136642456,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5744138014822014,
+      "grad_norm": 0.4501434862613678,
+      "learning_rate": 8e-05,
+      "loss": 1.8617,
+      "step": 2364
+    },
+    {
+      "epoch": 0.574656785323776,
+      "grad_norm": 0.41189295053482056,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5748997691653505,
+      "grad_norm": 0.4356091320514679,
+      "learning_rate": 8e-05,
+      "loss": 1.7326,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5751427530069251,
+      "grad_norm": 0.4309876263141632,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5753857368484996,
+      "grad_norm": 0.4038586914539337,
+      "learning_rate": 8e-05,
+      "loss": 1.6726,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5756287206900741,
+      "grad_norm": 0.4412710666656494,
+      "learning_rate": 8e-05,
+      "loss": 1.8463,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5758717045316486,
+      "grad_norm": 0.40325218439102173,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5761146883732232,
+      "grad_norm": 0.430065393447876,
+      "learning_rate": 8e-05,
+      "loss": 1.5624,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5763576722147977,
+      "grad_norm": 0.41245678067207336,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5766006560563722,
+      "grad_norm": 0.4184976816177368,
+      "learning_rate": 8e-05,
+      "loss": 1.7117,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5768436398979467,
+      "grad_norm": 0.43841052055358887,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5770866237395214,
+      "grad_norm": 0.45614078640937805,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5773296075810959,
+      "grad_norm": 0.43804284930229187,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5775725914226704,
+      "grad_norm": 0.4361899495124817,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5778155752642449,
+      "grad_norm": 0.4476282298564911,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5780585591058195,
+      "grad_norm": 0.42004987597465515,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 2379
+    },
+    {
+      "epoch": 0.578301542947394,
+      "grad_norm": 0.42843756079673767,
+      "learning_rate": 8e-05,
+      "loss": 1.6565,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5785445267889685,
+      "grad_norm": 0.4132315516471863,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5787875106305431,
+      "grad_norm": 0.4317929148674011,
+      "learning_rate": 8e-05,
+      "loss": 1.7078,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5790304944721176,
+      "grad_norm": 0.4433698058128357,
+      "learning_rate": 8e-05,
+      "loss": 1.7943,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5792734783136921,
+      "grad_norm": 0.4094293713569641,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5795164621552666,
+      "grad_norm": 0.41789159178733826,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5797594459968413,
+      "grad_norm": 0.4471970200538635,
+      "learning_rate": 8e-05,
+      "loss": 1.8259,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5800024298384158,
+      "grad_norm": 0.42537564039230347,
+      "learning_rate": 8e-05,
+      "loss": 1.7668,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5802454136799903,
+      "grad_norm": 0.4170583486557007,
+      "learning_rate": 8e-05,
+      "loss": 1.6746,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5804883975215648,
+      "grad_norm": 0.4144807457923889,
+      "learning_rate": 8e-05,
+      "loss": 1.5514,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5807313813631394,
+      "grad_norm": 0.4408377707004547,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5809743652047139,
+      "grad_norm": 0.4207456409931183,
+      "learning_rate": 8e-05,
+      "loss": 1.7122,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5812173490462884,
+      "grad_norm": 0.4438258707523346,
+      "learning_rate": 8e-05,
+      "loss": 1.5765,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5814603328878629,
+      "grad_norm": 0.39592891931533813,
+      "learning_rate": 8e-05,
+      "loss": 1.5903,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5817033167294375,
+      "grad_norm": 0.4536016285419464,
+      "learning_rate": 8e-05,
+      "loss": 1.6733,
+      "step": 2394
+    },
+    {
+      "epoch": 0.581946300571012,
+      "grad_norm": 0.4351329207420349,
+      "learning_rate": 8e-05,
+      "loss": 1.639,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5821892844125865,
+      "grad_norm": 0.4078857898712158,
+      "learning_rate": 8e-05,
+      "loss": 1.6014,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5824322682541611,
+      "grad_norm": 0.4060279428958893,
+      "learning_rate": 8e-05,
+      "loss": 1.6871,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5826752520957357,
+      "grad_norm": 0.41745471954345703,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5829182359373102,
+      "grad_norm": 0.4796887934207916,
+      "learning_rate": 8e-05,
+      "loss": 1.6946,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5831612197788847,
+      "grad_norm": 0.4440170228481293,
+      "learning_rate": 8e-05,
+      "loss": 1.7422,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5834042036204592,
+      "grad_norm": 0.41775283217430115,
+      "learning_rate": 8e-05,
+      "loss": 1.7525,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5836471874620338,
+      "grad_norm": 0.4444386065006256,
+      "learning_rate": 8e-05,
+      "loss": 1.793,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5838901713036083,
+      "grad_norm": 0.41325974464416504,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5841331551451828,
+      "grad_norm": 0.43009766936302185,
+      "learning_rate": 8e-05,
+      "loss": 1.7568,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5843761389867573,
+      "grad_norm": 0.43368902802467346,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 2405
+    },
+    {
+      "epoch": 0.584619122828332,
+      "grad_norm": 0.425471693277359,
+      "learning_rate": 8e-05,
+      "loss": 1.7077,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5848621066699065,
+      "grad_norm": 0.4581122100353241,
+      "learning_rate": 8e-05,
+      "loss": 1.8081,
+      "step": 2407
+    },
+    {
+      "epoch": 0.585105090511481,
+      "grad_norm": 0.41706401109695435,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5853480743530555,
+      "grad_norm": 0.46168407797813416,
+      "learning_rate": 8e-05,
+      "loss": 1.6429,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5855910581946301,
+      "grad_norm": 0.440673828125,
+      "learning_rate": 8e-05,
+      "loss": 1.7059,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5858340420362046,
+      "grad_norm": 0.4305659830570221,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5860770258777791,
+      "grad_norm": 0.40800726413726807,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5863200097193536,
+      "grad_norm": 0.43414685130119324,
+      "learning_rate": 8e-05,
+      "loss": 1.7174,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5865629935609282,
+      "grad_norm": 0.42639219760894775,
+      "learning_rate": 8e-05,
+      "loss": 1.7039,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5868059774025027,
+      "grad_norm": 0.4156681001186371,
+      "learning_rate": 8e-05,
+      "loss": 1.5807,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5870489612440772,
+      "grad_norm": 0.40939226746559143,
+      "learning_rate": 8e-05,
+      "loss": 1.518,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5872919450856519,
+      "grad_norm": 0.4400941729545593,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5875349289272264,
+      "grad_norm": 0.42164963483810425,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5877779127688009,
+      "grad_norm": 0.4202132821083069,
+      "learning_rate": 8e-05,
+      "loss": 1.6606,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5880208966103754,
+      "grad_norm": 0.4152732491493225,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 2420
+    },
+    {
+      "epoch": 0.58826388045195,
+      "grad_norm": 0.4240267276763916,
+      "learning_rate": 8e-05,
+      "loss": 1.7374,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5885068642935245,
+      "grad_norm": 0.4398930072784424,
+      "learning_rate": 8e-05,
+      "loss": 1.7702,
+      "step": 2422
+    },
+    {
+      "epoch": 0.588749848135099,
+      "grad_norm": 0.4304884076118469,
+      "learning_rate": 8e-05,
+      "loss": 1.7784,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5889928319766735,
+      "grad_norm": 0.4243101477622986,
+      "learning_rate": 8e-05,
+      "loss": 1.7612,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5892358158182481,
+      "grad_norm": 0.4187234044075012,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5894787996598226,
+      "grad_norm": 0.41048288345336914,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5897217835013971,
+      "grad_norm": 0.4052138328552246,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5899647673429717,
+      "grad_norm": 0.434247225522995,
+      "learning_rate": 8e-05,
+      "loss": 1.7819,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5902077511845463,
+      "grad_norm": 0.46596571803092957,
+      "learning_rate": 8e-05,
+      "loss": 1.8564,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5904507350261208,
+      "grad_norm": 0.44565877318382263,
+      "learning_rate": 8e-05,
+      "loss": 1.6353,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5906937188676953,
+      "grad_norm": 0.4320891201496124,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5909367027092698,
+      "grad_norm": 0.42095208168029785,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5911796865508444,
+      "grad_norm": 0.4223933517932892,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5914226703924189,
+      "grad_norm": 0.41740289330482483,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5916656542339934,
+      "grad_norm": 0.43194350600242615,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5919086380755679,
+      "grad_norm": 0.4439544975757599,
+      "learning_rate": 8e-05,
+      "loss": 1.6439,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5921516219171425,
+      "grad_norm": 0.43901556730270386,
+      "learning_rate": 8e-05,
+      "loss": 1.6926,
+      "step": 2437
+    },
+    {
+      "epoch": 0.592394605758717,
+      "grad_norm": 0.42874231934547424,
+      "learning_rate": 8e-05,
+      "loss": 1.7017,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5926375896002916,
+      "grad_norm": 0.4368149936199188,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5928805734418661,
+      "grad_norm": 0.43485620617866516,
+      "learning_rate": 8e-05,
+      "loss": 1.8416,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5931235572834407,
+      "grad_norm": 0.4173147678375244,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5933665411250152,
+      "grad_norm": 0.44863858819007874,
+      "learning_rate": 8e-05,
+      "loss": 1.7654,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5936095249665897,
+      "grad_norm": 0.4292725622653961,
+      "learning_rate": 8e-05,
+      "loss": 1.4832,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5938525088081642,
+      "grad_norm": 0.43171802163124084,
+      "learning_rate": 8e-05,
+      "loss": 1.7029,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5940954926497388,
+      "grad_norm": 0.44493812322616577,
+      "learning_rate": 8e-05,
+      "loss": 1.8242,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5943384764913133,
+      "grad_norm": 0.432722806930542,
+      "learning_rate": 8e-05,
+      "loss": 1.7377,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5945814603328878,
+      "grad_norm": 0.43153059482574463,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5948244441744625,
+      "grad_norm": 0.41324567794799805,
+      "learning_rate": 8e-05,
+      "loss": 1.5993,
+      "step": 2448
+    },
+    {
+      "epoch": 0.595067428016037,
+      "grad_norm": 0.45020240545272827,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5953104118576115,
+      "grad_norm": 0.4177494943141937,
+      "learning_rate": 8e-05,
+      "loss": 1.6298,
+      "step": 2450
+    },
+    {
+      "epoch": 0.595553395699186,
+      "grad_norm": 0.4175281226634979,
+      "learning_rate": 8e-05,
+      "loss": 1.6733,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5957963795407606,
+      "grad_norm": 0.42487746477127075,
+      "learning_rate": 8e-05,
+      "loss": 1.6801,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5960393633823351,
+      "grad_norm": 0.42074668407440186,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5962823472239096,
+      "grad_norm": 0.435325026512146,
+      "learning_rate": 8e-05,
+      "loss": 1.7868,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5965253310654841,
+      "grad_norm": 0.4328100085258484,
+      "learning_rate": 8e-05,
+      "loss": 1.7692,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5967683149070587,
+      "grad_norm": 0.45602723956108093,
+      "learning_rate": 8e-05,
+      "loss": 1.6837,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5970112987486332,
+      "grad_norm": 0.42551445960998535,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5972542825902077,
+      "grad_norm": 0.5649949908256531,
+      "learning_rate": 8e-05,
+      "loss": 1.8318,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5974972664317822,
+      "grad_norm": 0.4395221471786499,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5977402502733569,
+      "grad_norm": 0.44669994711875916,
+      "learning_rate": 8e-05,
+      "loss": 1.8846,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5979832341149314,
+      "grad_norm": 0.450961172580719,
+      "learning_rate": 8e-05,
+      "loss": 1.9402,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5982262179565059,
+      "grad_norm": 0.4219827651977539,
+      "learning_rate": 8e-05,
+      "loss": 1.5673,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5984692017980804,
+      "grad_norm": 0.4143693149089813,
+      "learning_rate": 8e-05,
+      "loss": 1.5848,
+      "step": 2463
+    },
+    {
+      "epoch": 0.598712185639655,
+      "grad_norm": 0.44943028688430786,
+      "learning_rate": 8e-05,
+      "loss": 1.7874,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5989551694812295,
+      "grad_norm": 0.41902869939804077,
+      "learning_rate": 8e-05,
+      "loss": 1.5856,
+      "step": 2465
+    },
+    {
+      "epoch": 0.599198153322804,
+      "grad_norm": 0.4259614050388336,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5994411371643785,
+      "grad_norm": 0.4326181411743164,
+      "learning_rate": 8e-05,
+      "loss": 1.7645,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5996841210059531,
+      "grad_norm": 0.48122575879096985,
+      "learning_rate": 8e-05,
+      "loss": 1.7447,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5999271048475276,
+      "grad_norm": 0.43446168303489685,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 2469
+    },
+    {
+      "epoch": 0.6001700886891022,
+      "grad_norm": 0.41794902086257935,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6004130725306767,
+      "grad_norm": 0.4562852382659912,
+      "learning_rate": 8e-05,
+      "loss": 1.8351,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6006560563722513,
+      "grad_norm": 0.41896143555641174,
+      "learning_rate": 8e-05,
+      "loss": 1.6621,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6008990402138258,
+      "grad_norm": 0.4381026029586792,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6011420240554003,
+      "grad_norm": 0.4462568461894989,
+      "learning_rate": 8e-05,
+      "loss": 1.7402,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6013850078969748,
+      "grad_norm": 0.4301934540271759,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6016279917385494,
+      "grad_norm": 0.42239949107170105,
+      "learning_rate": 8e-05,
+      "loss": 1.6468,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6018709755801239,
+      "grad_norm": 0.4313346743583679,
+      "learning_rate": 8e-05,
+      "loss": 1.6929,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6021139594216984,
+      "grad_norm": 0.4273393452167511,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6023569432632729,
+      "grad_norm": 0.41722846031188965,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6025999271048476,
+      "grad_norm": 0.44882383942604065,
+      "learning_rate": 8e-05,
+      "loss": 1.72,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6028429109464221,
+      "grad_norm": 0.4344826638698578,
+      "learning_rate": 8e-05,
+      "loss": 1.8675,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6030858947879966,
+      "grad_norm": 0.43989020586013794,
+      "learning_rate": 8e-05,
+      "loss": 1.7202,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6033288786295712,
+      "grad_norm": 0.4062393307685852,
+      "learning_rate": 8e-05,
+      "loss": 1.5496,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6035718624711457,
+      "grad_norm": 0.4321100413799286,
+      "learning_rate": 8e-05,
+      "loss": 1.7848,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6038148463127202,
+      "grad_norm": 0.42457231879234314,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6040578301542947,
+      "grad_norm": 0.4348156154155731,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6043008139958693,
+      "grad_norm": 0.4834965467453003,
+      "learning_rate": 8e-05,
+      "loss": 1.7267,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6045437978374438,
+      "grad_norm": 0.43507274985313416,
+      "learning_rate": 8e-05,
+      "loss": 1.7981,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6047867816790183,
+      "grad_norm": 0.42029669880867004,
+      "learning_rate": 8e-05,
+      "loss": 1.8782,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6050297655205928,
+      "grad_norm": 0.4108349084854126,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6052727493621675,
+      "grad_norm": 0.42134028673171997,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 2491
+    },
+    {
+      "epoch": 0.605515733203742,
+      "grad_norm": 0.41485196352005005,
+      "learning_rate": 8e-05,
+      "loss": 1.629,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6057587170453165,
+      "grad_norm": 0.43490925431251526,
+      "learning_rate": 8e-05,
+      "loss": 1.7514,
+      "step": 2493
+    },
+    {
+      "epoch": 0.606001700886891,
+      "grad_norm": 0.4186338782310486,
+      "learning_rate": 8e-05,
+      "loss": 1.7008,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6062446847284656,
+      "grad_norm": 0.4107659161090851,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6064876685700401,
+      "grad_norm": 0.42910102009773254,
+      "learning_rate": 8e-05,
+      "loss": 1.7501,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6067306524116146,
+      "grad_norm": 0.43301084637641907,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6069736362531891,
+      "grad_norm": 0.42027756571769714,
+      "learning_rate": 8e-05,
+      "loss": 1.7223,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6072166200947637,
+      "grad_norm": 0.415526807308197,
+      "learning_rate": 8e-05,
+      "loss": 1.6176,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6074596039363382,
+      "grad_norm": 0.42387959361076355,
+      "learning_rate": 8e-05,
+      "loss": 1.7902,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6077025877779128,
+      "grad_norm": 0.4080601930618286,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6079455716194873,
+      "grad_norm": 0.4546234607696533,
+      "learning_rate": 8e-05,
+      "loss": 1.8758,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6081885554610619,
+      "grad_norm": 0.41828829050064087,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6084315393026364,
+      "grad_norm": 0.4156382381916046,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6086745231442109,
+      "grad_norm": 0.4166845679283142,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6089175069857854,
+      "grad_norm": 0.464033842086792,
+      "learning_rate": 8e-05,
+      "loss": 1.6495,
+      "step": 2506
+    },
+    {
+      "epoch": 0.60916049082736,
+      "grad_norm": 0.4293297231197357,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6094034746689345,
+      "grad_norm": 0.41994819045066833,
+      "learning_rate": 8e-05,
+      "loss": 1.5575,
+      "step": 2508
+    },
+    {
+      "epoch": 0.609646458510509,
+      "grad_norm": 0.4435938000679016,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6098894423520835,
+      "grad_norm": 0.419653058052063,
+      "learning_rate": 8e-05,
+      "loss": 1.6947,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6101324261936582,
+      "grad_norm": 0.41150185465812683,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6103754100352327,
+      "grad_norm": 0.47121474146842957,
+      "learning_rate": 8e-05,
+      "loss": 1.7947,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6106183938768072,
+      "grad_norm": 0.42757055163383484,
+      "learning_rate": 8e-05,
+      "loss": 1.796,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6108613777183818,
+      "grad_norm": 0.41179606318473816,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6111043615599563,
+      "grad_norm": 0.4639357626438141,
+      "learning_rate": 8e-05,
+      "loss": 1.8249,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6113473454015308,
+      "grad_norm": 0.42395976185798645,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6115903292431053,
+      "grad_norm": 0.4099504053592682,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6118333130846799,
+      "grad_norm": 0.43916642665863037,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6120762969262544,
+      "grad_norm": 0.42398327589035034,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6123192807678289,
+      "grad_norm": 0.4526452124118805,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6125622646094034,
+      "grad_norm": 0.43443700671195984,
+      "learning_rate": 8e-05,
+      "loss": 1.82,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6128052484509781,
+      "grad_norm": 0.41617026925086975,
+      "learning_rate": 8e-05,
+      "loss": 1.6455,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6130482322925526,
+      "grad_norm": 0.4145825505256653,
+      "learning_rate": 8e-05,
+      "loss": 1.6597,
+      "step": 2523
+    },
+    {
+      "epoch": 0.6132912161341271,
+      "grad_norm": 0.4272093176841736,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6135341999757016,
+      "grad_norm": 0.4282555878162384,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6137771838172762,
+      "grad_norm": 0.4502201974391937,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6140201676588507,
+      "grad_norm": 0.4259174168109894,
+      "learning_rate": 8e-05,
+      "loss": 1.7968,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6142631515004252,
+      "grad_norm": 0.43686801195144653,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6145061353419997,
+      "grad_norm": 0.4490336775779724,
+      "learning_rate": 8e-05,
+      "loss": 1.7273,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6147491191835743,
+      "grad_norm": 0.4362275004386902,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6149921030251488,
+      "grad_norm": 0.45337826013565063,
+      "learning_rate": 8e-05,
+      "loss": 1.7809,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6152350868667233,
+      "grad_norm": 0.4103049337863922,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6154780707082979,
+      "grad_norm": 0.4356194734573364,
+      "learning_rate": 8e-05,
+      "loss": 1.7924,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6157210545498725,
+      "grad_norm": 0.42592617869377136,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 2534
+    },
+    {
+      "epoch": 0.615964038391447,
+      "grad_norm": 0.42202380299568176,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6162070222330215,
+      "grad_norm": 0.4229716658592224,
+      "learning_rate": 8e-05,
+      "loss": 1.6049,
+      "step": 2536
+    },
+    {
+      "epoch": 0.616450006074596,
+      "grad_norm": 0.44110849499702454,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6166929899161706,
+      "grad_norm": 0.45031967759132385,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6169359737577451,
+      "grad_norm": 0.43879231810569763,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6171789575993196,
+      "grad_norm": 0.4405655264854431,
+      "learning_rate": 8e-05,
+      "loss": 1.7393,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6174219414408941,
+      "grad_norm": 0.44394829869270325,
+      "learning_rate": 8e-05,
+      "loss": 1.7661,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6176649252824687,
+      "grad_norm": 0.4244672358036041,
+      "learning_rate": 8e-05,
+      "loss": 1.7738,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6179079091240433,
+      "grad_norm": 0.4294050931930542,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6181508929656178,
+      "grad_norm": 0.42670539021492004,
+      "learning_rate": 8e-05,
+      "loss": 1.8418,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6183938768071923,
+      "grad_norm": 0.4144493639469147,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6186368606487669,
+      "grad_norm": 0.44547638297080994,
+      "learning_rate": 8e-05,
+      "loss": 1.662,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6188798444903414,
+      "grad_norm": 0.42733410000801086,
+      "learning_rate": 8e-05,
+      "loss": 1.6567,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6191228283319159,
+      "grad_norm": 0.4180489182472229,
+      "learning_rate": 8e-05,
+      "loss": 1.6989,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6193658121734905,
+      "grad_norm": 0.4086499810218811,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 2549
+    },
+    {
+      "epoch": 0.619608796015065,
+      "grad_norm": 0.3928138315677643,
+      "learning_rate": 8e-05,
+      "loss": 1.4558,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6198517798566395,
+      "grad_norm": 0.4183439016342163,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 2551
+    },
+    {
+      "epoch": 0.620094763698214,
+      "grad_norm": 0.44228413701057434,
+      "learning_rate": 8e-05,
+      "loss": 1.7191,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6203377475397887,
+      "grad_norm": 0.42684802412986755,
+      "learning_rate": 8e-05,
+      "loss": 1.7245,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6205807313813632,
+      "grad_norm": 0.4153320789337158,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6208237152229377,
+      "grad_norm": 0.4317556917667389,
+      "learning_rate": 8e-05,
+      "loss": 1.7533,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6210666990645122,
+      "grad_norm": 0.423536479473114,
+      "learning_rate": 8e-05,
+      "loss": 1.696,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6213096829060868,
+      "grad_norm": 0.4165915548801422,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6215526667476613,
+      "grad_norm": 0.4175946116447449,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6217956505892358,
+      "grad_norm": 0.4565088450908661,
+      "learning_rate": 8e-05,
+      "loss": 1.8072,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6220386344308103,
+      "grad_norm": 0.47681909799575806,
+      "learning_rate": 8e-05,
+      "loss": 1.7637,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6222816182723849,
+      "grad_norm": 0.4348503351211548,
+      "learning_rate": 8e-05,
+      "loss": 1.4782,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6225246021139594,
+      "grad_norm": 0.4135608375072479,
+      "learning_rate": 8e-05,
+      "loss": 1.6008,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6227675859555339,
+      "grad_norm": 0.4383459985256195,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6230105697971084,
+      "grad_norm": 0.4476924240589142,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6232535536386831,
+      "grad_norm": 0.4379726052284241,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6234965374802576,
+      "grad_norm": 0.4011707007884979,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6237395213218321,
+      "grad_norm": 0.439129501581192,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6239825051634066,
+      "grad_norm": 0.46827462315559387,
+      "learning_rate": 8e-05,
+      "loss": 1.8474,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6242254890049812,
+      "grad_norm": 0.43110692501068115,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6244684728465557,
+      "grad_norm": 0.42442917823791504,
+      "learning_rate": 8e-05,
+      "loss": 1.703,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6247114566881302,
+      "grad_norm": 0.4448035657405853,
+      "learning_rate": 8e-05,
+      "loss": 1.787,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6249544405297047,
+      "grad_norm": 0.39536356925964355,
+      "learning_rate": 8e-05,
+      "loss": 1.6137,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6251974243712793,
+      "grad_norm": 0.42168813943862915,
+      "learning_rate": 8e-05,
+      "loss": 1.7345,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6254404082128538,
+      "grad_norm": 0.4354855716228485,
+      "learning_rate": 8e-05,
+      "loss": 1.7339,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6256833920544284,
+      "grad_norm": 0.4181711673736572,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6259263758960029,
+      "grad_norm": 0.42490074038505554,
+      "learning_rate": 8e-05,
+      "loss": 1.5755,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6261693597375775,
+      "grad_norm": 0.4245617091655731,
+      "learning_rate": 8e-05,
+      "loss": 1.6848,
+      "step": 2577
+    },
+    {
+      "epoch": 0.626412343579152,
+      "grad_norm": 0.44003593921661377,
+      "learning_rate": 8e-05,
+      "loss": 1.7697,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6266553274207265,
+      "grad_norm": 0.4520803987979889,
+      "learning_rate": 8e-05,
+      "loss": 1.7213,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6268983112623011,
+      "grad_norm": 0.4083026647567749,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6271412951038756,
+      "grad_norm": 0.4308316111564636,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6273842789454501,
+      "grad_norm": 0.44732677936553955,
+      "learning_rate": 8e-05,
+      "loss": 1.69,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6276272627870246,
+      "grad_norm": 0.41122570633888245,
+      "learning_rate": 8e-05,
+      "loss": 1.7143,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6278702466285992,
+      "grad_norm": 0.42767059803009033,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6281132304701738,
+      "grad_norm": 0.4482073187828064,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6283562143117483,
+      "grad_norm": 0.425933837890625,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6285991981533228,
+      "grad_norm": 0.4497233033180237,
+      "learning_rate": 8e-05,
+      "loss": 1.7595,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6288421819948974,
+      "grad_norm": 0.4143991470336914,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6290851658364719,
+      "grad_norm": 0.4549991190433502,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6293281496780464,
+      "grad_norm": 0.43654417991638184,
+      "learning_rate": 8e-05,
+      "loss": 1.7862,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6295711335196209,
+      "grad_norm": 0.4487234354019165,
+      "learning_rate": 8e-05,
+      "loss": 1.7315,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6298141173611955,
+      "grad_norm": 0.45858246088027954,
+      "learning_rate": 8e-05,
+      "loss": 1.7013,
+      "step": 2592
+    },
+    {
+      "epoch": 0.63005710120277,
+      "grad_norm": 0.42146363854408264,
+      "learning_rate": 8e-05,
+      "loss": 1.6246,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6303000850443445,
+      "grad_norm": 0.40874940156936646,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 2594
+    },
+    {
+      "epoch": 0.630543068885919,
+      "grad_norm": 0.4485381245613098,
+      "learning_rate": 8e-05,
+      "loss": 1.8109,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6307860527274937,
+      "grad_norm": 0.4378395676612854,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6310290365690682,
+      "grad_norm": 0.42810145020484924,
+      "learning_rate": 8e-05,
+      "loss": 1.7321,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6312720204106427,
+      "grad_norm": 0.43177729845046997,
+      "learning_rate": 8e-05,
+      "loss": 1.7266,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6315150042522172,
+      "grad_norm": 0.4171009957790375,
+      "learning_rate": 8e-05,
+      "loss": 1.4937,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6317579880937918,
+      "grad_norm": 0.42278575897216797,
+      "learning_rate": 8e-05,
+      "loss": 1.7061,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6320009719353663,
+      "grad_norm": 0.4374859929084778,
+      "learning_rate": 8e-05,
+      "loss": 1.6875,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6322439557769408,
+      "grad_norm": 0.43232500553131104,
+      "learning_rate": 8e-05,
+      "loss": 1.7732,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6324869396185153,
+      "grad_norm": 0.4480953812599182,
+      "learning_rate": 8e-05,
+      "loss": 1.833,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6327299234600899,
+      "grad_norm": 0.428558886051178,
+      "learning_rate": 8e-05,
+      "loss": 1.7832,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6329729073016644,
+      "grad_norm": 0.43188753724098206,
+      "learning_rate": 8e-05,
+      "loss": 1.789,
+      "step": 2605
+    },
+    {
+      "epoch": 0.633215891143239,
+      "grad_norm": 0.44204968214035034,
+      "learning_rate": 8e-05,
+      "loss": 1.7752,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6334588749848135,
+      "grad_norm": 0.4403454661369324,
+      "learning_rate": 8e-05,
+      "loss": 1.7196,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6337018588263881,
+      "grad_norm": 0.42113178968429565,
+      "learning_rate": 8e-05,
+      "loss": 1.7242,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6339448426679626,
+      "grad_norm": 0.4443967640399933,
+      "learning_rate": 8e-05,
+      "loss": 1.7155,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6341878265095371,
+      "grad_norm": 0.4284020662307739,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6344308103511116,
+      "grad_norm": 0.42479994893074036,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6346737941926862,
+      "grad_norm": 0.40834593772888184,
+      "learning_rate": 8e-05,
+      "loss": 1.516,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6349167780342607,
+      "grad_norm": 0.4117913842201233,
+      "learning_rate": 8e-05,
+      "loss": 1.7414,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6351597618758352,
+      "grad_norm": 0.42295315861701965,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6354027457174098,
+      "grad_norm": 0.4250132143497467,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6356457295589844,
+      "grad_norm": 0.42362743616104126,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6358887134005589,
+      "grad_norm": 0.437267005443573,
+      "learning_rate": 8e-05,
+      "loss": 1.6675,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6361316972421334,
+      "grad_norm": 0.4110496938228607,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 2618
+    },
+    {
+      "epoch": 0.636374681083708,
+      "grad_norm": 0.4345325529575348,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6366176649252825,
+      "grad_norm": 0.4358530640602112,
+      "learning_rate": 8e-05,
+      "loss": 1.7169,
+      "step": 2620
+    },
+    {
+      "epoch": 0.636860648766857,
+      "grad_norm": 0.42882612347602844,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6371036326084315,
+      "grad_norm": 0.4442894458770752,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6373466164500061,
+      "grad_norm": 0.4323555827140808,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6375896002915806,
+      "grad_norm": 0.446566641330719,
+      "learning_rate": 8e-05,
+      "loss": 1.7632,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6378325841331551,
+      "grad_norm": 0.47879257798194885,
+      "learning_rate": 8e-05,
+      "loss": 1.8424,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6380755679747296,
+      "grad_norm": 0.4168838560581207,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6383185518163043,
+      "grad_norm": 0.4289827048778534,
+      "learning_rate": 8e-05,
+      "loss": 1.549,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6385615356578788,
+      "grad_norm": 0.45622771978378296,
+      "learning_rate": 8e-05,
+      "loss": 1.7579,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6388045194994533,
+      "grad_norm": 0.4165874421596527,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6390475033410278,
+      "grad_norm": 0.439090758562088,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6392904871826024,
+      "grad_norm": 0.4318310022354126,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6395334710241769,
+      "grad_norm": 0.41731974482536316,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6397764548657514,
+      "grad_norm": 0.4407217800617218,
+      "learning_rate": 8e-05,
+      "loss": 1.7437,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6400194387073259,
+      "grad_norm": 0.4362947642803192,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6402624225489005,
+      "grad_norm": 0.4433870017528534,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 2635
+    },
+    {
+      "epoch": 0.640505406390475,
+      "grad_norm": 0.40365076065063477,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6407483902320495,
+      "grad_norm": 0.43577131628990173,
+      "learning_rate": 8e-05,
+      "loss": 1.7692,
+      "step": 2637
+    },
+    {
+      "epoch": 0.640991374073624,
+      "grad_norm": 0.418148398399353,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6412343579151987,
+      "grad_norm": 0.4208431839942932,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6414773417567732,
+      "grad_norm": 0.44447189569473267,
+      "learning_rate": 8e-05,
+      "loss": 1.7173,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6417203255983477,
+      "grad_norm": 0.41978487372398376,
+      "learning_rate": 8e-05,
+      "loss": 1.4916,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6419633094399222,
+      "grad_norm": 0.42766228318214417,
+      "learning_rate": 8e-05,
+      "loss": 1.464,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6422062932814968,
+      "grad_norm": 0.4409187436103821,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6424492771230713,
+      "grad_norm": 0.4471896290779114,
+      "learning_rate": 8e-05,
+      "loss": 1.5874,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6426922609646458,
+      "grad_norm": 0.438822478055954,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6429352448062204,
+      "grad_norm": 0.44489672780036926,
+      "learning_rate": 8e-05,
+      "loss": 1.694,
+      "step": 2646
+    },
+    {
+      "epoch": 0.643178228647795,
+      "grad_norm": 0.4189907908439636,
+      "learning_rate": 8e-05,
+      "loss": 1.6755,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6434212124893695,
+      "grad_norm": 0.43250712752342224,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 2648
+    },
+    {
+      "epoch": 0.643664196330944,
+      "grad_norm": 0.45793941617012024,
+      "learning_rate": 8e-05,
+      "loss": 1.7371,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6439071801725186,
+      "grad_norm": 0.4285843074321747,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6441501640140931,
+      "grad_norm": 0.42590782046318054,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6443931478556676,
+      "grad_norm": 0.4645673632621765,
+      "learning_rate": 8e-05,
+      "loss": 1.7451,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6446361316972421,
+      "grad_norm": 0.41977444291114807,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6448791155388167,
+      "grad_norm": 0.42946362495422363,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6451220993803912,
+      "grad_norm": 0.44838717579841614,
+      "learning_rate": 8e-05,
+      "loss": 1.7525,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6453650832219657,
+      "grad_norm": 0.4370400607585907,
+      "learning_rate": 8e-05,
+      "loss": 1.749,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6456080670635402,
+      "grad_norm": 0.4621409773826599,
+      "learning_rate": 8e-05,
+      "loss": 1.6799,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6458510509051149,
+      "grad_norm": 0.45565065741539,
+      "learning_rate": 8e-05,
+      "loss": 1.9007,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6460940347466894,
+      "grad_norm": 0.4766846001148224,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6463370185882639,
+      "grad_norm": 0.41699716448783875,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6465800024298384,
+      "grad_norm": 0.42942753434181213,
+      "learning_rate": 8e-05,
+      "loss": 1.6951,
+      "step": 2661
+    },
+    {
+      "epoch": 0.646822986271413,
+      "grad_norm": 0.4418134391307831,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6470659701129875,
+      "grad_norm": 0.4244852066040039,
+      "learning_rate": 8e-05,
+      "loss": 1.6953,
+      "step": 2663
+    },
+    {
+      "epoch": 0.647308953954562,
+      "grad_norm": 0.4480370283126831,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6475519377961365,
+      "grad_norm": 0.45050179958343506,
+      "learning_rate": 8e-05,
+      "loss": 1.7932,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6477949216377111,
+      "grad_norm": 0.43569415807724,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6480379054792856,
+      "grad_norm": 0.4113406836986542,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6482808893208601,
+      "grad_norm": 0.45200860500335693,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6485238731624346,
+      "grad_norm": 0.4367867708206177,
+      "learning_rate": 8e-05,
+      "loss": 1.5387,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6487668570040093,
+      "grad_norm": 0.42735886573791504,
+      "learning_rate": 8e-05,
+      "loss": 1.7624,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6490098408455838,
+      "grad_norm": 0.4208768904209137,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6492528246871583,
+      "grad_norm": 0.4291721284389496,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6494958085287328,
+      "grad_norm": 0.45863476395606995,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6497387923703074,
+      "grad_norm": 0.4188409149646759,
+      "learning_rate": 8e-05,
+      "loss": 1.7296,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6499817762118819,
+      "grad_norm": 0.43191903829574585,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6502247600534564,
+      "grad_norm": 0.4402519762516022,
+      "learning_rate": 8e-05,
+      "loss": 1.7769,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6504677438950309,
+      "grad_norm": 0.42813950777053833,
+      "learning_rate": 8e-05,
+      "loss": 1.6624,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6507107277366055,
+      "grad_norm": 0.4296523928642273,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 2678
+    },
+    {
+      "epoch": 0.65095371157818,
+      "grad_norm": 0.42331957817077637,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6511966954197546,
+      "grad_norm": 0.42995479702949524,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6514396792613292,
+      "grad_norm": 0.4499923884868622,
+      "learning_rate": 8e-05,
+      "loss": 1.696,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6516826631029037,
+      "grad_norm": 0.4375508427619934,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6519256469444782,
+      "grad_norm": 0.4245275557041168,
+      "learning_rate": 8e-05,
+      "loss": 1.7766,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6521686307860527,
+      "grad_norm": 0.44000858068466187,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6524116146276273,
+      "grad_norm": 0.4305912256240845,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6526545984692018,
+      "grad_norm": 0.46314767003059387,
+      "learning_rate": 8e-05,
+      "loss": 1.7763,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6528975823107763,
+      "grad_norm": 0.4215608537197113,
+      "learning_rate": 8e-05,
+      "loss": 1.7841,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6531405661523508,
+      "grad_norm": 0.42683127522468567,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6533835499939254,
+      "grad_norm": 0.42152634263038635,
+      "learning_rate": 8e-05,
+      "loss": 1.6997,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6536265338355,
+      "grad_norm": 0.43715110421180725,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6538695176770745,
+      "grad_norm": 0.4029892385005951,
+      "learning_rate": 8e-05,
+      "loss": 1.5506,
+      "step": 2691
+    },
+    {
+      "epoch": 0.654112501518649,
+      "grad_norm": 0.42743903398513794,
+      "learning_rate": 8e-05,
+      "loss": 1.5413,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6543554853602236,
+      "grad_norm": 0.4199978709220886,
+      "learning_rate": 8e-05,
+      "loss": 1.6717,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6545984692017981,
+      "grad_norm": 0.440282940864563,
+      "learning_rate": 8e-05,
+      "loss": 1.7336,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6548414530433726,
+      "grad_norm": 0.4389897286891937,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6550844368849471,
+      "grad_norm": 0.45226407051086426,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6553274207265217,
+      "grad_norm": 0.42576125264167786,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6555704045680962,
+      "grad_norm": 0.43796688318252563,
+      "learning_rate": 8e-05,
+      "loss": 1.7955,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6558133884096707,
+      "grad_norm": 0.4207051396369934,
+      "learning_rate": 8e-05,
+      "loss": 1.7579,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6560563722512452,
+      "grad_norm": 0.4578252136707306,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6562993560928199,
+      "grad_norm": 0.45163866877555847,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6565423399343944,
+      "grad_norm": 0.4318283200263977,
+      "learning_rate": 8e-05,
+      "loss": 1.7664,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6567853237759689,
+      "grad_norm": 0.42416471242904663,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6570283076175434,
+      "grad_norm": 0.4698314070701599,
+      "learning_rate": 8e-05,
+      "loss": 1.82,
+      "step": 2704
+    },
+    {
+      "epoch": 0.657271291459118,
+      "grad_norm": 0.4339340627193451,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6575142753006925,
+      "grad_norm": 0.4280409812927246,
+      "learning_rate": 8e-05,
+      "loss": 1.8196,
+      "step": 2706
+    },
+    {
+      "epoch": 0.657757259142267,
+      "grad_norm": 0.44138187170028687,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6580002429838415,
+      "grad_norm": 0.42129242420196533,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6582432268254161,
+      "grad_norm": 0.41588228940963745,
+      "learning_rate": 8e-05,
+      "loss": 1.6204,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6584862106669906,
+      "grad_norm": 0.4260967969894409,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6587291945085652,
+      "grad_norm": 0.43918874859809875,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6589721783501398,
+      "grad_norm": 0.46208980679512024,
+      "learning_rate": 8e-05,
+      "loss": 1.6909,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6592151621917143,
+      "grad_norm": 0.42393893003463745,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6594581460332888,
+      "grad_norm": 0.40289050340652466,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6597011298748633,
+      "grad_norm": 0.4082167446613312,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6599441137164379,
+      "grad_norm": 0.4172167479991913,
+      "learning_rate": 8e-05,
+      "loss": 1.5536,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6601870975580124,
+      "grad_norm": 0.4407787024974823,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6604300813995869,
+      "grad_norm": 0.4522525370121002,
+      "learning_rate": 8e-05,
+      "loss": 1.8383,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6606730652411614,
+      "grad_norm": 0.4222140312194824,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 2719
+    },
+    {
+      "epoch": 0.660916049082736,
+      "grad_norm": 0.418477863073349,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6611590329243106,
+      "grad_norm": 0.4419880211353302,
+      "learning_rate": 8e-05,
+      "loss": 1.8229,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6614020167658851,
+      "grad_norm": 0.4308634102344513,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6616450006074596,
+      "grad_norm": 0.41491010785102844,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6618879844490342,
+      "grad_norm": 0.42494139075279236,
+      "learning_rate": 8e-05,
+      "loss": 1.5578,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6621309682906087,
+      "grad_norm": 0.42618733644485474,
+      "learning_rate": 8e-05,
+      "loss": 1.6443,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6623739521321832,
+      "grad_norm": 0.4543378949165344,
+      "learning_rate": 8e-05,
+      "loss": 1.5653,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6626169359737577,
+      "grad_norm": 0.42057380080223083,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6628599198153323,
+      "grad_norm": 0.4596848785877228,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6631029036569068,
+      "grad_norm": 0.4214380979537964,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6633458874984813,
+      "grad_norm": 0.4242684543132782,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6635888713400558,
+      "grad_norm": 0.4381091296672821,
+      "learning_rate": 8e-05,
+      "loss": 1.7094,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6638318551816305,
+      "grad_norm": 0.4281906187534332,
+      "learning_rate": 8e-05,
+      "loss": 1.7558,
+      "step": 2732
+    },
+    {
+      "epoch": 0.664074839023205,
+      "grad_norm": 0.4531438648700714,
+      "learning_rate": 8e-05,
+      "loss": 1.7281,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6643178228647795,
+      "grad_norm": 0.42967137694358826,
+      "learning_rate": 8e-05,
+      "loss": 1.757,
+      "step": 2734
+    },
+    {
+      "epoch": 0.664560806706354,
+      "grad_norm": 0.40972012281417847,
+      "learning_rate": 8e-05,
+      "loss": 1.6486,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6648037905479286,
+      "grad_norm": 0.41229069232940674,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6650467743895031,
+      "grad_norm": 0.4281129240989685,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6652897582310776,
+      "grad_norm": 0.4217357635498047,
+      "learning_rate": 8e-05,
+      "loss": 1.6954,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6655327420726521,
+      "grad_norm": 0.4949724078178406,
+      "learning_rate": 8e-05,
+      "loss": 1.6552,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6657757259142267,
+      "grad_norm": 0.4355374276638031,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6660187097558012,
+      "grad_norm": 0.41408905386924744,
+      "learning_rate": 8e-05,
+      "loss": 1.5543,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6662616935973757,
+      "grad_norm": 0.4655771255493164,
+      "learning_rate": 8e-05,
+      "loss": 1.895,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6665046774389503,
+      "grad_norm": 0.4114040732383728,
+      "learning_rate": 8e-05,
+      "loss": 1.6217,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6667476612805249,
+      "grad_norm": 0.467652827501297,
+      "learning_rate": 8e-05,
+      "loss": 1.8087,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6669906451220994,
+      "grad_norm": 0.41278281807899475,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6672336289636739,
+      "grad_norm": 0.4434947073459625,
+      "learning_rate": 8e-05,
+      "loss": 1.8071,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6674766128052485,
+      "grad_norm": 0.4336857795715332,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 2747
+    },
+    {
+      "epoch": 0.667719596646823,
+      "grad_norm": 0.44751599431037903,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6679625804883975,
+      "grad_norm": 0.4325321912765503,
+      "learning_rate": 8e-05,
+      "loss": 1.7023,
+      "step": 2749
+    },
+    {
+      "epoch": 0.668205564329972,
+      "grad_norm": 0.4493902921676636,
+      "learning_rate": 8e-05,
+      "loss": 1.8001,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6684485481715466,
+      "grad_norm": 0.4371159076690674,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6686915320131211,
+      "grad_norm": 0.45062997937202454,
+      "learning_rate": 8e-05,
+      "loss": 1.6929,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6689345158546957,
+      "grad_norm": 0.43468889594078064,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6691774996962702,
+      "grad_norm": 0.4319245517253876,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6694204835378448,
+      "grad_norm": 0.41071975231170654,
+      "learning_rate": 8e-05,
+      "loss": 1.6132,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6696634673794193,
+      "grad_norm": 0.41352805495262146,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6699064512209938,
+      "grad_norm": 0.4360056519508362,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6701494350625683,
+      "grad_norm": 0.423724502325058,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6703924189041429,
+      "grad_norm": 0.4352973699569702,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6706354027457174,
+      "grad_norm": 0.44685518741607666,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6708783865872919,
+      "grad_norm": 0.4301108717918396,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6711213704288664,
+      "grad_norm": 0.4443804621696472,
+      "learning_rate": 8e-05,
+      "loss": 1.6424,
+      "step": 2762
+    },
+    {
+      "epoch": 0.671364354270441,
+      "grad_norm": 0.4080183804035187,
+      "learning_rate": 8e-05,
+      "loss": 1.6006,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6716073381120156,
+      "grad_norm": 0.4252590537071228,
+      "learning_rate": 8e-05,
+      "loss": 1.6456,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6718503219535901,
+      "grad_norm": 0.4212346374988556,
+      "learning_rate": 8e-05,
+      "loss": 1.7311,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6720933057951646,
+      "grad_norm": 0.436878502368927,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6723362896367392,
+      "grad_norm": 0.4279247224330902,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6725792734783137,
+      "grad_norm": 0.41586393117904663,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6728222573198882,
+      "grad_norm": 0.41080671548843384,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6730652411614627,
+      "grad_norm": 0.44032683968544006,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6733082250030373,
+      "grad_norm": 0.4310445487499237,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6735512088446118,
+      "grad_norm": 0.4293959438800812,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6737941926861863,
+      "grad_norm": 0.4403887093067169,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6740371765277609,
+      "grad_norm": 0.4439060389995575,
+      "learning_rate": 8e-05,
+      "loss": 1.6539,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6742801603693355,
+      "grad_norm": 0.4334704279899597,
+      "learning_rate": 8e-05,
+      "loss": 1.847,
+      "step": 2775
+    },
+    {
+      "epoch": 0.67452314421091,
+      "grad_norm": 0.419173926115036,
+      "learning_rate": 8e-05,
+      "loss": 1.5736,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6747661280524845,
+      "grad_norm": 0.44975632429122925,
+      "learning_rate": 8e-05,
+      "loss": 1.7237,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6750091118940591,
+      "grad_norm": 0.405448853969574,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6752520957356336,
+      "grad_norm": 0.459410160779953,
+      "learning_rate": 8e-05,
+      "loss": 1.9279,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6754950795772081,
+      "grad_norm": 0.4405708611011505,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6757380634187826,
+      "grad_norm": 0.410300612449646,
+      "learning_rate": 8e-05,
+      "loss": 1.668,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6759810472603572,
+      "grad_norm": 0.4388057291507721,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6762240311019317,
+      "grad_norm": 0.4449143707752228,
+      "learning_rate": 8e-05,
+      "loss": 1.7522,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6764670149435063,
+      "grad_norm": 0.43264544010162354,
+      "learning_rate": 8e-05,
+      "loss": 1.6467,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6767099987850808,
+      "grad_norm": 0.4240364134311676,
+      "learning_rate": 8e-05,
+      "loss": 1.6614,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6769529826266554,
+      "grad_norm": 0.42735329270362854,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6771959664682299,
+      "grad_norm": 0.42375391721725464,
+      "learning_rate": 8e-05,
+      "loss": 1.7181,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6774389503098044,
+      "grad_norm": 0.41570213437080383,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6776819341513789,
+      "grad_norm": 0.41946732997894287,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6779249179929535,
+      "grad_norm": 0.427961528301239,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 2790
+    },
+    {
+      "epoch": 0.678167901834528,
+      "grad_norm": 0.4631393551826477,
+      "learning_rate": 8e-05,
+      "loss": 1.7629,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6784108856761025,
+      "grad_norm": 0.43700042366981506,
+      "learning_rate": 8e-05,
+      "loss": 1.8111,
+      "step": 2792
+    },
+    {
+      "epoch": 0.678653869517677,
+      "grad_norm": 0.4172798693180084,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6788968533592517,
+      "grad_norm": 0.44530415534973145,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6791398372008262,
+      "grad_norm": 0.45082876086235046,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6793828210424007,
+      "grad_norm": 0.4352196753025055,
+      "learning_rate": 8e-05,
+      "loss": 1.8301,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6796258048839752,
+      "grad_norm": 0.41215893626213074,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6798687887255498,
+      "grad_norm": 0.43125149607658386,
+      "learning_rate": 8e-05,
+      "loss": 1.7772,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6801117725671243,
+      "grad_norm": 0.41493120789527893,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6803547564086988,
+      "grad_norm": 0.42573511600494385,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6805977402502733,
+      "grad_norm": 0.41968226432800293,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6808407240918479,
+      "grad_norm": 0.4159198999404907,
+      "learning_rate": 8e-05,
+      "loss": 1.4322,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6810837079334224,
+      "grad_norm": 0.4502333998680115,
+      "learning_rate": 8e-05,
+      "loss": 1.7811,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6813266917749969,
+      "grad_norm": 0.42658501863479614,
+      "learning_rate": 8e-05,
+      "loss": 1.6671,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6815696756165714,
+      "grad_norm": 0.4698079526424408,
+      "learning_rate": 8e-05,
+      "loss": 1.8551,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6818126594581461,
+      "grad_norm": 0.44726017117500305,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6820556432997206,
+      "grad_norm": 0.44822147488594055,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6822986271412951,
+      "grad_norm": 0.4212280809879303,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6825416109828696,
+      "grad_norm": 0.424966424703598,
+      "learning_rate": 8e-05,
+      "loss": 1.5315,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6827845948244442,
+      "grad_norm": 0.4459429979324341,
+      "learning_rate": 8e-05,
+      "loss": 1.7628,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6830275786660187,
+      "grad_norm": 0.41798895597457886,
+      "learning_rate": 8e-05,
+      "loss": 1.531,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6832705625075932,
+      "grad_norm": 0.45187270641326904,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6835135463491678,
+      "grad_norm": 0.43178221583366394,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6837565301907423,
+      "grad_norm": 0.42000076174736023,
+      "learning_rate": 8e-05,
+      "loss": 1.5952,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6839995140323168,
+      "grad_norm": 0.4335983395576477,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6842424978738914,
+      "grad_norm": 0.4407869279384613,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 2816
+    },
+    {
+      "epoch": 0.684485481715466,
+      "grad_norm": 0.44517838954925537,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6847284655570405,
+      "grad_norm": 0.41948360204696655,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 2818
+    },
+    {
+      "epoch": 0.684971449398615,
+      "grad_norm": 0.4196469783782959,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6852144332401895,
+      "grad_norm": 0.44996392726898193,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6854574170817641,
+      "grad_norm": 0.42877084016799927,
+      "learning_rate": 8e-05,
+      "loss": 1.8095,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6857004009233386,
+      "grad_norm": 0.42822638154029846,
+      "learning_rate": 8e-05,
+      "loss": 1.7041,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6859433847649131,
+      "grad_norm": 0.43635883927345276,
+      "learning_rate": 8e-05,
+      "loss": 1.7196,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6861863686064876,
+      "grad_norm": 0.4505825936794281,
+      "learning_rate": 8e-05,
+      "loss": 1.6865,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6864293524480622,
+      "grad_norm": 0.4453728199005127,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6866723362896368,
+      "grad_norm": 0.42800915241241455,
+      "learning_rate": 8e-05,
+      "loss": 1.6662,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6869153201312113,
+      "grad_norm": 0.43462935090065,
+      "learning_rate": 8e-05,
+      "loss": 1.7734,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6871583039727858,
+      "grad_norm": 0.4350861608982086,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6874012878143604,
+      "grad_norm": 0.4432491660118103,
+      "learning_rate": 8e-05,
+      "loss": 1.7994,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6876442716559349,
+      "grad_norm": 0.4268428087234497,
+      "learning_rate": 8e-05,
+      "loss": 1.6798,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6878872554975094,
+      "grad_norm": 0.425647109746933,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6881302393390839,
+      "grad_norm": 0.42241379618644714,
+      "learning_rate": 8e-05,
+      "loss": 1.5958,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6883732231806585,
+      "grad_norm": 0.44079574942588806,
+      "learning_rate": 8e-05,
+      "loss": 1.6557,
+      "step": 2833
+    },
+    {
+      "epoch": 0.688616207022233,
+      "grad_norm": 0.454862117767334,
+      "learning_rate": 8e-05,
+      "loss": 1.7568,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6888591908638075,
+      "grad_norm": 0.4520788788795471,
+      "learning_rate": 8e-05,
+      "loss": 1.6873,
+      "step": 2835
+    },
+    {
+      "epoch": 0.689102174705382,
+      "grad_norm": 0.42827796936035156,
+      "learning_rate": 8e-05,
+      "loss": 1.6534,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6893451585469567,
+      "grad_norm": 0.45492035150527954,
+      "learning_rate": 8e-05,
+      "loss": 1.7898,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6895881423885312,
+      "grad_norm": 0.43842339515686035,
+      "learning_rate": 8e-05,
+      "loss": 1.653,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6898311262301057,
+      "grad_norm": 0.410273939371109,
+      "learning_rate": 8e-05,
+      "loss": 1.5635,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6900741100716802,
+      "grad_norm": 0.393926203250885,
+      "learning_rate": 8e-05,
+      "loss": 1.5678,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6903170939132548,
+      "grad_norm": 0.4179787337779999,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6905600777548293,
+      "grad_norm": 0.4206624925136566,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6908030615964038,
+      "grad_norm": 0.42536622285842896,
+      "learning_rate": 8e-05,
+      "loss": 1.7121,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6910460454379784,
+      "grad_norm": 0.4455561935901642,
+      "learning_rate": 8e-05,
+      "loss": 1.7256,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6912890292795529,
+      "grad_norm": 0.43917906284332275,
+      "learning_rate": 8e-05,
+      "loss": 1.6017,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6915320131211274,
+      "grad_norm": 0.4152160584926605,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 2846
+    },
+    {
+      "epoch": 0.691774996962702,
+      "grad_norm": 0.43503397703170776,
+      "learning_rate": 8e-05,
+      "loss": 1.602,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6920179808042766,
+      "grad_norm": 0.464792937040329,
+      "learning_rate": 8e-05,
+      "loss": 1.8872,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6922609646458511,
+      "grad_norm": 0.4447360038757324,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6925039484874256,
+      "grad_norm": 0.4328846037387848,
+      "learning_rate": 8e-05,
+      "loss": 1.8175,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6927469323290001,
+      "grad_norm": 0.42637479305267334,
+      "learning_rate": 8e-05,
+      "loss": 1.8801,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6929899161705747,
+      "grad_norm": 0.4414830505847931,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6932329000121492,
+      "grad_norm": 0.4152732491493225,
+      "learning_rate": 8e-05,
+      "loss": 1.6778,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6934758838537237,
+      "grad_norm": 0.41810381412506104,
+      "learning_rate": 8e-05,
+      "loss": 1.7476,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6937188676952982,
+      "grad_norm": 0.4494502544403076,
+      "learning_rate": 8e-05,
+      "loss": 1.8524,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6939618515368728,
+      "grad_norm": 0.4219508767127991,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6942048353784473,
+      "grad_norm": 0.41995659470558167,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6944478192200219,
+      "grad_norm": 0.42812418937683105,
+      "learning_rate": 8e-05,
+      "loss": 1.7401,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6946908030615964,
+      "grad_norm": 0.42091864347457886,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 2859
+    },
+    {
+      "epoch": 0.694933786903171,
+      "grad_norm": 0.4300447702407837,
+      "learning_rate": 8e-05,
+      "loss": 1.7704,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6951767707447455,
+      "grad_norm": 0.4214186668395996,
+      "learning_rate": 8e-05,
+      "loss": 1.6359,
+      "step": 2861
+    },
+    {
+      "epoch": 0.69541975458632,
+      "grad_norm": 0.41487830877304077,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6956627384278945,
+      "grad_norm": 0.45346730947494507,
+      "learning_rate": 8e-05,
+      "loss": 1.7881,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6959057222694691,
+      "grad_norm": 0.43338078260421753,
+      "learning_rate": 8e-05,
+      "loss": 1.7183,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6961487061110436,
+      "grad_norm": 0.42131903767585754,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6963916899526181,
+      "grad_norm": 0.4134434759616852,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6966346737941926,
+      "grad_norm": 0.40478745102882385,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6968776576357673,
+      "grad_norm": 0.4301908016204834,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6971206414773418,
+      "grad_norm": 0.4128112196922302,
+      "learning_rate": 8e-05,
+      "loss": 1.5737,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6973636253189163,
+      "grad_norm": 0.4255407452583313,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6976066091604908,
+      "grad_norm": 0.4428727328777313,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6978495930020654,
+      "grad_norm": 0.4516085684299469,
+      "learning_rate": 8e-05,
+      "loss": 1.7717,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6980925768436399,
+      "grad_norm": 0.4255984127521515,
+      "learning_rate": 8e-05,
+      "loss": 1.7439,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6983355606852144,
+      "grad_norm": 0.43330180644989014,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6985785445267889,
+      "grad_norm": 0.4267866909503937,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6988215283683635,
+      "grad_norm": 0.43833673000335693,
+      "learning_rate": 8e-05,
+      "loss": 1.7438,
+      "step": 2876
+    },
+    {
+      "epoch": 0.699064512209938,
+      "grad_norm": 0.4407341778278351,
+      "learning_rate": 8e-05,
+      "loss": 1.8281,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6993074960515125,
+      "grad_norm": 0.43140849471092224,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6995504798930872,
+      "grad_norm": 0.42264214158058167,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6997934637346617,
+      "grad_norm": 0.44713887572288513,
+      "learning_rate": 8e-05,
+      "loss": 1.6725,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7000364475762362,
+      "grad_norm": 0.40955549478530884,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 2881
+    },
+    {
+      "epoch": 0.7002794314178107,
+      "grad_norm": 0.4211174547672272,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 2882
+    },
+    {
+      "epoch": 0.7005224152593853,
+      "grad_norm": 0.4128093719482422,
+      "learning_rate": 8e-05,
+      "loss": 1.5771,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7007653991009598,
+      "grad_norm": 0.4348185956478119,
+      "learning_rate": 8e-05,
+      "loss": 1.668,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7010083829425343,
+      "grad_norm": 0.40509769320487976,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7012513667841088,
+      "grad_norm": 0.4402783513069153,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7014943506256834,
+      "grad_norm": 0.4376557469367981,
+      "learning_rate": 8e-05,
+      "loss": 1.616,
+      "step": 2887
+    },
+    {
+      "epoch": 0.701737334467258,
+      "grad_norm": 0.4178268015384674,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7019803183088325,
+      "grad_norm": 0.45564791560173035,
+      "learning_rate": 8e-05,
+      "loss": 1.9045,
+      "step": 2889
+    },
+    {
+      "epoch": 0.702223302150407,
+      "grad_norm": 0.4381805658340454,
+      "learning_rate": 8e-05,
+      "loss": 1.8043,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7024662859919816,
+      "grad_norm": 0.4490160644054413,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7027092698335561,
+      "grad_norm": 0.43252772092819214,
+      "learning_rate": 8e-05,
+      "loss": 1.7327,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7029522536751306,
+      "grad_norm": 0.42838671803474426,
+      "learning_rate": 8e-05,
+      "loss": 1.606,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7031952375167051,
+      "grad_norm": 0.4459609091281891,
+      "learning_rate": 8e-05,
+      "loss": 1.6448,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7034382213582797,
+      "grad_norm": 0.43652117252349854,
+      "learning_rate": 8e-05,
+      "loss": 1.7875,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7036812051998542,
+      "grad_norm": 0.4296235740184784,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7039241890414287,
+      "grad_norm": 0.41683951020240784,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7041671728830032,
+      "grad_norm": 0.44808632135391235,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7044101567245779,
+      "grad_norm": 0.42905285954475403,
+      "learning_rate": 8e-05,
+      "loss": 1.6716,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7046531405661524,
+      "grad_norm": 0.43325552344322205,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7048961244077269,
+      "grad_norm": 0.44431445002555847,
+      "learning_rate": 8e-05,
+      "loss": 1.6893,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7051391082493014,
+      "grad_norm": 0.42025554180145264,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 2902
+    },
+    {
+      "epoch": 0.705382092090876,
+      "grad_norm": 0.4343130588531494,
+      "learning_rate": 8e-05,
+      "loss": 1.661,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7056250759324505,
+      "grad_norm": 0.43845680356025696,
+      "learning_rate": 8e-05,
+      "loss": 1.7959,
+      "step": 2904
+    },
+    {
+      "epoch": 0.705868059774025,
+      "grad_norm": 0.4372212588787079,
+      "learning_rate": 8e-05,
+      "loss": 1.6534,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7061110436155995,
+      "grad_norm": 0.41018640995025635,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7063540274571741,
+      "grad_norm": 0.43775278329849243,
+      "learning_rate": 8e-05,
+      "loss": 1.7379,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7065970112987486,
+      "grad_norm": 0.4367392361164093,
+      "learning_rate": 8e-05,
+      "loss": 1.6879,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7068399951403231,
+      "grad_norm": 0.4314105212688446,
+      "learning_rate": 8e-05,
+      "loss": 1.8476,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7070829789818978,
+      "grad_norm": 0.485707551240921,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7073259628234723,
+      "grad_norm": 0.43190839886665344,
+      "learning_rate": 8e-05,
+      "loss": 1.6327,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7075689466650468,
+      "grad_norm": 0.4512678384780884,
+      "learning_rate": 8e-05,
+      "loss": 1.813,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7078119305066213,
+      "grad_norm": 0.4097670614719391,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7080549143481959,
+      "grad_norm": 0.44747909903526306,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7082978981897704,
+      "grad_norm": 0.43108290433883667,
+      "learning_rate": 8e-05,
+      "loss": 1.6285,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7085408820313449,
+      "grad_norm": 0.4371739625930786,
+      "learning_rate": 8e-05,
+      "loss": 1.7504,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7087838658729194,
+      "grad_norm": 0.42412033677101135,
+      "learning_rate": 8e-05,
+      "loss": 1.7164,
+      "step": 2917
+    },
+    {
+      "epoch": 0.709026849714494,
+      "grad_norm": 0.4221256375312805,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7092698335560685,
+      "grad_norm": 0.4389481246471405,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 2919
+    },
+    {
+      "epoch": 0.709512817397643,
+      "grad_norm": 0.42696237564086914,
+      "learning_rate": 8e-05,
+      "loss": 1.657,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7097558012392176,
+      "grad_norm": 0.44028544425964355,
+      "learning_rate": 8e-05,
+      "loss": 1.5956,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7099987850807922,
+      "grad_norm": 0.4438371956348419,
+      "learning_rate": 8e-05,
+      "loss": 1.6568,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7102417689223667,
+      "grad_norm": 0.46004337072372437,
+      "learning_rate": 8e-05,
+      "loss": 1.8141,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7104847527639412,
+      "grad_norm": 0.4228387773036957,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7107277366055157,
+      "grad_norm": 0.42438724637031555,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7109707204470903,
+      "grad_norm": 0.42045244574546814,
+      "learning_rate": 8e-05,
+      "loss": 1.6598,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7112137042886648,
+      "grad_norm": 0.41184768080711365,
+      "learning_rate": 8e-05,
+      "loss": 1.6808,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7114566881302393,
+      "grad_norm": 0.4627285897731781,
+      "learning_rate": 8e-05,
+      "loss": 1.8629,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7116996719718138,
+      "grad_norm": 0.4347243905067444,
+      "learning_rate": 8e-05,
+      "loss": 1.5495,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7119426558133884,
+      "grad_norm": 0.43453219532966614,
+      "learning_rate": 8e-05,
+      "loss": 1.6426,
+      "step": 2930
+    },
+    {
+      "epoch": 0.712185639654963,
+      "grad_norm": 0.4535922706127167,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7124286234965375,
+      "grad_norm": 0.42448949813842773,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 2932
+    },
+    {
+      "epoch": 0.712671607338112,
+      "grad_norm": 0.43342697620391846,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7129145911796866,
+      "grad_norm": 0.4601576030254364,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7131575750212611,
+      "grad_norm": 0.424893856048584,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7134005588628356,
+      "grad_norm": 0.4176729619503021,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7136435427044101,
+      "grad_norm": 0.4509654939174652,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7138865265459847,
+      "grad_norm": 0.4501565098762512,
+      "learning_rate": 8e-05,
+      "loss": 1.7936,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7141295103875592,
+      "grad_norm": 0.41188111901283264,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7143724942291337,
+      "grad_norm": 0.427900105714798,
+      "learning_rate": 8e-05,
+      "loss": 1.736,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7146154780707082,
+      "grad_norm": 0.440960168838501,
+      "learning_rate": 8e-05,
+      "loss": 1.733,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7148584619122829,
+      "grad_norm": 0.44734588265419006,
+      "learning_rate": 8e-05,
+      "loss": 1.753,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7151014457538574,
+      "grad_norm": 0.43494489789009094,
+      "learning_rate": 8e-05,
+      "loss": 1.6996,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7153444295954319,
+      "grad_norm": 0.4065965712070465,
+      "learning_rate": 8e-05,
+      "loss": 1.6001,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7155874134370065,
+      "grad_norm": 0.4299347996711731,
+      "learning_rate": 8e-05,
+      "loss": 1.6479,
+      "step": 2945
+    },
+    {
+      "epoch": 0.715830397278581,
+      "grad_norm": 0.4462677240371704,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7160733811201555,
+      "grad_norm": 0.4489884674549103,
+      "learning_rate": 8e-05,
+      "loss": 1.7024,
+      "step": 2947
+    },
+    {
+      "epoch": 0.71631636496173,
+      "grad_norm": 0.4260941743850708,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7165593488033046,
+      "grad_norm": 0.43853750824928284,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7168023326448791,
+      "grad_norm": 0.42961418628692627,
+      "learning_rate": 8e-05,
+      "loss": 1.7548,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7170453164864536,
+      "grad_norm": 0.4214061200618744,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7172883003280281,
+      "grad_norm": 0.4244808256626129,
+      "learning_rate": 8e-05,
+      "loss": 1.554,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7175312841696028,
+      "grad_norm": 0.43740496039390564,
+      "learning_rate": 8e-05,
+      "loss": 1.6833,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7177742680111773,
+      "grad_norm": 0.46729227900505066,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7180172518527518,
+      "grad_norm": 0.42788493633270264,
+      "learning_rate": 8e-05,
+      "loss": 1.6844,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7182602356943263,
+      "grad_norm": 0.4276179373264313,
+      "learning_rate": 8e-05,
+      "loss": 1.757,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7185032195359009,
+      "grad_norm": 0.4147915840148926,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7187462033774754,
+      "grad_norm": 0.40998321771621704,
+      "learning_rate": 8e-05,
+      "loss": 1.5085,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7189891872190499,
+      "grad_norm": 0.424966037273407,
+      "learning_rate": 8e-05,
+      "loss": 1.7004,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7192321710606244,
+      "grad_norm": 0.43425485491752625,
+      "learning_rate": 8e-05,
+      "loss": 1.7853,
+      "step": 2960
+    },
+    {
+      "epoch": 0.719475154902199,
+      "grad_norm": 0.41092637181282043,
+      "learning_rate": 8e-05,
+      "loss": 1.6375,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7197181387437736,
+      "grad_norm": 0.46216538548469543,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7199611225853481,
+      "grad_norm": 0.44380462169647217,
+      "learning_rate": 8e-05,
+      "loss": 1.7846,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7202041064269226,
+      "grad_norm": 0.4513455331325531,
+      "learning_rate": 8e-05,
+      "loss": 1.7916,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7204470902684972,
+      "grad_norm": 0.44574451446533203,
+      "learning_rate": 8e-05,
+      "loss": 1.7439,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7206900741100717,
+      "grad_norm": 0.4227807819843292,
+      "learning_rate": 8e-05,
+      "loss": 1.6335,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7209330579516462,
+      "grad_norm": 0.41064074635505676,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7211760417932207,
+      "grad_norm": 0.42750582098960876,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7214190256347953,
+      "grad_norm": 0.4347093403339386,
+      "learning_rate": 8e-05,
+      "loss": 1.7491,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7216620094763698,
+      "grad_norm": 0.4427696466445923,
+      "learning_rate": 8e-05,
+      "loss": 1.8455,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7219049933179443,
+      "grad_norm": 0.42750856280326843,
+      "learning_rate": 8e-05,
+      "loss": 1.5269,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7221479771595188,
+      "grad_norm": 0.44155558943748474,
+      "learning_rate": 8e-05,
+      "loss": 1.692,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7223909610010935,
+      "grad_norm": 0.41951698064804077,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 2973
+    },
+    {
+      "epoch": 0.722633944842668,
+      "grad_norm": 0.42673179507255554,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7228769286842425,
+      "grad_norm": 0.42192110419273376,
+      "learning_rate": 8e-05,
+      "loss": 1.7488,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7231199125258171,
+      "grad_norm": 0.468248575925827,
+      "learning_rate": 8e-05,
+      "loss": 1.7992,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7233628963673916,
+      "grad_norm": 0.4387557804584503,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7236058802089661,
+      "grad_norm": 0.4511510729789734,
+      "learning_rate": 8e-05,
+      "loss": 1.7328,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7238488640505406,
+      "grad_norm": 0.43469998240470886,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7240918478921152,
+      "grad_norm": 0.4242485761642456,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7243348317336897,
+      "grad_norm": 0.41362321376800537,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 2981
+    },
+    {
+      "epoch": 0.7245778155752642,
+      "grad_norm": 0.43939733505249023,
+      "learning_rate": 8e-05,
+      "loss": 1.717,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7248207994168387,
+      "grad_norm": 0.4643685519695282,
+      "learning_rate": 8e-05,
+      "loss": 1.8142,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7250637832584134,
+      "grad_norm": 0.44100499153137207,
+      "learning_rate": 8e-05,
+      "loss": 1.5271,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7253067670999879,
+      "grad_norm": 0.45618927478790283,
+      "learning_rate": 8e-05,
+      "loss": 1.7367,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7255497509415624,
+      "grad_norm": 0.4324251115322113,
+      "learning_rate": 8e-05,
+      "loss": 1.7618,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7257927347831369,
+      "grad_norm": 0.43455609679222107,
+      "learning_rate": 8e-05,
+      "loss": 1.6849,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7260357186247115,
+      "grad_norm": 0.4199559986591339,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 2988
+    },
+    {
+      "epoch": 0.726278702466286,
+      "grad_norm": 0.44585034251213074,
+      "learning_rate": 8e-05,
+      "loss": 1.8438,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7265216863078605,
+      "grad_norm": 0.43811097741127014,
+      "learning_rate": 8e-05,
+      "loss": 1.6698,
+      "step": 2990
+    },
+    {
+      "epoch": 0.726764670149435,
+      "grad_norm": 0.48330649733543396,
+      "learning_rate": 8e-05,
+      "loss": 1.7734,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7270076539910096,
+      "grad_norm": 0.42586347460746765,
+      "learning_rate": 8e-05,
+      "loss": 1.7094,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7272506378325841,
+      "grad_norm": 0.4502209722995758,
+      "learning_rate": 8e-05,
+      "loss": 1.7918,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7274936216741587,
+      "grad_norm": 0.4350888431072235,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7277366055157332,
+      "grad_norm": 0.4330673813819885,
+      "learning_rate": 8e-05,
+      "loss": 1.6156,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7279795893573078,
+      "grad_norm": 0.41999706625938416,
+      "learning_rate": 8e-05,
+      "loss": 1.6932,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7282225731988823,
+      "grad_norm": 0.41733643412590027,
+      "learning_rate": 8e-05,
+      "loss": 1.6847,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7284655570404568,
+      "grad_norm": 0.43410784006118774,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7287085408820313,
+      "grad_norm": 0.42501723766326904,
+      "learning_rate": 8e-05,
+      "loss": 1.726,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7289515247236059,
+      "grad_norm": 0.41584810614585876,
+      "learning_rate": 8e-05,
+      "loss": 1.6048,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7291945085651804,
+      "grad_norm": 0.4186627268791199,
+      "learning_rate": 8e-05,
+      "loss": 1.5412,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7294374924067549,
+      "grad_norm": 0.4327656030654907,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7296804762483294,
+      "grad_norm": 0.40826526284217834,
+      "learning_rate": 8e-05,
+      "loss": 1.6837,
+      "step": 3003
+    },
+    {
+      "epoch": 0.729923460089904,
+      "grad_norm": 0.40665408968925476,
+      "learning_rate": 8e-05,
+      "loss": 1.5319,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7301664439314786,
+      "grad_norm": 0.4290464520454407,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7304094277730531,
+      "grad_norm": 0.42804446816444397,
+      "learning_rate": 8e-05,
+      "loss": 1.6235,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7306524116146276,
+      "grad_norm": 0.4496900737285614,
+      "learning_rate": 8e-05,
+      "loss": 1.7438,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7308953954562022,
+      "grad_norm": 0.4434277415275574,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7311383792977767,
+      "grad_norm": 0.4178725481033325,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7313813631393512,
+      "grad_norm": 0.4228324592113495,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7316243469809258,
+      "grad_norm": 0.42910534143447876,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7318673308225003,
+      "grad_norm": 0.42847323417663574,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7321103146640748,
+      "grad_norm": 0.41808050870895386,
+      "learning_rate": 8e-05,
+      "loss": 1.6401,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7323532985056493,
+      "grad_norm": 0.4299980401992798,
+      "learning_rate": 8e-05,
+      "loss": 1.7154,
+      "step": 3014
+    },
+    {
+      "epoch": 0.732596282347224,
+      "grad_norm": 0.4240909516811371,
+      "learning_rate": 8e-05,
+      "loss": 1.6776,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7328392661887985,
+      "grad_norm": 0.4330441653728485,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 3016
+    },
+    {
+      "epoch": 0.733082250030373,
+      "grad_norm": 0.44527700543403625,
+      "learning_rate": 8e-05,
+      "loss": 1.7607,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7333252338719475,
+      "grad_norm": 0.4417915642261505,
+      "learning_rate": 8e-05,
+      "loss": 1.7774,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7335682177135221,
+      "grad_norm": 0.441995769739151,
+      "learning_rate": 8e-05,
+      "loss": 1.6627,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7338112015550966,
+      "grad_norm": 0.4372912645339966,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7340541853966711,
+      "grad_norm": 0.4331246316432953,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7342971692382456,
+      "grad_norm": 0.4364757835865021,
+      "learning_rate": 8e-05,
+      "loss": 1.7941,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7345401530798202,
+      "grad_norm": 0.4373632073402405,
+      "learning_rate": 8e-05,
+      "loss": 1.6754,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7347831369213947,
+      "grad_norm": 0.42427098751068115,
+      "learning_rate": 8e-05,
+      "loss": 1.3536,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7350261207629692,
+      "grad_norm": 0.45470893383026123,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7352691046045438,
+      "grad_norm": 0.43789541721343994,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7355120884461184,
+      "grad_norm": 0.4273962378501892,
+      "learning_rate": 8e-05,
+      "loss": 1.587,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7357550722876929,
+      "grad_norm": 0.4530348479747772,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7359980561292674,
+      "grad_norm": 0.42630070447921753,
+      "learning_rate": 8e-05,
+      "loss": 1.6689,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7362410399708419,
+      "grad_norm": 0.4710538685321808,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7364840238124165,
+      "grad_norm": 0.4394691288471222,
+      "learning_rate": 8e-05,
+      "loss": 1.7098,
+      "step": 3031
+    },
+    {
+      "epoch": 0.736727007653991,
+      "grad_norm": 0.45098766684532166,
+      "learning_rate": 8e-05,
+      "loss": 1.6402,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7369699914955655,
+      "grad_norm": 0.4359957277774811,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 3033
+    },
+    {
+      "epoch": 0.73721297533714,
+      "grad_norm": 0.4104204773902893,
+      "learning_rate": 8e-05,
+      "loss": 1.4112,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7374559591787146,
+      "grad_norm": 0.4279259741306305,
+      "learning_rate": 8e-05,
+      "loss": 1.6271,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7376989430202892,
+      "grad_norm": 0.42476096749305725,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7379419268618637,
+      "grad_norm": 0.43152886629104614,
+      "learning_rate": 8e-05,
+      "loss": 1.5593,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7381849107034382,
+      "grad_norm": 0.43815726041793823,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7384278945450128,
+      "grad_norm": 0.4386484622955322,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7386708783865873,
+      "grad_norm": 0.46938014030456543,
+      "learning_rate": 8e-05,
+      "loss": 1.7885,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7389138622281618,
+      "grad_norm": 0.43240004777908325,
+      "learning_rate": 8e-05,
+      "loss": 1.7559,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7391568460697364,
+      "grad_norm": 0.4541640281677246,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7393998299113109,
+      "grad_norm": 0.44231152534484863,
+      "learning_rate": 8e-05,
+      "loss": 1.7713,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7396428137528854,
+      "grad_norm": 0.41579318046569824,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7398857975944599,
+      "grad_norm": 0.4093439280986786,
+      "learning_rate": 8e-05,
+      "loss": 1.6525,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7401287814360346,
+      "grad_norm": 0.427938312292099,
+      "learning_rate": 8e-05,
+      "loss": 1.7584,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7403717652776091,
+      "grad_norm": 0.41801324486732483,
+      "learning_rate": 8e-05,
+      "loss": 1.5418,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7406147491191836,
+      "grad_norm": 0.4497818946838379,
+      "learning_rate": 8e-05,
+      "loss": 1.8233,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7408577329607581,
+      "grad_norm": 0.44359248876571655,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7411007168023327,
+      "grad_norm": 0.4452890157699585,
+      "learning_rate": 8e-05,
+      "loss": 1.8363,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7413437006439072,
+      "grad_norm": 0.40395015478134155,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7415866844854817,
+      "grad_norm": 0.44283580780029297,
+      "learning_rate": 8e-05,
+      "loss": 1.7597,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7418296683270562,
+      "grad_norm": 0.41657838225364685,
+      "learning_rate": 8e-05,
+      "loss": 1.5566,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7420726521686308,
+      "grad_norm": 0.4326663911342621,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7423156360102053,
+      "grad_norm": 0.44257473945617676,
+      "learning_rate": 8e-05,
+      "loss": 1.7503,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7425586198517798,
+      "grad_norm": 0.4406522810459137,
+      "learning_rate": 8e-05,
+      "loss": 1.7226,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7428016036933544,
+      "grad_norm": 0.42228004336357117,
+      "learning_rate": 8e-05,
+      "loss": 1.6661,
+      "step": 3057
+    },
+    {
+      "epoch": 0.743044587534929,
+      "grad_norm": 0.463959664106369,
+      "learning_rate": 8e-05,
+      "loss": 1.7243,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7432875713765035,
+      "grad_norm": 0.437701940536499,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 3059
+    },
+    {
+      "epoch": 0.743530555218078,
+      "grad_norm": 0.42580461502075195,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7437735390596525,
+      "grad_norm": 0.4411897361278534,
+      "learning_rate": 8e-05,
+      "loss": 1.7185,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7440165229012271,
+      "grad_norm": 0.4475369155406952,
+      "learning_rate": 8e-05,
+      "loss": 1.5211,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7442595067428016,
+      "grad_norm": 0.43206387758255005,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7445024905843761,
+      "grad_norm": 0.4343615770339966,
+      "learning_rate": 8e-05,
+      "loss": 1.7415,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7447454744259506,
+      "grad_norm": 0.44774171710014343,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7449884582675252,
+      "grad_norm": 0.464191734790802,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7452314421090998,
+      "grad_norm": 0.43260547518730164,
+      "learning_rate": 8e-05,
+      "loss": 1.7925,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7454744259506743,
+      "grad_norm": 0.4078400433063507,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7457174097922488,
+      "grad_norm": 0.46859222650527954,
+      "learning_rate": 8e-05,
+      "loss": 1.6666,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7459603936338234,
+      "grad_norm": 0.45912933349609375,
+      "learning_rate": 8e-05,
+      "loss": 1.8639,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7462033774753979,
+      "grad_norm": 0.4265579581260681,
+      "learning_rate": 8e-05,
+      "loss": 1.756,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7464463613169724,
+      "grad_norm": 0.4226361811161041,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7466893451585469,
+      "grad_norm": 0.4646567106246948,
+      "learning_rate": 8e-05,
+      "loss": 1.7741,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7469323290001215,
+      "grad_norm": 0.43147245049476624,
+      "learning_rate": 8e-05,
+      "loss": 1.7334,
+      "step": 3074
+    },
+    {
+      "epoch": 0.747175312841696,
+      "grad_norm": 0.42696601152420044,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7474182966832705,
+      "grad_norm": 0.4488403797149658,
+      "learning_rate": 8e-05,
+      "loss": 1.7166,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7476612805248452,
+      "grad_norm": 0.4200011193752289,
+      "learning_rate": 8e-05,
+      "loss": 1.6413,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7479042643664197,
+      "grad_norm": 0.42939868569374084,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7481472482079942,
+      "grad_norm": 0.43149253726005554,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7483902320495687,
+      "grad_norm": 0.42696458101272583,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7486332158911433,
+      "grad_norm": 0.4416796863079071,
+      "learning_rate": 8e-05,
+      "loss": 1.7001,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7488761997327178,
+      "grad_norm": 0.44286996126174927,
+      "learning_rate": 8e-05,
+      "loss": 1.6675,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7491191835742923,
+      "grad_norm": 0.43856683373451233,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7493621674158668,
+      "grad_norm": 0.42399197816848755,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7496051512574414,
+      "grad_norm": 0.4241805970668793,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7498481350990159,
+      "grad_norm": 0.45851489901542664,
+      "learning_rate": 8e-05,
+      "loss": 1.8442,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7500911189405904,
+      "grad_norm": 0.46060052514076233,
+      "learning_rate": 8e-05,
+      "loss": 1.6845,
+      "step": 3087
+    },
+    {
+      "epoch": 0.750334102782165,
+      "grad_norm": 0.44973620772361755,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7505770866237396,
+      "grad_norm": 0.44633743166923523,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7508200704653141,
+      "grad_norm": 0.4270108938217163,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7510630543068886,
+      "grad_norm": 0.4253171384334564,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7513060381484631,
+      "grad_norm": 0.44977208971977234,
+      "learning_rate": 8e-05,
+      "loss": 1.7461,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7515490219900377,
+      "grad_norm": 0.4196910262107849,
+      "learning_rate": 8e-05,
+      "loss": 1.6844,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7517920058316122,
+      "grad_norm": 0.43074601888656616,
+      "learning_rate": 8e-05,
+      "loss": 1.6475,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7520349896731867,
+      "grad_norm": 0.4223160743713379,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7522779735147612,
+      "grad_norm": 0.4304984509944916,
+      "learning_rate": 8e-05,
+      "loss": 1.7682,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7525209573563358,
+      "grad_norm": 0.4316643476486206,
+      "learning_rate": 8e-05,
+      "loss": 1.7005,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7527639411979103,
+      "grad_norm": 0.4488762617111206,
+      "learning_rate": 8e-05,
+      "loss": 1.7032,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7530069250394849,
+      "grad_norm": 0.4390949308872223,
+      "learning_rate": 8e-05,
+      "loss": 1.5894,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7532499088810594,
+      "grad_norm": 0.44235989451408386,
+      "learning_rate": 8e-05,
+      "loss": 1.7045,
+      "step": 3100
+    },
+    {
+      "epoch": 0.753492892722634,
+      "grad_norm": 0.46370869874954224,
+      "learning_rate": 8e-05,
+      "loss": 1.8514,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7537358765642085,
+      "grad_norm": 0.4291474223136902,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 3102
+    },
+    {
+      "epoch": 0.753978860405783,
+      "grad_norm": 0.43415555357933044,
+      "learning_rate": 8e-05,
+      "loss": 1.779,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7542218442473575,
+      "grad_norm": 0.40880295634269714,
+      "learning_rate": 8e-05,
+      "loss": 1.5144,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7544648280889321,
+      "grad_norm": 0.4524337351322174,
+      "learning_rate": 8e-05,
+      "loss": 1.7613,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7547078119305066,
+      "grad_norm": 0.42361971735954285,
+      "learning_rate": 8e-05,
+      "loss": 1.3676,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7549507957720811,
+      "grad_norm": 0.446455180644989,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7551937796136557,
+      "grad_norm": 0.43976864218711853,
+      "learning_rate": 8e-05,
+      "loss": 1.7526,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7554367634552303,
+      "grad_norm": 0.4394785165786743,
+      "learning_rate": 8e-05,
+      "loss": 1.6729,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7556797472968048,
+      "grad_norm": 0.42967545986175537,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7559227311383793,
+      "grad_norm": 0.42354637384414673,
+      "learning_rate": 8e-05,
+      "loss": 1.6038,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7561657149799539,
+      "grad_norm": 0.45152711868286133,
+      "learning_rate": 8e-05,
+      "loss": 1.8139,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7564086988215284,
+      "grad_norm": 0.434503972530365,
+      "learning_rate": 8e-05,
+      "loss": 1.5586,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7566516826631029,
+      "grad_norm": 0.4203075170516968,
+      "learning_rate": 8e-05,
+      "loss": 1.7284,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7568946665046774,
+      "grad_norm": 0.44629746675491333,
+      "learning_rate": 8e-05,
+      "loss": 1.866,
+      "step": 3115
+    },
+    {
+      "epoch": 0.757137650346252,
+      "grad_norm": 0.4523211121559143,
+      "learning_rate": 8e-05,
+      "loss": 1.5863,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7573806341878265,
+      "grad_norm": 0.4517972469329834,
+      "learning_rate": 8e-05,
+      "loss": 1.6627,
+      "step": 3117
+    },
+    {
+      "epoch": 0.757623618029401,
+      "grad_norm": 0.41653215885162354,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7578666018709755,
+      "grad_norm": 0.44581496715545654,
+      "learning_rate": 8e-05,
+      "loss": 1.7048,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7581095857125502,
+      "grad_norm": 0.4377984404563904,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7583525695541247,
+      "grad_norm": 0.45606938004493713,
+      "learning_rate": 8e-05,
+      "loss": 1.8151,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7585955533956992,
+      "grad_norm": 0.4326377213001251,
+      "learning_rate": 8e-05,
+      "loss": 1.7019,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7588385372372737,
+      "grad_norm": 0.42819643020629883,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7590815210788483,
+      "grad_norm": 0.44774511456489563,
+      "learning_rate": 8e-05,
+      "loss": 1.7564,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7593245049204228,
+      "grad_norm": 0.440573513507843,
+      "learning_rate": 8e-05,
+      "loss": 1.7528,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7595674887619973,
+      "grad_norm": 0.4250212609767914,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7598104726035718,
+      "grad_norm": 0.44139567017555237,
+      "learning_rate": 8e-05,
+      "loss": 1.7481,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7600534564451464,
+      "grad_norm": 0.43515217304229736,
+      "learning_rate": 8e-05,
+      "loss": 1.6272,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7602964402867209,
+      "grad_norm": 0.4104181230068207,
+      "learning_rate": 8e-05,
+      "loss": 1.6047,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7605394241282954,
+      "grad_norm": 0.4388481676578522,
+      "learning_rate": 8e-05,
+      "loss": 1.7303,
+      "step": 3130
+    },
+    {
+      "epoch": 0.76078240796987,
+      "grad_norm": 0.4457818567752838,
+      "learning_rate": 8e-05,
+      "loss": 1.5728,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7610253918114446,
+      "grad_norm": 0.4531154930591583,
+      "learning_rate": 8e-05,
+      "loss": 1.6563,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7612683756530191,
+      "grad_norm": 0.4454151391983032,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7615113594945936,
+      "grad_norm": 0.4301358759403229,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7617543433361681,
+      "grad_norm": 0.4423365890979767,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7619973271777427,
+      "grad_norm": 0.44255703687667847,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7622403110193172,
+      "grad_norm": 0.4629448354244232,
+      "learning_rate": 8e-05,
+      "loss": 1.7387,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7624832948608917,
+      "grad_norm": 0.4280916154384613,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7627262787024662,
+      "grad_norm": 0.45554330945014954,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7629692625440408,
+      "grad_norm": 0.4304593801498413,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7632122463856154,
+      "grad_norm": 0.43380171060562134,
+      "learning_rate": 8e-05,
+      "loss": 1.6719,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7634552302271899,
+      "grad_norm": 0.44206365942955017,
+      "learning_rate": 8e-05,
+      "loss": 1.7635,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7636982140687645,
+      "grad_norm": 0.42580685019493103,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 3143
+    },
+    {
+      "epoch": 0.763941197910339,
+      "grad_norm": 0.44004255533218384,
+      "learning_rate": 8e-05,
+      "loss": 1.6932,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7641841817519135,
+      "grad_norm": 0.43854469060897827,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 3145
+    },
+    {
+      "epoch": 0.764427165593488,
+      "grad_norm": 0.41173112392425537,
+      "learning_rate": 8e-05,
+      "loss": 1.4801,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7646701494350626,
+      "grad_norm": 0.4296555817127228,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7649131332766371,
+      "grad_norm": 0.4560660719871521,
+      "learning_rate": 8e-05,
+      "loss": 1.7847,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7651561171182116,
+      "grad_norm": 0.4360103905200958,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7653991009597861,
+      "grad_norm": 0.4256157875061035,
+      "learning_rate": 8e-05,
+      "loss": 1.7381,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7656420848013608,
+      "grad_norm": 0.42012447118759155,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7658850686429353,
+      "grad_norm": 0.45626184344291687,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7661280524845098,
+      "grad_norm": 0.42164766788482666,
+      "learning_rate": 8e-05,
+      "loss": 1.5111,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7663710363260843,
+      "grad_norm": 0.42887359857559204,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7666140201676589,
+      "grad_norm": 0.44160687923431396,
+      "learning_rate": 8e-05,
+      "loss": 1.7249,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7668570040092334,
+      "grad_norm": 0.4335710406303406,
+      "learning_rate": 8e-05,
+      "loss": 1.7495,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7670999878508079,
+      "grad_norm": 0.4323950409889221,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7673429716923824,
+      "grad_norm": 0.43719926476478577,
+      "learning_rate": 8e-05,
+      "loss": 1.7419,
+      "step": 3158
+    },
+    {
+      "epoch": 0.767585955533957,
+      "grad_norm": 0.42044708132743835,
+      "learning_rate": 8e-05,
+      "loss": 1.7561,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7678289393755315,
+      "grad_norm": 0.449849933385849,
+      "learning_rate": 8e-05,
+      "loss": 1.7707,
+      "step": 3160
+    },
+    {
+      "epoch": 0.768071923217106,
+      "grad_norm": 0.42334234714508057,
+      "learning_rate": 8e-05,
+      "loss": 1.7083,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7683149070586806,
+      "grad_norm": 0.39916276931762695,
+      "learning_rate": 8e-05,
+      "loss": 1.4826,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7685578909002552,
+      "grad_norm": 0.4090518057346344,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7688008747418297,
+      "grad_norm": 0.45016583800315857,
+      "learning_rate": 8e-05,
+      "loss": 1.8095,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7690438585834042,
+      "grad_norm": 0.4227582812309265,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7692868424249787,
+      "grad_norm": 0.42541226744651794,
+      "learning_rate": 8e-05,
+      "loss": 1.5962,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7695298262665533,
+      "grad_norm": 0.4386545419692993,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7697728101081278,
+      "grad_norm": 0.43350985646247864,
+      "learning_rate": 8e-05,
+      "loss": 1.6121,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7700157939497023,
+      "grad_norm": 0.4352431297302246,
+      "learning_rate": 8e-05,
+      "loss": 1.7684,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7702587777912768,
+      "grad_norm": 0.4648544490337372,
+      "learning_rate": 8e-05,
+      "loss": 1.8036,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7705017616328514,
+      "grad_norm": 0.45580625534057617,
+      "learning_rate": 8e-05,
+      "loss": 1.7472,
+      "step": 3171
+    },
+    {
+      "epoch": 0.770744745474426,
+      "grad_norm": 0.43300363421440125,
+      "learning_rate": 8e-05,
+      "loss": 1.6563,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7709877293160005,
+      "grad_norm": 0.45607712864875793,
+      "learning_rate": 8e-05,
+      "loss": 1.7953,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7712307131575751,
+      "grad_norm": 0.4287468492984772,
+      "learning_rate": 8e-05,
+      "loss": 1.6988,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7714736969991496,
+      "grad_norm": 0.4145059883594513,
+      "learning_rate": 8e-05,
+      "loss": 1.473,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7717166808407241,
+      "grad_norm": 0.44404569268226624,
+      "learning_rate": 8e-05,
+      "loss": 1.6417,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7719596646822986,
+      "grad_norm": 0.4310551881790161,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7722026485238732,
+      "grad_norm": 0.44795772433280945,
+      "learning_rate": 8e-05,
+      "loss": 1.8037,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7724456323654477,
+      "grad_norm": 0.4437018930912018,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7726886162070222,
+      "grad_norm": 0.4506959319114685,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7729316000485967,
+      "grad_norm": 0.4333343505859375,
+      "learning_rate": 8e-05,
+      "loss": 1.6426,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7731745838901714,
+      "grad_norm": 0.413421094417572,
+      "learning_rate": 8e-05,
+      "loss": 1.4378,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7734175677317459,
+      "grad_norm": 0.4241701662540436,
+      "learning_rate": 8e-05,
+      "loss": 1.6437,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7736605515733204,
+      "grad_norm": 0.4190889596939087,
+      "learning_rate": 8e-05,
+      "loss": 1.6533,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7739035354148949,
+      "grad_norm": 0.41836678981781006,
+      "learning_rate": 8e-05,
+      "loss": 1.5168,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7741465192564695,
+      "grad_norm": 0.4503277838230133,
+      "learning_rate": 8e-05,
+      "loss": 1.8305,
+      "step": 3186
+    },
+    {
+      "epoch": 0.774389503098044,
+      "grad_norm": 0.4403499662876129,
+      "learning_rate": 8e-05,
+      "loss": 1.7695,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7746324869396185,
+      "grad_norm": 0.4220961034297943,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 3188
+    },
+    {
+      "epoch": 0.774875470781193,
+      "grad_norm": 0.43033790588378906,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7751184546227676,
+      "grad_norm": 0.4296756982803345,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7753614384643421,
+      "grad_norm": 0.4394514262676239,
+      "learning_rate": 8e-05,
+      "loss": 1.8112,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7756044223059166,
+      "grad_norm": 0.45250973105430603,
+      "learning_rate": 8e-05,
+      "loss": 1.767,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7758474061474911,
+      "grad_norm": 0.4266514778137207,
+      "learning_rate": 8e-05,
+      "loss": 1.6097,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7760903899890658,
+      "grad_norm": 0.44215261936187744,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7763333738306403,
+      "grad_norm": 0.5195906758308411,
+      "learning_rate": 8e-05,
+      "loss": 1.6998,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7765763576722148,
+      "grad_norm": 0.43021008372306824,
+      "learning_rate": 8e-05,
+      "loss": 1.7268,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7768193415137893,
+      "grad_norm": 0.4463469088077545,
+      "learning_rate": 8e-05,
+      "loss": 1.6953,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7770623253553639,
+      "grad_norm": 0.4469200074672699,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7773053091969384,
+      "grad_norm": 0.44574251770973206,
+      "learning_rate": 8e-05,
+      "loss": 1.7205,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7775482930385129,
+      "grad_norm": 0.4170497953891754,
+      "learning_rate": 8e-05,
+      "loss": 1.7066,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7777912768800874,
+      "grad_norm": 0.42403534054756165,
+      "learning_rate": 8e-05,
+      "loss": 1.6283,
+      "step": 3201
+    },
+    {
+      "epoch": 0.778034260721662,
+      "grad_norm": 0.4213058054447174,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7782772445632365,
+      "grad_norm": 0.434829443693161,
+      "learning_rate": 8e-05,
+      "loss": 1.6177,
+      "step": 3203
+    },
+    {
+      "epoch": 0.778520228404811,
+      "grad_norm": 0.3992372453212738,
+      "learning_rate": 8e-05,
+      "loss": 1.5494,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7787632122463856,
+      "grad_norm": 0.41837990283966064,
+      "learning_rate": 8e-05,
+      "loss": 1.6525,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7790061960879602,
+      "grad_norm": 0.42778345942497253,
+      "learning_rate": 8e-05,
+      "loss": 1.5802,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7792491799295347,
+      "grad_norm": 0.4075702726840973,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7794921637711092,
+      "grad_norm": 0.4612067639827728,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7797351476126838,
+      "grad_norm": 0.4795321524143219,
+      "learning_rate": 8e-05,
+      "loss": 1.724,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7799781314542583,
+      "grad_norm": 0.4442792534828186,
+      "learning_rate": 8e-05,
+      "loss": 1.8113,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7802211152958328,
+      "grad_norm": 0.4258017838001251,
+      "learning_rate": 8e-05,
+      "loss": 1.6777,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7804640991374073,
+      "grad_norm": 0.4336091876029968,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 3212
+    },
+    {
+      "epoch": 0.780707082978982,
+      "grad_norm": 0.4324623644351959,
+      "learning_rate": 8e-05,
+      "loss": 1.7787,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7809500668205565,
+      "grad_norm": 0.4305349886417389,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 3214
+    },
+    {
+      "epoch": 0.781193050662131,
+      "grad_norm": 0.4140620827674866,
+      "learning_rate": 8e-05,
+      "loss": 1.4616,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7814360345037055,
+      "grad_norm": 0.4296761155128479,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7816790183452801,
+      "grad_norm": 0.4193679094314575,
+      "learning_rate": 8e-05,
+      "loss": 1.4924,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7819220021868546,
+      "grad_norm": 0.4483497738838196,
+      "learning_rate": 8e-05,
+      "loss": 1.5761,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7821649860284291,
+      "grad_norm": 0.4483656585216522,
+      "learning_rate": 8e-05,
+      "loss": 1.7218,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7824079698700036,
+      "grad_norm": 0.43820837140083313,
+      "learning_rate": 8e-05,
+      "loss": 1.7265,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7826509537115782,
+      "grad_norm": 0.4398447573184967,
+      "learning_rate": 8e-05,
+      "loss": 1.6056,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7828939375531527,
+      "grad_norm": 0.4669850468635559,
+      "learning_rate": 8e-05,
+      "loss": 1.7997,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7831369213947272,
+      "grad_norm": 0.46327561140060425,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7833799052363017,
+      "grad_norm": 0.46118444204330444,
+      "learning_rate": 8e-05,
+      "loss": 1.808,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7836228890778764,
+      "grad_norm": 0.4404383599758148,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7838658729194509,
+      "grad_norm": 0.4344371259212494,
+      "learning_rate": 8e-05,
+      "loss": 1.547,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7841088567610254,
+      "grad_norm": 0.42703741788864136,
+      "learning_rate": 8e-05,
+      "loss": 1.5573,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7843518406025999,
+      "grad_norm": 0.43044766783714294,
+      "learning_rate": 8e-05,
+      "loss": 1.7356,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7845948244441745,
+      "grad_norm": 0.41756537556648254,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 3229
+    },
+    {
+      "epoch": 0.784837808285749,
+      "grad_norm": 0.45819297432899475,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7850807921273235,
+      "grad_norm": 0.4413856863975525,
+      "learning_rate": 8e-05,
+      "loss": 1.7679,
+      "step": 3231
+    },
+    {
+      "epoch": 0.785323775968898,
+      "grad_norm": 0.4363550543785095,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7855667598104726,
+      "grad_norm": 0.4228973686695099,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7858097436520471,
+      "grad_norm": 0.46946898102760315,
+      "learning_rate": 8e-05,
+      "loss": 1.6207,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7860527274936217,
+      "grad_norm": 0.4272329807281494,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7862957113351962,
+      "grad_norm": 0.4374501407146454,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7865386951767708,
+      "grad_norm": 0.4361892640590668,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7867816790183453,
+      "grad_norm": 0.4813070297241211,
+      "learning_rate": 8e-05,
+      "loss": 1.9879,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7870246628599198,
+      "grad_norm": 0.44477158784866333,
+      "learning_rate": 8e-05,
+      "loss": 1.7072,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7872676467014944,
+      "grad_norm": 0.4351281225681305,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7875106305430689,
+      "grad_norm": 0.40754440426826477,
+      "learning_rate": 8e-05,
+      "loss": 1.5814,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7877536143846434,
+      "grad_norm": 0.458784282207489,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7879965982262179,
+      "grad_norm": 0.41207802295684814,
+      "learning_rate": 8e-05,
+      "loss": 1.5416,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7882395820677925,
+      "grad_norm": 0.41917839646339417,
+      "learning_rate": 8e-05,
+      "loss": 1.4672,
+      "step": 3244
+    },
+    {
+      "epoch": 0.788482565909367,
+      "grad_norm": 0.43034201860427856,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7887255497509416,
+      "grad_norm": 0.4569275975227356,
+      "learning_rate": 8e-05,
+      "loss": 1.6384,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7889685335925161,
+      "grad_norm": 0.429230272769928,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7892115174340907,
+      "grad_norm": 0.43947574496269226,
+      "learning_rate": 8e-05,
+      "loss": 1.5794,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7894545012756652,
+      "grad_norm": 0.4172064960002899,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7896974851172397,
+      "grad_norm": 0.4189060628414154,
+      "learning_rate": 8e-05,
+      "loss": 1.6668,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7899404689588142,
+      "grad_norm": 0.4146542549133301,
+      "learning_rate": 8e-05,
+      "loss": 1.5907,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7901834528003888,
+      "grad_norm": 0.4231383204460144,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7904264366419633,
+      "grad_norm": 0.44231081008911133,
+      "learning_rate": 8e-05,
+      "loss": 1.7581,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7906694204835378,
+      "grad_norm": 0.4153212308883667,
+      "learning_rate": 8e-05,
+      "loss": 1.6327,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7909124043251123,
+      "grad_norm": 0.4461350739002228,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 3255
+    },
+    {
+      "epoch": 0.791155388166687,
+      "grad_norm": 0.441765695810318,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7913983720082615,
+      "grad_norm": 0.4294467270374298,
+      "learning_rate": 8e-05,
+      "loss": 1.7552,
+      "step": 3257
+    },
+    {
+      "epoch": 0.791641355849836,
+      "grad_norm": 0.42167702317237854,
+      "learning_rate": 8e-05,
+      "loss": 1.7205,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7918843396914105,
+      "grad_norm": 0.4077739715576172,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7921273235329851,
+      "grad_norm": 0.4522324502468109,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7923703073745596,
+      "grad_norm": 0.43168991804122925,
+      "learning_rate": 8e-05,
+      "loss": 1.6474,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7926132912161341,
+      "grad_norm": 0.4323519170284271,
+      "learning_rate": 8e-05,
+      "loss": 1.7567,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7928562750577086,
+      "grad_norm": 0.4322536587715149,
+      "learning_rate": 8e-05,
+      "loss": 1.7115,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7930992588992832,
+      "grad_norm": 0.4630602300167084,
+      "learning_rate": 8e-05,
+      "loss": 1.6845,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7933422427408577,
+      "grad_norm": 0.4112022817134857,
+      "learning_rate": 8e-05,
+      "loss": 1.5128,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7935852265824322,
+      "grad_norm": 0.46825671195983887,
+      "learning_rate": 8e-05,
+      "loss": 1.8214,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7938282104240068,
+      "grad_norm": 0.46098440885543823,
+      "learning_rate": 8e-05,
+      "loss": 1.8533,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7940711942655814,
+      "grad_norm": 0.41676875948905945,
+      "learning_rate": 8e-05,
+      "loss": 1.4883,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7943141781071559,
+      "grad_norm": 0.43512633442878723,
+      "learning_rate": 8e-05,
+      "loss": 1.5937,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7945571619487304,
+      "grad_norm": 0.4570131003856659,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7948001457903049,
+      "grad_norm": 0.4561043679714203,
+      "learning_rate": 8e-05,
+      "loss": 1.8821,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7950431296318795,
+      "grad_norm": 0.4399861991405487,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 3272
+    },
+    {
+      "epoch": 0.795286113473454,
+      "grad_norm": 0.41973876953125,
+      "learning_rate": 8e-05,
+      "loss": 1.6697,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7955290973150285,
+      "grad_norm": 0.4508086144924164,
+      "learning_rate": 8e-05,
+      "loss": 1.7316,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7957720811566031,
+      "grad_norm": 0.4166657626628876,
+      "learning_rate": 8e-05,
+      "loss": 1.5692,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7960150649981776,
+      "grad_norm": 0.45421019196510315,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7962580488397522,
+      "grad_norm": 0.43804869055747986,
+      "learning_rate": 8e-05,
+      "loss": 1.5893,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7965010326813267,
+      "grad_norm": 0.43674108386039734,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7967440165229013,
+      "grad_norm": 0.4358680248260498,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7969870003644758,
+      "grad_norm": 0.4250020682811737,
+      "learning_rate": 8e-05,
+      "loss": 1.5795,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7972299842060503,
+      "grad_norm": 0.42309874296188354,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7974729680476248,
+      "grad_norm": 0.4376814365386963,
+      "learning_rate": 8e-05,
+      "loss": 1.6276,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7977159518891994,
+      "grad_norm": 0.41417595744132996,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7979589357307739,
+      "grad_norm": 0.4405534565448761,
+      "learning_rate": 8e-05,
+      "loss": 1.7535,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7982019195723484,
+      "grad_norm": 0.43022340536117554,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7984449034139229,
+      "grad_norm": 0.4190630316734314,
+      "learning_rate": 8e-05,
+      "loss": 1.6549,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7986878872554976,
+      "grad_norm": 0.45187270641326904,
+      "learning_rate": 8e-05,
+      "loss": 1.7265,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7989308710970721,
+      "grad_norm": 0.4494510591030121,
+      "learning_rate": 8e-05,
+      "loss": 1.681,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7991738549386466,
+      "grad_norm": 0.43418967723846436,
+      "learning_rate": 8e-05,
+      "loss": 1.7563,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7994168387802211,
+      "grad_norm": 0.46451711654663086,
+      "learning_rate": 8e-05,
+      "loss": 1.7573,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7996598226217957,
+      "grad_norm": 0.43015822768211365,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7999028064633702,
+      "grad_norm": 0.4084012806415558,
+      "learning_rate": 8e-05,
+      "loss": 1.5914,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8001457903049447,
+      "grad_norm": 0.42597126960754395,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8003887741465192,
+      "grad_norm": 0.4301321804523468,
+      "learning_rate": 8e-05,
+      "loss": 1.5777,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8006317579880938,
+      "grad_norm": 0.42764273285865784,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8008747418296683,
+      "grad_norm": 0.43151944875717163,
+      "learning_rate": 8e-05,
+      "loss": 1.5817,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8011177256712428,
+      "grad_norm": 0.4275348484516144,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8013607095128173,
+      "grad_norm": 0.4620358943939209,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 3298
+    },
+    {
+      "epoch": 0.801603693354392,
+      "grad_norm": 0.4136797785758972,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8018466771959665,
+      "grad_norm": 0.4344896078109741,
+      "learning_rate": 8e-05,
+      "loss": 1.7279,
+      "step": 3300
+    },
+    {
+      "epoch": 0.802089661037541,
+      "grad_norm": 0.4327981770038605,
+      "learning_rate": 8e-05,
+      "loss": 1.6992,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8023326448791155,
+      "grad_norm": 0.43653398752212524,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8025756287206901,
+      "grad_norm": 0.3973734974861145,
+      "learning_rate": 8e-05,
+      "loss": 1.3584,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8028186125622646,
+      "grad_norm": 0.4527641534805298,
+      "learning_rate": 8e-05,
+      "loss": 1.7162,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8030615964038391,
+      "grad_norm": 0.4239053726196289,
+      "learning_rate": 8e-05,
+      "loss": 1.6993,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8033045802454137,
+      "grad_norm": 0.40652742981910706,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8035475640869882,
+      "grad_norm": 0.4082186222076416,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8037905479285627,
+      "grad_norm": 0.43346378207206726,
+      "learning_rate": 8e-05,
+      "loss": 1.707,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8040335317701373,
+      "grad_norm": 0.4594515562057495,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8042765156117119,
+      "grad_norm": 0.4392218291759491,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8045194994532864,
+      "grad_norm": 0.4414903223514557,
+      "learning_rate": 8e-05,
+      "loss": 1.6698,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8047624832948609,
+      "grad_norm": 0.43082642555236816,
+      "learning_rate": 8e-05,
+      "loss": 1.7035,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8050054671364354,
+      "grad_norm": 0.43654537200927734,
+      "learning_rate": 8e-05,
+      "loss": 1.7559,
+      "step": 3313
+    },
+    {
+      "epoch": 0.80524845097801,
+      "grad_norm": 0.4272270202636719,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8054914348195845,
+      "grad_norm": 0.44494035840034485,
+      "learning_rate": 8e-05,
+      "loss": 1.6353,
+      "step": 3315
+    },
+    {
+      "epoch": 0.805734418661159,
+      "grad_norm": 0.4131324291229248,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8059774025027335,
+      "grad_norm": 0.432865172624588,
+      "learning_rate": 8e-05,
+      "loss": 1.6906,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8062203863443081,
+      "grad_norm": 0.4416402578353882,
+      "learning_rate": 8e-05,
+      "loss": 1.7476,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8064633701858827,
+      "grad_norm": 0.42770132422447205,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8067063540274572,
+      "grad_norm": 0.4086218476295471,
+      "learning_rate": 8e-05,
+      "loss": 1.5342,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8069493378690317,
+      "grad_norm": 0.43067216873168945,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8071923217106063,
+      "grad_norm": 0.4209299087524414,
+      "learning_rate": 8e-05,
+      "loss": 1.7403,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8074353055521808,
+      "grad_norm": 0.41420069336891174,
+      "learning_rate": 8e-05,
+      "loss": 1.5766,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8076782893937553,
+      "grad_norm": 0.43512585759162903,
+      "learning_rate": 8e-05,
+      "loss": 1.7533,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8079212732353298,
+      "grad_norm": 0.4354604184627533,
+      "learning_rate": 8e-05,
+      "loss": 1.7222,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8081642570769044,
+      "grad_norm": 0.42139172554016113,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8084072409184789,
+      "grad_norm": 0.4381217658519745,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8086502247600534,
+      "grad_norm": 0.4311729669570923,
+      "learning_rate": 8e-05,
+      "loss": 1.5146,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8088932086016279,
+      "grad_norm": 0.4573782682418823,
+      "learning_rate": 8e-05,
+      "loss": 1.7139,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8091361924432026,
+      "grad_norm": 0.42885854840278625,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8093791762847771,
+      "grad_norm": 0.4713359475135803,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8096221601263516,
+      "grad_norm": 0.4129873514175415,
+      "learning_rate": 8e-05,
+      "loss": 1.7101,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8098651439679261,
+      "grad_norm": 0.4083806574344635,
+      "learning_rate": 8e-05,
+      "loss": 1.4683,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8101081278095007,
+      "grad_norm": 0.44849252700805664,
+      "learning_rate": 8e-05,
+      "loss": 1.7669,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8103511116510752,
+      "grad_norm": 0.42854878306388855,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8105940954926497,
+      "grad_norm": 0.4260036051273346,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8108370793342242,
+      "grad_norm": 0.4234296679496765,
+      "learning_rate": 8e-05,
+      "loss": 1.6486,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8110800631757988,
+      "grad_norm": 0.4228782057762146,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8113230470173733,
+      "grad_norm": 0.41907837986946106,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 3339
+    },
+    {
+      "epoch": 0.8115660308589479,
+      "grad_norm": 0.43353110551834106,
+      "learning_rate": 8e-05,
+      "loss": 1.5796,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8118090147005225,
+      "grad_norm": 0.4422175884246826,
+      "learning_rate": 8e-05,
+      "loss": 1.6667,
+      "step": 3341
+    },
+    {
+      "epoch": 0.812051998542097,
+      "grad_norm": 0.4439120888710022,
+      "learning_rate": 8e-05,
+      "loss": 1.5937,
+      "step": 3342
+    },
+    {
+      "epoch": 0.8122949823836715,
+      "grad_norm": 0.4479275047779083,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 3343
+    },
+    {
+      "epoch": 0.812537966225246,
+      "grad_norm": 0.45979902148246765,
+      "learning_rate": 8e-05,
+      "loss": 1.8641,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8127809500668206,
+      "grad_norm": 0.43276870250701904,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8130239339083951,
+      "grad_norm": 0.43319928646087646,
+      "learning_rate": 8e-05,
+      "loss": 1.7601,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8132669177499696,
+      "grad_norm": 0.4369921088218689,
+      "learning_rate": 8e-05,
+      "loss": 1.6858,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8135099015915441,
+      "grad_norm": 0.4362873435020447,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8137528854331187,
+      "grad_norm": 0.43402206897735596,
+      "learning_rate": 8e-05,
+      "loss": 1.7516,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8139958692746933,
+      "grad_norm": 0.44945472478866577,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8142388531162678,
+      "grad_norm": 0.45130228996276855,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8144818369578423,
+      "grad_norm": 0.452012300491333,
+      "learning_rate": 8e-05,
+      "loss": 1.7471,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8147248207994169,
+      "grad_norm": 0.42558014392852783,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8149678046409914,
+      "grad_norm": 0.47840672731399536,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8152107884825659,
+      "grad_norm": 0.4429165720939636,
+      "learning_rate": 8e-05,
+      "loss": 1.6445,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8154537723241404,
+      "grad_norm": 0.4439137279987335,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 3356
+    },
+    {
+      "epoch": 0.815696756165715,
+      "grad_norm": 0.43262672424316406,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8159397400072895,
+      "grad_norm": 0.43009153008461,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 3358
+    },
+    {
+      "epoch": 0.816182723848864,
+      "grad_norm": 0.4142119586467743,
+      "learning_rate": 8e-05,
+      "loss": 1.4287,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8164257076904385,
+      "grad_norm": 0.4471481740474701,
+      "learning_rate": 8e-05,
+      "loss": 1.7858,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8166686915320132,
+      "grad_norm": 0.4283752143383026,
+      "learning_rate": 8e-05,
+      "loss": 1.6464,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8169116753735877,
+      "grad_norm": 0.43131816387176514,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8171546592151622,
+      "grad_norm": 0.4510865807533264,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8173976430567367,
+      "grad_norm": 0.4215385615825653,
+      "learning_rate": 8e-05,
+      "loss": 1.6327,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8176406268983113,
+      "grad_norm": 0.4275309145450592,
+      "learning_rate": 8e-05,
+      "loss": 1.7252,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8178836107398858,
+      "grad_norm": 0.43687903881073,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8181265945814603,
+      "grad_norm": 0.41109180450439453,
+      "learning_rate": 8e-05,
+      "loss": 1.6543,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8183695784230348,
+      "grad_norm": 0.4381401836872101,
+      "learning_rate": 8e-05,
+      "loss": 1.6447,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8186125622646094,
+      "grad_norm": 0.4212642014026642,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8188555461061839,
+      "grad_norm": 0.433035284280777,
+      "learning_rate": 8e-05,
+      "loss": 1.5327,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8190985299477584,
+      "grad_norm": 0.43482905626296997,
+      "learning_rate": 8e-05,
+      "loss": 1.7041,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8193415137893331,
+      "grad_norm": 0.44731637835502625,
+      "learning_rate": 8e-05,
+      "loss": 1.8042,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8195844976309076,
+      "grad_norm": 0.4338732063770294,
+      "learning_rate": 8e-05,
+      "loss": 1.6448,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8198274814724821,
+      "grad_norm": 0.45476001501083374,
+      "learning_rate": 8e-05,
+      "loss": 1.795,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8200704653140566,
+      "grad_norm": 0.41525211930274963,
+      "learning_rate": 8e-05,
+      "loss": 1.5472,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8203134491556312,
+      "grad_norm": 0.40967607498168945,
+      "learning_rate": 8e-05,
+      "loss": 1.5924,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8205564329972057,
+      "grad_norm": 0.48753684759140015,
+      "learning_rate": 8e-05,
+      "loss": 1.7171,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8207994168387802,
+      "grad_norm": 0.4343242943286896,
+      "learning_rate": 8e-05,
+      "loss": 1.6293,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8210424006803547,
+      "grad_norm": 0.43552348017692566,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8212853845219293,
+      "grad_norm": 0.4191743731498718,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8215283683635038,
+      "grad_norm": 0.4408535957336426,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8217713522050784,
+      "grad_norm": 0.4226479232311249,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8220143360466529,
+      "grad_norm": 0.4483393430709839,
+      "learning_rate": 8e-05,
+      "loss": 1.5848,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8222573198882275,
+      "grad_norm": 0.4412248134613037,
+      "learning_rate": 8e-05,
+      "loss": 1.7877,
+      "step": 3384
+    },
+    {
+      "epoch": 0.822500303729802,
+      "grad_norm": 0.4477544128894806,
+      "learning_rate": 8e-05,
+      "loss": 1.7404,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8227432875713765,
+      "grad_norm": 0.44891446828842163,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 3386
+    },
+    {
+      "epoch": 0.822986271412951,
+      "grad_norm": 0.41483914852142334,
+      "learning_rate": 8e-05,
+      "loss": 1.583,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8232292552545256,
+      "grad_norm": 0.4404264986515045,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8234722390961001,
+      "grad_norm": 0.40554916858673096,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8237152229376746,
+      "grad_norm": 0.47212427854537964,
+      "learning_rate": 8e-05,
+      "loss": 1.8484,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8239582067792491,
+      "grad_norm": 0.43676480650901794,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8242011906208238,
+      "grad_norm": 0.430423378944397,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 3392
+    },
+    {
+      "epoch": 0.8244441744623983,
+      "grad_norm": 0.4706956744194031,
+      "learning_rate": 8e-05,
+      "loss": 1.7765,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8246871583039728,
+      "grad_norm": 0.44008731842041016,
+      "learning_rate": 8e-05,
+      "loss": 1.7287,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8249301421455473,
+      "grad_norm": 0.4169744849205017,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8251731259871219,
+      "grad_norm": 0.44858214259147644,
+      "learning_rate": 8e-05,
+      "loss": 1.6916,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8254161098286964,
+      "grad_norm": 0.4487575888633728,
+      "learning_rate": 8e-05,
+      "loss": 1.7307,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8256590936702709,
+      "grad_norm": 0.412962943315506,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8259020775118454,
+      "grad_norm": 0.43837735056877136,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 3399
+    },
+    {
+      "epoch": 0.82614506135342,
+      "grad_norm": 0.4435512125492096,
+      "learning_rate": 8e-05,
+      "loss": 1.7219,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8263880451949945,
+      "grad_norm": 0.43162834644317627,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 3401
+    },
+    {
+      "epoch": 0.826631029036569,
+      "grad_norm": 0.42454180121421814,
+      "learning_rate": 8e-05,
+      "loss": 1.6329,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8268740128781435,
+      "grad_norm": 0.4324725270271301,
+      "learning_rate": 8e-05,
+      "loss": 1.5425,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8271169967197182,
+      "grad_norm": 0.45470333099365234,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8273599805612927,
+      "grad_norm": 0.4303080141544342,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8276029644028672,
+      "grad_norm": 0.4435342252254486,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8278459482444418,
+      "grad_norm": 0.42477113008499146,
+      "learning_rate": 8e-05,
+      "loss": 1.5211,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8280889320860163,
+      "grad_norm": 0.4563555121421814,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8283319159275908,
+      "grad_norm": 0.43649882078170776,
+      "learning_rate": 8e-05,
+      "loss": 1.7101,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8285748997691653,
+      "grad_norm": 0.4296419620513916,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8288178836107399,
+      "grad_norm": 0.44226181507110596,
+      "learning_rate": 8e-05,
+      "loss": 1.7859,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8290608674523144,
+      "grad_norm": 0.43503424525260925,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 3412
+    },
+    {
+      "epoch": 0.829303851293889,
+      "grad_norm": 0.44227948784828186,
+      "learning_rate": 8e-05,
+      "loss": 1.7656,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8295468351354635,
+      "grad_norm": 0.42249754071235657,
+      "learning_rate": 8e-05,
+      "loss": 1.5291,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8297898189770381,
+      "grad_norm": 0.43921294808387756,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8300328028186126,
+      "grad_norm": 0.44947415590286255,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8302757866601871,
+      "grad_norm": 0.4385533332824707,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8305187705017616,
+      "grad_norm": 0.42582717537879944,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8307617543433362,
+      "grad_norm": 0.4467014968395233,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8310047381849107,
+      "grad_norm": 0.4120327830314636,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8312477220264852,
+      "grad_norm": 0.4282192885875702,
+      "learning_rate": 8e-05,
+      "loss": 1.6958,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8314907058680597,
+      "grad_norm": 0.43069887161254883,
+      "learning_rate": 8e-05,
+      "loss": 1.5908,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8317336897096343,
+      "grad_norm": 0.42623797059059143,
+      "learning_rate": 8e-05,
+      "loss": 1.7026,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8319766735512089,
+      "grad_norm": 0.43028900027275085,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8322196573927834,
+      "grad_norm": 0.47078806161880493,
+      "learning_rate": 8e-05,
+      "loss": 1.7937,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8324626412343579,
+      "grad_norm": 0.41601136326789856,
+      "learning_rate": 8e-05,
+      "loss": 1.4749,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8327056250759325,
+      "grad_norm": 0.4471643269062042,
+      "learning_rate": 8e-05,
+      "loss": 1.5608,
+      "step": 3427
+    },
+    {
+      "epoch": 0.832948608917507,
+      "grad_norm": 0.431229829788208,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8331915927590815,
+      "grad_norm": 0.4478210508823395,
+      "learning_rate": 8e-05,
+      "loss": 1.7758,
+      "step": 3429
+    },
+    {
+      "epoch": 0.833434576600656,
+      "grad_norm": 0.46986401081085205,
+      "learning_rate": 8e-05,
+      "loss": 1.8681,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8336775604422306,
+      "grad_norm": 0.45001447200775146,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8339205442838051,
+      "grad_norm": 0.4690040647983551,
+      "learning_rate": 8e-05,
+      "loss": 1.7207,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8341635281253796,
+      "grad_norm": 0.4155740439891815,
+      "learning_rate": 8e-05,
+      "loss": 1.5578,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8344065119669541,
+      "grad_norm": 0.4345017969608307,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8346494958085288,
+      "grad_norm": 0.4351257085800171,
+      "learning_rate": 8e-05,
+      "loss": 1.5172,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8348924796501033,
+      "grad_norm": 0.41618382930755615,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8351354634916778,
+      "grad_norm": 0.45230191946029663,
+      "learning_rate": 8e-05,
+      "loss": 1.7703,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8353784473332524,
+      "grad_norm": 0.46118733286857605,
+      "learning_rate": 8e-05,
+      "loss": 1.7368,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8356214311748269,
+      "grad_norm": 0.4292798340320587,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8358644150164014,
+      "grad_norm": 0.4418924152851105,
+      "learning_rate": 8e-05,
+      "loss": 1.791,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8361073988579759,
+      "grad_norm": 0.501727819442749,
+      "learning_rate": 8e-05,
+      "loss": 1.6064,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8363503826995505,
+      "grad_norm": 0.4515193998813629,
+      "learning_rate": 8e-05,
+      "loss": 1.8082,
+      "step": 3442
+    },
+    {
+      "epoch": 0.836593366541125,
+      "grad_norm": 0.447311669588089,
+      "learning_rate": 8e-05,
+      "loss": 1.7821,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8368363503826995,
+      "grad_norm": 0.40453997254371643,
+      "learning_rate": 8e-05,
+      "loss": 1.6365,
+      "step": 3444
+    },
+    {
+      "epoch": 0.837079334224274,
+      "grad_norm": 0.4350511133670807,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8373223180658487,
+      "grad_norm": 0.4293648302555084,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8375653019074232,
+      "grad_norm": 0.45656466484069824,
+      "learning_rate": 8e-05,
+      "loss": 1.761,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8378082857489977,
+      "grad_norm": 0.43648022413253784,
+      "learning_rate": 8e-05,
+      "loss": 1.5547,
+      "step": 3448
+    },
+    {
+      "epoch": 0.8380512695905722,
+      "grad_norm": 0.4508563280105591,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8382942534321468,
+      "grad_norm": 0.4058380722999573,
+      "learning_rate": 8e-05,
+      "loss": 1.5605,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8385372372737213,
+      "grad_norm": 0.4371514916419983,
+      "learning_rate": 8e-05,
+      "loss": 1.684,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8387802211152958,
+      "grad_norm": 0.4577498733997345,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8390232049568703,
+      "grad_norm": 0.4394805133342743,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 3453
+    },
+    {
+      "epoch": 0.839266188798445,
+      "grad_norm": 0.4270798861980438,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8395091726400195,
+      "grad_norm": 0.4307389259338379,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 3455
+    },
+    {
+      "epoch": 0.839752156481594,
+      "grad_norm": 0.46529629826545715,
+      "learning_rate": 8e-05,
+      "loss": 1.8229,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8399951403231685,
+      "grad_norm": 0.4350111484527588,
+      "learning_rate": 8e-05,
+      "loss": 1.6939,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8402381241647431,
+      "grad_norm": 0.43640774488449097,
+      "learning_rate": 8e-05,
+      "loss": 1.8225,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8404811080063176,
+      "grad_norm": 0.43147000670433044,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8407240918478921,
+      "grad_norm": 0.4396747350692749,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8409670756894666,
+      "grad_norm": 0.43728652596473694,
+      "learning_rate": 8e-05,
+      "loss": 1.7202,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8412100595310412,
+      "grad_norm": 0.4332346022129059,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8414530433726157,
+      "grad_norm": 0.44474124908447266,
+      "learning_rate": 8e-05,
+      "loss": 1.749,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8416960272141902,
+      "grad_norm": 0.4193655252456665,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8419390110557647,
+      "grad_norm": 0.43819692730903625,
+      "learning_rate": 8e-05,
+      "loss": 1.8205,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8421819948973394,
+      "grad_norm": 0.4616236388683319,
+      "learning_rate": 8e-05,
+      "loss": 1.6482,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8424249787389139,
+      "grad_norm": 0.4479854106903076,
+      "learning_rate": 8e-05,
+      "loss": 1.7734,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8426679625804884,
+      "grad_norm": 0.4341032803058624,
+      "learning_rate": 8e-05,
+      "loss": 1.7028,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8429109464220629,
+      "grad_norm": 0.4444016218185425,
+      "learning_rate": 8e-05,
+      "loss": 1.7333,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8431539302636375,
+      "grad_norm": 0.4501021206378937,
+      "learning_rate": 8e-05,
+      "loss": 1.7082,
+      "step": 3470
+    },
+    {
+      "epoch": 0.843396914105212,
+      "grad_norm": 0.42559748888015747,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8436398979467865,
+      "grad_norm": 0.4550130367279053,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8438828817883611,
+      "grad_norm": 0.4333694279193878,
+      "learning_rate": 8e-05,
+      "loss": 1.597,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8441258656299356,
+      "grad_norm": 0.44606897234916687,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8443688494715101,
+      "grad_norm": 0.4887733459472656,
+      "learning_rate": 8e-05,
+      "loss": 1.4676,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8446118333130846,
+      "grad_norm": 0.4485006332397461,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8448548171546593,
+      "grad_norm": 0.425587922334671,
+      "learning_rate": 8e-05,
+      "loss": 1.7125,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8450978009962338,
+      "grad_norm": 0.43635955452919006,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8453407848378083,
+      "grad_norm": 0.42515599727630615,
+      "learning_rate": 8e-05,
+      "loss": 1.4841,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8455837686793828,
+      "grad_norm": 0.43514877557754517,
+      "learning_rate": 8e-05,
+      "loss": 1.703,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8458267525209574,
+      "grad_norm": 0.4396819472312927,
+      "learning_rate": 8e-05,
+      "loss": 1.744,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8460697363625319,
+      "grad_norm": 0.44516438245773315,
+      "learning_rate": 8e-05,
+      "loss": 1.5858,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8463127202041064,
+      "grad_norm": 0.4290750324726105,
+      "learning_rate": 8e-05,
+      "loss": 1.5797,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8465557040456809,
+      "grad_norm": 0.4428640305995941,
+      "learning_rate": 8e-05,
+      "loss": 1.7581,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8467986878872555,
+      "grad_norm": 0.4393298327922821,
+      "learning_rate": 8e-05,
+      "loss": 1.7098,
+      "step": 3485
+    },
+    {
+      "epoch": 0.84704167172883,
+      "grad_norm": 0.4792869985103607,
+      "learning_rate": 8e-05,
+      "loss": 1.7111,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8472846555704046,
+      "grad_norm": 0.43274813890457153,
+      "learning_rate": 8e-05,
+      "loss": 1.7038,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8475276394119791,
+      "grad_norm": 0.44146275520324707,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8477706232535537,
+      "grad_norm": 0.42511728405952454,
+      "learning_rate": 8e-05,
+      "loss": 1.6475,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8480136070951282,
+      "grad_norm": 0.43266725540161133,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8482565909367027,
+      "grad_norm": 0.42569735646247864,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8484995747782772,
+      "grad_norm": 0.45071840286254883,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8487425586198518,
+      "grad_norm": 0.4464120864868164,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8489855424614263,
+      "grad_norm": 0.4615325629711151,
+      "learning_rate": 8e-05,
+      "loss": 1.8194,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8492285263030008,
+      "grad_norm": 0.43416813015937805,
+      "learning_rate": 8e-05,
+      "loss": 1.748,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8494715101445753,
+      "grad_norm": 0.44300299882888794,
+      "learning_rate": 8e-05,
+      "loss": 1.7749,
+      "step": 3496
+    },
+    {
+      "epoch": 0.84971449398615,
+      "grad_norm": 0.4522843658924103,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8499574778277245,
+      "grad_norm": 0.4314980208873749,
+      "learning_rate": 8e-05,
+      "loss": 1.7155,
+      "step": 3498
+    },
+    {
+      "epoch": 0.850200461669299,
+      "grad_norm": 0.44239771366119385,
+      "learning_rate": 8e-05,
+      "loss": 1.6547,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8504434455108735,
+      "grad_norm": 0.41549259424209595,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8506864293524481,
+      "grad_norm": 0.4617140293121338,
+      "learning_rate": 8e-05,
+      "loss": 1.8236,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8509294131940226,
+      "grad_norm": 0.4486651122570038,
+      "learning_rate": 8e-05,
+      "loss": 1.8024,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8511723970355971,
+      "grad_norm": 0.4224928319454193,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8514153808771717,
+      "grad_norm": 0.43449920415878296,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8516583647187462,
+      "grad_norm": 0.42443782091140747,
+      "learning_rate": 8e-05,
+      "loss": 1.686,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8519013485603207,
+      "grad_norm": 0.45661354064941406,
+      "learning_rate": 8e-05,
+      "loss": 1.9398,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8521443324018952,
+      "grad_norm": 0.43229126930236816,
+      "learning_rate": 8e-05,
+      "loss": 1.7264,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8523873162434699,
+      "grad_norm": 0.43250468373298645,
+      "learning_rate": 8e-05,
+      "loss": 1.7005,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8526303000850444,
+      "grad_norm": 0.4313049614429474,
+      "learning_rate": 8e-05,
+      "loss": 1.7562,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8528732839266189,
+      "grad_norm": 0.41590237617492676,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8531162677681934,
+      "grad_norm": 0.4380669891834259,
+      "learning_rate": 8e-05,
+      "loss": 1.6323,
+      "step": 3511
+    },
+    {
+      "epoch": 0.853359251609768,
+      "grad_norm": 0.4281664788722992,
+      "learning_rate": 8e-05,
+      "loss": 1.5646,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8536022354513425,
+      "grad_norm": 0.4371143877506256,
+      "learning_rate": 8e-05,
+      "loss": 1.7443,
+      "step": 3513
+    },
+    {
+      "epoch": 0.853845219292917,
+      "grad_norm": 0.46331271529197693,
+      "learning_rate": 8e-05,
+      "loss": 1.882,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8540882031344915,
+      "grad_norm": 0.493507444858551,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8543311869760661,
+      "grad_norm": 0.43589046597480774,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8545741708176406,
+      "grad_norm": 0.43377333879470825,
+      "learning_rate": 8e-05,
+      "loss": 1.6779,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8548171546592152,
+      "grad_norm": 0.4429006278514862,
+      "learning_rate": 8e-05,
+      "loss": 1.8276,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8550601385007897,
+      "grad_norm": 0.4460608959197998,
+      "learning_rate": 8e-05,
+      "loss": 1.5152,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8553031223423643,
+      "grad_norm": 0.41911131143569946,
+      "learning_rate": 8e-05,
+      "loss": 1.6234,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8555461061839388,
+      "grad_norm": 0.4320312440395355,
+      "learning_rate": 8e-05,
+      "loss": 1.6858,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8557890900255133,
+      "grad_norm": 0.43006858229637146,
+      "learning_rate": 8e-05,
+      "loss": 1.6759,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8560320738670878,
+      "grad_norm": 0.4599211513996124,
+      "learning_rate": 8e-05,
+      "loss": 1.625,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8562750577086624,
+      "grad_norm": 0.4371275007724762,
+      "learning_rate": 8e-05,
+      "loss": 1.6268,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8565180415502369,
+      "grad_norm": 0.4260992109775543,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8567610253918114,
+      "grad_norm": 0.4542173743247986,
+      "learning_rate": 8e-05,
+      "loss": 1.5979,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8570040092333859,
+      "grad_norm": 0.428114116191864,
+      "learning_rate": 8e-05,
+      "loss": 1.5764,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8572469930749606,
+      "grad_norm": 0.4457513988018036,
+      "learning_rate": 8e-05,
+      "loss": 1.7459,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8574899769165351,
+      "grad_norm": 0.4513632357120514,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8577329607581096,
+      "grad_norm": 0.44085678458213806,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8579759445996841,
+      "grad_norm": 0.4464363157749176,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8582189284412587,
+      "grad_norm": 0.43660178780555725,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8584619122828332,
+      "grad_norm": 0.4252590239048004,
+      "learning_rate": 8e-05,
+      "loss": 1.6293,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8587048961244077,
+      "grad_norm": 0.4417291581630707,
+      "learning_rate": 8e-05,
+      "loss": 1.8289,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8589478799659822,
+      "grad_norm": 0.4230963885784149,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8591908638075568,
+      "grad_norm": 0.45418643951416016,
+      "learning_rate": 8e-05,
+      "loss": 1.782,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8594338476491313,
+      "grad_norm": 0.4169585108757019,
+      "learning_rate": 8e-05,
+      "loss": 1.5834,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8596768314907058,
+      "grad_norm": 0.4516275227069855,
+      "learning_rate": 8e-05,
+      "loss": 1.6867,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8599198153322805,
+      "grad_norm": 0.4633653461933136,
+      "learning_rate": 8e-05,
+      "loss": 1.837,
+      "step": 3539
+    },
+    {
+      "epoch": 0.860162799173855,
+      "grad_norm": 0.4480876922607422,
+      "learning_rate": 8e-05,
+      "loss": 1.5903,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8604057830154295,
+      "grad_norm": 0.4733363091945648,
+      "learning_rate": 8e-05,
+      "loss": 1.7856,
+      "step": 3541
+    },
+    {
+      "epoch": 0.860648766857004,
+      "grad_norm": 0.43148073554039,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8608917506985786,
+      "grad_norm": 0.4488212764263153,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8611347345401531,
+      "grad_norm": 0.4222504794597626,
+      "learning_rate": 8e-05,
+      "loss": 1.5933,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8613777183817276,
+      "grad_norm": 0.4252151846885681,
+      "learning_rate": 8e-05,
+      "loss": 1.6984,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8616207022233021,
+      "grad_norm": 0.4198188781738281,
+      "learning_rate": 8e-05,
+      "loss": 1.6801,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8618636860648767,
+      "grad_norm": 0.4226166903972626,
+      "learning_rate": 8e-05,
+      "loss": 1.6448,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8621066699064512,
+      "grad_norm": 0.47494763135910034,
+      "learning_rate": 8e-05,
+      "loss": 1.691,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8623496537480257,
+      "grad_norm": 0.4533672034740448,
+      "learning_rate": 8e-05,
+      "loss": 1.5716,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8625926375896003,
+      "grad_norm": 0.4749249517917633,
+      "learning_rate": 8e-05,
+      "loss": 1.8267,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8628356214311749,
+      "grad_norm": 0.4386540651321411,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8630786052727494,
+      "grad_norm": 0.44479119777679443,
+      "learning_rate": 8e-05,
+      "loss": 1.7502,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8633215891143239,
+      "grad_norm": 0.44304460287094116,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8635645729558984,
+      "grad_norm": 0.4387069046497345,
+      "learning_rate": 8e-05,
+      "loss": 1.6295,
+      "step": 3554
+    },
+    {
+      "epoch": 0.863807556797473,
+      "grad_norm": 0.4467419981956482,
+      "learning_rate": 8e-05,
+      "loss": 1.8038,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8640505406390475,
+      "grad_norm": 0.4909220039844513,
+      "learning_rate": 8e-05,
+      "loss": 1.7801,
+      "step": 3556
+    },
+    {
+      "epoch": 0.864293524480622,
+      "grad_norm": 0.45934250950813293,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8645365083221965,
+      "grad_norm": 0.4486108124256134,
+      "learning_rate": 8e-05,
+      "loss": 1.6907,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8647794921637711,
+      "grad_norm": 0.45225340127944946,
+      "learning_rate": 8e-05,
+      "loss": 1.7553,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8650224760053457,
+      "grad_norm": 0.45665428042411804,
+      "learning_rate": 8e-05,
+      "loss": 1.7456,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8652654598469202,
+      "grad_norm": 0.4246043562889099,
+      "learning_rate": 8e-05,
+      "loss": 1.7728,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8655084436884947,
+      "grad_norm": 0.42815276980400085,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8657514275300693,
+      "grad_norm": 0.4386288821697235,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8659944113716438,
+      "grad_norm": 0.4750242531299591,
+      "learning_rate": 8e-05,
+      "loss": 1.965,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8662373952132183,
+      "grad_norm": 0.42668020725250244,
+      "learning_rate": 8e-05,
+      "loss": 1.6067,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8664803790547928,
+      "grad_norm": 0.445643812417984,
+      "learning_rate": 8e-05,
+      "loss": 1.7384,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8667233628963674,
+      "grad_norm": 0.4247991442680359,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8669663467379419,
+      "grad_norm": 0.4792838990688324,
+      "learning_rate": 8e-05,
+      "loss": 1.8569,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8672093305795164,
+      "grad_norm": 0.48669686913490295,
+      "learning_rate": 8e-05,
+      "loss": 1.7687,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8674523144210909,
+      "grad_norm": 0.45670074224472046,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8676952982626656,
+      "grad_norm": 0.43432384729385376,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8679382821042401,
+      "grad_norm": 0.43173617124557495,
+      "learning_rate": 8e-05,
+      "loss": 1.785,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8681812659458146,
+      "grad_norm": 0.4332762062549591,
+      "learning_rate": 8e-05,
+      "loss": 1.6047,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8684242497873892,
+      "grad_norm": 0.4563746154308319,
+      "learning_rate": 8e-05,
+      "loss": 1.8216,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8686672336289637,
+      "grad_norm": 0.43840107321739197,
+      "learning_rate": 8e-05,
+      "loss": 1.7353,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8689102174705382,
+      "grad_norm": 0.42819610238075256,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8691532013121127,
+      "grad_norm": 0.5060098767280579,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8693961851536873,
+      "grad_norm": 0.42244431376457214,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8696391689952618,
+      "grad_norm": 0.4488561153411865,
+      "learning_rate": 8e-05,
+      "loss": 1.686,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8698821528368363,
+      "grad_norm": 0.4538997411727905,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8701251366784108,
+      "grad_norm": 0.45665764808654785,
+      "learning_rate": 8e-05,
+      "loss": 1.7813,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8703681205199855,
+      "grad_norm": 0.43001365661621094,
+      "learning_rate": 8e-05,
+      "loss": 1.6761,
+      "step": 3582
+    },
+    {
+      "epoch": 0.87061110436156,
+      "grad_norm": 0.4255593717098236,
+      "learning_rate": 8e-05,
+      "loss": 1.5781,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8708540882031345,
+      "grad_norm": 0.4350177049636841,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 3584
+    },
+    {
+      "epoch": 0.871097072044709,
+      "grad_norm": 0.456264853477478,
+      "learning_rate": 8e-05,
+      "loss": 1.6987,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8713400558862836,
+      "grad_norm": 0.4561658799648285,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8715830397278581,
+      "grad_norm": 0.4243806004524231,
+      "learning_rate": 8e-05,
+      "loss": 1.5037,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8718260235694326,
+      "grad_norm": 0.4443802237510681,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8720690074110071,
+      "grad_norm": 0.44609907269477844,
+      "learning_rate": 8e-05,
+      "loss": 1.7685,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8723119912525817,
+      "grad_norm": 0.41593265533447266,
+      "learning_rate": 8e-05,
+      "loss": 1.5893,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8725549750941562,
+      "grad_norm": 0.4418441951274872,
+      "learning_rate": 8e-05,
+      "loss": 1.665,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8727979589357308,
+      "grad_norm": 0.44337302446365356,
+      "learning_rate": 8e-05,
+      "loss": 1.7126,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8730409427773053,
+      "grad_norm": 0.44593533873558044,
+      "learning_rate": 8e-05,
+      "loss": 1.7709,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8732839266188799,
+      "grad_norm": 0.4367687702178955,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8735269104604544,
+      "grad_norm": 0.4335482716560364,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8737698943020289,
+      "grad_norm": 0.4354555308818817,
+      "learning_rate": 8e-05,
+      "loss": 1.6476,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8740128781436034,
+      "grad_norm": 0.4317670464515686,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 3597
+    },
+    {
+      "epoch": 0.874255861985178,
+      "grad_norm": 0.4350283741950989,
+      "learning_rate": 8e-05,
+      "loss": 1.7135,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8744988458267525,
+      "grad_norm": 0.4702574610710144,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 3599
+    },
+    {
+      "epoch": 0.874741829668327,
+      "grad_norm": 0.45515093207359314,
+      "learning_rate": 8e-05,
+      "loss": 1.7553,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8749848135099015,
+      "grad_norm": 0.41463178396224976,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8752277973514762,
+      "grad_norm": 0.4467906057834625,
+      "learning_rate": 8e-05,
+      "loss": 1.5914,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8754707811930507,
+      "grad_norm": 0.44887083768844604,
+      "learning_rate": 8e-05,
+      "loss": 1.7091,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8757137650346252,
+      "grad_norm": 0.43566155433654785,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8759567488761998,
+      "grad_norm": 0.42790085077285767,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8761997327177743,
+      "grad_norm": 0.4351292848587036,
+      "learning_rate": 8e-05,
+      "loss": 1.8028,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8764427165593488,
+      "grad_norm": 0.4314468502998352,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8766857004009233,
+      "grad_norm": 0.43339359760284424,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8769286842424979,
+      "grad_norm": 0.4185611307621002,
+      "learning_rate": 8e-05,
+      "loss": 1.5387,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8771716680840724,
+      "grad_norm": 0.41385525465011597,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8774146519256469,
+      "grad_norm": 0.479339063167572,
+      "learning_rate": 8e-05,
+      "loss": 1.6947,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8776576357672214,
+      "grad_norm": 0.41456007957458496,
+      "learning_rate": 8e-05,
+      "loss": 1.6084,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8779006196087961,
+      "grad_norm": 0.41950544714927673,
+      "learning_rate": 8e-05,
+      "loss": 1.6337,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8781436034503706,
+      "grad_norm": 0.4560462534427643,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8783865872919451,
+      "grad_norm": 0.393917977809906,
+      "learning_rate": 8e-05,
+      "loss": 1.2595,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8786295711335196,
+      "grad_norm": 0.5065532922744751,
+      "learning_rate": 8e-05,
+      "loss": 1.7088,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8788725549750942,
+      "grad_norm": 0.4163370728492737,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8791155388166687,
+      "grad_norm": 0.4419529139995575,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8793585226582432,
+      "grad_norm": 0.4359494149684906,
+      "learning_rate": 8e-05,
+      "loss": 1.7483,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8796015064998177,
+      "grad_norm": 0.4330137073993683,
+      "learning_rate": 8e-05,
+      "loss": 1.6973,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8798444903413923,
+      "grad_norm": 0.4298569858074188,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8800874741829668,
+      "grad_norm": 0.4495445787906647,
+      "learning_rate": 8e-05,
+      "loss": 1.6498,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8803304580245414,
+      "grad_norm": 0.43694964051246643,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8805734418661159,
+      "grad_norm": 0.4324922263622284,
+      "learning_rate": 8e-05,
+      "loss": 1.7576,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8808164257076905,
+      "grad_norm": 0.46163323521614075,
+      "learning_rate": 8e-05,
+      "loss": 1.8227,
+      "step": 3625
+    },
+    {
+      "epoch": 0.881059409549265,
+      "grad_norm": 0.45882123708724976,
+      "learning_rate": 8e-05,
+      "loss": 1.6859,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8813023933908395,
+      "grad_norm": 0.4467215836048126,
+      "learning_rate": 8e-05,
+      "loss": 1.746,
+      "step": 3627
+    },
+    {
+      "epoch": 0.881545377232414,
+      "grad_norm": 0.4481157064437866,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8817883610739886,
+      "grad_norm": 0.41644763946533203,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8820313449155631,
+      "grad_norm": 0.42902112007141113,
+      "learning_rate": 8e-05,
+      "loss": 1.6572,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8822743287571376,
+      "grad_norm": 0.43686214089393616,
+      "learning_rate": 8e-05,
+      "loss": 1.8661,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8825173125987121,
+      "grad_norm": 0.41906967759132385,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8827602964402868,
+      "grad_norm": 0.4393027424812317,
+      "learning_rate": 8e-05,
+      "loss": 1.7538,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8830032802818613,
+      "grad_norm": 0.4225122928619385,
+      "learning_rate": 8e-05,
+      "loss": 1.6193,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8832462641234358,
+      "grad_norm": 0.42653805017471313,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8834892479650103,
+      "grad_norm": 0.43028491735458374,
+      "learning_rate": 8e-05,
+      "loss": 1.6115,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8837322318065849,
+      "grad_norm": 0.4372873902320862,
+      "learning_rate": 8e-05,
+      "loss": 1.6515,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8839752156481594,
+      "grad_norm": 0.4355347454547882,
+      "learning_rate": 8e-05,
+      "loss": 1.7952,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8842181994897339,
+      "grad_norm": 0.4396231472492218,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8844611833313085,
+      "grad_norm": 0.46660298109054565,
+      "learning_rate": 8e-05,
+      "loss": 1.6354,
+      "step": 3640
+    },
+    {
+      "epoch": 0.884704167172883,
+      "grad_norm": 0.43321511149406433,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8849471510144575,
+      "grad_norm": 0.4424004554748535,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 3642
+    },
+    {
+      "epoch": 0.885190134856032,
+      "grad_norm": 0.43963947892189026,
+      "learning_rate": 8e-05,
+      "loss": 1.7556,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8854331186976067,
+      "grad_norm": 0.45685961842536926,
+      "learning_rate": 8e-05,
+      "loss": 1.6713,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8856761025391812,
+      "grad_norm": 0.42240655422210693,
+      "learning_rate": 8e-05,
+      "loss": 1.482,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8859190863807557,
+      "grad_norm": 0.43038034439086914,
+      "learning_rate": 8e-05,
+      "loss": 1.6276,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8861620702223302,
+      "grad_norm": 0.45398852229118347,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8864050540639048,
+      "grad_norm": 0.43309852480888367,
+      "learning_rate": 8e-05,
+      "loss": 1.6535,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8866480379054793,
+      "grad_norm": 0.4446822702884674,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8868910217470538,
+      "grad_norm": 0.450141966342926,
+      "learning_rate": 8e-05,
+      "loss": 1.8129,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8871340055886283,
+      "grad_norm": 0.4411889910697937,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8873769894302029,
+      "grad_norm": 0.41618266701698303,
+      "learning_rate": 8e-05,
+      "loss": 1.5527,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8876199732717774,
+      "grad_norm": 0.41980019211769104,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 3653
+    },
+    {
+      "epoch": 0.887862957113352,
+      "grad_norm": 0.4287503957748413,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8881059409549265,
+      "grad_norm": 0.44084519147872925,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8883489247965011,
+      "grad_norm": 0.4462060034275055,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8885919086380756,
+      "grad_norm": 0.45298057794570923,
+      "learning_rate": 8e-05,
+      "loss": 1.6811,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8888348924796501,
+      "grad_norm": 0.46214690804481506,
+      "learning_rate": 8e-05,
+      "loss": 1.6546,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8890778763212246,
+      "grad_norm": 0.4398306608200073,
+      "learning_rate": 8e-05,
+      "loss": 1.7311,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8893208601627992,
+      "grad_norm": 0.4483547508716583,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8895638440043737,
+      "grad_norm": 0.43408456444740295,
+      "learning_rate": 8e-05,
+      "loss": 1.7453,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8898068278459482,
+      "grad_norm": 0.5019899010658264,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8900498116875227,
+      "grad_norm": 0.4173068106174469,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8902927955290973,
+      "grad_norm": 0.43129992485046387,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8905357793706719,
+      "grad_norm": 0.43201178312301636,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8907787632122464,
+      "grad_norm": 0.46675631403923035,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8910217470538209,
+      "grad_norm": 0.4165835678577423,
+      "learning_rate": 8e-05,
+      "loss": 1.4372,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8912647308953955,
+      "grad_norm": 0.46898940205574036,
+      "learning_rate": 8e-05,
+      "loss": 1.7336,
+      "step": 3668
+    },
+    {
+      "epoch": 0.89150771473697,
+      "grad_norm": 0.44775640964508057,
+      "learning_rate": 8e-05,
+      "loss": 1.5412,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8917506985785445,
+      "grad_norm": 0.425492525100708,
+      "learning_rate": 8e-05,
+      "loss": 1.6023,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8919936824201191,
+      "grad_norm": 0.4394069314002991,
+      "learning_rate": 8e-05,
+      "loss": 1.7818,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8922366662616936,
+      "grad_norm": 0.4408743977546692,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8924796501032681,
+      "grad_norm": 0.4505542814731598,
+      "learning_rate": 8e-05,
+      "loss": 1.7657,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8927226339448426,
+      "grad_norm": 0.42818546295166016,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8929656177864173,
+      "grad_norm": 0.4732562303543091,
+      "learning_rate": 8e-05,
+      "loss": 1.6668,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8932086016279918,
+      "grad_norm": 0.4384468197822571,
+      "learning_rate": 8e-05,
+      "loss": 1.6128,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8934515854695663,
+      "grad_norm": 0.4543495178222656,
+      "learning_rate": 8e-05,
+      "loss": 1.7599,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8936945693111408,
+      "grad_norm": 0.44212764501571655,
+      "learning_rate": 8e-05,
+      "loss": 1.7095,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8939375531527154,
+      "grad_norm": 0.4205838441848755,
+      "learning_rate": 8e-05,
+      "loss": 1.5524,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8941805369942899,
+      "grad_norm": 0.43754810094833374,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8944235208358644,
+      "grad_norm": 0.4554198980331421,
+      "learning_rate": 8e-05,
+      "loss": 1.7246,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8946665046774389,
+      "grad_norm": 0.4411252439022064,
+      "learning_rate": 8e-05,
+      "loss": 1.6917,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8949094885190135,
+      "grad_norm": 0.46217888593673706,
+      "learning_rate": 8e-05,
+      "loss": 1.7745,
+      "step": 3683
+    },
+    {
+      "epoch": 0.895152472360588,
+      "grad_norm": 0.4497666358947754,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8953954562021625,
+      "grad_norm": 0.44245076179504395,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 3685
+    },
+    {
+      "epoch": 0.895638440043737,
+      "grad_norm": 0.44537630677223206,
+      "learning_rate": 8e-05,
+      "loss": 1.6339,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8958814238853117,
+      "grad_norm": 0.43668246269226074,
+      "learning_rate": 8e-05,
+      "loss": 1.7544,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8961244077268862,
+      "grad_norm": 0.45592814683914185,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8963673915684607,
+      "grad_norm": 0.43188872933387756,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8966103754100352,
+      "grad_norm": 0.4650302231311798,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8968533592516098,
+      "grad_norm": 0.4456447958946228,
+      "learning_rate": 8e-05,
+      "loss": 1.7843,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8970963430931843,
+      "grad_norm": 0.4261976182460785,
+      "learning_rate": 8e-05,
+      "loss": 1.6349,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8973393269347588,
+      "grad_norm": 0.46011608839035034,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8975823107763333,
+      "grad_norm": 0.41910019516944885,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8978252946179079,
+      "grad_norm": 0.4435359537601471,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8980682784594824,
+      "grad_norm": 0.41694116592407227,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 3696
+    },
+    {
+      "epoch": 0.898311262301057,
+      "grad_norm": 0.46052736043930054,
+      "learning_rate": 8e-05,
+      "loss": 1.7167,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8985542461426315,
+      "grad_norm": 0.44669756293296814,
+      "learning_rate": 8e-05,
+      "loss": 1.6857,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8987972299842061,
+      "grad_norm": 0.44350555539131165,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8990402138257806,
+      "grad_norm": 0.44076672196388245,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8992831976673551,
+      "grad_norm": 0.4463120698928833,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8995261815089296,
+      "grad_norm": 0.4212900996208191,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8997691653505042,
+      "grad_norm": 0.41470015048980713,
+      "learning_rate": 8e-05,
+      "loss": 1.4742,
+      "step": 3703
+    },
+    {
+      "epoch": 0.9000121491920787,
+      "grad_norm": 0.4319457709789276,
+      "learning_rate": 8e-05,
+      "loss": 1.6531,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9002551330336532,
+      "grad_norm": 0.4450264275074005,
+      "learning_rate": 8e-05,
+      "loss": 1.7263,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9004981168752278,
+      "grad_norm": 0.43627074360847473,
+      "learning_rate": 8e-05,
+      "loss": 1.5114,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9007411007168024,
+      "grad_norm": 0.4284435510635376,
+      "learning_rate": 8e-05,
+      "loss": 1.7223,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9009840845583769,
+      "grad_norm": 0.43454113602638245,
+      "learning_rate": 8e-05,
+      "loss": 1.7091,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9012270683999514,
+      "grad_norm": 0.4481728672981262,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 3709
+    },
+    {
+      "epoch": 0.901470052241526,
+      "grad_norm": 0.437235027551651,
+      "learning_rate": 8e-05,
+      "loss": 1.7211,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9017130360831005,
+      "grad_norm": 0.4523758888244629,
+      "learning_rate": 8e-05,
+      "loss": 1.8167,
+      "step": 3711
+    },
+    {
+      "epoch": 0.901956019924675,
+      "grad_norm": 0.4398941993713379,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9021990037662495,
+      "grad_norm": 0.43979698419570923,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9024419876078241,
+      "grad_norm": 0.4364776909351349,
+      "learning_rate": 8e-05,
+      "loss": 1.7197,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9026849714493986,
+      "grad_norm": 0.4454684853553772,
+      "learning_rate": 8e-05,
+      "loss": 1.7948,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9029279552909731,
+      "grad_norm": 0.420951247215271,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9031709391325476,
+      "grad_norm": 0.4175846576690674,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9034139229741223,
+      "grad_norm": 0.45951104164123535,
+      "learning_rate": 8e-05,
+      "loss": 1.6589,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9036569068156968,
+      "grad_norm": 0.42241957783699036,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9038998906572713,
+      "grad_norm": 0.4215924143791199,
+      "learning_rate": 8e-05,
+      "loss": 1.6034,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9041428744988458,
+      "grad_norm": 0.4260345995426178,
+      "learning_rate": 8e-05,
+      "loss": 1.5909,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9043858583404204,
+      "grad_norm": 0.41836705803871155,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9046288421819949,
+      "grad_norm": 0.441590815782547,
+      "learning_rate": 8e-05,
+      "loss": 1.7337,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9048718260235694,
+      "grad_norm": 0.4375176727771759,
+      "learning_rate": 8e-05,
+      "loss": 1.8204,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9051148098651439,
+      "grad_norm": 0.4322124421596527,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9053577937067185,
+      "grad_norm": 0.44348740577697754,
+      "learning_rate": 8e-05,
+      "loss": 1.8074,
+      "step": 3726
+    },
+    {
+      "epoch": 0.905600777548293,
+      "grad_norm": 0.47257333993911743,
+      "learning_rate": 8e-05,
+      "loss": 1.8493,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9058437613898676,
+      "grad_norm": 0.4179985523223877,
+      "learning_rate": 8e-05,
+      "loss": 1.6793,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9060867452314421,
+      "grad_norm": 0.4662730395793915,
+      "learning_rate": 8e-05,
+      "loss": 1.762,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9063297290730167,
+      "grad_norm": 0.436396062374115,
+      "learning_rate": 8e-05,
+      "loss": 1.7257,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9065727129145912,
+      "grad_norm": 0.43950334191322327,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9068156967561657,
+      "grad_norm": 0.4187745153903961,
+      "learning_rate": 8e-05,
+      "loss": 1.6314,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9070586805977402,
+      "grad_norm": 0.4142078459262848,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9073016644393148,
+      "grad_norm": 0.4899173676967621,
+      "learning_rate": 8e-05,
+      "loss": 1.7804,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9075446482808893,
+      "grad_norm": 0.43366000056266785,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9077876321224638,
+      "grad_norm": 0.48566362261772156,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9080306159640384,
+      "grad_norm": 0.4804273843765259,
+      "learning_rate": 8e-05,
+      "loss": 1.7643,
+      "step": 3737
+    },
+    {
+      "epoch": 0.908273599805613,
+      "grad_norm": 0.47491735219955444,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9085165836471875,
+      "grad_norm": 0.4293172359466553,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 3739
+    },
+    {
+      "epoch": 0.908759567488762,
+      "grad_norm": 0.4463214874267578,
+      "learning_rate": 8e-05,
+      "loss": 1.7855,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9090025513303366,
+      "grad_norm": 0.45213109254837036,
+      "learning_rate": 8e-05,
+      "loss": 1.6432,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9092455351719111,
+      "grad_norm": 0.4258787930011749,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9094885190134856,
+      "grad_norm": 0.4446050524711609,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9097315028550601,
+      "grad_norm": 0.4585287868976593,
+      "learning_rate": 8e-05,
+      "loss": 1.6366,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9099744866966347,
+      "grad_norm": 0.4389406442642212,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9102174705382092,
+      "grad_norm": 0.42453545331954956,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9104604543797837,
+      "grad_norm": 0.4588576555252075,
+      "learning_rate": 8e-05,
+      "loss": 1.6588,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9107034382213582,
+      "grad_norm": 0.43968674540519714,
+      "learning_rate": 8e-05,
+      "loss": 1.8274,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9109464220629329,
+      "grad_norm": 0.441427081823349,
+      "learning_rate": 8e-05,
+      "loss": 1.6406,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9111894059045074,
+      "grad_norm": 0.45133528113365173,
+      "learning_rate": 8e-05,
+      "loss": 1.804,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9114323897460819,
+      "grad_norm": 0.46313828229904175,
+      "learning_rate": 8e-05,
+      "loss": 1.6836,
+      "step": 3751
+    },
+    {
+      "epoch": 0.9116753735876564,
+      "grad_norm": 0.4399878978729248,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 3752
+    },
+    {
+      "epoch": 0.911918357429231,
+      "grad_norm": 0.42163991928100586,
+      "learning_rate": 8e-05,
+      "loss": 1.6599,
+      "step": 3753
+    },
+    {
+      "epoch": 0.9121613412708055,
+      "grad_norm": 0.43801045417785645,
+      "learning_rate": 8e-05,
+      "loss": 1.8133,
+      "step": 3754
+    },
+    {
+      "epoch": 0.91240432511238,
+      "grad_norm": 0.4394350051879883,
+      "learning_rate": 8e-05,
+      "loss": 1.7488,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9126473089539545,
+      "grad_norm": 0.43560341000556946,
+      "learning_rate": 8e-05,
+      "loss": 1.8907,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9128902927955291,
+      "grad_norm": 0.4244707524776459,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9131332766371036,
+      "grad_norm": 0.4295383393764496,
+      "learning_rate": 8e-05,
+      "loss": 1.6266,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9133762604786781,
+      "grad_norm": 0.44377464056015015,
+      "learning_rate": 8e-05,
+      "loss": 1.7126,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9136192443202527,
+      "grad_norm": 0.4204627573490143,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9138622281618273,
+      "grad_norm": 0.4427950382232666,
+      "learning_rate": 8e-05,
+      "loss": 1.7198,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9141052120034018,
+      "grad_norm": 0.4773518443107605,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9143481958449763,
+      "grad_norm": 0.42238178849220276,
+      "learning_rate": 8e-05,
+      "loss": 1.5658,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9145911796865508,
+      "grad_norm": 0.46549883484840393,
+      "learning_rate": 8e-05,
+      "loss": 1.6457,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9148341635281254,
+      "grad_norm": 0.46497073769569397,
+      "learning_rate": 8e-05,
+      "loss": 1.7181,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9150771473696999,
+      "grad_norm": 0.4723299741744995,
+      "learning_rate": 8e-05,
+      "loss": 1.7338,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9153201312112744,
+      "grad_norm": 0.4395950734615326,
+      "learning_rate": 8e-05,
+      "loss": 1.7546,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9155631150528489,
+      "grad_norm": 0.4463095963001251,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9158060988944235,
+      "grad_norm": 0.44808605313301086,
+      "learning_rate": 8e-05,
+      "loss": 1.7304,
+      "step": 3769
+    },
+    {
+      "epoch": 0.916049082735998,
+      "grad_norm": 0.4596715271472931,
+      "learning_rate": 8e-05,
+      "loss": 1.6726,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9162920665775726,
+      "grad_norm": 0.4164433777332306,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9165350504191472,
+      "grad_norm": 0.41679054498672485,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 3772
+    },
+    {
+      "epoch": 0.9167780342607217,
+      "grad_norm": 0.4454224705696106,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9170210181022962,
+      "grad_norm": 0.4797179400920868,
+      "learning_rate": 8e-05,
+      "loss": 1.5807,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9172640019438707,
+      "grad_norm": 0.4299633204936981,
+      "learning_rate": 8e-05,
+      "loss": 1.7411,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9175069857854453,
+      "grad_norm": 0.4576341211795807,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9177499696270198,
+      "grad_norm": 0.4510483145713806,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9179929534685943,
+      "grad_norm": 0.4450870454311371,
+      "learning_rate": 8e-05,
+      "loss": 1.5899,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9182359373101688,
+      "grad_norm": 0.44307419657707214,
+      "learning_rate": 8e-05,
+      "loss": 1.6936,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9184789211517435,
+      "grad_norm": 0.448016881942749,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 3780
+    },
+    {
+      "epoch": 0.918721904993318,
+      "grad_norm": 0.43140777945518494,
+      "learning_rate": 8e-05,
+      "loss": 1.6801,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9189648888348925,
+      "grad_norm": 0.46103036403656006,
+      "learning_rate": 8e-05,
+      "loss": 1.8138,
+      "step": 3782
+    },
+    {
+      "epoch": 0.919207872676467,
+      "grad_norm": 0.4282258152961731,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9194508565180416,
+      "grad_norm": 0.4454348683357239,
+      "learning_rate": 8e-05,
+      "loss": 1.6373,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9196938403596161,
+      "grad_norm": 0.44090795516967773,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9199368242011906,
+      "grad_norm": 0.4316701591014862,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9201798080427651,
+      "grad_norm": 0.45041531324386597,
+      "learning_rate": 8e-05,
+      "loss": 1.6981,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9204227918843397,
+      "grad_norm": 0.4414864778518677,
+      "learning_rate": 8e-05,
+      "loss": 1.6654,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9206657757259142,
+      "grad_norm": 0.4503396451473236,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9209087595674887,
+      "grad_norm": 0.4543728232383728,
+      "learning_rate": 8e-05,
+      "loss": 1.7385,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9211517434090633,
+      "grad_norm": 0.4209250509738922,
+      "learning_rate": 8e-05,
+      "loss": 1.6865,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9213947272506379,
+      "grad_norm": 0.4681580364704132,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9216377110922124,
+      "grad_norm": 0.45268356800079346,
+      "learning_rate": 8e-05,
+      "loss": 1.7109,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9218806949337869,
+      "grad_norm": 0.47112980484962463,
+      "learning_rate": 8e-05,
+      "loss": 1.8866,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9221236787753614,
+      "grad_norm": 0.4518403112888336,
+      "learning_rate": 8e-05,
+      "loss": 1.7379,
+      "step": 3795
+    },
+    {
+      "epoch": 0.922366662616936,
+      "grad_norm": 0.4558570086956024,
+      "learning_rate": 8e-05,
+      "loss": 1.6351,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9226096464585105,
+      "grad_norm": 0.43270429968833923,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 3797
+    },
+    {
+      "epoch": 0.922852630300085,
+      "grad_norm": 0.41840115189552307,
+      "learning_rate": 8e-05,
+      "loss": 1.7104,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9230956141416595,
+      "grad_norm": 0.4156055450439453,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9233385979832341,
+      "grad_norm": 0.46905747056007385,
+      "learning_rate": 8e-05,
+      "loss": 1.8143,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9235815818248087,
+      "grad_norm": 0.42408761382102966,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9238245656663832,
+      "grad_norm": 0.4440987706184387,
+      "learning_rate": 8e-05,
+      "loss": 1.7802,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9240675495079578,
+      "grad_norm": 0.45483335852622986,
+      "learning_rate": 8e-05,
+      "loss": 1.836,
+      "step": 3803
+    },
+    {
+      "epoch": 0.9243105333495323,
+      "grad_norm": 0.43841424584388733,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9245535171911068,
+      "grad_norm": 0.4519685208797455,
+      "learning_rate": 8e-05,
+      "loss": 1.8092,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9247965010326813,
+      "grad_norm": 0.4293490946292877,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9250394848742559,
+      "grad_norm": 0.4349623918533325,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 3807
+    },
+    {
+      "epoch": 0.9252824687158304,
+      "grad_norm": 0.4470785856246948,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 3808
+    },
+    {
+      "epoch": 0.9255254525574049,
+      "grad_norm": 0.4260224997997284,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9257684363989794,
+      "grad_norm": 0.4532127380371094,
+      "learning_rate": 8e-05,
+      "loss": 1.7822,
+      "step": 3810
+    },
+    {
+      "epoch": 0.926011420240554,
+      "grad_norm": 0.44793978333473206,
+      "learning_rate": 8e-05,
+      "loss": 1.7247,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9262544040821286,
+      "grad_norm": 0.42789098620414734,
+      "learning_rate": 8e-05,
+      "loss": 1.7186,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9264973879237031,
+      "grad_norm": 0.43468183279037476,
+      "learning_rate": 8e-05,
+      "loss": 1.7638,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9267403717652776,
+      "grad_norm": 0.4288477301597595,
+      "learning_rate": 8e-05,
+      "loss": 1.5528,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9269833556068522,
+      "grad_norm": 0.4307689368724823,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9272263394484267,
+      "grad_norm": 0.44754841923713684,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9274693232900012,
+      "grad_norm": 0.41433072090148926,
+      "learning_rate": 8e-05,
+      "loss": 1.4898,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9277123071315757,
+      "grad_norm": 0.4276430904865265,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9279552909731503,
+      "grad_norm": 0.4168943166732788,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9281982748147248,
+      "grad_norm": 0.42808476090431213,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9284412586562993,
+      "grad_norm": 0.4125967025756836,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9286842424978738,
+      "grad_norm": 0.44815272092819214,
+      "learning_rate": 8e-05,
+      "loss": 1.675,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9289272263394485,
+      "grad_norm": 0.4305836260318756,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 3823
+    },
+    {
+      "epoch": 0.929170210181023,
+      "grad_norm": 0.44415712356567383,
+      "learning_rate": 8e-05,
+      "loss": 1.7498,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9294131940225975,
+      "grad_norm": 0.4288206100463867,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 3825
+    },
+    {
+      "epoch": 0.929656177864172,
+      "grad_norm": 0.4303172528743744,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9298991617057466,
+      "grad_norm": 0.42203474044799805,
+      "learning_rate": 8e-05,
+      "loss": 1.5863,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9301421455473211,
+      "grad_norm": 0.44495272636413574,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9303851293888956,
+      "grad_norm": 0.42819342017173767,
+      "learning_rate": 8e-05,
+      "loss": 1.4838,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9306281132304701,
+      "grad_norm": 0.462415874004364,
+      "learning_rate": 8e-05,
+      "loss": 1.7729,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9308710970720447,
+      "grad_norm": 0.44062116742134094,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9311140809136192,
+      "grad_norm": 0.4618380665779114,
+      "learning_rate": 8e-05,
+      "loss": 1.7803,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9313570647551938,
+      "grad_norm": 0.4144555926322937,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9316000485967683,
+      "grad_norm": 0.4324909746646881,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9318430324383429,
+      "grad_norm": 0.46868637204170227,
+      "learning_rate": 8e-05,
+      "loss": 1.7552,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9320860162799174,
+      "grad_norm": 0.4231494963169098,
+      "learning_rate": 8e-05,
+      "loss": 1.6184,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9323290001214919,
+      "grad_norm": 0.45624321699142456,
+      "learning_rate": 8e-05,
+      "loss": 1.7291,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9325719839630665,
+      "grad_norm": 0.4282325804233551,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 3838
+    },
+    {
+      "epoch": 0.932814967804641,
+      "grad_norm": 0.4465329647064209,
+      "learning_rate": 8e-05,
+      "loss": 1.8064,
+      "step": 3839
+    },
+    {
+      "epoch": 0.9330579516462155,
+      "grad_norm": 0.43566209077835083,
+      "learning_rate": 8e-05,
+      "loss": 1.7131,
+      "step": 3840
+    },
+    {
+      "epoch": 0.93330093548779,
+      "grad_norm": 0.4469618499279022,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9335439193293646,
+      "grad_norm": 0.4326150715351105,
+      "learning_rate": 8e-05,
+      "loss": 1.679,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9337869031709392,
+      "grad_norm": 0.4418710470199585,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9340298870125137,
+      "grad_norm": 0.4272327423095703,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9342728708540882,
+      "grad_norm": 0.40438759326934814,
+      "learning_rate": 8e-05,
+      "loss": 1.4998,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9345158546956628,
+      "grad_norm": 0.45255959033966064,
+      "learning_rate": 8e-05,
+      "loss": 1.8988,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9347588385372373,
+      "grad_norm": 0.429398775100708,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9350018223788118,
+      "grad_norm": 0.43159204721450806,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9352448062203863,
+      "grad_norm": 0.4394102990627289,
+      "learning_rate": 8e-05,
+      "loss": 1.801,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9354877900619609,
+      "grad_norm": 0.5241539478302002,
+      "learning_rate": 8e-05,
+      "loss": 1.861,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9357307739035354,
+      "grad_norm": 0.42874231934547424,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9359737577451099,
+      "grad_norm": 0.4416431784629822,
+      "learning_rate": 8e-05,
+      "loss": 1.7436,
+      "step": 3852
+    },
+    {
+      "epoch": 0.9362167415866844,
+      "grad_norm": 0.4375641942024231,
+      "learning_rate": 8e-05,
+      "loss": 1.6546,
+      "step": 3853
+    },
+    {
+      "epoch": 0.9364597254282591,
+      "grad_norm": 0.45421335101127625,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9367027092698336,
+      "grad_norm": 0.43949729204177856,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9369456931114081,
+      "grad_norm": 0.41936421394348145,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9371886769529826,
+      "grad_norm": 0.4321996867656708,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9374316607945572,
+      "grad_norm": 0.4504789710044861,
+      "learning_rate": 8e-05,
+      "loss": 1.7593,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9376746446361317,
+      "grad_norm": 0.4356359541416168,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9379176284777062,
+      "grad_norm": 0.4679346978664398,
+      "learning_rate": 8e-05,
+      "loss": 1.7676,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9381606123192807,
+      "grad_norm": 0.4798983335494995,
+      "learning_rate": 8e-05,
+      "loss": 1.8808,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9384035961608553,
+      "grad_norm": 0.41880279779434204,
+      "learning_rate": 8e-05,
+      "loss": 1.4909,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9386465800024298,
+      "grad_norm": 0.4715595841407776,
+      "learning_rate": 8e-05,
+      "loss": 1.8462,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9388895638440043,
+      "grad_norm": 0.4375055134296417,
+      "learning_rate": 8e-05,
+      "loss": 1.7915,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9391325476855789,
+      "grad_norm": 0.43628376722335815,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9393755315271535,
+      "grad_norm": 0.4589753746986389,
+      "learning_rate": 8e-05,
+      "loss": 1.8512,
+      "step": 3866
+    },
+    {
+      "epoch": 0.939618515368728,
+      "grad_norm": 0.4741183817386627,
+      "learning_rate": 8e-05,
+      "loss": 1.9218,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9398614992103025,
+      "grad_norm": 0.4436606764793396,
+      "learning_rate": 8e-05,
+      "loss": 1.7764,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9401044830518771,
+      "grad_norm": 0.4502270817756653,
+      "learning_rate": 8e-05,
+      "loss": 1.6611,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9403474668934516,
+      "grad_norm": 0.4147776663303375,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9405904507350261,
+      "grad_norm": 0.44295233488082886,
+      "learning_rate": 8e-05,
+      "loss": 1.7112,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9408334345766006,
+      "grad_norm": 0.43181100487709045,
+      "learning_rate": 8e-05,
+      "loss": 1.6158,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9410764184181752,
+      "grad_norm": 0.4307020306587219,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9413194022597497,
+      "grad_norm": 0.451726496219635,
+      "learning_rate": 8e-05,
+      "loss": 1.8012,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9415623861013243,
+      "grad_norm": 0.42303499579429626,
+      "learning_rate": 8e-05,
+      "loss": 1.6685,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9418053699428988,
+      "grad_norm": 0.4335882365703583,
+      "learning_rate": 8e-05,
+      "loss": 1.7952,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9420483537844734,
+      "grad_norm": 0.46532154083251953,
+      "learning_rate": 8e-05,
+      "loss": 1.7478,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9422913376260479,
+      "grad_norm": 0.43461593985557556,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9425343214676224,
+      "grad_norm": 0.4340265989303589,
+      "learning_rate": 8e-05,
+      "loss": 1.7306,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9427773053091969,
+      "grad_norm": 0.4898430407047272,
+      "learning_rate": 8e-05,
+      "loss": 1.8592,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9430202891507715,
+      "grad_norm": 0.47340986132621765,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 3881
+    },
+    {
+      "epoch": 0.943263272992346,
+      "grad_norm": 0.41468220949172974,
+      "learning_rate": 8e-05,
+      "loss": 1.4892,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9435062568339205,
+      "grad_norm": 0.4307178258895874,
+      "learning_rate": 8e-05,
+      "loss": 1.643,
+      "step": 3883
+    },
+    {
+      "epoch": 0.943749240675495,
+      "grad_norm": 0.40235522389411926,
+      "learning_rate": 8e-05,
+      "loss": 1.476,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9439922245170697,
+      "grad_norm": 0.4372292757034302,
+      "learning_rate": 8e-05,
+      "loss": 1.7015,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9442352083586442,
+      "grad_norm": 0.4738166034221649,
+      "learning_rate": 8e-05,
+      "loss": 1.6367,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9444781922002187,
+      "grad_norm": 0.441791296005249,
+      "learning_rate": 8e-05,
+      "loss": 1.6415,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9447211760417932,
+      "grad_norm": 0.47080785036087036,
+      "learning_rate": 8e-05,
+      "loss": 1.6769,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9449641598833678,
+      "grad_norm": 0.4464341700077057,
+      "learning_rate": 8e-05,
+      "loss": 1.7424,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9452071437249423,
+      "grad_norm": 0.44120916724205017,
+      "learning_rate": 8e-05,
+      "loss": 1.7818,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9454501275665168,
+      "grad_norm": 0.4526466727256775,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9456931114080913,
+      "grad_norm": 0.42678770422935486,
+      "learning_rate": 8e-05,
+      "loss": 1.628,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9459360952496659,
+      "grad_norm": 0.44390010833740234,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9461790790912404,
+      "grad_norm": 0.43124493956565857,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 3894
+    },
+    {
+      "epoch": 0.946422062932815,
+      "grad_norm": 0.4441266357898712,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9466650467743895,
+      "grad_norm": 0.43033742904663086,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9469080306159641,
+      "grad_norm": 0.4417329728603363,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9471510144575386,
+      "grad_norm": 0.4591132700443268,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9473939982991131,
+      "grad_norm": 0.4692525863647461,
+      "learning_rate": 8e-05,
+      "loss": 1.7429,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9476369821406876,
+      "grad_norm": 0.41225874423980713,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9478799659822622,
+      "grad_norm": 0.4399300813674927,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9481229498238367,
+      "grad_norm": 0.4374392032623291,
+      "learning_rate": 8e-05,
+      "loss": 1.5804,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9483659336654112,
+      "grad_norm": 0.43579840660095215,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9486089175069858,
+      "grad_norm": 0.44514647126197815,
+      "learning_rate": 8e-05,
+      "loss": 1.6871,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9488519013485603,
+      "grad_norm": 0.4231220483779907,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9490948851901349,
+      "grad_norm": 0.42387354373931885,
+      "learning_rate": 8e-05,
+      "loss": 1.5549,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9493378690317094,
+      "grad_norm": 0.4361368417739868,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 3907
+    },
+    {
+      "epoch": 0.949580852873284,
+      "grad_norm": 0.45002809166908264,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9498238367148585,
+      "grad_norm": 0.4487747251987457,
+      "learning_rate": 8e-05,
+      "loss": 1.6767,
+      "step": 3909
+    },
+    {
+      "epoch": 0.950066820556433,
+      "grad_norm": 0.4275915026664734,
+      "learning_rate": 8e-05,
+      "loss": 1.7244,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9503098043980075,
+      "grad_norm": 0.4531268775463104,
+      "learning_rate": 8e-05,
+      "loss": 1.792,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9505527882395821,
+      "grad_norm": 0.4469793438911438,
+      "learning_rate": 8e-05,
+      "loss": 1.6944,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9507957720811566,
+      "grad_norm": 0.47670620679855347,
+      "learning_rate": 8e-05,
+      "loss": 1.8372,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9510387559227311,
+      "grad_norm": 0.4273851215839386,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9512817397643056,
+      "grad_norm": 0.43185049295425415,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9515247236058803,
+      "grad_norm": 0.45217883586883545,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9517677074474548,
+      "grad_norm": 0.46018561720848083,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9520106912890293,
+      "grad_norm": 0.4508971571922302,
+      "learning_rate": 8e-05,
+      "loss": 1.6576,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9522536751306038,
+      "grad_norm": 0.45906463265419006,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9524966589721784,
+      "grad_norm": 0.45052576065063477,
+      "learning_rate": 8e-05,
+      "loss": 1.6121,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9527396428137529,
+      "grad_norm": 0.4389657974243164,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9529826266553274,
+      "grad_norm": 0.44085535407066345,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9532256104969019,
+      "grad_norm": 0.42999693751335144,
+      "learning_rate": 8e-05,
+      "loss": 1.5155,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9534685943384765,
+      "grad_norm": 0.44076135754585266,
+      "learning_rate": 8e-05,
+      "loss": 1.6239,
+      "step": 3924
+    },
+    {
+      "epoch": 0.953711578180051,
+      "grad_norm": 0.4463903307914734,
+      "learning_rate": 8e-05,
+      "loss": 1.7302,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9539545620216255,
+      "grad_norm": 0.4058784246444702,
+      "learning_rate": 8e-05,
+      "loss": 1.5399,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9541975458632,
+      "grad_norm": 0.4075396955013275,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9544405297047747,
+      "grad_norm": 0.4768963158130646,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9546835135463492,
+      "grad_norm": 0.44637706875801086,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9549264973879237,
+      "grad_norm": 0.4580327868461609,
+      "learning_rate": 8e-05,
+      "loss": 1.7074,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9551694812294982,
+      "grad_norm": 0.4614556133747101,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9554124650710728,
+      "grad_norm": 0.4647141993045807,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9556554489126473,
+      "grad_norm": 0.46834543347358704,
+      "learning_rate": 8e-05,
+      "loss": 1.8585,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9558984327542218,
+      "grad_norm": 0.4443167746067047,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9561414165957964,
+      "grad_norm": 0.4312688410282135,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9563844004373709,
+      "grad_norm": 0.43566352128982544,
+      "learning_rate": 8e-05,
+      "loss": 1.6819,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9566273842789454,
+      "grad_norm": 0.4570044279098511,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 3937
+    },
+    {
+      "epoch": 0.95687036812052,
+      "grad_norm": 0.41685932874679565,
+      "learning_rate": 8e-05,
+      "loss": 1.4497,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9571133519620946,
+      "grad_norm": 0.45033618807792664,
+      "learning_rate": 8e-05,
+      "loss": 1.6775,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9573563358036691,
+      "grad_norm": 0.42552435398101807,
+      "learning_rate": 8e-05,
+      "loss": 1.6552,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9575993196452436,
+      "grad_norm": 0.4351639449596405,
+      "learning_rate": 8e-05,
+      "loss": 1.6942,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9578423034868181,
+      "grad_norm": 0.4668383002281189,
+      "learning_rate": 8e-05,
+      "loss": 1.7461,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9580852873283927,
+      "grad_norm": 0.4245462715625763,
+      "learning_rate": 8e-05,
+      "loss": 1.6618,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9583282711699672,
+      "grad_norm": 0.4555746614933014,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9585712550115417,
+      "grad_norm": 0.45726481080055237,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9588142388531162,
+      "grad_norm": 0.4600939154624939,
+      "learning_rate": 8e-05,
+      "loss": 1.7468,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9590572226946908,
+      "grad_norm": 0.4585624039173126,
+      "learning_rate": 8e-05,
+      "loss": 1.8054,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9593002065362654,
+      "grad_norm": 0.44232597947120667,
+      "learning_rate": 8e-05,
+      "loss": 1.7861,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9595431903778399,
+      "grad_norm": 0.4450703561306,
+      "learning_rate": 8e-05,
+      "loss": 1.7983,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9597861742194144,
+      "grad_norm": 0.4300907254219055,
+      "learning_rate": 8e-05,
+      "loss": 1.6232,
+      "step": 3950
+    },
+    {
+      "epoch": 0.960029158060989,
+      "grad_norm": 0.44462138414382935,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9602721419025635,
+      "grad_norm": 0.4591369032859802,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 3952
+    },
+    {
+      "epoch": 0.960515125744138,
+      "grad_norm": 0.4237477481365204,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9607581095857125,
+      "grad_norm": 0.4457152485847473,
+      "learning_rate": 8e-05,
+      "loss": 1.6112,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9610010934272871,
+      "grad_norm": 0.4420410394668579,
+      "learning_rate": 8e-05,
+      "loss": 1.7181,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9612440772688616,
+      "grad_norm": 0.43850862979888916,
+      "learning_rate": 8e-05,
+      "loss": 1.6103,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9614870611104361,
+      "grad_norm": 0.4496399164199829,
+      "learning_rate": 8e-05,
+      "loss": 1.7205,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9617300449520106,
+      "grad_norm": 0.4551542401313782,
+      "learning_rate": 8e-05,
+      "loss": 1.6197,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9619730287935853,
+      "grad_norm": 0.46938806772232056,
+      "learning_rate": 8e-05,
+      "loss": 1.7807,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9622160126351598,
+      "grad_norm": 0.4382041394710541,
+      "learning_rate": 8e-05,
+      "loss": 1.5484,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9624589964767343,
+      "grad_norm": 0.4262560307979584,
+      "learning_rate": 8e-05,
+      "loss": 1.7816,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9627019803183088,
+      "grad_norm": 0.44450390338897705,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9629449641598834,
+      "grad_norm": 0.42746517062187195,
+      "learning_rate": 8e-05,
+      "loss": 1.5365,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9631879480014579,
+      "grad_norm": 0.4284069836139679,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9634309318430324,
+      "grad_norm": 0.4214082658290863,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9636739156846069,
+      "grad_norm": 0.46706804633140564,
+      "learning_rate": 8e-05,
+      "loss": 1.7296,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9639168995261815,
+      "grad_norm": 0.44554388523101807,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 3967
+    },
+    {
+      "epoch": 0.964159883367756,
+      "grad_norm": 0.4561922550201416,
+      "learning_rate": 8e-05,
+      "loss": 1.6789,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9644028672093305,
+      "grad_norm": 0.4489273726940155,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9646458510509052,
+      "grad_norm": 0.4501919448375702,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9648888348924797,
+      "grad_norm": 0.4532281458377838,
+      "learning_rate": 8e-05,
+      "loss": 1.883,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9651318187340542,
+      "grad_norm": 0.4407244920730591,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9653748025756287,
+      "grad_norm": 0.49307096004486084,
+      "learning_rate": 8e-05,
+      "loss": 1.8904,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9656177864172033,
+      "grad_norm": 0.42403313517570496,
+      "learning_rate": 8e-05,
+      "loss": 1.6812,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9658607702587778,
+      "grad_norm": 0.4292251765727997,
+      "learning_rate": 8e-05,
+      "loss": 1.6418,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9661037541003523,
+      "grad_norm": 0.40331390500068665,
+      "learning_rate": 8e-05,
+      "loss": 1.6009,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9663467379419268,
+      "grad_norm": 0.4302370846271515,
+      "learning_rate": 8e-05,
+      "loss": 1.7002,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9665897217835014,
+      "grad_norm": 0.42405349016189575,
+      "learning_rate": 8e-05,
+      "loss": 1.571,
+      "step": 3978
+    },
+    {
+      "epoch": 0.966832705625076,
+      "grad_norm": 0.459858775138855,
+      "learning_rate": 8e-05,
+      "loss": 1.7839,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9670756894666505,
+      "grad_norm": 0.42601335048675537,
+      "learning_rate": 8e-05,
+      "loss": 1.698,
+      "step": 3980
+    },
+    {
+      "epoch": 0.967318673308225,
+      "grad_norm": 0.4495616555213928,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9675616571497996,
+      "grad_norm": 0.4324414134025574,
+      "learning_rate": 8e-05,
+      "loss": 1.6636,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9678046409913741,
+      "grad_norm": 0.4413282871246338,
+      "learning_rate": 8e-05,
+      "loss": 1.578,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9680476248329486,
+      "grad_norm": 0.4280273914337158,
+      "learning_rate": 8e-05,
+      "loss": 1.7238,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9682906086745231,
+      "grad_norm": 0.4312518239021301,
+      "learning_rate": 8e-05,
+      "loss": 1.5317,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9685335925160977,
+      "grad_norm": 0.4215475618839264,
+      "learning_rate": 8e-05,
+      "loss": 1.4422,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9687765763576722,
+      "grad_norm": 0.46008285880088806,
+      "learning_rate": 8e-05,
+      "loss": 1.7972,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9690195601992467,
+      "grad_norm": 0.42137765884399414,
+      "learning_rate": 8e-05,
+      "loss": 1.606,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9692625440408212,
+      "grad_norm": 0.4334642291069031,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9695055278823959,
+      "grad_norm": 0.45535993576049805,
+      "learning_rate": 8e-05,
+      "loss": 1.6945,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9697485117239704,
+      "grad_norm": 0.4822055697441101,
+      "learning_rate": 8e-05,
+      "loss": 1.79,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9699914955655449,
+      "grad_norm": 0.42554226517677307,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9702344794071194,
+      "grad_norm": 0.4345811605453491,
+      "learning_rate": 8e-05,
+      "loss": 1.5143,
+      "step": 3993
+    },
+    {
+      "epoch": 0.970477463248694,
+      "grad_norm": 0.42335784435272217,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9707204470902685,
+      "grad_norm": 0.4500880539417267,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 3995
+    },
+    {
+      "epoch": 0.970963430931843,
+      "grad_norm": 0.4286655783653259,
+      "learning_rate": 8e-05,
+      "loss": 1.6573,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9712064147734175,
+      "grad_norm": 0.44797059893608093,
+      "learning_rate": 8e-05,
+      "loss": 1.6719,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9714493986149921,
+      "grad_norm": 0.4294436275959015,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9716923824565666,
+      "grad_norm": 0.4353179633617401,
+      "learning_rate": 8e-05,
+      "loss": 1.7842,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9719353662981411,
+      "grad_norm": 0.4457424581050873,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9721783501397158,
+      "grad_norm": 0.4767199456691742,
+      "learning_rate": 8e-05,
+      "loss": 1.6238,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9724213339812903,
+      "grad_norm": 0.4406599998474121,
+      "learning_rate": 8e-05,
+      "loss": 1.728,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9726643178228648,
+      "grad_norm": 0.41611844301223755,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9729073016644393,
+      "grad_norm": 0.4021282494068146,
+      "learning_rate": 8e-05,
+      "loss": 1.6468,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9731502855060139,
+      "grad_norm": 0.4623206853866577,
+      "learning_rate": 8e-05,
+      "loss": 1.8006,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9733932693475884,
+      "grad_norm": 0.43975308537483215,
+      "learning_rate": 8e-05,
+      "loss": 1.6068,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9736362531891629,
+      "grad_norm": 0.4245419204235077,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9738792370307374,
+      "grad_norm": 0.44398415088653564,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 4008
+    },
+    {
+      "epoch": 0.974122220872312,
+      "grad_norm": 0.4648004174232483,
+      "learning_rate": 8e-05,
+      "loss": 1.7698,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9743652047138865,
+      "grad_norm": 0.4504488408565521,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 4010
+    },
+    {
+      "epoch": 0.974608188555461,
+      "grad_norm": 0.42932868003845215,
+      "learning_rate": 8e-05,
+      "loss": 1.7208,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9748511723970356,
+      "grad_norm": 0.4134998023509979,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9750941562386102,
+      "grad_norm": 0.4421544671058655,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9753371400801847,
+      "grad_norm": 0.42658910155296326,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9755801239217592,
+      "grad_norm": 0.4376913607120514,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9758231077633337,
+      "grad_norm": 0.45027095079421997,
+      "learning_rate": 8e-05,
+      "loss": 1.7413,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9760660916049083,
+      "grad_norm": 0.4712527096271515,
+      "learning_rate": 8e-05,
+      "loss": 1.5724,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9763090754464828,
+      "grad_norm": 0.4317570626735687,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9765520592880573,
+      "grad_norm": 0.4263004660606384,
+      "learning_rate": 8e-05,
+      "loss": 1.7571,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9767950431296318,
+      "grad_norm": 0.4242319166660309,
+      "learning_rate": 8e-05,
+      "loss": 1.7998,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9770380269712065,
+      "grad_norm": 0.4449961483478546,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 4021
+    },
+    {
+      "epoch": 0.977281010812781,
+      "grad_norm": 0.43382737040519714,
+      "learning_rate": 8e-05,
+      "loss": 1.5696,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9775239946543555,
+      "grad_norm": 0.43983548879623413,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 4023
+    },
+    {
+      "epoch": 0.97776697849593,
+      "grad_norm": 0.4574201703071594,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9780099623375046,
+      "grad_norm": 0.45253363251686096,
+      "learning_rate": 8e-05,
+      "loss": 1.7293,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9782529461790791,
+      "grad_norm": 0.473109632730484,
+      "learning_rate": 8e-05,
+      "loss": 1.9134,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9784959300206536,
+      "grad_norm": 0.43669694662094116,
+      "learning_rate": 8e-05,
+      "loss": 1.6908,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9787389138622281,
+      "grad_norm": 0.42991501092910767,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9789818977038027,
+      "grad_norm": 0.435087651014328,
+      "learning_rate": 8e-05,
+      "loss": 1.5382,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9792248815453772,
+      "grad_norm": 0.4384804368019104,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9794678653869517,
+      "grad_norm": 0.4694783389568329,
+      "learning_rate": 8e-05,
+      "loss": 1.6915,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9797108492285262,
+      "grad_norm": 0.4328809082508087,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9799538330701009,
+      "grad_norm": 0.45652657747268677,
+      "learning_rate": 8e-05,
+      "loss": 1.687,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9801968169116754,
+      "grad_norm": 0.45094016194343567,
+      "learning_rate": 8e-05,
+      "loss": 1.607,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9804398007532499,
+      "grad_norm": 0.4609827399253845,
+      "learning_rate": 8e-05,
+      "loss": 1.6923,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9806827845948245,
+      "grad_norm": 0.429403156042099,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 4036
+    },
+    {
+      "epoch": 0.980925768436399,
+      "grad_norm": 0.44226983189582825,
+      "learning_rate": 8e-05,
+      "loss": 1.7299,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9811687522779735,
+      "grad_norm": 0.42094704508781433,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 4038
+    },
+    {
+      "epoch": 0.981411736119548,
+      "grad_norm": 0.4412955045700073,
+      "learning_rate": 8e-05,
+      "loss": 1.6089,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9816547199611226,
+      "grad_norm": 0.4200059771537781,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9818977038026971,
+      "grad_norm": 0.4401897192001343,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9821406876442716,
+      "grad_norm": 0.4275898337364197,
+      "learning_rate": 8e-05,
+      "loss": 1.7538,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9823836714858462,
+      "grad_norm": 0.45740917325019836,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9826266553274208,
+      "grad_norm": 0.439273476600647,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9828696391689953,
+      "grad_norm": 0.4535593092441559,
+      "learning_rate": 8e-05,
+      "loss": 1.6918,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9831126230105698,
+      "grad_norm": 0.4854368269443512,
+      "learning_rate": 8e-05,
+      "loss": 1.6927,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9833556068521443,
+      "grad_norm": 0.42863789200782776,
+      "learning_rate": 8e-05,
+      "loss": 1.7046,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9835985906937189,
+      "grad_norm": 0.4711615741252899,
+      "learning_rate": 8e-05,
+      "loss": 1.8057,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9838415745352934,
+      "grad_norm": 0.4448649287223816,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9840845583768679,
+      "grad_norm": 0.451811820268631,
+      "learning_rate": 8e-05,
+      "loss": 1.7449,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9843275422184424,
+      "grad_norm": 0.5398558974266052,
+      "learning_rate": 8e-05,
+      "loss": 1.7791,
+      "step": 4051
+    },
+    {
+      "epoch": 0.984570526060017,
+      "grad_norm": 0.44177329540252686,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9848135099015916,
+      "grad_norm": 0.45150044560432434,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9850564937431661,
+      "grad_norm": 0.43011075258255005,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9852994775847406,
+      "grad_norm": 0.4788955748081207,
+      "learning_rate": 8e-05,
+      "loss": 1.6562,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9855424614263152,
+      "grad_norm": 0.44742095470428467,
+      "learning_rate": 8e-05,
+      "loss": 1.6291,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9857854452678897,
+      "grad_norm": 0.46386680006980896,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9860284291094642,
+      "grad_norm": 0.47185713052749634,
+      "learning_rate": 8e-05,
+      "loss": 1.8072,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9862714129510387,
+      "grad_norm": 0.4229238033294678,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9865143967926133,
+      "grad_norm": 0.45910993218421936,
+      "learning_rate": 8e-05,
+      "loss": 1.6978,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9867573806341878,
+      "grad_norm": 0.4431948661804199,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9870003644757623,
+      "grad_norm": 0.4927588105201721,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9872433483173368,
+      "grad_norm": 0.44559353590011597,
+      "learning_rate": 8e-05,
+      "loss": 1.7012,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9874863321589115,
+      "grad_norm": 0.46200689673423767,
+      "learning_rate": 8e-05,
+      "loss": 1.7815,
+      "step": 4064
+    },
+    {
+      "epoch": 0.987729316000486,
+      "grad_norm": 0.42653772234916687,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9879722998420605,
+      "grad_norm": 0.4286855161190033,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9882152836836351,
+      "grad_norm": 0.4451533555984497,
+      "learning_rate": 8e-05,
+      "loss": 1.7342,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9884582675252096,
+      "grad_norm": 0.4371572434902191,
+      "learning_rate": 8e-05,
+      "loss": 1.6921,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9887012513667841,
+      "grad_norm": 0.44632038474082947,
+      "learning_rate": 8e-05,
+      "loss": 1.7375,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9889442352083586,
+      "grad_norm": 0.42659518122673035,
+      "learning_rate": 8e-05,
+      "loss": 1.5387,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9891872190499332,
+      "grad_norm": 0.4391607940196991,
+      "learning_rate": 8e-05,
+      "loss": 1.6069,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9894302028915077,
+      "grad_norm": 0.4430285096168518,
+      "learning_rate": 8e-05,
+      "loss": 1.7348,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9896731867330822,
+      "grad_norm": 0.44522327184677124,
+      "learning_rate": 8e-05,
+      "loss": 1.7549,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9899161705746568,
+      "grad_norm": 0.45676812529563904,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9901591544162314,
+      "grad_norm": 0.4433252215385437,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9904021382578059,
+      "grad_norm": 0.4801267683506012,
+      "learning_rate": 8e-05,
+      "loss": 1.7097,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9906451220993804,
+      "grad_norm": 0.45244312286376953,
+      "learning_rate": 8e-05,
+      "loss": 1.7259,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9908881059409549,
+      "grad_norm": 0.42037010192871094,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9911310897825295,
+      "grad_norm": 0.41117605566978455,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 4079
+    },
+    {
+      "epoch": 0.991374073624104,
+      "grad_norm": 0.43240705132484436,
+      "learning_rate": 8e-05,
+      "loss": 1.6446,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9916170574656785,
+      "grad_norm": 0.4600159227848053,
+      "learning_rate": 8e-05,
+      "loss": 1.7091,
+      "step": 4081
+    },
+    {
+      "epoch": 0.991860041307253,
+      "grad_norm": 0.4207499921321869,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9921030251488276,
+      "grad_norm": 0.4696202278137207,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9923460089904022,
+      "grad_norm": 0.4340289533138275,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9925889928319767,
+      "grad_norm": 0.4257737994194031,
+      "learning_rate": 8e-05,
+      "loss": 1.5115,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9928319766735512,
+      "grad_norm": 0.41688215732574463,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9930749605151258,
+      "grad_norm": 0.4607880115509033,
+      "learning_rate": 8e-05,
+      "loss": 1.5914,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9933179443567003,
+      "grad_norm": 0.4354183077812195,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9935609281982748,
+      "grad_norm": 0.43185368180274963,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9938039120398493,
+      "grad_norm": 0.43680867552757263,
+      "learning_rate": 8e-05,
+      "loss": 1.6987,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9940468958814239,
+      "grad_norm": 0.5325239300727844,
+      "learning_rate": 8e-05,
+      "loss": 1.7633,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9942898797229984,
+      "grad_norm": 0.4470573961734772,
+      "learning_rate": 8e-05,
+      "loss": 1.6782,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9945328635645729,
+      "grad_norm": 0.45053988695144653,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9947758474061474,
+      "grad_norm": 0.4452193081378937,
+      "learning_rate": 8e-05,
+      "loss": 1.7554,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9950188312477221,
+      "grad_norm": 0.434201180934906,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9952618150892966,
+      "grad_norm": 0.4479779899120331,
+      "learning_rate": 8e-05,
+      "loss": 1.6102,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9955047989308711,
+      "grad_norm": 0.42550167441368103,
+      "learning_rate": 8e-05,
+      "loss": 1.7215,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9957477827724456,
+      "grad_norm": 0.4378780424594879,
+      "learning_rate": 8e-05,
+      "loss": 1.7572,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9959907666140202,
+      "grad_norm": 0.4461440145969391,
+      "learning_rate": 8e-05,
+      "loss": 1.6853,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9962337504555947,
+      "grad_norm": 0.4413752853870392,
+      "learning_rate": 8e-05,
+      "loss": 1.4906,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9964767342971692,
+      "grad_norm": 0.45257726311683655,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9967197181387438,
+      "grad_norm": 0.457055926322937,
+      "learning_rate": 8e-05,
+      "loss": 1.5767,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9969627019803183,
+      "grad_norm": 0.45002129673957825,
+      "learning_rate": 8e-05,
+      "loss": 1.6607,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9972056858218928,
+      "grad_norm": 0.4914070665836334,
+      "learning_rate": 8e-05,
+      "loss": 1.875,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9974486696634673,
+      "grad_norm": 0.43596139550209045,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 4105
+    },
+    {
+      "epoch": 0.997691653505042,
+      "grad_norm": 0.44118353724479675,
+      "learning_rate": 8e-05,
+      "loss": 1.7497,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9979346373466165,
+      "grad_norm": 0.44389355182647705,
+      "learning_rate": 8e-05,
+      "loss": 1.6199,
+      "step": 4107
+    },
+    {
+      "epoch": 0.998177621188191,
+      "grad_norm": 0.4240395426750183,
+      "learning_rate": 8e-05,
+      "loss": 1.5962,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9984206050297655,
+      "grad_norm": 0.41717231273651123,
+      "learning_rate": 8e-05,
+      "loss": 1.6061,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9986635888713401,
+      "grad_norm": 0.4284403622150421,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9989065727129146,
+      "grad_norm": 0.47413235902786255,
+      "learning_rate": 8e-05,
+      "loss": 1.8638,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9991495565544891,
+      "grad_norm": 0.4562981128692627,
+      "learning_rate": 8e-05,
+      "loss": 1.7536,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9993925403960636,
+      "grad_norm": 0.46152573823928833,
+      "learning_rate": 8e-05,
+      "loss": 1.8031,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9996355242376382,
+      "grad_norm": 0.43792322278022766,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9998785080792127,
+      "grad_norm": 0.4393792152404785,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 4115
+    },
+    {
+      "epoch": 1.0002429838415745,
+      "grad_norm": 0.4479010999202728,
+      "learning_rate": 8e-05,
+      "loss": 1.4698,
+      "step": 4116
+    },
+    {
+      "epoch": 1.000485967683149,
+      "grad_norm": 0.44824063777923584,
+      "learning_rate": 8e-05,
+      "loss": 1.4875,
+      "step": 4117
+    },
+    {
+      "epoch": 1.0007289515247235,
+      "grad_norm": 0.4471682608127594,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0009719353662982,
+      "grad_norm": 0.460489422082901,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 4119
+    },
+    {
+      "epoch": 1.0012149192078728,
+      "grad_norm": 0.520616352558136,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0014579030494473,
+      "grad_norm": 0.4648950695991516,
+      "learning_rate": 8e-05,
+      "loss": 1.5131,
+      "step": 4121
+    },
+    {
+      "epoch": 1.0017008868910218,
+      "grad_norm": 0.43558400869369507,
+      "learning_rate": 8e-05,
+      "loss": 1.4174,
+      "step": 4122
+    },
+    {
+      "epoch": 1.0019438707325963,
+      "grad_norm": 0.4830182194709778,
+      "learning_rate": 8e-05,
+      "loss": 1.5579,
+      "step": 4123
+    },
+    {
+      "epoch": 1.0021868545741708,
+      "grad_norm": 0.5019348859786987,
+      "learning_rate": 8e-05,
+      "loss": 1.5043,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0024298384157453,
+      "grad_norm": 0.5495268106460571,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0026728222573198,
+      "grad_norm": 0.4628094434738159,
+      "learning_rate": 8e-05,
+      "loss": 1.4746,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0029158060988945,
+      "grad_norm": 0.503748893737793,
+      "learning_rate": 8e-05,
+      "loss": 1.5289,
+      "step": 4127
+    },
+    {
+      "epoch": 1.003158789940469,
+      "grad_norm": 0.4710977375507355,
+      "learning_rate": 8e-05,
+      "loss": 1.5804,
+      "step": 4128
+    },
+    {
+      "epoch": 1.0034017737820435,
+      "grad_norm": 0.49901896715164185,
+      "learning_rate": 8e-05,
+      "loss": 1.4244,
+      "step": 4129
+    },
+    {
+      "epoch": 1.003644757623618,
+      "grad_norm": 0.475931316614151,
+      "learning_rate": 8e-05,
+      "loss": 1.5235,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0038877414651926,
+      "grad_norm": 0.4804648458957672,
+      "learning_rate": 8e-05,
+      "loss": 1.5823,
+      "step": 4131
+    },
+    {
+      "epoch": 1.004130725306767,
+      "grad_norm": 0.4405800402164459,
+      "learning_rate": 8e-05,
+      "loss": 1.5178,
+      "step": 4132
+    },
+    {
+      "epoch": 1.0043737091483416,
+      "grad_norm": 0.4687287509441376,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 4133
+    },
+    {
+      "epoch": 1.004616692989916,
+      "grad_norm": 0.43883663415908813,
+      "learning_rate": 8e-05,
+      "loss": 1.45,
+      "step": 4134
+    },
+    {
+      "epoch": 1.0048596768314908,
+      "grad_norm": 0.5092957615852356,
+      "learning_rate": 8e-05,
+      "loss": 1.6468,
+      "step": 4135
+    },
+    {
+      "epoch": 1.0051026606730653,
+      "grad_norm": 0.4919053912162781,
+      "learning_rate": 8e-05,
+      "loss": 1.6888,
+      "step": 4136
+    },
+    {
+      "epoch": 1.0053456445146398,
+      "grad_norm": 0.45662403106689453,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 4137
+    },
+    {
+      "epoch": 1.0055886283562143,
+      "grad_norm": 0.4511517584323883,
+      "learning_rate": 8e-05,
+      "loss": 1.5302,
+      "step": 4138
+    },
+    {
+      "epoch": 1.0058316121977888,
+      "grad_norm": 0.4731735289096832,
+      "learning_rate": 8e-05,
+      "loss": 1.5312,
+      "step": 4139
+    },
+    {
+      "epoch": 1.0060745960393633,
+      "grad_norm": 0.45367908477783203,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0063175798809378,
+      "grad_norm": 0.4449901282787323,
+      "learning_rate": 8e-05,
+      "loss": 1.4836,
+      "step": 4141
+    },
+    {
+      "epoch": 1.0065605637225123,
+      "grad_norm": 0.4540715515613556,
+      "learning_rate": 8e-05,
+      "loss": 1.5418,
+      "step": 4142
+    },
+    {
+      "epoch": 1.006803547564087,
+      "grad_norm": 0.46869245171546936,
+      "learning_rate": 8e-05,
+      "loss": 1.5505,
+      "step": 4143
+    },
+    {
+      "epoch": 1.0070465314056616,
+      "grad_norm": 0.45588356256484985,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 4144
+    },
+    {
+      "epoch": 1.007289515247236,
+      "grad_norm": 0.4640970826148987,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 4145
+    },
+    {
+      "epoch": 1.0075324990888106,
+      "grad_norm": 0.45515191555023193,
+      "learning_rate": 8e-05,
+      "loss": 1.4815,
+      "step": 4146
+    },
+    {
+      "epoch": 1.007775482930385,
+      "grad_norm": 0.4766829013824463,
+      "learning_rate": 8e-05,
+      "loss": 1.4471,
+      "step": 4147
+    },
+    {
+      "epoch": 1.0080184667719596,
+      "grad_norm": 0.45073580741882324,
+      "learning_rate": 8e-05,
+      "loss": 1.4595,
+      "step": 4148
+    },
+    {
+      "epoch": 1.0082614506135341,
+      "grad_norm": 0.48333314061164856,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 4149
+    },
+    {
+      "epoch": 1.0085044344551088,
+      "grad_norm": 0.44914644956588745,
+      "learning_rate": 8e-05,
+      "loss": 1.4672,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0087474182966834,
+      "grad_norm": 0.4693595767021179,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 4151
+    },
+    {
+      "epoch": 1.0089904021382579,
+      "grad_norm": 0.503622829914093,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 4152
+    },
+    {
+      "epoch": 1.0092333859798324,
+      "grad_norm": 0.4690110385417938,
+      "learning_rate": 8e-05,
+      "loss": 1.4948,
+      "step": 4153
+    },
+    {
+      "epoch": 1.0094763698214069,
+      "grad_norm": 0.46136602759361267,
+      "learning_rate": 8e-05,
+      "loss": 1.4775,
+      "step": 4154
+    },
+    {
+      "epoch": 1.0097193536629814,
+      "grad_norm": 0.4447779655456543,
+      "learning_rate": 8e-05,
+      "loss": 1.4163,
+      "step": 4155
+    },
+    {
+      "epoch": 1.0099623375045559,
+      "grad_norm": 0.4570035934448242,
+      "learning_rate": 8e-05,
+      "loss": 1.5822,
+      "step": 4156
+    },
+    {
+      "epoch": 1.0102053213461304,
+      "grad_norm": 0.466577410697937,
+      "learning_rate": 8e-05,
+      "loss": 1.5538,
+      "step": 4157
+    },
+    {
+      "epoch": 1.0104483051877051,
+      "grad_norm": 0.4630834460258484,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 4158
+    },
+    {
+      "epoch": 1.0106912890292796,
+      "grad_norm": 0.45885714888572693,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 4159
+    },
+    {
+      "epoch": 1.0109342728708541,
+      "grad_norm": 0.4896791875362396,
+      "learning_rate": 8e-05,
+      "loss": 1.4257,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0111772567124286,
+      "grad_norm": 0.4490632712841034,
+      "learning_rate": 8e-05,
+      "loss": 1.4346,
+      "step": 4161
+    },
+    {
+      "epoch": 1.0114202405540031,
+      "grad_norm": 0.4708739221096039,
+      "learning_rate": 8e-05,
+      "loss": 1.4962,
+      "step": 4162
+    },
+    {
+      "epoch": 1.0116632243955777,
+      "grad_norm": 0.5062789916992188,
+      "learning_rate": 8e-05,
+      "loss": 1.5963,
+      "step": 4163
+    },
+    {
+      "epoch": 1.0119062082371522,
+      "grad_norm": 0.47871389985084534,
+      "learning_rate": 8e-05,
+      "loss": 1.5161,
+      "step": 4164
+    },
+    {
+      "epoch": 1.0121491920787267,
+      "grad_norm": 0.49338245391845703,
+      "learning_rate": 8e-05,
+      "loss": 1.5657,
+      "step": 4165
+    },
+    {
+      "epoch": 1.0123921759203014,
+      "grad_norm": 0.49327173829078674,
+      "learning_rate": 8e-05,
+      "loss": 1.4956,
+      "step": 4166
+    },
+    {
+      "epoch": 1.012635159761876,
+      "grad_norm": 0.5024109482765198,
+      "learning_rate": 8e-05,
+      "loss": 1.6569,
+      "step": 4167
+    },
+    {
+      "epoch": 1.0128781436034504,
+      "grad_norm": 0.44707122445106506,
+      "learning_rate": 8e-05,
+      "loss": 1.401,
+      "step": 4168
+    },
+    {
+      "epoch": 1.013121127445025,
+      "grad_norm": 0.47040215134620667,
+      "learning_rate": 8e-05,
+      "loss": 1.5129,
+      "step": 4169
+    },
+    {
+      "epoch": 1.0133641112865994,
+      "grad_norm": 0.5325475335121155,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 4170
+    },
+    {
+      "epoch": 1.013607095128174,
+      "grad_norm": 0.49752360582351685,
+      "learning_rate": 8e-05,
+      "loss": 1.6174,
+      "step": 4171
+    },
+    {
+      "epoch": 1.0138500789697484,
+      "grad_norm": 0.43803757429122925,
+      "learning_rate": 8e-05,
+      "loss": 1.535,
+      "step": 4172
+    },
+    {
+      "epoch": 1.014093062811323,
+      "grad_norm": 0.4945295453071594,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 4173
+    },
+    {
+      "epoch": 1.0143360466528977,
+      "grad_norm": 0.5018302798271179,
+      "learning_rate": 8e-05,
+      "loss": 1.6294,
+      "step": 4174
+    },
+    {
+      "epoch": 1.0145790304944722,
+      "grad_norm": 0.4580550789833069,
+      "learning_rate": 8e-05,
+      "loss": 1.4831,
+      "step": 4175
+    },
+    {
+      "epoch": 1.0148220143360467,
+      "grad_norm": 0.48314008116722107,
+      "learning_rate": 8e-05,
+      "loss": 1.716,
+      "step": 4176
+    },
+    {
+      "epoch": 1.0150649981776212,
+      "grad_norm": 0.5108659863471985,
+      "learning_rate": 8e-05,
+      "loss": 1.8082,
+      "step": 4177
+    },
+    {
+      "epoch": 1.0153079820191957,
+      "grad_norm": 0.47268712520599365,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 4178
+    },
+    {
+      "epoch": 1.0155509658607702,
+      "grad_norm": 0.4512706995010376,
+      "learning_rate": 8e-05,
+      "loss": 1.625,
+      "step": 4179
+    },
+    {
+      "epoch": 1.0157939497023447,
+      "grad_norm": 0.4682491719722748,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0160369335439192,
+      "grad_norm": 0.4897303581237793,
+      "learning_rate": 8e-05,
+      "loss": 1.5166,
+      "step": 4181
+    },
+    {
+      "epoch": 1.016279917385494,
+      "grad_norm": 0.49180689454078674,
+      "learning_rate": 8e-05,
+      "loss": 1.631,
+      "step": 4182
+    },
+    {
+      "epoch": 1.0165229012270685,
+      "grad_norm": 0.45297256112098694,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 4183
+    },
+    {
+      "epoch": 1.016765885068643,
+      "grad_norm": 0.48363837599754333,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 4184
+    },
+    {
+      "epoch": 1.0170088689102175,
+      "grad_norm": 0.4772539734840393,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 4185
+    },
+    {
+      "epoch": 1.017251852751792,
+      "grad_norm": 0.47336944937705994,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 4186
+    },
+    {
+      "epoch": 1.0174948365933665,
+      "grad_norm": 0.4368659257888794,
+      "learning_rate": 8e-05,
+      "loss": 1.4888,
+      "step": 4187
+    },
+    {
+      "epoch": 1.017737820434941,
+      "grad_norm": 0.44567447900772095,
+      "learning_rate": 8e-05,
+      "loss": 1.3732,
+      "step": 4188
+    },
+    {
+      "epoch": 1.0179808042765157,
+      "grad_norm": 0.4640052020549774,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 4189
+    },
+    {
+      "epoch": 1.0182237881180902,
+      "grad_norm": 0.47982051968574524,
+      "learning_rate": 8e-05,
+      "loss": 1.5876,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0184667719596647,
+      "grad_norm": 0.44950589537620544,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 4191
+    },
+    {
+      "epoch": 1.0187097558012392,
+      "grad_norm": 0.4855518639087677,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 4192
+    },
+    {
+      "epoch": 1.0189527396428137,
+      "grad_norm": 0.4857620894908905,
+      "learning_rate": 8e-05,
+      "loss": 1.4661,
+      "step": 4193
+    },
+    {
+      "epoch": 1.0191957234843882,
+      "grad_norm": 0.438090056180954,
+      "learning_rate": 8e-05,
+      "loss": 1.5679,
+      "step": 4194
+    },
+    {
+      "epoch": 1.0194387073259628,
+      "grad_norm": 0.43027639389038086,
+      "learning_rate": 8e-05,
+      "loss": 1.4108,
+      "step": 4195
+    },
+    {
+      "epoch": 1.0196816911675373,
+      "grad_norm": 0.4818543791770935,
+      "learning_rate": 8e-05,
+      "loss": 1.6295,
+      "step": 4196
+    },
+    {
+      "epoch": 1.019924675009112,
+      "grad_norm": 0.4437964856624603,
+      "learning_rate": 8e-05,
+      "loss": 1.4359,
+      "step": 4197
+    },
+    {
+      "epoch": 1.0201676588506865,
+      "grad_norm": 0.46072128415107727,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 4198
+    },
+    {
+      "epoch": 1.020410642692261,
+      "grad_norm": 0.4764827787876129,
+      "learning_rate": 8e-05,
+      "loss": 1.5975,
+      "step": 4199
+    },
+    {
+      "epoch": 1.0206536265338355,
+      "grad_norm": 0.42438018321990967,
+      "learning_rate": 8e-05,
+      "loss": 1.2514,
+      "step": 4200
+    },
+    {
+      "epoch": 1.02089661037541,
+      "grad_norm": 0.43238332867622375,
+      "learning_rate": 8e-05,
+      "loss": 1.4977,
+      "step": 4201
+    },
+    {
+      "epoch": 1.0211395942169845,
+      "grad_norm": 0.4710623621940613,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 4202
+    },
+    {
+      "epoch": 1.021382578058559,
+      "grad_norm": 0.4809209406375885,
+      "learning_rate": 8e-05,
+      "loss": 1.5465,
+      "step": 4203
+    },
+    {
+      "epoch": 1.0216255619001335,
+      "grad_norm": 0.4535318911075592,
+      "learning_rate": 8e-05,
+      "loss": 1.3782,
+      "step": 4204
+    },
+    {
+      "epoch": 1.0218685457417083,
+      "grad_norm": 0.5450059175491333,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 4205
+    },
+    {
+      "epoch": 1.0221115295832828,
+      "grad_norm": 0.49440786242485046,
+      "learning_rate": 8e-05,
+      "loss": 1.4713,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0223545134248573,
+      "grad_norm": 0.4647062420845032,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 4207
+    },
+    {
+      "epoch": 1.0225974972664318,
+      "grad_norm": 0.47729116678237915,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 4208
+    },
+    {
+      "epoch": 1.0228404811080063,
+      "grad_norm": 0.49122363328933716,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 4209
+    },
+    {
+      "epoch": 1.0230834649495808,
+      "grad_norm": 0.477539986371994,
+      "learning_rate": 8e-05,
+      "loss": 1.3751,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0233264487911553,
+      "grad_norm": 0.4731014370918274,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 4211
+    },
+    {
+      "epoch": 1.0235694326327298,
+      "grad_norm": 0.5083236694335938,
+      "learning_rate": 8e-05,
+      "loss": 1.3147,
+      "step": 4212
+    },
+    {
+      "epoch": 1.0238124164743045,
+      "grad_norm": 0.491510808467865,
+      "learning_rate": 8e-05,
+      "loss": 1.4507,
+      "step": 4213
+    },
+    {
+      "epoch": 1.024055400315879,
+      "grad_norm": 0.4652339816093445,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 4214
+    },
+    {
+      "epoch": 1.0242983841574536,
+      "grad_norm": 0.4640326499938965,
+      "learning_rate": 8e-05,
+      "loss": 1.5527,
+      "step": 4215
+    },
+    {
+      "epoch": 1.024541367999028,
+      "grad_norm": 0.4783758521080017,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 4216
+    },
+    {
+      "epoch": 1.0247843518406026,
+      "grad_norm": 0.48230019211769104,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 4217
+    },
+    {
+      "epoch": 1.025027335682177,
+      "grad_norm": 0.4410772919654846,
+      "learning_rate": 8e-05,
+      "loss": 1.3945,
+      "step": 4218
+    },
+    {
+      "epoch": 1.0252703195237516,
+      "grad_norm": 0.45225733518600464,
+      "learning_rate": 8e-05,
+      "loss": 1.4848,
+      "step": 4219
+    },
+    {
+      "epoch": 1.0255133033653263,
+      "grad_norm": 0.4797126352787018,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 4220
+    },
+    {
+      "epoch": 1.0257562872069008,
+      "grad_norm": 0.482758104801178,
+      "learning_rate": 8e-05,
+      "loss": 1.4169,
+      "step": 4221
+    },
+    {
+      "epoch": 1.0259992710484753,
+      "grad_norm": 0.46662724018096924,
+      "learning_rate": 8e-05,
+      "loss": 1.564,
+      "step": 4222
+    },
+    {
+      "epoch": 1.0262422548900498,
+      "grad_norm": 0.47064152359962463,
+      "learning_rate": 8e-05,
+      "loss": 1.5344,
+      "step": 4223
+    },
+    {
+      "epoch": 1.0264852387316243,
+      "grad_norm": 0.4672752618789673,
+      "learning_rate": 8e-05,
+      "loss": 1.5456,
+      "step": 4224
+    },
+    {
+      "epoch": 1.0267282225731988,
+      "grad_norm": 0.4895230829715729,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 4225
+    },
+    {
+      "epoch": 1.0269712064147734,
+      "grad_norm": 0.45517873764038086,
+      "learning_rate": 8e-05,
+      "loss": 1.4696,
+      "step": 4226
+    },
+    {
+      "epoch": 1.0272141902563479,
+      "grad_norm": 0.47745412588119507,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 4227
+    },
+    {
+      "epoch": 1.0274571740979226,
+      "grad_norm": 0.45787400007247925,
+      "learning_rate": 8e-05,
+      "loss": 1.4856,
+      "step": 4228
+    },
+    {
+      "epoch": 1.027700157939497,
+      "grad_norm": 0.4915253520011902,
+      "learning_rate": 8e-05,
+      "loss": 1.5872,
+      "step": 4229
+    },
+    {
+      "epoch": 1.0279431417810716,
+      "grad_norm": 0.5053480267524719,
+      "learning_rate": 8e-05,
+      "loss": 1.6959,
+      "step": 4230
+    },
+    {
+      "epoch": 1.028186125622646,
+      "grad_norm": 0.4676015079021454,
+      "learning_rate": 8e-05,
+      "loss": 1.5137,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0284291094642206,
+      "grad_norm": 0.4821891486644745,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 4232
+    },
+    {
+      "epoch": 1.0286720933057951,
+      "grad_norm": 0.4566926062107086,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 4233
+    },
+    {
+      "epoch": 1.0289150771473696,
+      "grad_norm": 0.4543267488479614,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 4234
+    },
+    {
+      "epoch": 1.0291580609889441,
+      "grad_norm": 0.4805215895175934,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0294010448305189,
+      "grad_norm": 0.5220195055007935,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 4236
+    },
+    {
+      "epoch": 1.0296440286720934,
+      "grad_norm": 0.46338599920272827,
+      "learning_rate": 8e-05,
+      "loss": 1.5631,
+      "step": 4237
+    },
+    {
+      "epoch": 1.0298870125136679,
+      "grad_norm": 0.5296505093574524,
+      "learning_rate": 8e-05,
+      "loss": 1.5533,
+      "step": 4238
+    },
+    {
+      "epoch": 1.0301299963552424,
+      "grad_norm": 0.44303831458091736,
+      "learning_rate": 8e-05,
+      "loss": 1.4405,
+      "step": 4239
+    },
+    {
+      "epoch": 1.030372980196817,
+      "grad_norm": 0.5055634379386902,
+      "learning_rate": 8e-05,
+      "loss": 1.5885,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0306159640383914,
+      "grad_norm": 0.5126274824142456,
+      "learning_rate": 8e-05,
+      "loss": 1.7298,
+      "step": 4241
+    },
+    {
+      "epoch": 1.030858947879966,
+      "grad_norm": 0.4947178065776825,
+      "learning_rate": 8e-05,
+      "loss": 1.6433,
+      "step": 4242
+    },
+    {
+      "epoch": 1.0311019317215404,
+      "grad_norm": 0.45463287830352783,
+      "learning_rate": 8e-05,
+      "loss": 1.4966,
+      "step": 4243
+    },
+    {
+      "epoch": 1.0313449155631151,
+      "grad_norm": 0.4832444489002228,
+      "learning_rate": 8e-05,
+      "loss": 1.4348,
+      "step": 4244
+    },
+    {
+      "epoch": 1.0315878994046896,
+      "grad_norm": 0.4687455892562866,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0318308832462642,
+      "grad_norm": 0.4908648729324341,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 4246
+    },
+    {
+      "epoch": 1.0320738670878387,
+      "grad_norm": 0.470570832490921,
+      "learning_rate": 8e-05,
+      "loss": 1.6202,
+      "step": 4247
+    },
+    {
+      "epoch": 1.0323168509294132,
+      "grad_norm": 0.45871421694755554,
+      "learning_rate": 8e-05,
+      "loss": 1.4582,
+      "step": 4248
+    },
+    {
+      "epoch": 1.0325598347709877,
+      "grad_norm": 0.48522108793258667,
+      "learning_rate": 8e-05,
+      "loss": 1.5274,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0328028186125622,
+      "grad_norm": 0.4908321499824524,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 4250
+    },
+    {
+      "epoch": 1.033045802454137,
+      "grad_norm": 0.47638770937919617,
+      "learning_rate": 8e-05,
+      "loss": 1.5615,
+      "step": 4251
+    },
+    {
+      "epoch": 1.0332887862957114,
+      "grad_norm": 0.47000938653945923,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 4252
+    },
+    {
+      "epoch": 1.033531770137286,
+      "grad_norm": 0.47024354338645935,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 4253
+    },
+    {
+      "epoch": 1.0337747539788604,
+      "grad_norm": 0.4773096740245819,
+      "learning_rate": 8e-05,
+      "loss": 1.5757,
+      "step": 4254
+    },
+    {
+      "epoch": 1.034017737820435,
+      "grad_norm": 0.468064546585083,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0342607216620094,
+      "grad_norm": 0.4880591928958893,
+      "learning_rate": 8e-05,
+      "loss": 1.518,
+      "step": 4256
+    },
+    {
+      "epoch": 1.034503705503584,
+      "grad_norm": 0.5044259428977966,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 4257
+    },
+    {
+      "epoch": 1.0347466893451585,
+      "grad_norm": 0.48384636640548706,
+      "learning_rate": 8e-05,
+      "loss": 1.5485,
+      "step": 4258
+    },
+    {
+      "epoch": 1.0349896731867332,
+      "grad_norm": 0.47855108976364136,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 4259
+    },
+    {
+      "epoch": 1.0352326570283077,
+      "grad_norm": 0.43458762764930725,
+      "learning_rate": 8e-05,
+      "loss": 1.2764,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0354756408698822,
+      "grad_norm": 0.4868089258670807,
+      "learning_rate": 8e-05,
+      "loss": 1.5718,
+      "step": 4261
+    },
+    {
+      "epoch": 1.0357186247114567,
+      "grad_norm": 0.4686622619628906,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 4262
+    },
+    {
+      "epoch": 1.0359616085530312,
+      "grad_norm": 0.47472110390663147,
+      "learning_rate": 8e-05,
+      "loss": 1.5018,
+      "step": 4263
+    },
+    {
+      "epoch": 1.0362045923946057,
+      "grad_norm": 0.46865585446357727,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 4264
+    },
+    {
+      "epoch": 1.0364475762361802,
+      "grad_norm": 0.46429798007011414,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0366905600777547,
+      "grad_norm": 0.46401089429855347,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 4266
+    },
+    {
+      "epoch": 1.0369335439193295,
+      "grad_norm": 0.4745710492134094,
+      "learning_rate": 8e-05,
+      "loss": 1.611,
+      "step": 4267
+    },
+    {
+      "epoch": 1.037176527760904,
+      "grad_norm": 0.4850454032421112,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 4268
+    },
+    {
+      "epoch": 1.0374195116024785,
+      "grad_norm": 0.5099414587020874,
+      "learning_rate": 8e-05,
+      "loss": 1.7188,
+      "step": 4269
+    },
+    {
+      "epoch": 1.037662495444053,
+      "grad_norm": 0.46733108162879944,
+      "learning_rate": 8e-05,
+      "loss": 1.5896,
+      "step": 4270
+    },
+    {
+      "epoch": 1.0379054792856275,
+      "grad_norm": 0.5027939081192017,
+      "learning_rate": 8e-05,
+      "loss": 1.5509,
+      "step": 4271
+    },
+    {
+      "epoch": 1.038148463127202,
+      "grad_norm": 0.49305927753448486,
+      "learning_rate": 8e-05,
+      "loss": 1.5527,
+      "step": 4272
+    },
+    {
+      "epoch": 1.0383914469687765,
+      "grad_norm": 0.4683418273925781,
+      "learning_rate": 8e-05,
+      "loss": 1.4095,
+      "step": 4273
+    },
+    {
+      "epoch": 1.038634430810351,
+      "grad_norm": 0.46370837092399597,
+      "learning_rate": 8e-05,
+      "loss": 1.6211,
+      "step": 4274
+    },
+    {
+      "epoch": 1.0388774146519257,
+      "grad_norm": 0.4825771749019623,
+      "learning_rate": 8e-05,
+      "loss": 1.4581,
+      "step": 4275
+    },
+    {
+      "epoch": 1.0391203984935002,
+      "grad_norm": 0.45184585452079773,
+      "learning_rate": 8e-05,
+      "loss": 1.5611,
+      "step": 4276
+    },
+    {
+      "epoch": 1.0393633823350747,
+      "grad_norm": 0.5146018266677856,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 4277
+    },
+    {
+      "epoch": 1.0396063661766493,
+      "grad_norm": 0.46783193945884705,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 4278
+    },
+    {
+      "epoch": 1.0398493500182238,
+      "grad_norm": 0.4465492069721222,
+      "learning_rate": 8e-05,
+      "loss": 1.5246,
+      "step": 4279
+    },
+    {
+      "epoch": 1.0400923338597983,
+      "grad_norm": 0.4659942090511322,
+      "learning_rate": 8e-05,
+      "loss": 1.5237,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0403353177013728,
+      "grad_norm": 0.4768926501274109,
+      "learning_rate": 8e-05,
+      "loss": 1.47,
+      "step": 4281
+    },
+    {
+      "epoch": 1.0405783015429475,
+      "grad_norm": 0.47466230392456055,
+      "learning_rate": 8e-05,
+      "loss": 1.6665,
+      "step": 4282
+    },
+    {
+      "epoch": 1.040821285384522,
+      "grad_norm": 0.5322874784469604,
+      "learning_rate": 8e-05,
+      "loss": 1.8133,
+      "step": 4283
+    },
+    {
+      "epoch": 1.0410642692260965,
+      "grad_norm": 0.4453260004520416,
+      "learning_rate": 8e-05,
+      "loss": 1.4616,
+      "step": 4284
+    },
+    {
+      "epoch": 1.041307253067671,
+      "grad_norm": 0.5126699209213257,
+      "learning_rate": 8e-05,
+      "loss": 1.6839,
+      "step": 4285
+    },
+    {
+      "epoch": 1.0415502369092455,
+      "grad_norm": 0.4786202311515808,
+      "learning_rate": 8e-05,
+      "loss": 1.6831,
+      "step": 4286
+    },
+    {
+      "epoch": 1.04179322075082,
+      "grad_norm": 0.5078117251396179,
+      "learning_rate": 8e-05,
+      "loss": 1.4928,
+      "step": 4287
+    },
+    {
+      "epoch": 1.0420362045923945,
+      "grad_norm": 0.4736517667770386,
+      "learning_rate": 8e-05,
+      "loss": 1.5996,
+      "step": 4288
+    },
+    {
+      "epoch": 1.042279188433969,
+      "grad_norm": 0.4480752944946289,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 4289
+    },
+    {
+      "epoch": 1.0425221722755438,
+      "grad_norm": 0.48934248089790344,
+      "learning_rate": 8e-05,
+      "loss": 1.5663,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0427651561171183,
+      "grad_norm": 0.4530789256095886,
+      "learning_rate": 8e-05,
+      "loss": 1.458,
+      "step": 4291
+    },
+    {
+      "epoch": 1.0430081399586928,
+      "grad_norm": 0.4945630431175232,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 4292
+    },
+    {
+      "epoch": 1.0432511238002673,
+      "grad_norm": 0.4807519018650055,
+      "learning_rate": 8e-05,
+      "loss": 1.5787,
+      "step": 4293
+    },
+    {
+      "epoch": 1.0434941076418418,
+      "grad_norm": 0.4709617495536804,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 4294
+    },
+    {
+      "epoch": 1.0437370914834163,
+      "grad_norm": 0.4870735704898834,
+      "learning_rate": 8e-05,
+      "loss": 1.6883,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0439800753249908,
+      "grad_norm": 0.49594011902809143,
+      "learning_rate": 8e-05,
+      "loss": 1.6386,
+      "step": 4296
+    },
+    {
+      "epoch": 1.0442230591665653,
+      "grad_norm": 0.4608069956302643,
+      "learning_rate": 8e-05,
+      "loss": 1.5377,
+      "step": 4297
+    },
+    {
+      "epoch": 1.04446604300814,
+      "grad_norm": 0.4574331045150757,
+      "learning_rate": 8e-05,
+      "loss": 1.4777,
+      "step": 4298
+    },
+    {
+      "epoch": 1.0447090268497146,
+      "grad_norm": 0.4482179284095764,
+      "learning_rate": 8e-05,
+      "loss": 1.3707,
+      "step": 4299
+    },
+    {
+      "epoch": 1.044952010691289,
+      "grad_norm": 0.4467381238937378,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0451949945328636,
+      "grad_norm": 0.4570399820804596,
+      "learning_rate": 8e-05,
+      "loss": 1.5234,
+      "step": 4301
+    },
+    {
+      "epoch": 1.045437978374438,
+      "grad_norm": 0.47596460580825806,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 4302
+    },
+    {
+      "epoch": 1.0456809622160126,
+      "grad_norm": 0.48790422081947327,
+      "learning_rate": 8e-05,
+      "loss": 1.5561,
+      "step": 4303
+    },
+    {
+      "epoch": 1.045923946057587,
+      "grad_norm": 0.46789875626564026,
+      "learning_rate": 8e-05,
+      "loss": 1.5122,
+      "step": 4304
+    },
+    {
+      "epoch": 1.0461669298991616,
+      "grad_norm": 0.5032777190208435,
+      "learning_rate": 8e-05,
+      "loss": 1.5137,
+      "step": 4305
+    },
+    {
+      "epoch": 1.0464099137407363,
+      "grad_norm": 0.49813881516456604,
+      "learning_rate": 8e-05,
+      "loss": 1.5339,
+      "step": 4306
+    },
+    {
+      "epoch": 1.0466528975823108,
+      "grad_norm": 0.545244574546814,
+      "learning_rate": 8e-05,
+      "loss": 1.6089,
+      "step": 4307
+    },
+    {
+      "epoch": 1.0468958814238853,
+      "grad_norm": 0.47614142298698425,
+      "learning_rate": 8e-05,
+      "loss": 1.4706,
+      "step": 4308
+    },
+    {
+      "epoch": 1.0471388652654599,
+      "grad_norm": 0.4766044616699219,
+      "learning_rate": 8e-05,
+      "loss": 1.5144,
+      "step": 4309
+    },
+    {
+      "epoch": 1.0473818491070344,
+      "grad_norm": 0.4749530553817749,
+      "learning_rate": 8e-05,
+      "loss": 1.5608,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0476248329486089,
+      "grad_norm": 0.46956250071525574,
+      "learning_rate": 8e-05,
+      "loss": 1.5053,
+      "step": 4311
+    },
+    {
+      "epoch": 1.0478678167901834,
+      "grad_norm": 0.5111247897148132,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 4312
+    },
+    {
+      "epoch": 1.048110800631758,
+      "grad_norm": 0.48100975155830383,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 4313
+    },
+    {
+      "epoch": 1.0483537844733326,
+      "grad_norm": 0.44856706261634827,
+      "learning_rate": 8e-05,
+      "loss": 1.5182,
+      "step": 4314
+    },
+    {
+      "epoch": 1.0485967683149071,
+      "grad_norm": 0.4658728241920471,
+      "learning_rate": 8e-05,
+      "loss": 1.3753,
+      "step": 4315
+    },
+    {
+      "epoch": 1.0488397521564816,
+      "grad_norm": 0.4779205620288849,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 4316
+    },
+    {
+      "epoch": 1.0490827359980561,
+      "grad_norm": 0.4791796803474426,
+      "learning_rate": 8e-05,
+      "loss": 1.3884,
+      "step": 4317
+    },
+    {
+      "epoch": 1.0493257198396306,
+      "grad_norm": 0.4321656823158264,
+      "learning_rate": 8e-05,
+      "loss": 1.247,
+      "step": 4318
+    },
+    {
+      "epoch": 1.0495687036812051,
+      "grad_norm": 0.44187963008880615,
+      "learning_rate": 8e-05,
+      "loss": 1.5298,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0498116875227796,
+      "grad_norm": 0.48076295852661133,
+      "learning_rate": 8e-05,
+      "loss": 1.6546,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0500546713643544,
+      "grad_norm": 0.4582403004169464,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 4321
+    },
+    {
+      "epoch": 1.0502976552059289,
+      "grad_norm": 0.48594361543655396,
+      "learning_rate": 8e-05,
+      "loss": 1.5171,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0505406390475034,
+      "grad_norm": 0.4511914849281311,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 4323
+    },
+    {
+      "epoch": 1.050783622889078,
+      "grad_norm": 0.4856415390968323,
+      "learning_rate": 8e-05,
+      "loss": 1.5995,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0510266067306524,
+      "grad_norm": 0.5015833973884583,
+      "learning_rate": 8e-05,
+      "loss": 1.5888,
+      "step": 4325
+    },
+    {
+      "epoch": 1.051269590572227,
+      "grad_norm": 0.47599563002586365,
+      "learning_rate": 8e-05,
+      "loss": 1.5856,
+      "step": 4326
+    },
+    {
+      "epoch": 1.0515125744138014,
+      "grad_norm": 0.4673370122909546,
+      "learning_rate": 8e-05,
+      "loss": 1.4406,
+      "step": 4327
+    },
+    {
+      "epoch": 1.051755558255376,
+      "grad_norm": 0.5171816945075989,
+      "learning_rate": 8e-05,
+      "loss": 1.6824,
+      "step": 4328
+    },
+    {
+      "epoch": 1.0519985420969507,
+      "grad_norm": 0.46786558628082275,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 4329
+    },
+    {
+      "epoch": 1.0522415259385252,
+      "grad_norm": 0.46182796359062195,
+      "learning_rate": 8e-05,
+      "loss": 1.5522,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0524845097800997,
+      "grad_norm": 0.50552898645401,
+      "learning_rate": 8e-05,
+      "loss": 1.6889,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0527274936216742,
+      "grad_norm": 0.5110907554626465,
+      "learning_rate": 8e-05,
+      "loss": 1.6351,
+      "step": 4332
+    },
+    {
+      "epoch": 1.0529704774632487,
+      "grad_norm": 0.4719978868961334,
+      "learning_rate": 8e-05,
+      "loss": 1.5663,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0532134613048232,
+      "grad_norm": 0.47473645210266113,
+      "learning_rate": 8e-05,
+      "loss": 1.5134,
+      "step": 4334
+    },
+    {
+      "epoch": 1.0534564451463977,
+      "grad_norm": 0.5138607025146484,
+      "learning_rate": 8e-05,
+      "loss": 1.5837,
+      "step": 4335
+    },
+    {
+      "epoch": 1.0536994289879722,
+      "grad_norm": 0.493149071931839,
+      "learning_rate": 8e-05,
+      "loss": 1.5098,
+      "step": 4336
+    },
+    {
+      "epoch": 1.053942412829547,
+      "grad_norm": 0.44648727774620056,
+      "learning_rate": 8e-05,
+      "loss": 1.4822,
+      "step": 4337
+    },
+    {
+      "epoch": 1.0541853966711214,
+      "grad_norm": 0.5008643269538879,
+      "learning_rate": 8e-05,
+      "loss": 1.5281,
+      "step": 4338
+    },
+    {
+      "epoch": 1.054428380512696,
+      "grad_norm": 0.5520813465118408,
+      "learning_rate": 8e-05,
+      "loss": 1.5225,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0546713643542704,
+      "grad_norm": 0.4723474085330963,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 4340
+    },
+    {
+      "epoch": 1.054914348195845,
+      "grad_norm": 0.46508780121803284,
+      "learning_rate": 8e-05,
+      "loss": 1.5321,
+      "step": 4341
+    },
+    {
+      "epoch": 1.0551573320374195,
+      "grad_norm": 0.4706136882305145,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 4342
+    },
+    {
+      "epoch": 1.055400315878994,
+      "grad_norm": 0.4866950809955597,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 4343
+    },
+    {
+      "epoch": 1.0556432997205687,
+      "grad_norm": 0.48028650879859924,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0558862835621432,
+      "grad_norm": 0.4771740436553955,
+      "learning_rate": 8e-05,
+      "loss": 1.4762,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0561292674037177,
+      "grad_norm": 0.527273416519165,
+      "learning_rate": 8e-05,
+      "loss": 1.66,
+      "step": 4346
+    },
+    {
+      "epoch": 1.0563722512452922,
+      "grad_norm": 0.5095369219779968,
+      "learning_rate": 8e-05,
+      "loss": 1.58,
+      "step": 4347
+    },
+    {
+      "epoch": 1.0566152350868667,
+      "grad_norm": 0.48933225870132446,
+      "learning_rate": 8e-05,
+      "loss": 1.6219,
+      "step": 4348
+    },
+    {
+      "epoch": 1.0568582189284412,
+      "grad_norm": 0.4847005009651184,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 4349
+    },
+    {
+      "epoch": 1.0571012027700157,
+      "grad_norm": 0.4733482599258423,
+      "learning_rate": 8e-05,
+      "loss": 1.4132,
+      "step": 4350
+    },
+    {
+      "epoch": 1.0573441866115902,
+      "grad_norm": 0.4872048795223236,
+      "learning_rate": 8e-05,
+      "loss": 1.4715,
+      "step": 4351
+    },
+    {
+      "epoch": 1.057587170453165,
+      "grad_norm": 0.4924294352531433,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 4352
+    },
+    {
+      "epoch": 1.0578301542947395,
+      "grad_norm": 0.47971946001052856,
+      "learning_rate": 8e-05,
+      "loss": 1.5573,
+      "step": 4353
+    },
+    {
+      "epoch": 1.058073138136314,
+      "grad_norm": 0.4633120894432068,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 4354
+    },
+    {
+      "epoch": 1.0583161219778885,
+      "grad_norm": 0.4915049374103546,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 4355
+    },
+    {
+      "epoch": 1.058559105819463,
+      "grad_norm": 0.47829174995422363,
+      "learning_rate": 8e-05,
+      "loss": 1.593,
+      "step": 4356
+    },
+    {
+      "epoch": 1.0588020896610375,
+      "grad_norm": 0.45118874311447144,
+      "learning_rate": 8e-05,
+      "loss": 1.457,
+      "step": 4357
+    },
+    {
+      "epoch": 1.059045073502612,
+      "grad_norm": 0.495598703622818,
+      "learning_rate": 8e-05,
+      "loss": 1.474,
+      "step": 4358
+    },
+    {
+      "epoch": 1.0592880573441865,
+      "grad_norm": 0.4829000234603882,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 4359
+    },
+    {
+      "epoch": 1.0595310411857612,
+      "grad_norm": 0.460509717464447,
+      "learning_rate": 8e-05,
+      "loss": 1.5296,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0597740250273358,
+      "grad_norm": 0.4830223619937897,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 4361
+    },
+    {
+      "epoch": 1.0600170088689103,
+      "grad_norm": 0.4810490608215332,
+      "learning_rate": 8e-05,
+      "loss": 1.5874,
+      "step": 4362
+    },
+    {
+      "epoch": 1.0602599927104848,
+      "grad_norm": 0.46496760845184326,
+      "learning_rate": 8e-05,
+      "loss": 1.4858,
+      "step": 4363
+    },
+    {
+      "epoch": 1.0605029765520593,
+      "grad_norm": 0.46511009335517883,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 4364
+    },
+    {
+      "epoch": 1.0607459603936338,
+      "grad_norm": 0.4688032865524292,
+      "learning_rate": 8e-05,
+      "loss": 1.4744,
+      "step": 4365
+    },
+    {
+      "epoch": 1.0609889442352083,
+      "grad_norm": 0.4910958409309387,
+      "learning_rate": 8e-05,
+      "loss": 1.5627,
+      "step": 4366
+    },
+    {
+      "epoch": 1.0612319280767828,
+      "grad_norm": 0.509813666343689,
+      "learning_rate": 8e-05,
+      "loss": 1.535,
+      "step": 4367
+    },
+    {
+      "epoch": 1.0614749119183575,
+      "grad_norm": 0.463809609413147,
+      "learning_rate": 8e-05,
+      "loss": 1.4222,
+      "step": 4368
+    },
+    {
+      "epoch": 1.061717895759932,
+      "grad_norm": 0.48283740878105164,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 4369
+    },
+    {
+      "epoch": 1.0619608796015065,
+      "grad_norm": 0.49186843633651733,
+      "learning_rate": 8e-05,
+      "loss": 1.4917,
+      "step": 4370
+    },
+    {
+      "epoch": 1.062203863443081,
+      "grad_norm": 0.43694913387298584,
+      "learning_rate": 8e-05,
+      "loss": 1.4355,
+      "step": 4371
+    },
+    {
+      "epoch": 1.0624468472846555,
+      "grad_norm": 0.47494521737098694,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 4372
+    },
+    {
+      "epoch": 1.06268983112623,
+      "grad_norm": 0.4725058674812317,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 4373
+    },
+    {
+      "epoch": 1.0629328149678046,
+      "grad_norm": 0.46237069368362427,
+      "learning_rate": 8e-05,
+      "loss": 1.3902,
+      "step": 4374
+    },
+    {
+      "epoch": 1.0631757988093793,
+      "grad_norm": 0.49597036838531494,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 4375
+    },
+    {
+      "epoch": 1.0634187826509538,
+      "grad_norm": 0.471748024225235,
+      "learning_rate": 8e-05,
+      "loss": 1.4223,
+      "step": 4376
+    },
+    {
+      "epoch": 1.0636617664925283,
+      "grad_norm": 0.47769811749458313,
+      "learning_rate": 8e-05,
+      "loss": 1.4648,
+      "step": 4377
+    },
+    {
+      "epoch": 1.0639047503341028,
+      "grad_norm": 0.5324870347976685,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 4378
+    },
+    {
+      "epoch": 1.0641477341756773,
+      "grad_norm": 0.5277772545814514,
+      "learning_rate": 8e-05,
+      "loss": 1.4978,
+      "step": 4379
+    },
+    {
+      "epoch": 1.0643907180172518,
+      "grad_norm": 0.49268168210983276,
+      "learning_rate": 8e-05,
+      "loss": 1.689,
+      "step": 4380
+    },
+    {
+      "epoch": 1.0646337018588263,
+      "grad_norm": 0.4566490352153778,
+      "learning_rate": 8e-05,
+      "loss": 1.4397,
+      "step": 4381
+    },
+    {
+      "epoch": 1.0648766857004008,
+      "grad_norm": 0.487904816865921,
+      "learning_rate": 8e-05,
+      "loss": 1.7033,
+      "step": 4382
+    },
+    {
+      "epoch": 1.0651196695419753,
+      "grad_norm": 0.515239417552948,
+      "learning_rate": 8e-05,
+      "loss": 1.6504,
+      "step": 4383
+    },
+    {
+      "epoch": 1.06536265338355,
+      "grad_norm": 0.4975634813308716,
+      "learning_rate": 8e-05,
+      "loss": 1.5569,
+      "step": 4384
+    },
+    {
+      "epoch": 1.0656056372251246,
+      "grad_norm": 0.5210979580879211,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 4385
+    },
+    {
+      "epoch": 1.065848621066699,
+      "grad_norm": 0.45190316438674927,
+      "learning_rate": 8e-05,
+      "loss": 1.4805,
+      "step": 4386
+    },
+    {
+      "epoch": 1.0660916049082736,
+      "grad_norm": 0.47176656126976013,
+      "learning_rate": 8e-05,
+      "loss": 1.5591,
+      "step": 4387
+    },
+    {
+      "epoch": 1.066334588749848,
+      "grad_norm": 0.49992209672927856,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 4388
+    },
+    {
+      "epoch": 1.0665775725914226,
+      "grad_norm": 0.4631107747554779,
+      "learning_rate": 8e-05,
+      "loss": 1.5754,
+      "step": 4389
+    },
+    {
+      "epoch": 1.0668205564329971,
+      "grad_norm": 0.4818519055843353,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 4390
+    },
+    {
+      "epoch": 1.0670635402745718,
+      "grad_norm": 0.46295106410980225,
+      "learning_rate": 8e-05,
+      "loss": 1.4639,
+      "step": 4391
+    },
+    {
+      "epoch": 1.0673065241161463,
+      "grad_norm": 0.48188602924346924,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 4392
+    },
+    {
+      "epoch": 1.0675495079577209,
+      "grad_norm": 0.4762363135814667,
+      "learning_rate": 8e-05,
+      "loss": 1.5416,
+      "step": 4393
+    },
+    {
+      "epoch": 1.0677924917992954,
+      "grad_norm": 0.4638352692127228,
+      "learning_rate": 8e-05,
+      "loss": 1.5432,
+      "step": 4394
+    },
+    {
+      "epoch": 1.0680354756408699,
+      "grad_norm": 0.49502280354499817,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 4395
+    },
+    {
+      "epoch": 1.0682784594824444,
+      "grad_norm": 0.4757213592529297,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 4396
+    },
+    {
+      "epoch": 1.0685214433240189,
+      "grad_norm": 0.4880535304546356,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 4397
+    },
+    {
+      "epoch": 1.0687644271655934,
+      "grad_norm": 0.4845835268497467,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 4398
+    },
+    {
+      "epoch": 1.0690074110071681,
+      "grad_norm": 0.4648076295852661,
+      "learning_rate": 8e-05,
+      "loss": 1.499,
+      "step": 4399
+    },
+    {
+      "epoch": 1.0692503948487426,
+      "grad_norm": 0.46832144260406494,
+      "learning_rate": 8e-05,
+      "loss": 1.5077,
+      "step": 4400
+    },
+    {
+      "epoch": 1.0694933786903171,
+      "grad_norm": 0.4800246059894562,
+      "learning_rate": 8e-05,
+      "loss": 1.5496,
+      "step": 4401
+    },
+    {
+      "epoch": 1.0697363625318916,
+      "grad_norm": 0.5600773096084595,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 4402
+    },
+    {
+      "epoch": 1.0699793463734661,
+      "grad_norm": 0.46718907356262207,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4403
+    },
+    {
+      "epoch": 1.0702223302150407,
+      "grad_norm": 0.4852052628993988,
+      "learning_rate": 8e-05,
+      "loss": 1.5128,
+      "step": 4404
+    },
+    {
+      "epoch": 1.0704653140566152,
+      "grad_norm": 0.48645520210266113,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 4405
+    },
+    {
+      "epoch": 1.0707082978981899,
+      "grad_norm": 0.49028751254081726,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 4406
+    },
+    {
+      "epoch": 1.0709512817397644,
+      "grad_norm": 0.47034451365470886,
+      "learning_rate": 8e-05,
+      "loss": 1.5148,
+      "step": 4407
+    },
+    {
+      "epoch": 1.071194265581339,
+      "grad_norm": 0.4791884124279022,
+      "learning_rate": 8e-05,
+      "loss": 1.4788,
+      "step": 4408
+    },
+    {
+      "epoch": 1.0714372494229134,
+      "grad_norm": 0.47847259044647217,
+      "learning_rate": 8e-05,
+      "loss": 1.5036,
+      "step": 4409
+    },
+    {
+      "epoch": 1.071680233264488,
+      "grad_norm": 0.47419115900993347,
+      "learning_rate": 8e-05,
+      "loss": 1.556,
+      "step": 4410
+    },
+    {
+      "epoch": 1.0719232171060624,
+      "grad_norm": 0.47514063119888306,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 4411
+    },
+    {
+      "epoch": 1.072166200947637,
+      "grad_norm": 0.47650012373924255,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 4412
+    },
+    {
+      "epoch": 1.0724091847892114,
+      "grad_norm": 0.4925329089164734,
+      "learning_rate": 8e-05,
+      "loss": 1.5493,
+      "step": 4413
+    },
+    {
+      "epoch": 1.072652168630786,
+      "grad_norm": 0.45949816703796387,
+      "learning_rate": 8e-05,
+      "loss": 1.4675,
+      "step": 4414
+    },
+    {
+      "epoch": 1.0728951524723607,
+      "grad_norm": 0.488942414522171,
+      "learning_rate": 8e-05,
+      "loss": 1.5727,
+      "step": 4415
+    },
+    {
+      "epoch": 1.0731381363139352,
+      "grad_norm": 0.48125630617141724,
+      "learning_rate": 8e-05,
+      "loss": 1.4637,
+      "step": 4416
+    },
+    {
+      "epoch": 1.0733811201555097,
+      "grad_norm": 0.5078297853469849,
+      "learning_rate": 8e-05,
+      "loss": 1.5801,
+      "step": 4417
+    },
+    {
+      "epoch": 1.0736241039970842,
+      "grad_norm": 0.4568662643432617,
+      "learning_rate": 8e-05,
+      "loss": 1.4168,
+      "step": 4418
+    },
+    {
+      "epoch": 1.0738670878386587,
+      "grad_norm": 0.4804803431034088,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 4419
+    },
+    {
+      "epoch": 1.0741100716802332,
+      "grad_norm": 0.4950558841228485,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 4420
+    },
+    {
+      "epoch": 1.0743530555218077,
+      "grad_norm": 0.49187353253364563,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 4421
+    },
+    {
+      "epoch": 1.0745960393633824,
+      "grad_norm": 0.488584965467453,
+      "learning_rate": 8e-05,
+      "loss": 1.623,
+      "step": 4422
+    },
+    {
+      "epoch": 1.074839023204957,
+      "grad_norm": 0.46431395411491394,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 4423
+    },
+    {
+      "epoch": 1.0750820070465315,
+      "grad_norm": 0.4693463146686554,
+      "learning_rate": 8e-05,
+      "loss": 1.5113,
+      "step": 4424
+    },
+    {
+      "epoch": 1.075324990888106,
+      "grad_norm": 0.498616486787796,
+      "learning_rate": 8e-05,
+      "loss": 1.6756,
+      "step": 4425
+    },
+    {
+      "epoch": 1.0755679747296805,
+      "grad_norm": 0.5458131432533264,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 4426
+    },
+    {
+      "epoch": 1.075810958571255,
+      "grad_norm": 0.4984363615512848,
+      "learning_rate": 8e-05,
+      "loss": 1.5537,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0760539424128295,
+      "grad_norm": 0.5386041402816772,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 4428
+    },
+    {
+      "epoch": 1.076296926254404,
+      "grad_norm": 0.4752993881702423,
+      "learning_rate": 8e-05,
+      "loss": 1.3911,
+      "step": 4429
+    },
+    {
+      "epoch": 1.0765399100959787,
+      "grad_norm": 0.5560598373413086,
+      "learning_rate": 8e-05,
+      "loss": 1.7479,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0767828939375532,
+      "grad_norm": 0.48651355504989624,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0770258777791277,
+      "grad_norm": 0.47595900297164917,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 4432
+    },
+    {
+      "epoch": 1.0772688616207022,
+      "grad_norm": 0.4925907850265503,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 4433
+    },
+    {
+      "epoch": 1.0775118454622767,
+      "grad_norm": 0.4877159595489502,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0777548293038512,
+      "grad_norm": 0.46722546219825745,
+      "learning_rate": 8e-05,
+      "loss": 1.4232,
+      "step": 4435
+    },
+    {
+      "epoch": 1.0779978131454258,
+      "grad_norm": 0.4926069676876068,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0782407969870003,
+      "grad_norm": 0.4899085760116577,
+      "learning_rate": 8e-05,
+      "loss": 1.4807,
+      "step": 4437
+    },
+    {
+      "epoch": 1.078483780828575,
+      "grad_norm": 0.4641192555427551,
+      "learning_rate": 8e-05,
+      "loss": 1.357,
+      "step": 4438
+    },
+    {
+      "epoch": 1.0787267646701495,
+      "grad_norm": 0.4559221863746643,
+      "learning_rate": 8e-05,
+      "loss": 1.3688,
+      "step": 4439
+    },
+    {
+      "epoch": 1.078969748511724,
+      "grad_norm": 0.4909510314464569,
+      "learning_rate": 8e-05,
+      "loss": 1.4455,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0792127323532985,
+      "grad_norm": 0.4498349726200104,
+      "learning_rate": 8e-05,
+      "loss": 1.4388,
+      "step": 4441
+    },
+    {
+      "epoch": 1.079455716194873,
+      "grad_norm": 0.4728998839855194,
+      "learning_rate": 8e-05,
+      "loss": 1.4769,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0796987000364475,
+      "grad_norm": 0.47789114713668823,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 4443
+    },
+    {
+      "epoch": 1.079941683878022,
+      "grad_norm": 0.4805726408958435,
+      "learning_rate": 8e-05,
+      "loss": 1.4045,
+      "step": 4444
+    },
+    {
+      "epoch": 1.0801846677195965,
+      "grad_norm": 0.5224361419677734,
+      "learning_rate": 8e-05,
+      "loss": 1.5644,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0804276515611713,
+      "grad_norm": 0.5380455851554871,
+      "learning_rate": 8e-05,
+      "loss": 1.5701,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0806706354027458,
+      "grad_norm": 0.47281455993652344,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 4447
+    },
+    {
+      "epoch": 1.0809136192443203,
+      "grad_norm": 0.49805375933647156,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 4448
+    },
+    {
+      "epoch": 1.0811566030858948,
+      "grad_norm": 0.4601954221725464,
+      "learning_rate": 8e-05,
+      "loss": 1.4289,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0813995869274693,
+      "grad_norm": 0.4668978750705719,
+      "learning_rate": 8e-05,
+      "loss": 1.4761,
+      "step": 4450
+    },
+    {
+      "epoch": 1.0816425707690438,
+      "grad_norm": 0.49863532185554504,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 4451
+    },
+    {
+      "epoch": 1.0818855546106183,
+      "grad_norm": 0.519084632396698,
+      "learning_rate": 8e-05,
+      "loss": 1.4202,
+      "step": 4452
+    },
+    {
+      "epoch": 1.082128538452193,
+      "grad_norm": 0.5036517977714539,
+      "learning_rate": 8e-05,
+      "loss": 1.5863,
+      "step": 4453
+    },
+    {
+      "epoch": 1.0823715222937675,
+      "grad_norm": 0.5504593253135681,
+      "learning_rate": 8e-05,
+      "loss": 1.5649,
+      "step": 4454
+    },
+    {
+      "epoch": 1.082614506135342,
+      "grad_norm": 0.4941757917404175,
+      "learning_rate": 8e-05,
+      "loss": 1.5163,
+      "step": 4455
+    },
+    {
+      "epoch": 1.0828574899769166,
+      "grad_norm": 0.5287080407142639,
+      "learning_rate": 8e-05,
+      "loss": 1.627,
+      "step": 4456
+    },
+    {
+      "epoch": 1.083100473818491,
+      "grad_norm": 0.4476391673088074,
+      "learning_rate": 8e-05,
+      "loss": 1.3939,
+      "step": 4457
+    },
+    {
+      "epoch": 1.0833434576600656,
+      "grad_norm": 0.4848242402076721,
+      "learning_rate": 8e-05,
+      "loss": 1.5046,
+      "step": 4458
+    },
+    {
+      "epoch": 1.08358644150164,
+      "grad_norm": 0.4548267126083374,
+      "learning_rate": 8e-05,
+      "loss": 1.4376,
+      "step": 4459
+    },
+    {
+      "epoch": 1.0838294253432146,
+      "grad_norm": 0.5123399496078491,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 4460
+    },
+    {
+      "epoch": 1.0840724091847893,
+      "grad_norm": 0.5134498476982117,
+      "learning_rate": 8e-05,
+      "loss": 1.6183,
+      "step": 4461
+    },
+    {
+      "epoch": 1.0843153930263638,
+      "grad_norm": 0.5067969560623169,
+      "learning_rate": 8e-05,
+      "loss": 1.6798,
+      "step": 4462
+    },
+    {
+      "epoch": 1.0845583768679383,
+      "grad_norm": 0.5087610483169556,
+      "learning_rate": 8e-05,
+      "loss": 1.7465,
+      "step": 4463
+    },
+    {
+      "epoch": 1.0848013607095128,
+      "grad_norm": 0.46484583616256714,
+      "learning_rate": 8e-05,
+      "loss": 1.5389,
+      "step": 4464
+    },
+    {
+      "epoch": 1.0850443445510873,
+      "grad_norm": 0.48163628578186035,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 4465
+    },
+    {
+      "epoch": 1.0852873283926618,
+      "grad_norm": 0.48081186413764954,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 4466
+    },
+    {
+      "epoch": 1.0855303122342363,
+      "grad_norm": 0.48678550124168396,
+      "learning_rate": 8e-05,
+      "loss": 1.5444,
+      "step": 4467
+    },
+    {
+      "epoch": 1.0857732960758109,
+      "grad_norm": 0.46936535835266113,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 4468
+    },
+    {
+      "epoch": 1.0860162799173856,
+      "grad_norm": 0.48902615904808044,
+      "learning_rate": 8e-05,
+      "loss": 1.4387,
+      "step": 4469
+    },
+    {
+      "epoch": 1.08625926375896,
+      "grad_norm": 0.482583612203598,
+      "learning_rate": 8e-05,
+      "loss": 1.4929,
+      "step": 4470
+    },
+    {
+      "epoch": 1.0865022476005346,
+      "grad_norm": 0.48313435912132263,
+      "learning_rate": 8e-05,
+      "loss": 1.6186,
+      "step": 4471
+    },
+    {
+      "epoch": 1.086745231442109,
+      "grad_norm": 0.4497871994972229,
+      "learning_rate": 8e-05,
+      "loss": 1.4113,
+      "step": 4472
+    },
+    {
+      "epoch": 1.0869882152836836,
+      "grad_norm": 0.5060304999351501,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 4473
+    },
+    {
+      "epoch": 1.0872311991252581,
+      "grad_norm": 0.5089784860610962,
+      "learning_rate": 8e-05,
+      "loss": 1.6301,
+      "step": 4474
+    },
+    {
+      "epoch": 1.0874741829668326,
+      "grad_norm": 0.4690403640270233,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 4475
+    },
+    {
+      "epoch": 1.0877171668084071,
+      "grad_norm": 0.47103267908096313,
+      "learning_rate": 8e-05,
+      "loss": 1.5793,
+      "step": 4476
+    },
+    {
+      "epoch": 1.0879601506499819,
+      "grad_norm": 0.4867348074913025,
+      "learning_rate": 8e-05,
+      "loss": 1.6146,
+      "step": 4477
+    },
+    {
+      "epoch": 1.0882031344915564,
+      "grad_norm": 0.48713359236717224,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 4478
+    },
+    {
+      "epoch": 1.0884461183331309,
+      "grad_norm": 0.5132150053977966,
+      "learning_rate": 8e-05,
+      "loss": 1.5353,
+      "step": 4479
+    },
+    {
+      "epoch": 1.0886891021747054,
+      "grad_norm": 0.4999169707298279,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 4480
+    },
+    {
+      "epoch": 1.0889320860162799,
+      "grad_norm": 0.47533559799194336,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 4481
+    },
+    {
+      "epoch": 1.0891750698578544,
+      "grad_norm": 0.5248066186904907,
+      "learning_rate": 8e-05,
+      "loss": 1.7323,
+      "step": 4482
+    },
+    {
+      "epoch": 1.089418053699429,
+      "grad_norm": 0.47552141547203064,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 4483
+    },
+    {
+      "epoch": 1.0896610375410036,
+      "grad_norm": 0.4716910719871521,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0899040213825781,
+      "grad_norm": 0.48882541060447693,
+      "learning_rate": 8e-05,
+      "loss": 1.4498,
+      "step": 4485
+    },
+    {
+      "epoch": 1.0901470052241526,
+      "grad_norm": 0.4709269404411316,
+      "learning_rate": 8e-05,
+      "loss": 1.5157,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0903899890657271,
+      "grad_norm": 0.49522435665130615,
+      "learning_rate": 8e-05,
+      "loss": 1.7509,
+      "step": 4487
+    },
+    {
+      "epoch": 1.0906329729073017,
+      "grad_norm": 0.49233612418174744,
+      "learning_rate": 8e-05,
+      "loss": 1.5535,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0908759567488762,
+      "grad_norm": 0.47601085901260376,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0911189405904507,
+      "grad_norm": 0.5018871426582336,
+      "learning_rate": 8e-05,
+      "loss": 1.4558,
+      "step": 4490
+    },
+    {
+      "epoch": 1.0913619244320252,
+      "grad_norm": 0.5203490257263184,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 4491
+    },
+    {
+      "epoch": 1.0916049082736,
+      "grad_norm": 0.5150461792945862,
+      "learning_rate": 8e-05,
+      "loss": 1.7546,
+      "step": 4492
+    },
+    {
+      "epoch": 1.0918478921151744,
+      "grad_norm": 0.47152963280677795,
+      "learning_rate": 8e-05,
+      "loss": 1.4318,
+      "step": 4493
+    },
+    {
+      "epoch": 1.092090875956749,
+      "grad_norm": 0.49111270904541016,
+      "learning_rate": 8e-05,
+      "loss": 1.66,
+      "step": 4494
+    },
+    {
+      "epoch": 1.0923338597983234,
+      "grad_norm": 0.45349034667015076,
+      "learning_rate": 8e-05,
+      "loss": 1.5311,
+      "step": 4495
+    },
+    {
+      "epoch": 1.092576843639898,
+      "grad_norm": 0.4733912944793701,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 4496
+    },
+    {
+      "epoch": 1.0928198274814724,
+      "grad_norm": 0.47600284218788147,
+      "learning_rate": 8e-05,
+      "loss": 1.4851,
+      "step": 4497
+    },
+    {
+      "epoch": 1.093062811323047,
+      "grad_norm": 0.49591273069381714,
+      "learning_rate": 8e-05,
+      "loss": 1.6283,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0933057951646215,
+      "grad_norm": 0.4981767237186432,
+      "learning_rate": 8e-05,
+      "loss": 1.5471,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0935487790061962,
+      "grad_norm": 0.48707085847854614,
+      "learning_rate": 8e-05,
+      "loss": 1.511,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0937917628477707,
+      "grad_norm": 0.5000172853469849,
+      "learning_rate": 8e-05,
+      "loss": 1.4935,
+      "step": 4501
+    },
+    {
+      "epoch": 1.0940347466893452,
+      "grad_norm": 0.5009807348251343,
+      "learning_rate": 8e-05,
+      "loss": 1.6049,
+      "step": 4502
+    },
+    {
+      "epoch": 1.0942777305309197,
+      "grad_norm": 0.4950929284095764,
+      "learning_rate": 8e-05,
+      "loss": 1.5867,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0945207143724942,
+      "grad_norm": 0.502869188785553,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 4504
+    },
+    {
+      "epoch": 1.0947636982140687,
+      "grad_norm": 0.47175437211990356,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0950066820556432,
+      "grad_norm": 0.5055964589118958,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 4506
+    },
+    {
+      "epoch": 1.0952496658972177,
+      "grad_norm": 0.48494982719421387,
+      "learning_rate": 8e-05,
+      "loss": 1.5012,
+      "step": 4507
+    },
+    {
+      "epoch": 1.0954926497387925,
+      "grad_norm": 0.49093398451805115,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 4508
+    },
+    {
+      "epoch": 1.095735633580367,
+      "grad_norm": 0.5152201652526855,
+      "learning_rate": 8e-05,
+      "loss": 1.642,
+      "step": 4509
+    },
+    {
+      "epoch": 1.0959786174219415,
+      "grad_norm": 0.47880688309669495,
+      "learning_rate": 8e-05,
+      "loss": 1.4671,
+      "step": 4510
+    },
+    {
+      "epoch": 1.096221601263516,
+      "grad_norm": 0.461619108915329,
+      "learning_rate": 8e-05,
+      "loss": 1.4682,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0964645851050905,
+      "grad_norm": 0.4974827170372009,
+      "learning_rate": 8e-05,
+      "loss": 1.7512,
+      "step": 4512
+    },
+    {
+      "epoch": 1.096707568946665,
+      "grad_norm": 0.49793532490730286,
+      "learning_rate": 8e-05,
+      "loss": 1.4524,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0969505527882395,
+      "grad_norm": 0.48256006836891174,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0971935366298142,
+      "grad_norm": 0.5167440176010132,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 4515
+    },
+    {
+      "epoch": 1.0974365204713887,
+      "grad_norm": 0.49837926030158997,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 4516
+    },
+    {
+      "epoch": 1.0976795043129632,
+      "grad_norm": 0.4728340804576874,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 4517
+    },
+    {
+      "epoch": 1.0979224881545377,
+      "grad_norm": 0.4873230755329132,
+      "learning_rate": 8e-05,
+      "loss": 1.5871,
+      "step": 4518
+    },
+    {
+      "epoch": 1.0981654719961123,
+      "grad_norm": 0.4752941131591797,
+      "learning_rate": 8e-05,
+      "loss": 1.4264,
+      "step": 4519
+    },
+    {
+      "epoch": 1.0984084558376868,
+      "grad_norm": 0.5011289715766907,
+      "learning_rate": 8e-05,
+      "loss": 1.557,
+      "step": 4520
+    },
+    {
+      "epoch": 1.0986514396792613,
+      "grad_norm": 0.46568232774734497,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 4521
+    },
+    {
+      "epoch": 1.0988944235208358,
+      "grad_norm": 0.5012978911399841,
+      "learning_rate": 8e-05,
+      "loss": 1.4441,
+      "step": 4522
+    },
+    {
+      "epoch": 1.0991374073624105,
+      "grad_norm": 0.486845463514328,
+      "learning_rate": 8e-05,
+      "loss": 1.5015,
+      "step": 4523
+    },
+    {
+      "epoch": 1.099380391203985,
+      "grad_norm": 0.4983042776584625,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 4524
+    },
+    {
+      "epoch": 1.0996233750455595,
+      "grad_norm": 0.488980233669281,
+      "learning_rate": 8e-05,
+      "loss": 1.5244,
+      "step": 4525
+    },
+    {
+      "epoch": 1.099866358887134,
+      "grad_norm": 0.5147802233695984,
+      "learning_rate": 8e-05,
+      "loss": 1.5413,
+      "step": 4526
+    },
+    {
+      "epoch": 1.1001093427287085,
+      "grad_norm": 0.49820229411125183,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 4527
+    },
+    {
+      "epoch": 1.100352326570283,
+      "grad_norm": 0.46373334527015686,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 4528
+    },
+    {
+      "epoch": 1.1005953104118575,
+      "grad_norm": 0.49564653635025024,
+      "learning_rate": 8e-05,
+      "loss": 1.5434,
+      "step": 4529
+    },
+    {
+      "epoch": 1.100838294253432,
+      "grad_norm": 0.5321706533432007,
+      "learning_rate": 8e-05,
+      "loss": 1.4043,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1010812780950068,
+      "grad_norm": 0.4796103239059448,
+      "learning_rate": 8e-05,
+      "loss": 1.4843,
+      "step": 4531
+    },
+    {
+      "epoch": 1.1013242619365813,
+      "grad_norm": 0.4949493408203125,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 4532
+    },
+    {
+      "epoch": 1.1015672457781558,
+      "grad_norm": 0.529646635055542,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 4533
+    },
+    {
+      "epoch": 1.1018102296197303,
+      "grad_norm": 0.49438926577568054,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 4534
+    },
+    {
+      "epoch": 1.1020532134613048,
+      "grad_norm": 0.49114930629730225,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 4535
+    },
+    {
+      "epoch": 1.1022961973028793,
+      "grad_norm": 0.5116463303565979,
+      "learning_rate": 8e-05,
+      "loss": 1.5122,
+      "step": 4536
+    },
+    {
+      "epoch": 1.1025391811444538,
+      "grad_norm": 0.4837435781955719,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 4537
+    },
+    {
+      "epoch": 1.1027821649860283,
+      "grad_norm": 0.4761262536048889,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 4538
+    },
+    {
+      "epoch": 1.103025148827603,
+      "grad_norm": 0.4860967993736267,
+      "learning_rate": 8e-05,
+      "loss": 1.4643,
+      "step": 4539
+    },
+    {
+      "epoch": 1.1032681326691776,
+      "grad_norm": 0.49716654419898987,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 4540
+    },
+    {
+      "epoch": 1.103511116510752,
+      "grad_norm": 0.4729624390602112,
+      "learning_rate": 8e-05,
+      "loss": 1.5015,
+      "step": 4541
+    },
+    {
+      "epoch": 1.1037541003523266,
+      "grad_norm": 0.4831719398498535,
+      "learning_rate": 8e-05,
+      "loss": 1.4918,
+      "step": 4542
+    },
+    {
+      "epoch": 1.103997084193901,
+      "grad_norm": 0.49242299795150757,
+      "learning_rate": 8e-05,
+      "loss": 1.6002,
+      "step": 4543
+    },
+    {
+      "epoch": 1.1042400680354756,
+      "grad_norm": 0.4934948682785034,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 4544
+    },
+    {
+      "epoch": 1.10448305187705,
+      "grad_norm": 0.4512420892715454,
+      "learning_rate": 8e-05,
+      "loss": 1.3936,
+      "step": 4545
+    },
+    {
+      "epoch": 1.1047260357186248,
+      "grad_norm": 0.4724448323249817,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 4546
+    },
+    {
+      "epoch": 1.1049690195601993,
+      "grad_norm": 0.5331870317459106,
+      "learning_rate": 8e-05,
+      "loss": 1.7108,
+      "step": 4547
+    },
+    {
+      "epoch": 1.1052120034017738,
+      "grad_norm": 0.4867265224456787,
+      "learning_rate": 8e-05,
+      "loss": 1.4867,
+      "step": 4548
+    },
+    {
+      "epoch": 1.1054549872433483,
+      "grad_norm": 0.4752415418624878,
+      "learning_rate": 8e-05,
+      "loss": 1.4557,
+      "step": 4549
+    },
+    {
+      "epoch": 1.1056979710849228,
+      "grad_norm": 0.5043330192565918,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1059409549264974,
+      "grad_norm": 0.49358829855918884,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 4551
+    },
+    {
+      "epoch": 1.1061839387680719,
+      "grad_norm": 0.4915328919887543,
+      "learning_rate": 8e-05,
+      "loss": 1.6624,
+      "step": 4552
+    },
+    {
+      "epoch": 1.1064269226096464,
+      "grad_norm": 0.4916263520717621,
+      "learning_rate": 8e-05,
+      "loss": 1.5191,
+      "step": 4553
+    },
+    {
+      "epoch": 1.106669906451221,
+      "grad_norm": 0.48078909516334534,
+      "learning_rate": 8e-05,
+      "loss": 1.7389,
+      "step": 4554
+    },
+    {
+      "epoch": 1.1069128902927956,
+      "grad_norm": 0.4844862222671509,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 4555
+    },
+    {
+      "epoch": 1.10715587413437,
+      "grad_norm": 0.4841192960739136,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 4556
+    },
+    {
+      "epoch": 1.1073988579759446,
+      "grad_norm": 0.4634784162044525,
+      "learning_rate": 8e-05,
+      "loss": 1.5504,
+      "step": 4557
+    },
+    {
+      "epoch": 1.1076418418175191,
+      "grad_norm": 0.4601224958896637,
+      "learning_rate": 8e-05,
+      "loss": 1.453,
+      "step": 4558
+    },
+    {
+      "epoch": 1.1078848256590936,
+      "grad_norm": 0.508124828338623,
+      "learning_rate": 8e-05,
+      "loss": 1.7133,
+      "step": 4559
+    },
+    {
+      "epoch": 1.1081278095006681,
+      "grad_norm": 0.4907205402851105,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1083707933422426,
+      "grad_norm": 0.4707714319229126,
+      "learning_rate": 8e-05,
+      "loss": 1.5149,
+      "step": 4561
+    },
+    {
+      "epoch": 1.1086137771838174,
+      "grad_norm": 0.49038493633270264,
+      "learning_rate": 8e-05,
+      "loss": 1.4785,
+      "step": 4562
+    },
+    {
+      "epoch": 1.1088567610253919,
+      "grad_norm": 0.47631633281707764,
+      "learning_rate": 8e-05,
+      "loss": 1.497,
+      "step": 4563
+    },
+    {
+      "epoch": 1.1090997448669664,
+      "grad_norm": 0.4876428544521332,
+      "learning_rate": 8e-05,
+      "loss": 1.5552,
+      "step": 4564
+    },
+    {
+      "epoch": 1.109342728708541,
+      "grad_norm": 0.4734289050102234,
+      "learning_rate": 8e-05,
+      "loss": 1.5294,
+      "step": 4565
+    },
+    {
+      "epoch": 1.1095857125501154,
+      "grad_norm": 0.4776040315628052,
+      "learning_rate": 8e-05,
+      "loss": 1.3865,
+      "step": 4566
+    },
+    {
+      "epoch": 1.10982869639169,
+      "grad_norm": 0.49185818433761597,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 4567
+    },
+    {
+      "epoch": 1.1100716802332644,
+      "grad_norm": 0.4999822676181793,
+      "learning_rate": 8e-05,
+      "loss": 1.5316,
+      "step": 4568
+    },
+    {
+      "epoch": 1.110314664074839,
+      "grad_norm": 0.5300837159156799,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 4569
+    },
+    {
+      "epoch": 1.1105576479164136,
+      "grad_norm": 0.5131652355194092,
+      "learning_rate": 8e-05,
+      "loss": 1.6467,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1108006317579882,
+      "grad_norm": 0.516015887260437,
+      "learning_rate": 8e-05,
+      "loss": 1.5594,
+      "step": 4571
+    },
+    {
+      "epoch": 1.1110436155995627,
+      "grad_norm": 0.4935130476951599,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 4572
+    },
+    {
+      "epoch": 1.1112865994411372,
+      "grad_norm": 0.48669302463531494,
+      "learning_rate": 8e-05,
+      "loss": 1.522,
+      "step": 4573
+    },
+    {
+      "epoch": 1.1115295832827117,
+      "grad_norm": 0.4898160398006439,
+      "learning_rate": 8e-05,
+      "loss": 1.5322,
+      "step": 4574
+    },
+    {
+      "epoch": 1.1117725671242862,
+      "grad_norm": 0.50649094581604,
+      "learning_rate": 8e-05,
+      "loss": 1.5424,
+      "step": 4575
+    },
+    {
+      "epoch": 1.1120155509658607,
+      "grad_norm": 0.5074909329414368,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 4576
+    },
+    {
+      "epoch": 1.1122585348074354,
+      "grad_norm": 0.47325778007507324,
+      "learning_rate": 8e-05,
+      "loss": 1.5479,
+      "step": 4577
+    },
+    {
+      "epoch": 1.11250151864901,
+      "grad_norm": 0.4595005512237549,
+      "learning_rate": 8e-05,
+      "loss": 1.4735,
+      "step": 4578
+    },
+    {
+      "epoch": 1.1127445024905844,
+      "grad_norm": 0.5001984238624573,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 4579
+    },
+    {
+      "epoch": 1.112987486332159,
+      "grad_norm": 0.4759065508842468,
+      "learning_rate": 8e-05,
+      "loss": 1.4116,
+      "step": 4580
+    },
+    {
+      "epoch": 1.1132304701737334,
+      "grad_norm": 0.48632240295410156,
+      "learning_rate": 8e-05,
+      "loss": 1.4756,
+      "step": 4581
+    },
+    {
+      "epoch": 1.113473454015308,
+      "grad_norm": 0.49018189311027527,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 4582
+    },
+    {
+      "epoch": 1.1137164378568825,
+      "grad_norm": 0.49138110876083374,
+      "learning_rate": 8e-05,
+      "loss": 1.4693,
+      "step": 4583
+    },
+    {
+      "epoch": 1.113959421698457,
+      "grad_norm": 0.5156691074371338,
+      "learning_rate": 8e-05,
+      "loss": 1.7581,
+      "step": 4584
+    },
+    {
+      "epoch": 1.1142024055400315,
+      "grad_norm": 0.5037038922309875,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 4585
+    },
+    {
+      "epoch": 1.1144453893816062,
+      "grad_norm": 0.48624318838119507,
+      "learning_rate": 8e-05,
+      "loss": 1.6369,
+      "step": 4586
+    },
+    {
+      "epoch": 1.1146883732231807,
+      "grad_norm": 0.47056812047958374,
+      "learning_rate": 8e-05,
+      "loss": 1.5755,
+      "step": 4587
+    },
+    {
+      "epoch": 1.1149313570647552,
+      "grad_norm": 0.4479929506778717,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 4588
+    },
+    {
+      "epoch": 1.1151743409063297,
+      "grad_norm": 0.5116077065467834,
+      "learning_rate": 8e-05,
+      "loss": 1.669,
+      "step": 4589
+    },
+    {
+      "epoch": 1.1154173247479042,
+      "grad_norm": 0.46570268273353577,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1156603085894787,
+      "grad_norm": 0.6732439994812012,
+      "learning_rate": 8e-05,
+      "loss": 1.6609,
+      "step": 4591
+    },
+    {
+      "epoch": 1.1159032924310532,
+      "grad_norm": 0.4752379357814789,
+      "learning_rate": 8e-05,
+      "loss": 1.5214,
+      "step": 4592
+    },
+    {
+      "epoch": 1.116146276272628,
+      "grad_norm": 0.5039642453193665,
+      "learning_rate": 8e-05,
+      "loss": 1.5707,
+      "step": 4593
+    },
+    {
+      "epoch": 1.1163892601142025,
+      "grad_norm": 0.4722294509410858,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 4594
+    },
+    {
+      "epoch": 1.116632243955777,
+      "grad_norm": 0.5270997285842896,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 4595
+    },
+    {
+      "epoch": 1.1168752277973515,
+      "grad_norm": 0.4860701858997345,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 4596
+    },
+    {
+      "epoch": 1.117118211638926,
+      "grad_norm": 0.4902597963809967,
+      "learning_rate": 8e-05,
+      "loss": 1.6449,
+      "step": 4597
+    },
+    {
+      "epoch": 1.1173611954805005,
+      "grad_norm": 0.4660976529121399,
+      "learning_rate": 8e-05,
+      "loss": 1.4471,
+      "step": 4598
+    },
+    {
+      "epoch": 1.117604179322075,
+      "grad_norm": 0.5060970783233643,
+      "learning_rate": 8e-05,
+      "loss": 1.6059,
+      "step": 4599
+    },
+    {
+      "epoch": 1.1178471631636495,
+      "grad_norm": 0.46083125472068787,
+      "learning_rate": 8e-05,
+      "loss": 1.4755,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1180901470052242,
+      "grad_norm": 0.4754008650779724,
+      "learning_rate": 8e-05,
+      "loss": 1.5176,
+      "step": 4601
+    },
+    {
+      "epoch": 1.1183331308467988,
+      "grad_norm": 0.4712495803833008,
+      "learning_rate": 8e-05,
+      "loss": 1.4506,
+      "step": 4602
+    },
+    {
+      "epoch": 1.1185761146883733,
+      "grad_norm": 0.46213024854660034,
+      "learning_rate": 8e-05,
+      "loss": 1.4938,
+      "step": 4603
+    },
+    {
+      "epoch": 1.1188190985299478,
+      "grad_norm": 0.5205399990081787,
+      "learning_rate": 8e-05,
+      "loss": 1.6813,
+      "step": 4604
+    },
+    {
+      "epoch": 1.1190620823715223,
+      "grad_norm": 0.47757062315940857,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 4605
+    },
+    {
+      "epoch": 1.1193050662130968,
+      "grad_norm": 0.4760383069515228,
+      "learning_rate": 8e-05,
+      "loss": 1.58,
+      "step": 4606
+    },
+    {
+      "epoch": 1.1195480500546713,
+      "grad_norm": 0.4909297525882721,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 4607
+    },
+    {
+      "epoch": 1.119791033896246,
+      "grad_norm": 0.50496906042099,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 4608
+    },
+    {
+      "epoch": 1.1200340177378205,
+      "grad_norm": 0.48355770111083984,
+      "learning_rate": 8e-05,
+      "loss": 1.5324,
+      "step": 4609
+    },
+    {
+      "epoch": 1.120277001579395,
+      "grad_norm": 0.4681239128112793,
+      "learning_rate": 8e-05,
+      "loss": 1.4916,
+      "step": 4610
+    },
+    {
+      "epoch": 1.1205199854209695,
+      "grad_norm": 0.499054878950119,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 4611
+    },
+    {
+      "epoch": 1.120762969262544,
+      "grad_norm": 0.504642128944397,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 4612
+    },
+    {
+      "epoch": 1.1210059531041185,
+      "grad_norm": 0.47570905089378357,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 4613
+    },
+    {
+      "epoch": 1.121248936945693,
+      "grad_norm": 0.4898286759853363,
+      "learning_rate": 8e-05,
+      "loss": 1.532,
+      "step": 4614
+    },
+    {
+      "epoch": 1.1214919207872676,
+      "grad_norm": 0.4712437391281128,
+      "learning_rate": 8e-05,
+      "loss": 1.5443,
+      "step": 4615
+    },
+    {
+      "epoch": 1.121734904628842,
+      "grad_norm": 0.4597233533859253,
+      "learning_rate": 8e-05,
+      "loss": 1.3605,
+      "step": 4616
+    },
+    {
+      "epoch": 1.1219778884704168,
+      "grad_norm": 0.4840410053730011,
+      "learning_rate": 8e-05,
+      "loss": 1.3845,
+      "step": 4617
+    },
+    {
+      "epoch": 1.1222208723119913,
+      "grad_norm": 0.4747578799724579,
+      "learning_rate": 8e-05,
+      "loss": 1.4681,
+      "step": 4618
+    },
+    {
+      "epoch": 1.1224638561535658,
+      "grad_norm": 0.48730745911598206,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 4619
+    },
+    {
+      "epoch": 1.1227068399951403,
+      "grad_norm": 0.49377936124801636,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1229498238367148,
+      "grad_norm": 0.47984492778778076,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 4621
+    },
+    {
+      "epoch": 1.1231928076782893,
+      "grad_norm": 0.505234956741333,
+      "learning_rate": 8e-05,
+      "loss": 1.5352,
+      "step": 4622
+    },
+    {
+      "epoch": 1.1234357915198638,
+      "grad_norm": 0.5102158784866333,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 4623
+    },
+    {
+      "epoch": 1.1236787753614386,
+      "grad_norm": 0.5234937071800232,
+      "learning_rate": 8e-05,
+      "loss": 1.7839,
+      "step": 4624
+    },
+    {
+      "epoch": 1.123921759203013,
+      "grad_norm": 0.48089370131492615,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 4625
+    },
+    {
+      "epoch": 1.1241647430445876,
+      "grad_norm": 0.4827926456928253,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 4626
+    },
+    {
+      "epoch": 1.124407726886162,
+      "grad_norm": 0.4984688460826874,
+      "learning_rate": 8e-05,
+      "loss": 1.4903,
+      "step": 4627
+    },
+    {
+      "epoch": 1.1246507107277366,
+      "grad_norm": 0.4781351685523987,
+      "learning_rate": 8e-05,
+      "loss": 1.4618,
+      "step": 4628
+    },
+    {
+      "epoch": 1.124893694569311,
+      "grad_norm": 0.4695113003253937,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 4629
+    },
+    {
+      "epoch": 1.1251366784108856,
+      "grad_norm": 0.474439412355423,
+      "learning_rate": 8e-05,
+      "loss": 1.5141,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1253796622524601,
+      "grad_norm": 0.4992530345916748,
+      "learning_rate": 8e-05,
+      "loss": 1.7756,
+      "step": 4631
+    },
+    {
+      "epoch": 1.1256226460940348,
+      "grad_norm": 0.48744991421699524,
+      "learning_rate": 8e-05,
+      "loss": 1.3358,
+      "step": 4632
+    },
+    {
+      "epoch": 1.1258656299356093,
+      "grad_norm": 0.463468462228775,
+      "learning_rate": 8e-05,
+      "loss": 1.5871,
+      "step": 4633
+    },
+    {
+      "epoch": 1.1261086137771839,
+      "grad_norm": 0.5052722096443176,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 4634
+    },
+    {
+      "epoch": 1.1263515976187584,
+      "grad_norm": 0.493437260389328,
+      "learning_rate": 8e-05,
+      "loss": 1.5739,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1265945814603329,
+      "grad_norm": 0.4789940416812897,
+      "learning_rate": 8e-05,
+      "loss": 1.5721,
+      "step": 4636
+    },
+    {
+      "epoch": 1.1268375653019074,
+      "grad_norm": 0.4616963565349579,
+      "learning_rate": 8e-05,
+      "loss": 1.4302,
+      "step": 4637
+    },
+    {
+      "epoch": 1.1270805491434819,
+      "grad_norm": 0.49398893117904663,
+      "learning_rate": 8e-05,
+      "loss": 1.501,
+      "step": 4638
+    },
+    {
+      "epoch": 1.1273235329850566,
+      "grad_norm": 0.5223192572593689,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 4639
+    },
+    {
+      "epoch": 1.1275665168266311,
+      "grad_norm": 0.48357000946998596,
+      "learning_rate": 8e-05,
+      "loss": 1.4637,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1278095006682056,
+      "grad_norm": 0.4698563814163208,
+      "learning_rate": 8e-05,
+      "loss": 1.4021,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1280524845097801,
+      "grad_norm": 0.5038408637046814,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 4642
+    },
+    {
+      "epoch": 1.1282954683513546,
+      "grad_norm": 0.47303229570388794,
+      "learning_rate": 8e-05,
+      "loss": 1.6553,
+      "step": 4643
+    },
+    {
+      "epoch": 1.1285384521929291,
+      "grad_norm": 0.490739107131958,
+      "learning_rate": 8e-05,
+      "loss": 1.5663,
+      "step": 4644
+    },
+    {
+      "epoch": 1.1287814360345036,
+      "grad_norm": 0.5153795480728149,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1290244198760782,
+      "grad_norm": 0.48572787642478943,
+      "learning_rate": 8e-05,
+      "loss": 1.4979,
+      "step": 4646
+    },
+    {
+      "epoch": 1.1292674037176527,
+      "grad_norm": 0.49244698882102966,
+      "learning_rate": 8e-05,
+      "loss": 1.5418,
+      "step": 4647
+    },
+    {
+      "epoch": 1.1295103875592274,
+      "grad_norm": 0.5016176700592041,
+      "learning_rate": 8e-05,
+      "loss": 1.5195,
+      "step": 4648
+    },
+    {
+      "epoch": 1.129753371400802,
+      "grad_norm": 0.5107845664024353,
+      "learning_rate": 8e-05,
+      "loss": 1.604,
+      "step": 4649
+    },
+    {
+      "epoch": 1.1299963552423764,
+      "grad_norm": 0.4940442740917206,
+      "learning_rate": 8e-05,
+      "loss": 1.4451,
+      "step": 4650
+    },
+    {
+      "epoch": 1.130239339083951,
+      "grad_norm": 0.5319364666938782,
+      "learning_rate": 8e-05,
+      "loss": 1.7937,
+      "step": 4651
+    },
+    {
+      "epoch": 1.1304823229255254,
+      "grad_norm": 0.47523507475852966,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 4652
+    },
+    {
+      "epoch": 1.1307253067671,
+      "grad_norm": 0.4858156144618988,
+      "learning_rate": 8e-05,
+      "loss": 1.4458,
+      "step": 4653
+    },
+    {
+      "epoch": 1.1309682906086744,
+      "grad_norm": 0.4847645163536072,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 4654
+    },
+    {
+      "epoch": 1.1312112744502492,
+      "grad_norm": 0.5123755931854248,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1314542582918237,
+      "grad_norm": 0.5030513405799866,
+      "learning_rate": 8e-05,
+      "loss": 1.5121,
+      "step": 4656
+    },
+    {
+      "epoch": 1.1316972421333982,
+      "grad_norm": 0.4845196008682251,
+      "learning_rate": 8e-05,
+      "loss": 1.5186,
+      "step": 4657
+    },
+    {
+      "epoch": 1.1319402259749727,
+      "grad_norm": 0.4732392132282257,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 4658
+    },
+    {
+      "epoch": 1.1321832098165472,
+      "grad_norm": 0.4974471628665924,
+      "learning_rate": 8e-05,
+      "loss": 1.5059,
+      "step": 4659
+    },
+    {
+      "epoch": 1.1324261936581217,
+      "grad_norm": 0.5079742074012756,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 4660
+    },
+    {
+      "epoch": 1.1326691774996962,
+      "grad_norm": 0.467381089925766,
+      "learning_rate": 8e-05,
+      "loss": 1.5502,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1329121613412707,
+      "grad_norm": 0.4882206618785858,
+      "learning_rate": 8e-05,
+      "loss": 1.5406,
+      "step": 4662
+    },
+    {
+      "epoch": 1.1331551451828454,
+      "grad_norm": 0.47715675830841064,
+      "learning_rate": 8e-05,
+      "loss": 1.4841,
+      "step": 4663
+    },
+    {
+      "epoch": 1.13339812902442,
+      "grad_norm": 0.48405516147613525,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 4664
+    },
+    {
+      "epoch": 1.1336411128659944,
+      "grad_norm": 0.4942147433757782,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 4665
+    },
+    {
+      "epoch": 1.133884096707569,
+      "grad_norm": 0.49110597372055054,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 4666
+    },
+    {
+      "epoch": 1.1341270805491435,
+      "grad_norm": 0.5043883919715881,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 4667
+    },
+    {
+      "epoch": 1.134370064390718,
+      "grad_norm": 0.48378950357437134,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 4668
+    },
+    {
+      "epoch": 1.1346130482322925,
+      "grad_norm": 0.49431201815605164,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 4669
+    },
+    {
+      "epoch": 1.1348560320738672,
+      "grad_norm": 0.48956161737442017,
+      "learning_rate": 8e-05,
+      "loss": 1.5101,
+      "step": 4670
+    },
+    {
+      "epoch": 1.1350990159154417,
+      "grad_norm": 0.5244024991989136,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 4671
+    },
+    {
+      "epoch": 1.1353419997570162,
+      "grad_norm": 0.49336352944374084,
+      "learning_rate": 8e-05,
+      "loss": 1.5354,
+      "step": 4672
+    },
+    {
+      "epoch": 1.1355849835985907,
+      "grad_norm": 0.4963606595993042,
+      "learning_rate": 8e-05,
+      "loss": 1.3996,
+      "step": 4673
+    },
+    {
+      "epoch": 1.1358279674401652,
+      "grad_norm": 0.47992652654647827,
+      "learning_rate": 8e-05,
+      "loss": 1.4554,
+      "step": 4674
+    },
+    {
+      "epoch": 1.1360709512817397,
+      "grad_norm": 0.4892204999923706,
+      "learning_rate": 8e-05,
+      "loss": 1.5391,
+      "step": 4675
+    },
+    {
+      "epoch": 1.1363139351233142,
+      "grad_norm": 0.5057500600814819,
+      "learning_rate": 8e-05,
+      "loss": 1.694,
+      "step": 4676
+    },
+    {
+      "epoch": 1.1365569189648888,
+      "grad_norm": 0.5248228907585144,
+      "learning_rate": 8e-05,
+      "loss": 1.7161,
+      "step": 4677
+    },
+    {
+      "epoch": 1.1367999028064633,
+      "grad_norm": 0.532977819442749,
+      "learning_rate": 8e-05,
+      "loss": 1.5346,
+      "step": 4678
+    },
+    {
+      "epoch": 1.137042886648038,
+      "grad_norm": 0.5113626718521118,
+      "learning_rate": 8e-05,
+      "loss": 1.6967,
+      "step": 4679
+    },
+    {
+      "epoch": 1.1372858704896125,
+      "grad_norm": 0.4692635238170624,
+      "learning_rate": 8e-05,
+      "loss": 1.5542,
+      "step": 4680
+    },
+    {
+      "epoch": 1.137528854331187,
+      "grad_norm": 0.49526554346084595,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 4681
+    },
+    {
+      "epoch": 1.1377718381727615,
+      "grad_norm": 0.5089171528816223,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 4682
+    },
+    {
+      "epoch": 1.138014822014336,
+      "grad_norm": 0.5060449242591858,
+      "learning_rate": 8e-05,
+      "loss": 1.6825,
+      "step": 4683
+    },
+    {
+      "epoch": 1.1382578058559105,
+      "grad_norm": 0.48942819237709045,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 4684
+    },
+    {
+      "epoch": 1.138500789697485,
+      "grad_norm": 0.4926145374774933,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 4685
+    },
+    {
+      "epoch": 1.1387437735390598,
+      "grad_norm": 0.5087803602218628,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 4686
+    },
+    {
+      "epoch": 1.1389867573806343,
+      "grad_norm": 0.48427438735961914,
+      "learning_rate": 8e-05,
+      "loss": 1.5031,
+      "step": 4687
+    },
+    {
+      "epoch": 1.1392297412222088,
+      "grad_norm": 0.5045105218887329,
+      "learning_rate": 8e-05,
+      "loss": 1.5909,
+      "step": 4688
+    },
+    {
+      "epoch": 1.1394727250637833,
+      "grad_norm": 0.4508696496486664,
+      "learning_rate": 8e-05,
+      "loss": 1.4449,
+      "step": 4689
+    },
+    {
+      "epoch": 1.1397157089053578,
+      "grad_norm": 0.49791356921195984,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1399586927469323,
+      "grad_norm": 0.4775407910346985,
+      "learning_rate": 8e-05,
+      "loss": 1.387,
+      "step": 4691
+    },
+    {
+      "epoch": 1.1402016765885068,
+      "grad_norm": 0.511302649974823,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 4692
+    },
+    {
+      "epoch": 1.1404446604300813,
+      "grad_norm": 0.45818546414375305,
+      "learning_rate": 8e-05,
+      "loss": 1.404,
+      "step": 4693
+    },
+    {
+      "epoch": 1.140687644271656,
+      "grad_norm": 0.49595290422439575,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 4694
+    },
+    {
+      "epoch": 1.1409306281132305,
+      "grad_norm": 0.5106500387191772,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 4695
+    },
+    {
+      "epoch": 1.141173611954805,
+      "grad_norm": 0.49878454208374023,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 4696
+    },
+    {
+      "epoch": 1.1414165957963796,
+      "grad_norm": 0.500397264957428,
+      "learning_rate": 8e-05,
+      "loss": 1.4855,
+      "step": 4697
+    },
+    {
+      "epoch": 1.141659579637954,
+      "grad_norm": 0.4937758445739746,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 4698
+    },
+    {
+      "epoch": 1.1419025634795286,
+      "grad_norm": 0.4662370979785919,
+      "learning_rate": 8e-05,
+      "loss": 1.5188,
+      "step": 4699
+    },
+    {
+      "epoch": 1.142145547321103,
+      "grad_norm": 0.5035045742988586,
+      "learning_rate": 8e-05,
+      "loss": 1.75,
+      "step": 4700
+    },
+    {
+      "epoch": 1.1423885311626778,
+      "grad_norm": 0.46537551283836365,
+      "learning_rate": 8e-05,
+      "loss": 1.5041,
+      "step": 4701
+    },
+    {
+      "epoch": 1.1426315150042523,
+      "grad_norm": 0.4963296055793762,
+      "learning_rate": 8e-05,
+      "loss": 1.6238,
+      "step": 4702
+    },
+    {
+      "epoch": 1.1428744988458268,
+      "grad_norm": 0.4910426139831543,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 4703
+    },
+    {
+      "epoch": 1.1431174826874013,
+      "grad_norm": 0.4825107455253601,
+      "learning_rate": 8e-05,
+      "loss": 1.4731,
+      "step": 4704
+    },
+    {
+      "epoch": 1.1433604665289758,
+      "grad_norm": 0.5301664471626282,
+      "learning_rate": 8e-05,
+      "loss": 1.5283,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1436034503705503,
+      "grad_norm": 0.4744507968425751,
+      "learning_rate": 8e-05,
+      "loss": 1.5014,
+      "step": 4706
+    },
+    {
+      "epoch": 1.1438464342121248,
+      "grad_norm": 0.518217921257019,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 4707
+    },
+    {
+      "epoch": 1.1440894180536993,
+      "grad_norm": 0.5055997967720032,
+      "learning_rate": 8e-05,
+      "loss": 1.4942,
+      "step": 4708
+    },
+    {
+      "epoch": 1.1443324018952739,
+      "grad_norm": 0.4914219081401825,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 4709
+    },
+    {
+      "epoch": 1.1445753857368486,
+      "grad_norm": 0.4821662902832031,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 4710
+    },
+    {
+      "epoch": 1.144818369578423,
+      "grad_norm": 0.46851927042007446,
+      "learning_rate": 8e-05,
+      "loss": 1.5849,
+      "step": 4711
+    },
+    {
+      "epoch": 1.1450613534199976,
+      "grad_norm": 0.4876309633255005,
+      "learning_rate": 8e-05,
+      "loss": 1.6333,
+      "step": 4712
+    },
+    {
+      "epoch": 1.145304337261572,
+      "grad_norm": 0.47622254490852356,
+      "learning_rate": 8e-05,
+      "loss": 1.4008,
+      "step": 4713
+    },
+    {
+      "epoch": 1.1455473211031466,
+      "grad_norm": 0.5219088792800903,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 4714
+    },
+    {
+      "epoch": 1.1457903049447211,
+      "grad_norm": 0.5406057238578796,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 4715
+    },
+    {
+      "epoch": 1.1460332887862956,
+      "grad_norm": 0.4521893262863159,
+      "learning_rate": 8e-05,
+      "loss": 1.4575,
+      "step": 4716
+    },
+    {
+      "epoch": 1.1462762726278704,
+      "grad_norm": 0.5123537182807922,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 4717
+    },
+    {
+      "epoch": 1.1465192564694449,
+      "grad_norm": 0.5037504434585571,
+      "learning_rate": 8e-05,
+      "loss": 1.6159,
+      "step": 4718
+    },
+    {
+      "epoch": 1.1467622403110194,
+      "grad_norm": 0.5007531642913818,
+      "learning_rate": 8e-05,
+      "loss": 1.4692,
+      "step": 4719
+    },
+    {
+      "epoch": 1.1470052241525939,
+      "grad_norm": 0.4687584936618805,
+      "learning_rate": 8e-05,
+      "loss": 1.4393,
+      "step": 4720
+    },
+    {
+      "epoch": 1.1472482079941684,
+      "grad_norm": 0.48607566952705383,
+      "learning_rate": 8e-05,
+      "loss": 1.6602,
+      "step": 4721
+    },
+    {
+      "epoch": 1.1474911918357429,
+      "grad_norm": 0.5033706426620483,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 4722
+    },
+    {
+      "epoch": 1.1477341756773174,
+      "grad_norm": 0.48292407393455505,
+      "learning_rate": 8e-05,
+      "loss": 1.5219,
+      "step": 4723
+    },
+    {
+      "epoch": 1.147977159518892,
+      "grad_norm": 0.4908571243286133,
+      "learning_rate": 8e-05,
+      "loss": 1.5095,
+      "step": 4724
+    },
+    {
+      "epoch": 1.1482201433604664,
+      "grad_norm": 0.5198318958282471,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 4725
+    },
+    {
+      "epoch": 1.1484631272020411,
+      "grad_norm": 0.522821307182312,
+      "learning_rate": 8e-05,
+      "loss": 1.5839,
+      "step": 4726
+    },
+    {
+      "epoch": 1.1487061110436156,
+      "grad_norm": 0.5029705166816711,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 4727
+    },
+    {
+      "epoch": 1.1489490948851901,
+      "grad_norm": 0.482683390378952,
+      "learning_rate": 8e-05,
+      "loss": 1.5952,
+      "step": 4728
+    },
+    {
+      "epoch": 1.1491920787267647,
+      "grad_norm": 0.4707898795604706,
+      "learning_rate": 8e-05,
+      "loss": 1.5077,
+      "step": 4729
+    },
+    {
+      "epoch": 1.1494350625683392,
+      "grad_norm": 0.49435025453567505,
+      "learning_rate": 8e-05,
+      "loss": 1.5126,
+      "step": 4730
+    },
+    {
+      "epoch": 1.1496780464099137,
+      "grad_norm": 0.48430174589157104,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 4731
+    },
+    {
+      "epoch": 1.1499210302514884,
+      "grad_norm": 0.5027957558631897,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 4732
+    },
+    {
+      "epoch": 1.150164014093063,
+      "grad_norm": 0.5057840347290039,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 4733
+    },
+    {
+      "epoch": 1.1504069979346374,
+      "grad_norm": 0.49302026629447937,
+      "learning_rate": 8e-05,
+      "loss": 1.5188,
+      "step": 4734
+    },
+    {
+      "epoch": 1.150649981776212,
+      "grad_norm": 0.4590870440006256,
+      "learning_rate": 8e-05,
+      "loss": 1.3998,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1508929656177864,
+      "grad_norm": 0.4875410199165344,
+      "learning_rate": 8e-05,
+      "loss": 1.4736,
+      "step": 4736
+    },
+    {
+      "epoch": 1.151135949459361,
+      "grad_norm": 0.4885874092578888,
+      "learning_rate": 8e-05,
+      "loss": 1.4192,
+      "step": 4737
+    },
+    {
+      "epoch": 1.1513789333009354,
+      "grad_norm": 0.48425978422164917,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 4738
+    },
+    {
+      "epoch": 1.15162191714251,
+      "grad_norm": 0.48759689927101135,
+      "learning_rate": 8e-05,
+      "loss": 1.5119,
+      "step": 4739
+    },
+    {
+      "epoch": 1.1518649009840844,
+      "grad_norm": 0.4728999435901642,
+      "learning_rate": 8e-05,
+      "loss": 1.5266,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1521078848256592,
+      "grad_norm": 0.4802074134349823,
+      "learning_rate": 8e-05,
+      "loss": 1.5267,
+      "step": 4741
+    },
+    {
+      "epoch": 1.1523508686672337,
+      "grad_norm": 0.4883311688899994,
+      "learning_rate": 8e-05,
+      "loss": 1.5796,
+      "step": 4742
+    },
+    {
+      "epoch": 1.1525938525088082,
+      "grad_norm": 0.5358640551567078,
+      "learning_rate": 8e-05,
+      "loss": 1.7944,
+      "step": 4743
+    },
+    {
+      "epoch": 1.1528368363503827,
+      "grad_norm": 0.47325432300567627,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 4744
+    },
+    {
+      "epoch": 1.1530798201919572,
+      "grad_norm": 0.4655816853046417,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 4745
+    },
+    {
+      "epoch": 1.1533228040335317,
+      "grad_norm": 0.4938674867153168,
+      "learning_rate": 8e-05,
+      "loss": 1.679,
+      "step": 4746
+    },
+    {
+      "epoch": 1.1535657878751062,
+      "grad_norm": 0.4819544851779938,
+      "learning_rate": 8e-05,
+      "loss": 1.4388,
+      "step": 4747
+    },
+    {
+      "epoch": 1.153808771716681,
+      "grad_norm": 0.4740254878997803,
+      "learning_rate": 8e-05,
+      "loss": 1.537,
+      "step": 4748
+    },
+    {
+      "epoch": 1.1540517555582555,
+      "grad_norm": 0.4750215411186218,
+      "learning_rate": 8e-05,
+      "loss": 1.4199,
+      "step": 4749
+    },
+    {
+      "epoch": 1.15429473939983,
+      "grad_norm": 0.4926520884037018,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1545377232414045,
+      "grad_norm": 0.5084801912307739,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 4751
+    },
+    {
+      "epoch": 1.154780707082979,
+      "grad_norm": 0.5159065127372742,
+      "learning_rate": 8e-05,
+      "loss": 1.4692,
+      "step": 4752
+    },
+    {
+      "epoch": 1.1550236909245535,
+      "grad_norm": 0.4797264337539673,
+      "learning_rate": 8e-05,
+      "loss": 1.4043,
+      "step": 4753
+    },
+    {
+      "epoch": 1.155266674766128,
+      "grad_norm": 0.4910602271556854,
+      "learning_rate": 8e-05,
+      "loss": 1.5412,
+      "step": 4754
+    },
+    {
+      "epoch": 1.1555096586077025,
+      "grad_norm": 0.4896821975708008,
+      "learning_rate": 8e-05,
+      "loss": 1.4604,
+      "step": 4755
+    },
+    {
+      "epoch": 1.155752642449277,
+      "grad_norm": 0.4575153589248657,
+      "learning_rate": 8e-05,
+      "loss": 1.3905,
+      "step": 4756
+    },
+    {
+      "epoch": 1.1559956262908517,
+      "grad_norm": 0.5031094551086426,
+      "learning_rate": 8e-05,
+      "loss": 1.6907,
+      "step": 4757
+    },
+    {
+      "epoch": 1.1562386101324262,
+      "grad_norm": 0.48745593428611755,
+      "learning_rate": 8e-05,
+      "loss": 1.5557,
+      "step": 4758
+    },
+    {
+      "epoch": 1.1564815939740007,
+      "grad_norm": 0.481069415807724,
+      "learning_rate": 8e-05,
+      "loss": 1.6297,
+      "step": 4759
+    },
+    {
+      "epoch": 1.1567245778155753,
+      "grad_norm": 0.49431103467941284,
+      "learning_rate": 8e-05,
+      "loss": 1.7607,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1569675616571498,
+      "grad_norm": 0.49356427788734436,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 4761
+    },
+    {
+      "epoch": 1.1572105454987243,
+      "grad_norm": 0.47872838377952576,
+      "learning_rate": 8e-05,
+      "loss": 1.5089,
+      "step": 4762
+    },
+    {
+      "epoch": 1.157453529340299,
+      "grad_norm": 0.4987201988697052,
+      "learning_rate": 8e-05,
+      "loss": 1.6603,
+      "step": 4763
+    },
+    {
+      "epoch": 1.1576965131818735,
+      "grad_norm": 0.49159857630729675,
+      "learning_rate": 8e-05,
+      "loss": 1.7045,
+      "step": 4764
+    },
+    {
+      "epoch": 1.157939497023448,
+      "grad_norm": 0.4916110634803772,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 4765
+    },
+    {
+      "epoch": 1.1581824808650225,
+      "grad_norm": 0.4734465181827545,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 4766
+    },
+    {
+      "epoch": 1.158425464706597,
+      "grad_norm": 0.49806684255599976,
+      "learning_rate": 8e-05,
+      "loss": 1.6987,
+      "step": 4767
+    },
+    {
+      "epoch": 1.1586684485481715,
+      "grad_norm": 0.49170219898223877,
+      "learning_rate": 8e-05,
+      "loss": 1.6501,
+      "step": 4768
+    },
+    {
+      "epoch": 1.158911432389746,
+      "grad_norm": 0.49472299218177795,
+      "learning_rate": 8e-05,
+      "loss": 1.606,
+      "step": 4769
+    },
+    {
+      "epoch": 1.1591544162313205,
+      "grad_norm": 0.5127074122428894,
+      "learning_rate": 8e-05,
+      "loss": 1.6337,
+      "step": 4770
+    },
+    {
+      "epoch": 1.159397400072895,
+      "grad_norm": 0.48259374499320984,
+      "learning_rate": 8e-05,
+      "loss": 1.4517,
+      "step": 4771
+    },
+    {
+      "epoch": 1.1596403839144698,
+      "grad_norm": 0.46888816356658936,
+      "learning_rate": 8e-05,
+      "loss": 1.4475,
+      "step": 4772
+    },
+    {
+      "epoch": 1.1598833677560443,
+      "grad_norm": 0.46909913420677185,
+      "learning_rate": 8e-05,
+      "loss": 1.5184,
+      "step": 4773
+    },
+    {
+      "epoch": 1.1601263515976188,
+      "grad_norm": 0.527323842048645,
+      "learning_rate": 8e-05,
+      "loss": 1.6817,
+      "step": 4774
+    },
+    {
+      "epoch": 1.1603693354391933,
+      "grad_norm": 0.4771113097667694,
+      "learning_rate": 8e-05,
+      "loss": 1.4259,
+      "step": 4775
+    },
+    {
+      "epoch": 1.1606123192807678,
+      "grad_norm": 0.5066484212875366,
+      "learning_rate": 8e-05,
+      "loss": 1.5221,
+      "step": 4776
+    },
+    {
+      "epoch": 1.1608553031223423,
+      "grad_norm": 0.47015535831451416,
+      "learning_rate": 8e-05,
+      "loss": 1.3532,
+      "step": 4777
+    },
+    {
+      "epoch": 1.1610982869639168,
+      "grad_norm": 0.46470826864242554,
+      "learning_rate": 8e-05,
+      "loss": 1.4373,
+      "step": 4778
+    },
+    {
+      "epoch": 1.1613412708054915,
+      "grad_norm": 0.49431952834129333,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 4779
+    },
+    {
+      "epoch": 1.161584254647066,
+      "grad_norm": 0.47251176834106445,
+      "learning_rate": 8e-05,
+      "loss": 1.5002,
+      "step": 4780
+    },
+    {
+      "epoch": 1.1618272384886406,
+      "grad_norm": 0.5086686611175537,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 4781
+    },
+    {
+      "epoch": 1.162070222330215,
+      "grad_norm": 0.47409072518348694,
+      "learning_rate": 8e-05,
+      "loss": 1.4988,
+      "step": 4782
+    },
+    {
+      "epoch": 1.1623132061717896,
+      "grad_norm": 0.47215718030929565,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 4783
+    },
+    {
+      "epoch": 1.162556190013364,
+      "grad_norm": 0.48927122354507446,
+      "learning_rate": 8e-05,
+      "loss": 1.508,
+      "step": 4784
+    },
+    {
+      "epoch": 1.1627991738549386,
+      "grad_norm": 0.5188973546028137,
+      "learning_rate": 8e-05,
+      "loss": 1.4858,
+      "step": 4785
+    },
+    {
+      "epoch": 1.163042157696513,
+      "grad_norm": 0.4959692656993866,
+      "learning_rate": 8e-05,
+      "loss": 1.5514,
+      "step": 4786
+    },
+    {
+      "epoch": 1.1632851415380876,
+      "grad_norm": 0.5571417808532715,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 4787
+    },
+    {
+      "epoch": 1.1635281253796623,
+      "grad_norm": 0.48458805680274963,
+      "learning_rate": 8e-05,
+      "loss": 1.5498,
+      "step": 4788
+    },
+    {
+      "epoch": 1.1637711092212368,
+      "grad_norm": 0.5064283013343811,
+      "learning_rate": 8e-05,
+      "loss": 1.6008,
+      "step": 4789
+    },
+    {
+      "epoch": 1.1640140930628113,
+      "grad_norm": 0.4732767641544342,
+      "learning_rate": 8e-05,
+      "loss": 1.4393,
+      "step": 4790
+    },
+    {
+      "epoch": 1.1642570769043858,
+      "grad_norm": 0.47133201360702515,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 4791
+    },
+    {
+      "epoch": 1.1645000607459604,
+      "grad_norm": 0.45641860365867615,
+      "learning_rate": 8e-05,
+      "loss": 1.5238,
+      "step": 4792
+    },
+    {
+      "epoch": 1.1647430445875349,
+      "grad_norm": 0.4953515827655792,
+      "learning_rate": 8e-05,
+      "loss": 1.5202,
+      "step": 4793
+    },
+    {
+      "epoch": 1.1649860284291096,
+      "grad_norm": 0.531728982925415,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 4794
+    },
+    {
+      "epoch": 1.165229012270684,
+      "grad_norm": 0.4854678511619568,
+      "learning_rate": 8e-05,
+      "loss": 1.552,
+      "step": 4795
+    },
+    {
+      "epoch": 1.1654719961122586,
+      "grad_norm": 0.49309593439102173,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 4796
+    },
+    {
+      "epoch": 1.165714979953833,
+      "grad_norm": 0.48511409759521484,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 4797
+    },
+    {
+      "epoch": 1.1659579637954076,
+      "grad_norm": 0.47902655601501465,
+      "learning_rate": 8e-05,
+      "loss": 1.4911,
+      "step": 4798
+    },
+    {
+      "epoch": 1.1662009476369821,
+      "grad_norm": 0.49756568670272827,
+      "learning_rate": 8e-05,
+      "loss": 1.601,
+      "step": 4799
+    },
+    {
+      "epoch": 1.1664439314785566,
+      "grad_norm": 0.5049137473106384,
+      "learning_rate": 8e-05,
+      "loss": 1.5576,
+      "step": 4800
+    },
+    {
+      "epoch": 1.1666869153201311,
+      "grad_norm": 0.5164820551872253,
+      "learning_rate": 8e-05,
+      "loss": 1.7143,
+      "step": 4801
+    },
+    {
+      "epoch": 1.1669298991617056,
+      "grad_norm": 0.49506452679634094,
+      "learning_rate": 8e-05,
+      "loss": 1.4987,
+      "step": 4802
+    },
+    {
+      "epoch": 1.1671728830032804,
+      "grad_norm": 0.5335606932640076,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 4803
+    },
+    {
+      "epoch": 1.1674158668448549,
+      "grad_norm": 0.510491669178009,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 4804
+    },
+    {
+      "epoch": 1.1676588506864294,
+      "grad_norm": 0.4835631847381592,
+      "learning_rate": 8e-05,
+      "loss": 1.4574,
+      "step": 4805
+    },
+    {
+      "epoch": 1.167901834528004,
+      "grad_norm": 0.4844507873058319,
+      "learning_rate": 8e-05,
+      "loss": 1.414,
+      "step": 4806
+    },
+    {
+      "epoch": 1.1681448183695784,
+      "grad_norm": 0.49851512908935547,
+      "learning_rate": 8e-05,
+      "loss": 1.6168,
+      "step": 4807
+    },
+    {
+      "epoch": 1.168387802211153,
+      "grad_norm": 0.4857181906700134,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 4808
+    },
+    {
+      "epoch": 1.1686307860527274,
+      "grad_norm": 0.4759085178375244,
+      "learning_rate": 8e-05,
+      "loss": 1.5165,
+      "step": 4809
+    },
+    {
+      "epoch": 1.1688737698943021,
+      "grad_norm": 0.5064108967781067,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 4810
+    },
+    {
+      "epoch": 1.1691167537358766,
+      "grad_norm": 0.47053155303001404,
+      "learning_rate": 8e-05,
+      "loss": 1.4361,
+      "step": 4811
+    },
+    {
+      "epoch": 1.1693597375774512,
+      "grad_norm": 0.4682636857032776,
+      "learning_rate": 8e-05,
+      "loss": 1.5549,
+      "step": 4812
+    },
+    {
+      "epoch": 1.1696027214190257,
+      "grad_norm": 0.47961997985839844,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 4813
+    },
+    {
+      "epoch": 1.1698457052606002,
+      "grad_norm": 0.4778006672859192,
+      "learning_rate": 8e-05,
+      "loss": 1.5625,
+      "step": 4814
+    },
+    {
+      "epoch": 1.1700886891021747,
+      "grad_norm": 0.47755634784698486,
+      "learning_rate": 8e-05,
+      "loss": 1.6369,
+      "step": 4815
+    },
+    {
+      "epoch": 1.1703316729437492,
+      "grad_norm": 0.49588897824287415,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 4816
+    },
+    {
+      "epoch": 1.1705746567853237,
+      "grad_norm": 0.5048012733459473,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 4817
+    },
+    {
+      "epoch": 1.1708176406268982,
+      "grad_norm": 0.5149305462837219,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 4818
+    },
+    {
+      "epoch": 1.171060624468473,
+      "grad_norm": 0.5121803879737854,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 4819
+    },
+    {
+      "epoch": 1.1713036083100474,
+      "grad_norm": 0.4805653989315033,
+      "learning_rate": 8e-05,
+      "loss": 1.5325,
+      "step": 4820
+    },
+    {
+      "epoch": 1.171546592151622,
+      "grad_norm": 0.5157588720321655,
+      "learning_rate": 8e-05,
+      "loss": 1.5215,
+      "step": 4821
+    },
+    {
+      "epoch": 1.1717895759931964,
+      "grad_norm": 0.505052387714386,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 4822
+    },
+    {
+      "epoch": 1.172032559834771,
+      "grad_norm": 0.49368759989738464,
+      "learning_rate": 8e-05,
+      "loss": 1.4065,
+      "step": 4823
+    },
+    {
+      "epoch": 1.1722755436763455,
+      "grad_norm": 0.5167792439460754,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 4824
+    },
+    {
+      "epoch": 1.17251852751792,
+      "grad_norm": 0.4676288366317749,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 4825
+    },
+    {
+      "epoch": 1.1727615113594947,
+      "grad_norm": 0.4993060827255249,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 4826
+    },
+    {
+      "epoch": 1.1730044952010692,
+      "grad_norm": 0.48399436473846436,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 4827
+    },
+    {
+      "epoch": 1.1732474790426437,
+      "grad_norm": 0.4882155954837799,
+      "learning_rate": 8e-05,
+      "loss": 1.5902,
+      "step": 4828
+    },
+    {
+      "epoch": 1.1734904628842182,
+      "grad_norm": 0.504315197467804,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 4829
+    },
+    {
+      "epoch": 1.1737334467257927,
+      "grad_norm": 0.5006396770477295,
+      "learning_rate": 8e-05,
+      "loss": 1.4772,
+      "step": 4830
+    },
+    {
+      "epoch": 1.1739764305673672,
+      "grad_norm": 0.4713663160800934,
+      "learning_rate": 8e-05,
+      "loss": 1.5249,
+      "step": 4831
+    },
+    {
+      "epoch": 1.1742194144089417,
+      "grad_norm": 0.5137341022491455,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 4832
+    },
+    {
+      "epoch": 1.1744623982505162,
+      "grad_norm": 0.45355480909347534,
+      "learning_rate": 8e-05,
+      "loss": 1.483,
+      "step": 4833
+    },
+    {
+      "epoch": 1.174705382092091,
+      "grad_norm": 0.5054458975791931,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 4834
+    },
+    {
+      "epoch": 1.1749483659336655,
+      "grad_norm": 0.48597925901412964,
+      "learning_rate": 8e-05,
+      "loss": 1.5643,
+      "step": 4835
+    },
+    {
+      "epoch": 1.17519134977524,
+      "grad_norm": 0.5169289112091064,
+      "learning_rate": 8e-05,
+      "loss": 1.6078,
+      "step": 4836
+    },
+    {
+      "epoch": 1.1754343336168145,
+      "grad_norm": 0.48553454875946045,
+      "learning_rate": 8e-05,
+      "loss": 1.5063,
+      "step": 4837
+    },
+    {
+      "epoch": 1.175677317458389,
+      "grad_norm": 0.48568373918533325,
+      "learning_rate": 8e-05,
+      "loss": 1.5226,
+      "step": 4838
+    },
+    {
+      "epoch": 1.1759203012999635,
+      "grad_norm": 0.46898719668388367,
+      "learning_rate": 8e-05,
+      "loss": 1.543,
+      "step": 4839
+    },
+    {
+      "epoch": 1.176163285141538,
+      "grad_norm": 0.505601704120636,
+      "learning_rate": 8e-05,
+      "loss": 1.5355,
+      "step": 4840
+    },
+    {
+      "epoch": 1.1764062689831127,
+      "grad_norm": 0.4963681995868683,
+      "learning_rate": 8e-05,
+      "loss": 1.6093,
+      "step": 4841
+    },
+    {
+      "epoch": 1.1766492528246872,
+      "grad_norm": 0.5220352411270142,
+      "learning_rate": 8e-05,
+      "loss": 1.6957,
+      "step": 4842
+    },
+    {
+      "epoch": 1.1768922366662617,
+      "grad_norm": 0.4835437834262848,
+      "learning_rate": 8e-05,
+      "loss": 1.6669,
+      "step": 4843
+    },
+    {
+      "epoch": 1.1771352205078363,
+      "grad_norm": 0.4715404808521271,
+      "learning_rate": 8e-05,
+      "loss": 1.4991,
+      "step": 4844
+    },
+    {
+      "epoch": 1.1773782043494108,
+      "grad_norm": 0.5124507546424866,
+      "learning_rate": 8e-05,
+      "loss": 1.4876,
+      "step": 4845
+    },
+    {
+      "epoch": 1.1776211881909853,
+      "grad_norm": 0.5009899735450745,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 4846
+    },
+    {
+      "epoch": 1.1778641720325598,
+      "grad_norm": 0.48378613591194153,
+      "learning_rate": 8e-05,
+      "loss": 1.5293,
+      "step": 4847
+    },
+    {
+      "epoch": 1.1781071558741343,
+      "grad_norm": 0.47757688164711,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 4848
+    },
+    {
+      "epoch": 1.1783501397157088,
+      "grad_norm": 0.48205479979515076,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 4849
+    },
+    {
+      "epoch": 1.1785931235572835,
+      "grad_norm": 0.4974353015422821,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 4850
+    },
+    {
+      "epoch": 1.178836107398858,
+      "grad_norm": 0.4831935167312622,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 4851
+    },
+    {
+      "epoch": 1.1790790912404325,
+      "grad_norm": 0.4777524173259735,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 4852
+    },
+    {
+      "epoch": 1.179322075082007,
+      "grad_norm": 0.5090939998626709,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 4853
+    },
+    {
+      "epoch": 1.1795650589235815,
+      "grad_norm": 0.48446887731552124,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 4854
+    },
+    {
+      "epoch": 1.179808042765156,
+      "grad_norm": 0.5030233263969421,
+      "learning_rate": 8e-05,
+      "loss": 1.5602,
+      "step": 4855
+    },
+    {
+      "epoch": 1.1800510266067306,
+      "grad_norm": 0.49351197481155396,
+      "learning_rate": 8e-05,
+      "loss": 1.5073,
+      "step": 4856
+    },
+    {
+      "epoch": 1.1802940104483053,
+      "grad_norm": 0.5247987508773804,
+      "learning_rate": 8e-05,
+      "loss": 1.4947,
+      "step": 4857
+    },
+    {
+      "epoch": 1.1805369942898798,
+      "grad_norm": 0.5140460729598999,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 4858
+    },
+    {
+      "epoch": 1.1807799781314543,
+      "grad_norm": 0.6042105555534363,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 4859
+    },
+    {
+      "epoch": 1.1810229619730288,
+      "grad_norm": 0.4886859357357025,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 4860
+    },
+    {
+      "epoch": 1.1812659458146033,
+      "grad_norm": 0.4755132496356964,
+      "learning_rate": 8e-05,
+      "loss": 1.4459,
+      "step": 4861
+    },
+    {
+      "epoch": 1.1815089296561778,
+      "grad_norm": 0.5182363986968994,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 4862
+    },
+    {
+      "epoch": 1.1817519134977523,
+      "grad_norm": 0.4874965250492096,
+      "learning_rate": 8e-05,
+      "loss": 1.4847,
+      "step": 4863
+    },
+    {
+      "epoch": 1.1819948973393268,
+      "grad_norm": 0.5293747186660767,
+      "learning_rate": 8e-05,
+      "loss": 1.4515,
+      "step": 4864
+    },
+    {
+      "epoch": 1.1822378811809016,
+      "grad_norm": 0.4768170118331909,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 4865
+    },
+    {
+      "epoch": 1.182480865022476,
+      "grad_norm": 0.4977954030036926,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 4866
+    },
+    {
+      "epoch": 1.1827238488640506,
+      "grad_norm": 0.4954514801502228,
+      "learning_rate": 8e-05,
+      "loss": 1.5559,
+      "step": 4867
+    },
+    {
+      "epoch": 1.182966832705625,
+      "grad_norm": 0.49372243881225586,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 4868
+    },
+    {
+      "epoch": 1.1832098165471996,
+      "grad_norm": 0.4866653382778168,
+      "learning_rate": 8e-05,
+      "loss": 1.5512,
+      "step": 4869
+    },
+    {
+      "epoch": 1.183452800388774,
+      "grad_norm": 0.4873413145542145,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 4870
+    },
+    {
+      "epoch": 1.1836957842303486,
+      "grad_norm": 0.5071091055870056,
+      "learning_rate": 8e-05,
+      "loss": 1.495,
+      "step": 4871
+    },
+    {
+      "epoch": 1.1839387680719233,
+      "grad_norm": 0.4897591769695282,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 4872
+    },
+    {
+      "epoch": 1.1841817519134978,
+      "grad_norm": 0.4870610237121582,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 4873
+    },
+    {
+      "epoch": 1.1844247357550723,
+      "grad_norm": 0.497335821390152,
+      "learning_rate": 8e-05,
+      "loss": 1.6769,
+      "step": 4874
+    },
+    {
+      "epoch": 1.1846677195966469,
+      "grad_norm": 0.5242437124252319,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 4875
+    },
+    {
+      "epoch": 1.1849107034382214,
+      "grad_norm": 0.4999677240848541,
+      "learning_rate": 8e-05,
+      "loss": 1.5262,
+      "step": 4876
+    },
+    {
+      "epoch": 1.1851536872797959,
+      "grad_norm": 0.4961715340614319,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 4877
+    },
+    {
+      "epoch": 1.1853966711213704,
+      "grad_norm": 0.4674530625343323,
+      "learning_rate": 8e-05,
+      "loss": 1.3375,
+      "step": 4878
+    },
+    {
+      "epoch": 1.1856396549629449,
+      "grad_norm": 0.5108457207679749,
+      "learning_rate": 8e-05,
+      "loss": 1.4413,
+      "step": 4879
+    },
+    {
+      "epoch": 1.1858826388045194,
+      "grad_norm": 0.5021309852600098,
+      "learning_rate": 8e-05,
+      "loss": 1.4955,
+      "step": 4880
+    },
+    {
+      "epoch": 1.1861256226460941,
+      "grad_norm": 0.4813995063304901,
+      "learning_rate": 8e-05,
+      "loss": 1.4969,
+      "step": 4881
+    },
+    {
+      "epoch": 1.1863686064876686,
+      "grad_norm": 0.5020331144332886,
+      "learning_rate": 8e-05,
+      "loss": 1.611,
+      "step": 4882
+    },
+    {
+      "epoch": 1.1866115903292431,
+      "grad_norm": 0.47132426500320435,
+      "learning_rate": 8e-05,
+      "loss": 1.4674,
+      "step": 4883
+    },
+    {
+      "epoch": 1.1868545741708176,
+      "grad_norm": 0.5042454600334167,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 4884
+    },
+    {
+      "epoch": 1.1870975580123921,
+      "grad_norm": 0.47689178586006165,
+      "learning_rate": 8e-05,
+      "loss": 1.4526,
+      "step": 4885
+    },
+    {
+      "epoch": 1.1873405418539666,
+      "grad_norm": 0.5274168252944946,
+      "learning_rate": 8e-05,
+      "loss": 1.6323,
+      "step": 4886
+    },
+    {
+      "epoch": 1.1875835256955412,
+      "grad_norm": 0.48611509799957275,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 4887
+    },
+    {
+      "epoch": 1.1878265095371159,
+      "grad_norm": 0.4856575131416321,
+      "learning_rate": 8e-05,
+      "loss": 1.3345,
+      "step": 4888
+    },
+    {
+      "epoch": 1.1880694933786904,
+      "grad_norm": 0.5001177787780762,
+      "learning_rate": 8e-05,
+      "loss": 1.5689,
+      "step": 4889
+    },
+    {
+      "epoch": 1.188312477220265,
+      "grad_norm": 0.47780704498291016,
+      "learning_rate": 8e-05,
+      "loss": 1.5362,
+      "step": 4890
+    },
+    {
+      "epoch": 1.1885554610618394,
+      "grad_norm": 0.501974880695343,
+      "learning_rate": 8e-05,
+      "loss": 1.6305,
+      "step": 4891
+    },
+    {
+      "epoch": 1.188798444903414,
+      "grad_norm": 0.5246593952178955,
+      "learning_rate": 8e-05,
+      "loss": 1.5964,
+      "step": 4892
+    },
+    {
+      "epoch": 1.1890414287449884,
+      "grad_norm": 0.5109293460845947,
+      "learning_rate": 8e-05,
+      "loss": 1.7221,
+      "step": 4893
+    },
+    {
+      "epoch": 1.189284412586563,
+      "grad_norm": 0.4667907655239105,
+      "learning_rate": 8e-05,
+      "loss": 1.4186,
+      "step": 4894
+    },
+    {
+      "epoch": 1.1895273964281374,
+      "grad_norm": 0.4972124397754669,
+      "learning_rate": 8e-05,
+      "loss": 1.4685,
+      "step": 4895
+    },
+    {
+      "epoch": 1.1897703802697122,
+      "grad_norm": 0.5442632436752319,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 4896
+    },
+    {
+      "epoch": 1.1900133641112867,
+      "grad_norm": 0.5054337382316589,
+      "learning_rate": 8e-05,
+      "loss": 1.5278,
+      "step": 4897
+    },
+    {
+      "epoch": 1.1902563479528612,
+      "grad_norm": 0.4648742973804474,
+      "learning_rate": 8e-05,
+      "loss": 1.5297,
+      "step": 4898
+    },
+    {
+      "epoch": 1.1904993317944357,
+      "grad_norm": 0.4926924407482147,
+      "learning_rate": 8e-05,
+      "loss": 1.5335,
+      "step": 4899
+    },
+    {
+      "epoch": 1.1907423156360102,
+      "grad_norm": 0.5294963717460632,
+      "learning_rate": 8e-05,
+      "loss": 1.5361,
+      "step": 4900
+    },
+    {
+      "epoch": 1.1909852994775847,
+      "grad_norm": 0.49669149518013,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 4901
+    },
+    {
+      "epoch": 1.1912282833191592,
+      "grad_norm": 0.48247039318084717,
+      "learning_rate": 8e-05,
+      "loss": 1.5437,
+      "step": 4902
+    },
+    {
+      "epoch": 1.191471267160734,
+      "grad_norm": 0.5074687600135803,
+      "learning_rate": 8e-05,
+      "loss": 1.5326,
+      "step": 4903
+    },
+    {
+      "epoch": 1.1917142510023084,
+      "grad_norm": 0.5312740206718445,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 4904
+    },
+    {
+      "epoch": 1.191957234843883,
+      "grad_norm": 0.5274022221565247,
+      "learning_rate": 8e-05,
+      "loss": 1.4245,
+      "step": 4905
+    },
+    {
+      "epoch": 1.1922002186854574,
+      "grad_norm": 0.5465565323829651,
+      "learning_rate": 8e-05,
+      "loss": 1.6664,
+      "step": 4906
+    },
+    {
+      "epoch": 1.192443202527032,
+      "grad_norm": 0.5211489200592041,
+      "learning_rate": 8e-05,
+      "loss": 1.4923,
+      "step": 4907
+    },
+    {
+      "epoch": 1.1926861863686065,
+      "grad_norm": 0.4922049641609192,
+      "learning_rate": 8e-05,
+      "loss": 1.5637,
+      "step": 4908
+    },
+    {
+      "epoch": 1.192929170210181,
+      "grad_norm": 0.520454466342926,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 4909
+    },
+    {
+      "epoch": 1.1931721540517555,
+      "grad_norm": 0.4792466461658478,
+      "learning_rate": 8e-05,
+      "loss": 1.4152,
+      "step": 4910
+    },
+    {
+      "epoch": 1.19341513789333,
+      "grad_norm": 0.49296805262565613,
+      "learning_rate": 8e-05,
+      "loss": 1.4799,
+      "step": 4911
+    },
+    {
+      "epoch": 1.1936581217349047,
+      "grad_norm": 0.4917871057987213,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 4912
+    },
+    {
+      "epoch": 1.1939011055764792,
+      "grad_norm": 0.5282992720603943,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 4913
+    },
+    {
+      "epoch": 1.1941440894180537,
+      "grad_norm": 0.5092350840568542,
+      "learning_rate": 8e-05,
+      "loss": 1.6624,
+      "step": 4914
+    },
+    {
+      "epoch": 1.1943870732596282,
+      "grad_norm": 0.5002424716949463,
+      "learning_rate": 8e-05,
+      "loss": 1.5224,
+      "step": 4915
+    },
+    {
+      "epoch": 1.1946300571012027,
+      "grad_norm": 0.4991507828235626,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 4916
+    },
+    {
+      "epoch": 1.1948730409427772,
+      "grad_norm": 0.4798358678817749,
+      "learning_rate": 8e-05,
+      "loss": 1.4741,
+      "step": 4917
+    },
+    {
+      "epoch": 1.1951160247843517,
+      "grad_norm": 0.48927584290504456,
+      "learning_rate": 8e-05,
+      "loss": 1.554,
+      "step": 4918
+    },
+    {
+      "epoch": 1.1953590086259265,
+      "grad_norm": 0.5085628628730774,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 4919
+    },
+    {
+      "epoch": 1.195601992467501,
+      "grad_norm": 0.4918489456176758,
+      "learning_rate": 8e-05,
+      "loss": 1.4535,
+      "step": 4920
+    },
+    {
+      "epoch": 1.1958449763090755,
+      "grad_norm": 0.5270163416862488,
+      "learning_rate": 8e-05,
+      "loss": 1.7107,
+      "step": 4921
+    },
+    {
+      "epoch": 1.19608796015065,
+      "grad_norm": 0.4945998191833496,
+      "learning_rate": 8e-05,
+      "loss": 1.6391,
+      "step": 4922
+    },
+    {
+      "epoch": 1.1963309439922245,
+      "grad_norm": 0.5241203308105469,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 4923
+    },
+    {
+      "epoch": 1.196573927833799,
+      "grad_norm": 0.5361198782920837,
+      "learning_rate": 8e-05,
+      "loss": 1.6939,
+      "step": 4924
+    },
+    {
+      "epoch": 1.1968169116753735,
+      "grad_norm": 0.5357650518417358,
+      "learning_rate": 8e-05,
+      "loss": 1.8823,
+      "step": 4925
+    },
+    {
+      "epoch": 1.197059895516948,
+      "grad_norm": 0.4970327317714691,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 4926
+    },
+    {
+      "epoch": 1.1973028793585228,
+      "grad_norm": 0.5429365634918213,
+      "learning_rate": 8e-05,
+      "loss": 1.8505,
+      "step": 4927
+    },
+    {
+      "epoch": 1.1975458632000973,
+      "grad_norm": 0.5144810676574707,
+      "learning_rate": 8e-05,
+      "loss": 1.5955,
+      "step": 4928
+    },
+    {
+      "epoch": 1.1977888470416718,
+      "grad_norm": 0.5169961452484131,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 4929
+    },
+    {
+      "epoch": 1.1980318308832463,
+      "grad_norm": 0.48858097195625305,
+      "learning_rate": 8e-05,
+      "loss": 1.5093,
+      "step": 4930
+    },
+    {
+      "epoch": 1.1982748147248208,
+      "grad_norm": 0.49152645468711853,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1985177985663953,
+      "grad_norm": 0.4759743809700012,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 4932
+    },
+    {
+      "epoch": 1.1987607824079698,
+      "grad_norm": 0.496655136346817,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 4933
+    },
+    {
+      "epoch": 1.1990037662495445,
+      "grad_norm": 0.4892829358577728,
+      "learning_rate": 8e-05,
+      "loss": 1.4914,
+      "step": 4934
+    },
+    {
+      "epoch": 1.199246750091119,
+      "grad_norm": 0.4914308488368988,
+      "learning_rate": 8e-05,
+      "loss": 1.5573,
+      "step": 4935
+    },
+    {
+      "epoch": 1.1994897339326935,
+      "grad_norm": 0.48438698053359985,
+      "learning_rate": 8e-05,
+      "loss": 1.5609,
+      "step": 4936
+    },
+    {
+      "epoch": 1.199732717774268,
+      "grad_norm": 0.5089436769485474,
+      "learning_rate": 8e-05,
+      "loss": 1.5749,
+      "step": 4937
+    },
+    {
+      "epoch": 1.1999757016158425,
+      "grad_norm": 0.5164045095443726,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 4938
+    },
+    {
+      "epoch": 1.200218685457417,
+      "grad_norm": 0.4851560890674591,
+      "learning_rate": 8e-05,
+      "loss": 1.4702,
+      "step": 4939
+    },
+    {
+      "epoch": 1.2004616692989916,
+      "grad_norm": 0.5073465704917908,
+      "learning_rate": 8e-05,
+      "loss": 1.4575,
+      "step": 4940
+    },
+    {
+      "epoch": 1.200704653140566,
+      "grad_norm": 0.518622875213623,
+      "learning_rate": 8e-05,
+      "loss": 1.5886,
+      "step": 4941
+    },
+    {
+      "epoch": 1.2009476369821406,
+      "grad_norm": 0.5092657804489136,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 4942
+    },
+    {
+      "epoch": 1.2011906208237153,
+      "grad_norm": 0.4822739362716675,
+      "learning_rate": 8e-05,
+      "loss": 1.4829,
+      "step": 4943
+    },
+    {
+      "epoch": 1.2014336046652898,
+      "grad_norm": 0.4861515462398529,
+      "learning_rate": 8e-05,
+      "loss": 1.4628,
+      "step": 4944
+    },
+    {
+      "epoch": 1.2016765885068643,
+      "grad_norm": 0.4897085726261139,
+      "learning_rate": 8e-05,
+      "loss": 1.4495,
+      "step": 4945
+    },
+    {
+      "epoch": 1.2019195723484388,
+      "grad_norm": 0.5587472319602966,
+      "learning_rate": 8e-05,
+      "loss": 1.7602,
+      "step": 4946
+    },
+    {
+      "epoch": 1.2021625561900133,
+      "grad_norm": 0.5280104279518127,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 4947
+    },
+    {
+      "epoch": 1.2024055400315878,
+      "grad_norm": 0.4955413043498993,
+      "learning_rate": 8e-05,
+      "loss": 1.6357,
+      "step": 4948
+    },
+    {
+      "epoch": 1.2026485238731623,
+      "grad_norm": 0.49848154187202454,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 4949
+    },
+    {
+      "epoch": 1.202891507714737,
+      "grad_norm": 0.5088046789169312,
+      "learning_rate": 8e-05,
+      "loss": 1.5664,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2031344915563116,
+      "grad_norm": 0.501869261264801,
+      "learning_rate": 8e-05,
+      "loss": 1.4975,
+      "step": 4951
+    },
+    {
+      "epoch": 1.203377475397886,
+      "grad_norm": 0.4985013008117676,
+      "learning_rate": 8e-05,
+      "loss": 1.6967,
+      "step": 4952
+    },
+    {
+      "epoch": 1.2036204592394606,
+      "grad_norm": 0.4872387945652008,
+      "learning_rate": 8e-05,
+      "loss": 1.4936,
+      "step": 4953
+    },
+    {
+      "epoch": 1.203863443081035,
+      "grad_norm": 0.5081937313079834,
+      "learning_rate": 8e-05,
+      "loss": 1.635,
+      "step": 4954
+    },
+    {
+      "epoch": 1.2041064269226096,
+      "grad_norm": 0.5119295716285706,
+      "learning_rate": 8e-05,
+      "loss": 1.5971,
+      "step": 4955
+    },
+    {
+      "epoch": 1.2043494107641841,
+      "grad_norm": 0.49729034304618835,
+      "learning_rate": 8e-05,
+      "loss": 1.6049,
+      "step": 4956
+    },
+    {
+      "epoch": 1.2045923946057586,
+      "grad_norm": 0.48517194390296936,
+      "learning_rate": 8e-05,
+      "loss": 1.531,
+      "step": 4957
+    },
+    {
+      "epoch": 1.2048353784473333,
+      "grad_norm": 0.48284244537353516,
+      "learning_rate": 8e-05,
+      "loss": 1.5503,
+      "step": 4958
+    },
+    {
+      "epoch": 1.2050783622889079,
+      "grad_norm": 0.5146550536155701,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 4959
+    },
+    {
+      "epoch": 1.2053213461304824,
+      "grad_norm": 0.5105664730072021,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2055643299720569,
+      "grad_norm": 0.519795298576355,
+      "learning_rate": 8e-05,
+      "loss": 1.6005,
+      "step": 4961
+    },
+    {
+      "epoch": 1.2058073138136314,
+      "grad_norm": 0.5250253677368164,
+      "learning_rate": 8e-05,
+      "loss": 1.7533,
+      "step": 4962
+    },
+    {
+      "epoch": 1.2060502976552059,
+      "grad_norm": 0.5135072469711304,
+      "learning_rate": 8e-05,
+      "loss": 1.5117,
+      "step": 4963
+    },
+    {
+      "epoch": 1.2062932814967804,
+      "grad_norm": 0.5141922831535339,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 4964
+    },
+    {
+      "epoch": 1.2065362653383551,
+      "grad_norm": 0.4801460802555084,
+      "learning_rate": 8e-05,
+      "loss": 1.568,
+      "step": 4965
+    },
+    {
+      "epoch": 1.2067792491799296,
+      "grad_norm": 0.5072303414344788,
+      "learning_rate": 8e-05,
+      "loss": 1.6064,
+      "step": 4966
+    },
+    {
+      "epoch": 1.2070222330215041,
+      "grad_norm": 0.5107424855232239,
+      "learning_rate": 8e-05,
+      "loss": 1.5204,
+      "step": 4967
+    },
+    {
+      "epoch": 1.2072652168630786,
+      "grad_norm": 0.490480899810791,
+      "learning_rate": 8e-05,
+      "loss": 1.5583,
+      "step": 4968
+    },
+    {
+      "epoch": 1.2075082007046531,
+      "grad_norm": 0.49908387660980225,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 4969
+    },
+    {
+      "epoch": 1.2077511845462277,
+      "grad_norm": 0.49837201833724976,
+      "learning_rate": 8e-05,
+      "loss": 1.3751,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2079941683878022,
+      "grad_norm": 0.48241540789604187,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 4971
+    },
+    {
+      "epoch": 1.2082371522293767,
+      "grad_norm": 0.46809297800064087,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 4972
+    },
+    {
+      "epoch": 1.2084801360709512,
+      "grad_norm": 0.5077289938926697,
+      "learning_rate": 8e-05,
+      "loss": 1.706,
+      "step": 4973
+    },
+    {
+      "epoch": 1.208723119912526,
+      "grad_norm": 0.4969845712184906,
+      "learning_rate": 8e-05,
+      "loss": 1.3129,
+      "step": 4974
+    },
+    {
+      "epoch": 1.2089661037541004,
+      "grad_norm": 0.503190279006958,
+      "learning_rate": 8e-05,
+      "loss": 1.5573,
+      "step": 4975
+    },
+    {
+      "epoch": 1.209209087595675,
+      "grad_norm": 0.4771680235862732,
+      "learning_rate": 8e-05,
+      "loss": 1.4189,
+      "step": 4976
+    },
+    {
+      "epoch": 1.2094520714372494,
+      "grad_norm": 0.4991047978401184,
+      "learning_rate": 8e-05,
+      "loss": 1.5444,
+      "step": 4977
+    },
+    {
+      "epoch": 1.209695055278824,
+      "grad_norm": 0.48780396580696106,
+      "learning_rate": 8e-05,
+      "loss": 1.4774,
+      "step": 4978
+    },
+    {
+      "epoch": 1.2099380391203984,
+      "grad_norm": 0.49210116267204285,
+      "learning_rate": 8e-05,
+      "loss": 1.3153,
+      "step": 4979
+    },
+    {
+      "epoch": 1.210181022961973,
+      "grad_norm": 0.5078326463699341,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2104240068035477,
+      "grad_norm": 0.4930858314037323,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 4981
+    },
+    {
+      "epoch": 1.2106669906451222,
+      "grad_norm": 0.489206999540329,
+      "learning_rate": 8e-05,
+      "loss": 1.4865,
+      "step": 4982
+    },
+    {
+      "epoch": 1.2109099744866967,
+      "grad_norm": 0.4972721338272095,
+      "learning_rate": 8e-05,
+      "loss": 1.5554,
+      "step": 4983
+    },
+    {
+      "epoch": 1.2111529583282712,
+      "grad_norm": 0.5139051675796509,
+      "learning_rate": 8e-05,
+      "loss": 1.5959,
+      "step": 4984
+    },
+    {
+      "epoch": 1.2113959421698457,
+      "grad_norm": 0.4978826940059662,
+      "learning_rate": 8e-05,
+      "loss": 1.5786,
+      "step": 4985
+    },
+    {
+      "epoch": 1.2116389260114202,
+      "grad_norm": 0.49588507413864136,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 4986
+    },
+    {
+      "epoch": 1.2118819098529947,
+      "grad_norm": 0.5012803077697754,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 4987
+    },
+    {
+      "epoch": 1.2121248936945692,
+      "grad_norm": 0.48741668462753296,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 4988
+    },
+    {
+      "epoch": 1.2123678775361437,
+      "grad_norm": 0.5667822360992432,
+      "learning_rate": 8e-05,
+      "loss": 1.6532,
+      "step": 4989
+    },
+    {
+      "epoch": 1.2126108613777185,
+      "grad_norm": 0.4834272563457489,
+      "learning_rate": 8e-05,
+      "loss": 1.4925,
+      "step": 4990
+    },
+    {
+      "epoch": 1.212853845219293,
+      "grad_norm": 0.513761579990387,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 4991
+    },
+    {
+      "epoch": 1.2130968290608675,
+      "grad_norm": 0.5211657285690308,
+      "learning_rate": 8e-05,
+      "loss": 1.7651,
+      "step": 4992
+    },
+    {
+      "epoch": 1.213339812902442,
+      "grad_norm": 0.4817950427532196,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 4993
+    },
+    {
+      "epoch": 1.2135827967440165,
+      "grad_norm": 0.48243704438209534,
+      "learning_rate": 8e-05,
+      "loss": 1.5526,
+      "step": 4994
+    },
+    {
+      "epoch": 1.213825780585591,
+      "grad_norm": 0.49206146597862244,
+      "learning_rate": 8e-05,
+      "loss": 1.4893,
+      "step": 4995
+    },
+    {
+      "epoch": 1.2140687644271657,
+      "grad_norm": 0.49845072627067566,
+      "learning_rate": 8e-05,
+      "loss": 1.3752,
+      "step": 4996
+    },
+    {
+      "epoch": 1.2143117482687402,
+      "grad_norm": 0.5011497139930725,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 4997
+    },
+    {
+      "epoch": 1.2145547321103147,
+      "grad_norm": 0.5142675638198853,
+      "learning_rate": 8e-05,
+      "loss": 1.6115,
+      "step": 4998
+    },
+    {
+      "epoch": 1.2147977159518892,
+      "grad_norm": 0.5021056532859802,
+      "learning_rate": 8e-05,
+      "loss": 1.5254,
+      "step": 4999
+    },
+    {
+      "epoch": 1.2150406997934637,
+      "grad_norm": 0.5042612552642822,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 9.5312890822656e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}