diff --git "a/checkpoint-11676/trainer_state.json" "b/checkpoint-11676/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-11676/trainer_state.json"
@@ -0,0 +1,81765 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 11676,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.564576909900651e-05,
+      "grad_norm": 2.1407182216644287,
+      "learning_rate": 0.0001,
+      "loss": 3.8027,
+      "step": 1
+    },
+    {
+      "epoch": 0.00017129153819801302,
+      "grad_norm": 1.3210175037384033,
+      "learning_rate": 0.0001,
+      "loss": 1.9741,
+      "step": 2
+    },
+    {
+      "epoch": 0.00025693730729701953,
+      "grad_norm": 0.7681771516799927,
+      "learning_rate": 0.0001,
+      "loss": 1.2891,
+      "step": 3
+    },
+    {
+      "epoch": 0.00034258307639602604,
+      "grad_norm": 0.5840440392494202,
+      "learning_rate": 0.0001,
+      "loss": 0.997,
+      "step": 4
+    },
+    {
+      "epoch": 0.00042822884549503255,
+      "grad_norm": 0.7029933333396912,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005138746145940391,
+      "grad_norm": 0.49301037192344666,
+      "learning_rate": 0.0001,
+      "loss": 0.87,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005995203836930455,
+      "grad_norm": 0.478528767824173,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006851661527920521,
+      "grad_norm": 0.543280303478241,
+      "learning_rate": 0.0001,
+      "loss": 1.4063,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007708119218910585,
+      "grad_norm": 0.6935099959373474,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008564576909900651,
+      "grad_norm": 1.4179168939590454,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009421034600890716,
+      "grad_norm": 1.2076539993286133,
+      "learning_rate": 0.0001,
+      "loss": 1.1754,
+      "step": 11
+    },
+    {
+      "epoch": 0.0010277492291880781,
+      "grad_norm": 1.1194623708724976,
+      "learning_rate": 0.0001,
+      "loss": 2.1792,
+      "step": 12
+    },
+    {
+      "epoch": 0.0011133949982870847,
+      "grad_norm": 0.5073934197425842,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 13
+    },
+    {
+      "epoch": 0.001199040767386091,
+      "grad_norm": 0.593657374382019,
+      "learning_rate": 0.0001,
+      "loss": 0.9863,
+      "step": 14
+    },
+    {
+      "epoch": 0.0012846865364850976,
+      "grad_norm": 0.7573930025100708,
+      "learning_rate": 0.0001,
+      "loss": 1.0136,
+      "step": 15
+    },
+    {
+      "epoch": 0.0013703323055841042,
+      "grad_norm": 0.8296635746955872,
+      "learning_rate": 0.0001,
+      "loss": 0.8687,
+      "step": 16
+    },
+    {
+      "epoch": 0.0014559780746831107,
+      "grad_norm": 0.3881508409976959,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 17
+    },
+    {
+      "epoch": 0.001541623843782117,
+      "grad_norm": 1.3019808530807495,
+      "learning_rate": 0.0001,
+      "loss": 0.7283,
+      "step": 18
+    },
+    {
+      "epoch": 0.0016272696128811237,
+      "grad_norm": 0.4842688739299774,
+      "learning_rate": 0.0001,
+      "loss": 0.8046,
+      "step": 19
+    },
+    {
+      "epoch": 0.0017129153819801302,
+      "grad_norm": 0.42655014991760254,
+      "learning_rate": 0.0001,
+      "loss": 1.0338,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017985611510791368,
+      "grad_norm": 0.7510253190994263,
+      "learning_rate": 0.0001,
+      "loss": 0.9,
+      "step": 21
+    },
+    {
+      "epoch": 0.0018842069201781431,
+      "grad_norm": 0.9847319722175598,
+      "learning_rate": 0.0001,
+      "loss": 1.1499,
+      "step": 22
+    },
+    {
+      "epoch": 0.00196985268927715,
+      "grad_norm": 1.018487811088562,
+      "learning_rate": 0.0001,
+      "loss": 1.1062,
+      "step": 23
+    },
+    {
+      "epoch": 0.0020554984583761563,
+      "grad_norm": 0.9762689471244812,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 24
+    },
+    {
+      "epoch": 0.0021411442274751626,
+      "grad_norm": 0.7377593517303467,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 25
+    },
+    {
+      "epoch": 0.0022267899965741694,
+      "grad_norm": 0.4469539523124695,
+      "learning_rate": 0.0001,
+      "loss": 1.109,
+      "step": 26
+    },
+    {
+      "epoch": 0.0023124357656731757,
+      "grad_norm": 0.9055679440498352,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 27
+    },
+    {
+      "epoch": 0.002398081534772182,
+      "grad_norm": 0.797278642654419,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 28
+    },
+    {
+      "epoch": 0.002483727303871189,
+      "grad_norm": 0.8931136131286621,
+      "learning_rate": 0.0001,
+      "loss": 1.3674,
+      "step": 29
+    },
+    {
+      "epoch": 0.0025693730729701952,
+      "grad_norm": 0.47705361247062683,
+      "learning_rate": 0.0001,
+      "loss": 0.9241,
+      "step": 30
+    },
+    {
+      "epoch": 0.0026550188420692016,
+      "grad_norm": 0.7573657035827637,
+      "learning_rate": 0.0001,
+      "loss": 0.9897,
+      "step": 31
+    },
+    {
+      "epoch": 0.0027406646111682084,
+      "grad_norm": 0.5225767493247986,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 32
+    },
+    {
+      "epoch": 0.0028263103802672147,
+      "grad_norm": 0.6606186628341675,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 33
+    },
+    {
+      "epoch": 0.0029119561493662215,
+      "grad_norm": 0.49695804715156555,
+      "learning_rate": 0.0001,
+      "loss": 1.0039,
+      "step": 34
+    },
+    {
+      "epoch": 0.002997601918465228,
+      "grad_norm": 0.5895708203315735,
+      "learning_rate": 0.0001,
+      "loss": 0.6813,
+      "step": 35
+    },
+    {
+      "epoch": 0.003083247687564234,
+      "grad_norm": 0.42022380232810974,
+      "learning_rate": 0.0001,
+      "loss": 1.0119,
+      "step": 36
+    },
+    {
+      "epoch": 0.003168893456663241,
+      "grad_norm": 0.788876473903656,
+      "learning_rate": 0.0001,
+      "loss": 0.8857,
+      "step": 37
+    },
+    {
+      "epoch": 0.0032545392257622473,
+      "grad_norm": 0.8509645462036133,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 38
+    },
+    {
+      "epoch": 0.0033401849948612537,
+      "grad_norm": 0.4381870925426483,
+      "learning_rate": 0.0001,
+      "loss": 0.892,
+      "step": 39
+    },
+    {
+      "epoch": 0.0034258307639602604,
+      "grad_norm": 0.3514087200164795,
+      "learning_rate": 0.0001,
+      "loss": 0.9203,
+      "step": 40
+    },
+    {
+      "epoch": 0.003511476533059267,
+      "grad_norm": 0.80808424949646,
+      "learning_rate": 0.0001,
+      "loss": 0.683,
+      "step": 41
+    },
+    {
+      "epoch": 0.0035971223021582736,
+      "grad_norm": 0.4480339586734772,
+      "learning_rate": 0.0001,
+      "loss": 0.7667,
+      "step": 42
+    },
+    {
+      "epoch": 0.00368276807125728,
+      "grad_norm": 0.8001617193222046,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 43
+    },
+    {
+      "epoch": 0.0037684138403562863,
+      "grad_norm": 0.7130466103553772,
+      "learning_rate": 0.0001,
+      "loss": 1.3825,
+      "step": 44
+    },
+    {
+      "epoch": 0.003854059609455293,
+      "grad_norm": 0.4300091564655304,
+      "learning_rate": 0.0001,
+      "loss": 1.3337,
+      "step": 45
+    },
+    {
+      "epoch": 0.0039397053785543,
+      "grad_norm": 0.35683491826057434,
+      "learning_rate": 0.0001,
+      "loss": 1.0643,
+      "step": 46
+    },
+    {
+      "epoch": 0.004025351147653306,
+      "grad_norm": 0.30761009454727173,
+      "learning_rate": 0.0001,
+      "loss": 0.3752,
+      "step": 47
+    },
+    {
+      "epoch": 0.0041109969167523125,
+      "grad_norm": 0.3835221230983734,
+      "learning_rate": 0.0001,
+      "loss": 1.0807,
+      "step": 48
+    },
+    {
+      "epoch": 0.004196642685851319,
+      "grad_norm": 0.5990667939186096,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 49
+    },
+    {
+      "epoch": 0.004282288454950325,
+      "grad_norm": 0.3940849006175995,
+      "learning_rate": 0.0001,
+      "loss": 1.1553,
+      "step": 50
+    },
+    {
+      "epoch": 0.004367934224049332,
+      "grad_norm": 0.7021299600601196,
+      "learning_rate": 0.0001,
+      "loss": 1.0361,
+      "step": 51
+    },
+    {
+      "epoch": 0.004453579993148339,
+      "grad_norm": 0.7237667441368103,
+      "learning_rate": 0.0001,
+      "loss": 1.1509,
+      "step": 52
+    },
+    {
+      "epoch": 0.004539225762247345,
+      "grad_norm": 0.5876308679580688,
+      "learning_rate": 0.0001,
+      "loss": 1.1973,
+      "step": 53
+    },
+    {
+      "epoch": 0.0046248715313463515,
+      "grad_norm": 0.40785667300224304,
+      "learning_rate": 0.0001,
+      "loss": 1.0063,
+      "step": 54
+    },
+    {
+      "epoch": 0.004710517300445358,
+      "grad_norm": 0.4537876844406128,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 55
+    },
+    {
+      "epoch": 0.004796163069544364,
+      "grad_norm": 0.7787575125694275,
+      "learning_rate": 0.0001,
+      "loss": 0.9507,
+      "step": 56
+    },
+    {
+      "epoch": 0.004881808838643371,
+      "grad_norm": 0.8675814270973206,
+      "learning_rate": 0.0001,
+      "loss": 1.0,
+      "step": 57
+    },
+    {
+      "epoch": 0.004967454607742378,
+      "grad_norm": 0.5529398918151855,
+      "learning_rate": 0.0001,
+      "loss": 0.8489,
+      "step": 58
+    },
+    {
+      "epoch": 0.005053100376841384,
+      "grad_norm": 1.0773961544036865,
+      "learning_rate": 0.0001,
+      "loss": 1.1833,
+      "step": 59
+    },
+    {
+      "epoch": 0.0051387461459403904,
+      "grad_norm": 0.7752442359924316,
+      "learning_rate": 0.0001,
+      "loss": 1.1044,
+      "step": 60
+    },
+    {
+      "epoch": 0.005224391915039397,
+      "grad_norm": 0.42013096809387207,
+      "learning_rate": 0.0001,
+      "loss": 0.8978,
+      "step": 61
+    },
+    {
+      "epoch": 0.005310037684138403,
+      "grad_norm": 0.3914511799812317,
+      "learning_rate": 0.0001,
+      "loss": 1.1049,
+      "step": 62
+    },
+    {
+      "epoch": 0.00539568345323741,
+      "grad_norm": 0.38735485076904297,
+      "learning_rate": 0.0001,
+      "loss": 0.7828,
+      "step": 63
+    },
+    {
+      "epoch": 0.005481329222336417,
+      "grad_norm": 0.4872879683971405,
+      "learning_rate": 0.0001,
+      "loss": 0.9119,
+      "step": 64
+    },
+    {
+      "epoch": 0.005566974991435423,
+      "grad_norm": 0.5608871579170227,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 65
+    },
+    {
+      "epoch": 0.005652620760534429,
+      "grad_norm": 0.7091379761695862,
+      "learning_rate": 0.0001,
+      "loss": 0.7954,
+      "step": 66
+    },
+    {
+      "epoch": 0.005738266529633436,
+      "grad_norm": 0.467446893453598,
+      "learning_rate": 0.0001,
+      "loss": 0.7632,
+      "step": 67
+    },
+    {
+      "epoch": 0.005823912298732443,
+      "grad_norm": 0.5544875860214233,
+      "learning_rate": 0.0001,
+      "loss": 0.4107,
+      "step": 68
+    },
+    {
+      "epoch": 0.005909558067831449,
+      "grad_norm": 0.7177286148071289,
+      "learning_rate": 0.0001,
+      "loss": 0.8183,
+      "step": 69
+    },
+    {
+      "epoch": 0.005995203836930456,
+      "grad_norm": 0.9233511686325073,
+      "learning_rate": 0.0001,
+      "loss": 1.4599,
+      "step": 70
+    },
+    {
+      "epoch": 0.006080849606029462,
+      "grad_norm": 0.3937545418739319,
+      "learning_rate": 0.0001,
+      "loss": 0.8716,
+      "step": 71
+    },
+    {
+      "epoch": 0.006166495375128468,
+      "grad_norm": 0.744172990322113,
+      "learning_rate": 0.0001,
+      "loss": 0.9806,
+      "step": 72
+    },
+    {
+      "epoch": 0.006252141144227476,
+      "grad_norm": 0.45919060707092285,
+      "learning_rate": 0.0001,
+      "loss": 0.754,
+      "step": 73
+    },
+    {
+      "epoch": 0.006337786913326482,
+      "grad_norm": 0.34135982394218445,
+      "learning_rate": 0.0001,
+      "loss": 1.2353,
+      "step": 74
+    },
+    {
+      "epoch": 0.006423432682425488,
+      "grad_norm": 0.5846899747848511,
+      "learning_rate": 0.0001,
+      "loss": 1.1905,
+      "step": 75
+    },
+    {
+      "epoch": 0.006509078451524495,
+      "grad_norm": 0.40213146805763245,
+      "learning_rate": 0.0001,
+      "loss": 1.1718,
+      "step": 76
+    },
+    {
+      "epoch": 0.006594724220623501,
+      "grad_norm": 0.7563181519508362,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 77
+    },
+    {
+      "epoch": 0.006680369989722507,
+      "grad_norm": 0.661890983581543,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 78
+    },
+    {
+      "epoch": 0.0067660157588215145,
+      "grad_norm": 0.34491482377052307,
+      "learning_rate": 0.0001,
+      "loss": 0.9082,
+      "step": 79
+    },
+    {
+      "epoch": 0.006851661527920521,
+      "grad_norm": 0.3401099443435669,
+      "learning_rate": 0.0001,
+      "loss": 1.217,
+      "step": 80
+    },
+    {
+      "epoch": 0.006937307297019527,
+      "grad_norm": 0.873363733291626,
+      "learning_rate": 0.0001,
+      "loss": 0.8439,
+      "step": 81
+    },
+    {
+      "epoch": 0.007022953066118534,
+      "grad_norm": 0.38689032196998596,
+      "learning_rate": 0.0001,
+      "loss": 1.2645,
+      "step": 82
+    },
+    {
+      "epoch": 0.00710859883521754,
+      "grad_norm": 0.5254368782043457,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 83
+    },
+    {
+      "epoch": 0.007194244604316547,
+      "grad_norm": 0.7664945721626282,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 84
+    },
+    {
+      "epoch": 0.0072798903734155535,
+      "grad_norm": 0.8685832619667053,
+      "learning_rate": 0.0001,
+      "loss": 1.1869,
+      "step": 85
+    },
+    {
+      "epoch": 0.00736553614251456,
+      "grad_norm": 0.44120535254478455,
+      "learning_rate": 0.0001,
+      "loss": 1.2131,
+      "step": 86
+    },
+    {
+      "epoch": 0.007451181911613566,
+      "grad_norm": 0.5277093052864075,
+      "learning_rate": 0.0001,
+      "loss": 1.1586,
+      "step": 87
+    },
+    {
+      "epoch": 0.0075368276807125725,
+      "grad_norm": 0.3953602910041809,
+      "learning_rate": 0.0001,
+      "loss": 0.6948,
+      "step": 88
+    },
+    {
+      "epoch": 0.007622473449811579,
+      "grad_norm": 0.46623554825782776,
+      "learning_rate": 0.0001,
+      "loss": 1.0533,
+      "step": 89
+    },
+    {
+      "epoch": 0.007708119218910586,
+      "grad_norm": 0.367775022983551,
+      "learning_rate": 0.0001,
+      "loss": 0.9576,
+      "step": 90
+    },
+    {
+      "epoch": 0.0077937649880095924,
+      "grad_norm": 0.38137781620025635,
+      "learning_rate": 0.0001,
+      "loss": 0.3673,
+      "step": 91
+    },
+    {
+      "epoch": 0.0078794107571086,
+      "grad_norm": 0.5191543698310852,
+      "learning_rate": 0.0001,
+      "loss": 0.5933,
+      "step": 92
+    },
+    {
+      "epoch": 0.007965056526207605,
+      "grad_norm": 0.6034082174301147,
+      "learning_rate": 0.0001,
+      "loss": 0.7739,
+      "step": 93
+    },
+    {
+      "epoch": 0.008050702295306612,
+      "grad_norm": 0.26119935512542725,
+      "learning_rate": 0.0001,
+      "loss": 0.4609,
+      "step": 94
+    },
+    {
+      "epoch": 0.008136348064405618,
+      "grad_norm": 0.7898549437522888,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 95
+    },
+    {
+      "epoch": 0.008221993833504625,
+      "grad_norm": 0.7954878807067871,
+      "learning_rate": 0.0001,
+      "loss": 0.9567,
+      "step": 96
+    },
+    {
+      "epoch": 0.00830763960260363,
+      "grad_norm": 0.5140252709388733,
+      "learning_rate": 0.0001,
+      "loss": 0.8883,
+      "step": 97
+    },
+    {
+      "epoch": 0.008393285371702638,
+      "grad_norm": 0.6057766675949097,
+      "learning_rate": 0.0001,
+      "loss": 1.3298,
+      "step": 98
+    },
+    {
+      "epoch": 0.008478931140801645,
+      "grad_norm": 1.0262789726257324,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 99
+    },
+    {
+      "epoch": 0.00856457690990065,
+      "grad_norm": 0.7110846042633057,
+      "learning_rate": 0.0001,
+      "loss": 0.9052,
+      "step": 100
+    },
+    {
+      "epoch": 0.008650222678999658,
+      "grad_norm": 0.5379086136817932,
+      "learning_rate": 0.0001,
+      "loss": 1.0668,
+      "step": 101
+    },
+    {
+      "epoch": 0.008735868448098663,
+      "grad_norm": 0.4733043611049652,
+      "learning_rate": 0.0001,
+      "loss": 1.1722,
+      "step": 102
+    },
+    {
+      "epoch": 0.00882151421719767,
+      "grad_norm": 0.5439091324806213,
+      "learning_rate": 0.0001,
+      "loss": 0.8257,
+      "step": 103
+    },
+    {
+      "epoch": 0.008907159986296678,
+      "grad_norm": 0.2571442723274231,
+      "learning_rate": 0.0001,
+      "loss": 0.4486,
+      "step": 104
+    },
+    {
+      "epoch": 0.008992805755395683,
+      "grad_norm": 0.8616091012954712,
+      "learning_rate": 0.0001,
+      "loss": 1.0288,
+      "step": 105
+    },
+    {
+      "epoch": 0.00907845152449469,
+      "grad_norm": 0.6443458795547485,
+      "learning_rate": 0.0001,
+      "loss": 0.6632,
+      "step": 106
+    },
+    {
+      "epoch": 0.009164097293593696,
+      "grad_norm": 0.46296751499176025,
+      "learning_rate": 0.0001,
+      "loss": 1.0785,
+      "step": 107
+    },
+    {
+      "epoch": 0.009249743062692703,
+      "grad_norm": 0.7110704779624939,
+      "learning_rate": 0.0001,
+      "loss": 1.173,
+      "step": 108
+    },
+    {
+      "epoch": 0.00933538883179171,
+      "grad_norm": 0.36797258257865906,
+      "learning_rate": 0.0001,
+      "loss": 0.5382,
+      "step": 109
+    },
+    {
+      "epoch": 0.009421034600890716,
+      "grad_norm": 0.6590583920478821,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 110
+    },
+    {
+      "epoch": 0.009506680369989723,
+      "grad_norm": 0.716927707195282,
+      "learning_rate": 0.0001,
+      "loss": 1.1377,
+      "step": 111
+    },
+    {
+      "epoch": 0.009592326139088728,
+      "grad_norm": 0.642394483089447,
+      "learning_rate": 0.0001,
+      "loss": 1.9344,
+      "step": 112
+    },
+    {
+      "epoch": 0.009677971908187736,
+      "grad_norm": 0.5497715473175049,
+      "learning_rate": 0.0001,
+      "loss": 1.231,
+      "step": 113
+    },
+    {
+      "epoch": 0.009763617677286743,
+      "grad_norm": 0.3022732734680176,
+      "learning_rate": 0.0001,
+      "loss": 0.598,
+      "step": 114
+    },
+    {
+      "epoch": 0.009849263446385748,
+      "grad_norm": 0.5338352918624878,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 115
+    },
+    {
+      "epoch": 0.009934909215484755,
+      "grad_norm": 0.3874720633029938,
+      "learning_rate": 0.0001,
+      "loss": 1.2088,
+      "step": 116
+    },
+    {
+      "epoch": 0.010020554984583761,
+      "grad_norm": 0.5399787425994873,
+      "learning_rate": 0.0001,
+      "loss": 1.1128,
+      "step": 117
+    },
+    {
+      "epoch": 0.010106200753682768,
+      "grad_norm": 0.29917436838150024,
+      "learning_rate": 0.0001,
+      "loss": 0.869,
+      "step": 118
+    },
+    {
+      "epoch": 0.010191846522781775,
+      "grad_norm": 0.3607337772846222,
+      "learning_rate": 0.0001,
+      "loss": 1.0886,
+      "step": 119
+    },
+    {
+      "epoch": 0.010277492291880781,
+      "grad_norm": 0.831825315952301,
+      "learning_rate": 0.0001,
+      "loss": 1.2983,
+      "step": 120
+    },
+    {
+      "epoch": 0.010363138060979788,
+      "grad_norm": 0.3613249957561493,
+      "learning_rate": 0.0001,
+      "loss": 0.7668,
+      "step": 121
+    },
+    {
+      "epoch": 0.010448783830078794,
+      "grad_norm": 0.4138905107975006,
+      "learning_rate": 0.0001,
+      "loss": 0.5799,
+      "step": 122
+    },
+    {
+      "epoch": 0.0105344295991778,
+      "grad_norm": 0.8549215197563171,
+      "learning_rate": 0.0001,
+      "loss": 1.0307,
+      "step": 123
+    },
+    {
+      "epoch": 0.010620075368276806,
+      "grad_norm": 0.775458037853241,
+      "learning_rate": 0.0001,
+      "loss": 0.8462,
+      "step": 124
+    },
+    {
+      "epoch": 0.010705721137375813,
+      "grad_norm": 0.7238227128982544,
+      "learning_rate": 0.0001,
+      "loss": 1.0358,
+      "step": 125
+    },
+    {
+      "epoch": 0.01079136690647482,
+      "grad_norm": 0.648390531539917,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 126
+    },
+    {
+      "epoch": 0.010877012675573826,
+      "grad_norm": 0.8453584909439087,
+      "learning_rate": 0.0001,
+      "loss": 1.143,
+      "step": 127
+    },
+    {
+      "epoch": 0.010962658444672833,
+      "grad_norm": 0.5273696780204773,
+      "learning_rate": 0.0001,
+      "loss": 1.1174,
+      "step": 128
+    },
+    {
+      "epoch": 0.011048304213771839,
+      "grad_norm": 0.2049604207277298,
+      "learning_rate": 0.0001,
+      "loss": 0.7414,
+      "step": 129
+    },
+    {
+      "epoch": 0.011133949982870846,
+      "grad_norm": 0.43656739592552185,
+      "learning_rate": 0.0001,
+      "loss": 1.2888,
+      "step": 130
+    },
+    {
+      "epoch": 0.011219595751969853,
+      "grad_norm": 0.2539294362068176,
+      "learning_rate": 0.0001,
+      "loss": 0.2864,
+      "step": 131
+    },
+    {
+      "epoch": 0.011305241521068859,
+      "grad_norm": 0.6131764650344849,
+      "learning_rate": 0.0001,
+      "loss": 0.8357,
+      "step": 132
+    },
+    {
+      "epoch": 0.011390887290167866,
+      "grad_norm": 0.5654326677322388,
+      "learning_rate": 0.0001,
+      "loss": 0.6084,
+      "step": 133
+    },
+    {
+      "epoch": 0.011476533059266871,
+      "grad_norm": 0.5491416454315186,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 134
+    },
+    {
+      "epoch": 0.011562178828365879,
+      "grad_norm": 0.5856392979621887,
+      "learning_rate": 0.0001,
+      "loss": 0.7991,
+      "step": 135
+    },
+    {
+      "epoch": 0.011647824597464886,
+      "grad_norm": 0.20719201862812042,
+      "learning_rate": 0.0001,
+      "loss": 0.4665,
+      "step": 136
+    },
+    {
+      "epoch": 0.011733470366563891,
+      "grad_norm": 0.5941038727760315,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 137
+    },
+    {
+      "epoch": 0.011819116135662899,
+      "grad_norm": 0.6062179803848267,
+      "learning_rate": 0.0001,
+      "loss": 0.6045,
+      "step": 138
+    },
+    {
+      "epoch": 0.011904761904761904,
+      "grad_norm": 0.4582969844341278,
+      "learning_rate": 0.0001,
+      "loss": 0.5921,
+      "step": 139
+    },
+    {
+      "epoch": 0.011990407673860911,
+      "grad_norm": 0.356458842754364,
+      "learning_rate": 0.0001,
+      "loss": 0.8571,
+      "step": 140
+    },
+    {
+      "epoch": 0.012076053442959919,
+      "grad_norm": 0.7242527604103088,
+      "learning_rate": 0.0001,
+      "loss": 0.8083,
+      "step": 141
+    },
+    {
+      "epoch": 0.012161699212058924,
+      "grad_norm": 0.5755905508995056,
+      "learning_rate": 0.0001,
+      "loss": 0.6744,
+      "step": 142
+    },
+    {
+      "epoch": 0.012247344981157931,
+      "grad_norm": 0.491874635219574,
+      "learning_rate": 0.0001,
+      "loss": 1.0257,
+      "step": 143
+    },
+    {
+      "epoch": 0.012332990750256937,
+      "grad_norm": 0.6088501811027527,
+      "learning_rate": 0.0001,
+      "loss": 0.898,
+      "step": 144
+    },
+    {
+      "epoch": 0.012418636519355944,
+      "grad_norm": 0.46991458535194397,
+      "learning_rate": 0.0001,
+      "loss": 1.0548,
+      "step": 145
+    },
+    {
+      "epoch": 0.012504282288454951,
+      "grad_norm": 0.751399040222168,
+      "learning_rate": 0.0001,
+      "loss": 1.0215,
+      "step": 146
+    },
+    {
+      "epoch": 0.012589928057553957,
+      "grad_norm": 0.5059388279914856,
+      "learning_rate": 0.0001,
+      "loss": 0.6213,
+      "step": 147
+    },
+    {
+      "epoch": 0.012675573826652964,
+      "grad_norm": 0.6079367995262146,
+      "learning_rate": 0.0001,
+      "loss": 1.0706,
+      "step": 148
+    },
+    {
+      "epoch": 0.01276121959575197,
+      "grad_norm": 0.6798315644264221,
+      "learning_rate": 0.0001,
+      "loss": 1.0163,
+      "step": 149
+    },
+    {
+      "epoch": 0.012846865364850977,
+      "grad_norm": 0.9851637482643127,
+      "learning_rate": 0.0001,
+      "loss": 0.9106,
+      "step": 150
+    },
+    {
+      "epoch": 0.012932511133949982,
+      "grad_norm": 0.3938959836959839,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 151
+    },
+    {
+      "epoch": 0.01301815690304899,
+      "grad_norm": 0.8171619772911072,
+      "learning_rate": 0.0001,
+      "loss": 1.4108,
+      "step": 152
+    },
+    {
+      "epoch": 0.013103802672147996,
+      "grad_norm": 0.44686540961265564,
+      "learning_rate": 0.0001,
+      "loss": 1.0186,
+      "step": 153
+    },
+    {
+      "epoch": 0.013189448441247002,
+      "grad_norm": 0.46538734436035156,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 154
+    },
+    {
+      "epoch": 0.01327509421034601,
+      "grad_norm": 0.6820229291915894,
+      "learning_rate": 0.0001,
+      "loss": 0.8673,
+      "step": 155
+    },
+    {
+      "epoch": 0.013360739979445015,
+      "grad_norm": 0.3036135733127594,
+      "learning_rate": 0.0001,
+      "loss": 0.5743,
+      "step": 156
+    },
+    {
+      "epoch": 0.013446385748544022,
+      "grad_norm": 0.605675995349884,
+      "learning_rate": 0.0001,
+      "loss": 0.5587,
+      "step": 157
+    },
+    {
+      "epoch": 0.013532031517643029,
+      "grad_norm": 0.4226544499397278,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 158
+    },
+    {
+      "epoch": 0.013617677286742035,
+      "grad_norm": 0.7682145237922668,
+      "learning_rate": 0.0001,
+      "loss": 1.3029,
+      "step": 159
+    },
+    {
+      "epoch": 0.013703323055841042,
+      "grad_norm": 0.3437565863132477,
+      "learning_rate": 0.0001,
+      "loss": 1.0674,
+      "step": 160
+    },
+    {
+      "epoch": 0.013788968824940047,
+      "grad_norm": 0.4411551058292389,
+      "learning_rate": 0.0001,
+      "loss": 0.7153,
+      "step": 161
+    },
+    {
+      "epoch": 0.013874614594039054,
+      "grad_norm": 0.08874484151601791,
+      "learning_rate": 0.0001,
+      "loss": 0.4686,
+      "step": 162
+    },
+    {
+      "epoch": 0.013960260363138062,
+      "grad_norm": 0.839324951171875,
+      "learning_rate": 0.0001,
+      "loss": 1.0394,
+      "step": 163
+    },
+    {
+      "epoch": 0.014045906132237067,
+      "grad_norm": 0.5573042035102844,
+      "learning_rate": 0.0001,
+      "loss": 0.4485,
+      "step": 164
+    },
+    {
+      "epoch": 0.014131551901336074,
+      "grad_norm": 0.39185696840286255,
+      "learning_rate": 0.0001,
+      "loss": 1.1059,
+      "step": 165
+    },
+    {
+      "epoch": 0.01421719767043508,
+      "grad_norm": 0.8128798604011536,
+      "learning_rate": 0.0001,
+      "loss": 1.2481,
+      "step": 166
+    },
+    {
+      "epoch": 0.014302843439534087,
+      "grad_norm": 0.3043704628944397,
+      "learning_rate": 0.0001,
+      "loss": 1.246,
+      "step": 167
+    },
+    {
+      "epoch": 0.014388489208633094,
+      "grad_norm": 0.5417680144309998,
+      "learning_rate": 0.0001,
+      "loss": 0.3517,
+      "step": 168
+    },
+    {
+      "epoch": 0.0144741349777321,
+      "grad_norm": 0.7914559841156006,
+      "learning_rate": 0.0001,
+      "loss": 0.7708,
+      "step": 169
+    },
+    {
+      "epoch": 0.014559780746831107,
+      "grad_norm": 0.44303083419799805,
+      "learning_rate": 0.0001,
+      "loss": 0.9307,
+      "step": 170
+    },
+    {
+      "epoch": 0.014645426515930112,
+      "grad_norm": 0.9249691963195801,
+      "learning_rate": 0.0001,
+      "loss": 0.9611,
+      "step": 171
+    },
+    {
+      "epoch": 0.01473107228502912,
+      "grad_norm": 0.2054964154958725,
+      "learning_rate": 0.0001,
+      "loss": 0.3366,
+      "step": 172
+    },
+    {
+      "epoch": 0.014816718054128127,
+      "grad_norm": 0.4519209563732147,
+      "learning_rate": 0.0001,
+      "loss": 1.0818,
+      "step": 173
+    },
+    {
+      "epoch": 0.014902363823227132,
+      "grad_norm": 0.7107318639755249,
+      "learning_rate": 0.0001,
+      "loss": 1.294,
+      "step": 174
+    },
+    {
+      "epoch": 0.01498800959232614,
+      "grad_norm": 0.23641115427017212,
+      "learning_rate": 0.0001,
+      "loss": 0.2355,
+      "step": 175
+    },
+    {
+      "epoch": 0.015073655361425145,
+      "grad_norm": 0.2915380001068115,
+      "learning_rate": 0.0001,
+      "loss": 0.8061,
+      "step": 176
+    },
+    {
+      "epoch": 0.015159301130524152,
+      "grad_norm": 0.4831894338130951,
+      "learning_rate": 0.0001,
+      "loss": 0.8409,
+      "step": 177
+    },
+    {
+      "epoch": 0.015244946899623158,
+      "grad_norm": 1.4710631370544434,
+      "learning_rate": 0.0001,
+      "loss": 1.2793,
+      "step": 178
+    },
+    {
+      "epoch": 0.015330592668722165,
+      "grad_norm": 0.4858090579509735,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 179
+    },
+    {
+      "epoch": 0.015416238437821172,
+      "grad_norm": 0.33147555589675903,
+      "learning_rate": 0.0001,
+      "loss": 1.135,
+      "step": 180
+    },
+    {
+      "epoch": 0.015501884206920178,
+      "grad_norm": 0.582261323928833,
+      "learning_rate": 0.0001,
+      "loss": 0.6809,
+      "step": 181
+    },
+    {
+      "epoch": 0.015587529976019185,
+      "grad_norm": 0.7910903096199036,
+      "learning_rate": 0.0001,
+      "loss": 0.7857,
+      "step": 182
+    },
+    {
+      "epoch": 0.01567317574511819,
+      "grad_norm": 0.5223270058631897,
+      "learning_rate": 0.0001,
+      "loss": 0.7698,
+      "step": 183
+    },
+    {
+      "epoch": 0.0157588215142172,
+      "grad_norm": 0.5761642456054688,
+      "learning_rate": 0.0001,
+      "loss": 0.9095,
+      "step": 184
+    },
+    {
+      "epoch": 0.015844467283316205,
+      "grad_norm": 0.37717190384864807,
+      "learning_rate": 0.0001,
+      "loss": 1.2333,
+      "step": 185
+    },
+    {
+      "epoch": 0.01593011305241521,
+      "grad_norm": 0.4468681514263153,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 186
+    },
+    {
+      "epoch": 0.016015758821514216,
+      "grad_norm": 0.7254468202590942,
+      "learning_rate": 0.0001,
+      "loss": 1.049,
+      "step": 187
+    },
+    {
+      "epoch": 0.016101404590613225,
+      "grad_norm": 0.8637517690658569,
+      "learning_rate": 0.0001,
+      "loss": 1.1828,
+      "step": 188
+    },
+    {
+      "epoch": 0.01618705035971223,
+      "grad_norm": 0.5594127774238586,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 189
+    },
+    {
+      "epoch": 0.016272696128811236,
+      "grad_norm": 0.5686207413673401,
+      "learning_rate": 0.0001,
+      "loss": 0.5952,
+      "step": 190
+    },
+    {
+      "epoch": 0.016358341897910245,
+      "grad_norm": 0.5976225733757019,
+      "learning_rate": 0.0001,
+      "loss": 0.5878,
+      "step": 191
+    },
+    {
+      "epoch": 0.01644398766700925,
+      "grad_norm": 0.47808587551116943,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 192
+    },
+    {
+      "epoch": 0.016529633436108256,
+      "grad_norm": 0.33799299597740173,
+      "learning_rate": 0.0001,
+      "loss": 1.0913,
+      "step": 193
+    },
+    {
+      "epoch": 0.01661527920520726,
+      "grad_norm": 0.7074211835861206,
+      "learning_rate": 0.0001,
+      "loss": 1.1447,
+      "step": 194
+    },
+    {
+      "epoch": 0.01670092497430627,
+      "grad_norm": 0.539174497127533,
+      "learning_rate": 0.0001,
+      "loss": 1.3739,
+      "step": 195
+    },
+    {
+      "epoch": 0.016786570743405275,
+      "grad_norm": 0.4937298893928528,
+      "learning_rate": 0.0001,
+      "loss": 0.7595,
+      "step": 196
+    },
+    {
+      "epoch": 0.01687221651250428,
+      "grad_norm": 0.5498679280281067,
+      "learning_rate": 0.0001,
+      "loss": 0.5036,
+      "step": 197
+    },
+    {
+      "epoch": 0.01695786228160329,
+      "grad_norm": 0.39891958236694336,
+      "learning_rate": 0.0001,
+      "loss": 0.9622,
+      "step": 198
+    },
+    {
+      "epoch": 0.017043508050702295,
+      "grad_norm": 0.3063664436340332,
+      "learning_rate": 0.0001,
+      "loss": 1.0043,
+      "step": 199
+    },
+    {
+      "epoch": 0.0171291538198013,
+      "grad_norm": 0.5752357840538025,
+      "learning_rate": 0.0001,
+      "loss": 1.1121,
+      "step": 200
+    },
+    {
+      "epoch": 0.01721479958890031,
+      "grad_norm": 0.37041792273521423,
+      "learning_rate": 0.0001,
+      "loss": 0.8652,
+      "step": 201
+    },
+    {
+      "epoch": 0.017300445357999315,
+      "grad_norm": 0.4936182200908661,
+      "learning_rate": 0.0001,
+      "loss": 0.796,
+      "step": 202
+    },
+    {
+      "epoch": 0.01738609112709832,
+      "grad_norm": 0.7927535772323608,
+      "learning_rate": 0.0001,
+      "loss": 0.8683,
+      "step": 203
+    },
+    {
+      "epoch": 0.017471736896197326,
+      "grad_norm": 0.6254543662071228,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 204
+    },
+    {
+      "epoch": 0.017557382665296335,
+      "grad_norm": 0.47740989923477173,
+      "learning_rate": 0.0001,
+      "loss": 0.8543,
+      "step": 205
+    },
+    {
+      "epoch": 0.01764302843439534,
+      "grad_norm": 0.29157665371894836,
+      "learning_rate": 0.0001,
+      "loss": 0.5248,
+      "step": 206
+    },
+    {
+      "epoch": 0.017728674203494346,
+      "grad_norm": 0.5519976615905762,
+      "learning_rate": 0.0001,
+      "loss": 0.6298,
+      "step": 207
+    },
+    {
+      "epoch": 0.017814319972593355,
+      "grad_norm": 0.5171421766281128,
+      "learning_rate": 0.0001,
+      "loss": 0.5369,
+      "step": 208
+    },
+    {
+      "epoch": 0.01789996574169236,
+      "grad_norm": 0.8833096027374268,
+      "learning_rate": 0.0001,
+      "loss": 1.1489,
+      "step": 209
+    },
+    {
+      "epoch": 0.017985611510791366,
+      "grad_norm": 0.7240824699401855,
+      "learning_rate": 0.0001,
+      "loss": 1.2896,
+      "step": 210
+    },
+    {
+      "epoch": 0.018071257279890375,
+      "grad_norm": 0.6497964262962341,
+      "learning_rate": 0.0001,
+      "loss": 0.931,
+      "step": 211
+    },
+    {
+      "epoch": 0.01815690304898938,
+      "grad_norm": 0.4419800341129303,
+      "learning_rate": 0.0001,
+      "loss": 1.0824,
+      "step": 212
+    },
+    {
+      "epoch": 0.018242548818088386,
+      "grad_norm": 0.6659957766532898,
+      "learning_rate": 0.0001,
+      "loss": 1.013,
+      "step": 213
+    },
+    {
+      "epoch": 0.01832819458718739,
+      "grad_norm": 0.5790843963623047,
+      "learning_rate": 0.0001,
+      "loss": 1.0472,
+      "step": 214
+    },
+    {
+      "epoch": 0.0184138403562864,
+      "grad_norm": 0.4730265438556671,
+      "learning_rate": 0.0001,
+      "loss": 0.4081,
+      "step": 215
+    },
+    {
+      "epoch": 0.018499486125385406,
+      "grad_norm": 0.5624915361404419,
+      "learning_rate": 0.0001,
+      "loss": 1.2608,
+      "step": 216
+    },
+    {
+      "epoch": 0.01858513189448441,
+      "grad_norm": 0.8119117021560669,
+      "learning_rate": 0.0001,
+      "loss": 1.2179,
+      "step": 217
+    },
+    {
+      "epoch": 0.01867077766358342,
+      "grad_norm": 0.6666969060897827,
+      "learning_rate": 0.0001,
+      "loss": 0.8262,
+      "step": 218
+    },
+    {
+      "epoch": 0.018756423432682426,
+      "grad_norm": 0.7499668002128601,
+      "learning_rate": 0.0001,
+      "loss": 0.9745,
+      "step": 219
+    },
+    {
+      "epoch": 0.01884206920178143,
+      "grad_norm": 0.515470564365387,
+      "learning_rate": 0.0001,
+      "loss": 1.3259,
+      "step": 220
+    },
+    {
+      "epoch": 0.018927714970880437,
+      "grad_norm": 0.3271944522857666,
+      "learning_rate": 0.0001,
+      "loss": 0.8315,
+      "step": 221
+    },
+    {
+      "epoch": 0.019013360739979446,
+      "grad_norm": 0.3094387650489807,
+      "learning_rate": 0.0001,
+      "loss": 0.7416,
+      "step": 222
+    },
+    {
+      "epoch": 0.01909900650907845,
+      "grad_norm": 0.8694062232971191,
+      "learning_rate": 0.0001,
+      "loss": 1.094,
+      "step": 223
+    },
+    {
+      "epoch": 0.019184652278177457,
+      "grad_norm": 0.42380091547966003,
+      "learning_rate": 0.0001,
+      "loss": 0.5898,
+      "step": 224
+    },
+    {
+      "epoch": 0.019270298047276466,
+      "grad_norm": 0.4785110354423523,
+      "learning_rate": 0.0001,
+      "loss": 0.9793,
+      "step": 225
+    },
+    {
+      "epoch": 0.01935594381637547,
+      "grad_norm": 0.3493095338344574,
+      "learning_rate": 0.0001,
+      "loss": 0.9848,
+      "step": 226
+    },
+    {
+      "epoch": 0.019441589585474477,
+      "grad_norm": 1.0580689907073975,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 227
+    },
+    {
+      "epoch": 0.019527235354573486,
+      "grad_norm": 0.6997901201248169,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 228
+    },
+    {
+      "epoch": 0.01961288112367249,
+      "grad_norm": 0.4807130992412567,
+      "learning_rate": 0.0001,
+      "loss": 1.1388,
+      "step": 229
+    },
+    {
+      "epoch": 0.019698526892771497,
+      "grad_norm": 0.4109513759613037,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 230
+    },
+    {
+      "epoch": 0.019784172661870502,
+      "grad_norm": 1.2867282629013062,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 231
+    },
+    {
+      "epoch": 0.01986981843096951,
+      "grad_norm": 0.7769408822059631,
+      "learning_rate": 0.0001,
+      "loss": 0.7099,
+      "step": 232
+    },
+    {
+      "epoch": 0.019955464200068516,
+      "grad_norm": 0.6913193464279175,
+      "learning_rate": 0.0001,
+      "loss": 1.1492,
+      "step": 233
+    },
+    {
+      "epoch": 0.020041109969167522,
+      "grad_norm": 0.869472861289978,
+      "learning_rate": 0.0001,
+      "loss": 0.8328,
+      "step": 234
+    },
+    {
+      "epoch": 0.02012675573826653,
+      "grad_norm": 0.41049104928970337,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 235
+    },
+    {
+      "epoch": 0.020212401507365536,
+      "grad_norm": 0.41176638007164,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 236
+    },
+    {
+      "epoch": 0.020298047276464542,
+      "grad_norm": 0.407853364944458,
+      "learning_rate": 0.0001,
+      "loss": 0.5924,
+      "step": 237
+    },
+    {
+      "epoch": 0.02038369304556355,
+      "grad_norm": 0.39904913306236267,
+      "learning_rate": 0.0001,
+      "loss": 0.5724,
+      "step": 238
+    },
+    {
+      "epoch": 0.020469338814662556,
+      "grad_norm": 0.45222654938697815,
+      "learning_rate": 0.0001,
+      "loss": 1.179,
+      "step": 239
+    },
+    {
+      "epoch": 0.020554984583761562,
+      "grad_norm": 0.6465727686882019,
+      "learning_rate": 0.0001,
+      "loss": 0.8601,
+      "step": 240
+    },
+    {
+      "epoch": 0.020640630352860567,
+      "grad_norm": 0.5312575697898865,
+      "learning_rate": 0.0001,
+      "loss": 0.8362,
+      "step": 241
+    },
+    {
+      "epoch": 0.020726276121959576,
+      "grad_norm": 0.3865196108818054,
+      "learning_rate": 0.0001,
+      "loss": 0.7759,
+      "step": 242
+    },
+    {
+      "epoch": 0.02081192189105858,
+      "grad_norm": 0.4300842583179474,
+      "learning_rate": 0.0001,
+      "loss": 1.1806,
+      "step": 243
+    },
+    {
+      "epoch": 0.020897567660157587,
+      "grad_norm": 0.3994327485561371,
+      "learning_rate": 0.0001,
+      "loss": 1.1467,
+      "step": 244
+    },
+    {
+      "epoch": 0.020983213429256596,
+      "grad_norm": 0.6881745457649231,
+      "learning_rate": 0.0001,
+      "loss": 1.3381,
+      "step": 245
+    },
+    {
+      "epoch": 0.0210688591983556,
+      "grad_norm": 0.8350423574447632,
+      "learning_rate": 0.0001,
+      "loss": 1.0344,
+      "step": 246
+    },
+    {
+      "epoch": 0.021154504967454607,
+      "grad_norm": 0.415759801864624,
+      "learning_rate": 0.0001,
+      "loss": 1.1478,
+      "step": 247
+    },
+    {
+      "epoch": 0.021240150736553613,
+      "grad_norm": 0.31463631987571716,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 248
+    },
+    {
+      "epoch": 0.02132579650565262,
+      "grad_norm": 0.7802398800849915,
+      "learning_rate": 0.0001,
+      "loss": 0.8622,
+      "step": 249
+    },
+    {
+      "epoch": 0.021411442274751627,
+      "grad_norm": 0.5020819306373596,
+      "learning_rate": 0.0001,
+      "loss": 1.2081,
+      "step": 250
+    },
+    {
+      "epoch": 0.021497088043850632,
+      "grad_norm": 0.6716035008430481,
+      "learning_rate": 0.0001,
+      "loss": 1.1046,
+      "step": 251
+    },
+    {
+      "epoch": 0.02158273381294964,
+      "grad_norm": 0.6020827889442444,
+      "learning_rate": 0.0001,
+      "loss": 0.8633,
+      "step": 252
+    },
+    {
+      "epoch": 0.021668379582048647,
+      "grad_norm": 0.5042803883552551,
+      "learning_rate": 0.0001,
+      "loss": 1.3067,
+      "step": 253
+    },
+    {
+      "epoch": 0.021754025351147652,
+      "grad_norm": 0.45201894640922546,
+      "learning_rate": 0.0001,
+      "loss": 1.2332,
+      "step": 254
+    },
+    {
+      "epoch": 0.02183967112024666,
+      "grad_norm": 0.5424230694770813,
+      "learning_rate": 0.0001,
+      "loss": 1.2226,
+      "step": 255
+    },
+    {
+      "epoch": 0.021925316889345667,
+      "grad_norm": 0.7786694169044495,
+      "learning_rate": 0.0001,
+      "loss": 0.9639,
+      "step": 256
+    },
+    {
+      "epoch": 0.022010962658444672,
+      "grad_norm": 0.644834041595459,
+      "learning_rate": 0.0001,
+      "loss": 0.6126,
+      "step": 257
+    },
+    {
+      "epoch": 0.022096608427543678,
+      "grad_norm": 0.504009485244751,
+      "learning_rate": 0.0001,
+      "loss": 1.2045,
+      "step": 258
+    },
+    {
+      "epoch": 0.022182254196642687,
+      "grad_norm": 0.5058109760284424,
+      "learning_rate": 0.0001,
+      "loss": 1.0075,
+      "step": 259
+    },
+    {
+      "epoch": 0.022267899965741692,
+      "grad_norm": 0.6785022616386414,
+      "learning_rate": 0.0001,
+      "loss": 0.86,
+      "step": 260
+    },
+    {
+      "epoch": 0.022353545734840698,
+      "grad_norm": 0.5328841209411621,
+      "learning_rate": 0.0001,
+      "loss": 1.2505,
+      "step": 261
+    },
+    {
+      "epoch": 0.022439191503939707,
+      "grad_norm": 0.5045756697654724,
+      "learning_rate": 0.0001,
+      "loss": 0.5717,
+      "step": 262
+    },
+    {
+      "epoch": 0.022524837273038712,
+      "grad_norm": 0.5665920376777649,
+      "learning_rate": 0.0001,
+      "loss": 0.9508,
+      "step": 263
+    },
+    {
+      "epoch": 0.022610483042137718,
+      "grad_norm": 0.35589876770973206,
+      "learning_rate": 0.0001,
+      "loss": 0.9058,
+      "step": 264
+    },
+    {
+      "epoch": 0.022696128811236727,
+      "grad_norm": 0.6246945858001709,
+      "learning_rate": 0.0001,
+      "loss": 1.3004,
+      "step": 265
+    },
+    {
+      "epoch": 0.022781774580335732,
+      "grad_norm": 0.2250419706106186,
+      "learning_rate": 0.0001,
+      "loss": 0.3967,
+      "step": 266
+    },
+    {
+      "epoch": 0.022867420349434738,
+      "grad_norm": 0.7356628179550171,
+      "learning_rate": 0.0001,
+      "loss": 1.1773,
+      "step": 267
+    },
+    {
+      "epoch": 0.022953066118533743,
+      "grad_norm": 1.3864586353302002,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 268
+    },
+    {
+      "epoch": 0.023038711887632752,
+      "grad_norm": 0.6121350526809692,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 269
+    },
+    {
+      "epoch": 0.023124357656731757,
+      "grad_norm": 0.501342236995697,
+      "learning_rate": 0.0001,
+      "loss": 1.0362,
+      "step": 270
+    },
+    {
+      "epoch": 0.023210003425830763,
+      "grad_norm": 0.43822377920150757,
+      "learning_rate": 0.0001,
+      "loss": 1.3093,
+      "step": 271
+    },
+    {
+      "epoch": 0.023295649194929772,
+      "grad_norm": 0.709287166595459,
+      "learning_rate": 0.0001,
+      "loss": 0.9299,
+      "step": 272
+    },
+    {
+      "epoch": 0.023381294964028777,
+      "grad_norm": 0.6833740472793579,
+      "learning_rate": 0.0001,
+      "loss": 1.156,
+      "step": 273
+    },
+    {
+      "epoch": 0.023466940733127783,
+      "grad_norm": 0.4841674864292145,
+      "learning_rate": 0.0001,
+      "loss": 1.2187,
+      "step": 274
+    },
+    {
+      "epoch": 0.02355258650222679,
+      "grad_norm": 0.4597908854484558,
+      "learning_rate": 0.0001,
+      "loss": 1.1378,
+      "step": 275
+    },
+    {
+      "epoch": 0.023638232271325797,
+      "grad_norm": 0.5069704055786133,
+      "learning_rate": 0.0001,
+      "loss": 0.8924,
+      "step": 276
+    },
+    {
+      "epoch": 0.023723878040424803,
+      "grad_norm": 0.11718951910734177,
+      "learning_rate": 0.0001,
+      "loss": 0.6172,
+      "step": 277
+    },
+    {
+      "epoch": 0.023809523809523808,
+      "grad_norm": 0.504993736743927,
+      "learning_rate": 0.0001,
+      "loss": 1.1368,
+      "step": 278
+    },
+    {
+      "epoch": 0.023895169578622817,
+      "grad_norm": 0.6087578535079956,
+      "learning_rate": 0.0001,
+      "loss": 0.6764,
+      "step": 279
+    },
+    {
+      "epoch": 0.023980815347721823,
+      "grad_norm": 0.09570781886577606,
+      "learning_rate": 0.0001,
+      "loss": 0.7354,
+      "step": 280
+    },
+    {
+      "epoch": 0.024066461116820828,
+      "grad_norm": 0.3590712249279022,
+      "learning_rate": 0.0001,
+      "loss": 0.622,
+      "step": 281
+    },
+    {
+      "epoch": 0.024152106885919837,
+      "grad_norm": 0.44447997212409973,
+      "learning_rate": 0.0001,
+      "loss": 1.1711,
+      "step": 282
+    },
+    {
+      "epoch": 0.024237752655018843,
+      "grad_norm": 0.6229795813560486,
+      "learning_rate": 0.0001,
+      "loss": 1.1667,
+      "step": 283
+    },
+    {
+      "epoch": 0.024323398424117848,
+      "grad_norm": 0.4002545475959778,
+      "learning_rate": 0.0001,
+      "loss": 0.9031,
+      "step": 284
+    },
+    {
+      "epoch": 0.024409044193216854,
+      "grad_norm": 0.31596484780311584,
+      "learning_rate": 0.0001,
+      "loss": 0.7523,
+      "step": 285
+    },
+    {
+      "epoch": 0.024494689962315862,
+      "grad_norm": 0.5556086897850037,
+      "learning_rate": 0.0001,
+      "loss": 0.6851,
+      "step": 286
+    },
+    {
+      "epoch": 0.024580335731414868,
+      "grad_norm": 0.7028096914291382,
+      "learning_rate": 0.0001,
+      "loss": 1.1994,
+      "step": 287
+    },
+    {
+      "epoch": 0.024665981500513873,
+      "grad_norm": 0.5197622179985046,
+      "learning_rate": 0.0001,
+      "loss": 1.1527,
+      "step": 288
+    },
+    {
+      "epoch": 0.024751627269612882,
+      "grad_norm": 0.6883754730224609,
+      "learning_rate": 0.0001,
+      "loss": 1.1901,
+      "step": 289
+    },
+    {
+      "epoch": 0.024837273038711888,
+      "grad_norm": 0.7573227882385254,
+      "learning_rate": 0.0001,
+      "loss": 0.9544,
+      "step": 290
+    },
+    {
+      "epoch": 0.024922918807810893,
+      "grad_norm": 0.6680029630661011,
+      "learning_rate": 0.0001,
+      "loss": 0.7262,
+      "step": 291
+    },
+    {
+      "epoch": 0.025008564576909902,
+      "grad_norm": 0.5889681577682495,
+      "learning_rate": 0.0001,
+      "loss": 0.5525,
+      "step": 292
+    },
+    {
+      "epoch": 0.025094210346008908,
+      "grad_norm": 0.5041593313217163,
+      "learning_rate": 0.0001,
+      "loss": 0.4736,
+      "step": 293
+    },
+    {
+      "epoch": 0.025179856115107913,
+      "grad_norm": 0.5605884790420532,
+      "learning_rate": 0.0001,
+      "loss": 0.8871,
+      "step": 294
+    },
+    {
+      "epoch": 0.02526550188420692,
+      "grad_norm": 0.8335102796554565,
+      "learning_rate": 0.0001,
+      "loss": 1.0955,
+      "step": 295
+    },
+    {
+      "epoch": 0.025351147653305928,
+      "grad_norm": 0.6045655608177185,
+      "learning_rate": 0.0001,
+      "loss": 0.6234,
+      "step": 296
+    },
+    {
+      "epoch": 0.025436793422404933,
+      "grad_norm": 0.6898568272590637,
+      "learning_rate": 0.0001,
+      "loss": 0.8652,
+      "step": 297
+    },
+    {
+      "epoch": 0.02552243919150394,
+      "grad_norm": 0.5171418190002441,
+      "learning_rate": 0.0001,
+      "loss": 0.9337,
+      "step": 298
+    },
+    {
+      "epoch": 0.025608084960602948,
+      "grad_norm": 0.49419280886650085,
+      "learning_rate": 0.0001,
+      "loss": 1.072,
+      "step": 299
+    },
+    {
+      "epoch": 0.025693730729701953,
+      "grad_norm": 0.7039000988006592,
+      "learning_rate": 0.0001,
+      "loss": 0.8097,
+      "step": 300
+    },
+    {
+      "epoch": 0.02577937649880096,
+      "grad_norm": 0.4807871878147125,
+      "learning_rate": 0.0001,
+      "loss": 0.9315,
+      "step": 301
+    },
+    {
+      "epoch": 0.025865022267899964,
+      "grad_norm": 0.6047524213790894,
+      "learning_rate": 0.0001,
+      "loss": 0.6267,
+      "step": 302
+    },
+    {
+      "epoch": 0.025950668036998973,
+      "grad_norm": 0.3583734333515167,
+      "learning_rate": 0.0001,
+      "loss": 0.7393,
+      "step": 303
+    },
+    {
+      "epoch": 0.02603631380609798,
+      "grad_norm": 0.3306781053543091,
+      "learning_rate": 0.0001,
+      "loss": 1.0749,
+      "step": 304
+    },
+    {
+      "epoch": 0.026121959575196984,
+      "grad_norm": 0.9206941723823547,
+      "learning_rate": 0.0001,
+      "loss": 1.0891,
+      "step": 305
+    },
+    {
+      "epoch": 0.026207605344295993,
+      "grad_norm": 0.6685647368431091,
+      "learning_rate": 0.0001,
+      "loss": 0.9671,
+      "step": 306
+    },
+    {
+      "epoch": 0.026293251113395,
+      "grad_norm": 0.837591826915741,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 307
+    },
+    {
+      "epoch": 0.026378896882494004,
+      "grad_norm": 0.9400232434272766,
+      "learning_rate": 0.0001,
+      "loss": 1.1678,
+      "step": 308
+    },
+    {
+      "epoch": 0.026464542651593013,
+      "grad_norm": 0.8441941142082214,
+      "learning_rate": 0.0001,
+      "loss": 1.2175,
+      "step": 309
+    },
+    {
+      "epoch": 0.02655018842069202,
+      "grad_norm": 0.6060966849327087,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 310
+    },
+    {
+      "epoch": 0.026635834189791024,
+      "grad_norm": 0.4277983605861664,
+      "learning_rate": 0.0001,
+      "loss": 1.2307,
+      "step": 311
+    },
+    {
+      "epoch": 0.02672147995889003,
+      "grad_norm": 0.5674152970314026,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 312
+    },
+    {
+      "epoch": 0.026807125727989038,
+      "grad_norm": 0.5077130198478699,
+      "learning_rate": 0.0001,
+      "loss": 0.9775,
+      "step": 313
+    },
+    {
+      "epoch": 0.026892771497088044,
+      "grad_norm": 0.6487033367156982,
+      "learning_rate": 0.0001,
+      "loss": 0.8983,
+      "step": 314
+    },
+    {
+      "epoch": 0.02697841726618705,
+      "grad_norm": 0.10633030533790588,
+      "learning_rate": 0.0001,
+      "loss": 0.5295,
+      "step": 315
+    },
+    {
+      "epoch": 0.027064063035286058,
+      "grad_norm": 0.6914339661598206,
+      "learning_rate": 0.0001,
+      "loss": 0.7745,
+      "step": 316
+    },
+    {
+      "epoch": 0.027149708804385064,
+      "grad_norm": 0.46108168363571167,
+      "learning_rate": 0.0001,
+      "loss": 0.8482,
+      "step": 317
+    },
+    {
+      "epoch": 0.02723535457348407,
+      "grad_norm": 0.6462963819503784,
+      "learning_rate": 0.0001,
+      "loss": 0.9206,
+      "step": 318
+    },
+    {
+      "epoch": 0.027321000342583078,
+      "grad_norm": 0.4174485504627228,
+      "learning_rate": 0.0001,
+      "loss": 0.3802,
+      "step": 319
+    },
+    {
+      "epoch": 0.027406646111682084,
+      "grad_norm": 0.4596540331840515,
+      "learning_rate": 0.0001,
+      "loss": 0.7645,
+      "step": 320
+    },
+    {
+      "epoch": 0.02749229188078109,
+      "grad_norm": 0.34406107664108276,
+      "learning_rate": 0.0001,
+      "loss": 0.9075,
+      "step": 321
+    },
+    {
+      "epoch": 0.027577937649880094,
+      "grad_norm": 0.8861792683601379,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 322
+    },
+    {
+      "epoch": 0.027663583418979103,
+      "grad_norm": 0.5875692963600159,
+      "learning_rate": 0.0001,
+      "loss": 0.8815,
+      "step": 323
+    },
+    {
+      "epoch": 0.02774922918807811,
+      "grad_norm": 0.21518488228321075,
+      "learning_rate": 0.0001,
+      "loss": 0.7764,
+      "step": 324
+    },
+    {
+      "epoch": 0.027834874957177114,
+      "grad_norm": 0.3777693212032318,
+      "learning_rate": 0.0001,
+      "loss": 0.7068,
+      "step": 325
+    },
+    {
+      "epoch": 0.027920520726276123,
+      "grad_norm": 0.8393121361732483,
+      "learning_rate": 0.0001,
+      "loss": 0.9568,
+      "step": 326
+    },
+    {
+      "epoch": 0.02800616649537513,
+      "grad_norm": 0.6936140656471252,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 327
+    },
+    {
+      "epoch": 0.028091812264474134,
+      "grad_norm": 0.4062822461128235,
+      "learning_rate": 0.0001,
+      "loss": 0.8734,
+      "step": 328
+    },
+    {
+      "epoch": 0.02817745803357314,
+      "grad_norm": 0.7193405628204346,
+      "learning_rate": 0.0001,
+      "loss": 1.1579,
+      "step": 329
+    },
+    {
+      "epoch": 0.02826310380267215,
+      "grad_norm": 0.8354955911636353,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 330
+    },
+    {
+      "epoch": 0.028348749571771154,
+      "grad_norm": 0.4427030086517334,
+      "learning_rate": 0.0001,
+      "loss": 1.2548,
+      "step": 331
+    },
+    {
+      "epoch": 0.02843439534087016,
+      "grad_norm": 0.6535590291023254,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 332
+    },
+    {
+      "epoch": 0.02852004110996917,
+      "grad_norm": 0.4367024004459381,
+      "learning_rate": 0.0001,
+      "loss": 1.3009,
+      "step": 333
+    },
+    {
+      "epoch": 0.028605686879068174,
+      "grad_norm": 0.5570236444473267,
+      "learning_rate": 0.0001,
+      "loss": 0.6066,
+      "step": 334
+    },
+    {
+      "epoch": 0.02869133264816718,
+      "grad_norm": 0.9001383781433105,
+      "learning_rate": 0.0001,
+      "loss": 1.2826,
+      "step": 335
+    },
+    {
+      "epoch": 0.02877697841726619,
+      "grad_norm": 0.5564957857131958,
+      "learning_rate": 0.0001,
+      "loss": 0.6452,
+      "step": 336
+    },
+    {
+      "epoch": 0.028862624186365194,
+      "grad_norm": 0.5033341646194458,
+      "learning_rate": 0.0001,
+      "loss": 0.8572,
+      "step": 337
+    },
+    {
+      "epoch": 0.0289482699554642,
+      "grad_norm": 0.533803403377533,
+      "learning_rate": 0.0001,
+      "loss": 0.5657,
+      "step": 338
+    },
+    {
+      "epoch": 0.029033915724563205,
+      "grad_norm": 0.4778352975845337,
+      "learning_rate": 0.0001,
+      "loss": 0.7338,
+      "step": 339
+    },
+    {
+      "epoch": 0.029119561493662214,
+      "grad_norm": 0.7118761539459229,
+      "learning_rate": 0.0001,
+      "loss": 0.8444,
+      "step": 340
+    },
+    {
+      "epoch": 0.02920520726276122,
+      "grad_norm": 0.7638094425201416,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 341
+    },
+    {
+      "epoch": 0.029290853031860225,
+      "grad_norm": 0.4045732617378235,
+      "learning_rate": 0.0001,
+      "loss": 1.1765,
+      "step": 342
+    },
+    {
+      "epoch": 0.029376498800959234,
+      "grad_norm": 0.45774418115615845,
+      "learning_rate": 0.0001,
+      "loss": 0.7738,
+      "step": 343
+    },
+    {
+      "epoch": 0.02946214457005824,
+      "grad_norm": 0.6258485317230225,
+      "learning_rate": 0.0001,
+      "loss": 0.8157,
+      "step": 344
+    },
+    {
+      "epoch": 0.029547790339157245,
+      "grad_norm": 0.7696349024772644,
+      "learning_rate": 0.0001,
+      "loss": 0.7756,
+      "step": 345
+    },
+    {
+      "epoch": 0.029633436108256254,
+      "grad_norm": 0.6918955445289612,
+      "learning_rate": 0.0001,
+      "loss": 0.687,
+      "step": 346
+    },
+    {
+      "epoch": 0.02971908187735526,
+      "grad_norm": 0.6549646854400635,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 347
+    },
+    {
+      "epoch": 0.029804727646454265,
+      "grad_norm": 0.6702892780303955,
+      "learning_rate": 0.0001,
+      "loss": 1.0774,
+      "step": 348
+    },
+    {
+      "epoch": 0.02989037341555327,
+      "grad_norm": 0.7989969849586487,
+      "learning_rate": 0.0001,
+      "loss": 0.8624,
+      "step": 349
+    },
+    {
+      "epoch": 0.02997601918465228,
+      "grad_norm": 0.6989643573760986,
+      "learning_rate": 0.0001,
+      "loss": 0.7234,
+      "step": 350
+    },
+    {
+      "epoch": 0.030061664953751285,
+      "grad_norm": 0.6974698901176453,
+      "learning_rate": 0.0001,
+      "loss": 1.125,
+      "step": 351
+    },
+    {
+      "epoch": 0.03014731072285029,
+      "grad_norm": 0.6903096437454224,
+      "learning_rate": 0.0001,
+      "loss": 0.7142,
+      "step": 352
+    },
+    {
+      "epoch": 0.0302329564919493,
+      "grad_norm": 0.576593816280365,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 353
+    },
+    {
+      "epoch": 0.030318602261048305,
+      "grad_norm": 0.5015482902526855,
+      "learning_rate": 0.0001,
+      "loss": 1.0341,
+      "step": 354
+    },
+    {
+      "epoch": 0.03040424803014731,
+      "grad_norm": 0.6241230368614197,
+      "learning_rate": 0.0001,
+      "loss": 1.0925,
+      "step": 355
+    },
+    {
+      "epoch": 0.030489893799246316,
+      "grad_norm": 0.5610554814338684,
+      "learning_rate": 0.0001,
+      "loss": 0.9772,
+      "step": 356
+    },
+    {
+      "epoch": 0.030575539568345324,
+      "grad_norm": 0.2536352276802063,
+      "learning_rate": 0.0001,
+      "loss": 0.3132,
+      "step": 357
+    },
+    {
+      "epoch": 0.03066118533744433,
+      "grad_norm": 0.696640133857727,
+      "learning_rate": 0.0001,
+      "loss": 0.9018,
+      "step": 358
+    },
+    {
+      "epoch": 0.030746831106543335,
+      "grad_norm": 1.0005807876586914,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 359
+    },
+    {
+      "epoch": 0.030832476875642344,
+      "grad_norm": 0.4865444600582123,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 360
+    },
+    {
+      "epoch": 0.03091812264474135,
+      "grad_norm": 0.5010689496994019,
+      "learning_rate": 0.0001,
+      "loss": 1.0946,
+      "step": 361
+    },
+    {
+      "epoch": 0.031003768413840355,
+      "grad_norm": 0.7550519108772278,
+      "learning_rate": 0.0001,
+      "loss": 1.323,
+      "step": 362
+    },
+    {
+      "epoch": 0.031089414182939364,
+      "grad_norm": 0.3914968967437744,
+      "learning_rate": 0.0001,
+      "loss": 0.3234,
+      "step": 363
+    },
+    {
+      "epoch": 0.03117505995203837,
+      "grad_norm": 0.482723206281662,
+      "learning_rate": 0.0001,
+      "loss": 1.5518,
+      "step": 364
+    },
+    {
+      "epoch": 0.03126070572113738,
+      "grad_norm": 0.42014849185943604,
+      "learning_rate": 0.0001,
+      "loss": 0.9158,
+      "step": 365
+    },
+    {
+      "epoch": 0.03134635149023638,
+      "grad_norm": 0.5904659032821655,
+      "learning_rate": 0.0001,
+      "loss": 1.1051,
+      "step": 366
+    },
+    {
+      "epoch": 0.03143199725933539,
+      "grad_norm": 0.5801008939743042,
+      "learning_rate": 0.0001,
+      "loss": 0.6024,
+      "step": 367
+    },
+    {
+      "epoch": 0.0315176430284344,
+      "grad_norm": 0.4220430254936218,
+      "learning_rate": 0.0001,
+      "loss": 1.2132,
+      "step": 368
+    },
+    {
+      "epoch": 0.0316032887975334,
+      "grad_norm": 0.38940495252609253,
+      "learning_rate": 0.0001,
+      "loss": 0.9426,
+      "step": 369
+    },
+    {
+      "epoch": 0.03168893456663241,
+      "grad_norm": 0.5514323711395264,
+      "learning_rate": 0.0001,
+      "loss": 0.9476,
+      "step": 370
+    },
+    {
+      "epoch": 0.03177458033573141,
+      "grad_norm": 0.6607529520988464,
+      "learning_rate": 0.0001,
+      "loss": 1.3996,
+      "step": 371
+    },
+    {
+      "epoch": 0.03186022610483042,
+      "grad_norm": 0.9876105189323425,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 372
+    },
+    {
+      "epoch": 0.03194587187392943,
+      "grad_norm": 0.28976908326148987,
+      "learning_rate": 0.0001,
+      "loss": 0.8423,
+      "step": 373
+    },
+    {
+      "epoch": 0.03203151764302843,
+      "grad_norm": 0.5586889386177063,
+      "learning_rate": 0.0001,
+      "loss": 1.3375,
+      "step": 374
+    },
+    {
+      "epoch": 0.03211716341212744,
+      "grad_norm": 0.4334045648574829,
+      "learning_rate": 0.0001,
+      "loss": 0.4244,
+      "step": 375
+    },
+    {
+      "epoch": 0.03220280918122645,
+      "grad_norm": 0.6791065335273743,
+      "learning_rate": 0.0001,
+      "loss": 1.0662,
+      "step": 376
+    },
+    {
+      "epoch": 0.03228845495032545,
+      "grad_norm": 0.25032666325569153,
+      "learning_rate": 0.0001,
+      "loss": 0.546,
+      "step": 377
+    },
+    {
+      "epoch": 0.03237410071942446,
+      "grad_norm": 0.45329850912094116,
+      "learning_rate": 0.0001,
+      "loss": 1.0206,
+      "step": 378
+    },
+    {
+      "epoch": 0.03245974648852347,
+      "grad_norm": 0.3927946388721466,
+      "learning_rate": 0.0001,
+      "loss": 0.6671,
+      "step": 379
+    },
+    {
+      "epoch": 0.03254539225762247,
+      "grad_norm": 0.4248335361480713,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 380
+    },
+    {
+      "epoch": 0.03263103802672148,
+      "grad_norm": 0.6347900629043579,
+      "learning_rate": 0.0001,
+      "loss": 0.8052,
+      "step": 381
+    },
+    {
+      "epoch": 0.03271668379582049,
+      "grad_norm": 0.22440668940544128,
+      "learning_rate": 0.0001,
+      "loss": 0.7707,
+      "step": 382
+    },
+    {
+      "epoch": 0.03280232956491949,
+      "grad_norm": 0.8590061664581299,
+      "learning_rate": 0.0001,
+      "loss": 1.3267,
+      "step": 383
+    },
+    {
+      "epoch": 0.0328879753340185,
+      "grad_norm": 0.7317745089530945,
+      "learning_rate": 0.0001,
+      "loss": 0.9754,
+      "step": 384
+    },
+    {
+      "epoch": 0.03297362110311751,
+      "grad_norm": 0.7000951766967773,
+      "learning_rate": 0.0001,
+      "loss": 0.7304,
+      "step": 385
+    },
+    {
+      "epoch": 0.03305926687221651,
+      "grad_norm": 0.727485179901123,
+      "learning_rate": 0.0001,
+      "loss": 1.0467,
+      "step": 386
+    },
+    {
+      "epoch": 0.03314491264131552,
+      "grad_norm": 0.41113173961639404,
+      "learning_rate": 0.0001,
+      "loss": 0.6925,
+      "step": 387
+    },
+    {
+      "epoch": 0.03323055841041452,
+      "grad_norm": 0.4411856532096863,
+      "learning_rate": 0.0001,
+      "loss": 0.9749,
+      "step": 388
+    },
+    {
+      "epoch": 0.03331620417951353,
+      "grad_norm": 0.5055375099182129,
+      "learning_rate": 0.0001,
+      "loss": 0.9865,
+      "step": 389
+    },
+    {
+      "epoch": 0.03340184994861254,
+      "grad_norm": 0.6266697645187378,
+      "learning_rate": 0.0001,
+      "loss": 0.8675,
+      "step": 390
+    },
+    {
+      "epoch": 0.03348749571771154,
+      "grad_norm": 0.8243045806884766,
+      "learning_rate": 0.0001,
+      "loss": 0.8834,
+      "step": 391
+    },
+    {
+      "epoch": 0.03357314148681055,
+      "grad_norm": 0.7042661905288696,
+      "learning_rate": 0.0001,
+      "loss": 1.3041,
+      "step": 392
+    },
+    {
+      "epoch": 0.03365878725590956,
+      "grad_norm": 0.859485387802124,
+      "learning_rate": 0.0001,
+      "loss": 1.1216,
+      "step": 393
+    },
+    {
+      "epoch": 0.03374443302500856,
+      "grad_norm": 1.173735499382019,
+      "learning_rate": 0.0001,
+      "loss": 0.958,
+      "step": 394
+    },
+    {
+      "epoch": 0.03383007879410757,
+      "grad_norm": 0.4391869008541107,
+      "learning_rate": 0.0001,
+      "loss": 0.958,
+      "step": 395
+    },
+    {
+      "epoch": 0.03391572456320658,
+      "grad_norm": 0.8439468145370483,
+      "learning_rate": 0.0001,
+      "loss": 1.037,
+      "step": 396
+    },
+    {
+      "epoch": 0.03400137033230558,
+      "grad_norm": 0.43442481756210327,
+      "learning_rate": 0.0001,
+      "loss": 1.1613,
+      "step": 397
+    },
+    {
+      "epoch": 0.03408701610140459,
+      "grad_norm": 0.8197125196456909,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 398
+    },
+    {
+      "epoch": 0.0341726618705036,
+      "grad_norm": 0.6496662497520447,
+      "learning_rate": 0.0001,
+      "loss": 1.3418,
+      "step": 399
+    },
+    {
+      "epoch": 0.0342583076396026,
+      "grad_norm": 0.3766140639781952,
+      "learning_rate": 0.0001,
+      "loss": 0.9654,
+      "step": 400
+    },
+    {
+      "epoch": 0.03434395340870161,
+      "grad_norm": 0.4241209030151367,
+      "learning_rate": 0.0001,
+      "loss": 0.6901,
+      "step": 401
+    },
+    {
+      "epoch": 0.03442959917780062,
+      "grad_norm": 0.51568204164505,
+      "learning_rate": 0.0001,
+      "loss": 0.5866,
+      "step": 402
+    },
+    {
+      "epoch": 0.03451524494689962,
+      "grad_norm": 0.5474198460578918,
+      "learning_rate": 0.0001,
+      "loss": 0.9354,
+      "step": 403
+    },
+    {
+      "epoch": 0.03460089071599863,
+      "grad_norm": 0.7140264511108398,
+      "learning_rate": 0.0001,
+      "loss": 1.0019,
+      "step": 404
+    },
+    {
+      "epoch": 0.03468653648509764,
+      "grad_norm": 0.7723624110221863,
+      "learning_rate": 0.0001,
+      "loss": 0.9943,
+      "step": 405
+    },
+    {
+      "epoch": 0.03477218225419664,
+      "grad_norm": 0.6564207077026367,
+      "learning_rate": 0.0001,
+      "loss": 1.0588,
+      "step": 406
+    },
+    {
+      "epoch": 0.03485782802329565,
+      "grad_norm": 0.4757721424102783,
+      "learning_rate": 0.0001,
+      "loss": 0.9938,
+      "step": 407
+    },
+    {
+      "epoch": 0.03494347379239465,
+      "grad_norm": 0.5616127252578735,
+      "learning_rate": 0.0001,
+      "loss": 0.4199,
+      "step": 408
+    },
+    {
+      "epoch": 0.03502911956149366,
+      "grad_norm": 0.6517274379730225,
+      "learning_rate": 0.0001,
+      "loss": 0.6378,
+      "step": 409
+    },
+    {
+      "epoch": 0.03511476533059267,
+      "grad_norm": 0.47479501366615295,
+      "learning_rate": 0.0001,
+      "loss": 1.1206,
+      "step": 410
+    },
+    {
+      "epoch": 0.03520041109969167,
+      "grad_norm": 0.8521145582199097,
+      "learning_rate": 0.0001,
+      "loss": 0.9796,
+      "step": 411
+    },
+    {
+      "epoch": 0.03528605686879068,
+      "grad_norm": 0.7911610007286072,
+      "learning_rate": 0.0001,
+      "loss": 0.958,
+      "step": 412
+    },
+    {
+      "epoch": 0.03537170263788969,
+      "grad_norm": 0.5996243357658386,
+      "learning_rate": 0.0001,
+      "loss": 0.8534,
+      "step": 413
+    },
+    {
+      "epoch": 0.03545734840698869,
+      "grad_norm": 0.5354576110839844,
+      "learning_rate": 0.0001,
+      "loss": 2.4002,
+      "step": 414
+    },
+    {
+      "epoch": 0.0355429941760877,
+      "grad_norm": 0.9011054635047913,
+      "learning_rate": 0.0001,
+      "loss": 1.2357,
+      "step": 415
+    },
+    {
+      "epoch": 0.03562863994518671,
+      "grad_norm": 0.7250404953956604,
+      "learning_rate": 0.0001,
+      "loss": 0.9395,
+      "step": 416
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 0.4891771972179413,
+      "learning_rate": 0.0001,
+      "loss": 1.2017,
+      "step": 417
+    },
+    {
+      "epoch": 0.03579993148338472,
+      "grad_norm": 0.9767696261405945,
+      "learning_rate": 0.0001,
+      "loss": 0.7088,
+      "step": 418
+    },
+    {
+      "epoch": 0.03588557725248373,
+      "grad_norm": 0.6457866430282593,
+      "learning_rate": 0.0001,
+      "loss": 1.2398,
+      "step": 419
+    },
+    {
+      "epoch": 0.03597122302158273,
+      "grad_norm": 0.872735857963562,
+      "learning_rate": 0.0001,
+      "loss": 1.2293,
+      "step": 420
+    },
+    {
+      "epoch": 0.03605686879068174,
+      "grad_norm": 0.43899470567703247,
+      "learning_rate": 0.0001,
+      "loss": 1.1084,
+      "step": 421
+    },
+    {
+      "epoch": 0.03614251455978075,
+      "grad_norm": 0.5257956385612488,
+      "learning_rate": 0.0001,
+      "loss": 0.8521,
+      "step": 422
+    },
+    {
+      "epoch": 0.03622816032887975,
+      "grad_norm": 0.6545289158821106,
+      "learning_rate": 0.0001,
+      "loss": 0.8094,
+      "step": 423
+    },
+    {
+      "epoch": 0.03631380609797876,
+      "grad_norm": 0.40801599621772766,
+      "learning_rate": 0.0001,
+      "loss": 0.5052,
+      "step": 424
+    },
+    {
+      "epoch": 0.03639945186707776,
+      "grad_norm": 0.7679464221000671,
+      "learning_rate": 0.0001,
+      "loss": 0.9975,
+      "step": 425
+    },
+    {
+      "epoch": 0.03648509763617677,
+      "grad_norm": 0.35565468668937683,
+      "learning_rate": 0.0001,
+      "loss": 0.653,
+      "step": 426
+    },
+    {
+      "epoch": 0.03657074340527578,
+      "grad_norm": 0.7221429347991943,
+      "learning_rate": 0.0001,
+      "loss": 0.8042,
+      "step": 427
+    },
+    {
+      "epoch": 0.03665638917437478,
+      "grad_norm": 0.6741929054260254,
+      "learning_rate": 0.0001,
+      "loss": 1.1648,
+      "step": 428
+    },
+    {
+      "epoch": 0.03674203494347379,
+      "grad_norm": 0.40309837460517883,
+      "learning_rate": 0.0001,
+      "loss": 1.0225,
+      "step": 429
+    },
+    {
+      "epoch": 0.0368276807125728,
+      "grad_norm": 0.4527295231819153,
+      "learning_rate": 0.0001,
+      "loss": 0.5335,
+      "step": 430
+    },
+    {
+      "epoch": 0.0369133264816718,
+      "grad_norm": 0.44008690118789673,
+      "learning_rate": 0.0001,
+      "loss": 0.8875,
+      "step": 431
+    },
+    {
+      "epoch": 0.03699897225077081,
+      "grad_norm": 0.17492040991783142,
+      "learning_rate": 0.0001,
+      "loss": 1.035,
+      "step": 432
+    },
+    {
+      "epoch": 0.03708461801986982,
+      "grad_norm": 0.5427473187446594,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 433
+    },
+    {
+      "epoch": 0.03717026378896882,
+      "grad_norm": 0.5211861729621887,
+      "learning_rate": 0.0001,
+      "loss": 1.1724,
+      "step": 434
+    },
+    {
+      "epoch": 0.03725590955806783,
+      "grad_norm": 0.4623158276081085,
+      "learning_rate": 0.0001,
+      "loss": 0.9242,
+      "step": 435
+    },
+    {
+      "epoch": 0.03734155532716684,
+      "grad_norm": 0.3550674617290497,
+      "learning_rate": 0.0001,
+      "loss": 1.238,
+      "step": 436
+    },
+    {
+      "epoch": 0.03742720109626584,
+      "grad_norm": 0.4117298722267151,
+      "learning_rate": 0.0001,
+      "loss": 0.517,
+      "step": 437
+    },
+    {
+      "epoch": 0.03751284686536485,
+      "grad_norm": 0.653476357460022,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 438
+    },
+    {
+      "epoch": 0.03759849263446386,
+      "grad_norm": 0.5731231570243835,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 439
+    },
+    {
+      "epoch": 0.03768413840356286,
+      "grad_norm": 0.5027527809143066,
+      "learning_rate": 0.0001,
+      "loss": 0.7619,
+      "step": 440
+    },
+    {
+      "epoch": 0.03776978417266187,
+      "grad_norm": 0.5321351289749146,
+      "learning_rate": 0.0001,
+      "loss": 0.8115,
+      "step": 441
+    },
+    {
+      "epoch": 0.037855429941760874,
+      "grad_norm": 0.49275532364845276,
+      "learning_rate": 0.0001,
+      "loss": 1.186,
+      "step": 442
+    },
+    {
+      "epoch": 0.03794107571085988,
+      "grad_norm": 0.8788185119628906,
+      "learning_rate": 0.0001,
+      "loss": 0.8827,
+      "step": 443
+    },
+    {
+      "epoch": 0.03802672147995889,
+      "grad_norm": 0.3941192328929901,
+      "learning_rate": 0.0001,
+      "loss": 0.581,
+      "step": 444
+    },
+    {
+      "epoch": 0.038112367249057894,
+      "grad_norm": 0.5700114369392395,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 445
+    },
+    {
+      "epoch": 0.0381980130181569,
+      "grad_norm": 0.7418696880340576,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 446
+    },
+    {
+      "epoch": 0.03828365878725591,
+      "grad_norm": 0.8263189196586609,
+      "learning_rate": 0.0001,
+      "loss": 0.6546,
+      "step": 447
+    },
+    {
+      "epoch": 0.03836930455635491,
+      "grad_norm": 0.6078072786331177,
+      "learning_rate": 0.0001,
+      "loss": 1.2803,
+      "step": 448
+    },
+    {
+      "epoch": 0.03845495032545392,
+      "grad_norm": 0.7105002999305725,
+      "learning_rate": 0.0001,
+      "loss": 1.1328,
+      "step": 449
+    },
+    {
+      "epoch": 0.03854059609455293,
+      "grad_norm": 0.6973564028739929,
+      "learning_rate": 0.0001,
+      "loss": 1.3042,
+      "step": 450
+    },
+    {
+      "epoch": 0.03862624186365193,
+      "grad_norm": 0.5898457169532776,
+      "learning_rate": 0.0001,
+      "loss": 1.1103,
+      "step": 451
+    },
+    {
+      "epoch": 0.03871188763275094,
+      "grad_norm": 0.4843125641345978,
+      "learning_rate": 0.0001,
+      "loss": 0.3792,
+      "step": 452
+    },
+    {
+      "epoch": 0.03879753340184995,
+      "grad_norm": 0.49520447850227356,
+      "learning_rate": 0.0001,
+      "loss": 0.5388,
+      "step": 453
+    },
+    {
+      "epoch": 0.03888317917094895,
+      "grad_norm": 0.5767064094543457,
+      "learning_rate": 0.0001,
+      "loss": 0.8436,
+      "step": 454
+    },
+    {
+      "epoch": 0.03896882494004796,
+      "grad_norm": 0.6450064778327942,
+      "learning_rate": 0.0001,
+      "loss": 0.6846,
+      "step": 455
+    },
+    {
+      "epoch": 0.03905447070914697,
+      "grad_norm": 0.5623563528060913,
+      "learning_rate": 0.0001,
+      "loss": 1.1936,
+      "step": 456
+    },
+    {
+      "epoch": 0.03914011647824597,
+      "grad_norm": 0.8697921633720398,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 457
+    },
+    {
+      "epoch": 0.03922576224734498,
+      "grad_norm": 0.4188801944255829,
+      "learning_rate": 0.0001,
+      "loss": 0.745,
+      "step": 458
+    },
+    {
+      "epoch": 0.03931140801644399,
+      "grad_norm": 0.7082808017730713,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 459
+    },
+    {
+      "epoch": 0.03939705378554299,
+      "grad_norm": 0.4368472099304199,
+      "learning_rate": 0.0001,
+      "loss": 1.1595,
+      "step": 460
+    },
+    {
+      "epoch": 0.039482699554642,
+      "grad_norm": 0.4652301073074341,
+      "learning_rate": 0.0001,
+      "loss": 1.2621,
+      "step": 461
+    },
+    {
+      "epoch": 0.039568345323741004,
+      "grad_norm": 0.9012389779090881,
+      "learning_rate": 0.0001,
+      "loss": 1.0063,
+      "step": 462
+    },
+    {
+      "epoch": 0.03965399109284001,
+      "grad_norm": 0.802412211894989,
+      "learning_rate": 0.0001,
+      "loss": 1.1037,
+      "step": 463
+    },
+    {
+      "epoch": 0.03973963686193902,
+      "grad_norm": 0.7641111612319946,
+      "learning_rate": 0.0001,
+      "loss": 0.6814,
+      "step": 464
+    },
+    {
+      "epoch": 0.039825282631038024,
+      "grad_norm": 0.3740234375,
+      "learning_rate": 0.0001,
+      "loss": 1.0604,
+      "step": 465
+    },
+    {
+      "epoch": 0.03991092840013703,
+      "grad_norm": 0.5811796188354492,
+      "learning_rate": 0.0001,
+      "loss": 0.962,
+      "step": 466
+    },
+    {
+      "epoch": 0.03999657416923604,
+      "grad_norm": 0.3392530083656311,
+      "learning_rate": 0.0001,
+      "loss": 1.0212,
+      "step": 467
+    },
+    {
+      "epoch": 0.040082219938335044,
+      "grad_norm": 0.6484715938568115,
+      "learning_rate": 0.0001,
+      "loss": 1.335,
+      "step": 468
+    },
+    {
+      "epoch": 0.04016786570743405,
+      "grad_norm": 0.41925591230392456,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 469
+    },
+    {
+      "epoch": 0.04025351147653306,
+      "grad_norm": 0.4993842840194702,
+      "learning_rate": 0.0001,
+      "loss": 1.0336,
+      "step": 470
+    },
+    {
+      "epoch": 0.040339157245632064,
+      "grad_norm": 0.30888500809669495,
+      "learning_rate": 0.0001,
+      "loss": 0.7685,
+      "step": 471
+    },
+    {
+      "epoch": 0.04042480301473107,
+      "grad_norm": 1.0496199131011963,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 472
+    },
+    {
+      "epoch": 0.04051044878383008,
+      "grad_norm": 0.6065738797187805,
+      "learning_rate": 0.0001,
+      "loss": 1.1105,
+      "step": 473
+    },
+    {
+      "epoch": 0.040596094552929084,
+      "grad_norm": 0.9578580856323242,
+      "learning_rate": 0.0001,
+      "loss": 0.8429,
+      "step": 474
+    },
+    {
+      "epoch": 0.04068174032202809,
+      "grad_norm": 0.1335602104663849,
+      "learning_rate": 0.0001,
+      "loss": 0.633,
+      "step": 475
+    },
+    {
+      "epoch": 0.0407673860911271,
+      "grad_norm": 0.7023565173149109,
+      "learning_rate": 0.0001,
+      "loss": 0.8708,
+      "step": 476
+    },
+    {
+      "epoch": 0.040853031860226104,
+      "grad_norm": 0.8819644451141357,
+      "learning_rate": 0.0001,
+      "loss": 0.8637,
+      "step": 477
+    },
+    {
+      "epoch": 0.04093867762932511,
+      "grad_norm": 0.5819785594940186,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 478
+    },
+    {
+      "epoch": 0.041024323398424115,
+      "grad_norm": 0.7691850066184998,
+      "learning_rate": 0.0001,
+      "loss": 1.1967,
+      "step": 479
+    },
+    {
+      "epoch": 0.041109969167523124,
+      "grad_norm": 0.30655497312545776,
+      "learning_rate": 0.0001,
+      "loss": 0.4427,
+      "step": 480
+    },
+    {
+      "epoch": 0.04119561493662213,
+      "grad_norm": 0.1481640785932541,
+      "learning_rate": 0.0001,
+      "loss": 0.8522,
+      "step": 481
+    },
+    {
+      "epoch": 0.041281260705721134,
+      "grad_norm": 0.4499081075191498,
+      "learning_rate": 0.0001,
+      "loss": 1.1948,
+      "step": 482
+    },
+    {
+      "epoch": 0.04136690647482014,
+      "grad_norm": 0.4240255653858185,
+      "learning_rate": 0.0001,
+      "loss": 1.2188,
+      "step": 483
+    },
+    {
+      "epoch": 0.04145255224391915,
+      "grad_norm": 0.46958959102630615,
+      "learning_rate": 0.0001,
+      "loss": 0.954,
+      "step": 484
+    },
+    {
+      "epoch": 0.041538198013018154,
+      "grad_norm": 0.21364186704158783,
+      "learning_rate": 0.0001,
+      "loss": 0.4943,
+      "step": 485
+    },
+    {
+      "epoch": 0.04162384378211716,
+      "grad_norm": 0.3383186459541321,
+      "learning_rate": 0.0001,
+      "loss": 0.6066,
+      "step": 486
+    },
+    {
+      "epoch": 0.04170948955121617,
+      "grad_norm": 0.214533731341362,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 487
+    },
+    {
+      "epoch": 0.041795135320315174,
+      "grad_norm": 0.6317199468612671,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 488
+    },
+    {
+      "epoch": 0.04188078108941418,
+      "grad_norm": 0.6468168497085571,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 489
+    },
+    {
+      "epoch": 0.04196642685851319,
+      "grad_norm": 0.6532948613166809,
+      "learning_rate": 0.0001,
+      "loss": 1.0875,
+      "step": 490
+    },
+    {
+      "epoch": 0.042052072627612194,
+      "grad_norm": 0.9758458733558655,
+      "learning_rate": 0.0001,
+      "loss": 0.929,
+      "step": 491
+    },
+    {
+      "epoch": 0.0421377183967112,
+      "grad_norm": 0.5082405209541321,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 492
+    },
+    {
+      "epoch": 0.04222336416581021,
+      "grad_norm": 0.4663846790790558,
+      "learning_rate": 0.0001,
+      "loss": 0.7972,
+      "step": 493
+    },
+    {
+      "epoch": 0.042309009934909214,
+      "grad_norm": 0.7232125401496887,
+      "learning_rate": 0.0001,
+      "loss": 0.8886,
+      "step": 494
+    },
+    {
+      "epoch": 0.04239465570400822,
+      "grad_norm": 0.7180168628692627,
+      "learning_rate": 0.0001,
+      "loss": 1.1871,
+      "step": 495
+    },
+    {
+      "epoch": 0.042480301473107225,
+      "grad_norm": 0.5493132472038269,
+      "learning_rate": 0.0001,
+      "loss": 1.0529,
+      "step": 496
+    },
+    {
+      "epoch": 0.042565947242206234,
+      "grad_norm": 0.6851405501365662,
+      "learning_rate": 0.0001,
+      "loss": 0.7064,
+      "step": 497
+    },
+    {
+      "epoch": 0.04265159301130524,
+      "grad_norm": 0.1943942755460739,
+      "learning_rate": 0.0001,
+      "loss": 1.009,
+      "step": 498
+    },
+    {
+      "epoch": 0.042737238780404245,
+      "grad_norm": 0.8079454302787781,
+      "learning_rate": 0.0001,
+      "loss": 0.5849,
+      "step": 499
+    },
+    {
+      "epoch": 0.042822884549503254,
+      "grad_norm": 0.3989928960800171,
+      "learning_rate": 0.0001,
+      "loss": 1.2172,
+      "step": 500
+    },
+    {
+      "epoch": 0.04290853031860226,
+      "grad_norm": 0.950165867805481,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 501
+    },
+    {
+      "epoch": 0.042994176087701265,
+      "grad_norm": 0.7703631520271301,
+      "learning_rate": 0.0001,
+      "loss": 0.8723,
+      "step": 502
+    },
+    {
+      "epoch": 0.043079821856800274,
+      "grad_norm": 0.4297791123390198,
+      "learning_rate": 0.0001,
+      "loss": 0.6843,
+      "step": 503
+    },
+    {
+      "epoch": 0.04316546762589928,
+      "grad_norm": 0.6445506811141968,
+      "learning_rate": 0.0001,
+      "loss": 0.8336,
+      "step": 504
+    },
+    {
+      "epoch": 0.043251113394998285,
+      "grad_norm": 0.3178686797618866,
+      "learning_rate": 0.0001,
+      "loss": 0.3974,
+      "step": 505
+    },
+    {
+      "epoch": 0.043336759164097294,
+      "grad_norm": 0.4348612427711487,
+      "learning_rate": 0.0001,
+      "loss": 1.3024,
+      "step": 506
+    },
+    {
+      "epoch": 0.0434224049331963,
+      "grad_norm": 0.4375927448272705,
+      "learning_rate": 0.0001,
+      "loss": 1.0365,
+      "step": 507
+    },
+    {
+      "epoch": 0.043508050702295305,
+      "grad_norm": 0.43996647000312805,
+      "learning_rate": 0.0001,
+      "loss": 0.9234,
+      "step": 508
+    },
+    {
+      "epoch": 0.043593696471394314,
+      "grad_norm": 0.4969867169857025,
+      "learning_rate": 0.0001,
+      "loss": 1.0743,
+      "step": 509
+    },
+    {
+      "epoch": 0.04367934224049332,
+      "grad_norm": 0.7977616190910339,
+      "learning_rate": 0.0001,
+      "loss": 0.8679,
+      "step": 510
+    },
+    {
+      "epoch": 0.043764988009592325,
+      "grad_norm": 0.6286057233810425,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 511
+    },
+    {
+      "epoch": 0.043850633778691334,
+      "grad_norm": 0.4764932692050934,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 512
+    },
+    {
+      "epoch": 0.04393627954779034,
+      "grad_norm": 0.7388997673988342,
+      "learning_rate": 0.0001,
+      "loss": 0.6908,
+      "step": 513
+    },
+    {
+      "epoch": 0.044021925316889345,
+      "grad_norm": 0.8288263082504272,
+      "learning_rate": 0.0001,
+      "loss": 0.8025,
+      "step": 514
+    },
+    {
+      "epoch": 0.044107571085988354,
+      "grad_norm": 0.5530857443809509,
+      "learning_rate": 0.0001,
+      "loss": 0.7601,
+      "step": 515
+    },
+    {
+      "epoch": 0.044193216855087356,
+      "grad_norm": 0.7314450740814209,
+      "learning_rate": 0.0001,
+      "loss": 0.7809,
+      "step": 516
+    },
+    {
+      "epoch": 0.044278862624186364,
+      "grad_norm": 0.5346459746360779,
+      "learning_rate": 0.0001,
+      "loss": 1.0484,
+      "step": 517
+    },
+    {
+      "epoch": 0.04436450839328537,
+      "grad_norm": 0.5049276351928711,
+      "learning_rate": 0.0001,
+      "loss": 1.2082,
+      "step": 518
+    },
+    {
+      "epoch": 0.044450154162384375,
+      "grad_norm": 0.2397557646036148,
+      "learning_rate": 0.0001,
+      "loss": 0.8265,
+      "step": 519
+    },
+    {
+      "epoch": 0.044535799931483384,
+      "grad_norm": 0.6342805624008179,
+      "learning_rate": 0.0001,
+      "loss": 0.6984,
+      "step": 520
+    },
+    {
+      "epoch": 0.04462144570058239,
+      "grad_norm": 0.4809201955795288,
+      "learning_rate": 0.0001,
+      "loss": 0.6096,
+      "step": 521
+    },
+    {
+      "epoch": 0.044707091469681395,
+      "grad_norm": 0.4146733582019806,
+      "learning_rate": 0.0001,
+      "loss": 1.0074,
+      "step": 522
+    },
+    {
+      "epoch": 0.044792737238780404,
+      "grad_norm": 0.45770740509033203,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 523
+    },
+    {
+      "epoch": 0.04487838300787941,
+      "grad_norm": 0.7746132016181946,
+      "learning_rate": 0.0001,
+      "loss": 0.9107,
+      "step": 524
+    },
+    {
+      "epoch": 0.044964028776978415,
+      "grad_norm": 0.6676192879676819,
+      "learning_rate": 0.0001,
+      "loss": 1.3361,
+      "step": 525
+    },
+    {
+      "epoch": 0.045049674546077424,
+      "grad_norm": 1.3022526502609253,
+      "learning_rate": 0.0001,
+      "loss": 1.1529,
+      "step": 526
+    },
+    {
+      "epoch": 0.04513532031517643,
+      "grad_norm": 0.9171550869941711,
+      "learning_rate": 0.0001,
+      "loss": 1.275,
+      "step": 527
+    },
+    {
+      "epoch": 0.045220966084275435,
+      "grad_norm": 0.35264453291893005,
+      "learning_rate": 0.0001,
+      "loss": 0.7331,
+      "step": 528
+    },
+    {
+      "epoch": 0.045306611853374444,
+      "grad_norm": 0.6137323379516602,
+      "learning_rate": 0.0001,
+      "loss": 0.9192,
+      "step": 529
+    },
+    {
+      "epoch": 0.04539225762247345,
+      "grad_norm": 0.7124908566474915,
+      "learning_rate": 0.0001,
+      "loss": 1.0865,
+      "step": 530
+    },
+    {
+      "epoch": 0.045477903391572455,
+      "grad_norm": 0.7161806225776672,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 531
+    },
+    {
+      "epoch": 0.045563549160671464,
+      "grad_norm": 0.7961249947547913,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 532
+    },
+    {
+      "epoch": 0.045649194929770466,
+      "grad_norm": 0.5496739149093628,
+      "learning_rate": 0.0001,
+      "loss": 0.5208,
+      "step": 533
+    },
+    {
+      "epoch": 0.045734840698869475,
+      "grad_norm": 0.5630668997764587,
+      "learning_rate": 0.0001,
+      "loss": 1.1378,
+      "step": 534
+    },
+    {
+      "epoch": 0.045820486467968484,
+      "grad_norm": 0.8745608329772949,
+      "learning_rate": 0.0001,
+      "loss": 1.1297,
+      "step": 535
+    },
+    {
+      "epoch": 0.045906132237067486,
+      "grad_norm": 1.0324448347091675,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 536
+    },
+    {
+      "epoch": 0.045991778006166495,
+      "grad_norm": 0.6493133902549744,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 537
+    },
+    {
+      "epoch": 0.046077423775265504,
+      "grad_norm": 0.5429645776748657,
+      "learning_rate": 0.0001,
+      "loss": 0.7418,
+      "step": 538
+    },
+    {
+      "epoch": 0.046163069544364506,
+      "grad_norm": 0.40769702196121216,
+      "learning_rate": 0.0001,
+      "loss": 0.788,
+      "step": 539
+    },
+    {
+      "epoch": 0.046248715313463515,
+      "grad_norm": 0.5612887740135193,
+      "learning_rate": 0.0001,
+      "loss": 0.7653,
+      "step": 540
+    },
+    {
+      "epoch": 0.046334361082562524,
+      "grad_norm": 0.7170293927192688,
+      "learning_rate": 0.0001,
+      "loss": 1.2769,
+      "step": 541
+    },
+    {
+      "epoch": 0.046420006851661526,
+      "grad_norm": 0.4823557138442993,
+      "learning_rate": 0.0001,
+      "loss": 0.946,
+      "step": 542
+    },
+    {
+      "epoch": 0.046505652620760535,
+      "grad_norm": 0.7161024212837219,
+      "learning_rate": 0.0001,
+      "loss": 0.7222,
+      "step": 543
+    },
+    {
+      "epoch": 0.046591298389859544,
+      "grad_norm": 0.5455880165100098,
+      "learning_rate": 0.0001,
+      "loss": 0.7517,
+      "step": 544
+    },
+    {
+      "epoch": 0.046676944158958546,
+      "grad_norm": 0.5424516797065735,
+      "learning_rate": 0.0001,
+      "loss": 1.1973,
+      "step": 545
+    },
+    {
+      "epoch": 0.046762589928057555,
+      "grad_norm": 0.7174406051635742,
+      "learning_rate": 0.0001,
+      "loss": 1.0134,
+      "step": 546
+    },
+    {
+      "epoch": 0.046848235697156564,
+      "grad_norm": 0.6882132887840271,
+      "learning_rate": 0.0001,
+      "loss": 0.7571,
+      "step": 547
+    },
+    {
+      "epoch": 0.046933881466255566,
+      "grad_norm": 0.3591873049736023,
+      "learning_rate": 0.0001,
+      "loss": 0.7287,
+      "step": 548
+    },
+    {
+      "epoch": 0.047019527235354575,
+      "grad_norm": 0.33525073528289795,
+      "learning_rate": 0.0001,
+      "loss": 0.3405,
+      "step": 549
+    },
+    {
+      "epoch": 0.04710517300445358,
+      "grad_norm": 0.7920336723327637,
+      "learning_rate": 0.0001,
+      "loss": 0.8064,
+      "step": 550
+    },
+    {
+      "epoch": 0.047190818773552586,
+      "grad_norm": 0.5283752083778381,
+      "learning_rate": 0.0001,
+      "loss": 0.7745,
+      "step": 551
+    },
+    {
+      "epoch": 0.047276464542651594,
+      "grad_norm": 0.5163866877555847,
+      "learning_rate": 0.0001,
+      "loss": 0.7737,
+      "step": 552
+    },
+    {
+      "epoch": 0.047362110311750596,
+      "grad_norm": 0.512323796749115,
+      "learning_rate": 0.0001,
+      "loss": 1.28,
+      "step": 553
+    },
+    {
+      "epoch": 0.047447756080849605,
+      "grad_norm": 0.5260597467422485,
+      "learning_rate": 0.0001,
+      "loss": 2.0652,
+      "step": 554
+    },
+    {
+      "epoch": 0.047533401849948614,
+      "grad_norm": 0.5083516836166382,
+      "learning_rate": 0.0001,
+      "loss": 1.314,
+      "step": 555
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.8033387660980225,
+      "learning_rate": 0.0001,
+      "loss": 0.8604,
+      "step": 556
+    },
+    {
+      "epoch": 0.047704693388146625,
+      "grad_norm": 0.410037636756897,
+      "learning_rate": 0.0001,
+      "loss": 0.9876,
+      "step": 557
+    },
+    {
+      "epoch": 0.047790339157245634,
+      "grad_norm": 0.6074786186218262,
+      "learning_rate": 0.0001,
+      "loss": 0.7822,
+      "step": 558
+    },
+    {
+      "epoch": 0.047875984926344636,
+      "grad_norm": 0.5502527356147766,
+      "learning_rate": 0.0001,
+      "loss": 0.9635,
+      "step": 559
+    },
+    {
+      "epoch": 0.047961630695443645,
+      "grad_norm": 0.5514336228370667,
+      "learning_rate": 0.0001,
+      "loss": 0.8256,
+      "step": 560
+    },
+    {
+      "epoch": 0.048047276464542654,
+      "grad_norm": 0.7425929307937622,
+      "learning_rate": 0.0001,
+      "loss": 0.5905,
+      "step": 561
+    },
+    {
+      "epoch": 0.048132922233641656,
+      "grad_norm": 0.7431812882423401,
+      "learning_rate": 0.0001,
+      "loss": 1.2105,
+      "step": 562
+    },
+    {
+      "epoch": 0.048218568002740665,
+      "grad_norm": 0.6089662313461304,
+      "learning_rate": 0.0001,
+      "loss": 1.3446,
+      "step": 563
+    },
+    {
+      "epoch": 0.048304213771839674,
+      "grad_norm": 0.4687812030315399,
+      "learning_rate": 0.0001,
+      "loss": 0.6924,
+      "step": 564
+    },
+    {
+      "epoch": 0.048389859540938676,
+      "grad_norm": 0.46229982376098633,
+      "learning_rate": 0.0001,
+      "loss": 0.8234,
+      "step": 565
+    },
+    {
+      "epoch": 0.048475505310037685,
+      "grad_norm": 0.47352489829063416,
+      "learning_rate": 0.0001,
+      "loss": 0.6484,
+      "step": 566
+    },
+    {
+      "epoch": 0.048561151079136694,
+      "grad_norm": 0.44238704442977905,
+      "learning_rate": 0.0001,
+      "loss": 0.7084,
+      "step": 567
+    },
+    {
+      "epoch": 0.048646796848235696,
+      "grad_norm": 0.8841325640678406,
+      "learning_rate": 0.0001,
+      "loss": 0.996,
+      "step": 568
+    },
+    {
+      "epoch": 0.048732442617334705,
+      "grad_norm": 0.6674210429191589,
+      "learning_rate": 0.0001,
+      "loss": 0.7883,
+      "step": 569
+    },
+    {
+      "epoch": 0.04881808838643371,
+      "grad_norm": 0.6739692091941833,
+      "learning_rate": 0.0001,
+      "loss": 1.2134,
+      "step": 570
+    },
+    {
+      "epoch": 0.048903734155532716,
+      "grad_norm": 0.43517518043518066,
+      "learning_rate": 0.0001,
+      "loss": 0.8225,
+      "step": 571
+    },
+    {
+      "epoch": 0.048989379924631725,
+      "grad_norm": 0.350155770778656,
+      "learning_rate": 0.0001,
+      "loss": 0.8601,
+      "step": 572
+    },
+    {
+      "epoch": 0.04907502569373073,
+      "grad_norm": 1.0000340938568115,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 573
+    },
+    {
+      "epoch": 0.049160671462829736,
+      "grad_norm": 0.5708431005477905,
+      "learning_rate": 0.0001,
+      "loss": 1.2757,
+      "step": 574
+    },
+    {
+      "epoch": 0.049246317231928745,
+      "grad_norm": 0.33851733803749084,
+      "learning_rate": 0.0001,
+      "loss": 0.5888,
+      "step": 575
+    },
+    {
+      "epoch": 0.04933196300102775,
+      "grad_norm": 0.6811726689338684,
+      "learning_rate": 0.0001,
+      "loss": 0.5214,
+      "step": 576
+    },
+    {
+      "epoch": 0.049417608770126756,
+      "grad_norm": 0.5684581995010376,
+      "learning_rate": 0.0001,
+      "loss": 0.9537,
+      "step": 577
+    },
+    {
+      "epoch": 0.049503254539225765,
+      "grad_norm": 0.5653926730155945,
+      "learning_rate": 0.0001,
+      "loss": 0.7855,
+      "step": 578
+    },
+    {
+      "epoch": 0.04958890030832477,
+      "grad_norm": 0.7356138229370117,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 579
+    },
+    {
+      "epoch": 0.049674546077423776,
+      "grad_norm": 0.721383273601532,
+      "learning_rate": 0.0001,
+      "loss": 1.1334,
+      "step": 580
+    },
+    {
+      "epoch": 0.049760191846522785,
+      "grad_norm": 0.4772796630859375,
+      "learning_rate": 0.0001,
+      "loss": 1.105,
+      "step": 581
+    },
+    {
+      "epoch": 0.04984583761562179,
+      "grad_norm": 0.3908754885196686,
+      "learning_rate": 0.0001,
+      "loss": 1.1653,
+      "step": 582
+    },
+    {
+      "epoch": 0.049931483384720796,
+      "grad_norm": 0.409101665019989,
+      "learning_rate": 0.0001,
+      "loss": 1.157,
+      "step": 583
+    },
+    {
+      "epoch": 0.050017129153819805,
+      "grad_norm": 0.632165253162384,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 584
+    },
+    {
+      "epoch": 0.05010277492291881,
+      "grad_norm": 0.8503615856170654,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 585
+    },
+    {
+      "epoch": 0.050188420692017816,
+      "grad_norm": 0.9421952962875366,
+      "learning_rate": 0.0001,
+      "loss": 1.0626,
+      "step": 586
+    },
+    {
+      "epoch": 0.05027406646111682,
+      "grad_norm": 0.26071715354919434,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 587
+    },
+    {
+      "epoch": 0.050359712230215826,
+      "grad_norm": 1.1106311082839966,
+      "learning_rate": 0.0001,
+      "loss": 1.0845,
+      "step": 588
+    },
+    {
+      "epoch": 0.050445357999314835,
+      "grad_norm": 0.7580122351646423,
+      "learning_rate": 0.0001,
+      "loss": 1.126,
+      "step": 589
+    },
+    {
+      "epoch": 0.05053100376841384,
+      "grad_norm": 0.38138842582702637,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 590
+    },
+    {
+      "epoch": 0.050616649537512846,
+      "grad_norm": 0.50161212682724,
+      "learning_rate": 0.0001,
+      "loss": 1.0425,
+      "step": 591
+    },
+    {
+      "epoch": 0.050702295306611855,
+      "grad_norm": 0.36723601818084717,
+      "learning_rate": 0.0001,
+      "loss": 1.1914,
+      "step": 592
+    },
+    {
+      "epoch": 0.05078794107571086,
+      "grad_norm": 0.9287137389183044,
+      "learning_rate": 0.0001,
+      "loss": 1.0933,
+      "step": 593
+    },
+    {
+      "epoch": 0.050873586844809866,
+      "grad_norm": 0.519532322883606,
+      "learning_rate": 0.0001,
+      "loss": 0.666,
+      "step": 594
+    },
+    {
+      "epoch": 0.050959232613908875,
+      "grad_norm": 0.6193188428878784,
+      "learning_rate": 0.0001,
+      "loss": 1.1225,
+      "step": 595
+    },
+    {
+      "epoch": 0.05104487838300788,
+      "grad_norm": 0.4000997543334961,
+      "learning_rate": 0.0001,
+      "loss": 0.7752,
+      "step": 596
+    },
+    {
+      "epoch": 0.051130524152106886,
+      "grad_norm": 0.7142748832702637,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 597
+    },
+    {
+      "epoch": 0.051216169921205895,
+      "grad_norm": 0.7288256883621216,
+      "learning_rate": 0.0001,
+      "loss": 0.7872,
+      "step": 598
+    },
+    {
+      "epoch": 0.0513018156903049,
+      "grad_norm": 0.5121176242828369,
+      "learning_rate": 0.0001,
+      "loss": 0.7621,
+      "step": 599
+    },
+    {
+      "epoch": 0.051387461459403906,
+      "grad_norm": 0.5608232021331787,
+      "learning_rate": 0.0001,
+      "loss": 0.5405,
+      "step": 600
+    },
+    {
+      "epoch": 0.051473107228502915,
+      "grad_norm": 0.43688416481018066,
+      "learning_rate": 0.0001,
+      "loss": 0.913,
+      "step": 601
+    },
+    {
+      "epoch": 0.05155875299760192,
+      "grad_norm": 0.7652155756950378,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 602
+    },
+    {
+      "epoch": 0.051644398766700926,
+      "grad_norm": 0.7221406698226929,
+      "learning_rate": 0.0001,
+      "loss": 0.6557,
+      "step": 603
+    },
+    {
+      "epoch": 0.05173004453579993,
+      "grad_norm": 0.7641146183013916,
+      "learning_rate": 0.0001,
+      "loss": 0.9725,
+      "step": 604
+    },
+    {
+      "epoch": 0.05181569030489894,
+      "grad_norm": 0.8286182284355164,
+      "learning_rate": 0.0001,
+      "loss": 1.187,
+      "step": 605
+    },
+    {
+      "epoch": 0.051901336073997946,
+      "grad_norm": 0.8967953324317932,
+      "learning_rate": 0.0001,
+      "loss": 1.9493,
+      "step": 606
+    },
+    {
+      "epoch": 0.05198698184309695,
+      "grad_norm": 0.676673173904419,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 607
+    },
+    {
+      "epoch": 0.05207262761219596,
+      "grad_norm": 0.47513341903686523,
+      "learning_rate": 0.0001,
+      "loss": 0.5757,
+      "step": 608
+    },
+    {
+      "epoch": 0.052158273381294966,
+      "grad_norm": 0.9002152681350708,
+      "learning_rate": 0.0001,
+      "loss": 1.0967,
+      "step": 609
+    },
+    {
+      "epoch": 0.05224391915039397,
+      "grad_norm": 0.6119277477264404,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 610
+    },
+    {
+      "epoch": 0.05232956491949298,
+      "grad_norm": 0.730566143989563,
+      "learning_rate": 0.0001,
+      "loss": 1.3725,
+      "step": 611
+    },
+    {
+      "epoch": 0.052415210688591986,
+      "grad_norm": 0.5837122201919556,
+      "learning_rate": 0.0001,
+      "loss": 0.8119,
+      "step": 612
+    },
+    {
+      "epoch": 0.05250085645769099,
+      "grad_norm": 0.31312111020088196,
+      "learning_rate": 0.0001,
+      "loss": 0.6153,
+      "step": 613
+    },
+    {
+      "epoch": 0.05258650222679,
+      "grad_norm": 0.24057519435882568,
+      "learning_rate": 0.0001,
+      "loss": 0.4573,
+      "step": 614
+    },
+    {
+      "epoch": 0.052672147995889006,
+      "grad_norm": 0.6130755543708801,
+      "learning_rate": 0.0001,
+      "loss": 0.6861,
+      "step": 615
+    },
+    {
+      "epoch": 0.05275779376498801,
+      "grad_norm": 0.6537296175956726,
+      "learning_rate": 0.0001,
+      "loss": 0.9261,
+      "step": 616
+    },
+    {
+      "epoch": 0.05284343953408702,
+      "grad_norm": 0.8185330033302307,
+      "learning_rate": 0.0001,
+      "loss": 0.6714,
+      "step": 617
+    },
+    {
+      "epoch": 0.052929085303186026,
+      "grad_norm": 0.7503707408905029,
+      "learning_rate": 0.0001,
+      "loss": 0.9731,
+      "step": 618
+    },
+    {
+      "epoch": 0.05301473107228503,
+      "grad_norm": 0.4717768132686615,
+      "learning_rate": 0.0001,
+      "loss": 0.7803,
+      "step": 619
+    },
+    {
+      "epoch": 0.05310037684138404,
+      "grad_norm": 0.5815584063529968,
+      "learning_rate": 0.0001,
+      "loss": 0.6787,
+      "step": 620
+    },
+    {
+      "epoch": 0.053186022610483046,
+      "grad_norm": 0.3396456241607666,
+      "learning_rate": 0.0001,
+      "loss": 0.6446,
+      "step": 621
+    },
+    {
+      "epoch": 0.05327166837958205,
+      "grad_norm": 0.7437363266944885,
+      "learning_rate": 0.0001,
+      "loss": 1.1131,
+      "step": 622
+    },
+    {
+      "epoch": 0.053357314148681056,
+      "grad_norm": 0.7141955494880676,
+      "learning_rate": 0.0001,
+      "loss": 1.1841,
+      "step": 623
+    },
+    {
+      "epoch": 0.05344295991778006,
+      "grad_norm": 0.7344996333122253,
+      "learning_rate": 0.0001,
+      "loss": 0.9056,
+      "step": 624
+    },
+    {
+      "epoch": 0.05352860568687907,
+      "grad_norm": 0.6827633380889893,
+      "learning_rate": 0.0001,
+      "loss": 1.1099,
+      "step": 625
+    },
+    {
+      "epoch": 0.053614251455978076,
+      "grad_norm": 0.7095338702201843,
+      "learning_rate": 0.0001,
+      "loss": 1.0634,
+      "step": 626
+    },
+    {
+      "epoch": 0.05369989722507708,
+      "grad_norm": 0.7032517194747925,
+      "learning_rate": 0.0001,
+      "loss": 1.1173,
+      "step": 627
+    },
+    {
+      "epoch": 0.05378554299417609,
+      "grad_norm": 0.502735435962677,
+      "learning_rate": 0.0001,
+      "loss": 1.3994,
+      "step": 628
+    },
+    {
+      "epoch": 0.053871188763275096,
+      "grad_norm": 0.7908719778060913,
+      "learning_rate": 0.0001,
+      "loss": 1.1773,
+      "step": 629
+    },
+    {
+      "epoch": 0.0539568345323741,
+      "grad_norm": 0.26638758182525635,
+      "learning_rate": 0.0001,
+      "loss": 0.2928,
+      "step": 630
+    },
+    {
+      "epoch": 0.05404248030147311,
+      "grad_norm": 0.11984463036060333,
+      "learning_rate": 0.0001,
+      "loss": 0.1306,
+      "step": 631
+    },
+    {
+      "epoch": 0.054128126070572116,
+      "grad_norm": 0.7586448192596436,
+      "learning_rate": 0.0001,
+      "loss": 1.0866,
+      "step": 632
+    },
+    {
+      "epoch": 0.05421377183967112,
+      "grad_norm": 0.764858603477478,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 633
+    },
+    {
+      "epoch": 0.05429941760877013,
+      "grad_norm": 0.7381802201271057,
+      "learning_rate": 0.0001,
+      "loss": 0.6784,
+      "step": 634
+    },
+    {
+      "epoch": 0.054385063377869136,
+      "grad_norm": 0.413194864988327,
+      "learning_rate": 0.0001,
+      "loss": 1.2175,
+      "step": 635
+    },
+    {
+      "epoch": 0.05447070914696814,
+      "grad_norm": 0.5560341477394104,
+      "learning_rate": 0.0001,
+      "loss": 0.5423,
+      "step": 636
+    },
+    {
+      "epoch": 0.05455635491606715,
+      "grad_norm": 0.37948688864707947,
+      "learning_rate": 0.0001,
+      "loss": 1.0205,
+      "step": 637
+    },
+    {
+      "epoch": 0.054642000685166156,
+      "grad_norm": 0.5710121393203735,
+      "learning_rate": 0.0001,
+      "loss": 0.5539,
+      "step": 638
+    },
+    {
+      "epoch": 0.05472764645426516,
+      "grad_norm": 0.2698993384838104,
+      "learning_rate": 0.0001,
+      "loss": 0.4214,
+      "step": 639
+    },
+    {
+      "epoch": 0.05481329222336417,
+      "grad_norm": 0.09341246634721756,
+      "learning_rate": 0.0001,
+      "loss": 0.7033,
+      "step": 640
+    },
+    {
+      "epoch": 0.05489893799246317,
+      "grad_norm": 0.6860104203224182,
+      "learning_rate": 0.0001,
+      "loss": 0.9515,
+      "step": 641
+    },
+    {
+      "epoch": 0.05498458376156218,
+      "grad_norm": 0.541595458984375,
+      "learning_rate": 0.0001,
+      "loss": 0.4938,
+      "step": 642
+    },
+    {
+      "epoch": 0.05507022953066119,
+      "grad_norm": 0.9524424076080322,
+      "learning_rate": 0.0001,
+      "loss": 0.9619,
+      "step": 643
+    },
+    {
+      "epoch": 0.05515587529976019,
+      "grad_norm": 0.6466501951217651,
+      "learning_rate": 0.0001,
+      "loss": 0.8038,
+      "step": 644
+    },
+    {
+      "epoch": 0.0552415210688592,
+      "grad_norm": 0.28683874011039734,
+      "learning_rate": 0.0001,
+      "loss": 0.413,
+      "step": 645
+    },
+    {
+      "epoch": 0.05532716683795821,
+      "grad_norm": 0.8117082715034485,
+      "learning_rate": 0.0001,
+      "loss": 1.1611,
+      "step": 646
+    },
+    {
+      "epoch": 0.05541281260705721,
+      "grad_norm": 0.36540883779525757,
+      "learning_rate": 0.0001,
+      "loss": 0.7767,
+      "step": 647
+    },
+    {
+      "epoch": 0.05549845837615622,
+      "grad_norm": 0.3318658173084259,
+      "learning_rate": 0.0001,
+      "loss": 0.8642,
+      "step": 648
+    },
+    {
+      "epoch": 0.05558410414525523,
+      "grad_norm": 0.2670716345310211,
+      "learning_rate": 0.0001,
+      "loss": 0.4008,
+      "step": 649
+    },
+    {
+      "epoch": 0.05566974991435423,
+      "grad_norm": 0.4115969240665436,
+      "learning_rate": 0.0001,
+      "loss": 0.84,
+      "step": 650
+    },
+    {
+      "epoch": 0.05575539568345324,
+      "grad_norm": 0.49050867557525635,
+      "learning_rate": 0.0001,
+      "loss": 0.9503,
+      "step": 651
+    },
+    {
+      "epoch": 0.05584104145255225,
+      "grad_norm": 0.5048901438713074,
+      "learning_rate": 0.0001,
+      "loss": 0.5715,
+      "step": 652
+    },
+    {
+      "epoch": 0.05592668722165125,
+      "grad_norm": 0.7526636719703674,
+      "learning_rate": 0.0001,
+      "loss": 1.2827,
+      "step": 653
+    },
+    {
+      "epoch": 0.05601233299075026,
+      "grad_norm": 0.6038935780525208,
+      "learning_rate": 0.0001,
+      "loss": 0.5659,
+      "step": 654
+    },
+    {
+      "epoch": 0.05609797875984927,
+      "grad_norm": 0.5475082993507385,
+      "learning_rate": 0.0001,
+      "loss": 1.0698,
+      "step": 655
+    },
+    {
+      "epoch": 0.05618362452894827,
+      "grad_norm": 0.4802013635635376,
+      "learning_rate": 0.0001,
+      "loss": 1.027,
+      "step": 656
+    },
+    {
+      "epoch": 0.05626927029804728,
+      "grad_norm": 0.9009151458740234,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 657
+    },
+    {
+      "epoch": 0.05635491606714628,
+      "grad_norm": 0.8832013010978699,
+      "learning_rate": 0.0001,
+      "loss": 1.3443,
+      "step": 658
+    },
+    {
+      "epoch": 0.05644056183624529,
+      "grad_norm": 0.41954532265663147,
+      "learning_rate": 0.0001,
+      "loss": 1.039,
+      "step": 659
+    },
+    {
+      "epoch": 0.0565262076053443,
+      "grad_norm": 0.6622366309165955,
+      "learning_rate": 0.0001,
+      "loss": 0.8604,
+      "step": 660
+    },
+    {
+      "epoch": 0.0566118533744433,
+      "grad_norm": 0.7279390692710876,
+      "learning_rate": 0.0001,
+      "loss": 2.3037,
+      "step": 661
+    },
+    {
+      "epoch": 0.05669749914354231,
+      "grad_norm": 0.8500781655311584,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 662
+    },
+    {
+      "epoch": 0.05678314491264132,
+      "grad_norm": 0.3745923340320587,
+      "learning_rate": 0.0001,
+      "loss": 1.2957,
+      "step": 663
+    },
+    {
+      "epoch": 0.05686879068174032,
+      "grad_norm": 0.44493216276168823,
+      "learning_rate": 0.0001,
+      "loss": 1.2107,
+      "step": 664
+    },
+    {
+      "epoch": 0.05695443645083933,
+      "grad_norm": 0.7223116159439087,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 665
+    },
+    {
+      "epoch": 0.05704008221993834,
+      "grad_norm": 0.6123448610305786,
+      "learning_rate": 0.0001,
+      "loss": 1.1125,
+      "step": 666
+    },
+    {
+      "epoch": 0.05712572798903734,
+      "grad_norm": 0.69273841381073,
+      "learning_rate": 0.0001,
+      "loss": 1.1405,
+      "step": 667
+    },
+    {
+      "epoch": 0.05721137375813635,
+      "grad_norm": 0.11701565235853195,
+      "learning_rate": 0.0001,
+      "loss": 0.8956,
+      "step": 668
+    },
+    {
+      "epoch": 0.05729701952723536,
+      "grad_norm": 0.896030604839325,
+      "learning_rate": 0.0001,
+      "loss": 1.4636,
+      "step": 669
+    },
+    {
+      "epoch": 0.05738266529633436,
+      "grad_norm": 0.523858368396759,
+      "learning_rate": 0.0001,
+      "loss": 1.1979,
+      "step": 670
+    },
+    {
+      "epoch": 0.05746831106543337,
+      "grad_norm": 0.7075115442276001,
+      "learning_rate": 0.0001,
+      "loss": 0.6799,
+      "step": 671
+    },
+    {
+      "epoch": 0.05755395683453238,
+      "grad_norm": 0.4868046045303345,
+      "learning_rate": 0.0001,
+      "loss": 1.0887,
+      "step": 672
+    },
+    {
+      "epoch": 0.05763960260363138,
+      "grad_norm": 0.756670355796814,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 673
+    },
+    {
+      "epoch": 0.05772524837273039,
+      "grad_norm": 0.7803313136100769,
+      "learning_rate": 0.0001,
+      "loss": 0.8577,
+      "step": 674
+    },
+    {
+      "epoch": 0.0578108941418294,
+      "grad_norm": 0.6669572591781616,
+      "learning_rate": 0.0001,
+      "loss": 0.7757,
+      "step": 675
+    },
+    {
+      "epoch": 0.0578965399109284,
+      "grad_norm": 0.4315584897994995,
+      "learning_rate": 0.0001,
+      "loss": 0.9284,
+      "step": 676
+    },
+    {
+      "epoch": 0.05798218568002741,
+      "grad_norm": 0.7851589322090149,
+      "learning_rate": 0.0001,
+      "loss": 1.2023,
+      "step": 677
+    },
+    {
+      "epoch": 0.05806783144912641,
+      "grad_norm": 0.9606513977050781,
+      "learning_rate": 0.0001,
+      "loss": 1.086,
+      "step": 678
+    },
+    {
+      "epoch": 0.05815347721822542,
+      "grad_norm": 0.5235942602157593,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 679
+    },
+    {
+      "epoch": 0.05823912298732443,
+      "grad_norm": 0.3514179587364197,
+      "learning_rate": 0.0001,
+      "loss": 0.7803,
+      "step": 680
+    },
+    {
+      "epoch": 0.05832476875642343,
+      "grad_norm": 0.6455320715904236,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 681
+    },
+    {
+      "epoch": 0.05841041452552244,
+      "grad_norm": 0.4473048746585846,
+      "learning_rate": 0.0001,
+      "loss": 0.4082,
+      "step": 682
+    },
+    {
+      "epoch": 0.05849606029462145,
+      "grad_norm": 0.5274837017059326,
+      "learning_rate": 0.0001,
+      "loss": 1.073,
+      "step": 683
+    },
+    {
+      "epoch": 0.05858170606372045,
+      "grad_norm": 0.25785553455352783,
+      "learning_rate": 0.0001,
+      "loss": 0.4428,
+      "step": 684
+    },
+    {
+      "epoch": 0.05866735183281946,
+      "grad_norm": 0.5197411179542542,
+      "learning_rate": 0.0001,
+      "loss": 1.2489,
+      "step": 685
+    },
+    {
+      "epoch": 0.05875299760191847,
+      "grad_norm": 0.6875081062316895,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 686
+    },
+    {
+      "epoch": 0.05883864337101747,
+      "grad_norm": 0.7779210209846497,
+      "learning_rate": 0.0001,
+      "loss": 0.5374,
+      "step": 687
+    },
+    {
+      "epoch": 0.05892428914011648,
+      "grad_norm": 0.6531057357788086,
+      "learning_rate": 0.0001,
+      "loss": 1.2008,
+      "step": 688
+    },
+    {
+      "epoch": 0.05900993490921549,
+      "grad_norm": 0.4662330746650696,
+      "learning_rate": 0.0001,
+      "loss": 0.661,
+      "step": 689
+    },
+    {
+      "epoch": 0.05909558067831449,
+      "grad_norm": 0.9418478012084961,
+      "learning_rate": 0.0001,
+      "loss": 1.2381,
+      "step": 690
+    },
+    {
+      "epoch": 0.0591812264474135,
+      "grad_norm": 0.32507359981536865,
+      "learning_rate": 0.0001,
+      "loss": 0.6726,
+      "step": 691
+    },
+    {
+      "epoch": 0.05926687221651251,
+      "grad_norm": 0.3598092794418335,
+      "learning_rate": 0.0001,
+      "loss": 0.4922,
+      "step": 692
+    },
+    {
+      "epoch": 0.05935251798561151,
+      "grad_norm": 0.18320761620998383,
+      "learning_rate": 0.0001,
+      "loss": 0.6214,
+      "step": 693
+    },
+    {
+      "epoch": 0.05943816375471052,
+      "grad_norm": 0.8022447824478149,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 694
+    },
+    {
+      "epoch": 0.05952380952380952,
+      "grad_norm": 0.5966013669967651,
+      "learning_rate": 0.0001,
+      "loss": 0.7725,
+      "step": 695
+    },
+    {
+      "epoch": 0.05960945529290853,
+      "grad_norm": 0.8023384809494019,
+      "learning_rate": 0.0001,
+      "loss": 1.2406,
+      "step": 696
+    },
+    {
+      "epoch": 0.05969510106200754,
+      "grad_norm": 0.5644010901451111,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 697
+    },
+    {
+      "epoch": 0.05978074683110654,
+      "grad_norm": 0.7467589974403381,
+      "learning_rate": 0.0001,
+      "loss": 0.9911,
+      "step": 698
+    },
+    {
+      "epoch": 0.05986639260020555,
+      "grad_norm": 0.5473981499671936,
+      "learning_rate": 0.0001,
+      "loss": 1.1076,
+      "step": 699
+    },
+    {
+      "epoch": 0.05995203836930456,
+      "grad_norm": 0.883334219455719,
+      "learning_rate": 0.0001,
+      "loss": 1.3539,
+      "step": 700
+    },
+    {
+      "epoch": 0.06003768413840356,
+      "grad_norm": 0.5174258351325989,
+      "learning_rate": 0.0001,
+      "loss": 1.2584,
+      "step": 701
+    },
+    {
+      "epoch": 0.06012332990750257,
+      "grad_norm": 0.4599548280239105,
+      "learning_rate": 0.0001,
+      "loss": 1.1593,
+      "step": 702
+    },
+    {
+      "epoch": 0.06020897567660158,
+      "grad_norm": 0.5445143580436707,
+      "learning_rate": 0.0001,
+      "loss": 0.6491,
+      "step": 703
+    },
+    {
+      "epoch": 0.06029462144570058,
+      "grad_norm": 0.7480692863464355,
+      "learning_rate": 0.0001,
+      "loss": 1.124,
+      "step": 704
+    },
+    {
+      "epoch": 0.06038026721479959,
+      "grad_norm": 0.6988990306854248,
+      "learning_rate": 0.0001,
+      "loss": 0.9379,
+      "step": 705
+    },
+    {
+      "epoch": 0.0604659129838986,
+      "grad_norm": 0.779815137386322,
+      "learning_rate": 0.0001,
+      "loss": 0.7462,
+      "step": 706
+    },
+    {
+      "epoch": 0.0605515587529976,
+      "grad_norm": 0.5516529083251953,
+      "learning_rate": 0.0001,
+      "loss": 0.9736,
+      "step": 707
+    },
+    {
+      "epoch": 0.06063720452209661,
+      "grad_norm": 0.4913274049758911,
+      "learning_rate": 0.0001,
+      "loss": 0.616,
+      "step": 708
+    },
+    {
+      "epoch": 0.06072285029119562,
+      "grad_norm": 0.8959532976150513,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 709
+    },
+    {
+      "epoch": 0.06080849606029462,
+      "grad_norm": 0.5943747758865356,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 710
+    },
+    {
+      "epoch": 0.06089414182939363,
+      "grad_norm": 0.6380194425582886,
+      "learning_rate": 0.0001,
+      "loss": 0.8943,
+      "step": 711
+    },
+    {
+      "epoch": 0.06097978759849263,
+      "grad_norm": 0.5472355484962463,
+      "learning_rate": 0.0001,
+      "loss": 1.2239,
+      "step": 712
+    },
+    {
+      "epoch": 0.06106543336759164,
+      "grad_norm": 0.6007921695709229,
+      "learning_rate": 0.0001,
+      "loss": 1.1554,
+      "step": 713
+    },
+    {
+      "epoch": 0.06115107913669065,
+      "grad_norm": 0.3976982831954956,
+      "learning_rate": 0.0001,
+      "loss": 0.3519,
+      "step": 714
+    },
+    {
+      "epoch": 0.06123672490578965,
+      "grad_norm": 0.3488907516002655,
+      "learning_rate": 0.0001,
+      "loss": 0.6904,
+      "step": 715
+    },
+    {
+      "epoch": 0.06132237067488866,
+      "grad_norm": 0.43919798731803894,
+      "learning_rate": 0.0001,
+      "loss": 1.1845,
+      "step": 716
+    },
+    {
+      "epoch": 0.06140801644398767,
+      "grad_norm": 0.4159376323223114,
+      "learning_rate": 0.0001,
+      "loss": 1.1382,
+      "step": 717
+    },
+    {
+      "epoch": 0.06149366221308667,
+      "grad_norm": 0.670722484588623,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 718
+    },
+    {
+      "epoch": 0.06157930798218568,
+      "grad_norm": 0.4925307035446167,
+      "learning_rate": 0.0001,
+      "loss": 1.0327,
+      "step": 719
+    },
+    {
+      "epoch": 0.06166495375128469,
+      "grad_norm": 0.8640962243080139,
+      "learning_rate": 0.0001,
+      "loss": 1.2606,
+      "step": 720
+    },
+    {
+      "epoch": 0.06175059952038369,
+      "grad_norm": 0.5233961343765259,
+      "learning_rate": 0.0001,
+      "loss": 1.1317,
+      "step": 721
+    },
+    {
+      "epoch": 0.0618362452894827,
+      "grad_norm": 0.8031330704689026,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 722
+    },
+    {
+      "epoch": 0.06192189105858171,
+      "grad_norm": 0.4983638823032379,
+      "learning_rate": 0.0001,
+      "loss": 1.3077,
+      "step": 723
+    },
+    {
+      "epoch": 0.06200753682768071,
+      "grad_norm": 0.10826991498470306,
+      "learning_rate": 0.0001,
+      "loss": 0.8397,
+      "step": 724
+    },
+    {
+      "epoch": 0.06209318259677972,
+      "grad_norm": 0.4660208821296692,
+      "learning_rate": 0.0001,
+      "loss": 0.9162,
+      "step": 725
+    },
+    {
+      "epoch": 0.06217882836587873,
+      "grad_norm": 0.4512110650539398,
+      "learning_rate": 0.0001,
+      "loss": 1.0768,
+      "step": 726
+    },
+    {
+      "epoch": 0.06226447413497773,
+      "grad_norm": 0.4914858937263489,
+      "learning_rate": 0.0001,
+      "loss": 0.5795,
+      "step": 727
+    },
+    {
+      "epoch": 0.06235011990407674,
+      "grad_norm": 0.37967580556869507,
+      "learning_rate": 0.0001,
+      "loss": 1.0532,
+      "step": 728
+    },
+    {
+      "epoch": 0.06243576567317575,
+      "grad_norm": 0.7114173769950867,
+      "learning_rate": 0.0001,
+      "loss": 1.0451,
+      "step": 729
+    },
+    {
+      "epoch": 0.06252141144227476,
+      "grad_norm": 0.610486626625061,
+      "learning_rate": 0.0001,
+      "loss": 1.1314,
+      "step": 730
+    },
+    {
+      "epoch": 0.06260705721137376,
+      "grad_norm": 0.30135577917099,
+      "learning_rate": 0.0001,
+      "loss": 0.2253,
+      "step": 731
+    },
+    {
+      "epoch": 0.06269270298047276,
+      "grad_norm": 0.440043568611145,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 732
+    },
+    {
+      "epoch": 0.06277834874957178,
+      "grad_norm": 0.38572466373443604,
+      "learning_rate": 0.0001,
+      "loss": 1.1866,
+      "step": 733
+    },
+    {
+      "epoch": 0.06286399451867078,
+      "grad_norm": 0.511793851852417,
+      "learning_rate": 0.0001,
+      "loss": 1.9273,
+      "step": 734
+    },
+    {
+      "epoch": 0.06294964028776978,
+      "grad_norm": 0.379835307598114,
+      "learning_rate": 0.0001,
+      "loss": 0.5529,
+      "step": 735
+    },
+    {
+      "epoch": 0.0630352860568688,
+      "grad_norm": 0.29254835844039917,
+      "learning_rate": 0.0001,
+      "loss": 0.4244,
+      "step": 736
+    },
+    {
+      "epoch": 0.0631209318259678,
+      "grad_norm": 0.612762451171875,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 737
+    },
+    {
+      "epoch": 0.0632065775950668,
+      "grad_norm": 0.5478065013885498,
+      "learning_rate": 0.0001,
+      "loss": 1.045,
+      "step": 738
+    },
+    {
+      "epoch": 0.0632922233641658,
+      "grad_norm": 0.8628940582275391,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 739
+    },
+    {
+      "epoch": 0.06337786913326482,
+      "grad_norm": 0.7388699650764465,
+      "learning_rate": 0.0001,
+      "loss": 0.9586,
+      "step": 740
+    },
+    {
+      "epoch": 0.06346351490236382,
+      "grad_norm": 0.6450649499893188,
+      "learning_rate": 0.0001,
+      "loss": 0.8133,
+      "step": 741
+    },
+    {
+      "epoch": 0.06354916067146282,
+      "grad_norm": 0.591132640838623,
+      "learning_rate": 0.0001,
+      "loss": 0.7923,
+      "step": 742
+    },
+    {
+      "epoch": 0.06363480644056184,
+      "grad_norm": 0.14519643783569336,
+      "learning_rate": 0.0001,
+      "loss": 0.7566,
+      "step": 743
+    },
+    {
+      "epoch": 0.06372045220966084,
+      "grad_norm": 0.7019482851028442,
+      "learning_rate": 0.0001,
+      "loss": 0.9629,
+      "step": 744
+    },
+    {
+      "epoch": 0.06380609797875984,
+      "grad_norm": 0.7347108125686646,
+      "learning_rate": 0.0001,
+      "loss": 1.9149,
+      "step": 745
+    },
+    {
+      "epoch": 0.06389174374785886,
+      "grad_norm": 0.5329987406730652,
+      "learning_rate": 0.0001,
+      "loss": 0.6807,
+      "step": 746
+    },
+    {
+      "epoch": 0.06397738951695786,
+      "grad_norm": 0.6297817826271057,
+      "learning_rate": 0.0001,
+      "loss": 0.9615,
+      "step": 747
+    },
+    {
+      "epoch": 0.06406303528605686,
+      "grad_norm": 0.33294835686683655,
+      "learning_rate": 0.0001,
+      "loss": 0.5601,
+      "step": 748
+    },
+    {
+      "epoch": 0.06414868105515588,
+      "grad_norm": 0.6073649525642395,
+      "learning_rate": 0.0001,
+      "loss": 0.6942,
+      "step": 749
+    },
+    {
+      "epoch": 0.06423432682425488,
+      "grad_norm": 0.5542661547660828,
+      "learning_rate": 0.0001,
+      "loss": 1.1243,
+      "step": 750
+    },
+    {
+      "epoch": 0.06431997259335388,
+      "grad_norm": 0.45431455969810486,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 751
+    },
+    {
+      "epoch": 0.0644056183624529,
+      "grad_norm": 0.8246678113937378,
+      "learning_rate": 0.0001,
+      "loss": 1.286,
+      "step": 752
+    },
+    {
+      "epoch": 0.0644912641315519,
+      "grad_norm": 0.35999032855033875,
+      "learning_rate": 0.0001,
+      "loss": 0.8359,
+      "step": 753
+    },
+    {
+      "epoch": 0.0645769099006509,
+      "grad_norm": 0.691268265247345,
+      "learning_rate": 0.0001,
+      "loss": 0.5405,
+      "step": 754
+    },
+    {
+      "epoch": 0.06466255566974992,
+      "grad_norm": 0.516490638256073,
+      "learning_rate": 0.0001,
+      "loss": 1.0306,
+      "step": 755
+    },
+    {
+      "epoch": 0.06474820143884892,
+      "grad_norm": 0.6640735864639282,
+      "learning_rate": 0.0001,
+      "loss": 0.6071,
+      "step": 756
+    },
+    {
+      "epoch": 0.06483384720794792,
+      "grad_norm": 0.5953500866889954,
+      "learning_rate": 0.0001,
+      "loss": 1.0756,
+      "step": 757
+    },
+    {
+      "epoch": 0.06491949297704694,
+      "grad_norm": 0.4002969264984131,
+      "learning_rate": 0.0001,
+      "loss": 1.0823,
+      "step": 758
+    },
+    {
+      "epoch": 0.06500513874614594,
+      "grad_norm": 0.7642786502838135,
+      "learning_rate": 0.0001,
+      "loss": 0.6784,
+      "step": 759
+    },
+    {
+      "epoch": 0.06509078451524494,
+      "grad_norm": 0.5364596247673035,
+      "learning_rate": 0.0001,
+      "loss": 1.0676,
+      "step": 760
+    },
+    {
+      "epoch": 0.06517643028434396,
+      "grad_norm": 0.7519243359565735,
+      "learning_rate": 0.0001,
+      "loss": 1.2291,
+      "step": 761
+    },
+    {
+      "epoch": 0.06526207605344296,
+      "grad_norm": 0.8378028273582458,
+      "learning_rate": 0.0001,
+      "loss": 1.2826,
+      "step": 762
+    },
+    {
+      "epoch": 0.06534772182254196,
+      "grad_norm": 0.29634398221969604,
+      "learning_rate": 0.0001,
+      "loss": 0.4851,
+      "step": 763
+    },
+    {
+      "epoch": 0.06543336759164098,
+      "grad_norm": 0.3588211238384247,
+      "learning_rate": 0.0001,
+      "loss": 1.0186,
+      "step": 764
+    },
+    {
+      "epoch": 0.06551901336073998,
+      "grad_norm": 0.43837299942970276,
+      "learning_rate": 0.0001,
+      "loss": 1.0262,
+      "step": 765
+    },
+    {
+      "epoch": 0.06560465912983898,
+      "grad_norm": 0.7049617171287537,
+      "learning_rate": 0.0001,
+      "loss": 1.3411,
+      "step": 766
+    },
+    {
+      "epoch": 0.065690304898938,
+      "grad_norm": 0.23054686188697815,
+      "learning_rate": 0.0001,
+      "loss": 0.2796,
+      "step": 767
+    },
+    {
+      "epoch": 0.065775950668037,
+      "grad_norm": 0.8812205791473389,
+      "learning_rate": 0.0001,
+      "loss": 0.7337,
+      "step": 768
+    },
+    {
+      "epoch": 0.065861596437136,
+      "grad_norm": 0.9103335738182068,
+      "learning_rate": 0.0001,
+      "loss": 0.8104,
+      "step": 769
+    },
+    {
+      "epoch": 0.06594724220623502,
+      "grad_norm": 0.3828534185886383,
+      "learning_rate": 0.0001,
+      "loss": 1.0763,
+      "step": 770
+    },
+    {
+      "epoch": 0.06603288797533402,
+      "grad_norm": 0.704101026058197,
+      "learning_rate": 0.0001,
+      "loss": 0.6878,
+      "step": 771
+    },
+    {
+      "epoch": 0.06611853374443302,
+      "grad_norm": 0.969886064529419,
+      "learning_rate": 0.0001,
+      "loss": 1.163,
+      "step": 772
+    },
+    {
+      "epoch": 0.06620417951353204,
+      "grad_norm": 0.731186032295227,
+      "learning_rate": 0.0001,
+      "loss": 1.1102,
+      "step": 773
+    },
+    {
+      "epoch": 0.06628982528263104,
+      "grad_norm": 0.7241857647895813,
+      "learning_rate": 0.0001,
+      "loss": 1.2541,
+      "step": 774
+    },
+    {
+      "epoch": 0.06637547105173004,
+      "grad_norm": 0.8718588352203369,
+      "learning_rate": 0.0001,
+      "loss": 1.0918,
+      "step": 775
+    },
+    {
+      "epoch": 0.06646111682082904,
+      "grad_norm": 0.7419017553329468,
+      "learning_rate": 0.0001,
+      "loss": 0.5329,
+      "step": 776
+    },
+    {
+      "epoch": 0.06654676258992806,
+      "grad_norm": 0.6680845618247986,
+      "learning_rate": 0.0001,
+      "loss": 0.8376,
+      "step": 777
+    },
+    {
+      "epoch": 0.06663240835902706,
+      "grad_norm": 0.5957100987434387,
+      "learning_rate": 0.0001,
+      "loss": 0.5792,
+      "step": 778
+    },
+    {
+      "epoch": 0.06671805412812606,
+      "grad_norm": 0.6547686457633972,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 779
+    },
+    {
+      "epoch": 0.06680369989722508,
+      "grad_norm": 0.8018797636032104,
+      "learning_rate": 0.0001,
+      "loss": 1.0717,
+      "step": 780
+    },
+    {
+      "epoch": 0.06688934566632408,
+      "grad_norm": 0.38116565346717834,
+      "learning_rate": 0.0001,
+      "loss": 0.7315,
+      "step": 781
+    },
+    {
+      "epoch": 0.06697499143542308,
+      "grad_norm": 0.48569175601005554,
+      "learning_rate": 0.0001,
+      "loss": 0.7583,
+      "step": 782
+    },
+    {
+      "epoch": 0.0670606372045221,
+      "grad_norm": 0.5432281494140625,
+      "learning_rate": 0.0001,
+      "loss": 0.7228,
+      "step": 783
+    },
+    {
+      "epoch": 0.0671462829736211,
+      "grad_norm": 0.6834201812744141,
+      "learning_rate": 0.0001,
+      "loss": 0.9193,
+      "step": 784
+    },
+    {
+      "epoch": 0.0672319287427201,
+      "grad_norm": 0.6835902333259583,
+      "learning_rate": 0.0001,
+      "loss": 0.7944,
+      "step": 785
+    },
+    {
+      "epoch": 0.06731757451181912,
+      "grad_norm": 0.3821081221103668,
+      "learning_rate": 0.0001,
+      "loss": 0.7964,
+      "step": 786
+    },
+    {
+      "epoch": 0.06740322028091812,
+      "grad_norm": 0.8369297981262207,
+      "learning_rate": 0.0001,
+      "loss": 1.0076,
+      "step": 787
+    },
+    {
+      "epoch": 0.06748886605001712,
+      "grad_norm": 0.5141885876655579,
+      "learning_rate": 0.0001,
+      "loss": 0.8779,
+      "step": 788
+    },
+    {
+      "epoch": 0.06757451181911614,
+      "grad_norm": 0.9722675681114197,
+      "learning_rate": 0.0001,
+      "loss": 1.133,
+      "step": 789
+    },
+    {
+      "epoch": 0.06766015758821514,
+      "grad_norm": 0.6366396546363831,
+      "learning_rate": 0.0001,
+      "loss": 1.2532,
+      "step": 790
+    },
+    {
+      "epoch": 0.06774580335731414,
+      "grad_norm": 0.4678216576576233,
+      "learning_rate": 0.0001,
+      "loss": 1.1393,
+      "step": 791
+    },
+    {
+      "epoch": 0.06783144912641316,
+      "grad_norm": 0.8692271113395691,
+      "learning_rate": 0.0001,
+      "loss": 1.8759,
+      "step": 792
+    },
+    {
+      "epoch": 0.06791709489551216,
+      "grad_norm": 0.8188161849975586,
+      "learning_rate": 0.0001,
+      "loss": 0.6768,
+      "step": 793
+    },
+    {
+      "epoch": 0.06800274066461116,
+      "grad_norm": 0.5889531373977661,
+      "learning_rate": 0.0001,
+      "loss": 0.7099,
+      "step": 794
+    },
+    {
+      "epoch": 0.06808838643371018,
+      "grad_norm": 0.41981014609336853,
+      "learning_rate": 0.0001,
+      "loss": 0.9419,
+      "step": 795
+    },
+    {
+      "epoch": 0.06817403220280918,
+      "grad_norm": 0.38716667890548706,
+      "learning_rate": 0.0001,
+      "loss": 0.8523,
+      "step": 796
+    },
+    {
+      "epoch": 0.06825967797190818,
+      "grad_norm": 1.367733120918274,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 797
+    },
+    {
+      "epoch": 0.0683453237410072,
+      "grad_norm": 0.5768160820007324,
+      "learning_rate": 0.0001,
+      "loss": 1.1274,
+      "step": 798
+    },
+    {
+      "epoch": 0.0684309695101062,
+      "grad_norm": 0.7135080695152283,
+      "learning_rate": 0.0001,
+      "loss": 0.9123,
+      "step": 799
+    },
+    {
+      "epoch": 0.0685166152792052,
+      "grad_norm": 0.7313363552093506,
+      "learning_rate": 0.0001,
+      "loss": 0.9921,
+      "step": 800
+    },
+    {
+      "epoch": 0.06860226104830422,
+      "grad_norm": 0.21910931169986725,
+      "learning_rate": 0.0001,
+      "loss": 0.6096,
+      "step": 801
+    },
+    {
+      "epoch": 0.06868790681740322,
+      "grad_norm": 0.6004666090011597,
+      "learning_rate": 0.0001,
+      "loss": 1.1416,
+      "step": 802
+    },
+    {
+      "epoch": 0.06877355258650222,
+      "grad_norm": 0.5246005058288574,
+      "learning_rate": 0.0001,
+      "loss": 0.8914,
+      "step": 803
+    },
+    {
+      "epoch": 0.06885919835560124,
+      "grad_norm": 0.5555751323699951,
+      "learning_rate": 0.0001,
+      "loss": 0.6826,
+      "step": 804
+    },
+    {
+      "epoch": 0.06894484412470024,
+      "grad_norm": 0.7161823511123657,
+      "learning_rate": 0.0001,
+      "loss": 0.9216,
+      "step": 805
+    },
+    {
+      "epoch": 0.06903048989379924,
+      "grad_norm": 0.8359095454216003,
+      "learning_rate": 0.0001,
+      "loss": 1.063,
+      "step": 806
+    },
+    {
+      "epoch": 0.06911613566289826,
+      "grad_norm": 0.4906369745731354,
+      "learning_rate": 0.0001,
+      "loss": 1.1167,
+      "step": 807
+    },
+    {
+      "epoch": 0.06920178143199726,
+      "grad_norm": 0.8382011651992798,
+      "learning_rate": 0.0001,
+      "loss": 0.8196,
+      "step": 808
+    },
+    {
+      "epoch": 0.06928742720109626,
+      "grad_norm": 0.741990864276886,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 809
+    },
+    {
+      "epoch": 0.06937307297019528,
+      "grad_norm": 0.4707683324813843,
+      "learning_rate": 0.0001,
+      "loss": 0.8979,
+      "step": 810
+    },
+    {
+      "epoch": 0.06945871873929428,
+      "grad_norm": 0.45771151781082153,
+      "learning_rate": 0.0001,
+      "loss": 0.895,
+      "step": 811
+    },
+    {
+      "epoch": 0.06954436450839328,
+      "grad_norm": 0.440483421087265,
+      "learning_rate": 0.0001,
+      "loss": 1.2474,
+      "step": 812
+    },
+    {
+      "epoch": 0.06963001027749229,
+      "grad_norm": 0.537780225276947,
+      "learning_rate": 0.0001,
+      "loss": 0.4896,
+      "step": 813
+    },
+    {
+      "epoch": 0.0697156560465913,
+      "grad_norm": 0.5036051869392395,
+      "learning_rate": 0.0001,
+      "loss": 1.1515,
+      "step": 814
+    },
+    {
+      "epoch": 0.0698013018156903,
+      "grad_norm": 0.5618871450424194,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 815
+    },
+    {
+      "epoch": 0.0698869475847893,
+      "grad_norm": 0.693444013595581,
+      "learning_rate": 0.0001,
+      "loss": 1.3048,
+      "step": 816
+    },
+    {
+      "epoch": 0.06997259335388832,
+      "grad_norm": 0.38126319646835327,
+      "learning_rate": 0.0001,
+      "loss": 0.6174,
+      "step": 817
+    },
+    {
+      "epoch": 0.07005823912298732,
+      "grad_norm": 0.923052191734314,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 818
+    },
+    {
+      "epoch": 0.07014388489208633,
+      "grad_norm": 0.5716583132743835,
+      "learning_rate": 0.0001,
+      "loss": 1.0039,
+      "step": 819
+    },
+    {
+      "epoch": 0.07022953066118534,
+      "grad_norm": 0.4875178933143616,
+      "learning_rate": 0.0001,
+      "loss": 0.8974,
+      "step": 820
+    },
+    {
+      "epoch": 0.07031517643028434,
+      "grad_norm": 0.6508381366729736,
+      "learning_rate": 0.0001,
+      "loss": 1.1639,
+      "step": 821
+    },
+    {
+      "epoch": 0.07040082219938334,
+      "grad_norm": 0.583893358707428,
+      "learning_rate": 0.0001,
+      "loss": 0.5684,
+      "step": 822
+    },
+    {
+      "epoch": 0.07048646796848236,
+      "grad_norm": 0.37333089113235474,
+      "learning_rate": 0.0001,
+      "loss": 1.0728,
+      "step": 823
+    },
+    {
+      "epoch": 0.07057211373758136,
+      "grad_norm": 0.6644774079322815,
+      "learning_rate": 0.0001,
+      "loss": 0.8195,
+      "step": 824
+    },
+    {
+      "epoch": 0.07065775950668036,
+      "grad_norm": 0.6433677077293396,
+      "learning_rate": 0.0001,
+      "loss": 0.9939,
+      "step": 825
+    },
+    {
+      "epoch": 0.07074340527577938,
+      "grad_norm": 0.8906939029693604,
+      "learning_rate": 0.0001,
+      "loss": 1.1392,
+      "step": 826
+    },
+    {
+      "epoch": 0.07082905104487838,
+      "grad_norm": 0.436755895614624,
+      "learning_rate": 0.0001,
+      "loss": 1.0758,
+      "step": 827
+    },
+    {
+      "epoch": 0.07091469681397738,
+      "grad_norm": 0.8091644048690796,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 828
+    },
+    {
+      "epoch": 0.0710003425830764,
+      "grad_norm": 0.5066967010498047,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 829
+    },
+    {
+      "epoch": 0.0710859883521754,
+      "grad_norm": 0.8127711415290833,
+      "learning_rate": 0.0001,
+      "loss": 0.8893,
+      "step": 830
+    },
+    {
+      "epoch": 0.0711716341212744,
+      "grad_norm": 0.4478503167629242,
+      "learning_rate": 0.0001,
+      "loss": 0.9564,
+      "step": 831
+    },
+    {
+      "epoch": 0.07125727989037342,
+      "grad_norm": 0.7653983235359192,
+      "learning_rate": 0.0001,
+      "loss": 0.7478,
+      "step": 832
+    },
+    {
+      "epoch": 0.07134292565947242,
+      "grad_norm": 0.5355436205863953,
+      "learning_rate": 0.0001,
+      "loss": 1.3645,
+      "step": 833
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 0.6038991808891296,
+      "learning_rate": 0.0001,
+      "loss": 0.6991,
+      "step": 834
+    },
+    {
+      "epoch": 0.07151421719767044,
+      "grad_norm": 0.6463876366615295,
+      "learning_rate": 0.0001,
+      "loss": 0.8556,
+      "step": 835
+    },
+    {
+      "epoch": 0.07159986296676944,
+      "grad_norm": 0.6774596571922302,
+      "learning_rate": 0.0001,
+      "loss": 0.773,
+      "step": 836
+    },
+    {
+      "epoch": 0.07168550873586844,
+      "grad_norm": 0.39566051959991455,
+      "learning_rate": 0.0001,
+      "loss": 0.7389,
+      "step": 837
+    },
+    {
+      "epoch": 0.07177115450496746,
+      "grad_norm": 0.387466698884964,
+      "learning_rate": 0.0001,
+      "loss": 0.643,
+      "step": 838
+    },
+    {
+      "epoch": 0.07185680027406646,
+      "grad_norm": 0.40937894582748413,
+      "learning_rate": 0.0001,
+      "loss": 0.5295,
+      "step": 839
+    },
+    {
+      "epoch": 0.07194244604316546,
+      "grad_norm": 0.9348215460777283,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 840
+    },
+    {
+      "epoch": 0.07202809181226448,
+      "grad_norm": 0.4981231689453125,
+      "learning_rate": 0.0001,
+      "loss": 1.1988,
+      "step": 841
+    },
+    {
+      "epoch": 0.07211373758136348,
+      "grad_norm": 0.4199872612953186,
+      "learning_rate": 0.0001,
+      "loss": 0.9243,
+      "step": 842
+    },
+    {
+      "epoch": 0.07219938335046248,
+      "grad_norm": 0.4927428066730499,
+      "learning_rate": 0.0001,
+      "loss": 1.2135,
+      "step": 843
+    },
+    {
+      "epoch": 0.0722850291195615,
+      "grad_norm": 0.25933536887168884,
+      "learning_rate": 0.0001,
+      "loss": 0.4757,
+      "step": 844
+    },
+    {
+      "epoch": 0.0723706748886605,
+      "grad_norm": 0.8011898994445801,
+      "learning_rate": 0.0001,
+      "loss": 0.6894,
+      "step": 845
+    },
+    {
+      "epoch": 0.0724563206577595,
+      "grad_norm": 0.40785232186317444,
+      "learning_rate": 0.0001,
+      "loss": 1.0742,
+      "step": 846
+    },
+    {
+      "epoch": 0.0725419664268585,
+      "grad_norm": 0.7812751531600952,
+      "learning_rate": 0.0001,
+      "loss": 0.7193,
+      "step": 847
+    },
+    {
+      "epoch": 0.07262761219595752,
+      "grad_norm": 0.946189820766449,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 848
+    },
+    {
+      "epoch": 0.07271325796505652,
+      "grad_norm": 0.41740885376930237,
+      "learning_rate": 0.0001,
+      "loss": 1.2399,
+      "step": 849
+    },
+    {
+      "epoch": 0.07279890373415553,
+      "grad_norm": 0.43415722250938416,
+      "learning_rate": 0.0001,
+      "loss": 0.8141,
+      "step": 850
+    },
+    {
+      "epoch": 0.07288454950325454,
+      "grad_norm": 0.9328998327255249,
+      "learning_rate": 0.0001,
+      "loss": 1.0011,
+      "step": 851
+    },
+    {
+      "epoch": 0.07297019527235354,
+      "grad_norm": 0.6934128999710083,
+      "learning_rate": 0.0001,
+      "loss": 0.8288,
+      "step": 852
+    },
+    {
+      "epoch": 0.07305584104145255,
+      "grad_norm": 0.758263111114502,
+      "learning_rate": 0.0001,
+      "loss": 1.0507,
+      "step": 853
+    },
+    {
+      "epoch": 0.07314148681055156,
+      "grad_norm": 0.7144489884376526,
+      "learning_rate": 0.0001,
+      "loss": 1.4524,
+      "step": 854
+    },
+    {
+      "epoch": 0.07322713257965056,
+      "grad_norm": 0.3901574909687042,
+      "learning_rate": 0.0001,
+      "loss": 1.0144,
+      "step": 855
+    },
+    {
+      "epoch": 0.07331277834874957,
+      "grad_norm": 0.7373013496398926,
+      "learning_rate": 0.0001,
+      "loss": 1.1106,
+      "step": 856
+    },
+    {
+      "epoch": 0.07339842411784858,
+      "grad_norm": 0.6429391503334045,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 857
+    },
+    {
+      "epoch": 0.07348406988694758,
+      "grad_norm": 0.6313352584838867,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 858
+    },
+    {
+      "epoch": 0.07356971565604659,
+      "grad_norm": 0.4766548275947571,
+      "learning_rate": 0.0001,
+      "loss": 0.9159,
+      "step": 859
+    },
+    {
+      "epoch": 0.0736553614251456,
+      "grad_norm": 0.4468488097190857,
+      "learning_rate": 0.0001,
+      "loss": 1.2077,
+      "step": 860
+    },
+    {
+      "epoch": 0.0737410071942446,
+      "grad_norm": 0.4127160906791687,
+      "learning_rate": 0.0001,
+      "loss": 0.9622,
+      "step": 861
+    },
+    {
+      "epoch": 0.0738266529633436,
+      "grad_norm": 0.7633454203605652,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 862
+    },
+    {
+      "epoch": 0.07391229873244262,
+      "grad_norm": 0.3900066316127777,
+      "learning_rate": 0.0001,
+      "loss": 0.9602,
+      "step": 863
+    },
+    {
+      "epoch": 0.07399794450154162,
+      "grad_norm": 0.5154496431350708,
+      "learning_rate": 0.0001,
+      "loss": 1.2783,
+      "step": 864
+    },
+    {
+      "epoch": 0.07408359027064063,
+      "grad_norm": 0.5742350220680237,
+      "learning_rate": 0.0001,
+      "loss": 1.2438,
+      "step": 865
+    },
+    {
+      "epoch": 0.07416923603973964,
+      "grad_norm": 0.4547298550605774,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 866
+    },
+    {
+      "epoch": 0.07425488180883864,
+      "grad_norm": 0.6464493870735168,
+      "learning_rate": 0.0001,
+      "loss": 0.8208,
+      "step": 867
+    },
+    {
+      "epoch": 0.07434052757793765,
+      "grad_norm": 0.775537371635437,
+      "learning_rate": 0.0001,
+      "loss": 0.9047,
+      "step": 868
+    },
+    {
+      "epoch": 0.07442617334703666,
+      "grad_norm": 0.49695518612861633,
+      "learning_rate": 0.0001,
+      "loss": 1.1954,
+      "step": 869
+    },
+    {
+      "epoch": 0.07451181911613566,
+      "grad_norm": 1.013782024383545,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 870
+    },
+    {
+      "epoch": 0.07459746488523467,
+      "grad_norm": 0.22894170880317688,
+      "learning_rate": 0.0001,
+      "loss": 0.5278,
+      "step": 871
+    },
+    {
+      "epoch": 0.07468311065433368,
+      "grad_norm": 0.49142855405807495,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 872
+    },
+    {
+      "epoch": 0.07476875642343268,
+      "grad_norm": 0.5611847043037415,
+      "learning_rate": 0.0001,
+      "loss": 1.1308,
+      "step": 873
+    },
+    {
+      "epoch": 0.07485440219253169,
+      "grad_norm": 0.4679000973701477,
+      "learning_rate": 0.0001,
+      "loss": 0.9727,
+      "step": 874
+    },
+    {
+      "epoch": 0.0749400479616307,
+      "grad_norm": 0.5103771686553955,
+      "learning_rate": 0.0001,
+      "loss": 0.6253,
+      "step": 875
+    },
+    {
+      "epoch": 0.0750256937307297,
+      "grad_norm": 0.69802325963974,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 876
+    },
+    {
+      "epoch": 0.0751113394998287,
+      "grad_norm": 1.296006202697754,
+      "learning_rate": 0.0001,
+      "loss": 0.9686,
+      "step": 877
+    },
+    {
+      "epoch": 0.07519698526892772,
+      "grad_norm": 0.4046657383441925,
+      "learning_rate": 0.0001,
+      "loss": 0.8203,
+      "step": 878
+    },
+    {
+      "epoch": 0.07528263103802672,
+      "grad_norm": 0.6303603649139404,
+      "learning_rate": 0.0001,
+      "loss": 1.1181,
+      "step": 879
+    },
+    {
+      "epoch": 0.07536827680712573,
+      "grad_norm": 0.4857475757598877,
+      "learning_rate": 0.0001,
+      "loss": 0.7241,
+      "step": 880
+    },
+    {
+      "epoch": 0.07545392257622474,
+      "grad_norm": 1.3066489696502686,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 881
+    },
+    {
+      "epoch": 0.07553956834532374,
+      "grad_norm": 0.4376460015773773,
+      "learning_rate": 0.0001,
+      "loss": 1.0616,
+      "step": 882
+    },
+    {
+      "epoch": 0.07562521411442275,
+      "grad_norm": 0.6235179305076599,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 883
+    },
+    {
+      "epoch": 0.07571085988352175,
+      "grad_norm": 0.615985095500946,
+      "learning_rate": 0.0001,
+      "loss": 0.6122,
+      "step": 884
+    },
+    {
+      "epoch": 0.07579650565262076,
+      "grad_norm": 0.5654960870742798,
+      "learning_rate": 0.0001,
+      "loss": 0.4799,
+      "step": 885
+    },
+    {
+      "epoch": 0.07588215142171977,
+      "grad_norm": 0.529895007610321,
+      "learning_rate": 0.0001,
+      "loss": 1.1926,
+      "step": 886
+    },
+    {
+      "epoch": 0.07596779719081877,
+      "grad_norm": 0.11664079129695892,
+      "learning_rate": 0.0001,
+      "loss": 0.6156,
+      "step": 887
+    },
+    {
+      "epoch": 0.07605344295991778,
+      "grad_norm": 0.21068789064884186,
+      "learning_rate": 0.0001,
+      "loss": 0.2947,
+      "step": 888
+    },
+    {
+      "epoch": 0.07613908872901679,
+      "grad_norm": 0.3726842999458313,
+      "learning_rate": 0.0001,
+      "loss": 0.536,
+      "step": 889
+    },
+    {
+      "epoch": 0.07622473449811579,
+      "grad_norm": 0.805418848991394,
+      "learning_rate": 0.0001,
+      "loss": 0.9484,
+      "step": 890
+    },
+    {
+      "epoch": 0.0763103802672148,
+      "grad_norm": 0.4783673882484436,
+      "learning_rate": 0.0001,
+      "loss": 1.0414,
+      "step": 891
+    },
+    {
+      "epoch": 0.0763960260363138,
+      "grad_norm": 0.6701018214225769,
+      "learning_rate": 0.0001,
+      "loss": 1.2352,
+      "step": 892
+    },
+    {
+      "epoch": 0.07648167180541281,
+      "grad_norm": 0.4860330820083618,
+      "learning_rate": 0.0001,
+      "loss": 1.1191,
+      "step": 893
+    },
+    {
+      "epoch": 0.07656731757451182,
+      "grad_norm": 0.45683127641677856,
+      "learning_rate": 0.0001,
+      "loss": 1.031,
+      "step": 894
+    },
+    {
+      "epoch": 0.07665296334361082,
+      "grad_norm": 0.5696792006492615,
+      "learning_rate": 0.0001,
+      "loss": 0.857,
+      "step": 895
+    },
+    {
+      "epoch": 0.07673860911270983,
+      "grad_norm": 0.7632311582565308,
+      "learning_rate": 0.0001,
+      "loss": 1.2269,
+      "step": 896
+    },
+    {
+      "epoch": 0.07682425488180884,
+      "grad_norm": 0.7396734356880188,
+      "learning_rate": 0.0001,
+      "loss": 0.7812,
+      "step": 897
+    },
+    {
+      "epoch": 0.07690990065090784,
+      "grad_norm": 0.5018053650856018,
+      "learning_rate": 0.0001,
+      "loss": 0.7166,
+      "step": 898
+    },
+    {
+      "epoch": 0.07699554642000685,
+      "grad_norm": 0.8555542826652527,
+      "learning_rate": 0.0001,
+      "loss": 0.8474,
+      "step": 899
+    },
+    {
+      "epoch": 0.07708119218910586,
+      "grad_norm": 0.8351611495018005,
+      "learning_rate": 0.0001,
+      "loss": 0.8971,
+      "step": 900
+    },
+    {
+      "epoch": 0.07716683795820486,
+      "grad_norm": 0.9218192100524902,
+      "learning_rate": 0.0001,
+      "loss": 0.79,
+      "step": 901
+    },
+    {
+      "epoch": 0.07725248372730387,
+      "grad_norm": 0.1965782344341278,
+      "learning_rate": 0.0001,
+      "loss": 0.3254,
+      "step": 902
+    },
+    {
+      "epoch": 0.07733812949640288,
+      "grad_norm": 0.6108563542366028,
+      "learning_rate": 0.0001,
+      "loss": 0.9455,
+      "step": 903
+    },
+    {
+      "epoch": 0.07742377526550188,
+      "grad_norm": 0.3751218318939209,
+      "learning_rate": 0.0001,
+      "loss": 0.6065,
+      "step": 904
+    },
+    {
+      "epoch": 0.07750942103460089,
+      "grad_norm": 0.6463737487792969,
+      "learning_rate": 0.0001,
+      "loss": 0.8039,
+      "step": 905
+    },
+    {
+      "epoch": 0.0775950668036999,
+      "grad_norm": 0.9913851618766785,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 906
+    },
+    {
+      "epoch": 0.0776807125727989,
+      "grad_norm": 0.33626440167427063,
+      "learning_rate": 0.0001,
+      "loss": 0.8875,
+      "step": 907
+    },
+    {
+      "epoch": 0.0777663583418979,
+      "grad_norm": 0.4942544400691986,
+      "learning_rate": 0.0001,
+      "loss": 1.2007,
+      "step": 908
+    },
+    {
+      "epoch": 0.07785200411099692,
+      "grad_norm": 0.7640135288238525,
+      "learning_rate": 0.0001,
+      "loss": 1.0435,
+      "step": 909
+    },
+    {
+      "epoch": 0.07793764988009592,
+      "grad_norm": 0.2753806412220001,
+      "learning_rate": 0.0001,
+      "loss": 0.3884,
+      "step": 910
+    },
+    {
+      "epoch": 0.07802329564919493,
+      "grad_norm": 0.6247597336769104,
+      "learning_rate": 0.0001,
+      "loss": 1.2545,
+      "step": 911
+    },
+    {
+      "epoch": 0.07810894141829394,
+      "grad_norm": 0.6455150842666626,
+      "learning_rate": 0.0001,
+      "loss": 0.6762,
+      "step": 912
+    },
+    {
+      "epoch": 0.07819458718739294,
+      "grad_norm": 0.4930016100406647,
+      "learning_rate": 0.0001,
+      "loss": 0.8216,
+      "step": 913
+    },
+    {
+      "epoch": 0.07828023295649195,
+      "grad_norm": 0.34173327684402466,
+      "learning_rate": 0.0001,
+      "loss": 1.0033,
+      "step": 914
+    },
+    {
+      "epoch": 0.07836587872559096,
+      "grad_norm": 0.5050802230834961,
+      "learning_rate": 0.0001,
+      "loss": 0.5085,
+      "step": 915
+    },
+    {
+      "epoch": 0.07845152449468996,
+      "grad_norm": 0.7302289605140686,
+      "learning_rate": 0.0001,
+      "loss": 0.7773,
+      "step": 916
+    },
+    {
+      "epoch": 0.07853717026378897,
+      "grad_norm": 0.9365038275718689,
+      "learning_rate": 0.0001,
+      "loss": 1.2478,
+      "step": 917
+    },
+    {
+      "epoch": 0.07862281603288798,
+      "grad_norm": 0.811701238155365,
+      "learning_rate": 0.0001,
+      "loss": 1.1504,
+      "step": 918
+    },
+    {
+      "epoch": 0.07870846180198698,
+      "grad_norm": 0.808403491973877,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 919
+    },
+    {
+      "epoch": 0.07879410757108599,
+      "grad_norm": 1.2311265468597412,
+      "learning_rate": 0.0001,
+      "loss": 1.1637,
+      "step": 920
+    },
+    {
+      "epoch": 0.07887975334018499,
+      "grad_norm": 0.8511678576469421,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 921
+    },
+    {
+      "epoch": 0.078965399109284,
+      "grad_norm": 0.6394311785697937,
+      "learning_rate": 0.0001,
+      "loss": 0.717,
+      "step": 922
+    },
+    {
+      "epoch": 0.079051044878383,
+      "grad_norm": 0.42978864908218384,
+      "learning_rate": 0.0001,
+      "loss": 0.6748,
+      "step": 923
+    },
+    {
+      "epoch": 0.07913669064748201,
+      "grad_norm": 0.7933021783828735,
+      "learning_rate": 0.0001,
+      "loss": 1.0266,
+      "step": 924
+    },
+    {
+      "epoch": 0.07922233641658102,
+      "grad_norm": 0.7150386571884155,
+      "learning_rate": 0.0001,
+      "loss": 0.8828,
+      "step": 925
+    },
+    {
+      "epoch": 0.07930798218568003,
+      "grad_norm": 0.25985783338546753,
+      "learning_rate": 0.0001,
+      "loss": 0.3507,
+      "step": 926
+    },
+    {
+      "epoch": 0.07939362795477903,
+      "grad_norm": 0.7921019196510315,
+      "learning_rate": 0.0001,
+      "loss": 0.8981,
+      "step": 927
+    },
+    {
+      "epoch": 0.07947927372387804,
+      "grad_norm": 0.3380737006664276,
+      "learning_rate": 0.0001,
+      "loss": 0.775,
+      "step": 928
+    },
+    {
+      "epoch": 0.07956491949297705,
+      "grad_norm": 0.3837883770465851,
+      "learning_rate": 0.0001,
+      "loss": 0.6864,
+      "step": 929
+    },
+    {
+      "epoch": 0.07965056526207605,
+      "grad_norm": 0.7232382297515869,
+      "learning_rate": 0.0001,
+      "loss": 0.6927,
+      "step": 930
+    },
+    {
+      "epoch": 0.07973621103117506,
+      "grad_norm": 0.6210137605667114,
+      "learning_rate": 0.0001,
+      "loss": 0.988,
+      "step": 931
+    },
+    {
+      "epoch": 0.07982185680027407,
+      "grad_norm": 0.4849126935005188,
+      "learning_rate": 0.0001,
+      "loss": 0.795,
+      "step": 932
+    },
+    {
+      "epoch": 0.07990750256937307,
+      "grad_norm": 0.7156875133514404,
+      "learning_rate": 0.0001,
+      "loss": 0.9895,
+      "step": 933
+    },
+    {
+      "epoch": 0.07999314833847208,
+      "grad_norm": 0.6183912754058838,
+      "learning_rate": 0.0001,
+      "loss": 1.0562,
+      "step": 934
+    },
+    {
+      "epoch": 0.08007879410757109,
+      "grad_norm": 0.5784626007080078,
+      "learning_rate": 0.0001,
+      "loss": 1.177,
+      "step": 935
+    },
+    {
+      "epoch": 0.08016443987667009,
+      "grad_norm": 0.08815716207027435,
+      "learning_rate": 0.0001,
+      "loss": 0.4954,
+      "step": 936
+    },
+    {
+      "epoch": 0.0802500856457691,
+      "grad_norm": 0.5837327837944031,
+      "learning_rate": 0.0001,
+      "loss": 0.8337,
+      "step": 937
+    },
+    {
+      "epoch": 0.0803357314148681,
+      "grad_norm": 0.4155324399471283,
+      "learning_rate": 0.0001,
+      "loss": 1.0432,
+      "step": 938
+    },
+    {
+      "epoch": 0.08042137718396711,
+      "grad_norm": 0.6682422757148743,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 939
+    },
+    {
+      "epoch": 0.08050702295306612,
+      "grad_norm": 0.5050842761993408,
+      "learning_rate": 0.0001,
+      "loss": 0.8681,
+      "step": 940
+    },
+    {
+      "epoch": 0.08059266872216513,
+      "grad_norm": 0.19968263804912567,
+      "learning_rate": 0.0001,
+      "loss": 0.3654,
+      "step": 941
+    },
+    {
+      "epoch": 0.08067831449126413,
+      "grad_norm": 0.5842058062553406,
+      "learning_rate": 0.0001,
+      "loss": 0.9531,
+      "step": 942
+    },
+    {
+      "epoch": 0.08076396026036314,
+      "grad_norm": 0.4657710790634155,
+      "learning_rate": 0.0001,
+      "loss": 0.9841,
+      "step": 943
+    },
+    {
+      "epoch": 0.08084960602946215,
+      "grad_norm": 0.8010748028755188,
+      "learning_rate": 0.0001,
+      "loss": 1.2836,
+      "step": 944
+    },
+    {
+      "epoch": 0.08093525179856115,
+      "grad_norm": 0.4060788154602051,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 945
+    },
+    {
+      "epoch": 0.08102089756766016,
+      "grad_norm": 0.5881567001342773,
+      "learning_rate": 0.0001,
+      "loss": 0.8339,
+      "step": 946
+    },
+    {
+      "epoch": 0.08110654333675917,
+      "grad_norm": 0.5082589983940125,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 947
+    },
+    {
+      "epoch": 0.08119218910585817,
+      "grad_norm": 0.8229041695594788,
+      "learning_rate": 0.0001,
+      "loss": 0.7343,
+      "step": 948
+    },
+    {
+      "epoch": 0.08127783487495718,
+      "grad_norm": 0.6522393822669983,
+      "learning_rate": 0.0001,
+      "loss": 0.6194,
+      "step": 949
+    },
+    {
+      "epoch": 0.08136348064405619,
+      "grad_norm": 0.6331807374954224,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 950
+    },
+    {
+      "epoch": 0.08144912641315519,
+      "grad_norm": 0.8660833239555359,
+      "learning_rate": 0.0001,
+      "loss": 0.8291,
+      "step": 951
+    },
+    {
+      "epoch": 0.0815347721822542,
+      "grad_norm": 0.2455417662858963,
+      "learning_rate": 0.0001,
+      "loss": 0.4575,
+      "step": 952
+    },
+    {
+      "epoch": 0.0816204179513532,
+      "grad_norm": 0.6345613598823547,
+      "learning_rate": 0.0001,
+      "loss": 0.6045,
+      "step": 953
+    },
+    {
+      "epoch": 0.08170606372045221,
+      "grad_norm": 0.48294079303741455,
+      "learning_rate": 0.0001,
+      "loss": 0.8398,
+      "step": 954
+    },
+    {
+      "epoch": 0.08179170948955121,
+      "grad_norm": 0.797317624092102,
+      "learning_rate": 0.0001,
+      "loss": 0.9908,
+      "step": 955
+    },
+    {
+      "epoch": 0.08187735525865023,
+      "grad_norm": 0.7200642228126526,
+      "learning_rate": 0.0001,
+      "loss": 0.9522,
+      "step": 956
+    },
+    {
+      "epoch": 0.08196300102774923,
+      "grad_norm": 0.5746588110923767,
+      "learning_rate": 0.0001,
+      "loss": 0.6211,
+      "step": 957
+    },
+    {
+      "epoch": 0.08204864679684823,
+      "grad_norm": 0.7502796649932861,
+      "learning_rate": 0.0001,
+      "loss": 0.7393,
+      "step": 958
+    },
+    {
+      "epoch": 0.08213429256594725,
+      "grad_norm": 0.7874694466590881,
+      "learning_rate": 0.0001,
+      "loss": 0.9129,
+      "step": 959
+    },
+    {
+      "epoch": 0.08221993833504625,
+      "grad_norm": 0.6751909255981445,
+      "learning_rate": 0.0001,
+      "loss": 0.9176,
+      "step": 960
+    },
+    {
+      "epoch": 0.08230558410414525,
+      "grad_norm": 0.6348690390586853,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 961
+    },
+    {
+      "epoch": 0.08239122987324426,
+      "grad_norm": 0.582118570804596,
+      "learning_rate": 0.0001,
+      "loss": 0.994,
+      "step": 962
+    },
+    {
+      "epoch": 0.08247687564234327,
+      "grad_norm": 0.6422193050384521,
+      "learning_rate": 0.0001,
+      "loss": 0.9124,
+      "step": 963
+    },
+    {
+      "epoch": 0.08256252141144227,
+      "grad_norm": 0.6998811960220337,
+      "learning_rate": 0.0001,
+      "loss": 1.1487,
+      "step": 964
+    },
+    {
+      "epoch": 0.08264816718054128,
+      "grad_norm": 0.7625442147254944,
+      "learning_rate": 0.0001,
+      "loss": 1.1922,
+      "step": 965
+    },
+    {
+      "epoch": 0.08273381294964029,
+      "grad_norm": 0.5084379315376282,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 966
+    },
+    {
+      "epoch": 0.08281945871873929,
+      "grad_norm": 0.7954206466674805,
+      "learning_rate": 0.0001,
+      "loss": 0.9603,
+      "step": 967
+    },
+    {
+      "epoch": 0.0829051044878383,
+      "grad_norm": 0.48077091574668884,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 968
+    },
+    {
+      "epoch": 0.0829907502569373,
+      "grad_norm": 0.7081925868988037,
+      "learning_rate": 0.0001,
+      "loss": 0.8163,
+      "step": 969
+    },
+    {
+      "epoch": 0.08307639602603631,
+      "grad_norm": 0.6027596592903137,
+      "learning_rate": 0.0001,
+      "loss": 1.0785,
+      "step": 970
+    },
+    {
+      "epoch": 0.08316204179513532,
+      "grad_norm": 0.5575534105300903,
+      "learning_rate": 0.0001,
+      "loss": 0.5956,
+      "step": 971
+    },
+    {
+      "epoch": 0.08324768756423433,
+      "grad_norm": 0.4398977756500244,
+      "learning_rate": 0.0001,
+      "loss": 0.4854,
+      "step": 972
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.6633754372596741,
+      "learning_rate": 0.0001,
+      "loss": 1.151,
+      "step": 973
+    },
+    {
+      "epoch": 0.08341897910243234,
+      "grad_norm": 0.5305631160736084,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 974
+    },
+    {
+      "epoch": 0.08350462487153135,
+      "grad_norm": 0.6855217218399048,
+      "learning_rate": 0.0001,
+      "loss": 0.5522,
+      "step": 975
+    },
+    {
+      "epoch": 0.08359027064063035,
+      "grad_norm": 1.059149980545044,
+      "learning_rate": 0.0001,
+      "loss": 1.0219,
+      "step": 976
+    },
+    {
+      "epoch": 0.08367591640972936,
+      "grad_norm": 0.722783088684082,
+      "learning_rate": 0.0001,
+      "loss": 0.7206,
+      "step": 977
+    },
+    {
+      "epoch": 0.08376156217882837,
+      "grad_norm": 0.9453747868537903,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 978
+    },
+    {
+      "epoch": 0.08384720794792737,
+      "grad_norm": 0.7011775374412537,
+      "learning_rate": 0.0001,
+      "loss": 1.0847,
+      "step": 979
+    },
+    {
+      "epoch": 0.08393285371702638,
+      "grad_norm": 0.5231440663337708,
+      "learning_rate": 0.0001,
+      "loss": 0.8279,
+      "step": 980
+    },
+    {
+      "epoch": 0.08401849948612539,
+      "grad_norm": 0.4079996943473816,
+      "learning_rate": 0.0001,
+      "loss": 1.0393,
+      "step": 981
+    },
+    {
+      "epoch": 0.08410414525522439,
+      "grad_norm": 0.7875049710273743,
+      "learning_rate": 0.0001,
+      "loss": 1.0859,
+      "step": 982
+    },
+    {
+      "epoch": 0.0841897910243234,
+      "grad_norm": 0.4530707597732544,
+      "learning_rate": 0.0001,
+      "loss": 1.3592,
+      "step": 983
+    },
+    {
+      "epoch": 0.0842754367934224,
+      "grad_norm": 0.4627353847026825,
+      "learning_rate": 0.0001,
+      "loss": 0.4144,
+      "step": 984
+    },
+    {
+      "epoch": 0.08436108256252141,
+      "grad_norm": 0.5388820171356201,
+      "learning_rate": 0.0001,
+      "loss": 0.7394,
+      "step": 985
+    },
+    {
+      "epoch": 0.08444672833162042,
+      "grad_norm": 0.5398839116096497,
+      "learning_rate": 0.0001,
+      "loss": 0.8194,
+      "step": 986
+    },
+    {
+      "epoch": 0.08453237410071943,
+      "grad_norm": 0.4179549515247345,
+      "learning_rate": 0.0001,
+      "loss": 0.5044,
+      "step": 987
+    },
+    {
+      "epoch": 0.08461801986981843,
+      "grad_norm": 0.5723353028297424,
+      "learning_rate": 0.0001,
+      "loss": 0.793,
+      "step": 988
+    },
+    {
+      "epoch": 0.08470366563891744,
+      "grad_norm": 0.30284568667411804,
+      "learning_rate": 0.0001,
+      "loss": 0.5974,
+      "step": 989
+    },
+    {
+      "epoch": 0.08478931140801645,
+      "grad_norm": 1.0655642747879028,
+      "learning_rate": 0.0001,
+      "loss": 1.2158,
+      "step": 990
+    },
+    {
+      "epoch": 0.08487495717711545,
+      "grad_norm": 0.5462648272514343,
+      "learning_rate": 0.0001,
+      "loss": 1.0001,
+      "step": 991
+    },
+    {
+      "epoch": 0.08496060294621445,
+      "grad_norm": 0.6394854187965393,
+      "learning_rate": 0.0001,
+      "loss": 0.5899,
+      "step": 992
+    },
+    {
+      "epoch": 0.08504624871531347,
+      "grad_norm": 0.39151835441589355,
+      "learning_rate": 0.0001,
+      "loss": 1.1296,
+      "step": 993
+    },
+    {
+      "epoch": 0.08513189448441247,
+      "grad_norm": 0.48908641934394836,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 994
+    },
+    {
+      "epoch": 0.08521754025351147,
+      "grad_norm": 0.31017032265663147,
+      "learning_rate": 0.0001,
+      "loss": 0.6812,
+      "step": 995
+    },
+    {
+      "epoch": 0.08530318602261049,
+      "grad_norm": 0.5440806150436401,
+      "learning_rate": 0.0001,
+      "loss": 1.0022,
+      "step": 996
+    },
+    {
+      "epoch": 0.08538883179170949,
+      "grad_norm": 0.8314542174339294,
+      "learning_rate": 0.0001,
+      "loss": 1.2286,
+      "step": 997
+    },
+    {
+      "epoch": 0.08547447756080849,
+      "grad_norm": 0.5683814287185669,
+      "learning_rate": 0.0001,
+      "loss": 0.5549,
+      "step": 998
+    },
+    {
+      "epoch": 0.0855601233299075,
+      "grad_norm": 0.34978365898132324,
+      "learning_rate": 0.0001,
+      "loss": 0.8306,
+      "step": 999
+    },
+    {
+      "epoch": 0.08564576909900651,
+      "grad_norm": 0.4802309572696686,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08573141486810551,
+      "grad_norm": 0.5553489327430725,
+      "learning_rate": 0.0001,
+      "loss": 0.629,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08581706063720453,
+      "grad_norm": 1.5385195016860962,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08590270640630353,
+      "grad_norm": 0.6714608073234558,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08598835217540253,
+      "grad_norm": 0.38380175828933716,
+      "learning_rate": 0.0001,
+      "loss": 0.7672,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08607399794450155,
+      "grad_norm": 0.4402846395969391,
+      "learning_rate": 0.0001,
+      "loss": 1.4101,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08615964371360055,
+      "grad_norm": 0.6550090312957764,
+      "learning_rate": 0.0001,
+      "loss": 1.0464,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08624528948269955,
+      "grad_norm": 0.7128333449363708,
+      "learning_rate": 0.0001,
+      "loss": 0.8737,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08633093525179857,
+      "grad_norm": 0.5751252770423889,
+      "learning_rate": 0.0001,
+      "loss": 0.4642,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08641658102089757,
+      "grad_norm": 0.43736574053764343,
+      "learning_rate": 0.0001,
+      "loss": 1.082,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08650222678999657,
+      "grad_norm": 1.003411054611206,
+      "learning_rate": 0.0001,
+      "loss": 0.7381,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08658787255909559,
+      "grad_norm": 0.4926562011241913,
+      "learning_rate": 0.0001,
+      "loss": 0.493,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08667351832819459,
+      "grad_norm": 1.0493180751800537,
+      "learning_rate": 0.0001,
+      "loss": 1.2176,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08675916409729359,
+      "grad_norm": 1.0703368186950684,
+      "learning_rate": 0.0001,
+      "loss": 1.0039,
+      "step": 1013
+    },
+    {
+      "epoch": 0.0868448098663926,
+      "grad_norm": 0.4535515606403351,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08693045563549161,
+      "grad_norm": 0.4421178996562958,
+      "learning_rate": 0.0001,
+      "loss": 1.1808,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08701610140459061,
+      "grad_norm": 0.6676187515258789,
+      "learning_rate": 0.0001,
+      "loss": 1.3869,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08710174717368963,
+      "grad_norm": 0.5785860419273376,
+      "learning_rate": 0.0001,
+      "loss": 1.1044,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08718739294278863,
+      "grad_norm": 0.656802773475647,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08727303871188763,
+      "grad_norm": 0.5320791602134705,
+      "learning_rate": 0.0001,
+      "loss": 1.0011,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08735868448098665,
+      "grad_norm": 0.6286681890487671,
+      "learning_rate": 0.0001,
+      "loss": 0.9386,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08744433025008565,
+      "grad_norm": 0.5750426054000854,
+      "learning_rate": 0.0001,
+      "loss": 1.1998,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08752997601918465,
+      "grad_norm": 0.6157073378562927,
+      "learning_rate": 0.0001,
+      "loss": 0.8775,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08761562178828367,
+      "grad_norm": 0.725510835647583,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08770126755738267,
+      "grad_norm": 0.32301175594329834,
+      "learning_rate": 0.0001,
+      "loss": 0.4786,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08778691332648167,
+      "grad_norm": 0.6885251998901367,
+      "learning_rate": 0.0001,
+      "loss": 0.9331,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08787255909558069,
+      "grad_norm": 0.7670713067054749,
+      "learning_rate": 0.0001,
+      "loss": 0.7893,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08795820486467969,
+      "grad_norm": 0.48184487223625183,
+      "learning_rate": 0.0001,
+      "loss": 0.7305,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08804385063377869,
+      "grad_norm": 0.39736372232437134,
+      "learning_rate": 0.0001,
+      "loss": 1.0192,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08812949640287769,
+      "grad_norm": 0.4631947875022888,
+      "learning_rate": 0.0001,
+      "loss": 0.5315,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08821514217197671,
+      "grad_norm": 0.5988773107528687,
+      "learning_rate": 0.0001,
+      "loss": 0.7688,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08830078794107571,
+      "grad_norm": 0.961003839969635,
+      "learning_rate": 0.0001,
+      "loss": 0.9081,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08838643371017471,
+      "grad_norm": 0.6528297066688538,
+      "learning_rate": 0.0001,
+      "loss": 1.2283,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08847207947927373,
+      "grad_norm": 0.8059707880020142,
+      "learning_rate": 0.0001,
+      "loss": 1.007,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08855772524837273,
+      "grad_norm": 0.20696330070495605,
+      "learning_rate": 0.0001,
+      "loss": 0.5793,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08864337101747173,
+      "grad_norm": 0.5308125615119934,
+      "learning_rate": 0.0001,
+      "loss": 1.2185,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08872901678657075,
+      "grad_norm": 0.8629925847053528,
+      "learning_rate": 0.0001,
+      "loss": 0.8356,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08881466255566975,
+      "grad_norm": 0.5124247670173645,
+      "learning_rate": 0.0001,
+      "loss": 0.9764,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08890030832476875,
+      "grad_norm": 0.8700270056724548,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08898595409386777,
+      "grad_norm": 0.6915403604507446,
+      "learning_rate": 0.0001,
+      "loss": 1.2004,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08907159986296677,
+      "grad_norm": 0.34258151054382324,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08915724563206577,
+      "grad_norm": 0.5206171274185181,
+      "learning_rate": 0.0001,
+      "loss": 0.9549,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08924289140116479,
+      "grad_norm": 0.5871332287788391,
+      "learning_rate": 0.0001,
+      "loss": 0.4629,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08932853717026379,
+      "grad_norm": 0.5201041102409363,
+      "learning_rate": 0.0001,
+      "loss": 1.0076,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08941418293936279,
+      "grad_norm": 0.7643946409225464,
+      "learning_rate": 0.0001,
+      "loss": 1.2086,
+      "step": 1044
+    },
+    {
+      "epoch": 0.0894998287084618,
+      "grad_norm": 0.4797765016555786,
+      "learning_rate": 0.0001,
+      "loss": 0.8049,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08958547447756081,
+      "grad_norm": 0.38143137097358704,
+      "learning_rate": 0.0001,
+      "loss": 0.5887,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08967112024665981,
+      "grad_norm": 0.525643527507782,
+      "learning_rate": 0.0001,
+      "loss": 1.1671,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08975676601575883,
+      "grad_norm": 0.632914662361145,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08984241178485783,
+      "grad_norm": 0.7107246518135071,
+      "learning_rate": 0.0001,
+      "loss": 0.8028,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08992805755395683,
+      "grad_norm": 0.5453270077705383,
+      "learning_rate": 0.0001,
+      "loss": 0.8611,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09001370332305585,
+      "grad_norm": 0.6207929253578186,
+      "learning_rate": 0.0001,
+      "loss": 1.0366,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09009934909215485,
+      "grad_norm": 0.6974309086799622,
+      "learning_rate": 0.0001,
+      "loss": 0.8802,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09018499486125385,
+      "grad_norm": 1.0271036624908447,
+      "learning_rate": 0.0001,
+      "loss": 1.2698,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09027064063035287,
+      "grad_norm": 0.9124029874801636,
+      "learning_rate": 0.0001,
+      "loss": 1.1997,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09035628639945187,
+      "grad_norm": 1.055849313735962,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09044193216855087,
+      "grad_norm": 0.7013433575630188,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09052757793764989,
+      "grad_norm": 0.18892507255077362,
+      "learning_rate": 0.0001,
+      "loss": 0.6758,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09061322370674889,
+      "grad_norm": 0.5814532041549683,
+      "learning_rate": 0.0001,
+      "loss": 0.923,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09069886947584789,
+      "grad_norm": 0.5401561260223389,
+      "learning_rate": 0.0001,
+      "loss": 0.6752,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0907845152449469,
+      "grad_norm": 1.2662837505340576,
+      "learning_rate": 0.0001,
+      "loss": 0.9886,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09087016101404591,
+      "grad_norm": 0.3990923762321472,
+      "learning_rate": 0.0001,
+      "loss": 1.1314,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09095580678314491,
+      "grad_norm": 1.0023123025894165,
+      "learning_rate": 0.0001,
+      "loss": 1.3506,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09104145255224393,
+      "grad_norm": 0.38437455892562866,
+      "learning_rate": 0.0001,
+      "loss": 1.0053,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09112709832134293,
+      "grad_norm": 0.44476234912872314,
+      "learning_rate": 0.0001,
+      "loss": 0.9432,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09121274409044193,
+      "grad_norm": 0.6860992908477783,
+      "learning_rate": 0.0001,
+      "loss": 1.1317,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09129838985954093,
+      "grad_norm": 0.7838432788848877,
+      "learning_rate": 0.0001,
+      "loss": 1.0786,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09138403562863995,
+      "grad_norm": 0.5170823931694031,
+      "learning_rate": 0.0001,
+      "loss": 1.3232,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09146968139773895,
+      "grad_norm": 0.5883544087409973,
+      "learning_rate": 0.0001,
+      "loss": 0.5774,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09155532716683795,
+      "grad_norm": 0.3745083808898926,
+      "learning_rate": 0.0001,
+      "loss": 0.5885,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09164097293593697,
+      "grad_norm": 0.5732969045639038,
+      "learning_rate": 0.0001,
+      "loss": 0.5303,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09172661870503597,
+      "grad_norm": 0.5663641095161438,
+      "learning_rate": 0.0001,
+      "loss": 0.8374,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09181226447413497,
+      "grad_norm": 1.1236745119094849,
+      "learning_rate": 0.0001,
+      "loss": 1.3413,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09189791024323399,
+      "grad_norm": 0.4893132746219635,
+      "learning_rate": 0.0001,
+      "loss": 0.7943,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09198355601233299,
+      "grad_norm": 0.7285077571868896,
+      "learning_rate": 0.0001,
+      "loss": 1.0871,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09206920178143199,
+      "grad_norm": 0.543105959892273,
+      "learning_rate": 0.0001,
+      "loss": 0.5856,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09215484755053101,
+      "grad_norm": 0.4857051968574524,
+      "learning_rate": 0.0001,
+      "loss": 0.7489,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09224049331963001,
+      "grad_norm": 0.4956655204296112,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09232613908872901,
+      "grad_norm": 0.7534564137458801,
+      "learning_rate": 0.0001,
+      "loss": 1.3809,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09241178485782803,
+      "grad_norm": 0.4704369306564331,
+      "learning_rate": 0.0001,
+      "loss": 0.8219,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09249743062692703,
+      "grad_norm": 0.3899940252304077,
+      "learning_rate": 0.0001,
+      "loss": 0.7871,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09258307639602603,
+      "grad_norm": 0.3113993704319,
+      "learning_rate": 0.0001,
+      "loss": 0.3584,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09266872216512505,
+      "grad_norm": 0.5223938226699829,
+      "learning_rate": 0.0001,
+      "loss": 1.2898,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09275436793422405,
+      "grad_norm": 0.494926780462265,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09284001370332305,
+      "grad_norm": 0.479973167181015,
+      "learning_rate": 0.0001,
+      "loss": 0.9199,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09292565947242207,
+      "grad_norm": 0.46554115414619446,
+      "learning_rate": 0.0001,
+      "loss": 1.2022,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09301130524152107,
+      "grad_norm": 0.6514759063720703,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09309695101062007,
+      "grad_norm": 0.5015600919723511,
+      "learning_rate": 0.0001,
+      "loss": 1.0071,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09318259677971909,
+      "grad_norm": 0.868553876876831,
+      "learning_rate": 0.0001,
+      "loss": 0.9557,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09326824254881809,
+      "grad_norm": 0.2644060552120209,
+      "learning_rate": 0.0001,
+      "loss": 0.799,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09335388831791709,
+      "grad_norm": 0.8650840520858765,
+      "learning_rate": 0.0001,
+      "loss": 1.2096,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09343953408701611,
+      "grad_norm": 0.701171875,
+      "learning_rate": 0.0001,
+      "loss": 0.5286,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09352517985611511,
+      "grad_norm": 1.0450388193130493,
+      "learning_rate": 0.0001,
+      "loss": 0.9166,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09361082562521411,
+      "grad_norm": 0.4745027422904968,
+      "learning_rate": 0.0001,
+      "loss": 1.2519,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09369647139431313,
+      "grad_norm": 0.4651343822479248,
+      "learning_rate": 0.0001,
+      "loss": 0.6758,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09378211716341213,
+      "grad_norm": 0.40313720703125,
+      "learning_rate": 0.0001,
+      "loss": 0.8609,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09386776293251113,
+      "grad_norm": 0.491588294506073,
+      "learning_rate": 0.0001,
+      "loss": 0.7351,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09395340870161015,
+      "grad_norm": 0.8162128925323486,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09403905447070915,
+      "grad_norm": 0.5428500771522522,
+      "learning_rate": 0.0001,
+      "loss": 1.0117,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09412470023980815,
+      "grad_norm": 0.8340853452682495,
+      "learning_rate": 0.0001,
+      "loss": 1.0266,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09421034600890715,
+      "grad_norm": 0.8068060874938965,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09429599177800617,
+      "grad_norm": 0.45487019419670105,
+      "learning_rate": 0.0001,
+      "loss": 1.1164,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09438163754710517,
+      "grad_norm": 0.6453647017478943,
+      "learning_rate": 0.0001,
+      "loss": 0.6399,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09446728331620417,
+      "grad_norm": 0.9477840065956116,
+      "learning_rate": 0.0001,
+      "loss": 0.9658,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09455292908530319,
+      "grad_norm": 0.4841224253177643,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09463857485440219,
+      "grad_norm": 0.44776827096939087,
+      "learning_rate": 0.0001,
+      "loss": 1.0369,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09472422062350119,
+      "grad_norm": 0.8516103625297546,
+      "learning_rate": 0.0001,
+      "loss": 0.9537,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09480986639260021,
+      "grad_norm": 0.4632716476917267,
+      "learning_rate": 0.0001,
+      "loss": 1.212,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09489551216169921,
+      "grad_norm": 0.5816608667373657,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09498115793079821,
+      "grad_norm": 0.6027618050575256,
+      "learning_rate": 0.0001,
+      "loss": 1.1689,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09506680369989723,
+      "grad_norm": 0.7434003353118896,
+      "learning_rate": 0.0001,
+      "loss": 1.196,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09515244946899623,
+      "grad_norm": 0.7487145662307739,
+      "learning_rate": 0.0001,
+      "loss": 1.0591,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.4608497619628906,
+      "learning_rate": 0.0001,
+      "loss": 0.7294,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09532374100719425,
+      "grad_norm": 0.3505392372608185,
+      "learning_rate": 0.0001,
+      "loss": 0.5315,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09540938677629325,
+      "grad_norm": 0.6243942379951477,
+      "learning_rate": 0.0001,
+      "loss": 0.6795,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09549503254539225,
+      "grad_norm": 0.6839644908905029,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09558067831449127,
+      "grad_norm": 0.7432211637496948,
+      "learning_rate": 0.0001,
+      "loss": 0.6622,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09566632408359027,
+      "grad_norm": 0.5899722576141357,
+      "learning_rate": 0.0001,
+      "loss": 1.1789,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09575196985268927,
+      "grad_norm": 0.6684199571609497,
+      "learning_rate": 0.0001,
+      "loss": 0.9692,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09583761562178829,
+      "grad_norm": 0.5850124359130859,
+      "learning_rate": 0.0001,
+      "loss": 1.0739,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09592326139088729,
+      "grad_norm": 0.5711740255355835,
+      "learning_rate": 0.0001,
+      "loss": 1.0178,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09600890715998629,
+      "grad_norm": 0.4995405972003937,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09609455292908531,
+      "grad_norm": 0.7545071840286255,
+      "learning_rate": 0.0001,
+      "loss": 0.9584,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09618019869818431,
+      "grad_norm": 0.35175928473472595,
+      "learning_rate": 0.0001,
+      "loss": 1.0893,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09626584446728331,
+      "grad_norm": 0.4317012131214142,
+      "learning_rate": 0.0001,
+      "loss": 0.6434,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09635149023638233,
+      "grad_norm": 0.6555061340332031,
+      "learning_rate": 0.0001,
+      "loss": 0.9048,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09643713600548133,
+      "grad_norm": 0.6775338649749756,
+      "learning_rate": 0.0001,
+      "loss": 1.1383,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09652278177458033,
+      "grad_norm": 0.8341861367225647,
+      "learning_rate": 0.0001,
+      "loss": 0.8058,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09660842754367935,
+      "grad_norm": 0.7120161056518555,
+      "learning_rate": 0.0001,
+      "loss": 1.9225,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09669407331277835,
+      "grad_norm": 0.3624630868434906,
+      "learning_rate": 0.0001,
+      "loss": 0.7629,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09677971908187735,
+      "grad_norm": 0.7108566164970398,
+      "learning_rate": 0.0001,
+      "loss": 0.9664,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09686536485097637,
+      "grad_norm": 0.6705402135848999,
+      "learning_rate": 0.0001,
+      "loss": 1.045,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09695101062007537,
+      "grad_norm": 0.8236182332038879,
+      "learning_rate": 0.0001,
+      "loss": 1.0983,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09703665638917437,
+      "grad_norm": 0.606749415397644,
+      "learning_rate": 0.0001,
+      "loss": 0.5686,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09712230215827339,
+      "grad_norm": 0.9091976881027222,
+      "learning_rate": 0.0001,
+      "loss": 2.0971,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09720794792737239,
+      "grad_norm": 0.5604475736618042,
+      "learning_rate": 0.0001,
+      "loss": 0.4142,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09729359369647139,
+      "grad_norm": 0.6072295308113098,
+      "learning_rate": 0.0001,
+      "loss": 1.1889,
+      "step": 1136
+    },
+    {
+      "epoch": 0.0973792394655704,
+      "grad_norm": 0.2327422797679901,
+      "learning_rate": 0.0001,
+      "loss": 0.7373,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09746488523466941,
+      "grad_norm": 0.4745529592037201,
+      "learning_rate": 0.0001,
+      "loss": 0.9717,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09755053100376841,
+      "grad_norm": 0.8615824580192566,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09763617677286741,
+      "grad_norm": 0.760391354560852,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09772182254196643,
+      "grad_norm": 1.037918210029602,
+      "learning_rate": 0.0001,
+      "loss": 1.0449,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09780746831106543,
+      "grad_norm": 0.5313969254493713,
+      "learning_rate": 0.0001,
+      "loss": 0.7027,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09789311408016443,
+      "grad_norm": 0.4745279848575592,
+      "learning_rate": 0.0001,
+      "loss": 0.5339,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09797875984926345,
+      "grad_norm": 0.6396012306213379,
+      "learning_rate": 0.0001,
+      "loss": 0.5883,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09806440561836245,
+      "grad_norm": 0.5977548956871033,
+      "learning_rate": 0.0001,
+      "loss": 0.5388,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09815005138746145,
+      "grad_norm": 0.24692459404468536,
+      "learning_rate": 0.0001,
+      "loss": 0.363,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09823569715656047,
+      "grad_norm": 0.7153545022010803,
+      "learning_rate": 0.0001,
+      "loss": 0.855,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09832134292565947,
+      "grad_norm": 0.4216623902320862,
+      "learning_rate": 0.0001,
+      "loss": 0.6691,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09840698869475847,
+      "grad_norm": 0.8085505366325378,
+      "learning_rate": 0.0001,
+      "loss": 1.1802,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09849263446385749,
+      "grad_norm": 0.8618689179420471,
+      "learning_rate": 0.0001,
+      "loss": 0.813,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09857828023295649,
+      "grad_norm": 0.7679622769355774,
+      "learning_rate": 0.0001,
+      "loss": 1.0172,
+      "step": 1151
+    },
+    {
+      "epoch": 0.0986639260020555,
+      "grad_norm": 0.8656859397888184,
+      "learning_rate": 0.0001,
+      "loss": 1.2656,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09874957177115451,
+      "grad_norm": 0.4888617694377899,
+      "learning_rate": 0.0001,
+      "loss": 1.0625,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09883521754025351,
+      "grad_norm": 0.5504559278488159,
+      "learning_rate": 0.0001,
+      "loss": 1.0883,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09892086330935251,
+      "grad_norm": 0.5380454063415527,
+      "learning_rate": 0.0001,
+      "loss": 0.6963,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09900650907845153,
+      "grad_norm": 0.7862430810928345,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09909215484755053,
+      "grad_norm": 0.5755479335784912,
+      "learning_rate": 0.0001,
+      "loss": 0.7606,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09917780061664953,
+      "grad_norm": 0.6811529397964478,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09926344638574855,
+      "grad_norm": 0.6169347763061523,
+      "learning_rate": 0.0001,
+      "loss": 0.6163,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09934909215484755,
+      "grad_norm": 0.7338910698890686,
+      "learning_rate": 0.0001,
+      "loss": 0.9105,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09943473792394655,
+      "grad_norm": 0.83475661277771,
+      "learning_rate": 0.0001,
+      "loss": 0.8255,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09952038369304557,
+      "grad_norm": 0.8817697167396545,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09960602946214457,
+      "grad_norm": 1.0031250715255737,
+      "learning_rate": 0.0001,
+      "loss": 0.7886,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09969167523124357,
+      "grad_norm": 0.6480106115341187,
+      "learning_rate": 0.0001,
+      "loss": 0.9562,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09977732100034259,
+      "grad_norm": 1.001660943031311,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09986296676944159,
+      "grad_norm": 0.5806012749671936,
+      "learning_rate": 0.0001,
+      "loss": 0.6835,
+      "step": 1166
+    },
+    {
+      "epoch": 0.0999486125385406,
+      "grad_norm": 0.5753374099731445,
+      "learning_rate": 0.0001,
+      "loss": 0.6934,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10003425830763961,
+      "grad_norm": 0.5635197162628174,
+      "learning_rate": 0.0001,
+      "loss": 0.6821,
+      "step": 1168
+    },
+    {
+      "epoch": 0.10011990407673861,
+      "grad_norm": 1.0802620649337769,
+      "learning_rate": 0.0001,
+      "loss": 1.2135,
+      "step": 1169
+    },
+    {
+      "epoch": 0.10020554984583761,
+      "grad_norm": 1.0480722188949585,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10029119561493663,
+      "grad_norm": 0.9768354296684265,
+      "learning_rate": 0.0001,
+      "loss": 0.8054,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10037684138403563,
+      "grad_norm": 0.18759919703006744,
+      "learning_rate": 0.0001,
+      "loss": 0.1775,
+      "step": 1172
+    },
+    {
+      "epoch": 0.10046248715313463,
+      "grad_norm": 0.6751762628555298,
+      "learning_rate": 0.0001,
+      "loss": 0.7431,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10054813292223364,
+      "grad_norm": 0.7398316860198975,
+      "learning_rate": 0.0001,
+      "loss": 0.9685,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10063377869133265,
+      "grad_norm": 0.853391170501709,
+      "learning_rate": 0.0001,
+      "loss": 1.3789,
+      "step": 1175
+    },
+    {
+      "epoch": 0.10071942446043165,
+      "grad_norm": 0.7244229316711426,
+      "learning_rate": 0.0001,
+      "loss": 0.8311,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10080507022953065,
+      "grad_norm": 0.5136174559593201,
+      "learning_rate": 0.0001,
+      "loss": 1.0987,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10089071599862967,
+      "grad_norm": 0.6271736025810242,
+      "learning_rate": 0.0001,
+      "loss": 0.8637,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10097636176772867,
+      "grad_norm": 0.6499254703521729,
+      "learning_rate": 0.0001,
+      "loss": 0.8429,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10106200753682767,
+      "grad_norm": 0.41788041591644287,
+      "learning_rate": 0.0001,
+      "loss": 0.654,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10114765330592669,
+      "grad_norm": 0.686737596988678,
+      "learning_rate": 0.0001,
+      "loss": 1.421,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10123329907502569,
+      "grad_norm": 0.4611888527870178,
+      "learning_rate": 0.0001,
+      "loss": 0.7299,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1013189448441247,
+      "grad_norm": 0.598088800907135,
+      "learning_rate": 0.0001,
+      "loss": 0.4814,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10140459061322371,
+      "grad_norm": 0.7726507186889648,
+      "learning_rate": 0.0001,
+      "loss": 0.8903,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10149023638232271,
+      "grad_norm": 0.6883323788642883,
+      "learning_rate": 0.0001,
+      "loss": 0.675,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10157588215142171,
+      "grad_norm": 0.587283194065094,
+      "learning_rate": 0.0001,
+      "loss": 0.9694,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10166152792052073,
+      "grad_norm": 0.5839325785636902,
+      "learning_rate": 0.0001,
+      "loss": 0.5147,
+      "step": 1187
+    },
+    {
+      "epoch": 0.10174717368961973,
+      "grad_norm": 0.5665099024772644,
+      "learning_rate": 0.0001,
+      "loss": 1.1773,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10183281945871873,
+      "grad_norm": 0.8837037086486816,
+      "learning_rate": 0.0001,
+      "loss": 0.8946,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10191846522781775,
+      "grad_norm": 0.896073579788208,
+      "learning_rate": 0.0001,
+      "loss": 0.7755,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10200411099691675,
+      "grad_norm": 0.6847692728042603,
+      "learning_rate": 0.0001,
+      "loss": 0.8667,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10208975676601575,
+      "grad_norm": 0.6889893412590027,
+      "learning_rate": 0.0001,
+      "loss": 0.7599,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10217540253511477,
+      "grad_norm": 0.39392268657684326,
+      "learning_rate": 0.0001,
+      "loss": 0.9515,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10226104830421377,
+      "grad_norm": 0.6163459420204163,
+      "learning_rate": 0.0001,
+      "loss": 0.7111,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10234669407331277,
+      "grad_norm": 0.6816166043281555,
+      "learning_rate": 0.0001,
+      "loss": 0.8339,
+      "step": 1195
+    },
+    {
+      "epoch": 0.10243233984241179,
+      "grad_norm": 0.4214668273925781,
+      "learning_rate": 0.0001,
+      "loss": 1.3078,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10251798561151079,
+      "grad_norm": 0.21523253619670868,
+      "learning_rate": 0.0001,
+      "loss": 0.3044,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1026036313806098,
+      "grad_norm": 0.6668765544891357,
+      "learning_rate": 0.0001,
+      "loss": 0.7527,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10268927714970881,
+      "grad_norm": 1.071770429611206,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 1199
+    },
+    {
+      "epoch": 0.10277492291880781,
+      "grad_norm": 0.4275430738925934,
+      "learning_rate": 0.0001,
+      "loss": 0.675,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10286056868790681,
+      "grad_norm": 0.44320356845855713,
+      "learning_rate": 0.0001,
+      "loss": 0.8697,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10294621445700583,
+      "grad_norm": 0.8827096223831177,
+      "learning_rate": 0.0001,
+      "loss": 1.2913,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10303186022610483,
+      "grad_norm": 0.4796529710292816,
+      "learning_rate": 0.0001,
+      "loss": 0.7041,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10311750599520383,
+      "grad_norm": 0.5892266631126404,
+      "learning_rate": 0.0001,
+      "loss": 0.549,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10320315176430285,
+      "grad_norm": 0.4565364122390747,
+      "learning_rate": 0.0001,
+      "loss": 1.3068,
+      "step": 1205
+    },
+    {
+      "epoch": 0.10328879753340185,
+      "grad_norm": 0.5942059755325317,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10337444330250085,
+      "grad_norm": 0.5390171408653259,
+      "learning_rate": 0.0001,
+      "loss": 0.6432,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10346008907159986,
+      "grad_norm": 0.4844352900981903,
+      "learning_rate": 0.0001,
+      "loss": 0.8859,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10354573484069887,
+      "grad_norm": 0.5696431994438171,
+      "learning_rate": 0.0001,
+      "loss": 0.5725,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10363138060979787,
+      "grad_norm": 0.9452969431877136,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10371702637889688,
+      "grad_norm": 0.3883347809314728,
+      "learning_rate": 0.0001,
+      "loss": 0.8294,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10380267214799589,
+      "grad_norm": 0.6587196588516235,
+      "learning_rate": 0.0001,
+      "loss": 0.7225,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1038883179170949,
+      "grad_norm": 0.6393395662307739,
+      "learning_rate": 0.0001,
+      "loss": 0.5665,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1039739636861939,
+      "grad_norm": 0.6027940511703491,
+      "learning_rate": 0.0001,
+      "loss": 0.5581,
+      "step": 1214
+    },
+    {
+      "epoch": 0.10405960945529291,
+      "grad_norm": 0.36656683683395386,
+      "learning_rate": 0.0001,
+      "loss": 0.8359,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10414525522439191,
+      "grad_norm": 0.5477709770202637,
+      "learning_rate": 0.0001,
+      "loss": 1.0677,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10423090099349092,
+      "grad_norm": 0.4814382791519165,
+      "learning_rate": 0.0001,
+      "loss": 0.3831,
+      "step": 1217
+    },
+    {
+      "epoch": 0.10431654676258993,
+      "grad_norm": 0.4923788905143738,
+      "learning_rate": 0.0001,
+      "loss": 0.6096,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10440219253168893,
+      "grad_norm": 0.4698505401611328,
+      "learning_rate": 0.0001,
+      "loss": 0.7709,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10448783830078794,
+      "grad_norm": 0.4595884084701538,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10457348406988695,
+      "grad_norm": 0.432646244764328,
+      "learning_rate": 0.0001,
+      "loss": 0.8691,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10465912983898595,
+      "grad_norm": 0.6978726387023926,
+      "learning_rate": 0.0001,
+      "loss": 1.0274,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10474477560808496,
+      "grad_norm": 0.6730514168739319,
+      "learning_rate": 0.0001,
+      "loss": 1.0043,
+      "step": 1223
+    },
+    {
+      "epoch": 0.10483042137718397,
+      "grad_norm": 0.544624388217926,
+      "learning_rate": 0.0001,
+      "loss": 0.9672,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10491606714628297,
+      "grad_norm": 0.7848732471466064,
+      "learning_rate": 0.0001,
+      "loss": 0.9658,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10500171291538198,
+      "grad_norm": 0.6367408037185669,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10508735868448099,
+      "grad_norm": 0.4748637080192566,
+      "learning_rate": 0.0001,
+      "loss": 1.1011,
+      "step": 1227
+    },
+    {
+      "epoch": 0.10517300445358,
+      "grad_norm": 0.6201930046081543,
+      "learning_rate": 0.0001,
+      "loss": 0.5333,
+      "step": 1228
+    },
+    {
+      "epoch": 0.105258650222679,
+      "grad_norm": 0.5208135843276978,
+      "learning_rate": 0.0001,
+      "loss": 0.6165,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10534429599177801,
+      "grad_norm": 0.39293739199638367,
+      "learning_rate": 0.0001,
+      "loss": 0.9986,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10542994176087701,
+      "grad_norm": 0.33050405979156494,
+      "learning_rate": 0.0001,
+      "loss": 0.7894,
+      "step": 1231
+    },
+    {
+      "epoch": 0.10551558752997602,
+      "grad_norm": 0.868964433670044,
+      "learning_rate": 0.0001,
+      "loss": 1.0026,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10560123329907503,
+      "grad_norm": 1.0397356748580933,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10568687906817403,
+      "grad_norm": 0.6320386528968811,
+      "learning_rate": 0.0001,
+      "loss": 0.9594,
+      "step": 1234
+    },
+    {
+      "epoch": 0.10577252483727304,
+      "grad_norm": 0.5339903235435486,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10585817060637205,
+      "grad_norm": 0.4601982831954956,
+      "learning_rate": 0.0001,
+      "loss": 0.7432,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10594381637547105,
+      "grad_norm": 0.5617156028747559,
+      "learning_rate": 0.0001,
+      "loss": 1.3226,
+      "step": 1237
+    },
+    {
+      "epoch": 0.10602946214457006,
+      "grad_norm": 0.8237478137016296,
+      "learning_rate": 0.0001,
+      "loss": 1.3267,
+      "step": 1238
+    },
+    {
+      "epoch": 0.10611510791366907,
+      "grad_norm": 0.7386891841888428,
+      "learning_rate": 0.0001,
+      "loss": 1.1447,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10620075368276807,
+      "grad_norm": 0.625260591506958,
+      "learning_rate": 0.0001,
+      "loss": 0.9139,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10628639945186708,
+      "grad_norm": 1.1350523233413696,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10637204522096609,
+      "grad_norm": 0.7634811401367188,
+      "learning_rate": 0.0001,
+      "loss": 0.8812,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1064576909900651,
+      "grad_norm": 0.5141278505325317,
+      "learning_rate": 0.0001,
+      "loss": 0.5734,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1065433367591641,
+      "grad_norm": 0.308571994304657,
+      "learning_rate": 0.0001,
+      "loss": 0.4454,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1066289825282631,
+      "grad_norm": 0.6633331179618835,
+      "learning_rate": 0.0001,
+      "loss": 0.979,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10671462829736211,
+      "grad_norm": 0.667979896068573,
+      "learning_rate": 0.0001,
+      "loss": 0.5717,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10680027406646111,
+      "grad_norm": 0.5181378722190857,
+      "learning_rate": 0.0001,
+      "loss": 0.9548,
+      "step": 1247
+    },
+    {
+      "epoch": 0.10688591983556012,
+      "grad_norm": 0.42694932222366333,
+      "learning_rate": 0.0001,
+      "loss": 1.0143,
+      "step": 1248
+    },
+    {
+      "epoch": 0.10697156560465913,
+      "grad_norm": 0.4613799750804901,
+      "learning_rate": 0.0001,
+      "loss": 1.0578,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10705721137375813,
+      "grad_norm": 0.702510416507721,
+      "learning_rate": 0.0001,
+      "loss": 0.7592,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 0.8924612402915955,
+      "learning_rate": 0.0001,
+      "loss": 0.8601,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10722850291195615,
+      "grad_norm": 1.1231863498687744,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10731414868105515,
+      "grad_norm": 0.5224771499633789,
+      "learning_rate": 0.0001,
+      "loss": 0.51,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10739979445015416,
+      "grad_norm": 0.7032666206359863,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 1254
+    },
+    {
+      "epoch": 0.10748544021925317,
+      "grad_norm": 0.7852336168289185,
+      "learning_rate": 0.0001,
+      "loss": 0.761,
+      "step": 1255
+    },
+    {
+      "epoch": 0.10757108598835217,
+      "grad_norm": 0.945933997631073,
+      "learning_rate": 0.0001,
+      "loss": 1.0549,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10765673175745118,
+      "grad_norm": 0.5108488202095032,
+      "learning_rate": 0.0001,
+      "loss": 1.1143,
+      "step": 1257
+    },
+    {
+      "epoch": 0.10774237752655019,
+      "grad_norm": 0.5425330996513367,
+      "learning_rate": 0.0001,
+      "loss": 0.5075,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1078280232956492,
+      "grad_norm": 0.6220599412918091,
+      "learning_rate": 0.0001,
+      "loss": 1.1989,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1079136690647482,
+      "grad_norm": 0.6614335179328918,
+      "learning_rate": 0.0001,
+      "loss": 0.64,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10799931483384721,
+      "grad_norm": 0.8392595648765564,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 1261
+    },
+    {
+      "epoch": 0.10808496060294621,
+      "grad_norm": 0.8706468939781189,
+      "learning_rate": 0.0001,
+      "loss": 0.8187,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10817060637204522,
+      "grad_norm": 0.9878280162811279,
+      "learning_rate": 0.0001,
+      "loss": 1.8837,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10825625214114423,
+      "grad_norm": 0.46001100540161133,
+      "learning_rate": 0.0001,
+      "loss": 0.7285,
+      "step": 1264
+    },
+    {
+      "epoch": 0.10834189791024323,
+      "grad_norm": 0.5787230134010315,
+      "learning_rate": 0.0001,
+      "loss": 0.9819,
+      "step": 1265
+    },
+    {
+      "epoch": 0.10842754367934224,
+      "grad_norm": 1.8707797527313232,
+      "learning_rate": 0.0001,
+      "loss": 0.7982,
+      "step": 1266
+    },
+    {
+      "epoch": 0.10851318944844125,
+      "grad_norm": 0.5431042313575745,
+      "learning_rate": 0.0001,
+      "loss": 1.1837,
+      "step": 1267
+    },
+    {
+      "epoch": 0.10859883521754025,
+      "grad_norm": 0.2986743450164795,
+      "learning_rate": 0.0001,
+      "loss": 0.5691,
+      "step": 1268
+    },
+    {
+      "epoch": 0.10868448098663926,
+      "grad_norm": 0.8680964112281799,
+      "learning_rate": 0.0001,
+      "loss": 1.9334,
+      "step": 1269
+    },
+    {
+      "epoch": 0.10877012675573827,
+      "grad_norm": 0.6668152213096619,
+      "learning_rate": 0.0001,
+      "loss": 0.6486,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10885577252483727,
+      "grad_norm": 0.14952519536018372,
+      "learning_rate": 0.0001,
+      "loss": 1.0083,
+      "step": 1271
+    },
+    {
+      "epoch": 0.10894141829393628,
+      "grad_norm": 0.14151696860790253,
+      "learning_rate": 0.0001,
+      "loss": 0.7986,
+      "step": 1272
+    },
+    {
+      "epoch": 0.10902706406303529,
+      "grad_norm": 0.6089078187942505,
+      "learning_rate": 0.0001,
+      "loss": 0.8945,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1091127098321343,
+      "grad_norm": 0.4123326539993286,
+      "learning_rate": 0.0001,
+      "loss": 1.1154,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1091983556012333,
+      "grad_norm": 0.9819856882095337,
+      "learning_rate": 0.0001,
+      "loss": 1.4257,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10928400137033231,
+      "grad_norm": 0.499267578125,
+      "learning_rate": 0.0001,
+      "loss": 1.05,
+      "step": 1276
+    },
+    {
+      "epoch": 0.10936964713943131,
+      "grad_norm": 0.7107354998588562,
+      "learning_rate": 0.0001,
+      "loss": 0.8518,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10945529290853032,
+      "grad_norm": 0.2929224669933319,
+      "learning_rate": 0.0001,
+      "loss": 0.3604,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10954093867762933,
+      "grad_norm": 0.4627382457256317,
+      "learning_rate": 0.0001,
+      "loss": 1.0035,
+      "step": 1279
+    },
+    {
+      "epoch": 0.10962658444672833,
+      "grad_norm": 0.4891338050365448,
+      "learning_rate": 0.0001,
+      "loss": 1.1881,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10971223021582734,
+      "grad_norm": 0.769540011882782,
+      "learning_rate": 0.0001,
+      "loss": 0.7136,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10979787598492634,
+      "grad_norm": 0.5845102071762085,
+      "learning_rate": 0.0001,
+      "loss": 1.1824,
+      "step": 1282
+    },
+    {
+      "epoch": 0.10988352175402535,
+      "grad_norm": 1.0055838823318481,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 1283
+    },
+    {
+      "epoch": 0.10996916752312436,
+      "grad_norm": 0.6747437119483948,
+      "learning_rate": 0.0001,
+      "loss": 0.8322,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11005481329222336,
+      "grad_norm": 0.6497412323951721,
+      "learning_rate": 0.0001,
+      "loss": 0.4978,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11014045906132237,
+      "grad_norm": 0.822253942489624,
+      "learning_rate": 0.0001,
+      "loss": 1.0196,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11022610483042138,
+      "grad_norm": 0.5153818726539612,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11031175059952038,
+      "grad_norm": 0.7231460809707642,
+      "learning_rate": 0.0001,
+      "loss": 0.9364,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1103973963686194,
+      "grad_norm": 0.661533534526825,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1104830421377184,
+      "grad_norm": 0.1828005164861679,
+      "learning_rate": 0.0001,
+      "loss": 1.0989,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1105686879068174,
+      "grad_norm": 1.0465052127838135,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11065433367591641,
+      "grad_norm": 0.4696097671985626,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11073997944501542,
+      "grad_norm": 0.3275883197784424,
+      "learning_rate": 0.0001,
+      "loss": 0.5029,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11082562521411442,
+      "grad_norm": 0.5031739473342896,
+      "learning_rate": 0.0001,
+      "loss": 1.0417,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11091127098321343,
+      "grad_norm": 1.0856668949127197,
+      "learning_rate": 0.0001,
+      "loss": 1.0177,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11099691675231244,
+      "grad_norm": 0.6928914189338684,
+      "learning_rate": 0.0001,
+      "loss": 0.7627,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11108256252141144,
+      "grad_norm": 0.70926833152771,
+      "learning_rate": 0.0001,
+      "loss": 0.9833,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11116820829051045,
+      "grad_norm": 0.6918324828147888,
+      "learning_rate": 0.0001,
+      "loss": 1.2266,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11125385405960946,
+      "grad_norm": 0.5835555791854858,
+      "learning_rate": 0.0001,
+      "loss": 0.8974,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11133949982870846,
+      "grad_norm": 0.7588602900505066,
+      "learning_rate": 0.0001,
+      "loss": 1.0488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11142514559780747,
+      "grad_norm": 0.6016455888748169,
+      "learning_rate": 0.0001,
+      "loss": 1.0413,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11151079136690648,
+      "grad_norm": 0.3280949890613556,
+      "learning_rate": 0.0001,
+      "loss": 0.4939,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11159643713600548,
+      "grad_norm": 0.595888078212738,
+      "learning_rate": 0.0001,
+      "loss": 0.3107,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1116820829051045,
+      "grad_norm": 1.0846636295318604,
+      "learning_rate": 0.0001,
+      "loss": 0.7777,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1117677286742035,
+      "grad_norm": 0.3594922721385956,
+      "learning_rate": 0.0001,
+      "loss": 0.601,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1118533744433025,
+      "grad_norm": 0.8278983235359192,
+      "learning_rate": 0.0001,
+      "loss": 1.3429,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11193902021240151,
+      "grad_norm": 0.5420074462890625,
+      "learning_rate": 0.0001,
+      "loss": 0.9031,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11202466598150052,
+      "grad_norm": 0.6183144450187683,
+      "learning_rate": 0.0001,
+      "loss": 1.2285,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11211031175059952,
+      "grad_norm": 0.4622572660446167,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11219595751969853,
+      "grad_norm": 0.8625423908233643,
+      "learning_rate": 0.0001,
+      "loss": 0.9869,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11228160328879754,
+      "grad_norm": 0.8596449494361877,
+      "learning_rate": 0.0001,
+      "loss": 1.0736,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11236724905789654,
+      "grad_norm": 0.41659095883369446,
+      "learning_rate": 0.0001,
+      "loss": 0.9848,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11245289482699555,
+      "grad_norm": 0.510106086730957,
+      "learning_rate": 0.0001,
+      "loss": 0.8472,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11253854059609456,
+      "grad_norm": 0.9060420989990234,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11262418636519356,
+      "grad_norm": 0.9440112709999084,
+      "learning_rate": 0.0001,
+      "loss": 1.2351,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11270983213429256,
+      "grad_norm": 0.6746245622634888,
+      "learning_rate": 0.0001,
+      "loss": 0.7442,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11279547790339157,
+      "grad_norm": 0.45736032724380493,
+      "learning_rate": 0.0001,
+      "loss": 0.8459,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11288112367249058,
+      "grad_norm": 0.557775616645813,
+      "learning_rate": 0.0001,
+      "loss": 1.3392,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11296676944158958,
+      "grad_norm": 0.49701645970344543,
+      "learning_rate": 0.0001,
+      "loss": 0.8793,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1130524152106886,
+      "grad_norm": 0.4296928346157074,
+      "learning_rate": 0.0001,
+      "loss": 1.0301,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1131380609797876,
+      "grad_norm": 0.40025362372398376,
+      "learning_rate": 0.0001,
+      "loss": 1.1349,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1132237067488866,
+      "grad_norm": 0.5786082148551941,
+      "learning_rate": 0.0001,
+      "loss": 0.8444,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11330935251798561,
+      "grad_norm": 0.5063909888267517,
+      "learning_rate": 0.0001,
+      "loss": 0.4564,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11339499828708462,
+      "grad_norm": 0.8922147750854492,
+      "learning_rate": 0.0001,
+      "loss": 1.0265,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11348064405618362,
+      "grad_norm": 0.5282382965087891,
+      "learning_rate": 0.0001,
+      "loss": 1.244,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11356628982528263,
+      "grad_norm": 0.5968488454818726,
+      "learning_rate": 0.0001,
+      "loss": 0.9075,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11365193559438164,
+      "grad_norm": 0.4750228226184845,
+      "learning_rate": 0.0001,
+      "loss": 1.0072,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11373758136348064,
+      "grad_norm": 0.6573008894920349,
+      "learning_rate": 0.0001,
+      "loss": 0.6771,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11382322713257965,
+      "grad_norm": 0.7809916734695435,
+      "learning_rate": 0.0001,
+      "loss": 0.7494,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11390887290167866,
+      "grad_norm": 0.2609162926673889,
+      "learning_rate": 0.0001,
+      "loss": 0.4723,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11399451867077766,
+      "grad_norm": 0.6242702603340149,
+      "learning_rate": 0.0001,
+      "loss": 0.7397,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11408016443987667,
+      "grad_norm": 0.7627384662628174,
+      "learning_rate": 0.0001,
+      "loss": 0.7838,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11416581020897568,
+      "grad_norm": 0.6918596029281616,
+      "learning_rate": 0.0001,
+      "loss": 0.8956,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11425145597807468,
+      "grad_norm": 0.20347988605499268,
+      "learning_rate": 0.0001,
+      "loss": 0.9246,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1143371017471737,
+      "grad_norm": 0.6751242280006409,
+      "learning_rate": 0.0001,
+      "loss": 0.8999,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1144227475162727,
+      "grad_norm": 0.6350564360618591,
+      "learning_rate": 0.0001,
+      "loss": 1.0632,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1145083932853717,
+      "grad_norm": 0.43810877203941345,
+      "learning_rate": 0.0001,
+      "loss": 1.0987,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11459403905447071,
+      "grad_norm": 0.6207727193832397,
+      "learning_rate": 0.0001,
+      "loss": 0.6884,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11467968482356972,
+      "grad_norm": 0.6860010027885437,
+      "learning_rate": 0.0001,
+      "loss": 0.8252,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11476533059266872,
+      "grad_norm": 0.7951822280883789,
+      "learning_rate": 0.0001,
+      "loss": 1.059,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11485097636176773,
+      "grad_norm": 0.49357670545578003,
+      "learning_rate": 0.0001,
+      "loss": 0.6785,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11493662213086674,
+      "grad_norm": 0.13679493963718414,
+      "learning_rate": 0.0001,
+      "loss": 0.8002,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11502226789996574,
+      "grad_norm": 0.4053069055080414,
+      "learning_rate": 0.0001,
+      "loss": 1.0413,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11510791366906475,
+      "grad_norm": 0.3280472755432129,
+      "learning_rate": 0.0001,
+      "loss": 0.4709,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11519355943816376,
+      "grad_norm": 0.498248815536499,
+      "learning_rate": 0.0001,
+      "loss": 0.6195,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11527920520726276,
+      "grad_norm": 0.7977742552757263,
+      "learning_rate": 0.0001,
+      "loss": 0.7236,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11536485097636177,
+      "grad_norm": 0.5906994938850403,
+      "learning_rate": 0.0001,
+      "loss": 1.041,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11545049674546078,
+      "grad_norm": 0.7648127675056458,
+      "learning_rate": 0.0001,
+      "loss": 1.0521,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11553614251455978,
+      "grad_norm": 0.7071904540061951,
+      "learning_rate": 0.0001,
+      "loss": 1.0978,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1156217882836588,
+      "grad_norm": 0.5561211109161377,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1157074340527578,
+      "grad_norm": 1.0050745010375977,
+      "learning_rate": 0.0001,
+      "loss": 1.6751,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1157930798218568,
+      "grad_norm": 0.5044364929199219,
+      "learning_rate": 0.0001,
+      "loss": 0.9794,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1158787255909558,
+      "grad_norm": 0.7034612894058228,
+      "learning_rate": 0.0001,
+      "loss": 0.6483,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11596437136005482,
+      "grad_norm": 0.6140638589859009,
+      "learning_rate": 0.0001,
+      "loss": 1.0113,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11605001712915382,
+      "grad_norm": 0.59046870470047,
+      "learning_rate": 0.0001,
+      "loss": 0.778,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11613566289825282,
+      "grad_norm": 0.6099108457565308,
+      "learning_rate": 0.0001,
+      "loss": 0.9818,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11622130866735184,
+      "grad_norm": 0.8091556429862976,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11630695443645084,
+      "grad_norm": 0.28905460238456726,
+      "learning_rate": 0.0001,
+      "loss": 0.5794,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11639260020554984,
+      "grad_norm": 0.8140268921852112,
+      "learning_rate": 0.0001,
+      "loss": 1.0364,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11647824597464886,
+      "grad_norm": 0.4717259407043457,
+      "learning_rate": 0.0001,
+      "loss": 0.9767,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11656389174374786,
+      "grad_norm": 0.8462554216384888,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11664953751284686,
+      "grad_norm": 0.5557670593261719,
+      "learning_rate": 0.0001,
+      "loss": 1.212,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11673518328194588,
+      "grad_norm": 0.6853979825973511,
+      "learning_rate": 0.0001,
+      "loss": 1.2738,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11682082905104488,
+      "grad_norm": 0.7876192331314087,
+      "learning_rate": 0.0001,
+      "loss": 0.8879,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11690647482014388,
+      "grad_norm": 0.6769497394561768,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1169921205892429,
+      "grad_norm": 0.9178217649459839,
+      "learning_rate": 0.0001,
+      "loss": 0.9161,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1170777663583419,
+      "grad_norm": 0.46165111660957336,
+      "learning_rate": 0.0001,
+      "loss": 1.1554,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1171634121274409,
+      "grad_norm": 0.5163291692733765,
+      "learning_rate": 0.0001,
+      "loss": 0.7503,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11724905789653992,
+      "grad_norm": 0.6218660473823547,
+      "learning_rate": 0.0001,
+      "loss": 1.1964,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11733470366563892,
+      "grad_norm": 0.47357305884361267,
+      "learning_rate": 0.0001,
+      "loss": 1.2336,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11742034943473792,
+      "grad_norm": 0.7580575942993164,
+      "learning_rate": 0.0001,
+      "loss": 1.0259,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11750599520383694,
+      "grad_norm": 0.837584912776947,
+      "learning_rate": 0.0001,
+      "loss": 0.9114,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11759164097293594,
+      "grad_norm": 0.19466304779052734,
+      "learning_rate": 0.0001,
+      "loss": 0.3811,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11767728674203494,
+      "grad_norm": 0.4720872938632965,
+      "learning_rate": 0.0001,
+      "loss": 0.9588,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11776293251113396,
+      "grad_norm": 0.6699342727661133,
+      "learning_rate": 0.0001,
+      "loss": 0.8006,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11784857828023296,
+      "grad_norm": 0.6279401779174805,
+      "learning_rate": 0.0001,
+      "loss": 1.0167,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11793422404933196,
+      "grad_norm": 0.9356536269187927,
+      "learning_rate": 0.0001,
+      "loss": 1.1778,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11801986981843098,
+      "grad_norm": 0.40872490406036377,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11810551558752998,
+      "grad_norm": 0.8230023384094238,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11819116135662898,
+      "grad_norm": 0.6694610714912415,
+      "learning_rate": 0.0001,
+      "loss": 0.6359,
+      "step": 1380
+    },
+    {
+      "epoch": 0.118276807125728,
+      "grad_norm": 0.7068077921867371,
+      "learning_rate": 0.0001,
+      "loss": 0.7733,
+      "step": 1381
+    },
+    {
+      "epoch": 0.118362452894827,
+      "grad_norm": 0.5020363330841064,
+      "learning_rate": 0.0001,
+      "loss": 1.146,
+      "step": 1382
+    },
+    {
+      "epoch": 0.118448098663926,
+      "grad_norm": 0.9289647340774536,
+      "learning_rate": 0.0001,
+      "loss": 1.0827,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11853374443302502,
+      "grad_norm": 0.4397963881492615,
+      "learning_rate": 0.0001,
+      "loss": 0.7295,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11861939020212402,
+      "grad_norm": 0.7010247111320496,
+      "learning_rate": 0.0001,
+      "loss": 0.6164,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11870503597122302,
+      "grad_norm": 0.8232924342155457,
+      "learning_rate": 0.0001,
+      "loss": 0.7411,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11879068174032203,
+      "grad_norm": 0.42585521936416626,
+      "learning_rate": 0.0001,
+      "loss": 0.7539,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11887632750942104,
+      "grad_norm": 0.5260921716690063,
+      "learning_rate": 0.0001,
+      "loss": 0.965,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11896197327852004,
+      "grad_norm": 0.21219849586486816,
+      "learning_rate": 0.0001,
+      "loss": 0.4537,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11904761904761904,
+      "grad_norm": 0.4833400845527649,
+      "learning_rate": 0.0001,
+      "loss": 0.5087,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11913326481671806,
+      "grad_norm": 0.39503902196884155,
+      "learning_rate": 0.0001,
+      "loss": 0.4833,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11921891058581706,
+      "grad_norm": 0.32731807231903076,
+      "learning_rate": 0.0001,
+      "loss": 0.526,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11930455635491606,
+      "grad_norm": 0.7116717100143433,
+      "learning_rate": 0.0001,
+      "loss": 0.717,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11939020212401508,
+      "grad_norm": 0.729756236076355,
+      "learning_rate": 0.0001,
+      "loss": 0.6374,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11947584789311408,
+      "grad_norm": 0.46690213680267334,
+      "learning_rate": 0.0001,
+      "loss": 0.7784,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11956149366221308,
+      "grad_norm": 0.643465518951416,
+      "learning_rate": 0.0001,
+      "loss": 0.5444,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1196471394313121,
+      "grad_norm": 0.9403294920921326,
+      "learning_rate": 0.0001,
+      "loss": 1.9794,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1197327852004111,
+      "grad_norm": 0.5710289478302002,
+      "learning_rate": 0.0001,
+      "loss": 1.2308,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1198184309695101,
+      "grad_norm": 0.5814175605773926,
+      "learning_rate": 0.0001,
+      "loss": 0.9535,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11990407673860912,
+      "grad_norm": 0.6103763580322266,
+      "learning_rate": 0.0001,
+      "loss": 0.61,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11998972250770812,
+      "grad_norm": 0.629924476146698,
+      "learning_rate": 0.0001,
+      "loss": 1.1306,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12007536827680712,
+      "grad_norm": 0.8482604622840881,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12016101404590614,
+      "grad_norm": 0.5386776328086853,
+      "learning_rate": 0.0001,
+      "loss": 0.9853,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12024665981500514,
+      "grad_norm": 0.6618002653121948,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12033230558410414,
+      "grad_norm": 0.6283786296844482,
+      "learning_rate": 0.0001,
+      "loss": 1.1123,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12041795135320316,
+      "grad_norm": 0.5722478628158569,
+      "learning_rate": 0.0001,
+      "loss": 1.0169,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12050359712230216,
+      "grad_norm": 0.7459848523139954,
+      "learning_rate": 0.0001,
+      "loss": 0.6127,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12058924289140116,
+      "grad_norm": 0.3811485469341278,
+      "learning_rate": 0.0001,
+      "loss": 0.7313,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12067488866050018,
+      "grad_norm": 0.5662948489189148,
+      "learning_rate": 0.0001,
+      "loss": 0.7813,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12076053442959918,
+      "grad_norm": 1.051496148109436,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12084618019869818,
+      "grad_norm": 0.7615790367126465,
+      "learning_rate": 0.0001,
+      "loss": 1.0646,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1209318259677972,
+      "grad_norm": 0.9175042510032654,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1210174717368962,
+      "grad_norm": 0.49467548727989197,
+      "learning_rate": 0.0001,
+      "loss": 0.5734,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1211031175059952,
+      "grad_norm": 0.8855686187744141,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12118876327509422,
+      "grad_norm": 0.6882718205451965,
+      "learning_rate": 0.0001,
+      "loss": 0.5995,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12127440904419322,
+      "grad_norm": 0.7760741710662842,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12136005481329222,
+      "grad_norm": 0.5504466891288757,
+      "learning_rate": 0.0001,
+      "loss": 0.9928,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12144570058239124,
+      "grad_norm": 0.40050122141838074,
+      "learning_rate": 0.0001,
+      "loss": 0.4174,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12153134635149024,
+      "grad_norm": 0.4629264175891876,
+      "learning_rate": 0.0001,
+      "loss": 0.8771,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12161699212058924,
+      "grad_norm": 0.7418511509895325,
+      "learning_rate": 0.0001,
+      "loss": 0.7623,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12170263788968826,
+      "grad_norm": 0.4909239709377289,
+      "learning_rate": 0.0001,
+      "loss": 0.772,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12178828365878726,
+      "grad_norm": 0.5730029344558716,
+      "learning_rate": 0.0001,
+      "loss": 1.1675,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12187392942788626,
+      "grad_norm": 0.8492511510848999,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12195957519698526,
+      "grad_norm": 0.5885708928108215,
+      "learning_rate": 0.0001,
+      "loss": 0.972,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12204522096608428,
+      "grad_norm": 0.6178143620491028,
+      "learning_rate": 0.0001,
+      "loss": 0.4793,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12213086673518328,
+      "grad_norm": 0.5950000286102295,
+      "learning_rate": 0.0001,
+      "loss": 0.6355,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12221651250428228,
+      "grad_norm": 0.5454826951026917,
+      "learning_rate": 0.0001,
+      "loss": 1.0737,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1223021582733813,
+      "grad_norm": 0.4723428785800934,
+      "learning_rate": 0.0001,
+      "loss": 0.7537,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1223878040424803,
+      "grad_norm": 0.6631142497062683,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1224734498115793,
+      "grad_norm": 0.6340517401695251,
+      "learning_rate": 0.0001,
+      "loss": 0.7021,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12255909558067832,
+      "grad_norm": 0.7209214568138123,
+      "learning_rate": 0.0001,
+      "loss": 1.1221,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12264474134977732,
+      "grad_norm": 0.43742433190345764,
+      "learning_rate": 0.0001,
+      "loss": 1.0621,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12273038711887632,
+      "grad_norm": 0.7425007224082947,
+      "learning_rate": 0.0001,
+      "loss": 0.7828,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12281603288797534,
+      "grad_norm": 0.5590226054191589,
+      "learning_rate": 0.0001,
+      "loss": 0.925,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12290167865707434,
+      "grad_norm": 0.8822212815284729,
+      "learning_rate": 0.0001,
+      "loss": 1.096,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12298732442617334,
+      "grad_norm": 0.30888551473617554,
+      "learning_rate": 0.0001,
+      "loss": 0.2014,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12307297019527236,
+      "grad_norm": 0.5047376155853271,
+      "learning_rate": 0.0001,
+      "loss": 1.1871,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12315861596437136,
+      "grad_norm": 0.7623229622840881,
+      "learning_rate": 0.0001,
+      "loss": 1.0859,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12324426173347036,
+      "grad_norm": 0.9144919514656067,
+      "learning_rate": 0.0001,
+      "loss": 0.4357,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12332990750256938,
+      "grad_norm": 0.6491893529891968,
+      "learning_rate": 0.0001,
+      "loss": 1.1522,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12341555327166838,
+      "grad_norm": 0.4043698012828827,
+      "learning_rate": 0.0001,
+      "loss": 1.2136,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12350119904076738,
+      "grad_norm": 0.45452919602394104,
+      "learning_rate": 0.0001,
+      "loss": 0.7996,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1235868448098664,
+      "grad_norm": 0.7932302951812744,
+      "learning_rate": 0.0001,
+      "loss": 1.3014,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1236724905789654,
+      "grad_norm": 0.7129805684089661,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1237581363480644,
+      "grad_norm": 0.48654213547706604,
+      "learning_rate": 0.0001,
+      "loss": 1.252,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12384378211716342,
+      "grad_norm": 0.6229697465896606,
+      "learning_rate": 0.0001,
+      "loss": 1.0221,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12392942788626242,
+      "grad_norm": 0.2353973537683487,
+      "learning_rate": 0.0001,
+      "loss": 0.7748,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12401507365536142,
+      "grad_norm": 0.7139957547187805,
+      "learning_rate": 0.0001,
+      "loss": 0.8115,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12410071942446044,
+      "grad_norm": 0.8170469999313354,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12418636519355944,
+      "grad_norm": 0.6148186922073364,
+      "learning_rate": 0.0001,
+      "loss": 1.2343,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12427201096265844,
+      "grad_norm": 0.45420700311660767,
+      "learning_rate": 0.0001,
+      "loss": 0.8978,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12435765673175746,
+      "grad_norm": 0.5886965990066528,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12444330250085646,
+      "grad_norm": 0.7324672341346741,
+      "learning_rate": 0.0001,
+      "loss": 0.8143,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12452894826995546,
+      "grad_norm": 0.623483419418335,
+      "learning_rate": 0.0001,
+      "loss": 1.305,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12461459403905448,
+      "grad_norm": 0.6213169693946838,
+      "learning_rate": 0.0001,
+      "loss": 0.5442,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12470023980815348,
+      "grad_norm": 0.6501922011375427,
+      "learning_rate": 0.0001,
+      "loss": 0.7983,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12478588557725248,
+      "grad_norm": 1.0657994747161865,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1248715313463515,
+      "grad_norm": 1.0657418966293335,
+      "learning_rate": 0.0001,
+      "loss": 1.0999,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1249571771154505,
+      "grad_norm": 0.734519362449646,
+      "learning_rate": 0.0001,
+      "loss": 0.8127,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12504282288454951,
+      "grad_norm": 0.4358311593532562,
+      "learning_rate": 0.0001,
+      "loss": 0.9534,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12512846865364852,
+      "grad_norm": 0.7647553086280823,
+      "learning_rate": 0.0001,
+      "loss": 0.704,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12521411442274752,
+      "grad_norm": 1.082554817199707,
+      "learning_rate": 0.0001,
+      "loss": 1.149,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12529976019184652,
+      "grad_norm": 0.4981421232223511,
+      "learning_rate": 0.0001,
+      "loss": 0.9487,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12538540596094552,
+      "grad_norm": 0.6032602787017822,
+      "learning_rate": 0.0001,
+      "loss": 0.6528,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12547105173004452,
+      "grad_norm": 0.5440607070922852,
+      "learning_rate": 0.0001,
+      "loss": 1.1337,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12555669749914355,
+      "grad_norm": 1.038310170173645,
+      "learning_rate": 0.0001,
+      "loss": 1.317,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12564234326824256,
+      "grad_norm": 0.5549770593643188,
+      "learning_rate": 0.0001,
+      "loss": 0.6211,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12572798903734156,
+      "grad_norm": 0.7957559823989868,
+      "learning_rate": 0.0001,
+      "loss": 0.653,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12581363480644056,
+      "grad_norm": 0.6332491636276245,
+      "learning_rate": 0.0001,
+      "loss": 0.5673,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12589928057553956,
+      "grad_norm": 0.48768311738967896,
+      "learning_rate": 0.0001,
+      "loss": 1.0909,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12598492634463856,
+      "grad_norm": 1.0346739292144775,
+      "learning_rate": 0.0001,
+      "loss": 0.8486,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1260705721137376,
+      "grad_norm": 0.6992138624191284,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1261562178828366,
+      "grad_norm": 0.3987424075603485,
+      "learning_rate": 0.0001,
+      "loss": 0.3529,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1262418636519356,
+      "grad_norm": 0.442565381526947,
+      "learning_rate": 0.0001,
+      "loss": 0.3231,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1263275094210346,
+      "grad_norm": 0.8141229152679443,
+      "learning_rate": 0.0001,
+      "loss": 1.0989,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1264131551901336,
+      "grad_norm": 0.5316410660743713,
+      "learning_rate": 0.0001,
+      "loss": 0.9937,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1264988009592326,
+      "grad_norm": 0.8684542179107666,
+      "learning_rate": 0.0001,
+      "loss": 1.1254,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1265844467283316,
+      "grad_norm": 0.6434370875358582,
+      "learning_rate": 0.0001,
+      "loss": 0.9632,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12667009249743064,
+      "grad_norm": 0.36928704380989075,
+      "learning_rate": 0.0001,
+      "loss": 0.8268,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12675573826652964,
+      "grad_norm": 0.4501432478427887,
+      "learning_rate": 0.0001,
+      "loss": 0.8041,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12684138403562864,
+      "grad_norm": 0.6698675155639648,
+      "learning_rate": 0.0001,
+      "loss": 0.8368,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12692702980472764,
+      "grad_norm": 0.7041205167770386,
+      "learning_rate": 0.0001,
+      "loss": 0.885,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12701267557382664,
+      "grad_norm": 0.5037266612052917,
+      "learning_rate": 0.0001,
+      "loss": 0.9891,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12709832134292565,
+      "grad_norm": 0.9233026504516602,
+      "learning_rate": 0.0001,
+      "loss": 0.9921,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12718396711202468,
+      "grad_norm": 0.3915286362171173,
+      "learning_rate": 0.0001,
+      "loss": 0.7655,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12726961288112368,
+      "grad_norm": 0.5465261340141296,
+      "learning_rate": 0.0001,
+      "loss": 1.0786,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12735525865022268,
+      "grad_norm": 0.39694520831108093,
+      "learning_rate": 0.0001,
+      "loss": 0.8547,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12744090441932168,
+      "grad_norm": 0.4870356321334839,
+      "learning_rate": 0.0001,
+      "loss": 0.8918,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12752655018842068,
+      "grad_norm": 0.8511807322502136,
+      "learning_rate": 0.0001,
+      "loss": 1.0646,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1276121959575197,
+      "grad_norm": 0.6451842784881592,
+      "learning_rate": 0.0001,
+      "loss": 0.834,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12769784172661872,
+      "grad_norm": 0.4858407974243164,
+      "learning_rate": 0.0001,
+      "loss": 0.3179,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12778348749571772,
+      "grad_norm": 0.5843254327774048,
+      "learning_rate": 0.0001,
+      "loss": 0.7679,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12786913326481672,
+      "grad_norm": 0.747596800327301,
+      "learning_rate": 0.0001,
+      "loss": 1.0576,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12795477903391572,
+      "grad_norm": 1.1192823648452759,
+      "learning_rate": 0.0001,
+      "loss": 1.2555,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12804042480301472,
+      "grad_norm": 0.7569806575775146,
+      "learning_rate": 0.0001,
+      "loss": 0.7831,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12812607057211373,
+      "grad_norm": 0.8191123008728027,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12821171634121276,
+      "grad_norm": 0.7036327719688416,
+      "learning_rate": 0.0001,
+      "loss": 0.5336,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12829736211031176,
+      "grad_norm": 0.7222660183906555,
+      "learning_rate": 0.0001,
+      "loss": 1.1497,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12838300787941076,
+      "grad_norm": 0.5286391377449036,
+      "learning_rate": 0.0001,
+      "loss": 1.0159,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12846865364850976,
+      "grad_norm": 0.6638986468315125,
+      "learning_rate": 0.0001,
+      "loss": 0.7988,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12855429941760876,
+      "grad_norm": 1.0295785665512085,
+      "learning_rate": 0.0001,
+      "loss": 0.8566,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12863994518670777,
+      "grad_norm": 0.5955945253372192,
+      "learning_rate": 0.0001,
+      "loss": 0.966,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1287255909558068,
+      "grad_norm": 0.4243564009666443,
+      "learning_rate": 0.0001,
+      "loss": 0.3282,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1288112367249058,
+      "grad_norm": 0.6961436867713928,
+      "learning_rate": 0.0001,
+      "loss": 1.3258,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1288968824940048,
+      "grad_norm": 0.8685358166694641,
+      "learning_rate": 0.0001,
+      "loss": 1.1629,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1289825282631038,
+      "grad_norm": 0.48196953535079956,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1290681740322028,
+      "grad_norm": 0.550537109375,
+      "learning_rate": 0.0001,
+      "loss": 0.3304,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1291538198013018,
+      "grad_norm": 0.8198084235191345,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12923946557040084,
+      "grad_norm": 0.6520404815673828,
+      "learning_rate": 0.0001,
+      "loss": 0.7094,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12932511133949984,
+      "grad_norm": 0.6228954195976257,
+      "learning_rate": 0.0001,
+      "loss": 1.4026,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12941075710859884,
+      "grad_norm": 0.5685157775878906,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12949640287769784,
+      "grad_norm": 0.6054257750511169,
+      "learning_rate": 0.0001,
+      "loss": 0.7425,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12958204864679684,
+      "grad_norm": 0.8037416338920593,
+      "learning_rate": 0.0001,
+      "loss": 0.6454,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12966769441589585,
+      "grad_norm": 0.6858931183815002,
+      "learning_rate": 0.0001,
+      "loss": 0.9679,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12975334018499485,
+      "grad_norm": 0.5170000195503235,
+      "learning_rate": 0.0001,
+      "loss": 0.8709,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12983898595409388,
+      "grad_norm": 0.8819013833999634,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12992463172319288,
+      "grad_norm": 0.6005105972290039,
+      "learning_rate": 0.0001,
+      "loss": 0.6405,
+      "step": 1517
+    },
+    {
+      "epoch": 0.13001027749229188,
+      "grad_norm": 0.24563227593898773,
+      "learning_rate": 0.0001,
+      "loss": 0.8841,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13009592326139088,
+      "grad_norm": 0.5996600985527039,
+      "learning_rate": 0.0001,
+      "loss": 1.0811,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13018156903048989,
+      "grad_norm": 0.6874776482582092,
+      "learning_rate": 0.0001,
+      "loss": 0.9933,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1302672147995889,
+      "grad_norm": 0.9664721488952637,
+      "learning_rate": 0.0001,
+      "loss": 0.992,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13035286056868792,
+      "grad_norm": 0.4617278277873993,
+      "learning_rate": 0.0001,
+      "loss": 1.0381,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13043850633778692,
+      "grad_norm": 0.47001418471336365,
+      "learning_rate": 0.0001,
+      "loss": 0.8833,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13052415210688592,
+      "grad_norm": 0.764961302280426,
+      "learning_rate": 0.0001,
+      "loss": 0.8803,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13060979787598492,
+      "grad_norm": 1.1379623413085938,
+      "learning_rate": 0.0001,
+      "loss": 1.2906,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13069544364508393,
+      "grad_norm": 0.9925264120101929,
+      "learning_rate": 0.0001,
+      "loss": 1.4527,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13078108941418293,
+      "grad_norm": 0.9171153903007507,
+      "learning_rate": 0.0001,
+      "loss": 1.1344,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13086673518328196,
+      "grad_norm": 0.6784950494766235,
+      "learning_rate": 0.0001,
+      "loss": 1.1877,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13095238095238096,
+      "grad_norm": 0.8708311915397644,
+      "learning_rate": 0.0001,
+      "loss": 0.8635,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13103802672147996,
+      "grad_norm": 0.6860393285751343,
+      "learning_rate": 0.0001,
+      "loss": 1.0953,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13112367249057896,
+      "grad_norm": 0.9775427579879761,
+      "learning_rate": 0.0001,
+      "loss": 0.9595,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13120931825967797,
+      "grad_norm": 0.7700759172439575,
+      "learning_rate": 0.0001,
+      "loss": 0.8027,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13129496402877697,
+      "grad_norm": 1.041894555091858,
+      "learning_rate": 0.0001,
+      "loss": 1.1111,
+      "step": 1533
+    },
+    {
+      "epoch": 0.131380609797876,
+      "grad_norm": 0.44255536794662476,
+      "learning_rate": 0.0001,
+      "loss": 0.3507,
+      "step": 1534
+    },
+    {
+      "epoch": 0.131466255566975,
+      "grad_norm": 0.4977659285068512,
+      "learning_rate": 0.0001,
+      "loss": 1.1347,
+      "step": 1535
+    },
+    {
+      "epoch": 0.131551901336074,
+      "grad_norm": 0.8843346834182739,
+      "learning_rate": 0.0001,
+      "loss": 0.9745,
+      "step": 1536
+    },
+    {
+      "epoch": 0.131637547105173,
+      "grad_norm": 0.11524787545204163,
+      "learning_rate": 0.0001,
+      "loss": 0.5856,
+      "step": 1537
+    },
+    {
+      "epoch": 0.131723192874272,
+      "grad_norm": 0.6384403109550476,
+      "learning_rate": 0.0001,
+      "loss": 0.5041,
+      "step": 1538
+    },
+    {
+      "epoch": 0.131808838643371,
+      "grad_norm": 0.7073954939842224,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13189448441247004,
+      "grad_norm": 0.4476461410522461,
+      "learning_rate": 0.0001,
+      "loss": 0.9454,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13198013018156904,
+      "grad_norm": 0.5538251996040344,
+      "learning_rate": 0.0001,
+      "loss": 1.2908,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13206577595066804,
+      "grad_norm": 0.7125557065010071,
+      "learning_rate": 0.0001,
+      "loss": 1.0291,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13215142171976704,
+      "grad_norm": 0.6489753127098083,
+      "learning_rate": 0.0001,
+      "loss": 0.7551,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13223706748886604,
+      "grad_norm": 0.5303845405578613,
+      "learning_rate": 0.0001,
+      "loss": 1.2291,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13232271325796505,
+      "grad_norm": 0.9153067469596863,
+      "learning_rate": 0.0001,
+      "loss": 1.171,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13240835902706408,
+      "grad_norm": 0.5424889326095581,
+      "learning_rate": 0.0001,
+      "loss": 0.9468,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13249400479616308,
+      "grad_norm": 0.9731685519218445,
+      "learning_rate": 0.0001,
+      "loss": 0.6569,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13257965056526208,
+      "grad_norm": 0.6738279461860657,
+      "learning_rate": 0.0001,
+      "loss": 1.0889,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13266529633436108,
+      "grad_norm": 0.8318383693695068,
+      "learning_rate": 0.0001,
+      "loss": 1.2673,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13275094210346008,
+      "grad_norm": 0.9484409689903259,
+      "learning_rate": 0.0001,
+      "loss": 1.0869,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1328365878725591,
+      "grad_norm": 0.5849506855010986,
+      "learning_rate": 0.0001,
+      "loss": 0.492,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1329222336416581,
+      "grad_norm": 0.5120412111282349,
+      "learning_rate": 0.0001,
+      "loss": 1.16,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13300787941075712,
+      "grad_norm": 1.1595613956451416,
+      "learning_rate": 0.0001,
+      "loss": 0.9965,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13309352517985612,
+      "grad_norm": 0.12458866834640503,
+      "learning_rate": 0.0001,
+      "loss": 0.4674,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13317917094895512,
+      "grad_norm": 0.6736694574356079,
+      "learning_rate": 0.0001,
+      "loss": 1.2198,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13326481671805412,
+      "grad_norm": 1.0090261697769165,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13335046248715313,
+      "grad_norm": 0.9140573143959045,
+      "learning_rate": 0.0001,
+      "loss": 1.2035,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13343610825625213,
+      "grad_norm": 0.6811180114746094,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13352175402535116,
+      "grad_norm": 0.7343337535858154,
+      "learning_rate": 0.0001,
+      "loss": 0.7712,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13360739979445016,
+      "grad_norm": 0.9875969886779785,
+      "learning_rate": 0.0001,
+      "loss": 0.9834,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13369304556354916,
+      "grad_norm": 0.39642030000686646,
+      "learning_rate": 0.0001,
+      "loss": 0.7531,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13377869133264816,
+      "grad_norm": 0.5674512982368469,
+      "learning_rate": 0.0001,
+      "loss": 1.3236,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13386433710174717,
+      "grad_norm": 0.7752519249916077,
+      "learning_rate": 0.0001,
+      "loss": 1.2067,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13394998287084617,
+      "grad_norm": 0.8780034184455872,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1340356286399452,
+      "grad_norm": 1.0622485876083374,
+      "learning_rate": 0.0001,
+      "loss": 1.1072,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1341212744090442,
+      "grad_norm": 0.8268039226531982,
+      "learning_rate": 0.0001,
+      "loss": 0.9517,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1342069201781432,
+      "grad_norm": 0.4274117648601532,
+      "learning_rate": 0.0001,
+      "loss": 1.2068,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1342925659472422,
+      "grad_norm": 0.9136738777160645,
+      "learning_rate": 0.0001,
+      "loss": 0.9221,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1343782117163412,
+      "grad_norm": 1.0467026233673096,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1344638574854402,
+      "grad_norm": 0.5356448292732239,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13454950325453924,
+      "grad_norm": 0.823401153087616,
+      "learning_rate": 0.0001,
+      "loss": 1.1666,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13463514902363824,
+      "grad_norm": 0.5706819891929626,
+      "learning_rate": 0.0001,
+      "loss": 0.9733,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13472079479273724,
+      "grad_norm": 0.8271875381469727,
+      "learning_rate": 0.0001,
+      "loss": 0.938,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13480644056183624,
+      "grad_norm": 0.7359247803688049,
+      "learning_rate": 0.0001,
+      "loss": 0.8443,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13489208633093525,
+      "grad_norm": 0.7144054174423218,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13497773210003425,
+      "grad_norm": 0.6399334669113159,
+      "learning_rate": 0.0001,
+      "loss": 1.1311,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13506337786913328,
+      "grad_norm": 0.563167929649353,
+      "learning_rate": 0.0001,
+      "loss": 0.6294,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13514902363823228,
+      "grad_norm": 0.5729923248291016,
+      "learning_rate": 0.0001,
+      "loss": 0.7269,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13523466940733128,
+      "grad_norm": 0.8072612881660461,
+      "learning_rate": 0.0001,
+      "loss": 1.074,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13532031517643028,
+      "grad_norm": 1.2577322721481323,
+      "learning_rate": 0.0001,
+      "loss": 0.8575,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13540596094552929,
+      "grad_norm": 0.5634222030639648,
+      "learning_rate": 0.0001,
+      "loss": 0.5524,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1354916067146283,
+      "grad_norm": 0.7289922833442688,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13557725248372732,
+      "grad_norm": 0.283830851316452,
+      "learning_rate": 0.0001,
+      "loss": 0.5945,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13566289825282632,
+      "grad_norm": 0.570106029510498,
+      "learning_rate": 0.0001,
+      "loss": 0.5162,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13574854402192532,
+      "grad_norm": 0.9642724394798279,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13583418979102432,
+      "grad_norm": 0.4728585481643677,
+      "learning_rate": 0.0001,
+      "loss": 1.0077,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13591983556012333,
+      "grad_norm": 0.5478368401527405,
+      "learning_rate": 0.0001,
+      "loss": 0.4826,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13600548132922233,
+      "grad_norm": 0.44264572858810425,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13609112709832133,
+      "grad_norm": 0.37857523560523987,
+      "learning_rate": 0.0001,
+      "loss": 0.7212,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13617677286742036,
+      "grad_norm": 1.1303529739379883,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13626241863651936,
+      "grad_norm": 0.6972150206565857,
+      "learning_rate": 0.0001,
+      "loss": 1.1213,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13634806440561836,
+      "grad_norm": 0.12623240053653717,
+      "learning_rate": 0.0001,
+      "loss": 0.5583,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13643371017471737,
+      "grad_norm": 0.7325114011764526,
+      "learning_rate": 0.0001,
+      "loss": 1.1203,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13651935594381637,
+      "grad_norm": 0.6996632814407349,
+      "learning_rate": 0.0001,
+      "loss": 0.7275,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13660500171291537,
+      "grad_norm": 0.7038044929504395,
+      "learning_rate": 0.0001,
+      "loss": 0.7356,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1366906474820144,
+      "grad_norm": 0.9104805588722229,
+      "learning_rate": 0.0001,
+      "loss": 1.1029,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1367762932511134,
+      "grad_norm": 0.5023924112319946,
+      "learning_rate": 0.0001,
+      "loss": 1.0914,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1368619390202124,
+      "grad_norm": 1.0536991357803345,
+      "learning_rate": 0.0001,
+      "loss": 0.5673,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1369475847893114,
+      "grad_norm": 0.5190120339393616,
+      "learning_rate": 0.0001,
+      "loss": 1.0524,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1370332305584104,
+      "grad_norm": 0.666867196559906,
+      "learning_rate": 0.0001,
+      "loss": 0.9209,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1371188763275094,
+      "grad_norm": 0.6555282473564148,
+      "learning_rate": 0.0001,
+      "loss": 1.3402,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13720452209660844,
+      "grad_norm": 0.6935417056083679,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13729016786570744,
+      "grad_norm": 0.3158670961856842,
+      "learning_rate": 0.0001,
+      "loss": 0.7513,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13737581363480644,
+      "grad_norm": 0.4669382572174072,
+      "learning_rate": 0.0001,
+      "loss": 0.4101,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13746145940390544,
+      "grad_norm": 0.8058091402053833,
+      "learning_rate": 0.0001,
+      "loss": 0.9174,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13754710517300445,
+      "grad_norm": 0.6820075511932373,
+      "learning_rate": 0.0001,
+      "loss": 0.7432,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13763275094210345,
+      "grad_norm": 0.7961389422416687,
+      "learning_rate": 0.0001,
+      "loss": 0.6409,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13771839671120248,
+      "grad_norm": 0.6787572503089905,
+      "learning_rate": 0.0001,
+      "loss": 0.9079,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13780404248030148,
+      "grad_norm": 0.7684647440910339,
+      "learning_rate": 0.0001,
+      "loss": 0.9466,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13788968824940048,
+      "grad_norm": 0.7331491112709045,
+      "learning_rate": 0.0001,
+      "loss": 0.8246,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13797533401849948,
+      "grad_norm": 0.6756778359413147,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1380609797875985,
+      "grad_norm": 0.6843546032905579,
+      "learning_rate": 0.0001,
+      "loss": 0.7101,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1381466255566975,
+      "grad_norm": 0.4348793923854828,
+      "learning_rate": 0.0001,
+      "loss": 0.786,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13823227132579652,
+      "grad_norm": 0.8262733817100525,
+      "learning_rate": 0.0001,
+      "loss": 2.0139,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13831791709489552,
+      "grad_norm": 0.7362405061721802,
+      "learning_rate": 0.0001,
+      "loss": 0.9781,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13840356286399452,
+      "grad_norm": 0.6190007328987122,
+      "learning_rate": 0.0001,
+      "loss": 0.9023,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13848920863309352,
+      "grad_norm": 0.611380398273468,
+      "learning_rate": 0.0001,
+      "loss": 1.15,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13857485440219253,
+      "grad_norm": 0.40790507197380066,
+      "learning_rate": 0.0001,
+      "loss": 0.2834,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13866050017129153,
+      "grad_norm": 0.3164280354976654,
+      "learning_rate": 0.0001,
+      "loss": 0.3172,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13874614594039056,
+      "grad_norm": 0.4469259977340698,
+      "learning_rate": 0.0001,
+      "loss": 0.5919,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13883179170948956,
+      "grad_norm": 0.6457232236862183,
+      "learning_rate": 0.0001,
+      "loss": 0.8708,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13891743747858856,
+      "grad_norm": 0.7680903673171997,
+      "learning_rate": 0.0001,
+      "loss": 0.9969,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13900308324768756,
+      "grad_norm": 0.9684513807296753,
+      "learning_rate": 0.0001,
+      "loss": 1.1957,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13908872901678657,
+      "grad_norm": 0.5689455270767212,
+      "learning_rate": 0.0001,
+      "loss": 0.4465,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13917437478588557,
+      "grad_norm": 0.48528775572776794,
+      "learning_rate": 0.0001,
+      "loss": 0.9284,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13926002055498457,
+      "grad_norm": 0.7671158909797668,
+      "learning_rate": 0.0001,
+      "loss": 1.0669,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1393456663240836,
+      "grad_norm": 0.37550660967826843,
+      "learning_rate": 0.0001,
+      "loss": 0.5139,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1394313120931826,
+      "grad_norm": 0.5239292979240417,
+      "learning_rate": 0.0001,
+      "loss": 1.1451,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1395169578622816,
+      "grad_norm": 0.717453122138977,
+      "learning_rate": 0.0001,
+      "loss": 0.7294,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1396026036313806,
+      "grad_norm": 0.9300013780593872,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1396882494004796,
+      "grad_norm": 1.0198454856872559,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1397738951695786,
+      "grad_norm": 0.7445101737976074,
+      "learning_rate": 0.0001,
+      "loss": 1.0976,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13985954093867764,
+      "grad_norm": 0.40338781476020813,
+      "learning_rate": 0.0001,
+      "loss": 0.4474,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13994518670777664,
+      "grad_norm": 0.6558519005775452,
+      "learning_rate": 0.0001,
+      "loss": 0.5628,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14003083247687564,
+      "grad_norm": 0.5181640386581421,
+      "learning_rate": 0.0001,
+      "loss": 0.4338,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14011647824597465,
+      "grad_norm": 0.7103620171546936,
+      "learning_rate": 0.0001,
+      "loss": 1.074,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14020212401507365,
+      "grad_norm": 0.4184480309486389,
+      "learning_rate": 0.0001,
+      "loss": 0.9089,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14028776978417265,
+      "grad_norm": 0.5259749889373779,
+      "learning_rate": 0.0001,
+      "loss": 0.7903,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14037341555327168,
+      "grad_norm": 0.5141268968582153,
+      "learning_rate": 0.0001,
+      "loss": 0.6419,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14045906132237068,
+      "grad_norm": 0.5543323755264282,
+      "learning_rate": 0.0001,
+      "loss": 0.5322,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14054470709146968,
+      "grad_norm": 0.8291405439376831,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14063035286056869,
+      "grad_norm": 0.9100068211555481,
+      "learning_rate": 0.0001,
+      "loss": 1.8179,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1407159986296677,
+      "grad_norm": 0.7242834568023682,
+      "learning_rate": 0.0001,
+      "loss": 1.1317,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1408016443987667,
+      "grad_norm": 0.7211915850639343,
+      "learning_rate": 0.0001,
+      "loss": 1.1008,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14088729016786572,
+      "grad_norm": 0.7096391320228577,
+      "learning_rate": 0.0001,
+      "loss": 0.9113,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14097293593696472,
+      "grad_norm": 0.6131205558776855,
+      "learning_rate": 0.0001,
+      "loss": 0.7602,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14105858170606372,
+      "grad_norm": 0.6371170282363892,
+      "learning_rate": 0.0001,
+      "loss": 1.1584,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14114422747516273,
+      "grad_norm": 0.6788474917411804,
+      "learning_rate": 0.0001,
+      "loss": 0.8464,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14122987324426173,
+      "grad_norm": 0.661794900894165,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14131551901336073,
+      "grad_norm": 0.29670724272727966,
+      "learning_rate": 0.0001,
+      "loss": 0.7341,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14140116478245976,
+      "grad_norm": 0.8316453099250793,
+      "learning_rate": 0.0001,
+      "loss": 1.0463,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14148681055155876,
+      "grad_norm": 0.6612321138381958,
+      "learning_rate": 0.0001,
+      "loss": 1.3653,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14157245632065776,
+      "grad_norm": 0.7581387758255005,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14165810208975677,
+      "grad_norm": 0.7885164618492126,
+      "learning_rate": 0.0001,
+      "loss": 0.8213,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14174374785885577,
+      "grad_norm": 0.4867507219314575,
+      "learning_rate": 0.0001,
+      "loss": 0.3188,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14182939362795477,
+      "grad_norm": 0.7161390781402588,
+      "learning_rate": 0.0001,
+      "loss": 1.0236,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1419150393970538,
+      "grad_norm": 0.6802307963371277,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1420006851661528,
+      "grad_norm": 1.0278455018997192,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1420863309352518,
+      "grad_norm": 1.1100010871887207,
+      "learning_rate": 0.0001,
+      "loss": 1.07,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1421719767043508,
+      "grad_norm": 0.5367780327796936,
+      "learning_rate": 0.0001,
+      "loss": 1.0796,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1422576224734498,
+      "grad_norm": 0.5661779046058655,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1423432682425488,
+      "grad_norm": 0.7001707553863525,
+      "learning_rate": 0.0001,
+      "loss": 1.1012,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1424289140116478,
+      "grad_norm": 0.5917421579360962,
+      "learning_rate": 0.0001,
+      "loss": 0.7157,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14251455978074684,
+      "grad_norm": 0.8316869735717773,
+      "learning_rate": 0.0001,
+      "loss": 1.1503,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14260020554984584,
+      "grad_norm": 0.7738096714019775,
+      "learning_rate": 0.0001,
+      "loss": 1.0188,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14268585131894485,
+      "grad_norm": 0.5467625856399536,
+      "learning_rate": 0.0001,
+      "loss": 1.0423,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14277149708804385,
+      "grad_norm": 0.9325948357582092,
+      "learning_rate": 0.0001,
+      "loss": 1.2642,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.9439052939414978,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14294278862624185,
+      "grad_norm": 0.642646849155426,
+      "learning_rate": 0.0001,
+      "loss": 0.7541,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14302843439534088,
+      "grad_norm": 0.6962234377861023,
+      "learning_rate": 0.0001,
+      "loss": 0.6798,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14311408016443988,
+      "grad_norm": 0.46604764461517334,
+      "learning_rate": 0.0001,
+      "loss": 0.6988,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14319972593353889,
+      "grad_norm": 0.7284623980522156,
+      "learning_rate": 0.0001,
+      "loss": 1.1442,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1432853717026379,
+      "grad_norm": 0.5294442772865295,
+      "learning_rate": 0.0001,
+      "loss": 0.5895,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1433710174717369,
+      "grad_norm": 0.4569492042064667,
+      "learning_rate": 0.0001,
+      "loss": 1.2732,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1434566632408359,
+      "grad_norm": 0.8076384663581848,
+      "learning_rate": 0.0001,
+      "loss": 1.1591,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14354230900993492,
+      "grad_norm": 0.9937357306480408,
+      "learning_rate": 0.0001,
+      "loss": 1.011,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14362795477903392,
+      "grad_norm": 0.5679730772972107,
+      "learning_rate": 0.0001,
+      "loss": 1.1464,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14371360054813292,
+      "grad_norm": 0.5820409059524536,
+      "learning_rate": 0.0001,
+      "loss": 0.6125,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14379924631723193,
+      "grad_norm": 0.5143232941627502,
+      "learning_rate": 0.0001,
+      "loss": 1.043,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 0.7303682565689087,
+      "learning_rate": 0.0001,
+      "loss": 0.9721,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14397053785542993,
+      "grad_norm": 0.837049126625061,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14405618362452896,
+      "grad_norm": 0.6082029342651367,
+      "learning_rate": 0.0001,
+      "loss": 1.2171,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14414182939362796,
+      "grad_norm": 0.4037919044494629,
+      "learning_rate": 0.0001,
+      "loss": 1.087,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14422747516272696,
+      "grad_norm": 0.49174588918685913,
+      "learning_rate": 0.0001,
+      "loss": 0.85,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14431312093182597,
+      "grad_norm": 0.832642674446106,
+      "learning_rate": 0.0001,
+      "loss": 0.8514,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14439876670092497,
+      "grad_norm": 0.6568020582199097,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14448441247002397,
+      "grad_norm": 0.5219976305961609,
+      "learning_rate": 0.0001,
+      "loss": 0.7538,
+      "step": 1687
+    },
+    {
+      "epoch": 0.144570058239123,
+      "grad_norm": 0.5740683078765869,
+      "learning_rate": 0.0001,
+      "loss": 1.0096,
+      "step": 1688
+    },
+    {
+      "epoch": 0.144655704008222,
+      "grad_norm": 0.5280925035476685,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 1689
+    },
+    {
+      "epoch": 0.144741349777321,
+      "grad_norm": 0.5498401522636414,
+      "learning_rate": 0.0001,
+      "loss": 0.9252,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14482699554642,
+      "grad_norm": 1.2621352672576904,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 1691
+    },
+    {
+      "epoch": 0.144912641315519,
+      "grad_norm": 1.1449522972106934,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 1692
+    },
+    {
+      "epoch": 0.144998287084618,
+      "grad_norm": 0.7366014122962952,
+      "learning_rate": 0.0001,
+      "loss": 1.0877,
+      "step": 1693
+    },
+    {
+      "epoch": 0.145083932853717,
+      "grad_norm": 0.5974490642547607,
+      "learning_rate": 0.0001,
+      "loss": 0.9866,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14516957862281604,
+      "grad_norm": 0.5991716980934143,
+      "learning_rate": 0.0001,
+      "loss": 1.0647,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14525522439191504,
+      "grad_norm": 1.055510401725769,
+      "learning_rate": 0.0001,
+      "loss": 1.2073,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14534087016101405,
+      "grad_norm": 0.39201903343200684,
+      "learning_rate": 0.0001,
+      "loss": 0.7602,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14542651593011305,
+      "grad_norm": 0.85997474193573,
+      "learning_rate": 0.0001,
+      "loss": 0.8793,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14551216169921205,
+      "grad_norm": 0.995727002620697,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14559780746831105,
+      "grad_norm": 0.5165157914161682,
+      "learning_rate": 0.0001,
+      "loss": 1.1525,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14568345323741008,
+      "grad_norm": 0.7696508169174194,
+      "learning_rate": 0.0001,
+      "loss": 0.562,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14576909900650908,
+      "grad_norm": 0.6804817914962769,
+      "learning_rate": 0.0001,
+      "loss": 1.0215,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1458547447756081,
+      "grad_norm": 0.5847865343093872,
+      "learning_rate": 0.0001,
+      "loss": 0.3646,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1459403905447071,
+      "grad_norm": 0.5005224943161011,
+      "learning_rate": 0.0001,
+      "loss": 1.0506,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1460260363138061,
+      "grad_norm": 0.5572129487991333,
+      "learning_rate": 0.0001,
+      "loss": 0.539,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1461116820829051,
+      "grad_norm": 1.2933365106582642,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14619732785200412,
+      "grad_norm": 0.7788717150688171,
+      "learning_rate": 0.0001,
+      "loss": 0.699,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14628297362110312,
+      "grad_norm": 0.6620541214942932,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14636861939020213,
+      "grad_norm": 0.5780394077301025,
+      "learning_rate": 0.0001,
+      "loss": 0.5608,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14645426515930113,
+      "grad_norm": 1.066033959388733,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14653991092840013,
+      "grad_norm": 0.45781615376472473,
+      "learning_rate": 0.0001,
+      "loss": 0.7503,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14662555669749913,
+      "grad_norm": 0.47087690234184265,
+      "learning_rate": 0.0001,
+      "loss": 0.4539,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14671120246659816,
+      "grad_norm": 1.0352656841278076,
+      "learning_rate": 0.0001,
+      "loss": 2.0793,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14679684823569716,
+      "grad_norm": 0.9318323135375977,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14688249400479617,
+      "grad_norm": 0.5470418930053711,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14696813977389517,
+      "grad_norm": 0.5146464109420776,
+      "learning_rate": 0.0001,
+      "loss": 0.9921,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14705378554299417,
+      "grad_norm": 0.40464362502098083,
+      "learning_rate": 0.0001,
+      "loss": 1.1424,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14713943131209317,
+      "grad_norm": 0.5526432394981384,
+      "learning_rate": 0.0001,
+      "loss": 0.8157,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1472250770811922,
+      "grad_norm": 0.7004134058952332,
+      "learning_rate": 0.0001,
+      "loss": 0.8387,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1473107228502912,
+      "grad_norm": 0.6484535932540894,
+      "learning_rate": 0.0001,
+      "loss": 0.7247,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1473963686193902,
+      "grad_norm": 0.884462833404541,
+      "learning_rate": 0.0001,
+      "loss": 1.3024,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1474820143884892,
+      "grad_norm": 0.6597541570663452,
+      "learning_rate": 0.0001,
+      "loss": 0.9798,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1475676601575882,
+      "grad_norm": 0.4758937358856201,
+      "learning_rate": 0.0001,
+      "loss": 0.9339,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1476533059266872,
+      "grad_norm": 0.5114342570304871,
+      "learning_rate": 0.0001,
+      "loss": 0.3907,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14773895169578624,
+      "grad_norm": 1.3790701627731323,
+      "learning_rate": 0.0001,
+      "loss": 0.8696,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14782459746488524,
+      "grad_norm": 0.7155749201774597,
+      "learning_rate": 0.0001,
+      "loss": 0.6594,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14791024323398425,
+      "grad_norm": 0.6895710229873657,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14799588900308325,
+      "grad_norm": 0.7365022301673889,
+      "learning_rate": 0.0001,
+      "loss": 1.1045,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14808153477218225,
+      "grad_norm": 0.5674819350242615,
+      "learning_rate": 0.0001,
+      "loss": 0.5776,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14816718054128125,
+      "grad_norm": 0.5641562938690186,
+      "learning_rate": 0.0001,
+      "loss": 0.9379,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14825282631038025,
+      "grad_norm": 0.4714890718460083,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14833847207947928,
+      "grad_norm": 0.9575049877166748,
+      "learning_rate": 0.0001,
+      "loss": 1.0679,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14842411784857829,
+      "grad_norm": 0.5785413980484009,
+      "learning_rate": 0.0001,
+      "loss": 1.1589,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1485097636176773,
+      "grad_norm": 0.5648821592330933,
+      "learning_rate": 0.0001,
+      "loss": 0.9848,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1485954093867763,
+      "grad_norm": 0.758966326713562,
+      "learning_rate": 0.0001,
+      "loss": 1.1872,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1486810551558753,
+      "grad_norm": 0.7043062448501587,
+      "learning_rate": 0.0001,
+      "loss": 0.9223,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1487667009249743,
+      "grad_norm": 1.0661917924880981,
+      "learning_rate": 0.0001,
+      "loss": 0.838,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14885234669407332,
+      "grad_norm": 0.7718272805213928,
+      "learning_rate": 0.0001,
+      "loss": 0.7906,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14893799246317233,
+      "grad_norm": 0.5734906792640686,
+      "learning_rate": 0.0001,
+      "loss": 0.5057,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14902363823227133,
+      "grad_norm": 0.5202139616012573,
+      "learning_rate": 0.0001,
+      "loss": 0.8331,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14910928400137033,
+      "grad_norm": 0.7494276165962219,
+      "learning_rate": 0.0001,
+      "loss": 0.9403,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14919492977046933,
+      "grad_norm": 0.6445489525794983,
+      "learning_rate": 0.0001,
+      "loss": 0.6536,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14928057553956833,
+      "grad_norm": 0.6550923585891724,
+      "learning_rate": 0.0001,
+      "loss": 0.7118,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14936622130866736,
+      "grad_norm": 0.6805226802825928,
+      "learning_rate": 0.0001,
+      "loss": 0.6878,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14945186707776636,
+      "grad_norm": 0.6547775864601135,
+      "learning_rate": 0.0001,
+      "loss": 0.627,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14953751284686537,
+      "grad_norm": 1.4202346801757812,
+      "learning_rate": 0.0001,
+      "loss": 1.1676,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14962315861596437,
+      "grad_norm": 0.8187187910079956,
+      "learning_rate": 0.0001,
+      "loss": 0.8331,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14970880438506337,
+      "grad_norm": 0.4761867821216583,
+      "learning_rate": 0.0001,
+      "loss": 0.5841,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14979445015416237,
+      "grad_norm": 1.9144070148468018,
+      "learning_rate": 0.0001,
+      "loss": 1.2229,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1498800959232614,
+      "grad_norm": 1.0356754064559937,
+      "learning_rate": 0.0001,
+      "loss": 0.9494,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1499657416923604,
+      "grad_norm": 0.5979669690132141,
+      "learning_rate": 0.0001,
+      "loss": 1.0827,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1500513874614594,
+      "grad_norm": 0.7048048377037048,
+      "learning_rate": 0.0001,
+      "loss": 0.6304,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1501370332305584,
+      "grad_norm": 0.4667990207672119,
+      "learning_rate": 0.0001,
+      "loss": 0.9479,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1502226789996574,
+      "grad_norm": 0.41659852862358093,
+      "learning_rate": 0.0001,
+      "loss": 0.6789,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1503083247687564,
+      "grad_norm": 0.8033080697059631,
+      "learning_rate": 0.0001,
+      "loss": 1.163,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15039397053785544,
+      "grad_norm": 0.9822734594345093,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15047961630695444,
+      "grad_norm": 0.8684450387954712,
+      "learning_rate": 0.0001,
+      "loss": 0.9903,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15056526207605345,
+      "grad_norm": 0.9772304892539978,
+      "learning_rate": 0.0001,
+      "loss": 1.0217,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15065090784515245,
+      "grad_norm": 0.8162287473678589,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15073655361425145,
+      "grad_norm": 0.792847752571106,
+      "learning_rate": 0.0001,
+      "loss": 1.296,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15082219938335045,
+      "grad_norm": 0.6087763905525208,
+      "learning_rate": 0.0001,
+      "loss": 1.0386,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15090784515244948,
+      "grad_norm": 1.025903344154358,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15099349092154848,
+      "grad_norm": 0.5262158513069153,
+      "learning_rate": 0.0001,
+      "loss": 0.8374,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1510791366906475,
+      "grad_norm": 0.7190976738929749,
+      "learning_rate": 0.0001,
+      "loss": 0.8181,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1511647824597465,
+      "grad_norm": 0.6235056519508362,
+      "learning_rate": 0.0001,
+      "loss": 0.9058,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1512504282288455,
+      "grad_norm": 1.1998549699783325,
+      "learning_rate": 0.0001,
+      "loss": 1.2244,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1513360739979445,
+      "grad_norm": 0.7461796402931213,
+      "learning_rate": 0.0001,
+      "loss": 0.8348,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1514217197670435,
+      "grad_norm": 0.8417841196060181,
+      "learning_rate": 0.0001,
+      "loss": 0.9418,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15150736553614252,
+      "grad_norm": 0.783588707447052,
+      "learning_rate": 0.0001,
+      "loss": 1.1211,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15159301130524153,
+      "grad_norm": 0.43426796793937683,
+      "learning_rate": 0.0001,
+      "loss": 0.6171,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15167865707434053,
+      "grad_norm": 0.9733153581619263,
+      "learning_rate": 0.0001,
+      "loss": 1.1578,
+      "step": 1771
+    },
+    {
+      "epoch": 0.15176430284343953,
+      "grad_norm": 0.5566218495368958,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15184994861253853,
+      "grad_norm": 0.45109638571739197,
+      "learning_rate": 0.0001,
+      "loss": 0.7449,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15193559438163753,
+      "grad_norm": 0.9458006024360657,
+      "learning_rate": 0.0001,
+      "loss": 0.9967,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15202124015073656,
+      "grad_norm": 0.8400782346725464,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 1775
+    },
+    {
+      "epoch": 0.15210688591983557,
+      "grad_norm": 0.544459342956543,
+      "learning_rate": 0.0001,
+      "loss": 1.0994,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15219253168893457,
+      "grad_norm": 0.6392741203308105,
+      "learning_rate": 0.0001,
+      "loss": 0.9471,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15227817745803357,
+      "grad_norm": 0.3089018762111664,
+      "learning_rate": 0.0001,
+      "loss": 0.3915,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15236382322713257,
+      "grad_norm": 1.0368905067443848,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15244946899623157,
+      "grad_norm": 0.6897367238998413,
+      "learning_rate": 0.0001,
+      "loss": 1.0957,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1525351147653306,
+      "grad_norm": 1.2239227294921875,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1526207605344296,
+      "grad_norm": 0.845046877861023,
+      "learning_rate": 0.0001,
+      "loss": 1.0057,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1527064063035286,
+      "grad_norm": 0.967596709728241,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1527920520726276,
+      "grad_norm": 0.41560694575309753,
+      "learning_rate": 0.0001,
+      "loss": 0.7226,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1528776978417266,
+      "grad_norm": 0.6150450110435486,
+      "learning_rate": 0.0001,
+      "loss": 0.6623,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15296334361082561,
+      "grad_norm": 0.6020082235336304,
+      "learning_rate": 0.0001,
+      "loss": 0.6918,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15304898937992464,
+      "grad_norm": 0.657332718372345,
+      "learning_rate": 0.0001,
+      "loss": 0.6454,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15313463514902365,
+      "grad_norm": 0.6391571164131165,
+      "learning_rate": 0.0001,
+      "loss": 0.4078,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15322028091812265,
+      "grad_norm": 0.7777669429779053,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15330592668722165,
+      "grad_norm": 0.5976850986480713,
+      "learning_rate": 0.0001,
+      "loss": 1.2884,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15339157245632065,
+      "grad_norm": 1.017464518547058,
+      "learning_rate": 0.0001,
+      "loss": 1.0394,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15347721822541965,
+      "grad_norm": 0.5791511535644531,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15356286399451868,
+      "grad_norm": 1.0686423778533936,
+      "learning_rate": 0.0001,
+      "loss": 1.2785,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15364850976361769,
+      "grad_norm": 0.7500899434089661,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1537341555327167,
+      "grad_norm": 0.7110128402709961,
+      "learning_rate": 0.0001,
+      "loss": 1.408,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1538198013018157,
+      "grad_norm": 0.5524047613143921,
+      "learning_rate": 0.0001,
+      "loss": 1.0738,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1539054470709147,
+      "grad_norm": 0.5694042444229126,
+      "learning_rate": 0.0001,
+      "loss": 0.664,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1539910928400137,
+      "grad_norm": 1.4163657426834106,
+      "learning_rate": 0.0001,
+      "loss": 1.0097,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15407673860911272,
+      "grad_norm": 1.0164870023727417,
+      "learning_rate": 0.0001,
+      "loss": 1.9596,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15416238437821173,
+      "grad_norm": 0.547529399394989,
+      "learning_rate": 0.0001,
+      "loss": 1.0282,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15424803014731073,
+      "grad_norm": 0.45616787672042847,
+      "learning_rate": 0.0001,
+      "loss": 1.0656,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15433367591640973,
+      "grad_norm": 0.6092583537101746,
+      "learning_rate": 0.0001,
+      "loss": 1.9343,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15441932168550873,
+      "grad_norm": 0.6355782747268677,
+      "learning_rate": 0.0001,
+      "loss": 1.126,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15450496745460773,
+      "grad_norm": 0.8996865153312683,
+      "learning_rate": 0.0001,
+      "loss": 1.2369,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15459061322370674,
+      "grad_norm": 0.6168944835662842,
+      "learning_rate": 0.0001,
+      "loss": 1.011,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15467625899280577,
+      "grad_norm": 0.6526392102241516,
+      "learning_rate": 0.0001,
+      "loss": 0.6368,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15476190476190477,
+      "grad_norm": 0.6572227478027344,
+      "learning_rate": 0.0001,
+      "loss": 0.9182,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15484755053100377,
+      "grad_norm": 0.4038061499595642,
+      "learning_rate": 0.0001,
+      "loss": 0.7063,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15493319630010277,
+      "grad_norm": 0.3309141993522644,
+      "learning_rate": 0.0001,
+      "loss": 0.5857,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15501884206920177,
+      "grad_norm": 0.4429363012313843,
+      "learning_rate": 0.0001,
+      "loss": 0.7512,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15510448783830078,
+      "grad_norm": 0.8737627267837524,
+      "learning_rate": 0.0001,
+      "loss": 0.7242,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1551901336073998,
+      "grad_norm": 1.0066404342651367,
+      "learning_rate": 0.0001,
+      "loss": 1.1713,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1552757793764988,
+      "grad_norm": 0.5962488055229187,
+      "learning_rate": 0.0001,
+      "loss": 1.036,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1553614251455978,
+      "grad_norm": 0.40972989797592163,
+      "learning_rate": 0.0001,
+      "loss": 0.846,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1554470709146968,
+      "grad_norm": 0.5725266337394714,
+      "learning_rate": 0.0001,
+      "loss": 1.1491,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1555327166837958,
+      "grad_norm": 0.5088080763816833,
+      "learning_rate": 0.0001,
+      "loss": 0.5198,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15561836245289482,
+      "grad_norm": 0.4627346098423004,
+      "learning_rate": 0.0001,
+      "loss": 0.6416,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15570400822199384,
+      "grad_norm": 0.44164159893989563,
+      "learning_rate": 0.0001,
+      "loss": 0.5323,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15578965399109285,
+      "grad_norm": 0.4143899083137512,
+      "learning_rate": 0.0001,
+      "loss": 1.1847,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15587529976019185,
+      "grad_norm": 0.8894643187522888,
+      "learning_rate": 0.0001,
+      "loss": 1.0856,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15596094552929085,
+      "grad_norm": 0.5150936841964722,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15604659129838985,
+      "grad_norm": 0.4960081875324249,
+      "learning_rate": 0.0001,
+      "loss": 0.9871,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15613223706748885,
+      "grad_norm": 1.2571134567260742,
+      "learning_rate": 0.0001,
+      "loss": 0.86,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15621788283658788,
+      "grad_norm": 0.9449315071105957,
+      "learning_rate": 0.0001,
+      "loss": 1.2202,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1563035286056869,
+      "grad_norm": 0.7565779089927673,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1563891743747859,
+      "grad_norm": 0.4499366283416748,
+      "learning_rate": 0.0001,
+      "loss": 0.5426,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1564748201438849,
+      "grad_norm": 0.802484393119812,
+      "learning_rate": 0.0001,
+      "loss": 1.4319,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1565604659129839,
+      "grad_norm": 0.6967034339904785,
+      "learning_rate": 0.0001,
+      "loss": 1.0404,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1566461116820829,
+      "grad_norm": 0.6104376912117004,
+      "learning_rate": 0.0001,
+      "loss": 0.7041,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15673175745118192,
+      "grad_norm": 0.40670108795166016,
+      "learning_rate": 0.0001,
+      "loss": 0.4099,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15681740322028093,
+      "grad_norm": 0.5689238905906677,
+      "learning_rate": 0.0001,
+      "loss": 1.3251,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15690304898937993,
+      "grad_norm": 0.8716426491737366,
+      "learning_rate": 0.0001,
+      "loss": 1.4511,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15698869475847893,
+      "grad_norm": 0.5731133222579956,
+      "learning_rate": 0.0001,
+      "loss": 0.477,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15707434052757793,
+      "grad_norm": 0.7231414318084717,
+      "learning_rate": 0.0001,
+      "loss": 0.988,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15715998629667693,
+      "grad_norm": 0.7582845091819763,
+      "learning_rate": 0.0001,
+      "loss": 0.7383,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15724563206577596,
+      "grad_norm": 0.7874528169631958,
+      "learning_rate": 0.0001,
+      "loss": 0.8392,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15733127783487497,
+      "grad_norm": 0.5552375912666321,
+      "learning_rate": 0.0001,
+      "loss": 0.4488,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15741692360397397,
+      "grad_norm": 0.6916781663894653,
+      "learning_rate": 0.0001,
+      "loss": 1.0891,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15750256937307297,
+      "grad_norm": 0.7632482051849365,
+      "learning_rate": 0.0001,
+      "loss": 1.3001,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15758821514217197,
+      "grad_norm": 0.8387570381164551,
+      "learning_rate": 0.0001,
+      "loss": 0.8826,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15767386091127097,
+      "grad_norm": 0.6292305588722229,
+      "learning_rate": 0.0001,
+      "loss": 0.6528,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15775950668036998,
+      "grad_norm": 0.17916204035282135,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 1842
+    },
+    {
+      "epoch": 0.157845152449469,
+      "grad_norm": 0.9034995436668396,
+      "learning_rate": 0.0001,
+      "loss": 1.1571,
+      "step": 1843
+    },
+    {
+      "epoch": 0.157930798218568,
+      "grad_norm": 0.6832068562507629,
+      "learning_rate": 0.0001,
+      "loss": 0.6611,
+      "step": 1844
+    },
+    {
+      "epoch": 0.158016443987667,
+      "grad_norm": 0.6378197073936462,
+      "learning_rate": 0.0001,
+      "loss": 1.013,
+      "step": 1845
+    },
+    {
+      "epoch": 0.158102089756766,
+      "grad_norm": 0.6338422298431396,
+      "learning_rate": 0.0001,
+      "loss": 0.9799,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15818773552586501,
+      "grad_norm": 0.658716082572937,
+      "learning_rate": 0.0001,
+      "loss": 1.3057,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15827338129496402,
+      "grad_norm": 0.8775179982185364,
+      "learning_rate": 0.0001,
+      "loss": 1.1639,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15835902706406305,
+      "grad_norm": 0.5270436406135559,
+      "learning_rate": 0.0001,
+      "loss": 1.0689,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15844467283316205,
+      "grad_norm": 0.5294325947761536,
+      "learning_rate": 0.0001,
+      "loss": 0.8839,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15853031860226105,
+      "grad_norm": 0.5462505221366882,
+      "learning_rate": 0.0001,
+      "loss": 1.0747,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15861596437136005,
+      "grad_norm": 0.8299121856689453,
+      "learning_rate": 0.0001,
+      "loss": 0.8729,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15870161014045905,
+      "grad_norm": 0.6417173743247986,
+      "learning_rate": 0.0001,
+      "loss": 1.2293,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15878725590955806,
+      "grad_norm": 0.3668093979358673,
+      "learning_rate": 0.0001,
+      "loss": 0.7453,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15887290167865709,
+      "grad_norm": 0.7317532300949097,
+      "learning_rate": 0.0001,
+      "loss": 1.0658,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1589585474477561,
+      "grad_norm": 0.8103631734848022,
+      "learning_rate": 0.0001,
+      "loss": 1.0522,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1590441932168551,
+      "grad_norm": 0.6766731142997742,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1591298389859541,
+      "grad_norm": 0.5270974636077881,
+      "learning_rate": 0.0001,
+      "loss": 1.0903,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1592154847550531,
+      "grad_norm": 0.8822274208068848,
+      "learning_rate": 0.0001,
+      "loss": 1.0553,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1593011305241521,
+      "grad_norm": 0.4138229787349701,
+      "learning_rate": 0.0001,
+      "loss": 0.5953,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15938677629325113,
+      "grad_norm": 0.5601732134819031,
+      "learning_rate": 0.0001,
+      "loss": 0.9824,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15947242206235013,
+      "grad_norm": 0.7231619954109192,
+      "learning_rate": 0.0001,
+      "loss": 0.8214,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15955806783144913,
+      "grad_norm": 0.3140365481376648,
+      "learning_rate": 0.0001,
+      "loss": 0.4712,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15964371360054813,
+      "grad_norm": 0.8450373411178589,
+      "learning_rate": 0.0001,
+      "loss": 1.0686,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15972935936964713,
+      "grad_norm": 0.7886221408843994,
+      "learning_rate": 0.0001,
+      "loss": 1.9251,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15981500513874614,
+      "grad_norm": 0.5686867833137512,
+      "learning_rate": 0.0001,
+      "loss": 1.0661,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15990065090784517,
+      "grad_norm": 0.393904447555542,
+      "learning_rate": 0.0001,
+      "loss": 1.0569,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15998629667694417,
+      "grad_norm": 0.9074073433876038,
+      "learning_rate": 0.0001,
+      "loss": 1.2113,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16007194244604317,
+      "grad_norm": 0.7394840121269226,
+      "learning_rate": 0.0001,
+      "loss": 1.1878,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16015758821514217,
+      "grad_norm": 0.9281919002532959,
+      "learning_rate": 0.0001,
+      "loss": 0.4506,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16024323398424117,
+      "grad_norm": 1.0201468467712402,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16032887975334018,
+      "grad_norm": 0.6468662023544312,
+      "learning_rate": 0.0001,
+      "loss": 0.7439,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1604145255224392,
+      "grad_norm": 0.564475953578949,
+      "learning_rate": 0.0001,
+      "loss": 0.9788,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1605001712915382,
+      "grad_norm": 1.1282659769058228,
+      "learning_rate": 0.0001,
+      "loss": 1.098,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1605858170606372,
+      "grad_norm": 0.7583338618278503,
+      "learning_rate": 0.0001,
+      "loss": 0.9852,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1606714628297362,
+      "grad_norm": 0.7645503282546997,
+      "learning_rate": 0.0001,
+      "loss": 0.9948,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1607571085988352,
+      "grad_norm": 0.6193332672119141,
+      "learning_rate": 0.0001,
+      "loss": 1.096,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16084275436793422,
+      "grad_norm": 0.5938931107521057,
+      "learning_rate": 0.0001,
+      "loss": 1.0909,
+      "step": 1878
+    },
+    {
+      "epoch": 0.16092840013703322,
+      "grad_norm": 0.6068596243858337,
+      "learning_rate": 0.0001,
+      "loss": 1.2257,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16101404590613225,
+      "grad_norm": 1.3978294134140015,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16109969167523125,
+      "grad_norm": 0.632841944694519,
+      "learning_rate": 0.0001,
+      "loss": 0.9222,
+      "step": 1881
+    },
+    {
+      "epoch": 0.16118533744433025,
+      "grad_norm": 0.6158598065376282,
+      "learning_rate": 0.0001,
+      "loss": 1.01,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16127098321342925,
+      "grad_norm": 0.42953330278396606,
+      "learning_rate": 0.0001,
+      "loss": 0.7202,
+      "step": 1883
+    },
+    {
+      "epoch": 0.16135662898252826,
+      "grad_norm": 0.9300392866134644,
+      "learning_rate": 0.0001,
+      "loss": 1.124,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16144227475162726,
+      "grad_norm": 0.5228770971298218,
+      "learning_rate": 0.0001,
+      "loss": 1.0866,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1615279205207263,
+      "grad_norm": 0.48615843057632446,
+      "learning_rate": 0.0001,
+      "loss": 0.7317,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1616135662898253,
+      "grad_norm": 0.5262506008148193,
+      "learning_rate": 0.0001,
+      "loss": 1.1848,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1616992120589243,
+      "grad_norm": 0.9539384841918945,
+      "learning_rate": 0.0001,
+      "loss": 1.1044,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1617848578280233,
+      "grad_norm": 0.3965965807437897,
+      "learning_rate": 0.0001,
+      "loss": 0.44,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1618705035971223,
+      "grad_norm": 0.7258191108703613,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1619561493662213,
+      "grad_norm": 0.4563417136669159,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 1891
+    },
+    {
+      "epoch": 0.16204179513532033,
+      "grad_norm": 0.599514901638031,
+      "learning_rate": 0.0001,
+      "loss": 1.2605,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16212744090441933,
+      "grad_norm": 0.6591214537620544,
+      "learning_rate": 0.0001,
+      "loss": 0.6622,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16221308667351833,
+      "grad_norm": 0.7228677868843079,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16229873244261733,
+      "grad_norm": 1.074849247932434,
+      "learning_rate": 0.0001,
+      "loss": 1.1803,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16238437821171633,
+      "grad_norm": 0.8289828300476074,
+      "learning_rate": 0.0001,
+      "loss": 1.0049,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16247002398081534,
+      "grad_norm": 0.5321722626686096,
+      "learning_rate": 0.0001,
+      "loss": 0.8344,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16255566974991437,
+      "grad_norm": 0.6109360456466675,
+      "learning_rate": 0.0001,
+      "loss": 0.5927,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16264131551901337,
+      "grad_norm": 0.6584043502807617,
+      "learning_rate": 0.0001,
+      "loss": 0.8911,
+      "step": 1899
+    },
+    {
+      "epoch": 0.16272696128811237,
+      "grad_norm": 0.6200428605079651,
+      "learning_rate": 0.0001,
+      "loss": 0.6885,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16281260705721137,
+      "grad_norm": 0.6532961130142212,
+      "learning_rate": 0.0001,
+      "loss": 0.8352,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16289825282631037,
+      "grad_norm": 0.43170294165611267,
+      "learning_rate": 0.0001,
+      "loss": 0.8774,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16298389859540938,
+      "grad_norm": 0.5497574806213379,
+      "learning_rate": 0.0001,
+      "loss": 0.906,
+      "step": 1903
+    },
+    {
+      "epoch": 0.1630695443645084,
+      "grad_norm": 0.14097286760807037,
+      "learning_rate": 0.0001,
+      "loss": 0.6575,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1631551901336074,
+      "grad_norm": 0.7427582144737244,
+      "learning_rate": 0.0001,
+      "loss": 0.826,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1632408359027064,
+      "grad_norm": 0.34382855892181396,
+      "learning_rate": 0.0001,
+      "loss": 0.6852,
+      "step": 1906
+    },
+    {
+      "epoch": 0.1633264816718054,
+      "grad_norm": 0.4861757755279541,
+      "learning_rate": 0.0001,
+      "loss": 1.0795,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16341212744090441,
+      "grad_norm": 0.6481981873512268,
+      "learning_rate": 0.0001,
+      "loss": 1.1344,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16349777321000342,
+      "grad_norm": 0.8723146319389343,
+      "learning_rate": 0.0001,
+      "loss": 1.1704,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16358341897910242,
+      "grad_norm": 0.3013754189014435,
+      "learning_rate": 0.0001,
+      "loss": 0.6577,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16366906474820145,
+      "grad_norm": 0.52156001329422,
+      "learning_rate": 0.0001,
+      "loss": 1.0261,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16375471051730045,
+      "grad_norm": 0.9132922291755676,
+      "learning_rate": 0.0001,
+      "loss": 1.7813,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16384035628639945,
+      "grad_norm": 0.5787392258644104,
+      "learning_rate": 0.0001,
+      "loss": 1.3613,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16392600205549845,
+      "grad_norm": 0.6375370621681213,
+      "learning_rate": 0.0001,
+      "loss": 1.2543,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16401164782459746,
+      "grad_norm": 0.6506583094596863,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16409729359369646,
+      "grad_norm": 0.4967314600944519,
+      "learning_rate": 0.0001,
+      "loss": 1.3368,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1641829393627955,
+      "grad_norm": 0.3647153079509735,
+      "learning_rate": 0.0001,
+      "loss": 0.5401,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1642685851318945,
+      "grad_norm": 0.5072687864303589,
+      "learning_rate": 0.0001,
+      "loss": 0.9301,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1643542309009935,
+      "grad_norm": 0.49589407444000244,
+      "learning_rate": 0.0001,
+      "loss": 1.236,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1644398766700925,
+      "grad_norm": 0.665603518486023,
+      "learning_rate": 0.0001,
+      "loss": 1.0041,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1645255224391915,
+      "grad_norm": 0.8334211111068726,
+      "learning_rate": 0.0001,
+      "loss": 1.0185,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1646111682082905,
+      "grad_norm": 0.49192431569099426,
+      "learning_rate": 0.0001,
+      "loss": 0.5431,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16469681397738953,
+      "grad_norm": 0.3450392484664917,
+      "learning_rate": 0.0001,
+      "loss": 0.4689,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16478245974648853,
+      "grad_norm": 0.6566447615623474,
+      "learning_rate": 0.0001,
+      "loss": 1.0508,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16486810551558753,
+      "grad_norm": 0.6242066621780396,
+      "learning_rate": 0.0001,
+      "loss": 0.9243,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16495375128468653,
+      "grad_norm": 0.5845173597335815,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16503939705378554,
+      "grad_norm": 1.2919541597366333,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16512504282288454,
+      "grad_norm": 1.177848219871521,
+      "learning_rate": 0.0001,
+      "loss": 1.3391,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16521068859198357,
+      "grad_norm": 0.4772847890853882,
+      "learning_rate": 0.0001,
+      "loss": 0.4374,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16529633436108257,
+      "grad_norm": 0.7099941968917847,
+      "learning_rate": 0.0001,
+      "loss": 1.1614,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16538198013018157,
+      "grad_norm": 0.927446722984314,
+      "learning_rate": 0.0001,
+      "loss": 1.0564,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16546762589928057,
+      "grad_norm": 0.4641363322734833,
+      "learning_rate": 0.0001,
+      "loss": 1.1652,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16555327166837958,
+      "grad_norm": 0.7703837752342224,
+      "learning_rate": 0.0001,
+      "loss": 1.1278,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16563891743747858,
+      "grad_norm": 1.6441923379898071,
+      "learning_rate": 0.0001,
+      "loss": 0.6969,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1657245632065776,
+      "grad_norm": 0.6871307492256165,
+      "learning_rate": 0.0001,
+      "loss": 0.9112,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1658102089756766,
+      "grad_norm": 0.7469786405563354,
+      "learning_rate": 0.0001,
+      "loss": 0.9818,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1658958547447756,
+      "grad_norm": 0.5763053894042969,
+      "learning_rate": 0.0001,
+      "loss": 1.0413,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1659815005138746,
+      "grad_norm": 0.6928269863128662,
+      "learning_rate": 0.0001,
+      "loss": 0.8282,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16606714628297362,
+      "grad_norm": 1.153565764427185,
+      "learning_rate": 0.0001,
+      "loss": 1.1304,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16615279205207262,
+      "grad_norm": 0.655245840549469,
+      "learning_rate": 0.0001,
+      "loss": 1.1809,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16623843782117165,
+      "grad_norm": 0.6685817241668701,
+      "learning_rate": 0.0001,
+      "loss": 1.1107,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16632408359027065,
+      "grad_norm": 0.29637548327445984,
+      "learning_rate": 0.0001,
+      "loss": 0.5155,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16640972935936965,
+      "grad_norm": 0.5317182540893555,
+      "learning_rate": 0.0001,
+      "loss": 0.8806,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16649537512846865,
+      "grad_norm": 0.46637019515037537,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16658102089756766,
+      "grad_norm": 0.6588178277015686,
+      "learning_rate": 0.0001,
+      "loss": 0.5701,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.8332459330558777,
+      "learning_rate": 0.0001,
+      "loss": 1.0089,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16675231243576566,
+      "grad_norm": 1.061610460281372,
+      "learning_rate": 0.0001,
+      "loss": 0.8693,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1668379582048647,
+      "grad_norm": 0.6095160245895386,
+      "learning_rate": 0.0001,
+      "loss": 0.9916,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1669236039739637,
+      "grad_norm": 0.48993512988090515,
+      "learning_rate": 0.0001,
+      "loss": 1.0218,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1670092497430627,
+      "grad_norm": 0.9240294694900513,
+      "learning_rate": 0.0001,
+      "loss": 1.06,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1670948955121617,
+      "grad_norm": 0.5198048949241638,
+      "learning_rate": 0.0001,
+      "loss": 0.9983,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1671805412812607,
+      "grad_norm": 0.7735904455184937,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1672661870503597,
+      "grad_norm": 0.7814980149269104,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16735183281945873,
+      "grad_norm": 0.3246230185031891,
+      "learning_rate": 0.0001,
+      "loss": 0.4848,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16743747858855773,
+      "grad_norm": 0.8117189407348633,
+      "learning_rate": 0.0001,
+      "loss": 0.6069,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16752312435765673,
+      "grad_norm": 0.8723741769790649,
+      "learning_rate": 0.0001,
+      "loss": 1.1539,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16760877012675574,
+      "grad_norm": 0.5941243767738342,
+      "learning_rate": 0.0001,
+      "loss": 0.7995,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16769441589585474,
+      "grad_norm": 0.6367655992507935,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16778006166495374,
+      "grad_norm": 0.5853484272956848,
+      "learning_rate": 0.0001,
+      "loss": 0.7399,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16786570743405277,
+      "grad_norm": 0.9774623513221741,
+      "learning_rate": 0.0001,
+      "loss": 1.2437,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16795135320315177,
+      "grad_norm": 0.7943732738494873,
+      "learning_rate": 0.0001,
+      "loss": 1.0375,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16803699897225077,
+      "grad_norm": 0.7805942893028259,
+      "learning_rate": 0.0001,
+      "loss": 0.7461,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16812264474134977,
+      "grad_norm": 0.5757686495780945,
+      "learning_rate": 0.0001,
+      "loss": 1.1567,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16820829051044878,
+      "grad_norm": 0.9094904661178589,
+      "learning_rate": 0.0001,
+      "loss": 0.8184,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16829393627954778,
+      "grad_norm": 0.8887466192245483,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1683795820486468,
+      "grad_norm": 0.8191368579864502,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1684652278177458,
+      "grad_norm": 0.6382026672363281,
+      "learning_rate": 0.0001,
+      "loss": 0.3182,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1685508735868448,
+      "grad_norm": 0.4104350209236145,
+      "learning_rate": 0.0001,
+      "loss": 0.9494,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16863651935594381,
+      "grad_norm": 1.082144021987915,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16872216512504282,
+      "grad_norm": 0.7151246070861816,
+      "learning_rate": 0.0001,
+      "loss": 1.0718,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16880781089414182,
+      "grad_norm": 0.6418654918670654,
+      "learning_rate": 0.0001,
+      "loss": 1.0093,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16889345666324085,
+      "grad_norm": 0.6194281578063965,
+      "learning_rate": 0.0001,
+      "loss": 1.0639,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16897910243233985,
+      "grad_norm": 0.6166749596595764,
+      "learning_rate": 0.0001,
+      "loss": 0.8928,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16906474820143885,
+      "grad_norm": 0.8767044544219971,
+      "learning_rate": 0.0001,
+      "loss": 1.2197,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16915039397053785,
+      "grad_norm": 0.7782441973686218,
+      "learning_rate": 0.0001,
+      "loss": 1.3777,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16923603973963686,
+      "grad_norm": 0.8020647168159485,
+      "learning_rate": 0.0001,
+      "loss": 0.7033,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16932168550873586,
+      "grad_norm": 0.5283359885215759,
+      "learning_rate": 0.0001,
+      "loss": 1.0665,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1694073312778349,
+      "grad_norm": 0.8201982975006104,
+      "learning_rate": 0.0001,
+      "loss": 0.7716,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1694929770469339,
+      "grad_norm": 0.4894029200077057,
+      "learning_rate": 0.0001,
+      "loss": 1.2882,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1695786228160329,
+      "grad_norm": 0.9349902868270874,
+      "learning_rate": 0.0001,
+      "loss": 0.8617,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1696642685851319,
+      "grad_norm": 0.7546578049659729,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1697499143542309,
+      "grad_norm": 0.45774996280670166,
+      "learning_rate": 0.0001,
+      "loss": 0.4573,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1698355601233299,
+      "grad_norm": 0.684953510761261,
+      "learning_rate": 0.0001,
+      "loss": 0.9552,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1699212058924289,
+      "grad_norm": 0.6434280276298523,
+      "learning_rate": 0.0001,
+      "loss": 0.8175,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17000685166152793,
+      "grad_norm": 0.7761788368225098,
+      "learning_rate": 0.0001,
+      "loss": 0.5802,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17009249743062693,
+      "grad_norm": 0.5791441798210144,
+      "learning_rate": 0.0001,
+      "loss": 0.6526,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17017814319972593,
+      "grad_norm": 0.5552605390548706,
+      "learning_rate": 0.0001,
+      "loss": 0.9258,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17026378896882494,
+      "grad_norm": 0.8165088891983032,
+      "learning_rate": 0.0001,
+      "loss": 1.0831,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17034943473792394,
+      "grad_norm": 0.6286104917526245,
+      "learning_rate": 0.0001,
+      "loss": 1.1867,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17043508050702294,
+      "grad_norm": 0.9610998630523682,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17052072627612197,
+      "grad_norm": 0.6979968547821045,
+      "learning_rate": 0.0001,
+      "loss": 0.9629,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17060637204522097,
+      "grad_norm": 0.4046269655227661,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17069201781431997,
+      "grad_norm": 0.34363922476768494,
+      "learning_rate": 0.0001,
+      "loss": 0.3985,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17077766358341898,
+      "grad_norm": 0.5722994208335876,
+      "learning_rate": 0.0001,
+      "loss": 1.3634,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17086330935251798,
+      "grad_norm": 0.5308395624160767,
+      "learning_rate": 0.0001,
+      "loss": 1.0377,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17094895512161698,
+      "grad_norm": 0.4517698884010315,
+      "learning_rate": 0.0001,
+      "loss": 0.8743,
+      "step": 1996
+    },
+    {
+      "epoch": 0.171034600890716,
+      "grad_norm": 0.837763786315918,
+      "learning_rate": 0.0001,
+      "loss": 0.8595,
+      "step": 1997
+    },
+    {
+      "epoch": 0.171120246659815,
+      "grad_norm": 0.7925101518630981,
+      "learning_rate": 0.0001,
+      "loss": 0.7454,
+      "step": 1998
+    },
+    {
+      "epoch": 0.171205892428914,
+      "grad_norm": 0.6719907522201538,
+      "learning_rate": 0.0001,
+      "loss": 0.8076,
+      "step": 1999
+    },
+    {
+      "epoch": 0.17129153819801302,
+      "grad_norm": 0.7669806480407715,
+      "learning_rate": 0.0001,
+      "loss": 0.7643,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17137718396711202,
+      "grad_norm": 0.5630922317504883,
+      "learning_rate": 0.0001,
+      "loss": 1.1077,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17146282973621102,
+      "grad_norm": 0.23849493265151978,
+      "learning_rate": 0.0001,
+      "loss": 0.2164,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17154847550531005,
+      "grad_norm": 0.2672555446624756,
+      "learning_rate": 0.0001,
+      "loss": 0.4307,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17163412127440905,
+      "grad_norm": 0.8363445401191711,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17171976704350805,
+      "grad_norm": 0.7441173791885376,
+      "learning_rate": 0.0001,
+      "loss": 0.893,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17180541281260706,
+      "grad_norm": 0.3280055522918701,
+      "learning_rate": 0.0001,
+      "loss": 0.7268,
+      "step": 2006
+    },
+    {
+      "epoch": 0.17189105858170606,
+      "grad_norm": 1.0770773887634277,
+      "learning_rate": 0.0001,
+      "loss": 1.1707,
+      "step": 2007
+    },
+    {
+      "epoch": 0.17197670435080506,
+      "grad_norm": 0.47306346893310547,
+      "learning_rate": 0.0001,
+      "loss": 0.7383,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1720623501199041,
+      "grad_norm": 0.6291091442108154,
+      "learning_rate": 0.0001,
+      "loss": 0.4898,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1721479958890031,
+      "grad_norm": 0.6340357661247253,
+      "learning_rate": 0.0001,
+      "loss": 0.6913,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1722336416581021,
+      "grad_norm": 0.892682671546936,
+      "learning_rate": 0.0001,
+      "loss": 0.8216,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1723192874272011,
+      "grad_norm": 0.5426292419433594,
+      "learning_rate": 0.0001,
+      "loss": 1.3939,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1724049331963001,
+      "grad_norm": 1.01527738571167,
+      "learning_rate": 0.0001,
+      "loss": 1.3692,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1724905789653991,
+      "grad_norm": 0.6671502590179443,
+      "learning_rate": 0.0001,
+      "loss": 1.0848,
+      "step": 2014
+    },
+    {
+      "epoch": 0.17257622473449813,
+      "grad_norm": 0.6632112264633179,
+      "learning_rate": 0.0001,
+      "loss": 1.28,
+      "step": 2015
+    },
+    {
+      "epoch": 0.17266187050359713,
+      "grad_norm": 0.8504112362861633,
+      "learning_rate": 0.0001,
+      "loss": 0.7353,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17274751627269613,
+      "grad_norm": 0.9046083688735962,
+      "learning_rate": 0.0001,
+      "loss": 1.0021,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17283316204179514,
+      "grad_norm": 0.4168566167354584,
+      "learning_rate": 0.0001,
+      "loss": 0.7074,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17291880781089414,
+      "grad_norm": 1.0953240394592285,
+      "learning_rate": 0.0001,
+      "loss": 0.5447,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17300445357999314,
+      "grad_norm": 0.6924676895141602,
+      "learning_rate": 0.0001,
+      "loss": 1.0293,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17309009934909214,
+      "grad_norm": 0.481184720993042,
+      "learning_rate": 0.0001,
+      "loss": 1.0003,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17317574511819117,
+      "grad_norm": 0.6314960718154907,
+      "learning_rate": 0.0001,
+      "loss": 0.5524,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17326139088729017,
+      "grad_norm": 0.5634593367576599,
+      "learning_rate": 0.0001,
+      "loss": 0.8736,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17334703665638918,
+      "grad_norm": 0.8486800789833069,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17343268242548818,
+      "grad_norm": 1.021767497062683,
+      "learning_rate": 0.0001,
+      "loss": 1.0882,
+      "step": 2025
+    },
+    {
+      "epoch": 0.17351832819458718,
+      "grad_norm": 0.812950074672699,
+      "learning_rate": 0.0001,
+      "loss": 0.8207,
+      "step": 2026
+    },
+    {
+      "epoch": 0.17360397396368618,
+      "grad_norm": 0.5406774282455444,
+      "learning_rate": 0.0001,
+      "loss": 0.6771,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1736896197327852,
+      "grad_norm": 0.6816397905349731,
+      "learning_rate": 0.0001,
+      "loss": 0.7737,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1737752655018842,
+      "grad_norm": 0.5797739624977112,
+      "learning_rate": 0.0001,
+      "loss": 0.8266,
+      "step": 2029
+    },
+    {
+      "epoch": 0.17386091127098321,
+      "grad_norm": 0.5218822956085205,
+      "learning_rate": 0.0001,
+      "loss": 0.8367,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17394655704008222,
+      "grad_norm": 0.32372552156448364,
+      "learning_rate": 0.0001,
+      "loss": 0.4668,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17403220280918122,
+      "grad_norm": 0.6823001503944397,
+      "learning_rate": 0.0001,
+      "loss": 1.0324,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17411784857828022,
+      "grad_norm": 0.38322463631629944,
+      "learning_rate": 0.0001,
+      "loss": 0.3895,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17420349434737925,
+      "grad_norm": 0.636540949344635,
+      "learning_rate": 0.0001,
+      "loss": 1.084,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17428914011647825,
+      "grad_norm": 0.5731691122055054,
+      "learning_rate": 0.0001,
+      "loss": 0.8898,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17437478588557725,
+      "grad_norm": 0.7462801337242126,
+      "learning_rate": 0.0001,
+      "loss": 0.7372,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17446043165467626,
+      "grad_norm": 0.6988315582275391,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17454607742377526,
+      "grad_norm": 0.4037798345088959,
+      "learning_rate": 0.0001,
+      "loss": 0.655,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17463172319287426,
+      "grad_norm": 1.1386915445327759,
+      "learning_rate": 0.0001,
+      "loss": 1.0427,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1747173689619733,
+      "grad_norm": 0.5954161882400513,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1748030147310723,
+      "grad_norm": 0.5123779773712158,
+      "learning_rate": 0.0001,
+      "loss": 1.1448,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1748886605001713,
+      "grad_norm": 0.6903687715530396,
+      "learning_rate": 0.0001,
+      "loss": 1.0507,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1749743062692703,
+      "grad_norm": 0.36272501945495605,
+      "learning_rate": 0.0001,
+      "loss": 0.3515,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1750599520383693,
+      "grad_norm": 0.6867143511772156,
+      "learning_rate": 0.0001,
+      "loss": 1.1278,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1751455978074683,
+      "grad_norm": 0.8417600393295288,
+      "learning_rate": 0.0001,
+      "loss": 1.0465,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17523124357656733,
+      "grad_norm": 0.4661330282688141,
+      "learning_rate": 0.0001,
+      "loss": 0.6985,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17531688934566633,
+      "grad_norm": 0.43858253955841064,
+      "learning_rate": 0.0001,
+      "loss": 0.9057,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17540253511476533,
+      "grad_norm": 1.1882177591323853,
+      "learning_rate": 0.0001,
+      "loss": 1.1171,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17548818088386434,
+      "grad_norm": 0.5994183421134949,
+      "learning_rate": 0.0001,
+      "loss": 0.6579,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17557382665296334,
+      "grad_norm": 0.9538695216178894,
+      "learning_rate": 0.0001,
+      "loss": 1.2293,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17565947242206234,
+      "grad_norm": 0.8608779907226562,
+      "learning_rate": 0.0001,
+      "loss": 0.7855,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17574511819116137,
+      "grad_norm": 0.7391785383224487,
+      "learning_rate": 0.0001,
+      "loss": 0.562,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17583076396026037,
+      "grad_norm": 0.734028697013855,
+      "learning_rate": 0.0001,
+      "loss": 0.7067,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17591640972935937,
+      "grad_norm": 0.77808678150177,
+      "learning_rate": 0.0001,
+      "loss": 0.8925,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17600205549845838,
+      "grad_norm": 0.42552870512008667,
+      "learning_rate": 0.0001,
+      "loss": 0.7171,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17608770126755738,
+      "grad_norm": 0.6438249945640564,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17617334703665638,
+      "grad_norm": 0.4712925851345062,
+      "learning_rate": 0.0001,
+      "loss": 1.0903,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17625899280575538,
+      "grad_norm": 0.8840556740760803,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1763446385748544,
+      "grad_norm": 0.5111938714981079,
+      "learning_rate": 0.0001,
+      "loss": 0.9875,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17643028434395341,
+      "grad_norm": 0.5906338691711426,
+      "learning_rate": 0.0001,
+      "loss": 1.0491,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17651593011305242,
+      "grad_norm": 0.1722915917634964,
+      "learning_rate": 0.0001,
+      "loss": 0.3228,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17660157588215142,
+      "grad_norm": 1.1484708786010742,
+      "learning_rate": 0.0001,
+      "loss": 1.1975,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17668722165125042,
+      "grad_norm": 0.7394465804100037,
+      "learning_rate": 0.0001,
+      "loss": 0.6327,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17677286742034942,
+      "grad_norm": 0.5671532154083252,
+      "learning_rate": 0.0001,
+      "loss": 0.542,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17685851318944845,
+      "grad_norm": 0.4679796099662781,
+      "learning_rate": 0.0001,
+      "loss": 0.5684,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17694415895854745,
+      "grad_norm": 0.4526395797729492,
+      "learning_rate": 0.0001,
+      "loss": 0.7232,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17702980472764646,
+      "grad_norm": 0.7846791744232178,
+      "learning_rate": 0.0001,
+      "loss": 1.357,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17711545049674546,
+      "grad_norm": 0.9349731802940369,
+      "learning_rate": 0.0001,
+      "loss": 0.8512,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17720109626584446,
+      "grad_norm": 0.6114161610603333,
+      "learning_rate": 0.0001,
+      "loss": 0.6832,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17728674203494346,
+      "grad_norm": 0.6855459809303284,
+      "learning_rate": 0.0001,
+      "loss": 0.8503,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1773723878040425,
+      "grad_norm": 0.925615668296814,
+      "learning_rate": 0.0001,
+      "loss": 0.8322,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1774580335731415,
+      "grad_norm": 0.7053787112236023,
+      "learning_rate": 0.0001,
+      "loss": 0.6506,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1775436793422405,
+      "grad_norm": 0.44820746779441833,
+      "learning_rate": 0.0001,
+      "loss": 0.5716,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1776293251113395,
+      "grad_norm": 0.18874676525592804,
+      "learning_rate": 0.0001,
+      "loss": 0.7424,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1777149708804385,
+      "grad_norm": 1.0347059965133667,
+      "learning_rate": 0.0001,
+      "loss": 1.2085,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1778006166495375,
+      "grad_norm": 0.8323296904563904,
+      "learning_rate": 0.0001,
+      "loss": 0.9802,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17788626241863653,
+      "grad_norm": 0.46172964572906494,
+      "learning_rate": 0.0001,
+      "loss": 0.7228,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17797190818773553,
+      "grad_norm": 0.7047109007835388,
+      "learning_rate": 0.0001,
+      "loss": 0.8246,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17805755395683454,
+      "grad_norm": 0.9850871562957764,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17814319972593354,
+      "grad_norm": 0.7467983365058899,
+      "learning_rate": 0.0001,
+      "loss": 0.9105,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17822884549503254,
+      "grad_norm": 0.90313321352005,
+      "learning_rate": 0.0001,
+      "loss": 1.4166,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17831449126413154,
+      "grad_norm": 0.6680160164833069,
+      "learning_rate": 0.0001,
+      "loss": 0.919,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17840013703323057,
+      "grad_norm": 1.0075277090072632,
+      "learning_rate": 0.0001,
+      "loss": 1.2593,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17848578280232957,
+      "grad_norm": 0.7641968727111816,
+      "learning_rate": 0.0001,
+      "loss": 0.8107,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 0.6211161613464355,
+      "learning_rate": 0.0001,
+      "loss": 1.2962,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17865707434052758,
+      "grad_norm": 0.49535709619522095,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17874272010962658,
+      "grad_norm": 0.6597256064414978,
+      "learning_rate": 0.0001,
+      "loss": 0.8577,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17882836587872558,
+      "grad_norm": 0.5833244323730469,
+      "learning_rate": 0.0001,
+      "loss": 0.5137,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1789140116478246,
+      "grad_norm": 0.4666082262992859,
+      "learning_rate": 0.0001,
+      "loss": 0.5558,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1789996574169236,
+      "grad_norm": 0.8222501277923584,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17908530318602262,
+      "grad_norm": 0.7697572112083435,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17917094895512162,
+      "grad_norm": 0.5619245767593384,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17925659472422062,
+      "grad_norm": 0.6369295120239258,
+      "learning_rate": 0.0001,
+      "loss": 0.5198,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17934224049331962,
+      "grad_norm": 0.814294695854187,
+      "learning_rate": 0.0001,
+      "loss": 1.2195,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17942788626241862,
+      "grad_norm": 0.4355331361293793,
+      "learning_rate": 0.0001,
+      "loss": 1.0697,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17951353203151765,
+      "grad_norm": 0.549958348274231,
+      "learning_rate": 0.0001,
+      "loss": 0.8276,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17959917780061666,
+      "grad_norm": 0.6133109331130981,
+      "learning_rate": 0.0001,
+      "loss": 0.5636,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17968482356971566,
+      "grad_norm": 0.37027445435523987,
+      "learning_rate": 0.0001,
+      "loss": 0.5975,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17977046933881466,
+      "grad_norm": 0.8112702965736389,
+      "learning_rate": 0.0001,
+      "loss": 1.0481,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17985611510791366,
+      "grad_norm": 0.37297356128692627,
+      "learning_rate": 0.0001,
+      "loss": 0.8248,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17994176087701266,
+      "grad_norm": 0.6897878646850586,
+      "learning_rate": 0.0001,
+      "loss": 0.5471,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1800274066461117,
+      "grad_norm": 0.9608831405639648,
+      "learning_rate": 0.0001,
+      "loss": 0.9038,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1801130524152107,
+      "grad_norm": 0.7798635959625244,
+      "learning_rate": 0.0001,
+      "loss": 1.02,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1801986981843097,
+      "grad_norm": 0.9896133542060852,
+      "learning_rate": 0.0001,
+      "loss": 1.162,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1802843439534087,
+      "grad_norm": 0.8650609850883484,
+      "learning_rate": 0.0001,
+      "loss": 0.7915,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1803699897225077,
+      "grad_norm": 1.0715320110321045,
+      "learning_rate": 0.0001,
+      "loss": 1.2891,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1804556354916067,
+      "grad_norm": 0.5535070300102234,
+      "learning_rate": 0.0001,
+      "loss": 1.0709,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18054128126070573,
+      "grad_norm": 0.9960615634918213,
+      "learning_rate": 0.0001,
+      "loss": 0.9997,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18062692702980473,
+      "grad_norm": 0.4177529215812683,
+      "learning_rate": 0.0001,
+      "loss": 0.7429,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18071257279890374,
+      "grad_norm": 0.5388562083244324,
+      "learning_rate": 0.0001,
+      "loss": 0.9601,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18079821856800274,
+      "grad_norm": 0.4772290289402008,
+      "learning_rate": 0.0001,
+      "loss": 0.7863,
+      "step": 2111
+    },
+    {
+      "epoch": 0.18088386433710174,
+      "grad_norm": 0.6974534392356873,
+      "learning_rate": 0.0001,
+      "loss": 0.8627,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18096951010620074,
+      "grad_norm": 0.3828803300857544,
+      "learning_rate": 0.0001,
+      "loss": 0.6418,
+      "step": 2113
+    },
+    {
+      "epoch": 0.18105515587529977,
+      "grad_norm": 0.5145190954208374,
+      "learning_rate": 0.0001,
+      "loss": 1.0026,
+      "step": 2114
+    },
+    {
+      "epoch": 0.18114080164439877,
+      "grad_norm": 0.7201874256134033,
+      "learning_rate": 0.0001,
+      "loss": 1.1087,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18122644741349778,
+      "grad_norm": 1.0721187591552734,
+      "learning_rate": 0.0001,
+      "loss": 1.0686,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18131209318259678,
+      "grad_norm": 0.6360383033752441,
+      "learning_rate": 0.0001,
+      "loss": 1.1386,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18139773895169578,
+      "grad_norm": 0.7045168876647949,
+      "learning_rate": 0.0001,
+      "loss": 0.9346,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18148338472079478,
+      "grad_norm": 0.7102216482162476,
+      "learning_rate": 0.0001,
+      "loss": 0.9799,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1815690304898938,
+      "grad_norm": 0.7184001803398132,
+      "learning_rate": 0.0001,
+      "loss": 0.759,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18165467625899281,
+      "grad_norm": 0.6941666603088379,
+      "learning_rate": 0.0001,
+      "loss": 1.0613,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18174032202809182,
+      "grad_norm": 0.5484546422958374,
+      "learning_rate": 0.0001,
+      "loss": 0.6622,
+      "step": 2122
+    },
+    {
+      "epoch": 0.18182596779719082,
+      "grad_norm": 0.7323927283287048,
+      "learning_rate": 0.0001,
+      "loss": 1.069,
+      "step": 2123
+    },
+    {
+      "epoch": 0.18191161356628982,
+      "grad_norm": 0.7285987138748169,
+      "learning_rate": 0.0001,
+      "loss": 1.2945,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18199725933538882,
+      "grad_norm": 0.9084867238998413,
+      "learning_rate": 0.0001,
+      "loss": 1.2844,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18208290510448785,
+      "grad_norm": 0.5710561275482178,
+      "learning_rate": 0.0001,
+      "loss": 0.7282,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18216855087358685,
+      "grad_norm": 0.48194605112075806,
+      "learning_rate": 0.0001,
+      "loss": 0.703,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18225419664268586,
+      "grad_norm": 0.5310478806495667,
+      "learning_rate": 0.0001,
+      "loss": 0.6935,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18233984241178486,
+      "grad_norm": 0.7127010822296143,
+      "learning_rate": 0.0001,
+      "loss": 0.8697,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18242548818088386,
+      "grad_norm": 0.4950988292694092,
+      "learning_rate": 0.0001,
+      "loss": 0.698,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18251113394998286,
+      "grad_norm": 0.4113163948059082,
+      "learning_rate": 0.0001,
+      "loss": 0.8229,
+      "step": 2131
+    },
+    {
+      "epoch": 0.18259677971908186,
+      "grad_norm": 0.5818114280700684,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1826824254881809,
+      "grad_norm": 0.4883345365524292,
+      "learning_rate": 0.0001,
+      "loss": 1.1415,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1827680712572799,
+      "grad_norm": 0.8412510752677917,
+      "learning_rate": 0.0001,
+      "loss": 0.8627,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1828537170263789,
+      "grad_norm": 0.5145347714424133,
+      "learning_rate": 0.0001,
+      "loss": 1.0273,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1829393627954779,
+      "grad_norm": 0.9578337073326111,
+      "learning_rate": 0.0001,
+      "loss": 1.1022,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1830250085645769,
+      "grad_norm": 0.6158467531204224,
+      "learning_rate": 0.0001,
+      "loss": 1.0853,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1831106543336759,
+      "grad_norm": 0.43713608384132385,
+      "learning_rate": 0.0001,
+      "loss": 0.6391,
+      "step": 2138
+    },
+    {
+      "epoch": 0.18319630010277493,
+      "grad_norm": 1.1173853874206543,
+      "learning_rate": 0.0001,
+      "loss": 1.362,
+      "step": 2139
+    },
+    {
+      "epoch": 0.18328194587187394,
+      "grad_norm": 0.41750699281692505,
+      "learning_rate": 0.0001,
+      "loss": 0.9401,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18336759164097294,
+      "grad_norm": 0.656234622001648,
+      "learning_rate": 0.0001,
+      "loss": 1.075,
+      "step": 2141
+    },
+    {
+      "epoch": 0.18345323741007194,
+      "grad_norm": 0.6206073760986328,
+      "learning_rate": 0.0001,
+      "loss": 0.6681,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18353888317917094,
+      "grad_norm": 0.7364947199821472,
+      "learning_rate": 0.0001,
+      "loss": 0.6949,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18362452894826994,
+      "grad_norm": 0.4486753046512604,
+      "learning_rate": 0.0001,
+      "loss": 0.4502,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18371017471736897,
+      "grad_norm": 0.5995922684669495,
+      "learning_rate": 0.0001,
+      "loss": 0.7918,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18379582048646798,
+      "grad_norm": 0.7133471965789795,
+      "learning_rate": 0.0001,
+      "loss": 1.0777,
+      "step": 2146
+    },
+    {
+      "epoch": 0.18388146625556698,
+      "grad_norm": 0.46537837386131287,
+      "learning_rate": 0.0001,
+      "loss": 0.9365,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18396711202466598,
+      "grad_norm": 0.7020965814590454,
+      "learning_rate": 0.0001,
+      "loss": 0.7021,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18405275779376498,
+      "grad_norm": 0.667150616645813,
+      "learning_rate": 0.0001,
+      "loss": 1.2425,
+      "step": 2149
+    },
+    {
+      "epoch": 0.18413840356286398,
+      "grad_norm": 0.44835448265075684,
+      "learning_rate": 0.0001,
+      "loss": 0.3696,
+      "step": 2150
+    },
+    {
+      "epoch": 0.184224049331963,
+      "grad_norm": 0.7901645302772522,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18430969510106202,
+      "grad_norm": 0.48186245560646057,
+      "learning_rate": 0.0001,
+      "loss": 0.5277,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18439534087016102,
+      "grad_norm": 0.6769862174987793,
+      "learning_rate": 0.0001,
+      "loss": 1.0338,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18448098663926002,
+      "grad_norm": 0.537036120891571,
+      "learning_rate": 0.0001,
+      "loss": 1.3451,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18456663240835902,
+      "grad_norm": 0.6258284449577332,
+      "learning_rate": 0.0001,
+      "loss": 1.2491,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18465227817745802,
+      "grad_norm": 0.709747314453125,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18473792394655705,
+      "grad_norm": 0.6910694241523743,
+      "learning_rate": 0.0001,
+      "loss": 1.3186,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18482356971565606,
+      "grad_norm": 0.8891716599464417,
+      "learning_rate": 0.0001,
+      "loss": 1.1326,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18490921548475506,
+      "grad_norm": 0.7153406739234924,
+      "learning_rate": 0.0001,
+      "loss": 1.1976,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18499486125385406,
+      "grad_norm": 0.8618939518928528,
+      "learning_rate": 0.0001,
+      "loss": 0.8341,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18508050702295306,
+      "grad_norm": 0.7132860422134399,
+      "learning_rate": 0.0001,
+      "loss": 1.034,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18516615279205206,
+      "grad_norm": 0.6360827088356018,
+      "learning_rate": 0.0001,
+      "loss": 1.169,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18525179856115107,
+      "grad_norm": 1.0209712982177734,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1853374443302501,
+      "grad_norm": 0.7295277118682861,
+      "learning_rate": 0.0001,
+      "loss": 1.2289,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1854230900993491,
+      "grad_norm": 0.6087251901626587,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1855087358684481,
+      "grad_norm": 0.5576160550117493,
+      "learning_rate": 0.0001,
+      "loss": 0.7838,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1855943816375471,
+      "grad_norm": 1.0947442054748535,
+      "learning_rate": 0.0001,
+      "loss": 1.987,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1856800274066461,
+      "grad_norm": 0.5223337411880493,
+      "learning_rate": 0.0001,
+      "loss": 1.3038,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1857656731757451,
+      "grad_norm": 0.6004317998886108,
+      "learning_rate": 0.0001,
+      "loss": 0.948,
+      "step": 2169
+    },
+    {
+      "epoch": 0.18585131894484413,
+      "grad_norm": 0.3591175377368927,
+      "learning_rate": 0.0001,
+      "loss": 0.5164,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18593696471394314,
+      "grad_norm": 0.4674661457538605,
+      "learning_rate": 0.0001,
+      "loss": 0.8265,
+      "step": 2171
+    },
+    {
+      "epoch": 0.18602261048304214,
+      "grad_norm": 0.5220275521278381,
+      "learning_rate": 0.0001,
+      "loss": 0.6004,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18610825625214114,
+      "grad_norm": 0.6542509198188782,
+      "learning_rate": 0.0001,
+      "loss": 0.9629,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18619390202124014,
+      "grad_norm": 0.7798675894737244,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18627954779033915,
+      "grad_norm": 0.7980674505233765,
+      "learning_rate": 0.0001,
+      "loss": 0.6431,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18636519355943817,
+      "grad_norm": 0.8368552923202515,
+      "learning_rate": 0.0001,
+      "loss": 0.9799,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18645083932853718,
+      "grad_norm": 0.9978295564651489,
+      "learning_rate": 0.0001,
+      "loss": 0.9077,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18653648509763618,
+      "grad_norm": 0.8951623439788818,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18662213086673518,
+      "grad_norm": 0.506655752658844,
+      "learning_rate": 0.0001,
+      "loss": 0.6631,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18670777663583418,
+      "grad_norm": 0.5198215246200562,
+      "learning_rate": 0.0001,
+      "loss": 0.9777,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18679342240493318,
+      "grad_norm": 0.799736499786377,
+      "learning_rate": 0.0001,
+      "loss": 0.9792,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18687906817403221,
+      "grad_norm": 0.25604695081710815,
+      "learning_rate": 0.0001,
+      "loss": 0.5212,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18696471394313122,
+      "grad_norm": 0.7735158205032349,
+      "learning_rate": 0.0001,
+      "loss": 0.9004,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18705035971223022,
+      "grad_norm": 1.0071239471435547,
+      "learning_rate": 0.0001,
+      "loss": 1.1029,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18713600548132922,
+      "grad_norm": 0.5411102771759033,
+      "learning_rate": 0.0001,
+      "loss": 0.4684,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18722165125042822,
+      "grad_norm": 0.4372774660587311,
+      "learning_rate": 0.0001,
+      "loss": 0.459,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18730729701952722,
+      "grad_norm": 0.16695144772529602,
+      "learning_rate": 0.0001,
+      "loss": 0.4827,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18739294278862625,
+      "grad_norm": 0.7340555787086487,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18747858855772526,
+      "grad_norm": 0.9255430102348328,
+      "learning_rate": 0.0001,
+      "loss": 1.1659,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18756423432682426,
+      "grad_norm": 0.3670545816421509,
+      "learning_rate": 0.0001,
+      "loss": 0.5705,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18764988009592326,
+      "grad_norm": 0.9785749316215515,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18773552586502226,
+      "grad_norm": 0.37568432092666626,
+      "learning_rate": 0.0001,
+      "loss": 0.5115,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18782117163412126,
+      "grad_norm": 1.1744623184204102,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1879068174032203,
+      "grad_norm": 0.3058803379535675,
+      "learning_rate": 0.0001,
+      "loss": 0.4379,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1879924631723193,
+      "grad_norm": 0.6447697877883911,
+      "learning_rate": 0.0001,
+      "loss": 0.9727,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1880781089414183,
+      "grad_norm": 0.4673742949962616,
+      "learning_rate": 0.0001,
+      "loss": 1.1021,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1881637547105173,
+      "grad_norm": 0.5421869158744812,
+      "learning_rate": 0.0001,
+      "loss": 0.4101,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1882494004796163,
+      "grad_norm": 0.9251261353492737,
+      "learning_rate": 0.0001,
+      "loss": 1.0306,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1883350462487153,
+      "grad_norm": 1.0523439645767212,
+      "learning_rate": 0.0001,
+      "loss": 1.2586,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1884206920178143,
+      "grad_norm": 0.43254587054252625,
+      "learning_rate": 0.0001,
+      "loss": 0.7619,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18850633778691334,
+      "grad_norm": 0.5344021916389465,
+      "learning_rate": 0.0001,
+      "loss": 0.6839,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18859198355601234,
+      "grad_norm": 0.9358989596366882,
+      "learning_rate": 0.0001,
+      "loss": 0.9281,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18867762932511134,
+      "grad_norm": 0.6990581154823303,
+      "learning_rate": 0.0001,
+      "loss": 1.3393,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18876327509421034,
+      "grad_norm": 0.7360828518867493,
+      "learning_rate": 0.0001,
+      "loss": 1.0131,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18884892086330934,
+      "grad_norm": 0.5534975528717041,
+      "learning_rate": 0.0001,
+      "loss": 1.2052,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18893456663240835,
+      "grad_norm": 0.5091391801834106,
+      "learning_rate": 0.0001,
+      "loss": 0.5772,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18902021240150738,
+      "grad_norm": 0.7084749937057495,
+      "learning_rate": 0.0001,
+      "loss": 0.7574,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18910585817060638,
+      "grad_norm": 0.6432274580001831,
+      "learning_rate": 0.0001,
+      "loss": 0.8243,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18919150393970538,
+      "grad_norm": 0.6606919169425964,
+      "learning_rate": 0.0001,
+      "loss": 0.944,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18927714970880438,
+      "grad_norm": 0.5596864819526672,
+      "learning_rate": 0.0001,
+      "loss": 1.2417,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18936279547790338,
+      "grad_norm": 0.9406799077987671,
+      "learning_rate": 0.0001,
+      "loss": 1.1478,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18944844124700239,
+      "grad_norm": 0.4687572121620178,
+      "learning_rate": 0.0001,
+      "loss": 1.1464,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18953408701610142,
+      "grad_norm": 0.4228557348251343,
+      "learning_rate": 0.0001,
+      "loss": 0.717,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18961973278520042,
+      "grad_norm": 0.518473744392395,
+      "learning_rate": 0.0001,
+      "loss": 0.6155,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18970537855429942,
+      "grad_norm": 0.4265695810317993,
+      "learning_rate": 0.0001,
+      "loss": 0.8003,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18979102432339842,
+      "grad_norm": 0.6186901330947876,
+      "learning_rate": 0.0001,
+      "loss": 0.5091,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18987667009249742,
+      "grad_norm": 1.310078740119934,
+      "learning_rate": 0.0001,
+      "loss": 1.1572,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18996231586159643,
+      "grad_norm": 0.7485141158103943,
+      "learning_rate": 0.0001,
+      "loss": 0.6799,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19004796163069546,
+      "grad_norm": 0.7560403347015381,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19013360739979446,
+      "grad_norm": 1.1999495029449463,
+      "learning_rate": 0.0001,
+      "loss": 1.0373,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19021925316889346,
+      "grad_norm": 0.7842257618904114,
+      "learning_rate": 0.0001,
+      "loss": 1.2027,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19030489893799246,
+      "grad_norm": 0.6981107592582703,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19039054470709146,
+      "grad_norm": 0.5129143595695496,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 2223
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.8973177075386047,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1905618362452895,
+      "grad_norm": 0.609063982963562,
+      "learning_rate": 0.0001,
+      "loss": 0.94,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1906474820143885,
+      "grad_norm": 0.9106993079185486,
+      "learning_rate": 0.0001,
+      "loss": 1.0824,
+      "step": 2226
+    },
+    {
+      "epoch": 0.1907331277834875,
+      "grad_norm": 0.5078130960464478,
+      "learning_rate": 0.0001,
+      "loss": 1.0077,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1908187735525865,
+      "grad_norm": 0.9392686486244202,
+      "learning_rate": 0.0001,
+      "loss": 1.0145,
+      "step": 2228
+    },
+    {
+      "epoch": 0.1909044193216855,
+      "grad_norm": 0.7049013376235962,
+      "learning_rate": 0.0001,
+      "loss": 0.7835,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1909900650907845,
+      "grad_norm": 0.42175424098968506,
+      "learning_rate": 0.0001,
+      "loss": 0.4582,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19107571085988354,
+      "grad_norm": 0.5853713750839233,
+      "learning_rate": 0.0001,
+      "loss": 1.0841,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19116135662898254,
+      "grad_norm": 0.7127283215522766,
+      "learning_rate": 0.0001,
+      "loss": 0.852,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19124700239808154,
+      "grad_norm": 0.7456316351890564,
+      "learning_rate": 0.0001,
+      "loss": 0.6932,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19133264816718054,
+      "grad_norm": 0.6254094839096069,
+      "learning_rate": 0.0001,
+      "loss": 0.6706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19141829393627954,
+      "grad_norm": 0.42474493384361267,
+      "learning_rate": 0.0001,
+      "loss": 0.9975,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19150393970537855,
+      "grad_norm": 1.4154940843582153,
+      "learning_rate": 0.0001,
+      "loss": 1.9067,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19158958547447755,
+      "grad_norm": 0.6188265681266785,
+      "learning_rate": 0.0001,
+      "loss": 0.9155,
+      "step": 2237
+    },
+    {
+      "epoch": 0.19167523124357658,
+      "grad_norm": 0.7119168043136597,
+      "learning_rate": 0.0001,
+      "loss": 1.1654,
+      "step": 2238
+    },
+    {
+      "epoch": 0.19176087701267558,
+      "grad_norm": 0.9859538674354553,
+      "learning_rate": 0.0001,
+      "loss": 0.9805,
+      "step": 2239
+    },
+    {
+      "epoch": 0.19184652278177458,
+      "grad_norm": 0.6992178559303284,
+      "learning_rate": 0.0001,
+      "loss": 0.5135,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19193216855087358,
+      "grad_norm": 0.8260892033576965,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 2241
+    },
+    {
+      "epoch": 0.19201781431997259,
+      "grad_norm": 0.47449079155921936,
+      "learning_rate": 0.0001,
+      "loss": 1.0889,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1921034600890716,
+      "grad_norm": 1.0394604206085205,
+      "learning_rate": 0.0001,
+      "loss": 1.0589,
+      "step": 2243
+    },
+    {
+      "epoch": 0.19218910585817062,
+      "grad_norm": 0.5670893788337708,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19227475162726962,
+      "grad_norm": 0.7505695223808289,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19236039739636862,
+      "grad_norm": 0.9373841881752014,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19244604316546762,
+      "grad_norm": 0.6305624842643738,
+      "learning_rate": 0.0001,
+      "loss": 0.4442,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19253168893456662,
+      "grad_norm": 0.9237645268440247,
+      "learning_rate": 0.0001,
+      "loss": 1.1578,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19261733470366563,
+      "grad_norm": 0.45526617765426636,
+      "learning_rate": 0.0001,
+      "loss": 0.6161,
+      "step": 2249
+    },
+    {
+      "epoch": 0.19270298047276466,
+      "grad_norm": 0.9112492799758911,
+      "learning_rate": 0.0001,
+      "loss": 1.2657,
+      "step": 2250
+    },
+    {
+      "epoch": 0.19278862624186366,
+      "grad_norm": 1.0892130136489868,
+      "learning_rate": 0.0001,
+      "loss": 1.1988,
+      "step": 2251
+    },
+    {
+      "epoch": 0.19287427201096266,
+      "grad_norm": 0.6592493057250977,
+      "learning_rate": 0.0001,
+      "loss": 0.774,
+      "step": 2252
+    },
+    {
+      "epoch": 0.19295991778006166,
+      "grad_norm": 0.5916958451271057,
+      "learning_rate": 0.0001,
+      "loss": 1.0322,
+      "step": 2253
+    },
+    {
+      "epoch": 0.19304556354916066,
+      "grad_norm": 0.8105832934379578,
+      "learning_rate": 0.0001,
+      "loss": 1.3615,
+      "step": 2254
+    },
+    {
+      "epoch": 0.19313120931825967,
+      "grad_norm": 0.6777892112731934,
+      "learning_rate": 0.0001,
+      "loss": 1.0959,
+      "step": 2255
+    },
+    {
+      "epoch": 0.1932168550873587,
+      "grad_norm": 0.8586673736572266,
+      "learning_rate": 0.0001,
+      "loss": 1.2148,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1933025008564577,
+      "grad_norm": 0.6299570202827454,
+      "learning_rate": 0.0001,
+      "loss": 1.1066,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1933881466255567,
+      "grad_norm": 0.7466683983802795,
+      "learning_rate": 0.0001,
+      "loss": 0.9089,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1934737923946557,
+      "grad_norm": 0.5690147876739502,
+      "learning_rate": 0.0001,
+      "loss": 0.7963,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1935594381637547,
+      "grad_norm": 0.8689637780189514,
+      "learning_rate": 0.0001,
+      "loss": 0.7809,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1936450839328537,
+      "grad_norm": 0.8832905888557434,
+      "learning_rate": 0.0001,
+      "loss": 0.9854,
+      "step": 2261
+    },
+    {
+      "epoch": 0.19373072970195274,
+      "grad_norm": 0.879408061504364,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 2262
+    },
+    {
+      "epoch": 0.19381637547105174,
+      "grad_norm": 0.9965304732322693,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19390202124015074,
+      "grad_norm": 0.5694089531898499,
+      "learning_rate": 0.0001,
+      "loss": 0.7654,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19398766700924974,
+      "grad_norm": 0.6875213384628296,
+      "learning_rate": 0.0001,
+      "loss": 0.7344,
+      "step": 2265
+    },
+    {
+      "epoch": 0.19407331277834874,
+      "grad_norm": 0.7105584144592285,
+      "learning_rate": 0.0001,
+      "loss": 0.9309,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19415895854744775,
+      "grad_norm": 0.4918626546859741,
+      "learning_rate": 0.0001,
+      "loss": 1.0255,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19424460431654678,
+      "grad_norm": 0.34096020460128784,
+      "learning_rate": 0.0001,
+      "loss": 0.4166,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19433025008564578,
+      "grad_norm": 0.4603935182094574,
+      "learning_rate": 0.0001,
+      "loss": 1.0368,
+      "step": 2269
+    },
+    {
+      "epoch": 0.19441589585474478,
+      "grad_norm": 0.47337111830711365,
+      "learning_rate": 0.0001,
+      "loss": 1.02,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19450154162384378,
+      "grad_norm": 0.5140810608863831,
+      "learning_rate": 0.0001,
+      "loss": 0.6096,
+      "step": 2271
+    },
+    {
+      "epoch": 0.19458718739294278,
+      "grad_norm": 0.584954559803009,
+      "learning_rate": 0.0001,
+      "loss": 1.0313,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1946728331620418,
+      "grad_norm": 0.9750040173530579,
+      "learning_rate": 0.0001,
+      "loss": 0.9723,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1947584789311408,
+      "grad_norm": 0.8779977560043335,
+      "learning_rate": 0.0001,
+      "loss": 1.1659,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19484412470023982,
+      "grad_norm": 0.6130079030990601,
+      "learning_rate": 0.0001,
+      "loss": 0.8386,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19492977046933882,
+      "grad_norm": 0.9618662595748901,
+      "learning_rate": 0.0001,
+      "loss": 1.0029,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19501541623843782,
+      "grad_norm": 1.0734412670135498,
+      "learning_rate": 0.0001,
+      "loss": 1.1271,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19510106200753682,
+      "grad_norm": 0.7053017020225525,
+      "learning_rate": 0.0001,
+      "loss": 0.7154,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19518670777663583,
+      "grad_norm": 0.4249109625816345,
+      "learning_rate": 0.0001,
+      "loss": 0.9242,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19527235354573483,
+      "grad_norm": 1.023342490196228,
+      "learning_rate": 0.0001,
+      "loss": 0.6594,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19535799931483386,
+      "grad_norm": 0.9739706516265869,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19544364508393286,
+      "grad_norm": 0.555195689201355,
+      "learning_rate": 0.0001,
+      "loss": 0.8984,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19552929085303186,
+      "grad_norm": 1.0568645000457764,
+      "learning_rate": 0.0001,
+      "loss": 1.3066,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19561493662213086,
+      "grad_norm": 0.4821378290653229,
+      "learning_rate": 0.0001,
+      "loss": 1.1295,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19570058239122987,
+      "grad_norm": 0.632229208946228,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 2285
+    },
+    {
+      "epoch": 0.19578622816032887,
+      "grad_norm": 1.0002727508544922,
+      "learning_rate": 0.0001,
+      "loss": 0.778,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1958718739294279,
+      "grad_norm": 0.6924866437911987,
+      "learning_rate": 0.0001,
+      "loss": 0.8031,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1959575196985269,
+      "grad_norm": 1.2338014841079712,
+      "learning_rate": 0.0001,
+      "loss": 1.1465,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1960431654676259,
+      "grad_norm": 1.0141777992248535,
+      "learning_rate": 0.0001,
+      "loss": 0.9019,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1961288112367249,
+      "grad_norm": 0.8999325037002563,
+      "learning_rate": 0.0001,
+      "loss": 1.2282,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1962144570058239,
+      "grad_norm": 0.5787315368652344,
+      "learning_rate": 0.0001,
+      "loss": 1.0036,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1963001027749229,
+      "grad_norm": 0.6354789733886719,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 2292
+    },
+    {
+      "epoch": 0.19638574854402194,
+      "grad_norm": 1.0098079442977905,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 2293
+    },
+    {
+      "epoch": 0.19647139431312094,
+      "grad_norm": 1.0193783044815063,
+      "learning_rate": 0.0001,
+      "loss": 1.1346,
+      "step": 2294
+    },
+    {
+      "epoch": 0.19655704008221994,
+      "grad_norm": 0.5413192510604858,
+      "learning_rate": 0.0001,
+      "loss": 0.9844,
+      "step": 2295
+    },
+    {
+      "epoch": 0.19664268585131894,
+      "grad_norm": 0.6788745522499084,
+      "learning_rate": 0.0001,
+      "loss": 1.1311,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19672833162041795,
+      "grad_norm": 0.6446637511253357,
+      "learning_rate": 0.0001,
+      "loss": 0.7514,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19681397738951695,
+      "grad_norm": 0.8456950783729553,
+      "learning_rate": 0.0001,
+      "loss": 1.1329,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19689962315861598,
+      "grad_norm": 0.4296552836894989,
+      "learning_rate": 0.0001,
+      "loss": 0.773,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19698526892771498,
+      "grad_norm": 0.8030391335487366,
+      "learning_rate": 0.0001,
+      "loss": 0.8899,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19707091469681398,
+      "grad_norm": 0.7183400392532349,
+      "learning_rate": 0.0001,
+      "loss": 0.7472,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19715656046591298,
+      "grad_norm": 0.8419056534767151,
+      "learning_rate": 0.0001,
+      "loss": 0.7262,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19724220623501199,
+      "grad_norm": 0.7363333106040955,
+      "learning_rate": 0.0001,
+      "loss": 1.1755,
+      "step": 2303
+    },
+    {
+      "epoch": 0.197327852004111,
+      "grad_norm": 0.5453713536262512,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19741349777321002,
+      "grad_norm": 0.7369690537452698,
+      "learning_rate": 0.0001,
+      "loss": 0.97,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19749914354230902,
+      "grad_norm": 0.8270426988601685,
+      "learning_rate": 0.0001,
+      "loss": 0.7707,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19758478931140802,
+      "grad_norm": 0.5114608407020569,
+      "learning_rate": 0.0001,
+      "loss": 0.5426,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19767043508050702,
+      "grad_norm": 1.1092476844787598,
+      "learning_rate": 0.0001,
+      "loss": 0.9789,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19775608084960603,
+      "grad_norm": 0.6683344841003418,
+      "learning_rate": 0.0001,
+      "loss": 0.9189,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19784172661870503,
+      "grad_norm": 0.41577279567718506,
+      "learning_rate": 0.0001,
+      "loss": 0.6717,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19792737238780403,
+      "grad_norm": 0.6462950110435486,
+      "learning_rate": 0.0001,
+      "loss": 0.4861,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19801301815690306,
+      "grad_norm": 0.6850558519363403,
+      "learning_rate": 0.0001,
+      "loss": 0.3916,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19809866392600206,
+      "grad_norm": 0.6994614601135254,
+      "learning_rate": 0.0001,
+      "loss": 1.1556,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19818430969510106,
+      "grad_norm": 0.6227368116378784,
+      "learning_rate": 0.0001,
+      "loss": 0.4821,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19826995546420006,
+      "grad_norm": 0.8006405830383301,
+      "learning_rate": 0.0001,
+      "loss": 0.9075,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19835560123329907,
+      "grad_norm": 0.8459920287132263,
+      "learning_rate": 0.0001,
+      "loss": 1.0181,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19844124700239807,
+      "grad_norm": 0.5277412533760071,
+      "learning_rate": 0.0001,
+      "loss": 1.2906,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1985268927714971,
+      "grad_norm": 0.6967654824256897,
+      "learning_rate": 0.0001,
+      "loss": 1.2,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1986125385405961,
+      "grad_norm": 0.5985580086708069,
+      "learning_rate": 0.0001,
+      "loss": 0.6807,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1986981843096951,
+      "grad_norm": 0.7495021224021912,
+      "learning_rate": 0.0001,
+      "loss": 0.7262,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1987838300787941,
+      "grad_norm": 0.5697104930877686,
+      "learning_rate": 0.0001,
+      "loss": 0.8861,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1988694758478931,
+      "grad_norm": 0.49789586663246155,
+      "learning_rate": 0.0001,
+      "loss": 0.9976,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1989551216169921,
+      "grad_norm": 0.9102198481559753,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19904076738609114,
+      "grad_norm": 0.5957935452461243,
+      "learning_rate": 0.0001,
+      "loss": 0.5226,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19912641315519014,
+      "grad_norm": 0.8059887290000916,
+      "learning_rate": 0.0001,
+      "loss": 1.1202,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19921205892428914,
+      "grad_norm": 0.8895174860954285,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19929770469338814,
+      "grad_norm": 0.3572797179222107,
+      "learning_rate": 0.0001,
+      "loss": 0.7356,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19938335046248715,
+      "grad_norm": 1.9537373781204224,
+      "learning_rate": 0.0001,
+      "loss": 1.1956,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19946899623158615,
+      "grad_norm": 0.5239080190658569,
+      "learning_rate": 0.0001,
+      "loss": 0.9737,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19955464200068518,
+      "grad_norm": 0.8686486482620239,
+      "learning_rate": 0.0001,
+      "loss": 1.0522,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19964028776978418,
+      "grad_norm": 0.5517935156822205,
+      "learning_rate": 0.0001,
+      "loss": 0.8755,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19972593353888318,
+      "grad_norm": 0.7372665405273438,
+      "learning_rate": 0.0001,
+      "loss": 1.0221,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19981157930798218,
+      "grad_norm": 0.5680809020996094,
+      "learning_rate": 0.0001,
+      "loss": 0.7409,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1998972250770812,
+      "grad_norm": 0.5854288935661316,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1999828708461802,
+      "grad_norm": 0.4764133095741272,
+      "learning_rate": 0.0001,
+      "loss": 0.8522,
+      "step": 2335
+    },
+    {
+      "epoch": 0.20006851661527922,
+      "grad_norm": 0.44850796461105347,
+      "learning_rate": 0.0001,
+      "loss": 1.2329,
+      "step": 2336
+    },
+    {
+      "epoch": 0.20015416238437822,
+      "grad_norm": 0.7064544558525085,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 2337
+    },
+    {
+      "epoch": 0.20023980815347722,
+      "grad_norm": 0.5171502828598022,
+      "learning_rate": 0.0001,
+      "loss": 0.5881,
+      "step": 2338
+    },
+    {
+      "epoch": 0.20032545392257622,
+      "grad_norm": 0.559375524520874,
+      "learning_rate": 0.0001,
+      "loss": 0.8819,
+      "step": 2339
+    },
+    {
+      "epoch": 0.20041109969167523,
+      "grad_norm": 0.37707245349884033,
+      "learning_rate": 0.0001,
+      "loss": 0.8112,
+      "step": 2340
+    },
+    {
+      "epoch": 0.20049674546077423,
+      "grad_norm": 0.8846973776817322,
+      "learning_rate": 0.0001,
+      "loss": 0.9867,
+      "step": 2341
+    },
+    {
+      "epoch": 0.20058239122987326,
+      "grad_norm": 0.5781739950180054,
+      "learning_rate": 0.0001,
+      "loss": 0.8896,
+      "step": 2342
+    },
+    {
+      "epoch": 0.20066803699897226,
+      "grad_norm": 0.8377025127410889,
+      "learning_rate": 0.0001,
+      "loss": 1.1963,
+      "step": 2343
+    },
+    {
+      "epoch": 0.20075368276807126,
+      "grad_norm": 0.5393242835998535,
+      "learning_rate": 0.0001,
+      "loss": 0.7982,
+      "step": 2344
+    },
+    {
+      "epoch": 0.20083932853717026,
+      "grad_norm": 0.5944969654083252,
+      "learning_rate": 0.0001,
+      "loss": 0.6568,
+      "step": 2345
+    },
+    {
+      "epoch": 0.20092497430626927,
+      "grad_norm": 0.839867115020752,
+      "learning_rate": 0.0001,
+      "loss": 1.032,
+      "step": 2346
+    },
+    {
+      "epoch": 0.20101062007536827,
+      "grad_norm": 0.48574066162109375,
+      "learning_rate": 0.0001,
+      "loss": 1.2428,
+      "step": 2347
+    },
+    {
+      "epoch": 0.20109626584446727,
+      "grad_norm": 0.8741891980171204,
+      "learning_rate": 0.0001,
+      "loss": 1.2265,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2011819116135663,
+      "grad_norm": 0.7596957087516785,
+      "learning_rate": 0.0001,
+      "loss": 0.7518,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2012675573826653,
+      "grad_norm": 0.6587006449699402,
+      "learning_rate": 0.0001,
+      "loss": 0.6583,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2013532031517643,
+      "grad_norm": 0.4362960457801819,
+      "learning_rate": 0.0001,
+      "loss": 0.3785,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2014388489208633,
+      "grad_norm": 0.9645147323608398,
+      "learning_rate": 0.0001,
+      "loss": 0.8181,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2015244946899623,
+      "grad_norm": 0.6279506683349609,
+      "learning_rate": 0.0001,
+      "loss": 0.7214,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2016101404590613,
+      "grad_norm": 0.5727840662002563,
+      "learning_rate": 0.0001,
+      "loss": 0.964,
+      "step": 2354
+    },
+    {
+      "epoch": 0.20169578622816034,
+      "grad_norm": 0.8253939747810364,
+      "learning_rate": 0.0001,
+      "loss": 0.8296,
+      "step": 2355
+    },
+    {
+      "epoch": 0.20178143199725934,
+      "grad_norm": 0.56620192527771,
+      "learning_rate": 0.0001,
+      "loss": 1.1398,
+      "step": 2356
+    },
+    {
+      "epoch": 0.20186707776635834,
+      "grad_norm": 0.10872380435466766,
+      "learning_rate": 0.0001,
+      "loss": 0.4272,
+      "step": 2357
+    },
+    {
+      "epoch": 0.20195272353545735,
+      "grad_norm": 0.5316913723945618,
+      "learning_rate": 0.0001,
+      "loss": 0.3648,
+      "step": 2358
+    },
+    {
+      "epoch": 0.20203836930455635,
+      "grad_norm": 0.7980967164039612,
+      "learning_rate": 0.0001,
+      "loss": 0.6374,
+      "step": 2359
+    },
+    {
+      "epoch": 0.20212401507365535,
+      "grad_norm": 0.5504451990127563,
+      "learning_rate": 0.0001,
+      "loss": 1.1246,
+      "step": 2360
+    },
+    {
+      "epoch": 0.20220966084275438,
+      "grad_norm": 0.5690616369247437,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 2361
+    },
+    {
+      "epoch": 0.20229530661185338,
+      "grad_norm": 0.5935251712799072,
+      "learning_rate": 0.0001,
+      "loss": 1.3137,
+      "step": 2362
+    },
+    {
+      "epoch": 0.20238095238095238,
+      "grad_norm": 0.5186508297920227,
+      "learning_rate": 0.0001,
+      "loss": 0.9285,
+      "step": 2363
+    },
+    {
+      "epoch": 0.20246659815005139,
+      "grad_norm": 0.7508563995361328,
+      "learning_rate": 0.0001,
+      "loss": 2.1831,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2025522439191504,
+      "grad_norm": 0.6714151501655579,
+      "learning_rate": 0.0001,
+      "loss": 0.9411,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2026378896882494,
+      "grad_norm": 0.38137638568878174,
+      "learning_rate": 0.0001,
+      "loss": 0.792,
+      "step": 2366
+    },
+    {
+      "epoch": 0.20272353545734842,
+      "grad_norm": 0.6428802609443665,
+      "learning_rate": 0.0001,
+      "loss": 1.0336,
+      "step": 2367
+    },
+    {
+      "epoch": 0.20280918122644742,
+      "grad_norm": 0.6338450312614441,
+      "learning_rate": 0.0001,
+      "loss": 0.8106,
+      "step": 2368
+    },
+    {
+      "epoch": 0.20289482699554642,
+      "grad_norm": 0.553137481212616,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 2369
+    },
+    {
+      "epoch": 0.20298047276464543,
+      "grad_norm": 0.416166752576828,
+      "learning_rate": 0.0001,
+      "loss": 0.5143,
+      "step": 2370
+    },
+    {
+      "epoch": 0.20306611853374443,
+      "grad_norm": 0.5472464561462402,
+      "learning_rate": 0.0001,
+      "loss": 0.8538,
+      "step": 2371
+    },
+    {
+      "epoch": 0.20315176430284343,
+      "grad_norm": 0.6779702305793762,
+      "learning_rate": 0.0001,
+      "loss": 0.6524,
+      "step": 2372
+    },
+    {
+      "epoch": 0.20323741007194246,
+      "grad_norm": 0.42521747946739197,
+      "learning_rate": 0.0001,
+      "loss": 0.5291,
+      "step": 2373
+    },
+    {
+      "epoch": 0.20332305584104146,
+      "grad_norm": 0.7929352521896362,
+      "learning_rate": 0.0001,
+      "loss": 1.2358,
+      "step": 2374
+    },
+    {
+      "epoch": 0.20340870161014046,
+      "grad_norm": 0.6603885889053345,
+      "learning_rate": 0.0001,
+      "loss": 0.7075,
+      "step": 2375
+    },
+    {
+      "epoch": 0.20349434737923947,
+      "grad_norm": 1.002261757850647,
+      "learning_rate": 0.0001,
+      "loss": 1.2252,
+      "step": 2376
+    },
+    {
+      "epoch": 0.20357999314833847,
+      "grad_norm": 0.8575955033302307,
+      "learning_rate": 0.0001,
+      "loss": 1.0397,
+      "step": 2377
+    },
+    {
+      "epoch": 0.20366563891743747,
+      "grad_norm": 0.27905499935150146,
+      "learning_rate": 0.0001,
+      "loss": 0.4792,
+      "step": 2378
+    },
+    {
+      "epoch": 0.20375128468653647,
+      "grad_norm": 0.7859793305397034,
+      "learning_rate": 0.0001,
+      "loss": 0.7004,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2038369304556355,
+      "grad_norm": 1.049201250076294,
+      "learning_rate": 0.0001,
+      "loss": 1.2389,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2039225762247345,
+      "grad_norm": 0.5715323686599731,
+      "learning_rate": 0.0001,
+      "loss": 0.88,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2040082219938335,
+      "grad_norm": 0.8204423785209656,
+      "learning_rate": 0.0001,
+      "loss": 1.0923,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2040938677629325,
+      "grad_norm": 0.6737041473388672,
+      "learning_rate": 0.0001,
+      "loss": 0.7488,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2041795135320315,
+      "grad_norm": 1.0734483003616333,
+      "learning_rate": 0.0001,
+      "loss": 1.3234,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2042651593011305,
+      "grad_norm": 0.7888963222503662,
+      "learning_rate": 0.0001,
+      "loss": 1.0457,
+      "step": 2385
+    },
+    {
+      "epoch": 0.20435080507022954,
+      "grad_norm": 0.7871268391609192,
+      "learning_rate": 0.0001,
+      "loss": 1.0872,
+      "step": 2386
+    },
+    {
+      "epoch": 0.20443645083932854,
+      "grad_norm": 0.8746002912521362,
+      "learning_rate": 0.0001,
+      "loss": 1.2261,
+      "step": 2387
+    },
+    {
+      "epoch": 0.20452209660842754,
+      "grad_norm": 0.8491378426551819,
+      "learning_rate": 0.0001,
+      "loss": 0.8353,
+      "step": 2388
+    },
+    {
+      "epoch": 0.20460774237752655,
+      "grad_norm": 0.48957696557044983,
+      "learning_rate": 0.0001,
+      "loss": 0.514,
+      "step": 2389
+    },
+    {
+      "epoch": 0.20469338814662555,
+      "grad_norm": 0.2811725437641144,
+      "learning_rate": 0.0001,
+      "loss": 0.4599,
+      "step": 2390
+    },
+    {
+      "epoch": 0.20477903391572455,
+      "grad_norm": 0.7925663590431213,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 2391
+    },
+    {
+      "epoch": 0.20486467968482358,
+      "grad_norm": 0.43385612964630127,
+      "learning_rate": 0.0001,
+      "loss": 0.9652,
+      "step": 2392
+    },
+    {
+      "epoch": 0.20495032545392258,
+      "grad_norm": 0.7627438902854919,
+      "learning_rate": 0.0001,
+      "loss": 1.0378,
+      "step": 2393
+    },
+    {
+      "epoch": 0.20503597122302158,
+      "grad_norm": 0.8170188069343567,
+      "learning_rate": 0.0001,
+      "loss": 1.2759,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2051216169921206,
+      "grad_norm": 0.7043712735176086,
+      "learning_rate": 0.0001,
+      "loss": 1.2085,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2052072627612196,
+      "grad_norm": 0.6933789849281311,
+      "learning_rate": 0.0001,
+      "loss": 1.249,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2052929085303186,
+      "grad_norm": 0.24092549085617065,
+      "learning_rate": 0.0001,
+      "loss": 0.3683,
+      "step": 2397
+    },
+    {
+      "epoch": 0.20537855429941762,
+      "grad_norm": 0.5231190323829651,
+      "learning_rate": 0.0001,
+      "loss": 0.7393,
+      "step": 2398
+    },
+    {
+      "epoch": 0.20546420006851662,
+      "grad_norm": 0.806840717792511,
+      "learning_rate": 0.0001,
+      "loss": 1.7467,
+      "step": 2399
+    },
+    {
+      "epoch": 0.20554984583761562,
+      "grad_norm": 0.7976426482200623,
+      "learning_rate": 0.0001,
+      "loss": 0.7095,
+      "step": 2400
+    },
+    {
+      "epoch": 0.20563549160671463,
+      "grad_norm": 0.720623254776001,
+      "learning_rate": 0.0001,
+      "loss": 0.9371,
+      "step": 2401
+    },
+    {
+      "epoch": 0.20572113737581363,
+      "grad_norm": 0.5279609560966492,
+      "learning_rate": 0.0001,
+      "loss": 0.3225,
+      "step": 2402
+    },
+    {
+      "epoch": 0.20580678314491263,
+      "grad_norm": 0.716867208480835,
+      "learning_rate": 0.0001,
+      "loss": 1.0666,
+      "step": 2403
+    },
+    {
+      "epoch": 0.20589242891401166,
+      "grad_norm": 0.5238314867019653,
+      "learning_rate": 0.0001,
+      "loss": 1.3133,
+      "step": 2404
+    },
+    {
+      "epoch": 0.20597807468311066,
+      "grad_norm": 0.8668951392173767,
+      "learning_rate": 0.0001,
+      "loss": 0.9297,
+      "step": 2405
+    },
+    {
+      "epoch": 0.20606372045220966,
+      "grad_norm": 0.6887823343276978,
+      "learning_rate": 0.0001,
+      "loss": 0.9651,
+      "step": 2406
+    },
+    {
+      "epoch": 0.20614936622130867,
+      "grad_norm": 0.7090550661087036,
+      "learning_rate": 0.0001,
+      "loss": 0.9225,
+      "step": 2407
+    },
+    {
+      "epoch": 0.20623501199040767,
+      "grad_norm": 0.4663130044937134,
+      "learning_rate": 0.0001,
+      "loss": 0.783,
+      "step": 2408
+    },
+    {
+      "epoch": 0.20632065775950667,
+      "grad_norm": 0.5062718987464905,
+      "learning_rate": 0.0001,
+      "loss": 0.867,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2064063035286057,
+      "grad_norm": 0.975720226764679,
+      "learning_rate": 0.0001,
+      "loss": 1.203,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2064919492977047,
+      "grad_norm": 0.6222168803215027,
+      "learning_rate": 0.0001,
+      "loss": 1.0429,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2065775950668037,
+      "grad_norm": 0.8442457914352417,
+      "learning_rate": 0.0001,
+      "loss": 1.2114,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2066632408359027,
+      "grad_norm": 0.6274807453155518,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2067488866050017,
+      "grad_norm": 0.5210537910461426,
+      "learning_rate": 0.0001,
+      "loss": 1.3541,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2068345323741007,
+      "grad_norm": 0.730255663394928,
+      "learning_rate": 0.0001,
+      "loss": 0.9345,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2069201781431997,
+      "grad_norm": 0.5785207152366638,
+      "learning_rate": 0.0001,
+      "loss": 0.9997,
+      "step": 2416
+    },
+    {
+      "epoch": 0.20700582391229874,
+      "grad_norm": 0.8463599681854248,
+      "learning_rate": 0.0001,
+      "loss": 1.0245,
+      "step": 2417
+    },
+    {
+      "epoch": 0.20709146968139774,
+      "grad_norm": 0.6571394205093384,
+      "learning_rate": 0.0001,
+      "loss": 0.5825,
+      "step": 2418
+    },
+    {
+      "epoch": 0.20717711545049675,
+      "grad_norm": 0.7708553075790405,
+      "learning_rate": 0.0001,
+      "loss": 1.0184,
+      "step": 2419
+    },
+    {
+      "epoch": 0.20726276121959575,
+      "grad_norm": 0.5965237617492676,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 2420
+    },
+    {
+      "epoch": 0.20734840698869475,
+      "grad_norm": 0.551938533782959,
+      "learning_rate": 0.0001,
+      "loss": 0.9135,
+      "step": 2421
+    },
+    {
+      "epoch": 0.20743405275779375,
+      "grad_norm": 0.9425605535507202,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 2422
+    },
+    {
+      "epoch": 0.20751969852689278,
+      "grad_norm": 0.4917559325695038,
+      "learning_rate": 0.0001,
+      "loss": 1.2129,
+      "step": 2423
+    },
+    {
+      "epoch": 0.20760534429599178,
+      "grad_norm": 0.4894367456436157,
+      "learning_rate": 0.0001,
+      "loss": 1.0306,
+      "step": 2424
+    },
+    {
+      "epoch": 0.20769099006509079,
+      "grad_norm": 0.5576333999633789,
+      "learning_rate": 0.0001,
+      "loss": 1.1505,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2077766358341898,
+      "grad_norm": 0.5824314951896667,
+      "learning_rate": 0.0001,
+      "loss": 1.148,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2078622816032888,
+      "grad_norm": 0.8896623253822327,
+      "learning_rate": 0.0001,
+      "loss": 1.2811,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2079479273723878,
+      "grad_norm": 1.1864899396896362,
+      "learning_rate": 0.0001,
+      "loss": 1.1311,
+      "step": 2428
+    },
+    {
+      "epoch": 0.20803357314148682,
+      "grad_norm": 0.6005644798278809,
+      "learning_rate": 0.0001,
+      "loss": 1.1156,
+      "step": 2429
+    },
+    {
+      "epoch": 0.20811921891058582,
+      "grad_norm": 0.5673446655273438,
+      "learning_rate": 0.0001,
+      "loss": 0.879,
+      "step": 2430
+    },
+    {
+      "epoch": 0.20820486467968483,
+      "grad_norm": 1.3155996799468994,
+      "learning_rate": 0.0001,
+      "loss": 1.2515,
+      "step": 2431
+    },
+    {
+      "epoch": 0.20829051044878383,
+      "grad_norm": 0.8506407737731934,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 2432
+    },
+    {
+      "epoch": 0.20837615621788283,
+      "grad_norm": 0.7196330428123474,
+      "learning_rate": 0.0001,
+      "loss": 1.064,
+      "step": 2433
+    },
+    {
+      "epoch": 0.20846180198698183,
+      "grad_norm": 0.6397178173065186,
+      "learning_rate": 0.0001,
+      "loss": 0.5488,
+      "step": 2434
+    },
+    {
+      "epoch": 0.20854744775608086,
+      "grad_norm": 0.4893302321434021,
+      "learning_rate": 0.0001,
+      "loss": 1.102,
+      "step": 2435
+    },
+    {
+      "epoch": 0.20863309352517986,
+      "grad_norm": 1.0160610675811768,
+      "learning_rate": 0.0001,
+      "loss": 1.1483,
+      "step": 2436
+    },
+    {
+      "epoch": 0.20871873929427887,
+      "grad_norm": 0.46453535556793213,
+      "learning_rate": 0.0001,
+      "loss": 0.4974,
+      "step": 2437
+    },
+    {
+      "epoch": 0.20880438506337787,
+      "grad_norm": 0.49540552496910095,
+      "learning_rate": 0.0001,
+      "loss": 0.3717,
+      "step": 2438
+    },
+    {
+      "epoch": 0.20889003083247687,
+      "grad_norm": 0.6269391775131226,
+      "learning_rate": 0.0001,
+      "loss": 1.4419,
+      "step": 2439
+    },
+    {
+      "epoch": 0.20897567660157587,
+      "grad_norm": 0.9501695036888123,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2090613223706749,
+      "grad_norm": 0.8228318691253662,
+      "learning_rate": 0.0001,
+      "loss": 1.2561,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2091469681397739,
+      "grad_norm": 0.8185625672340393,
+      "learning_rate": 0.0001,
+      "loss": 0.7597,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2092326139088729,
+      "grad_norm": 1.0248770713806152,
+      "learning_rate": 0.0001,
+      "loss": 1.095,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2093182596779719,
+      "grad_norm": 0.8818730711936951,
+      "learning_rate": 0.0001,
+      "loss": 0.761,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2094039054470709,
+      "grad_norm": 0.8697795271873474,
+      "learning_rate": 0.0001,
+      "loss": 1.1054,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2094895512161699,
+      "grad_norm": 0.4705776870250702,
+      "learning_rate": 0.0001,
+      "loss": 0.9703,
+      "step": 2446
+    },
+    {
+      "epoch": 0.20957519698526894,
+      "grad_norm": 0.7976250052452087,
+      "learning_rate": 0.0001,
+      "loss": 1.2529,
+      "step": 2447
+    },
+    {
+      "epoch": 0.20966084275436794,
+      "grad_norm": 0.600182831287384,
+      "learning_rate": 0.0001,
+      "loss": 1.2479,
+      "step": 2448
+    },
+    {
+      "epoch": 0.20974648852346695,
+      "grad_norm": 0.6620160937309265,
+      "learning_rate": 0.0001,
+      "loss": 0.8345,
+      "step": 2449
+    },
+    {
+      "epoch": 0.20983213429256595,
+      "grad_norm": 0.4058929681777954,
+      "learning_rate": 0.0001,
+      "loss": 0.3901,
+      "step": 2450
+    },
+    {
+      "epoch": 0.20991778006166495,
+      "grad_norm": 0.6226440668106079,
+      "learning_rate": 0.0001,
+      "loss": 1.0681,
+      "step": 2451
+    },
+    {
+      "epoch": 0.21000342583076395,
+      "grad_norm": 0.415536105632782,
+      "learning_rate": 0.0001,
+      "loss": 0.755,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21008907159986295,
+      "grad_norm": 0.7010723948478699,
+      "learning_rate": 0.0001,
+      "loss": 0.911,
+      "step": 2453
+    },
+    {
+      "epoch": 0.21017471736896198,
+      "grad_norm": 0.6934406161308289,
+      "learning_rate": 0.0001,
+      "loss": 0.4563,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21026036313806098,
+      "grad_norm": 0.15937013924121857,
+      "learning_rate": 0.0001,
+      "loss": 0.5501,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21034600890716,
+      "grad_norm": 0.6352848410606384,
+      "learning_rate": 0.0001,
+      "loss": 0.4078,
+      "step": 2456
+    },
+    {
+      "epoch": 0.210431654676259,
+      "grad_norm": 0.873002290725708,
+      "learning_rate": 0.0001,
+      "loss": 1.8689,
+      "step": 2457
+    },
+    {
+      "epoch": 0.210517300445358,
+      "grad_norm": 0.5335526466369629,
+      "learning_rate": 0.0001,
+      "loss": 0.8091,
+      "step": 2458
+    },
+    {
+      "epoch": 0.210602946214457,
+      "grad_norm": 0.5837013721466064,
+      "learning_rate": 0.0001,
+      "loss": 0.613,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21068859198355602,
+      "grad_norm": 0.8050652742385864,
+      "learning_rate": 0.0001,
+      "loss": 0.521,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21077423775265502,
+      "grad_norm": 0.810369074344635,
+      "learning_rate": 0.0001,
+      "loss": 0.9301,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21085988352175403,
+      "grad_norm": 0.5661274790763855,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21094552929085303,
+      "grad_norm": 0.6218873262405396,
+      "learning_rate": 0.0001,
+      "loss": 0.6405,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21103117505995203,
+      "grad_norm": 1.0180445909500122,
+      "learning_rate": 0.0001,
+      "loss": 0.9745,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21111682082905103,
+      "grad_norm": 0.544060468673706,
+      "learning_rate": 0.0001,
+      "loss": 0.7134,
+      "step": 2465
+    },
+    {
+      "epoch": 0.21120246659815006,
+      "grad_norm": 0.28936195373535156,
+      "learning_rate": 0.0001,
+      "loss": 0.6315,
+      "step": 2466
+    },
+    {
+      "epoch": 0.21128811236724906,
+      "grad_norm": 0.325719952583313,
+      "learning_rate": 0.0001,
+      "loss": 1.0679,
+      "step": 2467
+    },
+    {
+      "epoch": 0.21137375813634807,
+      "grad_norm": 0.4708036184310913,
+      "learning_rate": 0.0001,
+      "loss": 0.3544,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21145940390544707,
+      "grad_norm": 0.8400583863258362,
+      "learning_rate": 0.0001,
+      "loss": 0.7624,
+      "step": 2469
+    },
+    {
+      "epoch": 0.21154504967454607,
+      "grad_norm": 0.5084882378578186,
+      "learning_rate": 0.0001,
+      "loss": 1.0929,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21163069544364507,
+      "grad_norm": 0.5214126110076904,
+      "learning_rate": 0.0001,
+      "loss": 0.954,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2117163412127441,
+      "grad_norm": 0.814346969127655,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2118019869818431,
+      "grad_norm": 0.6169758439064026,
+      "learning_rate": 0.0001,
+      "loss": 0.8346,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2118876327509421,
+      "grad_norm": 1.0014954805374146,
+      "learning_rate": 0.0001,
+      "loss": 1.3247,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2119732785200411,
+      "grad_norm": 0.7037865519523621,
+      "learning_rate": 0.0001,
+      "loss": 1.1146,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2120589242891401,
+      "grad_norm": 0.3697775602340698,
+      "learning_rate": 0.0001,
+      "loss": 0.387,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2121445700582391,
+      "grad_norm": 0.8027715682983398,
+      "learning_rate": 0.0001,
+      "loss": 0.9391,
+      "step": 2477
+    },
+    {
+      "epoch": 0.21223021582733814,
+      "grad_norm": 0.6004588603973389,
+      "learning_rate": 0.0001,
+      "loss": 0.4505,
+      "step": 2478
+    },
+    {
+      "epoch": 0.21231586159643714,
+      "grad_norm": 0.6318751573562622,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 2479
+    },
+    {
+      "epoch": 0.21240150736553615,
+      "grad_norm": 0.7008181214332581,
+      "learning_rate": 0.0001,
+      "loss": 0.9365,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21248715313463515,
+      "grad_norm": 0.7087741494178772,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 2481
+    },
+    {
+      "epoch": 0.21257279890373415,
+      "grad_norm": 0.7008094787597656,
+      "learning_rate": 0.0001,
+      "loss": 0.82,
+      "step": 2482
+    },
+    {
+      "epoch": 0.21265844467283315,
+      "grad_norm": 1.032899260520935,
+      "learning_rate": 0.0001,
+      "loss": 1.0904,
+      "step": 2483
+    },
+    {
+      "epoch": 0.21274409044193218,
+      "grad_norm": 0.7010834813117981,
+      "learning_rate": 0.0001,
+      "loss": 0.9342,
+      "step": 2484
+    },
+    {
+      "epoch": 0.21282973621103118,
+      "grad_norm": 0.29541081190109253,
+      "learning_rate": 0.0001,
+      "loss": 0.5293,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2129153819801302,
+      "grad_norm": 0.4540451765060425,
+      "learning_rate": 0.0001,
+      "loss": 0.967,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2130010277492292,
+      "grad_norm": 0.811636745929718,
+      "learning_rate": 0.0001,
+      "loss": 0.7304,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2130866735183282,
+      "grad_norm": 0.6492415070533752,
+      "learning_rate": 0.0001,
+      "loss": 1.3318,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2131723192874272,
+      "grad_norm": 0.6688328981399536,
+      "learning_rate": 0.0001,
+      "loss": 0.7096,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2132579650565262,
+      "grad_norm": 0.8949735760688782,
+      "learning_rate": 0.0001,
+      "loss": 0.8392,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21334361082562522,
+      "grad_norm": 0.8359949588775635,
+      "learning_rate": 0.0001,
+      "loss": 0.9598,
+      "step": 2491
+    },
+    {
+      "epoch": 0.21342925659472423,
+      "grad_norm": 0.35091179609298706,
+      "learning_rate": 0.0001,
+      "loss": 0.3214,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21351490236382323,
+      "grad_norm": 1.1070080995559692,
+      "learning_rate": 0.0001,
+      "loss": 1.1172,
+      "step": 2493
+    },
+    {
+      "epoch": 0.21360054813292223,
+      "grad_norm": 0.5251277089118958,
+      "learning_rate": 0.0001,
+      "loss": 0.8567,
+      "step": 2494
+    },
+    {
+      "epoch": 0.21368619390202123,
+      "grad_norm": 0.499639630317688,
+      "learning_rate": 0.0001,
+      "loss": 1.2945,
+      "step": 2495
+    },
+    {
+      "epoch": 0.21377183967112023,
+      "grad_norm": 0.754711925983429,
+      "learning_rate": 0.0001,
+      "loss": 0.7549,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21385748544021926,
+      "grad_norm": 0.8328573703765869,
+      "learning_rate": 0.0001,
+      "loss": 1.0768,
+      "step": 2497
+    },
+    {
+      "epoch": 0.21394313120931827,
+      "grad_norm": 0.6577131152153015,
+      "learning_rate": 0.0001,
+      "loss": 1.0544,
+      "step": 2498
+    },
+    {
+      "epoch": 0.21402877697841727,
+      "grad_norm": 0.8434766530990601,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 2499
+    },
+    {
+      "epoch": 0.21411442274751627,
+      "grad_norm": 0.48015642166137695,
+      "learning_rate": 0.0001,
+      "loss": 0.7719,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21420006851661527,
+      "grad_norm": 0.4269872009754181,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 2501
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.7513227462768555,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2143713600548133,
+      "grad_norm": 0.4831150472164154,
+      "learning_rate": 0.0001,
+      "loss": 0.6513,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2144570058239123,
+      "grad_norm": 1.1485960483551025,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2145426515930113,
+      "grad_norm": 0.709112823009491,
+      "learning_rate": 0.0001,
+      "loss": 1.0861,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2146282973621103,
+      "grad_norm": 0.526587188243866,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2147139431312093,
+      "grad_norm": 0.8737342357635498,
+      "learning_rate": 0.0001,
+      "loss": 0.706,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2147995889003083,
+      "grad_norm": 1.119794249534607,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 2508
+    },
+    {
+      "epoch": 0.21488523466940734,
+      "grad_norm": 0.5183132886886597,
+      "learning_rate": 0.0001,
+      "loss": 0.9617,
+      "step": 2509
+    },
+    {
+      "epoch": 0.21497088043850635,
+      "grad_norm": 0.6347960233688354,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21505652620760535,
+      "grad_norm": 0.6949610114097595,
+      "learning_rate": 0.0001,
+      "loss": 1.1885,
+      "step": 2511
+    },
+    {
+      "epoch": 0.21514217197670435,
+      "grad_norm": 0.9597961902618408,
+      "learning_rate": 0.0001,
+      "loss": 1.1311,
+      "step": 2512
+    },
+    {
+      "epoch": 0.21522781774580335,
+      "grad_norm": 0.9095984697341919,
+      "learning_rate": 0.0001,
+      "loss": 0.7709,
+      "step": 2513
+    },
+    {
+      "epoch": 0.21531346351490235,
+      "grad_norm": 0.6644256711006165,
+      "learning_rate": 0.0001,
+      "loss": 1.0158,
+      "step": 2514
+    },
+    {
+      "epoch": 0.21539910928400138,
+      "grad_norm": 0.8798538446426392,
+      "learning_rate": 0.0001,
+      "loss": 0.8763,
+      "step": 2515
+    },
+    {
+      "epoch": 0.21548475505310039,
+      "grad_norm": 0.6769435405731201,
+      "learning_rate": 0.0001,
+      "loss": 0.9965,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2155704008221994,
+      "grad_norm": 0.5459786653518677,
+      "learning_rate": 0.0001,
+      "loss": 1.1403,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2156560465912984,
+      "grad_norm": 0.7016478776931763,
+      "learning_rate": 0.0001,
+      "loss": 1.0574,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2157416923603974,
+      "grad_norm": 0.6253093481063843,
+      "learning_rate": 0.0001,
+      "loss": 1.0048,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2158273381294964,
+      "grad_norm": 0.4033289849758148,
+      "learning_rate": 0.0001,
+      "loss": 0.5533,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21591298389859542,
+      "grad_norm": 1.121386170387268,
+      "learning_rate": 0.0001,
+      "loss": 1.2796,
+      "step": 2521
+    },
+    {
+      "epoch": 0.21599862966769443,
+      "grad_norm": 0.7210914492607117,
+      "learning_rate": 0.0001,
+      "loss": 0.7384,
+      "step": 2522
+    },
+    {
+      "epoch": 0.21608427543679343,
+      "grad_norm": 0.5668298602104187,
+      "learning_rate": 0.0001,
+      "loss": 1.0168,
+      "step": 2523
+    },
+    {
+      "epoch": 0.21616992120589243,
+      "grad_norm": 0.5169442296028137,
+      "learning_rate": 0.0001,
+      "loss": 1.3101,
+      "step": 2524
+    },
+    {
+      "epoch": 0.21625556697499143,
+      "grad_norm": 0.6568666696548462,
+      "learning_rate": 0.0001,
+      "loss": 1.1393,
+      "step": 2525
+    },
+    {
+      "epoch": 0.21634121274409043,
+      "grad_norm": 0.5770350098609924,
+      "learning_rate": 0.0001,
+      "loss": 1.1576,
+      "step": 2526
+    },
+    {
+      "epoch": 0.21642685851318944,
+      "grad_norm": 0.6878929734230042,
+      "learning_rate": 0.0001,
+      "loss": 1.0098,
+      "step": 2527
+    },
+    {
+      "epoch": 0.21651250428228846,
+      "grad_norm": 1.062159776687622,
+      "learning_rate": 0.0001,
+      "loss": 0.8221,
+      "step": 2528
+    },
+    {
+      "epoch": 0.21659815005138747,
+      "grad_norm": 0.34046757221221924,
+      "learning_rate": 0.0001,
+      "loss": 0.7684,
+      "step": 2529
+    },
+    {
+      "epoch": 0.21668379582048647,
+      "grad_norm": 0.9847444891929626,
+      "learning_rate": 0.0001,
+      "loss": 0.6461,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21676944158958547,
+      "grad_norm": 1.2064244747161865,
+      "learning_rate": 0.0001,
+      "loss": 0.8817,
+      "step": 2531
+    },
+    {
+      "epoch": 0.21685508735868447,
+      "grad_norm": 0.2731172442436218,
+      "learning_rate": 0.0001,
+      "loss": 0.2498,
+      "step": 2532
+    },
+    {
+      "epoch": 0.21694073312778347,
+      "grad_norm": 0.8961516618728638,
+      "learning_rate": 0.0001,
+      "loss": 0.9277,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2170263788968825,
+      "grad_norm": 0.6136212348937988,
+      "learning_rate": 0.0001,
+      "loss": 1.0686,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2171120246659815,
+      "grad_norm": 1.0040104389190674,
+      "learning_rate": 0.0001,
+      "loss": 1.4053,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2171976704350805,
+      "grad_norm": 0.6500115394592285,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2172833162041795,
+      "grad_norm": 0.4064091444015503,
+      "learning_rate": 0.0001,
+      "loss": 0.9495,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2173689619732785,
+      "grad_norm": 0.7345540523529053,
+      "learning_rate": 0.0001,
+      "loss": 0.9316,
+      "step": 2538
+    },
+    {
+      "epoch": 0.21745460774237751,
+      "grad_norm": 0.919444739818573,
+      "learning_rate": 0.0001,
+      "loss": 1.1268,
+      "step": 2539
+    },
+    {
+      "epoch": 0.21754025351147654,
+      "grad_norm": 0.503440260887146,
+      "learning_rate": 0.0001,
+      "loss": 1.0391,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21762589928057555,
+      "grad_norm": 0.8314846158027649,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 2541
+    },
+    {
+      "epoch": 0.21771154504967455,
+      "grad_norm": 0.6271324753761292,
+      "learning_rate": 0.0001,
+      "loss": 0.8296,
+      "step": 2542
+    },
+    {
+      "epoch": 0.21779719081877355,
+      "grad_norm": 0.6477083563804626,
+      "learning_rate": 0.0001,
+      "loss": 0.9445,
+      "step": 2543
+    },
+    {
+      "epoch": 0.21788283658787255,
+      "grad_norm": 0.5233659148216248,
+      "learning_rate": 0.0001,
+      "loss": 1.3425,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21796848235697155,
+      "grad_norm": 0.7149603366851807,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21805412812607058,
+      "grad_norm": 1.1675821542739868,
+      "learning_rate": 0.0001,
+      "loss": 1.9671,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2181397738951696,
+      "grad_norm": 0.6186656951904297,
+      "learning_rate": 0.0001,
+      "loss": 0.6839,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2182254196642686,
+      "grad_norm": 0.6949894428253174,
+      "learning_rate": 0.0001,
+      "loss": 1.0215,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2183110654333676,
+      "grad_norm": 0.4193609654903412,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2183967112024666,
+      "grad_norm": 0.49367403984069824,
+      "learning_rate": 0.0001,
+      "loss": 0.9354,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2184823569715656,
+      "grad_norm": 0.6383912563323975,
+      "learning_rate": 0.0001,
+      "loss": 0.9701,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21856800274066462,
+      "grad_norm": 0.7145264148712158,
+      "learning_rate": 0.0001,
+      "loss": 0.6443,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21865364850976363,
+      "grad_norm": 1.0531916618347168,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21873929427886263,
+      "grad_norm": 0.9574797749519348,
+      "learning_rate": 0.0001,
+      "loss": 1.0597,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21882494004796163,
+      "grad_norm": 0.7070325016975403,
+      "learning_rate": 0.0001,
+      "loss": 0.6067,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21891058581706063,
+      "grad_norm": 0.39756128191947937,
+      "learning_rate": 0.0001,
+      "loss": 0.9902,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21899623158615963,
+      "grad_norm": 0.5281681418418884,
+      "learning_rate": 0.0001,
+      "loss": 1.1065,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21908187735525866,
+      "grad_norm": 0.44648686051368713,
+      "learning_rate": 0.0001,
+      "loss": 0.4752,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21916752312435767,
+      "grad_norm": 0.962680459022522,
+      "learning_rate": 0.0001,
+      "loss": 0.5178,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21925316889345667,
+      "grad_norm": 1.2362395524978638,
+      "learning_rate": 0.0001,
+      "loss": 0.7038,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21933881466255567,
+      "grad_norm": 0.9817659258842468,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21942446043165467,
+      "grad_norm": 0.9602881669998169,
+      "learning_rate": 0.0001,
+      "loss": 0.9948,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21951010620075367,
+      "grad_norm": 0.4155106544494629,
+      "learning_rate": 0.0001,
+      "loss": 0.8587,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21959575196985268,
+      "grad_norm": 0.4459531307220459,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2196813977389517,
+      "grad_norm": 0.6696712970733643,
+      "learning_rate": 0.0001,
+      "loss": 0.533,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2197670435080507,
+      "grad_norm": 0.538658618927002,
+      "learning_rate": 0.0001,
+      "loss": 1.0672,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2198526892771497,
+      "grad_norm": 0.8996044397354126,
+      "learning_rate": 0.0001,
+      "loss": 0.8715,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2199383350462487,
+      "grad_norm": 0.9772889614105225,
+      "learning_rate": 0.0001,
+      "loss": 0.8873,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22002398081534771,
+      "grad_norm": 0.7712833285331726,
+      "learning_rate": 0.0001,
+      "loss": 0.7402,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22010962658444672,
+      "grad_norm": 0.7803371548652649,
+      "learning_rate": 0.0001,
+      "loss": 0.5363,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22019527235354575,
+      "grad_norm": 0.7573778033256531,
+      "learning_rate": 0.0001,
+      "loss": 0.7118,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22028091812264475,
+      "grad_norm": 0.8000575304031372,
+      "learning_rate": 0.0001,
+      "loss": 0.9956,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22036656389174375,
+      "grad_norm": 0.4268200099468231,
+      "learning_rate": 0.0001,
+      "loss": 0.6154,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22045220966084275,
+      "grad_norm": 0.6368035078048706,
+      "learning_rate": 0.0001,
+      "loss": 1.0051,
+      "step": 2574
+    },
+    {
+      "epoch": 0.22053785542994175,
+      "grad_norm": 0.7205711007118225,
+      "learning_rate": 0.0001,
+      "loss": 0.7023,
+      "step": 2575
+    },
+    {
+      "epoch": 0.22062350119904076,
+      "grad_norm": 1.0784815549850464,
+      "learning_rate": 0.0001,
+      "loss": 0.8713,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22070914696813979,
+      "grad_norm": 1.0732301473617554,
+      "learning_rate": 0.0001,
+      "loss": 0.8097,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2207947927372388,
+      "grad_norm": 0.9392873644828796,
+      "learning_rate": 0.0001,
+      "loss": 0.9874,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2208804385063378,
+      "grad_norm": 1.1252259016036987,
+      "learning_rate": 0.0001,
+      "loss": 1.0372,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2209660842754368,
+      "grad_norm": 0.6531725525856018,
+      "learning_rate": 0.0001,
+      "loss": 0.4906,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2210517300445358,
+      "grad_norm": 0.7280725240707397,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2211373758136348,
+      "grad_norm": 0.6957252621650696,
+      "learning_rate": 0.0001,
+      "loss": 0.6516,
+      "step": 2582
+    },
+    {
+      "epoch": 0.22122302158273383,
+      "grad_norm": 0.693309485912323,
+      "learning_rate": 0.0001,
+      "loss": 1.0727,
+      "step": 2583
+    },
+    {
+      "epoch": 0.22130866735183283,
+      "grad_norm": 1.206518530845642,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22139431312093183,
+      "grad_norm": 0.9570721387863159,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22147995889003083,
+      "grad_norm": 0.2310360074043274,
+      "learning_rate": 0.0001,
+      "loss": 0.7405,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22156560465912983,
+      "grad_norm": 0.5621093511581421,
+      "learning_rate": 0.0001,
+      "loss": 0.5474,
+      "step": 2587
+    },
+    {
+      "epoch": 0.22165125042822884,
+      "grad_norm": 1.1206847429275513,
+      "learning_rate": 0.0001,
+      "loss": 1.0804,
+      "step": 2588
+    },
+    {
+      "epoch": 0.22173689619732787,
+      "grad_norm": 0.9489470720291138,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 2589
+    },
+    {
+      "epoch": 0.22182254196642687,
+      "grad_norm": 0.7575852870941162,
+      "learning_rate": 0.0001,
+      "loss": 1.1891,
+      "step": 2590
+    },
+    {
+      "epoch": 0.22190818773552587,
+      "grad_norm": 1.0652724504470825,
+      "learning_rate": 0.0001,
+      "loss": 0.5829,
+      "step": 2591
+    },
+    {
+      "epoch": 0.22199383350462487,
+      "grad_norm": 0.3949885964393616,
+      "learning_rate": 0.0001,
+      "loss": 0.2589,
+      "step": 2592
+    },
+    {
+      "epoch": 0.22207947927372387,
+      "grad_norm": 1.0714274644851685,
+      "learning_rate": 0.0001,
+      "loss": 1.1974,
+      "step": 2593
+    },
+    {
+      "epoch": 0.22216512504282288,
+      "grad_norm": 0.6371351480484009,
+      "learning_rate": 0.0001,
+      "loss": 0.7164,
+      "step": 2594
+    },
+    {
+      "epoch": 0.22225077081192188,
+      "grad_norm": 0.5054485201835632,
+      "learning_rate": 0.0001,
+      "loss": 0.7663,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2223364165810209,
+      "grad_norm": 0.5325796604156494,
+      "learning_rate": 0.0001,
+      "loss": 1.2151,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2224220623501199,
+      "grad_norm": 0.48135730624198914,
+      "learning_rate": 0.0001,
+      "loss": 0.9188,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2225077081192189,
+      "grad_norm": 0.9029459953308105,
+      "learning_rate": 0.0001,
+      "loss": 1.2357,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2225933538883179,
+      "grad_norm": 0.6153978705406189,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 2599
+    },
+    {
+      "epoch": 0.22267899965741692,
+      "grad_norm": 0.7585981488227844,
+      "learning_rate": 0.0001,
+      "loss": 0.8257,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22276464542651592,
+      "grad_norm": 0.5993663668632507,
+      "learning_rate": 0.0001,
+      "loss": 1.0696,
+      "step": 2601
+    },
+    {
+      "epoch": 0.22285029119561495,
+      "grad_norm": 0.7512819766998291,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 2602
+    },
+    {
+      "epoch": 0.22293593696471395,
+      "grad_norm": 0.837230920791626,
+      "learning_rate": 0.0001,
+      "loss": 0.9701,
+      "step": 2603
+    },
+    {
+      "epoch": 0.22302158273381295,
+      "grad_norm": 0.527381956577301,
+      "learning_rate": 0.0001,
+      "loss": 1.0361,
+      "step": 2604
+    },
+    {
+      "epoch": 0.22310722850291195,
+      "grad_norm": 0.5158360600471497,
+      "learning_rate": 0.0001,
+      "loss": 0.8857,
+      "step": 2605
+    },
+    {
+      "epoch": 0.22319287427201095,
+      "grad_norm": 0.4970020055770874,
+      "learning_rate": 0.0001,
+      "loss": 1.0368,
+      "step": 2606
+    },
+    {
+      "epoch": 0.22327852004110996,
+      "grad_norm": 0.49405404925346375,
+      "learning_rate": 0.0001,
+      "loss": 0.8273,
+      "step": 2607
+    },
+    {
+      "epoch": 0.223364165810209,
+      "grad_norm": 0.9038569331169128,
+      "learning_rate": 0.0001,
+      "loss": 1.161,
+      "step": 2608
+    },
+    {
+      "epoch": 0.223449811579308,
+      "grad_norm": 0.7680282592773438,
+      "learning_rate": 0.0001,
+      "loss": 1.0042,
+      "step": 2609
+    },
+    {
+      "epoch": 0.223535457348407,
+      "grad_norm": 0.4833921492099762,
+      "learning_rate": 0.0001,
+      "loss": 0.3751,
+      "step": 2610
+    },
+    {
+      "epoch": 0.223621103117506,
+      "grad_norm": 0.6842895746231079,
+      "learning_rate": 0.0001,
+      "loss": 0.8853,
+      "step": 2611
+    },
+    {
+      "epoch": 0.223706748886605,
+      "grad_norm": 0.587503969669342,
+      "learning_rate": 0.0001,
+      "loss": 1.0361,
+      "step": 2612
+    },
+    {
+      "epoch": 0.223792394655704,
+      "grad_norm": 0.6925882697105408,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 2613
+    },
+    {
+      "epoch": 0.22387804042480303,
+      "grad_norm": 0.4596776068210602,
+      "learning_rate": 0.0001,
+      "loss": 0.9524,
+      "step": 2614
+    },
+    {
+      "epoch": 0.22396368619390203,
+      "grad_norm": 0.8958559632301331,
+      "learning_rate": 0.0001,
+      "loss": 1.22,
+      "step": 2615
+    },
+    {
+      "epoch": 0.22404933196300103,
+      "grad_norm": 0.8892364501953125,
+      "learning_rate": 0.0001,
+      "loss": 1.3947,
+      "step": 2616
+    },
+    {
+      "epoch": 0.22413497773210003,
+      "grad_norm": 1.029266119003296,
+      "learning_rate": 0.0001,
+      "loss": 1.4312,
+      "step": 2617
+    },
+    {
+      "epoch": 0.22422062350119903,
+      "grad_norm": 1.090889573097229,
+      "learning_rate": 0.0001,
+      "loss": 0.8757,
+      "step": 2618
+    },
+    {
+      "epoch": 0.22430626927029804,
+      "grad_norm": 0.5686971545219421,
+      "learning_rate": 0.0001,
+      "loss": 1.1566,
+      "step": 2619
+    },
+    {
+      "epoch": 0.22439191503939707,
+      "grad_norm": 0.668198823928833,
+      "learning_rate": 0.0001,
+      "loss": 1.0982,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22447756080849607,
+      "grad_norm": 0.8301640152931213,
+      "learning_rate": 0.0001,
+      "loss": 1.0791,
+      "step": 2621
+    },
+    {
+      "epoch": 0.22456320657759507,
+      "grad_norm": 0.8333268165588379,
+      "learning_rate": 0.0001,
+      "loss": 0.9327,
+      "step": 2622
+    },
+    {
+      "epoch": 0.22464885234669407,
+      "grad_norm": 0.572799563407898,
+      "learning_rate": 0.0001,
+      "loss": 1.0302,
+      "step": 2623
+    },
+    {
+      "epoch": 0.22473449811579307,
+      "grad_norm": 0.571378231048584,
+      "learning_rate": 0.0001,
+      "loss": 1.1739,
+      "step": 2624
+    },
+    {
+      "epoch": 0.22482014388489208,
+      "grad_norm": 0.877548336982727,
+      "learning_rate": 0.0001,
+      "loss": 1.2397,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2249057896539911,
+      "grad_norm": 0.7569597363471985,
+      "learning_rate": 0.0001,
+      "loss": 1.2063,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2249914354230901,
+      "grad_norm": 0.27462509274482727,
+      "learning_rate": 0.0001,
+      "loss": 1.2009,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2250770811921891,
+      "grad_norm": 0.7196272015571594,
+      "learning_rate": 0.0001,
+      "loss": 0.9072,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2251627269612881,
+      "grad_norm": 0.7621464729309082,
+      "learning_rate": 0.0001,
+      "loss": 1.0089,
+      "step": 2629
+    },
+    {
+      "epoch": 0.22524837273038711,
+      "grad_norm": 0.638896644115448,
+      "learning_rate": 0.0001,
+      "loss": 1.2335,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22533401849948612,
+      "grad_norm": 0.692456066608429,
+      "learning_rate": 0.0001,
+      "loss": 1.2598,
+      "step": 2631
+    },
+    {
+      "epoch": 0.22541966426858512,
+      "grad_norm": 1.2654119729995728,
+      "learning_rate": 0.0001,
+      "loss": 0.8212,
+      "step": 2632
+    },
+    {
+      "epoch": 0.22550531003768415,
+      "grad_norm": 0.8487153649330139,
+      "learning_rate": 0.0001,
+      "loss": 0.8015,
+      "step": 2633
+    },
+    {
+      "epoch": 0.22559095580678315,
+      "grad_norm": 0.9032750129699707,
+      "learning_rate": 0.0001,
+      "loss": 1.0409,
+      "step": 2634
+    },
+    {
+      "epoch": 0.22567660157588215,
+      "grad_norm": 0.4845799505710602,
+      "learning_rate": 0.0001,
+      "loss": 0.7292,
+      "step": 2635
+    },
+    {
+      "epoch": 0.22576224734498115,
+      "grad_norm": 0.6052261590957642,
+      "learning_rate": 0.0001,
+      "loss": 1.0521,
+      "step": 2636
+    },
+    {
+      "epoch": 0.22584789311408016,
+      "grad_norm": 0.5403116941452026,
+      "learning_rate": 0.0001,
+      "loss": 1.0938,
+      "step": 2637
+    },
+    {
+      "epoch": 0.22593353888317916,
+      "grad_norm": 0.7104957103729248,
+      "learning_rate": 0.0001,
+      "loss": 0.5555,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2260191846522782,
+      "grad_norm": 0.6124822497367859,
+      "learning_rate": 0.0001,
+      "loss": 0.828,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2261048304213772,
+      "grad_norm": 0.9255884289741516,
+      "learning_rate": 0.0001,
+      "loss": 0.8604,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2261904761904762,
+      "grad_norm": 0.7312043309211731,
+      "learning_rate": 0.0001,
+      "loss": 1.168,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2262761219595752,
+      "grad_norm": 0.9141709804534912,
+      "learning_rate": 0.0001,
+      "loss": 1.0157,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2263617677286742,
+      "grad_norm": 0.4937070608139038,
+      "learning_rate": 0.0001,
+      "loss": 1.0682,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2264474134977732,
+      "grad_norm": 0.6677408814430237,
+      "learning_rate": 0.0001,
+      "loss": 0.7233,
+      "step": 2644
+    },
+    {
+      "epoch": 0.22653305926687223,
+      "grad_norm": 0.6801018714904785,
+      "learning_rate": 0.0001,
+      "loss": 0.4155,
+      "step": 2645
+    },
+    {
+      "epoch": 0.22661870503597123,
+      "grad_norm": 0.8754667043685913,
+      "learning_rate": 0.0001,
+      "loss": 0.4989,
+      "step": 2646
+    },
+    {
+      "epoch": 0.22670435080507023,
+      "grad_norm": 1.2783877849578857,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 2647
+    },
+    {
+      "epoch": 0.22678999657416923,
+      "grad_norm": 0.5938459634780884,
+      "learning_rate": 0.0001,
+      "loss": 0.8941,
+      "step": 2648
+    },
+    {
+      "epoch": 0.22687564234326824,
+      "grad_norm": 0.5961527824401855,
+      "learning_rate": 0.0001,
+      "loss": 0.6873,
+      "step": 2649
+    },
+    {
+      "epoch": 0.22696128811236724,
+      "grad_norm": 0.6821664571762085,
+      "learning_rate": 0.0001,
+      "loss": 0.6424,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22704693388146627,
+      "grad_norm": 0.5953925251960754,
+      "learning_rate": 0.0001,
+      "loss": 0.4964,
+      "step": 2651
+    },
+    {
+      "epoch": 0.22713257965056527,
+      "grad_norm": 0.6600239276885986,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 2652
+    },
+    {
+      "epoch": 0.22721822541966427,
+      "grad_norm": 0.8155249357223511,
+      "learning_rate": 0.0001,
+      "loss": 1.1466,
+      "step": 2653
+    },
+    {
+      "epoch": 0.22730387118876327,
+      "grad_norm": 0.9563654065132141,
+      "learning_rate": 0.0001,
+      "loss": 1.1244,
+      "step": 2654
+    },
+    {
+      "epoch": 0.22738951695786228,
+      "grad_norm": 0.9187497496604919,
+      "learning_rate": 0.0001,
+      "loss": 1.0576,
+      "step": 2655
+    },
+    {
+      "epoch": 0.22747516272696128,
+      "grad_norm": 0.7556740641593933,
+      "learning_rate": 0.0001,
+      "loss": 0.845,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2275608084960603,
+      "grad_norm": 0.6652969717979431,
+      "learning_rate": 0.0001,
+      "loss": 0.8127,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2276464542651593,
+      "grad_norm": 0.45529448986053467,
+      "learning_rate": 0.0001,
+      "loss": 0.3085,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2277321000342583,
+      "grad_norm": 0.737090528011322,
+      "learning_rate": 0.0001,
+      "loss": 1.0906,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2278177458033573,
+      "grad_norm": 0.9976766705513,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22790339157245632,
+      "grad_norm": 1.0442959070205688,
+      "learning_rate": 0.0001,
+      "loss": 1.5064,
+      "step": 2661
+    },
+    {
+      "epoch": 0.22798903734155532,
+      "grad_norm": 0.4908403158187866,
+      "learning_rate": 0.0001,
+      "loss": 0.7018,
+      "step": 2662
+    },
+    {
+      "epoch": 0.22807468311065435,
+      "grad_norm": 0.8906158208847046,
+      "learning_rate": 0.0001,
+      "loss": 0.9819,
+      "step": 2663
+    },
+    {
+      "epoch": 0.22816032887975335,
+      "grad_norm": 0.40375861525535583,
+      "learning_rate": 0.0001,
+      "loss": 0.426,
+      "step": 2664
+    },
+    {
+      "epoch": 0.22824597464885235,
+      "grad_norm": 0.4185880720615387,
+      "learning_rate": 0.0001,
+      "loss": 0.6556,
+      "step": 2665
+    },
+    {
+      "epoch": 0.22833162041795135,
+      "grad_norm": 0.8400716185569763,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 2666
+    },
+    {
+      "epoch": 0.22841726618705036,
+      "grad_norm": 0.7044552564620972,
+      "learning_rate": 0.0001,
+      "loss": 0.9224,
+      "step": 2667
+    },
+    {
+      "epoch": 0.22850291195614936,
+      "grad_norm": 0.4208148717880249,
+      "learning_rate": 0.0001,
+      "loss": 0.6504,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22858855772524836,
+      "grad_norm": 0.54139244556427,
+      "learning_rate": 0.0001,
+      "loss": 0.9313,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2286742034943474,
+      "grad_norm": 0.6799348592758179,
+      "learning_rate": 0.0001,
+      "loss": 0.6912,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2287598492634464,
+      "grad_norm": 0.773669421672821,
+      "learning_rate": 0.0001,
+      "loss": 0.6054,
+      "step": 2671
+    },
+    {
+      "epoch": 0.2288454950325454,
+      "grad_norm": 0.979598879814148,
+      "learning_rate": 0.0001,
+      "loss": 1.145,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2289311408016444,
+      "grad_norm": 0.6313305497169495,
+      "learning_rate": 0.0001,
+      "loss": 1.0703,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2290167865707434,
+      "grad_norm": 0.43706074357032776,
+      "learning_rate": 0.0001,
+      "loss": 1.0869,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2291024323398424,
+      "grad_norm": 0.45411285758018494,
+      "learning_rate": 0.0001,
+      "loss": 0.479,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22918807810894143,
+      "grad_norm": 0.7789355516433716,
+      "learning_rate": 0.0001,
+      "loss": 0.7059,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22927372387804043,
+      "grad_norm": 0.6519642472267151,
+      "learning_rate": 0.0001,
+      "loss": 0.7884,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22935936964713943,
+      "grad_norm": 0.47284701466560364,
+      "learning_rate": 0.0001,
+      "loss": 0.8413,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22944501541623843,
+      "grad_norm": 0.4226182699203491,
+      "learning_rate": 0.0001,
+      "loss": 0.7978,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22953066118533744,
+      "grad_norm": 0.28141483664512634,
+      "learning_rate": 0.0001,
+      "loss": 0.4473,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22961630695443644,
+      "grad_norm": 0.31214138865470886,
+      "learning_rate": 0.0001,
+      "loss": 0.6415,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22970195272353547,
+      "grad_norm": 0.6970632076263428,
+      "learning_rate": 0.0001,
+      "loss": 1.1906,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22978759849263447,
+      "grad_norm": 1.011612892150879,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22987324426173347,
+      "grad_norm": 0.7250677943229675,
+      "learning_rate": 0.0001,
+      "loss": 0.7445,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22995889003083247,
+      "grad_norm": 0.24802090227603912,
+      "learning_rate": 0.0001,
+      "loss": 0.621,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23004453579993148,
+      "grad_norm": 0.5658782124519348,
+      "learning_rate": 0.0001,
+      "loss": 0.6877,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23013018156903048,
+      "grad_norm": 0.7631900906562805,
+      "learning_rate": 0.0001,
+      "loss": 0.8229,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2302158273381295,
+      "grad_norm": 0.868381679058075,
+      "learning_rate": 0.0001,
+      "loss": 0.9671,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2303014731072285,
+      "grad_norm": 0.3826766610145569,
+      "learning_rate": 0.0001,
+      "loss": 0.5854,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2303871188763275,
+      "grad_norm": 0.7667944431304932,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23047276464542651,
+      "grad_norm": 0.9081211090087891,
+      "learning_rate": 0.0001,
+      "loss": 1.2636,
+      "step": 2691
+    },
+    {
+      "epoch": 0.23055841041452552,
+      "grad_norm": 1.0207940340042114,
+      "learning_rate": 0.0001,
+      "loss": 1.223,
+      "step": 2692
+    },
+    {
+      "epoch": 0.23064405618362452,
+      "grad_norm": 0.9874982237815857,
+      "learning_rate": 0.0001,
+      "loss": 1.2121,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23072970195272355,
+      "grad_norm": 0.5293700098991394,
+      "learning_rate": 0.0001,
+      "loss": 1.078,
+      "step": 2694
+    },
+    {
+      "epoch": 0.23081534772182255,
+      "grad_norm": 0.7850746512413025,
+      "learning_rate": 0.0001,
+      "loss": 1.1085,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23090099349092155,
+      "grad_norm": 0.7266283631324768,
+      "learning_rate": 0.0001,
+      "loss": 1.1101,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23098663926002055,
+      "grad_norm": 0.653465211391449,
+      "learning_rate": 0.0001,
+      "loss": 0.6497,
+      "step": 2697
+    },
+    {
+      "epoch": 0.23107228502911956,
+      "grad_norm": 0.5573545098304749,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 2698
+    },
+    {
+      "epoch": 0.23115793079821856,
+      "grad_norm": 0.8605012893676758,
+      "learning_rate": 0.0001,
+      "loss": 0.9595,
+      "step": 2699
+    },
+    {
+      "epoch": 0.2312435765673176,
+      "grad_norm": 0.8325172662734985,
+      "learning_rate": 0.0001,
+      "loss": 0.6206,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2313292223364166,
+      "grad_norm": 0.8544961810112,
+      "learning_rate": 0.0001,
+      "loss": 1.0515,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2314148681055156,
+      "grad_norm": 0.7333258390426636,
+      "learning_rate": 0.0001,
+      "loss": 0.9879,
+      "step": 2702
+    },
+    {
+      "epoch": 0.2315005138746146,
+      "grad_norm": 0.9569529294967651,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2315861596437136,
+      "grad_norm": 0.8538923263549805,
+      "learning_rate": 0.0001,
+      "loss": 1.2036,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2316718054128126,
+      "grad_norm": 0.7008359432220459,
+      "learning_rate": 0.0001,
+      "loss": 0.951,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2317574511819116,
+      "grad_norm": 0.532710075378418,
+      "learning_rate": 0.0001,
+      "loss": 0.9814,
+      "step": 2706
+    },
+    {
+      "epoch": 0.23184309695101063,
+      "grad_norm": 1.3219298124313354,
+      "learning_rate": 0.0001,
+      "loss": 1.3012,
+      "step": 2707
+    },
+    {
+      "epoch": 0.23192874272010963,
+      "grad_norm": 0.9309573173522949,
+      "learning_rate": 0.0001,
+      "loss": 0.9981,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23201438848920863,
+      "grad_norm": 0.5840291380882263,
+      "learning_rate": 0.0001,
+      "loss": 0.8207,
+      "step": 2709
+    },
+    {
+      "epoch": 0.23210003425830764,
+      "grad_norm": 0.5365191698074341,
+      "learning_rate": 0.0001,
+      "loss": 0.9726,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23218568002740664,
+      "grad_norm": 0.6442887783050537,
+      "learning_rate": 0.0001,
+      "loss": 0.8382,
+      "step": 2711
+    },
+    {
+      "epoch": 0.23227132579650564,
+      "grad_norm": 0.49317651987075806,
+      "learning_rate": 0.0001,
+      "loss": 0.6553,
+      "step": 2712
+    },
+    {
+      "epoch": 0.23235697156560467,
+      "grad_norm": 0.7540099620819092,
+      "learning_rate": 0.0001,
+      "loss": 0.504,
+      "step": 2713
+    },
+    {
+      "epoch": 0.23244261733470367,
+      "grad_norm": 0.8942608833312988,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 2714
+    },
+    {
+      "epoch": 0.23252826310380267,
+      "grad_norm": 1.0262843370437622,
+      "learning_rate": 0.0001,
+      "loss": 1.3929,
+      "step": 2715
+    },
+    {
+      "epoch": 0.23261390887290168,
+      "grad_norm": 1.7905356884002686,
+      "learning_rate": 0.0001,
+      "loss": 0.8651,
+      "step": 2716
+    },
+    {
+      "epoch": 0.23269955464200068,
+      "grad_norm": 0.8882350921630859,
+      "learning_rate": 0.0001,
+      "loss": 1.088,
+      "step": 2717
+    },
+    {
+      "epoch": 0.23278520041109968,
+      "grad_norm": 0.6196362972259521,
+      "learning_rate": 0.0001,
+      "loss": 0.4814,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2328708461801987,
+      "grad_norm": 0.8286392092704773,
+      "learning_rate": 0.0001,
+      "loss": 1.1855,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2329564919492977,
+      "grad_norm": 0.48828884959220886,
+      "learning_rate": 0.0001,
+      "loss": 1.1375,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2330421377183967,
+      "grad_norm": 0.9269441366195679,
+      "learning_rate": 0.0001,
+      "loss": 0.8472,
+      "step": 2721
+    },
+    {
+      "epoch": 0.23312778348749572,
+      "grad_norm": 0.45634910464286804,
+      "learning_rate": 0.0001,
+      "loss": 0.4682,
+      "step": 2722
+    },
+    {
+      "epoch": 0.23321342925659472,
+      "grad_norm": 0.8153235912322998,
+      "learning_rate": 0.0001,
+      "loss": 1.1874,
+      "step": 2723
+    },
+    {
+      "epoch": 0.23329907502569372,
+      "grad_norm": 0.46905726194381714,
+      "learning_rate": 0.0001,
+      "loss": 0.4357,
+      "step": 2724
+    },
+    {
+      "epoch": 0.23338472079479275,
+      "grad_norm": 0.6939133405685425,
+      "learning_rate": 0.0001,
+      "loss": 0.7458,
+      "step": 2725
+    },
+    {
+      "epoch": 0.23347036656389175,
+      "grad_norm": 0.8943004608154297,
+      "learning_rate": 0.0001,
+      "loss": 1.8336,
+      "step": 2726
+    },
+    {
+      "epoch": 0.23355601233299075,
+      "grad_norm": 0.9586155414581299,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 2727
+    },
+    {
+      "epoch": 0.23364165810208976,
+      "grad_norm": 0.8214272260665894,
+      "learning_rate": 0.0001,
+      "loss": 0.5744,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23372730387118876,
+      "grad_norm": 0.5177655816078186,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 2729
+    },
+    {
+      "epoch": 0.23381294964028776,
+      "grad_norm": 0.7192922830581665,
+      "learning_rate": 0.0001,
+      "loss": 0.5954,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2338985954093868,
+      "grad_norm": 0.5382031798362732,
+      "learning_rate": 0.0001,
+      "loss": 0.8586,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2339842411784858,
+      "grad_norm": 0.6629506349563599,
+      "learning_rate": 0.0001,
+      "loss": 1.0518,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2340698869475848,
+      "grad_norm": 0.5748872756958008,
+      "learning_rate": 0.0001,
+      "loss": 1.0173,
+      "step": 2733
+    },
+    {
+      "epoch": 0.2341555327166838,
+      "grad_norm": 0.9187731146812439,
+      "learning_rate": 0.0001,
+      "loss": 1.0562,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2342411784857828,
+      "grad_norm": 0.6577361226081848,
+      "learning_rate": 0.0001,
+      "loss": 0.5015,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2343268242548818,
+      "grad_norm": 0.5844483375549316,
+      "learning_rate": 0.0001,
+      "loss": 1.0323,
+      "step": 2736
+    },
+    {
+      "epoch": 0.23441247002398083,
+      "grad_norm": 0.4004537761211395,
+      "learning_rate": 0.0001,
+      "loss": 0.8783,
+      "step": 2737
+    },
+    {
+      "epoch": 0.23449811579307983,
+      "grad_norm": 0.9027756452560425,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 2738
+    },
+    {
+      "epoch": 0.23458376156217883,
+      "grad_norm": 0.6122171878814697,
+      "learning_rate": 0.0001,
+      "loss": 0.8816,
+      "step": 2739
+    },
+    {
+      "epoch": 0.23466940733127784,
+      "grad_norm": 0.6435300707817078,
+      "learning_rate": 0.0001,
+      "loss": 0.7326,
+      "step": 2740
+    },
+    {
+      "epoch": 0.23475505310037684,
+      "grad_norm": 0.6443653106689453,
+      "learning_rate": 0.0001,
+      "loss": 0.9909,
+      "step": 2741
+    },
+    {
+      "epoch": 0.23484069886947584,
+      "grad_norm": 0.4817613959312439,
+      "learning_rate": 0.0001,
+      "loss": 0.8639,
+      "step": 2742
+    },
+    {
+      "epoch": 0.23492634463857484,
+      "grad_norm": 0.3998481333255768,
+      "learning_rate": 0.0001,
+      "loss": 0.5512,
+      "step": 2743
+    },
+    {
+      "epoch": 0.23501199040767387,
+      "grad_norm": 0.5061914920806885,
+      "learning_rate": 0.0001,
+      "loss": 1.361,
+      "step": 2744
+    },
+    {
+      "epoch": 0.23509763617677287,
+      "grad_norm": 0.5775543451309204,
+      "learning_rate": 0.0001,
+      "loss": 1.0369,
+      "step": 2745
+    },
+    {
+      "epoch": 0.23518328194587187,
+      "grad_norm": 0.7374217510223389,
+      "learning_rate": 0.0001,
+      "loss": 0.6374,
+      "step": 2746
+    },
+    {
+      "epoch": 0.23526892771497088,
+      "grad_norm": 0.8332074284553528,
+      "learning_rate": 0.0001,
+      "loss": 0.8021,
+      "step": 2747
+    },
+    {
+      "epoch": 0.23535457348406988,
+      "grad_norm": 0.8420785069465637,
+      "learning_rate": 0.0001,
+      "loss": 0.961,
+      "step": 2748
+    },
+    {
+      "epoch": 0.23544021925316888,
+      "grad_norm": 0.4702248275279999,
+      "learning_rate": 0.0001,
+      "loss": 0.9049,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2355258650222679,
+      "grad_norm": 0.3176955580711365,
+      "learning_rate": 0.0001,
+      "loss": 0.4711,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2356115107913669,
+      "grad_norm": 1.0283794403076172,
+      "learning_rate": 0.0001,
+      "loss": 1.0098,
+      "step": 2751
+    },
+    {
+      "epoch": 0.23569715656046591,
+      "grad_norm": 0.5972887873649597,
+      "learning_rate": 0.0001,
+      "loss": 1.1321,
+      "step": 2752
+    },
+    {
+      "epoch": 0.23578280232956492,
+      "grad_norm": 0.9662367105484009,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 2753
+    },
+    {
+      "epoch": 0.23586844809866392,
+      "grad_norm": 0.8384140729904175,
+      "learning_rate": 0.0001,
+      "loss": 0.8279,
+      "step": 2754
+    },
+    {
+      "epoch": 0.23595409386776292,
+      "grad_norm": 0.9710468053817749,
+      "learning_rate": 0.0001,
+      "loss": 1.7646,
+      "step": 2755
+    },
+    {
+      "epoch": 0.23603973963686195,
+      "grad_norm": 0.7472357749938965,
+      "learning_rate": 0.0001,
+      "loss": 0.942,
+      "step": 2756
+    },
+    {
+      "epoch": 0.23612538540596095,
+      "grad_norm": 0.5659598708152771,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 2757
+    },
+    {
+      "epoch": 0.23621103117505995,
+      "grad_norm": 0.4873392581939697,
+      "learning_rate": 0.0001,
+      "loss": 1.214,
+      "step": 2758
+    },
+    {
+      "epoch": 0.23629667694415896,
+      "grad_norm": 0.8677089214324951,
+      "learning_rate": 0.0001,
+      "loss": 0.9162,
+      "step": 2759
+    },
+    {
+      "epoch": 0.23638232271325796,
+      "grad_norm": 0.5613845586776733,
+      "learning_rate": 0.0001,
+      "loss": 0.8416,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23646796848235696,
+      "grad_norm": 0.8352916240692139,
+      "learning_rate": 0.0001,
+      "loss": 0.9595,
+      "step": 2761
+    },
+    {
+      "epoch": 0.236553614251456,
+      "grad_norm": 0.6560889482498169,
+      "learning_rate": 0.0001,
+      "loss": 0.8453,
+      "step": 2762
+    },
+    {
+      "epoch": 0.236639260020555,
+      "grad_norm": 0.6653005480766296,
+      "learning_rate": 0.0001,
+      "loss": 0.7313,
+      "step": 2763
+    },
+    {
+      "epoch": 0.236724905789654,
+      "grad_norm": 1.3541948795318604,
+      "learning_rate": 0.0001,
+      "loss": 1.8089,
+      "step": 2764
+    },
+    {
+      "epoch": 0.236810551558753,
+      "grad_norm": 0.5895563960075378,
+      "learning_rate": 0.0001,
+      "loss": 1.0158,
+      "step": 2765
+    },
+    {
+      "epoch": 0.236896197327852,
+      "grad_norm": 0.4853476583957672,
+      "learning_rate": 0.0001,
+      "loss": 1.0371,
+      "step": 2766
+    },
+    {
+      "epoch": 0.236981843096951,
+      "grad_norm": 0.6107394099235535,
+      "learning_rate": 0.0001,
+      "loss": 0.6134,
+      "step": 2767
+    },
+    {
+      "epoch": 0.23706748886605003,
+      "grad_norm": 0.7674142718315125,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 2768
+    },
+    {
+      "epoch": 0.23715313463514903,
+      "grad_norm": 1.0179474353790283,
+      "learning_rate": 0.0001,
+      "loss": 1.0402,
+      "step": 2769
+    },
+    {
+      "epoch": 0.23723878040424803,
+      "grad_norm": 0.5794686675071716,
+      "learning_rate": 0.0001,
+      "loss": 0.5437,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23732442617334704,
+      "grad_norm": 0.37972283363342285,
+      "learning_rate": 0.0001,
+      "loss": 0.2221,
+      "step": 2771
+    },
+    {
+      "epoch": 0.23741007194244604,
+      "grad_norm": 0.7641513347625732,
+      "learning_rate": 0.0001,
+      "loss": 0.3404,
+      "step": 2772
+    },
+    {
+      "epoch": 0.23749571771154504,
+      "grad_norm": 0.7958416938781738,
+      "learning_rate": 0.0001,
+      "loss": 0.8183,
+      "step": 2773
+    },
+    {
+      "epoch": 0.23758136348064407,
+      "grad_norm": 1.2704416513442993,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 2774
+    },
+    {
+      "epoch": 0.23766700924974307,
+      "grad_norm": 0.592178463935852,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 2775
+    },
+    {
+      "epoch": 0.23775265501884207,
+      "grad_norm": 0.8446080684661865,
+      "learning_rate": 0.0001,
+      "loss": 0.9771,
+      "step": 2776
+    },
+    {
+      "epoch": 0.23783830078794108,
+      "grad_norm": 0.6882047653198242,
+      "learning_rate": 0.0001,
+      "loss": 0.8667,
+      "step": 2777
+    },
+    {
+      "epoch": 0.23792394655704008,
+      "grad_norm": 0.6356030106544495,
+      "learning_rate": 0.0001,
+      "loss": 1.1589,
+      "step": 2778
+    },
+    {
+      "epoch": 0.23800959232613908,
+      "grad_norm": 0.8521285653114319,
+      "learning_rate": 0.0001,
+      "loss": 1.0808,
+      "step": 2779
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 1.072920799255371,
+      "learning_rate": 0.0001,
+      "loss": 1.2562,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2381808838643371,
+      "grad_norm": 0.43651482462882996,
+      "learning_rate": 0.0001,
+      "loss": 0.8007,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2382665296334361,
+      "grad_norm": 0.5570762157440186,
+      "learning_rate": 0.0001,
+      "loss": 1.1787,
+      "step": 2782
+    },
+    {
+      "epoch": 0.23835217540253512,
+      "grad_norm": 0.6365900635719299,
+      "learning_rate": 0.0001,
+      "loss": 1.1725,
+      "step": 2783
+    },
+    {
+      "epoch": 0.23843782117163412,
+      "grad_norm": 0.5156111121177673,
+      "learning_rate": 0.0001,
+      "loss": 1.0638,
+      "step": 2784
+    },
+    {
+      "epoch": 0.23852346694073312,
+      "grad_norm": 0.7581338286399841,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 2785
+    },
+    {
+      "epoch": 0.23860911270983212,
+      "grad_norm": 0.5942611694335938,
+      "learning_rate": 0.0001,
+      "loss": 0.4515,
+      "step": 2786
+    },
+    {
+      "epoch": 0.23869475847893115,
+      "grad_norm": 0.17524956166744232,
+      "learning_rate": 0.0001,
+      "loss": 0.754,
+      "step": 2787
+    },
+    {
+      "epoch": 0.23878040424803015,
+      "grad_norm": 0.49114659428596497,
+      "learning_rate": 0.0001,
+      "loss": 1.1731,
+      "step": 2788
+    },
+    {
+      "epoch": 0.23886605001712916,
+      "grad_norm": 0.5908578634262085,
+      "learning_rate": 0.0001,
+      "loss": 0.9968,
+      "step": 2789
+    },
+    {
+      "epoch": 0.23895169578622816,
+      "grad_norm": 0.6716576814651489,
+      "learning_rate": 0.0001,
+      "loss": 1.3344,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23903734155532716,
+      "grad_norm": 0.9739160537719727,
+      "learning_rate": 0.0001,
+      "loss": 0.9555,
+      "step": 2791
+    },
+    {
+      "epoch": 0.23912298732442616,
+      "grad_norm": 0.39779043197631836,
+      "learning_rate": 0.0001,
+      "loss": 0.7217,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2392086330935252,
+      "grad_norm": 0.13220281898975372,
+      "learning_rate": 0.0001,
+      "loss": 0.3864,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2392942788626242,
+      "grad_norm": 1.2748634815216064,
+      "learning_rate": 0.0001,
+      "loss": 1.0044,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2393799246317232,
+      "grad_norm": 0.5180392861366272,
+      "learning_rate": 0.0001,
+      "loss": 0.8591,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2394655704008222,
+      "grad_norm": 1.3750905990600586,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2395512161699212,
+      "grad_norm": 3.02921986579895,
+      "learning_rate": 0.0001,
+      "loss": 1.0157,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2396368619390202,
+      "grad_norm": 13.809231758117676,
+      "learning_rate": 0.0001,
+      "loss": 0.6833,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23972250770811923,
+      "grad_norm": 2.707033395767212,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23980815347721823,
+      "grad_norm": 2.7679731845855713,
+      "learning_rate": 0.0001,
+      "loss": 0.9808,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23989379924631724,
+      "grad_norm": 1.2237247228622437,
+      "learning_rate": 0.0001,
+      "loss": 1.3277,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23997944501541624,
+      "grad_norm": 0.5886673927307129,
+      "learning_rate": 0.0001,
+      "loss": 0.8196,
+      "step": 2802
+    },
+    {
+      "epoch": 0.24006509078451524,
+      "grad_norm": 0.9756793975830078,
+      "learning_rate": 0.0001,
+      "loss": 0.8888,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24015073655361424,
+      "grad_norm": 0.7976711988449097,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24023638232271327,
+      "grad_norm": 0.8156786561012268,
+      "learning_rate": 0.0001,
+      "loss": 0.9962,
+      "step": 2805
+    },
+    {
+      "epoch": 0.24032202809181227,
+      "grad_norm": 0.7472038865089417,
+      "learning_rate": 0.0001,
+      "loss": 1.0692,
+      "step": 2806
+    },
+    {
+      "epoch": 0.24040767386091128,
+      "grad_norm": 1.0213667154312134,
+      "learning_rate": 0.0001,
+      "loss": 0.9834,
+      "step": 2807
+    },
+    {
+      "epoch": 0.24049331963001028,
+      "grad_norm": 0.48621413111686707,
+      "learning_rate": 0.0001,
+      "loss": 0.6346,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24057896539910928,
+      "grad_norm": 0.41296887397766113,
+      "learning_rate": 0.0001,
+      "loss": 0.8169,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24066461116820828,
+      "grad_norm": 0.7422424554824829,
+      "learning_rate": 0.0001,
+      "loss": 0.9861,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24075025693730728,
+      "grad_norm": 0.4903729259967804,
+      "learning_rate": 0.0001,
+      "loss": 1.7416,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2408359027064063,
+      "grad_norm": 0.40224406123161316,
+      "learning_rate": 0.0001,
+      "loss": 0.6037,
+      "step": 2812
+    },
+    {
+      "epoch": 0.24092154847550531,
+      "grad_norm": 0.4680914878845215,
+      "learning_rate": 0.0001,
+      "loss": 0.9425,
+      "step": 2813
+    },
+    {
+      "epoch": 0.24100719424460432,
+      "grad_norm": 0.7645402550697327,
+      "learning_rate": 0.0001,
+      "loss": 0.9688,
+      "step": 2814
+    },
+    {
+      "epoch": 0.24109284001370332,
+      "grad_norm": 0.30069154500961304,
+      "learning_rate": 0.0001,
+      "loss": 0.3833,
+      "step": 2815
+    },
+    {
+      "epoch": 0.24117848578280232,
+      "grad_norm": 0.5857062339782715,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24126413155190132,
+      "grad_norm": 0.7462190985679626,
+      "learning_rate": 0.0001,
+      "loss": 1.2361,
+      "step": 2817
+    },
+    {
+      "epoch": 0.24134977732100035,
+      "grad_norm": 0.9654137492179871,
+      "learning_rate": 0.0001,
+      "loss": 1.2059,
+      "step": 2818
+    },
+    {
+      "epoch": 0.24143542309009935,
+      "grad_norm": 0.5709378719329834,
+      "learning_rate": 0.0001,
+      "loss": 0.7123,
+      "step": 2819
+    },
+    {
+      "epoch": 0.24152106885919836,
+      "grad_norm": 0.6208949685096741,
+      "learning_rate": 0.0001,
+      "loss": 0.6668,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24160671462829736,
+      "grad_norm": 0.6341590881347656,
+      "learning_rate": 0.0001,
+      "loss": 0.6939,
+      "step": 2821
+    },
+    {
+      "epoch": 0.24169236039739636,
+      "grad_norm": 0.4755454957485199,
+      "learning_rate": 0.0001,
+      "loss": 1.1197,
+      "step": 2822
+    },
+    {
+      "epoch": 0.24177800616649536,
+      "grad_norm": 0.49806591868400574,
+      "learning_rate": 0.0001,
+      "loss": 1.0205,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2418636519355944,
+      "grad_norm": 0.587247371673584,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2419492977046934,
+      "grad_norm": 1.3278383016586304,
+      "learning_rate": 0.0001,
+      "loss": 0.642,
+      "step": 2825
+    },
+    {
+      "epoch": 0.2420349434737924,
+      "grad_norm": 0.6107705235481262,
+      "learning_rate": 0.0001,
+      "loss": 1.1075,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2421205892428914,
+      "grad_norm": 0.4897710382938385,
+      "learning_rate": 0.0001,
+      "loss": 1.1454,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2422062350119904,
+      "grad_norm": 0.6415708065032959,
+      "learning_rate": 0.0001,
+      "loss": 0.8358,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2422918807810894,
+      "grad_norm": 0.7723029851913452,
+      "learning_rate": 0.0001,
+      "loss": 1.3029,
+      "step": 2829
+    },
+    {
+      "epoch": 0.24237752655018843,
+      "grad_norm": 0.48541566729545593,
+      "learning_rate": 0.0001,
+      "loss": 1.0317,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24246317231928743,
+      "grad_norm": 0.8341834545135498,
+      "learning_rate": 0.0001,
+      "loss": 0.9384,
+      "step": 2831
+    },
+    {
+      "epoch": 0.24254881808838644,
+      "grad_norm": 1.1768743991851807,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24263446385748544,
+      "grad_norm": 0.821630597114563,
+      "learning_rate": 0.0001,
+      "loss": 0.662,
+      "step": 2833
+    },
+    {
+      "epoch": 0.24272010962658444,
+      "grad_norm": 0.766329824924469,
+      "learning_rate": 0.0001,
+      "loss": 0.9483,
+      "step": 2834
+    },
+    {
+      "epoch": 0.24280575539568344,
+      "grad_norm": 1.0224735736846924,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 2835
+    },
+    {
+      "epoch": 0.24289140116478247,
+      "grad_norm": 0.8299707770347595,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24297704693388147,
+      "grad_norm": 1.0668805837631226,
+      "learning_rate": 0.0001,
+      "loss": 1.0009,
+      "step": 2837
+    },
+    {
+      "epoch": 0.24306269270298048,
+      "grad_norm": 0.6121595501899719,
+      "learning_rate": 0.0001,
+      "loss": 1.1873,
+      "step": 2838
+    },
+    {
+      "epoch": 0.24314833847207948,
+      "grad_norm": 0.5923357605934143,
+      "learning_rate": 0.0001,
+      "loss": 1.0326,
+      "step": 2839
+    },
+    {
+      "epoch": 0.24323398424117848,
+      "grad_norm": 0.6868860721588135,
+      "learning_rate": 0.0001,
+      "loss": 0.584,
+      "step": 2840
+    },
+    {
+      "epoch": 0.24331963001027748,
+      "grad_norm": 0.7867839932441711,
+      "learning_rate": 0.0001,
+      "loss": 1.2102,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2434052757793765,
+      "grad_norm": 0.763146162033081,
+      "learning_rate": 0.0001,
+      "loss": 1.1693,
+      "step": 2842
+    },
+    {
+      "epoch": 0.24349092154847551,
+      "grad_norm": 0.560850203037262,
+      "learning_rate": 0.0001,
+      "loss": 0.3457,
+      "step": 2843
+    },
+    {
+      "epoch": 0.24357656731757452,
+      "grad_norm": 1.5615473985671997,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 2844
+    },
+    {
+      "epoch": 0.24366221308667352,
+      "grad_norm": 0.45998716354370117,
+      "learning_rate": 0.0001,
+      "loss": 0.8175,
+      "step": 2845
+    },
+    {
+      "epoch": 0.24374785885577252,
+      "grad_norm": 0.7715473175048828,
+      "learning_rate": 0.0001,
+      "loss": 0.8226,
+      "step": 2846
+    },
+    {
+      "epoch": 0.24383350462487152,
+      "grad_norm": 0.7539876699447632,
+      "learning_rate": 0.0001,
+      "loss": 0.9563,
+      "step": 2847
+    },
+    {
+      "epoch": 0.24391915039397052,
+      "grad_norm": 0.5398236513137817,
+      "learning_rate": 0.0001,
+      "loss": 1.0649,
+      "step": 2848
+    },
+    {
+      "epoch": 0.24400479616306955,
+      "grad_norm": 0.7279013991355896,
+      "learning_rate": 0.0001,
+      "loss": 1.2361,
+      "step": 2849
+    },
+    {
+      "epoch": 0.24409044193216856,
+      "grad_norm": 0.48720577359199524,
+      "learning_rate": 0.0001,
+      "loss": 0.4703,
+      "step": 2850
+    },
+    {
+      "epoch": 0.24417608770126756,
+      "grad_norm": 0.46012771129608154,
+      "learning_rate": 0.0001,
+      "loss": 0.6318,
+      "step": 2851
+    },
+    {
+      "epoch": 0.24426173347036656,
+      "grad_norm": 0.9932137727737427,
+      "learning_rate": 0.0001,
+      "loss": 1.2852,
+      "step": 2852
+    },
+    {
+      "epoch": 0.24434737923946556,
+      "grad_norm": 0.872929036617279,
+      "learning_rate": 0.0001,
+      "loss": 1.0246,
+      "step": 2853
+    },
+    {
+      "epoch": 0.24443302500856456,
+      "grad_norm": 0.7130566835403442,
+      "learning_rate": 0.0001,
+      "loss": 0.7617,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2445186707776636,
+      "grad_norm": 0.7201546430587769,
+      "learning_rate": 0.0001,
+      "loss": 0.5996,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2446043165467626,
+      "grad_norm": 0.7950826287269592,
+      "learning_rate": 0.0001,
+      "loss": 0.9268,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2446899623158616,
+      "grad_norm": 0.8085317015647888,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2447756080849606,
+      "grad_norm": 0.8635661005973816,
+      "learning_rate": 0.0001,
+      "loss": 1.3317,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2448612538540596,
+      "grad_norm": 0.5327932834625244,
+      "learning_rate": 0.0001,
+      "loss": 1.1969,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2449468996231586,
+      "grad_norm": 0.6855432987213135,
+      "learning_rate": 0.0001,
+      "loss": 1.0868,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24503254539225763,
+      "grad_norm": 0.7303252816200256,
+      "learning_rate": 0.0001,
+      "loss": 0.977,
+      "step": 2861
+    },
+    {
+      "epoch": 0.24511819116135664,
+      "grad_norm": 0.9202411770820618,
+      "learning_rate": 0.0001,
+      "loss": 0.7659,
+      "step": 2862
+    },
+    {
+      "epoch": 0.24520383693045564,
+      "grad_norm": 0.7298071980476379,
+      "learning_rate": 0.0001,
+      "loss": 1.2618,
+      "step": 2863
+    },
+    {
+      "epoch": 0.24528948269955464,
+      "grad_norm": 0.710647463798523,
+      "learning_rate": 0.0001,
+      "loss": 0.7984,
+      "step": 2864
+    },
+    {
+      "epoch": 0.24537512846865364,
+      "grad_norm": 0.75152987241745,
+      "learning_rate": 0.0001,
+      "loss": 1.0428,
+      "step": 2865
+    },
+    {
+      "epoch": 0.24546077423775264,
+      "grad_norm": 0.4583593010902405,
+      "learning_rate": 0.0001,
+      "loss": 0.8541,
+      "step": 2866
+    },
+    {
+      "epoch": 0.24554642000685167,
+      "grad_norm": 0.9549511075019836,
+      "learning_rate": 0.0001,
+      "loss": 1.4104,
+      "step": 2867
+    },
+    {
+      "epoch": 0.24563206577595068,
+      "grad_norm": 1.0078797340393066,
+      "learning_rate": 0.0001,
+      "loss": 1.809,
+      "step": 2868
+    },
+    {
+      "epoch": 0.24571771154504968,
+      "grad_norm": 0.5875250697135925,
+      "learning_rate": 0.0001,
+      "loss": 1.0359,
+      "step": 2869
+    },
+    {
+      "epoch": 0.24580335731414868,
+      "grad_norm": 0.897290825843811,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24588900308324768,
+      "grad_norm": 0.9664297103881836,
+      "learning_rate": 0.0001,
+      "loss": 0.9489,
+      "step": 2871
+    },
+    {
+      "epoch": 0.24597464885234668,
+      "grad_norm": 0.8447188138961792,
+      "learning_rate": 0.0001,
+      "loss": 0.8689,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2460602946214457,
+      "grad_norm": 0.8113739490509033,
+      "learning_rate": 0.0001,
+      "loss": 0.6241,
+      "step": 2873
+    },
+    {
+      "epoch": 0.24614594039054472,
+      "grad_norm": 0.39094212651252747,
+      "learning_rate": 0.0001,
+      "loss": 0.923,
+      "step": 2874
+    },
+    {
+      "epoch": 0.24623158615964372,
+      "grad_norm": 0.8395467400550842,
+      "learning_rate": 0.0001,
+      "loss": 0.9702,
+      "step": 2875
+    },
+    {
+      "epoch": 0.24631723192874272,
+      "grad_norm": 0.49253785610198975,
+      "learning_rate": 0.0001,
+      "loss": 0.3885,
+      "step": 2876
+    },
+    {
+      "epoch": 0.24640287769784172,
+      "grad_norm": 0.42888906598091125,
+      "learning_rate": 0.0001,
+      "loss": 1.0546,
+      "step": 2877
+    },
+    {
+      "epoch": 0.24648852346694072,
+      "grad_norm": 0.6202597618103027,
+      "learning_rate": 0.0001,
+      "loss": 0.7019,
+      "step": 2878
+    },
+    {
+      "epoch": 0.24657416923603975,
+      "grad_norm": 0.421184778213501,
+      "learning_rate": 0.0001,
+      "loss": 1.1226,
+      "step": 2879
+    },
+    {
+      "epoch": 0.24665981500513876,
+      "grad_norm": 0.7588629126548767,
+      "learning_rate": 0.0001,
+      "loss": 0.8412,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24674546077423776,
+      "grad_norm": 0.48882269859313965,
+      "learning_rate": 0.0001,
+      "loss": 0.8665,
+      "step": 2881
+    },
+    {
+      "epoch": 0.24683110654333676,
+      "grad_norm": 0.5910060405731201,
+      "learning_rate": 0.0001,
+      "loss": 0.8118,
+      "step": 2882
+    },
+    {
+      "epoch": 0.24691675231243576,
+      "grad_norm": 1.2897777557373047,
+      "learning_rate": 0.0001,
+      "loss": 0.8802,
+      "step": 2883
+    },
+    {
+      "epoch": 0.24700239808153476,
+      "grad_norm": 0.8293715119361877,
+      "learning_rate": 0.0001,
+      "loss": 1.0304,
+      "step": 2884
+    },
+    {
+      "epoch": 0.24708804385063377,
+      "grad_norm": 0.8450093865394592,
+      "learning_rate": 0.0001,
+      "loss": 0.96,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2471736896197328,
+      "grad_norm": 0.2193930745124817,
+      "learning_rate": 0.0001,
+      "loss": 0.396,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2472593353888318,
+      "grad_norm": 0.9829198718070984,
+      "learning_rate": 0.0001,
+      "loss": 0.8315,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2473449811579308,
+      "grad_norm": 1.0117591619491577,
+      "learning_rate": 0.0001,
+      "loss": 1.1272,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2474306269270298,
+      "grad_norm": 0.6146866679191589,
+      "learning_rate": 0.0001,
+      "loss": 0.9059,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2475162726961288,
+      "grad_norm": 0.6333936452865601,
+      "learning_rate": 0.0001,
+      "loss": 0.4763,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2476019184652278,
+      "grad_norm": 0.4674029052257538,
+      "learning_rate": 0.0001,
+      "loss": 0.232,
+      "step": 2891
+    },
+    {
+      "epoch": 0.24768756423432683,
+      "grad_norm": 0.7818970680236816,
+      "learning_rate": 0.0001,
+      "loss": 0.7828,
+      "step": 2892
+    },
+    {
+      "epoch": 0.24777321000342584,
+      "grad_norm": 0.8044243454933167,
+      "learning_rate": 0.0001,
+      "loss": 0.8553,
+      "step": 2893
+    },
+    {
+      "epoch": 0.24785885577252484,
+      "grad_norm": 0.6841492056846619,
+      "learning_rate": 0.0001,
+      "loss": 0.9995,
+      "step": 2894
+    },
+    {
+      "epoch": 0.24794450154162384,
+      "grad_norm": 0.8769542574882507,
+      "learning_rate": 0.0001,
+      "loss": 0.6028,
+      "step": 2895
+    },
+    {
+      "epoch": 0.24803014731072284,
+      "grad_norm": 0.5087369680404663,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 2896
+    },
+    {
+      "epoch": 0.24811579307982184,
+      "grad_norm": 0.8980422616004944,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 2897
+    },
+    {
+      "epoch": 0.24820143884892087,
+      "grad_norm": 0.658931314945221,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 2898
+    },
+    {
+      "epoch": 0.24828708461801988,
+      "grad_norm": 0.7636980414390564,
+      "learning_rate": 0.0001,
+      "loss": 1.2551,
+      "step": 2899
+    },
+    {
+      "epoch": 0.24837273038711888,
+      "grad_norm": 0.5540590882301331,
+      "learning_rate": 0.0001,
+      "loss": 0.538,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24845837615621788,
+      "grad_norm": 0.7448364496231079,
+      "learning_rate": 0.0001,
+      "loss": 1.2902,
+      "step": 2901
+    },
+    {
+      "epoch": 0.24854402192531688,
+      "grad_norm": 1.0145132541656494,
+      "learning_rate": 0.0001,
+      "loss": 1.7825,
+      "step": 2902
+    },
+    {
+      "epoch": 0.24862966769441588,
+      "grad_norm": 0.9014068245887756,
+      "learning_rate": 0.0001,
+      "loss": 1.324,
+      "step": 2903
+    },
+    {
+      "epoch": 0.24871531346351491,
+      "grad_norm": 0.6351029872894287,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 2904
+    },
+    {
+      "epoch": 0.24880095923261392,
+      "grad_norm": 0.5205368399620056,
+      "learning_rate": 0.0001,
+      "loss": 0.6351,
+      "step": 2905
+    },
+    {
+      "epoch": 0.24888660500171292,
+      "grad_norm": 0.8780915141105652,
+      "learning_rate": 0.0001,
+      "loss": 0.8946,
+      "step": 2906
+    },
+    {
+      "epoch": 0.24897225077081192,
+      "grad_norm": 1.1195614337921143,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 2907
+    },
+    {
+      "epoch": 0.24905789653991092,
+      "grad_norm": 0.8405762314796448,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 2908
+    },
+    {
+      "epoch": 0.24914354230900992,
+      "grad_norm": 0.26373451948165894,
+      "learning_rate": 0.0001,
+      "loss": 0.3996,
+      "step": 2909
+    },
+    {
+      "epoch": 0.24922918807810895,
+      "grad_norm": 0.6144184470176697,
+      "learning_rate": 0.0001,
+      "loss": 0.946,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24931483384720796,
+      "grad_norm": 0.7674356698989868,
+      "learning_rate": 0.0001,
+      "loss": 1.1249,
+      "step": 2911
+    },
+    {
+      "epoch": 0.24940047961630696,
+      "grad_norm": 0.6043383479118347,
+      "learning_rate": 0.0001,
+      "loss": 0.6265,
+      "step": 2912
+    },
+    {
+      "epoch": 0.24948612538540596,
+      "grad_norm": 0.821098804473877,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 2913
+    },
+    {
+      "epoch": 0.24957177115450496,
+      "grad_norm": 0.29250624775886536,
+      "learning_rate": 0.0001,
+      "loss": 0.4941,
+      "step": 2914
+    },
+    {
+      "epoch": 0.24965741692360396,
+      "grad_norm": 0.7046856880187988,
+      "learning_rate": 0.0001,
+      "loss": 0.8006,
+      "step": 2915
+    },
+    {
+      "epoch": 0.249743062692703,
+      "grad_norm": 0.5772303342819214,
+      "learning_rate": 0.0001,
+      "loss": 0.9461,
+      "step": 2916
+    },
+    {
+      "epoch": 0.249828708461802,
+      "grad_norm": 0.7293463945388794,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 2917
+    },
+    {
+      "epoch": 0.249914354230901,
+      "grad_norm": 0.3594537377357483,
+      "learning_rate": 0.0001,
+      "loss": 0.715,
+      "step": 2918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8762264251708984,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 2919
+    },
+    {
+      "epoch": 0.25008564576909903,
+      "grad_norm": 0.6971782445907593,
+      "learning_rate": 0.0001,
+      "loss": 0.8503,
+      "step": 2920
+    },
+    {
+      "epoch": 0.250171291538198,
+      "grad_norm": 0.8027974963188171,
+      "learning_rate": 0.0001,
+      "loss": 1.0593,
+      "step": 2921
+    },
+    {
+      "epoch": 0.25025693730729703,
+      "grad_norm": 0.5110678672790527,
+      "learning_rate": 0.0001,
+      "loss": 0.4855,
+      "step": 2922
+    },
+    {
+      "epoch": 0.250342583076396,
+      "grad_norm": 0.899661123752594,
+      "learning_rate": 0.0001,
+      "loss": 0.7923,
+      "step": 2923
+    },
+    {
+      "epoch": 0.25042822884549504,
+      "grad_norm": 0.44862374663352966,
+      "learning_rate": 0.0001,
+      "loss": 0.7013,
+      "step": 2924
+    },
+    {
+      "epoch": 0.250513874614594,
+      "grad_norm": 0.6644240021705627,
+      "learning_rate": 0.0001,
+      "loss": 0.7972,
+      "step": 2925
+    },
+    {
+      "epoch": 0.25059952038369304,
+      "grad_norm": 0.7676885724067688,
+      "learning_rate": 0.0001,
+      "loss": 0.8055,
+      "step": 2926
+    },
+    {
+      "epoch": 0.25068516615279207,
+      "grad_norm": 0.5632569789886475,
+      "learning_rate": 0.0001,
+      "loss": 1.086,
+      "step": 2927
+    },
+    {
+      "epoch": 0.25077081192189105,
+      "grad_norm": 1.0404545068740845,
+      "learning_rate": 0.0001,
+      "loss": 1.1638,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2508564576909901,
+      "grad_norm": 0.946870744228363,
+      "learning_rate": 0.0001,
+      "loss": 0.7048,
+      "step": 2929
+    },
+    {
+      "epoch": 0.25094210346008905,
+      "grad_norm": 0.9062106013298035,
+      "learning_rate": 0.0001,
+      "loss": 1.0067,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2510277492291881,
+      "grad_norm": 0.482864111661911,
+      "learning_rate": 0.0001,
+      "loss": 0.8816,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2511133949982871,
+      "grad_norm": 0.6500808000564575,
+      "learning_rate": 0.0001,
+      "loss": 0.9958,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2511990407673861,
+      "grad_norm": 0.7934048771858215,
+      "learning_rate": 0.0001,
+      "loss": 0.8067,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2512846865364851,
+      "grad_norm": 0.3263147473335266,
+      "learning_rate": 0.0001,
+      "loss": 0.8188,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2513703323055841,
+      "grad_norm": 0.8267553448677063,
+      "learning_rate": 0.0001,
+      "loss": 0.7151,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2514559780746831,
+      "grad_norm": 0.6467897891998291,
+      "learning_rate": 0.0001,
+      "loss": 1.1589,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2515416238437821,
+      "grad_norm": 0.48625195026397705,
+      "learning_rate": 0.0001,
+      "loss": 0.9856,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2516272696128811,
+      "grad_norm": 0.5035710334777832,
+      "learning_rate": 0.0001,
+      "loss": 0.5919,
+      "step": 2938
+    },
+    {
+      "epoch": 0.25171291538198015,
+      "grad_norm": 0.9908263087272644,
+      "learning_rate": 0.0001,
+      "loss": 0.7675,
+      "step": 2939
+    },
+    {
+      "epoch": 0.2517985611510791,
+      "grad_norm": 0.2713707685470581,
+      "learning_rate": 0.0001,
+      "loss": 0.5645,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25188420692017816,
+      "grad_norm": 0.8962688446044922,
+      "learning_rate": 0.0001,
+      "loss": 0.9576,
+      "step": 2941
+    },
+    {
+      "epoch": 0.25196985268927713,
+      "grad_norm": 0.6739730834960938,
+      "learning_rate": 0.0001,
+      "loss": 1.4811,
+      "step": 2942
+    },
+    {
+      "epoch": 0.25205549845837616,
+      "grad_norm": 0.7265098094940186,
+      "learning_rate": 0.0001,
+      "loss": 0.5847,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2521411442274752,
+      "grad_norm": 0.5943591594696045,
+      "learning_rate": 0.0001,
+      "loss": 1.0557,
+      "step": 2944
+    },
+    {
+      "epoch": 0.25222678999657416,
+      "grad_norm": 0.8644501566886902,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2523124357656732,
+      "grad_norm": 0.509766161441803,
+      "learning_rate": 0.0001,
+      "loss": 0.7203,
+      "step": 2946
+    },
+    {
+      "epoch": 0.25239808153477217,
+      "grad_norm": 0.7091003060340881,
+      "learning_rate": 0.0001,
+      "loss": 1.0536,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2524837273038712,
+      "grad_norm": 0.18341127038002014,
+      "learning_rate": 0.0001,
+      "loss": 0.5628,
+      "step": 2948
+    },
+    {
+      "epoch": 0.25256937307297017,
+      "grad_norm": 0.6788491010665894,
+      "learning_rate": 0.0001,
+      "loss": 0.6888,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2526550188420692,
+      "grad_norm": 0.4629046618938446,
+      "learning_rate": 0.0001,
+      "loss": 1.1627,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25274066461116823,
+      "grad_norm": 0.6498672366142273,
+      "learning_rate": 0.0001,
+      "loss": 1.005,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2528263103802672,
+      "grad_norm": 0.9916602969169617,
+      "learning_rate": 0.0001,
+      "loss": 1.0557,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25291195614936623,
+      "grad_norm": 0.6407415866851807,
+      "learning_rate": 0.0001,
+      "loss": 1.0677,
+      "step": 2953
+    },
+    {
+      "epoch": 0.2529976019184652,
+      "grad_norm": 0.8398866057395935,
+      "learning_rate": 0.0001,
+      "loss": 0.6339,
+      "step": 2954
+    },
+    {
+      "epoch": 0.25308324768756424,
+      "grad_norm": 0.5972257852554321,
+      "learning_rate": 0.0001,
+      "loss": 0.5334,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2531688934566632,
+      "grad_norm": 1.0064492225646973,
+      "learning_rate": 0.0001,
+      "loss": 1.3959,
+      "step": 2956
+    },
+    {
+      "epoch": 0.25325453922576224,
+      "grad_norm": 0.4949345886707306,
+      "learning_rate": 0.0001,
+      "loss": 0.9898,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2533401849948613,
+      "grad_norm": 0.4682876765727997,
+      "learning_rate": 0.0001,
+      "loss": 1.0046,
+      "step": 2958
+    },
+    {
+      "epoch": 0.25342583076396025,
+      "grad_norm": 0.26560521125793457,
+      "learning_rate": 0.0001,
+      "loss": 0.5451,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2535114765330593,
+      "grad_norm": 1.1055831909179688,
+      "learning_rate": 0.0001,
+      "loss": 1.1032,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25359712230215825,
+      "grad_norm": 0.6737390160560608,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2536827680712573,
+      "grad_norm": 0.18598349392414093,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2537684138403563,
+      "grad_norm": 0.9833742380142212,
+      "learning_rate": 0.0001,
+      "loss": 1.0852,
+      "step": 2963
+    },
+    {
+      "epoch": 0.2538540596094553,
+      "grad_norm": 0.30864182114601135,
+      "learning_rate": 0.0001,
+      "loss": 0.3421,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2539397053785543,
+      "grad_norm": 0.8364406228065491,
+      "learning_rate": 0.0001,
+      "loss": 1.0618,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2540253511476533,
+      "grad_norm": 0.8940327167510986,
+      "learning_rate": 0.0001,
+      "loss": 1.0111,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2541109969167523,
+      "grad_norm": 0.9576913118362427,
+      "learning_rate": 0.0001,
+      "loss": 1.0223,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2541966426858513,
+      "grad_norm": 0.6924036741256714,
+      "learning_rate": 0.0001,
+      "loss": 0.7834,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2542822884549503,
+      "grad_norm": 0.6171330213546753,
+      "learning_rate": 0.0001,
+      "loss": 1.1398,
+      "step": 2969
+    },
+    {
+      "epoch": 0.25436793422404935,
+      "grad_norm": 0.8565890192985535,
+      "learning_rate": 0.0001,
+      "loss": 0.9864,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2544535799931483,
+      "grad_norm": 0.9871692061424255,
+      "learning_rate": 0.0001,
+      "loss": 0.705,
+      "step": 2971
+    },
+    {
+      "epoch": 0.25453922576224736,
+      "grad_norm": 0.09416119754314423,
+      "learning_rate": 0.0001,
+      "loss": 0.106,
+      "step": 2972
+    },
+    {
+      "epoch": 0.25462487153134633,
+      "grad_norm": 1.0123094320297241,
+      "learning_rate": 0.0001,
+      "loss": 0.5155,
+      "step": 2973
+    },
+    {
+      "epoch": 0.25471051730044536,
+      "grad_norm": 0.6552871465682983,
+      "learning_rate": 0.0001,
+      "loss": 0.838,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2547961630695444,
+      "grad_norm": 1.0212332010269165,
+      "learning_rate": 0.0001,
+      "loss": 1.3569,
+      "step": 2975
+    },
+    {
+      "epoch": 0.25488180883864336,
+      "grad_norm": 0.8341555595397949,
+      "learning_rate": 0.0001,
+      "loss": 0.8616,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2549674546077424,
+      "grad_norm": 0.429613322019577,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 2977
+    },
+    {
+      "epoch": 0.25505310037684137,
+      "grad_norm": 0.8392235636711121,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 2978
+    },
+    {
+      "epoch": 0.2551387461459404,
+      "grad_norm": 0.4730878472328186,
+      "learning_rate": 0.0001,
+      "loss": 0.3296,
+      "step": 2979
+    },
+    {
+      "epoch": 0.2552243919150394,
+      "grad_norm": 1.176443099975586,
+      "learning_rate": 0.0001,
+      "loss": 1.1151,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2553100376841384,
+      "grad_norm": 0.8795576095581055,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 2981
+    },
+    {
+      "epoch": 0.25539568345323743,
+      "grad_norm": 0.5527130365371704,
+      "learning_rate": 0.0001,
+      "loss": 0.8794,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2554813292223364,
+      "grad_norm": 0.8837311863899231,
+      "learning_rate": 0.0001,
+      "loss": 0.8942,
+      "step": 2983
+    },
+    {
+      "epoch": 0.25556697499143544,
+      "grad_norm": 0.8536129593849182,
+      "learning_rate": 0.0001,
+      "loss": 1.2552,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2556526207605344,
+      "grad_norm": 0.7124508023262024,
+      "learning_rate": 0.0001,
+      "loss": 0.8603,
+      "step": 2985
+    },
+    {
+      "epoch": 0.25573826652963344,
+      "grad_norm": 0.7617152333259583,
+      "learning_rate": 0.0001,
+      "loss": 1.1059,
+      "step": 2986
+    },
+    {
+      "epoch": 0.25582391229873247,
+      "grad_norm": 0.12364605069160461,
+      "learning_rate": 0.0001,
+      "loss": 0.5111,
+      "step": 2987
+    },
+    {
+      "epoch": 0.25590955806783144,
+      "grad_norm": 0.7431224584579468,
+      "learning_rate": 0.0001,
+      "loss": 0.8164,
+      "step": 2988
+    },
+    {
+      "epoch": 0.2559952038369305,
+      "grad_norm": 0.8262582421302795,
+      "learning_rate": 0.0001,
+      "loss": 1.2795,
+      "step": 2989
+    },
+    {
+      "epoch": 0.25608084960602945,
+      "grad_norm": 0.5319197773933411,
+      "learning_rate": 0.0001,
+      "loss": 0.8822,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2561664953751285,
+      "grad_norm": 0.517071545124054,
+      "learning_rate": 0.0001,
+      "loss": 0.6396,
+      "step": 2991
+    },
+    {
+      "epoch": 0.25625214114422745,
+      "grad_norm": 0.5123437643051147,
+      "learning_rate": 0.0001,
+      "loss": 0.8882,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2563377869133265,
+      "grad_norm": 0.7947025895118713,
+      "learning_rate": 0.0001,
+      "loss": 0.9432,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2564234326824255,
+      "grad_norm": 0.9249466061592102,
+      "learning_rate": 0.0001,
+      "loss": 1.1529,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2565090784515245,
+      "grad_norm": 0.9496250748634338,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2565947242206235,
+      "grad_norm": 0.49116477370262146,
+      "learning_rate": 0.0001,
+      "loss": 1.0193,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2566803699897225,
+      "grad_norm": 0.7939097881317139,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2567660157588215,
+      "grad_norm": 0.7404831647872925,
+      "learning_rate": 0.0001,
+      "loss": 1.0424,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2568516615279205,
+      "grad_norm": 0.7147152423858643,
+      "learning_rate": 0.0001,
+      "loss": 0.7427,
+      "step": 2999
+    },
+    {
+      "epoch": 0.2569373072970195,
+      "grad_norm": 0.531198263168335,
+      "learning_rate": 0.0001,
+      "loss": 1.1325,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25702295306611855,
+      "grad_norm": 0.8997960686683655,
+      "learning_rate": 0.0001,
+      "loss": 1.1802,
+      "step": 3001
+    },
+    {
+      "epoch": 0.25710859883521753,
+      "grad_norm": 1.0455670356750488,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 3002
+    },
+    {
+      "epoch": 0.25719424460431656,
+      "grad_norm": 0.6826450228691101,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 3003
+    },
+    {
+      "epoch": 0.25727989037341553,
+      "grad_norm": 0.5907893776893616,
+      "learning_rate": 0.0001,
+      "loss": 0.5326,
+      "step": 3004
+    },
+    {
+      "epoch": 0.25736553614251456,
+      "grad_norm": 0.7237275838851929,
+      "learning_rate": 0.0001,
+      "loss": 0.7184,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2574511819116136,
+      "grad_norm": 0.6879096627235413,
+      "learning_rate": 0.0001,
+      "loss": 0.9207,
+      "step": 3006
+    },
+    {
+      "epoch": 0.25753682768071257,
+      "grad_norm": 0.48776060342788696,
+      "learning_rate": 0.0001,
+      "loss": 0.8364,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2576224734498116,
+      "grad_norm": 0.9570364356040955,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 3008
+    },
+    {
+      "epoch": 0.25770811921891057,
+      "grad_norm": 0.9456592202186584,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2577937649880096,
+      "grad_norm": 0.4240400493144989,
+      "learning_rate": 0.0001,
+      "loss": 0.7597,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2578794107571086,
+      "grad_norm": 1.0741080045700073,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2579650565262076,
+      "grad_norm": 0.3811372220516205,
+      "learning_rate": 0.0001,
+      "loss": 0.4108,
+      "step": 3012
+    },
+    {
+      "epoch": 0.25805070229530663,
+      "grad_norm": 0.6508479118347168,
+      "learning_rate": 0.0001,
+      "loss": 0.5276,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2581363480644056,
+      "grad_norm": 0.8377147316932678,
+      "learning_rate": 0.0001,
+      "loss": 1.2042,
+      "step": 3014
+    },
+    {
+      "epoch": 0.25822199383350464,
+      "grad_norm": 0.520591676235199,
+      "learning_rate": 0.0001,
+      "loss": 0.929,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2583076396026036,
+      "grad_norm": 0.6867049336433411,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 3016
+    },
+    {
+      "epoch": 0.25839328537170264,
+      "grad_norm": 0.6920384764671326,
+      "learning_rate": 0.0001,
+      "loss": 0.6129,
+      "step": 3017
+    },
+    {
+      "epoch": 0.25847893114080167,
+      "grad_norm": 0.1770748645067215,
+      "learning_rate": 0.0001,
+      "loss": 0.6174,
+      "step": 3018
+    },
+    {
+      "epoch": 0.25856457690990065,
+      "grad_norm": 0.9026081562042236,
+      "learning_rate": 0.0001,
+      "loss": 1.0724,
+      "step": 3019
+    },
+    {
+      "epoch": 0.2586502226789997,
+      "grad_norm": 0.8595598340034485,
+      "learning_rate": 0.0001,
+      "loss": 0.8423,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25873586844809865,
+      "grad_norm": 0.9105491638183594,
+      "learning_rate": 0.0001,
+      "loss": 1.3221,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2588215142171977,
+      "grad_norm": 1.004754662513733,
+      "learning_rate": 0.0001,
+      "loss": 1.0518,
+      "step": 3022
+    },
+    {
+      "epoch": 0.25890715998629665,
+      "grad_norm": 0.5511451363563538,
+      "learning_rate": 0.0001,
+      "loss": 0.6044,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2589928057553957,
+      "grad_norm": 0.8452308773994446,
+      "learning_rate": 0.0001,
+      "loss": 1.1236,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2590784515244947,
+      "grad_norm": 0.4153958261013031,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2591640972935937,
+      "grad_norm": 1.3324213027954102,
+      "learning_rate": 0.0001,
+      "loss": 0.6659,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2592497430626927,
+      "grad_norm": 0.3339775502681732,
+      "learning_rate": 0.0001,
+      "loss": 0.7905,
+      "step": 3027
+    },
+    {
+      "epoch": 0.2593353888317917,
+      "grad_norm": 0.7438167333602905,
+      "learning_rate": 0.0001,
+      "loss": 0.8685,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2594210346008907,
+      "grad_norm": 0.6129335761070251,
+      "learning_rate": 0.0001,
+      "loss": 1.1647,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2595066803699897,
+      "grad_norm": 0.8044155240058899,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2595923261390887,
+      "grad_norm": 0.5331215262413025,
+      "learning_rate": 0.0001,
+      "loss": 0.4427,
+      "step": 3031
+    },
+    {
+      "epoch": 0.25967797190818775,
+      "grad_norm": 0.5819835662841797,
+      "learning_rate": 0.0001,
+      "loss": 1.2789,
+      "step": 3032
+    },
+    {
+      "epoch": 0.25976361767728673,
+      "grad_norm": 0.830643355846405,
+      "learning_rate": 0.0001,
+      "loss": 0.7615,
+      "step": 3033
+    },
+    {
+      "epoch": 0.25984926344638576,
+      "grad_norm": 0.6030675768852234,
+      "learning_rate": 0.0001,
+      "loss": 0.6722,
+      "step": 3034
+    },
+    {
+      "epoch": 0.25993490921548473,
+      "grad_norm": 0.7380208373069763,
+      "learning_rate": 0.0001,
+      "loss": 1.0955,
+      "step": 3035
+    },
+    {
+      "epoch": 0.26002055498458376,
+      "grad_norm": 0.9988999962806702,
+      "learning_rate": 0.0001,
+      "loss": 1.3453,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2601062007536828,
+      "grad_norm": 0.6191974878311157,
+      "learning_rate": 0.0001,
+      "loss": 0.9949,
+      "step": 3037
+    },
+    {
+      "epoch": 0.26019184652278177,
+      "grad_norm": 0.6586062908172607,
+      "learning_rate": 0.0001,
+      "loss": 1.037,
+      "step": 3038
+    },
+    {
+      "epoch": 0.2602774922918808,
+      "grad_norm": 0.25922343134880066,
+      "learning_rate": 0.0001,
+      "loss": 0.2054,
+      "step": 3039
+    },
+    {
+      "epoch": 0.26036313806097977,
+      "grad_norm": 0.4577895700931549,
+      "learning_rate": 0.0001,
+      "loss": 0.8469,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2604487838300788,
+      "grad_norm": 0.898105800151825,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2605344295991778,
+      "grad_norm": 0.5355742573738098,
+      "learning_rate": 0.0001,
+      "loss": 1.0435,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2606200753682768,
+      "grad_norm": 0.6814472675323486,
+      "learning_rate": 0.0001,
+      "loss": 0.7313,
+      "step": 3043
+    },
+    {
+      "epoch": 0.26070572113737583,
+      "grad_norm": 0.782407820224762,
+      "learning_rate": 0.0001,
+      "loss": 0.9102,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2607913669064748,
+      "grad_norm": 0.8521183133125305,
+      "learning_rate": 0.0001,
+      "loss": 1.113,
+      "step": 3045
+    },
+    {
+      "epoch": 0.26087701267557384,
+      "grad_norm": 0.892219603061676,
+      "learning_rate": 0.0001,
+      "loss": 0.8578,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2609626584446728,
+      "grad_norm": 0.8312287330627441,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 3047
+    },
+    {
+      "epoch": 0.26104830421377184,
+      "grad_norm": 0.6402986645698547,
+      "learning_rate": 0.0001,
+      "loss": 1.0736,
+      "step": 3048
+    },
+    {
+      "epoch": 0.26113394998287087,
+      "grad_norm": 0.9627200365066528,
+      "learning_rate": 0.0001,
+      "loss": 1.4356,
+      "step": 3049
+    },
+    {
+      "epoch": 0.26121959575196985,
+      "grad_norm": 1.1531264781951904,
+      "learning_rate": 0.0001,
+      "loss": 1.0533,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2613052415210689,
+      "grad_norm": 0.8403530716896057,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 3051
+    },
+    {
+      "epoch": 0.26139088729016785,
+      "grad_norm": 0.8241884708404541,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2614765330592669,
+      "grad_norm": 0.5514559745788574,
+      "learning_rate": 0.0001,
+      "loss": 1.2201,
+      "step": 3053
+    },
+    {
+      "epoch": 0.26156217882836585,
+      "grad_norm": 2.209044933319092,
+      "learning_rate": 0.0001,
+      "loss": 0.8548,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2616478245974649,
+      "grad_norm": 0.5867525339126587,
+      "learning_rate": 0.0001,
+      "loss": 0.4868,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2617334703665639,
+      "grad_norm": 0.6651480197906494,
+      "learning_rate": 0.0001,
+      "loss": 1.2688,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2618191161356629,
+      "grad_norm": 0.6299558281898499,
+      "learning_rate": 0.0001,
+      "loss": 0.541,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2619047619047619,
+      "grad_norm": 0.49468737840652466,
+      "learning_rate": 0.0001,
+      "loss": 0.9228,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2619904076738609,
+      "grad_norm": 0.6626302003860474,
+      "learning_rate": 0.0001,
+      "loss": 0.835,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2620760534429599,
+      "grad_norm": 1.0498543977737427,
+      "learning_rate": 0.0001,
+      "loss": 1.3833,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26216169921205895,
+      "grad_norm": 0.7682993412017822,
+      "learning_rate": 0.0001,
+      "loss": 0.8394,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2622473449811579,
+      "grad_norm": 0.9635684490203857,
+      "learning_rate": 0.0001,
+      "loss": 1.1383,
+      "step": 3062
+    },
+    {
+      "epoch": 0.26233299075025696,
+      "grad_norm": 0.7608566284179688,
+      "learning_rate": 0.0001,
+      "loss": 1.0016,
+      "step": 3063
+    },
+    {
+      "epoch": 0.26241863651935593,
+      "grad_norm": 0.7692232131958008,
+      "learning_rate": 0.0001,
+      "loss": 1.2693,
+      "step": 3064
+    },
+    {
+      "epoch": 0.26250428228845496,
+      "grad_norm": 0.3967738449573517,
+      "learning_rate": 0.0001,
+      "loss": 0.7867,
+      "step": 3065
+    },
+    {
+      "epoch": 0.26258992805755393,
+      "grad_norm": 0.4984910488128662,
+      "learning_rate": 0.0001,
+      "loss": 0.9073,
+      "step": 3066
+    },
+    {
+      "epoch": 0.26267557382665296,
+      "grad_norm": 0.5208864808082581,
+      "learning_rate": 0.0001,
+      "loss": 0.638,
+      "step": 3067
+    },
+    {
+      "epoch": 0.262761219595752,
+      "grad_norm": 0.5428648591041565,
+      "learning_rate": 0.0001,
+      "loss": 1.2305,
+      "step": 3068
+    },
+    {
+      "epoch": 0.26284686536485097,
+      "grad_norm": 0.4511657953262329,
+      "learning_rate": 0.0001,
+      "loss": 0.7034,
+      "step": 3069
+    },
+    {
+      "epoch": 0.26293251113395,
+      "grad_norm": 0.5032827258110046,
+      "learning_rate": 0.0001,
+      "loss": 0.8083,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26301815690304897,
+      "grad_norm": 0.8258193135261536,
+      "learning_rate": 0.0001,
+      "loss": 1.2348,
+      "step": 3071
+    },
+    {
+      "epoch": 0.263103802672148,
+      "grad_norm": 0.6904740929603577,
+      "learning_rate": 0.0001,
+      "loss": 0.7047,
+      "step": 3072
+    },
+    {
+      "epoch": 0.263189448441247,
+      "grad_norm": 0.5115885138511658,
+      "learning_rate": 0.0001,
+      "loss": 1.07,
+      "step": 3073
+    },
+    {
+      "epoch": 0.263275094210346,
+      "grad_norm": 0.39273250102996826,
+      "learning_rate": 0.0001,
+      "loss": 0.5028,
+      "step": 3074
+    },
+    {
+      "epoch": 0.26336073997944504,
+      "grad_norm": 0.6126481890678406,
+      "learning_rate": 0.0001,
+      "loss": 1.2721,
+      "step": 3075
+    },
+    {
+      "epoch": 0.263446385748544,
+      "grad_norm": 0.9701017737388611,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 3076
+    },
+    {
+      "epoch": 0.26353203151764304,
+      "grad_norm": 0.48778998851776123,
+      "learning_rate": 0.0001,
+      "loss": 0.9067,
+      "step": 3077
+    },
+    {
+      "epoch": 0.263617677286742,
+      "grad_norm": 0.46998289227485657,
+      "learning_rate": 0.0001,
+      "loss": 1.0008,
+      "step": 3078
+    },
+    {
+      "epoch": 0.26370332305584104,
+      "grad_norm": 0.4578712284564972,
+      "learning_rate": 0.0001,
+      "loss": 0.9819,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2637889688249401,
+      "grad_norm": 0.7881773710250854,
+      "learning_rate": 0.0001,
+      "loss": 1.2753,
+      "step": 3080
+    },
+    {
+      "epoch": 0.26387461459403905,
+      "grad_norm": 0.4878506660461426,
+      "learning_rate": 0.0001,
+      "loss": 0.7267,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2639602603631381,
+      "grad_norm": 0.4687492251396179,
+      "learning_rate": 0.0001,
+      "loss": 0.8746,
+      "step": 3082
+    },
+    {
+      "epoch": 0.26404590613223705,
+      "grad_norm": 0.5791462063789368,
+      "learning_rate": 0.0001,
+      "loss": 0.8766,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2641315519013361,
+      "grad_norm": 0.4255748689174652,
+      "learning_rate": 0.0001,
+      "loss": 1.0128,
+      "step": 3084
+    },
+    {
+      "epoch": 0.26421719767043506,
+      "grad_norm": 0.6884052157402039,
+      "learning_rate": 0.0001,
+      "loss": 0.9742,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2643028434395341,
+      "grad_norm": 1.0000877380371094,
+      "learning_rate": 0.0001,
+      "loss": 0.925,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2643884892086331,
+      "grad_norm": 0.43873754143714905,
+      "learning_rate": 0.0001,
+      "loss": 0.5816,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2644741349777321,
+      "grad_norm": 1.1135571002960205,
+      "learning_rate": 0.0001,
+      "loss": 1.2719,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2645597807468311,
+      "grad_norm": 0.49156421422958374,
+      "learning_rate": 0.0001,
+      "loss": 0.6824,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2646454265159301,
+      "grad_norm": 0.7844637036323547,
+      "learning_rate": 0.0001,
+      "loss": 0.7145,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2647310722850291,
+      "grad_norm": 0.5189628005027771,
+      "learning_rate": 0.0001,
+      "loss": 1.348,
+      "step": 3091
+    },
+    {
+      "epoch": 0.26481671805412815,
+      "grad_norm": 0.5521788001060486,
+      "learning_rate": 0.0001,
+      "loss": 0.9505,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2649023638232271,
+      "grad_norm": 0.650807797908783,
+      "learning_rate": 0.0001,
+      "loss": 1.2385,
+      "step": 3093
+    },
+    {
+      "epoch": 0.26498800959232616,
+      "grad_norm": 0.7351845502853394,
+      "learning_rate": 0.0001,
+      "loss": 1.1805,
+      "step": 3094
+    },
+    {
+      "epoch": 0.26507365536142513,
+      "grad_norm": 0.9065636396408081,
+      "learning_rate": 0.0001,
+      "loss": 1.0661,
+      "step": 3095
+    },
+    {
+      "epoch": 0.26515930113052416,
+      "grad_norm": 0.6640109419822693,
+      "learning_rate": 0.0001,
+      "loss": 0.7667,
+      "step": 3096
+    },
+    {
+      "epoch": 0.26524494689962314,
+      "grad_norm": 0.912059485912323,
+      "learning_rate": 0.0001,
+      "loss": 1.2141,
+      "step": 3097
+    },
+    {
+      "epoch": 0.26533059266872216,
+      "grad_norm": 0.6506069898605347,
+      "learning_rate": 0.0001,
+      "loss": 1.1581,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2654162384378212,
+      "grad_norm": 0.2771294414997101,
+      "learning_rate": 0.0001,
+      "loss": 0.5099,
+      "step": 3099
+    },
+    {
+      "epoch": 0.26550188420692017,
+      "grad_norm": 0.49498820304870605,
+      "learning_rate": 0.0001,
+      "loss": 0.8283,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2655875299760192,
+      "grad_norm": 1.2200114727020264,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2656731757451182,
+      "grad_norm": 0.7270843386650085,
+      "learning_rate": 0.0001,
+      "loss": 0.7426,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2657588215142172,
+      "grad_norm": 0.4101960062980652,
+      "learning_rate": 0.0001,
+      "loss": 0.4559,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2658444672833162,
+      "grad_norm": 0.6920380592346191,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2659301130524152,
+      "grad_norm": 0.5781320333480835,
+      "learning_rate": 0.0001,
+      "loss": 0.5813,
+      "step": 3105
+    },
+    {
+      "epoch": 0.26601575882151424,
+      "grad_norm": 0.5702509880065918,
+      "learning_rate": 0.0001,
+      "loss": 1.0529,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2661014045906132,
+      "grad_norm": 0.8364969491958618,
+      "learning_rate": 0.0001,
+      "loss": 1.3064,
+      "step": 3107
+    },
+    {
+      "epoch": 0.26618705035971224,
+      "grad_norm": 0.8221661448478699,
+      "learning_rate": 0.0001,
+      "loss": 2.3407,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2662726961288112,
+      "grad_norm": 1.0458447933197021,
+      "learning_rate": 0.0001,
+      "loss": 1.1313,
+      "step": 3109
+    },
+    {
+      "epoch": 0.26635834189791024,
+      "grad_norm": 0.6957823038101196,
+      "learning_rate": 0.0001,
+      "loss": 1.1736,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2664439876670093,
+      "grad_norm": 0.8653143048286438,
+      "learning_rate": 0.0001,
+      "loss": 0.9246,
+      "step": 3111
+    },
+    {
+      "epoch": 0.26652963343610825,
+      "grad_norm": 0.8871896266937256,
+      "learning_rate": 0.0001,
+      "loss": 0.8082,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2666152792052073,
+      "grad_norm": 0.5301217436790466,
+      "learning_rate": 0.0001,
+      "loss": 1.1372,
+      "step": 3113
+    },
+    {
+      "epoch": 0.26670092497430625,
+      "grad_norm": 0.7873694896697998,
+      "learning_rate": 0.0001,
+      "loss": 0.6557,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2667865707434053,
+      "grad_norm": 0.7629202008247375,
+      "learning_rate": 0.0001,
+      "loss": 0.6928,
+      "step": 3115
+    },
+    {
+      "epoch": 0.26687221651250426,
+      "grad_norm": 0.5969181656837463,
+      "learning_rate": 0.0001,
+      "loss": 0.4802,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2669578622816033,
+      "grad_norm": 1.0844197273254395,
+      "learning_rate": 0.0001,
+      "loss": 1.2586,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2670435080507023,
+      "grad_norm": 1.0392102003097534,
+      "learning_rate": 0.0001,
+      "loss": 0.7039,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2671291538198013,
+      "grad_norm": 0.50434809923172,
+      "learning_rate": 0.0001,
+      "loss": 0.3654,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2672147995889003,
+      "grad_norm": 0.9094035029411316,
+      "learning_rate": 0.0001,
+      "loss": 0.8347,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2673004453579993,
+      "grad_norm": 0.8484794497489929,
+      "learning_rate": 0.0001,
+      "loss": 0.8618,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2673860911270983,
+      "grad_norm": 0.6673020720481873,
+      "learning_rate": 0.0001,
+      "loss": 0.7334,
+      "step": 3122
+    },
+    {
+      "epoch": 0.26747173689619735,
+      "grad_norm": 1.145127773284912,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 3123
+    },
+    {
+      "epoch": 0.26755738266529633,
+      "grad_norm": 0.4808180630207062,
+      "learning_rate": 0.0001,
+      "loss": 1.1252,
+      "step": 3124
+    },
+    {
+      "epoch": 0.26764302843439536,
+      "grad_norm": 0.8213830590248108,
+      "learning_rate": 0.0001,
+      "loss": 1.0831,
+      "step": 3125
+    },
+    {
+      "epoch": 0.26772867420349433,
+      "grad_norm": 0.7156901359558105,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 3126
+    },
+    {
+      "epoch": 0.26781431997259336,
+      "grad_norm": 0.9539074301719666,
+      "learning_rate": 0.0001,
+      "loss": 1.0129,
+      "step": 3127
+    },
+    {
+      "epoch": 0.26789996574169234,
+      "grad_norm": 0.74574875831604,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 3128
+    },
+    {
+      "epoch": 0.26798561151079137,
+      "grad_norm": 0.5781568288803101,
+      "learning_rate": 0.0001,
+      "loss": 0.737,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2680712572798904,
+      "grad_norm": 0.24588008224964142,
+      "learning_rate": 0.0001,
+      "loss": 0.295,
+      "step": 3130
+    },
+    {
+      "epoch": 0.26815690304898937,
+      "grad_norm": 0.9583401679992676,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2682425488180884,
+      "grad_norm": 0.7424822449684143,
+      "learning_rate": 0.0001,
+      "loss": 1.282,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2683281945871874,
+      "grad_norm": 0.5934191346168518,
+      "learning_rate": 0.0001,
+      "loss": 1.0459,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2684138403562864,
+      "grad_norm": 0.22484298050403595,
+      "learning_rate": 0.0001,
+      "loss": 0.4915,
+      "step": 3134
+    },
+    {
+      "epoch": 0.26849948612538543,
+      "grad_norm": 1.694745421409607,
+      "learning_rate": 0.0001,
+      "loss": 0.9538,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2685851318944844,
+      "grad_norm": 0.8062807321548462,
+      "learning_rate": 0.0001,
+      "loss": 0.7351,
+      "step": 3136
+    },
+    {
+      "epoch": 0.26867077766358344,
+      "grad_norm": 0.4125975966453552,
+      "learning_rate": 0.0001,
+      "loss": 0.4983,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2687564234326824,
+      "grad_norm": 0.251983106136322,
+      "learning_rate": 0.0001,
+      "loss": 0.3747,
+      "step": 3138
+    },
+    {
+      "epoch": 0.26884206920178144,
+      "grad_norm": 1.029518723487854,
+      "learning_rate": 0.0001,
+      "loss": 0.9611,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2689277149708804,
+      "grad_norm": 0.7116233706474304,
+      "learning_rate": 0.0001,
+      "loss": 1.0857,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26901336073997945,
+      "grad_norm": 0.31565675139427185,
+      "learning_rate": 0.0001,
+      "loss": 0.4489,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2690990065090785,
+      "grad_norm": 1.0670076608657837,
+      "learning_rate": 0.0001,
+      "loss": 0.96,
+      "step": 3142
+    },
+    {
+      "epoch": 0.26918465227817745,
+      "grad_norm": 1.6910876035690308,
+      "learning_rate": 0.0001,
+      "loss": 0.9337,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2692702980472765,
+      "grad_norm": 0.5172990560531616,
+      "learning_rate": 0.0001,
+      "loss": 0.7261,
+      "step": 3144
+    },
+    {
+      "epoch": 0.26935594381637545,
+      "grad_norm": 0.7340489029884338,
+      "learning_rate": 0.0001,
+      "loss": 1.0188,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2694415895854745,
+      "grad_norm": 1.036390781402588,
+      "learning_rate": 0.0001,
+      "loss": 1.2303,
+      "step": 3146
+    },
+    {
+      "epoch": 0.26952723535457346,
+      "grad_norm": 0.959773600101471,
+      "learning_rate": 0.0001,
+      "loss": 1.2365,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2696128811236725,
+      "grad_norm": 0.8836159110069275,
+      "learning_rate": 0.0001,
+      "loss": 1.1802,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2696985268927715,
+      "grad_norm": 0.5863701701164246,
+      "learning_rate": 0.0001,
+      "loss": 0.688,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2697841726618705,
+      "grad_norm": 0.5969390273094177,
+      "learning_rate": 0.0001,
+      "loss": 0.8337,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2698698184309695,
+      "grad_norm": 0.5339203476905823,
+      "learning_rate": 0.0001,
+      "loss": 0.6951,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2699554642000685,
+      "grad_norm": 1.1627899408340454,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2700411099691675,
+      "grad_norm": 1.5190892219543457,
+      "learning_rate": 0.0001,
+      "loss": 0.7562,
+      "step": 3153
+    },
+    {
+      "epoch": 0.27012675573826656,
+      "grad_norm": 0.5530638098716736,
+      "learning_rate": 0.0001,
+      "loss": 1.0376,
+      "step": 3154
+    },
+    {
+      "epoch": 0.27021240150736553,
+      "grad_norm": 1.0479520559310913,
+      "learning_rate": 0.0001,
+      "loss": 1.3096,
+      "step": 3155
+    },
+    {
+      "epoch": 0.27029804727646456,
+      "grad_norm": 0.8687689900398254,
+      "learning_rate": 0.0001,
+      "loss": 1.0517,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27038369304556353,
+      "grad_norm": 0.5471723675727844,
+      "learning_rate": 0.0001,
+      "loss": 0.4583,
+      "step": 3157
+    },
+    {
+      "epoch": 0.27046933881466256,
+      "grad_norm": 0.8596060872077942,
+      "learning_rate": 0.0001,
+      "loss": 0.8346,
+      "step": 3158
+    },
+    {
+      "epoch": 0.27055498458376154,
+      "grad_norm": 0.7760963439941406,
+      "learning_rate": 0.0001,
+      "loss": 0.7917,
+      "step": 3159
+    },
+    {
+      "epoch": 0.27064063035286057,
+      "grad_norm": 0.722392201423645,
+      "learning_rate": 0.0001,
+      "loss": 0.6145,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2707262761219596,
+      "grad_norm": 1.09742271900177,
+      "learning_rate": 0.0001,
+      "loss": 0.6572,
+      "step": 3161
+    },
+    {
+      "epoch": 0.27081192189105857,
+      "grad_norm": 0.35765698552131653,
+      "learning_rate": 0.0001,
+      "loss": 0.3959,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2708975676601576,
+      "grad_norm": 0.6139572858810425,
+      "learning_rate": 0.0001,
+      "loss": 1.0756,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2709832134292566,
+      "grad_norm": 0.44297298789024353,
+      "learning_rate": 0.0001,
+      "loss": 0.4827,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2710688591983556,
+      "grad_norm": 0.7757145166397095,
+      "learning_rate": 0.0001,
+      "loss": 0.678,
+      "step": 3165
+    },
+    {
+      "epoch": 0.27115450496745463,
+      "grad_norm": 0.5910109281539917,
+      "learning_rate": 0.0001,
+      "loss": 0.416,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2712401507365536,
+      "grad_norm": 0.8059303760528564,
+      "learning_rate": 0.0001,
+      "loss": 1.1891,
+      "step": 3167
+    },
+    {
+      "epoch": 0.27132579650565264,
+      "grad_norm": 0.6706194281578064,
+      "learning_rate": 0.0001,
+      "loss": 0.9019,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2714114422747516,
+      "grad_norm": 0.5384389758110046,
+      "learning_rate": 0.0001,
+      "loss": 1.168,
+      "step": 3169
+    },
+    {
+      "epoch": 0.27149708804385064,
+      "grad_norm": 0.9333133101463318,
+      "learning_rate": 0.0001,
+      "loss": 1.4203,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2715827338129496,
+      "grad_norm": 0.5429137349128723,
+      "learning_rate": 0.0001,
+      "loss": 0.8371,
+      "step": 3171
+    },
+    {
+      "epoch": 0.27166837958204865,
+      "grad_norm": 0.8900122046470642,
+      "learning_rate": 0.0001,
+      "loss": 0.9381,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2717540253511477,
+      "grad_norm": 0.802410364151001,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 3173
+    },
+    {
+      "epoch": 0.27183967112024665,
+      "grad_norm": 0.9178720116615295,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2719253168893457,
+      "grad_norm": 0.44315823912620544,
+      "learning_rate": 0.0001,
+      "loss": 0.7018,
+      "step": 3175
+    },
+    {
+      "epoch": 0.27201096265844465,
+      "grad_norm": 0.8188782930374146,
+      "learning_rate": 0.0001,
+      "loss": 0.8355,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2720966084275437,
+      "grad_norm": 1.053568720817566,
+      "learning_rate": 0.0001,
+      "loss": 1.3015,
+      "step": 3177
+    },
+    {
+      "epoch": 0.27218225419664266,
+      "grad_norm": 0.6771734952926636,
+      "learning_rate": 0.0001,
+      "loss": 0.5746,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2722678999657417,
+      "grad_norm": 0.5446050763130188,
+      "learning_rate": 0.0001,
+      "loss": 0.41,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2723535457348407,
+      "grad_norm": 0.8969122171401978,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2724391915039397,
+      "grad_norm": 0.8449813723564148,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2725248372730387,
+      "grad_norm": 0.5709275603294373,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2726104830421377,
+      "grad_norm": 0.887831449508667,
+      "learning_rate": 0.0001,
+      "loss": 0.8543,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2726961288112367,
+      "grad_norm": 0.5525292754173279,
+      "learning_rate": 0.0001,
+      "loss": 1.3199,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27278177458033576,
+      "grad_norm": 1.4959871768951416,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 3185
+    },
+    {
+      "epoch": 0.27286742034943473,
+      "grad_norm": 0.4544351100921631,
+      "learning_rate": 0.0001,
+      "loss": 0.9084,
+      "step": 3186
+    },
+    {
+      "epoch": 0.27295306611853376,
+      "grad_norm": 0.6453256011009216,
+      "learning_rate": 0.0001,
+      "loss": 0.5759,
+      "step": 3187
+    },
+    {
+      "epoch": 0.27303871188763273,
+      "grad_norm": 0.8645716905593872,
+      "learning_rate": 0.0001,
+      "loss": 0.6724,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27312435765673176,
+      "grad_norm": 0.9780831336975098,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 3189
+    },
+    {
+      "epoch": 0.27321000342583074,
+      "grad_norm": 1.0640151500701904,
+      "learning_rate": 0.0001,
+      "loss": 0.6408,
+      "step": 3190
+    },
+    {
+      "epoch": 0.27329564919492977,
+      "grad_norm": 0.8310284614562988,
+      "learning_rate": 0.0001,
+      "loss": 0.8101,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2733812949640288,
+      "grad_norm": 0.7759308815002441,
+      "learning_rate": 0.0001,
+      "loss": 0.6357,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2734669407331278,
+      "grad_norm": 0.4437793493270874,
+      "learning_rate": 0.0001,
+      "loss": 0.8529,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2735525865022268,
+      "grad_norm": 0.6949102878570557,
+      "learning_rate": 0.0001,
+      "loss": 0.5542,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2736382322713258,
+      "grad_norm": 0.8831070065498352,
+      "learning_rate": 0.0001,
+      "loss": 1.0741,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2737238780404248,
+      "grad_norm": 1.3807108402252197,
+      "learning_rate": 0.0001,
+      "loss": 1.9105,
+      "step": 3196
+    },
+    {
+      "epoch": 0.27380952380952384,
+      "grad_norm": 0.7997007369995117,
+      "learning_rate": 0.0001,
+      "loss": 1.0683,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2738951695786228,
+      "grad_norm": 1.0351195335388184,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 3198
+    },
+    {
+      "epoch": 0.27398081534772184,
+      "grad_norm": 1.1896787881851196,
+      "learning_rate": 0.0001,
+      "loss": 0.9658,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2740664611168208,
+      "grad_norm": 0.7553027272224426,
+      "learning_rate": 0.0001,
+      "loss": 1.085,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27415210688591984,
+      "grad_norm": 1.1486461162567139,
+      "learning_rate": 0.0001,
+      "loss": 0.8352,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2742377526550188,
+      "grad_norm": 0.963746190071106,
+      "learning_rate": 0.0001,
+      "loss": 0.8781,
+      "step": 3202
+    },
+    {
+      "epoch": 0.27432339842411785,
+      "grad_norm": 0.7503437995910645,
+      "learning_rate": 0.0001,
+      "loss": 0.4673,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2744090441932169,
+      "grad_norm": 0.3861800730228424,
+      "learning_rate": 0.0001,
+      "loss": 0.8299,
+      "step": 3204
+    },
+    {
+      "epoch": 0.27449468996231585,
+      "grad_norm": 0.7762646675109863,
+      "learning_rate": 0.0001,
+      "loss": 1.3374,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2745803357314149,
+      "grad_norm": 0.523257315158844,
+      "learning_rate": 0.0001,
+      "loss": 1.1836,
+      "step": 3206
+    },
+    {
+      "epoch": 0.27466598150051386,
+      "grad_norm": 0.23178623616695404,
+      "learning_rate": 0.0001,
+      "loss": 0.4338,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2747516272696129,
+      "grad_norm": 0.96977299451828,
+      "learning_rate": 0.0001,
+      "loss": 1.7796,
+      "step": 3208
+    },
+    {
+      "epoch": 0.27483727303871186,
+      "grad_norm": 0.45208173990249634,
+      "learning_rate": 0.0001,
+      "loss": 0.7256,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2749229188078109,
+      "grad_norm": 0.6145665049552917,
+      "learning_rate": 0.0001,
+      "loss": 1.2078,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2750085645769099,
+      "grad_norm": 1.1067196130752563,
+      "learning_rate": 0.0001,
+      "loss": 1.1566,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2750942103460089,
+      "grad_norm": 0.9697567224502563,
+      "learning_rate": 0.0001,
+      "loss": 1.2327,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2751798561151079,
+      "grad_norm": 0.8728991746902466,
+      "learning_rate": 0.0001,
+      "loss": 1.0608,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2752655018842069,
+      "grad_norm": 0.9746297597885132,
+      "learning_rate": 0.0001,
+      "loss": 0.8211,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2753511476533059,
+      "grad_norm": 0.5097201466560364,
+      "learning_rate": 0.0001,
+      "loss": 1.1447,
+      "step": 3215
+    },
+    {
+      "epoch": 0.27543679342240496,
+      "grad_norm": 0.8138169646263123,
+      "learning_rate": 0.0001,
+      "loss": 1.3893,
+      "step": 3216
+    },
+    {
+      "epoch": 0.27552243919150393,
+      "grad_norm": 0.4546966552734375,
+      "learning_rate": 0.0001,
+      "loss": 0.7318,
+      "step": 3217
+    },
+    {
+      "epoch": 0.27560808496060296,
+      "grad_norm": 1.0839340686798096,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 3218
+    },
+    {
+      "epoch": 0.27569373072970194,
+      "grad_norm": 0.975492537021637,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 3219
+    },
+    {
+      "epoch": 0.27577937649880097,
+      "grad_norm": 0.4369092881679535,
+      "learning_rate": 0.0001,
+      "loss": 1.2724,
+      "step": 3220
+    },
+    {
+      "epoch": 0.27586502226789994,
+      "grad_norm": 0.7677700519561768,
+      "learning_rate": 0.0001,
+      "loss": 0.7614,
+      "step": 3221
+    },
+    {
+      "epoch": 0.27595066803699897,
+      "grad_norm": 0.4597131609916687,
+      "learning_rate": 0.0001,
+      "loss": 0.8566,
+      "step": 3222
+    },
+    {
+      "epoch": 0.276036313806098,
+      "grad_norm": 1.3881884813308716,
+      "learning_rate": 0.0001,
+      "loss": 0.5724,
+      "step": 3223
+    },
+    {
+      "epoch": 0.276121959575197,
+      "grad_norm": 0.33862608671188354,
+      "learning_rate": 0.0001,
+      "loss": 0.6164,
+      "step": 3224
+    },
+    {
+      "epoch": 0.276207605344296,
+      "grad_norm": 0.8398133516311646,
+      "learning_rate": 0.0001,
+      "loss": 0.7343,
+      "step": 3225
+    },
+    {
+      "epoch": 0.276293251113395,
+      "grad_norm": 0.49041247367858887,
+      "learning_rate": 0.0001,
+      "loss": 0.6413,
+      "step": 3226
+    },
+    {
+      "epoch": 0.276378896882494,
+      "grad_norm": 0.7547625303268433,
+      "learning_rate": 0.0001,
+      "loss": 0.9602,
+      "step": 3227
+    },
+    {
+      "epoch": 0.27646454265159304,
+      "grad_norm": 0.4023687541484833,
+      "learning_rate": 0.0001,
+      "loss": 0.7526,
+      "step": 3228
+    },
+    {
+      "epoch": 0.276550188420692,
+      "grad_norm": 0.4783082902431488,
+      "learning_rate": 0.0001,
+      "loss": 0.848,
+      "step": 3229
+    },
+    {
+      "epoch": 0.27663583418979104,
+      "grad_norm": 0.4853856563568115,
+      "learning_rate": 0.0001,
+      "loss": 0.4668,
+      "step": 3230
+    },
+    {
+      "epoch": 0.27672147995889,
+      "grad_norm": 0.6101393699645996,
+      "learning_rate": 0.0001,
+      "loss": 1.1207,
+      "step": 3231
+    },
+    {
+      "epoch": 0.27680712572798905,
+      "grad_norm": 0.3965330123901367,
+      "learning_rate": 0.0001,
+      "loss": 0.65,
+      "step": 3232
+    },
+    {
+      "epoch": 0.276892771497088,
+      "grad_norm": 1.186692714691162,
+      "learning_rate": 0.0001,
+      "loss": 1.4758,
+      "step": 3233
+    },
+    {
+      "epoch": 0.27697841726618705,
+      "grad_norm": 0.6069204807281494,
+      "learning_rate": 0.0001,
+      "loss": 0.5645,
+      "step": 3234
+    },
+    {
+      "epoch": 0.2770640630352861,
+      "grad_norm": 0.7329856753349304,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 3235
+    },
+    {
+      "epoch": 0.27714970880438505,
+      "grad_norm": 0.9600014686584473,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2772353545734841,
+      "grad_norm": 0.9769136309623718,
+      "learning_rate": 0.0001,
+      "loss": 0.7795,
+      "step": 3237
+    },
+    {
+      "epoch": 0.27732100034258306,
+      "grad_norm": 0.8553221821784973,
+      "learning_rate": 0.0001,
+      "loss": 0.9991,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2774066461116821,
+      "grad_norm": 1.2145850658416748,
+      "learning_rate": 0.0001,
+      "loss": 0.7054,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2774922918807811,
+      "grad_norm": 0.8756325244903564,
+      "learning_rate": 0.0001,
+      "loss": 1.3067,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2775779376498801,
+      "grad_norm": 0.7387400269508362,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2776635834189791,
+      "grad_norm": 0.5804352760314941,
+      "learning_rate": 0.0001,
+      "loss": 0.8988,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2777492291880781,
+      "grad_norm": 0.8135124444961548,
+      "learning_rate": 0.0001,
+      "loss": 0.9691,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2778348749571771,
+      "grad_norm": 0.7275405526161194,
+      "learning_rate": 0.0001,
+      "loss": 0.7859,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2779205207262761,
+      "grad_norm": 0.43681108951568604,
+      "learning_rate": 0.0001,
+      "loss": 0.4603,
+      "step": 3245
+    },
+    {
+      "epoch": 0.27800616649537513,
+      "grad_norm": 0.9157546758651733,
+      "learning_rate": 0.0001,
+      "loss": 0.6271,
+      "step": 3246
+    },
+    {
+      "epoch": 0.27809181226447416,
+      "grad_norm": 0.5564626455307007,
+      "learning_rate": 0.0001,
+      "loss": 1.1212,
+      "step": 3247
+    },
+    {
+      "epoch": 0.27817745803357313,
+      "grad_norm": 0.6209586262702942,
+      "learning_rate": 0.0001,
+      "loss": 0.8236,
+      "step": 3248
+    },
+    {
+      "epoch": 0.27826310380267216,
+      "grad_norm": 0.6077439188957214,
+      "learning_rate": 0.0001,
+      "loss": 1.0154,
+      "step": 3249
+    },
+    {
+      "epoch": 0.27834874957177114,
+      "grad_norm": 0.8536509871482849,
+      "learning_rate": 0.0001,
+      "loss": 1.0565,
+      "step": 3250
+    },
+    {
+      "epoch": 0.27843439534087017,
+      "grad_norm": 0.3889530897140503,
+      "learning_rate": 0.0001,
+      "loss": 0.5411,
+      "step": 3251
+    },
+    {
+      "epoch": 0.27852004110996914,
+      "grad_norm": 0.4277670979499817,
+      "learning_rate": 0.0001,
+      "loss": 0.6461,
+      "step": 3252
+    },
+    {
+      "epoch": 0.27860568687906817,
+      "grad_norm": 0.9533857107162476,
+      "learning_rate": 0.0001,
+      "loss": 0.6653,
+      "step": 3253
+    },
+    {
+      "epoch": 0.2786913326481672,
+      "grad_norm": 0.623180627822876,
+      "learning_rate": 0.0001,
+      "loss": 0.9191,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2787769784172662,
+      "grad_norm": 1.2005881071090698,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2788626241863652,
+      "grad_norm": 0.6327788829803467,
+      "learning_rate": 0.0001,
+      "loss": 1.1594,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2789482699554642,
+      "grad_norm": 0.7716575860977173,
+      "learning_rate": 0.0001,
+      "loss": 0.7328,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2790339157245632,
+      "grad_norm": 0.8673377633094788,
+      "learning_rate": 0.0001,
+      "loss": 1.1037,
+      "step": 3258
+    },
+    {
+      "epoch": 0.27911956149366224,
+      "grad_norm": 0.9249511361122131,
+      "learning_rate": 0.0001,
+      "loss": 0.8667,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2792052072627612,
+      "grad_norm": 0.734154224395752,
+      "learning_rate": 0.0001,
+      "loss": 0.6241,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27929085303186024,
+      "grad_norm": 0.5546109080314636,
+      "learning_rate": 0.0001,
+      "loss": 1.0915,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2793764988009592,
+      "grad_norm": 0.6849035620689392,
+      "learning_rate": 0.0001,
+      "loss": 0.9357,
+      "step": 3262
+    },
+    {
+      "epoch": 0.27946214457005825,
+      "grad_norm": 0.9678376913070679,
+      "learning_rate": 0.0001,
+      "loss": 0.9433,
+      "step": 3263
+    },
+    {
+      "epoch": 0.2795477903391572,
+      "grad_norm": 0.5924509167671204,
+      "learning_rate": 0.0001,
+      "loss": 0.792,
+      "step": 3264
+    },
+    {
+      "epoch": 0.27963343610825625,
+      "grad_norm": 0.68061763048172,
+      "learning_rate": 0.0001,
+      "loss": 1.1148,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2797190818773553,
+      "grad_norm": 0.8122154474258423,
+      "learning_rate": 0.0001,
+      "loss": 0.7702,
+      "step": 3266
+    },
+    {
+      "epoch": 0.27980472764645425,
+      "grad_norm": 0.9229497313499451,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2798903734155533,
+      "grad_norm": 1.090920090675354,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 3268
+    },
+    {
+      "epoch": 0.27997601918465226,
+      "grad_norm": 0.8063322305679321,
+      "learning_rate": 0.0001,
+      "loss": 1.1934,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2800616649537513,
+      "grad_norm": 0.8127689361572266,
+      "learning_rate": 0.0001,
+      "loss": 1.1971,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2801473107228503,
+      "grad_norm": 1.3411794900894165,
+      "learning_rate": 0.0001,
+      "loss": 1.3494,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2802329564919493,
+      "grad_norm": 0.7567200064659119,
+      "learning_rate": 0.0001,
+      "loss": 0.6823,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2803186022610483,
+      "grad_norm": 0.8441646695137024,
+      "learning_rate": 0.0001,
+      "loss": 1.1448,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2804042480301473,
+      "grad_norm": 1.0293142795562744,
+      "learning_rate": 0.0001,
+      "loss": 1.0084,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2804898937992463,
+      "grad_norm": 0.1597876399755478,
+      "learning_rate": 0.0001,
+      "loss": 0.5975,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2805755395683453,
+      "grad_norm": 0.9104405641555786,
+      "learning_rate": 0.0001,
+      "loss": 1.1965,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28066118533744433,
+      "grad_norm": 0.8602784276008606,
+      "learning_rate": 0.0001,
+      "loss": 1.132,
+      "step": 3277
+    },
+    {
+      "epoch": 0.28074683110654336,
+      "grad_norm": 0.6016204357147217,
+      "learning_rate": 0.0001,
+      "loss": 1.1254,
+      "step": 3278
+    },
+    {
+      "epoch": 0.28083247687564233,
+      "grad_norm": 0.9423284530639648,
+      "learning_rate": 0.0001,
+      "loss": 0.91,
+      "step": 3279
+    },
+    {
+      "epoch": 0.28091812264474136,
+      "grad_norm": 0.40621456503868103,
+      "learning_rate": 0.0001,
+      "loss": 0.5957,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28100376841384034,
+      "grad_norm": 0.440426766872406,
+      "learning_rate": 0.0001,
+      "loss": 0.8106,
+      "step": 3281
+    },
+    {
+      "epoch": 0.28108941418293937,
+      "grad_norm": 0.9497649073600769,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 3282
+    },
+    {
+      "epoch": 0.28117505995203834,
+      "grad_norm": 0.7142874002456665,
+      "learning_rate": 0.0001,
+      "loss": 1.2975,
+      "step": 3283
+    },
+    {
+      "epoch": 0.28126070572113737,
+      "grad_norm": 0.853411078453064,
+      "learning_rate": 0.0001,
+      "loss": 0.813,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2813463514902364,
+      "grad_norm": 0.7255558967590332,
+      "learning_rate": 0.0001,
+      "loss": 1.2731,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2814319972593354,
+      "grad_norm": 0.9229912757873535,
+      "learning_rate": 0.0001,
+      "loss": 0.9995,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2815176430284344,
+      "grad_norm": 0.4610539674758911,
+      "learning_rate": 0.0001,
+      "loss": 0.9468,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2816032887975334,
+      "grad_norm": 0.6597099900245667,
+      "learning_rate": 0.0001,
+      "loss": 0.4236,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2816889345666324,
+      "grad_norm": 0.950468897819519,
+      "learning_rate": 0.0001,
+      "loss": 0.7011,
+      "step": 3289
+    },
+    {
+      "epoch": 0.28177458033573144,
+      "grad_norm": 0.6704849600791931,
+      "learning_rate": 0.0001,
+      "loss": 1.1179,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2818602261048304,
+      "grad_norm": 0.4435824155807495,
+      "learning_rate": 0.0001,
+      "loss": 0.3296,
+      "step": 3291
+    },
+    {
+      "epoch": 0.28194587187392944,
+      "grad_norm": 0.7625576853752136,
+      "learning_rate": 0.0001,
+      "loss": 1.0721,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2820315176430284,
+      "grad_norm": 0.9373893737792969,
+      "learning_rate": 0.0001,
+      "loss": 1.2695,
+      "step": 3293
+    },
+    {
+      "epoch": 0.28211716341212745,
+      "grad_norm": 0.6876699924468994,
+      "learning_rate": 0.0001,
+      "loss": 0.6153,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2822028091812264,
+      "grad_norm": 0.8141164183616638,
+      "learning_rate": 0.0001,
+      "loss": 0.6131,
+      "step": 3295
+    },
+    {
+      "epoch": 0.28228845495032545,
+      "grad_norm": 0.9192612171173096,
+      "learning_rate": 0.0001,
+      "loss": 1.3747,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2823741007194245,
+      "grad_norm": 0.4867872893810272,
+      "learning_rate": 0.0001,
+      "loss": 0.9775,
+      "step": 3297
+    },
+    {
+      "epoch": 0.28245974648852346,
+      "grad_norm": 1.0176634788513184,
+      "learning_rate": 0.0001,
+      "loss": 0.8849,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2825453922576225,
+      "grad_norm": 0.8420210480690002,
+      "learning_rate": 0.0001,
+      "loss": 1.2497,
+      "step": 3299
+    },
+    {
+      "epoch": 0.28263103802672146,
+      "grad_norm": 0.7384817004203796,
+      "learning_rate": 0.0001,
+      "loss": 1.2197,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2827166837958205,
+      "grad_norm": 0.6272978186607361,
+      "learning_rate": 0.0001,
+      "loss": 0.7036,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2828023295649195,
+      "grad_norm": 0.5906031727790833,
+      "learning_rate": 0.0001,
+      "loss": 1.0729,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2828879753340185,
+      "grad_norm": 0.7619339227676392,
+      "learning_rate": 0.0001,
+      "loss": 1.0008,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2829736211031175,
+      "grad_norm": 1.7522646188735962,
+      "learning_rate": 0.0001,
+      "loss": 1.2599,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2830592668722165,
+      "grad_norm": 0.5877655744552612,
+      "learning_rate": 0.0001,
+      "loss": 1.0015,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2831449126413155,
+      "grad_norm": 0.36292555928230286,
+      "learning_rate": 0.0001,
+      "loss": 0.6743,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2832305584104145,
+      "grad_norm": 0.5050802230834961,
+      "learning_rate": 0.0001,
+      "loss": 1.0686,
+      "step": 3307
+    },
+    {
+      "epoch": 0.28331620417951353,
+      "grad_norm": 0.5790701508522034,
+      "learning_rate": 0.0001,
+      "loss": 0.724,
+      "step": 3308
+    },
+    {
+      "epoch": 0.28340184994861256,
+      "grad_norm": 0.6444624662399292,
+      "learning_rate": 0.0001,
+      "loss": 1.1113,
+      "step": 3309
+    },
+    {
+      "epoch": 0.28348749571771154,
+      "grad_norm": 0.40587857365608215,
+      "learning_rate": 0.0001,
+      "loss": 1.2513,
+      "step": 3310
+    },
+    {
+      "epoch": 0.28357314148681056,
+      "grad_norm": 1.3657066822052002,
+      "learning_rate": 0.0001,
+      "loss": 1.8609,
+      "step": 3311
+    },
+    {
+      "epoch": 0.28365878725590954,
+      "grad_norm": 0.4221973121166229,
+      "learning_rate": 0.0001,
+      "loss": 0.7293,
+      "step": 3312
+    },
+    {
+      "epoch": 0.28374443302500857,
+      "grad_norm": 0.5109499096870422,
+      "learning_rate": 0.0001,
+      "loss": 0.8036,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2838300787941076,
+      "grad_norm": 0.85152268409729,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2839157245632066,
+      "grad_norm": 1.08304762840271,
+      "learning_rate": 0.0001,
+      "loss": 1.2699,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2840013703323056,
+      "grad_norm": 0.8052664399147034,
+      "learning_rate": 0.0001,
+      "loss": 0.6914,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2840870161014046,
+      "grad_norm": 0.49574801325798035,
+      "learning_rate": 0.0001,
+      "loss": 0.2792,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2841726618705036,
+      "grad_norm": 0.9136214256286621,
+      "learning_rate": 0.0001,
+      "loss": 1.3053,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2842583076396026,
+      "grad_norm": 0.8570126891136169,
+      "learning_rate": 0.0001,
+      "loss": 1.4147,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2843439534087016,
+      "grad_norm": 0.9274875521659851,
+      "learning_rate": 0.0001,
+      "loss": 0.825,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28442959917780064,
+      "grad_norm": 1.3084990978240967,
+      "learning_rate": 0.0001,
+      "loss": 2.0402,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2845152449468996,
+      "grad_norm": 0.734194815158844,
+      "learning_rate": 0.0001,
+      "loss": 0.9104,
+      "step": 3322
+    },
+    {
+      "epoch": 0.28460089071599864,
+      "grad_norm": 0.4273173213005066,
+      "learning_rate": 0.0001,
+      "loss": 0.606,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2846865364850976,
+      "grad_norm": 1.386075496673584,
+      "learning_rate": 0.0001,
+      "loss": 1.1208,
+      "step": 3324
+    },
+    {
+      "epoch": 0.28477218225419665,
+      "grad_norm": 0.5813913941383362,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2848578280232956,
+      "grad_norm": 0.6011453866958618,
+      "learning_rate": 0.0001,
+      "loss": 0.9517,
+      "step": 3326
+    },
+    {
+      "epoch": 0.28494347379239465,
+      "grad_norm": 0.3258308470249176,
+      "learning_rate": 0.0001,
+      "loss": 0.4402,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2850291195614937,
+      "grad_norm": 0.6993041038513184,
+      "learning_rate": 0.0001,
+      "loss": 0.732,
+      "step": 3328
+    },
+    {
+      "epoch": 0.28511476533059266,
+      "grad_norm": 1.0141106843948364,
+      "learning_rate": 0.0001,
+      "loss": 1.3147,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2852004110996917,
+      "grad_norm": 0.49041059613227844,
+      "learning_rate": 0.0001,
+      "loss": 0.7272,
+      "step": 3330
+    },
+    {
+      "epoch": 0.28528605686879066,
+      "grad_norm": 1.1090794801712036,
+      "learning_rate": 0.0001,
+      "loss": 0.9972,
+      "step": 3331
+    },
+    {
+      "epoch": 0.2853717026378897,
+      "grad_norm": 0.8872842192649841,
+      "learning_rate": 0.0001,
+      "loss": 1.0774,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2854573484069887,
+      "grad_norm": 0.47178053855895996,
+      "learning_rate": 0.0001,
+      "loss": 0.8119,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2855429941760877,
+      "grad_norm": 0.6544367671012878,
+      "learning_rate": 0.0001,
+      "loss": 1.0477,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2856286399451867,
+      "grad_norm": 1.0644080638885498,
+      "learning_rate": 0.0001,
+      "loss": 1.1098,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.9029351472854614,
+      "learning_rate": 0.0001,
+      "loss": 1.3584,
+      "step": 3336
+    },
+    {
+      "epoch": 0.28579993148338473,
+      "grad_norm": 1.0936691761016846,
+      "learning_rate": 0.0001,
+      "loss": 1.1084,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2858855772524837,
+      "grad_norm": 0.6634782552719116,
+      "learning_rate": 0.0001,
+      "loss": 0.4161,
+      "step": 3338
+    },
+    {
+      "epoch": 0.28597122302158273,
+      "grad_norm": 0.5184661746025085,
+      "learning_rate": 0.0001,
+      "loss": 1.0046,
+      "step": 3339
+    },
+    {
+      "epoch": 0.28605686879068176,
+      "grad_norm": 0.7893422245979309,
+      "learning_rate": 0.0001,
+      "loss": 1.067,
+      "step": 3340
+    },
+    {
+      "epoch": 0.28614251455978074,
+      "grad_norm": 0.5579618811607361,
+      "learning_rate": 0.0001,
+      "loss": 0.7673,
+      "step": 3341
+    },
+    {
+      "epoch": 0.28622816032887977,
+      "grad_norm": 0.7333992719650269,
+      "learning_rate": 0.0001,
+      "loss": 0.7693,
+      "step": 3342
+    },
+    {
+      "epoch": 0.28631380609797874,
+      "grad_norm": 0.7497597932815552,
+      "learning_rate": 0.0001,
+      "loss": 1.1877,
+      "step": 3343
+    },
+    {
+      "epoch": 0.28639945186707777,
+      "grad_norm": 0.9375843405723572,
+      "learning_rate": 0.0001,
+      "loss": 1.0697,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2864850976361768,
+      "grad_norm": 0.7780022025108337,
+      "learning_rate": 0.0001,
+      "loss": 0.6053,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2865707434052758,
+      "grad_norm": 0.7891181707382202,
+      "learning_rate": 0.0001,
+      "loss": 0.9886,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2866563891743748,
+      "grad_norm": 0.6181745529174805,
+      "learning_rate": 0.0001,
+      "loss": 1.2422,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2867420349434738,
+      "grad_norm": 0.836759626865387,
+      "learning_rate": 0.0001,
+      "loss": 1.0873,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2868276807125728,
+      "grad_norm": 0.9662436842918396,
+      "learning_rate": 0.0001,
+      "loss": 0.8405,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2869133264816718,
+      "grad_norm": 0.5771559476852417,
+      "learning_rate": 0.0001,
+      "loss": 0.3937,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2869989722507708,
+      "grad_norm": 1.0003854036331177,
+      "learning_rate": 0.0001,
+      "loss": 1.0767,
+      "step": 3351
+    },
+    {
+      "epoch": 0.28708461801986984,
+      "grad_norm": 0.8331012725830078,
+      "learning_rate": 0.0001,
+      "loss": 0.9043,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2871702637889688,
+      "grad_norm": 0.461577832698822,
+      "learning_rate": 0.0001,
+      "loss": 0.8538,
+      "step": 3353
+    },
+    {
+      "epoch": 0.28725590955806785,
+      "grad_norm": 0.6814552545547485,
+      "learning_rate": 0.0001,
+      "loss": 1.8042,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2873415553271668,
+      "grad_norm": 0.3697505593299866,
+      "learning_rate": 0.0001,
+      "loss": 0.7209,
+      "step": 3355
+    },
+    {
+      "epoch": 0.28742720109626585,
+      "grad_norm": 0.6393977403640747,
+      "learning_rate": 0.0001,
+      "loss": 0.7765,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2875128468653648,
+      "grad_norm": 1.1150652170181274,
+      "learning_rate": 0.0001,
+      "loss": 0.9478,
+      "step": 3357
+    },
+    {
+      "epoch": 0.28759849263446385,
+      "grad_norm": 1.1992329359054565,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2876841384035629,
+      "grad_norm": 0.8227225542068481,
+      "learning_rate": 0.0001,
+      "loss": 0.7398,
+      "step": 3359
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.7954972982406616,
+      "learning_rate": 0.0001,
+      "loss": 0.9291,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2878554299417609,
+      "grad_norm": 0.719910204410553,
+      "learning_rate": 0.0001,
+      "loss": 0.8317,
+      "step": 3361
+    },
+    {
+      "epoch": 0.28794107571085986,
+      "grad_norm": 0.5127081274986267,
+      "learning_rate": 0.0001,
+      "loss": 1.0661,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2880267214799589,
+      "grad_norm": 0.9425294399261475,
+      "learning_rate": 0.0001,
+      "loss": 0.7861,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2881123672490579,
+      "grad_norm": 0.8609326481819153,
+      "learning_rate": 0.0001,
+      "loss": 1.2889,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2881980130181569,
+      "grad_norm": 0.4913707971572876,
+      "learning_rate": 0.0001,
+      "loss": 0.8089,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2882836587872559,
+      "grad_norm": 0.4479687213897705,
+      "learning_rate": 0.0001,
+      "loss": 0.5873,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2883693045563549,
+      "grad_norm": 0.8025332093238831,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 3367
+    },
+    {
+      "epoch": 0.28845495032545393,
+      "grad_norm": 0.9674707651138306,
+      "learning_rate": 0.0001,
+      "loss": 0.8108,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2885405960945529,
+      "grad_norm": 0.7094219326972961,
+      "learning_rate": 0.0001,
+      "loss": 0.683,
+      "step": 3369
+    },
+    {
+      "epoch": 0.28862624186365193,
+      "grad_norm": 0.9695871472358704,
+      "learning_rate": 0.0001,
+      "loss": 1.1042,
+      "step": 3370
+    },
+    {
+      "epoch": 0.28871188763275096,
+      "grad_norm": 0.8946298360824585,
+      "learning_rate": 0.0001,
+      "loss": 0.858,
+      "step": 3371
+    },
+    {
+      "epoch": 0.28879753340184994,
+      "grad_norm": 0.6023060083389282,
+      "learning_rate": 0.0001,
+      "loss": 1.0401,
+      "step": 3372
+    },
+    {
+      "epoch": 0.28888317917094897,
+      "grad_norm": 0.4576970934867859,
+      "learning_rate": 0.0001,
+      "loss": 0.7829,
+      "step": 3373
+    },
+    {
+      "epoch": 0.28896882494004794,
+      "grad_norm": 0.1787348836660385,
+      "learning_rate": 0.0001,
+      "loss": 0.7237,
+      "step": 3374
+    },
+    {
+      "epoch": 0.28905447070914697,
+      "grad_norm": 1.080391526222229,
+      "learning_rate": 0.0001,
+      "loss": 1.2696,
+      "step": 3375
+    },
+    {
+      "epoch": 0.289140116478246,
+      "grad_norm": 0.556543231010437,
+      "learning_rate": 0.0001,
+      "loss": 0.8413,
+      "step": 3376
+    },
+    {
+      "epoch": 0.289225762247345,
+      "grad_norm": 0.8265848159790039,
+      "learning_rate": 0.0001,
+      "loss": 0.9284,
+      "step": 3377
+    },
+    {
+      "epoch": 0.289311408016444,
+      "grad_norm": 1.1561189889907837,
+      "learning_rate": 0.0001,
+      "loss": 1.2184,
+      "step": 3378
+    },
+    {
+      "epoch": 0.289397053785543,
+      "grad_norm": 0.7847104668617249,
+      "learning_rate": 0.0001,
+      "loss": 1.0048,
+      "step": 3379
+    },
+    {
+      "epoch": 0.289482699554642,
+      "grad_norm": 0.86273592710495,
+      "learning_rate": 0.0001,
+      "loss": 1.1866,
+      "step": 3380
+    },
+    {
+      "epoch": 0.289568345323741,
+      "grad_norm": 1.116898775100708,
+      "learning_rate": 0.0001,
+      "loss": 0.5073,
+      "step": 3381
+    },
+    {
+      "epoch": 0.28965399109284,
+      "grad_norm": 1.4579551219940186,
+      "learning_rate": 0.0001,
+      "loss": 0.7035,
+      "step": 3382
+    },
+    {
+      "epoch": 0.28973963686193904,
+      "grad_norm": 0.9902433753013611,
+      "learning_rate": 0.0001,
+      "loss": 1.2229,
+      "step": 3383
+    },
+    {
+      "epoch": 0.289825282631038,
+      "grad_norm": 0.5121811628341675,
+      "learning_rate": 0.0001,
+      "loss": 0.3384,
+      "step": 3384
+    },
+    {
+      "epoch": 0.28991092840013705,
+      "grad_norm": 0.5996220111846924,
+      "learning_rate": 0.0001,
+      "loss": 0.7,
+      "step": 3385
+    },
+    {
+      "epoch": 0.289996574169236,
+      "grad_norm": 0.851335346698761,
+      "learning_rate": 0.0001,
+      "loss": 1.0961,
+      "step": 3386
+    },
+    {
+      "epoch": 0.29008221993833505,
+      "grad_norm": 0.6123542189598083,
+      "learning_rate": 0.0001,
+      "loss": 0.9103,
+      "step": 3387
+    },
+    {
+      "epoch": 0.290167865707434,
+      "grad_norm": 0.43083280324935913,
+      "learning_rate": 0.0001,
+      "loss": 0.5744,
+      "step": 3388
+    },
+    {
+      "epoch": 0.29025351147653305,
+      "grad_norm": 0.772055447101593,
+      "learning_rate": 0.0001,
+      "loss": 0.6297,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2903391572456321,
+      "grad_norm": 0.41605037450790405,
+      "learning_rate": 0.0001,
+      "loss": 0.342,
+      "step": 3390
+    },
+    {
+      "epoch": 0.29042480301473106,
+      "grad_norm": 0.5119863152503967,
+      "learning_rate": 0.0001,
+      "loss": 0.7285,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2905104487838301,
+      "grad_norm": 0.4551796317100525,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29059609455292906,
+      "grad_norm": 0.6782887578010559,
+      "learning_rate": 0.0001,
+      "loss": 1.2193,
+      "step": 3393
+    },
+    {
+      "epoch": 0.2906817403220281,
+      "grad_norm": 1.1197854280471802,
+      "learning_rate": 0.0001,
+      "loss": 1.0553,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2907673860911271,
+      "grad_norm": 0.5718299150466919,
+      "learning_rate": 0.0001,
+      "loss": 1.4604,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2908530318602261,
+      "grad_norm": 1.1886217594146729,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2909386776293251,
+      "grad_norm": 0.7862503528594971,
+      "learning_rate": 0.0001,
+      "loss": 0.7309,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2910243233984241,
+      "grad_norm": 0.5198236703872681,
+      "learning_rate": 0.0001,
+      "loss": 0.511,
+      "step": 3398
+    },
+    {
+      "epoch": 0.29110996916752313,
+      "grad_norm": 0.8423651456832886,
+      "learning_rate": 0.0001,
+      "loss": 1.0494,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2911956149366221,
+      "grad_norm": 0.6052988171577454,
+      "learning_rate": 0.0001,
+      "loss": 1.1405,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29128126070572113,
+      "grad_norm": 1.1685277223587036,
+      "learning_rate": 0.0001,
+      "loss": 0.7833,
+      "step": 3401
+    },
+    {
+      "epoch": 0.29136690647482016,
+      "grad_norm": 0.7983923554420471,
+      "learning_rate": 0.0001,
+      "loss": 0.8061,
+      "step": 3402
+    },
+    {
+      "epoch": 0.29145255224391914,
+      "grad_norm": 0.5118330717086792,
+      "learning_rate": 0.0001,
+      "loss": 0.7982,
+      "step": 3403
+    },
+    {
+      "epoch": 0.29153819801301817,
+      "grad_norm": 0.8321371078491211,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29162384378211714,
+      "grad_norm": 0.7357081770896912,
+      "learning_rate": 0.0001,
+      "loss": 1.0569,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2917094895512162,
+      "grad_norm": 0.5373146533966064,
+      "learning_rate": 0.0001,
+      "loss": 0.6975,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2917951353203152,
+      "grad_norm": 0.6367455720901489,
+      "learning_rate": 0.0001,
+      "loss": 0.5131,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2918807810894142,
+      "grad_norm": 0.5644450187683105,
+      "learning_rate": 0.0001,
+      "loss": 0.6862,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2919664268585132,
+      "grad_norm": 0.5015959739685059,
+      "learning_rate": 0.0001,
+      "loss": 1.0206,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2920520726276122,
+      "grad_norm": 0.5417486429214478,
+      "learning_rate": 0.0001,
+      "loss": 0.9694,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2921377183967112,
+      "grad_norm": 0.7998532652854919,
+      "learning_rate": 0.0001,
+      "loss": 0.6314,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2922233641658102,
+      "grad_norm": 0.6136223673820496,
+      "learning_rate": 0.0001,
+      "loss": 1.3991,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2923090099349092,
+      "grad_norm": 0.8849233984947205,
+      "learning_rate": 0.0001,
+      "loss": 1.9906,
+      "step": 3413
+    },
+    {
+      "epoch": 0.29239465570400824,
+      "grad_norm": 0.9164513349533081,
+      "learning_rate": 0.0001,
+      "loss": 0.6743,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2924803014731072,
+      "grad_norm": 0.776380181312561,
+      "learning_rate": 0.0001,
+      "loss": 0.6274,
+      "step": 3415
+    },
+    {
+      "epoch": 0.29256594724220625,
+      "grad_norm": 0.5512747764587402,
+      "learning_rate": 0.0001,
+      "loss": 0.5695,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2926515930113052,
+      "grad_norm": 0.8714987635612488,
+      "learning_rate": 0.0001,
+      "loss": 0.8074,
+      "step": 3417
+    },
+    {
+      "epoch": 0.29273723878040425,
+      "grad_norm": 0.6272192597389221,
+      "learning_rate": 0.0001,
+      "loss": 0.6812,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2928228845495033,
+      "grad_norm": 0.913780152797699,
+      "learning_rate": 0.0001,
+      "loss": 1.0808,
+      "step": 3419
+    },
+    {
+      "epoch": 0.29290853031860226,
+      "grad_norm": 0.4971335530281067,
+      "learning_rate": 0.0001,
+      "loss": 0.8086,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2929941760877013,
+      "grad_norm": 0.6149380803108215,
+      "learning_rate": 0.0001,
+      "loss": 0.4829,
+      "step": 3421
+    },
+    {
+      "epoch": 0.29307982185680026,
+      "grad_norm": 1.209468960762024,
+      "learning_rate": 0.0001,
+      "loss": 1.0068,
+      "step": 3422
+    },
+    {
+      "epoch": 0.2931654676258993,
+      "grad_norm": 0.9825507998466492,
+      "learning_rate": 0.0001,
+      "loss": 0.8651,
+      "step": 3423
+    },
+    {
+      "epoch": 0.29325111339499826,
+      "grad_norm": 0.6189348697662354,
+      "learning_rate": 0.0001,
+      "loss": 1.302,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2933367591640973,
+      "grad_norm": 0.6384595036506653,
+      "learning_rate": 0.0001,
+      "loss": 1.2532,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2934224049331963,
+      "grad_norm": 0.3744337260723114,
+      "learning_rate": 0.0001,
+      "loss": 0.8164,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2935080507022953,
+      "grad_norm": 0.6619053483009338,
+      "learning_rate": 0.0001,
+      "loss": 0.4104,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2935936964713943,
+      "grad_norm": 0.5032241940498352,
+      "learning_rate": 0.0001,
+      "loss": 1.107,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2936793422404933,
+      "grad_norm": 1.0680584907531738,
+      "learning_rate": 0.0001,
+      "loss": 0.7274,
+      "step": 3429
+    },
+    {
+      "epoch": 0.29376498800959233,
+      "grad_norm": 0.7240095734596252,
+      "learning_rate": 0.0001,
+      "loss": 1.278,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2938506337786913,
+      "grad_norm": 0.9952961206436157,
+      "learning_rate": 0.0001,
+      "loss": 0.8947,
+      "step": 3431
+    },
+    {
+      "epoch": 0.29393627954779034,
+      "grad_norm": 0.7070422768592834,
+      "learning_rate": 0.0001,
+      "loss": 0.4722,
+      "step": 3432
+    },
+    {
+      "epoch": 0.29402192531688937,
+      "grad_norm": 0.8203014135360718,
+      "learning_rate": 0.0001,
+      "loss": 1.0161,
+      "step": 3433
+    },
+    {
+      "epoch": 0.29410757108598834,
+      "grad_norm": 0.8565195798873901,
+      "learning_rate": 0.0001,
+      "loss": 0.8998,
+      "step": 3434
+    },
+    {
+      "epoch": 0.29419321685508737,
+      "grad_norm": 0.9683007001876831,
+      "learning_rate": 0.0001,
+      "loss": 1.3019,
+      "step": 3435
+    },
+    {
+      "epoch": 0.29427886262418634,
+      "grad_norm": 0.8478633165359497,
+      "learning_rate": 0.0001,
+      "loss": 0.949,
+      "step": 3436
+    },
+    {
+      "epoch": 0.2943645083932854,
+      "grad_norm": 0.1419105976819992,
+      "learning_rate": 0.0001,
+      "loss": 0.0552,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2944501541623844,
+      "grad_norm": 0.7162604928016663,
+      "learning_rate": 0.0001,
+      "loss": 1.0135,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2945357999314834,
+      "grad_norm": 0.5454469323158264,
+      "learning_rate": 0.0001,
+      "loss": 0.5627,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2946214457005824,
+      "grad_norm": 0.7348208427429199,
+      "learning_rate": 0.0001,
+      "loss": 0.8449,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2947070914696814,
+      "grad_norm": 0.5429595112800598,
+      "learning_rate": 0.0001,
+      "loss": 0.7411,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2947927372387804,
+      "grad_norm": 0.6807927489280701,
+      "learning_rate": 0.0001,
+      "loss": 1.2975,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2948783830078794,
+      "grad_norm": 0.6168951988220215,
+      "learning_rate": 0.0001,
+      "loss": 1.0643,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2949640287769784,
+      "grad_norm": 0.8328611850738525,
+      "learning_rate": 0.0001,
+      "loss": 0.8242,
+      "step": 3444
+    },
+    {
+      "epoch": 0.29504967454607745,
+      "grad_norm": 0.6014366745948792,
+      "learning_rate": 0.0001,
+      "loss": 1.2431,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2951353203151764,
+      "grad_norm": 0.44359949231147766,
+      "learning_rate": 0.0001,
+      "loss": 0.7928,
+      "step": 3446
+    },
+    {
+      "epoch": 0.29522096608427545,
+      "grad_norm": 0.7576709985733032,
+      "learning_rate": 0.0001,
+      "loss": 1.3106,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2953066118533744,
+      "grad_norm": 0.8144282698631287,
+      "learning_rate": 0.0001,
+      "loss": 0.8805,
+      "step": 3448
+    },
+    {
+      "epoch": 0.29539225762247345,
+      "grad_norm": 0.6306112408638,
+      "learning_rate": 0.0001,
+      "loss": 1.1745,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2954779033915725,
+      "grad_norm": 0.6411952376365662,
+      "learning_rate": 0.0001,
+      "loss": 1.2424,
+      "step": 3450
+    },
+    {
+      "epoch": 0.29556354916067146,
+      "grad_norm": 0.7585483193397522,
+      "learning_rate": 0.0001,
+      "loss": 0.8884,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2956491949297705,
+      "grad_norm": 0.9177693724632263,
+      "learning_rate": 0.0001,
+      "loss": 0.8716,
+      "step": 3452
+    },
+    {
+      "epoch": 0.29573484069886946,
+      "grad_norm": 0.21694734692573547,
+      "learning_rate": 0.0001,
+      "loss": 0.2946,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2958204864679685,
+      "grad_norm": 0.6056800484657288,
+      "learning_rate": 0.0001,
+      "loss": 0.6184,
+      "step": 3454
+    },
+    {
+      "epoch": 0.29590613223706747,
+      "grad_norm": 0.9702003002166748,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2959917780061665,
+      "grad_norm": 0.5211352705955505,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2960774237752655,
+      "grad_norm": 0.724011242389679,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2961630695443645,
+      "grad_norm": 0.8947895765304565,
+      "learning_rate": 0.0001,
+      "loss": 1.0878,
+      "step": 3458
+    },
+    {
+      "epoch": 0.29624871531346353,
+      "grad_norm": 0.9594396948814392,
+      "learning_rate": 0.0001,
+      "loss": 0.5543,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2963343610825625,
+      "grad_norm": 0.25633737444877625,
+      "learning_rate": 0.0001,
+      "loss": 0.3869,
+      "step": 3460
+    },
+    {
+      "epoch": 0.29642000685166153,
+      "grad_norm": 0.4652431905269623,
+      "learning_rate": 0.0001,
+      "loss": 0.9857,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2965056526207605,
+      "grad_norm": 0.8645159602165222,
+      "learning_rate": 0.0001,
+      "loss": 1.2852,
+      "step": 3462
+    },
+    {
+      "epoch": 0.29659129838985954,
+      "grad_norm": 0.6995427012443542,
+      "learning_rate": 0.0001,
+      "loss": 1.9841,
+      "step": 3463
+    },
+    {
+      "epoch": 0.29667694415895857,
+      "grad_norm": 1.3758527040481567,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 3464
+    },
+    {
+      "epoch": 0.29676258992805754,
+      "grad_norm": 0.12265872210264206,
+      "learning_rate": 0.0001,
+      "loss": 0.46,
+      "step": 3465
+    },
+    {
+      "epoch": 0.29684823569715657,
+      "grad_norm": 0.4522158205509186,
+      "learning_rate": 0.0001,
+      "loss": 1.0269,
+      "step": 3466
+    },
+    {
+      "epoch": 0.29693388146625554,
+      "grad_norm": 0.7362008690834045,
+      "learning_rate": 0.0001,
+      "loss": 1.33,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2970195272353546,
+      "grad_norm": 0.7437427639961243,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2971051730044536,
+      "grad_norm": 0.8866015672683716,
+      "learning_rate": 0.0001,
+      "loss": 1.067,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2971908187735526,
+      "grad_norm": 0.554294228553772,
+      "learning_rate": 0.0001,
+      "loss": 1.1658,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2972764645426516,
+      "grad_norm": 0.6605127453804016,
+      "learning_rate": 0.0001,
+      "loss": 0.7404,
+      "step": 3471
+    },
+    {
+      "epoch": 0.2973621103117506,
+      "grad_norm": 0.7691430449485779,
+      "learning_rate": 0.0001,
+      "loss": 0.7664,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2974477560808496,
+      "grad_norm": 0.7630181312561035,
+      "learning_rate": 0.0001,
+      "loss": 0.7944,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2975334018499486,
+      "grad_norm": 0.5747367739677429,
+      "learning_rate": 0.0001,
+      "loss": 0.5468,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2976190476190476,
+      "grad_norm": 0.7061007618904114,
+      "learning_rate": 0.0001,
+      "loss": 0.5722,
+      "step": 3475
+    },
+    {
+      "epoch": 0.29770469338814665,
+      "grad_norm": 0.9969678521156311,
+      "learning_rate": 0.0001,
+      "loss": 0.8437,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2977903391572456,
+      "grad_norm": 0.9336510300636292,
+      "learning_rate": 0.0001,
+      "loss": 1.1471,
+      "step": 3477
+    },
+    {
+      "epoch": 0.29787598492634465,
+      "grad_norm": 0.3857652246952057,
+      "learning_rate": 0.0001,
+      "loss": 0.7975,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2979616306954436,
+      "grad_norm": 0.7161991596221924,
+      "learning_rate": 0.0001,
+      "loss": 1.1243,
+      "step": 3479
+    },
+    {
+      "epoch": 0.29804727646454265,
+      "grad_norm": 0.6520877480506897,
+      "learning_rate": 0.0001,
+      "loss": 1.0896,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2981329222336417,
+      "grad_norm": 0.6157839298248291,
+      "learning_rate": 0.0001,
+      "loss": 1.2345,
+      "step": 3481
+    },
+    {
+      "epoch": 0.29821856800274066,
+      "grad_norm": 0.8407564163208008,
+      "learning_rate": 0.0001,
+      "loss": 1.3287,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2983042137718397,
+      "grad_norm": 0.5692492127418518,
+      "learning_rate": 0.0001,
+      "loss": 0.8462,
+      "step": 3483
+    },
+    {
+      "epoch": 0.29838985954093866,
+      "grad_norm": 0.8280079960823059,
+      "learning_rate": 0.0001,
+      "loss": 1.1422,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2984755053100377,
+      "grad_norm": 0.8274717330932617,
+      "learning_rate": 0.0001,
+      "loss": 0.9114,
+      "step": 3485
+    },
+    {
+      "epoch": 0.29856115107913667,
+      "grad_norm": 0.759829580783844,
+      "learning_rate": 0.0001,
+      "loss": 0.968,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2986467968482357,
+      "grad_norm": 0.47255805134773254,
+      "learning_rate": 0.0001,
+      "loss": 1.2086,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2987324426173347,
+      "grad_norm": 0.9635027647018433,
+      "learning_rate": 0.0001,
+      "loss": 0.8035,
+      "step": 3488
+    },
+    {
+      "epoch": 0.2988180883864337,
+      "grad_norm": 0.5073122978210449,
+      "learning_rate": 0.0001,
+      "loss": 0.9248,
+      "step": 3489
+    },
+    {
+      "epoch": 0.29890373415553273,
+      "grad_norm": 0.9469984769821167,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2989893799246317,
+      "grad_norm": 0.44273141026496887,
+      "learning_rate": 0.0001,
+      "loss": 1.0419,
+      "step": 3491
+    },
+    {
+      "epoch": 0.29907502569373073,
+      "grad_norm": 0.7623530030250549,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 3492
+    },
+    {
+      "epoch": 0.29916067146282976,
+      "grad_norm": 0.7210238575935364,
+      "learning_rate": 0.0001,
+      "loss": 0.8594,
+      "step": 3493
+    },
+    {
+      "epoch": 0.29924631723192874,
+      "grad_norm": 0.4530341625213623,
+      "learning_rate": 0.0001,
+      "loss": 0.8258,
+      "step": 3494
+    },
+    {
+      "epoch": 0.29933196300102777,
+      "grad_norm": 0.5161679983139038,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 3495
+    },
+    {
+      "epoch": 0.29941760877012674,
+      "grad_norm": 0.8118231892585754,
+      "learning_rate": 0.0001,
+      "loss": 0.7719,
+      "step": 3496
+    },
+    {
+      "epoch": 0.29950325453922577,
+      "grad_norm": 0.6741995811462402,
+      "learning_rate": 0.0001,
+      "loss": 0.7975,
+      "step": 3497
+    },
+    {
+      "epoch": 0.29958890030832475,
+      "grad_norm": 0.1970827877521515,
+      "learning_rate": 0.0001,
+      "loss": 0.9176,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2996745460774238,
+      "grad_norm": 0.8698933124542236,
+      "learning_rate": 0.0001,
+      "loss": 0.7817,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2997601918465228,
+      "grad_norm": 0.6597597002983093,
+      "learning_rate": 0.0001,
+      "loss": 0.8451,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2998458376156218,
+      "grad_norm": 0.44999632239341736,
+      "learning_rate": 0.0001,
+      "loss": 0.7894,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2999314833847208,
+      "grad_norm": 0.8150225281715393,
+      "learning_rate": 0.0001,
+      "loss": 1.2986,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3000171291538198,
+      "grad_norm": 0.7894243597984314,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3001027749229188,
+      "grad_norm": 0.5680282115936279,
+      "learning_rate": 0.0001,
+      "loss": 0.7807,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3001884206920178,
+      "grad_norm": 0.43767330050468445,
+      "learning_rate": 0.0001,
+      "loss": 1.0954,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3002740664611168,
+      "grad_norm": 0.4981527626514435,
+      "learning_rate": 0.0001,
+      "loss": 0.867,
+      "step": 3506
+    },
+    {
+      "epoch": 0.30035971223021585,
+      "grad_norm": 0.8080961108207703,
+      "learning_rate": 0.0001,
+      "loss": 0.719,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3004453579993148,
+      "grad_norm": 1.207919716835022,
+      "learning_rate": 0.0001,
+      "loss": 1.3316,
+      "step": 3508
+    },
+    {
+      "epoch": 0.30053100376841385,
+      "grad_norm": 0.7896754741668701,
+      "learning_rate": 0.0001,
+      "loss": 1.1736,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3006166495375128,
+      "grad_norm": 0.8684563040733337,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 3510
+    },
+    {
+      "epoch": 0.30070229530661186,
+      "grad_norm": 0.7705377340316772,
+      "learning_rate": 0.0001,
+      "loss": 0.5829,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3007879410757109,
+      "grad_norm": 0.8800756335258484,
+      "learning_rate": 0.0001,
+      "loss": 0.6819,
+      "step": 3512
+    },
+    {
+      "epoch": 0.30087358684480986,
+      "grad_norm": 0.6221142411231995,
+      "learning_rate": 0.0001,
+      "loss": 0.7087,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3009592326139089,
+      "grad_norm": 0.6022440791130066,
+      "learning_rate": 0.0001,
+      "loss": 1.2001,
+      "step": 3514
+    },
+    {
+      "epoch": 0.30104487838300786,
+      "grad_norm": 0.6525847911834717,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3011305241521069,
+      "grad_norm": 1.0390877723693848,
+      "learning_rate": 0.0001,
+      "loss": 0.8299,
+      "step": 3516
+    },
+    {
+      "epoch": 0.30121616992120587,
+      "grad_norm": 0.21614256501197815,
+      "learning_rate": 0.0001,
+      "loss": 0.8536,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3013018156903049,
+      "grad_norm": 0.5654253363609314,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3013874614594039,
+      "grad_norm": 0.4631986916065216,
+      "learning_rate": 0.0001,
+      "loss": 0.4574,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3014731072285029,
+      "grad_norm": 1.5328978300094604,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 3520
+    },
+    {
+      "epoch": 0.30155875299760193,
+      "grad_norm": 0.575782299041748,
+      "learning_rate": 0.0001,
+      "loss": 0.9619,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3016443987667009,
+      "grad_norm": 0.6897450685501099,
+      "learning_rate": 0.0001,
+      "loss": 1.0339,
+      "step": 3522
+    },
+    {
+      "epoch": 0.30173004453579994,
+      "grad_norm": 0.8311665058135986,
+      "learning_rate": 0.0001,
+      "loss": 0.8128,
+      "step": 3523
+    },
+    {
+      "epoch": 0.30181569030489896,
+      "grad_norm": 0.4948771893978119,
+      "learning_rate": 0.0001,
+      "loss": 0.5104,
+      "step": 3524
+    },
+    {
+      "epoch": 0.30190133607399794,
+      "grad_norm": 0.287619024515152,
+      "learning_rate": 0.0001,
+      "loss": 0.4474,
+      "step": 3525
+    },
+    {
+      "epoch": 0.30198698184309697,
+      "grad_norm": 0.853660523891449,
+      "learning_rate": 0.0001,
+      "loss": 1.8569,
+      "step": 3526
+    },
+    {
+      "epoch": 0.30207262761219594,
+      "grad_norm": 1.094225525856018,
+      "learning_rate": 0.0001,
+      "loss": 1.2469,
+      "step": 3527
+    },
+    {
+      "epoch": 0.302158273381295,
+      "grad_norm": 1.1731698513031006,
+      "learning_rate": 0.0001,
+      "loss": 0.8789,
+      "step": 3528
+    },
+    {
+      "epoch": 0.30224391915039395,
+      "grad_norm": 0.6542361378669739,
+      "learning_rate": 0.0001,
+      "loss": 0.627,
+      "step": 3529
+    },
+    {
+      "epoch": 0.302329564919493,
+      "grad_norm": 0.772493302822113,
+      "learning_rate": 0.0001,
+      "loss": 1.1848,
+      "step": 3530
+    },
+    {
+      "epoch": 0.302415210688592,
+      "grad_norm": 0.5317011475563049,
+      "learning_rate": 0.0001,
+      "loss": 0.6341,
+      "step": 3531
+    },
+    {
+      "epoch": 0.302500856457691,
+      "grad_norm": 0.6810627579689026,
+      "learning_rate": 0.0001,
+      "loss": 0.6919,
+      "step": 3532
+    },
+    {
+      "epoch": 0.30258650222679,
+      "grad_norm": 0.5998390913009644,
+      "learning_rate": 0.0001,
+      "loss": 0.8212,
+      "step": 3533
+    },
+    {
+      "epoch": 0.302672147995889,
+      "grad_norm": 0.6813536882400513,
+      "learning_rate": 0.0001,
+      "loss": 2.609,
+      "step": 3534
+    },
+    {
+      "epoch": 0.302757793764988,
+      "grad_norm": 0.8369669318199158,
+      "learning_rate": 0.0001,
+      "loss": 1.2941,
+      "step": 3535
+    },
+    {
+      "epoch": 0.302843439534087,
+      "grad_norm": 0.5256216526031494,
+      "learning_rate": 0.0001,
+      "loss": 0.761,
+      "step": 3536
+    },
+    {
+      "epoch": 0.302929085303186,
+      "grad_norm": 0.825709879398346,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 3537
+    },
+    {
+      "epoch": 0.30301473107228505,
+      "grad_norm": 0.7418362498283386,
+      "learning_rate": 0.0001,
+      "loss": 0.7214,
+      "step": 3538
+    },
+    {
+      "epoch": 0.303100376841384,
+      "grad_norm": 0.7957780361175537,
+      "learning_rate": 0.0001,
+      "loss": 0.9825,
+      "step": 3539
+    },
+    {
+      "epoch": 0.30318602261048305,
+      "grad_norm": 0.9832172393798828,
+      "learning_rate": 0.0001,
+      "loss": 0.8048,
+      "step": 3540
+    },
+    {
+      "epoch": 0.303271668379582,
+      "grad_norm": 0.5729221105575562,
+      "learning_rate": 0.0001,
+      "loss": 0.8791,
+      "step": 3541
+    },
+    {
+      "epoch": 0.30335731414868106,
+      "grad_norm": 0.7211901545524597,
+      "learning_rate": 0.0001,
+      "loss": 1.2676,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3034429599177801,
+      "grad_norm": 0.6072773337364197,
+      "learning_rate": 0.0001,
+      "loss": 1.2732,
+      "step": 3543
+    },
+    {
+      "epoch": 0.30352860568687906,
+      "grad_norm": 0.7346463203430176,
+      "learning_rate": 0.0001,
+      "loss": 1.1882,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3036142514559781,
+      "grad_norm": 0.6986691951751709,
+      "learning_rate": 0.0001,
+      "loss": 0.6915,
+      "step": 3545
+    },
+    {
+      "epoch": 0.30369989722507706,
+      "grad_norm": 1.0081342458724976,
+      "learning_rate": 0.0001,
+      "loss": 1.24,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3037855429941761,
+      "grad_norm": 0.5814337134361267,
+      "learning_rate": 0.0001,
+      "loss": 1.0963,
+      "step": 3547
+    },
+    {
+      "epoch": 0.30387118876327507,
+      "grad_norm": 0.4785625636577606,
+      "learning_rate": 0.0001,
+      "loss": 0.8001,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3039568345323741,
+      "grad_norm": 0.9172512888908386,
+      "learning_rate": 0.0001,
+      "loss": 1.1955,
+      "step": 3549
+    },
+    {
+      "epoch": 0.30404248030147313,
+      "grad_norm": 0.516446590423584,
+      "learning_rate": 0.0001,
+      "loss": 1.2624,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3041281260705721,
+      "grad_norm": 0.6678693890571594,
+      "learning_rate": 0.0001,
+      "loss": 0.3914,
+      "step": 3551
+    },
+    {
+      "epoch": 0.30421377183967113,
+      "grad_norm": 0.8318419456481934,
+      "learning_rate": 0.0001,
+      "loss": 1.2274,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3042994176087701,
+      "grad_norm": 0.7440193891525269,
+      "learning_rate": 0.0001,
+      "loss": 1.1964,
+      "step": 3553
+    },
+    {
+      "epoch": 0.30438506337786914,
+      "grad_norm": 1.1463382244110107,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 3554
+    },
+    {
+      "epoch": 0.30447070914696817,
+      "grad_norm": 0.830697238445282,
+      "learning_rate": 0.0001,
+      "loss": 0.9218,
+      "step": 3555
+    },
+    {
+      "epoch": 0.30455635491606714,
+      "grad_norm": 0.8912988901138306,
+      "learning_rate": 0.0001,
+      "loss": 1.0353,
+      "step": 3556
+    },
+    {
+      "epoch": 0.30464200068516617,
+      "grad_norm": 0.8989576697349548,
+      "learning_rate": 0.0001,
+      "loss": 0.8648,
+      "step": 3557
+    },
+    {
+      "epoch": 0.30472764645426514,
+      "grad_norm": 0.22979995608329773,
+      "learning_rate": 0.0001,
+      "loss": 0.571,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3048132922233642,
+      "grad_norm": 0.9936613440513611,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 3559
+    },
+    {
+      "epoch": 0.30489893799246315,
+      "grad_norm": 0.5507964491844177,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3049845837615622,
+      "grad_norm": 0.7809141874313354,
+      "learning_rate": 0.0001,
+      "loss": 0.8097,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3050702295306612,
+      "grad_norm": 0.48809581995010376,
+      "learning_rate": 0.0001,
+      "loss": 0.7362,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3051558752997602,
+      "grad_norm": 0.4376184344291687,
+      "learning_rate": 0.0001,
+      "loss": 0.7214,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3052415210688592,
+      "grad_norm": 0.9387377500534058,
+      "learning_rate": 0.0001,
+      "loss": 0.6668,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3053271668379582,
+      "grad_norm": 0.9207602739334106,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3054128126070572,
+      "grad_norm": 0.6674703359603882,
+      "learning_rate": 0.0001,
+      "loss": 0.9026,
+      "step": 3566
+    },
+    {
+      "epoch": 0.30549845837615625,
+      "grad_norm": 0.4968853294849396,
+      "learning_rate": 0.0001,
+      "loss": 0.6197,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3055841041452552,
+      "grad_norm": 1.2805712223052979,
+      "learning_rate": 0.0001,
+      "loss": 1.1345,
+      "step": 3568
+    },
+    {
+      "epoch": 0.30566974991435425,
+      "grad_norm": 0.6311193704605103,
+      "learning_rate": 0.0001,
+      "loss": 0.7731,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3057553956834532,
+      "grad_norm": 0.5524489879608154,
+      "learning_rate": 0.0001,
+      "loss": 0.72,
+      "step": 3570
+    },
+    {
+      "epoch": 0.30584104145255225,
+      "grad_norm": 0.3157130181789398,
+      "learning_rate": 0.0001,
+      "loss": 0.432,
+      "step": 3571
+    },
+    {
+      "epoch": 0.30592668722165123,
+      "grad_norm": 0.898093044757843,
+      "learning_rate": 0.0001,
+      "loss": 1.1632,
+      "step": 3572
+    },
+    {
+      "epoch": 0.30601233299075026,
+      "grad_norm": 0.7655217051506042,
+      "learning_rate": 0.0001,
+      "loss": 0.6417,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3060979787598493,
+      "grad_norm": 0.970911979675293,
+      "learning_rate": 0.0001,
+      "loss": 0.9529,
+      "step": 3574
+    },
+    {
+      "epoch": 0.30618362452894826,
+      "grad_norm": 0.7306068539619446,
+      "learning_rate": 0.0001,
+      "loss": 0.8764,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3062692702980473,
+      "grad_norm": 0.5369967222213745,
+      "learning_rate": 0.0001,
+      "loss": 0.6921,
+      "step": 3576
+    },
+    {
+      "epoch": 0.30635491606714627,
+      "grad_norm": 0.5995248556137085,
+      "learning_rate": 0.0001,
+      "loss": 0.9466,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3064405618362453,
+      "grad_norm": 0.7149795889854431,
+      "learning_rate": 0.0001,
+      "loss": 0.4966,
+      "step": 3578
+    },
+    {
+      "epoch": 0.30652620760534427,
+      "grad_norm": 1.1679513454437256,
+      "learning_rate": 0.0001,
+      "loss": 2.1295,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3066118533744433,
+      "grad_norm": 0.7388948798179626,
+      "learning_rate": 0.0001,
+      "loss": 1.1292,
+      "step": 3580
+    },
+    {
+      "epoch": 0.30669749914354233,
+      "grad_norm": 0.8412114381790161,
+      "learning_rate": 0.0001,
+      "loss": 1.185,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3067831449126413,
+      "grad_norm": 0.8203011751174927,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 3582
+    },
+    {
+      "epoch": 0.30686879068174033,
+      "grad_norm": 0.5874364972114563,
+      "learning_rate": 0.0001,
+      "loss": 1.0744,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3069544364508393,
+      "grad_norm": 0.671749472618103,
+      "learning_rate": 0.0001,
+      "loss": 1.2714,
+      "step": 3584
+    },
+    {
+      "epoch": 0.30704008221993834,
+      "grad_norm": 0.6547632813453674,
+      "learning_rate": 0.0001,
+      "loss": 1.1946,
+      "step": 3585
+    },
+    {
+      "epoch": 0.30712572798903737,
+      "grad_norm": 0.7425119280815125,
+      "learning_rate": 0.0001,
+      "loss": 0.9471,
+      "step": 3586
+    },
+    {
+      "epoch": 0.30721137375813634,
+      "grad_norm": 0.43978217244148254,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 3587
+    },
+    {
+      "epoch": 0.30729701952723537,
+      "grad_norm": 0.4663841724395752,
+      "learning_rate": 0.0001,
+      "loss": 0.5888,
+      "step": 3588
+    },
+    {
+      "epoch": 0.30738266529633435,
+      "grad_norm": 0.47186121344566345,
+      "learning_rate": 0.0001,
+      "loss": 0.8373,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3074683110654334,
+      "grad_norm": 0.36277490854263306,
+      "learning_rate": 0.0001,
+      "loss": 0.5673,
+      "step": 3590
+    },
+    {
+      "epoch": 0.30755395683453235,
+      "grad_norm": 0.8434424996376038,
+      "learning_rate": 0.0001,
+      "loss": 1.2531,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3076396026036314,
+      "grad_norm": 0.6882156729698181,
+      "learning_rate": 0.0001,
+      "loss": 0.8128,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3077252483727304,
+      "grad_norm": 0.5674391984939575,
+      "learning_rate": 0.0001,
+      "loss": 0.426,
+      "step": 3593
+    },
+    {
+      "epoch": 0.3078108941418294,
+      "grad_norm": 0.6983845829963684,
+      "learning_rate": 0.0001,
+      "loss": 0.6465,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3078965399109284,
+      "grad_norm": 0.44205209612846375,
+      "learning_rate": 0.0001,
+      "loss": 0.6559,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3079821856800274,
+      "grad_norm": 0.5453655123710632,
+      "learning_rate": 0.0001,
+      "loss": 0.3974,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3080678314491264,
+      "grad_norm": 0.4549030661582947,
+      "learning_rate": 0.0001,
+      "loss": 0.3041,
+      "step": 3597
+    },
+    {
+      "epoch": 0.30815347721822545,
+      "grad_norm": 0.7134392261505127,
+      "learning_rate": 0.0001,
+      "loss": 0.6338,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3082391229873244,
+      "grad_norm": 0.5644324421882629,
+      "learning_rate": 0.0001,
+      "loss": 1.0345,
+      "step": 3599
+    },
+    {
+      "epoch": 0.30832476875642345,
+      "grad_norm": 0.8180065155029297,
+      "learning_rate": 0.0001,
+      "loss": 0.6017,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3084104145255224,
+      "grad_norm": 1.0156810283660889,
+      "learning_rate": 0.0001,
+      "loss": 0.9485,
+      "step": 3601
+    },
+    {
+      "epoch": 0.30849606029462145,
+      "grad_norm": 0.6007433533668518,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 3602
+    },
+    {
+      "epoch": 0.30858170606372043,
+      "grad_norm": 0.6276346445083618,
+      "learning_rate": 0.0001,
+      "loss": 1.2693,
+      "step": 3603
+    },
+    {
+      "epoch": 0.30866735183281946,
+      "grad_norm": 0.8717529773712158,
+      "learning_rate": 0.0001,
+      "loss": 0.9402,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3087529976019185,
+      "grad_norm": 0.6992687582969666,
+      "learning_rate": 0.0001,
+      "loss": 0.8515,
+      "step": 3605
+    },
+    {
+      "epoch": 0.30883864337101746,
+      "grad_norm": 0.6992174386978149,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3089242891401165,
+      "grad_norm": 0.47502705454826355,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 3607
+    },
+    {
+      "epoch": 0.30900993490921547,
+      "grad_norm": 0.8907037377357483,
+      "learning_rate": 0.0001,
+      "loss": 1.1011,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3090955806783145,
+      "grad_norm": 0.7458038330078125,
+      "learning_rate": 0.0001,
+      "loss": 1.0082,
+      "step": 3609
+    },
+    {
+      "epoch": 0.30918122644741347,
+      "grad_norm": 0.771414577960968,
+      "learning_rate": 0.0001,
+      "loss": 0.8739,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3092668722165125,
+      "grad_norm": 0.9047467708587646,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 3611
+    },
+    {
+      "epoch": 0.30935251798561153,
+      "grad_norm": 0.6717650890350342,
+      "learning_rate": 0.0001,
+      "loss": 1.0625,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3094381637547105,
+      "grad_norm": 0.9382766485214233,
+      "learning_rate": 0.0001,
+      "loss": 0.7404,
+      "step": 3613
+    },
+    {
+      "epoch": 0.30952380952380953,
+      "grad_norm": 0.8774073719978333,
+      "learning_rate": 0.0001,
+      "loss": 0.8633,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3096094552929085,
+      "grad_norm": 0.9392250776290894,
+      "learning_rate": 0.0001,
+      "loss": 0.6296,
+      "step": 3615
+    },
+    {
+      "epoch": 0.30969510106200754,
+      "grad_norm": 0.7033014893531799,
+      "learning_rate": 0.0001,
+      "loss": 0.4793,
+      "step": 3616
+    },
+    {
+      "epoch": 0.30978074683110657,
+      "grad_norm": 0.9049376249313354,
+      "learning_rate": 0.0001,
+      "loss": 1.0724,
+      "step": 3617
+    },
+    {
+      "epoch": 0.30986639260020554,
+      "grad_norm": 0.7775086760520935,
+      "learning_rate": 0.0001,
+      "loss": 0.8238,
+      "step": 3618
+    },
+    {
+      "epoch": 0.30995203836930457,
+      "grad_norm": 0.7605637907981873,
+      "learning_rate": 0.0001,
+      "loss": 1.2011,
+      "step": 3619
+    },
+    {
+      "epoch": 0.31003768413840355,
+      "grad_norm": 0.5035964250564575,
+      "learning_rate": 0.0001,
+      "loss": 0.8561,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3101233299075026,
+      "grad_norm": 0.6520989537239075,
+      "learning_rate": 0.0001,
+      "loss": 0.6263,
+      "step": 3621
+    },
+    {
+      "epoch": 0.31020897567660155,
+      "grad_norm": 0.5161353945732117,
+      "learning_rate": 0.0001,
+      "loss": 0.8295,
+      "step": 3622
+    },
+    {
+      "epoch": 0.3102946214457006,
+      "grad_norm": 0.9571381211280823,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3103802672147996,
+      "grad_norm": 0.7427163124084473,
+      "learning_rate": 0.0001,
+      "loss": 1.0649,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3104659129838986,
+      "grad_norm": 0.5890627503395081,
+      "learning_rate": 0.0001,
+      "loss": 1.2678,
+      "step": 3625
+    },
+    {
+      "epoch": 0.3105515587529976,
+      "grad_norm": 1.2415893077850342,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3106372045220966,
+      "grad_norm": 0.8589575290679932,
+      "learning_rate": 0.0001,
+      "loss": 0.8258,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3107228502911956,
+      "grad_norm": 0.5650753974914551,
+      "learning_rate": 0.0001,
+      "loss": 1.1234,
+      "step": 3628
+    },
+    {
+      "epoch": 0.31080849606029465,
+      "grad_norm": 1.0353001356124878,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3108941418293936,
+      "grad_norm": 0.9908233284950256,
+      "learning_rate": 0.0001,
+      "loss": 0.987,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31097978759849265,
+      "grad_norm": 0.600673496723175,
+      "learning_rate": 0.0001,
+      "loss": 0.5391,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3110654333675916,
+      "grad_norm": 1.0430487394332886,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 3632
+    },
+    {
+      "epoch": 0.31115107913669066,
+      "grad_norm": 0.8626402020454407,
+      "learning_rate": 0.0001,
+      "loss": 0.8926,
+      "step": 3633
+    },
+    {
+      "epoch": 0.31123672490578963,
+      "grad_norm": 0.604469895362854,
+      "learning_rate": 0.0001,
+      "loss": 1.3179,
+      "step": 3634
+    },
+    {
+      "epoch": 0.31132237067488866,
+      "grad_norm": 0.6403637528419495,
+      "learning_rate": 0.0001,
+      "loss": 0.415,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3114080164439877,
+      "grad_norm": 0.582991898059845,
+      "learning_rate": 0.0001,
+      "loss": 1.2381,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31149366221308666,
+      "grad_norm": 0.4791080951690674,
+      "learning_rate": 0.0001,
+      "loss": 0.7821,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3115793079821857,
+      "grad_norm": 1.068909764289856,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 3638
+    },
+    {
+      "epoch": 0.31166495375128467,
+      "grad_norm": 0.12282731384038925,
+      "learning_rate": 0.0001,
+      "loss": 0.3808,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3117505995203837,
+      "grad_norm": 0.5277617573738098,
+      "learning_rate": 0.0001,
+      "loss": 0.4891,
+      "step": 3640
+    },
+    {
+      "epoch": 0.31183624528948267,
+      "grad_norm": 0.4984751045703888,
+      "learning_rate": 0.0001,
+      "loss": 0.8137,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3119218910585817,
+      "grad_norm": 0.4361402690410614,
+      "learning_rate": 0.0001,
+      "loss": 0.7338,
+      "step": 3642
+    },
+    {
+      "epoch": 0.31200753682768073,
+      "grad_norm": 0.45227041840553284,
+      "learning_rate": 0.0001,
+      "loss": 0.6765,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3120931825967797,
+      "grad_norm": 0.5058034658432007,
+      "learning_rate": 0.0001,
+      "loss": 0.5545,
+      "step": 3644
+    },
+    {
+      "epoch": 0.31217882836587874,
+      "grad_norm": 1.4158971309661865,
+      "learning_rate": 0.0001,
+      "loss": 0.7605,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3122644741349777,
+      "grad_norm": 0.31450825929641724,
+      "learning_rate": 0.0001,
+      "loss": 0.22,
+      "step": 3646
+    },
+    {
+      "epoch": 0.31235011990407674,
+      "grad_norm": 0.6041990518569946,
+      "learning_rate": 0.0001,
+      "loss": 0.9587,
+      "step": 3647
+    },
+    {
+      "epoch": 0.31243576567317577,
+      "grad_norm": 0.9062447547912598,
+      "learning_rate": 0.0001,
+      "loss": 1.1642,
+      "step": 3648
+    },
+    {
+      "epoch": 0.31252141144227474,
+      "grad_norm": 1.002585768699646,
+      "learning_rate": 0.0001,
+      "loss": 0.8689,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3126070572113738,
+      "grad_norm": 0.8955001831054688,
+      "learning_rate": 0.0001,
+      "loss": 1.8695,
+      "step": 3650
+    },
+    {
+      "epoch": 0.31269270298047275,
+      "grad_norm": 0.7442821860313416,
+      "learning_rate": 0.0001,
+      "loss": 0.9174,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3127783487495718,
+      "grad_norm": 0.24285157024860382,
+      "learning_rate": 0.0001,
+      "loss": 0.4539,
+      "step": 3652
+    },
+    {
+      "epoch": 0.31286399451867075,
+      "grad_norm": 0.5099822878837585,
+      "learning_rate": 0.0001,
+      "loss": 0.9419,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3129496402877698,
+      "grad_norm": 0.600045382976532,
+      "learning_rate": 0.0001,
+      "loss": 1.3377,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3130352860568688,
+      "grad_norm": 0.6149832010269165,
+      "learning_rate": 0.0001,
+      "loss": 0.9757,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3131209318259678,
+      "grad_norm": 0.575124204158783,
+      "learning_rate": 0.0001,
+      "loss": 0.6875,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3132065775950668,
+      "grad_norm": 0.48899710178375244,
+      "learning_rate": 0.0001,
+      "loss": 0.314,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3132922233641658,
+      "grad_norm": 0.6116422414779663,
+      "learning_rate": 0.0001,
+      "loss": 0.6332,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3133778691332648,
+      "grad_norm": 0.7413456439971924,
+      "learning_rate": 0.0001,
+      "loss": 1.0951,
+      "step": 3659
+    },
+    {
+      "epoch": 0.31346351490236385,
+      "grad_norm": 0.9466413855552673,
+      "learning_rate": 0.0001,
+      "loss": 0.7807,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3135491606714628,
+      "grad_norm": 0.22997193038463593,
+      "learning_rate": 0.0001,
+      "loss": 0.1966,
+      "step": 3661
+    },
+    {
+      "epoch": 0.31363480644056185,
+      "grad_norm": 0.7447431087493896,
+      "learning_rate": 0.0001,
+      "loss": 0.9401,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3137204522096608,
+      "grad_norm": 0.8448193669319153,
+      "learning_rate": 0.0001,
+      "loss": 1.1721,
+      "step": 3663
+    },
+    {
+      "epoch": 0.31380609797875986,
+      "grad_norm": 0.569908618927002,
+      "learning_rate": 0.0001,
+      "loss": 1.2214,
+      "step": 3664
+    },
+    {
+      "epoch": 0.31389174374785883,
+      "grad_norm": 1.8450249433517456,
+      "learning_rate": 0.0001,
+      "loss": 1.0949,
+      "step": 3665
+    },
+    {
+      "epoch": 0.31397738951695786,
+      "grad_norm": 0.635596513748169,
+      "learning_rate": 0.0001,
+      "loss": 0.8415,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3140630352860569,
+      "grad_norm": 0.5139503479003906,
+      "learning_rate": 0.0001,
+      "loss": 0.9311,
+      "step": 3667
+    },
+    {
+      "epoch": 0.31414868105515587,
+      "grad_norm": 1.091674566268921,
+      "learning_rate": 0.0001,
+      "loss": 0.856,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3142343268242549,
+      "grad_norm": 1.1924262046813965,
+      "learning_rate": 0.0001,
+      "loss": 1.4221,
+      "step": 3669
+    },
+    {
+      "epoch": 0.31431997259335387,
+      "grad_norm": 0.6868544816970825,
+      "learning_rate": 0.0001,
+      "loss": 1.22,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3144056183624529,
+      "grad_norm": 0.9702930450439453,
+      "learning_rate": 0.0001,
+      "loss": 0.7669,
+      "step": 3671
+    },
+    {
+      "epoch": 0.31449126413155193,
+      "grad_norm": 0.7182427048683167,
+      "learning_rate": 0.0001,
+      "loss": 0.8275,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3145769099006509,
+      "grad_norm": 0.9181588292121887,
+      "learning_rate": 0.0001,
+      "loss": 1.2945,
+      "step": 3673
+    },
+    {
+      "epoch": 0.31466255566974993,
+      "grad_norm": 0.6148859858512878,
+      "learning_rate": 0.0001,
+      "loss": 1.1148,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3147482014388489,
+      "grad_norm": 0.7504810690879822,
+      "learning_rate": 0.0001,
+      "loss": 0.6357,
+      "step": 3675
+    },
+    {
+      "epoch": 0.31483384720794794,
+      "grad_norm": 0.8354966044425964,
+      "learning_rate": 0.0001,
+      "loss": 0.6923,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3149194929770469,
+      "grad_norm": 0.48352640867233276,
+      "learning_rate": 0.0001,
+      "loss": 0.7378,
+      "step": 3677
+    },
+    {
+      "epoch": 0.31500513874614594,
+      "grad_norm": 0.7618727087974548,
+      "learning_rate": 0.0001,
+      "loss": 0.8132,
+      "step": 3678
+    },
+    {
+      "epoch": 0.31509078451524497,
+      "grad_norm": 0.3741157352924347,
+      "learning_rate": 0.0001,
+      "loss": 0.551,
+      "step": 3679
+    },
+    {
+      "epoch": 0.31517643028434394,
+      "grad_norm": 0.6182940006256104,
+      "learning_rate": 0.0001,
+      "loss": 0.5911,
+      "step": 3680
+    },
+    {
+      "epoch": 0.315262076053443,
+      "grad_norm": 0.5289412140846252,
+      "learning_rate": 0.0001,
+      "loss": 1.2283,
+      "step": 3681
+    },
+    {
+      "epoch": 0.31534772182254195,
+      "grad_norm": 0.40847837924957275,
+      "learning_rate": 0.0001,
+      "loss": 0.6634,
+      "step": 3682
+    },
+    {
+      "epoch": 0.315433367591641,
+      "grad_norm": 0.6680887937545776,
+      "learning_rate": 0.0001,
+      "loss": 0.6959,
+      "step": 3683
+    },
+    {
+      "epoch": 0.31551901336073995,
+      "grad_norm": 0.5459215044975281,
+      "learning_rate": 0.0001,
+      "loss": 1.2106,
+      "step": 3684
+    },
+    {
+      "epoch": 0.315604659129839,
+      "grad_norm": 0.4102620780467987,
+      "learning_rate": 0.0001,
+      "loss": 0.8328,
+      "step": 3685
+    },
+    {
+      "epoch": 0.315690304898938,
+      "grad_norm": 0.7284419536590576,
+      "learning_rate": 0.0001,
+      "loss": 1.2386,
+      "step": 3686
+    },
+    {
+      "epoch": 0.315775950668037,
+      "grad_norm": 0.791623592376709,
+      "learning_rate": 0.0001,
+      "loss": 0.6935,
+      "step": 3687
+    },
+    {
+      "epoch": 0.315861596437136,
+      "grad_norm": 0.8076296448707581,
+      "learning_rate": 0.0001,
+      "loss": 1.3295,
+      "step": 3688
+    },
+    {
+      "epoch": 0.315947242206235,
+      "grad_norm": 1.0915331840515137,
+      "learning_rate": 0.0001,
+      "loss": 0.9574,
+      "step": 3689
+    },
+    {
+      "epoch": 0.316032887975334,
+      "grad_norm": 0.7402876615524292,
+      "learning_rate": 0.0001,
+      "loss": 0.7833,
+      "step": 3690
+    },
+    {
+      "epoch": 0.31611853374443305,
+      "grad_norm": 0.7304937839508057,
+      "learning_rate": 0.0001,
+      "loss": 0.763,
+      "step": 3691
+    },
+    {
+      "epoch": 0.316204179513532,
+      "grad_norm": 0.6177854537963867,
+      "learning_rate": 0.0001,
+      "loss": 0.6017,
+      "step": 3692
+    },
+    {
+      "epoch": 0.31628982528263105,
+      "grad_norm": 0.6845636963844299,
+      "learning_rate": 0.0001,
+      "loss": 0.69,
+      "step": 3693
+    },
+    {
+      "epoch": 0.31637547105173003,
+      "grad_norm": 0.5614719390869141,
+      "learning_rate": 0.0001,
+      "loss": 1.0684,
+      "step": 3694
+    },
+    {
+      "epoch": 0.31646111682082906,
+      "grad_norm": 0.5324113965034485,
+      "learning_rate": 0.0001,
+      "loss": 0.6905,
+      "step": 3695
+    },
+    {
+      "epoch": 0.31654676258992803,
+      "grad_norm": 0.8020150065422058,
+      "learning_rate": 0.0001,
+      "loss": 1.0711,
+      "step": 3696
+    },
+    {
+      "epoch": 0.31663240835902706,
+      "grad_norm": 0.5792848467826843,
+      "learning_rate": 0.0001,
+      "loss": 1.0119,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3167180541281261,
+      "grad_norm": 0.9131925106048584,
+      "learning_rate": 0.0001,
+      "loss": 1.1187,
+      "step": 3698
+    },
+    {
+      "epoch": 0.31680369989722507,
+      "grad_norm": 0.7345327734947205,
+      "learning_rate": 0.0001,
+      "loss": 0.8519,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3168893456663241,
+      "grad_norm": 0.9916974306106567,
+      "learning_rate": 0.0001,
+      "loss": 0.924,
+      "step": 3700
+    },
+    {
+      "epoch": 0.31697499143542307,
+      "grad_norm": 0.6918055415153503,
+      "learning_rate": 0.0001,
+      "loss": 0.5148,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3170606372045221,
+      "grad_norm": 0.5105330348014832,
+      "learning_rate": 0.0001,
+      "loss": 0.9509,
+      "step": 3702
+    },
+    {
+      "epoch": 0.31714628297362113,
+      "grad_norm": 0.9193611741065979,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3172319287427201,
+      "grad_norm": 0.6233958005905151,
+      "learning_rate": 0.0001,
+      "loss": 1.1429,
+      "step": 3704
+    },
+    {
+      "epoch": 0.31731757451181913,
+      "grad_norm": 0.5441810488700867,
+      "learning_rate": 0.0001,
+      "loss": 0.987,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3174032202809181,
+      "grad_norm": 0.5569643378257751,
+      "learning_rate": 0.0001,
+      "loss": 1.0452,
+      "step": 3706
+    },
+    {
+      "epoch": 0.31748886605001714,
+      "grad_norm": 0.6064330339431763,
+      "learning_rate": 0.0001,
+      "loss": 0.9687,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3175745118191161,
+      "grad_norm": 0.44742488861083984,
+      "learning_rate": 0.0001,
+      "loss": 1.2631,
+      "step": 3708
+    },
+    {
+      "epoch": 0.31766015758821514,
+      "grad_norm": 0.8212966918945312,
+      "learning_rate": 0.0001,
+      "loss": 0.9796,
+      "step": 3709
+    },
+    {
+      "epoch": 0.31774580335731417,
+      "grad_norm": 0.7089031338691711,
+      "learning_rate": 0.0001,
+      "loss": 1.0379,
+      "step": 3710
+    },
+    {
+      "epoch": 0.31783144912641315,
+      "grad_norm": 0.7585715651512146,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3179170948955122,
+      "grad_norm": 0.7831366658210754,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 3712
+    },
+    {
+      "epoch": 0.31800274066461115,
+      "grad_norm": 0.871525764465332,
+      "learning_rate": 0.0001,
+      "loss": 0.9928,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3180883864337102,
+      "grad_norm": 0.5307945013046265,
+      "learning_rate": 0.0001,
+      "loss": 0.4858,
+      "step": 3714
+    },
+    {
+      "epoch": 0.31817403220280915,
+      "grad_norm": 0.8107026219367981,
+      "learning_rate": 0.0001,
+      "loss": 0.8226,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3182596779719082,
+      "grad_norm": 0.6329191327095032,
+      "learning_rate": 0.0001,
+      "loss": 0.5118,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3183453237410072,
+      "grad_norm": 0.7691789269447327,
+      "learning_rate": 0.0001,
+      "loss": 0.7375,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3184309695101062,
+      "grad_norm": 0.953406035900116,
+      "learning_rate": 0.0001,
+      "loss": 1.0204,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3185166152792052,
+      "grad_norm": 0.7879107594490051,
+      "learning_rate": 0.0001,
+      "loss": 0.9145,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3186022610483042,
+      "grad_norm": 0.9656384587287903,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3186879068174032,
+      "grad_norm": 0.7235584855079651,
+      "learning_rate": 0.0001,
+      "loss": 0.6919,
+      "step": 3721
+    },
+    {
+      "epoch": 0.31877355258650225,
+      "grad_norm": 0.7868753671646118,
+      "learning_rate": 0.0001,
+      "loss": 0.983,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3188591983556012,
+      "grad_norm": 0.6900597214698792,
+      "learning_rate": 0.0001,
+      "loss": 1.0346,
+      "step": 3723
+    },
+    {
+      "epoch": 0.31894484412470026,
+      "grad_norm": 0.9670583605766296,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 3724
+    },
+    {
+      "epoch": 0.31903048989379923,
+      "grad_norm": 0.7796013951301575,
+      "learning_rate": 0.0001,
+      "loss": 1.049,
+      "step": 3725
+    },
+    {
+      "epoch": 0.31911613566289826,
+      "grad_norm": 0.8091456294059753,
+      "learning_rate": 0.0001,
+      "loss": 1.0961,
+      "step": 3726
+    },
+    {
+      "epoch": 0.31920178143199723,
+      "grad_norm": 0.8388510346412659,
+      "learning_rate": 0.0001,
+      "loss": 1.1967,
+      "step": 3727
+    },
+    {
+      "epoch": 0.31928742720109626,
+      "grad_norm": 0.3916582763195038,
+      "learning_rate": 0.0001,
+      "loss": 0.6764,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3193730729701953,
+      "grad_norm": 0.851215124130249,
+      "learning_rate": 0.0001,
+      "loss": 0.7575,
+      "step": 3729
+    },
+    {
+      "epoch": 0.31945871873929427,
+      "grad_norm": 0.6986595392227173,
+      "learning_rate": 0.0001,
+      "loss": 0.4201,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3195443645083933,
+      "grad_norm": 1.0586532354354858,
+      "learning_rate": 0.0001,
+      "loss": 1.3994,
+      "step": 3731
+    },
+    {
+      "epoch": 0.31963001027749227,
+      "grad_norm": 0.3230210244655609,
+      "learning_rate": 0.0001,
+      "loss": 0.4887,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3197156560465913,
+      "grad_norm": 0.8453580141067505,
+      "learning_rate": 0.0001,
+      "loss": 0.8489,
+      "step": 3733
+    },
+    {
+      "epoch": 0.31980130181569033,
+      "grad_norm": 0.48909544944763184,
+      "learning_rate": 0.0001,
+      "loss": 0.9304,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3198869475847893,
+      "grad_norm": 0.9039849042892456,
+      "learning_rate": 0.0001,
+      "loss": 1.1778,
+      "step": 3735
+    },
+    {
+      "epoch": 0.31997259335388833,
+      "grad_norm": 1.095834732055664,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3200582391229873,
+      "grad_norm": 0.8143795728683472,
+      "learning_rate": 0.0001,
+      "loss": 0.8126,
+      "step": 3737
+    },
+    {
+      "epoch": 0.32014388489208634,
+      "grad_norm": 0.6612892746925354,
+      "learning_rate": 0.0001,
+      "loss": 0.8215,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3202295306611853,
+      "grad_norm": 0.6733485460281372,
+      "learning_rate": 0.0001,
+      "loss": 0.6989,
+      "step": 3739
+    },
+    {
+      "epoch": 0.32031517643028434,
+      "grad_norm": 0.5490016937255859,
+      "learning_rate": 0.0001,
+      "loss": 1.1166,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3204008221993834,
+      "grad_norm": 0.3255758583545685,
+      "learning_rate": 0.0001,
+      "loss": 0.4791,
+      "step": 3741
+    },
+    {
+      "epoch": 0.32048646796848235,
+      "grad_norm": 0.7885031700134277,
+      "learning_rate": 0.0001,
+      "loss": 0.6232,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3205721137375814,
+      "grad_norm": 0.8022674322128296,
+      "learning_rate": 0.0001,
+      "loss": 0.8424,
+      "step": 3743
+    },
+    {
+      "epoch": 0.32065775950668035,
+      "grad_norm": 0.5397855639457703,
+      "learning_rate": 0.0001,
+      "loss": 0.879,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3207434052757794,
+      "grad_norm": 0.6100704073905945,
+      "learning_rate": 0.0001,
+      "loss": 0.9788,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3208290510448784,
+      "grad_norm": 0.5918869376182556,
+      "learning_rate": 0.0001,
+      "loss": 0.8343,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3209146968139774,
+      "grad_norm": 0.8446809649467468,
+      "learning_rate": 0.0001,
+      "loss": 0.6386,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3210003425830764,
+      "grad_norm": 0.9247546195983887,
+      "learning_rate": 0.0001,
+      "loss": 1.0976,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3210859883521754,
+      "grad_norm": 0.6264172196388245,
+      "learning_rate": 0.0001,
+      "loss": 0.6861,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3211716341212744,
+      "grad_norm": 0.7469913363456726,
+      "learning_rate": 0.0001,
+      "loss": 1.0805,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3212572798903734,
+      "grad_norm": 0.8562371730804443,
+      "learning_rate": 0.0001,
+      "loss": 0.8434,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3213429256594724,
+      "grad_norm": 0.7589596509933472,
+      "learning_rate": 0.0001,
+      "loss": 2.3645,
+      "step": 3752
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 1.2243458032608032,
+      "learning_rate": 0.0001,
+      "loss": 1.1922,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3215142171976704,
+      "grad_norm": 0.7484469413757324,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 3754
+    },
+    {
+      "epoch": 0.32159986296676946,
+      "grad_norm": 0.909461498260498,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 3755
+    },
+    {
+      "epoch": 0.32168550873586843,
+      "grad_norm": 1.2133209705352783,
+      "learning_rate": 0.0001,
+      "loss": 1.0772,
+      "step": 3756
+    },
+    {
+      "epoch": 0.32177115450496746,
+      "grad_norm": 1.3357479572296143,
+      "learning_rate": 0.0001,
+      "loss": 0.8104,
+      "step": 3757
+    },
+    {
+      "epoch": 0.32185680027406643,
+      "grad_norm": 0.7081220149993896,
+      "learning_rate": 0.0001,
+      "loss": 1.1525,
+      "step": 3758
+    },
+    {
+      "epoch": 0.32194244604316546,
+      "grad_norm": 1.010072946548462,
+      "learning_rate": 0.0001,
+      "loss": 1.3661,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3220280918122645,
+      "grad_norm": 0.5929631590843201,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32211373758136347,
+      "grad_norm": 0.46801069378852844,
+      "learning_rate": 0.0001,
+      "loss": 0.5651,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3221993833504625,
+      "grad_norm": 0.7301025986671448,
+      "learning_rate": 0.0001,
+      "loss": 0.8098,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3222850291195615,
+      "grad_norm": 0.7163842916488647,
+      "learning_rate": 0.0001,
+      "loss": 1.0727,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3223706748886605,
+      "grad_norm": 0.9388201832771301,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 3764
+    },
+    {
+      "epoch": 0.32245632065775953,
+      "grad_norm": 0.532342791557312,
+      "learning_rate": 0.0001,
+      "loss": 0.9302,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3225419664268585,
+      "grad_norm": 0.770913302898407,
+      "learning_rate": 0.0001,
+      "loss": 0.8806,
+      "step": 3766
+    },
+    {
+      "epoch": 0.32262761219595754,
+      "grad_norm": 0.6837062239646912,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3227132579650565,
+      "grad_norm": 0.6178735494613647,
+      "learning_rate": 0.0001,
+      "loss": 1.2845,
+      "step": 3768
+    },
+    {
+      "epoch": 0.32279890373415554,
+      "grad_norm": 1.0344032049179077,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3228845495032545,
+      "grad_norm": 0.5388871431350708,
+      "learning_rate": 0.0001,
+      "loss": 1.0154,
+      "step": 3770
+    },
+    {
+      "epoch": 0.32297019527235354,
+      "grad_norm": 0.737572431564331,
+      "learning_rate": 0.0001,
+      "loss": 0.7035,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3230558410414526,
+      "grad_norm": 0.47039732336997986,
+      "learning_rate": 0.0001,
+      "loss": 0.8406,
+      "step": 3772
+    },
+    {
+      "epoch": 0.32314148681055155,
+      "grad_norm": 0.5450010895729065,
+      "learning_rate": 0.0001,
+      "loss": 0.455,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3232271325796506,
+      "grad_norm": 0.9729748368263245,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 3774
+    },
+    {
+      "epoch": 0.32331277834874955,
+      "grad_norm": 0.4914383590221405,
+      "learning_rate": 0.0001,
+      "loss": 0.2099,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3233984241178486,
+      "grad_norm": 0.6080459356307983,
+      "learning_rate": 0.0001,
+      "loss": 0.6126,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3234840698869476,
+      "grad_norm": 0.7361956238746643,
+      "learning_rate": 0.0001,
+      "loss": 0.908,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3235697156560466,
+      "grad_norm": 0.6435142159461975,
+      "learning_rate": 0.0001,
+      "loss": 0.9297,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3236553614251456,
+      "grad_norm": 0.6273334622383118,
+      "learning_rate": 0.0001,
+      "loss": 0.7496,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3237410071942446,
+      "grad_norm": 0.8656031489372253,
+      "learning_rate": 0.0001,
+      "loss": 1.0811,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3238266529633436,
+      "grad_norm": 0.6780563592910767,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3239122987324426,
+      "grad_norm": 1.088423728942871,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3239979445015416,
+      "grad_norm": 0.593212366104126,
+      "learning_rate": 0.0001,
+      "loss": 0.9175,
+      "step": 3783
+    },
+    {
+      "epoch": 0.32408359027064065,
+      "grad_norm": 0.514229953289032,
+      "learning_rate": 0.0001,
+      "loss": 1.1992,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3241692360397396,
+      "grad_norm": 0.6757786870002747,
+      "learning_rate": 0.0001,
+      "loss": 1.182,
+      "step": 3785
+    },
+    {
+      "epoch": 0.32425488180883866,
+      "grad_norm": 0.4515257477760315,
+      "learning_rate": 0.0001,
+      "loss": 0.5571,
+      "step": 3786
+    },
+    {
+      "epoch": 0.32434052757793763,
+      "grad_norm": 0.526136040687561,
+      "learning_rate": 0.0001,
+      "loss": 0.8915,
+      "step": 3787
+    },
+    {
+      "epoch": 0.32442617334703666,
+      "grad_norm": 1.1321384906768799,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 3788
+    },
+    {
+      "epoch": 0.32451181911613564,
+      "grad_norm": 0.9213374853134155,
+      "learning_rate": 0.0001,
+      "loss": 1.8688,
+      "step": 3789
+    },
+    {
+      "epoch": 0.32459746488523467,
+      "grad_norm": 0.5081727504730225,
+      "learning_rate": 0.0001,
+      "loss": 0.9121,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3246831106543337,
+      "grad_norm": 0.7579797506332397,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 3791
+    },
+    {
+      "epoch": 0.32476875642343267,
+      "grad_norm": 0.5302383303642273,
+      "learning_rate": 0.0001,
+      "loss": 0.7465,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3248544021925317,
+      "grad_norm": 0.614389181137085,
+      "learning_rate": 0.0001,
+      "loss": 1.3555,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3249400479616307,
+      "grad_norm": 0.816831111907959,
+      "learning_rate": 0.0001,
+      "loss": 1.0504,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3250256937307297,
+      "grad_norm": 0.45557695627212524,
+      "learning_rate": 0.0001,
+      "loss": 0.322,
+      "step": 3795
+    },
+    {
+      "epoch": 0.32511133949982873,
+      "grad_norm": 0.5030039548873901,
+      "learning_rate": 0.0001,
+      "loss": 0.4344,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3251969852689277,
+      "grad_norm": 0.6149501800537109,
+      "learning_rate": 0.0001,
+      "loss": 1.1322,
+      "step": 3797
+    },
+    {
+      "epoch": 0.32528263103802674,
+      "grad_norm": 0.6883942484855652,
+      "learning_rate": 0.0001,
+      "loss": 0.4148,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3253682768071257,
+      "grad_norm": 0.4377252161502838,
+      "learning_rate": 0.0001,
+      "loss": 0.6341,
+      "step": 3799
+    },
+    {
+      "epoch": 0.32545392257622474,
+      "grad_norm": 0.7551379799842834,
+      "learning_rate": 0.0001,
+      "loss": 1.1261,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3255395683453237,
+      "grad_norm": 0.5753685832023621,
+      "learning_rate": 0.0001,
+      "loss": 0.9465,
+      "step": 3801
+    },
+    {
+      "epoch": 0.32562521411442275,
+      "grad_norm": 0.4683603346347809,
+      "learning_rate": 0.0001,
+      "loss": 0.9132,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3257108598835218,
+      "grad_norm": 0.7039430737495422,
+      "learning_rate": 0.0001,
+      "loss": 0.9673,
+      "step": 3803
+    },
+    {
+      "epoch": 0.32579650565262075,
+      "grad_norm": 1.1104389429092407,
+      "learning_rate": 0.0001,
+      "loss": 1.2686,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3258821514217198,
+      "grad_norm": 0.1898360699415207,
+      "learning_rate": 0.0001,
+      "loss": 0.547,
+      "step": 3805
+    },
+    {
+      "epoch": 0.32596779719081875,
+      "grad_norm": 0.9104442596435547,
+      "learning_rate": 0.0001,
+      "loss": 1.1648,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3260534429599178,
+      "grad_norm": 0.8727749586105347,
+      "learning_rate": 0.0001,
+      "loss": 1.0028,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3261390887290168,
+      "grad_norm": 0.6141719818115234,
+      "learning_rate": 0.0001,
+      "loss": 0.6139,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3262247344981158,
+      "grad_norm": 0.6981468796730042,
+      "learning_rate": 0.0001,
+      "loss": 0.7678,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3263103802672148,
+      "grad_norm": 0.49097341299057007,
+      "learning_rate": 0.0001,
+      "loss": 0.8862,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3263960260363138,
+      "grad_norm": 0.5728621482849121,
+      "learning_rate": 0.0001,
+      "loss": 1.4725,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3264816718054128,
+      "grad_norm": 0.8966420292854309,
+      "learning_rate": 0.0001,
+      "loss": 1.1839,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3265673175745118,
+      "grad_norm": 0.6024606823921204,
+      "learning_rate": 0.0001,
+      "loss": 1.1137,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3266529633436108,
+      "grad_norm": 0.8919455409049988,
+      "learning_rate": 0.0001,
+      "loss": 0.8171,
+      "step": 3814
+    },
+    {
+      "epoch": 0.32673860911270985,
+      "grad_norm": 0.7159332036972046,
+      "learning_rate": 0.0001,
+      "loss": 0.6608,
+      "step": 3815
+    },
+    {
+      "epoch": 0.32682425488180883,
+      "grad_norm": 0.7059553861618042,
+      "learning_rate": 0.0001,
+      "loss": 0.8874,
+      "step": 3816
+    },
+    {
+      "epoch": 0.32690990065090786,
+      "grad_norm": 0.7782920002937317,
+      "learning_rate": 0.0001,
+      "loss": 0.6475,
+      "step": 3817
+    },
+    {
+      "epoch": 0.32699554642000683,
+      "grad_norm": 0.7523946762084961,
+      "learning_rate": 0.0001,
+      "loss": 0.8958,
+      "step": 3818
+    },
+    {
+      "epoch": 0.32708119218910586,
+      "grad_norm": 0.683193027973175,
+      "learning_rate": 0.0001,
+      "loss": 0.5259,
+      "step": 3819
+    },
+    {
+      "epoch": 0.32716683795820484,
+      "grad_norm": 0.8811415433883667,
+      "learning_rate": 0.0001,
+      "loss": 0.9044,
+      "step": 3820
+    },
+    {
+      "epoch": 0.32725248372730387,
+      "grad_norm": 0.7754276990890503,
+      "learning_rate": 0.0001,
+      "loss": 1.2221,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3273381294964029,
+      "grad_norm": 0.5997880101203918,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 3822
+    },
+    {
+      "epoch": 0.32742377526550187,
+      "grad_norm": 0.6031416654586792,
+      "learning_rate": 0.0001,
+      "loss": 0.6844,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3275094210346009,
+      "grad_norm": 0.8124034404754639,
+      "learning_rate": 0.0001,
+      "loss": 1.0528,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3275950668036999,
+      "grad_norm": 0.5417327880859375,
+      "learning_rate": 0.0001,
+      "loss": 0.9325,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3276807125727989,
+      "grad_norm": 0.5731030106544495,
+      "learning_rate": 0.0001,
+      "loss": 0.4845,
+      "step": 3826
+    },
+    {
+      "epoch": 0.32776635834189793,
+      "grad_norm": 0.8526995778083801,
+      "learning_rate": 0.0001,
+      "loss": 1.0919,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3278520041109969,
+      "grad_norm": 0.8200701475143433,
+      "learning_rate": 0.0001,
+      "loss": 1.1514,
+      "step": 3828
+    },
+    {
+      "epoch": 0.32793764988009594,
+      "grad_norm": 0.5541043281555176,
+      "learning_rate": 0.0001,
+      "loss": 0.5101,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3280232956491949,
+      "grad_norm": 0.6272509694099426,
+      "learning_rate": 0.0001,
+      "loss": 1.018,
+      "step": 3830
+    },
+    {
+      "epoch": 0.32810894141829394,
+      "grad_norm": 0.9193716049194336,
+      "learning_rate": 0.0001,
+      "loss": 1.0895,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3281945871873929,
+      "grad_norm": 0.49844229221343994,
+      "learning_rate": 0.0001,
+      "loss": 0.718,
+      "step": 3832
+    },
+    {
+      "epoch": 0.32828023295649195,
+      "grad_norm": 0.5249109864234924,
+      "learning_rate": 0.0001,
+      "loss": 0.7352,
+      "step": 3833
+    },
+    {
+      "epoch": 0.328365878725591,
+      "grad_norm": 0.598171591758728,
+      "learning_rate": 0.0001,
+      "loss": 0.4243,
+      "step": 3834
+    },
+    {
+      "epoch": 0.32845152449468995,
+      "grad_norm": 0.6438761353492737,
+      "learning_rate": 0.0001,
+      "loss": 0.9617,
+      "step": 3835
+    },
+    {
+      "epoch": 0.328537170263789,
+      "grad_norm": 0.7174655795097351,
+      "learning_rate": 0.0001,
+      "loss": 0.794,
+      "step": 3836
+    },
+    {
+      "epoch": 0.32862281603288795,
+      "grad_norm": 0.7540953159332275,
+      "learning_rate": 0.0001,
+      "loss": 0.9088,
+      "step": 3837
+    },
+    {
+      "epoch": 0.328708461801987,
+      "grad_norm": 0.5865548253059387,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 3838
+    },
+    {
+      "epoch": 0.328794107571086,
+      "grad_norm": 1.0358495712280273,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 3839
+    },
+    {
+      "epoch": 0.328879753340185,
+      "grad_norm": 0.7782305479049683,
+      "learning_rate": 0.0001,
+      "loss": 0.9724,
+      "step": 3840
+    },
+    {
+      "epoch": 0.328965399109284,
+      "grad_norm": 0.9762192964553833,
+      "learning_rate": 0.0001,
+      "loss": 1.1192,
+      "step": 3841
+    },
+    {
+      "epoch": 0.329051044878383,
+      "grad_norm": 0.7602148652076721,
+      "learning_rate": 0.0001,
+      "loss": 0.6512,
+      "step": 3842
+    },
+    {
+      "epoch": 0.329136690647482,
+      "grad_norm": 0.5229492783546448,
+      "learning_rate": 0.0001,
+      "loss": 0.632,
+      "step": 3843
+    },
+    {
+      "epoch": 0.329222336416581,
+      "grad_norm": 0.5522335171699524,
+      "learning_rate": 0.0001,
+      "loss": 0.6893,
+      "step": 3844
+    },
+    {
+      "epoch": 0.32930798218568,
+      "grad_norm": 1.0075739622116089,
+      "learning_rate": 0.0001,
+      "loss": 0.9616,
+      "step": 3845
+    },
+    {
+      "epoch": 0.32939362795477906,
+      "grad_norm": 0.688080370426178,
+      "learning_rate": 0.0001,
+      "loss": 0.8345,
+      "step": 3846
+    },
+    {
+      "epoch": 0.32947927372387803,
+      "grad_norm": 0.8217177987098694,
+      "learning_rate": 0.0001,
+      "loss": 1.172,
+      "step": 3847
+    },
+    {
+      "epoch": 0.32956491949297706,
+      "grad_norm": 1.0206615924835205,
+      "learning_rate": 0.0001,
+      "loss": 1.0629,
+      "step": 3848
+    },
+    {
+      "epoch": 0.32965056526207603,
+      "grad_norm": 0.8249554634094238,
+      "learning_rate": 0.0001,
+      "loss": 0.6054,
+      "step": 3849
+    },
+    {
+      "epoch": 0.32973621103117506,
+      "grad_norm": 0.831951916217804,
+      "learning_rate": 0.0001,
+      "loss": 1.1603,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3298218568002741,
+      "grad_norm": 0.1794513761997223,
+      "learning_rate": 0.0001,
+      "loss": 0.5432,
+      "step": 3851
+    },
+    {
+      "epoch": 0.32990750256937307,
+      "grad_norm": 0.9661657214164734,
+      "learning_rate": 0.0001,
+      "loss": 1.1541,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3299931483384721,
+      "grad_norm": 0.9469389319419861,
+      "learning_rate": 0.0001,
+      "loss": 0.9025,
+      "step": 3853
+    },
+    {
+      "epoch": 0.33007879410757107,
+      "grad_norm": 1.0742616653442383,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3301644398766701,
+      "grad_norm": 0.7952379584312439,
+      "learning_rate": 0.0001,
+      "loss": 0.9873,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3302500856457691,
+      "grad_norm": 0.6943824887275696,
+      "learning_rate": 0.0001,
+      "loss": 0.9221,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3303357314148681,
+      "grad_norm": 1.0735204219818115,
+      "learning_rate": 0.0001,
+      "loss": 0.8759,
+      "step": 3857
+    },
+    {
+      "epoch": 0.33042137718396714,
+      "grad_norm": 0.6727892160415649,
+      "learning_rate": 0.0001,
+      "loss": 0.7017,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3305070229530661,
+      "grad_norm": 0.9199410080909729,
+      "learning_rate": 0.0001,
+      "loss": 1.1695,
+      "step": 3859
+    },
+    {
+      "epoch": 0.33059266872216514,
+      "grad_norm": 0.5937519073486328,
+      "learning_rate": 0.0001,
+      "loss": 0.9725,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3306783144912641,
+      "grad_norm": 0.47855842113494873,
+      "learning_rate": 0.0001,
+      "loss": 1.2207,
+      "step": 3861
+    },
+    {
+      "epoch": 0.33076396026036314,
+      "grad_norm": 0.9464132189750671,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3308496060294621,
+      "grad_norm": 0.6330770254135132,
+      "learning_rate": 0.0001,
+      "loss": 0.9511,
+      "step": 3863
+    },
+    {
+      "epoch": 0.33093525179856115,
+      "grad_norm": 0.8735906481742859,
+      "learning_rate": 0.0001,
+      "loss": 1.2107,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3310208975676602,
+      "grad_norm": 0.73003089427948,
+      "learning_rate": 0.0001,
+      "loss": 0.9781,
+      "step": 3865
+    },
+    {
+      "epoch": 0.33110654333675915,
+      "grad_norm": 0.8138616681098938,
+      "learning_rate": 0.0001,
+      "loss": 1.0967,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3311921891058582,
+      "grad_norm": 0.8736640214920044,
+      "learning_rate": 0.0001,
+      "loss": 1.0382,
+      "step": 3867
+    },
+    {
+      "epoch": 0.33127783487495716,
+      "grad_norm": 0.6422914862632751,
+      "learning_rate": 0.0001,
+      "loss": 1.0714,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3313634806440562,
+      "grad_norm": 0.6863909363746643,
+      "learning_rate": 0.0001,
+      "loss": 0.5571,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3314491264131552,
+      "grad_norm": 0.5890761017799377,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3315347721822542,
+      "grad_norm": 1.1748697757720947,
+      "learning_rate": 0.0001,
+      "loss": 1.3556,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3316204179513532,
+      "grad_norm": 0.9623372554779053,
+      "learning_rate": 0.0001,
+      "loss": 0.849,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3317060637204522,
+      "grad_norm": 0.7072551250457764,
+      "learning_rate": 0.0001,
+      "loss": 1.2449,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3317917094895512,
+      "grad_norm": 1.018304467201233,
+      "learning_rate": 0.0001,
+      "loss": 1.06,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3318773552586502,
+      "grad_norm": 0.49030572175979614,
+      "learning_rate": 0.0001,
+      "loss": 1.0049,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3319630010277492,
+      "grad_norm": 1.4897379875183105,
+      "learning_rate": 0.0001,
+      "loss": 1.3245,
+      "step": 3876
+    },
+    {
+      "epoch": 0.33204864679684826,
+      "grad_norm": 0.49515843391418457,
+      "learning_rate": 0.0001,
+      "loss": 0.9785,
+      "step": 3877
+    },
+    {
+      "epoch": 0.33213429256594723,
+      "grad_norm": 0.6601856350898743,
+      "learning_rate": 0.0001,
+      "loss": 1.0668,
+      "step": 3878
+    },
+    {
+      "epoch": 0.33221993833504626,
+      "grad_norm": 0.8079005479812622,
+      "learning_rate": 0.0001,
+      "loss": 0.7947,
+      "step": 3879
+    },
+    {
+      "epoch": 0.33230558410414524,
+      "grad_norm": 0.9091882109642029,
+      "learning_rate": 0.0001,
+      "loss": 0.9571,
+      "step": 3880
+    },
+    {
+      "epoch": 0.33239122987324426,
+      "grad_norm": 0.8855841755867004,
+      "learning_rate": 0.0001,
+      "loss": 1.019,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3324768756423433,
+      "grad_norm": 0.5262647271156311,
+      "learning_rate": 0.0001,
+      "loss": 1.0535,
+      "step": 3882
+    },
+    {
+      "epoch": 0.33256252141144227,
+      "grad_norm": 0.9334031939506531,
+      "learning_rate": 0.0001,
+      "loss": 1.2061,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3326481671805413,
+      "grad_norm": 0.6548319458961487,
+      "learning_rate": 0.0001,
+      "loss": 1.2493,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3327338129496403,
+      "grad_norm": 0.7709169983863831,
+      "learning_rate": 0.0001,
+      "loss": 0.5941,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3328194587187393,
+      "grad_norm": 0.69833904504776,
+      "learning_rate": 0.0001,
+      "loss": 1.2767,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3329051044878383,
+      "grad_norm": 0.7542062401771545,
+      "learning_rate": 0.0001,
+      "loss": 0.6339,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3329907502569373,
+      "grad_norm": 0.6139936447143555,
+      "learning_rate": 0.0001,
+      "loss": 0.8694,
+      "step": 3888
+    },
+    {
+      "epoch": 0.33307639602603634,
+      "grad_norm": 0.4004305601119995,
+      "learning_rate": 0.0001,
+      "loss": 0.7795,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3331620417951353,
+      "grad_norm": 0.41088610887527466,
+      "learning_rate": 0.0001,
+      "loss": 0.6812,
+      "step": 3890
+    },
+    {
+      "epoch": 0.33324768756423434,
+      "grad_norm": 0.6681734919548035,
+      "learning_rate": 0.0001,
+      "loss": 0.5745,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.9178585410118103,
+      "learning_rate": 0.0001,
+      "loss": 0.9818,
+      "step": 3892
+    },
+    {
+      "epoch": 0.33341897910243234,
+      "grad_norm": 1.257815957069397,
+      "learning_rate": 0.0001,
+      "loss": 0.928,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3335046248715313,
+      "grad_norm": 0.7917136549949646,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 3894
+    },
+    {
+      "epoch": 0.33359027064063035,
+      "grad_norm": 0.6909205317497253,
+      "learning_rate": 0.0001,
+      "loss": 1.0775,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3336759164097294,
+      "grad_norm": 1.1097546815872192,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 3896
+    },
+    {
+      "epoch": 0.33376156217882835,
+      "grad_norm": 0.7206329107284546,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3338472079479274,
+      "grad_norm": 0.8126260042190552,
+      "learning_rate": 0.0001,
+      "loss": 1.1002,
+      "step": 3898
+    },
+    {
+      "epoch": 0.33393285371702636,
+      "grad_norm": 0.9475092887878418,
+      "learning_rate": 0.0001,
+      "loss": 0.8994,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3340184994861254,
+      "grad_norm": 1.0356593132019043,
+      "learning_rate": 0.0001,
+      "loss": 0.8885,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3341041452552244,
+      "grad_norm": 0.5530816316604614,
+      "learning_rate": 0.0001,
+      "loss": 1.0453,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3341897910243234,
+      "grad_norm": 0.7647601962089539,
+      "learning_rate": 0.0001,
+      "loss": 0.8357,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3342754367934224,
+      "grad_norm": 0.6523106098175049,
+      "learning_rate": 0.0001,
+      "loss": 0.6459,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3343610825625214,
+      "grad_norm": 0.5767570734024048,
+      "learning_rate": 0.0001,
+      "loss": 0.841,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3344467283316204,
+      "grad_norm": 1.2437281608581543,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3345323741007194,
+      "grad_norm": 1.0356162786483765,
+      "learning_rate": 0.0001,
+      "loss": 1.1551,
+      "step": 3906
+    },
+    {
+      "epoch": 0.33461801986981843,
+      "grad_norm": 0.7824986577033997,
+      "learning_rate": 0.0001,
+      "loss": 0.9844,
+      "step": 3907
+    },
+    {
+      "epoch": 0.33470366563891746,
+      "grad_norm": 0.9615926146507263,
+      "learning_rate": 0.0001,
+      "loss": 0.9991,
+      "step": 3908
+    },
+    {
+      "epoch": 0.33478931140801643,
+      "grad_norm": 0.487847238779068,
+      "learning_rate": 0.0001,
+      "loss": 1.2769,
+      "step": 3909
+    },
+    {
+      "epoch": 0.33487495717711546,
+      "grad_norm": 0.6530437469482422,
+      "learning_rate": 0.0001,
+      "loss": 1.9412,
+      "step": 3910
+    },
+    {
+      "epoch": 0.33496060294621444,
+      "grad_norm": 0.4872656762599945,
+      "learning_rate": 0.0001,
+      "loss": 0.8196,
+      "step": 3911
+    },
+    {
+      "epoch": 0.33504624871531347,
+      "grad_norm": 0.9041366577148438,
+      "learning_rate": 0.0001,
+      "loss": 2.0552,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3351318944844125,
+      "grad_norm": 0.8484565615653992,
+      "learning_rate": 0.0001,
+      "loss": 1.2705,
+      "step": 3913
+    },
+    {
+      "epoch": 0.33521754025351147,
+      "grad_norm": 0.52808678150177,
+      "learning_rate": 0.0001,
+      "loss": 0.9024,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3353031860226105,
+      "grad_norm": 0.9721782207489014,
+      "learning_rate": 0.0001,
+      "loss": 1.2094,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3353888317917095,
+      "grad_norm": 0.898870050907135,
+      "learning_rate": 0.0001,
+      "loss": 0.8889,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3354744775608085,
+      "grad_norm": 0.5932486057281494,
+      "learning_rate": 0.0001,
+      "loss": 0.6795,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3355601233299075,
+      "grad_norm": 1.004218339920044,
+      "learning_rate": 0.0001,
+      "loss": 1.1979,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3356457690990065,
+      "grad_norm": 0.7841537594795227,
+      "learning_rate": 0.0001,
+      "loss": 1.0737,
+      "step": 3919
+    },
+    {
+      "epoch": 0.33573141486810554,
+      "grad_norm": 0.4874781668186188,
+      "learning_rate": 0.0001,
+      "loss": 0.8666,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3358170606372045,
+      "grad_norm": 0.7889251112937927,
+      "learning_rate": 0.0001,
+      "loss": 1.1422,
+      "step": 3921
+    },
+    {
+      "epoch": 0.33590270640630354,
+      "grad_norm": 0.5411005616188049,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3359883521754025,
+      "grad_norm": 0.8598918318748474,
+      "learning_rate": 0.0001,
+      "loss": 1.154,
+      "step": 3923
+    },
+    {
+      "epoch": 0.33607399794450155,
+      "grad_norm": 0.8684966564178467,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3361596437136006,
+      "grad_norm": 0.7983951568603516,
+      "learning_rate": 0.0001,
+      "loss": 0.5227,
+      "step": 3925
+    },
+    {
+      "epoch": 0.33624528948269955,
+      "grad_norm": 0.9223345518112183,
+      "learning_rate": 0.0001,
+      "loss": 1.2644,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3363309352517986,
+      "grad_norm": 0.3728843033313751,
+      "learning_rate": 0.0001,
+      "loss": 0.4775,
+      "step": 3927
+    },
+    {
+      "epoch": 0.33641658102089755,
+      "grad_norm": 0.8178536891937256,
+      "learning_rate": 0.0001,
+      "loss": 0.7439,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3365022267899966,
+      "grad_norm": 0.7334458827972412,
+      "learning_rate": 0.0001,
+      "loss": 0.5774,
+      "step": 3929
+    },
+    {
+      "epoch": 0.33658787255909556,
+      "grad_norm": 0.8880513906478882,
+      "learning_rate": 0.0001,
+      "loss": 0.6644,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3366735183281946,
+      "grad_norm": 0.8781514167785645,
+      "learning_rate": 0.0001,
+      "loss": 1.0322,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3367591640972936,
+      "grad_norm": 0.28036344051361084,
+      "learning_rate": 0.0001,
+      "loss": 0.3351,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3368448098663926,
+      "grad_norm": 0.5518988370895386,
+      "learning_rate": 0.0001,
+      "loss": 0.9739,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3369304556354916,
+      "grad_norm": 0.8683523535728455,
+      "learning_rate": 0.0001,
+      "loss": 1.2059,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3370161014045906,
+      "grad_norm": 0.9729846119880676,
+      "learning_rate": 0.0001,
+      "loss": 0.789,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3371017471736896,
+      "grad_norm": 0.6070700883865356,
+      "learning_rate": 0.0001,
+      "loss": 0.8583,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3371873929427886,
+      "grad_norm": 0.604752779006958,
+      "learning_rate": 0.0001,
+      "loss": 0.5687,
+      "step": 3937
+    },
+    {
+      "epoch": 0.33727303871188763,
+      "grad_norm": 0.5227699875831604,
+      "learning_rate": 0.0001,
+      "loss": 0.758,
+      "step": 3938
+    },
+    {
+      "epoch": 0.33735868448098666,
+      "grad_norm": 1.1349425315856934,
+      "learning_rate": 0.0001,
+      "loss": 1.2405,
+      "step": 3939
+    },
+    {
+      "epoch": 0.33744433025008563,
+      "grad_norm": 0.9215121865272522,
+      "learning_rate": 0.0001,
+      "loss": 1.1554,
+      "step": 3940
+    },
+    {
+      "epoch": 0.33752997601918466,
+      "grad_norm": 0.6267814636230469,
+      "learning_rate": 0.0001,
+      "loss": 0.9786,
+      "step": 3941
+    },
+    {
+      "epoch": 0.33761562178828364,
+      "grad_norm": 0.6530206203460693,
+      "learning_rate": 0.0001,
+      "loss": 0.9129,
+      "step": 3942
+    },
+    {
+      "epoch": 0.33770126755738267,
+      "grad_norm": 0.6874387860298157,
+      "learning_rate": 0.0001,
+      "loss": 0.5862,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3377869133264817,
+      "grad_norm": 0.688974916934967,
+      "learning_rate": 0.0001,
+      "loss": 0.5867,
+      "step": 3944
+    },
+    {
+      "epoch": 0.33787255909558067,
+      "grad_norm": 0.8037420511245728,
+      "learning_rate": 0.0001,
+      "loss": 0.9099,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3379582048646797,
+      "grad_norm": 0.8654597997665405,
+      "learning_rate": 0.0001,
+      "loss": 1.0343,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3380438506337787,
+      "grad_norm": 0.7876628041267395,
+      "learning_rate": 0.0001,
+      "loss": 0.4209,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3381294964028777,
+      "grad_norm": 0.8509734869003296,
+      "learning_rate": 0.0001,
+      "loss": 1.0058,
+      "step": 3948
+    },
+    {
+      "epoch": 0.3382151421719767,
+      "grad_norm": 0.6083712577819824,
+      "learning_rate": 0.0001,
+      "loss": 1.242,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3383007879410757,
+      "grad_norm": 0.7830711007118225,
+      "learning_rate": 0.0001,
+      "loss": 0.7811,
+      "step": 3950
+    },
+    {
+      "epoch": 0.33838643371017474,
+      "grad_norm": 0.7679637670516968,
+      "learning_rate": 0.0001,
+      "loss": 0.4525,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3384720794792737,
+      "grad_norm": 0.9170171618461609,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 3952
+    },
+    {
+      "epoch": 0.33855772524837274,
+      "grad_norm": 0.70868980884552,
+      "learning_rate": 0.0001,
+      "loss": 0.5832,
+      "step": 3953
+    },
+    {
+      "epoch": 0.3386433710174717,
+      "grad_norm": 0.8790260553359985,
+      "learning_rate": 0.0001,
+      "loss": 0.7498,
+      "step": 3954
+    },
+    {
+      "epoch": 0.33872901678657075,
+      "grad_norm": 0.9015652537345886,
+      "learning_rate": 0.0001,
+      "loss": 0.681,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3388146625556698,
+      "grad_norm": 0.8810988068580627,
+      "learning_rate": 0.0001,
+      "loss": 0.9733,
+      "step": 3956
+    },
+    {
+      "epoch": 0.33890030832476875,
+      "grad_norm": 0.5957056283950806,
+      "learning_rate": 0.0001,
+      "loss": 0.943,
+      "step": 3957
+    },
+    {
+      "epoch": 0.3389859540938678,
+      "grad_norm": 1.1350635290145874,
+      "learning_rate": 0.0001,
+      "loss": 1.887,
+      "step": 3958
+    },
+    {
+      "epoch": 0.33907159986296675,
+      "grad_norm": 0.5361334681510925,
+      "learning_rate": 0.0001,
+      "loss": 0.9243,
+      "step": 3959
+    },
+    {
+      "epoch": 0.3391572456320658,
+      "grad_norm": 0.4713767170906067,
+      "learning_rate": 0.0001,
+      "loss": 0.6573,
+      "step": 3960
+    },
+    {
+      "epoch": 0.33924289140116476,
+      "grad_norm": 0.47690021991729736,
+      "learning_rate": 0.0001,
+      "loss": 0.6044,
+      "step": 3961
+    },
+    {
+      "epoch": 0.3393285371702638,
+      "grad_norm": 0.7301964163780212,
+      "learning_rate": 0.0001,
+      "loss": 0.8948,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3394141829393628,
+      "grad_norm": 0.74989914894104,
+      "learning_rate": 0.0001,
+      "loss": 1.1063,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3394998287084618,
+      "grad_norm": 0.7357795238494873,
+      "learning_rate": 0.0001,
+      "loss": 0.5051,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3395854744775608,
+      "grad_norm": 1.1512796878814697,
+      "learning_rate": 0.0001,
+      "loss": 1.1803,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3396711202466598,
+      "grad_norm": 0.8430600166320801,
+      "learning_rate": 0.0001,
+      "loss": 0.8781,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3397567660157588,
+      "grad_norm": 0.8658788204193115,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3398424117848578,
+      "grad_norm": 1.0237457752227783,
+      "learning_rate": 0.0001,
+      "loss": 0.7897,
+      "step": 3968
+    },
+    {
+      "epoch": 0.33992805755395683,
+      "grad_norm": 0.8562004566192627,
+      "learning_rate": 0.0001,
+      "loss": 0.5784,
+      "step": 3969
+    },
+    {
+      "epoch": 0.34001370332305586,
+      "grad_norm": 0.6326751708984375,
+      "learning_rate": 0.0001,
+      "loss": 0.8949,
+      "step": 3970
+    },
+    {
+      "epoch": 0.34009934909215483,
+      "grad_norm": 0.7176500558853149,
+      "learning_rate": 0.0001,
+      "loss": 0.7405,
+      "step": 3971
+    },
+    {
+      "epoch": 0.34018499486125386,
+      "grad_norm": 0.9423809051513672,
+      "learning_rate": 0.0001,
+      "loss": 0.6742,
+      "step": 3972
+    },
+    {
+      "epoch": 0.34027064063035284,
+      "grad_norm": 0.8202533721923828,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 3973
+    },
+    {
+      "epoch": 0.34035628639945187,
+      "grad_norm": 0.4277772903442383,
+      "learning_rate": 0.0001,
+      "loss": 0.7548,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3404419321685509,
+      "grad_norm": 0.6507940292358398,
+      "learning_rate": 0.0001,
+      "loss": 1.1197,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3405275779376499,
+      "grad_norm": 0.4601222276687622,
+      "learning_rate": 0.0001,
+      "loss": 0.6762,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3406132237067489,
+      "grad_norm": 0.33624789118766785,
+      "learning_rate": 0.0001,
+      "loss": 0.2151,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3406988694758479,
+      "grad_norm": 1.9455010890960693,
+      "learning_rate": 0.0001,
+      "loss": 0.8174,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3407845152449469,
+      "grad_norm": 0.8049925565719604,
+      "learning_rate": 0.0001,
+      "loss": 0.9357,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3408701610140459,
+      "grad_norm": 0.9561393857002258,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3409558067831449,
+      "grad_norm": 0.947756826877594,
+      "learning_rate": 0.0001,
+      "loss": 1.1078,
+      "step": 3981
+    },
+    {
+      "epoch": 0.34104145255224394,
+      "grad_norm": 0.8774806261062622,
+      "learning_rate": 0.0001,
+      "loss": 1.3261,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3411270983213429,
+      "grad_norm": 0.6265371441841125,
+      "learning_rate": 0.0001,
+      "loss": 0.7191,
+      "step": 3983
+    },
+    {
+      "epoch": 0.34121274409044194,
+      "grad_norm": 0.5308070778846741,
+      "learning_rate": 0.0001,
+      "loss": 0.9071,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3412983898595409,
+      "grad_norm": 1.8665146827697754,
+      "learning_rate": 0.0001,
+      "loss": 1.0092,
+      "step": 3985
+    },
+    {
+      "epoch": 0.34138403562863995,
+      "grad_norm": 0.768906831741333,
+      "learning_rate": 0.0001,
+      "loss": 0.7092,
+      "step": 3986
+    },
+    {
+      "epoch": 0.341469681397739,
+      "grad_norm": 0.5732473731040955,
+      "learning_rate": 0.0001,
+      "loss": 0.6387,
+      "step": 3987
+    },
+    {
+      "epoch": 0.34155532716683795,
+      "grad_norm": 0.5373100638389587,
+      "learning_rate": 0.0001,
+      "loss": 1.0138,
+      "step": 3988
+    },
+    {
+      "epoch": 0.341640972935937,
+      "grad_norm": 0.9027242660522461,
+      "learning_rate": 0.0001,
+      "loss": 0.7127,
+      "step": 3989
+    },
+    {
+      "epoch": 0.34172661870503596,
+      "grad_norm": 1.0391813516616821,
+      "learning_rate": 0.0001,
+      "loss": 1.0706,
+      "step": 3990
+    },
+    {
+      "epoch": 0.341812264474135,
+      "grad_norm": 0.6813167929649353,
+      "learning_rate": 0.0001,
+      "loss": 1.1324,
+      "step": 3991
+    },
+    {
+      "epoch": 0.34189791024323396,
+      "grad_norm": 0.5015556216239929,
+      "learning_rate": 0.0001,
+      "loss": 0.9678,
+      "step": 3992
+    },
+    {
+      "epoch": 0.341983556012333,
+      "grad_norm": 0.8745372891426086,
+      "learning_rate": 0.0001,
+      "loss": 1.9459,
+      "step": 3993
+    },
+    {
+      "epoch": 0.342069201781432,
+      "grad_norm": 1.3848177194595337,
+      "learning_rate": 0.0001,
+      "loss": 0.6843,
+      "step": 3994
+    },
+    {
+      "epoch": 0.342154847550531,
+      "grad_norm": 1.2184122800827026,
+      "learning_rate": 0.0001,
+      "loss": 0.9778,
+      "step": 3995
+    },
+    {
+      "epoch": 0.34224049331963,
+      "grad_norm": 0.751976490020752,
+      "learning_rate": 0.0001,
+      "loss": 0.9213,
+      "step": 3996
+    },
+    {
+      "epoch": 0.342326139088729,
+      "grad_norm": 1.828227162361145,
+      "learning_rate": 0.0001,
+      "loss": 1.837,
+      "step": 3997
+    },
+    {
+      "epoch": 0.342411784857828,
+      "grad_norm": 1.0308113098144531,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 3998
+    },
+    {
+      "epoch": 0.34249743062692706,
+      "grad_norm": 0.522477924823761,
+      "learning_rate": 0.0001,
+      "loss": 0.8174,
+      "step": 3999
+    },
+    {
+      "epoch": 0.34258307639602603,
+      "grad_norm": 0.6187266707420349,
+      "learning_rate": 0.0001,
+      "loss": 0.8689,
+      "step": 4000
+    },
+    {
+      "epoch": 0.34266872216512506,
+      "grad_norm": 0.5349617600440979,
+      "learning_rate": 0.0001,
+      "loss": 0.3496,
+      "step": 4001
+    },
+    {
+      "epoch": 0.34275436793422404,
+      "grad_norm": 0.5224820971488953,
+      "learning_rate": 0.0001,
+      "loss": 0.8083,
+      "step": 4002
+    },
+    {
+      "epoch": 0.34284001370332307,
+      "grad_norm": 0.5646684169769287,
+      "learning_rate": 0.0001,
+      "loss": 0.954,
+      "step": 4003
+    },
+    {
+      "epoch": 0.34292565947242204,
+      "grad_norm": 1.0254480838775635,
+      "learning_rate": 0.0001,
+      "loss": 0.8716,
+      "step": 4004
+    },
+    {
+      "epoch": 0.34301130524152107,
+      "grad_norm": 1.292759656906128,
+      "learning_rate": 0.0001,
+      "loss": 0.7957,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3430969510106201,
+      "grad_norm": 1.0352989435195923,
+      "learning_rate": 0.0001,
+      "loss": 1.8547,
+      "step": 4006
+    },
+    {
+      "epoch": 0.3431825967797191,
+      "grad_norm": 0.816444456577301,
+      "learning_rate": 0.0001,
+      "loss": 0.7898,
+      "step": 4007
+    },
+    {
+      "epoch": 0.3432682425488181,
+      "grad_norm": 0.6494966149330139,
+      "learning_rate": 0.0001,
+      "loss": 0.5218,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3433538883179171,
+      "grad_norm": 0.8565108180046082,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 4009
+    },
+    {
+      "epoch": 0.3434395340870161,
+      "grad_norm": 1.052125096321106,
+      "learning_rate": 0.0001,
+      "loss": 1.13,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3435251798561151,
+      "grad_norm": 0.9971240162849426,
+      "learning_rate": 0.0001,
+      "loss": 1.1173,
+      "step": 4011
+    },
+    {
+      "epoch": 0.3436108256252141,
+      "grad_norm": 1.015007734298706,
+      "learning_rate": 0.0001,
+      "loss": 0.8365,
+      "step": 4012
+    },
+    {
+      "epoch": 0.34369647139431314,
+      "grad_norm": 0.7017309069633484,
+      "learning_rate": 0.0001,
+      "loss": 0.8321,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3437821171634121,
+      "grad_norm": 0.6485611200332642,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 4014
+    },
+    {
+      "epoch": 0.34386776293251115,
+      "grad_norm": 0.9753700494766235,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3439534087016101,
+      "grad_norm": 0.7744503021240234,
+      "learning_rate": 0.0001,
+      "loss": 0.4345,
+      "step": 4016
+    },
+    {
+      "epoch": 0.34403905447070915,
+      "grad_norm": 0.9724608063697815,
+      "learning_rate": 0.0001,
+      "loss": 0.7912,
+      "step": 4017
+    },
+    {
+      "epoch": 0.3441247002398082,
+      "grad_norm": 0.6229115128517151,
+      "learning_rate": 0.0001,
+      "loss": 1.0964,
+      "step": 4018
+    },
+    {
+      "epoch": 0.34421034600890715,
+      "grad_norm": 0.5874398350715637,
+      "learning_rate": 0.0001,
+      "loss": 1.071,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3442959917780062,
+      "grad_norm": 1.1775118112564087,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34438163754710516,
+      "grad_norm": 0.6143108606338501,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3444672833162042,
+      "grad_norm": 1.2165679931640625,
+      "learning_rate": 0.0001,
+      "loss": 2.3497,
+      "step": 4022
+    },
+    {
+      "epoch": 0.34455292908530316,
+      "grad_norm": 0.6680158376693726,
+      "learning_rate": 0.0001,
+      "loss": 0.9398,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3446385748544022,
+      "grad_norm": 0.5912964940071106,
+      "learning_rate": 0.0001,
+      "loss": 0.7645,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3447242206235012,
+      "grad_norm": 0.9859731197357178,
+      "learning_rate": 0.0001,
+      "loss": 0.8576,
+      "step": 4025
+    },
+    {
+      "epoch": 0.3448098663926002,
+      "grad_norm": 0.6119368076324463,
+      "learning_rate": 0.0001,
+      "loss": 1.1366,
+      "step": 4026
+    },
+    {
+      "epoch": 0.3448955121616992,
+      "grad_norm": 0.8237451314926147,
+      "learning_rate": 0.0001,
+      "loss": 0.9035,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3449811579307982,
+      "grad_norm": 0.5500839948654175,
+      "learning_rate": 0.0001,
+      "loss": 1.2635,
+      "step": 4028
+    },
+    {
+      "epoch": 0.34506680369989723,
+      "grad_norm": 0.9050427675247192,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 4029
+    },
+    {
+      "epoch": 0.34515244946899626,
+      "grad_norm": 0.877797544002533,
+      "learning_rate": 0.0001,
+      "loss": 0.799,
+      "step": 4030
+    },
+    {
+      "epoch": 0.34523809523809523,
+      "grad_norm": 0.9689050316810608,
+      "learning_rate": 0.0001,
+      "loss": 1.2273,
+      "step": 4031
+    },
+    {
+      "epoch": 0.34532374100719426,
+      "grad_norm": 0.8013272285461426,
+      "learning_rate": 0.0001,
+      "loss": 0.7996,
+      "step": 4032
+    },
+    {
+      "epoch": 0.34540938677629324,
+      "grad_norm": 1.0830535888671875,
+      "learning_rate": 0.0001,
+      "loss": 1.0406,
+      "step": 4033
+    },
+    {
+      "epoch": 0.34549503254539227,
+      "grad_norm": 0.9053192734718323,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 4034
+    },
+    {
+      "epoch": 0.34558067831449124,
+      "grad_norm": 0.7942925095558167,
+      "learning_rate": 0.0001,
+      "loss": 1.1857,
+      "step": 4035
+    },
+    {
+      "epoch": 0.34566632408359027,
+      "grad_norm": 0.7211294770240784,
+      "learning_rate": 0.0001,
+      "loss": 0.6543,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3457519698526893,
+      "grad_norm": 0.7774462103843689,
+      "learning_rate": 0.0001,
+      "loss": 0.978,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3458376156217883,
+      "grad_norm": 0.8164578676223755,
+      "learning_rate": 0.0001,
+      "loss": 1.0215,
+      "step": 4038
+    },
+    {
+      "epoch": 0.3459232613908873,
+      "grad_norm": 0.7308622002601624,
+      "learning_rate": 0.0001,
+      "loss": 0.8893,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3460089071599863,
+      "grad_norm": 0.7131825685501099,
+      "learning_rate": 0.0001,
+      "loss": 1.101,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3460945529290853,
+      "grad_norm": 0.5010089874267578,
+      "learning_rate": 0.0001,
+      "loss": 0.7145,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3461801986981843,
+      "grad_norm": 0.7951499223709106,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3462658444672833,
+      "grad_norm": 0.6462355852127075,
+      "learning_rate": 0.0001,
+      "loss": 0.4399,
+      "step": 4043
+    },
+    {
+      "epoch": 0.34635149023638234,
+      "grad_norm": 0.47562697529792786,
+      "learning_rate": 0.0001,
+      "loss": 1.1688,
+      "step": 4044
+    },
+    {
+      "epoch": 0.3464371360054813,
+      "grad_norm": 0.5887297987937927,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 4045
+    },
+    {
+      "epoch": 0.34652278177458035,
+      "grad_norm": 0.6867591142654419,
+      "learning_rate": 0.0001,
+      "loss": 0.4381,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3466084275436793,
+      "grad_norm": 0.7125188708305359,
+      "learning_rate": 0.0001,
+      "loss": 0.5123,
+      "step": 4047
+    },
+    {
+      "epoch": 0.34669407331277835,
+      "grad_norm": 0.4860018491744995,
+      "learning_rate": 0.0001,
+      "loss": 0.7389,
+      "step": 4048
+    },
+    {
+      "epoch": 0.3467797190818774,
+      "grad_norm": 1.0190547704696655,
+      "learning_rate": 0.0001,
+      "loss": 2.2234,
+      "step": 4049
+    },
+    {
+      "epoch": 0.34686536485097635,
+      "grad_norm": 0.42208993434906006,
+      "learning_rate": 0.0001,
+      "loss": 0.9283,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3469510106200754,
+      "grad_norm": 1.1803189516067505,
+      "learning_rate": 0.0001,
+      "loss": 1.1667,
+      "step": 4051
+    },
+    {
+      "epoch": 0.34703665638917436,
+      "grad_norm": 0.9657424092292786,
+      "learning_rate": 0.0001,
+      "loss": 0.8283,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3471223021582734,
+      "grad_norm": 0.8891724348068237,
+      "learning_rate": 0.0001,
+      "loss": 0.3298,
+      "step": 4053
+    },
+    {
+      "epoch": 0.34720794792737236,
+      "grad_norm": 0.6233811974525452,
+      "learning_rate": 0.0001,
+      "loss": 0.6147,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3472935936964714,
+      "grad_norm": 0.6370161771774292,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 4055
+    },
+    {
+      "epoch": 0.3473792394655704,
+      "grad_norm": 0.6170233488082886,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 4056
+    },
+    {
+      "epoch": 0.3474648852346694,
+      "grad_norm": 0.29066580533981323,
+      "learning_rate": 0.0001,
+      "loss": 0.3253,
+      "step": 4057
+    },
+    {
+      "epoch": 0.3475505310037684,
+      "grad_norm": 0.6716935038566589,
+      "learning_rate": 0.0001,
+      "loss": 1.0356,
+      "step": 4058
+    },
+    {
+      "epoch": 0.3476361767728674,
+      "grad_norm": 0.848132312297821,
+      "learning_rate": 0.0001,
+      "loss": 0.6008,
+      "step": 4059
+    },
+    {
+      "epoch": 0.34772182254196643,
+      "grad_norm": 0.9426493644714355,
+      "learning_rate": 0.0001,
+      "loss": 1.2066,
+      "step": 4060
+    },
+    {
+      "epoch": 0.34780746831106546,
+      "grad_norm": 0.7190262675285339,
+      "learning_rate": 0.0001,
+      "loss": 0.7487,
+      "step": 4061
+    },
+    {
+      "epoch": 0.34789311408016443,
+      "grad_norm": 0.6437575817108154,
+      "learning_rate": 0.0001,
+      "loss": 1.157,
+      "step": 4062
+    },
+    {
+      "epoch": 0.34797875984926346,
+      "grad_norm": 0.9057676792144775,
+      "learning_rate": 0.0001,
+      "loss": 0.4437,
+      "step": 4063
+    },
+    {
+      "epoch": 0.34806440561836244,
+      "grad_norm": 0.4914810359477997,
+      "learning_rate": 0.0001,
+      "loss": 1.1101,
+      "step": 4064
+    },
+    {
+      "epoch": 0.34815005138746147,
+      "grad_norm": 0.8682899475097656,
+      "learning_rate": 0.0001,
+      "loss": 0.8077,
+      "step": 4065
+    },
+    {
+      "epoch": 0.34823569715656044,
+      "grad_norm": 0.4183109402656555,
+      "learning_rate": 0.0001,
+      "loss": 0.3698,
+      "step": 4066
+    },
+    {
+      "epoch": 0.34832134292565947,
+      "grad_norm": 0.9369683265686035,
+      "learning_rate": 0.0001,
+      "loss": 1.0674,
+      "step": 4067
+    },
+    {
+      "epoch": 0.3484069886947585,
+      "grad_norm": 0.8611548542976379,
+      "learning_rate": 0.0001,
+      "loss": 0.9382,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3484926344638575,
+      "grad_norm": 0.6702811121940613,
+      "learning_rate": 0.0001,
+      "loss": 0.696,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3485782802329565,
+      "grad_norm": 0.4214920997619629,
+      "learning_rate": 0.0001,
+      "loss": 0.8397,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3486639260020555,
+      "grad_norm": 0.6876782774925232,
+      "learning_rate": 0.0001,
+      "loss": 1.0126,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3487495717711545,
+      "grad_norm": 0.9173919558525085,
+      "learning_rate": 0.0001,
+      "loss": 1.1963,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3488352175402535,
+      "grad_norm": 1.714719295501709,
+      "learning_rate": 0.0001,
+      "loss": 0.4336,
+      "step": 4073
+    },
+    {
+      "epoch": 0.3489208633093525,
+      "grad_norm": 0.49976685643196106,
+      "learning_rate": 0.0001,
+      "loss": 0.8816,
+      "step": 4074
+    },
+    {
+      "epoch": 0.34900650907845154,
+      "grad_norm": 0.620921790599823,
+      "learning_rate": 0.0001,
+      "loss": 0.5172,
+      "step": 4075
+    },
+    {
+      "epoch": 0.3490921548475505,
+      "grad_norm": 0.9734196662902832,
+      "learning_rate": 0.0001,
+      "loss": 0.941,
+      "step": 4076
+    },
+    {
+      "epoch": 0.34917780061664955,
+      "grad_norm": 0.8336032629013062,
+      "learning_rate": 0.0001,
+      "loss": 1.1979,
+      "step": 4077
+    },
+    {
+      "epoch": 0.3492634463857485,
+      "grad_norm": 0.6602168679237366,
+      "learning_rate": 0.0001,
+      "loss": 0.5301,
+      "step": 4078
+    },
+    {
+      "epoch": 0.34934909215484755,
+      "grad_norm": 1.3560254573822021,
+      "learning_rate": 0.0001,
+      "loss": 1.2303,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3494347379239466,
+      "grad_norm": 0.692051112651825,
+      "learning_rate": 0.0001,
+      "loss": 0.6624,
+      "step": 4080
+    },
+    {
+      "epoch": 0.34952038369304556,
+      "grad_norm": 0.6030526757240295,
+      "learning_rate": 0.0001,
+      "loss": 1.2609,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3496060294621446,
+      "grad_norm": 0.6631998419761658,
+      "learning_rate": 0.0001,
+      "loss": 0.8855,
+      "step": 4082
+    },
+    {
+      "epoch": 0.34969167523124356,
+      "grad_norm": 0.5373520255088806,
+      "learning_rate": 0.0001,
+      "loss": 1.2762,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3497773210003426,
+      "grad_norm": 0.7373189926147461,
+      "learning_rate": 0.0001,
+      "loss": 0.9575,
+      "step": 4084
+    },
+    {
+      "epoch": 0.34986296676944156,
+      "grad_norm": 0.6201328039169312,
+      "learning_rate": 0.0001,
+      "loss": 0.6062,
+      "step": 4085
+    },
+    {
+      "epoch": 0.3499486125385406,
+      "grad_norm": 0.8653696179389954,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3500342583076396,
+      "grad_norm": 1.1566648483276367,
+      "learning_rate": 0.0001,
+      "loss": 0.9034,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3501199040767386,
+      "grad_norm": 0.8023694753646851,
+      "learning_rate": 0.0001,
+      "loss": 1.2987,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3502055498458376,
+      "grad_norm": 0.9010756015777588,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3502911956149366,
+      "grad_norm": 0.4192590117454529,
+      "learning_rate": 0.0001,
+      "loss": 0.4987,
+      "step": 4090
+    },
+    {
+      "epoch": 0.35037684138403563,
+      "grad_norm": 0.9243385195732117,
+      "learning_rate": 0.0001,
+      "loss": 0.9539,
+      "step": 4091
+    },
+    {
+      "epoch": 0.35046248715313466,
+      "grad_norm": 0.481720894575119,
+      "learning_rate": 0.0001,
+      "loss": 1.0776,
+      "step": 4092
+    },
+    {
+      "epoch": 0.35054813292223364,
+      "grad_norm": 1.6090478897094727,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 4093
+    },
+    {
+      "epoch": 0.35063377869133266,
+      "grad_norm": 0.7361171245574951,
+      "learning_rate": 0.0001,
+      "loss": 0.7256,
+      "step": 4094
+    },
+    {
+      "epoch": 0.35071942446043164,
+      "grad_norm": 0.7051979899406433,
+      "learning_rate": 0.0001,
+      "loss": 0.4322,
+      "step": 4095
+    },
+    {
+      "epoch": 0.35080507022953067,
+      "grad_norm": 0.543821394443512,
+      "learning_rate": 0.0001,
+      "loss": 0.7794,
+      "step": 4096
+    },
+    {
+      "epoch": 0.35089071599862964,
+      "grad_norm": 0.7300189137458801,
+      "learning_rate": 0.0001,
+      "loss": 1.2237,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3509763617677287,
+      "grad_norm": 0.760587215423584,
+      "learning_rate": 0.0001,
+      "loss": 0.9734,
+      "step": 4098
+    },
+    {
+      "epoch": 0.3510620075368277,
+      "grad_norm": 0.7106562256813049,
+      "learning_rate": 0.0001,
+      "loss": 1.047,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3511476533059267,
+      "grad_norm": 0.5059508085250854,
+      "learning_rate": 0.0001,
+      "loss": 0.7709,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3512332990750257,
+      "grad_norm": 0.2585921883583069,
+      "learning_rate": 0.0001,
+      "loss": 0.217,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3513189448441247,
+      "grad_norm": 1.0118603706359863,
+      "learning_rate": 0.0001,
+      "loss": 1.2964,
+      "step": 4102
+    },
+    {
+      "epoch": 0.3514045906132237,
+      "grad_norm": 0.9240988492965698,
+      "learning_rate": 0.0001,
+      "loss": 0.4716,
+      "step": 4103
+    },
+    {
+      "epoch": 0.35149023638232274,
+      "grad_norm": 0.8946094512939453,
+      "learning_rate": 0.0001,
+      "loss": 1.2482,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3515758821514217,
+      "grad_norm": 1.116896152496338,
+      "learning_rate": 0.0001,
+      "loss": 0.8218,
+      "step": 4105
+    },
+    {
+      "epoch": 0.35166152792052074,
+      "grad_norm": 1.056064248085022,
+      "learning_rate": 0.0001,
+      "loss": 1.1181,
+      "step": 4106
+    },
+    {
+      "epoch": 0.3517471736896197,
+      "grad_norm": 0.8802167773246765,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 4107
+    },
+    {
+      "epoch": 0.35183281945871875,
+      "grad_norm": 0.6236728429794312,
+      "learning_rate": 0.0001,
+      "loss": 0.7062,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3519184652278177,
+      "grad_norm": 0.4646722674369812,
+      "learning_rate": 0.0001,
+      "loss": 0.2651,
+      "step": 4109
+    },
+    {
+      "epoch": 0.35200411099691675,
+      "grad_norm": 0.8585432767868042,
+      "learning_rate": 0.0001,
+      "loss": 0.6136,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3520897567660158,
+      "grad_norm": 0.6398283243179321,
+      "learning_rate": 0.0001,
+      "loss": 1.1618,
+      "step": 4111
+    },
+    {
+      "epoch": 0.35217540253511476,
+      "grad_norm": 0.7555648684501648,
+      "learning_rate": 0.0001,
+      "loss": 1.1148,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3522610483042138,
+      "grad_norm": 0.5805606842041016,
+      "learning_rate": 0.0001,
+      "loss": 0.8005,
+      "step": 4113
+    },
+    {
+      "epoch": 0.35234669407331276,
+      "grad_norm": 0.8109303712844849,
+      "learning_rate": 0.0001,
+      "loss": 0.5748,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3524323398424118,
+      "grad_norm": 0.8463156223297119,
+      "learning_rate": 0.0001,
+      "loss": 1.8567,
+      "step": 4115
+    },
+    {
+      "epoch": 0.35251798561151076,
+      "grad_norm": 0.8424477577209473,
+      "learning_rate": 0.0001,
+      "loss": 1.1862,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3526036313806098,
+      "grad_norm": 0.8469288349151611,
+      "learning_rate": 0.0001,
+      "loss": 0.6853,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3526892771497088,
+      "grad_norm": 0.8780702352523804,
+      "learning_rate": 0.0001,
+      "loss": 1.2755,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3527749229188078,
+      "grad_norm": 1.1485599279403687,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 4119
+    },
+    {
+      "epoch": 0.35286056868790683,
+      "grad_norm": 0.523274302482605,
+      "learning_rate": 0.0001,
+      "loss": 0.4672,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3529462144570058,
+      "grad_norm": 0.7294040322303772,
+      "learning_rate": 0.0001,
+      "loss": 0.8341,
+      "step": 4121
+    },
+    {
+      "epoch": 0.35303186022610483,
+      "grad_norm": 0.6842169165611267,
+      "learning_rate": 0.0001,
+      "loss": 0.4771,
+      "step": 4122
+    },
+    {
+      "epoch": 0.35311750599520386,
+      "grad_norm": 0.8127782940864563,
+      "learning_rate": 0.0001,
+      "loss": 1.0568,
+      "step": 4123
+    },
+    {
+      "epoch": 0.35320315176430284,
+      "grad_norm": 0.8958213925361633,
+      "learning_rate": 0.0001,
+      "loss": 1.2454,
+      "step": 4124
+    },
+    {
+      "epoch": 0.35328879753340187,
+      "grad_norm": 0.5580642819404602,
+      "learning_rate": 0.0001,
+      "loss": 0.7627,
+      "step": 4125
+    },
+    {
+      "epoch": 0.35337444330250084,
+      "grad_norm": 0.975152313709259,
+      "learning_rate": 0.0001,
+      "loss": 0.7056,
+      "step": 4126
+    },
+    {
+      "epoch": 0.35346008907159987,
+      "grad_norm": 0.4083564579486847,
+      "learning_rate": 0.0001,
+      "loss": 0.6467,
+      "step": 4127
+    },
+    {
+      "epoch": 0.35354573484069884,
+      "grad_norm": 1.1590626239776611,
+      "learning_rate": 0.0001,
+      "loss": 0.8654,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3536313806097979,
+      "grad_norm": 0.5646111965179443,
+      "learning_rate": 0.0001,
+      "loss": 0.8526,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3537170263788969,
+      "grad_norm": 0.47636672854423523,
+      "learning_rate": 0.0001,
+      "loss": 0.4691,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3538026721479959,
+      "grad_norm": 0.6311143636703491,
+      "learning_rate": 0.0001,
+      "loss": 1.1639,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3538883179170949,
+      "grad_norm": 0.27588754892349243,
+      "learning_rate": 0.0001,
+      "loss": 0.2998,
+      "step": 4132
+    },
+    {
+      "epoch": 0.3539739636861939,
+      "grad_norm": 0.7439038753509521,
+      "learning_rate": 0.0001,
+      "loss": 0.7139,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3540596094552929,
+      "grad_norm": 0.3795234262943268,
+      "learning_rate": 0.0001,
+      "loss": 0.4386,
+      "step": 4134
+    },
+    {
+      "epoch": 0.35414525522439194,
+      "grad_norm": 0.6824955940246582,
+      "learning_rate": 0.0001,
+      "loss": 1.1339,
+      "step": 4135
+    },
+    {
+      "epoch": 0.3542309009934909,
+      "grad_norm": 0.37208059430122375,
+      "learning_rate": 0.0001,
+      "loss": 0.3916,
+      "step": 4136
+    },
+    {
+      "epoch": 0.35431654676258995,
+      "grad_norm": 0.45681852102279663,
+      "learning_rate": 0.0001,
+      "loss": 0.3436,
+      "step": 4137
+    },
+    {
+      "epoch": 0.3544021925316889,
+      "grad_norm": 1.1760334968566895,
+      "learning_rate": 0.0001,
+      "loss": 0.9564,
+      "step": 4138
+    },
+    {
+      "epoch": 0.35448783830078795,
+      "grad_norm": 0.8902950882911682,
+      "learning_rate": 0.0001,
+      "loss": 0.7681,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3545734840698869,
+      "grad_norm": 1.246111273765564,
+      "learning_rate": 0.0001,
+      "loss": 1.0147,
+      "step": 4140
+    },
+    {
+      "epoch": 0.35465912983898595,
+      "grad_norm": 0.7233940958976746,
+      "learning_rate": 0.0001,
+      "loss": 1.0837,
+      "step": 4141
+    },
+    {
+      "epoch": 0.354744775608085,
+      "grad_norm": 0.5701032280921936,
+      "learning_rate": 0.0001,
+      "loss": 0.9548,
+      "step": 4142
+    },
+    {
+      "epoch": 0.35483042137718396,
+      "grad_norm": 0.8604149222373962,
+      "learning_rate": 0.0001,
+      "loss": 0.8408,
+      "step": 4143
+    },
+    {
+      "epoch": 0.354916067146283,
+      "grad_norm": 0.31538572907447815,
+      "learning_rate": 0.0001,
+      "loss": 0.2442,
+      "step": 4144
+    },
+    {
+      "epoch": 0.35500171291538196,
+      "grad_norm": 0.7324234247207642,
+      "learning_rate": 0.0001,
+      "loss": 0.6254,
+      "step": 4145
+    },
+    {
+      "epoch": 0.355087358684481,
+      "grad_norm": 0.5884763598442078,
+      "learning_rate": 0.0001,
+      "loss": 0.6982,
+      "step": 4146
+    },
+    {
+      "epoch": 0.35517300445357997,
+      "grad_norm": 1.146898865699768,
+      "learning_rate": 0.0001,
+      "loss": 1.3361,
+      "step": 4147
+    },
+    {
+      "epoch": 0.355258650222679,
+      "grad_norm": 0.8050804138183594,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 4148
+    },
+    {
+      "epoch": 0.355344295991778,
+      "grad_norm": 0.8954089879989624,
+      "learning_rate": 0.0001,
+      "loss": 0.8517,
+      "step": 4149
+    },
+    {
+      "epoch": 0.355429941760877,
+      "grad_norm": 0.852520227432251,
+      "learning_rate": 0.0001,
+      "loss": 0.7212,
+      "step": 4150
+    },
+    {
+      "epoch": 0.35551558752997603,
+      "grad_norm": 0.6608217358589172,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 4151
+    },
+    {
+      "epoch": 0.355601233299075,
+      "grad_norm": 0.5919408798217773,
+      "learning_rate": 0.0001,
+      "loss": 1.1827,
+      "step": 4152
+    },
+    {
+      "epoch": 0.35568687906817403,
+      "grad_norm": 1.0631439685821533,
+      "learning_rate": 0.0001,
+      "loss": 0.7659,
+      "step": 4153
+    },
+    {
+      "epoch": 0.35577252483727306,
+      "grad_norm": 0.5413939356803894,
+      "learning_rate": 0.0001,
+      "loss": 1.0217,
+      "step": 4154
+    },
+    {
+      "epoch": 0.35585817060637204,
+      "grad_norm": 0.8098914623260498,
+      "learning_rate": 0.0001,
+      "loss": 0.6704,
+      "step": 4155
+    },
+    {
+      "epoch": 0.35594381637547107,
+      "grad_norm": 0.7227965593338013,
+      "learning_rate": 0.0001,
+      "loss": 1.0962,
+      "step": 4156
+    },
+    {
+      "epoch": 0.35602946214457004,
+      "grad_norm": 0.8757303357124329,
+      "learning_rate": 0.0001,
+      "loss": 0.877,
+      "step": 4157
+    },
+    {
+      "epoch": 0.35611510791366907,
+      "grad_norm": 0.8811019062995911,
+      "learning_rate": 0.0001,
+      "loss": 1.1782,
+      "step": 4158
+    },
+    {
+      "epoch": 0.35620075368276805,
+      "grad_norm": 0.7800834774971008,
+      "learning_rate": 0.0001,
+      "loss": 0.7135,
+      "step": 4159
+    },
+    {
+      "epoch": 0.3562863994518671,
+      "grad_norm": 0.5435731410980225,
+      "learning_rate": 0.0001,
+      "loss": 1.2815,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3563720452209661,
+      "grad_norm": 0.8036850690841675,
+      "learning_rate": 0.0001,
+      "loss": 0.8129,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3564576909900651,
+      "grad_norm": 0.7638561129570007,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3565433367591641,
+      "grad_norm": 0.6291007399559021,
+      "learning_rate": 0.0001,
+      "loss": 1.0437,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3566289825282631,
+      "grad_norm": 0.5007753372192383,
+      "learning_rate": 0.0001,
+      "loss": 1.0315,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3567146282973621,
+      "grad_norm": 0.7626506090164185,
+      "learning_rate": 0.0001,
+      "loss": 0.8568,
+      "step": 4165
+    },
+    {
+      "epoch": 0.35680027406646114,
+      "grad_norm": 0.9917731881141663,
+      "learning_rate": 0.0001,
+      "loss": 0.656,
+      "step": 4166
+    },
+    {
+      "epoch": 0.3568859198355601,
+      "grad_norm": 1.6501092910766602,
+      "learning_rate": 0.0001,
+      "loss": 0.9101,
+      "step": 4167
+    },
+    {
+      "epoch": 0.35697156560465915,
+      "grad_norm": 0.8279223442077637,
+      "learning_rate": 0.0001,
+      "loss": 0.9061,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3570572113737581,
+      "grad_norm": 0.7459834218025208,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 4169
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.204209804534912,
+      "learning_rate": 0.0001,
+      "loss": 0.7541,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3572285029119561,
+      "grad_norm": 0.5438406467437744,
+      "learning_rate": 0.0001,
+      "loss": 0.8121,
+      "step": 4171
+    },
+    {
+      "epoch": 0.35731414868105515,
+      "grad_norm": 0.8597432971000671,
+      "learning_rate": 0.0001,
+      "loss": 0.5269,
+      "step": 4172
+    },
+    {
+      "epoch": 0.3573997944501542,
+      "grad_norm": 0.8132796287536621,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 4173
+    },
+    {
+      "epoch": 0.35748544021925316,
+      "grad_norm": 0.8870833516120911,
+      "learning_rate": 0.0001,
+      "loss": 1.8509,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3575710859883522,
+      "grad_norm": 0.80549156665802,
+      "learning_rate": 0.0001,
+      "loss": 0.9011,
+      "step": 4175
+    },
+    {
+      "epoch": 0.35765673175745116,
+      "grad_norm": 1.0354965925216675,
+      "learning_rate": 0.0001,
+      "loss": 0.8223,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3577423775265502,
+      "grad_norm": 0.6001464128494263,
+      "learning_rate": 0.0001,
+      "loss": 1.26,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3578280232956492,
+      "grad_norm": 1.0888597965240479,
+      "learning_rate": 0.0001,
+      "loss": 0.6604,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3579136690647482,
+      "grad_norm": 0.5751863718032837,
+      "learning_rate": 0.0001,
+      "loss": 0.9451,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3579993148338472,
+      "grad_norm": 0.5362673401832581,
+      "learning_rate": 0.0001,
+      "loss": 0.8242,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3580849606029462,
+      "grad_norm": 1.0013110637664795,
+      "learning_rate": 0.0001,
+      "loss": 1.299,
+      "step": 4181
+    },
+    {
+      "epoch": 0.35817060637204523,
+      "grad_norm": 0.7189188599586487,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3582562521411442,
+      "grad_norm": 0.45895054936408997,
+      "learning_rate": 0.0001,
+      "loss": 0.829,
+      "step": 4183
+    },
+    {
+      "epoch": 0.35834189791024323,
+      "grad_norm": 0.27016645669937134,
+      "learning_rate": 0.0001,
+      "loss": 0.4431,
+      "step": 4184
+    },
+    {
+      "epoch": 0.35842754367934226,
+      "grad_norm": 1.0713868141174316,
+      "learning_rate": 0.0001,
+      "loss": 0.9636,
+      "step": 4185
+    },
+    {
+      "epoch": 0.35851318944844124,
+      "grad_norm": 1.0373444557189941,
+      "learning_rate": 0.0001,
+      "loss": 1.1876,
+      "step": 4186
+    },
+    {
+      "epoch": 0.35859883521754027,
+      "grad_norm": 0.9533973932266235,
+      "learning_rate": 0.0001,
+      "loss": 0.7435,
+      "step": 4187
+    },
+    {
+      "epoch": 0.35868448098663924,
+      "grad_norm": 0.7671564221382141,
+      "learning_rate": 0.0001,
+      "loss": 0.5712,
+      "step": 4188
+    },
+    {
+      "epoch": 0.3587701267557383,
+      "grad_norm": 0.5648828148841858,
+      "learning_rate": 0.0001,
+      "loss": 1.0236,
+      "step": 4189
+    },
+    {
+      "epoch": 0.35885577252483725,
+      "grad_norm": 0.8919347524642944,
+      "learning_rate": 0.0001,
+      "loss": 1.0812,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3589414182939363,
+      "grad_norm": 0.8249903917312622,
+      "learning_rate": 0.0001,
+      "loss": 1.0554,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3590270640630353,
+      "grad_norm": 0.5883195400238037,
+      "learning_rate": 0.0001,
+      "loss": 0.9327,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3591127098321343,
+      "grad_norm": 0.5847822427749634,
+      "learning_rate": 0.0001,
+      "loss": 1.0613,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3591983556012333,
+      "grad_norm": 0.5266925096511841,
+      "learning_rate": 0.0001,
+      "loss": 0.8223,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3592840013703323,
+      "grad_norm": 0.3541956841945648,
+      "learning_rate": 0.0001,
+      "loss": 0.4633,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3593696471394313,
+      "grad_norm": 0.817772388458252,
+      "learning_rate": 0.0001,
+      "loss": 1.142,
+      "step": 4196
+    },
+    {
+      "epoch": 0.35945529290853034,
+      "grad_norm": 0.9116843342781067,
+      "learning_rate": 0.0001,
+      "loss": 0.8319,
+      "step": 4197
+    },
+    {
+      "epoch": 0.3595409386776293,
+      "grad_norm": 0.8306120038032532,
+      "learning_rate": 0.0001,
+      "loss": 0.766,
+      "step": 4198
+    },
+    {
+      "epoch": 0.35962658444672835,
+      "grad_norm": 0.7080026865005493,
+      "learning_rate": 0.0001,
+      "loss": 0.6282,
+      "step": 4199
+    },
+    {
+      "epoch": 0.3597122302158273,
+      "grad_norm": 0.7801399827003479,
+      "learning_rate": 0.0001,
+      "loss": 0.5893,
+      "step": 4200
+    },
+    {
+      "epoch": 0.35979787598492635,
+      "grad_norm": 1.1438735723495483,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3598835217540253,
+      "grad_norm": 0.5674352645874023,
+      "learning_rate": 0.0001,
+      "loss": 1.1094,
+      "step": 4202
+    },
+    {
+      "epoch": 0.35996916752312436,
+      "grad_norm": 0.5292955636978149,
+      "learning_rate": 0.0001,
+      "loss": 0.6532,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3600548132922234,
+      "grad_norm": 0.6668105125427246,
+      "learning_rate": 0.0001,
+      "loss": 0.6626,
+      "step": 4204
+    },
+    {
+      "epoch": 0.36014045906132236,
+      "grad_norm": 0.7583181858062744,
+      "learning_rate": 0.0001,
+      "loss": 0.8742,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3602261048304214,
+      "grad_norm": 0.5207677483558655,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 4206
+    },
+    {
+      "epoch": 0.36031175059952036,
+      "grad_norm": 0.7525460124015808,
+      "learning_rate": 0.0001,
+      "loss": 0.6979,
+      "step": 4207
+    },
+    {
+      "epoch": 0.3603973963686194,
+      "grad_norm": 0.4266641438007355,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3604830421377184,
+      "grad_norm": 0.7495178580284119,
+      "learning_rate": 0.0001,
+      "loss": 0.6856,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3605686879068174,
+      "grad_norm": 0.9838277101516724,
+      "learning_rate": 0.0001,
+      "loss": 1.2412,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3606543336759164,
+      "grad_norm": 0.2723329961299896,
+      "learning_rate": 0.0001,
+      "loss": 0.2393,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3607399794450154,
+      "grad_norm": 0.8937063217163086,
+      "learning_rate": 0.0001,
+      "loss": 0.639,
+      "step": 4212
+    },
+    {
+      "epoch": 0.36082562521411443,
+      "grad_norm": 1.8375933170318604,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3609112709832134,
+      "grad_norm": 1.0150622129440308,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 4214
+    },
+    {
+      "epoch": 0.36099691675231244,
+      "grad_norm": 0.8161620497703552,
+      "learning_rate": 0.0001,
+      "loss": 1.2901,
+      "step": 4215
+    },
+    {
+      "epoch": 0.36108256252141147,
+      "grad_norm": 0.48862481117248535,
+      "learning_rate": 0.0001,
+      "loss": 1.2572,
+      "step": 4216
+    },
+    {
+      "epoch": 0.36116820829051044,
+      "grad_norm": 0.40610989928245544,
+      "learning_rate": 0.0001,
+      "loss": 0.5929,
+      "step": 4217
+    },
+    {
+      "epoch": 0.36125385405960947,
+      "grad_norm": 0.7160743474960327,
+      "learning_rate": 0.0001,
+      "loss": 1.0074,
+      "step": 4218
+    },
+    {
+      "epoch": 0.36133949982870844,
+      "grad_norm": 0.9572138786315918,
+      "learning_rate": 0.0001,
+      "loss": 1.0919,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3614251455978075,
+      "grad_norm": 0.448503702878952,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 4220
+    },
+    {
+      "epoch": 0.36151079136690645,
+      "grad_norm": 0.7802540063858032,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3615964371360055,
+      "grad_norm": 0.7668935060501099,
+      "learning_rate": 0.0001,
+      "loss": 0.793,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3616820829051045,
+      "grad_norm": 0.7332763671875,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3617677286742035,
+      "grad_norm": 0.7515878677368164,
+      "learning_rate": 0.0001,
+      "loss": 1.1671,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3618533744433025,
+      "grad_norm": 0.7180070877075195,
+      "learning_rate": 0.0001,
+      "loss": 0.6891,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3619390202124015,
+      "grad_norm": 0.5700474977493286,
+      "learning_rate": 0.0001,
+      "loss": 0.6182,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3620246659815005,
+      "grad_norm": 1.0120551586151123,
+      "learning_rate": 0.0001,
+      "loss": 1.0558,
+      "step": 4227
+    },
+    {
+      "epoch": 0.36211031175059955,
+      "grad_norm": 0.7953854203224182,
+      "learning_rate": 0.0001,
+      "loss": 0.7589,
+      "step": 4228
+    },
+    {
+      "epoch": 0.3621959575196985,
+      "grad_norm": 0.6131635904312134,
+      "learning_rate": 0.0001,
+      "loss": 1.0395,
+      "step": 4229
+    },
+    {
+      "epoch": 0.36228160328879755,
+      "grad_norm": 0.19624949991703033,
+      "learning_rate": 0.0001,
+      "loss": 0.4027,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3623672490578965,
+      "grad_norm": 0.4744391143321991,
+      "learning_rate": 0.0001,
+      "loss": 0.9588,
+      "step": 4231
+    },
+    {
+      "epoch": 0.36245289482699555,
+      "grad_norm": 0.5886574983596802,
+      "learning_rate": 0.0001,
+      "loss": 1.1045,
+      "step": 4232
+    },
+    {
+      "epoch": 0.3625385405960945,
+      "grad_norm": 0.5858904123306274,
+      "learning_rate": 0.0001,
+      "loss": 0.8663,
+      "step": 4233
+    },
+    {
+      "epoch": 0.36262418636519356,
+      "grad_norm": 1.1184523105621338,
+      "learning_rate": 0.0001,
+      "loss": 1.2493,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3627098321342926,
+      "grad_norm": 0.4497387111186981,
+      "learning_rate": 0.0001,
+      "loss": 0.9161,
+      "step": 4235
+    },
+    {
+      "epoch": 0.36279547790339156,
+      "grad_norm": 1.055209994316101,
+      "learning_rate": 0.0001,
+      "loss": 0.4926,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3628811236724906,
+      "grad_norm": 0.1434563547372818,
+      "learning_rate": 0.0001,
+      "loss": 0.4364,
+      "step": 4237
+    },
+    {
+      "epoch": 0.36296676944158957,
+      "grad_norm": 0.5646766424179077,
+      "learning_rate": 0.0001,
+      "loss": 0.9021,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3630524152106886,
+      "grad_norm": 0.7206990122795105,
+      "learning_rate": 0.0001,
+      "loss": 0.963,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3631380609797876,
+      "grad_norm": 0.8740995526313782,
+      "learning_rate": 0.0001,
+      "loss": 1.0548,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3632237067488866,
+      "grad_norm": 0.7945768237113953,
+      "learning_rate": 0.0001,
+      "loss": 1.2375,
+      "step": 4241
+    },
+    {
+      "epoch": 0.36330935251798563,
+      "grad_norm": 0.7565231919288635,
+      "learning_rate": 0.0001,
+      "loss": 0.5378,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3633949982870846,
+      "grad_norm": 1.027437448501587,
+      "learning_rate": 0.0001,
+      "loss": 1.7661,
+      "step": 4243
+    },
+    {
+      "epoch": 0.36348064405618363,
+      "grad_norm": 0.8112614750862122,
+      "learning_rate": 0.0001,
+      "loss": 1.1558,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3635662898252826,
+      "grad_norm": 0.5973889231681824,
+      "learning_rate": 0.0001,
+      "loss": 1.2182,
+      "step": 4245
+    },
+    {
+      "epoch": 0.36365193559438164,
+      "grad_norm": 0.5586645007133484,
+      "learning_rate": 0.0001,
+      "loss": 0.9313,
+      "step": 4246
+    },
+    {
+      "epoch": 0.36373758136348067,
+      "grad_norm": 0.8851279616355896,
+      "learning_rate": 0.0001,
+      "loss": 0.7579,
+      "step": 4247
+    },
+    {
+      "epoch": 0.36382322713257964,
+      "grad_norm": 0.9658660292625427,
+      "learning_rate": 0.0001,
+      "loss": 1.1219,
+      "step": 4248
+    },
+    {
+      "epoch": 0.36390887290167867,
+      "grad_norm": 0.7822650671005249,
+      "learning_rate": 0.0001,
+      "loss": 1.0828,
+      "step": 4249
+    },
+    {
+      "epoch": 0.36399451867077764,
+      "grad_norm": 0.6306983828544617,
+      "learning_rate": 0.0001,
+      "loss": 1.1414,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3640801644398767,
+      "grad_norm": 0.8990561962127686,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3641658102089757,
+      "grad_norm": 0.41985800862312317,
+      "learning_rate": 0.0001,
+      "loss": 0.9378,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3642514559780747,
+      "grad_norm": 0.38207754492759705,
+      "learning_rate": 0.0001,
+      "loss": 0.4468,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3643371017471737,
+      "grad_norm": 0.4400835633277893,
+      "learning_rate": 0.0001,
+      "loss": 0.6061,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3644227475162727,
+      "grad_norm": 0.6254514455795288,
+      "learning_rate": 0.0001,
+      "loss": 0.7673,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3645083932853717,
+      "grad_norm": 0.7078502774238586,
+      "learning_rate": 0.0001,
+      "loss": 0.7424,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3645940390544707,
+      "grad_norm": 0.6861881613731384,
+      "learning_rate": 0.0001,
+      "loss": 0.87,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3646796848235697,
+      "grad_norm": 0.6823412179946899,
+      "learning_rate": 0.0001,
+      "loss": 0.7993,
+      "step": 4258
+    },
+    {
+      "epoch": 0.36476533059266875,
+      "grad_norm": 0.6354880332946777,
+      "learning_rate": 0.0001,
+      "loss": 1.0745,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3648509763617677,
+      "grad_norm": 0.32278454303741455,
+      "learning_rate": 0.0001,
+      "loss": 0.6459,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36493662213086675,
+      "grad_norm": 0.9127703309059143,
+      "learning_rate": 0.0001,
+      "loss": 0.9276,
+      "step": 4261
+    },
+    {
+      "epoch": 0.3650222678999657,
+      "grad_norm": 0.967166543006897,
+      "learning_rate": 0.0001,
+      "loss": 0.6045,
+      "step": 4262
+    },
+    {
+      "epoch": 0.36510791366906475,
+      "grad_norm": 0.7298080325126648,
+      "learning_rate": 0.0001,
+      "loss": 1.1967,
+      "step": 4263
+    },
+    {
+      "epoch": 0.36519355943816373,
+      "grad_norm": 0.9527549743652344,
+      "learning_rate": 0.0001,
+      "loss": 0.9152,
+      "step": 4264
+    },
+    {
+      "epoch": 0.36527920520726276,
+      "grad_norm": 0.8499334454536438,
+      "learning_rate": 0.0001,
+      "loss": 0.92,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3653648509763618,
+      "grad_norm": 0.7671193480491638,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 4266
+    },
+    {
+      "epoch": 0.36545049674546076,
+      "grad_norm": 0.9776320457458496,
+      "learning_rate": 0.0001,
+      "loss": 1.0866,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3655361425145598,
+      "grad_norm": 0.7178165316581726,
+      "learning_rate": 0.0001,
+      "loss": 1.3167,
+      "step": 4268
+    },
+    {
+      "epoch": 0.36562178828365877,
+      "grad_norm": 1.0559412240982056,
+      "learning_rate": 0.0001,
+      "loss": 1.3497,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3657074340527578,
+      "grad_norm": 0.4658542573451996,
+      "learning_rate": 0.0001,
+      "loss": 0.3285,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3657930798218568,
+      "grad_norm": 0.7567940950393677,
+      "learning_rate": 0.0001,
+      "loss": 1.0631,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3658787255909558,
+      "grad_norm": 0.9560396075248718,
+      "learning_rate": 0.0001,
+      "loss": 1.117,
+      "step": 4272
+    },
+    {
+      "epoch": 0.36596437136005483,
+      "grad_norm": 1.1149698495864868,
+      "learning_rate": 0.0001,
+      "loss": 0.9461,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3660500171291538,
+      "grad_norm": 0.8752434253692627,
+      "learning_rate": 0.0001,
+      "loss": 1.2307,
+      "step": 4274
+    },
+    {
+      "epoch": 0.36613566289825283,
+      "grad_norm": 0.7316733598709106,
+      "learning_rate": 0.0001,
+      "loss": 0.8331,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3662213086673518,
+      "grad_norm": 0.45480361580848694,
+      "learning_rate": 0.0001,
+      "loss": 0.9518,
+      "step": 4276
+    },
+    {
+      "epoch": 0.36630695443645084,
+      "grad_norm": 0.8557897806167603,
+      "learning_rate": 0.0001,
+      "loss": 0.4131,
+      "step": 4277
+    },
+    {
+      "epoch": 0.36639260020554987,
+      "grad_norm": 0.9462363123893738,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 4278
+    },
+    {
+      "epoch": 0.36647824597464884,
+      "grad_norm": 0.7567299008369446,
+      "learning_rate": 0.0001,
+      "loss": 0.7903,
+      "step": 4279
+    },
+    {
+      "epoch": 0.36656389174374787,
+      "grad_norm": 0.5732178688049316,
+      "learning_rate": 0.0001,
+      "loss": 0.9893,
+      "step": 4280
+    },
+    {
+      "epoch": 0.36664953751284685,
+      "grad_norm": 0.8692439794540405,
+      "learning_rate": 0.0001,
+      "loss": 0.6657,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3667351832819459,
+      "grad_norm": 0.8267269134521484,
+      "learning_rate": 0.0001,
+      "loss": 1.4007,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3668208290510449,
+      "grad_norm": 0.5979217290878296,
+      "learning_rate": 0.0001,
+      "loss": 1.0009,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3669064748201439,
+      "grad_norm": 0.17960350215435028,
+      "learning_rate": 0.0001,
+      "loss": 0.1915,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3669921205892429,
+      "grad_norm": 0.4837811589241028,
+      "learning_rate": 0.0001,
+      "loss": 0.482,
+      "step": 4285
+    },
+    {
+      "epoch": 0.3670777663583419,
+      "grad_norm": 0.9512040615081787,
+      "learning_rate": 0.0001,
+      "loss": 1.0701,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3671634121274409,
+      "grad_norm": 0.7674461603164673,
+      "learning_rate": 0.0001,
+      "loss": 1.3001,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3672490578965399,
+      "grad_norm": 0.7340047955513,
+      "learning_rate": 0.0001,
+      "loss": 0.4787,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3673347036656389,
+      "grad_norm": 0.9580519795417786,
+      "learning_rate": 0.0001,
+      "loss": 1.047,
+      "step": 4289
+    },
+    {
+      "epoch": 0.36742034943473795,
+      "grad_norm": 0.790567934513092,
+      "learning_rate": 0.0001,
+      "loss": 0.6817,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3675059952038369,
+      "grad_norm": 0.9530838131904602,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 4291
+    },
+    {
+      "epoch": 0.36759164097293595,
+      "grad_norm": 0.9290636777877808,
+      "learning_rate": 0.0001,
+      "loss": 0.7075,
+      "step": 4292
+    },
+    {
+      "epoch": 0.3676772867420349,
+      "grad_norm": 0.7065321803092957,
+      "learning_rate": 0.0001,
+      "loss": 1.2899,
+      "step": 4293
+    },
+    {
+      "epoch": 0.36776293251113396,
+      "grad_norm": 0.5259181261062622,
+      "learning_rate": 0.0001,
+      "loss": 0.6992,
+      "step": 4294
+    },
+    {
+      "epoch": 0.36784857828023293,
+      "grad_norm": 0.7635881900787354,
+      "learning_rate": 0.0001,
+      "loss": 0.8552,
+      "step": 4295
+    },
+    {
+      "epoch": 0.36793422404933196,
+      "grad_norm": 0.40282702445983887,
+      "learning_rate": 0.0001,
+      "loss": 0.7085,
+      "step": 4296
+    },
+    {
+      "epoch": 0.368019869818431,
+      "grad_norm": 0.6608101725578308,
+      "learning_rate": 0.0001,
+      "loss": 0.7819,
+      "step": 4297
+    },
+    {
+      "epoch": 0.36810551558752996,
+      "grad_norm": 0.8980430364608765,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 4298
+    },
+    {
+      "epoch": 0.368191161356629,
+      "grad_norm": 0.5739223957061768,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 4299
+    },
+    {
+      "epoch": 0.36827680712572797,
+      "grad_norm": 0.7401938438415527,
+      "learning_rate": 0.0001,
+      "loss": 0.8463,
+      "step": 4300
+    },
+    {
+      "epoch": 0.368362452894827,
+      "grad_norm": 0.5203717947006226,
+      "learning_rate": 0.0001,
+      "loss": 1.2103,
+      "step": 4301
+    },
+    {
+      "epoch": 0.368448098663926,
+      "grad_norm": 0.8352279663085938,
+      "learning_rate": 0.0001,
+      "loss": 0.828,
+      "step": 4302
+    },
+    {
+      "epoch": 0.368533744433025,
+      "grad_norm": 0.9778397679328918,
+      "learning_rate": 0.0001,
+      "loss": 1.314,
+      "step": 4303
+    },
+    {
+      "epoch": 0.36861939020212403,
+      "grad_norm": 0.8681238889694214,
+      "learning_rate": 0.0001,
+      "loss": 1.2374,
+      "step": 4304
+    },
+    {
+      "epoch": 0.368705035971223,
+      "grad_norm": 0.6224347352981567,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 4305
+    },
+    {
+      "epoch": 0.36879068174032203,
+      "grad_norm": 0.5909730792045593,
+      "learning_rate": 0.0001,
+      "loss": 1.2536,
+      "step": 4306
+    },
+    {
+      "epoch": 0.368876327509421,
+      "grad_norm": 0.7421393990516663,
+      "learning_rate": 0.0001,
+      "loss": 0.7885,
+      "step": 4307
+    },
+    {
+      "epoch": 0.36896197327852004,
+      "grad_norm": 1.1369951963424683,
+      "learning_rate": 0.0001,
+      "loss": 1.2438,
+      "step": 4308
+    },
+    {
+      "epoch": 0.36904761904761907,
+      "grad_norm": 0.46131935715675354,
+      "learning_rate": 0.0001,
+      "loss": 0.3599,
+      "step": 4309
+    },
+    {
+      "epoch": 0.36913326481671804,
+      "grad_norm": 0.9641860127449036,
+      "learning_rate": 0.0001,
+      "loss": 1.2362,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3692189105858171,
+      "grad_norm": 0.5246390700340271,
+      "learning_rate": 0.0001,
+      "loss": 0.7016,
+      "step": 4311
+    },
+    {
+      "epoch": 0.36930455635491605,
+      "grad_norm": 0.4664073586463928,
+      "learning_rate": 0.0001,
+      "loss": 1.0003,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3693902021240151,
+      "grad_norm": 1.3066420555114746,
+      "learning_rate": 0.0001,
+      "loss": 0.9157,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3694758478931141,
+      "grad_norm": 0.7705223560333252,
+      "learning_rate": 0.0001,
+      "loss": 1.0892,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3695614936622131,
+      "grad_norm": 1.1344407796859741,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3696471394313121,
+      "grad_norm": 0.6395446062088013,
+      "learning_rate": 0.0001,
+      "loss": 1.0987,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3697327852004111,
+      "grad_norm": 0.7780296206474304,
+      "learning_rate": 0.0001,
+      "loss": 0.8569,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3698184309695101,
+      "grad_norm": 0.6162005662918091,
+      "learning_rate": 0.0001,
+      "loss": 0.8546,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3699040767386091,
+      "grad_norm": 0.6337170600891113,
+      "learning_rate": 0.0001,
+      "loss": 1.1008,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3699897225077081,
+      "grad_norm": 0.4716821014881134,
+      "learning_rate": 0.0001,
+      "loss": 0.809,
+      "step": 4320
+    },
+    {
+      "epoch": 0.37007536827680715,
+      "grad_norm": 0.5528703331947327,
+      "learning_rate": 0.0001,
+      "loss": 0.8384,
+      "step": 4321
+    },
+    {
+      "epoch": 0.3701610140459061,
+      "grad_norm": 0.7572903037071228,
+      "learning_rate": 0.0001,
+      "loss": 0.6452,
+      "step": 4322
+    },
+    {
+      "epoch": 0.37024665981500515,
+      "grad_norm": 0.49905234575271606,
+      "learning_rate": 0.0001,
+      "loss": 1.2439,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3703323055841041,
+      "grad_norm": 0.5932380557060242,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 4324
+    },
+    {
+      "epoch": 0.37041795135320316,
+      "grad_norm": 0.7899297475814819,
+      "learning_rate": 0.0001,
+      "loss": 0.6751,
+      "step": 4325
+    },
+    {
+      "epoch": 0.37050359712230213,
+      "grad_norm": 0.7377115488052368,
+      "learning_rate": 0.0001,
+      "loss": 0.962,
+      "step": 4326
+    },
+    {
+      "epoch": 0.37058924289140116,
+      "grad_norm": 0.5952841639518738,
+      "learning_rate": 0.0001,
+      "loss": 1.2534,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3706748886605002,
+      "grad_norm": 0.8705366253852844,
+      "learning_rate": 0.0001,
+      "loss": 1.2872,
+      "step": 4328
+    },
+    {
+      "epoch": 0.37076053442959916,
+      "grad_norm": 0.6204879283905029,
+      "learning_rate": 0.0001,
+      "loss": 1.2798,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3708461801986982,
+      "grad_norm": 1.147934079170227,
+      "learning_rate": 0.0001,
+      "loss": 0.9512,
+      "step": 4330
+    },
+    {
+      "epoch": 0.37093182596779717,
+      "grad_norm": 1.2792991399765015,
+      "learning_rate": 0.0001,
+      "loss": 1.1843,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3710174717368962,
+      "grad_norm": 0.6934433579444885,
+      "learning_rate": 0.0001,
+      "loss": 0.8979,
+      "step": 4332
+    },
+    {
+      "epoch": 0.37110311750599523,
+      "grad_norm": 0.3724888861179352,
+      "learning_rate": 0.0001,
+      "loss": 0.6972,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3711887632750942,
+      "grad_norm": 1.0403355360031128,
+      "learning_rate": 0.0001,
+      "loss": 2.092,
+      "step": 4334
+    },
+    {
+      "epoch": 0.37127440904419323,
+      "grad_norm": 0.6000156998634338,
+      "learning_rate": 0.0001,
+      "loss": 1.1936,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3713600548132922,
+      "grad_norm": 0.42689821124076843,
+      "learning_rate": 0.0001,
+      "loss": 0.6774,
+      "step": 4336
+    },
+    {
+      "epoch": 0.37144570058239124,
+      "grad_norm": 0.6877188086509705,
+      "learning_rate": 0.0001,
+      "loss": 0.9644,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3715313463514902,
+      "grad_norm": 0.7271714806556702,
+      "learning_rate": 0.0001,
+      "loss": 0.9686,
+      "step": 4338
+    },
+    {
+      "epoch": 0.37161699212058924,
+      "grad_norm": 0.5020119547843933,
+      "learning_rate": 0.0001,
+      "loss": 0.9281,
+      "step": 4339
+    },
+    {
+      "epoch": 0.37170263788968827,
+      "grad_norm": 1.379259467124939,
+      "learning_rate": 0.0001,
+      "loss": 0.8061,
+      "step": 4340
+    },
+    {
+      "epoch": 0.37178828365878724,
+      "grad_norm": 0.6795976161956787,
+      "learning_rate": 0.0001,
+      "loss": 0.6001,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3718739294278863,
+      "grad_norm": 0.809497594833374,
+      "learning_rate": 0.0001,
+      "loss": 0.6213,
+      "step": 4342
+    },
+    {
+      "epoch": 0.37195957519698525,
+      "grad_norm": 0.860130786895752,
+      "learning_rate": 0.0001,
+      "loss": 0.9503,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3720452209660843,
+      "grad_norm": 1.0829377174377441,
+      "learning_rate": 0.0001,
+      "loss": 1.1382,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3721308667351833,
+      "grad_norm": 0.6189896464347839,
+      "learning_rate": 0.0001,
+      "loss": 1.0006,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3722165125042823,
+      "grad_norm": 0.7119210362434387,
+      "learning_rate": 0.0001,
+      "loss": 0.8044,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3723021582733813,
+      "grad_norm": 1.146198034286499,
+      "learning_rate": 0.0001,
+      "loss": 1.1202,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3723878040424803,
+      "grad_norm": 0.6901896595954895,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3724734498115793,
+      "grad_norm": 0.5831952095031738,
+      "learning_rate": 0.0001,
+      "loss": 1.1333,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3725590955806783,
+      "grad_norm": 1.0424180030822754,
+      "learning_rate": 0.0001,
+      "loss": 1.0742,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3726447413497773,
+      "grad_norm": 0.48688092827796936,
+      "learning_rate": 0.0001,
+      "loss": 0.5484,
+      "step": 4351
+    },
+    {
+      "epoch": 0.37273038711887635,
+      "grad_norm": 0.5693936944007874,
+      "learning_rate": 0.0001,
+      "loss": 0.8438,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3728160328879753,
+      "grad_norm": 0.6551782488822937,
+      "learning_rate": 0.0001,
+      "loss": 0.6114,
+      "step": 4353
+    },
+    {
+      "epoch": 0.37290167865707435,
+      "grad_norm": 1.0742565393447876,
+      "learning_rate": 0.0001,
+      "loss": 0.7283,
+      "step": 4354
+    },
+    {
+      "epoch": 0.37298732442617333,
+      "grad_norm": 1.0069721937179565,
+      "learning_rate": 0.0001,
+      "loss": 1.0057,
+      "step": 4355
+    },
+    {
+      "epoch": 0.37307297019527236,
+      "grad_norm": 0.8876773715019226,
+      "learning_rate": 0.0001,
+      "loss": 0.6022,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3731586159643714,
+      "grad_norm": 0.7124550342559814,
+      "learning_rate": 0.0001,
+      "loss": 0.8687,
+      "step": 4357
+    },
+    {
+      "epoch": 0.37324426173347036,
+      "grad_norm": 0.7256026268005371,
+      "learning_rate": 0.0001,
+      "loss": 0.5481,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3733299075025694,
+      "grad_norm": 0.8689727783203125,
+      "learning_rate": 0.0001,
+      "loss": 1.1578,
+      "step": 4359
+    },
+    {
+      "epoch": 0.37341555327166837,
+      "grad_norm": 0.7123323082923889,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3735011990407674,
+      "grad_norm": 0.8352968096733093,
+      "learning_rate": 0.0001,
+      "loss": 0.4654,
+      "step": 4361
+    },
+    {
+      "epoch": 0.37358684480986637,
+      "grad_norm": 0.9375721216201782,
+      "learning_rate": 0.0001,
+      "loss": 1.3132,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3736724905789654,
+      "grad_norm": 1.1492063999176025,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 4363
+    },
+    {
+      "epoch": 0.37375813634806443,
+      "grad_norm": 0.7962020635604858,
+      "learning_rate": 0.0001,
+      "loss": 1.1809,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3738437821171634,
+      "grad_norm": 0.5822604298591614,
+      "learning_rate": 0.0001,
+      "loss": 0.8692,
+      "step": 4365
+    },
+    {
+      "epoch": 0.37392942788626243,
+      "grad_norm": 0.7190683484077454,
+      "learning_rate": 0.0001,
+      "loss": 1.856,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3740150736553614,
+      "grad_norm": 0.7805909514427185,
+      "learning_rate": 0.0001,
+      "loss": 0.8496,
+      "step": 4367
+    },
+    {
+      "epoch": 0.37410071942446044,
+      "grad_norm": 0.6150520443916321,
+      "learning_rate": 0.0001,
+      "loss": 0.9826,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3741863651935594,
+      "grad_norm": 0.8803643584251404,
+      "learning_rate": 0.0001,
+      "loss": 1.1208,
+      "step": 4369
+    },
+    {
+      "epoch": 0.37427201096265844,
+      "grad_norm": 0.2836211323738098,
+      "learning_rate": 0.0001,
+      "loss": 0.3196,
+      "step": 4370
+    },
+    {
+      "epoch": 0.37435765673175747,
+      "grad_norm": 0.5991128087043762,
+      "learning_rate": 0.0001,
+      "loss": 1.2309,
+      "step": 4371
+    },
+    {
+      "epoch": 0.37444330250085645,
+      "grad_norm": 0.6316978335380554,
+      "learning_rate": 0.0001,
+      "loss": 0.6349,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3745289482699555,
+      "grad_norm": 0.24827709794044495,
+      "learning_rate": 0.0001,
+      "loss": 0.2271,
+      "step": 4373
+    },
+    {
+      "epoch": 0.37461459403905445,
+      "grad_norm": 0.46660685539245605,
+      "learning_rate": 0.0001,
+      "loss": 0.6649,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3747002398081535,
+      "grad_norm": 1.0341535806655884,
+      "learning_rate": 0.0001,
+      "loss": 1.0513,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3747858855772525,
+      "grad_norm": 0.7211129665374756,
+      "learning_rate": 0.0001,
+      "loss": 0.7154,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3748715313463515,
+      "grad_norm": 0.9095459580421448,
+      "learning_rate": 0.0001,
+      "loss": 0.9355,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3749571771154505,
+      "grad_norm": 0.909996509552002,
+      "learning_rate": 0.0001,
+      "loss": 0.6517,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3750428228845495,
+      "grad_norm": 0.7158880829811096,
+      "learning_rate": 0.0001,
+      "loss": 1.2991,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3751284686536485,
+      "grad_norm": 0.329862505197525,
+      "learning_rate": 0.0001,
+      "loss": 0.6684,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3752141144227475,
+      "grad_norm": 0.860834538936615,
+      "learning_rate": 0.0001,
+      "loss": 0.9983,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3752997601918465,
+      "grad_norm": 0.7530023455619812,
+      "learning_rate": 0.0001,
+      "loss": 0.5745,
+      "step": 4382
+    },
+    {
+      "epoch": 0.37538540596094555,
+      "grad_norm": 0.7902788519859314,
+      "learning_rate": 0.0001,
+      "loss": 1.3292,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3754710517300445,
+      "grad_norm": 1.0080277919769287,
+      "learning_rate": 0.0001,
+      "loss": 1.0663,
+      "step": 4384
+    },
+    {
+      "epoch": 0.37555669749914355,
+      "grad_norm": 1.0101979970932007,
+      "learning_rate": 0.0001,
+      "loss": 1.0578,
+      "step": 4385
+    },
+    {
+      "epoch": 0.37564234326824253,
+      "grad_norm": 0.6622142791748047,
+      "learning_rate": 0.0001,
+      "loss": 1.1105,
+      "step": 4386
+    },
+    {
+      "epoch": 0.37572798903734156,
+      "grad_norm": 0.6193706393241882,
+      "learning_rate": 0.0001,
+      "loss": 0.8903,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3758136348064406,
+      "grad_norm": 0.4472377598285675,
+      "learning_rate": 0.0001,
+      "loss": 0.4948,
+      "step": 4388
+    },
+    {
+      "epoch": 0.37589928057553956,
+      "grad_norm": 0.4211956560611725,
+      "learning_rate": 0.0001,
+      "loss": 0.5375,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3759849263446386,
+      "grad_norm": 0.777411937713623,
+      "learning_rate": 0.0001,
+      "loss": 0.6294,
+      "step": 4390
+    },
+    {
+      "epoch": 0.37607057211373757,
+      "grad_norm": 0.8787093758583069,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3761562178828366,
+      "grad_norm": 0.8858389258384705,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 4392
+    },
+    {
+      "epoch": 0.37624186365193557,
+      "grad_norm": 0.7805777788162231,
+      "learning_rate": 0.0001,
+      "loss": 0.8887,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3763275094210346,
+      "grad_norm": 1.2278327941894531,
+      "learning_rate": 0.0001,
+      "loss": 1.082,
+      "step": 4394
+    },
+    {
+      "epoch": 0.37641315519013363,
+      "grad_norm": 0.6153329610824585,
+      "learning_rate": 0.0001,
+      "loss": 0.7386,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3764988009592326,
+      "grad_norm": 0.5305188894271851,
+      "learning_rate": 0.0001,
+      "loss": 0.8893,
+      "step": 4396
+    },
+    {
+      "epoch": 0.37658444672833163,
+      "grad_norm": 0.7167041301727295,
+      "learning_rate": 0.0001,
+      "loss": 0.908,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3766700924974306,
+      "grad_norm": 0.5072420239448547,
+      "learning_rate": 0.0001,
+      "loss": 0.6898,
+      "step": 4398
+    },
+    {
+      "epoch": 0.37675573826652964,
+      "grad_norm": 1.025693655014038,
+      "learning_rate": 0.0001,
+      "loss": 1.3239,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3768413840356286,
+      "grad_norm": 0.5753906965255737,
+      "learning_rate": 0.0001,
+      "loss": 1.0029,
+      "step": 4400
+    },
+    {
+      "epoch": 0.37692702980472764,
+      "grad_norm": 1.1159428358078003,
+      "learning_rate": 0.0001,
+      "loss": 0.621,
+      "step": 4401
+    },
+    {
+      "epoch": 0.37701267557382667,
+      "grad_norm": 0.8539507985115051,
+      "learning_rate": 0.0001,
+      "loss": 0.4774,
+      "step": 4402
+    },
+    {
+      "epoch": 0.37709832134292565,
+      "grad_norm": 0.6412431001663208,
+      "learning_rate": 0.0001,
+      "loss": 0.9056,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3771839671120247,
+      "grad_norm": 0.8844560980796814,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 4404
+    },
+    {
+      "epoch": 0.37726961288112365,
+      "grad_norm": 0.39477285742759705,
+      "learning_rate": 0.0001,
+      "loss": 0.5054,
+      "step": 4405
+    },
+    {
+      "epoch": 0.3773552586502227,
+      "grad_norm": 0.9094131588935852,
+      "learning_rate": 0.0001,
+      "loss": 1.2406,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3774409044193217,
+      "grad_norm": 0.8282626867294312,
+      "learning_rate": 0.0001,
+      "loss": 1.2144,
+      "step": 4407
+    },
+    {
+      "epoch": 0.3775265501884207,
+      "grad_norm": 0.9664351344108582,
+      "learning_rate": 0.0001,
+      "loss": 1.107,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3776121959575197,
+      "grad_norm": 1.120078682899475,
+      "learning_rate": 0.0001,
+      "loss": 1.2799,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3776978417266187,
+      "grad_norm": 1.2770822048187256,
+      "learning_rate": 0.0001,
+      "loss": 1.3855,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3777834874957177,
+      "grad_norm": 0.795743465423584,
+      "learning_rate": 0.0001,
+      "loss": 1.3127,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3778691332648167,
+      "grad_norm": 1.0157018899917603,
+      "learning_rate": 0.0001,
+      "loss": 0.9651,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3779547790339157,
+      "grad_norm": 0.6703539490699768,
+      "learning_rate": 0.0001,
+      "loss": 0.531,
+      "step": 4413
+    },
+    {
+      "epoch": 0.37804042480301475,
+      "grad_norm": 0.7847596406936646,
+      "learning_rate": 0.0001,
+      "loss": 1.1919,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3781260705721137,
+      "grad_norm": 1.9421255588531494,
+      "learning_rate": 0.0001,
+      "loss": 1.3951,
+      "step": 4415
+    },
+    {
+      "epoch": 0.37821171634121276,
+      "grad_norm": 0.5599547624588013,
+      "learning_rate": 0.0001,
+      "loss": 0.9525,
+      "step": 4416
+    },
+    {
+      "epoch": 0.37829736211031173,
+      "grad_norm": 0.9637709259986877,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 4417
+    },
+    {
+      "epoch": 0.37838300787941076,
+      "grad_norm": 0.9517964124679565,
+      "learning_rate": 0.0001,
+      "loss": 0.6876,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3784686536485098,
+      "grad_norm": 0.863549530506134,
+      "learning_rate": 0.0001,
+      "loss": 0.7793,
+      "step": 4419
+    },
+    {
+      "epoch": 0.37855429941760876,
+      "grad_norm": 0.7247411608695984,
+      "learning_rate": 0.0001,
+      "loss": 0.7445,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3786399451867078,
+      "grad_norm": 0.5494673252105713,
+      "learning_rate": 0.0001,
+      "loss": 0.3877,
+      "step": 4421
+    },
+    {
+      "epoch": 0.37872559095580677,
+      "grad_norm": 1.1537550687789917,
+      "learning_rate": 0.0001,
+      "loss": 1.3024,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3788112367249058,
+      "grad_norm": 0.9864112138748169,
+      "learning_rate": 0.0001,
+      "loss": 1.0764,
+      "step": 4423
+    },
+    {
+      "epoch": 0.37889688249400477,
+      "grad_norm": 0.3139295279979706,
+      "learning_rate": 0.0001,
+      "loss": 0.414,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3789825282631038,
+      "grad_norm": 0.5587508678436279,
+      "learning_rate": 0.0001,
+      "loss": 0.7353,
+      "step": 4425
+    },
+    {
+      "epoch": 0.37906817403220283,
+      "grad_norm": 0.6192237734794617,
+      "learning_rate": 0.0001,
+      "loss": 0.778,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3791538198013018,
+      "grad_norm": 0.9608550667762756,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 4427
+    },
+    {
+      "epoch": 0.37923946557040084,
+      "grad_norm": 0.7479259371757507,
+      "learning_rate": 0.0001,
+      "loss": 0.8822,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3793251113394998,
+      "grad_norm": 0.607132613658905,
+      "learning_rate": 0.0001,
+      "loss": 1.0813,
+      "step": 4429
+    },
+    {
+      "epoch": 0.37941075710859884,
+      "grad_norm": 0.648796558380127,
+      "learning_rate": 0.0001,
+      "loss": 0.5722,
+      "step": 4430
+    },
+    {
+      "epoch": 0.37949640287769787,
+      "grad_norm": 0.6882171630859375,
+      "learning_rate": 0.0001,
+      "loss": 0.5866,
+      "step": 4431
+    },
+    {
+      "epoch": 0.37958204864679684,
+      "grad_norm": 0.9630005359649658,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3796676944158959,
+      "grad_norm": 0.7160454988479614,
+      "learning_rate": 0.0001,
+      "loss": 0.9836,
+      "step": 4433
+    },
+    {
+      "epoch": 0.37975334018499485,
+      "grad_norm": 0.722834050655365,
+      "learning_rate": 0.0001,
+      "loss": 0.3609,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3798389859540939,
+      "grad_norm": 0.7097123861312866,
+      "learning_rate": 0.0001,
+      "loss": 0.5701,
+      "step": 4435
+    },
+    {
+      "epoch": 0.37992463172319285,
+      "grad_norm": 1.0543904304504395,
+      "learning_rate": 0.0001,
+      "loss": 0.644,
+      "step": 4436
+    },
+    {
+      "epoch": 0.3800102774922919,
+      "grad_norm": 1.0580089092254639,
+      "learning_rate": 0.0001,
+      "loss": 1.0132,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3800959232613909,
+      "grad_norm": 0.7000226378440857,
+      "learning_rate": 0.0001,
+      "loss": 1.1034,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3801815690304899,
+      "grad_norm": 0.4916914999485016,
+      "learning_rate": 0.0001,
+      "loss": 0.7979,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3802672147995889,
+      "grad_norm": 1.0451921224594116,
+      "learning_rate": 0.0001,
+      "loss": 1.4119,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3803528605686879,
+      "grad_norm": 1.3989946842193604,
+      "learning_rate": 0.0001,
+      "loss": 0.9127,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3804385063377869,
+      "grad_norm": 1.4696277379989624,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3805241521068859,
+      "grad_norm": 0.499041885137558,
+      "learning_rate": 0.0001,
+      "loss": 0.736,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3806097978759849,
+      "grad_norm": 1.087377667427063,
+      "learning_rate": 0.0001,
+      "loss": 0.7357,
+      "step": 4444
+    },
+    {
+      "epoch": 0.38069544364508395,
+      "grad_norm": 0.9536212086677551,
+      "learning_rate": 0.0001,
+      "loss": 1.0573,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3807810894141829,
+      "grad_norm": 0.5344764590263367,
+      "learning_rate": 0.0001,
+      "loss": 0.6439,
+      "step": 4446
+    },
+    {
+      "epoch": 0.38086673518328196,
+      "grad_norm": 1.0042290687561035,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 4447
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.9826169013977051,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 4448
+    },
+    {
+      "epoch": 0.38103802672147996,
+      "grad_norm": 0.691528856754303,
+      "learning_rate": 0.0001,
+      "loss": 0.6607,
+      "step": 4449
+    },
+    {
+      "epoch": 0.381123672490579,
+      "grad_norm": 0.7830841541290283,
+      "learning_rate": 0.0001,
+      "loss": 1.3268,
+      "step": 4450
+    },
+    {
+      "epoch": 0.38120931825967797,
+      "grad_norm": 0.5972996354103088,
+      "learning_rate": 0.0001,
+      "loss": 1.2383,
+      "step": 4451
+    },
+    {
+      "epoch": 0.381294964028777,
+      "grad_norm": 0.5763824582099915,
+      "learning_rate": 0.0001,
+      "loss": 1.0056,
+      "step": 4452
+    },
+    {
+      "epoch": 0.38138060979787597,
+      "grad_norm": 0.9030980467796326,
+      "learning_rate": 0.0001,
+      "loss": 0.6953,
+      "step": 4453
+    },
+    {
+      "epoch": 0.381466255566975,
+      "grad_norm": 0.6539404988288879,
+      "learning_rate": 0.0001,
+      "loss": 0.5853,
+      "step": 4454
+    },
+    {
+      "epoch": 0.381551901336074,
+      "grad_norm": 0.6973399519920349,
+      "learning_rate": 0.0001,
+      "loss": 0.4869,
+      "step": 4455
+    },
+    {
+      "epoch": 0.381637547105173,
+      "grad_norm": 0.5098214745521545,
+      "learning_rate": 0.0001,
+      "loss": 0.4378,
+      "step": 4456
+    },
+    {
+      "epoch": 0.38172319287427203,
+      "grad_norm": 0.8589546084403992,
+      "learning_rate": 0.0001,
+      "loss": 0.9513,
+      "step": 4457
+    },
+    {
+      "epoch": 0.381808838643371,
+      "grad_norm": 0.6640045642852783,
+      "learning_rate": 0.0001,
+      "loss": 0.5562,
+      "step": 4458
+    },
+    {
+      "epoch": 0.38189448441247004,
+      "grad_norm": 0.7775278091430664,
+      "learning_rate": 0.0001,
+      "loss": 0.9564,
+      "step": 4459
+    },
+    {
+      "epoch": 0.381980130181569,
+      "grad_norm": 1.0751577615737915,
+      "learning_rate": 0.0001,
+      "loss": 1.3707,
+      "step": 4460
+    },
+    {
+      "epoch": 0.38206577595066804,
+      "grad_norm": 0.7275714874267578,
+      "learning_rate": 0.0001,
+      "loss": 1.2693,
+      "step": 4461
+    },
+    {
+      "epoch": 0.38215142171976707,
+      "grad_norm": 1.1548186540603638,
+      "learning_rate": 0.0001,
+      "loss": 1.1652,
+      "step": 4462
+    },
+    {
+      "epoch": 0.38223706748886604,
+      "grad_norm": 0.867953360080719,
+      "learning_rate": 0.0001,
+      "loss": 0.8211,
+      "step": 4463
+    },
+    {
+      "epoch": 0.3823227132579651,
+      "grad_norm": 0.6783629655838013,
+      "learning_rate": 0.0001,
+      "loss": 0.7404,
+      "step": 4464
+    },
+    {
+      "epoch": 0.38240835902706405,
+      "grad_norm": 0.6260953545570374,
+      "learning_rate": 0.0001,
+      "loss": 1.064,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3824940047961631,
+      "grad_norm": 0.7301231622695923,
+      "learning_rate": 0.0001,
+      "loss": 0.6055,
+      "step": 4466
+    },
+    {
+      "epoch": 0.38257965056526205,
+      "grad_norm": 0.735485315322876,
+      "learning_rate": 0.0001,
+      "loss": 0.6424,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3826652963343611,
+      "grad_norm": 0.9170037508010864,
+      "learning_rate": 0.0001,
+      "loss": 1.1212,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3827509421034601,
+      "grad_norm": 0.6853671073913574,
+      "learning_rate": 0.0001,
+      "loss": 0.8357,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3828365878725591,
+      "grad_norm": 0.6551205515861511,
+      "learning_rate": 0.0001,
+      "loss": 0.7564,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3829222336416581,
+      "grad_norm": 0.6467787623405457,
+      "learning_rate": 0.0001,
+      "loss": 1.2826,
+      "step": 4471
+    },
+    {
+      "epoch": 0.3830078794107571,
+      "grad_norm": 0.7404801845550537,
+      "learning_rate": 0.0001,
+      "loss": 0.8395,
+      "step": 4472
+    },
+    {
+      "epoch": 0.3830935251798561,
+      "grad_norm": 1.0335991382598877,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3831791709489551,
+      "grad_norm": 0.7233255505561829,
+      "learning_rate": 0.0001,
+      "loss": 0.6001,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3832648167180541,
+      "grad_norm": 0.7076817750930786,
+      "learning_rate": 0.0001,
+      "loss": 0.9029,
+      "step": 4475
+    },
+    {
+      "epoch": 0.38335046248715315,
+      "grad_norm": 0.6875801086425781,
+      "learning_rate": 0.0001,
+      "loss": 0.8215,
+      "step": 4476
+    },
+    {
+      "epoch": 0.38343610825625213,
+      "grad_norm": 0.7818890810012817,
+      "learning_rate": 0.0001,
+      "loss": 0.9971,
+      "step": 4477
+    },
+    {
+      "epoch": 0.38352175402535116,
+      "grad_norm": 0.5975618362426758,
+      "learning_rate": 0.0001,
+      "loss": 1.0991,
+      "step": 4478
+    },
+    {
+      "epoch": 0.38360739979445013,
+      "grad_norm": 0.6661436557769775,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 4479
+    },
+    {
+      "epoch": 0.38369304556354916,
+      "grad_norm": 0.6493008732795715,
+      "learning_rate": 0.0001,
+      "loss": 1.1064,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3837786913326482,
+      "grad_norm": 0.47999000549316406,
+      "learning_rate": 0.0001,
+      "loss": 0.6755,
+      "step": 4481
+    },
+    {
+      "epoch": 0.38386433710174717,
+      "grad_norm": 1.3395329713821411,
+      "learning_rate": 0.0001,
+      "loss": 0.8116,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3839499828708462,
+      "grad_norm": 0.6749282479286194,
+      "learning_rate": 0.0001,
+      "loss": 0.9517,
+      "step": 4483
+    },
+    {
+      "epoch": 0.38403562863994517,
+      "grad_norm": 0.799188494682312,
+      "learning_rate": 0.0001,
+      "loss": 0.6121,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3841212744090442,
+      "grad_norm": 0.5424755215644836,
+      "learning_rate": 0.0001,
+      "loss": 0.9608,
+      "step": 4485
+    },
+    {
+      "epoch": 0.3842069201781432,
+      "grad_norm": 0.6288254857063293,
+      "learning_rate": 0.0001,
+      "loss": 0.9267,
+      "step": 4486
+    },
+    {
+      "epoch": 0.3842925659472422,
+      "grad_norm": 0.7635599970817566,
+      "learning_rate": 0.0001,
+      "loss": 0.6747,
+      "step": 4487
+    },
+    {
+      "epoch": 0.38437821171634123,
+      "grad_norm": 1.0328621864318848,
+      "learning_rate": 0.0001,
+      "loss": 0.9957,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3844638574854402,
+      "grad_norm": 0.8343757390975952,
+      "learning_rate": 0.0001,
+      "loss": 1.0193,
+      "step": 4489
+    },
+    {
+      "epoch": 0.38454950325453924,
+      "grad_norm": 0.6381348967552185,
+      "learning_rate": 0.0001,
+      "loss": 0.6562,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3846351490236382,
+      "grad_norm": 0.7582728862762451,
+      "learning_rate": 0.0001,
+      "loss": 0.885,
+      "step": 4491
+    },
+    {
+      "epoch": 0.38472079479273724,
+      "grad_norm": 0.907049834728241,
+      "learning_rate": 0.0001,
+      "loss": 0.8764,
+      "step": 4492
+    },
+    {
+      "epoch": 0.38480644056183627,
+      "grad_norm": 0.5461777448654175,
+      "learning_rate": 0.0001,
+      "loss": 0.8564,
+      "step": 4493
+    },
+    {
+      "epoch": 0.38489208633093525,
+      "grad_norm": 1.1111866235733032,
+      "learning_rate": 0.0001,
+      "loss": 0.8675,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3849777321000343,
+      "grad_norm": 0.5274984240531921,
+      "learning_rate": 0.0001,
+      "loss": 1.1904,
+      "step": 4495
+    },
+    {
+      "epoch": 0.38506337786913325,
+      "grad_norm": 0.8169320821762085,
+      "learning_rate": 0.0001,
+      "loss": 1.0444,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3851490236382323,
+      "grad_norm": 0.7949111461639404,
+      "learning_rate": 0.0001,
+      "loss": 0.5133,
+      "step": 4497
+    },
+    {
+      "epoch": 0.38523466940733125,
+      "grad_norm": 0.4539337456226349,
+      "learning_rate": 0.0001,
+      "loss": 0.3727,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3853203151764303,
+      "grad_norm": 1.0375703573226929,
+      "learning_rate": 0.0001,
+      "loss": 0.8633,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3854059609455293,
+      "grad_norm": 0.483709454536438,
+      "learning_rate": 0.0001,
+      "loss": 0.8429,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3854916067146283,
+      "grad_norm": 0.7662063837051392,
+      "learning_rate": 0.0001,
+      "loss": 0.9912,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3855772524837273,
+      "grad_norm": 1.0457968711853027,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 4502
+    },
+    {
+      "epoch": 0.3856628982528263,
+      "grad_norm": 0.7847151160240173,
+      "learning_rate": 0.0001,
+      "loss": 1.0998,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3857485440219253,
+      "grad_norm": 0.34697046875953674,
+      "learning_rate": 0.0001,
+      "loss": 0.4746,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3858341897910243,
+      "grad_norm": 0.5511472821235657,
+      "learning_rate": 0.0001,
+      "loss": 0.3574,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3859198355601233,
+      "grad_norm": 0.693153977394104,
+      "learning_rate": 0.0001,
+      "loss": 0.9412,
+      "step": 4506
+    },
+    {
+      "epoch": 0.38600548132922236,
+      "grad_norm": 1.1349132061004639,
+      "learning_rate": 0.0001,
+      "loss": 1.0059,
+      "step": 4507
+    },
+    {
+      "epoch": 0.38609112709832133,
+      "grad_norm": 0.8025808334350586,
+      "learning_rate": 0.0001,
+      "loss": 1.0587,
+      "step": 4508
+    },
+    {
+      "epoch": 0.38617677286742036,
+      "grad_norm": 1.3186403512954712,
+      "learning_rate": 0.0001,
+      "loss": 1.2318,
+      "step": 4509
+    },
+    {
+      "epoch": 0.38626241863651933,
+      "grad_norm": 0.8022989630699158,
+      "learning_rate": 0.0001,
+      "loss": 0.6984,
+      "step": 4510
+    },
+    {
+      "epoch": 0.38634806440561836,
+      "grad_norm": 0.8802483677864075,
+      "learning_rate": 0.0001,
+      "loss": 1.3133,
+      "step": 4511
+    },
+    {
+      "epoch": 0.3864337101747174,
+      "grad_norm": 1.0235623121261597,
+      "learning_rate": 0.0001,
+      "loss": 1.1108,
+      "step": 4512
+    },
+    {
+      "epoch": 0.38651935594381637,
+      "grad_norm": 0.5933769345283508,
+      "learning_rate": 0.0001,
+      "loss": 0.6471,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3866050017129154,
+      "grad_norm": 0.7012772560119629,
+      "learning_rate": 0.0001,
+      "loss": 0.621,
+      "step": 4514
+    },
+    {
+      "epoch": 0.38669064748201437,
+      "grad_norm": 0.8313301801681519,
+      "learning_rate": 0.0001,
+      "loss": 1.136,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3867762932511134,
+      "grad_norm": 0.425899863243103,
+      "learning_rate": 0.0001,
+      "loss": 0.2586,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3868619390202124,
+      "grad_norm": 0.7371231913566589,
+      "learning_rate": 0.0001,
+      "loss": 0.976,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3869475847893114,
+      "grad_norm": 0.5862066745758057,
+      "learning_rate": 0.0001,
+      "loss": 1.0863,
+      "step": 4518
+    },
+    {
+      "epoch": 0.38703323055841043,
+      "grad_norm": 1.2632347345352173,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3871188763275094,
+      "grad_norm": 0.8903829455375671,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 4520
+    },
+    {
+      "epoch": 0.38720452209660844,
+      "grad_norm": 0.3292497396469116,
+      "learning_rate": 0.0001,
+      "loss": 0.5628,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3872901678657074,
+      "grad_norm": 0.7471050024032593,
+      "learning_rate": 0.0001,
+      "loss": 1.0024,
+      "step": 4522
+    },
+    {
+      "epoch": 0.38737581363480644,
+      "grad_norm": 0.7371292114257812,
+      "learning_rate": 0.0001,
+      "loss": 0.763,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3874614594039055,
+      "grad_norm": 0.6767558455467224,
+      "learning_rate": 0.0001,
+      "loss": 1.1172,
+      "step": 4524
+    },
+    {
+      "epoch": 0.38754710517300445,
+      "grad_norm": 1.1880598068237305,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3876327509421035,
+      "grad_norm": 0.8318496942520142,
+      "learning_rate": 0.0001,
+      "loss": 0.8416,
+      "step": 4526
+    },
+    {
+      "epoch": 0.38771839671120245,
+      "grad_norm": 0.7031965255737305,
+      "learning_rate": 0.0001,
+      "loss": 0.6045,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3878040424803015,
+      "grad_norm": 0.6509886384010315,
+      "learning_rate": 0.0001,
+      "loss": 1.144,
+      "step": 4528
+    },
+    {
+      "epoch": 0.38788968824940045,
+      "grad_norm": 0.6569823622703552,
+      "learning_rate": 0.0001,
+      "loss": 0.4935,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3879753340184995,
+      "grad_norm": 0.7238826155662537,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3880609797875985,
+      "grad_norm": 0.19032521545886993,
+      "learning_rate": 0.0001,
+      "loss": 0.7101,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3881466255566975,
+      "grad_norm": 0.6859326958656311,
+      "learning_rate": 0.0001,
+      "loss": 0.7432,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3882322713257965,
+      "grad_norm": 0.6488766074180603,
+      "learning_rate": 0.0001,
+      "loss": 0.8979,
+      "step": 4533
+    },
+    {
+      "epoch": 0.3883179170948955,
+      "grad_norm": 0.5580387115478516,
+      "learning_rate": 0.0001,
+      "loss": 0.3617,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3884035628639945,
+      "grad_norm": 0.6670434474945068,
+      "learning_rate": 0.0001,
+      "loss": 1.0355,
+      "step": 4535
+    },
+    {
+      "epoch": 0.38848920863309355,
+      "grad_norm": 1.2255841493606567,
+      "learning_rate": 0.0001,
+      "loss": 1.0009,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3885748544021925,
+      "grad_norm": 0.7784870862960815,
+      "learning_rate": 0.0001,
+      "loss": 0.3845,
+      "step": 4537
+    },
+    {
+      "epoch": 0.38866050017129156,
+      "grad_norm": 0.9181932806968689,
+      "learning_rate": 0.0001,
+      "loss": 0.8085,
+      "step": 4538
+    },
+    {
+      "epoch": 0.38874614594039053,
+      "grad_norm": 1.114902377128601,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 4539
+    },
+    {
+      "epoch": 0.38883179170948956,
+      "grad_norm": 0.6498281955718994,
+      "learning_rate": 0.0001,
+      "loss": 0.662,
+      "step": 4540
+    },
+    {
+      "epoch": 0.38891743747858853,
+      "grad_norm": 1.1569596529006958,
+      "learning_rate": 0.0001,
+      "loss": 0.7862,
+      "step": 4541
+    },
+    {
+      "epoch": 0.38900308324768756,
+      "grad_norm": 0.8542700409889221,
+      "learning_rate": 0.0001,
+      "loss": 0.6093,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3890887290167866,
+      "grad_norm": 0.6080235838890076,
+      "learning_rate": 0.0001,
+      "loss": 1.0807,
+      "step": 4543
+    },
+    {
+      "epoch": 0.38917437478588557,
+      "grad_norm": 0.4939481317996979,
+      "learning_rate": 0.0001,
+      "loss": 0.8924,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3892600205549846,
+      "grad_norm": 0.9777079820632935,
+      "learning_rate": 0.0001,
+      "loss": 1.3546,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3893456663240836,
+      "grad_norm": 0.48583459854125977,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3894313120931826,
+      "grad_norm": 0.5963572263717651,
+      "learning_rate": 0.0001,
+      "loss": 1.0859,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3895169578622816,
+      "grad_norm": 0.675934374332428,
+      "learning_rate": 0.0001,
+      "loss": 1.0697,
+      "step": 4548
+    },
+    {
+      "epoch": 0.3896026036313806,
+      "grad_norm": 0.48818981647491455,
+      "learning_rate": 0.0001,
+      "loss": 0.875,
+      "step": 4549
+    },
+    {
+      "epoch": 0.38968824940047964,
+      "grad_norm": 0.9124215245246887,
+      "learning_rate": 0.0001,
+      "loss": 1.1191,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3897738951695786,
+      "grad_norm": 0.4134933054447174,
+      "learning_rate": 0.0001,
+      "loss": 0.5308,
+      "step": 4551
+    },
+    {
+      "epoch": 0.38985954093867764,
+      "grad_norm": 0.7630157470703125,
+      "learning_rate": 0.0001,
+      "loss": 1.0,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3899451867077766,
+      "grad_norm": 1.1511067152023315,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 4553
+    },
+    {
+      "epoch": 0.39003083247687564,
+      "grad_norm": 0.7789441347122192,
+      "learning_rate": 0.0001,
+      "loss": 0.7616,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3901164782459747,
+      "grad_norm": 0.5671170949935913,
+      "learning_rate": 0.0001,
+      "loss": 1.0592,
+      "step": 4555
+    },
+    {
+      "epoch": 0.39020212401507365,
+      "grad_norm": 0.4965318739414215,
+      "learning_rate": 0.0001,
+      "loss": 1.0451,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3902877697841727,
+      "grad_norm": 0.7542549967765808,
+      "learning_rate": 0.0001,
+      "loss": 0.7632,
+      "step": 4557
+    },
+    {
+      "epoch": 0.39037341555327165,
+      "grad_norm": 0.612166702747345,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 4558
+    },
+    {
+      "epoch": 0.3904590613223707,
+      "grad_norm": 0.3363237679004669,
+      "learning_rate": 0.0001,
+      "loss": 0.6084,
+      "step": 4559
+    },
+    {
+      "epoch": 0.39054470709146966,
+      "grad_norm": 0.8014049530029297,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3906303528605687,
+      "grad_norm": 0.46430957317352295,
+      "learning_rate": 0.0001,
+      "loss": 0.4791,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3907159986296677,
+      "grad_norm": 0.8592241406440735,
+      "learning_rate": 0.0001,
+      "loss": 1.2844,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3908016443987667,
+      "grad_norm": 0.7636821866035461,
+      "learning_rate": 0.0001,
+      "loss": 0.7578,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3908872901678657,
+      "grad_norm": 0.6857783198356628,
+      "learning_rate": 0.0001,
+      "loss": 2.0162,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3909729359369647,
+      "grad_norm": 0.48289498686790466,
+      "learning_rate": 0.0001,
+      "loss": 0.8729,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3910585817060637,
+      "grad_norm": 0.7892183065414429,
+      "learning_rate": 0.0001,
+      "loss": 0.8407,
+      "step": 4566
+    },
+    {
+      "epoch": 0.39114422747516275,
+      "grad_norm": 1.0878921747207642,
+      "learning_rate": 0.0001,
+      "loss": 0.9803,
+      "step": 4567
+    },
+    {
+      "epoch": 0.3912298732442617,
+      "grad_norm": 1.041853427886963,
+      "learning_rate": 0.0001,
+      "loss": 0.9621,
+      "step": 4568
+    },
+    {
+      "epoch": 0.39131551901336076,
+      "grad_norm": 1.2361027002334595,
+      "learning_rate": 0.0001,
+      "loss": 1.1109,
+      "step": 4569
+    },
+    {
+      "epoch": 0.39140116478245973,
+      "grad_norm": 0.39380449056625366,
+      "learning_rate": 0.0001,
+      "loss": 0.5085,
+      "step": 4570
+    },
+    {
+      "epoch": 0.39148681055155876,
+      "grad_norm": 1.2779080867767334,
+      "learning_rate": 0.0001,
+      "loss": 1.2102,
+      "step": 4571
+    },
+    {
+      "epoch": 0.39157245632065774,
+      "grad_norm": 0.9902412295341492,
+      "learning_rate": 0.0001,
+      "loss": 1.3534,
+      "step": 4572
+    },
+    {
+      "epoch": 0.39165810208975677,
+      "grad_norm": 1.0671193599700928,
+      "learning_rate": 0.0001,
+      "loss": 0.7,
+      "step": 4573
+    },
+    {
+      "epoch": 0.3917437478588558,
+      "grad_norm": 1.0607682466506958,
+      "learning_rate": 0.0001,
+      "loss": 0.8685,
+      "step": 4574
+    },
+    {
+      "epoch": 0.39182939362795477,
+      "grad_norm": 0.9212198853492737,
+      "learning_rate": 0.0001,
+      "loss": 0.8897,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3919150393970538,
+      "grad_norm": 0.5906891226768494,
+      "learning_rate": 0.0001,
+      "loss": 0.8179,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3920006851661528,
+      "grad_norm": 1.0687217712402344,
+      "learning_rate": 0.0001,
+      "loss": 1.0655,
+      "step": 4577
+    },
+    {
+      "epoch": 0.3920863309352518,
+      "grad_norm": 0.7661501169204712,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3921719767043508,
+      "grad_norm": 1.0922819375991821,
+      "learning_rate": 0.0001,
+      "loss": 1.0824,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3922576224734498,
+      "grad_norm": 0.5054302215576172,
+      "learning_rate": 0.0001,
+      "loss": 1.2347,
+      "step": 4580
+    },
+    {
+      "epoch": 0.39234326824254884,
+      "grad_norm": 0.8717608451843262,
+      "learning_rate": 0.0001,
+      "loss": 1.0292,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3924289140116478,
+      "grad_norm": 0.8337442874908447,
+      "learning_rate": 0.0001,
+      "loss": 0.6861,
+      "step": 4582
+    },
+    {
+      "epoch": 0.39251455978074684,
+      "grad_norm": 0.6601764559745789,
+      "learning_rate": 0.0001,
+      "loss": 0.9751,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3926002055498458,
+      "grad_norm": 0.6832159161567688,
+      "learning_rate": 0.0001,
+      "loss": 0.8863,
+      "step": 4584
+    },
+    {
+      "epoch": 0.39268585131894485,
+      "grad_norm": 0.5074254274368286,
+      "learning_rate": 0.0001,
+      "loss": 0.8644,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3927714970880439,
+      "grad_norm": 0.5838516354560852,
+      "learning_rate": 0.0001,
+      "loss": 1.0144,
+      "step": 4586
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 0.6094033718109131,
+      "learning_rate": 0.0001,
+      "loss": 1.022,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3929427886262419,
+      "grad_norm": 0.5771576166152954,
+      "learning_rate": 0.0001,
+      "loss": 1.169,
+      "step": 4588
+    },
+    {
+      "epoch": 0.39302843439534085,
+      "grad_norm": 0.8678125739097595,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3931140801644399,
+      "grad_norm": 0.7306006550788879,
+      "learning_rate": 0.0001,
+      "loss": 0.701,
+      "step": 4590
+    },
+    {
+      "epoch": 0.39319972593353886,
+      "grad_norm": 0.6020304560661316,
+      "learning_rate": 0.0001,
+      "loss": 0.4449,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3932853717026379,
+      "grad_norm": 0.47850239276885986,
+      "learning_rate": 0.0001,
+      "loss": 0.813,
+      "step": 4592
+    },
+    {
+      "epoch": 0.3933710174717369,
+      "grad_norm": 0.733709990978241,
+      "learning_rate": 0.0001,
+      "loss": 1.749,
+      "step": 4593
+    },
+    {
+      "epoch": 0.3934566632408359,
+      "grad_norm": 1.03411865234375,
+      "learning_rate": 0.0001,
+      "loss": 1.0229,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3935423090099349,
+      "grad_norm": 0.809097409248352,
+      "learning_rate": 0.0001,
+      "loss": 0.7588,
+      "step": 4595
+    },
+    {
+      "epoch": 0.3936279547790339,
+      "grad_norm": 0.8277100324630737,
+      "learning_rate": 0.0001,
+      "loss": 1.0199,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3937136005481329,
+      "grad_norm": 1.2887567281723022,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 4597
+    },
+    {
+      "epoch": 0.39379924631723195,
+      "grad_norm": 0.5222105383872986,
+      "learning_rate": 0.0001,
+      "loss": 1.0151,
+      "step": 4598
+    },
+    {
+      "epoch": 0.39388489208633093,
+      "grad_norm": 1.014047384262085,
+      "learning_rate": 0.0001,
+      "loss": 1.1067,
+      "step": 4599
+    },
+    {
+      "epoch": 0.39397053785542996,
+      "grad_norm": 1.089269995689392,
+      "learning_rate": 0.0001,
+      "loss": 1.056,
+      "step": 4600
+    },
+    {
+      "epoch": 0.39405618362452893,
+      "grad_norm": 1.0358595848083496,
+      "learning_rate": 0.0001,
+      "loss": 1.1711,
+      "step": 4601
+    },
+    {
+      "epoch": 0.39414182939362796,
+      "grad_norm": 0.6309729218482971,
+      "learning_rate": 0.0001,
+      "loss": 0.3448,
+      "step": 4602
+    },
+    {
+      "epoch": 0.39422747516272694,
+      "grad_norm": 1.5494352579116821,
+      "learning_rate": 0.0001,
+      "loss": 1.0269,
+      "step": 4603
+    },
+    {
+      "epoch": 0.39431312093182597,
+      "grad_norm": 0.31894877552986145,
+      "learning_rate": 0.0001,
+      "loss": 0.3972,
+      "step": 4604
+    },
+    {
+      "epoch": 0.394398766700925,
+      "grad_norm": 0.861693799495697,
+      "learning_rate": 0.0001,
+      "loss": 0.57,
+      "step": 4605
+    },
+    {
+      "epoch": 0.39448441247002397,
+      "grad_norm": 1.1014204025268555,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 4606
+    },
+    {
+      "epoch": 0.394570058239123,
+      "grad_norm": 0.805317223072052,
+      "learning_rate": 0.0001,
+      "loss": 1.094,
+      "step": 4607
+    },
+    {
+      "epoch": 0.394655704008222,
+      "grad_norm": 1.253690481185913,
+      "learning_rate": 0.0001,
+      "loss": 1.1523,
+      "step": 4608
+    },
+    {
+      "epoch": 0.394741349777321,
+      "grad_norm": 0.7644290328025818,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 4609
+    },
+    {
+      "epoch": 0.39482699554642003,
+      "grad_norm": 0.5651330947875977,
+      "learning_rate": 0.0001,
+      "loss": 1.0064,
+      "step": 4610
+    },
+    {
+      "epoch": 0.394912641315519,
+      "grad_norm": 0.37226855754852295,
+      "learning_rate": 0.0001,
+      "loss": 0.4898,
+      "step": 4611
+    },
+    {
+      "epoch": 0.39499828708461804,
+      "grad_norm": 0.7233548760414124,
+      "learning_rate": 0.0001,
+      "loss": 1.0942,
+      "step": 4612
+    },
+    {
+      "epoch": 0.395083932853717,
+      "grad_norm": 0.8628584146499634,
+      "learning_rate": 0.0001,
+      "loss": 0.7877,
+      "step": 4613
+    },
+    {
+      "epoch": 0.39516957862281604,
+      "grad_norm": 0.8992170095443726,
+      "learning_rate": 0.0001,
+      "loss": 1.026,
+      "step": 4614
+    },
+    {
+      "epoch": 0.395255224391915,
+      "grad_norm": 1.1055327653884888,
+      "learning_rate": 0.0001,
+      "loss": 1.3081,
+      "step": 4615
+    },
+    {
+      "epoch": 0.39534087016101405,
+      "grad_norm": 0.5466746091842651,
+      "learning_rate": 0.0001,
+      "loss": 1.1112,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3954265159301131,
+      "grad_norm": 0.7546491622924805,
+      "learning_rate": 0.0001,
+      "loss": 0.7513,
+      "step": 4617
+    },
+    {
+      "epoch": 0.39551216169921205,
+      "grad_norm": 1.0645891427993774,
+      "learning_rate": 0.0001,
+      "loss": 1.0628,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3955978074683111,
+      "grad_norm": 0.5530161261558533,
+      "learning_rate": 0.0001,
+      "loss": 0.6487,
+      "step": 4619
+    },
+    {
+      "epoch": 0.39568345323741005,
+      "grad_norm": 0.6212671995162964,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3957690990065091,
+      "grad_norm": 1.1372240781784058,
+      "learning_rate": 0.0001,
+      "loss": 1.1499,
+      "step": 4621
+    },
+    {
+      "epoch": 0.39585474477560806,
+      "grad_norm": 0.9341187477111816,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3959403905447071,
+      "grad_norm": 0.5911365747451782,
+      "learning_rate": 0.0001,
+      "loss": 0.2433,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3960260363138061,
+      "grad_norm": 0.5436334013938904,
+      "learning_rate": 0.0001,
+      "loss": 1.0255,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3961116820829051,
+      "grad_norm": 0.7289847731590271,
+      "learning_rate": 0.0001,
+      "loss": 0.8335,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3961973278520041,
+      "grad_norm": 0.8696873188018799,
+      "learning_rate": 0.0001,
+      "loss": 1.0272,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3962829736211031,
+      "grad_norm": 0.9004592895507812,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3963686193902021,
+      "grad_norm": 0.2920278310775757,
+      "learning_rate": 0.0001,
+      "loss": 0.2724,
+      "step": 4628
+    },
+    {
+      "epoch": 0.39645426515930116,
+      "grad_norm": 0.7239041924476624,
+      "learning_rate": 0.0001,
+      "loss": 0.4639,
+      "step": 4629
+    },
+    {
+      "epoch": 0.39653991092840013,
+      "grad_norm": 0.5800740122795105,
+      "learning_rate": 0.0001,
+      "loss": 1.2922,
+      "step": 4630
+    },
+    {
+      "epoch": 0.39662555669749916,
+      "grad_norm": 0.9600546956062317,
+      "learning_rate": 0.0001,
+      "loss": 0.9038,
+      "step": 4631
+    },
+    {
+      "epoch": 0.39671120246659813,
+      "grad_norm": 0.9565032124519348,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 4632
+    },
+    {
+      "epoch": 0.39679684823569716,
+      "grad_norm": 0.5583470463752747,
+      "learning_rate": 0.0001,
+      "loss": 1.1337,
+      "step": 4633
+    },
+    {
+      "epoch": 0.39688249400479614,
+      "grad_norm": 0.3596307337284088,
+      "learning_rate": 0.0001,
+      "loss": 0.4076,
+      "step": 4634
+    },
+    {
+      "epoch": 0.39696813977389517,
+      "grad_norm": 0.7809063196182251,
+      "learning_rate": 0.0001,
+      "loss": 1.1821,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3970537855429942,
+      "grad_norm": 0.6668574810028076,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 4636
+    },
+    {
+      "epoch": 0.39713943131209317,
+      "grad_norm": 0.8529664278030396,
+      "learning_rate": 0.0001,
+      "loss": 1.0778,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3972250770811922,
+      "grad_norm": 1.8954570293426514,
+      "learning_rate": 0.0001,
+      "loss": 0.5056,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3973107228502912,
+      "grad_norm": 0.7033258676528931,
+      "learning_rate": 0.0001,
+      "loss": 0.9847,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3973963686193902,
+      "grad_norm": 1.364327311515808,
+      "learning_rate": 0.0001,
+      "loss": 1.396,
+      "step": 4640
+    },
+    {
+      "epoch": 0.39748201438848924,
+      "grad_norm": 1.5456807613372803,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3975676601575882,
+      "grad_norm": 0.8810327053070068,
+      "learning_rate": 0.0001,
+      "loss": 1.3555,
+      "step": 4642
+    },
+    {
+      "epoch": 0.39765330592668724,
+      "grad_norm": 0.8105685114860535,
+      "learning_rate": 0.0001,
+      "loss": 0.9103,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3977389516957862,
+      "grad_norm": 0.6442139148712158,
+      "learning_rate": 0.0001,
+      "loss": 1.3171,
+      "step": 4644
+    },
+    {
+      "epoch": 0.39782459746488524,
+      "grad_norm": 1.0930134057998657,
+      "learning_rate": 0.0001,
+      "loss": 0.9574,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3979102432339842,
+      "grad_norm": 1.1811105012893677,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 4646
+    },
+    {
+      "epoch": 0.39799588900308325,
+      "grad_norm": 0.8172235488891602,
+      "learning_rate": 0.0001,
+      "loss": 0.8297,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3980815347721823,
+      "grad_norm": 0.7806366682052612,
+      "learning_rate": 0.0001,
+      "loss": 0.8406,
+      "step": 4648
+    },
+    {
+      "epoch": 0.39816718054128125,
+      "grad_norm": 0.7758505344390869,
+      "learning_rate": 0.0001,
+      "loss": 0.7801,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3982528263103803,
+      "grad_norm": 0.42445358633995056,
+      "learning_rate": 0.0001,
+      "loss": 0.6815,
+      "step": 4650
+    },
+    {
+      "epoch": 0.39833847207947926,
+      "grad_norm": 0.7530607581138611,
+      "learning_rate": 0.0001,
+      "loss": 0.6894,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3984241178485783,
+      "grad_norm": 0.47612929344177246,
+      "learning_rate": 0.0001,
+      "loss": 0.6395,
+      "step": 4652
+    },
+    {
+      "epoch": 0.39850976361767726,
+      "grad_norm": 0.8760970234870911,
+      "learning_rate": 0.0001,
+      "loss": 0.8124,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3985954093867763,
+      "grad_norm": 0.5735673308372498,
+      "learning_rate": 0.0001,
+      "loss": 0.7104,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3986810551558753,
+      "grad_norm": 0.6057902574539185,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3987667009249743,
+      "grad_norm": 1.0075510740280151,
+      "learning_rate": 0.0001,
+      "loss": 1.9085,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3988523466940733,
+      "grad_norm": 0.4877760112285614,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3989379924631723,
+      "grad_norm": 0.8807364106178284,
+      "learning_rate": 0.0001,
+      "loss": 1.1637,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3990236382322713,
+      "grad_norm": 0.8752393126487732,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 4659
+    },
+    {
+      "epoch": 0.39910928400137036,
+      "grad_norm": 1.0317091941833496,
+      "learning_rate": 0.0001,
+      "loss": 1.3793,
+      "step": 4660
+    },
+    {
+      "epoch": 0.39919492977046933,
+      "grad_norm": 0.6151207089424133,
+      "learning_rate": 0.0001,
+      "loss": 1.126,
+      "step": 4661
+    },
+    {
+      "epoch": 0.39928057553956836,
+      "grad_norm": 0.5901468396186829,
+      "learning_rate": 0.0001,
+      "loss": 0.8294,
+      "step": 4662
+    },
+    {
+      "epoch": 0.39936622130866734,
+      "grad_norm": 0.5240193009376526,
+      "learning_rate": 0.0001,
+      "loss": 0.5696,
+      "step": 4663
+    },
+    {
+      "epoch": 0.39945186707776636,
+      "grad_norm": 1.3308981657028198,
+      "learning_rate": 0.0001,
+      "loss": 1.0729,
+      "step": 4664
+    },
+    {
+      "epoch": 0.39953751284686534,
+      "grad_norm": 0.867911696434021,
+      "learning_rate": 0.0001,
+      "loss": 1.1626,
+      "step": 4665
+    },
+    {
+      "epoch": 0.39962315861596437,
+      "grad_norm": 0.49453040957450867,
+      "learning_rate": 0.0001,
+      "loss": 0.6337,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3997088043850634,
+      "grad_norm": 1.635938048362732,
+      "learning_rate": 0.0001,
+      "loss": 1.1634,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3997944501541624,
+      "grad_norm": 0.5850088596343994,
+      "learning_rate": 0.0001,
+      "loss": 1.0551,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3998800959232614,
+      "grad_norm": 1.0036717653274536,
+      "learning_rate": 0.0001,
+      "loss": 1.2145,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3999657416923604,
+      "grad_norm": 0.7474343776702881,
+      "learning_rate": 0.0001,
+      "loss": 1.2264,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4000513874614594,
+      "grad_norm": 0.6406418681144714,
+      "learning_rate": 0.0001,
+      "loss": 0.6577,
+      "step": 4671
+    },
+    {
+      "epoch": 0.40013703323055844,
+      "grad_norm": 0.8581080436706543,
+      "learning_rate": 0.0001,
+      "loss": 1.2986,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4002226789996574,
+      "grad_norm": 0.9454993605613708,
+      "learning_rate": 0.0001,
+      "loss": 2.2725,
+      "step": 4673
+    },
+    {
+      "epoch": 0.40030832476875644,
+      "grad_norm": 1.055612325668335,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4003939705378554,
+      "grad_norm": 0.9071846008300781,
+      "learning_rate": 0.0001,
+      "loss": 0.795,
+      "step": 4675
+    },
+    {
+      "epoch": 0.40047961630695444,
+      "grad_norm": 0.6660609245300293,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4005652620760534,
+      "grad_norm": 0.5846877694129944,
+      "learning_rate": 0.0001,
+      "loss": 0.4768,
+      "step": 4677
+    },
+    {
+      "epoch": 0.40065090784515245,
+      "grad_norm": 0.6096349954605103,
+      "learning_rate": 0.0001,
+      "loss": 0.5209,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4007365536142515,
+      "grad_norm": 0.954208493232727,
+      "learning_rate": 0.0001,
+      "loss": 0.9977,
+      "step": 4679
+    },
+    {
+      "epoch": 0.40082219938335045,
+      "grad_norm": 0.46659907698631287,
+      "learning_rate": 0.0001,
+      "loss": 0.814,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4009078451524495,
+      "grad_norm": 0.6562754511833191,
+      "learning_rate": 0.0001,
+      "loss": 1.3329,
+      "step": 4681
+    },
+    {
+      "epoch": 0.40099349092154846,
+      "grad_norm": 0.6506263613700867,
+      "learning_rate": 0.0001,
+      "loss": 1.1644,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4010791366906475,
+      "grad_norm": 0.5695533752441406,
+      "learning_rate": 0.0001,
+      "loss": 1.0826,
+      "step": 4683
+    },
+    {
+      "epoch": 0.4011647824597465,
+      "grad_norm": 0.6140940189361572,
+      "learning_rate": 0.0001,
+      "loss": 1.0419,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4012504282288455,
+      "grad_norm": 0.8675922155380249,
+      "learning_rate": 0.0001,
+      "loss": 1.1168,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4013360739979445,
+      "grad_norm": 0.8234465718269348,
+      "learning_rate": 0.0001,
+      "loss": 0.398,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4014217197670435,
+      "grad_norm": 0.7795776724815369,
+      "learning_rate": 0.0001,
+      "loss": 1.0022,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4015073655361425,
+      "grad_norm": 0.600053071975708,
+      "learning_rate": 0.0001,
+      "loss": 0.7602,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4015930113052415,
+      "grad_norm": 0.5815989375114441,
+      "learning_rate": 0.0001,
+      "loss": 0.5935,
+      "step": 4689
+    },
+    {
+      "epoch": 0.40167865707434053,
+      "grad_norm": 0.9028575420379639,
+      "learning_rate": 0.0001,
+      "loss": 0.6038,
+      "step": 4690
+    },
+    {
+      "epoch": 0.40176430284343956,
+      "grad_norm": 0.5208615064620972,
+      "learning_rate": 0.0001,
+      "loss": 0.5167,
+      "step": 4691
+    },
+    {
+      "epoch": 0.40184994861253853,
+      "grad_norm": 1.1257128715515137,
+      "learning_rate": 0.0001,
+      "loss": 1.0744,
+      "step": 4692
+    },
+    {
+      "epoch": 0.40193559438163756,
+      "grad_norm": 0.263847678899765,
+      "learning_rate": 0.0001,
+      "loss": 0.4176,
+      "step": 4693
+    },
+    {
+      "epoch": 0.40202124015073654,
+      "grad_norm": 0.7006592750549316,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 4694
+    },
+    {
+      "epoch": 0.40210688591983557,
+      "grad_norm": 0.5425605773925781,
+      "learning_rate": 0.0001,
+      "loss": 0.4914,
+      "step": 4695
+    },
+    {
+      "epoch": 0.40219253168893454,
+      "grad_norm": 0.5710000395774841,
+      "learning_rate": 0.0001,
+      "loss": 0.7134,
+      "step": 4696
+    },
+    {
+      "epoch": 0.40227817745803357,
+      "grad_norm": 0.3322749137878418,
+      "learning_rate": 0.0001,
+      "loss": 0.3709,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4023638232271326,
+      "grad_norm": 0.920876681804657,
+      "learning_rate": 0.0001,
+      "loss": 1.2138,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4024494689962316,
+      "grad_norm": 1.0299707651138306,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4025351147653306,
+      "grad_norm": 1.1246761083602905,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4026207605344296,
+      "grad_norm": 0.7256906628608704,
+      "learning_rate": 0.0001,
+      "loss": 0.7124,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4027064063035286,
+      "grad_norm": 0.9372466802597046,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 4702
+    },
+    {
+      "epoch": 0.40279205207262764,
+      "grad_norm": 0.7934892773628235,
+      "learning_rate": 0.0001,
+      "loss": 0.6842,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4028776978417266,
+      "grad_norm": 0.6845753192901611,
+      "learning_rate": 0.0001,
+      "loss": 1.158,
+      "step": 4704
+    },
+    {
+      "epoch": 0.40296334361082564,
+      "grad_norm": 0.5154064297676086,
+      "learning_rate": 0.0001,
+      "loss": 1.0893,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4030489893799246,
+      "grad_norm": 0.9355843663215637,
+      "learning_rate": 0.0001,
+      "loss": 0.7204,
+      "step": 4706
+    },
+    {
+      "epoch": 0.40313463514902365,
+      "grad_norm": 0.8520399928092957,
+      "learning_rate": 0.0001,
+      "loss": 0.7624,
+      "step": 4707
+    },
+    {
+      "epoch": 0.4032202809181226,
+      "grad_norm": 1.0806037187576294,
+      "learning_rate": 0.0001,
+      "loss": 1.1424,
+      "step": 4708
+    },
+    {
+      "epoch": 0.40330592668722165,
+      "grad_norm": 0.5055833458900452,
+      "learning_rate": 0.0001,
+      "loss": 0.9077,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4033915724563207,
+      "grad_norm": 0.8784946203231812,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 4710
+    },
+    {
+      "epoch": 0.40347721822541965,
+      "grad_norm": 0.8600638508796692,
+      "learning_rate": 0.0001,
+      "loss": 0.9311,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4035628639945187,
+      "grad_norm": 0.6108738780021667,
+      "learning_rate": 0.0001,
+      "loss": 1.0174,
+      "step": 4712
+    },
+    {
+      "epoch": 0.40364850976361766,
+      "grad_norm": 1.2778152227401733,
+      "learning_rate": 0.0001,
+      "loss": 1.1631,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4037341555327167,
+      "grad_norm": 0.5548070669174194,
+      "learning_rate": 0.0001,
+      "loss": 0.9558,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4038198013018157,
+      "grad_norm": 0.5294012427330017,
+      "learning_rate": 0.0001,
+      "loss": 0.5324,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4039054470709147,
+      "grad_norm": 0.648986279964447,
+      "learning_rate": 0.0001,
+      "loss": 0.7916,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4039910928400137,
+      "grad_norm": 0.983928918838501,
+      "learning_rate": 0.0001,
+      "loss": 1.7935,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4040767386091127,
+      "grad_norm": 1.42551589012146,
+      "learning_rate": 0.0001,
+      "loss": 1.2819,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4041623843782117,
+      "grad_norm": 0.9866476655006409,
+      "learning_rate": 0.0001,
+      "loss": 0.8961,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4042480301473107,
+      "grad_norm": 1.0790560245513916,
+      "learning_rate": 0.0001,
+      "loss": 0.725,
+      "step": 4720
+    },
+    {
+      "epoch": 0.40433367591640973,
+      "grad_norm": 0.6365882158279419,
+      "learning_rate": 0.0001,
+      "loss": 0.7405,
+      "step": 4721
+    },
+    {
+      "epoch": 0.40441932168550876,
+      "grad_norm": 0.7853591442108154,
+      "learning_rate": 0.0001,
+      "loss": 1.112,
+      "step": 4722
+    },
+    {
+      "epoch": 0.40450496745460773,
+      "grad_norm": 0.7692604064941406,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 4723
+    },
+    {
+      "epoch": 0.40459061322370676,
+      "grad_norm": 0.5621455907821655,
+      "learning_rate": 0.0001,
+      "loss": 0.595,
+      "step": 4724
+    },
+    {
+      "epoch": 0.40467625899280574,
+      "grad_norm": 0.8602408766746521,
+      "learning_rate": 0.0001,
+      "loss": 0.8191,
+      "step": 4725
+    },
+    {
+      "epoch": 0.40476190476190477,
+      "grad_norm": 0.6032326221466064,
+      "learning_rate": 0.0001,
+      "loss": 1.0206,
+      "step": 4726
+    },
+    {
+      "epoch": 0.40484755053100374,
+      "grad_norm": 0.9670132398605347,
+      "learning_rate": 0.0001,
+      "loss": 0.8202,
+      "step": 4727
+    },
+    {
+      "epoch": 0.40493319630010277,
+      "grad_norm": 1.0621006488800049,
+      "learning_rate": 0.0001,
+      "loss": 1.0393,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4050188420692018,
+      "grad_norm": 0.7413795590400696,
+      "learning_rate": 0.0001,
+      "loss": 0.817,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4051044878383008,
+      "grad_norm": 0.679947018623352,
+      "learning_rate": 0.0001,
+      "loss": 0.8868,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4051901336073998,
+      "grad_norm": 0.5018595457077026,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4052757793764988,
+      "grad_norm": 0.46053364872932434,
+      "learning_rate": 0.0001,
+      "loss": 0.8007,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4053614251455978,
+      "grad_norm": 0.5956481099128723,
+      "learning_rate": 0.0001,
+      "loss": 0.5954,
+      "step": 4733
+    },
+    {
+      "epoch": 0.40544707091469684,
+      "grad_norm": 0.7805513143539429,
+      "learning_rate": 0.0001,
+      "loss": 0.8125,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4055327166837958,
+      "grad_norm": 0.5876202583312988,
+      "learning_rate": 0.0001,
+      "loss": 0.8228,
+      "step": 4735
+    },
+    {
+      "epoch": 0.40561836245289484,
+      "grad_norm": 0.4977235794067383,
+      "learning_rate": 0.0001,
+      "loss": 0.4537,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4057040082219938,
+      "grad_norm": 1.0475867986679077,
+      "learning_rate": 0.0001,
+      "loss": 1.1442,
+      "step": 4737
+    },
+    {
+      "epoch": 0.40578965399109285,
+      "grad_norm": 0.8012874126434326,
+      "learning_rate": 0.0001,
+      "loss": 0.7527,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4058752997601918,
+      "grad_norm": 0.7495686411857605,
+      "learning_rate": 0.0001,
+      "loss": 1.337,
+      "step": 4739
+    },
+    {
+      "epoch": 0.40596094552929085,
+      "grad_norm": 1.046452522277832,
+      "learning_rate": 0.0001,
+      "loss": 0.8463,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4060465912983899,
+      "grad_norm": 0.6658312678337097,
+      "learning_rate": 0.0001,
+      "loss": 0.566,
+      "step": 4741
+    },
+    {
+      "epoch": 0.40613223706748885,
+      "grad_norm": 0.732349693775177,
+      "learning_rate": 0.0001,
+      "loss": 0.8609,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4062178828365879,
+      "grad_norm": 0.9622615575790405,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 4743
+    },
+    {
+      "epoch": 0.40630352860568686,
+      "grad_norm": 0.5880265831947327,
+      "learning_rate": 0.0001,
+      "loss": 1.0426,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4063891743747859,
+      "grad_norm": 0.8791993260383606,
+      "learning_rate": 0.0001,
+      "loss": 0.7847,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4064748201438849,
+      "grad_norm": 0.674696683883667,
+      "learning_rate": 0.0001,
+      "loss": 0.6024,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4065604659129839,
+      "grad_norm": 0.8567352890968323,
+      "learning_rate": 0.0001,
+      "loss": 1.141,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4066461116820829,
+      "grad_norm": 0.9515517354011536,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4067317574511819,
+      "grad_norm": 0.33974307775497437,
+      "learning_rate": 0.0001,
+      "loss": 0.4199,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4068174032202809,
+      "grad_norm": 1.349760890007019,
+      "learning_rate": 0.0001,
+      "loss": 1.3708,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4069030489893799,
+      "grad_norm": 0.4225471317768097,
+      "learning_rate": 0.0001,
+      "loss": 0.6172,
+      "step": 4751
+    },
+    {
+      "epoch": 0.40698869475847893,
+      "grad_norm": 0.564050018787384,
+      "learning_rate": 0.0001,
+      "loss": 1.3571,
+      "step": 4752
+    },
+    {
+      "epoch": 0.40707434052757796,
+      "grad_norm": 0.7333709597587585,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 4753
+    },
+    {
+      "epoch": 0.40715998629667693,
+      "grad_norm": 0.9997852444648743,
+      "learning_rate": 0.0001,
+      "loss": 0.9839,
+      "step": 4754
+    },
+    {
+      "epoch": 0.40724563206577596,
+      "grad_norm": 1.109707236289978,
+      "learning_rate": 0.0001,
+      "loss": 1.067,
+      "step": 4755
+    },
+    {
+      "epoch": 0.40733127783487494,
+      "grad_norm": 1.1939632892608643,
+      "learning_rate": 0.0001,
+      "loss": 1.0646,
+      "step": 4756
+    },
+    {
+      "epoch": 0.40741692360397397,
+      "grad_norm": 0.5917473435401917,
+      "learning_rate": 0.0001,
+      "loss": 1.0715,
+      "step": 4757
+    },
+    {
+      "epoch": 0.40750256937307294,
+      "grad_norm": 0.7969913482666016,
+      "learning_rate": 0.0001,
+      "loss": 0.868,
+      "step": 4758
+    },
+    {
+      "epoch": 0.407588215142172,
+      "grad_norm": 0.825420618057251,
+      "learning_rate": 0.0001,
+      "loss": 1.0067,
+      "step": 4759
+    },
+    {
+      "epoch": 0.407673860911271,
+      "grad_norm": 0.5955457091331482,
+      "learning_rate": 0.0001,
+      "loss": 0.3359,
+      "step": 4760
+    },
+    {
+      "epoch": 0.40775950668037,
+      "grad_norm": 0.9615036845207214,
+      "learning_rate": 0.0001,
+      "loss": 0.9878,
+      "step": 4761
+    },
+    {
+      "epoch": 0.407845152449469,
+      "grad_norm": 0.8857546448707581,
+      "learning_rate": 0.0001,
+      "loss": 0.817,
+      "step": 4762
+    },
+    {
+      "epoch": 0.407930798218568,
+      "grad_norm": 0.5422261357307434,
+      "learning_rate": 0.0001,
+      "loss": 0.9962,
+      "step": 4763
+    },
+    {
+      "epoch": 0.408016443987667,
+      "grad_norm": 0.5367940664291382,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 4764
+    },
+    {
+      "epoch": 0.40810208975676604,
+      "grad_norm": 0.3836544156074524,
+      "learning_rate": 0.0001,
+      "loss": 0.5626,
+      "step": 4765
+    },
+    {
+      "epoch": 0.408187735525865,
+      "grad_norm": 0.80195552110672,
+      "learning_rate": 0.0001,
+      "loss": 0.9746,
+      "step": 4766
+    },
+    {
+      "epoch": 0.40827338129496404,
+      "grad_norm": 0.823832631111145,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 4767
+    },
+    {
+      "epoch": 0.408359027064063,
+      "grad_norm": 0.3882998824119568,
+      "learning_rate": 0.0001,
+      "loss": 0.3698,
+      "step": 4768
+    },
+    {
+      "epoch": 0.40844467283316205,
+      "grad_norm": 0.9336508512496948,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 4769
+    },
+    {
+      "epoch": 0.408530318602261,
+      "grad_norm": 0.30474674701690674,
+      "learning_rate": 0.0001,
+      "loss": 0.1798,
+      "step": 4770
+    },
+    {
+      "epoch": 0.40861596437136005,
+      "grad_norm": 0.4108959436416626,
+      "learning_rate": 0.0001,
+      "loss": 0.7269,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4087016101404591,
+      "grad_norm": 0.5815849900245667,
+      "learning_rate": 0.0001,
+      "loss": 0.6935,
+      "step": 4772
+    },
+    {
+      "epoch": 0.40878725590955806,
+      "grad_norm": 0.6974014639854431,
+      "learning_rate": 0.0001,
+      "loss": 1.1668,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4088729016786571,
+      "grad_norm": 0.8680015206336975,
+      "learning_rate": 0.0001,
+      "loss": 1.3538,
+      "step": 4774
+    },
+    {
+      "epoch": 0.40895854744775606,
+      "grad_norm": 0.6772201657295227,
+      "learning_rate": 0.0001,
+      "loss": 0.5148,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4090441932168551,
+      "grad_norm": 0.6210677027702332,
+      "learning_rate": 0.0001,
+      "loss": 1.2745,
+      "step": 4776
+    },
+    {
+      "epoch": 0.4091298389859541,
+      "grad_norm": 0.5809104442596436,
+      "learning_rate": 0.0001,
+      "loss": 0.9663,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4092154847550531,
+      "grad_norm": 0.8782781362533569,
+      "learning_rate": 0.0001,
+      "loss": 0.8452,
+      "step": 4778
+    },
+    {
+      "epoch": 0.4093011305241521,
+      "grad_norm": 0.8534605503082275,
+      "learning_rate": 0.0001,
+      "loss": 1.104,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4093867762932511,
+      "grad_norm": 1.0196315050125122,
+      "learning_rate": 0.0001,
+      "loss": 0.8467,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4094724220623501,
+      "grad_norm": 0.7992936372756958,
+      "learning_rate": 0.0001,
+      "loss": 0.9929,
+      "step": 4781
+    },
+    {
+      "epoch": 0.4095580678314491,
+      "grad_norm": 0.9101520776748657,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 4782
+    },
+    {
+      "epoch": 0.40964371360054813,
+      "grad_norm": 0.908094048500061,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 4783
+    },
+    {
+      "epoch": 0.40972935936964716,
+      "grad_norm": 0.6142090559005737,
+      "learning_rate": 0.0001,
+      "loss": 0.6673,
+      "step": 4784
+    },
+    {
+      "epoch": 0.40981500513874614,
+      "grad_norm": 0.849695086479187,
+      "learning_rate": 0.0001,
+      "loss": 0.8706,
+      "step": 4785
+    },
+    {
+      "epoch": 0.40990065090784517,
+      "grad_norm": 0.7411463856697083,
+      "learning_rate": 0.0001,
+      "loss": 0.6763,
+      "step": 4786
+    },
+    {
+      "epoch": 0.40998629667694414,
+      "grad_norm": 0.6136890649795532,
+      "learning_rate": 0.0001,
+      "loss": 0.606,
+      "step": 4787
+    },
+    {
+      "epoch": 0.41007194244604317,
+      "grad_norm": 0.9911336898803711,
+      "learning_rate": 0.0001,
+      "loss": 1.327,
+      "step": 4788
+    },
+    {
+      "epoch": 0.4101575882151422,
+      "grad_norm": 0.6774811744689941,
+      "learning_rate": 0.0001,
+      "loss": 0.9921,
+      "step": 4789
+    },
+    {
+      "epoch": 0.4102432339842412,
+      "grad_norm": 1.0288456678390503,
+      "learning_rate": 0.0001,
+      "loss": 0.9708,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4103288797533402,
+      "grad_norm": 1.102091908454895,
+      "learning_rate": 0.0001,
+      "loss": 1.9224,
+      "step": 4791
+    },
+    {
+      "epoch": 0.4104145255224392,
+      "grad_norm": 0.8956152200698853,
+      "learning_rate": 0.0001,
+      "loss": 1.3865,
+      "step": 4792
+    },
+    {
+      "epoch": 0.4105001712915382,
+      "grad_norm": 0.511534571647644,
+      "learning_rate": 0.0001,
+      "loss": 0.5067,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4105858170606372,
+      "grad_norm": 0.5150948762893677,
+      "learning_rate": 0.0001,
+      "loss": 0.7151,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4106714628297362,
+      "grad_norm": 0.4655480682849884,
+      "learning_rate": 0.0001,
+      "loss": 0.5174,
+      "step": 4795
+    },
+    {
+      "epoch": 0.41075710859883524,
+      "grad_norm": 1.0245698690414429,
+      "learning_rate": 0.0001,
+      "loss": 1.3088,
+      "step": 4796
+    },
+    {
+      "epoch": 0.4108427543679342,
+      "grad_norm": 0.6836179494857788,
+      "learning_rate": 0.0001,
+      "loss": 0.8803,
+      "step": 4797
+    },
+    {
+      "epoch": 0.41092840013703325,
+      "grad_norm": 0.7698058485984802,
+      "learning_rate": 0.0001,
+      "loss": 0.5699,
+      "step": 4798
+    },
+    {
+      "epoch": 0.4110140459061322,
+      "grad_norm": 0.5398633480072021,
+      "learning_rate": 0.0001,
+      "loss": 1.2264,
+      "step": 4799
+    },
+    {
+      "epoch": 0.41109969167523125,
+      "grad_norm": 0.4123792350292206,
+      "learning_rate": 0.0001,
+      "loss": 0.4976,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4111853374443302,
+      "grad_norm": 0.5514782667160034,
+      "learning_rate": 0.0001,
+      "loss": 1.0804,
+      "step": 4801
+    },
+    {
+      "epoch": 0.41127098321342925,
+      "grad_norm": 0.7499353289604187,
+      "learning_rate": 0.0001,
+      "loss": 0.9769,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4113566289825283,
+      "grad_norm": 0.6346805691719055,
+      "learning_rate": 0.0001,
+      "loss": 1.0147,
+      "step": 4803
+    },
+    {
+      "epoch": 0.41144227475162726,
+      "grad_norm": 1.230237364768982,
+      "learning_rate": 0.0001,
+      "loss": 0.8736,
+      "step": 4804
+    },
+    {
+      "epoch": 0.4115279205207263,
+      "grad_norm": 0.8271334767341614,
+      "learning_rate": 0.0001,
+      "loss": 2.0739,
+      "step": 4805
+    },
+    {
+      "epoch": 0.41161356628982526,
+      "grad_norm": 0.6314065456390381,
+      "learning_rate": 0.0001,
+      "loss": 1.2157,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4116992120589243,
+      "grad_norm": 0.5785664319992065,
+      "learning_rate": 0.0001,
+      "loss": 1.0515,
+      "step": 4807
+    },
+    {
+      "epoch": 0.4117848578280233,
+      "grad_norm": 0.633587121963501,
+      "learning_rate": 0.0001,
+      "loss": 1.3026,
+      "step": 4808
+    },
+    {
+      "epoch": 0.4118705035971223,
+      "grad_norm": 1.255781650543213,
+      "learning_rate": 0.0001,
+      "loss": 1.3215,
+      "step": 4809
+    },
+    {
+      "epoch": 0.4119561493662213,
+      "grad_norm": 0.9662954807281494,
+      "learning_rate": 0.0001,
+      "loss": 0.8632,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4120417951353203,
+      "grad_norm": 0.4153375029563904,
+      "learning_rate": 0.0001,
+      "loss": 0.5638,
+      "step": 4811
+    },
+    {
+      "epoch": 0.41212744090441933,
+      "grad_norm": 0.5849943161010742,
+      "learning_rate": 0.0001,
+      "loss": 1.0342,
+      "step": 4812
+    },
+    {
+      "epoch": 0.4122130866735183,
+      "grad_norm": 0.3467383086681366,
+      "learning_rate": 0.0001,
+      "loss": 0.3866,
+      "step": 4813
+    },
+    {
+      "epoch": 0.41229873244261733,
+      "grad_norm": 0.6256757378578186,
+      "learning_rate": 0.0001,
+      "loss": 0.7906,
+      "step": 4814
+    },
+    {
+      "epoch": 0.41238437821171636,
+      "grad_norm": 0.9225805997848511,
+      "learning_rate": 0.0001,
+      "loss": 0.9083,
+      "step": 4815
+    },
+    {
+      "epoch": 0.41247002398081534,
+      "grad_norm": 0.9588562846183777,
+      "learning_rate": 0.0001,
+      "loss": 0.8682,
+      "step": 4816
+    },
+    {
+      "epoch": 0.41255566974991437,
+      "grad_norm": 0.9014851450920105,
+      "learning_rate": 0.0001,
+      "loss": 0.8549,
+      "step": 4817
+    },
+    {
+      "epoch": 0.41264131551901334,
+      "grad_norm": 0.6757388114929199,
+      "learning_rate": 0.0001,
+      "loss": 0.642,
+      "step": 4818
+    },
+    {
+      "epoch": 0.41272696128811237,
+      "grad_norm": 1.2625830173492432,
+      "learning_rate": 0.0001,
+      "loss": 1.2712,
+      "step": 4819
+    },
+    {
+      "epoch": 0.4128126070572114,
+      "grad_norm": 1.1886014938354492,
+      "learning_rate": 0.0001,
+      "loss": 1.2282,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4128982528263104,
+      "grad_norm": 0.625144898891449,
+      "learning_rate": 0.0001,
+      "loss": 0.9054,
+      "step": 4821
+    },
+    {
+      "epoch": 0.4129838985954094,
+      "grad_norm": 0.6743766665458679,
+      "learning_rate": 0.0001,
+      "loss": 1.2054,
+      "step": 4822
+    },
+    {
+      "epoch": 0.4130695443645084,
+      "grad_norm": 0.6010375618934631,
+      "learning_rate": 0.0001,
+      "loss": 1.2571,
+      "step": 4823
+    },
+    {
+      "epoch": 0.4131551901336074,
+      "grad_norm": 0.9817209243774414,
+      "learning_rate": 0.0001,
+      "loss": 0.9196,
+      "step": 4824
+    },
+    {
+      "epoch": 0.4132408359027064,
+      "grad_norm": 0.9223926663398743,
+      "learning_rate": 0.0001,
+      "loss": 0.8016,
+      "step": 4825
+    },
+    {
+      "epoch": 0.4133264816718054,
+      "grad_norm": 1.0903384685516357,
+      "learning_rate": 0.0001,
+      "loss": 1.2022,
+      "step": 4826
+    },
+    {
+      "epoch": 0.41341212744090444,
+      "grad_norm": 0.5218858122825623,
+      "learning_rate": 0.0001,
+      "loss": 0.9036,
+      "step": 4827
+    },
+    {
+      "epoch": 0.4134977732100034,
+      "grad_norm": 0.8701545000076294,
+      "learning_rate": 0.0001,
+      "loss": 0.9398,
+      "step": 4828
+    },
+    {
+      "epoch": 0.41358341897910245,
+      "grad_norm": 0.8790535926818848,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 4829
+    },
+    {
+      "epoch": 0.4136690647482014,
+      "grad_norm": 0.5486084222793579,
+      "learning_rate": 0.0001,
+      "loss": 0.937,
+      "step": 4830
+    },
+    {
+      "epoch": 0.41375471051730045,
+      "grad_norm": 0.5181458592414856,
+      "learning_rate": 0.0001,
+      "loss": 0.7286,
+      "step": 4831
+    },
+    {
+      "epoch": 0.4138403562863994,
+      "grad_norm": 0.6698970794677734,
+      "learning_rate": 0.0001,
+      "loss": 0.7589,
+      "step": 4832
+    },
+    {
+      "epoch": 0.41392600205549845,
+      "grad_norm": 0.7379876971244812,
+      "learning_rate": 0.0001,
+      "loss": 0.6894,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4140116478245975,
+      "grad_norm": 0.7622514963150024,
+      "learning_rate": 0.0001,
+      "loss": 0.7245,
+      "step": 4834
+    },
+    {
+      "epoch": 0.41409729359369646,
+      "grad_norm": 0.7159548997879028,
+      "learning_rate": 0.0001,
+      "loss": 1.0879,
+      "step": 4835
+    },
+    {
+      "epoch": 0.4141829393627955,
+      "grad_norm": 0.7443615794181824,
+      "learning_rate": 0.0001,
+      "loss": 1.2806,
+      "step": 4836
+    },
+    {
+      "epoch": 0.41426858513189446,
+      "grad_norm": 0.6587781310081482,
+      "learning_rate": 0.0001,
+      "loss": 0.9128,
+      "step": 4837
+    },
+    {
+      "epoch": 0.4143542309009935,
+      "grad_norm": 1.0642433166503906,
+      "learning_rate": 0.0001,
+      "loss": 1.0229,
+      "step": 4838
+    },
+    {
+      "epoch": 0.4144398766700925,
+      "grad_norm": 0.6529123783111572,
+      "learning_rate": 0.0001,
+      "loss": 1.0011,
+      "step": 4839
+    },
+    {
+      "epoch": 0.4145255224391915,
+      "grad_norm": 1.0801596641540527,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4146111682082905,
+      "grad_norm": 0.8151669502258301,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 4841
+    },
+    {
+      "epoch": 0.4146968139773895,
+      "grad_norm": 0.5226008892059326,
+      "learning_rate": 0.0001,
+      "loss": 0.4151,
+      "step": 4842
+    },
+    {
+      "epoch": 0.41478245974648853,
+      "grad_norm": 0.534711480140686,
+      "learning_rate": 0.0001,
+      "loss": 0.5392,
+      "step": 4843
+    },
+    {
+      "epoch": 0.4148681055155875,
+      "grad_norm": 0.5981283187866211,
+      "learning_rate": 0.0001,
+      "loss": 0.9266,
+      "step": 4844
+    },
+    {
+      "epoch": 0.41495375128468653,
+      "grad_norm": 0.776588499546051,
+      "learning_rate": 0.0001,
+      "loss": 0.7578,
+      "step": 4845
+    },
+    {
+      "epoch": 0.41503939705378556,
+      "grad_norm": 0.8144716620445251,
+      "learning_rate": 0.0001,
+      "loss": 0.8729,
+      "step": 4846
+    },
+    {
+      "epoch": 0.41512504282288454,
+      "grad_norm": 0.5427016615867615,
+      "learning_rate": 0.0001,
+      "loss": 0.4228,
+      "step": 4847
+    },
+    {
+      "epoch": 0.41521068859198357,
+      "grad_norm": 0.5503108501434326,
+      "learning_rate": 0.0001,
+      "loss": 0.975,
+      "step": 4848
+    },
+    {
+      "epoch": 0.41529633436108254,
+      "grad_norm": 0.6140769720077515,
+      "learning_rate": 0.0001,
+      "loss": 1.0278,
+      "step": 4849
+    },
+    {
+      "epoch": 0.41538198013018157,
+      "grad_norm": 0.9646977186203003,
+      "learning_rate": 0.0001,
+      "loss": 0.755,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4154676258992806,
+      "grad_norm": 1.728556752204895,
+      "learning_rate": 0.0001,
+      "loss": 1.8799,
+      "step": 4851
+    },
+    {
+      "epoch": 0.4155532716683796,
+      "grad_norm": 0.5780028104782104,
+      "learning_rate": 0.0001,
+      "loss": 0.2917,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4156389174374786,
+      "grad_norm": 0.7768418192863464,
+      "learning_rate": 0.0001,
+      "loss": 0.7955,
+      "step": 4853
+    },
+    {
+      "epoch": 0.4157245632065776,
+      "grad_norm": 0.9126242995262146,
+      "learning_rate": 0.0001,
+      "loss": 0.7782,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4158102089756766,
+      "grad_norm": 1.182665228843689,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 4855
+    },
+    {
+      "epoch": 0.4158958547447756,
+      "grad_norm": 0.6804608106613159,
+      "learning_rate": 0.0001,
+      "loss": 0.8419,
+      "step": 4856
+    },
+    {
+      "epoch": 0.4159815005138746,
+      "grad_norm": 0.4479224681854248,
+      "learning_rate": 0.0001,
+      "loss": 0.6853,
+      "step": 4857
+    },
+    {
+      "epoch": 0.41606714628297364,
+      "grad_norm": 0.37083977460861206,
+      "learning_rate": 0.0001,
+      "loss": 0.6936,
+      "step": 4858
+    },
+    {
+      "epoch": 0.4161527920520726,
+      "grad_norm": 0.6267076730728149,
+      "learning_rate": 0.0001,
+      "loss": 0.8641,
+      "step": 4859
+    },
+    {
+      "epoch": 0.41623843782117165,
+      "grad_norm": 0.6890312433242798,
+      "learning_rate": 0.0001,
+      "loss": 1.0085,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4163240835902706,
+      "grad_norm": 0.39787915349006653,
+      "learning_rate": 0.0001,
+      "loss": 0.5135,
+      "step": 4861
+    },
+    {
+      "epoch": 0.41640972935936965,
+      "grad_norm": 1.097380518913269,
+      "learning_rate": 0.0001,
+      "loss": 1.2615,
+      "step": 4862
+    },
+    {
+      "epoch": 0.4164953751284687,
+      "grad_norm": 0.6059827208518982,
+      "learning_rate": 0.0001,
+      "loss": 0.9943,
+      "step": 4863
+    },
+    {
+      "epoch": 0.41658102089756766,
+      "grad_norm": 0.8206390142440796,
+      "learning_rate": 0.0001,
+      "loss": 1.1336,
+      "step": 4864
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 1.340116262435913,
+      "learning_rate": 0.0001,
+      "loss": 1.2129,
+      "step": 4865
+    },
+    {
+      "epoch": 0.41675231243576566,
+      "grad_norm": 0.4258352816104889,
+      "learning_rate": 0.0001,
+      "loss": 0.4306,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4168379582048647,
+      "grad_norm": 0.8214362263679504,
+      "learning_rate": 0.0001,
+      "loss": 0.4739,
+      "step": 4867
+    },
+    {
+      "epoch": 0.41692360397396366,
+      "grad_norm": 0.7263301610946655,
+      "learning_rate": 0.0001,
+      "loss": 0.7504,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4170092497430627,
+      "grad_norm": 0.38173362612724304,
+      "learning_rate": 0.0001,
+      "loss": 0.2868,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4170948955121617,
+      "grad_norm": 0.6128078103065491,
+      "learning_rate": 0.0001,
+      "loss": 1.2854,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4171805412812607,
+      "grad_norm": 0.8821213841438293,
+      "learning_rate": 0.0001,
+      "loss": 0.9361,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4172661870503597,
+      "grad_norm": 0.5300100445747375,
+      "learning_rate": 0.0001,
+      "loss": 0.4109,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4173518328194587,
+      "grad_norm": 0.771249532699585,
+      "learning_rate": 0.0001,
+      "loss": 1.1679,
+      "step": 4873
+    },
+    {
+      "epoch": 0.41743747858855773,
+      "grad_norm": 0.9766458868980408,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4175231243576567,
+      "grad_norm": 0.6116730570793152,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 4875
+    },
+    {
+      "epoch": 0.41760877012675574,
+      "grad_norm": 0.6073743104934692,
+      "learning_rate": 0.0001,
+      "loss": 1.0055,
+      "step": 4876
+    },
+    {
+      "epoch": 0.41769441589585476,
+      "grad_norm": 1.0617859363555908,
+      "learning_rate": 0.0001,
+      "loss": 1.1146,
+      "step": 4877
+    },
+    {
+      "epoch": 0.41778006166495374,
+      "grad_norm": 0.9133240580558777,
+      "learning_rate": 0.0001,
+      "loss": 1.2225,
+      "step": 4878
+    },
+    {
+      "epoch": 0.41786570743405277,
+      "grad_norm": 0.5818451046943665,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 4879
+    },
+    {
+      "epoch": 0.41795135320315174,
+      "grad_norm": 0.7972307801246643,
+      "learning_rate": 0.0001,
+      "loss": 1.148,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4180369989722508,
+      "grad_norm": 0.20233863592147827,
+      "learning_rate": 0.0001,
+      "loss": 0.4527,
+      "step": 4881
+    },
+    {
+      "epoch": 0.4181226447413498,
+      "grad_norm": 0.8355880975723267,
+      "learning_rate": 0.0001,
+      "loss": 1.009,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4182082905104488,
+      "grad_norm": 0.7149040102958679,
+      "learning_rate": 0.0001,
+      "loss": 0.7829,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4182939362795478,
+      "grad_norm": 0.9224740862846375,
+      "learning_rate": 0.0001,
+      "loss": 1.1153,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4183795820486468,
+      "grad_norm": 0.959021806716919,
+      "learning_rate": 0.0001,
+      "loss": 1.1774,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4184652278177458,
+      "grad_norm": 0.6371318697929382,
+      "learning_rate": 0.0001,
+      "loss": 0.9819,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4185508735868448,
+      "grad_norm": 0.5775449872016907,
+      "learning_rate": 0.0001,
+      "loss": 0.9934,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4186365193559438,
+      "grad_norm": 0.5953220129013062,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 4888
+    },
+    {
+      "epoch": 0.41872216512504284,
+      "grad_norm": 1.082646369934082,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 4889
+    },
+    {
+      "epoch": 0.4188078108941418,
+      "grad_norm": 0.9546902179718018,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 4890
+    },
+    {
+      "epoch": 0.41889345666324085,
+      "grad_norm": 0.8489063382148743,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4189791024323398,
+      "grad_norm": 0.7451435327529907,
+      "learning_rate": 0.0001,
+      "loss": 0.7676,
+      "step": 4892
+    },
+    {
+      "epoch": 0.41906474820143885,
+      "grad_norm": 0.49771782755851746,
+      "learning_rate": 0.0001,
+      "loss": 0.5408,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4191503939705379,
+      "grad_norm": 0.743211030960083,
+      "learning_rate": 0.0001,
+      "loss": 1.3669,
+      "step": 4894
+    },
+    {
+      "epoch": 0.41923603973963686,
+      "grad_norm": 0.719571053981781,
+      "learning_rate": 0.0001,
+      "loss": 0.8317,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4193216855087359,
+      "grad_norm": 0.4479481279850006,
+      "learning_rate": 0.0001,
+      "loss": 0.675,
+      "step": 4896
+    },
+    {
+      "epoch": 0.41940733127783486,
+      "grad_norm": 0.4411208927631378,
+      "learning_rate": 0.0001,
+      "loss": 0.5622,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4194929770469339,
+      "grad_norm": 0.3956540524959564,
+      "learning_rate": 0.0001,
+      "loss": 0.8365,
+      "step": 4898
+    },
+    {
+      "epoch": 0.41957862281603286,
+      "grad_norm": 0.6580791473388672,
+      "learning_rate": 0.0001,
+      "loss": 1.1216,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4196642685851319,
+      "grad_norm": 0.4637317359447479,
+      "learning_rate": 0.0001,
+      "loss": 0.6085,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4197499143542309,
+      "grad_norm": 0.9323273301124573,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4198355601233299,
+      "grad_norm": 0.9609602093696594,
+      "learning_rate": 0.0001,
+      "loss": 0.9817,
+      "step": 4902
+    },
+    {
+      "epoch": 0.41992120589242893,
+      "grad_norm": 0.7337010502815247,
+      "learning_rate": 0.0001,
+      "loss": 0.9661,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4200068516615279,
+      "grad_norm": 0.9506608843803406,
+      "learning_rate": 0.0001,
+      "loss": 1.2534,
+      "step": 4904
+    },
+    {
+      "epoch": 0.42009249743062693,
+      "grad_norm": 0.5800967812538147,
+      "learning_rate": 0.0001,
+      "loss": 0.8857,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4201781431997259,
+      "grad_norm": 0.8489867448806763,
+      "learning_rate": 0.0001,
+      "loss": 0.9397,
+      "step": 4906
+    },
+    {
+      "epoch": 0.42026378896882494,
+      "grad_norm": 1.137114405632019,
+      "learning_rate": 0.0001,
+      "loss": 1.0767,
+      "step": 4907
+    },
+    {
+      "epoch": 0.42034943473792397,
+      "grad_norm": 0.3939639925956726,
+      "learning_rate": 0.0001,
+      "loss": 0.2713,
+      "step": 4908
+    },
+    {
+      "epoch": 0.42043508050702294,
+      "grad_norm": 1.2718749046325684,
+      "learning_rate": 0.0001,
+      "loss": 1.1809,
+      "step": 4909
+    },
+    {
+      "epoch": 0.42052072627612197,
+      "grad_norm": 1.094857931137085,
+      "learning_rate": 0.0001,
+      "loss": 0.7154,
+      "step": 4910
+    },
+    {
+      "epoch": 0.42060637204522094,
+      "grad_norm": 0.668690025806427,
+      "learning_rate": 0.0001,
+      "loss": 1.3473,
+      "step": 4911
+    },
+    {
+      "epoch": 0.42069201781432,
+      "grad_norm": 0.7815335988998413,
+      "learning_rate": 0.0001,
+      "loss": 0.9575,
+      "step": 4912
+    },
+    {
+      "epoch": 0.420777663583419,
+      "grad_norm": 0.4631480574607849,
+      "learning_rate": 0.0001,
+      "loss": 0.2829,
+      "step": 4913
+    },
+    {
+      "epoch": 0.420863309352518,
+      "grad_norm": 0.773600697517395,
+      "learning_rate": 0.0001,
+      "loss": 1.0654,
+      "step": 4914
+    },
+    {
+      "epoch": 0.420948955121617,
+      "grad_norm": 0.8620867133140564,
+      "learning_rate": 0.0001,
+      "loss": 1.1526,
+      "step": 4915
+    },
+    {
+      "epoch": 0.421034600890716,
+      "grad_norm": 1.1834889650344849,
+      "learning_rate": 0.0001,
+      "loss": 0.9813,
+      "step": 4916
+    },
+    {
+      "epoch": 0.421120246659815,
+      "grad_norm": 0.8075287938117981,
+      "learning_rate": 0.0001,
+      "loss": 0.3295,
+      "step": 4917
+    },
+    {
+      "epoch": 0.421205892428914,
+      "grad_norm": 0.8361770510673523,
+      "learning_rate": 0.0001,
+      "loss": 0.6905,
+      "step": 4918
+    },
+    {
+      "epoch": 0.421291538198013,
+      "grad_norm": 0.7326636910438538,
+      "learning_rate": 0.0001,
+      "loss": 1.1679,
+      "step": 4919
+    },
+    {
+      "epoch": 0.42137718396711205,
+      "grad_norm": 1.0205061435699463,
+      "learning_rate": 0.0001,
+      "loss": 1.0525,
+      "step": 4920
+    },
+    {
+      "epoch": 0.421462829736211,
+      "grad_norm": 0.8619720339775085,
+      "learning_rate": 0.0001,
+      "loss": 1.0322,
+      "step": 4921
+    },
+    {
+      "epoch": 0.42154847550531005,
+      "grad_norm": 0.5263166427612305,
+      "learning_rate": 0.0001,
+      "loss": 0.6673,
+      "step": 4922
+    },
+    {
+      "epoch": 0.421634121274409,
+      "grad_norm": 0.6021981835365295,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 4923
+    },
+    {
+      "epoch": 0.42171976704350805,
+      "grad_norm": 0.8122313618659973,
+      "learning_rate": 0.0001,
+      "loss": 0.7122,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4218054128126071,
+      "grad_norm": 0.41988658905029297,
+      "learning_rate": 0.0001,
+      "loss": 0.9377,
+      "step": 4925
+    },
+    {
+      "epoch": 0.42189105858170606,
+      "grad_norm": 0.7852214574813843,
+      "learning_rate": 0.0001,
+      "loss": 1.206,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4219767043508051,
+      "grad_norm": 0.5542019605636597,
+      "learning_rate": 0.0001,
+      "loss": 0.8887,
+      "step": 4927
+    },
+    {
+      "epoch": 0.42206235011990406,
+      "grad_norm": 0.3556737005710602,
+      "learning_rate": 0.0001,
+      "loss": 0.7175,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4221479958890031,
+      "grad_norm": 0.5931001305580139,
+      "learning_rate": 0.0001,
+      "loss": 1.3628,
+      "step": 4929
+    },
+    {
+      "epoch": 0.42223364165810207,
+      "grad_norm": 0.4447270333766937,
+      "learning_rate": 0.0001,
+      "loss": 0.8581,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4223192874272011,
+      "grad_norm": 0.727512001991272,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4224049331963001,
+      "grad_norm": 1.010209083557129,
+      "learning_rate": 0.0001,
+      "loss": 0.5934,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4224905789653991,
+      "grad_norm": 1.1458560228347778,
+      "learning_rate": 0.0001,
+      "loss": 1.2598,
+      "step": 4933
+    },
+    {
+      "epoch": 0.42257622473449813,
+      "grad_norm": 0.596380889415741,
+      "learning_rate": 0.0001,
+      "loss": 0.663,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4226618705035971,
+      "grad_norm": 1.569285273551941,
+      "learning_rate": 0.0001,
+      "loss": 1.8692,
+      "step": 4935
+    },
+    {
+      "epoch": 0.42274751627269613,
+      "grad_norm": 0.9456716179847717,
+      "learning_rate": 0.0001,
+      "loss": 1.0478,
+      "step": 4936
+    },
+    {
+      "epoch": 0.42283316204179516,
+      "grad_norm": 0.5477632284164429,
+      "learning_rate": 0.0001,
+      "loss": 0.7428,
+      "step": 4937
+    },
+    {
+      "epoch": 0.42291880781089414,
+      "grad_norm": 0.7100539803504944,
+      "learning_rate": 0.0001,
+      "loss": 0.4083,
+      "step": 4938
+    },
+    {
+      "epoch": 0.42300445357999317,
+      "grad_norm": 0.5411487221717834,
+      "learning_rate": 0.0001,
+      "loss": 0.9155,
+      "step": 4939
+    },
+    {
+      "epoch": 0.42309009934909214,
+      "grad_norm": 0.3996324837207794,
+      "learning_rate": 0.0001,
+      "loss": 0.7752,
+      "step": 4940
+    },
+    {
+      "epoch": 0.42317574511819117,
+      "grad_norm": 0.6735579967498779,
+      "learning_rate": 0.0001,
+      "loss": 0.9127,
+      "step": 4941
+    },
+    {
+      "epoch": 0.42326139088729015,
+      "grad_norm": 0.8267484903335571,
+      "learning_rate": 0.0001,
+      "loss": 1.2993,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4233470366563892,
+      "grad_norm": 0.9370483756065369,
+      "learning_rate": 0.0001,
+      "loss": 0.8144,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4234326824254882,
+      "grad_norm": 0.9961097836494446,
+      "learning_rate": 0.0001,
+      "loss": 0.696,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4235183281945872,
+      "grad_norm": 0.7715698480606079,
+      "learning_rate": 0.0001,
+      "loss": 0.8411,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4236039739636862,
+      "grad_norm": 1.1814497709274292,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4236896197327852,
+      "grad_norm": 0.629518985748291,
+      "learning_rate": 0.0001,
+      "loss": 0.9582,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4237752655018842,
+      "grad_norm": 0.8328909873962402,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4238609112709832,
+      "grad_norm": 0.746147096157074,
+      "learning_rate": 0.0001,
+      "loss": 0.7755,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4239465570400822,
+      "grad_norm": 0.33126309514045715,
+      "learning_rate": 0.0001,
+      "loss": 0.9965,
+      "step": 4950
+    },
+    {
+      "epoch": 0.42403220280918125,
+      "grad_norm": 1.0228137969970703,
+      "learning_rate": 0.0001,
+      "loss": 1.1312,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4241178485782802,
+      "grad_norm": 1.0051860809326172,
+      "learning_rate": 0.0001,
+      "loss": 1.202,
+      "step": 4952
+    },
+    {
+      "epoch": 0.42420349434737925,
+      "grad_norm": 1.1207969188690186,
+      "learning_rate": 0.0001,
+      "loss": 1.0283,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4242891401164782,
+      "grad_norm": 0.5988748669624329,
+      "learning_rate": 0.0001,
+      "loss": 1.1097,
+      "step": 4954
+    },
+    {
+      "epoch": 0.42437478588557725,
+      "grad_norm": 0.8390317559242249,
+      "learning_rate": 0.0001,
+      "loss": 0.8089,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4244604316546763,
+      "grad_norm": 0.7568089365959167,
+      "learning_rate": 0.0001,
+      "loss": 0.9509,
+      "step": 4956
+    },
+    {
+      "epoch": 0.42454607742377526,
+      "grad_norm": 1.6415657997131348,
+      "learning_rate": 0.0001,
+      "loss": 1.0271,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4246317231928743,
+      "grad_norm": 0.9049854874610901,
+      "learning_rate": 0.0001,
+      "loss": 0.7515,
+      "step": 4958
+    },
+    {
+      "epoch": 0.42471736896197326,
+      "grad_norm": 0.6392579078674316,
+      "learning_rate": 0.0001,
+      "loss": 0.5024,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4248030147310723,
+      "grad_norm": 0.9667561650276184,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 4960
+    },
+    {
+      "epoch": 0.42488866050017127,
+      "grad_norm": 1.3640738725662231,
+      "learning_rate": 0.0001,
+      "loss": 0.9783,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4249743062692703,
+      "grad_norm": 0.5308374166488647,
+      "learning_rate": 0.0001,
+      "loss": 0.6954,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4250599520383693,
+      "grad_norm": 0.6368821263313293,
+      "learning_rate": 0.0001,
+      "loss": 1.1046,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4251455978074683,
+      "grad_norm": 0.6549354791641235,
+      "learning_rate": 0.0001,
+      "loss": 0.7476,
+      "step": 4964
+    },
+    {
+      "epoch": 0.42523124357656733,
+      "grad_norm": 0.6300658583641052,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4253168893456663,
+      "grad_norm": 0.6212546229362488,
+      "learning_rate": 0.0001,
+      "loss": 1.251,
+      "step": 4966
+    },
+    {
+      "epoch": 0.42540253511476533,
+      "grad_norm": 0.8162621259689331,
+      "learning_rate": 0.0001,
+      "loss": 0.8917,
+      "step": 4967
+    },
+    {
+      "epoch": 0.42548818088386436,
+      "grad_norm": 0.36425986886024475,
+      "learning_rate": 0.0001,
+      "loss": 0.6544,
+      "step": 4968
+    },
+    {
+      "epoch": 0.42557382665296334,
+      "grad_norm": 0.6810424327850342,
+      "learning_rate": 0.0001,
+      "loss": 0.9631,
+      "step": 4969
+    },
+    {
+      "epoch": 0.42565947242206237,
+      "grad_norm": 0.6944541335105896,
+      "learning_rate": 0.0001,
+      "loss": 0.8046,
+      "step": 4970
+    },
+    {
+      "epoch": 0.42574511819116134,
+      "grad_norm": 0.6129245162010193,
+      "learning_rate": 0.0001,
+      "loss": 0.8199,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4258307639602604,
+      "grad_norm": 0.9350326657295227,
+      "learning_rate": 0.0001,
+      "loss": 0.8855,
+      "step": 4972
+    },
+    {
+      "epoch": 0.42591640972935935,
+      "grad_norm": 0.807665228843689,
+      "learning_rate": 0.0001,
+      "loss": 0.7822,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4260020554984584,
+      "grad_norm": 0.657364010810852,
+      "learning_rate": 0.0001,
+      "loss": 1.2184,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4260877012675574,
+      "grad_norm": 0.5882235765457153,
+      "learning_rate": 0.0001,
+      "loss": 0.6303,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4261733470366564,
+      "grad_norm": 0.7688015103340149,
+      "learning_rate": 0.0001,
+      "loss": 1.2158,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4262589928057554,
+      "grad_norm": 0.844916820526123,
+      "learning_rate": 0.0001,
+      "loss": 0.9881,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4263446385748544,
+      "grad_norm": 0.3135867416858673,
+      "learning_rate": 0.0001,
+      "loss": 0.5101,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4264302843439534,
+      "grad_norm": 0.9592832922935486,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4265159301130524,
+      "grad_norm": 0.4565441906452179,
+      "learning_rate": 0.0001,
+      "loss": 0.4567,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4266015758821514,
+      "grad_norm": 1.2633647918701172,
+      "learning_rate": 0.0001,
+      "loss": 1.3125,
+      "step": 4981
+    },
+    {
+      "epoch": 0.42668722165125045,
+      "grad_norm": 0.7220638990402222,
+      "learning_rate": 0.0001,
+      "loss": 0.8467,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4267728674203494,
+      "grad_norm": 0.7451822757720947,
+      "learning_rate": 0.0001,
+      "loss": 0.7169,
+      "step": 4983
+    },
+    {
+      "epoch": 0.42685851318944845,
+      "grad_norm": 0.6742476224899292,
+      "learning_rate": 0.0001,
+      "loss": 0.8993,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4269441589585474,
+      "grad_norm": 0.8833948373794556,
+      "learning_rate": 0.0001,
+      "loss": 0.8943,
+      "step": 4985
+    },
+    {
+      "epoch": 0.42702980472764646,
+      "grad_norm": 0.7418994307518005,
+      "learning_rate": 0.0001,
+      "loss": 0.7224,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4271154504967455,
+      "grad_norm": 0.48298391699790955,
+      "learning_rate": 0.0001,
+      "loss": 0.6624,
+      "step": 4987
+    },
+    {
+      "epoch": 0.42720109626584446,
+      "grad_norm": 0.47877535223960876,
+      "learning_rate": 0.0001,
+      "loss": 0.6522,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4272867420349435,
+      "grad_norm": 1.0981204509735107,
+      "learning_rate": 0.0001,
+      "loss": 1.3739,
+      "step": 4989
+    },
+    {
+      "epoch": 0.42737238780404246,
+      "grad_norm": 0.9482793807983398,
+      "learning_rate": 0.0001,
+      "loss": 1.2267,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4274580335731415,
+      "grad_norm": 1.0246636867523193,
+      "learning_rate": 0.0001,
+      "loss": 0.5371,
+      "step": 4991
+    },
+    {
+      "epoch": 0.42754367934224047,
+      "grad_norm": 0.8783990144729614,
+      "learning_rate": 0.0001,
+      "loss": 1.0653,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4276293251113395,
+      "grad_norm": 0.443251371383667,
+      "learning_rate": 0.0001,
+      "loss": 0.6919,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4277149708804385,
+      "grad_norm": 0.6505297422409058,
+      "learning_rate": 0.0001,
+      "loss": 1.0855,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4278006166495375,
+      "grad_norm": 0.5624368190765381,
+      "learning_rate": 0.0001,
+      "loss": 1.1081,
+      "step": 4995
+    },
+    {
+      "epoch": 0.42788626241863653,
+      "grad_norm": 1.035270094871521,
+      "learning_rate": 0.0001,
+      "loss": 0.9497,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4279719081877355,
+      "grad_norm": 0.8548805117607117,
+      "learning_rate": 0.0001,
+      "loss": 1.1226,
+      "step": 4997
+    },
+    {
+      "epoch": 0.42805755395683454,
+      "grad_norm": 0.8374253511428833,
+      "learning_rate": 0.0001,
+      "loss": 1.0171,
+      "step": 4998
+    },
+    {
+      "epoch": 0.42814319972593357,
+      "grad_norm": 1.0585509538650513,
+      "learning_rate": 0.0001,
+      "loss": 1.0523,
+      "step": 4999
+    },
+    {
+      "epoch": 0.42822884549503254,
+      "grad_norm": 1.1307402849197388,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 5000
+    },
+    {
+      "epoch": 0.42831449126413157,
+      "grad_norm": 0.5529690980911255,
+      "learning_rate": 0.0001,
+      "loss": 0.9256,
+      "step": 5001
+    },
+    {
+      "epoch": 0.42840013703323054,
+      "grad_norm": 0.5463783740997314,
+      "learning_rate": 0.0001,
+      "loss": 0.4066,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4284857828023296,
+      "grad_norm": 1.1216672658920288,
+      "learning_rate": 0.0001,
+      "loss": 0.865,
+      "step": 5003
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.5025882124900818,
+      "learning_rate": 0.0001,
+      "loss": 0.8002,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4286570743405276,
+      "grad_norm": 0.9099591374397278,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4287427201096266,
+      "grad_norm": 0.9607415795326233,
+      "learning_rate": 0.0001,
+      "loss": 1.0682,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4288283658787256,
+      "grad_norm": 0.870597243309021,
+      "learning_rate": 0.0001,
+      "loss": 1.0992,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4289140116478246,
+      "grad_norm": 0.7233594655990601,
+      "learning_rate": 0.0001,
+      "loss": 0.677,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4289996574169236,
+      "grad_norm": 0.6168982982635498,
+      "learning_rate": 0.0001,
+      "loss": 0.5457,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4290853031860226,
+      "grad_norm": 0.8613752722740173,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4291709489551216,
+      "grad_norm": 0.6907917857170105,
+      "learning_rate": 0.0001,
+      "loss": 0.9499,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4292565947242206,
+      "grad_norm": 0.6315884590148926,
+      "learning_rate": 0.0001,
+      "loss": 1.2915,
+      "step": 5012
+    },
+    {
+      "epoch": 0.42934224049331965,
+      "grad_norm": 0.6695212721824646,
+      "learning_rate": 0.0001,
+      "loss": 0.8685,
+      "step": 5013
+    },
+    {
+      "epoch": 0.4294278862624186,
+      "grad_norm": 0.8135017156600952,
+      "learning_rate": 0.0001,
+      "loss": 0.5009,
+      "step": 5014
+    },
+    {
+      "epoch": 0.42951353203151765,
+      "grad_norm": 0.26842427253723145,
+      "learning_rate": 0.0001,
+      "loss": 0.3942,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4295991778006166,
+      "grad_norm": 0.6817671060562134,
+      "learning_rate": 0.0001,
+      "loss": 0.8396,
+      "step": 5016
+    },
+    {
+      "epoch": 0.42968482356971566,
+      "grad_norm": 0.46507540345191956,
+      "learning_rate": 0.0001,
+      "loss": 0.8311,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4297704693388147,
+      "grad_norm": 0.9281595945358276,
+      "learning_rate": 0.0001,
+      "loss": 1.1694,
+      "step": 5018
+    },
+    {
+      "epoch": 0.42985611510791366,
+      "grad_norm": 1.2368531227111816,
+      "learning_rate": 0.0001,
+      "loss": 0.8573,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4299417608770127,
+      "grad_norm": 0.8121311664581299,
+      "learning_rate": 0.0001,
+      "loss": 0.9064,
+      "step": 5020
+    },
+    {
+      "epoch": 0.43002740664611167,
+      "grad_norm": 0.9345277547836304,
+      "learning_rate": 0.0001,
+      "loss": 0.7981,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4301130524152107,
+      "grad_norm": 0.4713897109031677,
+      "learning_rate": 0.0001,
+      "loss": 0.8053,
+      "step": 5022
+    },
+    {
+      "epoch": 0.43019869818430967,
+      "grad_norm": 0.4234749376773834,
+      "learning_rate": 0.0001,
+      "loss": 0.8148,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4302843439534087,
+      "grad_norm": 0.35560014843940735,
+      "learning_rate": 0.0001,
+      "loss": 0.3527,
+      "step": 5024
+    },
+    {
+      "epoch": 0.43036998972250773,
+      "grad_norm": 0.708313524723053,
+      "learning_rate": 0.0001,
+      "loss": 0.6785,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4304556354916067,
+      "grad_norm": 0.5515590906143188,
+      "learning_rate": 0.0001,
+      "loss": 0.9572,
+      "step": 5026
+    },
+    {
+      "epoch": 0.43054128126070573,
+      "grad_norm": 0.7815543413162231,
+      "learning_rate": 0.0001,
+      "loss": 0.6799,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4306269270298047,
+      "grad_norm": 0.5492229461669922,
+      "learning_rate": 0.0001,
+      "loss": 1.0496,
+      "step": 5028
+    },
+    {
+      "epoch": 0.43071257279890374,
+      "grad_norm": 0.5032923221588135,
+      "learning_rate": 0.0001,
+      "loss": 0.9657,
+      "step": 5029
+    },
+    {
+      "epoch": 0.43079821856800277,
+      "grad_norm": 0.5832013487815857,
+      "learning_rate": 0.0001,
+      "loss": 0.8006,
+      "step": 5030
+    },
+    {
+      "epoch": 0.43088386433710174,
+      "grad_norm": 0.8489106893539429,
+      "learning_rate": 0.0001,
+      "loss": 0.9833,
+      "step": 5031
+    },
+    {
+      "epoch": 0.43096951010620077,
+      "grad_norm": 0.6536113023757935,
+      "learning_rate": 0.0001,
+      "loss": 1.1871,
+      "step": 5032
+    },
+    {
+      "epoch": 0.43105515587529974,
+      "grad_norm": 0.8006150126457214,
+      "learning_rate": 0.0001,
+      "loss": 0.9102,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4311408016443988,
+      "grad_norm": 0.5807569026947021,
+      "learning_rate": 0.0001,
+      "loss": 0.6938,
+      "step": 5034
+    },
+    {
+      "epoch": 0.43122644741349775,
+      "grad_norm": 1.0335670709609985,
+      "learning_rate": 0.0001,
+      "loss": 0.9353,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4313120931825968,
+      "grad_norm": 0.5630851984024048,
+      "learning_rate": 0.0001,
+      "loss": 0.3026,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4313977389516958,
+      "grad_norm": 0.9316971302032471,
+      "learning_rate": 0.0001,
+      "loss": 1.2468,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4314833847207948,
+      "grad_norm": 0.9177730083465576,
+      "learning_rate": 0.0001,
+      "loss": 0.9501,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4315690304898938,
+      "grad_norm": 0.9283683896064758,
+      "learning_rate": 0.0001,
+      "loss": 0.9506,
+      "step": 5039
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 0.6443682909011841,
+      "learning_rate": 0.0001,
+      "loss": 1.0773,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4317403220280918,
+      "grad_norm": 0.7128401398658752,
+      "learning_rate": 0.0001,
+      "loss": 0.9491,
+      "step": 5041
+    },
+    {
+      "epoch": 0.43182596779719085,
+      "grad_norm": 0.8814131021499634,
+      "learning_rate": 0.0001,
+      "loss": 0.9637,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4319116135662898,
+      "grad_norm": 0.5539996027946472,
+      "learning_rate": 0.0001,
+      "loss": 0.8333,
+      "step": 5043
+    },
+    {
+      "epoch": 0.43199725933538885,
+      "grad_norm": 1.376700520515442,
+      "learning_rate": 0.0001,
+      "loss": 1.1452,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4320829051044878,
+      "grad_norm": 1.3639726638793945,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 5045
+    },
+    {
+      "epoch": 0.43216855087358685,
+      "grad_norm": 0.9052367210388184,
+      "learning_rate": 0.0001,
+      "loss": 0.6351,
+      "step": 5046
+    },
+    {
+      "epoch": 0.43225419664268583,
+      "grad_norm": 0.8403195738792419,
+      "learning_rate": 0.0001,
+      "loss": 0.85,
+      "step": 5047
+    },
+    {
+      "epoch": 0.43233984241178486,
+      "grad_norm": 1.094456672668457,
+      "learning_rate": 0.0001,
+      "loss": 0.8279,
+      "step": 5048
+    },
+    {
+      "epoch": 0.4324254881808839,
+      "grad_norm": 0.6767417788505554,
+      "learning_rate": 0.0001,
+      "loss": 0.7062,
+      "step": 5049
+    },
+    {
+      "epoch": 0.43251113394998286,
+      "grad_norm": 0.7313541769981384,
+      "learning_rate": 0.0001,
+      "loss": 1.0354,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4325967797190819,
+      "grad_norm": 0.9313742518424988,
+      "learning_rate": 0.0001,
+      "loss": 0.4428,
+      "step": 5051
+    },
+    {
+      "epoch": 0.43268242548818087,
+      "grad_norm": 1.0090910196304321,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4327680712572799,
+      "grad_norm": 0.8533984422683716,
+      "learning_rate": 0.0001,
+      "loss": 0.8396,
+      "step": 5053
+    },
+    {
+      "epoch": 0.43285371702637887,
+      "grad_norm": 0.5378702282905579,
+      "learning_rate": 0.0001,
+      "loss": 0.9902,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4329393627954779,
+      "grad_norm": 1.1595308780670166,
+      "learning_rate": 0.0001,
+      "loss": 1.0446,
+      "step": 5055
+    },
+    {
+      "epoch": 0.43302500856457693,
+      "grad_norm": 0.6055369973182678,
+      "learning_rate": 0.0001,
+      "loss": 1.1474,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4331106543336759,
+      "grad_norm": 0.7738845348358154,
+      "learning_rate": 0.0001,
+      "loss": 1.0401,
+      "step": 5057
+    },
+    {
+      "epoch": 0.43319630010277493,
+      "grad_norm": 0.5216959714889526,
+      "learning_rate": 0.0001,
+      "loss": 0.969,
+      "step": 5058
+    },
+    {
+      "epoch": 0.4332819458718739,
+      "grad_norm": 0.9084863662719727,
+      "learning_rate": 0.0001,
+      "loss": 1.0696,
+      "step": 5059
+    },
+    {
+      "epoch": 0.43336759164097294,
+      "grad_norm": 0.5597068667411804,
+      "learning_rate": 0.0001,
+      "loss": 0.8772,
+      "step": 5060
+    },
+    {
+      "epoch": 0.43345323741007197,
+      "grad_norm": 0.7475374937057495,
+      "learning_rate": 0.0001,
+      "loss": 0.8833,
+      "step": 5061
+    },
+    {
+      "epoch": 0.43353888317917094,
+      "grad_norm": 0.48957380652427673,
+      "learning_rate": 0.0001,
+      "loss": 0.699,
+      "step": 5062
+    },
+    {
+      "epoch": 0.43362452894826997,
+      "grad_norm": 0.8545600771903992,
+      "learning_rate": 0.0001,
+      "loss": 1.0147,
+      "step": 5063
+    },
+    {
+      "epoch": 0.43371017471736895,
+      "grad_norm": 1.024044394493103,
+      "learning_rate": 0.0001,
+      "loss": 0.7235,
+      "step": 5064
+    },
+    {
+      "epoch": 0.433795820486468,
+      "grad_norm": 0.9296140670776367,
+      "learning_rate": 0.0001,
+      "loss": 0.8209,
+      "step": 5065
+    },
+    {
+      "epoch": 0.43388146625556695,
+      "grad_norm": 0.97967529296875,
+      "learning_rate": 0.0001,
+      "loss": 0.8114,
+      "step": 5066
+    },
+    {
+      "epoch": 0.433967112024666,
+      "grad_norm": 0.7206858396530151,
+      "learning_rate": 0.0001,
+      "loss": 0.8685,
+      "step": 5067
+    },
+    {
+      "epoch": 0.434052757793765,
+      "grad_norm": 0.3246588706970215,
+      "learning_rate": 0.0001,
+      "loss": 0.4914,
+      "step": 5068
+    },
+    {
+      "epoch": 0.434138403562864,
+      "grad_norm": 0.6005561947822571,
+      "learning_rate": 0.0001,
+      "loss": 0.8574,
+      "step": 5069
+    },
+    {
+      "epoch": 0.434224049331963,
+      "grad_norm": 0.5830866098403931,
+      "learning_rate": 0.0001,
+      "loss": 1.0132,
+      "step": 5070
+    },
+    {
+      "epoch": 0.434309695101062,
+      "grad_norm": 0.5594747066497803,
+      "learning_rate": 0.0001,
+      "loss": 1.0142,
+      "step": 5071
+    },
+    {
+      "epoch": 0.434395340870161,
+      "grad_norm": 0.7726355791091919,
+      "learning_rate": 0.0001,
+      "loss": 0.7942,
+      "step": 5072
+    },
+    {
+      "epoch": 0.43448098663926005,
+      "grad_norm": 0.5480542778968811,
+      "learning_rate": 0.0001,
+      "loss": 1.0958,
+      "step": 5073
+    },
+    {
+      "epoch": 0.434566632408359,
+      "grad_norm": 0.7503798007965088,
+      "learning_rate": 0.0001,
+      "loss": 0.6316,
+      "step": 5074
+    },
+    {
+      "epoch": 0.43465227817745805,
+      "grad_norm": 1.1786887645721436,
+      "learning_rate": 0.0001,
+      "loss": 0.7704,
+      "step": 5075
+    },
+    {
+      "epoch": 0.434737923946557,
+      "grad_norm": 0.4569133520126343,
+      "learning_rate": 0.0001,
+      "loss": 0.655,
+      "step": 5076
+    },
+    {
+      "epoch": 0.43482356971565606,
+      "grad_norm": 0.8212758302688599,
+      "learning_rate": 0.0001,
+      "loss": 0.67,
+      "step": 5077
+    },
+    {
+      "epoch": 0.43490921548475503,
+      "grad_norm": 1.1137181520462036,
+      "learning_rate": 0.0001,
+      "loss": 1.3011,
+      "step": 5078
+    },
+    {
+      "epoch": 0.43499486125385406,
+      "grad_norm": 1.021937370300293,
+      "learning_rate": 0.0001,
+      "loss": 1.2782,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4350805070229531,
+      "grad_norm": 0.6656599640846252,
+      "learning_rate": 0.0001,
+      "loss": 0.7059,
+      "step": 5080
+    },
+    {
+      "epoch": 0.43516615279205206,
+      "grad_norm": 0.5080909132957458,
+      "learning_rate": 0.0001,
+      "loss": 0.915,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4352517985611511,
+      "grad_norm": 0.6813841462135315,
+      "learning_rate": 0.0001,
+      "loss": 0.9236,
+      "step": 5082
+    },
+    {
+      "epoch": 0.43533744433025007,
+      "grad_norm": 1.0503475666046143,
+      "learning_rate": 0.0001,
+      "loss": 1.1824,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4354230900993491,
+      "grad_norm": 0.8719230890274048,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 5084
+    },
+    {
+      "epoch": 0.43550873586844807,
+      "grad_norm": 0.6874189972877502,
+      "learning_rate": 0.0001,
+      "loss": 0.5978,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4355943816375471,
+      "grad_norm": 0.8941506743431091,
+      "learning_rate": 0.0001,
+      "loss": 1.3171,
+      "step": 5086
+    },
+    {
+      "epoch": 0.43568002740664613,
+      "grad_norm": 0.5679627060890198,
+      "learning_rate": 0.0001,
+      "loss": 0.6752,
+      "step": 5087
+    },
+    {
+      "epoch": 0.4357656731757451,
+      "grad_norm": 0.7988864183425903,
+      "learning_rate": 0.0001,
+      "loss": 1.2226,
+      "step": 5088
+    },
+    {
+      "epoch": 0.43585131894484413,
+      "grad_norm": 0.8279475569725037,
+      "learning_rate": 0.0001,
+      "loss": 0.7665,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4359369647139431,
+      "grad_norm": 0.518819272518158,
+      "learning_rate": 0.0001,
+      "loss": 0.8401,
+      "step": 5090
+    },
+    {
+      "epoch": 0.43602261048304214,
+      "grad_norm": 0.5706796646118164,
+      "learning_rate": 0.0001,
+      "loss": 0.8423,
+      "step": 5091
+    },
+    {
+      "epoch": 0.43610825625214117,
+      "grad_norm": 1.4381864070892334,
+      "learning_rate": 0.0001,
+      "loss": 1.2129,
+      "step": 5092
+    },
+    {
+      "epoch": 0.43619390202124014,
+      "grad_norm": 0.4656867980957031,
+      "learning_rate": 0.0001,
+      "loss": 0.2508,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4362795477903392,
+      "grad_norm": 0.6324881911277771,
+      "learning_rate": 0.0001,
+      "loss": 0.8954,
+      "step": 5094
+    },
+    {
+      "epoch": 0.43636519355943815,
+      "grad_norm": 0.7485379576683044,
+      "learning_rate": 0.0001,
+      "loss": 0.9849,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4364508393285372,
+      "grad_norm": 0.6472955942153931,
+      "learning_rate": 0.0001,
+      "loss": 1.0847,
+      "step": 5096
+    },
+    {
+      "epoch": 0.43653648509763615,
+      "grad_norm": 0.7215884923934937,
+      "learning_rate": 0.0001,
+      "loss": 0.8998,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4366221308667352,
+      "grad_norm": 1.4356807470321655,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4367077766358342,
+      "grad_norm": 0.7744952440261841,
+      "learning_rate": 0.0001,
+      "loss": 0.8916,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4367934224049332,
+      "grad_norm": 0.8533983826637268,
+      "learning_rate": 0.0001,
+      "loss": 0.738,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4368790681740322,
+      "grad_norm": 1.0332037210464478,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 5101
+    },
+    {
+      "epoch": 0.4369647139431312,
+      "grad_norm": 0.7516844272613525,
+      "learning_rate": 0.0001,
+      "loss": 0.7177,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4370503597122302,
+      "grad_norm": 1.1199145317077637,
+      "learning_rate": 0.0001,
+      "loss": 1.2051,
+      "step": 5103
+    },
+    {
+      "epoch": 0.43713600548132925,
+      "grad_norm": 0.779935896396637,
+      "learning_rate": 0.0001,
+      "loss": 0.6753,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4372216512504282,
+      "grad_norm": 1.0785032510757446,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 5105
+    },
+    {
+      "epoch": 0.43730729701952725,
+      "grad_norm": 0.4780031442642212,
+      "learning_rate": 0.0001,
+      "loss": 0.9541,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4373929427886262,
+      "grad_norm": 0.6790104508399963,
+      "learning_rate": 0.0001,
+      "loss": 0.4584,
+      "step": 5107
+    },
+    {
+      "epoch": 0.43747858855772526,
+      "grad_norm": 0.8826776742935181,
+      "learning_rate": 0.0001,
+      "loss": 0.7938,
+      "step": 5108
+    },
+    {
+      "epoch": 0.43756423432682423,
+      "grad_norm": 0.8540782928466797,
+      "learning_rate": 0.0001,
+      "loss": 0.4946,
+      "step": 5109
+    },
+    {
+      "epoch": 0.43764988009592326,
+      "grad_norm": 0.9622606039047241,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4377355258650223,
+      "grad_norm": 0.5316354036331177,
+      "learning_rate": 0.0001,
+      "loss": 0.794,
+      "step": 5111
+    },
+    {
+      "epoch": 0.43782117163412126,
+      "grad_norm": 0.3106464445590973,
+      "learning_rate": 0.0001,
+      "loss": 0.452,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4379068174032203,
+      "grad_norm": 0.7392357587814331,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 5113
+    },
+    {
+      "epoch": 0.43799246317231927,
+      "grad_norm": 1.1135973930358887,
+      "learning_rate": 0.0001,
+      "loss": 1.1929,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4380781089414183,
+      "grad_norm": 0.8074598908424377,
+      "learning_rate": 0.0001,
+      "loss": 0.926,
+      "step": 5115
+    },
+    {
+      "epoch": 0.43816375471051733,
+      "grad_norm": 0.4313228726387024,
+      "learning_rate": 0.0001,
+      "loss": 0.4735,
+      "step": 5116
+    },
+    {
+      "epoch": 0.4382494004796163,
+      "grad_norm": 0.8000640869140625,
+      "learning_rate": 0.0001,
+      "loss": 0.6435,
+      "step": 5117
+    },
+    {
+      "epoch": 0.43833504624871533,
+      "grad_norm": 0.7567710280418396,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4384206920178143,
+      "grad_norm": 1.0796022415161133,
+      "learning_rate": 0.0001,
+      "loss": 1.2197,
+      "step": 5119
+    },
+    {
+      "epoch": 0.43850633778691334,
+      "grad_norm": 0.6777092814445496,
+      "learning_rate": 0.0001,
+      "loss": 0.437,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4385919835560123,
+      "grad_norm": 0.46093061566352844,
+      "learning_rate": 0.0001,
+      "loss": 0.478,
+      "step": 5121
+    },
+    {
+      "epoch": 0.43867762932511134,
+      "grad_norm": 0.34502115845680237,
+      "learning_rate": 0.0001,
+      "loss": 0.4269,
+      "step": 5122
+    },
+    {
+      "epoch": 0.43876327509421037,
+      "grad_norm": 1.4046014547348022,
+      "learning_rate": 0.0001,
+      "loss": 1.0264,
+      "step": 5123
+    },
+    {
+      "epoch": 0.43884892086330934,
+      "grad_norm": 0.5915937423706055,
+      "learning_rate": 0.0001,
+      "loss": 1.1709,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4389345666324084,
+      "grad_norm": 0.822709858417511,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 5125
+    },
+    {
+      "epoch": 0.43902021240150735,
+      "grad_norm": 1.1052027940750122,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4391058581706064,
+      "grad_norm": 0.235282301902771,
+      "learning_rate": 0.0001,
+      "loss": 0.2136,
+      "step": 5127
+    },
+    {
+      "epoch": 0.43919150393970535,
+      "grad_norm": 0.9361926913261414,
+      "learning_rate": 0.0001,
+      "loss": 0.8683,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4392771497088044,
+      "grad_norm": 0.733879804611206,
+      "learning_rate": 0.0001,
+      "loss": 0.9755,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4393627954779034,
+      "grad_norm": 0.4796643555164337,
+      "learning_rate": 0.0001,
+      "loss": 0.7121,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4394484412470024,
+      "grad_norm": 0.9258108139038086,
+      "learning_rate": 0.0001,
+      "loss": 1.005,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4395340870161014,
+      "grad_norm": 0.4768390357494354,
+      "learning_rate": 0.0001,
+      "loss": 0.5321,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4396197327852004,
+      "grad_norm": 1.1036423444747925,
+      "learning_rate": 0.0001,
+      "loss": 1.2221,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4397053785542994,
+      "grad_norm": 0.5799343585968018,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 5134
+    },
+    {
+      "epoch": 0.43979102432339845,
+      "grad_norm": 1.140048623085022,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4398766700924974,
+      "grad_norm": 0.7873648405075073,
+      "learning_rate": 0.0001,
+      "loss": 1.3811,
+      "step": 5136
+    },
+    {
+      "epoch": 0.43996231586159645,
+      "grad_norm": 1.124009370803833,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 5137
+    },
+    {
+      "epoch": 0.44004796163069543,
+      "grad_norm": 0.7869529128074646,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 5138
+    },
+    {
+      "epoch": 0.44013360739979446,
+      "grad_norm": 1.2214473485946655,
+      "learning_rate": 0.0001,
+      "loss": 1.3739,
+      "step": 5139
+    },
+    {
+      "epoch": 0.44021925316889343,
+      "grad_norm": 1.0419243574142456,
+      "learning_rate": 0.0001,
+      "loss": 0.6161,
+      "step": 5140
+    },
+    {
+      "epoch": 0.44030489893799246,
+      "grad_norm": 0.7438523173332214,
+      "learning_rate": 0.0001,
+      "loss": 1.2503,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4403905447070915,
+      "grad_norm": 0.49661314487457275,
+      "learning_rate": 0.0001,
+      "loss": 0.4583,
+      "step": 5142
+    },
+    {
+      "epoch": 0.44047619047619047,
+      "grad_norm": 1.239661455154419,
+      "learning_rate": 0.0001,
+      "loss": 1.1045,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4405618362452895,
+      "grad_norm": 0.3978006839752197,
+      "learning_rate": 0.0001,
+      "loss": 0.5849,
+      "step": 5144
+    },
+    {
+      "epoch": 0.44064748201438847,
+      "grad_norm": 0.8872053623199463,
+      "learning_rate": 0.0001,
+      "loss": 1.3224,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4407331277834875,
+      "grad_norm": 0.8778040409088135,
+      "learning_rate": 0.0001,
+      "loss": 1.1379,
+      "step": 5146
+    },
+    {
+      "epoch": 0.44081877355258653,
+      "grad_norm": 0.8882962465286255,
+      "learning_rate": 0.0001,
+      "loss": 1.2911,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4409044193216855,
+      "grad_norm": 0.737378716468811,
+      "learning_rate": 0.0001,
+      "loss": 0.9489,
+      "step": 5148
+    },
+    {
+      "epoch": 0.44099006509078453,
+      "grad_norm": 0.7551065683364868,
+      "learning_rate": 0.0001,
+      "loss": 1.0962,
+      "step": 5149
+    },
+    {
+      "epoch": 0.4410757108598835,
+      "grad_norm": 0.8190209269523621,
+      "learning_rate": 0.0001,
+      "loss": 1.0543,
+      "step": 5150
+    },
+    {
+      "epoch": 0.44116135662898254,
+      "grad_norm": 1.0962241888046265,
+      "learning_rate": 0.0001,
+      "loss": 1.1666,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4412470023980815,
+      "grad_norm": 0.8965091705322266,
+      "learning_rate": 0.0001,
+      "loss": 1.1325,
+      "step": 5152
+    },
+    {
+      "epoch": 0.44133264816718054,
+      "grad_norm": 0.8958146572113037,
+      "learning_rate": 0.0001,
+      "loss": 1.2201,
+      "step": 5153
+    },
+    {
+      "epoch": 0.44141829393627957,
+      "grad_norm": 0.6693230867385864,
+      "learning_rate": 0.0001,
+      "loss": 0.682,
+      "step": 5154
+    },
+    {
+      "epoch": 0.44150393970537855,
+      "grad_norm": 1.4122955799102783,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4415895854744776,
+      "grad_norm": 0.9499876499176025,
+      "learning_rate": 0.0001,
+      "loss": 0.9128,
+      "step": 5156
+    },
+    {
+      "epoch": 0.44167523124357655,
+      "grad_norm": 1.6484360694885254,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4417608770126756,
+      "grad_norm": 1.1629599332809448,
+      "learning_rate": 0.0001,
+      "loss": 1.2109,
+      "step": 5158
+    },
+    {
+      "epoch": 0.44184652278177455,
+      "grad_norm": 0.8290032744407654,
+      "learning_rate": 0.0001,
+      "loss": 0.9897,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4419321685508736,
+      "grad_norm": 0.5760030150413513,
+      "learning_rate": 0.0001,
+      "loss": 0.5691,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4420178143199726,
+      "grad_norm": 0.5693584084510803,
+      "learning_rate": 0.0001,
+      "loss": 1.1455,
+      "step": 5161
+    },
+    {
+      "epoch": 0.4421034600890716,
+      "grad_norm": 1.0241460800170898,
+      "learning_rate": 0.0001,
+      "loss": 1.0386,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4421891058581706,
+      "grad_norm": 0.7619434595108032,
+      "learning_rate": 0.0001,
+      "loss": 0.4834,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4422747516272696,
+      "grad_norm": 0.5789291858673096,
+      "learning_rate": 0.0001,
+      "loss": 0.4137,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4423603973963686,
+      "grad_norm": 0.5727609395980835,
+      "learning_rate": 0.0001,
+      "loss": 1.0645,
+      "step": 5165
+    },
+    {
+      "epoch": 0.44244604316546765,
+      "grad_norm": 0.6380529999732971,
+      "learning_rate": 0.0001,
+      "loss": 0.5328,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4425316889345666,
+      "grad_norm": 0.8677741289138794,
+      "learning_rate": 0.0001,
+      "loss": 0.9482,
+      "step": 5167
+    },
+    {
+      "epoch": 0.44261733470366565,
+      "grad_norm": 0.8193942904472351,
+      "learning_rate": 0.0001,
+      "loss": 1.1732,
+      "step": 5168
+    },
+    {
+      "epoch": 0.44270298047276463,
+      "grad_norm": 1.0240463018417358,
+      "learning_rate": 0.0001,
+      "loss": 0.7911,
+      "step": 5169
+    },
+    {
+      "epoch": 0.44278862624186366,
+      "grad_norm": 1.3590604066848755,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 5170
+    },
+    {
+      "epoch": 0.44287427201096263,
+      "grad_norm": 1.1942747831344604,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 5171
+    },
+    {
+      "epoch": 0.44295991778006166,
+      "grad_norm": 0.8407852053642273,
+      "learning_rate": 0.0001,
+      "loss": 0.6393,
+      "step": 5172
+    },
+    {
+      "epoch": 0.4430455635491607,
+      "grad_norm": 0.9525219798088074,
+      "learning_rate": 0.0001,
+      "loss": 1.0132,
+      "step": 5173
+    },
+    {
+      "epoch": 0.44313120931825967,
+      "grad_norm": 0.3860262334346771,
+      "learning_rate": 0.0001,
+      "loss": 0.6798,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4432168550873587,
+      "grad_norm": 0.6700161695480347,
+      "learning_rate": 0.0001,
+      "loss": 1.1453,
+      "step": 5175
+    },
+    {
+      "epoch": 0.44330250085645767,
+      "grad_norm": 0.8820961713790894,
+      "learning_rate": 0.0001,
+      "loss": 0.7716,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4433881466255567,
+      "grad_norm": 0.8276817798614502,
+      "learning_rate": 0.0001,
+      "loss": 1.9109,
+      "step": 5177
+    },
+    {
+      "epoch": 0.44347379239465573,
+      "grad_norm": 0.6646428108215332,
+      "learning_rate": 0.0001,
+      "loss": 1.0844,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4435594381637547,
+      "grad_norm": 0.8467397093772888,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 5179
+    },
+    {
+      "epoch": 0.44364508393285373,
+      "grad_norm": 1.652514100074768,
+      "learning_rate": 0.0001,
+      "loss": 1.395,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4437307297019527,
+      "grad_norm": 0.38841426372528076,
+      "learning_rate": 0.0001,
+      "loss": 0.7596,
+      "step": 5181
+    },
+    {
+      "epoch": 0.44381637547105174,
+      "grad_norm": 0.6483497619628906,
+      "learning_rate": 0.0001,
+      "loss": 1.1153,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4439020212401507,
+      "grad_norm": 0.6711593270301819,
+      "learning_rate": 0.0001,
+      "loss": 1.0036,
+      "step": 5183
+    },
+    {
+      "epoch": 0.44398766700924974,
+      "grad_norm": 1.1292256116867065,
+      "learning_rate": 0.0001,
+      "loss": 0.9098,
+      "step": 5184
+    },
+    {
+      "epoch": 0.44407331277834877,
+      "grad_norm": 0.6638242602348328,
+      "learning_rate": 0.0001,
+      "loss": 0.3006,
+      "step": 5185
+    },
+    {
+      "epoch": 0.44415895854744775,
+      "grad_norm": 0.5851790904998779,
+      "learning_rate": 0.0001,
+      "loss": 0.9194,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4442446043165468,
+      "grad_norm": 0.7370529174804688,
+      "learning_rate": 0.0001,
+      "loss": 1.0248,
+      "step": 5187
+    },
+    {
+      "epoch": 0.44433025008564575,
+      "grad_norm": 0.5424681305885315,
+      "learning_rate": 0.0001,
+      "loss": 0.4012,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4444158958547448,
+      "grad_norm": 0.7091993689537048,
+      "learning_rate": 0.0001,
+      "loss": 0.7719,
+      "step": 5189
+    },
+    {
+      "epoch": 0.44450154162384375,
+      "grad_norm": 1.0523958206176758,
+      "learning_rate": 0.0001,
+      "loss": 0.7222,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4445871873929428,
+      "grad_norm": 0.8030400276184082,
+      "learning_rate": 0.0001,
+      "loss": 1.0674,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4446728331620418,
+      "grad_norm": 0.7316297888755798,
+      "learning_rate": 0.0001,
+      "loss": 0.9022,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4447584789311408,
+      "grad_norm": 0.8569072484970093,
+      "learning_rate": 0.0001,
+      "loss": 1.1492,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4448441247002398,
+      "grad_norm": 0.8445351719856262,
+      "learning_rate": 0.0001,
+      "loss": 1.2012,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4449297704693388,
+      "grad_norm": 1.2187871932983398,
+      "learning_rate": 0.0001,
+      "loss": 1.2473,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4450154162384378,
+      "grad_norm": 0.8608721494674683,
+      "learning_rate": 0.0001,
+      "loss": 0.9295,
+      "step": 5196
+    },
+    {
+      "epoch": 0.44510106200753685,
+      "grad_norm": 0.9048358798027039,
+      "learning_rate": 0.0001,
+      "loss": 1.2871,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4451867077766358,
+      "grad_norm": 0.7928715348243713,
+      "learning_rate": 0.0001,
+      "loss": 0.9553,
+      "step": 5198
+    },
+    {
+      "epoch": 0.44527235354573486,
+      "grad_norm": 0.5582913160324097,
+      "learning_rate": 0.0001,
+      "loss": 0.5947,
+      "step": 5199
+    },
+    {
+      "epoch": 0.44535799931483383,
+      "grad_norm": 0.9858488440513611,
+      "learning_rate": 0.0001,
+      "loss": 0.986,
+      "step": 5200
+    },
+    {
+      "epoch": 0.44544364508393286,
+      "grad_norm": 0.7285011410713196,
+      "learning_rate": 0.0001,
+      "loss": 0.7681,
+      "step": 5201
+    },
+    {
+      "epoch": 0.44552929085303183,
+      "grad_norm": 0.7926207184791565,
+      "learning_rate": 0.0001,
+      "loss": 0.7277,
+      "step": 5202
+    },
+    {
+      "epoch": 0.44561493662213086,
+      "grad_norm": 0.583648681640625,
+      "learning_rate": 0.0001,
+      "loss": 0.4972,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4457005823912299,
+      "grad_norm": 0.859714150428772,
+      "learning_rate": 0.0001,
+      "loss": 1.0671,
+      "step": 5204
+    },
+    {
+      "epoch": 0.44578622816032887,
+      "grad_norm": 0.5596405863761902,
+      "learning_rate": 0.0001,
+      "loss": 0.4221,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4458718739294279,
+      "grad_norm": 1.1424980163574219,
+      "learning_rate": 0.0001,
+      "loss": 1.7242,
+      "step": 5206
+    },
+    {
+      "epoch": 0.44595751969852687,
+      "grad_norm": 0.8765913844108582,
+      "learning_rate": 0.0001,
+      "loss": 1.0936,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4460431654676259,
+      "grad_norm": 0.570728063583374,
+      "learning_rate": 0.0001,
+      "loss": 1.1455,
+      "step": 5208
+    },
+    {
+      "epoch": 0.44612881123672493,
+      "grad_norm": 0.3040480315685272,
+      "learning_rate": 0.0001,
+      "loss": 0.5078,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4462144570058239,
+      "grad_norm": 0.7160006761550903,
+      "learning_rate": 0.0001,
+      "loss": 0.989,
+      "step": 5210
+    },
+    {
+      "epoch": 0.44630010277492294,
+      "grad_norm": 0.6872061491012573,
+      "learning_rate": 0.0001,
+      "loss": 1.2596,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4463857485440219,
+      "grad_norm": 0.3891212046146393,
+      "learning_rate": 0.0001,
+      "loss": 0.6434,
+      "step": 5212
+    },
+    {
+      "epoch": 0.44647139431312094,
+      "grad_norm": 1.212595820426941,
+      "learning_rate": 0.0001,
+      "loss": 0.787,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4465570400822199,
+      "grad_norm": 0.9821494221687317,
+      "learning_rate": 0.0001,
+      "loss": 0.7174,
+      "step": 5214
+    },
+    {
+      "epoch": 0.44664268585131894,
+      "grad_norm": 0.6363126635551453,
+      "learning_rate": 0.0001,
+      "loss": 1.2085,
+      "step": 5215
+    },
+    {
+      "epoch": 0.446728331620418,
+      "grad_norm": 0.7241467833518982,
+      "learning_rate": 0.0001,
+      "loss": 0.5319,
+      "step": 5216
+    },
+    {
+      "epoch": 0.44681397738951695,
+      "grad_norm": 0.5958557724952698,
+      "learning_rate": 0.0001,
+      "loss": 0.6873,
+      "step": 5217
+    },
+    {
+      "epoch": 0.446899623158616,
+      "grad_norm": 0.606508731842041,
+      "learning_rate": 0.0001,
+      "loss": 0.9198,
+      "step": 5218
+    },
+    {
+      "epoch": 0.44698526892771495,
+      "grad_norm": 1.5176845788955688,
+      "learning_rate": 0.0001,
+      "loss": 0.9618,
+      "step": 5219
+    },
+    {
+      "epoch": 0.447070914696814,
+      "grad_norm": 0.4207318425178528,
+      "learning_rate": 0.0001,
+      "loss": 0.4452,
+      "step": 5220
+    },
+    {
+      "epoch": 0.447156560465913,
+      "grad_norm": 0.7019359469413757,
+      "learning_rate": 0.0001,
+      "loss": 0.77,
+      "step": 5221
+    },
+    {
+      "epoch": 0.447242206235012,
+      "grad_norm": 0.7785884737968445,
+      "learning_rate": 0.0001,
+      "loss": 0.5843,
+      "step": 5222
+    },
+    {
+      "epoch": 0.447327852004111,
+      "grad_norm": 0.8901249766349792,
+      "learning_rate": 0.0001,
+      "loss": 1.3084,
+      "step": 5223
+    },
+    {
+      "epoch": 0.44741349777321,
+      "grad_norm": 0.7551905512809753,
+      "learning_rate": 0.0001,
+      "loss": 1.0587,
+      "step": 5224
+    },
+    {
+      "epoch": 0.447499143542309,
+      "grad_norm": 0.6091492772102356,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 5225
+    },
+    {
+      "epoch": 0.447584789311408,
+      "grad_norm": 1.0955369472503662,
+      "learning_rate": 0.0001,
+      "loss": 0.822,
+      "step": 5226
+    },
+    {
+      "epoch": 0.447670435080507,
+      "grad_norm": 0.44863709807395935,
+      "learning_rate": 0.0001,
+      "loss": 0.8535,
+      "step": 5227
+    },
+    {
+      "epoch": 0.44775608084960605,
+      "grad_norm": 0.5350818037986755,
+      "learning_rate": 0.0001,
+      "loss": 0.766,
+      "step": 5228
+    },
+    {
+      "epoch": 0.447841726618705,
+      "grad_norm": 0.5543168783187866,
+      "learning_rate": 0.0001,
+      "loss": 1.0263,
+      "step": 5229
+    },
+    {
+      "epoch": 0.44792737238780406,
+      "grad_norm": 0.7326213121414185,
+      "learning_rate": 0.0001,
+      "loss": 1.3095,
+      "step": 5230
+    },
+    {
+      "epoch": 0.44801301815690303,
+      "grad_norm": 0.8018301725387573,
+      "learning_rate": 0.0001,
+      "loss": 0.7092,
+      "step": 5231
+    },
+    {
+      "epoch": 0.44809866392600206,
+      "grad_norm": 1.0500853061676025,
+      "learning_rate": 0.0001,
+      "loss": 1.1476,
+      "step": 5232
+    },
+    {
+      "epoch": 0.44818430969510104,
+      "grad_norm": 0.7815940380096436,
+      "learning_rate": 0.0001,
+      "loss": 1.2801,
+      "step": 5233
+    },
+    {
+      "epoch": 0.44826995546420006,
+      "grad_norm": 1.1621636152267456,
+      "learning_rate": 0.0001,
+      "loss": 1.3049,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4483556012332991,
+      "grad_norm": 0.6897354125976562,
+      "learning_rate": 0.0001,
+      "loss": 0.5507,
+      "step": 5235
+    },
+    {
+      "epoch": 0.44844124700239807,
+      "grad_norm": 0.9008827209472656,
+      "learning_rate": 0.0001,
+      "loss": 1.031,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4485268927714971,
+      "grad_norm": 0.896937906742096,
+      "learning_rate": 0.0001,
+      "loss": 0.9653,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4486125385405961,
+      "grad_norm": 0.5659229159355164,
+      "learning_rate": 0.0001,
+      "loss": 0.9082,
+      "step": 5238
+    },
+    {
+      "epoch": 0.4486981843096951,
+      "grad_norm": 1.0294079780578613,
+      "learning_rate": 0.0001,
+      "loss": 1.1023,
+      "step": 5239
+    },
+    {
+      "epoch": 0.44878383007879413,
+      "grad_norm": 0.535696804523468,
+      "learning_rate": 0.0001,
+      "loss": 0.1882,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4488694758478931,
+      "grad_norm": 1.100218415260315,
+      "learning_rate": 0.0001,
+      "loss": 0.9496,
+      "step": 5241
+    },
+    {
+      "epoch": 0.44895512161699214,
+      "grad_norm": 0.5484207272529602,
+      "learning_rate": 0.0001,
+      "loss": 0.897,
+      "step": 5242
+    },
+    {
+      "epoch": 0.4490407673860911,
+      "grad_norm": 0.6512097120285034,
+      "learning_rate": 0.0001,
+      "loss": 0.9883,
+      "step": 5243
+    },
+    {
+      "epoch": 0.44912641315519014,
+      "grad_norm": 1.0658724308013916,
+      "learning_rate": 0.0001,
+      "loss": 0.995,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4492120589242891,
+      "grad_norm": 0.47368836402893066,
+      "learning_rate": 0.0001,
+      "loss": 0.921,
+      "step": 5245
+    },
+    {
+      "epoch": 0.44929770469338814,
+      "grad_norm": 0.706858217716217,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4493833504624872,
+      "grad_norm": 0.8250250220298767,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 5247
+    },
+    {
+      "epoch": 0.44946899623158615,
+      "grad_norm": 1.2295176982879639,
+      "learning_rate": 0.0001,
+      "loss": 1.3005,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4495546420006852,
+      "grad_norm": 1.4577884674072266,
+      "learning_rate": 0.0001,
+      "loss": 1.1587,
+      "step": 5249
+    },
+    {
+      "epoch": 0.44964028776978415,
+      "grad_norm": 0.4679050147533417,
+      "learning_rate": 0.0001,
+      "loss": 0.2742,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4497259335388832,
+      "grad_norm": 0.8448243737220764,
+      "learning_rate": 0.0001,
+      "loss": 1.2731,
+      "step": 5251
+    },
+    {
+      "epoch": 0.4498115793079822,
+      "grad_norm": 0.5438274145126343,
+      "learning_rate": 0.0001,
+      "loss": 0.3373,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4498972250770812,
+      "grad_norm": 0.5822242498397827,
+      "learning_rate": 0.0001,
+      "loss": 0.7071,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4499828708461802,
+      "grad_norm": 0.6707018613815308,
+      "learning_rate": 0.0001,
+      "loss": 0.6493,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4500685166152792,
+      "grad_norm": 0.8304451704025269,
+      "learning_rate": 0.0001,
+      "loss": 1.0259,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4501541623843782,
+      "grad_norm": 0.735582709312439,
+      "learning_rate": 0.0001,
+      "loss": 1.0496,
+      "step": 5256
+    },
+    {
+      "epoch": 0.4502398081534772,
+      "grad_norm": 0.8116124272346497,
+      "learning_rate": 0.0001,
+      "loss": 1.2712,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4503254539225762,
+      "grad_norm": 1.114503026008606,
+      "learning_rate": 0.0001,
+      "loss": 0.9833,
+      "step": 5258
+    },
+    {
+      "epoch": 0.45041109969167525,
+      "grad_norm": 0.798523485660553,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 5259
+    },
+    {
+      "epoch": 0.45049674546077423,
+      "grad_norm": 0.9432134032249451,
+      "learning_rate": 0.0001,
+      "loss": 1.1052,
+      "step": 5260
+    },
+    {
+      "epoch": 0.45058239122987326,
+      "grad_norm": 1.0866038799285889,
+      "learning_rate": 0.0001,
+      "loss": 1.3491,
+      "step": 5261
+    },
+    {
+      "epoch": 0.45066803699897223,
+      "grad_norm": 1.0869801044464111,
+      "learning_rate": 0.0001,
+      "loss": 0.9516,
+      "step": 5262
+    },
+    {
+      "epoch": 0.45075368276807126,
+      "grad_norm": 0.5659863352775574,
+      "learning_rate": 0.0001,
+      "loss": 1.0183,
+      "step": 5263
+    },
+    {
+      "epoch": 0.45083932853717024,
+      "grad_norm": 1.4984863996505737,
+      "learning_rate": 0.0001,
+      "loss": 0.8324,
+      "step": 5264
+    },
+    {
+      "epoch": 0.45092497430626927,
+      "grad_norm": 0.7123324275016785,
+      "learning_rate": 0.0001,
+      "loss": 1.2838,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4510106200753683,
+      "grad_norm": 0.6710283160209656,
+      "learning_rate": 0.0001,
+      "loss": 1.0136,
+      "step": 5266
+    },
+    {
+      "epoch": 0.45109626584446727,
+      "grad_norm": 1.0215113162994385,
+      "learning_rate": 0.0001,
+      "loss": 0.7496,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4511819116135663,
+      "grad_norm": 0.796882688999176,
+      "learning_rate": 0.0001,
+      "loss": 0.7579,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4512675573826653,
+      "grad_norm": 0.6056481599807739,
+      "learning_rate": 0.0001,
+      "loss": 1.0205,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4513532031517643,
+      "grad_norm": 0.49826452136039734,
+      "learning_rate": 0.0001,
+      "loss": 0.7777,
+      "step": 5270
+    },
+    {
+      "epoch": 0.45143884892086333,
+      "grad_norm": 0.32785919308662415,
+      "learning_rate": 0.0001,
+      "loss": 0.4937,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4515244946899623,
+      "grad_norm": 0.5710442066192627,
+      "learning_rate": 0.0001,
+      "loss": 0.5376,
+      "step": 5272
+    },
+    {
+      "epoch": 0.45161014045906134,
+      "grad_norm": 1.2856475114822388,
+      "learning_rate": 0.0001,
+      "loss": 1.0135,
+      "step": 5273
+    },
+    {
+      "epoch": 0.4516957862281603,
+      "grad_norm": 0.8061153888702393,
+      "learning_rate": 0.0001,
+      "loss": 1.2968,
+      "step": 5274
+    },
+    {
+      "epoch": 0.45178143199725934,
+      "grad_norm": 0.5980777144432068,
+      "learning_rate": 0.0001,
+      "loss": 1.0756,
+      "step": 5275
+    },
+    {
+      "epoch": 0.4518670777663583,
+      "grad_norm": 0.8072212934494019,
+      "learning_rate": 0.0001,
+      "loss": 0.8544,
+      "step": 5276
+    },
+    {
+      "epoch": 0.45195272353545735,
+      "grad_norm": 0.9048564434051514,
+      "learning_rate": 0.0001,
+      "loss": 1.1386,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4520383693045564,
+      "grad_norm": 0.9546582102775574,
+      "learning_rate": 0.0001,
+      "loss": 1.0657,
+      "step": 5278
+    },
+    {
+      "epoch": 0.45212401507365535,
+      "grad_norm": 0.3927929699420929,
+      "learning_rate": 0.0001,
+      "loss": 0.4524,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4522096608427544,
+      "grad_norm": 0.8346943855285645,
+      "learning_rate": 0.0001,
+      "loss": 1.3882,
+      "step": 5280
+    },
+    {
+      "epoch": 0.45229530661185335,
+      "grad_norm": 1.152428388595581,
+      "learning_rate": 0.0001,
+      "loss": 0.6672,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4523809523809524,
+      "grad_norm": 1.07718026638031,
+      "learning_rate": 0.0001,
+      "loss": 0.7768,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4524665981500514,
+      "grad_norm": 0.59132981300354,
+      "learning_rate": 0.0001,
+      "loss": 0.4365,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4525522439191504,
+      "grad_norm": 1.0956568717956543,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4526378896882494,
+      "grad_norm": 0.6826682090759277,
+      "learning_rate": 0.0001,
+      "loss": 0.9334,
+      "step": 5285
+    },
+    {
+      "epoch": 0.4527235354573484,
+      "grad_norm": 0.5696227550506592,
+      "learning_rate": 0.0001,
+      "loss": 0.6915,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4528091812264474,
+      "grad_norm": 0.6407721638679504,
+      "learning_rate": 0.0001,
+      "loss": 0.4683,
+      "step": 5287
+    },
+    {
+      "epoch": 0.4528948269955464,
+      "grad_norm": 0.6787132024765015,
+      "learning_rate": 0.0001,
+      "loss": 1.2255,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4529804727646454,
+      "grad_norm": 0.8375104069709778,
+      "learning_rate": 0.0001,
+      "loss": 0.7394,
+      "step": 5289
+    },
+    {
+      "epoch": 0.45306611853374446,
+      "grad_norm": 0.77740079164505,
+      "learning_rate": 0.0001,
+      "loss": 0.7233,
+      "step": 5290
+    },
+    {
+      "epoch": 0.45315176430284343,
+      "grad_norm": 0.9251188039779663,
+      "learning_rate": 0.0001,
+      "loss": 0.9413,
+      "step": 5291
+    },
+    {
+      "epoch": 0.45323741007194246,
+      "grad_norm": 0.9389598965644836,
+      "learning_rate": 0.0001,
+      "loss": 0.903,
+      "step": 5292
+    },
+    {
+      "epoch": 0.45332305584104143,
+      "grad_norm": 0.826842188835144,
+      "learning_rate": 0.0001,
+      "loss": 1.0726,
+      "step": 5293
+    },
+    {
+      "epoch": 0.45340870161014046,
+      "grad_norm": 0.9180453419685364,
+      "learning_rate": 0.0001,
+      "loss": 1.0983,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4534943473792395,
+      "grad_norm": 0.9370865225791931,
+      "learning_rate": 0.0001,
+      "loss": 1.2595,
+      "step": 5295
+    },
+    {
+      "epoch": 0.45357999314833847,
+      "grad_norm": 0.7188237905502319,
+      "learning_rate": 0.0001,
+      "loss": 0.7502,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4536656389174375,
+      "grad_norm": 0.3654862344264984,
+      "learning_rate": 0.0001,
+      "loss": 0.601,
+      "step": 5297
+    },
+    {
+      "epoch": 0.45375128468653647,
+      "grad_norm": 0.8795923590660095,
+      "learning_rate": 0.0001,
+      "loss": 0.9799,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4538369304556355,
+      "grad_norm": 0.7908286452293396,
+      "learning_rate": 0.0001,
+      "loss": 1.0523,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4539225762247345,
+      "grad_norm": 0.8246570825576782,
+      "learning_rate": 0.0001,
+      "loss": 0.8371,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4540082219938335,
+      "grad_norm": 0.7109977602958679,
+      "learning_rate": 0.0001,
+      "loss": 1.1106,
+      "step": 5301
+    },
+    {
+      "epoch": 0.45409386776293253,
+      "grad_norm": 0.4421456754207611,
+      "learning_rate": 0.0001,
+      "loss": 0.5791,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4541795135320315,
+      "grad_norm": 0.8845040202140808,
+      "learning_rate": 0.0001,
+      "loss": 0.9859,
+      "step": 5303
+    },
+    {
+      "epoch": 0.45426515930113054,
+      "grad_norm": 0.9342193007469177,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4543508050702295,
+      "grad_norm": 0.6825669407844543,
+      "learning_rate": 0.0001,
+      "loss": 1.1967,
+      "step": 5305
+    },
+    {
+      "epoch": 0.45443645083932854,
+      "grad_norm": 0.683047890663147,
+      "learning_rate": 0.0001,
+      "loss": 0.3532,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4545220966084275,
+      "grad_norm": 0.6245560050010681,
+      "learning_rate": 0.0001,
+      "loss": 1.3514,
+      "step": 5307
+    },
+    {
+      "epoch": 0.45460774237752655,
+      "grad_norm": 0.715580940246582,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4546933881466256,
+      "grad_norm": 1.2651145458221436,
+      "learning_rate": 0.0001,
+      "loss": 1.1225,
+      "step": 5309
+    },
+    {
+      "epoch": 0.45477903391572455,
+      "grad_norm": 0.8996610045433044,
+      "learning_rate": 0.0001,
+      "loss": 0.89,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4548646796848236,
+      "grad_norm": 0.5729882121086121,
+      "learning_rate": 0.0001,
+      "loss": 0.5396,
+      "step": 5311
+    },
+    {
+      "epoch": 0.45495032545392255,
+      "grad_norm": 0.8477318286895752,
+      "learning_rate": 0.0001,
+      "loss": 0.611,
+      "step": 5312
+    },
+    {
+      "epoch": 0.4550359712230216,
+      "grad_norm": 0.4712734520435333,
+      "learning_rate": 0.0001,
+      "loss": 0.7434,
+      "step": 5313
+    },
+    {
+      "epoch": 0.4551216169921206,
+      "grad_norm": 0.9265299439430237,
+      "learning_rate": 0.0001,
+      "loss": 1.1018,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4552072627612196,
+      "grad_norm": 0.6502110958099365,
+      "learning_rate": 0.0001,
+      "loss": 1.1528,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4552929085303186,
+      "grad_norm": 1.0098990201950073,
+      "learning_rate": 0.0001,
+      "loss": 1.1585,
+      "step": 5316
+    },
+    {
+      "epoch": 0.4553785542994176,
+      "grad_norm": 0.9462293386459351,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 5317
+    },
+    {
+      "epoch": 0.4554642000685166,
+      "grad_norm": 0.8270633220672607,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 5318
+    },
+    {
+      "epoch": 0.4555498458376156,
+      "grad_norm": 1.1979153156280518,
+      "learning_rate": 0.0001,
+      "loss": 1.0053,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4556354916067146,
+      "grad_norm": 0.6839941740036011,
+      "learning_rate": 0.0001,
+      "loss": 1.2584,
+      "step": 5320
+    },
+    {
+      "epoch": 0.45572113737581366,
+      "grad_norm": 0.94957435131073,
+      "learning_rate": 0.0001,
+      "loss": 1.2566,
+      "step": 5321
+    },
+    {
+      "epoch": 0.45580678314491263,
+      "grad_norm": 0.860059916973114,
+      "learning_rate": 0.0001,
+      "loss": 0.9571,
+      "step": 5322
+    },
+    {
+      "epoch": 0.45589242891401166,
+      "grad_norm": 0.9245743751525879,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 5323
+    },
+    {
+      "epoch": 0.45597807468311063,
+      "grad_norm": 0.8262640237808228,
+      "learning_rate": 0.0001,
+      "loss": 0.8886,
+      "step": 5324
+    },
+    {
+      "epoch": 0.45606372045220966,
+      "grad_norm": 0.8270162343978882,
+      "learning_rate": 0.0001,
+      "loss": 1.2189,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4561493662213087,
+      "grad_norm": 0.9227482676506042,
+      "learning_rate": 0.0001,
+      "loss": 1.2689,
+      "step": 5326
+    },
+    {
+      "epoch": 0.45623501199040767,
+      "grad_norm": 0.43882080912590027,
+      "learning_rate": 0.0001,
+      "loss": 0.5017,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4563206577595067,
+      "grad_norm": 1.257652759552002,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 5328
+    },
+    {
+      "epoch": 0.4564063035286057,
+      "grad_norm": 0.7130914330482483,
+      "learning_rate": 0.0001,
+      "loss": 1.1426,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4564919492977047,
+      "grad_norm": 1.6487200260162354,
+      "learning_rate": 0.0001,
+      "loss": 1.3533,
+      "step": 5330
+    },
+    {
+      "epoch": 0.4565775950668037,
+      "grad_norm": 0.8394613265991211,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 5331
+    },
+    {
+      "epoch": 0.4566632408359027,
+      "grad_norm": 0.6072746515274048,
+      "learning_rate": 0.0001,
+      "loss": 0.9266,
+      "step": 5332
+    },
+    {
+      "epoch": 0.45674888660500174,
+      "grad_norm": 0.6091123223304749,
+      "learning_rate": 0.0001,
+      "loss": 0.5959,
+      "step": 5333
+    },
+    {
+      "epoch": 0.4568345323741007,
+      "grad_norm": 0.5220286250114441,
+      "learning_rate": 0.0001,
+      "loss": 0.8932,
+      "step": 5334
+    },
+    {
+      "epoch": 0.45692017814319974,
+      "grad_norm": 0.7222042679786682,
+      "learning_rate": 0.0001,
+      "loss": 0.899,
+      "step": 5335
+    },
+    {
+      "epoch": 0.4570058239122987,
+      "grad_norm": 1.2534902095794678,
+      "learning_rate": 0.0001,
+      "loss": 1.0862,
+      "step": 5336
+    },
+    {
+      "epoch": 0.45709146968139774,
+      "grad_norm": 0.7459235191345215,
+      "learning_rate": 0.0001,
+      "loss": 0.9069,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4571771154504967,
+      "grad_norm": 0.8489087224006653,
+      "learning_rate": 0.0001,
+      "loss": 1.2086,
+      "step": 5338
+    },
+    {
+      "epoch": 0.45726276121959575,
+      "grad_norm": 0.8036373853683472,
+      "learning_rate": 0.0001,
+      "loss": 0.6648,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4573484069886948,
+      "grad_norm": 0.8200138211250305,
+      "learning_rate": 0.0001,
+      "loss": 0.9775,
+      "step": 5340
+    },
+    {
+      "epoch": 0.45743405275779375,
+      "grad_norm": 0.9946182370185852,
+      "learning_rate": 0.0001,
+      "loss": 0.7327,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4575196985268928,
+      "grad_norm": 0.7371546626091003,
+      "learning_rate": 0.0001,
+      "loss": 1.096,
+      "step": 5342
+    },
+    {
+      "epoch": 0.45760534429599176,
+      "grad_norm": 0.6765991449356079,
+      "learning_rate": 0.0001,
+      "loss": 0.5951,
+      "step": 5343
+    },
+    {
+      "epoch": 0.4576909900650908,
+      "grad_norm": 0.8633242249488831,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4577766358341898,
+      "grad_norm": 0.7026433944702148,
+      "learning_rate": 0.0001,
+      "loss": 0.6874,
+      "step": 5345
+    },
+    {
+      "epoch": 0.4578622816032888,
+      "grad_norm": 0.6741832494735718,
+      "learning_rate": 0.0001,
+      "loss": 1.7086,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4579479273723878,
+      "grad_norm": 0.6693118214607239,
+      "learning_rate": 0.0001,
+      "loss": 1.1548,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4580335731414868,
+      "grad_norm": 1.0476969480514526,
+      "learning_rate": 0.0001,
+      "loss": 1.3185,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4581192189105858,
+      "grad_norm": 1.3199197053909302,
+      "learning_rate": 0.0001,
+      "loss": 1.0256,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4582048646796848,
+      "grad_norm": 1.0498019456863403,
+      "learning_rate": 0.0001,
+      "loss": 0.8753,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4582905104487838,
+      "grad_norm": 0.5283469557762146,
+      "learning_rate": 0.0001,
+      "loss": 0.7035,
+      "step": 5351
+    },
+    {
+      "epoch": 0.45837615621788286,
+      "grad_norm": 0.5822468996047974,
+      "learning_rate": 0.0001,
+      "loss": 0.7814,
+      "step": 5352
+    },
+    {
+      "epoch": 0.45846180198698183,
+      "grad_norm": 0.8508754968643188,
+      "learning_rate": 0.0001,
+      "loss": 1.2331,
+      "step": 5353
+    },
+    {
+      "epoch": 0.45854744775608086,
+      "grad_norm": 0.9478009343147278,
+      "learning_rate": 0.0001,
+      "loss": 0.7676,
+      "step": 5354
+    },
+    {
+      "epoch": 0.45863309352517984,
+      "grad_norm": 0.64399254322052,
+      "learning_rate": 0.0001,
+      "loss": 0.6677,
+      "step": 5355
+    },
+    {
+      "epoch": 0.45871873929427887,
+      "grad_norm": 0.8066491484642029,
+      "learning_rate": 0.0001,
+      "loss": 0.8641,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4588043850633779,
+      "grad_norm": 0.7969967722892761,
+      "learning_rate": 0.0001,
+      "loss": 0.9787,
+      "step": 5357
+    },
+    {
+      "epoch": 0.45889003083247687,
+      "grad_norm": 0.77696692943573,
+      "learning_rate": 0.0001,
+      "loss": 0.4865,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4589756766015759,
+      "grad_norm": 0.493387907743454,
+      "learning_rate": 0.0001,
+      "loss": 0.8096,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4590613223706749,
+      "grad_norm": 0.9941874146461487,
+      "learning_rate": 0.0001,
+      "loss": 0.9496,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4591469681397739,
+      "grad_norm": 0.737909197807312,
+      "learning_rate": 0.0001,
+      "loss": 0.4824,
+      "step": 5361
+    },
+    {
+      "epoch": 0.4592326139088729,
+      "grad_norm": 0.8129358887672424,
+      "learning_rate": 0.0001,
+      "loss": 0.5246,
+      "step": 5362
+    },
+    {
+      "epoch": 0.4593182596779719,
+      "grad_norm": 0.6544095277786255,
+      "learning_rate": 0.0001,
+      "loss": 1.032,
+      "step": 5363
+    },
+    {
+      "epoch": 0.45940390544707094,
+      "grad_norm": 1.0910247564315796,
+      "learning_rate": 0.0001,
+      "loss": 0.9621,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4594895512161699,
+      "grad_norm": 0.28788867592811584,
+      "learning_rate": 0.0001,
+      "loss": 0.8394,
+      "step": 5365
+    },
+    {
+      "epoch": 0.45957519698526894,
+      "grad_norm": 0.7097026705741882,
+      "learning_rate": 0.0001,
+      "loss": 1.131,
+      "step": 5366
+    },
+    {
+      "epoch": 0.4596608427543679,
+      "grad_norm": 0.9650702476501465,
+      "learning_rate": 0.0001,
+      "loss": 1.1308,
+      "step": 5367
+    },
+    {
+      "epoch": 0.45974648852346695,
+      "grad_norm": 0.7887323498725891,
+      "learning_rate": 0.0001,
+      "loss": 0.8272,
+      "step": 5368
+    },
+    {
+      "epoch": 0.459832134292566,
+      "grad_norm": 0.7706044912338257,
+      "learning_rate": 0.0001,
+      "loss": 0.3793,
+      "step": 5369
+    },
+    {
+      "epoch": 0.45991778006166495,
+      "grad_norm": 1.1365162134170532,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 5370
+    },
+    {
+      "epoch": 0.460003425830764,
+      "grad_norm": 1.1501942873001099,
+      "learning_rate": 0.0001,
+      "loss": 1.022,
+      "step": 5371
+    },
+    {
+      "epoch": 0.46008907159986295,
+      "grad_norm": 0.5043150186538696,
+      "learning_rate": 0.0001,
+      "loss": 1.2628,
+      "step": 5372
+    },
+    {
+      "epoch": 0.460174717368962,
+      "grad_norm": 0.7335601449012756,
+      "learning_rate": 0.0001,
+      "loss": 0.5907,
+      "step": 5373
+    },
+    {
+      "epoch": 0.46026036313806096,
+      "grad_norm": 0.8210372924804688,
+      "learning_rate": 0.0001,
+      "loss": 0.9437,
+      "step": 5374
+    },
+    {
+      "epoch": 0.46034600890716,
+      "grad_norm": 0.9423237442970276,
+      "learning_rate": 0.0001,
+      "loss": 1.7927,
+      "step": 5375
+    },
+    {
+      "epoch": 0.460431654676259,
+      "grad_norm": 0.7448026537895203,
+      "learning_rate": 0.0001,
+      "loss": 0.5262,
+      "step": 5376
+    },
+    {
+      "epoch": 0.460517300445358,
+      "grad_norm": 1.2313847541809082,
+      "learning_rate": 0.0001,
+      "loss": 0.9,
+      "step": 5377
+    },
+    {
+      "epoch": 0.460602946214457,
+      "grad_norm": 1.2910646200180054,
+      "learning_rate": 0.0001,
+      "loss": 1.378,
+      "step": 5378
+    },
+    {
+      "epoch": 0.460688591983556,
+      "grad_norm": 0.645210862159729,
+      "learning_rate": 0.0001,
+      "loss": 1.0696,
+      "step": 5379
+    },
+    {
+      "epoch": 0.460774237752655,
+      "grad_norm": 0.44960442185401917,
+      "learning_rate": 0.0001,
+      "loss": 0.4275,
+      "step": 5380
+    },
+    {
+      "epoch": 0.460859883521754,
+      "grad_norm": 0.602227509021759,
+      "learning_rate": 0.0001,
+      "loss": 0.4195,
+      "step": 5381
+    },
+    {
+      "epoch": 0.46094552929085303,
+      "grad_norm": 0.7907888889312744,
+      "learning_rate": 0.0001,
+      "loss": 0.8731,
+      "step": 5382
+    },
+    {
+      "epoch": 0.46103117505995206,
+      "grad_norm": 0.6154989004135132,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 5383
+    },
+    {
+      "epoch": 0.46111682082905103,
+      "grad_norm": 0.7359213829040527,
+      "learning_rate": 0.0001,
+      "loss": 0.9436,
+      "step": 5384
+    },
+    {
+      "epoch": 0.46120246659815006,
+      "grad_norm": 0.5902386903762817,
+      "learning_rate": 0.0001,
+      "loss": 0.8883,
+      "step": 5385
+    },
+    {
+      "epoch": 0.46128811236724904,
+      "grad_norm": 1.1160893440246582,
+      "learning_rate": 0.0001,
+      "loss": 1.2578,
+      "step": 5386
+    },
+    {
+      "epoch": 0.46137375813634807,
+      "grad_norm": 0.6537150740623474,
+      "learning_rate": 0.0001,
+      "loss": 0.6855,
+      "step": 5387
+    },
+    {
+      "epoch": 0.4614594039054471,
+      "grad_norm": 0.5408684015274048,
+      "learning_rate": 0.0001,
+      "loss": 0.5702,
+      "step": 5388
+    },
+    {
+      "epoch": 0.46154504967454607,
+      "grad_norm": 0.5850176215171814,
+      "learning_rate": 0.0001,
+      "loss": 0.9929,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4616306954436451,
+      "grad_norm": 1.3646230697631836,
+      "learning_rate": 0.0001,
+      "loss": 2.0456,
+      "step": 5390
+    },
+    {
+      "epoch": 0.4617163412127441,
+      "grad_norm": 1.5131759643554688,
+      "learning_rate": 0.0001,
+      "loss": 0.955,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4618019869818431,
+      "grad_norm": 1.306979775428772,
+      "learning_rate": 0.0001,
+      "loss": 1.3328,
+      "step": 5392
+    },
+    {
+      "epoch": 0.4618876327509421,
+      "grad_norm": 0.7747806906700134,
+      "learning_rate": 0.0001,
+      "loss": 0.6815,
+      "step": 5393
+    },
+    {
+      "epoch": 0.4619732785200411,
+      "grad_norm": 0.6086723208427429,
+      "learning_rate": 0.0001,
+      "loss": 0.9145,
+      "step": 5394
+    },
+    {
+      "epoch": 0.46205892428914014,
+      "grad_norm": 0.5125333666801453,
+      "learning_rate": 0.0001,
+      "loss": 0.8983,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4621445700582391,
+      "grad_norm": 0.8700965046882629,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 5396
+    },
+    {
+      "epoch": 0.46223021582733814,
+      "grad_norm": 0.7331364154815674,
+      "learning_rate": 0.0001,
+      "loss": 1.1306,
+      "step": 5397
+    },
+    {
+      "epoch": 0.4623158615964371,
+      "grad_norm": 0.6777184009552002,
+      "learning_rate": 0.0001,
+      "loss": 0.9924,
+      "step": 5398
+    },
+    {
+      "epoch": 0.46240150736553615,
+      "grad_norm": 0.8091490864753723,
+      "learning_rate": 0.0001,
+      "loss": 0.7502,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4624871531346352,
+      "grad_norm": 0.7515975832939148,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 5400
+    },
+    {
+      "epoch": 0.46257279890373415,
+      "grad_norm": 0.49315565824508667,
+      "learning_rate": 0.0001,
+      "loss": 0.8138,
+      "step": 5401
+    },
+    {
+      "epoch": 0.4626584446728332,
+      "grad_norm": 0.7816513776779175,
+      "learning_rate": 0.0001,
+      "loss": 0.5586,
+      "step": 5402
+    },
+    {
+      "epoch": 0.46274409044193215,
+      "grad_norm": 1.2166188955307007,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4628297362110312,
+      "grad_norm": 0.7984596490859985,
+      "learning_rate": 0.0001,
+      "loss": 1.1741,
+      "step": 5404
+    },
+    {
+      "epoch": 0.46291538198013016,
+      "grad_norm": 1.2968701124191284,
+      "learning_rate": 0.0001,
+      "loss": 2.383,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4630010277492292,
+      "grad_norm": 0.8423162698745728,
+      "learning_rate": 0.0001,
+      "loss": 1.433,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4630866735183282,
+      "grad_norm": 0.9475583434104919,
+      "learning_rate": 0.0001,
+      "loss": 0.9779,
+      "step": 5407
+    },
+    {
+      "epoch": 0.4631723192874272,
+      "grad_norm": 1.019040584564209,
+      "learning_rate": 0.0001,
+      "loss": 1.1651,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4632579650565262,
+      "grad_norm": 0.9824590086936951,
+      "learning_rate": 0.0001,
+      "loss": 0.8904,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4633436108256252,
+      "grad_norm": 1.0805083513259888,
+      "learning_rate": 0.0001,
+      "loss": 1.1919,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4634292565947242,
+      "grad_norm": 0.5338455438613892,
+      "learning_rate": 0.0001,
+      "loss": 1.0792,
+      "step": 5411
+    },
+    {
+      "epoch": 0.4635149023638232,
+      "grad_norm": 0.9126402139663696,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 5412
+    },
+    {
+      "epoch": 0.46360054813292223,
+      "grad_norm": 0.37973496317863464,
+      "learning_rate": 0.0001,
+      "loss": 0.2973,
+      "step": 5413
+    },
+    {
+      "epoch": 0.46368619390202126,
+      "grad_norm": 0.7478768229484558,
+      "learning_rate": 0.0001,
+      "loss": 0.9853,
+      "step": 5414
+    },
+    {
+      "epoch": 0.46377183967112023,
+      "grad_norm": 1.3337688446044922,
+      "learning_rate": 0.0001,
+      "loss": 0.9818,
+      "step": 5415
+    },
+    {
+      "epoch": 0.46385748544021926,
+      "grad_norm": 0.7889738082885742,
+      "learning_rate": 0.0001,
+      "loss": 0.802,
+      "step": 5416
+    },
+    {
+      "epoch": 0.46394313120931824,
+      "grad_norm": 0.7688711285591125,
+      "learning_rate": 0.0001,
+      "loss": 0.8193,
+      "step": 5417
+    },
+    {
+      "epoch": 0.46402877697841727,
+      "grad_norm": 0.28557392954826355,
+      "learning_rate": 0.0001,
+      "loss": 0.1219,
+      "step": 5418
+    },
+    {
+      "epoch": 0.4641144227475163,
+      "grad_norm": 0.5092398524284363,
+      "learning_rate": 0.0001,
+      "loss": 0.3432,
+      "step": 5419
+    },
+    {
+      "epoch": 0.46420006851661527,
+      "grad_norm": 1.274568796157837,
+      "learning_rate": 0.0001,
+      "loss": 1.1157,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 0.7594764828681946,
+      "learning_rate": 0.0001,
+      "loss": 0.7536,
+      "step": 5421
+    },
+    {
+      "epoch": 0.4643713600548133,
+      "grad_norm": 0.8560181856155396,
+      "learning_rate": 0.0001,
+      "loss": 1.2197,
+      "step": 5422
+    },
+    {
+      "epoch": 0.4644570058239123,
+      "grad_norm": 1.136411428451538,
+      "learning_rate": 0.0001,
+      "loss": 0.7835,
+      "step": 5423
+    },
+    {
+      "epoch": 0.4645426515930113,
+      "grad_norm": 0.5110612511634827,
+      "learning_rate": 0.0001,
+      "loss": 0.803,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4646282973621103,
+      "grad_norm": 0.8704935908317566,
+      "learning_rate": 0.0001,
+      "loss": 1.0301,
+      "step": 5425
+    },
+    {
+      "epoch": 0.46471394313120934,
+      "grad_norm": 0.6429837942123413,
+      "learning_rate": 0.0001,
+      "loss": 0.928,
+      "step": 5426
+    },
+    {
+      "epoch": 0.4647995889003083,
+      "grad_norm": 0.3755650520324707,
+      "learning_rate": 0.0001,
+      "loss": 0.7632,
+      "step": 5427
+    },
+    {
+      "epoch": 0.46488523466940734,
+      "grad_norm": 0.5072048306465149,
+      "learning_rate": 0.0001,
+      "loss": 0.5122,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4649708804385063,
+      "grad_norm": 0.5230625867843628,
+      "learning_rate": 0.0001,
+      "loss": 0.9128,
+      "step": 5429
+    },
+    {
+      "epoch": 0.46505652620760535,
+      "grad_norm": 0.6222766637802124,
+      "learning_rate": 0.0001,
+      "loss": 0.3732,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4651421719767044,
+      "grad_norm": 1.5711126327514648,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 5431
+    },
+    {
+      "epoch": 0.46522781774580335,
+      "grad_norm": 0.8096694946289062,
+      "learning_rate": 0.0001,
+      "loss": 0.675,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4653134635149024,
+      "grad_norm": 1.0742621421813965,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 5433
+    },
+    {
+      "epoch": 0.46539910928400136,
+      "grad_norm": 0.9613361954689026,
+      "learning_rate": 0.0001,
+      "loss": 0.9373,
+      "step": 5434
+    },
+    {
+      "epoch": 0.4654847550531004,
+      "grad_norm": 0.8535910844802856,
+      "learning_rate": 0.0001,
+      "loss": 0.9896,
+      "step": 5435
+    },
+    {
+      "epoch": 0.46557040082219936,
+      "grad_norm": 0.7590482234954834,
+      "learning_rate": 0.0001,
+      "loss": 0.7386,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4656560465912984,
+      "grad_norm": 0.7000344395637512,
+      "learning_rate": 0.0001,
+      "loss": 0.9267,
+      "step": 5437
+    },
+    {
+      "epoch": 0.4657416923603974,
+      "grad_norm": 0.7021924257278442,
+      "learning_rate": 0.0001,
+      "loss": 1.1248,
+      "step": 5438
+    },
+    {
+      "epoch": 0.4658273381294964,
+      "grad_norm": 0.5073518753051758,
+      "learning_rate": 0.0001,
+      "loss": 1.2639,
+      "step": 5439
+    },
+    {
+      "epoch": 0.4659129838985954,
+      "grad_norm": 0.4574916362762451,
+      "learning_rate": 0.0001,
+      "loss": 0.5455,
+      "step": 5440
+    },
+    {
+      "epoch": 0.4659986296676944,
+      "grad_norm": 0.856311559677124,
+      "learning_rate": 0.0001,
+      "loss": 0.9557,
+      "step": 5441
+    },
+    {
+      "epoch": 0.4660842754367934,
+      "grad_norm": 0.6358059048652649,
+      "learning_rate": 0.0001,
+      "loss": 0.9836,
+      "step": 5442
+    },
+    {
+      "epoch": 0.4661699212058924,
+      "grad_norm": 1.212785243988037,
+      "learning_rate": 0.0001,
+      "loss": 0.8018,
+      "step": 5443
+    },
+    {
+      "epoch": 0.46625556697499143,
+      "grad_norm": 1.0824590921401978,
+      "learning_rate": 0.0001,
+      "loss": 1.1081,
+      "step": 5444
+    },
+    {
+      "epoch": 0.46634121274409046,
+      "grad_norm": 0.802953839302063,
+      "learning_rate": 0.0001,
+      "loss": 0.6609,
+      "step": 5445
+    },
+    {
+      "epoch": 0.46642685851318944,
+      "grad_norm": 1.3993231058120728,
+      "learning_rate": 0.0001,
+      "loss": 0.9776,
+      "step": 5446
+    },
+    {
+      "epoch": 0.46651250428228846,
+      "grad_norm": 0.6292349100112915,
+      "learning_rate": 0.0001,
+      "loss": 0.6468,
+      "step": 5447
+    },
+    {
+      "epoch": 0.46659815005138744,
+      "grad_norm": 0.5677114129066467,
+      "learning_rate": 0.0001,
+      "loss": 0.9872,
+      "step": 5448
+    },
+    {
+      "epoch": 0.46668379582048647,
+      "grad_norm": 0.9804188013076782,
+      "learning_rate": 0.0001,
+      "loss": 2.3351,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4667694415895855,
+      "grad_norm": 0.7721236944198608,
+      "learning_rate": 0.0001,
+      "loss": 0.818,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4668550873586845,
+      "grad_norm": 0.6802047491073608,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4669407331277835,
+      "grad_norm": 0.5128448605537415,
+      "learning_rate": 0.0001,
+      "loss": 0.5097,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4670263788968825,
+      "grad_norm": 0.6285668015480042,
+      "learning_rate": 0.0001,
+      "loss": 0.9419,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4671120246659815,
+      "grad_norm": 0.6128541231155396,
+      "learning_rate": 0.0001,
+      "loss": 0.632,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4671976704350805,
+      "grad_norm": 0.5676799416542053,
+      "learning_rate": 0.0001,
+      "loss": 0.8765,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4672833162041795,
+      "grad_norm": 0.8326805233955383,
+      "learning_rate": 0.0001,
+      "loss": 0.9205,
+      "step": 5456
+    },
+    {
+      "epoch": 0.46736896197327854,
+      "grad_norm": 0.7122310400009155,
+      "learning_rate": 0.0001,
+      "loss": 0.4932,
+      "step": 5457
+    },
+    {
+      "epoch": 0.4674546077423775,
+      "grad_norm": 0.35310614109039307,
+      "learning_rate": 0.0001,
+      "loss": 0.5721,
+      "step": 5458
+    },
+    {
+      "epoch": 0.46754025351147654,
+      "grad_norm": 1.2231868505477905,
+      "learning_rate": 0.0001,
+      "loss": 0.7595,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4676258992805755,
+      "grad_norm": 0.8476229906082153,
+      "learning_rate": 0.0001,
+      "loss": 0.7855,
+      "step": 5460
+    },
+    {
+      "epoch": 0.46771154504967455,
+      "grad_norm": 1.0238169431686401,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 5461
+    },
+    {
+      "epoch": 0.4677971908187736,
+      "grad_norm": 0.8419808149337769,
+      "learning_rate": 0.0001,
+      "loss": 0.6265,
+      "step": 5462
+    },
+    {
+      "epoch": 0.46788283658787255,
+      "grad_norm": 0.8018752932548523,
+      "learning_rate": 0.0001,
+      "loss": 0.7974,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4679684823569716,
+      "grad_norm": 1.2051020860671997,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 5464
+    },
+    {
+      "epoch": 0.46805412812607056,
+      "grad_norm": 0.5942273139953613,
+      "learning_rate": 0.0001,
+      "loss": 1.0684,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4681397738951696,
+      "grad_norm": 0.5791336297988892,
+      "learning_rate": 0.0001,
+      "loss": 1.2008,
+      "step": 5466
+    },
+    {
+      "epoch": 0.46822541966426856,
+      "grad_norm": 0.9270151257514954,
+      "learning_rate": 0.0001,
+      "loss": 0.9425,
+      "step": 5467
+    },
+    {
+      "epoch": 0.4683110654333676,
+      "grad_norm": 1.0576789379119873,
+      "learning_rate": 0.0001,
+      "loss": 1.1994,
+      "step": 5468
+    },
+    {
+      "epoch": 0.4683967112024666,
+      "grad_norm": 0.9623100757598877,
+      "learning_rate": 0.0001,
+      "loss": 1.0352,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4684823569715656,
+      "grad_norm": 0.8564298152923584,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4685680027406646,
+      "grad_norm": 0.9056994318962097,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 5471
+    },
+    {
+      "epoch": 0.4686536485097636,
+      "grad_norm": 0.7099137902259827,
+      "learning_rate": 0.0001,
+      "loss": 1.2074,
+      "step": 5472
+    },
+    {
+      "epoch": 0.46873929427886263,
+      "grad_norm": 0.5843431353569031,
+      "learning_rate": 0.0001,
+      "loss": 1.2737,
+      "step": 5473
+    },
+    {
+      "epoch": 0.46882494004796166,
+      "grad_norm": 0.7069448828697205,
+      "learning_rate": 0.0001,
+      "loss": 1.1911,
+      "step": 5474
+    },
+    {
+      "epoch": 0.46891058581706063,
+      "grad_norm": 1.1088424921035767,
+      "learning_rate": 0.0001,
+      "loss": 1.0504,
+      "step": 5475
+    },
+    {
+      "epoch": 0.46899623158615966,
+      "grad_norm": 0.8017354607582092,
+      "learning_rate": 0.0001,
+      "loss": 1.045,
+      "step": 5476
+    },
+    {
+      "epoch": 0.46908187735525864,
+      "grad_norm": 0.6458702087402344,
+      "learning_rate": 0.0001,
+      "loss": 0.9885,
+      "step": 5477
+    },
+    {
+      "epoch": 0.46916752312435767,
+      "grad_norm": 1.0039829015731812,
+      "learning_rate": 0.0001,
+      "loss": 1.0977,
+      "step": 5478
+    },
+    {
+      "epoch": 0.46925316889345664,
+      "grad_norm": 0.7875403165817261,
+      "learning_rate": 0.0001,
+      "loss": 1.1705,
+      "step": 5479
+    },
+    {
+      "epoch": 0.46933881466255567,
+      "grad_norm": 1.0828933715820312,
+      "learning_rate": 0.0001,
+      "loss": 1.3628,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4694244604316547,
+      "grad_norm": 1.25056791305542,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4695101062007537,
+      "grad_norm": 0.8471863865852356,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4695957519698527,
+      "grad_norm": 0.8063057065010071,
+      "learning_rate": 0.0001,
+      "loss": 0.777,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4696813977389517,
+      "grad_norm": 0.8483702540397644,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4697670435080507,
+      "grad_norm": 0.6532863974571228,
+      "learning_rate": 0.0001,
+      "loss": 1.2041,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4698526892771497,
+      "grad_norm": 0.7209187150001526,
+      "learning_rate": 0.0001,
+      "loss": 1.3018,
+      "step": 5486
+    },
+    {
+      "epoch": 0.4699383350462487,
+      "grad_norm": 0.8328231573104858,
+      "learning_rate": 0.0001,
+      "loss": 0.7258,
+      "step": 5487
+    },
+    {
+      "epoch": 0.47002398081534774,
+      "grad_norm": 1.637444019317627,
+      "learning_rate": 0.0001,
+      "loss": 1.1777,
+      "step": 5488
+    },
+    {
+      "epoch": 0.4701096265844467,
+      "grad_norm": 0.38962531089782715,
+      "learning_rate": 0.0001,
+      "loss": 0.3411,
+      "step": 5489
+    },
+    {
+      "epoch": 0.47019527235354575,
+      "grad_norm": 0.9280734658241272,
+      "learning_rate": 0.0001,
+      "loss": 0.7937,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4702809181226447,
+      "grad_norm": 0.7014694809913635,
+      "learning_rate": 0.0001,
+      "loss": 0.6228,
+      "step": 5491
+    },
+    {
+      "epoch": 0.47036656389174375,
+      "grad_norm": 0.6590593457221985,
+      "learning_rate": 0.0001,
+      "loss": 0.6609,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4704522096608428,
+      "grad_norm": 1.135981559753418,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 5493
+    },
+    {
+      "epoch": 0.47053785542994175,
+      "grad_norm": 0.9213424324989319,
+      "learning_rate": 0.0001,
+      "loss": 1.0326,
+      "step": 5494
+    },
+    {
+      "epoch": 0.4706235011990408,
+      "grad_norm": 0.6565827131271362,
+      "learning_rate": 0.0001,
+      "loss": 0.695,
+      "step": 5495
+    },
+    {
+      "epoch": 0.47070914696813976,
+      "grad_norm": 0.3243500888347626,
+      "learning_rate": 0.0001,
+      "loss": 0.2169,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4707947927372388,
+      "grad_norm": 0.808733344078064,
+      "learning_rate": 0.0001,
+      "loss": 0.7457,
+      "step": 5497
+    },
+    {
+      "epoch": 0.47088043850633776,
+      "grad_norm": 0.5352258086204529,
+      "learning_rate": 0.0001,
+      "loss": 0.766,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4709660842754368,
+      "grad_norm": 0.33549121022224426,
+      "learning_rate": 0.0001,
+      "loss": 0.3831,
+      "step": 5499
+    },
+    {
+      "epoch": 0.4710517300445358,
+      "grad_norm": 1.4426608085632324,
+      "learning_rate": 0.0001,
+      "loss": 1.2035,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4711373758136348,
+      "grad_norm": 0.8714738488197327,
+      "learning_rate": 0.0001,
+      "loss": 1.2549,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4712230215827338,
+      "grad_norm": 0.8760849237442017,
+      "learning_rate": 0.0001,
+      "loss": 0.6122,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4713086673518328,
+      "grad_norm": 0.6405802965164185,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 5503
+    },
+    {
+      "epoch": 0.47139431312093183,
+      "grad_norm": 0.7547061443328857,
+      "learning_rate": 0.0001,
+      "loss": 1.2471,
+      "step": 5504
+    },
+    {
+      "epoch": 0.47147995889003086,
+      "grad_norm": 1.0399761199951172,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 5505
+    },
+    {
+      "epoch": 0.47156560465912983,
+      "grad_norm": 1.0945743322372437,
+      "learning_rate": 0.0001,
+      "loss": 1.1012,
+      "step": 5506
+    },
+    {
+      "epoch": 0.47165125042822886,
+      "grad_norm": 0.7986792325973511,
+      "learning_rate": 0.0001,
+      "loss": 1.265,
+      "step": 5507
+    },
+    {
+      "epoch": 0.47173689619732784,
+      "grad_norm": 0.7990564703941345,
+      "learning_rate": 0.0001,
+      "loss": 0.9439,
+      "step": 5508
+    },
+    {
+      "epoch": 0.47182254196642687,
+      "grad_norm": 0.498226523399353,
+      "learning_rate": 0.0001,
+      "loss": 0.8655,
+      "step": 5509
+    },
+    {
+      "epoch": 0.47190818773552584,
+      "grad_norm": 0.780541718006134,
+      "learning_rate": 0.0001,
+      "loss": 0.8414,
+      "step": 5510
+    },
+    {
+      "epoch": 0.47199383350462487,
+      "grad_norm": 0.9144472479820251,
+      "learning_rate": 0.0001,
+      "loss": 1.2008,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4720794792737239,
+      "grad_norm": 0.612694263458252,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4721651250428229,
+      "grad_norm": 0.5463518500328064,
+      "learning_rate": 0.0001,
+      "loss": 0.5327,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4722507708119219,
+      "grad_norm": 0.7803296446800232,
+      "learning_rate": 0.0001,
+      "loss": 1.0784,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4723364165810209,
+      "grad_norm": 0.9960771203041077,
+      "learning_rate": 0.0001,
+      "loss": 0.9205,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4724220623501199,
+      "grad_norm": 0.876353919506073,
+      "learning_rate": 0.0001,
+      "loss": 2.0722,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4725077081192189,
+      "grad_norm": 0.9164867401123047,
+      "learning_rate": 0.0001,
+      "loss": 1.2973,
+      "step": 5517
+    },
+    {
+      "epoch": 0.4725933538883179,
+      "grad_norm": 0.616395115852356,
+      "learning_rate": 0.0001,
+      "loss": 1.0344,
+      "step": 5518
+    },
+    {
+      "epoch": 0.47267899965741694,
+      "grad_norm": 0.7066731452941895,
+      "learning_rate": 0.0001,
+      "loss": 0.4791,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4727646454265159,
+      "grad_norm": 0.7075924277305603,
+      "learning_rate": 0.0001,
+      "loss": 0.7046,
+      "step": 5520
+    },
+    {
+      "epoch": 0.47285029119561495,
+      "grad_norm": 1.4381016492843628,
+      "learning_rate": 0.0001,
+      "loss": 1.3667,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4729359369647139,
+      "grad_norm": 1.169844388961792,
+      "learning_rate": 0.0001,
+      "loss": 0.9516,
+      "step": 5522
+    },
+    {
+      "epoch": 0.47302158273381295,
+      "grad_norm": 0.1990370899438858,
+      "learning_rate": 0.0001,
+      "loss": 0.6222,
+      "step": 5523
+    },
+    {
+      "epoch": 0.473107228502912,
+      "grad_norm": 0.6350782513618469,
+      "learning_rate": 0.0001,
+      "loss": 0.5428,
+      "step": 5524
+    },
+    {
+      "epoch": 0.47319287427201095,
+      "grad_norm": 0.9012671113014221,
+      "learning_rate": 0.0001,
+      "loss": 2.0359,
+      "step": 5525
+    },
+    {
+      "epoch": 0.47327852004111,
+      "grad_norm": 0.6050360798835754,
+      "learning_rate": 0.0001,
+      "loss": 0.749,
+      "step": 5526
+    },
+    {
+      "epoch": 0.47336416581020896,
+      "grad_norm": 0.9158898591995239,
+      "learning_rate": 0.0001,
+      "loss": 1.1991,
+      "step": 5527
+    },
+    {
+      "epoch": 0.473449811579308,
+      "grad_norm": 0.7533619403839111,
+      "learning_rate": 0.0001,
+      "loss": 1.2628,
+      "step": 5528
+    },
+    {
+      "epoch": 0.47353545734840696,
+      "grad_norm": 0.5148915648460388,
+      "learning_rate": 0.0001,
+      "loss": 0.6063,
+      "step": 5529
+    },
+    {
+      "epoch": 0.473621103117506,
+      "grad_norm": 0.7934363484382629,
+      "learning_rate": 0.0001,
+      "loss": 1.3665,
+      "step": 5530
+    },
+    {
+      "epoch": 0.473706748886605,
+      "grad_norm": 1.1983412504196167,
+      "learning_rate": 0.0001,
+      "loss": 1.0386,
+      "step": 5531
+    },
+    {
+      "epoch": 0.473792394655704,
+      "grad_norm": 0.49752065539360046,
+      "learning_rate": 0.0001,
+      "loss": 0.6067,
+      "step": 5532
+    },
+    {
+      "epoch": 0.473878040424803,
+      "grad_norm": 0.9410013556480408,
+      "learning_rate": 0.0001,
+      "loss": 0.8213,
+      "step": 5533
+    },
+    {
+      "epoch": 0.473963686193902,
+      "grad_norm": 0.6040640473365784,
+      "learning_rate": 0.0001,
+      "loss": 0.9955,
+      "step": 5534
+    },
+    {
+      "epoch": 0.47404933196300103,
+      "grad_norm": 0.5479767918586731,
+      "learning_rate": 0.0001,
+      "loss": 1.0423,
+      "step": 5535
+    },
+    {
+      "epoch": 0.47413497773210006,
+      "grad_norm": 1.5133213996887207,
+      "learning_rate": 0.0001,
+      "loss": 1.3062,
+      "step": 5536
+    },
+    {
+      "epoch": 0.47422062350119903,
+      "grad_norm": 0.9495038390159607,
+      "learning_rate": 0.0001,
+      "loss": 1.0977,
+      "step": 5537
+    },
+    {
+      "epoch": 0.47430626927029806,
+      "grad_norm": 0.38028666377067566,
+      "learning_rate": 0.0001,
+      "loss": 0.614,
+      "step": 5538
+    },
+    {
+      "epoch": 0.47439191503939704,
+      "grad_norm": 0.6909717321395874,
+      "learning_rate": 0.0001,
+      "loss": 1.0528,
+      "step": 5539
+    },
+    {
+      "epoch": 0.47447756080849607,
+      "grad_norm": 0.9911840558052063,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 5540
+    },
+    {
+      "epoch": 0.47456320657759504,
+      "grad_norm": 0.8426598310470581,
+      "learning_rate": 0.0001,
+      "loss": 1.0987,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4746488523466941,
+      "grad_norm": 0.7198424339294434,
+      "learning_rate": 0.0001,
+      "loss": 0.2968,
+      "step": 5542
+    },
+    {
+      "epoch": 0.4747344981157931,
+      "grad_norm": 1.502017617225647,
+      "learning_rate": 0.0001,
+      "loss": 1.3064,
+      "step": 5543
+    },
+    {
+      "epoch": 0.4748201438848921,
+      "grad_norm": 0.9198643565177917,
+      "learning_rate": 0.0001,
+      "loss": 0.9065,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4749057896539911,
+      "grad_norm": 0.6979465484619141,
+      "learning_rate": 0.0001,
+      "loss": 0.9501,
+      "step": 5545
+    },
+    {
+      "epoch": 0.4749914354230901,
+      "grad_norm": 0.6610729694366455,
+      "learning_rate": 0.0001,
+      "loss": 1.1372,
+      "step": 5546
+    },
+    {
+      "epoch": 0.4750770811921891,
+      "grad_norm": 0.6517041325569153,
+      "learning_rate": 0.0001,
+      "loss": 0.9613,
+      "step": 5547
+    },
+    {
+      "epoch": 0.47516272696128814,
+      "grad_norm": 0.5684115290641785,
+      "learning_rate": 0.0001,
+      "loss": 1.0266,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4752483727303871,
+      "grad_norm": 0.6530154943466187,
+      "learning_rate": 0.0001,
+      "loss": 1.1258,
+      "step": 5549
+    },
+    {
+      "epoch": 0.47533401849948614,
+      "grad_norm": 0.7830435037612915,
+      "learning_rate": 0.0001,
+      "loss": 0.6843,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4754196642685851,
+      "grad_norm": 0.968303382396698,
+      "learning_rate": 0.0001,
+      "loss": 1.1055,
+      "step": 5551
+    },
+    {
+      "epoch": 0.47550531003768415,
+      "grad_norm": 0.5228111743927002,
+      "learning_rate": 0.0001,
+      "loss": 1.1721,
+      "step": 5552
+    },
+    {
+      "epoch": 0.4755909558067831,
+      "grad_norm": 0.5955407023429871,
+      "learning_rate": 0.0001,
+      "loss": 1.1255,
+      "step": 5553
+    },
+    {
+      "epoch": 0.47567660157588215,
+      "grad_norm": 0.6512063145637512,
+      "learning_rate": 0.0001,
+      "loss": 0.7661,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4757622473449812,
+      "grad_norm": 0.9550678730010986,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 5555
+    },
+    {
+      "epoch": 0.47584789311408016,
+      "grad_norm": 0.8326413035392761,
+      "learning_rate": 0.0001,
+      "loss": 0.742,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4759335388831792,
+      "grad_norm": 1.317162275314331,
+      "learning_rate": 0.0001,
+      "loss": 1.1422,
+      "step": 5557
+    },
+    {
+      "epoch": 0.47601918465227816,
+      "grad_norm": 0.6209298372268677,
+      "learning_rate": 0.0001,
+      "loss": 0.6604,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4761048304213772,
+      "grad_norm": 0.8988860249519348,
+      "learning_rate": 0.0001,
+      "loss": 1.122,
+      "step": 5559
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 2.548673629760742,
+      "learning_rate": 0.0001,
+      "loss": 1.0936,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4762761219595752,
+      "grad_norm": 0.5571726560592651,
+      "learning_rate": 0.0001,
+      "loss": 1.0306,
+      "step": 5561
+    },
+    {
+      "epoch": 0.4763617677286742,
+      "grad_norm": 1.6550867557525635,
+      "learning_rate": 0.0001,
+      "loss": 0.5822,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4764474134977732,
+      "grad_norm": 0.8612803220748901,
+      "learning_rate": 0.0001,
+      "loss": 1.1401,
+      "step": 5563
+    },
+    {
+      "epoch": 0.4765330592668722,
+      "grad_norm": 0.749382734298706,
+      "learning_rate": 0.0001,
+      "loss": 0.5411,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4766187050359712,
+      "grad_norm": 0.5716632604598999,
+      "learning_rate": 0.0001,
+      "loss": 1.0332,
+      "step": 5565
+    },
+    {
+      "epoch": 0.47670435080507023,
+      "grad_norm": 0.8735801577568054,
+      "learning_rate": 0.0001,
+      "loss": 0.6967,
+      "step": 5566
+    },
+    {
+      "epoch": 0.47678999657416926,
+      "grad_norm": 0.3602047860622406,
+      "learning_rate": 0.0001,
+      "loss": 0.49,
+      "step": 5567
+    },
+    {
+      "epoch": 0.47687564234326824,
+      "grad_norm": 0.7314115166664124,
+      "learning_rate": 0.0001,
+      "loss": 0.6081,
+      "step": 5568
+    },
+    {
+      "epoch": 0.47696128811236727,
+      "grad_norm": 0.5891883373260498,
+      "learning_rate": 0.0001,
+      "loss": 0.9563,
+      "step": 5569
+    },
+    {
+      "epoch": 0.47704693388146624,
+      "grad_norm": 1.1036527156829834,
+      "learning_rate": 0.0001,
+      "loss": 1.8235,
+      "step": 5570
+    },
+    {
+      "epoch": 0.47713257965056527,
+      "grad_norm": 0.8392348885536194,
+      "learning_rate": 0.0001,
+      "loss": 0.9666,
+      "step": 5571
+    },
+    {
+      "epoch": 0.47721822541966424,
+      "grad_norm": 0.5828944444656372,
+      "learning_rate": 0.0001,
+      "loss": 0.6346,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4773038711887633,
+      "grad_norm": 0.8014712929725647,
+      "learning_rate": 0.0001,
+      "loss": 0.8257,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4773895169578623,
+      "grad_norm": 0.24977387487888336,
+      "learning_rate": 0.0001,
+      "loss": 0.493,
+      "step": 5574
+    },
+    {
+      "epoch": 0.4774751627269613,
+      "grad_norm": 0.518244206905365,
+      "learning_rate": 0.0001,
+      "loss": 0.6625,
+      "step": 5575
+    },
+    {
+      "epoch": 0.4775608084960603,
+      "grad_norm": 0.737521767616272,
+      "learning_rate": 0.0001,
+      "loss": 0.7186,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4776464542651593,
+      "grad_norm": 1.3990963697433472,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4777321000342583,
+      "grad_norm": 0.6170826554298401,
+      "learning_rate": 0.0001,
+      "loss": 1.1951,
+      "step": 5578
+    },
+    {
+      "epoch": 0.47781774580335734,
+      "grad_norm": 0.5436701774597168,
+      "learning_rate": 0.0001,
+      "loss": 0.853,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4779033915724563,
+      "grad_norm": 0.5375548005104065,
+      "learning_rate": 0.0001,
+      "loss": 0.4,
+      "step": 5580
+    },
+    {
+      "epoch": 0.47798903734155535,
+      "grad_norm": 0.9266979098320007,
+      "learning_rate": 0.0001,
+      "loss": 1.7033,
+      "step": 5581
+    },
+    {
+      "epoch": 0.4780746831106543,
+      "grad_norm": 0.4226469397544861,
+      "learning_rate": 0.0001,
+      "loss": 0.6123,
+      "step": 5582
+    },
+    {
+      "epoch": 0.47816032887975335,
+      "grad_norm": 0.7027469277381897,
+      "learning_rate": 0.0001,
+      "loss": 0.597,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4782459746488523,
+      "grad_norm": 1.6554720401763916,
+      "learning_rate": 0.0001,
+      "loss": 1.1601,
+      "step": 5584
+    },
+    {
+      "epoch": 0.47833162041795135,
+      "grad_norm": 1.3223038911819458,
+      "learning_rate": 0.0001,
+      "loss": 1.1826,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4784172661870504,
+      "grad_norm": 0.9181719422340393,
+      "learning_rate": 0.0001,
+      "loss": 1.0321,
+      "step": 5586
+    },
+    {
+      "epoch": 0.47850291195614936,
+      "grad_norm": 0.6639939546585083,
+      "learning_rate": 0.0001,
+      "loss": 1.2272,
+      "step": 5587
+    },
+    {
+      "epoch": 0.4785885577252484,
+      "grad_norm": 0.8241331577301025,
+      "learning_rate": 0.0001,
+      "loss": 1.0159,
+      "step": 5588
+    },
+    {
+      "epoch": 0.47867420349434736,
+      "grad_norm": 0.9676554799079895,
+      "learning_rate": 0.0001,
+      "loss": 0.6691,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4787598492634464,
+      "grad_norm": 0.2952415943145752,
+      "learning_rate": 0.0001,
+      "loss": 0.5767,
+      "step": 5590
+    },
+    {
+      "epoch": 0.47884549503254537,
+      "grad_norm": 0.266501247882843,
+      "learning_rate": 0.0001,
+      "loss": 0.1152,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4789311408016444,
+      "grad_norm": 0.25507408380508423,
+      "learning_rate": 0.0001,
+      "loss": 0.4784,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4790167865707434,
+      "grad_norm": 0.6790238618850708,
+      "learning_rate": 0.0001,
+      "loss": 1.0751,
+      "step": 5593
+    },
+    {
+      "epoch": 0.4791024323398424,
+      "grad_norm": 0.7136505842208862,
+      "learning_rate": 0.0001,
+      "loss": 0.6971,
+      "step": 5594
+    },
+    {
+      "epoch": 0.47918807810894143,
+      "grad_norm": 0.6493948101997375,
+      "learning_rate": 0.0001,
+      "loss": 0.8418,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4792737238780404,
+      "grad_norm": 0.48331883549690247,
+      "learning_rate": 0.0001,
+      "loss": 0.7194,
+      "step": 5596
+    },
+    {
+      "epoch": 0.47935936964713943,
+      "grad_norm": 0.4128072261810303,
+      "learning_rate": 0.0001,
+      "loss": 0.4308,
+      "step": 5597
+    },
+    {
+      "epoch": 0.47944501541623846,
+      "grad_norm": 0.6158770322799683,
+      "learning_rate": 0.0001,
+      "loss": 0.89,
+      "step": 5598
+    },
+    {
+      "epoch": 0.47953066118533744,
+      "grad_norm": 0.9983496069908142,
+      "learning_rate": 0.0001,
+      "loss": 1.0329,
+      "step": 5599
+    },
+    {
+      "epoch": 0.47961630695443647,
+      "grad_norm": 0.724435567855835,
+      "learning_rate": 0.0001,
+      "loss": 0.9138,
+      "step": 5600
+    },
+    {
+      "epoch": 0.47970195272353544,
+      "grad_norm": 1.0018072128295898,
+      "learning_rate": 0.0001,
+      "loss": 0.7865,
+      "step": 5601
+    },
+    {
+      "epoch": 0.47978759849263447,
+      "grad_norm": 0.41748863458633423,
+      "learning_rate": 0.0001,
+      "loss": 0.5627,
+      "step": 5602
+    },
+    {
+      "epoch": 0.47987324426173344,
+      "grad_norm": 1.6466511487960815,
+      "learning_rate": 0.0001,
+      "loss": 1.2726,
+      "step": 5603
+    },
+    {
+      "epoch": 0.4799588900308325,
+      "grad_norm": 1.0288585424423218,
+      "learning_rate": 0.0001,
+      "loss": 0.9776,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4800445357999315,
+      "grad_norm": 0.6333812475204468,
+      "learning_rate": 0.0001,
+      "loss": 0.4709,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4801301815690305,
+      "grad_norm": 1.1505107879638672,
+      "learning_rate": 0.0001,
+      "loss": 1.957,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4802158273381295,
+      "grad_norm": 0.9128137230873108,
+      "learning_rate": 0.0001,
+      "loss": 0.9588,
+      "step": 5607
+    },
+    {
+      "epoch": 0.4803014731072285,
+      "grad_norm": 1.058381199836731,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4803871188763275,
+      "grad_norm": 1.1195893287658691,
+      "learning_rate": 0.0001,
+      "loss": 1.9752,
+      "step": 5609
+    },
+    {
+      "epoch": 0.48047276464542654,
+      "grad_norm": 0.9425456523895264,
+      "learning_rate": 0.0001,
+      "loss": 1.1469,
+      "step": 5610
+    },
+    {
+      "epoch": 0.4805584104145255,
+      "grad_norm": 0.8954184055328369,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 5611
+    },
+    {
+      "epoch": 0.48064405618362455,
+      "grad_norm": 0.9247467517852783,
+      "learning_rate": 0.0001,
+      "loss": 1.0072,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4807297019527235,
+      "grad_norm": 0.6497198939323425,
+      "learning_rate": 0.0001,
+      "loss": 0.8574,
+      "step": 5613
+    },
+    {
+      "epoch": 0.48081534772182255,
+      "grad_norm": 0.6011865139007568,
+      "learning_rate": 0.0001,
+      "loss": 0.6343,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4809009934909215,
+      "grad_norm": 1.0637507438659668,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 5615
+    },
+    {
+      "epoch": 0.48098663926002055,
+      "grad_norm": 0.8417260050773621,
+      "learning_rate": 0.0001,
+      "loss": 1.0712,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4810722850291196,
+      "grad_norm": 1.1285250186920166,
+      "learning_rate": 0.0001,
+      "loss": 0.7305,
+      "step": 5617
+    },
+    {
+      "epoch": 0.48115793079821856,
+      "grad_norm": 1.207551121711731,
+      "learning_rate": 0.0001,
+      "loss": 1.8532,
+      "step": 5618
+    },
+    {
+      "epoch": 0.4812435765673176,
+      "grad_norm": 0.6714890599250793,
+      "learning_rate": 0.0001,
+      "loss": 0.4455,
+      "step": 5619
+    },
+    {
+      "epoch": 0.48132922233641656,
+      "grad_norm": 0.9959033131599426,
+      "learning_rate": 0.0001,
+      "loss": 1.2275,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4814148681055156,
+      "grad_norm": 1.0146955251693726,
+      "learning_rate": 0.0001,
+      "loss": 1.356,
+      "step": 5621
+    },
+    {
+      "epoch": 0.48150051387461457,
+      "grad_norm": 0.8975377082824707,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 5622
+    },
+    {
+      "epoch": 0.4815861596437136,
+      "grad_norm": 0.8409011960029602,
+      "learning_rate": 0.0001,
+      "loss": 0.7228,
+      "step": 5623
+    },
+    {
+      "epoch": 0.4816718054128126,
+      "grad_norm": 0.8037174940109253,
+      "learning_rate": 0.0001,
+      "loss": 1.0037,
+      "step": 5624
+    },
+    {
+      "epoch": 0.4817574511819116,
+      "grad_norm": 0.4129940867424011,
+      "learning_rate": 0.0001,
+      "loss": 0.5613,
+      "step": 5625
+    },
+    {
+      "epoch": 0.48184309695101063,
+      "grad_norm": 0.5585822463035583,
+      "learning_rate": 0.0001,
+      "loss": 0.5065,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4819287427201096,
+      "grad_norm": 0.7155659794807434,
+      "learning_rate": 0.0001,
+      "loss": 1.1223,
+      "step": 5627
+    },
+    {
+      "epoch": 0.48201438848920863,
+      "grad_norm": 0.6426607370376587,
+      "learning_rate": 0.0001,
+      "loss": 0.3388,
+      "step": 5628
+    },
+    {
+      "epoch": 0.48210003425830766,
+      "grad_norm": 0.287300169467926,
+      "learning_rate": 0.0001,
+      "loss": 0.6451,
+      "step": 5629
+    },
+    {
+      "epoch": 0.48218568002740664,
+      "grad_norm": 1.3172732591629028,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 5630
+    },
+    {
+      "epoch": 0.48227132579650567,
+      "grad_norm": 0.7514084577560425,
+      "learning_rate": 0.0001,
+      "loss": 0.7021,
+      "step": 5631
+    },
+    {
+      "epoch": 0.48235697156560464,
+      "grad_norm": 0.5626547932624817,
+      "learning_rate": 0.0001,
+      "loss": 1.2867,
+      "step": 5632
+    },
+    {
+      "epoch": 0.48244261733470367,
+      "grad_norm": 0.6535382270812988,
+      "learning_rate": 0.0001,
+      "loss": 1.1938,
+      "step": 5633
+    },
+    {
+      "epoch": 0.48252826310380265,
+      "grad_norm": 0.43917426466941833,
+      "learning_rate": 0.0001,
+      "loss": 0.984,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4826139088729017,
+      "grad_norm": 0.6779139041900635,
+      "learning_rate": 0.0001,
+      "loss": 0.9017,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4826995546420007,
+      "grad_norm": 0.5424606800079346,
+      "learning_rate": 0.0001,
+      "loss": 0.4435,
+      "step": 5636
+    },
+    {
+      "epoch": 0.4827852004110997,
+      "grad_norm": 0.7156080603599548,
+      "learning_rate": 0.0001,
+      "loss": 0.9807,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4828708461801987,
+      "grad_norm": 0.6766181588172913,
+      "learning_rate": 0.0001,
+      "loss": 0.6131,
+      "step": 5638
+    },
+    {
+      "epoch": 0.4829564919492977,
+      "grad_norm": 1.026989221572876,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 5639
+    },
+    {
+      "epoch": 0.4830421377183967,
+      "grad_norm": 0.2988794445991516,
+      "learning_rate": 0.0001,
+      "loss": 0.4291,
+      "step": 5640
+    },
+    {
+      "epoch": 0.48312778348749574,
+      "grad_norm": 0.7486392259597778,
+      "learning_rate": 0.0001,
+      "loss": 1.0604,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4832134292565947,
+      "grad_norm": 0.9850327968597412,
+      "learning_rate": 0.0001,
+      "loss": 1.4223,
+      "step": 5642
+    },
+    {
+      "epoch": 0.48329907502569375,
+      "grad_norm": 1.4433577060699463,
+      "learning_rate": 0.0001,
+      "loss": 1.1431,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4833847207947927,
+      "grad_norm": 0.7421269416809082,
+      "learning_rate": 0.0001,
+      "loss": 0.6964,
+      "step": 5644
+    },
+    {
+      "epoch": 0.48347036656389175,
+      "grad_norm": 0.8776100277900696,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 5645
+    },
+    {
+      "epoch": 0.4835560123329907,
+      "grad_norm": 0.5927689671516418,
+      "learning_rate": 0.0001,
+      "loss": 1.0621,
+      "step": 5646
+    },
+    {
+      "epoch": 0.48364165810208976,
+      "grad_norm": 0.5920923948287964,
+      "learning_rate": 0.0001,
+      "loss": 1.1304,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4837273038711888,
+      "grad_norm": 0.542470395565033,
+      "learning_rate": 0.0001,
+      "loss": 0.4557,
+      "step": 5648
+    },
+    {
+      "epoch": 0.48381294964028776,
+      "grad_norm": 0.6021775007247925,
+      "learning_rate": 0.0001,
+      "loss": 0.9135,
+      "step": 5649
+    },
+    {
+      "epoch": 0.4838985954093868,
+      "grad_norm": 0.5582777857780457,
+      "learning_rate": 0.0001,
+      "loss": 1.0674,
+      "step": 5650
+    },
+    {
+      "epoch": 0.48398424117848576,
+      "grad_norm": 1.1918606758117676,
+      "learning_rate": 0.0001,
+      "loss": 1.0989,
+      "step": 5651
+    },
+    {
+      "epoch": 0.4840698869475848,
+      "grad_norm": 0.7769346237182617,
+      "learning_rate": 0.0001,
+      "loss": 1.0068,
+      "step": 5652
+    },
+    {
+      "epoch": 0.4841555327166838,
+      "grad_norm": 1.000518560409546,
+      "learning_rate": 0.0001,
+      "loss": 0.9553,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4842411784857828,
+      "grad_norm": 0.5920243263244629,
+      "learning_rate": 0.0001,
+      "loss": 0.4371,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4843268242548818,
+      "grad_norm": 0.8383395671844482,
+      "learning_rate": 0.0001,
+      "loss": 0.8968,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4844124700239808,
+      "grad_norm": 0.5983134508132935,
+      "learning_rate": 0.0001,
+      "loss": 0.9691,
+      "step": 5656
+    },
+    {
+      "epoch": 0.48449811579307983,
+      "grad_norm": 0.8890658617019653,
+      "learning_rate": 0.0001,
+      "loss": 1.1365,
+      "step": 5657
+    },
+    {
+      "epoch": 0.4845837615621788,
+      "grad_norm": 0.8623133897781372,
+      "learning_rate": 0.0001,
+      "loss": 0.8017,
+      "step": 5658
+    },
+    {
+      "epoch": 0.48466940733127784,
+      "grad_norm": 0.8688929677009583,
+      "learning_rate": 0.0001,
+      "loss": 0.7764,
+      "step": 5659
+    },
+    {
+      "epoch": 0.48475505310037686,
+      "grad_norm": 1.1476109027862549,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 5660
+    },
+    {
+      "epoch": 0.48484069886947584,
+      "grad_norm": 1.0181968212127686,
+      "learning_rate": 0.0001,
+      "loss": 0.9098,
+      "step": 5661
+    },
+    {
+      "epoch": 0.48492634463857487,
+      "grad_norm": 0.7198486924171448,
+      "learning_rate": 0.0001,
+      "loss": 1.0517,
+      "step": 5662
+    },
+    {
+      "epoch": 0.48501199040767384,
+      "grad_norm": 0.7406960725784302,
+      "learning_rate": 0.0001,
+      "loss": 1.1536,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4850976361767729,
+      "grad_norm": 0.596300482749939,
+      "learning_rate": 0.0001,
+      "loss": 0.4448,
+      "step": 5664
+    },
+    {
+      "epoch": 0.48518328194587185,
+      "grad_norm": 0.5348647832870483,
+      "learning_rate": 0.0001,
+      "loss": 0.5389,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4852689277149709,
+      "grad_norm": 1.0215376615524292,
+      "learning_rate": 0.0001,
+      "loss": 0.8348,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4853545734840699,
+      "grad_norm": 1.0488349199295044,
+      "learning_rate": 0.0001,
+      "loss": 1.055,
+      "step": 5667
+    },
+    {
+      "epoch": 0.4854402192531689,
+      "grad_norm": 1.155617356300354,
+      "learning_rate": 0.0001,
+      "loss": 1.1863,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4855258650222679,
+      "grad_norm": 0.3982382118701935,
+      "learning_rate": 0.0001,
+      "loss": 0.4974,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4856115107913669,
+      "grad_norm": 0.5811617970466614,
+      "learning_rate": 0.0001,
+      "loss": 0.9293,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4856971565604659,
+      "grad_norm": 0.7270127534866333,
+      "learning_rate": 0.0001,
+      "loss": 0.5512,
+      "step": 5671
+    },
+    {
+      "epoch": 0.48578280232956494,
+      "grad_norm": 0.3857720196247101,
+      "learning_rate": 0.0001,
+      "loss": 0.8215,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4858684480986639,
+      "grad_norm": 1.1889805793762207,
+      "learning_rate": 0.0001,
+      "loss": 1.2172,
+      "step": 5673
+    },
+    {
+      "epoch": 0.48595409386776295,
+      "grad_norm": 0.5107437372207642,
+      "learning_rate": 0.0001,
+      "loss": 0.9306,
+      "step": 5674
+    },
+    {
+      "epoch": 0.4860397396368619,
+      "grad_norm": 0.8536267280578613,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 5675
+    },
+    {
+      "epoch": 0.48612538540596095,
+      "grad_norm": 0.7672193050384521,
+      "learning_rate": 0.0001,
+      "loss": 0.7441,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4862110311750599,
+      "grad_norm": 0.6489520072937012,
+      "learning_rate": 0.0001,
+      "loss": 1.2065,
+      "step": 5677
+    },
+    {
+      "epoch": 0.48629667694415896,
+      "grad_norm": 0.8134575486183167,
+      "learning_rate": 0.0001,
+      "loss": 0.7209,
+      "step": 5678
+    },
+    {
+      "epoch": 0.486382322713258,
+      "grad_norm": 0.6673193573951721,
+      "learning_rate": 0.0001,
+      "loss": 0.9266,
+      "step": 5679
+    },
+    {
+      "epoch": 0.48646796848235696,
+      "grad_norm": 0.90538090467453,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 5680
+    },
+    {
+      "epoch": 0.486553614251456,
+      "grad_norm": 0.5864711403846741,
+      "learning_rate": 0.0001,
+      "loss": 0.9798,
+      "step": 5681
+    },
+    {
+      "epoch": 0.48663926002055496,
+      "grad_norm": 1.0190099477767944,
+      "learning_rate": 0.0001,
+      "loss": 1.112,
+      "step": 5682
+    },
+    {
+      "epoch": 0.486724905789654,
+      "grad_norm": 0.5046374201774597,
+      "learning_rate": 0.0001,
+      "loss": 0.2625,
+      "step": 5683
+    },
+    {
+      "epoch": 0.486810551558753,
+      "grad_norm": 0.8188319206237793,
+      "learning_rate": 0.0001,
+      "loss": 1.0201,
+      "step": 5684
+    },
+    {
+      "epoch": 0.486896197327852,
+      "grad_norm": 0.8234528303146362,
+      "learning_rate": 0.0001,
+      "loss": 0.6444,
+      "step": 5685
+    },
+    {
+      "epoch": 0.48698184309695103,
+      "grad_norm": 0.9835221767425537,
+      "learning_rate": 0.0001,
+      "loss": 1.1708,
+      "step": 5686
+    },
+    {
+      "epoch": 0.48706748886605,
+      "grad_norm": 0.5845721364021301,
+      "learning_rate": 0.0001,
+      "loss": 0.8637,
+      "step": 5687
+    },
+    {
+      "epoch": 0.48715313463514903,
+      "grad_norm": 1.6668882369995117,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 5688
+    },
+    {
+      "epoch": 0.487238780404248,
+      "grad_norm": 1.0766241550445557,
+      "learning_rate": 0.0001,
+      "loss": 1.0986,
+      "step": 5689
+    },
+    {
+      "epoch": 0.48732442617334704,
+      "grad_norm": 0.7560464143753052,
+      "learning_rate": 0.0001,
+      "loss": 0.6548,
+      "step": 5690
+    },
+    {
+      "epoch": 0.48741007194244607,
+      "grad_norm": 0.5991166234016418,
+      "learning_rate": 0.0001,
+      "loss": 0.6914,
+      "step": 5691
+    },
+    {
+      "epoch": 0.48749571771154504,
+      "grad_norm": 0.6639272570610046,
+      "learning_rate": 0.0001,
+      "loss": 0.6735,
+      "step": 5692
+    },
+    {
+      "epoch": 0.48758136348064407,
+      "grad_norm": 0.7735525369644165,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 5693
+    },
+    {
+      "epoch": 0.48766700924974304,
+      "grad_norm": 1.0284606218338013,
+      "learning_rate": 0.0001,
+      "loss": 1.1598,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4877526550188421,
+      "grad_norm": 0.9737710356712341,
+      "learning_rate": 0.0001,
+      "loss": 0.8837,
+      "step": 5695
+    },
+    {
+      "epoch": 0.48783830078794105,
+      "grad_norm": 1.0970410108566284,
+      "learning_rate": 0.0001,
+      "loss": 1.1604,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4879239465570401,
+      "grad_norm": 0.6569715142250061,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 5697
+    },
+    {
+      "epoch": 0.4880095923261391,
+      "grad_norm": 0.4801821708679199,
+      "learning_rate": 0.0001,
+      "loss": 0.6545,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4880952380952381,
+      "grad_norm": 0.7176531553268433,
+      "learning_rate": 0.0001,
+      "loss": 0.5888,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4881808838643371,
+      "grad_norm": 0.6023481488227844,
+      "learning_rate": 0.0001,
+      "loss": 1.1109,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4882665296334361,
+      "grad_norm": 0.7917184829711914,
+      "learning_rate": 0.0001,
+      "loss": 1.3526,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4883521754025351,
+      "grad_norm": 0.5151416063308716,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 5702
+    },
+    {
+      "epoch": 0.48843782117163415,
+      "grad_norm": 0.7499685883522034,
+      "learning_rate": 0.0001,
+      "loss": 1.0542,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4885234669407331,
+      "grad_norm": 0.9750015139579773,
+      "learning_rate": 0.0001,
+      "loss": 0.805,
+      "step": 5704
+    },
+    {
+      "epoch": 0.48860911270983215,
+      "grad_norm": 0.5351232886314392,
+      "learning_rate": 0.0001,
+      "loss": 0.8849,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4886947584789311,
+      "grad_norm": 1.229606032371521,
+      "learning_rate": 0.0001,
+      "loss": 0.7273,
+      "step": 5706
+    },
+    {
+      "epoch": 0.48878040424803015,
+      "grad_norm": 0.7012339234352112,
+      "learning_rate": 0.0001,
+      "loss": 1.104,
+      "step": 5707
+    },
+    {
+      "epoch": 0.48886605001712913,
+      "grad_norm": 1.0103918313980103,
+      "learning_rate": 0.0001,
+      "loss": 1.2246,
+      "step": 5708
+    },
+    {
+      "epoch": 0.48895169578622816,
+      "grad_norm": 1.0886043310165405,
+      "learning_rate": 0.0001,
+      "loss": 0.8718,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4890373415553272,
+      "grad_norm": 1.0608543157577515,
+      "learning_rate": 0.0001,
+      "loss": 1.1511,
+      "step": 5710
+    },
+    {
+      "epoch": 0.48912298732442616,
+      "grad_norm": 0.5774617791175842,
+      "learning_rate": 0.0001,
+      "loss": 0.9073,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4892086330935252,
+      "grad_norm": 0.8840686082839966,
+      "learning_rate": 0.0001,
+      "loss": 1.0253,
+      "step": 5712
+    },
+    {
+      "epoch": 0.48929427886262417,
+      "grad_norm": 0.5464500188827515,
+      "learning_rate": 0.0001,
+      "loss": 1.1315,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4893799246317232,
+      "grad_norm": 0.6254547238349915,
+      "learning_rate": 0.0001,
+      "loss": 0.996,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4894655704008222,
+      "grad_norm": 0.5251661539077759,
+      "learning_rate": 0.0001,
+      "loss": 0.3528,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4895512161699212,
+      "grad_norm": 1.215438961982727,
+      "learning_rate": 0.0001,
+      "loss": 1.1956,
+      "step": 5716
+    },
+    {
+      "epoch": 0.48963686193902023,
+      "grad_norm": 0.6098703145980835,
+      "learning_rate": 0.0001,
+      "loss": 0.6189,
+      "step": 5717
+    },
+    {
+      "epoch": 0.4897225077081192,
+      "grad_norm": 0.3834235966205597,
+      "learning_rate": 0.0001,
+      "loss": 0.5876,
+      "step": 5718
+    },
+    {
+      "epoch": 0.48980815347721823,
+      "grad_norm": 0.8686200380325317,
+      "learning_rate": 0.0001,
+      "loss": 1.2219,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4898937992463172,
+      "grad_norm": 1.1383130550384521,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 5720
+    },
+    {
+      "epoch": 0.48997944501541624,
+      "grad_norm": 0.788934051990509,
+      "learning_rate": 0.0001,
+      "loss": 0.8057,
+      "step": 5721
+    },
+    {
+      "epoch": 0.49006509078451527,
+      "grad_norm": 0.6541995406150818,
+      "learning_rate": 0.0001,
+      "loss": 1.1523,
+      "step": 5722
+    },
+    {
+      "epoch": 0.49015073655361424,
+      "grad_norm": 1.0248087644577026,
+      "learning_rate": 0.0001,
+      "loss": 0.879,
+      "step": 5723
+    },
+    {
+      "epoch": 0.49023638232271327,
+      "grad_norm": 0.8418538570404053,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 5724
+    },
+    {
+      "epoch": 0.49032202809181225,
+      "grad_norm": 0.6474932432174683,
+      "learning_rate": 0.0001,
+      "loss": 0.4445,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4904076738609113,
+      "grad_norm": 0.6242976784706116,
+      "learning_rate": 0.0001,
+      "loss": 0.6814,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4904933196300103,
+      "grad_norm": 1.0238099098205566,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 5727
+    },
+    {
+      "epoch": 0.4905789653991093,
+      "grad_norm": 0.6541022062301636,
+      "learning_rate": 0.0001,
+      "loss": 1.2658,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4906646111682083,
+      "grad_norm": 1.559415578842163,
+      "learning_rate": 0.0001,
+      "loss": 0.5871,
+      "step": 5729
+    },
+    {
+      "epoch": 0.4907502569373073,
+      "grad_norm": 0.665458619594574,
+      "learning_rate": 0.0001,
+      "loss": 0.9803,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4908359027064063,
+      "grad_norm": 0.8236952424049377,
+      "learning_rate": 0.0001,
+      "loss": 0.7733,
+      "step": 5731
+    },
+    {
+      "epoch": 0.4909215484755053,
+      "grad_norm": 0.5140729546546936,
+      "learning_rate": 0.0001,
+      "loss": 0.6976,
+      "step": 5732
+    },
+    {
+      "epoch": 0.4910071942446043,
+      "grad_norm": 0.6456547379493713,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 5733
+    },
+    {
+      "epoch": 0.49109284001370335,
+      "grad_norm": 0.9146756529808044,
+      "learning_rate": 0.0001,
+      "loss": 1.3801,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4911784857828023,
+      "grad_norm": 0.606023371219635,
+      "learning_rate": 0.0001,
+      "loss": 0.8851,
+      "step": 5735
+    },
+    {
+      "epoch": 0.49126413155190135,
+      "grad_norm": 0.8737720251083374,
+      "learning_rate": 0.0001,
+      "loss": 1.1525,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4913497773210003,
+      "grad_norm": 0.7685626149177551,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 5737
+    },
+    {
+      "epoch": 0.49143542309009935,
+      "grad_norm": 0.8794716596603394,
+      "learning_rate": 0.0001,
+      "loss": 0.8137,
+      "step": 5738
+    },
+    {
+      "epoch": 0.49152106885919833,
+      "grad_norm": 0.5516017079353333,
+      "learning_rate": 0.0001,
+      "loss": 1.1842,
+      "step": 5739
+    },
+    {
+      "epoch": 0.49160671462829736,
+      "grad_norm": 0.9533675312995911,
+      "learning_rate": 0.0001,
+      "loss": 1.0107,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4916923603973964,
+      "grad_norm": 0.7284535765647888,
+      "learning_rate": 0.0001,
+      "loss": 0.7563,
+      "step": 5741
+    },
+    {
+      "epoch": 0.49177800616649536,
+      "grad_norm": 0.39138349890708923,
+      "learning_rate": 0.0001,
+      "loss": 0.5442,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4918636519355944,
+      "grad_norm": 0.9420462250709534,
+      "learning_rate": 0.0001,
+      "loss": 0.7336,
+      "step": 5743
+    },
+    {
+      "epoch": 0.49194929770469337,
+      "grad_norm": 0.9571136832237244,
+      "learning_rate": 0.0001,
+      "loss": 1.1865,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4920349434737924,
+      "grad_norm": 0.7962754964828491,
+      "learning_rate": 0.0001,
+      "loss": 1.1962,
+      "step": 5745
+    },
+    {
+      "epoch": 0.4921205892428914,
+      "grad_norm": 0.9117836952209473,
+      "learning_rate": 0.0001,
+      "loss": 0.9908,
+      "step": 5746
+    },
+    {
+      "epoch": 0.4922062350119904,
+      "grad_norm": 0.7800178527832031,
+      "learning_rate": 0.0001,
+      "loss": 0.8039,
+      "step": 5747
+    },
+    {
+      "epoch": 0.49229188078108943,
+      "grad_norm": 1.0994230508804321,
+      "learning_rate": 0.0001,
+      "loss": 1.2403,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4923775265501884,
+      "grad_norm": 0.6270629167556763,
+      "learning_rate": 0.0001,
+      "loss": 0.553,
+      "step": 5749
+    },
+    {
+      "epoch": 0.49246317231928743,
+      "grad_norm": 0.47766104340553284,
+      "learning_rate": 0.0001,
+      "loss": 0.835,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4925488180883864,
+      "grad_norm": 1.2992877960205078,
+      "learning_rate": 0.0001,
+      "loss": 1.3043,
+      "step": 5751
+    },
+    {
+      "epoch": 0.49263446385748544,
+      "grad_norm": 0.6541818976402283,
+      "learning_rate": 0.0001,
+      "loss": 0.7351,
+      "step": 5752
+    },
+    {
+      "epoch": 0.49272010962658447,
+      "grad_norm": 0.8513002991676331,
+      "learning_rate": 0.0001,
+      "loss": 0.8358,
+      "step": 5753
+    },
+    {
+      "epoch": 0.49280575539568344,
+      "grad_norm": 0.5467233061790466,
+      "learning_rate": 0.0001,
+      "loss": 0.5696,
+      "step": 5754
+    },
+    {
+      "epoch": 0.49289140116478247,
+      "grad_norm": 1.001258134841919,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 5755
+    },
+    {
+      "epoch": 0.49297704693388145,
+      "grad_norm": 0.9054346680641174,
+      "learning_rate": 0.0001,
+      "loss": 1.1288,
+      "step": 5756
+    },
+    {
+      "epoch": 0.4930626927029805,
+      "grad_norm": 0.48785045742988586,
+      "learning_rate": 0.0001,
+      "loss": 0.8895,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4931483384720795,
+      "grad_norm": 1.1307014226913452,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4932339842411785,
+      "grad_norm": 0.7377840876579285,
+      "learning_rate": 0.0001,
+      "loss": 0.7505,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4933196300102775,
+      "grad_norm": 0.3530736267566681,
+      "learning_rate": 0.0001,
+      "loss": 0.2244,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4934052757793765,
+      "grad_norm": 0.7069528698921204,
+      "learning_rate": 0.0001,
+      "loss": 1.2794,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4934909215484755,
+      "grad_norm": 1.2965278625488281,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4935765673175745,
+      "grad_norm": 0.9232261776924133,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4936622130866735,
+      "grad_norm": 0.6841115355491638,
+      "learning_rate": 0.0001,
+      "loss": 1.1517,
+      "step": 5764
+    },
+    {
+      "epoch": 0.49374785885577255,
+      "grad_norm": 0.6944504380226135,
+      "learning_rate": 0.0001,
+      "loss": 0.7316,
+      "step": 5765
+    },
+    {
+      "epoch": 0.4938335046248715,
+      "grad_norm": 0.7975636124610901,
+      "learning_rate": 0.0001,
+      "loss": 1.1095,
+      "step": 5766
+    },
+    {
+      "epoch": 0.49391915039397055,
+      "grad_norm": 0.6086589097976685,
+      "learning_rate": 0.0001,
+      "loss": 1.1263,
+      "step": 5767
+    },
+    {
+      "epoch": 0.4940047961630695,
+      "grad_norm": 0.41647976636886597,
+      "learning_rate": 0.0001,
+      "loss": 0.8668,
+      "step": 5768
+    },
+    {
+      "epoch": 0.49409044193216856,
+      "grad_norm": 1.3860942125320435,
+      "learning_rate": 0.0001,
+      "loss": 1.0161,
+      "step": 5769
+    },
+    {
+      "epoch": 0.49417608770126753,
+      "grad_norm": 0.49466845393180847,
+      "learning_rate": 0.0001,
+      "loss": 1.1282,
+      "step": 5770
+    },
+    {
+      "epoch": 0.49426173347036656,
+      "grad_norm": 1.1412264108657837,
+      "learning_rate": 0.0001,
+      "loss": 1.0781,
+      "step": 5771
+    },
+    {
+      "epoch": 0.4943473792394656,
+      "grad_norm": 0.5446130037307739,
+      "learning_rate": 0.0001,
+      "loss": 1.0421,
+      "step": 5772
+    },
+    {
+      "epoch": 0.49443302500856456,
+      "grad_norm": 0.8636963963508606,
+      "learning_rate": 0.0001,
+      "loss": 0.7941,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4945186707776636,
+      "grad_norm": 0.6271157264709473,
+      "learning_rate": 0.0001,
+      "loss": 0.9598,
+      "step": 5774
+    },
+    {
+      "epoch": 0.49460431654676257,
+      "grad_norm": 0.3194127380847931,
+      "learning_rate": 0.0001,
+      "loss": 0.2824,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4946899623158616,
+      "grad_norm": 0.8788530230522156,
+      "learning_rate": 0.0001,
+      "loss": 0.5764,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4947756080849606,
+      "grad_norm": 0.5982626080513,
+      "learning_rate": 0.0001,
+      "loss": 1.055,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4948612538540596,
+      "grad_norm": 1.0793815851211548,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 5778
+    },
+    {
+      "epoch": 0.49494689962315863,
+      "grad_norm": 0.8388663530349731,
+      "learning_rate": 0.0001,
+      "loss": 0.9807,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4950325453922576,
+      "grad_norm": 0.7949279546737671,
+      "learning_rate": 0.0001,
+      "loss": 0.6465,
+      "step": 5780
+    },
+    {
+      "epoch": 0.49511819116135664,
+      "grad_norm": 0.6162511110305786,
+      "learning_rate": 0.0001,
+      "loss": 0.9968,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4952038369304556,
+      "grad_norm": 0.6025516986846924,
+      "learning_rate": 0.0001,
+      "loss": 1.0275,
+      "step": 5782
+    },
+    {
+      "epoch": 0.49528948269955464,
+      "grad_norm": 0.8944854736328125,
+      "learning_rate": 0.0001,
+      "loss": 1.1213,
+      "step": 5783
+    },
+    {
+      "epoch": 0.49537512846865367,
+      "grad_norm": 0.8817099332809448,
+      "learning_rate": 0.0001,
+      "loss": 0.8164,
+      "step": 5784
+    },
+    {
+      "epoch": 0.49546077423775264,
+      "grad_norm": 0.9183173775672913,
+      "learning_rate": 0.0001,
+      "loss": 1.0353,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4955464200068517,
+      "grad_norm": 1.411128282546997,
+      "learning_rate": 0.0001,
+      "loss": 1.0968,
+      "step": 5786
+    },
+    {
+      "epoch": 0.49563206577595065,
+      "grad_norm": 0.6617475152015686,
+      "learning_rate": 0.0001,
+      "loss": 1.1181,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4957177115450497,
+      "grad_norm": 0.8116161823272705,
+      "learning_rate": 0.0001,
+      "loss": 1.3231,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4958033573141487,
+      "grad_norm": 0.8673804402351379,
+      "learning_rate": 0.0001,
+      "loss": 0.9857,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4958890030832477,
+      "grad_norm": 0.7433578968048096,
+      "learning_rate": 0.0001,
+      "loss": 0.6929,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4959746488523467,
+      "grad_norm": 0.7977319359779358,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4960602946214457,
+      "grad_norm": 0.8550966382026672,
+      "learning_rate": 0.0001,
+      "loss": 0.8877,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4961459403905447,
+      "grad_norm": 0.6295143365859985,
+      "learning_rate": 0.0001,
+      "loss": 1.0268,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4962315861596437,
+      "grad_norm": 0.880431056022644,
+      "learning_rate": 0.0001,
+      "loss": 0.7415,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4963172319287427,
+      "grad_norm": 0.8900250792503357,
+      "learning_rate": 0.0001,
+      "loss": 0.815,
+      "step": 5795
+    },
+    {
+      "epoch": 0.49640287769784175,
+      "grad_norm": 0.9012637138366699,
+      "learning_rate": 0.0001,
+      "loss": 1.2217,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4964885234669407,
+      "grad_norm": 0.8173678517341614,
+      "learning_rate": 0.0001,
+      "loss": 0.992,
+      "step": 5797
+    },
+    {
+      "epoch": 0.49657416923603975,
+      "grad_norm": 0.5138055682182312,
+      "learning_rate": 0.0001,
+      "loss": 0.7199,
+      "step": 5798
+    },
+    {
+      "epoch": 0.4966598150051387,
+      "grad_norm": 0.546964704990387,
+      "learning_rate": 0.0001,
+      "loss": 0.7792,
+      "step": 5799
+    },
+    {
+      "epoch": 0.49674546077423776,
+      "grad_norm": 0.2630160450935364,
+      "learning_rate": 0.0001,
+      "loss": 0.2512,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4968311065433368,
+      "grad_norm": 0.6691546440124512,
+      "learning_rate": 0.0001,
+      "loss": 1.1981,
+      "step": 5801
+    },
+    {
+      "epoch": 0.49691675231243576,
+      "grad_norm": 0.6658236980438232,
+      "learning_rate": 0.0001,
+      "loss": 1.0437,
+      "step": 5802
+    },
+    {
+      "epoch": 0.4970023980815348,
+      "grad_norm": 1.2132118940353394,
+      "learning_rate": 0.0001,
+      "loss": 1.1783,
+      "step": 5803
+    },
+    {
+      "epoch": 0.49708804385063377,
+      "grad_norm": 0.9886226058006287,
+      "learning_rate": 0.0001,
+      "loss": 1.3179,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4971736896197328,
+      "grad_norm": 0.893600583076477,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 5805
+    },
+    {
+      "epoch": 0.49725933538883177,
+      "grad_norm": 0.9973822236061096,
+      "learning_rate": 0.0001,
+      "loss": 1.0581,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4973449811579308,
+      "grad_norm": 0.6645148396492004,
+      "learning_rate": 0.0001,
+      "loss": 1.0364,
+      "step": 5807
+    },
+    {
+      "epoch": 0.49743062692702983,
+      "grad_norm": 0.6677850484848022,
+      "learning_rate": 0.0001,
+      "loss": 1.0915,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4975162726961288,
+      "grad_norm": 1.5003076791763306,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 5809
+    },
+    {
+      "epoch": 0.49760191846522783,
+      "grad_norm": 0.6976088285446167,
+      "learning_rate": 0.0001,
+      "loss": 0.9712,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4976875642343268,
+      "grad_norm": 0.6094777584075928,
+      "learning_rate": 0.0001,
+      "loss": 1.1787,
+      "step": 5811
+    },
+    {
+      "epoch": 0.49777321000342584,
+      "grad_norm": 0.751645028591156,
+      "learning_rate": 0.0001,
+      "loss": 0.8708,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4978588557725248,
+      "grad_norm": 0.636905312538147,
+      "learning_rate": 0.0001,
+      "loss": 0.9506,
+      "step": 5813
+    },
+    {
+      "epoch": 0.49794450154162384,
+      "grad_norm": 0.6534644365310669,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 5814
+    },
+    {
+      "epoch": 0.49803014731072287,
+      "grad_norm": 0.6085953712463379,
+      "learning_rate": 0.0001,
+      "loss": 0.5721,
+      "step": 5815
+    },
+    {
+      "epoch": 0.49811579307982184,
+      "grad_norm": 0.4165063500404358,
+      "learning_rate": 0.0001,
+      "loss": 0.5289,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4982014388489209,
+      "grad_norm": 1.0724955797195435,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 5817
+    },
+    {
+      "epoch": 0.49828708461801985,
+      "grad_norm": 0.6442351341247559,
+      "learning_rate": 0.0001,
+      "loss": 1.1587,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4983727303871189,
+      "grad_norm": 0.8955477476119995,
+      "learning_rate": 0.0001,
+      "loss": 1.3754,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4984583761562179,
+      "grad_norm": 0.7058886885643005,
+      "learning_rate": 0.0001,
+      "loss": 0.6661,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4985440219253169,
+      "grad_norm": 0.8309382200241089,
+      "learning_rate": 0.0001,
+      "loss": 0.8577,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4986296676944159,
+      "grad_norm": 1.1961333751678467,
+      "learning_rate": 0.0001,
+      "loss": 0.8046,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4987153134635149,
+      "grad_norm": 0.7699986100196838,
+      "learning_rate": 0.0001,
+      "loss": 1.3266,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4988009592326139,
+      "grad_norm": 0.6105479598045349,
+      "learning_rate": 0.0001,
+      "loss": 0.6157,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4988866050017129,
+      "grad_norm": 0.6160721182823181,
+      "learning_rate": 0.0001,
+      "loss": 0.9441,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4989722507708119,
+      "grad_norm": 1.1475261449813843,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 5826
+    },
+    {
+      "epoch": 0.49905789653991095,
+      "grad_norm": 0.5895989537239075,
+      "learning_rate": 0.0001,
+      "loss": 1.0808,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4991435423090099,
+      "grad_norm": 1.022177815437317,
+      "learning_rate": 0.0001,
+      "loss": 1.1315,
+      "step": 5828
+    },
+    {
+      "epoch": 0.49922918807810895,
+      "grad_norm": 0.399362176656723,
+      "learning_rate": 0.0001,
+      "loss": 0.288,
+      "step": 5829
+    },
+    {
+      "epoch": 0.49931483384720793,
+      "grad_norm": 1.1360052824020386,
+      "learning_rate": 0.0001,
+      "loss": 1.2317,
+      "step": 5830
+    },
+    {
+      "epoch": 0.49940047961630696,
+      "grad_norm": 0.5358473658561707,
+      "learning_rate": 0.0001,
+      "loss": 1.019,
+      "step": 5831
+    },
+    {
+      "epoch": 0.499486125385406,
+      "grad_norm": 0.8200550079345703,
+      "learning_rate": 0.0001,
+      "loss": 0.5449,
+      "step": 5832
+    },
+    {
+      "epoch": 0.49957177115450496,
+      "grad_norm": 0.5058261156082153,
+      "learning_rate": 0.0001,
+      "loss": 0.2908,
+      "step": 5833
+    },
+    {
+      "epoch": 0.499657416923604,
+      "grad_norm": 0.7566227316856384,
+      "learning_rate": 0.0001,
+      "loss": 0.5665,
+      "step": 5834
+    },
+    {
+      "epoch": 0.49974306269270297,
+      "grad_norm": 0.38788190484046936,
+      "learning_rate": 0.0001,
+      "loss": 0.3211,
+      "step": 5835
+    },
+    {
+      "epoch": 0.499828708461802,
+      "grad_norm": 0.6876218914985657,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 5836
+    },
+    {
+      "epoch": 0.49991435423090097,
+      "grad_norm": 1.0310086011886597,
+      "learning_rate": 0.0001,
+      "loss": 1.1247,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6756823062896729,
+      "learning_rate": 0.0001,
+      "loss": 1.1641,
+      "step": 5838
+    },
+    {
+      "epoch": 0.500085645769099,
+      "grad_norm": 0.9354893565177917,
+      "learning_rate": 0.0001,
+      "loss": 0.8826,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5001712915381981,
+      "grad_norm": 1.1756224632263184,
+      "learning_rate": 0.0001,
+      "loss": 0.8167,
+      "step": 5840
+    },
+    {
+      "epoch": 0.500256937307297,
+      "grad_norm": 0.5654330849647522,
+      "learning_rate": 0.0001,
+      "loss": 0.8686,
+      "step": 5841
+    },
+    {
+      "epoch": 0.500342583076396,
+      "grad_norm": 0.8241333961486816,
+      "learning_rate": 0.0001,
+      "loss": 0.7799,
+      "step": 5842
+    },
+    {
+      "epoch": 0.500428228845495,
+      "grad_norm": 0.96967613697052,
+      "learning_rate": 0.0001,
+      "loss": 1.1597,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5005138746145941,
+      "grad_norm": 0.9478892683982849,
+      "learning_rate": 0.0001,
+      "loss": 1.0844,
+      "step": 5844
+    },
+    {
+      "epoch": 0.500599520383693,
+      "grad_norm": 0.6748441457748413,
+      "learning_rate": 0.0001,
+      "loss": 1.0556,
+      "step": 5845
+    },
+    {
+      "epoch": 0.500685166152792,
+      "grad_norm": 0.7425933480262756,
+      "learning_rate": 0.0001,
+      "loss": 0.9574,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5007708119218911,
+      "grad_norm": 1.3981232643127441,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5008564576909901,
+      "grad_norm": 1.0331090688705444,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 5848
+    },
+    {
+      "epoch": 0.500942103460089,
+      "grad_norm": 0.7965897917747498,
+      "learning_rate": 0.0001,
+      "loss": 0.7545,
+      "step": 5849
+    },
+    {
+      "epoch": 0.501027749229188,
+      "grad_norm": 0.6504186391830444,
+      "learning_rate": 0.0001,
+      "loss": 1.01,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5011133949982871,
+      "grad_norm": 1.1582683324813843,
+      "learning_rate": 0.0001,
+      "loss": 0.9841,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5011990407673861,
+      "grad_norm": 0.37405115365982056,
+      "learning_rate": 0.0001,
+      "loss": 0.5822,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5012846865364851,
+      "grad_norm": 0.5569260716438293,
+      "learning_rate": 0.0001,
+      "loss": 0.9109,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5013703323055841,
+      "grad_norm": 1.1612204313278198,
+      "learning_rate": 0.0001,
+      "loss": 0.5282,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5014559780746831,
+      "grad_norm": 0.324550062417984,
+      "learning_rate": 0.0001,
+      "loss": 0.1084,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5015416238437821,
+      "grad_norm": 0.8313344120979309,
+      "learning_rate": 0.0001,
+      "loss": 0.7895,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5016272696128812,
+      "grad_norm": 0.807465672492981,
+      "learning_rate": 0.0001,
+      "loss": 0.8141,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5017129153819802,
+      "grad_norm": 1.2387615442276,
+      "learning_rate": 0.0001,
+      "loss": 1.1831,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5017985611510791,
+      "grad_norm": 0.6001763343811035,
+      "learning_rate": 0.0001,
+      "loss": 0.4858,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5018842069201781,
+      "grad_norm": 0.8169418573379517,
+      "learning_rate": 0.0001,
+      "loss": 1.2204,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5019698526892772,
+      "grad_norm": 0.649381160736084,
+      "learning_rate": 0.0001,
+      "loss": 0.7644,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5020554984583762,
+      "grad_norm": 0.5743280649185181,
+      "learning_rate": 0.0001,
+      "loss": 1.0966,
+      "step": 5862
+    },
+    {
+      "epoch": 0.5021411442274751,
+      "grad_norm": 1.1451730728149414,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5022267899965742,
+      "grad_norm": 0.9714857935905457,
+      "learning_rate": 0.0001,
+      "loss": 0.783,
+      "step": 5864
+    },
+    {
+      "epoch": 0.5023124357656732,
+      "grad_norm": 1.0528827905654907,
+      "learning_rate": 0.0001,
+      "loss": 0.9854,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5023980815347722,
+      "grad_norm": 0.7576881647109985,
+      "learning_rate": 0.0001,
+      "loss": 0.8398,
+      "step": 5866
+    },
+    {
+      "epoch": 0.5024837273038711,
+      "grad_norm": 1.0439414978027344,
+      "learning_rate": 0.0001,
+      "loss": 0.8794,
+      "step": 5867
+    },
+    {
+      "epoch": 0.5025693730729702,
+      "grad_norm": 0.7455350160598755,
+      "learning_rate": 0.0001,
+      "loss": 0.7344,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5026550188420692,
+      "grad_norm": 0.9458174705505371,
+      "learning_rate": 0.0001,
+      "loss": 1.176,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5027406646111682,
+      "grad_norm": 0.9833660125732422,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5028263103802673,
+      "grad_norm": 0.6223528981208801,
+      "learning_rate": 0.0001,
+      "loss": 0.9199,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5029119561493662,
+      "grad_norm": 0.7934913635253906,
+      "learning_rate": 0.0001,
+      "loss": 1.3435,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5029976019184652,
+      "grad_norm": 0.6990629434585571,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5030832476875642,
+      "grad_norm": 0.973822832107544,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5031688934566633,
+      "grad_norm": 1.1568291187286377,
+      "learning_rate": 0.0001,
+      "loss": 1.9971,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5032545392257622,
+      "grad_norm": 1.0234981775283813,
+      "learning_rate": 0.0001,
+      "loss": 1.1818,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5033401849948612,
+      "grad_norm": 0.7278028726577759,
+      "learning_rate": 0.0001,
+      "loss": 1.3563,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5034258307639603,
+      "grad_norm": 1.0407941341400146,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5035114765330593,
+      "grad_norm": 0.4539453685283661,
+      "learning_rate": 0.0001,
+      "loss": 0.3927,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5035971223021583,
+      "grad_norm": 0.8166367411613464,
+      "learning_rate": 0.0001,
+      "loss": 1.0546,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5036827680712572,
+      "grad_norm": 0.8655957579612732,
+      "learning_rate": 0.0001,
+      "loss": 1.0998,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5037684138403563,
+      "grad_norm": 0.9500569105148315,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5038540596094553,
+      "grad_norm": 1.0782804489135742,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5039397053785543,
+      "grad_norm": 1.7914961576461792,
+      "learning_rate": 0.0001,
+      "loss": 1.1635,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5040253511476533,
+      "grad_norm": 0.8860687613487244,
+      "learning_rate": 0.0001,
+      "loss": 0.6338,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5041109969167523,
+      "grad_norm": 0.6772682070732117,
+      "learning_rate": 0.0001,
+      "loss": 1.0836,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5041966426858513,
+      "grad_norm": 1.230628252029419,
+      "learning_rate": 0.0001,
+      "loss": 1.1677,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5042822884549504,
+      "grad_norm": 0.7088826894760132,
+      "learning_rate": 0.0001,
+      "loss": 0.9973,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5043679342240494,
+      "grad_norm": 1.1550877094268799,
+      "learning_rate": 0.0001,
+      "loss": 0.7334,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5044535799931483,
+      "grad_norm": 0.5767970681190491,
+      "learning_rate": 0.0001,
+      "loss": 1.0402,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5045392257622473,
+      "grad_norm": 0.44289442896842957,
+      "learning_rate": 0.0001,
+      "loss": 0.958,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5046248715313464,
+      "grad_norm": 0.6259189248085022,
+      "learning_rate": 0.0001,
+      "loss": 0.6949,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5047105173004454,
+      "grad_norm": 0.5851679444313049,
+      "learning_rate": 0.0001,
+      "loss": 0.6834,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5047961630695443,
+      "grad_norm": 0.8618548512458801,
+      "learning_rate": 0.0001,
+      "loss": 1.0683,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5048818088386434,
+      "grad_norm": 1.1376477479934692,
+      "learning_rate": 0.0001,
+      "loss": 0.8105,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5049674546077424,
+      "grad_norm": 1.4625635147094727,
+      "learning_rate": 0.0001,
+      "loss": 1.2444,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5050531003768414,
+      "grad_norm": 0.897398829460144,
+      "learning_rate": 0.0001,
+      "loss": 0.8848,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5051387461459403,
+      "grad_norm": 0.5814123153686523,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5052243919150394,
+      "grad_norm": 0.6909059882164001,
+      "learning_rate": 0.0001,
+      "loss": 1.0826,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5053100376841384,
+      "grad_norm": 0.49616536498069763,
+      "learning_rate": 0.0001,
+      "loss": 0.9971,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5053956834532374,
+      "grad_norm": 0.6246143579483032,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5054813292223365,
+      "grad_norm": 0.5309974551200867,
+      "learning_rate": 0.0001,
+      "loss": 0.901,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5055669749914354,
+      "grad_norm": 0.868180513381958,
+      "learning_rate": 0.0001,
+      "loss": 0.7902,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5056526207605344,
+      "grad_norm": 0.9781628847122192,
+      "learning_rate": 0.0001,
+      "loss": 0.7782,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5057382665296334,
+      "grad_norm": 0.5587007403373718,
+      "learning_rate": 0.0001,
+      "loss": 0.8799,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5058239122987325,
+      "grad_norm": 0.4745989441871643,
+      "learning_rate": 0.0001,
+      "loss": 0.4048,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5059095580678314,
+      "grad_norm": 1.11418616771698,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5059952038369304,
+      "grad_norm": 0.5102430582046509,
+      "learning_rate": 0.0001,
+      "loss": 0.6035,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5060808496060295,
+      "grad_norm": 0.408741295337677,
+      "learning_rate": 0.0001,
+      "loss": 0.8341,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5061664953751285,
+      "grad_norm": 0.358875036239624,
+      "learning_rate": 0.0001,
+      "loss": 0.2945,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5062521411442275,
+      "grad_norm": 0.9516525268554688,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5063377869133264,
+      "grad_norm": 0.7605562210083008,
+      "learning_rate": 0.0001,
+      "loss": 0.594,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5064234326824255,
+      "grad_norm": 1.1954560279846191,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5065090784515245,
+      "grad_norm": 0.7738981246948242,
+      "learning_rate": 0.0001,
+      "loss": 1.1878,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5065947242206235,
+      "grad_norm": 1.2315337657928467,
+      "learning_rate": 0.0001,
+      "loss": 1.1552,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5066803699897225,
+      "grad_norm": 0.9902182221412659,
+      "learning_rate": 0.0001,
+      "loss": 1.0733,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5067660157588215,
+      "grad_norm": 1.0875972509384155,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5068516615279205,
+      "grad_norm": 0.8811930418014526,
+      "learning_rate": 0.0001,
+      "loss": 0.8643,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5069373072970196,
+      "grad_norm": 0.5278341770172119,
+      "learning_rate": 0.0001,
+      "loss": 0.8901,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5070229530661186,
+      "grad_norm": 0.38562512397766113,
+      "learning_rate": 0.0001,
+      "loss": 0.4172,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5071085988352175,
+      "grad_norm": 0.8026906847953796,
+      "learning_rate": 0.0001,
+      "loss": 0.8581,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5071942446043165,
+      "grad_norm": 0.3141734302043915,
+      "learning_rate": 0.0001,
+      "loss": 0.4073,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5072798903734156,
+      "grad_norm": 0.793424665927887,
+      "learning_rate": 0.0001,
+      "loss": 1.1333,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5073655361425146,
+      "grad_norm": 1.0521159172058105,
+      "learning_rate": 0.0001,
+      "loss": 1.0765,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5074511819116135,
+      "grad_norm": 0.7650701403617859,
+      "learning_rate": 0.0001,
+      "loss": 0.7904,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5075368276807126,
+      "grad_norm": 1.0421805381774902,
+      "learning_rate": 0.0001,
+      "loss": 1.1618,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5076224734498116,
+      "grad_norm": 0.8244573473930359,
+      "learning_rate": 0.0001,
+      "loss": 0.7757,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5077081192189106,
+      "grad_norm": 1.1196900606155396,
+      "learning_rate": 0.0001,
+      "loss": 1.1008,
+      "step": 5928
+    },
+    {
+      "epoch": 0.5077937649880095,
+      "grad_norm": 0.852834939956665,
+      "learning_rate": 0.0001,
+      "loss": 0.7218,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5078794107571086,
+      "grad_norm": 0.9222322702407837,
+      "learning_rate": 0.0001,
+      "loss": 1.0672,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5079650565262076,
+      "grad_norm": 1.1983267068862915,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5080507022953066,
+      "grad_norm": 1.0746763944625854,
+      "learning_rate": 0.0001,
+      "loss": 1.8431,
+      "step": 5932
+    },
+    {
+      "epoch": 0.5081363480644057,
+      "grad_norm": 0.7199786305427551,
+      "learning_rate": 0.0001,
+      "loss": 0.7824,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5082219938335046,
+      "grad_norm": 0.5938516855239868,
+      "learning_rate": 0.0001,
+      "loss": 0.49,
+      "step": 5934
+    },
+    {
+      "epoch": 0.5083076396026036,
+      "grad_norm": 1.5062823295593262,
+      "learning_rate": 0.0001,
+      "loss": 0.9732,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5083932853717026,
+      "grad_norm": 0.5586996078491211,
+      "learning_rate": 0.0001,
+      "loss": 0.5391,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5084789311408017,
+      "grad_norm": 0.6262711882591248,
+      "learning_rate": 0.0001,
+      "loss": 0.8729,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5085645769099006,
+      "grad_norm": 0.8174694180488586,
+      "learning_rate": 0.0001,
+      "loss": 0.8025,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5086502226789996,
+      "grad_norm": 0.6151589751243591,
+      "learning_rate": 0.0001,
+      "loss": 1.1367,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5087358684480987,
+      "grad_norm": 0.9262535572052002,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5088215142171977,
+      "grad_norm": 0.5328113436698914,
+      "learning_rate": 0.0001,
+      "loss": 0.8635,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5089071599862967,
+      "grad_norm": 0.7352861762046814,
+      "learning_rate": 0.0001,
+      "loss": 1.0112,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5089928057553957,
+      "grad_norm": 0.6943427324295044,
+      "learning_rate": 0.0001,
+      "loss": 1.0534,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5090784515244947,
+      "grad_norm": 1.1733678579330444,
+      "learning_rate": 0.0001,
+      "loss": 0.888,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5091640972935937,
+      "grad_norm": 0.5697802901268005,
+      "learning_rate": 0.0001,
+      "loss": 1.2649,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5092497430626927,
+      "grad_norm": 0.7265795469284058,
+      "learning_rate": 0.0001,
+      "loss": 0.8047,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5093353888317917,
+      "grad_norm": 0.9106659889221191,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5094210346008907,
+      "grad_norm": 0.9665881395339966,
+      "learning_rate": 0.0001,
+      "loss": 0.8379,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5095066803699897,
+      "grad_norm": 1.0903408527374268,
+      "learning_rate": 0.0001,
+      "loss": 1.3064,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5095923261390888,
+      "grad_norm": 0.6641898155212402,
+      "learning_rate": 0.0001,
+      "loss": 0.5339,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5096779719081878,
+      "grad_norm": 0.686638593673706,
+      "learning_rate": 0.0001,
+      "loss": 1.3027,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5097636176772867,
+      "grad_norm": 0.9261236786842346,
+      "learning_rate": 0.0001,
+      "loss": 1.1163,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5098492634463857,
+      "grad_norm": 0.7841808199882507,
+      "learning_rate": 0.0001,
+      "loss": 1.2218,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5099349092154848,
+      "grad_norm": 0.6010593175888062,
+      "learning_rate": 0.0001,
+      "loss": 0.7949,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5100205549845838,
+      "grad_norm": 0.5729544758796692,
+      "learning_rate": 0.0001,
+      "loss": 0.6229,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5101062007536827,
+      "grad_norm": 0.7680389881134033,
+      "learning_rate": 0.0001,
+      "loss": 0.648,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5101918465227818,
+      "grad_norm": 0.852963387966156,
+      "learning_rate": 0.0001,
+      "loss": 0.825,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5102774922918808,
+      "grad_norm": 0.8562958836555481,
+      "learning_rate": 0.0001,
+      "loss": 0.6201,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5103631380609798,
+      "grad_norm": 1.0085214376449585,
+      "learning_rate": 0.0001,
+      "loss": 1.2083,
+      "step": 5959
+    },
+    {
+      "epoch": 0.5104487838300787,
+      "grad_norm": 0.659272313117981,
+      "learning_rate": 0.0001,
+      "loss": 0.8524,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5105344295991778,
+      "grad_norm": 0.9019708633422852,
+      "learning_rate": 0.0001,
+      "loss": 0.9859,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5106200753682768,
+      "grad_norm": 0.8270323276519775,
+      "learning_rate": 0.0001,
+      "loss": 1.3029,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5107057211373758,
+      "grad_norm": 1.5466822385787964,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 5963
+    },
+    {
+      "epoch": 0.5107913669064749,
+      "grad_norm": 0.4455975890159607,
+      "learning_rate": 0.0001,
+      "loss": 0.7822,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5108770126755738,
+      "grad_norm": 1.0497636795043945,
+      "learning_rate": 0.0001,
+      "loss": 2.201,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5109626584446728,
+      "grad_norm": 0.6277524828910828,
+      "learning_rate": 0.0001,
+      "loss": 1.0548,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5110483042137718,
+      "grad_norm": 0.6096447110176086,
+      "learning_rate": 0.0001,
+      "loss": 1.0517,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5111339499828709,
+      "grad_norm": 1.1108351945877075,
+      "learning_rate": 0.0001,
+      "loss": 1.1273,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5112195957519698,
+      "grad_norm": 0.7292288541793823,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5113052415210688,
+      "grad_norm": 0.4705936014652252,
+      "learning_rate": 0.0001,
+      "loss": 1.0202,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5113908872901679,
+      "grad_norm": 1.3248143196105957,
+      "learning_rate": 0.0001,
+      "loss": 0.9126,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5114765330592669,
+      "grad_norm": 0.4357629120349884,
+      "learning_rate": 0.0001,
+      "loss": 0.5148,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5115621788283659,
+      "grad_norm": 0.9458514451980591,
+      "learning_rate": 0.0001,
+      "loss": 1.7828,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5116478245974649,
+      "grad_norm": 0.9188002943992615,
+      "learning_rate": 0.0001,
+      "loss": 1.0836,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5117334703665639,
+      "grad_norm": 1.088136076927185,
+      "learning_rate": 0.0001,
+      "loss": 0.9223,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5118191161356629,
+      "grad_norm": 1.0987064838409424,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5119047619047619,
+      "grad_norm": 0.9495627284049988,
+      "learning_rate": 0.0001,
+      "loss": 1.0961,
+      "step": 5977
+    },
+    {
+      "epoch": 0.511990407673861,
+      "grad_norm": 1.0392324924468994,
+      "learning_rate": 0.0001,
+      "loss": 1.0163,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5120760534429599,
+      "grad_norm": 1.387008547782898,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5121616992120589,
+      "grad_norm": 0.5608152151107788,
+      "learning_rate": 0.0001,
+      "loss": 1.3357,
+      "step": 5980
+    },
+    {
+      "epoch": 0.512247344981158,
+      "grad_norm": 0.7131004929542542,
+      "learning_rate": 0.0001,
+      "loss": 1.2511,
+      "step": 5981
+    },
+    {
+      "epoch": 0.512332990750257,
+      "grad_norm": 1.644274353981018,
+      "learning_rate": 0.0001,
+      "loss": 0.858,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5124186365193559,
+      "grad_norm": 0.9941288232803345,
+      "learning_rate": 0.0001,
+      "loss": 0.915,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5125042822884549,
+      "grad_norm": 0.6245418190956116,
+      "learning_rate": 0.0001,
+      "loss": 0.9639,
+      "step": 5984
+    },
+    {
+      "epoch": 0.512589928057554,
+      "grad_norm": 0.6836696267127991,
+      "learning_rate": 0.0001,
+      "loss": 1.101,
+      "step": 5985
+    },
+    {
+      "epoch": 0.512675573826653,
+      "grad_norm": 0.46983182430267334,
+      "learning_rate": 0.0001,
+      "loss": 0.5209,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5127612195957519,
+      "grad_norm": 1.0748683214187622,
+      "learning_rate": 0.0001,
+      "loss": 1.2603,
+      "step": 5987
+    },
+    {
+      "epoch": 0.512846865364851,
+      "grad_norm": 0.6538723707199097,
+      "learning_rate": 0.0001,
+      "loss": 0.5607,
+      "step": 5988
+    },
+    {
+      "epoch": 0.51293251113395,
+      "grad_norm": 0.5200199484825134,
+      "learning_rate": 0.0001,
+      "loss": 0.9114,
+      "step": 5989
+    },
+    {
+      "epoch": 0.513018156903049,
+      "grad_norm": 0.6280938982963562,
+      "learning_rate": 0.0001,
+      "loss": 0.5502,
+      "step": 5990
+    },
+    {
+      "epoch": 0.513103802672148,
+      "grad_norm": 1.0686701536178589,
+      "learning_rate": 0.0001,
+      "loss": 1.0526,
+      "step": 5991
+    },
+    {
+      "epoch": 0.513189448441247,
+      "grad_norm": 0.777652382850647,
+      "learning_rate": 0.0001,
+      "loss": 0.6253,
+      "step": 5992
+    },
+    {
+      "epoch": 0.513275094210346,
+      "grad_norm": 0.6483321189880371,
+      "learning_rate": 0.0001,
+      "loss": 1.0964,
+      "step": 5993
+    },
+    {
+      "epoch": 0.513360739979445,
+      "grad_norm": 0.55298912525177,
+      "learning_rate": 0.0001,
+      "loss": 1.0923,
+      "step": 5994
+    },
+    {
+      "epoch": 0.5134463857485441,
+      "grad_norm": 1.4989031553268433,
+      "learning_rate": 0.0001,
+      "loss": 1.153,
+      "step": 5995
+    },
+    {
+      "epoch": 0.513532031517643,
+      "grad_norm": 0.7619442343711853,
+      "learning_rate": 0.0001,
+      "loss": 0.7807,
+      "step": 5996
+    },
+    {
+      "epoch": 0.513617677286742,
+      "grad_norm": 0.6579172015190125,
+      "learning_rate": 0.0001,
+      "loss": 0.8466,
+      "step": 5997
+    },
+    {
+      "epoch": 0.513703323055841,
+      "grad_norm": 0.9630225896835327,
+      "learning_rate": 0.0001,
+      "loss": 1.0361,
+      "step": 5998
+    },
+    {
+      "epoch": 0.5137889688249401,
+      "grad_norm": 0.7024539113044739,
+      "learning_rate": 0.0001,
+      "loss": 0.6188,
+      "step": 5999
+    },
+    {
+      "epoch": 0.513874614594039,
+      "grad_norm": 1.491180181503296,
+      "learning_rate": 0.0001,
+      "loss": 0.8746,
+      "step": 6000
+    },
+    {
+      "epoch": 0.513960260363138,
+      "grad_norm": 0.46239954233169556,
+      "learning_rate": 0.0001,
+      "loss": 0.3551,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5140459061322371,
+      "grad_norm": 0.8119739890098572,
+      "learning_rate": 0.0001,
+      "loss": 0.8746,
+      "step": 6002
+    },
+    {
+      "epoch": 0.5141315519013361,
+      "grad_norm": 0.7159539461135864,
+      "learning_rate": 0.0001,
+      "loss": 0.7062,
+      "step": 6003
+    },
+    {
+      "epoch": 0.5142171976704351,
+      "grad_norm": 1.0735424757003784,
+      "learning_rate": 0.0001,
+      "loss": 0.6902,
+      "step": 6004
+    },
+    {
+      "epoch": 0.5143028434395341,
+      "grad_norm": 0.8766245245933533,
+      "learning_rate": 0.0001,
+      "loss": 1.0153,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5143884892086331,
+      "grad_norm": 1.0319677591323853,
+      "learning_rate": 0.0001,
+      "loss": 1.0189,
+      "step": 6006
+    },
+    {
+      "epoch": 0.5144741349777321,
+      "grad_norm": 0.6851661801338196,
+      "learning_rate": 0.0001,
+      "loss": 0.9146,
+      "step": 6007
+    },
+    {
+      "epoch": 0.5145597807468311,
+      "grad_norm": 0.7592886686325073,
+      "learning_rate": 0.0001,
+      "loss": 0.7998,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5146454265159301,
+      "grad_norm": 0.5554174780845642,
+      "learning_rate": 0.0001,
+      "loss": 0.6051,
+      "step": 6009
+    },
+    {
+      "epoch": 0.5147310722850291,
+      "grad_norm": 0.886280357837677,
+      "learning_rate": 0.0001,
+      "loss": 1.3083,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5148167180541281,
+      "grad_norm": 0.5851724147796631,
+      "learning_rate": 0.0001,
+      "loss": 0.4167,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5149023638232272,
+      "grad_norm": 0.9096223711967468,
+      "learning_rate": 0.0001,
+      "loss": 0.9219,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5149880095923262,
+      "grad_norm": 0.5984715223312378,
+      "learning_rate": 0.0001,
+      "loss": 0.9353,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5150736553614251,
+      "grad_norm": 0.5777068734169006,
+      "learning_rate": 0.0001,
+      "loss": 0.9228,
+      "step": 6014
+    },
+    {
+      "epoch": 0.5151593011305241,
+      "grad_norm": 0.7172765135765076,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5152449468996232,
+      "grad_norm": 0.5140990018844604,
+      "learning_rate": 0.0001,
+      "loss": 0.973,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5153305926687222,
+      "grad_norm": 0.42758694291114807,
+      "learning_rate": 0.0001,
+      "loss": 0.3772,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5154162384378211,
+      "grad_norm": 0.8788056969642639,
+      "learning_rate": 0.0001,
+      "loss": 1.2359,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5155018842069202,
+      "grad_norm": 0.8502061367034912,
+      "learning_rate": 0.0001,
+      "loss": 0.9846,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5155875299760192,
+      "grad_norm": 0.6751374006271362,
+      "learning_rate": 0.0001,
+      "loss": 1.2265,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5156731757451182,
+      "grad_norm": 0.795085608959198,
+      "learning_rate": 0.0001,
+      "loss": 1.0112,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5157588215142171,
+      "grad_norm": 1.135378360748291,
+      "learning_rate": 0.0001,
+      "loss": 0.8649,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5158444672833162,
+      "grad_norm": 1.0241308212280273,
+      "learning_rate": 0.0001,
+      "loss": 1.312,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5159301130524152,
+      "grad_norm": 0.6046897172927856,
+      "learning_rate": 0.0001,
+      "loss": 1.2089,
+      "step": 6024
+    },
+    {
+      "epoch": 0.5160157588215142,
+      "grad_norm": 0.36005839705467224,
+      "learning_rate": 0.0001,
+      "loss": 0.1568,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5161014045906133,
+      "grad_norm": 0.9275981187820435,
+      "learning_rate": 0.0001,
+      "loss": 0.802,
+      "step": 6026
+    },
+    {
+      "epoch": 0.5161870503597122,
+      "grad_norm": 0.7925171256065369,
+      "learning_rate": 0.0001,
+      "loss": 1.1441,
+      "step": 6027
+    },
+    {
+      "epoch": 0.5162726961288112,
+      "grad_norm": 0.2713598906993866,
+      "learning_rate": 0.0001,
+      "loss": 0.2047,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5163583418979102,
+      "grad_norm": 0.6563075184822083,
+      "learning_rate": 0.0001,
+      "loss": 1.2081,
+      "step": 6029
+    },
+    {
+      "epoch": 0.5164439876670093,
+      "grad_norm": 0.5925971269607544,
+      "learning_rate": 0.0001,
+      "loss": 1.0034,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5165296334361082,
+      "grad_norm": 0.6389843225479126,
+      "learning_rate": 0.0001,
+      "loss": 0.667,
+      "step": 6031
+    },
+    {
+      "epoch": 0.5166152792052072,
+      "grad_norm": 0.6342236995697021,
+      "learning_rate": 0.0001,
+      "loss": 0.5879,
+      "step": 6032
+    },
+    {
+      "epoch": 0.5167009249743063,
+      "grad_norm": 0.805618405342102,
+      "learning_rate": 0.0001,
+      "loss": 0.9297,
+      "step": 6033
+    },
+    {
+      "epoch": 0.5167865707434053,
+      "grad_norm": 0.893006443977356,
+      "learning_rate": 0.0001,
+      "loss": 1.2226,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5168722165125043,
+      "grad_norm": 0.903538703918457,
+      "learning_rate": 0.0001,
+      "loss": 0.5369,
+      "step": 6035
+    },
+    {
+      "epoch": 0.5169578622816033,
+      "grad_norm": 0.7285211086273193,
+      "learning_rate": 0.0001,
+      "loss": 0.6098,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5170435080507023,
+      "grad_norm": 0.8404451608657837,
+      "learning_rate": 0.0001,
+      "loss": 0.7957,
+      "step": 6037
+    },
+    {
+      "epoch": 0.5171291538198013,
+      "grad_norm": 0.6208089590072632,
+      "learning_rate": 0.0001,
+      "loss": 0.5199,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5172147995889003,
+      "grad_norm": 0.9013316035270691,
+      "learning_rate": 0.0001,
+      "loss": 0.7927,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5173004453579994,
+      "grad_norm": 0.5919445157051086,
+      "learning_rate": 0.0001,
+      "loss": 0.9688,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5173860911270983,
+      "grad_norm": 1.1397289037704468,
+      "learning_rate": 0.0001,
+      "loss": 2.1771,
+      "step": 6041
+    },
+    {
+      "epoch": 0.5174717368961973,
+      "grad_norm": 1.1371204853057861,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5175573826652964,
+      "grad_norm": 0.7905836701393127,
+      "learning_rate": 0.0001,
+      "loss": 0.5525,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5176430284343954,
+      "grad_norm": 0.9153354167938232,
+      "learning_rate": 0.0001,
+      "loss": 0.8081,
+      "step": 6044
+    },
+    {
+      "epoch": 0.5177286742034943,
+      "grad_norm": 0.9178486466407776,
+      "learning_rate": 0.0001,
+      "loss": 0.6874,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5178143199725933,
+      "grad_norm": 0.5084734559059143,
+      "learning_rate": 0.0001,
+      "loss": 0.9178,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5178999657416924,
+      "grad_norm": 1.184606671333313,
+      "learning_rate": 0.0001,
+      "loss": 0.9919,
+      "step": 6047
+    },
+    {
+      "epoch": 0.5179856115107914,
+      "grad_norm": 0.7210314869880676,
+      "learning_rate": 0.0001,
+      "loss": 0.6971,
+      "step": 6048
+    },
+    {
+      "epoch": 0.5180712572798903,
+      "grad_norm": 0.6225554943084717,
+      "learning_rate": 0.0001,
+      "loss": 0.5281,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5181569030489894,
+      "grad_norm": 0.6301462650299072,
+      "learning_rate": 0.0001,
+      "loss": 0.508,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5182425488180884,
+      "grad_norm": 1.2904566526412964,
+      "learning_rate": 0.0001,
+      "loss": 1.3171,
+      "step": 6051
+    },
+    {
+      "epoch": 0.5183281945871874,
+      "grad_norm": 0.9720475077629089,
+      "learning_rate": 0.0001,
+      "loss": 0.8409,
+      "step": 6052
+    },
+    {
+      "epoch": 0.5184138403562863,
+      "grad_norm": 0.8356928825378418,
+      "learning_rate": 0.0001,
+      "loss": 1.3227,
+      "step": 6053
+    },
+    {
+      "epoch": 0.5184994861253854,
+      "grad_norm": 0.9746174812316895,
+      "learning_rate": 0.0001,
+      "loss": 1.2787,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5185851318944844,
+      "grad_norm": 0.6246987581253052,
+      "learning_rate": 0.0001,
+      "loss": 0.934,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5186707776635834,
+      "grad_norm": 0.7436205744743347,
+      "learning_rate": 0.0001,
+      "loss": 0.4035,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5187564234326825,
+      "grad_norm": 0.879176676273346,
+      "learning_rate": 0.0001,
+      "loss": 1.0324,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5188420692017814,
+      "grad_norm": 0.8010760545730591,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5189277149708804,
+      "grad_norm": 0.5964545607566833,
+      "learning_rate": 0.0001,
+      "loss": 1.1033,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5190133607399794,
+      "grad_norm": 0.7466155886650085,
+      "learning_rate": 0.0001,
+      "loss": 0.7451,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5190990065090785,
+      "grad_norm": 0.9114970564842224,
+      "learning_rate": 0.0001,
+      "loss": 1.0347,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5191846522781774,
+      "grad_norm": 0.6018864512443542,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5192702980472764,
+      "grad_norm": 0.952886164188385,
+      "learning_rate": 0.0001,
+      "loss": 1.0369,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5193559438163755,
+      "grad_norm": 0.7140659689903259,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5194415895854745,
+      "grad_norm": 0.899229884147644,
+      "learning_rate": 0.0001,
+      "loss": 0.5978,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5195272353545735,
+      "grad_norm": 0.7698892951011658,
+      "learning_rate": 0.0001,
+      "loss": 0.5342,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5196128811236725,
+      "grad_norm": 1.3356425762176514,
+      "learning_rate": 0.0001,
+      "loss": 0.5471,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5196985268927715,
+      "grad_norm": 0.39604341983795166,
+      "learning_rate": 0.0001,
+      "loss": 0.5111,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5197841726618705,
+      "grad_norm": 0.7917873859405518,
+      "learning_rate": 0.0001,
+      "loss": 0.9416,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5198698184309695,
+      "grad_norm": 0.7827185392379761,
+      "learning_rate": 0.0001,
+      "loss": 0.8637,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5199554642000686,
+      "grad_norm": 0.966883659362793,
+      "learning_rate": 0.0001,
+      "loss": 1.1849,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5200411099691675,
+      "grad_norm": 0.5535942316055298,
+      "learning_rate": 0.0001,
+      "loss": 1.1019,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5201267557382665,
+      "grad_norm": 0.700545608997345,
+      "learning_rate": 0.0001,
+      "loss": 0.816,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5202124015073656,
+      "grad_norm": 1.1916145086288452,
+      "learning_rate": 0.0001,
+      "loss": 1.1313,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5202980472764646,
+      "grad_norm": 0.6538026928901672,
+      "learning_rate": 0.0001,
+      "loss": 1.2993,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5203836930455635,
+      "grad_norm": 0.8158594965934753,
+      "learning_rate": 0.0001,
+      "loss": 0.7544,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5204693388146625,
+      "grad_norm": 1.0632644891738892,
+      "learning_rate": 0.0001,
+      "loss": 1.1912,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5205549845837616,
+      "grad_norm": 0.2716015577316284,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5206406303528606,
+      "grad_norm": 0.8193739652633667,
+      "learning_rate": 0.0001,
+      "loss": 0.9133,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5207262761219595,
+      "grad_norm": 0.6077334880828857,
+      "learning_rate": 0.0001,
+      "loss": 1.0437,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5208119218910586,
+      "grad_norm": 0.37221744656562805,
+      "learning_rate": 0.0001,
+      "loss": 0.6192,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5208975676601576,
+      "grad_norm": 0.711104154586792,
+      "learning_rate": 0.0001,
+      "loss": 0.7803,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5209832134292566,
+      "grad_norm": 0.9011251330375671,
+      "learning_rate": 0.0001,
+      "loss": 0.7739,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5210688591983555,
+      "grad_norm": 0.6499476432800293,
+      "learning_rate": 0.0001,
+      "loss": 1.2552,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5211545049674546,
+      "grad_norm": 0.7005913853645325,
+      "learning_rate": 0.0001,
+      "loss": 0.7875,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5212401507365536,
+      "grad_norm": 0.5287420749664307,
+      "learning_rate": 0.0001,
+      "loss": 0.8177,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5213257965056526,
+      "grad_norm": 1.0610438585281372,
+      "learning_rate": 0.0001,
+      "loss": 1.0788,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5214114422747517,
+      "grad_norm": 0.5723148584365845,
+      "learning_rate": 0.0001,
+      "loss": 1.8753,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5214970880438506,
+      "grad_norm": 0.917660117149353,
+      "learning_rate": 0.0001,
+      "loss": 0.9838,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5215827338129496,
+      "grad_norm": 0.3571094572544098,
+      "learning_rate": 0.0001,
+      "loss": 0.509,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5216683795820486,
+      "grad_norm": 0.8390600681304932,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5217540253511477,
+      "grad_norm": 0.5333647131919861,
+      "learning_rate": 0.0001,
+      "loss": 1.0686,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5218396711202467,
+      "grad_norm": 0.7930417060852051,
+      "learning_rate": 0.0001,
+      "loss": 1.1603,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5219253168893456,
+      "grad_norm": 1.3022254705429077,
+      "learning_rate": 0.0001,
+      "loss": 1.0733,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5220109626584447,
+      "grad_norm": 0.4484078288078308,
+      "learning_rate": 0.0001,
+      "loss": 0.3629,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5220966084275437,
+      "grad_norm": 0.5165913105010986,
+      "learning_rate": 0.0001,
+      "loss": 0.5893,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5221822541966427,
+      "grad_norm": 0.8814609050750732,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5222678999657417,
+      "grad_norm": 0.707400918006897,
+      "learning_rate": 0.0001,
+      "loss": 0.8265,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5223535457348407,
+      "grad_norm": 0.5861172080039978,
+      "learning_rate": 0.0001,
+      "loss": 1.2626,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5224391915039397,
+      "grad_norm": 1.0190396308898926,
+      "learning_rate": 0.0001,
+      "loss": 1.2626,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5225248372730387,
+      "grad_norm": 0.3592463433742523,
+      "learning_rate": 0.0001,
+      "loss": 0.6226,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5226104830421378,
+      "grad_norm": 0.5317867994308472,
+      "learning_rate": 0.0001,
+      "loss": 0.8633,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5226961288112367,
+      "grad_norm": 0.3948874771595001,
+      "learning_rate": 0.0001,
+      "loss": 0.7478,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5227817745803357,
+      "grad_norm": 0.7264795303344727,
+      "learning_rate": 0.0001,
+      "loss": 0.7134,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5228674203494348,
+      "grad_norm": 0.7374923229217529,
+      "learning_rate": 0.0001,
+      "loss": 1.3321,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5229530661185338,
+      "grad_norm": 0.6728997230529785,
+      "learning_rate": 0.0001,
+      "loss": 0.7178,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5230387118876327,
+      "grad_norm": 0.713257908821106,
+      "learning_rate": 0.0001,
+      "loss": 0.9565,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5231243576567317,
+      "grad_norm": 0.3750254213809967,
+      "learning_rate": 0.0001,
+      "loss": 0.5746,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5232100034258308,
+      "grad_norm": 0.7563744783401489,
+      "learning_rate": 0.0001,
+      "loss": 0.8623,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5232956491949298,
+      "grad_norm": 0.5428711771965027,
+      "learning_rate": 0.0001,
+      "loss": 0.7176,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5233812949640287,
+      "grad_norm": 1.0136387348175049,
+      "learning_rate": 0.0001,
+      "loss": 0.9006,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5234669407331278,
+      "grad_norm": 0.528102457523346,
+      "learning_rate": 0.0001,
+      "loss": 0.9172,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5235525865022268,
+      "grad_norm": 0.7717801332473755,
+      "learning_rate": 0.0001,
+      "loss": 0.8805,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5236382322713258,
+      "grad_norm": 0.8153849244117737,
+      "learning_rate": 0.0001,
+      "loss": 0.8033,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5237238780404248,
+      "grad_norm": 0.604077935218811,
+      "learning_rate": 0.0001,
+      "loss": 0.9163,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.7109283804893494,
+      "learning_rate": 0.0001,
+      "loss": 0.8841,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5238951695786228,
+      "grad_norm": 0.5991655588150024,
+      "learning_rate": 0.0001,
+      "loss": 0.8454,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5239808153477218,
+      "grad_norm": 0.3101174831390381,
+      "learning_rate": 0.0001,
+      "loss": 0.2592,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5240664611168209,
+      "grad_norm": 0.9692459106445312,
+      "learning_rate": 0.0001,
+      "loss": 1.1553,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5241521068859198,
+      "grad_norm": 0.614037811756134,
+      "learning_rate": 0.0001,
+      "loss": 0.7609,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5242377526550188,
+      "grad_norm": 0.9139115214347839,
+      "learning_rate": 0.0001,
+      "loss": 0.8246,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5243233984241179,
+      "grad_norm": 0.8269458413124084,
+      "learning_rate": 0.0001,
+      "loss": 0.8506,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5244090441932169,
+      "grad_norm": 0.4538160264492035,
+      "learning_rate": 0.0001,
+      "loss": 0.7165,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5244946899623159,
+      "grad_norm": 0.5138024687767029,
+      "learning_rate": 0.0001,
+      "loss": 0.7692,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5245803357314148,
+      "grad_norm": 0.47186607122421265,
+      "learning_rate": 0.0001,
+      "loss": 0.8125,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5246659815005139,
+      "grad_norm": 0.6614543795585632,
+      "learning_rate": 0.0001,
+      "loss": 1.059,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5247516272696129,
+      "grad_norm": 0.987076461315155,
+      "learning_rate": 0.0001,
+      "loss": 0.752,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5248372730387119,
+      "grad_norm": 1.1550955772399902,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 6128
+    },
+    {
+      "epoch": 0.524922918807811,
+      "grad_norm": 0.5719786286354065,
+      "learning_rate": 0.0001,
+      "loss": 1.092,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5250085645769099,
+      "grad_norm": 0.4805266857147217,
+      "learning_rate": 0.0001,
+      "loss": 0.9275,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5250942103460089,
+      "grad_norm": 0.9483142495155334,
+      "learning_rate": 0.0001,
+      "loss": 1.2809,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5251798561151079,
+      "grad_norm": 0.5780265927314758,
+      "learning_rate": 0.0001,
+      "loss": 1.0458,
+      "step": 6132
+    },
+    {
+      "epoch": 0.525265501884207,
+      "grad_norm": 0.8506825566291809,
+      "learning_rate": 0.0001,
+      "loss": 0.3178,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5253511476533059,
+      "grad_norm": 1.130171775817871,
+      "learning_rate": 0.0001,
+      "loss": 1.2144,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5254367934224049,
+      "grad_norm": 0.8522285223007202,
+      "learning_rate": 0.0001,
+      "loss": 0.7208,
+      "step": 6135
+    },
+    {
+      "epoch": 0.525522439191504,
+      "grad_norm": 0.7596500515937805,
+      "learning_rate": 0.0001,
+      "loss": 0.6088,
+      "step": 6136
+    },
+    {
+      "epoch": 0.525608084960603,
+      "grad_norm": 0.9712073802947998,
+      "learning_rate": 0.0001,
+      "loss": 0.8477,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5256937307297019,
+      "grad_norm": 0.2751576900482178,
+      "learning_rate": 0.0001,
+      "loss": 0.1374,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5257793764988009,
+      "grad_norm": 0.6378294825553894,
+      "learning_rate": 0.0001,
+      "loss": 0.8963,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5258650222679,
+      "grad_norm": 1.0930739641189575,
+      "learning_rate": 0.0001,
+      "loss": 0.9247,
+      "step": 6140
+    },
+    {
+      "epoch": 0.525950668036999,
+      "grad_norm": 0.9457883834838867,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5260363138060979,
+      "grad_norm": 0.9462238550186157,
+      "learning_rate": 0.0001,
+      "loss": 0.983,
+      "step": 6142
+    },
+    {
+      "epoch": 0.526121959575197,
+      "grad_norm": 0.8092339038848877,
+      "learning_rate": 0.0001,
+      "loss": 0.8613,
+      "step": 6143
+    },
+    {
+      "epoch": 0.526207605344296,
+      "grad_norm": 0.6743729114532471,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 6144
+    },
+    {
+      "epoch": 0.526293251113395,
+      "grad_norm": 0.8061973452568054,
+      "learning_rate": 0.0001,
+      "loss": 0.4951,
+      "step": 6145
+    },
+    {
+      "epoch": 0.526378896882494,
+      "grad_norm": 0.44764241576194763,
+      "learning_rate": 0.0001,
+      "loss": 0.6207,
+      "step": 6146
+    },
+    {
+      "epoch": 0.526464542651593,
+      "grad_norm": 0.8505096435546875,
+      "learning_rate": 0.0001,
+      "loss": 0.787,
+      "step": 6147
+    },
+    {
+      "epoch": 0.526550188420692,
+      "grad_norm": 0.5975065231323242,
+      "learning_rate": 0.0001,
+      "loss": 0.7413,
+      "step": 6148
+    },
+    {
+      "epoch": 0.526635834189791,
+      "grad_norm": 0.6378376483917236,
+      "learning_rate": 0.0001,
+      "loss": 1.2696,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5267214799588901,
+      "grad_norm": 0.4318898022174835,
+      "learning_rate": 0.0001,
+      "loss": 0.1785,
+      "step": 6150
+    },
+    {
+      "epoch": 0.526807125727989,
+      "grad_norm": 0.790093183517456,
+      "learning_rate": 0.0001,
+      "loss": 0.5821,
+      "step": 6151
+    },
+    {
+      "epoch": 0.526892771497088,
+      "grad_norm": 0.3164728283882141,
+      "learning_rate": 0.0001,
+      "loss": 0.1385,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5269784172661871,
+      "grad_norm": 0.9044681191444397,
+      "learning_rate": 0.0001,
+      "loss": 1.1663,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5270640630352861,
+      "grad_norm": 0.7582246661186218,
+      "learning_rate": 0.0001,
+      "loss": 0.8251,
+      "step": 6154
+    },
+    {
+      "epoch": 0.527149708804385,
+      "grad_norm": 0.7924333214759827,
+      "learning_rate": 0.0001,
+      "loss": 1.0497,
+      "step": 6155
+    },
+    {
+      "epoch": 0.527235354573484,
+      "grad_norm": 0.29058635234832764,
+      "learning_rate": 0.0001,
+      "loss": 0.6616,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5273210003425831,
+      "grad_norm": 1.029937744140625,
+      "learning_rate": 0.0001,
+      "loss": 1.009,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5274066461116821,
+      "grad_norm": 0.7759717106819153,
+      "learning_rate": 0.0001,
+      "loss": 1.0193,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5274922918807811,
+      "grad_norm": 0.7780102491378784,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5275779376498801,
+      "grad_norm": 0.8441569209098816,
+      "learning_rate": 0.0001,
+      "loss": 1.133,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5276635834189791,
+      "grad_norm": 0.9741737246513367,
+      "learning_rate": 0.0001,
+      "loss": 1.3516,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5277492291880781,
+      "grad_norm": 1.3123496770858765,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5278348749571771,
+      "grad_norm": 1.191143274307251,
+      "learning_rate": 0.0001,
+      "loss": 1.9193,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5279205207262762,
+      "grad_norm": 0.7092328667640686,
+      "learning_rate": 0.0001,
+      "loss": 0.3994,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5280061664953751,
+      "grad_norm": 1.2203807830810547,
+      "learning_rate": 0.0001,
+      "loss": 1.1471,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5280918122644741,
+      "grad_norm": 0.8014639616012573,
+      "learning_rate": 0.0001,
+      "loss": 0.4154,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5281774580335732,
+      "grad_norm": 1.0879405736923218,
+      "learning_rate": 0.0001,
+      "loss": 1.2283,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5282631038026722,
+      "grad_norm": 0.6519467830657959,
+      "learning_rate": 0.0001,
+      "loss": 0.6309,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5283487495717711,
+      "grad_norm": 1.0931260585784912,
+      "learning_rate": 0.0001,
+      "loss": 1.1087,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5284343953408701,
+      "grad_norm": 0.7885266542434692,
+      "learning_rate": 0.0001,
+      "loss": 1.1945,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5285200411099692,
+      "grad_norm": 0.9294074177742004,
+      "learning_rate": 0.0001,
+      "loss": 1.1505,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5286056868790682,
+      "grad_norm": 0.8509870767593384,
+      "learning_rate": 0.0001,
+      "loss": 1.1738,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5286913326481671,
+      "grad_norm": 1.1464639902114868,
+      "learning_rate": 0.0001,
+      "loss": 0.8972,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5287769784172662,
+      "grad_norm": 0.7666948437690735,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5288626241863652,
+      "grad_norm": 0.8291814923286438,
+      "learning_rate": 0.0001,
+      "loss": 1.1975,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5289482699554642,
+      "grad_norm": 1.1574949026107788,
+      "learning_rate": 0.0001,
+      "loss": 1.3298,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5290339157245632,
+      "grad_norm": 0.928378701210022,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5291195614936622,
+      "grad_norm": 0.17639267444610596,
+      "learning_rate": 0.0001,
+      "loss": 0.3703,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5292052072627612,
+      "grad_norm": 0.6769609451293945,
+      "learning_rate": 0.0001,
+      "loss": 0.79,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5292908530318602,
+      "grad_norm": 0.8804629445075989,
+      "learning_rate": 0.0001,
+      "loss": 1.2335,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5293764988009593,
+      "grad_norm": 0.7050837874412537,
+      "learning_rate": 0.0001,
+      "loss": 0.7697,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5294621445700582,
+      "grad_norm": 0.5306723117828369,
+      "learning_rate": 0.0001,
+      "loss": 0.4013,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5295477903391572,
+      "grad_norm": 0.8144797086715698,
+      "learning_rate": 0.0001,
+      "loss": 1.1867,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5296334361082563,
+      "grad_norm": 1.0535444021224976,
+      "learning_rate": 0.0001,
+      "loss": 1.08,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5297190818773553,
+      "grad_norm": 0.6550827026367188,
+      "learning_rate": 0.0001,
+      "loss": 1.0136,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5298047276464543,
+      "grad_norm": 0.9748328924179077,
+      "learning_rate": 0.0001,
+      "loss": 1.3092,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5298903734155532,
+      "grad_norm": 1.0596011877059937,
+      "learning_rate": 0.0001,
+      "loss": 0.8823,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5299760191846523,
+      "grad_norm": 0.6751880645751953,
+      "learning_rate": 0.0001,
+      "loss": 0.7794,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5300616649537513,
+      "grad_norm": 0.8408435583114624,
+      "learning_rate": 0.0001,
+      "loss": 1.2789,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5301473107228503,
+      "grad_norm": 0.8205633759498596,
+      "learning_rate": 0.0001,
+      "loss": 1.1775,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5302329564919493,
+      "grad_norm": 0.8470243215560913,
+      "learning_rate": 0.0001,
+      "loss": 0.6035,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5303186022610483,
+      "grad_norm": 0.5372868776321411,
+      "learning_rate": 0.0001,
+      "loss": 1.1982,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5304042480301473,
+      "grad_norm": 1.129098892211914,
+      "learning_rate": 0.0001,
+      "loss": 1.1285,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5304898937992463,
+      "grad_norm": 0.8368631601333618,
+      "learning_rate": 0.0001,
+      "loss": 1.09,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5305755395683454,
+      "grad_norm": 1.0364115238189697,
+      "learning_rate": 0.0001,
+      "loss": 1.3082,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5306611853374443,
+      "grad_norm": 0.7079994678497314,
+      "learning_rate": 0.0001,
+      "loss": 0.7345,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5307468311065433,
+      "grad_norm": 0.661487877368927,
+      "learning_rate": 0.0001,
+      "loss": 0.674,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5308324768756424,
+      "grad_norm": 1.0043047666549683,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5309181226447414,
+      "grad_norm": 0.6024109721183777,
+      "learning_rate": 0.0001,
+      "loss": 1.0812,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5310037684138403,
+      "grad_norm": 0.8442049622535706,
+      "learning_rate": 0.0001,
+      "loss": 0.9074,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5310894141829393,
+      "grad_norm": 1.174730658531189,
+      "learning_rate": 0.0001,
+      "loss": 0.758,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5311750599520384,
+      "grad_norm": 0.6045871376991272,
+      "learning_rate": 0.0001,
+      "loss": 0.272,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5312607057211374,
+      "grad_norm": 0.43512415885925293,
+      "learning_rate": 0.0001,
+      "loss": 0.2413,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5313463514902363,
+      "grad_norm": 0.24141569435596466,
+      "learning_rate": 0.0001,
+      "loss": 0.3355,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5314319972593354,
+      "grad_norm": 0.9774969816207886,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5315176430284344,
+      "grad_norm": 0.6235119104385376,
+      "learning_rate": 0.0001,
+      "loss": 0.7936,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5316032887975334,
+      "grad_norm": 0.6814229488372803,
+      "learning_rate": 0.0001,
+      "loss": 0.8926,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5316889345666324,
+      "grad_norm": 0.788509726524353,
+      "learning_rate": 0.0001,
+      "loss": 1.1053,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5317745803357314,
+      "grad_norm": 0.6833407282829285,
+      "learning_rate": 0.0001,
+      "loss": 0.4468,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5318602261048304,
+      "grad_norm": 0.5710582733154297,
+      "learning_rate": 0.0001,
+      "loss": 0.9037,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5319458718739294,
+      "grad_norm": 1.03251314163208,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5320315176430285,
+      "grad_norm": 0.1896325647830963,
+      "learning_rate": 0.0001,
+      "loss": 0.3849,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5321171634121274,
+      "grad_norm": 0.6570830345153809,
+      "learning_rate": 0.0001,
+      "loss": 0.6762,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5322028091812264,
+      "grad_norm": 0.7949010133743286,
+      "learning_rate": 0.0001,
+      "loss": 0.5282,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5322884549503255,
+      "grad_norm": 0.9441889524459839,
+      "learning_rate": 0.0001,
+      "loss": 1.0015,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5323741007194245,
+      "grad_norm": 0.6936191320419312,
+      "learning_rate": 0.0001,
+      "loss": 1.0405,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5324597464885235,
+      "grad_norm": 0.8208956122398376,
+      "learning_rate": 0.0001,
+      "loss": 0.8315,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5325453922576224,
+      "grad_norm": 0.8210596442222595,
+      "learning_rate": 0.0001,
+      "loss": 0.6055,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5326310380267215,
+      "grad_norm": 0.7527268528938293,
+      "learning_rate": 0.0001,
+      "loss": 1.3526,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5327166837958205,
+      "grad_norm": 0.3456796705722809,
+      "learning_rate": 0.0001,
+      "loss": 0.2179,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5328023295649195,
+      "grad_norm": 1.1349291801452637,
+      "learning_rate": 0.0001,
+      "loss": 1.1059,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5328879753340185,
+      "grad_norm": 1.0762274265289307,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5329736211031175,
+      "grad_norm": 0.34134146571159363,
+      "learning_rate": 0.0001,
+      "loss": 0.2688,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5330592668722165,
+      "grad_norm": 2.2490427494049072,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5331449126413155,
+      "grad_norm": 0.8678284287452698,
+      "learning_rate": 0.0001,
+      "loss": 1.0619,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5332305584104146,
+      "grad_norm": 0.7688842415809631,
+      "learning_rate": 0.0001,
+      "loss": 0.5086,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5333162041795135,
+      "grad_norm": 1.2643146514892578,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5334018499486125,
+      "grad_norm": 1.042758822441101,
+      "learning_rate": 0.0001,
+      "loss": 0.8318,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5334874957177116,
+      "grad_norm": 0.9063992500305176,
+      "learning_rate": 0.0001,
+      "loss": 0.5674,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5335731414868106,
+      "grad_norm": 1.0708907842636108,
+      "learning_rate": 0.0001,
+      "loss": 1.0766,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5336587872559095,
+      "grad_norm": 0.8735249638557434,
+      "learning_rate": 0.0001,
+      "loss": 1.0908,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5337444330250085,
+      "grad_norm": 0.9538981914520264,
+      "learning_rate": 0.0001,
+      "loss": 0.9843,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5338300787941076,
+      "grad_norm": 0.7851930856704712,
+      "learning_rate": 0.0001,
+      "loss": 0.9655,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5339157245632066,
+      "grad_norm": 0.9926151037216187,
+      "learning_rate": 0.0001,
+      "loss": 0.399,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5340013703323055,
+      "grad_norm": 1.0620639324188232,
+      "learning_rate": 0.0001,
+      "loss": 0.9906,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5340870161014046,
+      "grad_norm": 0.7699184417724609,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5341726618705036,
+      "grad_norm": 0.7604926824569702,
+      "learning_rate": 0.0001,
+      "loss": 0.9596,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5342583076396026,
+      "grad_norm": 0.32718244194984436,
+      "learning_rate": 0.0001,
+      "loss": 0.4273,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5343439534087016,
+      "grad_norm": 0.9129083156585693,
+      "learning_rate": 0.0001,
+      "loss": 1.1892,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5344295991778006,
+      "grad_norm": 0.7842795848846436,
+      "learning_rate": 0.0001,
+      "loss": 1.2972,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5345152449468996,
+      "grad_norm": 1.133847951889038,
+      "learning_rate": 0.0001,
+      "loss": 1.0157,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5346008907159986,
+      "grad_norm": 1.08368718624115,
+      "learning_rate": 0.0001,
+      "loss": 1.2053,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5346865364850977,
+      "grad_norm": 0.753576397895813,
+      "learning_rate": 0.0001,
+      "loss": 0.765,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5347721822541966,
+      "grad_norm": 0.9537203311920166,
+      "learning_rate": 0.0001,
+      "loss": 0.8756,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5348578280232956,
+      "grad_norm": 0.937401294708252,
+      "learning_rate": 0.0001,
+      "loss": 1.1105,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5349434737923947,
+      "grad_norm": 0.3398419916629791,
+      "learning_rate": 0.0001,
+      "loss": 0.3849,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5350291195614937,
+      "grad_norm": 0.9199445247650146,
+      "learning_rate": 0.0001,
+      "loss": 0.8521,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5351147653305927,
+      "grad_norm": 0.7681429982185364,
+      "learning_rate": 0.0001,
+      "loss": 1.2344,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5352004110996916,
+      "grad_norm": 0.2898103892803192,
+      "learning_rate": 0.0001,
+      "loss": 0.6288,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5352860568687907,
+      "grad_norm": 0.5399866104125977,
+      "learning_rate": 0.0001,
+      "loss": 0.734,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5353717026378897,
+      "grad_norm": 0.9556615352630615,
+      "learning_rate": 0.0001,
+      "loss": 0.7481,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5354573484069887,
+      "grad_norm": 0.934872567653656,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5355429941760877,
+      "grad_norm": 1.0027291774749756,
+      "learning_rate": 0.0001,
+      "loss": 0.539,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5356286399451867,
+      "grad_norm": 0.6300255656242371,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.8383276462554932,
+      "learning_rate": 0.0001,
+      "loss": 0.9858,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5357999314833847,
+      "grad_norm": 1.6010655164718628,
+      "learning_rate": 0.0001,
+      "loss": 1.3668,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5358855772524838,
+      "grad_norm": 0.6973920464515686,
+      "learning_rate": 0.0001,
+      "loss": 1.0644,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5359712230215827,
+      "grad_norm": 0.5502693057060242,
+      "learning_rate": 0.0001,
+      "loss": 0.9827,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5360568687906817,
+      "grad_norm": 0.6268861889839172,
+      "learning_rate": 0.0001,
+      "loss": 0.9703,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5361425145597808,
+      "grad_norm": 0.9796669483184814,
+      "learning_rate": 0.0001,
+      "loss": 0.6906,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5362281603288798,
+      "grad_norm": 0.6662825345993042,
+      "learning_rate": 0.0001,
+      "loss": 0.6771,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5363138060979787,
+      "grad_norm": 0.885465681552887,
+      "learning_rate": 0.0001,
+      "loss": 1.0532,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5363994518670777,
+      "grad_norm": 0.47403576970100403,
+      "learning_rate": 0.0001,
+      "loss": 1.0112,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5364850976361768,
+      "grad_norm": 1.2302337884902954,
+      "learning_rate": 0.0001,
+      "loss": 1.229,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5365707434052758,
+      "grad_norm": 0.38480138778686523,
+      "learning_rate": 0.0001,
+      "loss": 0.5475,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5366563891743747,
+      "grad_norm": 0.970120906829834,
+      "learning_rate": 0.0001,
+      "loss": 0.9716,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5367420349434738,
+      "grad_norm": 0.9712870121002197,
+      "learning_rate": 0.0001,
+      "loss": 1.2359,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5368276807125728,
+      "grad_norm": 0.9739351868629456,
+      "learning_rate": 0.0001,
+      "loss": 0.632,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5369133264816718,
+      "grad_norm": 0.7982573509216309,
+      "learning_rate": 0.0001,
+      "loss": 1.113,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5369989722507709,
+      "grad_norm": 0.7910007238388062,
+      "learning_rate": 0.0001,
+      "loss": 0.5376,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5370846180198698,
+      "grad_norm": 0.5094560384750366,
+      "learning_rate": 0.0001,
+      "loss": 0.6803,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5371702637889688,
+      "grad_norm": 0.5736317038536072,
+      "learning_rate": 0.0001,
+      "loss": 0.8975,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5372559095580678,
+      "grad_norm": 0.9039039611816406,
+      "learning_rate": 0.0001,
+      "loss": 0.6651,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5373415553271669,
+      "grad_norm": 0.8505333065986633,
+      "learning_rate": 0.0001,
+      "loss": 0.7673,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5374272010962658,
+      "grad_norm": 0.5798442363739014,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5375128468653648,
+      "grad_norm": 0.8711569309234619,
+      "learning_rate": 0.0001,
+      "loss": 0.9982,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5375984926344639,
+      "grad_norm": 0.3459269404411316,
+      "learning_rate": 0.0001,
+      "loss": 0.4508,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5376841384035629,
+      "grad_norm": 0.9019712805747986,
+      "learning_rate": 0.0001,
+      "loss": 1.1028,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5377697841726619,
+      "grad_norm": 0.8714526891708374,
+      "learning_rate": 0.0001,
+      "loss": 0.6431,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5378554299417608,
+      "grad_norm": 0.6949652433395386,
+      "learning_rate": 0.0001,
+      "loss": 1.934,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5379410757108599,
+      "grad_norm": 0.9579134583473206,
+      "learning_rate": 0.0001,
+      "loss": 1.1072,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5380267214799589,
+      "grad_norm": 0.9410560727119446,
+      "learning_rate": 0.0001,
+      "loss": 0.6763,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5381123672490579,
+      "grad_norm": 0.7835625410079956,
+      "learning_rate": 0.0001,
+      "loss": 0.814,
+      "step": 6283
+    },
+    {
+      "epoch": 0.538198013018157,
+      "grad_norm": 0.6139087080955505,
+      "learning_rate": 0.0001,
+      "loss": 1.2929,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5382836587872559,
+      "grad_norm": 0.7231254577636719,
+      "learning_rate": 0.0001,
+      "loss": 0.6804,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5383693045563549,
+      "grad_norm": 0.8783027529716492,
+      "learning_rate": 0.0001,
+      "loss": 1.0485,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5384549503254539,
+      "grad_norm": 0.7183824777603149,
+      "learning_rate": 0.0001,
+      "loss": 1.1665,
+      "step": 6287
+    },
+    {
+      "epoch": 0.538540596094553,
+      "grad_norm": 0.6977509260177612,
+      "learning_rate": 0.0001,
+      "loss": 0.6888,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5386262418636519,
+      "grad_norm": 0.8218511343002319,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5387118876327509,
+      "grad_norm": 0.8916334509849548,
+      "learning_rate": 0.0001,
+      "loss": 0.7989,
+      "step": 6290
+    },
+    {
+      "epoch": 0.53879753340185,
+      "grad_norm": 0.5729463696479797,
+      "learning_rate": 0.0001,
+      "loss": 1.0373,
+      "step": 6291
+    },
+    {
+      "epoch": 0.538883179170949,
+      "grad_norm": 0.8708328008651733,
+      "learning_rate": 0.0001,
+      "loss": 0.8635,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5389688249400479,
+      "grad_norm": 0.4861520826816559,
+      "learning_rate": 0.0001,
+      "loss": 0.4772,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5390544707091469,
+      "grad_norm": 0.668601930141449,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 6294
+    },
+    {
+      "epoch": 0.539140116478246,
+      "grad_norm": 0.6588530540466309,
+      "learning_rate": 0.0001,
+      "loss": 0.8488,
+      "step": 6295
+    },
+    {
+      "epoch": 0.539225762247345,
+      "grad_norm": 0.9198886156082153,
+      "learning_rate": 0.0001,
+      "loss": 1.2372,
+      "step": 6296
+    },
+    {
+      "epoch": 0.539311408016444,
+      "grad_norm": 0.5690991878509521,
+      "learning_rate": 0.0001,
+      "loss": 0.5809,
+      "step": 6297
+    },
+    {
+      "epoch": 0.539397053785543,
+      "grad_norm": 0.7045687437057495,
+      "learning_rate": 0.0001,
+      "loss": 0.9146,
+      "step": 6298
+    },
+    {
+      "epoch": 0.539482699554642,
+      "grad_norm": 0.7011622190475464,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 6299
+    },
+    {
+      "epoch": 0.539568345323741,
+      "grad_norm": 0.6996843218803406,
+      "learning_rate": 0.0001,
+      "loss": 0.9948,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5396539910928401,
+      "grad_norm": 0.602756679058075,
+      "learning_rate": 0.0001,
+      "loss": 0.4766,
+      "step": 6301
+    },
+    {
+      "epoch": 0.539739636861939,
+      "grad_norm": 0.8476176857948303,
+      "learning_rate": 0.0001,
+      "loss": 0.7951,
+      "step": 6302
+    },
+    {
+      "epoch": 0.539825282631038,
+      "grad_norm": 1.3042851686477661,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 6303
+    },
+    {
+      "epoch": 0.539910928400137,
+      "grad_norm": 0.5598807334899902,
+      "learning_rate": 0.0001,
+      "loss": 0.899,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5399965741692361,
+      "grad_norm": 0.704115629196167,
+      "learning_rate": 0.0001,
+      "loss": 1.0278,
+      "step": 6305
+    },
+    {
+      "epoch": 0.540082219938335,
+      "grad_norm": 0.8402760028839111,
+      "learning_rate": 0.0001,
+      "loss": 0.9329,
+      "step": 6306
+    },
+    {
+      "epoch": 0.540167865707434,
+      "grad_norm": 0.5876890420913696,
+      "learning_rate": 0.0001,
+      "loss": 0.6597,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5402535114765331,
+      "grad_norm": 0.9252611398696899,
+      "learning_rate": 0.0001,
+      "loss": 0.8558,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5403391572456321,
+      "grad_norm": 0.8542856574058533,
+      "learning_rate": 0.0001,
+      "loss": 1.3684,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5404248030147311,
+      "grad_norm": 1.361355185508728,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 6310
+    },
+    {
+      "epoch": 0.54051044878383,
+      "grad_norm": 0.7852384448051453,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5405960945529291,
+      "grad_norm": 0.9107966423034668,
+      "learning_rate": 0.0001,
+      "loss": 0.7557,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5406817403220281,
+      "grad_norm": 0.8110779523849487,
+      "learning_rate": 0.0001,
+      "loss": 0.9571,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5407673860911271,
+      "grad_norm": 1.1439085006713867,
+      "learning_rate": 0.0001,
+      "loss": 1.1551,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5408530318602262,
+      "grad_norm": 0.6568119525909424,
+      "learning_rate": 0.0001,
+      "loss": 1.061,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5409386776293251,
+      "grad_norm": 0.8914629220962524,
+      "learning_rate": 0.0001,
+      "loss": 0.6374,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5410243233984241,
+      "grad_norm": 1.266400933265686,
+      "learning_rate": 0.0001,
+      "loss": 0.9491,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5411099691675231,
+      "grad_norm": 1.0814120769500732,
+      "learning_rate": 0.0001,
+      "loss": 1.1032,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5411956149366222,
+      "grad_norm": 0.5667025446891785,
+      "learning_rate": 0.0001,
+      "loss": 1.0245,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5412812607057211,
+      "grad_norm": 1.1575380563735962,
+      "learning_rate": 0.0001,
+      "loss": 1.2718,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5413669064748201,
+      "grad_norm": 1.0784029960632324,
+      "learning_rate": 0.0001,
+      "loss": 0.9979,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5414525522439192,
+      "grad_norm": 0.7050232291221619,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5415381980130182,
+      "grad_norm": 1.0877654552459717,
+      "learning_rate": 0.0001,
+      "loss": 1.1121,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5416238437821171,
+      "grad_norm": 1.2125033140182495,
+      "learning_rate": 0.0001,
+      "loss": 1.8833,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5417094895512161,
+      "grad_norm": 1.2327327728271484,
+      "learning_rate": 0.0001,
+      "loss": 1.2967,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5417951353203152,
+      "grad_norm": 0.8374308943748474,
+      "learning_rate": 0.0001,
+      "loss": 0.6454,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5418807810894142,
+      "grad_norm": 1.075110912322998,
+      "learning_rate": 0.0001,
+      "loss": 0.8901,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5419664268585132,
+      "grad_norm": 1.0262956619262695,
+      "learning_rate": 0.0001,
+      "loss": 0.6157,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5420520726276122,
+      "grad_norm": 0.36954164505004883,
+      "learning_rate": 0.0001,
+      "loss": 0.1694,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5421377183967112,
+      "grad_norm": 0.9314480423927307,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5422233641658102,
+      "grad_norm": 0.7536929249763489,
+      "learning_rate": 0.0001,
+      "loss": 1.182,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5423090099349093,
+      "grad_norm": 0.6530731320381165,
+      "learning_rate": 0.0001,
+      "loss": 1.0976,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5423946557040082,
+      "grad_norm": 0.9641848206520081,
+      "learning_rate": 0.0001,
+      "loss": 0.9839,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5424803014731072,
+      "grad_norm": 0.6719341278076172,
+      "learning_rate": 0.0001,
+      "loss": 1.2408,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5425659472422062,
+      "grad_norm": 0.7267654538154602,
+      "learning_rate": 0.0001,
+      "loss": 0.9591,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5426515930113053,
+      "grad_norm": 0.8339561820030212,
+      "learning_rate": 0.0001,
+      "loss": 1.1621,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5427372387804043,
+      "grad_norm": 0.8477576375007629,
+      "learning_rate": 0.0001,
+      "loss": 0.8741,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5428228845495032,
+      "grad_norm": 0.6448758244514465,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5429085303186023,
+      "grad_norm": 0.778948187828064,
+      "learning_rate": 0.0001,
+      "loss": 0.9687,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5429941760877013,
+      "grad_norm": 0.5818303823471069,
+      "learning_rate": 0.0001,
+      "loss": 0.9356,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5430798218568003,
+      "grad_norm": 0.8526086807250977,
+      "learning_rate": 0.0001,
+      "loss": 0.668,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5431654676258992,
+      "grad_norm": 0.9976934194564819,
+      "learning_rate": 0.0001,
+      "loss": 0.9684,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5432511133949983,
+      "grad_norm": 0.6510010957717896,
+      "learning_rate": 0.0001,
+      "loss": 1.0816,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5433367591640973,
+      "grad_norm": 1.0828437805175781,
+      "learning_rate": 0.0001,
+      "loss": 1.0223,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5434224049331963,
+      "grad_norm": 0.5642203092575073,
+      "learning_rate": 0.0001,
+      "loss": 0.974,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5435080507022954,
+      "grad_norm": 1.0965496301651,
+      "learning_rate": 0.0001,
+      "loss": 1.1129,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5435936964713943,
+      "grad_norm": 1.0925159454345703,
+      "learning_rate": 0.0001,
+      "loss": 1.3901,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5436793422404933,
+      "grad_norm": 0.7794240713119507,
+      "learning_rate": 0.0001,
+      "loss": 0.7,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5437649880095923,
+      "grad_norm": 0.7849947214126587,
+      "learning_rate": 0.0001,
+      "loss": 1.3766,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5438506337786914,
+      "grad_norm": 0.8455568552017212,
+      "learning_rate": 0.0001,
+      "loss": 0.7272,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5439362795477903,
+      "grad_norm": 0.6965721249580383,
+      "learning_rate": 0.0001,
+      "loss": 0.6508,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5440219253168893,
+      "grad_norm": 1.0255223512649536,
+      "learning_rate": 0.0001,
+      "loss": 0.5922,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5441075710859884,
+      "grad_norm": 0.9161310791969299,
+      "learning_rate": 0.0001,
+      "loss": 1.2074,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5441932168550874,
+      "grad_norm": 0.9254721999168396,
+      "learning_rate": 0.0001,
+      "loss": 0.8557,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5442788626241863,
+      "grad_norm": 0.9302588701248169,
+      "learning_rate": 0.0001,
+      "loss": 0.8595,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5443645083932853,
+      "grad_norm": 0.8771781325340271,
+      "learning_rate": 0.0001,
+      "loss": 1.0169,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5444501541623844,
+      "grad_norm": 1.0721633434295654,
+      "learning_rate": 0.0001,
+      "loss": 1.1013,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5445357999314834,
+      "grad_norm": 1.2802969217300415,
+      "learning_rate": 0.0001,
+      "loss": 0.8672,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5446214457005824,
+      "grad_norm": 1.1304607391357422,
+      "learning_rate": 0.0001,
+      "loss": 0.8666,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5447070914696814,
+      "grad_norm": 0.6544949412345886,
+      "learning_rate": 0.0001,
+      "loss": 1.1355,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5447927372387804,
+      "grad_norm": 2.0387136936187744,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5448783830078794,
+      "grad_norm": 0.7772772908210754,
+      "learning_rate": 0.0001,
+      "loss": 0.7439,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5449640287769785,
+      "grad_norm": 0.9238820672035217,
+      "learning_rate": 0.0001,
+      "loss": 0.6251,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5450496745460774,
+      "grad_norm": 0.7720708847045898,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5451353203151764,
+      "grad_norm": 0.5969306230545044,
+      "learning_rate": 0.0001,
+      "loss": 0.9011,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5452209660842754,
+      "grad_norm": 1.2798914909362793,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5453066118533745,
+      "grad_norm": 0.6582180857658386,
+      "learning_rate": 0.0001,
+      "loss": 1.0772,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5453922576224735,
+      "grad_norm": 0.8136601448059082,
+      "learning_rate": 0.0001,
+      "loss": 0.7335,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5454779033915724,
+      "grad_norm": 0.723143458366394,
+      "learning_rate": 0.0001,
+      "loss": 0.9701,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5455635491606715,
+      "grad_norm": 0.9596467614173889,
+      "learning_rate": 0.0001,
+      "loss": 0.6654,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5456491949297705,
+      "grad_norm": 0.837440550327301,
+      "learning_rate": 0.0001,
+      "loss": 1.2459,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5457348406988695,
+      "grad_norm": 0.6952359080314636,
+      "learning_rate": 0.0001,
+      "loss": 1.085,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5458204864679684,
+      "grad_norm": 0.8808367848396301,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5459061322370675,
+      "grad_norm": 1.256890058517456,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5459917780061665,
+      "grad_norm": 0.5811572670936584,
+      "learning_rate": 0.0001,
+      "loss": 0.8476,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5460774237752655,
+      "grad_norm": 0.7938216328620911,
+      "learning_rate": 0.0001,
+      "loss": 1.2643,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5461630695443646,
+      "grad_norm": 0.42881450057029724,
+      "learning_rate": 0.0001,
+      "loss": 0.4684,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5462487153134635,
+      "grad_norm": 0.6096703410148621,
+      "learning_rate": 0.0001,
+      "loss": 0.6256,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5463343610825625,
+      "grad_norm": 1.1100718975067139,
+      "learning_rate": 0.0001,
+      "loss": 0.7629,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5464200068516615,
+      "grad_norm": 1.1198110580444336,
+      "learning_rate": 0.0001,
+      "loss": 1.1358,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5465056526207606,
+      "grad_norm": 0.7284631729125977,
+      "learning_rate": 0.0001,
+      "loss": 1.0276,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5465912983898595,
+      "grad_norm": 0.6708879470825195,
+      "learning_rate": 0.0001,
+      "loss": 0.6086,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5466769441589585,
+      "grad_norm": 0.793653130531311,
+      "learning_rate": 0.0001,
+      "loss": 0.86,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5467625899280576,
+      "grad_norm": 0.6552831530570984,
+      "learning_rate": 0.0001,
+      "loss": 0.9133,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5468482356971566,
+      "grad_norm": 0.9513140916824341,
+      "learning_rate": 0.0001,
+      "loss": 1.1521,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5469338814662555,
+      "grad_norm": 0.7006797790527344,
+      "learning_rate": 0.0001,
+      "loss": 0.7567,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5470195272353545,
+      "grad_norm": 0.7423794269561768,
+      "learning_rate": 0.0001,
+      "loss": 0.8335,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5471051730044536,
+      "grad_norm": 0.8612438440322876,
+      "learning_rate": 0.0001,
+      "loss": 0.8558,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5471908187735526,
+      "grad_norm": 0.8051362037658691,
+      "learning_rate": 0.0001,
+      "loss": 1.2707,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5472764645426516,
+      "grad_norm": 0.9603021144866943,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5473621103117506,
+      "grad_norm": 0.7870426774024963,
+      "learning_rate": 0.0001,
+      "loss": 0.7857,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5474477560808496,
+      "grad_norm": 1.2743629217147827,
+      "learning_rate": 0.0001,
+      "loss": 1.0419,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5475334018499486,
+      "grad_norm": 0.8625341653823853,
+      "learning_rate": 0.0001,
+      "loss": 1.2503,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5476190476190477,
+      "grad_norm": 0.5734780430793762,
+      "learning_rate": 0.0001,
+      "loss": 0.433,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5477046933881466,
+      "grad_norm": 0.6672755479812622,
+      "learning_rate": 0.0001,
+      "loss": 0.6518,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5477903391572456,
+      "grad_norm": 1.3684619665145874,
+      "learning_rate": 0.0001,
+      "loss": 1.7727,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5478759849263446,
+      "grad_norm": 1.3285044431686401,
+      "learning_rate": 0.0001,
+      "loss": 1.036,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5479616306954437,
+      "grad_norm": 0.8018969893455505,
+      "learning_rate": 0.0001,
+      "loss": 0.5908,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5480472764645427,
+      "grad_norm": 0.9651978611946106,
+      "learning_rate": 0.0001,
+      "loss": 0.9652,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5481329222336416,
+      "grad_norm": 0.36337679624557495,
+      "learning_rate": 0.0001,
+      "loss": 0.5313,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5482185680027407,
+      "grad_norm": 0.5836355686187744,
+      "learning_rate": 0.0001,
+      "loss": 1.1781,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5483042137718397,
+      "grad_norm": 0.5113340020179749,
+      "learning_rate": 0.0001,
+      "loss": 0.6533,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5483898595409387,
+      "grad_norm": 0.37029701471328735,
+      "learning_rate": 0.0001,
+      "loss": 0.1487,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5484755053100376,
+      "grad_norm": 0.6826488375663757,
+      "learning_rate": 0.0001,
+      "loss": 0.7598,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5485611510791367,
+      "grad_norm": 0.7648865580558777,
+      "learning_rate": 0.0001,
+      "loss": 1.0048,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5486467968482357,
+      "grad_norm": 1.0607887506484985,
+      "learning_rate": 0.0001,
+      "loss": 1.8727,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5487324426173347,
+      "grad_norm": 1.0197746753692627,
+      "learning_rate": 0.0001,
+      "loss": 1.067,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5488180883864338,
+      "grad_norm": 0.6069988012313843,
+      "learning_rate": 0.0001,
+      "loss": 0.5316,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5489037341555327,
+      "grad_norm": 0.6975987553596497,
+      "learning_rate": 0.0001,
+      "loss": 1.2634,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5489893799246317,
+      "grad_norm": 0.5586996674537659,
+      "learning_rate": 0.0001,
+      "loss": 0.9819,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5490750256937307,
+      "grad_norm": 0.5114778876304626,
+      "learning_rate": 0.0001,
+      "loss": 0.9307,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5491606714628298,
+      "grad_norm": 0.596436083316803,
+      "learning_rate": 0.0001,
+      "loss": 1.0881,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5492463172319287,
+      "grad_norm": 1.2005715370178223,
+      "learning_rate": 0.0001,
+      "loss": 1.2261,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5493319630010277,
+      "grad_norm": 0.8865706324577332,
+      "learning_rate": 0.0001,
+      "loss": 0.7543,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5494176087701268,
+      "grad_norm": 1.0031697750091553,
+      "learning_rate": 0.0001,
+      "loss": 0.804,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5495032545392258,
+      "grad_norm": 0.678606390953064,
+      "learning_rate": 0.0001,
+      "loss": 0.4514,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5495889003083247,
+      "grad_norm": 0.7665726542472839,
+      "learning_rate": 0.0001,
+      "loss": 0.7748,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5496745460774237,
+      "grad_norm": 1.229155421257019,
+      "learning_rate": 0.0001,
+      "loss": 0.7313,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5497601918465228,
+      "grad_norm": 0.7102965712547302,
+      "learning_rate": 0.0001,
+      "loss": 0.8187,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5498458376156218,
+      "grad_norm": 0.6390897035598755,
+      "learning_rate": 0.0001,
+      "loss": 0.961,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5499314833847208,
+      "grad_norm": 0.8443874716758728,
+      "learning_rate": 0.0001,
+      "loss": 0.9499,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5500171291538198,
+      "grad_norm": 0.8043901324272156,
+      "learning_rate": 0.0001,
+      "loss": 0.6499,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5501027749229188,
+      "grad_norm": 0.862535297870636,
+      "learning_rate": 0.0001,
+      "loss": 1.0325,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5501884206920178,
+      "grad_norm": 0.9152210354804993,
+      "learning_rate": 0.0001,
+      "loss": 1.3122,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5502740664611169,
+      "grad_norm": 0.8350108861923218,
+      "learning_rate": 0.0001,
+      "loss": 1.0137,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5503597122302158,
+      "grad_norm": 1.1895016431808472,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5504453579993148,
+      "grad_norm": 0.6964080929756165,
+      "learning_rate": 0.0001,
+      "loss": 0.5796,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5505310037684138,
+      "grad_norm": 0.764220118522644,
+      "learning_rate": 0.0001,
+      "loss": 1.0497,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5506166495375129,
+      "grad_norm": 1.10202956199646,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5507022953066119,
+      "grad_norm": 0.6813096404075623,
+      "learning_rate": 0.0001,
+      "loss": 0.7335,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5507879410757108,
+      "grad_norm": 0.756445050239563,
+      "learning_rate": 0.0001,
+      "loss": 0.4275,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5508735868448099,
+      "grad_norm": 0.6755313873291016,
+      "learning_rate": 0.0001,
+      "loss": 1.1966,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5509592326139089,
+      "grad_norm": 0.894339919090271,
+      "learning_rate": 0.0001,
+      "loss": 1.2267,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5510448783830079,
+      "grad_norm": 0.5435007810592651,
+      "learning_rate": 0.0001,
+      "loss": 0.642,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5511305241521068,
+      "grad_norm": 0.7071726322174072,
+      "learning_rate": 0.0001,
+      "loss": 0.636,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5512161699212059,
+      "grad_norm": 0.6490583419799805,
+      "learning_rate": 0.0001,
+      "loss": 1.0427,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5513018156903049,
+      "grad_norm": 0.8256914019584656,
+      "learning_rate": 0.0001,
+      "loss": 0.5653,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5513874614594039,
+      "grad_norm": 0.5307016968727112,
+      "learning_rate": 0.0001,
+      "loss": 0.6675,
+      "step": 6438
+    },
+    {
+      "epoch": 0.551473107228503,
+      "grad_norm": 0.2755190432071686,
+      "learning_rate": 0.0001,
+      "loss": 0.3078,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5515587529976019,
+      "grad_norm": 0.552716076374054,
+      "learning_rate": 0.0001,
+      "loss": 0.5211,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5516443987667009,
+      "grad_norm": 0.7615090012550354,
+      "learning_rate": 0.0001,
+      "loss": 0.7451,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5517300445357999,
+      "grad_norm": 0.7170068621635437,
+      "learning_rate": 0.0001,
+      "loss": 0.6424,
+      "step": 6442
+    },
+    {
+      "epoch": 0.551815690304899,
+      "grad_norm": 0.9502476453781128,
+      "learning_rate": 0.0001,
+      "loss": 1.0451,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5519013360739979,
+      "grad_norm": 0.8709639310836792,
+      "learning_rate": 0.0001,
+      "loss": 1.0715,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5519869818430969,
+      "grad_norm": 0.31799885630607605,
+      "learning_rate": 0.0001,
+      "loss": 0.4718,
+      "step": 6445
+    },
+    {
+      "epoch": 0.552072627612196,
+      "grad_norm": 1.043389081954956,
+      "learning_rate": 0.0001,
+      "loss": 1.2565,
+      "step": 6446
+    },
+    {
+      "epoch": 0.552158273381295,
+      "grad_norm": 0.8447428345680237,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 6447
+    },
+    {
+      "epoch": 0.552243919150394,
+      "grad_norm": 0.7170526385307312,
+      "learning_rate": 0.0001,
+      "loss": 0.8906,
+      "step": 6448
+    },
+    {
+      "epoch": 0.552329564919493,
+      "grad_norm": 0.598723828792572,
+      "learning_rate": 0.0001,
+      "loss": 1.1351,
+      "step": 6449
+    },
+    {
+      "epoch": 0.552415210688592,
+      "grad_norm": 0.8403136134147644,
+      "learning_rate": 0.0001,
+      "loss": 0.939,
+      "step": 6450
+    },
+    {
+      "epoch": 0.552500856457691,
+      "grad_norm": 1.1694540977478027,
+      "learning_rate": 0.0001,
+      "loss": 1.1203,
+      "step": 6451
+    },
+    {
+      "epoch": 0.55258650222679,
+      "grad_norm": 0.8605048060417175,
+      "learning_rate": 0.0001,
+      "loss": 0.6447,
+      "step": 6452
+    },
+    {
+      "epoch": 0.552672147995889,
+      "grad_norm": 0.6793586611747742,
+      "learning_rate": 0.0001,
+      "loss": 0.9577,
+      "step": 6453
+    },
+    {
+      "epoch": 0.552757793764988,
+      "grad_norm": 0.9546366333961487,
+      "learning_rate": 0.0001,
+      "loss": 0.7529,
+      "step": 6454
+    },
+    {
+      "epoch": 0.552843439534087,
+      "grad_norm": 0.8330127000808716,
+      "learning_rate": 0.0001,
+      "loss": 0.7922,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5529290853031861,
+      "grad_norm": 1.2231162786483765,
+      "learning_rate": 0.0001,
+      "loss": 1.3245,
+      "step": 6456
+    },
+    {
+      "epoch": 0.553014731072285,
+      "grad_norm": 1.2937203645706177,
+      "learning_rate": 0.0001,
+      "loss": 1.0387,
+      "step": 6457
+    },
+    {
+      "epoch": 0.553100376841384,
+      "grad_norm": 0.7726119160652161,
+      "learning_rate": 0.0001,
+      "loss": 0.6449,
+      "step": 6458
+    },
+    {
+      "epoch": 0.553186022610483,
+      "grad_norm": 0.7756932973861694,
+      "learning_rate": 0.0001,
+      "loss": 1.0596,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5532716683795821,
+      "grad_norm": 0.951482355594635,
+      "learning_rate": 0.0001,
+      "loss": 0.9969,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5533573141486811,
+      "grad_norm": 1.192322850227356,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 6461
+    },
+    {
+      "epoch": 0.55344295991778,
+      "grad_norm": 0.9233120679855347,
+      "learning_rate": 0.0001,
+      "loss": 0.7214,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5535286056868791,
+      "grad_norm": 0.5903610587120056,
+      "learning_rate": 0.0001,
+      "loss": 1.1878,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5536142514559781,
+      "grad_norm": 0.5823624730110168,
+      "learning_rate": 0.0001,
+      "loss": 0.9198,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5536998972250771,
+      "grad_norm": 0.6103382110595703,
+      "learning_rate": 0.0001,
+      "loss": 0.6173,
+      "step": 6465
+    },
+    {
+      "epoch": 0.553785542994176,
+      "grad_norm": 0.9364888668060303,
+      "learning_rate": 0.0001,
+      "loss": 0.8955,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5538711887632751,
+      "grad_norm": 0.7030479311943054,
+      "learning_rate": 0.0001,
+      "loss": 1.1034,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5539568345323741,
+      "grad_norm": 0.7820113897323608,
+      "learning_rate": 0.0001,
+      "loss": 0.5182,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5540424803014731,
+      "grad_norm": 1.0140708684921265,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5541281260705722,
+      "grad_norm": 0.676881730556488,
+      "learning_rate": 0.0001,
+      "loss": 0.8373,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5542137718396711,
+      "grad_norm": 0.9701757431030273,
+      "learning_rate": 0.0001,
+      "loss": 1.0504,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5542994176087701,
+      "grad_norm": 0.9054971933364868,
+      "learning_rate": 0.0001,
+      "loss": 1.1014,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5543850633778691,
+      "grad_norm": 1.3390227556228638,
+      "learning_rate": 0.0001,
+      "loss": 2.0842,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5544707091469682,
+      "grad_norm": 1.5130199193954468,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5545563549160671,
+      "grad_norm": 1.2085517644882202,
+      "learning_rate": 0.0001,
+      "loss": 1.1888,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5546420006851661,
+      "grad_norm": 0.49603399634361267,
+      "learning_rate": 0.0001,
+      "loss": 0.4503,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5547276464542652,
+      "grad_norm": 0.7969306707382202,
+      "learning_rate": 0.0001,
+      "loss": 0.9791,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5548132922233642,
+      "grad_norm": 0.35986578464508057,
+      "learning_rate": 0.0001,
+      "loss": 0.1238,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5548989379924631,
+      "grad_norm": 0.4883843958377838,
+      "learning_rate": 0.0001,
+      "loss": 0.4771,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5549845837615622,
+      "grad_norm": 0.8279018402099609,
+      "learning_rate": 0.0001,
+      "loss": 0.9384,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5550702295306612,
+      "grad_norm": 0.9071462750434875,
+      "learning_rate": 0.0001,
+      "loss": 0.8184,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5551558752997602,
+      "grad_norm": 1.114243984222412,
+      "learning_rate": 0.0001,
+      "loss": 1.1941,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5552415210688592,
+      "grad_norm": 0.8429853916168213,
+      "learning_rate": 0.0001,
+      "loss": 1.0866,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5553271668379582,
+      "grad_norm": 0.9152540564537048,
+      "learning_rate": 0.0001,
+      "loss": 0.8678,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5554128126070572,
+      "grad_norm": 1.2196807861328125,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5554984583761562,
+      "grad_norm": 0.9653230309486389,
+      "learning_rate": 0.0001,
+      "loss": 1.0924,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5555841041452553,
+      "grad_norm": 0.8747971057891846,
+      "learning_rate": 0.0001,
+      "loss": 0.7177,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5556697499143542,
+      "grad_norm": 0.877303957939148,
+      "learning_rate": 0.0001,
+      "loss": 0.5658,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5557553956834532,
+      "grad_norm": 0.6857127547264099,
+      "learning_rate": 0.0001,
+      "loss": 0.6156,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5558410414525522,
+      "grad_norm": 1.0107793807983398,
+      "learning_rate": 0.0001,
+      "loss": 0.9612,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5559266872216513,
+      "grad_norm": 0.5263829231262207,
+      "learning_rate": 0.0001,
+      "loss": 0.7271,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5560123329907503,
+      "grad_norm": 1.3715711832046509,
+      "learning_rate": 0.0001,
+      "loss": 1.3804,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5560979787598492,
+      "grad_norm": 0.8305419683456421,
+      "learning_rate": 0.0001,
+      "loss": 1.0255,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5561836245289483,
+      "grad_norm": 0.7396612167358398,
+      "learning_rate": 0.0001,
+      "loss": 0.5869,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5562692702980473,
+      "grad_norm": 1.0425032377243042,
+      "learning_rate": 0.0001,
+      "loss": 0.6889,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5563549160671463,
+      "grad_norm": 0.9543383717536926,
+      "learning_rate": 0.0001,
+      "loss": 1.3388,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5564405618362452,
+      "grad_norm": 1.0920405387878418,
+      "learning_rate": 0.0001,
+      "loss": 1.1815,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5565262076053443,
+      "grad_norm": 1.0351366996765137,
+      "learning_rate": 0.0001,
+      "loss": 1.3252,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5566118533744433,
+      "grad_norm": 0.8790441155433655,
+      "learning_rate": 0.0001,
+      "loss": 0.8532,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5566974991435423,
+      "grad_norm": 1.3378087282180786,
+      "learning_rate": 0.0001,
+      "loss": 0.7171,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5567831449126414,
+      "grad_norm": 1.0298430919647217,
+      "learning_rate": 0.0001,
+      "loss": 0.3676,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5568687906817403,
+      "grad_norm": 0.5512800812721252,
+      "learning_rate": 0.0001,
+      "loss": 0.428,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5569544364508393,
+      "grad_norm": 0.8122586011886597,
+      "learning_rate": 0.0001,
+      "loss": 1.0488,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5570400822199383,
+      "grad_norm": 1.1016842126846313,
+      "learning_rate": 0.0001,
+      "loss": 1.1905,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5571257279890374,
+      "grad_norm": 0.873515784740448,
+      "learning_rate": 0.0001,
+      "loss": 0.9848,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5572113737581363,
+      "grad_norm": 0.741508960723877,
+      "learning_rate": 0.0001,
+      "loss": 1.1439,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5572970195272353,
+      "grad_norm": 0.8743225336074829,
+      "learning_rate": 0.0001,
+      "loss": 0.925,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5573826652963344,
+      "grad_norm": 1.1357065439224243,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5574683110654334,
+      "grad_norm": 0.41512542963027954,
+      "learning_rate": 0.0001,
+      "loss": 0.372,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5575539568345323,
+      "grad_norm": 0.7484045624732971,
+      "learning_rate": 0.0001,
+      "loss": 0.7581,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5576396026036314,
+      "grad_norm": 1.0769673585891724,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5577252483727304,
+      "grad_norm": 0.7488591074943542,
+      "learning_rate": 0.0001,
+      "loss": 1.1144,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5578108941418294,
+      "grad_norm": 0.9005342125892639,
+      "learning_rate": 0.0001,
+      "loss": 1.0047,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5578965399109284,
+      "grad_norm": 0.5469145178794861,
+      "learning_rate": 0.0001,
+      "loss": 0.9318,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5579821856800274,
+      "grad_norm": 0.36950987577438354,
+      "learning_rate": 0.0001,
+      "loss": 0.2208,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5580678314491264,
+      "grad_norm": 0.7242395281791687,
+      "learning_rate": 0.0001,
+      "loss": 1.2739,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5581534772182254,
+      "grad_norm": 1.1185740232467651,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5582391229873245,
+      "grad_norm": 0.5193781852722168,
+      "learning_rate": 0.0001,
+      "loss": 1.0303,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5583247687564235,
+      "grad_norm": 0.5836544036865234,
+      "learning_rate": 0.0001,
+      "loss": 0.984,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5584104145255224,
+      "grad_norm": 0.6267916560173035,
+      "learning_rate": 0.0001,
+      "loss": 0.7314,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5584960602946214,
+      "grad_norm": 0.85222989320755,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5585817060637205,
+      "grad_norm": 0.9715830683708191,
+      "learning_rate": 0.0001,
+      "loss": 0.749,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5586673518328195,
+      "grad_norm": 0.9463797211647034,
+      "learning_rate": 0.0001,
+      "loss": 1.2274,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5587529976019184,
+      "grad_norm": 0.9598461389541626,
+      "learning_rate": 0.0001,
+      "loss": 0.9272,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5588386433710175,
+      "grad_norm": 0.7341274619102478,
+      "learning_rate": 0.0001,
+      "loss": 0.6887,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5589242891401165,
+      "grad_norm": 0.8349555730819702,
+      "learning_rate": 0.0001,
+      "loss": 0.8476,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5590099349092155,
+      "grad_norm": 0.6488535404205322,
+      "learning_rate": 0.0001,
+      "loss": 1.0009,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5590955806783144,
+      "grad_norm": 0.9305251240730286,
+      "learning_rate": 0.0001,
+      "loss": 1.3255,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5591812264474135,
+      "grad_norm": 0.5543324947357178,
+      "learning_rate": 0.0001,
+      "loss": 0.8988,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5592668722165125,
+      "grad_norm": 0.9162471294403076,
+      "learning_rate": 0.0001,
+      "loss": 0.4389,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5593525179856115,
+      "grad_norm": 0.5559175610542297,
+      "learning_rate": 0.0001,
+      "loss": 0.6168,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5594381637547106,
+      "grad_norm": 0.6011418700218201,
+      "learning_rate": 0.0001,
+      "loss": 0.9295,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5595238095238095,
+      "grad_norm": 1.0374879837036133,
+      "learning_rate": 0.0001,
+      "loss": 0.9243,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5596094552929085,
+      "grad_norm": 0.8783213496208191,
+      "learning_rate": 0.0001,
+      "loss": 0.8574,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5596951010620075,
+      "grad_norm": 1.1644067764282227,
+      "learning_rate": 0.0001,
+      "loss": 1.033,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5597807468311066,
+      "grad_norm": 1.0933395624160767,
+      "learning_rate": 0.0001,
+      "loss": 1.1978,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5598663926002055,
+      "grad_norm": 1.2957764863967896,
+      "learning_rate": 0.0001,
+      "loss": 1.1835,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5599520383693045,
+      "grad_norm": 0.6097084283828735,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5600376841384036,
+      "grad_norm": 0.615729570388794,
+      "learning_rate": 0.0001,
+      "loss": 0.9264,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5601233299075026,
+      "grad_norm": 0.9703981876373291,
+      "learning_rate": 0.0001,
+      "loss": 1.1092,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5602089756766016,
+      "grad_norm": 0.8293263912200928,
+      "learning_rate": 0.0001,
+      "loss": 1.0059,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5602946214457006,
+      "grad_norm": 0.8499071598052979,
+      "learning_rate": 0.0001,
+      "loss": 1.2088,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5603802672147996,
+      "grad_norm": 0.5335692167282104,
+      "learning_rate": 0.0001,
+      "loss": 0.4495,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5604659129838986,
+      "grad_norm": 0.7588483095169067,
+      "learning_rate": 0.0001,
+      "loss": 1.0935,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5605515587529976,
+      "grad_norm": 0.5301291942596436,
+      "learning_rate": 0.0001,
+      "loss": 0.2355,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5606372045220966,
+      "grad_norm": 0.9795524477958679,
+      "learning_rate": 0.0001,
+      "loss": 0.9523,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5607228502911956,
+      "grad_norm": 0.7576870918273926,
+      "learning_rate": 0.0001,
+      "loss": 0.6448,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5608084960602946,
+      "grad_norm": 1.0033676624298096,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5608941418293937,
+      "grad_norm": 0.7580384612083435,
+      "learning_rate": 0.0001,
+      "loss": 1.0371,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5609797875984927,
+      "grad_norm": 0.6465206146240234,
+      "learning_rate": 0.0001,
+      "loss": 0.5518,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5610654333675916,
+      "grad_norm": 0.69085294008255,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5611510791366906,
+      "grad_norm": 0.7467508316040039,
+      "learning_rate": 0.0001,
+      "loss": 1.154,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5612367249057897,
+      "grad_norm": 0.8179501891136169,
+      "learning_rate": 0.0001,
+      "loss": 0.9416,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5613223706748887,
+      "grad_norm": 0.6567110419273376,
+      "learning_rate": 0.0001,
+      "loss": 0.487,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5614080164439876,
+      "grad_norm": 0.8180222511291504,
+      "learning_rate": 0.0001,
+      "loss": 0.6809,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5614936622130867,
+      "grad_norm": 0.8552481532096863,
+      "learning_rate": 0.0001,
+      "loss": 0.7078,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5615793079821857,
+      "grad_norm": 1.0056865215301514,
+      "learning_rate": 0.0001,
+      "loss": 0.9794,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5616649537512847,
+      "grad_norm": 0.6089522242546082,
+      "learning_rate": 0.0001,
+      "loss": 0.6505,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5617505995203836,
+      "grad_norm": 1.0399435758590698,
+      "learning_rate": 0.0001,
+      "loss": 0.7804,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5618362452894827,
+      "grad_norm": 0.847899317741394,
+      "learning_rate": 0.0001,
+      "loss": 1.1067,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5619218910585817,
+      "grad_norm": 0.5916684865951538,
+      "learning_rate": 0.0001,
+      "loss": 1.0648,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5620075368276807,
+      "grad_norm": 0.6953430771827698,
+      "learning_rate": 0.0001,
+      "loss": 0.9692,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5620931825967798,
+      "grad_norm": 0.8901475667953491,
+      "learning_rate": 0.0001,
+      "loss": 0.7351,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5621788283658787,
+      "grad_norm": 1.0760502815246582,
+      "learning_rate": 0.0001,
+      "loss": 1.0239,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5622644741349777,
+      "grad_norm": 0.617929995059967,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5623501199040767,
+      "grad_norm": 0.3730836808681488,
+      "learning_rate": 0.0001,
+      "loss": 0.4538,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5624357656731758,
+      "grad_norm": 0.5617478489875793,
+      "learning_rate": 0.0001,
+      "loss": 0.843,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5625214114422747,
+      "grad_norm": 0.7916732430458069,
+      "learning_rate": 0.0001,
+      "loss": 0.8356,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5626070572113737,
+      "grad_norm": 0.6838459372520447,
+      "learning_rate": 0.0001,
+      "loss": 0.4354,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5626927029804728,
+      "grad_norm": 0.8750597834587097,
+      "learning_rate": 0.0001,
+      "loss": 1.1469,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5627783487495718,
+      "grad_norm": 1.1070609092712402,
+      "learning_rate": 0.0001,
+      "loss": 1.01,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5628639945186708,
+      "grad_norm": 0.8982393145561218,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5629496402877698,
+      "grad_norm": 1.1004505157470703,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5630352860568688,
+      "grad_norm": 0.7784290313720703,
+      "learning_rate": 0.0001,
+      "loss": 0.5612,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5631209318259678,
+      "grad_norm": 1.495064377784729,
+      "learning_rate": 0.0001,
+      "loss": 1.1991,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5632065775950668,
+      "grad_norm": 1.3933497667312622,
+      "learning_rate": 0.0001,
+      "loss": 0.9013,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5632922233641658,
+      "grad_norm": 0.6554699540138245,
+      "learning_rate": 0.0001,
+      "loss": 0.542,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5633778691332648,
+      "grad_norm": 0.6536192297935486,
+      "learning_rate": 0.0001,
+      "loss": 0.5036,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5634635149023638,
+      "grad_norm": 0.8941277861595154,
+      "learning_rate": 0.0001,
+      "loss": 0.7723,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5635491606714629,
+      "grad_norm": 0.5942602157592773,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5636348064405619,
+      "grad_norm": 0.4448130428791046,
+      "learning_rate": 0.0001,
+      "loss": 0.1734,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5637204522096608,
+      "grad_norm": 0.7609057426452637,
+      "learning_rate": 0.0001,
+      "loss": 0.6903,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5638060979787598,
+      "grad_norm": 1.012851595878601,
+      "learning_rate": 0.0001,
+      "loss": 1.1902,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5638917437478589,
+      "grad_norm": 0.5101674795150757,
+      "learning_rate": 0.0001,
+      "loss": 0.6284,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5639773895169579,
+      "grad_norm": 1.4944874048233032,
+      "learning_rate": 0.0001,
+      "loss": 1.3678,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5640630352860568,
+      "grad_norm": 0.7113536596298218,
+      "learning_rate": 0.0001,
+      "loss": 0.9118,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5641486810551559,
+      "grad_norm": 0.7060819864273071,
+      "learning_rate": 0.0001,
+      "loss": 1.1274,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5642343268242549,
+      "grad_norm": 0.906292736530304,
+      "learning_rate": 0.0001,
+      "loss": 0.9515,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5643199725933539,
+      "grad_norm": 0.18395549058914185,
+      "learning_rate": 0.0001,
+      "loss": 0.1711,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5644056183624528,
+      "grad_norm": 0.5305095911026001,
+      "learning_rate": 0.0001,
+      "loss": 0.6804,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5644912641315519,
+      "grad_norm": 0.5950195789337158,
+      "learning_rate": 0.0001,
+      "loss": 1.0036,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5645769099006509,
+      "grad_norm": 0.9122858643531799,
+      "learning_rate": 0.0001,
+      "loss": 1.0823,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5646625556697499,
+      "grad_norm": 0.8528048992156982,
+      "learning_rate": 0.0001,
+      "loss": 1.0935,
+      "step": 6593
+    },
+    {
+      "epoch": 0.564748201438849,
+      "grad_norm": 0.6206998229026794,
+      "learning_rate": 0.0001,
+      "loss": 0.8402,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5648338472079479,
+      "grad_norm": 0.64557284116745,
+      "learning_rate": 0.0001,
+      "loss": 1.0705,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5649194929770469,
+      "grad_norm": 0.5371809005737305,
+      "learning_rate": 0.0001,
+      "loss": 0.7933,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5650051387461459,
+      "grad_norm": 0.7939035296440125,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 6597
+    },
+    {
+      "epoch": 0.565090784515245,
+      "grad_norm": 0.5585407614707947,
+      "learning_rate": 0.0001,
+      "loss": 0.3344,
+      "step": 6598
+    },
+    {
+      "epoch": 0.565176430284344,
+      "grad_norm": 0.693450927734375,
+      "learning_rate": 0.0001,
+      "loss": 1.0635,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5652620760534429,
+      "grad_norm": 0.7479416131973267,
+      "learning_rate": 0.0001,
+      "loss": 1.1547,
+      "step": 6600
+    },
+    {
+      "epoch": 0.565347721822542,
+      "grad_norm": 0.5982543230056763,
+      "learning_rate": 0.0001,
+      "loss": 0.5354,
+      "step": 6601
+    },
+    {
+      "epoch": 0.565433367591641,
+      "grad_norm": 0.9960576295852661,
+      "learning_rate": 0.0001,
+      "loss": 0.9767,
+      "step": 6602
+    },
+    {
+      "epoch": 0.56551901336074,
+      "grad_norm": 0.6828094124794006,
+      "learning_rate": 0.0001,
+      "loss": 1.0023,
+      "step": 6603
+    },
+    {
+      "epoch": 0.565604659129839,
+      "grad_norm": 0.7860559225082397,
+      "learning_rate": 0.0001,
+      "loss": 1.1811,
+      "step": 6604
+    },
+    {
+      "epoch": 0.565690304898938,
+      "grad_norm": 0.8057470321655273,
+      "learning_rate": 0.0001,
+      "loss": 1.1339,
+      "step": 6605
+    },
+    {
+      "epoch": 0.565775950668037,
+      "grad_norm": 0.891311526298523,
+      "learning_rate": 0.0001,
+      "loss": 0.6702,
+      "step": 6606
+    },
+    {
+      "epoch": 0.565861596437136,
+      "grad_norm": 0.2708924114704132,
+      "learning_rate": 0.0001,
+      "loss": 0.1283,
+      "step": 6607
+    },
+    {
+      "epoch": 0.565947242206235,
+      "grad_norm": 0.7948108315467834,
+      "learning_rate": 0.0001,
+      "loss": 0.7924,
+      "step": 6608
+    },
+    {
+      "epoch": 0.566032887975334,
+      "grad_norm": 0.6172263026237488,
+      "learning_rate": 0.0001,
+      "loss": 1.1321,
+      "step": 6609
+    },
+    {
+      "epoch": 0.566118533744433,
+      "grad_norm": 1.092839002609253,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5662041795135321,
+      "grad_norm": 1.016952395439148,
+      "learning_rate": 0.0001,
+      "loss": 1.3708,
+      "step": 6611
+    },
+    {
+      "epoch": 0.566289825282631,
+      "grad_norm": 0.7899763584136963,
+      "learning_rate": 0.0001,
+      "loss": 0.9524,
+      "step": 6612
+    },
+    {
+      "epoch": 0.56637547105173,
+      "grad_norm": 0.8330665826797485,
+      "learning_rate": 0.0001,
+      "loss": 0.8634,
+      "step": 6613
+    },
+    {
+      "epoch": 0.566461116820829,
+      "grad_norm": 0.8010527491569519,
+      "learning_rate": 0.0001,
+      "loss": 1.2105,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5665467625899281,
+      "grad_norm": 0.9961677193641663,
+      "learning_rate": 0.0001,
+      "loss": 0.9209,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5666324083590271,
+      "grad_norm": 0.5826826691627502,
+      "learning_rate": 0.0001,
+      "loss": 0.6982,
+      "step": 6616
+    },
+    {
+      "epoch": 0.566718054128126,
+      "grad_norm": 0.8044235110282898,
+      "learning_rate": 0.0001,
+      "loss": 0.8017,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5668036998972251,
+      "grad_norm": 0.9132981896400452,
+      "learning_rate": 0.0001,
+      "loss": 0.921,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5668893456663241,
+      "grad_norm": 0.6519424915313721,
+      "learning_rate": 0.0001,
+      "loss": 1.2166,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5669749914354231,
+      "grad_norm": 1.218218207359314,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 6620
+    },
+    {
+      "epoch": 0.567060637204522,
+      "grad_norm": 1.0772734880447388,
+      "learning_rate": 0.0001,
+      "loss": 1.3219,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5671462829736211,
+      "grad_norm": 0.5741967558860779,
+      "learning_rate": 0.0001,
+      "loss": 0.321,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5672319287427201,
+      "grad_norm": 0.9782358407974243,
+      "learning_rate": 0.0001,
+      "loss": 0.907,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5673175745118191,
+      "grad_norm": 0.43141797184944153,
+      "learning_rate": 0.0001,
+      "loss": 0.2742,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5674032202809182,
+      "grad_norm": 1.1220719814300537,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5674888660500171,
+      "grad_norm": 0.6083734035491943,
+      "learning_rate": 0.0001,
+      "loss": 0.1631,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5675745118191161,
+      "grad_norm": 0.4449094831943512,
+      "learning_rate": 0.0001,
+      "loss": 0.7546,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5676601575882152,
+      "grad_norm": 0.8429003357887268,
+      "learning_rate": 0.0001,
+      "loss": 1.0537,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5677458033573142,
+      "grad_norm": 0.722315788269043,
+      "learning_rate": 0.0001,
+      "loss": 1.2233,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5678314491264131,
+      "grad_norm": 1.0146409273147583,
+      "learning_rate": 0.0001,
+      "loss": 1.1134,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5679170948955121,
+      "grad_norm": 1.1096349954605103,
+      "learning_rate": 0.0001,
+      "loss": 1.0913,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5680027406646112,
+      "grad_norm": 0.6740328073501587,
+      "learning_rate": 0.0001,
+      "loss": 0.9734,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5680883864337102,
+      "grad_norm": 0.5741242170333862,
+      "learning_rate": 0.0001,
+      "loss": 0.6383,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5681740322028092,
+      "grad_norm": 0.714082658290863,
+      "learning_rate": 0.0001,
+      "loss": 0.5943,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5682596779719082,
+      "grad_norm": 1.1327532529830933,
+      "learning_rate": 0.0001,
+      "loss": 1.4178,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5683453237410072,
+      "grad_norm": 0.6537178158760071,
+      "learning_rate": 0.0001,
+      "loss": 0.4276,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5684309695101062,
+      "grad_norm": 0.7800969481468201,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5685166152792052,
+      "grad_norm": 0.7301328778266907,
+      "learning_rate": 0.0001,
+      "loss": 1.2168,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5686022610483042,
+      "grad_norm": 0.7238157987594604,
+      "learning_rate": 0.0001,
+      "loss": 1.1719,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5686879068174032,
+      "grad_norm": 1.0703537464141846,
+      "learning_rate": 0.0001,
+      "loss": 1.191,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5687735525865022,
+      "grad_norm": 0.9336387515068054,
+      "learning_rate": 0.0001,
+      "loss": 1.7841,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5688591983556013,
+      "grad_norm": 1.0601216554641724,
+      "learning_rate": 0.0001,
+      "loss": 0.6137,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5689448441247003,
+      "grad_norm": 0.7315763235092163,
+      "learning_rate": 0.0001,
+      "loss": 0.9508,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5690304898937992,
+      "grad_norm": 0.6479259729385376,
+      "learning_rate": 0.0001,
+      "loss": 0.8462,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5691161356628982,
+      "grad_norm": 0.8332936763763428,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5692017814319973,
+      "grad_norm": 0.7088634967803955,
+      "learning_rate": 0.0001,
+      "loss": 0.6616,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5692874272010963,
+      "grad_norm": 0.478023499250412,
+      "learning_rate": 0.0001,
+      "loss": 0.8813,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5693730729701952,
+      "grad_norm": 0.9943012595176697,
+      "learning_rate": 0.0001,
+      "loss": 2.0404,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5694587187392943,
+      "grad_norm": 0.881984531879425,
+      "learning_rate": 0.0001,
+      "loss": 1.0678,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5695443645083933,
+      "grad_norm": 0.8569051027297974,
+      "learning_rate": 0.0001,
+      "loss": 0.969,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5696300102774923,
+      "grad_norm": 0.43304872512817383,
+      "learning_rate": 0.0001,
+      "loss": 0.8694,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5697156560465912,
+      "grad_norm": 0.6292818188667297,
+      "learning_rate": 0.0001,
+      "loss": 1.0629,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5698013018156903,
+      "grad_norm": 0.9326842427253723,
+      "learning_rate": 0.0001,
+      "loss": 1.2066,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5698869475847893,
+      "grad_norm": 1.0460646152496338,
+      "learning_rate": 0.0001,
+      "loss": 1.1526,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5699725933538883,
+      "grad_norm": 0.8736068606376648,
+      "learning_rate": 0.0001,
+      "loss": 0.8563,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5700582391229874,
+      "grad_norm": 1.2656033039093018,
+      "learning_rate": 0.0001,
+      "loss": 1.9049,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5701438848920863,
+      "grad_norm": 1.0450910329818726,
+      "learning_rate": 0.0001,
+      "loss": 1.2366,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5702295306611853,
+      "grad_norm": 0.37856820225715637,
+      "learning_rate": 0.0001,
+      "loss": 0.414,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5703151764302844,
+      "grad_norm": 0.9019595384597778,
+      "learning_rate": 0.0001,
+      "loss": 0.8004,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5704008221993834,
+      "grad_norm": 0.5741859078407288,
+      "learning_rate": 0.0001,
+      "loss": 0.8665,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5704864679684823,
+      "grad_norm": 0.8853818774223328,
+      "learning_rate": 0.0001,
+      "loss": 1.0246,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5705721137375813,
+      "grad_norm": 0.3398793339729309,
+      "learning_rate": 0.0001,
+      "loss": 0.3595,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5706577595066804,
+      "grad_norm": 0.8362144231796265,
+      "learning_rate": 0.0001,
+      "loss": 0.9795,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5707434052757794,
+      "grad_norm": 1.681220293045044,
+      "learning_rate": 0.0001,
+      "loss": 1.1746,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5708290510448784,
+      "grad_norm": 0.6038786172866821,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5709146968139774,
+      "grad_norm": 0.9283973574638367,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5710003425830764,
+      "grad_norm": 0.9329904913902283,
+      "learning_rate": 0.0001,
+      "loss": 0.7639,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5710859883521754,
+      "grad_norm": 0.7578496932983398,
+      "learning_rate": 0.0001,
+      "loss": 0.9979,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5711716341212744,
+      "grad_norm": 0.6860520839691162,
+      "learning_rate": 0.0001,
+      "loss": 0.5593,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5712572798903734,
+      "grad_norm": 0.6657906174659729,
+      "learning_rate": 0.0001,
+      "loss": 0.5204,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5713429256594724,
+      "grad_norm": 0.43561556935310364,
+      "learning_rate": 0.0001,
+      "loss": 0.2816,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.8950371742248535,
+      "learning_rate": 0.0001,
+      "loss": 1.039,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5715142171976705,
+      "grad_norm": 0.37152236700057983,
+      "learning_rate": 0.0001,
+      "loss": 0.3403,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5715998629667695,
+      "grad_norm": 0.9337782263755798,
+      "learning_rate": 0.0001,
+      "loss": 0.8868,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5716855087358684,
+      "grad_norm": 0.8063766360282898,
+      "learning_rate": 0.0001,
+      "loss": 1.3228,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5717711545049674,
+      "grad_norm": 1.038413166999817,
+      "learning_rate": 0.0001,
+      "loss": 0.6298,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5718568002740665,
+      "grad_norm": 0.8721783757209778,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5719424460431655,
+      "grad_norm": 0.7382345199584961,
+      "learning_rate": 0.0001,
+      "loss": 0.5239,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5720280918122644,
+      "grad_norm": 0.8672158122062683,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5721137375813635,
+      "grad_norm": 0.4235042333602905,
+      "learning_rate": 0.0001,
+      "loss": 0.6682,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5721993833504625,
+      "grad_norm": 0.6470679044723511,
+      "learning_rate": 0.0001,
+      "loss": 1.1359,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5722850291195615,
+      "grad_norm": 0.4772200882434845,
+      "learning_rate": 0.0001,
+      "loss": 0.5129,
+      "step": 6682
+    },
+    {
+      "epoch": 0.5723706748886604,
+      "grad_norm": 0.8952112197875977,
+      "learning_rate": 0.0001,
+      "loss": 1.078,
+      "step": 6683
+    },
+    {
+      "epoch": 0.5724563206577595,
+      "grad_norm": 0.764445424079895,
+      "learning_rate": 0.0001,
+      "loss": 0.6342,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5725419664268585,
+      "grad_norm": 0.7705364227294922,
+      "learning_rate": 0.0001,
+      "loss": 1.2609,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5726276121959575,
+      "grad_norm": 0.8274341225624084,
+      "learning_rate": 0.0001,
+      "loss": 0.7062,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5727132579650566,
+      "grad_norm": 0.610056459903717,
+      "learning_rate": 0.0001,
+      "loss": 0.6323,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5727989037341555,
+      "grad_norm": 0.7053898572921753,
+      "learning_rate": 0.0001,
+      "loss": 0.4351,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5728845495032545,
+      "grad_norm": 1.6544584035873413,
+      "learning_rate": 0.0001,
+      "loss": 0.6629,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5729701952723536,
+      "grad_norm": 0.8106437921524048,
+      "learning_rate": 0.0001,
+      "loss": 1.2447,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5730558410414526,
+      "grad_norm": 0.35199275612831116,
+      "learning_rate": 0.0001,
+      "loss": 0.7608,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5731414868105515,
+      "grad_norm": 1.0048716068267822,
+      "learning_rate": 0.0001,
+      "loss": 1.1847,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5732271325796505,
+      "grad_norm": 0.8044989109039307,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5733127783487496,
+      "grad_norm": 0.9926363825798035,
+      "learning_rate": 0.0001,
+      "loss": 1.8578,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5733984241178486,
+      "grad_norm": 1.1016262769699097,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5734840698869476,
+      "grad_norm": 0.8609286546707153,
+      "learning_rate": 0.0001,
+      "loss": 0.8356,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5735697156560466,
+      "grad_norm": 0.7801516056060791,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5736553614251456,
+      "grad_norm": 1.0352696180343628,
+      "learning_rate": 0.0001,
+      "loss": 1.0294,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5737410071942446,
+      "grad_norm": 0.5102863907814026,
+      "learning_rate": 0.0001,
+      "loss": 1.0729,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5738266529633436,
+      "grad_norm": 1.024735927581787,
+      "learning_rate": 0.0001,
+      "loss": 1.138,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5739122987324426,
+      "grad_norm": 0.8784698247909546,
+      "learning_rate": 0.0001,
+      "loss": 1.311,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5739979445015416,
+      "grad_norm": 0.8365779519081116,
+      "learning_rate": 0.0001,
+      "loss": 0.7697,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5740835902706406,
+      "grad_norm": 0.7858773469924927,
+      "learning_rate": 0.0001,
+      "loss": 0.6418,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5741692360397397,
+      "grad_norm": 1.2033839225769043,
+      "learning_rate": 0.0001,
+      "loss": 0.7408,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5742548818088387,
+      "grad_norm": 1.2339961528778076,
+      "learning_rate": 0.0001,
+      "loss": 0.9888,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5743405275779376,
+      "grad_norm": 0.7334771752357483,
+      "learning_rate": 0.0001,
+      "loss": 1.844,
+      "step": 6706
+    },
+    {
+      "epoch": 0.5744261733470366,
+      "grad_norm": 0.8349814414978027,
+      "learning_rate": 0.0001,
+      "loss": 0.5542,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5745118191161357,
+      "grad_norm": 1.0709072351455688,
+      "learning_rate": 0.0001,
+      "loss": 1.0055,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5745974648852347,
+      "grad_norm": 0.9415644407272339,
+      "learning_rate": 0.0001,
+      "loss": 0.9982,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5746831106543336,
+      "grad_norm": 0.7747213840484619,
+      "learning_rate": 0.0001,
+      "loss": 0.6586,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5747687564234327,
+      "grad_norm": 0.7721061706542969,
+      "learning_rate": 0.0001,
+      "loss": 0.7918,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5748544021925317,
+      "grad_norm": 0.6077190637588501,
+      "learning_rate": 0.0001,
+      "loss": 0.6392,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5749400479616307,
+      "grad_norm": 1.4358783960342407,
+      "learning_rate": 0.0001,
+      "loss": 1.3439,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5750256937307296,
+      "grad_norm": 0.4609130620956421,
+      "learning_rate": 0.0001,
+      "loss": 0.3849,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5751113394998287,
+      "grad_norm": 0.9163593649864197,
+      "learning_rate": 0.0001,
+      "loss": 0.8444,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5751969852689277,
+      "grad_norm": 0.7800784111022949,
+      "learning_rate": 0.0001,
+      "loss": 1.0164,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5752826310380267,
+      "grad_norm": 0.5653655529022217,
+      "learning_rate": 0.0001,
+      "loss": 0.4074,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5753682768071258,
+      "grad_norm": 1.1587409973144531,
+      "learning_rate": 0.0001,
+      "loss": 0.5884,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5754539225762247,
+      "grad_norm": 0.8632098436355591,
+      "learning_rate": 0.0001,
+      "loss": 0.8488,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.6782427430152893,
+      "learning_rate": 0.0001,
+      "loss": 1.0439,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5756252141144228,
+      "grad_norm": 0.7687835097312927,
+      "learning_rate": 0.0001,
+      "loss": 1.1845,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5757108598835218,
+      "grad_norm": 0.7004750967025757,
+      "learning_rate": 0.0001,
+      "loss": 0.9606,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5757965056526207,
+      "grad_norm": 0.5813886523246765,
+      "learning_rate": 0.0001,
+      "loss": 0.9263,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5758821514217197,
+      "grad_norm": 0.47908684611320496,
+      "learning_rate": 0.0001,
+      "loss": 0.7548,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5759677971908188,
+      "grad_norm": 1.1832523345947266,
+      "learning_rate": 0.0001,
+      "loss": 1.2168,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5760534429599178,
+      "grad_norm": 0.9367848038673401,
+      "learning_rate": 0.0001,
+      "loss": 1.0902,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5761390887290168,
+      "grad_norm": 1.021573781967163,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5762247344981158,
+      "grad_norm": 1.0361422300338745,
+      "learning_rate": 0.0001,
+      "loss": 1.3555,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5763103802672148,
+      "grad_norm": 0.8426274657249451,
+      "learning_rate": 0.0001,
+      "loss": 1.2936,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5763960260363138,
+      "grad_norm": 0.6110818982124329,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5764816718054128,
+      "grad_norm": 0.5030013918876648,
+      "learning_rate": 0.0001,
+      "loss": 0.9416,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5765673175745119,
+      "grad_norm": 0.704984188079834,
+      "learning_rate": 0.0001,
+      "loss": 1.2853,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5766529633436108,
+      "grad_norm": 1.1829030513763428,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5767386091127098,
+      "grad_norm": 0.740940511226654,
+      "learning_rate": 0.0001,
+      "loss": 0.9635,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5768242548818089,
+      "grad_norm": 1.0143400430679321,
+      "learning_rate": 0.0001,
+      "loss": 0.7195,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5769099006509079,
+      "grad_norm": 0.6799222230911255,
+      "learning_rate": 0.0001,
+      "loss": 1.1517,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5769955464200068,
+      "grad_norm": 1.121921181678772,
+      "learning_rate": 0.0001,
+      "loss": 1.8574,
+      "step": 6737
+    },
+    {
+      "epoch": 0.5770811921891058,
+      "grad_norm": 0.8318312168121338,
+      "learning_rate": 0.0001,
+      "loss": 1.1981,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5771668379582049,
+      "grad_norm": 0.9170000553131104,
+      "learning_rate": 0.0001,
+      "loss": 0.7323,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5772524837273039,
+      "grad_norm": 0.889627993106842,
+      "learning_rate": 0.0001,
+      "loss": 1.229,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5773381294964028,
+      "grad_norm": 0.7651299238204956,
+      "learning_rate": 0.0001,
+      "loss": 0.9856,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5774237752655019,
+      "grad_norm": 0.7052779793739319,
+      "learning_rate": 0.0001,
+      "loss": 1.265,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5775094210346009,
+      "grad_norm": 0.692042887210846,
+      "learning_rate": 0.0001,
+      "loss": 1.0995,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5775950668036999,
+      "grad_norm": 1.035134196281433,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5776807125727988,
+      "grad_norm": 0.5033268928527832,
+      "learning_rate": 0.0001,
+      "loss": 0.861,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5777663583418979,
+      "grad_norm": 0.9839335680007935,
+      "learning_rate": 0.0001,
+      "loss": 0.7138,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5778520041109969,
+      "grad_norm": 0.38994038105010986,
+      "learning_rate": 0.0001,
+      "loss": 0.4095,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5779376498800959,
+      "grad_norm": 0.633240282535553,
+      "learning_rate": 0.0001,
+      "loss": 1.0589,
+      "step": 6748
+    },
+    {
+      "epoch": 0.578023295649195,
+      "grad_norm": 0.5011651515960693,
+      "learning_rate": 0.0001,
+      "loss": 0.4105,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5781089414182939,
+      "grad_norm": 0.7531744837760925,
+      "learning_rate": 0.0001,
+      "loss": 0.45,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5781945871873929,
+      "grad_norm": 0.5166212916374207,
+      "learning_rate": 0.0001,
+      "loss": 0.6798,
+      "step": 6751
+    },
+    {
+      "epoch": 0.578280232956492,
+      "grad_norm": 0.5973391532897949,
+      "learning_rate": 0.0001,
+      "loss": 1.0252,
+      "step": 6752
+    },
+    {
+      "epoch": 0.578365878725591,
+      "grad_norm": 0.9972236752510071,
+      "learning_rate": 0.0001,
+      "loss": 0.7912,
+      "step": 6753
+    },
+    {
+      "epoch": 0.57845152449469,
+      "grad_norm": 0.999031662940979,
+      "learning_rate": 0.0001,
+      "loss": 0.8447,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5785371702637889,
+      "grad_norm": 1.1688791513442993,
+      "learning_rate": 0.0001,
+      "loss": 0.8507,
+      "step": 6755
+    },
+    {
+      "epoch": 0.578622816032888,
+      "grad_norm": 0.7879659533500671,
+      "learning_rate": 0.0001,
+      "loss": 1.1724,
+      "step": 6756
+    },
+    {
+      "epoch": 0.578708461801987,
+      "grad_norm": 0.6457107663154602,
+      "learning_rate": 0.0001,
+      "loss": 1.0153,
+      "step": 6757
+    },
+    {
+      "epoch": 0.578794107571086,
+      "grad_norm": 0.8573365211486816,
+      "learning_rate": 0.0001,
+      "loss": 0.9209,
+      "step": 6758
+    },
+    {
+      "epoch": 0.578879753340185,
+      "grad_norm": 1.1955839395523071,
+      "learning_rate": 0.0001,
+      "loss": 1.2651,
+      "step": 6759
+    },
+    {
+      "epoch": 0.578965399109284,
+      "grad_norm": 0.3991374373435974,
+      "learning_rate": 0.0001,
+      "loss": 0.4338,
+      "step": 6760
+    },
+    {
+      "epoch": 0.579051044878383,
+      "grad_norm": 1.1058260202407837,
+      "learning_rate": 0.0001,
+      "loss": 0.9019,
+      "step": 6761
+    },
+    {
+      "epoch": 0.579136690647482,
+      "grad_norm": 0.830984890460968,
+      "learning_rate": 0.0001,
+      "loss": 0.7239,
+      "step": 6762
+    },
+    {
+      "epoch": 0.579222336416581,
+      "grad_norm": 0.7179774641990662,
+      "learning_rate": 0.0001,
+      "loss": 0.8119,
+      "step": 6763
+    },
+    {
+      "epoch": 0.57930798218568,
+      "grad_norm": 0.8354492783546448,
+      "learning_rate": 0.0001,
+      "loss": 0.7504,
+      "step": 6764
+    },
+    {
+      "epoch": 0.579393627954779,
+      "grad_norm": 0.5445770621299744,
+      "learning_rate": 0.0001,
+      "loss": 0.471,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5794792737238781,
+      "grad_norm": 1.3949190378189087,
+      "learning_rate": 0.0001,
+      "loss": 0.665,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5795649194929771,
+      "grad_norm": 0.9101109504699707,
+      "learning_rate": 0.0001,
+      "loss": 0.807,
+      "step": 6767
+    },
+    {
+      "epoch": 0.579650565262076,
+      "grad_norm": 0.9709841012954712,
+      "learning_rate": 0.0001,
+      "loss": 0.8124,
+      "step": 6768
+    },
+    {
+      "epoch": 0.579736211031175,
+      "grad_norm": 0.7266142964363098,
+      "learning_rate": 0.0001,
+      "loss": 1.001,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5798218568002741,
+      "grad_norm": 0.8649279475212097,
+      "learning_rate": 0.0001,
+      "loss": 0.9039,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5799075025693731,
+      "grad_norm": 0.8043094873428345,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 6771
+    },
+    {
+      "epoch": 0.579993148338472,
+      "grad_norm": 0.5972099900245667,
+      "learning_rate": 0.0001,
+      "loss": 0.8559,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5800787941075711,
+      "grad_norm": 1.11875581741333,
+      "learning_rate": 0.0001,
+      "loss": 1.1792,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5801644398766701,
+      "grad_norm": 0.6945333480834961,
+      "learning_rate": 0.0001,
+      "loss": 0.5858,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5802500856457691,
+      "grad_norm": 1.1895091533660889,
+      "learning_rate": 0.0001,
+      "loss": 1.9568,
+      "step": 6775
+    },
+    {
+      "epoch": 0.580335731414868,
+      "grad_norm": 0.7014080882072449,
+      "learning_rate": 0.0001,
+      "loss": 0.4938,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5804213771839671,
+      "grad_norm": 0.45980048179626465,
+      "learning_rate": 0.0001,
+      "loss": 0.277,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5805070229530661,
+      "grad_norm": 0.9580221176147461,
+      "learning_rate": 0.0001,
+      "loss": 1.2725,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5805926687221651,
+      "grad_norm": 0.7437206506729126,
+      "learning_rate": 0.0001,
+      "loss": 1.2099,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5806783144912642,
+      "grad_norm": 0.6749523282051086,
+      "learning_rate": 0.0001,
+      "loss": 1.1053,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5807639602603631,
+      "grad_norm": 0.30198588967323303,
+      "learning_rate": 0.0001,
+      "loss": 0.1422,
+      "step": 6781
+    },
+    {
+      "epoch": 0.5808496060294621,
+      "grad_norm": 0.5522127151489258,
+      "learning_rate": 0.0001,
+      "loss": 0.4609,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5809352517985612,
+      "grad_norm": 0.27789658308029175,
+      "learning_rate": 0.0001,
+      "loss": 0.4793,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5810208975676602,
+      "grad_norm": 0.9591198563575745,
+      "learning_rate": 0.0001,
+      "loss": 1.299,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5811065433367592,
+      "grad_norm": 1.0855746269226074,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5811921891058581,
+      "grad_norm": 0.31672385334968567,
+      "learning_rate": 0.0001,
+      "loss": 0.3112,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5812778348749572,
+      "grad_norm": 0.6267746686935425,
+      "learning_rate": 0.0001,
+      "loss": 0.8062,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5813634806440562,
+      "grad_norm": 0.5990134477615356,
+      "learning_rate": 0.0001,
+      "loss": 0.5865,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5814491264131552,
+      "grad_norm": 0.7134597897529602,
+      "learning_rate": 0.0001,
+      "loss": 1.1396,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5815347721822542,
+      "grad_norm": 0.7958407998085022,
+      "learning_rate": 0.0001,
+      "loss": 1.1749,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5816204179513532,
+      "grad_norm": 0.5945923328399658,
+      "learning_rate": 0.0001,
+      "loss": 0.6679,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5817060637204522,
+      "grad_norm": 1.1702975034713745,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 6792
+    },
+    {
+      "epoch": 0.5817917094895512,
+      "grad_norm": 1.2783780097961426,
+      "learning_rate": 0.0001,
+      "loss": 1.3229,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5818773552586503,
+      "grad_norm": 0.7634915709495544,
+      "learning_rate": 0.0001,
+      "loss": 0.8555,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5819630010277492,
+      "grad_norm": 1.10080885887146,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5820486467968482,
+      "grad_norm": 1.0345234870910645,
+      "learning_rate": 0.0001,
+      "loss": 0.6729,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5821342925659473,
+      "grad_norm": 0.7880010008811951,
+      "learning_rate": 0.0001,
+      "loss": 0.9255,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5822199383350463,
+      "grad_norm": 0.7886226177215576,
+      "learning_rate": 0.0001,
+      "loss": 1.1817,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5823055841041452,
+      "grad_norm": 0.5572845339775085,
+      "learning_rate": 0.0001,
+      "loss": 0.7879,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5823912298732442,
+      "grad_norm": 0.8106064200401306,
+      "learning_rate": 0.0001,
+      "loss": 1.1709,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5824768756423433,
+      "grad_norm": 0.5630437731742859,
+      "learning_rate": 0.0001,
+      "loss": 0.7833,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5825625214114423,
+      "grad_norm": 0.8426217436790466,
+      "learning_rate": 0.0001,
+      "loss": 0.9901,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5826481671805412,
+      "grad_norm": 0.7787497639656067,
+      "learning_rate": 0.0001,
+      "loss": 0.653,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5827338129496403,
+      "grad_norm": 0.9327191114425659,
+      "learning_rate": 0.0001,
+      "loss": 0.9485,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5828194587187393,
+      "grad_norm": 1.3683362007141113,
+      "learning_rate": 0.0001,
+      "loss": 1.349,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5829051044878383,
+      "grad_norm": 0.6845794916152954,
+      "learning_rate": 0.0001,
+      "loss": 0.8795,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5829907502569374,
+      "grad_norm": 0.6331135630607605,
+      "learning_rate": 0.0001,
+      "loss": 1.2452,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5830763960260363,
+      "grad_norm": 0.42422494292259216,
+      "learning_rate": 0.0001,
+      "loss": 0.5333,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5831620417951353,
+      "grad_norm": 0.46139946579933167,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5832476875642343,
+      "grad_norm": 0.7821620106697083,
+      "learning_rate": 0.0001,
+      "loss": 0.7867,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 1.1876546144485474,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5834189791024323,
+      "grad_norm": 0.7139398455619812,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5835046248715313,
+      "grad_norm": 1.0336298942565918,
+      "learning_rate": 0.0001,
+      "loss": 1.3711,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5835902706406304,
+      "grad_norm": 1.116819977760315,
+      "learning_rate": 0.0001,
+      "loss": 1.0929,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5836759164097294,
+      "grad_norm": 0.7604418992996216,
+      "learning_rate": 0.0001,
+      "loss": 0.8234,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5837615621788284,
+      "grad_norm": 0.7280924916267395,
+      "learning_rate": 0.0001,
+      "loss": 1.2616,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5838472079479273,
+      "grad_norm": 0.8608729839324951,
+      "learning_rate": 0.0001,
+      "loss": 1.2018,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5839328537170264,
+      "grad_norm": 0.6765090823173523,
+      "learning_rate": 0.0001,
+      "loss": 0.4027,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5840184994861254,
+      "grad_norm": 0.8920789361000061,
+      "learning_rate": 0.0001,
+      "loss": 1.1391,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5841041452552244,
+      "grad_norm": 2.5512542724609375,
+      "learning_rate": 0.0001,
+      "loss": 2.5733,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5841897910243234,
+      "grad_norm": 1.1843560934066772,
+      "learning_rate": 0.0001,
+      "loss": 0.9465,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5842754367934224,
+      "grad_norm": 0.5902701616287231,
+      "learning_rate": 0.0001,
+      "loss": 0.8776,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5843610825625214,
+      "grad_norm": 0.6526244878768921,
+      "learning_rate": 0.0001,
+      "loss": 0.7699,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5844467283316204,
+      "grad_norm": 0.7151080369949341,
+      "learning_rate": 0.0001,
+      "loss": 0.6329,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5845323741007195,
+      "grad_norm": 0.7815601229667664,
+      "learning_rate": 0.0001,
+      "loss": 0.7015,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5846180198698184,
+      "grad_norm": 1.2917917966842651,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5847036656389174,
+      "grad_norm": 0.7335100769996643,
+      "learning_rate": 0.0001,
+      "loss": 0.884,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5847893114080165,
+      "grad_norm": 0.7748374938964844,
+      "learning_rate": 0.0001,
+      "loss": 0.7188,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5848749571771155,
+      "grad_norm": 0.8950281143188477,
+      "learning_rate": 0.0001,
+      "loss": 1.1585,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5849606029462144,
+      "grad_norm": 0.7288409471511841,
+      "learning_rate": 0.0001,
+      "loss": 0.5116,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5850462487153134,
+      "grad_norm": 0.31879115104675293,
+      "learning_rate": 0.0001,
+      "loss": 0.4267,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5851318944844125,
+      "grad_norm": 0.8342686891555786,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5852175402535115,
+      "grad_norm": 0.7130057215690613,
+      "learning_rate": 0.0001,
+      "loss": 0.7215,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5853031860226104,
+      "grad_norm": 0.760628879070282,
+      "learning_rate": 0.0001,
+      "loss": 0.9761,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5853888317917095,
+      "grad_norm": 1.3680734634399414,
+      "learning_rate": 0.0001,
+      "loss": 1.1481,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5854744775608085,
+      "grad_norm": 1.0184653997421265,
+      "learning_rate": 0.0001,
+      "loss": 1.2332,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5855601233299075,
+      "grad_norm": 0.9992694854736328,
+      "learning_rate": 0.0001,
+      "loss": 1.3049,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5856457690990066,
+      "grad_norm": 0.8705909252166748,
+      "learning_rate": 0.0001,
+      "loss": 0.8651,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5857314148681055,
+      "grad_norm": 0.8709957003593445,
+      "learning_rate": 0.0001,
+      "loss": 0.7861,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5858170606372045,
+      "grad_norm": 0.6122743487358093,
+      "learning_rate": 0.0001,
+      "loss": 0.414,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5859027064063035,
+      "grad_norm": 0.8207975029945374,
+      "learning_rate": 0.0001,
+      "loss": 1.018,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5859883521754026,
+      "grad_norm": 0.6296021938323975,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5860739979445015,
+      "grad_norm": 0.2606130838394165,
+      "learning_rate": 0.0001,
+      "loss": 0.3607,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5861596437136005,
+      "grad_norm": 0.7255527377128601,
+      "learning_rate": 0.0001,
+      "loss": 1.2582,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5862452894826996,
+      "grad_norm": 0.8514520525932312,
+      "learning_rate": 0.0001,
+      "loss": 0.6915,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5863309352517986,
+      "grad_norm": 0.7340638041496277,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5864165810208976,
+      "grad_norm": 0.31416627764701843,
+      "learning_rate": 0.0001,
+      "loss": 0.1049,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5865022267899965,
+      "grad_norm": 0.7624481916427612,
+      "learning_rate": 0.0001,
+      "loss": 0.854,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5865878725590956,
+      "grad_norm": 0.6855506896972656,
+      "learning_rate": 0.0001,
+      "loss": 0.5869,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5866735183281946,
+      "grad_norm": 1.024787425994873,
+      "learning_rate": 0.0001,
+      "loss": 0.8193,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5867591640972936,
+      "grad_norm": 0.7933021187782288,
+      "learning_rate": 0.0001,
+      "loss": 1.0104,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5868448098663926,
+      "grad_norm": 0.7048773169517517,
+      "learning_rate": 0.0001,
+      "loss": 0.9862,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5869304556354916,
+      "grad_norm": 0.5890714526176453,
+      "learning_rate": 0.0001,
+      "loss": 0.838,
+      "step": 6853
+    },
+    {
+      "epoch": 0.5870161014045906,
+      "grad_norm": 0.5019266605377197,
+      "learning_rate": 0.0001,
+      "loss": 1.0873,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5871017471736896,
+      "grad_norm": 1.4815300703048706,
+      "learning_rate": 0.0001,
+      "loss": 1.8474,
+      "step": 6855
+    },
+    {
+      "epoch": 0.5871873929427887,
+      "grad_norm": 0.6716006994247437,
+      "learning_rate": 0.0001,
+      "loss": 1.0316,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5872730387118876,
+      "grad_norm": 0.9073118567466736,
+      "learning_rate": 0.0001,
+      "loss": 1.197,
+      "step": 6857
+    },
+    {
+      "epoch": 0.5873586844809866,
+      "grad_norm": 0.5417948365211487,
+      "learning_rate": 0.0001,
+      "loss": 0.5494,
+      "step": 6858
+    },
+    {
+      "epoch": 0.5874443302500857,
+      "grad_norm": 0.8490779399871826,
+      "learning_rate": 0.0001,
+      "loss": 0.9431,
+      "step": 6859
+    },
+    {
+      "epoch": 0.5875299760191847,
+      "grad_norm": 0.7992382049560547,
+      "learning_rate": 0.0001,
+      "loss": 0.6632,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5876156217882836,
+      "grad_norm": 0.9763277173042297,
+      "learning_rate": 0.0001,
+      "loss": 1.045,
+      "step": 6861
+    },
+    {
+      "epoch": 0.5877012675573826,
+      "grad_norm": 0.6619424819946289,
+      "learning_rate": 0.0001,
+      "loss": 0.5165,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5877869133264817,
+      "grad_norm": 0.6789329648017883,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5878725590955807,
+      "grad_norm": 0.4576415419578552,
+      "learning_rate": 0.0001,
+      "loss": 0.688,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5879582048646796,
+      "grad_norm": 0.769723117351532,
+      "learning_rate": 0.0001,
+      "loss": 0.8529,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5880438506337787,
+      "grad_norm": 1.0998363494873047,
+      "learning_rate": 0.0001,
+      "loss": 0.7493,
+      "step": 6866
+    },
+    {
+      "epoch": 0.5881294964028777,
+      "grad_norm": 1.3928632736206055,
+      "learning_rate": 0.0001,
+      "loss": 1.0146,
+      "step": 6867
+    },
+    {
+      "epoch": 0.5882151421719767,
+      "grad_norm": 0.7364118099212646,
+      "learning_rate": 0.0001,
+      "loss": 0.6868,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5883007879410758,
+      "grad_norm": 1.070141077041626,
+      "learning_rate": 0.0001,
+      "loss": 1.9457,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5883864337101747,
+      "grad_norm": 0.8676923513412476,
+      "learning_rate": 0.0001,
+      "loss": 0.7184,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5884720794792737,
+      "grad_norm": 0.9618691802024841,
+      "learning_rate": 0.0001,
+      "loss": 0.8875,
+      "step": 6871
+    },
+    {
+      "epoch": 0.5885577252483727,
+      "grad_norm": 1.2224812507629395,
+      "learning_rate": 0.0001,
+      "loss": 1.1079,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5886433710174718,
+      "grad_norm": 0.6365410089492798,
+      "learning_rate": 0.0001,
+      "loss": 0.7604,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5887290167865707,
+      "grad_norm": 0.6253990530967712,
+      "learning_rate": 0.0001,
+      "loss": 1.1778,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5888146625556697,
+      "grad_norm": 0.7758371829986572,
+      "learning_rate": 0.0001,
+      "loss": 1.0189,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5889003083247688,
+      "grad_norm": 1.1464190483093262,
+      "learning_rate": 0.0001,
+      "loss": 1.3748,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5889859540938678,
+      "grad_norm": 0.3332884609699249,
+      "learning_rate": 0.0001,
+      "loss": 0.6289,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5890715998629668,
+      "grad_norm": 0.62575763463974,
+      "learning_rate": 0.0001,
+      "loss": 0.9533,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5891572456320657,
+      "grad_norm": 0.7957264184951782,
+      "learning_rate": 0.0001,
+      "loss": 0.7849,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5892428914011648,
+      "grad_norm": 0.44261831045150757,
+      "learning_rate": 0.0001,
+      "loss": 0.5981,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5893285371702638,
+      "grad_norm": 0.5193822383880615,
+      "learning_rate": 0.0001,
+      "loss": 1.0336,
+      "step": 6881
+    },
+    {
+      "epoch": 0.5894141829393628,
+      "grad_norm": 1.2468317747116089,
+      "learning_rate": 0.0001,
+      "loss": 1.8152,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5894998287084618,
+      "grad_norm": 1.0012816190719604,
+      "learning_rate": 0.0001,
+      "loss": 0.8793,
+      "step": 6883
+    },
+    {
+      "epoch": 0.5895854744775608,
+      "grad_norm": 0.5679765343666077,
+      "learning_rate": 0.0001,
+      "loss": 0.6118,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5896711202466598,
+      "grad_norm": 1.2661877870559692,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5897567660157588,
+      "grad_norm": 1.001263976097107,
+      "learning_rate": 0.0001,
+      "loss": 1.0379,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5898424117848579,
+      "grad_norm": 0.5323554277420044,
+      "learning_rate": 0.0001,
+      "loss": 0.9935,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5899280575539568,
+      "grad_norm": 0.608736515045166,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5900137033230558,
+      "grad_norm": 0.3475283086299896,
+      "learning_rate": 0.0001,
+      "loss": 0.8198,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5900993490921549,
+      "grad_norm": 0.9028075933456421,
+      "learning_rate": 0.0001,
+      "loss": 0.8019,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5901849948612539,
+      "grad_norm": 0.8595422506332397,
+      "learning_rate": 0.0001,
+      "loss": 0.8963,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5902706406303528,
+      "grad_norm": 0.46092143654823303,
+      "learning_rate": 0.0001,
+      "loss": 0.9101,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5903562863994518,
+      "grad_norm": 0.5316591858863831,
+      "learning_rate": 0.0001,
+      "loss": 1.0498,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5904419321685509,
+      "grad_norm": 0.8483090996742249,
+      "learning_rate": 0.0001,
+      "loss": 0.262,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5905275779376499,
+      "grad_norm": 0.9513596892356873,
+      "learning_rate": 0.0001,
+      "loss": 1.0963,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5906132237067488,
+      "grad_norm": 0.6923518776893616,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5906988694758479,
+      "grad_norm": 1.3546772003173828,
+      "learning_rate": 0.0001,
+      "loss": 0.8739,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5907845152449469,
+      "grad_norm": 0.6106842756271362,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5908701610140459,
+      "grad_norm": 1.161518931388855,
+      "learning_rate": 0.0001,
+      "loss": 1.2968,
+      "step": 6899
+    },
+    {
+      "epoch": 0.590955806783145,
+      "grad_norm": 1.5059902667999268,
+      "learning_rate": 0.0001,
+      "loss": 1.2064,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5910414525522439,
+      "grad_norm": 1.1038014888763428,
+      "learning_rate": 0.0001,
+      "loss": 0.9688,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5911270983213429,
+      "grad_norm": 0.6613191366195679,
+      "learning_rate": 0.0001,
+      "loss": 1.1007,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5912127440904419,
+      "grad_norm": 0.7321470379829407,
+      "learning_rate": 0.0001,
+      "loss": 1.2013,
+      "step": 6903
+    },
+    {
+      "epoch": 0.591298389859541,
+      "grad_norm": 0.6192826628684998,
+      "learning_rate": 0.0001,
+      "loss": 0.8639,
+      "step": 6904
+    },
+    {
+      "epoch": 0.59138403562864,
+      "grad_norm": 0.35973361134529114,
+      "learning_rate": 0.0001,
+      "loss": 0.5662,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5914696813977389,
+      "grad_norm": 0.43506449460983276,
+      "learning_rate": 0.0001,
+      "loss": 0.4838,
+      "step": 6906
+    },
+    {
+      "epoch": 0.591555327166838,
+      "grad_norm": 0.5374320149421692,
+      "learning_rate": 0.0001,
+      "loss": 1.0008,
+      "step": 6907
+    },
+    {
+      "epoch": 0.591640972935937,
+      "grad_norm": 0.37916553020477295,
+      "learning_rate": 0.0001,
+      "loss": 0.2854,
+      "step": 6908
+    },
+    {
+      "epoch": 0.591726618705036,
+      "grad_norm": 1.1755951642990112,
+      "learning_rate": 0.0001,
+      "loss": 1.0061,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5918122644741349,
+      "grad_norm": 0.7688059210777283,
+      "learning_rate": 0.0001,
+      "loss": 0.7423,
+      "step": 6910
+    },
+    {
+      "epoch": 0.591897910243234,
+      "grad_norm": 0.5941349864006042,
+      "learning_rate": 0.0001,
+      "loss": 0.8372,
+      "step": 6911
+    },
+    {
+      "epoch": 0.591983556012333,
+      "grad_norm": 0.41770124435424805,
+      "learning_rate": 0.0001,
+      "loss": 0.6756,
+      "step": 6912
+    },
+    {
+      "epoch": 0.592069201781432,
+      "grad_norm": 0.6456491947174072,
+      "learning_rate": 0.0001,
+      "loss": 0.9917,
+      "step": 6913
+    },
+    {
+      "epoch": 0.592154847550531,
+      "grad_norm": 0.6951163411140442,
+      "learning_rate": 0.0001,
+      "loss": 0.751,
+      "step": 6914
+    },
+    {
+      "epoch": 0.59224049331963,
+      "grad_norm": 0.4965023696422577,
+      "learning_rate": 0.0001,
+      "loss": 0.9745,
+      "step": 6915
+    },
+    {
+      "epoch": 0.592326139088729,
+      "grad_norm": 1.339392900466919,
+      "learning_rate": 0.0001,
+      "loss": 1.3313,
+      "step": 6916
+    },
+    {
+      "epoch": 0.592411784857828,
+      "grad_norm": 0.5044484734535217,
+      "learning_rate": 0.0001,
+      "loss": 0.8054,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5924974306269271,
+      "grad_norm": 0.7672072649002075,
+      "learning_rate": 0.0001,
+      "loss": 0.6453,
+      "step": 6918
+    },
+    {
+      "epoch": 0.592583076396026,
+      "grad_norm": 0.3837575614452362,
+      "learning_rate": 0.0001,
+      "loss": 0.6687,
+      "step": 6919
+    },
+    {
+      "epoch": 0.592668722165125,
+      "grad_norm": 0.8068817853927612,
+      "learning_rate": 0.0001,
+      "loss": 0.7479,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5927543679342241,
+      "grad_norm": 1.0062692165374756,
+      "learning_rate": 0.0001,
+      "loss": 1.1077,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5928400137033231,
+      "grad_norm": 1.2002915143966675,
+      "learning_rate": 0.0001,
+      "loss": 1.0191,
+      "step": 6922
+    },
+    {
+      "epoch": 0.592925659472422,
+      "grad_norm": 0.7871105670928955,
+      "learning_rate": 0.0001,
+      "loss": 1.2886,
+      "step": 6923
+    },
+    {
+      "epoch": 0.593011305241521,
+      "grad_norm": 0.855661153793335,
+      "learning_rate": 0.0001,
+      "loss": 0.7693,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5930969510106201,
+      "grad_norm": 0.9053385257720947,
+      "learning_rate": 0.0001,
+      "loss": 0.7801,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5931825967797191,
+      "grad_norm": 0.43984323740005493,
+      "learning_rate": 0.0001,
+      "loss": 0.4632,
+      "step": 6926
+    },
+    {
+      "epoch": 0.593268242548818,
+      "grad_norm": 0.38939085602760315,
+      "learning_rate": 0.0001,
+      "loss": 0.3522,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5933538883179171,
+      "grad_norm": 0.81789231300354,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5934395340870161,
+      "grad_norm": 0.4486599564552307,
+      "learning_rate": 0.0001,
+      "loss": 0.7006,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5935251798561151,
+      "grad_norm": 1.0911959409713745,
+      "learning_rate": 0.0001,
+      "loss": 1.0054,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5936108256252142,
+      "grad_norm": 1.3807880878448486,
+      "learning_rate": 0.0001,
+      "loss": 0.9753,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5936964713943131,
+      "grad_norm": 0.7539426684379578,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5937821171634121,
+      "grad_norm": 0.4809465706348419,
+      "learning_rate": 0.0001,
+      "loss": 0.9748,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5938677629325111,
+      "grad_norm": 0.8980976343154907,
+      "learning_rate": 0.0001,
+      "loss": 0.9267,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5939534087016102,
+      "grad_norm": 0.8647889494895935,
+      "learning_rate": 0.0001,
+      "loss": 0.5766,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5940390544707091,
+      "grad_norm": 1.2122018337249756,
+      "learning_rate": 0.0001,
+      "loss": 0.9469,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5941247002398081,
+      "grad_norm": 0.9272986650466919,
+      "learning_rate": 0.0001,
+      "loss": 0.7948,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5942103460089072,
+      "grad_norm": 0.9985275864601135,
+      "learning_rate": 0.0001,
+      "loss": 1.1345,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5942959917780062,
+      "grad_norm": 1.165825366973877,
+      "learning_rate": 0.0001,
+      "loss": 0.8463,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5943816375471052,
+      "grad_norm": 0.6248428225517273,
+      "learning_rate": 0.0001,
+      "loss": 1.2917,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5944672833162041,
+      "grad_norm": 0.5513609051704407,
+      "learning_rate": 0.0001,
+      "loss": 1.0227,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5945529290853032,
+      "grad_norm": 0.5581822991371155,
+      "learning_rate": 0.0001,
+      "loss": 0.8524,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5946385748544022,
+      "grad_norm": 0.7079516649246216,
+      "learning_rate": 0.0001,
+      "loss": 0.6197,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5947242206235012,
+      "grad_norm": 0.7246556878089905,
+      "learning_rate": 0.0001,
+      "loss": 0.6447,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5948098663926003,
+      "grad_norm": 0.5432363152503967,
+      "learning_rate": 0.0001,
+      "loss": 0.7238,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5948955121616992,
+      "grad_norm": 1.0817711353302002,
+      "learning_rate": 0.0001,
+      "loss": 0.809,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5949811579307982,
+      "grad_norm": 0.7122589945793152,
+      "learning_rate": 0.0001,
+      "loss": 1.0127,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5950668036998972,
+      "grad_norm": 0.5313768982887268,
+      "learning_rate": 0.0001,
+      "loss": 0.9514,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5951524494689963,
+      "grad_norm": 0.8826805353164673,
+      "learning_rate": 0.0001,
+      "loss": 1.1308,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5952380952380952,
+      "grad_norm": 0.7050142288208008,
+      "learning_rate": 0.0001,
+      "loss": 0.5153,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5953237410071942,
+      "grad_norm": 0.7682812213897705,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5954093867762933,
+      "grad_norm": 1.335827112197876,
+      "learning_rate": 0.0001,
+      "loss": 1.312,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5954950325453923,
+      "grad_norm": 0.5364965200424194,
+      "learning_rate": 0.0001,
+      "loss": 1.0081,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5955806783144912,
+      "grad_norm": 0.40142208337783813,
+      "learning_rate": 0.0001,
+      "loss": 0.3048,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5956663240835903,
+      "grad_norm": 0.8470568060874939,
+      "learning_rate": 0.0001,
+      "loss": 0.863,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5957519698526893,
+      "grad_norm": 0.5841603875160217,
+      "learning_rate": 0.0001,
+      "loss": 1.2581,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5958376156217883,
+      "grad_norm": 0.539711058139801,
+      "learning_rate": 0.0001,
+      "loss": 1.2951,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5959232613908872,
+      "grad_norm": 0.8107927441596985,
+      "learning_rate": 0.0001,
+      "loss": 1.2759,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5960089071599863,
+      "grad_norm": 0.401509165763855,
+      "learning_rate": 0.0001,
+      "loss": 0.403,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5960945529290853,
+      "grad_norm": 0.36657437682151794,
+      "learning_rate": 0.0001,
+      "loss": 0.2913,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5961801986981843,
+      "grad_norm": 1.0047030448913574,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5962658444672834,
+      "grad_norm": 0.8654273748397827,
+      "learning_rate": 0.0001,
+      "loss": 1.2016,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5963514902363823,
+      "grad_norm": 1.1594429016113281,
+      "learning_rate": 0.0001,
+      "loss": 1.3521,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5964371360054813,
+      "grad_norm": 0.9119693636894226,
+      "learning_rate": 0.0001,
+      "loss": 0.9987,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5965227817745803,
+      "grad_norm": 1.0462267398834229,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5966084275436794,
+      "grad_norm": 0.6473318934440613,
+      "learning_rate": 0.0001,
+      "loss": 1.1732,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5966940733127784,
+      "grad_norm": 1.1282601356506348,
+      "learning_rate": 0.0001,
+      "loss": 1.2845,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5967797190818773,
+      "grad_norm": 0.23158809542655945,
+      "learning_rate": 0.0001,
+      "loss": 0.8294,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5968653648509764,
+      "grad_norm": 0.7095812559127808,
+      "learning_rate": 0.0001,
+      "loss": 0.7865,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5969510106200754,
+      "grad_norm": 0.7955691814422607,
+      "learning_rate": 0.0001,
+      "loss": 0.7486,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5970366563891744,
+      "grad_norm": 1.037381887435913,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5971223021582733,
+      "grad_norm": 0.5927301645278931,
+      "learning_rate": 0.0001,
+      "loss": 0.9215,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5972079479273724,
+      "grad_norm": 0.35146552324295044,
+      "learning_rate": 0.0001,
+      "loss": 0.5061,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5972935936964714,
+      "grad_norm": 0.7127518057823181,
+      "learning_rate": 0.0001,
+      "loss": 1.0913,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5973792394655704,
+      "grad_norm": 0.3979513347148895,
+      "learning_rate": 0.0001,
+      "loss": 0.4698,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5974648852346695,
+      "grad_norm": 0.2864982485771179,
+      "learning_rate": 0.0001,
+      "loss": 0.5376,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5975505310037684,
+      "grad_norm": 0.7705459594726562,
+      "learning_rate": 0.0001,
+      "loss": 1.3366,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5976361767728674,
+      "grad_norm": 0.5490134954452515,
+      "learning_rate": 0.0001,
+      "loss": 0.7702,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5977218225419664,
+      "grad_norm": 0.9080960154533386,
+      "learning_rate": 0.0001,
+      "loss": 1.2704,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5978074683110655,
+      "grad_norm": 0.28979915380477905,
+      "learning_rate": 0.0001,
+      "loss": 0.1302,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5978931140801644,
+      "grad_norm": 1.5808790922164917,
+      "learning_rate": 0.0001,
+      "loss": 1.0618,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5979787598492634,
+      "grad_norm": 0.6078392863273621,
+      "learning_rate": 0.0001,
+      "loss": 0.8051,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5980644056183625,
+      "grad_norm": 0.4687974452972412,
+      "learning_rate": 0.0001,
+      "loss": 1.234,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5981500513874615,
+      "grad_norm": 0.47788363695144653,
+      "learning_rate": 0.0001,
+      "loss": 0.8071,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5982356971565604,
+      "grad_norm": 1.1237332820892334,
+      "learning_rate": 0.0001,
+      "loss": 1.0209,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5983213429256595,
+      "grad_norm": 0.45560187101364136,
+      "learning_rate": 0.0001,
+      "loss": 0.5734,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5984069886947585,
+      "grad_norm": 0.7780702114105225,
+      "learning_rate": 0.0001,
+      "loss": 1.1459,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5984926344638575,
+      "grad_norm": 0.9079281687736511,
+      "learning_rate": 0.0001,
+      "loss": 0.7362,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5985782802329565,
+      "grad_norm": 0.5931071043014526,
+      "learning_rate": 0.0001,
+      "loss": 0.7717,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5986639260020555,
+      "grad_norm": 1.272708773612976,
+      "learning_rate": 0.0001,
+      "loss": 0.7048,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5987495717711545,
+      "grad_norm": 0.8855377435684204,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5988352175402535,
+      "grad_norm": 1.0307962894439697,
+      "learning_rate": 0.0001,
+      "loss": 0.756,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5989208633093526,
+      "grad_norm": 0.36119675636291504,
+      "learning_rate": 0.0001,
+      "loss": 0.4468,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5990065090784515,
+      "grad_norm": 0.8384900689125061,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5990921548475505,
+      "grad_norm": 0.9512903690338135,
+      "learning_rate": 0.0001,
+      "loss": 0.813,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5991778006166495,
+      "grad_norm": 0.6884494423866272,
+      "learning_rate": 0.0001,
+      "loss": 1.2112,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5992634463857486,
+      "grad_norm": 2.0204455852508545,
+      "learning_rate": 0.0001,
+      "loss": 0.9875,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5993490921548476,
+      "grad_norm": 0.7076169848442078,
+      "learning_rate": 0.0001,
+      "loss": 1.2971,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5994347379239465,
+      "grad_norm": 0.7170209884643555,
+      "learning_rate": 0.0001,
+      "loss": 0.8132,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5995203836930456,
+      "grad_norm": 0.9948247075080872,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5996060294621446,
+      "grad_norm": 0.7759745121002197,
+      "learning_rate": 0.0001,
+      "loss": 0.6377,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5996916752312436,
+      "grad_norm": 0.568347692489624,
+      "learning_rate": 0.0001,
+      "loss": 0.5392,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5997773210003425,
+      "grad_norm": 0.7315685749053955,
+      "learning_rate": 0.0001,
+      "loss": 0.9809,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5998629667694416,
+      "grad_norm": 0.6466290354728699,
+      "learning_rate": 0.0001,
+      "loss": 0.5685,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5999486125385406,
+      "grad_norm": 1.074722170829773,
+      "learning_rate": 0.0001,
+      "loss": 0.6944,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6000342583076396,
+      "grad_norm": 0.7292446494102478,
+      "learning_rate": 0.0001,
+      "loss": 0.7399,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6001199040767387,
+      "grad_norm": 0.6902074813842773,
+      "learning_rate": 0.0001,
+      "loss": 1.1289,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6002055498458376,
+      "grad_norm": 0.8817043900489807,
+      "learning_rate": 0.0001,
+      "loss": 1.8597,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6002911956149366,
+      "grad_norm": 0.7631065845489502,
+      "learning_rate": 0.0001,
+      "loss": 0.74,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6003768413840356,
+      "grad_norm": 0.6240226030349731,
+      "learning_rate": 0.0001,
+      "loss": 0.604,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6004624871531347,
+      "grad_norm": 0.949397087097168,
+      "learning_rate": 0.0001,
+      "loss": 1.2975,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6005481329222336,
+      "grad_norm": 0.7764929533004761,
+      "learning_rate": 0.0001,
+      "loss": 1.0446,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6006337786913326,
+      "grad_norm": 1.0705606937408447,
+      "learning_rate": 0.0001,
+      "loss": 1.0833,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6007194244604317,
+      "grad_norm": 0.8104068040847778,
+      "learning_rate": 0.0001,
+      "loss": 0.534,
+      "step": 7014
+    },
+    {
+      "epoch": 0.6008050702295307,
+      "grad_norm": 0.7638000845909119,
+      "learning_rate": 0.0001,
+      "loss": 1.009,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6008907159986296,
+      "grad_norm": 0.7891031503677368,
+      "learning_rate": 0.0001,
+      "loss": 1.8786,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6009763617677287,
+      "grad_norm": 0.8209484219551086,
+      "learning_rate": 0.0001,
+      "loss": 1.1646,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6010620075368277,
+      "grad_norm": 0.9053457379341125,
+      "learning_rate": 0.0001,
+      "loss": 1.1447,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6011476533059267,
+      "grad_norm": 0.6042333841323853,
+      "learning_rate": 0.0001,
+      "loss": 0.79,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6012332990750257,
+      "grad_norm": 0.5478500127792358,
+      "learning_rate": 0.0001,
+      "loss": 0.9656,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6013189448441247,
+      "grad_norm": 0.69715815782547,
+      "learning_rate": 0.0001,
+      "loss": 0.814,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6014045906132237,
+      "grad_norm": 0.8609411120414734,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 7022
+    },
+    {
+      "epoch": 0.6014902363823227,
+      "grad_norm": 0.4905543327331543,
+      "learning_rate": 0.0001,
+      "loss": 1.0478,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6015758821514218,
+      "grad_norm": 0.9294548034667969,
+      "learning_rate": 0.0001,
+      "loss": 0.954,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6016615279205207,
+      "grad_norm": 0.6629143953323364,
+      "learning_rate": 0.0001,
+      "loss": 1.0958,
+      "step": 7025
+    },
+    {
+      "epoch": 0.6017471736896197,
+      "grad_norm": 0.5513284802436829,
+      "learning_rate": 0.0001,
+      "loss": 0.3868,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6018328194587187,
+      "grad_norm": 0.651427686214447,
+      "learning_rate": 0.0001,
+      "loss": 0.4999,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6019184652278178,
+      "grad_norm": 1.145808458328247,
+      "learning_rate": 0.0001,
+      "loss": 0.8846,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6020041109969168,
+      "grad_norm": 0.5697384476661682,
+      "learning_rate": 0.0001,
+      "loss": 0.9324,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6020897567660157,
+      "grad_norm": 0.47026121616363525,
+      "learning_rate": 0.0001,
+      "loss": 0.6557,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6021754025351148,
+      "grad_norm": 0.6859575510025024,
+      "learning_rate": 0.0001,
+      "loss": 0.817,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6022610483042138,
+      "grad_norm": 0.5234060883522034,
+      "learning_rate": 0.0001,
+      "loss": 0.9104,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6023466940733128,
+      "grad_norm": 0.5013370513916016,
+      "learning_rate": 0.0001,
+      "loss": 1.0363,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6024323398424117,
+      "grad_norm": 0.9581459760665894,
+      "learning_rate": 0.0001,
+      "loss": 0.9368,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6025179856115108,
+      "grad_norm": 0.6825016140937805,
+      "learning_rate": 0.0001,
+      "loss": 0.7522,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6026036313806098,
+      "grad_norm": 1.2893192768096924,
+      "learning_rate": 0.0001,
+      "loss": 1.0833,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6026892771497088,
+      "grad_norm": 1.055524230003357,
+      "learning_rate": 0.0001,
+      "loss": 1.1906,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6027749229188079,
+      "grad_norm": 1.0108163356781006,
+      "learning_rate": 0.0001,
+      "loss": 0.7317,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6028605686879068,
+      "grad_norm": 0.6165544390678406,
+      "learning_rate": 0.0001,
+      "loss": 0.7772,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6029462144570058,
+      "grad_norm": 0.8182818293571472,
+      "learning_rate": 0.0001,
+      "loss": 0.4181,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6030318602261048,
+      "grad_norm": 0.6050503253936768,
+      "learning_rate": 0.0001,
+      "loss": 1.0209,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6031175059952039,
+      "grad_norm": 0.3576950430870056,
+      "learning_rate": 0.0001,
+      "loss": 0.3791,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6032031517643028,
+      "grad_norm": 0.5460241436958313,
+      "learning_rate": 0.0001,
+      "loss": 0.8733,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6032887975334018,
+      "grad_norm": 1.0887633562088013,
+      "learning_rate": 0.0001,
+      "loss": 0.789,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6033744433025009,
+      "grad_norm": 1.2367185354232788,
+      "learning_rate": 0.0001,
+      "loss": 1.0664,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6034600890715999,
+      "grad_norm": 0.5567166209220886,
+      "learning_rate": 0.0001,
+      "loss": 1.0598,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6035457348406988,
+      "grad_norm": 0.39939072728157043,
+      "learning_rate": 0.0001,
+      "loss": 0.5282,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6036313806097979,
+      "grad_norm": 0.9372262358665466,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6037170263788969,
+      "grad_norm": 0.8682109117507935,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6038026721479959,
+      "grad_norm": 0.6587437391281128,
+      "learning_rate": 0.0001,
+      "loss": 0.8547,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6038883179170949,
+      "grad_norm": 0.9629835486412048,
+      "learning_rate": 0.0001,
+      "loss": 1.12,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6039739636861939,
+      "grad_norm": 1.1825692653656006,
+      "learning_rate": 0.0001,
+      "loss": 1.0703,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6040596094552929,
+      "grad_norm": 0.4507846534252167,
+      "learning_rate": 0.0001,
+      "loss": 0.4814,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6041452552243919,
+      "grad_norm": 0.7914722561836243,
+      "learning_rate": 0.0001,
+      "loss": 1.1067,
+      "step": 7054
+    },
+    {
+      "epoch": 0.604230900993491,
+      "grad_norm": 0.839918851852417,
+      "learning_rate": 0.0001,
+      "loss": 0.8421,
+      "step": 7055
+    },
+    {
+      "epoch": 0.60431654676259,
+      "grad_norm": 0.1596875935792923,
+      "learning_rate": 0.0001,
+      "loss": 0.2451,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6044021925316889,
+      "grad_norm": 0.7439535856246948,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6044878383007879,
+      "grad_norm": 1.102774739265442,
+      "learning_rate": 0.0001,
+      "loss": 0.9601,
+      "step": 7058
+    },
+    {
+      "epoch": 0.604573484069887,
+      "grad_norm": 1.258415937423706,
+      "learning_rate": 0.0001,
+      "loss": 1.1146,
+      "step": 7059
+    },
+    {
+      "epoch": 0.604659129838986,
+      "grad_norm": 0.9734809994697571,
+      "learning_rate": 0.0001,
+      "loss": 0.9168,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6047447756080849,
+      "grad_norm": 0.6693723797798157,
+      "learning_rate": 0.0001,
+      "loss": 0.981,
+      "step": 7061
+    },
+    {
+      "epoch": 0.604830421377184,
+      "grad_norm": 0.8408697843551636,
+      "learning_rate": 0.0001,
+      "loss": 1.0979,
+      "step": 7062
+    },
+    {
+      "epoch": 0.604916067146283,
+      "grad_norm": 1.2881200313568115,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 7063
+    },
+    {
+      "epoch": 0.605001712915382,
+      "grad_norm": 0.620795488357544,
+      "learning_rate": 0.0001,
+      "loss": 1.1655,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6050873586844809,
+      "grad_norm": 0.9315448999404907,
+      "learning_rate": 0.0001,
+      "loss": 1.194,
+      "step": 7065
+    },
+    {
+      "epoch": 0.60517300445358,
+      "grad_norm": 0.705968976020813,
+      "learning_rate": 0.0001,
+      "loss": 1.0711,
+      "step": 7066
+    },
+    {
+      "epoch": 0.605258650222679,
+      "grad_norm": 0.9586004614830017,
+      "learning_rate": 0.0001,
+      "loss": 0.929,
+      "step": 7067
+    },
+    {
+      "epoch": 0.605344295991778,
+      "grad_norm": 0.8957716226577759,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 7068
+    },
+    {
+      "epoch": 0.605429941760877,
+      "grad_norm": 0.6891859173774719,
+      "learning_rate": 0.0001,
+      "loss": 0.7327,
+      "step": 7069
+    },
+    {
+      "epoch": 0.605515587529976,
+      "grad_norm": 0.5773548483848572,
+      "learning_rate": 0.0001,
+      "loss": 0.5024,
+      "step": 7070
+    },
+    {
+      "epoch": 0.605601233299075,
+      "grad_norm": 0.634196400642395,
+      "learning_rate": 0.0001,
+      "loss": 0.721,
+      "step": 7071
+    },
+    {
+      "epoch": 0.605686879068174,
+      "grad_norm": 0.7832228541374207,
+      "learning_rate": 0.0001,
+      "loss": 1.2682,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6057725248372731,
+      "grad_norm": 0.7709258794784546,
+      "learning_rate": 0.0001,
+      "loss": 1.1465,
+      "step": 7073
+    },
+    {
+      "epoch": 0.605858170606372,
+      "grad_norm": 0.5469086766242981,
+      "learning_rate": 0.0001,
+      "loss": 0.9171,
+      "step": 7074
+    },
+    {
+      "epoch": 0.605943816375471,
+      "grad_norm": 0.8071468472480774,
+      "learning_rate": 0.0001,
+      "loss": 0.8644,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6060294621445701,
+      "grad_norm": 0.7378748059272766,
+      "learning_rate": 0.0001,
+      "loss": 0.9129,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6061151079136691,
+      "grad_norm": 1.0834769010543823,
+      "learning_rate": 0.0001,
+      "loss": 1.259,
+      "step": 7077
+    },
+    {
+      "epoch": 0.606200753682768,
+      "grad_norm": 0.7960401773452759,
+      "learning_rate": 0.0001,
+      "loss": 1.3098,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6062863994518671,
+      "grad_norm": 1.6897752285003662,
+      "learning_rate": 0.0001,
+      "loss": 1.1535,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6063720452209661,
+      "grad_norm": 0.813714861869812,
+      "learning_rate": 0.0001,
+      "loss": 0.7464,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6064576909900651,
+      "grad_norm": 0.737843930721283,
+      "learning_rate": 0.0001,
+      "loss": 0.6347,
+      "step": 7081
+    },
+    {
+      "epoch": 0.606543336759164,
+      "grad_norm": 0.7493770122528076,
+      "learning_rate": 0.0001,
+      "loss": 1.3717,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6066289825282631,
+      "grad_norm": 1.2308869361877441,
+      "learning_rate": 0.0001,
+      "loss": 0.9283,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6067146282973621,
+      "grad_norm": 0.8281447887420654,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6068002740664611,
+      "grad_norm": 0.7705833315849304,
+      "learning_rate": 0.0001,
+      "loss": 1.0325,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6068859198355602,
+      "grad_norm": 0.9798144102096558,
+      "learning_rate": 0.0001,
+      "loss": 1.2604,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6069715656046591,
+      "grad_norm": 0.8370644450187683,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6070572113737581,
+      "grad_norm": 0.967917263507843,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 0.7047057747840881,
+      "learning_rate": 0.0001,
+      "loss": 0.8129,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6072285029119562,
+      "grad_norm": 0.9638987183570862,
+      "learning_rate": 0.0001,
+      "loss": 1.0687,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6073141486810552,
+      "grad_norm": 1.2425748109817505,
+      "learning_rate": 0.0001,
+      "loss": 1.0662,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6073997944501541,
+      "grad_norm": 0.983052670955658,
+      "learning_rate": 0.0001,
+      "loss": 0.9785,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6074854402192532,
+      "grad_norm": 1.1480014324188232,
+      "learning_rate": 0.0001,
+      "loss": 1.2579,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6075710859883522,
+      "grad_norm": 0.56418776512146,
+      "learning_rate": 0.0001,
+      "loss": 0.749,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6076567317574512,
+      "grad_norm": 0.8940640687942505,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6077423775265501,
+      "grad_norm": 1.0222464799880981,
+      "learning_rate": 0.0001,
+      "loss": 1.2303,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6078280232956492,
+      "grad_norm": 0.7509710192680359,
+      "learning_rate": 0.0001,
+      "loss": 0.9446,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6079136690647482,
+      "grad_norm": 1.1454700231552124,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6079993148338472,
+      "grad_norm": 0.7778225541114807,
+      "learning_rate": 0.0001,
+      "loss": 0.7929,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6080849606029463,
+      "grad_norm": 1.0197185277938843,
+      "learning_rate": 0.0001,
+      "loss": 0.9743,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6081706063720452,
+      "grad_norm": 0.5685485005378723,
+      "learning_rate": 0.0001,
+      "loss": 1.0325,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6082562521411442,
+      "grad_norm": 1.0434389114379883,
+      "learning_rate": 0.0001,
+      "loss": 1.1144,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6083418979102432,
+      "grad_norm": 0.7761505246162415,
+      "learning_rate": 0.0001,
+      "loss": 1.1422,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6084275436793423,
+      "grad_norm": 0.8836236000061035,
+      "learning_rate": 0.0001,
+      "loss": 1.3262,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6085131894484412,
+      "grad_norm": 2.0205323696136475,
+      "learning_rate": 0.0001,
+      "loss": 2.2555,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6085988352175402,
+      "grad_norm": 0.5678834319114685,
+      "learning_rate": 0.0001,
+      "loss": 0.5134,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6086844809866393,
+      "grad_norm": 1.0516102313995361,
+      "learning_rate": 0.0001,
+      "loss": 1.1957,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6087701267557383,
+      "grad_norm": 0.855266809463501,
+      "learning_rate": 0.0001,
+      "loss": 0.998,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6088557725248372,
+      "grad_norm": 1.4198116064071655,
+      "learning_rate": 0.0001,
+      "loss": 1.4221,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6089414182939363,
+      "grad_norm": 0.6219750642776489,
+      "learning_rate": 0.0001,
+      "loss": 0.7236,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6090270640630353,
+      "grad_norm": 0.7207213044166565,
+      "learning_rate": 0.0001,
+      "loss": 0.7559,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6091127098321343,
+      "grad_norm": 0.39913058280944824,
+      "learning_rate": 0.0001,
+      "loss": 0.5006,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6091983556012333,
+      "grad_norm": 0.7896507978439331,
+      "learning_rate": 0.0001,
+      "loss": 0.6032,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6092840013703323,
+      "grad_norm": 0.8177736401557922,
+      "learning_rate": 0.0001,
+      "loss": 0.7769,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6093696471394313,
+      "grad_norm": 1.1534035205841064,
+      "learning_rate": 0.0001,
+      "loss": 1.3643,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6094552929085303,
+      "grad_norm": 0.7231832146644592,
+      "learning_rate": 0.0001,
+      "loss": 0.9614,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6095409386776294,
+      "grad_norm": 0.4694613516330719,
+      "learning_rate": 0.0001,
+      "loss": 0.8406,
+      "step": 7117
+    },
+    {
+      "epoch": 0.6096265844467283,
+      "grad_norm": 0.7514703869819641,
+      "learning_rate": 0.0001,
+      "loss": 0.7811,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6097122302158273,
+      "grad_norm": 0.5422842502593994,
+      "learning_rate": 0.0001,
+      "loss": 0.1644,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6097978759849263,
+      "grad_norm": 0.7524746060371399,
+      "learning_rate": 0.0001,
+      "loss": 0.5764,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6098835217540254,
+      "grad_norm": 0.6397144794464111,
+      "learning_rate": 0.0001,
+      "loss": 1.0189,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6099691675231244,
+      "grad_norm": 1.0697141885757446,
+      "learning_rate": 0.0001,
+      "loss": 0.4766,
+      "step": 7122
+    },
+    {
+      "epoch": 0.6100548132922233,
+      "grad_norm": 0.8491172790527344,
+      "learning_rate": 0.0001,
+      "loss": 0.9857,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6101404590613224,
+      "grad_norm": 0.8861572742462158,
+      "learning_rate": 0.0001,
+      "loss": 0.8967,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6102261048304214,
+      "grad_norm": 0.7698114514350891,
+      "learning_rate": 0.0001,
+      "loss": 0.9558,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6103117505995204,
+      "grad_norm": 0.7055954933166504,
+      "learning_rate": 0.0001,
+      "loss": 0.872,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6103973963686193,
+      "grad_norm": 0.9793805480003357,
+      "learning_rate": 0.0001,
+      "loss": 1.1103,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6104830421377184,
+      "grad_norm": 0.8262122869491577,
+      "learning_rate": 0.0001,
+      "loss": 0.7365,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6105686879068174,
+      "grad_norm": 0.7402160167694092,
+      "learning_rate": 0.0001,
+      "loss": 0.4238,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6106543336759164,
+      "grad_norm": 0.7796196341514587,
+      "learning_rate": 0.0001,
+      "loss": 0.754,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6107399794450155,
+      "grad_norm": 0.8923584222793579,
+      "learning_rate": 0.0001,
+      "loss": 0.5497,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6108256252141144,
+      "grad_norm": 0.6793668866157532,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6109112709832134,
+      "grad_norm": 1.3350656032562256,
+      "learning_rate": 0.0001,
+      "loss": 0.9507,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6109969167523125,
+      "grad_norm": 0.5930936336517334,
+      "learning_rate": 0.0001,
+      "loss": 1.0066,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6110825625214115,
+      "grad_norm": 0.9106727242469788,
+      "learning_rate": 0.0001,
+      "loss": 0.9125,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6111682082905104,
+      "grad_norm": 0.4558521807193756,
+      "learning_rate": 0.0001,
+      "loss": 0.7763,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6112538540596094,
+      "grad_norm": 0.9421074390411377,
+      "learning_rate": 0.0001,
+      "loss": 0.9457,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6113394998287085,
+      "grad_norm": 0.8263735771179199,
+      "learning_rate": 0.0001,
+      "loss": 0.847,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6114251455978075,
+      "grad_norm": 1.0190616846084595,
+      "learning_rate": 0.0001,
+      "loss": 1.2556,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6115107913669064,
+      "grad_norm": 0.8979155421257019,
+      "learning_rate": 0.0001,
+      "loss": 0.9296,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6115964371360055,
+      "grad_norm": 0.7033807039260864,
+      "learning_rate": 0.0001,
+      "loss": 0.5753,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6116820829051045,
+      "grad_norm": 0.5355032086372375,
+      "learning_rate": 0.0001,
+      "loss": 0.6087,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6117677286742035,
+      "grad_norm": 1.0398849248886108,
+      "learning_rate": 0.0001,
+      "loss": 0.7742,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6118533744433025,
+      "grad_norm": 1.2732462882995605,
+      "learning_rate": 0.0001,
+      "loss": 1.0825,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6119390202124015,
+      "grad_norm": 1.1682463884353638,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6120246659815005,
+      "grad_norm": 0.8902857899665833,
+      "learning_rate": 0.0001,
+      "loss": 1.0743,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6121103117505995,
+      "grad_norm": 1.1598684787750244,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6121959575196986,
+      "grad_norm": 0.7648635506629944,
+      "learning_rate": 0.0001,
+      "loss": 1.0364,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6122816032887975,
+      "grad_norm": 0.9208410382270813,
+      "learning_rate": 0.0001,
+      "loss": 1.0224,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6123672490578965,
+      "grad_norm": 1.037514328956604,
+      "learning_rate": 0.0001,
+      "loss": 0.8373,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6124528948269955,
+      "grad_norm": 0.6622211337089539,
+      "learning_rate": 0.0001,
+      "loss": 0.4733,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6125385405960946,
+      "grad_norm": 1.0057358741760254,
+      "learning_rate": 0.0001,
+      "loss": 1.1038,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6126241863651936,
+      "grad_norm": 0.7321262359619141,
+      "learning_rate": 0.0001,
+      "loss": 0.6355,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6127098321342925,
+      "grad_norm": 0.226361483335495,
+      "learning_rate": 0.0001,
+      "loss": 0.2565,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6127954779033916,
+      "grad_norm": 0.571540892124176,
+      "learning_rate": 0.0001,
+      "loss": 1.1196,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6128811236724906,
+      "grad_norm": 1.099760890007019,
+      "learning_rate": 0.0001,
+      "loss": 0.7782,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6129667694415896,
+      "grad_norm": 0.504781186580658,
+      "learning_rate": 0.0001,
+      "loss": 0.821,
+      "step": 7157
+    },
+    {
+      "epoch": 0.6130524152106885,
+      "grad_norm": 0.7961287498474121,
+      "learning_rate": 0.0001,
+      "loss": 0.9275,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6131380609797876,
+      "grad_norm": 0.579876720905304,
+      "learning_rate": 0.0001,
+      "loss": 0.4977,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6132237067488866,
+      "grad_norm": 0.8673494458198547,
+      "learning_rate": 0.0001,
+      "loss": 1.0013,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6133093525179856,
+      "grad_norm": 0.7289370894432068,
+      "learning_rate": 0.0001,
+      "loss": 1.0272,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6133949982870847,
+      "grad_norm": 0.7599378228187561,
+      "learning_rate": 0.0001,
+      "loss": 0.7347,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6134806440561836,
+      "grad_norm": 0.6954655647277832,
+      "learning_rate": 0.0001,
+      "loss": 1.2608,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6135662898252826,
+      "grad_norm": 0.8425039052963257,
+      "learning_rate": 0.0001,
+      "loss": 1.1486,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6136519355943817,
+      "grad_norm": 0.30655816197395325,
+      "learning_rate": 0.0001,
+      "loss": 0.1989,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6137375813634807,
+      "grad_norm": 1.5734928846359253,
+      "learning_rate": 0.0001,
+      "loss": 1.1702,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6138232271325796,
+      "grad_norm": 0.9099913835525513,
+      "learning_rate": 0.0001,
+      "loss": 0.8162,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6139088729016786,
+      "grad_norm": 0.8867801427841187,
+      "learning_rate": 0.0001,
+      "loss": 0.5339,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6139945186707777,
+      "grad_norm": 0.5865547060966492,
+      "learning_rate": 0.0001,
+      "loss": 0.8718,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6140801644398767,
+      "grad_norm": 1.0975793600082397,
+      "learning_rate": 0.0001,
+      "loss": 1.0156,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6141658102089756,
+      "grad_norm": 0.5222142934799194,
+      "learning_rate": 0.0001,
+      "loss": 0.6859,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6142514559780747,
+      "grad_norm": 1.015579104423523,
+      "learning_rate": 0.0001,
+      "loss": 1.0181,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6143371017471737,
+      "grad_norm": 1.209083914756775,
+      "learning_rate": 0.0001,
+      "loss": 1.2008,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6144227475162727,
+      "grad_norm": 0.7028567790985107,
+      "learning_rate": 0.0001,
+      "loss": 0.8519,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6145083932853717,
+      "grad_norm": 0.7440289855003357,
+      "learning_rate": 0.0001,
+      "loss": 1.2532,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6145940390544707,
+      "grad_norm": 0.38341522216796875,
+      "learning_rate": 0.0001,
+      "loss": 0.4837,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6146796848235697,
+      "grad_norm": 0.7631319165229797,
+      "learning_rate": 0.0001,
+      "loss": 0.8615,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6147653305926687,
+      "grad_norm": 0.7053281664848328,
+      "learning_rate": 0.0001,
+      "loss": 0.5975,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6148509763617678,
+      "grad_norm": 0.9238978624343872,
+      "learning_rate": 0.0001,
+      "loss": 0.8076,
+      "step": 7179
+    },
+    {
+      "epoch": 0.6149366221308668,
+      "grad_norm": 0.5642279982566833,
+      "learning_rate": 0.0001,
+      "loss": 0.488,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6150222678999657,
+      "grad_norm": 0.8314937949180603,
+      "learning_rate": 0.0001,
+      "loss": 0.9685,
+      "step": 7181
+    },
+    {
+      "epoch": 0.6151079136690647,
+      "grad_norm": 0.66522216796875,
+      "learning_rate": 0.0001,
+      "loss": 0.9569,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6151935594381638,
+      "grad_norm": 0.6167471408843994,
+      "learning_rate": 0.0001,
+      "loss": 1.1612,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6152792052072628,
+      "grad_norm": 0.9965395331382751,
+      "learning_rate": 0.0001,
+      "loss": 0.6568,
+      "step": 7184
+    },
+    {
+      "epoch": 0.6153648509763617,
+      "grad_norm": 0.884418785572052,
+      "learning_rate": 0.0001,
+      "loss": 1.1636,
+      "step": 7185
+    },
+    {
+      "epoch": 0.6154504967454608,
+      "grad_norm": 1.1195147037506104,
+      "learning_rate": 0.0001,
+      "loss": 0.9478,
+      "step": 7186
+    },
+    {
+      "epoch": 0.6155361425145598,
+      "grad_norm": 0.5688483119010925,
+      "learning_rate": 0.0001,
+      "loss": 0.5016,
+      "step": 7187
+    },
+    {
+      "epoch": 0.6156217882836588,
+      "grad_norm": 0.7726809978485107,
+      "learning_rate": 0.0001,
+      "loss": 0.7425,
+      "step": 7188
+    },
+    {
+      "epoch": 0.6157074340527577,
+      "grad_norm": 0.6762937903404236,
+      "learning_rate": 0.0001,
+      "loss": 0.4836,
+      "step": 7189
+    },
+    {
+      "epoch": 0.6157930798218568,
+      "grad_norm": 1.057231068611145,
+      "learning_rate": 0.0001,
+      "loss": 1.2436,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6158787255909558,
+      "grad_norm": 0.8360273241996765,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6159643713600548,
+      "grad_norm": 1.0110887289047241,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6160500171291539,
+      "grad_norm": 0.4690276086330414,
+      "learning_rate": 0.0001,
+      "loss": 0.2162,
+      "step": 7193
+    },
+    {
+      "epoch": 0.6161356628982528,
+      "grad_norm": 0.6942659616470337,
+      "learning_rate": 0.0001,
+      "loss": 0.8077,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6162213086673518,
+      "grad_norm": 1.4454057216644287,
+      "learning_rate": 0.0001,
+      "loss": 0.3929,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6163069544364509,
+      "grad_norm": 0.960157573223114,
+      "learning_rate": 0.0001,
+      "loss": 0.6802,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6163926002055499,
+      "grad_norm": 0.8453572988510132,
+      "learning_rate": 0.0001,
+      "loss": 1.1433,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6164782459746488,
+      "grad_norm": 0.3582194745540619,
+      "learning_rate": 0.0001,
+      "loss": 0.1595,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6165638917437478,
+      "grad_norm": 0.7610119581222534,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 7199
+    },
+    {
+      "epoch": 0.6166495375128469,
+      "grad_norm": 0.894481897354126,
+      "learning_rate": 0.0001,
+      "loss": 0.825,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6167351832819459,
+      "grad_norm": 1.01326584815979,
+      "learning_rate": 0.0001,
+      "loss": 1.2323,
+      "step": 7201
+    },
+    {
+      "epoch": 0.6168208290510448,
+      "grad_norm": 0.8442440032958984,
+      "learning_rate": 0.0001,
+      "loss": 0.9109,
+      "step": 7202
+    },
+    {
+      "epoch": 0.6169064748201439,
+      "grad_norm": 0.6756128072738647,
+      "learning_rate": 0.0001,
+      "loss": 1.2671,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6169921205892429,
+      "grad_norm": 0.9442592263221741,
+      "learning_rate": 0.0001,
+      "loss": 0.8245,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6170777663583419,
+      "grad_norm": 0.5541669130325317,
+      "learning_rate": 0.0001,
+      "loss": 0.8219,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6171634121274409,
+      "grad_norm": 0.9895410537719727,
+      "learning_rate": 0.0001,
+      "loss": 1.1134,
+      "step": 7206
+    },
+    {
+      "epoch": 0.6172490578965399,
+      "grad_norm": 0.9832032322883606,
+      "learning_rate": 0.0001,
+      "loss": 1.057,
+      "step": 7207
+    },
+    {
+      "epoch": 0.6173347036656389,
+      "grad_norm": 1.0303969383239746,
+      "learning_rate": 0.0001,
+      "loss": 0.9135,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6174203494347379,
+      "grad_norm": 0.8384649157524109,
+      "learning_rate": 0.0001,
+      "loss": 1.0708,
+      "step": 7209
+    },
+    {
+      "epoch": 0.617505995203837,
+      "grad_norm": 1.1098719835281372,
+      "learning_rate": 0.0001,
+      "loss": 1.1875,
+      "step": 7210
+    },
+    {
+      "epoch": 0.617591640972936,
+      "grad_norm": 1.1680737733840942,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6176772867420349,
+      "grad_norm": 0.6825666427612305,
+      "learning_rate": 0.0001,
+      "loss": 0.5707,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6177629325111339,
+      "grad_norm": 0.9996074438095093,
+      "learning_rate": 0.0001,
+      "loss": 0.6612,
+      "step": 7213
+    },
+    {
+      "epoch": 0.617848578280233,
+      "grad_norm": 0.983212411403656,
+      "learning_rate": 0.0001,
+      "loss": 1.2871,
+      "step": 7214
+    },
+    {
+      "epoch": 0.617934224049332,
+      "grad_norm": 0.5123722553253174,
+      "learning_rate": 0.0001,
+      "loss": 0.6603,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6180198698184309,
+      "grad_norm": 0.872926652431488,
+      "learning_rate": 0.0001,
+      "loss": 0.6524,
+      "step": 7216
+    },
+    {
+      "epoch": 0.61810551558753,
+      "grad_norm": 0.6033281683921814,
+      "learning_rate": 0.0001,
+      "loss": 0.664,
+      "step": 7217
+    },
+    {
+      "epoch": 0.618191161356629,
+      "grad_norm": 0.6720301508903503,
+      "learning_rate": 0.0001,
+      "loss": 0.7087,
+      "step": 7218
+    },
+    {
+      "epoch": 0.618276807125728,
+      "grad_norm": 0.7276901006698608,
+      "learning_rate": 0.0001,
+      "loss": 1.0013,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6183624528948269,
+      "grad_norm": 0.9375104904174805,
+      "learning_rate": 0.0001,
+      "loss": 0.709,
+      "step": 7220
+    },
+    {
+      "epoch": 0.618448098663926,
+      "grad_norm": 0.687114953994751,
+      "learning_rate": 0.0001,
+      "loss": 0.7471,
+      "step": 7221
+    },
+    {
+      "epoch": 0.618533744433025,
+      "grad_norm": 0.706307590007782,
+      "learning_rate": 0.0001,
+      "loss": 0.6594,
+      "step": 7222
+    },
+    {
+      "epoch": 0.618619390202124,
+      "grad_norm": 0.8839142322540283,
+      "learning_rate": 0.0001,
+      "loss": 1.1197,
+      "step": 7223
+    },
+    {
+      "epoch": 0.6187050359712231,
+      "grad_norm": 0.7893907427787781,
+      "learning_rate": 0.0001,
+      "loss": 0.3282,
+      "step": 7224
+    },
+    {
+      "epoch": 0.618790681740322,
+      "grad_norm": 0.7214511036872864,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 7225
+    },
+    {
+      "epoch": 0.618876327509421,
+      "grad_norm": 0.6083884835243225,
+      "learning_rate": 0.0001,
+      "loss": 0.6134,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6189619732785201,
+      "grad_norm": 0.4049798846244812,
+      "learning_rate": 0.0001,
+      "loss": 0.5971,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 1.0939589738845825,
+      "learning_rate": 0.0001,
+      "loss": 1.1093,
+      "step": 7228
+    },
+    {
+      "epoch": 0.619133264816718,
+      "grad_norm": 0.9432108402252197,
+      "learning_rate": 0.0001,
+      "loss": 0.9369,
+      "step": 7229
+    },
+    {
+      "epoch": 0.619218910585817,
+      "grad_norm": 1.7669771909713745,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6193045563549161,
+      "grad_norm": 0.8504917621612549,
+      "learning_rate": 0.0001,
+      "loss": 1.2348,
+      "step": 7231
+    },
+    {
+      "epoch": 0.6193902021240151,
+      "grad_norm": 0.8311629891395569,
+      "learning_rate": 0.0001,
+      "loss": 0.9874,
+      "step": 7232
+    },
+    {
+      "epoch": 0.619475847893114,
+      "grad_norm": 1.1124396324157715,
+      "learning_rate": 0.0001,
+      "loss": 1.1588,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6195614936622131,
+      "grad_norm": 0.6862830519676208,
+      "learning_rate": 0.0001,
+      "loss": 0.5001,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6196471394313121,
+      "grad_norm": 1.1187832355499268,
+      "learning_rate": 0.0001,
+      "loss": 0.9407,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6197327852004111,
+      "grad_norm": 1.0536829233169556,
+      "learning_rate": 0.0001,
+      "loss": 0.6806,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6198184309695101,
+      "grad_norm": 0.3369216024875641,
+      "learning_rate": 0.0001,
+      "loss": 0.4836,
+      "step": 7237
+    },
+    {
+      "epoch": 0.6199040767386091,
+      "grad_norm": 0.8195139765739441,
+      "learning_rate": 0.0001,
+      "loss": 1.0299,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6199897225077081,
+      "grad_norm": 0.833943784236908,
+      "learning_rate": 0.0001,
+      "loss": 1.2882,
+      "step": 7239
+    },
+    {
+      "epoch": 0.6200753682768071,
+      "grad_norm": 0.9482589364051819,
+      "learning_rate": 0.0001,
+      "loss": 0.9488,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6201610140459062,
+      "grad_norm": 1.106203556060791,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6202466598150052,
+      "grad_norm": 0.7302289605140686,
+      "learning_rate": 0.0001,
+      "loss": 0.4631,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6203323055841041,
+      "grad_norm": 1.0314648151397705,
+      "learning_rate": 0.0001,
+      "loss": 0.7846,
+      "step": 7243
+    },
+    {
+      "epoch": 0.6204179513532031,
+      "grad_norm": 0.7653214931488037,
+      "learning_rate": 0.0001,
+      "loss": 0.52,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6205035971223022,
+      "grad_norm": 0.673358142375946,
+      "learning_rate": 0.0001,
+      "loss": 0.5636,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6205892428914012,
+      "grad_norm": 1.1547832489013672,
+      "learning_rate": 0.0001,
+      "loss": 1.3476,
+      "step": 7246
+    },
+    {
+      "epoch": 0.6206748886605001,
+      "grad_norm": 0.7195034027099609,
+      "learning_rate": 0.0001,
+      "loss": 1.147,
+      "step": 7247
+    },
+    {
+      "epoch": 0.6207605344295992,
+      "grad_norm": 0.5641812086105347,
+      "learning_rate": 0.0001,
+      "loss": 0.9891,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6208461801986982,
+      "grad_norm": 0.48762643337249756,
+      "learning_rate": 0.0001,
+      "loss": 0.5225,
+      "step": 7249
+    },
+    {
+      "epoch": 0.6209318259677972,
+      "grad_norm": 1.2486449480056763,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6210174717368961,
+      "grad_norm": 1.121297001838684,
+      "learning_rate": 0.0001,
+      "loss": 1.0638,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6211031175059952,
+      "grad_norm": 1.1462829113006592,
+      "learning_rate": 0.0001,
+      "loss": 0.8541,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6211887632750942,
+      "grad_norm": 0.792739748954773,
+      "learning_rate": 0.0001,
+      "loss": 0.7871,
+      "step": 7253
+    },
+    {
+      "epoch": 0.6212744090441932,
+      "grad_norm": 0.6997836232185364,
+      "learning_rate": 0.0001,
+      "loss": 0.3691,
+      "step": 7254
+    },
+    {
+      "epoch": 0.6213600548132923,
+      "grad_norm": 0.8574251532554626,
+      "learning_rate": 0.0001,
+      "loss": 0.9438,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6214457005823912,
+      "grad_norm": 0.6510566473007202,
+      "learning_rate": 0.0001,
+      "loss": 0.7766,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6215313463514902,
+      "grad_norm": 1.0315278768539429,
+      "learning_rate": 0.0001,
+      "loss": 0.9207,
+      "step": 7257
+    },
+    {
+      "epoch": 0.6216169921205893,
+      "grad_norm": 1.0000832080841064,
+      "learning_rate": 0.0001,
+      "loss": 1.0963,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6217026378896883,
+      "grad_norm": 0.719559907913208,
+      "learning_rate": 0.0001,
+      "loss": 0.7616,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6217882836587872,
+      "grad_norm": 1.1377297639846802,
+      "learning_rate": 0.0001,
+      "loss": 0.909,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6218739294278862,
+      "grad_norm": 0.32465749979019165,
+      "learning_rate": 0.0001,
+      "loss": 0.4749,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6219595751969853,
+      "grad_norm": 1.0465503931045532,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 7262
+    },
+    {
+      "epoch": 0.6220452209660843,
+      "grad_norm": 0.6171438097953796,
+      "learning_rate": 0.0001,
+      "loss": 0.7559,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6221308667351833,
+      "grad_norm": 0.6180439591407776,
+      "learning_rate": 0.0001,
+      "loss": 0.9453,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6222165125042823,
+      "grad_norm": 0.5517540574073792,
+      "learning_rate": 0.0001,
+      "loss": 0.9553,
+      "step": 7265
+    },
+    {
+      "epoch": 0.6223021582733813,
+      "grad_norm": 0.9106813669204712,
+      "learning_rate": 0.0001,
+      "loss": 1.186,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6223878040424803,
+      "grad_norm": 0.5654793381690979,
+      "learning_rate": 0.0001,
+      "loss": 0.9353,
+      "step": 7267
+    },
+    {
+      "epoch": 0.6224734498115793,
+      "grad_norm": 0.8907931447029114,
+      "learning_rate": 0.0001,
+      "loss": 0.7696,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6225590955806783,
+      "grad_norm": 0.6461849212646484,
+      "learning_rate": 0.0001,
+      "loss": 0.49,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6226447413497773,
+      "grad_norm": 1.1702014207839966,
+      "learning_rate": 0.0001,
+      "loss": 0.9742,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6227303871188763,
+      "grad_norm": 1.1390634775161743,
+      "learning_rate": 0.0001,
+      "loss": 1.1026,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6228160328879754,
+      "grad_norm": 1.0720528364181519,
+      "learning_rate": 0.0001,
+      "loss": 1.2061,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6229016786570744,
+      "grad_norm": 0.5823259353637695,
+      "learning_rate": 0.0001,
+      "loss": 0.4547,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6229873244261733,
+      "grad_norm": 1.0503008365631104,
+      "learning_rate": 0.0001,
+      "loss": 1.1078,
+      "step": 7274
+    },
+    {
+      "epoch": 0.6230729701952723,
+      "grad_norm": 0.5006273984909058,
+      "learning_rate": 0.0001,
+      "loss": 0.9312,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6231586159643714,
+      "grad_norm": 1.029545783996582,
+      "learning_rate": 0.0001,
+      "loss": 0.9696,
+      "step": 7276
+    },
+    {
+      "epoch": 0.6232442617334704,
+      "grad_norm": 1.0128273963928223,
+      "learning_rate": 0.0001,
+      "loss": 0.8578,
+      "step": 7277
+    },
+    {
+      "epoch": 0.6233299075025693,
+      "grad_norm": 0.9018886089324951,
+      "learning_rate": 0.0001,
+      "loss": 1.2223,
+      "step": 7278
+    },
+    {
+      "epoch": 0.6234155532716684,
+      "grad_norm": 0.41331368684768677,
+      "learning_rate": 0.0001,
+      "loss": 0.275,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6235011990407674,
+      "grad_norm": 0.880259096622467,
+      "learning_rate": 0.0001,
+      "loss": 1.1886,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6235868448098664,
+      "grad_norm": 0.6082117557525635,
+      "learning_rate": 0.0001,
+      "loss": 1.2372,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6236724905789653,
+      "grad_norm": 0.6120486855506897,
+      "learning_rate": 0.0001,
+      "loss": 0.2871,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6237581363480644,
+      "grad_norm": 0.8310970067977905,
+      "learning_rate": 0.0001,
+      "loss": 0.7091,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6238437821171634,
+      "grad_norm": 0.514252781867981,
+      "learning_rate": 0.0001,
+      "loss": 0.873,
+      "step": 7284
+    },
+    {
+      "epoch": 0.6239294278862624,
+      "grad_norm": 0.7923002243041992,
+      "learning_rate": 0.0001,
+      "loss": 0.6931,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6240150736553615,
+      "grad_norm": 1.165787935256958,
+      "learning_rate": 0.0001,
+      "loss": 1.078,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6241007194244604,
+      "grad_norm": 0.6720137000083923,
+      "learning_rate": 0.0001,
+      "loss": 0.96,
+      "step": 7287
+    },
+    {
+      "epoch": 0.6241863651935594,
+      "grad_norm": 0.19401752948760986,
+      "learning_rate": 0.0001,
+      "loss": 0.2032,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6242720109626585,
+      "grad_norm": 0.8464502096176147,
+      "learning_rate": 0.0001,
+      "loss": 0.7548,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6243576567317575,
+      "grad_norm": 1.2159130573272705,
+      "learning_rate": 0.0001,
+      "loss": 0.7848,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6244433025008564,
+      "grad_norm": 0.7477132678031921,
+      "learning_rate": 0.0001,
+      "loss": 0.9056,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6245289482699554,
+      "grad_norm": 0.9693397283554077,
+      "learning_rate": 0.0001,
+      "loss": 0.6619,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6246145940390545,
+      "grad_norm": 0.6880479454994202,
+      "learning_rate": 0.0001,
+      "loss": 0.8111,
+      "step": 7293
+    },
+    {
+      "epoch": 0.6247002398081535,
+      "grad_norm": 0.8869489431381226,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6247858855772525,
+      "grad_norm": 0.7024097442626953,
+      "learning_rate": 0.0001,
+      "loss": 0.6213,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6248715313463515,
+      "grad_norm": 0.7150831818580627,
+      "learning_rate": 0.0001,
+      "loss": 1.1473,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6249571771154505,
+      "grad_norm": 1.9632437229156494,
+      "learning_rate": 0.0001,
+      "loss": 1.1654,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6250428228845495,
+      "grad_norm": 1.0561491250991821,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6251284686536485,
+      "grad_norm": 0.9382666349411011,
+      "learning_rate": 0.0001,
+      "loss": 0.9653,
+      "step": 7299
+    },
+    {
+      "epoch": 0.6252141144227475,
+      "grad_norm": 0.8456935882568359,
+      "learning_rate": 0.0001,
+      "loss": 0.8599,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6252997601918465,
+      "grad_norm": 0.944604218006134,
+      "learning_rate": 0.0001,
+      "loss": 0.8975,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6253854059609455,
+      "grad_norm": 0.9482534527778625,
+      "learning_rate": 0.0001,
+      "loss": 0.9318,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6254710517300446,
+      "grad_norm": 0.6836103200912476,
+      "learning_rate": 0.0001,
+      "loss": 1.119,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6255566974991436,
+      "grad_norm": 1.542005181312561,
+      "learning_rate": 0.0001,
+      "loss": 1.026,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6256423432682425,
+      "grad_norm": 0.8735882639884949,
+      "learning_rate": 0.0001,
+      "loss": 1.434,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6257279890373415,
+      "grad_norm": 0.8581831455230713,
+      "learning_rate": 0.0001,
+      "loss": 0.7168,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6258136348064406,
+      "grad_norm": 0.7944645285606384,
+      "learning_rate": 0.0001,
+      "loss": 0.9536,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6258992805755396,
+      "grad_norm": 1.0766985416412354,
+      "learning_rate": 0.0001,
+      "loss": 1.1833,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6259849263446385,
+      "grad_norm": 0.5451133251190186,
+      "learning_rate": 0.0001,
+      "loss": 0.5988,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6260705721137376,
+      "grad_norm": 0.6243048310279846,
+      "learning_rate": 0.0001,
+      "loss": 0.8781,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6261562178828366,
+      "grad_norm": 0.6374447345733643,
+      "learning_rate": 0.0001,
+      "loss": 0.7196,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6262418636519356,
+      "grad_norm": 0.6957812905311584,
+      "learning_rate": 0.0001,
+      "loss": 1.1663,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6263275094210347,
+      "grad_norm": 0.9047351479530334,
+      "learning_rate": 0.0001,
+      "loss": 0.6905,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6264131551901336,
+      "grad_norm": 1.3206897974014282,
+      "learning_rate": 0.0001,
+      "loss": 1.2293,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6264988009592326,
+      "grad_norm": 1.1333189010620117,
+      "learning_rate": 0.0001,
+      "loss": 0.9785,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6265844467283316,
+      "grad_norm": 1.1130790710449219,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6266700924974307,
+      "grad_norm": 0.7904829978942871,
+      "learning_rate": 0.0001,
+      "loss": 1.1144,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6267557382665296,
+      "grad_norm": 0.9784208536148071,
+      "learning_rate": 0.0001,
+      "loss": 1.065,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6268413840356286,
+      "grad_norm": 0.5981972217559814,
+      "learning_rate": 0.0001,
+      "loss": 0.8618,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6269270298047277,
+      "grad_norm": 0.7238190770149231,
+      "learning_rate": 0.0001,
+      "loss": 1.2934,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6270126755738267,
+      "grad_norm": 0.5128633975982666,
+      "learning_rate": 0.0001,
+      "loss": 0.6308,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6270983213429256,
+      "grad_norm": 0.9749186038970947,
+      "learning_rate": 0.0001,
+      "loss": 1.2687,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6271839671120246,
+      "grad_norm": 1.3764547109603882,
+      "learning_rate": 0.0001,
+      "loss": 0.9594,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6272696128811237,
+      "grad_norm": 1.4454389810562134,
+      "learning_rate": 0.0001,
+      "loss": 1.1418,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6273552586502227,
+      "grad_norm": 1.0257917642593384,
+      "learning_rate": 0.0001,
+      "loss": 1.0125,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6274409044193217,
+      "grad_norm": 0.7528908848762512,
+      "learning_rate": 0.0001,
+      "loss": 0.7426,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6275265501884207,
+      "grad_norm": 0.7032840251922607,
+      "learning_rate": 0.0001,
+      "loss": 1.0155,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6276121959575197,
+      "grad_norm": 0.9555079936981201,
+      "learning_rate": 0.0001,
+      "loss": 1.0487,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6276978417266187,
+      "grad_norm": 0.5304853320121765,
+      "learning_rate": 0.0001,
+      "loss": 1.0575,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6277834874957177,
+      "grad_norm": 0.5691076517105103,
+      "learning_rate": 0.0001,
+      "loss": 0.7654,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6278691332648167,
+      "grad_norm": 0.9598084092140198,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6279547790339157,
+      "grad_norm": 1.1162930727005005,
+      "learning_rate": 0.0001,
+      "loss": 0.9551,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6280404248030147,
+      "grad_norm": 0.853754997253418,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6281260705721138,
+      "grad_norm": 0.7944943904876709,
+      "learning_rate": 0.0001,
+      "loss": 0.8978,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6282117163412128,
+      "grad_norm": 0.7785724401473999,
+      "learning_rate": 0.0001,
+      "loss": 0.8316,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6282973621103117,
+      "grad_norm": 0.8534624576568604,
+      "learning_rate": 0.0001,
+      "loss": 1.3805,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6283830078794107,
+      "grad_norm": 0.6261518597602844,
+      "learning_rate": 0.0001,
+      "loss": 0.61,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6284686536485098,
+      "grad_norm": 1.095715045928955,
+      "learning_rate": 0.0001,
+      "loss": 1.1366,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6285542994176088,
+      "grad_norm": 0.9035392999649048,
+      "learning_rate": 0.0001,
+      "loss": 1.2894,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6286399451867077,
+      "grad_norm": 0.7569425106048584,
+      "learning_rate": 0.0001,
+      "loss": 0.5753,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6287255909558068,
+      "grad_norm": 0.9357876777648926,
+      "learning_rate": 0.0001,
+      "loss": 0.5166,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6288112367249058,
+      "grad_norm": 0.85343998670578,
+      "learning_rate": 0.0001,
+      "loss": 0.9876,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6288968824940048,
+      "grad_norm": 1.0652940273284912,
+      "learning_rate": 0.0001,
+      "loss": 1.1698,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6289825282631039,
+      "grad_norm": 1.0429601669311523,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6290681740322028,
+      "grad_norm": 0.9348326325416565,
+      "learning_rate": 0.0001,
+      "loss": 0.6106,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6291538198013018,
+      "grad_norm": 0.7339234948158264,
+      "learning_rate": 0.0001,
+      "loss": 0.6208,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6292394655704008,
+      "grad_norm": 0.6307599544525146,
+      "learning_rate": 0.0001,
+      "loss": 1.0833,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6293251113394999,
+      "grad_norm": 0.5872017741203308,
+      "learning_rate": 0.0001,
+      "loss": 1.0152,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6294107571085988,
+      "grad_norm": 0.4902920722961426,
+      "learning_rate": 0.0001,
+      "loss": 0.5918,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6294964028776978,
+      "grad_norm": 0.8303737640380859,
+      "learning_rate": 0.0001,
+      "loss": 0.4391,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6295820486467969,
+      "grad_norm": 0.6459687948226929,
+      "learning_rate": 0.0001,
+      "loss": 0.6522,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6296676944158959,
+      "grad_norm": 0.9672994613647461,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6297533401849948,
+      "grad_norm": 0.92167067527771,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6298389859540938,
+      "grad_norm": 0.5689713954925537,
+      "learning_rate": 0.0001,
+      "loss": 0.4364,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6299246317231929,
+      "grad_norm": 0.8699139356613159,
+      "learning_rate": 0.0001,
+      "loss": 1.1094,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6300102774922919,
+      "grad_norm": 0.6037547588348389,
+      "learning_rate": 0.0001,
+      "loss": 0.9766,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6300959232613909,
+      "grad_norm": 0.8530517220497131,
+      "learning_rate": 0.0001,
+      "loss": 0.6552,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6301815690304899,
+      "grad_norm": 0.5515087842941284,
+      "learning_rate": 0.0001,
+      "loss": 0.4386,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6302672147995889,
+      "grad_norm": 0.5513121485710144,
+      "learning_rate": 0.0001,
+      "loss": 1.2706,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6303528605686879,
+      "grad_norm": 0.8420184254646301,
+      "learning_rate": 0.0001,
+      "loss": 0.8825,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6304385063377869,
+      "grad_norm": 0.5805354714393616,
+      "learning_rate": 0.0001,
+      "loss": 0.3553,
+      "step": 7361
+    },
+    {
+      "epoch": 0.630524152106886,
+      "grad_norm": 0.812544584274292,
+      "learning_rate": 0.0001,
+      "loss": 1.1054,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6306097978759849,
+      "grad_norm": 0.6569427847862244,
+      "learning_rate": 0.0001,
+      "loss": 0.5631,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6306954436450839,
+      "grad_norm": 0.7859117984771729,
+      "learning_rate": 0.0001,
+      "loss": 0.7183,
+      "step": 7364
+    },
+    {
+      "epoch": 0.630781089414183,
+      "grad_norm": 0.6776217818260193,
+      "learning_rate": 0.0001,
+      "loss": 1.073,
+      "step": 7365
+    },
+    {
+      "epoch": 0.630866735183282,
+      "grad_norm": 0.5223105549812317,
+      "learning_rate": 0.0001,
+      "loss": 0.6563,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6309523809523809,
+      "grad_norm": 0.9593638181686401,
+      "learning_rate": 0.0001,
+      "loss": 1.3763,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6310380267214799,
+      "grad_norm": 0.6519347429275513,
+      "learning_rate": 0.0001,
+      "loss": 0.4449,
+      "step": 7368
+    },
+    {
+      "epoch": 0.631123672490579,
+      "grad_norm": 1.2849982976913452,
+      "learning_rate": 0.0001,
+      "loss": 0.8567,
+      "step": 7369
+    },
+    {
+      "epoch": 0.631209318259678,
+      "grad_norm": 1.1017799377441406,
+      "learning_rate": 0.0001,
+      "loss": 1.2324,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6312949640287769,
+      "grad_norm": 0.6981381177902222,
+      "learning_rate": 0.0001,
+      "loss": 1.1624,
+      "step": 7371
+    },
+    {
+      "epoch": 0.631380609797876,
+      "grad_norm": 0.6126965880393982,
+      "learning_rate": 0.0001,
+      "loss": 0.3369,
+      "step": 7372
+    },
+    {
+      "epoch": 0.631466255566975,
+      "grad_norm": 1.0516971349716187,
+      "learning_rate": 0.0001,
+      "loss": 1.0756,
+      "step": 7373
+    },
+    {
+      "epoch": 0.631551901336074,
+      "grad_norm": 0.7178974747657776,
+      "learning_rate": 0.0001,
+      "loss": 0.9854,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6316375471051731,
+      "grad_norm": 0.6332042217254639,
+      "learning_rate": 0.0001,
+      "loss": 0.951,
+      "step": 7375
+    },
+    {
+      "epoch": 0.631723192874272,
+      "grad_norm": 0.9185196161270142,
+      "learning_rate": 0.0001,
+      "loss": 0.9122,
+      "step": 7376
+    },
+    {
+      "epoch": 0.631808838643371,
+      "grad_norm": 0.5692428946495056,
+      "learning_rate": 0.0001,
+      "loss": 0.693,
+      "step": 7377
+    },
+    {
+      "epoch": 0.63189448441247,
+      "grad_norm": 1.2323869466781616,
+      "learning_rate": 0.0001,
+      "loss": 1.1968,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6319801301815691,
+      "grad_norm": 0.9828149676322937,
+      "learning_rate": 0.0001,
+      "loss": 0.9072,
+      "step": 7379
+    },
+    {
+      "epoch": 0.632065775950668,
+      "grad_norm": 0.4714638590812683,
+      "learning_rate": 0.0001,
+      "loss": 0.787,
+      "step": 7380
+    },
+    {
+      "epoch": 0.632151421719767,
+      "grad_norm": 1.1425918340682983,
+      "learning_rate": 0.0001,
+      "loss": 0.9399,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6322370674888661,
+      "grad_norm": 1.0284031629562378,
+      "learning_rate": 0.0001,
+      "loss": 0.8167,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6323227132579651,
+      "grad_norm": 0.8637199401855469,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 7383
+    },
+    {
+      "epoch": 0.632408359027064,
+      "grad_norm": 0.5582469701766968,
+      "learning_rate": 0.0001,
+      "loss": 0.7992,
+      "step": 7384
+    },
+    {
+      "epoch": 0.632494004796163,
+      "grad_norm": 0.672616183757782,
+      "learning_rate": 0.0001,
+      "loss": 1.0797,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6325796505652621,
+      "grad_norm": 0.8128583431243896,
+      "learning_rate": 0.0001,
+      "loss": 0.7594,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6326652963343611,
+      "grad_norm": 0.6889652013778687,
+      "learning_rate": 0.0001,
+      "loss": 1.0865,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6327509421034601,
+      "grad_norm": 1.023848056793213,
+      "learning_rate": 0.0001,
+      "loss": 1.0126,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6328365878725591,
+      "grad_norm": 0.8641898036003113,
+      "learning_rate": 0.0001,
+      "loss": 1.2343,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6329222336416581,
+      "grad_norm": 1.0266669988632202,
+      "learning_rate": 0.0001,
+      "loss": 0.6922,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6330078794107571,
+      "grad_norm": 1.0649604797363281,
+      "learning_rate": 0.0001,
+      "loss": 0.7216,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6330935251798561,
+      "grad_norm": 0.9045429825782776,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6331791709489552,
+      "grad_norm": 1.062514305114746,
+      "learning_rate": 0.0001,
+      "loss": 0.9122,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6332648167180541,
+      "grad_norm": 0.8040652871131897,
+      "learning_rate": 0.0001,
+      "loss": 0.8603,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6333504624871531,
+      "grad_norm": 0.7809293270111084,
+      "learning_rate": 0.0001,
+      "loss": 0.8078,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6334361082562522,
+      "grad_norm": 1.1721916198730469,
+      "learning_rate": 0.0001,
+      "loss": 0.8048,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6335217540253512,
+      "grad_norm": 0.863947331905365,
+      "learning_rate": 0.0001,
+      "loss": 1.027,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6336073997944501,
+      "grad_norm": 0.8445721864700317,
+      "learning_rate": 0.0001,
+      "loss": 1.1514,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6336930455635491,
+      "grad_norm": 0.8423464894294739,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6337786913326482,
+      "grad_norm": 0.8025368452072144,
+      "learning_rate": 0.0001,
+      "loss": 1.1468,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6338643371017472,
+      "grad_norm": 0.5804991126060486,
+      "learning_rate": 0.0001,
+      "loss": 0.5838,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6339499828708461,
+      "grad_norm": 0.8023805618286133,
+      "learning_rate": 0.0001,
+      "loss": 0.9896,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6340356286399452,
+      "grad_norm": 0.7305936813354492,
+      "learning_rate": 0.0001,
+      "loss": 0.4653,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6341212744090442,
+      "grad_norm": 0.7396258115768433,
+      "learning_rate": 0.0001,
+      "loss": 0.6654,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6342069201781432,
+      "grad_norm": 0.8341403007507324,
+      "learning_rate": 0.0001,
+      "loss": 1.0663,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6342925659472423,
+      "grad_norm": 1.0478709936141968,
+      "learning_rate": 0.0001,
+      "loss": 1.0967,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6343782117163412,
+      "grad_norm": 1.0175738334655762,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6344638574854402,
+      "grad_norm": 1.2262365818023682,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6345495032545392,
+      "grad_norm": 1.132767915725708,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6346351490236383,
+      "grad_norm": 0.5971997380256653,
+      "learning_rate": 0.0001,
+      "loss": 0.733,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6347207947927372,
+      "grad_norm": 0.5769020915031433,
+      "learning_rate": 0.0001,
+      "loss": 0.9043,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6348064405618362,
+      "grad_norm": 1.0269757509231567,
+      "learning_rate": 0.0001,
+      "loss": 1.8934,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6348920863309353,
+      "grad_norm": 1.5504083633422852,
+      "learning_rate": 0.0001,
+      "loss": 1.0811,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6349777321000343,
+      "grad_norm": 0.9799243807792664,
+      "learning_rate": 0.0001,
+      "loss": 1.021,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6350633778691332,
+      "grad_norm": 0.994730532169342,
+      "learning_rate": 0.0001,
+      "loss": 1.1171,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6351490236382322,
+      "grad_norm": 0.6667338013648987,
+      "learning_rate": 0.0001,
+      "loss": 1.2239,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6352346694073313,
+      "grad_norm": 0.6362695097923279,
+      "learning_rate": 0.0001,
+      "loss": 1.0909,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6353203151764303,
+      "grad_norm": 0.8649466633796692,
+      "learning_rate": 0.0001,
+      "loss": 0.8966,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6354059609455293,
+      "grad_norm": 0.6924716234207153,
+      "learning_rate": 0.0001,
+      "loss": 0.9796,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6354916067146283,
+      "grad_norm": 0.6474699974060059,
+      "learning_rate": 0.0001,
+      "loss": 1.309,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6355772524837273,
+      "grad_norm": 0.8882156014442444,
+      "learning_rate": 0.0001,
+      "loss": 0.9779,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6356628982528263,
+      "grad_norm": 0.9559507966041565,
+      "learning_rate": 0.0001,
+      "loss": 0.7813,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6357485440219253,
+      "grad_norm": 0.6466600298881531,
+      "learning_rate": 0.0001,
+      "loss": 1.1242,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6358341897910244,
+      "grad_norm": 0.6954374313354492,
+      "learning_rate": 0.0001,
+      "loss": 1.3581,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6359198355601233,
+      "grad_norm": 0.6046079993247986,
+      "learning_rate": 0.0001,
+      "loss": 0.7378,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6360054813292223,
+      "grad_norm": 0.7275225520133972,
+      "learning_rate": 0.0001,
+      "loss": 1.0283,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6360911270983214,
+      "grad_norm": 0.9381139278411865,
+      "learning_rate": 0.0001,
+      "loss": 0.9733,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6361767728674204,
+      "grad_norm": 0.6921767592430115,
+      "learning_rate": 0.0001,
+      "loss": 0.6811,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6362624186365193,
+      "grad_norm": 0.7208137512207031,
+      "learning_rate": 0.0001,
+      "loss": 0.9568,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6363480644056183,
+      "grad_norm": 0.7545326948165894,
+      "learning_rate": 0.0001,
+      "loss": 1.3066,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6364337101747174,
+      "grad_norm": 0.5571404099464417,
+      "learning_rate": 0.0001,
+      "loss": 0.9573,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6365193559438164,
+      "grad_norm": 0.8095611333847046,
+      "learning_rate": 0.0001,
+      "loss": 1.111,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6366050017129153,
+      "grad_norm": 1.1837818622589111,
+      "learning_rate": 0.0001,
+      "loss": 1.2526,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6366906474820144,
+      "grad_norm": 0.7577601671218872,
+      "learning_rate": 0.0001,
+      "loss": 1.1977,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6367762932511134,
+      "grad_norm": 0.5957587957382202,
+      "learning_rate": 0.0001,
+      "loss": 0.4555,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6368619390202124,
+      "grad_norm": 0.6851762533187866,
+      "learning_rate": 0.0001,
+      "loss": 0.5755,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6369475847893115,
+      "grad_norm": 1.1712689399719238,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6370332305584104,
+      "grad_norm": 1.220287799835205,
+      "learning_rate": 0.0001,
+      "loss": 1.4314,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6371188763275094,
+      "grad_norm": 0.9565568566322327,
+      "learning_rate": 0.0001,
+      "loss": 0.8757,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6372045220966084,
+      "grad_norm": 0.8043931126594543,
+      "learning_rate": 0.0001,
+      "loss": 0.7469,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6372901678657075,
+      "grad_norm": 0.581591010093689,
+      "learning_rate": 0.0001,
+      "loss": 0.84,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6373758136348064,
+      "grad_norm": 1.8764991760253906,
+      "learning_rate": 0.0001,
+      "loss": 0.9811,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6374614594039054,
+      "grad_norm": 0.41980600357055664,
+      "learning_rate": 0.0001,
+      "loss": 0.5857,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6375471051730045,
+      "grad_norm": 1.064895510673523,
+      "learning_rate": 0.0001,
+      "loss": 0.9462,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6376327509421035,
+      "grad_norm": 0.6478850245475769,
+      "learning_rate": 0.0001,
+      "loss": 1.1205,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6377183967112025,
+      "grad_norm": 0.6980164647102356,
+      "learning_rate": 0.0001,
+      "loss": 1.2451,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6378040424803014,
+      "grad_norm": 0.7937341332435608,
+      "learning_rate": 0.0001,
+      "loss": 1.0269,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6378896882494005,
+      "grad_norm": 0.7270026803016663,
+      "learning_rate": 0.0001,
+      "loss": 0.7549,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6379753340184995,
+      "grad_norm": 0.5944691300392151,
+      "learning_rate": 0.0001,
+      "loss": 0.827,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6380609797875985,
+      "grad_norm": 0.6812103986740112,
+      "learning_rate": 0.0001,
+      "loss": 0.5533,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6381466255566975,
+      "grad_norm": 0.8219323754310608,
+      "learning_rate": 0.0001,
+      "loss": 0.7708,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6382322713257965,
+      "grad_norm": 0.3026890456676483,
+      "learning_rate": 0.0001,
+      "loss": 0.3208,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6383179170948955,
+      "grad_norm": 0.8539815545082092,
+      "learning_rate": 0.0001,
+      "loss": 1.1613,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6384035628639945,
+      "grad_norm": 0.39192360639572144,
+      "learning_rate": 0.0001,
+      "loss": 0.6196,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6384892086330936,
+      "grad_norm": 1.445845365524292,
+      "learning_rate": 0.0001,
+      "loss": 1.4002,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6385748544021925,
+      "grad_norm": 0.5968790054321289,
+      "learning_rate": 0.0001,
+      "loss": 1.0062,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6386605001712915,
+      "grad_norm": 0.8331459760665894,
+      "learning_rate": 0.0001,
+      "loss": 1.2233,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6387461459403906,
+      "grad_norm": 1.224468469619751,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6388317917094896,
+      "grad_norm": 1.2109973430633545,
+      "learning_rate": 0.0001,
+      "loss": 1.1767,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6389174374785885,
+      "grad_norm": 1.3329466581344604,
+      "learning_rate": 0.0001,
+      "loss": 1.0604,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6390030832476875,
+      "grad_norm": 0.8457967638969421,
+      "learning_rate": 0.0001,
+      "loss": 1.078,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6390887290167866,
+      "grad_norm": 1.2611650228500366,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6391743747858856,
+      "grad_norm": 1.0636454820632935,
+      "learning_rate": 0.0001,
+      "loss": 0.9884,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6392600205549845,
+      "grad_norm": 0.4007361829280853,
+      "learning_rate": 0.0001,
+      "loss": 0.8511,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6393456663240836,
+      "grad_norm": 0.8870645761489868,
+      "learning_rate": 0.0001,
+      "loss": 1.3003,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6394313120931826,
+      "grad_norm": 0.5657684803009033,
+      "learning_rate": 0.0001,
+      "loss": 0.5828,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6395169578622816,
+      "grad_norm": 1.0877231359481812,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6396026036313807,
+      "grad_norm": 0.4250383973121643,
+      "learning_rate": 0.0001,
+      "loss": 0.1857,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6396882494004796,
+      "grad_norm": 0.6987583637237549,
+      "learning_rate": 0.0001,
+      "loss": 0.4909,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6397738951695786,
+      "grad_norm": 0.9640466570854187,
+      "learning_rate": 0.0001,
+      "loss": 1.2511,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6398595409386776,
+      "grad_norm": 0.4959496259689331,
+      "learning_rate": 0.0001,
+      "loss": 0.4056,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6399451867077767,
+      "grad_norm": 0.9061046838760376,
+      "learning_rate": 0.0001,
+      "loss": 1.0884,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6400308324768756,
+      "grad_norm": 0.5431190729141235,
+      "learning_rate": 0.0001,
+      "loss": 0.9354,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6401164782459746,
+      "grad_norm": 0.9721535444259644,
+      "learning_rate": 0.0001,
+      "loss": 1.9698,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6402021240150737,
+      "grad_norm": 0.31012436747550964,
+      "learning_rate": 0.0001,
+      "loss": 0.4364,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6402877697841727,
+      "grad_norm": 0.9195960164070129,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6403734155532717,
+      "grad_norm": 0.5350064635276794,
+      "learning_rate": 0.0001,
+      "loss": 0.309,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6404590613223706,
+      "grad_norm": 0.7986957430839539,
+      "learning_rate": 0.0001,
+      "loss": 0.688,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6405447070914697,
+      "grad_norm": 0.7152588963508606,
+      "learning_rate": 0.0001,
+      "loss": 1.1278,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6406303528605687,
+      "grad_norm": 0.5875113606452942,
+      "learning_rate": 0.0001,
+      "loss": 1.287,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6407159986296677,
+      "grad_norm": 1.0576565265655518,
+      "learning_rate": 0.0001,
+      "loss": 0.8335,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6408016443987667,
+      "grad_norm": 0.766553521156311,
+      "learning_rate": 0.0001,
+      "loss": 0.713,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6408872901678657,
+      "grad_norm": 0.9001264572143555,
+      "learning_rate": 0.0001,
+      "loss": 0.8757,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6409729359369647,
+      "grad_norm": 0.956609308719635,
+      "learning_rate": 0.0001,
+      "loss": 1.0192,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6410585817060637,
+      "grad_norm": 0.4969010353088379,
+      "learning_rate": 0.0001,
+      "loss": 0.8688,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6411442274751628,
+      "grad_norm": 0.861411988735199,
+      "learning_rate": 0.0001,
+      "loss": 0.4548,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6412298732442617,
+      "grad_norm": 0.7598739266395569,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6413155190133607,
+      "grad_norm": 0.5293703675270081,
+      "learning_rate": 0.0001,
+      "loss": 1.1012,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6414011647824598,
+      "grad_norm": 0.327858030796051,
+      "learning_rate": 0.0001,
+      "loss": 0.357,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6414868105515588,
+      "grad_norm": 0.7935364842414856,
+      "learning_rate": 0.0001,
+      "loss": 0.7216,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6415724563206577,
+      "grad_norm": 0.5695441961288452,
+      "learning_rate": 0.0001,
+      "loss": 0.8625,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6416581020897568,
+      "grad_norm": 1.058430790901184,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6417437478588558,
+      "grad_norm": 0.5217165350914001,
+      "learning_rate": 0.0001,
+      "loss": 0.7279,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6418293936279548,
+      "grad_norm": 1.150910496711731,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6419150393970537,
+      "grad_norm": 0.6798585653305054,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6420006851661528,
+      "grad_norm": 0.9352955222129822,
+      "learning_rate": 0.0001,
+      "loss": 0.7543,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6420863309352518,
+      "grad_norm": 0.7230253219604492,
+      "learning_rate": 0.0001,
+      "loss": 1.2566,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6421719767043508,
+      "grad_norm": 1.21560800075531,
+      "learning_rate": 0.0001,
+      "loss": 1.1371,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6422576224734499,
+      "grad_norm": 0.8278729915618896,
+      "learning_rate": 0.0001,
+      "loss": 0.9702,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6423432682425488,
+      "grad_norm": 0.9593185186386108,
+      "learning_rate": 0.0001,
+      "loss": 1.092,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6424289140116478,
+      "grad_norm": 0.70489901304245,
+      "learning_rate": 0.0001,
+      "loss": 0.6557,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6425145597807468,
+      "grad_norm": 1.02269446849823,
+      "learning_rate": 0.0001,
+      "loss": 1.3855,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6426002055498459,
+      "grad_norm": 0.6504704356193542,
+      "learning_rate": 0.0001,
+      "loss": 1.0878,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6426858513189448,
+      "grad_norm": 0.9607993960380554,
+      "learning_rate": 0.0001,
+      "loss": 0.612,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6427714970880438,
+      "grad_norm": 0.8225135803222656,
+      "learning_rate": 0.0001,
+      "loss": 0.6899,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.0243226289749146,
+      "learning_rate": 0.0001,
+      "loss": 1.0727,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6429427886262419,
+      "grad_norm": 0.5414093136787415,
+      "learning_rate": 0.0001,
+      "loss": 0.5111,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6430284343953409,
+      "grad_norm": 1.0133734941482544,
+      "learning_rate": 0.0001,
+      "loss": 1.3613,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6431140801644398,
+      "grad_norm": 0.634095311164856,
+      "learning_rate": 0.0001,
+      "loss": 0.362,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6431997259335389,
+      "grad_norm": 0.9219125509262085,
+      "learning_rate": 0.0001,
+      "loss": 0.908,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6432853717026379,
+      "grad_norm": 1.3184831142425537,
+      "learning_rate": 0.0001,
+      "loss": 0.9645,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6433710174717369,
+      "grad_norm": 0.6162232160568237,
+      "learning_rate": 0.0001,
+      "loss": 0.9698,
+      "step": 7512
+    },
+    {
+      "epoch": 0.643456663240836,
+      "grad_norm": 0.6295791268348694,
+      "learning_rate": 0.0001,
+      "loss": 0.4764,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6435423090099349,
+      "grad_norm": 0.8520804643630981,
+      "learning_rate": 0.0001,
+      "loss": 0.6367,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6436279547790339,
+      "grad_norm": 0.5705785155296326,
+      "learning_rate": 0.0001,
+      "loss": 0.7697,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6437136005481329,
+      "grad_norm": 0.9900500178337097,
+      "learning_rate": 0.0001,
+      "loss": 0.8853,
+      "step": 7516
+    },
+    {
+      "epoch": 0.643799246317232,
+      "grad_norm": 0.8652080297470093,
+      "learning_rate": 0.0001,
+      "loss": 0.7859,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6438848920863309,
+      "grad_norm": 1.134945034980774,
+      "learning_rate": 0.0001,
+      "loss": 1.0401,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6439705378554299,
+      "grad_norm": 0.8079750537872314,
+      "learning_rate": 0.0001,
+      "loss": 0.9415,
+      "step": 7519
+    },
+    {
+      "epoch": 0.644056183624529,
+      "grad_norm": 0.8030602931976318,
+      "learning_rate": 0.0001,
+      "loss": 1.0627,
+      "step": 7520
+    },
+    {
+      "epoch": 0.644141829393628,
+      "grad_norm": 0.8531151413917542,
+      "learning_rate": 0.0001,
+      "loss": 0.8771,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6442274751627269,
+      "grad_norm": 0.9853979349136353,
+      "learning_rate": 0.0001,
+      "loss": 0.9287,
+      "step": 7522
+    },
+    {
+      "epoch": 0.644313120931826,
+      "grad_norm": 1.01267409324646,
+      "learning_rate": 0.0001,
+      "loss": 1.3232,
+      "step": 7523
+    },
+    {
+      "epoch": 0.644398766700925,
+      "grad_norm": 0.7332214117050171,
+      "learning_rate": 0.0001,
+      "loss": 1.131,
+      "step": 7524
+    },
+    {
+      "epoch": 0.644484412470024,
+      "grad_norm": 1.1906027793884277,
+      "learning_rate": 0.0001,
+      "loss": 0.7888,
+      "step": 7525
+    },
+    {
+      "epoch": 0.644570058239123,
+      "grad_norm": 0.6338390707969666,
+      "learning_rate": 0.0001,
+      "loss": 0.6422,
+      "step": 7526
+    },
+    {
+      "epoch": 0.644655704008222,
+      "grad_norm": 0.5808262228965759,
+      "learning_rate": 0.0001,
+      "loss": 0.7119,
+      "step": 7527
+    },
+    {
+      "epoch": 0.644741349777321,
+      "grad_norm": 0.6875061988830566,
+      "learning_rate": 0.0001,
+      "loss": 0.4487,
+      "step": 7528
+    },
+    {
+      "epoch": 0.64482699554642,
+      "grad_norm": 0.7466690540313721,
+      "learning_rate": 0.0001,
+      "loss": 0.8172,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6449126413155191,
+      "grad_norm": 0.639363706111908,
+      "learning_rate": 0.0001,
+      "loss": 0.8625,
+      "step": 7530
+    },
+    {
+      "epoch": 0.644998287084618,
+      "grad_norm": 0.9869226217269897,
+      "learning_rate": 0.0001,
+      "loss": 0.8839,
+      "step": 7531
+    },
+    {
+      "epoch": 0.645083932853717,
+      "grad_norm": 0.6521191000938416,
+      "learning_rate": 0.0001,
+      "loss": 0.2769,
+      "step": 7532
+    },
+    {
+      "epoch": 0.645169578622816,
+      "grad_norm": 0.6516343355178833,
+      "learning_rate": 0.0001,
+      "loss": 1.0329,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6452552243919151,
+      "grad_norm": 0.8494449257850647,
+      "learning_rate": 0.0001,
+      "loss": 1.2314,
+      "step": 7534
+    },
+    {
+      "epoch": 0.645340870161014,
+      "grad_norm": 0.7462759613990784,
+      "learning_rate": 0.0001,
+      "loss": 1.1224,
+      "step": 7535
+    },
+    {
+      "epoch": 0.645426515930113,
+      "grad_norm": 0.9669510126113892,
+      "learning_rate": 0.0001,
+      "loss": 1.0411,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6455121616992121,
+      "grad_norm": 0.6502089500427246,
+      "learning_rate": 0.0001,
+      "loss": 0.8165,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6455978074683111,
+      "grad_norm": 1.1386070251464844,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 7538
+    },
+    {
+      "epoch": 0.64568345323741,
+      "grad_norm": 0.4936411380767822,
+      "learning_rate": 0.0001,
+      "loss": 0.8188,
+      "step": 7539
+    },
+    {
+      "epoch": 0.645769099006509,
+      "grad_norm": 0.9725869297981262,
+      "learning_rate": 0.0001,
+      "loss": 1.3114,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6458547447756081,
+      "grad_norm": 0.8701868653297424,
+      "learning_rate": 0.0001,
+      "loss": 1.1737,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6459403905447071,
+      "grad_norm": 1.1311283111572266,
+      "learning_rate": 0.0001,
+      "loss": 1.29,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6460260363138061,
+      "grad_norm": 1.1118136644363403,
+      "learning_rate": 0.0001,
+      "loss": 0.7456,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6461116820829051,
+      "grad_norm": 0.593003511428833,
+      "learning_rate": 0.0001,
+      "loss": 0.8338,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6461973278520041,
+      "grad_norm": 0.541266679763794,
+      "learning_rate": 0.0001,
+      "loss": 0.1736,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6462829736211031,
+      "grad_norm": 0.8497434854507446,
+      "learning_rate": 0.0001,
+      "loss": 0.736,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6463686193902021,
+      "grad_norm": 0.6529626250267029,
+      "learning_rate": 0.0001,
+      "loss": 0.9372,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6464542651593012,
+      "grad_norm": 0.9244458079338074,
+      "learning_rate": 0.0001,
+      "loss": 0.7654,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6465399109284001,
+      "grad_norm": 1.088392734527588,
+      "learning_rate": 0.0001,
+      "loss": 1.3249,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6466255566974991,
+      "grad_norm": 1.0195978879928589,
+      "learning_rate": 0.0001,
+      "loss": 1.0032,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6467112024665982,
+      "grad_norm": 0.5449867844581604,
+      "learning_rate": 0.0001,
+      "loss": 0.9608,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6467968482356972,
+      "grad_norm": 0.7150182723999023,
+      "learning_rate": 0.0001,
+      "loss": 0.5749,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6468824940047961,
+      "grad_norm": 0.7590324878692627,
+      "learning_rate": 0.0001,
+      "loss": 0.6637,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6469681397738952,
+      "grad_norm": 0.6762273907661438,
+      "learning_rate": 0.0001,
+      "loss": 1.2756,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6470537855429942,
+      "grad_norm": 1.0352487564086914,
+      "learning_rate": 0.0001,
+      "loss": 0.9572,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6471394313120932,
+      "grad_norm": 0.9681846499443054,
+      "learning_rate": 0.0001,
+      "loss": 1.1466,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6472250770811921,
+      "grad_norm": 0.6370871663093567,
+      "learning_rate": 0.0001,
+      "loss": 0.728,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6473107228502912,
+      "grad_norm": 0.5325378179550171,
+      "learning_rate": 0.0001,
+      "loss": 0.7773,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6473963686193902,
+      "grad_norm": 0.5018674731254578,
+      "learning_rate": 0.0001,
+      "loss": 0.6126,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6474820143884892,
+      "grad_norm": 0.8144243359565735,
+      "learning_rate": 0.0001,
+      "loss": 0.9628,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6475676601575883,
+      "grad_norm": 0.8589040637016296,
+      "learning_rate": 0.0001,
+      "loss": 0.7103,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6476533059266872,
+      "grad_norm": 1.2450083494186401,
+      "learning_rate": 0.0001,
+      "loss": 1.0612,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6477389516957862,
+      "grad_norm": 1.085924506187439,
+      "learning_rate": 0.0001,
+      "loss": 1.0233,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6478245974648852,
+      "grad_norm": 0.7461416721343994,
+      "learning_rate": 0.0001,
+      "loss": 0.9502,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6479102432339843,
+      "grad_norm": 0.8869432210922241,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6479958890030832,
+      "grad_norm": 1.2936046123504639,
+      "learning_rate": 0.0001,
+      "loss": 1.0066,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6480815347721822,
+      "grad_norm": 1.1546531915664673,
+      "learning_rate": 0.0001,
+      "loss": 1.1898,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6481671805412813,
+      "grad_norm": 1.098912000656128,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6482528263103803,
+      "grad_norm": 0.9997996687889099,
+      "learning_rate": 0.0001,
+      "loss": 1.0933,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6483384720794793,
+      "grad_norm": 1.23857843875885,
+      "learning_rate": 0.0001,
+      "loss": 1.108,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6484241178485782,
+      "grad_norm": 1.390872836112976,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6485097636176773,
+      "grad_norm": 1.2171235084533691,
+      "learning_rate": 0.0001,
+      "loss": 1.0574,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6485954093867763,
+      "grad_norm": 1.2960644960403442,
+      "learning_rate": 0.0001,
+      "loss": 1.0528,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6486810551558753,
+      "grad_norm": 0.9944027066230774,
+      "learning_rate": 0.0001,
+      "loss": 0.9012,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6487667009249743,
+      "grad_norm": 0.9691668748855591,
+      "learning_rate": 0.0001,
+      "loss": 0.9667,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6488523466940733,
+      "grad_norm": 0.9199744462966919,
+      "learning_rate": 0.0001,
+      "loss": 1.0263,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6489379924631723,
+      "grad_norm": 1.1250938177108765,
+      "learning_rate": 0.0001,
+      "loss": 0.7873,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6490236382322713,
+      "grad_norm": 0.8017643094062805,
+      "learning_rate": 0.0001,
+      "loss": 0.2755,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6491092840013704,
+      "grad_norm": 0.663375735282898,
+      "learning_rate": 0.0001,
+      "loss": 1.2821,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6491949297704693,
+      "grad_norm": 0.5567197799682617,
+      "learning_rate": 0.0001,
+      "loss": 0.4739,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6492805755395683,
+      "grad_norm": 0.7662072777748108,
+      "learning_rate": 0.0001,
+      "loss": 0.9131,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6493662213086674,
+      "grad_norm": 0.8155107498168945,
+      "learning_rate": 0.0001,
+      "loss": 0.877,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6494518670777664,
+      "grad_norm": 0.8748714923858643,
+      "learning_rate": 0.0001,
+      "loss": 0.8686,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6495375128468653,
+      "grad_norm": 0.8866846561431885,
+      "learning_rate": 0.0001,
+      "loss": 1.2176,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6496231586159644,
+      "grad_norm": 1.3232405185699463,
+      "learning_rate": 0.0001,
+      "loss": 1.0527,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6497088043850634,
+      "grad_norm": 1.317531704902649,
+      "learning_rate": 0.0001,
+      "loss": 0.6555,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6497944501541624,
+      "grad_norm": 0.700070858001709,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6498800959232613,
+      "grad_norm": 0.7687492966651917,
+      "learning_rate": 0.0001,
+      "loss": 0.5871,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6499657416923604,
+      "grad_norm": 1.0735210180282593,
+      "learning_rate": 0.0001,
+      "loss": 1.0469,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6500513874614594,
+      "grad_norm": 1.1214724779129028,
+      "learning_rate": 0.0001,
+      "loss": 0.8529,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6501370332305584,
+      "grad_norm": 0.8446028232574463,
+      "learning_rate": 0.0001,
+      "loss": 0.9915,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6502226789996575,
+      "grad_norm": 0.7938650250434875,
+      "learning_rate": 0.0001,
+      "loss": 1.089,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6503083247687564,
+      "grad_norm": 0.27957871556282043,
+      "learning_rate": 0.0001,
+      "loss": 0.5264,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6503939705378554,
+      "grad_norm": 0.7463667392730713,
+      "learning_rate": 0.0001,
+      "loss": 1.1461,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6504796163069544,
+      "grad_norm": 1.3047034740447998,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6505652620760535,
+      "grad_norm": 0.6931648254394531,
+      "learning_rate": 0.0001,
+      "loss": 1.0127,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6506509078451524,
+      "grad_norm": 0.5534169673919678,
+      "learning_rate": 0.0001,
+      "loss": 0.1955,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6507365536142514,
+      "grad_norm": 0.2951107919216156,
+      "learning_rate": 0.0001,
+      "loss": 0.0927,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6508221993833505,
+      "grad_norm": 0.4326786398887634,
+      "learning_rate": 0.0001,
+      "loss": 0.4792,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6509078451524495,
+      "grad_norm": 1.1611665487289429,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6509934909215485,
+      "grad_norm": 0.7243531346321106,
+      "learning_rate": 0.0001,
+      "loss": 0.5809,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6510791366906474,
+      "grad_norm": 0.957494854927063,
+      "learning_rate": 0.0001,
+      "loss": 0.8878,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6511647824597465,
+      "grad_norm": 1.134367823600769,
+      "learning_rate": 0.0001,
+      "loss": 0.9995,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6512504282288455,
+      "grad_norm": 1.2383956909179688,
+      "learning_rate": 0.0001,
+      "loss": 2.2526,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6513360739979445,
+      "grad_norm": 1.0699437856674194,
+      "learning_rate": 0.0001,
+      "loss": 0.502,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6514217197670435,
+      "grad_norm": 1.152877926826477,
+      "learning_rate": 0.0001,
+      "loss": 1.1875,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6515073655361425,
+      "grad_norm": 0.8147044777870178,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6515930113052415,
+      "grad_norm": 0.5655791163444519,
+      "learning_rate": 0.0001,
+      "loss": 0.9952,
+      "step": 7608
+    },
+    {
+      "epoch": 0.6516786570743405,
+      "grad_norm": 0.5897744297981262,
+      "learning_rate": 0.0001,
+      "loss": 1.0177,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6517643028434396,
+      "grad_norm": 1.3615835905075073,
+      "learning_rate": 0.0001,
+      "loss": 1.2787,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6518499486125385,
+      "grad_norm": 1.0843838453292847,
+      "learning_rate": 0.0001,
+      "loss": 0.9169,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6519355943816375,
+      "grad_norm": 1.4068183898925781,
+      "learning_rate": 0.0001,
+      "loss": 0.9028,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6520212401507366,
+      "grad_norm": 0.6971965432167053,
+      "learning_rate": 0.0001,
+      "loss": 1.1279,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6521068859198356,
+      "grad_norm": 0.9237734079360962,
+      "learning_rate": 0.0001,
+      "loss": 1.1575,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6521925316889345,
+      "grad_norm": 0.8477229475975037,
+      "learning_rate": 0.0001,
+      "loss": 0.8296,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6522781774580336,
+      "grad_norm": 0.6019101142883301,
+      "learning_rate": 0.0001,
+      "loss": 0.4365,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6523638232271326,
+      "grad_norm": 0.5217714905738831,
+      "learning_rate": 0.0001,
+      "loss": 1.0201,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6524494689962316,
+      "grad_norm": 0.6574777960777283,
+      "learning_rate": 0.0001,
+      "loss": 1.1205,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6525351147653305,
+      "grad_norm": 0.6253805160522461,
+      "learning_rate": 0.0001,
+      "loss": 1.1898,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6526207605344296,
+      "grad_norm": 0.9481054544448853,
+      "learning_rate": 0.0001,
+      "loss": 1.252,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6527064063035286,
+      "grad_norm": 0.47903066873550415,
+      "learning_rate": 0.0001,
+      "loss": 0.8224,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6527920520726276,
+      "grad_norm": 0.984918475151062,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6528776978417267,
+      "grad_norm": 0.9245924353599548,
+      "learning_rate": 0.0001,
+      "loss": 1.0853,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6529633436108256,
+      "grad_norm": 0.6790052652359009,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6530489893799246,
+      "grad_norm": 0.7182889580726624,
+      "learning_rate": 0.0001,
+      "loss": 0.9817,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6531346351490236,
+      "grad_norm": 0.6240759491920471,
+      "learning_rate": 0.0001,
+      "loss": 0.5209,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6532202809181227,
+      "grad_norm": 0.5919815301895142,
+      "learning_rate": 0.0001,
+      "loss": 0.8401,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6533059266872216,
+      "grad_norm": 1.0559461116790771,
+      "learning_rate": 0.0001,
+      "loss": 1.3713,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6533915724563206,
+      "grad_norm": 0.7846957445144653,
+      "learning_rate": 0.0001,
+      "loss": 0.9195,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6534772182254197,
+      "grad_norm": 0.5740596055984497,
+      "learning_rate": 0.0001,
+      "loss": 1.2024,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6535628639945187,
+      "grad_norm": 0.3405897915363312,
+      "learning_rate": 0.0001,
+      "loss": 0.932,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6536485097636177,
+      "grad_norm": 0.7021228671073914,
+      "learning_rate": 0.0001,
+      "loss": 0.5245,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6537341555327166,
+      "grad_norm": 0.46083155274391174,
+      "learning_rate": 0.0001,
+      "loss": 0.6691,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6538198013018157,
+      "grad_norm": 0.7924981117248535,
+      "learning_rate": 0.0001,
+      "loss": 1.1493,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6539054470709147,
+      "grad_norm": 0.2967544496059418,
+      "learning_rate": 0.0001,
+      "loss": 0.3364,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6539910928400137,
+      "grad_norm": 0.7324005365371704,
+      "learning_rate": 0.0001,
+      "loss": 0.7836,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6540767386091128,
+      "grad_norm": 1.083558201789856,
+      "learning_rate": 0.0001,
+      "loss": 1.1567,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6541623843782117,
+      "grad_norm": 0.8451089859008789,
+      "learning_rate": 0.0001,
+      "loss": 0.8984,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6542480301473107,
+      "grad_norm": 1.2655646800994873,
+      "learning_rate": 0.0001,
+      "loss": 1.0615,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6543336759164097,
+      "grad_norm": 0.9758747220039368,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6544193216855088,
+      "grad_norm": 0.8275357484817505,
+      "learning_rate": 0.0001,
+      "loss": 0.8159,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6545049674546077,
+      "grad_norm": 1.0765528678894043,
+      "learning_rate": 0.0001,
+      "loss": 1.1241,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6545906132237067,
+      "grad_norm": 0.6830666661262512,
+      "learning_rate": 0.0001,
+      "loss": 1.0875,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6546762589928058,
+      "grad_norm": 0.9611583948135376,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6547619047619048,
+      "grad_norm": 0.6304017901420593,
+      "learning_rate": 0.0001,
+      "loss": 0.8793,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6548475505310037,
+      "grad_norm": 0.7917098999023438,
+      "learning_rate": 0.0001,
+      "loss": 0.9015,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6549331963001028,
+      "grad_norm": 0.9093272686004639,
+      "learning_rate": 0.0001,
+      "loss": 0.7906,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6550188420692018,
+      "grad_norm": 0.8505716919898987,
+      "learning_rate": 0.0001,
+      "loss": 1.002,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6551044878383008,
+      "grad_norm": 1.2098933458328247,
+      "learning_rate": 0.0001,
+      "loss": 1.1077,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6551901336073997,
+      "grad_norm": 1.1004496812820435,
+      "learning_rate": 0.0001,
+      "loss": 1.1926,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6552757793764988,
+      "grad_norm": 1.1970980167388916,
+      "learning_rate": 0.0001,
+      "loss": 0.8993,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6553614251455978,
+      "grad_norm": 0.776637613773346,
+      "learning_rate": 0.0001,
+      "loss": 0.8511,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6554470709146968,
+      "grad_norm": 0.7690945863723755,
+      "learning_rate": 0.0001,
+      "loss": 1.2271,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6555327166837959,
+      "grad_norm": 0.6606077551841736,
+      "learning_rate": 0.0001,
+      "loss": 1.0273,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6556183624528948,
+      "grad_norm": 0.8473818302154541,
+      "learning_rate": 0.0001,
+      "loss": 0.9338,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6557040082219938,
+      "grad_norm": 0.848793625831604,
+      "learning_rate": 0.0001,
+      "loss": 0.8737,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6557896539910928,
+      "grad_norm": 0.5399657487869263,
+      "learning_rate": 0.0001,
+      "loss": 0.3818,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6558752997601919,
+      "grad_norm": 1.1779463291168213,
+      "learning_rate": 0.0001,
+      "loss": 1.9762,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6559609455292909,
+      "grad_norm": 0.9821550846099854,
+      "learning_rate": 0.0001,
+      "loss": 0.8556,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6560465912983898,
+      "grad_norm": 0.8414906859397888,
+      "learning_rate": 0.0001,
+      "loss": 0.9583,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6561322370674889,
+      "grad_norm": 0.869047999382019,
+      "learning_rate": 0.0001,
+      "loss": 0.8666,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6562178828365879,
+      "grad_norm": 0.5675357580184937,
+      "learning_rate": 0.0001,
+      "loss": 0.5834,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6563035286056869,
+      "grad_norm": 0.6728357672691345,
+      "learning_rate": 0.0001,
+      "loss": 0.862,
+      "step": 7663
+    },
+    {
+      "epoch": 0.6563891743747858,
+      "grad_norm": 0.8977221250534058,
+      "learning_rate": 0.0001,
+      "loss": 0.8613,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6564748201438849,
+      "grad_norm": 1.2974798679351807,
+      "learning_rate": 0.0001,
+      "loss": 2.3464,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6565604659129839,
+      "grad_norm": 0.9136859178543091,
+      "learning_rate": 0.0001,
+      "loss": 0.8447,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6566461116820829,
+      "grad_norm": 0.6282453536987305,
+      "learning_rate": 0.0001,
+      "loss": 0.5524,
+      "step": 7667
+    },
+    {
+      "epoch": 0.656731757451182,
+      "grad_norm": 0.621978223323822,
+      "learning_rate": 0.0001,
+      "loss": 1.3635,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6568174032202809,
+      "grad_norm": 0.7287351489067078,
+      "learning_rate": 0.0001,
+      "loss": 1.1618,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6569030489893799,
+      "grad_norm": 0.8739994764328003,
+      "learning_rate": 0.0001,
+      "loss": 0.9148,
+      "step": 7670
+    },
+    {
+      "epoch": 0.656988694758479,
+      "grad_norm": 0.6221050024032593,
+      "learning_rate": 0.0001,
+      "loss": 1.1407,
+      "step": 7671
+    },
+    {
+      "epoch": 0.657074340527578,
+      "grad_norm": 1.1568074226379395,
+      "learning_rate": 0.0001,
+      "loss": 1.2657,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6571599862966769,
+      "grad_norm": 0.7939939498901367,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6572456320657759,
+      "grad_norm": 0.5565939545631409,
+      "learning_rate": 0.0001,
+      "loss": 0.7771,
+      "step": 7674
+    },
+    {
+      "epoch": 0.657331277834875,
+      "grad_norm": 0.4786389172077179,
+      "learning_rate": 0.0001,
+      "loss": 0.3207,
+      "step": 7675
+    },
+    {
+      "epoch": 0.657416923603974,
+      "grad_norm": 1.0794832706451416,
+      "learning_rate": 0.0001,
+      "loss": 0.9797,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6575025693730729,
+      "grad_norm": 0.9811509251594543,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 7677
+    },
+    {
+      "epoch": 0.657588215142172,
+      "grad_norm": 0.6299436688423157,
+      "learning_rate": 0.0001,
+      "loss": 0.2892,
+      "step": 7678
+    },
+    {
+      "epoch": 0.657673860911271,
+      "grad_norm": 1.005570411682129,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 7679
+    },
+    {
+      "epoch": 0.65775950668037,
+      "grad_norm": 0.6464515924453735,
+      "learning_rate": 0.0001,
+      "loss": 0.4799,
+      "step": 7680
+    },
+    {
+      "epoch": 0.657845152449469,
+      "grad_norm": 0.5958821177482605,
+      "learning_rate": 0.0001,
+      "loss": 0.8949,
+      "step": 7681
+    },
+    {
+      "epoch": 0.657930798218568,
+      "grad_norm": 0.8325895667076111,
+      "learning_rate": 0.0001,
+      "loss": 0.7548,
+      "step": 7682
+    },
+    {
+      "epoch": 0.658016443987667,
+      "grad_norm": 0.5957608819007874,
+      "learning_rate": 0.0001,
+      "loss": 0.9347,
+      "step": 7683
+    },
+    {
+      "epoch": 0.658102089756766,
+      "grad_norm": 1.4761167764663696,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6581877355258651,
+      "grad_norm": 1.7095550298690796,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 7685
+    },
+    {
+      "epoch": 0.658273381294964,
+      "grad_norm": 0.9418708682060242,
+      "learning_rate": 0.0001,
+      "loss": 0.8338,
+      "step": 7686
+    },
+    {
+      "epoch": 0.658359027064063,
+      "grad_norm": 0.9782050848007202,
+      "learning_rate": 0.0001,
+      "loss": 1.1192,
+      "step": 7687
+    },
+    {
+      "epoch": 0.658444672833162,
+      "grad_norm": 0.9163830280303955,
+      "learning_rate": 0.0001,
+      "loss": 1.0863,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6585303186022611,
+      "grad_norm": 1.1561591625213623,
+      "learning_rate": 0.0001,
+      "loss": 1.2611,
+      "step": 7689
+    },
+    {
+      "epoch": 0.65861596437136,
+      "grad_norm": 1.3425654172897339,
+      "learning_rate": 0.0001,
+      "loss": 0.5227,
+      "step": 7690
+    },
+    {
+      "epoch": 0.658701610140459,
+      "grad_norm": 1.1221041679382324,
+      "learning_rate": 0.0001,
+      "loss": 0.8494,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6587872559095581,
+      "grad_norm": 1.2281582355499268,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6588729016786571,
+      "grad_norm": 0.7982869744300842,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6589585474477561,
+      "grad_norm": 0.8972541093826294,
+      "learning_rate": 0.0001,
+      "loss": 0.8257,
+      "step": 7694
+    },
+    {
+      "epoch": 0.659044193216855,
+      "grad_norm": 0.9425706267356873,
+      "learning_rate": 0.0001,
+      "loss": 0.8995,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6591298389859541,
+      "grad_norm": 0.8433093428611755,
+      "learning_rate": 0.0001,
+      "loss": 1.0018,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6592154847550531,
+      "grad_norm": 1.3085788488388062,
+      "learning_rate": 0.0001,
+      "loss": 0.9374,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6593011305241521,
+      "grad_norm": 0.43470123410224915,
+      "learning_rate": 0.0001,
+      "loss": 0.9206,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6593867762932512,
+      "grad_norm": 1.1719597578048706,
+      "learning_rate": 0.0001,
+      "loss": 1.023,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6594724220623501,
+      "grad_norm": 0.6586341857910156,
+      "learning_rate": 0.0001,
+      "loss": 0.4491,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6595580678314491,
+      "grad_norm": 1.2072604894638062,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6596437136005482,
+      "grad_norm": 1.2066916227340698,
+      "learning_rate": 0.0001,
+      "loss": 2.8501,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6597293593696472,
+      "grad_norm": 0.7572288513183594,
+      "learning_rate": 0.0001,
+      "loss": 0.8516,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6598150051387461,
+      "grad_norm": 1.5517146587371826,
+      "learning_rate": 0.0001,
+      "loss": 1.1438,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6599006509078451,
+      "grad_norm": 0.9407103061676025,
+      "learning_rate": 0.0001,
+      "loss": 0.8164,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6599862966769442,
+      "grad_norm": 0.8316556215286255,
+      "learning_rate": 0.0001,
+      "loss": 0.9332,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6600719424460432,
+      "grad_norm": 0.6559456586837769,
+      "learning_rate": 0.0001,
+      "loss": 0.9559,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6601575882151421,
+      "grad_norm": 0.388164222240448,
+      "learning_rate": 0.0001,
+      "loss": 0.4281,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6602432339842412,
+      "grad_norm": 0.38388094305992126,
+      "learning_rate": 0.0001,
+      "loss": 0.3512,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6603288797533402,
+      "grad_norm": 0.5776779055595398,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6604145255224392,
+      "grad_norm": 0.7446640729904175,
+      "learning_rate": 0.0001,
+      "loss": 0.513,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6605001712915382,
+      "grad_norm": 0.9330748319625854,
+      "learning_rate": 0.0001,
+      "loss": 0.7253,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6605858170606372,
+      "grad_norm": 1.0988819599151611,
+      "learning_rate": 0.0001,
+      "loss": 1.0417,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6606714628297362,
+      "grad_norm": 0.5922281742095947,
+      "learning_rate": 0.0001,
+      "loss": 1.1119,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6607571085988352,
+      "grad_norm": 0.9469457268714905,
+      "learning_rate": 0.0001,
+      "loss": 1.0164,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6608427543679343,
+      "grad_norm": 0.7138570547103882,
+      "learning_rate": 0.0001,
+      "loss": 0.8969,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6609284001370332,
+      "grad_norm": 0.615116536617279,
+      "learning_rate": 0.0001,
+      "loss": 0.951,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6610140459061322,
+      "grad_norm": 0.4704378843307495,
+      "learning_rate": 0.0001,
+      "loss": 0.9852,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6610996916752312,
+      "grad_norm": 0.6337037086486816,
+      "learning_rate": 0.0001,
+      "loss": 1.021,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6611853374443303,
+      "grad_norm": 1.0805782079696655,
+      "learning_rate": 0.0001,
+      "loss": 1.2738,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6612709832134293,
+      "grad_norm": 0.6045351028442383,
+      "learning_rate": 0.0001,
+      "loss": 0.8088,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6613566289825282,
+      "grad_norm": 0.9829623103141785,
+      "learning_rate": 0.0001,
+      "loss": 0.8747,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6614422747516273,
+      "grad_norm": 1.0078191757202148,
+      "learning_rate": 0.0001,
+      "loss": 1.1517,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6615279205207263,
+      "grad_norm": 0.6809642314910889,
+      "learning_rate": 0.0001,
+      "loss": 0.8148,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6616135662898253,
+      "grad_norm": 0.6549866199493408,
+      "learning_rate": 0.0001,
+      "loss": 0.989,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6616992120589242,
+      "grad_norm": 0.6810969114303589,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6617848578280233,
+      "grad_norm": 0.8142023682594299,
+      "learning_rate": 0.0001,
+      "loss": 1.0635,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6618705035971223,
+      "grad_norm": 0.7406982779502869,
+      "learning_rate": 0.0001,
+      "loss": 1.3397,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6619561493662213,
+      "grad_norm": 1.0061653852462769,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6620417951353204,
+      "grad_norm": 1.2801361083984375,
+      "learning_rate": 0.0001,
+      "loss": 1.2415,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6621274409044193,
+      "grad_norm": 0.6659467816352844,
+      "learning_rate": 0.0001,
+      "loss": 0.777,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6622130866735183,
+      "grad_norm": 0.8075217604637146,
+      "learning_rate": 0.0001,
+      "loss": 1.2059,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6622987324426174,
+      "grad_norm": 1.009833574295044,
+      "learning_rate": 0.0001,
+      "loss": 1.3725,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6623843782117164,
+      "grad_norm": 0.95920729637146,
+      "learning_rate": 0.0001,
+      "loss": 1.0348,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6624700239808153,
+      "grad_norm": 0.6936363577842712,
+      "learning_rate": 0.0001,
+      "loss": 1.1536,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6625556697499143,
+      "grad_norm": 0.7120153307914734,
+      "learning_rate": 0.0001,
+      "loss": 1.3044,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6626413155190134,
+      "grad_norm": 0.8489973545074463,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6627269612881124,
+      "grad_norm": 0.8242935538291931,
+      "learning_rate": 0.0001,
+      "loss": 1.1108,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6628126070572113,
+      "grad_norm": 0.6684483289718628,
+      "learning_rate": 0.0001,
+      "loss": 1.2364,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6628982528263104,
+      "grad_norm": 0.8998829126358032,
+      "learning_rate": 0.0001,
+      "loss": 1.0922,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6629838985954094,
+      "grad_norm": 1.2911702394485474,
+      "learning_rate": 0.0001,
+      "loss": 1.251,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6630695443645084,
+      "grad_norm": 1.275686264038086,
+      "learning_rate": 0.0001,
+      "loss": 1.0291,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6631551901336074,
+      "grad_norm": 0.5748265385627747,
+      "learning_rate": 0.0001,
+      "loss": 1.2254,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6632408359027064,
+      "grad_norm": 0.574734091758728,
+      "learning_rate": 0.0001,
+      "loss": 0.7558,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6633264816718054,
+      "grad_norm": 0.6906788945198059,
+      "learning_rate": 0.0001,
+      "loss": 0.8316,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6634121274409044,
+      "grad_norm": 0.8428770899772644,
+      "learning_rate": 0.0001,
+      "loss": 1.029,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6634977732100035,
+      "grad_norm": 0.553987979888916,
+      "learning_rate": 0.0001,
+      "loss": 0.7218,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6635834189791024,
+      "grad_norm": 1.0273921489715576,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6636690647482014,
+      "grad_norm": 0.9431573152542114,
+      "learning_rate": 0.0001,
+      "loss": 0.8095,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6637547105173004,
+      "grad_norm": 0.9321008920669556,
+      "learning_rate": 0.0001,
+      "loss": 1.0687,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6638403562863995,
+      "grad_norm": 0.6014727354049683,
+      "learning_rate": 0.0001,
+      "loss": 0.8107,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6639260020554985,
+      "grad_norm": 0.40086865425109863,
+      "learning_rate": 0.0001,
+      "loss": 0.507,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6640116478245974,
+      "grad_norm": 0.46368515491485596,
+      "learning_rate": 0.0001,
+      "loss": 0.5681,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6640972935936965,
+      "grad_norm": 0.7705414891242981,
+      "learning_rate": 0.0001,
+      "loss": 1.1255,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6641829393627955,
+      "grad_norm": 0.6915990710258484,
+      "learning_rate": 0.0001,
+      "loss": 0.8953,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6642685851318945,
+      "grad_norm": 0.5828676223754883,
+      "learning_rate": 0.0001,
+      "loss": 1.1748,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6643542309009934,
+      "grad_norm": 0.9710167050361633,
+      "learning_rate": 0.0001,
+      "loss": 1.3414,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6644398766700925,
+      "grad_norm": 0.7303417921066284,
+      "learning_rate": 0.0001,
+      "loss": 0.7209,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6645255224391915,
+      "grad_norm": 1.1301360130310059,
+      "learning_rate": 0.0001,
+      "loss": 1.0525,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6646111682082905,
+      "grad_norm": 0.7901946306228638,
+      "learning_rate": 0.0001,
+      "loss": 0.5199,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6646968139773896,
+      "grad_norm": 0.9083805084228516,
+      "learning_rate": 0.0001,
+      "loss": 1.0866,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6647824597464885,
+      "grad_norm": 0.3904183506965637,
+      "learning_rate": 0.0001,
+      "loss": 0.5394,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6648681055155875,
+      "grad_norm": 0.6157308220863342,
+      "learning_rate": 0.0001,
+      "loss": 1.0918,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6649537512846866,
+      "grad_norm": 0.5219336152076721,
+      "learning_rate": 0.0001,
+      "loss": 0.7444,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6650393970537856,
+      "grad_norm": 0.6601976752281189,
+      "learning_rate": 0.0001,
+      "loss": 1.0058,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6651250428228845,
+      "grad_norm": 0.5522967576980591,
+      "learning_rate": 0.0001,
+      "loss": 0.7804,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6652106885919835,
+      "grad_norm": 0.8783413171768188,
+      "learning_rate": 0.0001,
+      "loss": 0.9846,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6652963343610826,
+      "grad_norm": 0.8403188586235046,
+      "learning_rate": 0.0001,
+      "loss": 0.8526,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6653819801301816,
+      "grad_norm": 0.6350783109664917,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6654676258992805,
+      "grad_norm": 0.768433153629303,
+      "learning_rate": 0.0001,
+      "loss": 0.9052,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6655532716683796,
+      "grad_norm": 0.6762741804122925,
+      "learning_rate": 0.0001,
+      "loss": 1.0146,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6656389174374786,
+      "grad_norm": 0.9722688794136047,
+      "learning_rate": 0.0001,
+      "loss": 1.2383,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6657245632065776,
+      "grad_norm": 1.0144453048706055,
+      "learning_rate": 0.0001,
+      "loss": 1.0272,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6658102089756766,
+      "grad_norm": 0.7095398902893066,
+      "learning_rate": 0.0001,
+      "loss": 1.1321,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6658958547447756,
+      "grad_norm": 0.8829271197319031,
+      "learning_rate": 0.0001,
+      "loss": 1.1486,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6659815005138746,
+      "grad_norm": 0.6277135014533997,
+      "learning_rate": 0.0001,
+      "loss": 0.8624,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6660671462829736,
+      "grad_norm": 0.9055166244506836,
+      "learning_rate": 0.0001,
+      "loss": 1.0833,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6661527920520727,
+      "grad_norm": 0.6958747506141663,
+      "learning_rate": 0.0001,
+      "loss": 0.4944,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6662384378211716,
+      "grad_norm": 0.7327143549919128,
+      "learning_rate": 0.0001,
+      "loss": 0.8566,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6663240835902706,
+      "grad_norm": 0.3701886832714081,
+      "learning_rate": 0.0001,
+      "loss": 0.4981,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6664097293593696,
+      "grad_norm": 0.5814071297645569,
+      "learning_rate": 0.0001,
+      "loss": 0.5394,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6664953751284687,
+      "grad_norm": 0.9065936803817749,
+      "learning_rate": 0.0001,
+      "loss": 0.9326,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6665810208975677,
+      "grad_norm": 1.6424376964569092,
+      "learning_rate": 0.0001,
+      "loss": 1.2021,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.4517441987991333,
+      "learning_rate": 0.0001,
+      "loss": 0.4469,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6667523124357657,
+      "grad_norm": 0.7980627417564392,
+      "learning_rate": 0.0001,
+      "loss": 0.8051,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6668379582048647,
+      "grad_norm": 0.820490837097168,
+      "learning_rate": 0.0001,
+      "loss": 1.2937,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6669236039739637,
+      "grad_norm": 0.8198030591011047,
+      "learning_rate": 0.0001,
+      "loss": 0.9774,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6670092497430626,
+      "grad_norm": 0.7518323659896851,
+      "learning_rate": 0.0001,
+      "loss": 1.2207,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6670948955121617,
+      "grad_norm": 0.8174818158149719,
+      "learning_rate": 0.0001,
+      "loss": 0.9096,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6671805412812607,
+      "grad_norm": 0.9354525804519653,
+      "learning_rate": 0.0001,
+      "loss": 0.5826,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6672661870503597,
+      "grad_norm": 1.69789719581604,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6673518328194588,
+      "grad_norm": 0.8482400178909302,
+      "learning_rate": 0.0001,
+      "loss": 0.822,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6674374785885577,
+      "grad_norm": 1.050839900970459,
+      "learning_rate": 0.0001,
+      "loss": 0.8967,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6675231243576567,
+      "grad_norm": 0.6774271130561829,
+      "learning_rate": 0.0001,
+      "loss": 1.2164,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6676087701267558,
+      "grad_norm": 0.7476747632026672,
+      "learning_rate": 0.0001,
+      "loss": 1.1629,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6676944158958548,
+      "grad_norm": 0.8024318814277649,
+      "learning_rate": 0.0001,
+      "loss": 0.7892,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6677800616649537,
+      "grad_norm": 0.9427232146263123,
+      "learning_rate": 0.0001,
+      "loss": 1.1585,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6678657074340527,
+      "grad_norm": 1.4155277013778687,
+      "learning_rate": 0.0001,
+      "loss": 1.0344,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6679513532031518,
+      "grad_norm": 0.7799866795539856,
+      "learning_rate": 0.0001,
+      "loss": 1.0549,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6680369989722508,
+      "grad_norm": 0.6835325360298157,
+      "learning_rate": 0.0001,
+      "loss": 1.0601,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6681226447413497,
+      "grad_norm": 0.8268097043037415,
+      "learning_rate": 0.0001,
+      "loss": 0.9736,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6682082905104488,
+      "grad_norm": 0.7290294170379639,
+      "learning_rate": 0.0001,
+      "loss": 1.0099,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6682939362795478,
+      "grad_norm": 0.7898291945457458,
+      "learning_rate": 0.0001,
+      "loss": 1.0053,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6683795820486468,
+      "grad_norm": 0.9823508858680725,
+      "learning_rate": 0.0001,
+      "loss": 0.9126,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6684652278177458,
+      "grad_norm": 1.1301010847091675,
+      "learning_rate": 0.0001,
+      "loss": 1.7715,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6685508735868448,
+      "grad_norm": 1.7915030717849731,
+      "learning_rate": 0.0001,
+      "loss": 0.6142,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6686365193559438,
+      "grad_norm": 0.9104704856872559,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6687221651250428,
+      "grad_norm": 0.6908785700798035,
+      "learning_rate": 0.0001,
+      "loss": 0.757,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6688078108941419,
+      "grad_norm": 0.5495688915252686,
+      "learning_rate": 0.0001,
+      "loss": 0.6185,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6688934566632408,
+      "grad_norm": 1.0990334749221802,
+      "learning_rate": 0.0001,
+      "loss": 1.2492,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6689791024323398,
+      "grad_norm": 1.0164541006088257,
+      "learning_rate": 0.0001,
+      "loss": 0.9617,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6690647482014388,
+      "grad_norm": 0.7368618249893188,
+      "learning_rate": 0.0001,
+      "loss": 1.2487,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6691503939705379,
+      "grad_norm": 1.1002901792526245,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6692360397396369,
+      "grad_norm": 0.8152766823768616,
+      "learning_rate": 0.0001,
+      "loss": 1.1116,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6693216855087358,
+      "grad_norm": 0.7938683032989502,
+      "learning_rate": 0.0001,
+      "loss": 0.6742,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6694073312778349,
+      "grad_norm": 0.246659055352211,
+      "learning_rate": 0.0001,
+      "loss": 0.333,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6694929770469339,
+      "grad_norm": 0.8463900685310364,
+      "learning_rate": 0.0001,
+      "loss": 1.0626,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6695786228160329,
+      "grad_norm": 0.5959887504577637,
+      "learning_rate": 0.0001,
+      "loss": 0.4691,
+      "step": 7818
+    },
+    {
+      "epoch": 0.669664268585132,
+      "grad_norm": 0.2767575979232788,
+      "learning_rate": 0.0001,
+      "loss": 0.4535,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6697499143542309,
+      "grad_norm": 1.355996012687683,
+      "learning_rate": 0.0001,
+      "loss": 0.7822,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6698355601233299,
+      "grad_norm": 1.106752634048462,
+      "learning_rate": 0.0001,
+      "loss": 1.1423,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6699212058924289,
+      "grad_norm": 0.707421064376831,
+      "learning_rate": 0.0001,
+      "loss": 1.2981,
+      "step": 7822
+    },
+    {
+      "epoch": 0.670006851661528,
+      "grad_norm": 0.430410772562027,
+      "learning_rate": 0.0001,
+      "loss": 0.7354,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6700924974306269,
+      "grad_norm": 0.626253068447113,
+      "learning_rate": 0.0001,
+      "loss": 0.941,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6701781431997259,
+      "grad_norm": 0.6350778341293335,
+      "learning_rate": 0.0001,
+      "loss": 1.1424,
+      "step": 7825
+    },
+    {
+      "epoch": 0.670263788968825,
+      "grad_norm": 0.7083667516708374,
+      "learning_rate": 0.0001,
+      "loss": 1.1657,
+      "step": 7826
+    },
+    {
+      "epoch": 0.670349434737924,
+      "grad_norm": 0.4931880831718445,
+      "learning_rate": 0.0001,
+      "loss": 0.1979,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6704350805070229,
+      "grad_norm": 0.7227419018745422,
+      "learning_rate": 0.0001,
+      "loss": 0.8656,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6705207262761219,
+      "grad_norm": 0.6682107448577881,
+      "learning_rate": 0.0001,
+      "loss": 1.0914,
+      "step": 7829
+    },
+    {
+      "epoch": 0.670606372045221,
+      "grad_norm": 1.0935138463974,
+      "learning_rate": 0.0001,
+      "loss": 1.3274,
+      "step": 7830
+    },
+    {
+      "epoch": 0.67069201781432,
+      "grad_norm": 1.041372537612915,
+      "learning_rate": 0.0001,
+      "loss": 0.7935,
+      "step": 7831
+    },
+    {
+      "epoch": 0.670777663583419,
+      "grad_norm": 0.8339431285858154,
+      "learning_rate": 0.0001,
+      "loss": 1.0882,
+      "step": 7832
+    },
+    {
+      "epoch": 0.670863309352518,
+      "grad_norm": 0.5557489991188049,
+      "learning_rate": 0.0001,
+      "loss": 0.6568,
+      "step": 7833
+    },
+    {
+      "epoch": 0.670948955121617,
+      "grad_norm": 0.7754536271095276,
+      "learning_rate": 0.0001,
+      "loss": 1.0563,
+      "step": 7834
+    },
+    {
+      "epoch": 0.671034600890716,
+      "grad_norm": 0.5296995043754578,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 7835
+    },
+    {
+      "epoch": 0.671120246659815,
+      "grad_norm": 0.6325036883354187,
+      "learning_rate": 0.0001,
+      "loss": 0.7535,
+      "step": 7836
+    },
+    {
+      "epoch": 0.671205892428914,
+      "grad_norm": 1.0798885822296143,
+      "learning_rate": 0.0001,
+      "loss": 1.3992,
+      "step": 7837
+    },
+    {
+      "epoch": 0.671291538198013,
+      "grad_norm": 0.6491509675979614,
+      "learning_rate": 0.0001,
+      "loss": 0.9861,
+      "step": 7838
+    },
+    {
+      "epoch": 0.671377183967112,
+      "grad_norm": 0.8879920244216919,
+      "learning_rate": 0.0001,
+      "loss": 0.7488,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6714628297362111,
+      "grad_norm": 0.6026825904846191,
+      "learning_rate": 0.0001,
+      "loss": 1.2141,
+      "step": 7840
+    },
+    {
+      "epoch": 0.67154847550531,
+      "grad_norm": 0.8580632209777832,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 7841
+    },
+    {
+      "epoch": 0.671634121274409,
+      "grad_norm": 0.9532682299613953,
+      "learning_rate": 0.0001,
+      "loss": 0.7453,
+      "step": 7842
+    },
+    {
+      "epoch": 0.671719767043508,
+      "grad_norm": 0.575933575630188,
+      "learning_rate": 0.0001,
+      "loss": 0.3956,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6718054128126071,
+      "grad_norm": 0.6608785390853882,
+      "learning_rate": 0.0001,
+      "loss": 0.6497,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6718910585817061,
+      "grad_norm": 1.0872150659561157,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 7845
+    },
+    {
+      "epoch": 0.671976704350805,
+      "grad_norm": 1.2211171388626099,
+      "learning_rate": 0.0001,
+      "loss": 1.898,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6720623501199041,
+      "grad_norm": 0.5162544846534729,
+      "learning_rate": 0.0001,
+      "loss": 0.7339,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6721479958890031,
+      "grad_norm": 0.8099302053451538,
+      "learning_rate": 0.0001,
+      "loss": 1.2336,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6722336416581021,
+      "grad_norm": 0.8313835263252258,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6723192874272012,
+      "grad_norm": 0.7004876732826233,
+      "learning_rate": 0.0001,
+      "loss": 0.6694,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6724049331963001,
+      "grad_norm": 0.2859377861022949,
+      "learning_rate": 0.0001,
+      "loss": 0.3628,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6724905789653991,
+      "grad_norm": 0.856419026851654,
+      "learning_rate": 0.0001,
+      "loss": 1.0518,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6725762247344981,
+      "grad_norm": 0.4634850025177002,
+      "learning_rate": 0.0001,
+      "loss": 0.4066,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6726618705035972,
+      "grad_norm": 2.3316450119018555,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6727475162726961,
+      "grad_norm": 0.7412492036819458,
+      "learning_rate": 0.0001,
+      "loss": 0.8245,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6728331620417951,
+      "grad_norm": 0.758648693561554,
+      "learning_rate": 0.0001,
+      "loss": 1.0822,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6729188078108942,
+      "grad_norm": 0.6700330972671509,
+      "learning_rate": 0.0001,
+      "loss": 1.0463,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6730044535799932,
+      "grad_norm": 0.4358261227607727,
+      "learning_rate": 0.0001,
+      "loss": 0.1214,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6730900993490921,
+      "grad_norm": 0.535088837146759,
+      "learning_rate": 0.0001,
+      "loss": 0.6505,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6731757451181911,
+      "grad_norm": 0.7277849912643433,
+      "learning_rate": 0.0001,
+      "loss": 0.7166,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6732613908872902,
+      "grad_norm": 0.643203616142273,
+      "learning_rate": 0.0001,
+      "loss": 0.6993,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6733470366563892,
+      "grad_norm": 1.1350139379501343,
+      "learning_rate": 0.0001,
+      "loss": 1.1512,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6734326824254881,
+      "grad_norm": 0.9564943909645081,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6735183281945872,
+      "grad_norm": 1.1520880460739136,
+      "learning_rate": 0.0001,
+      "loss": 0.6383,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6736039739636862,
+      "grad_norm": 0.5930637121200562,
+      "learning_rate": 0.0001,
+      "loss": 0.4352,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6736896197327852,
+      "grad_norm": 0.7624469995498657,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6737752655018842,
+      "grad_norm": 1.409673810005188,
+      "learning_rate": 0.0001,
+      "loss": 1.1708,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6738609112709832,
+      "grad_norm": 0.4599895179271698,
+      "learning_rate": 0.0001,
+      "loss": 0.5169,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6739465570400822,
+      "grad_norm": 1.0830656290054321,
+      "learning_rate": 0.0001,
+      "loss": 1.0608,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6740322028091812,
+      "grad_norm": 0.9018991589546204,
+      "learning_rate": 0.0001,
+      "loss": 0.953,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6741178485782803,
+      "grad_norm": 0.6379257440567017,
+      "learning_rate": 0.0001,
+      "loss": 1.0511,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6742034943473793,
+      "grad_norm": 1.1614946126937866,
+      "learning_rate": 0.0001,
+      "loss": 1.1045,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6742891401164782,
+      "grad_norm": 0.5551683306694031,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6743747858855772,
+      "grad_norm": 1.1062079668045044,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6744604316546763,
+      "grad_norm": 0.6166083216667175,
+      "learning_rate": 0.0001,
+      "loss": 0.4977,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6745460774237753,
+      "grad_norm": 0.5086182355880737,
+      "learning_rate": 0.0001,
+      "loss": 0.582,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6746317231928742,
+      "grad_norm": 0.5998806953430176,
+      "learning_rate": 0.0001,
+      "loss": 0.9542,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6747173689619733,
+      "grad_norm": 1.0318044424057007,
+      "learning_rate": 0.0001,
+      "loss": 0.7245,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6748030147310723,
+      "grad_norm": 0.56178879737854,
+      "learning_rate": 0.0001,
+      "loss": 0.528,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6748886605001713,
+      "grad_norm": 0.28523150086402893,
+      "learning_rate": 0.0001,
+      "loss": 0.2102,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6749743062692704,
+      "grad_norm": 1.0909301042556763,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6750599520383693,
+      "grad_norm": 0.37899520993232727,
+      "learning_rate": 0.0001,
+      "loss": 0.4497,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6751455978074683,
+      "grad_norm": 0.6949790716171265,
+      "learning_rate": 0.0001,
+      "loss": 0.7429,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6752312435765673,
+      "grad_norm": 0.6974218487739563,
+      "learning_rate": 0.0001,
+      "loss": 0.9803,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6753168893456664,
+      "grad_norm": 0.579366147518158,
+      "learning_rate": 0.0001,
+      "loss": 0.5397,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6754025351147653,
+      "grad_norm": 0.5251124501228333,
+      "learning_rate": 0.0001,
+      "loss": 0.9983,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6754881808838643,
+      "grad_norm": 0.7776161432266235,
+      "learning_rate": 0.0001,
+      "loss": 1.3102,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6755738266529634,
+      "grad_norm": 0.6341519355773926,
+      "learning_rate": 0.0001,
+      "loss": 0.6448,
+      "step": 7888
+    },
+    {
+      "epoch": 0.6756594724220624,
+      "grad_norm": 0.533046305179596,
+      "learning_rate": 0.0001,
+      "loss": 0.8153,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6757451181911613,
+      "grad_norm": 0.7608333230018616,
+      "learning_rate": 0.0001,
+      "loss": 0.6059,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6758307639602603,
+      "grad_norm": 0.7180198431015015,
+      "learning_rate": 0.0001,
+      "loss": 0.8834,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6759164097293594,
+      "grad_norm": 0.692480206489563,
+      "learning_rate": 0.0001,
+      "loss": 1.3191,
+      "step": 7892
+    },
+    {
+      "epoch": 0.6760020554984584,
+      "grad_norm": 0.6395915150642395,
+      "learning_rate": 0.0001,
+      "loss": 0.8351,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6760877012675574,
+      "grad_norm": 1.031226396560669,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6761733470366564,
+      "grad_norm": 1.093033790588379,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6762589928057554,
+      "grad_norm": 0.5448176264762878,
+      "learning_rate": 0.0001,
+      "loss": 0.3643,
+      "step": 7896
+    },
+    {
+      "epoch": 0.6763446385748544,
+      "grad_norm": 1.014332890510559,
+      "learning_rate": 0.0001,
+      "loss": 1.2315,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6764302843439534,
+      "grad_norm": 0.8813629746437073,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6765159301130524,
+      "grad_norm": 1.1641566753387451,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6766015758821514,
+      "grad_norm": 0.6034997701644897,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6766872216512504,
+      "grad_norm": 0.815203845500946,
+      "learning_rate": 0.0001,
+      "loss": 0.939,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6767728674203495,
+      "grad_norm": 0.6934569478034973,
+      "learning_rate": 0.0001,
+      "loss": 0.999,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6768585131894485,
+      "grad_norm": 0.5350626707077026,
+      "learning_rate": 0.0001,
+      "loss": 0.5963,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6769441589585474,
+      "grad_norm": 1.1951621770858765,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6770298047276464,
+      "grad_norm": 0.5450400114059448,
+      "learning_rate": 0.0001,
+      "loss": 0.1753,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6771154504967455,
+      "grad_norm": 1.333815097808838,
+      "learning_rate": 0.0001,
+      "loss": 1.0613,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6772010962658445,
+      "grad_norm": 0.6217408776283264,
+      "learning_rate": 0.0001,
+      "loss": 0.9341,
+      "step": 7907
+    },
+    {
+      "epoch": 0.6772867420349434,
+      "grad_norm": 0.554595947265625,
+      "learning_rate": 0.0001,
+      "loss": 1.0386,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6773723878040425,
+      "grad_norm": 0.6860207319259644,
+      "learning_rate": 0.0001,
+      "loss": 0.7098,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6774580335731415,
+      "grad_norm": 1.0966235399246216,
+      "learning_rate": 0.0001,
+      "loss": 0.6534,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6775436793422405,
+      "grad_norm": 1.205595850944519,
+      "learning_rate": 0.0001,
+      "loss": 1.0627,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6776293251113396,
+      "grad_norm": 1.3146849870681763,
+      "learning_rate": 0.0001,
+      "loss": 2.1796,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6777149708804385,
+      "grad_norm": 0.6186277866363525,
+      "learning_rate": 0.0001,
+      "loss": 0.5942,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6778006166495375,
+      "grad_norm": 0.7706251740455627,
+      "learning_rate": 0.0001,
+      "loss": 0.6515,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6778862624186365,
+      "grad_norm": 1.1959699392318726,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6779719081877356,
+      "grad_norm": 0.8474946618080139,
+      "learning_rate": 0.0001,
+      "loss": 0.8376,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6780575539568345,
+      "grad_norm": 1.501881718635559,
+      "learning_rate": 0.0001,
+      "loss": 0.9442,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6781431997259335,
+      "grad_norm": 0.5123477578163147,
+      "learning_rate": 0.0001,
+      "loss": 0.4909,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6782288454950326,
+      "grad_norm": 0.8036922812461853,
+      "learning_rate": 0.0001,
+      "loss": 0.8128,
+      "step": 7919
+    },
+    {
+      "epoch": 0.6783144912641316,
+      "grad_norm": 0.9528365135192871,
+      "learning_rate": 0.0001,
+      "loss": 0.8767,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6784001370332305,
+      "grad_norm": 1.0542346239089966,
+      "learning_rate": 0.0001,
+      "loss": 1.2767,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6784857828023295,
+      "grad_norm": 0.5875850319862366,
+      "learning_rate": 0.0001,
+      "loss": 0.2829,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 0.7848888635635376,
+      "learning_rate": 0.0001,
+      "loss": 0.6998,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6786570743405276,
+      "grad_norm": 0.6574537754058838,
+      "learning_rate": 0.0001,
+      "loss": 1.0579,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6787427201096266,
+      "grad_norm": 0.7797162532806396,
+      "learning_rate": 0.0001,
+      "loss": 0.7071,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6788283658787256,
+      "grad_norm": 0.7138040661811829,
+      "learning_rate": 0.0001,
+      "loss": 1.0173,
+      "step": 7926
+    },
+    {
+      "epoch": 0.6789140116478246,
+      "grad_norm": 0.7631307244300842,
+      "learning_rate": 0.0001,
+      "loss": 1.1604,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6789996574169236,
+      "grad_norm": 1.331123948097229,
+      "learning_rate": 0.0001,
+      "loss": 0.8469,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6790853031860226,
+      "grad_norm": 1.0098873376846313,
+      "learning_rate": 0.0001,
+      "loss": 1.9466,
+      "step": 7929
+    },
+    {
+      "epoch": 0.6791709489551216,
+      "grad_norm": 1.0650442838668823,
+      "learning_rate": 0.0001,
+      "loss": 1.3219,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6792565947242206,
+      "grad_norm": 1.002076506614685,
+      "learning_rate": 0.0001,
+      "loss": 1.0984,
+      "step": 7931
+    },
+    {
+      "epoch": 0.6793422404933196,
+      "grad_norm": 1.2275643348693848,
+      "learning_rate": 0.0001,
+      "loss": 1.2652,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6794278862624187,
+      "grad_norm": 1.3350635766983032,
+      "learning_rate": 0.0001,
+      "loss": 0.7703,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6795135320315177,
+      "grad_norm": 1.21095609664917,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6795991778006166,
+      "grad_norm": 1.398032546043396,
+      "learning_rate": 0.0001,
+      "loss": 1.2245,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6796848235697156,
+      "grad_norm": 0.6274577975273132,
+      "learning_rate": 0.0001,
+      "loss": 0.7194,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6797704693388147,
+      "grad_norm": 0.5837054252624512,
+      "learning_rate": 0.0001,
+      "loss": 0.8375,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6798561151079137,
+      "grad_norm": 0.8843039870262146,
+      "learning_rate": 0.0001,
+      "loss": 1.1613,
+      "step": 7938
+    },
+    {
+      "epoch": 0.6799417608770126,
+      "grad_norm": 0.852579653263092,
+      "learning_rate": 0.0001,
+      "loss": 0.5168,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6800274066461117,
+      "grad_norm": 1.0465435981750488,
+      "learning_rate": 0.0001,
+      "loss": 1.0016,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6801130524152107,
+      "grad_norm": 0.9000258445739746,
+      "learning_rate": 0.0001,
+      "loss": 0.4219,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6801986981843097,
+      "grad_norm": 0.6562278866767883,
+      "learning_rate": 0.0001,
+      "loss": 0.6621,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6802843439534088,
+      "grad_norm": 1.3136217594146729,
+      "learning_rate": 0.0001,
+      "loss": 1.3409,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6803699897225077,
+      "grad_norm": 0.7760037183761597,
+      "learning_rate": 0.0001,
+      "loss": 0.7855,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6804556354916067,
+      "grad_norm": 0.8817828893661499,
+      "learning_rate": 0.0001,
+      "loss": 1.0435,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6805412812607057,
+      "grad_norm": 1.1062430143356323,
+      "learning_rate": 0.0001,
+      "loss": 0.75,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6806269270298048,
+      "grad_norm": 0.7562169432640076,
+      "learning_rate": 0.0001,
+      "loss": 1.221,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6807125727989037,
+      "grad_norm": 1.1076816320419312,
+      "learning_rate": 0.0001,
+      "loss": 1.0815,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6807982185680027,
+      "grad_norm": 0.844016969203949,
+      "learning_rate": 0.0001,
+      "loss": 0.9252,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6808838643371018,
+      "grad_norm": 0.8175241947174072,
+      "learning_rate": 0.0001,
+      "loss": 0.9009,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6809695101062008,
+      "grad_norm": 0.8921611905097961,
+      "learning_rate": 0.0001,
+      "loss": 1.0614,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6810551558752997,
+      "grad_norm": 1.089821219444275,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6811408016443987,
+      "grad_norm": 1.091033935546875,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6812264474134978,
+      "grad_norm": 0.25962111353874207,
+      "learning_rate": 0.0001,
+      "loss": 0.3249,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6813120931825968,
+      "grad_norm": 0.7406303286552429,
+      "learning_rate": 0.0001,
+      "loss": 1.2396,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6813977389516958,
+      "grad_norm": 0.29838910698890686,
+      "learning_rate": 0.0001,
+      "loss": 0.3419,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6814833847207948,
+      "grad_norm": 0.9556125402450562,
+      "learning_rate": 0.0001,
+      "loss": 0.9881,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6815690304898938,
+      "grad_norm": 0.7329519391059875,
+      "learning_rate": 0.0001,
+      "loss": 1.1383,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6816546762589928,
+      "grad_norm": 0.921617865562439,
+      "learning_rate": 0.0001,
+      "loss": 0.6861,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6817403220280918,
+      "grad_norm": 0.6274394392967224,
+      "learning_rate": 0.0001,
+      "loss": 0.5561,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6818259677971908,
+      "grad_norm": 0.7916021943092346,
+      "learning_rate": 0.0001,
+      "loss": 0.6924,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6819116135662898,
+      "grad_norm": 1.5611653327941895,
+      "learning_rate": 0.0001,
+      "loss": 1.2964,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6819972593353888,
+      "grad_norm": 0.9337931275367737,
+      "learning_rate": 0.0001,
+      "loss": 0.9694,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6820829051044879,
+      "grad_norm": 1.0455467700958252,
+      "learning_rate": 0.0001,
+      "loss": 0.907,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6821685508735869,
+      "grad_norm": 1.0880913734436035,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6822541966426858,
+      "grad_norm": 0.38289400935173035,
+      "learning_rate": 0.0001,
+      "loss": 0.407,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6823398424117848,
+      "grad_norm": 0.8944470882415771,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6824254881808839,
+      "grad_norm": 0.6732245087623596,
+      "learning_rate": 0.0001,
+      "loss": 1.2234,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6825111339499829,
+      "grad_norm": 1.584082841873169,
+      "learning_rate": 0.0001,
+      "loss": 1.2481,
+      "step": 7969
+    },
+    {
+      "epoch": 0.6825967797190818,
+      "grad_norm": 0.6763570308685303,
+      "learning_rate": 0.0001,
+      "loss": 0.7468,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6826824254881809,
+      "grad_norm": 0.5743582248687744,
+      "learning_rate": 0.0001,
+      "loss": 0.4613,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6827680712572799,
+      "grad_norm": 0.8540172576904297,
+      "learning_rate": 0.0001,
+      "loss": 0.9288,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6828537170263789,
+      "grad_norm": 0.8274142742156982,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 7973
+    },
+    {
+      "epoch": 0.682939362795478,
+      "grad_norm": 0.9762735366821289,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6830250085645769,
+      "grad_norm": 1.2052226066589355,
+      "learning_rate": 0.0001,
+      "loss": 0.7324,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6831106543336759,
+      "grad_norm": 1.3383748531341553,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6831963001027749,
+      "grad_norm": 1.2557731866836548,
+      "learning_rate": 0.0001,
+      "loss": 1.3207,
+      "step": 7977
+    },
+    {
+      "epoch": 0.683281945871874,
+      "grad_norm": 0.863298773765564,
+      "learning_rate": 0.0001,
+      "loss": 1.2545,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6833675916409729,
+      "grad_norm": 0.8722476363182068,
+      "learning_rate": 0.0001,
+      "loss": 1.1172,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6834532374100719,
+      "grad_norm": 1.1781283617019653,
+      "learning_rate": 0.0001,
+      "loss": 0.9822,
+      "step": 7980
+    },
+    {
+      "epoch": 0.683538883179171,
+      "grad_norm": 0.6666592359542847,
+      "learning_rate": 0.0001,
+      "loss": 1.0085,
+      "step": 7981
+    },
+    {
+      "epoch": 0.68362452894827,
+      "grad_norm": 0.8909516930580139,
+      "learning_rate": 0.0001,
+      "loss": 1.116,
+      "step": 7982
+    },
+    {
+      "epoch": 0.683710174717369,
+      "grad_norm": 0.9032033681869507,
+      "learning_rate": 0.0001,
+      "loss": 0.5777,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6837958204864679,
+      "grad_norm": 0.6664184927940369,
+      "learning_rate": 0.0001,
+      "loss": 0.6539,
+      "step": 7984
+    },
+    {
+      "epoch": 0.683881466255567,
+      "grad_norm": 0.7337456941604614,
+      "learning_rate": 0.0001,
+      "loss": 1.1897,
+      "step": 7985
+    },
+    {
+      "epoch": 0.683967112024666,
+      "grad_norm": 0.6670850515365601,
+      "learning_rate": 0.0001,
+      "loss": 0.6366,
+      "step": 7986
+    },
+    {
+      "epoch": 0.684052757793765,
+      "grad_norm": 0.6831005215644836,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 7987
+    },
+    {
+      "epoch": 0.684138403562864,
+      "grad_norm": 0.9407575726509094,
+      "learning_rate": 0.0001,
+      "loss": 0.6282,
+      "step": 7988
+    },
+    {
+      "epoch": 0.684224049331963,
+      "grad_norm": 0.5608586072921753,
+      "learning_rate": 0.0001,
+      "loss": 0.8327,
+      "step": 7989
+    },
+    {
+      "epoch": 0.684309695101062,
+      "grad_norm": 1.437313199043274,
+      "learning_rate": 0.0001,
+      "loss": 0.635,
+      "step": 7990
+    },
+    {
+      "epoch": 0.684395340870161,
+      "grad_norm": 0.7221481800079346,
+      "learning_rate": 0.0001,
+      "loss": 0.6232,
+      "step": 7991
+    },
+    {
+      "epoch": 0.68448098663926,
+      "grad_norm": 0.7718965411186218,
+      "learning_rate": 0.0001,
+      "loss": 1.2554,
+      "step": 7992
+    },
+    {
+      "epoch": 0.684566632408359,
+      "grad_norm": 0.8702987432479858,
+      "learning_rate": 0.0001,
+      "loss": 1.4007,
+      "step": 7993
+    },
+    {
+      "epoch": 0.684652278177458,
+      "grad_norm": 1.0756090879440308,
+      "learning_rate": 0.0001,
+      "loss": 1.2132,
+      "step": 7994
+    },
+    {
+      "epoch": 0.6847379239465571,
+      "grad_norm": 0.4820309281349182,
+      "learning_rate": 0.0001,
+      "loss": 0.8776,
+      "step": 7995
+    },
+    {
+      "epoch": 0.684823569715656,
+      "grad_norm": 0.6127198934555054,
+      "learning_rate": 0.0001,
+      "loss": 0.6339,
+      "step": 7996
+    },
+    {
+      "epoch": 0.684909215484755,
+      "grad_norm": 0.5280531048774719,
+      "learning_rate": 0.0001,
+      "loss": 0.5653,
+      "step": 7997
+    },
+    {
+      "epoch": 0.6849948612538541,
+      "grad_norm": 0.6651861667633057,
+      "learning_rate": 0.0001,
+      "loss": 1.197,
+      "step": 7998
+    },
+    {
+      "epoch": 0.6850805070229531,
+      "grad_norm": 0.7665339112281799,
+      "learning_rate": 0.0001,
+      "loss": 1.0141,
+      "step": 7999
+    },
+    {
+      "epoch": 0.6851661527920521,
+      "grad_norm": 0.8754515051841736,
+      "learning_rate": 0.0001,
+      "loss": 0.6795,
+      "step": 8000
+    },
+    {
+      "epoch": 0.685251798561151,
+      "grad_norm": 0.5802514553070068,
+      "learning_rate": 0.0001,
+      "loss": 0.3705,
+      "step": 8001
+    },
+    {
+      "epoch": 0.6853374443302501,
+      "grad_norm": 1.1554185152053833,
+      "learning_rate": 0.0001,
+      "loss": 1.2553,
+      "step": 8002
+    },
+    {
+      "epoch": 0.6854230900993491,
+      "grad_norm": 0.9153088927268982,
+      "learning_rate": 0.0001,
+      "loss": 1.1473,
+      "step": 8003
+    },
+    {
+      "epoch": 0.6855087358684481,
+      "grad_norm": 1.025997519493103,
+      "learning_rate": 0.0001,
+      "loss": 1.1706,
+      "step": 8004
+    },
+    {
+      "epoch": 0.6855943816375472,
+      "grad_norm": 0.8191499710083008,
+      "learning_rate": 0.0001,
+      "loss": 0.4523,
+      "step": 8005
+    },
+    {
+      "epoch": 0.6856800274066461,
+      "grad_norm": 1.025879979133606,
+      "learning_rate": 0.0001,
+      "loss": 1.1051,
+      "step": 8006
+    },
+    {
+      "epoch": 0.6857656731757451,
+      "grad_norm": 0.7944890260696411,
+      "learning_rate": 0.0001,
+      "loss": 0.7037,
+      "step": 8007
+    },
+    {
+      "epoch": 0.6858513189448441,
+      "grad_norm": 0.5486347675323486,
+      "learning_rate": 0.0001,
+      "loss": 0.5662,
+      "step": 8008
+    },
+    {
+      "epoch": 0.6859369647139432,
+      "grad_norm": 0.9222549200057983,
+      "learning_rate": 0.0001,
+      "loss": 1.1232,
+      "step": 8009
+    },
+    {
+      "epoch": 0.6860226104830421,
+      "grad_norm": 0.7136363387107849,
+      "learning_rate": 0.0001,
+      "loss": 0.9564,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6861082562521411,
+      "grad_norm": 1.0706267356872559,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 8011
+    },
+    {
+      "epoch": 0.6861939020212402,
+      "grad_norm": 0.9239405989646912,
+      "learning_rate": 0.0001,
+      "loss": 0.8648,
+      "step": 8012
+    },
+    {
+      "epoch": 0.6862795477903392,
+      "grad_norm": 1.0235801935195923,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 8013
+    },
+    {
+      "epoch": 0.6863651935594381,
+      "grad_norm": 0.5163418650627136,
+      "learning_rate": 0.0001,
+      "loss": 0.672,
+      "step": 8014
+    },
+    {
+      "epoch": 0.6864508393285371,
+      "grad_norm": 0.7815303206443787,
+      "learning_rate": 0.0001,
+      "loss": 1.1071,
+      "step": 8015
+    },
+    {
+      "epoch": 0.6865364850976362,
+      "grad_norm": 0.5402613282203674,
+      "learning_rate": 0.0001,
+      "loss": 0.7592,
+      "step": 8016
+    },
+    {
+      "epoch": 0.6866221308667352,
+      "grad_norm": 0.8178493976593018,
+      "learning_rate": 0.0001,
+      "loss": 0.6075,
+      "step": 8017
+    },
+    {
+      "epoch": 0.6867077766358342,
+      "grad_norm": 1.265343427658081,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 8018
+    },
+    {
+      "epoch": 0.6867934224049332,
+      "grad_norm": 0.667504608631134,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 8019
+    },
+    {
+      "epoch": 0.6868790681740322,
+      "grad_norm": 0.7695797681808472,
+      "learning_rate": 0.0001,
+      "loss": 0.7859,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6869647139431312,
+      "grad_norm": 0.9763283729553223,
+      "learning_rate": 0.0001,
+      "loss": 0.9166,
+      "step": 8021
+    },
+    {
+      "epoch": 0.6870503597122302,
+      "grad_norm": 0.6980285048484802,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 8022
+    },
+    {
+      "epoch": 0.6871360054813292,
+      "grad_norm": 1.0822428464889526,
+      "learning_rate": 0.0001,
+      "loss": 0.8891,
+      "step": 8023
+    },
+    {
+      "epoch": 0.6872216512504282,
+      "grad_norm": 1.2837536334991455,
+      "learning_rate": 0.0001,
+      "loss": 0.8887,
+      "step": 8024
+    },
+    {
+      "epoch": 0.6873072970195272,
+      "grad_norm": 0.8008579611778259,
+      "learning_rate": 0.0001,
+      "loss": 1.2709,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6873929427886263,
+      "grad_norm": 0.8854803442955017,
+      "learning_rate": 0.0001,
+      "loss": 1.1495,
+      "step": 8026
+    },
+    {
+      "epoch": 0.6874785885577253,
+      "grad_norm": 0.7141470909118652,
+      "learning_rate": 0.0001,
+      "loss": 0.9409,
+      "step": 8027
+    },
+    {
+      "epoch": 0.6875642343268242,
+      "grad_norm": 0.9534090757369995,
+      "learning_rate": 0.0001,
+      "loss": 0.731,
+      "step": 8028
+    },
+    {
+      "epoch": 0.6876498800959233,
+      "grad_norm": 0.6924055218696594,
+      "learning_rate": 0.0001,
+      "loss": 0.8628,
+      "step": 8029
+    },
+    {
+      "epoch": 0.6877355258650223,
+      "grad_norm": 0.9518143534660339,
+      "learning_rate": 0.0001,
+      "loss": 1.0178,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6878211716341213,
+      "grad_norm": 0.8779423236846924,
+      "learning_rate": 0.0001,
+      "loss": 1.6804,
+      "step": 8031
+    },
+    {
+      "epoch": 0.6879068174032202,
+      "grad_norm": 0.9515752196311951,
+      "learning_rate": 0.0001,
+      "loss": 1.1373,
+      "step": 8032
+    },
+    {
+      "epoch": 0.6879924631723193,
+      "grad_norm": 0.7800067663192749,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 8033
+    },
+    {
+      "epoch": 0.6880781089414183,
+      "grad_norm": 1.176525592803955,
+      "learning_rate": 0.0001,
+      "loss": 1.3545,
+      "step": 8034
+    },
+    {
+      "epoch": 0.6881637547105173,
+      "grad_norm": 0.37632235884666443,
+      "learning_rate": 0.0001,
+      "loss": 0.4209,
+      "step": 8035
+    },
+    {
+      "epoch": 0.6882494004796164,
+      "grad_norm": 0.7566089034080505,
+      "learning_rate": 0.0001,
+      "loss": 0.7411,
+      "step": 8036
+    },
+    {
+      "epoch": 0.6883350462487153,
+      "grad_norm": 1.0247067213058472,
+      "learning_rate": 0.0001,
+      "loss": 0.7655,
+      "step": 8037
+    },
+    {
+      "epoch": 0.6884206920178143,
+      "grad_norm": 1.0313074588775635,
+      "learning_rate": 0.0001,
+      "loss": 1.092,
+      "step": 8038
+    },
+    {
+      "epoch": 0.6885063377869133,
+      "grad_norm": 0.8517690896987915,
+      "learning_rate": 0.0001,
+      "loss": 0.9722,
+      "step": 8039
+    },
+    {
+      "epoch": 0.6885919835560124,
+      "grad_norm": 0.984847903251648,
+      "learning_rate": 0.0001,
+      "loss": 1.2496,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6886776293251113,
+      "grad_norm": 1.7496427297592163,
+      "learning_rate": 0.0001,
+      "loss": 1.2534,
+      "step": 8041
+    },
+    {
+      "epoch": 0.6887632750942103,
+      "grad_norm": 0.802437961101532,
+      "learning_rate": 0.0001,
+      "loss": 0.5315,
+      "step": 8042
+    },
+    {
+      "epoch": 0.6888489208633094,
+      "grad_norm": 1.4126074314117432,
+      "learning_rate": 0.0001,
+      "loss": 1.2559,
+      "step": 8043
+    },
+    {
+      "epoch": 0.6889345666324084,
+      "grad_norm": 0.5976142883300781,
+      "learning_rate": 0.0001,
+      "loss": 0.9148,
+      "step": 8044
+    },
+    {
+      "epoch": 0.6890202124015073,
+      "grad_norm": 0.6746236085891724,
+      "learning_rate": 0.0001,
+      "loss": 1.0408,
+      "step": 8045
+    },
+    {
+      "epoch": 0.6891058581706063,
+      "grad_norm": 0.7025140523910522,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 8046
+    },
+    {
+      "epoch": 0.6891915039397054,
+      "grad_norm": 1.1325973272323608,
+      "learning_rate": 0.0001,
+      "loss": 1.0284,
+      "step": 8047
+    },
+    {
+      "epoch": 0.6892771497088044,
+      "grad_norm": 0.9242881536483765,
+      "learning_rate": 0.0001,
+      "loss": 0.7987,
+      "step": 8048
+    },
+    {
+      "epoch": 0.6893627954779034,
+      "grad_norm": 0.5900736451148987,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 8049
+    },
+    {
+      "epoch": 0.6894484412470024,
+      "grad_norm": 0.6058974266052246,
+      "learning_rate": 0.0001,
+      "loss": 1.1683,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6895340870161014,
+      "grad_norm": 0.8993467688560486,
+      "learning_rate": 0.0001,
+      "loss": 0.7241,
+      "step": 8051
+    },
+    {
+      "epoch": 0.6896197327852004,
+      "grad_norm": 1.0706961154937744,
+      "learning_rate": 0.0001,
+      "loss": 1.3176,
+      "step": 8052
+    },
+    {
+      "epoch": 0.6897053785542994,
+      "grad_norm": 0.6177393198013306,
+      "learning_rate": 0.0001,
+      "loss": 0.9829,
+      "step": 8053
+    },
+    {
+      "epoch": 0.6897910243233984,
+      "grad_norm": 0.3917364776134491,
+      "learning_rate": 0.0001,
+      "loss": 0.4584,
+      "step": 8054
+    },
+    {
+      "epoch": 0.6898766700924974,
+      "grad_norm": 1.0003212690353394,
+      "learning_rate": 0.0001,
+      "loss": 0.522,
+      "step": 8055
+    },
+    {
+      "epoch": 0.6899623158615964,
+      "grad_norm": 0.549935519695282,
+      "learning_rate": 0.0001,
+      "loss": 1.0616,
+      "step": 8056
+    },
+    {
+      "epoch": 0.6900479616306955,
+      "grad_norm": 1.3138242959976196,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 8057
+    },
+    {
+      "epoch": 0.6901336073997945,
+      "grad_norm": 0.90514075756073,
+      "learning_rate": 0.0001,
+      "loss": 1.0982,
+      "step": 8058
+    },
+    {
+      "epoch": 0.6902192531688934,
+      "grad_norm": 0.619576096534729,
+      "learning_rate": 0.0001,
+      "loss": 1.2166,
+      "step": 8059
+    },
+    {
+      "epoch": 0.6903048989379925,
+      "grad_norm": 1.1131598949432373,
+      "learning_rate": 0.0001,
+      "loss": 1.3185,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6903905447070915,
+      "grad_norm": 1.0947893857955933,
+      "learning_rate": 0.0001,
+      "loss": 0.7141,
+      "step": 8061
+    },
+    {
+      "epoch": 0.6904761904761905,
+      "grad_norm": 0.5667213201522827,
+      "learning_rate": 0.0001,
+      "loss": 1.1867,
+      "step": 8062
+    },
+    {
+      "epoch": 0.6905618362452894,
+      "grad_norm": 0.8494309186935425,
+      "learning_rate": 0.0001,
+      "loss": 0.5597,
+      "step": 8063
+    },
+    {
+      "epoch": 0.6906474820143885,
+      "grad_norm": 0.9518434405326843,
+      "learning_rate": 0.0001,
+      "loss": 0.3288,
+      "step": 8064
+    },
+    {
+      "epoch": 0.6907331277834875,
+      "grad_norm": 0.7854611277580261,
+      "learning_rate": 0.0001,
+      "loss": 1.0947,
+      "step": 8065
+    },
+    {
+      "epoch": 0.6908187735525865,
+      "grad_norm": 1.0192670822143555,
+      "learning_rate": 0.0001,
+      "loss": 1.2564,
+      "step": 8066
+    },
+    {
+      "epoch": 0.6909044193216856,
+      "grad_norm": 0.872220516204834,
+      "learning_rate": 0.0001,
+      "loss": 0.6827,
+      "step": 8067
+    },
+    {
+      "epoch": 0.6909900650907845,
+      "grad_norm": 0.7572110891342163,
+      "learning_rate": 0.0001,
+      "loss": 0.534,
+      "step": 8068
+    },
+    {
+      "epoch": 0.6910757108598835,
+      "grad_norm": 0.9249520301818848,
+      "learning_rate": 0.0001,
+      "loss": 1.1282,
+      "step": 8069
+    },
+    {
+      "epoch": 0.6911613566289825,
+      "grad_norm": 1.087942123413086,
+      "learning_rate": 0.0001,
+      "loss": 0.6477,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6912470023980816,
+      "grad_norm": 0.8615002632141113,
+      "learning_rate": 0.0001,
+      "loss": 0.9215,
+      "step": 8071
+    },
+    {
+      "epoch": 0.6913326481671805,
+      "grad_norm": 0.9302707314491272,
+      "learning_rate": 0.0001,
+      "loss": 0.7703,
+      "step": 8072
+    },
+    {
+      "epoch": 0.6914182939362795,
+      "grad_norm": 1.0399943590164185,
+      "learning_rate": 0.0001,
+      "loss": 0.6344,
+      "step": 8073
+    },
+    {
+      "epoch": 0.6915039397053786,
+      "grad_norm": 1.3429605960845947,
+      "learning_rate": 0.0001,
+      "loss": 1.248,
+      "step": 8074
+    },
+    {
+      "epoch": 0.6915895854744776,
+      "grad_norm": 1.0074509382247925,
+      "learning_rate": 0.0001,
+      "loss": 0.8839,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6916752312435765,
+      "grad_norm": 0.49957898259162903,
+      "learning_rate": 0.0001,
+      "loss": 0.8616,
+      "step": 8076
+    },
+    {
+      "epoch": 0.6917608770126755,
+      "grad_norm": 1.0756094455718994,
+      "learning_rate": 0.0001,
+      "loss": 1.3135,
+      "step": 8077
+    },
+    {
+      "epoch": 0.6918465227817746,
+      "grad_norm": 0.6887243390083313,
+      "learning_rate": 0.0001,
+      "loss": 0.928,
+      "step": 8078
+    },
+    {
+      "epoch": 0.6919321685508736,
+      "grad_norm": 0.7180985808372498,
+      "learning_rate": 0.0001,
+      "loss": 1.2439,
+      "step": 8079
+    },
+    {
+      "epoch": 0.6920178143199726,
+      "grad_norm": 1.1320688724517822,
+      "learning_rate": 0.0001,
+      "loss": 0.6991,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6921034600890716,
+      "grad_norm": 0.871765673160553,
+      "learning_rate": 0.0001,
+      "loss": 0.2643,
+      "step": 8081
+    },
+    {
+      "epoch": 0.6921891058581706,
+      "grad_norm": 0.5721661448478699,
+      "learning_rate": 0.0001,
+      "loss": 0.8087,
+      "step": 8082
+    },
+    {
+      "epoch": 0.6922747516272696,
+      "grad_norm": 1.0635148286819458,
+      "learning_rate": 0.0001,
+      "loss": 1.1416,
+      "step": 8083
+    },
+    {
+      "epoch": 0.6923603973963686,
+      "grad_norm": 0.7985076904296875,
+      "learning_rate": 0.0001,
+      "loss": 1.0838,
+      "step": 8084
+    },
+    {
+      "epoch": 0.6924460431654677,
+      "grad_norm": 0.8378544449806213,
+      "learning_rate": 0.0001,
+      "loss": 1.2854,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6925316889345666,
+      "grad_norm": 0.5118926763534546,
+      "learning_rate": 0.0001,
+      "loss": 1.1619,
+      "step": 8086
+    },
+    {
+      "epoch": 0.6926173347036656,
+      "grad_norm": 0.8267455697059631,
+      "learning_rate": 0.0001,
+      "loss": 0.9227,
+      "step": 8087
+    },
+    {
+      "epoch": 0.6927029804727647,
+      "grad_norm": 0.8779959082603455,
+      "learning_rate": 0.0001,
+      "loss": 0.9373,
+      "step": 8088
+    },
+    {
+      "epoch": 0.6927886262418637,
+      "grad_norm": 0.7912876009941101,
+      "learning_rate": 0.0001,
+      "loss": 0.9831,
+      "step": 8089
+    },
+    {
+      "epoch": 0.6928742720109626,
+      "grad_norm": 0.9020532369613647,
+      "learning_rate": 0.0001,
+      "loss": 0.9581,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6929599177800617,
+      "grad_norm": 0.7153639197349548,
+      "learning_rate": 0.0001,
+      "loss": 0.4765,
+      "step": 8091
+    },
+    {
+      "epoch": 0.6930455635491607,
+      "grad_norm": 1.0509922504425049,
+      "learning_rate": 0.0001,
+      "loss": 1.276,
+      "step": 8092
+    },
+    {
+      "epoch": 0.6931312093182597,
+      "grad_norm": 0.5672833919525146,
+      "learning_rate": 0.0001,
+      "loss": 0.9694,
+      "step": 8093
+    },
+    {
+      "epoch": 0.6932168550873586,
+      "grad_norm": 0.525810956954956,
+      "learning_rate": 0.0001,
+      "loss": 0.2135,
+      "step": 8094
+    },
+    {
+      "epoch": 0.6933025008564577,
+      "grad_norm": 0.9352670907974243,
+      "learning_rate": 0.0001,
+      "loss": 1.276,
+      "step": 8095
+    },
+    {
+      "epoch": 0.6933881466255567,
+      "grad_norm": 0.774359405040741,
+      "learning_rate": 0.0001,
+      "loss": 1.2486,
+      "step": 8096
+    },
+    {
+      "epoch": 0.6934737923946557,
+      "grad_norm": 0.735844612121582,
+      "learning_rate": 0.0001,
+      "loss": 1.9107,
+      "step": 8097
+    },
+    {
+      "epoch": 0.6935594381637548,
+      "grad_norm": 1.2306641340255737,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 8098
+    },
+    {
+      "epoch": 0.6936450839328537,
+      "grad_norm": 0.7132367491722107,
+      "learning_rate": 0.0001,
+      "loss": 1.3847,
+      "step": 8099
+    },
+    {
+      "epoch": 0.6937307297019527,
+      "grad_norm": 1.285535216331482,
+      "learning_rate": 0.0001,
+      "loss": 1.0551,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6938163754710517,
+      "grad_norm": 0.6416744589805603,
+      "learning_rate": 0.0001,
+      "loss": 0.4609,
+      "step": 8101
+    },
+    {
+      "epoch": 0.6939020212401508,
+      "grad_norm": 0.5656083226203918,
+      "learning_rate": 0.0001,
+      "loss": 1.1304,
+      "step": 8102
+    },
+    {
+      "epoch": 0.6939876670092497,
+      "grad_norm": 0.9156350493431091,
+      "learning_rate": 0.0001,
+      "loss": 0.7209,
+      "step": 8103
+    },
+    {
+      "epoch": 0.6940733127783487,
+      "grad_norm": 0.5752999782562256,
+      "learning_rate": 0.0001,
+      "loss": 0.6159,
+      "step": 8104
+    },
+    {
+      "epoch": 0.6941589585474478,
+      "grad_norm": 0.5255430340766907,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6942446043165468,
+      "grad_norm": 0.547609806060791,
+      "learning_rate": 0.0001,
+      "loss": 0.492,
+      "step": 8106
+    },
+    {
+      "epoch": 0.6943302500856458,
+      "grad_norm": 0.8411626815795898,
+      "learning_rate": 0.0001,
+      "loss": 1.1262,
+      "step": 8107
+    },
+    {
+      "epoch": 0.6944158958547447,
+      "grad_norm": 0.644541323184967,
+      "learning_rate": 0.0001,
+      "loss": 1.0483,
+      "step": 8108
+    },
+    {
+      "epoch": 0.6945015416238438,
+      "grad_norm": 0.9116525053977966,
+      "learning_rate": 0.0001,
+      "loss": 0.7992,
+      "step": 8109
+    },
+    {
+      "epoch": 0.6945871873929428,
+      "grad_norm": 0.46233069896698,
+      "learning_rate": 0.0001,
+      "loss": 0.2579,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6946728331620418,
+      "grad_norm": 0.9624473452568054,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 8111
+    },
+    {
+      "epoch": 0.6947584789311408,
+      "grad_norm": 1.2274088859558105,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 8112
+    },
+    {
+      "epoch": 0.6948441247002398,
+      "grad_norm": 0.8253557682037354,
+      "learning_rate": 0.0001,
+      "loss": 0.7066,
+      "step": 8113
+    },
+    {
+      "epoch": 0.6949297704693388,
+      "grad_norm": 0.9089987277984619,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 8114
+    },
+    {
+      "epoch": 0.6950154162384378,
+      "grad_norm": 0.5750834345817566,
+      "learning_rate": 0.0001,
+      "loss": 0.936,
+      "step": 8115
+    },
+    {
+      "epoch": 0.6951010620075369,
+      "grad_norm": 1.3187369108200073,
+      "learning_rate": 0.0001,
+      "loss": 1.133,
+      "step": 8116
+    },
+    {
+      "epoch": 0.6951867077766358,
+      "grad_norm": 0.8922926783561707,
+      "learning_rate": 0.0001,
+      "loss": 0.9197,
+      "step": 8117
+    },
+    {
+      "epoch": 0.6952723535457348,
+      "grad_norm": 0.9719269275665283,
+      "learning_rate": 0.0001,
+      "loss": 0.8771,
+      "step": 8118
+    },
+    {
+      "epoch": 0.6953579993148339,
+      "grad_norm": 1.1650378704071045,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 8119
+    },
+    {
+      "epoch": 0.6954436450839329,
+      "grad_norm": 0.7795016765594482,
+      "learning_rate": 0.0001,
+      "loss": 0.7225,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6955292908530318,
+      "grad_norm": 1.025094985961914,
+      "learning_rate": 0.0001,
+      "loss": 1.2254,
+      "step": 8121
+    },
+    {
+      "epoch": 0.6956149366221309,
+      "grad_norm": 0.8067141175270081,
+      "learning_rate": 0.0001,
+      "loss": 1.3244,
+      "step": 8122
+    },
+    {
+      "epoch": 0.6957005823912299,
+      "grad_norm": 0.9172433614730835,
+      "learning_rate": 0.0001,
+      "loss": 0.7573,
+      "step": 8123
+    },
+    {
+      "epoch": 0.6957862281603289,
+      "grad_norm": 0.6621445417404175,
+      "learning_rate": 0.0001,
+      "loss": 0.5073,
+      "step": 8124
+    },
+    {
+      "epoch": 0.6958718739294278,
+      "grad_norm": 0.97459477186203,
+      "learning_rate": 0.0001,
+      "loss": 1.1896,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6959575196985269,
+      "grad_norm": 1.0381896495819092,
+      "learning_rate": 0.0001,
+      "loss": 0.9832,
+      "step": 8126
+    },
+    {
+      "epoch": 0.6960431654676259,
+      "grad_norm": 0.3977573812007904,
+      "learning_rate": 0.0001,
+      "loss": 0.7449,
+      "step": 8127
+    },
+    {
+      "epoch": 0.6961288112367249,
+      "grad_norm": 0.7312228679656982,
+      "learning_rate": 0.0001,
+      "loss": 0.9668,
+      "step": 8128
+    },
+    {
+      "epoch": 0.696214457005824,
+      "grad_norm": 1.0579332113265991,
+      "learning_rate": 0.0001,
+      "loss": 0.9412,
+      "step": 8129
+    },
+    {
+      "epoch": 0.6963001027749229,
+      "grad_norm": 1.1626601219177246,
+      "learning_rate": 0.0001,
+      "loss": 1.2029,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6963857485440219,
+      "grad_norm": 0.5843555927276611,
+      "learning_rate": 0.0001,
+      "loss": 0.8748,
+      "step": 8131
+    },
+    {
+      "epoch": 0.6964713943131209,
+      "grad_norm": 0.6484670042991638,
+      "learning_rate": 0.0001,
+      "loss": 1.2434,
+      "step": 8132
+    },
+    {
+      "epoch": 0.69655704008222,
+      "grad_norm": 0.7831952571868896,
+      "learning_rate": 0.0001,
+      "loss": 0.7083,
+      "step": 8133
+    },
+    {
+      "epoch": 0.6966426858513189,
+      "grad_norm": 1.0525215864181519,
+      "learning_rate": 0.0001,
+      "loss": 1.7642,
+      "step": 8134
+    },
+    {
+      "epoch": 0.6967283316204179,
+      "grad_norm": 0.9662362337112427,
+      "learning_rate": 0.0001,
+      "loss": 1.2783,
+      "step": 8135
+    },
+    {
+      "epoch": 0.696813977389517,
+      "grad_norm": 0.9095815420150757,
+      "learning_rate": 0.0001,
+      "loss": 0.7742,
+      "step": 8136
+    },
+    {
+      "epoch": 0.696899623158616,
+      "grad_norm": 0.4570182263851166,
+      "learning_rate": 0.0001,
+      "loss": 0.4717,
+      "step": 8137
+    },
+    {
+      "epoch": 0.696985268927715,
+      "grad_norm": 0.4171493649482727,
+      "learning_rate": 0.0001,
+      "loss": 0.5617,
+      "step": 8138
+    },
+    {
+      "epoch": 0.6970709146968139,
+      "grad_norm": 1.422377586364746,
+      "learning_rate": 0.0001,
+      "loss": 1.2434,
+      "step": 8139
+    },
+    {
+      "epoch": 0.697156560465913,
+      "grad_norm": 0.8981227278709412,
+      "learning_rate": 0.0001,
+      "loss": 0.7621,
+      "step": 8140
+    },
+    {
+      "epoch": 0.697242206235012,
+      "grad_norm": 0.9309478402137756,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 8141
+    },
+    {
+      "epoch": 0.697327852004111,
+      "grad_norm": 0.6460554599761963,
+      "learning_rate": 0.0001,
+      "loss": 1.1847,
+      "step": 8142
+    },
+    {
+      "epoch": 0.69741349777321,
+      "grad_norm": 0.9679590463638306,
+      "learning_rate": 0.0001,
+      "loss": 1.1478,
+      "step": 8143
+    },
+    {
+      "epoch": 0.697499143542309,
+      "grad_norm": 0.6526610851287842,
+      "learning_rate": 0.0001,
+      "loss": 0.4759,
+      "step": 8144
+    },
+    {
+      "epoch": 0.697584789311408,
+      "grad_norm": 0.925771176815033,
+      "learning_rate": 0.0001,
+      "loss": 1.1844,
+      "step": 8145
+    },
+    {
+      "epoch": 0.697670435080507,
+      "grad_norm": 0.941368579864502,
+      "learning_rate": 0.0001,
+      "loss": 0.9836,
+      "step": 8146
+    },
+    {
+      "epoch": 0.697756080849606,
+      "grad_norm": 0.6188156008720398,
+      "learning_rate": 0.0001,
+      "loss": 0.8859,
+      "step": 8147
+    },
+    {
+      "epoch": 0.697841726618705,
+      "grad_norm": 0.6831346750259399,
+      "learning_rate": 0.0001,
+      "loss": 0.3945,
+      "step": 8148
+    },
+    {
+      "epoch": 0.697927372387804,
+      "grad_norm": 0.3937978744506836,
+      "learning_rate": 0.0001,
+      "loss": 0.337,
+      "step": 8149
+    },
+    {
+      "epoch": 0.6980130181569031,
+      "grad_norm": 0.8706310987472534,
+      "learning_rate": 0.0001,
+      "loss": 0.9476,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6980986639260021,
+      "grad_norm": 0.3897257149219513,
+      "learning_rate": 0.0001,
+      "loss": 0.4236,
+      "step": 8151
+    },
+    {
+      "epoch": 0.698184309695101,
+      "grad_norm": 0.32987111806869507,
+      "learning_rate": 0.0001,
+      "loss": 0.677,
+      "step": 8152
+    },
+    {
+      "epoch": 0.6982699554642001,
+      "grad_norm": 0.8687098026275635,
+      "learning_rate": 0.0001,
+      "loss": 1.2162,
+      "step": 8153
+    },
+    {
+      "epoch": 0.6983556012332991,
+      "grad_norm": 0.6727737188339233,
+      "learning_rate": 0.0001,
+      "loss": 0.9301,
+      "step": 8154
+    },
+    {
+      "epoch": 0.6984412470023981,
+      "grad_norm": 1.5037829875946045,
+      "learning_rate": 0.0001,
+      "loss": 1.1028,
+      "step": 8155
+    },
+    {
+      "epoch": 0.698526892771497,
+      "grad_norm": 0.5224624872207642,
+      "learning_rate": 0.0001,
+      "loss": 0.4027,
+      "step": 8156
+    },
+    {
+      "epoch": 0.6986125385405961,
+      "grad_norm": 0.6409405469894409,
+      "learning_rate": 0.0001,
+      "loss": 1.0924,
+      "step": 8157
+    },
+    {
+      "epoch": 0.6986981843096951,
+      "grad_norm": 1.2889357805252075,
+      "learning_rate": 0.0001,
+      "loss": 1.2803,
+      "step": 8158
+    },
+    {
+      "epoch": 0.6987838300787941,
+      "grad_norm": 1.1073633432388306,
+      "learning_rate": 0.0001,
+      "loss": 1.1014,
+      "step": 8159
+    },
+    {
+      "epoch": 0.6988694758478932,
+      "grad_norm": 0.4761982262134552,
+      "learning_rate": 0.0001,
+      "loss": 0.3441,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6989551216169921,
+      "grad_norm": 1.0558353662490845,
+      "learning_rate": 0.0001,
+      "loss": 0.9283,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6990407673860911,
+      "grad_norm": 0.9461495280265808,
+      "learning_rate": 0.0001,
+      "loss": 0.8561,
+      "step": 8162
+    },
+    {
+      "epoch": 0.6991264131551901,
+      "grad_norm": 0.8583139777183533,
+      "learning_rate": 0.0001,
+      "loss": 0.4964,
+      "step": 8163
+    },
+    {
+      "epoch": 0.6992120589242892,
+      "grad_norm": 0.6600015163421631,
+      "learning_rate": 0.0001,
+      "loss": 0.9593,
+      "step": 8164
+    },
+    {
+      "epoch": 0.6992977046933881,
+      "grad_norm": 0.766814112663269,
+      "learning_rate": 0.0001,
+      "loss": 1.303,
+      "step": 8165
+    },
+    {
+      "epoch": 0.6993833504624871,
+      "grad_norm": 0.7587482929229736,
+      "learning_rate": 0.0001,
+      "loss": 1.1276,
+      "step": 8166
+    },
+    {
+      "epoch": 0.6994689962315862,
+      "grad_norm": 1.8744012117385864,
+      "learning_rate": 0.0001,
+      "loss": 2.0981,
+      "step": 8167
+    },
+    {
+      "epoch": 0.6995546420006852,
+      "grad_norm": 1.1164873838424683,
+      "learning_rate": 0.0001,
+      "loss": 0.6332,
+      "step": 8168
+    },
+    {
+      "epoch": 0.6996402877697842,
+      "grad_norm": 0.78631591796875,
+      "learning_rate": 0.0001,
+      "loss": 0.7877,
+      "step": 8169
+    },
+    {
+      "epoch": 0.6997259335388831,
+      "grad_norm": 1.0634671449661255,
+      "learning_rate": 0.0001,
+      "loss": 0.8257,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6998115793079822,
+      "grad_norm": 0.8273914456367493,
+      "learning_rate": 0.0001,
+      "loss": 1.0417,
+      "step": 8171
+    },
+    {
+      "epoch": 0.6998972250770812,
+      "grad_norm": 1.0933618545532227,
+      "learning_rate": 0.0001,
+      "loss": 1.2672,
+      "step": 8172
+    },
+    {
+      "epoch": 0.6999828708461802,
+      "grad_norm": 0.5399419069290161,
+      "learning_rate": 0.0001,
+      "loss": 1.2324,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7000685166152792,
+      "grad_norm": 0.865102231502533,
+      "learning_rate": 0.0001,
+      "loss": 0.7582,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7001541623843782,
+      "grad_norm": 1.3085362911224365,
+      "learning_rate": 0.0001,
+      "loss": 1.1881,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7002398081534772,
+      "grad_norm": 0.7090051770210266,
+      "learning_rate": 0.0001,
+      "loss": 0.8862,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7003254539225763,
+      "grad_norm": 1.1585747003555298,
+      "learning_rate": 0.0001,
+      "loss": 1.1942,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7004110996916753,
+      "grad_norm": 1.0173089504241943,
+      "learning_rate": 0.0001,
+      "loss": 1.2619,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7004967454607742,
+      "grad_norm": 0.6725343465805054,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7005823912298732,
+      "grad_norm": 0.6150527000427246,
+      "learning_rate": 0.0001,
+      "loss": 1.1841,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7006680369989723,
+      "grad_norm": 1.1988683938980103,
+      "learning_rate": 0.0001,
+      "loss": 0.8935,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7007536827680713,
+      "grad_norm": 0.9881510138511658,
+      "learning_rate": 0.0001,
+      "loss": 1.1201,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7008393285371702,
+      "grad_norm": 0.4495965540409088,
+      "learning_rate": 0.0001,
+      "loss": 0.4436,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7009249743062693,
+      "grad_norm": 0.8888470530509949,
+      "learning_rate": 0.0001,
+      "loss": 1.0025,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7010106200753683,
+      "grad_norm": 0.9759475588798523,
+      "learning_rate": 0.0001,
+      "loss": 1.0085,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7010962658444673,
+      "grad_norm": 0.9071581959724426,
+      "learning_rate": 0.0001,
+      "loss": 0.6023,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7011819116135662,
+      "grad_norm": 0.5798145532608032,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 8187
+    },
+    {
+      "epoch": 0.7012675573826653,
+      "grad_norm": 1.2804642915725708,
+      "learning_rate": 0.0001,
+      "loss": 0.798,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7013532031517643,
+      "grad_norm": 0.7263706922531128,
+      "learning_rate": 0.0001,
+      "loss": 1.0061,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7014388489208633,
+      "grad_norm": 0.976760745048523,
+      "learning_rate": 0.0001,
+      "loss": 0.7869,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7015244946899624,
+      "grad_norm": 0.9007558226585388,
+      "learning_rate": 0.0001,
+      "loss": 1.1406,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7016101404590613,
+      "grad_norm": 0.8151209950447083,
+      "learning_rate": 0.0001,
+      "loss": 0.8888,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7016957862281603,
+      "grad_norm": 0.6870827078819275,
+      "learning_rate": 0.0001,
+      "loss": 0.6969,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7017814319972593,
+      "grad_norm": 1.022139072418213,
+      "learning_rate": 0.0001,
+      "loss": 1.1135,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7018670777663584,
+      "grad_norm": 0.9501768946647644,
+      "learning_rate": 0.0001,
+      "loss": 1.0789,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7019527235354573,
+      "grad_norm": 0.7608470916748047,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7020383693045563,
+      "grad_norm": 0.6039759516716003,
+      "learning_rate": 0.0001,
+      "loss": 0.7938,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7021240150736554,
+      "grad_norm": 0.758667528629303,
+      "learning_rate": 0.0001,
+      "loss": 0.6105,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7022096608427544,
+      "grad_norm": 2.3841660022735596,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7022953066118534,
+      "grad_norm": 0.6358631253242493,
+      "learning_rate": 0.0001,
+      "loss": 0.6187,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7023809523809523,
+      "grad_norm": 0.4454493820667267,
+      "learning_rate": 0.0001,
+      "loss": 0.7079,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7024665981500514,
+      "grad_norm": 0.6635354161262512,
+      "learning_rate": 0.0001,
+      "loss": 0.5477,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7025522439191504,
+      "grad_norm": 0.4483833611011505,
+      "learning_rate": 0.0001,
+      "loss": 0.7969,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7026378896882494,
+      "grad_norm": 0.8132144808769226,
+      "learning_rate": 0.0001,
+      "loss": 0.9805,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7027235354573484,
+      "grad_norm": 1.0367504358291626,
+      "learning_rate": 0.0001,
+      "loss": 1.141,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7028091812264474,
+      "grad_norm": 0.6992629766464233,
+      "learning_rate": 0.0001,
+      "loss": 1.2734,
+      "step": 8206
+    },
+    {
+      "epoch": 0.7028948269955464,
+      "grad_norm": 0.7823894619941711,
+      "learning_rate": 0.0001,
+      "loss": 0.5336,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7029804727646455,
+      "grad_norm": 0.9185014367103577,
+      "learning_rate": 0.0001,
+      "loss": 0.8394,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7030661185337445,
+      "grad_norm": 0.5716201663017273,
+      "learning_rate": 0.0001,
+      "loss": 0.939,
+      "step": 8209
+    },
+    {
+      "epoch": 0.7031517643028434,
+      "grad_norm": 1.110278606414795,
+      "learning_rate": 0.0001,
+      "loss": 0.939,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7032374100719424,
+      "grad_norm": 0.3214237093925476,
+      "learning_rate": 0.0001,
+      "loss": 0.5411,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7033230558410415,
+      "grad_norm": 1.216571569442749,
+      "learning_rate": 0.0001,
+      "loss": 1.3339,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7034087016101405,
+      "grad_norm": 1.0580286979675293,
+      "learning_rate": 0.0001,
+      "loss": 1.371,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7034943473792394,
+      "grad_norm": 1.0210195779800415,
+      "learning_rate": 0.0001,
+      "loss": 1.1089,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7035799931483385,
+      "grad_norm": 0.6617929935455322,
+      "learning_rate": 0.0001,
+      "loss": 0.992,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7036656389174375,
+      "grad_norm": 1.0080877542495728,
+      "learning_rate": 0.0001,
+      "loss": 1.1848,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7037512846865365,
+      "grad_norm": 0.6263553500175476,
+      "learning_rate": 0.0001,
+      "loss": 0.7507,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7038369304556354,
+      "grad_norm": 0.6124815940856934,
+      "learning_rate": 0.0001,
+      "loss": 0.6134,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7039225762247345,
+      "grad_norm": 0.8333303928375244,
+      "learning_rate": 0.0001,
+      "loss": 1.0165,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7040082219938335,
+      "grad_norm": 1.1388883590698242,
+      "learning_rate": 0.0001,
+      "loss": 1.249,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7040938677629325,
+      "grad_norm": 0.43450936675071716,
+      "learning_rate": 0.0001,
+      "loss": 0.1402,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7041795135320316,
+      "grad_norm": 0.736934244632721,
+      "learning_rate": 0.0001,
+      "loss": 0.9229,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7042651593011305,
+      "grad_norm": 0.7361476421356201,
+      "learning_rate": 0.0001,
+      "loss": 0.4978,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7043508050702295,
+      "grad_norm": 1.1324750185012817,
+      "learning_rate": 0.0001,
+      "loss": 0.7491,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7044364508393285,
+      "grad_norm": 1.1064505577087402,
+      "learning_rate": 0.0001,
+      "loss": 1.0002,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7045220966084276,
+      "grad_norm": 0.9246273040771484,
+      "learning_rate": 0.0001,
+      "loss": 1.1617,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7046077423775265,
+      "grad_norm": 0.3548925220966339,
+      "learning_rate": 0.0001,
+      "loss": 0.5981,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7046933881466255,
+      "grad_norm": 0.9739626049995422,
+      "learning_rate": 0.0001,
+      "loss": 1.0405,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7047790339157246,
+      "grad_norm": 1.0371860265731812,
+      "learning_rate": 0.0001,
+      "loss": 0.6204,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7048646796848236,
+      "grad_norm": 0.605276346206665,
+      "learning_rate": 0.0001,
+      "loss": 0.9543,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7049503254539226,
+      "grad_norm": 0.8252348303794861,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7050359712230215,
+      "grad_norm": 0.5492702126502991,
+      "learning_rate": 0.0001,
+      "loss": 1.0288,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7051216169921206,
+      "grad_norm": 0.9693685173988342,
+      "learning_rate": 0.0001,
+      "loss": 1.1182,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7052072627612196,
+      "grad_norm": 0.5984851717948914,
+      "learning_rate": 0.0001,
+      "loss": 1.065,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7052929085303186,
+      "grad_norm": 0.7221837043762207,
+      "learning_rate": 0.0001,
+      "loss": 0.6125,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7053785542994176,
+      "grad_norm": 0.9023817181587219,
+      "learning_rate": 0.0001,
+      "loss": 0.774,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7054642000685166,
+      "grad_norm": 1.0007911920547485,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7055498458376156,
+      "grad_norm": 0.6175064444541931,
+      "learning_rate": 0.0001,
+      "loss": 0.87,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7056354916067147,
+      "grad_norm": 1.0158884525299072,
+      "learning_rate": 0.0001,
+      "loss": 0.9868,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7057211373758137,
+      "grad_norm": 1.3997613191604614,
+      "learning_rate": 0.0001,
+      "loss": 1.1732,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7058067831449126,
+      "grad_norm": 0.8979126811027527,
+      "learning_rate": 0.0001,
+      "loss": 1.1479,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7058924289140116,
+      "grad_norm": 0.8543203473091125,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7059780746831107,
+      "grad_norm": 0.8383392095565796,
+      "learning_rate": 0.0001,
+      "loss": 1.1716,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7060637204522097,
+      "grad_norm": 1.082988977432251,
+      "learning_rate": 0.0001,
+      "loss": 1.0417,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7061493662213086,
+      "grad_norm": 0.5691854357719421,
+      "learning_rate": 0.0001,
+      "loss": 0.862,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7062350119904077,
+      "grad_norm": 0.8592280745506287,
+      "learning_rate": 0.0001,
+      "loss": 0.8116,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7063206577595067,
+      "grad_norm": 0.7740523219108582,
+      "learning_rate": 0.0001,
+      "loss": 0.916,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7064063035286057,
+      "grad_norm": 0.8398024439811707,
+      "learning_rate": 0.0001,
+      "loss": 0.823,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7064919492977046,
+      "grad_norm": 0.8193003535270691,
+      "learning_rate": 0.0001,
+      "loss": 1.1212,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7065775950668037,
+      "grad_norm": 0.3142823874950409,
+      "learning_rate": 0.0001,
+      "loss": 0.6457,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7066632408359027,
+      "grad_norm": 0.8758176565170288,
+      "learning_rate": 0.0001,
+      "loss": 1.2729,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7067488866050017,
+      "grad_norm": 1.3178203105926514,
+      "learning_rate": 0.0001,
+      "loss": 0.8161,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7068345323741008,
+      "grad_norm": 0.7015073895454407,
+      "learning_rate": 0.0001,
+      "loss": 0.8584,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7069201781431997,
+      "grad_norm": 0.9654026627540588,
+      "learning_rate": 0.0001,
+      "loss": 1.0486,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7070058239122987,
+      "grad_norm": 0.8344144225120544,
+      "learning_rate": 0.0001,
+      "loss": 1.0594,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7070914696813977,
+      "grad_norm": 1.2234480381011963,
+      "learning_rate": 0.0001,
+      "loss": 1.3041,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7071771154504968,
+      "grad_norm": 0.9688882231712341,
+      "learning_rate": 0.0001,
+      "loss": 1.2589,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7072627612195957,
+      "grad_norm": 0.8506436944007874,
+      "learning_rate": 0.0001,
+      "loss": 0.4962,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7073484069886947,
+      "grad_norm": 0.3935694098472595,
+      "learning_rate": 0.0001,
+      "loss": 0.1054,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7074340527577938,
+      "grad_norm": 0.7904990911483765,
+      "learning_rate": 0.0001,
+      "loss": 1.2655,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7075196985268928,
+      "grad_norm": 0.8935215473175049,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7076053442959918,
+      "grad_norm": 0.6091339588165283,
+      "learning_rate": 0.0001,
+      "loss": 0.4784,
+      "step": 8262
+    },
+    {
+      "epoch": 0.7076909900650907,
+      "grad_norm": 0.7253555655479431,
+      "learning_rate": 0.0001,
+      "loss": 0.9192,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7077766358341898,
+      "grad_norm": 0.606824517250061,
+      "learning_rate": 0.0001,
+      "loss": 0.8726,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7078622816032888,
+      "grad_norm": 0.55972820520401,
+      "learning_rate": 0.0001,
+      "loss": 0.559,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7079479273723878,
+      "grad_norm": 0.5523128509521484,
+      "learning_rate": 0.0001,
+      "loss": 1.0094,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7080335731414868,
+      "grad_norm": 0.8288094401359558,
+      "learning_rate": 0.0001,
+      "loss": 1.1862,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7081192189105858,
+      "grad_norm": 0.4320540726184845,
+      "learning_rate": 0.0001,
+      "loss": 0.1528,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7082048646796848,
+      "grad_norm": 0.6609402894973755,
+      "learning_rate": 0.0001,
+      "loss": 0.7718,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7082905104487839,
+      "grad_norm": 1.0696948766708374,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7083761562178829,
+      "grad_norm": 0.9726970195770264,
+      "learning_rate": 0.0001,
+      "loss": 0.9596,
+      "step": 8271
+    },
+    {
+      "epoch": 0.7084618019869818,
+      "grad_norm": 0.6593026518821716,
+      "learning_rate": 0.0001,
+      "loss": 1.0785,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7085474477560808,
+      "grad_norm": 1.5641887187957764,
+      "learning_rate": 0.0001,
+      "loss": 2.1328,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7086330935251799,
+      "grad_norm": 0.6272464394569397,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7087187392942789,
+      "grad_norm": 0.8310005068778992,
+      "learning_rate": 0.0001,
+      "loss": 0.9924,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7088043850633778,
+      "grad_norm": 0.8221049308776855,
+      "learning_rate": 0.0001,
+      "loss": 0.1846,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7088900308324769,
+      "grad_norm": 1.0560626983642578,
+      "learning_rate": 0.0001,
+      "loss": 0.6608,
+      "step": 8277
+    },
+    {
+      "epoch": 0.7089756766015759,
+      "grad_norm": 0.7895159721374512,
+      "learning_rate": 0.0001,
+      "loss": 0.8589,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7090613223706749,
+      "grad_norm": 0.904167652130127,
+      "learning_rate": 0.0001,
+      "loss": 1.0094,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7091469681397738,
+      "grad_norm": 0.6526954174041748,
+      "learning_rate": 0.0001,
+      "loss": 1.3005,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7092326139088729,
+      "grad_norm": 0.8164382576942444,
+      "learning_rate": 0.0001,
+      "loss": 0.7658,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7093182596779719,
+      "grad_norm": 0.6605362296104431,
+      "learning_rate": 0.0001,
+      "loss": 1.3673,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7094039054470709,
+      "grad_norm": 0.8769996166229248,
+      "learning_rate": 0.0001,
+      "loss": 1.0015,
+      "step": 8283
+    },
+    {
+      "epoch": 0.70948955121617,
+      "grad_norm": 1.7654283046722412,
+      "learning_rate": 0.0001,
+      "loss": 1.0507,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7095751969852689,
+      "grad_norm": 0.6711273789405823,
+      "learning_rate": 0.0001,
+      "loss": 1.1451,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7096608427543679,
+      "grad_norm": 0.6300133466720581,
+      "learning_rate": 0.0001,
+      "loss": 1.1983,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7097464885234669,
+      "grad_norm": 0.9404494166374207,
+      "learning_rate": 0.0001,
+      "loss": 1.3491,
+      "step": 8287
+    },
+    {
+      "epoch": 0.709832134292566,
+      "grad_norm": 0.7625806331634521,
+      "learning_rate": 0.0001,
+      "loss": 1.1528,
+      "step": 8288
+    },
+    {
+      "epoch": 0.709917780061665,
+      "grad_norm": 0.4398385286331177,
+      "learning_rate": 0.0001,
+      "loss": 0.4996,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7100034258307639,
+      "grad_norm": 0.7179349660873413,
+      "learning_rate": 0.0001,
+      "loss": 0.8356,
+      "step": 8290
+    },
+    {
+      "epoch": 0.710089071599863,
+      "grad_norm": 0.8868502974510193,
+      "learning_rate": 0.0001,
+      "loss": 0.9638,
+      "step": 8291
+    },
+    {
+      "epoch": 0.710174717368962,
+      "grad_norm": 0.9616005420684814,
+      "learning_rate": 0.0001,
+      "loss": 1.0144,
+      "step": 8292
+    },
+    {
+      "epoch": 0.710260363138061,
+      "grad_norm": 0.9926223158836365,
+      "learning_rate": 0.0001,
+      "loss": 0.9097,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7103460089071599,
+      "grad_norm": 0.8507649898529053,
+      "learning_rate": 0.0001,
+      "loss": 0.8659,
+      "step": 8294
+    },
+    {
+      "epoch": 0.710431654676259,
+      "grad_norm": 0.8417763113975525,
+      "learning_rate": 0.0001,
+      "loss": 0.6844,
+      "step": 8295
+    },
+    {
+      "epoch": 0.710517300445358,
+      "grad_norm": 0.7476323246955872,
+      "learning_rate": 0.0001,
+      "loss": 0.9162,
+      "step": 8296
+    },
+    {
+      "epoch": 0.710602946214457,
+      "grad_norm": 1.3748782873153687,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 8297
+    },
+    {
+      "epoch": 0.710688591983556,
+      "grad_norm": 0.5317225456237793,
+      "learning_rate": 0.0001,
+      "loss": 0.472,
+      "step": 8298
+    },
+    {
+      "epoch": 0.710774237752655,
+      "grad_norm": 0.6363984942436218,
+      "learning_rate": 0.0001,
+      "loss": 0.6003,
+      "step": 8299
+    },
+    {
+      "epoch": 0.710859883521754,
+      "grad_norm": 0.6307455897331238,
+      "learning_rate": 0.0001,
+      "loss": 0.7936,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7109455292908531,
+      "grad_norm": 1.153670310974121,
+      "learning_rate": 0.0001,
+      "loss": 1.0265,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7110311750599521,
+      "grad_norm": 0.7719588875770569,
+      "learning_rate": 0.0001,
+      "loss": 0.6873,
+      "step": 8302
+    },
+    {
+      "epoch": 0.711116820829051,
+      "grad_norm": 0.8246824145317078,
+      "learning_rate": 0.0001,
+      "loss": 1.3603,
+      "step": 8303
+    },
+    {
+      "epoch": 0.71120246659815,
+      "grad_norm": 0.8671765327453613,
+      "learning_rate": 0.0001,
+      "loss": 1.0296,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7112881123672491,
+      "grad_norm": 0.7888370156288147,
+      "learning_rate": 0.0001,
+      "loss": 1.9497,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7113737581363481,
+      "grad_norm": 0.7641797661781311,
+      "learning_rate": 0.0001,
+      "loss": 0.8601,
+      "step": 8306
+    },
+    {
+      "epoch": 0.711459403905447,
+      "grad_norm": 1.1452747583389282,
+      "learning_rate": 0.0001,
+      "loss": 1.1029,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7115450496745461,
+      "grad_norm": 0.9133027791976929,
+      "learning_rate": 0.0001,
+      "loss": 0.8442,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7116306954436451,
+      "grad_norm": 0.5478026270866394,
+      "learning_rate": 0.0001,
+      "loss": 1.1844,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7117163412127441,
+      "grad_norm": 1.027406930923462,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 8310
+    },
+    {
+      "epoch": 0.711801986981843,
+      "grad_norm": 0.5557610392570496,
+      "learning_rate": 0.0001,
+      "loss": 1.0147,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7118876327509421,
+      "grad_norm": 0.7944310307502747,
+      "learning_rate": 0.0001,
+      "loss": 1.1438,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7119732785200411,
+      "grad_norm": 1.3290586471557617,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7120589242891401,
+      "grad_norm": 0.6724480390548706,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7121445700582392,
+      "grad_norm": 1.8072699308395386,
+      "learning_rate": 0.0001,
+      "loss": 1.0041,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7122302158273381,
+      "grad_norm": 0.6654115319252014,
+      "learning_rate": 0.0001,
+      "loss": 0.4192,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7123158615964371,
+      "grad_norm": 0.7904651165008545,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7124015073655361,
+      "grad_norm": 0.7363705039024353,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7124871531346352,
+      "grad_norm": 1.1315892934799194,
+      "learning_rate": 0.0001,
+      "loss": 0.9125,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7125727989037342,
+      "grad_norm": 0.5578249096870422,
+      "learning_rate": 0.0001,
+      "loss": 0.7386,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7126584446728331,
+      "grad_norm": 0.9721454381942749,
+      "learning_rate": 0.0001,
+      "loss": 0.948,
+      "step": 8321
+    },
+    {
+      "epoch": 0.7127440904419322,
+      "grad_norm": 1.1675609350204468,
+      "learning_rate": 0.0001,
+      "loss": 1.1437,
+      "step": 8322
+    },
+    {
+      "epoch": 0.7128297362110312,
+      "grad_norm": 0.734417736530304,
+      "learning_rate": 0.0001,
+      "loss": 0.7655,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7129153819801302,
+      "grad_norm": 0.5855022668838501,
+      "learning_rate": 0.0001,
+      "loss": 0.8696,
+      "step": 8324
+    },
+    {
+      "epoch": 0.7130010277492291,
+      "grad_norm": 0.764681339263916,
+      "learning_rate": 0.0001,
+      "loss": 0.531,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7130866735183282,
+      "grad_norm": 0.5496594905853271,
+      "learning_rate": 0.0001,
+      "loss": 0.5358,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7131723192874272,
+      "grad_norm": 1.319923758506775,
+      "learning_rate": 0.0001,
+      "loss": 0.9824,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7132579650565262,
+      "grad_norm": 0.4631485641002655,
+      "learning_rate": 0.0001,
+      "loss": 0.3339,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7133436108256253,
+      "grad_norm": 0.7166049480438232,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7134292565947242,
+      "grad_norm": 0.8441925048828125,
+      "learning_rate": 0.0001,
+      "loss": 0.9998,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7135149023638232,
+      "grad_norm": 0.7656216025352478,
+      "learning_rate": 0.0001,
+      "loss": 0.9525,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7136005481329223,
+      "grad_norm": 0.8516676425933838,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7136861939020213,
+      "grad_norm": 0.8232327103614807,
+      "learning_rate": 0.0001,
+      "loss": 0.6612,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7137718396711202,
+      "grad_norm": 0.7940201163291931,
+      "learning_rate": 0.0001,
+      "loss": 0.8584,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7138574854402192,
+      "grad_norm": 0.692077100276947,
+      "learning_rate": 0.0001,
+      "loss": 0.6725,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7139431312093183,
+      "grad_norm": 0.6541866660118103,
+      "learning_rate": 0.0001,
+      "loss": 1.0574,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7140287769784173,
+      "grad_norm": 0.8815173506736755,
+      "learning_rate": 0.0001,
+      "loss": 1.1605,
+      "step": 8337
+    },
+    {
+      "epoch": 0.7141144227475162,
+      "grad_norm": 0.7585886120796204,
+      "learning_rate": 0.0001,
+      "loss": 0.8995,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7142000685166153,
+      "grad_norm": 0.9040355682373047,
+      "learning_rate": 0.0001,
+      "loss": 0.9936,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.7881430983543396,
+      "learning_rate": 0.0001,
+      "loss": 1.0565,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7143713600548133,
+      "grad_norm": 0.7520089149475098,
+      "learning_rate": 0.0001,
+      "loss": 0.9183,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7144570058239123,
+      "grad_norm": 0.5986270308494568,
+      "learning_rate": 0.0001,
+      "loss": 0.8278,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7145426515930113,
+      "grad_norm": 0.5622308850288391,
+      "learning_rate": 0.0001,
+      "loss": 0.8967,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7146282973621103,
+      "grad_norm": 0.6901191473007202,
+      "learning_rate": 0.0001,
+      "loss": 1.1208,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7147139431312093,
+      "grad_norm": 1.15199875831604,
+      "learning_rate": 0.0001,
+      "loss": 1.1756,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7147995889003084,
+      "grad_norm": 0.8052424788475037,
+      "learning_rate": 0.0001,
+      "loss": 0.6241,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7148852346694073,
+      "grad_norm": 0.6956855654716492,
+      "learning_rate": 0.0001,
+      "loss": 1.074,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7149708804385063,
+      "grad_norm": 0.7384576797485352,
+      "learning_rate": 0.0001,
+      "loss": 0.7276,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7150565262076053,
+      "grad_norm": 1.1197199821472168,
+      "learning_rate": 0.0001,
+      "loss": 0.7756,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7151421719767044,
+      "grad_norm": 0.5488266944885254,
+      "learning_rate": 0.0001,
+      "loss": 1.0452,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7152278177458034,
+      "grad_norm": 1.8919092416763306,
+      "learning_rate": 0.0001,
+      "loss": 1.0155,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7153134635149023,
+      "grad_norm": 0.8849449157714844,
+      "learning_rate": 0.0001,
+      "loss": 0.8265,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7153991092840014,
+      "grad_norm": 0.4820089042186737,
+      "learning_rate": 0.0001,
+      "loss": 0.604,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7154847550531004,
+      "grad_norm": 0.771303117275238,
+      "learning_rate": 0.0001,
+      "loss": 1.0299,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7155704008221994,
+      "grad_norm": 0.7400087714195251,
+      "learning_rate": 0.0001,
+      "loss": 0.8409,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7156560465912984,
+      "grad_norm": 0.8022162914276123,
+      "learning_rate": 0.0001,
+      "loss": 1.0124,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7157416923603974,
+      "grad_norm": 0.7750237584114075,
+      "learning_rate": 0.0001,
+      "loss": 0.8125,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7158273381294964,
+      "grad_norm": 0.7955005764961243,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7159129838985954,
+      "grad_norm": 0.7366554141044617,
+      "learning_rate": 0.0001,
+      "loss": 1.0682,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7159986296676945,
+      "grad_norm": 1.6223981380462646,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7160842754367934,
+      "grad_norm": 1.479447364807129,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7161699212058924,
+      "grad_norm": 1.061464786529541,
+      "learning_rate": 0.0001,
+      "loss": 0.7621,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7162555669749915,
+      "grad_norm": 1.0522844791412354,
+      "learning_rate": 0.0001,
+      "loss": 0.8882,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7163412127440905,
+      "grad_norm": 0.5072388648986816,
+      "learning_rate": 0.0001,
+      "loss": 0.5081,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7164268585131894,
+      "grad_norm": 1.1244897842407227,
+      "learning_rate": 0.0001,
+      "loss": 1.0938,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7165125042822884,
+      "grad_norm": 0.8408418893814087,
+      "learning_rate": 0.0001,
+      "loss": 0.9132,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7165981500513875,
+      "grad_norm": 0.6791041493415833,
+      "learning_rate": 0.0001,
+      "loss": 1.3212,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7166837958204865,
+      "grad_norm": 0.6176796555519104,
+      "learning_rate": 0.0001,
+      "loss": 0.8291,
+      "step": 8368
+    },
+    {
+      "epoch": 0.7167694415895854,
+      "grad_norm": 0.7076791524887085,
+      "learning_rate": 0.0001,
+      "loss": 1.1943,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7168550873586845,
+      "grad_norm": 0.5616340041160583,
+      "learning_rate": 0.0001,
+      "loss": 0.5402,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7169407331277835,
+      "grad_norm": 0.8429402709007263,
+      "learning_rate": 0.0001,
+      "loss": 0.5769,
+      "step": 8371
+    },
+    {
+      "epoch": 0.7170263788968825,
+      "grad_norm": 0.5943896174430847,
+      "learning_rate": 0.0001,
+      "loss": 1.2749,
+      "step": 8372
+    },
+    {
+      "epoch": 0.7171120246659815,
+      "grad_norm": 0.6636331677436829,
+      "learning_rate": 0.0001,
+      "loss": 1.0527,
+      "step": 8373
+    },
+    {
+      "epoch": 0.7171976704350805,
+      "grad_norm": 1.3676162958145142,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 8374
+    },
+    {
+      "epoch": 0.7172833162041795,
+      "grad_norm": 0.5466113090515137,
+      "learning_rate": 0.0001,
+      "loss": 0.712,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7173689619732785,
+      "grad_norm": 1.1390727758407593,
+      "learning_rate": 0.0001,
+      "loss": 1.3633,
+      "step": 8376
+    },
+    {
+      "epoch": 0.7174546077423776,
+      "grad_norm": 0.7228270173072815,
+      "learning_rate": 0.0001,
+      "loss": 0.8589,
+      "step": 8377
+    },
+    {
+      "epoch": 0.7175402535114765,
+      "grad_norm": 1.0208450555801392,
+      "learning_rate": 0.0001,
+      "loss": 0.8867,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7176258992805755,
+      "grad_norm": 0.9643282890319824,
+      "learning_rate": 0.0001,
+      "loss": 0.9136,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7177115450496745,
+      "grad_norm": 0.32769688963890076,
+      "learning_rate": 0.0001,
+      "loss": 0.625,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7177971908187736,
+      "grad_norm": 0.5806612372398376,
+      "learning_rate": 0.0001,
+      "loss": 0.9308,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7178828365878726,
+      "grad_norm": 0.635493278503418,
+      "learning_rate": 0.0001,
+      "loss": 0.9471,
+      "step": 8382
+    },
+    {
+      "epoch": 0.7179684823569715,
+      "grad_norm": 0.7557166218757629,
+      "learning_rate": 0.0001,
+      "loss": 0.8535,
+      "step": 8383
+    },
+    {
+      "epoch": 0.7180541281260706,
+      "grad_norm": 0.9226289391517639,
+      "learning_rate": 0.0001,
+      "loss": 0.976,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7181397738951696,
+      "grad_norm": 0.3197271227836609,
+      "learning_rate": 0.0001,
+      "loss": 0.508,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7182254196642686,
+      "grad_norm": 0.8626756072044373,
+      "learning_rate": 0.0001,
+      "loss": 0.9488,
+      "step": 8386
+    },
+    {
+      "epoch": 0.7183110654333676,
+      "grad_norm": 0.5527372360229492,
+      "learning_rate": 0.0001,
+      "loss": 1.0118,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7183967112024666,
+      "grad_norm": 0.5014415383338928,
+      "learning_rate": 0.0001,
+      "loss": 1.0296,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7184823569715656,
+      "grad_norm": 0.6377484202384949,
+      "learning_rate": 0.0001,
+      "loss": 1.3587,
+      "step": 8389
+    },
+    {
+      "epoch": 0.7185680027406646,
+      "grad_norm": 0.7359502911567688,
+      "learning_rate": 0.0001,
+      "loss": 1.142,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7186536485097637,
+      "grad_norm": 0.612975537776947,
+      "learning_rate": 0.0001,
+      "loss": 1.1152,
+      "step": 8391
+    },
+    {
+      "epoch": 0.7187392942788626,
+      "grad_norm": 1.3551099300384521,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 8392
+    },
+    {
+      "epoch": 0.7188249400479616,
+      "grad_norm": 0.4138217866420746,
+      "learning_rate": 0.0001,
+      "loss": 0.156,
+      "step": 8393
+    },
+    {
+      "epoch": 0.7189105858170607,
+      "grad_norm": 0.41704261302948,
+      "learning_rate": 0.0001,
+      "loss": 0.482,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7189962315861597,
+      "grad_norm": 0.7013681530952454,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7190818773552586,
+      "grad_norm": 0.6837418079376221,
+      "learning_rate": 0.0001,
+      "loss": 0.6348,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7191675231243576,
+      "grad_norm": 0.9113123416900635,
+      "learning_rate": 0.0001,
+      "loss": 0.8579,
+      "step": 8397
+    },
+    {
+      "epoch": 0.7192531688934567,
+      "grad_norm": 0.9259822368621826,
+      "learning_rate": 0.0001,
+      "loss": 0.5025,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7193388146625557,
+      "grad_norm": 1.1430116891860962,
+      "learning_rate": 0.0001,
+      "loss": 1.1984,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 0.3566345274448395,
+      "learning_rate": 0.0001,
+      "loss": 0.188,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7195101062007537,
+      "grad_norm": 1.0896841287612915,
+      "learning_rate": 0.0001,
+      "loss": 1.1634,
+      "step": 8401
+    },
+    {
+      "epoch": 0.7195957519698527,
+      "grad_norm": 1.1116070747375488,
+      "learning_rate": 0.0001,
+      "loss": 0.9282,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7196813977389517,
+      "grad_norm": 0.8076801896095276,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7197670435080507,
+      "grad_norm": 1.192254900932312,
+      "learning_rate": 0.0001,
+      "loss": 1.0248,
+      "step": 8404
+    },
+    {
+      "epoch": 0.7198526892771497,
+      "grad_norm": 0.978227972984314,
+      "learning_rate": 0.0001,
+      "loss": 1.1563,
+      "step": 8405
+    },
+    {
+      "epoch": 0.7199383350462487,
+      "grad_norm": 0.6698260307312012,
+      "learning_rate": 0.0001,
+      "loss": 1.1501,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7200239808153477,
+      "grad_norm": 0.7969732880592346,
+      "learning_rate": 0.0001,
+      "loss": 0.8506,
+      "step": 8407
+    },
+    {
+      "epoch": 0.7201096265844468,
+      "grad_norm": 0.9596307277679443,
+      "learning_rate": 0.0001,
+      "loss": 1.2402,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7201952723535457,
+      "grad_norm": 0.5285815596580505,
+      "learning_rate": 0.0001,
+      "loss": 0.8609,
+      "step": 8409
+    },
+    {
+      "epoch": 0.7202809181226447,
+      "grad_norm": 0.9273360371589661,
+      "learning_rate": 0.0001,
+      "loss": 0.855,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7203665638917437,
+      "grad_norm": 0.7632514834403992,
+      "learning_rate": 0.0001,
+      "loss": 1.1832,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7204522096608428,
+      "grad_norm": 0.4373334050178528,
+      "learning_rate": 0.0001,
+      "loss": 0.3769,
+      "step": 8412
+    },
+    {
+      "epoch": 0.7205378554299418,
+      "grad_norm": 0.8069186806678772,
+      "learning_rate": 0.0001,
+      "loss": 0.7241,
+      "step": 8413
+    },
+    {
+      "epoch": 0.7206235011990407,
+      "grad_norm": 0.6654577255249023,
+      "learning_rate": 0.0001,
+      "loss": 1.0068,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7207091469681398,
+      "grad_norm": 0.6990348696708679,
+      "learning_rate": 0.0001,
+      "loss": 0.6564,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7207947927372388,
+      "grad_norm": 1.3754725456237793,
+      "learning_rate": 0.0001,
+      "loss": 1.11,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7208804385063378,
+      "grad_norm": 1.1826971769332886,
+      "learning_rate": 0.0001,
+      "loss": 1.0687,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7209660842754368,
+      "grad_norm": 0.7558387517929077,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7210517300445358,
+      "grad_norm": 0.807471752166748,
+      "learning_rate": 0.0001,
+      "loss": 0.7846,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7211373758136348,
+      "grad_norm": 0.7242648005485535,
+      "learning_rate": 0.0001,
+      "loss": 1.3507,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7212230215827338,
+      "grad_norm": 1.0088399648666382,
+      "learning_rate": 0.0001,
+      "loss": 0.8127,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7213086673518329,
+      "grad_norm": 0.9205088019371033,
+      "learning_rate": 0.0001,
+      "loss": 0.7884,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7213943131209318,
+      "grad_norm": 0.7358270287513733,
+      "learning_rate": 0.0001,
+      "loss": 0.9646,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7214799588900308,
+      "grad_norm": 1.3821401596069336,
+      "learning_rate": 0.0001,
+      "loss": 1.1798,
+      "step": 8424
+    },
+    {
+      "epoch": 0.7215656046591299,
+      "grad_norm": 0.5742711424827576,
+      "learning_rate": 0.0001,
+      "loss": 0.7004,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7216512504282289,
+      "grad_norm": 1.394580602645874,
+      "learning_rate": 0.0001,
+      "loss": 1.0356,
+      "step": 8426
+    },
+    {
+      "epoch": 0.7217368961973278,
+      "grad_norm": 1.199459195137024,
+      "learning_rate": 0.0001,
+      "loss": 1.1784,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7218225419664268,
+      "grad_norm": 1.3843107223510742,
+      "learning_rate": 0.0001,
+      "loss": 0.8924,
+      "step": 8428
+    },
+    {
+      "epoch": 0.7219081877355259,
+      "grad_norm": 1.1341389417648315,
+      "learning_rate": 0.0001,
+      "loss": 1.2888,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7219938335046249,
+      "grad_norm": 0.8250232934951782,
+      "learning_rate": 0.0001,
+      "loss": 1.028,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7220794792737238,
+      "grad_norm": 0.7950568199157715,
+      "learning_rate": 0.0001,
+      "loss": 1.128,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7221651250428229,
+      "grad_norm": 0.9514407515525818,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7222507708119219,
+      "grad_norm": 0.551051676273346,
+      "learning_rate": 0.0001,
+      "loss": 0.9475,
+      "step": 8433
+    },
+    {
+      "epoch": 0.7223364165810209,
+      "grad_norm": 0.7711202502250671,
+      "learning_rate": 0.0001,
+      "loss": 0.8228,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7224220623501199,
+      "grad_norm": 0.9384329915046692,
+      "learning_rate": 0.0001,
+      "loss": 1.1988,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7225077081192189,
+      "grad_norm": 0.5082053542137146,
+      "learning_rate": 0.0001,
+      "loss": 1.083,
+      "step": 8436
+    },
+    {
+      "epoch": 0.7225933538883179,
+      "grad_norm": 0.5915307402610779,
+      "learning_rate": 0.0001,
+      "loss": 1.0355,
+      "step": 8437
+    },
+    {
+      "epoch": 0.7226789996574169,
+      "grad_norm": 0.6343019604682922,
+      "learning_rate": 0.0001,
+      "loss": 0.8948,
+      "step": 8438
+    },
+    {
+      "epoch": 0.722764645426516,
+      "grad_norm": 1.1287202835083008,
+      "learning_rate": 0.0001,
+      "loss": 0.6732,
+      "step": 8439
+    },
+    {
+      "epoch": 0.722850291195615,
+      "grad_norm": 0.9396056532859802,
+      "learning_rate": 0.0001,
+      "loss": 1.0766,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7229359369647139,
+      "grad_norm": 0.9359802603721619,
+      "learning_rate": 0.0001,
+      "loss": 0.7435,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7230215827338129,
+      "grad_norm": 1.0730243921279907,
+      "learning_rate": 0.0001,
+      "loss": 1.0068,
+      "step": 8442
+    },
+    {
+      "epoch": 0.723107228502912,
+      "grad_norm": 0.7871019244194031,
+      "learning_rate": 0.0001,
+      "loss": 1.1825,
+      "step": 8443
+    },
+    {
+      "epoch": 0.723192874272011,
+      "grad_norm": 0.5477461218833923,
+      "learning_rate": 0.0001,
+      "loss": 0.4909,
+      "step": 8444
+    },
+    {
+      "epoch": 0.7232785200411099,
+      "grad_norm": 0.5399050116539001,
+      "learning_rate": 0.0001,
+      "loss": 0.9059,
+      "step": 8445
+    },
+    {
+      "epoch": 0.723364165810209,
+      "grad_norm": 0.8783190846443176,
+      "learning_rate": 0.0001,
+      "loss": 1.1898,
+      "step": 8446
+    },
+    {
+      "epoch": 0.723449811579308,
+      "grad_norm": 0.8963746428489685,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 8447
+    },
+    {
+      "epoch": 0.723535457348407,
+      "grad_norm": 0.8666284680366516,
+      "learning_rate": 0.0001,
+      "loss": 1.0677,
+      "step": 8448
+    },
+    {
+      "epoch": 0.723621103117506,
+      "grad_norm": 0.7694497108459473,
+      "learning_rate": 0.0001,
+      "loss": 0.6294,
+      "step": 8449
+    },
+    {
+      "epoch": 0.723706748886605,
+      "grad_norm": 0.8744204640388489,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 8450
+    },
+    {
+      "epoch": 0.723792394655704,
+      "grad_norm": 0.9377502799034119,
+      "learning_rate": 0.0001,
+      "loss": 0.7986,
+      "step": 8451
+    },
+    {
+      "epoch": 0.723878040424803,
+      "grad_norm": 1.0650368928909302,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7239636861939021,
+      "grad_norm": 0.853525698184967,
+      "learning_rate": 0.0001,
+      "loss": 0.6814,
+      "step": 8453
+    },
+    {
+      "epoch": 0.724049331963001,
+      "grad_norm": 0.6146976947784424,
+      "learning_rate": 0.0001,
+      "loss": 0.413,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7241349777321,
+      "grad_norm": 0.5875634551048279,
+      "learning_rate": 0.0001,
+      "loss": 0.6979,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7242206235011991,
+      "grad_norm": 1.1355270147323608,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7243062692702981,
+      "grad_norm": 0.695375919342041,
+      "learning_rate": 0.0001,
+      "loss": 1.1092,
+      "step": 8457
+    },
+    {
+      "epoch": 0.724391915039397,
+      "grad_norm": 0.8084471821784973,
+      "learning_rate": 0.0001,
+      "loss": 0.7726,
+      "step": 8458
+    },
+    {
+      "epoch": 0.724477560808496,
+      "grad_norm": 1.1442811489105225,
+      "learning_rate": 0.0001,
+      "loss": 0.9738,
+      "step": 8459
+    },
+    {
+      "epoch": 0.7245632065775951,
+      "grad_norm": 0.9940797090530396,
+      "learning_rate": 0.0001,
+      "loss": 1.171,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7246488523466941,
+      "grad_norm": 0.8730963468551636,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 8461
+    },
+    {
+      "epoch": 0.724734498115793,
+      "grad_norm": 0.7137035727500916,
+      "learning_rate": 0.0001,
+      "loss": 1.1578,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7248201438848921,
+      "grad_norm": 0.7843567728996277,
+      "learning_rate": 0.0001,
+      "loss": 0.8008,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7249057896539911,
+      "grad_norm": 0.9889667630195618,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7249914354230901,
+      "grad_norm": 0.5948835611343384,
+      "learning_rate": 0.0001,
+      "loss": 0.42,
+      "step": 8465
+    },
+    {
+      "epoch": 0.725077081192189,
+      "grad_norm": 0.5516878366470337,
+      "learning_rate": 0.0001,
+      "loss": 0.5817,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7251627269612881,
+      "grad_norm": 0.8420388102531433,
+      "learning_rate": 0.0001,
+      "loss": 0.9538,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7252483727303871,
+      "grad_norm": 0.8330066800117493,
+      "learning_rate": 0.0001,
+      "loss": 0.4708,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7253340184994861,
+      "grad_norm": 0.6165844798088074,
+      "learning_rate": 0.0001,
+      "loss": 0.375,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7254196642685852,
+      "grad_norm": 0.7951841950416565,
+      "learning_rate": 0.0001,
+      "loss": 1.2166,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7255053100376841,
+      "grad_norm": 0.6353521347045898,
+      "learning_rate": 0.0001,
+      "loss": 0.8658,
+      "step": 8471
+    },
+    {
+      "epoch": 0.7255909558067831,
+      "grad_norm": 0.5282878279685974,
+      "learning_rate": 0.0001,
+      "loss": 0.1388,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7256766015758821,
+      "grad_norm": 1.3240272998809814,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 8473
+    },
+    {
+      "epoch": 0.7257622473449812,
+      "grad_norm": 0.7801118493080139,
+      "learning_rate": 0.0001,
+      "loss": 0.4505,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7258478931140802,
+      "grad_norm": 0.7269506454467773,
+      "learning_rate": 0.0001,
+      "loss": 0.2585,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7259335388831791,
+      "grad_norm": 0.4374140202999115,
+      "learning_rate": 0.0001,
+      "loss": 0.3417,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7260191846522782,
+      "grad_norm": 0.9412896037101746,
+      "learning_rate": 0.0001,
+      "loss": 1.1407,
+      "step": 8477
+    },
+    {
+      "epoch": 0.7261048304213772,
+      "grad_norm": 0.864763617515564,
+      "learning_rate": 0.0001,
+      "loss": 0.7969,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7261904761904762,
+      "grad_norm": 0.6886641979217529,
+      "learning_rate": 0.0001,
+      "loss": 0.9597,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7262761219595752,
+      "grad_norm": 0.7054807543754578,
+      "learning_rate": 0.0001,
+      "loss": 1.1133,
+      "step": 8480
+    },
+    {
+      "epoch": 0.7263617677286742,
+      "grad_norm": 0.7144836187362671,
+      "learning_rate": 0.0001,
+      "loss": 0.7349,
+      "step": 8481
+    },
+    {
+      "epoch": 0.7264474134977732,
+      "grad_norm": 0.8287653923034668,
+      "learning_rate": 0.0001,
+      "loss": 1.1158,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7265330592668722,
+      "grad_norm": 0.7848095893859863,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 8483
+    },
+    {
+      "epoch": 0.7266187050359713,
+      "grad_norm": 1.2223725318908691,
+      "learning_rate": 0.0001,
+      "loss": 1.1074,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7267043508050702,
+      "grad_norm": 0.9615859985351562,
+      "learning_rate": 0.0001,
+      "loss": 1.4348,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7267899965741692,
+      "grad_norm": 0.4114951193332672,
+      "learning_rate": 0.0001,
+      "loss": 0.2683,
+      "step": 8486
+    },
+    {
+      "epoch": 0.7268756423432683,
+      "grad_norm": 0.8919564485549927,
+      "learning_rate": 0.0001,
+      "loss": 1.0332,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7269612881123673,
+      "grad_norm": 1.0243709087371826,
+      "learning_rate": 0.0001,
+      "loss": 1.3641,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7270469338814662,
+      "grad_norm": 0.6501665711402893,
+      "learning_rate": 0.0001,
+      "loss": 1.2535,
+      "step": 8489
+    },
+    {
+      "epoch": 0.7271325796505652,
+      "grad_norm": 0.7502941489219666,
+      "learning_rate": 0.0001,
+      "loss": 1.1403,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7272182254196643,
+      "grad_norm": 0.6610739231109619,
+      "learning_rate": 0.0001,
+      "loss": 1.0913,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7273038711887633,
+      "grad_norm": 0.9571748971939087,
+      "learning_rate": 0.0001,
+      "loss": 1.3431,
+      "step": 8492
+    },
+    {
+      "epoch": 0.7273895169578622,
+      "grad_norm": 0.7964262962341309,
+      "learning_rate": 0.0001,
+      "loss": 1.0218,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7274751627269613,
+      "grad_norm": 1.0167871713638306,
+      "learning_rate": 0.0001,
+      "loss": 1.0531,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7275608084960603,
+      "grad_norm": 0.790817141532898,
+      "learning_rate": 0.0001,
+      "loss": 1.0074,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7276464542651593,
+      "grad_norm": 0.6810256838798523,
+      "learning_rate": 0.0001,
+      "loss": 1.1173,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7277321000342583,
+      "grad_norm": 1.1931861639022827,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7278177458033573,
+      "grad_norm": 0.5930211544036865,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7279033915724563,
+      "grad_norm": 1.0836644172668457,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7279890373415553,
+      "grad_norm": 0.8992727994918823,
+      "learning_rate": 0.0001,
+      "loss": 0.9642,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7280746831106544,
+      "grad_norm": 1.0260931253433228,
+      "learning_rate": 0.0001,
+      "loss": 2.0118,
+      "step": 8501
+    },
+    {
+      "epoch": 0.7281603288797533,
+      "grad_norm": 0.9495205283164978,
+      "learning_rate": 0.0001,
+      "loss": 1.1824,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7282459746488523,
+      "grad_norm": 0.9924813508987427,
+      "learning_rate": 0.0001,
+      "loss": 0.5413,
+      "step": 8503
+    },
+    {
+      "epoch": 0.7283316204179514,
+      "grad_norm": 0.7242381572723389,
+      "learning_rate": 0.0001,
+      "loss": 0.8525,
+      "step": 8504
+    },
+    {
+      "epoch": 0.7284172661870504,
+      "grad_norm": 0.5985684394836426,
+      "learning_rate": 0.0001,
+      "loss": 0.5467,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7285029119561494,
+      "grad_norm": 0.32107284665107727,
+      "learning_rate": 0.0001,
+      "loss": 0.3522,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7285885577252483,
+      "grad_norm": 0.9281655550003052,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7286742034943474,
+      "grad_norm": 0.7659258842468262,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7287598492634464,
+      "grad_norm": 1.1383925676345825,
+      "learning_rate": 0.0001,
+      "loss": 0.8686,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7288454950325454,
+      "grad_norm": 0.37362542748451233,
+      "learning_rate": 0.0001,
+      "loss": 0.4539,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7289311408016445,
+      "grad_norm": 1.0538822412490845,
+      "learning_rate": 0.0001,
+      "loss": 1.3399,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7290167865707434,
+      "grad_norm": 0.6074616312980652,
+      "learning_rate": 0.0001,
+      "loss": 0.4527,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7291024323398424,
+      "grad_norm": 1.01919686794281,
+      "learning_rate": 0.0001,
+      "loss": 0.4058,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7291880781089414,
+      "grad_norm": 1.19564950466156,
+      "learning_rate": 0.0001,
+      "loss": 2.0893,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7292737238780405,
+      "grad_norm": 0.6074126958847046,
+      "learning_rate": 0.0001,
+      "loss": 1.0499,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7293593696471394,
+      "grad_norm": 0.9641960859298706,
+      "learning_rate": 0.0001,
+      "loss": 1.2727,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7294450154162384,
+      "grad_norm": 1.1049820184707642,
+      "learning_rate": 0.0001,
+      "loss": 1.1417,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7295306611853375,
+      "grad_norm": 0.8748407959938049,
+      "learning_rate": 0.0001,
+      "loss": 0.5001,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7296163069544365,
+      "grad_norm": 0.5391489267349243,
+      "learning_rate": 0.0001,
+      "loss": 0.709,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7297019527235354,
+      "grad_norm": 0.6369848847389221,
+      "learning_rate": 0.0001,
+      "loss": 0.3641,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7297875984926344,
+      "grad_norm": 0.7380988001823425,
+      "learning_rate": 0.0001,
+      "loss": 0.8262,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7298732442617335,
+      "grad_norm": 0.8781270384788513,
+      "learning_rate": 0.0001,
+      "loss": 1.2724,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7299588900308325,
+      "grad_norm": 0.9479106664657593,
+      "learning_rate": 0.0001,
+      "loss": 0.33,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7300445357999314,
+      "grad_norm": 1.1288049221038818,
+      "learning_rate": 0.0001,
+      "loss": 1.3473,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7301301815690305,
+      "grad_norm": 1.3787415027618408,
+      "learning_rate": 0.0001,
+      "loss": 1.2261,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7302158273381295,
+      "grad_norm": 0.870119035243988,
+      "learning_rate": 0.0001,
+      "loss": 0.6838,
+      "step": 8526
+    },
+    {
+      "epoch": 0.7303014731072285,
+      "grad_norm": 1.23395836353302,
+      "learning_rate": 0.0001,
+      "loss": 1.8929,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7303871188763275,
+      "grad_norm": 0.6885380148887634,
+      "learning_rate": 0.0001,
+      "loss": 0.9712,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7304727646454265,
+      "grad_norm": 0.6197518706321716,
+      "learning_rate": 0.0001,
+      "loss": 0.8708,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7305584104145255,
+      "grad_norm": 1.1269514560699463,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7306440561836245,
+      "grad_norm": 0.8016595244407654,
+      "learning_rate": 0.0001,
+      "loss": 0.5237,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7307297019527236,
+      "grad_norm": 0.353957861661911,
+      "learning_rate": 0.0001,
+      "loss": 0.4087,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7308153477218226,
+      "grad_norm": 0.5205323100090027,
+      "learning_rate": 0.0001,
+      "loss": 0.5083,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7309009934909215,
+      "grad_norm": 0.37026718258857727,
+      "learning_rate": 0.0001,
+      "loss": 0.4432,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7309866392600206,
+      "grad_norm": 1.0738346576690674,
+      "learning_rate": 0.0001,
+      "loss": 0.7122,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7310722850291196,
+      "grad_norm": 1.063364028930664,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7311579307982186,
+      "grad_norm": 1.084298014640808,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7312435765673175,
+      "grad_norm": 1.156079888343811,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7313292223364166,
+      "grad_norm": 0.986843466758728,
+      "learning_rate": 0.0001,
+      "loss": 0.8818,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7314148681055156,
+      "grad_norm": 0.4758853018283844,
+      "learning_rate": 0.0001,
+      "loss": 0.8891,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7315005138746146,
+      "grad_norm": 0.7932077646255493,
+      "learning_rate": 0.0001,
+      "loss": 1.0388,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7315861596437137,
+      "grad_norm": 0.7586220502853394,
+      "learning_rate": 0.0001,
+      "loss": 1.2012,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7316718054128126,
+      "grad_norm": 1.7791955471038818,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7317574511819116,
+      "grad_norm": 1.1170423030853271,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7318430969510106,
+      "grad_norm": 0.6729744672775269,
+      "learning_rate": 0.0001,
+      "loss": 1.1224,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7319287427201097,
+      "grad_norm": 0.6506988406181335,
+      "learning_rate": 0.0001,
+      "loss": 1.314,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7320143884892086,
+      "grad_norm": 0.7010005116462708,
+      "learning_rate": 0.0001,
+      "loss": 0.993,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7321000342583076,
+      "grad_norm": 1.616136074066162,
+      "learning_rate": 0.0001,
+      "loss": 1.1027,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7321856800274067,
+      "grad_norm": 0.8405584692955017,
+      "learning_rate": 0.0001,
+      "loss": 1.0832,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7322713257965057,
+      "grad_norm": 0.4775457978248596,
+      "learning_rate": 0.0001,
+      "loss": 0.2057,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7323569715656046,
+      "grad_norm": 1.0647741556167603,
+      "learning_rate": 0.0001,
+      "loss": 0.8814,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7324426173347036,
+      "grad_norm": 1.267869234085083,
+      "learning_rate": 0.0001,
+      "loss": 1.1273,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7325282631038027,
+      "grad_norm": 0.4648052155971527,
+      "learning_rate": 0.0001,
+      "loss": 0.6238,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7326139088729017,
+      "grad_norm": 1.1691733598709106,
+      "learning_rate": 0.0001,
+      "loss": 1.1029,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7326995546420006,
+      "grad_norm": 0.49745607376098633,
+      "learning_rate": 0.0001,
+      "loss": 0.5633,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7327852004110997,
+      "grad_norm": 0.7985617518424988,
+      "learning_rate": 0.0001,
+      "loss": 0.8075,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7328708461801987,
+      "grad_norm": 0.7840003371238708,
+      "learning_rate": 0.0001,
+      "loss": 0.7172,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7329564919492977,
+      "grad_norm": 0.9149075746536255,
+      "learning_rate": 0.0001,
+      "loss": 0.8475,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7330421377183967,
+      "grad_norm": 0.862050473690033,
+      "learning_rate": 0.0001,
+      "loss": 1.2138,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7331277834874957,
+      "grad_norm": 0.9189406037330627,
+      "learning_rate": 0.0001,
+      "loss": 1.1224,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7332134292565947,
+      "grad_norm": 0.5108150839805603,
+      "learning_rate": 0.0001,
+      "loss": 0.9715,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7332990750256937,
+      "grad_norm": 0.8296217322349548,
+      "learning_rate": 0.0001,
+      "loss": 0.8671,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7333847207947928,
+      "grad_norm": 1.0412993431091309,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7334703665638918,
+      "grad_norm": 0.9884964823722839,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7335560123329907,
+      "grad_norm": 0.9098817110061646,
+      "learning_rate": 0.0001,
+      "loss": 1.2277,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7336416581020898,
+      "grad_norm": 0.5240215063095093,
+      "learning_rate": 0.0001,
+      "loss": 0.8808,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7337273038711888,
+      "grad_norm": 0.24801112711429596,
+      "learning_rate": 0.0001,
+      "loss": 0.2159,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7338129496402878,
+      "grad_norm": 0.7436892986297607,
+      "learning_rate": 0.0001,
+      "loss": 1.1465,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7338985954093867,
+      "grad_norm": 1.20950186252594,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7339842411784858,
+      "grad_norm": 1.065714955329895,
+      "learning_rate": 0.0001,
+      "loss": 0.5082,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7340698869475848,
+      "grad_norm": 1.2825539112091064,
+      "learning_rate": 0.0001,
+      "loss": 1.1239,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7341555327166838,
+      "grad_norm": 0.6805071830749512,
+      "learning_rate": 0.0001,
+      "loss": 0.8919,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7342411784857829,
+      "grad_norm": 0.767120897769928,
+      "learning_rate": 0.0001,
+      "loss": 0.8731,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7343268242548818,
+      "grad_norm": 1.159533977508545,
+      "learning_rate": 0.0001,
+      "loss": 1.2027,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7344124700239808,
+      "grad_norm": 0.672326385974884,
+      "learning_rate": 0.0001,
+      "loss": 0.4744,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7344981157930798,
+      "grad_norm": 1.2867143154144287,
+      "learning_rate": 0.0001,
+      "loss": 1.0662,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7345837615621789,
+      "grad_norm": 0.8526127934455872,
+      "learning_rate": 0.0001,
+      "loss": 1.271,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7346694073312778,
+      "grad_norm": 0.6471309661865234,
+      "learning_rate": 0.0001,
+      "loss": 0.9652,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7347550531003768,
+      "grad_norm": 0.7824822664260864,
+      "learning_rate": 0.0001,
+      "loss": 1.0373,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7348406988694759,
+      "grad_norm": 1.1811211109161377,
+      "learning_rate": 0.0001,
+      "loss": 1.1804,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7349263446385749,
+      "grad_norm": 1.2302933931350708,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7350119904076738,
+      "grad_norm": 0.7933406829833984,
+      "learning_rate": 0.0001,
+      "loss": 1.151,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7350976361767728,
+      "grad_norm": 1.1990076303482056,
+      "learning_rate": 0.0001,
+      "loss": 1.3793,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7351832819458719,
+      "grad_norm": 0.6301919221878052,
+      "learning_rate": 0.0001,
+      "loss": 0.7015,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7352689277149709,
+      "grad_norm": 0.3587115406990051,
+      "learning_rate": 0.0001,
+      "loss": 0.3723,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7353545734840699,
+      "grad_norm": 0.31488358974456787,
+      "learning_rate": 0.0001,
+      "loss": 0.4144,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7354402192531689,
+      "grad_norm": 1.403894305229187,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7355258650222679,
+      "grad_norm": 0.9064099788665771,
+      "learning_rate": 0.0001,
+      "loss": 1.1349,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7356115107913669,
+      "grad_norm": 0.7143677473068237,
+      "learning_rate": 0.0001,
+      "loss": 0.3129,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7356971565604659,
+      "grad_norm": 0.7792229056358337,
+      "learning_rate": 0.0001,
+      "loss": 1.0239,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7357828023295649,
+      "grad_norm": 0.7509090900421143,
+      "learning_rate": 0.0001,
+      "loss": 0.4991,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7358684480986639,
+      "grad_norm": 1.0535849332809448,
+      "learning_rate": 0.0001,
+      "loss": 1.0796,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7359540938677629,
+      "grad_norm": 0.5481096506118774,
+      "learning_rate": 0.0001,
+      "loss": 0.7239,
+      "step": 8593
+    },
+    {
+      "epoch": 0.736039739636862,
+      "grad_norm": 1.1976640224456787,
+      "learning_rate": 0.0001,
+      "loss": 1.0626,
+      "step": 8594
+    },
+    {
+      "epoch": 0.736125385405961,
+      "grad_norm": 0.6552754640579224,
+      "learning_rate": 0.0001,
+      "loss": 0.6672,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7362110311750599,
+      "grad_norm": 0.7168229818344116,
+      "learning_rate": 0.0001,
+      "loss": 0.5178,
+      "step": 8596
+    },
+    {
+      "epoch": 0.736296676944159,
+      "grad_norm": 1.1671112775802612,
+      "learning_rate": 0.0001,
+      "loss": 1.376,
+      "step": 8597
+    },
+    {
+      "epoch": 0.736382322713258,
+      "grad_norm": 0.5688337087631226,
+      "learning_rate": 0.0001,
+      "loss": 0.8995,
+      "step": 8598
+    },
+    {
+      "epoch": 0.736467968482357,
+      "grad_norm": 0.5679592490196228,
+      "learning_rate": 0.0001,
+      "loss": 0.6487,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7365536142514559,
+      "grad_norm": 0.9993517398834229,
+      "learning_rate": 0.0001,
+      "loss": 1.2165,
+      "step": 8600
+    },
+    {
+      "epoch": 0.736639260020555,
+      "grad_norm": 1.1469398736953735,
+      "learning_rate": 0.0001,
+      "loss": 0.6419,
+      "step": 8601
+    },
+    {
+      "epoch": 0.736724905789654,
+      "grad_norm": 0.6262681484222412,
+      "learning_rate": 0.0001,
+      "loss": 0.9433,
+      "step": 8602
+    },
+    {
+      "epoch": 0.736810551558753,
+      "grad_norm": 0.7156298756599426,
+      "learning_rate": 0.0001,
+      "loss": 0.5203,
+      "step": 8603
+    },
+    {
+      "epoch": 0.736896197327852,
+      "grad_norm": 1.2639405727386475,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 8604
+    },
+    {
+      "epoch": 0.736981843096951,
+      "grad_norm": 1.0552133321762085,
+      "learning_rate": 0.0001,
+      "loss": 0.9837,
+      "step": 8605
+    },
+    {
+      "epoch": 0.73706748886605,
+      "grad_norm": 0.9868779182434082,
+      "learning_rate": 0.0001,
+      "loss": 0.9451,
+      "step": 8606
+    },
+    {
+      "epoch": 0.737153134635149,
+      "grad_norm": 0.8522647023200989,
+      "learning_rate": 0.0001,
+      "loss": 0.7763,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7372387804042481,
+      "grad_norm": 0.7940419316291809,
+      "learning_rate": 0.0001,
+      "loss": 0.8037,
+      "step": 8608
+    },
+    {
+      "epoch": 0.737324426173347,
+      "grad_norm": 0.6239277124404907,
+      "learning_rate": 0.0001,
+      "loss": 0.8071,
+      "step": 8609
+    },
+    {
+      "epoch": 0.737410071942446,
+      "grad_norm": 0.7125552892684937,
+      "learning_rate": 0.0001,
+      "loss": 1.0424,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7374957177115451,
+      "grad_norm": 0.5199329853057861,
+      "learning_rate": 0.0001,
+      "loss": 1.0436,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7375813634806441,
+      "grad_norm": 1.0758838653564453,
+      "learning_rate": 0.0001,
+      "loss": 1.0976,
+      "step": 8612
+    },
+    {
+      "epoch": 0.737667009249743,
+      "grad_norm": 1.4832500219345093,
+      "learning_rate": 0.0001,
+      "loss": 1.1201,
+      "step": 8613
+    },
+    {
+      "epoch": 0.737752655018842,
+      "grad_norm": 0.9377415180206299,
+      "learning_rate": 0.0001,
+      "loss": 0.767,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7378383007879411,
+      "grad_norm": 0.613138735294342,
+      "learning_rate": 0.0001,
+      "loss": 0.4639,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7379239465570401,
+      "grad_norm": 1.463179111480713,
+      "learning_rate": 0.0001,
+      "loss": 2.1173,
+      "step": 8616
+    },
+    {
+      "epoch": 0.738009592326139,
+      "grad_norm": 0.7131281495094299,
+      "learning_rate": 0.0001,
+      "loss": 0.6343,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7380952380952381,
+      "grad_norm": 0.7177920937538147,
+      "learning_rate": 0.0001,
+      "loss": 0.9259,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7381808838643371,
+      "grad_norm": 1.2609951496124268,
+      "learning_rate": 0.0001,
+      "loss": 1.0394,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7382665296334361,
+      "grad_norm": 0.935879111289978,
+      "learning_rate": 0.0001,
+      "loss": 1.4114,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7383521754025351,
+      "grad_norm": 0.3233628571033478,
+      "learning_rate": 0.0001,
+      "loss": 0.135,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7384378211716341,
+      "grad_norm": 0.9327612519264221,
+      "learning_rate": 0.0001,
+      "loss": 0.7161,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7385234669407331,
+      "grad_norm": 0.8407263159751892,
+      "learning_rate": 0.0001,
+      "loss": 0.8001,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7386091127098321,
+      "grad_norm": 0.8486418128013611,
+      "learning_rate": 0.0001,
+      "loss": 0.9484,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7386947584789312,
+      "grad_norm": 0.7453429698944092,
+      "learning_rate": 0.0001,
+      "loss": 1.1585,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7387804042480302,
+      "grad_norm": 0.5930347442626953,
+      "learning_rate": 0.0001,
+      "loss": 0.5642,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7388660500171291,
+      "grad_norm": 1.357508659362793,
+      "learning_rate": 0.0001,
+      "loss": 0.7861,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7389516957862282,
+      "grad_norm": 1.1276041269302368,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7390373415553272,
+      "grad_norm": 0.9836664199829102,
+      "learning_rate": 0.0001,
+      "loss": 0.7886,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7391229873244262,
+      "grad_norm": 0.9016251564025879,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7392086330935251,
+      "grad_norm": 0.94005286693573,
+      "learning_rate": 0.0001,
+      "loss": 0.8425,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7392942788626242,
+      "grad_norm": 1.0963302850723267,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7393799246317232,
+      "grad_norm": 1.1101466417312622,
+      "learning_rate": 0.0001,
+      "loss": 1.1309,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7394655704008222,
+      "grad_norm": 0.6989795565605164,
+      "learning_rate": 0.0001,
+      "loss": 1.1254,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7395512161699213,
+      "grad_norm": 1.1911354064941406,
+      "learning_rate": 0.0001,
+      "loss": 1.1862,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7396368619390202,
+      "grad_norm": 1.204333782196045,
+      "learning_rate": 0.0001,
+      "loss": 1.0256,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7397225077081192,
+      "grad_norm": 0.8383384346961975,
+      "learning_rate": 0.0001,
+      "loss": 1.0592,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7398081534772182,
+      "grad_norm": 0.7262300848960876,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7398937992463173,
+      "grad_norm": 0.9809969663619995,
+      "learning_rate": 0.0001,
+      "loss": 1.1236,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7399794450154162,
+      "grad_norm": 0.6116905212402344,
+      "learning_rate": 0.0001,
+      "loss": 0.6416,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7400650907845152,
+      "grad_norm": 0.9271953105926514,
+      "learning_rate": 0.0001,
+      "loss": 1.241,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7401507365536143,
+      "grad_norm": 0.8968188166618347,
+      "learning_rate": 0.0001,
+      "loss": 0.7991,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7402363823227133,
+      "grad_norm": 0.648825466632843,
+      "learning_rate": 0.0001,
+      "loss": 0.6316,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7403220280918122,
+      "grad_norm": 0.6983048915863037,
+      "learning_rate": 0.0001,
+      "loss": 1.2052,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7404076738609112,
+      "grad_norm": 0.952521800994873,
+      "learning_rate": 0.0001,
+      "loss": 0.8933,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7404933196300103,
+      "grad_norm": 0.8015673756599426,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7405789653991093,
+      "grad_norm": 0.8753068447113037,
+      "learning_rate": 0.0001,
+      "loss": 1.064,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7406646111682083,
+      "grad_norm": 1.0815714597702026,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7407502569373073,
+      "grad_norm": 0.5517322421073914,
+      "learning_rate": 0.0001,
+      "loss": 0.8928,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7408359027064063,
+      "grad_norm": 0.569942057132721,
+      "learning_rate": 0.0001,
+      "loss": 0.4949,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7409215484755053,
+      "grad_norm": 0.8239222764968872,
+      "learning_rate": 0.0001,
+      "loss": 0.5933,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7410071942446043,
+      "grad_norm": 0.7870293259620667,
+      "learning_rate": 0.0001,
+      "loss": 1.1243,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7410928400137033,
+      "grad_norm": 1.0358235836029053,
+      "learning_rate": 0.0001,
+      "loss": 1.1157,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7411784857828023,
+      "grad_norm": 0.715076744556427,
+      "learning_rate": 0.0001,
+      "loss": 1.1397,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7412641315519013,
+      "grad_norm": 1.1187764406204224,
+      "learning_rate": 0.0001,
+      "loss": 1.0355,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7413497773210004,
+      "grad_norm": 0.9408628940582275,
+      "learning_rate": 0.0001,
+      "loss": 0.886,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7414354230900994,
+      "grad_norm": 1.3163888454437256,
+      "learning_rate": 0.0001,
+      "loss": 0.7927,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7415210688591983,
+      "grad_norm": 1.3079153299331665,
+      "learning_rate": 0.0001,
+      "loss": 1.2499,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7416067146282974,
+      "grad_norm": 0.9892915487289429,
+      "learning_rate": 0.0001,
+      "loss": 0.9418,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7416923603973964,
+      "grad_norm": 0.6752997040748596,
+      "learning_rate": 0.0001,
+      "loss": 0.8489,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7417780061664954,
+      "grad_norm": 0.7837368845939636,
+      "learning_rate": 0.0001,
+      "loss": 1.1944,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7418636519355943,
+      "grad_norm": 1.1363797187805176,
+      "learning_rate": 0.0001,
+      "loss": 1.3404,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7419492977046934,
+      "grad_norm": 1.14016592502594,
+      "learning_rate": 0.0001,
+      "loss": 0.9835,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7420349434737924,
+      "grad_norm": 0.7487924695014954,
+      "learning_rate": 0.0001,
+      "loss": 0.9915,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7421205892428914,
+      "grad_norm": 1.0421836376190186,
+      "learning_rate": 0.0001,
+      "loss": 1.0204,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7422062350119905,
+      "grad_norm": 0.841944694519043,
+      "learning_rate": 0.0001,
+      "loss": 0.6978,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7422918807810894,
+      "grad_norm": 1.4532848596572876,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7423775265501884,
+      "grad_norm": 0.8344189524650574,
+      "learning_rate": 0.0001,
+      "loss": 1.2354,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7424631723192874,
+      "grad_norm": 0.9028735160827637,
+      "learning_rate": 0.0001,
+      "loss": 0.8182,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7425488180883865,
+      "grad_norm": 0.8252964019775391,
+      "learning_rate": 0.0001,
+      "loss": 0.8766,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7426344638574854,
+      "grad_norm": 0.9535041451454163,
+      "learning_rate": 0.0001,
+      "loss": 0.872,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7427201096265844,
+      "grad_norm": 1.3913151025772095,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7428057553956835,
+      "grad_norm": 1.3766770362854004,
+      "learning_rate": 0.0001,
+      "loss": 1.3193,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7428914011647825,
+      "grad_norm": 0.7959471344947815,
+      "learning_rate": 0.0001,
+      "loss": 0.8562,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7429770469338814,
+      "grad_norm": 1.378125548362732,
+      "learning_rate": 0.0001,
+      "loss": 0.7494,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7430626927029804,
+      "grad_norm": 0.9892541766166687,
+      "learning_rate": 0.0001,
+      "loss": 1.0643,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7431483384720795,
+      "grad_norm": 0.728908896446228,
+      "learning_rate": 0.0001,
+      "loss": 0.7543,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7432339842411785,
+      "grad_norm": 0.3601706326007843,
+      "learning_rate": 0.0001,
+      "loss": 0.6096,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7433196300102775,
+      "grad_norm": 0.8482088446617126,
+      "learning_rate": 0.0001,
+      "loss": 1.1228,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7434052757793765,
+      "grad_norm": 0.6901963353157043,
+      "learning_rate": 0.0001,
+      "loss": 1.1314,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7434909215484755,
+      "grad_norm": 0.6800196766853333,
+      "learning_rate": 0.0001,
+      "loss": 1.2062,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7435765673175745,
+      "grad_norm": 1.153091549873352,
+      "learning_rate": 0.0001,
+      "loss": 1.263,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7436622130866736,
+      "grad_norm": 0.4123636782169342,
+      "learning_rate": 0.0001,
+      "loss": 0.4466,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7437478588557725,
+      "grad_norm": 0.5844290852546692,
+      "learning_rate": 0.0001,
+      "loss": 0.7827,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7438335046248715,
+      "grad_norm": 1.1081832647323608,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7439191503939705,
+      "grad_norm": 0.7579890489578247,
+      "learning_rate": 0.0001,
+      "loss": 0.7321,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7440047961630696,
+      "grad_norm": 0.7201582193374634,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7440904419321686,
+      "grad_norm": 0.9126925468444824,
+      "learning_rate": 0.0001,
+      "loss": 0.7659,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7441760877012675,
+      "grad_norm": 0.5133847594261169,
+      "learning_rate": 0.0001,
+      "loss": 0.4596,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7442617334703666,
+      "grad_norm": 0.6733587384223938,
+      "learning_rate": 0.0001,
+      "loss": 1.0317,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7443473792394656,
+      "grad_norm": 1.1693042516708374,
+      "learning_rate": 0.0001,
+      "loss": 1.2247,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7444330250085646,
+      "grad_norm": 0.6810747385025024,
+      "learning_rate": 0.0001,
+      "loss": 1.1831,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7445186707776635,
+      "grad_norm": 0.6701362133026123,
+      "learning_rate": 0.0001,
+      "loss": 1.3014,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7446043165467626,
+      "grad_norm": 0.9544815421104431,
+      "learning_rate": 0.0001,
+      "loss": 1.2052,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7446899623158616,
+      "grad_norm": 0.630055844783783,
+      "learning_rate": 0.0001,
+      "loss": 0.5382,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7447756080849606,
+      "grad_norm": 1.1171045303344727,
+      "learning_rate": 0.0001,
+      "loss": 0.8044,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7448612538540597,
+      "grad_norm": 0.6382997035980225,
+      "learning_rate": 0.0001,
+      "loss": 0.5326,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7449468996231586,
+      "grad_norm": 0.8571712970733643,
+      "learning_rate": 0.0001,
+      "loss": 1.1247,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7450325453922576,
+      "grad_norm": 0.8320066332817078,
+      "learning_rate": 0.0001,
+      "loss": 0.9708,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7451181911613566,
+      "grad_norm": 0.9371066093444824,
+      "learning_rate": 0.0001,
+      "loss": 1.3671,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7452038369304557,
+      "grad_norm": 0.8613821268081665,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7452894826995546,
+      "grad_norm": 1.4107128381729126,
+      "learning_rate": 0.0001,
+      "loss": 1.3385,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7453751284686536,
+      "grad_norm": 0.8099102973937988,
+      "learning_rate": 0.0001,
+      "loss": 0.8488,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7454607742377527,
+      "grad_norm": 0.7408746480941772,
+      "learning_rate": 0.0001,
+      "loss": 1.2948,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7455464200068517,
+      "grad_norm": 1.0811121463775635,
+      "learning_rate": 0.0001,
+      "loss": 1.0597,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7456320657759506,
+      "grad_norm": 0.40463486313819885,
+      "learning_rate": 0.0001,
+      "loss": 0.1156,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7457177115450496,
+      "grad_norm": 1.013086199760437,
+      "learning_rate": 0.0001,
+      "loss": 0.9277,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7458033573141487,
+      "grad_norm": 1.1732221841812134,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7458890030832477,
+      "grad_norm": 0.5662166476249695,
+      "learning_rate": 0.0001,
+      "loss": 0.9137,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7459746488523467,
+      "grad_norm": 0.4930499494075775,
+      "learning_rate": 0.0001,
+      "loss": 0.7207,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7460602946214457,
+      "grad_norm": 0.5753157138824463,
+      "learning_rate": 0.0001,
+      "loss": 0.384,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7461459403905447,
+      "grad_norm": 0.9861400723457336,
+      "learning_rate": 0.0001,
+      "loss": 1.2712,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7462315861596437,
+      "grad_norm": 0.9747478365898132,
+      "learning_rate": 0.0001,
+      "loss": 0.9365,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7463172319287428,
+      "grad_norm": 0.5336748957633972,
+      "learning_rate": 0.0001,
+      "loss": 0.7419,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7464028776978417,
+      "grad_norm": 1.050990343093872,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7464885234669407,
+      "grad_norm": 0.9842614531517029,
+      "learning_rate": 0.0001,
+      "loss": 0.6605,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7465741692360397,
+      "grad_norm": 0.6047006249427795,
+      "learning_rate": 0.0001,
+      "loss": 0.9607,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7466598150051388,
+      "grad_norm": 1.18828547000885,
+      "learning_rate": 0.0001,
+      "loss": 1.3035,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7467454607742378,
+      "grad_norm": 0.8280365467071533,
+      "learning_rate": 0.0001,
+      "loss": 0.7945,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7468311065433367,
+      "grad_norm": 0.5867172479629517,
+      "learning_rate": 0.0001,
+      "loss": 0.3393,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7469167523124358,
+      "grad_norm": 0.7095484137535095,
+      "learning_rate": 0.0001,
+      "loss": 0.6321,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7470023980815348,
+      "grad_norm": 0.7709778547286987,
+      "learning_rate": 0.0001,
+      "loss": 0.5666,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7470880438506338,
+      "grad_norm": 0.7831560969352722,
+      "learning_rate": 0.0001,
+      "loss": 1.3551,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7471736896197327,
+      "grad_norm": 0.9105431437492371,
+      "learning_rate": 0.0001,
+      "loss": 0.877,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7472593353888318,
+      "grad_norm": 0.6682350039482117,
+      "learning_rate": 0.0001,
+      "loss": 1.0948,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7473449811579308,
+      "grad_norm": 0.8384807705879211,
+      "learning_rate": 0.0001,
+      "loss": 1.1512,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7474306269270298,
+      "grad_norm": 0.8254141211509705,
+      "learning_rate": 0.0001,
+      "loss": 1.25,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7475162726961289,
+      "grad_norm": 0.6865155696868896,
+      "learning_rate": 0.0001,
+      "loss": 0.4097,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7476019184652278,
+      "grad_norm": 1.6585661172866821,
+      "learning_rate": 0.0001,
+      "loss": 1.8709,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7476875642343268,
+      "grad_norm": 0.8665419816970825,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7477732100034258,
+      "grad_norm": 1.0182595252990723,
+      "learning_rate": 0.0001,
+      "loss": 0.9208,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7478588557725249,
+      "grad_norm": 0.7368389368057251,
+      "learning_rate": 0.0001,
+      "loss": 0.9615,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7479445015416238,
+      "grad_norm": 0.9690853953361511,
+      "learning_rate": 0.0001,
+      "loss": 1.0931,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7480301473107228,
+      "grad_norm": 1.009142518043518,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7481157930798219,
+      "grad_norm": 1.234938144683838,
+      "learning_rate": 0.0001,
+      "loss": 1.057,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7482014388489209,
+      "grad_norm": 0.7695930004119873,
+      "learning_rate": 0.0001,
+      "loss": 1.1732,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7482870846180198,
+      "grad_norm": 0.7516759634017944,
+      "learning_rate": 0.0001,
+      "loss": 1.0641,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7483727303871188,
+      "grad_norm": 0.8572515845298767,
+      "learning_rate": 0.0001,
+      "loss": 0.5031,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7484583761562179,
+      "grad_norm": 0.7287034392356873,
+      "learning_rate": 0.0001,
+      "loss": 1.3468,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7485440219253169,
+      "grad_norm": 0.8905504941940308,
+      "learning_rate": 0.0001,
+      "loss": 0.6016,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7486296676944159,
+      "grad_norm": 0.7791568636894226,
+      "learning_rate": 0.0001,
+      "loss": 1.0264,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7487153134635149,
+      "grad_norm": 0.7440105080604553,
+      "learning_rate": 0.0001,
+      "loss": 0.9008,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7488009592326139,
+      "grad_norm": 0.9672698378562927,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7488866050017129,
+      "grad_norm": 1.621350646018982,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 8744
+    },
+    {
+      "epoch": 0.748972250770812,
+      "grad_norm": 0.9459127187728882,
+      "learning_rate": 0.0001,
+      "loss": 0.9929,
+      "step": 8745
+    },
+    {
+      "epoch": 0.749057896539911,
+      "grad_norm": 0.4032800495624542,
+      "learning_rate": 0.0001,
+      "loss": 0.9361,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7491435423090099,
+      "grad_norm": 0.8052759170532227,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7492291880781089,
+      "grad_norm": 1.016761064529419,
+      "learning_rate": 0.0001,
+      "loss": 0.6907,
+      "step": 8748
+    },
+    {
+      "epoch": 0.749314833847208,
+      "grad_norm": 1.3156813383102417,
+      "learning_rate": 0.0001,
+      "loss": 0.9369,
+      "step": 8749
+    },
+    {
+      "epoch": 0.749400479616307,
+      "grad_norm": 1.190130352973938,
+      "learning_rate": 0.0001,
+      "loss": 1.1809,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7494861253854059,
+      "grad_norm": 0.6982339024543762,
+      "learning_rate": 0.0001,
+      "loss": 1.0082,
+      "step": 8751
+    },
+    {
+      "epoch": 0.749571771154505,
+      "grad_norm": 0.9580833315849304,
+      "learning_rate": 0.0001,
+      "loss": 0.6104,
+      "step": 8752
+    },
+    {
+      "epoch": 0.749657416923604,
+      "grad_norm": 0.9675077199935913,
+      "learning_rate": 0.0001,
+      "loss": 0.7791,
+      "step": 8753
+    },
+    {
+      "epoch": 0.749743062692703,
+      "grad_norm": 1.0684641599655151,
+      "learning_rate": 0.0001,
+      "loss": 0.6428,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7498287084618019,
+      "grad_norm": 0.6657161116600037,
+      "learning_rate": 0.0001,
+      "loss": 1.4153,
+      "step": 8755
+    },
+    {
+      "epoch": 0.749914354230901,
+      "grad_norm": 0.9057447910308838,
+      "learning_rate": 0.0001,
+      "loss": 0.5116,
+      "step": 8756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8038339614868164,
+      "learning_rate": 0.0001,
+      "loss": 1.3122,
+      "step": 8757
+    },
+    {
+      "epoch": 0.750085645769099,
+      "grad_norm": 0.5746389031410217,
+      "learning_rate": 0.0001,
+      "loss": 0.5715,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7501712915381981,
+      "grad_norm": 0.832206666469574,
+      "learning_rate": 0.0001,
+      "loss": 1.2179,
+      "step": 8759
+    },
+    {
+      "epoch": 0.750256937307297,
+      "grad_norm": 1.1059813499450684,
+      "learning_rate": 0.0001,
+      "loss": 0.6953,
+      "step": 8760
+    },
+    {
+      "epoch": 0.750342583076396,
+      "grad_norm": 1.1003234386444092,
+      "learning_rate": 0.0001,
+      "loss": 1.0301,
+      "step": 8761
+    },
+    {
+      "epoch": 0.750428228845495,
+      "grad_norm": 0.5313746333122253,
+      "learning_rate": 0.0001,
+      "loss": 0.8953,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7505138746145941,
+      "grad_norm": 0.8279226422309875,
+      "learning_rate": 0.0001,
+      "loss": 0.9647,
+      "step": 8763
+    },
+    {
+      "epoch": 0.750599520383693,
+      "grad_norm": 0.5718855857849121,
+      "learning_rate": 0.0001,
+      "loss": 0.7132,
+      "step": 8764
+    },
+    {
+      "epoch": 0.750685166152792,
+      "grad_norm": 0.7121441960334778,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7507708119218911,
+      "grad_norm": 1.4012573957443237,
+      "learning_rate": 0.0001,
+      "loss": 0.9294,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7508564576909901,
+      "grad_norm": 1.0922493934631348,
+      "learning_rate": 0.0001,
+      "loss": 0.8851,
+      "step": 8767
+    },
+    {
+      "epoch": 0.750942103460089,
+      "grad_norm": 0.3692774772644043,
+      "learning_rate": 0.0001,
+      "loss": 0.2483,
+      "step": 8768
+    },
+    {
+      "epoch": 0.751027749229188,
+      "grad_norm": 0.44385313987731934,
+      "learning_rate": 0.0001,
+      "loss": 0.5841,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7511133949982871,
+      "grad_norm": 0.8860092759132385,
+      "learning_rate": 0.0001,
+      "loss": 0.8619,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7511990407673861,
+      "grad_norm": 0.9144753813743591,
+      "learning_rate": 0.0001,
+      "loss": 0.8958,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7512846865364851,
+      "grad_norm": 1.0138827562332153,
+      "learning_rate": 0.0001,
+      "loss": 1.1696,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7513703323055841,
+      "grad_norm": 0.7034373879432678,
+      "learning_rate": 0.0001,
+      "loss": 0.7502,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7514559780746831,
+      "grad_norm": 0.749322235584259,
+      "learning_rate": 0.0001,
+      "loss": 1.0582,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7515416238437821,
+      "grad_norm": 0.6448805928230286,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7516272696128812,
+      "grad_norm": 1.0351574420928955,
+      "learning_rate": 0.0001,
+      "loss": 1.0232,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7517129153819802,
+      "grad_norm": 0.5617785453796387,
+      "learning_rate": 0.0001,
+      "loss": 0.9803,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7517985611510791,
+      "grad_norm": 1.0231915712356567,
+      "learning_rate": 0.0001,
+      "loss": 0.849,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7518842069201781,
+      "grad_norm": 0.6498993635177612,
+      "learning_rate": 0.0001,
+      "loss": 0.8546,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7519698526892772,
+      "grad_norm": 0.972949743270874,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7520554984583762,
+      "grad_norm": 0.915507435798645,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7521411442274751,
+      "grad_norm": 0.4702945947647095,
+      "learning_rate": 0.0001,
+      "loss": 0.4933,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7522267899965742,
+      "grad_norm": 1.3932605981826782,
+      "learning_rate": 0.0001,
+      "loss": 1.4047,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7523124357656732,
+      "grad_norm": 0.9451853632926941,
+      "learning_rate": 0.0001,
+      "loss": 1.1712,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7523980815347722,
+      "grad_norm": 0.8418645858764648,
+      "learning_rate": 0.0001,
+      "loss": 1.0671,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7524837273038711,
+      "grad_norm": 0.7910413146018982,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7525693730729702,
+      "grad_norm": 1.3543150424957275,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7526550188420692,
+      "grad_norm": 0.6409228444099426,
+      "learning_rate": 0.0001,
+      "loss": 1.0647,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7527406646111682,
+      "grad_norm": 0.8034427762031555,
+      "learning_rate": 0.0001,
+      "loss": 0.8094,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7528263103802673,
+      "grad_norm": 0.767596423625946,
+      "learning_rate": 0.0001,
+      "loss": 1.4183,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7529119561493662,
+      "grad_norm": 1.0532041788101196,
+      "learning_rate": 0.0001,
+      "loss": 0.5707,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7529976019184652,
+      "grad_norm": 0.6075825095176697,
+      "learning_rate": 0.0001,
+      "loss": 1.235,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7530832476875642,
+      "grad_norm": 0.7449963688850403,
+      "learning_rate": 0.0001,
+      "loss": 0.6975,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7531688934566633,
+      "grad_norm": 0.9320348501205444,
+      "learning_rate": 0.0001,
+      "loss": 1.182,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7532545392257622,
+      "grad_norm": 0.8250600695610046,
+      "learning_rate": 0.0001,
+      "loss": 0.6089,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7533401849948612,
+      "grad_norm": 1.2909502983093262,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7534258307639603,
+      "grad_norm": 0.8067193627357483,
+      "learning_rate": 0.0001,
+      "loss": 0.97,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7535114765330593,
+      "grad_norm": 0.7327744960784912,
+      "learning_rate": 0.0001,
+      "loss": 0.2604,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7535971223021583,
+      "grad_norm": 0.9592378735542297,
+      "learning_rate": 0.0001,
+      "loss": 1.0036,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7536827680712572,
+      "grad_norm": 0.9641855955123901,
+      "learning_rate": 0.0001,
+      "loss": 1.2026,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7537684138403563,
+      "grad_norm": 1.0288074016571045,
+      "learning_rate": 0.0001,
+      "loss": 0.8706,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7538540596094553,
+      "grad_norm": 0.9330957531929016,
+      "learning_rate": 0.0001,
+      "loss": 0.7088,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7539397053785543,
+      "grad_norm": 1.0053375959396362,
+      "learning_rate": 0.0001,
+      "loss": 0.7825,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7540253511476533,
+      "grad_norm": 0.7488108277320862,
+      "learning_rate": 0.0001,
+      "loss": 1.0284,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7541109969167523,
+      "grad_norm": 0.6438100934028625,
+      "learning_rate": 0.0001,
+      "loss": 0.5007,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7541966426858513,
+      "grad_norm": 1.165541648864746,
+      "learning_rate": 0.0001,
+      "loss": 1.7653,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7542822884549504,
+      "grad_norm": 0.9234575033187866,
+      "learning_rate": 0.0001,
+      "loss": 1.2379,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7543679342240494,
+      "grad_norm": 0.7340512275695801,
+      "learning_rate": 0.0001,
+      "loss": 0.9483,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7544535799931483,
+      "grad_norm": 0.8041210174560547,
+      "learning_rate": 0.0001,
+      "loss": 1.2229,
+      "step": 8809
+    },
+    {
+      "epoch": 0.7545392257622473,
+      "grad_norm": 0.9504506587982178,
+      "learning_rate": 0.0001,
+      "loss": 0.9392,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7546248715313464,
+      "grad_norm": 1.0577155351638794,
+      "learning_rate": 0.0001,
+      "loss": 1.8361,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7547105173004454,
+      "grad_norm": 0.786949872970581,
+      "learning_rate": 0.0001,
+      "loss": 0.8308,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7547961630695443,
+      "grad_norm": 0.7435232400894165,
+      "learning_rate": 0.0001,
+      "loss": 1.2397,
+      "step": 8813
+    },
+    {
+      "epoch": 0.7548818088386434,
+      "grad_norm": 0.9951729774475098,
+      "learning_rate": 0.0001,
+      "loss": 1.1387,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7549674546077424,
+      "grad_norm": 0.7988361120223999,
+      "learning_rate": 0.0001,
+      "loss": 0.8861,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7550531003768414,
+      "grad_norm": 0.3633357882499695,
+      "learning_rate": 0.0001,
+      "loss": 0.1327,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7551387461459403,
+      "grad_norm": 0.6324083209037781,
+      "learning_rate": 0.0001,
+      "loss": 1.058,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7552243919150394,
+      "grad_norm": 1.1266148090362549,
+      "learning_rate": 0.0001,
+      "loss": 1.3496,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7553100376841384,
+      "grad_norm": 0.7978406548500061,
+      "learning_rate": 0.0001,
+      "loss": 0.7994,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7553956834532374,
+      "grad_norm": 0.5866055488586426,
+      "learning_rate": 0.0001,
+      "loss": 1.1767,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7554813292223365,
+      "grad_norm": 0.6194197535514832,
+      "learning_rate": 0.0001,
+      "loss": 1.0224,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7555669749914354,
+      "grad_norm": 1.183215618133545,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7556526207605344,
+      "grad_norm": 0.8914014101028442,
+      "learning_rate": 0.0001,
+      "loss": 0.5259,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7557382665296334,
+      "grad_norm": 1.117340087890625,
+      "learning_rate": 0.0001,
+      "loss": 1.0759,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7558239122987325,
+      "grad_norm": 0.8740167617797852,
+      "learning_rate": 0.0001,
+      "loss": 0.8606,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7559095580678314,
+      "grad_norm": 1.0150768756866455,
+      "learning_rate": 0.0001,
+      "loss": 1.17,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7559952038369304,
+      "grad_norm": 0.7053858041763306,
+      "learning_rate": 0.0001,
+      "loss": 1.0222,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7560808496060295,
+      "grad_norm": 0.886874794960022,
+      "learning_rate": 0.0001,
+      "loss": 0.8704,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7561664953751285,
+      "grad_norm": 1.1431279182434082,
+      "learning_rate": 0.0001,
+      "loss": 1.0533,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7562521411442275,
+      "grad_norm": 0.9239665865898132,
+      "learning_rate": 0.0001,
+      "loss": 0.9134,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7563377869133264,
+      "grad_norm": 0.5763223767280579,
+      "learning_rate": 0.0001,
+      "loss": 0.5776,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7564234326824255,
+      "grad_norm": 0.6673739552497864,
+      "learning_rate": 0.0001,
+      "loss": 0.65,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7565090784515245,
+      "grad_norm": 0.7692014575004578,
+      "learning_rate": 0.0001,
+      "loss": 1.0509,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7565947242206235,
+      "grad_norm": 1.2942986488342285,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7566803699897225,
+      "grad_norm": 0.5932096242904663,
+      "learning_rate": 0.0001,
+      "loss": 0.9935,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7567660157588215,
+      "grad_norm": 1.0773334503173828,
+      "learning_rate": 0.0001,
+      "loss": 1.0565,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7568516615279205,
+      "grad_norm": 0.5991517901420593,
+      "learning_rate": 0.0001,
+      "loss": 1.0798,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7569373072970196,
+      "grad_norm": 0.594413161277771,
+      "learning_rate": 0.0001,
+      "loss": 0.7125,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7570229530661186,
+      "grad_norm": 1.0836925506591797,
+      "learning_rate": 0.0001,
+      "loss": 1.0794,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7571085988352175,
+      "grad_norm": 0.9124338626861572,
+      "learning_rate": 0.0001,
+      "loss": 1.2931,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7571942446043165,
+      "grad_norm": 0.751033365726471,
+      "learning_rate": 0.0001,
+      "loss": 0.9026,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7572798903734156,
+      "grad_norm": 0.5554829835891724,
+      "learning_rate": 0.0001,
+      "loss": 0.4651,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7573655361425146,
+      "grad_norm": 0.5819361805915833,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7574511819116135,
+      "grad_norm": 0.6291237473487854,
+      "learning_rate": 0.0001,
+      "loss": 0.7036,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7575368276807126,
+      "grad_norm": 0.7270961999893188,
+      "learning_rate": 0.0001,
+      "loss": 1.3971,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7576224734498116,
+      "grad_norm": 0.7837800979614258,
+      "learning_rate": 0.0001,
+      "loss": 1.3107,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7577081192189106,
+      "grad_norm": 0.5551730394363403,
+      "learning_rate": 0.0001,
+      "loss": 1.0149,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7577937649880095,
+      "grad_norm": 1.1670211553573608,
+      "learning_rate": 0.0001,
+      "loss": 1.0219,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7578794107571086,
+      "grad_norm": 0.7427196502685547,
+      "learning_rate": 0.0001,
+      "loss": 0.7087,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7579650565262076,
+      "grad_norm": 0.5578095316886902,
+      "learning_rate": 0.0001,
+      "loss": 0.8485,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7580507022953066,
+      "grad_norm": 0.9079173803329468,
+      "learning_rate": 0.0001,
+      "loss": 0.948,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7581363480644057,
+      "grad_norm": 0.4047374427318573,
+      "learning_rate": 0.0001,
+      "loss": 0.4118,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7582219938335046,
+      "grad_norm": 1.3857603073120117,
+      "learning_rate": 0.0001,
+      "loss": 1.8419,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7583076396026036,
+      "grad_norm": 0.3910633325576782,
+      "learning_rate": 0.0001,
+      "loss": 0.3698,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7583932853717026,
+      "grad_norm": 0.25815635919570923,
+      "learning_rate": 0.0001,
+      "loss": 0.4298,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7584789311408017,
+      "grad_norm": 1.002934217453003,
+      "learning_rate": 0.0001,
+      "loss": 0.9272,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7585645769099006,
+      "grad_norm": 0.8405870795249939,
+      "learning_rate": 0.0001,
+      "loss": 0.6488,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7586502226789996,
+      "grad_norm": 0.5748960971832275,
+      "learning_rate": 0.0001,
+      "loss": 0.528,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7587358684480987,
+      "grad_norm": 0.9736438393592834,
+      "learning_rate": 0.0001,
+      "loss": 0.7725,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7588215142171977,
+      "grad_norm": 0.6621454954147339,
+      "learning_rate": 0.0001,
+      "loss": 0.9173,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7589071599862967,
+      "grad_norm": 1.0215580463409424,
+      "learning_rate": 0.0001,
+      "loss": 0.6924,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7589928057553957,
+      "grad_norm": 0.6479471325874329,
+      "learning_rate": 0.0001,
+      "loss": 1.1528,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7590784515244947,
+      "grad_norm": 0.9969015121459961,
+      "learning_rate": 0.0001,
+      "loss": 1.0122,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7591640972935937,
+      "grad_norm": 0.631708025932312,
+      "learning_rate": 0.0001,
+      "loss": 0.8018,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7592497430626927,
+      "grad_norm": 0.9046694040298462,
+      "learning_rate": 0.0001,
+      "loss": 0.7668,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7593353888317917,
+      "grad_norm": 0.8551928400993347,
+      "learning_rate": 0.0001,
+      "loss": 0.6801,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7594210346008907,
+      "grad_norm": 1.502137303352356,
+      "learning_rate": 0.0001,
+      "loss": 1.1145,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7595066803699897,
+      "grad_norm": 0.5309129357337952,
+      "learning_rate": 0.0001,
+      "loss": 0.3449,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7595923261390888,
+      "grad_norm": 1.182918906211853,
+      "learning_rate": 0.0001,
+      "loss": 0.8539,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7596779719081878,
+      "grad_norm": 1.2236875295639038,
+      "learning_rate": 0.0001,
+      "loss": 1.1105,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7597636176772867,
+      "grad_norm": 0.7851039171218872,
+      "learning_rate": 0.0001,
+      "loss": 0.8325,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7598492634463857,
+      "grad_norm": 1.0875988006591797,
+      "learning_rate": 0.0001,
+      "loss": 1.8845,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7599349092154848,
+      "grad_norm": 0.5457683205604553,
+      "learning_rate": 0.0001,
+      "loss": 0.9896,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7600205549845838,
+      "grad_norm": 1.5984314680099487,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7601062007536827,
+      "grad_norm": 0.9863172769546509,
+      "learning_rate": 0.0001,
+      "loss": 1.1107,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7601918465227818,
+      "grad_norm": 0.9743345975875854,
+      "learning_rate": 0.0001,
+      "loss": 0.886,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7602774922918808,
+      "grad_norm": 1.1317558288574219,
+      "learning_rate": 0.0001,
+      "loss": 1.0242,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7603631380609798,
+      "grad_norm": 0.6135861873626709,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7604487838300787,
+      "grad_norm": 1.5134141445159912,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7605344295991778,
+      "grad_norm": 1.1389250755310059,
+      "learning_rate": 0.0001,
+      "loss": 1.2327,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7606200753682768,
+      "grad_norm": 1.2369376420974731,
+      "learning_rate": 0.0001,
+      "loss": 0.7465,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7607057211373758,
+      "grad_norm": 0.5685446262359619,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7607913669064749,
+      "grad_norm": 0.8683727383613586,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7608770126755738,
+      "grad_norm": 0.8827126622200012,
+      "learning_rate": 0.0001,
+      "loss": 0.9547,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7609626584446728,
+      "grad_norm": 0.719866156578064,
+      "learning_rate": 0.0001,
+      "loss": 1.1239,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7610483042137718,
+      "grad_norm": 0.953847348690033,
+      "learning_rate": 0.0001,
+      "loss": 0.9611,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7611339499828709,
+      "grad_norm": 1.1982332468032837,
+      "learning_rate": 0.0001,
+      "loss": 0.7963,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7612195957519698,
+      "grad_norm": 0.842888593673706,
+      "learning_rate": 0.0001,
+      "loss": 0.8377,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7613052415210688,
+      "grad_norm": 0.7377665638923645,
+      "learning_rate": 0.0001,
+      "loss": 0.6691,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7613908872901679,
+      "grad_norm": 0.411740779876709,
+      "learning_rate": 0.0001,
+      "loss": 0.4719,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7614765330592669,
+      "grad_norm": 0.7185056805610657,
+      "learning_rate": 0.0001,
+      "loss": 0.9907,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7615621788283659,
+      "grad_norm": 0.5652514100074768,
+      "learning_rate": 0.0001,
+      "loss": 0.5488,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7616478245974649,
+      "grad_norm": 0.6480763554573059,
+      "learning_rate": 0.0001,
+      "loss": 0.8004,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7617334703665639,
+      "grad_norm": 0.997398316860199,
+      "learning_rate": 0.0001,
+      "loss": 1.0492,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7618191161356629,
+      "grad_norm": 0.8708875179290771,
+      "learning_rate": 0.0001,
+      "loss": 0.878,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.49595075845718384,
+      "learning_rate": 0.0001,
+      "loss": 0.5575,
+      "step": 8896
+    },
+    {
+      "epoch": 0.761990407673861,
+      "grad_norm": 0.568340003490448,
+      "learning_rate": 0.0001,
+      "loss": 0.9061,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7620760534429599,
+      "grad_norm": 1.0088541507720947,
+      "learning_rate": 0.0001,
+      "loss": 1.9449,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7621616992120589,
+      "grad_norm": 0.5163686871528625,
+      "learning_rate": 0.0001,
+      "loss": 0.8487,
+      "step": 8899
+    },
+    {
+      "epoch": 0.762247344981158,
+      "grad_norm": 0.6535877585411072,
+      "learning_rate": 0.0001,
+      "loss": 0.5766,
+      "step": 8900
+    },
+    {
+      "epoch": 0.762332990750257,
+      "grad_norm": 0.8741235136985779,
+      "learning_rate": 0.0001,
+      "loss": 1.1716,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7624186365193559,
+      "grad_norm": 1.15413498878479,
+      "learning_rate": 0.0001,
+      "loss": 1.2029,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7625042822884549,
+      "grad_norm": 0.45298102498054504,
+      "learning_rate": 0.0001,
+      "loss": 0.0977,
+      "step": 8903
+    },
+    {
+      "epoch": 0.762589928057554,
+      "grad_norm": 1.1900149583816528,
+      "learning_rate": 0.0001,
+      "loss": 0.8868,
+      "step": 8904
+    },
+    {
+      "epoch": 0.762675573826653,
+      "grad_norm": 0.8650181293487549,
+      "learning_rate": 0.0001,
+      "loss": 1.0409,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7627612195957519,
+      "grad_norm": 1.0183863639831543,
+      "learning_rate": 0.0001,
+      "loss": 0.9721,
+      "step": 8906
+    },
+    {
+      "epoch": 0.762846865364851,
+      "grad_norm": 1.118916392326355,
+      "learning_rate": 0.0001,
+      "loss": 1.1513,
+      "step": 8907
+    },
+    {
+      "epoch": 0.76293251113395,
+      "grad_norm": 1.1490775346755981,
+      "learning_rate": 0.0001,
+      "loss": 1.3021,
+      "step": 8908
+    },
+    {
+      "epoch": 0.763018156903049,
+      "grad_norm": 0.6929804682731628,
+      "learning_rate": 0.0001,
+      "loss": 0.8818,
+      "step": 8909
+    },
+    {
+      "epoch": 0.763103802672148,
+      "grad_norm": 0.5730358958244324,
+      "learning_rate": 0.0001,
+      "loss": 0.9293,
+      "step": 8910
+    },
+    {
+      "epoch": 0.763189448441247,
+      "grad_norm": 1.1385064125061035,
+      "learning_rate": 0.0001,
+      "loss": 1.3773,
+      "step": 8911
+    },
+    {
+      "epoch": 0.763275094210346,
+      "grad_norm": 0.8587868213653564,
+      "learning_rate": 0.0001,
+      "loss": 1.0949,
+      "step": 8912
+    },
+    {
+      "epoch": 0.763360739979445,
+      "grad_norm": 1.0367546081542969,
+      "learning_rate": 0.0001,
+      "loss": 0.5514,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7634463857485441,
+      "grad_norm": 0.6199944615364075,
+      "learning_rate": 0.0001,
+      "loss": 1.0758,
+      "step": 8914
+    },
+    {
+      "epoch": 0.763532031517643,
+      "grad_norm": 0.7042656540870667,
+      "learning_rate": 0.0001,
+      "loss": 1.0026,
+      "step": 8915
+    },
+    {
+      "epoch": 0.763617677286742,
+      "grad_norm": 0.4609256684780121,
+      "learning_rate": 0.0001,
+      "loss": 0.7146,
+      "step": 8916
+    },
+    {
+      "epoch": 0.763703323055841,
+      "grad_norm": 0.5424200296401978,
+      "learning_rate": 0.0001,
+      "loss": 0.433,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7637889688249401,
+      "grad_norm": 0.7561718821525574,
+      "learning_rate": 0.0001,
+      "loss": 1.2728,
+      "step": 8918
+    },
+    {
+      "epoch": 0.763874614594039,
+      "grad_norm": 0.7193705439567566,
+      "learning_rate": 0.0001,
+      "loss": 1.2133,
+      "step": 8919
+    },
+    {
+      "epoch": 0.763960260363138,
+      "grad_norm": 0.6188835501670837,
+      "learning_rate": 0.0001,
+      "loss": 0.4972,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7640459061322371,
+      "grad_norm": 0.9731792211532593,
+      "learning_rate": 0.0001,
+      "loss": 1.2196,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7641315519013361,
+      "grad_norm": 2.419196128845215,
+      "learning_rate": 0.0001,
+      "loss": 2.3535,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7642171976704351,
+      "grad_norm": 0.3605031371116638,
+      "learning_rate": 0.0001,
+      "loss": 0.1,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7643028434395341,
+      "grad_norm": 0.3909345269203186,
+      "learning_rate": 0.0001,
+      "loss": 0.509,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7643884892086331,
+      "grad_norm": 0.7393956780433655,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7644741349777321,
+      "grad_norm": 0.8747111558914185,
+      "learning_rate": 0.0001,
+      "loss": 1.2694,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7645597807468311,
+      "grad_norm": 0.8537681698799133,
+      "learning_rate": 0.0001,
+      "loss": 0.9891,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7646454265159301,
+      "grad_norm": 0.8762173652648926,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7647310722850291,
+      "grad_norm": 0.31992313265800476,
+      "learning_rate": 0.0001,
+      "loss": 0.5238,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7648167180541281,
+      "grad_norm": 1.6843622922897339,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7649023638232272,
+      "grad_norm": 1.0075702667236328,
+      "learning_rate": 0.0001,
+      "loss": 0.7268,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7649880095923262,
+      "grad_norm": 0.9802848696708679,
+      "learning_rate": 0.0001,
+      "loss": 0.7066,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7650736553614251,
+      "grad_norm": 0.7549925446510315,
+      "learning_rate": 0.0001,
+      "loss": 0.5849,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7651593011305241,
+      "grad_norm": 0.6192092895507812,
+      "learning_rate": 0.0001,
+      "loss": 0.5565,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7652449468996232,
+      "grad_norm": 0.9994215369224548,
+      "learning_rate": 0.0001,
+      "loss": 1.0747,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7653305926687222,
+      "grad_norm": 1.0030579566955566,
+      "learning_rate": 0.0001,
+      "loss": 1.0494,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7654162384378211,
+      "grad_norm": 1.0066769123077393,
+      "learning_rate": 0.0001,
+      "loss": 0.7945,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7655018842069202,
+      "grad_norm": 1.1165063381195068,
+      "learning_rate": 0.0001,
+      "loss": 0.8311,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7655875299760192,
+      "grad_norm": 0.4356134533882141,
+      "learning_rate": 0.0001,
+      "loss": 0.5451,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7656731757451182,
+      "grad_norm": 0.595764696598053,
+      "learning_rate": 0.0001,
+      "loss": 0.5144,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7657588215142171,
+      "grad_norm": 0.4760047197341919,
+      "learning_rate": 0.0001,
+      "loss": 0.9787,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7658444672833162,
+      "grad_norm": 1.0060269832611084,
+      "learning_rate": 0.0001,
+      "loss": 1.2637,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7659301130524152,
+      "grad_norm": 0.6793139576911926,
+      "learning_rate": 0.0001,
+      "loss": 0.5794,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7660157588215142,
+      "grad_norm": 0.9032101631164551,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7661014045906133,
+      "grad_norm": 0.8795917630195618,
+      "learning_rate": 0.0001,
+      "loss": 1.1521,
+      "step": 8945
+    },
+    {
+      "epoch": 0.7661870503597122,
+      "grad_norm": 0.6995624303817749,
+      "learning_rate": 0.0001,
+      "loss": 1.1493,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7662726961288112,
+      "grad_norm": 0.5586099624633789,
+      "learning_rate": 0.0001,
+      "loss": 1.2469,
+      "step": 8947
+    },
+    {
+      "epoch": 0.7663583418979102,
+      "grad_norm": 2.0854711532592773,
+      "learning_rate": 0.0001,
+      "loss": 0.5299,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7664439876670093,
+      "grad_norm": 0.6823059320449829,
+      "learning_rate": 0.0001,
+      "loss": 1.1836,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7665296334361082,
+      "grad_norm": 0.8214395642280579,
+      "learning_rate": 0.0001,
+      "loss": 0.7939,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7666152792052072,
+      "grad_norm": 0.6766270399093628,
+      "learning_rate": 0.0001,
+      "loss": 1.2668,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7667009249743063,
+      "grad_norm": 1.06308114528656,
+      "learning_rate": 0.0001,
+      "loss": 1.1896,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7667865707434053,
+      "grad_norm": 2.007598638534546,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 8953
+    },
+    {
+      "epoch": 0.7668722165125043,
+      "grad_norm": 0.8292940855026245,
+      "learning_rate": 0.0001,
+      "loss": 1.3207,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7669578622816033,
+      "grad_norm": 0.4174482524394989,
+      "learning_rate": 0.0001,
+      "loss": 0.4795,
+      "step": 8955
+    },
+    {
+      "epoch": 0.7670435080507023,
+      "grad_norm": 1.0807329416275024,
+      "learning_rate": 0.0001,
+      "loss": 1.2112,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7671291538198013,
+      "grad_norm": 0.808191180229187,
+      "learning_rate": 0.0001,
+      "loss": 0.9416,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7672147995889003,
+      "grad_norm": 0.811796247959137,
+      "learning_rate": 0.0001,
+      "loss": 0.9343,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7673004453579994,
+      "grad_norm": 1.2420172691345215,
+      "learning_rate": 0.0001,
+      "loss": 1.4033,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7673860911270983,
+      "grad_norm": 1.0000157356262207,
+      "learning_rate": 0.0001,
+      "loss": 0.6263,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7674717368961973,
+      "grad_norm": 1.1116094589233398,
+      "learning_rate": 0.0001,
+      "loss": 0.6595,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7675573826652964,
+      "grad_norm": 0.3300705552101135,
+      "learning_rate": 0.0001,
+      "loss": 0.4502,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7676430284343954,
+      "grad_norm": 0.7388720512390137,
+      "learning_rate": 0.0001,
+      "loss": 1.1786,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7677286742034943,
+      "grad_norm": 1.1063454151153564,
+      "learning_rate": 0.0001,
+      "loss": 0.8385,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7678143199725933,
+      "grad_norm": 0.9206770062446594,
+      "learning_rate": 0.0001,
+      "loss": 1.2284,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7678999657416924,
+      "grad_norm": 1.098550796508789,
+      "learning_rate": 0.0001,
+      "loss": 1.4011,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7679856115107914,
+      "grad_norm": 0.9100479483604431,
+      "learning_rate": 0.0001,
+      "loss": 1.0534,
+      "step": 8967
+    },
+    {
+      "epoch": 0.7680712572798903,
+      "grad_norm": 1.4717662334442139,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7681569030489894,
+      "grad_norm": 0.7530730366706848,
+      "learning_rate": 0.0001,
+      "loss": 1.0611,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7682425488180884,
+      "grad_norm": 1.1312845945358276,
+      "learning_rate": 0.0001,
+      "loss": 0.4636,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7683281945871874,
+      "grad_norm": 0.6903303265571594,
+      "learning_rate": 0.0001,
+      "loss": 1.234,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7684138403562863,
+      "grad_norm": 0.8970790505409241,
+      "learning_rate": 0.0001,
+      "loss": 0.593,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7684994861253854,
+      "grad_norm": 0.8317780494689941,
+      "learning_rate": 0.0001,
+      "loss": 0.6574,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7685851318944844,
+      "grad_norm": 1.1199860572814941,
+      "learning_rate": 0.0001,
+      "loss": 1.1946,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7686707776635834,
+      "grad_norm": 0.8499816656112671,
+      "learning_rate": 0.0001,
+      "loss": 0.2474,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7687564234326825,
+      "grad_norm": 0.9903639554977417,
+      "learning_rate": 0.0001,
+      "loss": 1.0846,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7688420692017814,
+      "grad_norm": 0.6057581305503845,
+      "learning_rate": 0.0001,
+      "loss": 1.0958,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7689277149708804,
+      "grad_norm": 1.2586557865142822,
+      "learning_rate": 0.0001,
+      "loss": 0.9118,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7690133607399794,
+      "grad_norm": 0.5316813588142395,
+      "learning_rate": 0.0001,
+      "loss": 0.7957,
+      "step": 8979
+    },
+    {
+      "epoch": 0.7690990065090785,
+      "grad_norm": 0.7293710708618164,
+      "learning_rate": 0.0001,
+      "loss": 0.6133,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7691846522781774,
+      "grad_norm": 0.5648038387298584,
+      "learning_rate": 0.0001,
+      "loss": 1.1049,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7692702980472764,
+      "grad_norm": 0.6023947596549988,
+      "learning_rate": 0.0001,
+      "loss": 0.9728,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7693559438163755,
+      "grad_norm": 1.0498764514923096,
+      "learning_rate": 0.0001,
+      "loss": 1.1772,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7694415895854745,
+      "grad_norm": 0.8606389760971069,
+      "learning_rate": 0.0001,
+      "loss": 1.099,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7695272353545735,
+      "grad_norm": 0.6241243481636047,
+      "learning_rate": 0.0001,
+      "loss": 0.8755,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7696128811236725,
+      "grad_norm": 0.7730876803398132,
+      "learning_rate": 0.0001,
+      "loss": 0.2782,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7696985268927715,
+      "grad_norm": 0.9633994102478027,
+      "learning_rate": 0.0001,
+      "loss": 1.0943,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7697841726618705,
+      "grad_norm": 1.3142565488815308,
+      "learning_rate": 0.0001,
+      "loss": 1.1522,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7698698184309695,
+      "grad_norm": 0.9084510803222656,
+      "learning_rate": 0.0001,
+      "loss": 0.2371,
+      "step": 8989
+    },
+    {
+      "epoch": 0.7699554642000686,
+      "grad_norm": 0.7625862956047058,
+      "learning_rate": 0.0001,
+      "loss": 1.1557,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7700411099691675,
+      "grad_norm": 0.8233332633972168,
+      "learning_rate": 0.0001,
+      "loss": 0.7848,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7701267557382665,
+      "grad_norm": 1.045209288597107,
+      "learning_rate": 0.0001,
+      "loss": 0.7714,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7702124015073656,
+      "grad_norm": 0.6869642734527588,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7702980472764646,
+      "grad_norm": 1.298593282699585,
+      "learning_rate": 0.0001,
+      "loss": 1.3088,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7703836930455635,
+      "grad_norm": 1.0722566843032837,
+      "learning_rate": 0.0001,
+      "loss": 1.318,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7704693388146625,
+      "grad_norm": 1.4042974710464478,
+      "learning_rate": 0.0001,
+      "loss": 0.8753,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7705549845837616,
+      "grad_norm": 0.9357549548149109,
+      "learning_rate": 0.0001,
+      "loss": 0.9101,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7706406303528606,
+      "grad_norm": 0.6612019538879395,
+      "learning_rate": 0.0001,
+      "loss": 0.7036,
+      "step": 8998
+    },
+    {
+      "epoch": 0.7707262761219595,
+      "grad_norm": 0.5481436848640442,
+      "learning_rate": 0.0001,
+      "loss": 0.673,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7708119218910586,
+      "grad_norm": 0.8770986199378967,
+      "learning_rate": 0.0001,
+      "loss": 0.5928,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7708975676601576,
+      "grad_norm": 0.7615388035774231,
+      "learning_rate": 0.0001,
+      "loss": 0.9055,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7709832134292566,
+      "grad_norm": 0.9568910002708435,
+      "learning_rate": 0.0001,
+      "loss": 1.2449,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7710688591983555,
+      "grad_norm": 0.7819222211837769,
+      "learning_rate": 0.0001,
+      "loss": 0.9947,
+      "step": 9003
+    },
+    {
+      "epoch": 0.7711545049674546,
+      "grad_norm": 0.9567471146583557,
+      "learning_rate": 0.0001,
+      "loss": 0.2907,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7712401507365536,
+      "grad_norm": 0.4540629982948303,
+      "learning_rate": 0.0001,
+      "loss": 0.9278,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7713257965056526,
+      "grad_norm": 1.1361511945724487,
+      "learning_rate": 0.0001,
+      "loss": 1.0324,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7714114422747517,
+      "grad_norm": 0.7202365398406982,
+      "learning_rate": 0.0001,
+      "loss": 0.7174,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7714970880438506,
+      "grad_norm": 1.2446542978286743,
+      "learning_rate": 0.0001,
+      "loss": 1.0159,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7715827338129496,
+      "grad_norm": 0.652557909488678,
+      "learning_rate": 0.0001,
+      "loss": 0.5874,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7716683795820486,
+      "grad_norm": 0.9135733842849731,
+      "learning_rate": 0.0001,
+      "loss": 1.2048,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7717540253511477,
+      "grad_norm": 0.5747937560081482,
+      "learning_rate": 0.0001,
+      "loss": 1.0349,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7718396711202467,
+      "grad_norm": 0.8039659261703491,
+      "learning_rate": 0.0001,
+      "loss": 1.3479,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7719253168893456,
+      "grad_norm": 0.3300723135471344,
+      "learning_rate": 0.0001,
+      "loss": 0.2835,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7720109626584447,
+      "grad_norm": 0.930689811706543,
+      "learning_rate": 0.0001,
+      "loss": 0.8705,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7720966084275437,
+      "grad_norm": 0.3517017960548401,
+      "learning_rate": 0.0001,
+      "loss": 0.2508,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7721822541966427,
+      "grad_norm": 1.527122974395752,
+      "learning_rate": 0.0001,
+      "loss": 1.2176,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7722678999657417,
+      "grad_norm": 0.7418503761291504,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7723535457348407,
+      "grad_norm": 0.6952468752861023,
+      "learning_rate": 0.0001,
+      "loss": 0.982,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7724391915039397,
+      "grad_norm": 0.49495527148246765,
+      "learning_rate": 0.0001,
+      "loss": 0.906,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7725248372730387,
+      "grad_norm": 0.8605284690856934,
+      "learning_rate": 0.0001,
+      "loss": 0.9488,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7726104830421378,
+      "grad_norm": 1.0788204669952393,
+      "learning_rate": 0.0001,
+      "loss": 1.0302,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7726961288112367,
+      "grad_norm": 0.6234711408615112,
+      "learning_rate": 0.0001,
+      "loss": 0.2027,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7727817745803357,
+      "grad_norm": 0.5187007784843445,
+      "learning_rate": 0.0001,
+      "loss": 1.2109,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7728674203494348,
+      "grad_norm": 0.5934009552001953,
+      "learning_rate": 0.0001,
+      "loss": 0.7054,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7729530661185338,
+      "grad_norm": 1.0613605976104736,
+      "learning_rate": 0.0001,
+      "loss": 0.6038,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7730387118876327,
+      "grad_norm": 0.5378656983375549,
+      "learning_rate": 0.0001,
+      "loss": 0.866,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7731243576567317,
+      "grad_norm": 0.9564335346221924,
+      "learning_rate": 0.0001,
+      "loss": 1.1954,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7732100034258308,
+      "grad_norm": 1.1319708824157715,
+      "learning_rate": 0.0001,
+      "loss": 1.083,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7732956491949298,
+      "grad_norm": 1.0924018621444702,
+      "learning_rate": 0.0001,
+      "loss": 0.6968,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7733812949640287,
+      "grad_norm": 0.6115031838417053,
+      "learning_rate": 0.0001,
+      "loss": 0.852,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7734669407331278,
+      "grad_norm": 1.2335180044174194,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7735525865022268,
+      "grad_norm": 0.7374507188796997,
+      "learning_rate": 0.0001,
+      "loss": 1.2956,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7736382322713258,
+      "grad_norm": 0.8664005994796753,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7737238780404248,
+      "grad_norm": 0.718346118927002,
+      "learning_rate": 0.0001,
+      "loss": 0.9095,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7738095238095238,
+      "grad_norm": 1.0330454111099243,
+      "learning_rate": 0.0001,
+      "loss": 1.1739,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7738951695786228,
+      "grad_norm": 0.4079849421977997,
+      "learning_rate": 0.0001,
+      "loss": 0.1223,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7739808153477218,
+      "grad_norm": 0.9179007411003113,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7740664611168209,
+      "grad_norm": 1.1808298826217651,
+      "learning_rate": 0.0001,
+      "loss": 1.0096,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7741521068859198,
+      "grad_norm": 0.7598329186439514,
+      "learning_rate": 0.0001,
+      "loss": 1.1166,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7742377526550188,
+      "grad_norm": 1.3721084594726562,
+      "learning_rate": 0.0001,
+      "loss": 1.1586,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7743233984241179,
+      "grad_norm": 0.8431118130683899,
+      "learning_rate": 0.0001,
+      "loss": 0.9527,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7744090441932169,
+      "grad_norm": 0.8208003044128418,
+      "learning_rate": 0.0001,
+      "loss": 0.3956,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7744946899623159,
+      "grad_norm": 1.1520808935165405,
+      "learning_rate": 0.0001,
+      "loss": 1.3111,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7745803357314148,
+      "grad_norm": 0.670417070388794,
+      "learning_rate": 0.0001,
+      "loss": 1.1912,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7746659815005139,
+      "grad_norm": 0.5385125279426575,
+      "learning_rate": 0.0001,
+      "loss": 0.511,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7747516272696129,
+      "grad_norm": 0.390947163105011,
+      "learning_rate": 0.0001,
+      "loss": 0.2568,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7748372730387119,
+      "grad_norm": 0.8668069243431091,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 9047
+    },
+    {
+      "epoch": 0.774922918807811,
+      "grad_norm": 0.5920236110687256,
+      "learning_rate": 0.0001,
+      "loss": 0.9272,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7750085645769099,
+      "grad_norm": 0.8642544746398926,
+      "learning_rate": 0.0001,
+      "loss": 1.2612,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7750942103460089,
+      "grad_norm": 0.5689441561698914,
+      "learning_rate": 0.0001,
+      "loss": 0.8215,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7751798561151079,
+      "grad_norm": 1.0043381452560425,
+      "learning_rate": 0.0001,
+      "loss": 1.097,
+      "step": 9051
+    },
+    {
+      "epoch": 0.775265501884207,
+      "grad_norm": 0.6145971417427063,
+      "learning_rate": 0.0001,
+      "loss": 0.8586,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7753511476533059,
+      "grad_norm": 0.4461723268032074,
+      "learning_rate": 0.0001,
+      "loss": 0.2907,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7754367934224049,
+      "grad_norm": 0.6003821492195129,
+      "learning_rate": 0.0001,
+      "loss": 1.0699,
+      "step": 9054
+    },
+    {
+      "epoch": 0.775522439191504,
+      "grad_norm": 0.7251746654510498,
+      "learning_rate": 0.0001,
+      "loss": 1.3047,
+      "step": 9055
+    },
+    {
+      "epoch": 0.775608084960603,
+      "grad_norm": 0.797837495803833,
+      "learning_rate": 0.0001,
+      "loss": 1.0565,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7756937307297019,
+      "grad_norm": 0.6026554107666016,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7757793764988009,
+      "grad_norm": 1.1058429479599,
+      "learning_rate": 0.0001,
+      "loss": 1.2874,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7758650222679,
+      "grad_norm": 0.6564393639564514,
+      "learning_rate": 0.0001,
+      "loss": 1.0999,
+      "step": 9059
+    },
+    {
+      "epoch": 0.775950668036999,
+      "grad_norm": 0.7916784286499023,
+      "learning_rate": 0.0001,
+      "loss": 1.1567,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7760363138060979,
+      "grad_norm": 0.8841283917427063,
+      "learning_rate": 0.0001,
+      "loss": 0.6337,
+      "step": 9061
+    },
+    {
+      "epoch": 0.776121959575197,
+      "grad_norm": 0.8337303400039673,
+      "learning_rate": 0.0001,
+      "loss": 1.0146,
+      "step": 9062
+    },
+    {
+      "epoch": 0.776207605344296,
+      "grad_norm": 0.946527361869812,
+      "learning_rate": 0.0001,
+      "loss": 0.9121,
+      "step": 9063
+    },
+    {
+      "epoch": 0.776293251113395,
+      "grad_norm": 0.6776830554008484,
+      "learning_rate": 0.0001,
+      "loss": 0.6703,
+      "step": 9064
+    },
+    {
+      "epoch": 0.776378896882494,
+      "grad_norm": 0.6982309818267822,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 9065
+    },
+    {
+      "epoch": 0.776464542651593,
+      "grad_norm": 0.6173207759857178,
+      "learning_rate": 0.0001,
+      "loss": 0.762,
+      "step": 9066
+    },
+    {
+      "epoch": 0.776550188420692,
+      "grad_norm": 0.952536404132843,
+      "learning_rate": 0.0001,
+      "loss": 0.7728,
+      "step": 9067
+    },
+    {
+      "epoch": 0.776635834189791,
+      "grad_norm": 0.5579339265823364,
+      "learning_rate": 0.0001,
+      "loss": 0.8014,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7767214799588901,
+      "grad_norm": 1.253414511680603,
+      "learning_rate": 0.0001,
+      "loss": 0.8932,
+      "step": 9069
+    },
+    {
+      "epoch": 0.776807125727989,
+      "grad_norm": 0.5577089786529541,
+      "learning_rate": 0.0001,
+      "loss": 0.8358,
+      "step": 9070
+    },
+    {
+      "epoch": 0.776892771497088,
+      "grad_norm": 1.2037031650543213,
+      "learning_rate": 0.0001,
+      "loss": 1.3099,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7769784172661871,
+      "grad_norm": 0.8769517540931702,
+      "learning_rate": 0.0001,
+      "loss": 0.8956,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7770640630352861,
+      "grad_norm": 0.8501620888710022,
+      "learning_rate": 0.0001,
+      "loss": 0.9545,
+      "step": 9073
+    },
+    {
+      "epoch": 0.777149708804385,
+      "grad_norm": 0.6644009351730347,
+      "learning_rate": 0.0001,
+      "loss": 0.8653,
+      "step": 9074
+    },
+    {
+      "epoch": 0.777235354573484,
+      "grad_norm": 0.6840207576751709,
+      "learning_rate": 0.0001,
+      "loss": 0.4958,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7773210003425831,
+      "grad_norm": 1.1389522552490234,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7774066461116821,
+      "grad_norm": 0.9153192043304443,
+      "learning_rate": 0.0001,
+      "loss": 0.8437,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7774922918807811,
+      "grad_norm": 1.1012499332427979,
+      "learning_rate": 0.0001,
+      "loss": 1.1328,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7775779376498801,
+      "grad_norm": 0.8436275124549866,
+      "learning_rate": 0.0001,
+      "loss": 0.9412,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7776635834189791,
+      "grad_norm": 0.775981068611145,
+      "learning_rate": 0.0001,
+      "loss": 0.604,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7777492291880781,
+      "grad_norm": 0.9218137264251709,
+      "learning_rate": 0.0001,
+      "loss": 0.7789,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7778348749571771,
+      "grad_norm": 0.9170219302177429,
+      "learning_rate": 0.0001,
+      "loss": 0.8172,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7779205207262762,
+      "grad_norm": 0.49465975165367126,
+      "learning_rate": 0.0001,
+      "loss": 0.49,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7780061664953751,
+      "grad_norm": 2.1742606163024902,
+      "learning_rate": 0.0001,
+      "loss": 0.5639,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7780918122644741,
+      "grad_norm": 0.6232671141624451,
+      "learning_rate": 0.0001,
+      "loss": 0.7006,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7781774580335732,
+      "grad_norm": 1.049675703048706,
+      "learning_rate": 0.0001,
+      "loss": 1.1004,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7782631038026722,
+      "grad_norm": 0.6396573185920715,
+      "learning_rate": 0.0001,
+      "loss": 0.68,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7783487495717711,
+      "grad_norm": 0.5859378576278687,
+      "learning_rate": 0.0001,
+      "loss": 0.5717,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7784343953408701,
+      "grad_norm": 0.8782861232757568,
+      "learning_rate": 0.0001,
+      "loss": 1.0718,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7785200411099692,
+      "grad_norm": 0.5185028910636902,
+      "learning_rate": 0.0001,
+      "loss": 0.138,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7786056868790682,
+      "grad_norm": 0.956028401851654,
+      "learning_rate": 0.0001,
+      "loss": 0.3334,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7786913326481671,
+      "grad_norm": 0.8842917084693909,
+      "learning_rate": 0.0001,
+      "loss": 0.8849,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7787769784172662,
+      "grad_norm": 1.082543969154358,
+      "learning_rate": 0.0001,
+      "loss": 0.7618,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7788626241863652,
+      "grad_norm": 0.8359021544456482,
+      "learning_rate": 0.0001,
+      "loss": 0.7881,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7789482699554642,
+      "grad_norm": 0.6339659094810486,
+      "learning_rate": 0.0001,
+      "loss": 0.433,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7790339157245632,
+      "grad_norm": 0.9249748587608337,
+      "learning_rate": 0.0001,
+      "loss": 1.3087,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7791195614936622,
+      "grad_norm": 0.7864447236061096,
+      "learning_rate": 0.0001,
+      "loss": 0.6364,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7792052072627612,
+      "grad_norm": 0.7056111693382263,
+      "learning_rate": 0.0001,
+      "loss": 1.2422,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7792908530318602,
+      "grad_norm": 1.032351016998291,
+      "learning_rate": 0.0001,
+      "loss": 1.3774,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7793764988009593,
+      "grad_norm": 0.819115936756134,
+      "learning_rate": 0.0001,
+      "loss": 1.2029,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7794621445700582,
+      "grad_norm": 0.8477752208709717,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7795477903391572,
+      "grad_norm": 1.2246438264846802,
+      "learning_rate": 0.0001,
+      "loss": 0.536,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7796334361082563,
+      "grad_norm": 0.8788439035415649,
+      "learning_rate": 0.0001,
+      "loss": 0.5368,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7797190818773553,
+      "grad_norm": 1.0072373151779175,
+      "learning_rate": 0.0001,
+      "loss": 0.874,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7798047276464543,
+      "grad_norm": 0.8210771679878235,
+      "learning_rate": 0.0001,
+      "loss": 0.7155,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7798903734155532,
+      "grad_norm": 1.3685805797576904,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7799760191846523,
+      "grad_norm": 0.6154951453208923,
+      "learning_rate": 0.0001,
+      "loss": 0.749,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7800616649537513,
+      "grad_norm": 1.181451439857483,
+      "learning_rate": 0.0001,
+      "loss": 0.8934,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7801473107228503,
+      "grad_norm": 1.1274080276489258,
+      "learning_rate": 0.0001,
+      "loss": 1.1963,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7802329564919493,
+      "grad_norm": 1.0225180387496948,
+      "learning_rate": 0.0001,
+      "loss": 1.099,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7803186022610483,
+      "grad_norm": 0.8519769310951233,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7804042480301473,
+      "grad_norm": 1.3202964067459106,
+      "learning_rate": 0.0001,
+      "loss": 1.2288,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7804898937992463,
+      "grad_norm": 0.7238753437995911,
+      "learning_rate": 0.0001,
+      "loss": 0.4846,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7805755395683454,
+      "grad_norm": 0.5777471661567688,
+      "learning_rate": 0.0001,
+      "loss": 0.3866,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7806611853374443,
+      "grad_norm": 0.9367586374282837,
+      "learning_rate": 0.0001,
+      "loss": 0.7135,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7807468311065433,
+      "grad_norm": 0.513486385345459,
+      "learning_rate": 0.0001,
+      "loss": 0.5711,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7808324768756424,
+      "grad_norm": 1.2267060279846191,
+      "learning_rate": 0.0001,
+      "loss": 1.1765,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7809181226447414,
+      "grad_norm": 1.011247992515564,
+      "learning_rate": 0.0001,
+      "loss": 0.933,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7810037684138403,
+      "grad_norm": 0.6862307786941528,
+      "learning_rate": 0.0001,
+      "loss": 1.2726,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7810894141829393,
+      "grad_norm": 0.8816829323768616,
+      "learning_rate": 0.0001,
+      "loss": 1.209,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7811750599520384,
+      "grad_norm": 0.5746612548828125,
+      "learning_rate": 0.0001,
+      "loss": 0.1463,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7812607057211374,
+      "grad_norm": 0.8125324249267578,
+      "learning_rate": 0.0001,
+      "loss": 0.735,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7813463514902363,
+      "grad_norm": 1.0770437717437744,
+      "learning_rate": 0.0001,
+      "loss": 0.9835,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7814319972593354,
+      "grad_norm": 0.9809699654579163,
+      "learning_rate": 0.0001,
+      "loss": 0.628,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7815176430284344,
+      "grad_norm": 0.7328506708145142,
+      "learning_rate": 0.0001,
+      "loss": 0.2116,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7816032887975334,
+      "grad_norm": 0.324362188577652,
+      "learning_rate": 0.0001,
+      "loss": 0.0962,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7816889345666324,
+      "grad_norm": 0.8599445223808289,
+      "learning_rate": 0.0001,
+      "loss": 1.1138,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7817745803357314,
+      "grad_norm": 0.4697704017162323,
+      "learning_rate": 0.0001,
+      "loss": 0.596,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7818602261048304,
+      "grad_norm": 0.9717550873756409,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7819458718739294,
+      "grad_norm": 0.7822465896606445,
+      "learning_rate": 0.0001,
+      "loss": 0.4643,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7820315176430285,
+      "grad_norm": 0.7532967329025269,
+      "learning_rate": 0.0001,
+      "loss": 0.5536,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7821171634121274,
+      "grad_norm": 0.6987689733505249,
+      "learning_rate": 0.0001,
+      "loss": 1.1389,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7822028091812264,
+      "grad_norm": 0.7227880954742432,
+      "learning_rate": 0.0001,
+      "loss": 1.1676,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7822884549503255,
+      "grad_norm": 1.1163058280944824,
+      "learning_rate": 0.0001,
+      "loss": 1.1066,
+      "step": 9134
+    },
+    {
+      "epoch": 0.7823741007194245,
+      "grad_norm": 0.7016303539276123,
+      "learning_rate": 0.0001,
+      "loss": 0.7254,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7824597464885235,
+      "grad_norm": 1.0448086261749268,
+      "learning_rate": 0.0001,
+      "loss": 0.7904,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7825453922576224,
+      "grad_norm": 0.8734720349311829,
+      "learning_rate": 0.0001,
+      "loss": 1.2663,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7826310380267215,
+      "grad_norm": 0.6194837093353271,
+      "learning_rate": 0.0001,
+      "loss": 0.4655,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7827166837958205,
+      "grad_norm": 0.8710647225379944,
+      "learning_rate": 0.0001,
+      "loss": 0.9314,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7828023295649195,
+      "grad_norm": 0.8470144867897034,
+      "learning_rate": 0.0001,
+      "loss": 0.7934,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7828879753340185,
+      "grad_norm": 0.6607686281204224,
+      "learning_rate": 0.0001,
+      "loss": 0.5346,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7829736211031175,
+      "grad_norm": 0.6822025179862976,
+      "learning_rate": 0.0001,
+      "loss": 1.0536,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7830592668722165,
+      "grad_norm": 1.0144407749176025,
+      "learning_rate": 0.0001,
+      "loss": 0.7752,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7831449126413155,
+      "grad_norm": 1.2049707174301147,
+      "learning_rate": 0.0001,
+      "loss": 1.1202,
+      "step": 9144
+    },
+    {
+      "epoch": 0.7832305584104146,
+      "grad_norm": 1.1758092641830444,
+      "learning_rate": 0.0001,
+      "loss": 0.8378,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7833162041795135,
+      "grad_norm": 0.5928599834442139,
+      "learning_rate": 0.0001,
+      "loss": 0.74,
+      "step": 9146
+    },
+    {
+      "epoch": 0.7834018499486125,
+      "grad_norm": 0.6167176365852356,
+      "learning_rate": 0.0001,
+      "loss": 1.2328,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7834874957177116,
+      "grad_norm": 0.415114164352417,
+      "learning_rate": 0.0001,
+      "loss": 0.6605,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7835731414868106,
+      "grad_norm": 1.198518991470337,
+      "learning_rate": 0.0001,
+      "loss": 0.7437,
+      "step": 9149
+    },
+    {
+      "epoch": 0.7836587872559095,
+      "grad_norm": 2.458049774169922,
+      "learning_rate": 0.0001,
+      "loss": 0.8261,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7837444330250085,
+      "grad_norm": 0.6638389229774475,
+      "learning_rate": 0.0001,
+      "loss": 0.4003,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7838300787941076,
+      "grad_norm": 0.811120867729187,
+      "learning_rate": 0.0001,
+      "loss": 1.3702,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7839157245632066,
+      "grad_norm": 1.1010502576828003,
+      "learning_rate": 0.0001,
+      "loss": 0.9021,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7840013703323055,
+      "grad_norm": 0.8507602214813232,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 9154
+    },
+    {
+      "epoch": 0.7840870161014046,
+      "grad_norm": 0.9363976716995239,
+      "learning_rate": 0.0001,
+      "loss": 0.7809,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7841726618705036,
+      "grad_norm": 0.9753392338752747,
+      "learning_rate": 0.0001,
+      "loss": 1.1181,
+      "step": 9156
+    },
+    {
+      "epoch": 0.7842583076396026,
+      "grad_norm": 0.5291316509246826,
+      "learning_rate": 0.0001,
+      "loss": 0.8553,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7843439534087016,
+      "grad_norm": 0.565546452999115,
+      "learning_rate": 0.0001,
+      "loss": 0.3481,
+      "step": 9158
+    },
+    {
+      "epoch": 0.7844295991778006,
+      "grad_norm": 0.5946458578109741,
+      "learning_rate": 0.0001,
+      "loss": 0.8326,
+      "step": 9159
+    },
+    {
+      "epoch": 0.7845152449468996,
+      "grad_norm": 0.5208998322486877,
+      "learning_rate": 0.0001,
+      "loss": 0.7888,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7846008907159986,
+      "grad_norm": 0.6884639263153076,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 9161
+    },
+    {
+      "epoch": 0.7846865364850977,
+      "grad_norm": 1.0978007316589355,
+      "learning_rate": 0.0001,
+      "loss": 0.85,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7847721822541966,
+      "grad_norm": 0.8432996273040771,
+      "learning_rate": 0.0001,
+      "loss": 0.9526,
+      "step": 9163
+    },
+    {
+      "epoch": 0.7848578280232956,
+      "grad_norm": 0.8001319169998169,
+      "learning_rate": 0.0001,
+      "loss": 0.7653,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7849434737923947,
+      "grad_norm": 1.2483159303665161,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7850291195614937,
+      "grad_norm": 0.8569945693016052,
+      "learning_rate": 0.0001,
+      "loss": 0.6881,
+      "step": 9166
+    },
+    {
+      "epoch": 0.7851147653305927,
+      "grad_norm": 1.36465585231781,
+      "learning_rate": 0.0001,
+      "loss": 0.4196,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7852004110996916,
+      "grad_norm": 0.8098609447479248,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 9168
+    },
+    {
+      "epoch": 0.7852860568687907,
+      "grad_norm": 0.7548666000366211,
+      "learning_rate": 0.0001,
+      "loss": 0.5425,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7853717026378897,
+      "grad_norm": 0.9950742125511169,
+      "learning_rate": 0.0001,
+      "loss": 1.0521,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7854573484069887,
+      "grad_norm": 0.8175922632217407,
+      "learning_rate": 0.0001,
+      "loss": 1.131,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7855429941760877,
+      "grad_norm": 0.7653370499610901,
+      "learning_rate": 0.0001,
+      "loss": 0.1784,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7856286399451867,
+      "grad_norm": 0.782753586769104,
+      "learning_rate": 0.0001,
+      "loss": 0.9141,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.6653229594230652,
+      "learning_rate": 0.0001,
+      "loss": 1.0998,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7857999314833847,
+      "grad_norm": 0.7362860441207886,
+      "learning_rate": 0.0001,
+      "loss": 0.81,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7858855772524838,
+      "grad_norm": 0.8818892240524292,
+      "learning_rate": 0.0001,
+      "loss": 0.6042,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7859712230215827,
+      "grad_norm": 0.5779796242713928,
+      "learning_rate": 0.0001,
+      "loss": 0.9114,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7860568687906817,
+      "grad_norm": 1.1549943685531616,
+      "learning_rate": 0.0001,
+      "loss": 0.9234,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7861425145597808,
+      "grad_norm": 0.8270366787910461,
+      "learning_rate": 0.0001,
+      "loss": 1.0188,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7862281603288798,
+      "grad_norm": 0.7502653002738953,
+      "learning_rate": 0.0001,
+      "loss": 0.7964,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7863138060979787,
+      "grad_norm": 0.9547973871231079,
+      "learning_rate": 0.0001,
+      "loss": 0.9923,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7863994518670777,
+      "grad_norm": 0.715717077255249,
+      "learning_rate": 0.0001,
+      "loss": 0.6736,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7864850976361768,
+      "grad_norm": 0.7900360226631165,
+      "learning_rate": 0.0001,
+      "loss": 1.0249,
+      "step": 9183
+    },
+    {
+      "epoch": 0.7865707434052758,
+      "grad_norm": 0.8799462914466858,
+      "learning_rate": 0.0001,
+      "loss": 0.5539,
+      "step": 9184
+    },
+    {
+      "epoch": 0.7866563891743747,
+      "grad_norm": 0.6087498068809509,
+      "learning_rate": 0.0001,
+      "loss": 1.0476,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7867420349434738,
+      "grad_norm": 0.7647240161895752,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7868276807125728,
+      "grad_norm": 0.5852156281471252,
+      "learning_rate": 0.0001,
+      "loss": 0.3975,
+      "step": 9187
+    },
+    {
+      "epoch": 0.7869133264816718,
+      "grad_norm": 0.6524302363395691,
+      "learning_rate": 0.0001,
+      "loss": 1.1958,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7869989722507709,
+      "grad_norm": 1.5235655307769775,
+      "learning_rate": 0.0001,
+      "loss": 1.1692,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7870846180198698,
+      "grad_norm": 0.6188966035842896,
+      "learning_rate": 0.0001,
+      "loss": 0.8037,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7871702637889688,
+      "grad_norm": 0.3986698389053345,
+      "learning_rate": 0.0001,
+      "loss": 0.8493,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7872559095580678,
+      "grad_norm": 0.8116810917854309,
+      "learning_rate": 0.0001,
+      "loss": 0.5898,
+      "step": 9192
+    },
+    {
+      "epoch": 0.7873415553271669,
+      "grad_norm": 0.7446384429931641,
+      "learning_rate": 0.0001,
+      "loss": 1.0635,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7874272010962658,
+      "grad_norm": 0.7540643811225891,
+      "learning_rate": 0.0001,
+      "loss": 1.1352,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7875128468653648,
+      "grad_norm": 1.2299855947494507,
+      "learning_rate": 0.0001,
+      "loss": 1.1738,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7875984926344639,
+      "grad_norm": 1.0997116565704346,
+      "learning_rate": 0.0001,
+      "loss": 1.3347,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7876841384035629,
+      "grad_norm": 1.2803295850753784,
+      "learning_rate": 0.0001,
+      "loss": 1.0871,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7877697841726619,
+      "grad_norm": 0.7493914365768433,
+      "learning_rate": 0.0001,
+      "loss": 1.2947,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7878554299417608,
+      "grad_norm": 0.6469742655754089,
+      "learning_rate": 0.0001,
+      "loss": 0.4906,
+      "step": 9199
+    },
+    {
+      "epoch": 0.7879410757108599,
+      "grad_norm": 1.138870358467102,
+      "learning_rate": 0.0001,
+      "loss": 1.2281,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7880267214799589,
+      "grad_norm": 0.5234661102294922,
+      "learning_rate": 0.0001,
+      "loss": 0.6539,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7881123672490579,
+      "grad_norm": 0.7123279571533203,
+      "learning_rate": 0.0001,
+      "loss": 0.9705,
+      "step": 9202
+    },
+    {
+      "epoch": 0.788198013018157,
+      "grad_norm": 0.825966477394104,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 9203
+    },
+    {
+      "epoch": 0.7882836587872559,
+      "grad_norm": 0.8149266242980957,
+      "learning_rate": 0.0001,
+      "loss": 0.6649,
+      "step": 9204
+    },
+    {
+      "epoch": 0.7883693045563549,
+      "grad_norm": 0.5627367496490479,
+      "learning_rate": 0.0001,
+      "loss": 0.9548,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7884549503254539,
+      "grad_norm": 0.8428074717521667,
+      "learning_rate": 0.0001,
+      "loss": 0.7145,
+      "step": 9206
+    },
+    {
+      "epoch": 0.788540596094553,
+      "grad_norm": 0.5149638652801514,
+      "learning_rate": 0.0001,
+      "loss": 0.8926,
+      "step": 9207
+    },
+    {
+      "epoch": 0.7886262418636519,
+      "grad_norm": 1.2695016860961914,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 9208
+    },
+    {
+      "epoch": 0.7887118876327509,
+      "grad_norm": 0.5359033942222595,
+      "learning_rate": 0.0001,
+      "loss": 0.4032,
+      "step": 9209
+    },
+    {
+      "epoch": 0.78879753340185,
+      "grad_norm": 0.6121941804885864,
+      "learning_rate": 0.0001,
+      "loss": 0.6433,
+      "step": 9210
+    },
+    {
+      "epoch": 0.788883179170949,
+      "grad_norm": 0.6746660470962524,
+      "learning_rate": 0.0001,
+      "loss": 1.1057,
+      "step": 9211
+    },
+    {
+      "epoch": 0.7889688249400479,
+      "grad_norm": 0.45555755496025085,
+      "learning_rate": 0.0001,
+      "loss": 0.7164,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7890544707091469,
+      "grad_norm": 0.8651643395423889,
+      "learning_rate": 0.0001,
+      "loss": 0.9575,
+      "step": 9213
+    },
+    {
+      "epoch": 0.789140116478246,
+      "grad_norm": 1.2232475280761719,
+      "learning_rate": 0.0001,
+      "loss": 0.6927,
+      "step": 9214
+    },
+    {
+      "epoch": 0.789225762247345,
+      "grad_norm": 0.8460260629653931,
+      "learning_rate": 0.0001,
+      "loss": 0.8259,
+      "step": 9215
+    },
+    {
+      "epoch": 0.789311408016444,
+      "grad_norm": 1.050052523612976,
+      "learning_rate": 0.0001,
+      "loss": 0.963,
+      "step": 9216
+    },
+    {
+      "epoch": 0.789397053785543,
+      "grad_norm": 0.6556635499000549,
+      "learning_rate": 0.0001,
+      "loss": 0.8663,
+      "step": 9217
+    },
+    {
+      "epoch": 0.789482699554642,
+      "grad_norm": 0.4373672604560852,
+      "learning_rate": 0.0001,
+      "loss": 0.6031,
+      "step": 9218
+    },
+    {
+      "epoch": 0.789568345323741,
+      "grad_norm": 0.9273231625556946,
+      "learning_rate": 0.0001,
+      "loss": 1.0921,
+      "step": 9219
+    },
+    {
+      "epoch": 0.7896539910928401,
+      "grad_norm": 1.2778228521347046,
+      "learning_rate": 0.0001,
+      "loss": 1.2185,
+      "step": 9220
+    },
+    {
+      "epoch": 0.789739636861939,
+      "grad_norm": 1.0774083137512207,
+      "learning_rate": 0.0001,
+      "loss": 1.2399,
+      "step": 9221
+    },
+    {
+      "epoch": 0.789825282631038,
+      "grad_norm": 0.5288732051849365,
+      "learning_rate": 0.0001,
+      "loss": 0.8039,
+      "step": 9222
+    },
+    {
+      "epoch": 0.789910928400137,
+      "grad_norm": 0.550058901309967,
+      "learning_rate": 0.0001,
+      "loss": 0.3916,
+      "step": 9223
+    },
+    {
+      "epoch": 0.7899965741692361,
+      "grad_norm": 0.7014310359954834,
+      "learning_rate": 0.0001,
+      "loss": 0.6941,
+      "step": 9224
+    },
+    {
+      "epoch": 0.790082219938335,
+      "grad_norm": 0.8124582767486572,
+      "learning_rate": 0.0001,
+      "loss": 0.9226,
+      "step": 9225
+    },
+    {
+      "epoch": 0.790167865707434,
+      "grad_norm": 0.9061583876609802,
+      "learning_rate": 0.0001,
+      "loss": 1.0845,
+      "step": 9226
+    },
+    {
+      "epoch": 0.7902535114765331,
+      "grad_norm": 0.9885817170143127,
+      "learning_rate": 0.0001,
+      "loss": 0.969,
+      "step": 9227
+    },
+    {
+      "epoch": 0.7903391572456321,
+      "grad_norm": 0.575226902961731,
+      "learning_rate": 0.0001,
+      "loss": 0.3267,
+      "step": 9228
+    },
+    {
+      "epoch": 0.7904248030147311,
+      "grad_norm": 1.0643738508224487,
+      "learning_rate": 0.0001,
+      "loss": 0.9116,
+      "step": 9229
+    },
+    {
+      "epoch": 0.79051044878383,
+      "grad_norm": 0.634292483329773,
+      "learning_rate": 0.0001,
+      "loss": 1.0688,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7905960945529291,
+      "grad_norm": 0.7217761278152466,
+      "learning_rate": 0.0001,
+      "loss": 0.7423,
+      "step": 9231
+    },
+    {
+      "epoch": 0.7906817403220281,
+      "grad_norm": 1.085455060005188,
+      "learning_rate": 0.0001,
+      "loss": 0.7923,
+      "step": 9232
+    },
+    {
+      "epoch": 0.7907673860911271,
+      "grad_norm": 0.731594979763031,
+      "learning_rate": 0.0001,
+      "loss": 0.7704,
+      "step": 9233
+    },
+    {
+      "epoch": 0.7908530318602262,
+      "grad_norm": 0.9454121589660645,
+      "learning_rate": 0.0001,
+      "loss": 1.3516,
+      "step": 9234
+    },
+    {
+      "epoch": 0.7909386776293251,
+      "grad_norm": 1.1539521217346191,
+      "learning_rate": 0.0001,
+      "loss": 0.6055,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7910243233984241,
+      "grad_norm": 0.23425522446632385,
+      "learning_rate": 0.0001,
+      "loss": 0.5839,
+      "step": 9236
+    },
+    {
+      "epoch": 0.7911099691675231,
+      "grad_norm": 0.32574260234832764,
+      "learning_rate": 0.0001,
+      "loss": 0.3868,
+      "step": 9237
+    },
+    {
+      "epoch": 0.7911956149366222,
+      "grad_norm": 0.7312561273574829,
+      "learning_rate": 0.0001,
+      "loss": 0.3814,
+      "step": 9238
+    },
+    {
+      "epoch": 0.7912812607057211,
+      "grad_norm": 1.1265498399734497,
+      "learning_rate": 0.0001,
+      "loss": 1.169,
+      "step": 9239
+    },
+    {
+      "epoch": 0.7913669064748201,
+      "grad_norm": 0.8533796668052673,
+      "learning_rate": 0.0001,
+      "loss": 0.9305,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7914525522439192,
+      "grad_norm": 1.0646069049835205,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 9241
+    },
+    {
+      "epoch": 0.7915381980130182,
+      "grad_norm": 0.7244068384170532,
+      "learning_rate": 0.0001,
+      "loss": 0.6538,
+      "step": 9242
+    },
+    {
+      "epoch": 0.7916238437821171,
+      "grad_norm": 0.5753114223480225,
+      "learning_rate": 0.0001,
+      "loss": 0.3601,
+      "step": 9243
+    },
+    {
+      "epoch": 0.7917094895512161,
+      "grad_norm": 0.43291857838630676,
+      "learning_rate": 0.0001,
+      "loss": 0.4523,
+      "step": 9244
+    },
+    {
+      "epoch": 0.7917951353203152,
+      "grad_norm": 0.7284010052680969,
+      "learning_rate": 0.0001,
+      "loss": 0.7744,
+      "step": 9245
+    },
+    {
+      "epoch": 0.7918807810894142,
+      "grad_norm": 0.8317292332649231,
+      "learning_rate": 0.0001,
+      "loss": 1.2221,
+      "step": 9246
+    },
+    {
+      "epoch": 0.7919664268585132,
+      "grad_norm": 0.880259096622467,
+      "learning_rate": 0.0001,
+      "loss": 0.9898,
+      "step": 9247
+    },
+    {
+      "epoch": 0.7920520726276122,
+      "grad_norm": 1.1463444232940674,
+      "learning_rate": 0.0001,
+      "loss": 1.2603,
+      "step": 9248
+    },
+    {
+      "epoch": 0.7921377183967112,
+      "grad_norm": 0.7467766404151917,
+      "learning_rate": 0.0001,
+      "loss": 1.196,
+      "step": 9249
+    },
+    {
+      "epoch": 0.7922233641658102,
+      "grad_norm": 0.3943456709384918,
+      "learning_rate": 0.0001,
+      "loss": 0.383,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7923090099349093,
+      "grad_norm": 0.8821214437484741,
+      "learning_rate": 0.0001,
+      "loss": 1.0875,
+      "step": 9251
+    },
+    {
+      "epoch": 0.7923946557040082,
+      "grad_norm": 0.593226432800293,
+      "learning_rate": 0.0001,
+      "loss": 1.122,
+      "step": 9252
+    },
+    {
+      "epoch": 0.7924803014731072,
+      "grad_norm": 0.9141805768013,
+      "learning_rate": 0.0001,
+      "loss": 1.3298,
+      "step": 9253
+    },
+    {
+      "epoch": 0.7925659472422062,
+      "grad_norm": 0.5360312461853027,
+      "learning_rate": 0.0001,
+      "loss": 0.2444,
+      "step": 9254
+    },
+    {
+      "epoch": 0.7926515930113053,
+      "grad_norm": 0.8524233102798462,
+      "learning_rate": 0.0001,
+      "loss": 0.9461,
+      "step": 9255
+    },
+    {
+      "epoch": 0.7927372387804043,
+      "grad_norm": 0.6104497909545898,
+      "learning_rate": 0.0001,
+      "loss": 1.0722,
+      "step": 9256
+    },
+    {
+      "epoch": 0.7928228845495032,
+      "grad_norm": 0.7275055050849915,
+      "learning_rate": 0.0001,
+      "loss": 1.0335,
+      "step": 9257
+    },
+    {
+      "epoch": 0.7929085303186023,
+      "grad_norm": 0.9758222103118896,
+      "learning_rate": 0.0001,
+      "loss": 0.7167,
+      "step": 9258
+    },
+    {
+      "epoch": 0.7929941760877013,
+      "grad_norm": 0.5707628726959229,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 9259
+    },
+    {
+      "epoch": 0.7930798218568003,
+      "grad_norm": 1.1170086860656738,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7931654676258992,
+      "grad_norm": 0.6163632273674011,
+      "learning_rate": 0.0001,
+      "loss": 0.9701,
+      "step": 9261
+    },
+    {
+      "epoch": 0.7932511133949983,
+      "grad_norm": 0.5881780385971069,
+      "learning_rate": 0.0001,
+      "loss": 0.6968,
+      "step": 9262
+    },
+    {
+      "epoch": 0.7933367591640973,
+      "grad_norm": 1.5895323753356934,
+      "learning_rate": 0.0001,
+      "loss": 0.9092,
+      "step": 9263
+    },
+    {
+      "epoch": 0.7934224049331963,
+      "grad_norm": 0.6840298771858215,
+      "learning_rate": 0.0001,
+      "loss": 0.8339,
+      "step": 9264
+    },
+    {
+      "epoch": 0.7935080507022954,
+      "grad_norm": 0.9453288912773132,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 9265
+    },
+    {
+      "epoch": 0.7935936964713943,
+      "grad_norm": 1.0026741027832031,
+      "learning_rate": 0.0001,
+      "loss": 0.3363,
+      "step": 9266
+    },
+    {
+      "epoch": 0.7936793422404933,
+      "grad_norm": 0.9367688894271851,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 9267
+    },
+    {
+      "epoch": 0.7937649880095923,
+      "grad_norm": 0.6371034979820251,
+      "learning_rate": 0.0001,
+      "loss": 0.9547,
+      "step": 9268
+    },
+    {
+      "epoch": 0.7938506337786914,
+      "grad_norm": 0.5128031373023987,
+      "learning_rate": 0.0001,
+      "loss": 0.7553,
+      "step": 9269
+    },
+    {
+      "epoch": 0.7939362795477903,
+      "grad_norm": 0.6325673460960388,
+      "learning_rate": 0.0001,
+      "loss": 0.9532,
+      "step": 9270
+    },
+    {
+      "epoch": 0.7940219253168893,
+      "grad_norm": 0.5160966515541077,
+      "learning_rate": 0.0001,
+      "loss": 0.5027,
+      "step": 9271
+    },
+    {
+      "epoch": 0.7941075710859884,
+      "grad_norm": 1.0368915796279907,
+      "learning_rate": 0.0001,
+      "loss": 0.861,
+      "step": 9272
+    },
+    {
+      "epoch": 0.7941932168550874,
+      "grad_norm": 1.3006367683410645,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 9273
+    },
+    {
+      "epoch": 0.7942788626241863,
+      "grad_norm": 0.6021485328674316,
+      "learning_rate": 0.0001,
+      "loss": 0.4241,
+      "step": 9274
+    },
+    {
+      "epoch": 0.7943645083932853,
+      "grad_norm": 0.727392315864563,
+      "learning_rate": 0.0001,
+      "loss": 1.4208,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7944501541623844,
+      "grad_norm": 0.7428408265113831,
+      "learning_rate": 0.0001,
+      "loss": 1.1912,
+      "step": 9276
+    },
+    {
+      "epoch": 0.7945357999314834,
+      "grad_norm": 0.7827781438827515,
+      "learning_rate": 0.0001,
+      "loss": 0.9326,
+      "step": 9277
+    },
+    {
+      "epoch": 0.7946214457005824,
+      "grad_norm": 0.8291465640068054,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 9278
+    },
+    {
+      "epoch": 0.7947070914696814,
+      "grad_norm": 0.9961708188056946,
+      "learning_rate": 0.0001,
+      "loss": 1.0057,
+      "step": 9279
+    },
+    {
+      "epoch": 0.7947927372387804,
+      "grad_norm": 1.115055799484253,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7948783830078794,
+      "grad_norm": 0.831710159778595,
+      "learning_rate": 0.0001,
+      "loss": 1.2643,
+      "step": 9281
+    },
+    {
+      "epoch": 0.7949640287769785,
+      "grad_norm": 0.8608905076980591,
+      "learning_rate": 0.0001,
+      "loss": 1.1466,
+      "step": 9282
+    },
+    {
+      "epoch": 0.7950496745460774,
+      "grad_norm": 1.06439208984375,
+      "learning_rate": 0.0001,
+      "loss": 0.8862,
+      "step": 9283
+    },
+    {
+      "epoch": 0.7951353203151764,
+      "grad_norm": 1.3833858966827393,
+      "learning_rate": 0.0001,
+      "loss": 1.2063,
+      "step": 9284
+    },
+    {
+      "epoch": 0.7952209660842754,
+      "grad_norm": 0.9574213624000549,
+      "learning_rate": 0.0001,
+      "loss": 1.0298,
+      "step": 9285
+    },
+    {
+      "epoch": 0.7953066118533745,
+      "grad_norm": 1.7159619331359863,
+      "learning_rate": 0.0001,
+      "loss": 0.9307,
+      "step": 9286
+    },
+    {
+      "epoch": 0.7953922576224735,
+      "grad_norm": 0.8310466408729553,
+      "learning_rate": 0.0001,
+      "loss": 0.7082,
+      "step": 9287
+    },
+    {
+      "epoch": 0.7954779033915724,
+      "grad_norm": 0.9463576674461365,
+      "learning_rate": 0.0001,
+      "loss": 1.1368,
+      "step": 9288
+    },
+    {
+      "epoch": 0.7955635491606715,
+      "grad_norm": 0.9926431775093079,
+      "learning_rate": 0.0001,
+      "loss": 0.5942,
+      "step": 9289
+    },
+    {
+      "epoch": 0.7956491949297705,
+      "grad_norm": 0.6566421389579773,
+      "learning_rate": 0.0001,
+      "loss": 0.9521,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7957348406988695,
+      "grad_norm": 0.907111406326294,
+      "learning_rate": 0.0001,
+      "loss": 0.9441,
+      "step": 9291
+    },
+    {
+      "epoch": 0.7958204864679684,
+      "grad_norm": 0.8071328401565552,
+      "learning_rate": 0.0001,
+      "loss": 1.244,
+      "step": 9292
+    },
+    {
+      "epoch": 0.7959061322370675,
+      "grad_norm": 0.7465983033180237,
+      "learning_rate": 0.0001,
+      "loss": 0.4819,
+      "step": 9293
+    },
+    {
+      "epoch": 0.7959917780061665,
+      "grad_norm": 1.0641295909881592,
+      "learning_rate": 0.0001,
+      "loss": 0.9522,
+      "step": 9294
+    },
+    {
+      "epoch": 0.7960774237752655,
+      "grad_norm": 0.7912877798080444,
+      "learning_rate": 0.0001,
+      "loss": 1.1192,
+      "step": 9295
+    },
+    {
+      "epoch": 0.7961630695443646,
+      "grad_norm": 1.1506074666976929,
+      "learning_rate": 0.0001,
+      "loss": 0.9795,
+      "step": 9296
+    },
+    {
+      "epoch": 0.7962487153134635,
+      "grad_norm": 1.0386265516281128,
+      "learning_rate": 0.0001,
+      "loss": 0.5985,
+      "step": 9297
+    },
+    {
+      "epoch": 0.7963343610825625,
+      "grad_norm": 0.9681318998336792,
+      "learning_rate": 0.0001,
+      "loss": 1.0167,
+      "step": 9298
+    },
+    {
+      "epoch": 0.7964200068516615,
+      "grad_norm": 0.9618659615516663,
+      "learning_rate": 0.0001,
+      "loss": 0.878,
+      "step": 9299
+    },
+    {
+      "epoch": 0.7965056526207606,
+      "grad_norm": 0.7406155467033386,
+      "learning_rate": 0.0001,
+      "loss": 0.9016,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7965912983898595,
+      "grad_norm": 0.6448941826820374,
+      "learning_rate": 0.0001,
+      "loss": 0.6587,
+      "step": 9301
+    },
+    {
+      "epoch": 0.7966769441589585,
+      "grad_norm": 0.6186447143554688,
+      "learning_rate": 0.0001,
+      "loss": 0.468,
+      "step": 9302
+    },
+    {
+      "epoch": 0.7967625899280576,
+      "grad_norm": 0.5937578082084656,
+      "learning_rate": 0.0001,
+      "loss": 0.1896,
+      "step": 9303
+    },
+    {
+      "epoch": 0.7968482356971566,
+      "grad_norm": 1.1056606769561768,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 9304
+    },
+    {
+      "epoch": 0.7969338814662555,
+      "grad_norm": 0.8510629534721375,
+      "learning_rate": 0.0001,
+      "loss": 1.0373,
+      "step": 9305
+    },
+    {
+      "epoch": 0.7970195272353545,
+      "grad_norm": 1.0637047290802002,
+      "learning_rate": 0.0001,
+      "loss": 1.1739,
+      "step": 9306
+    },
+    {
+      "epoch": 0.7971051730044536,
+      "grad_norm": 0.9385107755661011,
+      "learning_rate": 0.0001,
+      "loss": 0.935,
+      "step": 9307
+    },
+    {
+      "epoch": 0.7971908187735526,
+      "grad_norm": 0.8358237743377686,
+      "learning_rate": 0.0001,
+      "loss": 1.0928,
+      "step": 9308
+    },
+    {
+      "epoch": 0.7972764645426516,
+      "grad_norm": 0.5300450325012207,
+      "learning_rate": 0.0001,
+      "loss": 0.6958,
+      "step": 9309
+    },
+    {
+      "epoch": 0.7973621103117506,
+      "grad_norm": 0.5753965973854065,
+      "learning_rate": 0.0001,
+      "loss": 0.8596,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7974477560808496,
+      "grad_norm": 1.2533162832260132,
+      "learning_rate": 0.0001,
+      "loss": 0.7275,
+      "step": 9311
+    },
+    {
+      "epoch": 0.7975334018499486,
+      "grad_norm": 0.7596814036369324,
+      "learning_rate": 0.0001,
+      "loss": 1.4438,
+      "step": 9312
+    },
+    {
+      "epoch": 0.7976190476190477,
+      "grad_norm": 0.65901780128479,
+      "learning_rate": 0.0001,
+      "loss": 0.6327,
+      "step": 9313
+    },
+    {
+      "epoch": 0.7977046933881466,
+      "grad_norm": 0.8698801398277283,
+      "learning_rate": 0.0001,
+      "loss": 1.1706,
+      "step": 9314
+    },
+    {
+      "epoch": 0.7977903391572456,
+      "grad_norm": 1.0370961427688599,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 9315
+    },
+    {
+      "epoch": 0.7978759849263446,
+      "grad_norm": 0.8425692915916443,
+      "learning_rate": 0.0001,
+      "loss": 0.6576,
+      "step": 9316
+    },
+    {
+      "epoch": 0.7979616306954437,
+      "grad_norm": 1.1103529930114746,
+      "learning_rate": 0.0001,
+      "loss": 1.0416,
+      "step": 9317
+    },
+    {
+      "epoch": 0.7980472764645427,
+      "grad_norm": 0.9759340882301331,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 9318
+    },
+    {
+      "epoch": 0.7981329222336416,
+      "grad_norm": 0.7674329280853271,
+      "learning_rate": 0.0001,
+      "loss": 0.8129,
+      "step": 9319
+    },
+    {
+      "epoch": 0.7982185680027407,
+      "grad_norm": 0.551967978477478,
+      "learning_rate": 0.0001,
+      "loss": 0.9349,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7983042137718397,
+      "grad_norm": 1.1311932802200317,
+      "learning_rate": 0.0001,
+      "loss": 1.4196,
+      "step": 9321
+    },
+    {
+      "epoch": 0.7983898595409387,
+      "grad_norm": 0.5335666537284851,
+      "learning_rate": 0.0001,
+      "loss": 0.5315,
+      "step": 9322
+    },
+    {
+      "epoch": 0.7984755053100376,
+      "grad_norm": 0.8338903784751892,
+      "learning_rate": 0.0001,
+      "loss": 0.7431,
+      "step": 9323
+    },
+    {
+      "epoch": 0.7985611510791367,
+      "grad_norm": 0.9770576357841492,
+      "learning_rate": 0.0001,
+      "loss": 0.5737,
+      "step": 9324
+    },
+    {
+      "epoch": 0.7986467968482357,
+      "grad_norm": 0.8570345640182495,
+      "learning_rate": 0.0001,
+      "loss": 0.5376,
+      "step": 9325
+    },
+    {
+      "epoch": 0.7987324426173347,
+      "grad_norm": 0.7541743516921997,
+      "learning_rate": 0.0001,
+      "loss": 1.2935,
+      "step": 9326
+    },
+    {
+      "epoch": 0.7988180883864338,
+      "grad_norm": 0.6369880437850952,
+      "learning_rate": 0.0001,
+      "loss": 0.675,
+      "step": 9327
+    },
+    {
+      "epoch": 0.7989037341555327,
+      "grad_norm": 0.6600217223167419,
+      "learning_rate": 0.0001,
+      "loss": 0.4597,
+      "step": 9328
+    },
+    {
+      "epoch": 0.7989893799246317,
+      "grad_norm": 0.9663341045379639,
+      "learning_rate": 0.0001,
+      "loss": 0.8306,
+      "step": 9329
+    },
+    {
+      "epoch": 0.7990750256937307,
+      "grad_norm": 0.8296036124229431,
+      "learning_rate": 0.0001,
+      "loss": 0.8679,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7991606714628298,
+      "grad_norm": 0.46340513229370117,
+      "learning_rate": 0.0001,
+      "loss": 0.8141,
+      "step": 9331
+    },
+    {
+      "epoch": 0.7992463172319287,
+      "grad_norm": 1.1405445337295532,
+      "learning_rate": 0.0001,
+      "loss": 1.2577,
+      "step": 9332
+    },
+    {
+      "epoch": 0.7993319630010277,
+      "grad_norm": 0.5362456440925598,
+      "learning_rate": 0.0001,
+      "loss": 0.8591,
+      "step": 9333
+    },
+    {
+      "epoch": 0.7994176087701268,
+      "grad_norm": 0.7578037977218628,
+      "learning_rate": 0.0001,
+      "loss": 0.5094,
+      "step": 9334
+    },
+    {
+      "epoch": 0.7995032545392258,
+      "grad_norm": 0.9015144109725952,
+      "learning_rate": 0.0001,
+      "loss": 1.2978,
+      "step": 9335
+    },
+    {
+      "epoch": 0.7995889003083247,
+      "grad_norm": 0.46168655157089233,
+      "learning_rate": 0.0001,
+      "loss": 0.8274,
+      "step": 9336
+    },
+    {
+      "epoch": 0.7996745460774237,
+      "grad_norm": 1.2259461879730225,
+      "learning_rate": 0.0001,
+      "loss": 1.2039,
+      "step": 9337
+    },
+    {
+      "epoch": 0.7997601918465228,
+      "grad_norm": 1.0499050617218018,
+      "learning_rate": 0.0001,
+      "loss": 0.8478,
+      "step": 9338
+    },
+    {
+      "epoch": 0.7998458376156218,
+      "grad_norm": 0.8064746260643005,
+      "learning_rate": 0.0001,
+      "loss": 1.345,
+      "step": 9339
+    },
+    {
+      "epoch": 0.7999314833847208,
+      "grad_norm": 1.2051513195037842,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8000171291538198,
+      "grad_norm": 1.1140934228897095,
+      "learning_rate": 0.0001,
+      "loss": 1.2013,
+      "step": 9341
+    },
+    {
+      "epoch": 0.8001027749229188,
+      "grad_norm": 0.9285039901733398,
+      "learning_rate": 0.0001,
+      "loss": 0.7646,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8001884206920178,
+      "grad_norm": 0.7240397930145264,
+      "learning_rate": 0.0001,
+      "loss": 1.0772,
+      "step": 9343
+    },
+    {
+      "epoch": 0.8002740664611169,
+      "grad_norm": 1.2033004760742188,
+      "learning_rate": 0.0001,
+      "loss": 0.7989,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8003597122302158,
+      "grad_norm": 0.9705157279968262,
+      "learning_rate": 0.0001,
+      "loss": 1.386,
+      "step": 9345
+    },
+    {
+      "epoch": 0.8004453579993148,
+      "grad_norm": 1.0045303106307983,
+      "learning_rate": 0.0001,
+      "loss": 0.8264,
+      "step": 9346
+    },
+    {
+      "epoch": 0.8005310037684138,
+      "grad_norm": 1.1338562965393066,
+      "learning_rate": 0.0001,
+      "loss": 0.8927,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8006166495375129,
+      "grad_norm": 0.8045526146888733,
+      "learning_rate": 0.0001,
+      "loss": 1.0313,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8007022953066119,
+      "grad_norm": 1.0017714500427246,
+      "learning_rate": 0.0001,
+      "loss": 0.6908,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8007879410757108,
+      "grad_norm": 0.8494471907615662,
+      "learning_rate": 0.0001,
+      "loss": 0.7663,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8008735868448099,
+      "grad_norm": 1.2732852697372437,
+      "learning_rate": 0.0001,
+      "loss": 1.4267,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8009592326139089,
+      "grad_norm": 1.1353853940963745,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8010448783830079,
+      "grad_norm": 0.9090720415115356,
+      "learning_rate": 0.0001,
+      "loss": 0.8949,
+      "step": 9353
+    },
+    {
+      "epoch": 0.8011305241521068,
+      "grad_norm": 0.7554200291633606,
+      "learning_rate": 0.0001,
+      "loss": 0.7709,
+      "step": 9354
+    },
+    {
+      "epoch": 0.8012161699212059,
+      "grad_norm": 0.7453703880310059,
+      "learning_rate": 0.0001,
+      "loss": 1.0473,
+      "step": 9355
+    },
+    {
+      "epoch": 0.8013018156903049,
+      "grad_norm": 1.0036020278930664,
+      "learning_rate": 0.0001,
+      "loss": 0.7609,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8013874614594039,
+      "grad_norm": 0.48324722051620483,
+      "learning_rate": 0.0001,
+      "loss": 0.3965,
+      "step": 9357
+    },
+    {
+      "epoch": 0.801473107228503,
+      "grad_norm": 0.6349208354949951,
+      "learning_rate": 0.0001,
+      "loss": 0.5415,
+      "step": 9358
+    },
+    {
+      "epoch": 0.8015587529976019,
+      "grad_norm": 0.7947945594787598,
+      "learning_rate": 0.0001,
+      "loss": 1.3123,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8016443987667009,
+      "grad_norm": 0.5706855654716492,
+      "learning_rate": 0.0001,
+      "loss": 0.8986,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8017300445357999,
+      "grad_norm": 1.3164361715316772,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 9361
+    },
+    {
+      "epoch": 0.801815690304899,
+      "grad_norm": 1.068043828010559,
+      "learning_rate": 0.0001,
+      "loss": 1.2903,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8019013360739979,
+      "grad_norm": 0.734844446182251,
+      "learning_rate": 0.0001,
+      "loss": 0.9567,
+      "step": 9363
+    },
+    {
+      "epoch": 0.8019869818430969,
+      "grad_norm": 0.9450734257698059,
+      "learning_rate": 0.0001,
+      "loss": 1.2378,
+      "step": 9364
+    },
+    {
+      "epoch": 0.802072627612196,
+      "grad_norm": 0.3666079044342041,
+      "learning_rate": 0.0001,
+      "loss": 0.4593,
+      "step": 9365
+    },
+    {
+      "epoch": 0.802158273381295,
+      "grad_norm": 0.849915623664856,
+      "learning_rate": 0.0001,
+      "loss": 0.6411,
+      "step": 9366
+    },
+    {
+      "epoch": 0.802243919150394,
+      "grad_norm": 0.931544303894043,
+      "learning_rate": 0.0001,
+      "loss": 0.8911,
+      "step": 9367
+    },
+    {
+      "epoch": 0.802329564919493,
+      "grad_norm": 1.204470157623291,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 9368
+    },
+    {
+      "epoch": 0.802415210688592,
+      "grad_norm": 0.9751660227775574,
+      "learning_rate": 0.0001,
+      "loss": 1.1545,
+      "step": 9369
+    },
+    {
+      "epoch": 0.802500856457691,
+      "grad_norm": 0.7843050360679626,
+      "learning_rate": 0.0001,
+      "loss": 1.2984,
+      "step": 9370
+    },
+    {
+      "epoch": 0.80258650222679,
+      "grad_norm": 0.9571824073791504,
+      "learning_rate": 0.0001,
+      "loss": 0.9397,
+      "step": 9371
+    },
+    {
+      "epoch": 0.802672147995889,
+      "grad_norm": 1.0273127555847168,
+      "learning_rate": 0.0001,
+      "loss": 1.3425,
+      "step": 9372
+    },
+    {
+      "epoch": 0.802757793764988,
+      "grad_norm": 0.7151604294776917,
+      "learning_rate": 0.0001,
+      "loss": 0.5881,
+      "step": 9373
+    },
+    {
+      "epoch": 0.802843439534087,
+      "grad_norm": 0.6232353448867798,
+      "learning_rate": 0.0001,
+      "loss": 0.7232,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8029290853031861,
+      "grad_norm": 1.0682258605957031,
+      "learning_rate": 0.0001,
+      "loss": 1.0698,
+      "step": 9375
+    },
+    {
+      "epoch": 0.803014731072285,
+      "grad_norm": 1.3501163721084595,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 9376
+    },
+    {
+      "epoch": 0.803100376841384,
+      "grad_norm": 0.7985648512840271,
+      "learning_rate": 0.0001,
+      "loss": 1.1528,
+      "step": 9377
+    },
+    {
+      "epoch": 0.803186022610483,
+      "grad_norm": 0.566157877445221,
+      "learning_rate": 0.0001,
+      "loss": 0.9356,
+      "step": 9378
+    },
+    {
+      "epoch": 0.8032716683795821,
+      "grad_norm": 1.1080700159072876,
+      "learning_rate": 0.0001,
+      "loss": 1.268,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8033573141486811,
+      "grad_norm": 0.6726261973381042,
+      "learning_rate": 0.0001,
+      "loss": 0.7468,
+      "step": 9380
+    },
+    {
+      "epoch": 0.80344295991778,
+      "grad_norm": 0.905696451663971,
+      "learning_rate": 0.0001,
+      "loss": 0.9918,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8035286056868791,
+      "grad_norm": 0.6901817321777344,
+      "learning_rate": 0.0001,
+      "loss": 1.0564,
+      "step": 9382
+    },
+    {
+      "epoch": 0.8036142514559781,
+      "grad_norm": 0.42380353808403015,
+      "learning_rate": 0.0001,
+      "loss": 0.1624,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8036998972250771,
+      "grad_norm": 1.0521661043167114,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 9384
+    },
+    {
+      "epoch": 0.803785542994176,
+      "grad_norm": 0.7640438675880432,
+      "learning_rate": 0.0001,
+      "loss": 1.0245,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8038711887632751,
+      "grad_norm": 0.638514518737793,
+      "learning_rate": 0.0001,
+      "loss": 1.0972,
+      "step": 9386
+    },
+    {
+      "epoch": 0.8039568345323741,
+      "grad_norm": 1.283604383468628,
+      "learning_rate": 0.0001,
+      "loss": 0.7756,
+      "step": 9387
+    },
+    {
+      "epoch": 0.8040424803014731,
+      "grad_norm": 0.9744076728820801,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8041281260705722,
+      "grad_norm": 0.7962610721588135,
+      "learning_rate": 0.0001,
+      "loss": 0.9871,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8042137718396711,
+      "grad_norm": 1.0669164657592773,
+      "learning_rate": 0.0001,
+      "loss": 1.7859,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8042994176087701,
+      "grad_norm": 0.941760778427124,
+      "learning_rate": 0.0001,
+      "loss": 0.9486,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8043850633778691,
+      "grad_norm": 0.5548285841941833,
+      "learning_rate": 0.0001,
+      "loss": 0.6419,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8044707091469682,
+      "grad_norm": 1.2042367458343506,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8045563549160671,
+      "grad_norm": 0.6941892504692078,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8046420006851661,
+      "grad_norm": 0.7118306756019592,
+      "learning_rate": 0.0001,
+      "loss": 0.8222,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8047276464542652,
+      "grad_norm": 1.8747124671936035,
+      "learning_rate": 0.0001,
+      "loss": 1.5485,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8048132922233642,
+      "grad_norm": 0.682378351688385,
+      "learning_rate": 0.0001,
+      "loss": 0.8492,
+      "step": 9397
+    },
+    {
+      "epoch": 0.8048989379924631,
+      "grad_norm": 1.0078529119491577,
+      "learning_rate": 0.0001,
+      "loss": 1.1353,
+      "step": 9398
+    },
+    {
+      "epoch": 0.8049845837615622,
+      "grad_norm": 0.4452633559703827,
+      "learning_rate": 0.0001,
+      "loss": 0.5413,
+      "step": 9399
+    },
+    {
+      "epoch": 0.8050702295306612,
+      "grad_norm": 0.8787591457366943,
+      "learning_rate": 0.0001,
+      "loss": 0.4494,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8051558752997602,
+      "grad_norm": 0.7992796301841736,
+      "learning_rate": 0.0001,
+      "loss": 1.312,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8052415210688592,
+      "grad_norm": 1.0605119466781616,
+      "learning_rate": 0.0001,
+      "loss": 1.1514,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8053271668379582,
+      "grad_norm": 0.980076253414154,
+      "learning_rate": 0.0001,
+      "loss": 1.1323,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8054128126070572,
+      "grad_norm": 0.5576825737953186,
+      "learning_rate": 0.0001,
+      "loss": 0.5656,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8054984583761562,
+      "grad_norm": 0.9291499257087708,
+      "learning_rate": 0.0001,
+      "loss": 1.1662,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8055841041452553,
+      "grad_norm": 0.8500435948371887,
+      "learning_rate": 0.0001,
+      "loss": 0.8144,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8056697499143542,
+      "grad_norm": 1.4124348163604736,
+      "learning_rate": 0.0001,
+      "loss": 1.3707,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8057553956834532,
+      "grad_norm": 0.6184667348861694,
+      "learning_rate": 0.0001,
+      "loss": 1.2171,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8058410414525522,
+      "grad_norm": 1.0558761358261108,
+      "learning_rate": 0.0001,
+      "loss": 1.2279,
+      "step": 9409
+    },
+    {
+      "epoch": 0.8059266872216513,
+      "grad_norm": 1.0584535598754883,
+      "learning_rate": 0.0001,
+      "loss": 0.9433,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8060123329907503,
+      "grad_norm": 0.9649565815925598,
+      "learning_rate": 0.0001,
+      "loss": 1.1482,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8060979787598492,
+      "grad_norm": 1.9515680074691772,
+      "learning_rate": 0.0001,
+      "loss": 0.921,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8061836245289483,
+      "grad_norm": 0.7637656331062317,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8062692702980473,
+      "grad_norm": 1.083452820777893,
+      "learning_rate": 0.0001,
+      "loss": 1.775,
+      "step": 9414
+    },
+    {
+      "epoch": 0.8063549160671463,
+      "grad_norm": 0.8531796336174011,
+      "learning_rate": 0.0001,
+      "loss": 1.0437,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8064405618362452,
+      "grad_norm": 0.7166566252708435,
+      "learning_rate": 0.0001,
+      "loss": 1.2408,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8065262076053443,
+      "grad_norm": 0.5535027384757996,
+      "learning_rate": 0.0001,
+      "loss": 0.8995,
+      "step": 9417
+    },
+    {
+      "epoch": 0.8066118533744433,
+      "grad_norm": 0.6290565729141235,
+      "learning_rate": 0.0001,
+      "loss": 0.2399,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8066974991435423,
+      "grad_norm": 1.1818490028381348,
+      "learning_rate": 0.0001,
+      "loss": 1.21,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8067831449126414,
+      "grad_norm": 0.7250170111656189,
+      "learning_rate": 0.0001,
+      "loss": 0.7274,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8068687906817403,
+      "grad_norm": 0.9543491005897522,
+      "learning_rate": 0.0001,
+      "loss": 1.0164,
+      "step": 9421
+    },
+    {
+      "epoch": 0.8069544364508393,
+      "grad_norm": 1.0616828203201294,
+      "learning_rate": 0.0001,
+      "loss": 1.214,
+      "step": 9422
+    },
+    {
+      "epoch": 0.8070400822199383,
+      "grad_norm": 0.5983845591545105,
+      "learning_rate": 0.0001,
+      "loss": 0.7556,
+      "step": 9423
+    },
+    {
+      "epoch": 0.8071257279890374,
+      "grad_norm": 0.9603384137153625,
+      "learning_rate": 0.0001,
+      "loss": 1.0864,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8072113737581363,
+      "grad_norm": 0.9437248706817627,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8072970195272353,
+      "grad_norm": 0.9980857372283936,
+      "learning_rate": 0.0001,
+      "loss": 0.6843,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8073826652963344,
+      "grad_norm": 0.8537907600402832,
+      "learning_rate": 0.0001,
+      "loss": 1.1432,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8074683110654334,
+      "grad_norm": 0.5671331882476807,
+      "learning_rate": 0.0001,
+      "loss": 0.7632,
+      "step": 9428
+    },
+    {
+      "epoch": 0.8075539568345323,
+      "grad_norm": 0.5932679176330566,
+      "learning_rate": 0.0001,
+      "loss": 0.8989,
+      "step": 9429
+    },
+    {
+      "epoch": 0.8076396026036314,
+      "grad_norm": 0.9075400233268738,
+      "learning_rate": 0.0001,
+      "loss": 1.2766,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8077252483727304,
+      "grad_norm": 1.0682244300842285,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 9431
+    },
+    {
+      "epoch": 0.8078108941418294,
+      "grad_norm": 1.3385237455368042,
+      "learning_rate": 0.0001,
+      "loss": 1.0303,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8078965399109284,
+      "grad_norm": 0.5919856429100037,
+      "learning_rate": 0.0001,
+      "loss": 0.8827,
+      "step": 9433
+    },
+    {
+      "epoch": 0.8079821856800274,
+      "grad_norm": 0.9512463808059692,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 9434
+    },
+    {
+      "epoch": 0.8080678314491264,
+      "grad_norm": 0.7418727278709412,
+      "learning_rate": 0.0001,
+      "loss": 1.0488,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8081534772182254,
+      "grad_norm": 1.1579939126968384,
+      "learning_rate": 0.0001,
+      "loss": 0.9636,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8082391229873245,
+      "grad_norm": 1.3758692741394043,
+      "learning_rate": 0.0001,
+      "loss": 1.0023,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8083247687564235,
+      "grad_norm": 0.3884992301464081,
+      "learning_rate": 0.0001,
+      "loss": 0.4463,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8084104145255224,
+      "grad_norm": 0.6601038575172424,
+      "learning_rate": 0.0001,
+      "loss": 1.1851,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8084960602946214,
+      "grad_norm": 0.6589560508728027,
+      "learning_rate": 0.0001,
+      "loss": 1.1166,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8085817060637205,
+      "grad_norm": 0.836138904094696,
+      "learning_rate": 0.0001,
+      "loss": 0.9047,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8086673518328195,
+      "grad_norm": 0.8542571067810059,
+      "learning_rate": 0.0001,
+      "loss": 0.7443,
+      "step": 9442
+    },
+    {
+      "epoch": 0.8087529976019184,
+      "grad_norm": 1.0118588209152222,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 9443
+    },
+    {
+      "epoch": 0.8088386433710175,
+      "grad_norm": 1.2326596975326538,
+      "learning_rate": 0.0001,
+      "loss": 1.2635,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8089242891401165,
+      "grad_norm": 0.7515220046043396,
+      "learning_rate": 0.0001,
+      "loss": 0.6367,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8090099349092155,
+      "grad_norm": 0.6307200193405151,
+      "learning_rate": 0.0001,
+      "loss": 0.4422,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8090955806783144,
+      "grad_norm": 0.5720088481903076,
+      "learning_rate": 0.0001,
+      "loss": 0.7184,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8091812264474135,
+      "grad_norm": 0.8794780969619751,
+      "learning_rate": 0.0001,
+      "loss": 0.797,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8092668722165125,
+      "grad_norm": 0.9479193687438965,
+      "learning_rate": 0.0001,
+      "loss": 1.377,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8093525179856115,
+      "grad_norm": 0.8476702570915222,
+      "learning_rate": 0.0001,
+      "loss": 0.6742,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8094381637547106,
+      "grad_norm": 1.4278607368469238,
+      "learning_rate": 0.0001,
+      "loss": 2.0329,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.8720712661743164,
+      "learning_rate": 0.0001,
+      "loss": 1.0569,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8096094552929085,
+      "grad_norm": 0.9206389784812927,
+      "learning_rate": 0.0001,
+      "loss": 0.842,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8096951010620075,
+      "grad_norm": 0.4452422261238098,
+      "learning_rate": 0.0001,
+      "loss": 0.6582,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8097807468311066,
+      "grad_norm": 1.3541351556777954,
+      "learning_rate": 0.0001,
+      "loss": 1.282,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8098663926002055,
+      "grad_norm": 0.4252638816833496,
+      "learning_rate": 0.0001,
+      "loss": 0.2087,
+      "step": 9456
+    },
+    {
+      "epoch": 0.8099520383693045,
+      "grad_norm": 0.5747126936912537,
+      "learning_rate": 0.0001,
+      "loss": 1.3519,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8100376841384036,
+      "grad_norm": 0.6949205994606018,
+      "learning_rate": 0.0001,
+      "loss": 1.0702,
+      "step": 9458
+    },
+    {
+      "epoch": 0.8101233299075026,
+      "grad_norm": 0.6453054547309875,
+      "learning_rate": 0.0001,
+      "loss": 0.5241,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8102089756766016,
+      "grad_norm": 1.0329124927520752,
+      "learning_rate": 0.0001,
+      "loss": 0.8352,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8102946214457006,
+      "grad_norm": 0.6610122323036194,
+      "learning_rate": 0.0001,
+      "loss": 0.9657,
+      "step": 9461
+    },
+    {
+      "epoch": 0.8103802672147996,
+      "grad_norm": 0.9872442483901978,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8104659129838986,
+      "grad_norm": 0.8182790875434875,
+      "learning_rate": 0.0001,
+      "loss": 1.3648,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8105515587529976,
+      "grad_norm": 0.647554874420166,
+      "learning_rate": 0.0001,
+      "loss": 0.9202,
+      "step": 9464
+    },
+    {
+      "epoch": 0.8106372045220966,
+      "grad_norm": 1.0223963260650635,
+      "learning_rate": 0.0001,
+      "loss": 0.9523,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8107228502911956,
+      "grad_norm": 0.8787111043930054,
+      "learning_rate": 0.0001,
+      "loss": 0.7623,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8108084960602946,
+      "grad_norm": 1.160585641860962,
+      "learning_rate": 0.0001,
+      "loss": 1.0459,
+      "step": 9467
+    },
+    {
+      "epoch": 0.8108941418293937,
+      "grad_norm": 0.643982470035553,
+      "learning_rate": 0.0001,
+      "loss": 0.2433,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8109797875984927,
+      "grad_norm": 1.1967406272888184,
+      "learning_rate": 0.0001,
+      "loss": 0.6687,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8110654333675916,
+      "grad_norm": 0.5259833335876465,
+      "learning_rate": 0.0001,
+      "loss": 1.1082,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8111510791366906,
+      "grad_norm": 0.7958382368087769,
+      "learning_rate": 0.0001,
+      "loss": 1.178,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8112367249057897,
+      "grad_norm": 0.9903834462165833,
+      "learning_rate": 0.0001,
+      "loss": 1.1063,
+      "step": 9472
+    },
+    {
+      "epoch": 0.8113223706748887,
+      "grad_norm": 0.7722984552383423,
+      "learning_rate": 0.0001,
+      "loss": 1.153,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8114080164439876,
+      "grad_norm": 0.8160942196846008,
+      "learning_rate": 0.0001,
+      "loss": 0.5884,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8114936622130867,
+      "grad_norm": 0.7726562023162842,
+      "learning_rate": 0.0001,
+      "loss": 0.8954,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8115793079821857,
+      "grad_norm": 1.1706643104553223,
+      "learning_rate": 0.0001,
+      "loss": 0.8125,
+      "step": 9476
+    },
+    {
+      "epoch": 0.8116649537512847,
+      "grad_norm": 0.6653918623924255,
+      "learning_rate": 0.0001,
+      "loss": 1.0943,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8117505995203836,
+      "grad_norm": 0.7811260223388672,
+      "learning_rate": 0.0001,
+      "loss": 0.4853,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8118362452894827,
+      "grad_norm": 0.8680320382118225,
+      "learning_rate": 0.0001,
+      "loss": 0.7806,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8119218910585817,
+      "grad_norm": 0.8160377740859985,
+      "learning_rate": 0.0001,
+      "loss": 0.5946,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8120075368276807,
+      "grad_norm": 0.7809337973594666,
+      "learning_rate": 0.0001,
+      "loss": 0.8657,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8120931825967798,
+      "grad_norm": 0.7526123523712158,
+      "learning_rate": 0.0001,
+      "loss": 0.9999,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8121788283658787,
+      "grad_norm": 1.1093931198120117,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8122644741349777,
+      "grad_norm": 1.0489888191223145,
+      "learning_rate": 0.0001,
+      "loss": 0.7683,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8123501199040767,
+      "grad_norm": 0.6112576723098755,
+      "learning_rate": 0.0001,
+      "loss": 0.9585,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8124357656731758,
+      "grad_norm": 1.0103150606155396,
+      "learning_rate": 0.0001,
+      "loss": 1.0014,
+      "step": 9486
+    },
+    {
+      "epoch": 0.8125214114422747,
+      "grad_norm": 0.7808998227119446,
+      "learning_rate": 0.0001,
+      "loss": 0.5156,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8126070572113737,
+      "grad_norm": 0.7551828622817993,
+      "learning_rate": 0.0001,
+      "loss": 1.2285,
+      "step": 9488
+    },
+    {
+      "epoch": 0.8126927029804728,
+      "grad_norm": 0.7493751645088196,
+      "learning_rate": 0.0001,
+      "loss": 0.8839,
+      "step": 9489
+    },
+    {
+      "epoch": 0.8127783487495718,
+      "grad_norm": 1.6108540296554565,
+      "learning_rate": 0.0001,
+      "loss": 1.3504,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8128639945186708,
+      "grad_norm": 1.2639893293380737,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 9491
+    },
+    {
+      "epoch": 0.8129496402877698,
+      "grad_norm": 1.0032919645309448,
+      "learning_rate": 0.0001,
+      "loss": 1.1501,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8130352860568688,
+      "grad_norm": 1.0771764516830444,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 9493
+    },
+    {
+      "epoch": 0.8131209318259678,
+      "grad_norm": 0.9125810265541077,
+      "learning_rate": 0.0001,
+      "loss": 0.8662,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8132065775950668,
+      "grad_norm": 1.5546103715896606,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8132922233641658,
+      "grad_norm": 0.7123640775680542,
+      "learning_rate": 0.0001,
+      "loss": 0.5491,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8133778691332648,
+      "grad_norm": 0.8378816843032837,
+      "learning_rate": 0.0001,
+      "loss": 0.9131,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8134635149023638,
+      "grad_norm": 0.6238937973976135,
+      "learning_rate": 0.0001,
+      "loss": 0.3156,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8135491606714629,
+      "grad_norm": 0.6010368466377258,
+      "learning_rate": 0.0001,
+      "loss": 1.1182,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8136348064405619,
+      "grad_norm": 1.4524800777435303,
+      "learning_rate": 0.0001,
+      "loss": 1.2799,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8137204522096608,
+      "grad_norm": 0.7800278663635254,
+      "learning_rate": 0.0001,
+      "loss": 0.6604,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8138060979787598,
+      "grad_norm": 0.6727455258369446,
+      "learning_rate": 0.0001,
+      "loss": 1.1609,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8138917437478589,
+      "grad_norm": 0.6212877631187439,
+      "learning_rate": 0.0001,
+      "loss": 1.059,
+      "step": 9503
+    },
+    {
+      "epoch": 0.8139773895169579,
+      "grad_norm": 0.8370538353919983,
+      "learning_rate": 0.0001,
+      "loss": 0.7109,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8140630352860568,
+      "grad_norm": 0.8367963433265686,
+      "learning_rate": 0.0001,
+      "loss": 1.0569,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8141486810551559,
+      "grad_norm": 1.2462847232818604,
+      "learning_rate": 0.0001,
+      "loss": 1.2679,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8142343268242549,
+      "grad_norm": 0.6048295497894287,
+      "learning_rate": 0.0001,
+      "loss": 0.7783,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8143199725933539,
+      "grad_norm": 0.7563977837562561,
+      "learning_rate": 0.0001,
+      "loss": 0.7057,
+      "step": 9508
+    },
+    {
+      "epoch": 0.8144056183624528,
+      "grad_norm": 0.7824932336807251,
+      "learning_rate": 0.0001,
+      "loss": 0.9387,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8144912641315519,
+      "grad_norm": 0.9277403950691223,
+      "learning_rate": 0.0001,
+      "loss": 1.0535,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8145769099006509,
+      "grad_norm": 0.8763672113418579,
+      "learning_rate": 0.0001,
+      "loss": 0.8547,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8146625556697499,
+      "grad_norm": 1.260624647140503,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 9512
+    },
+    {
+      "epoch": 0.814748201438849,
+      "grad_norm": 0.6315179467201233,
+      "learning_rate": 0.0001,
+      "loss": 1.0816,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8148338472079479,
+      "grad_norm": 0.9903106093406677,
+      "learning_rate": 0.0001,
+      "loss": 1.0382,
+      "step": 9514
+    },
+    {
+      "epoch": 0.8149194929770469,
+      "grad_norm": 0.943383514881134,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8150051387461459,
+      "grad_norm": 1.7696456909179688,
+      "learning_rate": 0.0001,
+      "loss": 1.1248,
+      "step": 9516
+    },
+    {
+      "epoch": 0.815090784515245,
+      "grad_norm": 1.2176768779754639,
+      "learning_rate": 0.0001,
+      "loss": 1.0556,
+      "step": 9517
+    },
+    {
+      "epoch": 0.815176430284344,
+      "grad_norm": 1.0617433786392212,
+      "learning_rate": 0.0001,
+      "loss": 1.0827,
+      "step": 9518
+    },
+    {
+      "epoch": 0.8152620760534429,
+      "grad_norm": 0.973888099193573,
+      "learning_rate": 0.0001,
+      "loss": 0.6649,
+      "step": 9519
+    },
+    {
+      "epoch": 0.815347721822542,
+      "grad_norm": 0.7880772352218628,
+      "learning_rate": 0.0001,
+      "loss": 0.9426,
+      "step": 9520
+    },
+    {
+      "epoch": 0.815433367591641,
+      "grad_norm": 0.9596991539001465,
+      "learning_rate": 0.0001,
+      "loss": 0.8597,
+      "step": 9521
+    },
+    {
+      "epoch": 0.81551901336074,
+      "grad_norm": 0.7050376534461975,
+      "learning_rate": 0.0001,
+      "loss": 0.7416,
+      "step": 9522
+    },
+    {
+      "epoch": 0.815604659129839,
+      "grad_norm": 0.7638078927993774,
+      "learning_rate": 0.0001,
+      "loss": 0.7009,
+      "step": 9523
+    },
+    {
+      "epoch": 0.815690304898938,
+      "grad_norm": 1.2253936529159546,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 9524
+    },
+    {
+      "epoch": 0.815775950668037,
+      "grad_norm": 0.8420498371124268,
+      "learning_rate": 0.0001,
+      "loss": 1.0926,
+      "step": 9525
+    },
+    {
+      "epoch": 0.815861596437136,
+      "grad_norm": 0.661850094795227,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 9526
+    },
+    {
+      "epoch": 0.815947242206235,
+      "grad_norm": 0.6868914365768433,
+      "learning_rate": 0.0001,
+      "loss": 1.1322,
+      "step": 9527
+    },
+    {
+      "epoch": 0.816032887975334,
+      "grad_norm": 0.25868767499923706,
+      "learning_rate": 0.0001,
+      "loss": 0.0977,
+      "step": 9528
+    },
+    {
+      "epoch": 0.816118533744433,
+      "grad_norm": 0.703075647354126,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 9529
+    },
+    {
+      "epoch": 0.8162041795135321,
+      "grad_norm": 0.5587443709373474,
+      "learning_rate": 0.0001,
+      "loss": 0.6441,
+      "step": 9530
+    },
+    {
+      "epoch": 0.816289825282631,
+      "grad_norm": 1.1823711395263672,
+      "learning_rate": 0.0001,
+      "loss": 1.1606,
+      "step": 9531
+    },
+    {
+      "epoch": 0.81637547105173,
+      "grad_norm": 0.6216550469398499,
+      "learning_rate": 0.0001,
+      "loss": 0.6565,
+      "step": 9532
+    },
+    {
+      "epoch": 0.816461116820829,
+      "grad_norm": 0.6574486494064331,
+      "learning_rate": 0.0001,
+      "loss": 0.8546,
+      "step": 9533
+    },
+    {
+      "epoch": 0.8165467625899281,
+      "grad_norm": 0.7007403373718262,
+      "learning_rate": 0.0001,
+      "loss": 1.0707,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8166324083590271,
+      "grad_norm": 0.7271286845207214,
+      "learning_rate": 0.0001,
+      "loss": 0.9389,
+      "step": 9535
+    },
+    {
+      "epoch": 0.816718054128126,
+      "grad_norm": 0.9000371098518372,
+      "learning_rate": 0.0001,
+      "loss": 0.7904,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8168036998972251,
+      "grad_norm": 0.6032791137695312,
+      "learning_rate": 0.0001,
+      "loss": 0.8131,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8168893456663241,
+      "grad_norm": 1.2656641006469727,
+      "learning_rate": 0.0001,
+      "loss": 0.9035,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8169749914354231,
+      "grad_norm": 0.7891127467155457,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 9539
+    },
+    {
+      "epoch": 0.817060637204522,
+      "grad_norm": 1.1543855667114258,
+      "learning_rate": 0.0001,
+      "loss": 1.1123,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8171462829736211,
+      "grad_norm": 1.381874680519104,
+      "learning_rate": 0.0001,
+      "loss": 1.1072,
+      "step": 9541
+    },
+    {
+      "epoch": 0.8172319287427201,
+      "grad_norm": 1.3615211248397827,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8173175745118191,
+      "grad_norm": 0.33209267258644104,
+      "learning_rate": 0.0001,
+      "loss": 0.6108,
+      "step": 9543
+    },
+    {
+      "epoch": 0.8174032202809182,
+      "grad_norm": 0.857854962348938,
+      "learning_rate": 0.0001,
+      "loss": 0.9964,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8174888660500171,
+      "grad_norm": 1.0952317714691162,
+      "learning_rate": 0.0001,
+      "loss": 3.2829,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8175745118191161,
+      "grad_norm": 0.5760693550109863,
+      "learning_rate": 0.0001,
+      "loss": 0.6493,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8176601575882152,
+      "grad_norm": 1.2445347309112549,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8177458033573142,
+      "grad_norm": 0.8649731874465942,
+      "learning_rate": 0.0001,
+      "loss": 0.9283,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8178314491264131,
+      "grad_norm": 0.7895470261573792,
+      "learning_rate": 0.0001,
+      "loss": 0.8711,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8179170948955121,
+      "grad_norm": 0.6463416218757629,
+      "learning_rate": 0.0001,
+      "loss": 1.239,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8180027406646112,
+      "grad_norm": 1.035321831703186,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8180883864337102,
+      "grad_norm": 0.6945057511329651,
+      "learning_rate": 0.0001,
+      "loss": 0.9089,
+      "step": 9552
+    },
+    {
+      "epoch": 0.8181740322028092,
+      "grad_norm": 0.9585725665092468,
+      "learning_rate": 0.0001,
+      "loss": 1.1814,
+      "step": 9553
+    },
+    {
+      "epoch": 0.8182596779719082,
+      "grad_norm": 0.4064914286136627,
+      "learning_rate": 0.0001,
+      "loss": 0.4906,
+      "step": 9554
+    },
+    {
+      "epoch": 0.8183453237410072,
+      "grad_norm": 0.7846525311470032,
+      "learning_rate": 0.0001,
+      "loss": 0.8901,
+      "step": 9555
+    },
+    {
+      "epoch": 0.8184309695101062,
+      "grad_norm": 1.089974045753479,
+      "learning_rate": 0.0001,
+      "loss": 1.1839,
+      "step": 9556
+    },
+    {
+      "epoch": 0.8185166152792052,
+      "grad_norm": 1.1302334070205688,
+      "learning_rate": 0.0001,
+      "loss": 1.0922,
+      "step": 9557
+    },
+    {
+      "epoch": 0.8186022610483042,
+      "grad_norm": 0.7125784158706665,
+      "learning_rate": 0.0001,
+      "loss": 0.9292,
+      "step": 9558
+    },
+    {
+      "epoch": 0.8186879068174032,
+      "grad_norm": 0.5317464470863342,
+      "learning_rate": 0.0001,
+      "loss": 0.8922,
+      "step": 9559
+    },
+    {
+      "epoch": 0.8187735525865022,
+      "grad_norm": 0.9077727198600769,
+      "learning_rate": 0.0001,
+      "loss": 1.3532,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8188591983556013,
+      "grad_norm": 1.2734583616256714,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 9561
+    },
+    {
+      "epoch": 0.8189448441247003,
+      "grad_norm": 1.3151381015777588,
+      "learning_rate": 0.0001,
+      "loss": 1.2882,
+      "step": 9562
+    },
+    {
+      "epoch": 0.8190304898937992,
+      "grad_norm": 1.5717791318893433,
+      "learning_rate": 0.0001,
+      "loss": 2.2052,
+      "step": 9563
+    },
+    {
+      "epoch": 0.8191161356628982,
+      "grad_norm": 1.0041558742523193,
+      "learning_rate": 0.0001,
+      "loss": 0.7741,
+      "step": 9564
+    },
+    {
+      "epoch": 0.8192017814319973,
+      "grad_norm": 0.6630010604858398,
+      "learning_rate": 0.0001,
+      "loss": 0.5626,
+      "step": 9565
+    },
+    {
+      "epoch": 0.8192874272010963,
+      "grad_norm": 0.8164867758750916,
+      "learning_rate": 0.0001,
+      "loss": 0.7469,
+      "step": 9566
+    },
+    {
+      "epoch": 0.8193730729701952,
+      "grad_norm": 0.7375938892364502,
+      "learning_rate": 0.0001,
+      "loss": 0.6775,
+      "step": 9567
+    },
+    {
+      "epoch": 0.8194587187392943,
+      "grad_norm": 0.864578127861023,
+      "learning_rate": 0.0001,
+      "loss": 0.8008,
+      "step": 9568
+    },
+    {
+      "epoch": 0.8195443645083933,
+      "grad_norm": 0.27972763776779175,
+      "learning_rate": 0.0001,
+      "loss": 0.7195,
+      "step": 9569
+    },
+    {
+      "epoch": 0.8196300102774923,
+      "grad_norm": 0.5953155159950256,
+      "learning_rate": 0.0001,
+      "loss": 0.399,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8197156560465912,
+      "grad_norm": 0.6958663463592529,
+      "learning_rate": 0.0001,
+      "loss": 0.595,
+      "step": 9571
+    },
+    {
+      "epoch": 0.8198013018156903,
+      "grad_norm": 1.0097373723983765,
+      "learning_rate": 0.0001,
+      "loss": 1.244,
+      "step": 9572
+    },
+    {
+      "epoch": 0.8198869475847893,
+      "grad_norm": 0.8613535761833191,
+      "learning_rate": 0.0001,
+      "loss": 1.136,
+      "step": 9573
+    },
+    {
+      "epoch": 0.8199725933538883,
+      "grad_norm": 0.5655787587165833,
+      "learning_rate": 0.0001,
+      "loss": 0.9519,
+      "step": 9574
+    },
+    {
+      "epoch": 0.8200582391229874,
+      "grad_norm": 0.7053650617599487,
+      "learning_rate": 0.0001,
+      "loss": 0.9532,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8201438848920863,
+      "grad_norm": 0.24521146714687347,
+      "learning_rate": 0.0001,
+      "loss": 0.1004,
+      "step": 9576
+    },
+    {
+      "epoch": 0.8202295306611853,
+      "grad_norm": 0.7087011337280273,
+      "learning_rate": 0.0001,
+      "loss": 0.5586,
+      "step": 9577
+    },
+    {
+      "epoch": 0.8203151764302844,
+      "grad_norm": 1.0707579851150513,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 9578
+    },
+    {
+      "epoch": 0.8204008221993834,
+      "grad_norm": 0.6716638803482056,
+      "learning_rate": 0.0001,
+      "loss": 0.4881,
+      "step": 9579
+    },
+    {
+      "epoch": 0.8204864679684823,
+      "grad_norm": 0.6533719897270203,
+      "learning_rate": 0.0001,
+      "loss": 0.7017,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8205721137375813,
+      "grad_norm": 1.2195518016815186,
+      "learning_rate": 0.0001,
+      "loss": 0.9376,
+      "step": 9581
+    },
+    {
+      "epoch": 0.8206577595066804,
+      "grad_norm": 0.9806455969810486,
+      "learning_rate": 0.0001,
+      "loss": 1.1548,
+      "step": 9582
+    },
+    {
+      "epoch": 0.8207434052757794,
+      "grad_norm": 1.000178575515747,
+      "learning_rate": 0.0001,
+      "loss": 1.0757,
+      "step": 9583
+    },
+    {
+      "epoch": 0.8208290510448784,
+      "grad_norm": 1.0750733613967896,
+      "learning_rate": 0.0001,
+      "loss": 1.0001,
+      "step": 9584
+    },
+    {
+      "epoch": 0.8209146968139774,
+      "grad_norm": 0.9535443782806396,
+      "learning_rate": 0.0001,
+      "loss": 0.8905,
+      "step": 9585
+    },
+    {
+      "epoch": 0.8210003425830764,
+      "grad_norm": 1.0355451107025146,
+      "learning_rate": 0.0001,
+      "loss": 0.5405,
+      "step": 9586
+    },
+    {
+      "epoch": 0.8210859883521754,
+      "grad_norm": 0.7210434675216675,
+      "learning_rate": 0.0001,
+      "loss": 0.4937,
+      "step": 9587
+    },
+    {
+      "epoch": 0.8211716341212744,
+      "grad_norm": 1.1883957386016846,
+      "learning_rate": 0.0001,
+      "loss": 0.819,
+      "step": 9588
+    },
+    {
+      "epoch": 0.8212572798903734,
+      "grad_norm": 0.6198357939720154,
+      "learning_rate": 0.0001,
+      "loss": 1.0477,
+      "step": 9589
+    },
+    {
+      "epoch": 0.8213429256594724,
+      "grad_norm": 0.2959338426589966,
+      "learning_rate": 0.0001,
+      "loss": 0.4862,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 0.9308714270591736,
+      "learning_rate": 0.0001,
+      "loss": 0.8023,
+      "step": 9591
+    },
+    {
+      "epoch": 0.8215142171976705,
+      "grad_norm": 1.0641944408416748,
+      "learning_rate": 0.0001,
+      "loss": 0.7555,
+      "step": 9592
+    },
+    {
+      "epoch": 0.8215998629667695,
+      "grad_norm": 0.9907900094985962,
+      "learning_rate": 0.0001,
+      "loss": 0.8173,
+      "step": 9593
+    },
+    {
+      "epoch": 0.8216855087358684,
+      "grad_norm": 0.9777377843856812,
+      "learning_rate": 0.0001,
+      "loss": 1.1795,
+      "step": 9594
+    },
+    {
+      "epoch": 0.8217711545049674,
+      "grad_norm": 2.1888833045959473,
+      "learning_rate": 0.0001,
+      "loss": 1.4188,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8218568002740665,
+      "grad_norm": 1.8388326168060303,
+      "learning_rate": 0.0001,
+      "loss": 2.2498,
+      "step": 9596
+    },
+    {
+      "epoch": 0.8219424460431655,
+      "grad_norm": 0.5462139248847961,
+      "learning_rate": 0.0001,
+      "loss": 0.685,
+      "step": 9597
+    },
+    {
+      "epoch": 0.8220280918122644,
+      "grad_norm": 1.4528800249099731,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 9598
+    },
+    {
+      "epoch": 0.8221137375813635,
+      "grad_norm": 0.9805005192756653,
+      "learning_rate": 0.0001,
+      "loss": 0.8696,
+      "step": 9599
+    },
+    {
+      "epoch": 0.8221993833504625,
+      "grad_norm": 0.9035658836364746,
+      "learning_rate": 0.0001,
+      "loss": 1.28,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8222850291195615,
+      "grad_norm": 1.0146327018737793,
+      "learning_rate": 0.0001,
+      "loss": 0.8151,
+      "step": 9601
+    },
+    {
+      "epoch": 0.8223706748886604,
+      "grad_norm": 0.448579341173172,
+      "learning_rate": 0.0001,
+      "loss": 0.526,
+      "step": 9602
+    },
+    {
+      "epoch": 0.8224563206577595,
+      "grad_norm": 0.9112615585327148,
+      "learning_rate": 0.0001,
+      "loss": 0.7022,
+      "step": 9603
+    },
+    {
+      "epoch": 0.8225419664268585,
+      "grad_norm": 1.1079403162002563,
+      "learning_rate": 0.0001,
+      "loss": 0.6413,
+      "step": 9604
+    },
+    {
+      "epoch": 0.8226276121959575,
+      "grad_norm": 0.862099289894104,
+      "learning_rate": 0.0001,
+      "loss": 1.0455,
+      "step": 9605
+    },
+    {
+      "epoch": 0.8227132579650566,
+      "grad_norm": 0.9171148538589478,
+      "learning_rate": 0.0001,
+      "loss": 1.2792,
+      "step": 9606
+    },
+    {
+      "epoch": 0.8227989037341555,
+      "grad_norm": 1.2209872007369995,
+      "learning_rate": 0.0001,
+      "loss": 1.4654,
+      "step": 9607
+    },
+    {
+      "epoch": 0.8228845495032545,
+      "grad_norm": 1.267250657081604,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 9608
+    },
+    {
+      "epoch": 0.8229701952723536,
+      "grad_norm": 0.9581796526908875,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 9609
+    },
+    {
+      "epoch": 0.8230558410414526,
+      "grad_norm": 0.7119987607002258,
+      "learning_rate": 0.0001,
+      "loss": 0.6983,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8231414868105515,
+      "grad_norm": 1.168896198272705,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 9611
+    },
+    {
+      "epoch": 0.8232271325796505,
+      "grad_norm": 1.1064563989639282,
+      "learning_rate": 0.0001,
+      "loss": 1.1275,
+      "step": 9612
+    },
+    {
+      "epoch": 0.8233127783487496,
+      "grad_norm": 0.639010488986969,
+      "learning_rate": 0.0001,
+      "loss": 1.065,
+      "step": 9613
+    },
+    {
+      "epoch": 0.8233984241178486,
+      "grad_norm": 1.2956280708312988,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 9614
+    },
+    {
+      "epoch": 0.8234840698869476,
+      "grad_norm": 0.7126975655555725,
+      "learning_rate": 0.0001,
+      "loss": 0.649,
+      "step": 9615
+    },
+    {
+      "epoch": 0.8235697156560466,
+      "grad_norm": 0.7622031569480896,
+      "learning_rate": 0.0001,
+      "loss": 0.7255,
+      "step": 9616
+    },
+    {
+      "epoch": 0.8236553614251456,
+      "grad_norm": 0.7344436049461365,
+      "learning_rate": 0.0001,
+      "loss": 0.8611,
+      "step": 9617
+    },
+    {
+      "epoch": 0.8237410071942446,
+      "grad_norm": 1.441341519355774,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 9618
+    },
+    {
+      "epoch": 0.8238266529633436,
+      "grad_norm": 1.0833202600479126,
+      "learning_rate": 0.0001,
+      "loss": 1.2581,
+      "step": 9619
+    },
+    {
+      "epoch": 0.8239122987324426,
+      "grad_norm": 0.7412538528442383,
+      "learning_rate": 0.0001,
+      "loss": 0.6243,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8239979445015416,
+      "grad_norm": 0.6908422708511353,
+      "learning_rate": 0.0001,
+      "loss": 0.6191,
+      "step": 9621
+    },
+    {
+      "epoch": 0.8240835902706406,
+      "grad_norm": 0.5591647028923035,
+      "learning_rate": 0.0001,
+      "loss": 0.5125,
+      "step": 9622
+    },
+    {
+      "epoch": 0.8241692360397397,
+      "grad_norm": 0.3153376281261444,
+      "learning_rate": 0.0001,
+      "loss": 0.3758,
+      "step": 9623
+    },
+    {
+      "epoch": 0.8242548818088387,
+      "grad_norm": 1.0434980392456055,
+      "learning_rate": 0.0001,
+      "loss": 0.9755,
+      "step": 9624
+    },
+    {
+      "epoch": 0.8243405275779376,
+      "grad_norm": 0.7534710764884949,
+      "learning_rate": 0.0001,
+      "loss": 0.597,
+      "step": 9625
+    },
+    {
+      "epoch": 0.8244261733470366,
+      "grad_norm": 0.9581044912338257,
+      "learning_rate": 0.0001,
+      "loss": 0.9741,
+      "step": 9626
+    },
+    {
+      "epoch": 0.8245118191161357,
+      "grad_norm": 1.4480794668197632,
+      "learning_rate": 0.0001,
+      "loss": 0.5505,
+      "step": 9627
+    },
+    {
+      "epoch": 0.8245974648852347,
+      "grad_norm": 0.7647348046302795,
+      "learning_rate": 0.0001,
+      "loss": 0.8527,
+      "step": 9628
+    },
+    {
+      "epoch": 0.8246831106543336,
+      "grad_norm": 0.7122359275817871,
+      "learning_rate": 0.0001,
+      "loss": 1.1821,
+      "step": 9629
+    },
+    {
+      "epoch": 0.8247687564234327,
+      "grad_norm": 0.7278649806976318,
+      "learning_rate": 0.0001,
+      "loss": 1.001,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8248544021925317,
+      "grad_norm": 1.5334540605545044,
+      "learning_rate": 0.0001,
+      "loss": 1.1367,
+      "step": 9631
+    },
+    {
+      "epoch": 0.8249400479616307,
+      "grad_norm": 0.3400205969810486,
+      "learning_rate": 0.0001,
+      "loss": 0.2035,
+      "step": 9632
+    },
+    {
+      "epoch": 0.8250256937307296,
+      "grad_norm": 0.6996507048606873,
+      "learning_rate": 0.0001,
+      "loss": 0.9231,
+      "step": 9633
+    },
+    {
+      "epoch": 0.8251113394998287,
+      "grad_norm": 0.6025702953338623,
+      "learning_rate": 0.0001,
+      "loss": 0.6539,
+      "step": 9634
+    },
+    {
+      "epoch": 0.8251969852689277,
+      "grad_norm": 1.1288474798202515,
+      "learning_rate": 0.0001,
+      "loss": 1.0115,
+      "step": 9635
+    },
+    {
+      "epoch": 0.8252826310380267,
+      "grad_norm": 1.2226530313491821,
+      "learning_rate": 0.0001,
+      "loss": 1.1462,
+      "step": 9636
+    },
+    {
+      "epoch": 0.8253682768071258,
+      "grad_norm": 0.8471972942352295,
+      "learning_rate": 0.0001,
+      "loss": 1.1515,
+      "step": 9637
+    },
+    {
+      "epoch": 0.8254539225762247,
+      "grad_norm": 0.9686079025268555,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 9638
+    },
+    {
+      "epoch": 0.8255395683453237,
+      "grad_norm": 0.9126067757606506,
+      "learning_rate": 0.0001,
+      "loss": 0.8251,
+      "step": 9639
+    },
+    {
+      "epoch": 0.8256252141144228,
+      "grad_norm": 1.0715841054916382,
+      "learning_rate": 0.0001,
+      "loss": 0.8573,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8257108598835218,
+      "grad_norm": 1.1947388648986816,
+      "learning_rate": 0.0001,
+      "loss": 1.1501,
+      "step": 9641
+    },
+    {
+      "epoch": 0.8257965056526207,
+      "grad_norm": 0.5507193207740784,
+      "learning_rate": 0.0001,
+      "loss": 0.4272,
+      "step": 9642
+    },
+    {
+      "epoch": 0.8258821514217197,
+      "grad_norm": 0.9019783735275269,
+      "learning_rate": 0.0001,
+      "loss": 0.8631,
+      "step": 9643
+    },
+    {
+      "epoch": 0.8259677971908188,
+      "grad_norm": 1.0180811882019043,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 9644
+    },
+    {
+      "epoch": 0.8260534429599178,
+      "grad_norm": 0.8115811944007874,
+      "learning_rate": 0.0001,
+      "loss": 1.2151,
+      "step": 9645
+    },
+    {
+      "epoch": 0.8261390887290168,
+      "grad_norm": 0.4547938108444214,
+      "learning_rate": 0.0001,
+      "loss": 0.3617,
+      "step": 9646
+    },
+    {
+      "epoch": 0.8262247344981158,
+      "grad_norm": 0.8139580488204956,
+      "learning_rate": 0.0001,
+      "loss": 0.7416,
+      "step": 9647
+    },
+    {
+      "epoch": 0.8263103802672148,
+      "grad_norm": 1.1976757049560547,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 9648
+    },
+    {
+      "epoch": 0.8263960260363138,
+      "grad_norm": 0.7764602899551392,
+      "learning_rate": 0.0001,
+      "loss": 0.7961,
+      "step": 9649
+    },
+    {
+      "epoch": 0.8264816718054128,
+      "grad_norm": 0.8737772703170776,
+      "learning_rate": 0.0001,
+      "loss": 0.8265,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8265673175745119,
+      "grad_norm": 0.5661895275115967,
+      "learning_rate": 0.0001,
+      "loss": 0.9153,
+      "step": 9651
+    },
+    {
+      "epoch": 0.8266529633436108,
+      "grad_norm": 0.4744296371936798,
+      "learning_rate": 0.0001,
+      "loss": 0.6518,
+      "step": 9652
+    },
+    {
+      "epoch": 0.8267386091127098,
+      "grad_norm": 1.0468389987945557,
+      "learning_rate": 0.0001,
+      "loss": 0.937,
+      "step": 9653
+    },
+    {
+      "epoch": 0.8268242548818089,
+      "grad_norm": 1.1316604614257812,
+      "learning_rate": 0.0001,
+      "loss": 0.8621,
+      "step": 9654
+    },
+    {
+      "epoch": 0.8269099006509079,
+      "grad_norm": 0.41332921385765076,
+      "learning_rate": 0.0001,
+      "loss": 0.1403,
+      "step": 9655
+    },
+    {
+      "epoch": 0.8269955464200068,
+      "grad_norm": 1.3547035455703735,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 9656
+    },
+    {
+      "epoch": 0.8270811921891058,
+      "grad_norm": 0.6850831508636475,
+      "learning_rate": 0.0001,
+      "loss": 0.8647,
+      "step": 9657
+    },
+    {
+      "epoch": 0.8271668379582049,
+      "grad_norm": 1.2909266948699951,
+      "learning_rate": 0.0001,
+      "loss": 0.9274,
+      "step": 9658
+    },
+    {
+      "epoch": 0.8272524837273039,
+      "grad_norm": 1.4007035493850708,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 9659
+    },
+    {
+      "epoch": 0.8273381294964028,
+      "grad_norm": 0.7131264805793762,
+      "learning_rate": 0.0001,
+      "loss": 0.6662,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8274237752655019,
+      "grad_norm": 0.9493045210838318,
+      "learning_rate": 0.0001,
+      "loss": 0.9346,
+      "step": 9661
+    },
+    {
+      "epoch": 0.8275094210346009,
+      "grad_norm": 0.7838862538337708,
+      "learning_rate": 0.0001,
+      "loss": 0.8038,
+      "step": 9662
+    },
+    {
+      "epoch": 0.8275950668036999,
+      "grad_norm": 0.617772102355957,
+      "learning_rate": 0.0001,
+      "loss": 0.5421,
+      "step": 9663
+    },
+    {
+      "epoch": 0.8276807125727988,
+      "grad_norm": 0.9514848589897156,
+      "learning_rate": 0.0001,
+      "loss": 0.8112,
+      "step": 9664
+    },
+    {
+      "epoch": 0.8277663583418979,
+      "grad_norm": 0.6313784718513489,
+      "learning_rate": 0.0001,
+      "loss": 1.1629,
+      "step": 9665
+    },
+    {
+      "epoch": 0.8278520041109969,
+      "grad_norm": 0.9652764201164246,
+      "learning_rate": 0.0001,
+      "loss": 0.9051,
+      "step": 9666
+    },
+    {
+      "epoch": 0.8279376498800959,
+      "grad_norm": 1.101338267326355,
+      "learning_rate": 0.0001,
+      "loss": 1.0909,
+      "step": 9667
+    },
+    {
+      "epoch": 0.828023295649195,
+      "grad_norm": 0.9487239122390747,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 9668
+    },
+    {
+      "epoch": 0.8281089414182939,
+      "grad_norm": 0.6421770453453064,
+      "learning_rate": 0.0001,
+      "loss": 0.3384,
+      "step": 9669
+    },
+    {
+      "epoch": 0.8281945871873929,
+      "grad_norm": 0.9862491488456726,
+      "learning_rate": 0.0001,
+      "loss": 1.0835,
+      "step": 9670
+    },
+    {
+      "epoch": 0.828280232956492,
+      "grad_norm": 0.9800370335578918,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 9671
+    },
+    {
+      "epoch": 0.828365878725591,
+      "grad_norm": 1.5602679252624512,
+      "learning_rate": 0.0001,
+      "loss": 1.0763,
+      "step": 9672
+    },
+    {
+      "epoch": 0.82845152449469,
+      "grad_norm": 1.4614876508712769,
+      "learning_rate": 0.0001,
+      "loss": 2.1788,
+      "step": 9673
+    },
+    {
+      "epoch": 0.8285371702637889,
+      "grad_norm": 0.6028294563293457,
+      "learning_rate": 0.0001,
+      "loss": 0.7971,
+      "step": 9674
+    },
+    {
+      "epoch": 0.828622816032888,
+      "grad_norm": 0.9943216443061829,
+      "learning_rate": 0.0001,
+      "loss": 0.8015,
+      "step": 9675
+    },
+    {
+      "epoch": 0.828708461801987,
+      "grad_norm": 0.8416003584861755,
+      "learning_rate": 0.0001,
+      "loss": 0.9121,
+      "step": 9676
+    },
+    {
+      "epoch": 0.828794107571086,
+      "grad_norm": 1.0797289609909058,
+      "learning_rate": 0.0001,
+      "loss": 1.1722,
+      "step": 9677
+    },
+    {
+      "epoch": 0.828879753340185,
+      "grad_norm": 1.188651204109192,
+      "learning_rate": 0.0001,
+      "loss": 1.1538,
+      "step": 9678
+    },
+    {
+      "epoch": 0.828965399109284,
+      "grad_norm": 0.6501845121383667,
+      "learning_rate": 0.0001,
+      "loss": 1.2441,
+      "step": 9679
+    },
+    {
+      "epoch": 0.829051044878383,
+      "grad_norm": 0.8039807677268982,
+      "learning_rate": 0.0001,
+      "loss": 0.9042,
+      "step": 9680
+    },
+    {
+      "epoch": 0.829136690647482,
+      "grad_norm": 1.317016363143921,
+      "learning_rate": 0.0001,
+      "loss": 1.1148,
+      "step": 9681
+    },
+    {
+      "epoch": 0.829222336416581,
+      "grad_norm": 0.7719148397445679,
+      "learning_rate": 0.0001,
+      "loss": 0.84,
+      "step": 9682
+    },
+    {
+      "epoch": 0.82930798218568,
+      "grad_norm": 0.7585822343826294,
+      "learning_rate": 0.0001,
+      "loss": 1.3816,
+      "step": 9683
+    },
+    {
+      "epoch": 0.829393627954779,
+      "grad_norm": 0.7153705358505249,
+      "learning_rate": 0.0001,
+      "loss": 0.8549,
+      "step": 9684
+    },
+    {
+      "epoch": 0.8294792737238781,
+      "grad_norm": 1.0266331434249878,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 9685
+    },
+    {
+      "epoch": 0.8295649194929771,
+      "grad_norm": 0.8700186610221863,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 9686
+    },
+    {
+      "epoch": 0.829650565262076,
+      "grad_norm": 1.030519723892212,
+      "learning_rate": 0.0001,
+      "loss": 1.2847,
+      "step": 9687
+    },
+    {
+      "epoch": 0.829736211031175,
+      "grad_norm": 0.6955069303512573,
+      "learning_rate": 0.0001,
+      "loss": 1.2902,
+      "step": 9688
+    },
+    {
+      "epoch": 0.8298218568002741,
+      "grad_norm": 1.0132602453231812,
+      "learning_rate": 0.0001,
+      "loss": 1.2048,
+      "step": 9689
+    },
+    {
+      "epoch": 0.8299075025693731,
+      "grad_norm": 0.7260561585426331,
+      "learning_rate": 0.0001,
+      "loss": 0.4575,
+      "step": 9690
+    },
+    {
+      "epoch": 0.829993148338472,
+      "grad_norm": 1.4194241762161255,
+      "learning_rate": 0.0001,
+      "loss": 1.1851,
+      "step": 9691
+    },
+    {
+      "epoch": 0.8300787941075711,
+      "grad_norm": 0.9363057613372803,
+      "learning_rate": 0.0001,
+      "loss": 0.9131,
+      "step": 9692
+    },
+    {
+      "epoch": 0.8301644398766701,
+      "grad_norm": 0.4822021722793579,
+      "learning_rate": 0.0001,
+      "loss": 0.8169,
+      "step": 9693
+    },
+    {
+      "epoch": 0.8302500856457691,
+      "grad_norm": 0.6123325824737549,
+      "learning_rate": 0.0001,
+      "loss": 1.0252,
+      "step": 9694
+    },
+    {
+      "epoch": 0.830335731414868,
+      "grad_norm": 0.9971012473106384,
+      "learning_rate": 0.0001,
+      "loss": 0.9817,
+      "step": 9695
+    },
+    {
+      "epoch": 0.8304213771839671,
+      "grad_norm": 0.6353170871734619,
+      "learning_rate": 0.0001,
+      "loss": 1.073,
+      "step": 9696
+    },
+    {
+      "epoch": 0.8305070229530661,
+      "grad_norm": 1.1716276407241821,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 9697
+    },
+    {
+      "epoch": 0.8305926687221651,
+      "grad_norm": 1.5776073932647705,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 9698
+    },
+    {
+      "epoch": 0.8306783144912642,
+      "grad_norm": 1.1226372718811035,
+      "learning_rate": 0.0001,
+      "loss": 0.7764,
+      "step": 9699
+    },
+    {
+      "epoch": 0.8307639602603631,
+      "grad_norm": 0.718097448348999,
+      "learning_rate": 0.0001,
+      "loss": 1.0677,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8308496060294621,
+      "grad_norm": 0.9873561859130859,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 9701
+    },
+    {
+      "epoch": 0.8309352517985612,
+      "grad_norm": 0.6983315944671631,
+      "learning_rate": 0.0001,
+      "loss": 0.7321,
+      "step": 9702
+    },
+    {
+      "epoch": 0.8310208975676602,
+      "grad_norm": 0.4158795177936554,
+      "learning_rate": 0.0001,
+      "loss": 0.4079,
+      "step": 9703
+    },
+    {
+      "epoch": 0.8311065433367592,
+      "grad_norm": 1.1429378986358643,
+      "learning_rate": 0.0001,
+      "loss": 1.0837,
+      "step": 9704
+    },
+    {
+      "epoch": 0.8311921891058581,
+      "grad_norm": 0.6762258410453796,
+      "learning_rate": 0.0001,
+      "loss": 1.1582,
+      "step": 9705
+    },
+    {
+      "epoch": 0.8312778348749572,
+      "grad_norm": 0.9777173399925232,
+      "learning_rate": 0.0001,
+      "loss": 1.1547,
+      "step": 9706
+    },
+    {
+      "epoch": 0.8313634806440562,
+      "grad_norm": 1.1333807706832886,
+      "learning_rate": 0.0001,
+      "loss": 0.9417,
+      "step": 9707
+    },
+    {
+      "epoch": 0.8314491264131552,
+      "grad_norm": 0.9282203912734985,
+      "learning_rate": 0.0001,
+      "loss": 2.0356,
+      "step": 9708
+    },
+    {
+      "epoch": 0.8315347721822542,
+      "grad_norm": 0.6391975283622742,
+      "learning_rate": 0.0001,
+      "loss": 0.7748,
+      "step": 9709
+    },
+    {
+      "epoch": 0.8316204179513532,
+      "grad_norm": 0.5739451050758362,
+      "learning_rate": 0.0001,
+      "loss": 1.0357,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8317060637204522,
+      "grad_norm": 0.8020213842391968,
+      "learning_rate": 0.0001,
+      "loss": 0.5645,
+      "step": 9711
+    },
+    {
+      "epoch": 0.8317917094895512,
+      "grad_norm": 0.7109288573265076,
+      "learning_rate": 0.0001,
+      "loss": 1.2939,
+      "step": 9712
+    },
+    {
+      "epoch": 0.8318773552586503,
+      "grad_norm": 0.9055540561676025,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 9713
+    },
+    {
+      "epoch": 0.8319630010277492,
+      "grad_norm": 0.9079368114471436,
+      "learning_rate": 0.0001,
+      "loss": 0.9398,
+      "step": 9714
+    },
+    {
+      "epoch": 0.8320486467968482,
+      "grad_norm": 0.9363124966621399,
+      "learning_rate": 0.0001,
+      "loss": 1.1794,
+      "step": 9715
+    },
+    {
+      "epoch": 0.8321342925659473,
+      "grad_norm": 0.9106072187423706,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 9716
+    },
+    {
+      "epoch": 0.8322199383350463,
+      "grad_norm": 0.7593501806259155,
+      "learning_rate": 0.0001,
+      "loss": 1.2102,
+      "step": 9717
+    },
+    {
+      "epoch": 0.8323055841041452,
+      "grad_norm": 0.9842453002929688,
+      "learning_rate": 0.0001,
+      "loss": 1.2735,
+      "step": 9718
+    },
+    {
+      "epoch": 0.8323912298732442,
+      "grad_norm": 0.849137544631958,
+      "learning_rate": 0.0001,
+      "loss": 1.3571,
+      "step": 9719
+    },
+    {
+      "epoch": 0.8324768756423433,
+      "grad_norm": 0.8209199905395508,
+      "learning_rate": 0.0001,
+      "loss": 0.9834,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8325625214114423,
+      "grad_norm": 1.058655858039856,
+      "learning_rate": 0.0001,
+      "loss": 1.0229,
+      "step": 9721
+    },
+    {
+      "epoch": 0.8326481671805412,
+      "grad_norm": 0.7291421890258789,
+      "learning_rate": 0.0001,
+      "loss": 0.7728,
+      "step": 9722
+    },
+    {
+      "epoch": 0.8327338129496403,
+      "grad_norm": 1.273083209991455,
+      "learning_rate": 0.0001,
+      "loss": 1.0649,
+      "step": 9723
+    },
+    {
+      "epoch": 0.8328194587187393,
+      "grad_norm": 0.9033750891685486,
+      "learning_rate": 0.0001,
+      "loss": 0.6996,
+      "step": 9724
+    },
+    {
+      "epoch": 0.8329051044878383,
+      "grad_norm": 1.330854892730713,
+      "learning_rate": 0.0001,
+      "loss": 0.9015,
+      "step": 9725
+    },
+    {
+      "epoch": 0.8329907502569374,
+      "grad_norm": 0.6314312219619751,
+      "learning_rate": 0.0001,
+      "loss": 1.3762,
+      "step": 9726
+    },
+    {
+      "epoch": 0.8330763960260363,
+      "grad_norm": 0.6405630111694336,
+      "learning_rate": 0.0001,
+      "loss": 1.0076,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8331620417951353,
+      "grad_norm": 0.34404608607292175,
+      "learning_rate": 0.0001,
+      "loss": 0.3687,
+      "step": 9728
+    },
+    {
+      "epoch": 0.8332476875642343,
+      "grad_norm": 1.321926474571228,
+      "learning_rate": 0.0001,
+      "loss": 1.186,
+      "step": 9729
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.3225572109222412,
+      "learning_rate": 0.0001,
+      "loss": 0.9408,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8334189791024323,
+      "grad_norm": 0.6933873891830444,
+      "learning_rate": 0.0001,
+      "loss": 1.325,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8335046248715313,
+      "grad_norm": 0.6363680958747864,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 9732
+    },
+    {
+      "epoch": 0.8335902706406304,
+      "grad_norm": 1.0435510873794556,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8336759164097294,
+      "grad_norm": 0.37453383207321167,
+      "learning_rate": 0.0001,
+      "loss": 0.2364,
+      "step": 9734
+    },
+    {
+      "epoch": 0.8337615621788284,
+      "grad_norm": 1.1632816791534424,
+      "learning_rate": 0.0001,
+      "loss": 1.1725,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8338472079479273,
+      "grad_norm": 0.44597557187080383,
+      "learning_rate": 0.0001,
+      "loss": 0.5546,
+      "step": 9736
+    },
+    {
+      "epoch": 0.8339328537170264,
+      "grad_norm": 0.9601131677627563,
+      "learning_rate": 0.0001,
+      "loss": 0.7552,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8340184994861254,
+      "grad_norm": 0.5560502409934998,
+      "learning_rate": 0.0001,
+      "loss": 0.8262,
+      "step": 9738
+    },
+    {
+      "epoch": 0.8341041452552244,
+      "grad_norm": 0.9269390106201172,
+      "learning_rate": 0.0001,
+      "loss": 1.0339,
+      "step": 9739
+    },
+    {
+      "epoch": 0.8341897910243234,
+      "grad_norm": 0.8771026134490967,
+      "learning_rate": 0.0001,
+      "loss": 1.1562,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8342754367934224,
+      "grad_norm": 0.6976728439331055,
+      "learning_rate": 0.0001,
+      "loss": 0.6668,
+      "step": 9741
+    },
+    {
+      "epoch": 0.8343610825625214,
+      "grad_norm": 0.7507012486457825,
+      "learning_rate": 0.0001,
+      "loss": 1.2545,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8344467283316204,
+      "grad_norm": 1.0424004793167114,
+      "learning_rate": 0.0001,
+      "loss": 1.0603,
+      "step": 9743
+    },
+    {
+      "epoch": 0.8345323741007195,
+      "grad_norm": 0.7230269312858582,
+      "learning_rate": 0.0001,
+      "loss": 0.7513,
+      "step": 9744
+    },
+    {
+      "epoch": 0.8346180198698184,
+      "grad_norm": 0.8066165447235107,
+      "learning_rate": 0.0001,
+      "loss": 1.3222,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8347036656389174,
+      "grad_norm": 0.5927751660346985,
+      "learning_rate": 0.0001,
+      "loss": 0.7744,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8347893114080165,
+      "grad_norm": 0.9437839388847351,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 9747
+    },
+    {
+      "epoch": 0.8348749571771155,
+      "grad_norm": 1.0163923501968384,
+      "learning_rate": 0.0001,
+      "loss": 0.8471,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8349606029462144,
+      "grad_norm": 1.0403530597686768,
+      "learning_rate": 0.0001,
+      "loss": 0.9691,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8350462487153134,
+      "grad_norm": 0.8001373410224915,
+      "learning_rate": 0.0001,
+      "loss": 0.7821,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8351318944844125,
+      "grad_norm": 0.768049418926239,
+      "learning_rate": 0.0001,
+      "loss": 0.8835,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8352175402535115,
+      "grad_norm": 1.2732949256896973,
+      "learning_rate": 0.0001,
+      "loss": 1.0706,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8353031860226104,
+      "grad_norm": 0.7640895843505859,
+      "learning_rate": 0.0001,
+      "loss": 0.6344,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8353888317917095,
+      "grad_norm": 0.6761962175369263,
+      "learning_rate": 0.0001,
+      "loss": 0.8464,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8354744775608085,
+      "grad_norm": 0.7654963135719299,
+      "learning_rate": 0.0001,
+      "loss": 0.679,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8355601233299075,
+      "grad_norm": 1.40985107421875,
+      "learning_rate": 0.0001,
+      "loss": 1.0297,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8356457690990066,
+      "grad_norm": 1.2996052503585815,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8357314148681055,
+      "grad_norm": 0.6908334493637085,
+      "learning_rate": 0.0001,
+      "loss": 0.9455,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8358170606372045,
+      "grad_norm": 0.7529284954071045,
+      "learning_rate": 0.0001,
+      "loss": 0.8034,
+      "step": 9759
+    },
+    {
+      "epoch": 0.8359027064063035,
+      "grad_norm": 0.7710176706314087,
+      "learning_rate": 0.0001,
+      "loss": 1.1763,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8359883521754026,
+      "grad_norm": 0.8056353330612183,
+      "learning_rate": 0.0001,
+      "loss": 1.0523,
+      "step": 9761
+    },
+    {
+      "epoch": 0.8360739979445015,
+      "grad_norm": 0.37599003314971924,
+      "learning_rate": 0.0001,
+      "loss": 0.4756,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8361596437136005,
+      "grad_norm": 0.6558787226676941,
+      "learning_rate": 0.0001,
+      "loss": 0.9801,
+      "step": 9763
+    },
+    {
+      "epoch": 0.8362452894826996,
+      "grad_norm": 0.616941511631012,
+      "learning_rate": 0.0001,
+      "loss": 0.8787,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8363309352517986,
+      "grad_norm": 0.7162683010101318,
+      "learning_rate": 0.0001,
+      "loss": 0.5498,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8364165810208976,
+      "grad_norm": 0.6420366168022156,
+      "learning_rate": 0.0001,
+      "loss": 1.0445,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8365022267899965,
+      "grad_norm": 0.9993038773536682,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8365878725590956,
+      "grad_norm": 0.7482195496559143,
+      "learning_rate": 0.0001,
+      "loss": 1.0592,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8366735183281946,
+      "grad_norm": 0.7069492936134338,
+      "learning_rate": 0.0001,
+      "loss": 0.254,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8367591640972936,
+      "grad_norm": 0.8462894558906555,
+      "learning_rate": 0.0001,
+      "loss": 1.2734,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8368448098663926,
+      "grad_norm": 0.7416204810142517,
+      "learning_rate": 0.0001,
+      "loss": 1.227,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8369304556354916,
+      "grad_norm": 0.6372088193893433,
+      "learning_rate": 0.0001,
+      "loss": 0.6385,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8370161014045906,
+      "grad_norm": 0.7135084271430969,
+      "learning_rate": 0.0001,
+      "loss": 0.7456,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8371017471736896,
+      "grad_norm": 0.5136961936950684,
+      "learning_rate": 0.0001,
+      "loss": 0.4465,
+      "step": 9774
+    },
+    {
+      "epoch": 0.8371873929427887,
+      "grad_norm": 0.9037790298461914,
+      "learning_rate": 0.0001,
+      "loss": 1.1852,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8372730387118876,
+      "grad_norm": 1.080518126487732,
+      "learning_rate": 0.0001,
+      "loss": 1.1433,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8373586844809866,
+      "grad_norm": 0.5192341804504395,
+      "learning_rate": 0.0001,
+      "loss": 0.392,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8374443302500857,
+      "grad_norm": 0.8916842341423035,
+      "learning_rate": 0.0001,
+      "loss": 1.0786,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8375299760191847,
+      "grad_norm": 0.7397857904434204,
+      "learning_rate": 0.0001,
+      "loss": 0.7043,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8376156217882836,
+      "grad_norm": 2.3882431983947754,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8377012675573826,
+      "grad_norm": 1.2520877122879028,
+      "learning_rate": 0.0001,
+      "loss": 1.1298,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8377869133264817,
+      "grad_norm": 0.6686891317367554,
+      "learning_rate": 0.0001,
+      "loss": 0.573,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8378725590955807,
+      "grad_norm": 1.0924577713012695,
+      "learning_rate": 0.0001,
+      "loss": 0.7337,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8379582048646796,
+      "grad_norm": 1.1110278367996216,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8380438506337787,
+      "grad_norm": 0.9840734601020813,
+      "learning_rate": 0.0001,
+      "loss": 0.2754,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8381294964028777,
+      "grad_norm": 0.5414174795150757,
+      "learning_rate": 0.0001,
+      "loss": 0.4431,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8382151421719767,
+      "grad_norm": 1.0190706253051758,
+      "learning_rate": 0.0001,
+      "loss": 0.9795,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8383007879410758,
+      "grad_norm": 0.7893204689025879,
+      "learning_rate": 0.0001,
+      "loss": 1.1309,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8383864337101747,
+      "grad_norm": 0.9094313979148865,
+      "learning_rate": 0.0001,
+      "loss": 1.2595,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8384720794792737,
+      "grad_norm": 0.8817195892333984,
+      "learning_rate": 0.0001,
+      "loss": 1.0968,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8385577252483727,
+      "grad_norm": 0.7512573003768921,
+      "learning_rate": 0.0001,
+      "loss": 0.7504,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8386433710174718,
+      "grad_norm": 1.156384825706482,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8387290167865707,
+      "grad_norm": 1.6894835233688354,
+      "learning_rate": 0.0001,
+      "loss": 0.8861,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8388146625556697,
+      "grad_norm": 0.9519880414009094,
+      "learning_rate": 0.0001,
+      "loss": 0.6296,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8389003083247688,
+      "grad_norm": 0.7120647430419922,
+      "learning_rate": 0.0001,
+      "loss": 0.3275,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8389859540938678,
+      "grad_norm": 0.6408304572105408,
+      "learning_rate": 0.0001,
+      "loss": 0.8673,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8390715998629668,
+      "grad_norm": 0.7116587162017822,
+      "learning_rate": 0.0001,
+      "loss": 0.6869,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8391572456320657,
+      "grad_norm": 0.8186467885971069,
+      "learning_rate": 0.0001,
+      "loss": 0.9359,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8392428914011648,
+      "grad_norm": 0.6771461963653564,
+      "learning_rate": 0.0001,
+      "loss": 1.2124,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8393285371702638,
+      "grad_norm": 0.6024441719055176,
+      "learning_rate": 0.0001,
+      "loss": 0.9808,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8394141829393628,
+      "grad_norm": 0.5683155655860901,
+      "learning_rate": 0.0001,
+      "loss": 0.5772,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8394998287084618,
+      "grad_norm": 0.7628741264343262,
+      "learning_rate": 0.0001,
+      "loss": 0.9556,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8395854744775608,
+      "grad_norm": 0.08772283792495728,
+      "learning_rate": 0.0001,
+      "loss": 0.1039,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8396711202466598,
+      "grad_norm": 1.2685428857803345,
+      "learning_rate": 0.0001,
+      "loss": 1.1483,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8397567660157588,
+      "grad_norm": 0.7540746927261353,
+      "learning_rate": 0.0001,
+      "loss": 1.0306,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8398424117848579,
+      "grad_norm": 1.0202332735061646,
+      "learning_rate": 0.0001,
+      "loss": 0.8749,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8399280575539568,
+      "grad_norm": 0.970513641834259,
+      "learning_rate": 0.0001,
+      "loss": 0.729,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8400137033230558,
+      "grad_norm": 1.0042035579681396,
+      "learning_rate": 0.0001,
+      "loss": 0.8238,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8400993490921549,
+      "grad_norm": 0.8418703079223633,
+      "learning_rate": 0.0001,
+      "loss": 0.9794,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8401849948612539,
+      "grad_norm": 0.9047480821609497,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8402706406303528,
+      "grad_norm": 0.8683605790138245,
+      "learning_rate": 0.0001,
+      "loss": 0.864,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8403562863994518,
+      "grad_norm": 0.8170899748802185,
+      "learning_rate": 0.0001,
+      "loss": 1.1283,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8404419321685509,
+      "grad_norm": 1.4880627393722534,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8405275779376499,
+      "grad_norm": 1.5669045448303223,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8406132237067488,
+      "grad_norm": 0.967214822769165,
+      "learning_rate": 0.0001,
+      "loss": 0.6781,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8406988694758479,
+      "grad_norm": 0.8750613331794739,
+      "learning_rate": 0.0001,
+      "loss": 1.0142,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8407845152449469,
+      "grad_norm": 1.2785487174987793,
+      "learning_rate": 0.0001,
+      "loss": 1.2307,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8408701610140459,
+      "grad_norm": 0.5094785690307617,
+      "learning_rate": 0.0001,
+      "loss": 0.7733,
+      "step": 9818
+    },
+    {
+      "epoch": 0.840955806783145,
+      "grad_norm": 0.8990716338157654,
+      "learning_rate": 0.0001,
+      "loss": 0.6033,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8410414525522439,
+      "grad_norm": 1.0070641040802002,
+      "learning_rate": 0.0001,
+      "loss": 0.9707,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8411270983213429,
+      "grad_norm": 0.9723562598228455,
+      "learning_rate": 0.0001,
+      "loss": 0.7893,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8412127440904419,
+      "grad_norm": 1.2897614240646362,
+      "learning_rate": 0.0001,
+      "loss": 0.8731,
+      "step": 9822
+    },
+    {
+      "epoch": 0.841298389859541,
+      "grad_norm": 0.9959054589271545,
+      "learning_rate": 0.0001,
+      "loss": 0.5618,
+      "step": 9823
+    },
+    {
+      "epoch": 0.84138403562864,
+      "grad_norm": 0.6382449269294739,
+      "learning_rate": 0.0001,
+      "loss": 0.3637,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8414696813977389,
+      "grad_norm": 0.5554924011230469,
+      "learning_rate": 0.0001,
+      "loss": 0.9145,
+      "step": 9825
+    },
+    {
+      "epoch": 0.841555327166838,
+      "grad_norm": 1.4682860374450684,
+      "learning_rate": 0.0001,
+      "loss": 1.2564,
+      "step": 9826
+    },
+    {
+      "epoch": 0.841640972935937,
+      "grad_norm": 1.3975017070770264,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 9827
+    },
+    {
+      "epoch": 0.841726618705036,
+      "grad_norm": 0.8769072890281677,
+      "learning_rate": 0.0001,
+      "loss": 1.0073,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8418122644741349,
+      "grad_norm": 0.6007905602455139,
+      "learning_rate": 0.0001,
+      "loss": 1.0816,
+      "step": 9829
+    },
+    {
+      "epoch": 0.841897910243234,
+      "grad_norm": 0.732247531414032,
+      "learning_rate": 0.0001,
+      "loss": 0.5715,
+      "step": 9830
+    },
+    {
+      "epoch": 0.841983556012333,
+      "grad_norm": 0.9063822627067566,
+      "learning_rate": 0.0001,
+      "loss": 0.6993,
+      "step": 9831
+    },
+    {
+      "epoch": 0.842069201781432,
+      "grad_norm": 0.7940922379493713,
+      "learning_rate": 0.0001,
+      "loss": 1.0293,
+      "step": 9832
+    },
+    {
+      "epoch": 0.842154847550531,
+      "grad_norm": 0.8628281950950623,
+      "learning_rate": 0.0001,
+      "loss": 1.0682,
+      "step": 9833
+    },
+    {
+      "epoch": 0.84224049331963,
+      "grad_norm": 0.6933965086936951,
+      "learning_rate": 0.0001,
+      "loss": 1.0352,
+      "step": 9834
+    },
+    {
+      "epoch": 0.842326139088729,
+      "grad_norm": 0.7311537861824036,
+      "learning_rate": 0.0001,
+      "loss": 1.2637,
+      "step": 9835
+    },
+    {
+      "epoch": 0.842411784857828,
+      "grad_norm": 0.5252571702003479,
+      "learning_rate": 0.0001,
+      "loss": 0.8726,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8424974306269271,
+      "grad_norm": 1.000626564025879,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 9837
+    },
+    {
+      "epoch": 0.842583076396026,
+      "grad_norm": 1.4036768674850464,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 9838
+    },
+    {
+      "epoch": 0.842668722165125,
+      "grad_norm": 1.2322627305984497,
+      "learning_rate": 0.0001,
+      "loss": 1.2939,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8427543679342241,
+      "grad_norm": 1.1661878824234009,
+      "learning_rate": 0.0001,
+      "loss": 0.8864,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8428400137033231,
+      "grad_norm": 0.5777583122253418,
+      "learning_rate": 0.0001,
+      "loss": 1.0299,
+      "step": 9841
+    },
+    {
+      "epoch": 0.842925659472422,
+      "grad_norm": 0.5892528295516968,
+      "learning_rate": 0.0001,
+      "loss": 0.9626,
+      "step": 9842
+    },
+    {
+      "epoch": 0.843011305241521,
+      "grad_norm": 0.969569981098175,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8430969510106201,
+      "grad_norm": 1.204911231994629,
+      "learning_rate": 0.0001,
+      "loss": 1.1024,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8431825967797191,
+      "grad_norm": 0.9129303097724915,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 9845
+    },
+    {
+      "epoch": 0.843268242548818,
+      "grad_norm": 1.0808717012405396,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8433538883179171,
+      "grad_norm": 0.7703279852867126,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8434395340870161,
+      "grad_norm": 0.8632581233978271,
+      "learning_rate": 0.0001,
+      "loss": 1.1354,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8435251798561151,
+      "grad_norm": 0.6780834794044495,
+      "learning_rate": 0.0001,
+      "loss": 0.8951,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8436108256252142,
+      "grad_norm": 1.0804728269577026,
+      "learning_rate": 0.0001,
+      "loss": 0.774,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8436964713943131,
+      "grad_norm": 1.2280231714248657,
+      "learning_rate": 0.0001,
+      "loss": 1.3601,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8437821171634121,
+      "grad_norm": 0.5537565350532532,
+      "learning_rate": 0.0001,
+      "loss": 0.7035,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8438677629325111,
+      "grad_norm": 0.8819207549095154,
+      "learning_rate": 0.0001,
+      "loss": 1.1654,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8439534087016102,
+      "grad_norm": 0.9208138585090637,
+      "learning_rate": 0.0001,
+      "loss": 0.6669,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8440390544707091,
+      "grad_norm": 1.3009816408157349,
+      "learning_rate": 0.0001,
+      "loss": 1.1822,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8441247002398081,
+      "grad_norm": 0.6945609450340271,
+      "learning_rate": 0.0001,
+      "loss": 0.7665,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8442103460089072,
+      "grad_norm": 0.5453943610191345,
+      "learning_rate": 0.0001,
+      "loss": 0.9366,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8442959917780062,
+      "grad_norm": 1.0071911811828613,
+      "learning_rate": 0.0001,
+      "loss": 0.9204,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8443816375471052,
+      "grad_norm": 0.5809407234191895,
+      "learning_rate": 0.0001,
+      "loss": 0.6142,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8444672833162041,
+      "grad_norm": 0.5423142313957214,
+      "learning_rate": 0.0001,
+      "loss": 0.9521,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8445529290853032,
+      "grad_norm": 0.7828460931777954,
+      "learning_rate": 0.0001,
+      "loss": 0.3804,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8446385748544022,
+      "grad_norm": 0.7024605870246887,
+      "learning_rate": 0.0001,
+      "loss": 0.9104,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8447242206235012,
+      "grad_norm": 0.7574268579483032,
+      "learning_rate": 0.0001,
+      "loss": 0.8896,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8448098663926003,
+      "grad_norm": 1.1696947813034058,
+      "learning_rate": 0.0001,
+      "loss": 1.4227,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8448955121616992,
+      "grad_norm": 1.196974754333496,
+      "learning_rate": 0.0001,
+      "loss": 1.1386,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8449811579307982,
+      "grad_norm": 0.9606952667236328,
+      "learning_rate": 0.0001,
+      "loss": 0.7425,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8450668036998972,
+      "grad_norm": 1.0681970119476318,
+      "learning_rate": 0.0001,
+      "loss": 0.6744,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8451524494689963,
+      "grad_norm": 0.6885577440261841,
+      "learning_rate": 0.0001,
+      "loss": 1.2595,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8452380952380952,
+      "grad_norm": 0.39081305265426636,
+      "learning_rate": 0.0001,
+      "loss": 0.3577,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8453237410071942,
+      "grad_norm": 0.7357162237167358,
+      "learning_rate": 0.0001,
+      "loss": 0.6712,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8454093867762933,
+      "grad_norm": 0.5469192862510681,
+      "learning_rate": 0.0001,
+      "loss": 0.349,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8454950325453923,
+      "grad_norm": 1.132682204246521,
+      "learning_rate": 0.0001,
+      "loss": 1.1194,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8455806783144912,
+      "grad_norm": 0.6549168229103088,
+      "learning_rate": 0.0001,
+      "loss": 0.9442,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8456663240835903,
+      "grad_norm": 0.5703808665275574,
+      "learning_rate": 0.0001,
+      "loss": 0.9947,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8457519698526893,
+      "grad_norm": 0.9279567003250122,
+      "learning_rate": 0.0001,
+      "loss": 1.2621,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8458376156217883,
+      "grad_norm": 0.729023277759552,
+      "learning_rate": 0.0001,
+      "loss": 0.685,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8459232613908872,
+      "grad_norm": 0.843154788017273,
+      "learning_rate": 0.0001,
+      "loss": 1.1832,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8460089071599863,
+      "grad_norm": 0.8370838761329651,
+      "learning_rate": 0.0001,
+      "loss": 1.1499,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8460945529290853,
+      "grad_norm": 1.0230859518051147,
+      "learning_rate": 0.0001,
+      "loss": 1.2056,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8461801986981843,
+      "grad_norm": 0.7530663013458252,
+      "learning_rate": 0.0001,
+      "loss": 0.4509,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8462658444672834,
+      "grad_norm": 1.2383469343185425,
+      "learning_rate": 0.0001,
+      "loss": 1.0759,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8463514902363823,
+      "grad_norm": 0.9778288006782532,
+      "learning_rate": 0.0001,
+      "loss": 0.7964,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8464371360054813,
+      "grad_norm": 1.4012153148651123,
+      "learning_rate": 0.0001,
+      "loss": 1.1359,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8465227817745803,
+      "grad_norm": 0.8210024833679199,
+      "learning_rate": 0.0001,
+      "loss": 1.321,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8466084275436794,
+      "grad_norm": 0.6288166046142578,
+      "learning_rate": 0.0001,
+      "loss": 0.8608,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8466940733127784,
+      "grad_norm": 0.4665488600730896,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8467797190818773,
+      "grad_norm": 0.9861795902252197,
+      "learning_rate": 0.0001,
+      "loss": 1.0326,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8468653648509764,
+      "grad_norm": 0.8381695747375488,
+      "learning_rate": 0.0001,
+      "loss": 0.8939,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8469510106200754,
+      "grad_norm": 0.7075218558311462,
+      "learning_rate": 0.0001,
+      "loss": 1.2349,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8470366563891744,
+      "grad_norm": 1.1967793703079224,
+      "learning_rate": 0.0001,
+      "loss": 0.9944,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8471223021582733,
+      "grad_norm": 0.8614391088485718,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8472079479273724,
+      "grad_norm": 0.7227154970169067,
+      "learning_rate": 0.0001,
+      "loss": 0.7948,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8472935936964714,
+      "grad_norm": 1.0234262943267822,
+      "learning_rate": 0.0001,
+      "loss": 1.1001,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8473792394655704,
+      "grad_norm": 0.7400779724121094,
+      "learning_rate": 0.0001,
+      "loss": 1.1007,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8474648852346695,
+      "grad_norm": 0.6570312976837158,
+      "learning_rate": 0.0001,
+      "loss": 0.552,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8475505310037684,
+      "grad_norm": 0.7460960149765015,
+      "learning_rate": 0.0001,
+      "loss": 0.9835,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8476361767728674,
+      "grad_norm": 0.6743166446685791,
+      "learning_rate": 0.0001,
+      "loss": 1.0538,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8477218225419664,
+      "grad_norm": 1.8477208614349365,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8478074683110655,
+      "grad_norm": 1.3557132482528687,
+      "learning_rate": 0.0001,
+      "loss": 1.0192,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8478931140801644,
+      "grad_norm": 0.7351353168487549,
+      "learning_rate": 0.0001,
+      "loss": 0.9436,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8479787598492634,
+      "grad_norm": 1.0314180850982666,
+      "learning_rate": 0.0001,
+      "loss": 1.1431,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8480644056183625,
+      "grad_norm": 1.0178865194320679,
+      "learning_rate": 0.0001,
+      "loss": 0.9004,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8481500513874615,
+      "grad_norm": 0.8016180992126465,
+      "learning_rate": 0.0001,
+      "loss": 0.7037,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8482356971565604,
+      "grad_norm": 1.4904981851577759,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8483213429256595,
+      "grad_norm": 0.6610416173934937,
+      "learning_rate": 0.0001,
+      "loss": 1.1503,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8484069886947585,
+      "grad_norm": 0.7583536505699158,
+      "learning_rate": 0.0001,
+      "loss": 1.2478,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8484926344638575,
+      "grad_norm": 0.38624078035354614,
+      "learning_rate": 0.0001,
+      "loss": 0.1628,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8485782802329565,
+      "grad_norm": 0.8495957851409912,
+      "learning_rate": 0.0001,
+      "loss": 0.7799,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8486639260020555,
+      "grad_norm": 1.047749638557434,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8487495717711545,
+      "grad_norm": 0.9380716681480408,
+      "learning_rate": 0.0001,
+      "loss": 1.0456,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8488352175402535,
+      "grad_norm": 0.6716987490653992,
+      "learning_rate": 0.0001,
+      "loss": 1.1891,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8489208633093526,
+      "grad_norm": 1.246104121208191,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8490065090784515,
+      "grad_norm": 0.422940731048584,
+      "learning_rate": 0.0001,
+      "loss": 0.4194,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8490921548475505,
+      "grad_norm": 0.8356874585151672,
+      "learning_rate": 0.0001,
+      "loss": 0.9659,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8491778006166495,
+      "grad_norm": 1.1003682613372803,
+      "learning_rate": 0.0001,
+      "loss": 1.4293,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8492634463857486,
+      "grad_norm": 0.43267151713371277,
+      "learning_rate": 0.0001,
+      "loss": 0.7518,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8493490921548476,
+      "grad_norm": 0.5563209056854248,
+      "learning_rate": 0.0001,
+      "loss": 0.4893,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8494347379239465,
+      "grad_norm": 0.9945854544639587,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8495203836930456,
+      "grad_norm": 0.9712049961090088,
+      "learning_rate": 0.0001,
+      "loss": 1.1266,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8496060294621446,
+      "grad_norm": 0.6948213577270508,
+      "learning_rate": 0.0001,
+      "loss": 0.5126,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8496916752312436,
+      "grad_norm": 1.1462990045547485,
+      "learning_rate": 0.0001,
+      "loss": 0.8903,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8497773210003425,
+      "grad_norm": 0.7263715863227844,
+      "learning_rate": 0.0001,
+      "loss": 0.8068,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8498629667694416,
+      "grad_norm": 1.1669702529907227,
+      "learning_rate": 0.0001,
+      "loss": 0.9692,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8499486125385406,
+      "grad_norm": 0.8165844082832336,
+      "learning_rate": 0.0001,
+      "loss": 0.9893,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8500342583076396,
+      "grad_norm": 0.885449230670929,
+      "learning_rate": 0.0001,
+      "loss": 1.1111,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8501199040767387,
+      "grad_norm": 0.6647763848304749,
+      "learning_rate": 0.0001,
+      "loss": 0.9791,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8502055498458376,
+      "grad_norm": 0.9539822340011597,
+      "learning_rate": 0.0001,
+      "loss": 0.7998,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8502911956149366,
+      "grad_norm": 0.7753962278366089,
+      "learning_rate": 0.0001,
+      "loss": 1.2861,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8503768413840356,
+      "grad_norm": 0.43431904911994934,
+      "learning_rate": 0.0001,
+      "loss": 0.222,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8504624871531347,
+      "grad_norm": 1.4788732528686523,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8505481329222336,
+      "grad_norm": 0.8147912621498108,
+      "learning_rate": 0.0001,
+      "loss": 1.1577,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8506337786913326,
+      "grad_norm": 0.9632115364074707,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8507194244604317,
+      "grad_norm": 1.16718590259552,
+      "learning_rate": 0.0001,
+      "loss": 0.9793,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8508050702295307,
+      "grad_norm": 0.930607259273529,
+      "learning_rate": 0.0001,
+      "loss": 1.1632,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8508907159986296,
+      "grad_norm": 1.2135519981384277,
+      "learning_rate": 0.0001,
+      "loss": 0.8855,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8509763617677287,
+      "grad_norm": 0.5568941235542297,
+      "learning_rate": 0.0001,
+      "loss": 0.8616,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8510620075368277,
+      "grad_norm": 1.343384027481079,
+      "learning_rate": 0.0001,
+      "loss": 1.1484,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8511476533059267,
+      "grad_norm": 0.7458193898200989,
+      "learning_rate": 0.0001,
+      "loss": 0.7389,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8512332990750257,
+      "grad_norm": 0.9220462441444397,
+      "learning_rate": 0.0001,
+      "loss": 1.0542,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8513189448441247,
+      "grad_norm": 0.9783229827880859,
+      "learning_rate": 0.0001,
+      "loss": 0.9065,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8514045906132237,
+      "grad_norm": 1.347597360610962,
+      "learning_rate": 0.0001,
+      "loss": 1.2042,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8514902363823227,
+      "grad_norm": 0.8329741954803467,
+      "learning_rate": 0.0001,
+      "loss": 1.3585,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8515758821514218,
+      "grad_norm": 0.2749738097190857,
+      "learning_rate": 0.0001,
+      "loss": 0.1809,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8516615279205207,
+      "grad_norm": 0.6645426154136658,
+      "learning_rate": 0.0001,
+      "loss": 0.8547,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8517471736896197,
+      "grad_norm": 1.011879563331604,
+      "learning_rate": 0.0001,
+      "loss": 1.0783,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8518328194587187,
+      "grad_norm": 0.8181348443031311,
+      "learning_rate": 0.0001,
+      "loss": 0.8792,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8519184652278178,
+      "grad_norm": 0.9197264909744263,
+      "learning_rate": 0.0001,
+      "loss": 0.979,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8520041109969168,
+      "grad_norm": 0.5920846462249756,
+      "learning_rate": 0.0001,
+      "loss": 0.5519,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8520897567660157,
+      "grad_norm": 1.1461305618286133,
+      "learning_rate": 0.0001,
+      "loss": 1.2759,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8521754025351148,
+      "grad_norm": 0.7353900074958801,
+      "learning_rate": 0.0001,
+      "loss": 1.1678,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8522610483042138,
+      "grad_norm": 0.799923837184906,
+      "learning_rate": 0.0001,
+      "loss": 1.1703,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8523466940733128,
+      "grad_norm": 1.3325282335281372,
+      "learning_rate": 0.0001,
+      "loss": 0.49,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8524323398424117,
+      "grad_norm": 0.7635164856910706,
+      "learning_rate": 0.0001,
+      "loss": 1.0283,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8525179856115108,
+      "grad_norm": 0.7835969924926758,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8526036313806098,
+      "grad_norm": 0.7395520806312561,
+      "learning_rate": 0.0001,
+      "loss": 1.2126,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8526892771497088,
+      "grad_norm": 0.9769800901412964,
+      "learning_rate": 0.0001,
+      "loss": 1.2616,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8527749229188079,
+      "grad_norm": 0.3885398209095001,
+      "learning_rate": 0.0001,
+      "loss": 0.303,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8528605686879068,
+      "grad_norm": 1.0790822505950928,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8529462144570058,
+      "grad_norm": 0.7668962478637695,
+      "learning_rate": 0.0001,
+      "loss": 0.5742,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8530318602261048,
+      "grad_norm": 0.6661116480827332,
+      "learning_rate": 0.0001,
+      "loss": 0.4504,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8531175059952039,
+      "grad_norm": 0.37372666597366333,
+      "learning_rate": 0.0001,
+      "loss": 0.4246,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8532031517643028,
+      "grad_norm": 0.9614447951316833,
+      "learning_rate": 0.0001,
+      "loss": 1.1524,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8532887975334018,
+      "grad_norm": 0.42850956320762634,
+      "learning_rate": 0.0001,
+      "loss": 0.5039,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8533744433025009,
+      "grad_norm": 0.8187220096588135,
+      "learning_rate": 0.0001,
+      "loss": 1.0624,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8534600890715999,
+      "grad_norm": 0.9887098073959351,
+      "learning_rate": 0.0001,
+      "loss": 0.327,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8535457348406988,
+      "grad_norm": 1.1603611707687378,
+      "learning_rate": 0.0001,
+      "loss": 1.426,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8536313806097979,
+      "grad_norm": 0.8611712455749512,
+      "learning_rate": 0.0001,
+      "loss": 0.8581,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8537170263788969,
+      "grad_norm": 0.6785863041877747,
+      "learning_rate": 0.0001,
+      "loss": 0.9824,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8538026721479959,
+      "grad_norm": 0.6144427061080933,
+      "learning_rate": 0.0001,
+      "loss": 0.5446,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8538883179170949,
+      "grad_norm": 0.504302442073822,
+      "learning_rate": 0.0001,
+      "loss": 0.596,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8539739636861939,
+      "grad_norm": 1.4385205507278442,
+      "learning_rate": 0.0001,
+      "loss": 1.1456,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8540596094552929,
+      "grad_norm": 0.8613168001174927,
+      "learning_rate": 0.0001,
+      "loss": 1.1588,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8541452552243919,
+      "grad_norm": 0.9472533464431763,
+      "learning_rate": 0.0001,
+      "loss": 1.1341,
+      "step": 9973
+    },
+    {
+      "epoch": 0.854230900993491,
+      "grad_norm": 0.9669912457466125,
+      "learning_rate": 0.0001,
+      "loss": 1.1664,
+      "step": 9974
+    },
+    {
+      "epoch": 0.85431654676259,
+      "grad_norm": 0.7889193892478943,
+      "learning_rate": 0.0001,
+      "loss": 0.93,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8544021925316889,
+      "grad_norm": 0.9257684350013733,
+      "learning_rate": 0.0001,
+      "loss": 1.2467,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8544878383007879,
+      "grad_norm": 1.7804150581359863,
+      "learning_rate": 0.0001,
+      "loss": 1.004,
+      "step": 9977
+    },
+    {
+      "epoch": 0.854573484069887,
+      "grad_norm": 0.9746298789978027,
+      "learning_rate": 0.0001,
+      "loss": 0.7548,
+      "step": 9978
+    },
+    {
+      "epoch": 0.854659129838986,
+      "grad_norm": 1.245076298713684,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8547447756080849,
+      "grad_norm": 1.0522781610488892,
+      "learning_rate": 0.0001,
+      "loss": 0.9388,
+      "step": 9980
+    },
+    {
+      "epoch": 0.854830421377184,
+      "grad_norm": 0.7883130311965942,
+      "learning_rate": 0.0001,
+      "loss": 0.9646,
+      "step": 9981
+    },
+    {
+      "epoch": 0.854916067146283,
+      "grad_norm": 0.7044639587402344,
+      "learning_rate": 0.0001,
+      "loss": 0.7244,
+      "step": 9982
+    },
+    {
+      "epoch": 0.855001712915382,
+      "grad_norm": 0.743251621723175,
+      "learning_rate": 0.0001,
+      "loss": 0.7505,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8550873586844809,
+      "grad_norm": 0.47834429144859314,
+      "learning_rate": 0.0001,
+      "loss": 0.6761,
+      "step": 9984
+    },
+    {
+      "epoch": 0.85517300445358,
+      "grad_norm": 0.8356747627258301,
+      "learning_rate": 0.0001,
+      "loss": 1.0745,
+      "step": 9985
+    },
+    {
+      "epoch": 0.855258650222679,
+      "grad_norm": 0.6880533695220947,
+      "learning_rate": 0.0001,
+      "loss": 0.9441,
+      "step": 9986
+    },
+    {
+      "epoch": 0.855344295991778,
+      "grad_norm": 0.8885425329208374,
+      "learning_rate": 0.0001,
+      "loss": 0.9807,
+      "step": 9987
+    },
+    {
+      "epoch": 0.855429941760877,
+      "grad_norm": 0.6495847105979919,
+      "learning_rate": 0.0001,
+      "loss": 0.9273,
+      "step": 9988
+    },
+    {
+      "epoch": 0.855515587529976,
+      "grad_norm": 0.6444048285484314,
+      "learning_rate": 0.0001,
+      "loss": 0.6542,
+      "step": 9989
+    },
+    {
+      "epoch": 0.855601233299075,
+      "grad_norm": 0.779570996761322,
+      "learning_rate": 0.0001,
+      "loss": 0.995,
+      "step": 9990
+    },
+    {
+      "epoch": 0.855686879068174,
+      "grad_norm": 0.6653901934623718,
+      "learning_rate": 0.0001,
+      "loss": 1.2888,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8557725248372731,
+      "grad_norm": 1.0049206018447876,
+      "learning_rate": 0.0001,
+      "loss": 1.2369,
+      "step": 9992
+    },
+    {
+      "epoch": 0.855858170606372,
+      "grad_norm": 0.9279667735099792,
+      "learning_rate": 0.0001,
+      "loss": 0.5442,
+      "step": 9993
+    },
+    {
+      "epoch": 0.855943816375471,
+      "grad_norm": 0.4166795015335083,
+      "learning_rate": 0.0001,
+      "loss": 0.2328,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8560294621445701,
+      "grad_norm": 1.203041911125183,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8561151079136691,
+      "grad_norm": 1.4012845754623413,
+      "learning_rate": 0.0001,
+      "loss": 1.1093,
+      "step": 9996
+    },
+    {
+      "epoch": 0.856200753682768,
+      "grad_norm": 1.1105231046676636,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8562863994518671,
+      "grad_norm": 1.2189409732818604,
+      "learning_rate": 0.0001,
+      "loss": 1.2345,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8563720452209661,
+      "grad_norm": 1.1618810892105103,
+      "learning_rate": 0.0001,
+      "loss": 0.8715,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8564576909900651,
+      "grad_norm": 0.8560646772384644,
+      "learning_rate": 0.0001,
+      "loss": 0.8677,
+      "step": 10000
+    },
+    {
+      "epoch": 0.856543336759164,
+      "grad_norm": 1.1334491968154907,
+      "learning_rate": 0.0001,
+      "loss": 0.7463,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8566289825282631,
+      "grad_norm": 0.9366280436515808,
+      "learning_rate": 0.0001,
+      "loss": 1.0208,
+      "step": 10002
+    },
+    {
+      "epoch": 0.8567146282973621,
+      "grad_norm": 0.5994080901145935,
+      "learning_rate": 0.0001,
+      "loss": 0.8013,
+      "step": 10003
+    },
+    {
+      "epoch": 0.8568002740664611,
+      "grad_norm": 0.7833908796310425,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8568859198355602,
+      "grad_norm": 0.5239083170890808,
+      "learning_rate": 0.0001,
+      "loss": 0.802,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8569715656046591,
+      "grad_norm": 0.6637773513793945,
+      "learning_rate": 0.0001,
+      "loss": 0.6999,
+      "step": 10006
+    },
+    {
+      "epoch": 0.8570572113737581,
+      "grad_norm": 0.7701103091239929,
+      "learning_rate": 0.0001,
+      "loss": 0.5867,
+      "step": 10007
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.9444966316223145,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 10008
+    },
+    {
+      "epoch": 0.8572285029119562,
+      "grad_norm": 0.7261773347854614,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8573141486810552,
+      "grad_norm": 0.9396772980690002,
+      "learning_rate": 0.0001,
+      "loss": 0.7551,
+      "step": 10010
+    },
+    {
+      "epoch": 0.8573997944501541,
+      "grad_norm": 0.8077096343040466,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 10011
+    },
+    {
+      "epoch": 0.8574854402192532,
+      "grad_norm": 0.6798161268234253,
+      "learning_rate": 0.0001,
+      "loss": 1.1416,
+      "step": 10012
+    },
+    {
+      "epoch": 0.8575710859883522,
+      "grad_norm": 0.47672536969184875,
+      "learning_rate": 0.0001,
+      "loss": 0.3791,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8576567317574512,
+      "grad_norm": 0.6856245398521423,
+      "learning_rate": 0.0001,
+      "loss": 0.9225,
+      "step": 10014
+    },
+    {
+      "epoch": 0.8577423775265501,
+      "grad_norm": 0.9021053910255432,
+      "learning_rate": 0.0001,
+      "loss": 1.077,
+      "step": 10015
+    },
+    {
+      "epoch": 0.8578280232956492,
+      "grad_norm": 0.7816261053085327,
+      "learning_rate": 0.0001,
+      "loss": 0.978,
+      "step": 10016
+    },
+    {
+      "epoch": 0.8579136690647482,
+      "grad_norm": 0.9314254522323608,
+      "learning_rate": 0.0001,
+      "loss": 0.9631,
+      "step": 10017
+    },
+    {
+      "epoch": 0.8579993148338472,
+      "grad_norm": 0.6566295623779297,
+      "learning_rate": 0.0001,
+      "loss": 1.1701,
+      "step": 10018
+    },
+    {
+      "epoch": 0.8580849606029463,
+      "grad_norm": 1.056299090385437,
+      "learning_rate": 0.0001,
+      "loss": 0.8791,
+      "step": 10019
+    },
+    {
+      "epoch": 0.8581706063720452,
+      "grad_norm": 0.23341424763202667,
+      "learning_rate": 0.0001,
+      "loss": 0.1182,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8582562521411442,
+      "grad_norm": 1.3595507144927979,
+      "learning_rate": 0.0001,
+      "loss": 1.2677,
+      "step": 10021
+    },
+    {
+      "epoch": 0.8583418979102432,
+      "grad_norm": 1.0793980360031128,
+      "learning_rate": 0.0001,
+      "loss": 1.2251,
+      "step": 10022
+    },
+    {
+      "epoch": 0.8584275436793423,
+      "grad_norm": 0.7370421290397644,
+      "learning_rate": 0.0001,
+      "loss": 0.7059,
+      "step": 10023
+    },
+    {
+      "epoch": 0.8585131894484412,
+      "grad_norm": 0.7904725074768066,
+      "learning_rate": 0.0001,
+      "loss": 0.893,
+      "step": 10024
+    },
+    {
+      "epoch": 0.8585988352175402,
+      "grad_norm": 1.3653994798660278,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 10025
+    },
+    {
+      "epoch": 0.8586844809866393,
+      "grad_norm": 1.1656608581542969,
+      "learning_rate": 0.0001,
+      "loss": 1.012,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8587701267557383,
+      "grad_norm": 1.2149062156677246,
+      "learning_rate": 0.0001,
+      "loss": 1.3595,
+      "step": 10027
+    },
+    {
+      "epoch": 0.8588557725248372,
+      "grad_norm": 1.138266921043396,
+      "learning_rate": 0.0001,
+      "loss": 0.9759,
+      "step": 10028
+    },
+    {
+      "epoch": 0.8589414182939363,
+      "grad_norm": 1.3249852657318115,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 10029
+    },
+    {
+      "epoch": 0.8590270640630353,
+      "grad_norm": 0.8875058889389038,
+      "learning_rate": 0.0001,
+      "loss": 1.2146,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8591127098321343,
+      "grad_norm": 0.7989923357963562,
+      "learning_rate": 0.0001,
+      "loss": 0.785,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8591983556012333,
+      "grad_norm": 0.7180217504501343,
+      "learning_rate": 0.0001,
+      "loss": 1.2924,
+      "step": 10032
+    },
+    {
+      "epoch": 0.8592840013703323,
+      "grad_norm": 1.039006233215332,
+      "learning_rate": 0.0001,
+      "loss": 0.6109,
+      "step": 10033
+    },
+    {
+      "epoch": 0.8593696471394313,
+      "grad_norm": 0.9345694780349731,
+      "learning_rate": 0.0001,
+      "loss": 0.7355,
+      "step": 10034
+    },
+    {
+      "epoch": 0.8594552929085303,
+      "grad_norm": 1.208327293395996,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 10035
+    },
+    {
+      "epoch": 0.8595409386776294,
+      "grad_norm": 1.5774720907211304,
+      "learning_rate": 0.0001,
+      "loss": 1.2633,
+      "step": 10036
+    },
+    {
+      "epoch": 0.8596265844467283,
+      "grad_norm": 0.8661912679672241,
+      "learning_rate": 0.0001,
+      "loss": 0.991,
+      "step": 10037
+    },
+    {
+      "epoch": 0.8597122302158273,
+      "grad_norm": 0.7547167539596558,
+      "learning_rate": 0.0001,
+      "loss": 1.0122,
+      "step": 10038
+    },
+    {
+      "epoch": 0.8597978759849263,
+      "grad_norm": 0.7321308255195618,
+      "learning_rate": 0.0001,
+      "loss": 1.3335,
+      "step": 10039
+    },
+    {
+      "epoch": 0.8598835217540254,
+      "grad_norm": 0.5192583203315735,
+      "learning_rate": 0.0001,
+      "loss": 1.1092,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8599691675231244,
+      "grad_norm": 0.6819077134132385,
+      "learning_rate": 0.0001,
+      "loss": 0.5904,
+      "step": 10041
+    },
+    {
+      "epoch": 0.8600548132922233,
+      "grad_norm": 0.7779615521430969,
+      "learning_rate": 0.0001,
+      "loss": 1.1937,
+      "step": 10042
+    },
+    {
+      "epoch": 0.8601404590613224,
+      "grad_norm": 1.0689561367034912,
+      "learning_rate": 0.0001,
+      "loss": 0.8175,
+      "step": 10043
+    },
+    {
+      "epoch": 0.8602261048304214,
+      "grad_norm": 0.9055615067481995,
+      "learning_rate": 0.0001,
+      "loss": 1.0943,
+      "step": 10044
+    },
+    {
+      "epoch": 0.8603117505995204,
+      "grad_norm": 1.4112628698349,
+      "learning_rate": 0.0001,
+      "loss": 0.5731,
+      "step": 10045
+    },
+    {
+      "epoch": 0.8603973963686193,
+      "grad_norm": 0.6782106161117554,
+      "learning_rate": 0.0001,
+      "loss": 0.7237,
+      "step": 10046
+    },
+    {
+      "epoch": 0.8604830421377184,
+      "grad_norm": 1.3129470348358154,
+      "learning_rate": 0.0001,
+      "loss": 1.0181,
+      "step": 10047
+    },
+    {
+      "epoch": 0.8605686879068174,
+      "grad_norm": 1.0494133234024048,
+      "learning_rate": 0.0001,
+      "loss": 1.074,
+      "step": 10048
+    },
+    {
+      "epoch": 0.8606543336759164,
+      "grad_norm": 0.6789389252662659,
+      "learning_rate": 0.0001,
+      "loss": 1.361,
+      "step": 10049
+    },
+    {
+      "epoch": 0.8607399794450155,
+      "grad_norm": 0.7528505921363831,
+      "learning_rate": 0.0001,
+      "loss": 1.1219,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8608256252141144,
+      "grad_norm": 0.942546546459198,
+      "learning_rate": 0.0001,
+      "loss": 0.9006,
+      "step": 10051
+    },
+    {
+      "epoch": 0.8609112709832134,
+      "grad_norm": 0.5561685562133789,
+      "learning_rate": 0.0001,
+      "loss": 1.0182,
+      "step": 10052
+    },
+    {
+      "epoch": 0.8609969167523125,
+      "grad_norm": 0.8359872102737427,
+      "learning_rate": 0.0001,
+      "loss": 0.7796,
+      "step": 10053
+    },
+    {
+      "epoch": 0.8610825625214115,
+      "grad_norm": 1.1051985025405884,
+      "learning_rate": 0.0001,
+      "loss": 1.3235,
+      "step": 10054
+    },
+    {
+      "epoch": 0.8611682082905104,
+      "grad_norm": 1.2225656509399414,
+      "learning_rate": 0.0001,
+      "loss": 1.1503,
+      "step": 10055
+    },
+    {
+      "epoch": 0.8612538540596094,
+      "grad_norm": 1.0812900066375732,
+      "learning_rate": 0.0001,
+      "loss": 1.0512,
+      "step": 10056
+    },
+    {
+      "epoch": 0.8613394998287085,
+      "grad_norm": 0.6007379293441772,
+      "learning_rate": 0.0001,
+      "loss": 0.7816,
+      "step": 10057
+    },
+    {
+      "epoch": 0.8614251455978075,
+      "grad_norm": 1.1273895502090454,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 10058
+    },
+    {
+      "epoch": 0.8615107913669064,
+      "grad_norm": 1.5573025941848755,
+      "learning_rate": 0.0001,
+      "loss": 1.9482,
+      "step": 10059
+    },
+    {
+      "epoch": 0.8615964371360055,
+      "grad_norm": 0.8268787264823914,
+      "learning_rate": 0.0001,
+      "loss": 0.7242,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8616820829051045,
+      "grad_norm": 0.8572856187820435,
+      "learning_rate": 0.0001,
+      "loss": 0.8605,
+      "step": 10061
+    },
+    {
+      "epoch": 0.8617677286742035,
+      "grad_norm": 0.9962348341941833,
+      "learning_rate": 0.0001,
+      "loss": 1.3876,
+      "step": 10062
+    },
+    {
+      "epoch": 0.8618533744433025,
+      "grad_norm": 1.2731298208236694,
+      "learning_rate": 0.0001,
+      "loss": 0.9631,
+      "step": 10063
+    },
+    {
+      "epoch": 0.8619390202124015,
+      "grad_norm": 1.1746459007263184,
+      "learning_rate": 0.0001,
+      "loss": 1.2536,
+      "step": 10064
+    },
+    {
+      "epoch": 0.8620246659815005,
+      "grad_norm": 0.736829400062561,
+      "learning_rate": 0.0001,
+      "loss": 0.917,
+      "step": 10065
+    },
+    {
+      "epoch": 0.8621103117505995,
+      "grad_norm": 1.0870239734649658,
+      "learning_rate": 0.0001,
+      "loss": 1.1632,
+      "step": 10066
+    },
+    {
+      "epoch": 0.8621959575196986,
+      "grad_norm": 1.3886439800262451,
+      "learning_rate": 0.0001,
+      "loss": 0.7251,
+      "step": 10067
+    },
+    {
+      "epoch": 0.8622816032887975,
+      "grad_norm": 0.6010820865631104,
+      "learning_rate": 0.0001,
+      "loss": 0.982,
+      "step": 10068
+    },
+    {
+      "epoch": 0.8623672490578965,
+      "grad_norm": 0.8117319345474243,
+      "learning_rate": 0.0001,
+      "loss": 1.1075,
+      "step": 10069
+    },
+    {
+      "epoch": 0.8624528948269955,
+      "grad_norm": 1.051358938217163,
+      "learning_rate": 0.0001,
+      "loss": 1.0055,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8625385405960946,
+      "grad_norm": 0.903662919998169,
+      "learning_rate": 0.0001,
+      "loss": 0.98,
+      "step": 10071
+    },
+    {
+      "epoch": 0.8626241863651936,
+      "grad_norm": 0.5443506836891174,
+      "learning_rate": 0.0001,
+      "loss": 0.9123,
+      "step": 10072
+    },
+    {
+      "epoch": 0.8627098321342925,
+      "grad_norm": 1.325753092765808,
+      "learning_rate": 0.0001,
+      "loss": 0.7511,
+      "step": 10073
+    },
+    {
+      "epoch": 0.8627954779033916,
+      "grad_norm": 0.6324295997619629,
+      "learning_rate": 0.0001,
+      "loss": 0.9505,
+      "step": 10074
+    },
+    {
+      "epoch": 0.8628811236724906,
+      "grad_norm": 0.8398393392562866,
+      "learning_rate": 0.0001,
+      "loss": 0.9329,
+      "step": 10075
+    },
+    {
+      "epoch": 0.8629667694415896,
+      "grad_norm": 1.1704521179199219,
+      "learning_rate": 0.0001,
+      "loss": 1.2498,
+      "step": 10076
+    },
+    {
+      "epoch": 0.8630524152106885,
+      "grad_norm": 1.007563591003418,
+      "learning_rate": 0.0001,
+      "loss": 0.8502,
+      "step": 10077
+    },
+    {
+      "epoch": 0.8631380609797876,
+      "grad_norm": 1.116469144821167,
+      "learning_rate": 0.0001,
+      "loss": 1.2365,
+      "step": 10078
+    },
+    {
+      "epoch": 0.8632237067488866,
+      "grad_norm": 0.9276856780052185,
+      "learning_rate": 0.0001,
+      "loss": 0.5653,
+      "step": 10079
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.9315510392189026,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8633949982870847,
+      "grad_norm": 0.5126023888587952,
+      "learning_rate": 0.0001,
+      "loss": 0.4233,
+      "step": 10081
+    },
+    {
+      "epoch": 0.8634806440561836,
+      "grad_norm": 0.9486197233200073,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 10082
+    },
+    {
+      "epoch": 0.8635662898252826,
+      "grad_norm": 0.7383543252944946,
+      "learning_rate": 0.0001,
+      "loss": 0.2472,
+      "step": 10083
+    },
+    {
+      "epoch": 0.8636519355943817,
+      "grad_norm": 0.76557856798172,
+      "learning_rate": 0.0001,
+      "loss": 1.0885,
+      "step": 10084
+    },
+    {
+      "epoch": 0.8637375813634807,
+      "grad_norm": 0.9554963707923889,
+      "learning_rate": 0.0001,
+      "loss": 1.0319,
+      "step": 10085
+    },
+    {
+      "epoch": 0.8638232271325796,
+      "grad_norm": 0.6834177374839783,
+      "learning_rate": 0.0001,
+      "loss": 0.8311,
+      "step": 10086
+    },
+    {
+      "epoch": 0.8639088729016786,
+      "grad_norm": 0.9908255338668823,
+      "learning_rate": 0.0001,
+      "loss": 1.1539,
+      "step": 10087
+    },
+    {
+      "epoch": 0.8639945186707777,
+      "grad_norm": 0.9155910015106201,
+      "learning_rate": 0.0001,
+      "loss": 1.0008,
+      "step": 10088
+    },
+    {
+      "epoch": 0.8640801644398767,
+      "grad_norm": 0.887785017490387,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 10089
+    },
+    {
+      "epoch": 0.8641658102089756,
+      "grad_norm": 1.0139455795288086,
+      "learning_rate": 0.0001,
+      "loss": 1.042,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8642514559780747,
+      "grad_norm": 1.3374961614608765,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 10091
+    },
+    {
+      "epoch": 0.8643371017471737,
+      "grad_norm": 1.2422674894332886,
+      "learning_rate": 0.0001,
+      "loss": 1.1006,
+      "step": 10092
+    },
+    {
+      "epoch": 0.8644227475162727,
+      "grad_norm": 0.8941219449043274,
+      "learning_rate": 0.0001,
+      "loss": 0.977,
+      "step": 10093
+    },
+    {
+      "epoch": 0.8645083932853717,
+      "grad_norm": 0.7224281430244446,
+      "learning_rate": 0.0001,
+      "loss": 0.3869,
+      "step": 10094
+    },
+    {
+      "epoch": 0.8645940390544707,
+      "grad_norm": 0.4861789047718048,
+      "learning_rate": 0.0001,
+      "loss": 0.5783,
+      "step": 10095
+    },
+    {
+      "epoch": 0.8646796848235697,
+      "grad_norm": 0.6995654106140137,
+      "learning_rate": 0.0001,
+      "loss": 0.7597,
+      "step": 10096
+    },
+    {
+      "epoch": 0.8647653305926687,
+      "grad_norm": 0.7096772789955139,
+      "learning_rate": 0.0001,
+      "loss": 0.9483,
+      "step": 10097
+    },
+    {
+      "epoch": 0.8648509763617678,
+      "grad_norm": 1.2055054903030396,
+      "learning_rate": 0.0001,
+      "loss": 1.1977,
+      "step": 10098
+    },
+    {
+      "epoch": 0.8649366221308668,
+      "grad_norm": 0.8064544796943665,
+      "learning_rate": 0.0001,
+      "loss": 0.6947,
+      "step": 10099
+    },
+    {
+      "epoch": 0.8650222678999657,
+      "grad_norm": 0.7597392201423645,
+      "learning_rate": 0.0001,
+      "loss": 0.8012,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8651079136690647,
+      "grad_norm": 0.8145788311958313,
+      "learning_rate": 0.0001,
+      "loss": 1.1352,
+      "step": 10101
+    },
+    {
+      "epoch": 0.8651935594381638,
+      "grad_norm": 0.6813762187957764,
+      "learning_rate": 0.0001,
+      "loss": 0.889,
+      "step": 10102
+    },
+    {
+      "epoch": 0.8652792052072628,
+      "grad_norm": 0.7722563743591309,
+      "learning_rate": 0.0001,
+      "loss": 0.8115,
+      "step": 10103
+    },
+    {
+      "epoch": 0.8653648509763617,
+      "grad_norm": 0.8164512515068054,
+      "learning_rate": 0.0001,
+      "loss": 1.3154,
+      "step": 10104
+    },
+    {
+      "epoch": 0.8654504967454608,
+      "grad_norm": 0.8505242466926575,
+      "learning_rate": 0.0001,
+      "loss": 0.6368,
+      "step": 10105
+    },
+    {
+      "epoch": 0.8655361425145598,
+      "grad_norm": 1.204026699066162,
+      "learning_rate": 0.0001,
+      "loss": 1.0414,
+      "step": 10106
+    },
+    {
+      "epoch": 0.8656217882836588,
+      "grad_norm": 0.7325835824012756,
+      "learning_rate": 0.0001,
+      "loss": 0.7439,
+      "step": 10107
+    },
+    {
+      "epoch": 0.8657074340527577,
+      "grad_norm": 0.8736917972564697,
+      "learning_rate": 0.0001,
+      "loss": 0.5779,
+      "step": 10108
+    },
+    {
+      "epoch": 0.8657930798218568,
+      "grad_norm": 0.8494018316268921,
+      "learning_rate": 0.0001,
+      "loss": 0.7834,
+      "step": 10109
+    },
+    {
+      "epoch": 0.8658787255909558,
+      "grad_norm": 1.0115816593170166,
+      "learning_rate": 0.0001,
+      "loss": 0.7,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8659643713600548,
+      "grad_norm": 1.2902806997299194,
+      "learning_rate": 0.0001,
+      "loss": 0.9237,
+      "step": 10111
+    },
+    {
+      "epoch": 0.8660500171291539,
+      "grad_norm": 0.8107749819755554,
+      "learning_rate": 0.0001,
+      "loss": 1.0811,
+      "step": 10112
+    },
+    {
+      "epoch": 0.8661356628982528,
+      "grad_norm": 0.9279272556304932,
+      "learning_rate": 0.0001,
+      "loss": 1.1034,
+      "step": 10113
+    },
+    {
+      "epoch": 0.8662213086673518,
+      "grad_norm": 0.8327707648277283,
+      "learning_rate": 0.0001,
+      "loss": 0.4324,
+      "step": 10114
+    },
+    {
+      "epoch": 0.8663069544364509,
+      "grad_norm": 1.289818286895752,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 10115
+    },
+    {
+      "epoch": 0.8663926002055499,
+      "grad_norm": 1.3582873344421387,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 10116
+    },
+    {
+      "epoch": 0.8664782459746488,
+      "grad_norm": 0.7290763258934021,
+      "learning_rate": 0.0001,
+      "loss": 0.5903,
+      "step": 10117
+    },
+    {
+      "epoch": 0.8665638917437478,
+      "grad_norm": 0.9157921075820923,
+      "learning_rate": 0.0001,
+      "loss": 0.9007,
+      "step": 10118
+    },
+    {
+      "epoch": 0.8666495375128469,
+      "grad_norm": 1.3398851156234741,
+      "learning_rate": 0.0001,
+      "loss": 1.162,
+      "step": 10119
+    },
+    {
+      "epoch": 0.8667351832819459,
+      "grad_norm": 0.9511755704879761,
+      "learning_rate": 0.0001,
+      "loss": 1.2337,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8668208290510448,
+      "grad_norm": 1.1179531812667847,
+      "learning_rate": 0.0001,
+      "loss": 1.1397,
+      "step": 10121
+    },
+    {
+      "epoch": 0.8669064748201439,
+      "grad_norm": 0.5429726839065552,
+      "learning_rate": 0.0001,
+      "loss": 0.3654,
+      "step": 10122
+    },
+    {
+      "epoch": 0.8669921205892429,
+      "grad_norm": 0.7462955713272095,
+      "learning_rate": 0.0001,
+      "loss": 0.7866,
+      "step": 10123
+    },
+    {
+      "epoch": 0.8670777663583419,
+      "grad_norm": 1.1451958417892456,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 10124
+    },
+    {
+      "epoch": 0.8671634121274409,
+      "grad_norm": 0.6979402899742126,
+      "learning_rate": 0.0001,
+      "loss": 0.6935,
+      "step": 10125
+    },
+    {
+      "epoch": 0.8672490578965399,
+      "grad_norm": 0.902165949344635,
+      "learning_rate": 0.0001,
+      "loss": 1.4511,
+      "step": 10126
+    },
+    {
+      "epoch": 0.8673347036656389,
+      "grad_norm": 1.2771577835083008,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 10127
+    },
+    {
+      "epoch": 0.8674203494347379,
+      "grad_norm": 0.8340098857879639,
+      "learning_rate": 0.0001,
+      "loss": 0.7481,
+      "step": 10128
+    },
+    {
+      "epoch": 0.867505995203837,
+      "grad_norm": 0.6240222454071045,
+      "learning_rate": 0.0001,
+      "loss": 0.9061,
+      "step": 10129
+    },
+    {
+      "epoch": 0.867591640972936,
+      "grad_norm": 0.2898057699203491,
+      "learning_rate": 0.0001,
+      "loss": 0.5295,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8676772867420349,
+      "grad_norm": 0.6559683084487915,
+      "learning_rate": 0.0001,
+      "loss": 0.9267,
+      "step": 10131
+    },
+    {
+      "epoch": 0.8677629325111339,
+      "grad_norm": 0.9797573685646057,
+      "learning_rate": 0.0001,
+      "loss": 1.1163,
+      "step": 10132
+    },
+    {
+      "epoch": 0.867848578280233,
+      "grad_norm": 0.829511821269989,
+      "learning_rate": 0.0001,
+      "loss": 1.0104,
+      "step": 10133
+    },
+    {
+      "epoch": 0.867934224049332,
+      "grad_norm": 1.1524680852890015,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 10134
+    },
+    {
+      "epoch": 0.8680198698184309,
+      "grad_norm": 0.9437550902366638,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 10135
+    },
+    {
+      "epoch": 0.86810551558753,
+      "grad_norm": 1.2062712907791138,
+      "learning_rate": 0.0001,
+      "loss": 1.0633,
+      "step": 10136
+    },
+    {
+      "epoch": 0.868191161356629,
+      "grad_norm": 1.0812649726867676,
+      "learning_rate": 0.0001,
+      "loss": 1.3017,
+      "step": 10137
+    },
+    {
+      "epoch": 0.868276807125728,
+      "grad_norm": 0.7041144967079163,
+      "learning_rate": 0.0001,
+      "loss": 0.9348,
+      "step": 10138
+    },
+    {
+      "epoch": 0.8683624528948269,
+      "grad_norm": 0.4856644570827484,
+      "learning_rate": 0.0001,
+      "loss": 0.4826,
+      "step": 10139
+    },
+    {
+      "epoch": 0.868448098663926,
+      "grad_norm": 0.5853478908538818,
+      "learning_rate": 0.0001,
+      "loss": 0.6515,
+      "step": 10140
+    },
+    {
+      "epoch": 0.868533744433025,
+      "grad_norm": 0.8236793279647827,
+      "learning_rate": 0.0001,
+      "loss": 0.5206,
+      "step": 10141
+    },
+    {
+      "epoch": 0.868619390202124,
+      "grad_norm": 0.9554690718650818,
+      "learning_rate": 0.0001,
+      "loss": 0.2652,
+      "step": 10142
+    },
+    {
+      "epoch": 0.8687050359712231,
+      "grad_norm": 0.6796308755874634,
+      "learning_rate": 0.0001,
+      "loss": 1.1529,
+      "step": 10143
+    },
+    {
+      "epoch": 0.868790681740322,
+      "grad_norm": 1.0171515941619873,
+      "learning_rate": 0.0001,
+      "loss": 1.236,
+      "step": 10144
+    },
+    {
+      "epoch": 0.868876327509421,
+      "grad_norm": 1.3083491325378418,
+      "learning_rate": 0.0001,
+      "loss": 1.1044,
+      "step": 10145
+    },
+    {
+      "epoch": 0.8689619732785201,
+      "grad_norm": 1.006118655204773,
+      "learning_rate": 0.0001,
+      "loss": 0.7328,
+      "step": 10146
+    },
+    {
+      "epoch": 0.8690476190476191,
+      "grad_norm": 0.5817467570304871,
+      "learning_rate": 0.0001,
+      "loss": 0.9927,
+      "step": 10147
+    },
+    {
+      "epoch": 0.869133264816718,
+      "grad_norm": 0.7284364104270935,
+      "learning_rate": 0.0001,
+      "loss": 1.1777,
+      "step": 10148
+    },
+    {
+      "epoch": 0.869218910585817,
+      "grad_norm": 0.6377183794975281,
+      "learning_rate": 0.0001,
+      "loss": 1.077,
+      "step": 10149
+    },
+    {
+      "epoch": 0.8693045563549161,
+      "grad_norm": 0.6921759843826294,
+      "learning_rate": 0.0001,
+      "loss": 0.8146,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8693902021240151,
+      "grad_norm": 1.1584585905075073,
+      "learning_rate": 0.0001,
+      "loss": 0.7177,
+      "step": 10151
+    },
+    {
+      "epoch": 0.869475847893114,
+      "grad_norm": 0.7537373900413513,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 10152
+    },
+    {
+      "epoch": 0.8695614936622131,
+      "grad_norm": 1.2372499704360962,
+      "learning_rate": 0.0001,
+      "loss": 1.0022,
+      "step": 10153
+    },
+    {
+      "epoch": 0.8696471394313121,
+      "grad_norm": 0.7706846594810486,
+      "learning_rate": 0.0001,
+      "loss": 0.6798,
+      "step": 10154
+    },
+    {
+      "epoch": 0.8697327852004111,
+      "grad_norm": 1.1987336874008179,
+      "learning_rate": 0.0001,
+      "loss": 0.9265,
+      "step": 10155
+    },
+    {
+      "epoch": 0.8698184309695101,
+      "grad_norm": 0.7132117748260498,
+      "learning_rate": 0.0001,
+      "loss": 1.2107,
+      "step": 10156
+    },
+    {
+      "epoch": 0.8699040767386091,
+      "grad_norm": 0.705583930015564,
+      "learning_rate": 0.0001,
+      "loss": 1.2826,
+      "step": 10157
+    },
+    {
+      "epoch": 0.8699897225077081,
+      "grad_norm": 0.894326388835907,
+      "learning_rate": 0.0001,
+      "loss": 1.0545,
+      "step": 10158
+    },
+    {
+      "epoch": 0.8700753682768071,
+      "grad_norm": 0.8116465210914612,
+      "learning_rate": 0.0001,
+      "loss": 0.6926,
+      "step": 10159
+    },
+    {
+      "epoch": 0.8701610140459062,
+      "grad_norm": 1.0868204832077026,
+      "learning_rate": 0.0001,
+      "loss": 0.834,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8702466598150052,
+      "grad_norm": 0.9885972142219543,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 10161
+    },
+    {
+      "epoch": 0.8703323055841041,
+      "grad_norm": 1.2373440265655518,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 10162
+    },
+    {
+      "epoch": 0.8704179513532031,
+      "grad_norm": 1.1013433933258057,
+      "learning_rate": 0.0001,
+      "loss": 0.7293,
+      "step": 10163
+    },
+    {
+      "epoch": 0.8705035971223022,
+      "grad_norm": 0.5925343036651611,
+      "learning_rate": 0.0001,
+      "loss": 1.0034,
+      "step": 10164
+    },
+    {
+      "epoch": 0.8705892428914012,
+      "grad_norm": 0.798409104347229,
+      "learning_rate": 0.0001,
+      "loss": 0.7617,
+      "step": 10165
+    },
+    {
+      "epoch": 0.8706748886605001,
+      "grad_norm": 1.1905858516693115,
+      "learning_rate": 0.0001,
+      "loss": 0.8582,
+      "step": 10166
+    },
+    {
+      "epoch": 0.8707605344295992,
+      "grad_norm": 1.2156519889831543,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 10167
+    },
+    {
+      "epoch": 0.8708461801986982,
+      "grad_norm": 0.7930446863174438,
+      "learning_rate": 0.0001,
+      "loss": 0.5368,
+      "step": 10168
+    },
+    {
+      "epoch": 0.8709318259677972,
+      "grad_norm": 0.6119534969329834,
+      "learning_rate": 0.0001,
+      "loss": 0.5185,
+      "step": 10169
+    },
+    {
+      "epoch": 0.8710174717368961,
+      "grad_norm": 0.9186002612113953,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8711031175059952,
+      "grad_norm": 1.2002860307693481,
+      "learning_rate": 0.0001,
+      "loss": 1.0102,
+      "step": 10171
+    },
+    {
+      "epoch": 0.8711887632750942,
+      "grad_norm": 0.6799396276473999,
+      "learning_rate": 0.0001,
+      "loss": 1.0602,
+      "step": 10172
+    },
+    {
+      "epoch": 0.8712744090441932,
+      "grad_norm": 0.8330284953117371,
+      "learning_rate": 0.0001,
+      "loss": 1.0238,
+      "step": 10173
+    },
+    {
+      "epoch": 0.8713600548132923,
+      "grad_norm": 0.8716406226158142,
+      "learning_rate": 0.0001,
+      "loss": 1.0754,
+      "step": 10174
+    },
+    {
+      "epoch": 0.8714457005823912,
+      "grad_norm": 0.8478915095329285,
+      "learning_rate": 0.0001,
+      "loss": 0.9213,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8715313463514902,
+      "grad_norm": 1.1345592737197876,
+      "learning_rate": 0.0001,
+      "loss": 1.2181,
+      "step": 10176
+    },
+    {
+      "epoch": 0.8716169921205893,
+      "grad_norm": 1.043449878692627,
+      "learning_rate": 0.0001,
+      "loss": 1.276,
+      "step": 10177
+    },
+    {
+      "epoch": 0.8717026378896883,
+      "grad_norm": 0.8176807761192322,
+      "learning_rate": 0.0001,
+      "loss": 1.2509,
+      "step": 10178
+    },
+    {
+      "epoch": 0.8717882836587872,
+      "grad_norm": 1.444312334060669,
+      "learning_rate": 0.0001,
+      "loss": 1.0386,
+      "step": 10179
+    },
+    {
+      "epoch": 0.8718739294278862,
+      "grad_norm": 1.0019983053207397,
+      "learning_rate": 0.0001,
+      "loss": 0.8603,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8719595751969853,
+      "grad_norm": 0.6414742469787598,
+      "learning_rate": 0.0001,
+      "loss": 0.9097,
+      "step": 10181
+    },
+    {
+      "epoch": 0.8720452209660843,
+      "grad_norm": 0.7481512427330017,
+      "learning_rate": 0.0001,
+      "loss": 0.9163,
+      "step": 10182
+    },
+    {
+      "epoch": 0.8721308667351833,
+      "grad_norm": 0.9311866760253906,
+      "learning_rate": 0.0001,
+      "loss": 0.9325,
+      "step": 10183
+    },
+    {
+      "epoch": 0.8722165125042823,
+      "grad_norm": 0.8178591132164001,
+      "learning_rate": 0.0001,
+      "loss": 1.0862,
+      "step": 10184
+    },
+    {
+      "epoch": 0.8723021582733813,
+      "grad_norm": 1.3257595300674438,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 10185
+    },
+    {
+      "epoch": 0.8723878040424803,
+      "grad_norm": 0.6666454076766968,
+      "learning_rate": 0.0001,
+      "loss": 1.0694,
+      "step": 10186
+    },
+    {
+      "epoch": 0.8724734498115793,
+      "grad_norm": 0.8725594878196716,
+      "learning_rate": 0.0001,
+      "loss": 1.1625,
+      "step": 10187
+    },
+    {
+      "epoch": 0.8725590955806783,
+      "grad_norm": 1.6387277841567993,
+      "learning_rate": 0.0001,
+      "loss": 1.0654,
+      "step": 10188
+    },
+    {
+      "epoch": 0.8726447413497773,
+      "grad_norm": 0.4826602637767792,
+      "learning_rate": 0.0001,
+      "loss": 0.7106,
+      "step": 10189
+    },
+    {
+      "epoch": 0.8727303871188763,
+      "grad_norm": 0.7309057116508484,
+      "learning_rate": 0.0001,
+      "loss": 1.0695,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8728160328879754,
+      "grad_norm": 0.9925447106361389,
+      "learning_rate": 0.0001,
+      "loss": 0.8053,
+      "step": 10191
+    },
+    {
+      "epoch": 0.8729016786570744,
+      "grad_norm": 0.6202601790428162,
+      "learning_rate": 0.0001,
+      "loss": 0.9701,
+      "step": 10192
+    },
+    {
+      "epoch": 0.8729873244261733,
+      "grad_norm": 0.5798255801200867,
+      "learning_rate": 0.0001,
+      "loss": 0.8027,
+      "step": 10193
+    },
+    {
+      "epoch": 0.8730729701952723,
+      "grad_norm": 0.6535800099372864,
+      "learning_rate": 0.0001,
+      "loss": 0.6889,
+      "step": 10194
+    },
+    {
+      "epoch": 0.8731586159643714,
+      "grad_norm": 0.907902717590332,
+      "learning_rate": 0.0001,
+      "loss": 0.6298,
+      "step": 10195
+    },
+    {
+      "epoch": 0.8732442617334704,
+      "grad_norm": 1.0092391967773438,
+      "learning_rate": 0.0001,
+      "loss": 0.8504,
+      "step": 10196
+    },
+    {
+      "epoch": 0.8733299075025693,
+      "grad_norm": 1.2122652530670166,
+      "learning_rate": 0.0001,
+      "loss": 0.9516,
+      "step": 10197
+    },
+    {
+      "epoch": 0.8734155532716684,
+      "grad_norm": 0.9041799902915955,
+      "learning_rate": 0.0001,
+      "loss": 1.0559,
+      "step": 10198
+    },
+    {
+      "epoch": 0.8735011990407674,
+      "grad_norm": 0.8592575192451477,
+      "learning_rate": 0.0001,
+      "loss": 0.7521,
+      "step": 10199
+    },
+    {
+      "epoch": 0.8735868448098664,
+      "grad_norm": 0.9126320481300354,
+      "learning_rate": 0.0001,
+      "loss": 1.1914,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8736724905789653,
+      "grad_norm": 0.8037563562393188,
+      "learning_rate": 0.0001,
+      "loss": 0.5422,
+      "step": 10201
+    },
+    {
+      "epoch": 0.8737581363480644,
+      "grad_norm": 0.5054523348808289,
+      "learning_rate": 0.0001,
+      "loss": 0.5123,
+      "step": 10202
+    },
+    {
+      "epoch": 0.8738437821171634,
+      "grad_norm": 0.8375506401062012,
+      "learning_rate": 0.0001,
+      "loss": 0.8448,
+      "step": 10203
+    },
+    {
+      "epoch": 0.8739294278862624,
+      "grad_norm": 1.7300459146499634,
+      "learning_rate": 0.0001,
+      "loss": 2.1806,
+      "step": 10204
+    },
+    {
+      "epoch": 0.8740150736553615,
+      "grad_norm": 0.7640141844749451,
+      "learning_rate": 0.0001,
+      "loss": 0.6838,
+      "step": 10205
+    },
+    {
+      "epoch": 0.8741007194244604,
+      "grad_norm": 0.8845239877700806,
+      "learning_rate": 0.0001,
+      "loss": 1.0845,
+      "step": 10206
+    },
+    {
+      "epoch": 0.8741863651935594,
+      "grad_norm": 1.1035317182540894,
+      "learning_rate": 0.0001,
+      "loss": 1.0844,
+      "step": 10207
+    },
+    {
+      "epoch": 0.8742720109626585,
+      "grad_norm": 1.477257251739502,
+      "learning_rate": 0.0001,
+      "loss": 0.8881,
+      "step": 10208
+    },
+    {
+      "epoch": 0.8743576567317575,
+      "grad_norm": 1.0674898624420166,
+      "learning_rate": 0.0001,
+      "loss": 1.168,
+      "step": 10209
+    },
+    {
+      "epoch": 0.8744433025008564,
+      "grad_norm": 0.8231119513511658,
+      "learning_rate": 0.0001,
+      "loss": 0.99,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8745289482699554,
+      "grad_norm": 0.38333839178085327,
+      "learning_rate": 0.0001,
+      "loss": 0.4882,
+      "step": 10211
+    },
+    {
+      "epoch": 0.8746145940390545,
+      "grad_norm": 1.2798635959625244,
+      "learning_rate": 0.0001,
+      "loss": 1.1357,
+      "step": 10212
+    },
+    {
+      "epoch": 0.8747002398081535,
+      "grad_norm": 0.36073875427246094,
+      "learning_rate": 0.0001,
+      "loss": 0.2623,
+      "step": 10213
+    },
+    {
+      "epoch": 0.8747858855772525,
+      "grad_norm": 0.636972963809967,
+      "learning_rate": 0.0001,
+      "loss": 1.0558,
+      "step": 10214
+    },
+    {
+      "epoch": 0.8748715313463515,
+      "grad_norm": 1.2577779293060303,
+      "learning_rate": 0.0001,
+      "loss": 0.5545,
+      "step": 10215
+    },
+    {
+      "epoch": 0.8749571771154505,
+      "grad_norm": 0.7451605796813965,
+      "learning_rate": 0.0001,
+      "loss": 0.7605,
+      "step": 10216
+    },
+    {
+      "epoch": 0.8750428228845495,
+      "grad_norm": 0.7505033016204834,
+      "learning_rate": 0.0001,
+      "loss": 1.143,
+      "step": 10217
+    },
+    {
+      "epoch": 0.8751284686536485,
+      "grad_norm": 1.0398789644241333,
+      "learning_rate": 0.0001,
+      "loss": 1.1531,
+      "step": 10218
+    },
+    {
+      "epoch": 0.8752141144227475,
+      "grad_norm": 0.9302664995193481,
+      "learning_rate": 0.0001,
+      "loss": 0.6399,
+      "step": 10219
+    },
+    {
+      "epoch": 0.8752997601918465,
+      "grad_norm": 0.8488762378692627,
+      "learning_rate": 0.0001,
+      "loss": 0.9322,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8753854059609455,
+      "grad_norm": 0.6485114097595215,
+      "learning_rate": 0.0001,
+      "loss": 0.9503,
+      "step": 10221
+    },
+    {
+      "epoch": 0.8754710517300446,
+      "grad_norm": 0.7768989205360413,
+      "learning_rate": 0.0001,
+      "loss": 0.787,
+      "step": 10222
+    },
+    {
+      "epoch": 0.8755566974991436,
+      "grad_norm": 0.9626396894454956,
+      "learning_rate": 0.0001,
+      "loss": 0.9154,
+      "step": 10223
+    },
+    {
+      "epoch": 0.8756423432682425,
+      "grad_norm": 0.7402014136314392,
+      "learning_rate": 0.0001,
+      "loss": 0.6217,
+      "step": 10224
+    },
+    {
+      "epoch": 0.8757279890373415,
+      "grad_norm": 1.8008317947387695,
+      "learning_rate": 0.0001,
+      "loss": 1.1822,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8758136348064406,
+      "grad_norm": 1.3239223957061768,
+      "learning_rate": 0.0001,
+      "loss": 0.5087,
+      "step": 10226
+    },
+    {
+      "epoch": 0.8758992805755396,
+      "grad_norm": 0.7169291377067566,
+      "learning_rate": 0.0001,
+      "loss": 0.6055,
+      "step": 10227
+    },
+    {
+      "epoch": 0.8759849263446385,
+      "grad_norm": 0.8275295495986938,
+      "learning_rate": 0.0001,
+      "loss": 1.1605,
+      "step": 10228
+    },
+    {
+      "epoch": 0.8760705721137376,
+      "grad_norm": 0.4556267559528351,
+      "learning_rate": 0.0001,
+      "loss": 0.1104,
+      "step": 10229
+    },
+    {
+      "epoch": 0.8761562178828366,
+      "grad_norm": 0.7655818462371826,
+      "learning_rate": 0.0001,
+      "loss": 1.0396,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8762418636519356,
+      "grad_norm": 0.9498430490493774,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 10231
+    },
+    {
+      "epoch": 0.8763275094210347,
+      "grad_norm": 0.9236242175102234,
+      "learning_rate": 0.0001,
+      "loss": 0.7639,
+      "step": 10232
+    },
+    {
+      "epoch": 0.8764131551901336,
+      "grad_norm": 0.8423366546630859,
+      "learning_rate": 0.0001,
+      "loss": 1.1093,
+      "step": 10233
+    },
+    {
+      "epoch": 0.8764988009592326,
+      "grad_norm": 0.3228733241558075,
+      "learning_rate": 0.0001,
+      "loss": 0.3487,
+      "step": 10234
+    },
+    {
+      "epoch": 0.8765844467283316,
+      "grad_norm": 0.9796724319458008,
+      "learning_rate": 0.0001,
+      "loss": 0.9923,
+      "step": 10235
+    },
+    {
+      "epoch": 0.8766700924974307,
+      "grad_norm": 1.4151864051818848,
+      "learning_rate": 0.0001,
+      "loss": 1.6989,
+      "step": 10236
+    },
+    {
+      "epoch": 0.8767557382665296,
+      "grad_norm": 0.8094688653945923,
+      "learning_rate": 0.0001,
+      "loss": 1.2115,
+      "step": 10237
+    },
+    {
+      "epoch": 0.8768413840356286,
+      "grad_norm": 0.8727339506149292,
+      "learning_rate": 0.0001,
+      "loss": 0.8863,
+      "step": 10238
+    },
+    {
+      "epoch": 0.8769270298047277,
+      "grad_norm": 0.5344792604446411,
+      "learning_rate": 0.0001,
+      "loss": 0.8606,
+      "step": 10239
+    },
+    {
+      "epoch": 0.8770126755738267,
+      "grad_norm": 0.8988091945648193,
+      "learning_rate": 0.0001,
+      "loss": 0.7033,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8770983213429256,
+      "grad_norm": 0.9207578301429749,
+      "learning_rate": 0.0001,
+      "loss": 0.6021,
+      "step": 10241
+    },
+    {
+      "epoch": 0.8771839671120246,
+      "grad_norm": 1.1810215711593628,
+      "learning_rate": 0.0001,
+      "loss": 0.9634,
+      "step": 10242
+    },
+    {
+      "epoch": 0.8772696128811237,
+      "grad_norm": 1.1142306327819824,
+      "learning_rate": 0.0001,
+      "loss": 0.7595,
+      "step": 10243
+    },
+    {
+      "epoch": 0.8773552586502227,
+      "grad_norm": 0.9705197811126709,
+      "learning_rate": 0.0001,
+      "loss": 0.8134,
+      "step": 10244
+    },
+    {
+      "epoch": 0.8774409044193217,
+      "grad_norm": 1.1316860914230347,
+      "learning_rate": 0.0001,
+      "loss": 1.9763,
+      "step": 10245
+    },
+    {
+      "epoch": 0.8775265501884207,
+      "grad_norm": 1.120012640953064,
+      "learning_rate": 0.0001,
+      "loss": 1.2448,
+      "step": 10246
+    },
+    {
+      "epoch": 0.8776121959575197,
+      "grad_norm": 0.6065239310264587,
+      "learning_rate": 0.0001,
+      "loss": 1.1762,
+      "step": 10247
+    },
+    {
+      "epoch": 0.8776978417266187,
+      "grad_norm": 0.549989640712738,
+      "learning_rate": 0.0001,
+      "loss": 0.8042,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8777834874957177,
+      "grad_norm": 0.3523915410041809,
+      "learning_rate": 0.0001,
+      "loss": 0.461,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8778691332648167,
+      "grad_norm": 0.7062180042266846,
+      "learning_rate": 0.0001,
+      "loss": 1.0081,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8779547790339157,
+      "grad_norm": 1.0030386447906494,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 10251
+    },
+    {
+      "epoch": 0.8780404248030147,
+      "grad_norm": 0.8672506213188171,
+      "learning_rate": 0.0001,
+      "loss": 0.9904,
+      "step": 10252
+    },
+    {
+      "epoch": 0.8781260705721138,
+      "grad_norm": 1.1042827367782593,
+      "learning_rate": 0.0001,
+      "loss": 1.077,
+      "step": 10253
+    },
+    {
+      "epoch": 0.8782117163412128,
+      "grad_norm": 0.37751612067222595,
+      "learning_rate": 0.0001,
+      "loss": 0.2818,
+      "step": 10254
+    },
+    {
+      "epoch": 0.8782973621103117,
+      "grad_norm": 1.5883558988571167,
+      "learning_rate": 0.0001,
+      "loss": 0.9406,
+      "step": 10255
+    },
+    {
+      "epoch": 0.8783830078794107,
+      "grad_norm": 1.1055240631103516,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 10256
+    },
+    {
+      "epoch": 0.8784686536485098,
+      "grad_norm": 0.8396492004394531,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 10257
+    },
+    {
+      "epoch": 0.8785542994176088,
+      "grad_norm": 1.128309726715088,
+      "learning_rate": 0.0001,
+      "loss": 0.9713,
+      "step": 10258
+    },
+    {
+      "epoch": 0.8786399451867077,
+      "grad_norm": 1.26515793800354,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 10259
+    },
+    {
+      "epoch": 0.8787255909558068,
+      "grad_norm": 0.9236963987350464,
+      "learning_rate": 0.0001,
+      "loss": 1.0548,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8788112367249058,
+      "grad_norm": 0.7180848717689514,
+      "learning_rate": 0.0001,
+      "loss": 0.8003,
+      "step": 10261
+    },
+    {
+      "epoch": 0.8788968824940048,
+      "grad_norm": 0.6827696561813354,
+      "learning_rate": 0.0001,
+      "loss": 0.9328,
+      "step": 10262
+    },
+    {
+      "epoch": 0.8789825282631039,
+      "grad_norm": 0.7411082983016968,
+      "learning_rate": 0.0001,
+      "loss": 1.1229,
+      "step": 10263
+    },
+    {
+      "epoch": 0.8790681740322028,
+      "grad_norm": 0.7002551555633545,
+      "learning_rate": 0.0001,
+      "loss": 1.0295,
+      "step": 10264
+    },
+    {
+      "epoch": 0.8791538198013018,
+      "grad_norm": 0.523635983467102,
+      "learning_rate": 0.0001,
+      "loss": 1.0171,
+      "step": 10265
+    },
+    {
+      "epoch": 0.8792394655704008,
+      "grad_norm": 0.8091793060302734,
+      "learning_rate": 0.0001,
+      "loss": 0.739,
+      "step": 10266
+    },
+    {
+      "epoch": 0.8793251113394999,
+      "grad_norm": 1.1096343994140625,
+      "learning_rate": 0.0001,
+      "loss": 0.9771,
+      "step": 10267
+    },
+    {
+      "epoch": 0.8794107571085988,
+      "grad_norm": 1.2177820205688477,
+      "learning_rate": 0.0001,
+      "loss": 1.2404,
+      "step": 10268
+    },
+    {
+      "epoch": 0.8794964028776978,
+      "grad_norm": 1.0916812419891357,
+      "learning_rate": 0.0001,
+      "loss": 1.0594,
+      "step": 10269
+    },
+    {
+      "epoch": 0.8795820486467969,
+      "grad_norm": 1.2835105657577515,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8796676944158959,
+      "grad_norm": 0.5645735263824463,
+      "learning_rate": 0.0001,
+      "loss": 0.9586,
+      "step": 10271
+    },
+    {
+      "epoch": 0.8797533401849948,
+      "grad_norm": 0.9726064801216125,
+      "learning_rate": 0.0001,
+      "loss": 0.8294,
+      "step": 10272
+    },
+    {
+      "epoch": 0.8798389859540938,
+      "grad_norm": 1.241855263710022,
+      "learning_rate": 0.0001,
+      "loss": 1.3969,
+      "step": 10273
+    },
+    {
+      "epoch": 0.8799246317231929,
+      "grad_norm": 0.40225204825401306,
+      "learning_rate": 0.0001,
+      "loss": 0.5386,
+      "step": 10274
+    },
+    {
+      "epoch": 0.8800102774922919,
+      "grad_norm": 0.6360182762145996,
+      "learning_rate": 0.0001,
+      "loss": 0.6504,
+      "step": 10275
+    },
+    {
+      "epoch": 0.8800959232613909,
+      "grad_norm": 1.0521876811981201,
+      "learning_rate": 0.0001,
+      "loss": 0.861,
+      "step": 10276
+    },
+    {
+      "epoch": 0.8801815690304899,
+      "grad_norm": 1.0265463590621948,
+      "learning_rate": 0.0001,
+      "loss": 1.1172,
+      "step": 10277
+    },
+    {
+      "epoch": 0.8802672147995889,
+      "grad_norm": 0.5611082315444946,
+      "learning_rate": 0.0001,
+      "loss": 0.8812,
+      "step": 10278
+    },
+    {
+      "epoch": 0.8803528605686879,
+      "grad_norm": 1.1726503372192383,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 10279
+    },
+    {
+      "epoch": 0.8804385063377869,
+      "grad_norm": 0.6184001564979553,
+      "learning_rate": 0.0001,
+      "loss": 0.9798,
+      "step": 10280
+    },
+    {
+      "epoch": 0.880524152106886,
+      "grad_norm": 0.41519424319267273,
+      "learning_rate": 0.0001,
+      "loss": 0.4797,
+      "step": 10281
+    },
+    {
+      "epoch": 0.8806097978759849,
+      "grad_norm": 1.2467455863952637,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 10282
+    },
+    {
+      "epoch": 0.8806954436450839,
+      "grad_norm": 1.338944911956787,
+      "learning_rate": 0.0001,
+      "loss": 0.7356,
+      "step": 10283
+    },
+    {
+      "epoch": 0.880781089414183,
+      "grad_norm": 0.6436741352081299,
+      "learning_rate": 0.0001,
+      "loss": 0.9125,
+      "step": 10284
+    },
+    {
+      "epoch": 0.880866735183282,
+      "grad_norm": 0.9677364230155945,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 10285
+    },
+    {
+      "epoch": 0.8809523809523809,
+      "grad_norm": 0.5877595543861389,
+      "learning_rate": 0.0001,
+      "loss": 1.1738,
+      "step": 10286
+    },
+    {
+      "epoch": 0.8810380267214799,
+      "grad_norm": 0.6236221194267273,
+      "learning_rate": 0.0001,
+      "loss": 0.5749,
+      "step": 10287
+    },
+    {
+      "epoch": 0.881123672490579,
+      "grad_norm": 0.9595468640327454,
+      "learning_rate": 0.0001,
+      "loss": 1.2331,
+      "step": 10288
+    },
+    {
+      "epoch": 0.881209318259678,
+      "grad_norm": 1.0146751403808594,
+      "learning_rate": 0.0001,
+      "loss": 0.933,
+      "step": 10289
+    },
+    {
+      "epoch": 0.8812949640287769,
+      "grad_norm": 0.748242199420929,
+      "learning_rate": 0.0001,
+      "loss": 1.0249,
+      "step": 10290
+    },
+    {
+      "epoch": 0.881380609797876,
+      "grad_norm": 0.53025883436203,
+      "learning_rate": 0.0001,
+      "loss": 0.1575,
+      "step": 10291
+    },
+    {
+      "epoch": 0.881466255566975,
+      "grad_norm": 0.8476935029029846,
+      "learning_rate": 0.0001,
+      "loss": 1.0556,
+      "step": 10292
+    },
+    {
+      "epoch": 0.881551901336074,
+      "grad_norm": 0.799673318862915,
+      "learning_rate": 0.0001,
+      "loss": 1.0768,
+      "step": 10293
+    },
+    {
+      "epoch": 0.8816375471051731,
+      "grad_norm": 1.1318846940994263,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 10294
+    },
+    {
+      "epoch": 0.881723192874272,
+      "grad_norm": 0.9422380924224854,
+      "learning_rate": 0.0001,
+      "loss": 0.8132,
+      "step": 10295
+    },
+    {
+      "epoch": 0.881808838643371,
+      "grad_norm": 1.211185336112976,
+      "learning_rate": 0.0001,
+      "loss": 1.387,
+      "step": 10296
+    },
+    {
+      "epoch": 0.88189448441247,
+      "grad_norm": 0.6065942049026489,
+      "learning_rate": 0.0001,
+      "loss": 0.7767,
+      "step": 10297
+    },
+    {
+      "epoch": 0.8819801301815691,
+      "grad_norm": 0.9667946696281433,
+      "learning_rate": 0.0001,
+      "loss": 1.1472,
+      "step": 10298
+    },
+    {
+      "epoch": 0.882065775950668,
+      "grad_norm": 0.8652786612510681,
+      "learning_rate": 0.0001,
+      "loss": 0.7684,
+      "step": 10299
+    },
+    {
+      "epoch": 0.882151421719767,
+      "grad_norm": 0.7407770752906799,
+      "learning_rate": 0.0001,
+      "loss": 1.0077,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8822370674888661,
+      "grad_norm": 0.21964897215366364,
+      "learning_rate": 0.0001,
+      "loss": 0.1258,
+      "step": 10301
+    },
+    {
+      "epoch": 0.8823227132579651,
+      "grad_norm": 1.301786184310913,
+      "learning_rate": 0.0001,
+      "loss": 1.0526,
+      "step": 10302
+    },
+    {
+      "epoch": 0.882408359027064,
+      "grad_norm": 1.6131818294525146,
+      "learning_rate": 0.0001,
+      "loss": 1.1373,
+      "step": 10303
+    },
+    {
+      "epoch": 0.882494004796163,
+      "grad_norm": 0.3041911721229553,
+      "learning_rate": 0.0001,
+      "loss": 0.2053,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8825796505652621,
+      "grad_norm": 0.5760833024978638,
+      "learning_rate": 0.0001,
+      "loss": 1.0074,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8826652963343611,
+      "grad_norm": 0.9048680663108826,
+      "learning_rate": 0.0001,
+      "loss": 0.8104,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8827509421034601,
+      "grad_norm": 0.7637243270874023,
+      "learning_rate": 0.0001,
+      "loss": 1.1019,
+      "step": 10307
+    },
+    {
+      "epoch": 0.8828365878725591,
+      "grad_norm": 0.9998234510421753,
+      "learning_rate": 0.0001,
+      "loss": 1.1945,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8829222336416581,
+      "grad_norm": 1.0712876319885254,
+      "learning_rate": 0.0001,
+      "loss": 0.79,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8830078794107571,
+      "grad_norm": 1.6003601551055908,
+      "learning_rate": 0.0001,
+      "loss": 0.6719,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8830935251798561,
+      "grad_norm": 1.0150550603866577,
+      "learning_rate": 0.0001,
+      "loss": 0.8038,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8831791709489552,
+      "grad_norm": 1.0880799293518066,
+      "learning_rate": 0.0001,
+      "loss": 1.1459,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8832648167180541,
+      "grad_norm": 0.8268613815307617,
+      "learning_rate": 0.0001,
+      "loss": 1.1373,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8833504624871531,
+      "grad_norm": 0.9225596189498901,
+      "learning_rate": 0.0001,
+      "loss": 1.3126,
+      "step": 10314
+    },
+    {
+      "epoch": 0.8834361082562522,
+      "grad_norm": 1.218306541442871,
+      "learning_rate": 0.0001,
+      "loss": 1.1488,
+      "step": 10315
+    },
+    {
+      "epoch": 0.8835217540253512,
+      "grad_norm": 0.7598164677619934,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8836073997944501,
+      "grad_norm": 0.7544660568237305,
+      "learning_rate": 0.0001,
+      "loss": 0.5926,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8836930455635491,
+      "grad_norm": 0.796830952167511,
+      "learning_rate": 0.0001,
+      "loss": 0.9029,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8837786913326482,
+      "grad_norm": 0.5223682522773743,
+      "learning_rate": 0.0001,
+      "loss": 0.1661,
+      "step": 10319
+    },
+    {
+      "epoch": 0.8838643371017472,
+      "grad_norm": 0.8505058288574219,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8839499828708461,
+      "grad_norm": 0.8271844387054443,
+      "learning_rate": 0.0001,
+      "loss": 1.0284,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8840356286399452,
+      "grad_norm": 0.5966514348983765,
+      "learning_rate": 0.0001,
+      "loss": 0.7552,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8841212744090442,
+      "grad_norm": 0.8607640266418457,
+      "learning_rate": 0.0001,
+      "loss": 0.8271,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8842069201781432,
+      "grad_norm": 0.8011165261268616,
+      "learning_rate": 0.0001,
+      "loss": 0.5227,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8842925659472423,
+      "grad_norm": 1.0269256830215454,
+      "learning_rate": 0.0001,
+      "loss": 0.8678,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8843782117163412,
+      "grad_norm": 0.9062498211860657,
+      "learning_rate": 0.0001,
+      "loss": 0.6684,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8844638574854402,
+      "grad_norm": 0.9211000204086304,
+      "learning_rate": 0.0001,
+      "loss": 0.6294,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8845495032545392,
+      "grad_norm": 0.8849647641181946,
+      "learning_rate": 0.0001,
+      "loss": 1.1168,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8846351490236383,
+      "grad_norm": 1.1658700704574585,
+      "learning_rate": 0.0001,
+      "loss": 1.2649,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8847207947927372,
+      "grad_norm": 0.8017857074737549,
+      "learning_rate": 0.0001,
+      "loss": 0.9649,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8848064405618362,
+      "grad_norm": 0.7239571213722229,
+      "learning_rate": 0.0001,
+      "loss": 0.6513,
+      "step": 10331
+    },
+    {
+      "epoch": 0.8848920863309353,
+      "grad_norm": 0.8927319049835205,
+      "learning_rate": 0.0001,
+      "loss": 1.0771,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8849777321000343,
+      "grad_norm": 0.5378254055976868,
+      "learning_rate": 0.0001,
+      "loss": 0.3666,
+      "step": 10333
+    },
+    {
+      "epoch": 0.8850633778691332,
+      "grad_norm": 1.2971224784851074,
+      "learning_rate": 0.0001,
+      "loss": 1.0155,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8851490236382322,
+      "grad_norm": 1.5203324556350708,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 10335
+    },
+    {
+      "epoch": 0.8852346694073313,
+      "grad_norm": 0.4651581346988678,
+      "learning_rate": 0.0001,
+      "loss": 0.1396,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8853203151764303,
+      "grad_norm": 0.9113432765007019,
+      "learning_rate": 0.0001,
+      "loss": 1.0943,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8854059609455293,
+      "grad_norm": 0.7332322001457214,
+      "learning_rate": 0.0001,
+      "loss": 0.7842,
+      "step": 10338
+    },
+    {
+      "epoch": 0.8854916067146283,
+      "grad_norm": 0.8863824605941772,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 10339
+    },
+    {
+      "epoch": 0.8855772524837273,
+      "grad_norm": 0.9449905753135681,
+      "learning_rate": 0.0001,
+      "loss": 0.9594,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8856628982528263,
+      "grad_norm": 1.1506856679916382,
+      "learning_rate": 0.0001,
+      "loss": 1.1842,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8857485440219253,
+      "grad_norm": 1.2629388570785522,
+      "learning_rate": 0.0001,
+      "loss": 1.3808,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8858341897910244,
+      "grad_norm": 0.7336286306381226,
+      "learning_rate": 0.0001,
+      "loss": 1.2599,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8859198355601233,
+      "grad_norm": 0.6917622685432434,
+      "learning_rate": 0.0001,
+      "loss": 0.7652,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8860054813292223,
+      "grad_norm": 0.8879480957984924,
+      "learning_rate": 0.0001,
+      "loss": 0.6944,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8860911270983214,
+      "grad_norm": 0.3196566104888916,
+      "learning_rate": 0.0001,
+      "loss": 0.4564,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8861767728674204,
+      "grad_norm": 2.3984482288360596,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8862624186365193,
+      "grad_norm": 1.1640434265136719,
+      "learning_rate": 0.0001,
+      "loss": 1.1671,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8863480644056183,
+      "grad_norm": 0.9118731021881104,
+      "learning_rate": 0.0001,
+      "loss": 1.0052,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8864337101747174,
+      "grad_norm": 0.8122690916061401,
+      "learning_rate": 0.0001,
+      "loss": 1.1834,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8865193559438164,
+      "grad_norm": 0.7002624869346619,
+      "learning_rate": 0.0001,
+      "loss": 1.048,
+      "step": 10351
+    },
+    {
+      "epoch": 0.8866050017129153,
+      "grad_norm": 0.6400678157806396,
+      "learning_rate": 0.0001,
+      "loss": 0.5402,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8866906474820144,
+      "grad_norm": 0.6459676623344421,
+      "learning_rate": 0.0001,
+      "loss": 1.0733,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8867762932511134,
+      "grad_norm": 0.6115622520446777,
+      "learning_rate": 0.0001,
+      "loss": 0.8861,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8868619390202124,
+      "grad_norm": 0.736456573009491,
+      "learning_rate": 0.0001,
+      "loss": 1.0311,
+      "step": 10355
+    },
+    {
+      "epoch": 0.8869475847893115,
+      "grad_norm": 0.6744664311408997,
+      "learning_rate": 0.0001,
+      "loss": 1.1904,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8870332305584104,
+      "grad_norm": 1.0871284008026123,
+      "learning_rate": 0.0001,
+      "loss": 0.9149,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8871188763275094,
+      "grad_norm": 1.3156107664108276,
+      "learning_rate": 0.0001,
+      "loss": 1.0019,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8872045220966084,
+      "grad_norm": 1.3825275897979736,
+      "learning_rate": 0.0001,
+      "loss": 1.0303,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8872901678657075,
+      "grad_norm": 0.7104515433311462,
+      "learning_rate": 0.0001,
+      "loss": 0.1932,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8873758136348064,
+      "grad_norm": 0.45579078793525696,
+      "learning_rate": 0.0001,
+      "loss": 0.4976,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8874614594039054,
+      "grad_norm": 1.0056015253067017,
+      "learning_rate": 0.0001,
+      "loss": 0.6704,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8875471051730045,
+      "grad_norm": 0.34291255474090576,
+      "learning_rate": 0.0001,
+      "loss": 0.3616,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8876327509421035,
+      "grad_norm": 1.3777108192443848,
+      "learning_rate": 0.0001,
+      "loss": 0.9443,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8877183967112025,
+      "grad_norm": 1.2132139205932617,
+      "learning_rate": 0.0001,
+      "loss": 1.0601,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8878040424803014,
+      "grad_norm": 1.0611553192138672,
+      "learning_rate": 0.0001,
+      "loss": 0.9586,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8878896882494005,
+      "grad_norm": 0.9053070545196533,
+      "learning_rate": 0.0001,
+      "loss": 1.2243,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8879753340184995,
+      "grad_norm": 1.5575318336486816,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8880609797875985,
+      "grad_norm": 0.6528362035751343,
+      "learning_rate": 0.0001,
+      "loss": 0.4593,
+      "step": 10369
+    },
+    {
+      "epoch": 0.8881466255566975,
+      "grad_norm": 1.6828663349151611,
+      "learning_rate": 0.0001,
+      "loss": 0.8431,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8882322713257965,
+      "grad_norm": 0.4599086046218872,
+      "learning_rate": 0.0001,
+      "loss": 0.4976,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8883179170948955,
+      "grad_norm": 0.6414572596549988,
+      "learning_rate": 0.0001,
+      "loss": 0.6326,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8884035628639945,
+      "grad_norm": 0.7538192272186279,
+      "learning_rate": 0.0001,
+      "loss": 0.667,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8884892086330936,
+      "grad_norm": 0.6098623871803284,
+      "learning_rate": 0.0001,
+      "loss": 0.7639,
+      "step": 10374
+    },
+    {
+      "epoch": 0.8885748544021925,
+      "grad_norm": 0.870356023311615,
+      "learning_rate": 0.0001,
+      "loss": 0.9159,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8886605001712915,
+      "grad_norm": 0.9272902607917786,
+      "learning_rate": 0.0001,
+      "loss": 1.0576,
+      "step": 10376
+    },
+    {
+      "epoch": 0.8887461459403906,
+      "grad_norm": 0.6514511704444885,
+      "learning_rate": 0.0001,
+      "loss": 0.8557,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8888317917094896,
+      "grad_norm": 1.259229063987732,
+      "learning_rate": 0.0001,
+      "loss": 1.3087,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8889174374785885,
+      "grad_norm": 0.6497792601585388,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8890030832476875,
+      "grad_norm": 0.38086166977882385,
+      "learning_rate": 0.0001,
+      "loss": 0.3813,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8890887290167866,
+      "grad_norm": 0.6368332505226135,
+      "learning_rate": 0.0001,
+      "loss": 1.0647,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8891743747858856,
+      "grad_norm": 0.7839457988739014,
+      "learning_rate": 0.0001,
+      "loss": 1.2728,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8892600205549845,
+      "grad_norm": 0.8283966183662415,
+      "learning_rate": 0.0001,
+      "loss": 0.8321,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8893456663240836,
+      "grad_norm": 0.8851462602615356,
+      "learning_rate": 0.0001,
+      "loss": 1.2495,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8894313120931826,
+      "grad_norm": 1.071713924407959,
+      "learning_rate": 0.0001,
+      "loss": 0.9048,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8895169578622816,
+      "grad_norm": 1.4304125308990479,
+      "learning_rate": 0.0001,
+      "loss": 0.8986,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8896026036313807,
+      "grad_norm": 0.6305707693099976,
+      "learning_rate": 0.0001,
+      "loss": 1.2372,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8896882494004796,
+      "grad_norm": 0.7920156121253967,
+      "learning_rate": 0.0001,
+      "loss": 0.919,
+      "step": 10388
+    },
+    {
+      "epoch": 0.8897738951695786,
+      "grad_norm": 0.5746107697486877,
+      "learning_rate": 0.0001,
+      "loss": 0.7225,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8898595409386776,
+      "grad_norm": 0.6949302554130554,
+      "learning_rate": 0.0001,
+      "loss": 0.6525,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8899451867077767,
+      "grad_norm": 0.39997509121894836,
+      "learning_rate": 0.0001,
+      "loss": 0.2611,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8900308324768756,
+      "grad_norm": 0.6208327412605286,
+      "learning_rate": 0.0001,
+      "loss": 0.9778,
+      "step": 10392
+    },
+    {
+      "epoch": 0.8901164782459746,
+      "grad_norm": 0.7141807079315186,
+      "learning_rate": 0.0001,
+      "loss": 1.2465,
+      "step": 10393
+    },
+    {
+      "epoch": 0.8902021240150737,
+      "grad_norm": 1.4690890312194824,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8902877697841727,
+      "grad_norm": 1.3288801908493042,
+      "learning_rate": 0.0001,
+      "loss": 0.8858,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8903734155532717,
+      "grad_norm": 0.67985600233078,
+      "learning_rate": 0.0001,
+      "loss": 0.9441,
+      "step": 10396
+    },
+    {
+      "epoch": 0.8904590613223706,
+      "grad_norm": 0.8696058988571167,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8905447070914697,
+      "grad_norm": 0.7724635601043701,
+      "learning_rate": 0.0001,
+      "loss": 0.8261,
+      "step": 10398
+    },
+    {
+      "epoch": 0.8906303528605687,
+      "grad_norm": 0.9615402221679688,
+      "learning_rate": 0.0001,
+      "loss": 0.9856,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8907159986296677,
+      "grad_norm": 0.7090144157409668,
+      "learning_rate": 0.0001,
+      "loss": 0.4082,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8908016443987667,
+      "grad_norm": 1.0364185571670532,
+      "learning_rate": 0.0001,
+      "loss": 0.5914,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8908872901678657,
+      "grad_norm": 0.9632070064544678,
+      "learning_rate": 0.0001,
+      "loss": 0.7562,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8909729359369647,
+      "grad_norm": 0.928074061870575,
+      "learning_rate": 0.0001,
+      "loss": 0.8036,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8910585817060637,
+      "grad_norm": 0.8258852958679199,
+      "learning_rate": 0.0001,
+      "loss": 0.7394,
+      "step": 10404
+    },
+    {
+      "epoch": 0.8911442274751628,
+      "grad_norm": 0.543941855430603,
+      "learning_rate": 0.0001,
+      "loss": 0.4613,
+      "step": 10405
+    },
+    {
+      "epoch": 0.8912298732442617,
+      "grad_norm": 0.8984708786010742,
+      "learning_rate": 0.0001,
+      "loss": 0.6937,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8913155190133607,
+      "grad_norm": 1.1308454275131226,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8914011647824598,
+      "grad_norm": 0.55525803565979,
+      "learning_rate": 0.0001,
+      "loss": 0.6212,
+      "step": 10408
+    },
+    {
+      "epoch": 0.8914868105515588,
+      "grad_norm": 1.176194190979004,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8915724563206577,
+      "grad_norm": 0.9643118977546692,
+      "learning_rate": 0.0001,
+      "loss": 0.764,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8916581020897568,
+      "grad_norm": 0.6724147200584412,
+      "learning_rate": 0.0001,
+      "loss": 0.8278,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8917437478588558,
+      "grad_norm": 0.7281259894371033,
+      "learning_rate": 0.0001,
+      "loss": 0.4989,
+      "step": 10412
+    },
+    {
+      "epoch": 0.8918293936279548,
+      "grad_norm": 0.6105765104293823,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8919150393970537,
+      "grad_norm": 1.5021957159042358,
+      "learning_rate": 0.0001,
+      "loss": 0.9197,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8920006851661528,
+      "grad_norm": 1.0889214277267456,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8920863309352518,
+      "grad_norm": 0.7698819041252136,
+      "learning_rate": 0.0001,
+      "loss": 0.9308,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8921719767043508,
+      "grad_norm": 1.2121680974960327,
+      "learning_rate": 0.0001,
+      "loss": 1.3401,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8922576224734499,
+      "grad_norm": 1.0687086582183838,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8923432682425488,
+      "grad_norm": 0.47414517402648926,
+      "learning_rate": 0.0001,
+      "loss": 0.4908,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8924289140116478,
+      "grad_norm": 0.5517957806587219,
+      "learning_rate": 0.0001,
+      "loss": 0.7015,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8925145597807468,
+      "grad_norm": 0.6556202173233032,
+      "learning_rate": 0.0001,
+      "loss": 0.7186,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8926002055498459,
+      "grad_norm": 1.335588812828064,
+      "learning_rate": 0.0001,
+      "loss": 1.0188,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8926858513189448,
+      "grad_norm": 0.42811375856399536,
+      "learning_rate": 0.0001,
+      "loss": 0.411,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8927714970880438,
+      "grad_norm": 1.2608295679092407,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 0.634296715259552,
+      "learning_rate": 0.0001,
+      "loss": 0.4969,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8929427886262419,
+      "grad_norm": 1.1034349203109741,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8930284343953409,
+      "grad_norm": 0.80524742603302,
+      "learning_rate": 0.0001,
+      "loss": 1.1159,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8931140801644398,
+      "grad_norm": 0.8258519768714905,
+      "learning_rate": 0.0001,
+      "loss": 1.1799,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8931997259335389,
+      "grad_norm": 0.6439557075500488,
+      "learning_rate": 0.0001,
+      "loss": 1.0016,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8932853717026379,
+      "grad_norm": 0.9534463882446289,
+      "learning_rate": 0.0001,
+      "loss": 0.7617,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8933710174717369,
+      "grad_norm": 0.5559413433074951,
+      "learning_rate": 0.0001,
+      "loss": 1.147,
+      "step": 10431
+    },
+    {
+      "epoch": 0.893456663240836,
+      "grad_norm": 0.6775586605072021,
+      "learning_rate": 0.0001,
+      "loss": 0.9311,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8935423090099349,
+      "grad_norm": 0.8371473550796509,
+      "learning_rate": 0.0001,
+      "loss": 0.8213,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8936279547790339,
+      "grad_norm": 1.1173804998397827,
+      "learning_rate": 0.0001,
+      "loss": 0.6843,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8937136005481329,
+      "grad_norm": 1.2251040935516357,
+      "learning_rate": 0.0001,
+      "loss": 0.8719,
+      "step": 10435
+    },
+    {
+      "epoch": 0.893799246317232,
+      "grad_norm": 0.6638129949569702,
+      "learning_rate": 0.0001,
+      "loss": 1.2348,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8938848920863309,
+      "grad_norm": 0.9095849394798279,
+      "learning_rate": 0.0001,
+      "loss": 0.6206,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8939705378554299,
+      "grad_norm": 1.4049204587936401,
+      "learning_rate": 0.0001,
+      "loss": 1.1277,
+      "step": 10438
+    },
+    {
+      "epoch": 0.894056183624529,
+      "grad_norm": 0.5760965347290039,
+      "learning_rate": 0.0001,
+      "loss": 0.5645,
+      "step": 10439
+    },
+    {
+      "epoch": 0.894141829393628,
+      "grad_norm": 0.7384447455406189,
+      "learning_rate": 0.0001,
+      "loss": 0.8612,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8942274751627269,
+      "grad_norm": 0.8026058077812195,
+      "learning_rate": 0.0001,
+      "loss": 0.8028,
+      "step": 10441
+    },
+    {
+      "epoch": 0.894313120931826,
+      "grad_norm": 0.5817800164222717,
+      "learning_rate": 0.0001,
+      "loss": 0.7341,
+      "step": 10442
+    },
+    {
+      "epoch": 0.894398766700925,
+      "grad_norm": 1.2617266178131104,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 10443
+    },
+    {
+      "epoch": 0.894484412470024,
+      "grad_norm": 1.6842715740203857,
+      "learning_rate": 0.0001,
+      "loss": 2.2507,
+      "step": 10444
+    },
+    {
+      "epoch": 0.894570058239123,
+      "grad_norm": 0.7142163515090942,
+      "learning_rate": 0.0001,
+      "loss": 0.8944,
+      "step": 10445
+    },
+    {
+      "epoch": 0.894655704008222,
+      "grad_norm": 0.4558975398540497,
+      "learning_rate": 0.0001,
+      "loss": 0.3978,
+      "step": 10446
+    },
+    {
+      "epoch": 0.894741349777321,
+      "grad_norm": 0.7449926137924194,
+      "learning_rate": 0.0001,
+      "loss": 0.706,
+      "step": 10447
+    },
+    {
+      "epoch": 0.89482699554642,
+      "grad_norm": 0.7984645366668701,
+      "learning_rate": 0.0001,
+      "loss": 1.2846,
+      "step": 10448
+    },
+    {
+      "epoch": 0.8949126413155191,
+      "grad_norm": 1.0345362424850464,
+      "learning_rate": 0.0001,
+      "loss": 1.275,
+      "step": 10449
+    },
+    {
+      "epoch": 0.894998287084618,
+      "grad_norm": 0.600214421749115,
+      "learning_rate": 0.0001,
+      "loss": 0.4466,
+      "step": 10450
+    },
+    {
+      "epoch": 0.895083932853717,
+      "grad_norm": 0.6411067843437195,
+      "learning_rate": 0.0001,
+      "loss": 0.5246,
+      "step": 10451
+    },
+    {
+      "epoch": 0.895169578622816,
+      "grad_norm": 0.7450178861618042,
+      "learning_rate": 0.0001,
+      "loss": 1.1548,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8952552243919151,
+      "grad_norm": 0.2759278416633606,
+      "learning_rate": 0.0001,
+      "loss": 0.3218,
+      "step": 10453
+    },
+    {
+      "epoch": 0.895340870161014,
+      "grad_norm": 0.38594016432762146,
+      "learning_rate": 0.0001,
+      "loss": 0.1749,
+      "step": 10454
+    },
+    {
+      "epoch": 0.895426515930113,
+      "grad_norm": 0.6417236924171448,
+      "learning_rate": 0.0001,
+      "loss": 1.1461,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8955121616992121,
+      "grad_norm": 1.2450140714645386,
+      "learning_rate": 0.0001,
+      "loss": 1.4136,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8955978074683111,
+      "grad_norm": 0.6934308409690857,
+      "learning_rate": 0.0001,
+      "loss": 1.1941,
+      "step": 10457
+    },
+    {
+      "epoch": 0.89568345323741,
+      "grad_norm": 0.5217545032501221,
+      "learning_rate": 0.0001,
+      "loss": 0.8176,
+      "step": 10458
+    },
+    {
+      "epoch": 0.895769099006509,
+      "grad_norm": 1.2325890064239502,
+      "learning_rate": 0.0001,
+      "loss": 0.578,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8958547447756081,
+      "grad_norm": 1.2245486974716187,
+      "learning_rate": 0.0001,
+      "loss": 1.3263,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8959403905447071,
+      "grad_norm": 0.7032846212387085,
+      "learning_rate": 0.0001,
+      "loss": 0.9533,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8960260363138061,
+      "grad_norm": 1.6481372117996216,
+      "learning_rate": 0.0001,
+      "loss": 0.8478,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8961116820829051,
+      "grad_norm": 1.6918529272079468,
+      "learning_rate": 0.0001,
+      "loss": 2.2161,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8961973278520041,
+      "grad_norm": 1.0662145614624023,
+      "learning_rate": 0.0001,
+      "loss": 0.9517,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8962829736211031,
+      "grad_norm": 1.2472121715545654,
+      "learning_rate": 0.0001,
+      "loss": 1.2931,
+      "step": 10465
+    },
+    {
+      "epoch": 0.8963686193902021,
+      "grad_norm": 1.2171539068222046,
+      "learning_rate": 0.0001,
+      "loss": 1.0978,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8964542651593012,
+      "grad_norm": 0.38867655396461487,
+      "learning_rate": 0.0001,
+      "loss": 0.5282,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8965399109284001,
+      "grad_norm": 0.7529162764549255,
+      "learning_rate": 0.0001,
+      "loss": 0.7208,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8966255566974991,
+      "grad_norm": 0.6115010380744934,
+      "learning_rate": 0.0001,
+      "loss": 1.0583,
+      "step": 10469
+    },
+    {
+      "epoch": 0.8967112024665982,
+      "grad_norm": 0.9437117576599121,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8967968482356972,
+      "grad_norm": 0.960708737373352,
+      "learning_rate": 0.0001,
+      "loss": 1.2833,
+      "step": 10471
+    },
+    {
+      "epoch": 0.8968824940047961,
+      "grad_norm": 0.8292139172554016,
+      "learning_rate": 0.0001,
+      "loss": 0.8474,
+      "step": 10472
+    },
+    {
+      "epoch": 0.8969681397738952,
+      "grad_norm": 1.6240816116333008,
+      "learning_rate": 0.0001,
+      "loss": 1.3027,
+      "step": 10473
+    },
+    {
+      "epoch": 0.8970537855429942,
+      "grad_norm": 0.9549742937088013,
+      "learning_rate": 0.0001,
+      "loss": 1.1839,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8971394313120932,
+      "grad_norm": 0.8562600612640381,
+      "learning_rate": 0.0001,
+      "loss": 0.8951,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8972250770811921,
+      "grad_norm": 0.9141826629638672,
+      "learning_rate": 0.0001,
+      "loss": 1.3378,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8973107228502912,
+      "grad_norm": 0.450592041015625,
+      "learning_rate": 0.0001,
+      "loss": 0.143,
+      "step": 10477
+    },
+    {
+      "epoch": 0.8973963686193902,
+      "grad_norm": 0.8821120858192444,
+      "learning_rate": 0.0001,
+      "loss": 0.7405,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8974820143884892,
+      "grad_norm": 0.7972191572189331,
+      "learning_rate": 0.0001,
+      "loss": 0.9363,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8975676601575883,
+      "grad_norm": 0.7020977139472961,
+      "learning_rate": 0.0001,
+      "loss": 1.0576,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8976533059266872,
+      "grad_norm": 0.523467481136322,
+      "learning_rate": 0.0001,
+      "loss": 0.9379,
+      "step": 10481
+    },
+    {
+      "epoch": 0.8977389516957862,
+      "grad_norm": 1.0174974203109741,
+      "learning_rate": 0.0001,
+      "loss": 1.0828,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8978245974648852,
+      "grad_norm": 0.6864518523216248,
+      "learning_rate": 0.0001,
+      "loss": 0.1612,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8979102432339843,
+      "grad_norm": 0.5815088152885437,
+      "learning_rate": 0.0001,
+      "loss": 0.9972,
+      "step": 10484
+    },
+    {
+      "epoch": 0.8979958890030832,
+      "grad_norm": 0.9714076519012451,
+      "learning_rate": 0.0001,
+      "loss": 1.3762,
+      "step": 10485
+    },
+    {
+      "epoch": 0.8980815347721822,
+      "grad_norm": 1.060370683670044,
+      "learning_rate": 0.0001,
+      "loss": 1.3193,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8981671805412813,
+      "grad_norm": 0.5502649545669556,
+      "learning_rate": 0.0001,
+      "loss": 0.4311,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8982528263103803,
+      "grad_norm": 1.1996469497680664,
+      "learning_rate": 0.0001,
+      "loss": 0.9577,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8983384720794793,
+      "grad_norm": 0.8159409761428833,
+      "learning_rate": 0.0001,
+      "loss": 0.7698,
+      "step": 10489
+    },
+    {
+      "epoch": 0.8984241178485782,
+      "grad_norm": 0.8474125266075134,
+      "learning_rate": 0.0001,
+      "loss": 0.8824,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8985097636176773,
+      "grad_norm": 0.6483515501022339,
+      "learning_rate": 0.0001,
+      "loss": 0.7977,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8985954093867763,
+      "grad_norm": 0.8101222515106201,
+      "learning_rate": 0.0001,
+      "loss": 0.6575,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8986810551558753,
+      "grad_norm": 0.5710697770118713,
+      "learning_rate": 0.0001,
+      "loss": 1.038,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8987667009249743,
+      "grad_norm": 1.1994684934616089,
+      "learning_rate": 0.0001,
+      "loss": 0.8084,
+      "step": 10494
+    },
+    {
+      "epoch": 0.8988523466940733,
+      "grad_norm": 1.1740427017211914,
+      "learning_rate": 0.0001,
+      "loss": 1.385,
+      "step": 10495
+    },
+    {
+      "epoch": 0.8989379924631723,
+      "grad_norm": 1.0721182823181152,
+      "learning_rate": 0.0001,
+      "loss": 1.886,
+      "step": 10496
+    },
+    {
+      "epoch": 0.8990236382322713,
+      "grad_norm": 0.6083436608314514,
+      "learning_rate": 0.0001,
+      "loss": 0.9874,
+      "step": 10497
+    },
+    {
+      "epoch": 0.8991092840013704,
+      "grad_norm": 0.4997051954269409,
+      "learning_rate": 0.0001,
+      "loss": 1.1279,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8991949297704693,
+      "grad_norm": 0.9152095913887024,
+      "learning_rate": 0.0001,
+      "loss": 1.1969,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8992805755395683,
+      "grad_norm": 0.6975170969963074,
+      "learning_rate": 0.0001,
+      "loss": 0.9985,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8993662213086674,
+      "grad_norm": 0.8609218001365662,
+      "learning_rate": 0.0001,
+      "loss": 0.6221,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8994518670777664,
+      "grad_norm": 1.0668667554855347,
+      "learning_rate": 0.0001,
+      "loss": 1.2746,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8995375128468653,
+      "grad_norm": 0.5596345663070679,
+      "learning_rate": 0.0001,
+      "loss": 0.5975,
+      "step": 10503
+    },
+    {
+      "epoch": 0.8996231586159644,
+      "grad_norm": 1.141230583190918,
+      "learning_rate": 0.0001,
+      "loss": 1.0234,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8997088043850634,
+      "grad_norm": 1.032350778579712,
+      "learning_rate": 0.0001,
+      "loss": 1.9743,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8997944501541624,
+      "grad_norm": 1.2195544242858887,
+      "learning_rate": 0.0001,
+      "loss": 1.2236,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8998800959232613,
+      "grad_norm": 0.16074466705322266,
+      "learning_rate": 0.0001,
+      "loss": 0.1107,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8999657416923604,
+      "grad_norm": 0.9030329585075378,
+      "learning_rate": 0.0001,
+      "loss": 0.8198,
+      "step": 10508
+    },
+    {
+      "epoch": 0.9000513874614594,
+      "grad_norm": 0.11858568340539932,
+      "learning_rate": 0.0001,
+      "loss": 0.0723,
+      "step": 10509
+    },
+    {
+      "epoch": 0.9001370332305584,
+      "grad_norm": 0.7616882920265198,
+      "learning_rate": 0.0001,
+      "loss": 0.6393,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9002226789996575,
+      "grad_norm": 0.5981311798095703,
+      "learning_rate": 0.0001,
+      "loss": 0.6851,
+      "step": 10511
+    },
+    {
+      "epoch": 0.9003083247687564,
+      "grad_norm": 0.6255185604095459,
+      "learning_rate": 0.0001,
+      "loss": 1.0888,
+      "step": 10512
+    },
+    {
+      "epoch": 0.9003939705378554,
+      "grad_norm": 2.34773588180542,
+      "learning_rate": 0.0001,
+      "loss": 1.2555,
+      "step": 10513
+    },
+    {
+      "epoch": 0.9004796163069544,
+      "grad_norm": 0.8657112717628479,
+      "learning_rate": 0.0001,
+      "loss": 0.7893,
+      "step": 10514
+    },
+    {
+      "epoch": 0.9005652620760535,
+      "grad_norm": 0.8247033953666687,
+      "learning_rate": 0.0001,
+      "loss": 0.9352,
+      "step": 10515
+    },
+    {
+      "epoch": 0.9006509078451524,
+      "grad_norm": 0.9549154043197632,
+      "learning_rate": 0.0001,
+      "loss": 0.9774,
+      "step": 10516
+    },
+    {
+      "epoch": 0.9007365536142514,
+      "grad_norm": 0.6633344292640686,
+      "learning_rate": 0.0001,
+      "loss": 0.9521,
+      "step": 10517
+    },
+    {
+      "epoch": 0.9008221993833505,
+      "grad_norm": 1.0978094339370728,
+      "learning_rate": 0.0001,
+      "loss": 1.199,
+      "step": 10518
+    },
+    {
+      "epoch": 0.9009078451524495,
+      "grad_norm": 0.9135652184486389,
+      "learning_rate": 0.0001,
+      "loss": 1.0599,
+      "step": 10519
+    },
+    {
+      "epoch": 0.9009934909215485,
+      "grad_norm": 0.837428867816925,
+      "learning_rate": 0.0001,
+      "loss": 0.8653,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9010791366906474,
+      "grad_norm": 0.8755746483802795,
+      "learning_rate": 0.0001,
+      "loss": 0.8688,
+      "step": 10521
+    },
+    {
+      "epoch": 0.9011647824597465,
+      "grad_norm": 1.1719028949737549,
+      "learning_rate": 0.0001,
+      "loss": 0.8615,
+      "step": 10522
+    },
+    {
+      "epoch": 0.9012504282288455,
+      "grad_norm": 0.7140466570854187,
+      "learning_rate": 0.0001,
+      "loss": 1.0331,
+      "step": 10523
+    },
+    {
+      "epoch": 0.9013360739979445,
+      "grad_norm": 0.6602419018745422,
+      "learning_rate": 0.0001,
+      "loss": 1.0299,
+      "step": 10524
+    },
+    {
+      "epoch": 0.9014217197670435,
+      "grad_norm": 0.8909021019935608,
+      "learning_rate": 0.0001,
+      "loss": 1.1551,
+      "step": 10525
+    },
+    {
+      "epoch": 0.9015073655361425,
+      "grad_norm": 0.7689779996871948,
+      "learning_rate": 0.0001,
+      "loss": 1.0143,
+      "step": 10526
+    },
+    {
+      "epoch": 0.9015930113052415,
+      "grad_norm": 1.323865532875061,
+      "learning_rate": 0.0001,
+      "loss": 1.9177,
+      "step": 10527
+    },
+    {
+      "epoch": 0.9016786570743405,
+      "grad_norm": 1.999805212020874,
+      "learning_rate": 0.0001,
+      "loss": 1.8317,
+      "step": 10528
+    },
+    {
+      "epoch": 0.9017643028434396,
+      "grad_norm": 0.8935039639472961,
+      "learning_rate": 0.0001,
+      "loss": 0.7402,
+      "step": 10529
+    },
+    {
+      "epoch": 0.9018499486125385,
+      "grad_norm": 0.6312384009361267,
+      "learning_rate": 0.0001,
+      "loss": 0.8985,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9019355943816375,
+      "grad_norm": 0.7873218059539795,
+      "learning_rate": 0.0001,
+      "loss": 0.8429,
+      "step": 10531
+    },
+    {
+      "epoch": 0.9020212401507366,
+      "grad_norm": 1.1457407474517822,
+      "learning_rate": 0.0001,
+      "loss": 1.277,
+      "step": 10532
+    },
+    {
+      "epoch": 0.9021068859198356,
+      "grad_norm": 0.6179249882698059,
+      "learning_rate": 0.0001,
+      "loss": 0.5064,
+      "step": 10533
+    },
+    {
+      "epoch": 0.9021925316889345,
+      "grad_norm": 0.6321704387664795,
+      "learning_rate": 0.0001,
+      "loss": 1.0424,
+      "step": 10534
+    },
+    {
+      "epoch": 0.9022781774580336,
+      "grad_norm": 1.5785778760910034,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 10535
+    },
+    {
+      "epoch": 0.9023638232271326,
+      "grad_norm": 0.3700374662876129,
+      "learning_rate": 0.0001,
+      "loss": 0.1574,
+      "step": 10536
+    },
+    {
+      "epoch": 0.9024494689962316,
+      "grad_norm": 0.7105832099914551,
+      "learning_rate": 0.0001,
+      "loss": 0.4626,
+      "step": 10537
+    },
+    {
+      "epoch": 0.9025351147653305,
+      "grad_norm": 0.9532680511474609,
+      "learning_rate": 0.0001,
+      "loss": 0.8456,
+      "step": 10538
+    },
+    {
+      "epoch": 0.9026207605344296,
+      "grad_norm": 1.3434724807739258,
+      "learning_rate": 0.0001,
+      "loss": 1.3429,
+      "step": 10539
+    },
+    {
+      "epoch": 0.9027064063035286,
+      "grad_norm": 0.8526580929756165,
+      "learning_rate": 0.0001,
+      "loss": 0.9885,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9027920520726276,
+      "grad_norm": 1.2469850778579712,
+      "learning_rate": 0.0001,
+      "loss": 1.2296,
+      "step": 10541
+    },
+    {
+      "epoch": 0.9028776978417267,
+      "grad_norm": 0.6864014863967896,
+      "learning_rate": 0.0001,
+      "loss": 0.754,
+      "step": 10542
+    },
+    {
+      "epoch": 0.9029633436108256,
+      "grad_norm": 0.5348075032234192,
+      "learning_rate": 0.0001,
+      "loss": 0.7173,
+      "step": 10543
+    },
+    {
+      "epoch": 0.9030489893799246,
+      "grad_norm": 0.6819708347320557,
+      "learning_rate": 0.0001,
+      "loss": 0.1355,
+      "step": 10544
+    },
+    {
+      "epoch": 0.9031346351490236,
+      "grad_norm": 0.8813194036483765,
+      "learning_rate": 0.0001,
+      "loss": 0.7833,
+      "step": 10545
+    },
+    {
+      "epoch": 0.9032202809181227,
+      "grad_norm": 0.41090062260627747,
+      "learning_rate": 0.0001,
+      "loss": 0.3742,
+      "step": 10546
+    },
+    {
+      "epoch": 0.9033059266872216,
+      "grad_norm": 0.9387096166610718,
+      "learning_rate": 0.0001,
+      "loss": 0.8337,
+      "step": 10547
+    },
+    {
+      "epoch": 0.9033915724563206,
+      "grad_norm": 0.839160144329071,
+      "learning_rate": 0.0001,
+      "loss": 0.6425,
+      "step": 10548
+    },
+    {
+      "epoch": 0.9034772182254197,
+      "grad_norm": 1.120407223701477,
+      "learning_rate": 0.0001,
+      "loss": 1.0102,
+      "step": 10549
+    },
+    {
+      "epoch": 0.9035628639945187,
+      "grad_norm": 1.3074640035629272,
+      "learning_rate": 0.0001,
+      "loss": 1.1783,
+      "step": 10550
+    },
+    {
+      "epoch": 0.9036485097636177,
+      "grad_norm": 1.220710039138794,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 10551
+    },
+    {
+      "epoch": 0.9037341555327166,
+      "grad_norm": 1.1899540424346924,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 10552
+    },
+    {
+      "epoch": 0.9038198013018157,
+      "grad_norm": 0.8126803636550903,
+      "learning_rate": 0.0001,
+      "loss": 1.2284,
+      "step": 10553
+    },
+    {
+      "epoch": 0.9039054470709147,
+      "grad_norm": 0.6941216588020325,
+      "learning_rate": 0.0001,
+      "loss": 1.2677,
+      "step": 10554
+    },
+    {
+      "epoch": 0.9039910928400137,
+      "grad_norm": 1.078118085861206,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 10555
+    },
+    {
+      "epoch": 0.9040767386091128,
+      "grad_norm": 1.173125147819519,
+      "learning_rate": 0.0001,
+      "loss": 1.148,
+      "step": 10556
+    },
+    {
+      "epoch": 0.9041623843782117,
+      "grad_norm": 0.7609752416610718,
+      "learning_rate": 0.0001,
+      "loss": 0.6327,
+      "step": 10557
+    },
+    {
+      "epoch": 0.9042480301473107,
+      "grad_norm": 1.407680869102478,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 10558
+    },
+    {
+      "epoch": 0.9043336759164097,
+      "grad_norm": 1.2703148126602173,
+      "learning_rate": 0.0001,
+      "loss": 0.8992,
+      "step": 10559
+    },
+    {
+      "epoch": 0.9044193216855088,
+      "grad_norm": 1.203023910522461,
+      "learning_rate": 0.0001,
+      "loss": 0.8884,
+      "step": 10560
+    },
+    {
+      "epoch": 0.9045049674546077,
+      "grad_norm": 0.8249909281730652,
+      "learning_rate": 0.0001,
+      "loss": 0.6741,
+      "step": 10561
+    },
+    {
+      "epoch": 0.9045906132237067,
+      "grad_norm": 0.7692618370056152,
+      "learning_rate": 0.0001,
+      "loss": 0.6478,
+      "step": 10562
+    },
+    {
+      "epoch": 0.9046762589928058,
+      "grad_norm": 0.7092819809913635,
+      "learning_rate": 0.0001,
+      "loss": 0.8918,
+      "step": 10563
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.6928096413612366,
+      "learning_rate": 0.0001,
+      "loss": 1.0157,
+      "step": 10564
+    },
+    {
+      "epoch": 0.9048475505310037,
+      "grad_norm": 0.5759556889533997,
+      "learning_rate": 0.0001,
+      "loss": 0.7145,
+      "step": 10565
+    },
+    {
+      "epoch": 0.9049331963001028,
+      "grad_norm": 1.5261483192443848,
+      "learning_rate": 0.0001,
+      "loss": 1.1064,
+      "step": 10566
+    },
+    {
+      "epoch": 0.9050188420692018,
+      "grad_norm": 0.9734504222869873,
+      "learning_rate": 0.0001,
+      "loss": 0.9499,
+      "step": 10567
+    },
+    {
+      "epoch": 0.9051044878383008,
+      "grad_norm": 0.7793055176734924,
+      "learning_rate": 0.0001,
+      "loss": 1.0488,
+      "step": 10568
+    },
+    {
+      "epoch": 0.9051901336073997,
+      "grad_norm": 0.5806634426116943,
+      "learning_rate": 0.0001,
+      "loss": 0.8574,
+      "step": 10569
+    },
+    {
+      "epoch": 0.9052757793764988,
+      "grad_norm": 1.1396450996398926,
+      "learning_rate": 0.0001,
+      "loss": 0.8664,
+      "step": 10570
+    },
+    {
+      "epoch": 0.9053614251455978,
+      "grad_norm": 0.683203399181366,
+      "learning_rate": 0.0001,
+      "loss": 0.7041,
+      "step": 10571
+    },
+    {
+      "epoch": 0.9054470709146968,
+      "grad_norm": 0.9220424890518188,
+      "learning_rate": 0.0001,
+      "loss": 0.5829,
+      "step": 10572
+    },
+    {
+      "epoch": 0.9055327166837959,
+      "grad_norm": 1.226509928703308,
+      "learning_rate": 0.0001,
+      "loss": 1.1933,
+      "step": 10573
+    },
+    {
+      "epoch": 0.9056183624528948,
+      "grad_norm": 1.0910919904708862,
+      "learning_rate": 0.0001,
+      "loss": 1.4677,
+      "step": 10574
+    },
+    {
+      "epoch": 0.9057040082219938,
+      "grad_norm": 1.1435027122497559,
+      "learning_rate": 0.0001,
+      "loss": 0.6133,
+      "step": 10575
+    },
+    {
+      "epoch": 0.9057896539910928,
+      "grad_norm": 1.524458885192871,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 10576
+    },
+    {
+      "epoch": 0.9058752997601919,
+      "grad_norm": 0.7383764386177063,
+      "learning_rate": 0.0001,
+      "loss": 0.4368,
+      "step": 10577
+    },
+    {
+      "epoch": 0.9059609455292909,
+      "grad_norm": 0.7228689789772034,
+      "learning_rate": 0.0001,
+      "loss": 1.1837,
+      "step": 10578
+    },
+    {
+      "epoch": 0.9060465912983898,
+      "grad_norm": 1.0512810945510864,
+      "learning_rate": 0.0001,
+      "loss": 1.2301,
+      "step": 10579
+    },
+    {
+      "epoch": 0.9061322370674889,
+      "grad_norm": 0.6696946620941162,
+      "learning_rate": 0.0001,
+      "loss": 1.0479,
+      "step": 10580
+    },
+    {
+      "epoch": 0.9062178828365879,
+      "grad_norm": 1.1908296346664429,
+      "learning_rate": 0.0001,
+      "loss": 1.0201,
+      "step": 10581
+    },
+    {
+      "epoch": 0.9063035286056869,
+      "grad_norm": 0.7096329927444458,
+      "learning_rate": 0.0001,
+      "loss": 1.2977,
+      "step": 10582
+    },
+    {
+      "epoch": 0.9063891743747858,
+      "grad_norm": 0.9643850922584534,
+      "learning_rate": 0.0001,
+      "loss": 0.9515,
+      "step": 10583
+    },
+    {
+      "epoch": 0.9064748201438849,
+      "grad_norm": 1.2512801885604858,
+      "learning_rate": 0.0001,
+      "loss": 0.6156,
+      "step": 10584
+    },
+    {
+      "epoch": 0.9065604659129839,
+      "grad_norm": 0.7609808444976807,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 10585
+    },
+    {
+      "epoch": 0.9066461116820829,
+      "grad_norm": 1.2630038261413574,
+      "learning_rate": 0.0001,
+      "loss": 1.3623,
+      "step": 10586
+    },
+    {
+      "epoch": 0.906731757451182,
+      "grad_norm": 1.2147728204727173,
+      "learning_rate": 0.0001,
+      "loss": 0.8638,
+      "step": 10587
+    },
+    {
+      "epoch": 0.9068174032202809,
+      "grad_norm": 0.8620423078536987,
+      "learning_rate": 0.0001,
+      "loss": 0.7487,
+      "step": 10588
+    },
+    {
+      "epoch": 0.9069030489893799,
+      "grad_norm": 1.2392226457595825,
+      "learning_rate": 0.0001,
+      "loss": 0.9543,
+      "step": 10589
+    },
+    {
+      "epoch": 0.906988694758479,
+      "grad_norm": 1.2420227527618408,
+      "learning_rate": 0.0001,
+      "loss": 1.152,
+      "step": 10590
+    },
+    {
+      "epoch": 0.907074340527578,
+      "grad_norm": 0.9536930918693542,
+      "learning_rate": 0.0001,
+      "loss": 0.974,
+      "step": 10591
+    },
+    {
+      "epoch": 0.9071599862966769,
+      "grad_norm": 0.6545112133026123,
+      "learning_rate": 0.0001,
+      "loss": 0.4848,
+      "step": 10592
+    },
+    {
+      "epoch": 0.9072456320657759,
+      "grad_norm": 0.4567387104034424,
+      "learning_rate": 0.0001,
+      "loss": 0.3035,
+      "step": 10593
+    },
+    {
+      "epoch": 0.907331277834875,
+      "grad_norm": 0.8897137641906738,
+      "learning_rate": 0.0001,
+      "loss": 0.8554,
+      "step": 10594
+    },
+    {
+      "epoch": 0.907416923603974,
+      "grad_norm": 0.6513718366622925,
+      "learning_rate": 0.0001,
+      "loss": 0.6389,
+      "step": 10595
+    },
+    {
+      "epoch": 0.9075025693730729,
+      "grad_norm": 0.7160093188285828,
+      "learning_rate": 0.0001,
+      "loss": 1.1007,
+      "step": 10596
+    },
+    {
+      "epoch": 0.907588215142172,
+      "grad_norm": 0.6682780981063843,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 10597
+    },
+    {
+      "epoch": 0.907673860911271,
+      "grad_norm": 1.9008746147155762,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 10598
+    },
+    {
+      "epoch": 0.90775950668037,
+      "grad_norm": 0.8680933713912964,
+      "learning_rate": 0.0001,
+      "loss": 0.837,
+      "step": 10599
+    },
+    {
+      "epoch": 0.907845152449469,
+      "grad_norm": 0.7484337091445923,
+      "learning_rate": 0.0001,
+      "loss": 1.195,
+      "step": 10600
+    },
+    {
+      "epoch": 0.907930798218568,
+      "grad_norm": 0.8322603106498718,
+      "learning_rate": 0.0001,
+      "loss": 0.6243,
+      "step": 10601
+    },
+    {
+      "epoch": 0.908016443987667,
+      "grad_norm": 0.801035463809967,
+      "learning_rate": 0.0001,
+      "loss": 0.9021,
+      "step": 10602
+    },
+    {
+      "epoch": 0.908102089756766,
+      "grad_norm": 1.0191398859024048,
+      "learning_rate": 0.0001,
+      "loss": 1.4356,
+      "step": 10603
+    },
+    {
+      "epoch": 0.9081877355258651,
+      "grad_norm": 0.681731104850769,
+      "learning_rate": 0.0001,
+      "loss": 0.7737,
+      "step": 10604
+    },
+    {
+      "epoch": 0.908273381294964,
+      "grad_norm": 1.2972103357315063,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 10605
+    },
+    {
+      "epoch": 0.908359027064063,
+      "grad_norm": 1.2868458032608032,
+      "learning_rate": 0.0001,
+      "loss": 1.2005,
+      "step": 10606
+    },
+    {
+      "epoch": 0.908444672833162,
+      "grad_norm": 0.5099084973335266,
+      "learning_rate": 0.0001,
+      "loss": 0.6791,
+      "step": 10607
+    },
+    {
+      "epoch": 0.9085303186022611,
+      "grad_norm": 0.7074721455574036,
+      "learning_rate": 0.0001,
+      "loss": 0.6246,
+      "step": 10608
+    },
+    {
+      "epoch": 0.90861596437136,
+      "grad_norm": 1.0005098581314087,
+      "learning_rate": 0.0001,
+      "loss": 0.6729,
+      "step": 10609
+    },
+    {
+      "epoch": 0.908701610140459,
+      "grad_norm": 0.7443298101425171,
+      "learning_rate": 0.0001,
+      "loss": 0.7359,
+      "step": 10610
+    },
+    {
+      "epoch": 0.9087872559095581,
+      "grad_norm": 1.2262204885482788,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 10611
+    },
+    {
+      "epoch": 0.9088729016786571,
+      "grad_norm": 0.618663489818573,
+      "learning_rate": 0.0001,
+      "loss": 0.7115,
+      "step": 10612
+    },
+    {
+      "epoch": 0.9089585474477561,
+      "grad_norm": 1.0589526891708374,
+      "learning_rate": 0.0001,
+      "loss": 0.8651,
+      "step": 10613
+    },
+    {
+      "epoch": 0.909044193216855,
+      "grad_norm": 0.6179689764976501,
+      "learning_rate": 0.0001,
+      "loss": 0.7987,
+      "step": 10614
+    },
+    {
+      "epoch": 0.9091298389859541,
+      "grad_norm": 0.694033145904541,
+      "learning_rate": 0.0001,
+      "loss": 1.1705,
+      "step": 10615
+    },
+    {
+      "epoch": 0.9092154847550531,
+      "grad_norm": 0.8992536664009094,
+      "learning_rate": 0.0001,
+      "loss": 0.8066,
+      "step": 10616
+    },
+    {
+      "epoch": 0.9093011305241521,
+      "grad_norm": 0.7382333874702454,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 10617
+    },
+    {
+      "epoch": 0.9093867762932512,
+      "grad_norm": 0.6674996614456177,
+      "learning_rate": 0.0001,
+      "loss": 0.3693,
+      "step": 10618
+    },
+    {
+      "epoch": 0.9094724220623501,
+      "grad_norm": 1.263257384300232,
+      "learning_rate": 0.0001,
+      "loss": 1.1589,
+      "step": 10619
+    },
+    {
+      "epoch": 0.9095580678314491,
+      "grad_norm": 1.2010304927825928,
+      "learning_rate": 0.0001,
+      "loss": 0.8961,
+      "step": 10620
+    },
+    {
+      "epoch": 0.9096437136005482,
+      "grad_norm": 1.0204113721847534,
+      "learning_rate": 0.0001,
+      "loss": 0.8676,
+      "step": 10621
+    },
+    {
+      "epoch": 0.9097293593696472,
+      "grad_norm": 0.7152019739151001,
+      "learning_rate": 0.0001,
+      "loss": 1.0545,
+      "step": 10622
+    },
+    {
+      "epoch": 0.9098150051387461,
+      "grad_norm": 1.1549179553985596,
+      "learning_rate": 0.0001,
+      "loss": 2.2392,
+      "step": 10623
+    },
+    {
+      "epoch": 0.9099006509078451,
+      "grad_norm": 0.6002699136734009,
+      "learning_rate": 0.0001,
+      "loss": 1.2206,
+      "step": 10624
+    },
+    {
+      "epoch": 0.9099862966769442,
+      "grad_norm": 0.9385048747062683,
+      "learning_rate": 0.0001,
+      "loss": 1.003,
+      "step": 10625
+    },
+    {
+      "epoch": 0.9100719424460432,
+      "grad_norm": 0.47440847754478455,
+      "learning_rate": 0.0001,
+      "loss": 0.4354,
+      "step": 10626
+    },
+    {
+      "epoch": 0.9101575882151421,
+      "grad_norm": 1.0206423997879028,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 10627
+    },
+    {
+      "epoch": 0.9102432339842412,
+      "grad_norm": 0.5826166272163391,
+      "learning_rate": 0.0001,
+      "loss": 0.4102,
+      "step": 10628
+    },
+    {
+      "epoch": 0.9103288797533402,
+      "grad_norm": 1.3508471250534058,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 10629
+    },
+    {
+      "epoch": 0.9104145255224392,
+      "grad_norm": 1.1910245418548584,
+      "learning_rate": 0.0001,
+      "loss": 1.9836,
+      "step": 10630
+    },
+    {
+      "epoch": 0.9105001712915382,
+      "grad_norm": 0.8744184970855713,
+      "learning_rate": 0.0001,
+      "loss": 0.9548,
+      "step": 10631
+    },
+    {
+      "epoch": 0.9105858170606372,
+      "grad_norm": 0.7851701378822327,
+      "learning_rate": 0.0001,
+      "loss": 0.6095,
+      "step": 10632
+    },
+    {
+      "epoch": 0.9106714628297362,
+      "grad_norm": 0.408584862947464,
+      "learning_rate": 0.0001,
+      "loss": 0.591,
+      "step": 10633
+    },
+    {
+      "epoch": 0.9107571085988352,
+      "grad_norm": 0.7725461721420288,
+      "learning_rate": 0.0001,
+      "loss": 0.9245,
+      "step": 10634
+    },
+    {
+      "epoch": 0.9108427543679343,
+      "grad_norm": 1.2961474657058716,
+      "learning_rate": 0.0001,
+      "loss": 1.03,
+      "step": 10635
+    },
+    {
+      "epoch": 0.9109284001370332,
+      "grad_norm": 0.1773337870836258,
+      "learning_rate": 0.0001,
+      "loss": 0.0535,
+      "step": 10636
+    },
+    {
+      "epoch": 0.9110140459061322,
+      "grad_norm": 0.6340424418449402,
+      "learning_rate": 0.0001,
+      "loss": 0.8449,
+      "step": 10637
+    },
+    {
+      "epoch": 0.9110996916752312,
+      "grad_norm": 0.7060760259628296,
+      "learning_rate": 0.0001,
+      "loss": 0.4533,
+      "step": 10638
+    },
+    {
+      "epoch": 0.9111853374443303,
+      "grad_norm": 1.1194998025894165,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 10639
+    },
+    {
+      "epoch": 0.9112709832134293,
+      "grad_norm": 0.8251051306724548,
+      "learning_rate": 0.0001,
+      "loss": 1.1752,
+      "step": 10640
+    },
+    {
+      "epoch": 0.9113566289825282,
+      "grad_norm": 1.341165542602539,
+      "learning_rate": 0.0001,
+      "loss": 1.283,
+      "step": 10641
+    },
+    {
+      "epoch": 0.9114422747516273,
+      "grad_norm": 0.8172129392623901,
+      "learning_rate": 0.0001,
+      "loss": 1.0439,
+      "step": 10642
+    },
+    {
+      "epoch": 0.9115279205207263,
+      "grad_norm": 0.6678515672683716,
+      "learning_rate": 0.0001,
+      "loss": 0.7976,
+      "step": 10643
+    },
+    {
+      "epoch": 0.9116135662898253,
+      "grad_norm": 0.8094120025634766,
+      "learning_rate": 0.0001,
+      "loss": 0.7024,
+      "step": 10644
+    },
+    {
+      "epoch": 0.9116992120589242,
+      "grad_norm": 0.6000279188156128,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 10645
+    },
+    {
+      "epoch": 0.9117848578280233,
+      "grad_norm": 0.7425779700279236,
+      "learning_rate": 0.0001,
+      "loss": 0.6268,
+      "step": 10646
+    },
+    {
+      "epoch": 0.9118705035971223,
+      "grad_norm": 0.9945883750915527,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 10647
+    },
+    {
+      "epoch": 0.9119561493662213,
+      "grad_norm": 0.3201678693294525,
+      "learning_rate": 0.0001,
+      "loss": 0.1975,
+      "step": 10648
+    },
+    {
+      "epoch": 0.9120417951353204,
+      "grad_norm": 0.8555192351341248,
+      "learning_rate": 0.0001,
+      "loss": 0.8743,
+      "step": 10649
+    },
+    {
+      "epoch": 0.9121274409044193,
+      "grad_norm": 0.6752687096595764,
+      "learning_rate": 0.0001,
+      "loss": 1.0732,
+      "step": 10650
+    },
+    {
+      "epoch": 0.9122130866735183,
+      "grad_norm": 1.201664686203003,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 10651
+    },
+    {
+      "epoch": 0.9122987324426174,
+      "grad_norm": 1.2445100545883179,
+      "learning_rate": 0.0001,
+      "loss": 0.9935,
+      "step": 10652
+    },
+    {
+      "epoch": 0.9123843782117164,
+      "grad_norm": 0.8229407668113708,
+      "learning_rate": 0.0001,
+      "loss": 0.9,
+      "step": 10653
+    },
+    {
+      "epoch": 0.9124700239808153,
+      "grad_norm": 0.9027330875396729,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 10654
+    },
+    {
+      "epoch": 0.9125556697499143,
+      "grad_norm": 0.7054086327552795,
+      "learning_rate": 0.0001,
+      "loss": 1.0032,
+      "step": 10655
+    },
+    {
+      "epoch": 0.9126413155190134,
+      "grad_norm": 0.8111066818237305,
+      "learning_rate": 0.0001,
+      "loss": 0.6953,
+      "step": 10656
+    },
+    {
+      "epoch": 0.9127269612881124,
+      "grad_norm": 0.5847154259681702,
+      "learning_rate": 0.0001,
+      "loss": 1.0247,
+      "step": 10657
+    },
+    {
+      "epoch": 0.9128126070572113,
+      "grad_norm": 0.8831062912940979,
+      "learning_rate": 0.0001,
+      "loss": 1.0318,
+      "step": 10658
+    },
+    {
+      "epoch": 0.9128982528263104,
+      "grad_norm": 0.5965173244476318,
+      "learning_rate": 0.0001,
+      "loss": 0.4698,
+      "step": 10659
+    },
+    {
+      "epoch": 0.9129838985954094,
+      "grad_norm": 0.7901139259338379,
+      "learning_rate": 0.0001,
+      "loss": 0.704,
+      "step": 10660
+    },
+    {
+      "epoch": 0.9130695443645084,
+      "grad_norm": 0.865538477897644,
+      "learning_rate": 0.0001,
+      "loss": 1.0899,
+      "step": 10661
+    },
+    {
+      "epoch": 0.9131551901336074,
+      "grad_norm": 0.7232416272163391,
+      "learning_rate": 0.0001,
+      "loss": 0.5654,
+      "step": 10662
+    },
+    {
+      "epoch": 0.9132408359027064,
+      "grad_norm": 1.4115650653839111,
+      "learning_rate": 0.0001,
+      "loss": 1.8048,
+      "step": 10663
+    },
+    {
+      "epoch": 0.9133264816718054,
+      "grad_norm": 0.517103910446167,
+      "learning_rate": 0.0001,
+      "loss": 0.1679,
+      "step": 10664
+    },
+    {
+      "epoch": 0.9134121274409044,
+      "grad_norm": 1.1115137338638306,
+      "learning_rate": 0.0001,
+      "loss": 1.0072,
+      "step": 10665
+    },
+    {
+      "epoch": 0.9134977732100035,
+      "grad_norm": 0.946090817451477,
+      "learning_rate": 0.0001,
+      "loss": 0.8353,
+      "step": 10666
+    },
+    {
+      "epoch": 0.9135834189791024,
+      "grad_norm": 0.9037908315658569,
+      "learning_rate": 0.0001,
+      "loss": 1.0673,
+      "step": 10667
+    },
+    {
+      "epoch": 0.9136690647482014,
+      "grad_norm": 1.2689919471740723,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 10668
+    },
+    {
+      "epoch": 0.9137547105173004,
+      "grad_norm": 0.533300518989563,
+      "learning_rate": 0.0001,
+      "loss": 0.3815,
+      "step": 10669
+    },
+    {
+      "epoch": 0.9138403562863995,
+      "grad_norm": 0.6576631665229797,
+      "learning_rate": 0.0001,
+      "loss": 0.8654,
+      "step": 10670
+    },
+    {
+      "epoch": 0.9139260020554985,
+      "grad_norm": 0.22882860898971558,
+      "learning_rate": 0.0001,
+      "loss": 0.3714,
+      "step": 10671
+    },
+    {
+      "epoch": 0.9140116478245974,
+      "grad_norm": 0.7913926839828491,
+      "learning_rate": 0.0001,
+      "loss": 1.2585,
+      "step": 10672
+    },
+    {
+      "epoch": 0.9140972935936965,
+      "grad_norm": 1.8209880590438843,
+      "learning_rate": 0.0001,
+      "loss": 1.2297,
+      "step": 10673
+    },
+    {
+      "epoch": 0.9141829393627955,
+      "grad_norm": 0.8750295042991638,
+      "learning_rate": 0.0001,
+      "loss": 0.9467,
+      "step": 10674
+    },
+    {
+      "epoch": 0.9142685851318945,
+      "grad_norm": 1.3790698051452637,
+      "learning_rate": 0.0001,
+      "loss": 1.3593,
+      "step": 10675
+    },
+    {
+      "epoch": 0.9143542309009934,
+      "grad_norm": 0.8789300322532654,
+      "learning_rate": 0.0001,
+      "loss": 1.0444,
+      "step": 10676
+    },
+    {
+      "epoch": 0.9144398766700925,
+      "grad_norm": 0.31727561354637146,
+      "learning_rate": 0.0001,
+      "loss": 0.1129,
+      "step": 10677
+    },
+    {
+      "epoch": 0.9145255224391915,
+      "grad_norm": 0.8607107400894165,
+      "learning_rate": 0.0001,
+      "loss": 1.0266,
+      "step": 10678
+    },
+    {
+      "epoch": 0.9146111682082905,
+      "grad_norm": 0.7553046345710754,
+      "learning_rate": 0.0001,
+      "loss": 0.8511,
+      "step": 10679
+    },
+    {
+      "epoch": 0.9146968139773896,
+      "grad_norm": 0.7708513736724854,
+      "learning_rate": 0.0001,
+      "loss": 0.5426,
+      "step": 10680
+    },
+    {
+      "epoch": 0.9147824597464885,
+      "grad_norm": 1.3504252433776855,
+      "learning_rate": 0.0001,
+      "loss": 1.2333,
+      "step": 10681
+    },
+    {
+      "epoch": 0.9148681055155875,
+      "grad_norm": 0.8180840611457825,
+      "learning_rate": 0.0001,
+      "loss": 1.2532,
+      "step": 10682
+    },
+    {
+      "epoch": 0.9149537512846866,
+      "grad_norm": 1.1773725748062134,
+      "learning_rate": 0.0001,
+      "loss": 1.2307,
+      "step": 10683
+    },
+    {
+      "epoch": 0.9150393970537856,
+      "grad_norm": 0.6654910445213318,
+      "learning_rate": 0.0001,
+      "loss": 0.6485,
+      "step": 10684
+    },
+    {
+      "epoch": 0.9151250428228845,
+      "grad_norm": 0.9227784276008606,
+      "learning_rate": 0.0001,
+      "loss": 1.1162,
+      "step": 10685
+    },
+    {
+      "epoch": 0.9152106885919835,
+      "grad_norm": 1.1509249210357666,
+      "learning_rate": 0.0001,
+      "loss": 0.869,
+      "step": 10686
+    },
+    {
+      "epoch": 0.9152963343610826,
+      "grad_norm": 1.0432201623916626,
+      "learning_rate": 0.0001,
+      "loss": 1.1444,
+      "step": 10687
+    },
+    {
+      "epoch": 0.9153819801301816,
+      "grad_norm": 0.8058709502220154,
+      "learning_rate": 0.0001,
+      "loss": 1.2107,
+      "step": 10688
+    },
+    {
+      "epoch": 0.9154676258992805,
+      "grad_norm": 1.214064359664917,
+      "learning_rate": 0.0001,
+      "loss": 1.0319,
+      "step": 10689
+    },
+    {
+      "epoch": 0.9155532716683796,
+      "grad_norm": 1.266386866569519,
+      "learning_rate": 0.0001,
+      "loss": 0.8029,
+      "step": 10690
+    },
+    {
+      "epoch": 0.9156389174374786,
+      "grad_norm": 0.8835861086845398,
+      "learning_rate": 0.0001,
+      "loss": 0.9285,
+      "step": 10691
+    },
+    {
+      "epoch": 0.9157245632065776,
+      "grad_norm": 0.7852265238761902,
+      "learning_rate": 0.0001,
+      "loss": 1.0411,
+      "step": 10692
+    },
+    {
+      "epoch": 0.9158102089756766,
+      "grad_norm": 1.0866369009017944,
+      "learning_rate": 0.0001,
+      "loss": 1.0111,
+      "step": 10693
+    },
+    {
+      "epoch": 0.9158958547447756,
+      "grad_norm": 0.534328281879425,
+      "learning_rate": 0.0001,
+      "loss": 0.4013,
+      "step": 10694
+    },
+    {
+      "epoch": 0.9159815005138746,
+      "grad_norm": 0.7097445726394653,
+      "learning_rate": 0.0001,
+      "loss": 1.1717,
+      "step": 10695
+    },
+    {
+      "epoch": 0.9160671462829736,
+      "grad_norm": 0.5772000551223755,
+      "learning_rate": 0.0001,
+      "loss": 0.8958,
+      "step": 10696
+    },
+    {
+      "epoch": 0.9161527920520727,
+      "grad_norm": 0.8526248931884766,
+      "learning_rate": 0.0001,
+      "loss": 0.5235,
+      "step": 10697
+    },
+    {
+      "epoch": 0.9162384378211716,
+      "grad_norm": 0.8604404330253601,
+      "learning_rate": 0.0001,
+      "loss": 1.3027,
+      "step": 10698
+    },
+    {
+      "epoch": 0.9163240835902706,
+      "grad_norm": 0.8388153314590454,
+      "learning_rate": 0.0001,
+      "loss": 0.9579,
+      "step": 10699
+    },
+    {
+      "epoch": 0.9164097293593696,
+      "grad_norm": 1.0173871517181396,
+      "learning_rate": 0.0001,
+      "loss": 0.8716,
+      "step": 10700
+    },
+    {
+      "epoch": 0.9164953751284687,
+      "grad_norm": 0.5515727996826172,
+      "learning_rate": 0.0001,
+      "loss": 0.6509,
+      "step": 10701
+    },
+    {
+      "epoch": 0.9165810208975677,
+      "grad_norm": 0.8224193453788757,
+      "learning_rate": 0.0001,
+      "loss": 1.0902,
+      "step": 10702
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 0.673329770565033,
+      "learning_rate": 0.0001,
+      "loss": 0.9509,
+      "step": 10703
+    },
+    {
+      "epoch": 0.9167523124357657,
+      "grad_norm": 0.8797872066497803,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 10704
+    },
+    {
+      "epoch": 0.9168379582048647,
+      "grad_norm": 0.902789294719696,
+      "learning_rate": 0.0001,
+      "loss": 0.8731,
+      "step": 10705
+    },
+    {
+      "epoch": 0.9169236039739637,
+      "grad_norm": 0.8766804337501526,
+      "learning_rate": 0.0001,
+      "loss": 1.2322,
+      "step": 10706
+    },
+    {
+      "epoch": 0.9170092497430626,
+      "grad_norm": 0.7969714999198914,
+      "learning_rate": 0.0001,
+      "loss": 1.1685,
+      "step": 10707
+    },
+    {
+      "epoch": 0.9170948955121617,
+      "grad_norm": 1.130534291267395,
+      "learning_rate": 0.0001,
+      "loss": 0.735,
+      "step": 10708
+    },
+    {
+      "epoch": 0.9171805412812607,
+      "grad_norm": 1.315017819404602,
+      "learning_rate": 0.0001,
+      "loss": 1.388,
+      "step": 10709
+    },
+    {
+      "epoch": 0.9172661870503597,
+      "grad_norm": 0.9480187892913818,
+      "learning_rate": 0.0001,
+      "loss": 1.0175,
+      "step": 10710
+    },
+    {
+      "epoch": 0.9173518328194588,
+      "grad_norm": 1.6589654684066772,
+      "learning_rate": 0.0001,
+      "loss": 1.2567,
+      "step": 10711
+    },
+    {
+      "epoch": 0.9174374785885577,
+      "grad_norm": 1.5067408084869385,
+      "learning_rate": 0.0001,
+      "loss": 0.6251,
+      "step": 10712
+    },
+    {
+      "epoch": 0.9175231243576567,
+      "grad_norm": 0.7670378088951111,
+      "learning_rate": 0.0001,
+      "loss": 1.0226,
+      "step": 10713
+    },
+    {
+      "epoch": 0.9176087701267558,
+      "grad_norm": 1.3213778734207153,
+      "learning_rate": 0.0001,
+      "loss": 0.8397,
+      "step": 10714
+    },
+    {
+      "epoch": 0.9176944158958548,
+      "grad_norm": 1.1507353782653809,
+      "learning_rate": 0.0001,
+      "loss": 0.7695,
+      "step": 10715
+    },
+    {
+      "epoch": 0.9177800616649537,
+      "grad_norm": 0.848016083240509,
+      "learning_rate": 0.0001,
+      "loss": 0.6302,
+      "step": 10716
+    },
+    {
+      "epoch": 0.9178657074340527,
+      "grad_norm": 0.8595674633979797,
+      "learning_rate": 0.0001,
+      "loss": 0.7031,
+      "step": 10717
+    },
+    {
+      "epoch": 0.9179513532031518,
+      "grad_norm": 0.8679869174957275,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 10718
+    },
+    {
+      "epoch": 0.9180369989722508,
+      "grad_norm": 0.6978803873062134,
+      "learning_rate": 0.0001,
+      "loss": 0.7038,
+      "step": 10719
+    },
+    {
+      "epoch": 0.9181226447413497,
+      "grad_norm": 0.6778854727745056,
+      "learning_rate": 0.0001,
+      "loss": 1.1926,
+      "step": 10720
+    },
+    {
+      "epoch": 0.9182082905104488,
+      "grad_norm": 1.0897223949432373,
+      "learning_rate": 0.0001,
+      "loss": 0.7322,
+      "step": 10721
+    },
+    {
+      "epoch": 0.9182939362795478,
+      "grad_norm": 0.9670572876930237,
+      "learning_rate": 0.0001,
+      "loss": 0.8901,
+      "step": 10722
+    },
+    {
+      "epoch": 0.9183795820486468,
+      "grad_norm": 0.8171031475067139,
+      "learning_rate": 0.0001,
+      "loss": 1.2112,
+      "step": 10723
+    },
+    {
+      "epoch": 0.9184652278177458,
+      "grad_norm": 0.5581685900688171,
+      "learning_rate": 0.0001,
+      "loss": 0.2144,
+      "step": 10724
+    },
+    {
+      "epoch": 0.9185508735868448,
+      "grad_norm": 0.7003543376922607,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 10725
+    },
+    {
+      "epoch": 0.9186365193559438,
+      "grad_norm": 0.8105494379997253,
+      "learning_rate": 0.0001,
+      "loss": 1.0149,
+      "step": 10726
+    },
+    {
+      "epoch": 0.9187221651250428,
+      "grad_norm": 0.9787182211875916,
+      "learning_rate": 0.0001,
+      "loss": 1.2546,
+      "step": 10727
+    },
+    {
+      "epoch": 0.9188078108941419,
+      "grad_norm": 0.775027871131897,
+      "learning_rate": 0.0001,
+      "loss": 0.6594,
+      "step": 10728
+    },
+    {
+      "epoch": 0.9188934566632408,
+      "grad_norm": 0.7788079977035522,
+      "learning_rate": 0.0001,
+      "loss": 0.9186,
+      "step": 10729
+    },
+    {
+      "epoch": 0.9189791024323398,
+      "grad_norm": 0.9612367153167725,
+      "learning_rate": 0.0001,
+      "loss": 1.3377,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9190647482014388,
+      "grad_norm": 0.8326892852783203,
+      "learning_rate": 0.0001,
+      "loss": 0.6488,
+      "step": 10731
+    },
+    {
+      "epoch": 0.9191503939705379,
+      "grad_norm": 1.1393117904663086,
+      "learning_rate": 0.0001,
+      "loss": 1.8632,
+      "step": 10732
+    },
+    {
+      "epoch": 0.9192360397396369,
+      "grad_norm": 0.9103503823280334,
+      "learning_rate": 0.0001,
+      "loss": 1.0363,
+      "step": 10733
+    },
+    {
+      "epoch": 0.9193216855087358,
+      "grad_norm": 0.6293918490409851,
+      "learning_rate": 0.0001,
+      "loss": 0.8806,
+      "step": 10734
+    },
+    {
+      "epoch": 0.9194073312778349,
+      "grad_norm": 0.9986556768417358,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 10735
+    },
+    {
+      "epoch": 0.9194929770469339,
+      "grad_norm": 0.7815371155738831,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 10736
+    },
+    {
+      "epoch": 0.9195786228160329,
+      "grad_norm": 1.0067360401153564,
+      "learning_rate": 0.0001,
+      "loss": 1.0179,
+      "step": 10737
+    },
+    {
+      "epoch": 0.919664268585132,
+      "grad_norm": 0.6491672396659851,
+      "learning_rate": 0.0001,
+      "loss": 0.9102,
+      "step": 10738
+    },
+    {
+      "epoch": 0.9197499143542309,
+      "grad_norm": 0.5230492949485779,
+      "learning_rate": 0.0001,
+      "loss": 0.8442,
+      "step": 10739
+    },
+    {
+      "epoch": 0.9198355601233299,
+      "grad_norm": 0.9889001846313477,
+      "learning_rate": 0.0001,
+      "loss": 0.806,
+      "step": 10740
+    },
+    {
+      "epoch": 0.9199212058924289,
+      "grad_norm": 1.479846715927124,
+      "learning_rate": 0.0001,
+      "loss": 1.2637,
+      "step": 10741
+    },
+    {
+      "epoch": 0.920006851661528,
+      "grad_norm": 1.2847141027450562,
+      "learning_rate": 0.0001,
+      "loss": 1.0663,
+      "step": 10742
+    },
+    {
+      "epoch": 0.9200924974306269,
+      "grad_norm": 1.0133187770843506,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 10743
+    },
+    {
+      "epoch": 0.9201781431997259,
+      "grad_norm": 1.2272825241088867,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 10744
+    },
+    {
+      "epoch": 0.920263788968825,
+      "grad_norm": 1.038925051689148,
+      "learning_rate": 0.0001,
+      "loss": 1.3211,
+      "step": 10745
+    },
+    {
+      "epoch": 0.920349434737924,
+      "grad_norm": 1.097294807434082,
+      "learning_rate": 0.0001,
+      "loss": 1.9393,
+      "step": 10746
+    },
+    {
+      "epoch": 0.9204350805070229,
+      "grad_norm": 0.7133694887161255,
+      "learning_rate": 0.0001,
+      "loss": 1.0457,
+      "step": 10747
+    },
+    {
+      "epoch": 0.9205207262761219,
+      "grad_norm": 0.7429875135421753,
+      "learning_rate": 0.0001,
+      "loss": 1.1866,
+      "step": 10748
+    },
+    {
+      "epoch": 0.920606372045221,
+      "grad_norm": 0.3685433566570282,
+      "learning_rate": 0.0001,
+      "loss": 0.2111,
+      "step": 10749
+    },
+    {
+      "epoch": 0.92069201781432,
+      "grad_norm": 0.7697542905807495,
+      "learning_rate": 0.0001,
+      "loss": 1.0633,
+      "step": 10750
+    },
+    {
+      "epoch": 0.920777663583419,
+      "grad_norm": 1.077475666999817,
+      "learning_rate": 0.0001,
+      "loss": 1.1024,
+      "step": 10751
+    },
+    {
+      "epoch": 0.920863309352518,
+      "grad_norm": 1.1500669717788696,
+      "learning_rate": 0.0001,
+      "loss": 0.9805,
+      "step": 10752
+    },
+    {
+      "epoch": 0.920948955121617,
+      "grad_norm": 1.3529186248779297,
+      "learning_rate": 0.0001,
+      "loss": 1.3046,
+      "step": 10753
+    },
+    {
+      "epoch": 0.921034600890716,
+      "grad_norm": 1.053628921508789,
+      "learning_rate": 0.0001,
+      "loss": 1.3419,
+      "step": 10754
+    },
+    {
+      "epoch": 0.921120246659815,
+      "grad_norm": 0.9260209798812866,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 10755
+    },
+    {
+      "epoch": 0.921205892428914,
+      "grad_norm": 1.305408239364624,
+      "learning_rate": 0.0001,
+      "loss": 0.8447,
+      "step": 10756
+    },
+    {
+      "epoch": 0.921291538198013,
+      "grad_norm": 0.9424445629119873,
+      "learning_rate": 0.0001,
+      "loss": 1.2228,
+      "step": 10757
+    },
+    {
+      "epoch": 0.921377183967112,
+      "grad_norm": 0.8419443964958191,
+      "learning_rate": 0.0001,
+      "loss": 1.2331,
+      "step": 10758
+    },
+    {
+      "epoch": 0.9214628297362111,
+      "grad_norm": 1.1803592443466187,
+      "learning_rate": 0.0001,
+      "loss": 1.1203,
+      "step": 10759
+    },
+    {
+      "epoch": 0.92154847550531,
+      "grad_norm": 0.4826277792453766,
+      "learning_rate": 0.0001,
+      "loss": 0.3685,
+      "step": 10760
+    },
+    {
+      "epoch": 0.921634121274409,
+      "grad_norm": 0.6664538979530334,
+      "learning_rate": 0.0001,
+      "loss": 1.1649,
+      "step": 10761
+    },
+    {
+      "epoch": 0.921719767043508,
+      "grad_norm": 1.2226828336715698,
+      "learning_rate": 0.0001,
+      "loss": 1.0078,
+      "step": 10762
+    },
+    {
+      "epoch": 0.9218054128126071,
+      "grad_norm": 0.7565990686416626,
+      "learning_rate": 0.0001,
+      "loss": 0.9321,
+      "step": 10763
+    },
+    {
+      "epoch": 0.9218910585817061,
+      "grad_norm": 0.6113494634628296,
+      "learning_rate": 0.0001,
+      "loss": 0.472,
+      "step": 10764
+    },
+    {
+      "epoch": 0.921976704350805,
+      "grad_norm": 1.4949008226394653,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 10765
+    },
+    {
+      "epoch": 0.9220623501199041,
+      "grad_norm": 1.3345504999160767,
+      "learning_rate": 0.0001,
+      "loss": 0.807,
+      "step": 10766
+    },
+    {
+      "epoch": 0.9221479958890031,
+      "grad_norm": 0.6946737170219421,
+      "learning_rate": 0.0001,
+      "loss": 1.1577,
+      "step": 10767
+    },
+    {
+      "epoch": 0.9222336416581021,
+      "grad_norm": 0.589728832244873,
+      "learning_rate": 0.0001,
+      "loss": 1.028,
+      "step": 10768
+    },
+    {
+      "epoch": 0.9223192874272012,
+      "grad_norm": 0.6135980486869812,
+      "learning_rate": 0.0001,
+      "loss": 0.9524,
+      "step": 10769
+    },
+    {
+      "epoch": 0.9224049331963001,
+      "grad_norm": 0.9924595355987549,
+      "learning_rate": 0.0001,
+      "loss": 1.0417,
+      "step": 10770
+    },
+    {
+      "epoch": 0.9224905789653991,
+      "grad_norm": 0.8780953884124756,
+      "learning_rate": 0.0001,
+      "loss": 0.8477,
+      "step": 10771
+    },
+    {
+      "epoch": 0.9225762247344981,
+      "grad_norm": 0.716914713382721,
+      "learning_rate": 0.0001,
+      "loss": 0.8715,
+      "step": 10772
+    },
+    {
+      "epoch": 0.9226618705035972,
+      "grad_norm": 0.668220579624176,
+      "learning_rate": 0.0001,
+      "loss": 0.2321,
+      "step": 10773
+    },
+    {
+      "epoch": 0.9227475162726961,
+      "grad_norm": 1.3898581266403198,
+      "learning_rate": 0.0001,
+      "loss": 0.8557,
+      "step": 10774
+    },
+    {
+      "epoch": 0.9228331620417951,
+      "grad_norm": 1.2596575021743774,
+      "learning_rate": 0.0001,
+      "loss": 1.1395,
+      "step": 10775
+    },
+    {
+      "epoch": 0.9229188078108942,
+      "grad_norm": 0.24727916717529297,
+      "learning_rate": 0.0001,
+      "loss": 0.2375,
+      "step": 10776
+    },
+    {
+      "epoch": 0.9230044535799932,
+      "grad_norm": 0.3560200333595276,
+      "learning_rate": 0.0001,
+      "loss": 0.5015,
+      "step": 10777
+    },
+    {
+      "epoch": 0.9230900993490921,
+      "grad_norm": 1.2498220205307007,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 10778
+    },
+    {
+      "epoch": 0.9231757451181911,
+      "grad_norm": 0.931330680847168,
+      "learning_rate": 0.0001,
+      "loss": 0.6769,
+      "step": 10779
+    },
+    {
+      "epoch": 0.9232613908872902,
+      "grad_norm": 0.9622438549995422,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 10780
+    },
+    {
+      "epoch": 0.9233470366563892,
+      "grad_norm": 0.7824144959449768,
+      "learning_rate": 0.0001,
+      "loss": 1.3195,
+      "step": 10781
+    },
+    {
+      "epoch": 0.9234326824254881,
+      "grad_norm": 1.6487141847610474,
+      "learning_rate": 0.0001,
+      "loss": 1.3225,
+      "step": 10782
+    },
+    {
+      "epoch": 0.9235183281945872,
+      "grad_norm": 1.2727549076080322,
+      "learning_rate": 0.0001,
+      "loss": 1.374,
+      "step": 10783
+    },
+    {
+      "epoch": 0.9236039739636862,
+      "grad_norm": 0.7886867523193359,
+      "learning_rate": 0.0001,
+      "loss": 1.2002,
+      "step": 10784
+    },
+    {
+      "epoch": 0.9236896197327852,
+      "grad_norm": 0.6301016211509705,
+      "learning_rate": 0.0001,
+      "loss": 1.0542,
+      "step": 10785
+    },
+    {
+      "epoch": 0.9237752655018842,
+      "grad_norm": 0.791573703289032,
+      "learning_rate": 0.0001,
+      "loss": 1.194,
+      "step": 10786
+    },
+    {
+      "epoch": 0.9238609112709832,
+      "grad_norm": 0.7299854159355164,
+      "learning_rate": 0.0001,
+      "loss": 0.9623,
+      "step": 10787
+    },
+    {
+      "epoch": 0.9239465570400822,
+      "grad_norm": 1.3698631525039673,
+      "learning_rate": 0.0001,
+      "loss": 1.0339,
+      "step": 10788
+    },
+    {
+      "epoch": 0.9240322028091812,
+      "grad_norm": 0.871344268321991,
+      "learning_rate": 0.0001,
+      "loss": 1.08,
+      "step": 10789
+    },
+    {
+      "epoch": 0.9241178485782803,
+      "grad_norm": 0.7527289986610413,
+      "learning_rate": 0.0001,
+      "loss": 1.1269,
+      "step": 10790
+    },
+    {
+      "epoch": 0.9242034943473793,
+      "grad_norm": 1.0797889232635498,
+      "learning_rate": 0.0001,
+      "loss": 0.9205,
+      "step": 10791
+    },
+    {
+      "epoch": 0.9242891401164782,
+      "grad_norm": 0.7942084670066833,
+      "learning_rate": 0.0001,
+      "loss": 1.4539,
+      "step": 10792
+    },
+    {
+      "epoch": 0.9243747858855772,
+      "grad_norm": 0.7484357357025146,
+      "learning_rate": 0.0001,
+      "loss": 1.2355,
+      "step": 10793
+    },
+    {
+      "epoch": 0.9244604316546763,
+      "grad_norm": 0.8452218174934387,
+      "learning_rate": 0.0001,
+      "loss": 0.8667,
+      "step": 10794
+    },
+    {
+      "epoch": 0.9245460774237753,
+      "grad_norm": 0.3319012224674225,
+      "learning_rate": 0.0001,
+      "loss": 0.1982,
+      "step": 10795
+    },
+    {
+      "epoch": 0.9246317231928742,
+      "grad_norm": 1.0485206842422485,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 10796
+    },
+    {
+      "epoch": 0.9247173689619733,
+      "grad_norm": 0.8756988048553467,
+      "learning_rate": 0.0001,
+      "loss": 0.5596,
+      "step": 10797
+    },
+    {
+      "epoch": 0.9248030147310723,
+      "grad_norm": 0.9822337031364441,
+      "learning_rate": 0.0001,
+      "loss": 1.056,
+      "step": 10798
+    },
+    {
+      "epoch": 0.9248886605001713,
+      "grad_norm": 0.6229698657989502,
+      "learning_rate": 0.0001,
+      "loss": 1.0012,
+      "step": 10799
+    },
+    {
+      "epoch": 0.9249743062692704,
+      "grad_norm": 0.23129093647003174,
+      "learning_rate": 0.0001,
+      "loss": 0.4787,
+      "step": 10800
+    },
+    {
+      "epoch": 0.9250599520383693,
+      "grad_norm": 0.5882505178451538,
+      "learning_rate": 0.0001,
+      "loss": 0.5562,
+      "step": 10801
+    },
+    {
+      "epoch": 0.9251455978074683,
+      "grad_norm": 1.0643784999847412,
+      "learning_rate": 0.0001,
+      "loss": 0.9725,
+      "step": 10802
+    },
+    {
+      "epoch": 0.9252312435765673,
+      "grad_norm": 0.5945513248443604,
+      "learning_rate": 0.0001,
+      "loss": 0.3539,
+      "step": 10803
+    },
+    {
+      "epoch": 0.9253168893456664,
+      "grad_norm": 1.2225199937820435,
+      "learning_rate": 0.0001,
+      "loss": 0.7286,
+      "step": 10804
+    },
+    {
+      "epoch": 0.9254025351147653,
+      "grad_norm": 1.1577205657958984,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 10805
+    },
+    {
+      "epoch": 0.9254881808838643,
+      "grad_norm": 0.9652645587921143,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 10806
+    },
+    {
+      "epoch": 0.9255738266529634,
+      "grad_norm": 0.7569334506988525,
+      "learning_rate": 0.0001,
+      "loss": 0.7625,
+      "step": 10807
+    },
+    {
+      "epoch": 0.9256594724220624,
+      "grad_norm": 0.5340976119041443,
+      "learning_rate": 0.0001,
+      "loss": 0.5484,
+      "step": 10808
+    },
+    {
+      "epoch": 0.9257451181911613,
+      "grad_norm": 0.9468872547149658,
+      "learning_rate": 0.0001,
+      "loss": 0.9121,
+      "step": 10809
+    },
+    {
+      "epoch": 0.9258307639602603,
+      "grad_norm": 0.8576138615608215,
+      "learning_rate": 0.0001,
+      "loss": 1.1064,
+      "step": 10810
+    },
+    {
+      "epoch": 0.9259164097293594,
+      "grad_norm": 0.8741364479064941,
+      "learning_rate": 0.0001,
+      "loss": 1.3524,
+      "step": 10811
+    },
+    {
+      "epoch": 0.9260020554984584,
+      "grad_norm": 0.8703165650367737,
+      "learning_rate": 0.0001,
+      "loss": 0.2848,
+      "step": 10812
+    },
+    {
+      "epoch": 0.9260877012675574,
+      "grad_norm": 0.7843953967094421,
+      "learning_rate": 0.0001,
+      "loss": 1.1843,
+      "step": 10813
+    },
+    {
+      "epoch": 0.9261733470366564,
+      "grad_norm": 1.5972179174423218,
+      "learning_rate": 0.0001,
+      "loss": 1.0117,
+      "step": 10814
+    },
+    {
+      "epoch": 0.9262589928057554,
+      "grad_norm": 0.804937481880188,
+      "learning_rate": 0.0001,
+      "loss": 1.2799,
+      "step": 10815
+    },
+    {
+      "epoch": 0.9263446385748544,
+      "grad_norm": 0.8690212368965149,
+      "learning_rate": 0.0001,
+      "loss": 1.0434,
+      "step": 10816
+    },
+    {
+      "epoch": 0.9264302843439534,
+      "grad_norm": 0.7940465807914734,
+      "learning_rate": 0.0001,
+      "loss": 0.9653,
+      "step": 10817
+    },
+    {
+      "epoch": 0.9265159301130524,
+      "grad_norm": 0.8775970339775085,
+      "learning_rate": 0.0001,
+      "loss": 0.6855,
+      "step": 10818
+    },
+    {
+      "epoch": 0.9266015758821514,
+      "grad_norm": 1.0334919691085815,
+      "learning_rate": 0.0001,
+      "loss": 0.9759,
+      "step": 10819
+    },
+    {
+      "epoch": 0.9266872216512504,
+      "grad_norm": 0.5823876261711121,
+      "learning_rate": 0.0001,
+      "loss": 0.5173,
+      "step": 10820
+    },
+    {
+      "epoch": 0.9267728674203495,
+      "grad_norm": 0.6938380002975464,
+      "learning_rate": 0.0001,
+      "loss": 0.9622,
+      "step": 10821
+    },
+    {
+      "epoch": 0.9268585131894485,
+      "grad_norm": 0.8832963109016418,
+      "learning_rate": 0.0001,
+      "loss": 0.9978,
+      "step": 10822
+    },
+    {
+      "epoch": 0.9269441589585474,
+      "grad_norm": 0.7077249884605408,
+      "learning_rate": 0.0001,
+      "loss": 0.9542,
+      "step": 10823
+    },
+    {
+      "epoch": 0.9270298047276464,
+      "grad_norm": 1.1908472776412964,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 10824
+    },
+    {
+      "epoch": 0.9271154504967455,
+      "grad_norm": 0.8399367928504944,
+      "learning_rate": 0.0001,
+      "loss": 1.0511,
+      "step": 10825
+    },
+    {
+      "epoch": 0.9272010962658445,
+      "grad_norm": 1.5117889642715454,
+      "learning_rate": 0.0001,
+      "loss": 1.1129,
+      "step": 10826
+    },
+    {
+      "epoch": 0.9272867420349434,
+      "grad_norm": 0.7639455199241638,
+      "learning_rate": 0.0001,
+      "loss": 1.1951,
+      "step": 10827
+    },
+    {
+      "epoch": 0.9273723878040425,
+      "grad_norm": 0.8791640996932983,
+      "learning_rate": 0.0001,
+      "loss": 0.7955,
+      "step": 10828
+    },
+    {
+      "epoch": 0.9274580335731415,
+      "grad_norm": 0.5630981922149658,
+      "learning_rate": 0.0001,
+      "loss": 0.8177,
+      "step": 10829
+    },
+    {
+      "epoch": 0.9275436793422405,
+      "grad_norm": 1.0007959604263306,
+      "learning_rate": 0.0001,
+      "loss": 0.7288,
+      "step": 10830
+    },
+    {
+      "epoch": 0.9276293251113396,
+      "grad_norm": 1.9870657920837402,
+      "learning_rate": 0.0001,
+      "loss": 0.8909,
+      "step": 10831
+    },
+    {
+      "epoch": 0.9277149708804385,
+      "grad_norm": 0.6437454223632812,
+      "learning_rate": 0.0001,
+      "loss": 0.5976,
+      "step": 10832
+    },
+    {
+      "epoch": 0.9278006166495375,
+      "grad_norm": 0.6606407761573792,
+      "learning_rate": 0.0001,
+      "loss": 0.5116,
+      "step": 10833
+    },
+    {
+      "epoch": 0.9278862624186365,
+      "grad_norm": 0.7339625954627991,
+      "learning_rate": 0.0001,
+      "loss": 0.7606,
+      "step": 10834
+    },
+    {
+      "epoch": 0.9279719081877356,
+      "grad_norm": 0.990536630153656,
+      "learning_rate": 0.0001,
+      "loss": 0.849,
+      "step": 10835
+    },
+    {
+      "epoch": 0.9280575539568345,
+      "grad_norm": 0.6647485494613647,
+      "learning_rate": 0.0001,
+      "loss": 0.5606,
+      "step": 10836
+    },
+    {
+      "epoch": 0.9281431997259335,
+      "grad_norm": 0.7931591272354126,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 10837
+    },
+    {
+      "epoch": 0.9282288454950326,
+      "grad_norm": 0.7058805823326111,
+      "learning_rate": 0.0001,
+      "loss": 0.8361,
+      "step": 10838
+    },
+    {
+      "epoch": 0.9283144912641316,
+      "grad_norm": 1.2345960140228271,
+      "learning_rate": 0.0001,
+      "loss": 1.1767,
+      "step": 10839
+    },
+    {
+      "epoch": 0.9284001370332305,
+      "grad_norm": 0.6791106462478638,
+      "learning_rate": 0.0001,
+      "loss": 1.3241,
+      "step": 10840
+    },
+    {
+      "epoch": 0.9284857828023295,
+      "grad_norm": 0.7877445220947266,
+      "learning_rate": 0.0001,
+      "loss": 0.4939,
+      "step": 10841
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 1.2280219793319702,
+      "learning_rate": 0.0001,
+      "loss": 1.3367,
+      "step": 10842
+    },
+    {
+      "epoch": 0.9286570743405276,
+      "grad_norm": 0.6135997176170349,
+      "learning_rate": 0.0001,
+      "loss": 0.4817,
+      "step": 10843
+    },
+    {
+      "epoch": 0.9287427201096266,
+      "grad_norm": 0.7466526627540588,
+      "learning_rate": 0.0001,
+      "loss": 1.267,
+      "step": 10844
+    },
+    {
+      "epoch": 0.9288283658787256,
+      "grad_norm": 0.8248903751373291,
+      "learning_rate": 0.0001,
+      "loss": 0.5566,
+      "step": 10845
+    },
+    {
+      "epoch": 0.9289140116478246,
+      "grad_norm": 2.330495834350586,
+      "learning_rate": 0.0001,
+      "loss": 0.8129,
+      "step": 10846
+    },
+    {
+      "epoch": 0.9289996574169236,
+      "grad_norm": 0.8837090134620667,
+      "learning_rate": 0.0001,
+      "loss": 0.6037,
+      "step": 10847
+    },
+    {
+      "epoch": 0.9290853031860226,
+      "grad_norm": 1.1087682247161865,
+      "learning_rate": 0.0001,
+      "loss": 0.8098,
+      "step": 10848
+    },
+    {
+      "epoch": 0.9291709489551216,
+      "grad_norm": 0.6494540572166443,
+      "learning_rate": 0.0001,
+      "loss": 0.4943,
+      "step": 10849
+    },
+    {
+      "epoch": 0.9292565947242206,
+      "grad_norm": 0.6289742588996887,
+      "learning_rate": 0.0001,
+      "loss": 0.4646,
+      "step": 10850
+    },
+    {
+      "epoch": 0.9293422404933196,
+      "grad_norm": 0.5964198112487793,
+      "learning_rate": 0.0001,
+      "loss": 0.8858,
+      "step": 10851
+    },
+    {
+      "epoch": 0.9294278862624187,
+      "grad_norm": 0.7647263407707214,
+      "learning_rate": 0.0001,
+      "loss": 0.4722,
+      "step": 10852
+    },
+    {
+      "epoch": 0.9295135320315177,
+      "grad_norm": 1.0775479078292847,
+      "learning_rate": 0.0001,
+      "loss": 0.9494,
+      "step": 10853
+    },
+    {
+      "epoch": 0.9295991778006166,
+      "grad_norm": 0.5898659825325012,
+      "learning_rate": 0.0001,
+      "loss": 1.0699,
+      "step": 10854
+    },
+    {
+      "epoch": 0.9296848235697156,
+      "grad_norm": 0.5267601609230042,
+      "learning_rate": 0.0001,
+      "loss": 0.5235,
+      "step": 10855
+    },
+    {
+      "epoch": 0.9297704693388147,
+      "grad_norm": 0.6521083116531372,
+      "learning_rate": 0.0001,
+      "loss": 0.6324,
+      "step": 10856
+    },
+    {
+      "epoch": 0.9298561151079137,
+      "grad_norm": 1.4624738693237305,
+      "learning_rate": 0.0001,
+      "loss": 1.0776,
+      "step": 10857
+    },
+    {
+      "epoch": 0.9299417608770126,
+      "grad_norm": 0.5047433376312256,
+      "learning_rate": 0.0001,
+      "loss": 0.6999,
+      "step": 10858
+    },
+    {
+      "epoch": 0.9300274066461117,
+      "grad_norm": 1.3092674016952515,
+      "learning_rate": 0.0001,
+      "loss": 1.237,
+      "step": 10859
+    },
+    {
+      "epoch": 0.9301130524152107,
+      "grad_norm": 0.7045691013336182,
+      "learning_rate": 0.0001,
+      "loss": 0.8362,
+      "step": 10860
+    },
+    {
+      "epoch": 0.9301986981843097,
+      "grad_norm": 1.0722391605377197,
+      "learning_rate": 0.0001,
+      "loss": 1.198,
+      "step": 10861
+    },
+    {
+      "epoch": 0.9302843439534088,
+      "grad_norm": 0.7994459271430969,
+      "learning_rate": 0.0001,
+      "loss": 0.8696,
+      "step": 10862
+    },
+    {
+      "epoch": 0.9303699897225077,
+      "grad_norm": 0.7976245880126953,
+      "learning_rate": 0.0001,
+      "loss": 1.0154,
+      "step": 10863
+    },
+    {
+      "epoch": 0.9304556354916067,
+      "grad_norm": 1.175114393234253,
+      "learning_rate": 0.0001,
+      "loss": 1.219,
+      "step": 10864
+    },
+    {
+      "epoch": 0.9305412812607057,
+      "grad_norm": 1.3174127340316772,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 10865
+    },
+    {
+      "epoch": 0.9306269270298048,
+      "grad_norm": 0.9168612957000732,
+      "learning_rate": 0.0001,
+      "loss": 0.9921,
+      "step": 10866
+    },
+    {
+      "epoch": 0.9307125727989037,
+      "grad_norm": 0.7132528424263,
+      "learning_rate": 0.0001,
+      "loss": 0.5201,
+      "step": 10867
+    },
+    {
+      "epoch": 0.9307982185680027,
+      "grad_norm": 1.195343255996704,
+      "learning_rate": 0.0001,
+      "loss": 1.2366,
+      "step": 10868
+    },
+    {
+      "epoch": 0.9308838643371018,
+      "grad_norm": 1.256960153579712,
+      "learning_rate": 0.0001,
+      "loss": 0.1989,
+      "step": 10869
+    },
+    {
+      "epoch": 0.9309695101062008,
+      "grad_norm": 0.7780666351318359,
+      "learning_rate": 0.0001,
+      "loss": 1.3115,
+      "step": 10870
+    },
+    {
+      "epoch": 0.9310551558752997,
+      "grad_norm": 0.5799373388290405,
+      "learning_rate": 0.0001,
+      "loss": 0.3055,
+      "step": 10871
+    },
+    {
+      "epoch": 0.9311408016443987,
+      "grad_norm": 0.36707401275634766,
+      "learning_rate": 0.0001,
+      "loss": 0.4552,
+      "step": 10872
+    },
+    {
+      "epoch": 0.9312264474134978,
+      "grad_norm": 1.0923163890838623,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 10873
+    },
+    {
+      "epoch": 0.9313120931825968,
+      "grad_norm": 0.685183048248291,
+      "learning_rate": 0.0001,
+      "loss": 0.2199,
+      "step": 10874
+    },
+    {
+      "epoch": 0.9313977389516958,
+      "grad_norm": 0.5984359383583069,
+      "learning_rate": 0.0001,
+      "loss": 1.0778,
+      "step": 10875
+    },
+    {
+      "epoch": 0.9314833847207948,
+      "grad_norm": 0.9010688662528992,
+      "learning_rate": 0.0001,
+      "loss": 0.6471,
+      "step": 10876
+    },
+    {
+      "epoch": 0.9315690304898938,
+      "grad_norm": 1.1164920330047607,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 10877
+    },
+    {
+      "epoch": 0.9316546762589928,
+      "grad_norm": 1.406424880027771,
+      "learning_rate": 0.0001,
+      "loss": 1.3703,
+      "step": 10878
+    },
+    {
+      "epoch": 0.9317403220280918,
+      "grad_norm": 0.6704549193382263,
+      "learning_rate": 0.0001,
+      "loss": 0.9915,
+      "step": 10879
+    },
+    {
+      "epoch": 0.9318259677971908,
+      "grad_norm": 0.9636950492858887,
+      "learning_rate": 0.0001,
+      "loss": 1.1062,
+      "step": 10880
+    },
+    {
+      "epoch": 0.9319116135662898,
+      "grad_norm": 0.6664835810661316,
+      "learning_rate": 0.0001,
+      "loss": 0.4447,
+      "step": 10881
+    },
+    {
+      "epoch": 0.9319972593353888,
+      "grad_norm": 0.744727373123169,
+      "learning_rate": 0.0001,
+      "loss": 0.776,
+      "step": 10882
+    },
+    {
+      "epoch": 0.9320829051044879,
+      "grad_norm": 1.1188607215881348,
+      "learning_rate": 0.0001,
+      "loss": 1.1899,
+      "step": 10883
+    },
+    {
+      "epoch": 0.9321685508735869,
+      "grad_norm": 1.3642823696136475,
+      "learning_rate": 0.0001,
+      "loss": 1.2975,
+      "step": 10884
+    },
+    {
+      "epoch": 0.9322541966426858,
+      "grad_norm": 1.4860433340072632,
+      "learning_rate": 0.0001,
+      "loss": 1.0086,
+      "step": 10885
+    },
+    {
+      "epoch": 0.9323398424117848,
+      "grad_norm": 1.1462475061416626,
+      "learning_rate": 0.0001,
+      "loss": 1.0474,
+      "step": 10886
+    },
+    {
+      "epoch": 0.9324254881808839,
+      "grad_norm": 0.9392229318618774,
+      "learning_rate": 0.0001,
+      "loss": 0.9569,
+      "step": 10887
+    },
+    {
+      "epoch": 0.9325111339499829,
+      "grad_norm": 0.8199727535247803,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 10888
+    },
+    {
+      "epoch": 0.9325967797190818,
+      "grad_norm": 1.0908629894256592,
+      "learning_rate": 0.0001,
+      "loss": 0.8831,
+      "step": 10889
+    },
+    {
+      "epoch": 0.9326824254881809,
+      "grad_norm": 0.7309316992759705,
+      "learning_rate": 0.0001,
+      "loss": 0.656,
+      "step": 10890
+    },
+    {
+      "epoch": 0.9327680712572799,
+      "grad_norm": 0.8265985250473022,
+      "learning_rate": 0.0001,
+      "loss": 1.1867,
+      "step": 10891
+    },
+    {
+      "epoch": 0.9328537170263789,
+      "grad_norm": 0.897254467010498,
+      "learning_rate": 0.0001,
+      "loss": 1.3422,
+      "step": 10892
+    },
+    {
+      "epoch": 0.932939362795478,
+      "grad_norm": 0.9270237684249878,
+      "learning_rate": 0.0001,
+      "loss": 1.0114,
+      "step": 10893
+    },
+    {
+      "epoch": 0.9330250085645769,
+      "grad_norm": 1.3078233003616333,
+      "learning_rate": 0.0001,
+      "loss": 1.3273,
+      "step": 10894
+    },
+    {
+      "epoch": 0.9331106543336759,
+      "grad_norm": 0.7259948253631592,
+      "learning_rate": 0.0001,
+      "loss": 0.7974,
+      "step": 10895
+    },
+    {
+      "epoch": 0.9331963001027749,
+      "grad_norm": 1.0207674503326416,
+      "learning_rate": 0.0001,
+      "loss": 1.0824,
+      "step": 10896
+    },
+    {
+      "epoch": 0.933281945871874,
+      "grad_norm": 0.9447816610336304,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 10897
+    },
+    {
+      "epoch": 0.9333675916409729,
+      "grad_norm": 0.7208675742149353,
+      "learning_rate": 0.0001,
+      "loss": 0.7628,
+      "step": 10898
+    },
+    {
+      "epoch": 0.9334532374100719,
+      "grad_norm": 1.1319624185562134,
+      "learning_rate": 0.0001,
+      "loss": 0.6074,
+      "step": 10899
+    },
+    {
+      "epoch": 0.933538883179171,
+      "grad_norm": 0.4333060085773468,
+      "learning_rate": 0.0001,
+      "loss": 0.8436,
+      "step": 10900
+    },
+    {
+      "epoch": 0.93362452894827,
+      "grad_norm": 1.1564470529556274,
+      "learning_rate": 0.0001,
+      "loss": 0.983,
+      "step": 10901
+    },
+    {
+      "epoch": 0.933710174717369,
+      "grad_norm": 0.7146464586257935,
+      "learning_rate": 0.0001,
+      "loss": 1.334,
+      "step": 10902
+    },
+    {
+      "epoch": 0.9337958204864679,
+      "grad_norm": 0.7897799611091614,
+      "learning_rate": 0.0001,
+      "loss": 1.1345,
+      "step": 10903
+    },
+    {
+      "epoch": 0.933881466255567,
+      "grad_norm": 1.467822790145874,
+      "learning_rate": 0.0001,
+      "loss": 0.8106,
+      "step": 10904
+    },
+    {
+      "epoch": 0.933967112024666,
+      "grad_norm": 0.8060843348503113,
+      "learning_rate": 0.0001,
+      "loss": 1.0812,
+      "step": 10905
+    },
+    {
+      "epoch": 0.934052757793765,
+      "grad_norm": 0.9946094751358032,
+      "learning_rate": 0.0001,
+      "loss": 1.2318,
+      "step": 10906
+    },
+    {
+      "epoch": 0.934138403562864,
+      "grad_norm": 1.0669655799865723,
+      "learning_rate": 0.0001,
+      "loss": 0.6708,
+      "step": 10907
+    },
+    {
+      "epoch": 0.934224049331963,
+      "grad_norm": 0.9473127722740173,
+      "learning_rate": 0.0001,
+      "loss": 1.008,
+      "step": 10908
+    },
+    {
+      "epoch": 0.934309695101062,
+      "grad_norm": 0.7439343929290771,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 10909
+    },
+    {
+      "epoch": 0.934395340870161,
+      "grad_norm": 0.9755323529243469,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 10910
+    },
+    {
+      "epoch": 0.93448098663926,
+      "grad_norm": 1.2609480619430542,
+      "learning_rate": 0.0001,
+      "loss": 1.125,
+      "step": 10911
+    },
+    {
+      "epoch": 0.934566632408359,
+      "grad_norm": 0.9875310063362122,
+      "learning_rate": 0.0001,
+      "loss": 0.935,
+      "step": 10912
+    },
+    {
+      "epoch": 0.934652278177458,
+      "grad_norm": 1.724806785583496,
+      "learning_rate": 0.0001,
+      "loss": 0.4986,
+      "step": 10913
+    },
+    {
+      "epoch": 0.9347379239465571,
+      "grad_norm": 0.3507496118545532,
+      "learning_rate": 0.0001,
+      "loss": 0.104,
+      "step": 10914
+    },
+    {
+      "epoch": 0.934823569715656,
+      "grad_norm": 0.8000248670578003,
+      "learning_rate": 0.0001,
+      "loss": 1.2369,
+      "step": 10915
+    },
+    {
+      "epoch": 0.934909215484755,
+      "grad_norm": 1.043148159980774,
+      "learning_rate": 0.0001,
+      "loss": 0.9726,
+      "step": 10916
+    },
+    {
+      "epoch": 0.9349948612538541,
+      "grad_norm": 0.668446958065033,
+      "learning_rate": 0.0001,
+      "loss": 1.0771,
+      "step": 10917
+    },
+    {
+      "epoch": 0.9350805070229531,
+      "grad_norm": 0.7826523184776306,
+      "learning_rate": 0.0001,
+      "loss": 0.959,
+      "step": 10918
+    },
+    {
+      "epoch": 0.9351661527920521,
+      "grad_norm": 0.6248810887336731,
+      "learning_rate": 0.0001,
+      "loss": 0.5595,
+      "step": 10919
+    },
+    {
+      "epoch": 0.935251798561151,
+      "grad_norm": 1.0423357486724854,
+      "learning_rate": 0.0001,
+      "loss": 0.9362,
+      "step": 10920
+    },
+    {
+      "epoch": 0.9353374443302501,
+      "grad_norm": 1.3636916875839233,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 10921
+    },
+    {
+      "epoch": 0.9354230900993491,
+      "grad_norm": 0.6542587280273438,
+      "learning_rate": 0.0001,
+      "loss": 1.2001,
+      "step": 10922
+    },
+    {
+      "epoch": 0.9355087358684481,
+      "grad_norm": 0.831851065158844,
+      "learning_rate": 0.0001,
+      "loss": 0.7514,
+      "step": 10923
+    },
+    {
+      "epoch": 0.9355943816375472,
+      "grad_norm": 0.7405803203582764,
+      "learning_rate": 0.0001,
+      "loss": 0.6382,
+      "step": 10924
+    },
+    {
+      "epoch": 0.9356800274066461,
+      "grad_norm": 0.6237040162086487,
+      "learning_rate": 0.0001,
+      "loss": 1.2213,
+      "step": 10925
+    },
+    {
+      "epoch": 0.9357656731757451,
+      "grad_norm": 0.900044858455658,
+      "learning_rate": 0.0001,
+      "loss": 0.806,
+      "step": 10926
+    },
+    {
+      "epoch": 0.9358513189448441,
+      "grad_norm": 0.808199405670166,
+      "learning_rate": 0.0001,
+      "loss": 0.9095,
+      "step": 10927
+    },
+    {
+      "epoch": 0.9359369647139432,
+      "grad_norm": 0.8167599439620972,
+      "learning_rate": 0.0001,
+      "loss": 0.5523,
+      "step": 10928
+    },
+    {
+      "epoch": 0.9360226104830421,
+      "grad_norm": 1.8206366300582886,
+      "learning_rate": 0.0001,
+      "loss": 1.2167,
+      "step": 10929
+    },
+    {
+      "epoch": 0.9361082562521411,
+      "grad_norm": 1.117928147315979,
+      "learning_rate": 0.0001,
+      "loss": 0.9061,
+      "step": 10930
+    },
+    {
+      "epoch": 0.9361939020212402,
+      "grad_norm": 0.9613667726516724,
+      "learning_rate": 0.0001,
+      "loss": 0.9918,
+      "step": 10931
+    },
+    {
+      "epoch": 0.9362795477903392,
+      "grad_norm": 0.5714908838272095,
+      "learning_rate": 0.0001,
+      "loss": 0.2763,
+      "step": 10932
+    },
+    {
+      "epoch": 0.9363651935594381,
+      "grad_norm": 0.7861186861991882,
+      "learning_rate": 0.0001,
+      "loss": 0.584,
+      "step": 10933
+    },
+    {
+      "epoch": 0.9364508393285371,
+      "grad_norm": 0.7310831546783447,
+      "learning_rate": 0.0001,
+      "loss": 0.1602,
+      "step": 10934
+    },
+    {
+      "epoch": 0.9365364850976362,
+      "grad_norm": 0.6026157140731812,
+      "learning_rate": 0.0001,
+      "loss": 1.0585,
+      "step": 10935
+    },
+    {
+      "epoch": 0.9366221308667352,
+      "grad_norm": 0.781816840171814,
+      "learning_rate": 0.0001,
+      "loss": 0.6875,
+      "step": 10936
+    },
+    {
+      "epoch": 0.9367077766358342,
+      "grad_norm": 0.7007602453231812,
+      "learning_rate": 0.0001,
+      "loss": 1.0777,
+      "step": 10937
+    },
+    {
+      "epoch": 0.9367934224049332,
+      "grad_norm": 0.6651493310928345,
+      "learning_rate": 0.0001,
+      "loss": 0.6615,
+      "step": 10938
+    },
+    {
+      "epoch": 0.9368790681740322,
+      "grad_norm": 1.1406958103179932,
+      "learning_rate": 0.0001,
+      "loss": 0.6775,
+      "step": 10939
+    },
+    {
+      "epoch": 0.9369647139431312,
+      "grad_norm": 0.7849461436271667,
+      "learning_rate": 0.0001,
+      "loss": 0.8385,
+      "step": 10940
+    },
+    {
+      "epoch": 0.9370503597122302,
+      "grad_norm": 1.1418557167053223,
+      "learning_rate": 0.0001,
+      "loss": 1.2621,
+      "step": 10941
+    },
+    {
+      "epoch": 0.9371360054813292,
+      "grad_norm": 0.7767273187637329,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 10942
+    },
+    {
+      "epoch": 0.9372216512504282,
+      "grad_norm": 1.3460228443145752,
+      "learning_rate": 0.0001,
+      "loss": 1.098,
+      "step": 10943
+    },
+    {
+      "epoch": 0.9373072970195272,
+      "grad_norm": 0.7769716382026672,
+      "learning_rate": 0.0001,
+      "loss": 0.7259,
+      "step": 10944
+    },
+    {
+      "epoch": 0.9373929427886263,
+      "grad_norm": 0.9387155175209045,
+      "learning_rate": 0.0001,
+      "loss": 0.5312,
+      "step": 10945
+    },
+    {
+      "epoch": 0.9374785885577253,
+      "grad_norm": 1.3415720462799072,
+      "learning_rate": 0.0001,
+      "loss": 1.0406,
+      "step": 10946
+    },
+    {
+      "epoch": 0.9375642343268242,
+      "grad_norm": 0.7550864815711975,
+      "learning_rate": 0.0001,
+      "loss": 1.356,
+      "step": 10947
+    },
+    {
+      "epoch": 0.9376498800959233,
+      "grad_norm": 0.6813321113586426,
+      "learning_rate": 0.0001,
+      "loss": 0.6487,
+      "step": 10948
+    },
+    {
+      "epoch": 0.9377355258650223,
+      "grad_norm": 0.66719651222229,
+      "learning_rate": 0.0001,
+      "loss": 0.5157,
+      "step": 10949
+    },
+    {
+      "epoch": 0.9378211716341213,
+      "grad_norm": 0.6302657127380371,
+      "learning_rate": 0.0001,
+      "loss": 0.9999,
+      "step": 10950
+    },
+    {
+      "epoch": 0.9379068174032202,
+      "grad_norm": 1.5602327585220337,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 10951
+    },
+    {
+      "epoch": 0.9379924631723193,
+      "grad_norm": 1.1223819255828857,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 10952
+    },
+    {
+      "epoch": 0.9380781089414183,
+      "grad_norm": 0.7426851987838745,
+      "learning_rate": 0.0001,
+      "loss": 0.5708,
+      "step": 10953
+    },
+    {
+      "epoch": 0.9381637547105173,
+      "grad_norm": 1.1046366691589355,
+      "learning_rate": 0.0001,
+      "loss": 0.9303,
+      "step": 10954
+    },
+    {
+      "epoch": 0.9382494004796164,
+      "grad_norm": 0.3221035897731781,
+      "learning_rate": 0.0001,
+      "loss": 0.4645,
+      "step": 10955
+    },
+    {
+      "epoch": 0.9383350462487153,
+      "grad_norm": 1.0941672325134277,
+      "learning_rate": 0.0001,
+      "loss": 1.1063,
+      "step": 10956
+    },
+    {
+      "epoch": 0.9384206920178143,
+      "grad_norm": 1.3735613822937012,
+      "learning_rate": 0.0001,
+      "loss": 0.9253,
+      "step": 10957
+    },
+    {
+      "epoch": 0.9385063377869133,
+      "grad_norm": 0.3338514268398285,
+      "learning_rate": 0.0001,
+      "loss": 0.0964,
+      "step": 10958
+    },
+    {
+      "epoch": 0.9385919835560124,
+      "grad_norm": 0.9418327808380127,
+      "learning_rate": 0.0001,
+      "loss": 1.025,
+      "step": 10959
+    },
+    {
+      "epoch": 0.9386776293251113,
+      "grad_norm": 0.8098942041397095,
+      "learning_rate": 0.0001,
+      "loss": 0.845,
+      "step": 10960
+    },
+    {
+      "epoch": 0.9387632750942103,
+      "grad_norm": 0.7849554419517517,
+      "learning_rate": 0.0001,
+      "loss": 0.9165,
+      "step": 10961
+    },
+    {
+      "epoch": 0.9388489208633094,
+      "grad_norm": 0.9862844944000244,
+      "learning_rate": 0.0001,
+      "loss": 0.9023,
+      "step": 10962
+    },
+    {
+      "epoch": 0.9389345666324084,
+      "grad_norm": 0.7482150197029114,
+      "learning_rate": 0.0001,
+      "loss": 0.8256,
+      "step": 10963
+    },
+    {
+      "epoch": 0.9390202124015073,
+      "grad_norm": 1.042069911956787,
+      "learning_rate": 0.0001,
+      "loss": 0.6806,
+      "step": 10964
+    },
+    {
+      "epoch": 0.9391058581706063,
+      "grad_norm": 0.6389273405075073,
+      "learning_rate": 0.0001,
+      "loss": 0.5248,
+      "step": 10965
+    },
+    {
+      "epoch": 0.9391915039397054,
+      "grad_norm": 0.7014659643173218,
+      "learning_rate": 0.0001,
+      "loss": 1.0454,
+      "step": 10966
+    },
+    {
+      "epoch": 0.9392771497088044,
+      "grad_norm": 0.7374102473258972,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 10967
+    },
+    {
+      "epoch": 0.9393627954779034,
+      "grad_norm": 0.7666456699371338,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 10968
+    },
+    {
+      "epoch": 0.9394484412470024,
+      "grad_norm": 3.166416883468628,
+      "learning_rate": 0.0001,
+      "loss": 1.0443,
+      "step": 10969
+    },
+    {
+      "epoch": 0.9395340870161014,
+      "grad_norm": 0.8007620573043823,
+      "learning_rate": 0.0001,
+      "loss": 1.3381,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9396197327852004,
+      "grad_norm": 0.7654953598976135,
+      "learning_rate": 0.0001,
+      "loss": 0.5616,
+      "step": 10971
+    },
+    {
+      "epoch": 0.9397053785542994,
+      "grad_norm": 0.9027271270751953,
+      "learning_rate": 0.0001,
+      "loss": 0.3911,
+      "step": 10972
+    },
+    {
+      "epoch": 0.9397910243233984,
+      "grad_norm": 0.6043803691864014,
+      "learning_rate": 0.0001,
+      "loss": 1.1788,
+      "step": 10973
+    },
+    {
+      "epoch": 0.9398766700924974,
+      "grad_norm": 0.7212535738945007,
+      "learning_rate": 0.0001,
+      "loss": 0.8439,
+      "step": 10974
+    },
+    {
+      "epoch": 0.9399623158615964,
+      "grad_norm": 1.3938263654708862,
+      "learning_rate": 0.0001,
+      "loss": 1.2629,
+      "step": 10975
+    },
+    {
+      "epoch": 0.9400479616306955,
+      "grad_norm": 0.7886857986450195,
+      "learning_rate": 0.0001,
+      "loss": 0.7112,
+      "step": 10976
+    },
+    {
+      "epoch": 0.9401336073997945,
+      "grad_norm": 0.6132733821868896,
+      "learning_rate": 0.0001,
+      "loss": 0.9665,
+      "step": 10977
+    },
+    {
+      "epoch": 0.9402192531688934,
+      "grad_norm": 0.7208647727966309,
+      "learning_rate": 0.0001,
+      "loss": 0.9444,
+      "step": 10978
+    },
+    {
+      "epoch": 0.9403048989379925,
+      "grad_norm": 1.0239734649658203,
+      "learning_rate": 0.0001,
+      "loss": 1.0662,
+      "step": 10979
+    },
+    {
+      "epoch": 0.9403905447070915,
+      "grad_norm": 1.1696442365646362,
+      "learning_rate": 0.0001,
+      "loss": 1.0495,
+      "step": 10980
+    },
+    {
+      "epoch": 0.9404761904761905,
+      "grad_norm": 0.7799210548400879,
+      "learning_rate": 0.0001,
+      "loss": 0.611,
+      "step": 10981
+    },
+    {
+      "epoch": 0.9405618362452894,
+      "grad_norm": 1.0290623903274536,
+      "learning_rate": 0.0001,
+      "loss": 1.2546,
+      "step": 10982
+    },
+    {
+      "epoch": 0.9406474820143885,
+      "grad_norm": 0.8532921671867371,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 10983
+    },
+    {
+      "epoch": 0.9407331277834875,
+      "grad_norm": 0.621740460395813,
+      "learning_rate": 0.0001,
+      "loss": 0.3493,
+      "step": 10984
+    },
+    {
+      "epoch": 0.9408187735525865,
+      "grad_norm": 1.0160542726516724,
+      "learning_rate": 0.0001,
+      "loss": 0.8447,
+      "step": 10985
+    },
+    {
+      "epoch": 0.9409044193216856,
+      "grad_norm": 0.7192440032958984,
+      "learning_rate": 0.0001,
+      "loss": 0.5424,
+      "step": 10986
+    },
+    {
+      "epoch": 0.9409900650907845,
+      "grad_norm": 1.2101560831069946,
+      "learning_rate": 0.0001,
+      "loss": 1.3026,
+      "step": 10987
+    },
+    {
+      "epoch": 0.9410757108598835,
+      "grad_norm": 0.763606607913971,
+      "learning_rate": 0.0001,
+      "loss": 0.9708,
+      "step": 10988
+    },
+    {
+      "epoch": 0.9411613566289825,
+      "grad_norm": 0.7860888242721558,
+      "learning_rate": 0.0001,
+      "loss": 0.7303,
+      "step": 10989
+    },
+    {
+      "epoch": 0.9412470023980816,
+      "grad_norm": 0.7266968488693237,
+      "learning_rate": 0.0001,
+      "loss": 0.7707,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9413326481671805,
+      "grad_norm": 1.1540780067443848,
+      "learning_rate": 0.0001,
+      "loss": 0.9249,
+      "step": 10991
+    },
+    {
+      "epoch": 0.9414182939362795,
+      "grad_norm": 1.0457814931869507,
+      "learning_rate": 0.0001,
+      "loss": 1.0688,
+      "step": 10992
+    },
+    {
+      "epoch": 0.9415039397053786,
+      "grad_norm": 1.201002836227417,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 10993
+    },
+    {
+      "epoch": 0.9415895854744776,
+      "grad_norm": 1.635982632637024,
+      "learning_rate": 0.0001,
+      "loss": 0.8033,
+      "step": 10994
+    },
+    {
+      "epoch": 0.9416752312435765,
+      "grad_norm": 1.0230764150619507,
+      "learning_rate": 0.0001,
+      "loss": 0.792,
+      "step": 10995
+    },
+    {
+      "epoch": 0.9417608770126755,
+      "grad_norm": 0.5079510807991028,
+      "learning_rate": 0.0001,
+      "loss": 0.2487,
+      "step": 10996
+    },
+    {
+      "epoch": 0.9418465227817746,
+      "grad_norm": 0.7731167674064636,
+      "learning_rate": 0.0001,
+      "loss": 0.7909,
+      "step": 10997
+    },
+    {
+      "epoch": 0.9419321685508736,
+      "grad_norm": 0.6042595505714417,
+      "learning_rate": 0.0001,
+      "loss": 0.7904,
+      "step": 10998
+    },
+    {
+      "epoch": 0.9420178143199726,
+      "grad_norm": 0.6691380739212036,
+      "learning_rate": 0.0001,
+      "loss": 1.0721,
+      "step": 10999
+    },
+    {
+      "epoch": 0.9421034600890716,
+      "grad_norm": 1.2166317701339722,
+      "learning_rate": 0.0001,
+      "loss": 1.4362,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9421891058581706,
+      "grad_norm": 1.076927900314331,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 11001
+    },
+    {
+      "epoch": 0.9422747516272696,
+      "grad_norm": 0.5799708366394043,
+      "learning_rate": 0.0001,
+      "loss": 0.9708,
+      "step": 11002
+    },
+    {
+      "epoch": 0.9423603973963686,
+      "grad_norm": 0.6198391318321228,
+      "learning_rate": 0.0001,
+      "loss": 0.9781,
+      "step": 11003
+    },
+    {
+      "epoch": 0.9424460431654677,
+      "grad_norm": 1.5378128290176392,
+      "learning_rate": 0.0001,
+      "loss": 1.335,
+      "step": 11004
+    },
+    {
+      "epoch": 0.9425316889345666,
+      "grad_norm": 0.9952632784843445,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 11005
+    },
+    {
+      "epoch": 0.9426173347036656,
+      "grad_norm": 0.7873629927635193,
+      "learning_rate": 0.0001,
+      "loss": 1.2507,
+      "step": 11006
+    },
+    {
+      "epoch": 0.9427029804727647,
+      "grad_norm": 0.45777347683906555,
+      "learning_rate": 0.0001,
+      "loss": 0.8722,
+      "step": 11007
+    },
+    {
+      "epoch": 0.9427886262418637,
+      "grad_norm": 0.6989631652832031,
+      "learning_rate": 0.0001,
+      "loss": 0.6079,
+      "step": 11008
+    },
+    {
+      "epoch": 0.9428742720109626,
+      "grad_norm": 0.7089950442314148,
+      "learning_rate": 0.0001,
+      "loss": 0.8922,
+      "step": 11009
+    },
+    {
+      "epoch": 0.9429599177800617,
+      "grad_norm": 0.5013858079910278,
+      "learning_rate": 0.0001,
+      "loss": 0.4316,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9430455635491607,
+      "grad_norm": 2.1196415424346924,
+      "learning_rate": 0.0001,
+      "loss": 0.95,
+      "step": 11011
+    },
+    {
+      "epoch": 0.9431312093182597,
+      "grad_norm": 0.9746466875076294,
+      "learning_rate": 0.0001,
+      "loss": 0.6886,
+      "step": 11012
+    },
+    {
+      "epoch": 0.9432168550873586,
+      "grad_norm": 1.3682781457901,
+      "learning_rate": 0.0001,
+      "loss": 1.1107,
+      "step": 11013
+    },
+    {
+      "epoch": 0.9433025008564577,
+      "grad_norm": 0.7561332583427429,
+      "learning_rate": 0.0001,
+      "loss": 1.1945,
+      "step": 11014
+    },
+    {
+      "epoch": 0.9433881466255567,
+      "grad_norm": 0.8900492787361145,
+      "learning_rate": 0.0001,
+      "loss": 0.8237,
+      "step": 11015
+    },
+    {
+      "epoch": 0.9434737923946557,
+      "grad_norm": 0.7589621543884277,
+      "learning_rate": 0.0001,
+      "loss": 0.6968,
+      "step": 11016
+    },
+    {
+      "epoch": 0.9435594381637548,
+      "grad_norm": 0.5158383846282959,
+      "learning_rate": 0.0001,
+      "loss": 0.9385,
+      "step": 11017
+    },
+    {
+      "epoch": 0.9436450839328537,
+      "grad_norm": 1.030346155166626,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 11018
+    },
+    {
+      "epoch": 0.9437307297019527,
+      "grad_norm": 0.7423628568649292,
+      "learning_rate": 0.0001,
+      "loss": 0.8638,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9438163754710517,
+      "grad_norm": 0.6056767106056213,
+      "learning_rate": 0.0001,
+      "loss": 0.9305,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9439020212401508,
+      "grad_norm": 0.7691469788551331,
+      "learning_rate": 0.0001,
+      "loss": 1.2571,
+      "step": 11021
+    },
+    {
+      "epoch": 0.9439876670092497,
+      "grad_norm": 1.175462245941162,
+      "learning_rate": 0.0001,
+      "loss": 1.2137,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9440733127783487,
+      "grad_norm": 1.040862798690796,
+      "learning_rate": 0.0001,
+      "loss": 0.8929,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9441589585474478,
+      "grad_norm": 0.742217481136322,
+      "learning_rate": 0.0001,
+      "loss": 0.6061,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9442446043165468,
+      "grad_norm": 0.9472373127937317,
+      "learning_rate": 0.0001,
+      "loss": 0.6451,
+      "step": 11025
+    },
+    {
+      "epoch": 0.9443302500856458,
+      "grad_norm": 1.3579673767089844,
+      "learning_rate": 0.0001,
+      "loss": 1.2515,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9444158958547447,
+      "grad_norm": 0.8623193502426147,
+      "learning_rate": 0.0001,
+      "loss": 0.9474,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9445015416238438,
+      "grad_norm": 0.8094980716705322,
+      "learning_rate": 0.0001,
+      "loss": 0.6051,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9445871873929428,
+      "grad_norm": 0.7629110813140869,
+      "learning_rate": 0.0001,
+      "loss": 1.0256,
+      "step": 11029
+    },
+    {
+      "epoch": 0.9446728331620418,
+      "grad_norm": 0.408272922039032,
+      "learning_rate": 0.0001,
+      "loss": 0.566,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9447584789311408,
+      "grad_norm": 0.8293506503105164,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9448441247002398,
+      "grad_norm": 1.3587278127670288,
+      "learning_rate": 0.0001,
+      "loss": 1.7739,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9449297704693388,
+      "grad_norm": 0.4132346510887146,
+      "learning_rate": 0.0001,
+      "loss": 0.4118,
+      "step": 11033
+    },
+    {
+      "epoch": 0.9450154162384378,
+      "grad_norm": 0.7144877314567566,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9451010620075369,
+      "grad_norm": 0.602925717830658,
+      "learning_rate": 0.0001,
+      "loss": 1.0276,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9451867077766358,
+      "grad_norm": 0.6488293409347534,
+      "learning_rate": 0.0001,
+      "loss": 0.5427,
+      "step": 11036
+    },
+    {
+      "epoch": 0.9452723535457348,
+      "grad_norm": 0.7957733869552612,
+      "learning_rate": 0.0001,
+      "loss": 1.028,
+      "step": 11037
+    },
+    {
+      "epoch": 0.9453579993148339,
+      "grad_norm": 0.48985403776168823,
+      "learning_rate": 0.0001,
+      "loss": 0.4388,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9454436450839329,
+      "grad_norm": 0.7614470720291138,
+      "learning_rate": 0.0001,
+      "loss": 0.998,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9455292908530318,
+      "grad_norm": 0.9828156232833862,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9456149366221309,
+      "grad_norm": 1.0144087076187134,
+      "learning_rate": 0.0001,
+      "loss": 0.7619,
+      "step": 11041
+    },
+    {
+      "epoch": 0.9457005823912299,
+      "grad_norm": 1.0463309288024902,
+      "learning_rate": 0.0001,
+      "loss": 0.5627,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9457862281603289,
+      "grad_norm": 1.132972240447998,
+      "learning_rate": 0.0001,
+      "loss": 1.0274,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9458718739294278,
+      "grad_norm": 0.8779407739639282,
+      "learning_rate": 0.0001,
+      "loss": 1.0137,
+      "step": 11044
+    },
+    {
+      "epoch": 0.9459575196985269,
+      "grad_norm": 0.7095499634742737,
+      "learning_rate": 0.0001,
+      "loss": 0.425,
+      "step": 11045
+    },
+    {
+      "epoch": 0.9460431654676259,
+      "grad_norm": 1.0567479133605957,
+      "learning_rate": 0.0001,
+      "loss": 0.625,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9461288112367249,
+      "grad_norm": 0.7905806303024292,
+      "learning_rate": 0.0001,
+      "loss": 0.8535,
+      "step": 11047
+    },
+    {
+      "epoch": 0.946214457005824,
+      "grad_norm": 0.924558162689209,
+      "learning_rate": 0.0001,
+      "loss": 1.8063,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9463001027749229,
+      "grad_norm": 1.0410491228103638,
+      "learning_rate": 0.0001,
+      "loss": 1.2013,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9463857485440219,
+      "grad_norm": 0.8967708945274353,
+      "learning_rate": 0.0001,
+      "loss": 1.2584,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9464713943131209,
+      "grad_norm": 0.9271420240402222,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 11051
+    },
+    {
+      "epoch": 0.94655704008222,
+      "grad_norm": 0.8425499200820923,
+      "learning_rate": 0.0001,
+      "loss": 0.7004,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9466426858513189,
+      "grad_norm": 1.2659543752670288,
+      "learning_rate": 0.0001,
+      "loss": 0.9232,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9467283316204179,
+      "grad_norm": 0.7034193277359009,
+      "learning_rate": 0.0001,
+      "loss": 0.8743,
+      "step": 11054
+    },
+    {
+      "epoch": 0.946813977389517,
+      "grad_norm": 1.4830031394958496,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 11055
+    },
+    {
+      "epoch": 0.946899623158616,
+      "grad_norm": 0.6773744225502014,
+      "learning_rate": 0.0001,
+      "loss": 0.8345,
+      "step": 11056
+    },
+    {
+      "epoch": 0.946985268927715,
+      "grad_norm": 0.8503670692443848,
+      "learning_rate": 0.0001,
+      "loss": 0.4874,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9470709146968139,
+      "grad_norm": 0.7710981965065002,
+      "learning_rate": 0.0001,
+      "loss": 0.6398,
+      "step": 11058
+    },
+    {
+      "epoch": 0.947156560465913,
+      "grad_norm": 0.9721851944923401,
+      "learning_rate": 0.0001,
+      "loss": 1.2089,
+      "step": 11059
+    },
+    {
+      "epoch": 0.947242206235012,
+      "grad_norm": 0.46215251088142395,
+      "learning_rate": 0.0001,
+      "loss": 0.4558,
+      "step": 11060
+    },
+    {
+      "epoch": 0.947327852004111,
+      "grad_norm": 0.6786835193634033,
+      "learning_rate": 0.0001,
+      "loss": 0.7254,
+      "step": 11061
+    },
+    {
+      "epoch": 0.94741349777321,
+      "grad_norm": 0.6992685794830322,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 11062
+    },
+    {
+      "epoch": 0.947499143542309,
+      "grad_norm": 1.3788594007492065,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 11063
+    },
+    {
+      "epoch": 0.947584789311408,
+      "grad_norm": 0.8464872241020203,
+      "learning_rate": 0.0001,
+      "loss": 0.8413,
+      "step": 11064
+    },
+    {
+      "epoch": 0.947670435080507,
+      "grad_norm": 0.6354055404663086,
+      "learning_rate": 0.0001,
+      "loss": 0.98,
+      "step": 11065
+    },
+    {
+      "epoch": 0.947756080849606,
+      "grad_norm": 0.9687389731407166,
+      "learning_rate": 0.0001,
+      "loss": 0.9014,
+      "step": 11066
+    },
+    {
+      "epoch": 0.947841726618705,
+      "grad_norm": 0.6371427774429321,
+      "learning_rate": 0.0001,
+      "loss": 0.5642,
+      "step": 11067
+    },
+    {
+      "epoch": 0.947927372387804,
+      "grad_norm": 1.1613246202468872,
+      "learning_rate": 0.0001,
+      "loss": 0.8479,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9480130181569031,
+      "grad_norm": 1.1414809226989746,
+      "learning_rate": 0.0001,
+      "loss": 1.0927,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9480986639260021,
+      "grad_norm": 1.069327712059021,
+      "learning_rate": 0.0001,
+      "loss": 1.3885,
+      "step": 11070
+    },
+    {
+      "epoch": 0.948184309695101,
+      "grad_norm": 0.7736017107963562,
+      "learning_rate": 0.0001,
+      "loss": 0.5224,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9482699554642001,
+      "grad_norm": 0.49713701009750366,
+      "learning_rate": 0.0001,
+      "loss": 0.5555,
+      "step": 11072
+    },
+    {
+      "epoch": 0.9483556012332991,
+      "grad_norm": 0.9540987014770508,
+      "learning_rate": 0.0001,
+      "loss": 0.9793,
+      "step": 11073
+    },
+    {
+      "epoch": 0.9484412470023981,
+      "grad_norm": 0.8871132135391235,
+      "learning_rate": 0.0001,
+      "loss": 1.3987,
+      "step": 11074
+    },
+    {
+      "epoch": 0.948526892771497,
+      "grad_norm": 0.9943392276763916,
+      "learning_rate": 0.0001,
+      "loss": 0.9767,
+      "step": 11075
+    },
+    {
+      "epoch": 0.9486125385405961,
+      "grad_norm": 0.7793281674385071,
+      "learning_rate": 0.0001,
+      "loss": 1.2171,
+      "step": 11076
+    },
+    {
+      "epoch": 0.9486981843096951,
+      "grad_norm": 0.8740864992141724,
+      "learning_rate": 0.0001,
+      "loss": 0.7466,
+      "step": 11077
+    },
+    {
+      "epoch": 0.9487838300787941,
+      "grad_norm": 1.0806246995925903,
+      "learning_rate": 0.0001,
+      "loss": 0.8071,
+      "step": 11078
+    },
+    {
+      "epoch": 0.9488694758478932,
+      "grad_norm": 0.5895525217056274,
+      "learning_rate": 0.0001,
+      "loss": 0.5946,
+      "step": 11079
+    },
+    {
+      "epoch": 0.9489551216169921,
+      "grad_norm": 0.6336055994033813,
+      "learning_rate": 0.0001,
+      "loss": 0.225,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9490407673860911,
+      "grad_norm": 0.7263014912605286,
+      "learning_rate": 0.0001,
+      "loss": 0.876,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9491264131551901,
+      "grad_norm": 1.198154330253601,
+      "learning_rate": 0.0001,
+      "loss": 1.2322,
+      "step": 11082
+    },
+    {
+      "epoch": 0.9492120589242892,
+      "grad_norm": 1.017783761024475,
+      "learning_rate": 0.0001,
+      "loss": 0.7157,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9492977046933881,
+      "grad_norm": 0.4764805734157562,
+      "learning_rate": 0.0001,
+      "loss": 0.5544,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9493833504624871,
+      "grad_norm": 0.8495124578475952,
+      "learning_rate": 0.0001,
+      "loss": 1.0556,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9494689962315862,
+      "grad_norm": 0.699312150478363,
+      "learning_rate": 0.0001,
+      "loss": 0.8832,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9495546420006852,
+      "grad_norm": 1.329125165939331,
+      "learning_rate": 0.0001,
+      "loss": 0.9689,
+      "step": 11087
+    },
+    {
+      "epoch": 0.9496402877697842,
+      "grad_norm": 0.7823395729064941,
+      "learning_rate": 0.0001,
+      "loss": 1.284,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9497259335388831,
+      "grad_norm": 1.2331492900848389,
+      "learning_rate": 0.0001,
+      "loss": 1.0014,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9498115793079822,
+      "grad_norm": 1.1922624111175537,
+      "learning_rate": 0.0001,
+      "loss": 1.1565,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9498972250770812,
+      "grad_norm": 0.29494306445121765,
+      "learning_rate": 0.0001,
+      "loss": 0.3921,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9499828708461802,
+      "grad_norm": 0.7870509624481201,
+      "learning_rate": 0.0001,
+      "loss": 0.4782,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9500685166152792,
+      "grad_norm": 0.794599175453186,
+      "learning_rate": 0.0001,
+      "loss": 0.9578,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9501541623843782,
+      "grad_norm": 0.7677463293075562,
+      "learning_rate": 0.0001,
+      "loss": 1.0475,
+      "step": 11094
+    },
+    {
+      "epoch": 0.9502398081534772,
+      "grad_norm": 1.044616937637329,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9503254539225763,
+      "grad_norm": 0.9977099299430847,
+      "learning_rate": 0.0001,
+      "loss": 0.8471,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9504110996916753,
+      "grad_norm": 0.8821357488632202,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9504967454607742,
+      "grad_norm": 0.596660852432251,
+      "learning_rate": 0.0001,
+      "loss": 0.6145,
+      "step": 11098
+    },
+    {
+      "epoch": 0.9505823912298732,
+      "grad_norm": 0.7727266550064087,
+      "learning_rate": 0.0001,
+      "loss": 0.6192,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9506680369989723,
+      "grad_norm": 1.314601182937622,
+      "learning_rate": 0.0001,
+      "loss": 1.3356,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9507536827680713,
+      "grad_norm": 0.6982491612434387,
+      "learning_rate": 0.0001,
+      "loss": 1.1502,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9508393285371702,
+      "grad_norm": 1.0001527070999146,
+      "learning_rate": 0.0001,
+      "loss": 0.8481,
+      "step": 11102
+    },
+    {
+      "epoch": 0.9509249743062693,
+      "grad_norm": 1.5708287954330444,
+      "learning_rate": 0.0001,
+      "loss": 1.0646,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9510106200753683,
+      "grad_norm": 0.7147097587585449,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9510962658444673,
+      "grad_norm": 0.9666454792022705,
+      "learning_rate": 0.0001,
+      "loss": 0.9053,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9511819116135662,
+      "grad_norm": 1.0474175214767456,
+      "learning_rate": 0.0001,
+      "loss": 0.7409,
+      "step": 11106
+    },
+    {
+      "epoch": 0.9512675573826653,
+      "grad_norm": 0.9960020780563354,
+      "learning_rate": 0.0001,
+      "loss": 0.9998,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9513532031517643,
+      "grad_norm": 0.4391898512840271,
+      "learning_rate": 0.0001,
+      "loss": 0.4482,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9514388489208633,
+      "grad_norm": 0.893037736415863,
+      "learning_rate": 0.0001,
+      "loss": 0.6798,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9515244946899624,
+      "grad_norm": 1.3611464500427246,
+      "learning_rate": 0.0001,
+      "loss": 1.8313,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9516101404590613,
+      "grad_norm": 0.7626347541809082,
+      "learning_rate": 0.0001,
+      "loss": 0.9016,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9516957862281603,
+      "grad_norm": 0.7515599131584167,
+      "learning_rate": 0.0001,
+      "loss": 0.46,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9517814319972593,
+      "grad_norm": 1.715595006942749,
+      "learning_rate": 0.0001,
+      "loss": 0.9032,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9518670777663584,
+      "grad_norm": 0.894315779209137,
+      "learning_rate": 0.0001,
+      "loss": 1.0467,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9519527235354573,
+      "grad_norm": 1.2879818677902222,
+      "learning_rate": 0.0001,
+      "loss": 1.0662,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9520383693045563,
+      "grad_norm": 1.3613840341567993,
+      "learning_rate": 0.0001,
+      "loss": 1.0811,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9521240150736554,
+      "grad_norm": 0.7844460010528564,
+      "learning_rate": 0.0001,
+      "loss": 0.8288,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9522096608427544,
+      "grad_norm": 1.546049952507019,
+      "learning_rate": 0.0001,
+      "loss": 0.7606,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9522953066118534,
+      "grad_norm": 0.8563426733016968,
+      "learning_rate": 0.0001,
+      "loss": 1.2059,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.120756983757019,
+      "learning_rate": 0.0001,
+      "loss": 1.3496,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9524665981500514,
+      "grad_norm": 0.7223706841468811,
+      "learning_rate": 0.0001,
+      "loss": 1.2049,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9525522439191504,
+      "grad_norm": 1.065109133720398,
+      "learning_rate": 0.0001,
+      "loss": 1.353,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9526378896882494,
+      "grad_norm": 1.1125341653823853,
+      "learning_rate": 0.0001,
+      "loss": 0.9944,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9527235354573484,
+      "grad_norm": 0.8678776025772095,
+      "learning_rate": 0.0001,
+      "loss": 0.947,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9528091812264474,
+      "grad_norm": 0.8020812273025513,
+      "learning_rate": 0.0001,
+      "loss": 0.4911,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9528948269955464,
+      "grad_norm": 1.1600972414016724,
+      "learning_rate": 0.0001,
+      "loss": 1.0282,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9529804727646455,
+      "grad_norm": 0.8623621463775635,
+      "learning_rate": 0.0001,
+      "loss": 0.3814,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9530661185337445,
+      "grad_norm": 0.30939000844955444,
+      "learning_rate": 0.0001,
+      "loss": 0.1792,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9531517643028434,
+      "grad_norm": 0.9492350816726685,
+      "learning_rate": 0.0001,
+      "loss": 0.7261,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9532374100719424,
+      "grad_norm": 1.0121853351593018,
+      "learning_rate": 0.0001,
+      "loss": 1.0297,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9533230558410415,
+      "grad_norm": 1.2518738508224487,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9534087016101405,
+      "grad_norm": 0.8799542784690857,
+      "learning_rate": 0.0001,
+      "loss": 0.7719,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9534943473792394,
+      "grad_norm": 1.147139072418213,
+      "learning_rate": 0.0001,
+      "loss": 1.3482,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9535799931483385,
+      "grad_norm": 1.8832610845565796,
+      "learning_rate": 0.0001,
+      "loss": 1.7141,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9536656389174375,
+      "grad_norm": 0.9861207008361816,
+      "learning_rate": 0.0001,
+      "loss": 0.6835,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9537512846865365,
+      "grad_norm": 1.3291308879852295,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9538369304556354,
+      "grad_norm": 0.9057320356369019,
+      "learning_rate": 0.0001,
+      "loss": 0.6041,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9539225762247345,
+      "grad_norm": 0.928021252155304,
+      "learning_rate": 0.0001,
+      "loss": 1.1418,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9540082219938335,
+      "grad_norm": 1.2820675373077393,
+      "learning_rate": 0.0001,
+      "loss": 0.9321,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9540938677629325,
+      "grad_norm": 0.5042296051979065,
+      "learning_rate": 0.0001,
+      "loss": 0.4882,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9541795135320316,
+      "grad_norm": 0.8458282351493835,
+      "learning_rate": 0.0001,
+      "loss": 1.1949,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9542651593011305,
+      "grad_norm": 0.9758810997009277,
+      "learning_rate": 0.0001,
+      "loss": 0.6129,
+      "step": 11142
+    },
+    {
+      "epoch": 0.9543508050702295,
+      "grad_norm": 4.656057834625244,
+      "learning_rate": 0.0001,
+      "loss": 0.7998,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9544364508393285,
+      "grad_norm": 0.846444308757782,
+      "learning_rate": 0.0001,
+      "loss": 0.2772,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9545220966084276,
+      "grad_norm": 0.6217826008796692,
+      "learning_rate": 0.0001,
+      "loss": 0.565,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9546077423775265,
+      "grad_norm": 0.7161499857902527,
+      "learning_rate": 0.0001,
+      "loss": 0.1361,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9546933881466255,
+      "grad_norm": 0.8380667567253113,
+      "learning_rate": 0.0001,
+      "loss": 1.2597,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9547790339157246,
+      "grad_norm": 1.1342580318450928,
+      "learning_rate": 0.0001,
+      "loss": 1.8869,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9548646796848236,
+      "grad_norm": 1.4118669033050537,
+      "learning_rate": 0.0001,
+      "loss": 1.8842,
+      "step": 11149
+    },
+    {
+      "epoch": 0.9549503254539226,
+      "grad_norm": 0.32505086064338684,
+      "learning_rate": 0.0001,
+      "loss": 0.3095,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9550359712230215,
+      "grad_norm": 1.382283091545105,
+      "learning_rate": 0.0001,
+      "loss": 1.1577,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9551216169921206,
+      "grad_norm": 1.546897530555725,
+      "learning_rate": 0.0001,
+      "loss": 1.9311,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9552072627612196,
+      "grad_norm": 0.7983377575874329,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9552929085303186,
+      "grad_norm": 1.0116060972213745,
+      "learning_rate": 0.0001,
+      "loss": 0.9094,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9553785542994176,
+      "grad_norm": 0.39040684700012207,
+      "learning_rate": 0.0001,
+      "loss": 0.3786,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9554642000685166,
+      "grad_norm": 0.873219907283783,
+      "learning_rate": 0.0001,
+      "loss": 1.1919,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9555498458376156,
+      "grad_norm": 1.0522431135177612,
+      "learning_rate": 0.0001,
+      "loss": 1.3382,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9556354916067147,
+      "grad_norm": 0.8334867358207703,
+      "learning_rate": 0.0001,
+      "loss": 1.4152,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9557211373758137,
+      "grad_norm": 0.873955488204956,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9558067831449126,
+      "grad_norm": 0.34988677501678467,
+      "learning_rate": 0.0001,
+      "loss": 0.3057,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9558924289140116,
+      "grad_norm": 0.631062924861908,
+      "learning_rate": 0.0001,
+      "loss": 1.0376,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9559780746831107,
+      "grad_norm": 0.6893855333328247,
+      "learning_rate": 0.0001,
+      "loss": 0.8753,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9560637204522097,
+      "grad_norm": 1.0509891510009766,
+      "learning_rate": 0.0001,
+      "loss": 0.7334,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9561493662213086,
+      "grad_norm": 0.825799286365509,
+      "learning_rate": 0.0001,
+      "loss": 1.2157,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9562350119904077,
+      "grad_norm": 0.732179582118988,
+      "learning_rate": 0.0001,
+      "loss": 0.6337,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9563206577595067,
+      "grad_norm": 0.6920359134674072,
+      "learning_rate": 0.0001,
+      "loss": 0.8091,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9564063035286057,
+      "grad_norm": 0.8551188707351685,
+      "learning_rate": 0.0001,
+      "loss": 1.1014,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9564919492977046,
+      "grad_norm": 0.5035691261291504,
+      "learning_rate": 0.0001,
+      "loss": 0.7591,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9565775950668037,
+      "grad_norm": 1.4677724838256836,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9566632408359027,
+      "grad_norm": 0.8399138450622559,
+      "learning_rate": 0.0001,
+      "loss": 0.701,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9567488866050017,
+      "grad_norm": 0.8500009775161743,
+      "learning_rate": 0.0001,
+      "loss": 0.6039,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9568345323741008,
+      "grad_norm": 0.8640254139900208,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9569201781431997,
+      "grad_norm": 1.2272790670394897,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9570058239122987,
+      "grad_norm": 0.5700595378875732,
+      "learning_rate": 0.0001,
+      "loss": 0.985,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9570914696813977,
+      "grad_norm": 1.4191890954971313,
+      "learning_rate": 0.0001,
+      "loss": 0.3791,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9571771154504968,
+      "grad_norm": 1.1473301649093628,
+      "learning_rate": 0.0001,
+      "loss": 1.1567,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9572627612195957,
+      "grad_norm": 1.1038917303085327,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9573484069886947,
+      "grad_norm": 0.9424799680709839,
+      "learning_rate": 0.0001,
+      "loss": 0.8745,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9574340527577938,
+      "grad_norm": 0.8862293362617493,
+      "learning_rate": 0.0001,
+      "loss": 1.085,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9575196985268928,
+      "grad_norm": 0.9800492525100708,
+      "learning_rate": 0.0001,
+      "loss": 1.1346,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9576053442959918,
+      "grad_norm": 0.8515070080757141,
+      "learning_rate": 0.0001,
+      "loss": 1.2417,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9576909900650907,
+      "grad_norm": 0.559107780456543,
+      "learning_rate": 0.0001,
+      "loss": 0.6018,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9577766358341898,
+      "grad_norm": 0.9500965476036072,
+      "learning_rate": 0.0001,
+      "loss": 1.2932,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9578622816032888,
+      "grad_norm": 1.0351301431655884,
+      "learning_rate": 0.0001,
+      "loss": 1.2396,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9579479273723878,
+      "grad_norm": 0.954852283000946,
+      "learning_rate": 0.0001,
+      "loss": 1.0507,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9580335731414868,
+      "grad_norm": 1.6133348941802979,
+      "learning_rate": 0.0001,
+      "loss": 1.0505,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9581192189105858,
+      "grad_norm": 0.7120471000671387,
+      "learning_rate": 0.0001,
+      "loss": 1.1711,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9582048646796848,
+      "grad_norm": 0.8858668804168701,
+      "learning_rate": 0.0001,
+      "loss": 0.8757,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9582905104487839,
+      "grad_norm": 1.147140622138977,
+      "learning_rate": 0.0001,
+      "loss": 0.9822,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9583761562178829,
+      "grad_norm": 0.7614222168922424,
+      "learning_rate": 0.0001,
+      "loss": 1.2349,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9584618019869818,
+      "grad_norm": 1.0736331939697266,
+      "learning_rate": 0.0001,
+      "loss": 1.2552,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9585474477560808,
+      "grad_norm": 0.8567391037940979,
+      "learning_rate": 0.0001,
+      "loss": 0.8023,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9586330935251799,
+      "grad_norm": 1.376670002937317,
+      "learning_rate": 0.0001,
+      "loss": 1.0638,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9587187392942789,
+      "grad_norm": 1.3092936277389526,
+      "learning_rate": 0.0001,
+      "loss": 2.147,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9588043850633778,
+      "grad_norm": 1.1340413093566895,
+      "learning_rate": 0.0001,
+      "loss": 1.0947,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9588900308324769,
+      "grad_norm": 0.7944868206977844,
+      "learning_rate": 0.0001,
+      "loss": 0.5698,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9589756766015759,
+      "grad_norm": 0.20536811649799347,
+      "learning_rate": 0.0001,
+      "loss": 0.263,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9590613223706749,
+      "grad_norm": 1.1207002401351929,
+      "learning_rate": 0.0001,
+      "loss": 1.3475,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9591469681397738,
+      "grad_norm": 0.4322524070739746,
+      "learning_rate": 0.0001,
+      "loss": 0.4545,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9592326139088729,
+      "grad_norm": 0.7530121207237244,
+      "learning_rate": 0.0001,
+      "loss": 1.047,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9593182596779719,
+      "grad_norm": 0.8659049272537231,
+      "learning_rate": 0.0001,
+      "loss": 0.6611,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9594039054470709,
+      "grad_norm": 0.8288974165916443,
+      "learning_rate": 0.0001,
+      "loss": 0.5945,
+      "step": 11202
+    },
+    {
+      "epoch": 0.95948955121617,
+      "grad_norm": 0.85310959815979,
+      "learning_rate": 0.0001,
+      "loss": 1.2491,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9595751969852689,
+      "grad_norm": 0.9564068913459778,
+      "learning_rate": 0.0001,
+      "loss": 0.9144,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9596608427543679,
+      "grad_norm": 1.1069273948669434,
+      "learning_rate": 0.0001,
+      "loss": 1.1898,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9597464885234669,
+      "grad_norm": 0.6496805548667908,
+      "learning_rate": 0.0001,
+      "loss": 0.455,
+      "step": 11206
+    },
+    {
+      "epoch": 0.959832134292566,
+      "grad_norm": 0.7321456074714661,
+      "learning_rate": 0.0001,
+      "loss": 0.9182,
+      "step": 11207
+    },
+    {
+      "epoch": 0.959917780061665,
+      "grad_norm": 0.5873109102249146,
+      "learning_rate": 0.0001,
+      "loss": 0.4987,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9600034258307639,
+      "grad_norm": 0.8230220675468445,
+      "learning_rate": 0.0001,
+      "loss": 0.7913,
+      "step": 11209
+    },
+    {
+      "epoch": 0.960089071599863,
+      "grad_norm": 0.6766443252563477,
+      "learning_rate": 0.0001,
+      "loss": 0.9502,
+      "step": 11210
+    },
+    {
+      "epoch": 0.960174717368962,
+      "grad_norm": 0.9253697991371155,
+      "learning_rate": 0.0001,
+      "loss": 0.606,
+      "step": 11211
+    },
+    {
+      "epoch": 0.960260363138061,
+      "grad_norm": 0.7431102395057678,
+      "learning_rate": 0.0001,
+      "loss": 1.1025,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9603460089071599,
+      "grad_norm": 0.8472195863723755,
+      "learning_rate": 0.0001,
+      "loss": 0.9044,
+      "step": 11213
+    },
+    {
+      "epoch": 0.960431654676259,
+      "grad_norm": 0.9976782202720642,
+      "learning_rate": 0.0001,
+      "loss": 1.0296,
+      "step": 11214
+    },
+    {
+      "epoch": 0.960517300445358,
+      "grad_norm": 1.2578622102737427,
+      "learning_rate": 0.0001,
+      "loss": 1.9238,
+      "step": 11215
+    },
+    {
+      "epoch": 0.960602946214457,
+      "grad_norm": 1.1004860401153564,
+      "learning_rate": 0.0001,
+      "loss": 0.9899,
+      "step": 11216
+    },
+    {
+      "epoch": 0.960688591983556,
+      "grad_norm": 1.162024736404419,
+      "learning_rate": 0.0001,
+      "loss": 0.3584,
+      "step": 11217
+    },
+    {
+      "epoch": 0.960774237752655,
+      "grad_norm": 0.544905424118042,
+      "learning_rate": 0.0001,
+      "loss": 0.8044,
+      "step": 11218
+    },
+    {
+      "epoch": 0.960859883521754,
+      "grad_norm": 0.8426231145858765,
+      "learning_rate": 0.0001,
+      "loss": 0.9658,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9609455292908531,
+      "grad_norm": 0.5849726796150208,
+      "learning_rate": 0.0001,
+      "loss": 1.2248,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9610311750599521,
+      "grad_norm": 0.9048805236816406,
+      "learning_rate": 0.0001,
+      "loss": 0.6538,
+      "step": 11221
+    },
+    {
+      "epoch": 0.961116820829051,
+      "grad_norm": 1.2222418785095215,
+      "learning_rate": 0.0001,
+      "loss": 1.3212,
+      "step": 11222
+    },
+    {
+      "epoch": 0.96120246659815,
+      "grad_norm": 0.7028616070747375,
+      "learning_rate": 0.0001,
+      "loss": 0.6265,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9612881123672491,
+      "grad_norm": 1.1788990497589111,
+      "learning_rate": 0.0001,
+      "loss": 1.0974,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9613737581363481,
+      "grad_norm": 0.664419412612915,
+      "learning_rate": 0.0001,
+      "loss": 1.1085,
+      "step": 11225
+    },
+    {
+      "epoch": 0.961459403905447,
+      "grad_norm": 1.5625102519989014,
+      "learning_rate": 0.0001,
+      "loss": 1.3455,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9615450496745461,
+      "grad_norm": 0.9169950485229492,
+      "learning_rate": 0.0001,
+      "loss": 0.9639,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9616306954436451,
+      "grad_norm": 1.0381009578704834,
+      "learning_rate": 0.0001,
+      "loss": 1.0708,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9617163412127441,
+      "grad_norm": 1.0533710718154907,
+      "learning_rate": 0.0001,
+      "loss": 0.6049,
+      "step": 11229
+    },
+    {
+      "epoch": 0.961801986981843,
+      "grad_norm": 1.2092151641845703,
+      "learning_rate": 0.0001,
+      "loss": 1.0103,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9618876327509421,
+      "grad_norm": 0.7721717357635498,
+      "learning_rate": 0.0001,
+      "loss": 0.7827,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9619732785200411,
+      "grad_norm": 0.994151771068573,
+      "learning_rate": 0.0001,
+      "loss": 1.1998,
+      "step": 11232
+    },
+    {
+      "epoch": 0.9620589242891401,
+      "grad_norm": 1.636218547821045,
+      "learning_rate": 0.0001,
+      "loss": 1.2139,
+      "step": 11233
+    },
+    {
+      "epoch": 0.9621445700582392,
+      "grad_norm": 0.7682728171348572,
+      "learning_rate": 0.0001,
+      "loss": 0.6608,
+      "step": 11234
+    },
+    {
+      "epoch": 0.9622302158273381,
+      "grad_norm": 0.7902467250823975,
+      "learning_rate": 0.0001,
+      "loss": 0.7743,
+      "step": 11235
+    },
+    {
+      "epoch": 0.9623158615964371,
+      "grad_norm": 0.6423425674438477,
+      "learning_rate": 0.0001,
+      "loss": 0.9212,
+      "step": 11236
+    },
+    {
+      "epoch": 0.9624015073655361,
+      "grad_norm": 1.2966251373291016,
+      "learning_rate": 0.0001,
+      "loss": 1.2361,
+      "step": 11237
+    },
+    {
+      "epoch": 0.9624871531346352,
+      "grad_norm": 0.997541606426239,
+      "learning_rate": 0.0001,
+      "loss": 1.2948,
+      "step": 11238
+    },
+    {
+      "epoch": 0.9625727989037342,
+      "grad_norm": 1.2889586687088013,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 11239
+    },
+    {
+      "epoch": 0.9626584446728331,
+      "grad_norm": 1.438310980796814,
+      "learning_rate": 0.0001,
+      "loss": 1.3012,
+      "step": 11240
+    },
+    {
+      "epoch": 0.9627440904419322,
+      "grad_norm": 0.995114266872406,
+      "learning_rate": 0.0001,
+      "loss": 0.6968,
+      "step": 11241
+    },
+    {
+      "epoch": 0.9628297362110312,
+      "grad_norm": 1.0249780416488647,
+      "learning_rate": 0.0001,
+      "loss": 1.1724,
+      "step": 11242
+    },
+    {
+      "epoch": 0.9629153819801302,
+      "grad_norm": 1.2364528179168701,
+      "learning_rate": 0.0001,
+      "loss": 0.2284,
+      "step": 11243
+    },
+    {
+      "epoch": 0.9630010277492291,
+      "grad_norm": 0.5764398574829102,
+      "learning_rate": 0.0001,
+      "loss": 1.2298,
+      "step": 11244
+    },
+    {
+      "epoch": 0.9630866735183282,
+      "grad_norm": 0.6529443860054016,
+      "learning_rate": 0.0001,
+      "loss": 0.7537,
+      "step": 11245
+    },
+    {
+      "epoch": 0.9631723192874272,
+      "grad_norm": 1.3382939100265503,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 11246
+    },
+    {
+      "epoch": 0.9632579650565262,
+      "grad_norm": 0.6406915187835693,
+      "learning_rate": 0.0001,
+      "loss": 0.5528,
+      "step": 11247
+    },
+    {
+      "epoch": 0.9633436108256253,
+      "grad_norm": 0.8218042254447937,
+      "learning_rate": 0.0001,
+      "loss": 0.7596,
+      "step": 11248
+    },
+    {
+      "epoch": 0.9634292565947242,
+      "grad_norm": 0.6921981573104858,
+      "learning_rate": 0.0001,
+      "loss": 0.8199,
+      "step": 11249
+    },
+    {
+      "epoch": 0.9635149023638232,
+      "grad_norm": 0.7679973840713501,
+      "learning_rate": 0.0001,
+      "loss": 0.9899,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9636005481329223,
+      "grad_norm": 0.41002681851387024,
+      "learning_rate": 0.0001,
+      "loss": 0.114,
+      "step": 11251
+    },
+    {
+      "epoch": 0.9636861939020213,
+      "grad_norm": 1.2849117517471313,
+      "learning_rate": 0.0001,
+      "loss": 1.3227,
+      "step": 11252
+    },
+    {
+      "epoch": 0.9637718396711202,
+      "grad_norm": 1.18439519405365,
+      "learning_rate": 0.0001,
+      "loss": 1.2796,
+      "step": 11253
+    },
+    {
+      "epoch": 0.9638574854402192,
+      "grad_norm": 0.9063783884048462,
+      "learning_rate": 0.0001,
+      "loss": 0.9051,
+      "step": 11254
+    },
+    {
+      "epoch": 0.9639431312093183,
+      "grad_norm": 0.8898060321807861,
+      "learning_rate": 0.0001,
+      "loss": 0.6426,
+      "step": 11255
+    },
+    {
+      "epoch": 0.9640287769784173,
+      "grad_norm": 0.4361731708049774,
+      "learning_rate": 0.0001,
+      "loss": 0.2809,
+      "step": 11256
+    },
+    {
+      "epoch": 0.9641144227475162,
+      "grad_norm": 0.7665619254112244,
+      "learning_rate": 0.0001,
+      "loss": 1.0924,
+      "step": 11257
+    },
+    {
+      "epoch": 0.9642000685166153,
+      "grad_norm": 0.7526699900627136,
+      "learning_rate": 0.0001,
+      "loss": 1.221,
+      "step": 11258
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 0.8959524035453796,
+      "learning_rate": 0.0001,
+      "loss": 0.9957,
+      "step": 11259
+    },
+    {
+      "epoch": 0.9643713600548133,
+      "grad_norm": 1.3583736419677734,
+      "learning_rate": 0.0001,
+      "loss": 1.4139,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9644570058239123,
+      "grad_norm": 0.9039846062660217,
+      "learning_rate": 0.0001,
+      "loss": 1.2016,
+      "step": 11261
+    },
+    {
+      "epoch": 0.9645426515930113,
+      "grad_norm": 1.00972318649292,
+      "learning_rate": 0.0001,
+      "loss": 0.9335,
+      "step": 11262
+    },
+    {
+      "epoch": 0.9646282973621103,
+      "grad_norm": 1.0299222469329834,
+      "learning_rate": 0.0001,
+      "loss": 0.8894,
+      "step": 11263
+    },
+    {
+      "epoch": 0.9647139431312093,
+      "grad_norm": 0.7791805267333984,
+      "learning_rate": 0.0001,
+      "loss": 0.5528,
+      "step": 11264
+    },
+    {
+      "epoch": 0.9647995889003084,
+      "grad_norm": 1.1882740259170532,
+      "learning_rate": 0.0001,
+      "loss": 1.308,
+      "step": 11265
+    },
+    {
+      "epoch": 0.9648852346694073,
+      "grad_norm": 0.4790668785572052,
+      "learning_rate": 0.0001,
+      "loss": 0.3673,
+      "step": 11266
+    },
+    {
+      "epoch": 0.9649708804385063,
+      "grad_norm": 0.7535160779953003,
+      "learning_rate": 0.0001,
+      "loss": 0.8987,
+      "step": 11267
+    },
+    {
+      "epoch": 0.9650565262076053,
+      "grad_norm": 1.0337495803833008,
+      "learning_rate": 0.0001,
+      "loss": 0.8933,
+      "step": 11268
+    },
+    {
+      "epoch": 0.9651421719767044,
+      "grad_norm": 1.1664800643920898,
+      "learning_rate": 0.0001,
+      "loss": 0.7795,
+      "step": 11269
+    },
+    {
+      "epoch": 0.9652278177458034,
+      "grad_norm": 0.8588460683822632,
+      "learning_rate": 0.0001,
+      "loss": 0.2395,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9653134635149023,
+      "grad_norm": 1.3333572149276733,
+      "learning_rate": 0.0001,
+      "loss": 0.646,
+      "step": 11271
+    },
+    {
+      "epoch": 0.9653991092840014,
+      "grad_norm": 1.1008433103561401,
+      "learning_rate": 0.0001,
+      "loss": 0.9345,
+      "step": 11272
+    },
+    {
+      "epoch": 0.9654847550531004,
+      "grad_norm": 0.5690298080444336,
+      "learning_rate": 0.0001,
+      "loss": 1.0881,
+      "step": 11273
+    },
+    {
+      "epoch": 0.9655704008221994,
+      "grad_norm": 1.405272364616394,
+      "learning_rate": 0.0001,
+      "loss": 1.2344,
+      "step": 11274
+    },
+    {
+      "epoch": 0.9656560465912984,
+      "grad_norm": 1.1117405891418457,
+      "learning_rate": 0.0001,
+      "loss": 1.0377,
+      "step": 11275
+    },
+    {
+      "epoch": 0.9657416923603974,
+      "grad_norm": 1.0451841354370117,
+      "learning_rate": 0.0001,
+      "loss": 1.1793,
+      "step": 11276
+    },
+    {
+      "epoch": 0.9658273381294964,
+      "grad_norm": 0.9931027889251709,
+      "learning_rate": 0.0001,
+      "loss": 0.8199,
+      "step": 11277
+    },
+    {
+      "epoch": 0.9659129838985954,
+      "grad_norm": 0.77096027135849,
+      "learning_rate": 0.0001,
+      "loss": 0.6483,
+      "step": 11278
+    },
+    {
+      "epoch": 0.9659986296676945,
+      "grad_norm": 0.5993596315383911,
+      "learning_rate": 0.0001,
+      "loss": 0.7186,
+      "step": 11279
+    },
+    {
+      "epoch": 0.9660842754367934,
+      "grad_norm": 1.431849479675293,
+      "learning_rate": 0.0001,
+      "loss": 1.4471,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9661699212058924,
+      "grad_norm": 0.9195043444633484,
+      "learning_rate": 0.0001,
+      "loss": 1.0835,
+      "step": 11281
+    },
+    {
+      "epoch": 0.9662555669749915,
+      "grad_norm": 0.9727979302406311,
+      "learning_rate": 0.0001,
+      "loss": 0.3835,
+      "step": 11282
+    },
+    {
+      "epoch": 0.9663412127440905,
+      "grad_norm": 0.7250005602836609,
+      "learning_rate": 0.0001,
+      "loss": 0.5612,
+      "step": 11283
+    },
+    {
+      "epoch": 0.9664268585131894,
+      "grad_norm": 0.9886427521705627,
+      "learning_rate": 0.0001,
+      "loss": 0.7912,
+      "step": 11284
+    },
+    {
+      "epoch": 0.9665125042822884,
+      "grad_norm": 0.6565956473350525,
+      "learning_rate": 0.0001,
+      "loss": 1.2295,
+      "step": 11285
+    },
+    {
+      "epoch": 0.9665981500513875,
+      "grad_norm": 0.6399570107460022,
+      "learning_rate": 0.0001,
+      "loss": 1.1382,
+      "step": 11286
+    },
+    {
+      "epoch": 0.9666837958204865,
+      "grad_norm": 0.648192822933197,
+      "learning_rate": 0.0001,
+      "loss": 0.824,
+      "step": 11287
+    },
+    {
+      "epoch": 0.9667694415895854,
+      "grad_norm": 1.1037497520446777,
+      "learning_rate": 0.0001,
+      "loss": 1.2091,
+      "step": 11288
+    },
+    {
+      "epoch": 0.9668550873586845,
+      "grad_norm": 0.9560112357139587,
+      "learning_rate": 0.0001,
+      "loss": 0.8849,
+      "step": 11289
+    },
+    {
+      "epoch": 0.9669407331277835,
+      "grad_norm": 0.836138129234314,
+      "learning_rate": 0.0001,
+      "loss": 1.2918,
+      "step": 11290
+    },
+    {
+      "epoch": 0.9670263788968825,
+      "grad_norm": 0.7435234785079956,
+      "learning_rate": 0.0001,
+      "loss": 0.6926,
+      "step": 11291
+    },
+    {
+      "epoch": 0.9671120246659815,
+      "grad_norm": 0.8253004550933838,
+      "learning_rate": 0.0001,
+      "loss": 1.171,
+      "step": 11292
+    },
+    {
+      "epoch": 0.9671976704350805,
+      "grad_norm": 0.6813583374023438,
+      "learning_rate": 0.0001,
+      "loss": 0.9724,
+      "step": 11293
+    },
+    {
+      "epoch": 0.9672833162041795,
+      "grad_norm": 0.8385945558547974,
+      "learning_rate": 0.0001,
+      "loss": 1.195,
+      "step": 11294
+    },
+    {
+      "epoch": 0.9673689619732785,
+      "grad_norm": 0.9046676754951477,
+      "learning_rate": 0.0001,
+      "loss": 0.6572,
+      "step": 11295
+    },
+    {
+      "epoch": 0.9674546077423776,
+      "grad_norm": 0.8521518707275391,
+      "learning_rate": 0.0001,
+      "loss": 0.8838,
+      "step": 11296
+    },
+    {
+      "epoch": 0.9675402535114765,
+      "grad_norm": 0.9234656095504761,
+      "learning_rate": 0.0001,
+      "loss": 0.5901,
+      "step": 11297
+    },
+    {
+      "epoch": 0.9676258992805755,
+      "grad_norm": 1.0838366746902466,
+      "learning_rate": 0.0001,
+      "loss": 0.9546,
+      "step": 11298
+    },
+    {
+      "epoch": 0.9677115450496745,
+      "grad_norm": 1.105960726737976,
+      "learning_rate": 0.0001,
+      "loss": 0.8866,
+      "step": 11299
+    },
+    {
+      "epoch": 0.9677971908187736,
+      "grad_norm": 0.3186955749988556,
+      "learning_rate": 0.0001,
+      "loss": 0.099,
+      "step": 11300
+    },
+    {
+      "epoch": 0.9678828365878726,
+      "grad_norm": 0.8273793458938599,
+      "learning_rate": 0.0001,
+      "loss": 1.0175,
+      "step": 11301
+    },
+    {
+      "epoch": 0.9679684823569715,
+      "grad_norm": 0.7776914834976196,
+      "learning_rate": 0.0001,
+      "loss": 0.1958,
+      "step": 11302
+    },
+    {
+      "epoch": 0.9680541281260706,
+      "grad_norm": 0.8203572630882263,
+      "learning_rate": 0.0001,
+      "loss": 0.8593,
+      "step": 11303
+    },
+    {
+      "epoch": 0.9681397738951696,
+      "grad_norm": 0.7968804836273193,
+      "learning_rate": 0.0001,
+      "loss": 0.8825,
+      "step": 11304
+    },
+    {
+      "epoch": 0.9682254196642686,
+      "grad_norm": 1.071002721786499,
+      "learning_rate": 0.0001,
+      "loss": 0.8699,
+      "step": 11305
+    },
+    {
+      "epoch": 0.9683110654333676,
+      "grad_norm": 0.8158132433891296,
+      "learning_rate": 0.0001,
+      "loss": 1.1362,
+      "step": 11306
+    },
+    {
+      "epoch": 0.9683967112024666,
+      "grad_norm": 0.9506875276565552,
+      "learning_rate": 0.0001,
+      "loss": 1.1161,
+      "step": 11307
+    },
+    {
+      "epoch": 0.9684823569715656,
+      "grad_norm": 1.0227159261703491,
+      "learning_rate": 0.0001,
+      "loss": 0.5119,
+      "step": 11308
+    },
+    {
+      "epoch": 0.9685680027406646,
+      "grad_norm": 1.544538974761963,
+      "learning_rate": 0.0001,
+      "loss": 0.6368,
+      "step": 11309
+    },
+    {
+      "epoch": 0.9686536485097637,
+      "grad_norm": 0.67411208152771,
+      "learning_rate": 0.0001,
+      "loss": 0.9291,
+      "step": 11310
+    },
+    {
+      "epoch": 0.9687392942788626,
+      "grad_norm": 0.9913407564163208,
+      "learning_rate": 0.0001,
+      "loss": 0.7905,
+      "step": 11311
+    },
+    {
+      "epoch": 0.9688249400479616,
+      "grad_norm": 1.0366617441177368,
+      "learning_rate": 0.0001,
+      "loss": 0.899,
+      "step": 11312
+    },
+    {
+      "epoch": 0.9689105858170607,
+      "grad_norm": 1.2001430988311768,
+      "learning_rate": 0.0001,
+      "loss": 1.3155,
+      "step": 11313
+    },
+    {
+      "epoch": 0.9689962315861597,
+      "grad_norm": 1.513871192932129,
+      "learning_rate": 0.0001,
+      "loss": 0.962,
+      "step": 11314
+    },
+    {
+      "epoch": 0.9690818773552586,
+      "grad_norm": 1.1172661781311035,
+      "learning_rate": 0.0001,
+      "loss": 1.2044,
+      "step": 11315
+    },
+    {
+      "epoch": 0.9691675231243576,
+      "grad_norm": 0.9200412631034851,
+      "learning_rate": 0.0001,
+      "loss": 1.2659,
+      "step": 11316
+    },
+    {
+      "epoch": 0.9692531688934567,
+      "grad_norm": 0.6031836271286011,
+      "learning_rate": 0.0001,
+      "loss": 0.9609,
+      "step": 11317
+    },
+    {
+      "epoch": 0.9693388146625557,
+      "grad_norm": 0.9111235737800598,
+      "learning_rate": 0.0001,
+      "loss": 1.0146,
+      "step": 11318
+    },
+    {
+      "epoch": 0.9694244604316546,
+      "grad_norm": 1.0114736557006836,
+      "learning_rate": 0.0001,
+      "loss": 0.6551,
+      "step": 11319
+    },
+    {
+      "epoch": 0.9695101062007537,
+      "grad_norm": 0.8449671268463135,
+      "learning_rate": 0.0001,
+      "loss": 1.1653,
+      "step": 11320
+    },
+    {
+      "epoch": 0.9695957519698527,
+      "grad_norm": 0.7910891175270081,
+      "learning_rate": 0.0001,
+      "loss": 0.5319,
+      "step": 11321
+    },
+    {
+      "epoch": 0.9696813977389517,
+      "grad_norm": 1.3105055093765259,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 11322
+    },
+    {
+      "epoch": 0.9697670435080507,
+      "grad_norm": 1.1570191383361816,
+      "learning_rate": 0.0001,
+      "loss": 1.2671,
+      "step": 11323
+    },
+    {
+      "epoch": 0.9698526892771497,
+      "grad_norm": 0.9820029735565186,
+      "learning_rate": 0.0001,
+      "loss": 0.8336,
+      "step": 11324
+    },
+    {
+      "epoch": 0.9699383350462487,
+      "grad_norm": 1.2118386030197144,
+      "learning_rate": 0.0001,
+      "loss": 1.3755,
+      "step": 11325
+    },
+    {
+      "epoch": 0.9700239808153477,
+      "grad_norm": 1.093338966369629,
+      "learning_rate": 0.0001,
+      "loss": 1.1298,
+      "step": 11326
+    },
+    {
+      "epoch": 0.9701096265844468,
+      "grad_norm": 1.1602429151535034,
+      "learning_rate": 0.0001,
+      "loss": 0.9177,
+      "step": 11327
+    },
+    {
+      "epoch": 0.9701952723535457,
+      "grad_norm": 0.6296625137329102,
+      "learning_rate": 0.0001,
+      "loss": 0.8545,
+      "step": 11328
+    },
+    {
+      "epoch": 0.9702809181226447,
+      "grad_norm": 0.8618013858795166,
+      "learning_rate": 0.0001,
+      "loss": 0.8457,
+      "step": 11329
+    },
+    {
+      "epoch": 0.9703665638917437,
+      "grad_norm": 0.5179263949394226,
+      "learning_rate": 0.0001,
+      "loss": 0.5323,
+      "step": 11330
+    },
+    {
+      "epoch": 0.9704522096608428,
+      "grad_norm": 0.816574215888977,
+      "learning_rate": 0.0001,
+      "loss": 0.8834,
+      "step": 11331
+    },
+    {
+      "epoch": 0.9705378554299418,
+      "grad_norm": 1.005911946296692,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 11332
+    },
+    {
+      "epoch": 0.9706235011990407,
+      "grad_norm": 2.9012227058410645,
+      "learning_rate": 0.0001,
+      "loss": 1.0913,
+      "step": 11333
+    },
+    {
+      "epoch": 0.9707091469681398,
+      "grad_norm": 1.1340389251708984,
+      "learning_rate": 0.0001,
+      "loss": 1.0429,
+      "step": 11334
+    },
+    {
+      "epoch": 0.9707947927372388,
+      "grad_norm": 0.41354766488075256,
+      "learning_rate": 0.0001,
+      "loss": 0.5028,
+      "step": 11335
+    },
+    {
+      "epoch": 0.9708804385063378,
+      "grad_norm": 1.0744998455047607,
+      "learning_rate": 0.0001,
+      "loss": 0.7292,
+      "step": 11336
+    },
+    {
+      "epoch": 0.9709660842754368,
+      "grad_norm": 1.0304107666015625,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 11337
+    },
+    {
+      "epoch": 0.9710517300445358,
+      "grad_norm": 0.847290575504303,
+      "learning_rate": 0.0001,
+      "loss": 0.4805,
+      "step": 11338
+    },
+    {
+      "epoch": 0.9711373758136348,
+      "grad_norm": 1.7744868993759155,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 11339
+    },
+    {
+      "epoch": 0.9712230215827338,
+      "grad_norm": 0.8226816058158875,
+      "learning_rate": 0.0001,
+      "loss": 1.2357,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9713086673518329,
+      "grad_norm": 2.0648326873779297,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 11341
+    },
+    {
+      "epoch": 0.9713943131209318,
+      "grad_norm": 1.1208957433700562,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 11342
+    },
+    {
+      "epoch": 0.9714799588900308,
+      "grad_norm": 0.9255852103233337,
+      "learning_rate": 0.0001,
+      "loss": 0.9754,
+      "step": 11343
+    },
+    {
+      "epoch": 0.9715656046591299,
+      "grad_norm": 0.679489254951477,
+      "learning_rate": 0.0001,
+      "loss": 1.002,
+      "step": 11344
+    },
+    {
+      "epoch": 0.9716512504282289,
+      "grad_norm": 1.1927310228347778,
+      "learning_rate": 0.0001,
+      "loss": 1.0891,
+      "step": 11345
+    },
+    {
+      "epoch": 0.9717368961973278,
+      "grad_norm": 0.7496615648269653,
+      "learning_rate": 0.0001,
+      "loss": 0.6093,
+      "step": 11346
+    },
+    {
+      "epoch": 0.9718225419664268,
+      "grad_norm": 1.0207024812698364,
+      "learning_rate": 0.0001,
+      "loss": 1.3707,
+      "step": 11347
+    },
+    {
+      "epoch": 0.9719081877355259,
+      "grad_norm": 0.7735506296157837,
+      "learning_rate": 0.0001,
+      "loss": 0.7155,
+      "step": 11348
+    },
+    {
+      "epoch": 0.9719938335046249,
+      "grad_norm": 1.1510285139083862,
+      "learning_rate": 0.0001,
+      "loss": 1.0472,
+      "step": 11349
+    },
+    {
+      "epoch": 0.9720794792737238,
+      "grad_norm": 0.7908464670181274,
+      "learning_rate": 0.0001,
+      "loss": 1.006,
+      "step": 11350
+    },
+    {
+      "epoch": 0.9721651250428229,
+      "grad_norm": 1.0796704292297363,
+      "learning_rate": 0.0001,
+      "loss": 0.8507,
+      "step": 11351
+    },
+    {
+      "epoch": 0.9722507708119219,
+      "grad_norm": 0.7342812418937683,
+      "learning_rate": 0.0001,
+      "loss": 0.8495,
+      "step": 11352
+    },
+    {
+      "epoch": 0.9723364165810209,
+      "grad_norm": 0.9705809950828552,
+      "learning_rate": 0.0001,
+      "loss": 1.0136,
+      "step": 11353
+    },
+    {
+      "epoch": 0.9724220623501199,
+      "grad_norm": 0.9375585317611694,
+      "learning_rate": 0.0001,
+      "loss": 0.3671,
+      "step": 11354
+    },
+    {
+      "epoch": 0.9725077081192189,
+      "grad_norm": 0.6355669498443604,
+      "learning_rate": 0.0001,
+      "loss": 0.4823,
+      "step": 11355
+    },
+    {
+      "epoch": 0.9725933538883179,
+      "grad_norm": 0.895858883857727,
+      "learning_rate": 0.0001,
+      "loss": 0.671,
+      "step": 11356
+    },
+    {
+      "epoch": 0.9726789996574169,
+      "grad_norm": 0.8166865706443787,
+      "learning_rate": 0.0001,
+      "loss": 0.6526,
+      "step": 11357
+    },
+    {
+      "epoch": 0.972764645426516,
+      "grad_norm": 0.7054765820503235,
+      "learning_rate": 0.0001,
+      "loss": 0.9982,
+      "step": 11358
+    },
+    {
+      "epoch": 0.972850291195615,
+      "grad_norm": 0.798946738243103,
+      "learning_rate": 0.0001,
+      "loss": 1.3871,
+      "step": 11359
+    },
+    {
+      "epoch": 0.9729359369647139,
+      "grad_norm": 0.917731523513794,
+      "learning_rate": 0.0001,
+      "loss": 1.1657,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9730215827338129,
+      "grad_norm": 1.262100338935852,
+      "learning_rate": 0.0001,
+      "loss": 1.0022,
+      "step": 11361
+    },
+    {
+      "epoch": 0.973107228502912,
+      "grad_norm": 0.7087904214859009,
+      "learning_rate": 0.0001,
+      "loss": 1.1314,
+      "step": 11362
+    },
+    {
+      "epoch": 0.973192874272011,
+      "grad_norm": 1.0240155458450317,
+      "learning_rate": 0.0001,
+      "loss": 1.1658,
+      "step": 11363
+    },
+    {
+      "epoch": 0.9732785200411099,
+      "grad_norm": 0.8505158424377441,
+      "learning_rate": 0.0001,
+      "loss": 0.1974,
+      "step": 11364
+    },
+    {
+      "epoch": 0.973364165810209,
+      "grad_norm": 1.2973623275756836,
+      "learning_rate": 0.0001,
+      "loss": 0.8226,
+      "step": 11365
+    },
+    {
+      "epoch": 0.973449811579308,
+      "grad_norm": 1.1416640281677246,
+      "learning_rate": 0.0001,
+      "loss": 1.2548,
+      "step": 11366
+    },
+    {
+      "epoch": 0.973535457348407,
+      "grad_norm": 0.5900280475616455,
+      "learning_rate": 0.0001,
+      "loss": 0.7122,
+      "step": 11367
+    },
+    {
+      "epoch": 0.973621103117506,
+      "grad_norm": 0.3850347399711609,
+      "learning_rate": 0.0001,
+      "loss": 0.4681,
+      "step": 11368
+    },
+    {
+      "epoch": 0.973706748886605,
+      "grad_norm": 0.9361905455589294,
+      "learning_rate": 0.0001,
+      "loss": 1.1481,
+      "step": 11369
+    },
+    {
+      "epoch": 0.973792394655704,
+      "grad_norm": 1.0830384492874146,
+      "learning_rate": 0.0001,
+      "loss": 1.3082,
+      "step": 11370
+    },
+    {
+      "epoch": 0.973878040424803,
+      "grad_norm": 0.8644310235977173,
+      "learning_rate": 0.0001,
+      "loss": 1.0757,
+      "step": 11371
+    },
+    {
+      "epoch": 0.9739636861939021,
+      "grad_norm": 1.283169150352478,
+      "learning_rate": 0.0001,
+      "loss": 0.6858,
+      "step": 11372
+    },
+    {
+      "epoch": 0.974049331963001,
+      "grad_norm": 1.007970929145813,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 11373
+    },
+    {
+      "epoch": 0.9741349777321,
+      "grad_norm": 0.6053493022918701,
+      "learning_rate": 0.0001,
+      "loss": 0.6447,
+      "step": 11374
+    },
+    {
+      "epoch": 0.9742206235011991,
+      "grad_norm": 1.0548992156982422,
+      "learning_rate": 0.0001,
+      "loss": 0.7141,
+      "step": 11375
+    },
+    {
+      "epoch": 0.9743062692702981,
+      "grad_norm": 0.8912431597709656,
+      "learning_rate": 0.0001,
+      "loss": 1.0183,
+      "step": 11376
+    },
+    {
+      "epoch": 0.974391915039397,
+      "grad_norm": 0.8271539807319641,
+      "learning_rate": 0.0001,
+      "loss": 0.9267,
+      "step": 11377
+    },
+    {
+      "epoch": 0.974477560808496,
+      "grad_norm": 0.39959418773651123,
+      "learning_rate": 0.0001,
+      "loss": 0.1139,
+      "step": 11378
+    },
+    {
+      "epoch": 0.9745632065775951,
+      "grad_norm": 1.7957217693328857,
+      "learning_rate": 0.0001,
+      "loss": 1.3278,
+      "step": 11379
+    },
+    {
+      "epoch": 0.9746488523466941,
+      "grad_norm": 0.8301962614059448,
+      "learning_rate": 0.0001,
+      "loss": 0.9495,
+      "step": 11380
+    },
+    {
+      "epoch": 0.974734498115793,
+      "grad_norm": 0.7024197578430176,
+      "learning_rate": 0.0001,
+      "loss": 1.3809,
+      "step": 11381
+    },
+    {
+      "epoch": 0.9748201438848921,
+      "grad_norm": 1.2326247692108154,
+      "learning_rate": 0.0001,
+      "loss": 1.0687,
+      "step": 11382
+    },
+    {
+      "epoch": 0.9749057896539911,
+      "grad_norm": 0.6628285050392151,
+      "learning_rate": 0.0001,
+      "loss": 1.1665,
+      "step": 11383
+    },
+    {
+      "epoch": 0.9749914354230901,
+      "grad_norm": 0.7974384427070618,
+      "learning_rate": 0.0001,
+      "loss": 0.6039,
+      "step": 11384
+    },
+    {
+      "epoch": 0.975077081192189,
+      "grad_norm": 0.9788844585418701,
+      "learning_rate": 0.0001,
+      "loss": 0.8803,
+      "step": 11385
+    },
+    {
+      "epoch": 0.9751627269612881,
+      "grad_norm": 0.7552551627159119,
+      "learning_rate": 0.0001,
+      "loss": 1.4027,
+      "step": 11386
+    },
+    {
+      "epoch": 0.9752483727303871,
+      "grad_norm": 0.38807007670402527,
+      "learning_rate": 0.0001,
+      "loss": 0.5586,
+      "step": 11387
+    },
+    {
+      "epoch": 0.9753340184994861,
+      "grad_norm": 1.2159099578857422,
+      "learning_rate": 0.0001,
+      "loss": 1.1544,
+      "step": 11388
+    },
+    {
+      "epoch": 0.9754196642685852,
+      "grad_norm": 1.0205228328704834,
+      "learning_rate": 0.0001,
+      "loss": 1.2556,
+      "step": 11389
+    },
+    {
+      "epoch": 0.9755053100376841,
+      "grad_norm": 0.6346865892410278,
+      "learning_rate": 0.0001,
+      "loss": 0.1611,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9755909558067831,
+      "grad_norm": 0.6430689096450806,
+      "learning_rate": 0.0001,
+      "loss": 1.0557,
+      "step": 11391
+    },
+    {
+      "epoch": 0.9756766015758821,
+      "grad_norm": 1.063765525817871,
+      "learning_rate": 0.0001,
+      "loss": 1.0592,
+      "step": 11392
+    },
+    {
+      "epoch": 0.9757622473449812,
+      "grad_norm": 0.6850374341011047,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 11393
+    },
+    {
+      "epoch": 0.9758478931140802,
+      "grad_norm": 1.7422598600387573,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 11394
+    },
+    {
+      "epoch": 0.9759335388831791,
+      "grad_norm": 1.194718360900879,
+      "learning_rate": 0.0001,
+      "loss": 1.0588,
+      "step": 11395
+    },
+    {
+      "epoch": 0.9760191846522782,
+      "grad_norm": 0.9049730896949768,
+      "learning_rate": 0.0001,
+      "loss": 0.4175,
+      "step": 11396
+    },
+    {
+      "epoch": 0.9761048304213772,
+      "grad_norm": 0.44699281454086304,
+      "learning_rate": 0.0001,
+      "loss": 0.4002,
+      "step": 11397
+    },
+    {
+      "epoch": 0.9761904761904762,
+      "grad_norm": 0.9896309971809387,
+      "learning_rate": 0.0001,
+      "loss": 1.1189,
+      "step": 11398
+    },
+    {
+      "epoch": 0.9762761219595752,
+      "grad_norm": 0.4222181737422943,
+      "learning_rate": 0.0001,
+      "loss": 0.1582,
+      "step": 11399
+    },
+    {
+      "epoch": 0.9763617677286742,
+      "grad_norm": 1.163392186164856,
+      "learning_rate": 0.0001,
+      "loss": 1.0723,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9764474134977732,
+      "grad_norm": 0.5336900949478149,
+      "learning_rate": 0.0001,
+      "loss": 0.7775,
+      "step": 11401
+    },
+    {
+      "epoch": 0.9765330592668722,
+      "grad_norm": 0.8469893932342529,
+      "learning_rate": 0.0001,
+      "loss": 0.5769,
+      "step": 11402
+    },
+    {
+      "epoch": 0.9766187050359713,
+      "grad_norm": 0.441363662481308,
+      "learning_rate": 0.0001,
+      "loss": 0.4572,
+      "step": 11403
+    },
+    {
+      "epoch": 0.9767043508050702,
+      "grad_norm": 1.0251952409744263,
+      "learning_rate": 0.0001,
+      "loss": 1.011,
+      "step": 11404
+    },
+    {
+      "epoch": 0.9767899965741692,
+      "grad_norm": 0.44874995946884155,
+      "learning_rate": 0.0001,
+      "loss": 0.4004,
+      "step": 11405
+    },
+    {
+      "epoch": 0.9768756423432683,
+      "grad_norm": 1.0508224964141846,
+      "learning_rate": 0.0001,
+      "loss": 0.6609,
+      "step": 11406
+    },
+    {
+      "epoch": 0.9769612881123673,
+      "grad_norm": 0.6498736143112183,
+      "learning_rate": 0.0001,
+      "loss": 0.8217,
+      "step": 11407
+    },
+    {
+      "epoch": 0.9770469338814662,
+      "grad_norm": 1.0230956077575684,
+      "learning_rate": 0.0001,
+      "loss": 0.7859,
+      "step": 11408
+    },
+    {
+      "epoch": 0.9771325796505652,
+      "grad_norm": 0.8851493000984192,
+      "learning_rate": 0.0001,
+      "loss": 1.2618,
+      "step": 11409
+    },
+    {
+      "epoch": 0.9772182254196643,
+      "grad_norm": 0.9444818496704102,
+      "learning_rate": 0.0001,
+      "loss": 1.081,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9773038711887633,
+      "grad_norm": 1.0034782886505127,
+      "learning_rate": 0.0001,
+      "loss": 0.9476,
+      "step": 11411
+    },
+    {
+      "epoch": 0.9773895169578622,
+      "grad_norm": 0.8375968337059021,
+      "learning_rate": 0.0001,
+      "loss": 1.1261,
+      "step": 11412
+    },
+    {
+      "epoch": 0.9774751627269613,
+      "grad_norm": 1.0844053030014038,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 11413
+    },
+    {
+      "epoch": 0.9775608084960603,
+      "grad_norm": 0.6264021992683411,
+      "learning_rate": 0.0001,
+      "loss": 1.1512,
+      "step": 11414
+    },
+    {
+      "epoch": 0.9776464542651593,
+      "grad_norm": 0.5738667845726013,
+      "learning_rate": 0.0001,
+      "loss": 0.9122,
+      "step": 11415
+    },
+    {
+      "epoch": 0.9777321000342583,
+      "grad_norm": 0.986858069896698,
+      "learning_rate": 0.0001,
+      "loss": 1.0204,
+      "step": 11416
+    },
+    {
+      "epoch": 0.9778177458033573,
+      "grad_norm": 0.6849302649497986,
+      "learning_rate": 0.0001,
+      "loss": 1.1871,
+      "step": 11417
+    },
+    {
+      "epoch": 0.9779033915724563,
+      "grad_norm": 0.5842999815940857,
+      "learning_rate": 0.0001,
+      "loss": 1.0534,
+      "step": 11418
+    },
+    {
+      "epoch": 0.9779890373415553,
+      "grad_norm": 1.0804508924484253,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 11419
+    },
+    {
+      "epoch": 0.9780746831106544,
+      "grad_norm": 0.5355669856071472,
+      "learning_rate": 0.0001,
+      "loss": 0.51,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9781603288797533,
+      "grad_norm": 1.2749834060668945,
+      "learning_rate": 0.0001,
+      "loss": 0.7633,
+      "step": 11421
+    },
+    {
+      "epoch": 0.9782459746488523,
+      "grad_norm": 1.0621834993362427,
+      "learning_rate": 0.0001,
+      "loss": 1.0431,
+      "step": 11422
+    },
+    {
+      "epoch": 0.9783316204179514,
+      "grad_norm": 1.1347012519836426,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 11423
+    },
+    {
+      "epoch": 0.9784172661870504,
+      "grad_norm": 1.152005672454834,
+      "learning_rate": 0.0001,
+      "loss": 0.9272,
+      "step": 11424
+    },
+    {
+      "epoch": 0.9785029119561494,
+      "grad_norm": 1.1896299123764038,
+      "learning_rate": 0.0001,
+      "loss": 2.1997,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9785885577252483,
+      "grad_norm": 1.0173671245574951,
+      "learning_rate": 0.0001,
+      "loss": 1.0052,
+      "step": 11426
+    },
+    {
+      "epoch": 0.9786742034943474,
+      "grad_norm": 1.0953742265701294,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 11427
+    },
+    {
+      "epoch": 0.9787598492634464,
+      "grad_norm": 0.5999398231506348,
+      "learning_rate": 0.0001,
+      "loss": 0.7692,
+      "step": 11428
+    },
+    {
+      "epoch": 0.9788454950325454,
+      "grad_norm": 0.7210646271705627,
+      "learning_rate": 0.0001,
+      "loss": 0.807,
+      "step": 11429
+    },
+    {
+      "epoch": 0.9789311408016445,
+      "grad_norm": 1.042235016822815,
+      "learning_rate": 0.0001,
+      "loss": 0.6249,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9790167865707434,
+      "grad_norm": 1.4168765544891357,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 11431
+    },
+    {
+      "epoch": 0.9791024323398424,
+      "grad_norm": 0.4791155457496643,
+      "learning_rate": 0.0001,
+      "loss": 0.759,
+      "step": 11432
+    },
+    {
+      "epoch": 0.9791880781089414,
+      "grad_norm": 0.6629249453544617,
+      "learning_rate": 0.0001,
+      "loss": 1.0985,
+      "step": 11433
+    },
+    {
+      "epoch": 0.9792737238780405,
+      "grad_norm": 1.0027848482131958,
+      "learning_rate": 0.0001,
+      "loss": 0.8118,
+      "step": 11434
+    },
+    {
+      "epoch": 0.9793593696471394,
+      "grad_norm": 0.8541488647460938,
+      "learning_rate": 0.0001,
+      "loss": 0.8908,
+      "step": 11435
+    },
+    {
+      "epoch": 0.9794450154162384,
+      "grad_norm": 0.962821364402771,
+      "learning_rate": 0.0001,
+      "loss": 0.9589,
+      "step": 11436
+    },
+    {
+      "epoch": 0.9795306611853375,
+      "grad_norm": 0.8302850723266602,
+      "learning_rate": 0.0001,
+      "loss": 0.8122,
+      "step": 11437
+    },
+    {
+      "epoch": 0.9796163069544365,
+      "grad_norm": 0.7731879949569702,
+      "learning_rate": 0.0001,
+      "loss": 0.9826,
+      "step": 11438
+    },
+    {
+      "epoch": 0.9797019527235354,
+      "grad_norm": 1.1895320415496826,
+      "learning_rate": 0.0001,
+      "loss": 2.0874,
+      "step": 11439
+    },
+    {
+      "epoch": 0.9797875984926344,
+      "grad_norm": 1.2276767492294312,
+      "learning_rate": 0.0001,
+      "loss": 0.8371,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9798732442617335,
+      "grad_norm": 0.9856301546096802,
+      "learning_rate": 0.0001,
+      "loss": 0.8675,
+      "step": 11441
+    },
+    {
+      "epoch": 0.9799588900308325,
+      "grad_norm": 0.6353697180747986,
+      "learning_rate": 0.0001,
+      "loss": 0.7725,
+      "step": 11442
+    },
+    {
+      "epoch": 0.9800445357999314,
+      "grad_norm": 0.8863317966461182,
+      "learning_rate": 0.0001,
+      "loss": 0.7156,
+      "step": 11443
+    },
+    {
+      "epoch": 0.9801301815690305,
+      "grad_norm": 0.7582398653030396,
+      "learning_rate": 0.0001,
+      "loss": 0.7154,
+      "step": 11444
+    },
+    {
+      "epoch": 0.9802158273381295,
+      "grad_norm": 0.7424359321594238,
+      "learning_rate": 0.0001,
+      "loss": 0.4898,
+      "step": 11445
+    },
+    {
+      "epoch": 0.9803014731072285,
+      "grad_norm": 0.8325217366218567,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 11446
+    },
+    {
+      "epoch": 0.9803871188763275,
+      "grad_norm": 0.5759202837944031,
+      "learning_rate": 0.0001,
+      "loss": 0.1567,
+      "step": 11447
+    },
+    {
+      "epoch": 0.9804727646454265,
+      "grad_norm": 0.9150540232658386,
+      "learning_rate": 0.0001,
+      "loss": 0.909,
+      "step": 11448
+    },
+    {
+      "epoch": 0.9805584104145255,
+      "grad_norm": 0.7626501321792603,
+      "learning_rate": 0.0001,
+      "loss": 1.1182,
+      "step": 11449
+    },
+    {
+      "epoch": 0.9806440561836245,
+      "grad_norm": 1.5515707731246948,
+      "learning_rate": 0.0001,
+      "loss": 0.9702,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9807297019527236,
+      "grad_norm": 0.3991280794143677,
+      "learning_rate": 0.0001,
+      "loss": 0.5422,
+      "step": 11451
+    },
+    {
+      "epoch": 0.9808153477218226,
+      "grad_norm": 0.985334575176239,
+      "learning_rate": 0.0001,
+      "loss": 0.8823,
+      "step": 11452
+    },
+    {
+      "epoch": 0.9809009934909215,
+      "grad_norm": 0.7611268758773804,
+      "learning_rate": 0.0001,
+      "loss": 0.8122,
+      "step": 11453
+    },
+    {
+      "epoch": 0.9809866392600206,
+      "grad_norm": 0.8247898817062378,
+      "learning_rate": 0.0001,
+      "loss": 1.267,
+      "step": 11454
+    },
+    {
+      "epoch": 0.9810722850291196,
+      "grad_norm": 0.8464422225952148,
+      "learning_rate": 0.0001,
+      "loss": 1.0341,
+      "step": 11455
+    },
+    {
+      "epoch": 0.9811579307982186,
+      "grad_norm": 1.4770721197128296,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 11456
+    },
+    {
+      "epoch": 0.9812435765673175,
+      "grad_norm": 0.3529617488384247,
+      "learning_rate": 0.0001,
+      "loss": 0.288,
+      "step": 11457
+    },
+    {
+      "epoch": 0.9813292223364166,
+      "grad_norm": 0.9641810655593872,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 11458
+    },
+    {
+      "epoch": 0.9814148681055156,
+      "grad_norm": 1.3149858713150024,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 11459
+    },
+    {
+      "epoch": 0.9815005138746146,
+      "grad_norm": 0.6154556274414062,
+      "learning_rate": 0.0001,
+      "loss": 0.6222,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9815861596437137,
+      "grad_norm": 0.9295340776443481,
+      "learning_rate": 0.0001,
+      "loss": 0.6777,
+      "step": 11461
+    },
+    {
+      "epoch": 0.9816718054128126,
+      "grad_norm": 1.0903316736221313,
+      "learning_rate": 0.0001,
+      "loss": 1.0527,
+      "step": 11462
+    },
+    {
+      "epoch": 0.9817574511819116,
+      "grad_norm": 0.8288399577140808,
+      "learning_rate": 0.0001,
+      "loss": 1.0227,
+      "step": 11463
+    },
+    {
+      "epoch": 0.9818430969510106,
+      "grad_norm": 0.9739673137664795,
+      "learning_rate": 0.0001,
+      "loss": 0.6674,
+      "step": 11464
+    },
+    {
+      "epoch": 0.9819287427201097,
+      "grad_norm": 1.0606138706207275,
+      "learning_rate": 0.0001,
+      "loss": 0.5205,
+      "step": 11465
+    },
+    {
+      "epoch": 0.9820143884892086,
+      "grad_norm": 0.7687507271766663,
+      "learning_rate": 0.0001,
+      "loss": 0.7358,
+      "step": 11466
+    },
+    {
+      "epoch": 0.9821000342583076,
+      "grad_norm": 1.4977388381958008,
+      "learning_rate": 0.0001,
+      "loss": 0.9951,
+      "step": 11467
+    },
+    {
+      "epoch": 0.9821856800274067,
+      "grad_norm": 1.089781403541565,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 11468
+    },
+    {
+      "epoch": 0.9822713257965057,
+      "grad_norm": 1.4566428661346436,
+      "learning_rate": 0.0001,
+      "loss": 0.8544,
+      "step": 11469
+    },
+    {
+      "epoch": 0.9823569715656046,
+      "grad_norm": 1.069410800933838,
+      "learning_rate": 0.0001,
+      "loss": 1.2385,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9824426173347036,
+      "grad_norm": 1.5246508121490479,
+      "learning_rate": 0.0001,
+      "loss": 1.1662,
+      "step": 11471
+    },
+    {
+      "epoch": 0.9825282631038027,
+      "grad_norm": 0.9177106618881226,
+      "learning_rate": 0.0001,
+      "loss": 1.2946,
+      "step": 11472
+    },
+    {
+      "epoch": 0.9826139088729017,
+      "grad_norm": 0.8915868401527405,
+      "learning_rate": 0.0001,
+      "loss": 0.8789,
+      "step": 11473
+    },
+    {
+      "epoch": 0.9826995546420006,
+      "grad_norm": 0.5618125796318054,
+      "learning_rate": 0.0001,
+      "loss": 0.1619,
+      "step": 11474
+    },
+    {
+      "epoch": 0.9827852004110997,
+      "grad_norm": 0.6592290997505188,
+      "learning_rate": 0.0001,
+      "loss": 0.674,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9828708461801987,
+      "grad_norm": 1.0697871446609497,
+      "learning_rate": 0.0001,
+      "loss": 1.1453,
+      "step": 11476
+    },
+    {
+      "epoch": 0.9829564919492977,
+      "grad_norm": 0.8195000290870667,
+      "learning_rate": 0.0001,
+      "loss": 0.4957,
+      "step": 11477
+    },
+    {
+      "epoch": 0.9830421377183967,
+      "grad_norm": 0.6788261532783508,
+      "learning_rate": 0.0001,
+      "loss": 1.2952,
+      "step": 11478
+    },
+    {
+      "epoch": 0.9831277834874957,
+      "grad_norm": 0.7014586329460144,
+      "learning_rate": 0.0001,
+      "loss": 1.0064,
+      "step": 11479
+    },
+    {
+      "epoch": 0.9832134292565947,
+      "grad_norm": 0.6348764896392822,
+      "learning_rate": 0.0001,
+      "loss": 1.3609,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9832990750256937,
+      "grad_norm": 1.2657550573349,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 11481
+    },
+    {
+      "epoch": 0.9833847207947928,
+      "grad_norm": 0.8315340876579285,
+      "learning_rate": 0.0001,
+      "loss": 0.8576,
+      "step": 11482
+    },
+    {
+      "epoch": 0.9834703665638918,
+      "grad_norm": 0.804084837436676,
+      "learning_rate": 0.0001,
+      "loss": 1.1438,
+      "step": 11483
+    },
+    {
+      "epoch": 0.9835560123329907,
+      "grad_norm": 1.1403474807739258,
+      "learning_rate": 0.0001,
+      "loss": 1.1666,
+      "step": 11484
+    },
+    {
+      "epoch": 0.9836416581020898,
+      "grad_norm": 0.6925166249275208,
+      "learning_rate": 0.0001,
+      "loss": 1.0139,
+      "step": 11485
+    },
+    {
+      "epoch": 0.9837273038711888,
+      "grad_norm": 0.5811964273452759,
+      "learning_rate": 0.0001,
+      "loss": 0.6568,
+      "step": 11486
+    },
+    {
+      "epoch": 0.9838129496402878,
+      "grad_norm": 0.9671300649642944,
+      "learning_rate": 0.0001,
+      "loss": 0.9766,
+      "step": 11487
+    },
+    {
+      "epoch": 0.9838985954093867,
+      "grad_norm": 0.7004082202911377,
+      "learning_rate": 0.0001,
+      "loss": 1.0493,
+      "step": 11488
+    },
+    {
+      "epoch": 0.9839842411784858,
+      "grad_norm": 0.7770243287086487,
+      "learning_rate": 0.0001,
+      "loss": 1.1123,
+      "step": 11489
+    },
+    {
+      "epoch": 0.9840698869475848,
+      "grad_norm": 0.771783173084259,
+      "learning_rate": 0.0001,
+      "loss": 1.2582,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9841555327166838,
+      "grad_norm": 0.6623798608779907,
+      "learning_rate": 0.0001,
+      "loss": 0.4609,
+      "step": 11491
+    },
+    {
+      "epoch": 0.9842411784857829,
+      "grad_norm": 0.688903272151947,
+      "learning_rate": 0.0001,
+      "loss": 0.8429,
+      "step": 11492
+    },
+    {
+      "epoch": 0.9843268242548818,
+      "grad_norm": 1.0342237949371338,
+      "learning_rate": 0.0001,
+      "loss": 0.8061,
+      "step": 11493
+    },
+    {
+      "epoch": 0.9844124700239808,
+      "grad_norm": 1.2111701965332031,
+      "learning_rate": 0.0001,
+      "loss": 1.1318,
+      "step": 11494
+    },
+    {
+      "epoch": 0.9844981157930798,
+      "grad_norm": 1.3690029382705688,
+      "learning_rate": 0.0001,
+      "loss": 0.8952,
+      "step": 11495
+    },
+    {
+      "epoch": 0.9845837615621789,
+      "grad_norm": 0.3196917474269867,
+      "learning_rate": 0.0001,
+      "loss": 0.1516,
+      "step": 11496
+    },
+    {
+      "epoch": 0.9846694073312778,
+      "grad_norm": 1.0641603469848633,
+      "learning_rate": 0.0001,
+      "loss": 0.9508,
+      "step": 11497
+    },
+    {
+      "epoch": 0.9847550531003768,
+      "grad_norm": 1.262039303779602,
+      "learning_rate": 0.0001,
+      "loss": 1.303,
+      "step": 11498
+    },
+    {
+      "epoch": 0.9848406988694759,
+      "grad_norm": 1.4591038227081299,
+      "learning_rate": 0.0001,
+      "loss": 1.0978,
+      "step": 11499
+    },
+    {
+      "epoch": 0.9849263446385749,
+      "grad_norm": 0.7569414377212524,
+      "learning_rate": 0.0001,
+      "loss": 0.6256,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9850119904076738,
+      "grad_norm": 0.9910289645195007,
+      "learning_rate": 0.0001,
+      "loss": 0.7525,
+      "step": 11501
+    },
+    {
+      "epoch": 0.9850976361767728,
+      "grad_norm": 0.7714381217956543,
+      "learning_rate": 0.0001,
+      "loss": 1.1473,
+      "step": 11502
+    },
+    {
+      "epoch": 0.9851832819458719,
+      "grad_norm": 0.9516531229019165,
+      "learning_rate": 0.0001,
+      "loss": 1.2348,
+      "step": 11503
+    },
+    {
+      "epoch": 0.9852689277149709,
+      "grad_norm": 1.3062411546707153,
+      "learning_rate": 0.0001,
+      "loss": 2.0348,
+      "step": 11504
+    },
+    {
+      "epoch": 0.9853545734840699,
+      "grad_norm": 0.7335470914840698,
+      "learning_rate": 0.0001,
+      "loss": 0.6139,
+      "step": 11505
+    },
+    {
+      "epoch": 0.9854402192531689,
+      "grad_norm": 1.2921226024627686,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 11506
+    },
+    {
+      "epoch": 0.9855258650222679,
+      "grad_norm": 0.8529534935951233,
+      "learning_rate": 0.0001,
+      "loss": 1.3811,
+      "step": 11507
+    },
+    {
+      "epoch": 0.9856115107913669,
+      "grad_norm": 1.6338824033737183,
+      "learning_rate": 0.0001,
+      "loss": 0.9663,
+      "step": 11508
+    },
+    {
+      "epoch": 0.9856971565604659,
+      "grad_norm": 0.8281322717666626,
+      "learning_rate": 0.0001,
+      "loss": 1.0676,
+      "step": 11509
+    },
+    {
+      "epoch": 0.9857828023295649,
+      "grad_norm": 0.8052711486816406,
+      "learning_rate": 0.0001,
+      "loss": 1.1738,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9858684480986639,
+      "grad_norm": 0.6132403612136841,
+      "learning_rate": 0.0001,
+      "loss": 0.5407,
+      "step": 11511
+    },
+    {
+      "epoch": 0.9859540938677629,
+      "grad_norm": 0.9065271019935608,
+      "learning_rate": 0.0001,
+      "loss": 0.9121,
+      "step": 11512
+    },
+    {
+      "epoch": 0.986039739636862,
+      "grad_norm": 0.6753214001655579,
+      "learning_rate": 0.0001,
+      "loss": 0.8876,
+      "step": 11513
+    },
+    {
+      "epoch": 0.986125385405961,
+      "grad_norm": 0.6940893530845642,
+      "learning_rate": 0.0001,
+      "loss": 0.6328,
+      "step": 11514
+    },
+    {
+      "epoch": 0.9862110311750599,
+      "grad_norm": 1.1148852109909058,
+      "learning_rate": 0.0001,
+      "loss": 0.8442,
+      "step": 11515
+    },
+    {
+      "epoch": 0.986296676944159,
+      "grad_norm": 1.0743604898452759,
+      "learning_rate": 0.0001,
+      "loss": 1.0875,
+      "step": 11516
+    },
+    {
+      "epoch": 0.986382322713258,
+      "grad_norm": 0.8662080764770508,
+      "learning_rate": 0.0001,
+      "loss": 1.0446,
+      "step": 11517
+    },
+    {
+      "epoch": 0.986467968482357,
+      "grad_norm": 1.3458592891693115,
+      "learning_rate": 0.0001,
+      "loss": 1.4397,
+      "step": 11518
+    },
+    {
+      "epoch": 0.9865536142514559,
+      "grad_norm": 0.8271464705467224,
+      "learning_rate": 0.0001,
+      "loss": 0.9555,
+      "step": 11519
+    },
+    {
+      "epoch": 0.986639260020555,
+      "grad_norm": 1.4628316164016724,
+      "learning_rate": 0.0001,
+      "loss": 1.037,
+      "step": 11520
+    },
+    {
+      "epoch": 0.986724905789654,
+      "grad_norm": 1.005035638809204,
+      "learning_rate": 0.0001,
+      "loss": 0.9575,
+      "step": 11521
+    },
+    {
+      "epoch": 0.986810551558753,
+      "grad_norm": 0.49305349588394165,
+      "learning_rate": 0.0001,
+      "loss": 0.411,
+      "step": 11522
+    },
+    {
+      "epoch": 0.986896197327852,
+      "grad_norm": 0.7829978466033936,
+      "learning_rate": 0.0001,
+      "loss": 1.2161,
+      "step": 11523
+    },
+    {
+      "epoch": 0.986981843096951,
+      "grad_norm": 0.8429348468780518,
+      "learning_rate": 0.0001,
+      "loss": 0.8573,
+      "step": 11524
+    },
+    {
+      "epoch": 0.98706748886605,
+      "grad_norm": 0.9965189695358276,
+      "learning_rate": 0.0001,
+      "loss": 0.5958,
+      "step": 11525
+    },
+    {
+      "epoch": 0.987153134635149,
+      "grad_norm": 0.5433588624000549,
+      "learning_rate": 0.0001,
+      "loss": 0.1758,
+      "step": 11526
+    },
+    {
+      "epoch": 0.9872387804042481,
+      "grad_norm": 0.7366058826446533,
+      "learning_rate": 0.0001,
+      "loss": 1.3267,
+      "step": 11527
+    },
+    {
+      "epoch": 0.987324426173347,
+      "grad_norm": 0.6178042888641357,
+      "learning_rate": 0.0001,
+      "loss": 0.9653,
+      "step": 11528
+    },
+    {
+      "epoch": 0.987410071942446,
+      "grad_norm": 1.2472699880599976,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 11529
+    },
+    {
+      "epoch": 0.9874957177115451,
+      "grad_norm": 1.1032323837280273,
+      "learning_rate": 0.0001,
+      "loss": 1.3093,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9875813634806441,
+      "grad_norm": 1.1598109006881714,
+      "learning_rate": 0.0001,
+      "loss": 1.0943,
+      "step": 11531
+    },
+    {
+      "epoch": 0.987667009249743,
+      "grad_norm": 1.0218505859375,
+      "learning_rate": 0.0001,
+      "loss": 1.1035,
+      "step": 11532
+    },
+    {
+      "epoch": 0.987752655018842,
+      "grad_norm": 1.4773674011230469,
+      "learning_rate": 0.0001,
+      "loss": 1.1982,
+      "step": 11533
+    },
+    {
+      "epoch": 0.9878383007879411,
+      "grad_norm": 0.9001504182815552,
+      "learning_rate": 0.0001,
+      "loss": 1.1787,
+      "step": 11534
+    },
+    {
+      "epoch": 0.9879239465570401,
+      "grad_norm": 0.6625956892967224,
+      "learning_rate": 0.0001,
+      "loss": 0.8929,
+      "step": 11535
+    },
+    {
+      "epoch": 0.988009592326139,
+      "grad_norm": 0.9912849068641663,
+      "learning_rate": 0.0001,
+      "loss": 1.1009,
+      "step": 11536
+    },
+    {
+      "epoch": 0.9880952380952381,
+      "grad_norm": 0.6673577427864075,
+      "learning_rate": 0.0001,
+      "loss": 1.0455,
+      "step": 11537
+    },
+    {
+      "epoch": 0.9881808838643371,
+      "grad_norm": 0.9628725051879883,
+      "learning_rate": 0.0001,
+      "loss": 0.7402,
+      "step": 11538
+    },
+    {
+      "epoch": 0.9882665296334361,
+      "grad_norm": 0.8494071364402771,
+      "learning_rate": 0.0001,
+      "loss": 1.0488,
+      "step": 11539
+    },
+    {
+      "epoch": 0.9883521754025351,
+      "grad_norm": 0.4705158770084381,
+      "learning_rate": 0.0001,
+      "loss": 0.2451,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9884378211716341,
+      "grad_norm": 0.5194993019104004,
+      "learning_rate": 0.0001,
+      "loss": 0.4477,
+      "step": 11541
+    },
+    {
+      "epoch": 0.9885234669407331,
+      "grad_norm": 1.1105163097381592,
+      "learning_rate": 0.0001,
+      "loss": 1.2443,
+      "step": 11542
+    },
+    {
+      "epoch": 0.9886091127098321,
+      "grad_norm": 0.7207448482513428,
+      "learning_rate": 0.0001,
+      "loss": 0.5927,
+      "step": 11543
+    },
+    {
+      "epoch": 0.9886947584789312,
+      "grad_norm": 0.9494970440864563,
+      "learning_rate": 0.0001,
+      "loss": 0.9027,
+      "step": 11544
+    },
+    {
+      "epoch": 0.9887804042480302,
+      "grad_norm": 0.7498946785926819,
+      "learning_rate": 0.0001,
+      "loss": 0.8755,
+      "step": 11545
+    },
+    {
+      "epoch": 0.9888660500171291,
+      "grad_norm": 1.3942359685897827,
+      "learning_rate": 0.0001,
+      "loss": 0.9955,
+      "step": 11546
+    },
+    {
+      "epoch": 0.9889516957862282,
+      "grad_norm": 0.3612481653690338,
+      "learning_rate": 0.0001,
+      "loss": 0.2967,
+      "step": 11547
+    },
+    {
+      "epoch": 0.9890373415553272,
+      "grad_norm": 0.9764088988304138,
+      "learning_rate": 0.0001,
+      "loss": 0.2145,
+      "step": 11548
+    },
+    {
+      "epoch": 0.9891229873244262,
+      "grad_norm": 0.6249983310699463,
+      "learning_rate": 0.0001,
+      "loss": 1.2144,
+      "step": 11549
+    },
+    {
+      "epoch": 0.9892086330935251,
+      "grad_norm": 1.135413646697998,
+      "learning_rate": 0.0001,
+      "loss": 0.8259,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9892942788626242,
+      "grad_norm": 0.7693428993225098,
+      "learning_rate": 0.0001,
+      "loss": 1.2771,
+      "step": 11551
+    },
+    {
+      "epoch": 0.9893799246317232,
+      "grad_norm": 1.0747216939926147,
+      "learning_rate": 0.0001,
+      "loss": 0.8686,
+      "step": 11552
+    },
+    {
+      "epoch": 0.9894655704008222,
+      "grad_norm": 1.4242078065872192,
+      "learning_rate": 0.0001,
+      "loss": 2.1913,
+      "step": 11553
+    },
+    {
+      "epoch": 0.9895512161699213,
+      "grad_norm": 0.7429977655410767,
+      "learning_rate": 0.0001,
+      "loss": 0.8845,
+      "step": 11554
+    },
+    {
+      "epoch": 0.9896368619390202,
+      "grad_norm": 1.053063988685608,
+      "learning_rate": 0.0001,
+      "loss": 1.0884,
+      "step": 11555
+    },
+    {
+      "epoch": 0.9897225077081192,
+      "grad_norm": 0.6141541600227356,
+      "learning_rate": 0.0001,
+      "loss": 0.877,
+      "step": 11556
+    },
+    {
+      "epoch": 0.9898081534772182,
+      "grad_norm": 0.7122877240180969,
+      "learning_rate": 0.0001,
+      "loss": 0.8375,
+      "step": 11557
+    },
+    {
+      "epoch": 0.9898937992463173,
+      "grad_norm": 0.9084086418151855,
+      "learning_rate": 0.0001,
+      "loss": 1.1283,
+      "step": 11558
+    },
+    {
+      "epoch": 0.9899794450154162,
+      "grad_norm": 0.9512540102005005,
+      "learning_rate": 0.0001,
+      "loss": 0.9716,
+      "step": 11559
+    },
+    {
+      "epoch": 0.9900650907845152,
+      "grad_norm": 0.8472195863723755,
+      "learning_rate": 0.0001,
+      "loss": 1.0327,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9901507365536143,
+      "grad_norm": 0.6887949109077454,
+      "learning_rate": 0.0001,
+      "loss": 0.3255,
+      "step": 11561
+    },
+    {
+      "epoch": 0.9902363823227133,
+      "grad_norm": 3.341338872909546,
+      "learning_rate": 0.0001,
+      "loss": 0.7568,
+      "step": 11562
+    },
+    {
+      "epoch": 0.9903220280918122,
+      "grad_norm": 0.948204517364502,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 11563
+    },
+    {
+      "epoch": 0.9904076738609112,
+      "grad_norm": 0.8420327305793762,
+      "learning_rate": 0.0001,
+      "loss": 1.0533,
+      "step": 11564
+    },
+    {
+      "epoch": 0.9904933196300103,
+      "grad_norm": 1.0995208024978638,
+      "learning_rate": 0.0001,
+      "loss": 0.7457,
+      "step": 11565
+    },
+    {
+      "epoch": 0.9905789653991093,
+      "grad_norm": 1.018399953842163,
+      "learning_rate": 0.0001,
+      "loss": 1.1438,
+      "step": 11566
+    },
+    {
+      "epoch": 0.9906646111682083,
+      "grad_norm": 1.0156399011611938,
+      "learning_rate": 0.0001,
+      "loss": 0.5161,
+      "step": 11567
+    },
+    {
+      "epoch": 0.9907502569373073,
+      "grad_norm": 1.149336576461792,
+      "learning_rate": 0.0001,
+      "loss": 1.2284,
+      "step": 11568
+    },
+    {
+      "epoch": 0.9908359027064063,
+      "grad_norm": 0.9477387070655823,
+      "learning_rate": 0.0001,
+      "loss": 0.2602,
+      "step": 11569
+    },
+    {
+      "epoch": 0.9909215484755053,
+      "grad_norm": 0.7329121828079224,
+      "learning_rate": 0.0001,
+      "loss": 0.5188,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9910071942446043,
+      "grad_norm": 1.2672537565231323,
+      "learning_rate": 0.0001,
+      "loss": 0.4428,
+      "step": 11571
+    },
+    {
+      "epoch": 0.9910928400137033,
+      "grad_norm": 0.7994662523269653,
+      "learning_rate": 0.0001,
+      "loss": 1.1647,
+      "step": 11572
+    },
+    {
+      "epoch": 0.9911784857828023,
+      "grad_norm": 0.46923062205314636,
+      "learning_rate": 0.0001,
+      "loss": 0.4551,
+      "step": 11573
+    },
+    {
+      "epoch": 0.9912641315519013,
+      "grad_norm": 0.8383017182350159,
+      "learning_rate": 0.0001,
+      "loss": 0.8018,
+      "step": 11574
+    },
+    {
+      "epoch": 0.9913497773210004,
+      "grad_norm": 1.1086522340774536,
+      "learning_rate": 0.0001,
+      "loss": 1.3297,
+      "step": 11575
+    },
+    {
+      "epoch": 0.9914354230900994,
+      "grad_norm": 0.7230445146560669,
+      "learning_rate": 0.0001,
+      "loss": 0.7804,
+      "step": 11576
+    },
+    {
+      "epoch": 0.9915210688591983,
+      "grad_norm": 1.205310583114624,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 11577
+    },
+    {
+      "epoch": 0.9916067146282974,
+      "grad_norm": 1.0107951164245605,
+      "learning_rate": 0.0001,
+      "loss": 0.6614,
+      "step": 11578
+    },
+    {
+      "epoch": 0.9916923603973964,
+      "grad_norm": 0.9482755661010742,
+      "learning_rate": 0.0001,
+      "loss": 1.08,
+      "step": 11579
+    },
+    {
+      "epoch": 0.9917780061664954,
+      "grad_norm": 0.6852154731750488,
+      "learning_rate": 0.0001,
+      "loss": 0.8673,
+      "step": 11580
+    },
+    {
+      "epoch": 0.9918636519355943,
+      "grad_norm": 0.8790484070777893,
+      "learning_rate": 0.0001,
+      "loss": 0.9621,
+      "step": 11581
+    },
+    {
+      "epoch": 0.9919492977046934,
+      "grad_norm": 1.0143035650253296,
+      "learning_rate": 0.0001,
+      "loss": 0.6817,
+      "step": 11582
+    },
+    {
+      "epoch": 0.9920349434737924,
+      "grad_norm": 0.7368716597557068,
+      "learning_rate": 0.0001,
+      "loss": 0.9763,
+      "step": 11583
+    },
+    {
+      "epoch": 0.9921205892428914,
+      "grad_norm": 1.266638159751892,
+      "learning_rate": 0.0001,
+      "loss": 0.6819,
+      "step": 11584
+    },
+    {
+      "epoch": 0.9922062350119905,
+      "grad_norm": 0.7726707458496094,
+      "learning_rate": 0.0001,
+      "loss": 0.8836,
+      "step": 11585
+    },
+    {
+      "epoch": 0.9922918807810894,
+      "grad_norm": 1.1751093864440918,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 11586
+    },
+    {
+      "epoch": 0.9923775265501884,
+      "grad_norm": 0.9653459191322327,
+      "learning_rate": 0.0001,
+      "loss": 1.0737,
+      "step": 11587
+    },
+    {
+      "epoch": 0.9924631723192874,
+      "grad_norm": 1.0259662866592407,
+      "learning_rate": 0.0001,
+      "loss": 1.3292,
+      "step": 11588
+    },
+    {
+      "epoch": 0.9925488180883865,
+      "grad_norm": 1.3720544576644897,
+      "learning_rate": 0.0001,
+      "loss": 0.7526,
+      "step": 11589
+    },
+    {
+      "epoch": 0.9926344638574854,
+      "grad_norm": 0.8030245304107666,
+      "learning_rate": 0.0001,
+      "loss": 1.3148,
+      "step": 11590
+    },
+    {
+      "epoch": 0.9927201096265844,
+      "grad_norm": 0.8892662525177002,
+      "learning_rate": 0.0001,
+      "loss": 1.3477,
+      "step": 11591
+    },
+    {
+      "epoch": 0.9928057553956835,
+      "grad_norm": 0.7448217272758484,
+      "learning_rate": 0.0001,
+      "loss": 1.0489,
+      "step": 11592
+    },
+    {
+      "epoch": 0.9928914011647825,
+      "grad_norm": 0.8326708078384399,
+      "learning_rate": 0.0001,
+      "loss": 0.7897,
+      "step": 11593
+    },
+    {
+      "epoch": 0.9929770469338814,
+      "grad_norm": 0.9860242009162903,
+      "learning_rate": 0.0001,
+      "loss": 1.0487,
+      "step": 11594
+    },
+    {
+      "epoch": 0.9930626927029804,
+      "grad_norm": 0.82135409116745,
+      "learning_rate": 0.0001,
+      "loss": 0.7154,
+      "step": 11595
+    },
+    {
+      "epoch": 0.9931483384720795,
+      "grad_norm": 1.430895209312439,
+      "learning_rate": 0.0001,
+      "loss": 1.0471,
+      "step": 11596
+    },
+    {
+      "epoch": 0.9932339842411785,
+      "grad_norm": 1.1373822689056396,
+      "learning_rate": 0.0001,
+      "loss": 0.8232,
+      "step": 11597
+    },
+    {
+      "epoch": 0.9933196300102775,
+      "grad_norm": 0.8384882211685181,
+      "learning_rate": 0.0001,
+      "loss": 0.9731,
+      "step": 11598
+    },
+    {
+      "epoch": 0.9934052757793765,
+      "grad_norm": 1.24127197265625,
+      "learning_rate": 0.0001,
+      "loss": 1.0899,
+      "step": 11599
+    },
+    {
+      "epoch": 0.9934909215484755,
+      "grad_norm": 1.0067191123962402,
+      "learning_rate": 0.0001,
+      "loss": 0.905,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9935765673175745,
+      "grad_norm": 0.6699666976928711,
+      "learning_rate": 0.0001,
+      "loss": 0.82,
+      "step": 11601
+    },
+    {
+      "epoch": 0.9936622130866736,
+      "grad_norm": 1.6638636589050293,
+      "learning_rate": 0.0001,
+      "loss": 1.0609,
+      "step": 11602
+    },
+    {
+      "epoch": 0.9937478588557725,
+      "grad_norm": 0.7741130590438843,
+      "learning_rate": 0.0001,
+      "loss": 0.9931,
+      "step": 11603
+    },
+    {
+      "epoch": 0.9938335046248715,
+      "grad_norm": 0.8984777927398682,
+      "learning_rate": 0.0001,
+      "loss": 1.0444,
+      "step": 11604
+    },
+    {
+      "epoch": 0.9939191503939705,
+      "grad_norm": 0.6259962916374207,
+      "learning_rate": 0.0001,
+      "loss": 0.9916,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9940047961630696,
+      "grad_norm": 1.5802946090698242,
+      "learning_rate": 0.0001,
+      "loss": 0.6423,
+      "step": 11606
+    },
+    {
+      "epoch": 0.9940904419321686,
+      "grad_norm": 1.5460307598114014,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 11607
+    },
+    {
+      "epoch": 0.9941760877012675,
+      "grad_norm": 1.2580955028533936,
+      "learning_rate": 0.0001,
+      "loss": 1.1993,
+      "step": 11608
+    },
+    {
+      "epoch": 0.9942617334703666,
+      "grad_norm": 1.0943286418914795,
+      "learning_rate": 0.0001,
+      "loss": 0.6665,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9943473792394656,
+      "grad_norm": 1.1472316980361938,
+      "learning_rate": 0.0001,
+      "loss": 1.2835,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9944330250085646,
+      "grad_norm": 0.9514975547790527,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 11611
+    },
+    {
+      "epoch": 0.9945186707776635,
+      "grad_norm": 0.7297724485397339,
+      "learning_rate": 0.0001,
+      "loss": 1.2124,
+      "step": 11612
+    },
+    {
+      "epoch": 0.9946043165467626,
+      "grad_norm": 0.6263510584831238,
+      "learning_rate": 0.0001,
+      "loss": 0.8125,
+      "step": 11613
+    },
+    {
+      "epoch": 0.9946899623158616,
+      "grad_norm": 1.2738659381866455,
+      "learning_rate": 0.0001,
+      "loss": 1.0538,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9947756080849606,
+      "grad_norm": 0.6973748803138733,
+      "learning_rate": 0.0001,
+      "loss": 0.8254,
+      "step": 11615
+    },
+    {
+      "epoch": 0.9948612538540597,
+      "grad_norm": 0.6945320963859558,
+      "learning_rate": 0.0001,
+      "loss": 0.77,
+      "step": 11616
+    },
+    {
+      "epoch": 0.9949468996231586,
+      "grad_norm": 0.5493613481521606,
+      "learning_rate": 0.0001,
+      "loss": 0.4365,
+      "step": 11617
+    },
+    {
+      "epoch": 0.9950325453922576,
+      "grad_norm": 1.0809597969055176,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 11618
+    },
+    {
+      "epoch": 0.9951181911613566,
+      "grad_norm": 0.9490041732788086,
+      "learning_rate": 0.0001,
+      "loss": 0.8531,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9952038369304557,
+      "grad_norm": 1.288097858428955,
+      "learning_rate": 0.0001,
+      "loss": 1.3233,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9952894826995546,
+      "grad_norm": 0.6708362102508545,
+      "learning_rate": 0.0001,
+      "loss": 0.9076,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9953751284686536,
+      "grad_norm": 0.7001467347145081,
+      "learning_rate": 0.0001,
+      "loss": 0.6306,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9954607742377527,
+      "grad_norm": 0.6968885064125061,
+      "learning_rate": 0.0001,
+      "loss": 1.0765,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9955464200068517,
+      "grad_norm": 0.659478485584259,
+      "learning_rate": 0.0001,
+      "loss": 1.2701,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9956320657759506,
+      "grad_norm": 0.8903671503067017,
+      "learning_rate": 0.0001,
+      "loss": 0.9037,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9957177115450496,
+      "grad_norm": 0.8211873769760132,
+      "learning_rate": 0.0001,
+      "loss": 0.821,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9958033573141487,
+      "grad_norm": 1.5835938453674316,
+      "learning_rate": 0.0001,
+      "loss": 0.5694,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9958890030832477,
+      "grad_norm": 0.7098378539085388,
+      "learning_rate": 0.0001,
+      "loss": 0.6647,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9959746488523467,
+      "grad_norm": 1.0053339004516602,
+      "learning_rate": 0.0001,
+      "loss": 0.974,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9960602946214457,
+      "grad_norm": 0.8807279467582703,
+      "learning_rate": 0.0001,
+      "loss": 0.2194,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9961459403905447,
+      "grad_norm": 1.0970162153244019,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9962315861596437,
+      "grad_norm": 0.9056877493858337,
+      "learning_rate": 0.0001,
+      "loss": 1.2103,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9963172319287428,
+      "grad_norm": 1.0168012380599976,
+      "learning_rate": 0.0001,
+      "loss": 1.0762,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9964028776978417,
+      "grad_norm": 0.6358001828193665,
+      "learning_rate": 0.0001,
+      "loss": 1.0313,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9964885234669407,
+      "grad_norm": 0.974793553352356,
+      "learning_rate": 0.0001,
+      "loss": 1.2767,
+      "step": 11635
+    },
+    {
+      "epoch": 0.9965741692360397,
+      "grad_norm": 1.158340573310852,
+      "learning_rate": 0.0001,
+      "loss": 0.9078,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9966598150051388,
+      "grad_norm": 1.0591456890106201,
+      "learning_rate": 0.0001,
+      "loss": 0.9664,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9967454607742378,
+      "grad_norm": 1.1090871095657349,
+      "learning_rate": 0.0001,
+      "loss": 0.9378,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9968311065433367,
+      "grad_norm": 0.46141257882118225,
+      "learning_rate": 0.0001,
+      "loss": 0.3215,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9969167523124358,
+      "grad_norm": 1.0616416931152344,
+      "learning_rate": 0.0001,
+      "loss": 1.0858,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9970023980815348,
+      "grad_norm": 0.5257631540298462,
+      "learning_rate": 0.0001,
+      "loss": 0.8873,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9970880438506338,
+      "grad_norm": 0.658038854598999,
+      "learning_rate": 0.0001,
+      "loss": 1.2631,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9971736896197327,
+      "grad_norm": 1.136543869972229,
+      "learning_rate": 0.0001,
+      "loss": 1.2002,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9972593353888318,
+      "grad_norm": 0.5972200036048889,
+      "learning_rate": 0.0001,
+      "loss": 0.446,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9973449811579308,
+      "grad_norm": 1.5871379375457764,
+      "learning_rate": 0.0001,
+      "loss": 0.757,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9974306269270298,
+      "grad_norm": 0.9062651991844177,
+      "learning_rate": 0.0001,
+      "loss": 0.8933,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9975162726961289,
+      "grad_norm": 1.3734855651855469,
+      "learning_rate": 0.0001,
+      "loss": 1.3482,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9976019184652278,
+      "grad_norm": 0.9668061137199402,
+      "learning_rate": 0.0001,
+      "loss": 1.0167,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9976875642343268,
+      "grad_norm": 0.6121673583984375,
+      "learning_rate": 0.0001,
+      "loss": 0.9433,
+      "step": 11649
+    },
+    {
+      "epoch": 0.9977732100034258,
+      "grad_norm": 1.3181300163269043,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9978588557725249,
+      "grad_norm": 1.750434398651123,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9979445015416238,
+      "grad_norm": 0.7693085074424744,
+      "learning_rate": 0.0001,
+      "loss": 0.8419,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9980301473107228,
+      "grad_norm": 0.8622987866401672,
+      "learning_rate": 0.0001,
+      "loss": 1.0936,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9981157930798219,
+      "grad_norm": 1.058423399925232,
+      "learning_rate": 0.0001,
+      "loss": 1.0595,
+      "step": 11654
+    },
+    {
+      "epoch": 0.9982014388489209,
+      "grad_norm": 1.2306995391845703,
+      "learning_rate": 0.0001,
+      "loss": 1.1465,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9982870846180198,
+      "grad_norm": 0.8465193510055542,
+      "learning_rate": 0.0001,
+      "loss": 1.3136,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9983727303871188,
+      "grad_norm": 1.6453828811645508,
+      "learning_rate": 0.0001,
+      "loss": 1.1953,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9984583761562179,
+      "grad_norm": 0.9480196237564087,
+      "learning_rate": 0.0001,
+      "loss": 0.9068,
+      "step": 11658
+    },
+    {
+      "epoch": 0.9985440219253169,
+      "grad_norm": 0.9643617272377014,
+      "learning_rate": 0.0001,
+      "loss": 0.993,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9986296676944159,
+      "grad_norm": 0.6934698820114136,
+      "learning_rate": 0.0001,
+      "loss": 1.215,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9987153134635149,
+      "grad_norm": 0.676835298538208,
+      "learning_rate": 0.0001,
+      "loss": 0.857,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9988009592326139,
+      "grad_norm": 1.043225646018982,
+      "learning_rate": 0.0001,
+      "loss": 0.7167,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9988866050017129,
+      "grad_norm": 0.19757167994976044,
+      "learning_rate": 0.0001,
+      "loss": 0.1818,
+      "step": 11663
+    },
+    {
+      "epoch": 0.998972250770812,
+      "grad_norm": 0.6293027400970459,
+      "learning_rate": 0.0001,
+      "loss": 0.3092,
+      "step": 11664
+    },
+    {
+      "epoch": 0.999057896539911,
+      "grad_norm": 0.9451879858970642,
+      "learning_rate": 0.0001,
+      "loss": 0.8021,
+      "step": 11665
+    },
+    {
+      "epoch": 0.9991435423090099,
+      "grad_norm": 0.4934714734554291,
+      "learning_rate": 0.0001,
+      "loss": 0.8891,
+      "step": 11666
+    },
+    {
+      "epoch": 0.9992291880781089,
+      "grad_norm": 0.9225949048995972,
+      "learning_rate": 0.0001,
+      "loss": 1.84,
+      "step": 11667
+    },
+    {
+      "epoch": 0.999314833847208,
+      "grad_norm": 0.9819358587265015,
+      "learning_rate": 0.0001,
+      "loss": 1.152,
+      "step": 11668
+    },
+    {
+      "epoch": 0.999400479616307,
+      "grad_norm": 0.8529909253120422,
+      "learning_rate": 0.0001,
+      "loss": 0.8262,
+      "step": 11669
+    },
+    {
+      "epoch": 0.9994861253854059,
+      "grad_norm": 0.46254822611808777,
+      "learning_rate": 0.0001,
+      "loss": 0.185,
+      "step": 11670
+    },
+    {
+      "epoch": 0.999571771154505,
+      "grad_norm": 0.7454558610916138,
+      "learning_rate": 0.0001,
+      "loss": 1.2472,
+      "step": 11671
+    },
+    {
+      "epoch": 0.999657416923604,
+      "grad_norm": 0.7211062908172607,
+      "learning_rate": 0.0001,
+      "loss": 0.8705,
+      "step": 11672
+    },
+    {
+      "epoch": 0.999743062692703,
+      "grad_norm": 1.315718173980713,
+      "learning_rate": 0.0001,
+      "loss": 1.3047,
+      "step": 11673
+    },
+    {
+      "epoch": 0.9998287084618019,
+      "grad_norm": 2.3761284351348877,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 11674
+    },
+    {
+      "epoch": 0.999914354230901,
+      "grad_norm": 0.6382873058319092,
+      "learning_rate": 0.0001,
+      "loss": 0.9659,
+      "step": 11675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8651794195175171,
+      "learning_rate": 0.0001,
+      "loss": 0.7182,
+      "step": 11676
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 35028,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 1.3653834526875648e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}